{ "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": 0.680244505405426, "learning_rate": 2.25e-07, "loss": 1.1603, "step": 10 }, { "grad_norm": 0.6091309189796448, "learning_rate": 4.75e-07, "loss": 1.1754, "step": 20 }, { "grad_norm": 0.5753999352455139, "learning_rate": 7.25e-07, "loss": 1.1549, "step": 30 }, { "grad_norm": 0.6570479869842529, "learning_rate": 9.75e-07, "loss": 1.1592, "step": 40 }, { "grad_norm": 0.5949161648750305, "learning_rate": 1.2250000000000001e-06, "loss": 1.1352, "step": 50 }, { "grad_norm": 0.6184219121932983, "learning_rate": 1.475e-06, "loss": 1.1495, "step": 60 }, { "grad_norm": 0.7696045637130737, "learning_rate": 1.7250000000000002e-06, "loss": 1.0967, "step": 70 }, { "grad_norm": 0.9022743701934814, "learning_rate": 1.975e-06, "loss": 1.0883, "step": 80 }, { "grad_norm": 0.4865836501121521, "learning_rate": 2.225e-06, "loss": 1.074, "step": 90 }, { "grad_norm": 0.4590373635292053, "learning_rate": 2.4750000000000004e-06, "loss": 1.088, "step": 100 }, { "grad_norm": 0.45875152945518494, "learning_rate": 2.725e-06, "loss": 1.0942, "step": 110 }, { "grad_norm": 0.6248760223388672, "learning_rate": 2.975e-06, "loss": 1.0942, "step": 120 }, { "grad_norm": 0.5599473714828491, "learning_rate": 3.225e-06, "loss": 1.0806, "step": 130 }, { "grad_norm": 0.5536745190620422, "learning_rate": 3.4750000000000006e-06, "loss": 1.0789, "step": 140 }, { "grad_norm": 0.5808615684509277, "learning_rate": 3.725e-06, "loss": 1.0734, "step": 150 }, { "grad_norm": 0.7449338436126709, "learning_rate": 3.975e-06, "loss": 1.0966, "step": 160 }, { "grad_norm": 0.8537556529045105, "learning_rate": 4.225e-06, "loss": 1.0788, "step": 170 }, { "grad_norm": 0.7458491921424866, "learning_rate": 4.475e-06, "loss": 1.0851, "step": 180 }, { "grad_norm": 0.8642412424087524, "learning_rate": 4.7250000000000005e-06, "loss": 1.0989, "step": 190 }, { "grad_norm": 0.7092909812927246, "learning_rate": 4.975000000000001e-06, "loss": 1.0521, "step": 200 }, { "grad_norm": 0.8721814155578613, "learning_rate": 5.225e-06, "loss": 1.0684, "step": 210 }, { "grad_norm": 0.8121271133422852, "learning_rate": 5.475e-06, "loss": 1.0731, "step": 220 }, { "grad_norm": 0.8235288858413696, "learning_rate": 5.725e-06, "loss": 1.0595, "step": 230 }, { "grad_norm": 0.8100305795669556, "learning_rate": 5.975e-06, "loss": 1.0636, "step": 240 }, { "grad_norm": 1.0030239820480347, "learning_rate": 6.2250000000000005e-06, "loss": 1.0662, "step": 250 }, { "grad_norm": 0.8188721537590027, "learning_rate": 6.475000000000001e-06, "loss": 1.0643, "step": 260 }, { "grad_norm": 1.0045833587646484, "learning_rate": 6.725000000000001e-06, "loss": 1.0325, "step": 270 }, { "grad_norm": 1.1029564142227173, "learning_rate": 6.975000000000001e-06, "loss": 1.0176, "step": 280 }, { "grad_norm": 1.0219018459320068, "learning_rate": 7.2249999999999994e-06, "loss": 0.9972, "step": 290 }, { "grad_norm": 1.5318355560302734, "learning_rate": 7.4750000000000004e-06, "loss": 0.9614, "step": 300 }, { "grad_norm": 1.4444018602371216, "learning_rate": 7.725e-06, "loss": 0.9317, "step": 310 }, { "grad_norm": 1.338832139968872, "learning_rate": 7.975e-06, "loss": 0.881, "step": 320 }, { "grad_norm": 2.5377678871154785, "learning_rate": 8.225e-06, "loss": 0.8602, "step": 330 }, { "grad_norm": 1.6988798379898071, "learning_rate": 8.475000000000001e-06, "loss": 0.8129, "step": 340 }, { "grad_norm": 1.98197340965271, "learning_rate": 8.725e-06, "loss": 0.7856, "step": 350 }, { "grad_norm": 2.2033467292785645, "learning_rate": 8.975e-06, "loss": 0.7228, "step": 360 }, { "grad_norm": 2.258134603500366, "learning_rate": 9.225e-06, "loss": 0.7087, "step": 370 }, { "grad_norm": 2.0371739864349365, "learning_rate": 9.475e-06, "loss": 0.6652, "step": 380 }, { "grad_norm": 2.1832895278930664, "learning_rate": 9.725000000000001e-06, "loss": 0.6464, "step": 390 }, { "grad_norm": 2.752384901046753, "learning_rate": 9.975e-06, "loss": 0.6279, "step": 400 }, { "grad_norm": 2.1112849712371826, "learning_rate": 1.0225e-05, "loss": 0.5717, "step": 410 }, { "grad_norm": 2.706686019897461, "learning_rate": 1.0475e-05, "loss": 0.529, "step": 420 }, { "grad_norm": 1.978775143623352, "learning_rate": 1.0725e-05, "loss": 0.5247, "step": 430 }, { "grad_norm": 2.175722599029541, "learning_rate": 1.0975e-05, "loss": 0.4793, "step": 440 }, { "grad_norm": 1.9270185232162476, "learning_rate": 1.1225e-05, "loss": 0.4356, "step": 450 }, { "grad_norm": 2.4831485748291016, "learning_rate": 1.1475000000000001e-05, "loss": 0.4164, "step": 460 }, { "grad_norm": 2.688061475753784, "learning_rate": 1.1725e-05, "loss": 0.4052, "step": 470 }, { "grad_norm": 2.2138259410858154, "learning_rate": 1.1975e-05, "loss": 0.3882, "step": 480 }, { "grad_norm": 2.7426412105560303, "learning_rate": 1.2225e-05, "loss": 0.3616, "step": 490 }, { "grad_norm": 3.224731206893921, "learning_rate": 1.2475e-05, "loss": 0.3167, "step": 500 }, { "grad_norm": 2.0572757720947266, "learning_rate": 1.2725000000000001e-05, "loss": 0.3276, "step": 510 }, { "grad_norm": 2.6166484355926514, "learning_rate": 1.2975e-05, "loss": 0.2961, "step": 520 }, { "grad_norm": 3.092761516571045, "learning_rate": 1.3225000000000001e-05, "loss": 0.2792, "step": 530 }, { "grad_norm": 1.6964805126190186, "learning_rate": 1.3475000000000002e-05, "loss": 0.2211, "step": 540 }, { "grad_norm": 2.4852511882781982, "learning_rate": 1.3725000000000002e-05, "loss": 0.26, "step": 550 }, { "grad_norm": 2.632524013519287, "learning_rate": 1.3975000000000003e-05, "loss": 0.2456, "step": 560 }, { "grad_norm": 2.477297782897949, "learning_rate": 1.4225e-05, "loss": 0.2023, "step": 570 }, { "grad_norm": 1.834025502204895, "learning_rate": 1.4475e-05, "loss": 0.2005, "step": 580 }, { "grad_norm": 2.5430939197540283, "learning_rate": 1.4725e-05, "loss": 0.1962, "step": 590 }, { "grad_norm": 2.6601715087890625, "learning_rate": 1.4975e-05, "loss": 0.2113, "step": 600 }, { "grad_norm": 2.2221667766571045, "learning_rate": 1.5225e-05, "loss": 0.1936, "step": 610 }, { "grad_norm": 2.8993051052093506, "learning_rate": 1.5475e-05, "loss": 0.1956, "step": 620 }, { "grad_norm": 2.1114587783813477, "learning_rate": 1.5725e-05, "loss": 0.192, "step": 630 }, { "grad_norm": 2.506571054458618, "learning_rate": 1.5975000000000002e-05, "loss": 0.1872, "step": 640 }, { "grad_norm": 2.442742109298706, "learning_rate": 1.6225e-05, "loss": 0.1675, "step": 650 }, { "grad_norm": 2.7277331352233887, "learning_rate": 1.6475e-05, "loss": 0.1933, "step": 660 }, { "grad_norm": 1.443422555923462, "learning_rate": 1.6725000000000003e-05, "loss": 0.1534, "step": 670 }, { "grad_norm": 2.146476984024048, "learning_rate": 1.6975000000000003e-05, "loss": 0.1742, "step": 680 }, { "grad_norm": 2.0798447132110596, "learning_rate": 1.7225e-05, "loss": 0.1606, "step": 690 }, { "grad_norm": 2.066296100616455, "learning_rate": 1.7475e-05, "loss": 0.1877, "step": 700 }, { "grad_norm": 2.0630435943603516, "learning_rate": 1.7725e-05, "loss": 0.1578, "step": 710 }, { "grad_norm": 2.27388858795166, "learning_rate": 1.7975e-05, "loss": 0.1693, "step": 720 }, { "grad_norm": 1.5914658308029175, "learning_rate": 1.8225e-05, "loss": 0.1627, "step": 730 }, { "grad_norm": 2.0790457725524902, "learning_rate": 1.8475000000000002e-05, "loss": 0.1276, "step": 740 }, { "grad_norm": 1.609139323234558, "learning_rate": 1.8725e-05, "loss": 0.1519, "step": 750 }, { "grad_norm": 1.8330469131469727, "learning_rate": 1.8975e-05, "loss": 0.1503, "step": 760 }, { "grad_norm": 1.6515507698059082, "learning_rate": 1.9225e-05, "loss": 0.1586, "step": 770 }, { "grad_norm": 1.27581787109375, "learning_rate": 1.9475000000000002e-05, "loss": 0.1947, "step": 780 }, { "grad_norm": 1.373388648033142, "learning_rate": 1.9725000000000002e-05, "loss": 0.1367, "step": 790 }, { "grad_norm": 1.685286283493042, "learning_rate": 1.9975e-05, "loss": 0.1409, "step": 800 }, { "grad_norm": 1.698001503944397, "learning_rate": 2.0225000000000004e-05, "loss": 0.1412, "step": 810 }, { "grad_norm": 1.6651697158813477, "learning_rate": 2.0475e-05, "loss": 0.1483, "step": 820 }, { "grad_norm": 2.101857900619507, "learning_rate": 2.0725e-05, "loss": 0.1652, "step": 830 }, { "grad_norm": 1.6212613582611084, "learning_rate": 2.0975e-05, "loss": 0.1537, "step": 840 }, { "grad_norm": 1.4796812534332275, "learning_rate": 2.1225e-05, "loss": 0.1558, "step": 850 }, { "grad_norm": 1.8637739419937134, "learning_rate": 2.1475e-05, "loss": 0.1368, "step": 860 }, { "grad_norm": 1.7156217098236084, "learning_rate": 2.1725e-05, "loss": 0.1571, "step": 870 }, { "grad_norm": 1.6239635944366455, "learning_rate": 2.1975000000000002e-05, "loss": 0.136, "step": 880 }, { "grad_norm": 1.5542595386505127, "learning_rate": 2.2225e-05, "loss": 0.1449, "step": 890 }, { "grad_norm": 1.5209914445877075, "learning_rate": 2.2475e-05, "loss": 0.1403, "step": 900 }, { "grad_norm": 1.4359207153320312, "learning_rate": 2.2725000000000003e-05, "loss": 0.138, "step": 910 }, { "grad_norm": 1.1700928211212158, "learning_rate": 2.2975000000000003e-05, "loss": 0.1284, "step": 920 }, { "grad_norm": 1.619882345199585, "learning_rate": 2.3225000000000002e-05, "loss": 0.1433, "step": 930 }, { "grad_norm": 1.6490432024002075, "learning_rate": 2.3475e-05, "loss": 0.1476, "step": 940 }, { "grad_norm": 1.5331430435180664, "learning_rate": 2.3725e-05, "loss": 0.1239, "step": 950 }, { "grad_norm": 1.2536495923995972, "learning_rate": 2.3975e-05, "loss": 0.1306, "step": 960 }, { "grad_norm": 1.7240979671478271, "learning_rate": 2.4225e-05, "loss": 0.1618, "step": 970 }, { "grad_norm": 1.235575795173645, "learning_rate": 2.4475000000000002e-05, "loss": 0.1241, "step": 980 }, { "grad_norm": 1.6119449138641357, "learning_rate": 2.4725e-05, "loss": 0.1227, "step": 990 }, { "grad_norm": 1.4238557815551758, "learning_rate": 2.4975e-05, "loss": 0.1284, "step": 1000 }, { "grad_norm": 1.0283904075622559, "learning_rate": 2.5225e-05, "loss": 0.1194, "step": 1010 }, { "grad_norm": 1.7176034450531006, "learning_rate": 2.5475e-05, "loss": 0.1415, "step": 1020 }, { "grad_norm": 1.4602965116500854, "learning_rate": 2.5725e-05, "loss": 0.1119, "step": 1030 }, { "grad_norm": 1.3930244445800781, "learning_rate": 2.5974999999999998e-05, "loss": 0.1371, "step": 1040 }, { "grad_norm": 1.2168962955474854, "learning_rate": 2.6225e-05, "loss": 0.1412, "step": 1050 }, { "grad_norm": 1.3418439626693726, "learning_rate": 2.6475e-05, "loss": 0.1523, "step": 1060 }, { "grad_norm": 1.7848503589630127, "learning_rate": 2.6725e-05, "loss": 0.1481, "step": 1070 }, { "grad_norm": 1.7762672901153564, "learning_rate": 2.6975000000000002e-05, "loss": 0.1662, "step": 1080 }, { "grad_norm": 1.559237003326416, "learning_rate": 2.7225e-05, "loss": 0.1422, "step": 1090 }, { "grad_norm": 1.2470636367797852, "learning_rate": 2.7475e-05, "loss": 0.1133, "step": 1100 }, { "grad_norm": 1.8816145658493042, "learning_rate": 2.7725e-05, "loss": 0.1573, "step": 1110 }, { "grad_norm": 1.241547703742981, "learning_rate": 2.7975000000000002e-05, "loss": 0.1083, "step": 1120 }, { "grad_norm": 1.3543808460235596, "learning_rate": 2.8225e-05, "loss": 0.1337, "step": 1130 }, { "grad_norm": 1.464935064315796, "learning_rate": 2.8475e-05, "loss": 0.1672, "step": 1140 }, { "grad_norm": 1.4301731586456299, "learning_rate": 2.8725e-05, "loss": 0.145, "step": 1150 }, { "grad_norm": 0.973242461681366, "learning_rate": 2.8975000000000003e-05, "loss": 0.1364, "step": 1160 }, { "grad_norm": 1.203488826751709, "learning_rate": 2.9225000000000002e-05, "loss": 0.1307, "step": 1170 }, { "grad_norm": 1.0987604856491089, "learning_rate": 2.9475e-05, "loss": 0.1151, "step": 1180 }, { "grad_norm": 1.3806569576263428, "learning_rate": 2.9725000000000004e-05, "loss": 0.1372, "step": 1190 }, { "grad_norm": 2.7105679512023926, "learning_rate": 2.9975000000000004e-05, "loss": 0.1319, "step": 1200 }, { "grad_norm": 2.189544439315796, "learning_rate": 3.0225000000000003e-05, "loss": 0.1392, "step": 1210 }, { "grad_norm": 1.2400239706039429, "learning_rate": 3.0475000000000002e-05, "loss": 0.1196, "step": 1220 }, { "grad_norm": 1.2847272157669067, "learning_rate": 3.0725e-05, "loss": 0.1337, "step": 1230 }, { "grad_norm": 0.8742586970329285, "learning_rate": 3.0975e-05, "loss": 0.1279, "step": 1240 }, { "grad_norm": 1.428936243057251, "learning_rate": 3.122500000000001e-05, "loss": 0.1124, "step": 1250 }, { "grad_norm": 1.5799204111099243, "learning_rate": 3.1475e-05, "loss": 0.1194, "step": 1260 }, { "grad_norm": 0.9705102443695068, "learning_rate": 3.1725e-05, "loss": 0.1069, "step": 1270 }, { "grad_norm": 0.9701875448226929, "learning_rate": 3.1975e-05, "loss": 0.1252, "step": 1280 }, { "grad_norm": 1.0518624782562256, "learning_rate": 3.2225e-05, "loss": 0.1155, "step": 1290 }, { "grad_norm": 0.806573748588562, "learning_rate": 3.2474999999999997e-05, "loss": 0.1107, "step": 1300 }, { "grad_norm": 0.9010583758354187, "learning_rate": 3.2725e-05, "loss": 0.1128, "step": 1310 }, { "grad_norm": 1.1856096982955933, "learning_rate": 3.2975e-05, "loss": 0.1195, "step": 1320 }, { "grad_norm": 0.877661943435669, "learning_rate": 3.3225e-05, "loss": 0.0977, "step": 1330 }, { "grad_norm": 1.0025099515914917, "learning_rate": 3.3475e-05, "loss": 0.0903, "step": 1340 }, { "grad_norm": 1.0183720588684082, "learning_rate": 3.3725e-05, "loss": 0.1293, "step": 1350 }, { "grad_norm": 1.462875485420227, "learning_rate": 3.3975e-05, "loss": 0.1227, "step": 1360 }, { "grad_norm": 1.2076387405395508, "learning_rate": 3.4225e-05, "loss": 0.1027, "step": 1370 }, { "grad_norm": 1.147117018699646, "learning_rate": 3.4475000000000005e-05, "loss": 0.1051, "step": 1380 }, { "grad_norm": 1.2798658609390259, "learning_rate": 3.4725000000000004e-05, "loss": 0.107, "step": 1390 }, { "grad_norm": 1.2464507818222046, "learning_rate": 3.4975e-05, "loss": 0.1054, "step": 1400 }, { "grad_norm": 1.0233255624771118, "learning_rate": 3.5225e-05, "loss": 0.1111, "step": 1410 }, { "grad_norm": 1.2810893058776855, "learning_rate": 3.5475e-05, "loss": 0.1116, "step": 1420 }, { "grad_norm": 1.0458263158798218, "learning_rate": 3.5725e-05, "loss": 0.0904, "step": 1430 }, { "grad_norm": 0.9468550682067871, "learning_rate": 3.5975e-05, "loss": 0.1086, "step": 1440 }, { "grad_norm": 0.9540814161300659, "learning_rate": 3.6225000000000006e-05, "loss": 0.1211, "step": 1450 }, { "grad_norm": 1.2106773853302002, "learning_rate": 3.6475000000000006e-05, "loss": 0.1158, "step": 1460 }, { "grad_norm": 0.9256830811500549, "learning_rate": 3.6725000000000005e-05, "loss": 0.1173, "step": 1470 }, { "grad_norm": 1.1543033123016357, "learning_rate": 3.6975000000000004e-05, "loss": 0.1065, "step": 1480 }, { "grad_norm": 0.9572845101356506, "learning_rate": 3.7225000000000004e-05, "loss": 0.105, "step": 1490 }, { "grad_norm": 0.9019548892974854, "learning_rate": 3.7475e-05, "loss": 0.1146, "step": 1500 }, { "grad_norm": 0.918161928653717, "learning_rate": 3.7725e-05, "loss": 0.0945, "step": 1510 }, { "grad_norm": 1.1651395559310913, "learning_rate": 3.7975e-05, "loss": 0.1097, "step": 1520 }, { "grad_norm": 0.9387496113777161, "learning_rate": 3.8225e-05, "loss": 0.1142, "step": 1530 }, { "grad_norm": 0.8228543996810913, "learning_rate": 3.8475e-05, "loss": 0.105, "step": 1540 }, { "grad_norm": 0.9993794560432434, "learning_rate": 3.8725e-05, "loss": 0.0975, "step": 1550 }, { "grad_norm": 0.7760846614837646, "learning_rate": 3.8975e-05, "loss": 0.1219, "step": 1560 }, { "grad_norm": 0.8031872510910034, "learning_rate": 3.9225e-05, "loss": 0.0962, "step": 1570 }, { "grad_norm": 1.0826778411865234, "learning_rate": 3.9475000000000004e-05, "loss": 0.1028, "step": 1580 }, { "grad_norm": 0.9071829319000244, "learning_rate": 3.9725e-05, "loss": 0.1129, "step": 1590 }, { "grad_norm": 0.9046777486801147, "learning_rate": 3.9975e-05, "loss": 0.1063, "step": 1600 }, { "grad_norm": 0.7598981857299805, "learning_rate": 4.0225e-05, "loss": 0.0863, "step": 1610 }, { "grad_norm": 0.8219106197357178, "learning_rate": 4.0475e-05, "loss": 0.1051, "step": 1620 }, { "grad_norm": 0.9724528193473816, "learning_rate": 4.0725e-05, "loss": 0.1279, "step": 1630 }, { "grad_norm": 0.9008668661117554, "learning_rate": 4.0975e-05, "loss": 0.1046, "step": 1640 }, { "grad_norm": 0.765158474445343, "learning_rate": 4.1225e-05, "loss": 0.1058, "step": 1650 }, { "grad_norm": 0.7991491556167603, "learning_rate": 4.1475000000000005e-05, "loss": 0.1076, "step": 1660 }, { "grad_norm": 0.9603814482688904, "learning_rate": 4.1725000000000005e-05, "loss": 0.0927, "step": 1670 }, { "grad_norm": 0.8996055126190186, "learning_rate": 4.1975000000000004e-05, "loss": 0.0962, "step": 1680 }, { "grad_norm": 0.8052534461021423, "learning_rate": 4.2225e-05, "loss": 0.0976, "step": 1690 }, { "grad_norm": 0.7835028767585754, "learning_rate": 4.2475e-05, "loss": 0.1023, "step": 1700 }, { "grad_norm": 1.238051414489746, "learning_rate": 4.2725e-05, "loss": 0.1231, "step": 1710 }, { "grad_norm": 1.0571870803833008, "learning_rate": 4.2975e-05, "loss": 0.1072, "step": 1720 }, { "grad_norm": 0.9382038116455078, "learning_rate": 4.322500000000001e-05, "loss": 0.1045, "step": 1730 }, { "grad_norm": 1.0474119186401367, "learning_rate": 4.3475000000000006e-05, "loss": 0.094, "step": 1740 }, { "grad_norm": 0.8908576369285583, "learning_rate": 4.3725000000000006e-05, "loss": 0.1028, "step": 1750 }, { "grad_norm": 1.0952162742614746, "learning_rate": 4.3975e-05, "loss": 0.1042, "step": 1760 }, { "grad_norm": 0.8382447957992554, "learning_rate": 4.4225e-05, "loss": 0.091, "step": 1770 }, { "grad_norm": 0.869783341884613, "learning_rate": 4.4475e-05, "loss": 0.1126, "step": 1780 }, { "grad_norm": 0.9035142064094543, "learning_rate": 4.4725e-05, "loss": 0.092, "step": 1790 }, { "grad_norm": 0.8749136924743652, "learning_rate": 4.4975e-05, "loss": 0.1001, "step": 1800 }, { "grad_norm": 0.722263514995575, "learning_rate": 4.5225e-05, "loss": 0.0922, "step": 1810 }, { "grad_norm": 1.1278636455535889, "learning_rate": 4.5475e-05, "loss": 0.1135, "step": 1820 }, { "grad_norm": 1.349082112312317, "learning_rate": 4.5725e-05, "loss": 0.1074, "step": 1830 }, { "grad_norm": 0.9080536961555481, "learning_rate": 4.5975e-05, "loss": 0.1128, "step": 1840 }, { "grad_norm": 0.9027018547058105, "learning_rate": 4.6225e-05, "loss": 0.1227, "step": 1850 }, { "grad_norm": 0.7627484798431396, "learning_rate": 4.6475000000000005e-05, "loss": 0.1252, "step": 1860 }, { "grad_norm": 0.7459473609924316, "learning_rate": 4.6725000000000004e-05, "loss": 0.093, "step": 1870 }, { "grad_norm": 0.9471239447593689, "learning_rate": 4.6975000000000003e-05, "loss": 0.0859, "step": 1880 }, { "grad_norm": 0.912121057510376, "learning_rate": 4.7225e-05, "loss": 0.0945, "step": 1890 }, { "grad_norm": 0.5690087080001831, "learning_rate": 4.7475e-05, "loss": 0.09, "step": 1900 }, { "grad_norm": 1.1898024082183838, "learning_rate": 4.7725e-05, "loss": 0.1169, "step": 1910 }, { "grad_norm": 0.6205836534500122, "learning_rate": 4.7975e-05, "loss": 0.0982, "step": 1920 }, { "grad_norm": 0.8883643746376038, "learning_rate": 4.822500000000001e-05, "loss": 0.0875, "step": 1930 }, { "grad_norm": 0.9483284950256348, "learning_rate": 4.8475000000000006e-05, "loss": 0.0967, "step": 1940 }, { "grad_norm": 0.6717948913574219, "learning_rate": 4.8725000000000005e-05, "loss": 0.0835, "step": 1950 }, { "grad_norm": 1.068524956703186, "learning_rate": 4.8975000000000005e-05, "loss": 0.1077, "step": 1960 }, { "grad_norm": 0.6906771063804626, "learning_rate": 4.9225000000000004e-05, "loss": 0.0864, "step": 1970 }, { "grad_norm": 1.0905823707580566, "learning_rate": 4.9475e-05, "loss": 0.1071, "step": 1980 }, { "grad_norm": 0.7720902562141418, "learning_rate": 4.9725e-05, "loss": 0.0817, "step": 1990 }, { "grad_norm": 0.7835018634796143, "learning_rate": 4.9975e-05, "loss": 0.107, "step": 2000 }, { "grad_norm": 0.7584051489830017, "learning_rate": 5.0225e-05, "loss": 0.1236, "step": 2010 }, { "grad_norm": 1.00912344455719, "learning_rate": 5.047500000000001e-05, "loss": 0.0982, "step": 2020 }, { "grad_norm": 0.9166877865791321, "learning_rate": 5.0725e-05, "loss": 0.0834, "step": 2030 }, { "grad_norm": 0.5230685472488403, "learning_rate": 5.0975000000000006e-05, "loss": 0.1065, "step": 2040 }, { "grad_norm": 0.8564101457595825, "learning_rate": 5.1225e-05, "loss": 0.1024, "step": 2050 }, { "grad_norm": 0.9942524433135986, "learning_rate": 5.1475000000000004e-05, "loss": 0.0956, "step": 2060 }, { "grad_norm": 0.8471402525901794, "learning_rate": 5.1725000000000004e-05, "loss": 0.0937, "step": 2070 }, { "grad_norm": 0.9816962480545044, "learning_rate": 5.197500000000001e-05, "loss": 0.1054, "step": 2080 }, { "grad_norm": 0.8807085156440735, "learning_rate": 5.2225e-05, "loss": 0.0978, "step": 2090 }, { "grad_norm": 0.8200457096099854, "learning_rate": 5.247500000000001e-05, "loss": 0.0994, "step": 2100 }, { "grad_norm": 0.5696097016334534, "learning_rate": 5.2725e-05, "loss": 0.1181, "step": 2110 }, { "grad_norm": 0.8671006560325623, "learning_rate": 5.297500000000001e-05, "loss": 0.1021, "step": 2120 }, { "grad_norm": 0.8476674556732178, "learning_rate": 5.3225e-05, "loss": 0.1012, "step": 2130 }, { "grad_norm": 0.7317443490028381, "learning_rate": 5.3475e-05, "loss": 0.08, "step": 2140 }, { "grad_norm": 0.8184857964515686, "learning_rate": 5.3725000000000005e-05, "loss": 0.1084, "step": 2150 }, { "grad_norm": 0.5154433250427246, "learning_rate": 5.3975e-05, "loss": 0.0887, "step": 2160 }, { "grad_norm": 0.5670997500419617, "learning_rate": 5.4225000000000003e-05, "loss": 0.0925, "step": 2170 }, { "grad_norm": 0.900976836681366, "learning_rate": 5.4474999999999996e-05, "loss": 0.1047, "step": 2180 }, { "grad_norm": 0.63798588514328, "learning_rate": 5.4725e-05, "loss": 0.0878, "step": 2190 }, { "grad_norm": 0.5306079983711243, "learning_rate": 5.4975e-05, "loss": 0.0839, "step": 2200 }, { "grad_norm": 0.590713381767273, "learning_rate": 5.522500000000001e-05, "loss": 0.0872, "step": 2210 }, { "grad_norm": 1.2493854761123657, "learning_rate": 5.5475e-05, "loss": 0.0915, "step": 2220 }, { "grad_norm": 0.7268276810646057, "learning_rate": 5.5725000000000006e-05, "loss": 0.0882, "step": 2230 }, { "grad_norm": 0.684813380241394, "learning_rate": 5.5975e-05, "loss": 0.0835, "step": 2240 }, { "grad_norm": 0.718123197555542, "learning_rate": 5.6225000000000005e-05, "loss": 0.0969, "step": 2250 }, { "grad_norm": 0.6042284369468689, "learning_rate": 5.6475e-05, "loss": 0.0856, "step": 2260 }, { "grad_norm": 0.5972579717636108, "learning_rate": 5.6725e-05, "loss": 0.0771, "step": 2270 }, { "grad_norm": 0.7482088804244995, "learning_rate": 5.6975e-05, "loss": 0.0792, "step": 2280 }, { "grad_norm": 0.9524646997451782, "learning_rate": 5.722500000000001e-05, "loss": 0.106, "step": 2290 }, { "grad_norm": 0.8603826761245728, "learning_rate": 5.7475e-05, "loss": 0.0838, "step": 2300 }, { "grad_norm": 0.7626532912254333, "learning_rate": 5.772500000000001e-05, "loss": 0.09, "step": 2310 }, { "grad_norm": 0.8651349544525146, "learning_rate": 5.7975e-05, "loss": 0.0729, "step": 2320 }, { "grad_norm": 0.7227286100387573, "learning_rate": 5.8225000000000006e-05, "loss": 0.1027, "step": 2330 }, { "grad_norm": 0.8561999797821045, "learning_rate": 5.8475000000000005e-05, "loss": 0.0915, "step": 2340 }, { "grad_norm": 0.7053679823875427, "learning_rate": 5.8725000000000004e-05, "loss": 0.096, "step": 2350 }, { "grad_norm": 0.7833651900291443, "learning_rate": 5.8975000000000004e-05, "loss": 0.0982, "step": 2360 }, { "grad_norm": 0.630891740322113, "learning_rate": 5.922500000000001e-05, "loss": 0.0873, "step": 2370 }, { "grad_norm": 0.6595512628555298, "learning_rate": 5.9475e-05, "loss": 0.0977, "step": 2380 }, { "grad_norm": 0.6549544930458069, "learning_rate": 5.9724999999999995e-05, "loss": 0.1031, "step": 2390 }, { "grad_norm": 0.6938033103942871, "learning_rate": 5.9975e-05, "loss": 0.0883, "step": 2400 }, { "grad_norm": 0.6153424978256226, "learning_rate": 6.0225e-05, "loss": 0.0865, "step": 2410 }, { "grad_norm": 0.6463296413421631, "learning_rate": 6.0475000000000006e-05, "loss": 0.1145, "step": 2420 }, { "grad_norm": 0.8225053548812866, "learning_rate": 6.0725e-05, "loss": 0.0953, "step": 2430 }, { "grad_norm": 0.7964223623275757, "learning_rate": 6.0975000000000005e-05, "loss": 0.0981, "step": 2440 }, { "grad_norm": 0.8822996020317078, "learning_rate": 6.1225e-05, "loss": 0.0852, "step": 2450 }, { "grad_norm": 0.642469048500061, "learning_rate": 6.1475e-05, "loss": 0.0786, "step": 2460 }, { "grad_norm": 0.652252733707428, "learning_rate": 6.1725e-05, "loss": 0.0823, "step": 2470 }, { "grad_norm": 0.6256334185600281, "learning_rate": 6.1975e-05, "loss": 0.1122, "step": 2480 }, { "grad_norm": 0.7826414704322815, "learning_rate": 6.2225e-05, "loss": 0.0825, "step": 2490 }, { "grad_norm": 0.6647141575813293, "learning_rate": 6.2475e-05, "loss": 0.0812, "step": 2500 }, { "grad_norm": 0.5930452346801758, "learning_rate": 6.2725e-05, "loss": 0.0692, "step": 2510 }, { "grad_norm": 0.7224600911140442, "learning_rate": 6.297500000000001e-05, "loss": 0.0814, "step": 2520 }, { "grad_norm": 0.604549765586853, "learning_rate": 6.3225e-05, "loss": 0.0768, "step": 2530 }, { "grad_norm": 0.6261560916900635, "learning_rate": 6.347500000000001e-05, "loss": 0.0822, "step": 2540 }, { "grad_norm": 0.7654407024383545, "learning_rate": 6.3725e-05, "loss": 0.0853, "step": 2550 }, { "grad_norm": 0.5477272272109985, "learning_rate": 6.397500000000001e-05, "loss": 0.0899, "step": 2560 }, { "grad_norm": 0.6548789143562317, "learning_rate": 6.4225e-05, "loss": 0.0807, "step": 2570 }, { "grad_norm": 0.6279175281524658, "learning_rate": 6.447500000000001e-05, "loss": 0.0803, "step": 2580 }, { "grad_norm": 0.6270503401756287, "learning_rate": 6.4725e-05, "loss": 0.0801, "step": 2590 }, { "grad_norm": 0.6899020671844482, "learning_rate": 6.497500000000001e-05, "loss": 0.0895, "step": 2600 }, { "grad_norm": 0.6946532130241394, "learning_rate": 6.5225e-05, "loss": 0.0891, "step": 2610 }, { "grad_norm": 0.6085049510002136, "learning_rate": 6.5475e-05, "loss": 0.0702, "step": 2620 }, { "grad_norm": 0.6051257252693176, "learning_rate": 6.5725e-05, "loss": 0.077, "step": 2630 }, { "grad_norm": 0.6420267224311829, "learning_rate": 6.5975e-05, "loss": 0.0939, "step": 2640 }, { "grad_norm": 0.5438717603683472, "learning_rate": 6.6225e-05, "loss": 0.0922, "step": 2650 }, { "grad_norm": 0.6388781666755676, "learning_rate": 6.6475e-05, "loss": 0.0945, "step": 2660 }, { "grad_norm": 0.8712764382362366, "learning_rate": 6.672500000000001e-05, "loss": 0.1004, "step": 2670 }, { "grad_norm": 1.0257817506790161, "learning_rate": 6.6975e-05, "loss": 0.113, "step": 2680 }, { "grad_norm": 0.7105761170387268, "learning_rate": 6.722500000000001e-05, "loss": 0.0857, "step": 2690 }, { "grad_norm": 0.6772826313972473, "learning_rate": 6.7475e-05, "loss": 0.0786, "step": 2700 }, { "grad_norm": 0.6383380889892578, "learning_rate": 6.7725e-05, "loss": 0.09, "step": 2710 }, { "grad_norm": 0.7018582224845886, "learning_rate": 6.7975e-05, "loss": 0.1019, "step": 2720 }, { "grad_norm": 0.6313619613647461, "learning_rate": 6.8225e-05, "loss": 0.0879, "step": 2730 }, { "grad_norm": 0.5272067785263062, "learning_rate": 6.8475e-05, "loss": 0.0803, "step": 2740 }, { "grad_norm": 0.7778469920158386, "learning_rate": 6.8725e-05, "loss": 0.0877, "step": 2750 }, { "grad_norm": 0.7492944002151489, "learning_rate": 6.8975e-05, "loss": 0.0811, "step": 2760 }, { "grad_norm": 0.5935215950012207, "learning_rate": 6.9225e-05, "loss": 0.0847, "step": 2770 }, { "grad_norm": 0.7024961709976196, "learning_rate": 6.9475e-05, "loss": 0.0899, "step": 2780 }, { "grad_norm": 0.49780595302581787, "learning_rate": 6.9725e-05, "loss": 0.084, "step": 2790 }, { "grad_norm": 0.5282698273658752, "learning_rate": 6.997500000000001e-05, "loss": 0.0756, "step": 2800 }, { "grad_norm": 0.5787959098815918, "learning_rate": 7.022500000000001e-05, "loss": 0.1042, "step": 2810 }, { "grad_norm": 0.735684335231781, "learning_rate": 7.0475e-05, "loss": 0.0928, "step": 2820 }, { "grad_norm": 0.7814189195632935, "learning_rate": 7.072500000000001e-05, "loss": 0.0802, "step": 2830 }, { "grad_norm": 0.43192628026008606, "learning_rate": 7.0975e-05, "loss": 0.0949, "step": 2840 }, { "grad_norm": 0.6132184863090515, "learning_rate": 7.122500000000001e-05, "loss": 0.0993, "step": 2850 }, { "grad_norm": 0.5912649035453796, "learning_rate": 7.1475e-05, "loss": 0.0809, "step": 2860 }, { "grad_norm": 0.6607341170310974, "learning_rate": 7.172500000000001e-05, "loss": 0.0764, "step": 2870 }, { "grad_norm": 0.605894923210144, "learning_rate": 7.1975e-05, "loss": 0.0704, "step": 2880 }, { "grad_norm": 0.6907094717025757, "learning_rate": 7.2225e-05, "loss": 0.0834, "step": 2890 }, { "grad_norm": 0.46912986040115356, "learning_rate": 7.2475e-05, "loss": 0.0911, "step": 2900 }, { "grad_norm": 0.681545615196228, "learning_rate": 7.272499999999999e-05, "loss": 0.0759, "step": 2910 }, { "grad_norm": 0.5072075128555298, "learning_rate": 7.2975e-05, "loss": 0.0866, "step": 2920 }, { "grad_norm": 0.602841854095459, "learning_rate": 7.3225e-05, "loss": 0.0684, "step": 2930 }, { "grad_norm": 0.6649315357208252, "learning_rate": 7.347500000000001e-05, "loss": 0.0685, "step": 2940 }, { "grad_norm": 0.7016647458076477, "learning_rate": 7.3725e-05, "loss": 0.0697, "step": 2950 }, { "grad_norm": 0.693024218082428, "learning_rate": 7.397500000000001e-05, "loss": 0.0833, "step": 2960 }, { "grad_norm": 0.48047009110450745, "learning_rate": 7.4225e-05, "loss": 0.0727, "step": 2970 }, { "grad_norm": 0.6501755118370056, "learning_rate": 7.447500000000001e-05, "loss": 0.0844, "step": 2980 }, { "grad_norm": 0.6456348299980164, "learning_rate": 7.4725e-05, "loss": 0.089, "step": 2990 }, { "grad_norm": 0.4695075750350952, "learning_rate": 7.4975e-05, "loss": 0.0895, "step": 3000 }, { "grad_norm": 0.4707720875740051, "learning_rate": 7.5225e-05, "loss": 0.0678, "step": 3010 }, { "grad_norm": 0.6672714352607727, "learning_rate": 7.5475e-05, "loss": 0.0791, "step": 3020 }, { "grad_norm": 0.6611412167549133, "learning_rate": 7.5725e-05, "loss": 0.0737, "step": 3030 }, { "grad_norm": 0.530724048614502, "learning_rate": 7.5975e-05, "loss": 0.0768, "step": 3040 }, { "grad_norm": 0.5026446580886841, "learning_rate": 7.6225e-05, "loss": 0.074, "step": 3050 }, { "grad_norm": 0.5563125610351562, "learning_rate": 7.6475e-05, "loss": 0.0751, "step": 3060 }, { "grad_norm": 0.6619698405265808, "learning_rate": 7.672500000000001e-05, "loss": 0.0656, "step": 3070 }, { "grad_norm": 0.589493453502655, "learning_rate": 7.697500000000001e-05, "loss": 0.0775, "step": 3080 }, { "grad_norm": 0.829303503036499, "learning_rate": 7.722500000000001e-05, "loss": 0.0974, "step": 3090 }, { "grad_norm": 0.601528525352478, "learning_rate": 7.747500000000001e-05, "loss": 0.0771, "step": 3100 }, { "grad_norm": 0.6697906255722046, "learning_rate": 7.7725e-05, "loss": 0.071, "step": 3110 }, { "grad_norm": 0.5676504969596863, "learning_rate": 7.797500000000001e-05, "loss": 0.0692, "step": 3120 }, { "grad_norm": 0.5201124548912048, "learning_rate": 7.8225e-05, "loss": 0.0768, "step": 3130 }, { "grad_norm": 0.5930479168891907, "learning_rate": 7.8475e-05, "loss": 0.0859, "step": 3140 }, { "grad_norm": 0.6003865003585815, "learning_rate": 7.8725e-05, "loss": 0.0739, "step": 3150 }, { "grad_norm": 0.6613808870315552, "learning_rate": 7.8975e-05, "loss": 0.0769, "step": 3160 }, { "grad_norm": 0.6118493676185608, "learning_rate": 7.9225e-05, "loss": 0.0726, "step": 3170 }, { "grad_norm": 0.4647340476512909, "learning_rate": 7.9475e-05, "loss": 0.0907, "step": 3180 }, { "grad_norm": 0.7966919541358948, "learning_rate": 7.9725e-05, "loss": 0.093, "step": 3190 }, { "grad_norm": 0.525335967540741, "learning_rate": 7.9975e-05, "loss": 0.0891, "step": 3200 }, { "grad_norm": 0.5029305219650269, "learning_rate": 8.022500000000001e-05, "loss": 0.0757, "step": 3210 }, { "grad_norm": 0.7103932499885559, "learning_rate": 8.0475e-05, "loss": 0.0995, "step": 3220 }, { "grad_norm": 0.5534277558326721, "learning_rate": 8.072500000000001e-05, "loss": 0.0953, "step": 3230 }, { "grad_norm": 0.6976082921028137, "learning_rate": 8.0975e-05, "loss": 0.0821, "step": 3240 }, { "grad_norm": 0.7198306322097778, "learning_rate": 8.122500000000001e-05, "loss": 0.0938, "step": 3250 }, { "grad_norm": 0.7189089059829712, "learning_rate": 8.1475e-05, "loss": 0.0855, "step": 3260 }, { "grad_norm": 0.7195263504981995, "learning_rate": 8.172500000000001e-05, "loss": 0.0829, "step": 3270 }, { "grad_norm": 0.39758312702178955, "learning_rate": 8.1975e-05, "loss": 0.0735, "step": 3280 }, { "grad_norm": 0.3971259891986847, "learning_rate": 8.2225e-05, "loss": 0.0715, "step": 3290 }, { "grad_norm": 0.5517513155937195, "learning_rate": 8.2475e-05, "loss": 0.0743, "step": 3300 }, { "grad_norm": 0.5765997767448425, "learning_rate": 8.2725e-05, "loss": 0.0838, "step": 3310 }, { "grad_norm": 0.6094714403152466, "learning_rate": 8.2975e-05, "loss": 0.0911, "step": 3320 }, { "grad_norm": 0.6877467632293701, "learning_rate": 8.3225e-05, "loss": 0.0754, "step": 3330 }, { "grad_norm": 0.3906482458114624, "learning_rate": 8.347500000000001e-05, "loss": 0.0718, "step": 3340 }, { "grad_norm": 0.534462034702301, "learning_rate": 8.3725e-05, "loss": 0.0734, "step": 3350 }, { "grad_norm": 0.713546633720398, "learning_rate": 8.397500000000001e-05, "loss": 0.0777, "step": 3360 }, { "grad_norm": 0.4991622269153595, "learning_rate": 8.422500000000001e-05, "loss": 0.0785, "step": 3370 }, { "grad_norm": 0.5688065886497498, "learning_rate": 8.447500000000001e-05, "loss": 0.1154, "step": 3380 }, { "grad_norm": 0.5358280539512634, "learning_rate": 8.4725e-05, "loss": 0.0694, "step": 3390 }, { "grad_norm": 0.6376456022262573, "learning_rate": 8.4975e-05, "loss": 0.0834, "step": 3400 }, { "grad_norm": 0.5692137479782104, "learning_rate": 8.5225e-05, "loss": 0.0761, "step": 3410 }, { "grad_norm": 0.6078459620475769, "learning_rate": 8.5475e-05, "loss": 0.068, "step": 3420 }, { "grad_norm": 0.6219914555549622, "learning_rate": 8.5725e-05, "loss": 0.0786, "step": 3430 }, { "grad_norm": 0.6457554697990417, "learning_rate": 8.5975e-05, "loss": 0.0944, "step": 3440 }, { "grad_norm": 0.6305280923843384, "learning_rate": 8.6225e-05, "loss": 0.0802, "step": 3450 }, { "grad_norm": 0.46669429540634155, "learning_rate": 8.6475e-05, "loss": 0.0729, "step": 3460 }, { "grad_norm": 0.5616320967674255, "learning_rate": 8.672500000000001e-05, "loss": 0.0904, "step": 3470 }, { "grad_norm": 0.5622738599777222, "learning_rate": 8.6975e-05, "loss": 0.0887, "step": 3480 }, { "grad_norm": 0.4849782884120941, "learning_rate": 8.7225e-05, "loss": 0.0863, "step": 3490 }, { "grad_norm": 0.46744203567504883, "learning_rate": 8.747500000000001e-05, "loss": 0.0747, "step": 3500 }, { "grad_norm": 0.5717671513557434, "learning_rate": 8.7725e-05, "loss": 0.0799, "step": 3510 }, { "grad_norm": 0.3920367956161499, "learning_rate": 8.797500000000001e-05, "loss": 0.0778, "step": 3520 }, { "grad_norm": 0.5349674820899963, "learning_rate": 8.8225e-05, "loss": 0.0762, "step": 3530 }, { "grad_norm": 0.45464086532592773, "learning_rate": 8.847500000000001e-05, "loss": 0.0721, "step": 3540 }, { "grad_norm": 0.4125288724899292, "learning_rate": 8.8725e-05, "loss": 0.0768, "step": 3550 }, { "grad_norm": 0.44126155972480774, "learning_rate": 8.897500000000001e-05, "loss": 0.061, "step": 3560 }, { "grad_norm": 0.5346813201904297, "learning_rate": 8.9225e-05, "loss": 0.0817, "step": 3570 }, { "grad_norm": 0.48566192388534546, "learning_rate": 8.9475e-05, "loss": 0.0867, "step": 3580 }, { "grad_norm": 0.5075343251228333, "learning_rate": 8.9725e-05, "loss": 0.0852, "step": 3590 }, { "grad_norm": 0.5118183493614197, "learning_rate": 8.9975e-05, "loss": 0.0783, "step": 3600 }, { "grad_norm": 0.34911447763442993, "learning_rate": 9.0225e-05, "loss": 0.0799, "step": 3610 }, { "grad_norm": 0.5489487648010254, "learning_rate": 9.0475e-05, "loss": 0.0757, "step": 3620 }, { "grad_norm": 0.6597517728805542, "learning_rate": 9.072500000000001e-05, "loss": 0.0635, "step": 3630 }, { "grad_norm": 0.7097198367118835, "learning_rate": 9.0975e-05, "loss": 0.0754, "step": 3640 }, { "grad_norm": 0.47328028082847595, "learning_rate": 9.122500000000001e-05, "loss": 0.0798, "step": 3650 }, { "grad_norm": 0.459682434797287, "learning_rate": 9.1475e-05, "loss": 0.076, "step": 3660 }, { "grad_norm": 0.6338856220245361, "learning_rate": 9.172500000000001e-05, "loss": 0.0898, "step": 3670 }, { "grad_norm": 0.7858336567878723, "learning_rate": 9.1975e-05, "loss": 0.0833, "step": 3680 }, { "grad_norm": 0.5354365110397339, "learning_rate": 9.2225e-05, "loss": 0.0811, "step": 3690 }, { "grad_norm": 0.4700893759727478, "learning_rate": 9.2475e-05, "loss": 0.0705, "step": 3700 }, { "grad_norm": 0.4933316111564636, "learning_rate": 9.2725e-05, "loss": 0.072, "step": 3710 }, { "grad_norm": 0.49771785736083984, "learning_rate": 9.2975e-05, "loss": 0.0805, "step": 3720 }, { "grad_norm": 0.4712162911891937, "learning_rate": 9.3225e-05, "loss": 0.0652, "step": 3730 }, { "grad_norm": 0.5891095399856567, "learning_rate": 9.3475e-05, "loss": 0.0708, "step": 3740 }, { "grad_norm": 0.5640130639076233, "learning_rate": 9.3725e-05, "loss": 0.0807, "step": 3750 }, { "grad_norm": 0.4983829855918884, "learning_rate": 9.397500000000001e-05, "loss": 0.0689, "step": 3760 }, { "grad_norm": 0.4472532868385315, "learning_rate": 9.422500000000001e-05, "loss": 0.0893, "step": 3770 }, { "grad_norm": 0.43757274746894836, "learning_rate": 9.4475e-05, "loss": 0.0848, "step": 3780 }, { "grad_norm": 0.5159354209899902, "learning_rate": 9.472500000000001e-05, "loss": 0.0705, "step": 3790 }, { "grad_norm": 0.5909969806671143, "learning_rate": 9.4975e-05, "loss": 0.1027, "step": 3800 }, { "grad_norm": 0.6230734586715698, "learning_rate": 9.522500000000001e-05, "loss": 0.0714, "step": 3810 }, { "grad_norm": 0.7051389217376709, "learning_rate": 9.5475e-05, "loss": 0.0794, "step": 3820 }, { "grad_norm": 0.45560190081596375, "learning_rate": 9.572500000000001e-05, "loss": 0.0677, "step": 3830 }, { "grad_norm": 0.564217209815979, "learning_rate": 9.5975e-05, "loss": 0.0664, "step": 3840 }, { "grad_norm": 0.5676417350769043, "learning_rate": 9.622500000000001e-05, "loss": 0.0835, "step": 3850 }, { "grad_norm": 0.5275018811225891, "learning_rate": 9.6475e-05, "loss": 0.0719, "step": 3860 }, { "grad_norm": 0.4657714366912842, "learning_rate": 9.6725e-05, "loss": 0.0627, "step": 3870 }, { "grad_norm": 0.6800485253334045, "learning_rate": 9.6975e-05, "loss": 0.0689, "step": 3880 }, { "grad_norm": 0.6019474267959595, "learning_rate": 9.7225e-05, "loss": 0.0879, "step": 3890 }, { "grad_norm": 0.5059846639633179, "learning_rate": 9.747500000000001e-05, "loss": 0.0805, "step": 3900 }, { "grad_norm": 0.344991534948349, "learning_rate": 9.7725e-05, "loss": 0.0728, "step": 3910 }, { "grad_norm": 0.5577927827835083, "learning_rate": 9.797500000000001e-05, "loss": 0.0877, "step": 3920 }, { "grad_norm": 0.4784092605113983, "learning_rate": 9.8225e-05, "loss": 0.068, "step": 3930 }, { "grad_norm": 0.5226242542266846, "learning_rate": 9.847500000000001e-05, "loss": 0.0826, "step": 3940 }, { "grad_norm": 0.5615122318267822, "learning_rate": 9.8725e-05, "loss": 0.0695, "step": 3950 }, { "grad_norm": 0.3639674186706543, "learning_rate": 9.897500000000001e-05, "loss": 0.0879, "step": 3960 }, { "grad_norm": 0.5316547155380249, "learning_rate": 9.9225e-05, "loss": 0.0701, "step": 3970 }, { "grad_norm": 0.43595749139785767, "learning_rate": 9.9475e-05, "loss": 0.0803, "step": 3980 }, { "grad_norm": 0.4905216693878174, "learning_rate": 9.9725e-05, "loss": 0.0761, "step": 3990 }, { "grad_norm": 0.6213178634643555, "learning_rate": 9.9975e-05, "loss": 0.0849, "step": 4000 }, { "grad_norm": 0.5712447166442871, "learning_rate": 9.999999653982884e-05, "loss": 0.0726, "step": 4010 }, { "grad_norm": 0.4938797056674957, "learning_rate": 9.999998457874392e-05, "loss": 0.078, "step": 4020 }, { "grad_norm": 0.6088446378707886, "learning_rate": 9.999996407402913e-05, "loss": 0.0814, "step": 4030 }, { "grad_norm": 0.39301398396492004, "learning_rate": 9.999993502568801e-05, "loss": 0.0656, "step": 4040 }, { "grad_norm": 0.5613072514533997, "learning_rate": 9.999989743372548e-05, "loss": 0.0745, "step": 4050 }, { "grad_norm": 0.5090293288230896, "learning_rate": 9.999985129814798e-05, "loss": 0.0699, "step": 4060 }, { "grad_norm": 0.458036869764328, "learning_rate": 9.99997966189634e-05, "loss": 0.0629, "step": 4070 }, { "grad_norm": 0.4278196096420288, "learning_rate": 9.999973339618107e-05, "loss": 0.0837, "step": 4080 }, { "grad_norm": 0.6164867281913757, "learning_rate": 9.999966162981179e-05, "loss": 0.0732, "step": 4090 }, { "grad_norm": 0.5170760154724121, "learning_rate": 9.999958131986784e-05, "loss": 0.069, "step": 4100 }, { "grad_norm": 0.5297067165374756, "learning_rate": 9.999949246636293e-05, "loss": 0.0822, "step": 4110 }, { "grad_norm": 0.47822171449661255, "learning_rate": 9.999939506931224e-05, "loss": 0.0784, "step": 4120 }, { "grad_norm": 0.5428102612495422, "learning_rate": 9.999928912873243e-05, "loss": 0.0716, "step": 4130 }, { "grad_norm": 0.44498932361602783, "learning_rate": 9.999917464464159e-05, "loss": 0.0802, "step": 4140 }, { "grad_norm": 0.5112330317497253, "learning_rate": 9.999905161705929e-05, "loss": 0.0835, "step": 4150 }, { "grad_norm": 0.5264190435409546, "learning_rate": 9.999892004600653e-05, "loss": 0.0692, "step": 4160 }, { "grad_norm": 0.4274420440196991, "learning_rate": 9.999877993150581e-05, "loss": 0.0704, "step": 4170 }, { "grad_norm": 0.5425940752029419, "learning_rate": 9.999863127358108e-05, "loss": 0.0674, "step": 4180 }, { "grad_norm": 0.47383183240890503, "learning_rate": 9.999847407225773e-05, "loss": 0.0738, "step": 4190 }, { "grad_norm": 0.6969469785690308, "learning_rate": 9.999830832756262e-05, "loss": 0.0725, "step": 4200 }, { "grad_norm": 0.4476061761379242, "learning_rate": 9.999813403952407e-05, "loss": 0.0702, "step": 4210 }, { "grad_norm": 0.7133558392524719, "learning_rate": 9.999795120817187e-05, "loss": 0.0936, "step": 4220 }, { "grad_norm": 0.46676766872406006, "learning_rate": 9.999775983353725e-05, "loss": 0.092, "step": 4230 }, { "grad_norm": 0.547597348690033, "learning_rate": 9.999755991565292e-05, "loss": 0.0804, "step": 4240 }, { "grad_norm": 0.5866601467132568, "learning_rate": 9.999735145455303e-05, "loss": 0.0855, "step": 4250 }, { "grad_norm": 0.5722534656524658, "learning_rate": 9.99971344502732e-05, "loss": 0.0834, "step": 4260 }, { "grad_norm": 0.481266587972641, "learning_rate": 9.999690890285053e-05, "loss": 0.0747, "step": 4270 }, { "grad_norm": 0.5264154076576233, "learning_rate": 9.999667481232356e-05, "loss": 0.0665, "step": 4280 }, { "grad_norm": 0.49708929657936096, "learning_rate": 9.999643217873225e-05, "loss": 0.0719, "step": 4290 }, { "grad_norm": 0.45181509852409363, "learning_rate": 9.999618100211809e-05, "loss": 0.0679, "step": 4300 }, { "grad_norm": 0.3943522870540619, "learning_rate": 9.999592128252402e-05, "loss": 0.0697, "step": 4310 }, { "grad_norm": 0.4263547956943512, "learning_rate": 9.999565301999437e-05, "loss": 0.0945, "step": 4320 }, { "grad_norm": 0.48950430750846863, "learning_rate": 9.999537621457502e-05, "loss": 0.0664, "step": 4330 }, { "grad_norm": 0.46132609248161316, "learning_rate": 9.999509086631323e-05, "loss": 0.0599, "step": 4340 }, { "grad_norm": 0.46957898139953613, "learning_rate": 9.99947969752578e-05, "loss": 0.0563, "step": 4350 }, { "grad_norm": 0.3828161656856537, "learning_rate": 9.999449454145891e-05, "loss": 0.0552, "step": 4360 }, { "grad_norm": 0.4312327206134796, "learning_rate": 9.999418356496827e-05, "loss": 0.0639, "step": 4370 }, { "grad_norm": 0.5646336078643799, "learning_rate": 9.999386404583899e-05, "loss": 0.0689, "step": 4380 }, { "grad_norm": 0.6735820174217224, "learning_rate": 9.999353598412568e-05, "loss": 0.0685, "step": 4390 }, { "grad_norm": 0.6168258786201477, "learning_rate": 9.999319937988442e-05, "loss": 0.0802, "step": 4400 }, { "grad_norm": 0.52671217918396, "learning_rate": 9.999285423317268e-05, "loss": 0.0666, "step": 4410 }, { "grad_norm": 0.4577769935131073, "learning_rate": 9.999250054404947e-05, "loss": 0.0868, "step": 4420 }, { "grad_norm": 0.48866117000579834, "learning_rate": 9.99921383125752e-05, "loss": 0.0709, "step": 4430 }, { "grad_norm": 0.43695497512817383, "learning_rate": 9.99917675388118e-05, "loss": 0.0713, "step": 4440 }, { "grad_norm": 0.5032289624214172, "learning_rate": 9.99913882228226e-05, "loss": 0.0828, "step": 4450 }, { "grad_norm": 0.535183846950531, "learning_rate": 9.999100036467242e-05, "loss": 0.0748, "step": 4460 }, { "grad_norm": 0.3468345105648041, "learning_rate": 9.999060396442753e-05, "loss": 0.0637, "step": 4470 }, { "grad_norm": 0.4700489044189453, "learning_rate": 9.999019902215566e-05, "loss": 0.0715, "step": 4480 }, { "grad_norm": 0.4966925382614136, "learning_rate": 9.998978553792602e-05, "loss": 0.0747, "step": 4490 }, { "grad_norm": 0.5608887672424316, "learning_rate": 9.998936351180926e-05, "loss": 0.0753, "step": 4500 }, { "grad_norm": 0.5167213678359985, "learning_rate": 9.998893294387747e-05, "loss": 0.0946, "step": 4510 }, { "grad_norm": 0.43413108587265015, "learning_rate": 9.998849383420426e-05, "loss": 0.0669, "step": 4520 }, { "grad_norm": 0.5757310390472412, "learning_rate": 9.998804618286465e-05, "loss": 0.0659, "step": 4530 }, { "grad_norm": 0.5909011363983154, "learning_rate": 9.99875899899351e-05, "loss": 0.0883, "step": 4540 }, { "grad_norm": 0.3669481873512268, "learning_rate": 9.99871252554936e-05, "loss": 0.0798, "step": 4550 }, { "grad_norm": 0.4152976870536804, "learning_rate": 9.998665197961955e-05, "loss": 0.0686, "step": 4560 }, { "grad_norm": 0.5814568400382996, "learning_rate": 9.998617016239379e-05, "loss": 0.0831, "step": 4570 }, { "grad_norm": 0.5006296634674072, "learning_rate": 9.998567980389869e-05, "loss": 0.0703, "step": 4580 }, { "grad_norm": 0.7360182404518127, "learning_rate": 9.998518090421802e-05, "loss": 0.0698, "step": 4590 }, { "grad_norm": 0.5398156046867371, "learning_rate": 9.998467346343703e-05, "loss": 0.0676, "step": 4600 }, { "grad_norm": 0.5550403594970703, "learning_rate": 9.998415748164243e-05, "loss": 0.0664, "step": 4610 }, { "grad_norm": 0.4512180685997009, "learning_rate": 9.998363295892238e-05, "loss": 0.0629, "step": 4620 }, { "grad_norm": 0.6895257830619812, "learning_rate": 9.998309989536652e-05, "loss": 0.0756, "step": 4630 }, { "grad_norm": 0.5178534984588623, "learning_rate": 9.998255829106593e-05, "loss": 0.0799, "step": 4640 }, { "grad_norm": 0.46622535586357117, "learning_rate": 9.998200814611316e-05, "loss": 0.0683, "step": 4650 }, { "grad_norm": 0.3376472592353821, "learning_rate": 9.998144946060219e-05, "loss": 0.0789, "step": 4660 }, { "grad_norm": 0.7337660193443298, "learning_rate": 9.998088223462852e-05, "loss": 0.0779, "step": 4670 }, { "grad_norm": 0.44438865780830383, "learning_rate": 9.998030646828905e-05, "loss": 0.0774, "step": 4680 }, { "grad_norm": 0.42340436577796936, "learning_rate": 9.997972216168217e-05, "loss": 0.0832, "step": 4690 }, { "grad_norm": 0.403886616230011, "learning_rate": 9.997912931490771e-05, "loss": 0.0711, "step": 4700 }, { "grad_norm": 0.4263748228549957, "learning_rate": 9.9978527928067e-05, "loss": 0.072, "step": 4710 }, { "grad_norm": 0.43720102310180664, "learning_rate": 9.997791800126277e-05, "loss": 0.0623, "step": 4720 }, { "grad_norm": 0.430374413728714, "learning_rate": 9.997729953459927e-05, "loss": 0.0876, "step": 4730 }, { "grad_norm": 0.3946286141872406, "learning_rate": 9.997667252818214e-05, "loss": 0.079, "step": 4740 }, { "grad_norm": 0.38153815269470215, "learning_rate": 9.997603698211855e-05, "loss": 0.0649, "step": 4750 }, { "grad_norm": 0.4520833492279053, "learning_rate": 9.99753928965171e-05, "loss": 0.0751, "step": 4760 }, { "grad_norm": 0.4603523910045624, "learning_rate": 9.997474027148781e-05, "loss": 0.0784, "step": 4770 }, { "grad_norm": 0.5559136271476746, "learning_rate": 9.997407910714223e-05, "loss": 0.0938, "step": 4780 }, { "grad_norm": 0.46750855445861816, "learning_rate": 9.997340940359332e-05, "loss": 0.0709, "step": 4790 }, { "grad_norm": 0.447765052318573, "learning_rate": 9.997273116095552e-05, "loss": 0.0719, "step": 4800 }, { "grad_norm": 0.45837026834487915, "learning_rate": 9.997204437934473e-05, "loss": 0.0704, "step": 4810 }, { "grad_norm": 0.5094914436340332, "learning_rate": 9.997134905887829e-05, "loss": 0.0758, "step": 4820 }, { "grad_norm": 0.4294559061527252, "learning_rate": 9.997064519967501e-05, "loss": 0.0697, "step": 4830 }, { "grad_norm": 0.4536200761795044, "learning_rate": 9.996993280185517e-05, "loss": 0.0725, "step": 4840 }, { "grad_norm": 0.3864877223968506, "learning_rate": 9.99692118655405e-05, "loss": 0.0685, "step": 4850 }, { "grad_norm": 0.3821287751197815, "learning_rate": 9.996848239085417e-05, "loss": 0.0649, "step": 4860 }, { "grad_norm": 0.39283987879753113, "learning_rate": 9.996774437792085e-05, "loss": 0.0663, "step": 4870 }, { "grad_norm": 0.5281031131744385, "learning_rate": 9.996699782686664e-05, "loss": 0.0605, "step": 4880 }, { "grad_norm": 0.5055439472198486, "learning_rate": 9.996624273781909e-05, "loss": 0.083, "step": 4890 }, { "grad_norm": 0.5166409015655518, "learning_rate": 9.996547911090725e-05, "loss": 0.0579, "step": 4900 }, { "grad_norm": 0.389724463224411, "learning_rate": 9.996470694626157e-05, "loss": 0.0659, "step": 4910 }, { "grad_norm": 0.4282790422439575, "learning_rate": 9.996392624401403e-05, "loss": 0.0822, "step": 4920 }, { "grad_norm": 0.5900306701660156, "learning_rate": 9.996313700429801e-05, "loss": 0.0706, "step": 4930 }, { "grad_norm": 0.4257619380950928, "learning_rate": 9.996233922724836e-05, "loss": 0.0642, "step": 4940 }, { "grad_norm": 0.3807382583618164, "learning_rate": 9.996153291300141e-05, "loss": 0.0715, "step": 4950 }, { "grad_norm": 0.30842137336730957, "learning_rate": 9.996071806169494e-05, "loss": 0.0682, "step": 4960 }, { "grad_norm": 0.37914982438087463, "learning_rate": 9.995989467346817e-05, "loss": 0.0763, "step": 4970 }, { "grad_norm": 0.5185093283653259, "learning_rate": 9.995906274846183e-05, "loss": 0.0665, "step": 4980 }, { "grad_norm": 0.42891693115234375, "learning_rate": 9.995822228681803e-05, "loss": 0.0632, "step": 4990 }, { "grad_norm": 0.3649630546569824, "learning_rate": 9.99573732886804e-05, "loss": 0.0667, "step": 5000 }, { "grad_norm": 0.34290575981140137, "learning_rate": 9.995651575419402e-05, "loss": 0.0635, "step": 5010 }, { "grad_norm": 0.5387507081031799, "learning_rate": 9.995564968350541e-05, "loss": 0.0642, "step": 5020 }, { "grad_norm": 0.4341564178466797, "learning_rate": 9.995477507676256e-05, "loss": 0.0631, "step": 5030 }, { "grad_norm": 0.39974406361579895, "learning_rate": 9.995389193411493e-05, "loss": 0.0671, "step": 5040 }, { "grad_norm": 0.4096246361732483, "learning_rate": 9.995300025571339e-05, "loss": 0.0774, "step": 5050 }, { "grad_norm": 0.48588827252388, "learning_rate": 9.995210004171034e-05, "loss": 0.0548, "step": 5060 }, { "grad_norm": 0.4246942698955536, "learning_rate": 9.995119129225956e-05, "loss": 0.0823, "step": 5070 }, { "grad_norm": 0.5857129096984863, "learning_rate": 9.995027400751637e-05, "loss": 0.0752, "step": 5080 }, { "grad_norm": 0.43735241889953613, "learning_rate": 9.994934818763751e-05, "loss": 0.0666, "step": 5090 }, { "grad_norm": 0.39751487970352173, "learning_rate": 9.994841383278115e-05, "loss": 0.0722, "step": 5100 }, { "grad_norm": 0.4003525376319885, "learning_rate": 9.994747094310695e-05, "loss": 0.0713, "step": 5110 }, { "grad_norm": 0.41466590762138367, "learning_rate": 9.994651951877604e-05, "loss": 0.083, "step": 5120 }, { "grad_norm": 0.5017175078392029, "learning_rate": 9.994555955995099e-05, "loss": 0.0874, "step": 5130 }, { "grad_norm": 0.4556835889816284, "learning_rate": 9.994459106679581e-05, "loss": 0.0728, "step": 5140 }, { "grad_norm": 0.46676796674728394, "learning_rate": 9.994361403947603e-05, "loss": 0.061, "step": 5150 }, { "grad_norm": 0.36432087421417236, "learning_rate": 9.994262847815854e-05, "loss": 0.0848, "step": 5160 }, { "grad_norm": 0.4110585153102875, "learning_rate": 9.99416343830118e-05, "loss": 0.0725, "step": 5170 }, { "grad_norm": 0.45471715927124023, "learning_rate": 9.994063175420565e-05, "loss": 0.0676, "step": 5180 }, { "grad_norm": 0.5486341714859009, "learning_rate": 9.99396205919114e-05, "loss": 0.0952, "step": 5190 }, { "grad_norm": 0.497489869594574, "learning_rate": 9.993860089630185e-05, "loss": 0.0577, "step": 5200 }, { "grad_norm": 0.41795822978019714, "learning_rate": 9.993757266755123e-05, "loss": 0.0734, "step": 5210 }, { "grad_norm": 0.45780766010284424, "learning_rate": 9.993653590583522e-05, "loss": 0.0759, "step": 5220 }, { "grad_norm": 0.4108163118362427, "learning_rate": 9.993549061133102e-05, "loss": 0.0657, "step": 5230 }, { "grad_norm": 0.5259591937065125, "learning_rate": 9.993443678421719e-05, "loss": 0.0738, "step": 5240 }, { "grad_norm": 0.4150567352771759, "learning_rate": 9.993337442467384e-05, "loss": 0.0837, "step": 5250 }, { "grad_norm": 0.500643253326416, "learning_rate": 9.993230353288248e-05, "loss": 0.0725, "step": 5260 }, { "grad_norm": 0.49974048137664795, "learning_rate": 9.993122410902608e-05, "loss": 0.0736, "step": 5270 }, { "grad_norm": 0.3610246479511261, "learning_rate": 9.993013615328912e-05, "loss": 0.0754, "step": 5280 }, { "grad_norm": 0.4070785343647003, "learning_rate": 9.992903966585747e-05, "loss": 0.0717, "step": 5290 }, { "grad_norm": 0.4465145468711853, "learning_rate": 9.992793464691852e-05, "loss": 0.058, "step": 5300 }, { "grad_norm": 0.4635448753833771, "learning_rate": 9.992682109666105e-05, "loss": 0.0558, "step": 5310 }, { "grad_norm": 0.52644282579422, "learning_rate": 9.992569901527538e-05, "loss": 0.0643, "step": 5320 }, { "grad_norm": 0.33734312653541565, "learning_rate": 9.99245684029532e-05, "loss": 0.0817, "step": 5330 }, { "grad_norm": 0.46451103687286377, "learning_rate": 9.992342925988774e-05, "loss": 0.0681, "step": 5340 }, { "grad_norm": 0.39653870463371277, "learning_rate": 9.992228158627361e-05, "loss": 0.0611, "step": 5350 }, { "grad_norm": 0.4331892728805542, "learning_rate": 9.992112538230693e-05, "loss": 0.0637, "step": 5360 }, { "grad_norm": 0.4936828911304474, "learning_rate": 9.991996064818527e-05, "loss": 0.0567, "step": 5370 }, { "grad_norm": 0.48331740498542786, "learning_rate": 9.991878738410768e-05, "loss": 0.063, "step": 5380 }, { "grad_norm": 0.38026171922683716, "learning_rate": 9.991760559027457e-05, "loss": 0.067, "step": 5390 }, { "grad_norm": 0.3678070306777954, "learning_rate": 9.991641526688793e-05, "loss": 0.0663, "step": 5400 }, { "grad_norm": 0.46542003750801086, "learning_rate": 9.991521641415113e-05, "loss": 0.0584, "step": 5410 }, { "grad_norm": 0.3434266746044159, "learning_rate": 9.991400903226904e-05, "loss": 0.0632, "step": 5420 }, { "grad_norm": 0.6439757347106934, "learning_rate": 9.991279312144794e-05, "loss": 0.0672, "step": 5430 }, { "grad_norm": 0.4004443883895874, "learning_rate": 9.991156868189564e-05, "loss": 0.0742, "step": 5440 }, { "grad_norm": 0.38030529022216797, "learning_rate": 9.991033571382131e-05, "loss": 0.0498, "step": 5450 }, { "grad_norm": 0.6089221835136414, "learning_rate": 9.990909421743569e-05, "loss": 0.064, "step": 5460 }, { "grad_norm": 0.5074700117111206, "learning_rate": 9.990784419295085e-05, "loss": 0.0595, "step": 5470 }, { "grad_norm": 0.42496299743652344, "learning_rate": 9.990658564058044e-05, "loss": 0.0671, "step": 5480 }, { "grad_norm": 0.3176672160625458, "learning_rate": 9.990531856053948e-05, "loss": 0.0575, "step": 5490 }, { "grad_norm": 0.43575534224510193, "learning_rate": 9.99040429530445e-05, "loss": 0.0658, "step": 5500 }, { "grad_norm": 0.34953397512435913, "learning_rate": 9.990275881831346e-05, "loss": 0.0655, "step": 5510 }, { "grad_norm": 0.4072035551071167, "learning_rate": 9.990146615656577e-05, "loss": 0.0609, "step": 5520 }, { "grad_norm": 0.3539828360080719, "learning_rate": 9.990016496802233e-05, "loss": 0.0614, "step": 5530 }, { "grad_norm": 0.4270588457584381, "learning_rate": 9.989885525290548e-05, "loss": 0.0533, "step": 5540 }, { "grad_norm": 0.31334108114242554, "learning_rate": 9.989753701143897e-05, "loss": 0.057, "step": 5550 }, { "grad_norm": 0.3486911654472351, "learning_rate": 9.989621024384812e-05, "loss": 0.0555, "step": 5560 }, { "grad_norm": 0.46441715955734253, "learning_rate": 9.989487495035959e-05, "loss": 0.056, "step": 5570 }, { "grad_norm": 0.43778228759765625, "learning_rate": 9.989353113120156e-05, "loss": 0.0691, "step": 5580 }, { "grad_norm": 0.4407021999359131, "learning_rate": 9.989217878660366e-05, "loss": 0.0611, "step": 5590 }, { "grad_norm": 0.41818732023239136, "learning_rate": 9.989081791679695e-05, "loss": 0.0681, "step": 5600 }, { "grad_norm": 0.426581472158432, "learning_rate": 9.988944852201397e-05, "loss": 0.0613, "step": 5610 }, { "grad_norm": 0.4369766116142273, "learning_rate": 9.988807060248873e-05, "loss": 0.0522, "step": 5620 }, { "grad_norm": 0.3242044448852539, "learning_rate": 9.988668415845665e-05, "loss": 0.0512, "step": 5630 }, { "grad_norm": 0.4033331274986267, "learning_rate": 9.988528919015466e-05, "loss": 0.0614, "step": 5640 }, { "grad_norm": 0.3740059435367584, "learning_rate": 9.988388569782112e-05, "loss": 0.0707, "step": 5650 }, { "grad_norm": 0.4235581159591675, "learning_rate": 9.988247368169583e-05, "loss": 0.0715, "step": 5660 }, { "grad_norm": 0.4755922853946686, "learning_rate": 9.988105314202007e-05, "loss": 0.0638, "step": 5670 }, { "grad_norm": 0.5229880213737488, "learning_rate": 9.987962407903659e-05, "loss": 0.0685, "step": 5680 }, { "grad_norm": 0.4417530298233032, "learning_rate": 9.987818649298957e-05, "loss": 0.0612, "step": 5690 }, { "grad_norm": 0.4175824224948883, "learning_rate": 9.987674038412465e-05, "loss": 0.0563, "step": 5700 }, { "grad_norm": 0.4663163423538208, "learning_rate": 9.987528575268891e-05, "loss": 0.0652, "step": 5710 }, { "grad_norm": 0.45519036054611206, "learning_rate": 9.987382259893095e-05, "loss": 0.0606, "step": 5720 }, { "grad_norm": 0.36827075481414795, "learning_rate": 9.987235092310074e-05, "loss": 0.0623, "step": 5730 }, { "grad_norm": 0.3840262293815613, "learning_rate": 9.987087072544978e-05, "loss": 0.0661, "step": 5740 }, { "grad_norm": 0.9068324565887451, "learning_rate": 9.9869382006231e-05, "loss": 0.0608, "step": 5750 }, { "grad_norm": 0.4124724566936493, "learning_rate": 9.986788476569875e-05, "loss": 0.0651, "step": 5760 }, { "grad_norm": 0.3530438542366028, "learning_rate": 9.986637900410887e-05, "loss": 0.0646, "step": 5770 }, { "grad_norm": 0.44121310114860535, "learning_rate": 9.986486472171869e-05, "loss": 0.0697, "step": 5780 }, { "grad_norm": 0.3260878920555115, "learning_rate": 9.986334191878692e-05, "loss": 0.0613, "step": 5790 }, { "grad_norm": 0.3432501554489136, "learning_rate": 9.986181059557378e-05, "loss": 0.0693, "step": 5800 }, { "grad_norm": 0.6470703482627869, "learning_rate": 9.986027075234094e-05, "loss": 0.0618, "step": 5810 }, { "grad_norm": 0.3856572210788727, "learning_rate": 9.985872238935152e-05, "loss": 0.0615, "step": 5820 }, { "grad_norm": 0.3654939532279968, "learning_rate": 9.985716550687008e-05, "loss": 0.0733, "step": 5830 }, { "grad_norm": 0.40768346190452576, "learning_rate": 9.985560010516264e-05, "loss": 0.0658, "step": 5840 }, { "grad_norm": 0.36548101902008057, "learning_rate": 9.985402618449668e-05, "loss": 0.0541, "step": 5850 }, { "grad_norm": 0.37310361862182617, "learning_rate": 9.985244374514118e-05, "loss": 0.0621, "step": 5860 }, { "grad_norm": 0.45410624146461487, "learning_rate": 9.985085278736651e-05, "loss": 0.0542, "step": 5870 }, { "grad_norm": 0.3614143133163452, "learning_rate": 9.984925331144452e-05, "loss": 0.0655, "step": 5880 }, { "grad_norm": 0.4253481328487396, "learning_rate": 9.984764531764851e-05, "loss": 0.0642, "step": 5890 }, { "grad_norm": 0.3187392055988312, "learning_rate": 9.984602880625326e-05, "loss": 0.0765, "step": 5900 }, { "grad_norm": 0.37191861867904663, "learning_rate": 9.9844403777535e-05, "loss": 0.0639, "step": 5910 }, { "grad_norm": 0.43022698163986206, "learning_rate": 9.984277023177135e-05, "loss": 0.0517, "step": 5920 }, { "grad_norm": 0.4515504240989685, "learning_rate": 9.984112816924148e-05, "loss": 0.055, "step": 5930 }, { "grad_norm": 0.42435938119888306, "learning_rate": 9.983947759022596e-05, "loss": 0.0639, "step": 5940 }, { "grad_norm": 0.3747391104698181, "learning_rate": 9.983781849500682e-05, "loss": 0.0563, "step": 5950 }, { "grad_norm": 0.41670462489128113, "learning_rate": 9.98361508838676e-05, "loss": 0.0535, "step": 5960 }, { "grad_norm": 0.4430261254310608, "learning_rate": 9.98344747570932e-05, "loss": 0.065, "step": 5970 }, { "grad_norm": 0.411371111869812, "learning_rate": 9.983279011497004e-05, "loss": 0.0586, "step": 5980 }, { "grad_norm": 0.3782680928707123, "learning_rate": 9.983109695778596e-05, "loss": 0.0737, "step": 5990 }, { "grad_norm": 0.40488001704216003, "learning_rate": 9.982939528583032e-05, "loss": 0.0831, "step": 6000 }, { "grad_norm": 0.41291069984436035, "learning_rate": 9.982768509939385e-05, "loss": 0.0621, "step": 6010 }, { "grad_norm": 0.3601927161216736, "learning_rate": 9.982596639876879e-05, "loss": 0.0571, "step": 6020 }, { "grad_norm": 0.40966159105300903, "learning_rate": 9.982423918424881e-05, "loss": 0.0615, "step": 6030 }, { "grad_norm": 0.43428510427474976, "learning_rate": 9.982250345612908e-05, "loss": 0.0589, "step": 6040 }, { "grad_norm": 0.33438003063201904, "learning_rate": 9.982075921470611e-05, "loss": 0.061, "step": 6050 }, { "grad_norm": 0.3507230579853058, "learning_rate": 9.981900646027802e-05, "loss": 0.0571, "step": 6060 }, { "grad_norm": 0.37596747279167175, "learning_rate": 9.981724519314425e-05, "loss": 0.0652, "step": 6070 }, { "grad_norm": 0.45413148403167725, "learning_rate": 9.981547541360581e-05, "loss": 0.063, "step": 6080 }, { "grad_norm": 0.6183632612228394, "learning_rate": 9.981369712196508e-05, "loss": 0.0638, "step": 6090 }, { "grad_norm": 0.3562270998954773, "learning_rate": 9.981191031852592e-05, "loss": 0.0744, "step": 6100 }, { "grad_norm": 0.4340735971927643, "learning_rate": 9.981011500359362e-05, "loss": 0.0523, "step": 6110 }, { "grad_norm": 0.3795546591281891, "learning_rate": 9.9808311177475e-05, "loss": 0.0581, "step": 6120 }, { "grad_norm": 0.30439892411231995, "learning_rate": 9.980649884047826e-05, "loss": 0.0491, "step": 6130 }, { "grad_norm": 0.3704807162284851, "learning_rate": 9.980467799291307e-05, "loss": 0.0586, "step": 6140 }, { "grad_norm": 0.3680281341075897, "learning_rate": 9.980284863509058e-05, "loss": 0.0818, "step": 6150 }, { "grad_norm": 0.4901278018951416, "learning_rate": 9.980101076732334e-05, "loss": 0.0778, "step": 6160 }, { "grad_norm": 0.40673866868019104, "learning_rate": 9.979916438992544e-05, "loss": 0.0658, "step": 6170 }, { "grad_norm": 0.3410937786102295, "learning_rate": 9.979730950321237e-05, "loss": 0.0566, "step": 6180 }, { "grad_norm": 0.37484195828437805, "learning_rate": 9.979544610750104e-05, "loss": 0.0701, "step": 6190 }, { "grad_norm": 0.34105631709098816, "learning_rate": 9.97935742031099e-05, "loss": 0.0556, "step": 6200 }, { "grad_norm": 0.45936524868011475, "learning_rate": 9.979169379035878e-05, "loss": 0.0657, "step": 6210 }, { "grad_norm": 0.4360050559043884, "learning_rate": 9.978980486956899e-05, "loss": 0.0604, "step": 6220 }, { "grad_norm": 0.5355028510093689, "learning_rate": 9.978790744106332e-05, "loss": 0.0601, "step": 6230 }, { "grad_norm": 0.363358736038208, "learning_rate": 9.978600150516594e-05, "loss": 0.0633, "step": 6240 }, { "grad_norm": 0.3425540328025818, "learning_rate": 9.978408706220259e-05, "loss": 0.0664, "step": 6250 }, { "grad_norm": 0.5144519209861755, "learning_rate": 9.978216411250032e-05, "loss": 0.0563, "step": 6260 }, { "grad_norm": 0.36563360691070557, "learning_rate": 9.978023265638778e-05, "loss": 0.0565, "step": 6270 }, { "grad_norm": 0.5762659311294556, "learning_rate": 9.977829269419495e-05, "loss": 0.0589, "step": 6280 }, { "grad_norm": 0.6267912983894348, "learning_rate": 9.977634422625335e-05, "loss": 0.0625, "step": 6290 }, { "grad_norm": 0.42200177907943726, "learning_rate": 9.97743872528959e-05, "loss": 0.0613, "step": 6300 }, { "grad_norm": 0.45263221859931946, "learning_rate": 9.9772421774457e-05, "loss": 0.0604, "step": 6310 }, { "grad_norm": 0.40229183435440063, "learning_rate": 9.977044779127252e-05, "loss": 0.057, "step": 6320 }, { "grad_norm": 0.39685896039009094, "learning_rate": 9.976846530367971e-05, "loss": 0.0544, "step": 6330 }, { "grad_norm": 0.4275648295879364, "learning_rate": 9.976647431201735e-05, "loss": 0.0738, "step": 6340 }, { "grad_norm": 0.30761396884918213, "learning_rate": 9.976447481662568e-05, "loss": 0.0684, "step": 6350 }, { "grad_norm": 0.5052717924118042, "learning_rate": 9.976246681784629e-05, "loss": 0.0679, "step": 6360 }, { "grad_norm": 0.38475117087364197, "learning_rate": 9.976045031602234e-05, "loss": 0.0601, "step": 6370 }, { "grad_norm": 0.49494898319244385, "learning_rate": 9.975842531149837e-05, "loss": 0.0619, "step": 6380 }, { "grad_norm": 0.41169247031211853, "learning_rate": 9.975639180462043e-05, "loss": 0.0712, "step": 6390 }, { "grad_norm": 0.3035963475704193, "learning_rate": 9.975434979573596e-05, "loss": 0.0639, "step": 6400 }, { "grad_norm": 0.3584347367286682, "learning_rate": 9.97522992851939e-05, "loss": 0.0523, "step": 6410 }, { "grad_norm": 0.3711678981781006, "learning_rate": 9.975024027334461e-05, "loss": 0.0663, "step": 6420 }, { "grad_norm": 0.660851776599884, "learning_rate": 9.974817276053993e-05, "loss": 0.0755, "step": 6430 }, { "grad_norm": 0.3393391966819763, "learning_rate": 9.974609674713315e-05, "loss": 0.0689, "step": 6440 }, { "grad_norm": 0.47390395402908325, "learning_rate": 9.9744012233479e-05, "loss": 0.0617, "step": 6450 }, { "grad_norm": 0.45689547061920166, "learning_rate": 9.974191921993366e-05, "loss": 0.0544, "step": 6460 }, { "grad_norm": 0.33881160616874695, "learning_rate": 9.973981770685474e-05, "loss": 0.0641, "step": 6470 }, { "grad_norm": 0.7075117826461792, "learning_rate": 9.97377076946014e-05, "loss": 0.0627, "step": 6480 }, { "grad_norm": 0.46797531843185425, "learning_rate": 9.973558918353412e-05, "loss": 0.0723, "step": 6490 }, { "grad_norm": 0.34433865547180176, "learning_rate": 9.973346217401494e-05, "loss": 0.0538, "step": 6500 }, { "grad_norm": 0.36045756936073303, "learning_rate": 9.973132666640726e-05, "loss": 0.0507, "step": 6510 }, { "grad_norm": 0.40544062852859497, "learning_rate": 9.972918266107602e-05, "loss": 0.0567, "step": 6520 }, { "grad_norm": 0.4313898980617523, "learning_rate": 9.972703015838756e-05, "loss": 0.0567, "step": 6530 }, { "grad_norm": 0.39562171697616577, "learning_rate": 9.97248691587097e-05, "loss": 0.064, "step": 6540 }, { "grad_norm": 0.4036118984222412, "learning_rate": 9.972269966241166e-05, "loss": 0.0587, "step": 6550 }, { "grad_norm": 0.41442418098449707, "learning_rate": 9.972052166986417e-05, "loss": 0.0589, "step": 6560 }, { "grad_norm": 0.4585121273994446, "learning_rate": 9.971833518143938e-05, "loss": 0.0702, "step": 6570 }, { "grad_norm": 0.3663182258605957, "learning_rate": 9.971614019751093e-05, "loss": 0.05, "step": 6580 }, { "grad_norm": 0.3913862109184265, "learning_rate": 9.971393671845383e-05, "loss": 0.0709, "step": 6590 }, { "grad_norm": 0.3846462368965149, "learning_rate": 9.971172474464464e-05, "loss": 0.0683, "step": 6600 }, { "grad_norm": 0.30814582109451294, "learning_rate": 9.97095042764613e-05, "loss": 0.0518, "step": 6610 }, { "grad_norm": 0.3343302011489868, "learning_rate": 9.970727531428324e-05, "loss": 0.0493, "step": 6620 }, { "grad_norm": 0.3831178843975067, "learning_rate": 9.970503785849132e-05, "loss": 0.0556, "step": 6630 }, { "grad_norm": 0.34204500913619995, "learning_rate": 9.970279190946788e-05, "loss": 0.0505, "step": 6640 }, { "grad_norm": 0.30864885449409485, "learning_rate": 9.970053746759667e-05, "loss": 0.0661, "step": 6650 }, { "grad_norm": 0.44034796953201294, "learning_rate": 9.969827453326292e-05, "loss": 0.052, "step": 6660 }, { "grad_norm": 0.38088953495025635, "learning_rate": 9.969600310685332e-05, "loss": 0.0533, "step": 6670 }, { "grad_norm": 0.3854491412639618, "learning_rate": 9.969372318875596e-05, "loss": 0.0594, "step": 6680 }, { "grad_norm": 0.3009752631187439, "learning_rate": 9.969143477936043e-05, "loss": 0.0606, "step": 6690 }, { "grad_norm": 0.4999152421951294, "learning_rate": 9.968913787905775e-05, "loss": 0.0602, "step": 6700 }, { "grad_norm": 0.4571070671081543, "learning_rate": 9.968683248824045e-05, "loss": 0.0625, "step": 6710 }, { "grad_norm": 0.4045286476612091, "learning_rate": 9.968451860730238e-05, "loss": 0.0706, "step": 6720 }, { "grad_norm": 0.36216792464256287, "learning_rate": 9.968219623663896e-05, "loss": 0.0605, "step": 6730 }, { "grad_norm": 0.4314101040363312, "learning_rate": 9.967986537664702e-05, "loss": 0.0763, "step": 6740 }, { "grad_norm": 0.5378241539001465, "learning_rate": 9.967752602772483e-05, "loss": 0.0557, "step": 6750 }, { "grad_norm": 0.3073936104774475, "learning_rate": 9.967517819027212e-05, "loss": 0.0555, "step": 6760 }, { "grad_norm": 0.5039275884628296, "learning_rate": 9.967282186469009e-05, "loss": 0.0535, "step": 6770 }, { "grad_norm": 0.340315580368042, "learning_rate": 9.967045705138135e-05, "loss": 0.0593, "step": 6780 }, { "grad_norm": 0.2814526855945587, "learning_rate": 9.966808375074998e-05, "loss": 0.0528, "step": 6790 }, { "grad_norm": 0.38120323419570923, "learning_rate": 9.966570196320154e-05, "loss": 0.0614, "step": 6800 }, { "grad_norm": 0.3242633640766144, "learning_rate": 9.966331168914299e-05, "loss": 0.0457, "step": 6810 }, { "grad_norm": 0.3372310996055603, "learning_rate": 9.966091292898277e-05, "loss": 0.0557, "step": 6820 }, { "grad_norm": 0.4032164216041565, "learning_rate": 9.965850568313076e-05, "loss": 0.0556, "step": 6830 }, { "grad_norm": 0.3934151828289032, "learning_rate": 9.965608995199827e-05, "loss": 0.0562, "step": 6840 }, { "grad_norm": 0.357907772064209, "learning_rate": 9.965366573599812e-05, "loss": 0.0561, "step": 6850 }, { "grad_norm": 0.42349350452423096, "learning_rate": 9.965123303554453e-05, "loss": 0.055, "step": 6860 }, { "grad_norm": 0.46693071722984314, "learning_rate": 9.964879185105317e-05, "loss": 0.0703, "step": 6870 }, { "grad_norm": 0.3569488227367401, "learning_rate": 9.964634218294119e-05, "loss": 0.0664, "step": 6880 }, { "grad_norm": 0.4093628525733948, "learning_rate": 9.964388403162714e-05, "loss": 0.0629, "step": 6890 }, { "grad_norm": 0.5046935677528381, "learning_rate": 9.96414173975311e-05, "loss": 0.0676, "step": 6900 }, { "grad_norm": 0.3853631913661957, "learning_rate": 9.963894228107451e-05, "loss": 0.0536, "step": 6910 }, { "grad_norm": 0.3939613401889801, "learning_rate": 9.963645868268032e-05, "loss": 0.0596, "step": 6920 }, { "grad_norm": 0.27533507347106934, "learning_rate": 9.963396660277289e-05, "loss": 0.0603, "step": 6930 }, { "grad_norm": 0.42146772146224976, "learning_rate": 9.963146604177807e-05, "loss": 0.0469, "step": 6940 }, { "grad_norm": 0.34088581800460815, "learning_rate": 9.962895700012311e-05, "loss": 0.0513, "step": 6950 }, { "grad_norm": 0.3640419840812683, "learning_rate": 9.962643947823677e-05, "loss": 0.0513, "step": 6960 }, { "grad_norm": 0.29886922240257263, "learning_rate": 9.962391347654921e-05, "loss": 0.0601, "step": 6970 }, { "grad_norm": 0.3659098148345947, "learning_rate": 9.962137899549204e-05, "loss": 0.0626, "step": 6980 }, { "grad_norm": 0.6428754925727844, "learning_rate": 9.961883603549835e-05, "loss": 0.0517, "step": 6990 }, { "grad_norm": 0.40283074975013733, "learning_rate": 9.961628459700267e-05, "loss": 0.0577, "step": 7000 }, { "grad_norm": 0.4167563021183014, "learning_rate": 9.961372468044095e-05, "loss": 0.0622, "step": 7010 }, { "grad_norm": 0.4557013511657715, "learning_rate": 9.961115628625062e-05, "loss": 0.0579, "step": 7020 }, { "grad_norm": 0.3862314820289612, "learning_rate": 9.960857941487056e-05, "loss": 0.0683, "step": 7030 }, { "grad_norm": 0.3550468981266022, "learning_rate": 9.960599406674106e-05, "loss": 0.053, "step": 7040 }, { "grad_norm": 0.4012375473976135, "learning_rate": 9.960340024230393e-05, "loss": 0.0493, "step": 7050 }, { "grad_norm": 0.4241991341114044, "learning_rate": 9.960079794200232e-05, "loss": 0.0542, "step": 7060 }, { "grad_norm": 0.41916361451148987, "learning_rate": 9.959818716628096e-05, "loss": 0.056, "step": 7070 }, { "grad_norm": 0.3261587619781494, "learning_rate": 9.95955679155859e-05, "loss": 0.0729, "step": 7080 }, { "grad_norm": 0.5133737325668335, "learning_rate": 9.959294019036472e-05, "loss": 0.0627, "step": 7090 }, { "grad_norm": 0.42227309942245483, "learning_rate": 9.959030399106646e-05, "loss": 0.056, "step": 7100 }, { "grad_norm": 0.4281708002090454, "learning_rate": 9.958765931814153e-05, "loss": 0.0637, "step": 7110 }, { "grad_norm": 0.3313122093677521, "learning_rate": 9.958500617204184e-05, "loss": 0.0446, "step": 7120 }, { "grad_norm": 0.3978215456008911, "learning_rate": 9.958234455322075e-05, "loss": 0.0539, "step": 7130 }, { "grad_norm": 0.45235756039619446, "learning_rate": 9.957967446213308e-05, "loss": 0.0682, "step": 7140 }, { "grad_norm": 0.4527306854724884, "learning_rate": 9.957699589923501e-05, "loss": 0.0675, "step": 7150 }, { "grad_norm": 0.537695050239563, "learning_rate": 9.957430886498431e-05, "loss": 0.0643, "step": 7160 }, { "grad_norm": 0.3796778619289398, "learning_rate": 9.957161335984008e-05, "loss": 0.0614, "step": 7170 }, { "grad_norm": 0.4141479432582855, "learning_rate": 9.956890938426291e-05, "loss": 0.0585, "step": 7180 }, { "grad_norm": 0.3226820230484009, "learning_rate": 9.956619693871482e-05, "loss": 0.0598, "step": 7190 }, { "grad_norm": 0.43006035685539246, "learning_rate": 9.956347602365934e-05, "loss": 0.0605, "step": 7200 }, { "grad_norm": 0.3233400583267212, "learning_rate": 9.956074663956135e-05, "loss": 0.0615, "step": 7210 }, { "grad_norm": 0.3821176588535309, "learning_rate": 9.955800878688726e-05, "loss": 0.0581, "step": 7220 }, { "grad_norm": 0.3703433573246002, "learning_rate": 9.955526246610489e-05, "loss": 0.0457, "step": 7230 }, { "grad_norm": 0.516550600528717, "learning_rate": 9.955250767768349e-05, "loss": 0.0574, "step": 7240 }, { "grad_norm": 0.34358271956443787, "learning_rate": 9.95497444220938e-05, "loss": 0.0591, "step": 7250 }, { "grad_norm": 0.37156131863594055, "learning_rate": 9.954697269980797e-05, "loss": 0.046, "step": 7260 }, { "grad_norm": 0.4163942337036133, "learning_rate": 9.954419251129962e-05, "loss": 0.0486, "step": 7270 }, { "grad_norm": 0.48533928394317627, "learning_rate": 9.95414038570438e-05, "loss": 0.0541, "step": 7280 }, { "grad_norm": 0.30100107192993164, "learning_rate": 9.953860673751703e-05, "loss": 0.0548, "step": 7290 }, { "grad_norm": 0.5212969779968262, "learning_rate": 9.953580115319725e-05, "loss": 0.0514, "step": 7300 }, { "grad_norm": 0.2951434552669525, "learning_rate": 9.953298710456387e-05, "loss": 0.0581, "step": 7310 }, { "grad_norm": 0.3975193202495575, "learning_rate": 9.953016459209771e-05, "loss": 0.0502, "step": 7320 }, { "grad_norm": 0.3470879793167114, "learning_rate": 9.952733361628108e-05, "loss": 0.0571, "step": 7330 }, { "grad_norm": 0.32232871651649475, "learning_rate": 9.952449417759772e-05, "loss": 0.0588, "step": 7340 }, { "grad_norm": 0.41001492738723755, "learning_rate": 9.952164627653279e-05, "loss": 0.0479, "step": 7350 }, { "grad_norm": 0.44576072692871094, "learning_rate": 9.951878991357292e-05, "loss": 0.0501, "step": 7360 }, { "grad_norm": 0.32141295075416565, "learning_rate": 9.951592508920622e-05, "loss": 0.0582, "step": 7370 }, { "grad_norm": 0.41778889298439026, "learning_rate": 9.951305180392219e-05, "loss": 0.0484, "step": 7380 }, { "grad_norm": 0.3311668038368225, "learning_rate": 9.951017005821178e-05, "loss": 0.0493, "step": 7390 }, { "grad_norm": 0.34922900795936584, "learning_rate": 9.95072798525674e-05, "loss": 0.0545, "step": 7400 }, { "grad_norm": 0.32163533568382263, "learning_rate": 9.950438118748293e-05, "loss": 0.0504, "step": 7410 }, { "grad_norm": 0.3569711446762085, "learning_rate": 9.950147406345366e-05, "loss": 0.0597, "step": 7420 }, { "grad_norm": 0.31682804226875305, "learning_rate": 9.949855848097635e-05, "loss": 0.0528, "step": 7430 }, { "grad_norm": 0.3436415493488312, "learning_rate": 9.949563444054916e-05, "loss": 0.0538, "step": 7440 }, { "grad_norm": 0.31496381759643555, "learning_rate": 9.949270194267178e-05, "loss": 0.0492, "step": 7450 }, { "grad_norm": 0.30810463428497314, "learning_rate": 9.948976098784526e-05, "loss": 0.0529, "step": 7460 }, { "grad_norm": 0.337283730506897, "learning_rate": 9.948681157657213e-05, "loss": 0.0544, "step": 7470 }, { "grad_norm": 0.41609182953834534, "learning_rate": 9.948385370935638e-05, "loss": 0.0651, "step": 7480 }, { "grad_norm": 0.4429175853729248, "learning_rate": 9.94808873867034e-05, "loss": 0.0549, "step": 7490 }, { "grad_norm": 0.35776376724243164, "learning_rate": 9.947791260912009e-05, "loss": 0.052, "step": 7500 }, { "grad_norm": 0.3798717260360718, "learning_rate": 9.947492937711474e-05, "loss": 0.0479, "step": 7510 }, { "grad_norm": 0.34027099609375, "learning_rate": 9.947193769119707e-05, "loss": 0.0449, "step": 7520 }, { "grad_norm": 0.33751416206359863, "learning_rate": 9.946893755187834e-05, "loss": 0.0594, "step": 7530 }, { "grad_norm": 0.29064056277275085, "learning_rate": 9.946592895967115e-05, "loss": 0.0545, "step": 7540 }, { "grad_norm": 0.29808276891708374, "learning_rate": 9.94629119150896e-05, "loss": 0.0487, "step": 7550 }, { "grad_norm": 0.38113898038864136, "learning_rate": 9.94598864186492e-05, "loss": 0.0524, "step": 7560 }, { "grad_norm": 0.48394641280174255, "learning_rate": 9.945685247086696e-05, "loss": 0.0631, "step": 7570 }, { "grad_norm": 0.309293270111084, "learning_rate": 9.945381007226129e-05, "loss": 0.0532, "step": 7580 }, { "grad_norm": 0.3380117118358612, "learning_rate": 9.945075922335203e-05, "loss": 0.0572, "step": 7590 }, { "grad_norm": 0.3213367164134979, "learning_rate": 9.944769992466049e-05, "loss": 0.0666, "step": 7600 }, { "grad_norm": 0.33812180161476135, "learning_rate": 9.944463217670945e-05, "loss": 0.0662, "step": 7610 }, { "grad_norm": 0.3283236622810364, "learning_rate": 9.944155598002307e-05, "loss": 0.063, "step": 7620 }, { "grad_norm": 0.3792947828769684, "learning_rate": 9.943847133512701e-05, "loss": 0.0607, "step": 7630 }, { "grad_norm": 0.43910494446754456, "learning_rate": 9.943537824254834e-05, "loss": 0.053, "step": 7640 }, { "grad_norm": 0.4166377782821655, "learning_rate": 9.943227670281559e-05, "loss": 0.0553, "step": 7650 }, { "grad_norm": 0.3308382034301758, "learning_rate": 9.942916671645873e-05, "loss": 0.0548, "step": 7660 }, { "grad_norm": 0.4193284809589386, "learning_rate": 9.942604828400916e-05, "loss": 0.0595, "step": 7670 }, { "grad_norm": 0.3331829309463501, "learning_rate": 9.942292140599975e-05, "loss": 0.0595, "step": 7680 }, { "grad_norm": 0.4003916084766388, "learning_rate": 9.94197860829648e-05, "loss": 0.0498, "step": 7690 }, { "grad_norm": 0.30441296100616455, "learning_rate": 9.941664231544004e-05, "loss": 0.0527, "step": 7700 }, { "grad_norm": 0.35167360305786133, "learning_rate": 9.941349010396264e-05, "loss": 0.0622, "step": 7710 }, { "grad_norm": 0.2878600060939789, "learning_rate": 9.941032944907125e-05, "loss": 0.0564, "step": 7720 }, { "grad_norm": 0.43430641293525696, "learning_rate": 9.940716035130596e-05, "loss": 0.0571, "step": 7730 }, { "grad_norm": 0.32715684175491333, "learning_rate": 9.940398281120821e-05, "loss": 0.0593, "step": 7740 }, { "grad_norm": 0.39518988132476807, "learning_rate": 9.940079682932102e-05, "loss": 0.048, "step": 7750 }, { "grad_norm": 0.37505313754081726, "learning_rate": 9.939760240618877e-05, "loss": 0.0538, "step": 7760 }, { "grad_norm": 0.39559951424598694, "learning_rate": 9.939439954235729e-05, "loss": 0.0579, "step": 7770 }, { "grad_norm": 0.30762413144111633, "learning_rate": 9.939118823837387e-05, "loss": 0.056, "step": 7780 }, { "grad_norm": 0.37749701738357544, "learning_rate": 9.938796849478725e-05, "loss": 0.0544, "step": 7790 }, { "grad_norm": 0.346614807844162, "learning_rate": 9.938474031214755e-05, "loss": 0.0527, "step": 7800 }, { "grad_norm": 0.3608200252056122, "learning_rate": 9.938150369100643e-05, "loss": 0.0507, "step": 7810 }, { "grad_norm": 0.4236980080604553, "learning_rate": 9.93782586319169e-05, "loss": 0.0467, "step": 7820 }, { "grad_norm": 0.3265879154205322, "learning_rate": 9.937500513543348e-05, "loss": 0.0496, "step": 7830 }, { "grad_norm": 0.3966817557811737, "learning_rate": 9.937174320211207e-05, "loss": 0.0469, "step": 7840 }, { "grad_norm": 0.5169612765312195, "learning_rate": 9.936847283251009e-05, "loss": 0.0549, "step": 7850 }, { "grad_norm": 0.5031573176383972, "learning_rate": 9.936519402718632e-05, "loss": 0.0592, "step": 7860 }, { "grad_norm": 0.45779916644096375, "learning_rate": 9.936190678670102e-05, "loss": 0.0584, "step": 7870 }, { "grad_norm": 0.38526663184165955, "learning_rate": 9.935861111161593e-05, "loss": 0.055, "step": 7880 }, { "grad_norm": 0.32685694098472595, "learning_rate": 9.935530700249416e-05, "loss": 0.0546, "step": 7890 }, { "grad_norm": 0.37867674231529236, "learning_rate": 9.935199445990028e-05, "loss": 0.0692, "step": 7900 }, { "grad_norm": 0.28184428811073303, "learning_rate": 9.934867348440033e-05, "loss": 0.061, "step": 7910 }, { "grad_norm": 0.3944634795188904, "learning_rate": 9.934534407656176e-05, "loss": 0.0755, "step": 7920 }, { "grad_norm": 0.382587194442749, "learning_rate": 9.93420062369535e-05, "loss": 0.0464, "step": 7930 }, { "grad_norm": 0.40115851163864136, "learning_rate": 9.933865996614589e-05, "loss": 0.0594, "step": 7940 }, { "grad_norm": 0.511740505695343, "learning_rate": 9.933530526471068e-05, "loss": 0.0548, "step": 7950 }, { "grad_norm": 0.3758091628551483, "learning_rate": 9.933194213322114e-05, "loss": 0.0491, "step": 7960 }, { "grad_norm": 0.3679521977901459, "learning_rate": 9.932857057225192e-05, "loss": 0.0521, "step": 7970 }, { "grad_norm": 0.4204922914505005, "learning_rate": 9.932519058237912e-05, "loss": 0.0534, "step": 7980 }, { "grad_norm": 0.3648950457572937, "learning_rate": 9.932180216418032e-05, "loss": 0.0509, "step": 7990 }, { "grad_norm": 0.3513130247592926, "learning_rate": 9.931840531823446e-05, "loss": 0.0451, "step": 8000 }, { "grad_norm": 0.3746764659881592, "learning_rate": 9.9315000045122e-05, "loss": 0.0437, "step": 8010 }, { "grad_norm": 0.39758339524269104, "learning_rate": 9.931158634542481e-05, "loss": 0.0511, "step": 8020 }, { "grad_norm": 0.328840047121048, "learning_rate": 9.930816421972617e-05, "loss": 0.0509, "step": 8030 }, { "grad_norm": 0.2631353437900543, "learning_rate": 9.930473366861086e-05, "loss": 0.0432, "step": 8040 }, { "grad_norm": 0.2785874009132385, "learning_rate": 9.930129469266505e-05, "loss": 0.0522, "step": 8050 }, { "grad_norm": 0.2696782350540161, "learning_rate": 9.929784729247638e-05, "loss": 0.0475, "step": 8060 }, { "grad_norm": 0.4547630846500397, "learning_rate": 9.929439146863389e-05, "loss": 0.048, "step": 8070 }, { "grad_norm": 0.43519312143325806, "learning_rate": 9.92909272217281e-05, "loss": 0.0619, "step": 8080 }, { "grad_norm": 0.33846089243888855, "learning_rate": 9.928745455235097e-05, "loss": 0.0624, "step": 8090 }, { "grad_norm": 0.29307299852371216, "learning_rate": 9.928397346109588e-05, "loss": 0.0573, "step": 8100 }, { "grad_norm": 0.37035617232322693, "learning_rate": 9.928048394855762e-05, "loss": 0.0593, "step": 8110 }, { "grad_norm": 0.42385563254356384, "learning_rate": 9.92769860153325e-05, "loss": 0.0504, "step": 8120 }, { "grad_norm": 0.40942370891571045, "learning_rate": 9.927347966201819e-05, "loss": 0.0718, "step": 8130 }, { "grad_norm": 0.36830365657806396, "learning_rate": 9.926996488921383e-05, "loss": 0.0618, "step": 8140 }, { "grad_norm": 0.3621976673603058, "learning_rate": 9.926644169752001e-05, "loss": 0.0539, "step": 8150 }, { "grad_norm": 0.37519371509552, "learning_rate": 9.926291008753875e-05, "loss": 0.0579, "step": 8160 }, { "grad_norm": 0.3033985197544098, "learning_rate": 9.92593700598735e-05, "loss": 0.0561, "step": 8170 }, { "grad_norm": 0.33790746331214905, "learning_rate": 9.925582161512915e-05, "loss": 0.0479, "step": 8180 }, { "grad_norm": 0.4538002014160156, "learning_rate": 9.925226475391205e-05, "loss": 0.056, "step": 8190 }, { "grad_norm": 0.4379037916660309, "learning_rate": 9.924869947682993e-05, "loss": 0.0496, "step": 8200 }, { "grad_norm": 0.27318406105041504, "learning_rate": 9.924512578449204e-05, "loss": 0.0419, "step": 8210 }, { "grad_norm": 0.387322336435318, "learning_rate": 9.924154367750901e-05, "loss": 0.0509, "step": 8220 }, { "grad_norm": 0.4076681435108185, "learning_rate": 9.923795315649293e-05, "loss": 0.0507, "step": 8230 }, { "grad_norm": 0.30978259444236755, "learning_rate": 9.92343542220573e-05, "loss": 0.0516, "step": 8240 }, { "grad_norm": 0.3714648485183716, "learning_rate": 9.92307468748171e-05, "loss": 0.0549, "step": 8250 }, { "grad_norm": 0.38031303882598877, "learning_rate": 9.922713111538873e-05, "loss": 0.0445, "step": 8260 }, { "grad_norm": 0.39603567123413086, "learning_rate": 9.922350694439003e-05, "loss": 0.0475, "step": 8270 }, { "grad_norm": 0.32475996017456055, "learning_rate": 9.921987436244024e-05, "loss": 0.0613, "step": 8280 }, { "grad_norm": 0.3620864450931549, "learning_rate": 9.921623337016008e-05, "loss": 0.0496, "step": 8290 }, { "grad_norm": 0.27994754910469055, "learning_rate": 9.921258396817172e-05, "loss": 0.0555, "step": 8300 }, { "grad_norm": 0.3251165747642517, "learning_rate": 9.920892615709874e-05, "loss": 0.046, "step": 8310 }, { "grad_norm": 0.41662904620170593, "learning_rate": 9.920525993756612e-05, "loss": 0.0433, "step": 8320 }, { "grad_norm": 0.3321804702281952, "learning_rate": 9.920158531020036e-05, "loss": 0.0488, "step": 8330 }, { "grad_norm": 0.3583967685699463, "learning_rate": 9.919790227562933e-05, "loss": 0.0489, "step": 8340 }, { "grad_norm": 0.3993895649909973, "learning_rate": 9.919421083448237e-05, "loss": 0.0515, "step": 8350 }, { "grad_norm": 0.3354372978210449, "learning_rate": 9.919051098739022e-05, "loss": 0.0443, "step": 8360 }, { "grad_norm": 0.3037347197532654, "learning_rate": 9.918680273498514e-05, "loss": 0.0495, "step": 8370 }, { "grad_norm": 0.30810463428497314, "learning_rate": 9.918308607790072e-05, "loss": 0.0554, "step": 8380 }, { "grad_norm": 0.4334182143211365, "learning_rate": 9.917936101677205e-05, "loss": 0.048, "step": 8390 }, { "grad_norm": 0.3910318911075592, "learning_rate": 9.917562755223564e-05, "loss": 0.0493, "step": 8400 }, { "grad_norm": 0.4079945981502533, "learning_rate": 9.917188568492944e-05, "loss": 0.0534, "step": 8410 }, { "grad_norm": 0.3748576045036316, "learning_rate": 9.916813541549283e-05, "loss": 0.049, "step": 8420 }, { "grad_norm": 0.4566931128501892, "learning_rate": 9.916437674456663e-05, "loss": 0.058, "step": 8430 }, { "grad_norm": 0.34162095189094543, "learning_rate": 9.916060967279308e-05, "loss": 0.053, "step": 8440 }, { "grad_norm": 0.217827707529068, "learning_rate": 9.91568342008159e-05, "loss": 0.0492, "step": 8450 }, { "grad_norm": 0.4149891436100006, "learning_rate": 9.915305032928019e-05, "loss": 0.0515, "step": 8460 }, { "grad_norm": 0.31543320417404175, "learning_rate": 9.914925805883253e-05, "loss": 0.043, "step": 8470 }, { "grad_norm": 0.4513542950153351, "learning_rate": 9.914545739012088e-05, "loss": 0.0475, "step": 8480 }, { "grad_norm": 0.33068451285362244, "learning_rate": 9.91416483237947e-05, "loss": 0.0514, "step": 8490 }, { "grad_norm": 0.47141632437705994, "learning_rate": 9.913783086050485e-05, "loss": 0.0577, "step": 8500 }, { "grad_norm": 0.4015023112297058, "learning_rate": 9.913400500090364e-05, "loss": 0.0477, "step": 8510 }, { "grad_norm": 0.3929557800292969, "learning_rate": 9.913017074564479e-05, "loss": 0.0414, "step": 8520 }, { "grad_norm": 0.3447268307209015, "learning_rate": 9.912632809538348e-05, "loss": 0.0562, "step": 8530 }, { "grad_norm": 0.46634602546691895, "learning_rate": 9.912247705077629e-05, "loss": 0.0539, "step": 8540 }, { "grad_norm": 0.3327675461769104, "learning_rate": 9.911861761248127e-05, "loss": 0.0546, "step": 8550 }, { "grad_norm": 0.26840031147003174, "learning_rate": 9.91147497811579e-05, "loss": 0.0529, "step": 8560 }, { "grad_norm": 0.31714120507240295, "learning_rate": 9.911087355746709e-05, "loss": 0.0754, "step": 8570 }, { "grad_norm": 0.3364338278770447, "learning_rate": 9.910698894207117e-05, "loss": 0.0615, "step": 8580 }, { "grad_norm": 0.3781903386116028, "learning_rate": 9.910309593563392e-05, "loss": 0.0533, "step": 8590 }, { "grad_norm": 0.37420734763145447, "learning_rate": 9.909919453882057e-05, "loss": 0.0367, "step": 8600 }, { "grad_norm": 0.40407419204711914, "learning_rate": 9.90952847522977e-05, "loss": 0.0445, "step": 8610 }, { "grad_norm": 0.38200896978378296, "learning_rate": 9.909136657673346e-05, "loss": 0.0496, "step": 8620 }, { "grad_norm": 0.33564770221710205, "learning_rate": 9.908744001279731e-05, "loss": 0.0453, "step": 8630 }, { "grad_norm": 0.3864067792892456, "learning_rate": 9.90835050611602e-05, "loss": 0.0538, "step": 8640 }, { "grad_norm": 0.33519962430000305, "learning_rate": 9.90795617224945e-05, "loss": 0.0609, "step": 8650 }, { "grad_norm": 0.3909360468387604, "learning_rate": 9.907560999747405e-05, "loss": 0.0495, "step": 8660 }, { "grad_norm": 0.3684551417827606, "learning_rate": 9.907164988677408e-05, "loss": 0.0464, "step": 8670 }, { "grad_norm": 0.36991599202156067, "learning_rate": 9.906768139107124e-05, "loss": 0.0486, "step": 8680 }, { "grad_norm": 0.33217257261276245, "learning_rate": 9.906370451104367e-05, "loss": 0.0507, "step": 8690 }, { "grad_norm": 0.4369140863418579, "learning_rate": 9.905971924737088e-05, "loss": 0.0528, "step": 8700 }, { "grad_norm": 0.29045313596725464, "learning_rate": 9.905572560073387e-05, "loss": 0.0592, "step": 8710 }, { "grad_norm": 0.3925102949142456, "learning_rate": 9.905172357181501e-05, "loss": 0.0632, "step": 8720 }, { "grad_norm": 0.308178186416626, "learning_rate": 9.904771316129817e-05, "loss": 0.0482, "step": 8730 }, { "grad_norm": 0.4043664336204529, "learning_rate": 9.904369436986862e-05, "loss": 0.0436, "step": 8740 }, { "grad_norm": 0.2915383577346802, "learning_rate": 9.903966719821303e-05, "loss": 0.0581, "step": 8750 }, { "grad_norm": 0.38067427277565, "learning_rate": 9.903563164701956e-05, "loss": 0.0485, "step": 8760 }, { "grad_norm": 0.38881370425224304, "learning_rate": 9.903158771697778e-05, "loss": 0.0568, "step": 8770 }, { "grad_norm": 0.3935610353946686, "learning_rate": 9.902753540877867e-05, "loss": 0.0587, "step": 8780 }, { "grad_norm": 0.3169468641281128, "learning_rate": 9.902347472311466e-05, "loss": 0.0502, "step": 8790 }, { "grad_norm": 0.35673362016677856, "learning_rate": 9.901940566067962e-05, "loss": 0.0461, "step": 8800 }, { "grad_norm": 0.6177525520324707, "learning_rate": 9.901532822216883e-05, "loss": 0.0527, "step": 8810 }, { "grad_norm": 0.47098612785339355, "learning_rate": 9.901124240827904e-05, "loss": 0.0521, "step": 8820 }, { "grad_norm": 0.584746241569519, "learning_rate": 9.900714821970835e-05, "loss": 0.0568, "step": 8830 }, { "grad_norm": 0.5027649998664856, "learning_rate": 9.900304565715641e-05, "loss": 0.0565, "step": 8840 }, { "grad_norm": 0.33472687005996704, "learning_rate": 9.899893472132419e-05, "loss": 0.0499, "step": 8850 }, { "grad_norm": 0.28795894980430603, "learning_rate": 9.899481541291415e-05, "loss": 0.0512, "step": 8860 }, { "grad_norm": 0.4284873604774475, "learning_rate": 9.899068773263016e-05, "loss": 0.0549, "step": 8870 }, { "grad_norm": 0.3568730056285858, "learning_rate": 9.898655168117754e-05, "loss": 0.0514, "step": 8880 }, { "grad_norm": 0.330951988697052, "learning_rate": 9.898240725926302e-05, "loss": 0.0534, "step": 8890 }, { "grad_norm": 0.32185986638069153, "learning_rate": 9.897825446759478e-05, "loss": 0.0567, "step": 8900 }, { "grad_norm": 0.3010598123073578, "learning_rate": 9.897409330688241e-05, "loss": 0.0498, "step": 8910 }, { "grad_norm": 0.4450047016143799, "learning_rate": 9.896992377783692e-05, "loss": 0.06, "step": 8920 }, { "grad_norm": 0.6578584313392639, "learning_rate": 9.89657458811708e-05, "loss": 0.0512, "step": 8930 }, { "grad_norm": 0.4378317892551422, "learning_rate": 9.896155961759792e-05, "loss": 0.0583, "step": 8940 }, { "grad_norm": 0.43145400285720825, "learning_rate": 9.895736498783361e-05, "loss": 0.0542, "step": 8950 }, { "grad_norm": 0.3364643156528473, "learning_rate": 9.895316199259462e-05, "loss": 0.0427, "step": 8960 }, { "grad_norm": 0.35319679975509644, "learning_rate": 9.894895063259909e-05, "loss": 0.0438, "step": 8970 }, { "grad_norm": 0.32563722133636475, "learning_rate": 9.894473090856667e-05, "loss": 0.0489, "step": 8980 }, { "grad_norm": 0.3075736463069916, "learning_rate": 9.894050282121839e-05, "loss": 0.0687, "step": 8990 }, { "grad_norm": 0.28349772095680237, "learning_rate": 9.893626637127668e-05, "loss": 0.0496, "step": 9000 }, { "grad_norm": 0.3078306019306183, "learning_rate": 9.893202155946546e-05, "loss": 0.043, "step": 9010 }, { "grad_norm": 0.4170679748058319, "learning_rate": 9.892776838651006e-05, "loss": 0.0458, "step": 9020 }, { "grad_norm": 0.3586791455745697, "learning_rate": 9.892350685313722e-05, "loss": 0.0545, "step": 9030 }, { "grad_norm": 0.24248795211315155, "learning_rate": 9.891923696007513e-05, "loss": 0.0561, "step": 9040 }, { "grad_norm": 0.508902370929718, "learning_rate": 9.891495870805336e-05, "loss": 0.0506, "step": 9050 }, { "grad_norm": 0.3638265132904053, "learning_rate": 9.891067209780298e-05, "loss": 0.0634, "step": 9060 }, { "grad_norm": 0.2789747416973114, "learning_rate": 9.890637713005646e-05, "loss": 0.0464, "step": 9070 }, { "grad_norm": 0.3553892970085144, "learning_rate": 9.890207380554767e-05, "loss": 0.0473, "step": 9080 }, { "grad_norm": 0.35728392004966736, "learning_rate": 9.889776212501196e-05, "loss": 0.0572, "step": 9090 }, { "grad_norm": 0.3190673291683197, "learning_rate": 9.889344208918605e-05, "loss": 0.0549, "step": 9100 }, { "grad_norm": 0.3739255368709564, "learning_rate": 9.888911369880812e-05, "loss": 0.0509, "step": 9110 }, { "grad_norm": 0.3693375587463379, "learning_rate": 9.888477695461777e-05, "loss": 0.0514, "step": 9120 }, { "grad_norm": 0.36199304461479187, "learning_rate": 9.888043185735607e-05, "loss": 0.0575, "step": 9130 }, { "grad_norm": 0.389072448015213, "learning_rate": 9.887607840776542e-05, "loss": 0.0459, "step": 9140 }, { "grad_norm": 0.5004851818084717, "learning_rate": 9.887171660658975e-05, "loss": 0.0661, "step": 9150 }, { "grad_norm": 0.2855278253555298, "learning_rate": 9.886734645457435e-05, "loss": 0.0501, "step": 9160 }, { "grad_norm": 0.4619764983654022, "learning_rate": 9.886296795246597e-05, "loss": 0.049, "step": 9170 }, { "grad_norm": 0.2773784399032593, "learning_rate": 9.885858110101276e-05, "loss": 0.0514, "step": 9180 }, { "grad_norm": 0.3910369277000427, "learning_rate": 9.885418590096434e-05, "loss": 0.0444, "step": 9190 }, { "grad_norm": 0.4144303798675537, "learning_rate": 9.88497823530717e-05, "loss": 0.0588, "step": 9200 }, { "grad_norm": 0.3427840769290924, "learning_rate": 9.884537045808732e-05, "loss": 0.0689, "step": 9210 }, { "grad_norm": 0.3539861738681793, "learning_rate": 9.884095021676502e-05, "loss": 0.0566, "step": 9220 }, { "grad_norm": 0.31533464789390564, "learning_rate": 9.883652162986017e-05, "loss": 0.0487, "step": 9230 }, { "grad_norm": 0.33105039596557617, "learning_rate": 9.883208469812943e-05, "loss": 0.0632, "step": 9240 }, { "grad_norm": 0.28895828127861023, "learning_rate": 9.882763942233098e-05, "loss": 0.0406, "step": 9250 }, { "grad_norm": 0.4967151880264282, "learning_rate": 9.882318580322441e-05, "loss": 0.0558, "step": 9260 }, { "grad_norm": 0.3110918700695038, "learning_rate": 9.881872384157067e-05, "loss": 0.0535, "step": 9270 }, { "grad_norm": 0.32740533351898193, "learning_rate": 9.881425353813225e-05, "loss": 0.0496, "step": 9280 }, { "grad_norm": 0.4557196795940399, "learning_rate": 9.880977489367296e-05, "loss": 0.0682, "step": 9290 }, { "grad_norm": 0.3272894024848938, "learning_rate": 9.88052879089581e-05, "loss": 0.061, "step": 9300 }, { "grad_norm": 0.4488731622695923, "learning_rate": 9.880079258475434e-05, "loss": 0.0504, "step": 9310 }, { "grad_norm": 0.4069036543369293, "learning_rate": 9.879628892182985e-05, "loss": 0.0611, "step": 9320 }, { "grad_norm": 0.4140278697013855, "learning_rate": 9.879177692095416e-05, "loss": 0.0507, "step": 9330 }, { "grad_norm": 0.46120139956474304, "learning_rate": 9.878725658289825e-05, "loss": 0.068, "step": 9340 }, { "grad_norm": 0.4046339690685272, "learning_rate": 9.878272790843454e-05, "loss": 0.0562, "step": 9350 }, { "grad_norm": 0.3509257137775421, "learning_rate": 9.877819089833682e-05, "loss": 0.0533, "step": 9360 }, { "grad_norm": 0.4976699650287628, "learning_rate": 9.877364555338038e-05, "loss": 0.067, "step": 9370 }, { "grad_norm": 0.5058965682983398, "learning_rate": 9.876909187434186e-05, "loss": 0.0515, "step": 9380 }, { "grad_norm": 0.41417956352233887, "learning_rate": 9.876452986199939e-05, "loss": 0.0551, "step": 9390 }, { "grad_norm": 0.3680115044116974, "learning_rate": 9.875995951713248e-05, "loss": 0.0616, "step": 9400 }, { "grad_norm": 0.3522776961326599, "learning_rate": 9.875538084052207e-05, "loss": 0.0621, "step": 9410 }, { "grad_norm": 0.6598427891731262, "learning_rate": 9.875079383295053e-05, "loss": 0.0568, "step": 9420 }, { "grad_norm": 0.43735989928245544, "learning_rate": 9.874619849520167e-05, "loss": 0.0461, "step": 9430 }, { "grad_norm": 0.4372156858444214, "learning_rate": 9.874159482806069e-05, "loss": 0.0518, "step": 9440 }, { "grad_norm": 0.29873141646385193, "learning_rate": 9.873698283231426e-05, "loss": 0.0459, "step": 9450 }, { "grad_norm": 0.4737343192100525, "learning_rate": 9.87323625087504e-05, "loss": 0.0529, "step": 9460 }, { "grad_norm": 0.4437967538833618, "learning_rate": 9.872773385815863e-05, "loss": 0.0699, "step": 9470 }, { "grad_norm": 0.35027357935905457, "learning_rate": 9.872309688132986e-05, "loss": 0.0472, "step": 9480 }, { "grad_norm": 0.3185220956802368, "learning_rate": 9.871845157905639e-05, "loss": 0.0452, "step": 9490 }, { "grad_norm": 0.3250884711742401, "learning_rate": 9.871379795213201e-05, "loss": 0.0472, "step": 9500 }, { "grad_norm": 0.25316300988197327, "learning_rate": 9.87091360013519e-05, "loss": 0.0511, "step": 9510 }, { "grad_norm": 0.42903223633766174, "learning_rate": 9.870446572751262e-05, "loss": 0.0627, "step": 9520 }, { "grad_norm": 0.36277586221694946, "learning_rate": 9.869978713141224e-05, "loss": 0.0497, "step": 9530 }, { "grad_norm": 0.4405355751514435, "learning_rate": 9.869510021385016e-05, "loss": 0.0581, "step": 9540 }, { "grad_norm": 0.275034099817276, "learning_rate": 9.869040497562727e-05, "loss": 0.0491, "step": 9550 }, { "grad_norm": 0.29239243268966675, "learning_rate": 9.868570141754587e-05, "loss": 0.0568, "step": 9560 }, { "grad_norm": 0.42845529317855835, "learning_rate": 9.868098954040965e-05, "loss": 0.0654, "step": 9570 }, { "grad_norm": 0.345284104347229, "learning_rate": 9.867626934502374e-05, "loss": 0.0613, "step": 9580 }, { "grad_norm": 0.3811255991458893, "learning_rate": 9.86715408321947e-05, "loss": 0.052, "step": 9590 }, { "grad_norm": 0.40435004234313965, "learning_rate": 9.86668040027305e-05, "loss": 0.0511, "step": 9600 }, { "grad_norm": 0.3463766276836395, "learning_rate": 9.866205885744053e-05, "loss": 0.0453, "step": 9610 }, { "grad_norm": 0.3778681457042694, "learning_rate": 9.865730539713563e-05, "loss": 0.0513, "step": 9620 }, { "grad_norm": 0.3811517357826233, "learning_rate": 9.8652543622628e-05, "loss": 0.0442, "step": 9630 }, { "grad_norm": 0.3369111716747284, "learning_rate": 9.864777353473132e-05, "loss": 0.044, "step": 9640 }, { "grad_norm": 0.2950879633426666, "learning_rate": 9.864299513426068e-05, "loss": 0.0495, "step": 9650 }, { "grad_norm": 0.35875117778778076, "learning_rate": 9.863820842203254e-05, "loss": 0.0477, "step": 9660 }, { "grad_norm": 0.3608224391937256, "learning_rate": 9.863341339886483e-05, "loss": 0.0528, "step": 9670 }, { "grad_norm": 0.3265710175037384, "learning_rate": 9.86286100655769e-05, "loss": 0.0422, "step": 9680 }, { "grad_norm": 0.5107349157333374, "learning_rate": 9.862379842298953e-05, "loss": 0.0516, "step": 9690 }, { "grad_norm": 0.4094040095806122, "learning_rate": 9.861897847192485e-05, "loss": 0.0452, "step": 9700 }, { "grad_norm": 0.43487754464149475, "learning_rate": 9.86141502132065e-05, "loss": 0.0429, "step": 9710 }, { "grad_norm": 0.28635621070861816, "learning_rate": 9.860931364765946e-05, "loss": 0.0521, "step": 9720 }, { "grad_norm": 0.5057131052017212, "learning_rate": 9.860446877611021e-05, "loss": 0.0619, "step": 9730 }, { "grad_norm": 0.41095593571662903, "learning_rate": 9.859961559938655e-05, "loss": 0.0492, "step": 9740 }, { "grad_norm": 0.38556644320487976, "learning_rate": 9.85947541183178e-05, "loss": 0.0476, "step": 9750 }, { "grad_norm": 0.3442963659763336, "learning_rate": 9.858988433373463e-05, "loss": 0.0527, "step": 9760 }, { "grad_norm": 0.3512479364871979, "learning_rate": 9.858500624646918e-05, "loss": 0.0505, "step": 9770 }, { "grad_norm": 0.377346932888031, "learning_rate": 9.858011985735497e-05, "loss": 0.0519, "step": 9780 }, { "grad_norm": 0.3493242561817169, "learning_rate": 9.857522516722693e-05, "loss": 0.0481, "step": 9790 }, { "grad_norm": 0.2956080436706543, "learning_rate": 9.857032217692145e-05, "loss": 0.0421, "step": 9800 }, { "grad_norm": 0.439777135848999, "learning_rate": 9.856541088727631e-05, "loss": 0.0507, "step": 9810 }, { "grad_norm": 0.513818085193634, "learning_rate": 9.856049129913072e-05, "loss": 0.0534, "step": 9820 }, { "grad_norm": 0.3789896070957184, "learning_rate": 9.85555634133253e-05, "loss": 0.0554, "step": 9830 }, { "grad_norm": 0.3242129385471344, "learning_rate": 9.855062723070208e-05, "loss": 0.0614, "step": 9840 }, { "grad_norm": 0.3325449526309967, "learning_rate": 9.854568275210454e-05, "loss": 0.0627, "step": 9850 }, { "grad_norm": 0.4713754951953888, "learning_rate": 9.854072997837754e-05, "loss": 0.0535, "step": 9860 }, { "grad_norm": 0.3784227967262268, "learning_rate": 9.853576891036737e-05, "loss": 0.0536, "step": 9870 }, { "grad_norm": 0.3985401391983032, "learning_rate": 9.853079954892177e-05, "loss": 0.0674, "step": 9880 }, { "grad_norm": 0.330801784992218, "learning_rate": 9.852582189488983e-05, "loss": 0.0526, "step": 9890 }, { "grad_norm": 0.34224358201026917, "learning_rate": 9.852083594912212e-05, "loss": 0.0587, "step": 9900 }, { "grad_norm": 0.4420894980430603, "learning_rate": 9.851584171247058e-05, "loss": 0.0601, "step": 9910 }, { "grad_norm": 0.3954101502895355, "learning_rate": 9.851083918578863e-05, "loss": 0.0645, "step": 9920 }, { "grad_norm": 0.35100340843200684, "learning_rate": 9.850582836993103e-05, "loss": 0.0539, "step": 9930 }, { "grad_norm": 0.33627814054489136, "learning_rate": 9.850080926575397e-05, "loss": 0.0483, "step": 9940 }, { "grad_norm": 0.34283021092414856, "learning_rate": 9.849578187411515e-05, "loss": 0.0413, "step": 9950 }, { "grad_norm": 0.3026321828365326, "learning_rate": 9.849074619587354e-05, "loss": 0.0481, "step": 9960 }, { "grad_norm": 0.5906696915626526, "learning_rate": 9.848570223188964e-05, "loss": 0.0634, "step": 9970 }, { "grad_norm": 0.3733430802822113, "learning_rate": 9.848064998302531e-05, "loss": 0.056, "step": 9980 }, { "grad_norm": 0.5166105628013611, "learning_rate": 9.847558945014386e-05, "loss": 0.0568, "step": 9990 }, { "grad_norm": 0.42684078216552734, "learning_rate": 9.847052063410996e-05, "loss": 0.0602, "step": 10000 }, { "grad_norm": 0.3495195508003235, "learning_rate": 9.846544353578977e-05, "loss": 0.0552, "step": 10010 }, { "grad_norm": 0.35331037640571594, "learning_rate": 9.846035815605081e-05, "loss": 0.0501, "step": 10020 }, { "grad_norm": 0.4665534496307373, "learning_rate": 9.845526449576204e-05, "loss": 0.0417, "step": 10030 }, { "grad_norm": 0.40700972080230713, "learning_rate": 9.845016255579383e-05, "loss": 0.0536, "step": 10040 }, { "grad_norm": 0.31612154841423035, "learning_rate": 9.844505233701794e-05, "loss": 0.0437, "step": 10050 }, { "grad_norm": 0.3721238076686859, "learning_rate": 9.843993384030757e-05, "loss": 0.0555, "step": 10060 }, { "grad_norm": 0.316663533449173, "learning_rate": 9.843480706653737e-05, "loss": 0.0424, "step": 10070 }, { "grad_norm": 0.4047912061214447, "learning_rate": 9.84296720165833e-05, "loss": 0.0536, "step": 10080 }, { "grad_norm": 0.29965710639953613, "learning_rate": 9.842452869132286e-05, "loss": 0.0502, "step": 10090 }, { "grad_norm": 0.35497361421585083, "learning_rate": 9.841937709163489e-05, "loss": 0.0446, "step": 10100 }, { "grad_norm": 0.28081750869750977, "learning_rate": 9.841421721839962e-05, "loss": 0.0425, "step": 10110 }, { "grad_norm": 0.27952778339385986, "learning_rate": 9.840904907249879e-05, "loss": 0.0467, "step": 10120 }, { "grad_norm": 0.4673019051551819, "learning_rate": 9.840387265481545e-05, "loss": 0.0523, "step": 10130 }, { "grad_norm": 0.289666086435318, "learning_rate": 9.839868796623411e-05, "loss": 0.0454, "step": 10140 }, { "grad_norm": 0.30146920680999756, "learning_rate": 9.839349500764072e-05, "loss": 0.0495, "step": 10150 }, { "grad_norm": 0.42299729585647583, "learning_rate": 9.83882937799226e-05, "loss": 0.0464, "step": 10160 }, { "grad_norm": 0.34245768189430237, "learning_rate": 9.838308428396849e-05, "loss": 0.044, "step": 10170 }, { "grad_norm": 0.34737277030944824, "learning_rate": 9.837786652066854e-05, "loss": 0.0534, "step": 10180 }, { "grad_norm": 0.25071200728416443, "learning_rate": 9.837264049091437e-05, "loss": 0.0536, "step": 10190 }, { "grad_norm": 0.4240446388721466, "learning_rate": 9.836740619559893e-05, "loss": 0.0496, "step": 10200 }, { "grad_norm": 0.40801459550857544, "learning_rate": 9.836216363561659e-05, "loss": 0.0504, "step": 10210 }, { "grad_norm": 0.36917173862457275, "learning_rate": 9.835691281186322e-05, "loss": 0.0445, "step": 10220 }, { "grad_norm": 0.3919163644313812, "learning_rate": 9.8351653725236e-05, "loss": 0.0567, "step": 10230 }, { "grad_norm": 0.2973801791667938, "learning_rate": 9.83463863766336e-05, "loss": 0.0502, "step": 10240 }, { "grad_norm": 0.3025505542755127, "learning_rate": 9.834111076695602e-05, "loss": 0.0461, "step": 10250 }, { "grad_norm": 0.35054925084114075, "learning_rate": 9.833582689710477e-05, "loss": 0.043, "step": 10260 }, { "grad_norm": 0.28263261914253235, "learning_rate": 9.833053476798268e-05, "loss": 0.0528, "step": 10270 }, { "grad_norm": 0.3033379912376404, "learning_rate": 9.832523438049404e-05, "loss": 0.0495, "step": 10280 }, { "grad_norm": 0.48042160272598267, "learning_rate": 9.831992573554454e-05, "loss": 0.0609, "step": 10290 }, { "grad_norm": 0.3670100271701813, "learning_rate": 9.831460883404128e-05, "loss": 0.0472, "step": 10300 }, { "grad_norm": 0.4418092668056488, "learning_rate": 9.830928367689278e-05, "loss": 0.0483, "step": 10310 }, { "grad_norm": 0.4056044816970825, "learning_rate": 9.830395026500896e-05, "loss": 0.0528, "step": 10320 }, { "grad_norm": 0.8279550075531006, "learning_rate": 9.829860859930115e-05, "loss": 0.0501, "step": 10330 }, { "grad_norm": 0.39085641503334045, "learning_rate": 9.829325868068212e-05, "loss": 0.0576, "step": 10340 }, { "grad_norm": 0.462886780500412, "learning_rate": 9.8287900510066e-05, "loss": 0.0596, "step": 10350 }, { "grad_norm": 0.3021717369556427, "learning_rate": 9.828253408836834e-05, "loss": 0.0538, "step": 10360 }, { "grad_norm": 0.4061485826969147, "learning_rate": 9.827715941650615e-05, "loss": 0.0491, "step": 10370 }, { "grad_norm": 0.28030362725257874, "learning_rate": 9.82717764953978e-05, "loss": 0.0518, "step": 10380 }, { "grad_norm": 0.3284587860107422, "learning_rate": 9.826638532596308e-05, "loss": 0.0521, "step": 10390 }, { "grad_norm": 0.32008296251296997, "learning_rate": 9.82609859091232e-05, "loss": 0.053, "step": 10400 }, { "grad_norm": 0.4022481143474579, "learning_rate": 9.825557824580076e-05, "loss": 0.0532, "step": 10410 }, { "grad_norm": 0.37448713183403015, "learning_rate": 9.82501623369198e-05, "loss": 0.0558, "step": 10420 }, { "grad_norm": 0.39767926931381226, "learning_rate": 9.824473818340574e-05, "loss": 0.0459, "step": 10430 }, { "grad_norm": 0.3630606234073639, "learning_rate": 9.823930578618541e-05, "loss": 0.0477, "step": 10440 }, { "grad_norm": 0.32950329780578613, "learning_rate": 9.823386514618709e-05, "loss": 0.0549, "step": 10450 }, { "grad_norm": 0.33731764554977417, "learning_rate": 9.82284162643404e-05, "loss": 0.0408, "step": 10460 }, { "grad_norm": 0.4062858819961548, "learning_rate": 9.822295914157642e-05, "loss": 0.0463, "step": 10470 }, { "grad_norm": 0.34708172082901, "learning_rate": 9.821749377882763e-05, "loss": 0.0426, "step": 10480 }, { "grad_norm": 0.21981193125247955, "learning_rate": 9.821202017702791e-05, "loss": 0.0393, "step": 10490 }, { "grad_norm": 0.3631404638290405, "learning_rate": 9.820653833711253e-05, "loss": 0.0407, "step": 10500 }, { "grad_norm": 0.3514969050884247, "learning_rate": 9.820104826001822e-05, "loss": 0.0543, "step": 10510 }, { "grad_norm": 0.3548215627670288, "learning_rate": 9.819554994668305e-05, "loss": 0.0436, "step": 10520 }, { "grad_norm": 0.37860292196273804, "learning_rate": 9.819004339804654e-05, "loss": 0.0552, "step": 10530 }, { "grad_norm": 0.36843812465667725, "learning_rate": 9.818452861504961e-05, "loss": 0.0487, "step": 10540 }, { "grad_norm": 0.39681872725486755, "learning_rate": 9.81790055986346e-05, "loss": 0.0418, "step": 10550 }, { "grad_norm": 0.4841613173484802, "learning_rate": 9.817347434974523e-05, "loss": 0.0434, "step": 10560 }, { "grad_norm": 0.5899572968482971, "learning_rate": 9.816793486932664e-05, "loss": 0.046, "step": 10570 }, { "grad_norm": 0.443198025226593, "learning_rate": 9.816238715832538e-05, "loss": 0.0531, "step": 10580 }, { "grad_norm": 0.31922540068626404, "learning_rate": 9.815683121768939e-05, "loss": 0.0459, "step": 10590 }, { "grad_norm": 0.4216412603855133, "learning_rate": 9.815126704836804e-05, "loss": 0.0576, "step": 10600 }, { "grad_norm": 0.2924049496650696, "learning_rate": 9.81456946513121e-05, "loss": 0.069, "step": 10610 }, { "grad_norm": 0.3710961639881134, "learning_rate": 9.814011402747373e-05, "loss": 0.0612, "step": 10620 }, { "grad_norm": 0.2365480363368988, "learning_rate": 9.813452517780651e-05, "loss": 0.042, "step": 10630 }, { "grad_norm": 0.2973170876502991, "learning_rate": 9.81289281032654e-05, "loss": 0.0553, "step": 10640 }, { "grad_norm": 0.3768279552459717, "learning_rate": 9.812332280480683e-05, "loss": 0.06, "step": 10650 }, { "grad_norm": 0.39910003542900085, "learning_rate": 9.811770928338854e-05, "loss": 0.0625, "step": 10660 }, { "grad_norm": 0.3890138566493988, "learning_rate": 9.811208753996979e-05, "loss": 0.054, "step": 10670 }, { "grad_norm": 0.39493826031684875, "learning_rate": 9.810645757551113e-05, "loss": 0.0434, "step": 10680 }, { "grad_norm": 0.3013497591018677, "learning_rate": 9.810081939097459e-05, "loss": 0.0421, "step": 10690 }, { "grad_norm": 0.41916877031326294, "learning_rate": 9.809517298732356e-05, "loss": 0.0689, "step": 10700 }, { "grad_norm": 0.32744961977005005, "learning_rate": 9.80895183655229e-05, "loss": 0.0512, "step": 10710 }, { "grad_norm": 0.2633541226387024, "learning_rate": 9.808385552653877e-05, "loss": 0.0451, "step": 10720 }, { "grad_norm": 0.2757241427898407, "learning_rate": 9.807818447133886e-05, "loss": 0.0419, "step": 10730 }, { "grad_norm": 0.3419446349143982, "learning_rate": 9.807250520089215e-05, "loss": 0.0399, "step": 10740 }, { "grad_norm": 0.33984777331352234, "learning_rate": 9.806681771616908e-05, "loss": 0.0503, "step": 10750 }, { "grad_norm": 0.3053440451622009, "learning_rate": 9.80611220181415e-05, "loss": 0.0537, "step": 10760 }, { "grad_norm": 0.47802621126174927, "learning_rate": 9.805541810778264e-05, "loss": 0.0565, "step": 10770 }, { "grad_norm": 0.4282189607620239, "learning_rate": 9.804970598606716e-05, "loss": 0.0504, "step": 10780 }, { "grad_norm": 0.372513085603714, "learning_rate": 9.804398565397106e-05, "loss": 0.0466, "step": 10790 }, { "grad_norm": 0.36559662222862244, "learning_rate": 9.803825711247183e-05, "loss": 0.0551, "step": 10800 }, { "grad_norm": 0.42134082317352295, "learning_rate": 9.803252036254831e-05, "loss": 0.0483, "step": 10810 }, { "grad_norm": 0.32085227966308594, "learning_rate": 9.802677540518076e-05, "loss": 0.0493, "step": 10820 }, { "grad_norm": 0.44576478004455566, "learning_rate": 9.802102224135081e-05, "loss": 0.0473, "step": 10830 }, { "grad_norm": 0.3359218239784241, "learning_rate": 9.801526087204155e-05, "loss": 0.0489, "step": 10840 }, { "grad_norm": 0.4236951768398285, "learning_rate": 9.800949129823743e-05, "loss": 0.0553, "step": 10850 }, { "grad_norm": 0.36919018626213074, "learning_rate": 9.80037135209243e-05, "loss": 0.0471, "step": 10860 }, { "grad_norm": 0.4655466377735138, "learning_rate": 9.799792754108946e-05, "loss": 0.0475, "step": 10870 }, { "grad_norm": 0.26976528763771057, "learning_rate": 9.799213335972152e-05, "loss": 0.0501, "step": 10880 }, { "grad_norm": 0.356387197971344, "learning_rate": 9.798633097781058e-05, "loss": 0.0458, "step": 10890 }, { "grad_norm": 0.30014747381210327, "learning_rate": 9.79805203963481e-05, "loss": 0.055, "step": 10900 }, { "grad_norm": 0.3814883530139923, "learning_rate": 9.797470161632697e-05, "loss": 0.0429, "step": 10910 }, { "grad_norm": 0.28289589285850525, "learning_rate": 9.796887463874145e-05, "loss": 0.0455, "step": 10920 }, { "grad_norm": 0.3320058286190033, "learning_rate": 9.796303946458718e-05, "loss": 0.0572, "step": 10930 }, { "grad_norm": 0.4353282153606415, "learning_rate": 9.795719609486127e-05, "loss": 0.0534, "step": 10940 }, { "grad_norm": 0.3353002667427063, "learning_rate": 9.795134453056219e-05, "loss": 0.0499, "step": 10950 }, { "grad_norm": 0.27636879682540894, "learning_rate": 9.794548477268979e-05, "loss": 0.0394, "step": 10960 }, { "grad_norm": 0.3269585371017456, "learning_rate": 9.793961682224537e-05, "loss": 0.0421, "step": 10970 }, { "grad_norm": 0.29007160663604736, "learning_rate": 9.793374068023156e-05, "loss": 0.0465, "step": 10980 }, { "grad_norm": 0.3287290036678314, "learning_rate": 9.792785634765247e-05, "loss": 0.0467, "step": 10990 }, { "grad_norm": 0.2557740807533264, "learning_rate": 9.792196382551357e-05, "loss": 0.0401, "step": 11000 }, { "grad_norm": 0.29265889525413513, "learning_rate": 9.791606311482171e-05, "loss": 0.0433, "step": 11010 }, { "grad_norm": 0.4183748662471771, "learning_rate": 9.791015421658518e-05, "loss": 0.0472, "step": 11020 }, { "grad_norm": 0.37188732624053955, "learning_rate": 9.790423713181362e-05, "loss": 0.041, "step": 11030 }, { "grad_norm": 0.2839495539665222, "learning_rate": 9.789831186151814e-05, "loss": 0.0399, "step": 11040 }, { "grad_norm": 0.3502497971057892, "learning_rate": 9.789237840671118e-05, "loss": 0.0404, "step": 11050 }, { "grad_norm": 0.36417362093925476, "learning_rate": 9.78864367684066e-05, "loss": 0.0489, "step": 11060 }, { "grad_norm": 0.3589928150177002, "learning_rate": 9.788048694761968e-05, "loss": 0.0416, "step": 11070 }, { "grad_norm": 0.31831714510917664, "learning_rate": 9.787452894536709e-05, "loss": 0.0457, "step": 11080 }, { "grad_norm": 0.371077299118042, "learning_rate": 9.786856276266685e-05, "loss": 0.043, "step": 11090 }, { "grad_norm": 0.3055412769317627, "learning_rate": 9.786258840053845e-05, "loss": 0.0491, "step": 11100 }, { "grad_norm": 0.28990915417671204, "learning_rate": 9.785660586000273e-05, "loss": 0.0507, "step": 11110 }, { "grad_norm": 0.22162391245365143, "learning_rate": 9.785061514208196e-05, "loss": 0.0422, "step": 11120 }, { "grad_norm": 0.27910763025283813, "learning_rate": 9.784461624779977e-05, "loss": 0.055, "step": 11130 }, { "grad_norm": 0.5411117076873779, "learning_rate": 9.783860917818123e-05, "loss": 0.046, "step": 11140 }, { "grad_norm": 0.3851487338542938, "learning_rate": 9.783259393425277e-05, "loss": 0.0353, "step": 11150 }, { "grad_norm": 0.37534502148628235, "learning_rate": 9.782657051704221e-05, "loss": 0.0442, "step": 11160 }, { "grad_norm": 0.2803233861923218, "learning_rate": 9.782053892757883e-05, "loss": 0.0388, "step": 11170 }, { "grad_norm": 0.40827926993370056, "learning_rate": 9.781449916689324e-05, "loss": 0.0429, "step": 11180 }, { "grad_norm": 0.399796724319458, "learning_rate": 9.780845123601746e-05, "loss": 0.0575, "step": 11190 }, { "grad_norm": 0.5310270190238953, "learning_rate": 9.780239513598492e-05, "loss": 0.0543, "step": 11200 }, { "grad_norm": 0.4282005727291107, "learning_rate": 9.779633086783047e-05, "loss": 0.0504, "step": 11210 }, { "grad_norm": 0.4026923179626465, "learning_rate": 9.779025843259031e-05, "loss": 0.0442, "step": 11220 }, { "grad_norm": 0.275418758392334, "learning_rate": 9.778417783130204e-05, "loss": 0.0557, "step": 11230 }, { "grad_norm": 0.28586575388908386, "learning_rate": 9.777808906500468e-05, "loss": 0.0519, "step": 11240 }, { "grad_norm": 0.33338287472724915, "learning_rate": 9.777199213473862e-05, "loss": 0.0389, "step": 11250 }, { "grad_norm": 0.36513379216194153, "learning_rate": 9.77658870415457e-05, "loss": 0.0424, "step": 11260 }, { "grad_norm": 0.3456001877784729, "learning_rate": 9.775977378646906e-05, "loss": 0.0433, "step": 11270 }, { "grad_norm": 0.4426993131637573, "learning_rate": 9.775365237055331e-05, "loss": 0.0622, "step": 11280 }, { "grad_norm": 0.36683234572410583, "learning_rate": 9.774752279484445e-05, "loss": 0.0459, "step": 11290 }, { "grad_norm": 0.2936476469039917, "learning_rate": 9.774138506038984e-05, "loss": 0.0436, "step": 11300 }, { "grad_norm": 0.34306401014328003, "learning_rate": 9.773523916823826e-05, "loss": 0.0452, "step": 11310 }, { "grad_norm": 0.28588950634002686, "learning_rate": 9.772908511943986e-05, "loss": 0.0481, "step": 11320 }, { "grad_norm": 0.39111921191215515, "learning_rate": 9.77229229150462e-05, "loss": 0.0487, "step": 11330 }, { "grad_norm": 0.3611043691635132, "learning_rate": 9.771675255611024e-05, "loss": 0.0419, "step": 11340 }, { "grad_norm": 0.3546876907348633, "learning_rate": 9.771057404368632e-05, "loss": 0.0658, "step": 11350 }, { "grad_norm": 0.3463813066482544, "learning_rate": 9.770438737883018e-05, "loss": 0.0527, "step": 11360 }, { "grad_norm": 0.3731580078601837, "learning_rate": 9.769819256259898e-05, "loss": 0.0419, "step": 11370 }, { "grad_norm": 0.38596147298812866, "learning_rate": 9.769198959605119e-05, "loss": 0.0454, "step": 11380 }, { "grad_norm": 0.3766334354877472, "learning_rate": 9.768577848024678e-05, "loss": 0.0592, "step": 11390 }, { "grad_norm": 0.425969660282135, "learning_rate": 9.767955921624702e-05, "loss": 0.056, "step": 11400 }, { "grad_norm": 0.4161304533481598, "learning_rate": 9.767333180511465e-05, "loss": 0.0387, "step": 11410 }, { "grad_norm": 0.31414711475372314, "learning_rate": 9.766709624791373e-05, "loss": 0.0477, "step": 11420 }, { "grad_norm": 0.2897292375564575, "learning_rate": 9.766085254570975e-05, "loss": 0.0419, "step": 11430 }, { "grad_norm": 0.33347582817077637, "learning_rate": 9.76546006995696e-05, "loss": 0.0428, "step": 11440 }, { "grad_norm": 0.3437913656234741, "learning_rate": 9.764834071056155e-05, "loss": 0.0439, "step": 11450 }, { "grad_norm": 0.3724147379398346, "learning_rate": 9.764207257975526e-05, "loss": 0.0505, "step": 11460 }, { "grad_norm": 0.37240201234817505, "learning_rate": 9.763579630822179e-05, "loss": 0.0457, "step": 11470 }, { "grad_norm": 0.29357174038887024, "learning_rate": 9.762951189703356e-05, "loss": 0.0373, "step": 11480 }, { "grad_norm": 0.38539597392082214, "learning_rate": 9.762321934726442e-05, "loss": 0.0474, "step": 11490 }, { "grad_norm": 0.35416415333747864, "learning_rate": 9.761691865998959e-05, "loss": 0.04, "step": 11500 }, { "grad_norm": 0.2504614591598511, "learning_rate": 9.76106098362857e-05, "loss": 0.0369, "step": 11510 }, { "grad_norm": 0.3257100284099579, "learning_rate": 9.760429287723072e-05, "loss": 0.0417, "step": 11520 }, { "grad_norm": 0.4685854911804199, "learning_rate": 9.759796778390406e-05, "loss": 0.0414, "step": 11530 }, { "grad_norm": 0.31230345368385315, "learning_rate": 9.759163455738653e-05, "loss": 0.0414, "step": 11540 }, { "grad_norm": 0.3518829643726349, "learning_rate": 9.75852931987603e-05, "loss": 0.0456, "step": 11550 }, { "grad_norm": 0.2963595986366272, "learning_rate": 9.757894370910891e-05, "loss": 0.0419, "step": 11560 }, { "grad_norm": 0.32662126421928406, "learning_rate": 9.757258608951733e-05, "loss": 0.036, "step": 11570 }, { "grad_norm": 0.4056953191757202, "learning_rate": 9.75662203410719e-05, "loss": 0.0444, "step": 11580 }, { "grad_norm": 0.3778083324432373, "learning_rate": 9.755984646486034e-05, "loss": 0.0477, "step": 11590 }, { "grad_norm": 0.32114166021347046, "learning_rate": 9.75534644619718e-05, "loss": 0.0429, "step": 11600 }, { "grad_norm": 0.2992969751358032, "learning_rate": 9.754707433349676e-05, "loss": 0.0518, "step": 11610 }, { "grad_norm": 0.3400469422340393, "learning_rate": 9.754067608052715e-05, "loss": 0.0488, "step": 11620 }, { "grad_norm": 0.32875797152519226, "learning_rate": 9.753426970415622e-05, "loss": 0.0638, "step": 11630 }, { "grad_norm": 0.3206680119037628, "learning_rate": 9.752785520547868e-05, "loss": 0.042, "step": 11640 }, { "grad_norm": 0.34829866886138916, "learning_rate": 9.752143258559056e-05, "loss": 0.0435, "step": 11650 }, { "grad_norm": 0.3216668963432312, "learning_rate": 9.751500184558933e-05, "loss": 0.0469, "step": 11660 }, { "grad_norm": 0.2766817808151245, "learning_rate": 9.750856298657383e-05, "loss": 0.0365, "step": 11670 }, { "grad_norm": 0.3313349485397339, "learning_rate": 9.750211600964428e-05, "loss": 0.0456, "step": 11680 }, { "grad_norm": 0.3590764105319977, "learning_rate": 9.749566091590226e-05, "loss": 0.0421, "step": 11690 }, { "grad_norm": 0.3243299424648285, "learning_rate": 9.748919770645083e-05, "loss": 0.0431, "step": 11700 }, { "grad_norm": 0.3203932046890259, "learning_rate": 9.748272638239432e-05, "loss": 0.0411, "step": 11710 }, { "grad_norm": 0.31107932329177856, "learning_rate": 9.747624694483855e-05, "loss": 0.0404, "step": 11720 }, { "grad_norm": 0.36770063638687134, "learning_rate": 9.746975939489065e-05, "loss": 0.0329, "step": 11730 }, { "grad_norm": 0.35681721568107605, "learning_rate": 9.746326373365918e-05, "loss": 0.0424, "step": 11740 }, { "grad_norm": 0.38313379883766174, "learning_rate": 9.745675996225403e-05, "loss": 0.0521, "step": 11750 }, { "grad_norm": 0.2758781611919403, "learning_rate": 9.745024808178657e-05, "loss": 0.0389, "step": 11760 }, { "grad_norm": 0.30446067452430725, "learning_rate": 9.744372809336947e-05, "loss": 0.0414, "step": 11770 }, { "grad_norm": 0.36436283588409424, "learning_rate": 9.743719999811682e-05, "loss": 0.0384, "step": 11780 }, { "grad_norm": 0.26536911725997925, "learning_rate": 9.743066379714412e-05, "loss": 0.0395, "step": 11790 }, { "grad_norm": 0.32424262166023254, "learning_rate": 9.74241194915682e-05, "loss": 0.0399, "step": 11800 }, { "grad_norm": 0.3132893741130829, "learning_rate": 9.741756708250731e-05, "loss": 0.0472, "step": 11810 }, { "grad_norm": 0.2843575179576874, "learning_rate": 9.741100657108109e-05, "loss": 0.0427, "step": 11820 }, { "grad_norm": 0.4425945281982422, "learning_rate": 9.740443795841054e-05, "loss": 0.0582, "step": 11830 }, { "grad_norm": 0.3193880319595337, "learning_rate": 9.739786124561805e-05, "loss": 0.0441, "step": 11840 }, { "grad_norm": 0.3475829064846039, "learning_rate": 9.73912764338274e-05, "loss": 0.0511, "step": 11850 }, { "grad_norm": 0.3363094925880432, "learning_rate": 9.738468352416377e-05, "loss": 0.0553, "step": 11860 }, { "grad_norm": 0.49890831112861633, "learning_rate": 9.737808251775369e-05, "loss": 0.0533, "step": 11870 }, { "grad_norm": 0.3265444040298462, "learning_rate": 9.737147341572512e-05, "loss": 0.0575, "step": 11880 }, { "grad_norm": 0.3624761402606964, "learning_rate": 9.736485621920735e-05, "loss": 0.0435, "step": 11890 }, { "grad_norm": 0.40738070011138916, "learning_rate": 9.735823092933108e-05, "loss": 0.0489, "step": 11900 }, { "grad_norm": 0.36625057458877563, "learning_rate": 9.735159754722838e-05, "loss": 0.0476, "step": 11910 }, { "grad_norm": 0.36292868852615356, "learning_rate": 9.734495607403275e-05, "loss": 0.0422, "step": 11920 }, { "grad_norm": 0.41501757502555847, "learning_rate": 9.733830651087901e-05, "loss": 0.0439, "step": 11930 }, { "grad_norm": 0.3528241515159607, "learning_rate": 9.733164885890338e-05, "loss": 0.0474, "step": 11940 }, { "grad_norm": 0.3627758324146271, "learning_rate": 9.732498311924349e-05, "loss": 0.0388, "step": 11950 }, { "grad_norm": 0.47243207693099976, "learning_rate": 9.731830929303833e-05, "loss": 0.0465, "step": 11960 }, { "grad_norm": 0.38192927837371826, "learning_rate": 9.731162738142827e-05, "loss": 0.0451, "step": 11970 }, { "grad_norm": 0.35687142610549927, "learning_rate": 9.730493738555506e-05, "loss": 0.0532, "step": 11980 }, { "grad_norm": 0.43470633029937744, "learning_rate": 9.729823930656186e-05, "loss": 0.0541, "step": 11990 }, { "grad_norm": 0.41955703496932983, "learning_rate": 9.729153314559316e-05, "loss": 0.0598, "step": 12000 }, { "grad_norm": 0.33013102412223816, "learning_rate": 9.728481890379486e-05, "loss": 0.0453, "step": 12010 }, { "grad_norm": 0.3313421905040741, "learning_rate": 9.727809658231428e-05, "loss": 0.0472, "step": 12020 }, { "grad_norm": 0.39004841446876526, "learning_rate": 9.727136618230003e-05, "loss": 0.0418, "step": 12030 }, { "grad_norm": 0.26272889971733093, "learning_rate": 9.726462770490219e-05, "loss": 0.0444, "step": 12040 }, { "grad_norm": 0.436764121055603, "learning_rate": 9.725788115127214e-05, "loss": 0.0492, "step": 12050 }, { "grad_norm": 0.39463910460472107, "learning_rate": 9.725112652256274e-05, "loss": 0.0472, "step": 12060 }, { "grad_norm": 0.3257252275943756, "learning_rate": 9.724436381992812e-05, "loss": 0.0587, "step": 12070 }, { "grad_norm": 0.34397441148757935, "learning_rate": 9.723759304452387e-05, "loss": 0.0475, "step": 12080 }, { "grad_norm": 0.3599284589290619, "learning_rate": 9.72308141975069e-05, "loss": 0.0402, "step": 12090 }, { "grad_norm": 0.35797879099845886, "learning_rate": 9.722402728003557e-05, "loss": 0.0475, "step": 12100 }, { "grad_norm": 0.28628823161125183, "learning_rate": 9.721723229326953e-05, "loss": 0.044, "step": 12110 }, { "grad_norm": 0.318559467792511, "learning_rate": 9.721042923836992e-05, "loss": 0.0418, "step": 12120 }, { "grad_norm": 0.3325667679309845, "learning_rate": 9.720361811649914e-05, "loss": 0.0481, "step": 12130 }, { "grad_norm": 0.39474964141845703, "learning_rate": 9.719679892882106e-05, "loss": 0.0442, "step": 12140 }, { "grad_norm": 0.33157142996788025, "learning_rate": 9.718997167650085e-05, "loss": 0.0456, "step": 12150 }, { "grad_norm": 0.45448270440101624, "learning_rate": 9.718313636070515e-05, "loss": 0.044, "step": 12160 }, { "grad_norm": 0.3138445019721985, "learning_rate": 9.717629298260192e-05, "loss": 0.0565, "step": 12170 }, { "grad_norm": 0.31891050934791565, "learning_rate": 9.716944154336047e-05, "loss": 0.0619, "step": 12180 }, { "grad_norm": 0.43575623631477356, "learning_rate": 9.716258204415157e-05, "loss": 0.0458, "step": 12190 }, { "grad_norm": 0.373239129781723, "learning_rate": 9.715571448614728e-05, "loss": 0.0528, "step": 12200 }, { "grad_norm": 0.4622630178928375, "learning_rate": 9.71488388705211e-05, "loss": 0.0657, "step": 12210 }, { "grad_norm": 0.3904314637184143, "learning_rate": 9.714195519844788e-05, "loss": 0.0508, "step": 12220 }, { "grad_norm": 0.4162587523460388, "learning_rate": 9.713506347110386e-05, "loss": 0.0504, "step": 12230 }, { "grad_norm": 0.43310606479644775, "learning_rate": 9.712816368966662e-05, "loss": 0.0504, "step": 12240 }, { "grad_norm": 0.5657839775085449, "learning_rate": 9.712125585531517e-05, "loss": 0.0539, "step": 12250 }, { "grad_norm": 0.29152023792266846, "learning_rate": 9.711433996922988e-05, "loss": 0.0574, "step": 12260 }, { "grad_norm": 0.4447205662727356, "learning_rate": 9.710741603259245e-05, "loss": 0.0472, "step": 12270 }, { "grad_norm": 0.4298403859138489, "learning_rate": 9.710048404658603e-05, "loss": 0.0489, "step": 12280 }, { "grad_norm": 0.2881176173686981, "learning_rate": 9.709354401239508e-05, "loss": 0.0489, "step": 12290 }, { "grad_norm": 0.38555124402046204, "learning_rate": 9.708659593120546e-05, "loss": 0.0422, "step": 12300 }, { "grad_norm": 0.4354504942893982, "learning_rate": 9.707963980420443e-05, "loss": 0.0463, "step": 12310 }, { "grad_norm": 0.3630059063434601, "learning_rate": 9.707267563258058e-05, "loss": 0.0519, "step": 12320 }, { "grad_norm": 0.4141741991043091, "learning_rate": 9.70657034175239e-05, "loss": 0.0462, "step": 12330 }, { "grad_norm": 0.3459510803222656, "learning_rate": 9.705872316022577e-05, "loss": 0.0524, "step": 12340 }, { "grad_norm": 0.4079864025115967, "learning_rate": 9.705173486187891e-05, "loss": 0.0492, "step": 12350 }, { "grad_norm": 0.30446121096611023, "learning_rate": 9.704473852367741e-05, "loss": 0.0503, "step": 12360 }, { "grad_norm": 0.38431647419929504, "learning_rate": 9.70377341468168e-05, "loss": 0.0504, "step": 12370 }, { "grad_norm": 0.311985582113266, "learning_rate": 9.703072173249389e-05, "loss": 0.045, "step": 12380 }, { "grad_norm": 0.3668471574783325, "learning_rate": 9.702370128190693e-05, "loss": 0.0552, "step": 12390 }, { "grad_norm": 0.28938186168670654, "learning_rate": 9.701667279625552e-05, "loss": 0.0549, "step": 12400 }, { "grad_norm": 0.30594027042388916, "learning_rate": 9.700963627674065e-05, "loss": 0.052, "step": 12410 }, { "grad_norm": 0.3133656084537506, "learning_rate": 9.700259172456466e-05, "loss": 0.0421, "step": 12420 }, { "grad_norm": 0.5056242346763611, "learning_rate": 9.699553914093124e-05, "loss": 0.0603, "step": 12430 }, { "grad_norm": 0.3237967789173126, "learning_rate": 9.698847852704553e-05, "loss": 0.0555, "step": 12440 }, { "grad_norm": 0.36306941509246826, "learning_rate": 9.6981409884114e-05, "loss": 0.0458, "step": 12450 }, { "grad_norm": 0.3454953730106354, "learning_rate": 9.697433321334443e-05, "loss": 0.0561, "step": 12460 }, { "grad_norm": 0.3221973180770874, "learning_rate": 9.696724851594607e-05, "loss": 0.046, "step": 12470 }, { "grad_norm": 0.41093355417251587, "learning_rate": 9.696015579312952e-05, "loss": 0.0423, "step": 12480 }, { "grad_norm": 0.25620710849761963, "learning_rate": 9.695305504610668e-05, "loss": 0.0405, "step": 12490 }, { "grad_norm": 0.2909619212150574, "learning_rate": 9.694594627609092e-05, "loss": 0.0428, "step": 12500 }, { "grad_norm": 0.2904984951019287, "learning_rate": 9.693882948429691e-05, "loss": 0.0515, "step": 12510 }, { "grad_norm": 0.28364354372024536, "learning_rate": 9.693170467194071e-05, "loss": 0.0464, "step": 12520 }, { "grad_norm": 0.32500553131103516, "learning_rate": 9.692457184023977e-05, "loss": 0.047, "step": 12530 }, { "grad_norm": 0.364516943693161, "learning_rate": 9.691743099041291e-05, "loss": 0.0575, "step": 12540 }, { "grad_norm": 0.37772104144096375, "learning_rate": 9.691028212368027e-05, "loss": 0.0438, "step": 12550 }, { "grad_norm": 0.2729596197605133, "learning_rate": 9.690312524126342e-05, "loss": 0.0402, "step": 12560 }, { "grad_norm": 0.22309359908103943, "learning_rate": 9.689596034438527e-05, "loss": 0.0417, "step": 12570 }, { "grad_norm": 0.3439357876777649, "learning_rate": 9.688878743427012e-05, "loss": 0.0409, "step": 12580 }, { "grad_norm": 0.39545682072639465, "learning_rate": 9.688160651214359e-05, "loss": 0.0491, "step": 12590 }, { "grad_norm": 0.3098262846469879, "learning_rate": 9.687441757923273e-05, "loss": 0.049, "step": 12600 }, { "grad_norm": 0.31624171137809753, "learning_rate": 9.68672206367659e-05, "loss": 0.0364, "step": 12610 }, { "grad_norm": 0.27494317293167114, "learning_rate": 9.686001568597291e-05, "loss": 0.0415, "step": 12620 }, { "grad_norm": 0.3211715519428253, "learning_rate": 9.685280272808486e-05, "loss": 0.0362, "step": 12630 }, { "grad_norm": 0.29154258966445923, "learning_rate": 9.684558176433424e-05, "loss": 0.0579, "step": 12640 }, { "grad_norm": 0.34085655212402344, "learning_rate": 9.683835279595495e-05, "loss": 0.0437, "step": 12650 }, { "grad_norm": 0.29835009574890137, "learning_rate": 9.683111582418216e-05, "loss": 0.0534, "step": 12660 }, { "grad_norm": 0.2951970100402832, "learning_rate": 9.682387085025254e-05, "loss": 0.0429, "step": 12670 }, { "grad_norm": 0.2472623735666275, "learning_rate": 9.681661787540401e-05, "loss": 0.0502, "step": 12680 }, { "grad_norm": 0.3316689431667328, "learning_rate": 9.680935690087593e-05, "loss": 0.0506, "step": 12690 }, { "grad_norm": 0.2872734069824219, "learning_rate": 9.680208792790901e-05, "loss": 0.0485, "step": 12700 }, { "grad_norm": 0.48771214485168457, "learning_rate": 9.679481095774529e-05, "loss": 0.0441, "step": 12710 }, { "grad_norm": 0.42065560817718506, "learning_rate": 9.678752599162822e-05, "loss": 0.0397, "step": 12720 }, { "grad_norm": 0.3312217593193054, "learning_rate": 9.678023303080259e-05, "loss": 0.0383, "step": 12730 }, { "grad_norm": 0.33977648615837097, "learning_rate": 9.677293207651459e-05, "loss": 0.0439, "step": 12740 }, { "grad_norm": 0.35622575879096985, "learning_rate": 9.676562313001173e-05, "loss": 0.0408, "step": 12750 }, { "grad_norm": 0.4062582850456238, "learning_rate": 9.675830619254293e-05, "loss": 0.0558, "step": 12760 }, { "grad_norm": 0.2888488471508026, "learning_rate": 9.675098126535843e-05, "loss": 0.0494, "step": 12770 }, { "grad_norm": 0.266402006149292, "learning_rate": 9.674364834970988e-05, "loss": 0.0454, "step": 12780 }, { "grad_norm": 0.3350980281829834, "learning_rate": 9.673630744685028e-05, "loss": 0.0462, "step": 12790 }, { "grad_norm": 0.32916101813316345, "learning_rate": 9.672895855803397e-05, "loss": 0.0457, "step": 12800 }, { "grad_norm": 0.2692308723926544, "learning_rate": 9.672160168451667e-05, "loss": 0.0464, "step": 12810 }, { "grad_norm": 0.31595486402511597, "learning_rate": 9.671423682755549e-05, "loss": 0.0561, "step": 12820 }, { "grad_norm": 0.3924204111099243, "learning_rate": 9.670686398840888e-05, "loss": 0.0476, "step": 12830 }, { "grad_norm": 0.24993355572223663, "learning_rate": 9.669948316833664e-05, "loss": 0.0402, "step": 12840 }, { "grad_norm": 0.3219819962978363, "learning_rate": 9.669209436859997e-05, "loss": 0.0452, "step": 12850 }, { "grad_norm": 0.3127584159374237, "learning_rate": 9.66846975904614e-05, "loss": 0.0468, "step": 12860 }, { "grad_norm": 0.38912639021873474, "learning_rate": 9.667729283518483e-05, "loss": 0.0438, "step": 12870 }, { "grad_norm": 0.3719359338283539, "learning_rate": 9.666988010403557e-05, "loss": 0.048, "step": 12880 }, { "grad_norm": 0.3753342032432556, "learning_rate": 9.66624593982802e-05, "loss": 0.0495, "step": 12890 }, { "grad_norm": 0.3350154757499695, "learning_rate": 9.665503071918675e-05, "loss": 0.0529, "step": 12900 }, { "grad_norm": 0.27409133315086365, "learning_rate": 9.664759406802456e-05, "loss": 0.0473, "step": 12910 }, { "grad_norm": 0.3622220456600189, "learning_rate": 9.664014944606437e-05, "loss": 0.0357, "step": 12920 }, { "grad_norm": 0.3451075553894043, "learning_rate": 9.663269685457822e-05, "loss": 0.0427, "step": 12930 }, { "grad_norm": 0.38000988960266113, "learning_rate": 9.662523629483962e-05, "loss": 0.0497, "step": 12940 }, { "grad_norm": 0.27478477358818054, "learning_rate": 9.661776776812333e-05, "loss": 0.0383, "step": 12950 }, { "grad_norm": 0.32443100214004517, "learning_rate": 9.661029127570553e-05, "loss": 0.0438, "step": 12960 }, { "grad_norm": 0.34358489513397217, "learning_rate": 9.660280681886373e-05, "loss": 0.0468, "step": 12970 }, { "grad_norm": 0.36191055178642273, "learning_rate": 9.659531439887685e-05, "loss": 0.0397, "step": 12980 }, { "grad_norm": 0.30668747425079346, "learning_rate": 9.658781401702511e-05, "loss": 0.0389, "step": 12990 }, { "grad_norm": 0.3575226962566376, "learning_rate": 9.658030567459015e-05, "loss": 0.0433, "step": 13000 }, { "grad_norm": 0.46970438957214355, "learning_rate": 9.65727893728549e-05, "loss": 0.0591, "step": 13010 }, { "grad_norm": 0.3548637628555298, "learning_rate": 9.656526511310375e-05, "loss": 0.0553, "step": 13020 }, { "grad_norm": 0.38597697019577026, "learning_rate": 9.655773289662233e-05, "loss": 0.0554, "step": 13030 }, { "grad_norm": 0.4039151072502136, "learning_rate": 9.655019272469772e-05, "loss": 0.0454, "step": 13040 }, { "grad_norm": 0.3631962835788727, "learning_rate": 9.654264459861832e-05, "loss": 0.0448, "step": 13050 }, { "grad_norm": 0.26876339316368103, "learning_rate": 9.653508851967391e-05, "loss": 0.0422, "step": 13060 }, { "grad_norm": 0.3754293918609619, "learning_rate": 9.65275244891556e-05, "loss": 0.0571, "step": 13070 }, { "grad_norm": 0.3124825358390808, "learning_rate": 9.651995250835591e-05, "loss": 0.0493, "step": 13080 }, { "grad_norm": 0.372050940990448, "learning_rate": 9.651237257856862e-05, "loss": 0.0497, "step": 13090 }, { "grad_norm": 0.26709800958633423, "learning_rate": 9.6504784701089e-05, "loss": 0.0451, "step": 13100 }, { "grad_norm": 0.23558051884174347, "learning_rate": 9.649718887721357e-05, "loss": 0.0336, "step": 13110 }, { "grad_norm": 0.29497194290161133, "learning_rate": 9.648958510824028e-05, "loss": 0.0358, "step": 13120 }, { "grad_norm": 0.3608657419681549, "learning_rate": 9.648197339546837e-05, "loss": 0.0446, "step": 13130 }, { "grad_norm": 0.24071206152439117, "learning_rate": 9.647435374019851e-05, "loss": 0.0463, "step": 13140 }, { "grad_norm": 0.33509376645088196, "learning_rate": 9.646672614373266e-05, "loss": 0.0438, "step": 13150 }, { "grad_norm": 0.23310814797878265, "learning_rate": 9.645909060737418e-05, "loss": 0.0355, "step": 13160 }, { "grad_norm": 0.314706414937973, "learning_rate": 9.645144713242778e-05, "loss": 0.041, "step": 13170 }, { "grad_norm": 0.29768475890159607, "learning_rate": 9.64437957201995e-05, "loss": 0.0418, "step": 13180 }, { "grad_norm": 0.3666340410709381, "learning_rate": 9.643613637199678e-05, "loss": 0.0423, "step": 13190 }, { "grad_norm": 0.4336912930011749, "learning_rate": 9.642846908912839e-05, "loss": 0.0507, "step": 13200 }, { "grad_norm": 0.2918311357498169, "learning_rate": 9.642079387290444e-05, "loss": 0.0369, "step": 13210 }, { "grad_norm": 0.35997524857521057, "learning_rate": 9.641311072463644e-05, "loss": 0.0413, "step": 13220 }, { "grad_norm": 0.34169188141822815, "learning_rate": 9.640541964563722e-05, "loss": 0.0417, "step": 13230 }, { "grad_norm": 0.31198790669441223, "learning_rate": 9.639772063722096e-05, "loss": 0.0489, "step": 13240 }, { "grad_norm": 0.2567351162433624, "learning_rate": 9.639001370070324e-05, "loss": 0.0367, "step": 13250 }, { "grad_norm": 0.3884032070636749, "learning_rate": 9.638229883740095e-05, "loss": 0.0433, "step": 13260 }, { "grad_norm": 0.37792864441871643, "learning_rate": 9.637457604863233e-05, "loss": 0.043, "step": 13270 }, { "grad_norm": 0.3219437897205353, "learning_rate": 9.636684533571703e-05, "loss": 0.0439, "step": 13280 }, { "grad_norm": 0.2953260838985443, "learning_rate": 9.635910669997599e-05, "loss": 0.0457, "step": 13290 }, { "grad_norm": 0.3480205833911896, "learning_rate": 9.635136014273154e-05, "loss": 0.0369, "step": 13300 }, { "grad_norm": 0.30794814229011536, "learning_rate": 9.634360566530735e-05, "loss": 0.045, "step": 13310 }, { "grad_norm": 0.378616601228714, "learning_rate": 9.633584326902845e-05, "loss": 0.0471, "step": 13320 }, { "grad_norm": 0.2996373772621155, "learning_rate": 9.632807295522124e-05, "loss": 0.039, "step": 13330 }, { "grad_norm": 0.33152738213539124, "learning_rate": 9.632029472521342e-05, "loss": 0.04, "step": 13340 }, { "grad_norm": 0.30899327993392944, "learning_rate": 9.631250858033409e-05, "loss": 0.0383, "step": 13350 }, { "grad_norm": 0.3310941457748413, "learning_rate": 9.630471452191371e-05, "loss": 0.043, "step": 13360 }, { "grad_norm": 0.2383221834897995, "learning_rate": 9.629691255128405e-05, "loss": 0.0448, "step": 13370 }, { "grad_norm": 0.4282112121582031, "learning_rate": 9.628910266977825e-05, "loss": 0.0426, "step": 13380 }, { "grad_norm": 0.39276108145713806, "learning_rate": 9.628128487873083e-05, "loss": 0.0485, "step": 13390 }, { "grad_norm": 0.2956819534301758, "learning_rate": 9.627345917947761e-05, "loss": 0.0524, "step": 13400 }, { "grad_norm": 0.38690176606178284, "learning_rate": 9.626562557335579e-05, "loss": 0.0476, "step": 13410 }, { "grad_norm": 0.2886183261871338, "learning_rate": 9.625778406170393e-05, "loss": 0.05, "step": 13420 }, { "grad_norm": 0.3312789499759674, "learning_rate": 9.624993464586193e-05, "loss": 0.0435, "step": 13430 }, { "grad_norm": 0.2757115066051483, "learning_rate": 9.624207732717105e-05, "loss": 0.0428, "step": 13440 }, { "grad_norm": 0.3725881576538086, "learning_rate": 9.623421210697386e-05, "loss": 0.0461, "step": 13450 }, { "grad_norm": 0.3861514925956726, "learning_rate": 9.622633898661434e-05, "loss": 0.0504, "step": 13460 }, { "grad_norm": 0.36901500821113586, "learning_rate": 9.621845796743778e-05, "loss": 0.0594, "step": 13470 }, { "grad_norm": 0.2722601294517517, "learning_rate": 9.621056905079082e-05, "loss": 0.0443, "step": 13480 }, { "grad_norm": 0.3083535432815552, "learning_rate": 9.620267223802149e-05, "loss": 0.0462, "step": 13490 }, { "grad_norm": 0.3184105157852173, "learning_rate": 9.619476753047911e-05, "loss": 0.0348, "step": 13500 }, { "grad_norm": 0.3001801371574402, "learning_rate": 9.618685492951438e-05, "loss": 0.0469, "step": 13510 }, { "grad_norm": 0.2540890574455261, "learning_rate": 9.617893443647938e-05, "loss": 0.0552, "step": 13520 }, { "grad_norm": 0.3080143630504608, "learning_rate": 9.617100605272746e-05, "loss": 0.0585, "step": 13530 }, { "grad_norm": 0.3723299503326416, "learning_rate": 9.616306977961338e-05, "loss": 0.0551, "step": 13540 }, { "grad_norm": 0.43122613430023193, "learning_rate": 9.615512561849326e-05, "loss": 0.0401, "step": 13550 }, { "grad_norm": 0.3590448200702667, "learning_rate": 9.61471735707245e-05, "loss": 0.0493, "step": 13560 }, { "grad_norm": 0.333884060382843, "learning_rate": 9.613921363766592e-05, "loss": 0.0419, "step": 13570 }, { "grad_norm": 0.28143972158432007, "learning_rate": 9.613124582067763e-05, "loss": 0.0467, "step": 13580 }, { "grad_norm": 0.5451198816299438, "learning_rate": 9.612327012112112e-05, "loss": 0.0576, "step": 13590 }, { "grad_norm": 0.3536711633205414, "learning_rate": 9.611528654035921e-05, "loss": 0.064, "step": 13600 }, { "grad_norm": 0.2894448935985565, "learning_rate": 9.610729507975611e-05, "loss": 0.0421, "step": 13610 }, { "grad_norm": 0.2953687608242035, "learning_rate": 9.609929574067731e-05, "loss": 0.0503, "step": 13620 }, { "grad_norm": 0.3571934700012207, "learning_rate": 9.609128852448967e-05, "loss": 0.0381, "step": 13630 }, { "grad_norm": 0.3148418664932251, "learning_rate": 9.608327343256143e-05, "loss": 0.0402, "step": 13640 }, { "grad_norm": 0.4277511239051819, "learning_rate": 9.607525046626216e-05, "loss": 0.0486, "step": 13650 }, { "grad_norm": 0.33104291558265686, "learning_rate": 9.606721962696272e-05, "loss": 0.0465, "step": 13660 }, { "grad_norm": 0.44772234559059143, "learning_rate": 9.60591809160354e-05, "loss": 0.0381, "step": 13670 }, { "grad_norm": 0.30874699354171753, "learning_rate": 9.605113433485378e-05, "loss": 0.0394, "step": 13680 }, { "grad_norm": 0.4158157408237457, "learning_rate": 9.604307988479279e-05, "loss": 0.0448, "step": 13690 }, { "grad_norm": 0.3432582914829254, "learning_rate": 9.603501756722876e-05, "loss": 0.0447, "step": 13700 }, { "grad_norm": 0.27144303917884827, "learning_rate": 9.602694738353927e-05, "loss": 0.0404, "step": 13710 }, { "grad_norm": 0.3113251328468323, "learning_rate": 9.601886933510331e-05, "loss": 0.0428, "step": 13720 }, { "grad_norm": 0.49520212411880493, "learning_rate": 9.60107834233012e-05, "loss": 0.0447, "step": 13730 }, { "grad_norm": 0.36239707469940186, "learning_rate": 9.60026896495146e-05, "loss": 0.0404, "step": 13740 }, { "grad_norm": 0.38429340720176697, "learning_rate": 9.599458801512652e-05, "loss": 0.0406, "step": 13750 }, { "grad_norm": 0.3965672254562378, "learning_rate": 9.598647852152129e-05, "loss": 0.059, "step": 13760 }, { "grad_norm": 0.21327653527259827, "learning_rate": 9.597836117008462e-05, "loss": 0.0409, "step": 13770 }, { "grad_norm": 0.30368611216545105, "learning_rate": 9.597023596220356e-05, "loss": 0.0409, "step": 13780 }, { "grad_norm": 0.31388694047927856, "learning_rate": 9.596210289926643e-05, "loss": 0.0461, "step": 13790 }, { "grad_norm": 0.46689680218696594, "learning_rate": 9.5953961982663e-05, "loss": 0.043, "step": 13800 }, { "grad_norm": 0.41367408633232117, "learning_rate": 9.594581321378431e-05, "loss": 0.057, "step": 13810 }, { "grad_norm": 0.3079356849193573, "learning_rate": 9.593765659402276e-05, "loss": 0.0557, "step": 13820 }, { "grad_norm": 0.34800735116004944, "learning_rate": 9.59294921247721e-05, "loss": 0.0492, "step": 13830 }, { "grad_norm": 0.3112814426422119, "learning_rate": 9.59213198074274e-05, "loss": 0.0504, "step": 13840 }, { "grad_norm": 0.26857995986938477, "learning_rate": 9.59131396433851e-05, "loss": 0.0382, "step": 13850 }, { "grad_norm": 0.3057907521724701, "learning_rate": 9.590495163404297e-05, "loss": 0.0464, "step": 13860 }, { "grad_norm": 0.3305908739566803, "learning_rate": 9.589675578080009e-05, "loss": 0.046, "step": 13870 }, { "grad_norm": 0.32642653584480286, "learning_rate": 9.588855208505694e-05, "loss": 0.0394, "step": 13880 }, { "grad_norm": 0.41251304745674133, "learning_rate": 9.588034054821529e-05, "loss": 0.0532, "step": 13890 }, { "grad_norm": 0.22171613574028015, "learning_rate": 9.587212117167826e-05, "loss": 0.0383, "step": 13900 }, { "grad_norm": 0.32490041851997375, "learning_rate": 9.586389395685033e-05, "loss": 0.0443, "step": 13910 }, { "grad_norm": 0.35846972465515137, "learning_rate": 9.585565890513733e-05, "loss": 0.0455, "step": 13920 }, { "grad_norm": 0.3718245327472687, "learning_rate": 9.584741601794636e-05, "loss": 0.0473, "step": 13930 }, { "grad_norm": 0.22678738832473755, "learning_rate": 9.58391652966859e-05, "loss": 0.0416, "step": 13940 }, { "grad_norm": 0.2833535075187683, "learning_rate": 9.583090674276583e-05, "loss": 0.0453, "step": 13950 }, { "grad_norm": 0.3448253870010376, "learning_rate": 9.582264035759726e-05, "loss": 0.0422, "step": 13960 }, { "grad_norm": 0.32688528299331665, "learning_rate": 9.58143661425927e-05, "loss": 0.0616, "step": 13970 }, { "grad_norm": 0.3600134253501892, "learning_rate": 9.580608409916601e-05, "loss": 0.0512, "step": 13980 }, { "grad_norm": 0.34884896874427795, "learning_rate": 9.579779422873233e-05, "loss": 0.0432, "step": 13990 }, { "grad_norm": 0.3122742176055908, "learning_rate": 9.578949653270819e-05, "loss": 0.0469, "step": 14000 }, { "grad_norm": 0.3632911741733551, "learning_rate": 9.578119101251144e-05, "loss": 0.0403, "step": 14010 }, { "grad_norm": 0.36928269267082214, "learning_rate": 9.577287766956127e-05, "loss": 0.0516, "step": 14020 }, { "grad_norm": 0.33224597573280334, "learning_rate": 9.57645565052782e-05, "loss": 0.0455, "step": 14030 }, { "grad_norm": 0.3590426445007324, "learning_rate": 9.575622752108407e-05, "loss": 0.0402, "step": 14040 }, { "grad_norm": 0.30154281854629517, "learning_rate": 9.57478907184021e-05, "loss": 0.0436, "step": 14050 }, { "grad_norm": 0.31423312425613403, "learning_rate": 9.573954609865681e-05, "loss": 0.0438, "step": 14060 }, { "grad_norm": 0.3468669354915619, "learning_rate": 9.573119366327408e-05, "loss": 0.0482, "step": 14070 }, { "grad_norm": 0.2867783308029175, "learning_rate": 9.57228334136811e-05, "loss": 0.0415, "step": 14080 }, { "grad_norm": 0.3117583394050598, "learning_rate": 9.571446535130641e-05, "loss": 0.0397, "step": 14090 }, { "grad_norm": 0.41518667340278625, "learning_rate": 9.570608947757988e-05, "loss": 0.0435, "step": 14100 }, { "grad_norm": 0.3581177592277527, "learning_rate": 9.569770579393274e-05, "loss": 0.0449, "step": 14110 }, { "grad_norm": 0.3474122881889343, "learning_rate": 9.56893143017975e-05, "loss": 0.0482, "step": 14120 }, { "grad_norm": 0.33579590916633606, "learning_rate": 9.568091500260806e-05, "loss": 0.0498, "step": 14130 }, { "grad_norm": 0.2927698493003845, "learning_rate": 9.567250789779961e-05, "loss": 0.0472, "step": 14140 }, { "grad_norm": 0.28455159068107605, "learning_rate": 9.566409298880872e-05, "loss": 0.0392, "step": 14150 }, { "grad_norm": 0.39179062843322754, "learning_rate": 9.565567027707326e-05, "loss": 0.0568, "step": 14160 }, { "grad_norm": 0.31773075461387634, "learning_rate": 9.56472397640324e-05, "loss": 0.0455, "step": 14170 }, { "grad_norm": 0.39538562297821045, "learning_rate": 9.563880145112675e-05, "loss": 0.0451, "step": 14180 }, { "grad_norm": 0.3726343810558319, "learning_rate": 9.563035533979814e-05, "loss": 0.0496, "step": 14190 }, { "grad_norm": 0.27921420335769653, "learning_rate": 9.562190143148981e-05, "loss": 0.0359, "step": 14200 }, { "grad_norm": 0.29811787605285645, "learning_rate": 9.561343972764627e-05, "loss": 0.0409, "step": 14210 }, { "grad_norm": 0.22383365035057068, "learning_rate": 9.560497022971343e-05, "loss": 0.0455, "step": 14220 }, { "grad_norm": 0.29469454288482666, "learning_rate": 9.559649293913847e-05, "loss": 0.0422, "step": 14230 }, { "grad_norm": 0.345781534910202, "learning_rate": 9.558800785736993e-05, "loss": 0.0402, "step": 14240 }, { "grad_norm": 0.29155293107032776, "learning_rate": 9.557951498585767e-05, "loss": 0.0429, "step": 14250 }, { "grad_norm": 0.27379298210144043, "learning_rate": 9.557101432605293e-05, "loss": 0.036, "step": 14260 }, { "grad_norm": 0.3511975407600403, "learning_rate": 9.556250587940818e-05, "loss": 0.0484, "step": 14270 }, { "grad_norm": 0.2788086533546448, "learning_rate": 9.555398964737734e-05, "loss": 0.041, "step": 14280 }, { "grad_norm": 0.2964950501918793, "learning_rate": 9.554546563141555e-05, "loss": 0.0391, "step": 14290 }, { "grad_norm": 0.2982643246650696, "learning_rate": 9.553693383297937e-05, "loss": 0.0415, "step": 14300 }, { "grad_norm": 0.31415221095085144, "learning_rate": 9.552839425352663e-05, "loss": 0.0369, "step": 14310 }, { "grad_norm": 0.24070928990840912, "learning_rate": 9.551984689451652e-05, "loss": 0.0504, "step": 14320 }, { "grad_norm": 0.3301495909690857, "learning_rate": 9.551129175740953e-05, "loss": 0.0401, "step": 14330 }, { "grad_norm": 0.4238455891609192, "learning_rate": 9.550272884366754e-05, "loss": 0.0395, "step": 14340 }, { "grad_norm": 0.37366729974746704, "learning_rate": 9.549415815475369e-05, "loss": 0.0425, "step": 14350 }, { "grad_norm": 0.22214873135089874, "learning_rate": 9.548557969213247e-05, "loss": 0.0339, "step": 14360 }, { "grad_norm": 0.3564934730529785, "learning_rate": 9.547699345726972e-05, "loss": 0.04, "step": 14370 }, { "grad_norm": 0.32796967029571533, "learning_rate": 9.546839945163257e-05, "loss": 0.0361, "step": 14380 }, { "grad_norm": 0.3662305772304535, "learning_rate": 9.545979767668953e-05, "loss": 0.0578, "step": 14390 }, { "grad_norm": 0.281246542930603, "learning_rate": 9.54511881339104e-05, "loss": 0.0423, "step": 14400 }, { "grad_norm": 0.3370741009712219, "learning_rate": 9.54425708247663e-05, "loss": 0.0421, "step": 14410 }, { "grad_norm": 0.29292240738868713, "learning_rate": 9.543394575072972e-05, "loss": 0.0417, "step": 14420 }, { "grad_norm": 0.30305537581443787, "learning_rate": 9.542531291327441e-05, "loss": 0.0452, "step": 14430 }, { "grad_norm": 0.2469734102487564, "learning_rate": 9.541667231387552e-05, "loss": 0.0398, "step": 14440 }, { "grad_norm": 0.35749372839927673, "learning_rate": 9.540802395400949e-05, "loss": 0.0443, "step": 14450 }, { "grad_norm": 0.3155733048915863, "learning_rate": 9.539936783515406e-05, "loss": 0.0369, "step": 14460 }, { "grad_norm": 0.3356267213821411, "learning_rate": 9.539070395878835e-05, "loss": 0.0536, "step": 14470 }, { "grad_norm": 0.3521187901496887, "learning_rate": 9.538203232639277e-05, "loss": 0.0406, "step": 14480 }, { "grad_norm": 0.36382272839546204, "learning_rate": 9.537335293944907e-05, "loss": 0.0447, "step": 14490 }, { "grad_norm": 0.3023393750190735, "learning_rate": 9.536466579944032e-05, "loss": 0.0502, "step": 14500 }, { "grad_norm": 0.2960761785507202, "learning_rate": 9.535597090785091e-05, "loss": 0.049, "step": 14510 }, { "grad_norm": 0.36794763803482056, "learning_rate": 9.534726826616656e-05, "loss": 0.0379, "step": 14520 }, { "grad_norm": 0.3515905439853668, "learning_rate": 9.53385578758743e-05, "loss": 0.0554, "step": 14530 }, { "grad_norm": 0.3105199635028839, "learning_rate": 9.532983973846252e-05, "loss": 0.0384, "step": 14540 }, { "grad_norm": 0.383594274520874, "learning_rate": 9.53211138554209e-05, "loss": 0.0382, "step": 14550 }, { "grad_norm": 0.2992456257343292, "learning_rate": 9.531238022824047e-05, "loss": 0.0421, "step": 14560 }, { "grad_norm": 0.2262265533208847, "learning_rate": 9.530363885841355e-05, "loss": 0.0351, "step": 14570 }, { "grad_norm": 0.31157711148262024, "learning_rate": 9.52948897474338e-05, "loss": 0.0452, "step": 14580 }, { "grad_norm": 0.33189088106155396, "learning_rate": 9.528613289679622e-05, "loss": 0.0541, "step": 14590 }, { "grad_norm": 0.367151141166687, "learning_rate": 9.52773683079971e-05, "loss": 0.048, "step": 14600 }, { "grad_norm": 0.22647739946842194, "learning_rate": 9.526859598253407e-05, "loss": 0.0343, "step": 14610 }, { "grad_norm": 0.26602989435195923, "learning_rate": 9.525981592190609e-05, "loss": 0.0353, "step": 14620 }, { "grad_norm": 0.29753196239471436, "learning_rate": 9.525102812761342e-05, "loss": 0.0339, "step": 14630 }, { "grad_norm": 0.25588712096214294, "learning_rate": 9.524223260115768e-05, "loss": 0.0349, "step": 14640 }, { "grad_norm": 0.33179011940956116, "learning_rate": 9.523342934404175e-05, "loss": 0.0429, "step": 14650 }, { "grad_norm": 0.2483186572790146, "learning_rate": 9.522461835776989e-05, "loss": 0.0413, "step": 14660 }, { "grad_norm": 0.37926164269447327, "learning_rate": 9.521579964384764e-05, "loss": 0.0445, "step": 14670 }, { "grad_norm": 0.35028281807899475, "learning_rate": 9.52069732037819e-05, "loss": 0.0358, "step": 14680 }, { "grad_norm": 0.3512571156024933, "learning_rate": 9.519813903908083e-05, "loss": 0.0409, "step": 14690 }, { "grad_norm": 0.30071476101875305, "learning_rate": 9.5189297151254e-05, "loss": 0.0471, "step": 14700 }, { "grad_norm": 0.5772396922111511, "learning_rate": 9.518044754181218e-05, "loss": 0.0558, "step": 14710 }, { "grad_norm": 0.21300528943538666, "learning_rate": 9.51715902122676e-05, "loss": 0.0429, "step": 14720 }, { "grad_norm": 0.3183485269546509, "learning_rate": 9.516272516413368e-05, "loss": 0.0484, "step": 14730 }, { "grad_norm": 0.27718329429626465, "learning_rate": 9.515385239892525e-05, "loss": 0.0605, "step": 14740 }, { "grad_norm": 0.3067921996116638, "learning_rate": 9.514497191815839e-05, "loss": 0.0507, "step": 14750 }, { "grad_norm": 0.3508376181125641, "learning_rate": 9.513608372335055e-05, "loss": 0.0515, "step": 14760 }, { "grad_norm": 0.3117135167121887, "learning_rate": 9.512718781602045e-05, "loss": 0.0453, "step": 14770 }, { "grad_norm": 0.23087945580482483, "learning_rate": 9.511828419768823e-05, "loss": 0.0411, "step": 14780 }, { "grad_norm": 0.38552334904670715, "learning_rate": 9.510937286987521e-05, "loss": 0.0374, "step": 14790 }, { "grad_norm": 0.36088523268699646, "learning_rate": 9.510045383410408e-05, "loss": 0.0449, "step": 14800 }, { "grad_norm": 0.2651139795780182, "learning_rate": 9.509152709189892e-05, "loss": 0.0398, "step": 14810 }, { "grad_norm": 0.3024272918701172, "learning_rate": 9.508259264478504e-05, "loss": 0.0429, "step": 14820 }, { "grad_norm": 0.2720775008201599, "learning_rate": 9.507365049428909e-05, "loss": 0.0555, "step": 14830 }, { "grad_norm": 0.35599544644355774, "learning_rate": 9.506470064193902e-05, "loss": 0.0629, "step": 14840 }, { "grad_norm": 0.2950197458267212, "learning_rate": 9.505574308926414e-05, "loss": 0.0514, "step": 14850 }, { "grad_norm": 0.3480195105075836, "learning_rate": 9.504677783779505e-05, "loss": 0.0429, "step": 14860 }, { "grad_norm": 0.3023340106010437, "learning_rate": 9.503780488906365e-05, "loss": 0.041, "step": 14870 }, { "grad_norm": 0.3582873046398163, "learning_rate": 9.502882424460319e-05, "loss": 0.0449, "step": 14880 }, { "grad_norm": 0.4048386514186859, "learning_rate": 9.501983590594821e-05, "loss": 0.0407, "step": 14890 }, { "grad_norm": 0.2806795537471771, "learning_rate": 9.501083987463455e-05, "loss": 0.0402, "step": 14900 }, { "grad_norm": 0.2773851752281189, "learning_rate": 9.500183615219942e-05, "loss": 0.0407, "step": 14910 }, { "grad_norm": 0.287642240524292, "learning_rate": 9.49928247401813e-05, "loss": 0.0414, "step": 14920 }, { "grad_norm": 0.3295131027698517, "learning_rate": 9.498380564011997e-05, "loss": 0.0378, "step": 14930 }, { "grad_norm": 0.25860241055488586, "learning_rate": 9.497477885355656e-05, "loss": 0.0452, "step": 14940 }, { "grad_norm": 0.34776929020881653, "learning_rate": 9.496574438203353e-05, "loss": 0.0475, "step": 14950 }, { "grad_norm": 0.3833771347999573, "learning_rate": 9.495670222709459e-05, "loss": 0.0499, "step": 14960 }, { "grad_norm": 0.3411467969417572, "learning_rate": 9.494765239028483e-05, "loss": 0.039, "step": 14970 }, { "grad_norm": 0.3360363841056824, "learning_rate": 9.493859487315057e-05, "loss": 0.042, "step": 14980 }, { "grad_norm": 0.26120465993881226, "learning_rate": 9.492952967723953e-05, "loss": 0.0418, "step": 14990 }, { "grad_norm": 0.319923996925354, "learning_rate": 9.492045680410068e-05, "loss": 0.0506, "step": 15000 }, { "grad_norm": 0.21544913947582245, "learning_rate": 9.491137625528436e-05, "loss": 0.038, "step": 15010 }, { "grad_norm": 0.3945525288581848, "learning_rate": 9.490228803234215e-05, "loss": 0.0413, "step": 15020 }, { "grad_norm": 0.4014047086238861, "learning_rate": 9.489319213682701e-05, "loss": 0.0396, "step": 15030 }, { "grad_norm": 0.29891613125801086, "learning_rate": 9.488408857029316e-05, "loss": 0.0537, "step": 15040 }, { "grad_norm": 0.3639589548110962, "learning_rate": 9.487497733429616e-05, "loss": 0.0525, "step": 15050 }, { "grad_norm": 0.33639347553253174, "learning_rate": 9.486585843039286e-05, "loss": 0.0495, "step": 15060 }, { "grad_norm": 0.2403927892446518, "learning_rate": 9.485673186014143e-05, "loss": 0.04, "step": 15070 }, { "grad_norm": 0.2683698534965515, "learning_rate": 9.484759762510137e-05, "loss": 0.0452, "step": 15080 }, { "grad_norm": 0.3253733813762665, "learning_rate": 9.483845572683346e-05, "loss": 0.0423, "step": 15090 }, { "grad_norm": 0.3159249722957611, "learning_rate": 9.48293061668998e-05, "loss": 0.0425, "step": 15100 }, { "grad_norm": 0.38768282532691956, "learning_rate": 9.48201489468638e-05, "loss": 0.042, "step": 15110 }, { "grad_norm": 0.287718802690506, "learning_rate": 9.481098406829016e-05, "loss": 0.0398, "step": 15120 }, { "grad_norm": 0.30348387360572815, "learning_rate": 9.480181153274495e-05, "loss": 0.0412, "step": 15130 }, { "grad_norm": 0.271451860666275, "learning_rate": 9.479263134179548e-05, "loss": 0.0352, "step": 15140 }, { "grad_norm": 0.39056557416915894, "learning_rate": 9.478344349701039e-05, "loss": 0.0494, "step": 15150 }, { "grad_norm": 0.3298358917236328, "learning_rate": 9.477424799995964e-05, "loss": 0.0479, "step": 15160 }, { "grad_norm": 0.28852927684783936, "learning_rate": 9.476504485221448e-05, "loss": 0.0499, "step": 15170 }, { "grad_norm": 0.302329957485199, "learning_rate": 9.475583405534748e-05, "loss": 0.0423, "step": 15180 }, { "grad_norm": 0.27220672369003296, "learning_rate": 9.474661561093251e-05, "loss": 0.0411, "step": 15190 }, { "grad_norm": 0.3942328691482544, "learning_rate": 9.473738952054478e-05, "loss": 0.049, "step": 15200 }, { "grad_norm": 0.35044366121292114, "learning_rate": 9.472815578576073e-05, "loss": 0.0434, "step": 15210 }, { "grad_norm": 0.430635541677475, "learning_rate": 9.471891440815817e-05, "loss": 0.0417, "step": 15220 }, { "grad_norm": 0.3710823953151703, "learning_rate": 9.470966538931621e-05, "loss": 0.0342, "step": 15230 }, { "grad_norm": 0.4162822961807251, "learning_rate": 9.470040873081525e-05, "loss": 0.0388, "step": 15240 }, { "grad_norm": 0.3947800397872925, "learning_rate": 9.469114443423698e-05, "loss": 0.0532, "step": 15250 }, { "grad_norm": 0.3617061674594879, "learning_rate": 9.468187250116445e-05, "loss": 0.0473, "step": 15260 }, { "grad_norm": 0.305177241563797, "learning_rate": 9.467259293318197e-05, "loss": 0.0468, "step": 15270 }, { "grad_norm": 0.4458027184009552, "learning_rate": 9.466330573187514e-05, "loss": 0.0465, "step": 15280 }, { "grad_norm": 0.3129308819770813, "learning_rate": 9.46540108988309e-05, "loss": 0.0517, "step": 15290 }, { "grad_norm": 0.32664456963539124, "learning_rate": 9.46447084356375e-05, "loss": 0.0452, "step": 15300 }, { "grad_norm": 0.33497846126556396, "learning_rate": 9.463539834388447e-05, "loss": 0.0501, "step": 15310 }, { "grad_norm": 0.3520512878894806, "learning_rate": 9.462608062516263e-05, "loss": 0.0376, "step": 15320 }, { "grad_norm": 0.32406896352767944, "learning_rate": 9.461675528106413e-05, "loss": 0.0379, "step": 15330 }, { "grad_norm": 0.29700323939323425, "learning_rate": 9.460742231318244e-05, "loss": 0.0366, "step": 15340 }, { "grad_norm": 0.3157297670841217, "learning_rate": 9.459808172311229e-05, "loss": 0.0465, "step": 15350 }, { "grad_norm": 0.32765424251556396, "learning_rate": 9.458873351244972e-05, "loss": 0.0381, "step": 15360 }, { "grad_norm": 0.32943296432495117, "learning_rate": 9.457937768279211e-05, "loss": 0.0482, "step": 15370 }, { "grad_norm": 0.2689002752304077, "learning_rate": 9.45700142357381e-05, "loss": 0.0401, "step": 15380 }, { "grad_norm": 0.330495685338974, "learning_rate": 9.456064317288765e-05, "loss": 0.0407, "step": 15390 }, { "grad_norm": 0.284259557723999, "learning_rate": 9.455126449584201e-05, "loss": 0.0427, "step": 15400 }, { "grad_norm": 0.2923649251461029, "learning_rate": 9.454187820620375e-05, "loss": 0.0393, "step": 15410 }, { "grad_norm": 0.3299538791179657, "learning_rate": 9.453248430557673e-05, "loss": 0.0434, "step": 15420 }, { "grad_norm": 0.26775500178337097, "learning_rate": 9.452308279556611e-05, "loss": 0.0615, "step": 15430 }, { "grad_norm": 0.35576796531677246, "learning_rate": 9.451367367777835e-05, "loss": 0.041, "step": 15440 }, { "grad_norm": 0.35054564476013184, "learning_rate": 9.450425695382122e-05, "loss": 0.0516, "step": 15450 }, { "grad_norm": 0.2752997577190399, "learning_rate": 9.449483262530375e-05, "loss": 0.0445, "step": 15460 }, { "grad_norm": 0.3264414072036743, "learning_rate": 9.448540069383633e-05, "loss": 0.0365, "step": 15470 }, { "grad_norm": 0.27788299322128296, "learning_rate": 9.447596116103061e-05, "loss": 0.0461, "step": 15480 }, { "grad_norm": 0.38429445028305054, "learning_rate": 9.446651402849955e-05, "loss": 0.0467, "step": 15490 }, { "grad_norm": 0.3142186105251312, "learning_rate": 9.44570592978574e-05, "loss": 0.0428, "step": 15500 }, { "grad_norm": 0.3411239683628082, "learning_rate": 9.444759697071972e-05, "loss": 0.0457, "step": 15510 }, { "grad_norm": 0.33642613887786865, "learning_rate": 9.443812704870336e-05, "loss": 0.0389, "step": 15520 }, { "grad_norm": 0.27719464898109436, "learning_rate": 9.442864953342649e-05, "loss": 0.0364, "step": 15530 }, { "grad_norm": 0.3371521830558777, "learning_rate": 9.441916442650852e-05, "loss": 0.0422, "step": 15540 }, { "grad_norm": 0.24962636828422546, "learning_rate": 9.440967172957023e-05, "loss": 0.0336, "step": 15550 }, { "grad_norm": 0.30904683470726013, "learning_rate": 9.440017144423364e-05, "loss": 0.0401, "step": 15560 }, { "grad_norm": 0.35357245802879333, "learning_rate": 9.439066357212209e-05, "loss": 0.0465, "step": 15570 }, { "grad_norm": 0.489637553691864, "learning_rate": 9.438114811486022e-05, "loss": 0.0501, "step": 15580 }, { "grad_norm": 0.22315093874931335, "learning_rate": 9.4371625074074e-05, "loss": 0.0388, "step": 15590 }, { "grad_norm": 0.35066142678260803, "learning_rate": 9.436209445139059e-05, "loss": 0.0403, "step": 15600 }, { "grad_norm": 0.43321576714515686, "learning_rate": 9.435255624843855e-05, "loss": 0.048, "step": 15610 }, { "grad_norm": 0.31790900230407715, "learning_rate": 9.43430104668477e-05, "loss": 0.0392, "step": 15620 }, { "grad_norm": 0.3145657777786255, "learning_rate": 9.433345710824914e-05, "loss": 0.0432, "step": 15630 }, { "grad_norm": 0.374864399433136, "learning_rate": 9.432389617427529e-05, "loss": 0.0438, "step": 15640 }, { "grad_norm": 0.32821089029312134, "learning_rate": 9.431432766655984e-05, "loss": 0.0442, "step": 15650 }, { "grad_norm": 0.3712776005268097, "learning_rate": 9.430475158673778e-05, "loss": 0.0496, "step": 15660 }, { "grad_norm": 0.31333181262016296, "learning_rate": 9.429516793644542e-05, "loss": 0.039, "step": 15670 }, { "grad_norm": 0.2951185703277588, "learning_rate": 9.428557671732034e-05, "loss": 0.0378, "step": 15680 }, { "grad_norm": 0.35491567850112915, "learning_rate": 9.42759779310014e-05, "loss": 0.0465, "step": 15690 }, { "grad_norm": 0.33582744002342224, "learning_rate": 9.426637157912879e-05, "loss": 0.0441, "step": 15700 }, { "grad_norm": 0.4237026870250702, "learning_rate": 9.425675766334397e-05, "loss": 0.0456, "step": 15710 }, { "grad_norm": 0.3069133460521698, "learning_rate": 9.424713618528968e-05, "loss": 0.0383, "step": 15720 }, { "grad_norm": 0.3006352484226227, "learning_rate": 9.423750714661e-05, "loss": 0.0454, "step": 15730 }, { "grad_norm": 0.3448623716831207, "learning_rate": 9.422787054895022e-05, "loss": 0.0385, "step": 15740 }, { "grad_norm": 0.3434298038482666, "learning_rate": 9.4218226393957e-05, "loss": 0.0518, "step": 15750 }, { "grad_norm": 0.37644752860069275, "learning_rate": 9.420857468327828e-05, "loss": 0.0442, "step": 15760 }, { "grad_norm": 0.266934335231781, "learning_rate": 9.419891541856323e-05, "loss": 0.0504, "step": 15770 }, { "grad_norm": 0.3617183268070221, "learning_rate": 9.41892486014624e-05, "loss": 0.0483, "step": 15780 }, { "grad_norm": 0.37142816185951233, "learning_rate": 9.417957423362756e-05, "loss": 0.0514, "step": 15790 }, { "grad_norm": 0.32222607731819153, "learning_rate": 9.416989231671178e-05, "loss": 0.0483, "step": 15800 }, { "grad_norm": 0.28842976689338684, "learning_rate": 9.416020285236946e-05, "loss": 0.0443, "step": 15810 }, { "grad_norm": 0.3107130527496338, "learning_rate": 9.415050584225626e-05, "loss": 0.0428, "step": 15820 }, { "grad_norm": 0.3208422362804413, "learning_rate": 9.414080128802914e-05, "loss": 0.0397, "step": 15830 }, { "grad_norm": 0.27449846267700195, "learning_rate": 9.413108919134632e-05, "loss": 0.0347, "step": 15840 }, { "grad_norm": 0.30793285369873047, "learning_rate": 9.412136955386734e-05, "loss": 0.0405, "step": 15850 }, { "grad_norm": 0.3974357545375824, "learning_rate": 9.411164237725303e-05, "loss": 0.043, "step": 15860 }, { "grad_norm": 0.26995158195495605, "learning_rate": 9.41019076631655e-05, "loss": 0.0427, "step": 15870 }, { "grad_norm": 0.37107595801353455, "learning_rate": 9.409216541326815e-05, "loss": 0.0421, "step": 15880 }, { "grad_norm": 0.23414525389671326, "learning_rate": 9.408241562922564e-05, "loss": 0.0447, "step": 15890 }, { "grad_norm": 0.3388262987136841, "learning_rate": 9.407265831270395e-05, "loss": 0.0383, "step": 15900 }, { "grad_norm": 0.279140830039978, "learning_rate": 9.406289346537035e-05, "loss": 0.0411, "step": 15910 }, { "grad_norm": 0.3506709337234497, "learning_rate": 9.405312108889339e-05, "loss": 0.0462, "step": 15920 }, { "grad_norm": 0.307807981967926, "learning_rate": 9.404334118494288e-05, "loss": 0.0499, "step": 15930 }, { "grad_norm": 0.32523009181022644, "learning_rate": 9.403355375518995e-05, "loss": 0.0427, "step": 15940 }, { "grad_norm": 0.3811091184616089, "learning_rate": 9.4023758801307e-05, "loss": 0.0385, "step": 15950 }, { "grad_norm": 0.5486855506896973, "learning_rate": 9.401395632496774e-05, "loss": 0.0427, "step": 15960 }, { "grad_norm": 0.2504093050956726, "learning_rate": 9.400414632784711e-05, "loss": 0.0484, "step": 15970 }, { "grad_norm": 0.280328631401062, "learning_rate": 9.39943288116214e-05, "loss": 0.043, "step": 15980 }, { "grad_norm": 0.38578304648399353, "learning_rate": 9.398450377796815e-05, "loss": 0.05, "step": 15990 }, { "grad_norm": 0.4229767918586731, "learning_rate": 9.397467122856616e-05, "loss": 0.0438, "step": 16000 }, { "grad_norm": 0.3467263877391815, "learning_rate": 9.396483116509558e-05, "loss": 0.0432, "step": 16010 }, { "grad_norm": 0.37100890278816223, "learning_rate": 9.39549835892378e-05, "loss": 0.0378, "step": 16020 }, { "grad_norm": 0.3561440706253052, "learning_rate": 9.39451285026755e-05, "loss": 0.0426, "step": 16030 }, { "grad_norm": 0.47684475779533386, "learning_rate": 9.393526590709262e-05, "loss": 0.0441, "step": 16040 }, { "grad_norm": 0.3566969633102417, "learning_rate": 9.392539580417444e-05, "loss": 0.0432, "step": 16050 }, { "grad_norm": 0.36831268668174744, "learning_rate": 9.391551819560747e-05, "loss": 0.0431, "step": 16060 }, { "grad_norm": 0.2806608974933624, "learning_rate": 9.390563308307955e-05, "loss": 0.0435, "step": 16070 }, { "grad_norm": 0.3070192337036133, "learning_rate": 9.389574046827974e-05, "loss": 0.0447, "step": 16080 }, { "grad_norm": 0.34824225306510925, "learning_rate": 9.388584035289845e-05, "loss": 0.0466, "step": 16090 }, { "grad_norm": 0.41158929467201233, "learning_rate": 9.387593273862732e-05, "loss": 0.0511, "step": 16100 }, { "grad_norm": 0.31697893142700195, "learning_rate": 9.386601762715929e-05, "loss": 0.0463, "step": 16110 }, { "grad_norm": 0.39503249526023865, "learning_rate": 9.38560950201886e-05, "loss": 0.0493, "step": 16120 }, { "grad_norm": 0.5879943370819092, "learning_rate": 9.384616491941071e-05, "loss": 0.0432, "step": 16130 }, { "grad_norm": 0.3558208644390106, "learning_rate": 9.383622732652245e-05, "loss": 0.0361, "step": 16140 }, { "grad_norm": 0.29379522800445557, "learning_rate": 9.382628224322187e-05, "loss": 0.0468, "step": 16150 }, { "grad_norm": 0.3380815386772156, "learning_rate": 9.381632967120829e-05, "loss": 0.0392, "step": 16160 }, { "grad_norm": 0.38068342208862305, "learning_rate": 9.380636961218235e-05, "loss": 0.0418, "step": 16170 }, { "grad_norm": 0.28796544671058655, "learning_rate": 9.379640206784597e-05, "loss": 0.0373, "step": 16180 }, { "grad_norm": 0.36596444249153137, "learning_rate": 9.378642703990229e-05, "loss": 0.0404, "step": 16190 }, { "grad_norm": 0.3039620816707611, "learning_rate": 9.37764445300558e-05, "loss": 0.0452, "step": 16200 }, { "grad_norm": 0.2615838348865509, "learning_rate": 9.376645454001222e-05, "loss": 0.0473, "step": 16210 }, { "grad_norm": 0.28700029850006104, "learning_rate": 9.375645707147858e-05, "loss": 0.0379, "step": 16220 }, { "grad_norm": 0.3078685402870178, "learning_rate": 9.374645212616316e-05, "loss": 0.0462, "step": 16230 }, { "grad_norm": 0.3243401348590851, "learning_rate": 9.373643970577555e-05, "loss": 0.0416, "step": 16240 }, { "grad_norm": 0.36983856558799744, "learning_rate": 9.372641981202659e-05, "loss": 0.0413, "step": 16250 }, { "grad_norm": 0.41347914934158325, "learning_rate": 9.37163924466284e-05, "loss": 0.0477, "step": 16260 }, { "grad_norm": 0.3030194342136383, "learning_rate": 9.370635761129438e-05, "loss": 0.0413, "step": 16270 }, { "grad_norm": 0.31857091188430786, "learning_rate": 9.36963153077392e-05, "loss": 0.0465, "step": 16280 }, { "grad_norm": 0.32468968629837036, "learning_rate": 9.368626553767888e-05, "loss": 0.0418, "step": 16290 }, { "grad_norm": 0.3416540026664734, "learning_rate": 9.367620830283057e-05, "loss": 0.0338, "step": 16300 }, { "grad_norm": 0.3311488926410675, "learning_rate": 9.366614360491281e-05, "loss": 0.0443, "step": 16310 }, { "grad_norm": 0.35677793622016907, "learning_rate": 9.365607144564539e-05, "loss": 0.0407, "step": 16320 }, { "grad_norm": 0.2849990427494049, "learning_rate": 9.364599182674934e-05, "loss": 0.0426, "step": 16330 }, { "grad_norm": 0.29887494444847107, "learning_rate": 9.3635904749947e-05, "loss": 0.0371, "step": 16340 }, { "grad_norm": 0.33544474840164185, "learning_rate": 9.362581021696202e-05, "loss": 0.0349, "step": 16350 }, { "grad_norm": 0.35628899931907654, "learning_rate": 9.361570822951921e-05, "loss": 0.0445, "step": 16360 }, { "grad_norm": 0.2582319378852844, "learning_rate": 9.360559878934476e-05, "loss": 0.0546, "step": 16370 }, { "grad_norm": 0.33235809206962585, "learning_rate": 9.359548189816611e-05, "loss": 0.0384, "step": 16380 }, { "grad_norm": 0.34933796525001526, "learning_rate": 9.358535755771193e-05, "loss": 0.0455, "step": 16390 }, { "grad_norm": 0.32721686363220215, "learning_rate": 9.357522576971221e-05, "loss": 0.0468, "step": 16400 }, { "grad_norm": 0.2630440294742584, "learning_rate": 9.356508653589819e-05, "loss": 0.0424, "step": 16410 }, { "grad_norm": 0.28859275579452515, "learning_rate": 9.355493985800237e-05, "loss": 0.0451, "step": 16420 }, { "grad_norm": 0.3209753930568695, "learning_rate": 9.354478573775857e-05, "loss": 0.0471, "step": 16430 }, { "grad_norm": 0.31036636233329773, "learning_rate": 9.353462417690186e-05, "loss": 0.044, "step": 16440 }, { "grad_norm": 0.3691061735153198, "learning_rate": 9.352445517716853e-05, "loss": 0.0378, "step": 16450 }, { "grad_norm": 0.31185129284858704, "learning_rate": 9.351427874029621e-05, "loss": 0.0403, "step": 16460 }, { "grad_norm": 0.3619175851345062, "learning_rate": 9.350409486802379e-05, "loss": 0.0376, "step": 16470 }, { "grad_norm": 0.38652369379997253, "learning_rate": 9.349390356209138e-05, "loss": 0.0529, "step": 16480 }, { "grad_norm": 0.30334845185279846, "learning_rate": 9.348370482424042e-05, "loss": 0.047, "step": 16490 }, { "grad_norm": 0.37459704279899597, "learning_rate": 9.347349865621357e-05, "loss": 0.0413, "step": 16500 }, { "grad_norm": 0.3510027527809143, "learning_rate": 9.346328505975481e-05, "loss": 0.0363, "step": 16510 }, { "grad_norm": 0.24696074426174164, "learning_rate": 9.345306403660936e-05, "loss": 0.0438, "step": 16520 }, { "grad_norm": 0.4397953152656555, "learning_rate": 9.344283558852371e-05, "loss": 0.044, "step": 16530 }, { "grad_norm": 0.3134211003780365, "learning_rate": 9.343259971724563e-05, "loss": 0.0383, "step": 16540 }, { "grad_norm": 0.29789426922798157, "learning_rate": 9.342235642452413e-05, "loss": 0.0411, "step": 16550 }, { "grad_norm": 0.3650570511817932, "learning_rate": 9.341210571210954e-05, "loss": 0.0472, "step": 16560 }, { "grad_norm": 0.2893986105918884, "learning_rate": 9.340184758175338e-05, "loss": 0.0389, "step": 16570 }, { "grad_norm": 0.24981795251369476, "learning_rate": 9.339158203520854e-05, "loss": 0.0443, "step": 16580 }, { "grad_norm": 0.21025384962558746, "learning_rate": 9.338130907422908e-05, "loss": 0.0339, "step": 16590 }, { "grad_norm": 0.2885726988315582, "learning_rate": 9.337102870057037e-05, "loss": 0.0335, "step": 16600 }, { "grad_norm": 0.2743820250034332, "learning_rate": 9.336074091598907e-05, "loss": 0.0362, "step": 16610 }, { "grad_norm": 0.25782448053359985, "learning_rate": 9.335044572224306e-05, "loss": 0.0307, "step": 16620 }, { "grad_norm": 0.3034486174583435, "learning_rate": 9.334014312109151e-05, "loss": 0.0332, "step": 16630 }, { "grad_norm": 0.4149041473865509, "learning_rate": 9.332983311429486e-05, "loss": 0.0437, "step": 16640 }, { "grad_norm": 0.3128279149532318, "learning_rate": 9.33195157036148e-05, "loss": 0.0384, "step": 16650 }, { "grad_norm": 0.34412282705307007, "learning_rate": 9.330919089081432e-05, "loss": 0.0368, "step": 16660 }, { "grad_norm": 0.30552423000335693, "learning_rate": 9.32988586776576e-05, "loss": 0.0596, "step": 16670 }, { "grad_norm": 0.4439716935157776, "learning_rate": 9.328851906591016e-05, "loss": 0.0358, "step": 16680 }, { "grad_norm": 0.2988627552986145, "learning_rate": 9.327817205733875e-05, "loss": 0.0422, "step": 16690 }, { "grad_norm": 0.2662118077278137, "learning_rate": 9.326781765371142e-05, "loss": 0.0346, "step": 16700 }, { "grad_norm": 0.3346235454082489, "learning_rate": 9.325745585679741e-05, "loss": 0.0402, "step": 16710 }, { "grad_norm": 0.2630904018878937, "learning_rate": 9.32470866683673e-05, "loss": 0.0512, "step": 16720 }, { "grad_norm": 0.34601593017578125, "learning_rate": 9.323671009019288e-05, "loss": 0.0514, "step": 16730 }, { "grad_norm": 0.3513593375682831, "learning_rate": 9.322632612404725e-05, "loss": 0.0371, "step": 16740 }, { "grad_norm": 0.290382444858551, "learning_rate": 9.321593477170471e-05, "loss": 0.0478, "step": 16750 }, { "grad_norm": 0.3042750358581543, "learning_rate": 9.320553603494088e-05, "loss": 0.0389, "step": 16760 }, { "grad_norm": 0.38500306010246277, "learning_rate": 9.319512991553261e-05, "loss": 0.0396, "step": 16770 }, { "grad_norm": 0.2805958688259125, "learning_rate": 9.318471641525803e-05, "loss": 0.0415, "step": 16780 }, { "grad_norm": 0.2956908345222473, "learning_rate": 9.317429553589652e-05, "loss": 0.0419, "step": 16790 }, { "grad_norm": 0.3236039876937866, "learning_rate": 9.316386727922873e-05, "loss": 0.0437, "step": 16800 }, { "grad_norm": 0.36326873302459717, "learning_rate": 9.315343164703656e-05, "loss": 0.0396, "step": 16810 }, { "grad_norm": 0.3245916962623596, "learning_rate": 9.314298864110316e-05, "loss": 0.041, "step": 16820 }, { "grad_norm": 0.4046924412250519, "learning_rate": 9.313253826321295e-05, "loss": 0.0393, "step": 16830 }, { "grad_norm": 0.27797505259513855, "learning_rate": 9.312208051515165e-05, "loss": 0.0371, "step": 16840 }, { "grad_norm": 0.3647770881652832, "learning_rate": 9.311161539870618e-05, "loss": 0.0409, "step": 16850 }, { "grad_norm": 0.24435853958129883, "learning_rate": 9.310114291566474e-05, "loss": 0.0438, "step": 16860 }, { "grad_norm": 0.3824923634529114, "learning_rate": 9.309066306781679e-05, "loss": 0.0327, "step": 16870 }, { "grad_norm": 0.2864387631416321, "learning_rate": 9.308017585695306e-05, "loss": 0.037, "step": 16880 }, { "grad_norm": 0.3126680552959442, "learning_rate": 9.306968128486552e-05, "loss": 0.0425, "step": 16890 }, { "grad_norm": 0.35515186190605164, "learning_rate": 9.30591793533474e-05, "loss": 0.0349, "step": 16900 }, { "grad_norm": 0.30528730154037476, "learning_rate": 9.304867006419321e-05, "loss": 0.036, "step": 16910 }, { "grad_norm": 0.26310795545578003, "learning_rate": 9.303815341919868e-05, "loss": 0.0378, "step": 16920 }, { "grad_norm": 0.3007355332374573, "learning_rate": 9.302762942016084e-05, "loss": 0.044, "step": 16930 }, { "grad_norm": 0.3377184271812439, "learning_rate": 9.301709806887792e-05, "loss": 0.0401, "step": 16940 }, { "grad_norm": 0.3938685357570648, "learning_rate": 9.300655936714948e-05, "loss": 0.0472, "step": 16950 }, { "grad_norm": 0.3267129361629486, "learning_rate": 9.299601331677627e-05, "loss": 0.0384, "step": 16960 }, { "grad_norm": 0.5238194465637207, "learning_rate": 9.298545991956033e-05, "loss": 0.0488, "step": 16970 }, { "grad_norm": 0.2898818552494049, "learning_rate": 9.297489917730493e-05, "loss": 0.0464, "step": 16980 }, { "grad_norm": 0.4203214943408966, "learning_rate": 9.296433109181464e-05, "loss": 0.0466, "step": 16990 }, { "grad_norm": 0.3649360239505768, "learning_rate": 9.295375566489523e-05, "loss": 0.0461, "step": 17000 }, { "grad_norm": 0.3542673885822296, "learning_rate": 9.294317289835379e-05, "loss": 0.0508, "step": 17010 }, { "grad_norm": 0.4557419419288635, "learning_rate": 9.293258279399859e-05, "loss": 0.0524, "step": 17020 }, { "grad_norm": 0.3722783327102661, "learning_rate": 9.292198535363919e-05, "loss": 0.0422, "step": 17030 }, { "grad_norm": 0.2995437681674957, "learning_rate": 9.291138057908641e-05, "loss": 0.0376, "step": 17040 }, { "grad_norm": 0.3678897023200989, "learning_rate": 9.290076847215234e-05, "loss": 0.038, "step": 17050 }, { "grad_norm": 0.4264276325702667, "learning_rate": 9.289014903465025e-05, "loss": 0.0532, "step": 17060 }, { "grad_norm": 0.3346143662929535, "learning_rate": 9.287952226839475e-05, "loss": 0.0342, "step": 17070 }, { "grad_norm": 0.31795141100883484, "learning_rate": 9.286888817520164e-05, "loss": 0.0416, "step": 17080 }, { "grad_norm": 0.3396551012992859, "learning_rate": 9.285824675688803e-05, "loss": 0.0433, "step": 17090 }, { "grad_norm": 0.2585776448249817, "learning_rate": 9.28475980152722e-05, "loss": 0.0413, "step": 17100 }, { "grad_norm": 0.27303385734558105, "learning_rate": 9.283694195217379e-05, "loss": 0.0438, "step": 17110 }, { "grad_norm": 0.4269823133945465, "learning_rate": 9.282627856941356e-05, "loss": 0.0411, "step": 17120 }, { "grad_norm": 0.37981244921684265, "learning_rate": 9.281560786881363e-05, "loss": 0.0444, "step": 17130 }, { "grad_norm": 0.35399892926216125, "learning_rate": 9.280492985219733e-05, "loss": 0.0409, "step": 17140 }, { "grad_norm": 0.32138457894325256, "learning_rate": 9.279424452138924e-05, "loss": 0.0422, "step": 17150 }, { "grad_norm": 0.38279688358306885, "learning_rate": 9.278355187821517e-05, "loss": 0.0426, "step": 17160 }, { "grad_norm": 0.3348556458950043, "learning_rate": 9.277285192450224e-05, "loss": 0.0468, "step": 17170 }, { "grad_norm": 0.39661622047424316, "learning_rate": 9.276214466207875e-05, "loss": 0.0408, "step": 17180 }, { "grad_norm": 0.4208698868751526, "learning_rate": 9.275143009277427e-05, "loss": 0.0437, "step": 17190 }, { "grad_norm": 0.3779715597629547, "learning_rate": 9.274070821841964e-05, "loss": 0.0418, "step": 17200 }, { "grad_norm": 0.3530980050563812, "learning_rate": 9.272997904084696e-05, "loss": 0.0399, "step": 17210 }, { "grad_norm": 0.3079186677932739, "learning_rate": 9.271924256188951e-05, "loss": 0.0366, "step": 17220 }, { "grad_norm": 0.3134459853172302, "learning_rate": 9.270849878338189e-05, "loss": 0.0403, "step": 17230 }, { "grad_norm": 0.2943434715270996, "learning_rate": 9.269774770715991e-05, "loss": 0.0394, "step": 17240 }, { "grad_norm": 0.3075118362903595, "learning_rate": 9.268698933506061e-05, "loss": 0.0499, "step": 17250 }, { "grad_norm": 0.32588446140289307, "learning_rate": 9.267622366892235e-05, "loss": 0.0403, "step": 17260 }, { "grad_norm": 0.2994861304759979, "learning_rate": 9.266545071058465e-05, "loss": 0.0416, "step": 17270 }, { "grad_norm": 0.43113797903060913, "learning_rate": 9.265467046188833e-05, "loss": 0.0351, "step": 17280 }, { "grad_norm": 0.2980540096759796, "learning_rate": 9.264388292467543e-05, "loss": 0.0472, "step": 17290 }, { "grad_norm": 0.3276672065258026, "learning_rate": 9.263308810078926e-05, "loss": 0.0418, "step": 17300 }, { "grad_norm": 0.23057927191257477, "learning_rate": 9.262228599207434e-05, "loss": 0.0368, "step": 17310 }, { "grad_norm": 0.2923731207847595, "learning_rate": 9.261147660037647e-05, "loss": 0.033, "step": 17320 }, { "grad_norm": 0.28963416814804077, "learning_rate": 9.26006599275427e-05, "loss": 0.0371, "step": 17330 }, { "grad_norm": 0.36166754364967346, "learning_rate": 9.258983597542124e-05, "loss": 0.047, "step": 17340 }, { "grad_norm": 0.22982488572597504, "learning_rate": 9.257900474586167e-05, "loss": 0.038, "step": 17350 }, { "grad_norm": 0.2528805732727051, "learning_rate": 9.256816624071471e-05, "loss": 0.0379, "step": 17360 }, { "grad_norm": 0.3232249319553375, "learning_rate": 9.25573204618324e-05, "loss": 0.0412, "step": 17370 }, { "grad_norm": 0.23439785838127136, "learning_rate": 9.254646741106796e-05, "loss": 0.0371, "step": 17380 }, { "grad_norm": 0.229953333735466, "learning_rate": 9.253560709027589e-05, "loss": 0.0326, "step": 17390 }, { "grad_norm": 0.42876893281936646, "learning_rate": 9.252473950131192e-05, "loss": 0.0452, "step": 17400 }, { "grad_norm": 0.26708275079727173, "learning_rate": 9.251386464603302e-05, "loss": 0.0337, "step": 17410 }, { "grad_norm": 0.4327963888645172, "learning_rate": 9.250298252629741e-05, "loss": 0.0447, "step": 17420 }, { "grad_norm": 0.4835042655467987, "learning_rate": 9.249209314396454e-05, "loss": 0.0374, "step": 17430 }, { "grad_norm": 0.26533040404319763, "learning_rate": 9.248119650089513e-05, "loss": 0.0422, "step": 17440 }, { "grad_norm": 0.34249347448349, "learning_rate": 9.247029259895108e-05, "loss": 0.0409, "step": 17450 }, { "grad_norm": 0.481590211391449, "learning_rate": 9.24593814399956e-05, "loss": 0.0473, "step": 17460 }, { "grad_norm": 0.33240070939064026, "learning_rate": 9.244846302589309e-05, "loss": 0.0409, "step": 17470 }, { "grad_norm": 0.32594406604766846, "learning_rate": 9.243753735850923e-05, "loss": 0.0439, "step": 17480 }, { "grad_norm": 0.3304900527000427, "learning_rate": 9.24266044397109e-05, "loss": 0.0387, "step": 17490 }, { "grad_norm": 0.3681481182575226, "learning_rate": 9.241566427136624e-05, "loss": 0.036, "step": 17500 }, { "grad_norm": 0.2898999750614166, "learning_rate": 9.240471685534463e-05, "loss": 0.0383, "step": 17510 }, { "grad_norm": 0.27274152636528015, "learning_rate": 9.239376219351667e-05, "loss": 0.0407, "step": 17520 }, { "grad_norm": 0.2940172553062439, "learning_rate": 9.238280028775425e-05, "loss": 0.0383, "step": 17530 }, { "grad_norm": 0.27539706230163574, "learning_rate": 9.237183113993041e-05, "loss": 0.0444, "step": 17540 }, { "grad_norm": 0.24695193767547607, "learning_rate": 9.236085475191952e-05, "loss": 0.034, "step": 17550 }, { "grad_norm": 0.3609212636947632, "learning_rate": 9.234987112559709e-05, "loss": 0.052, "step": 17560 }, { "grad_norm": 0.28479793667793274, "learning_rate": 9.233888026283999e-05, "loss": 0.0434, "step": 17570 }, { "grad_norm": 0.3835596442222595, "learning_rate": 9.232788216552619e-05, "loss": 0.0448, "step": 17580 }, { "grad_norm": 0.2775576710700989, "learning_rate": 9.231687683553502e-05, "loss": 0.043, "step": 17590 }, { "grad_norm": 0.4775995910167694, "learning_rate": 9.230586427474698e-05, "loss": 0.0544, "step": 17600 }, { "grad_norm": 0.298453152179718, "learning_rate": 9.229484448504379e-05, "loss": 0.0452, "step": 17610 }, { "grad_norm": 0.344101220369339, "learning_rate": 9.228381746830843e-05, "loss": 0.0397, "step": 17620 }, { "grad_norm": 0.28978899121284485, "learning_rate": 9.227278322642514e-05, "loss": 0.0391, "step": 17630 }, { "grad_norm": 0.3306983709335327, "learning_rate": 9.226174176127937e-05, "loss": 0.0397, "step": 17640 }, { "grad_norm": 0.2195616513490677, "learning_rate": 9.22506930747578e-05, "loss": 0.0345, "step": 17650 }, { "grad_norm": 0.3581613600254059, "learning_rate": 9.223963716874831e-05, "loss": 0.0455, "step": 17660 }, { "grad_norm": 0.2639351487159729, "learning_rate": 9.222857404514012e-05, "loss": 0.0372, "step": 17670 }, { "grad_norm": 0.3453105688095093, "learning_rate": 9.221750370582355e-05, "loss": 0.0414, "step": 17680 }, { "grad_norm": 0.29965171217918396, "learning_rate": 9.220642615269028e-05, "loss": 0.0372, "step": 17690 }, { "grad_norm": 0.29599910974502563, "learning_rate": 9.219534138763311e-05, "loss": 0.0419, "step": 17700 }, { "grad_norm": 0.28672587871551514, "learning_rate": 9.218424941254613e-05, "loss": 0.045, "step": 17710 }, { "grad_norm": 0.4353128671646118, "learning_rate": 9.217315022932468e-05, "loss": 0.0413, "step": 17720 }, { "grad_norm": 0.3842564821243286, "learning_rate": 9.216204383986528e-05, "loss": 0.0423, "step": 17730 }, { "grad_norm": 0.3143433630466461, "learning_rate": 9.215093024606574e-05, "loss": 0.0498, "step": 17740 }, { "grad_norm": 0.41524070501327515, "learning_rate": 9.213980944982506e-05, "loss": 0.0583, "step": 17750 }, { "grad_norm": 0.3129536807537079, "learning_rate": 9.212868145304346e-05, "loss": 0.0438, "step": 17760 }, { "grad_norm": 0.38175612688064575, "learning_rate": 9.211754625762241e-05, "loss": 0.0384, "step": 17770 }, { "grad_norm": 0.47345566749572754, "learning_rate": 9.210640386546463e-05, "loss": 0.0478, "step": 17780 }, { "grad_norm": 0.3057158887386322, "learning_rate": 9.209525427847405e-05, "loss": 0.0318, "step": 17790 }, { "grad_norm": 0.3572303056716919, "learning_rate": 9.208409749855583e-05, "loss": 0.0387, "step": 17800 }, { "grad_norm": 0.2947794198989868, "learning_rate": 9.207293352761633e-05, "loss": 0.0421, "step": 17810 }, { "grad_norm": 0.5777736306190491, "learning_rate": 9.206176236756319e-05, "loss": 0.0412, "step": 17820 }, { "grad_norm": 0.2769997715950012, "learning_rate": 9.205058402030525e-05, "loss": 0.0411, "step": 17830 }, { "grad_norm": 0.3976726233959198, "learning_rate": 9.203939848775259e-05, "loss": 0.0351, "step": 17840 }, { "grad_norm": 0.2704472243785858, "learning_rate": 9.202820577181652e-05, "loss": 0.0353, "step": 17850 }, { "grad_norm": 0.2702574133872986, "learning_rate": 9.201700587440953e-05, "loss": 0.0383, "step": 17860 }, { "grad_norm": 0.3391551375389099, "learning_rate": 9.200579879744544e-05, "loss": 0.0385, "step": 17870 }, { "grad_norm": 0.32895052433013916, "learning_rate": 9.199458454283918e-05, "loss": 0.0383, "step": 17880 }, { "grad_norm": 0.25900256633758545, "learning_rate": 9.198336311250697e-05, "loss": 0.0413, "step": 17890 }, { "grad_norm": 0.40781858563423157, "learning_rate": 9.197213450836626e-05, "loss": 0.0339, "step": 17900 }, { "grad_norm": 0.2940504550933838, "learning_rate": 9.19608987323357e-05, "loss": 0.0386, "step": 17910 }, { "grad_norm": 0.40177515149116516, "learning_rate": 9.194965578633517e-05, "loss": 0.032, "step": 17920 }, { "grad_norm": 0.21601904928684235, "learning_rate": 9.193840567228582e-05, "loss": 0.0381, "step": 17930 }, { "grad_norm": 0.34659403562545776, "learning_rate": 9.192714839210994e-05, "loss": 0.0349, "step": 17940 }, { "grad_norm": 0.3357577919960022, "learning_rate": 9.19158839477311e-05, "loss": 0.0449, "step": 17950 }, { "grad_norm": 0.3625031113624573, "learning_rate": 9.190461234107411e-05, "loss": 0.0441, "step": 17960 }, { "grad_norm": 0.3140789866447449, "learning_rate": 9.189333357406496e-05, "loss": 0.0356, "step": 17970 }, { "grad_norm": 0.2656749486923218, "learning_rate": 9.188204764863089e-05, "loss": 0.0471, "step": 17980 }, { "grad_norm": 0.2835037410259247, "learning_rate": 9.187075456670033e-05, "loss": 0.0331, "step": 17990 }, { "grad_norm": 0.29655033349990845, "learning_rate": 9.1859454330203e-05, "loss": 0.0417, "step": 18000 }, { "grad_norm": 0.26779094338417053, "learning_rate": 9.18481469410698e-05, "loss": 0.0481, "step": 18010 }, { "grad_norm": 0.3321940302848816, "learning_rate": 9.183683240123281e-05, "loss": 0.0367, "step": 18020 }, { "grad_norm": 0.2546979486942291, "learning_rate": 9.182551071262541e-05, "loss": 0.0487, "step": 18030 }, { "grad_norm": 0.33776021003723145, "learning_rate": 9.181418187718218e-05, "loss": 0.0453, "step": 18040 }, { "grad_norm": 0.4099810719490051, "learning_rate": 9.180284589683888e-05, "loss": 0.0387, "step": 18050 }, { "grad_norm": 0.26497870683670044, "learning_rate": 9.17915027735325e-05, "loss": 0.0378, "step": 18060 }, { "grad_norm": 0.29518091678619385, "learning_rate": 9.178015250920133e-05, "loss": 0.035, "step": 18070 }, { "grad_norm": 0.3782236576080322, "learning_rate": 9.176879510578477e-05, "loss": 0.0378, "step": 18080 }, { "grad_norm": 0.26639795303344727, "learning_rate": 9.17574305652235e-05, "loss": 0.0421, "step": 18090 }, { "grad_norm": 0.41015130281448364, "learning_rate": 9.174605888945942e-05, "loss": 0.0423, "step": 18100 }, { "grad_norm": 0.2721092104911804, "learning_rate": 9.173468008043564e-05, "loss": 0.0376, "step": 18110 }, { "grad_norm": 0.409267395734787, "learning_rate": 9.172329414009648e-05, "loss": 0.0441, "step": 18120 }, { "grad_norm": 0.32053399085998535, "learning_rate": 9.171190107038747e-05, "loss": 0.0423, "step": 18130 }, { "grad_norm": 0.31938764452934265, "learning_rate": 9.170050087325541e-05, "loss": 0.034, "step": 18140 }, { "grad_norm": 0.21190781891345978, "learning_rate": 9.168909355064824e-05, "loss": 0.0389, "step": 18150 }, { "grad_norm": 0.2944002151489258, "learning_rate": 9.167767910451519e-05, "loss": 0.0364, "step": 18160 }, { "grad_norm": 0.3137376010417938, "learning_rate": 9.166625753680669e-05, "loss": 0.0372, "step": 18170 }, { "grad_norm": 0.2770487368106842, "learning_rate": 9.165482884947431e-05, "loss": 0.032, "step": 18180 }, { "grad_norm": 0.2542206346988678, "learning_rate": 9.164339304447098e-05, "loss": 0.039, "step": 18190 }, { "grad_norm": 0.2584398090839386, "learning_rate": 9.163195012375072e-05, "loss": 0.0283, "step": 18200 }, { "grad_norm": 0.29365304112434387, "learning_rate": 9.16205000892688e-05, "loss": 0.0365, "step": 18210 }, { "grad_norm": 0.40694424510002136, "learning_rate": 9.160904294298175e-05, "loss": 0.0326, "step": 18220 }, { "grad_norm": 0.30803704261779785, "learning_rate": 9.159757868684727e-05, "loss": 0.0385, "step": 18230 }, { "grad_norm": 0.42075595259666443, "learning_rate": 9.15861073228243e-05, "loss": 0.0408, "step": 18240 }, { "grad_norm": 0.3687211573123932, "learning_rate": 9.157462885287296e-05, "loss": 0.0446, "step": 18250 }, { "grad_norm": 0.2588743269443512, "learning_rate": 9.156314327895461e-05, "loss": 0.0352, "step": 18260 }, { "grad_norm": 0.27652522921562195, "learning_rate": 9.155165060303185e-05, "loss": 0.0413, "step": 18270 }, { "grad_norm": 0.27205201983451843, "learning_rate": 9.154015082706841e-05, "loss": 0.0347, "step": 18280 }, { "grad_norm": 0.324604868888855, "learning_rate": 9.152864395302936e-05, "loss": 0.0364, "step": 18290 }, { "grad_norm": 0.32145318388938904, "learning_rate": 9.151712998288085e-05, "loss": 0.0357, "step": 18300 }, { "grad_norm": 0.24692341685295105, "learning_rate": 9.150560891859031e-05, "loss": 0.0413, "step": 18310 }, { "grad_norm": 0.3380054831504822, "learning_rate": 9.14940807621264e-05, "loss": 0.0453, "step": 18320 }, { "grad_norm": 0.35342469811439514, "learning_rate": 9.148254551545894e-05, "loss": 0.0366, "step": 18330 }, { "grad_norm": 0.3021315038204193, "learning_rate": 9.147100318055901e-05, "loss": 0.0433, "step": 18340 }, { "grad_norm": 0.4389913082122803, "learning_rate": 9.145945375939888e-05, "loss": 0.0401, "step": 18350 }, { "grad_norm": 0.34006887674331665, "learning_rate": 9.144789725395203e-05, "loss": 0.0385, "step": 18360 }, { "grad_norm": 0.3935396671295166, "learning_rate": 9.14363336661931e-05, "loss": 0.0348, "step": 18370 }, { "grad_norm": 0.30712956190109253, "learning_rate": 9.142476299809806e-05, "loss": 0.0416, "step": 18380 }, { "grad_norm": 0.3084118962287903, "learning_rate": 9.1413185251644e-05, "loss": 0.036, "step": 18390 }, { "grad_norm": 0.46003326773643494, "learning_rate": 9.140160042880923e-05, "loss": 0.0481, "step": 18400 }, { "grad_norm": 0.3663806617259979, "learning_rate": 9.139000853157327e-05, "loss": 0.0435, "step": 18410 }, { "grad_norm": 0.29877445101737976, "learning_rate": 9.137840956191688e-05, "loss": 0.041, "step": 18420 }, { "grad_norm": 0.2766457498073578, "learning_rate": 9.136680352182199e-05, "loss": 0.0366, "step": 18430 }, { "grad_norm": 0.254743754863739, "learning_rate": 9.135519041327177e-05, "loss": 0.0374, "step": 18440 }, { "grad_norm": 0.30507656931877136, "learning_rate": 9.134357023825058e-05, "loss": 0.048, "step": 18450 }, { "grad_norm": 0.3879513144493103, "learning_rate": 9.133194299874398e-05, "loss": 0.0378, "step": 18460 }, { "grad_norm": 0.3071949779987335, "learning_rate": 9.132030869673876e-05, "loss": 0.0445, "step": 18470 }, { "grad_norm": 0.32368704676628113, "learning_rate": 9.130866733422288e-05, "loss": 0.0449, "step": 18480 }, { "grad_norm": 0.37028053402900696, "learning_rate": 9.129701891318556e-05, "loss": 0.0367, "step": 18490 }, { "grad_norm": 0.2904614806175232, "learning_rate": 9.128536343561718e-05, "loss": 0.0465, "step": 18500 }, { "grad_norm": 0.32387736439704895, "learning_rate": 9.127370090350934e-05, "loss": 0.0421, "step": 18510 }, { "grad_norm": 0.3470765948295593, "learning_rate": 9.126203131885487e-05, "loss": 0.0405, "step": 18520 }, { "grad_norm": 0.33689969778060913, "learning_rate": 9.125035468364775e-05, "loss": 0.046, "step": 18530 }, { "grad_norm": 0.32673394680023193, "learning_rate": 9.123867099988322e-05, "loss": 0.0382, "step": 18540 }, { "grad_norm": 0.3067881464958191, "learning_rate": 9.122698026955769e-05, "loss": 0.0394, "step": 18550 }, { "grad_norm": 0.3337791860103607, "learning_rate": 9.12152824946688e-05, "loss": 0.0539, "step": 18560 }, { "grad_norm": 0.4121781587600708, "learning_rate": 9.120357767721538e-05, "loss": 0.0415, "step": 18570 }, { "grad_norm": 0.3396424949169159, "learning_rate": 9.119186581919745e-05, "loss": 0.035, "step": 18580 }, { "grad_norm": 0.31414729356765747, "learning_rate": 9.118014692261624e-05, "loss": 0.0389, "step": 18590 }, { "grad_norm": 0.38145777583122253, "learning_rate": 9.116842098947422e-05, "loss": 0.0511, "step": 18600 }, { "grad_norm": 0.49241530895233154, "learning_rate": 9.115668802177499e-05, "loss": 0.0343, "step": 18610 }, { "grad_norm": 0.3410378098487854, "learning_rate": 9.114494802152342e-05, "loss": 0.0503, "step": 18620 }, { "grad_norm": 0.4708850681781769, "learning_rate": 9.113320099072555e-05, "loss": 0.0671, "step": 18630 }, { "grad_norm": 0.43267616629600525, "learning_rate": 9.112144693138864e-05, "loss": 0.0426, "step": 18640 }, { "grad_norm": 0.3231451213359833, "learning_rate": 9.110968584552111e-05, "loss": 0.0379, "step": 18650 }, { "grad_norm": 0.3881978988647461, "learning_rate": 9.109791773513264e-05, "loss": 0.0385, "step": 18660 }, { "grad_norm": 0.3294752538204193, "learning_rate": 9.108614260223403e-05, "loss": 0.044, "step": 18670 }, { "grad_norm": 0.3947228193283081, "learning_rate": 9.107436044883738e-05, "loss": 0.0416, "step": 18680 }, { "grad_norm": 0.32973217964172363, "learning_rate": 9.10625712769559e-05, "loss": 0.0401, "step": 18690 }, { "grad_norm": 0.3119606673717499, "learning_rate": 9.105077508860406e-05, "loss": 0.0381, "step": 18700 }, { "grad_norm": 0.4167226552963257, "learning_rate": 9.103897188579751e-05, "loss": 0.0371, "step": 18710 }, { "grad_norm": 0.3390364348888397, "learning_rate": 9.102716167055308e-05, "loss": 0.0321, "step": 18720 }, { "grad_norm": 0.3813320994377136, "learning_rate": 9.10153444448888e-05, "loss": 0.0643, "step": 18730 }, { "grad_norm": 0.2854875326156616, "learning_rate": 9.100352021082393e-05, "loss": 0.0342, "step": 18740 }, { "grad_norm": 0.3299148678779602, "learning_rate": 9.099168897037891e-05, "loss": 0.0323, "step": 18750 }, { "grad_norm": 0.2869228422641754, "learning_rate": 9.097985072557538e-05, "loss": 0.0377, "step": 18760 }, { "grad_norm": 0.23262691497802734, "learning_rate": 9.096800547843615e-05, "loss": 0.0569, "step": 18770 }, { "grad_norm": 0.4672508239746094, "learning_rate": 9.095615323098526e-05, "loss": 0.0483, "step": 18780 }, { "grad_norm": 0.38343802094459534, "learning_rate": 9.094429398524795e-05, "loss": 0.037, "step": 18790 }, { "grad_norm": 0.31013673543930054, "learning_rate": 9.093242774325061e-05, "loss": 0.037, "step": 18800 }, { "grad_norm": 0.25060105323791504, "learning_rate": 9.092055450702088e-05, "loss": 0.0373, "step": 18810 }, { "grad_norm": 0.35947081446647644, "learning_rate": 9.090867427858756e-05, "loss": 0.0411, "step": 18820 }, { "grad_norm": 0.4023818075656891, "learning_rate": 9.089678705998066e-05, "loss": 0.0375, "step": 18830 }, { "grad_norm": 0.32333898544311523, "learning_rate": 9.088489285323139e-05, "loss": 0.0384, "step": 18840 }, { "grad_norm": 0.25305649638175964, "learning_rate": 9.087299166037212e-05, "loss": 0.0368, "step": 18850 }, { "grad_norm": 0.32381975650787354, "learning_rate": 9.086108348343647e-05, "loss": 0.0433, "step": 18860 }, { "grad_norm": 0.28441745042800903, "learning_rate": 9.08491683244592e-05, "loss": 0.0377, "step": 18870 }, { "grad_norm": 0.19513966143131256, "learning_rate": 9.08372461854763e-05, "loss": 0.0333, "step": 18880 }, { "grad_norm": 0.2572403848171234, "learning_rate": 9.082531706852492e-05, "loss": 0.0317, "step": 18890 }, { "grad_norm": 0.3252226412296295, "learning_rate": 9.081338097564342e-05, "loss": 0.0502, "step": 18900 }, { "grad_norm": 0.2806633710861206, "learning_rate": 9.080143790887137e-05, "loss": 0.0356, "step": 18910 }, { "grad_norm": 0.27678531408309937, "learning_rate": 9.07894878702495e-05, "loss": 0.0318, "step": 18920 }, { "grad_norm": 0.32796618342399597, "learning_rate": 9.077753086181974e-05, "loss": 0.0429, "step": 18930 }, { "grad_norm": 0.32099974155426025, "learning_rate": 9.076556688562524e-05, "loss": 0.0344, "step": 18940 }, { "grad_norm": 0.34819185733795166, "learning_rate": 9.075359594371029e-05, "loss": 0.037, "step": 18950 }, { "grad_norm": 0.22678406536579132, "learning_rate": 9.07416180381204e-05, "loss": 0.0377, "step": 18960 }, { "grad_norm": 0.3711540699005127, "learning_rate": 9.072963317090228e-05, "loss": 0.0358, "step": 18970 }, { "grad_norm": 0.34270986914634705, "learning_rate": 9.071764134410382e-05, "loss": 0.0383, "step": 18980 }, { "grad_norm": 0.3251079320907593, "learning_rate": 9.070564255977407e-05, "loss": 0.0365, "step": 18990 }, { "grad_norm": 0.30465176701545715, "learning_rate": 9.06936368199633e-05, "loss": 0.0373, "step": 19000 }, { "grad_norm": 0.26810991764068604, "learning_rate": 9.0681624126723e-05, "loss": 0.0426, "step": 19010 }, { "grad_norm": 0.2850198745727539, "learning_rate": 9.066960448210576e-05, "loss": 0.0429, "step": 19020 }, { "grad_norm": 0.34243834018707275, "learning_rate": 9.065757788816543e-05, "loss": 0.0333, "step": 19030 }, { "grad_norm": 0.247733935713768, "learning_rate": 9.064554434695705e-05, "loss": 0.0426, "step": 19040 }, { "grad_norm": 0.3148338496685028, "learning_rate": 9.063350386053677e-05, "loss": 0.0451, "step": 19050 }, { "grad_norm": 0.37860020995140076, "learning_rate": 9.062145643096202e-05, "loss": 0.0408, "step": 19060 }, { "grad_norm": 0.3016754984855652, "learning_rate": 9.060940206029136e-05, "loss": 0.0414, "step": 19070 }, { "grad_norm": 0.3918987214565277, "learning_rate": 9.059734075058457e-05, "loss": 0.0374, "step": 19080 }, { "grad_norm": 0.2806384265422821, "learning_rate": 9.058527250390257e-05, "loss": 0.0388, "step": 19090 }, { "grad_norm": 0.32128486037254333, "learning_rate": 9.057319732230752e-05, "loss": 0.0429, "step": 19100 }, { "grad_norm": 0.30593690276145935, "learning_rate": 9.056111520786273e-05, "loss": 0.037, "step": 19110 }, { "grad_norm": 0.2823489308357239, "learning_rate": 9.054902616263268e-05, "loss": 0.0346, "step": 19120 }, { "grad_norm": 0.3509661555290222, "learning_rate": 9.05369301886831e-05, "loss": 0.042, "step": 19130 }, { "grad_norm": 0.32721853256225586, "learning_rate": 9.052482728808083e-05, "loss": 0.0328, "step": 19140 }, { "grad_norm": 0.3153187334537506, "learning_rate": 9.051271746289391e-05, "loss": 0.0273, "step": 19150 }, { "grad_norm": 0.29226216673851013, "learning_rate": 9.050060071519162e-05, "loss": 0.0366, "step": 19160 }, { "grad_norm": 0.4100629687309265, "learning_rate": 9.048847704704437e-05, "loss": 0.0329, "step": 19170 }, { "grad_norm": 0.29443714022636414, "learning_rate": 9.047634646052376e-05, "loss": 0.0363, "step": 19180 }, { "grad_norm": 0.43735459446907043, "learning_rate": 9.046420895770256e-05, "loss": 0.0528, "step": 19190 }, { "grad_norm": 0.3749738931655884, "learning_rate": 9.045206454065473e-05, "loss": 0.0366, "step": 19200 }, { "grad_norm": 0.2929086685180664, "learning_rate": 9.043991321145546e-05, "loss": 0.0316, "step": 19210 }, { "grad_norm": 0.29731485247612, "learning_rate": 9.042775497218105e-05, "loss": 0.0309, "step": 19220 }, { "grad_norm": 0.24493631720542908, "learning_rate": 9.041558982490901e-05, "loss": 0.0362, "step": 19230 }, { "grad_norm": 0.27940240502357483, "learning_rate": 9.040341777171805e-05, "loss": 0.0353, "step": 19240 }, { "grad_norm": 0.3456284999847412, "learning_rate": 9.039123881468802e-05, "loss": 0.0356, "step": 19250 }, { "grad_norm": 0.33453822135925293, "learning_rate": 9.037905295589998e-05, "loss": 0.0348, "step": 19260 }, { "grad_norm": 0.32331743836402893, "learning_rate": 9.036686019743617e-05, "loss": 0.0307, "step": 19270 }, { "grad_norm": 0.3314591646194458, "learning_rate": 9.035466054137997e-05, "loss": 0.0339, "step": 19280 }, { "grad_norm": 0.40271657705307007, "learning_rate": 9.0342453989816e-05, "loss": 0.0424, "step": 19290 }, { "grad_norm": 0.24742592871189117, "learning_rate": 9.033024054483e-05, "loss": 0.035, "step": 19300 }, { "grad_norm": 0.32474371790885925, "learning_rate": 9.031802020850894e-05, "loss": 0.0357, "step": 19310 }, { "grad_norm": 0.30157196521759033, "learning_rate": 9.030579298294092e-05, "loss": 0.0454, "step": 19320 }, { "grad_norm": 0.31771403551101685, "learning_rate": 9.029355887021524e-05, "loss": 0.0468, "step": 19330 }, { "grad_norm": 0.28655391931533813, "learning_rate": 9.028131787242238e-05, "loss": 0.0511, "step": 19340 }, { "grad_norm": 0.2166988104581833, "learning_rate": 9.026906999165399e-05, "loss": 0.0374, "step": 19350 }, { "grad_norm": 0.27378273010253906, "learning_rate": 9.025681523000291e-05, "loss": 0.029, "step": 19360 }, { "grad_norm": 0.357963889837265, "learning_rate": 9.024455358956315e-05, "loss": 0.0364, "step": 19370 }, { "grad_norm": 0.30884772539138794, "learning_rate": 9.023228507242984e-05, "loss": 0.0415, "step": 19380 }, { "grad_norm": 0.30099499225616455, "learning_rate": 9.022000968069937e-05, "loss": 0.0376, "step": 19390 }, { "grad_norm": 0.306153804063797, "learning_rate": 9.020772741646928e-05, "loss": 0.0361, "step": 19400 }, { "grad_norm": 0.26508983969688416, "learning_rate": 9.019543828183826e-05, "loss": 0.0318, "step": 19410 }, { "grad_norm": 0.837009072303772, "learning_rate": 9.018314227890616e-05, "loss": 0.0341, "step": 19420 }, { "grad_norm": 0.4345988631248474, "learning_rate": 9.017083940977408e-05, "loss": 0.054, "step": 19430 }, { "grad_norm": 0.30107954144477844, "learning_rate": 9.015852967654422e-05, "loss": 0.0416, "step": 19440 }, { "grad_norm": 0.2455548346042633, "learning_rate": 9.014621308131996e-05, "loss": 0.0346, "step": 19450 }, { "grad_norm": 0.2125491201877594, "learning_rate": 9.01338896262059e-05, "loss": 0.0361, "step": 19460 }, { "grad_norm": 0.24895811080932617, "learning_rate": 9.012155931330777e-05, "loss": 0.0425, "step": 19470 }, { "grad_norm": 0.3877929747104645, "learning_rate": 9.010922214473246e-05, "loss": 0.0343, "step": 19480 }, { "grad_norm": 0.3184635043144226, "learning_rate": 9.009687812258808e-05, "loss": 0.046, "step": 19490 }, { "grad_norm": 0.2539103627204895, "learning_rate": 9.00845272489839e-05, "loss": 0.0294, "step": 19500 }, { "grad_norm": 0.30561792850494385, "learning_rate": 9.007216952603031e-05, "loss": 0.0408, "step": 19510 }, { "grad_norm": 0.39285531640052795, "learning_rate": 9.005980495583894e-05, "loss": 0.0333, "step": 19520 }, { "grad_norm": 0.3131166696548462, "learning_rate": 9.004743354052252e-05, "loss": 0.0347, "step": 19530 }, { "grad_norm": 0.30740901827812195, "learning_rate": 9.003505528219503e-05, "loss": 0.037, "step": 19540 }, { "grad_norm": 0.24242593348026276, "learning_rate": 9.002267018297154e-05, "loss": 0.0392, "step": 19550 }, { "grad_norm": 0.31255480647087097, "learning_rate": 9.001027824496834e-05, "loss": 0.0429, "step": 19560 }, { "grad_norm": 0.34423699975013733, "learning_rate": 8.999787947030287e-05, "loss": 0.0348, "step": 19570 }, { "grad_norm": 0.23315167427062988, "learning_rate": 8.998547386109376e-05, "loss": 0.037, "step": 19580 }, { "grad_norm": 0.26624298095703125, "learning_rate": 8.997306141946073e-05, "loss": 0.0307, "step": 19590 }, { "grad_norm": 0.30420979857444763, "learning_rate": 8.996064214752481e-05, "loss": 0.0388, "step": 19600 }, { "grad_norm": 0.3860834538936615, "learning_rate": 8.994821604740806e-05, "loss": 0.0318, "step": 19610 }, { "grad_norm": 0.4733683466911316, "learning_rate": 8.993578312123377e-05, "loss": 0.034, "step": 19620 }, { "grad_norm": 0.23516152799129486, "learning_rate": 8.992334337112639e-05, "loss": 0.0354, "step": 19630 }, { "grad_norm": 0.32047608494758606, "learning_rate": 8.991089679921154e-05, "loss": 0.0326, "step": 19640 }, { "grad_norm": 0.3196501135826111, "learning_rate": 8.989844340761599e-05, "loss": 0.035, "step": 19650 }, { "grad_norm": 0.3534032106399536, "learning_rate": 8.988598319846768e-05, "loss": 0.0336, "step": 19660 }, { "grad_norm": 0.2150549590587616, "learning_rate": 8.987351617389574e-05, "loss": 0.0342, "step": 19670 }, { "grad_norm": 0.25530877709388733, "learning_rate": 8.98610423360304e-05, "loss": 0.0291, "step": 19680 }, { "grad_norm": 0.17028450965881348, "learning_rate": 8.984856168700317e-05, "loss": 0.0401, "step": 19690 }, { "grad_norm": 0.2797783315181732, "learning_rate": 8.983607422894658e-05, "loss": 0.035, "step": 19700 }, { "grad_norm": 0.3416748642921448, "learning_rate": 8.982357996399442e-05, "loss": 0.0425, "step": 19710 }, { "grad_norm": 0.3166443407535553, "learning_rate": 8.981107889428164e-05, "loss": 0.0366, "step": 19720 }, { "grad_norm": 0.24706529080867767, "learning_rate": 8.979857102194428e-05, "loss": 0.0323, "step": 19730 }, { "grad_norm": 0.2845945358276367, "learning_rate": 8.978605634911968e-05, "loss": 0.0339, "step": 19740 }, { "grad_norm": 0.3223609924316406, "learning_rate": 8.977353487794616e-05, "loss": 0.0299, "step": 19750 }, { "grad_norm": 0.30876293778419495, "learning_rate": 8.976100661056334e-05, "loss": 0.0395, "step": 19760 }, { "grad_norm": 0.23110564053058624, "learning_rate": 8.974847154911197e-05, "loss": 0.028, "step": 19770 }, { "grad_norm": 0.502733051776886, "learning_rate": 8.973592969573393e-05, "loss": 0.0405, "step": 19780 }, { "grad_norm": 0.35816606879234314, "learning_rate": 8.972338105257228e-05, "loss": 0.0432, "step": 19790 }, { "grad_norm": 0.24770647287368774, "learning_rate": 8.971082562177125e-05, "loss": 0.033, "step": 19800 }, { "grad_norm": 0.26814714074134827, "learning_rate": 8.96982634054762e-05, "loss": 0.0381, "step": 19810 }, { "grad_norm": 0.31523028016090393, "learning_rate": 8.96856944058337e-05, "loss": 0.0348, "step": 19820 }, { "grad_norm": 0.2952667474746704, "learning_rate": 8.967311862499144e-05, "loss": 0.0314, "step": 19830 }, { "grad_norm": 0.2866097688674927, "learning_rate": 8.966053606509825e-05, "loss": 0.0326, "step": 19840 }, { "grad_norm": 0.40925779938697815, "learning_rate": 8.964794672830417e-05, "loss": 0.0436, "step": 19850 }, { "grad_norm": 0.32756850123405457, "learning_rate": 8.963535061676038e-05, "loss": 0.0502, "step": 19860 }, { "grad_norm": 0.5186451077461243, "learning_rate": 8.962274773261918e-05, "loss": 0.0331, "step": 19870 }, { "grad_norm": 0.3904823660850525, "learning_rate": 8.961013807803409e-05, "loss": 0.0459, "step": 19880 }, { "grad_norm": 0.2761234939098358, "learning_rate": 8.959752165515973e-05, "loss": 0.0404, "step": 19890 }, { "grad_norm": 0.30626368522644043, "learning_rate": 8.958489846615193e-05, "loss": 0.0568, "step": 19900 }, { "grad_norm": 0.3254440426826477, "learning_rate": 8.957226851316762e-05, "loss": 0.036, "step": 19910 }, { "grad_norm": 0.24204596877098083, "learning_rate": 8.955963179836493e-05, "loss": 0.0336, "step": 19920 }, { "grad_norm": 0.2852480411529541, "learning_rate": 8.954698832390312e-05, "loss": 0.0322, "step": 19930 }, { "grad_norm": 0.28817218542099, "learning_rate": 8.953433809194263e-05, "loss": 0.0334, "step": 19940 }, { "grad_norm": 0.3579556345939636, "learning_rate": 8.9521681104645e-05, "loss": 0.0323, "step": 19950 }, { "grad_norm": 0.33513128757476807, "learning_rate": 8.9509017364173e-05, "loss": 0.0358, "step": 19960 }, { "grad_norm": 0.3544406294822693, "learning_rate": 8.949634687269052e-05, "loss": 0.0381, "step": 19970 }, { "grad_norm": 0.3096669614315033, "learning_rate": 8.948366963236259e-05, "loss": 0.0431, "step": 19980 }, { "grad_norm": 0.2577334940433502, "learning_rate": 8.947098564535538e-05, "loss": 0.0374, "step": 19990 }, { "grad_norm": 0.31935498118400574, "learning_rate": 8.945829491383627e-05, "loss": 0.0361, "step": 20000 }, { "grad_norm": 0.30783289670944214, "learning_rate": 8.944559743997374e-05, "loss": 0.036, "step": 20010 }, { "grad_norm": 0.3454737961292267, "learning_rate": 8.943289322593746e-05, "loss": 0.0347, "step": 20020 }, { "grad_norm": 0.3042621612548828, "learning_rate": 8.942018227389821e-05, "loss": 0.0398, "step": 20030 }, { "grad_norm": 0.35156604647636414, "learning_rate": 8.940746458602795e-05, "loss": 0.0506, "step": 20040 }, { "grad_norm": 0.35134467482566833, "learning_rate": 8.939474016449979e-05, "loss": 0.0426, "step": 20050 }, { "grad_norm": 0.263167142868042, "learning_rate": 8.938200901148799e-05, "loss": 0.0312, "step": 20060 }, { "grad_norm": 0.38283756375312805, "learning_rate": 8.936927112916795e-05, "loss": 0.0491, "step": 20070 }, { "grad_norm": 0.2988850474357605, "learning_rate": 8.935652651971622e-05, "loss": 0.0342, "step": 20080 }, { "grad_norm": 0.272814005613327, "learning_rate": 8.934377518531052e-05, "loss": 0.0418, "step": 20090 }, { "grad_norm": 0.29016241431236267, "learning_rate": 8.933101712812967e-05, "loss": 0.037, "step": 20100 }, { "grad_norm": 0.26307496428489685, "learning_rate": 8.931825235035374e-05, "loss": 0.0371, "step": 20110 }, { "grad_norm": 0.28380200266838074, "learning_rate": 8.930548085416382e-05, "loss": 0.0349, "step": 20120 }, { "grad_norm": 0.2932038903236389, "learning_rate": 8.92927026417422e-05, "loss": 0.0363, "step": 20130 }, { "grad_norm": 0.3769814968109131, "learning_rate": 8.92799177152724e-05, "loss": 0.0315, "step": 20140 }, { "grad_norm": 0.27802059054374695, "learning_rate": 8.926712607693895e-05, "loss": 0.0345, "step": 20150 }, { "grad_norm": 0.3013020157814026, "learning_rate": 8.925432772892762e-05, "loss": 0.0388, "step": 20160 }, { "grad_norm": 0.57244473695755, "learning_rate": 8.924152267342529e-05, "loss": 0.0455, "step": 20170 }, { "grad_norm": 0.2996712028980255, "learning_rate": 8.922871091261998e-05, "loss": 0.0421, "step": 20180 }, { "grad_norm": 0.21786107122898102, "learning_rate": 8.92158924487009e-05, "loss": 0.0354, "step": 20190 }, { "grad_norm": 0.3000074326992035, "learning_rate": 8.920306728385834e-05, "loss": 0.0328, "step": 20200 }, { "grad_norm": 0.2789457142353058, "learning_rate": 8.919023542028379e-05, "loss": 0.0318, "step": 20210 }, { "grad_norm": 0.30565372109413147, "learning_rate": 8.917739686016988e-05, "loss": 0.0405, "step": 20220 }, { "grad_norm": 0.23068755865097046, "learning_rate": 8.916455160571033e-05, "loss": 0.0418, "step": 20230 }, { "grad_norm": 0.31822288036346436, "learning_rate": 8.915169965910008e-05, "loss": 0.0347, "step": 20240 }, { "grad_norm": 0.26844075322151184, "learning_rate": 8.913884102253514e-05, "loss": 0.0292, "step": 20250 }, { "grad_norm": 0.23308177292346954, "learning_rate": 8.912597569821273e-05, "loss": 0.0305, "step": 20260 }, { "grad_norm": 0.3058111071586609, "learning_rate": 8.911310368833118e-05, "loss": 0.0323, "step": 20270 }, { "grad_norm": 0.2647559940814972, "learning_rate": 8.910022499508994e-05, "loss": 0.0274, "step": 20280 }, { "grad_norm": 0.26139524579048157, "learning_rate": 8.908733962068965e-05, "loss": 0.0488, "step": 20290 }, { "grad_norm": 0.4295404255390167, "learning_rate": 8.907444756733207e-05, "loss": 0.0375, "step": 20300 }, { "grad_norm": 0.28283146023750305, "learning_rate": 8.906154883722006e-05, "loss": 0.0405, "step": 20310 }, { "grad_norm": 0.39670422673225403, "learning_rate": 8.904864343255773e-05, "loss": 0.0358, "step": 20320 }, { "grad_norm": 0.23402105271816254, "learning_rate": 8.90357313555502e-05, "loss": 0.0305, "step": 20330 }, { "grad_norm": 0.28372102975845337, "learning_rate": 8.90228126084038e-05, "loss": 0.0366, "step": 20340 }, { "grad_norm": 0.3906462490558624, "learning_rate": 8.900988719332601e-05, "loss": 0.0408, "step": 20350 }, { "grad_norm": 0.2809843420982361, "learning_rate": 8.899695511252542e-05, "loss": 0.0339, "step": 20360 }, { "grad_norm": 0.33318573236465454, "learning_rate": 8.898401636821176e-05, "loss": 0.0405, "step": 20370 }, { "grad_norm": 0.2807134687900543, "learning_rate": 8.897107096259593e-05, "loss": 0.0494, "step": 20380 }, { "grad_norm": 0.31255632638931274, "learning_rate": 8.895811889788994e-05, "loss": 0.0326, "step": 20390 }, { "grad_norm": 0.23122583329677582, "learning_rate": 8.894516017630692e-05, "loss": 0.0323, "step": 20400 }, { "grad_norm": 0.352802038192749, "learning_rate": 8.893219480006118e-05, "loss": 0.0392, "step": 20410 }, { "grad_norm": 0.2741740047931671, "learning_rate": 8.891922277136817e-05, "loss": 0.0328, "step": 20420 }, { "grad_norm": 0.30201515555381775, "learning_rate": 8.890624409244441e-05, "loss": 0.0337, "step": 20430 }, { "grad_norm": 0.25004029273986816, "learning_rate": 8.889325876550763e-05, "loss": 0.0353, "step": 20440 }, { "grad_norm": 0.4078768491744995, "learning_rate": 8.888026679277666e-05, "loss": 0.037, "step": 20450 }, { "grad_norm": 0.17812934517860413, "learning_rate": 8.886726817647147e-05, "loss": 0.0546, "step": 20460 }, { "grad_norm": 0.23783381283283234, "learning_rate": 8.885426291881319e-05, "loss": 0.0464, "step": 20470 }, { "grad_norm": 0.2939341962337494, "learning_rate": 8.884125102202401e-05, "loss": 0.038, "step": 20480 }, { "grad_norm": 0.36367568373680115, "learning_rate": 8.882823248832736e-05, "loss": 0.0365, "step": 20490 }, { "grad_norm": 0.3273192346096039, "learning_rate": 8.881520731994772e-05, "loss": 0.0367, "step": 20500 }, { "grad_norm": 0.3671839237213135, "learning_rate": 8.880217551911077e-05, "loss": 0.0393, "step": 20510 }, { "grad_norm": 0.3264642059803009, "learning_rate": 8.878913708804323e-05, "loss": 0.0386, "step": 20520 }, { "grad_norm": 0.36774492263793945, "learning_rate": 8.877609202897308e-05, "loss": 0.0402, "step": 20530 }, { "grad_norm": 0.31476297974586487, "learning_rate": 8.876304034412933e-05, "loss": 0.0282, "step": 20540 }, { "grad_norm": 0.2976148724555969, "learning_rate": 8.874998203574214e-05, "loss": 0.0387, "step": 20550 }, { "grad_norm": 0.32902979850769043, "learning_rate": 8.873691710604284e-05, "loss": 0.0361, "step": 20560 }, { "grad_norm": 0.3230041563510895, "learning_rate": 8.872384555726387e-05, "loss": 0.0375, "step": 20570 }, { "grad_norm": 0.3228917717933655, "learning_rate": 8.871076739163878e-05, "loss": 0.0396, "step": 20580 }, { "grad_norm": 0.42334988713264465, "learning_rate": 8.86976826114023e-05, "loss": 0.0405, "step": 20590 }, { "grad_norm": 0.25431695580482483, "learning_rate": 8.868459121879023e-05, "loss": 0.0428, "step": 20600 }, { "grad_norm": 0.3498999774456024, "learning_rate": 8.867149321603956e-05, "loss": 0.0433, "step": 20610 }, { "grad_norm": 0.27620089054107666, "learning_rate": 8.865838860538835e-05, "loss": 0.0351, "step": 20620 }, { "grad_norm": 0.3643154203891754, "learning_rate": 8.864527738907585e-05, "loss": 0.0373, "step": 20630 }, { "grad_norm": 0.309037983417511, "learning_rate": 8.863215956934239e-05, "loss": 0.0432, "step": 20640 }, { "grad_norm": 0.33938103914260864, "learning_rate": 8.861903514842947e-05, "loss": 0.0328, "step": 20650 }, { "grad_norm": 0.3019244372844696, "learning_rate": 8.860590412857966e-05, "loss": 0.033, "step": 20660 }, { "grad_norm": 0.3058741092681885, "learning_rate": 8.85927665120367e-05, "loss": 0.0306, "step": 20670 }, { "grad_norm": 0.3263258635997772, "learning_rate": 8.857962230104546e-05, "loss": 0.0387, "step": 20680 }, { "grad_norm": 0.2852556109428406, "learning_rate": 8.856647149785193e-05, "loss": 0.0325, "step": 20690 }, { "grad_norm": 0.3041239082813263, "learning_rate": 8.855331410470322e-05, "loss": 0.0394, "step": 20700 }, { "grad_norm": 0.3601406216621399, "learning_rate": 8.854015012384756e-05, "loss": 0.0336, "step": 20710 }, { "grad_norm": 0.23865923285484314, "learning_rate": 8.852697955753433e-05, "loss": 0.0404, "step": 20720 }, { "grad_norm": 0.3193061053752899, "learning_rate": 8.851380240801399e-05, "loss": 0.0403, "step": 20730 }, { "grad_norm": 0.3417566120624542, "learning_rate": 8.850061867753818e-05, "loss": 0.0321, "step": 20740 }, { "grad_norm": 0.335542768239975, "learning_rate": 8.848742836835963e-05, "loss": 0.0386, "step": 20750 }, { "grad_norm": 0.23748624324798584, "learning_rate": 8.847423148273221e-05, "loss": 0.0361, "step": 20760 }, { "grad_norm": 0.33003026247024536, "learning_rate": 8.846102802291092e-05, "loss": 0.0434, "step": 20770 }, { "grad_norm": 0.3342144787311554, "learning_rate": 8.844781799115183e-05, "loss": 0.0402, "step": 20780 }, { "grad_norm": 0.26986464858055115, "learning_rate": 8.84346013897122e-05, "loss": 0.0319, "step": 20790 }, { "grad_norm": 0.2517906129360199, "learning_rate": 8.842137822085038e-05, "loss": 0.0415, "step": 20800 }, { "grad_norm": 0.2633362114429474, "learning_rate": 8.840814848682585e-05, "loss": 0.0372, "step": 20810 }, { "grad_norm": 0.27369698882102966, "learning_rate": 8.83949121898992e-05, "loss": 0.0455, "step": 20820 }, { "grad_norm": 0.2782422602176666, "learning_rate": 8.838166933233217e-05, "loss": 0.033, "step": 20830 }, { "grad_norm": 0.25599589943885803, "learning_rate": 8.83684199163876e-05, "loss": 0.0369, "step": 20840 }, { "grad_norm": 0.31130293011665344, "learning_rate": 8.835516394432943e-05, "loss": 0.0374, "step": 20850 }, { "grad_norm": 0.2813878655433655, "learning_rate": 8.834190141842276e-05, "loss": 0.0377, "step": 20860 }, { "grad_norm": 0.32252445816993713, "learning_rate": 8.83286323409338e-05, "loss": 0.034, "step": 20870 }, { "grad_norm": 0.4962053894996643, "learning_rate": 8.831535671412986e-05, "loss": 0.0397, "step": 20880 }, { "grad_norm": 0.263572633266449, "learning_rate": 8.830207454027938e-05, "loss": 0.0398, "step": 20890 }, { "grad_norm": 0.3006613254547119, "learning_rate": 8.828878582165192e-05, "loss": 0.0359, "step": 20900 }, { "grad_norm": 0.235648512840271, "learning_rate": 8.827549056051818e-05, "loss": 0.0371, "step": 20910 }, { "grad_norm": 0.26727989315986633, "learning_rate": 8.826218875914993e-05, "loss": 0.0331, "step": 20920 }, { "grad_norm": 0.32267120480537415, "learning_rate": 8.82488804198201e-05, "loss": 0.0292, "step": 20930 }, { "grad_norm": 0.27187857031822205, "learning_rate": 8.82355655448027e-05, "loss": 0.0314, "step": 20940 }, { "grad_norm": 0.3600943982601166, "learning_rate": 8.822224413637293e-05, "loss": 0.0367, "step": 20950 }, { "grad_norm": 0.26844722032546997, "learning_rate": 8.820891619680697e-05, "loss": 0.0321, "step": 20960 }, { "grad_norm": 0.3304896056652069, "learning_rate": 8.819558172838227e-05, "loss": 0.0377, "step": 20970 }, { "grad_norm": 0.26331645250320435, "learning_rate": 8.818224073337731e-05, "loss": 0.0339, "step": 20980 }, { "grad_norm": 0.24007532000541687, "learning_rate": 8.816889321407169e-05, "loss": 0.035, "step": 20990 }, { "grad_norm": 0.2887406349182129, "learning_rate": 8.815553917274615e-05, "loss": 0.0353, "step": 21000 }, { "grad_norm": 0.2690851390361786, "learning_rate": 8.81421786116825e-05, "loss": 0.0397, "step": 21010 }, { "grad_norm": 0.3441906273365021, "learning_rate": 8.812881153316373e-05, "loss": 0.0329, "step": 21020 }, { "grad_norm": 0.28343909978866577, "learning_rate": 8.81154379394739e-05, "loss": 0.0399, "step": 21030 }, { "grad_norm": 0.27233782410621643, "learning_rate": 8.810205783289818e-05, "loss": 0.0351, "step": 21040 }, { "grad_norm": 0.3400655686855316, "learning_rate": 8.808867121572286e-05, "loss": 0.04, "step": 21050 }, { "grad_norm": 0.3849300146102905, "learning_rate": 8.807527809023537e-05, "loss": 0.0461, "step": 21060 }, { "grad_norm": 0.3364846408367157, "learning_rate": 8.80618784587242e-05, "loss": 0.0417, "step": 21070 }, { "grad_norm": 0.27245041728019714, "learning_rate": 8.804847232347902e-05, "loss": 0.0329, "step": 21080 }, { "grad_norm": 0.2724340260028839, "learning_rate": 8.803505968679054e-05, "loss": 0.0508, "step": 21090 }, { "grad_norm": 0.3865889608860016, "learning_rate": 8.802164055095061e-05, "loss": 0.0436, "step": 21100 }, { "grad_norm": 0.2637442350387573, "learning_rate": 8.80082149182522e-05, "loss": 0.0387, "step": 21110 }, { "grad_norm": 0.26598411798477173, "learning_rate": 8.79947827909894e-05, "loss": 0.0337, "step": 21120 }, { "grad_norm": 0.277285099029541, "learning_rate": 8.798134417145738e-05, "loss": 0.0408, "step": 21130 }, { "grad_norm": 0.39078086614608765, "learning_rate": 8.796789906195243e-05, "loss": 0.0408, "step": 21140 }, { "grad_norm": 0.2639884054660797, "learning_rate": 8.795444746477195e-05, "loss": 0.0415, "step": 21150 }, { "grad_norm": 0.25309404730796814, "learning_rate": 8.794098938221446e-05, "loss": 0.0352, "step": 21160 }, { "grad_norm": 0.2757874131202698, "learning_rate": 8.792752481657957e-05, "loss": 0.0371, "step": 21170 }, { "grad_norm": 0.3996371626853943, "learning_rate": 8.791405377016802e-05, "loss": 0.0348, "step": 21180 }, { "grad_norm": 0.33210545778274536, "learning_rate": 8.790057624528163e-05, "loss": 0.0412, "step": 21190 }, { "grad_norm": 0.4105249047279358, "learning_rate": 8.788709224422333e-05, "loss": 0.0388, "step": 21200 }, { "grad_norm": 0.3054462671279907, "learning_rate": 8.787360176929717e-05, "loss": 0.0405, "step": 21210 }, { "grad_norm": 0.26398345828056335, "learning_rate": 8.786010482280834e-05, "loss": 0.0392, "step": 21220 }, { "grad_norm": 0.28446874022483826, "learning_rate": 8.784660140706306e-05, "loss": 0.0401, "step": 21230 }, { "grad_norm": 0.3569846749305725, "learning_rate": 8.783309152436872e-05, "loss": 0.0341, "step": 21240 }, { "grad_norm": 0.27202504873275757, "learning_rate": 8.781957517703375e-05, "loss": 0.0397, "step": 21250 }, { "grad_norm": 0.2946307063102722, "learning_rate": 8.780605236736776e-05, "loss": 0.0433, "step": 21260 }, { "grad_norm": 0.28817805647850037, "learning_rate": 8.779252309768142e-05, "loss": 0.0426, "step": 21270 }, { "grad_norm": 0.264355331659317, "learning_rate": 8.777898737028652e-05, "loss": 0.0406, "step": 21280 }, { "grad_norm": 0.28709688782691956, "learning_rate": 8.776544518749591e-05, "loss": 0.0375, "step": 21290 }, { "grad_norm": 0.35923585295677185, "learning_rate": 8.775189655162364e-05, "loss": 0.0331, "step": 21300 }, { "grad_norm": 0.33654212951660156, "learning_rate": 8.773834146498474e-05, "loss": 0.033, "step": 21310 }, { "grad_norm": 0.25800806283950806, "learning_rate": 8.772477992989545e-05, "loss": 0.0335, "step": 21320 }, { "grad_norm": 0.4049616754055023, "learning_rate": 8.771121194867304e-05, "loss": 0.0388, "step": 21330 }, { "grad_norm": 0.22181953489780426, "learning_rate": 8.769763752363589e-05, "loss": 0.0331, "step": 21340 }, { "grad_norm": 0.26993870735168457, "learning_rate": 8.768405665710352e-05, "loss": 0.0378, "step": 21350 }, { "grad_norm": 0.2978156507015228, "learning_rate": 8.767046935139655e-05, "loss": 0.0494, "step": 21360 }, { "grad_norm": 0.33837467432022095, "learning_rate": 8.765687560883666e-05, "loss": 0.0343, "step": 21370 }, { "grad_norm": 0.3291105329990387, "learning_rate": 8.764327543174664e-05, "loss": 0.031, "step": 21380 }, { "grad_norm": 0.2291313111782074, "learning_rate": 8.762966882245038e-05, "loss": 0.0354, "step": 21390 }, { "grad_norm": 0.3562714755535126, "learning_rate": 8.761605578327291e-05, "loss": 0.0444, "step": 21400 }, { "grad_norm": 0.367666631937027, "learning_rate": 8.76024363165403e-05, "loss": 0.0494, "step": 21410 }, { "grad_norm": 0.3008807599544525, "learning_rate": 8.758881042457976e-05, "loss": 0.0412, "step": 21420 }, { "grad_norm": 0.30358704924583435, "learning_rate": 8.757517810971957e-05, "loss": 0.0338, "step": 21430 }, { "grad_norm": 0.20244735479354858, "learning_rate": 8.756153937428913e-05, "loss": 0.0362, "step": 21440 }, { "grad_norm": 0.26523467898368835, "learning_rate": 8.754789422061889e-05, "loss": 0.0437, "step": 21450 }, { "grad_norm": 0.2251775711774826, "learning_rate": 8.753424265104052e-05, "loss": 0.0349, "step": 21460 }, { "grad_norm": 0.4323016107082367, "learning_rate": 8.752058466788659e-05, "loss": 0.0499, "step": 21470 }, { "grad_norm": 0.31253260374069214, "learning_rate": 8.750692027349097e-05, "loss": 0.0327, "step": 21480 }, { "grad_norm": 0.3330309987068176, "learning_rate": 8.749324947018847e-05, "loss": 0.0329, "step": 21490 }, { "grad_norm": 0.2731366753578186, "learning_rate": 8.747957226031507e-05, "loss": 0.0343, "step": 21500 }, { "grad_norm": 0.2769990861415863, "learning_rate": 8.746588864620787e-05, "loss": 0.0332, "step": 21510 }, { "grad_norm": 0.35263097286224365, "learning_rate": 8.745219863020498e-05, "loss": 0.0376, "step": 21520 }, { "grad_norm": 0.26317235827445984, "learning_rate": 8.743850221464564e-05, "loss": 0.0354, "step": 21530 }, { "grad_norm": 0.27112722396850586, "learning_rate": 8.742479940187026e-05, "loss": 0.0341, "step": 21540 }, { "grad_norm": 0.2957565188407898, "learning_rate": 8.74110901942202e-05, "loss": 0.0344, "step": 21550 }, { "grad_norm": 0.23131375014781952, "learning_rate": 8.739737459403803e-05, "loss": 0.0344, "step": 21560 }, { "grad_norm": 0.3641199767589569, "learning_rate": 8.738365260366737e-05, "loss": 0.0392, "step": 21570 }, { "grad_norm": 0.39117109775543213, "learning_rate": 8.736992422545292e-05, "loss": 0.0385, "step": 21580 }, { "grad_norm": 0.244814932346344, "learning_rate": 8.73561894617405e-05, "loss": 0.0416, "step": 21590 }, { "grad_norm": 0.28489798307418823, "learning_rate": 8.734244831487697e-05, "loss": 0.0376, "step": 21600 }, { "grad_norm": 0.32259443402290344, "learning_rate": 8.732870078721035e-05, "loss": 0.0426, "step": 21610 }, { "grad_norm": 0.2981700599193573, "learning_rate": 8.731494688108972e-05, "loss": 0.041, "step": 21620 }, { "grad_norm": 0.3291621804237366, "learning_rate": 8.730118659886523e-05, "loss": 0.0418, "step": 21630 }, { "grad_norm": 0.3391321897506714, "learning_rate": 8.728741994288814e-05, "loss": 0.0369, "step": 21640 }, { "grad_norm": 0.3333194851875305, "learning_rate": 8.727364691551079e-05, "loss": 0.0393, "step": 21650 }, { "grad_norm": 0.3680875301361084, "learning_rate": 8.725986751908661e-05, "loss": 0.0361, "step": 21660 }, { "grad_norm": 0.2711705267429352, "learning_rate": 8.724608175597016e-05, "loss": 0.0326, "step": 21670 }, { "grad_norm": 0.23551340401172638, "learning_rate": 8.723228962851699e-05, "loss": 0.0294, "step": 21680 }, { "grad_norm": 0.24802768230438232, "learning_rate": 8.721849113908385e-05, "loss": 0.0345, "step": 21690 }, { "grad_norm": 0.2545539438724518, "learning_rate": 8.720468629002848e-05, "loss": 0.0338, "step": 21700 }, { "grad_norm": 0.28654056787490845, "learning_rate": 8.719087508370978e-05, "loss": 0.0338, "step": 21710 }, { "grad_norm": 0.40694594383239746, "learning_rate": 8.717705752248772e-05, "loss": 0.0437, "step": 21720 }, { "grad_norm": 0.29760468006134033, "learning_rate": 8.71632336087233e-05, "loss": 0.0312, "step": 21730 }, { "grad_norm": 0.2460617870092392, "learning_rate": 8.71494033447787e-05, "loss": 0.037, "step": 21740 }, { "grad_norm": 0.29773983359336853, "learning_rate": 8.713556673301708e-05, "loss": 0.0295, "step": 21750 }, { "grad_norm": 0.29728010296821594, "learning_rate": 8.712172377580278e-05, "loss": 0.0436, "step": 21760 }, { "grad_norm": 0.32247084379196167, "learning_rate": 8.710787447550114e-05, "loss": 0.036, "step": 21770 }, { "grad_norm": 0.30334970355033875, "learning_rate": 8.70940188344787e-05, "loss": 0.0509, "step": 21780 }, { "grad_norm": 0.31890806555747986, "learning_rate": 8.708015685510293e-05, "loss": 0.0297, "step": 21790 }, { "grad_norm": 0.24424569308757782, "learning_rate": 8.706628853974252e-05, "loss": 0.0389, "step": 21800 }, { "grad_norm": 0.26136454939842224, "learning_rate": 8.705241389076715e-05, "loss": 0.0309, "step": 21810 }, { "grad_norm": 0.3891229033470154, "learning_rate": 8.703853291054764e-05, "loss": 0.0299, "step": 21820 }, { "grad_norm": 0.3300064206123352, "learning_rate": 8.702464560145587e-05, "loss": 0.0393, "step": 21830 }, { "grad_norm": 0.3766447603702545, "learning_rate": 8.701075196586476e-05, "loss": 0.0361, "step": 21840 }, { "grad_norm": 0.2462700754404068, "learning_rate": 8.699685200614842e-05, "loss": 0.0343, "step": 21850 }, { "grad_norm": 0.29803580045700073, "learning_rate": 8.698294572468193e-05, "loss": 0.041, "step": 21860 }, { "grad_norm": 0.22846268117427826, "learning_rate": 8.696903312384148e-05, "loss": 0.0366, "step": 21870 }, { "grad_norm": 0.3240954279899597, "learning_rate": 8.695511420600439e-05, "loss": 0.0331, "step": 21880 }, { "grad_norm": 0.2794431149959564, "learning_rate": 8.694118897354901e-05, "loss": 0.0328, "step": 21890 }, { "grad_norm": 0.32422932982444763, "learning_rate": 8.692725742885478e-05, "loss": 0.0288, "step": 21900 }, { "grad_norm": 0.40493351221084595, "learning_rate": 8.691331957430221e-05, "loss": 0.0426, "step": 21910 }, { "grad_norm": 0.3187209963798523, "learning_rate": 8.68993754122729e-05, "loss": 0.0402, "step": 21920 }, { "grad_norm": 0.2515866458415985, "learning_rate": 8.688542494514955e-05, "loss": 0.0337, "step": 21930 }, { "grad_norm": 0.3234502375125885, "learning_rate": 8.68714681753159e-05, "loss": 0.0388, "step": 21940 }, { "grad_norm": 0.3030560314655304, "learning_rate": 8.685750510515676e-05, "loss": 0.0323, "step": 21950 }, { "grad_norm": 0.40465226769447327, "learning_rate": 8.684353573705805e-05, "loss": 0.0354, "step": 21960 }, { "grad_norm": 0.25174444913864136, "learning_rate": 8.682956007340677e-05, "loss": 0.0321, "step": 21970 }, { "grad_norm": 0.26007965207099915, "learning_rate": 8.681557811659095e-05, "loss": 0.0332, "step": 21980 }, { "grad_norm": 0.3205330967903137, "learning_rate": 8.680158986899974e-05, "loss": 0.035, "step": 21990 }, { "grad_norm": 0.39024001359939575, "learning_rate": 8.678759533302335e-05, "loss": 0.0364, "step": 22000 }, { "grad_norm": 0.2963496148586273, "learning_rate": 8.677359451105308e-05, "loss": 0.0367, "step": 22010 }, { "grad_norm": 0.24787621200084686, "learning_rate": 8.675958740548123e-05, "loss": 0.0348, "step": 22020 }, { "grad_norm": 0.21604247391223907, "learning_rate": 8.674557401870129e-05, "loss": 0.0378, "step": 22030 }, { "grad_norm": 0.442342072725296, "learning_rate": 8.673155435310775e-05, "loss": 0.039, "step": 22040 }, { "grad_norm": 0.4912762939929962, "learning_rate": 8.671752841109617e-05, "loss": 0.0349, "step": 22050 }, { "grad_norm": 0.36063307523727417, "learning_rate": 8.670349619506321e-05, "loss": 0.0331, "step": 22060 }, { "grad_norm": 0.26238179206848145, "learning_rate": 8.66894577074066e-05, "loss": 0.0353, "step": 22070 }, { "grad_norm": 0.24833227694034576, "learning_rate": 8.667541295052513e-05, "loss": 0.0353, "step": 22080 }, { "grad_norm": 0.2865019738674164, "learning_rate": 8.666136192681865e-05, "loss": 0.0412, "step": 22090 }, { "grad_norm": 0.26748204231262207, "learning_rate": 8.664730463868811e-05, "loss": 0.0368, "step": 22100 }, { "grad_norm": 0.3723875880241394, "learning_rate": 8.663324108853552e-05, "loss": 0.0495, "step": 22110 }, { "grad_norm": 0.3863280415534973, "learning_rate": 8.661917127876395e-05, "loss": 0.0472, "step": 22120 }, { "grad_norm": 0.4864349663257599, "learning_rate": 8.660509521177754e-05, "loss": 0.0389, "step": 22130 }, { "grad_norm": 0.2858513295650482, "learning_rate": 8.65910128899815e-05, "loss": 0.0354, "step": 22140 }, { "grad_norm": 0.3269266188144684, "learning_rate": 8.657692431578214e-05, "loss": 0.0352, "step": 22150 }, { "grad_norm": 0.22031661868095398, "learning_rate": 8.656282949158679e-05, "loss": 0.0312, "step": 22160 }, { "grad_norm": 0.24903449416160583, "learning_rate": 8.654872841980388e-05, "loss": 0.0283, "step": 22170 }, { "grad_norm": 0.25626838207244873, "learning_rate": 8.653462110284289e-05, "loss": 0.0347, "step": 22180 }, { "grad_norm": 0.1803896576166153, "learning_rate": 8.652050754311437e-05, "loss": 0.0326, "step": 22190 }, { "grad_norm": 0.30109402537345886, "learning_rate": 8.650638774302995e-05, "loss": 0.0329, "step": 22200 }, { "grad_norm": 0.31661897897720337, "learning_rate": 8.649226170500233e-05, "loss": 0.0357, "step": 22210 }, { "grad_norm": 0.24800361692905426, "learning_rate": 8.647812943144524e-05, "loss": 0.0302, "step": 22220 }, { "grad_norm": 0.28521451354026794, "learning_rate": 8.646399092477351e-05, "loss": 0.0313, "step": 22230 }, { "grad_norm": 0.3213386833667755, "learning_rate": 8.644984618740301e-05, "loss": 0.0368, "step": 22240 }, { "grad_norm": 0.3416197597980499, "learning_rate": 8.643569522175073e-05, "loss": 0.0273, "step": 22250 }, { "grad_norm": 0.4017043709754944, "learning_rate": 8.642153803023463e-05, "loss": 0.0378, "step": 22260 }, { "grad_norm": 0.3181597888469696, "learning_rate": 8.640737461527383e-05, "loss": 0.0289, "step": 22270 }, { "grad_norm": 0.33427268266677856, "learning_rate": 8.639320497928845e-05, "loss": 0.0348, "step": 22280 }, { "grad_norm": 0.27115505933761597, "learning_rate": 8.637902912469969e-05, "loss": 0.0342, "step": 22290 }, { "grad_norm": 0.29408586025238037, "learning_rate": 8.636484705392982e-05, "loss": 0.0331, "step": 22300 }, { "grad_norm": 0.2924649119377136, "learning_rate": 8.635065876940216e-05, "loss": 0.0342, "step": 22310 }, { "grad_norm": 0.31017857789993286, "learning_rate": 8.633646427354112e-05, "loss": 0.0457, "step": 22320 }, { "grad_norm": 0.2547626793384552, "learning_rate": 8.632226356877213e-05, "loss": 0.0337, "step": 22330 }, { "grad_norm": 0.18828055262565613, "learning_rate": 8.630805665752173e-05, "loss": 0.0474, "step": 22340 }, { "grad_norm": 0.2629557251930237, "learning_rate": 8.629384354221748e-05, "loss": 0.0379, "step": 22350 }, { "grad_norm": 0.22743594646453857, "learning_rate": 8.627962422528797e-05, "loss": 0.0312, "step": 22360 }, { "grad_norm": 0.2446315586566925, "learning_rate": 8.626539870916296e-05, "loss": 0.0427, "step": 22370 }, { "grad_norm": 0.31970247626304626, "learning_rate": 8.625116699627317e-05, "loss": 0.0382, "step": 22380 }, { "grad_norm": 0.30017876625061035, "learning_rate": 8.623692908905041e-05, "loss": 0.0282, "step": 22390 }, { "grad_norm": 0.3015287220478058, "learning_rate": 8.622268498992755e-05, "loss": 0.0433, "step": 22400 }, { "grad_norm": 0.36004385352134705, "learning_rate": 8.620843470133851e-05, "loss": 0.0493, "step": 22410 }, { "grad_norm": 0.3681640923023224, "learning_rate": 8.619417822571829e-05, "loss": 0.0378, "step": 22420 }, { "grad_norm": 0.3251231014728546, "learning_rate": 8.617991556550292e-05, "loss": 0.0333, "step": 22430 }, { "grad_norm": 0.34662362933158875, "learning_rate": 8.616564672312952e-05, "loss": 0.0339, "step": 22440 }, { "grad_norm": 0.3656077980995178, "learning_rate": 8.61513717010362e-05, "loss": 0.0384, "step": 22450 }, { "grad_norm": 0.6713355183601379, "learning_rate": 8.613709050166221e-05, "loss": 0.0369, "step": 22460 }, { "grad_norm": 0.2700645625591278, "learning_rate": 8.61228031274478e-05, "loss": 0.0371, "step": 22470 }, { "grad_norm": 0.4220218062400818, "learning_rate": 8.610850958083431e-05, "loss": 0.0401, "step": 22480 }, { "grad_norm": 0.33382853865623474, "learning_rate": 8.609420986426409e-05, "loss": 0.0343, "step": 22490 }, { "grad_norm": 0.36537978053092957, "learning_rate": 8.60799039801806e-05, "loss": 0.0383, "step": 22500 }, { "grad_norm": 0.40510183572769165, "learning_rate": 8.606559193102828e-05, "loss": 0.0313, "step": 22510 }, { "grad_norm": 0.32920393347740173, "learning_rate": 8.605127371925273e-05, "loss": 0.0458, "step": 22520 }, { "grad_norm": 0.2883625030517578, "learning_rate": 8.603694934730047e-05, "loss": 0.0377, "step": 22530 }, { "grad_norm": 0.30118632316589355, "learning_rate": 8.602261881761919e-05, "loss": 0.0275, "step": 22540 }, { "grad_norm": 0.23238371312618256, "learning_rate": 8.600828213265759e-05, "loss": 0.0412, "step": 22550 }, { "grad_norm": 0.3519538938999176, "learning_rate": 8.599393929486539e-05, "loss": 0.0501, "step": 22560 }, { "grad_norm": 0.3663569688796997, "learning_rate": 8.59795903066934e-05, "loss": 0.0339, "step": 22570 }, { "grad_norm": 0.25022321939468384, "learning_rate": 8.596523517059347e-05, "loss": 0.0314, "step": 22580 }, { "grad_norm": 0.4487135410308838, "learning_rate": 8.59508738890185e-05, "loss": 0.0331, "step": 22590 }, { "grad_norm": 0.3790486752986908, "learning_rate": 8.593650646442246e-05, "loss": 0.0364, "step": 22600 }, { "grad_norm": 0.30601799488067627, "learning_rate": 8.59221328992603e-05, "loss": 0.0315, "step": 22610 }, { "grad_norm": 0.32445889711380005, "learning_rate": 8.590775319598813e-05, "loss": 0.0326, "step": 22620 }, { "grad_norm": 0.26498284935951233, "learning_rate": 8.589336735706301e-05, "loss": 0.0331, "step": 22630 }, { "grad_norm": 0.24703627824783325, "learning_rate": 8.587897538494307e-05, "loss": 0.0408, "step": 22640 }, { "grad_norm": 0.3093606233596802, "learning_rate": 8.586457728208756e-05, "loss": 0.0325, "step": 22650 }, { "grad_norm": 0.38700637221336365, "learning_rate": 8.585017305095667e-05, "loss": 0.0364, "step": 22660 }, { "grad_norm": 0.3726568818092346, "learning_rate": 8.583576269401173e-05, "loss": 0.0317, "step": 22670 }, { "grad_norm": 0.2782250642776489, "learning_rate": 8.582134621371504e-05, "loss": 0.0352, "step": 22680 }, { "grad_norm": 0.31813710927963257, "learning_rate": 8.580692361253e-05, "loss": 0.0367, "step": 22690 }, { "grad_norm": 0.3681449890136719, "learning_rate": 8.579249489292104e-05, "loss": 0.0326, "step": 22700 }, { "grad_norm": 0.3170640766620636, "learning_rate": 8.577806005735363e-05, "loss": 0.0434, "step": 22710 }, { "grad_norm": 0.2892415225505829, "learning_rate": 8.576361910829429e-05, "loss": 0.0266, "step": 22720 }, { "grad_norm": 0.2631549835205078, "learning_rate": 8.574917204821057e-05, "loss": 0.0329, "step": 22730 }, { "grad_norm": 0.24809882044792175, "learning_rate": 8.57347188795711e-05, "loss": 0.0293, "step": 22740 }, { "grad_norm": 0.25914496183395386, "learning_rate": 8.572025960484551e-05, "loss": 0.0427, "step": 22750 }, { "grad_norm": 0.4659270644187927, "learning_rate": 8.57057942265045e-05, "loss": 0.0526, "step": 22760 }, { "grad_norm": 0.26649829745292664, "learning_rate": 8.569132274701984e-05, "loss": 0.0339, "step": 22770 }, { "grad_norm": 0.38720375299453735, "learning_rate": 8.567684516886427e-05, "loss": 0.0355, "step": 22780 }, { "grad_norm": 0.2699924111366272, "learning_rate": 8.56623614945116e-05, "loss": 0.031, "step": 22790 }, { "grad_norm": 0.3119056820869446, "learning_rate": 8.564787172643675e-05, "loss": 0.0333, "step": 22800 }, { "grad_norm": 0.38502073287963867, "learning_rate": 8.563337586711559e-05, "loss": 0.0418, "step": 22810 }, { "grad_norm": 0.28378769755363464, "learning_rate": 8.561887391902506e-05, "loss": 0.0384, "step": 22820 }, { "grad_norm": 0.35892415046691895, "learning_rate": 8.560436588464316e-05, "loss": 0.0336, "step": 22830 }, { "grad_norm": 0.2975686192512512, "learning_rate": 8.55898517664489e-05, "loss": 0.0421, "step": 22840 }, { "grad_norm": 0.2901003956794739, "learning_rate": 8.557533156692236e-05, "loss": 0.0467, "step": 22850 }, { "grad_norm": 0.297583669424057, "learning_rate": 8.556080528854467e-05, "loss": 0.0298, "step": 22860 }, { "grad_norm": 0.30371689796447754, "learning_rate": 8.554627293379791e-05, "loss": 0.0353, "step": 22870 }, { "grad_norm": 0.35640037059783936, "learning_rate": 8.553173450516531e-05, "loss": 0.0431, "step": 22880 }, { "grad_norm": 0.24004489183425903, "learning_rate": 8.551719000513108e-05, "loss": 0.0328, "step": 22890 }, { "grad_norm": 0.3667817711830139, "learning_rate": 8.550263943618049e-05, "loss": 0.0459, "step": 22900 }, { "grad_norm": 0.24736644327640533, "learning_rate": 8.54880828007998e-05, "loss": 0.0378, "step": 22910 }, { "grad_norm": 0.3226330280303955, "learning_rate": 8.547352010147637e-05, "loss": 0.0389, "step": 22920 }, { "grad_norm": 0.22775372862815857, "learning_rate": 8.545895134069855e-05, "loss": 0.0329, "step": 22930 }, { "grad_norm": 0.4295380413532257, "learning_rate": 8.544437652095576e-05, "loss": 0.0353, "step": 22940 }, { "grad_norm": 0.4258117973804474, "learning_rate": 8.542979564473843e-05, "loss": 0.0454, "step": 22950 }, { "grad_norm": 0.30308616161346436, "learning_rate": 8.541520871453802e-05, "loss": 0.0353, "step": 22960 }, { "grad_norm": 0.36650410294532776, "learning_rate": 8.540061573284705e-05, "loss": 0.0356, "step": 22970 }, { "grad_norm": 0.2755577266216278, "learning_rate": 8.538601670215906e-05, "loss": 0.0336, "step": 22980 }, { "grad_norm": 0.3732624053955078, "learning_rate": 8.537141162496864e-05, "loss": 0.0421, "step": 22990 }, { "grad_norm": 0.24400046467781067, "learning_rate": 8.535680050377137e-05, "loss": 0.0359, "step": 23000 }, { "grad_norm": 0.3780549168586731, "learning_rate": 8.534218334106391e-05, "loss": 0.036, "step": 23010 }, { "grad_norm": 0.32341545820236206, "learning_rate": 8.532756013934393e-05, "loss": 0.0346, "step": 23020 }, { "grad_norm": 0.3238082826137543, "learning_rate": 8.531293090111012e-05, "loss": 0.0397, "step": 23030 }, { "grad_norm": 0.33884352445602417, "learning_rate": 8.529829562886225e-05, "loss": 0.0358, "step": 23040 }, { "grad_norm": 0.3580041527748108, "learning_rate": 8.528365432510105e-05, "loss": 0.0421, "step": 23050 }, { "grad_norm": 0.27885472774505615, "learning_rate": 8.526900699232833e-05, "loss": 0.0342, "step": 23060 }, { "grad_norm": 0.2782742381095886, "learning_rate": 8.525435363304695e-05, "loss": 0.0348, "step": 23070 }, { "grad_norm": 0.33839842677116394, "learning_rate": 8.523969424976072e-05, "loss": 0.0363, "step": 23080 }, { "grad_norm": 0.2242579460144043, "learning_rate": 8.522502884497457e-05, "loss": 0.0311, "step": 23090 }, { "grad_norm": 0.3520868420600891, "learning_rate": 8.521035742119437e-05, "loss": 0.0418, "step": 23100 }, { "grad_norm": 0.2821533679962158, "learning_rate": 8.519567998092712e-05, "loss": 0.0417, "step": 23110 }, { "grad_norm": 0.31984326243400574, "learning_rate": 8.518099652668075e-05, "loss": 0.0343, "step": 23120 }, { "grad_norm": 0.3037870228290558, "learning_rate": 8.516630706096429e-05, "loss": 0.0317, "step": 23130 }, { "grad_norm": 0.3579377830028534, "learning_rate": 8.515161158628773e-05, "loss": 0.0285, "step": 23140 }, { "grad_norm": 0.28524914383888245, "learning_rate": 8.513691010516216e-05, "loss": 0.0498, "step": 23150 }, { "grad_norm": 0.2573293149471283, "learning_rate": 8.512220262009966e-05, "loss": 0.0291, "step": 23160 }, { "grad_norm": 0.316336452960968, "learning_rate": 8.510748913361332e-05, "loss": 0.0365, "step": 23170 }, { "grad_norm": 0.2804018557071686, "learning_rate": 8.509276964821726e-05, "loss": 0.0306, "step": 23180 }, { "grad_norm": 0.2855237126350403, "learning_rate": 8.507804416642669e-05, "loss": 0.0356, "step": 23190 }, { "grad_norm": 0.35665762424468994, "learning_rate": 8.506331269075774e-05, "loss": 0.0328, "step": 23200 }, { "grad_norm": 0.3117693364620209, "learning_rate": 8.504857522372765e-05, "loss": 0.0315, "step": 23210 }, { "grad_norm": 0.3406231999397278, "learning_rate": 8.503383176785461e-05, "loss": 0.0329, "step": 23220 }, { "grad_norm": 0.32557976245880127, "learning_rate": 8.501908232565792e-05, "loss": 0.0339, "step": 23230 }, { "grad_norm": 0.21339863538742065, "learning_rate": 8.50043268996578e-05, "loss": 0.0315, "step": 23240 }, { "grad_norm": 0.3301340937614441, "learning_rate": 8.498956549237562e-05, "loss": 0.0367, "step": 23250 }, { "grad_norm": 0.32414504885673523, "learning_rate": 8.497479810633366e-05, "loss": 0.0295, "step": 23260 }, { "grad_norm": 0.2502712905406952, "learning_rate": 8.496002474405525e-05, "loss": 0.0339, "step": 23270 }, { "grad_norm": 0.3046214282512665, "learning_rate": 8.494524540806478e-05, "loss": 0.031, "step": 23280 }, { "grad_norm": 0.35779067873954773, "learning_rate": 8.493046010088761e-05, "loss": 0.0413, "step": 23290 }, { "grad_norm": 0.3002491891384125, "learning_rate": 8.491566882505018e-05, "loss": 0.032, "step": 23300 }, { "grad_norm": 0.35366135835647583, "learning_rate": 8.490087158307988e-05, "loss": 0.0392, "step": 23310 }, { "grad_norm": 0.27309396862983704, "learning_rate": 8.488606837750518e-05, "loss": 0.0347, "step": 23320 }, { "grad_norm": 0.2951575815677643, "learning_rate": 8.487125921085552e-05, "loss": 0.0351, "step": 23330 }, { "grad_norm": 0.3203965127468109, "learning_rate": 8.485644408566141e-05, "loss": 0.0367, "step": 23340 }, { "grad_norm": 0.2843395173549652, "learning_rate": 8.484162300445431e-05, "loss": 0.0298, "step": 23350 }, { "grad_norm": 0.3113275468349457, "learning_rate": 8.482679596976676e-05, "loss": 0.0302, "step": 23360 }, { "grad_norm": 0.28291308879852295, "learning_rate": 8.48119629841323e-05, "loss": 0.0418, "step": 23370 }, { "grad_norm": 0.2894894778728485, "learning_rate": 8.479712405008547e-05, "loss": 0.0396, "step": 23380 }, { "grad_norm": 0.23843473196029663, "learning_rate": 8.478227917016184e-05, "loss": 0.0395, "step": 23390 }, { "grad_norm": 0.22442400455474854, "learning_rate": 8.476742834689801e-05, "loss": 0.0447, "step": 23400 }, { "grad_norm": 0.3275262713432312, "learning_rate": 8.475257158283157e-05, "loss": 0.0428, "step": 23410 }, { "grad_norm": 0.35718080401420593, "learning_rate": 8.473770888050112e-05, "loss": 0.0296, "step": 23420 }, { "grad_norm": 0.27065321803092957, "learning_rate": 8.47228402424463e-05, "loss": 0.0371, "step": 23430 }, { "grad_norm": 0.29870325326919556, "learning_rate": 8.470796567120775e-05, "loss": 0.0336, "step": 23440 }, { "grad_norm": 0.3473934531211853, "learning_rate": 8.469308516932714e-05, "loss": 0.0371, "step": 23450 }, { "grad_norm": 0.21155092120170593, "learning_rate": 8.467819873934714e-05, "loss": 0.0298, "step": 23460 }, { "grad_norm": 0.296421080827713, "learning_rate": 8.466330638381143e-05, "loss": 0.0305, "step": 23470 }, { "grad_norm": 0.30201882123947144, "learning_rate": 8.464840810526469e-05, "loss": 0.0305, "step": 23480 }, { "grad_norm": 0.39564254879951477, "learning_rate": 8.463350390625264e-05, "loss": 0.0398, "step": 23490 }, { "grad_norm": 0.25201743841171265, "learning_rate": 8.4618593789322e-05, "loss": 0.0528, "step": 23500 }, { "grad_norm": 0.3088453710079193, "learning_rate": 8.46036777570205e-05, "loss": 0.0396, "step": 23510 }, { "grad_norm": 0.23174230754375458, "learning_rate": 8.458875581189688e-05, "loss": 0.0561, "step": 23520 }, { "grad_norm": 0.5187253952026367, "learning_rate": 8.457382795650092e-05, "loss": 0.039, "step": 23530 }, { "grad_norm": 0.3086748719215393, "learning_rate": 8.455889419338335e-05, "loss": 0.0353, "step": 23540 }, { "grad_norm": 0.28287842869758606, "learning_rate": 8.454395452509593e-05, "loss": 0.0317, "step": 23550 }, { "grad_norm": 0.3136320412158966, "learning_rate": 8.452900895419146e-05, "loss": 0.0338, "step": 23560 }, { "grad_norm": 0.27225691080093384, "learning_rate": 8.451405748322376e-05, "loss": 0.0344, "step": 23570 }, { "grad_norm": 0.37389442324638367, "learning_rate": 8.449910011474759e-05, "loss": 0.0354, "step": 23580 }, { "grad_norm": 0.24858306348323822, "learning_rate": 8.448413685131876e-05, "loss": 0.0418, "step": 23590 }, { "grad_norm": 0.25770607590675354, "learning_rate": 8.446916769549407e-05, "loss": 0.0375, "step": 23600 }, { "grad_norm": 0.37837740778923035, "learning_rate": 8.445419264983136e-05, "loss": 0.0357, "step": 23610 }, { "grad_norm": 0.2815304398536682, "learning_rate": 8.443921171688947e-05, "loss": 0.0326, "step": 23620 }, { "grad_norm": 0.3416707217693329, "learning_rate": 8.442422489922819e-05, "loss": 0.0301, "step": 23630 }, { "grad_norm": 0.2576686441898346, "learning_rate": 8.440923219940838e-05, "loss": 0.0384, "step": 23640 }, { "grad_norm": 0.34431666135787964, "learning_rate": 8.439423361999189e-05, "loss": 0.0405, "step": 23650 }, { "grad_norm": 0.23645137250423431, "learning_rate": 8.437922916354155e-05, "loss": 0.0305, "step": 23660 }, { "grad_norm": 0.3708024024963379, "learning_rate": 8.436421883262123e-05, "loss": 0.0322, "step": 23670 }, { "grad_norm": 0.23450514674186707, "learning_rate": 8.434920262979577e-05, "loss": 0.0353, "step": 23680 }, { "grad_norm": 0.29668909311294556, "learning_rate": 8.433418055763104e-05, "loss": 0.0293, "step": 23690 }, { "grad_norm": 0.2832639813423157, "learning_rate": 8.431915261869389e-05, "loss": 0.0343, "step": 23700 }, { "grad_norm": 0.3142010271549225, "learning_rate": 8.43041188155522e-05, "loss": 0.034, "step": 23710 }, { "grad_norm": 0.31762921810150146, "learning_rate": 8.428907915077481e-05, "loss": 0.0332, "step": 23720 }, { "grad_norm": 0.3354977071285248, "learning_rate": 8.42740336269316e-05, "loss": 0.038, "step": 23730 }, { "grad_norm": 0.31045693159103394, "learning_rate": 8.425898224659345e-05, "loss": 0.0273, "step": 23740 }, { "grad_norm": 0.3105261027812958, "learning_rate": 8.42439250123322e-05, "loss": 0.0287, "step": 23750 }, { "grad_norm": 0.25783583521842957, "learning_rate": 8.422886192672076e-05, "loss": 0.0307, "step": 23760 }, { "grad_norm": 0.2603603005409241, "learning_rate": 8.421379299233297e-05, "loss": 0.0309, "step": 23770 }, { "grad_norm": 0.34373873472213745, "learning_rate": 8.419871821174371e-05, "loss": 0.0332, "step": 23780 }, { "grad_norm": 0.3254634439945221, "learning_rate": 8.418363758752884e-05, "loss": 0.035, "step": 23790 }, { "grad_norm": 0.2676459550857544, "learning_rate": 8.416855112226523e-05, "loss": 0.0329, "step": 23800 }, { "grad_norm": 0.31694114208221436, "learning_rate": 8.415345881853075e-05, "loss": 0.0319, "step": 23810 }, { "grad_norm": 0.22481094300746918, "learning_rate": 8.413836067890426e-05, "loss": 0.0307, "step": 23820 }, { "grad_norm": 0.3183846175670624, "learning_rate": 8.41232567059656e-05, "loss": 0.0295, "step": 23830 }, { "grad_norm": 0.21483144164085388, "learning_rate": 8.410814690229565e-05, "loss": 0.0327, "step": 23840 }, { "grad_norm": 0.3219496011734009, "learning_rate": 8.409303127047626e-05, "loss": 0.0292, "step": 23850 }, { "grad_norm": 0.5203520655632019, "learning_rate": 8.407790981309028e-05, "loss": 0.034, "step": 23860 }, { "grad_norm": 0.35283443331718445, "learning_rate": 8.406278253272153e-05, "loss": 0.0368, "step": 23870 }, { "grad_norm": 0.29441240429878235, "learning_rate": 8.404764943195487e-05, "loss": 0.0297, "step": 23880 }, { "grad_norm": 0.37449324131011963, "learning_rate": 8.403251051337613e-05, "loss": 0.0348, "step": 23890 }, { "grad_norm": 0.2667444348335266, "learning_rate": 8.401736577957214e-05, "loss": 0.0321, "step": 23900 }, { "grad_norm": 0.30086174607276917, "learning_rate": 8.40022152331307e-05, "loss": 0.0334, "step": 23910 }, { "grad_norm": 0.22248628735542297, "learning_rate": 8.398705887664064e-05, "loss": 0.04, "step": 23920 }, { "grad_norm": 0.26612791419029236, "learning_rate": 8.397189671269177e-05, "loss": 0.0297, "step": 23930 }, { "grad_norm": 0.33366936445236206, "learning_rate": 8.395672874387488e-05, "loss": 0.0318, "step": 23940 }, { "grad_norm": 0.2456272393465042, "learning_rate": 8.394155497278177e-05, "loss": 0.031, "step": 23950 }, { "grad_norm": 0.36843380331993103, "learning_rate": 8.392637540200523e-05, "loss": 0.0479, "step": 23960 }, { "grad_norm": 0.28894147276878357, "learning_rate": 8.391119003413902e-05, "loss": 0.0321, "step": 23970 }, { "grad_norm": 0.3396269977092743, "learning_rate": 8.38959988717779e-05, "loss": 0.0298, "step": 23980 }, { "grad_norm": 0.3449958264827728, "learning_rate": 8.388080191751764e-05, "loss": 0.0354, "step": 23990 }, { "grad_norm": 0.2965090572834015, "learning_rate": 8.386559917395496e-05, "loss": 0.0341, "step": 24000 }, { "grad_norm": 0.20066748559474945, "learning_rate": 8.385039064368761e-05, "loss": 0.0295, "step": 24010 }, { "grad_norm": 0.18844416737556458, "learning_rate": 8.383517632931431e-05, "loss": 0.0309, "step": 24020 }, { "grad_norm": 0.47693929076194763, "learning_rate": 8.381995623343477e-05, "loss": 0.0347, "step": 24030 }, { "grad_norm": 0.2701527178287506, "learning_rate": 8.380473035864968e-05, "loss": 0.0352, "step": 24040 }, { "grad_norm": 0.33929339051246643, "learning_rate": 8.378949870756076e-05, "loss": 0.0361, "step": 24050 }, { "grad_norm": 0.2727424204349518, "learning_rate": 8.377426128277063e-05, "loss": 0.0365, "step": 24060 }, { "grad_norm": 0.3053097426891327, "learning_rate": 8.375901808688298e-05, "loss": 0.0306, "step": 24070 }, { "grad_norm": 0.281878799200058, "learning_rate": 8.374376912250246e-05, "loss": 0.0408, "step": 24080 }, { "grad_norm": 0.21534471213817596, "learning_rate": 8.372851439223468e-05, "loss": 0.0286, "step": 24090 }, { "grad_norm": 0.27465805411338806, "learning_rate": 8.371325389868627e-05, "loss": 0.0367, "step": 24100 }, { "grad_norm": 0.2333792895078659, "learning_rate": 8.369798764446482e-05, "loss": 0.0293, "step": 24110 }, { "grad_norm": 0.31275948882102966, "learning_rate": 8.368271563217893e-05, "loss": 0.0267, "step": 24120 }, { "grad_norm": 0.28632014989852905, "learning_rate": 8.366743786443817e-05, "loss": 0.0382, "step": 24130 }, { "grad_norm": 0.38110285997390747, "learning_rate": 8.365215434385309e-05, "loss": 0.04, "step": 24140 }, { "grad_norm": 0.30963507294654846, "learning_rate": 8.36368650730352e-05, "loss": 0.0285, "step": 24150 }, { "grad_norm": 0.2702116072177887, "learning_rate": 8.362157005459705e-05, "loss": 0.0325, "step": 24160 }, { "grad_norm": 0.29460519552230835, "learning_rate": 8.360626929115213e-05, "loss": 0.0312, "step": 24170 }, { "grad_norm": 0.2891446352005005, "learning_rate": 8.359096278531492e-05, "loss": 0.0346, "step": 24180 }, { "grad_norm": 0.2947854995727539, "learning_rate": 8.357565053970088e-05, "loss": 0.0316, "step": 24190 }, { "grad_norm": 0.28985562920570374, "learning_rate": 8.356033255692647e-05, "loss": 0.026, "step": 24200 }, { "grad_norm": 0.2614394426345825, "learning_rate": 8.354500883960911e-05, "loss": 0.0306, "step": 24210 }, { "grad_norm": 0.27016639709472656, "learning_rate": 8.352967939036717e-05, "loss": 0.0327, "step": 24220 }, { "grad_norm": 0.30850276350975037, "learning_rate": 8.35143442118201e-05, "loss": 0.0395, "step": 24230 }, { "grad_norm": 0.2588769197463989, "learning_rate": 8.349900330658819e-05, "loss": 0.0284, "step": 24240 }, { "grad_norm": 0.6413942575454712, "learning_rate": 8.348365667729284e-05, "loss": 0.0365, "step": 24250 }, { "grad_norm": 0.3368961811065674, "learning_rate": 8.346830432655633e-05, "loss": 0.0284, "step": 24260 }, { "grad_norm": 0.34082165360450745, "learning_rate": 8.345294625700195e-05, "loss": 0.0342, "step": 24270 }, { "grad_norm": 0.3441898822784424, "learning_rate": 8.343758247125402e-05, "loss": 0.0345, "step": 24280 }, { "grad_norm": 0.28039056062698364, "learning_rate": 8.342221297193776e-05, "loss": 0.0354, "step": 24290 }, { "grad_norm": 0.30706149339675903, "learning_rate": 8.34068377616794e-05, "loss": 0.0398, "step": 24300 }, { "grad_norm": 0.3733520805835724, "learning_rate": 8.339145684310615e-05, "loss": 0.0371, "step": 24310 }, { "grad_norm": 0.33492666482925415, "learning_rate": 8.337607021884618e-05, "loss": 0.0326, "step": 24320 }, { "grad_norm": 0.22069741785526276, "learning_rate": 8.336067789152867e-05, "loss": 0.0299, "step": 24330 }, { "grad_norm": 0.275486558675766, "learning_rate": 8.334527986378369e-05, "loss": 0.0371, "step": 24340 }, { "grad_norm": 0.29865574836730957, "learning_rate": 8.332987613824239e-05, "loss": 0.0415, "step": 24350 }, { "grad_norm": 0.20417161285877228, "learning_rate": 8.331446671753685e-05, "loss": 0.031, "step": 24360 }, { "grad_norm": 0.2077489197254181, "learning_rate": 8.329905160430007e-05, "loss": 0.0326, "step": 24370 }, { "grad_norm": 0.2468155026435852, "learning_rate": 8.328363080116611e-05, "loss": 0.0266, "step": 24380 }, { "grad_norm": 0.21460668742656708, "learning_rate": 8.326820431076997e-05, "loss": 0.0291, "step": 24390 }, { "grad_norm": 0.2892761528491974, "learning_rate": 8.325277213574759e-05, "loss": 0.0382, "step": 24400 }, { "grad_norm": 0.250262051820755, "learning_rate": 8.32373342787359e-05, "loss": 0.0256, "step": 24410 }, { "grad_norm": 0.2813965082168579, "learning_rate": 8.322189074237285e-05, "loss": 0.0317, "step": 24420 }, { "grad_norm": 0.2810046076774597, "learning_rate": 8.32064415292973e-05, "loss": 0.0333, "step": 24430 }, { "grad_norm": 0.19930805265903473, "learning_rate": 8.319098664214907e-05, "loss": 0.0234, "step": 24440 }, { "grad_norm": 0.29790782928466797, "learning_rate": 8.3175526083569e-05, "loss": 0.0371, "step": 24450 }, { "grad_norm": 0.2683330774307251, "learning_rate": 8.316005985619889e-05, "loss": 0.0308, "step": 24460 }, { "grad_norm": 0.3084149658679962, "learning_rate": 8.314458796268147e-05, "loss": 0.0351, "step": 24470 }, { "grad_norm": 0.31699228286743164, "learning_rate": 8.312911040566047e-05, "loss": 0.0322, "step": 24480 }, { "grad_norm": 0.2293955385684967, "learning_rate": 8.31136271877806e-05, "loss": 0.0286, "step": 24490 }, { "grad_norm": 0.34256231784820557, "learning_rate": 8.309813831168748e-05, "loss": 0.0338, "step": 24500 }, { "grad_norm": 0.25269198417663574, "learning_rate": 8.308264378002777e-05, "loss": 0.0275, "step": 24510 }, { "grad_norm": 0.264904648065567, "learning_rate": 8.306714359544906e-05, "loss": 0.037, "step": 24520 }, { "grad_norm": 0.311322957277298, "learning_rate": 8.30516377605999e-05, "loss": 0.0297, "step": 24530 }, { "grad_norm": 0.27221912145614624, "learning_rate": 8.30361262781298e-05, "loss": 0.036, "step": 24540 }, { "grad_norm": 0.48558348417282104, "learning_rate": 8.302060915068924e-05, "loss": 0.036, "step": 24550 }, { "grad_norm": 0.19956864416599274, "learning_rate": 8.300508638092972e-05, "loss": 0.0276, "step": 24560 }, { "grad_norm": 0.2833218574523926, "learning_rate": 8.298955797150361e-05, "loss": 0.0278, "step": 24570 }, { "grad_norm": 0.32602658867836, "learning_rate": 8.297402392506433e-05, "loss": 0.0309, "step": 24580 }, { "grad_norm": 0.5111072063446045, "learning_rate": 8.295848424426617e-05, "loss": 0.0268, "step": 24590 }, { "grad_norm": 0.2542596161365509, "learning_rate": 8.29429389317645e-05, "loss": 0.029, "step": 24600 }, { "grad_norm": 0.2887077331542969, "learning_rate": 8.292738799021556e-05, "loss": 0.0352, "step": 24610 }, { "grad_norm": 0.22250311076641083, "learning_rate": 8.291183142227656e-05, "loss": 0.0302, "step": 24620 }, { "grad_norm": 0.18686677515506744, "learning_rate": 8.289626923060572e-05, "loss": 0.0374, "step": 24630 }, { "grad_norm": 0.2183740884065628, "learning_rate": 8.288070141786218e-05, "loss": 0.0356, "step": 24640 }, { "grad_norm": 0.35894399881362915, "learning_rate": 8.286512798670605e-05, "loss": 0.0295, "step": 24650 }, { "grad_norm": 0.2520447075366974, "learning_rate": 8.284954893979842e-05, "loss": 0.0337, "step": 24660 }, { "grad_norm": 0.27666816115379333, "learning_rate": 8.283396427980131e-05, "loss": 0.0349, "step": 24670 }, { "grad_norm": 0.3094097673892975, "learning_rate": 8.281837400937771e-05, "loss": 0.0338, "step": 24680 }, { "grad_norm": 0.23405714333057404, "learning_rate": 8.28027781311916e-05, "loss": 0.0363, "step": 24690 }, { "grad_norm": 0.26731187105178833, "learning_rate": 8.278717664790785e-05, "loss": 0.0324, "step": 24700 }, { "grad_norm": 0.29389357566833496, "learning_rate": 8.277156956219234e-05, "loss": 0.0315, "step": 24710 }, { "grad_norm": 0.3238888680934906, "learning_rate": 8.275595687671189e-05, "loss": 0.0326, "step": 24720 }, { "grad_norm": 0.37008488178253174, "learning_rate": 8.27403385941343e-05, "loss": 0.0363, "step": 24730 }, { "grad_norm": 0.26333796977996826, "learning_rate": 8.272471471712828e-05, "loss": 0.0342, "step": 24740 }, { "grad_norm": 0.2551065683364868, "learning_rate": 8.270908524836355e-05, "loss": 0.0361, "step": 24750 }, { "grad_norm": 0.5028234124183655, "learning_rate": 8.269345019051074e-05, "loss": 0.0406, "step": 24760 }, { "grad_norm": 0.2773648798465729, "learning_rate": 8.267780954624147e-05, "loss": 0.0357, "step": 24770 }, { "grad_norm": 0.29509100317955017, "learning_rate": 8.266216331822827e-05, "loss": 0.0356, "step": 24780 }, { "grad_norm": 0.2139730453491211, "learning_rate": 8.264651150914469e-05, "loss": 0.0305, "step": 24790 }, { "grad_norm": 0.2534809410572052, "learning_rate": 8.263085412166517e-05, "loss": 0.0262, "step": 24800 }, { "grad_norm": 0.25900113582611084, "learning_rate": 8.261519115846514e-05, "loss": 0.0322, "step": 24810 }, { "grad_norm": 0.20885100960731506, "learning_rate": 8.259952262222096e-05, "loss": 0.0294, "step": 24820 }, { "grad_norm": 0.26285818219184875, "learning_rate": 8.258384851560997e-05, "loss": 0.0337, "step": 24830 }, { "grad_norm": 0.4134880602359772, "learning_rate": 8.256816884131044e-05, "loss": 0.0354, "step": 24840 }, { "grad_norm": 0.3019576072692871, "learning_rate": 8.255248360200159e-05, "loss": 0.0317, "step": 24850 }, { "grad_norm": 0.23184695839881897, "learning_rate": 8.253679280036359e-05, "loss": 0.0262, "step": 24860 }, { "grad_norm": 0.273769736289978, "learning_rate": 8.252109643907762e-05, "loss": 0.0464, "step": 24870 }, { "grad_norm": 0.31938761472702026, "learning_rate": 8.250539452082569e-05, "loss": 0.0277, "step": 24880 }, { "grad_norm": 0.27494680881500244, "learning_rate": 8.248968704829087e-05, "loss": 0.0314, "step": 24890 }, { "grad_norm": 0.2885187864303589, "learning_rate": 8.247397402415714e-05, "loss": 0.0394, "step": 24900 }, { "grad_norm": 0.2829539477825165, "learning_rate": 8.24582554511094e-05, "loss": 0.0304, "step": 24910 }, { "grad_norm": 0.24420975148677826, "learning_rate": 8.244253133183355e-05, "loss": 0.029, "step": 24920 }, { "grad_norm": 0.3492058515548706, "learning_rate": 8.24268016690164e-05, "loss": 0.0307, "step": 24930 }, { "grad_norm": 0.22265154123306274, "learning_rate": 8.241106646534571e-05, "loss": 0.0353, "step": 24940 }, { "grad_norm": 0.3207402229309082, "learning_rate": 8.23953257235102e-05, "loss": 0.0341, "step": 24950 }, { "grad_norm": 0.26225629448890686, "learning_rate": 8.237957944619956e-05, "loss": 0.033, "step": 24960 }, { "grad_norm": 0.24455982446670532, "learning_rate": 8.236382763610437e-05, "loss": 0.0286, "step": 24970 }, { "grad_norm": 0.35920268297195435, "learning_rate": 8.234807029591619e-05, "loss": 0.0303, "step": 24980 }, { "grad_norm": 0.2871720790863037, "learning_rate": 8.233230742832752e-05, "loss": 0.0472, "step": 24990 }, { "grad_norm": 0.39794856309890747, "learning_rate": 8.231653903603178e-05, "loss": 0.0369, "step": 25000 }, { "grad_norm": 0.3352513611316681, "learning_rate": 8.23007651217234e-05, "loss": 0.0359, "step": 25010 }, { "grad_norm": 0.3257955014705658, "learning_rate": 8.228498568809769e-05, "loss": 0.0356, "step": 25020 }, { "grad_norm": 0.23040002584457397, "learning_rate": 8.22692007378509e-05, "loss": 0.0362, "step": 25030 }, { "grad_norm": 0.2589449882507324, "learning_rate": 8.225341027368028e-05, "loss": 0.0316, "step": 25040 }, { "grad_norm": 0.44811776280403137, "learning_rate": 8.223761429828399e-05, "loss": 0.0343, "step": 25050 }, { "grad_norm": 0.26829421520233154, "learning_rate": 8.22218128143611e-05, "loss": 0.0321, "step": 25060 }, { "grad_norm": 0.25760549306869507, "learning_rate": 8.220600582461166e-05, "loss": 0.0321, "step": 25070 }, { "grad_norm": 0.27449971437454224, "learning_rate": 8.219019333173668e-05, "loss": 0.0298, "step": 25080 }, { "grad_norm": 0.2517065405845642, "learning_rate": 8.217437533843805e-05, "loss": 0.0356, "step": 25090 }, { "grad_norm": 0.2439025640487671, "learning_rate": 8.215855184741867e-05, "loss": 0.0325, "step": 25100 }, { "grad_norm": 0.26501280069351196, "learning_rate": 8.21427228613823e-05, "loss": 0.03, "step": 25110 }, { "grad_norm": 0.23148640990257263, "learning_rate": 8.21268883830337e-05, "loss": 0.0312, "step": 25120 }, { "grad_norm": 0.3049479126930237, "learning_rate": 8.211104841507855e-05, "loss": 0.0433, "step": 25130 }, { "grad_norm": 0.27198976278305054, "learning_rate": 8.209520296022346e-05, "loss": 0.0344, "step": 25140 }, { "grad_norm": 0.31366145610809326, "learning_rate": 8.207935202117599e-05, "loss": 0.0355, "step": 25150 }, { "grad_norm": 0.328447163105011, "learning_rate": 8.206349560064463e-05, "loss": 0.0328, "step": 25160 }, { "grad_norm": 0.36003515124320984, "learning_rate": 8.204763370133881e-05, "loss": 0.0294, "step": 25170 }, { "grad_norm": 0.31508541107177734, "learning_rate": 8.203176632596892e-05, "loss": 0.03, "step": 25180 }, { "grad_norm": 0.4370879828929901, "learning_rate": 8.20158934772462e-05, "loss": 0.0389, "step": 25190 }, { "grad_norm": 0.2644104063510895, "learning_rate": 8.200001515788294e-05, "loss": 0.0242, "step": 25200 }, { "grad_norm": 0.339889794588089, "learning_rate": 8.198413137059228e-05, "loss": 0.0356, "step": 25210 }, { "grad_norm": 0.30586427450180054, "learning_rate": 8.196824211808835e-05, "loss": 0.029, "step": 25220 }, { "grad_norm": 0.32826775312423706, "learning_rate": 8.195234740308617e-05, "loss": 0.0331, "step": 25230 }, { "grad_norm": 0.32424792647361755, "learning_rate": 8.193644722830171e-05, "loss": 0.0243, "step": 25240 }, { "grad_norm": 0.30925241112709045, "learning_rate": 8.19205415964519e-05, "loss": 0.0265, "step": 25250 }, { "grad_norm": 0.3037685751914978, "learning_rate": 8.190463051025456e-05, "loss": 0.0283, "step": 25260 }, { "grad_norm": 0.2505023181438446, "learning_rate": 8.188871397242843e-05, "loss": 0.0305, "step": 25270 }, { "grad_norm": 0.21755078434944153, "learning_rate": 8.187279198569326e-05, "loss": 0.0329, "step": 25280 }, { "grad_norm": 0.23019038140773773, "learning_rate": 8.185686455276966e-05, "loss": 0.0308, "step": 25290 }, { "grad_norm": 0.32705411314964294, "learning_rate": 8.184093167637921e-05, "loss": 0.0322, "step": 25300 }, { "grad_norm": 0.2555890381336212, "learning_rate": 8.182499335924437e-05, "loss": 0.0342, "step": 25310 }, { "grad_norm": 0.22705523669719696, "learning_rate": 8.18090496040886e-05, "loss": 0.0264, "step": 25320 }, { "grad_norm": 0.2629956603050232, "learning_rate": 8.179310041363621e-05, "loss": 0.0375, "step": 25330 }, { "grad_norm": 0.2517438232898712, "learning_rate": 8.17771457906125e-05, "loss": 0.0275, "step": 25340 }, { "grad_norm": 0.25995323061943054, "learning_rate": 8.176118573774371e-05, "loss": 0.0288, "step": 25350 }, { "grad_norm": 0.26054009795188904, "learning_rate": 8.174522025775692e-05, "loss": 0.0275, "step": 25360 }, { "grad_norm": 0.31448453664779663, "learning_rate": 8.172924935338022e-05, "loss": 0.0295, "step": 25370 }, { "grad_norm": 0.18810871243476868, "learning_rate": 8.171327302734262e-05, "loss": 0.0247, "step": 25380 }, { "grad_norm": 0.21784541010856628, "learning_rate": 8.169729128237401e-05, "loss": 0.026, "step": 25390 }, { "grad_norm": 0.3226662278175354, "learning_rate": 8.168130412120525e-05, "loss": 0.0269, "step": 25400 }, { "grad_norm": 0.29069116711616516, "learning_rate": 8.16653115465681e-05, "loss": 0.032, "step": 25410 }, { "grad_norm": 0.2996164858341217, "learning_rate": 8.164931356119526e-05, "loss": 0.0246, "step": 25420 }, { "grad_norm": 0.3794037103652954, "learning_rate": 8.163331016782032e-05, "loss": 0.0394, "step": 25430 }, { "grad_norm": 0.27970874309539795, "learning_rate": 8.161730136917785e-05, "loss": 0.0268, "step": 25440 }, { "grad_norm": 0.23840615153312683, "learning_rate": 8.160128716800333e-05, "loss": 0.0386, "step": 25450 }, { "grad_norm": 0.3186179995536804, "learning_rate": 8.158526756703313e-05, "loss": 0.0282, "step": 25460 }, { "grad_norm": 0.30678921937942505, "learning_rate": 8.156924256900455e-05, "loss": 0.0338, "step": 25470 }, { "grad_norm": 0.28276050090789795, "learning_rate": 8.155321217665584e-05, "loss": 0.033, "step": 25480 }, { "grad_norm": 0.23345354199409485, "learning_rate": 8.153717639272614e-05, "loss": 0.0272, "step": 25490 }, { "grad_norm": 0.3316623866558075, "learning_rate": 8.152113521995555e-05, "loss": 0.0343, "step": 25500 }, { "grad_norm": 0.2715848684310913, "learning_rate": 8.150508866108505e-05, "loss": 0.0279, "step": 25510 }, { "grad_norm": 0.24494513869285583, "learning_rate": 8.148903671885657e-05, "loss": 0.0299, "step": 25520 }, { "grad_norm": 0.28286096453666687, "learning_rate": 8.147297939601292e-05, "loss": 0.0374, "step": 25530 }, { "grad_norm": 0.25076037645339966, "learning_rate": 8.145691669529792e-05, "loss": 0.0305, "step": 25540 }, { "grad_norm": 0.21322105824947357, "learning_rate": 8.144084861945618e-05, "loss": 0.028, "step": 25550 }, { "grad_norm": 0.282100111246109, "learning_rate": 8.142477517123333e-05, "loss": 0.0318, "step": 25560 }, { "grad_norm": 0.3077922463417053, "learning_rate": 8.140869635337586e-05, "loss": 0.0307, "step": 25570 }, { "grad_norm": 0.266065388917923, "learning_rate": 8.139261216863123e-05, "loss": 0.0289, "step": 25580 }, { "grad_norm": 0.2663378417491913, "learning_rate": 8.137652261974776e-05, "loss": 0.0303, "step": 25590 }, { "grad_norm": 0.3138018250465393, "learning_rate": 8.136042770947472e-05, "loss": 0.0295, "step": 25600 }, { "grad_norm": 0.26496773958206177, "learning_rate": 8.134432744056228e-05, "loss": 0.0324, "step": 25610 }, { "grad_norm": 0.3304949700832367, "learning_rate": 8.132822181576158e-05, "loss": 0.0402, "step": 25620 }, { "grad_norm": 0.29457974433898926, "learning_rate": 8.131211083782459e-05, "loss": 0.0345, "step": 25630 }, { "grad_norm": 0.31795719265937805, "learning_rate": 8.129599450950424e-05, "loss": 0.0314, "step": 25640 }, { "grad_norm": 0.24230322241783142, "learning_rate": 8.127987283355438e-05, "loss": 0.0375, "step": 25650 }, { "grad_norm": 0.278136283159256, "learning_rate": 8.126374581272976e-05, "loss": 0.0266, "step": 25660 }, { "grad_norm": 0.28614696860313416, "learning_rate": 8.124761344978605e-05, "loss": 0.041, "step": 25670 }, { "grad_norm": 0.2769566774368286, "learning_rate": 8.12314757474798e-05, "loss": 0.0298, "step": 25680 }, { "grad_norm": 0.27784791588783264, "learning_rate": 8.121533270856856e-05, "loss": 0.0283, "step": 25690 }, { "grad_norm": 0.26788726449012756, "learning_rate": 8.119918433581069e-05, "loss": 0.0288, "step": 25700 }, { "grad_norm": 0.35372641682624817, "learning_rate": 8.118303063196551e-05, "loss": 0.0322, "step": 25710 }, { "grad_norm": 0.35659751296043396, "learning_rate": 8.116687159979326e-05, "loss": 0.031, "step": 25720 }, { "grad_norm": 0.22395187616348267, "learning_rate": 8.115070724205508e-05, "loss": 0.0346, "step": 25730 }, { "grad_norm": 0.3388454020023346, "learning_rate": 8.113453756151296e-05, "loss": 0.0305, "step": 25740 }, { "grad_norm": 0.30230608582496643, "learning_rate": 8.111836256092995e-05, "loss": 0.0302, "step": 25750 }, { "grad_norm": 0.18835484981536865, "learning_rate": 8.110218224306985e-05, "loss": 0.0334, "step": 25760 }, { "grad_norm": 0.27039065957069397, "learning_rate": 8.108599661069745e-05, "loss": 0.0269, "step": 25770 }, { "grad_norm": 0.2969982624053955, "learning_rate": 8.106980566657845e-05, "loss": 0.0304, "step": 25780 }, { "grad_norm": 0.2704072594642639, "learning_rate": 8.10536094134794e-05, "loss": 0.0299, "step": 25790 }, { "grad_norm": 0.3607017397880554, "learning_rate": 8.103740785416783e-05, "loss": 0.0367, "step": 25800 }, { "grad_norm": 0.2533703148365021, "learning_rate": 8.102120099141212e-05, "loss": 0.0328, "step": 25810 }, { "grad_norm": 0.3532574474811554, "learning_rate": 8.100498882798163e-05, "loss": 0.0285, "step": 25820 }, { "grad_norm": 0.3083517551422119, "learning_rate": 8.09887713666465e-05, "loss": 0.0299, "step": 25830 }, { "grad_norm": 0.2622796893119812, "learning_rate": 8.09725486101779e-05, "loss": 0.0353, "step": 25840 }, { "grad_norm": 0.3576100468635559, "learning_rate": 8.095632056134784e-05, "loss": 0.0339, "step": 25850 }, { "grad_norm": 0.3078608810901642, "learning_rate": 8.094008722292925e-05, "loss": 0.0268, "step": 25860 }, { "grad_norm": 0.26467666029930115, "learning_rate": 8.092384859769598e-05, "loss": 0.0338, "step": 25870 }, { "grad_norm": 0.33493947982788086, "learning_rate": 8.090760468842275e-05, "loss": 0.042, "step": 25880 }, { "grad_norm": 0.2636958062648773, "learning_rate": 8.089135549788521e-05, "loss": 0.0313, "step": 25890 }, { "grad_norm": 0.31573614478111267, "learning_rate": 8.087510102885987e-05, "loss": 0.0258, "step": 25900 }, { "grad_norm": 0.25136467814445496, "learning_rate": 8.085884128412422e-05, "loss": 0.0355, "step": 25910 }, { "grad_norm": 0.263031542301178, "learning_rate": 8.084257626645659e-05, "loss": 0.0302, "step": 25920 }, { "grad_norm": 0.2977589964866638, "learning_rate": 8.08263059786362e-05, "loss": 0.0304, "step": 25930 }, { "grad_norm": 0.21982406079769135, "learning_rate": 8.081003042344325e-05, "loss": 0.026, "step": 25940 }, { "grad_norm": 0.25214847922325134, "learning_rate": 8.079374960365872e-05, "loss": 0.0284, "step": 25950 }, { "grad_norm": 0.2695353329181671, "learning_rate": 8.077746352206463e-05, "loss": 0.0314, "step": 25960 }, { "grad_norm": 0.3037709593772888, "learning_rate": 8.076117218144377e-05, "loss": 0.0291, "step": 25970 }, { "grad_norm": 0.30155032873153687, "learning_rate": 8.074487558457991e-05, "loss": 0.0281, "step": 25980 }, { "grad_norm": 0.2428794503211975, "learning_rate": 8.072857373425768e-05, "loss": 0.0289, "step": 25990 }, { "grad_norm": 0.31886667013168335, "learning_rate": 8.071226663326264e-05, "loss": 0.0296, "step": 26000 }, { "grad_norm": 0.33867147564888, "learning_rate": 8.069595428438121e-05, "loss": 0.0372, "step": 26010 }, { "grad_norm": 0.29598763585090637, "learning_rate": 8.067963669040072e-05, "loss": 0.0259, "step": 26020 }, { "grad_norm": 0.35034120082855225, "learning_rate": 8.066331385410942e-05, "loss": 0.0318, "step": 26030 }, { "grad_norm": 0.2717263698577881, "learning_rate": 8.064698577829641e-05, "loss": 0.0367, "step": 26040 }, { "grad_norm": 0.2564922869205475, "learning_rate": 8.063065246575175e-05, "loss": 0.0312, "step": 26050 }, { "grad_norm": 0.3055976331233978, "learning_rate": 8.061431391926631e-05, "loss": 0.0277, "step": 26060 }, { "grad_norm": 0.2894918918609619, "learning_rate": 8.059797014163195e-05, "loss": 0.0336, "step": 26070 }, { "grad_norm": 0.3052563965320587, "learning_rate": 8.058162113564133e-05, "loss": 0.0455, "step": 26080 }, { "grad_norm": 0.2795119881629944, "learning_rate": 8.056526690408806e-05, "loss": 0.0341, "step": 26090 }, { "grad_norm": 0.3514683246612549, "learning_rate": 8.054890744976666e-05, "loss": 0.0389, "step": 26100 }, { "grad_norm": 0.25081831216812134, "learning_rate": 8.053254277547248e-05, "loss": 0.0262, "step": 26110 }, { "grad_norm": 0.19892717897891998, "learning_rate": 8.051617288400182e-05, "loss": 0.0292, "step": 26120 }, { "grad_norm": 0.21810290217399597, "learning_rate": 8.049979777815182e-05, "loss": 0.0312, "step": 26130 }, { "grad_norm": 0.28788384795188904, "learning_rate": 8.048341746072054e-05, "loss": 0.0302, "step": 26140 }, { "grad_norm": 0.26155951619148254, "learning_rate": 8.046703193450696e-05, "loss": 0.044, "step": 26150 }, { "grad_norm": 0.34866222739219666, "learning_rate": 8.04506412023109e-05, "loss": 0.0283, "step": 26160 }, { "grad_norm": 0.20310015976428986, "learning_rate": 8.043424526693306e-05, "loss": 0.0286, "step": 26170 }, { "grad_norm": 0.2722871005535126, "learning_rate": 8.04178441311751e-05, "loss": 0.0308, "step": 26180 }, { "grad_norm": 0.25988176465034485, "learning_rate": 8.04014377978395e-05, "loss": 0.0296, "step": 26190 }, { "grad_norm": 0.2846917510032654, "learning_rate": 8.038502626972967e-05, "loss": 0.0333, "step": 26200 }, { "grad_norm": 0.3168918490409851, "learning_rate": 8.036860954964989e-05, "loss": 0.0374, "step": 26210 }, { "grad_norm": 0.24201498925685883, "learning_rate": 8.035218764040531e-05, "loss": 0.0331, "step": 26220 }, { "grad_norm": 0.21703983843326569, "learning_rate": 8.033576054480199e-05, "loss": 0.0291, "step": 26230 }, { "grad_norm": 0.3728594481945038, "learning_rate": 8.031932826564688e-05, "loss": 0.0309, "step": 26240 }, { "grad_norm": 0.2161438763141632, "learning_rate": 8.030289080574782e-05, "loss": 0.037, "step": 26250 }, { "grad_norm": 0.2552071213722229, "learning_rate": 8.028644816791349e-05, "loss": 0.033, "step": 26260 }, { "grad_norm": 0.2800721824169159, "learning_rate": 8.027000035495351e-05, "loss": 0.0253, "step": 26270 }, { "grad_norm": 0.39868733286857605, "learning_rate": 8.025354736967836e-05, "loss": 0.0488, "step": 26280 }, { "grad_norm": 0.2664601504802704, "learning_rate": 8.023708921489941e-05, "loss": 0.0465, "step": 26290 }, { "grad_norm": 0.30347391963005066, "learning_rate": 8.022062589342887e-05, "loss": 0.0395, "step": 26300 }, { "grad_norm": 0.26483720541000366, "learning_rate": 8.020415740807993e-05, "loss": 0.0423, "step": 26310 }, { "grad_norm": 0.24409198760986328, "learning_rate": 8.018768376166656e-05, "loss": 0.0328, "step": 26320 }, { "grad_norm": 0.25802329182624817, "learning_rate": 8.017120495700368e-05, "loss": 0.0371, "step": 26330 }, { "grad_norm": 0.2932744026184082, "learning_rate": 8.015472099690704e-05, "loss": 0.0342, "step": 26340 }, { "grad_norm": 0.2209317982196808, "learning_rate": 8.013823188419332e-05, "loss": 0.0276, "step": 26350 }, { "grad_norm": 0.262692391872406, "learning_rate": 8.012173762168006e-05, "loss": 0.0321, "step": 26360 }, { "grad_norm": 0.2369125932455063, "learning_rate": 8.010523821218567e-05, "loss": 0.0299, "step": 26370 }, { "grad_norm": 0.302130788564682, "learning_rate": 8.008873365852945e-05, "loss": 0.0322, "step": 26380 }, { "grad_norm": 0.2704709470272064, "learning_rate": 8.007222396353157e-05, "loss": 0.0328, "step": 26390 }, { "grad_norm": 0.3083808124065399, "learning_rate": 8.00557091300131e-05, "loss": 0.03, "step": 26400 }, { "grad_norm": 0.22933508455753326, "learning_rate": 8.003918916079597e-05, "loss": 0.0306, "step": 26410 }, { "grad_norm": 0.4884653389453888, "learning_rate": 8.002266405870298e-05, "loss": 0.0331, "step": 26420 }, { "grad_norm": 0.25874754786491394, "learning_rate": 8.000613382655782e-05, "loss": 0.0336, "step": 26430 }, { "grad_norm": 0.27675557136535645, "learning_rate": 7.998959846718505e-05, "loss": 0.0332, "step": 26440 }, { "grad_norm": 0.21485772728919983, "learning_rate": 7.997305798341012e-05, "loss": 0.0324, "step": 26450 }, { "grad_norm": 0.3541629910469055, "learning_rate": 7.995651237805937e-05, "loss": 0.039, "step": 26460 }, { "grad_norm": 0.2650720775127411, "learning_rate": 7.993996165395996e-05, "loss": 0.0305, "step": 26470 }, { "grad_norm": 0.3325348496437073, "learning_rate": 7.992340581393996e-05, "loss": 0.0311, "step": 26480 }, { "grad_norm": 0.2828216850757599, "learning_rate": 7.990684486082831e-05, "loss": 0.0357, "step": 26490 }, { "grad_norm": 0.3734305500984192, "learning_rate": 7.989027879745482e-05, "loss": 0.0417, "step": 26500 }, { "grad_norm": 0.24767443537712097, "learning_rate": 7.98737076266502e-05, "loss": 0.0421, "step": 26510 }, { "grad_norm": 0.26369765400886536, "learning_rate": 7.985713135124598e-05, "loss": 0.0406, "step": 26520 }, { "grad_norm": 0.31220367550849915, "learning_rate": 7.98405499740746e-05, "loss": 0.0365, "step": 26530 }, { "grad_norm": 0.2981390357017517, "learning_rate": 7.98239634979694e-05, "loss": 0.0395, "step": 26540 }, { "grad_norm": 0.3605949878692627, "learning_rate": 7.98073719257645e-05, "loss": 0.0389, "step": 26550 }, { "grad_norm": 0.37845608592033386, "learning_rate": 7.979077526029499e-05, "loss": 0.0449, "step": 26560 }, { "grad_norm": 0.39044031500816345, "learning_rate": 7.977417350439675e-05, "loss": 0.032, "step": 26570 }, { "grad_norm": 0.37401309609413147, "learning_rate": 7.97575666609066e-05, "loss": 0.0318, "step": 26580 }, { "grad_norm": 0.3074793219566345, "learning_rate": 7.974095473266216e-05, "loss": 0.0351, "step": 26590 }, { "grad_norm": 0.43403634428977966, "learning_rate": 7.972433772250198e-05, "loss": 0.0326, "step": 26600 }, { "grad_norm": 0.2985951602458954, "learning_rate": 7.970771563326544e-05, "loss": 0.0331, "step": 26610 }, { "grad_norm": 0.3097088634967804, "learning_rate": 7.96910884677928e-05, "loss": 0.0366, "step": 26620 }, { "grad_norm": 0.43390798568725586, "learning_rate": 7.967445622892523e-05, "loss": 0.0373, "step": 26630 }, { "grad_norm": 0.28456148505210876, "learning_rate": 7.965781891950465e-05, "loss": 0.0335, "step": 26640 }, { "grad_norm": 0.23387475311756134, "learning_rate": 7.964117654237397e-05, "loss": 0.026, "step": 26650 }, { "grad_norm": 0.3129494786262512, "learning_rate": 7.962452910037692e-05, "loss": 0.0392, "step": 26660 }, { "grad_norm": 0.2942778468132019, "learning_rate": 7.96078765963581e-05, "loss": 0.0334, "step": 26670 }, { "grad_norm": 0.24277141690254211, "learning_rate": 7.95912190331629e-05, "loss": 0.0354, "step": 26680 }, { "grad_norm": 0.2556321918964386, "learning_rate": 7.957455641363772e-05, "loss": 0.0354, "step": 26690 }, { "grad_norm": 0.30870741605758667, "learning_rate": 7.955788874062968e-05, "loss": 0.0431, "step": 26700 }, { "grad_norm": 0.31012988090515137, "learning_rate": 7.95412160169869e-05, "loss": 0.0367, "step": 26710 }, { "grad_norm": 0.33241015672683716, "learning_rate": 7.952453824555824e-05, "loss": 0.0322, "step": 26720 }, { "grad_norm": 0.2695174217224121, "learning_rate": 7.95078554291935e-05, "loss": 0.0313, "step": 26730 }, { "grad_norm": 0.2363610863685608, "learning_rate": 7.94911675707433e-05, "loss": 0.0305, "step": 26740 }, { "grad_norm": 0.2505609393119812, "learning_rate": 7.947447467305915e-05, "loss": 0.0395, "step": 26750 }, { "grad_norm": 0.287634015083313, "learning_rate": 7.94577767389934e-05, "loss": 0.0368, "step": 26760 }, { "grad_norm": 0.30816125869750977, "learning_rate": 7.944107377139928e-05, "loss": 0.0312, "step": 26770 }, { "grad_norm": 0.3127407729625702, "learning_rate": 7.942436577313088e-05, "loss": 0.0334, "step": 26780 }, { "grad_norm": 0.2614028751850128, "learning_rate": 7.940765274704312e-05, "loss": 0.0328, "step": 26790 }, { "grad_norm": 0.3697313964366913, "learning_rate": 7.939093469599181e-05, "loss": 0.0332, "step": 26800 }, { "grad_norm": 0.316038578748703, "learning_rate": 7.93742116228336e-05, "loss": 0.0304, "step": 26810 }, { "grad_norm": 0.3121386468410492, "learning_rate": 7.935748353042602e-05, "loss": 0.0316, "step": 26820 }, { "grad_norm": 0.29978588223457336, "learning_rate": 7.934075042162744e-05, "loss": 0.0291, "step": 26830 }, { "grad_norm": 0.31475114822387695, "learning_rate": 7.932401229929705e-05, "loss": 0.0419, "step": 26840 }, { "grad_norm": 0.28759101033210754, "learning_rate": 7.9307269166295e-05, "loss": 0.0345, "step": 26850 }, { "grad_norm": 0.26322513818740845, "learning_rate": 7.92905210254822e-05, "loss": 0.0305, "step": 26860 }, { "grad_norm": 0.31598421931266785, "learning_rate": 7.927376787972045e-05, "loss": 0.033, "step": 26870 }, { "grad_norm": 0.28173762559890747, "learning_rate": 7.92570097318724e-05, "loss": 0.0353, "step": 26880 }, { "grad_norm": 0.27417904138565063, "learning_rate": 7.924024658480158e-05, "loss": 0.0302, "step": 26890 }, { "grad_norm": 0.2647964060306549, "learning_rate": 7.922347844137233e-05, "loss": 0.0337, "step": 26900 }, { "grad_norm": 0.2976597547531128, "learning_rate": 7.92067053044499e-05, "loss": 0.0352, "step": 26910 }, { "grad_norm": 0.46688786149024963, "learning_rate": 7.918992717690031e-05, "loss": 0.035, "step": 26920 }, { "grad_norm": 0.2884725332260132, "learning_rate": 7.917314406159053e-05, "loss": 0.0346, "step": 26930 }, { "grad_norm": 0.3582327365875244, "learning_rate": 7.915635596138832e-05, "loss": 0.0425, "step": 26940 }, { "grad_norm": 0.2325042188167572, "learning_rate": 7.913956287916228e-05, "loss": 0.0376, "step": 26950 }, { "grad_norm": 0.2838636338710785, "learning_rate": 7.912276481778193e-05, "loss": 0.0271, "step": 26960 }, { "grad_norm": 0.24338102340698242, "learning_rate": 7.910596178011759e-05, "loss": 0.0364, "step": 26970 }, { "grad_norm": 0.23119018971920013, "learning_rate": 7.908915376904043e-05, "loss": 0.0289, "step": 26980 }, { "grad_norm": 0.24362339079380035, "learning_rate": 7.907234078742247e-05, "loss": 0.035, "step": 26990 }, { "grad_norm": 0.21863360702991486, "learning_rate": 7.90555228381366e-05, "loss": 0.0332, "step": 27000 }, { "grad_norm": 0.23589420318603516, "learning_rate": 7.903869992405656e-05, "loss": 0.0312, "step": 27010 }, { "grad_norm": 0.2590305507183075, "learning_rate": 7.902187204805691e-05, "loss": 0.0339, "step": 27020 }, { "grad_norm": 0.268661767244339, "learning_rate": 7.900503921301308e-05, "loss": 0.0281, "step": 27030 }, { "grad_norm": 0.2351028025150299, "learning_rate": 7.898820142180133e-05, "loss": 0.0345, "step": 27040 }, { "grad_norm": 0.26859816908836365, "learning_rate": 7.897135867729879e-05, "loss": 0.032, "step": 27050 }, { "grad_norm": 0.2218165099620819, "learning_rate": 7.89545109823834e-05, "loss": 0.0239, "step": 27060 }, { "grad_norm": 0.3225618302822113, "learning_rate": 7.8937658339934e-05, "loss": 0.0305, "step": 27070 }, { "grad_norm": 0.3736024796962738, "learning_rate": 7.892080075283026e-05, "loss": 0.0323, "step": 27080 }, { "grad_norm": 0.2505651116371155, "learning_rate": 7.890393822395263e-05, "loss": 0.0352, "step": 27090 }, { "grad_norm": 0.3670938313007355, "learning_rate": 7.88870707561825e-05, "loss": 0.0319, "step": 27100 }, { "grad_norm": 0.2993334233760834, "learning_rate": 7.887019835240203e-05, "loss": 0.0323, "step": 27110 }, { "grad_norm": 0.30188611149787903, "learning_rate": 7.885332101549427e-05, "loss": 0.0357, "step": 27120 }, { "grad_norm": 0.3413648009300232, "learning_rate": 7.883643874834308e-05, "loss": 0.0297, "step": 27130 }, { "grad_norm": 0.30086544156074524, "learning_rate": 7.881955155383321e-05, "loss": 0.0381, "step": 27140 }, { "grad_norm": 0.581155002117157, "learning_rate": 7.880265943485017e-05, "loss": 0.034, "step": 27150 }, { "grad_norm": 0.3079436123371124, "learning_rate": 7.878576239428038e-05, "loss": 0.029, "step": 27160 }, { "grad_norm": 0.21148788928985596, "learning_rate": 7.87688604350111e-05, "loss": 0.0341, "step": 27170 }, { "grad_norm": 0.3872378170490265, "learning_rate": 7.875195355993042e-05, "loss": 0.0384, "step": 27180 }, { "grad_norm": 0.29458072781562805, "learning_rate": 7.873504177192724e-05, "loss": 0.0361, "step": 27190 }, { "grad_norm": 0.30794668197631836, "learning_rate": 7.87181250738913e-05, "loss": 0.0409, "step": 27200 }, { "grad_norm": 0.2546442449092865, "learning_rate": 7.870120346871324e-05, "loss": 0.0311, "step": 27210 }, { "grad_norm": 0.31719648838043213, "learning_rate": 7.86842769592845e-05, "loss": 0.0303, "step": 27220 }, { "grad_norm": 0.3092624247074127, "learning_rate": 7.866734554849732e-05, "loss": 0.0285, "step": 27230 }, { "grad_norm": 0.2739826738834381, "learning_rate": 7.865040923924486e-05, "loss": 0.0275, "step": 27240 }, { "grad_norm": 0.3482922613620758, "learning_rate": 7.863346803442104e-05, "loss": 0.036, "step": 27250 }, { "grad_norm": 0.3147919476032257, "learning_rate": 7.861652193692067e-05, "loss": 0.0277, "step": 27260 }, { "grad_norm": 0.22108004987239838, "learning_rate": 7.859957094963937e-05, "loss": 0.0261, "step": 27270 }, { "grad_norm": 0.2830458879470825, "learning_rate": 7.858261507547357e-05, "loss": 0.0337, "step": 27280 }, { "grad_norm": 0.21870671212673187, "learning_rate": 7.856565431732061e-05, "loss": 0.0277, "step": 27290 }, { "grad_norm": 0.2374696582555771, "learning_rate": 7.854868867807859e-05, "loss": 0.0245, "step": 27300 }, { "grad_norm": 0.30467304587364197, "learning_rate": 7.85317181606465e-05, "loss": 0.0321, "step": 27310 }, { "grad_norm": 0.32927608489990234, "learning_rate": 7.85147427679241e-05, "loss": 0.0254, "step": 27320 }, { "grad_norm": 0.2984445095062256, "learning_rate": 7.849776250281205e-05, "loss": 0.0267, "step": 27330 }, { "grad_norm": 0.17534780502319336, "learning_rate": 7.84807773682118e-05, "loss": 0.0291, "step": 27340 }, { "grad_norm": 0.2936968207359314, "learning_rate": 7.846378736702565e-05, "loss": 0.0321, "step": 27350 }, { "grad_norm": 0.35357925295829773, "learning_rate": 7.844679250215671e-05, "loss": 0.0288, "step": 27360 }, { "grad_norm": 0.28325948119163513, "learning_rate": 7.842979277650898e-05, "loss": 0.0263, "step": 27370 }, { "grad_norm": 0.2865024507045746, "learning_rate": 7.84127881929872e-05, "loss": 0.0287, "step": 27380 }, { "grad_norm": 0.2728317975997925, "learning_rate": 7.839577875449704e-05, "loss": 0.0284, "step": 27390 }, { "grad_norm": 0.4019046127796173, "learning_rate": 7.837876446394489e-05, "loss": 0.0273, "step": 27400 }, { "grad_norm": 0.22953271865844727, "learning_rate": 7.836174532423805e-05, "loss": 0.03, "step": 27410 }, { "grad_norm": 0.27163660526275635, "learning_rate": 7.834472133828466e-05, "loss": 0.0293, "step": 27420 }, { "grad_norm": 0.32201501727104187, "learning_rate": 7.832769250899359e-05, "loss": 0.0373, "step": 27430 }, { "grad_norm": 0.36713799834251404, "learning_rate": 7.831065883927464e-05, "loss": 0.0376, "step": 27440 }, { "grad_norm": 0.27913179993629456, "learning_rate": 7.829362033203841e-05, "loss": 0.0342, "step": 27450 }, { "grad_norm": 0.25805357098579407, "learning_rate": 7.827657699019628e-05, "loss": 0.0315, "step": 27460 }, { "grad_norm": 0.34157794713974, "learning_rate": 7.825952881666052e-05, "loss": 0.0456, "step": 27470 }, { "grad_norm": 0.4135296940803528, "learning_rate": 7.824247581434418e-05, "loss": 0.0361, "step": 27480 }, { "grad_norm": 0.3420858383178711, "learning_rate": 7.822541798616116e-05, "loss": 0.031, "step": 27490 }, { "grad_norm": 0.2719346582889557, "learning_rate": 7.820835533502617e-05, "loss": 0.039, "step": 27500 }, { "grad_norm": 0.2991262674331665, "learning_rate": 7.819128786385475e-05, "loss": 0.0389, "step": 27510 }, { "grad_norm": 0.32364144921302795, "learning_rate": 7.817421557556329e-05, "loss": 0.0358, "step": 27520 }, { "grad_norm": 0.2867718040943146, "learning_rate": 7.815713847306893e-05, "loss": 0.0391, "step": 27530 }, { "grad_norm": 0.29788827896118164, "learning_rate": 7.81400565592897e-05, "loss": 0.0304, "step": 27540 }, { "grad_norm": 0.3878949284553528, "learning_rate": 7.812296983714444e-05, "loss": 0.0339, "step": 27550 }, { "grad_norm": 0.2468845695257187, "learning_rate": 7.810587830955281e-05, "loss": 0.0286, "step": 27560 }, { "grad_norm": 0.26484352350234985, "learning_rate": 7.808878197943528e-05, "loss": 0.03, "step": 27570 }, { "grad_norm": 0.19415408372879028, "learning_rate": 7.807168084971312e-05, "loss": 0.026, "step": 27580 }, { "grad_norm": 0.22094443440437317, "learning_rate": 7.805457492330849e-05, "loss": 0.0293, "step": 27590 }, { "grad_norm": 0.3437802493572235, "learning_rate": 7.803746420314428e-05, "loss": 0.0345, "step": 27600 }, { "grad_norm": 0.2875461280345917, "learning_rate": 7.802034869214428e-05, "loss": 0.0353, "step": 27610 }, { "grad_norm": 0.3732738196849823, "learning_rate": 7.800322839323303e-05, "loss": 0.0334, "step": 27620 }, { "grad_norm": 0.2181515395641327, "learning_rate": 7.798610330933593e-05, "loss": 0.0295, "step": 27630 }, { "grad_norm": 0.23311735689640045, "learning_rate": 7.796897344337922e-05, "loss": 0.0277, "step": 27640 }, { "grad_norm": 0.3114853501319885, "learning_rate": 7.795183879828989e-05, "loss": 0.0404, "step": 27650 }, { "grad_norm": 0.24912172555923462, "learning_rate": 7.793469937699579e-05, "loss": 0.0424, "step": 27660 }, { "grad_norm": 0.27798688411712646, "learning_rate": 7.791755518242558e-05, "loss": 0.0343, "step": 27670 }, { "grad_norm": 0.35273832082748413, "learning_rate": 7.790040621750876e-05, "loss": 0.0315, "step": 27680 }, { "grad_norm": 0.46261581778526306, "learning_rate": 7.788325248517558e-05, "loss": 0.0289, "step": 27690 }, { "grad_norm": 0.2593851685523987, "learning_rate": 7.786609398835715e-05, "loss": 0.0373, "step": 27700 }, { "grad_norm": 0.2806662917137146, "learning_rate": 7.784893072998541e-05, "loss": 0.0248, "step": 27710 }, { "grad_norm": 0.19223307073116302, "learning_rate": 7.783176271299306e-05, "loss": 0.0269, "step": 27720 }, { "grad_norm": 0.22535640001296997, "learning_rate": 7.781458994031368e-05, "loss": 0.0267, "step": 27730 }, { "grad_norm": 0.24313010275363922, "learning_rate": 7.779741241488161e-05, "loss": 0.0299, "step": 27740 }, { "grad_norm": 0.282747745513916, "learning_rate": 7.7780230139632e-05, "loss": 0.0298, "step": 27750 }, { "grad_norm": 0.253767728805542, "learning_rate": 7.776304311750087e-05, "loss": 0.0264, "step": 27760 }, { "grad_norm": 0.20470058917999268, "learning_rate": 7.7745851351425e-05, "loss": 0.0407, "step": 27770 }, { "grad_norm": 0.258922278881073, "learning_rate": 7.772865484434197e-05, "loss": 0.0277, "step": 27780 }, { "grad_norm": 0.2677960991859436, "learning_rate": 7.77114535991902e-05, "loss": 0.0334, "step": 27790 }, { "grad_norm": 0.262631893157959, "learning_rate": 7.769424761890893e-05, "loss": 0.0284, "step": 27800 }, { "grad_norm": 0.3803779184818268, "learning_rate": 7.767703690643817e-05, "loss": 0.0323, "step": 27810 }, { "grad_norm": 0.17270781099796295, "learning_rate": 7.76598214647188e-05, "loss": 0.024, "step": 27820 }, { "grad_norm": 0.22033357620239258, "learning_rate": 7.764260129669241e-05, "loss": 0.026, "step": 27830 }, { "grad_norm": 0.2586399018764496, "learning_rate": 7.76253764053015e-05, "loss": 0.0329, "step": 27840 }, { "grad_norm": 0.3040378987789154, "learning_rate": 7.760814679348932e-05, "loss": 0.0302, "step": 27850 }, { "grad_norm": 0.2583538889884949, "learning_rate": 7.759091246419992e-05, "loss": 0.0241, "step": 27860 }, { "grad_norm": 0.24132981896400452, "learning_rate": 7.75736734203782e-05, "loss": 0.0269, "step": 27870 }, { "grad_norm": 0.28095588088035583, "learning_rate": 7.755642966496985e-05, "loss": 0.0438, "step": 27880 }, { "grad_norm": 0.2903136610984802, "learning_rate": 7.753918120092132e-05, "loss": 0.0321, "step": 27890 }, { "grad_norm": 0.18647043406963348, "learning_rate": 7.752192803117993e-05, "loss": 0.0281, "step": 27900 }, { "grad_norm": 0.27505990862846375, "learning_rate": 7.750467015869377e-05, "loss": 0.0304, "step": 27910 }, { "grad_norm": 0.299981027841568, "learning_rate": 7.748740758641174e-05, "loss": 0.0314, "step": 27920 }, { "grad_norm": 0.27450984716415405, "learning_rate": 7.74701403172835e-05, "loss": 0.0312, "step": 27930 }, { "grad_norm": 0.27993419766426086, "learning_rate": 7.745286835425962e-05, "loss": 0.0296, "step": 27940 }, { "grad_norm": 0.31537100672721863, "learning_rate": 7.743559170029138e-05, "loss": 0.0296, "step": 27950 }, { "grad_norm": 0.29109397530555725, "learning_rate": 7.741831035833087e-05, "loss": 0.0397, "step": 27960 }, { "grad_norm": 0.18744446337223053, "learning_rate": 7.740102433133102e-05, "loss": 0.0288, "step": 27970 }, { "grad_norm": 0.25001809000968933, "learning_rate": 7.738373362224553e-05, "loss": 0.0245, "step": 27980 }, { "grad_norm": 0.3497280478477478, "learning_rate": 7.73664382340289e-05, "loss": 0.028, "step": 27990 }, { "grad_norm": 0.24000978469848633, "learning_rate": 7.734913816963647e-05, "loss": 0.0348, "step": 28000 }, { "grad_norm": 0.34472641348838806, "learning_rate": 7.73318334320243e-05, "loss": 0.0255, "step": 28010 }, { "grad_norm": 0.3269921541213989, "learning_rate": 7.731452402414934e-05, "loss": 0.038, "step": 28020 }, { "grad_norm": 0.29051679372787476, "learning_rate": 7.729720994896928e-05, "loss": 0.0294, "step": 28030 }, { "grad_norm": 0.2485845386981964, "learning_rate": 7.727989120944262e-05, "loss": 0.0311, "step": 28040 }, { "grad_norm": 0.24055466055870056, "learning_rate": 7.726256780852865e-05, "loss": 0.0293, "step": 28050 }, { "grad_norm": 0.2962172329425812, "learning_rate": 7.724523974918749e-05, "loss": 0.0296, "step": 28060 }, { "grad_norm": 0.35147783160209656, "learning_rate": 7.722790703438002e-05, "loss": 0.0326, "step": 28070 }, { "grad_norm": 0.22654889523983002, "learning_rate": 7.72105696670679e-05, "loss": 0.0364, "step": 28080 }, { "grad_norm": 0.3198586702346802, "learning_rate": 7.719322765021364e-05, "loss": 0.0331, "step": 28090 }, { "grad_norm": 0.3198360800743103, "learning_rate": 7.717588098678051e-05, "loss": 0.0337, "step": 28100 }, { "grad_norm": 0.2739424705505371, "learning_rate": 7.715852967973258e-05, "loss": 0.0322, "step": 28110 }, { "grad_norm": 0.2608492970466614, "learning_rate": 7.714117373203474e-05, "loss": 0.0338, "step": 28120 }, { "grad_norm": 0.27436578273773193, "learning_rate": 7.712381314665259e-05, "loss": 0.0306, "step": 28130 }, { "grad_norm": 0.31643107533454895, "learning_rate": 7.710644792655261e-05, "loss": 0.033, "step": 28140 }, { "grad_norm": 0.31301823258399963, "learning_rate": 7.708907807470207e-05, "loss": 0.041, "step": 28150 }, { "grad_norm": 0.25487083196640015, "learning_rate": 7.707170359406896e-05, "loss": 0.0311, "step": 28160 }, { "grad_norm": 0.3187040388584137, "learning_rate": 7.705432448762213e-05, "loss": 0.0333, "step": 28170 }, { "grad_norm": 0.31367820501327515, "learning_rate": 7.703694075833117e-05, "loss": 0.0422, "step": 28180 }, { "grad_norm": 0.3628843128681183, "learning_rate": 7.70195524091665e-05, "loss": 0.0539, "step": 28190 }, { "grad_norm": 0.3183954954147339, "learning_rate": 7.70021594430993e-05, "loss": 0.0373, "step": 28200 }, { "grad_norm": 0.28527194261550903, "learning_rate": 7.698476186310157e-05, "loss": 0.0332, "step": 28210 }, { "grad_norm": 0.3798556625843048, "learning_rate": 7.696735967214608e-05, "loss": 0.0332, "step": 28220 }, { "grad_norm": 0.3337184190750122, "learning_rate": 7.694995287320636e-05, "loss": 0.0355, "step": 28230 }, { "grad_norm": 0.30490967631340027, "learning_rate": 7.693254146925679e-05, "loss": 0.0351, "step": 28240 }, { "grad_norm": 0.47139695286750793, "learning_rate": 7.691512546327251e-05, "loss": 0.037, "step": 28250 }, { "grad_norm": 0.3151182234287262, "learning_rate": 7.689770485822939e-05, "loss": 0.036, "step": 28260 }, { "grad_norm": 0.3184099793434143, "learning_rate": 7.688027965710416e-05, "loss": 0.0345, "step": 28270 }, { "grad_norm": 0.38942837715148926, "learning_rate": 7.686284986287433e-05, "loss": 0.0339, "step": 28280 }, { "grad_norm": 0.2124606817960739, "learning_rate": 7.684541547851817e-05, "loss": 0.0296, "step": 28290 }, { "grad_norm": 0.2378241866827011, "learning_rate": 7.68279765070147e-05, "loss": 0.031, "step": 28300 }, { "grad_norm": 0.18935862183570862, "learning_rate": 7.68105329513438e-05, "loss": 0.0316, "step": 28310 }, { "grad_norm": 0.27907440066337585, "learning_rate": 7.67930848144861e-05, "loss": 0.0288, "step": 28320 }, { "grad_norm": 0.20126135647296906, "learning_rate": 7.6775632099423e-05, "loss": 0.0343, "step": 28330 }, { "grad_norm": 0.33491790294647217, "learning_rate": 7.675817480913667e-05, "loss": 0.0301, "step": 28340 }, { "grad_norm": 0.23388832807540894, "learning_rate": 7.674071294661011e-05, "loss": 0.0282, "step": 28350 }, { "grad_norm": 0.2761387825012207, "learning_rate": 7.672324651482707e-05, "loss": 0.0276, "step": 28360 }, { "grad_norm": 0.3080836534500122, "learning_rate": 7.670577551677209e-05, "loss": 0.0295, "step": 28370 }, { "grad_norm": 0.3808286190032959, "learning_rate": 7.668829995543047e-05, "loss": 0.0328, "step": 28380 }, { "grad_norm": 0.2609280049800873, "learning_rate": 7.667081983378832e-05, "loss": 0.0362, "step": 28390 }, { "grad_norm": 0.31702175736427307, "learning_rate": 7.66533351548325e-05, "loss": 0.0277, "step": 28400 }, { "grad_norm": 0.39791953563690186, "learning_rate": 7.663584592155069e-05, "loss": 0.0355, "step": 28410 }, { "grad_norm": 0.327437162399292, "learning_rate": 7.661835213693129e-05, "loss": 0.0356, "step": 28420 }, { "grad_norm": 0.3060523271560669, "learning_rate": 7.660085380396353e-05, "loss": 0.0332, "step": 28430 }, { "grad_norm": 0.2870880961418152, "learning_rate": 7.658335092563738e-05, "loss": 0.0333, "step": 28440 }, { "grad_norm": 0.18822067975997925, "learning_rate": 7.656584350494362e-05, "loss": 0.0293, "step": 28450 }, { "grad_norm": 0.2947266399860382, "learning_rate": 7.654833154487378e-05, "loss": 0.0287, "step": 28460 }, { "grad_norm": 0.22490398585796356, "learning_rate": 7.653081504842017e-05, "loss": 0.0272, "step": 28470 }, { "grad_norm": 0.3008618652820587, "learning_rate": 7.65132940185759e-05, "loss": 0.0383, "step": 28480 }, { "grad_norm": 0.25082093477249146, "learning_rate": 7.649576845833481e-05, "loss": 0.0249, "step": 28490 }, { "grad_norm": 0.33433762192726135, "learning_rate": 7.647823837069156e-05, "loss": 0.0316, "step": 28500 }, { "grad_norm": 0.33876851201057434, "learning_rate": 7.646070375864156e-05, "loss": 0.0303, "step": 28510 }, { "grad_norm": 0.2514645755290985, "learning_rate": 7.644316462518099e-05, "loss": 0.0334, "step": 28520 }, { "grad_norm": 0.28337493538856506, "learning_rate": 7.642562097330679e-05, "loss": 0.0338, "step": 28530 }, { "grad_norm": 0.3473219573497772, "learning_rate": 7.640807280601671e-05, "loss": 0.0335, "step": 28540 }, { "grad_norm": 0.22428859770298004, "learning_rate": 7.639052012630927e-05, "loss": 0.0377, "step": 28550 }, { "grad_norm": 0.23957210779190063, "learning_rate": 7.63729629371837e-05, "loss": 0.0336, "step": 28560 }, { "grad_norm": 0.36055150628089905, "learning_rate": 7.635540124164009e-05, "loss": 0.041, "step": 28570 }, { "grad_norm": 0.28244879841804504, "learning_rate": 7.633783504267922e-05, "loss": 0.0376, "step": 28580 }, { "grad_norm": 0.33619922399520874, "learning_rate": 7.632026434330269e-05, "loss": 0.0313, "step": 28590 }, { "grad_norm": 0.2600797116756439, "learning_rate": 7.630268914651282e-05, "loss": 0.0276, "step": 28600 }, { "grad_norm": 0.2698369026184082, "learning_rate": 7.628510945531278e-05, "loss": 0.0397, "step": 28610 }, { "grad_norm": 0.24970389902591705, "learning_rate": 7.626752527270641e-05, "loss": 0.037, "step": 28620 }, { "grad_norm": 0.2718091607093811, "learning_rate": 7.62499366016984e-05, "loss": 0.0309, "step": 28630 }, { "grad_norm": 0.3144467771053314, "learning_rate": 7.623234344529416e-05, "loss": 0.0346, "step": 28640 }, { "grad_norm": 0.3224661350250244, "learning_rate": 7.62147458064999e-05, "loss": 0.0347, "step": 28650 }, { "grad_norm": 0.3453907072544098, "learning_rate": 7.619714368832254e-05, "loss": 0.0372, "step": 28660 }, { "grad_norm": 0.3842419683933258, "learning_rate": 7.61795370937698e-05, "loss": 0.0282, "step": 28670 }, { "grad_norm": 0.23362700641155243, "learning_rate": 7.61619260258502e-05, "loss": 0.0306, "step": 28680 }, { "grad_norm": 0.2951200604438782, "learning_rate": 7.614431048757298e-05, "loss": 0.0336, "step": 28690 }, { "grad_norm": 0.3115476369857788, "learning_rate": 7.612669048194814e-05, "loss": 0.0297, "step": 28700 }, { "grad_norm": 0.2896186113357544, "learning_rate": 7.610906601198646e-05, "loss": 0.0317, "step": 28710 }, { "grad_norm": 0.22415845096111298, "learning_rate": 7.60914370806995e-05, "loss": 0.0333, "step": 28720 }, { "grad_norm": 0.30079540610313416, "learning_rate": 7.607380369109953e-05, "loss": 0.0469, "step": 28730 }, { "grad_norm": 0.20962604880332947, "learning_rate": 7.605616584619961e-05, "loss": 0.026, "step": 28740 }, { "grad_norm": 0.25456419587135315, "learning_rate": 7.603852354901362e-05, "loss": 0.0333, "step": 28750 }, { "grad_norm": 0.3243805468082428, "learning_rate": 7.602087680255609e-05, "loss": 0.0416, "step": 28760 }, { "grad_norm": 0.3447541892528534, "learning_rate": 7.600322560984238e-05, "loss": 0.023, "step": 28770 }, { "grad_norm": 0.2920767366886139, "learning_rate": 7.598556997388863e-05, "loss": 0.0325, "step": 28780 }, { "grad_norm": 0.3890122175216675, "learning_rate": 7.596790989771166e-05, "loss": 0.0361, "step": 28790 }, { "grad_norm": 0.23735590279102325, "learning_rate": 7.595024538432914e-05, "loss": 0.0313, "step": 28800 }, { "grad_norm": 0.2727576196193695, "learning_rate": 7.59325764367594e-05, "loss": 0.0245, "step": 28810 }, { "grad_norm": 0.24876080453395844, "learning_rate": 7.59149030580216e-05, "loss": 0.031, "step": 28820 }, { "grad_norm": 0.25220391154289246, "learning_rate": 7.589722525113562e-05, "loss": 0.0273, "step": 28830 }, { "grad_norm": 0.2857358753681183, "learning_rate": 7.587954301912216e-05, "loss": 0.0281, "step": 28840 }, { "grad_norm": 0.33956316113471985, "learning_rate": 7.586185636500263e-05, "loss": 0.0311, "step": 28850 }, { "grad_norm": 0.22181758284568787, "learning_rate": 7.584416529179914e-05, "loss": 0.0308, "step": 28860 }, { "grad_norm": 0.30115774273872375, "learning_rate": 7.582646980253465e-05, "loss": 0.0318, "step": 28870 }, { "grad_norm": 0.20046617090702057, "learning_rate": 7.580876990023282e-05, "loss": 0.037, "step": 28880 }, { "grad_norm": 0.24866873025894165, "learning_rate": 7.579106558791809e-05, "loss": 0.0286, "step": 28890 }, { "grad_norm": 0.22881944477558136, "learning_rate": 7.577335686861565e-05, "loss": 0.0314, "step": 28900 }, { "grad_norm": 0.2628772258758545, "learning_rate": 7.575564374535141e-05, "loss": 0.03, "step": 28910 }, { "grad_norm": 0.24683919548988342, "learning_rate": 7.573792622115207e-05, "loss": 0.0298, "step": 28920 }, { "grad_norm": 0.27813997864723206, "learning_rate": 7.572020429904507e-05, "loss": 0.0307, "step": 28930 }, { "grad_norm": 0.27772361040115356, "learning_rate": 7.570247798205861e-05, "loss": 0.031, "step": 28940 }, { "grad_norm": 0.23539671301841736, "learning_rate": 7.568474727322164e-05, "loss": 0.0312, "step": 28950 }, { "grad_norm": 0.342631459236145, "learning_rate": 7.566701217556384e-05, "loss": 0.0337, "step": 28960 }, { "grad_norm": 0.21857425570487976, "learning_rate": 7.564927269211564e-05, "loss": 0.028, "step": 28970 }, { "grad_norm": 0.20466060936450958, "learning_rate": 7.563152882590824e-05, "loss": 0.0242, "step": 28980 }, { "grad_norm": 0.22638443112373352, "learning_rate": 7.56137805799736e-05, "loss": 0.0285, "step": 28990 }, { "grad_norm": 0.25924330949783325, "learning_rate": 7.559602795734439e-05, "loss": 0.0338, "step": 29000 }, { "grad_norm": 0.2921043634414673, "learning_rate": 7.557827096105408e-05, "loss": 0.0299, "step": 29010 }, { "grad_norm": 0.21990694105625153, "learning_rate": 7.55605095941368e-05, "loss": 0.0324, "step": 29020 }, { "grad_norm": 0.3184763789176941, "learning_rate": 7.554274385962752e-05, "loss": 0.0339, "step": 29030 }, { "grad_norm": 0.3381718695163727, "learning_rate": 7.552497376056191e-05, "loss": 0.0363, "step": 29040 }, { "grad_norm": 0.2513328194618225, "learning_rate": 7.550719929997639e-05, "loss": 0.0342, "step": 29050 }, { "grad_norm": 0.27821630239486694, "learning_rate": 7.548942048090813e-05, "loss": 0.0346, "step": 29060 }, { "grad_norm": 0.2623501121997833, "learning_rate": 7.547163730639506e-05, "loss": 0.0342, "step": 29070 }, { "grad_norm": 0.24352286756038666, "learning_rate": 7.545384977947583e-05, "loss": 0.0307, "step": 29080 }, { "grad_norm": 0.25564879179000854, "learning_rate": 7.543605790318981e-05, "loss": 0.0264, "step": 29090 }, { "grad_norm": 0.2906785011291504, "learning_rate": 7.54182616805772e-05, "loss": 0.033, "step": 29100 }, { "grad_norm": 0.2863135039806366, "learning_rate": 7.540046111467885e-05, "loss": 0.0277, "step": 29110 }, { "grad_norm": 0.21774862706661224, "learning_rate": 7.53826562085364e-05, "loss": 0.0309, "step": 29120 }, { "grad_norm": 0.35459011793136597, "learning_rate": 7.536484696519221e-05, "loss": 0.028, "step": 29130 }, { "grad_norm": 0.30925649404525757, "learning_rate": 7.534703338768942e-05, "loss": 0.0262, "step": 29140 }, { "grad_norm": 0.31771349906921387, "learning_rate": 7.532921547907185e-05, "loss": 0.0334, "step": 29150 }, { "grad_norm": 0.2732851207256317, "learning_rate": 7.531139324238412e-05, "loss": 0.0364, "step": 29160 }, { "grad_norm": 0.22895854711532593, "learning_rate": 7.529356668067157e-05, "loss": 0.0409, "step": 29170 }, { "grad_norm": 0.276614248752594, "learning_rate": 7.527573579698023e-05, "loss": 0.0334, "step": 29180 }, { "grad_norm": 0.23894129693508148, "learning_rate": 7.525790059435693e-05, "loss": 0.0318, "step": 29190 }, { "grad_norm": 0.26262587308883667, "learning_rate": 7.524006107584926e-05, "loss": 0.0303, "step": 29200 }, { "grad_norm": 0.3457246422767639, "learning_rate": 7.522221724450544e-05, "loss": 0.0308, "step": 29210 }, { "grad_norm": 0.22151829302310944, "learning_rate": 7.520436910337451e-05, "loss": 0.039, "step": 29220 }, { "grad_norm": 0.2884620428085327, "learning_rate": 7.518651665550627e-05, "loss": 0.0334, "step": 29230 }, { "grad_norm": 0.2511797249317169, "learning_rate": 7.516865990395117e-05, "loss": 0.0407, "step": 29240 }, { "grad_norm": 0.30297866463661194, "learning_rate": 7.515079885176047e-05, "loss": 0.0334, "step": 29250 }, { "grad_norm": 0.2745342552661896, "learning_rate": 7.513293350198612e-05, "loss": 0.0308, "step": 29260 }, { "grad_norm": 0.2882441580295563, "learning_rate": 7.511506385768081e-05, "loss": 0.0315, "step": 29270 }, { "grad_norm": 0.28425833582878113, "learning_rate": 7.509718992189801e-05, "loss": 0.0329, "step": 29280 }, { "grad_norm": 0.2885096073150635, "learning_rate": 7.507931169769182e-05, "loss": 0.0323, "step": 29290 }, { "grad_norm": 0.3476753830909729, "learning_rate": 7.506142918811722e-05, "loss": 0.0285, "step": 29300 }, { "grad_norm": 0.263902485370636, "learning_rate": 7.504354239622978e-05, "loss": 0.0345, "step": 29310 }, { "grad_norm": 0.2399648278951645, "learning_rate": 7.50256513250859e-05, "loss": 0.0323, "step": 29320 }, { "grad_norm": 0.23833337426185608, "learning_rate": 7.500775597774265e-05, "loss": 0.027, "step": 29330 }, { "grad_norm": 0.3844814598560333, "learning_rate": 7.498985635725788e-05, "loss": 0.0328, "step": 29340 }, { "grad_norm": 0.25958818197250366, "learning_rate": 7.497195246669012e-05, "loss": 0.0364, "step": 29350 }, { "grad_norm": 0.3540421426296234, "learning_rate": 7.495404430909868e-05, "loss": 0.0307, "step": 29360 }, { "grad_norm": 0.4384176433086395, "learning_rate": 7.493613188754356e-05, "loss": 0.037, "step": 29370 }, { "grad_norm": 0.47866523265838623, "learning_rate": 7.49182152050855e-05, "loss": 0.0305, "step": 29380 }, { "grad_norm": 0.3859719932079315, "learning_rate": 7.490029426478598e-05, "loss": 0.0277, "step": 29390 }, { "grad_norm": 0.3431636095046997, "learning_rate": 7.488236906970719e-05, "loss": 0.0368, "step": 29400 }, { "grad_norm": 0.4725162982940674, "learning_rate": 7.486443962291207e-05, "loss": 0.0347, "step": 29410 }, { "grad_norm": 0.31420862674713135, "learning_rate": 7.484650592746424e-05, "loss": 0.0284, "step": 29420 }, { "grad_norm": 0.28752651810646057, "learning_rate": 7.482856798642811e-05, "loss": 0.032, "step": 29430 }, { "grad_norm": 0.2605203688144684, "learning_rate": 7.481062580286878e-05, "loss": 0.0299, "step": 29440 }, { "grad_norm": 0.2132844626903534, "learning_rate": 7.479267937985208e-05, "loss": 0.0332, "step": 29450 }, { "grad_norm": 0.21601521968841553, "learning_rate": 7.477472872044456e-05, "loss": 0.0227, "step": 29460 }, { "grad_norm": 0.32113558053970337, "learning_rate": 7.475677382771347e-05, "loss": 0.0337, "step": 29470 }, { "grad_norm": 0.26516464352607727, "learning_rate": 7.473881470472683e-05, "loss": 0.0383, "step": 29480 }, { "grad_norm": 0.3673427402973175, "learning_rate": 7.472085135455337e-05, "loss": 0.0352, "step": 29490 }, { "grad_norm": 0.2549360692501068, "learning_rate": 7.470288378026256e-05, "loss": 0.0339, "step": 29500 }, { "grad_norm": 0.3279927968978882, "learning_rate": 7.468491198492451e-05, "loss": 0.0302, "step": 29510 }, { "grad_norm": 0.4816288948059082, "learning_rate": 7.466693597161013e-05, "loss": 0.0313, "step": 29520 }, { "grad_norm": 0.27081191539764404, "learning_rate": 7.464895574339104e-05, "loss": 0.0308, "step": 29530 }, { "grad_norm": 0.3634816110134125, "learning_rate": 7.463097130333958e-05, "loss": 0.0402, "step": 29540 }, { "grad_norm": 0.2984289228916168, "learning_rate": 7.461298265452876e-05, "loss": 0.0297, "step": 29550 }, { "grad_norm": 0.21839436888694763, "learning_rate": 7.459498980003239e-05, "loss": 0.0258, "step": 29560 }, { "grad_norm": 0.31643781065940857, "learning_rate": 7.457699274292493e-05, "loss": 0.0323, "step": 29570 }, { "grad_norm": 0.24129897356033325, "learning_rate": 7.455899148628159e-05, "loss": 0.0352, "step": 29580 }, { "grad_norm": 0.21965132653713226, "learning_rate": 7.45409860331783e-05, "loss": 0.0282, "step": 29590 }, { "grad_norm": 0.3063414990901947, "learning_rate": 7.452297638669169e-05, "loss": 0.0286, "step": 29600 }, { "grad_norm": 0.22341352701187134, "learning_rate": 7.450496254989911e-05, "loss": 0.0263, "step": 29610 }, { "grad_norm": 0.31011247634887695, "learning_rate": 7.448694452587866e-05, "loss": 0.0291, "step": 29620 }, { "grad_norm": 0.24384628236293793, "learning_rate": 7.44689223177091e-05, "loss": 0.0315, "step": 29630 }, { "grad_norm": 0.41663363575935364, "learning_rate": 7.445089592846994e-05, "loss": 0.028, "step": 29640 }, { "grad_norm": 0.2909717857837677, "learning_rate": 7.443286536124141e-05, "loss": 0.0351, "step": 29650 }, { "grad_norm": 0.20813414454460144, "learning_rate": 7.441483061910443e-05, "loss": 0.0335, "step": 29660 }, { "grad_norm": 0.29764434695243835, "learning_rate": 7.439679170514064e-05, "loss": 0.0335, "step": 29670 }, { "grad_norm": 0.25518596172332764, "learning_rate": 7.43787486224324e-05, "loss": 0.0321, "step": 29680 }, { "grad_norm": 0.21107633411884308, "learning_rate": 7.436070137406276e-05, "loss": 0.0333, "step": 29690 }, { "grad_norm": 0.25882092118263245, "learning_rate": 7.434264996311556e-05, "loss": 0.0299, "step": 29700 }, { "grad_norm": 0.2892390191555023, "learning_rate": 7.432459439267525e-05, "loss": 0.0321, "step": 29710 }, { "grad_norm": 0.3729091286659241, "learning_rate": 7.430653466582701e-05, "loss": 0.0315, "step": 29720 }, { "grad_norm": 0.26657605171203613, "learning_rate": 7.42884707856568e-05, "loss": 0.0438, "step": 29730 }, { "grad_norm": 0.2496701031923294, "learning_rate": 7.427040275525122e-05, "loss": 0.0274, "step": 29740 }, { "grad_norm": 0.32093575596809387, "learning_rate": 7.42523305776976e-05, "loss": 0.0384, "step": 29750 }, { "grad_norm": 0.22309640049934387, "learning_rate": 7.4234254256084e-05, "loss": 0.0301, "step": 29760 }, { "grad_norm": 0.23163436353206635, "learning_rate": 7.421617379349915e-05, "loss": 0.0359, "step": 29770 }, { "grad_norm": 0.1763388216495514, "learning_rate": 7.41980891930325e-05, "loss": 0.0331, "step": 29780 }, { "grad_norm": 0.2677633762359619, "learning_rate": 7.418000045777425e-05, "loss": 0.0279, "step": 29790 }, { "grad_norm": 0.2582930326461792, "learning_rate": 7.416190759081523e-05, "loss": 0.0319, "step": 29800 }, { "grad_norm": 0.25885382294654846, "learning_rate": 7.414381059524704e-05, "loss": 0.0245, "step": 29810 }, { "grad_norm": 0.3004094958305359, "learning_rate": 7.412570947416195e-05, "loss": 0.028, "step": 29820 }, { "grad_norm": 0.3281594216823578, "learning_rate": 7.410760423065295e-05, "loss": 0.0376, "step": 29830 }, { "grad_norm": 0.23627859354019165, "learning_rate": 7.408949486781372e-05, "loss": 0.0392, "step": 29840 }, { "grad_norm": 0.30100369453430176, "learning_rate": 7.407138138873868e-05, "loss": 0.0338, "step": 29850 }, { "grad_norm": 0.32306089997291565, "learning_rate": 7.405326379652292e-05, "loss": 0.03, "step": 29860 }, { "grad_norm": 0.3399423062801361, "learning_rate": 7.403514209426222e-05, "loss": 0.0264, "step": 29870 }, { "grad_norm": 0.3038252592086792, "learning_rate": 7.40170162850531e-05, "loss": 0.0296, "step": 29880 }, { "grad_norm": 0.26348748803138733, "learning_rate": 7.399888637199278e-05, "loss": 0.0269, "step": 29890 }, { "grad_norm": 0.3611593544483185, "learning_rate": 7.398075235817914e-05, "loss": 0.0419, "step": 29900 }, { "grad_norm": 0.2885189950466156, "learning_rate": 7.39626142467108e-05, "loss": 0.0328, "step": 29910 }, { "grad_norm": 0.2510036528110504, "learning_rate": 7.394447204068706e-05, "loss": 0.0312, "step": 29920 }, { "grad_norm": 0.25382402539253235, "learning_rate": 7.392632574320793e-05, "loss": 0.0367, "step": 29930 }, { "grad_norm": 0.3429942727088928, "learning_rate": 7.390817535737411e-05, "loss": 0.0328, "step": 29940 }, { "grad_norm": 0.2213744968175888, "learning_rate": 7.389002088628703e-05, "loss": 0.0269, "step": 29950 }, { "grad_norm": 0.25072431564331055, "learning_rate": 7.387186233304877e-05, "loss": 0.0281, "step": 29960 }, { "grad_norm": 0.21797041594982147, "learning_rate": 7.385369970076212e-05, "loss": 0.0286, "step": 29970 }, { "grad_norm": 0.250139445066452, "learning_rate": 7.38355329925306e-05, "loss": 0.029, "step": 29980 }, { "grad_norm": 0.28255999088287354, "learning_rate": 7.381736221145838e-05, "loss": 0.0279, "step": 29990 }, { "grad_norm": 0.33780500292778015, "learning_rate": 7.37991873606504e-05, "loss": 0.0256, "step": 30000 }, { "grad_norm": 0.3276961147785187, "learning_rate": 7.378100844321218e-05, "loss": 0.0268, "step": 30010 }, { "grad_norm": 0.3065875768661499, "learning_rate": 7.376282546225004e-05, "loss": 0.0282, "step": 30020 }, { "grad_norm": 0.2229153960943222, "learning_rate": 7.374463842087094e-05, "loss": 0.0266, "step": 30030 }, { "grad_norm": 0.2852281928062439, "learning_rate": 7.372644732218254e-05, "loss": 0.026, "step": 30040 }, { "grad_norm": 0.3299334943294525, "learning_rate": 7.370825216929322e-05, "loss": 0.0382, "step": 30050 }, { "grad_norm": 0.23403365910053253, "learning_rate": 7.369005296531205e-05, "loss": 0.03, "step": 30060 }, { "grad_norm": 0.2559160590171814, "learning_rate": 7.367184971334873e-05, "loss": 0.03, "step": 30070 }, { "grad_norm": 0.3595258891582489, "learning_rate": 7.365364241651371e-05, "loss": 0.0278, "step": 30080 }, { "grad_norm": 0.2896117568016052, "learning_rate": 7.363543107791815e-05, "loss": 0.0243, "step": 30090 }, { "grad_norm": 0.24355250597000122, "learning_rate": 7.361721570067384e-05, "loss": 0.027, "step": 30100 }, { "grad_norm": 0.3383019268512726, "learning_rate": 7.359899628789331e-05, "loss": 0.035, "step": 30110 }, { "grad_norm": 0.32286980748176575, "learning_rate": 7.358077284268974e-05, "loss": 0.0308, "step": 30120 }, { "grad_norm": 0.23078106343746185, "learning_rate": 7.356254536817702e-05, "loss": 0.0288, "step": 30130 }, { "grad_norm": 0.38072821497917175, "learning_rate": 7.354431386746973e-05, "loss": 0.0351, "step": 30140 }, { "grad_norm": 0.3144843876361847, "learning_rate": 7.352607834368316e-05, "loss": 0.0283, "step": 30150 }, { "grad_norm": 0.26345211267471313, "learning_rate": 7.350783879993324e-05, "loss": 0.026, "step": 30160 }, { "grad_norm": 0.34977594017982483, "learning_rate": 7.348959523933658e-05, "loss": 0.032, "step": 30170 }, { "grad_norm": 0.2821301221847534, "learning_rate": 7.347134766501057e-05, "loss": 0.0282, "step": 30180 }, { "grad_norm": 0.431769460439682, "learning_rate": 7.345309608007315e-05, "loss": 0.0276, "step": 30190 }, { "grad_norm": 0.24938197433948517, "learning_rate": 7.343484048764308e-05, "loss": 0.0298, "step": 30200 }, { "grad_norm": 0.36573147773742676, "learning_rate": 7.341658089083972e-05, "loss": 0.0291, "step": 30210 }, { "grad_norm": 0.1674434095621109, "learning_rate": 7.339831729278313e-05, "loss": 0.0278, "step": 30220 }, { "grad_norm": 0.2597414553165436, "learning_rate": 7.338004969659404e-05, "loss": 0.0247, "step": 30230 }, { "grad_norm": 0.22154484689235687, "learning_rate": 7.336177810539391e-05, "loss": 0.0305, "step": 30240 }, { "grad_norm": 0.2953173518180847, "learning_rate": 7.334350252230485e-05, "loss": 0.0241, "step": 30250 }, { "grad_norm": 0.2540891766548157, "learning_rate": 7.332522295044965e-05, "loss": 0.0207, "step": 30260 }, { "grad_norm": 0.28381478786468506, "learning_rate": 7.33069393929518e-05, "loss": 0.0288, "step": 30270 }, { "grad_norm": 0.2892569303512573, "learning_rate": 7.328865185293545e-05, "loss": 0.0347, "step": 30280 }, { "grad_norm": 0.20309974253177643, "learning_rate": 7.327036033352546e-05, "loss": 0.0314, "step": 30290 }, { "grad_norm": 0.21553321182727814, "learning_rate": 7.325206483784733e-05, "loss": 0.0251, "step": 30300 }, { "grad_norm": 0.2173530012369156, "learning_rate": 7.323376536902724e-05, "loss": 0.0357, "step": 30310 }, { "grad_norm": 0.35277798771858215, "learning_rate": 7.321546193019213e-05, "loss": 0.0352, "step": 30320 }, { "grad_norm": 0.2363574504852295, "learning_rate": 7.31971545244695e-05, "loss": 0.0271, "step": 30330 }, { "grad_norm": 0.3282005488872528, "learning_rate": 7.31788431549876e-05, "loss": 0.0283, "step": 30340 }, { "grad_norm": 0.27289050817489624, "learning_rate": 7.316052782487534e-05, "loss": 0.0213, "step": 30350 }, { "grad_norm": 0.27664339542388916, "learning_rate": 7.314220853726234e-05, "loss": 0.0342, "step": 30360 }, { "grad_norm": 0.26949095726013184, "learning_rate": 7.312388529527884e-05, "loss": 0.03, "step": 30370 }, { "grad_norm": 0.2616163194179535, "learning_rate": 7.310555810205577e-05, "loss": 0.029, "step": 30380 }, { "grad_norm": 0.2093876451253891, "learning_rate": 7.308722696072476e-05, "loss": 0.0274, "step": 30390 }, { "grad_norm": 0.23587006330490112, "learning_rate": 7.306889187441811e-05, "loss": 0.0277, "step": 30400 }, { "grad_norm": 0.2007780373096466, "learning_rate": 7.305055284626876e-05, "loss": 0.0236, "step": 30410 }, { "grad_norm": 0.27219611406326294, "learning_rate": 7.303220987941037e-05, "loss": 0.0259, "step": 30420 }, { "grad_norm": 0.29577136039733887, "learning_rate": 7.301386297697726e-05, "loss": 0.0292, "step": 30430 }, { "grad_norm": 0.35655590891838074, "learning_rate": 7.299551214210438e-05, "loss": 0.0306, "step": 30440 }, { "grad_norm": 0.23490315675735474, "learning_rate": 7.297715737792738e-05, "loss": 0.0264, "step": 30450 }, { "grad_norm": 0.2748432159423828, "learning_rate": 7.295879868758265e-05, "loss": 0.0253, "step": 30460 }, { "grad_norm": 0.26764944195747375, "learning_rate": 7.294043607420713e-05, "loss": 0.0289, "step": 30470 }, { "grad_norm": 0.3045448064804077, "learning_rate": 7.292206954093852e-05, "loss": 0.0286, "step": 30480 }, { "grad_norm": 0.2593781352043152, "learning_rate": 7.290369909091515e-05, "loss": 0.0278, "step": 30490 }, { "grad_norm": 0.2662826180458069, "learning_rate": 7.2885324727276e-05, "loss": 0.0343, "step": 30500 }, { "grad_norm": 0.2436831295490265, "learning_rate": 7.286694645316076e-05, "loss": 0.0264, "step": 30510 }, { "grad_norm": 0.31879252195358276, "learning_rate": 7.284856427170982e-05, "loss": 0.0261, "step": 30520 }, { "grad_norm": 0.28374752402305603, "learning_rate": 7.283017818606414e-05, "loss": 0.0308, "step": 30530 }, { "grad_norm": 0.304994136095047, "learning_rate": 7.28117881993654e-05, "loss": 0.0303, "step": 30540 }, { "grad_norm": 0.24409636855125427, "learning_rate": 7.279339431475598e-05, "loss": 0.025, "step": 30550 }, { "grad_norm": 0.2956182062625885, "learning_rate": 7.277499653537887e-05, "loss": 0.0268, "step": 30560 }, { "grad_norm": 0.254334956407547, "learning_rate": 7.275659486437776e-05, "loss": 0.035, "step": 30570 }, { "grad_norm": 0.2108837217092514, "learning_rate": 7.273818930489695e-05, "loss": 0.0322, "step": 30580 }, { "grad_norm": 0.2182997763156891, "learning_rate": 7.271977986008151e-05, "loss": 0.0277, "step": 30590 }, { "grad_norm": 0.2904035449028015, "learning_rate": 7.270136653307705e-05, "loss": 0.0267, "step": 30600 }, { "grad_norm": 0.3003275692462921, "learning_rate": 7.268294932702994e-05, "loss": 0.0304, "step": 30610 }, { "grad_norm": 0.25644686818122864, "learning_rate": 7.266452824508719e-05, "loss": 0.0292, "step": 30620 }, { "grad_norm": 0.21508671343326569, "learning_rate": 7.264610329039643e-05, "loss": 0.0253, "step": 30630 }, { "grad_norm": 0.2417657971382141, "learning_rate": 7.262767446610599e-05, "loss": 0.0264, "step": 30640 }, { "grad_norm": 0.3626754879951477, "learning_rate": 7.260924177536485e-05, "loss": 0.0354, "step": 30650 }, { "grad_norm": 0.23417814075946808, "learning_rate": 7.259080522132265e-05, "loss": 0.0284, "step": 30660 }, { "grad_norm": 0.29174259305000305, "learning_rate": 7.257236480712972e-05, "loss": 0.0293, "step": 30670 }, { "grad_norm": 0.2839754819869995, "learning_rate": 7.255392053593697e-05, "loss": 0.0275, "step": 30680 }, { "grad_norm": 0.28474149107933044, "learning_rate": 7.253547241089607e-05, "loss": 0.0255, "step": 30690 }, { "grad_norm": 0.21874679625034332, "learning_rate": 7.251702043515927e-05, "loss": 0.0295, "step": 30700 }, { "grad_norm": 0.23598641157150269, "learning_rate": 7.249856461187952e-05, "loss": 0.0223, "step": 30710 }, { "grad_norm": 0.2714079022407532, "learning_rate": 7.248010494421042e-05, "loss": 0.0285, "step": 30720 }, { "grad_norm": 0.20932766795158386, "learning_rate": 7.246164143530622e-05, "loss": 0.0356, "step": 30730 }, { "grad_norm": 0.29685428738594055, "learning_rate": 7.244317408832181e-05, "loss": 0.027, "step": 30740 }, { "grad_norm": 0.34558773040771484, "learning_rate": 7.242470290641279e-05, "loss": 0.0272, "step": 30750 }, { "grad_norm": 0.20719937980175018, "learning_rate": 7.240622789273536e-05, "loss": 0.0282, "step": 30760 }, { "grad_norm": 0.23308604955673218, "learning_rate": 7.238774905044638e-05, "loss": 0.0261, "step": 30770 }, { "grad_norm": 0.25725051760673523, "learning_rate": 7.236926638270341e-05, "loss": 0.0305, "step": 30780 }, { "grad_norm": 0.28570693731307983, "learning_rate": 7.23507798926646e-05, "loss": 0.0264, "step": 30790 }, { "grad_norm": 0.2857683300971985, "learning_rate": 7.23322895834888e-05, "loss": 0.0268, "step": 30800 }, { "grad_norm": 0.22070659697055817, "learning_rate": 7.231379545833552e-05, "loss": 0.0264, "step": 30810 }, { "grad_norm": 0.2902523875236511, "learning_rate": 7.229529752036487e-05, "loss": 0.0305, "step": 30820 }, { "grad_norm": 0.2542291581630707, "learning_rate": 7.227679577273765e-05, "loss": 0.0318, "step": 30830 }, { "grad_norm": 0.3473320007324219, "learning_rate": 7.225829021861529e-05, "loss": 0.0341, "step": 30840 }, { "grad_norm": 0.22449113428592682, "learning_rate": 7.223978086115992e-05, "loss": 0.0232, "step": 30850 }, { "grad_norm": 0.29555293917655945, "learning_rate": 7.222126770353425e-05, "loss": 0.0355, "step": 30860 }, { "grad_norm": 0.2296013981103897, "learning_rate": 7.22027507489017e-05, "loss": 0.0362, "step": 30870 }, { "grad_norm": 0.3191843330860138, "learning_rate": 7.218423000042627e-05, "loss": 0.0295, "step": 30880 }, { "grad_norm": 0.2578499913215637, "learning_rate": 7.216570546127268e-05, "loss": 0.0278, "step": 30890 }, { "grad_norm": 0.29787570238113403, "learning_rate": 7.214717713460626e-05, "loss": 0.0323, "step": 30900 }, { "grad_norm": 0.21796946227550507, "learning_rate": 7.2128645023593e-05, "loss": 0.029, "step": 30910 }, { "grad_norm": 0.44720908999443054, "learning_rate": 7.211010913139951e-05, "loss": 0.0311, "step": 30920 }, { "grad_norm": 0.2635688781738281, "learning_rate": 7.209156946119308e-05, "loss": 0.0267, "step": 30930 }, { "grad_norm": 0.3513869345188141, "learning_rate": 7.207302601614166e-05, "loss": 0.0294, "step": 30940 }, { "grad_norm": 0.2666938602924347, "learning_rate": 7.205447879941378e-05, "loss": 0.0339, "step": 30950 }, { "grad_norm": 0.5494781732559204, "learning_rate": 7.203592781417866e-05, "loss": 0.0293, "step": 30960 }, { "grad_norm": 0.611892580986023, "learning_rate": 7.201737306360617e-05, "loss": 0.038, "step": 30970 }, { "grad_norm": 0.2701118588447571, "learning_rate": 7.19988145508668e-05, "loss": 0.0316, "step": 30980 }, { "grad_norm": 0.3171060383319855, "learning_rate": 7.198025227913168e-05, "loss": 0.0289, "step": 30990 }, { "grad_norm": 0.2939682900905609, "learning_rate": 7.196168625157261e-05, "loss": 0.0308, "step": 31000 }, { "grad_norm": 0.31401747465133667, "learning_rate": 7.194311647136201e-05, "loss": 0.0263, "step": 31010 }, { "grad_norm": 0.33419573307037354, "learning_rate": 7.192454294167297e-05, "loss": 0.0264, "step": 31020 }, { "grad_norm": 0.34059008955955505, "learning_rate": 7.190596566567917e-05, "loss": 0.0271, "step": 31030 }, { "grad_norm": 0.26939794421195984, "learning_rate": 7.188738464655496e-05, "loss": 0.0359, "step": 31040 }, { "grad_norm": 0.27507656812667847, "learning_rate": 7.186879988747533e-05, "loss": 0.027, "step": 31050 }, { "grad_norm": 0.28371766209602356, "learning_rate": 7.185021139161592e-05, "loss": 0.0268, "step": 31060 }, { "grad_norm": 0.2806054651737213, "learning_rate": 7.1831619162153e-05, "loss": 0.0343, "step": 31070 }, { "grad_norm": 0.2671743929386139, "learning_rate": 7.181302320226345e-05, "loss": 0.028, "step": 31080 }, { "grad_norm": 0.32053959369659424, "learning_rate": 7.179442351512482e-05, "loss": 0.0322, "step": 31090 }, { "grad_norm": 0.24430081248283386, "learning_rate": 7.177582010391528e-05, "loss": 0.0298, "step": 31100 }, { "grad_norm": 0.32479792833328247, "learning_rate": 7.175721297181366e-05, "loss": 0.0317, "step": 31110 }, { "grad_norm": 0.3678549528121948, "learning_rate": 7.173860212199942e-05, "loss": 0.0329, "step": 31120 }, { "grad_norm": 0.2378084510564804, "learning_rate": 7.171998755765263e-05, "loss": 0.0338, "step": 31130 }, { "grad_norm": 0.26626378297805786, "learning_rate": 7.170136928195398e-05, "loss": 0.0289, "step": 31140 }, { "grad_norm": 0.24084711074829102, "learning_rate": 7.168274729808489e-05, "loss": 0.0315, "step": 31150 }, { "grad_norm": 0.513900876045227, "learning_rate": 7.166412160922728e-05, "loss": 0.0302, "step": 31160 }, { "grad_norm": 0.2838513255119324, "learning_rate": 7.164549221856382e-05, "loss": 0.0309, "step": 31170 }, { "grad_norm": 0.2589426040649414, "learning_rate": 7.162685912927775e-05, "loss": 0.0414, "step": 31180 }, { "grad_norm": 0.3917986750602722, "learning_rate": 7.160822234455294e-05, "loss": 0.0305, "step": 31190 }, { "grad_norm": 0.19317494332790375, "learning_rate": 7.158958186757391e-05, "loss": 0.0229, "step": 31200 }, { "grad_norm": 0.3184877336025238, "learning_rate": 7.157093770152582e-05, "loss": 0.031, "step": 31210 }, { "grad_norm": 0.2528682053089142, "learning_rate": 7.155228984959446e-05, "loss": 0.028, "step": 31220 }, { "grad_norm": 0.2846381664276123, "learning_rate": 7.153363831496621e-05, "loss": 0.0316, "step": 31230 }, { "grad_norm": 0.3069770038127899, "learning_rate": 7.151498310082811e-05, "loss": 0.0374, "step": 31240 }, { "grad_norm": 0.3127930760383606, "learning_rate": 7.149632421036784e-05, "loss": 0.0332, "step": 31250 }, { "grad_norm": 0.2955954372882843, "learning_rate": 7.147766164677369e-05, "loss": 0.0448, "step": 31260 }, { "grad_norm": 0.23947034776210785, "learning_rate": 7.145899541323459e-05, "loss": 0.0247, "step": 31270 }, { "grad_norm": 0.21943794190883636, "learning_rate": 7.144032551294007e-05, "loss": 0.0346, "step": 31280 }, { "grad_norm": 0.21181094646453857, "learning_rate": 7.14216519490803e-05, "loss": 0.0296, "step": 31290 }, { "grad_norm": 0.3391711711883545, "learning_rate": 7.140297472484609e-05, "loss": 0.0417, "step": 31300 }, { "grad_norm": 0.23496544361114502, "learning_rate": 7.138429384342891e-05, "loss": 0.0449, "step": 31310 }, { "grad_norm": 0.37311995029449463, "learning_rate": 7.136560930802074e-05, "loss": 0.0299, "step": 31320 }, { "grad_norm": 0.2495623528957367, "learning_rate": 7.134692112181431e-05, "loss": 0.0299, "step": 31330 }, { "grad_norm": 0.2791009247303009, "learning_rate": 7.13282292880029e-05, "loss": 0.0296, "step": 31340 }, { "grad_norm": 0.259838730096817, "learning_rate": 7.130953380978043e-05, "loss": 0.033, "step": 31350 }, { "grad_norm": 0.25146859884262085, "learning_rate": 7.129083469034144e-05, "loss": 0.0345, "step": 31360 }, { "grad_norm": 0.41366735100746155, "learning_rate": 7.127213193288112e-05, "loss": 0.0355, "step": 31370 }, { "grad_norm": 0.3378807008266449, "learning_rate": 7.125342554059522e-05, "loss": 0.0246, "step": 31380 }, { "grad_norm": 0.23911885917186737, "learning_rate": 7.12347155166802e-05, "loss": 0.0293, "step": 31390 }, { "grad_norm": 0.33117809891700745, "learning_rate": 7.121600186433306e-05, "loss": 0.0269, "step": 31400 }, { "grad_norm": 0.26861703395843506, "learning_rate": 7.119728458675148e-05, "loss": 0.0284, "step": 31410 }, { "grad_norm": 0.24926051497459412, "learning_rate": 7.117856368713369e-05, "loss": 0.0245, "step": 31420 }, { "grad_norm": 0.20481446385383606, "learning_rate": 7.115983916867861e-05, "loss": 0.0283, "step": 31430 }, { "grad_norm": 0.332110196352005, "learning_rate": 7.114111103458574e-05, "loss": 0.0275, "step": 31440 }, { "grad_norm": 0.2897534668445587, "learning_rate": 7.11223792880552e-05, "loss": 0.0273, "step": 31450 }, { "grad_norm": 0.2168746292591095, "learning_rate": 7.110364393228773e-05, "loss": 0.0264, "step": 31460 }, { "grad_norm": 0.23734305799007416, "learning_rate": 7.108490497048471e-05, "loss": 0.0233, "step": 31470 }, { "grad_norm": 0.21045956015586853, "learning_rate": 7.10661624058481e-05, "loss": 0.0277, "step": 31480 }, { "grad_norm": 0.18505308032035828, "learning_rate": 7.10474162415805e-05, "loss": 0.0306, "step": 31490 }, { "grad_norm": 0.3102905750274658, "learning_rate": 7.102866648088511e-05, "loss": 0.0322, "step": 31500 }, { "grad_norm": 0.33894410729408264, "learning_rate": 7.100991312696576e-05, "loss": 0.0262, "step": 31510 }, { "grad_norm": 0.23454907536506653, "learning_rate": 7.099115618302686e-05, "loss": 0.0299, "step": 31520 }, { "grad_norm": 0.2351417988538742, "learning_rate": 7.097239565227349e-05, "loss": 0.0329, "step": 31530 }, { "grad_norm": 0.24905405938625336, "learning_rate": 7.09536315379113e-05, "loss": 0.0252, "step": 31540 }, { "grad_norm": 0.3243683874607086, "learning_rate": 7.093486384314656e-05, "loss": 0.0282, "step": 31550 }, { "grad_norm": 0.19364559650421143, "learning_rate": 7.091609257118616e-05, "loss": 0.0297, "step": 31560 }, { "grad_norm": 0.41800782084465027, "learning_rate": 7.08973177252376e-05, "loss": 0.0314, "step": 31570 }, { "grad_norm": 0.3024318814277649, "learning_rate": 7.087853930850898e-05, "loss": 0.0292, "step": 31580 }, { "grad_norm": 0.21653510630130768, "learning_rate": 7.085975732420903e-05, "loss": 0.0271, "step": 31590 }, { "grad_norm": 0.20893016457557678, "learning_rate": 7.084097177554706e-05, "loss": 0.0236, "step": 31600 }, { "grad_norm": 0.31407779455184937, "learning_rate": 7.082218266573301e-05, "loss": 0.0325, "step": 31610 }, { "grad_norm": 0.30769479274749756, "learning_rate": 7.080338999797743e-05, "loss": 0.0323, "step": 31620 }, { "grad_norm": 0.363827645778656, "learning_rate": 7.07845937754915e-05, "loss": 0.029, "step": 31630 }, { "grad_norm": 0.384625107049942, "learning_rate": 7.076579400148693e-05, "loss": 0.0395, "step": 31640 }, { "grad_norm": 0.27306368947029114, "learning_rate": 7.074699067917611e-05, "loss": 0.0366, "step": 31650 }, { "grad_norm": 0.23787645995616913, "learning_rate": 7.072818381177201e-05, "loss": 0.03, "step": 31660 }, { "grad_norm": 0.2458057701587677, "learning_rate": 7.070937340248823e-05, "loss": 0.0317, "step": 31670 }, { "grad_norm": 0.262188583612442, "learning_rate": 7.069055945453893e-05, "loss": 0.0315, "step": 31680 }, { "grad_norm": 0.28374066948890686, "learning_rate": 7.067174197113892e-05, "loss": 0.0286, "step": 31690 }, { "grad_norm": 0.3331913352012634, "learning_rate": 7.065292095550355e-05, "loss": 0.0292, "step": 31700 }, { "grad_norm": 0.2698207497596741, "learning_rate": 7.063409641084887e-05, "loss": 0.0281, "step": 31710 }, { "grad_norm": 0.21376974880695343, "learning_rate": 7.061526834039145e-05, "loss": 0.0331, "step": 31720 }, { "grad_norm": 0.26217353343963623, "learning_rate": 7.05964367473485e-05, "loss": 0.0278, "step": 31730 }, { "grad_norm": 0.2625194787979126, "learning_rate": 7.057760163493783e-05, "loss": 0.0315, "step": 31740 }, { "grad_norm": 0.2895132899284363, "learning_rate": 7.055876300637783e-05, "loss": 0.0297, "step": 31750 }, { "grad_norm": 0.35462242364883423, "learning_rate": 7.053992086488753e-05, "loss": 0.0377, "step": 31760 }, { "grad_norm": 0.28457674384117126, "learning_rate": 7.052107521368651e-05, "loss": 0.0407, "step": 31770 }, { "grad_norm": 0.29271239042282104, "learning_rate": 7.0502226055995e-05, "loss": 0.0415, "step": 31780 }, { "grad_norm": 0.27486759424209595, "learning_rate": 7.048337339503379e-05, "loss": 0.0443, "step": 31790 }, { "grad_norm": 0.3086239993572235, "learning_rate": 7.046451723402427e-05, "loss": 0.0318, "step": 31800 }, { "grad_norm": 0.19937652349472046, "learning_rate": 7.044565757618848e-05, "loss": 0.0436, "step": 31810 }, { "grad_norm": 0.3637087941169739, "learning_rate": 7.042679442474899e-05, "loss": 0.03, "step": 31820 }, { "grad_norm": 0.27251484990119934, "learning_rate": 7.040792778292902e-05, "loss": 0.0401, "step": 31830 }, { "grad_norm": 0.33161425590515137, "learning_rate": 7.038905765395234e-05, "loss": 0.0382, "step": 31840 }, { "grad_norm": 0.27005526423454285, "learning_rate": 7.037018404104334e-05, "loss": 0.0369, "step": 31850 }, { "grad_norm": 0.23061102628707886, "learning_rate": 7.035130694742702e-05, "loss": 0.0265, "step": 31860 }, { "grad_norm": 0.22404076159000397, "learning_rate": 7.033242637632897e-05, "loss": 0.0321, "step": 31870 }, { "grad_norm": 0.2637149393558502, "learning_rate": 7.031354233097534e-05, "loss": 0.0356, "step": 31880 }, { "grad_norm": 0.31594252586364746, "learning_rate": 7.029465481459289e-05, "loss": 0.0419, "step": 31890 }, { "grad_norm": 0.21018370985984802, "learning_rate": 7.027576383040898e-05, "loss": 0.0269, "step": 31900 }, { "grad_norm": 0.2512417137622833, "learning_rate": 7.025686938165159e-05, "loss": 0.0392, "step": 31910 }, { "grad_norm": 0.34610113501548767, "learning_rate": 7.023797147154924e-05, "loss": 0.0476, "step": 31920 }, { "grad_norm": 0.34370186924934387, "learning_rate": 7.021907010333111e-05, "loss": 0.0279, "step": 31930 }, { "grad_norm": 0.2872489392757416, "learning_rate": 7.020016528022685e-05, "loss": 0.0275, "step": 31940 }, { "grad_norm": 0.23462025821208954, "learning_rate": 7.018125700546683e-05, "loss": 0.0365, "step": 31950 }, { "grad_norm": 0.28772008419036865, "learning_rate": 7.016234528228196e-05, "loss": 0.0355, "step": 31960 }, { "grad_norm": 0.33570221066474915, "learning_rate": 7.014343011390372e-05, "loss": 0.0286, "step": 31970 }, { "grad_norm": 0.2579125463962555, "learning_rate": 7.01245115035642e-05, "loss": 0.0245, "step": 31980 }, { "grad_norm": 0.23987172544002533, "learning_rate": 7.010558945449606e-05, "loss": 0.0388, "step": 31990 }, { "grad_norm": 0.2393573820590973, "learning_rate": 7.008666396993258e-05, "loss": 0.0234, "step": 32000 }, { "grad_norm": 0.23225173354148865, "learning_rate": 7.006773505310759e-05, "loss": 0.0347, "step": 32010 }, { "grad_norm": 0.1922873556613922, "learning_rate": 7.004880270725553e-05, "loss": 0.0279, "step": 32020 }, { "grad_norm": 0.3021600842475891, "learning_rate": 7.002986693561144e-05, "loss": 0.0308, "step": 32030 }, { "grad_norm": 0.3273591697216034, "learning_rate": 7.001092774141089e-05, "loss": 0.0333, "step": 32040 }, { "grad_norm": 0.2222432792186737, "learning_rate": 6.999198512789009e-05, "loss": 0.0295, "step": 32050 }, { "grad_norm": 0.22304901480674744, "learning_rate": 6.997303909828584e-05, "loss": 0.0332, "step": 32060 }, { "grad_norm": 0.3231596350669861, "learning_rate": 6.995408965583544e-05, "loss": 0.0323, "step": 32070 }, { "grad_norm": 0.26321524381637573, "learning_rate": 6.993513680377688e-05, "loss": 0.0288, "step": 32080 }, { "grad_norm": 0.2868462800979614, "learning_rate": 6.991618054534868e-05, "loss": 0.0333, "step": 32090 }, { "grad_norm": 0.23955468833446503, "learning_rate": 6.989722088378991e-05, "loss": 0.0292, "step": 32100 }, { "grad_norm": 0.30331897735595703, "learning_rate": 6.987825782234027e-05, "loss": 0.0301, "step": 32110 }, { "grad_norm": 0.309026300907135, "learning_rate": 6.985929136424006e-05, "loss": 0.0275, "step": 32120 }, { "grad_norm": 0.27074095606803894, "learning_rate": 6.984032151273012e-05, "loss": 0.032, "step": 32130 }, { "grad_norm": 0.245544895529747, "learning_rate": 6.982134827105186e-05, "loss": 0.0282, "step": 32140 }, { "grad_norm": 0.25514811277389526, "learning_rate": 6.980237164244729e-05, "loss": 0.0255, "step": 32150 }, { "grad_norm": 0.23547501862049103, "learning_rate": 6.9783391630159e-05, "loss": 0.0264, "step": 32160 }, { "grad_norm": 0.25403231382369995, "learning_rate": 6.976440823743015e-05, "loss": 0.0319, "step": 32170 }, { "grad_norm": 0.29603976011276245, "learning_rate": 6.974542146750451e-05, "loss": 0.0274, "step": 32180 }, { "grad_norm": 0.2091415971517563, "learning_rate": 6.972643132362637e-05, "loss": 0.0241, "step": 32190 }, { "grad_norm": 0.20952442288398743, "learning_rate": 6.970743780904064e-05, "loss": 0.0305, "step": 32200 }, { "grad_norm": 0.2297494113445282, "learning_rate": 6.968844092699277e-05, "loss": 0.0292, "step": 32210 }, { "grad_norm": 0.2634357810020447, "learning_rate": 6.966944068072883e-05, "loss": 0.0284, "step": 32220 }, { "grad_norm": 0.26181209087371826, "learning_rate": 6.965043707349545e-05, "loss": 0.0257, "step": 32230 }, { "grad_norm": 0.4525354504585266, "learning_rate": 6.963143010853982e-05, "loss": 0.0329, "step": 32240 }, { "grad_norm": 0.25501275062561035, "learning_rate": 6.961241978910971e-05, "loss": 0.0333, "step": 32250 }, { "grad_norm": 0.26553788781166077, "learning_rate": 6.959340611845344e-05, "loss": 0.0266, "step": 32260 }, { "grad_norm": 0.31905069947242737, "learning_rate": 6.957438909981995e-05, "loss": 0.0279, "step": 32270 }, { "grad_norm": 0.21225395798683167, "learning_rate": 6.955536873645872e-05, "loss": 0.0232, "step": 32280 }, { "grad_norm": 0.2578237056732178, "learning_rate": 6.953634503161982e-05, "loss": 0.026, "step": 32290 }, { "grad_norm": 0.23348848521709442, "learning_rate": 6.951731798855387e-05, "loss": 0.0265, "step": 32300 }, { "grad_norm": 0.20767925679683685, "learning_rate": 6.949828761051208e-05, "loss": 0.034, "step": 32310 }, { "grad_norm": 0.22502857446670532, "learning_rate": 6.947925390074622e-05, "loss": 0.0251, "step": 32320 }, { "grad_norm": 0.22535714507102966, "learning_rate": 6.946021686250863e-05, "loss": 0.0252, "step": 32330 }, { "grad_norm": 0.26340654492378235, "learning_rate": 6.94411764990522e-05, "loss": 0.027, "step": 32340 }, { "grad_norm": 0.21516625583171844, "learning_rate": 6.942213281363044e-05, "loss": 0.0239, "step": 32350 }, { "grad_norm": 0.18349681794643402, "learning_rate": 6.940308580949737e-05, "loss": 0.0248, "step": 32360 }, { "grad_norm": 0.2225746065378189, "learning_rate": 6.93840354899076e-05, "loss": 0.0275, "step": 32370 }, { "grad_norm": 0.19038614630699158, "learning_rate": 6.936498185811633e-05, "loss": 0.0262, "step": 32380 }, { "grad_norm": 0.22536753118038177, "learning_rate": 6.93459249173793e-05, "loss": 0.0227, "step": 32390 }, { "grad_norm": 0.27949726581573486, "learning_rate": 6.932686467095279e-05, "loss": 0.028, "step": 32400 }, { "grad_norm": 0.30995649099349976, "learning_rate": 6.930780112209373e-05, "loss": 0.0274, "step": 32410 }, { "grad_norm": 0.2567163109779358, "learning_rate": 6.92887342740595e-05, "loss": 0.0279, "step": 32420 }, { "grad_norm": 0.2252110242843628, "learning_rate": 6.926966413010816e-05, "loss": 0.0362, "step": 32430 }, { "grad_norm": 0.27396848797798157, "learning_rate": 6.925059069349824e-05, "loss": 0.0242, "step": 32440 }, { "grad_norm": 0.2170693725347519, "learning_rate": 6.923151396748886e-05, "loss": 0.029, "step": 32450 }, { "grad_norm": 0.22654755413532257, "learning_rate": 6.921243395533974e-05, "loss": 0.0233, "step": 32460 }, { "grad_norm": 0.3012421727180481, "learning_rate": 6.919335066031109e-05, "loss": 0.0345, "step": 32470 }, { "grad_norm": 0.28933432698249817, "learning_rate": 6.917426408566379e-05, "loss": 0.0328, "step": 32480 }, { "grad_norm": 0.32019945979118347, "learning_rate": 6.915517423465916e-05, "loss": 0.0302, "step": 32490 }, { "grad_norm": 0.2301032990217209, "learning_rate": 6.913608111055914e-05, "loss": 0.0296, "step": 32500 }, { "grad_norm": 0.25388798117637634, "learning_rate": 6.911698471662623e-05, "loss": 0.0266, "step": 32510 }, { "grad_norm": 0.23563289642333984, "learning_rate": 6.90978850561235e-05, "loss": 0.0243, "step": 32520 }, { "grad_norm": 0.22498998045921326, "learning_rate": 6.907878213231454e-05, "loss": 0.0221, "step": 32530 }, { "grad_norm": 0.3087436556816101, "learning_rate": 6.90596759484635e-05, "loss": 0.0322, "step": 32540 }, { "grad_norm": 0.26691940426826477, "learning_rate": 6.904056650783514e-05, "loss": 0.0375, "step": 32550 }, { "grad_norm": 0.22562715411186218, "learning_rate": 6.902145381369471e-05, "loss": 0.0336, "step": 32560 }, { "grad_norm": 0.19305641949176788, "learning_rate": 6.900233786930808e-05, "loss": 0.0286, "step": 32570 }, { "grad_norm": 0.3002738058567047, "learning_rate": 6.898321867794161e-05, "loss": 0.0282, "step": 32580 }, { "grad_norm": 0.19395166635513306, "learning_rate": 6.896409624286226e-05, "loss": 0.0271, "step": 32590 }, { "grad_norm": 0.28286683559417725, "learning_rate": 6.894497056733754e-05, "loss": 0.0303, "step": 32600 }, { "grad_norm": 0.21295343339443207, "learning_rate": 6.89258416546355e-05, "loss": 0.0261, "step": 32610 }, { "grad_norm": 0.22309610247612, "learning_rate": 6.890670950802474e-05, "loss": 0.0307, "step": 32620 }, { "grad_norm": 0.2859935760498047, "learning_rate": 6.88875741307744e-05, "loss": 0.0259, "step": 32630 }, { "grad_norm": 0.3106079399585724, "learning_rate": 6.886843552615425e-05, "loss": 0.0311, "step": 32640 }, { "grad_norm": 0.20196163654327393, "learning_rate": 6.884929369743451e-05, "loss": 0.0264, "step": 32650 }, { "grad_norm": 0.2846902310848236, "learning_rate": 6.8830148647886e-05, "loss": 0.0267, "step": 32660 }, { "grad_norm": 0.28179046511650085, "learning_rate": 6.88110003807801e-05, "loss": 0.0353, "step": 32670 }, { "grad_norm": 0.2966645359992981, "learning_rate": 6.87918488993887e-05, "loss": 0.029, "step": 32680 }, { "grad_norm": 0.29555103182792664, "learning_rate": 6.877269420698431e-05, "loss": 0.0326, "step": 32690 }, { "grad_norm": 0.3103927969932556, "learning_rate": 6.875353630683989e-05, "loss": 0.034, "step": 32700 }, { "grad_norm": 0.4120451509952545, "learning_rate": 6.873437520222905e-05, "loss": 0.0393, "step": 32710 }, { "grad_norm": 0.24308224022388458, "learning_rate": 6.871521089642585e-05, "loss": 0.0268, "step": 32720 }, { "grad_norm": 0.22069497406482697, "learning_rate": 6.869604339270498e-05, "loss": 0.0283, "step": 32730 }, { "grad_norm": 0.2887929081916809, "learning_rate": 6.867687269434164e-05, "loss": 0.0323, "step": 32740 }, { "grad_norm": 0.2763567268848419, "learning_rate": 6.865769880461156e-05, "loss": 0.0272, "step": 32750 }, { "grad_norm": 0.1717955768108368, "learning_rate": 6.863852172679104e-05, "loss": 0.0282, "step": 32760 }, { "grad_norm": 0.32923999428749084, "learning_rate": 6.861934146415693e-05, "loss": 0.0314, "step": 32770 }, { "grad_norm": 0.2741430401802063, "learning_rate": 6.86001580199866e-05, "loss": 0.0368, "step": 32780 }, { "grad_norm": 0.26999804377555847, "learning_rate": 6.858097139755798e-05, "loss": 0.0338, "step": 32790 }, { "grad_norm": 0.2973218858242035, "learning_rate": 6.856178160014955e-05, "loss": 0.0339, "step": 32800 }, { "grad_norm": 0.21348458528518677, "learning_rate": 6.85425886310403e-05, "loss": 0.0293, "step": 32810 }, { "grad_norm": 0.2685355246067047, "learning_rate": 6.852339249350979e-05, "loss": 0.0305, "step": 32820 }, { "grad_norm": 0.27222082018852234, "learning_rate": 6.850419319083812e-05, "loss": 0.0395, "step": 32830 }, { "grad_norm": 0.24112601578235626, "learning_rate": 6.848499072630592e-05, "loss": 0.0249, "step": 32840 }, { "grad_norm": 0.3003871440887451, "learning_rate": 6.846578510319439e-05, "loss": 0.0327, "step": 32850 }, { "grad_norm": 0.21703284978866577, "learning_rate": 6.844657632478519e-05, "loss": 0.0326, "step": 32860 }, { "grad_norm": 0.2652033269405365, "learning_rate": 6.842736439436063e-05, "loss": 0.0261, "step": 32870 }, { "grad_norm": 0.28410032391548157, "learning_rate": 6.84081493152035e-05, "loss": 0.033, "step": 32880 }, { "grad_norm": 0.2767336666584015, "learning_rate": 6.83889310905971e-05, "loss": 0.0253, "step": 32890 }, { "grad_norm": 0.21727225184440613, "learning_rate": 6.836970972382533e-05, "loss": 0.0314, "step": 32900 }, { "grad_norm": 0.21799688041210175, "learning_rate": 6.835048521817257e-05, "loss": 0.0291, "step": 32910 }, { "grad_norm": 0.2568223178386688, "learning_rate": 6.833125757692379e-05, "loss": 0.0348, "step": 32920 }, { "grad_norm": 0.2179923951625824, "learning_rate": 6.831202680336441e-05, "loss": 0.0249, "step": 32930 }, { "grad_norm": 0.28357160091400146, "learning_rate": 6.829279290078052e-05, "loss": 0.0257, "step": 32940 }, { "grad_norm": 0.31103408336639404, "learning_rate": 6.827355587245863e-05, "loss": 0.0283, "step": 32950 }, { "grad_norm": 0.25269895792007446, "learning_rate": 6.82543157216858e-05, "loss": 0.026, "step": 32960 }, { "grad_norm": 0.19733257591724396, "learning_rate": 6.823507245174969e-05, "loss": 0.024, "step": 32970 }, { "grad_norm": 0.264529824256897, "learning_rate": 6.821582606593841e-05, "loss": 0.0254, "step": 32980 }, { "grad_norm": 0.29717713594436646, "learning_rate": 6.81965765675407e-05, "loss": 0.0276, "step": 32990 }, { "grad_norm": 0.2957804203033447, "learning_rate": 6.81773239598457e-05, "loss": 0.0275, "step": 33000 }, { "grad_norm": 0.30737727880477905, "learning_rate": 6.815806824614319e-05, "loss": 0.0267, "step": 33010 }, { "grad_norm": 0.24976898729801178, "learning_rate": 6.813880942972343e-05, "loss": 0.031, "step": 33020 }, { "grad_norm": 0.26925820112228394, "learning_rate": 6.811954751387726e-05, "loss": 0.0273, "step": 33030 }, { "grad_norm": 0.2566337585449219, "learning_rate": 6.810028250189598e-05, "loss": 0.0308, "step": 33040 }, { "grad_norm": 0.27785637974739075, "learning_rate": 6.808101439707147e-05, "loss": 0.0373, "step": 33050 }, { "grad_norm": 0.3238111138343811, "learning_rate": 6.806174320269609e-05, "loss": 0.029, "step": 33060 }, { "grad_norm": 0.22027340531349182, "learning_rate": 6.804246892206281e-05, "loss": 0.0285, "step": 33070 }, { "grad_norm": 0.26593273878097534, "learning_rate": 6.802319155846506e-05, "loss": 0.0418, "step": 33080 }, { "grad_norm": 0.25695160031318665, "learning_rate": 6.800391111519679e-05, "loss": 0.0254, "step": 33090 }, { "grad_norm": 0.2190997302532196, "learning_rate": 6.798462759555253e-05, "loss": 0.0344, "step": 33100 }, { "grad_norm": 0.26043206453323364, "learning_rate": 6.79653410028273e-05, "loss": 0.0272, "step": 33110 }, { "grad_norm": 0.22395116090774536, "learning_rate": 6.794605134031663e-05, "loss": 0.0287, "step": 33120 }, { "grad_norm": 0.2684347629547119, "learning_rate": 6.792675861131661e-05, "loss": 0.0269, "step": 33130 }, { "grad_norm": 0.27385783195495605, "learning_rate": 6.790746281912386e-05, "loss": 0.0328, "step": 33140 }, { "grad_norm": 0.24321399629116058, "learning_rate": 6.788816396703546e-05, "loss": 0.0199, "step": 33150 }, { "grad_norm": 0.3418519198894501, "learning_rate": 6.78688620583491e-05, "loss": 0.0276, "step": 33160 }, { "grad_norm": 0.3211621344089508, "learning_rate": 6.784955709636292e-05, "loss": 0.0257, "step": 33170 }, { "grad_norm": 0.30519071221351624, "learning_rate": 6.783024908437564e-05, "loss": 0.0277, "step": 33180 }, { "grad_norm": 0.2641049325466156, "learning_rate": 6.781093802568641e-05, "loss": 0.029, "step": 33190 }, { "grad_norm": 0.1588832437992096, "learning_rate": 6.779162392359504e-05, "loss": 0.022, "step": 33200 }, { "grad_norm": 0.2328253537416458, "learning_rate": 6.777230678140172e-05, "loss": 0.0278, "step": 33210 }, { "grad_norm": 0.28180500864982605, "learning_rate": 6.775298660240726e-05, "loss": 0.0274, "step": 33220 }, { "grad_norm": 0.3690125346183777, "learning_rate": 6.773366338991292e-05, "loss": 0.0388, "step": 33230 }, { "grad_norm": 0.26591551303863525, "learning_rate": 6.771433714722052e-05, "loss": 0.0263, "step": 33240 }, { "grad_norm": 0.2877618074417114, "learning_rate": 6.769500787763239e-05, "loss": 0.0321, "step": 33250 }, { "grad_norm": 0.19900231063365936, "learning_rate": 6.76756755844514e-05, "loss": 0.0276, "step": 33260 }, { "grad_norm": 0.2585359513759613, "learning_rate": 6.765634027098087e-05, "loss": 0.0271, "step": 33270 }, { "grad_norm": 0.3139580190181732, "learning_rate": 6.763700194052468e-05, "loss": 0.0324, "step": 33280 }, { "grad_norm": 0.2452988177537918, "learning_rate": 6.761766059638723e-05, "loss": 0.0242, "step": 33290 }, { "grad_norm": 0.32760363817214966, "learning_rate": 6.759831624187345e-05, "loss": 0.0303, "step": 33300 }, { "grad_norm": 0.3568437099456787, "learning_rate": 6.757896888028871e-05, "loss": 0.0299, "step": 33310 }, { "grad_norm": 0.18263140320777893, "learning_rate": 6.7559618514939e-05, "loss": 0.0264, "step": 33320 }, { "grad_norm": 0.27503809332847595, "learning_rate": 6.754026514913073e-05, "loss": 0.0254, "step": 33330 }, { "grad_norm": 0.32963934540748596, "learning_rate": 6.752090878617087e-05, "loss": 0.0331, "step": 33340 }, { "grad_norm": 0.2649499177932739, "learning_rate": 6.75015494293669e-05, "loss": 0.0264, "step": 33350 }, { "grad_norm": 0.2815909683704376, "learning_rate": 6.74821870820268e-05, "loss": 0.0236, "step": 33360 }, { "grad_norm": 0.19418960809707642, "learning_rate": 6.746282174745907e-05, "loss": 0.0309, "step": 33370 }, { "grad_norm": 0.30871909856796265, "learning_rate": 6.744345342897271e-05, "loss": 0.0348, "step": 33380 }, { "grad_norm": 0.2535719871520996, "learning_rate": 6.742408212987724e-05, "loss": 0.0321, "step": 33390 }, { "grad_norm": 0.22918210923671722, "learning_rate": 6.740470785348269e-05, "loss": 0.0219, "step": 33400 }, { "grad_norm": 0.31903865933418274, "learning_rate": 6.738533060309958e-05, "loss": 0.029, "step": 33410 }, { "grad_norm": 0.2880958914756775, "learning_rate": 6.736595038203894e-05, "loss": 0.0222, "step": 33420 }, { "grad_norm": 0.20910225808620453, "learning_rate": 6.734656719361236e-05, "loss": 0.027, "step": 33430 }, { "grad_norm": 0.24067342281341553, "learning_rate": 6.732718104113189e-05, "loss": 0.0251, "step": 33440 }, { "grad_norm": 0.31719639897346497, "learning_rate": 6.730779192791006e-05, "loss": 0.0227, "step": 33450 }, { "grad_norm": 0.26111382246017456, "learning_rate": 6.728839985725997e-05, "loss": 0.0249, "step": 33460 }, { "grad_norm": 0.2365211546421051, "learning_rate": 6.726900483249517e-05, "loss": 0.025, "step": 33470 }, { "grad_norm": 0.24989672005176544, "learning_rate": 6.724960685692976e-05, "loss": 0.0239, "step": 33480 }, { "grad_norm": 0.24946065247058868, "learning_rate": 6.723020593387833e-05, "loss": 0.0253, "step": 33490 }, { "grad_norm": 0.32837894558906555, "learning_rate": 6.721080206665593e-05, "loss": 0.0278, "step": 33500 }, { "grad_norm": 0.23176661133766174, "learning_rate": 6.719139525857819e-05, "loss": 0.0319, "step": 33510 }, { "grad_norm": 0.2648065388202667, "learning_rate": 6.717198551296117e-05, "loss": 0.0255, "step": 33520 }, { "grad_norm": 0.23141971230506897, "learning_rate": 6.715257283312148e-05, "loss": 0.0267, "step": 33530 }, { "grad_norm": 0.3022921085357666, "learning_rate": 6.713315722237623e-05, "loss": 0.0262, "step": 33540 }, { "grad_norm": 0.29465481638908386, "learning_rate": 6.7113738684043e-05, "loss": 0.0353, "step": 33550 }, { "grad_norm": 0.27197396755218506, "learning_rate": 6.709431722143989e-05, "loss": 0.029, "step": 33560 }, { "grad_norm": 0.2209663987159729, "learning_rate": 6.70748928378855e-05, "loss": 0.0332, "step": 33570 }, { "grad_norm": 0.23720605671405792, "learning_rate": 6.705546553669891e-05, "loss": 0.0245, "step": 33580 }, { "grad_norm": 0.23910924792289734, "learning_rate": 6.703603532119974e-05, "loss": 0.0236, "step": 33590 }, { "grad_norm": 0.21962833404541016, "learning_rate": 6.701660219470808e-05, "loss": 0.0245, "step": 33600 }, { "grad_norm": 0.25697779655456543, "learning_rate": 6.69971661605445e-05, "loss": 0.0237, "step": 33610 }, { "grad_norm": 0.2929518222808838, "learning_rate": 6.697772722203008e-05, "loss": 0.0282, "step": 33620 }, { "grad_norm": 0.28135162591934204, "learning_rate": 6.695828538248643e-05, "loss": 0.0254, "step": 33630 }, { "grad_norm": 0.2784777283668518, "learning_rate": 6.693884064523563e-05, "loss": 0.0284, "step": 33640 }, { "grad_norm": 0.229168102145195, "learning_rate": 6.691939301360023e-05, "loss": 0.024, "step": 33650 }, { "grad_norm": 0.24006159603595734, "learning_rate": 6.689994249090333e-05, "loss": 0.0248, "step": 33660 }, { "grad_norm": 0.2926398515701294, "learning_rate": 6.688048908046845e-05, "loss": 0.0296, "step": 33670 }, { "grad_norm": 0.20606710016727448, "learning_rate": 6.686103278561969e-05, "loss": 0.0249, "step": 33680 }, { "grad_norm": 0.3018878400325775, "learning_rate": 6.684157360968156e-05, "loss": 0.0312, "step": 33690 }, { "grad_norm": 0.25460144877433777, "learning_rate": 6.682211155597911e-05, "loss": 0.0247, "step": 33700 }, { "grad_norm": 0.2035936415195465, "learning_rate": 6.680264662783789e-05, "loss": 0.0274, "step": 33710 }, { "grad_norm": 0.187309131026268, "learning_rate": 6.678317882858391e-05, "loss": 0.0264, "step": 33720 }, { "grad_norm": 0.2186030000448227, "learning_rate": 6.67637081615437e-05, "loss": 0.0261, "step": 33730 }, { "grad_norm": 0.167196586728096, "learning_rate": 6.674423463004427e-05, "loss": 0.0233, "step": 33740 }, { "grad_norm": 0.2552645802497864, "learning_rate": 6.672475823741308e-05, "loss": 0.0248, "step": 33750 }, { "grad_norm": 0.2752331793308258, "learning_rate": 6.670527898697811e-05, "loss": 0.028, "step": 33760 }, { "grad_norm": 0.2531323730945587, "learning_rate": 6.668579688206788e-05, "loss": 0.0284, "step": 33770 }, { "grad_norm": 0.29977017641067505, "learning_rate": 6.666631192601131e-05, "loss": 0.0265, "step": 33780 }, { "grad_norm": 0.25846272706985474, "learning_rate": 6.664682412213785e-05, "loss": 0.0274, "step": 33790 }, { "grad_norm": 0.3455706536769867, "learning_rate": 6.662733347377745e-05, "loss": 0.0297, "step": 33800 }, { "grad_norm": 0.24231648445129395, "learning_rate": 6.660783998426051e-05, "loss": 0.0237, "step": 33810 }, { "grad_norm": 0.18899933993816376, "learning_rate": 6.658834365691794e-05, "loss": 0.0273, "step": 33820 }, { "grad_norm": 0.24848540127277374, "learning_rate": 6.656884449508115e-05, "loss": 0.0281, "step": 33830 }, { "grad_norm": 0.24778462946414948, "learning_rate": 6.654934250208198e-05, "loss": 0.0288, "step": 33840 }, { "grad_norm": 0.29626306891441345, "learning_rate": 6.65298376812528e-05, "loss": 0.0271, "step": 33850 }, { "grad_norm": 0.221525639295578, "learning_rate": 6.651033003592646e-05, "loss": 0.0301, "step": 33860 }, { "grad_norm": 0.4133853316307068, "learning_rate": 6.649081956943626e-05, "loss": 0.0243, "step": 33870 }, { "grad_norm": 0.3023349642753601, "learning_rate": 6.647130628511604e-05, "loss": 0.0285, "step": 33880 }, { "grad_norm": 0.29796090722084045, "learning_rate": 6.645179018630005e-05, "loss": 0.0299, "step": 33890 }, { "grad_norm": 0.2905804216861725, "learning_rate": 6.643227127632309e-05, "loss": 0.0235, "step": 33900 }, { "grad_norm": 0.3207780718803406, "learning_rate": 6.641274955852038e-05, "loss": 0.0313, "step": 33910 }, { "grad_norm": 0.3661394417285919, "learning_rate": 6.639322503622768e-05, "loss": 0.0499, "step": 33920 }, { "grad_norm": 0.26703163981437683, "learning_rate": 6.637369771278116e-05, "loss": 0.0252, "step": 33930 }, { "grad_norm": 0.5102611780166626, "learning_rate": 6.635416759151751e-05, "loss": 0.0245, "step": 33940 }, { "grad_norm": 0.2768203914165497, "learning_rate": 6.633463467577394e-05, "loss": 0.0423, "step": 33950 }, { "grad_norm": 0.2501097321510315, "learning_rate": 6.631509896888803e-05, "loss": 0.0267, "step": 33960 }, { "grad_norm": 0.300212025642395, "learning_rate": 6.629556047419794e-05, "loss": 0.0245, "step": 33970 }, { "grad_norm": 0.2838364839553833, "learning_rate": 6.627601919504223e-05, "loss": 0.0224, "step": 33980 }, { "grad_norm": 0.23152391612529755, "learning_rate": 6.625647513476001e-05, "loss": 0.0218, "step": 33990 }, { "grad_norm": 0.3009125888347626, "learning_rate": 6.62369282966908e-05, "loss": 0.0311, "step": 34000 }, { "grad_norm": 0.25253039598464966, "learning_rate": 6.621737868417464e-05, "loss": 0.0286, "step": 34010 }, { "grad_norm": 0.2216651439666748, "learning_rate": 6.619782630055198e-05, "loss": 0.0255, "step": 34020 }, { "grad_norm": 0.23047709465026855, "learning_rate": 6.617827114916382e-05, "loss": 0.026, "step": 34030 }, { "grad_norm": 0.2537482976913452, "learning_rate": 6.615871323335161e-05, "loss": 0.0308, "step": 34040 }, { "grad_norm": 0.1647738665342331, "learning_rate": 6.613915255645725e-05, "loss": 0.0258, "step": 34050 }, { "grad_norm": 0.2832794487476349, "learning_rate": 6.611958912182312e-05, "loss": 0.0253, "step": 34060 }, { "grad_norm": 0.18148484826087952, "learning_rate": 6.610002293279207e-05, "loss": 0.0232, "step": 34070 }, { "grad_norm": 0.2832314968109131, "learning_rate": 6.608045399270746e-05, "loss": 0.0269, "step": 34080 }, { "grad_norm": 0.20207512378692627, "learning_rate": 6.606088230491304e-05, "loss": 0.0251, "step": 34090 }, { "grad_norm": 0.27953994274139404, "learning_rate": 6.604130787275312e-05, "loss": 0.0312, "step": 34100 }, { "grad_norm": 0.20752836763858795, "learning_rate": 6.602173069957242e-05, "loss": 0.0229, "step": 34110 }, { "grad_norm": 0.2339184284210205, "learning_rate": 6.600215078871612e-05, "loss": 0.0238, "step": 34120 }, { "grad_norm": 0.3225305676460266, "learning_rate": 6.598256814352992e-05, "loss": 0.0251, "step": 34130 }, { "grad_norm": 0.2620900869369507, "learning_rate": 6.596298276735995e-05, "loss": 0.0196, "step": 34140 }, { "grad_norm": 0.23092655837535858, "learning_rate": 6.594339466355282e-05, "loss": 0.0275, "step": 34150 }, { "grad_norm": 0.24316127598285675, "learning_rate": 6.592380383545558e-05, "loss": 0.0252, "step": 34160 }, { "grad_norm": 0.24399027228355408, "learning_rate": 6.590421028641577e-05, "loss": 0.0309, "step": 34170 }, { "grad_norm": 0.3294280767440796, "learning_rate": 6.588461401978143e-05, "loss": 0.0301, "step": 34180 }, { "grad_norm": 0.220940500497818, "learning_rate": 6.586501503890099e-05, "loss": 0.0271, "step": 34190 }, { "grad_norm": 0.24240267276763916, "learning_rate": 6.584541334712338e-05, "loss": 0.0296, "step": 34200 }, { "grad_norm": 0.25136473774909973, "learning_rate": 6.582580894779802e-05, "loss": 0.03, "step": 34210 }, { "grad_norm": 0.21138979494571686, "learning_rate": 6.580620184427473e-05, "loss": 0.0251, "step": 34220 }, { "grad_norm": 0.16924524307250977, "learning_rate": 6.578659203990385e-05, "loss": 0.0246, "step": 34230 }, { "grad_norm": 0.32435593008995056, "learning_rate": 6.576697953803615e-05, "loss": 0.0255, "step": 34240 }, { "grad_norm": 0.27558791637420654, "learning_rate": 6.57473643420229e-05, "loss": 0.026, "step": 34250 }, { "grad_norm": 0.3347831666469574, "learning_rate": 6.572774645521574e-05, "loss": 0.0304, "step": 34260 }, { "grad_norm": 0.3120850920677185, "learning_rate": 6.570812588096688e-05, "loss": 0.0341, "step": 34270 }, { "grad_norm": 0.2401723712682724, "learning_rate": 6.568850262262893e-05, "loss": 0.0322, "step": 34280 }, { "grad_norm": 0.3082127571105957, "learning_rate": 6.566887668355497e-05, "loss": 0.0262, "step": 34290 }, { "grad_norm": 0.2389736771583557, "learning_rate": 6.564924806709851e-05, "loss": 0.0238, "step": 34300 }, { "grad_norm": 0.2953590154647827, "learning_rate": 6.562961677661359e-05, "loss": 0.0302, "step": 34310 }, { "grad_norm": 0.3137045204639435, "learning_rate": 6.56099828154546e-05, "loss": 0.0251, "step": 34320 }, { "grad_norm": 0.27200815081596375, "learning_rate": 6.55903461869765e-05, "loss": 0.0258, "step": 34330 }, { "grad_norm": 0.19397415220737457, "learning_rate": 6.557070689453465e-05, "loss": 0.0243, "step": 34340 }, { "grad_norm": 0.19303907454013824, "learning_rate": 6.555106494148482e-05, "loss": 0.0345, "step": 34350 }, { "grad_norm": 0.30016982555389404, "learning_rate": 6.553142033118333e-05, "loss": 0.0242, "step": 34360 }, { "grad_norm": 0.2315594106912613, "learning_rate": 6.551177306698688e-05, "loss": 0.0303, "step": 34370 }, { "grad_norm": 0.34032103419303894, "learning_rate": 6.549212315225267e-05, "loss": 0.032, "step": 34380 }, { "grad_norm": 0.25532257556915283, "learning_rate": 6.547247059033833e-05, "loss": 0.0308, "step": 34390 }, { "grad_norm": 0.28505998849868774, "learning_rate": 6.545281538460193e-05, "loss": 0.0278, "step": 34400 }, { "grad_norm": 0.19500593841075897, "learning_rate": 6.543315753840202e-05, "loss": 0.0262, "step": 34410 }, { "grad_norm": 0.37303218245506287, "learning_rate": 6.541349705509758e-05, "loss": 0.0261, "step": 34420 }, { "grad_norm": 0.2808196544647217, "learning_rate": 6.539383393804805e-05, "loss": 0.0274, "step": 34430 }, { "grad_norm": 0.32938337326049805, "learning_rate": 6.537416819061333e-05, "loss": 0.0299, "step": 34440 }, { "grad_norm": 0.2594938278198242, "learning_rate": 6.535449981615375e-05, "loss": 0.0287, "step": 34450 }, { "grad_norm": 0.2536807656288147, "learning_rate": 6.53348288180301e-05, "loss": 0.0274, "step": 34460 }, { "grad_norm": 0.22995008528232574, "learning_rate": 6.531515519960361e-05, "loss": 0.0347, "step": 34470 }, { "grad_norm": 0.2113136351108551, "learning_rate": 6.529547896423597e-05, "loss": 0.0264, "step": 34480 }, { "grad_norm": 0.20314081013202667, "learning_rate": 6.52758001152893e-05, "loss": 0.0305, "step": 34490 }, { "grad_norm": 0.21589656174182892, "learning_rate": 6.525611865612618e-05, "loss": 0.024, "step": 34500 }, { "grad_norm": 0.24582798779010773, "learning_rate": 6.523643459010966e-05, "loss": 0.0277, "step": 34510 }, { "grad_norm": 0.3180359899997711, "learning_rate": 6.521674792060317e-05, "loss": 0.0259, "step": 34520 }, { "grad_norm": 0.3303970396518707, "learning_rate": 6.519705865097063e-05, "loss": 0.0313, "step": 34530 }, { "grad_norm": 0.28941813111305237, "learning_rate": 6.517736678457641e-05, "loss": 0.0245, "step": 34540 }, { "grad_norm": 0.20486503839492798, "learning_rate": 6.515767232478534e-05, "loss": 0.0286, "step": 34550 }, { "grad_norm": 0.26221364736557007, "learning_rate": 6.51379752749626e-05, "loss": 0.0213, "step": 34560 }, { "grad_norm": 0.24527201056480408, "learning_rate": 6.511827563847393e-05, "loss": 0.0258, "step": 34570 }, { "grad_norm": 0.3815433084964752, "learning_rate": 6.509857341868542e-05, "loss": 0.0321, "step": 34580 }, { "grad_norm": 0.28745442628860474, "learning_rate": 6.507886861896367e-05, "loss": 0.0237, "step": 34590 }, { "grad_norm": 0.2127615511417389, "learning_rate": 6.505916124267567e-05, "loss": 0.0302, "step": 34600 }, { "grad_norm": 0.30943766236305237, "learning_rate": 6.503945129318891e-05, "loss": 0.0268, "step": 34610 }, { "grad_norm": 0.2885616719722748, "learning_rate": 6.501973877387122e-05, "loss": 0.0248, "step": 34620 }, { "grad_norm": 0.3020613193511963, "learning_rate": 6.500002368809098e-05, "loss": 0.027, "step": 34630 }, { "grad_norm": 0.22967737913131714, "learning_rate": 6.498030603921694e-05, "loss": 0.0252, "step": 34640 }, { "grad_norm": 0.35503682494163513, "learning_rate": 6.496058583061832e-05, "loss": 0.03, "step": 34650 }, { "grad_norm": 0.29165783524513245, "learning_rate": 6.494086306566475e-05, "loss": 0.0229, "step": 34660 }, { "grad_norm": 0.24982605874538422, "learning_rate": 6.492113774772632e-05, "loss": 0.0283, "step": 34670 }, { "grad_norm": 0.23629270493984222, "learning_rate": 6.490140988017354e-05, "loss": 0.0306, "step": 34680 }, { "grad_norm": 0.21478720009326935, "learning_rate": 6.488167946637736e-05, "loss": 0.0271, "step": 34690 }, { "grad_norm": 0.2568235397338867, "learning_rate": 6.486194650970915e-05, "loss": 0.027, "step": 34700 }, { "grad_norm": 0.41351771354675293, "learning_rate": 6.48422110135408e-05, "loss": 0.0282, "step": 34710 }, { "grad_norm": 0.2376459687948227, "learning_rate": 6.482247298124451e-05, "loss": 0.0357, "step": 34720 }, { "grad_norm": 0.2800801694393158, "learning_rate": 6.480273241619297e-05, "loss": 0.0261, "step": 34730 }, { "grad_norm": 0.2815338373184204, "learning_rate": 6.478298932175933e-05, "loss": 0.028, "step": 34740 }, { "grad_norm": 0.23846714198589325, "learning_rate": 6.476324370131712e-05, "loss": 0.0214, "step": 34750 }, { "grad_norm": 0.3425474464893341, "learning_rate": 6.474349555824036e-05, "loss": 0.0296, "step": 34760 }, { "grad_norm": 0.27812251448631287, "learning_rate": 6.472374489590342e-05, "loss": 0.0201, "step": 34770 }, { "grad_norm": 0.16748474538326263, "learning_rate": 6.470399171768118e-05, "loss": 0.0357, "step": 34780 }, { "grad_norm": 0.23564282059669495, "learning_rate": 6.468423602694891e-05, "loss": 0.0282, "step": 34790 }, { "grad_norm": 0.20935390889644623, "learning_rate": 6.466447782708232e-05, "loss": 0.0214, "step": 34800 }, { "grad_norm": 0.2752099335193634, "learning_rate": 6.464471712145754e-05, "loss": 0.0225, "step": 34810 }, { "grad_norm": 0.1934427171945572, "learning_rate": 6.462495391345114e-05, "loss": 0.0286, "step": 34820 }, { "grad_norm": 0.2095882147550583, "learning_rate": 6.46051882064401e-05, "loss": 0.0254, "step": 34830 }, { "grad_norm": 0.2468099594116211, "learning_rate": 6.458542000380186e-05, "loss": 0.0215, "step": 34840 }, { "grad_norm": 0.22255463898181915, "learning_rate": 6.456564930891424e-05, "loss": 0.0255, "step": 34850 }, { "grad_norm": 0.28085756301879883, "learning_rate": 6.454587612515555e-05, "loss": 0.033, "step": 34860 }, { "grad_norm": 0.27631911635398865, "learning_rate": 6.452610045590444e-05, "loss": 0.0321, "step": 34870 }, { "grad_norm": 0.2502408027648926, "learning_rate": 6.450632230454005e-05, "loss": 0.0249, "step": 34880 }, { "grad_norm": 0.2849808633327484, "learning_rate": 6.448654167444195e-05, "loss": 0.0302, "step": 34890 }, { "grad_norm": 0.26849132776260376, "learning_rate": 6.446675856899005e-05, "loss": 0.0256, "step": 34900 }, { "grad_norm": 0.19674982130527496, "learning_rate": 6.444697299156481e-05, "loss": 0.0261, "step": 34910 }, { "grad_norm": 0.27977681159973145, "learning_rate": 6.442718494554701e-05, "loss": 0.0263, "step": 34920 }, { "grad_norm": 0.27309679985046387, "learning_rate": 6.440739443431787e-05, "loss": 0.0222, "step": 34930 }, { "grad_norm": 0.20818231999874115, "learning_rate": 6.438760146125906e-05, "loss": 0.0229, "step": 34940 }, { "grad_norm": 0.2725761830806732, "learning_rate": 6.436780602975267e-05, "loss": 0.0336, "step": 34950 }, { "grad_norm": 0.25087010860443115, "learning_rate": 6.43480081431812e-05, "loss": 0.0292, "step": 34960 }, { "grad_norm": 0.26878464221954346, "learning_rate": 6.432820780492756e-05, "loss": 0.0276, "step": 34970 }, { "grad_norm": 0.2351534366607666, "learning_rate": 6.430840501837506e-05, "loss": 0.029, "step": 34980 }, { "grad_norm": 0.23121769726276398, "learning_rate": 6.428859978690748e-05, "loss": 0.0275, "step": 34990 }, { "grad_norm": 0.2261735498905182, "learning_rate": 6.426879211390901e-05, "loss": 0.026, "step": 35000 }, { "grad_norm": 0.21963922679424286, "learning_rate": 6.424898200276422e-05, "loss": 0.0285, "step": 35010 }, { "grad_norm": 0.20957644283771515, "learning_rate": 6.42291694568581e-05, "loss": 0.0252, "step": 35020 }, { "grad_norm": 0.31117990612983704, "learning_rate": 6.42093544795761e-05, "loss": 0.0278, "step": 35030 }, { "grad_norm": 0.2496708184480667, "learning_rate": 6.418953707430403e-05, "loss": 0.0269, "step": 35040 }, { "grad_norm": 0.4689496159553528, "learning_rate": 6.416971724442819e-05, "loss": 0.0226, "step": 35050 }, { "grad_norm": 0.30294373631477356, "learning_rate": 6.414989499333519e-05, "loss": 0.0257, "step": 35060 }, { "grad_norm": 0.16870389878749847, "learning_rate": 6.413007032441214e-05, "loss": 0.0229, "step": 35070 }, { "grad_norm": 0.24581226706504822, "learning_rate": 6.411024324104653e-05, "loss": 0.0302, "step": 35080 }, { "grad_norm": 0.2853797674179077, "learning_rate": 6.409041374662628e-05, "loss": 0.0292, "step": 35090 }, { "grad_norm": 0.22959887981414795, "learning_rate": 6.407058184453967e-05, "loss": 0.0361, "step": 35100 }, { "grad_norm": 0.2273571640253067, "learning_rate": 6.405074753817548e-05, "loss": 0.0241, "step": 35110 }, { "grad_norm": 0.23432090878486633, "learning_rate": 6.40309108309228e-05, "loss": 0.0238, "step": 35120 }, { "grad_norm": 0.25631850957870483, "learning_rate": 6.401107172617122e-05, "loss": 0.026, "step": 35130 }, { "grad_norm": 0.2938585579395294, "learning_rate": 6.399123022731068e-05, "loss": 0.0386, "step": 35140 }, { "grad_norm": 0.3077790439128876, "learning_rate": 6.397138633773157e-05, "loss": 0.0223, "step": 35150 }, { "grad_norm": 0.2930184006690979, "learning_rate": 6.395154006082463e-05, "loss": 0.0233, "step": 35160 }, { "grad_norm": 0.22907418012619019, "learning_rate": 6.393169139998109e-05, "loss": 0.0293, "step": 35170 }, { "grad_norm": 0.29554352164268494, "learning_rate": 6.39118403585925e-05, "loss": 0.0251, "step": 35180 }, { "grad_norm": 0.2434626668691635, "learning_rate": 6.38919869400509e-05, "loss": 0.0236, "step": 35190 }, { "grad_norm": 0.16647930443286896, "learning_rate": 6.387213114774865e-05, "loss": 0.0246, "step": 35200 }, { "grad_norm": 0.23942720890045166, "learning_rate": 6.385227298507863e-05, "loss": 0.0455, "step": 35210 }, { "grad_norm": 0.22620044648647308, "learning_rate": 6.3832412455434e-05, "loss": 0.0241, "step": 35220 }, { "grad_norm": 0.24660877883434296, "learning_rate": 6.381254956220841e-05, "loss": 0.0262, "step": 35230 }, { "grad_norm": 0.20509977638721466, "learning_rate": 6.379268430879586e-05, "loss": 0.0231, "step": 35240 }, { "grad_norm": 0.26580682396888733, "learning_rate": 6.37728166985908e-05, "loss": 0.0241, "step": 35250 }, { "grad_norm": 0.2489284873008728, "learning_rate": 6.375294673498804e-05, "loss": 0.023, "step": 35260 }, { "grad_norm": 0.1899222731590271, "learning_rate": 6.373307442138284e-05, "loss": 0.026, "step": 35270 }, { "grad_norm": 0.2195344865322113, "learning_rate": 6.371319976117081e-05, "loss": 0.0341, "step": 35280 }, { "grad_norm": 0.20540930330753326, "learning_rate": 6.3693322757748e-05, "loss": 0.0233, "step": 35290 }, { "grad_norm": 0.1823911815881729, "learning_rate": 6.367344341451086e-05, "loss": 0.026, "step": 35300 }, { "grad_norm": 0.17474867403507233, "learning_rate": 6.36535617348562e-05, "loss": 0.0266, "step": 35310 }, { "grad_norm": 0.2581818699836731, "learning_rate": 6.363367772218128e-05, "loss": 0.0282, "step": 35320 }, { "grad_norm": 0.23585712909698486, "learning_rate": 6.36137913798837e-05, "loss": 0.0255, "step": 35330 }, { "grad_norm": 0.29646775126457214, "learning_rate": 6.359390271136151e-05, "loss": 0.0217, "step": 35340 }, { "grad_norm": 0.25081613659858704, "learning_rate": 6.357401172001314e-05, "loss": 0.0258, "step": 35350 }, { "grad_norm": 0.186674103140831, "learning_rate": 6.355411840923742e-05, "loss": 0.0258, "step": 35360 }, { "grad_norm": 0.23682951927185059, "learning_rate": 6.353422278243358e-05, "loss": 0.0243, "step": 35370 }, { "grad_norm": 0.2567042410373688, "learning_rate": 6.351432484300121e-05, "loss": 0.0316, "step": 35380 }, { "grad_norm": 0.2098313271999359, "learning_rate": 6.349442459434036e-05, "loss": 0.0306, "step": 35390 }, { "grad_norm": 0.5104613304138184, "learning_rate": 6.34745220398514e-05, "loss": 0.032, "step": 35400 }, { "grad_norm": 0.3096320331096649, "learning_rate": 6.345461718293518e-05, "loss": 0.0274, "step": 35410 }, { "grad_norm": 0.30030757188796997, "learning_rate": 6.343471002699286e-05, "loss": 0.0365, "step": 35420 }, { "grad_norm": 0.30518293380737305, "learning_rate": 6.341480057542602e-05, "loss": 0.0286, "step": 35430 }, { "grad_norm": 0.2612636983394623, "learning_rate": 6.339488883163667e-05, "loss": 0.0279, "step": 35440 }, { "grad_norm": 0.23811431229114532, "learning_rate": 6.337497479902716e-05, "loss": 0.0245, "step": 35450 }, { "grad_norm": 0.25924038887023926, "learning_rate": 6.335505848100027e-05, "loss": 0.0285, "step": 35460 }, { "grad_norm": 0.27272775769233704, "learning_rate": 6.333513988095915e-05, "loss": 0.0236, "step": 35470 }, { "grad_norm": 0.39802882075309753, "learning_rate": 6.331521900230735e-05, "loss": 0.0265, "step": 35480 }, { "grad_norm": 0.2721974849700928, "learning_rate": 6.329529584844878e-05, "loss": 0.0269, "step": 35490 }, { "grad_norm": 0.3574790358543396, "learning_rate": 6.327537042278777e-05, "loss": 0.0239, "step": 35500 }, { "grad_norm": 0.2030261605978012, "learning_rate": 6.325544272872905e-05, "loss": 0.0263, "step": 35510 }, { "grad_norm": 0.2814108431339264, "learning_rate": 6.323551276967771e-05, "loss": 0.0303, "step": 35520 }, { "grad_norm": 0.19462358951568604, "learning_rate": 6.321558054903922e-05, "loss": 0.0215, "step": 35530 }, { "grad_norm": 0.2514280676841736, "learning_rate": 6.319564607021947e-05, "loss": 0.0249, "step": 35540 }, { "grad_norm": 0.20513427257537842, "learning_rate": 6.31757093366247e-05, "loss": 0.0235, "step": 35550 }, { "grad_norm": 0.2679421901702881, "learning_rate": 6.315577035166154e-05, "loss": 0.0249, "step": 35560 }, { "grad_norm": 0.2227233201265335, "learning_rate": 6.313582911873708e-05, "loss": 0.025, "step": 35570 }, { "grad_norm": 0.2207248955965042, "learning_rate": 6.311588564125865e-05, "loss": 0.0233, "step": 35580 }, { "grad_norm": 0.3177667260169983, "learning_rate": 6.30959399226341e-05, "loss": 0.0348, "step": 35590 }, { "grad_norm": 0.26272183656692505, "learning_rate": 6.30759919662716e-05, "loss": 0.0406, "step": 35600 }, { "grad_norm": 0.24949581921100616, "learning_rate": 6.30560417755797e-05, "loss": 0.029, "step": 35610 }, { "grad_norm": 0.2002916932106018, "learning_rate": 6.303608935396735e-05, "loss": 0.0204, "step": 35620 }, { "grad_norm": 0.2313961535692215, "learning_rate": 6.301613470484386e-05, "loss": 0.0268, "step": 35630 }, { "grad_norm": 0.2227611243724823, "learning_rate": 6.299617783161893e-05, "loss": 0.0268, "step": 35640 }, { "grad_norm": 0.22784866392612457, "learning_rate": 6.297621873770266e-05, "loss": 0.0252, "step": 35650 }, { "grad_norm": 0.19393444061279297, "learning_rate": 6.29562574265055e-05, "loss": 0.0227, "step": 35660 }, { "grad_norm": 0.2990390658378601, "learning_rate": 6.293629390143834e-05, "loss": 0.0302, "step": 35670 }, { "grad_norm": 0.2927452325820923, "learning_rate": 6.291632816591232e-05, "loss": 0.027, "step": 35680 }, { "grad_norm": 0.23090875148773193, "learning_rate": 6.28963602233391e-05, "loss": 0.0241, "step": 35690 }, { "grad_norm": 0.29526379704475403, "learning_rate": 6.287639007713062e-05, "loss": 0.0315, "step": 35700 }, { "grad_norm": 0.21612747013568878, "learning_rate": 6.285641773069926e-05, "loss": 0.0259, "step": 35710 }, { "grad_norm": 0.2343745231628418, "learning_rate": 6.283644318745773e-05, "loss": 0.0229, "step": 35720 }, { "grad_norm": 0.3907983601093292, "learning_rate": 6.281646645081912e-05, "loss": 0.0327, "step": 35730 }, { "grad_norm": 0.2370777428150177, "learning_rate": 6.279648752419693e-05, "loss": 0.0271, "step": 35740 }, { "grad_norm": 0.32202744483947754, "learning_rate": 6.2776506411005e-05, "loss": 0.0298, "step": 35750 }, { "grad_norm": 0.3948768675327301, "learning_rate": 6.275652311465758e-05, "loss": 0.0269, "step": 35760 }, { "grad_norm": 0.19663174450397491, "learning_rate": 6.273653763856926e-05, "loss": 0.025, "step": 35770 }, { "grad_norm": 0.22207613289356232, "learning_rate": 6.271654998615501e-05, "loss": 0.0281, "step": 35780 }, { "grad_norm": 0.3117921054363251, "learning_rate": 6.269656016083013e-05, "loss": 0.0214, "step": 35790 }, { "grad_norm": 0.2863466143608093, "learning_rate": 6.267656816601038e-05, "loss": 0.0336, "step": 35800 }, { "grad_norm": 0.23366312682628632, "learning_rate": 6.265657400511185e-05, "loss": 0.0339, "step": 35810 }, { "grad_norm": 0.26900172233581543, "learning_rate": 6.263657768155098e-05, "loss": 0.0309, "step": 35820 }, { "grad_norm": 0.23301012814044952, "learning_rate": 6.261657919874457e-05, "loss": 0.0291, "step": 35830 }, { "grad_norm": 0.2086150199174881, "learning_rate": 6.259657856010986e-05, "loss": 0.0288, "step": 35840 }, { "grad_norm": 0.3320339322090149, "learning_rate": 6.257657576906439e-05, "loss": 0.0254, "step": 35850 }, { "grad_norm": 0.25416040420532227, "learning_rate": 6.255657082902609e-05, "loss": 0.0222, "step": 35860 }, { "grad_norm": 0.2624105215072632, "learning_rate": 6.253656374341325e-05, "loss": 0.0218, "step": 35870 }, { "grad_norm": 0.2979528605937958, "learning_rate": 6.251655451564457e-05, "loss": 0.0289, "step": 35880 }, { "grad_norm": 0.18359129130840302, "learning_rate": 6.249654314913902e-05, "loss": 0.0292, "step": 35890 }, { "grad_norm": 0.3071705996990204, "learning_rate": 6.247652964731604e-05, "loss": 0.0227, "step": 35900 }, { "grad_norm": 0.3079209327697754, "learning_rate": 6.245651401359537e-05, "loss": 0.0374, "step": 35910 }, { "grad_norm": 0.2589524984359741, "learning_rate": 6.243649625139715e-05, "loss": 0.0274, "step": 35920 }, { "grad_norm": 0.3712480068206787, "learning_rate": 6.241647636414185e-05, "loss": 0.0252, "step": 35930 }, { "grad_norm": 0.28703969717025757, "learning_rate": 6.239645435525034e-05, "loss": 0.0321, "step": 35940 }, { "grad_norm": 0.3587968945503235, "learning_rate": 6.237643022814381e-05, "loss": 0.0323, "step": 35950 }, { "grad_norm": 0.23028545081615448, "learning_rate": 6.235640398624386e-05, "loss": 0.0229, "step": 35960 }, { "grad_norm": 0.2999163866043091, "learning_rate": 6.233637563297243e-05, "loss": 0.0323, "step": 35970 }, { "grad_norm": 0.2392408698797226, "learning_rate": 6.23163451717518e-05, "loss": 0.0264, "step": 35980 }, { "grad_norm": 0.2474985271692276, "learning_rate": 6.229631260600463e-05, "loss": 0.026, "step": 35990 }, { "grad_norm": 0.3495500683784485, "learning_rate": 6.227627793915392e-05, "loss": 0.0332, "step": 36000 }, { "grad_norm": 0.24621234834194183, "learning_rate": 6.225624117462309e-05, "loss": 0.0293, "step": 36010 }, { "grad_norm": 0.23286253213882446, "learning_rate": 6.223620231583586e-05, "loss": 0.0249, "step": 36020 }, { "grad_norm": 0.20104673504829407, "learning_rate": 6.221616136621629e-05, "loss": 0.021, "step": 36030 }, { "grad_norm": 0.2970496714115143, "learning_rate": 6.219611832918887e-05, "loss": 0.0312, "step": 36040 }, { "grad_norm": 0.24301572144031525, "learning_rate": 6.217607320817838e-05, "loss": 0.0266, "step": 36050 }, { "grad_norm": 0.27118274569511414, "learning_rate": 6.215602600661001e-05, "loss": 0.0253, "step": 36060 }, { "grad_norm": 0.2674344778060913, "learning_rate": 6.213597672790925e-05, "loss": 0.0257, "step": 36070 }, { "grad_norm": 0.1892387866973877, "learning_rate": 6.2115925375502e-05, "loss": 0.0284, "step": 36080 }, { "grad_norm": 0.18028254806995392, "learning_rate": 6.209587195281447e-05, "loss": 0.0286, "step": 36090 }, { "grad_norm": 0.28980791568756104, "learning_rate": 6.207581646327324e-05, "loss": 0.027, "step": 36100 }, { "grad_norm": 0.2978242337703705, "learning_rate": 6.205575891030526e-05, "loss": 0.0236, "step": 36110 }, { "grad_norm": 0.29353079199790955, "learning_rate": 6.203569929733781e-05, "loss": 0.0241, "step": 36120 }, { "grad_norm": 0.28006213903427124, "learning_rate": 6.201563762779852e-05, "loss": 0.0228, "step": 36130 }, { "grad_norm": 0.21795569360256195, "learning_rate": 6.199557390511538e-05, "loss": 0.0236, "step": 36140 }, { "grad_norm": 0.24119289219379425, "learning_rate": 6.197550813271675e-05, "loss": 0.0253, "step": 36150 }, { "grad_norm": 0.21025963127613068, "learning_rate": 6.195544031403131e-05, "loss": 0.0255, "step": 36160 }, { "grad_norm": 0.2868543565273285, "learning_rate": 6.19353704524881e-05, "loss": 0.0241, "step": 36170 }, { "grad_norm": 0.3322809934616089, "learning_rate": 6.191529855151652e-05, "loss": 0.0252, "step": 36180 }, { "grad_norm": 0.22553835809230804, "learning_rate": 6.189522461454629e-05, "loss": 0.0284, "step": 36190 }, { "grad_norm": 0.22113391757011414, "learning_rate": 6.187514864500752e-05, "loss": 0.0266, "step": 36200 }, { "grad_norm": 0.3071187436580658, "learning_rate": 6.185507064633062e-05, "loss": 0.0276, "step": 36210 }, { "grad_norm": 0.26438412070274353, "learning_rate": 6.18349906219464e-05, "loss": 0.0304, "step": 36220 }, { "grad_norm": 0.282530277967453, "learning_rate": 6.181490857528596e-05, "loss": 0.0266, "step": 36230 }, { "grad_norm": 0.24948444962501526, "learning_rate": 6.179482450978077e-05, "loss": 0.0254, "step": 36240 }, { "grad_norm": 0.2196018546819687, "learning_rate": 6.177473842886269e-05, "loss": 0.0223, "step": 36250 }, { "grad_norm": 0.24730199575424194, "learning_rate": 6.175465033596382e-05, "loss": 0.03, "step": 36260 }, { "grad_norm": 0.2035292237997055, "learning_rate": 6.173456023451671e-05, "loss": 0.0236, "step": 36270 }, { "grad_norm": 0.2999590039253235, "learning_rate": 6.171446812795422e-05, "loss": 0.0309, "step": 36280 }, { "grad_norm": 0.18012158572673798, "learning_rate": 6.169437401970949e-05, "loss": 0.0343, "step": 36290 }, { "grad_norm": 0.24546967446804047, "learning_rate": 6.16742779132161e-05, "loss": 0.0219, "step": 36300 }, { "grad_norm": 0.24674135446548462, "learning_rate": 6.165417981190789e-05, "loss": 0.0261, "step": 36310 }, { "grad_norm": 0.2551463842391968, "learning_rate": 6.16340797192191e-05, "loss": 0.0248, "step": 36320 }, { "grad_norm": 0.24338798224925995, "learning_rate": 6.161397763858427e-05, "loss": 0.0267, "step": 36330 }, { "grad_norm": 0.289508581161499, "learning_rate": 6.159387357343834e-05, "loss": 0.0213, "step": 36340 }, { "grad_norm": 0.2682068645954132, "learning_rate": 6.157376752721648e-05, "loss": 0.0314, "step": 36350 }, { "grad_norm": 0.8310303688049316, "learning_rate": 6.155365950335428e-05, "loss": 0.0296, "step": 36360 }, { "grad_norm": 0.3198423385620117, "learning_rate": 6.153354950528768e-05, "loss": 0.0252, "step": 36370 }, { "grad_norm": 0.2925565242767334, "learning_rate": 6.151343753645293e-05, "loss": 0.0327, "step": 36380 }, { "grad_norm": 0.337454617023468, "learning_rate": 6.149332360028657e-05, "loss": 0.0356, "step": 36390 }, { "grad_norm": 0.26541218161582947, "learning_rate": 6.147320770022555e-05, "loss": 0.034, "step": 36400 }, { "grad_norm": 0.2747735381126404, "learning_rate": 6.145308983970715e-05, "loss": 0.0316, "step": 36410 }, { "grad_norm": 0.2995680868625641, "learning_rate": 6.143297002216892e-05, "loss": 0.0246, "step": 36420 }, { "grad_norm": 0.27500325441360474, "learning_rate": 6.141284825104882e-05, "loss": 0.032, "step": 36430 }, { "grad_norm": 0.2919892370700836, "learning_rate": 6.13927245297851e-05, "loss": 0.0318, "step": 36440 }, { "grad_norm": 0.16369792819023132, "learning_rate": 6.137259886181633e-05, "loss": 0.0289, "step": 36450 }, { "grad_norm": 0.26154616475105286, "learning_rate": 6.135247125058145e-05, "loss": 0.0303, "step": 36460 }, { "grad_norm": 0.20879559218883514, "learning_rate": 6.133234169951974e-05, "loss": 0.0199, "step": 36470 }, { "grad_norm": 0.23277881741523743, "learning_rate": 6.131221021207078e-05, "loss": 0.0269, "step": 36480 }, { "grad_norm": 0.19796344637870789, "learning_rate": 6.129207679167448e-05, "loss": 0.0214, "step": 36490 }, { "grad_norm": 0.17285315692424774, "learning_rate": 6.127194144177109e-05, "loss": 0.0217, "step": 36500 }, { "grad_norm": 0.24937209486961365, "learning_rate": 6.125180416580118e-05, "loss": 0.0231, "step": 36510 }, { "grad_norm": 0.20648224651813507, "learning_rate": 6.123166496720571e-05, "loss": 0.0214, "step": 36520 }, { "grad_norm": 0.2677588164806366, "learning_rate": 6.121152384942588e-05, "loss": 0.0277, "step": 36530 }, { "grad_norm": 0.23449133336544037, "learning_rate": 6.119138081590324e-05, "loss": 0.0265, "step": 36540 }, { "grad_norm": 0.24857290089130402, "learning_rate": 6.117123587007971e-05, "loss": 0.0257, "step": 36550 }, { "grad_norm": 0.2278802990913391, "learning_rate": 6.11510890153975e-05, "loss": 0.0342, "step": 36560 }, { "grad_norm": 0.35255417227745056, "learning_rate": 6.113094025529916e-05, "loss": 0.0231, "step": 36570 }, { "grad_norm": 0.291609525680542, "learning_rate": 6.111078959322757e-05, "loss": 0.0351, "step": 36580 }, { "grad_norm": 0.2168435901403427, "learning_rate": 6.109063703262592e-05, "loss": 0.031, "step": 36590 }, { "grad_norm": 0.24670562148094177, "learning_rate": 6.107048257693772e-05, "loss": 0.0292, "step": 36600 }, { "grad_norm": 0.2762887477874756, "learning_rate": 6.105032622960683e-05, "loss": 0.0259, "step": 36610 }, { "grad_norm": 0.24680261313915253, "learning_rate": 6.103016799407743e-05, "loss": 0.0234, "step": 36620 }, { "grad_norm": 0.34503456950187683, "learning_rate": 6.1010007873793984e-05, "loss": 0.0288, "step": 36630 }, { "grad_norm": 0.20872929692268372, "learning_rate": 6.098984587220131e-05, "loss": 0.0235, "step": 36640 }, { "grad_norm": 0.20679044723510742, "learning_rate": 6.096968199274456e-05, "loss": 0.0261, "step": 36650 }, { "grad_norm": 0.2126832902431488, "learning_rate": 6.0949516238869166e-05, "loss": 0.0242, "step": 36660 }, { "grad_norm": 0.2254353165626526, "learning_rate": 6.092934861402092e-05, "loss": 0.0292, "step": 36670 }, { "grad_norm": 0.2721526026725769, "learning_rate": 6.0909179121645924e-05, "loss": 0.0321, "step": 36680 }, { "grad_norm": 0.36553138494491577, "learning_rate": 6.0889007765190576e-05, "loss": 0.035, "step": 36690 }, { "grad_norm": 0.17393121123313904, "learning_rate": 6.086883454810162e-05, "loss": 0.0212, "step": 36700 }, { "grad_norm": 0.29065436124801636, "learning_rate": 6.0848659473826084e-05, "loss": 0.0282, "step": 36710 }, { "grad_norm": 0.37807127833366394, "learning_rate": 6.082848254581138e-05, "loss": 0.0331, "step": 36720 }, { "grad_norm": 0.3161041736602783, "learning_rate": 6.080830376750517e-05, "loss": 0.0306, "step": 36730 }, { "grad_norm": 0.2161943018436432, "learning_rate": 6.0788123142355445e-05, "loss": 0.0251, "step": 36740 }, { "grad_norm": 0.28869274258613586, "learning_rate": 6.076794067381052e-05, "loss": 0.0243, "step": 36750 }, { "grad_norm": 0.2599569857120514, "learning_rate": 6.074775636531905e-05, "loss": 0.0269, "step": 36760 }, { "grad_norm": 0.2561613917350769, "learning_rate": 6.072757022032997e-05, "loss": 0.0279, "step": 36770 }, { "grad_norm": 0.2294132560491562, "learning_rate": 6.070738224229253e-05, "loss": 0.0304, "step": 36780 }, { "grad_norm": 0.23498868942260742, "learning_rate": 6.0687192434656314e-05, "loss": 0.0291, "step": 36790 }, { "grad_norm": 0.27936798334121704, "learning_rate": 6.066700080087121e-05, "loss": 0.0266, "step": 36800 }, { "grad_norm": 0.29224881529808044, "learning_rate": 6.0646807344387424e-05, "loss": 0.0275, "step": 36810 }, { "grad_norm": 0.2966535687446594, "learning_rate": 6.062661206865543e-05, "loss": 0.033, "step": 36820 }, { "grad_norm": 0.27454379200935364, "learning_rate": 6.06064149771261e-05, "loss": 0.0248, "step": 36830 }, { "grad_norm": 0.2387627214193344, "learning_rate": 6.058621607325051e-05, "loss": 0.0286, "step": 36840 }, { "grad_norm": 0.2810632586479187, "learning_rate": 6.056601536048014e-05, "loss": 0.0296, "step": 36850 }, { "grad_norm": 0.24581685662269592, "learning_rate": 6.0545812842266725e-05, "loss": 0.0378, "step": 36860 }, { "grad_norm": 0.2629590928554535, "learning_rate": 6.052560852206232e-05, "loss": 0.0258, "step": 36870 }, { "grad_norm": 0.2984349727630615, "learning_rate": 6.05054024033193e-05, "loss": 0.035, "step": 36880 }, { "grad_norm": 0.22739581763744354, "learning_rate": 6.048519448949032e-05, "loss": 0.0245, "step": 36890 }, { "grad_norm": 0.3346940279006958, "learning_rate": 6.046498478402839e-05, "loss": 0.0307, "step": 36900 }, { "grad_norm": 0.23381365835666656, "learning_rate": 6.044477329038677e-05, "loss": 0.0287, "step": 36910 }, { "grad_norm": 0.3267211318016052, "learning_rate": 6.042456001201906e-05, "loss": 0.0248, "step": 36920 }, { "grad_norm": 0.1951674073934555, "learning_rate": 6.040434495237917e-05, "loss": 0.023, "step": 36930 }, { "grad_norm": 0.273380845785141, "learning_rate": 6.0384128114921256e-05, "loss": 0.0261, "step": 36940 }, { "grad_norm": 0.285327285528183, "learning_rate": 6.036390950309987e-05, "loss": 0.0207, "step": 36950 }, { "grad_norm": 0.2220388650894165, "learning_rate": 6.0343689120369805e-05, "loss": 0.0281, "step": 36960 }, { "grad_norm": 0.287211149930954, "learning_rate": 6.032346697018616e-05, "loss": 0.0233, "step": 36970 }, { "grad_norm": 0.2461145967245102, "learning_rate": 6.0303243056004375e-05, "loss": 0.0311, "step": 36980 }, { "grad_norm": 0.22101692855358124, "learning_rate": 6.0283017381280136e-05, "loss": 0.0256, "step": 36990 }, { "grad_norm": 0.26850461959838867, "learning_rate": 6.026278994946948e-05, "loss": 0.0248, "step": 37000 }, { "grad_norm": 0.19902414083480835, "learning_rate": 6.02425607640287e-05, "loss": 0.0266, "step": 37010 }, { "grad_norm": 0.26303815841674805, "learning_rate": 6.022232982841441e-05, "loss": 0.03, "step": 37020 }, { "grad_norm": 0.29237937927246094, "learning_rate": 6.020209714608355e-05, "loss": 0.0253, "step": 37030 }, { "grad_norm": 0.26491355895996094, "learning_rate": 6.018186272049331e-05, "loss": 0.0242, "step": 37040 }, { "grad_norm": 0.20827797055244446, "learning_rate": 6.01616265551012e-05, "loss": 0.0296, "step": 37050 }, { "grad_norm": 0.3468197286128998, "learning_rate": 6.014138865336503e-05, "loss": 0.0265, "step": 37060 }, { "grad_norm": 0.28154635429382324, "learning_rate": 6.0121149018742905e-05, "loss": 0.0271, "step": 37070 }, { "grad_norm": 0.19185052812099457, "learning_rate": 6.010090765469325e-05, "loss": 0.024, "step": 37080 }, { "grad_norm": 0.2096242755651474, "learning_rate": 6.008066456467473e-05, "loss": 0.0251, "step": 37090 }, { "grad_norm": 0.20966292917728424, "learning_rate": 6.0060419752146335e-05, "loss": 0.0263, "step": 37100 }, { "grad_norm": 0.33286193013191223, "learning_rate": 6.0040173220567353e-05, "loss": 0.0242, "step": 37110 }, { "grad_norm": 0.23320958018302917, "learning_rate": 6.001992497339737e-05, "loss": 0.0236, "step": 37120 }, { "grad_norm": 0.3026627004146576, "learning_rate": 5.999967501409626e-05, "loss": 0.0306, "step": 37130 }, { "grad_norm": 0.2911577820777893, "learning_rate": 5.997942334612418e-05, "loss": 0.0223, "step": 37140 }, { "grad_norm": 0.15792936086654663, "learning_rate": 5.995916997294158e-05, "loss": 0.0189, "step": 37150 }, { "grad_norm": 0.2209194302558899, "learning_rate": 5.9938914898009214e-05, "loss": 0.025, "step": 37160 }, { "grad_norm": 0.22814129292964935, "learning_rate": 5.991865812478813e-05, "loss": 0.0216, "step": 37170 }, { "grad_norm": 0.20208679139614105, "learning_rate": 5.989839965673964e-05, "loss": 0.036, "step": 37180 }, { "grad_norm": 0.1665835976600647, "learning_rate": 5.987813949732539e-05, "loss": 0.0238, "step": 37190 }, { "grad_norm": 0.24992786347866058, "learning_rate": 5.9857877650007255e-05, "loss": 0.0218, "step": 37200 }, { "grad_norm": 0.19046738743782043, "learning_rate": 5.983761411824744e-05, "loss": 0.0202, "step": 37210 }, { "grad_norm": 0.1909829080104828, "learning_rate": 5.981734890550844e-05, "loss": 0.0258, "step": 37220 }, { "grad_norm": 0.18373402953147888, "learning_rate": 5.979708201525301e-05, "loss": 0.0254, "step": 37230 }, { "grad_norm": 0.2588195502758026, "learning_rate": 5.977681345094422e-05, "loss": 0.0283, "step": 37240 }, { "grad_norm": 0.294393926858902, "learning_rate": 5.97565432160454e-05, "loss": 0.0248, "step": 37250 }, { "grad_norm": 0.33662137389183044, "learning_rate": 5.9736271314020186e-05, "loss": 0.0237, "step": 37260 }, { "grad_norm": 0.2716183662414551, "learning_rate": 5.971599774833251e-05, "loss": 0.0279, "step": 37270 }, { "grad_norm": 0.30783385038375854, "learning_rate": 5.9695722522446525e-05, "loss": 0.0286, "step": 37280 }, { "grad_norm": 0.2801205515861511, "learning_rate": 5.9675445639826765e-05, "loss": 0.0271, "step": 37290 }, { "grad_norm": 0.3959498703479767, "learning_rate": 5.965516710393796e-05, "loss": 0.028, "step": 37300 }, { "grad_norm": 0.34853455424308777, "learning_rate": 5.963488691824516e-05, "loss": 0.0346, "step": 37310 }, { "grad_norm": 0.22419773042201996, "learning_rate": 5.96146050862137e-05, "loss": 0.0276, "step": 37320 }, { "grad_norm": 0.2447245717048645, "learning_rate": 5.959432161130919e-05, "loss": 0.0227, "step": 37330 }, { "grad_norm": 0.28463974595069885, "learning_rate": 5.9574036496997545e-05, "loss": 0.0289, "step": 37340 }, { "grad_norm": 0.28091397881507874, "learning_rate": 5.955374974674488e-05, "loss": 0.0261, "step": 37350 }, { "grad_norm": 0.2769912779331207, "learning_rate": 5.9533461364017696e-05, "loss": 0.0278, "step": 37360 }, { "grad_norm": 0.2058529555797577, "learning_rate": 5.9513171352282716e-05, "loss": 0.0248, "step": 37370 }, { "grad_norm": 0.24107196927070618, "learning_rate": 5.949287971500692e-05, "loss": 0.0254, "step": 37380 }, { "grad_norm": 0.19707131385803223, "learning_rate": 5.947258645565762e-05, "loss": 0.0223, "step": 37390 }, { "grad_norm": 0.17302973568439484, "learning_rate": 5.945229157770237e-05, "loss": 0.0233, "step": 37400 }, { "grad_norm": 0.19644863903522491, "learning_rate": 5.9431995084609006e-05, "loss": 0.0279, "step": 37410 }, { "grad_norm": 0.40652671456336975, "learning_rate": 5.941169697984564e-05, "loss": 0.0249, "step": 37420 }, { "grad_norm": 0.32618603110313416, "learning_rate": 5.9391397266880675e-05, "loss": 0.0244, "step": 37430 }, { "grad_norm": 0.30509164929389954, "learning_rate": 5.937109594918279e-05, "loss": 0.039, "step": 37440 }, { "grad_norm": 0.34967902302742004, "learning_rate": 5.9350793030220884e-05, "loss": 0.0334, "step": 37450 }, { "grad_norm": 0.33965209126472473, "learning_rate": 5.933048851346421e-05, "loss": 0.0257, "step": 37460 }, { "grad_norm": 0.2145891785621643, "learning_rate": 5.931018240238224e-05, "loss": 0.0363, "step": 37470 }, { "grad_norm": 0.3994004428386688, "learning_rate": 5.928987470044471e-05, "loss": 0.0245, "step": 37480 }, { "grad_norm": 0.33639904856681824, "learning_rate": 5.9269565411121695e-05, "loss": 0.0287, "step": 37490 }, { "grad_norm": 0.21750710904598236, "learning_rate": 5.924925453788347e-05, "loss": 0.0285, "step": 37500 }, { "grad_norm": 0.24408693611621857, "learning_rate": 5.92289420842006e-05, "loss": 0.0326, "step": 37510 }, { "grad_norm": 0.3536801040172577, "learning_rate": 5.9208628053543945e-05, "loss": 0.031, "step": 37520 }, { "grad_norm": 0.2585051357746124, "learning_rate": 5.918831244938462e-05, "loss": 0.0243, "step": 37530 }, { "grad_norm": 0.24074730277061462, "learning_rate": 5.9167995275194e-05, "loss": 0.0268, "step": 37540 }, { "grad_norm": 0.24223379790782928, "learning_rate": 5.914767653444373e-05, "loss": 0.0276, "step": 37550 }, { "grad_norm": 0.4097133278846741, "learning_rate": 5.912735623060572e-05, "loss": 0.0306, "step": 37560 }, { "grad_norm": 0.26197922229766846, "learning_rate": 5.910703436715217e-05, "loss": 0.0239, "step": 37570 }, { "grad_norm": 0.31591102480888367, "learning_rate": 5.908671094755552e-05, "loss": 0.0241, "step": 37580 }, { "grad_norm": 0.24386779963970184, "learning_rate": 5.906638597528851e-05, "loss": 0.0223, "step": 37590 }, { "grad_norm": 0.3455274701118469, "learning_rate": 5.9046059453824076e-05, "loss": 0.0276, "step": 37600 }, { "grad_norm": 0.40863561630249023, "learning_rate": 5.9025731386635505e-05, "loss": 0.0284, "step": 37610 }, { "grad_norm": 0.28069570660591125, "learning_rate": 5.900540177719629e-05, "loss": 0.0221, "step": 37620 }, { "grad_norm": 0.2811638116836548, "learning_rate": 5.898507062898021e-05, "loss": 0.0261, "step": 37630 }, { "grad_norm": 0.3134680986404419, "learning_rate": 5.8964737945461316e-05, "loss": 0.0275, "step": 37640 }, { "grad_norm": 0.24165934324264526, "learning_rate": 5.8944403730113885e-05, "loss": 0.0263, "step": 37650 }, { "grad_norm": 0.1930476725101471, "learning_rate": 5.892406798641248e-05, "loss": 0.031, "step": 37660 }, { "grad_norm": 0.23580235242843628, "learning_rate": 5.890373071783193e-05, "loss": 0.0239, "step": 37670 }, { "grad_norm": 0.27601706981658936, "learning_rate": 5.888339192784732e-05, "loss": 0.027, "step": 37680 }, { "grad_norm": 0.21410106122493744, "learning_rate": 5.8863051619934003e-05, "loss": 0.0247, "step": 37690 }, { "grad_norm": 0.3021269738674164, "learning_rate": 5.8842709797567554e-05, "loss": 0.0247, "step": 37700 }, { "grad_norm": 0.25688210129737854, "learning_rate": 5.8822366464223855e-05, "loss": 0.0208, "step": 37710 }, { "grad_norm": 0.19968973100185394, "learning_rate": 5.880202162337901e-05, "loss": 0.0273, "step": 37720 }, { "grad_norm": 0.20486503839492798, "learning_rate": 5.8781675278509405e-05, "loss": 0.0202, "step": 37730 }, { "grad_norm": 0.23150351643562317, "learning_rate": 5.8761327433091696e-05, "loss": 0.0228, "step": 37740 }, { "grad_norm": 0.21898673474788666, "learning_rate": 5.874097809060275e-05, "loss": 0.0341, "step": 37750 }, { "grad_norm": 0.1678084135055542, "learning_rate": 5.87206272545197e-05, "loss": 0.035, "step": 37760 }, { "grad_norm": 0.24827612936496735, "learning_rate": 5.8700274928319955e-05, "loss": 0.0275, "step": 37770 }, { "grad_norm": 0.35732313990592957, "learning_rate": 5.867992111548118e-05, "loss": 0.0243, "step": 37780 }, { "grad_norm": 0.21991689503192902, "learning_rate": 5.865956581948131e-05, "loss": 0.0216, "step": 37790 }, { "grad_norm": 0.21083562076091766, "learning_rate": 5.863920904379845e-05, "loss": 0.0214, "step": 37800 }, { "grad_norm": 0.21619641780853271, "learning_rate": 5.8618850791911064e-05, "loss": 0.0277, "step": 37810 }, { "grad_norm": 0.30013108253479004, "learning_rate": 5.859849106729779e-05, "loss": 0.0285, "step": 37820 }, { "grad_norm": 0.2553364038467407, "learning_rate": 5.857812987343758e-05, "loss": 0.0281, "step": 37830 }, { "grad_norm": 0.2965609133243561, "learning_rate": 5.855776721380957e-05, "loss": 0.0289, "step": 37840 }, { "grad_norm": 0.2187345176935196, "learning_rate": 5.8537403091893217e-05, "loss": 0.0247, "step": 37850 }, { "grad_norm": 0.3050580620765686, "learning_rate": 5.851703751116816e-05, "loss": 0.0282, "step": 37860 }, { "grad_norm": 0.24664899706840515, "learning_rate": 5.8496670475114336e-05, "loss": 0.0276, "step": 37870 }, { "grad_norm": 0.2379629909992218, "learning_rate": 5.84763019872119e-05, "loss": 0.0322, "step": 37880 }, { "grad_norm": 0.4044206440448761, "learning_rate": 5.845593205094131e-05, "loss": 0.0252, "step": 37890 }, { "grad_norm": 0.205423504114151, "learning_rate": 5.843556066978318e-05, "loss": 0.0243, "step": 37900 }, { "grad_norm": 0.2882479429244995, "learning_rate": 5.8415187847218455e-05, "loss": 0.0265, "step": 37910 }, { "grad_norm": 0.29219338297843933, "learning_rate": 5.839481358672827e-05, "loss": 0.0256, "step": 37920 }, { "grad_norm": 0.24220095574855804, "learning_rate": 5.837443789179407e-05, "loss": 0.029, "step": 37930 }, { "grad_norm": 0.23246735334396362, "learning_rate": 5.8354060765897445e-05, "loss": 0.0251, "step": 37940 }, { "grad_norm": 0.3465777039527893, "learning_rate": 5.8333682212520334e-05, "loss": 0.0342, "step": 37950 }, { "grad_norm": 0.23701393604278564, "learning_rate": 5.831330223514486e-05, "loss": 0.0262, "step": 37960 }, { "grad_norm": 0.2459220290184021, "learning_rate": 5.8292920837253396e-05, "loss": 0.0275, "step": 37970 }, { "grad_norm": 0.1875702142715454, "learning_rate": 5.827253802232857e-05, "loss": 0.0208, "step": 37980 }, { "grad_norm": 0.2727009356021881, "learning_rate": 5.825215379385327e-05, "loss": 0.0215, "step": 37990 }, { "grad_norm": 0.2717352509498596, "learning_rate": 5.823176815531057e-05, "loss": 0.0391, "step": 38000 }, { "grad_norm": 0.5404653549194336, "learning_rate": 5.8211381110183826e-05, "loss": 0.0354, "step": 38010 }, { "grad_norm": 0.2710607647895813, "learning_rate": 5.8190992661956645e-05, "loss": 0.0214, "step": 38020 }, { "grad_norm": 0.2040564864873886, "learning_rate": 5.817060281411284e-05, "loss": 0.0273, "step": 38030 }, { "grad_norm": 0.2022283524274826, "learning_rate": 5.815021157013647e-05, "loss": 0.0266, "step": 38040 }, { "grad_norm": 0.24055823683738708, "learning_rate": 5.8129818933511856e-05, "loss": 0.0355, "step": 38050 }, { "grad_norm": 0.25589054822921753, "learning_rate": 5.8109424907723544e-05, "loss": 0.0261, "step": 38060 }, { "grad_norm": 0.20582814514636993, "learning_rate": 5.80890294962563e-05, "loss": 0.0212, "step": 38070 }, { "grad_norm": 0.25702425837516785, "learning_rate": 5.806863270259515e-05, "loss": 0.0335, "step": 38080 }, { "grad_norm": 0.30961477756500244, "learning_rate": 5.804823453022536e-05, "loss": 0.0221, "step": 38090 }, { "grad_norm": 0.25166431069374084, "learning_rate": 5.80278349826324e-05, "loss": 0.0296, "step": 38100 }, { "grad_norm": 0.17402836680412292, "learning_rate": 5.8007434063302014e-05, "loss": 0.0226, "step": 38110 }, { "grad_norm": 0.278627872467041, "learning_rate": 5.7987031775720136e-05, "loss": 0.0231, "step": 38120 }, { "grad_norm": 0.22945967316627502, "learning_rate": 5.7966628123372976e-05, "loss": 0.0347, "step": 38130 }, { "grad_norm": 0.36203253269195557, "learning_rate": 5.7946223109746956e-05, "loss": 0.0347, "step": 38140 }, { "grad_norm": 0.2757031321525574, "learning_rate": 5.7925816738328754e-05, "loss": 0.0238, "step": 38150 }, { "grad_norm": 0.22357068955898285, "learning_rate": 5.790540901260521e-05, "loss": 0.026, "step": 38160 }, { "grad_norm": 0.2299269586801529, "learning_rate": 5.788499993606351e-05, "loss": 0.0276, "step": 38170 }, { "grad_norm": 0.24930621683597565, "learning_rate": 5.786458951219096e-05, "loss": 0.0274, "step": 38180 }, { "grad_norm": 0.28045207262039185, "learning_rate": 5.784417774447517e-05, "loss": 0.0335, "step": 38190 }, { "grad_norm": 0.1481834053993225, "learning_rate": 5.782376463640393e-05, "loss": 0.026, "step": 38200 }, { "grad_norm": 0.34216564893722534, "learning_rate": 5.780335019146531e-05, "loss": 0.0249, "step": 38210 }, { "grad_norm": 0.24524465203285217, "learning_rate": 5.778293441314755e-05, "loss": 0.0251, "step": 38220 }, { "grad_norm": 0.2812635898590088, "learning_rate": 5.776251730493917e-05, "loss": 0.0306, "step": 38230 }, { "grad_norm": 0.1893427073955536, "learning_rate": 5.774209887032887e-05, "loss": 0.0234, "step": 38240 }, { "grad_norm": 0.23714952170848846, "learning_rate": 5.772167911280565e-05, "loss": 0.0216, "step": 38250 }, { "grad_norm": 0.2114265263080597, "learning_rate": 5.770125803585864e-05, "loss": 0.0219, "step": 38260 }, { "grad_norm": 0.2587665319442749, "learning_rate": 5.768083564297726e-05, "loss": 0.0326, "step": 38270 }, { "grad_norm": 0.3069012463092804, "learning_rate": 5.766041193765114e-05, "loss": 0.0269, "step": 38280 }, { "grad_norm": 0.1975715458393097, "learning_rate": 5.763998692337015e-05, "loss": 0.024, "step": 38290 }, { "grad_norm": 0.26444128155708313, "learning_rate": 5.761956060362433e-05, "loss": 0.0285, "step": 38300 }, { "grad_norm": 0.2848256826400757, "learning_rate": 5.7599132981904e-05, "loss": 0.026, "step": 38310 }, { "grad_norm": 0.29662275314331055, "learning_rate": 5.75787040616997e-05, "loss": 0.0339, "step": 38320 }, { "grad_norm": 0.26083099842071533, "learning_rate": 5.755827384650212e-05, "loss": 0.0244, "step": 38330 }, { "grad_norm": 0.3128877580165863, "learning_rate": 5.753784233980228e-05, "loss": 0.0293, "step": 38340 }, { "grad_norm": 0.24329310655593872, "learning_rate": 5.751740954509135e-05, "loss": 0.0248, "step": 38350 }, { "grad_norm": 0.24539531767368317, "learning_rate": 5.7496975465860715e-05, "loss": 0.024, "step": 38360 }, { "grad_norm": 0.2078266739845276, "learning_rate": 5.747654010560202e-05, "loss": 0.0256, "step": 38370 }, { "grad_norm": 0.17832940816879272, "learning_rate": 5.7456103467807097e-05, "loss": 0.0248, "step": 38380 }, { "grad_norm": 0.24483147263526917, "learning_rate": 5.7435665555968046e-05, "loss": 0.0237, "step": 38390 }, { "grad_norm": 0.18743732571601868, "learning_rate": 5.74152263735771e-05, "loss": 0.0276, "step": 38400 }, { "grad_norm": 0.285200297832489, "learning_rate": 5.739478592412677e-05, "loss": 0.0245, "step": 38410 }, { "grad_norm": 0.350635290145874, "learning_rate": 5.7374344211109766e-05, "loss": 0.0334, "step": 38420 }, { "grad_norm": 0.3023850917816162, "learning_rate": 5.735390123801904e-05, "loss": 0.0248, "step": 38430 }, { "grad_norm": 0.27255672216415405, "learning_rate": 5.7333457008347704e-05, "loss": 0.0287, "step": 38440 }, { "grad_norm": 0.359133243560791, "learning_rate": 5.7313011525589156e-05, "loss": 0.0261, "step": 38450 }, { "grad_norm": 0.20853176712989807, "learning_rate": 5.729256479323694e-05, "loss": 0.0248, "step": 38460 }, { "grad_norm": 0.22426725924015045, "learning_rate": 5.727211681478485e-05, "loss": 0.0243, "step": 38470 }, { "grad_norm": 0.18348416686058044, "learning_rate": 5.7251667593726886e-05, "loss": 0.029, "step": 38480 }, { "grad_norm": 0.20792293548583984, "learning_rate": 5.723121713355728e-05, "loss": 0.0393, "step": 38490 }, { "grad_norm": 0.21178627014160156, "learning_rate": 5.721076543777044e-05, "loss": 0.0278, "step": 38500 }, { "grad_norm": 0.3034716546535492, "learning_rate": 5.7190312509860986e-05, "loss": 0.0231, "step": 38510 }, { "grad_norm": 0.29441893100738525, "learning_rate": 5.716985835332379e-05, "loss": 0.0249, "step": 38520 }, { "grad_norm": 0.22531729936599731, "learning_rate": 5.714940297165389e-05, "loss": 0.0231, "step": 38530 }, { "grad_norm": 0.22180940210819244, "learning_rate": 5.712894636834656e-05, "loss": 0.0268, "step": 38540 }, { "grad_norm": 0.28493744134902954, "learning_rate": 5.7108488546897276e-05, "loss": 0.0227, "step": 38550 }, { "grad_norm": 0.18935197591781616, "learning_rate": 5.708802951080172e-05, "loss": 0.0246, "step": 38560 }, { "grad_norm": 0.2564079165458679, "learning_rate": 5.706756926355576e-05, "loss": 0.0255, "step": 38570 }, { "grad_norm": 0.28899601101875305, "learning_rate": 5.704710780865554e-05, "loss": 0.0251, "step": 38580 }, { "grad_norm": 0.2903411388397217, "learning_rate": 5.7026645149597325e-05, "loss": 0.0237, "step": 38590 }, { "grad_norm": 0.2885664105415344, "learning_rate": 5.700618128987764e-05, "loss": 0.0294, "step": 38600 }, { "grad_norm": 0.17725886404514313, "learning_rate": 5.698571623299317e-05, "loss": 0.0246, "step": 38610 }, { "grad_norm": 0.2832558751106262, "learning_rate": 5.696524998244086e-05, "loss": 0.0199, "step": 38620 }, { "grad_norm": 0.21860577166080475, "learning_rate": 5.6944782541717836e-05, "loss": 0.0319, "step": 38630 }, { "grad_norm": 0.36871543526649475, "learning_rate": 5.69243139143214e-05, "loss": 0.0272, "step": 38640 }, { "grad_norm": 0.2365344762802124, "learning_rate": 5.6903844103749125e-05, "loss": 0.031, "step": 38650 }, { "grad_norm": 0.22056803107261658, "learning_rate": 5.688337311349869e-05, "loss": 0.0201, "step": 38660 }, { "grad_norm": 0.30115067958831787, "learning_rate": 5.6862900947068074e-05, "loss": 0.0266, "step": 38670 }, { "grad_norm": 0.3192759156227112, "learning_rate": 5.6842427607955374e-05, "loss": 0.0317, "step": 38680 }, { "grad_norm": 0.2912736237049103, "learning_rate": 5.682195309965893e-05, "loss": 0.0226, "step": 38690 }, { "grad_norm": 0.25301483273506165, "learning_rate": 5.6801477425677294e-05, "loss": 0.0303, "step": 38700 }, { "grad_norm": 0.27415111660957336, "learning_rate": 5.678100058950917e-05, "loss": 0.0261, "step": 38710 }, { "grad_norm": 0.28773659467697144, "learning_rate": 5.676052259465352e-05, "loss": 0.0273, "step": 38720 }, { "grad_norm": 0.25864264369010925, "learning_rate": 5.674004344460945e-05, "loss": 0.0225, "step": 38730 }, { "grad_norm": 0.24028092622756958, "learning_rate": 5.6719563142876295e-05, "loss": 0.0206, "step": 38740 }, { "grad_norm": 0.24019482731819153, "learning_rate": 5.669908169295359e-05, "loss": 0.0233, "step": 38750 }, { "grad_norm": 0.24787291884422302, "learning_rate": 5.667859909834105e-05, "loss": 0.0258, "step": 38760 }, { "grad_norm": 0.2579908072948456, "learning_rate": 5.6658115362538565e-05, "loss": 0.0254, "step": 38770 }, { "grad_norm": 0.26861169934272766, "learning_rate": 5.663763048904628e-05, "loss": 0.0242, "step": 38780 }, { "grad_norm": 0.32774555683135986, "learning_rate": 5.661714448136447e-05, "loss": 0.0209, "step": 38790 }, { "grad_norm": 0.2509014904499054, "learning_rate": 5.659665734299366e-05, "loss": 0.0247, "step": 38800 }, { "grad_norm": 0.2129548043012619, "learning_rate": 5.6576169077434516e-05, "loss": 0.028, "step": 38810 }, { "grad_norm": 0.236337810754776, "learning_rate": 5.6555679688187944e-05, "loss": 0.0253, "step": 38820 }, { "grad_norm": 0.2933162450790405, "learning_rate": 5.6535189178755e-05, "loss": 0.0257, "step": 38830 }, { "grad_norm": 0.3403928875923157, "learning_rate": 5.651469755263698e-05, "loss": 0.0248, "step": 38840 }, { "grad_norm": 0.21837425231933594, "learning_rate": 5.6494204813335316e-05, "loss": 0.0338, "step": 38850 }, { "grad_norm": 0.2513546347618103, "learning_rate": 5.647371096435168e-05, "loss": 0.0232, "step": 38860 }, { "grad_norm": 0.26377809047698975, "learning_rate": 5.645321600918788e-05, "loss": 0.0212, "step": 38870 }, { "grad_norm": 0.27928876876831055, "learning_rate": 5.643271995134597e-05, "loss": 0.0303, "step": 38880 }, { "grad_norm": 0.258841872215271, "learning_rate": 5.641222279432814e-05, "loss": 0.0247, "step": 38890 }, { "grad_norm": 0.29736119508743286, "learning_rate": 5.6391724541636834e-05, "loss": 0.0242, "step": 38900 }, { "grad_norm": 0.2541307806968689, "learning_rate": 5.6371225196774605e-05, "loss": 0.0264, "step": 38910 }, { "grad_norm": 0.3107912242412567, "learning_rate": 5.635072476324423e-05, "loss": 0.0287, "step": 38920 }, { "grad_norm": 0.2594354748725891, "learning_rate": 5.63302232445487e-05, "loss": 0.0251, "step": 38930 }, { "grad_norm": 0.29162949323654175, "learning_rate": 5.6309720644191144e-05, "loss": 0.0306, "step": 38940 }, { "grad_norm": 0.25314557552337646, "learning_rate": 5.628921696567491e-05, "loss": 0.0216, "step": 38950 }, { "grad_norm": 0.23253272473812103, "learning_rate": 5.62687122125035e-05, "loss": 0.0244, "step": 38960 }, { "grad_norm": 0.26060932874679565, "learning_rate": 5.624820638818062e-05, "loss": 0.0231, "step": 38970 }, { "grad_norm": 0.1898089200258255, "learning_rate": 5.6227699496210164e-05, "loss": 0.0322, "step": 38980 }, { "grad_norm": 0.34272703528404236, "learning_rate": 5.6207191540096195e-05, "loss": 0.028, "step": 38990 }, { "grad_norm": 0.31161463260650635, "learning_rate": 5.618668252334296e-05, "loss": 0.0217, "step": 39000 }, { "grad_norm": 0.24728882312774658, "learning_rate": 5.616617244945488e-05, "loss": 0.0254, "step": 39010 }, { "grad_norm": 0.20778857171535492, "learning_rate": 5.614566132193656e-05, "loss": 0.025, "step": 39020 }, { "grad_norm": 0.2330426424741745, "learning_rate": 5.612514914429282e-05, "loss": 0.0274, "step": 39030 }, { "grad_norm": 0.2953805923461914, "learning_rate": 5.610463592002863e-05, "loss": 0.0257, "step": 39040 }, { "grad_norm": 0.31242021918296814, "learning_rate": 5.608412165264909e-05, "loss": 0.0326, "step": 39050 }, { "grad_norm": 0.24090301990509033, "learning_rate": 5.606360634565959e-05, "loss": 0.0233, "step": 39060 }, { "grad_norm": 0.2542940676212311, "learning_rate": 5.604309000256559e-05, "loss": 0.025, "step": 39070 }, { "grad_norm": 0.2645052969455719, "learning_rate": 5.6022572626872785e-05, "loss": 0.0239, "step": 39080 }, { "grad_norm": 0.453640341758728, "learning_rate": 5.600205422208704e-05, "loss": 0.0271, "step": 39090 }, { "grad_norm": 0.2364870309829712, "learning_rate": 5.59815347917144e-05, "loss": 0.0277, "step": 39100 }, { "grad_norm": 0.1881992220878601, "learning_rate": 5.596101433926103e-05, "loss": 0.0257, "step": 39110 }, { "grad_norm": 0.2562711238861084, "learning_rate": 5.5940492868233364e-05, "loss": 0.0213, "step": 39120 }, { "grad_norm": 0.25085195899009705, "learning_rate": 5.591997038213793e-05, "loss": 0.0292, "step": 39130 }, { "grad_norm": 0.21778686344623566, "learning_rate": 5.5899446884481475e-05, "loss": 0.0256, "step": 39140 }, { "grad_norm": 0.1661652773618698, "learning_rate": 5.5878922378770906e-05, "loss": 0.0388, "step": 39150 }, { "grad_norm": 0.22349712252616882, "learning_rate": 5.5858396868513285e-05, "loss": 0.0267, "step": 39160 }, { "grad_norm": 0.21105988323688507, "learning_rate": 5.583787035721586e-05, "loss": 0.0214, "step": 39170 }, { "grad_norm": 0.3074043095111847, "learning_rate": 5.581734284838606e-05, "loss": 0.0231, "step": 39180 }, { "grad_norm": 0.27386951446533203, "learning_rate": 5.579681434553147e-05, "loss": 0.0245, "step": 39190 }, { "grad_norm": 0.23832857608795166, "learning_rate": 5.5776284852159854e-05, "loss": 0.0244, "step": 39200 }, { "grad_norm": 0.27450406551361084, "learning_rate": 5.575575437177913e-05, "loss": 0.034, "step": 39210 }, { "grad_norm": 0.34599176049232483, "learning_rate": 5.573522290789742e-05, "loss": 0.0222, "step": 39220 }, { "grad_norm": 0.18469642102718353, "learning_rate": 5.571469046402298e-05, "loss": 0.0249, "step": 39230 }, { "grad_norm": 0.2167750895023346, "learning_rate": 5.5694157043664205e-05, "loss": 0.0302, "step": 39240 }, { "grad_norm": 0.2130608856678009, "learning_rate": 5.567362265032975e-05, "loss": 0.0196, "step": 39250 }, { "grad_norm": 0.2525631785392761, "learning_rate": 5.565308728752836e-05, "loss": 0.0249, "step": 39260 }, { "grad_norm": 0.26810434460639954, "learning_rate": 5.5632550958768956e-05, "loss": 0.022, "step": 39270 }, { "grad_norm": 0.33919456601142883, "learning_rate": 5.5612013667560636e-05, "loss": 0.0273, "step": 39280 }, { "grad_norm": 0.27551308274269104, "learning_rate": 5.5591475417412676e-05, "loss": 0.042, "step": 39290 }, { "grad_norm": 0.22822032868862152, "learning_rate": 5.557093621183451e-05, "loss": 0.0246, "step": 39300 }, { "grad_norm": 0.25242266058921814, "learning_rate": 5.55503960543357e-05, "loss": 0.024, "step": 39310 }, { "grad_norm": 0.2513446807861328, "learning_rate": 5.552985494842601e-05, "loss": 0.0235, "step": 39320 }, { "grad_norm": 0.20795810222625732, "learning_rate": 5.550931289761534e-05, "loss": 0.023, "step": 39330 }, { "grad_norm": 0.276505708694458, "learning_rate": 5.548876990541378e-05, "loss": 0.0236, "step": 39340 }, { "grad_norm": 0.30963268876075745, "learning_rate": 5.5468225975331564e-05, "loss": 0.0308, "step": 39350 }, { "grad_norm": 0.3476678729057312, "learning_rate": 5.544768111087909e-05, "loss": 0.0266, "step": 39360 }, { "grad_norm": 0.18614216148853302, "learning_rate": 5.54271353155669e-05, "loss": 0.026, "step": 39370 }, { "grad_norm": 0.2940744161605835, "learning_rate": 5.5406588592905715e-05, "loss": 0.0259, "step": 39380 }, { "grad_norm": 0.23748256266117096, "learning_rate": 5.5386040946406416e-05, "loss": 0.0273, "step": 39390 }, { "grad_norm": 0.34455689787864685, "learning_rate": 5.536549237958004e-05, "loss": 0.0342, "step": 39400 }, { "grad_norm": 0.22228804230690002, "learning_rate": 5.5344942895937744e-05, "loss": 0.0359, "step": 39410 }, { "grad_norm": 0.20389333367347717, "learning_rate": 5.5324392498990904e-05, "loss": 0.0204, "step": 39420 }, { "grad_norm": 0.2800128161907196, "learning_rate": 5.5303841192251e-05, "loss": 0.024, "step": 39430 }, { "grad_norm": 0.23090164363384247, "learning_rate": 5.52832889792297e-05, "loss": 0.0241, "step": 39440 }, { "grad_norm": 0.22593693435192108, "learning_rate": 5.526273586343881e-05, "loss": 0.0243, "step": 39450 }, { "grad_norm": 0.21425281465053558, "learning_rate": 5.5242181848390306e-05, "loss": 0.0215, "step": 39460 }, { "grad_norm": 0.3500121533870697, "learning_rate": 5.5221626937596285e-05, "loss": 0.02, "step": 39470 }, { "grad_norm": 0.26942184567451477, "learning_rate": 5.520107113456903e-05, "loss": 0.0261, "step": 39480 }, { "grad_norm": 0.2874290347099304, "learning_rate": 5.5180514442820974e-05, "loss": 0.0351, "step": 39490 }, { "grad_norm": 0.26770591735839844, "learning_rate": 5.515995686586469e-05, "loss": 0.02, "step": 39500 }, { "grad_norm": 0.24819223582744598, "learning_rate": 5.5139398407212916e-05, "loss": 0.0202, "step": 39510 }, { "grad_norm": 0.25103864073753357, "learning_rate": 5.511883907037849e-05, "loss": 0.0226, "step": 39520 }, { "grad_norm": 0.22805754840373993, "learning_rate": 5.509827885887449e-05, "loss": 0.0236, "step": 39530 }, { "grad_norm": 0.2515983581542969, "learning_rate": 5.507771777621406e-05, "loss": 0.0243, "step": 39540 }, { "grad_norm": 0.16867198050022125, "learning_rate": 5.505715582591052e-05, "loss": 0.0332, "step": 39550 }, { "grad_norm": 0.2685464024543762, "learning_rate": 5.50365930114774e-05, "loss": 0.0298, "step": 39560 }, { "grad_norm": 0.20890894532203674, "learning_rate": 5.5016029336428255e-05, "loss": 0.0182, "step": 39570 }, { "grad_norm": 0.22015880048274994, "learning_rate": 5.49954648042769e-05, "loss": 0.026, "step": 39580 }, { "grad_norm": 0.3321133255958557, "learning_rate": 5.4974899418537226e-05, "loss": 0.0267, "step": 39590 }, { "grad_norm": 0.1604824960231781, "learning_rate": 5.4954333182723316e-05, "loss": 0.0225, "step": 39600 }, { "grad_norm": 0.27409547567367554, "learning_rate": 5.493376610034937e-05, "loss": 0.0201, "step": 39610 }, { "grad_norm": 0.25996074080467224, "learning_rate": 5.4913198174929735e-05, "loss": 0.0271, "step": 39620 }, { "grad_norm": 0.20150816440582275, "learning_rate": 5.489262940997889e-05, "loss": 0.0211, "step": 39630 }, { "grad_norm": 0.23119677603244781, "learning_rate": 5.487205980901151e-05, "loss": 0.0342, "step": 39640 }, { "grad_norm": 0.2041826993227005, "learning_rate": 5.485148937554234e-05, "loss": 0.0385, "step": 39650 }, { "grad_norm": 0.256046324968338, "learning_rate": 5.483091811308635e-05, "loss": 0.0261, "step": 39660 }, { "grad_norm": 0.2140437513589859, "learning_rate": 5.4810346025158564e-05, "loss": 0.0232, "step": 39670 }, { "grad_norm": 0.2202618271112442, "learning_rate": 5.478977311527421e-05, "loss": 0.0226, "step": 39680 }, { "grad_norm": 0.2275109440088272, "learning_rate": 5.476919938694863e-05, "loss": 0.021, "step": 39690 }, { "grad_norm": 0.38144928216934204, "learning_rate": 5.474862484369733e-05, "loss": 0.0194, "step": 39700 }, { "grad_norm": 0.2287474125623703, "learning_rate": 5.472804948903589e-05, "loss": 0.0228, "step": 39710 }, { "grad_norm": 0.46554532647132874, "learning_rate": 5.470747332648013e-05, "loss": 0.0207, "step": 39720 }, { "grad_norm": 0.18529100716114044, "learning_rate": 5.468689635954591e-05, "loss": 0.0299, "step": 39730 }, { "grad_norm": 0.2866687774658203, "learning_rate": 5.46663185917493e-05, "loss": 0.0235, "step": 39740 }, { "grad_norm": 0.2671871483325958, "learning_rate": 5.464574002660645e-05, "loss": 0.0249, "step": 39750 }, { "grad_norm": 0.19197769463062286, "learning_rate": 5.4625160667633724e-05, "loss": 0.0227, "step": 39760 }, { "grad_norm": 0.17819435894489288, "learning_rate": 5.4604580518347505e-05, "loss": 0.0261, "step": 39770 }, { "grad_norm": 0.2678981423377991, "learning_rate": 5.458399958226442e-05, "loss": 0.0272, "step": 39780 }, { "grad_norm": 0.3014848828315735, "learning_rate": 5.456341786290119e-05, "loss": 0.0223, "step": 39790 }, { "grad_norm": 0.25826895236968994, "learning_rate": 5.454283536377465e-05, "loss": 0.0308, "step": 39800 }, { "grad_norm": 0.24155747890472412, "learning_rate": 5.452225208840179e-05, "loss": 0.0228, "step": 39810 }, { "grad_norm": 0.171408012509346, "learning_rate": 5.450166804029975e-05, "loss": 0.0283, "step": 39820 }, { "grad_norm": 0.19623930752277374, "learning_rate": 5.448108322298574e-05, "loss": 0.023, "step": 39830 }, { "grad_norm": 0.271111398935318, "learning_rate": 5.446049763997717e-05, "loss": 0.026, "step": 39840 }, { "grad_norm": 0.3008683919906616, "learning_rate": 5.4439911294791546e-05, "loss": 0.0237, "step": 39850 }, { "grad_norm": 0.23338215053081512, "learning_rate": 5.441932419094652e-05, "loss": 0.0225, "step": 39860 }, { "grad_norm": 0.2245514839887619, "learning_rate": 5.439873633195985e-05, "loss": 0.0247, "step": 39870 }, { "grad_norm": 0.18562155961990356, "learning_rate": 5.437814772134947e-05, "loss": 0.0296, "step": 39880 }, { "grad_norm": 0.4548243582248688, "learning_rate": 5.4357558362633366e-05, "loss": 0.0295, "step": 39890 }, { "grad_norm": 0.27032116055488586, "learning_rate": 5.4336968259329726e-05, "loss": 0.0205, "step": 39900 }, { "grad_norm": 0.22262446582317352, "learning_rate": 5.431637741495681e-05, "loss": 0.0261, "step": 39910 }, { "grad_norm": 0.21297134459018707, "learning_rate": 5.429578583303307e-05, "loss": 0.0313, "step": 39920 }, { "grad_norm": 0.1850089132785797, "learning_rate": 5.427519351707701e-05, "loss": 0.0283, "step": 39930 }, { "grad_norm": 0.2970845103263855, "learning_rate": 5.42546004706073e-05, "loss": 0.0283, "step": 39940 }, { "grad_norm": 0.21687312424182892, "learning_rate": 5.4234006697142735e-05, "loss": 0.0225, "step": 39950 }, { "grad_norm": 0.2152254283428192, "learning_rate": 5.421341220020224e-05, "loss": 0.0239, "step": 39960 }, { "grad_norm": 0.28521764278411865, "learning_rate": 5.419281698330482e-05, "loss": 0.0293, "step": 39970 }, { "grad_norm": 0.20144140720367432, "learning_rate": 5.4172221049969665e-05, "loss": 0.0314, "step": 39980 }, { "grad_norm": 0.2877312898635864, "learning_rate": 5.415162440371604e-05, "loss": 0.0254, "step": 39990 }, { "grad_norm": 0.24613256752490997, "learning_rate": 5.413102704806334e-05, "loss": 0.0278, "step": 40000 }, { "grad_norm": 0.25900787115097046, "learning_rate": 5.41104289865311e-05, "loss": 0.0252, "step": 40010 }, { "grad_norm": 0.19441764056682587, "learning_rate": 5.408983022263898e-05, "loss": 0.0245, "step": 40020 }, { "grad_norm": 0.26822805404663086, "learning_rate": 5.406923075990673e-05, "loss": 0.0306, "step": 40030 }, { "grad_norm": 0.2935137450695038, "learning_rate": 5.404863060185423e-05, "loss": 0.0236, "step": 40040 }, { "grad_norm": 0.22429987788200378, "learning_rate": 5.40280297520015e-05, "loss": 0.0261, "step": 40050 }, { "grad_norm": 0.1899399310350418, "learning_rate": 5.400742821386865e-05, "loss": 0.024, "step": 40060 }, { "grad_norm": 0.23177774250507355, "learning_rate": 5.398682599097592e-05, "loss": 0.0271, "step": 40070 }, { "grad_norm": 0.23581717908382416, "learning_rate": 5.396622308684367e-05, "loss": 0.0231, "step": 40080 }, { "grad_norm": 0.3230368196964264, "learning_rate": 5.394561950499236e-05, "loss": 0.0274, "step": 40090 }, { "grad_norm": 0.2236519604921341, "learning_rate": 5.39250152489426e-05, "loss": 0.0242, "step": 40100 }, { "grad_norm": 0.18903698027133942, "learning_rate": 5.390441032221507e-05, "loss": 0.0248, "step": 40110 }, { "grad_norm": 0.22496892511844635, "learning_rate": 5.388380472833062e-05, "loss": 0.0247, "step": 40120 }, { "grad_norm": 0.22693045437335968, "learning_rate": 5.386319847081016e-05, "loss": 0.0265, "step": 40130 }, { "grad_norm": 0.26595011353492737, "learning_rate": 5.384259155317473e-05, "loss": 0.0293, "step": 40140 }, { "grad_norm": 0.34321263432502747, "learning_rate": 5.38219839789455e-05, "loss": 0.0318, "step": 40150 }, { "grad_norm": 0.21399952471256256, "learning_rate": 5.380137575164376e-05, "loss": 0.0295, "step": 40160 }, { "grad_norm": 0.2705830931663513, "learning_rate": 5.378076687479085e-05, "loss": 0.0241, "step": 40170 }, { "grad_norm": 0.3470790982246399, "learning_rate": 5.3760157351908305e-05, "loss": 0.0255, "step": 40180 }, { "grad_norm": 0.2900254726409912, "learning_rate": 5.373954718651768e-05, "loss": 0.0294, "step": 40190 }, { "grad_norm": 0.23908279836177826, "learning_rate": 5.371893638214074e-05, "loss": 0.0214, "step": 40200 }, { "grad_norm": 0.29819807410240173, "learning_rate": 5.369832494229927e-05, "loss": 0.0262, "step": 40210 }, { "grad_norm": 0.3336276113986969, "learning_rate": 5.367771287051524e-05, "loss": 0.0294, "step": 40220 }, { "grad_norm": 0.17678117752075195, "learning_rate": 5.365710017031065e-05, "loss": 0.0207, "step": 40230 }, { "grad_norm": 0.1870393306016922, "learning_rate": 5.3636486845207654e-05, "loss": 0.0248, "step": 40240 }, { "grad_norm": 0.3161413371562958, "learning_rate": 5.3615872898728526e-05, "loss": 0.0218, "step": 40250 }, { "grad_norm": 0.36678093671798706, "learning_rate": 5.3595258334395614e-05, "loss": 0.0295, "step": 40260 }, { "grad_norm": 0.2705114185810089, "learning_rate": 5.3574643155731396e-05, "loss": 0.0331, "step": 40270 }, { "grad_norm": 0.1918153315782547, "learning_rate": 5.3554027366258405e-05, "loss": 0.023, "step": 40280 }, { "grad_norm": 0.1637725532054901, "learning_rate": 5.3533410969499355e-05, "loss": 0.0264, "step": 40290 }, { "grad_norm": 0.19226031005382538, "learning_rate": 5.3512793968977e-05, "loss": 0.0248, "step": 40300 }, { "grad_norm": 0.26109328866004944, "learning_rate": 5.349217636821423e-05, "loss": 0.0254, "step": 40310 }, { "grad_norm": 0.18753786385059357, "learning_rate": 5.347155817073405e-05, "loss": 0.0232, "step": 40320 }, { "grad_norm": 0.2935178577899933, "learning_rate": 5.34509393800595e-05, "loss": 0.0235, "step": 40330 }, { "grad_norm": 0.28534361720085144, "learning_rate": 5.343031999971381e-05, "loss": 0.0212, "step": 40340 }, { "grad_norm": 0.1602344959974289, "learning_rate": 5.340970003322026e-05, "loss": 0.0234, "step": 40350 }, { "grad_norm": 0.2749179005622864, "learning_rate": 5.33890794841022e-05, "loss": 0.0208, "step": 40360 }, { "grad_norm": 0.20267318189144135, "learning_rate": 5.336845835588318e-05, "loss": 0.0249, "step": 40370 }, { "grad_norm": 0.22313019633293152, "learning_rate": 5.334783665208674e-05, "loss": 0.0254, "step": 40380 }, { "grad_norm": 0.23484568297863007, "learning_rate": 5.332721437623657e-05, "loss": 0.0241, "step": 40390 }, { "grad_norm": 0.23934023082256317, "learning_rate": 5.3306591531856464e-05, "loss": 0.0244, "step": 40400 }, { "grad_norm": 0.27539291977882385, "learning_rate": 5.3285968122470295e-05, "loss": 0.0225, "step": 40410 }, { "grad_norm": 0.2275850921869278, "learning_rate": 5.326534415160207e-05, "loss": 0.0303, "step": 40420 }, { "grad_norm": 0.20086950063705444, "learning_rate": 5.324471962277582e-05, "loss": 0.027, "step": 40430 }, { "grad_norm": 0.25418218970298767, "learning_rate": 5.3224094539515746e-05, "loss": 0.0265, "step": 40440 }, { "grad_norm": 0.2545555830001831, "learning_rate": 5.3203468905346076e-05, "loss": 0.0227, "step": 40450 }, { "grad_norm": 0.30543142557144165, "learning_rate": 5.3182842723791195e-05, "loss": 0.0246, "step": 40460 }, { "grad_norm": 0.2186293751001358, "learning_rate": 5.316221599837554e-05, "loss": 0.0227, "step": 40470 }, { "grad_norm": 0.28532639145851135, "learning_rate": 5.314158873262366e-05, "loss": 0.0293, "step": 40480 }, { "grad_norm": 0.19769826531410217, "learning_rate": 5.312096093006018e-05, "loss": 0.0248, "step": 40490 }, { "grad_norm": 0.1957734078168869, "learning_rate": 5.3100332594209846e-05, "loss": 0.0211, "step": 40500 }, { "grad_norm": 0.2282785326242447, "learning_rate": 5.3079703728597454e-05, "loss": 0.0255, "step": 40510 }, { "grad_norm": 0.18593361973762512, "learning_rate": 5.305907433674794e-05, "loss": 0.0246, "step": 40520 }, { "grad_norm": 0.23133523762226105, "learning_rate": 5.3038444422186284e-05, "loss": 0.0227, "step": 40530 }, { "grad_norm": 0.17258211970329285, "learning_rate": 5.301781398843757e-05, "loss": 0.0207, "step": 40540 }, { "grad_norm": 0.19520094990730286, "learning_rate": 5.299718303902699e-05, "loss": 0.0239, "step": 40550 }, { "grad_norm": 0.2691743075847626, "learning_rate": 5.297655157747979e-05, "loss": 0.0274, "step": 40560 }, { "grad_norm": 0.3157854676246643, "learning_rate": 5.295591960732136e-05, "loss": 0.026, "step": 40570 }, { "grad_norm": 0.2869786322116852, "learning_rate": 5.293528713207708e-05, "loss": 0.0244, "step": 40580 }, { "grad_norm": 0.2352631390094757, "learning_rate": 5.291465415527253e-05, "loss": 0.0244, "step": 40590 }, { "grad_norm": 0.3073740601539612, "learning_rate": 5.2894020680433296e-05, "loss": 0.026, "step": 40600 }, { "grad_norm": 0.24010121822357178, "learning_rate": 5.287338671108507e-05, "loss": 0.025, "step": 40610 }, { "grad_norm": 0.287369042634964, "learning_rate": 5.285275225075367e-05, "loss": 0.0221, "step": 40620 }, { "grad_norm": 0.22520549595355988, "learning_rate": 5.283211730296492e-05, "loss": 0.0217, "step": 40630 }, { "grad_norm": 0.29262790083885193, "learning_rate": 5.281148187124477e-05, "loss": 0.024, "step": 40640 }, { "grad_norm": 0.26948851346969604, "learning_rate": 5.279084595911927e-05, "loss": 0.0238, "step": 40650 }, { "grad_norm": 0.23663976788520813, "learning_rate": 5.277020957011453e-05, "loss": 0.0324, "step": 40660 }, { "grad_norm": 0.24255585670471191, "learning_rate": 5.274957270775673e-05, "loss": 0.0327, "step": 40670 }, { "grad_norm": 0.27877089381217957, "learning_rate": 5.2728935375572164e-05, "loss": 0.0223, "step": 40680 }, { "grad_norm": 0.22005175054073334, "learning_rate": 5.2708297577087176e-05, "loss": 0.0253, "step": 40690 }, { "grad_norm": 0.2905212342739105, "learning_rate": 5.2687659315828196e-05, "loss": 0.0218, "step": 40700 }, { "grad_norm": 0.27459362149238586, "learning_rate": 5.266702059532175e-05, "loss": 0.0219, "step": 40710 }, { "grad_norm": 0.24669578671455383, "learning_rate": 5.264638141909444e-05, "loss": 0.0292, "step": 40720 }, { "grad_norm": 0.2700371742248535, "learning_rate": 5.2625741790672925e-05, "loss": 0.0286, "step": 40730 }, { "grad_norm": 0.2829788327217102, "learning_rate": 5.2605101713583935e-05, "loss": 0.0242, "step": 40740 }, { "grad_norm": 0.18504415452480316, "learning_rate": 5.2584461191354315e-05, "loss": 0.0299, "step": 40750 }, { "grad_norm": 0.1688944697380066, "learning_rate": 5.2563820227510964e-05, "loss": 0.0254, "step": 40760 }, { "grad_norm": 0.19852100312709808, "learning_rate": 5.2543178825580864e-05, "loss": 0.023, "step": 40770 }, { "grad_norm": 0.3413665294647217, "learning_rate": 5.2522536989091054e-05, "loss": 0.0317, "step": 40780 }, { "grad_norm": 0.2179325819015503, "learning_rate": 5.2501894721568655e-05, "loss": 0.0271, "step": 40790 }, { "grad_norm": 0.2275945544242859, "learning_rate": 5.248125202654089e-05, "loss": 0.0223, "step": 40800 }, { "grad_norm": 0.17266494035720825, "learning_rate": 5.246060890753501e-05, "loss": 0.0312, "step": 40810 }, { "grad_norm": 0.2697032690048218, "learning_rate": 5.243996536807837e-05, "loss": 0.023, "step": 40820 }, { "grad_norm": 0.18522831797599792, "learning_rate": 5.2419321411698384e-05, "loss": 0.0217, "step": 40830 }, { "grad_norm": 0.2834755778312683, "learning_rate": 5.239867704192253e-05, "loss": 0.0278, "step": 40840 }, { "grad_norm": 0.2708193063735962, "learning_rate": 5.237803226227838e-05, "loss": 0.0196, "step": 40850 }, { "grad_norm": 0.22052296996116638, "learning_rate": 5.235738707629354e-05, "loss": 0.0223, "step": 40860 }, { "grad_norm": 0.23618583381175995, "learning_rate": 5.233674148749575e-05, "loss": 0.0226, "step": 40870 }, { "grad_norm": 0.20443327724933624, "learning_rate": 5.231609549941272e-05, "loss": 0.0244, "step": 40880 }, { "grad_norm": 0.17910805344581604, "learning_rate": 5.229544911557233e-05, "loss": 0.0255, "step": 40890 }, { "grad_norm": 0.1578531265258789, "learning_rate": 5.227480233950246e-05, "loss": 0.0242, "step": 40900 }, { "grad_norm": 0.2896192967891693, "learning_rate": 5.22541551747311e-05, "loss": 0.0279, "step": 40910 }, { "grad_norm": 0.3072662949562073, "learning_rate": 5.223350762478626e-05, "loss": 0.0219, "step": 40920 }, { "grad_norm": 0.1599823385477066, "learning_rate": 5.221285969319606e-05, "loss": 0.023, "step": 40930 }, { "grad_norm": 0.2508962154388428, "learning_rate": 5.219221138348865e-05, "loss": 0.022, "step": 40940 }, { "grad_norm": 0.2161540538072586, "learning_rate": 5.217156269919228e-05, "loss": 0.0241, "step": 40950 }, { "grad_norm": 0.19615201652050018, "learning_rate": 5.215091364383523e-05, "loss": 0.0234, "step": 40960 }, { "grad_norm": 0.25581076741218567, "learning_rate": 5.213026422094588e-05, "loss": 0.0242, "step": 40970 }, { "grad_norm": 0.21687184274196625, "learning_rate": 5.210961443405262e-05, "loss": 0.0204, "step": 40980 }, { "grad_norm": 0.23311872780323029, "learning_rate": 5.208896428668396e-05, "loss": 0.0201, "step": 40990 }, { "grad_norm": 0.27675074338912964, "learning_rate": 5.206831378236845e-05, "loss": 0.0238, "step": 41000 }, { "grad_norm": 0.1924545168876648, "learning_rate": 5.2047662924634666e-05, "loss": 0.0181, "step": 41010 }, { "grad_norm": 0.23500795662403107, "learning_rate": 5.20270117170113e-05, "loss": 0.0201, "step": 41020 }, { "grad_norm": 0.28940364718437195, "learning_rate": 5.200636016302707e-05, "loss": 0.0322, "step": 41030 }, { "grad_norm": 0.2201693207025528, "learning_rate": 5.1985708266210754e-05, "loss": 0.0271, "step": 41040 }, { "grad_norm": 0.22622378170490265, "learning_rate": 5.1965056030091206e-05, "loss": 0.0264, "step": 41050 }, { "grad_norm": 0.2646385431289673, "learning_rate": 5.194440345819732e-05, "loss": 0.0222, "step": 41060 }, { "grad_norm": 0.33418944478034973, "learning_rate": 5.1923750554058084e-05, "loss": 0.0274, "step": 41070 }, { "grad_norm": 0.12691137194633484, "learning_rate": 5.1903097321202476e-05, "loss": 0.0201, "step": 41080 }, { "grad_norm": 0.3028745651245117, "learning_rate": 5.18824437631596e-05, "loss": 0.0329, "step": 41090 }, { "grad_norm": 0.17888243496418, "learning_rate": 5.186178988345856e-05, "loss": 0.0242, "step": 41100 }, { "grad_norm": 0.2473394125699997, "learning_rate": 5.184113568562855e-05, "loss": 0.0262, "step": 41110 }, { "grad_norm": 0.323782354593277, "learning_rate": 5.18204811731988e-05, "loss": 0.0322, "step": 41120 }, { "grad_norm": 0.3253421485424042, "learning_rate": 5.179982634969862e-05, "loss": 0.025, "step": 41130 }, { "grad_norm": 0.38791102170944214, "learning_rate": 5.177917121865734e-05, "loss": 0.0288, "step": 41140 }, { "grad_norm": 0.24707946181297302, "learning_rate": 5.1758515783604346e-05, "loss": 0.0231, "step": 41150 }, { "grad_norm": 0.3787594139575958, "learning_rate": 5.173786004806911e-05, "loss": 0.0326, "step": 41160 }, { "grad_norm": 0.3177487850189209, "learning_rate": 5.1717204015581135e-05, "loss": 0.0227, "step": 41170 }, { "grad_norm": 0.19160154461860657, "learning_rate": 5.169654768966994e-05, "loss": 0.0228, "step": 41180 }, { "grad_norm": 0.2599118649959564, "learning_rate": 5.1675891073865156e-05, "loss": 0.0249, "step": 41190 }, { "grad_norm": 0.14552515745162964, "learning_rate": 5.1655234171696424e-05, "loss": 0.0227, "step": 41200 }, { "grad_norm": 0.21989025175571442, "learning_rate": 5.163457698669343e-05, "loss": 0.0244, "step": 41210 }, { "grad_norm": 0.23998881876468658, "learning_rate": 5.1613919522385936e-05, "loss": 0.0212, "step": 41220 }, { "grad_norm": 0.2878166735172272, "learning_rate": 5.1593261782303746e-05, "loss": 0.0322, "step": 41230 }, { "grad_norm": 0.2690696716308594, "learning_rate": 5.157260376997669e-05, "loss": 0.0241, "step": 41240 }, { "grad_norm": 0.20320898294448853, "learning_rate": 5.155194548893466e-05, "loss": 0.028, "step": 41250 }, { "grad_norm": 0.20345894992351532, "learning_rate": 5.153128694270759e-05, "loss": 0.0329, "step": 41260 }, { "grad_norm": 0.25555431842803955, "learning_rate": 5.151062813482548e-05, "loss": 0.0314, "step": 41270 }, { "grad_norm": 0.19753490388393402, "learning_rate": 5.148996906881832e-05, "loss": 0.0249, "step": 41280 }, { "grad_norm": 0.33647194504737854, "learning_rate": 5.1469309748216196e-05, "loss": 0.0222, "step": 41290 }, { "grad_norm": 0.22821944952011108, "learning_rate": 5.144865017654923e-05, "loss": 0.0208, "step": 41300 }, { "grad_norm": 0.22631652653217316, "learning_rate": 5.1427990357347556e-05, "loss": 0.0253, "step": 41310 }, { "grad_norm": 0.26092731952667236, "learning_rate": 5.140733029414139e-05, "loss": 0.0252, "step": 41320 }, { "grad_norm": 0.2524973154067993, "learning_rate": 5.138666999046098e-05, "loss": 0.0223, "step": 41330 }, { "grad_norm": 0.24356868863105774, "learning_rate": 5.136600944983658e-05, "loss": 0.0197, "step": 41340 }, { "grad_norm": 0.2722688913345337, "learning_rate": 5.134534867579853e-05, "loss": 0.0271, "step": 41350 }, { "grad_norm": 0.3325786888599396, "learning_rate": 5.1324687671877184e-05, "loss": 0.0219, "step": 41360 }, { "grad_norm": 0.20042070746421814, "learning_rate": 5.130402644160296e-05, "loss": 0.0267, "step": 41370 }, { "grad_norm": 0.2096908837556839, "learning_rate": 5.128336498850628e-05, "loss": 0.0258, "step": 41380 }, { "grad_norm": 0.17200136184692383, "learning_rate": 5.126270331611761e-05, "loss": 0.0206, "step": 41390 }, { "grad_norm": 0.2671923339366913, "learning_rate": 5.124204142796748e-05, "loss": 0.0216, "step": 41400 }, { "grad_norm": 0.18494871258735657, "learning_rate": 5.122137932758644e-05, "loss": 0.0218, "step": 41410 }, { "grad_norm": 0.22376835346221924, "learning_rate": 5.1200717018505074e-05, "loss": 0.0253, "step": 41420 }, { "grad_norm": 0.15439699590206146, "learning_rate": 5.118005450425403e-05, "loss": 0.0236, "step": 41430 }, { "grad_norm": 0.9512053728103638, "learning_rate": 5.115939178836392e-05, "loss": 0.0187, "step": 41440 }, { "grad_norm": 0.2987869381904602, "learning_rate": 5.113872887436547e-05, "loss": 0.028, "step": 41450 }, { "grad_norm": 0.25677359104156494, "learning_rate": 5.11180657657894e-05, "loss": 0.0289, "step": 41460 }, { "grad_norm": 0.2963982820510864, "learning_rate": 5.109740246616649e-05, "loss": 0.0255, "step": 41470 }, { "grad_norm": 0.3468051552772522, "learning_rate": 5.10767389790275e-05, "loss": 0.0353, "step": 41480 }, { "grad_norm": 0.2787207365036011, "learning_rate": 5.1056075307903265e-05, "loss": 0.0265, "step": 41490 }, { "grad_norm": 0.17729617655277252, "learning_rate": 5.1035411456324634e-05, "loss": 0.0243, "step": 41500 }, { "grad_norm": 1.0061630010604858, "learning_rate": 5.1014747427822504e-05, "loss": 0.0277, "step": 41510 }, { "grad_norm": 0.29045963287353516, "learning_rate": 5.0994083225927804e-05, "loss": 0.0321, "step": 41520 }, { "grad_norm": 0.3046354055404663, "learning_rate": 5.0973418854171475e-05, "loss": 0.0337, "step": 41530 }, { "grad_norm": 0.16879796981811523, "learning_rate": 5.095275431608447e-05, "loss": 0.0278, "step": 41540 }, { "grad_norm": 0.2663620412349701, "learning_rate": 5.093208961519782e-05, "loss": 0.0256, "step": 41550 }, { "grad_norm": 0.26641467213630676, "learning_rate": 5.091142475504255e-05, "loss": 0.0243, "step": 41560 }, { "grad_norm": 0.3304576873779297, "learning_rate": 5.089075973914971e-05, "loss": 0.0305, "step": 41570 }, { "grad_norm": 0.28271210193634033, "learning_rate": 5.08700945710504e-05, "loss": 0.0329, "step": 41580 }, { "grad_norm": 0.26036903262138367, "learning_rate": 5.0849429254275714e-05, "loss": 0.0251, "step": 41590 }, { "grad_norm": 0.19538885354995728, "learning_rate": 5.0828763792356804e-05, "loss": 0.0213, "step": 41600 }, { "grad_norm": 0.32398444414138794, "learning_rate": 5.080809818882483e-05, "loss": 0.0293, "step": 41610 }, { "grad_norm": 0.25097569823265076, "learning_rate": 5.078743244721097e-05, "loss": 0.0256, "step": 41620 }, { "grad_norm": 0.28496477007865906, "learning_rate": 5.0766766571046455e-05, "loss": 0.0277, "step": 41630 }, { "grad_norm": 0.18741117417812347, "learning_rate": 5.07461005638625e-05, "loss": 0.0183, "step": 41640 }, { "grad_norm": 0.23184341192245483, "learning_rate": 5.072543442919037e-05, "loss": 0.0254, "step": 41650 }, { "grad_norm": 0.22388134896755219, "learning_rate": 5.070476817056132e-05, "loss": 0.0306, "step": 41660 }, { "grad_norm": 0.2137264907360077, "learning_rate": 5.068410179150668e-05, "loss": 0.0226, "step": 41670 }, { "grad_norm": 0.21273426711559296, "learning_rate": 5.066343529555775e-05, "loss": 0.0203, "step": 41680 }, { "grad_norm": 0.21612797677516937, "learning_rate": 5.064276868624587e-05, "loss": 0.0209, "step": 41690 }, { "grad_norm": 0.2283157855272293, "learning_rate": 5.0622101967102396e-05, "loss": 0.0229, "step": 41700 }, { "grad_norm": 0.2089608758687973, "learning_rate": 5.0601435141658714e-05, "loss": 0.0216, "step": 41710 }, { "grad_norm": 0.2309696078300476, "learning_rate": 5.058076821344622e-05, "loss": 0.0234, "step": 41720 }, { "grad_norm": 0.2596016824245453, "learning_rate": 5.056010118599632e-05, "loss": 0.0287, "step": 41730 }, { "grad_norm": 0.21844547986984253, "learning_rate": 5.053943406284044e-05, "loss": 0.0206, "step": 41740 }, { "grad_norm": 0.2241986095905304, "learning_rate": 5.051876684751006e-05, "loss": 0.0256, "step": 41750 }, { "grad_norm": 0.19167399406433105, "learning_rate": 5.0498099543536584e-05, "loss": 0.0195, "step": 41760 }, { "grad_norm": 0.20332562923431396, "learning_rate": 5.047743215445152e-05, "loss": 0.0245, "step": 41770 }, { "grad_norm": 0.3046394884586334, "learning_rate": 5.045676468378637e-05, "loss": 0.0229, "step": 41780 }, { "grad_norm": 0.21707914769649506, "learning_rate": 5.0436097135072626e-05, "loss": 0.0213, "step": 41790 }, { "grad_norm": 0.1769774854183197, "learning_rate": 5.041542951184181e-05, "loss": 0.0212, "step": 41800 }, { "grad_norm": 0.19181685149669647, "learning_rate": 5.039476181762545e-05, "loss": 0.0206, "step": 41810 }, { "grad_norm": 0.18710918724536896, "learning_rate": 5.037409405595508e-05, "loss": 0.0168, "step": 41820 }, { "grad_norm": 0.2416795790195465, "learning_rate": 5.035342623036229e-05, "loss": 0.026, "step": 41830 }, { "grad_norm": 0.25062644481658936, "learning_rate": 5.033275834437862e-05, "loss": 0.0224, "step": 41840 }, { "grad_norm": 0.27491632103919983, "learning_rate": 5.031209040153564e-05, "loss": 0.0217, "step": 41850 }, { "grad_norm": 0.23825182020664215, "learning_rate": 5.0291422405364955e-05, "loss": 0.0273, "step": 41860 }, { "grad_norm": 0.1788502186536789, "learning_rate": 5.0270754359398133e-05, "loss": 0.0198, "step": 41870 }, { "grad_norm": 0.23026974499225616, "learning_rate": 5.025008626716682e-05, "loss": 0.0187, "step": 41880 }, { "grad_norm": 0.2406621277332306, "learning_rate": 5.0229418132202585e-05, "loss": 0.0212, "step": 41890 }, { "grad_norm": 0.32566046714782715, "learning_rate": 5.020874995803707e-05, "loss": 0.0253, "step": 41900 }, { "grad_norm": 0.2253076285123825, "learning_rate": 5.01880817482019e-05, "loss": 0.0263, "step": 41910 }, { "grad_norm": 0.32332566380500793, "learning_rate": 5.01674135062287e-05, "loss": 0.0216, "step": 41920 }, { "grad_norm": 0.2559194266796112, "learning_rate": 5.0146745235649115e-05, "loss": 0.0267, "step": 41930 }, { "grad_norm": 0.22603723406791687, "learning_rate": 5.0126076939994795e-05, "loss": 0.0208, "step": 41940 }, { "grad_norm": 0.28321075439453125, "learning_rate": 5.010540862279736e-05, "loss": 0.0299, "step": 41950 }, { "grad_norm": 0.33969995379447937, "learning_rate": 5.008474028758846e-05, "loss": 0.0248, "step": 41960 }, { "grad_norm": 0.20461268723011017, "learning_rate": 5.0064071937899784e-05, "loss": 0.0259, "step": 41970 }, { "grad_norm": 0.32031646370887756, "learning_rate": 5.004340357726296e-05, "loss": 0.0273, "step": 41980 }, { "grad_norm": 0.24647627770900726, "learning_rate": 5.002273520920965e-05, "loss": 0.0289, "step": 41990 }, { "grad_norm": 0.265777587890625, "learning_rate": 5.000206683727151e-05, "loss": 0.023, "step": 42000 }, { "grad_norm": 0.31613653898239136, "learning_rate": 4.998139846498021e-05, "loss": 0.0216, "step": 42010 }, { "grad_norm": 0.23687364161014557, "learning_rate": 4.9960730095867405e-05, "loss": 0.0232, "step": 42020 }, { "grad_norm": 0.35257741808891296, "learning_rate": 4.9940061733464755e-05, "loss": 0.0274, "step": 42030 }, { "grad_norm": 0.13819310069084167, "learning_rate": 4.991939338130392e-05, "loss": 0.0304, "step": 42040 }, { "grad_norm": 0.26351025700569153, "learning_rate": 4.989872504291653e-05, "loss": 0.0305, "step": 42050 }, { "grad_norm": 0.25111719965934753, "learning_rate": 4.9878056721834273e-05, "loss": 0.0238, "step": 42060 }, { "grad_norm": 0.23778097331523895, "learning_rate": 4.98573884215888e-05, "loss": 0.0251, "step": 42070 }, { "grad_norm": 0.2524837553501129, "learning_rate": 4.9836720145711715e-05, "loss": 0.0255, "step": 42080 }, { "grad_norm": 0.26749101281166077, "learning_rate": 4.9816051897734725e-05, "loss": 0.0276, "step": 42090 }, { "grad_norm": 0.23641228675842285, "learning_rate": 4.979538368118941e-05, "loss": 0.0255, "step": 42100 }, { "grad_norm": 0.15552805364131927, "learning_rate": 4.9774715499607446e-05, "loss": 0.0223, "step": 42110 }, { "grad_norm": 0.18734590709209442, "learning_rate": 4.9754047356520416e-05, "loss": 0.027, "step": 42120 }, { "grad_norm": 0.19536104798316956, "learning_rate": 4.973337925545999e-05, "loss": 0.0232, "step": 42130 }, { "grad_norm": 0.1942860186100006, "learning_rate": 4.9712711199957746e-05, "loss": 0.0193, "step": 42140 }, { "grad_norm": 0.24527859687805176, "learning_rate": 4.969204319354531e-05, "loss": 0.0228, "step": 42150 }, { "grad_norm": 0.2399703562259674, "learning_rate": 4.9671375239754267e-05, "loss": 0.0241, "step": 42160 }, { "grad_norm": 0.2920941412448883, "learning_rate": 4.9650707342116195e-05, "loss": 0.0245, "step": 42170 }, { "grad_norm": 0.23205050826072693, "learning_rate": 4.96300395041627e-05, "loss": 0.0217, "step": 42180 }, { "grad_norm": 0.18873900175094604, "learning_rate": 4.960937172942532e-05, "loss": 0.021, "step": 42190 }, { "grad_norm": 0.2276620864868164, "learning_rate": 4.958870402143565e-05, "loss": 0.0211, "step": 42200 }, { "grad_norm": 0.19800135493278503, "learning_rate": 4.9568036383725186e-05, "loss": 0.0227, "step": 42210 }, { "grad_norm": 0.2206467092037201, "learning_rate": 4.9547368819825516e-05, "loss": 0.0233, "step": 42220 }, { "grad_norm": 0.2602522671222687, "learning_rate": 4.952670133326812e-05, "loss": 0.0264, "step": 42230 }, { "grad_norm": 0.32670342922210693, "learning_rate": 4.950603392758453e-05, "loss": 0.0279, "step": 42240 }, { "grad_norm": 0.14788936078548431, "learning_rate": 4.948536660630621e-05, "loss": 0.0213, "step": 42250 }, { "grad_norm": 0.21663898229599, "learning_rate": 4.9464699372964685e-05, "loss": 0.0269, "step": 42260 }, { "grad_norm": 0.1907845139503479, "learning_rate": 4.9444032231091395e-05, "loss": 0.0248, "step": 42270 }, { "grad_norm": 0.24171817302703857, "learning_rate": 4.9423365184217776e-05, "loss": 0.0212, "step": 42280 }, { "grad_norm": 0.295995831489563, "learning_rate": 4.940269823587529e-05, "loss": 0.0245, "step": 42290 }, { "grad_norm": 0.2679312527179718, "learning_rate": 4.938203138959533e-05, "loss": 0.0209, "step": 42300 }, { "grad_norm": 0.2867226004600525, "learning_rate": 4.9361364648909325e-05, "loss": 0.0227, "step": 42310 }, { "grad_norm": 0.2528705596923828, "learning_rate": 4.9340698017348605e-05, "loss": 0.0238, "step": 42320 }, { "grad_norm": 0.316751629114151, "learning_rate": 4.932003149844458e-05, "loss": 0.0239, "step": 42330 }, { "grad_norm": 0.27228885889053345, "learning_rate": 4.929936509572857e-05, "loss": 0.027, "step": 42340 }, { "grad_norm": 0.2395476996898651, "learning_rate": 4.927869881273191e-05, "loss": 0.0234, "step": 42350 }, { "grad_norm": 0.3006269037723541, "learning_rate": 4.9258032652985894e-05, "loss": 0.0254, "step": 42360 }, { "grad_norm": 0.19538839161396027, "learning_rate": 4.9237366620021786e-05, "loss": 0.0237, "step": 42370 }, { "grad_norm": 0.26405149698257446, "learning_rate": 4.921670071737089e-05, "loss": 0.021, "step": 42380 }, { "grad_norm": 0.26784858107566833, "learning_rate": 4.91960349485644e-05, "loss": 0.0279, "step": 42390 }, { "grad_norm": 0.1579703986644745, "learning_rate": 4.9175369317133566e-05, "loss": 0.0202, "step": 42400 }, { "grad_norm": 0.24928011000156403, "learning_rate": 4.915470382660953e-05, "loss": 0.0223, "step": 42410 }, { "grad_norm": 0.17672666907310486, "learning_rate": 4.9134038480523524e-05, "loss": 0.0229, "step": 42420 }, { "grad_norm": 0.22273409366607666, "learning_rate": 4.911337328240664e-05, "loss": 0.0246, "step": 42430 }, { "grad_norm": 0.2084130197763443, "learning_rate": 4.909270823579003e-05, "loss": 0.0222, "step": 42440 }, { "grad_norm": 0.2277413308620453, "learning_rate": 4.907204334420476e-05, "loss": 0.0245, "step": 42450 }, { "grad_norm": 0.26340481638908386, "learning_rate": 4.9051378611181884e-05, "loss": 0.0257, "step": 42460 }, { "grad_norm": 0.29296639561653137, "learning_rate": 4.9030714040252486e-05, "loss": 0.0271, "step": 42470 }, { "grad_norm": 0.2644171118736267, "learning_rate": 4.901004963494752e-05, "loss": 0.0181, "step": 42480 }, { "grad_norm": 0.2544196844100952, "learning_rate": 4.898938539879802e-05, "loss": 0.0289, "step": 42490 }, { "grad_norm": 0.2978544533252716, "learning_rate": 4.8968721335334904e-05, "loss": 0.0281, "step": 42500 }, { "grad_norm": 0.19035054743289948, "learning_rate": 4.894805744808912e-05, "loss": 0.0254, "step": 42510 }, { "grad_norm": 0.20959559082984924, "learning_rate": 4.8927393740591524e-05, "loss": 0.0203, "step": 42520 }, { "grad_norm": 0.2812841236591339, "learning_rate": 4.890673021637302e-05, "loss": 0.026, "step": 42530 }, { "grad_norm": 0.17884299159049988, "learning_rate": 4.888606687896442e-05, "loss": 0.0207, "step": 42540 }, { "grad_norm": 0.19512322545051575, "learning_rate": 4.886540373189652e-05, "loss": 0.023, "step": 42550 }, { "grad_norm": 0.17606714367866516, "learning_rate": 4.88447407787001e-05, "loss": 0.0229, "step": 42560 }, { "grad_norm": 0.3027251660823822, "learning_rate": 4.8824078022905864e-05, "loss": 0.0235, "step": 42570 }, { "grad_norm": 0.27609360218048096, "learning_rate": 4.880341546804455e-05, "loss": 0.0203, "step": 42580 }, { "grad_norm": 0.25510314106941223, "learning_rate": 4.878275311764679e-05, "loss": 0.024, "step": 42590 }, { "grad_norm": 0.2688657343387604, "learning_rate": 4.8762090975243254e-05, "loss": 0.0238, "step": 42600 }, { "grad_norm": 0.18977779150009155, "learning_rate": 4.874142904436448e-05, "loss": 0.0194, "step": 42610 }, { "grad_norm": 0.28879237174987793, "learning_rate": 4.8720767328541094e-05, "loss": 0.0336, "step": 42620 }, { "grad_norm": 0.28213217854499817, "learning_rate": 4.870010583130358e-05, "loss": 0.0274, "step": 42630 }, { "grad_norm": 0.2948354184627533, "learning_rate": 4.867944455618241e-05, "loss": 0.0234, "step": 42640 }, { "grad_norm": 0.21943335235118866, "learning_rate": 4.865878350670807e-05, "loss": 0.0252, "step": 42650 }, { "grad_norm": 0.304230660200119, "learning_rate": 4.8638122686410914e-05, "loss": 0.0255, "step": 42660 }, { "grad_norm": 0.2788854241371155, "learning_rate": 4.861746209882137e-05, "loss": 0.0198, "step": 42670 }, { "grad_norm": 0.2668493688106537, "learning_rate": 4.859680174746972e-05, "loss": 0.0215, "step": 42680 }, { "grad_norm": 0.24736052751541138, "learning_rate": 4.857614163588629e-05, "loss": 0.0261, "step": 42690 }, { "grad_norm": 0.32397565245628357, "learning_rate": 4.855548176760131e-05, "loss": 0.0219, "step": 42700 }, { "grad_norm": 0.2506268620491028, "learning_rate": 4.8534822146144986e-05, "loss": 0.0237, "step": 42710 }, { "grad_norm": 0.2377041131258011, "learning_rate": 4.851416277504749e-05, "loss": 0.0317, "step": 42720 }, { "grad_norm": 0.2759930491447449, "learning_rate": 4.8493503657838923e-05, "loss": 0.0215, "step": 42730 }, { "grad_norm": 0.4493981897830963, "learning_rate": 4.84728447980494e-05, "loss": 0.0219, "step": 42740 }, { "grad_norm": 0.201244056224823, "learning_rate": 4.8452186199208914e-05, "loss": 0.0248, "step": 42750 }, { "grad_norm": 0.22850051522254944, "learning_rate": 4.843152786484749e-05, "loss": 0.0263, "step": 42760 }, { "grad_norm": 0.2594640851020813, "learning_rate": 4.8410869798495035e-05, "loss": 0.0227, "step": 42770 }, { "grad_norm": 0.22753837704658508, "learning_rate": 4.8390212003681486e-05, "loss": 0.0249, "step": 42780 }, { "grad_norm": 0.29299530386924744, "learning_rate": 4.836955448393667e-05, "loss": 0.0261, "step": 42790 }, { "grad_norm": 0.19724000990390778, "learning_rate": 4.8348897242790406e-05, "loss": 0.0207, "step": 42800 }, { "grad_norm": 0.18835213780403137, "learning_rate": 4.832824028377243e-05, "loss": 0.0181, "step": 42810 }, { "grad_norm": 0.19398289918899536, "learning_rate": 4.830758361041249e-05, "loss": 0.0217, "step": 42820 }, { "grad_norm": 0.2532748878002167, "learning_rate": 4.828692722624022e-05, "loss": 0.0224, "step": 42830 }, { "grad_norm": 0.28217729926109314, "learning_rate": 4.826627113478522e-05, "loss": 0.0377, "step": 42840 }, { "grad_norm": 0.32298362255096436, "learning_rate": 4.824561533957708e-05, "loss": 0.0206, "step": 42850 }, { "grad_norm": 0.23698487877845764, "learning_rate": 4.822495984414527e-05, "loss": 0.0216, "step": 42860 }, { "grad_norm": 0.2812434136867523, "learning_rate": 4.8204304652019304e-05, "loss": 0.026, "step": 42870 }, { "grad_norm": 0.2658621370792389, "learning_rate": 4.8183649766728537e-05, "loss": 0.0242, "step": 42880 }, { "grad_norm": 0.295095294713974, "learning_rate": 4.816299519180237e-05, "loss": 0.0255, "step": 42890 }, { "grad_norm": 0.20644311606884003, "learning_rate": 4.8142340930770066e-05, "loss": 0.0215, "step": 42900 }, { "grad_norm": 0.2412225306034088, "learning_rate": 4.81216869871609e-05, "loss": 0.0195, "step": 42910 }, { "grad_norm": 0.2917287349700928, "learning_rate": 4.8101033364504064e-05, "loss": 0.022, "step": 42920 }, { "grad_norm": 0.22070477902889252, "learning_rate": 4.808038006632866e-05, "loss": 0.027, "step": 42930 }, { "grad_norm": 0.21454614400863647, "learning_rate": 4.805972709616383e-05, "loss": 0.0242, "step": 42940 }, { "grad_norm": 0.25930529832839966, "learning_rate": 4.8039074457538556e-05, "loss": 0.02, "step": 42950 }, { "grad_norm": 0.2735125720500946, "learning_rate": 4.801842215398184e-05, "loss": 0.0243, "step": 42960 }, { "grad_norm": 0.3040405213832855, "learning_rate": 4.799777018902256e-05, "loss": 0.0251, "step": 42970 }, { "grad_norm": 0.2601567804813385, "learning_rate": 4.797711856618961e-05, "loss": 0.0187, "step": 42980 }, { "grad_norm": 0.24098746478557587, "learning_rate": 4.7956467289011765e-05, "loss": 0.0218, "step": 42990 }, { "grad_norm": 0.3702254593372345, "learning_rate": 4.793581636101778e-05, "loss": 0.0227, "step": 43000 }, { "grad_norm": 0.27884626388549805, "learning_rate": 4.7915165785736326e-05, "loss": 0.0207, "step": 43010 }, { "grad_norm": 0.247185617685318, "learning_rate": 4.789451556669599e-05, "loss": 0.0207, "step": 43020 }, { "grad_norm": 0.19218677282333374, "learning_rate": 4.7873865707425384e-05, "loss": 0.0274, "step": 43030 }, { "grad_norm": 0.20423591136932373, "learning_rate": 4.7853216211452975e-05, "loss": 0.0263, "step": 43040 }, { "grad_norm": 0.28037312626838684, "learning_rate": 4.78325670823072e-05, "loss": 0.02, "step": 43050 }, { "grad_norm": 0.20182503759860992, "learning_rate": 4.781191832351641e-05, "loss": 0.0223, "step": 43060 }, { "grad_norm": 0.23788152635097504, "learning_rate": 4.7791269938608965e-05, "loss": 0.022, "step": 43070 }, { "grad_norm": 0.19954262673854828, "learning_rate": 4.777062193111305e-05, "loss": 0.0222, "step": 43080 }, { "grad_norm": 0.16852936148643494, "learning_rate": 4.77499743045569e-05, "loss": 0.0226, "step": 43090 }, { "grad_norm": 0.24094383418560028, "learning_rate": 4.7729327062468596e-05, "loss": 0.0221, "step": 43100 }, { "grad_norm": 0.2584010064601898, "learning_rate": 4.770868020837619e-05, "loss": 0.0249, "step": 43110 }, { "grad_norm": 0.27528098225593567, "learning_rate": 4.768803374580768e-05, "loss": 0.0187, "step": 43120 }, { "grad_norm": 0.21864451467990875, "learning_rate": 4.766738767829094e-05, "loss": 0.0224, "step": 43130 }, { "grad_norm": 0.288633793592453, "learning_rate": 4.764674200935388e-05, "loss": 0.0236, "step": 43140 }, { "grad_norm": 0.1837245225906372, "learning_rate": 4.762609674252424e-05, "loss": 0.0215, "step": 43150 }, { "grad_norm": 0.2826879024505615, "learning_rate": 4.760545188132974e-05, "loss": 0.0313, "step": 43160 }, { "grad_norm": 0.23233598470687866, "learning_rate": 4.758480742929801e-05, "loss": 0.0226, "step": 43170 }, { "grad_norm": 0.1899067610502243, "learning_rate": 4.756416338995664e-05, "loss": 0.0295, "step": 43180 }, { "grad_norm": 0.22935634851455688, "learning_rate": 4.7543519766833145e-05, "loss": 0.0219, "step": 43190 }, { "grad_norm": 0.1874462217092514, "learning_rate": 4.752287656345492e-05, "loss": 0.0224, "step": 43200 }, { "grad_norm": 0.20841804146766663, "learning_rate": 4.7502233783349355e-05, "loss": 0.0238, "step": 43210 }, { "grad_norm": 0.18910779058933258, "learning_rate": 4.7481591430043694e-05, "loss": 0.023, "step": 43220 }, { "grad_norm": 0.26750242710113525, "learning_rate": 4.746094950706521e-05, "loss": 0.0259, "step": 43230 }, { "grad_norm": 0.23556195199489594, "learning_rate": 4.744030801794099e-05, "loss": 0.0216, "step": 43240 }, { "grad_norm": 0.25395509600639343, "learning_rate": 4.741966696619813e-05, "loss": 0.0272, "step": 43250 }, { "grad_norm": 0.22444109618663788, "learning_rate": 4.739902635536361e-05, "loss": 0.0246, "step": 43260 }, { "grad_norm": 0.23223648965358734, "learning_rate": 4.737838618896436e-05, "loss": 0.0183, "step": 43270 }, { "grad_norm": 0.2166605144739151, "learning_rate": 4.7357746470527203e-05, "loss": 0.025, "step": 43280 }, { "grad_norm": 0.2646403908729553, "learning_rate": 4.73371072035789e-05, "loss": 0.0279, "step": 43290 }, { "grad_norm": 0.24597978591918945, "learning_rate": 4.731646839164616e-05, "loss": 0.0218, "step": 43300 }, { "grad_norm": 0.2493031620979309, "learning_rate": 4.729583003825556e-05, "loss": 0.0221, "step": 43310 }, { "grad_norm": 0.24766381084918976, "learning_rate": 4.727519214693367e-05, "loss": 0.0257, "step": 43320 }, { "grad_norm": 0.3203088343143463, "learning_rate": 4.725455472120689e-05, "loss": 0.0279, "step": 43330 }, { "grad_norm": 0.2330351024866104, "learning_rate": 4.723391776460164e-05, "loss": 0.0255, "step": 43340 }, { "grad_norm": 0.1998222917318344, "learning_rate": 4.7213281280644186e-05, "loss": 0.0218, "step": 43350 }, { "grad_norm": 0.2595415413379669, "learning_rate": 4.719264527286075e-05, "loss": 0.0215, "step": 43360 }, { "grad_norm": 0.2832825779914856, "learning_rate": 4.717200974477744e-05, "loss": 0.0278, "step": 43370 }, { "grad_norm": 0.4754774868488312, "learning_rate": 4.715137469992034e-05, "loss": 0.0251, "step": 43380 }, { "grad_norm": 0.22943495213985443, "learning_rate": 4.713074014181539e-05, "loss": 0.0229, "step": 43390 }, { "grad_norm": 0.393993079662323, "learning_rate": 4.711010607398846e-05, "loss": 0.0202, "step": 43400 }, { "grad_norm": 0.3174806833267212, "learning_rate": 4.708947249996538e-05, "loss": 0.0226, "step": 43410 }, { "grad_norm": 0.224710151553154, "learning_rate": 4.706883942327183e-05, "loss": 0.0225, "step": 43420 }, { "grad_norm": 0.2650555670261383, "learning_rate": 4.704820684743347e-05, "loss": 0.0238, "step": 43430 }, { "grad_norm": 0.2299470156431198, "learning_rate": 4.702757477597581e-05, "loss": 0.0243, "step": 43440 }, { "grad_norm": 0.2785743474960327, "learning_rate": 4.700694321242435e-05, "loss": 0.0199, "step": 43450 }, { "grad_norm": 0.30006760358810425, "learning_rate": 4.698631216030441e-05, "loss": 0.0255, "step": 43460 }, { "grad_norm": 0.258103609085083, "learning_rate": 4.6965681623141314e-05, "loss": 0.0223, "step": 43470 }, { "grad_norm": 0.2002338171005249, "learning_rate": 4.694505160446024e-05, "loss": 0.0193, "step": 43480 }, { "grad_norm": 0.230246439576149, "learning_rate": 4.692442210778628e-05, "loss": 0.0241, "step": 43490 }, { "grad_norm": 0.32496100664138794, "learning_rate": 4.690379313664448e-05, "loss": 0.0206, "step": 43500 }, { "grad_norm": 0.265316367149353, "learning_rate": 4.688316469455973e-05, "loss": 0.0256, "step": 43510 }, { "grad_norm": 0.33445534110069275, "learning_rate": 4.6862536785056914e-05, "loss": 0.0203, "step": 43520 }, { "grad_norm": 0.24488528072834015, "learning_rate": 4.684190941166073e-05, "loss": 0.02, "step": 43530 }, { "grad_norm": 0.25683385133743286, "learning_rate": 4.682128257789587e-05, "loss": 0.0211, "step": 43540 }, { "grad_norm": 0.22167356312274933, "learning_rate": 4.680065628728687e-05, "loss": 0.0192, "step": 43550 }, { "grad_norm": 0.25036364793777466, "learning_rate": 4.678003054335822e-05, "loss": 0.0264, "step": 43560 }, { "grad_norm": 0.23368215560913086, "learning_rate": 4.675940534963428e-05, "loss": 0.0277, "step": 43570 }, { "grad_norm": 0.2902870774269104, "learning_rate": 4.673878070963931e-05, "loss": 0.0242, "step": 43580 }, { "grad_norm": 0.15792939066886902, "learning_rate": 4.671815662689756e-05, "loss": 0.0186, "step": 43590 }, { "grad_norm": 0.2457592785358429, "learning_rate": 4.669753310493306e-05, "loss": 0.0223, "step": 43600 }, { "grad_norm": 0.2334604114294052, "learning_rate": 4.6676910147269845e-05, "loss": 0.0229, "step": 43610 }, { "grad_norm": 0.1949252337217331, "learning_rate": 4.665628775743177e-05, "loss": 0.0193, "step": 43620 }, { "grad_norm": 0.29031234979629517, "learning_rate": 4.66356659389427e-05, "loss": 0.0194, "step": 43630 }, { "grad_norm": 0.21007417142391205, "learning_rate": 4.661504469532629e-05, "loss": 0.0193, "step": 43640 }, { "grad_norm": 0.2885582745075226, "learning_rate": 4.659442403010617e-05, "loss": 0.0187, "step": 43650 }, { "grad_norm": 0.30424535274505615, "learning_rate": 4.6573803946805845e-05, "loss": 0.0181, "step": 43660 }, { "grad_norm": 0.18901115655899048, "learning_rate": 4.6553184448948695e-05, "loss": 0.0242, "step": 43670 }, { "grad_norm": 0.2302614152431488, "learning_rate": 4.653256554005807e-05, "loss": 0.0205, "step": 43680 }, { "grad_norm": 0.18510600924491882, "learning_rate": 4.6511947223657145e-05, "loss": 0.0245, "step": 43690 }, { "grad_norm": 0.2592456340789795, "learning_rate": 4.649132950326906e-05, "loss": 0.0243, "step": 43700 }, { "grad_norm": 0.2178872525691986, "learning_rate": 4.647071238241679e-05, "loss": 0.0223, "step": 43710 }, { "grad_norm": 0.22993682324886322, "learning_rate": 4.6450095864623264e-05, "loss": 0.0194, "step": 43720 }, { "grad_norm": 0.45034289360046387, "learning_rate": 4.642947995341123e-05, "loss": 0.0184, "step": 43730 }, { "grad_norm": 0.24849426746368408, "learning_rate": 4.640886465230345e-05, "loss": 0.0213, "step": 43740 }, { "grad_norm": 0.22488071024417877, "learning_rate": 4.6388249964822485e-05, "loss": 0.0237, "step": 43750 }, { "grad_norm": 0.1886899322271347, "learning_rate": 4.63676358944908e-05, "loss": 0.0202, "step": 43760 }, { "grad_norm": 0.3138142228126526, "learning_rate": 4.634702244483081e-05, "loss": 0.0264, "step": 43770 }, { "grad_norm": 0.26837509870529175, "learning_rate": 4.632640961936475e-05, "loss": 0.0212, "step": 43780 }, { "grad_norm": 0.20501528680324554, "learning_rate": 4.6305797421614835e-05, "loss": 0.0232, "step": 43790 }, { "grad_norm": 0.18672509491443634, "learning_rate": 4.6285185855103105e-05, "loss": 0.0246, "step": 43800 }, { "grad_norm": 0.24651078879833221, "learning_rate": 4.626457492335151e-05, "loss": 0.0235, "step": 43810 }, { "grad_norm": 0.13798131048679352, "learning_rate": 4.624396462988188e-05, "loss": 0.0193, "step": 43820 }, { "grad_norm": 0.21830898523330688, "learning_rate": 4.6223354978215985e-05, "loss": 0.0223, "step": 43830 }, { "grad_norm": 0.1960180252790451, "learning_rate": 4.620274597187544e-05, "loss": 0.0218, "step": 43840 }, { "grad_norm": 0.27385857701301575, "learning_rate": 4.6182137614381736e-05, "loss": 0.0193, "step": 43850 }, { "grad_norm": 0.1760118305683136, "learning_rate": 4.616152990925631e-05, "loss": 0.0202, "step": 43860 }, { "grad_norm": 0.1551445722579956, "learning_rate": 4.614092286002041e-05, "loss": 0.0207, "step": 43870 }, { "grad_norm": 0.2590339481830597, "learning_rate": 4.612031647019528e-05, "loss": 0.0229, "step": 43880 }, { "grad_norm": 0.2949222922325134, "learning_rate": 4.609971074330193e-05, "loss": 0.025, "step": 43890 }, { "grad_norm": 0.19935955107212067, "learning_rate": 4.607910568286136e-05, "loss": 0.0257, "step": 43900 }, { "grad_norm": 0.2221204787492752, "learning_rate": 4.605850129239438e-05, "loss": 0.0215, "step": 43910 }, { "grad_norm": 0.20221351087093353, "learning_rate": 4.603789757542174e-05, "loss": 0.0183, "step": 43920 }, { "grad_norm": 0.3140532374382019, "learning_rate": 4.601729453546404e-05, "loss": 0.0258, "step": 43930 }, { "grad_norm": 0.201470285654068, "learning_rate": 4.599669217604177e-05, "loss": 0.0186, "step": 43940 }, { "grad_norm": 0.1875544786453247, "learning_rate": 4.597609050067532e-05, "loss": 0.0257, "step": 43950 }, { "grad_norm": 0.2392701953649521, "learning_rate": 4.595548951288495e-05, "loss": 0.0236, "step": 43960 }, { "grad_norm": 0.2684270143508911, "learning_rate": 4.593488921619081e-05, "loss": 0.0284, "step": 43970 }, { "grad_norm": 0.2557547092437744, "learning_rate": 4.591428961411289e-05, "loss": 0.0271, "step": 43980 }, { "grad_norm": 0.23900732398033142, "learning_rate": 4.589369071017117e-05, "loss": 0.0283, "step": 43990 }, { "grad_norm": 0.17385779321193695, "learning_rate": 4.587309250788538e-05, "loss": 0.0257, "step": 44000 }, { "grad_norm": 0.18162745237350464, "learning_rate": 4.585249501077522e-05, "loss": 0.0213, "step": 44010 }, { "grad_norm": 0.19502492249011993, "learning_rate": 4.583189822236021e-05, "loss": 0.0241, "step": 44020 }, { "grad_norm": 0.2524573802947998, "learning_rate": 4.5811302146159816e-05, "loss": 0.0214, "step": 44030 }, { "grad_norm": 0.1938321590423584, "learning_rate": 4.579070678569332e-05, "loss": 0.0215, "step": 44040 }, { "grad_norm": 0.32822030782699585, "learning_rate": 4.5770112144479904e-05, "loss": 0.0264, "step": 44050 }, { "grad_norm": 0.2629268765449524, "learning_rate": 4.5749518226038645e-05, "loss": 0.0226, "step": 44060 }, { "grad_norm": 0.3047578036785126, "learning_rate": 4.572892503388845e-05, "loss": 0.0202, "step": 44070 }, { "grad_norm": 0.15459521114826202, "learning_rate": 4.570833257154817e-05, "loss": 0.0239, "step": 44080 }, { "grad_norm": 0.2219657599925995, "learning_rate": 4.568774084253646e-05, "loss": 0.0274, "step": 44090 }, { "grad_norm": 0.20557032525539398, "learning_rate": 4.566714985037191e-05, "loss": 0.0254, "step": 44100 }, { "grad_norm": 0.2817263603210449, "learning_rate": 4.564655959857295e-05, "loss": 0.0242, "step": 44110 }, { "grad_norm": 0.3183901309967041, "learning_rate": 4.5625970090657893e-05, "loss": 0.0216, "step": 44120 }, { "grad_norm": 0.15627439320087433, "learning_rate": 4.560538133014491e-05, "loss": 0.0217, "step": 44130 }, { "grad_norm": 0.17049001157283783, "learning_rate": 4.5584793320552055e-05, "loss": 0.0173, "step": 44140 }, { "grad_norm": 0.26490160822868347, "learning_rate": 4.556420606539728e-05, "loss": 0.0214, "step": 44150 }, { "grad_norm": 0.15099835395812988, "learning_rate": 4.554361956819836e-05, "loss": 0.0208, "step": 44160 }, { "grad_norm": 0.1970723569393158, "learning_rate": 4.552303383247299e-05, "loss": 0.0186, "step": 44170 }, { "grad_norm": 0.2784380614757538, "learning_rate": 4.5502448861738664e-05, "loss": 0.0222, "step": 44180 }, { "grad_norm": 0.26166385412216187, "learning_rate": 4.548186465951284e-05, "loss": 0.032, "step": 44190 }, { "grad_norm": 0.19193509221076965, "learning_rate": 4.546128122931277e-05, "loss": 0.0234, "step": 44200 }, { "grad_norm": 0.24071131646633148, "learning_rate": 4.544069857465561e-05, "loss": 0.0216, "step": 44210 }, { "grad_norm": 0.23504778742790222, "learning_rate": 4.542011669905837e-05, "loss": 0.0192, "step": 44220 }, { "grad_norm": 0.25191813707351685, "learning_rate": 4.53995356060379e-05, "loss": 0.0188, "step": 44230 }, { "grad_norm": 0.23466989398002625, "learning_rate": 4.5378955299110994e-05, "loss": 0.0216, "step": 44240 }, { "grad_norm": 0.18849937617778778, "learning_rate": 4.5358375781794233e-05, "loss": 0.0212, "step": 44250 }, { "grad_norm": 0.15033257007598877, "learning_rate": 4.533779705760411e-05, "loss": 0.0244, "step": 44260 }, { "grad_norm": 0.24729062616825104, "learning_rate": 4.5317219130056934e-05, "loss": 0.0209, "step": 44270 }, { "grad_norm": 0.2986305058002472, "learning_rate": 4.5296642002668964e-05, "loss": 0.0234, "step": 44280 }, { "grad_norm": 0.28116822242736816, "learning_rate": 4.5276065678956216e-05, "loss": 0.0274, "step": 44290 }, { "grad_norm": 0.27295181155204773, "learning_rate": 4.525549016243466e-05, "loss": 0.0208, "step": 44300 }, { "grad_norm": 0.272645503282547, "learning_rate": 4.523491545662008e-05, "loss": 0.0201, "step": 44310 }, { "grad_norm": 0.23518091440200806, "learning_rate": 4.5214341565028104e-05, "loss": 0.024, "step": 44320 }, { "grad_norm": 0.33026769757270813, "learning_rate": 4.519376849117428e-05, "loss": 0.0211, "step": 44330 }, { "grad_norm": 0.2781289219856262, "learning_rate": 4.517319623857395e-05, "loss": 0.0271, "step": 44340 }, { "grad_norm": 0.2029438465833664, "learning_rate": 4.5152624810742394e-05, "loss": 0.0219, "step": 44350 }, { "grad_norm": 0.22947244346141815, "learning_rate": 4.5132054211194664e-05, "loss": 0.0206, "step": 44360 }, { "grad_norm": 0.2985520660877228, "learning_rate": 4.511148444344574e-05, "loss": 0.0192, "step": 44370 }, { "grad_norm": 0.2681902050971985, "learning_rate": 4.509091551101041e-05, "loss": 0.0285, "step": 44380 }, { "grad_norm": 0.3114836812019348, "learning_rate": 4.507034741740338e-05, "loss": 0.0219, "step": 44390 }, { "grad_norm": 0.2939586341381073, "learning_rate": 4.5049780166139145e-05, "loss": 0.0255, "step": 44400 }, { "grad_norm": 0.25500571727752686, "learning_rate": 4.5029213760732075e-05, "loss": 0.0315, "step": 44410 }, { "grad_norm": 0.211030051112175, "learning_rate": 4.5008648204696434e-05, "loss": 0.0257, "step": 44420 }, { "grad_norm": 0.22414720058441162, "learning_rate": 4.4988083501546284e-05, "loss": 0.0245, "step": 44430 }, { "grad_norm": 0.14919424057006836, "learning_rate": 4.4967519654795606e-05, "loss": 0.0212, "step": 44440 }, { "grad_norm": 0.24872088432312012, "learning_rate": 4.494695666795816e-05, "loss": 0.0193, "step": 44450 }, { "grad_norm": 0.2607574760913849, "learning_rate": 4.4926394544547644e-05, "loss": 0.0228, "step": 44460 }, { "grad_norm": 0.25041699409484863, "learning_rate": 4.490583328807752e-05, "loss": 0.0251, "step": 44470 }, { "grad_norm": 0.21529537439346313, "learning_rate": 4.488527290206117e-05, "loss": 0.0209, "step": 44480 }, { "grad_norm": 0.17517875134944916, "learning_rate": 4.48647133900118e-05, "loss": 0.0202, "step": 44490 }, { "grad_norm": 0.19431956112384796, "learning_rate": 4.484415475544244e-05, "loss": 0.0247, "step": 44500 }, { "grad_norm": 0.1996307671070099, "learning_rate": 4.4823597001866035e-05, "loss": 0.0227, "step": 44510 }, { "grad_norm": 0.23879481852054596, "learning_rate": 4.480304013279532e-05, "loss": 0.0207, "step": 44520 }, { "grad_norm": 0.1955704391002655, "learning_rate": 4.478248415174292e-05, "loss": 0.0202, "step": 44530 }, { "grad_norm": 0.2539553642272949, "learning_rate": 4.476192906222126e-05, "loss": 0.0225, "step": 44540 }, { "grad_norm": 0.24202966690063477, "learning_rate": 4.474137486774268e-05, "loss": 0.0338, "step": 44550 }, { "grad_norm": 0.2104383409023285, "learning_rate": 4.4720821571819296e-05, "loss": 0.0235, "step": 44560 }, { "grad_norm": 0.234160378575325, "learning_rate": 4.470026917796314e-05, "loss": 0.0279, "step": 44570 }, { "grad_norm": 0.22070899605751038, "learning_rate": 4.4679717689686005e-05, "loss": 0.0361, "step": 44580 }, { "grad_norm": 0.19393470883369446, "learning_rate": 4.465916711049963e-05, "loss": 0.0195, "step": 44590 }, { "grad_norm": 0.12687042355537415, "learning_rate": 4.4638617443915524e-05, "loss": 0.0225, "step": 44600 }, { "grad_norm": 0.24704980850219727, "learning_rate": 4.4618068693445055e-05, "loss": 0.0225, "step": 44610 }, { "grad_norm": 0.2823928892612457, "learning_rate": 4.459752086259946e-05, "loss": 0.0231, "step": 44620 }, { "grad_norm": 0.25911253690719604, "learning_rate": 4.457697395488977e-05, "loss": 0.0243, "step": 44630 }, { "grad_norm": 0.23010633885860443, "learning_rate": 4.455642797382693e-05, "loss": 0.0303, "step": 44640 }, { "grad_norm": 0.32908469438552856, "learning_rate": 4.4535882922921643e-05, "loss": 0.0236, "step": 44650 }, { "grad_norm": 0.29047563672065735, "learning_rate": 4.451533880568455e-05, "loss": 0.027, "step": 44660 }, { "grad_norm": 0.2508256435394287, "learning_rate": 4.449479562562603e-05, "loss": 0.0233, "step": 44670 }, { "grad_norm": 0.24426455795764923, "learning_rate": 4.447425338625639e-05, "loss": 0.0237, "step": 44680 }, { "grad_norm": 0.31480491161346436, "learning_rate": 4.4453712091085705e-05, "loss": 0.0221, "step": 44690 }, { "grad_norm": 0.20505960285663605, "learning_rate": 4.443317174362392e-05, "loss": 0.0259, "step": 44700 }, { "grad_norm": 0.1656040996313095, "learning_rate": 4.441263234738085e-05, "loss": 0.0221, "step": 44710 }, { "grad_norm": 0.21567532420158386, "learning_rate": 4.4392093905866086e-05, "loss": 0.0166, "step": 44720 }, { "grad_norm": 0.22808164358139038, "learning_rate": 4.437155642258911e-05, "loss": 0.0225, "step": 44730 }, { "grad_norm": 0.2585277855396271, "learning_rate": 4.4351019901059177e-05, "loss": 0.0266, "step": 44740 }, { "grad_norm": 0.24810339510440826, "learning_rate": 4.433048434478547e-05, "loss": 0.0314, "step": 44750 }, { "grad_norm": 0.19175416231155396, "learning_rate": 4.4309949757276906e-05, "loss": 0.0237, "step": 44760 }, { "grad_norm": 0.22268211841583252, "learning_rate": 4.428941614204233e-05, "loss": 0.0277, "step": 44770 }, { "grad_norm": 0.33567944169044495, "learning_rate": 4.426888350259034e-05, "loss": 0.0219, "step": 44780 }, { "grad_norm": 0.2586279511451721, "learning_rate": 4.424835184242939e-05, "loss": 0.0338, "step": 44790 }, { "grad_norm": 0.18468694388866425, "learning_rate": 4.422782116506784e-05, "loss": 0.0197, "step": 44800 }, { "grad_norm": 0.2654852867126465, "learning_rate": 4.4207291474013754e-05, "loss": 0.0245, "step": 44810 }, { "grad_norm": 0.2841859757900238, "learning_rate": 4.418676277277514e-05, "loss": 0.0308, "step": 44820 }, { "grad_norm": 0.1961567997932434, "learning_rate": 4.416623506485976e-05, "loss": 0.0252, "step": 44830 }, { "grad_norm": 0.26724016666412354, "learning_rate": 4.414570835377527e-05, "loss": 0.0222, "step": 44840 }, { "grad_norm": 0.21623532474040985, "learning_rate": 4.412518264302909e-05, "loss": 0.0225, "step": 44850 }, { "grad_norm": 0.2307257503271103, "learning_rate": 4.4104657936128535e-05, "loss": 0.0192, "step": 44860 }, { "grad_norm": 0.21588608622550964, "learning_rate": 4.408413423658071e-05, "loss": 0.0254, "step": 44870 }, { "grad_norm": 0.3296002745628357, "learning_rate": 4.406361154789253e-05, "loss": 0.0223, "step": 44880 }, { "grad_norm": 0.2666826546192169, "learning_rate": 4.4043089873570776e-05, "loss": 0.0277, "step": 44890 }, { "grad_norm": 0.24349208176136017, "learning_rate": 4.4022569217122025e-05, "loss": 0.0226, "step": 44900 }, { "grad_norm": 0.2715148627758026, "learning_rate": 4.400204958205274e-05, "loss": 0.0236, "step": 44910 }, { "grad_norm": 0.2524968087673187, "learning_rate": 4.3981530971869125e-05, "loss": 0.0277, "step": 44920 }, { "grad_norm": 0.20581452548503876, "learning_rate": 4.396101339007727e-05, "loss": 0.0225, "step": 44930 }, { "grad_norm": 0.17731913924217224, "learning_rate": 4.394049684018304e-05, "loss": 0.022, "step": 44940 }, { "grad_norm": 0.2315225899219513, "learning_rate": 4.3919981325692186e-05, "loss": 0.0245, "step": 44950 }, { "grad_norm": 0.210836261510849, "learning_rate": 4.389946685011024e-05, "loss": 0.0172, "step": 44960 }, { "grad_norm": 0.2040986716747284, "learning_rate": 4.387895341694255e-05, "loss": 0.0291, "step": 44970 }, { "grad_norm": 0.31195613741874695, "learning_rate": 4.3858441029694324e-05, "loss": 0.0286, "step": 44980 }, { "grad_norm": 0.20173315703868866, "learning_rate": 4.3837929691870527e-05, "loss": 0.0194, "step": 44990 }, { "grad_norm": 0.29207783937454224, "learning_rate": 4.381741940697604e-05, "loss": 0.0251, "step": 45000 }, { "grad_norm": 0.22321903705596924, "learning_rate": 4.379691017851547e-05, "loss": 0.0187, "step": 45010 }, { "grad_norm": 0.2999417185783386, "learning_rate": 4.3776402009993304e-05, "loss": 0.0271, "step": 45020 }, { "grad_norm": 0.23126673698425293, "learning_rate": 4.3755894904913794e-05, "loss": 0.02, "step": 45030 }, { "grad_norm": 0.312435120344162, "learning_rate": 4.373538886678109e-05, "loss": 0.0224, "step": 45040 }, { "grad_norm": 0.21536852419376373, "learning_rate": 4.371488389909909e-05, "loss": 0.0179, "step": 45050 }, { "grad_norm": 0.22771546244621277, "learning_rate": 4.3694380005371515e-05, "loss": 0.0249, "step": 45060 }, { "grad_norm": 0.27884528040885925, "learning_rate": 4.367387718910196e-05, "loss": 0.0204, "step": 45070 }, { "grad_norm": 0.29629838466644287, "learning_rate": 4.3653375453793764e-05, "loss": 0.0232, "step": 45080 }, { "grad_norm": 0.24930439889431, "learning_rate": 4.3632874802950136e-05, "loss": 0.0214, "step": 45090 }, { "grad_norm": 0.22198516130447388, "learning_rate": 4.3612375240074034e-05, "loss": 0.0229, "step": 45100 }, { "grad_norm": 0.22815130650997162, "learning_rate": 4.3591876768668325e-05, "loss": 0.0228, "step": 45110 }, { "grad_norm": 0.19595201313495636, "learning_rate": 4.3571379392235605e-05, "loss": 0.0201, "step": 45120 }, { "grad_norm": 0.25183698534965515, "learning_rate": 4.3550883114278335e-05, "loss": 0.0283, "step": 45130 }, { "grad_norm": 0.16945038735866547, "learning_rate": 4.353038793829876e-05, "loss": 0.0196, "step": 45140 }, { "grad_norm": 0.22700929641723633, "learning_rate": 4.350989386779891e-05, "loss": 0.0221, "step": 45150 }, { "grad_norm": 0.22334933280944824, "learning_rate": 4.3489400906280724e-05, "loss": 0.0329, "step": 45160 }, { "grad_norm": 0.23096632957458496, "learning_rate": 4.3468909057245845e-05, "loss": 0.0238, "step": 45170 }, { "grad_norm": 0.35524415969848633, "learning_rate": 4.3448418324195794e-05, "loss": 0.0226, "step": 45180 }, { "grad_norm": 0.45134374499320984, "learning_rate": 4.342792871063184e-05, "loss": 0.0215, "step": 45190 }, { "grad_norm": 0.2957838773727417, "learning_rate": 4.3407440220055145e-05, "loss": 0.0241, "step": 45200 }, { "grad_norm": 0.23554329574108124, "learning_rate": 4.33869528559666e-05, "loss": 0.0187, "step": 45210 }, { "grad_norm": 0.24442274868488312, "learning_rate": 4.336646662186696e-05, "loss": 0.0202, "step": 45220 }, { "grad_norm": 0.2498571127653122, "learning_rate": 4.334598152125672e-05, "loss": 0.0251, "step": 45230 }, { "grad_norm": 0.2047460973262787, "learning_rate": 4.3325497557636276e-05, "loss": 0.0296, "step": 45240 }, { "grad_norm": 0.19733980298042297, "learning_rate": 4.330501473450574e-05, "loss": 0.0212, "step": 45250 }, { "grad_norm": 0.26957419514656067, "learning_rate": 4.328453305536507e-05, "loss": 0.0186, "step": 45260 }, { "grad_norm": 0.21843191981315613, "learning_rate": 4.326405252371404e-05, "loss": 0.0226, "step": 45270 }, { "grad_norm": 0.2141331285238266, "learning_rate": 4.324357314305221e-05, "loss": 0.022, "step": 45280 }, { "grad_norm": 0.21606343984603882, "learning_rate": 4.3223094916878945e-05, "loss": 0.0234, "step": 45290 }, { "grad_norm": 0.20978443324565887, "learning_rate": 4.320261784869338e-05, "loss": 0.0302, "step": 45300 }, { "grad_norm": 0.19023491442203522, "learning_rate": 4.318214194199455e-05, "loss": 0.027, "step": 45310 }, { "grad_norm": 0.25994375348091125, "learning_rate": 4.316166720028118e-05, "loss": 0.0163, "step": 45320 }, { "grad_norm": 0.25863349437713623, "learning_rate": 4.3141193627051864e-05, "loss": 0.0232, "step": 45330 }, { "grad_norm": 0.2755209803581238, "learning_rate": 4.312072122580496e-05, "loss": 0.0239, "step": 45340 }, { "grad_norm": 0.1843230128288269, "learning_rate": 4.3100250000038646e-05, "loss": 0.0224, "step": 45350 }, { "grad_norm": 0.273656964302063, "learning_rate": 4.307977995325091e-05, "loss": 0.0302, "step": 45360 }, { "grad_norm": 0.23663872480392456, "learning_rate": 4.30593110889395e-05, "loss": 0.0238, "step": 45370 }, { "grad_norm": 0.22167134284973145, "learning_rate": 4.3038843410602016e-05, "loss": 0.0224, "step": 45380 }, { "grad_norm": 0.19868528842926025, "learning_rate": 4.3018376921735774e-05, "loss": 0.0227, "step": 45390 }, { "grad_norm": 0.14827270805835724, "learning_rate": 4.299791162583799e-05, "loss": 0.0206, "step": 45400 }, { "grad_norm": 0.1956629455089569, "learning_rate": 4.29774475264056e-05, "loss": 0.0224, "step": 45410 }, { "grad_norm": 0.22372214496135712, "learning_rate": 4.2956984626935365e-05, "loss": 0.0176, "step": 45420 }, { "grad_norm": 0.18423686921596527, "learning_rate": 4.293652293092383e-05, "loss": 0.0171, "step": 45430 }, { "grad_norm": 0.21736100316047668, "learning_rate": 4.2916062441867324e-05, "loss": 0.0244, "step": 45440 }, { "grad_norm": 0.26608866453170776, "learning_rate": 4.289560316326201e-05, "loss": 0.0266, "step": 45450 }, { "grad_norm": 0.23928497731685638, "learning_rate": 4.28751450986038e-05, "loss": 0.0268, "step": 45460 }, { "grad_norm": 0.22554899752140045, "learning_rate": 4.2854688251388444e-05, "loss": 0.0211, "step": 45470 }, { "grad_norm": 0.28935757279396057, "learning_rate": 4.2834232625111425e-05, "loss": 0.0253, "step": 45480 }, { "grad_norm": 0.17087410390377045, "learning_rate": 4.2813778223268086e-05, "loss": 0.0211, "step": 45490 }, { "grad_norm": 0.29962965846061707, "learning_rate": 4.2793325049353477e-05, "loss": 0.0208, "step": 45500 }, { "grad_norm": 0.2730606198310852, "learning_rate": 4.2772873106862535e-05, "loss": 0.0238, "step": 45510 }, { "grad_norm": 0.22231921553611755, "learning_rate": 4.275242239928993e-05, "loss": 0.0225, "step": 45520 }, { "grad_norm": 0.215668186545372, "learning_rate": 4.273197293013009e-05, "loss": 0.0172, "step": 45530 }, { "grad_norm": 0.21470944583415985, "learning_rate": 4.271152470287731e-05, "loss": 0.0167, "step": 45540 }, { "grad_norm": 0.3175458014011383, "learning_rate": 4.2691077721025594e-05, "loss": 0.0226, "step": 45550 }, { "grad_norm": 0.17038887739181519, "learning_rate": 4.267063198806883e-05, "loss": 0.0213, "step": 45560 }, { "grad_norm": 0.19720439612865448, "learning_rate": 4.2650187507500574e-05, "loss": 0.0181, "step": 45570 }, { "grad_norm": 0.15164914727210999, "learning_rate": 4.2629744282814275e-05, "loss": 0.0202, "step": 45580 }, { "grad_norm": 0.24384741485118866, "learning_rate": 4.2609302317503074e-05, "loss": 0.0187, "step": 45590 }, { "grad_norm": 0.19807656109333038, "learning_rate": 4.258886161505999e-05, "loss": 0.022, "step": 45600 }, { "grad_norm": 0.2333332747220993, "learning_rate": 4.2568422178977775e-05, "loss": 0.0231, "step": 45610 }, { "grad_norm": 0.1784185916185379, "learning_rate": 4.254798401274894e-05, "loss": 0.0247, "step": 45620 }, { "grad_norm": 0.2907821536064148, "learning_rate": 4.252754711986583e-05, "loss": 0.0206, "step": 45630 }, { "grad_norm": 0.2617238461971283, "learning_rate": 4.250711150382052e-05, "loss": 0.0198, "step": 45640 }, { "grad_norm": 0.1701219379901886, "learning_rate": 4.248667716810495e-05, "loss": 0.0231, "step": 45650 }, { "grad_norm": 0.28631433844566345, "learning_rate": 4.246624411621074e-05, "loss": 0.0182, "step": 45660 }, { "grad_norm": 0.2578663229942322, "learning_rate": 4.244581235162938e-05, "loss": 0.0204, "step": 45670 }, { "grad_norm": 0.2110958993434906, "learning_rate": 4.2425381877852075e-05, "loss": 0.0258, "step": 45680 }, { "grad_norm": 0.25734296441078186, "learning_rate": 4.2404952698369856e-05, "loss": 0.0196, "step": 45690 }, { "grad_norm": 0.22195492684841156, "learning_rate": 4.238452481667349e-05, "loss": 0.0222, "step": 45700 }, { "grad_norm": 0.2516726851463318, "learning_rate": 4.2364098236253526e-05, "loss": 0.0251, "step": 45710 }, { "grad_norm": 0.2622956335544586, "learning_rate": 4.2343672960600356e-05, "loss": 0.0217, "step": 45720 }, { "grad_norm": 0.19818814098834991, "learning_rate": 4.232324899320406e-05, "loss": 0.0173, "step": 45730 }, { "grad_norm": 0.17766335606575012, "learning_rate": 4.230282633755457e-05, "loss": 0.0204, "step": 45740 }, { "grad_norm": 0.2599754333496094, "learning_rate": 4.2282404997141515e-05, "loss": 0.0224, "step": 45750 }, { "grad_norm": 0.18868590891361237, "learning_rate": 4.2261984975454397e-05, "loss": 0.0233, "step": 45760 }, { "grad_norm": 0.2165953516960144, "learning_rate": 4.224156627598239e-05, "loss": 0.0201, "step": 45770 }, { "grad_norm": 0.19109371304512024, "learning_rate": 4.222114890221453e-05, "loss": 0.0224, "step": 45780 }, { "grad_norm": 0.22282037138938904, "learning_rate": 4.2200732857639546e-05, "loss": 0.0298, "step": 45790 }, { "grad_norm": 0.19907057285308838, "learning_rate": 4.2180318145746035e-05, "loss": 0.0198, "step": 45800 }, { "grad_norm": 0.2603458762168884, "learning_rate": 4.215990477002227e-05, "loss": 0.0261, "step": 45810 }, { "grad_norm": 0.28821393847465515, "learning_rate": 4.2139492733956356e-05, "loss": 0.0191, "step": 45820 }, { "grad_norm": 0.17887815833091736, "learning_rate": 4.211908204103615e-05, "loss": 0.0281, "step": 45830 }, { "grad_norm": 0.1905060112476349, "learning_rate": 4.2098672694749265e-05, "loss": 0.0321, "step": 45840 }, { "grad_norm": 0.14237862825393677, "learning_rate": 4.2078264698583133e-05, "loss": 0.0226, "step": 45850 }, { "grad_norm": 0.19010807573795319, "learning_rate": 4.205785805602488e-05, "loss": 0.0209, "step": 45860 }, { "grad_norm": 0.2559988498687744, "learning_rate": 4.203745277056149e-05, "loss": 0.0223, "step": 45870 }, { "grad_norm": 0.21488621830940247, "learning_rate": 4.201704884567964e-05, "loss": 0.0203, "step": 45880 }, { "grad_norm": 0.21661654114723206, "learning_rate": 4.1996646284865816e-05, "loss": 0.0276, "step": 45890 }, { "grad_norm": 0.1997603327035904, "learning_rate": 4.197624509160625e-05, "loss": 0.022, "step": 45900 }, { "grad_norm": 0.2209995537996292, "learning_rate": 4.195584526938692e-05, "loss": 0.0184, "step": 45910 }, { "grad_norm": 0.19462764263153076, "learning_rate": 4.193544682169365e-05, "loss": 0.0187, "step": 45920 }, { "grad_norm": 0.21809633076190948, "learning_rate": 4.1915049752011946e-05, "loss": 0.0187, "step": 45930 }, { "grad_norm": 0.21248267590999603, "learning_rate": 4.189465406382712e-05, "loss": 0.0184, "step": 45940 }, { "grad_norm": 0.2856793999671936, "learning_rate": 4.187425976062422e-05, "loss": 0.0283, "step": 45950 }, { "grad_norm": 0.23924772441387177, "learning_rate": 4.18538668458881e-05, "loss": 0.0221, "step": 45960 }, { "grad_norm": 0.20340614020824432, "learning_rate": 4.183347532310333e-05, "loss": 0.0214, "step": 45970 }, { "grad_norm": 0.16560357809066772, "learning_rate": 4.181308519575429e-05, "loss": 0.022, "step": 45980 }, { "grad_norm": 0.25806793570518494, "learning_rate": 4.179269646732507e-05, "loss": 0.0311, "step": 45990 }, { "grad_norm": 0.22705715894699097, "learning_rate": 4.177230914129954e-05, "loss": 0.0235, "step": 46000 }, { "grad_norm": 0.21756383776664734, "learning_rate": 4.175192322116136e-05, "loss": 0.0239, "step": 46010 }, { "grad_norm": 0.3002835512161255, "learning_rate": 4.173153871039391e-05, "loss": 0.0198, "step": 46020 }, { "grad_norm": 0.2279147356748581, "learning_rate": 4.171115561248036e-05, "loss": 0.0245, "step": 46030 }, { "grad_norm": 0.21613679826259613, "learning_rate": 4.16907739309036e-05, "loss": 0.0222, "step": 46040 }, { "grad_norm": 0.27742716670036316, "learning_rate": 4.167039366914633e-05, "loss": 0.0213, "step": 46050 }, { "grad_norm": 0.24591508507728577, "learning_rate": 4.165001483069096e-05, "loss": 0.0224, "step": 46060 }, { "grad_norm": 0.24919934570789337, "learning_rate": 4.16296374190197e-05, "loss": 0.028, "step": 46070 }, { "grad_norm": 0.2681907117366791, "learning_rate": 4.1609261437614464e-05, "loss": 0.0266, "step": 46080 }, { "grad_norm": 0.36096784472465515, "learning_rate": 4.158888688995696e-05, "loss": 0.0288, "step": 46090 }, { "grad_norm": 0.25778380036354065, "learning_rate": 4.1568513779528645e-05, "loss": 0.0241, "step": 46100 }, { "grad_norm": 0.23109744489192963, "learning_rate": 4.1548142109810704e-05, "loss": 0.0293, "step": 46110 }, { "grad_norm": 0.49173301458358765, "learning_rate": 4.152777188428414e-05, "loss": 0.0231, "step": 46120 }, { "grad_norm": 0.2195121943950653, "learning_rate": 4.1507403106429646e-05, "loss": 0.0225, "step": 46130 }, { "grad_norm": 0.3203425705432892, "learning_rate": 4.148703577972768e-05, "loss": 0.0217, "step": 46140 }, { "grad_norm": 0.22659948468208313, "learning_rate": 4.146666990765846e-05, "loss": 0.0199, "step": 46150 }, { "grad_norm": 0.21241113543510437, "learning_rate": 4.1446305493701986e-05, "loss": 0.022, "step": 46160 }, { "grad_norm": 0.22709569334983826, "learning_rate": 4.142594254133796e-05, "loss": 0.024, "step": 46170 }, { "grad_norm": 0.33024799823760986, "learning_rate": 4.1405581054045835e-05, "loss": 0.0279, "step": 46180 }, { "grad_norm": 0.23044320940971375, "learning_rate": 4.1385221035304864e-05, "loss": 0.0207, "step": 46190 }, { "grad_norm": 0.2576715052127838, "learning_rate": 4.136486248859398e-05, "loss": 0.0241, "step": 46200 }, { "grad_norm": 0.23550832271575928, "learning_rate": 4.1344505417391955e-05, "loss": 0.0247, "step": 46210 }, { "grad_norm": 0.26181983947753906, "learning_rate": 4.132414982517721e-05, "loss": 0.0183, "step": 46220 }, { "grad_norm": 0.24856087565422058, "learning_rate": 4.130379571542798e-05, "loss": 0.0224, "step": 46230 }, { "grad_norm": 0.23847362399101257, "learning_rate": 4.12834430916222e-05, "loss": 0.022, "step": 46240 }, { "grad_norm": 0.2290397584438324, "learning_rate": 4.126309195723763e-05, "loss": 0.0233, "step": 46250 }, { "grad_norm": 0.1889280378818512, "learning_rate": 4.124274231575168e-05, "loss": 0.0263, "step": 46260 }, { "grad_norm": 0.2892962694168091, "learning_rate": 4.122239417064154e-05, "loss": 0.0204, "step": 46270 }, { "grad_norm": 0.18497319519519806, "learning_rate": 4.1202047525384184e-05, "loss": 0.0197, "step": 46280 }, { "grad_norm": 0.21050052344799042, "learning_rate": 4.118170238345627e-05, "loss": 0.0194, "step": 46290 }, { "grad_norm": 0.18021222949028015, "learning_rate": 4.1161358748334256e-05, "loss": 0.0205, "step": 46300 }, { "grad_norm": 0.23037654161453247, "learning_rate": 4.1141016623494266e-05, "loss": 0.0204, "step": 46310 }, { "grad_norm": 0.3073374629020691, "learning_rate": 4.112067601241227e-05, "loss": 0.0193, "step": 46320 }, { "grad_norm": 0.1940980851650238, "learning_rate": 4.110033691856387e-05, "loss": 0.0156, "step": 46330 }, { "grad_norm": 0.24418054521083832, "learning_rate": 4.107999934542451e-05, "loss": 0.0259, "step": 46340 }, { "grad_norm": 0.2022758424282074, "learning_rate": 4.105966329646928e-05, "loss": 0.0215, "step": 46350 }, { "grad_norm": 0.23715654015541077, "learning_rate": 4.103932877517308e-05, "loss": 0.0216, "step": 46360 }, { "grad_norm": 0.31026750802993774, "learning_rate": 4.101899578501052e-05, "loss": 0.0196, "step": 46370 }, { "grad_norm": 0.22969192266464233, "learning_rate": 4.099866432945595e-05, "loss": 0.0207, "step": 46380 }, { "grad_norm": 0.294805109500885, "learning_rate": 4.097833441198346e-05, "loss": 0.0285, "step": 46390 }, { "grad_norm": 0.18083801865577698, "learning_rate": 4.095800603606685e-05, "loss": 0.02, "step": 46400 }, { "grad_norm": 0.23442493379116058, "learning_rate": 4.093767920517975e-05, "loss": 0.0211, "step": 46410 }, { "grad_norm": 0.1727195531129837, "learning_rate": 4.091735392279539e-05, "loss": 0.0236, "step": 46420 }, { "grad_norm": 0.16458995640277863, "learning_rate": 4.089703019238685e-05, "loss": 0.0202, "step": 46430 }, { "grad_norm": 0.2037719339132309, "learning_rate": 4.0876708017426866e-05, "loss": 0.0219, "step": 46440 }, { "grad_norm": 0.1915651112794876, "learning_rate": 4.085638740138798e-05, "loss": 0.02, "step": 46450 }, { "grad_norm": 0.2698497772216797, "learning_rate": 4.08360683477424e-05, "loss": 0.0218, "step": 46460 }, { "grad_norm": 0.25689029693603516, "learning_rate": 4.0815750859962085e-05, "loss": 0.0255, "step": 46470 }, { "grad_norm": 0.3336833119392395, "learning_rate": 4.079543494151879e-05, "loss": 0.0287, "step": 46480 }, { "grad_norm": 0.21381257474422455, "learning_rate": 4.07751205958839e-05, "loss": 0.0201, "step": 46490 }, { "grad_norm": 0.2090362012386322, "learning_rate": 4.0754807826528615e-05, "loss": 0.0183, "step": 46500 }, { "grad_norm": 0.29146528244018555, "learning_rate": 4.073449663692379e-05, "loss": 0.0259, "step": 46510 }, { "grad_norm": 0.23792648315429688, "learning_rate": 4.07141870305401e-05, "loss": 0.0186, "step": 46520 }, { "grad_norm": 0.2858662009239197, "learning_rate": 4.0693879010847866e-05, "loss": 0.0374, "step": 46530 }, { "grad_norm": 0.3971719443798065, "learning_rate": 4.06735725813172e-05, "loss": 0.0234, "step": 46540 }, { "grad_norm": 0.2160727083683014, "learning_rate": 4.0653267745417903e-05, "loss": 0.0201, "step": 46550 }, { "grad_norm": 0.23922157287597656, "learning_rate": 4.063296450661949e-05, "loss": 0.0185, "step": 46560 }, { "grad_norm": 0.25035062432289124, "learning_rate": 4.061266286839128e-05, "loss": 0.0226, "step": 46570 }, { "grad_norm": 0.21222957968711853, "learning_rate": 4.0592362834202225e-05, "loss": 0.0247, "step": 46580 }, { "grad_norm": 0.2144732028245926, "learning_rate": 4.057206440752107e-05, "loss": 0.0184, "step": 46590 }, { "grad_norm": 0.22983554005622864, "learning_rate": 4.0551767591816245e-05, "loss": 0.0209, "step": 46600 }, { "grad_norm": 0.27368900179862976, "learning_rate": 4.0531472390555935e-05, "loss": 0.0256, "step": 46610 }, { "grad_norm": 0.2826428711414337, "learning_rate": 4.051117880720802e-05, "loss": 0.022, "step": 46620 }, { "grad_norm": 0.27091115713119507, "learning_rate": 4.049088684524015e-05, "loss": 0.0178, "step": 46630 }, { "grad_norm": 0.13362106680870056, "learning_rate": 4.0470596508119636e-05, "loss": 0.0175, "step": 46640 }, { "grad_norm": 0.2137472927570343, "learning_rate": 4.0450307799313524e-05, "loss": 0.0243, "step": 46650 }, { "grad_norm": 0.20649389922618866, "learning_rate": 4.0430020722288656e-05, "loss": 0.0235, "step": 46660 }, { "grad_norm": 0.18821440637111664, "learning_rate": 4.040973528051148e-05, "loss": 0.0212, "step": 46670 }, { "grad_norm": 0.20694448053836823, "learning_rate": 4.038945147744827e-05, "loss": 0.0219, "step": 46680 }, { "grad_norm": 0.23329269886016846, "learning_rate": 4.0369169316564945e-05, "loss": 0.0212, "step": 46690 }, { "grad_norm": 0.3804820477962494, "learning_rate": 4.034888880132718e-05, "loss": 0.0217, "step": 46700 }, { "grad_norm": 0.29239824414253235, "learning_rate": 4.032860993520035e-05, "loss": 0.0234, "step": 46710 }, { "grad_norm": 0.35243502259254456, "learning_rate": 4.030833272164959e-05, "loss": 0.02, "step": 46720 }, { "grad_norm": 0.2158471643924713, "learning_rate": 4.0288057164139705e-05, "loss": 0.0246, "step": 46730 }, { "grad_norm": 0.1972220093011856, "learning_rate": 4.0267783266135205e-05, "loss": 0.0192, "step": 46740 }, { "grad_norm": 0.2497343271970749, "learning_rate": 4.024751103110039e-05, "loss": 0.0243, "step": 46750 }, { "grad_norm": 0.21288934350013733, "learning_rate": 4.0227240462499176e-05, "loss": 0.025, "step": 46760 }, { "grad_norm": 0.18156032264232635, "learning_rate": 4.020697156379531e-05, "loss": 0.0206, "step": 46770 }, { "grad_norm": 0.199901282787323, "learning_rate": 4.018670433845215e-05, "loss": 0.0228, "step": 46780 }, { "grad_norm": 0.19237110018730164, "learning_rate": 4.016643878993284e-05, "loss": 0.0204, "step": 46790 }, { "grad_norm": 0.24862977862358093, "learning_rate": 4.014617492170017e-05, "loss": 0.0241, "step": 46800 }, { "grad_norm": 0.26752519607543945, "learning_rate": 4.0125912737216726e-05, "loss": 0.0358, "step": 46810 }, { "grad_norm": 0.25244879722595215, "learning_rate": 4.0105652239944735e-05, "loss": 0.0224, "step": 46820 }, { "grad_norm": 0.19911323487758636, "learning_rate": 4.0085393433346144e-05, "loss": 0.0239, "step": 46830 }, { "grad_norm": 0.18117418885231018, "learning_rate": 4.006513632088268e-05, "loss": 0.0165, "step": 46840 }, { "grad_norm": 0.168525829911232, "learning_rate": 4.004488090601567e-05, "loss": 0.0188, "step": 46850 }, { "grad_norm": 0.23072762787342072, "learning_rate": 4.002462719220626e-05, "loss": 0.0195, "step": 46860 }, { "grad_norm": 0.15533719956874847, "learning_rate": 4.000437518291522e-05, "loss": 0.0208, "step": 46870 }, { "grad_norm": 0.3811061978340149, "learning_rate": 3.9984124881603094e-05, "loss": 0.02, "step": 46880 }, { "grad_norm": 0.2054446041584015, "learning_rate": 3.9963876291730086e-05, "loss": 0.0179, "step": 46890 }, { "grad_norm": 0.19130229949951172, "learning_rate": 3.994362941675614e-05, "loss": 0.0176, "step": 46900 }, { "grad_norm": 0.16106794774532318, "learning_rate": 3.992338426014088e-05, "loss": 0.0184, "step": 46910 }, { "grad_norm": 0.22015756368637085, "learning_rate": 3.9903140825343636e-05, "loss": 0.0206, "step": 46920 }, { "grad_norm": 0.22219794988632202, "learning_rate": 3.98828991158235e-05, "loss": 0.0191, "step": 46930 }, { "grad_norm": 0.19697386026382446, "learning_rate": 3.9862659135039185e-05, "loss": 0.0235, "step": 46940 }, { "grad_norm": 0.22089555859565735, "learning_rate": 3.984242088644918e-05, "loss": 0.0203, "step": 46950 }, { "grad_norm": 0.19440379738807678, "learning_rate": 3.9822184373511615e-05, "loss": 0.0175, "step": 46960 }, { "grad_norm": 0.269436240196228, "learning_rate": 3.98019495996844e-05, "loss": 0.0223, "step": 46970 }, { "grad_norm": 0.2601858377456665, "learning_rate": 3.978171656842507e-05, "loss": 0.019, "step": 46980 }, { "grad_norm": 0.30602020025253296, "learning_rate": 3.976148528319091e-05, "loss": 0.0196, "step": 46990 }, { "grad_norm": 0.24289679527282715, "learning_rate": 3.974125574743888e-05, "loss": 0.0203, "step": 47000 }, { "grad_norm": 0.26869311928749084, "learning_rate": 3.9721027964625686e-05, "loss": 0.0224, "step": 47010 }, { "grad_norm": 0.2680859863758087, "learning_rate": 3.9700801938207676e-05, "loss": 0.0189, "step": 47020 }, { "grad_norm": 0.23310710489749908, "learning_rate": 3.9680577671640916e-05, "loss": 0.0193, "step": 47030 }, { "grad_norm": 0.17733530700206757, "learning_rate": 3.966035516838121e-05, "loss": 0.0232, "step": 47040 }, { "grad_norm": 0.17641526460647583, "learning_rate": 3.9640134431884014e-05, "loss": 0.0163, "step": 47050 }, { "grad_norm": 0.14463044703006744, "learning_rate": 3.961991546560451e-05, "loss": 0.015, "step": 47060 }, { "grad_norm": 0.2597326934337616, "learning_rate": 3.959969827299753e-05, "loss": 0.0198, "step": 47070 }, { "grad_norm": 0.2941421568393707, "learning_rate": 3.9579482857517684e-05, "loss": 0.0184, "step": 47080 }, { "grad_norm": 0.15658697485923767, "learning_rate": 3.955926922261921e-05, "loss": 0.0164, "step": 47090 }, { "grad_norm": 0.21405602991580963, "learning_rate": 3.9539057371756084e-05, "loss": 0.0201, "step": 47100 }, { "grad_norm": 0.21574577689170837, "learning_rate": 3.951884730838195e-05, "loss": 0.02, "step": 47110 }, { "grad_norm": 0.24561752378940582, "learning_rate": 3.949863903595012e-05, "loss": 0.0156, "step": 47120 }, { "grad_norm": 0.17198170721530914, "learning_rate": 3.947843255791369e-05, "loss": 0.0176, "step": 47130 }, { "grad_norm": 0.28971460461616516, "learning_rate": 3.9458227877725364e-05, "loss": 0.0213, "step": 47140 }, { "grad_norm": 0.3028586804866791, "learning_rate": 3.943802499883758e-05, "loss": 0.0183, "step": 47150 }, { "grad_norm": 0.24988658726215363, "learning_rate": 3.9417823924702437e-05, "loss": 0.0221, "step": 47160 }, { "grad_norm": 0.21951687335968018, "learning_rate": 3.939762465877178e-05, "loss": 0.0236, "step": 47170 }, { "grad_norm": 0.16432315111160278, "learning_rate": 3.937742720449708e-05, "loss": 0.0214, "step": 47180 }, { "grad_norm": 0.17358779907226562, "learning_rate": 3.9357231565329563e-05, "loss": 0.0165, "step": 47190 }, { "grad_norm": 0.13759122788906097, "learning_rate": 3.933703774472008e-05, "loss": 0.0165, "step": 47200 }, { "grad_norm": 0.2303924709558487, "learning_rate": 3.93168457461192e-05, "loss": 0.0144, "step": 47210 }, { "grad_norm": 0.2256454974412918, "learning_rate": 3.9296655572977216e-05, "loss": 0.0203, "step": 47220 }, { "grad_norm": 0.22804929316043854, "learning_rate": 3.927646722874404e-05, "loss": 0.0208, "step": 47230 }, { "grad_norm": 0.23650115728378296, "learning_rate": 3.925628071686934e-05, "loss": 0.0186, "step": 47240 }, { "grad_norm": 0.18952307105064392, "learning_rate": 3.9236096040802415e-05, "loss": 0.0188, "step": 47250 }, { "grad_norm": 0.2994770109653473, "learning_rate": 3.9215913203992294e-05, "loss": 0.0281, "step": 47260 }, { "grad_norm": 0.3381701111793518, "learning_rate": 3.9195732209887645e-05, "loss": 0.0193, "step": 47270 }, { "grad_norm": 0.21858084201812744, "learning_rate": 3.9175553061936875e-05, "loss": 0.0224, "step": 47280 }, { "grad_norm": 0.2020547240972519, "learning_rate": 3.9155375763588045e-05, "loss": 0.0253, "step": 47290 }, { "grad_norm": 0.3677554726600647, "learning_rate": 3.913520031828889e-05, "loss": 0.0162, "step": 47300 }, { "grad_norm": 0.25584229826927185, "learning_rate": 3.911502672948685e-05, "loss": 0.0224, "step": 47310 }, { "grad_norm": 0.3176930844783783, "learning_rate": 3.9094855000629014e-05, "loss": 0.019, "step": 47320 }, { "grad_norm": 0.20570828020572662, "learning_rate": 3.907468513516223e-05, "loss": 0.0162, "step": 47330 }, { "grad_norm": 0.21468383073806763, "learning_rate": 3.905451713653294e-05, "loss": 0.0155, "step": 47340 }, { "grad_norm": 0.20603854954242706, "learning_rate": 3.903435100818731e-05, "loss": 0.0187, "step": 47350 }, { "grad_norm": 0.24483807384967804, "learning_rate": 3.901418675357117e-05, "loss": 0.0188, "step": 47360 }, { "grad_norm": 0.19218911230564117, "learning_rate": 3.8994024376130075e-05, "loss": 0.0266, "step": 47370 }, { "grad_norm": 0.24533027410507202, "learning_rate": 3.8973863879309194e-05, "loss": 0.0209, "step": 47380 }, { "grad_norm": 0.18207523226737976, "learning_rate": 3.8953705266553394e-05, "loss": 0.0155, "step": 47390 }, { "grad_norm": 0.2561637759208679, "learning_rate": 3.893354854130727e-05, "loss": 0.0252, "step": 47400 }, { "grad_norm": 0.28536269068717957, "learning_rate": 3.8913393707015006e-05, "loss": 0.02, "step": 47410 }, { "grad_norm": 0.2551190257072449, "learning_rate": 3.889324076712056e-05, "loss": 0.0234, "step": 47420 }, { "grad_norm": 0.2103908509016037, "learning_rate": 3.8873089725067476e-05, "loss": 0.0219, "step": 47430 }, { "grad_norm": 0.23208051919937134, "learning_rate": 3.885294058429905e-05, "loss": 0.0271, "step": 47440 }, { "grad_norm": 0.24795739352703094, "learning_rate": 3.8832793348258206e-05, "loss": 0.0206, "step": 47450 }, { "grad_norm": 0.20732265710830688, "learning_rate": 3.881264802038756e-05, "loss": 0.0292, "step": 47460 }, { "grad_norm": 0.25541239976882935, "learning_rate": 3.879250460412939e-05, "loss": 0.0274, "step": 47470 }, { "grad_norm": 0.36541470885276794, "learning_rate": 3.8772363102925644e-05, "loss": 0.02, "step": 47480 }, { "grad_norm": 0.2121492326259613, "learning_rate": 3.875222352021798e-05, "loss": 0.0172, "step": 47490 }, { "grad_norm": 0.19074565172195435, "learning_rate": 3.8732085859447686e-05, "loss": 0.0247, "step": 47500 }, { "grad_norm": 0.20247185230255127, "learning_rate": 3.871195012405575e-05, "loss": 0.0164, "step": 47510 }, { "grad_norm": 0.2617652714252472, "learning_rate": 3.869181631748278e-05, "loss": 0.0194, "step": 47520 }, { "grad_norm": 0.23542912304401398, "learning_rate": 3.867168444316915e-05, "loss": 0.0177, "step": 47530 }, { "grad_norm": 0.21404701471328735, "learning_rate": 3.865155450455481e-05, "loss": 0.0187, "step": 47540 }, { "grad_norm": 0.2269924432039261, "learning_rate": 3.8631426505079426e-05, "loss": 0.0198, "step": 47550 }, { "grad_norm": 0.30251598358154297, "learning_rate": 3.8611300448182304e-05, "loss": 0.029, "step": 47560 }, { "grad_norm": 0.35716012120246887, "learning_rate": 3.859117633730248e-05, "loss": 0.0192, "step": 47570 }, { "grad_norm": 0.29742491245269775, "learning_rate": 3.857105417587858e-05, "loss": 0.0205, "step": 47580 }, { "grad_norm": 0.2557927072048187, "learning_rate": 3.855093396734894e-05, "loss": 0.0197, "step": 47590 }, { "grad_norm": 0.19306831061840057, "learning_rate": 3.8530815715151545e-05, "loss": 0.0302, "step": 47600 }, { "grad_norm": 0.19948741793632507, "learning_rate": 3.851069942272405e-05, "loss": 0.0174, "step": 47610 }, { "grad_norm": 0.2975043058395386, "learning_rate": 3.849058509350382e-05, "loss": 0.0212, "step": 47620 }, { "grad_norm": 0.2403826266527176, "learning_rate": 3.8470472730927783e-05, "loss": 0.0259, "step": 47630 }, { "grad_norm": 0.13783055543899536, "learning_rate": 3.845036233843264e-05, "loss": 0.0178, "step": 47640 }, { "grad_norm": 0.21584878861904144, "learning_rate": 3.843025391945469e-05, "loss": 0.0206, "step": 47650 }, { "grad_norm": 0.23007334768772125, "learning_rate": 3.841014747742992e-05, "loss": 0.0201, "step": 47660 }, { "grad_norm": 0.2309180200099945, "learning_rate": 3.839004301579397e-05, "loss": 0.0165, "step": 47670 }, { "grad_norm": 0.23461692035198212, "learning_rate": 3.8369940537982097e-05, "loss": 0.0213, "step": 47680 }, { "grad_norm": 0.21036943793296814, "learning_rate": 3.834984004742933e-05, "loss": 0.0173, "step": 47690 }, { "grad_norm": 0.20195217430591583, "learning_rate": 3.832974154757026e-05, "loss": 0.0171, "step": 47700 }, { "grad_norm": 0.24760518968105316, "learning_rate": 3.830964504183919e-05, "loss": 0.0159, "step": 47710 }, { "grad_norm": 0.29805290699005127, "learning_rate": 3.828955053367003e-05, "loss": 0.0184, "step": 47720 }, { "grad_norm": 0.23475104570388794, "learning_rate": 3.826945802649642e-05, "loss": 0.0183, "step": 47730 }, { "grad_norm": 0.2021169811487198, "learning_rate": 3.824936752375159e-05, "loss": 0.0237, "step": 47740 }, { "grad_norm": 0.24515259265899658, "learning_rate": 3.822927902886848e-05, "loss": 0.0205, "step": 47750 }, { "grad_norm": 0.2205110788345337, "learning_rate": 3.8209192545279653e-05, "loss": 0.0186, "step": 47760 }, { "grad_norm": 0.15858639776706696, "learning_rate": 3.8189108076417326e-05, "loss": 0.016, "step": 47770 }, { "grad_norm": 0.20002196729183197, "learning_rate": 3.816902562571342e-05, "loss": 0.0216, "step": 47780 }, { "grad_norm": 0.22242483496665955, "learning_rate": 3.814894519659944e-05, "loss": 0.0147, "step": 47790 }, { "grad_norm": 0.19978588819503784, "learning_rate": 3.812886679250661e-05, "loss": 0.0167, "step": 47800 }, { "grad_norm": 0.19311054050922394, "learning_rate": 3.810879041686575e-05, "loss": 0.0174, "step": 47810 }, { "grad_norm": 0.2704499661922455, "learning_rate": 3.808871607310741e-05, "loss": 0.0216, "step": 47820 }, { "grad_norm": 0.2612363398075104, "learning_rate": 3.806864376466169e-05, "loss": 0.025, "step": 47830 }, { "grad_norm": 0.11758649349212646, "learning_rate": 3.804857349495845e-05, "loss": 0.0157, "step": 47840 }, { "grad_norm": 0.2590559720993042, "learning_rate": 3.802850526742713e-05, "loss": 0.0335, "step": 47850 }, { "grad_norm": 0.2594400942325592, "learning_rate": 3.800843908549683e-05, "loss": 0.0165, "step": 47860 }, { "grad_norm": 0.1738976687192917, "learning_rate": 3.7988374952596325e-05, "loss": 0.0151, "step": 47870 }, { "grad_norm": 0.21298709511756897, "learning_rate": 3.7968312872154e-05, "loss": 0.0194, "step": 47880 }, { "grad_norm": 0.22484524548053741, "learning_rate": 3.7948252847597965e-05, "loss": 0.0196, "step": 47890 }, { "grad_norm": 0.2815242409706116, "learning_rate": 3.7928194882355885e-05, "loss": 0.0185, "step": 47900 }, { "grad_norm": 0.16900821030139923, "learning_rate": 3.790813897985515e-05, "loss": 0.0177, "step": 47910 }, { "grad_norm": 0.32336142659187317, "learning_rate": 3.7888085143522726e-05, "loss": 0.0229, "step": 47920 }, { "grad_norm": 0.2021731734275818, "learning_rate": 3.7868033376785314e-05, "loss": 0.0221, "step": 47930 }, { "grad_norm": 0.2086973786354065, "learning_rate": 3.784798368306919e-05, "loss": 0.0192, "step": 47940 }, { "grad_norm": 0.1960480511188507, "learning_rate": 3.782793606580029e-05, "loss": 0.0161, "step": 47950 }, { "grad_norm": 0.18924041092395782, "learning_rate": 3.7807890528404205e-05, "loss": 0.0218, "step": 47960 }, { "grad_norm": 0.1690511703491211, "learning_rate": 3.778784707430616e-05, "loss": 0.0181, "step": 47970 }, { "grad_norm": 0.21778051555156708, "learning_rate": 3.776780570693107e-05, "loss": 0.0199, "step": 47980 }, { "grad_norm": 0.155976340174675, "learning_rate": 3.774776642970342e-05, "loss": 0.0224, "step": 47990 }, { "grad_norm": 0.24898356199264526, "learning_rate": 3.77277292460474e-05, "loss": 0.0165, "step": 48000 }, { "grad_norm": 0.26062050461769104, "learning_rate": 3.770769415938678e-05, "loss": 0.0188, "step": 48010 }, { "grad_norm": 0.18476004898548126, "learning_rate": 3.768766117314506e-05, "loss": 0.0247, "step": 48020 }, { "grad_norm": 0.22088730335235596, "learning_rate": 3.76676302907453e-05, "loss": 0.0178, "step": 48030 }, { "grad_norm": 0.2088860720396042, "learning_rate": 3.764760151561021e-05, "loss": 0.0205, "step": 48040 }, { "grad_norm": 0.19155487418174744, "learning_rate": 3.76275748511622e-05, "loss": 0.0172, "step": 48050 }, { "grad_norm": 0.15297581255435944, "learning_rate": 3.7607550300823255e-05, "loss": 0.0227, "step": 48060 }, { "grad_norm": 0.1984419822692871, "learning_rate": 3.7587527868015044e-05, "loss": 0.0295, "step": 48070 }, { "grad_norm": 0.28322431445121765, "learning_rate": 3.756750755615881e-05, "loss": 0.0217, "step": 48080 }, { "grad_norm": 0.17035171389579773, "learning_rate": 3.754748936867553e-05, "loss": 0.0224, "step": 48090 }, { "grad_norm": 0.2745186686515808, "learning_rate": 3.752747330898573e-05, "loss": 0.0202, "step": 48100 }, { "grad_norm": 0.2158540040254593, "learning_rate": 3.750745938050962e-05, "loss": 0.0207, "step": 48110 }, { "grad_norm": 0.21669553220272064, "learning_rate": 3.7487447586667025e-05, "loss": 0.0237, "step": 48120 }, { "grad_norm": 0.2546209692955017, "learning_rate": 3.7467437930877426e-05, "loss": 0.0165, "step": 48130 }, { "grad_norm": 0.2120831459760666, "learning_rate": 3.744743041655992e-05, "loss": 0.0196, "step": 48140 }, { "grad_norm": 0.230891615152359, "learning_rate": 3.742742504713324e-05, "loss": 0.0191, "step": 48150 }, { "grad_norm": 0.1648813635110855, "learning_rate": 3.740742182601576e-05, "loss": 0.0215, "step": 48160 }, { "grad_norm": 0.2784195840358734, "learning_rate": 3.7387420756625464e-05, "loss": 0.0246, "step": 48170 }, { "grad_norm": 0.18349583446979523, "learning_rate": 3.736742184238002e-05, "loss": 0.0207, "step": 48180 }, { "grad_norm": 0.19203756749629974, "learning_rate": 3.7347425086696684e-05, "loss": 0.0218, "step": 48190 }, { "grad_norm": 0.308791846036911, "learning_rate": 3.732743049299235e-05, "loss": 0.023, "step": 48200 }, { "grad_norm": 0.18005816638469696, "learning_rate": 3.730743806468354e-05, "loss": 0.0236, "step": 48210 }, { "grad_norm": 0.24289825558662415, "learning_rate": 3.7287447805186436e-05, "loss": 0.0176, "step": 48220 }, { "grad_norm": 0.21974511444568634, "learning_rate": 3.726745971791682e-05, "loss": 0.0231, "step": 48230 }, { "grad_norm": 0.24944008886814117, "learning_rate": 3.724747380629008e-05, "loss": 0.0282, "step": 48240 }, { "grad_norm": 0.2008690983057022, "learning_rate": 3.72274900737213e-05, "loss": 0.021, "step": 48250 }, { "grad_norm": 0.18081800639629364, "learning_rate": 3.7207508523625123e-05, "loss": 0.0192, "step": 48260 }, { "grad_norm": 0.19201812148094177, "learning_rate": 3.718752915941588e-05, "loss": 0.0191, "step": 48270 }, { "grad_norm": 0.2800067663192749, "learning_rate": 3.7167551984507464e-05, "loss": 0.0185, "step": 48280 }, { "grad_norm": 0.18294274806976318, "learning_rate": 3.714757700231346e-05, "loss": 0.0191, "step": 48290 }, { "grad_norm": 0.18737280368804932, "learning_rate": 3.712760421624703e-05, "loss": 0.0171, "step": 48300 }, { "grad_norm": 0.21216993033885956, "learning_rate": 3.710763362972099e-05, "loss": 0.0191, "step": 48310 }, { "grad_norm": 0.16782991588115692, "learning_rate": 3.708766524614774e-05, "loss": 0.0243, "step": 48320 }, { "grad_norm": 0.16531746089458466, "learning_rate": 3.7067699068939335e-05, "loss": 0.0183, "step": 48330 }, { "grad_norm": 0.2070423811674118, "learning_rate": 3.704773510150748e-05, "loss": 0.0204, "step": 48340 }, { "grad_norm": 0.2219659835100174, "learning_rate": 3.702777334726344e-05, "loss": 0.0236, "step": 48350 }, { "grad_norm": 0.16890014708042145, "learning_rate": 3.7007813809618164e-05, "loss": 0.0141, "step": 48360 }, { "grad_norm": 0.1807698905467987, "learning_rate": 3.6987856491982145e-05, "loss": 0.0241, "step": 48370 }, { "grad_norm": 0.26143133640289307, "learning_rate": 3.696790139776558e-05, "loss": 0.0221, "step": 48380 }, { "grad_norm": 0.23961903154850006, "learning_rate": 3.6947948530378235e-05, "loss": 0.0247, "step": 48390 }, { "grad_norm": 0.19568577408790588, "learning_rate": 3.6927997893229516e-05, "loss": 0.0204, "step": 48400 }, { "grad_norm": 0.23860347270965576, "learning_rate": 3.690804948972845e-05, "loss": 0.0174, "step": 48410 }, { "grad_norm": 0.17036966979503632, "learning_rate": 3.688810332328363e-05, "loss": 0.0204, "step": 48420 }, { "grad_norm": 0.19815151393413544, "learning_rate": 3.686815939730336e-05, "loss": 0.0298, "step": 48430 }, { "grad_norm": 0.23573115468025208, "learning_rate": 3.684821771519548e-05, "loss": 0.0196, "step": 48440 }, { "grad_norm": 0.2064853459596634, "learning_rate": 3.68282782803675e-05, "loss": 0.0197, "step": 48450 }, { "grad_norm": 0.22736512124538422, "learning_rate": 3.6808341096226504e-05, "loss": 0.0223, "step": 48460 }, { "grad_norm": 0.2680538296699524, "learning_rate": 3.678840616617924e-05, "loss": 0.0225, "step": 48470 }, { "grad_norm": 0.2941966652870178, "learning_rate": 3.6768473493632e-05, "loss": 0.0208, "step": 48480 }, { "grad_norm": 0.23379667103290558, "learning_rate": 3.6748543081990783e-05, "loss": 0.0223, "step": 48490 }, { "grad_norm": 0.23463965952396393, "learning_rate": 3.672861493466112e-05, "loss": 0.0193, "step": 48500 }, { "grad_norm": 0.25783905386924744, "learning_rate": 3.670868905504818e-05, "loss": 0.0206, "step": 48510 }, { "grad_norm": 0.2446351945400238, "learning_rate": 3.6688765446556784e-05, "loss": 0.0209, "step": 48520 }, { "grad_norm": 0.21967075765132904, "learning_rate": 3.6668844112591276e-05, "loss": 0.0243, "step": 48530 }, { "grad_norm": 0.25409621000289917, "learning_rate": 3.664892505655573e-05, "loss": 0.0195, "step": 48540 }, { "grad_norm": 0.21028892695903778, "learning_rate": 3.662900828185373e-05, "loss": 0.0212, "step": 48550 }, { "grad_norm": 0.17486202716827393, "learning_rate": 3.6609093791888516e-05, "loss": 0.0181, "step": 48560 }, { "grad_norm": 0.22554412484169006, "learning_rate": 3.658918159006292e-05, "loss": 0.0252, "step": 48570 }, { "grad_norm": 0.22753262519836426, "learning_rate": 3.656927167977942e-05, "loss": 0.0233, "step": 48580 }, { "grad_norm": 0.21952007710933685, "learning_rate": 3.654936406444006e-05, "loss": 0.0184, "step": 48590 }, { "grad_norm": 0.2115437090396881, "learning_rate": 3.65294587474465e-05, "loss": 0.0195, "step": 48600 }, { "grad_norm": 0.22305358946323395, "learning_rate": 3.650955573220002e-05, "loss": 0.0459, "step": 48610 }, { "grad_norm": 0.3077966570854187, "learning_rate": 3.648965502210149e-05, "loss": 0.0201, "step": 48620 }, { "grad_norm": 0.21495915949344635, "learning_rate": 3.646975662055142e-05, "loss": 0.0191, "step": 48630 }, { "grad_norm": 0.1453429013490677, "learning_rate": 3.644986053094987e-05, "loss": 0.0199, "step": 48640 }, { "grad_norm": 0.23928578197956085, "learning_rate": 3.642996675669659e-05, "loss": 0.024, "step": 48650 }, { "grad_norm": 0.16847778856754303, "learning_rate": 3.641007530119083e-05, "loss": 0.0198, "step": 48660 }, { "grad_norm": 0.269185870885849, "learning_rate": 3.639018616783153e-05, "loss": 0.0228, "step": 48670 }, { "grad_norm": 0.2555221617221832, "learning_rate": 3.637029936001719e-05, "loss": 0.0186, "step": 48680 }, { "grad_norm": 0.21294251084327698, "learning_rate": 3.6350414881145886e-05, "loss": 0.0234, "step": 48690 }, { "grad_norm": 0.3935580849647522, "learning_rate": 3.6330532734615386e-05, "loss": 0.017, "step": 48700 }, { "grad_norm": 0.37699732184410095, "learning_rate": 3.6310652923822975e-05, "loss": 0.0263, "step": 48710 }, { "grad_norm": 0.1506902575492859, "learning_rate": 3.629077545216558e-05, "loss": 0.0204, "step": 48720 }, { "grad_norm": 0.2347857803106308, "learning_rate": 3.627090032303969e-05, "loss": 0.0225, "step": 48730 }, { "grad_norm": 0.17741625010967255, "learning_rate": 3.625102753984146e-05, "loss": 0.0241, "step": 48740 }, { "grad_norm": 0.21140891313552856, "learning_rate": 3.623115710596659e-05, "loss": 0.0196, "step": 48750 }, { "grad_norm": 0.2514093816280365, "learning_rate": 3.6211289024810395e-05, "loss": 0.0229, "step": 48760 }, { "grad_norm": 0.20867490768432617, "learning_rate": 3.619142329976777e-05, "loss": 0.0182, "step": 48770 }, { "grad_norm": 0.18379583954811096, "learning_rate": 3.6171559934233247e-05, "loss": 0.0173, "step": 48780 }, { "grad_norm": 0.27242130041122437, "learning_rate": 3.615169893160093e-05, "loss": 0.0197, "step": 48790 }, { "grad_norm": 0.19637207686901093, "learning_rate": 3.61318402952645e-05, "loss": 0.0189, "step": 48800 }, { "grad_norm": 0.29237890243530273, "learning_rate": 3.6111984028617285e-05, "loss": 0.0225, "step": 48810 }, { "grad_norm": 0.21349671483039856, "learning_rate": 3.6092130135052134e-05, "loss": 0.0196, "step": 48820 }, { "grad_norm": 0.2609817683696747, "learning_rate": 3.6072278617961584e-05, "loss": 0.0249, "step": 48830 }, { "grad_norm": 0.23697073757648468, "learning_rate": 3.605242948073767e-05, "loss": 0.0231, "step": 48840 }, { "grad_norm": 0.21959026157855988, "learning_rate": 3.603258272677212e-05, "loss": 0.021, "step": 48850 }, { "grad_norm": 0.1996917724609375, "learning_rate": 3.601273835945616e-05, "loss": 0.0194, "step": 48860 }, { "grad_norm": 0.22810807824134827, "learning_rate": 3.5992896382180664e-05, "loss": 0.0208, "step": 48870 }, { "grad_norm": 0.18959785997867584, "learning_rate": 3.597305679833609e-05, "loss": 0.0252, "step": 48880 }, { "grad_norm": 0.19575147330760956, "learning_rate": 3.595321961131245e-05, "loss": 0.0184, "step": 48890 }, { "grad_norm": 0.22364908456802368, "learning_rate": 3.593338482449942e-05, "loss": 0.0256, "step": 48900 }, { "grad_norm": 0.20803987979888916, "learning_rate": 3.591355244128618e-05, "loss": 0.0313, "step": 48910 }, { "grad_norm": 0.2233557105064392, "learning_rate": 3.589372246506158e-05, "loss": 0.0265, "step": 48920 }, { "grad_norm": 0.1659872978925705, "learning_rate": 3.5873894899213984e-05, "loss": 0.0209, "step": 48930 }, { "grad_norm": 0.2035577893257141, "learning_rate": 3.5854069747131416e-05, "loss": 0.0213, "step": 48940 }, { "grad_norm": 0.16903328895568848, "learning_rate": 3.583424701220143e-05, "loss": 0.0223, "step": 48950 }, { "grad_norm": 0.2803443670272827, "learning_rate": 3.581442669781121e-05, "loss": 0.0222, "step": 48960 }, { "grad_norm": 0.24445252120494843, "learning_rate": 3.579460880734749e-05, "loss": 0.0231, "step": 48970 }, { "grad_norm": 0.19821959733963013, "learning_rate": 3.577479334419657e-05, "loss": 0.0207, "step": 48980 }, { "grad_norm": 0.1795222908258438, "learning_rate": 3.575498031174444e-05, "loss": 0.0207, "step": 48990 }, { "grad_norm": 0.22849874198436737, "learning_rate": 3.573516971337657e-05, "loss": 0.022, "step": 49000 }, { "grad_norm": 0.2965744137763977, "learning_rate": 3.5715361552478046e-05, "loss": 0.0292, "step": 49010 }, { "grad_norm": 0.46257492899894714, "learning_rate": 3.5695555832433536e-05, "loss": 0.0191, "step": 49020 }, { "grad_norm": 0.29288777709007263, "learning_rate": 3.5675752556627325e-05, "loss": 0.0207, "step": 49030 }, { "grad_norm": 0.21494147181510925, "learning_rate": 3.565595172844322e-05, "loss": 0.0201, "step": 49040 }, { "grad_norm": 0.3139961361885071, "learning_rate": 3.5636153351264666e-05, "loss": 0.0237, "step": 49050 }, { "grad_norm": 0.29292556643486023, "learning_rate": 3.5616357428474655e-05, "loss": 0.019, "step": 49060 }, { "grad_norm": 0.17074547708034515, "learning_rate": 3.559656396345575e-05, "loss": 0.0227, "step": 49070 }, { "grad_norm": 0.1753588616847992, "learning_rate": 3.5576772959590146e-05, "loss": 0.0223, "step": 49080 }, { "grad_norm": 0.2066275179386139, "learning_rate": 3.5556984420259545e-05, "loss": 0.0266, "step": 49090 }, { "grad_norm": 0.20636895298957825, "learning_rate": 3.5537198348845305e-05, "loss": 0.0185, "step": 49100 }, { "grad_norm": 0.18864569067955017, "learning_rate": 3.551741474872831e-05, "loss": 0.0224, "step": 49110 }, { "grad_norm": 0.26760995388031006, "learning_rate": 3.549763362328903e-05, "loss": 0.0233, "step": 49120 }, { "grad_norm": 0.37311506271362305, "learning_rate": 3.5477854975907515e-05, "loss": 0.0227, "step": 49130 }, { "grad_norm": 0.14592424035072327, "learning_rate": 3.5458078809963416e-05, "loss": 0.0218, "step": 49140 }, { "grad_norm": 0.23954279720783234, "learning_rate": 3.543830512883594e-05, "loss": 0.0202, "step": 49150 }, { "grad_norm": 0.2368459403514862, "learning_rate": 3.5418533935903824e-05, "loss": 0.0213, "step": 49160 }, { "grad_norm": 0.23769065737724304, "learning_rate": 3.539876523454547e-05, "loss": 0.0224, "step": 49170 }, { "grad_norm": 0.42764076590538025, "learning_rate": 3.537899902813878e-05, "loss": 0.0227, "step": 49180 }, { "grad_norm": 0.1965939849615097, "learning_rate": 3.5359235320061293e-05, "loss": 0.0192, "step": 49190 }, { "grad_norm": 0.2050393521785736, "learning_rate": 3.533947411369003e-05, "loss": 0.0174, "step": 49200 }, { "grad_norm": 0.2556493282318115, "learning_rate": 3.53197154124017e-05, "loss": 0.0228, "step": 49210 }, { "grad_norm": 0.1546536087989807, "learning_rate": 3.52999592195725e-05, "loss": 0.0209, "step": 49220 }, { "grad_norm": 0.20004163682460785, "learning_rate": 3.5280205538578224e-05, "loss": 0.0192, "step": 49230 }, { "grad_norm": 0.29225796461105347, "learning_rate": 3.5260454372794236e-05, "loss": 0.0211, "step": 49240 }, { "grad_norm": 0.20444372296333313, "learning_rate": 3.524070572559545e-05, "loss": 0.0193, "step": 49250 }, { "grad_norm": 0.2465996891260147, "learning_rate": 3.5220959600356395e-05, "loss": 0.0212, "step": 49260 }, { "grad_norm": 0.17652545869350433, "learning_rate": 3.5201216000451145e-05, "loss": 0.0213, "step": 49270 }, { "grad_norm": 0.1687338650226593, "learning_rate": 3.5181474929253335e-05, "loss": 0.0198, "step": 49280 }, { "grad_norm": 0.23643457889556885, "learning_rate": 3.516173639013615e-05, "loss": 0.0203, "step": 49290 }, { "grad_norm": 0.11174808442592621, "learning_rate": 3.5142000386472406e-05, "loss": 0.0209, "step": 49300 }, { "grad_norm": 0.21345220506191254, "learning_rate": 3.5122266921634427e-05, "loss": 0.0175, "step": 49310 }, { "grad_norm": 0.2557864189147949, "learning_rate": 3.510253599899413e-05, "loss": 0.0207, "step": 49320 }, { "grad_norm": 0.21953804790973663, "learning_rate": 3.5082807621922965e-05, "loss": 0.017, "step": 49330 }, { "grad_norm": 0.2477736473083496, "learning_rate": 3.506308179379201e-05, "loss": 0.0212, "step": 49340 }, { "grad_norm": 0.19143033027648926, "learning_rate": 3.5043358517971844e-05, "loss": 0.0198, "step": 49350 }, { "grad_norm": 0.18943411111831665, "learning_rate": 3.502363779783264e-05, "loss": 0.0173, "step": 49360 }, { "grad_norm": 0.23086726665496826, "learning_rate": 3.500391963674415e-05, "loss": 0.0172, "step": 49370 }, { "grad_norm": 0.2579137980937958, "learning_rate": 3.4984204038075615e-05, "loss": 0.0246, "step": 49380 }, { "grad_norm": 0.38383445143699646, "learning_rate": 3.496449100519595e-05, "loss": 0.0232, "step": 49390 }, { "grad_norm": 0.16589032113552094, "learning_rate": 3.494478054147354e-05, "loss": 0.0202, "step": 49400 }, { "grad_norm": 0.16550500690937042, "learning_rate": 3.4925072650276395e-05, "loss": 0.0175, "step": 49410 }, { "grad_norm": 0.22109156847000122, "learning_rate": 3.4905367334972016e-05, "loss": 0.0181, "step": 49420 }, { "grad_norm": 0.24810342490673065, "learning_rate": 3.488566459892752e-05, "loss": 0.0168, "step": 49430 }, { "grad_norm": 0.2512975335121155, "learning_rate": 3.4865964445509585e-05, "loss": 0.0174, "step": 49440 }, { "grad_norm": 0.20537713170051575, "learning_rate": 3.484626687808438e-05, "loss": 0.0177, "step": 49450 }, { "grad_norm": 0.1735881268978119, "learning_rate": 3.4826571900017735e-05, "loss": 0.0156, "step": 49460 }, { "grad_norm": 0.21860764920711517, "learning_rate": 3.480687951467495e-05, "loss": 0.0161, "step": 49470 }, { "grad_norm": 0.21581946313381195, "learning_rate": 3.4787189725420925e-05, "loss": 0.0204, "step": 49480 }, { "grad_norm": 0.23698626458644867, "learning_rate": 3.4767502535620086e-05, "loss": 0.0162, "step": 49490 }, { "grad_norm": 0.2113831639289856, "learning_rate": 3.474781794863648e-05, "loss": 0.0253, "step": 49500 }, { "grad_norm": 0.2353714257478714, "learning_rate": 3.472813596783363e-05, "loss": 0.0239, "step": 49510 }, { "grad_norm": 0.20439940690994263, "learning_rate": 3.470845659657466e-05, "loss": 0.0241, "step": 49520 }, { "grad_norm": 0.17545658349990845, "learning_rate": 3.468877983822223e-05, "loss": 0.0216, "step": 49530 }, { "grad_norm": 0.16657161712646484, "learning_rate": 3.466910569613855e-05, "loss": 0.0191, "step": 49540 }, { "grad_norm": 0.22173592448234558, "learning_rate": 3.464943417368542e-05, "loss": 0.0217, "step": 49550 }, { "grad_norm": 0.21844403445720673, "learning_rate": 3.462976527422415e-05, "loss": 0.018, "step": 49560 }, { "grad_norm": 0.2357415407896042, "learning_rate": 3.461009900111562e-05, "loss": 0.0207, "step": 49570 }, { "grad_norm": 0.21675054728984833, "learning_rate": 3.459043535772023e-05, "loss": 0.0237, "step": 49580 }, { "grad_norm": 0.21046583354473114, "learning_rate": 3.4570774347398014e-05, "loss": 0.0236, "step": 49590 }, { "grad_norm": 0.2335464209318161, "learning_rate": 3.4551115973508454e-05, "loss": 0.0235, "step": 49600 }, { "grad_norm": 0.36494430899620056, "learning_rate": 3.453146023941066e-05, "loss": 0.0243, "step": 49610 }, { "grad_norm": 0.22415439784526825, "learning_rate": 3.451180714846325e-05, "loss": 0.0226, "step": 49620 }, { "grad_norm": 0.17588357627391815, "learning_rate": 3.449215670402438e-05, "loss": 0.021, "step": 49630 }, { "grad_norm": 0.19933748245239258, "learning_rate": 3.447250890945181e-05, "loss": 0.0204, "step": 49640 }, { "grad_norm": 0.22929202020168304, "learning_rate": 3.4452863768102754e-05, "loss": 0.0235, "step": 49650 }, { "grad_norm": 0.2693992853164673, "learning_rate": 3.443322128333409e-05, "loss": 0.0204, "step": 49660 }, { "grad_norm": 0.27027812600135803, "learning_rate": 3.441358145850215e-05, "loss": 0.0245, "step": 49670 }, { "grad_norm": 0.20961570739746094, "learning_rate": 3.439394429696286e-05, "loss": 0.0186, "step": 49680 }, { "grad_norm": 0.17254118621349335, "learning_rate": 3.4374309802071644e-05, "loss": 0.0205, "step": 49690 }, { "grad_norm": 0.2028156965970993, "learning_rate": 3.435467797718353e-05, "loss": 0.0216, "step": 49700 }, { "grad_norm": 0.2316363900899887, "learning_rate": 3.433504882565306e-05, "loss": 0.0195, "step": 49710 }, { "grad_norm": 0.23073206841945648, "learning_rate": 3.43154223508343e-05, "loss": 0.0228, "step": 49720 }, { "grad_norm": 0.17081716656684875, "learning_rate": 3.429579855608089e-05, "loss": 0.0205, "step": 49730 }, { "grad_norm": 0.18316353857517242, "learning_rate": 3.427617744474597e-05, "loss": 0.0222, "step": 49740 }, { "grad_norm": 0.1688084900379181, "learning_rate": 3.425655902018231e-05, "loss": 0.0211, "step": 49750 }, { "grad_norm": 0.19481535255908966, "learning_rate": 3.423694328574211e-05, "loss": 0.0201, "step": 49760 }, { "grad_norm": 0.19555658102035522, "learning_rate": 3.42173302447772e-05, "loss": 0.0167, "step": 49770 }, { "grad_norm": 0.21514497697353363, "learning_rate": 3.419771990063886e-05, "loss": 0.0176, "step": 49780 }, { "grad_norm": 0.173617884516716, "learning_rate": 3.417811225667803e-05, "loss": 0.0184, "step": 49790 }, { "grad_norm": 0.1845203936100006, "learning_rate": 3.415850731624508e-05, "loss": 0.0197, "step": 49800 }, { "grad_norm": 0.222451850771904, "learning_rate": 3.4138905082689945e-05, "loss": 0.0212, "step": 49810 }, { "grad_norm": 0.31105107069015503, "learning_rate": 3.4119305559362145e-05, "loss": 0.0235, "step": 49820 }, { "grad_norm": 0.20054727792739868, "learning_rate": 3.4099708749610684e-05, "loss": 0.0193, "step": 49830 }, { "grad_norm": 0.26480963826179504, "learning_rate": 3.408011465678413e-05, "loss": 0.0225, "step": 49840 }, { "grad_norm": 0.21645233035087585, "learning_rate": 3.406052328423055e-05, "loss": 0.025, "step": 49850 }, { "grad_norm": 0.2560725808143616, "learning_rate": 3.4040934635297615e-05, "loss": 0.0202, "step": 49860 }, { "grad_norm": 0.2260015904903412, "learning_rate": 3.4021348713332466e-05, "loss": 0.0159, "step": 49870 }, { "grad_norm": 0.21276316046714783, "learning_rate": 3.4001765521681807e-05, "loss": 0.0203, "step": 49880 }, { "grad_norm": 0.2293197065591812, "learning_rate": 3.398218506369188e-05, "loss": 0.0309, "step": 49890 }, { "grad_norm": 0.18298786878585815, "learning_rate": 3.3962607342708404e-05, "loss": 0.0196, "step": 49900 }, { "grad_norm": 0.20563724637031555, "learning_rate": 3.394303236207673e-05, "loss": 0.0197, "step": 49910 }, { "grad_norm": 0.21904392540454865, "learning_rate": 3.392346012514166e-05, "loss": 0.0219, "step": 49920 }, { "grad_norm": 0.2263067364692688, "learning_rate": 3.390389063524757e-05, "loss": 0.019, "step": 49930 }, { "grad_norm": 0.1800534576177597, "learning_rate": 3.3884323895738324e-05, "loss": 0.0159, "step": 49940 }, { "grad_norm": 0.2992246747016907, "learning_rate": 3.386475990995738e-05, "loss": 0.021, "step": 49950 }, { "grad_norm": 0.20960405468940735, "learning_rate": 3.384519868124765e-05, "loss": 0.0191, "step": 49960 }, { "grad_norm": 0.15351687371730804, "learning_rate": 3.3825640212951645e-05, "loss": 0.0162, "step": 49970 }, { "grad_norm": 0.2622966766357422, "learning_rate": 3.380608450841134e-05, "loss": 0.019, "step": 49980 }, { "grad_norm": 0.24825280904769897, "learning_rate": 3.3786531570968305e-05, "loss": 0.0184, "step": 49990 }, { "grad_norm": 0.2799130976200104, "learning_rate": 3.3766981403963584e-05, "loss": 0.0223, "step": 50000 }, { "grad_norm": 0.29496100544929504, "learning_rate": 3.374743401073775e-05, "loss": 0.021, "step": 50010 }, { "grad_norm": 0.2979333996772766, "learning_rate": 3.372788939463095e-05, "loss": 0.0158, "step": 50020 }, { "grad_norm": 0.2345207780599594, "learning_rate": 3.370834755898281e-05, "loss": 0.0181, "step": 50030 }, { "grad_norm": 0.29031023383140564, "learning_rate": 3.3688808507132493e-05, "loss": 0.0212, "step": 50040 }, { "grad_norm": 0.19546283781528473, "learning_rate": 3.3669272242418685e-05, "loss": 0.0208, "step": 50050 }, { "grad_norm": 0.39548856019973755, "learning_rate": 3.364973876817961e-05, "loss": 0.0218, "step": 50060 }, { "grad_norm": 0.2743503451347351, "learning_rate": 3.363020808775299e-05, "loss": 0.0176, "step": 50070 }, { "grad_norm": 0.28548845648765564, "learning_rate": 3.361068020447611e-05, "loss": 0.0208, "step": 50080 }, { "grad_norm": 0.19841772317886353, "learning_rate": 3.3591155121685724e-05, "loss": 0.0169, "step": 50090 }, { "grad_norm": 0.18143044412136078, "learning_rate": 3.357163284271814e-05, "loss": 0.0256, "step": 50100 }, { "grad_norm": 0.16934746503829956, "learning_rate": 3.355211337090919e-05, "loss": 0.0183, "step": 50110 }, { "grad_norm": 0.2105877846479416, "learning_rate": 3.353259670959421e-05, "loss": 0.0216, "step": 50120 }, { "grad_norm": 0.18882213532924652, "learning_rate": 3.351308286210808e-05, "loss": 0.0175, "step": 50130 }, { "grad_norm": 0.27831777930259705, "learning_rate": 3.3493571831785156e-05, "loss": 0.0197, "step": 50140 }, { "grad_norm": 0.1934499442577362, "learning_rate": 3.347406362195936e-05, "loss": 0.018, "step": 50150 }, { "grad_norm": 0.2563333511352539, "learning_rate": 3.345455823596411e-05, "loss": 0.0201, "step": 50160 }, { "grad_norm": 0.15826009213924408, "learning_rate": 3.3435055677132346e-05, "loss": 0.019, "step": 50170 }, { "grad_norm": 0.23463256657123566, "learning_rate": 3.3415555948796505e-05, "loss": 0.0198, "step": 50180 }, { "grad_norm": 0.2743038833141327, "learning_rate": 3.3396059054288556e-05, "loss": 0.0169, "step": 50190 }, { "grad_norm": 0.19006840884685516, "learning_rate": 3.3376564996940015e-05, "loss": 0.0278, "step": 50200 }, { "grad_norm": 0.1785580813884735, "learning_rate": 3.3357073780081836e-05, "loss": 0.0181, "step": 50210 }, { "grad_norm": 0.17141452431678772, "learning_rate": 3.333758540704459e-05, "loss": 0.0171, "step": 50220 }, { "grad_norm": 0.18854404985904694, "learning_rate": 3.331809988115827e-05, "loss": 0.0174, "step": 50230 }, { "grad_norm": 0.29810935258865356, "learning_rate": 3.329861720575244e-05, "loss": 0.0198, "step": 50240 }, { "grad_norm": 0.24957600235939026, "learning_rate": 3.3279137384156126e-05, "loss": 0.0144, "step": 50250 }, { "grad_norm": 0.1981736123561859, "learning_rate": 3.3259660419697934e-05, "loss": 0.0224, "step": 50260 }, { "grad_norm": 0.14662054181098938, "learning_rate": 3.3240186315705926e-05, "loss": 0.0232, "step": 50270 }, { "grad_norm": 0.2945895195007324, "learning_rate": 3.322071507550769e-05, "loss": 0.0254, "step": 50280 }, { "grad_norm": 0.22416432201862335, "learning_rate": 3.320124670243033e-05, "loss": 0.0199, "step": 50290 }, { "grad_norm": 0.1957649290561676, "learning_rate": 3.318178119980045e-05, "loss": 0.0179, "step": 50300 }, { "grad_norm": 0.25152304768562317, "learning_rate": 3.316231857094421e-05, "loss": 0.0185, "step": 50310 }, { "grad_norm": 0.26882055401802063, "learning_rate": 3.31428588191872e-05, "loss": 0.02, "step": 50320 }, { "grad_norm": 0.26677268743515015, "learning_rate": 3.312340194785458e-05, "loss": 0.0204, "step": 50330 }, { "grad_norm": 0.27174603939056396, "learning_rate": 3.3103947960270975e-05, "loss": 0.0177, "step": 50340 }, { "grad_norm": 0.15765583515167236, "learning_rate": 3.308449685976058e-05, "loss": 0.014, "step": 50350 }, { "grad_norm": 0.24107177555561066, "learning_rate": 3.3065048649647024e-05, "loss": 0.0207, "step": 50360 }, { "grad_norm": 0.21650934219360352, "learning_rate": 3.304560333325348e-05, "loss": 0.0177, "step": 50370 }, { "grad_norm": 0.22741955518722534, "learning_rate": 3.3026160913902624e-05, "loss": 0.022, "step": 50380 }, { "grad_norm": 0.29561498761177063, "learning_rate": 3.300672139491662e-05, "loss": 0.0218, "step": 50390 }, { "grad_norm": 0.15976367890834808, "learning_rate": 3.298728477961717e-05, "loss": 0.0182, "step": 50400 }, { "grad_norm": 0.20145221054553986, "learning_rate": 3.296785107132544e-05, "loss": 0.0233, "step": 50410 }, { "grad_norm": 0.19161592423915863, "learning_rate": 3.2948420273362156e-05, "loss": 0.0151, "step": 50420 }, { "grad_norm": 0.1693459302186966, "learning_rate": 3.292899238904747e-05, "loss": 0.0193, "step": 50430 }, { "grad_norm": 0.2434946894645691, "learning_rate": 3.29095674217011e-05, "loss": 0.0207, "step": 50440 }, { "grad_norm": 0.2767046093940735, "learning_rate": 3.289014537464224e-05, "loss": 0.0222, "step": 50450 }, { "grad_norm": 0.18280377984046936, "learning_rate": 3.287072625118955e-05, "loss": 0.0216, "step": 50460 }, { "grad_norm": 0.280787855386734, "learning_rate": 3.285131005466129e-05, "loss": 0.0176, "step": 50470 }, { "grad_norm": 0.2147223800420761, "learning_rate": 3.2831896788375105e-05, "loss": 0.0162, "step": 50480 }, { "grad_norm": 0.16935448348522186, "learning_rate": 3.281248645564822e-05, "loss": 0.0164, "step": 50490 }, { "grad_norm": 0.2224137783050537, "learning_rate": 3.2793079059797306e-05, "loss": 0.0209, "step": 50500 }, { "grad_norm": 0.21151544153690338, "learning_rate": 3.277367460413859e-05, "loss": 0.0218, "step": 50510 }, { "grad_norm": 0.21735148131847382, "learning_rate": 3.275427309198773e-05, "loss": 0.026, "step": 50520 }, { "grad_norm": 0.32703161239624023, "learning_rate": 3.273487452665993e-05, "loss": 0.0193, "step": 50530 }, { "grad_norm": 0.2875167429447174, "learning_rate": 3.271547891146986e-05, "loss": 0.0226, "step": 50540 }, { "grad_norm": 0.2632473409175873, "learning_rate": 3.269608624973173e-05, "loss": 0.0165, "step": 50550 }, { "grad_norm": 0.20589418709278107, "learning_rate": 3.26766965447592e-05, "loss": 0.0215, "step": 50560 }, { "grad_norm": 0.2140364944934845, "learning_rate": 3.2657309799865424e-05, "loss": 0.0227, "step": 50570 }, { "grad_norm": 0.2013484537601471, "learning_rate": 3.2637926018363084e-05, "loss": 0.0179, "step": 50580 }, { "grad_norm": 0.2414168417453766, "learning_rate": 3.261854520356432e-05, "loss": 0.0227, "step": 50590 }, { "grad_norm": 0.1518883854150772, "learning_rate": 3.2599167358780816e-05, "loss": 0.0186, "step": 50600 }, { "grad_norm": 0.26301413774490356, "learning_rate": 3.2579792487323676e-05, "loss": 0.0186, "step": 50610 }, { "grad_norm": 0.30988267064094543, "learning_rate": 3.256042059250358e-05, "loss": 0.023, "step": 50620 }, { "grad_norm": 0.31308630108833313, "learning_rate": 3.254105167763062e-05, "loss": 0.0269, "step": 50630 }, { "grad_norm": 0.23103725910186768, "learning_rate": 3.252168574601443e-05, "loss": 0.0218, "step": 50640 }, { "grad_norm": 0.16750799119472504, "learning_rate": 3.2502322800964115e-05, "loss": 0.0198, "step": 50650 }, { "grad_norm": 0.2220226377248764, "learning_rate": 3.248296284578825e-05, "loss": 0.0217, "step": 50660 }, { "grad_norm": 0.24517503380775452, "learning_rate": 3.246360588379497e-05, "loss": 0.0238, "step": 50670 }, { "grad_norm": 0.2408270239830017, "learning_rate": 3.24442519182918e-05, "loss": 0.0249, "step": 50680 }, { "grad_norm": 0.17079102993011475, "learning_rate": 3.2424900952585845e-05, "loss": 0.0211, "step": 50690 }, { "grad_norm": 0.21217775344848633, "learning_rate": 3.240555298998362e-05, "loss": 0.0177, "step": 50700 }, { "grad_norm": 0.2382299304008484, "learning_rate": 3.2386208033791194e-05, "loss": 0.0217, "step": 50710 }, { "grad_norm": 0.24026651680469513, "learning_rate": 3.2366866087314064e-05, "loss": 0.0237, "step": 50720 }, { "grad_norm": 0.23607881367206573, "learning_rate": 3.234752715385727e-05, "loss": 0.0236, "step": 50730 }, { "grad_norm": 0.1864776611328125, "learning_rate": 3.232819123672529e-05, "loss": 0.0246, "step": 50740 }, { "grad_norm": 0.38041403889656067, "learning_rate": 3.2308858339222076e-05, "loss": 0.0274, "step": 50750 }, { "grad_norm": 0.3620881736278534, "learning_rate": 3.2289528464651144e-05, "loss": 0.0272, "step": 50760 }, { "grad_norm": 0.22419722378253937, "learning_rate": 3.22702016163154e-05, "loss": 0.0242, "step": 50770 }, { "grad_norm": 0.28164923191070557, "learning_rate": 3.2250877797517295e-05, "loss": 0.0207, "step": 50780 }, { "grad_norm": 0.23088152706623077, "learning_rate": 3.223155701155872e-05, "loss": 0.0215, "step": 50790 }, { "grad_norm": 0.23657412827014923, "learning_rate": 3.2212239261741105e-05, "loss": 0.0295, "step": 50800 }, { "grad_norm": 0.1868809312582016, "learning_rate": 3.219292455136528e-05, "loss": 0.0232, "step": 50810 }, { "grad_norm": 0.2262631505727768, "learning_rate": 3.217361288373165e-05, "loss": 0.0163, "step": 50820 }, { "grad_norm": 0.21550551056861877, "learning_rate": 3.215430426214002e-05, "loss": 0.0223, "step": 50830 }, { "grad_norm": 0.215159073472023, "learning_rate": 3.2134998689889705e-05, "loss": 0.0227, "step": 50840 }, { "grad_norm": 0.2057182341814041, "learning_rate": 3.211569617027952e-05, "loss": 0.0201, "step": 50850 }, { "grad_norm": 0.15272480249404907, "learning_rate": 3.209639670660769e-05, "loss": 0.0215, "step": 50860 }, { "grad_norm": 0.25208741426467896, "learning_rate": 3.207710030217202e-05, "loss": 0.0225, "step": 50870 }, { "grad_norm": 0.2518027424812317, "learning_rate": 3.205780696026972e-05, "loss": 0.0207, "step": 50880 }, { "grad_norm": 0.18933506309986115, "learning_rate": 3.203851668419749e-05, "loss": 0.0181, "step": 50890 }, { "grad_norm": 0.21929499506950378, "learning_rate": 3.201922947725149e-05, "loss": 0.0205, "step": 50900 }, { "grad_norm": 0.22985921800136566, "learning_rate": 3.199994534272742e-05, "loss": 0.0212, "step": 50910 }, { "grad_norm": 0.17810583114624023, "learning_rate": 3.1980664283920394e-05, "loss": 0.0231, "step": 50920 }, { "grad_norm": 0.19764013588428497, "learning_rate": 3.196138630412499e-05, "loss": 0.0199, "step": 50930 }, { "grad_norm": 0.25616079568862915, "learning_rate": 3.194211140663532e-05, "loss": 0.0181, "step": 50940 }, { "grad_norm": 0.2055714726448059, "learning_rate": 3.1922839594744914e-05, "loss": 0.0236, "step": 50950 }, { "grad_norm": 0.218446746468544, "learning_rate": 3.190357087174683e-05, "loss": 0.0218, "step": 50960 }, { "grad_norm": 0.16458149254322052, "learning_rate": 3.188430524093353e-05, "loss": 0.0175, "step": 50970 }, { "grad_norm": 0.1945965588092804, "learning_rate": 3.1865042705597014e-05, "loss": 0.0196, "step": 50980 }, { "grad_norm": 0.2156543880701065, "learning_rate": 3.184578326902868e-05, "loss": 0.0179, "step": 50990 }, { "grad_norm": 0.18483462929725647, "learning_rate": 3.182652693451949e-05, "loss": 0.0226, "step": 51000 }, { "grad_norm": 0.1585584431886673, "learning_rate": 3.180727370535981e-05, "loss": 0.0201, "step": 51010 }, { "grad_norm": 0.22990407049655914, "learning_rate": 3.178802358483945e-05, "loss": 0.0208, "step": 51020 }, { "grad_norm": 0.17091616988182068, "learning_rate": 3.17687765762478e-05, "loss": 0.0173, "step": 51030 }, { "grad_norm": 0.1777070015668869, "learning_rate": 3.174953268287357e-05, "loss": 0.0239, "step": 51040 }, { "grad_norm": 0.2607332468032837, "learning_rate": 3.173029190800508e-05, "loss": 0.0199, "step": 51050 }, { "grad_norm": 0.1796814203262329, "learning_rate": 3.171105425493e-05, "loss": 0.0264, "step": 51060 }, { "grad_norm": 0.22117261588573456, "learning_rate": 3.169181972693557e-05, "loss": 0.0163, "step": 51070 }, { "grad_norm": 0.26259854435920715, "learning_rate": 3.1672588327308404e-05, "loss": 0.0197, "step": 51080 }, { "grad_norm": 0.25832831859588623, "learning_rate": 3.1653360059334646e-05, "loss": 0.0248, "step": 51090 }, { "grad_norm": 0.34688088297843933, "learning_rate": 3.163413492629985e-05, "loss": 0.034, "step": 51100 }, { "grad_norm": 0.23441901803016663, "learning_rate": 3.161491293148912e-05, "loss": 0.0205, "step": 51110 }, { "grad_norm": 0.2036290168762207, "learning_rate": 3.1595694078186925e-05, "loss": 0.0203, "step": 51120 }, { "grad_norm": 0.2592296004295349, "learning_rate": 3.157647836967724e-05, "loss": 0.0181, "step": 51130 }, { "grad_norm": 0.24865686893463135, "learning_rate": 3.155726580924354e-05, "loss": 0.0198, "step": 51140 }, { "grad_norm": 0.21314053237438202, "learning_rate": 3.1538056400168676e-05, "loss": 0.0298, "step": 51150 }, { "grad_norm": 0.209075465798378, "learning_rate": 3.151885014573506e-05, "loss": 0.0258, "step": 51160 }, { "grad_norm": 0.2320469170808792, "learning_rate": 3.149964704922448e-05, "loss": 0.0313, "step": 51170 }, { "grad_norm": 0.23031647503376007, "learning_rate": 3.148044711391825e-05, "loss": 0.0189, "step": 51180 }, { "grad_norm": 0.21055275201797485, "learning_rate": 3.146125034309707e-05, "loss": 0.0198, "step": 51190 }, { "grad_norm": 0.23371359705924988, "learning_rate": 3.1442056740041195e-05, "loss": 0.0248, "step": 51200 }, { "grad_norm": 0.23476842045783997, "learning_rate": 3.142286630803026e-05, "loss": 0.0307, "step": 51210 }, { "grad_norm": 0.19478538632392883, "learning_rate": 3.140367905034337e-05, "loss": 0.0215, "step": 51220 }, { "grad_norm": 0.23742036521434784, "learning_rate": 3.138449497025914e-05, "loss": 0.0238, "step": 51230 }, { "grad_norm": 0.3359092175960541, "learning_rate": 3.136531407105557e-05, "loss": 0.0231, "step": 51240 }, { "grad_norm": 0.21624824404716492, "learning_rate": 3.1346136356010184e-05, "loss": 0.0271, "step": 51250 }, { "grad_norm": 0.23347458243370056, "learning_rate": 3.132696182839988e-05, "loss": 0.0183, "step": 51260 }, { "grad_norm": 0.2356213480234146, "learning_rate": 3.130779049150111e-05, "loss": 0.0261, "step": 51270 }, { "grad_norm": 0.18174685537815094, "learning_rate": 3.128862234858971e-05, "loss": 0.0193, "step": 51280 }, { "grad_norm": 0.28925198316574097, "learning_rate": 3.1269457402941005e-05, "loss": 0.0218, "step": 51290 }, { "grad_norm": 0.32825547456741333, "learning_rate": 3.125029565782974e-05, "loss": 0.0218, "step": 51300 }, { "grad_norm": 0.18226097524166107, "learning_rate": 3.123113711653013e-05, "loss": 0.0201, "step": 51310 }, { "grad_norm": 0.15983417630195618, "learning_rate": 3.121198178231587e-05, "loss": 0.0213, "step": 51320 }, { "grad_norm": 0.23145119845867157, "learning_rate": 3.119282965846006e-05, "loss": 0.0192, "step": 51330 }, { "grad_norm": 0.20545078814029694, "learning_rate": 3.1173680748235303e-05, "loss": 0.0201, "step": 51340 }, { "grad_norm": 0.2572181224822998, "learning_rate": 3.115453505491358e-05, "loss": 0.0198, "step": 51350 }, { "grad_norm": 0.15600250661373138, "learning_rate": 3.1135392581766404e-05, "loss": 0.0179, "step": 51360 }, { "grad_norm": 0.2702554762363434, "learning_rate": 3.111625333206467e-05, "loss": 0.0194, "step": 51370 }, { "grad_norm": 0.19630946218967438, "learning_rate": 3.109711730907881e-05, "loss": 0.0151, "step": 51380 }, { "grad_norm": 0.2357814610004425, "learning_rate": 3.107798451607859e-05, "loss": 0.0197, "step": 51390 }, { "grad_norm": 0.1695411205291748, "learning_rate": 3.105885495633328e-05, "loss": 0.0198, "step": 51400 }, { "grad_norm": 0.18049050867557526, "learning_rate": 3.103972863311163e-05, "loss": 0.0167, "step": 51410 }, { "grad_norm": 0.2836810350418091, "learning_rate": 3.102060554968178e-05, "loss": 0.0186, "step": 51420 }, { "grad_norm": 0.17617572844028473, "learning_rate": 3.100148570931137e-05, "loss": 0.0201, "step": 51430 }, { "grad_norm": 0.34194493293762207, "learning_rate": 3.098236911526744e-05, "loss": 0.0224, "step": 51440 }, { "grad_norm": 0.23488552868366241, "learning_rate": 3.0963255770816495e-05, "loss": 0.0158, "step": 51450 }, { "grad_norm": 0.1877233386039734, "learning_rate": 3.094414567922447e-05, "loss": 0.0245, "step": 51460 }, { "grad_norm": 0.15900316834449768, "learning_rate": 3.092503884375678e-05, "loss": 0.0228, "step": 51470 }, { "grad_norm": 0.2397780418395996, "learning_rate": 3.0905935267678255e-05, "loss": 0.0211, "step": 51480 }, { "grad_norm": 0.17265893518924713, "learning_rate": 3.0886834954253166e-05, "loss": 0.018, "step": 51490 }, { "grad_norm": 0.21598514914512634, "learning_rate": 3.086773790674524e-05, "loss": 0.0174, "step": 51500 }, { "grad_norm": 0.22056646645069122, "learning_rate": 3.0848644128417617e-05, "loss": 0.0227, "step": 51510 }, { "grad_norm": 0.2146686464548111, "learning_rate": 3.082955362253294e-05, "loss": 0.021, "step": 51520 }, { "grad_norm": 0.21865509450435638, "learning_rate": 3.081046639235323e-05, "loss": 0.0214, "step": 51530 }, { "grad_norm": 0.24428723752498627, "learning_rate": 3.0791382441139986e-05, "loss": 0.015, "step": 51540 }, { "grad_norm": 0.16739723086357117, "learning_rate": 3.077230177215411e-05, "loss": 0.0219, "step": 51550 }, { "grad_norm": 0.1381993591785431, "learning_rate": 3.0753224388656e-05, "loss": 0.0169, "step": 51560 }, { "grad_norm": 0.20983633399009705, "learning_rate": 3.073415029390544e-05, "loss": 0.0217, "step": 51570 }, { "grad_norm": 0.22470304369926453, "learning_rate": 3.071507949116166e-05, "loss": 0.019, "step": 51580 }, { "grad_norm": 0.29626384377479553, "learning_rate": 3.069601198368337e-05, "loss": 0.0183, "step": 51590 }, { "grad_norm": 0.2553234100341797, "learning_rate": 3.067694777472864e-05, "loss": 0.0187, "step": 51600 }, { "grad_norm": 0.17400875687599182, "learning_rate": 3.065788686755508e-05, "loss": 0.0158, "step": 51610 }, { "grad_norm": 0.25997957587242126, "learning_rate": 3.063882926541961e-05, "loss": 0.0204, "step": 51620 }, { "grad_norm": 0.22851000726222992, "learning_rate": 3.061977497157872e-05, "loss": 0.0193, "step": 51630 }, { "grad_norm": 0.25270408391952515, "learning_rate": 3.0600723989288224e-05, "loss": 0.0178, "step": 51640 }, { "grad_norm": 0.2596634030342102, "learning_rate": 3.058167632180343e-05, "loss": 0.0191, "step": 51650 }, { "grad_norm": 0.1338062286376953, "learning_rate": 3.0562631972379074e-05, "loss": 0.0222, "step": 51660 }, { "grad_norm": 0.19277013838291168, "learning_rate": 3.0543590944269276e-05, "loss": 0.0185, "step": 51670 }, { "grad_norm": 0.2653143107891083, "learning_rate": 3.052455324072766e-05, "loss": 0.0252, "step": 51680 }, { "grad_norm": 0.2537694573402405, "learning_rate": 3.0505518865007244e-05, "loss": 0.0189, "step": 51690 }, { "grad_norm": 0.23295575380325317, "learning_rate": 3.048648782036048e-05, "loss": 0.0176, "step": 51700 }, { "grad_norm": 0.2112608551979065, "learning_rate": 3.0467460110039236e-05, "loss": 0.0202, "step": 51710 }, { "grad_norm": 0.20747137069702148, "learning_rate": 3.0448435737294866e-05, "loss": 0.0208, "step": 51720 }, { "grad_norm": 0.22828730940818787, "learning_rate": 3.042941470537808e-05, "loss": 0.0169, "step": 51730 }, { "grad_norm": 0.25595733523368835, "learning_rate": 3.0410397017539072e-05, "loss": 0.0185, "step": 51740 }, { "grad_norm": 0.19661423563957214, "learning_rate": 3.039138267702742e-05, "loss": 0.0165, "step": 51750 }, { "grad_norm": 0.1563737392425537, "learning_rate": 3.0372371687092193e-05, "loss": 0.0228, "step": 51760 }, { "grad_norm": 0.2676403820514679, "learning_rate": 3.0353364050981823e-05, "loss": 0.0195, "step": 51770 }, { "grad_norm": 0.1607760190963745, "learning_rate": 3.033435977194418e-05, "loss": 0.0191, "step": 51780 }, { "grad_norm": 0.19453714787960052, "learning_rate": 3.0315358853226618e-05, "loss": 0.0199, "step": 51790 }, { "grad_norm": 0.19843238592147827, "learning_rate": 3.0296361298075844e-05, "loss": 0.0162, "step": 51800 }, { "grad_norm": 0.204395592212677, "learning_rate": 3.027736710973803e-05, "loss": 0.0196, "step": 51810 }, { "grad_norm": 0.1419314444065094, "learning_rate": 3.025837629145875e-05, "loss": 0.0166, "step": 51820 }, { "grad_norm": 0.22619648277759552, "learning_rate": 3.0239388846483048e-05, "loss": 0.0167, "step": 51830 }, { "grad_norm": 0.1580168455839157, "learning_rate": 3.022040477805532e-05, "loss": 0.0236, "step": 51840 }, { "grad_norm": 0.17358757555484772, "learning_rate": 3.020142408941946e-05, "loss": 0.0164, "step": 51850 }, { "grad_norm": 0.17154376208782196, "learning_rate": 3.018244678381873e-05, "loss": 0.0195, "step": 51860 }, { "grad_norm": 0.21822988986968994, "learning_rate": 3.0163472864495812e-05, "loss": 0.0217, "step": 51870 }, { "grad_norm": 0.15345819294452667, "learning_rate": 3.014450233469287e-05, "loss": 0.0174, "step": 51880 }, { "grad_norm": 0.25061044096946716, "learning_rate": 3.0125535197651422e-05, "loss": 0.0268, "step": 51890 }, { "grad_norm": 0.139227494597435, "learning_rate": 3.0106571456612448e-05, "loss": 0.0182, "step": 51900 }, { "grad_norm": 0.19393764436244965, "learning_rate": 3.0087611114816305e-05, "loss": 0.0248, "step": 51910 }, { "grad_norm": 0.22336214780807495, "learning_rate": 3.006865417550284e-05, "loss": 0.0193, "step": 51920 }, { "grad_norm": 0.28313344717025757, "learning_rate": 3.0049700641911242e-05, "loss": 0.0226, "step": 51930 }, { "grad_norm": 0.17777103185653687, "learning_rate": 3.0030750517280175e-05, "loss": 0.0179, "step": 51940 }, { "grad_norm": 0.25139090418815613, "learning_rate": 3.0011803804847682e-05, "loss": 0.0213, "step": 51950 }, { "grad_norm": 0.23597636818885803, "learning_rate": 2.9992860507851224e-05, "loss": 0.0219, "step": 51960 }, { "grad_norm": 0.20927853882312775, "learning_rate": 2.9973920629527723e-05, "loss": 0.0165, "step": 51970 }, { "grad_norm": 0.2148795872926712, "learning_rate": 2.9954984173113453e-05, "loss": 0.0178, "step": 51980 }, { "grad_norm": 0.24790522456169128, "learning_rate": 2.993605114184418e-05, "loss": 0.0201, "step": 51990 }, { "grad_norm": 0.2938438951969147, "learning_rate": 2.9917121538955005e-05, "loss": 0.0271, "step": 52000 }, { "grad_norm": 0.2719651758670807, "learning_rate": 2.98981953676805e-05, "loss": 0.0279, "step": 52010 }, { "grad_norm": 0.1767842322587967, "learning_rate": 2.9879272631254594e-05, "loss": 0.0193, "step": 52020 }, { "grad_norm": 0.23042474687099457, "learning_rate": 2.986035333291073e-05, "loss": 0.0182, "step": 52030 }, { "grad_norm": 0.1680702418088913, "learning_rate": 2.9841437475881646e-05, "loss": 0.0262, "step": 52040 }, { "grad_norm": 0.2449064999818802, "learning_rate": 2.9822525063399555e-05, "loss": 0.0209, "step": 52050 }, { "grad_norm": 0.1821565181016922, "learning_rate": 2.9803616098696087e-05, "loss": 0.0209, "step": 52060 }, { "grad_norm": 0.1985255479812622, "learning_rate": 2.9784710585002228e-05, "loss": 0.0216, "step": 52070 }, { "grad_norm": 0.2761216461658478, "learning_rate": 2.9765808525548467e-05, "loss": 0.0217, "step": 52080 }, { "grad_norm": 0.1921723484992981, "learning_rate": 2.974690992356461e-05, "loss": 0.0174, "step": 52090 }, { "grad_norm": 0.17575502395629883, "learning_rate": 2.9728014782279934e-05, "loss": 0.0236, "step": 52100 }, { "grad_norm": 0.17704541981220245, "learning_rate": 2.970912310492307e-05, "loss": 0.0196, "step": 52110 }, { "grad_norm": 0.16464018821716309, "learning_rate": 2.9690234894722134e-05, "loss": 0.0158, "step": 52120 }, { "grad_norm": 0.13527826964855194, "learning_rate": 2.9671350154904577e-05, "loss": 0.0222, "step": 52130 }, { "grad_norm": 0.18406963348388672, "learning_rate": 2.9652468888697282e-05, "loss": 0.0205, "step": 52140 }, { "grad_norm": 0.21448679268360138, "learning_rate": 2.9633591099326562e-05, "loss": 0.0197, "step": 52150 }, { "grad_norm": 0.18926815688610077, "learning_rate": 2.9614716790018072e-05, "loss": 0.0178, "step": 52160 }, { "grad_norm": 0.23943649232387543, "learning_rate": 2.959584596399697e-05, "loss": 0.0154, "step": 52170 }, { "grad_norm": 0.22817131876945496, "learning_rate": 2.9576978624487717e-05, "loss": 0.024, "step": 52180 }, { "grad_norm": 0.28411322832107544, "learning_rate": 2.9558114774714263e-05, "loss": 0.0177, "step": 52190 }, { "grad_norm": 0.3469076156616211, "learning_rate": 2.9539254417899897e-05, "loss": 0.0233, "step": 52200 }, { "grad_norm": 0.23278921842575073, "learning_rate": 2.9520397557267365e-05, "loss": 0.018, "step": 52210 }, { "grad_norm": 0.2107095718383789, "learning_rate": 2.9501544196038765e-05, "loss": 0.0194, "step": 52220 }, { "grad_norm": 0.22694283723831177, "learning_rate": 2.9482694337435614e-05, "loss": 0.0257, "step": 52230 }, { "grad_norm": 0.1381070762872696, "learning_rate": 2.946384798467887e-05, "loss": 0.0169, "step": 52240 }, { "grad_norm": 0.22283370792865753, "learning_rate": 2.9445005140988824e-05, "loss": 0.016, "step": 52250 }, { "grad_norm": 0.21970602869987488, "learning_rate": 2.942616580958524e-05, "loss": 0.0207, "step": 52260 }, { "grad_norm": 0.2422495186328888, "learning_rate": 2.9407329993687193e-05, "loss": 0.0177, "step": 52270 }, { "grad_norm": 0.24904488027095795, "learning_rate": 2.938849769651326e-05, "loss": 0.0188, "step": 52280 }, { "grad_norm": 0.19471955299377441, "learning_rate": 2.9369668921281345e-05, "loss": 0.0196, "step": 52290 }, { "grad_norm": 0.22231779992580414, "learning_rate": 2.9350843671208773e-05, "loss": 0.0195, "step": 52300 }, { "grad_norm": 0.20935148000717163, "learning_rate": 2.9332021949512245e-05, "loss": 0.0204, "step": 52310 }, { "grad_norm": 0.25766289234161377, "learning_rate": 2.9313203759407908e-05, "loss": 0.021, "step": 52320 }, { "grad_norm": 0.13646318018436432, "learning_rate": 2.929438910411127e-05, "loss": 0.0306, "step": 52330 }, { "grad_norm": 0.26476195454597473, "learning_rate": 2.927557798683722e-05, "loss": 0.0214, "step": 52340 }, { "grad_norm": 0.18414899706840515, "learning_rate": 2.9256770410800095e-05, "loss": 0.0155, "step": 52350 }, { "grad_norm": 0.2832992970943451, "learning_rate": 2.9237966379213554e-05, "loss": 0.0189, "step": 52360 }, { "grad_norm": 0.20553596317768097, "learning_rate": 2.9219165895290736e-05, "loss": 0.0197, "step": 52370 }, { "grad_norm": 0.1765483319759369, "learning_rate": 2.9200368962244083e-05, "loss": 0.0228, "step": 52380 }, { "grad_norm": 0.17854535579681396, "learning_rate": 2.9181575583285536e-05, "loss": 0.0167, "step": 52390 }, { "grad_norm": 0.22429214417934418, "learning_rate": 2.916278576162632e-05, "loss": 0.0208, "step": 52400 }, { "grad_norm": 0.31789013743400574, "learning_rate": 2.9143999500477116e-05, "loss": 0.0187, "step": 52410 }, { "grad_norm": 0.20449422299861908, "learning_rate": 2.9125216803048004e-05, "loss": 0.023, "step": 52420 }, { "grad_norm": 0.22238792479038239, "learning_rate": 2.9106437672548403e-05, "loss": 0.0175, "step": 52430 }, { "grad_norm": 0.16516229510307312, "learning_rate": 2.9087662112187154e-05, "loss": 0.0168, "step": 52440 }, { "grad_norm": 0.2578878402709961, "learning_rate": 2.9068890125172498e-05, "loss": 0.0162, "step": 52450 }, { "grad_norm": 0.26417532563209534, "learning_rate": 2.9050121714712058e-05, "loss": 0.0282, "step": 52460 }, { "grad_norm": 0.24834424257278442, "learning_rate": 2.9031356884012835e-05, "loss": 0.0198, "step": 52470 }, { "grad_norm": 0.2597562372684479, "learning_rate": 2.9012595636281247e-05, "loss": 0.0249, "step": 52480 }, { "grad_norm": 0.15542997419834137, "learning_rate": 2.8993837974723016e-05, "loss": 0.0314, "step": 52490 }, { "grad_norm": 0.19511853158473969, "learning_rate": 2.8975083902543394e-05, "loss": 0.0189, "step": 52500 }, { "grad_norm": 0.17424315214157104, "learning_rate": 2.895633342294688e-05, "loss": 0.0208, "step": 52510 }, { "grad_norm": 0.23854486644268036, "learning_rate": 2.893758653913744e-05, "loss": 0.0196, "step": 52520 }, { "grad_norm": 0.22309239208698273, "learning_rate": 2.8918843254318423e-05, "loss": 0.0183, "step": 52530 }, { "grad_norm": 0.2794658839702606, "learning_rate": 2.8900103571692483e-05, "loss": 0.0216, "step": 52540 }, { "grad_norm": 0.20536138117313385, "learning_rate": 2.8881367494461796e-05, "loss": 0.0187, "step": 52550 }, { "grad_norm": 0.17824918031692505, "learning_rate": 2.886263502582779e-05, "loss": 0.0209, "step": 52560 }, { "grad_norm": 0.201629176735878, "learning_rate": 2.8843906168991353e-05, "loss": 0.0198, "step": 52570 }, { "grad_norm": 0.1930060088634491, "learning_rate": 2.8825180927152728e-05, "loss": 0.0212, "step": 52580 }, { "grad_norm": 0.20203816890716553, "learning_rate": 2.8806459303511547e-05, "loss": 0.0156, "step": 52590 }, { "grad_norm": 0.2508186101913452, "learning_rate": 2.8787741301266852e-05, "loss": 0.0191, "step": 52600 }, { "grad_norm": 0.2334776222705841, "learning_rate": 2.876902692361697e-05, "loss": 0.0177, "step": 52610 }, { "grad_norm": 0.21660450100898743, "learning_rate": 2.875031617375976e-05, "loss": 0.0157, "step": 52620 }, { "grad_norm": 0.23081882297992706, "learning_rate": 2.8731609054892318e-05, "loss": 0.0157, "step": 52630 }, { "grad_norm": 0.2040141224861145, "learning_rate": 2.871290557021119e-05, "loss": 0.0167, "step": 52640 }, { "grad_norm": 0.38902172446250916, "learning_rate": 2.8694205722912298e-05, "loss": 0.0212, "step": 52650 }, { "grad_norm": 0.2312503159046173, "learning_rate": 2.8675509516190936e-05, "loss": 0.0209, "step": 52660 }, { "grad_norm": 0.5034071803092957, "learning_rate": 2.8656816953241772e-05, "loss": 0.0241, "step": 52670 }, { "grad_norm": 0.2570095360279083, "learning_rate": 2.863812803725887e-05, "loss": 0.0186, "step": 52680 }, { "grad_norm": 0.2536652684211731, "learning_rate": 2.8619442771435623e-05, "loss": 0.017, "step": 52690 }, { "grad_norm": 0.20332174003124237, "learning_rate": 2.860076115896484e-05, "loss": 0.0259, "step": 52700 }, { "grad_norm": 0.17340776324272156, "learning_rate": 2.8582083203038712e-05, "loss": 0.0213, "step": 52710 }, { "grad_norm": 0.16459888219833374, "learning_rate": 2.8563408906848778e-05, "loss": 0.0192, "step": 52720 }, { "grad_norm": 0.21098147332668304, "learning_rate": 2.8544738273585993e-05, "loss": 0.0167, "step": 52730 }, { "grad_norm": 0.20442476868629456, "learning_rate": 2.8526071306440595e-05, "loss": 0.0149, "step": 52740 }, { "grad_norm": 0.21527846157550812, "learning_rate": 2.850740800860233e-05, "loss": 0.0188, "step": 52750 }, { "grad_norm": 0.18026474118232727, "learning_rate": 2.8488748383260178e-05, "loss": 0.0168, "step": 52760 }, { "grad_norm": 0.15454241633415222, "learning_rate": 2.8470092433602635e-05, "loss": 0.0241, "step": 52770 }, { "grad_norm": 0.14356818795204163, "learning_rate": 2.845144016281745e-05, "loss": 0.0161, "step": 52780 }, { "grad_norm": 0.22765325009822845, "learning_rate": 2.8432791574091744e-05, "loss": 0.0285, "step": 52790 }, { "grad_norm": 0.15821047127246857, "learning_rate": 2.8414146670612134e-05, "loss": 0.0172, "step": 52800 }, { "grad_norm": 0.16729140281677246, "learning_rate": 2.8395505455564446e-05, "loss": 0.0191, "step": 52810 }, { "grad_norm": 0.2439875304698944, "learning_rate": 2.837686793213403e-05, "loss": 0.0186, "step": 52820 }, { "grad_norm": 0.22940050065517426, "learning_rate": 2.8358234103505477e-05, "loss": 0.0224, "step": 52830 }, { "grad_norm": 0.2693231403827667, "learning_rate": 2.8339603972862806e-05, "loss": 0.0181, "step": 52840 }, { "grad_norm": 0.23184996843338013, "learning_rate": 2.83209775433894e-05, "loss": 0.0197, "step": 52850 }, { "grad_norm": 0.1359613835811615, "learning_rate": 2.8302354818268013e-05, "loss": 0.0187, "step": 52860 }, { "grad_norm": 0.2127571702003479, "learning_rate": 2.8283735800680754e-05, "loss": 0.0219, "step": 52870 }, { "grad_norm": 0.18960346281528473, "learning_rate": 2.826512049380913e-05, "loss": 0.0179, "step": 52880 }, { "grad_norm": 0.22192969918251038, "learning_rate": 2.8246508900833935e-05, "loss": 0.0209, "step": 52890 }, { "grad_norm": 0.14097975194454193, "learning_rate": 2.822790102493541e-05, "loss": 0.0207, "step": 52900 }, { "grad_norm": 0.16711612045764923, "learning_rate": 2.820929686929314e-05, "loss": 0.0182, "step": 52910 }, { "grad_norm": 0.1780460923910141, "learning_rate": 2.819069643708605e-05, "loss": 0.0181, "step": 52920 }, { "grad_norm": 0.2457633912563324, "learning_rate": 2.8172099731492484e-05, "loss": 0.0212, "step": 52930 }, { "grad_norm": 0.18528807163238525, "learning_rate": 2.815350675569003e-05, "loss": 0.028, "step": 52940 }, { "grad_norm": 0.1983967423439026, "learning_rate": 2.8134917512855807e-05, "loss": 0.0217, "step": 52950 }, { "grad_norm": 0.39943233132362366, "learning_rate": 2.8116332006166134e-05, "loss": 0.0166, "step": 52960 }, { "grad_norm": 0.27433887124061584, "learning_rate": 2.809775023879685e-05, "loss": 0.0261, "step": 52970 }, { "grad_norm": 0.20547492802143097, "learning_rate": 2.807917221392299e-05, "loss": 0.0221, "step": 52980 }, { "grad_norm": 0.19269579648971558, "learning_rate": 2.8060597934719067e-05, "loss": 0.0178, "step": 52990 }, { "grad_norm": 0.18992814421653748, "learning_rate": 2.8042027404358927e-05, "loss": 0.022, "step": 53000 }, { "grad_norm": 0.17601265013217926, "learning_rate": 2.802346062601572e-05, "loss": 0.0196, "step": 53010 }, { "grad_norm": 0.2747848331928253, "learning_rate": 2.800489760286207e-05, "loss": 0.0261, "step": 53020 }, { "grad_norm": 0.3500610888004303, "learning_rate": 2.7986338338069834e-05, "loss": 0.0206, "step": 53030 }, { "grad_norm": 0.2541186213493347, "learning_rate": 2.7967782834810297e-05, "loss": 0.0236, "step": 53040 }, { "grad_norm": 0.29698508977890015, "learning_rate": 2.794923109625409e-05, "loss": 0.0207, "step": 53050 }, { "grad_norm": 0.24252402782440186, "learning_rate": 2.7930683125571205e-05, "loss": 0.0159, "step": 53060 }, { "grad_norm": 0.23701860010623932, "learning_rate": 2.7912138925930997e-05, "loss": 0.0221, "step": 53070 }, { "grad_norm": 0.18927742540836334, "learning_rate": 2.7893598500502117e-05, "loss": 0.0328, "step": 53080 }, { "grad_norm": 0.2774314284324646, "learning_rate": 2.7875061852452644e-05, "loss": 0.0212, "step": 53090 }, { "grad_norm": 0.2922809422016144, "learning_rate": 2.7856528984949982e-05, "loss": 0.0232, "step": 53100 }, { "grad_norm": 0.16137972474098206, "learning_rate": 2.7837999901160888e-05, "loss": 0.0185, "step": 53110 }, { "grad_norm": 0.22184011340141296, "learning_rate": 2.7819474604251484e-05, "loss": 0.017, "step": 53120 }, { "grad_norm": 0.15747563540935516, "learning_rate": 2.780095309738725e-05, "loss": 0.0225, "step": 53130 }, { "grad_norm": 0.1946982890367508, "learning_rate": 2.778243538373294e-05, "loss": 0.0163, "step": 53140 }, { "grad_norm": 0.15829747915267944, "learning_rate": 2.7763921466452826e-05, "loss": 0.0197, "step": 53150 }, { "grad_norm": 0.21035930514335632, "learning_rate": 2.7745411348710336e-05, "loss": 0.0247, "step": 53160 }, { "grad_norm": 0.3104405105113983, "learning_rate": 2.7726905033668395e-05, "loss": 0.0245, "step": 53170 }, { "grad_norm": 0.19809149205684662, "learning_rate": 2.7708402524489214e-05, "loss": 0.0233, "step": 53180 }, { "grad_norm": 0.18763960897922516, "learning_rate": 2.7689903824334364e-05, "loss": 0.0191, "step": 53190 }, { "grad_norm": 0.2699180841445923, "learning_rate": 2.7671408936364785e-05, "loss": 0.0214, "step": 53200 }, { "grad_norm": 0.2634789049625397, "learning_rate": 2.765291786374069e-05, "loss": 0.0185, "step": 53210 }, { "grad_norm": 0.1707092523574829, "learning_rate": 2.7634430609621775e-05, "loss": 0.0162, "step": 53220 }, { "grad_norm": 0.23936885595321655, "learning_rate": 2.7615947177166956e-05, "loss": 0.0186, "step": 53230 }, { "grad_norm": 0.29787662625312805, "learning_rate": 2.7597467569534553e-05, "loss": 0.0149, "step": 53240 }, { "grad_norm": 0.1930379569530487, "learning_rate": 2.757899178988226e-05, "loss": 0.0183, "step": 53250 }, { "grad_norm": 0.14700491726398468, "learning_rate": 2.7560519841367005e-05, "loss": 0.018, "step": 53260 }, { "grad_norm": 0.1824825257062912, "learning_rate": 2.7542051727145237e-05, "loss": 0.0177, "step": 53270 }, { "grad_norm": 0.2568514049053192, "learning_rate": 2.7523587450372578e-05, "loss": 0.0229, "step": 53280 }, { "grad_norm": 0.2146046906709671, "learning_rate": 2.750512701420409e-05, "loss": 0.0179, "step": 53290 }, { "grad_norm": 0.3069945275783539, "learning_rate": 2.7486670421794158e-05, "loss": 0.0205, "step": 53300 }, { "grad_norm": 0.35014262795448303, "learning_rate": 2.7468217676296515e-05, "loss": 0.0216, "step": 53310 }, { "grad_norm": 0.21141016483306885, "learning_rate": 2.7449768780864226e-05, "loss": 0.0191, "step": 53320 }, { "grad_norm": 0.18290956318378448, "learning_rate": 2.7431323738649724e-05, "loss": 0.0178, "step": 53330 }, { "grad_norm": 0.28292787075042725, "learning_rate": 2.7412882552804713e-05, "loss": 0.0202, "step": 53340 }, { "grad_norm": 0.16967979073524475, "learning_rate": 2.739444522648032e-05, "loss": 0.015, "step": 53350 }, { "grad_norm": 0.12817703187465668, "learning_rate": 2.737601176282697e-05, "loss": 0.0173, "step": 53360 }, { "grad_norm": 0.194911390542984, "learning_rate": 2.735758216499445e-05, "loss": 0.0187, "step": 53370 }, { "grad_norm": 0.1617957055568695, "learning_rate": 2.7339156436131864e-05, "loss": 0.0166, "step": 53380 }, { "grad_norm": 0.17110975086688995, "learning_rate": 2.7320734579387663e-05, "loss": 0.0191, "step": 53390 }, { "grad_norm": 0.2584983706474304, "learning_rate": 2.730231659790966e-05, "loss": 0.015, "step": 53400 }, { "grad_norm": 0.25437167286872864, "learning_rate": 2.728390249484494e-05, "loss": 0.0232, "step": 53410 }, { "grad_norm": 0.20692318677902222, "learning_rate": 2.7265492273340032e-05, "loss": 0.0184, "step": 53420 }, { "grad_norm": 0.22096730768680573, "learning_rate": 2.7247085936540678e-05, "loss": 0.015, "step": 53430 }, { "grad_norm": 0.18793043494224548, "learning_rate": 2.722868348759205e-05, "loss": 0.0179, "step": 53440 }, { "grad_norm": 0.21153584122657776, "learning_rate": 2.7210284929638635e-05, "loss": 0.0161, "step": 53450 }, { "grad_norm": 0.18275722861289978, "learning_rate": 2.7191890265824183e-05, "loss": 0.0191, "step": 53460 }, { "grad_norm": 0.221204936504364, "learning_rate": 2.7173499499291926e-05, "loss": 0.0191, "step": 53470 }, { "grad_norm": 0.23216909170150757, "learning_rate": 2.7155112633184277e-05, "loss": 0.0206, "step": 53480 }, { "grad_norm": 0.2934727966785431, "learning_rate": 2.713672967064307e-05, "loss": 0.0203, "step": 53490 }, { "grad_norm": 0.17117133736610413, "learning_rate": 2.711835061480945e-05, "loss": 0.0166, "step": 53500 }, { "grad_norm": 0.2122427225112915, "learning_rate": 2.7099975468823896e-05, "loss": 0.0216, "step": 53510 }, { "grad_norm": 0.2681918740272522, "learning_rate": 2.708160423582622e-05, "loss": 0.0181, "step": 53520 }, { "grad_norm": 0.3260733187198639, "learning_rate": 2.706323691895557e-05, "loss": 0.0225, "step": 53530 }, { "grad_norm": 0.2496149092912674, "learning_rate": 2.70448735213504e-05, "loss": 0.0168, "step": 53540 }, { "grad_norm": 0.22824734449386597, "learning_rate": 2.702651404614852e-05, "loss": 0.0228, "step": 53550 }, { "grad_norm": 0.2621277868747711, "learning_rate": 2.7008158496487056e-05, "loss": 0.0182, "step": 53560 }, { "grad_norm": 0.19874423742294312, "learning_rate": 2.6989806875502487e-05, "loss": 0.0179, "step": 53570 }, { "grad_norm": 0.20420388877391815, "learning_rate": 2.6971459186330584e-05, "loss": 0.0165, "step": 53580 }, { "grad_norm": 0.2169874906539917, "learning_rate": 2.695311543210648e-05, "loss": 0.0175, "step": 53590 }, { "grad_norm": 0.2090858817100525, "learning_rate": 2.693477561596463e-05, "loss": 0.0165, "step": 53600 }, { "grad_norm": 0.22347363829612732, "learning_rate": 2.6916439741038756e-05, "loss": 0.0182, "step": 53610 }, { "grad_norm": 0.24849078059196472, "learning_rate": 2.689810781046203e-05, "loss": 0.0204, "step": 53620 }, { "grad_norm": 0.21691688895225525, "learning_rate": 2.6879779827366823e-05, "loss": 0.0203, "step": 53630 }, { "grad_norm": 0.27737563848495483, "learning_rate": 2.6861455794884904e-05, "loss": 0.019, "step": 53640 }, { "grad_norm": 0.2073315531015396, "learning_rate": 2.6843135716147373e-05, "loss": 0.0177, "step": 53650 }, { "grad_norm": 0.25876420736312866, "learning_rate": 2.6824819594284556e-05, "loss": 0.0208, "step": 53660 }, { "grad_norm": 0.23056842386722565, "learning_rate": 2.6806507432426275e-05, "loss": 0.0205, "step": 53670 }, { "grad_norm": 0.1969492882490158, "learning_rate": 2.6788199233701512e-05, "loss": 0.0222, "step": 53680 }, { "grad_norm": 0.187312051653862, "learning_rate": 2.6769895001238652e-05, "loss": 0.0239, "step": 53690 }, { "grad_norm": 0.19663730263710022, "learning_rate": 2.67515947381654e-05, "loss": 0.0173, "step": 53700 }, { "grad_norm": 0.20287053287029266, "learning_rate": 2.6733298447608745e-05, "loss": 0.0184, "step": 53710 }, { "grad_norm": 0.2670707106590271, "learning_rate": 2.6715006132695074e-05, "loss": 0.0231, "step": 53720 }, { "grad_norm": 0.22508381307125092, "learning_rate": 2.6696717796549985e-05, "loss": 0.0293, "step": 53730 }, { "grad_norm": 0.28978487849235535, "learning_rate": 2.667843344229848e-05, "loss": 0.0176, "step": 53740 }, { "grad_norm": 0.2924948036670685, "learning_rate": 2.6660153073064846e-05, "loss": 0.0211, "step": 53750 }, { "grad_norm": 0.2861941158771515, "learning_rate": 2.664187669197271e-05, "loss": 0.0236, "step": 53760 }, { "grad_norm": 0.169092059135437, "learning_rate": 2.6623604302145005e-05, "loss": 0.0174, "step": 53770 }, { "grad_norm": 0.26917773485183716, "learning_rate": 2.6605335906703975e-05, "loss": 0.0197, "step": 53780 }, { "grad_norm": 0.17309673130512238, "learning_rate": 2.65870715087712e-05, "loss": 0.0221, "step": 53790 }, { "grad_norm": 0.2603498697280884, "learning_rate": 2.6568811111467573e-05, "loss": 0.0206, "step": 53800 }, { "grad_norm": 0.16156402230262756, "learning_rate": 2.6550554717913258e-05, "loss": 0.0189, "step": 53810 }, { "grad_norm": 0.2667025327682495, "learning_rate": 2.6532302331227805e-05, "loss": 0.0175, "step": 53820 }, { "grad_norm": 0.24221891164779663, "learning_rate": 2.651405395453004e-05, "loss": 0.0193, "step": 53830 }, { "grad_norm": 0.2016695886850357, "learning_rate": 2.6495809590938115e-05, "loss": 0.0234, "step": 53840 }, { "grad_norm": 0.13158315420150757, "learning_rate": 2.64775692435695e-05, "loss": 0.0199, "step": 53850 }, { "grad_norm": 0.1835896223783493, "learning_rate": 2.6459332915540928e-05, "loss": 0.0162, "step": 53860 }, { "grad_norm": 0.2678047716617584, "learning_rate": 2.644110060996856e-05, "loss": 0.0242, "step": 53870 }, { "grad_norm": 0.11245625466108322, "learning_rate": 2.642287232996774e-05, "loss": 0.0169, "step": 53880 }, { "grad_norm": 0.15124689042568207, "learning_rate": 2.6404648078653205e-05, "loss": 0.0135, "step": 53890 }, { "grad_norm": 0.21900515258312225, "learning_rate": 2.6386427859139002e-05, "loss": 0.0186, "step": 53900 }, { "grad_norm": 0.27588361501693726, "learning_rate": 2.63682116745384e-05, "loss": 0.0194, "step": 53910 }, { "grad_norm": 0.20378614962100983, "learning_rate": 2.6349999527964138e-05, "loss": 0.017, "step": 53920 }, { "grad_norm": 0.23754024505615234, "learning_rate": 2.63317914225281e-05, "loss": 0.0226, "step": 53930 }, { "grad_norm": 0.18798309564590454, "learning_rate": 2.6313587361341585e-05, "loss": 0.0234, "step": 53940 }, { "grad_norm": 0.32809001207351685, "learning_rate": 2.6295387347515165e-05, "loss": 0.0249, "step": 53950 }, { "grad_norm": 0.19636529684066772, "learning_rate": 2.6277191384158727e-05, "loss": 0.021, "step": 53960 }, { "grad_norm": 0.1966591328382492, "learning_rate": 2.6258999474381457e-05, "loss": 0.0226, "step": 53970 }, { "grad_norm": 0.23216502368450165, "learning_rate": 2.624081162129186e-05, "loss": 0.0174, "step": 53980 }, { "grad_norm": 0.20748740434646606, "learning_rate": 2.6222627827997765e-05, "loss": 0.0298, "step": 53990 }, { "grad_norm": 0.27436769008636475, "learning_rate": 2.6204448097606236e-05, "loss": 0.0182, "step": 54000 }, { "grad_norm": 0.29605215787887573, "learning_rate": 2.6186272433223726e-05, "loss": 0.0207, "step": 54010 }, { "grad_norm": 0.18174761533737183, "learning_rate": 2.6168100837955943e-05, "loss": 0.0175, "step": 54020 }, { "grad_norm": 0.22232085466384888, "learning_rate": 2.6149933314907926e-05, "loss": 0.0284, "step": 54030 }, { "grad_norm": 0.2893534302711487, "learning_rate": 2.6131769867184e-05, "loss": 0.0228, "step": 54040 }, { "grad_norm": 0.13402526080608368, "learning_rate": 2.611361049788783e-05, "loss": 0.0179, "step": 54050 }, { "grad_norm": 0.2594326436519623, "learning_rate": 2.6095455210122292e-05, "loss": 0.0169, "step": 54060 }, { "grad_norm": 0.4122079014778137, "learning_rate": 2.6077304006989712e-05, "loss": 0.015, "step": 54070 }, { "grad_norm": 0.20106182992458344, "learning_rate": 2.6059156891591562e-05, "loss": 0.0177, "step": 54080 }, { "grad_norm": 0.19463756680488586, "learning_rate": 2.6041013867028718e-05, "loss": 0.0169, "step": 54090 }, { "grad_norm": 0.19937987625598907, "learning_rate": 2.6022874936401347e-05, "loss": 0.0163, "step": 54100 }, { "grad_norm": 0.3683682978153229, "learning_rate": 2.6004740102808832e-05, "loss": 0.0179, "step": 54110 }, { "grad_norm": 0.20160095393657684, "learning_rate": 2.598660936935e-05, "loss": 0.0166, "step": 54120 }, { "grad_norm": 0.22864288091659546, "learning_rate": 2.5968482739122845e-05, "loss": 0.0218, "step": 54130 }, { "grad_norm": 0.24518203735351562, "learning_rate": 2.595036021522472e-05, "loss": 0.0181, "step": 54140 }, { "grad_norm": 0.24087662994861603, "learning_rate": 2.5932241800752278e-05, "loss": 0.0197, "step": 54150 }, { "grad_norm": 0.3358241021633148, "learning_rate": 2.5914127498801453e-05, "loss": 0.0321, "step": 54160 }, { "grad_norm": 0.20679500699043274, "learning_rate": 2.5896017312467497e-05, "loss": 0.0183, "step": 54170 }, { "grad_norm": 0.2428034394979477, "learning_rate": 2.587791124484493e-05, "loss": 0.0165, "step": 54180 }, { "grad_norm": 0.23541828989982605, "learning_rate": 2.5859809299027615e-05, "loss": 0.0215, "step": 54190 }, { "grad_norm": 0.2795184850692749, "learning_rate": 2.5841711478108632e-05, "loss": 0.0251, "step": 54200 }, { "grad_norm": 0.2666715681552887, "learning_rate": 2.582361778518043e-05, "loss": 0.0227, "step": 54210 }, { "grad_norm": 0.24489742517471313, "learning_rate": 2.580552822333472e-05, "loss": 0.0214, "step": 54220 }, { "grad_norm": 0.220693439245224, "learning_rate": 2.578744279566252e-05, "loss": 0.0175, "step": 54230 }, { "grad_norm": 0.2096511274576187, "learning_rate": 2.576936150525413e-05, "loss": 0.0239, "step": 54240 }, { "grad_norm": 0.2193983495235443, "learning_rate": 2.5751284355199168e-05, "loss": 0.0156, "step": 54250 }, { "grad_norm": 0.214910089969635, "learning_rate": 2.573321134858646e-05, "loss": 0.0167, "step": 54260 }, { "grad_norm": 0.19843441247940063, "learning_rate": 2.5715142488504286e-05, "loss": 0.0169, "step": 54270 }, { "grad_norm": 0.15116719901561737, "learning_rate": 2.5697077778040042e-05, "loss": 0.0196, "step": 54280 }, { "grad_norm": 0.28065189719200134, "learning_rate": 2.5679017220280522e-05, "loss": 0.0268, "step": 54290 }, { "grad_norm": 0.17663584649562836, "learning_rate": 2.5660960818311796e-05, "loss": 0.0194, "step": 54300 }, { "grad_norm": 0.2629784047603607, "learning_rate": 2.564290857521915e-05, "loss": 0.0187, "step": 54310 }, { "grad_norm": 0.21844987571239471, "learning_rate": 2.5624860494087298e-05, "loss": 0.0199, "step": 54320 }, { "grad_norm": 0.17408935725688934, "learning_rate": 2.5606816578000115e-05, "loss": 0.0228, "step": 54330 }, { "grad_norm": 0.2253086417913437, "learning_rate": 2.558877683004082e-05, "loss": 0.0192, "step": 54340 }, { "grad_norm": 0.19861595332622528, "learning_rate": 2.557074125329192e-05, "loss": 0.0165, "step": 54350 }, { "grad_norm": 0.24973560869693756, "learning_rate": 2.5552709850835195e-05, "loss": 0.0183, "step": 54360 }, { "grad_norm": 0.2600748836994171, "learning_rate": 2.5534682625751738e-05, "loss": 0.0171, "step": 54370 }, { "grad_norm": 0.3283443748950958, "learning_rate": 2.551665958112186e-05, "loss": 0.0249, "step": 54380 }, { "grad_norm": 0.2578452527523041, "learning_rate": 2.549864072002527e-05, "loss": 0.0163, "step": 54390 }, { "grad_norm": 0.31059324741363525, "learning_rate": 2.5480626045540858e-05, "loss": 0.0249, "step": 54400 }, { "grad_norm": 0.20077835023403168, "learning_rate": 2.546261556074684e-05, "loss": 0.0177, "step": 54410 }, { "grad_norm": 0.20565393567085266, "learning_rate": 2.5444609268720726e-05, "loss": 0.0179, "step": 54420 }, { "grad_norm": 0.2571931481361389, "learning_rate": 2.5426607172539297e-05, "loss": 0.0179, "step": 54430 }, { "grad_norm": 0.16447527706623077, "learning_rate": 2.5408609275278617e-05, "loss": 0.0225, "step": 54440 }, { "grad_norm": 0.22080351412296295, "learning_rate": 2.5390615580014055e-05, "loss": 0.0169, "step": 54450 }, { "grad_norm": 0.15509457886219025, "learning_rate": 2.5372626089820207e-05, "loss": 0.0209, "step": 54460 }, { "grad_norm": 0.19270232319831848, "learning_rate": 2.5354640807770997e-05, "loss": 0.0163, "step": 54470 }, { "grad_norm": 0.2468322366476059, "learning_rate": 2.5336659736939622e-05, "loss": 0.0219, "step": 54480 }, { "grad_norm": 0.30695441365242004, "learning_rate": 2.5318682880398554e-05, "loss": 0.0209, "step": 54490 }, { "grad_norm": 0.17959527671337128, "learning_rate": 2.530071024121956e-05, "loss": 0.0199, "step": 54500 }, { "grad_norm": 0.2856273055076599, "learning_rate": 2.5282741822473627e-05, "loss": 0.0181, "step": 54510 }, { "grad_norm": 0.28591418266296387, "learning_rate": 2.526477762723114e-05, "loss": 0.0202, "step": 54520 }, { "grad_norm": 0.22711032629013062, "learning_rate": 2.5246817658561618e-05, "loss": 0.021, "step": 54530 }, { "grad_norm": 0.19524182379245758, "learning_rate": 2.5228861919533965e-05, "loss": 0.021, "step": 54540 }, { "grad_norm": 0.17864523828029633, "learning_rate": 2.5210910413216326e-05, "loss": 0.0264, "step": 54550 }, { "grad_norm": 0.23825015127658844, "learning_rate": 2.5192963142676086e-05, "loss": 0.0146, "step": 54560 }, { "grad_norm": 0.21612292528152466, "learning_rate": 2.517502011098001e-05, "loss": 0.0184, "step": 54570 }, { "grad_norm": 0.1743631809949875, "learning_rate": 2.5157081321193987e-05, "loss": 0.0181, "step": 54580 }, { "grad_norm": 0.17684213817119598, "learning_rate": 2.5139146776383356e-05, "loss": 0.0175, "step": 54590 }, { "grad_norm": 0.16816847026348114, "learning_rate": 2.5121216479612575e-05, "loss": 0.0171, "step": 54600 }, { "grad_norm": 0.2950396239757538, "learning_rate": 2.510329043394546e-05, "loss": 0.0253, "step": 54610 }, { "grad_norm": 0.2773053050041199, "learning_rate": 2.508536864244508e-05, "loss": 0.0224, "step": 54620 }, { "grad_norm": 0.21835748851299286, "learning_rate": 2.5067451108173778e-05, "loss": 0.0175, "step": 54630 }, { "grad_norm": 0.1651262491941452, "learning_rate": 2.5049537834193204e-05, "loss": 0.0171, "step": 54640 }, { "grad_norm": 0.17028936743736267, "learning_rate": 2.5031628823564194e-05, "loss": 0.0177, "step": 54650 }, { "grad_norm": 0.1940453052520752, "learning_rate": 2.5013724079346933e-05, "loss": 0.0172, "step": 54660 }, { "grad_norm": 0.21633079648017883, "learning_rate": 2.4995823604600854e-05, "loss": 0.0149, "step": 54670 }, { "grad_norm": 0.1646433174610138, "learning_rate": 2.497792740238465e-05, "loss": 0.016, "step": 54680 }, { "grad_norm": 0.25815349817276, "learning_rate": 2.49600354757563e-05, "loss": 0.0237, "step": 54690 }, { "grad_norm": 0.20389433205127716, "learning_rate": 2.494214782777306e-05, "loss": 0.0184, "step": 54700 }, { "grad_norm": 0.18624824285507202, "learning_rate": 2.4924264461491386e-05, "loss": 0.0183, "step": 54710 }, { "grad_norm": 0.2308913767337799, "learning_rate": 2.4906385379967133e-05, "loss": 0.0213, "step": 54720 }, { "grad_norm": 0.2205028235912323, "learning_rate": 2.4888510586255285e-05, "loss": 0.0163, "step": 54730 }, { "grad_norm": 0.1705205887556076, "learning_rate": 2.487064008341018e-05, "loss": 0.0194, "step": 54740 }, { "grad_norm": 0.22369854152202606, "learning_rate": 2.4852773874485407e-05, "loss": 0.0161, "step": 54750 }, { "grad_norm": 0.20817053318023682, "learning_rate": 2.483491196253377e-05, "loss": 0.015, "step": 54760 }, { "grad_norm": 0.33672335743904114, "learning_rate": 2.4817054350607443e-05, "loss": 0.0256, "step": 54770 }, { "grad_norm": 0.22737418115139008, "learning_rate": 2.4799201041757742e-05, "loss": 0.0169, "step": 54780 }, { "grad_norm": 0.2698836922645569, "learning_rate": 2.4781352039035373e-05, "loss": 0.0222, "step": 54790 }, { "grad_norm": 0.2640700340270996, "learning_rate": 2.4763507345490194e-05, "loss": 0.0168, "step": 54800 }, { "grad_norm": 0.19690406322479248, "learning_rate": 2.4745666964171386e-05, "loss": 0.0186, "step": 54810 }, { "grad_norm": 0.23719671368598938, "learning_rate": 2.4727830898127387e-05, "loss": 0.0181, "step": 54820 }, { "grad_norm": 0.16034123301506042, "learning_rate": 2.4709999150405895e-05, "loss": 0.0255, "step": 54830 }, { "grad_norm": 0.24065391719341278, "learning_rate": 2.469217172405388e-05, "loss": 0.0212, "step": 54840 }, { "grad_norm": 0.21176710724830627, "learning_rate": 2.4674348622117527e-05, "loss": 0.0171, "step": 54850 }, { "grad_norm": 0.21353060007095337, "learning_rate": 2.465652984764234e-05, "loss": 0.0141, "step": 54860 }, { "grad_norm": 0.2063320279121399, "learning_rate": 2.4638715403673056e-05, "loss": 0.0176, "step": 54870 }, { "grad_norm": 0.25255894660949707, "learning_rate": 2.4620905293253676e-05, "loss": 0.0206, "step": 54880 }, { "grad_norm": 0.22951331734657288, "learning_rate": 2.4603099519427463e-05, "loss": 0.0189, "step": 54890 }, { "grad_norm": 0.21815156936645508, "learning_rate": 2.458529808523695e-05, "loss": 0.017, "step": 54900 }, { "grad_norm": 0.18282759189605713, "learning_rate": 2.456750099372387e-05, "loss": 0.0133, "step": 54910 }, { "grad_norm": 0.26296132802963257, "learning_rate": 2.4549708247929327e-05, "loss": 0.0233, "step": 54920 }, { "grad_norm": 0.19791565835475922, "learning_rate": 2.4531919850893554e-05, "loss": 0.0158, "step": 54930 }, { "grad_norm": 0.22409144043922424, "learning_rate": 2.4514135805656125e-05, "loss": 0.0126, "step": 54940 }, { "grad_norm": 0.1969069540500641, "learning_rate": 2.449635611525587e-05, "loss": 0.0224, "step": 54950 }, { "grad_norm": 0.2511272430419922, "learning_rate": 2.447858078273079e-05, "loss": 0.0179, "step": 54960 }, { "grad_norm": 0.2177768051624298, "learning_rate": 2.4460809811118275e-05, "loss": 0.0244, "step": 54970 }, { "grad_norm": 0.2500154376029968, "learning_rate": 2.444304320345483e-05, "loss": 0.0183, "step": 54980 }, { "grad_norm": 0.2110934853553772, "learning_rate": 2.4425280962776354e-05, "loss": 0.0209, "step": 54990 }, { "grad_norm": 0.18943454325199127, "learning_rate": 2.4407523092117875e-05, "loss": 0.0186, "step": 55000 }, { "grad_norm": 0.20947393774986267, "learning_rate": 2.438976959451374e-05, "loss": 0.0164, "step": 55010 }, { "grad_norm": 0.2677375078201294, "learning_rate": 2.4372020472997565e-05, "loss": 0.0177, "step": 55020 }, { "grad_norm": 0.18552495539188385, "learning_rate": 2.4354275730602122e-05, "loss": 0.0143, "step": 55030 }, { "grad_norm": 0.2248649150133133, "learning_rate": 2.433653537035958e-05, "loss": 0.0191, "step": 55040 }, { "grad_norm": 0.2557401955127716, "learning_rate": 2.431879939530123e-05, "loss": 0.0179, "step": 55050 }, { "grad_norm": 0.2831006646156311, "learning_rate": 2.4301067808457684e-05, "loss": 0.0179, "step": 55060 }, { "grad_norm": 0.2665477395057678, "learning_rate": 2.428334061285878e-05, "loss": 0.02, "step": 55070 }, { "grad_norm": 0.2592308521270752, "learning_rate": 2.426561781153361e-05, "loss": 0.0181, "step": 55080 }, { "grad_norm": 0.2559945285320282, "learning_rate": 2.424789940751052e-05, "loss": 0.0178, "step": 55090 }, { "grad_norm": 0.1966884881258011, "learning_rate": 2.423018540381712e-05, "loss": 0.0188, "step": 55100 }, { "grad_norm": 0.15452897548675537, "learning_rate": 2.421247580348021e-05, "loss": 0.0198, "step": 55110 }, { "grad_norm": 0.21903376281261444, "learning_rate": 2.4194770609525897e-05, "loss": 0.0149, "step": 55120 }, { "grad_norm": 0.2188851237297058, "learning_rate": 2.417706982497951e-05, "loss": 0.0209, "step": 55130 }, { "grad_norm": 0.17790891230106354, "learning_rate": 2.415937345286563e-05, "loss": 0.0217, "step": 55140 }, { "grad_norm": 0.2646494209766388, "learning_rate": 2.4141681496208087e-05, "loss": 0.018, "step": 55150 }, { "grad_norm": 0.20152884721755981, "learning_rate": 2.4123993958029946e-05, "loss": 0.0161, "step": 55160 }, { "grad_norm": 0.3101872205734253, "learning_rate": 2.4106310841353548e-05, "loss": 0.0177, "step": 55170 }, { "grad_norm": 0.26019027829170227, "learning_rate": 2.4088632149200398e-05, "loss": 0.0254, "step": 55180 }, { "grad_norm": 0.24458836019039154, "learning_rate": 2.4070957884591367e-05, "loss": 0.0187, "step": 55190 }, { "grad_norm": 0.39321452379226685, "learning_rate": 2.4053288050546464e-05, "loss": 0.0229, "step": 55200 }, { "grad_norm": 0.15631385147571564, "learning_rate": 2.403562265008498e-05, "loss": 0.0192, "step": 55210 }, { "grad_norm": 0.26193109154701233, "learning_rate": 2.4017961686225483e-05, "loss": 0.0221, "step": 55220 }, { "grad_norm": 0.19286805391311646, "learning_rate": 2.400030516198568e-05, "loss": 0.0304, "step": 55230 }, { "grad_norm": 0.2831328809261322, "learning_rate": 2.3982653080382673e-05, "loss": 0.0203, "step": 55240 }, { "grad_norm": 0.1791898012161255, "learning_rate": 2.396500544443266e-05, "loss": 0.0175, "step": 55250 }, { "grad_norm": 0.2542022168636322, "learning_rate": 2.3947362257151156e-05, "loss": 0.0184, "step": 55260 }, { "grad_norm": 0.19860340654850006, "learning_rate": 2.39297235215529e-05, "loss": 0.0193, "step": 55270 }, { "grad_norm": 0.23554258048534393, "learning_rate": 2.3912089240651873e-05, "loss": 0.0213, "step": 55280 }, { "grad_norm": 0.18785114586353302, "learning_rate": 2.389445941746129e-05, "loss": 0.0234, "step": 55290 }, { "grad_norm": 0.2371276170015335, "learning_rate": 2.3876834054993625e-05, "loss": 0.0173, "step": 55300 }, { "grad_norm": 0.24821078777313232, "learning_rate": 2.3859213156260522e-05, "loss": 0.0166, "step": 55310 }, { "grad_norm": 0.24197202920913696, "learning_rate": 2.3841596724272948e-05, "loss": 0.019, "step": 55320 }, { "grad_norm": 0.1993938833475113, "learning_rate": 2.382398476204106e-05, "loss": 0.0201, "step": 55330 }, { "grad_norm": 0.21869149804115295, "learning_rate": 2.3806377272574254e-05, "loss": 0.0171, "step": 55340 }, { "grad_norm": 0.24957236647605896, "learning_rate": 2.3788774258881174e-05, "loss": 0.019, "step": 55350 }, { "grad_norm": 0.22381018102169037, "learning_rate": 2.37711757239697e-05, "loss": 0.0186, "step": 55360 }, { "grad_norm": 0.17486654222011566, "learning_rate": 2.3753581670846954e-05, "loss": 0.0168, "step": 55370 }, { "grad_norm": 0.1260983794927597, "learning_rate": 2.3735992102519216e-05, "loss": 0.0189, "step": 55380 }, { "grad_norm": 0.21657583117485046, "learning_rate": 2.371840702199215e-05, "loss": 0.0171, "step": 55390 }, { "grad_norm": 0.2800918519496918, "learning_rate": 2.3700826432270494e-05, "loss": 0.0175, "step": 55400 }, { "grad_norm": 0.22857826948165894, "learning_rate": 2.3683250336358326e-05, "loss": 0.026, "step": 55410 }, { "grad_norm": 0.1667640060186386, "learning_rate": 2.3665678737258923e-05, "loss": 0.0227, "step": 55420 }, { "grad_norm": 0.20858979225158691, "learning_rate": 2.3648111637974745e-05, "loss": 0.0148, "step": 55430 }, { "grad_norm": 0.12478408217430115, "learning_rate": 2.36305490415076e-05, "loss": 0.0213, "step": 55440 }, { "grad_norm": 0.23947636783123016, "learning_rate": 2.3612990950858392e-05, "loss": 0.0155, "step": 55450 }, { "grad_norm": 0.09995324909687042, "learning_rate": 2.359543736902735e-05, "loss": 0.016, "step": 55460 }, { "grad_norm": 0.23247359693050385, "learning_rate": 2.3577888299013896e-05, "loss": 0.0156, "step": 55470 }, { "grad_norm": 0.30914968252182007, "learning_rate": 2.3560343743816683e-05, "loss": 0.0164, "step": 55480 }, { "grad_norm": 0.2045898288488388, "learning_rate": 2.354280370643362e-05, "loss": 0.0234, "step": 55490 }, { "grad_norm": 0.2007661610841751, "learning_rate": 2.3525268189861777e-05, "loss": 0.0166, "step": 55500 }, { "grad_norm": 0.2785988450050354, "learning_rate": 2.3507737197097513e-05, "loss": 0.0185, "step": 55510 }, { "grad_norm": 0.2341448813676834, "learning_rate": 2.34902107311364e-05, "loss": 0.0174, "step": 55520 }, { "grad_norm": 0.24190156161785126, "learning_rate": 2.3472688794973225e-05, "loss": 0.0191, "step": 55530 }, { "grad_norm": 0.14037732779979706, "learning_rate": 2.3455171391602016e-05, "loss": 0.0182, "step": 55540 }, { "grad_norm": 0.2342316210269928, "learning_rate": 2.343765852401601e-05, "loss": 0.0182, "step": 55550 }, { "grad_norm": 0.15678592026233673, "learning_rate": 2.342015019520768e-05, "loss": 0.0203, "step": 55560 }, { "grad_norm": 0.18347428739070892, "learning_rate": 2.3402646408168742e-05, "loss": 0.0156, "step": 55570 }, { "grad_norm": 0.24384935200214386, "learning_rate": 2.3385147165890074e-05, "loss": 0.0223, "step": 55580 }, { "grad_norm": 0.21916121244430542, "learning_rate": 2.336765247136184e-05, "loss": 0.0172, "step": 55590 }, { "grad_norm": 0.24936938285827637, "learning_rate": 2.33501623275734e-05, "loss": 0.0224, "step": 55600 }, { "grad_norm": 0.20752136409282684, "learning_rate": 2.333267673751334e-05, "loss": 0.0154, "step": 55610 }, { "grad_norm": 0.24069079756736755, "learning_rate": 2.33151957041695e-05, "loss": 0.0184, "step": 55620 }, { "grad_norm": 0.20704224705696106, "learning_rate": 2.329771923052884e-05, "loss": 0.0197, "step": 55630 }, { "grad_norm": 0.19923648238182068, "learning_rate": 2.3280247319577697e-05, "loss": 0.0175, "step": 55640 }, { "grad_norm": 0.18159247934818268, "learning_rate": 2.3262779974301473e-05, "loss": 0.0209, "step": 55650 }, { "grad_norm": 0.25104016065597534, "learning_rate": 2.3245317197684895e-05, "loss": 0.0177, "step": 55660 }, { "grad_norm": 0.2230958789587021, "learning_rate": 2.322785899271188e-05, "loss": 0.0175, "step": 55670 }, { "grad_norm": 0.2749019265174866, "learning_rate": 2.3210405362365507e-05, "loss": 0.0228, "step": 55680 }, { "grad_norm": 0.2455778568983078, "learning_rate": 2.31929563096282e-05, "loss": 0.0176, "step": 55690 }, { "grad_norm": 0.20268507301807404, "learning_rate": 2.317551183748146e-05, "loss": 0.016, "step": 55700 }, { "grad_norm": 0.24250249564647675, "learning_rate": 2.3158071948906103e-05, "loss": 0.015, "step": 55710 }, { "grad_norm": 0.18216510117053986, "learning_rate": 2.314063664688212e-05, "loss": 0.0167, "step": 55720 }, { "grad_norm": 0.2041366696357727, "learning_rate": 2.3123205934388725e-05, "loss": 0.0171, "step": 55730 }, { "grad_norm": 0.20625872910022736, "learning_rate": 2.310577981440436e-05, "loss": 0.0249, "step": 55740 }, { "grad_norm": 0.183100163936615, "learning_rate": 2.3088358289906665e-05, "loss": 0.0187, "step": 55750 }, { "grad_norm": 0.2165064513683319, "learning_rate": 2.307094136387252e-05, "loss": 0.0164, "step": 55760 }, { "grad_norm": 0.28896069526672363, "learning_rate": 2.305352903927796e-05, "loss": 0.0231, "step": 55770 }, { "grad_norm": 0.3464776873588562, "learning_rate": 2.303612131909831e-05, "loss": 0.0183, "step": 55780 }, { "grad_norm": 0.17303180694580078, "learning_rate": 2.3018718206308054e-05, "loss": 0.0242, "step": 55790 }, { "grad_norm": 0.2510527968406677, "learning_rate": 2.3001319703880925e-05, "loss": 0.0191, "step": 55800 }, { "grad_norm": 0.22096720337867737, "learning_rate": 2.2983925814789835e-05, "loss": 0.0161, "step": 55810 }, { "grad_norm": 0.21428857743740082, "learning_rate": 2.2966536542006957e-05, "loss": 0.0186, "step": 55820 }, { "grad_norm": 0.18519511818885803, "learning_rate": 2.294915188850358e-05, "loss": 0.017, "step": 55830 }, { "grad_norm": 0.1579272300004959, "learning_rate": 2.2931771857250333e-05, "loss": 0.0173, "step": 55840 }, { "grad_norm": 0.2147720456123352, "learning_rate": 2.2914396451216946e-05, "loss": 0.0164, "step": 55850 }, { "grad_norm": 0.25441640615463257, "learning_rate": 2.2897025673372412e-05, "loss": 0.0161, "step": 55860 }, { "grad_norm": 0.19631044566631317, "learning_rate": 2.287965952668494e-05, "loss": 0.0191, "step": 55870 }, { "grad_norm": 0.2224825918674469, "learning_rate": 2.2862298014121873e-05, "loss": 0.0208, "step": 55880 }, { "grad_norm": 0.2073650360107422, "learning_rate": 2.28449411386499e-05, "loss": 0.0189, "step": 55890 }, { "grad_norm": 0.16784636676311493, "learning_rate": 2.282758890323477e-05, "loss": 0.0163, "step": 55900 }, { "grad_norm": 0.20324182510375977, "learning_rate": 2.2810241310841528e-05, "loss": 0.0196, "step": 55910 }, { "grad_norm": 0.23457925021648407, "learning_rate": 2.2792898364434412e-05, "loss": 0.0171, "step": 55920 }, { "grad_norm": 0.17941847443580627, "learning_rate": 2.2775560066976843e-05, "loss": 0.0195, "step": 55930 }, { "grad_norm": 0.16454271972179413, "learning_rate": 2.275822642143147e-05, "loss": 0.02, "step": 55940 }, { "grad_norm": 0.21413981914520264, "learning_rate": 2.274089743076014e-05, "loss": 0.0154, "step": 55950 }, { "grad_norm": 0.24373380839824677, "learning_rate": 2.2723573097923917e-05, "loss": 0.0169, "step": 55960 }, { "grad_norm": 0.164934903383255, "learning_rate": 2.2706253425883024e-05, "loss": 0.0143, "step": 55970 }, { "grad_norm": 0.24864013493061066, "learning_rate": 2.2688938417596933e-05, "loss": 0.0181, "step": 55980 }, { "grad_norm": 0.22737360000610352, "learning_rate": 2.2671628076024305e-05, "loss": 0.0191, "step": 55990 }, { "grad_norm": 0.124465212225914, "learning_rate": 2.2654322404123008e-05, "loss": 0.0148, "step": 56000 }, { "grad_norm": 0.13729657232761383, "learning_rate": 2.2637021404850105e-05, "loss": 0.0137, "step": 56010 }, { "grad_norm": 0.25656577944755554, "learning_rate": 2.2619725081161885e-05, "loss": 0.0211, "step": 56020 }, { "grad_norm": 0.22050535678863525, "learning_rate": 2.260243343601376e-05, "loss": 0.0168, "step": 56030 }, { "grad_norm": 0.1829289346933365, "learning_rate": 2.2585146472360473e-05, "loss": 0.0197, "step": 56040 }, { "grad_norm": 0.23927675187587738, "learning_rate": 2.2567864193155834e-05, "loss": 0.0158, "step": 56050 }, { "grad_norm": 0.20588365197181702, "learning_rate": 2.255058660135294e-05, "loss": 0.0169, "step": 56060 }, { "grad_norm": 0.25735747814178467, "learning_rate": 2.253331369990407e-05, "loss": 0.0189, "step": 56070 }, { "grad_norm": 0.19850490987300873, "learning_rate": 2.251604549176063e-05, "loss": 0.0206, "step": 56080 }, { "grad_norm": 0.14958827197551727, "learning_rate": 2.249878197987337e-05, "loss": 0.017, "step": 56090 }, { "grad_norm": 0.1581542044878006, "learning_rate": 2.2481523167192087e-05, "loss": 0.0176, "step": 56100 }, { "grad_norm": 0.2018420547246933, "learning_rate": 2.2464269056665867e-05, "loss": 0.0152, "step": 56110 }, { "grad_norm": 0.18138404190540314, "learning_rate": 2.2447019651242958e-05, "loss": 0.0231, "step": 56120 }, { "grad_norm": 0.2512330114841461, "learning_rate": 2.2429774953870818e-05, "loss": 0.0171, "step": 56130 }, { "grad_norm": 0.1942356824874878, "learning_rate": 2.241253496749611e-05, "loss": 0.019, "step": 56140 }, { "grad_norm": 0.20590351521968842, "learning_rate": 2.2395299695064614e-05, "loss": 0.0158, "step": 56150 }, { "grad_norm": 0.17031243443489075, "learning_rate": 2.237806913952145e-05, "loss": 0.0172, "step": 56160 }, { "grad_norm": 0.19009411334991455, "learning_rate": 2.2360843303810798e-05, "loss": 0.0194, "step": 56170 }, { "grad_norm": 0.24766935408115387, "learning_rate": 2.2343622190876084e-05, "loss": 0.0162, "step": 56180 }, { "grad_norm": 0.2541278302669525, "learning_rate": 2.2326405803659935e-05, "loss": 0.0182, "step": 56190 }, { "grad_norm": 0.24322113394737244, "learning_rate": 2.230919414510416e-05, "loss": 0.0174, "step": 56200 }, { "grad_norm": 0.2744081914424896, "learning_rate": 2.229198721814976e-05, "loss": 0.0195, "step": 56210 }, { "grad_norm": 0.2689063847064972, "learning_rate": 2.2274785025736948e-05, "loss": 0.0169, "step": 56220 }, { "grad_norm": 0.23715749382972717, "learning_rate": 2.225758757080507e-05, "loss": 0.0219, "step": 56230 }, { "grad_norm": 0.2525266706943512, "learning_rate": 2.2240394856292723e-05, "loss": 0.0197, "step": 56240 }, { "grad_norm": 0.20739230513572693, "learning_rate": 2.2223206885137664e-05, "loss": 0.0153, "step": 56250 }, { "grad_norm": 0.23740626871585846, "learning_rate": 2.2206023660276853e-05, "loss": 0.0201, "step": 56260 }, { "grad_norm": 0.219722181558609, "learning_rate": 2.218884518464645e-05, "loss": 0.0198, "step": 56270 }, { "grad_norm": 0.23104092478752136, "learning_rate": 2.2171671461181732e-05, "loss": 0.021, "step": 56280 }, { "grad_norm": 0.1988341361284256, "learning_rate": 2.2154502492817292e-05, "loss": 0.0198, "step": 56290 }, { "grad_norm": 0.22768400609493256, "learning_rate": 2.2137338282486782e-05, "loss": 0.0154, "step": 56300 }, { "grad_norm": 0.18104618787765503, "learning_rate": 2.2120178833123113e-05, "loss": 0.0181, "step": 56310 }, { "grad_norm": 0.24669800698757172, "learning_rate": 2.210302414765838e-05, "loss": 0.0172, "step": 56320 }, { "grad_norm": 0.22803571820259094, "learning_rate": 2.2085874229023808e-05, "loss": 0.0161, "step": 56330 }, { "grad_norm": 0.3350299298763275, "learning_rate": 2.2068729080149907e-05, "loss": 0.0165, "step": 56340 }, { "grad_norm": 0.2331397980451584, "learning_rate": 2.205158870396625e-05, "loss": 0.0201, "step": 56350 }, { "grad_norm": 0.24624258279800415, "learning_rate": 2.2034453103401732e-05, "loss": 0.017, "step": 56360 }, { "grad_norm": 0.16875596344470978, "learning_rate": 2.20173222813843e-05, "loss": 0.0122, "step": 56370 }, { "grad_norm": 0.18145522475242615, "learning_rate": 2.200019624084116e-05, "loss": 0.0185, "step": 56380 }, { "grad_norm": 0.2003035992383957, "learning_rate": 2.1983074984698687e-05, "loss": 0.0153, "step": 56390 }, { "grad_norm": 0.15864725410938263, "learning_rate": 2.1965958515882433e-05, "loss": 0.0214, "step": 56400 }, { "grad_norm": 0.15032584965229034, "learning_rate": 2.1948846837317162e-05, "loss": 0.0149, "step": 56410 }, { "grad_norm": 0.23387598991394043, "learning_rate": 2.1931739951926738e-05, "loss": 0.0141, "step": 56420 }, { "grad_norm": 0.24090246856212616, "learning_rate": 2.191463786263429e-05, "loss": 0.0137, "step": 56430 }, { "grad_norm": 0.22037525475025177, "learning_rate": 2.1897540572362095e-05, "loss": 0.0177, "step": 56440 }, { "grad_norm": 0.2028399109840393, "learning_rate": 2.1880448084031614e-05, "loss": 0.0163, "step": 56450 }, { "grad_norm": 0.16401690244674683, "learning_rate": 2.1863360400563482e-05, "loss": 0.017, "step": 56460 }, { "grad_norm": 0.2159382551908493, "learning_rate": 2.184627752487754e-05, "loss": 0.0178, "step": 56470 }, { "grad_norm": 0.18532989919185638, "learning_rate": 2.1829199459892725e-05, "loss": 0.0163, "step": 56480 }, { "grad_norm": 0.1825369894504547, "learning_rate": 2.1812126208527282e-05, "loss": 0.0184, "step": 56490 }, { "grad_norm": 0.1444239467382431, "learning_rate": 2.179505777369852e-05, "loss": 0.0189, "step": 56500 }, { "grad_norm": 0.16006416082382202, "learning_rate": 2.1777994158322974e-05, "loss": 0.0146, "step": 56510 }, { "grad_norm": 0.29755982756614685, "learning_rate": 2.176093536531637e-05, "loss": 0.0224, "step": 56520 }, { "grad_norm": 0.14472399652004242, "learning_rate": 2.1743881397593537e-05, "loss": 0.0183, "step": 56530 }, { "grad_norm": 0.2795226573944092, "learning_rate": 2.1726832258068595e-05, "loss": 0.0251, "step": 56540 }, { "grad_norm": 0.2590635418891907, "learning_rate": 2.170978794965472e-05, "loss": 0.019, "step": 56550 }, { "grad_norm": 0.19270694255828857, "learning_rate": 2.169274847526438e-05, "loss": 0.0156, "step": 56560 }, { "grad_norm": 0.24724073708057404, "learning_rate": 2.1675713837809103e-05, "loss": 0.0184, "step": 56570 }, { "grad_norm": 0.2668430209159851, "learning_rate": 2.1658684040199655e-05, "loss": 0.0167, "step": 56580 }, { "grad_norm": 0.18767262995243073, "learning_rate": 2.1641659085345974e-05, "loss": 0.0142, "step": 56590 }, { "grad_norm": 0.18700198829174042, "learning_rate": 2.1624638976157154e-05, "loss": 0.015, "step": 56600 }, { "grad_norm": 0.18111002445220947, "learning_rate": 2.1607623715541476e-05, "loss": 0.0229, "step": 56610 }, { "grad_norm": 0.2879781723022461, "learning_rate": 2.159061330640636e-05, "loss": 0.0164, "step": 56620 }, { "grad_norm": 0.16044387221336365, "learning_rate": 2.1573607751658425e-05, "loss": 0.0149, "step": 56630 }, { "grad_norm": 0.1836564987897873, "learning_rate": 2.155660705420347e-05, "loss": 0.0184, "step": 56640 }, { "grad_norm": 0.1995168924331665, "learning_rate": 2.153961121694644e-05, "loss": 0.0162, "step": 56650 }, { "grad_norm": 0.1995195895433426, "learning_rate": 2.152262024279145e-05, "loss": 0.0251, "step": 56660 }, { "grad_norm": 0.3275933265686035, "learning_rate": 2.150563413464183e-05, "loss": 0.0184, "step": 56670 }, { "grad_norm": 0.19510020315647125, "learning_rate": 2.1488652895399974e-05, "loss": 0.0147, "step": 56680 }, { "grad_norm": 0.24517162144184113, "learning_rate": 2.1471676527967587e-05, "loss": 0.0165, "step": 56690 }, { "grad_norm": 0.17483170330524445, "learning_rate": 2.1454705035245414e-05, "loss": 0.0204, "step": 56700 }, { "grad_norm": 0.36119750142097473, "learning_rate": 2.143773842013343e-05, "loss": 0.015, "step": 56710 }, { "grad_norm": 0.24292953312397003, "learning_rate": 2.1420776685530796e-05, "loss": 0.0186, "step": 56720 }, { "grad_norm": 0.3049335181713104, "learning_rate": 2.1403819834335742e-05, "loss": 0.015, "step": 56730 }, { "grad_norm": 0.25915583968162537, "learning_rate": 2.1386867869445808e-05, "loss": 0.0203, "step": 56740 }, { "grad_norm": 0.159223273396492, "learning_rate": 2.1369920793757548e-05, "loss": 0.0238, "step": 56750 }, { "grad_norm": 0.16870437562465668, "learning_rate": 2.1352978610166828e-05, "loss": 0.0157, "step": 56760 }, { "grad_norm": 0.24649769067764282, "learning_rate": 2.1336041321568546e-05, "loss": 0.0195, "step": 56770 }, { "grad_norm": 0.23693165183067322, "learning_rate": 2.131910893085684e-05, "loss": 0.0146, "step": 56780 }, { "grad_norm": 0.24442905187606812, "learning_rate": 2.1302181440925012e-05, "loss": 0.0189, "step": 56790 }, { "grad_norm": 0.1985820084810257, "learning_rate": 2.128525885466546e-05, "loss": 0.0163, "step": 56800 }, { "grad_norm": 0.2554973363876343, "learning_rate": 2.1268341174969847e-05, "loss": 0.0197, "step": 56810 }, { "grad_norm": 0.13974152505397797, "learning_rate": 2.1251428404728894e-05, "loss": 0.017, "step": 56820 }, { "grad_norm": 0.17161674797534943, "learning_rate": 2.123452054683256e-05, "loss": 0.0164, "step": 56830 }, { "grad_norm": 0.22503986954689026, "learning_rate": 2.1217617604169927e-05, "loss": 0.0224, "step": 56840 }, { "grad_norm": 0.21221202611923218, "learning_rate": 2.120071957962924e-05, "loss": 0.0198, "step": 56850 }, { "grad_norm": 0.16068144142627716, "learning_rate": 2.1183826476097917e-05, "loss": 0.0209, "step": 56860 }, { "grad_norm": 0.19016337394714355, "learning_rate": 2.116693829646254e-05, "loss": 0.0161, "step": 56870 }, { "grad_norm": 0.22125039994716644, "learning_rate": 2.1150055043608806e-05, "loss": 0.0156, "step": 56880 }, { "grad_norm": 0.24327658116817474, "learning_rate": 2.113317672042162e-05, "loss": 0.0222, "step": 56890 }, { "grad_norm": 0.19409623742103577, "learning_rate": 2.1116303329785025e-05, "loss": 0.0165, "step": 56900 }, { "grad_norm": 0.2425948977470398, "learning_rate": 2.109943487458222e-05, "loss": 0.0206, "step": 56910 }, { "grad_norm": 0.24750380218029022, "learning_rate": 2.1082571357695574e-05, "loss": 0.0178, "step": 56920 }, { "grad_norm": 0.1568816900253296, "learning_rate": 2.1065712782006557e-05, "loss": 0.017, "step": 56930 }, { "grad_norm": 0.22737626731395721, "learning_rate": 2.104885915039591e-05, "loss": 0.0154, "step": 56940 }, { "grad_norm": 0.1953335553407669, "learning_rate": 2.103201046574338e-05, "loss": 0.0174, "step": 56950 }, { "grad_norm": 0.25084954500198364, "learning_rate": 2.1015166730928022e-05, "loss": 0.0155, "step": 56960 }, { "grad_norm": 0.23863285779953003, "learning_rate": 2.0998327948827912e-05, "loss": 0.0182, "step": 56970 }, { "grad_norm": 0.19262301921844482, "learning_rate": 2.0981494122320363e-05, "loss": 0.0204, "step": 56980 }, { "grad_norm": 0.22124211490154266, "learning_rate": 2.0964665254281822e-05, "loss": 0.0191, "step": 56990 }, { "grad_norm": 0.18996091187000275, "learning_rate": 2.094784134758784e-05, "loss": 0.0178, "step": 57000 }, { "grad_norm": 0.22347335517406464, "learning_rate": 2.0931022405113226e-05, "loss": 0.0232, "step": 57010 }, { "grad_norm": 0.2277429848909378, "learning_rate": 2.091420842973183e-05, "loss": 0.0212, "step": 57020 }, { "grad_norm": 0.1870007961988449, "learning_rate": 2.0897399424316715e-05, "loss": 0.0151, "step": 57030 }, { "grad_norm": 0.22509095072746277, "learning_rate": 2.0880595391740078e-05, "loss": 0.0181, "step": 57040 }, { "grad_norm": 0.2270914912223816, "learning_rate": 2.0863796334873277e-05, "loss": 0.0208, "step": 57050 }, { "grad_norm": 0.13467389345169067, "learning_rate": 2.08470022565868e-05, "loss": 0.0165, "step": 57060 }, { "grad_norm": 0.13282513618469238, "learning_rate": 2.0830213159750317e-05, "loss": 0.0141, "step": 57070 }, { "grad_norm": 0.251415491104126, "learning_rate": 2.0813429047232596e-05, "loss": 0.0177, "step": 57080 }, { "grad_norm": 0.19237083196640015, "learning_rate": 2.0796649921901594e-05, "loss": 0.0182, "step": 57090 }, { "grad_norm": 0.18180496990680695, "learning_rate": 2.077987578662441e-05, "loss": 0.0163, "step": 57100 }, { "grad_norm": 0.25552114844322205, "learning_rate": 2.0763106644267277e-05, "loss": 0.0221, "step": 57110 }, { "grad_norm": 0.23602883517742157, "learning_rate": 2.0746342497695607e-05, "loss": 0.0132, "step": 57120 }, { "grad_norm": 0.2627354860305786, "learning_rate": 2.0729583349773886e-05, "loss": 0.0147, "step": 57130 }, { "grad_norm": 0.2517901360988617, "learning_rate": 2.0712829203365853e-05, "loss": 0.0161, "step": 57140 }, { "grad_norm": 0.2079482227563858, "learning_rate": 2.0696080061334267e-05, "loss": 0.0162, "step": 57150 }, { "grad_norm": 0.19217759370803833, "learning_rate": 2.067933592654117e-05, "loss": 0.0163, "step": 57160 }, { "grad_norm": 0.22409681975841522, "learning_rate": 2.066259680184763e-05, "loss": 0.0181, "step": 57170 }, { "grad_norm": 0.2464347630739212, "learning_rate": 2.0645862690113908e-05, "loss": 0.0196, "step": 57180 }, { "grad_norm": 0.16110877692699432, "learning_rate": 2.0629133594199436e-05, "loss": 0.0179, "step": 57190 }, { "grad_norm": 0.2991843521595001, "learning_rate": 2.0612409516962704e-05, "loss": 0.0169, "step": 57200 }, { "grad_norm": 0.2782362699508667, "learning_rate": 2.0595690461261467e-05, "loss": 0.0174, "step": 57210 }, { "grad_norm": 0.25403282046318054, "learning_rate": 2.0578976429952503e-05, "loss": 0.0171, "step": 57220 }, { "grad_norm": 0.2429574728012085, "learning_rate": 2.0562267425891802e-05, "loss": 0.0185, "step": 57230 }, { "grad_norm": 0.17948471009731293, "learning_rate": 2.0545563451934467e-05, "loss": 0.0184, "step": 57240 }, { "grad_norm": 0.30679234862327576, "learning_rate": 2.0528864510934764e-05, "loss": 0.0218, "step": 57250 }, { "grad_norm": 0.22737878561019897, "learning_rate": 2.0512170605746096e-05, "loss": 0.0254, "step": 57260 }, { "grad_norm": 0.13044749200344086, "learning_rate": 2.0495481739220963e-05, "loss": 0.0197, "step": 57270 }, { "grad_norm": 0.16057957708835602, "learning_rate": 2.0478797914211045e-05, "loss": 0.0206, "step": 57280 }, { "grad_norm": 0.22530518472194672, "learning_rate": 2.046211913356716e-05, "loss": 0.021, "step": 57290 }, { "grad_norm": 0.20706066489219666, "learning_rate": 2.0445445400139247e-05, "loss": 0.0222, "step": 57300 }, { "grad_norm": 0.3057023584842682, "learning_rate": 2.0428776716776405e-05, "loss": 0.018, "step": 57310 }, { "grad_norm": 0.26193249225616455, "learning_rate": 2.0412113086326856e-05, "loss": 0.0227, "step": 57320 }, { "grad_norm": 0.21698370575904846, "learning_rate": 2.0395454511637918e-05, "loss": 0.0159, "step": 57330 }, { "grad_norm": 0.23403963446617126, "learning_rate": 2.037880099555616e-05, "loss": 0.019, "step": 57340 }, { "grad_norm": 0.23526032269001007, "learning_rate": 2.0362152540927144e-05, "loss": 0.0159, "step": 57350 }, { "grad_norm": 0.1842561811208725, "learning_rate": 2.0345509150595666e-05, "loss": 0.029, "step": 57360 }, { "grad_norm": 0.1487543284893036, "learning_rate": 2.0328870827405617e-05, "loss": 0.0183, "step": 57370 }, { "grad_norm": 0.1970384120941162, "learning_rate": 2.0312237574200043e-05, "loss": 0.0189, "step": 57380 }, { "grad_norm": 0.2931434214115143, "learning_rate": 2.029560939382112e-05, "loss": 0.0169, "step": 57390 }, { "grad_norm": 0.2389085739850998, "learning_rate": 2.0278986289110097e-05, "loss": 0.0156, "step": 57400 }, { "grad_norm": 0.19440418481826782, "learning_rate": 2.0262368262907484e-05, "loss": 0.016, "step": 57410 }, { "grad_norm": 0.20942777395248413, "learning_rate": 2.024575531805279e-05, "loss": 0.0219, "step": 57420 }, { "grad_norm": 0.23926004767417908, "learning_rate": 2.0229147457384735e-05, "loss": 0.0151, "step": 57430 }, { "grad_norm": 0.17945758998394012, "learning_rate": 2.0212544683741157e-05, "loss": 0.0318, "step": 57440 }, { "grad_norm": 0.19688041508197784, "learning_rate": 2.0195946999958976e-05, "loss": 0.0229, "step": 57450 }, { "grad_norm": 0.2212364822626114, "learning_rate": 2.017935440887434e-05, "loss": 0.0209, "step": 57460 }, { "grad_norm": 0.21998314559459686, "learning_rate": 2.0162766913322423e-05, "loss": 0.0153, "step": 57470 }, { "grad_norm": 0.24715067446231842, "learning_rate": 2.0146184516137588e-05, "loss": 0.0196, "step": 57480 }, { "grad_norm": 0.17814770340919495, "learning_rate": 2.012960722015332e-05, "loss": 0.0188, "step": 57490 }, { "grad_norm": 0.21590064465999603, "learning_rate": 2.0113035028202214e-05, "loss": 0.0165, "step": 57500 }, { "grad_norm": 0.19727763533592224, "learning_rate": 2.009646794311602e-05, "loss": 0.0213, "step": 57510 }, { "grad_norm": 0.1082933321595192, "learning_rate": 2.007990596772559e-05, "loss": 0.019, "step": 57520 }, { "grad_norm": 0.21150515973567963, "learning_rate": 2.0063349104860923e-05, "loss": 0.0178, "step": 57530 }, { "grad_norm": 0.2692560851573944, "learning_rate": 2.0046797357351116e-05, "loss": 0.027, "step": 57540 }, { "grad_norm": 0.19495384395122528, "learning_rate": 2.0030250728024412e-05, "loss": 0.0146, "step": 57550 }, { "grad_norm": 0.15862195193767548, "learning_rate": 2.001370921970819e-05, "loss": 0.0136, "step": 57560 }, { "grad_norm": 0.1480596959590912, "learning_rate": 1.9997172835228932e-05, "loss": 0.0154, "step": 57570 }, { "grad_norm": 0.23352620005607605, "learning_rate": 1.9980641577412262e-05, "loss": 0.018, "step": 57580 }, { "grad_norm": 0.2107628583908081, "learning_rate": 1.9964115449082925e-05, "loss": 0.0196, "step": 57590 }, { "grad_norm": 0.30103129148483276, "learning_rate": 1.9947594453064742e-05, "loss": 0.0199, "step": 57600 }, { "grad_norm": 0.24339617788791656, "learning_rate": 1.9931078592180774e-05, "loss": 0.0185, "step": 57610 }, { "grad_norm": 0.21860340237617493, "learning_rate": 1.9914567869253065e-05, "loss": 0.0155, "step": 57620 }, { "grad_norm": 0.2807215750217438, "learning_rate": 1.989806228710287e-05, "loss": 0.0253, "step": 57630 }, { "grad_norm": 0.24835330247879028, "learning_rate": 1.9881561848550555e-05, "loss": 0.016, "step": 57640 }, { "grad_norm": 0.23091250658035278, "learning_rate": 1.9865066556415544e-05, "loss": 0.021, "step": 57650 }, { "grad_norm": 0.21172161400318146, "learning_rate": 1.98485764135165e-05, "loss": 0.0174, "step": 57660 }, { "grad_norm": 0.1748986542224884, "learning_rate": 1.983209142267109e-05, "loss": 0.0156, "step": 57670 }, { "grad_norm": 0.19644595682621002, "learning_rate": 1.9815611586696165e-05, "loss": 0.016, "step": 57680 }, { "grad_norm": 0.15396499633789062, "learning_rate": 1.9799136908407667e-05, "loss": 0.0162, "step": 57690 }, { "grad_norm": 0.18091315031051636, "learning_rate": 1.9782667390620678e-05, "loss": 0.0157, "step": 57700 }, { "grad_norm": 0.25599953532218933, "learning_rate": 1.976620303614939e-05, "loss": 0.0155, "step": 57710 }, { "grad_norm": 0.22565503418445587, "learning_rate": 1.9749743847807108e-05, "loss": 0.0156, "step": 57720 }, { "grad_norm": 0.19448113441467285, "learning_rate": 1.9733289828406272e-05, "loss": 0.0167, "step": 57730 }, { "grad_norm": 0.2169630080461502, "learning_rate": 1.9716840980758382e-05, "loss": 0.0184, "step": 57740 }, { "grad_norm": 0.1865512728691101, "learning_rate": 1.9700397307674134e-05, "loss": 0.0168, "step": 57750 }, { "grad_norm": 0.2516012489795685, "learning_rate": 1.968395881196328e-05, "loss": 0.0206, "step": 57760 }, { "grad_norm": 0.19871248304843903, "learning_rate": 1.966752549643473e-05, "loss": 0.019, "step": 57770 }, { "grad_norm": 0.25830116868019104, "learning_rate": 1.965109736389647e-05, "loss": 0.0154, "step": 57780 }, { "grad_norm": 0.15963050723075867, "learning_rate": 1.9634674417155645e-05, "loss": 0.0156, "step": 57790 }, { "grad_norm": 0.2943350672721863, "learning_rate": 1.9618256659018434e-05, "loss": 0.0344, "step": 57800 }, { "grad_norm": 0.3164530098438263, "learning_rate": 1.9601844092290257e-05, "loss": 0.0201, "step": 57810 }, { "grad_norm": 0.19763383269309998, "learning_rate": 1.9585436719775512e-05, "loss": 0.0164, "step": 57820 }, { "grad_norm": 0.20183135569095612, "learning_rate": 1.9569034544277793e-05, "loss": 0.018, "step": 57830 }, { "grad_norm": 0.22800874710083008, "learning_rate": 1.9552637568599798e-05, "loss": 0.0168, "step": 57840 }, { "grad_norm": 0.1801283061504364, "learning_rate": 1.953624579554327e-05, "loss": 0.0251, "step": 57850 }, { "grad_norm": 0.19465988874435425, "learning_rate": 1.951985922790918e-05, "loss": 0.0222, "step": 57860 }, { "grad_norm": 0.18107886612415314, "learning_rate": 1.9503477868497505e-05, "loss": 0.0147, "step": 57870 }, { "grad_norm": 0.1758020520210266, "learning_rate": 1.9487101720107375e-05, "loss": 0.021, "step": 57880 }, { "grad_norm": 0.16170410811901093, "learning_rate": 1.9470730785537032e-05, "loss": 0.0169, "step": 57890 }, { "grad_norm": 0.2610560655593872, "learning_rate": 1.9454365067583823e-05, "loss": 0.0223, "step": 57900 }, { "grad_norm": 0.17359745502471924, "learning_rate": 1.9438004569044215e-05, "loss": 0.0135, "step": 57910 }, { "grad_norm": 0.17614054679870605, "learning_rate": 1.9421649292713724e-05, "loss": 0.0334, "step": 57920 }, { "grad_norm": 0.2590886354446411, "learning_rate": 1.9405299241387076e-05, "loss": 0.0201, "step": 57930 }, { "grad_norm": 0.20722678303718567, "learning_rate": 1.9388954417858007e-05, "loss": 0.0152, "step": 57940 }, { "grad_norm": 0.46190163493156433, "learning_rate": 1.9372614824919417e-05, "loss": 0.0177, "step": 57950 }, { "grad_norm": 0.2082720249891281, "learning_rate": 1.9356280465363284e-05, "loss": 0.0221, "step": 57960 }, { "grad_norm": 0.17020995914936066, "learning_rate": 1.9339951341980723e-05, "loss": 0.0161, "step": 57970 }, { "grad_norm": 0.271916002035141, "learning_rate": 1.9323627457561916e-05, "loss": 0.0171, "step": 57980 }, { "grad_norm": 0.2124875783920288, "learning_rate": 1.9307308814896198e-05, "loss": 0.0138, "step": 57990 }, { "grad_norm": 0.1797889918088913, "learning_rate": 1.9290995416771935e-05, "loss": 0.0149, "step": 58000 }, { "grad_norm": 0.17163711786270142, "learning_rate": 1.9274687265976665e-05, "loss": 0.0176, "step": 58010 }, { "grad_norm": 0.18809467554092407, "learning_rate": 1.9258384365297e-05, "loss": 0.0195, "step": 58020 }, { "grad_norm": 0.23348762094974518, "learning_rate": 1.924208671751866e-05, "loss": 0.0147, "step": 58030 }, { "grad_norm": 0.326587438583374, "learning_rate": 1.9225794325426492e-05, "loss": 0.0199, "step": 58040 }, { "grad_norm": 0.19508373737335205, "learning_rate": 1.920950719180436e-05, "loss": 0.0168, "step": 58050 }, { "grad_norm": 0.09508918225765228, "learning_rate": 1.919322531943536e-05, "loss": 0.0169, "step": 58060 }, { "grad_norm": 0.18730846047401428, "learning_rate": 1.917694871110157e-05, "loss": 0.0171, "step": 58070 }, { "grad_norm": 0.22782756388187408, "learning_rate": 1.9160677369584234e-05, "loss": 0.0159, "step": 58080 }, { "grad_norm": 0.21315981447696686, "learning_rate": 1.9144411297663694e-05, "loss": 0.0156, "step": 58090 }, { "grad_norm": 0.15446287393569946, "learning_rate": 1.9128150498119328e-05, "loss": 0.0183, "step": 58100 }, { "grad_norm": 0.21294529736042023, "learning_rate": 1.9111894973729726e-05, "loss": 0.0152, "step": 58110 }, { "grad_norm": 0.2211812436580658, "learning_rate": 1.9095644727272454e-05, "loss": 0.0165, "step": 58120 }, { "grad_norm": 0.23996654152870178, "learning_rate": 1.907939976152429e-05, "loss": 0.0196, "step": 58130 }, { "grad_norm": 0.29005539417266846, "learning_rate": 1.906316007926101e-05, "loss": 0.0207, "step": 58140 }, { "grad_norm": 0.20557531714439392, "learning_rate": 1.904692568325755e-05, "loss": 0.0168, "step": 58150 }, { "grad_norm": 0.26180070638656616, "learning_rate": 1.9030696576287925e-05, "loss": 0.0189, "step": 58160 }, { "grad_norm": 0.22042164206504822, "learning_rate": 1.9014472761125242e-05, "loss": 0.0146, "step": 58170 }, { "grad_norm": 0.28157612681388855, "learning_rate": 1.899825424054172e-05, "loss": 0.0202, "step": 58180 }, { "grad_norm": 0.36719784140586853, "learning_rate": 1.898204101730863e-05, "loss": 0.0172, "step": 58190 }, { "grad_norm": 0.1479831486940384, "learning_rate": 1.8965833094196394e-05, "loss": 0.0289, "step": 58200 }, { "grad_norm": 0.21011090278625488, "learning_rate": 1.8949630473974495e-05, "loss": 0.0156, "step": 58210 }, { "grad_norm": 0.24694859981536865, "learning_rate": 1.8933433159411517e-05, "loss": 0.0189, "step": 58220 }, { "grad_norm": 0.33725976943969727, "learning_rate": 1.891724115327514e-05, "loss": 0.0187, "step": 58230 }, { "grad_norm": 0.1701289862394333, "learning_rate": 1.8901054458332152e-05, "loss": 0.0208, "step": 58240 }, { "grad_norm": 0.20813336968421936, "learning_rate": 1.8884873077348364e-05, "loss": 0.0224, "step": 58250 }, { "grad_norm": 0.21481840312480927, "learning_rate": 1.88686970130888e-05, "loss": 0.0204, "step": 58260 }, { "grad_norm": 0.1820106953382492, "learning_rate": 1.8852526268317456e-05, "loss": 0.0198, "step": 58270 }, { "grad_norm": 0.20683196187019348, "learning_rate": 1.883636084579749e-05, "loss": 0.0174, "step": 58280 }, { "grad_norm": 0.2712518870830536, "learning_rate": 1.8820200748291145e-05, "loss": 0.0209, "step": 58290 }, { "grad_norm": 0.22229307889938354, "learning_rate": 1.8804045978559686e-05, "loss": 0.0189, "step": 58300 }, { "grad_norm": 0.2225358784198761, "learning_rate": 1.8787896539363594e-05, "loss": 0.0182, "step": 58310 }, { "grad_norm": 0.22330395877361298, "learning_rate": 1.8771752433462298e-05, "loss": 0.0166, "step": 58320 }, { "grad_norm": 0.24545025825500488, "learning_rate": 1.8755613663614447e-05, "loss": 0.0194, "step": 58330 }, { "grad_norm": 0.2579481899738312, "learning_rate": 1.873948023257767e-05, "loss": 0.0184, "step": 58340 }, { "grad_norm": 0.2471776306629181, "learning_rate": 1.8723352143108747e-05, "loss": 0.0172, "step": 58350 }, { "grad_norm": 0.2624565362930298, "learning_rate": 1.870722939796352e-05, "loss": 0.0176, "step": 58360 }, { "grad_norm": 0.24638350307941437, "learning_rate": 1.869111199989693e-05, "loss": 0.0177, "step": 58370 }, { "grad_norm": 0.18288730084896088, "learning_rate": 1.8674999951663018e-05, "loss": 0.0196, "step": 58380 }, { "grad_norm": 0.16010835766792297, "learning_rate": 1.8658893256014857e-05, "loss": 0.0205, "step": 58390 }, { "grad_norm": 0.28509196639060974, "learning_rate": 1.8642791915704655e-05, "loss": 0.0152, "step": 58400 }, { "grad_norm": 0.23088431358337402, "learning_rate": 1.8626695933483697e-05, "loss": 0.0161, "step": 58410 }, { "grad_norm": 0.17605999112129211, "learning_rate": 1.8610605312102335e-05, "loss": 0.0186, "step": 58420 }, { "grad_norm": 0.37842074036598206, "learning_rate": 1.8594520054310032e-05, "loss": 0.0248, "step": 58430 }, { "grad_norm": 0.15788722038269043, "learning_rate": 1.8578440162855326e-05, "loss": 0.0187, "step": 58440 }, { "grad_norm": 0.22867533564567566, "learning_rate": 1.8562365640485774e-05, "loss": 0.0205, "step": 58450 }, { "grad_norm": 0.17890256643295288, "learning_rate": 1.854629648994815e-05, "loss": 0.013, "step": 58460 }, { "grad_norm": 0.1600891351699829, "learning_rate": 1.8530232713988183e-05, "loss": 0.0161, "step": 58470 }, { "grad_norm": 0.25214388966560364, "learning_rate": 1.851417431535074e-05, "loss": 0.0173, "step": 58480 }, { "grad_norm": 0.2544655203819275, "learning_rate": 1.8498121296779785e-05, "loss": 0.0287, "step": 58490 }, { "grad_norm": 0.25022420287132263, "learning_rate": 1.848207366101829e-05, "loss": 0.0234, "step": 58500 }, { "grad_norm": 0.22710062563419342, "learning_rate": 1.8466031410808422e-05, "loss": 0.0177, "step": 58510 }, { "grad_norm": 0.19443339109420776, "learning_rate": 1.844999454889129e-05, "loss": 0.0222, "step": 58520 }, { "grad_norm": 0.19594889879226685, "learning_rate": 1.8433963078007237e-05, "loss": 0.022, "step": 58530 }, { "grad_norm": 0.1637980043888092, "learning_rate": 1.841793700089554e-05, "loss": 0.0184, "step": 58540 }, { "grad_norm": 0.21911051869392395, "learning_rate": 1.8401916320294644e-05, "loss": 0.0205, "step": 58550 }, { "grad_norm": 0.19493557512760162, "learning_rate": 1.838590103894205e-05, "loss": 0.0158, "step": 58560 }, { "grad_norm": 0.1926119476556778, "learning_rate": 1.8369891159574292e-05, "loss": 0.0156, "step": 58570 }, { "grad_norm": 0.20092914998531342, "learning_rate": 1.835388668492708e-05, "loss": 0.0208, "step": 58580 }, { "grad_norm": 0.20334585011005402, "learning_rate": 1.8337887617735095e-05, "loss": 0.0191, "step": 58590 }, { "grad_norm": 0.2526811957359314, "learning_rate": 1.8321893960732157e-05, "loss": 0.0167, "step": 58600 }, { "grad_norm": 0.2466556429862976, "learning_rate": 1.8305905716651138e-05, "loss": 0.0176, "step": 58610 }, { "grad_norm": 0.20445303618907928, "learning_rate": 1.8289922888224e-05, "loss": 0.0174, "step": 58620 }, { "grad_norm": 0.3157833218574524, "learning_rate": 1.8273945478181765e-05, "loss": 0.0221, "step": 58630 }, { "grad_norm": 0.31708645820617676, "learning_rate": 1.8257973489254558e-05, "loss": 0.0156, "step": 58640 }, { "grad_norm": 0.12850482761859894, "learning_rate": 1.824200692417152e-05, "loss": 0.0183, "step": 58650 }, { "grad_norm": 0.18467694520950317, "learning_rate": 1.8226045785660912e-05, "loss": 0.0171, "step": 58660 }, { "grad_norm": 0.2529491186141968, "learning_rate": 1.821009007645006e-05, "loss": 0.0192, "step": 58670 }, { "grad_norm": 0.13335716724395752, "learning_rate": 1.8194139799265357e-05, "loss": 0.0141, "step": 58680 }, { "grad_norm": 0.2898976504802704, "learning_rate": 1.8178194956832295e-05, "loss": 0.0197, "step": 58690 }, { "grad_norm": 0.14816914498806, "learning_rate": 1.8162255551875346e-05, "loss": 0.0162, "step": 58700 }, { "grad_norm": 0.23040492832660675, "learning_rate": 1.8146321587118194e-05, "loss": 0.0225, "step": 58710 }, { "grad_norm": 0.18230867385864258, "learning_rate": 1.8130393065283448e-05, "loss": 0.0133, "step": 58720 }, { "grad_norm": 0.1495443880558014, "learning_rate": 1.8114469989092925e-05, "loss": 0.017, "step": 58730 }, { "grad_norm": 0.21357710659503937, "learning_rate": 1.8098552361267397e-05, "loss": 0.0144, "step": 58740 }, { "grad_norm": 0.17690430581569672, "learning_rate": 1.8082640184526763e-05, "loss": 0.0168, "step": 58750 }, { "grad_norm": 0.1802937537431717, "learning_rate": 1.806673346158999e-05, "loss": 0.0159, "step": 58760 }, { "grad_norm": 0.18017782270908356, "learning_rate": 1.8050832195175067e-05, "loss": 0.0179, "step": 58770 }, { "grad_norm": 0.1958143562078476, "learning_rate": 1.8034936387999136e-05, "loss": 0.017, "step": 58780 }, { "grad_norm": 0.2582966387271881, "learning_rate": 1.8019046042778315e-05, "loss": 0.017, "step": 58790 }, { "grad_norm": 0.16462914645671844, "learning_rate": 1.800316116222785e-05, "loss": 0.0154, "step": 58800 }, { "grad_norm": 0.1660953313112259, "learning_rate": 1.7987281749062018e-05, "loss": 0.0354, "step": 58810 }, { "grad_norm": 0.12031499296426773, "learning_rate": 1.7971407805994195e-05, "loss": 0.0226, "step": 58820 }, { "grad_norm": 0.16289225220680237, "learning_rate": 1.7955539335736787e-05, "loss": 0.0191, "step": 58830 }, { "grad_norm": 0.2015806883573532, "learning_rate": 1.7939676341001304e-05, "loss": 0.0216, "step": 58840 }, { "grad_norm": 0.18642443418502808, "learning_rate": 1.7923818824498275e-05, "loss": 0.0155, "step": 58850 }, { "grad_norm": 0.28504660725593567, "learning_rate": 1.7907966788937315e-05, "loss": 0.0194, "step": 58860 }, { "grad_norm": 0.17413051426410675, "learning_rate": 1.7892120237027116e-05, "loss": 0.0194, "step": 58870 }, { "grad_norm": 0.17765633761882782, "learning_rate": 1.7876279171475413e-05, "loss": 0.0173, "step": 58880 }, { "grad_norm": 0.21027126908302307, "learning_rate": 1.7860443594989028e-05, "loss": 0.0128, "step": 58890 }, { "grad_norm": 0.23481197655200958, "learning_rate": 1.784461351027379e-05, "loss": 0.0182, "step": 58900 }, { "grad_norm": 0.17664632201194763, "learning_rate": 1.7828788920034677e-05, "loss": 0.021, "step": 58910 }, { "grad_norm": 0.24086502194404602, "learning_rate": 1.7812969826975623e-05, "loss": 0.0188, "step": 58920 }, { "grad_norm": 0.1998586803674698, "learning_rate": 1.7797156233799738e-05, "loss": 0.0179, "step": 58930 }, { "grad_norm": 0.16227297484874725, "learning_rate": 1.7781348143209094e-05, "loss": 0.0158, "step": 58940 }, { "grad_norm": 0.15337687730789185, "learning_rate": 1.7765545557904862e-05, "loss": 0.025, "step": 58950 }, { "grad_norm": 0.26501530408859253, "learning_rate": 1.7749748480587302e-05, "loss": 0.0171, "step": 58960 }, { "grad_norm": 0.13282524049282074, "learning_rate": 1.773395691395564e-05, "loss": 0.0131, "step": 58970 }, { "grad_norm": 0.17186444997787476, "learning_rate": 1.7718170860708305e-05, "loss": 0.0127, "step": 58980 }, { "grad_norm": 0.18968814611434937, "learning_rate": 1.770239032354264e-05, "loss": 0.0155, "step": 58990 }, { "grad_norm": 0.24985544383525848, "learning_rate": 1.7686615305155124e-05, "loss": 0.0243, "step": 59000 }, { "grad_norm": 0.168217271566391, "learning_rate": 1.767084580824128e-05, "loss": 0.0133, "step": 59010 }, { "grad_norm": 0.32423582673072815, "learning_rate": 1.765508183549569e-05, "loss": 0.0241, "step": 59020 }, { "grad_norm": 0.19092898070812225, "learning_rate": 1.763932338961199e-05, "loss": 0.017, "step": 59030 }, { "grad_norm": 0.18856219947338104, "learning_rate": 1.762357047328284e-05, "loss": 0.0163, "step": 59040 }, { "grad_norm": 0.294756680727005, "learning_rate": 1.7607823089199997e-05, "loss": 0.0191, "step": 59050 }, { "grad_norm": 0.19564436376094818, "learning_rate": 1.7592081240054265e-05, "loss": 0.0171, "step": 59060 }, { "grad_norm": 0.1742846816778183, "learning_rate": 1.7576344928535488e-05, "loss": 0.0202, "step": 59070 }, { "grad_norm": 0.21140415966510773, "learning_rate": 1.7560614157332573e-05, "loss": 0.0197, "step": 59080 }, { "grad_norm": 0.16089263558387756, "learning_rate": 1.7544888929133495e-05, "loss": 0.0199, "step": 59090 }, { "grad_norm": 0.17064915597438812, "learning_rate": 1.752916924662522e-05, "loss": 0.0184, "step": 59100 }, { "grad_norm": 0.1241026222705841, "learning_rate": 1.751345511249387e-05, "loss": 0.0132, "step": 59110 }, { "grad_norm": 0.16675540804862976, "learning_rate": 1.749774652942452e-05, "loss": 0.0153, "step": 59120 }, { "grad_norm": 0.18357698619365692, "learning_rate": 1.748204350010135e-05, "loss": 0.0175, "step": 59130 }, { "grad_norm": 0.1485394835472107, "learning_rate": 1.7466346027207574e-05, "loss": 0.0148, "step": 59140 }, { "grad_norm": 0.18482257425785065, "learning_rate": 1.745065411342547e-05, "loss": 0.0156, "step": 59150 }, { "grad_norm": 0.21776708960533142, "learning_rate": 1.7434967761436366e-05, "loss": 0.0211, "step": 59160 }, { "grad_norm": 0.2647944986820221, "learning_rate": 1.741928697392058e-05, "loss": 0.0174, "step": 59170 }, { "grad_norm": 0.266257107257843, "learning_rate": 1.7403611753557597e-05, "loss": 0.0153, "step": 59180 }, { "grad_norm": 0.19907236099243164, "learning_rate": 1.738794210302584e-05, "loss": 0.014, "step": 59190 }, { "grad_norm": 0.14609549939632416, "learning_rate": 1.7372278025002837e-05, "loss": 0.013, "step": 59200 }, { "grad_norm": 0.1879860907793045, "learning_rate": 1.7356619522165164e-05, "loss": 0.0176, "step": 59210 }, { "grad_norm": 0.19856412708759308, "learning_rate": 1.7340966597188378e-05, "loss": 0.0175, "step": 59220 }, { "grad_norm": 0.16116510331630707, "learning_rate": 1.732531925274722e-05, "loss": 0.0163, "step": 59230 }, { "grad_norm": 0.22668549418449402, "learning_rate": 1.7309677491515318e-05, "loss": 0.0159, "step": 59240 }, { "grad_norm": 0.2136567234992981, "learning_rate": 1.7294041316165455e-05, "loss": 0.0148, "step": 59250 }, { "grad_norm": 0.2281016707420349, "learning_rate": 1.727841072936942e-05, "loss": 0.0174, "step": 59260 }, { "grad_norm": 0.26021355390548706, "learning_rate": 1.7262785733798058e-05, "loss": 0.0143, "step": 59270 }, { "grad_norm": 0.23153434693813324, "learning_rate": 1.7247166332121246e-05, "loss": 0.0171, "step": 59280 }, { "grad_norm": 0.24640004336833954, "learning_rate": 1.7231552527007933e-05, "loss": 0.0188, "step": 59290 }, { "grad_norm": 0.24898360669612885, "learning_rate": 1.721594432112606e-05, "loss": 0.0154, "step": 59300 }, { "grad_norm": 0.24285303056240082, "learning_rate": 1.7200341717142655e-05, "loss": 0.024, "step": 59310 }, { "grad_norm": 0.2863440215587616, "learning_rate": 1.718474471772377e-05, "loss": 0.0176, "step": 59320 }, { "grad_norm": 0.204206645488739, "learning_rate": 1.7169153325534528e-05, "loss": 0.0152, "step": 59330 }, { "grad_norm": 0.25539925694465637, "learning_rate": 1.7153567543239045e-05, "loss": 0.016, "step": 59340 }, { "grad_norm": 0.11900859326124191, "learning_rate": 1.7137987373500525e-05, "loss": 0.0133, "step": 59350 }, { "grad_norm": 0.21496537327766418, "learning_rate": 1.7122412818981198e-05, "loss": 0.0195, "step": 59360 }, { "grad_norm": 0.25487321615219116, "learning_rate": 1.7106843882342283e-05, "loss": 0.0204, "step": 59370 }, { "grad_norm": 0.1092621237039566, "learning_rate": 1.709128056624415e-05, "loss": 0.0144, "step": 59380 }, { "grad_norm": 0.26630544662475586, "learning_rate": 1.70757228733461e-05, "loss": 0.0148, "step": 59390 }, { "grad_norm": 0.1473165601491928, "learning_rate": 1.706017080630653e-05, "loss": 0.0165, "step": 59400 }, { "grad_norm": 0.30417171120643616, "learning_rate": 1.7044624367782873e-05, "loss": 0.0188, "step": 59410 }, { "grad_norm": 0.21143262088298798, "learning_rate": 1.7029083560431553e-05, "loss": 0.0157, "step": 59420 }, { "grad_norm": 0.30514299869537354, "learning_rate": 1.701354838690813e-05, "loss": 0.0224, "step": 59430 }, { "grad_norm": 0.20556025207042694, "learning_rate": 1.6998018849867087e-05, "loss": 0.0172, "step": 59440 }, { "grad_norm": 0.24087131023406982, "learning_rate": 1.698249495196202e-05, "loss": 0.0219, "step": 59450 }, { "grad_norm": 0.2885900139808655, "learning_rate": 1.6966976695845528e-05, "loss": 0.0212, "step": 59460 }, { "grad_norm": 0.1480933576822281, "learning_rate": 1.6951464084169268e-05, "loss": 0.0134, "step": 59470 }, { "grad_norm": 0.23285110294818878, "learning_rate": 1.6935957119583916e-05, "loss": 0.0203, "step": 59480 }, { "grad_norm": 0.14680659770965576, "learning_rate": 1.6920455804739205e-05, "loss": 0.0153, "step": 59490 }, { "grad_norm": 0.20469722151756287, "learning_rate": 1.6904960142283855e-05, "loss": 0.0182, "step": 59500 }, { "grad_norm": 0.14118777215480804, "learning_rate": 1.6889470134865666e-05, "loss": 0.0142, "step": 59510 }, { "grad_norm": 0.24191632866859436, "learning_rate": 1.687398578513145e-05, "loss": 0.0163, "step": 59520 }, { "grad_norm": 0.17937302589416504, "learning_rate": 1.6858507095727066e-05, "loss": 0.0176, "step": 59530 }, { "grad_norm": 0.2210313379764557, "learning_rate": 1.6843034069297403e-05, "loss": 0.0157, "step": 59540 }, { "grad_norm": 0.2144046276807785, "learning_rate": 1.682756670848637e-05, "loss": 0.0274, "step": 59550 }, { "grad_norm": 0.2032436728477478, "learning_rate": 1.6812105015936937e-05, "loss": 0.0136, "step": 59560 }, { "grad_norm": 0.28537824749946594, "learning_rate": 1.6796648994291027e-05, "loss": 0.0255, "step": 59570 }, { "grad_norm": 0.18756185472011566, "learning_rate": 1.678119864618973e-05, "loss": 0.0134, "step": 59580 }, { "grad_norm": 0.16823667287826538, "learning_rate": 1.6765753974273023e-05, "loss": 0.0132, "step": 59590 }, { "grad_norm": 0.1639021337032318, "learning_rate": 1.675031498118001e-05, "loss": 0.014, "step": 59600 }, { "grad_norm": 0.178565114736557, "learning_rate": 1.6734881669548804e-05, "loss": 0.0131, "step": 59610 }, { "grad_norm": 0.23125305771827698, "learning_rate": 1.671945404201648e-05, "loss": 0.0199, "step": 59620 }, { "grad_norm": 0.25028517842292786, "learning_rate": 1.670403210121927e-05, "loss": 0.0177, "step": 59630 }, { "grad_norm": 0.20266640186309814, "learning_rate": 1.6688615849792312e-05, "loss": 0.0162, "step": 59640 }, { "grad_norm": 0.3023778796195984, "learning_rate": 1.6673205290369832e-05, "loss": 0.02, "step": 59650 }, { "grad_norm": 0.27691707015037537, "learning_rate": 1.6657800425585086e-05, "loss": 0.0193, "step": 59660 }, { "grad_norm": 0.23424799740314484, "learning_rate": 1.6642401258070327e-05, "loss": 0.0284, "step": 59670 }, { "grad_norm": 0.2528204917907715, "learning_rate": 1.6627007790456884e-05, "loss": 0.0258, "step": 59680 }, { "grad_norm": 0.2254931628704071, "learning_rate": 1.6611620025375037e-05, "loss": 0.0143, "step": 59690 }, { "grad_norm": 0.24392841756343842, "learning_rate": 1.6596237965454154e-05, "loss": 0.0226, "step": 59700 }, { "grad_norm": 0.28336864709854126, "learning_rate": 1.6580861613322607e-05, "loss": 0.0203, "step": 59710 }, { "grad_norm": 0.09190506488084793, "learning_rate": 1.6565490971607796e-05, "loss": 0.0129, "step": 59720 }, { "grad_norm": 0.15641140937805176, "learning_rate": 1.655012604293614e-05, "loss": 0.0166, "step": 59730 }, { "grad_norm": 0.21344119310379028, "learning_rate": 1.653476682993309e-05, "loss": 0.0186, "step": 59740 }, { "grad_norm": 0.1707119643688202, "learning_rate": 1.651941333522311e-05, "loss": 0.0155, "step": 59750 }, { "grad_norm": 0.2425418347120285, "learning_rate": 1.6504065561429715e-05, "loss": 0.0208, "step": 59760 }, { "grad_norm": 0.2673836946487427, "learning_rate": 1.6488723511175385e-05, "loss": 0.0239, "step": 59770 }, { "grad_norm": 0.2237272709608078, "learning_rate": 1.6473387187081668e-05, "loss": 0.0158, "step": 59780 }, { "grad_norm": 0.1702231466770172, "learning_rate": 1.6458056591769123e-05, "loss": 0.0152, "step": 59790 }, { "grad_norm": 0.21669620275497437, "learning_rate": 1.6442731727857335e-05, "loss": 0.0144, "step": 59800 }, { "grad_norm": 0.17671480774879456, "learning_rate": 1.642741259796492e-05, "loss": 0.016, "step": 59810 }, { "grad_norm": 0.18274234235286713, "learning_rate": 1.641209920470944e-05, "loss": 0.0181, "step": 59820 }, { "grad_norm": 0.17895644903182983, "learning_rate": 1.639679155070762e-05, "loss": 0.0137, "step": 59830 }, { "grad_norm": 0.21065831184387207, "learning_rate": 1.6381489638575048e-05, "loss": 0.0152, "step": 59840 }, { "grad_norm": 0.15291163325309753, "learning_rate": 1.636619347092643e-05, "loss": 0.0147, "step": 59850 }, { "grad_norm": 0.22873589396476746, "learning_rate": 1.6350903050375476e-05, "loss": 0.0139, "step": 59860 }, { "grad_norm": 0.34435412287712097, "learning_rate": 1.6335618379534856e-05, "loss": 0.0165, "step": 59870 }, { "grad_norm": 0.16093860566616058, "learning_rate": 1.6320339461016364e-05, "loss": 0.015, "step": 59880 }, { "grad_norm": 0.17262324690818787, "learning_rate": 1.6305066297430687e-05, "loss": 0.0161, "step": 59890 }, { "grad_norm": 0.17289112508296967, "learning_rate": 1.6289798891387654e-05, "loss": 0.0205, "step": 59900 }, { "grad_norm": 0.1996227204799652, "learning_rate": 1.6274537245495995e-05, "loss": 0.0142, "step": 59910 }, { "grad_norm": 0.11783894151449203, "learning_rate": 1.6259281362363527e-05, "loss": 0.0149, "step": 59920 }, { "grad_norm": 0.19715481996536255, "learning_rate": 1.6244031244597068e-05, "loss": 0.0204, "step": 59930 }, { "grad_norm": 0.21228821575641632, "learning_rate": 1.622878689480244e-05, "loss": 0.0144, "step": 59940 }, { "grad_norm": 0.27161887288093567, "learning_rate": 1.6213548315584498e-05, "loss": 0.0188, "step": 59950 }, { "grad_norm": 0.22378021478652954, "learning_rate": 1.6198315509547074e-05, "loss": 0.0173, "step": 59960 }, { "grad_norm": 0.16492916643619537, "learning_rate": 1.6183088479293056e-05, "loss": 0.0158, "step": 59970 }, { "grad_norm": 0.18041062355041504, "learning_rate": 1.6167867227424316e-05, "loss": 0.0151, "step": 59980 }, { "grad_norm": 0.23373469710350037, "learning_rate": 1.615265175654176e-05, "loss": 0.0199, "step": 59990 }, { "grad_norm": 0.21182310581207275, "learning_rate": 1.613744206924529e-05, "loss": 0.0178, "step": 60000 }, { "grad_norm": 0.21966154873371124, "learning_rate": 1.6122238168133845e-05, "loss": 0.0191, "step": 60010 }, { "grad_norm": 0.20493148267269135, "learning_rate": 1.6107040055805305e-05, "loss": 0.0157, "step": 60020 }, { "grad_norm": 0.22915548086166382, "learning_rate": 1.6091847734856673e-05, "loss": 0.017, "step": 60030 }, { "grad_norm": 0.2019706815481186, "learning_rate": 1.6076661207883865e-05, "loss": 0.0137, "step": 60040 }, { "grad_norm": 0.19016647338867188, "learning_rate": 1.6061480477481848e-05, "loss": 0.0143, "step": 60050 }, { "grad_norm": 0.2507743239402771, "learning_rate": 1.604630554624461e-05, "loss": 0.0149, "step": 60060 }, { "grad_norm": 0.14958156645298004, "learning_rate": 1.603113641676509e-05, "loss": 0.0151, "step": 60070 }, { "grad_norm": 0.14000657200813293, "learning_rate": 1.6015973091635338e-05, "loss": 0.0133, "step": 60080 }, { "grad_norm": 0.09902694821357727, "learning_rate": 1.6000815573446288e-05, "loss": 0.0175, "step": 60090 }, { "grad_norm": 0.2120448797941208, "learning_rate": 1.5985663864788002e-05, "loss": 0.0239, "step": 60100 }, { "grad_norm": 0.1968858540058136, "learning_rate": 1.597051796824946e-05, "loss": 0.0194, "step": 60110 }, { "grad_norm": 0.2523452043533325, "learning_rate": 1.5955377886418682e-05, "loss": 0.0168, "step": 60120 }, { "grad_norm": 0.2027314454317093, "learning_rate": 1.5940243621882704e-05, "loss": 0.0149, "step": 60130 }, { "grad_norm": 0.19843971729278564, "learning_rate": 1.5925115177227555e-05, "loss": 0.0171, "step": 60140 }, { "grad_norm": 0.2566685676574707, "learning_rate": 1.5909992555038288e-05, "loss": 0.0167, "step": 60150 }, { "grad_norm": 0.2207227498292923, "learning_rate": 1.5894875757898912e-05, "loss": 0.0202, "step": 60160 }, { "grad_norm": 0.1823665052652359, "learning_rate": 1.58797647883925e-05, "loss": 0.0159, "step": 60170 }, { "grad_norm": 0.21002614498138428, "learning_rate": 1.586465964910109e-05, "loss": 0.0203, "step": 60180 }, { "grad_norm": 0.18048329651355743, "learning_rate": 1.5849560342605734e-05, "loss": 0.0173, "step": 60190 }, { "grad_norm": 0.29333850741386414, "learning_rate": 1.583446687148651e-05, "loss": 0.0178, "step": 60200 }, { "grad_norm": 0.29577937722206116, "learning_rate": 1.581937923832248e-05, "loss": 0.0156, "step": 60210 }, { "grad_norm": 0.19710156321525574, "learning_rate": 1.5804297445691663e-05, "loss": 0.015, "step": 60220 }, { "grad_norm": 0.2274140566587448, "learning_rate": 1.578922149617119e-05, "loss": 0.0198, "step": 60230 }, { "grad_norm": 0.17300812900066376, "learning_rate": 1.5774151392337084e-05, "loss": 0.0142, "step": 60240 }, { "grad_norm": 0.24006855487823486, "learning_rate": 1.575908713676442e-05, "loss": 0.0205, "step": 60250 }, { "grad_norm": 0.23669449985027313, "learning_rate": 1.5744028732027293e-05, "loss": 0.0183, "step": 60260 }, { "grad_norm": 0.18913687765598297, "learning_rate": 1.5728976180698723e-05, "loss": 0.0166, "step": 60270 }, { "grad_norm": 0.23898546397686005, "learning_rate": 1.5713929485350836e-05, "loss": 0.0176, "step": 60280 }, { "grad_norm": 0.215615913271904, "learning_rate": 1.569888864855465e-05, "loss": 0.0172, "step": 60290 }, { "grad_norm": 0.19646425545215607, "learning_rate": 1.568385367288028e-05, "loss": 0.014, "step": 60300 }, { "grad_norm": 0.17886768281459808, "learning_rate": 1.5668824560896755e-05, "loss": 0.0168, "step": 60310 }, { "grad_norm": 0.235406756401062, "learning_rate": 1.5653801315172156e-05, "loss": 0.0149, "step": 60320 }, { "grad_norm": 0.21598698198795319, "learning_rate": 1.5638783938273554e-05, "loss": 0.0157, "step": 60330 }, { "grad_norm": 0.1470567137002945, "learning_rate": 1.5623772432766966e-05, "loss": 0.0203, "step": 60340 }, { "grad_norm": 0.2190689742565155, "learning_rate": 1.5608766801217507e-05, "loss": 0.0168, "step": 60350 }, { "grad_norm": 0.16438117623329163, "learning_rate": 1.5593767046189184e-05, "loss": 0.0131, "step": 60360 }, { "grad_norm": 0.2147335708141327, "learning_rate": 1.5578773170245064e-05, "loss": 0.0145, "step": 60370 }, { "grad_norm": 0.23567502200603485, "learning_rate": 1.5563785175947182e-05, "loss": 0.0175, "step": 60380 }, { "grad_norm": 0.1411905437707901, "learning_rate": 1.5548803065856582e-05, "loss": 0.0144, "step": 60390 }, { "grad_norm": 0.16463959217071533, "learning_rate": 1.55338268425333e-05, "loss": 0.0149, "step": 60400 }, { "grad_norm": 0.18766465783119202, "learning_rate": 1.5518856508536373e-05, "loss": 0.017, "step": 60410 }, { "grad_norm": 0.2759256362915039, "learning_rate": 1.5503892066423786e-05, "loss": 0.0156, "step": 60420 }, { "grad_norm": 0.22895470261573792, "learning_rate": 1.548893351875258e-05, "loss": 0.0259, "step": 60430 }, { "grad_norm": 0.1895693838596344, "learning_rate": 1.547398086807875e-05, "loss": 0.0187, "step": 60440 }, { "grad_norm": 0.26706644892692566, "learning_rate": 1.5459034116957305e-05, "loss": 0.0239, "step": 60450 }, { "grad_norm": 0.151311993598938, "learning_rate": 1.544409326794225e-05, "loss": 0.0241, "step": 60460 }, { "grad_norm": 0.20975087583065033, "learning_rate": 1.542915832358651e-05, "loss": 0.0185, "step": 60470 }, { "grad_norm": 0.2374381422996521, "learning_rate": 1.541422928644214e-05, "loss": 0.0149, "step": 60480 }, { "grad_norm": 0.23311714828014374, "learning_rate": 1.5399306159060024e-05, "loss": 0.0225, "step": 60490 }, { "grad_norm": 0.2000369131565094, "learning_rate": 1.5384388943990185e-05, "loss": 0.0146, "step": 60500 }, { "grad_norm": 0.18688571453094482, "learning_rate": 1.5369477643781526e-05, "loss": 0.0138, "step": 60510 }, { "grad_norm": 0.15951207280158997, "learning_rate": 1.5354572260981985e-05, "loss": 0.0182, "step": 60520 }, { "grad_norm": 0.1608525663614273, "learning_rate": 1.533967279813851e-05, "loss": 0.013, "step": 60530 }, { "grad_norm": 0.1764872521162033, "learning_rate": 1.5324779257796956e-05, "loss": 0.018, "step": 60540 }, { "grad_norm": 0.20759601891040802, "learning_rate": 1.5309891642502293e-05, "loss": 0.0162, "step": 60550 }, { "grad_norm": 0.2644270658493042, "learning_rate": 1.5295009954798357e-05, "loss": 0.0139, "step": 60560 }, { "grad_norm": 0.21285971999168396, "learning_rate": 1.5280134197228036e-05, "loss": 0.0159, "step": 60570 }, { "grad_norm": 0.1573653370141983, "learning_rate": 1.526526437233319e-05, "loss": 0.0191, "step": 60580 }, { "grad_norm": 0.18799474835395813, "learning_rate": 1.525040048265466e-05, "loss": 0.0189, "step": 60590 }, { "grad_norm": 0.13264653086662292, "learning_rate": 1.5235542530732288e-05, "loss": 0.0151, "step": 60600 }, { "grad_norm": 0.2419906109571457, "learning_rate": 1.5220690519104901e-05, "loss": 0.017, "step": 60610 }, { "grad_norm": 0.17811864614486694, "learning_rate": 1.5205844450310275e-05, "loss": 0.0141, "step": 60620 }, { "grad_norm": 0.2872205376625061, "learning_rate": 1.519100432688521e-05, "loss": 0.018, "step": 60630 }, { "grad_norm": 0.22543446719646454, "learning_rate": 1.5176170151365476e-05, "loss": 0.02, "step": 60640 }, { "grad_norm": 0.29700812697410583, "learning_rate": 1.5161341926285833e-05, "loss": 0.0121, "step": 60650 }, { "grad_norm": 0.1498064249753952, "learning_rate": 1.5146519654180025e-05, "loss": 0.0116, "step": 60660 }, { "grad_norm": 0.2417827546596527, "learning_rate": 1.5131703337580739e-05, "loss": 0.019, "step": 60670 }, { "grad_norm": 0.1402806043624878, "learning_rate": 1.5116892979019731e-05, "loss": 0.02, "step": 60680 }, { "grad_norm": 0.2002529352903366, "learning_rate": 1.5102088581027623e-05, "loss": 0.0142, "step": 60690 }, { "grad_norm": 0.3033509850502014, "learning_rate": 1.5087290146134154e-05, "loss": 0.0151, "step": 60700 }, { "grad_norm": 0.19998428225517273, "learning_rate": 1.5072497676867915e-05, "loss": 0.0165, "step": 60710 }, { "grad_norm": 0.1454237550497055, "learning_rate": 1.505771117575655e-05, "loss": 0.0181, "step": 60720 }, { "grad_norm": 0.12648358941078186, "learning_rate": 1.5042930645326691e-05, "loss": 0.0138, "step": 60730 }, { "grad_norm": 0.2107861191034317, "learning_rate": 1.5028156088103879e-05, "loss": 0.0187, "step": 60740 }, { "grad_norm": 0.2587502896785736, "learning_rate": 1.5013387506612735e-05, "loss": 0.0227, "step": 60750 }, { "grad_norm": 0.1443253606557846, "learning_rate": 1.4998624903376767e-05, "loss": 0.0194, "step": 60760 }, { "grad_norm": 0.196036234498024, "learning_rate": 1.4983868280918517e-05, "loss": 0.0181, "step": 60770 }, { "grad_norm": 0.21871612966060638, "learning_rate": 1.4969117641759478e-05, "loss": 0.0208, "step": 60780 }, { "grad_norm": 0.19310890138149261, "learning_rate": 1.495437298842014e-05, "loss": 0.0167, "step": 60790 }, { "grad_norm": 0.1985475718975067, "learning_rate": 1.4939634323419976e-05, "loss": 0.0187, "step": 60800 }, { "grad_norm": 0.25549551844596863, "learning_rate": 1.4924901649277384e-05, "loss": 0.0169, "step": 60810 }, { "grad_norm": 0.1858428716659546, "learning_rate": 1.4910174968509793e-05, "loss": 0.0154, "step": 60820 }, { "grad_norm": 0.19315269589424133, "learning_rate": 1.4895454283633598e-05, "loss": 0.0158, "step": 60830 }, { "grad_norm": 0.30811807513237, "learning_rate": 1.488073959716415e-05, "loss": 0.0186, "step": 60840 }, { "grad_norm": 0.16098494827747345, "learning_rate": 1.4866030911615791e-05, "loss": 0.0195, "step": 60850 }, { "grad_norm": 0.12378516048192978, "learning_rate": 1.4851328229501849e-05, "loss": 0.0163, "step": 60860 }, { "grad_norm": 0.16906532645225525, "learning_rate": 1.4836631553334562e-05, "loss": 0.016, "step": 60870 }, { "grad_norm": 0.2474488765001297, "learning_rate": 1.4821940885625251e-05, "loss": 0.017, "step": 60880 }, { "grad_norm": 0.2686633765697479, "learning_rate": 1.4807256228884109e-05, "loss": 0.0201, "step": 60890 }, { "grad_norm": 0.1256820410490036, "learning_rate": 1.4792577585620353e-05, "loss": 0.0143, "step": 60900 }, { "grad_norm": 0.17629234492778778, "learning_rate": 1.4777904958342164e-05, "loss": 0.0212, "step": 60910 }, { "grad_norm": 0.18928565084934235, "learning_rate": 1.4763238349556691e-05, "loss": 0.0156, "step": 60920 }, { "grad_norm": 0.2503184974193573, "learning_rate": 1.4748577761770072e-05, "loss": 0.0224, "step": 60930 }, { "grad_norm": 0.15974503755569458, "learning_rate": 1.4733923197487354e-05, "loss": 0.0181, "step": 60940 }, { "grad_norm": 0.20674440264701843, "learning_rate": 1.4719274659212662e-05, "loss": 0.0181, "step": 60950 }, { "grad_norm": 0.29429057240486145, "learning_rate": 1.4704632149448983e-05, "loss": 0.0154, "step": 60960 }, { "grad_norm": 0.21313141286373138, "learning_rate": 1.4689995670698343e-05, "loss": 0.025, "step": 60970 }, { "grad_norm": 0.2502671778202057, "learning_rate": 1.4675365225461728e-05, "loss": 0.0181, "step": 60980 }, { "grad_norm": 0.19402702152729034, "learning_rate": 1.4660740816239032e-05, "loss": 0.0153, "step": 60990 }, { "grad_norm": 0.18497087061405182, "learning_rate": 1.464612244552923e-05, "loss": 0.0182, "step": 61000 }, { "grad_norm": 0.2163255512714386, "learning_rate": 1.4631510115830161e-05, "loss": 0.0143, "step": 61010 }, { "grad_norm": 0.24152733385562897, "learning_rate": 1.4616903829638679e-05, "loss": 0.0189, "step": 61020 }, { "grad_norm": 0.17094449698925018, "learning_rate": 1.46023035894506e-05, "loss": 0.017, "step": 61030 }, { "grad_norm": 0.18551433086395264, "learning_rate": 1.4587709397760713e-05, "loss": 0.0157, "step": 61040 }, { "grad_norm": 0.1371183544397354, "learning_rate": 1.4573121257062755e-05, "loss": 0.0169, "step": 61050 }, { "grad_norm": 0.2959197759628296, "learning_rate": 1.4558539169849472e-05, "loss": 0.0217, "step": 61060 }, { "grad_norm": 0.23811273276805878, "learning_rate": 1.4543963138612499e-05, "loss": 0.0155, "step": 61070 }, { "grad_norm": 0.24874669313430786, "learning_rate": 1.4529393165842498e-05, "loss": 0.0178, "step": 61080 }, { "grad_norm": 0.21536526083946228, "learning_rate": 1.4514829254029084e-05, "loss": 0.0154, "step": 61090 }, { "grad_norm": 0.19721685349941254, "learning_rate": 1.450027140566083e-05, "loss": 0.0163, "step": 61100 }, { "grad_norm": 0.21558021008968353, "learning_rate": 1.4485719623225268e-05, "loss": 0.0152, "step": 61110 }, { "grad_norm": 0.5245439410209656, "learning_rate": 1.4471173909208912e-05, "loss": 0.0162, "step": 61120 }, { "grad_norm": 0.18141654133796692, "learning_rate": 1.4456634266097236e-05, "loss": 0.0173, "step": 61130 }, { "grad_norm": 0.13733132183551788, "learning_rate": 1.444210069637461e-05, "loss": 0.0235, "step": 61140 }, { "grad_norm": 0.21659304201602936, "learning_rate": 1.4427573202524502e-05, "loss": 0.0148, "step": 61150 }, { "grad_norm": 0.20611251890659332, "learning_rate": 1.4413051787029208e-05, "loss": 0.0209, "step": 61160 }, { "grad_norm": 0.2064555436372757, "learning_rate": 1.439853645237006e-05, "loss": 0.0201, "step": 61170 }, { "grad_norm": 0.19309495389461517, "learning_rate": 1.4384027201027344e-05, "loss": 0.0144, "step": 61180 }, { "grad_norm": 0.24342043697834015, "learning_rate": 1.4369524035480253e-05, "loss": 0.0173, "step": 61190 }, { "grad_norm": 0.2108064740896225, "learning_rate": 1.4355026958207035e-05, "loss": 0.0155, "step": 61200 }, { "grad_norm": 0.24605080485343933, "learning_rate": 1.434053597168481e-05, "loss": 0.0169, "step": 61210 }, { "grad_norm": 0.23681628704071045, "learning_rate": 1.432605107838969e-05, "loss": 0.0167, "step": 61220 }, { "grad_norm": 0.26850590109825134, "learning_rate": 1.4311572280796759e-05, "loss": 0.0167, "step": 61230 }, { "grad_norm": 0.232632115483284, "learning_rate": 1.4297099581380047e-05, "loss": 0.0179, "step": 61240 }, { "grad_norm": 0.17070196568965912, "learning_rate": 1.4282632982612538e-05, "loss": 0.022, "step": 61250 }, { "grad_norm": 0.14953076839447021, "learning_rate": 1.4268172486966197e-05, "loss": 0.0234, "step": 61260 }, { "grad_norm": 0.1619069129228592, "learning_rate": 1.4253718096911894e-05, "loss": 0.0165, "step": 61270 }, { "grad_norm": 0.19454126060009003, "learning_rate": 1.4239269814919504e-05, "loss": 0.0203, "step": 61280 }, { "grad_norm": 0.16806109249591827, "learning_rate": 1.4224827643457843e-05, "loss": 0.0237, "step": 61290 }, { "grad_norm": 0.2623615860939026, "learning_rate": 1.4210391584994686e-05, "loss": 0.0144, "step": 61300 }, { "grad_norm": 0.20010262727737427, "learning_rate": 1.4195961641996763e-05, "loss": 0.0188, "step": 61310 }, { "grad_norm": 0.2511085867881775, "learning_rate": 1.4181537816929751e-05, "loss": 0.0156, "step": 61320 }, { "grad_norm": 0.2290218323469162, "learning_rate": 1.4167120112258303e-05, "loss": 0.0203, "step": 61330 }, { "grad_norm": 0.16788579523563385, "learning_rate": 1.4152708530445963e-05, "loss": 0.0135, "step": 61340 }, { "grad_norm": 0.12261530756950378, "learning_rate": 1.4138303073955344e-05, "loss": 0.0151, "step": 61350 }, { "grad_norm": 0.18636906147003174, "learning_rate": 1.4123903745247897e-05, "loss": 0.019, "step": 61360 }, { "grad_norm": 0.13964581489562988, "learning_rate": 1.4109510546784082e-05, "loss": 0.0139, "step": 61370 }, { "grad_norm": 0.14644606411457062, "learning_rate": 1.4095123481023325e-05, "loss": 0.0197, "step": 61380 }, { "grad_norm": 0.1959879994392395, "learning_rate": 1.4080742550423936e-05, "loss": 0.0164, "step": 61390 }, { "grad_norm": 0.15191882848739624, "learning_rate": 1.4066367757443278e-05, "loss": 0.0155, "step": 61400 }, { "grad_norm": 0.18477192521095276, "learning_rate": 1.4051999104537572e-05, "loss": 0.0156, "step": 61410 }, { "grad_norm": 0.18195396661758423, "learning_rate": 1.4037636594162034e-05, "loss": 0.0157, "step": 61420 }, { "grad_norm": 0.16702193021774292, "learning_rate": 1.402328022877083e-05, "loss": 0.0135, "step": 61430 }, { "grad_norm": 0.23671787977218628, "learning_rate": 1.4008930010817073e-05, "loss": 0.0168, "step": 61440 }, { "grad_norm": 0.2834005057811737, "learning_rate": 1.3994585942752831e-05, "loss": 0.0145, "step": 61450 }, { "grad_norm": 0.25897422432899475, "learning_rate": 1.398024802702909e-05, "loss": 0.0135, "step": 61460 }, { "grad_norm": 0.1845325231552124, "learning_rate": 1.3965916266095819e-05, "loss": 0.0145, "step": 61470 }, { "grad_norm": 0.21437901258468628, "learning_rate": 1.395159066240192e-05, "loss": 0.0161, "step": 61480 }, { "grad_norm": 0.1777670532464981, "learning_rate": 1.3937271218395248e-05, "loss": 0.0151, "step": 61490 }, { "grad_norm": 0.255302369594574, "learning_rate": 1.3922957936522613e-05, "loss": 0.0172, "step": 61500 }, { "grad_norm": 0.17505842447280884, "learning_rate": 1.3908650819229758e-05, "loss": 0.0202, "step": 61510 }, { "grad_norm": 0.24177512526512146, "learning_rate": 1.389434986896137e-05, "loss": 0.0151, "step": 61520 }, { "grad_norm": 0.24014925956726074, "learning_rate": 1.388005508816112e-05, "loss": 0.0147, "step": 61530 }, { "grad_norm": 0.16551145911216736, "learning_rate": 1.3865766479271557e-05, "loss": 0.0162, "step": 61540 }, { "grad_norm": 0.1844620555639267, "learning_rate": 1.3851484044734225e-05, "loss": 0.0224, "step": 61550 }, { "grad_norm": 0.22063252329826355, "learning_rate": 1.3837207786989608e-05, "loss": 0.0228, "step": 61560 }, { "grad_norm": 0.12947174906730652, "learning_rate": 1.382293770847713e-05, "loss": 0.0152, "step": 61570 }, { "grad_norm": 0.2054828554391861, "learning_rate": 1.3808673811635159e-05, "loss": 0.0164, "step": 61580 }, { "grad_norm": 0.44682520627975464, "learning_rate": 1.3794416098900975e-05, "loss": 0.0179, "step": 61590 }, { "grad_norm": 0.17603491246700287, "learning_rate": 1.3780164572710886e-05, "loss": 0.0198, "step": 61600 }, { "grad_norm": 0.1410815417766571, "learning_rate": 1.3765919235500035e-05, "loss": 0.0182, "step": 61610 }, { "grad_norm": 0.2082669585943222, "learning_rate": 1.375168008970259e-05, "loss": 0.017, "step": 61620 }, { "grad_norm": 0.3473336100578308, "learning_rate": 1.3737447137751635e-05, "loss": 0.0141, "step": 61630 }, { "grad_norm": 0.2055324912071228, "learning_rate": 1.372322038207915e-05, "loss": 0.0127, "step": 61640 }, { "grad_norm": 0.2323116958141327, "learning_rate": 1.3708999825116159e-05, "loss": 0.0233, "step": 61650 }, { "grad_norm": 0.20508815348148346, "learning_rate": 1.3694785469292526e-05, "loss": 0.0201, "step": 61660 }, { "grad_norm": 0.1585892289876938, "learning_rate": 1.3680577317037101e-05, "loss": 0.0156, "step": 61670 }, { "grad_norm": 0.2181752622127533, "learning_rate": 1.3666375370777678e-05, "loss": 0.0154, "step": 61680 }, { "grad_norm": 0.19057364761829376, "learning_rate": 1.3652179632940981e-05, "loss": 0.0164, "step": 61690 }, { "grad_norm": 0.3105722665786743, "learning_rate": 1.3637990105952664e-05, "loss": 0.0175, "step": 61700 }, { "grad_norm": 0.2136056274175644, "learning_rate": 1.3623806792237337e-05, "loss": 0.0163, "step": 61710 }, { "grad_norm": 0.4420730471611023, "learning_rate": 1.3609629694218551e-05, "loss": 0.0175, "step": 61720 }, { "grad_norm": 0.3618493378162384, "learning_rate": 1.3595458814318762e-05, "loss": 0.0152, "step": 61730 }, { "grad_norm": 0.24398395419120789, "learning_rate": 1.3581294154959389e-05, "loss": 0.0198, "step": 61740 }, { "grad_norm": 0.2130240648984909, "learning_rate": 1.3567135718560792e-05, "loss": 0.0151, "step": 61750 }, { "grad_norm": 0.1806739866733551, "learning_rate": 1.3552983507542261e-05, "loss": 0.015, "step": 61760 }, { "grad_norm": 0.16751836240291595, "learning_rate": 1.3538837524322023e-05, "loss": 0.0185, "step": 61770 }, { "grad_norm": 0.16725994646549225, "learning_rate": 1.3524697771317251e-05, "loss": 0.023, "step": 61780 }, { "grad_norm": 0.1859607845544815, "learning_rate": 1.3510564250943997e-05, "loss": 0.0136, "step": 61790 }, { "grad_norm": 0.22930432856082916, "learning_rate": 1.3496436965617353e-05, "loss": 0.014, "step": 61800 }, { "grad_norm": 0.1995401382446289, "learning_rate": 1.3482315917751243e-05, "loss": 0.0178, "step": 61810 }, { "grad_norm": 0.13631413877010345, "learning_rate": 1.3468201109758582e-05, "loss": 0.011, "step": 61820 }, { "grad_norm": 0.1981038749217987, "learning_rate": 1.3454092544051222e-05, "loss": 0.0162, "step": 61830 }, { "grad_norm": 0.19831706583499908, "learning_rate": 1.3439990223039878e-05, "loss": 0.0158, "step": 61840 }, { "grad_norm": 0.17555077373981476, "learning_rate": 1.3425894149134321e-05, "loss": 0.0143, "step": 61850 }, { "grad_norm": 0.22875182330608368, "learning_rate": 1.3411804324743132e-05, "loss": 0.0204, "step": 61860 }, { "grad_norm": 0.2615443170070648, "learning_rate": 1.3397720752273896e-05, "loss": 0.0191, "step": 61870 }, { "grad_norm": 0.1969469040632248, "learning_rate": 1.3383643434133108e-05, "loss": 0.014, "step": 61880 }, { "grad_norm": 0.22425441443920135, "learning_rate": 1.3369572372726196e-05, "loss": 0.0183, "step": 61890 }, { "grad_norm": 0.18902161717414856, "learning_rate": 1.3355507570457525e-05, "loss": 0.0159, "step": 61900 }, { "grad_norm": 0.4484115242958069, "learning_rate": 1.334144902973038e-05, "loss": 0.0138, "step": 61910 }, { "grad_norm": 0.3346421718597412, "learning_rate": 1.3327396752946997e-05, "loss": 0.0158, "step": 61920 }, { "grad_norm": 0.1268218606710434, "learning_rate": 1.3313350742508496e-05, "loss": 0.0161, "step": 61930 }, { "grad_norm": 0.1507650762796402, "learning_rate": 1.3299311000814973e-05, "loss": 0.0169, "step": 61940 }, { "grad_norm": 0.13100741803646088, "learning_rate": 1.3285277530265432e-05, "loss": 0.0199, "step": 61950 }, { "grad_norm": 0.24124738574028015, "learning_rate": 1.3271250333257813e-05, "loss": 0.0162, "step": 61960 }, { "grad_norm": 0.21945267915725708, "learning_rate": 1.3257229412188977e-05, "loss": 0.0137, "step": 61970 }, { "grad_norm": 0.2503073811531067, "learning_rate": 1.3243214769454726e-05, "loss": 0.0145, "step": 61980 }, { "grad_norm": 0.15878671407699585, "learning_rate": 1.3229206407449751e-05, "loss": 0.0137, "step": 61990 }, { "grad_norm": 0.15268751978874207, "learning_rate": 1.3215204328567742e-05, "loss": 0.0175, "step": 62000 }, { "grad_norm": 0.1725747436285019, "learning_rate": 1.3201208535201232e-05, "loss": 0.0139, "step": 62010 }, { "grad_norm": 0.13241852819919586, "learning_rate": 1.3187219029741737e-05, "loss": 0.015, "step": 62020 }, { "grad_norm": 0.24881373345851898, "learning_rate": 1.3173235814579693e-05, "loss": 0.0131, "step": 62030 }, { "grad_norm": 0.21657998859882355, "learning_rate": 1.3159258892104398e-05, "loss": 0.0161, "step": 62040 }, { "grad_norm": 0.39272087812423706, "learning_rate": 1.3145288264704198e-05, "loss": 0.0184, "step": 62050 }, { "grad_norm": 0.30233079195022583, "learning_rate": 1.3131323934766237e-05, "loss": 0.0147, "step": 62060 }, { "grad_norm": 0.14344963431358337, "learning_rate": 1.311736590467666e-05, "loss": 0.0152, "step": 62070 }, { "grad_norm": 0.2078486829996109, "learning_rate": 1.3103414176820505e-05, "loss": 0.0166, "step": 62080 }, { "grad_norm": 0.1432906836271286, "learning_rate": 1.308946875358174e-05, "loss": 0.0229, "step": 62090 }, { "grad_norm": 0.17367714643478394, "learning_rate": 1.3075529637343276e-05, "loss": 0.0248, "step": 62100 }, { "grad_norm": 0.17995095252990723, "learning_rate": 1.3061596830486883e-05, "loss": 0.0167, "step": 62110 }, { "grad_norm": 0.2064812332391739, "learning_rate": 1.3047670335393353e-05, "loss": 0.0143, "step": 62120 }, { "grad_norm": 0.19355496764183044, "learning_rate": 1.3033750154442298e-05, "loss": 0.0182, "step": 62130 }, { "grad_norm": 0.18956358730793, "learning_rate": 1.3019836290012316e-05, "loss": 0.015, "step": 62140 }, { "grad_norm": 0.40149804949760437, "learning_rate": 1.30059287444809e-05, "loss": 0.0175, "step": 62150 }, { "grad_norm": 0.22959603369235992, "learning_rate": 1.299202752022447e-05, "loss": 0.0243, "step": 62160 }, { "grad_norm": 0.17687253654003143, "learning_rate": 1.2978132619618371e-05, "loss": 0.0138, "step": 62170 }, { "grad_norm": 0.18357837200164795, "learning_rate": 1.2964244045036866e-05, "loss": 0.017, "step": 62180 }, { "grad_norm": 0.17272858321666718, "learning_rate": 1.295036179885311e-05, "loss": 0.0144, "step": 62190 }, { "grad_norm": 0.15389162302017212, "learning_rate": 1.293648588343922e-05, "loss": 0.02, "step": 62200 }, { "grad_norm": 0.23445644974708557, "learning_rate": 1.2922616301166196e-05, "loss": 0.0152, "step": 62210 }, { "grad_norm": 0.22188985347747803, "learning_rate": 1.2908753054403976e-05, "loss": 0.0144, "step": 62220 }, { "grad_norm": 0.191792294383049, "learning_rate": 1.2894896145521429e-05, "loss": 0.0124, "step": 62230 }, { "grad_norm": 0.18021710216999054, "learning_rate": 1.2881045576886275e-05, "loss": 0.0189, "step": 62240 }, { "grad_norm": 0.26056158542633057, "learning_rate": 1.2867201350865254e-05, "loss": 0.0157, "step": 62250 }, { "grad_norm": 0.2043089121580124, "learning_rate": 1.2853363469823914e-05, "loss": 0.0165, "step": 62260 }, { "grad_norm": 0.20116260647773743, "learning_rate": 1.283953193612682e-05, "loss": 0.0129, "step": 62270 }, { "grad_norm": 0.20985887944698334, "learning_rate": 1.2825706752137372e-05, "loss": 0.0161, "step": 62280 }, { "grad_norm": 0.1741168051958084, "learning_rate": 1.2811887920217896e-05, "loss": 0.0157, "step": 62290 }, { "grad_norm": 0.21789196133613586, "learning_rate": 1.2798075442729707e-05, "loss": 0.0137, "step": 62300 }, { "grad_norm": 0.20152994990348816, "learning_rate": 1.2784269322032922e-05, "loss": 0.0142, "step": 62310 }, { "grad_norm": 0.14224231243133545, "learning_rate": 1.2770469560486686e-05, "loss": 0.0167, "step": 62320 }, { "grad_norm": 0.18706628680229187, "learning_rate": 1.2756676160448956e-05, "loss": 0.0167, "step": 62330 }, { "grad_norm": 0.2538165748119354, "learning_rate": 1.2742889124276663e-05, "loss": 0.0161, "step": 62340 }, { "grad_norm": 0.19025148451328278, "learning_rate": 1.2729108454325639e-05, "loss": 0.0152, "step": 62350 }, { "grad_norm": 0.1781177818775177, "learning_rate": 1.2715334152950614e-05, "loss": 0.0176, "step": 62360 }, { "grad_norm": 0.12197450548410416, "learning_rate": 1.2701566222505246e-05, "loss": 0.0146, "step": 62370 }, { "grad_norm": 0.2771981954574585, "learning_rate": 1.2687804665342107e-05, "loss": 0.0153, "step": 62380 }, { "grad_norm": 0.230501189827919, "learning_rate": 1.267404948381265e-05, "loss": 0.0173, "step": 62390 }, { "grad_norm": 0.29146528244018555, "learning_rate": 1.2660300680267267e-05, "loss": 0.012, "step": 62400 }, { "grad_norm": 0.23849450051784515, "learning_rate": 1.2646558257055257e-05, "loss": 0.014, "step": 62410 }, { "grad_norm": 0.2689754366874695, "learning_rate": 1.263282221652482e-05, "loss": 0.0176, "step": 62420 }, { "grad_norm": 0.09997429698705673, "learning_rate": 1.2619092561023088e-05, "loss": 0.0154, "step": 62430 }, { "grad_norm": 0.1402023881673813, "learning_rate": 1.2605369292896036e-05, "loss": 0.0175, "step": 62440 }, { "grad_norm": 0.20159074664115906, "learning_rate": 1.2591652414488658e-05, "loss": 0.0164, "step": 62450 }, { "grad_norm": 0.17199018597602844, "learning_rate": 1.2577941928144732e-05, "loss": 0.0167, "step": 62460 }, { "grad_norm": 0.1694256216287613, "learning_rate": 1.256423783620706e-05, "loss": 0.0175, "step": 62470 }, { "grad_norm": 0.15801681578159332, "learning_rate": 1.2550540141017264e-05, "loss": 0.014, "step": 62480 }, { "grad_norm": 0.17531001567840576, "learning_rate": 1.253684884491591e-05, "loss": 0.0112, "step": 62490 }, { "grad_norm": 0.2576708197593689, "learning_rate": 1.2523163950242483e-05, "loss": 0.0116, "step": 62500 }, { "grad_norm": 0.15144819021224976, "learning_rate": 1.2509485459335313e-05, "loss": 0.0158, "step": 62510 }, { "grad_norm": 0.1641121208667755, "learning_rate": 1.2495813374531739e-05, "loss": 0.0144, "step": 62520 }, { "grad_norm": 0.13356739282608032, "learning_rate": 1.2482147698167907e-05, "loss": 0.0123, "step": 62530 }, { "grad_norm": 0.24675752222537994, "learning_rate": 1.2468488432578911e-05, "loss": 0.0171, "step": 62540 }, { "grad_norm": 0.16601720452308655, "learning_rate": 1.2454835580098761e-05, "loss": 0.0153, "step": 62550 }, { "grad_norm": 0.17468056082725525, "learning_rate": 1.2441189143060338e-05, "loss": 0.0149, "step": 62560 }, { "grad_norm": 0.15034784376621246, "learning_rate": 1.242754912379548e-05, "loss": 0.0151, "step": 62570 }, { "grad_norm": 0.23112255334854126, "learning_rate": 1.2413915524634844e-05, "loss": 0.0301, "step": 62580 }, { "grad_norm": 0.1758115589618683, "learning_rate": 1.2400288347908073e-05, "loss": 0.0162, "step": 62590 }, { "grad_norm": 0.2019752860069275, "learning_rate": 1.2386667595943663e-05, "loss": 0.0132, "step": 62600 }, { "grad_norm": 0.163578599691391, "learning_rate": 1.2373053271069035e-05, "loss": 0.0126, "step": 62610 }, { "grad_norm": 0.1694779396057129, "learning_rate": 1.2359445375610501e-05, "loss": 0.0169, "step": 62620 }, { "grad_norm": 0.147226944565773, "learning_rate": 1.2345843911893301e-05, "loss": 0.0146, "step": 62630 }, { "grad_norm": 0.19665563106536865, "learning_rate": 1.2332248882241498e-05, "loss": 0.0222, "step": 62640 }, { "grad_norm": 0.2618882358074188, "learning_rate": 1.2318660288978178e-05, "loss": 0.0151, "step": 62650 }, { "grad_norm": 0.3595556914806366, "learning_rate": 1.2305078134425213e-05, "loss": 0.0164, "step": 62660 }, { "grad_norm": 0.19879752397537231, "learning_rate": 1.2291502420903434e-05, "loss": 0.0148, "step": 62670 }, { "grad_norm": 0.12427208572626114, "learning_rate": 1.2277933150732567e-05, "loss": 0.0142, "step": 62680 }, { "grad_norm": 0.14567632973194122, "learning_rate": 1.2264370326231216e-05, "loss": 0.0139, "step": 62690 }, { "grad_norm": 0.2872503399848938, "learning_rate": 1.2250813949716927e-05, "loss": 0.0152, "step": 62700 }, { "grad_norm": 0.14738084375858307, "learning_rate": 1.2237264023506063e-05, "loss": 0.0174, "step": 62710 }, { "grad_norm": 0.21606460213661194, "learning_rate": 1.2223720549913987e-05, "loss": 0.0149, "step": 62720 }, { "grad_norm": 0.2721754312515259, "learning_rate": 1.221018353125487e-05, "loss": 0.0134, "step": 62730 }, { "grad_norm": 0.12849734723567963, "learning_rate": 1.2196652969841837e-05, "loss": 0.0161, "step": 62740 }, { "grad_norm": 0.20193365216255188, "learning_rate": 1.2183128867986904e-05, "loss": 0.0191, "step": 62750 }, { "grad_norm": 0.18480153381824493, "learning_rate": 1.2169611228000927e-05, "loss": 0.0168, "step": 62760 }, { "grad_norm": 0.17642416059970856, "learning_rate": 1.2156100052193752e-05, "loss": 0.016, "step": 62770 }, { "grad_norm": 0.1796257346868515, "learning_rate": 1.214259534287403e-05, "loss": 0.014, "step": 62780 }, { "grad_norm": 0.16286328434944153, "learning_rate": 1.2129097102349363e-05, "loss": 0.0125, "step": 62790 }, { "grad_norm": 0.11143908649682999, "learning_rate": 1.2115605332926227e-05, "loss": 0.0154, "step": 62800 }, { "grad_norm": 0.16823254525661469, "learning_rate": 1.2102120036910003e-05, "loss": 0.0174, "step": 62810 }, { "grad_norm": 0.18472084403038025, "learning_rate": 1.2088641216604956e-05, "loss": 0.0169, "step": 62820 }, { "grad_norm": 0.11242426186800003, "learning_rate": 1.2075168874314264e-05, "loss": 0.0135, "step": 62830 }, { "grad_norm": 0.14112508296966553, "learning_rate": 1.2061703012339942e-05, "loss": 0.0126, "step": 62840 }, { "grad_norm": 0.2774806022644043, "learning_rate": 1.204824363298297e-05, "loss": 0.0167, "step": 62850 }, { "grad_norm": 0.18572048842906952, "learning_rate": 1.2034790738543173e-05, "loss": 0.0171, "step": 62860 }, { "grad_norm": 0.10968659818172455, "learning_rate": 1.2021344331319284e-05, "loss": 0.0153, "step": 62870 }, { "grad_norm": 0.17434242367744446, "learning_rate": 1.200790441360894e-05, "loss": 0.0176, "step": 62880 }, { "grad_norm": 0.16175420582294464, "learning_rate": 1.199447098770864e-05, "loss": 0.0143, "step": 62890 }, { "grad_norm": 0.15818332135677338, "learning_rate": 1.198104405591381e-05, "loss": 0.0181, "step": 62900 }, { "grad_norm": 0.25378715991973877, "learning_rate": 1.1967623620518697e-05, "loss": 0.0132, "step": 62910 }, { "grad_norm": 0.2383406013250351, "learning_rate": 1.1954209683816554e-05, "loss": 0.0159, "step": 62920 }, { "grad_norm": 0.19727906584739685, "learning_rate": 1.1940802248099402e-05, "loss": 0.0161, "step": 62930 }, { "grad_norm": 0.13640643656253815, "learning_rate": 1.1927401315658232e-05, "loss": 0.0143, "step": 62940 }, { "grad_norm": 0.1831909865140915, "learning_rate": 1.1914006888782898e-05, "loss": 0.0139, "step": 62950 }, { "grad_norm": 0.1568412482738495, "learning_rate": 1.1900618969762106e-05, "loss": 0.0178, "step": 62960 }, { "grad_norm": 0.22320427000522614, "learning_rate": 1.188723756088354e-05, "loss": 0.0163, "step": 62970 }, { "grad_norm": 0.18469665944576263, "learning_rate": 1.1873862664433672e-05, "loss": 0.0154, "step": 62980 }, { "grad_norm": 0.17865031957626343, "learning_rate": 1.1860494282697927e-05, "loss": 0.0159, "step": 62990 }, { "grad_norm": 0.20763617753982544, "learning_rate": 1.1847132417960588e-05, "loss": 0.013, "step": 63000 }, { "grad_norm": 0.13063989579677582, "learning_rate": 1.1833777072504832e-05, "loss": 0.0165, "step": 63010 }, { "grad_norm": 0.1664033681154251, "learning_rate": 1.1820428248612731e-05, "loss": 0.0166, "step": 63020 }, { "grad_norm": 0.26585477590560913, "learning_rate": 1.1807085948565245e-05, "loss": 0.0191, "step": 63030 }, { "grad_norm": 0.12704600393772125, "learning_rate": 1.1793750174642172e-05, "loss": 0.0153, "step": 63040 }, { "grad_norm": 0.12342232465744019, "learning_rate": 1.1780420929122254e-05, "loss": 0.017, "step": 63050 }, { "grad_norm": 0.24032816290855408, "learning_rate": 1.176709821428309e-05, "loss": 0.0133, "step": 63060 }, { "grad_norm": 0.26250845193862915, "learning_rate": 1.1753782032401173e-05, "loss": 0.0205, "step": 63070 }, { "grad_norm": 0.19155195355415344, "learning_rate": 1.1740472385751866e-05, "loss": 0.0164, "step": 63080 }, { "grad_norm": 0.15747365355491638, "learning_rate": 1.172716927660943e-05, "loss": 0.0158, "step": 63090 }, { "grad_norm": 0.2299298793077469, "learning_rate": 1.1713872707247015e-05, "loss": 0.0184, "step": 63100 }, { "grad_norm": 0.19150075316429138, "learning_rate": 1.1700582679936595e-05, "loss": 0.0141, "step": 63110 }, { "grad_norm": 0.21155676245689392, "learning_rate": 1.1687299196949136e-05, "loss": 0.0216, "step": 63120 }, { "grad_norm": 0.19511517882347107, "learning_rate": 1.1674022260554374e-05, "loss": 0.0143, "step": 63130 }, { "grad_norm": 0.17627838253974915, "learning_rate": 1.1660751873020987e-05, "loss": 0.0133, "step": 63140 }, { "grad_norm": 0.2534862160682678, "learning_rate": 1.1647488036616538e-05, "loss": 0.017, "step": 63150 }, { "grad_norm": 0.16620619595050812, "learning_rate": 1.1634230753607417e-05, "loss": 0.0183, "step": 63160 }, { "grad_norm": 0.27422231435775757, "learning_rate": 1.1620980026258982e-05, "loss": 0.0157, "step": 63170 }, { "grad_norm": 0.29242148995399475, "learning_rate": 1.1607735856835373e-05, "loss": 0.0154, "step": 63180 }, { "grad_norm": 0.16041839122772217, "learning_rate": 1.1594498247599677e-05, "loss": 0.0149, "step": 63190 }, { "grad_norm": 0.18621079623699188, "learning_rate": 1.158126720081384e-05, "loss": 0.0244, "step": 63200 }, { "grad_norm": 0.1421411782503128, "learning_rate": 1.156804271873868e-05, "loss": 0.0143, "step": 63210 }, { "grad_norm": 0.1680021435022354, "learning_rate": 1.1554824803633924e-05, "loss": 0.0181, "step": 63220 }, { "grad_norm": 0.2711457312107086, "learning_rate": 1.154161345775811e-05, "loss": 0.015, "step": 63230 }, { "grad_norm": 0.20860391855239868, "learning_rate": 1.1528408683368724e-05, "loss": 0.0191, "step": 63240 }, { "grad_norm": 0.22435027360916138, "learning_rate": 1.1515210482722088e-05, "loss": 0.0177, "step": 63250 }, { "grad_norm": 0.24761714041233063, "learning_rate": 1.150201885807342e-05, "loss": 0.0215, "step": 63260 }, { "grad_norm": 0.16549235582351685, "learning_rate": 1.1488833811676807e-05, "loss": 0.0156, "step": 63270 }, { "grad_norm": 0.23533213138580322, "learning_rate": 1.1475655345785213e-05, "loss": 0.0151, "step": 63280 }, { "grad_norm": 0.22802093625068665, "learning_rate": 1.1462483462650481e-05, "loss": 0.0154, "step": 63290 }, { "grad_norm": 0.29346850514411926, "learning_rate": 1.1449318164523331e-05, "loss": 0.0189, "step": 63300 }, { "grad_norm": 0.21004967391490936, "learning_rate": 1.1436159453653334e-05, "loss": 0.0127, "step": 63310 }, { "grad_norm": 0.1486353576183319, "learning_rate": 1.1423007332288955e-05, "loss": 0.0217, "step": 63320 }, { "grad_norm": 0.2287837415933609, "learning_rate": 1.1409861802677546e-05, "loss": 0.0182, "step": 63330 }, { "grad_norm": 0.25664275884628296, "learning_rate": 1.1396722867065313e-05, "loss": 0.016, "step": 63340 }, { "grad_norm": 0.21643923223018646, "learning_rate": 1.1383590527697352e-05, "loss": 0.0188, "step": 63350 }, { "grad_norm": 0.1481342613697052, "learning_rate": 1.137046478681758e-05, "loss": 0.0179, "step": 63360 }, { "grad_norm": 0.1843622475862503, "learning_rate": 1.1357345646668888e-05, "loss": 0.0308, "step": 63370 }, { "grad_norm": 0.18184448778629303, "learning_rate": 1.1344233109492924e-05, "loss": 0.017, "step": 63380 }, { "grad_norm": 0.18680155277252197, "learning_rate": 1.1331127177530292e-05, "loss": 0.0143, "step": 63390 }, { "grad_norm": 0.20478737354278564, "learning_rate": 1.1318027853020441e-05, "loss": 0.0136, "step": 63400 }, { "grad_norm": 0.19693957269191742, "learning_rate": 1.1304935138201645e-05, "loss": 0.014, "step": 63410 }, { "grad_norm": 0.18255005776882172, "learning_rate": 1.1291849035311153e-05, "loss": 0.0155, "step": 63420 }, { "grad_norm": 0.14544756710529327, "learning_rate": 1.1278769546584972e-05, "loss": 0.0165, "step": 63430 }, { "grad_norm": 0.2634446918964386, "learning_rate": 1.1265696674258052e-05, "loss": 0.0157, "step": 63440 }, { "grad_norm": 0.1888623982667923, "learning_rate": 1.1252630420564186e-05, "loss": 0.0152, "step": 63450 }, { "grad_norm": 0.22596324980258942, "learning_rate": 1.1239570787736036e-05, "loss": 0.0178, "step": 63460 }, { "grad_norm": 0.16666962206363678, "learning_rate": 1.1226517778005135e-05, "loss": 0.016, "step": 63470 }, { "grad_norm": 0.14746937155723572, "learning_rate": 1.1213471393601893e-05, "loss": 0.0167, "step": 63480 }, { "grad_norm": 0.17327232658863068, "learning_rate": 1.1200431636755587e-05, "loss": 0.0211, "step": 63490 }, { "grad_norm": 0.26784974336624146, "learning_rate": 1.1187398509694336e-05, "loss": 0.0236, "step": 63500 }, { "grad_norm": 0.16822826862335205, "learning_rate": 1.1174372014645146e-05, "loss": 0.0132, "step": 63510 }, { "grad_norm": 0.22611401975154877, "learning_rate": 1.1161352153833899e-05, "loss": 0.0223, "step": 63520 }, { "grad_norm": 0.39613333344459534, "learning_rate": 1.1148338929485325e-05, "loss": 0.0187, "step": 63530 }, { "grad_norm": 0.1568169742822647, "learning_rate": 1.113533234382304e-05, "loss": 0.012, "step": 63540 }, { "grad_norm": 0.13343384861946106, "learning_rate": 1.1122332399069513e-05, "loss": 0.0158, "step": 63550 }, { "grad_norm": 0.2064649909734726, "learning_rate": 1.1109339097446047e-05, "loss": 0.0174, "step": 63560 }, { "grad_norm": 0.21716946363449097, "learning_rate": 1.1096352441172897e-05, "loss": 0.0224, "step": 63570 }, { "grad_norm": 0.26865315437316895, "learning_rate": 1.1083372432469086e-05, "loss": 0.0154, "step": 63580 }, { "grad_norm": 0.2051560878753662, "learning_rate": 1.107039907355255e-05, "loss": 0.021, "step": 63590 }, { "grad_norm": 0.23031459748744965, "learning_rate": 1.1057432366640103e-05, "loss": 0.016, "step": 63600 }, { "grad_norm": 0.2655002772808075, "learning_rate": 1.1044472313947352e-05, "loss": 0.0194, "step": 63610 }, { "grad_norm": 0.1057710126042366, "learning_rate": 1.1031518917688877e-05, "loss": 0.0177, "step": 63620 }, { "grad_norm": 0.1975502073764801, "learning_rate": 1.1018572180078007e-05, "loss": 0.0224, "step": 63630 }, { "grad_norm": 0.2999148964881897, "learning_rate": 1.1005632103327018e-05, "loss": 0.0146, "step": 63640 }, { "grad_norm": 0.19876471161842346, "learning_rate": 1.0992698689646996e-05, "loss": 0.0181, "step": 63650 }, { "grad_norm": 0.19221659004688263, "learning_rate": 1.0979771941247919e-05, "loss": 0.0217, "step": 63660 }, { "grad_norm": 0.16339312493801117, "learning_rate": 1.0966851860338611e-05, "loss": 0.0176, "step": 63670 }, { "grad_norm": 0.18170100450515747, "learning_rate": 1.0953938449126766e-05, "loss": 0.0161, "step": 63680 }, { "grad_norm": 0.3491118848323822, "learning_rate": 1.0941031709818933e-05, "loss": 0.0268, "step": 63690 }, { "grad_norm": 0.22600246965885162, "learning_rate": 1.0928131644620509e-05, "loss": 0.0227, "step": 63700 }, { "grad_norm": 0.1925051063299179, "learning_rate": 1.0915238255735766e-05, "loss": 0.0188, "step": 63710 }, { "grad_norm": 0.16191010177135468, "learning_rate": 1.0902351545367833e-05, "loss": 0.0152, "step": 63720 }, { "grad_norm": 0.2364455908536911, "learning_rate": 1.0889471515718702e-05, "loss": 0.0164, "step": 63730 }, { "grad_norm": 0.15576328337192535, "learning_rate": 1.087659816898921e-05, "loss": 0.0129, "step": 63740 }, { "grad_norm": 0.2720010578632355, "learning_rate": 1.0863731507379082e-05, "loss": 0.0159, "step": 63750 }, { "grad_norm": 0.27218660712242126, "learning_rate": 1.0850871533086827e-05, "loss": 0.018, "step": 63760 }, { "grad_norm": 0.191084086894989, "learning_rate": 1.0838018248309927e-05, "loss": 0.0151, "step": 63770 }, { "grad_norm": 0.22740082442760468, "learning_rate": 1.0825171655244615e-05, "loss": 0.0206, "step": 63780 }, { "grad_norm": 0.22912466526031494, "learning_rate": 1.0812331756086025e-05, "loss": 0.0143, "step": 63790 }, { "grad_norm": 0.21518371999263763, "learning_rate": 1.079949855302817e-05, "loss": 0.0177, "step": 63800 }, { "grad_norm": 0.2778126001358032, "learning_rate": 1.0786672048263852e-05, "loss": 0.0191, "step": 63810 }, { "grad_norm": 0.19157689809799194, "learning_rate": 1.0773852243984817e-05, "loss": 0.0181, "step": 63820 }, { "grad_norm": 0.21058180928230286, "learning_rate": 1.0761039142381586e-05, "loss": 0.024, "step": 63830 }, { "grad_norm": 0.21980543434619904, "learning_rate": 1.0748232745643577e-05, "loss": 0.0149, "step": 63840 }, { "grad_norm": 0.1786145269870758, "learning_rate": 1.0735433055959055e-05, "loss": 0.0225, "step": 63850 }, { "grad_norm": 0.23811514675617218, "learning_rate": 1.0722640075515133e-05, "loss": 0.0178, "step": 63860 }, { "grad_norm": 0.1336848884820938, "learning_rate": 1.0709853806497795e-05, "loss": 0.0146, "step": 63870 }, { "grad_norm": 0.17536602914333344, "learning_rate": 1.0697074251091831e-05, "loss": 0.0152, "step": 63880 }, { "grad_norm": 0.40368157625198364, "learning_rate": 1.0684301411480962e-05, "loss": 0.0172, "step": 63890 }, { "grad_norm": 0.20804230868816376, "learning_rate": 1.067153528984769e-05, "loss": 0.0127, "step": 63900 }, { "grad_norm": 0.17823952436447144, "learning_rate": 1.0658775888373395e-05, "loss": 0.0159, "step": 63910 }, { "grad_norm": 0.21749676764011383, "learning_rate": 1.0646023209238314e-05, "loss": 0.022, "step": 63920 }, { "grad_norm": 0.12420140206813812, "learning_rate": 1.0633277254621537e-05, "loss": 0.0149, "step": 63930 }, { "grad_norm": 0.09985232353210449, "learning_rate": 1.0620538026700994e-05, "loss": 0.0133, "step": 63940 }, { "grad_norm": 0.2257099449634552, "learning_rate": 1.0607805527653486e-05, "loss": 0.0178, "step": 63950 }, { "grad_norm": 0.14813202619552612, "learning_rate": 1.059507975965462e-05, "loss": 0.0132, "step": 63960 }, { "grad_norm": 0.1556185632944107, "learning_rate": 1.0582360724878898e-05, "loss": 0.0117, "step": 63970 }, { "grad_norm": 0.22743181884288788, "learning_rate": 1.0569648425499651e-05, "loss": 0.0179, "step": 63980 }, { "grad_norm": 0.18753406405448914, "learning_rate": 1.0556942863689063e-05, "loss": 0.0171, "step": 63990 }, { "grad_norm": 0.1924988478422165, "learning_rate": 1.054424404161819e-05, "loss": 0.0134, "step": 64000 }, { "grad_norm": 0.14175531268119812, "learning_rate": 1.053155196145686e-05, "loss": 0.0141, "step": 64010 }, { "grad_norm": 0.15639500319957733, "learning_rate": 1.0518866625373863e-05, "loss": 0.0137, "step": 64020 }, { "grad_norm": 0.14814327657222748, "learning_rate": 1.0506188035536735e-05, "loss": 0.0158, "step": 64030 }, { "grad_norm": 0.1612798571586609, "learning_rate": 1.0493516194111919e-05, "loss": 0.0145, "step": 64040 }, { "grad_norm": 0.16376903653144836, "learning_rate": 1.0480851103264688e-05, "loss": 0.012, "step": 64050 }, { "grad_norm": 0.10939878970384598, "learning_rate": 1.046819276515913e-05, "loss": 0.0118, "step": 64060 }, { "grad_norm": 0.2198995053768158, "learning_rate": 1.0455541181958256e-05, "loss": 0.012, "step": 64070 }, { "grad_norm": 0.16843003034591675, "learning_rate": 1.0442896355823822e-05, "loss": 0.0143, "step": 64080 }, { "grad_norm": 0.16612091660499573, "learning_rate": 1.0430258288916539e-05, "loss": 0.0137, "step": 64090 }, { "grad_norm": 0.1612771451473236, "learning_rate": 1.0417626983395868e-05, "loss": 0.0154, "step": 64100 }, { "grad_norm": 0.15362408757209778, "learning_rate": 1.0405002441420165e-05, "loss": 0.0138, "step": 64110 }, { "grad_norm": 0.1878128945827484, "learning_rate": 1.039238466514662e-05, "loss": 0.0155, "step": 64120 }, { "grad_norm": 0.13100261986255646, "learning_rate": 1.0379773656731262e-05, "loss": 0.0183, "step": 64130 }, { "grad_norm": 0.25210854411125183, "learning_rate": 1.0367169418328986e-05, "loss": 0.0141, "step": 64140 }, { "grad_norm": 0.21002882719039917, "learning_rate": 1.0354571952093484e-05, "loss": 0.02, "step": 64150 }, { "grad_norm": 0.13611260056495667, "learning_rate": 1.034198126017733e-05, "loss": 0.013, "step": 64160 }, { "grad_norm": 0.2022136002779007, "learning_rate": 1.0329397344731928e-05, "loss": 0.0178, "step": 64170 }, { "grad_norm": 0.21711058914661407, "learning_rate": 1.0316820207907524e-05, "loss": 0.0182, "step": 64180 }, { "grad_norm": 0.1464618742465973, "learning_rate": 1.0304249851853215e-05, "loss": 0.0176, "step": 64190 }, { "grad_norm": 0.13241301476955414, "learning_rate": 1.0291686278716933e-05, "loss": 0.0157, "step": 64200 }, { "grad_norm": 0.16126516461372375, "learning_rate": 1.0279129490645418e-05, "loss": 0.0143, "step": 64210 }, { "grad_norm": 0.20678485929965973, "learning_rate": 1.0266579489784328e-05, "loss": 0.0123, "step": 64220 }, { "grad_norm": 0.1601167768239975, "learning_rate": 1.0254036278278084e-05, "loss": 0.0149, "step": 64230 }, { "grad_norm": 0.11048018932342529, "learning_rate": 1.0241499858269982e-05, "loss": 0.011, "step": 64240 }, { "grad_norm": 0.23367895185947418, "learning_rate": 1.0228970231902169e-05, "loss": 0.0137, "step": 64250 }, { "grad_norm": 0.23115570843219757, "learning_rate": 1.0216447401315582e-05, "loss": 0.0169, "step": 64260 }, { "grad_norm": 0.24009796977043152, "learning_rate": 1.020393136865007e-05, "loss": 0.011, "step": 64270 }, { "grad_norm": 0.24511978030204773, "learning_rate": 1.0191422136044242e-05, "loss": 0.0225, "step": 64280 }, { "grad_norm": 0.15773852169513702, "learning_rate": 1.017891970563563e-05, "loss": 0.0155, "step": 64290 }, { "grad_norm": 0.20908485352993011, "learning_rate": 1.0166424079560516e-05, "loss": 0.0208, "step": 64300 }, { "grad_norm": 0.174277663230896, "learning_rate": 1.0153935259954078e-05, "loss": 0.0151, "step": 64310 }, { "grad_norm": 0.20728977024555206, "learning_rate": 1.0141453248950311e-05, "loss": 0.0181, "step": 64320 }, { "grad_norm": 0.24001646041870117, "learning_rate": 1.0128978048682054e-05, "loss": 0.0187, "step": 64330 }, { "grad_norm": 0.3182049095630646, "learning_rate": 1.0116509661280982e-05, "loss": 0.0227, "step": 64340 }, { "grad_norm": 0.21086058020591736, "learning_rate": 1.0104048088877576e-05, "loss": 0.014, "step": 64350 }, { "grad_norm": 0.09845943003892899, "learning_rate": 1.0091593333601201e-05, "loss": 0.017, "step": 64360 }, { "grad_norm": 0.2328292578458786, "learning_rate": 1.0079145397580031e-05, "loss": 0.0156, "step": 64370 }, { "grad_norm": 0.22817116975784302, "learning_rate": 1.006670428294107e-05, "loss": 0.0234, "step": 64380 }, { "grad_norm": 0.16488094627857208, "learning_rate": 1.0054269991810166e-05, "loss": 0.0158, "step": 64390 }, { "grad_norm": 0.14386479556560516, "learning_rate": 1.0041842526312024e-05, "loss": 0.0159, "step": 64400 }, { "grad_norm": 0.1869511604309082, "learning_rate": 1.0029421888570101e-05, "loss": 0.0128, "step": 64410 }, { "grad_norm": 0.2285587638616562, "learning_rate": 1.0017008080706813e-05, "loss": 0.0237, "step": 64420 }, { "grad_norm": 0.12873686850070953, "learning_rate": 1.0004601104843287e-05, "loss": 0.0141, "step": 64430 }, { "grad_norm": 0.12138952314853668, "learning_rate": 9.992200963099562e-06, "loss": 0.0161, "step": 64440 }, { "grad_norm": 0.19760312139987946, "learning_rate": 9.979807657594486e-06, "loss": 0.0159, "step": 64450 }, { "grad_norm": 0.20783185958862305, "learning_rate": 9.967421190445703e-06, "loss": 0.0167, "step": 64460 }, { "grad_norm": 0.19216325879096985, "learning_rate": 9.955041563769769e-06, "loss": 0.0146, "step": 64470 }, { "grad_norm": 0.16013319790363312, "learning_rate": 9.942668779681974e-06, "loss": 0.0205, "step": 64480 }, { "grad_norm": 0.21441221237182617, "learning_rate": 9.930302840296541e-06, "loss": 0.014, "step": 64490 }, { "grad_norm": 0.17998012900352478, "learning_rate": 9.917943747726426e-06, "loss": 0.0147, "step": 64500 }, { "grad_norm": 0.20684629678726196, "learning_rate": 9.905591504083484e-06, "loss": 0.0137, "step": 64510 }, { "grad_norm": 0.21892905235290527, "learning_rate": 9.893246111478382e-06, "loss": 0.017, "step": 64520 }, { "grad_norm": 0.227031409740448, "learning_rate": 9.88090757202057e-06, "loss": 0.0145, "step": 64530 }, { "grad_norm": 0.21219711005687714, "learning_rate": 9.868575887818421e-06, "loss": 0.0195, "step": 64540 }, { "grad_norm": 0.1541087031364441, "learning_rate": 9.856251060979044e-06, "loss": 0.016, "step": 64550 }, { "grad_norm": 0.15644507110118866, "learning_rate": 9.843933093608426e-06, "loss": 0.0188, "step": 64560 }, { "grad_norm": 0.21295283734798431, "learning_rate": 9.831621987811368e-06, "loss": 0.0208, "step": 64570 }, { "grad_norm": 0.1886330544948578, "learning_rate": 9.819317745691509e-06, "loss": 0.0125, "step": 64580 }, { "grad_norm": 0.16098658740520477, "learning_rate": 9.8070203693513e-06, "loss": 0.0147, "step": 64590 }, { "grad_norm": 0.20786407589912415, "learning_rate": 9.794729860892048e-06, "loss": 0.0154, "step": 64600 }, { "grad_norm": 0.23696091771125793, "learning_rate": 9.782446222413827e-06, "loss": 0.0195, "step": 64610 }, { "grad_norm": 0.22770844399929047, "learning_rate": 9.770169456015598e-06, "loss": 0.0202, "step": 64620 }, { "grad_norm": 0.14891645312309265, "learning_rate": 9.75789956379512e-06, "loss": 0.0151, "step": 64630 }, { "grad_norm": 0.1713750809431076, "learning_rate": 9.74563654784898e-06, "loss": 0.0148, "step": 64640 }, { "grad_norm": 0.17615042626857758, "learning_rate": 9.733380410272596e-06, "loss": 0.0143, "step": 64650 }, { "grad_norm": 0.18308918178081512, "learning_rate": 9.721131153160207e-06, "loss": 0.0136, "step": 64660 }, { "grad_norm": 0.1489996314048767, "learning_rate": 9.708888778604886e-06, "loss": 0.0144, "step": 64670 }, { "grad_norm": 0.16836406290531158, "learning_rate": 9.696653288698488e-06, "loss": 0.015, "step": 64680 }, { "grad_norm": 0.1524864286184311, "learning_rate": 9.684424685531763e-06, "loss": 0.0243, "step": 64690 }, { "grad_norm": 0.19473575055599213, "learning_rate": 9.672202971194216e-06, "loss": 0.019, "step": 64700 }, { "grad_norm": 0.2691539525985718, "learning_rate": 9.659988147774213e-06, "loss": 0.0174, "step": 64710 }, { "grad_norm": 0.15310053527355194, "learning_rate": 9.647780217358942e-06, "loss": 0.0128, "step": 64720 }, { "grad_norm": 0.17810824513435364, "learning_rate": 9.635579182034376e-06, "loss": 0.0159, "step": 64730 }, { "grad_norm": 0.27248018980026245, "learning_rate": 9.623385043885386e-06, "loss": 0.021, "step": 64740 }, { "grad_norm": 0.18857213854789734, "learning_rate": 9.61119780499557e-06, "loss": 0.0129, "step": 64750 }, { "grad_norm": 0.25817933678627014, "learning_rate": 9.599017467447418e-06, "loss": 0.0154, "step": 64760 }, { "grad_norm": 0.17064742743968964, "learning_rate": 9.586844033322206e-06, "loss": 0.0144, "step": 64770 }, { "grad_norm": 0.2691056728363037, "learning_rate": 9.574677504700052e-06, "loss": 0.019, "step": 64780 }, { "grad_norm": 0.20167994499206543, "learning_rate": 9.562517883659877e-06, "loss": 0.023, "step": 64790 }, { "grad_norm": 0.22143253684043884, "learning_rate": 9.55036517227944e-06, "loss": 0.0138, "step": 64800 }, { "grad_norm": 0.20853163301944733, "learning_rate": 9.538219372635282e-06, "loss": 0.0137, "step": 64810 }, { "grad_norm": 0.16812048852443695, "learning_rate": 9.526080486802802e-06, "loss": 0.0132, "step": 64820 }, { "grad_norm": 0.200808584690094, "learning_rate": 9.513948516856203e-06, "loss": 0.0151, "step": 64830 }, { "grad_norm": 0.23371700942516327, "learning_rate": 9.501823464868504e-06, "loss": 0.0125, "step": 64840 }, { "grad_norm": 0.21310462057590485, "learning_rate": 9.489705332911547e-06, "loss": 0.0121, "step": 64850 }, { "grad_norm": 0.22562211751937866, "learning_rate": 9.477594123055994e-06, "loss": 0.0161, "step": 64860 }, { "grad_norm": 0.19935360550880432, "learning_rate": 9.465489837371321e-06, "loss": 0.0182, "step": 64870 }, { "grad_norm": 0.17815083265304565, "learning_rate": 9.453392477925794e-06, "loss": 0.0142, "step": 64880 }, { "grad_norm": 0.16726768016815186, "learning_rate": 9.441302046786566e-06, "loss": 0.0147, "step": 64890 }, { "grad_norm": 0.191597118973732, "learning_rate": 9.429218546019519e-06, "loss": 0.0131, "step": 64900 }, { "grad_norm": 0.29461899399757385, "learning_rate": 9.41714197768941e-06, "loss": 0.0217, "step": 64910 }, { "grad_norm": 0.24873290956020355, "learning_rate": 9.405072343859805e-06, "loss": 0.0169, "step": 64920 }, { "grad_norm": 0.27026689052581787, "learning_rate": 9.393009646593043e-06, "loss": 0.0153, "step": 64930 }, { "grad_norm": 0.15817922353744507, "learning_rate": 9.38095388795035e-06, "loss": 0.0154, "step": 64940 }, { "grad_norm": 0.16590185463428497, "learning_rate": 9.36890506999169e-06, "loss": 0.0172, "step": 64950 }, { "grad_norm": 0.1656460464000702, "learning_rate": 9.356863194775894e-06, "loss": 0.0202, "step": 64960 }, { "grad_norm": 0.23955342173576355, "learning_rate": 9.344828264360583e-06, "loss": 0.017, "step": 64970 }, { "grad_norm": 0.13448455929756165, "learning_rate": 9.332800280802201e-06, "loss": 0.0158, "step": 64980 }, { "grad_norm": 0.1594506949186325, "learning_rate": 9.32077924615602e-06, "loss": 0.0126, "step": 64990 }, { "grad_norm": 0.18907928466796875, "learning_rate": 9.308765162476063e-06, "loss": 0.0157, "step": 65000 }, { "grad_norm": 0.26946502923965454, "learning_rate": 9.296758031815239e-06, "loss": 0.0173, "step": 65010 }, { "grad_norm": 0.2608993351459503, "learning_rate": 9.284757856225229e-06, "loss": 0.0161, "step": 65020 }, { "grad_norm": 0.172334685921669, "learning_rate": 9.272764637756538e-06, "loss": 0.0167, "step": 65030 }, { "grad_norm": 0.23797987401485443, "learning_rate": 9.260778378458479e-06, "loss": 0.019, "step": 65040 }, { "grad_norm": 0.31478267908096313, "learning_rate": 9.248799080379172e-06, "loss": 0.0162, "step": 65050 }, { "grad_norm": 0.1784188598394394, "learning_rate": 9.236826745565558e-06, "loss": 0.0194, "step": 65060 }, { "grad_norm": 0.23009249567985535, "learning_rate": 9.224861376063388e-06, "loss": 0.0153, "step": 65070 }, { "grad_norm": 0.13458727300167084, "learning_rate": 9.212902973917192e-06, "loss": 0.0111, "step": 65080 }, { "grad_norm": 0.24171075224876404, "learning_rate": 9.20095154117035e-06, "loss": 0.0136, "step": 65090 }, { "grad_norm": 0.19878527522087097, "learning_rate": 9.189007079865036e-06, "loss": 0.0136, "step": 65100 }, { "grad_norm": 0.181329146027565, "learning_rate": 9.177069592042226e-06, "loss": 0.0158, "step": 65110 }, { "grad_norm": 0.21804559230804443, "learning_rate": 9.165139079741724e-06, "loss": 0.0189, "step": 65120 }, { "grad_norm": 0.24553436040878296, "learning_rate": 9.153215545002098e-06, "loss": 0.0209, "step": 65130 }, { "grad_norm": 0.15850062668323517, "learning_rate": 9.141298989860798e-06, "loss": 0.0226, "step": 65140 }, { "grad_norm": 0.23487162590026855, "learning_rate": 9.129389416353994e-06, "loss": 0.0156, "step": 65150 }, { "grad_norm": 0.1717662513256073, "learning_rate": 9.11748682651673e-06, "loss": 0.0117, "step": 65160 }, { "grad_norm": 0.18205681443214417, "learning_rate": 9.105591222382837e-06, "loss": 0.0152, "step": 65170 }, { "grad_norm": 0.24598437547683716, "learning_rate": 9.093702605984915e-06, "loss": 0.0125, "step": 65180 }, { "grad_norm": 0.21177954971790314, "learning_rate": 9.081820979354455e-06, "loss": 0.0169, "step": 65190 }, { "grad_norm": 0.17327387630939484, "learning_rate": 9.069946344521663e-06, "loss": 0.0142, "step": 65200 }, { "grad_norm": 0.2190670520067215, "learning_rate": 9.058078703515598e-06, "loss": 0.0148, "step": 65210 }, { "grad_norm": 0.1865990310907364, "learning_rate": 9.046218058364125e-06, "loss": 0.0184, "step": 65220 }, { "grad_norm": 0.2295663058757782, "learning_rate": 9.034364411093893e-06, "loss": 0.014, "step": 65230 }, { "grad_norm": 0.2368665337562561, "learning_rate": 9.022517763730371e-06, "loss": 0.0151, "step": 65240 }, { "grad_norm": 0.1196628138422966, "learning_rate": 9.010678118297827e-06, "loss": 0.0141, "step": 65250 }, { "grad_norm": 0.21531209349632263, "learning_rate": 8.998845476819345e-06, "loss": 0.0157, "step": 65260 }, { "grad_norm": 0.19544243812561035, "learning_rate": 8.987019841316773e-06, "loss": 0.0162, "step": 65270 }, { "grad_norm": 0.2572264075279236, "learning_rate": 8.975201213810802e-06, "loss": 0.0125, "step": 65280 }, { "grad_norm": 0.17800965905189514, "learning_rate": 8.963389596320915e-06, "loss": 0.0138, "step": 65290 }, { "grad_norm": 0.20389394462108612, "learning_rate": 8.951584990865391e-06, "loss": 0.0136, "step": 65300 }, { "grad_norm": 0.16795358061790466, "learning_rate": 8.939787399461319e-06, "loss": 0.0179, "step": 65310 }, { "grad_norm": 0.29019632935523987, "learning_rate": 8.927996824124591e-06, "loss": 0.0156, "step": 65320 }, { "grad_norm": 0.2561527490615845, "learning_rate": 8.916213266869854e-06, "loss": 0.0156, "step": 65330 }, { "grad_norm": 0.16543008387088776, "learning_rate": 8.904436729710658e-06, "loss": 0.0151, "step": 65340 }, { "grad_norm": 0.1683453470468521, "learning_rate": 8.892667214659245e-06, "loss": 0.0126, "step": 65350 }, { "grad_norm": 0.18368840217590332, "learning_rate": 8.880904723726713e-06, "loss": 0.017, "step": 65360 }, { "grad_norm": 0.23568850755691528, "learning_rate": 8.869149258922971e-06, "loss": 0.0127, "step": 65370 }, { "grad_norm": 0.09724895656108856, "learning_rate": 8.857400822256662e-06, "loss": 0.0141, "step": 65380 }, { "grad_norm": 0.1985107660293579, "learning_rate": 8.845659415735324e-06, "loss": 0.0195, "step": 65390 }, { "grad_norm": 0.16833354532718658, "learning_rate": 8.83392504136521e-06, "loss": 0.0147, "step": 65400 }, { "grad_norm": 0.15182003378868103, "learning_rate": 8.822197701151407e-06, "loss": 0.0175, "step": 65410 }, { "grad_norm": 0.25621598958969116, "learning_rate": 8.810477397097804e-06, "loss": 0.0163, "step": 65420 }, { "grad_norm": 0.10551828891038895, "learning_rate": 8.798764131207077e-06, "loss": 0.0179, "step": 65430 }, { "grad_norm": 0.19676662981510162, "learning_rate": 8.787057905480706e-06, "loss": 0.0157, "step": 65440 }, { "grad_norm": 0.21885955333709717, "learning_rate": 8.775358721918958e-06, "loss": 0.0151, "step": 65450 }, { "grad_norm": 0.19249103963375092, "learning_rate": 8.763666582520923e-06, "loss": 0.0156, "step": 65460 }, { "grad_norm": 0.18161842226982117, "learning_rate": 8.751981489284445e-06, "loss": 0.015, "step": 65470 }, { "grad_norm": 0.2345002144575119, "learning_rate": 8.740303444206188e-06, "loss": 0.017, "step": 65480 }, { "grad_norm": 0.20345532894134521, "learning_rate": 8.72863244928162e-06, "loss": 0.0128, "step": 65490 }, { "grad_norm": 0.12153277546167374, "learning_rate": 8.716968506504991e-06, "loss": 0.0222, "step": 65500 }, { "grad_norm": 0.19986507296562195, "learning_rate": 8.70531161786935e-06, "loss": 0.0151, "step": 65510 }, { "grad_norm": 0.19647054374217987, "learning_rate": 8.693661785366558e-06, "loss": 0.0215, "step": 65520 }, { "grad_norm": 0.15876805782318115, "learning_rate": 8.682019010987208e-06, "loss": 0.0189, "step": 65530 }, { "grad_norm": 0.1631588190793991, "learning_rate": 8.670383296720786e-06, "loss": 0.0133, "step": 65540 }, { "grad_norm": 0.21458449959754944, "learning_rate": 8.658754644555478e-06, "loss": 0.0139, "step": 65550 }, { "grad_norm": 0.11775410920381546, "learning_rate": 8.647133056478313e-06, "loss": 0.0146, "step": 65560 }, { "grad_norm": 0.1960308998823166, "learning_rate": 8.635518534475123e-06, "loss": 0.016, "step": 65570 }, { "grad_norm": 0.21889255940914154, "learning_rate": 8.623911080530467e-06, "loss": 0.019, "step": 65580 }, { "grad_norm": 0.15299928188323975, "learning_rate": 8.61231069662779e-06, "loss": 0.0159, "step": 65590 }, { "grad_norm": 0.1791744977235794, "learning_rate": 8.600717384749252e-06, "loss": 0.0155, "step": 65600 }, { "grad_norm": 0.19435963034629822, "learning_rate": 8.58913114687584e-06, "loss": 0.0139, "step": 65610 }, { "grad_norm": 0.22627869248390198, "learning_rate": 8.57755198498732e-06, "loss": 0.0173, "step": 65620 }, { "grad_norm": 0.2730368673801422, "learning_rate": 8.565979901062265e-06, "loss": 0.0228, "step": 65630 }, { "grad_norm": 0.173024982213974, "learning_rate": 8.554414897078033e-06, "loss": 0.0135, "step": 65640 }, { "grad_norm": 0.17195917665958405, "learning_rate": 8.542856975010727e-06, "loss": 0.0137, "step": 65650 }, { "grad_norm": 0.17951864004135132, "learning_rate": 8.531306136835337e-06, "loss": 0.0133, "step": 65660 }, { "grad_norm": 0.1616295725107193, "learning_rate": 8.519762384525548e-06, "loss": 0.0134, "step": 65670 }, { "grad_norm": 0.24099162220954895, "learning_rate": 8.508225720053875e-06, "loss": 0.0163, "step": 65680 }, { "grad_norm": 0.16383622586727142, "learning_rate": 8.496696145391625e-06, "loss": 0.0159, "step": 65690 }, { "grad_norm": 0.20873114466667175, "learning_rate": 8.485173662508889e-06, "loss": 0.013, "step": 65700 }, { "grad_norm": 0.19539900124073029, "learning_rate": 8.473658273374536e-06, "loss": 0.0158, "step": 65710 }, { "grad_norm": 0.19921232759952545, "learning_rate": 8.462149979956253e-06, "loss": 0.0106, "step": 65720 }, { "grad_norm": 0.16822388768196106, "learning_rate": 8.450648784220461e-06, "loss": 0.0146, "step": 65730 }, { "grad_norm": 0.15140356123447418, "learning_rate": 8.439154688132417e-06, "loss": 0.0164, "step": 65740 }, { "grad_norm": 0.2735850512981415, "learning_rate": 8.427667693656143e-06, "loss": 0.0136, "step": 65750 }, { "grad_norm": 0.18627797067165375, "learning_rate": 8.416187802754454e-06, "loss": 0.0149, "step": 65760 }, { "grad_norm": 0.1840154528617859, "learning_rate": 8.404715017388965e-06, "loss": 0.0117, "step": 65770 }, { "grad_norm": 0.16288498044013977, "learning_rate": 8.393249339520015e-06, "loss": 0.0187, "step": 65780 }, { "grad_norm": 0.10476832091808319, "learning_rate": 8.381790771106834e-06, "loss": 0.0119, "step": 65790 }, { "grad_norm": 0.29086238145828247, "learning_rate": 8.370339314107339e-06, "loss": 0.0131, "step": 65800 }, { "grad_norm": 0.18709784746170044, "learning_rate": 8.358894970478281e-06, "loss": 0.0157, "step": 65810 }, { "grad_norm": 0.19352379441261292, "learning_rate": 8.347457742175196e-06, "loss": 0.0166, "step": 65820 }, { "grad_norm": 0.21134944260120392, "learning_rate": 8.33602763115236e-06, "loss": 0.0129, "step": 65830 }, { "grad_norm": 0.16246038675308228, "learning_rate": 8.324604639362916e-06, "loss": 0.0223, "step": 65840 }, { "grad_norm": 0.15516509115695953, "learning_rate": 8.31318876875869e-06, "loss": 0.0144, "step": 65850 }, { "grad_norm": 0.20528411865234375, "learning_rate": 8.30178002129039e-06, "loss": 0.0131, "step": 65860 }, { "grad_norm": 0.14007939398288727, "learning_rate": 8.290378398907423e-06, "loss": 0.0142, "step": 65870 }, { "grad_norm": 0.23983240127563477, "learning_rate": 8.278983903558029e-06, "loss": 0.0183, "step": 65880 }, { "grad_norm": 0.3553548753261566, "learning_rate": 8.26759653718921e-06, "loss": 0.0152, "step": 65890 }, { "grad_norm": 0.194975808262825, "learning_rate": 8.25621630174676e-06, "loss": 0.0159, "step": 65900 }, { "grad_norm": 0.13223859667778015, "learning_rate": 8.244843199175261e-06, "loss": 0.0221, "step": 65910 }, { "grad_norm": 0.16273684799671173, "learning_rate": 8.233477231418046e-06, "loss": 0.0108, "step": 65920 }, { "grad_norm": 0.2524968087673187, "learning_rate": 8.22211840041725e-06, "loss": 0.0146, "step": 65930 }, { "grad_norm": 0.22550739347934723, "learning_rate": 8.210766708113792e-06, "loss": 0.018, "step": 65940 }, { "grad_norm": 0.222733736038208, "learning_rate": 8.199422156447367e-06, "loss": 0.0174, "step": 65950 }, { "grad_norm": 0.19775091111660004, "learning_rate": 8.188084747356451e-06, "loss": 0.0157, "step": 65960 }, { "grad_norm": 0.1972685307264328, "learning_rate": 8.176754482778299e-06, "loss": 0.0154, "step": 65970 }, { "grad_norm": 0.2022722363471985, "learning_rate": 8.165431364648918e-06, "loss": 0.0166, "step": 65980 }, { "grad_norm": 0.21282044053077698, "learning_rate": 8.154115394903162e-06, "loss": 0.0133, "step": 65990 }, { "grad_norm": 0.21026000380516052, "learning_rate": 8.142806575474582e-06, "loss": 0.0146, "step": 66000 }, { "grad_norm": 0.23986758291721344, "learning_rate": 8.131504908295562e-06, "loss": 0.017, "step": 66010 }, { "grad_norm": 0.18162797391414642, "learning_rate": 8.120210395297262e-06, "loss": 0.0131, "step": 66020 }, { "grad_norm": 0.15240679681301117, "learning_rate": 8.108923038409565e-06, "loss": 0.0154, "step": 66030 }, { "grad_norm": 0.241294264793396, "learning_rate": 8.097642839561226e-06, "loss": 0.015, "step": 66040 }, { "grad_norm": 0.2238077074289322, "learning_rate": 8.08636980067966e-06, "loss": 0.018, "step": 66050 }, { "grad_norm": 0.20478284358978271, "learning_rate": 8.075103923691186e-06, "loss": 0.0105, "step": 66060 }, { "grad_norm": 0.16436876356601715, "learning_rate": 8.063845210520793e-06, "loss": 0.0117, "step": 66070 }, { "grad_norm": 0.14686834812164307, "learning_rate": 8.052593663092295e-06, "loss": 0.0171, "step": 66080 }, { "grad_norm": 0.2541458010673523, "learning_rate": 8.04134928332827e-06, "loss": 0.0155, "step": 66090 }, { "grad_norm": 0.11234168708324432, "learning_rate": 8.030112073150086e-06, "loss": 0.0196, "step": 66100 }, { "grad_norm": 0.20003058016300201, "learning_rate": 8.018882034477881e-06, "loss": 0.0125, "step": 66110 }, { "grad_norm": 0.21237090229988098, "learning_rate": 8.007659169230541e-06, "loss": 0.0109, "step": 66120 }, { "grad_norm": 0.20530524849891663, "learning_rate": 7.996443479325755e-06, "loss": 0.0113, "step": 66130 }, { "grad_norm": 0.15372151136398315, "learning_rate": 7.985234966679977e-06, "loss": 0.0177, "step": 66140 }, { "grad_norm": 0.1423938274383545, "learning_rate": 7.974033633208438e-06, "loss": 0.0163, "step": 66150 }, { "grad_norm": 0.12944214046001434, "learning_rate": 7.962839480825135e-06, "loss": 0.0126, "step": 66160 }, { "grad_norm": 0.18760259449481964, "learning_rate": 7.951652511442858e-06, "loss": 0.019, "step": 66170 }, { "grad_norm": 0.2121400237083435, "learning_rate": 7.940472726973125e-06, "loss": 0.015, "step": 66180 }, { "grad_norm": 0.15102462470531464, "learning_rate": 7.929300129326289e-06, "loss": 0.0142, "step": 66190 }, { "grad_norm": 0.33082523941993713, "learning_rate": 7.91813472041142e-06, "loss": 0.018, "step": 66200 }, { "grad_norm": 0.15491454303264618, "learning_rate": 7.906976502136376e-06, "loss": 0.0159, "step": 66210 }, { "grad_norm": 0.26609763503074646, "learning_rate": 7.89582547640782e-06, "loss": 0.0151, "step": 66220 }, { "grad_norm": 0.17135605216026306, "learning_rate": 7.884681645131115e-06, "loss": 0.0126, "step": 66230 }, { "grad_norm": 0.22541218996047974, "learning_rate": 7.87354501021048e-06, "loss": 0.0117, "step": 66240 }, { "grad_norm": 0.18465641140937805, "learning_rate": 7.86241557354882e-06, "loss": 0.0144, "step": 66250 }, { "grad_norm": 0.14818373322486877, "learning_rate": 7.8512933370479e-06, "loss": 0.0145, "step": 66260 }, { "grad_norm": 0.23340071737766266, "learning_rate": 7.840178302608158e-06, "loss": 0.0155, "step": 66270 }, { "grad_norm": 0.19419154524803162, "learning_rate": 7.82907047212888e-06, "loss": 0.0142, "step": 66280 }, { "grad_norm": 0.24426664412021637, "learning_rate": 7.81796984750809e-06, "loss": 0.0177, "step": 66290 }, { "grad_norm": 0.14867158234119415, "learning_rate": 7.806876430642546e-06, "loss": 0.0156, "step": 66300 }, { "grad_norm": 0.17934070527553558, "learning_rate": 7.795790223427862e-06, "loss": 0.0152, "step": 66310 }, { "grad_norm": 0.1992979198694229, "learning_rate": 7.784711227758324e-06, "loss": 0.0174, "step": 66320 }, { "grad_norm": 0.2055814266204834, "learning_rate": 7.773639445527053e-06, "loss": 0.0158, "step": 66330 }, { "grad_norm": 0.19190460443496704, "learning_rate": 7.762574878625905e-06, "loss": 0.0132, "step": 66340 }, { "grad_norm": 0.226257786154747, "learning_rate": 7.751517528945513e-06, "loss": 0.0126, "step": 66350 }, { "grad_norm": 0.17773285508155823, "learning_rate": 7.740467398375278e-06, "loss": 0.0142, "step": 66360 }, { "grad_norm": 0.2654050588607788, "learning_rate": 7.729424488803378e-06, "loss": 0.0138, "step": 66370 }, { "grad_norm": 0.14572612941265106, "learning_rate": 7.71838880211671e-06, "loss": 0.0219, "step": 66380 }, { "grad_norm": 0.1764792948961258, "learning_rate": 7.707360340200997e-06, "loss": 0.0132, "step": 66390 }, { "grad_norm": 0.15966658294200897, "learning_rate": 7.696339104940697e-06, "loss": 0.0148, "step": 66400 }, { "grad_norm": 0.22830960154533386, "learning_rate": 7.685325098219038e-06, "loss": 0.0154, "step": 66410 }, { "grad_norm": 0.16281019151210785, "learning_rate": 7.674318321918017e-06, "loss": 0.0174, "step": 66420 }, { "grad_norm": 0.22358381748199463, "learning_rate": 7.663318777918366e-06, "loss": 0.0179, "step": 66430 }, { "grad_norm": 0.16392193734645844, "learning_rate": 7.652326468099647e-06, "loss": 0.0173, "step": 66440 }, { "grad_norm": 0.21520429849624634, "learning_rate": 7.641341394340096e-06, "loss": 0.0131, "step": 66450 }, { "grad_norm": 0.3268183767795563, "learning_rate": 7.630363558516818e-06, "loss": 0.0157, "step": 66460 }, { "grad_norm": 0.213829904794693, "learning_rate": 7.619392962505578e-06, "loss": 0.0157, "step": 66470 }, { "grad_norm": 0.1526528000831604, "learning_rate": 7.6084296081809725e-06, "loss": 0.0184, "step": 66480 }, { "grad_norm": 0.22440701723098755, "learning_rate": 7.597473497416347e-06, "loss": 0.0178, "step": 66490 }, { "grad_norm": 0.15040595829486847, "learning_rate": 7.586524632083764e-06, "loss": 0.016, "step": 66500 }, { "grad_norm": 0.24540220201015472, "learning_rate": 7.5755830140541326e-06, "loss": 0.0154, "step": 66510 }, { "grad_norm": 0.1532428115606308, "learning_rate": 7.5646486451970475e-06, "loss": 0.0136, "step": 66520 }, { "grad_norm": 0.24530456960201263, "learning_rate": 7.553721527380897e-06, "loss": 0.0177, "step": 66530 }, { "grad_norm": 0.11697620898485184, "learning_rate": 7.542801662472826e-06, "loss": 0.0128, "step": 66540 }, { "grad_norm": 0.16507084667682648, "learning_rate": 7.5318890523387475e-06, "loss": 0.0137, "step": 66550 }, { "grad_norm": 0.20300862193107605, "learning_rate": 7.52098369884332e-06, "loss": 0.0164, "step": 66560 }, { "grad_norm": 0.19756095111370087, "learning_rate": 7.510085603849992e-06, "loss": 0.0166, "step": 66570 }, { "grad_norm": 0.2335950881242752, "learning_rate": 7.499194769220918e-06, "loss": 0.0153, "step": 66580 }, { "grad_norm": 0.20002181828022003, "learning_rate": 7.48831119681706e-06, "loss": 0.0187, "step": 66590 }, { "grad_norm": 0.20220176875591278, "learning_rate": 7.477434888498119e-06, "loss": 0.0156, "step": 66600 }, { "grad_norm": 0.19715595245361328, "learning_rate": 7.466565846122564e-06, "loss": 0.0197, "step": 66610 }, { "grad_norm": 0.17607325315475464, "learning_rate": 7.455704071547626e-06, "loss": 0.0141, "step": 66620 }, { "grad_norm": 0.25383052229881287, "learning_rate": 7.444849566629247e-06, "loss": 0.014, "step": 66630 }, { "grad_norm": 0.2616034746170044, "learning_rate": 7.434002333222212e-06, "loss": 0.0127, "step": 66640 }, { "grad_norm": 0.24106790125370026, "learning_rate": 7.423162373179976e-06, "loss": 0.017, "step": 66650 }, { "grad_norm": 0.15817572176456451, "learning_rate": 7.412329688354835e-06, "loss": 0.0131, "step": 66660 }, { "grad_norm": 0.19649481773376465, "learning_rate": 7.40150428059776e-06, "loss": 0.0148, "step": 66670 }, { "grad_norm": 0.2224709838628769, "learning_rate": 7.3906861517585354e-06, "loss": 0.0157, "step": 66680 }, { "grad_norm": 0.16114361584186554, "learning_rate": 7.37987530368569e-06, "loss": 0.015, "step": 66690 }, { "grad_norm": 0.15998457372188568, "learning_rate": 7.369071738226474e-06, "loss": 0.0146, "step": 66700 }, { "grad_norm": 0.11596288532018661, "learning_rate": 7.358275457226954e-06, "loss": 0.0129, "step": 66710 }, { "grad_norm": 0.21446222066879272, "learning_rate": 7.347486462531899e-06, "loss": 0.0132, "step": 66720 }, { "grad_norm": 0.16732925176620483, "learning_rate": 7.336704755984858e-06, "loss": 0.0122, "step": 66730 }, { "grad_norm": 0.15865162014961243, "learning_rate": 7.325930339428133e-06, "loss": 0.0183, "step": 66740 }, { "grad_norm": 0.15295538306236267, "learning_rate": 7.315163214702769e-06, "loss": 0.0157, "step": 66750 }, { "grad_norm": 0.21184279024600983, "learning_rate": 7.304403383648595e-06, "loss": 0.0154, "step": 66760 }, { "grad_norm": 0.16049925982952118, "learning_rate": 7.293650848104139e-06, "loss": 0.014, "step": 66770 }, { "grad_norm": 0.29258105158805847, "learning_rate": 7.2829056099067374e-06, "loss": 0.013, "step": 66780 }, { "grad_norm": 0.17465589940547943, "learning_rate": 7.2721676708924494e-06, "loss": 0.018, "step": 66790 }, { "grad_norm": 0.4418568015098572, "learning_rate": 7.261437032896096e-06, "loss": 0.0288, "step": 66800 }, { "grad_norm": 0.25448381900787354, "learning_rate": 7.250713697751255e-06, "loss": 0.0148, "step": 66810 }, { "grad_norm": 0.14737087488174438, "learning_rate": 7.239997667290255e-06, "loss": 0.0132, "step": 66820 }, { "grad_norm": 0.1646401286125183, "learning_rate": 7.2292889433441425e-06, "loss": 0.0144, "step": 66830 }, { "grad_norm": 0.19968144595623016, "learning_rate": 7.218587527742793e-06, "loss": 0.0113, "step": 66840 }, { "grad_norm": 0.173960879445076, "learning_rate": 7.207893422314749e-06, "loss": 0.0151, "step": 66850 }, { "grad_norm": 0.17073826491832733, "learning_rate": 7.1972066288873545e-06, "loss": 0.0128, "step": 66860 }, { "grad_norm": 0.29696038365364075, "learning_rate": 7.186527149286687e-06, "loss": 0.0287, "step": 66870 }, { "grad_norm": 0.24828208982944489, "learning_rate": 7.175854985337576e-06, "loss": 0.0158, "step": 66880 }, { "grad_norm": 0.18761521577835083, "learning_rate": 7.16519013886362e-06, "loss": 0.0171, "step": 66890 }, { "grad_norm": 0.2868260145187378, "learning_rate": 7.154532611687109e-06, "loss": 0.0127, "step": 66900 }, { "grad_norm": 0.17424431443214417, "learning_rate": 7.143882405629177e-06, "loss": 0.0135, "step": 66910 }, { "grad_norm": 0.25181844830513, "learning_rate": 7.133239522509605e-06, "loss": 0.0132, "step": 66920 }, { "grad_norm": 0.1340962052345276, "learning_rate": 7.1226039641469956e-06, "loss": 0.0136, "step": 66930 }, { "grad_norm": 0.2691079080104828, "learning_rate": 7.111975732358678e-06, "loss": 0.018, "step": 66940 }, { "grad_norm": 0.1117033064365387, "learning_rate": 7.101354828960693e-06, "loss": 0.012, "step": 66950 }, { "grad_norm": 0.19673915207386017, "learning_rate": 7.090741255767918e-06, "loss": 0.0125, "step": 66960 }, { "grad_norm": 0.1778993010520935, "learning_rate": 7.080135014593875e-06, "loss": 0.0154, "step": 66970 }, { "grad_norm": 0.2061992585659027, "learning_rate": 7.069536107250896e-06, "loss": 0.0147, "step": 66980 }, { "grad_norm": 0.18307755887508392, "learning_rate": 7.058944535550049e-06, "loss": 0.0112, "step": 66990 }, { "grad_norm": 0.22497409582138062, "learning_rate": 7.048360301301138e-06, "loss": 0.0176, "step": 67000 }, { "grad_norm": 0.2852220833301544, "learning_rate": 7.03778340631272e-06, "loss": 0.0156, "step": 67010 }, { "grad_norm": 0.21958553791046143, "learning_rate": 7.0272138523921e-06, "loss": 0.0142, "step": 67020 }, { "grad_norm": 0.17946037650108337, "learning_rate": 7.016651641345334e-06, "loss": 0.018, "step": 67030 }, { "grad_norm": 0.20285199582576752, "learning_rate": 7.0060967749772e-06, "loss": 0.0154, "step": 67040 }, { "grad_norm": 0.3549937307834625, "learning_rate": 6.995549255091238e-06, "loss": 0.0258, "step": 67050 }, { "grad_norm": 0.19034931063652039, "learning_rate": 6.98500908348973e-06, "loss": 0.0171, "step": 67060 }, { "grad_norm": 0.17383970320224762, "learning_rate": 6.974476261973711e-06, "loss": 0.0173, "step": 67070 }, { "grad_norm": 0.1491393744945526, "learning_rate": 6.963950792342949e-06, "loss": 0.0149, "step": 67080 }, { "grad_norm": 0.14897462725639343, "learning_rate": 6.9534326763959715e-06, "loss": 0.0128, "step": 67090 }, { "grad_norm": 0.3920814096927643, "learning_rate": 6.942921915929995e-06, "loss": 0.018, "step": 67100 }, { "grad_norm": 0.12286458164453506, "learning_rate": 6.9324185127410735e-06, "loss": 0.0154, "step": 67110 }, { "grad_norm": 0.193686842918396, "learning_rate": 6.921922468623921e-06, "loss": 0.0146, "step": 67120 }, { "grad_norm": 0.25178930163383484, "learning_rate": 6.911433785372023e-06, "loss": 0.0146, "step": 67130 }, { "grad_norm": 0.155283585190773, "learning_rate": 6.900952464777632e-06, "loss": 0.0181, "step": 67140 }, { "grad_norm": 0.2987256348133087, "learning_rate": 6.8904785086316815e-06, "loss": 0.0163, "step": 67150 }, { "grad_norm": 0.2671062648296356, "learning_rate": 6.880011918723927e-06, "loss": 0.0138, "step": 67160 }, { "grad_norm": 0.3453499972820282, "learning_rate": 6.86955269684279e-06, "loss": 0.0142, "step": 67170 }, { "grad_norm": 0.2211114764213562, "learning_rate": 6.859100844775473e-06, "loss": 0.0205, "step": 67180 }, { "grad_norm": 0.15973250567913055, "learning_rate": 6.84865636430792e-06, "loss": 0.0175, "step": 67190 }, { "grad_norm": 0.1870642602443695, "learning_rate": 6.838219257224804e-06, "loss": 0.0118, "step": 67200 }, { "grad_norm": 0.21165381371974945, "learning_rate": 6.827789525309536e-06, "loss": 0.0157, "step": 67210 }, { "grad_norm": 0.28562524914741516, "learning_rate": 6.81736717034428e-06, "loss": 0.0172, "step": 67220 }, { "grad_norm": 0.15166422724723816, "learning_rate": 6.806952194109933e-06, "loss": 0.0255, "step": 67230 }, { "grad_norm": 0.17440040409564972, "learning_rate": 6.796544598386112e-06, "loss": 0.013, "step": 67240 }, { "grad_norm": 0.24784404039382935, "learning_rate": 6.786144384951204e-06, "loss": 0.019, "step": 67250 }, { "grad_norm": 0.15253892540931702, "learning_rate": 6.775751555582322e-06, "loss": 0.0159, "step": 67260 }, { "grad_norm": 0.13396112620830536, "learning_rate": 6.76536611205531e-06, "loss": 0.0163, "step": 67270 }, { "grad_norm": 0.14825491607189178, "learning_rate": 6.754988056144762e-06, "loss": 0.0183, "step": 67280 }, { "grad_norm": 0.19967448711395264, "learning_rate": 6.744617389624014e-06, "loss": 0.0138, "step": 67290 }, { "grad_norm": 0.2516539692878723, "learning_rate": 6.734254114265087e-06, "loss": 0.0195, "step": 67300 }, { "grad_norm": 0.1905534714460373, "learning_rate": 6.723898231838843e-06, "loss": 0.0136, "step": 67310 }, { "grad_norm": 0.14199529588222504, "learning_rate": 6.713549744114766e-06, "loss": 0.0153, "step": 67320 }, { "grad_norm": 0.13476687669754028, "learning_rate": 6.703208652861159e-06, "loss": 0.014, "step": 67330 }, { "grad_norm": 0.18185780942440033, "learning_rate": 6.69287495984503e-06, "loss": 0.0161, "step": 67340 }, { "grad_norm": 0.2252354770898819, "learning_rate": 6.6825486668321e-06, "loss": 0.0141, "step": 67350 }, { "grad_norm": 0.1650613695383072, "learning_rate": 6.672229775586886e-06, "loss": 0.0166, "step": 67360 }, { "grad_norm": 0.16249103844165802, "learning_rate": 6.661918287872576e-06, "loss": 0.015, "step": 67370 }, { "grad_norm": 0.2495257705450058, "learning_rate": 6.6516142054511346e-06, "loss": 0.0167, "step": 67380 }, { "grad_norm": 0.18058204650878906, "learning_rate": 6.641317530083241e-06, "loss": 0.0138, "step": 67390 }, { "grad_norm": 0.17554694414138794, "learning_rate": 6.631028263528322e-06, "loss": 0.0148, "step": 67400 }, { "grad_norm": 0.24198442697525024, "learning_rate": 6.620746407544537e-06, "loss": 0.0135, "step": 67410 }, { "grad_norm": 0.22429168224334717, "learning_rate": 6.610471963888742e-06, "loss": 0.0156, "step": 67420 }, { "grad_norm": 0.15878570079803467, "learning_rate": 6.600204934316606e-06, "loss": 0.0113, "step": 67430 }, { "grad_norm": 0.19763723015785217, "learning_rate": 6.589945320582452e-06, "loss": 0.014, "step": 67440 }, { "grad_norm": 0.20925411581993103, "learning_rate": 6.579693124439374e-06, "loss": 0.0127, "step": 67450 }, { "grad_norm": 0.2532055675983429, "learning_rate": 6.56944834763919e-06, "loss": 0.0137, "step": 67460 }, { "grad_norm": 0.20568139851093292, "learning_rate": 6.5592109919324575e-06, "loss": 0.0146, "step": 67470 }, { "grad_norm": 0.14210592210292816, "learning_rate": 6.54898105906846e-06, "loss": 0.0173, "step": 67480 }, { "grad_norm": 0.13260917365550995, "learning_rate": 6.5387585507952155e-06, "loss": 0.02, "step": 67490 }, { "grad_norm": 0.24230986833572388, "learning_rate": 6.528543468859461e-06, "loss": 0.0152, "step": 67500 }, { "grad_norm": 0.25389713048934937, "learning_rate": 6.518335815006682e-06, "loss": 0.0183, "step": 67510 }, { "grad_norm": 0.2922894358634949, "learning_rate": 6.5081355909810845e-06, "loss": 0.015, "step": 67520 }, { "grad_norm": 0.14356474578380585, "learning_rate": 6.497942798525608e-06, "loss": 0.0187, "step": 67530 }, { "grad_norm": 0.23878830671310425, "learning_rate": 6.487757439381936e-06, "loss": 0.0189, "step": 67540 }, { "grad_norm": 0.16693419218063354, "learning_rate": 6.477579515290433e-06, "loss": 0.0155, "step": 67550 }, { "grad_norm": 0.16270135343074799, "learning_rate": 6.467409027990273e-06, "loss": 0.0183, "step": 67560 }, { "grad_norm": 0.19531303644180298, "learning_rate": 6.457245979219279e-06, "loss": 0.0164, "step": 67570 }, { "grad_norm": 0.23074619472026825, "learning_rate": 6.447090370714054e-06, "loss": 0.0211, "step": 67580 }, { "grad_norm": 0.14608781039714813, "learning_rate": 6.436942204209917e-06, "loss": 0.0153, "step": 67590 }, { "grad_norm": 0.10440567135810852, "learning_rate": 6.4268014814408804e-06, "loss": 0.0185, "step": 67600 }, { "grad_norm": 0.35008302330970764, "learning_rate": 6.41666820413977e-06, "loss": 0.0129, "step": 67610 }, { "grad_norm": 0.15249353647232056, "learning_rate": 6.406542374038033e-06, "loss": 0.0177, "step": 67620 }, { "grad_norm": 0.17695216834545135, "learning_rate": 6.396423992865935e-06, "loss": 0.0114, "step": 67630 }, { "grad_norm": 0.13977161049842834, "learning_rate": 6.386313062352412e-06, "loss": 0.015, "step": 67640 }, { "grad_norm": 0.20580406486988068, "learning_rate": 6.376209584225152e-06, "loss": 0.0118, "step": 67650 }, { "grad_norm": 0.1831127107143402, "learning_rate": 6.366113560210557e-06, "loss": 0.0121, "step": 67660 }, { "grad_norm": 0.228846475481987, "learning_rate": 6.356024992033766e-06, "loss": 0.0165, "step": 67670 }, { "grad_norm": 0.15578614175319672, "learning_rate": 6.345943881418648e-06, "loss": 0.0125, "step": 67680 }, { "grad_norm": 0.18909496068954468, "learning_rate": 6.33587023008777e-06, "loss": 0.0128, "step": 67690 }, { "grad_norm": 0.16189226508140564, "learning_rate": 6.32580403976245e-06, "loss": 0.0128, "step": 67700 }, { "grad_norm": 0.20708070695400238, "learning_rate": 6.31574531216273e-06, "loss": 0.0143, "step": 67710 }, { "grad_norm": 0.25312715768814087, "learning_rate": 6.305694049007371e-06, "loss": 0.0179, "step": 67720 }, { "grad_norm": 0.19302719831466675, "learning_rate": 6.2956502520138575e-06, "loss": 0.0148, "step": 67730 }, { "grad_norm": 0.163643017411232, "learning_rate": 6.285613922898409e-06, "loss": 0.0129, "step": 67740 }, { "grad_norm": 0.19018638134002686, "learning_rate": 6.275585063375927e-06, "loss": 0.0175, "step": 67750 }, { "grad_norm": 0.21483063697814941, "learning_rate": 6.265563675160113e-06, "loss": 0.0138, "step": 67760 }, { "grad_norm": 0.08405397087335587, "learning_rate": 6.25554975996332e-06, "loss": 0.014, "step": 67770 }, { "grad_norm": 0.15927937626838684, "learning_rate": 6.245543319496661e-06, "loss": 0.0119, "step": 67780 }, { "grad_norm": 0.22037753462791443, "learning_rate": 6.2355443554699685e-06, "loss": 0.0156, "step": 67790 }, { "grad_norm": 0.12080900371074677, "learning_rate": 6.225552869591766e-06, "loss": 0.012, "step": 67800 }, { "grad_norm": 0.2554139792919159, "learning_rate": 6.215568863569365e-06, "loss": 0.0155, "step": 67810 }, { "grad_norm": 0.22327277064323425, "learning_rate": 6.205592339108712e-06, "loss": 0.0189, "step": 67820 }, { "grad_norm": 0.2066807895898819, "learning_rate": 6.195623297914577e-06, "loss": 0.0144, "step": 67830 }, { "grad_norm": 0.23024074733257294, "learning_rate": 6.185661741690357e-06, "loss": 0.0144, "step": 67840 }, { "grad_norm": 0.18280048668384552, "learning_rate": 6.1757076721382145e-06, "loss": 0.018, "step": 67850 }, { "grad_norm": 0.21308356523513794, "learning_rate": 6.165761090959038e-06, "loss": 0.0169, "step": 67860 }, { "grad_norm": 0.2705490291118622, "learning_rate": 6.155821999852424e-06, "loss": 0.0162, "step": 67870 }, { "grad_norm": 0.12497026473283768, "learning_rate": 6.145890400516696e-06, "loss": 0.0133, "step": 67880 }, { "grad_norm": 0.14411574602127075, "learning_rate": 6.1359662946488816e-06, "loss": 0.0139, "step": 67890 }, { "grad_norm": 0.2320912927389145, "learning_rate": 6.12604968394474e-06, "loss": 0.0122, "step": 67900 }, { "grad_norm": 0.17562420666217804, "learning_rate": 6.11614057009875e-06, "loss": 0.0158, "step": 67910 }, { "grad_norm": 0.1792379766702652, "learning_rate": 6.106238954804111e-06, "loss": 0.0156, "step": 67920 }, { "grad_norm": 0.2326149344444275, "learning_rate": 6.096344839752738e-06, "loss": 0.0151, "step": 67930 }, { "grad_norm": 0.2111196517944336, "learning_rate": 6.086458226635278e-06, "loss": 0.0177, "step": 67940 }, { "grad_norm": 0.185400128364563, "learning_rate": 6.076579117141046e-06, "loss": 0.0231, "step": 67950 }, { "grad_norm": 0.17969746887683868, "learning_rate": 6.066707512958153e-06, "loss": 0.0152, "step": 67960 }, { "grad_norm": 0.2090587317943573, "learning_rate": 6.056843415773361e-06, "loss": 0.0153, "step": 67970 }, { "grad_norm": 0.1843237429857254, "learning_rate": 6.0469868272721776e-06, "loss": 0.0154, "step": 67980 }, { "grad_norm": 0.16543130576610565, "learning_rate": 6.037137749138844e-06, "loss": 0.0185, "step": 67990 }, { "grad_norm": 0.2501780688762665, "learning_rate": 6.027296183056252e-06, "loss": 0.0202, "step": 68000 }, { "grad_norm": 0.2147654891014099, "learning_rate": 6.017462130706114e-06, "loss": 0.0139, "step": 68010 }, { "grad_norm": 0.21229736506938934, "learning_rate": 6.0076355937687516e-06, "loss": 0.0181, "step": 68020 }, { "grad_norm": 0.2582361400127411, "learning_rate": 5.9978165739232925e-06, "loss": 0.0128, "step": 68030 }, { "grad_norm": 0.17298762500286102, "learning_rate": 5.988005072847508e-06, "loss": 0.0157, "step": 68040 }, { "grad_norm": 0.19592903554439545, "learning_rate": 5.978201092217928e-06, "loss": 0.0239, "step": 68050 }, { "grad_norm": 0.16250118613243103, "learning_rate": 5.9684046337097895e-06, "loss": 0.0118, "step": 68060 }, { "grad_norm": 0.15096645057201385, "learning_rate": 5.958615698997016e-06, "loss": 0.0155, "step": 68070 }, { "grad_norm": 0.17005354166030884, "learning_rate": 5.948834289752303e-06, "loss": 0.014, "step": 68080 }, { "grad_norm": 0.21348167955875397, "learning_rate": 5.939060407646996e-06, "loss": 0.0188, "step": 68090 }, { "grad_norm": 0.13865166902542114, "learning_rate": 5.929294054351198e-06, "loss": 0.0162, "step": 68100 }, { "grad_norm": 0.16391004621982574, "learning_rate": 5.919535231533707e-06, "loss": 0.012, "step": 68110 }, { "grad_norm": 0.25859612226486206, "learning_rate": 5.909783940862046e-06, "loss": 0.0155, "step": 68120 }, { "grad_norm": 0.17862732708454132, "learning_rate": 5.900040184002436e-06, "loss": 0.0144, "step": 68130 }, { "grad_norm": 0.21627956628799438, "learning_rate": 5.890303962619831e-06, "loss": 0.0147, "step": 68140 }, { "grad_norm": 0.23331551253795624, "learning_rate": 5.88057527837787e-06, "loss": 0.0147, "step": 68150 }, { "grad_norm": 0.17608554661273956, "learning_rate": 5.870854132938924e-06, "loss": 0.0144, "step": 68160 }, { "grad_norm": 0.16887161135673523, "learning_rate": 5.861140527964071e-06, "loss": 0.0178, "step": 68170 }, { "grad_norm": 0.17524857819080353, "learning_rate": 5.851434465113098e-06, "loss": 0.011, "step": 68180 }, { "grad_norm": 0.18556471168994904, "learning_rate": 5.841735946044524e-06, "loss": 0.0164, "step": 68190 }, { "grad_norm": 0.18196283280849457, "learning_rate": 5.832044972415523e-06, "loss": 0.0146, "step": 68200 }, { "grad_norm": 0.16740913689136505, "learning_rate": 5.8223615458820655e-06, "loss": 0.0121, "step": 68210 }, { "grad_norm": 0.12298661470413208, "learning_rate": 5.812685668098733e-06, "loss": 0.014, "step": 68220 }, { "grad_norm": 0.1491413414478302, "learning_rate": 5.803017340718913e-06, "loss": 0.0158, "step": 68230 }, { "grad_norm": 0.2212340533733368, "learning_rate": 5.793356565394636e-06, "loss": 0.0137, "step": 68240 }, { "grad_norm": 0.2172156423330307, "learning_rate": 5.783703343776658e-06, "loss": 0.0129, "step": 68250 }, { "grad_norm": 0.11198902875185013, "learning_rate": 5.774057677514477e-06, "loss": 0.0119, "step": 68260 }, { "grad_norm": 0.11850633472204208, "learning_rate": 5.764419568256235e-06, "loss": 0.015, "step": 68270 }, { "grad_norm": 0.21147136390209198, "learning_rate": 5.754789017648865e-06, "loss": 0.0266, "step": 68280 }, { "grad_norm": 0.1947086900472641, "learning_rate": 5.745166027337934e-06, "loss": 0.0181, "step": 68290 }, { "grad_norm": 0.21214619278907776, "learning_rate": 5.735550598967754e-06, "loss": 0.0159, "step": 68300 }, { "grad_norm": 0.1575518548488617, "learning_rate": 5.725942734181339e-06, "loss": 0.0142, "step": 68310 }, { "grad_norm": 0.276309609413147, "learning_rate": 5.7163424346204146e-06, "loss": 0.0147, "step": 68320 }, { "grad_norm": 0.1671127825975418, "learning_rate": 5.706749701925407e-06, "loss": 0.0141, "step": 68330 }, { "grad_norm": 0.15082356333732605, "learning_rate": 5.697164537735461e-06, "loss": 0.014, "step": 68340 }, { "grad_norm": 0.19512587785720825, "learning_rate": 5.687586943688406e-06, "loss": 0.0116, "step": 68350 }, { "grad_norm": 0.1926662027835846, "learning_rate": 5.678016921420792e-06, "loss": 0.0161, "step": 68360 }, { "grad_norm": 0.17920389771461487, "learning_rate": 5.668454472567875e-06, "loss": 0.0219, "step": 68370 }, { "grad_norm": 0.2844723165035248, "learning_rate": 5.658899598763617e-06, "loss": 0.0179, "step": 68380 }, { "grad_norm": 0.18204499781131744, "learning_rate": 5.6493523016407035e-06, "loss": 0.0132, "step": 68390 }, { "grad_norm": 0.1555922031402588, "learning_rate": 5.6398125828304615e-06, "loss": 0.0169, "step": 68400 }, { "grad_norm": 0.18603013455867767, "learning_rate": 5.630280443963015e-06, "loss": 0.015, "step": 68410 }, { "grad_norm": 0.19309429824352264, "learning_rate": 5.620755886667112e-06, "loss": 0.0159, "step": 68420 }, { "grad_norm": 0.22053541243076324, "learning_rate": 5.611238912570271e-06, "loss": 0.0141, "step": 68430 }, { "grad_norm": 0.24265889823436737, "learning_rate": 5.601729523298649e-06, "loss": 0.0196, "step": 68440 }, { "grad_norm": 0.19929251074790955, "learning_rate": 5.592227720477161e-06, "loss": 0.017, "step": 68450 }, { "grad_norm": 0.20975814759731293, "learning_rate": 5.582733505729415e-06, "loss": 0.0188, "step": 68460 }, { "grad_norm": 0.21779954433441162, "learning_rate": 5.573246880677668e-06, "loss": 0.0162, "step": 68470 }, { "grad_norm": 0.22799348831176758, "learning_rate": 5.56376784694298e-06, "loss": 0.0167, "step": 68480 }, { "grad_norm": 0.1905393749475479, "learning_rate": 5.554296406145027e-06, "loss": 0.0184, "step": 68490 }, { "grad_norm": 0.19423195719718933, "learning_rate": 5.544832559902219e-06, "loss": 0.0149, "step": 68500 }, { "grad_norm": 0.2674110233783722, "learning_rate": 5.53537630983168e-06, "loss": 0.0195, "step": 68510 }, { "grad_norm": 0.15339910984039307, "learning_rate": 5.5259276575492125e-06, "loss": 0.0192, "step": 68520 }, { "grad_norm": 0.21240858733654022, "learning_rate": 5.516486604669357e-06, "loss": 0.0212, "step": 68530 }, { "grad_norm": 0.1513875424861908, "learning_rate": 5.507053152805303e-06, "loss": 0.0187, "step": 68540 }, { "grad_norm": 0.1596221774816513, "learning_rate": 5.497627303568975e-06, "loss": 0.0221, "step": 68550 }, { "grad_norm": 0.13232333958148956, "learning_rate": 5.488209058571003e-06, "loss": 0.0113, "step": 68560 }, { "grad_norm": 0.14711394906044006, "learning_rate": 5.4787984194207054e-06, "loss": 0.0139, "step": 68570 }, { "grad_norm": 0.1958189159631729, "learning_rate": 5.469395387726095e-06, "loss": 0.0128, "step": 68580 }, { "grad_norm": 0.20165535807609558, "learning_rate": 5.45999996509391e-06, "loss": 0.0152, "step": 68590 }, { "grad_norm": 0.13474565744400024, "learning_rate": 5.450612153129536e-06, "loss": 0.0144, "step": 68600 }, { "grad_norm": 0.11631802469491959, "learning_rate": 5.4412319534371426e-06, "loss": 0.0174, "step": 68610 }, { "grad_norm": 0.21966198086738586, "learning_rate": 5.431859367619513e-06, "loss": 0.0145, "step": 68620 }, { "grad_norm": 0.2419716715812683, "learning_rate": 5.422494397278172e-06, "loss": 0.0158, "step": 68630 }, { "grad_norm": 0.1942725032567978, "learning_rate": 5.413137044013344e-06, "loss": 0.0184, "step": 68640 }, { "grad_norm": 0.1628865897655487, "learning_rate": 5.403787309423941e-06, "loss": 0.0157, "step": 68650 }, { "grad_norm": 0.13809125125408173, "learning_rate": 5.3944451951075835e-06, "loss": 0.0118, "step": 68660 }, { "grad_norm": 0.15822052955627441, "learning_rate": 5.385110702660562e-06, "loss": 0.0158, "step": 68670 }, { "grad_norm": 0.1343589574098587, "learning_rate": 5.375783833677922e-06, "loss": 0.0164, "step": 68680 }, { "grad_norm": 0.17359226942062378, "learning_rate": 5.36646458975334e-06, "loss": 0.0135, "step": 68690 }, { "grad_norm": 0.1899709403514862, "learning_rate": 5.3571529724792294e-06, "loss": 0.0128, "step": 68700 }, { "grad_norm": 0.22925014793872833, "learning_rate": 5.347848983446702e-06, "loss": 0.0172, "step": 68710 }, { "grad_norm": 0.21922631561756134, "learning_rate": 5.3385526242455185e-06, "loss": 0.017, "step": 68720 }, { "grad_norm": 0.12276845425367355, "learning_rate": 5.329263896464226e-06, "loss": 0.0157, "step": 68730 }, { "grad_norm": 0.13592955470085144, "learning_rate": 5.3199828016899715e-06, "loss": 0.0156, "step": 68740 }, { "grad_norm": 0.17928913235664368, "learning_rate": 5.310709341508657e-06, "loss": 0.0151, "step": 68750 }, { "grad_norm": 0.23928415775299072, "learning_rate": 5.301443517504861e-06, "loss": 0.0142, "step": 68760 }, { "grad_norm": 0.250709593296051, "learning_rate": 5.292185331261862e-06, "loss": 0.0121, "step": 68770 }, { "grad_norm": 0.20804661512374878, "learning_rate": 5.28293478436162e-06, "loss": 0.0142, "step": 68780 }, { "grad_norm": 0.32190558314323425, "learning_rate": 5.273691878384829e-06, "loss": 0.0175, "step": 68790 }, { "grad_norm": 0.18655380606651306, "learning_rate": 5.264456614910812e-06, "loss": 0.0152, "step": 68800 }, { "grad_norm": 0.20539414882659912, "learning_rate": 5.255228995517647e-06, "loss": 0.027, "step": 68810 }, { "grad_norm": 0.27975502610206604, "learning_rate": 5.246009021782067e-06, "loss": 0.0147, "step": 68820 }, { "grad_norm": 0.23537670075893402, "learning_rate": 5.2367966952795225e-06, "loss": 0.0208, "step": 68830 }, { "grad_norm": 0.1934446543455124, "learning_rate": 5.2275920175841485e-06, "loss": 0.0126, "step": 68840 }, { "grad_norm": 0.16674473881721497, "learning_rate": 5.218394990268771e-06, "loss": 0.0145, "step": 68850 }, { "grad_norm": 0.24803480505943298, "learning_rate": 5.209205614904916e-06, "loss": 0.0152, "step": 68860 }, { "grad_norm": 0.24711313843727112, "learning_rate": 5.200023893062777e-06, "loss": 0.015, "step": 68870 }, { "grad_norm": 0.14233459532260895, "learning_rate": 5.190849826311289e-06, "loss": 0.0107, "step": 68880 }, { "grad_norm": 0.2273312360048294, "learning_rate": 5.181683416218025e-06, "loss": 0.0205, "step": 68890 }, { "grad_norm": 0.16244377195835114, "learning_rate": 5.172524664349276e-06, "loss": 0.0121, "step": 68900 }, { "grad_norm": 0.13752597570419312, "learning_rate": 5.1633735722700416e-06, "loss": 0.0133, "step": 68910 }, { "grad_norm": 0.1383407562971115, "learning_rate": 5.154230141543958e-06, "loss": 0.013, "step": 68920 }, { "grad_norm": 0.21078947186470032, "learning_rate": 5.145094373733433e-06, "loss": 0.0148, "step": 68930 }, { "grad_norm": 0.22264130413532257, "learning_rate": 5.135966270399478e-06, "loss": 0.0104, "step": 68940 }, { "grad_norm": 0.24835146963596344, "learning_rate": 5.126845833101857e-06, "loss": 0.0181, "step": 68950 }, { "grad_norm": 0.17318716645240784, "learning_rate": 5.1177330633989994e-06, "loss": 0.0125, "step": 68960 }, { "grad_norm": 0.2499004751443863, "learning_rate": 5.108627962848033e-06, "loss": 0.0252, "step": 68970 }, { "grad_norm": 0.14786747097969055, "learning_rate": 5.099530533004759e-06, "loss": 0.0164, "step": 68980 }, { "grad_norm": 0.1954839676618576, "learning_rate": 5.090440775423699e-06, "loss": 0.0118, "step": 68990 }, { "grad_norm": 0.18465331196784973, "learning_rate": 5.081358691658022e-06, "loss": 0.0134, "step": 69000 }, { "grad_norm": 0.24483925104141235, "learning_rate": 5.072284283259621e-06, "loss": 0.0159, "step": 69010 }, { "grad_norm": 0.18563084304332733, "learning_rate": 5.063217551779053e-06, "loss": 0.0157, "step": 69020 }, { "grad_norm": 0.1903349608182907, "learning_rate": 5.054158498765588e-06, "loss": 0.0119, "step": 69030 }, { "grad_norm": 0.33056938648223877, "learning_rate": 5.0451071257671625e-06, "loss": 0.0167, "step": 69040 }, { "grad_norm": 0.22192105650901794, "learning_rate": 5.036063434330407e-06, "loss": 0.0165, "step": 69050 }, { "grad_norm": 0.23838233947753906, "learning_rate": 5.027027426000652e-06, "loss": 0.0127, "step": 69060 }, { "grad_norm": 0.12011348456144333, "learning_rate": 5.017999102321886e-06, "loss": 0.0123, "step": 69070 }, { "grad_norm": 0.28102004528045654, "learning_rate": 5.0089784648368224e-06, "loss": 0.0159, "step": 69080 }, { "grad_norm": 0.2017531543970108, "learning_rate": 4.999965515086829e-06, "loss": 0.0126, "step": 69090 }, { "grad_norm": 0.12036529928445816, "learning_rate": 4.990960254611976e-06, "loss": 0.0126, "step": 69100 }, { "grad_norm": 0.27311229705810547, "learning_rate": 4.9819626849510194e-06, "loss": 0.0154, "step": 69110 }, { "grad_norm": 0.18756696581840515, "learning_rate": 4.97297280764138e-06, "loss": 0.0225, "step": 69120 }, { "grad_norm": 0.23420901596546173, "learning_rate": 4.963990624219211e-06, "loss": 0.0149, "step": 69130 }, { "grad_norm": 0.15537536144256592, "learning_rate": 4.955016136219298e-06, "loss": 0.014, "step": 69140 }, { "grad_norm": 0.372511088848114, "learning_rate": 4.946049345175146e-06, "loss": 0.0115, "step": 69150 }, { "grad_norm": 0.24009740352630615, "learning_rate": 4.937090252618937e-06, "loss": 0.0169, "step": 69160 }, { "grad_norm": 0.10764310508966446, "learning_rate": 4.928138860081521e-06, "loss": 0.0189, "step": 69170 }, { "grad_norm": 0.17209595441818237, "learning_rate": 4.919195169092472e-06, "loss": 0.0164, "step": 69180 }, { "grad_norm": 0.1899401843547821, "learning_rate": 4.910259181179994e-06, "loss": 0.0156, "step": 69190 }, { "grad_norm": 0.21612389385700226, "learning_rate": 4.901330897871015e-06, "loss": 0.0169, "step": 69200 }, { "grad_norm": 0.1838444322347641, "learning_rate": 4.8924103206911375e-06, "loss": 0.0108, "step": 69210 }, { "grad_norm": 0.2764389216899872, "learning_rate": 4.883497451164637e-06, "loss": 0.0193, "step": 69220 }, { "grad_norm": 0.17045055329799652, "learning_rate": 4.874592290814484e-06, "loss": 0.0117, "step": 69230 }, { "grad_norm": 0.3888951241970062, "learning_rate": 4.865694841162327e-06, "loss": 0.0219, "step": 69240 }, { "grad_norm": 0.18344466388225555, "learning_rate": 4.856805103728496e-06, "loss": 0.0153, "step": 69250 }, { "grad_norm": 0.2750413119792938, "learning_rate": 4.847923080032007e-06, "loss": 0.0126, "step": 69260 }, { "grad_norm": 0.14528168737888336, "learning_rate": 4.8390487715905486e-06, "loss": 0.0112, "step": 69270 }, { "grad_norm": 0.12825973331928253, "learning_rate": 4.8301821799205e-06, "loss": 0.0135, "step": 69280 }, { "grad_norm": 0.14174695312976837, "learning_rate": 4.821323306536918e-06, "loss": 0.0122, "step": 69290 }, { "grad_norm": 0.1730862408876419, "learning_rate": 4.8124721529535455e-06, "loss": 0.0147, "step": 69300 }, { "grad_norm": 0.18652471899986267, "learning_rate": 4.803628720682807e-06, "loss": 0.0197, "step": 69310 }, { "grad_norm": 0.19206099212169647, "learning_rate": 4.794793011235776e-06, "loss": 0.0158, "step": 69320 }, { "grad_norm": 0.1061013862490654, "learning_rate": 4.78596502612228e-06, "loss": 0.0188, "step": 69330 }, { "grad_norm": 0.1690930873155594, "learning_rate": 4.777144766850738e-06, "loss": 0.0163, "step": 69340 }, { "grad_norm": 0.2776566445827484, "learning_rate": 4.768332234928313e-06, "loss": 0.0156, "step": 69350 }, { "grad_norm": 0.15552206337451935, "learning_rate": 4.759527431860828e-06, "loss": 0.0139, "step": 69360 }, { "grad_norm": 0.18966428935527802, "learning_rate": 4.750730359152755e-06, "loss": 0.0167, "step": 69370 }, { "grad_norm": 0.33100637793540955, "learning_rate": 4.741941018307311e-06, "loss": 0.0135, "step": 69380 }, { "grad_norm": 0.23563367128372192, "learning_rate": 4.733159410826321e-06, "loss": 0.0186, "step": 69390 }, { "grad_norm": 0.15566040575504303, "learning_rate": 4.724385538210357e-06, "loss": 0.0135, "step": 69400 }, { "grad_norm": 0.16858616471290588, "learning_rate": 4.715619401958599e-06, "loss": 0.0137, "step": 69410 }, { "grad_norm": 0.27935993671417236, "learning_rate": 4.706861003568958e-06, "loss": 0.0178, "step": 69420 }, { "grad_norm": 0.16024436056613922, "learning_rate": 4.698110344538003e-06, "loss": 0.0153, "step": 69430 }, { "grad_norm": 0.18361777067184448, "learning_rate": 4.689367426360975e-06, "loss": 0.0127, "step": 69440 }, { "grad_norm": 0.22270992398262024, "learning_rate": 4.680632250531819e-06, "loss": 0.014, "step": 69450 }, { "grad_norm": 0.20387543737888336, "learning_rate": 4.671904818543115e-06, "loss": 0.014, "step": 69460 }, { "grad_norm": 0.24655558168888092, "learning_rate": 4.66318513188615e-06, "loss": 0.0143, "step": 69470 }, { "grad_norm": 0.17382115125656128, "learning_rate": 4.654473192050884e-06, "loss": 0.0129, "step": 69480 }, { "grad_norm": 0.18912780284881592, "learning_rate": 4.64576900052594e-06, "loss": 0.0156, "step": 69490 }, { "grad_norm": 0.26334497332572937, "learning_rate": 4.637072558798638e-06, "loss": 0.0143, "step": 69500 }, { "grad_norm": 0.22533561289310455, "learning_rate": 4.628383868354969e-06, "loss": 0.0198, "step": 69510 }, { "grad_norm": 0.2240401655435562, "learning_rate": 4.6197029306795595e-06, "loss": 0.0118, "step": 69520 }, { "grad_norm": 0.3597559928894043, "learning_rate": 4.611029747255779e-06, "loss": 0.0124, "step": 69530 }, { "grad_norm": 0.213384211063385, "learning_rate": 4.6023643195656164e-06, "loss": 0.0168, "step": 69540 }, { "grad_norm": 0.15240813791751862, "learning_rate": 4.593706649089768e-06, "loss": 0.0137, "step": 69550 }, { "grad_norm": 0.20220792293548584, "learning_rate": 4.585056737307597e-06, "loss": 0.0138, "step": 69560 }, { "grad_norm": 0.16154153645038605, "learning_rate": 4.576414585697103e-06, "loss": 0.0138, "step": 69570 }, { "grad_norm": 0.2519243061542511, "learning_rate": 4.567780195735044e-06, "loss": 0.0136, "step": 69580 }, { "grad_norm": 0.1341867446899414, "learning_rate": 4.559153568896757e-06, "loss": 0.0129, "step": 69590 }, { "grad_norm": 0.2661396861076355, "learning_rate": 4.550534706656329e-06, "loss": 0.0137, "step": 69600 }, { "grad_norm": 0.15466362237930298, "learning_rate": 4.541923610486465e-06, "loss": 0.0162, "step": 69610 }, { "grad_norm": 0.23844388127326965, "learning_rate": 4.533320281858578e-06, "loss": 0.0131, "step": 69620 }, { "grad_norm": 0.17464683949947357, "learning_rate": 4.52472472224274e-06, "loss": 0.0172, "step": 69630 }, { "grad_norm": 0.15210238099098206, "learning_rate": 4.51613693310769e-06, "loss": 0.015, "step": 69640 }, { "grad_norm": 0.21132706105709076, "learning_rate": 4.507556915920868e-06, "loss": 0.0144, "step": 69650 }, { "grad_norm": 0.15555346012115479, "learning_rate": 4.498984672148332e-06, "loss": 0.0116, "step": 69660 }, { "grad_norm": 0.16442468762397766, "learning_rate": 4.490420203254864e-06, "loss": 0.016, "step": 69670 }, { "grad_norm": 0.2760392725467682, "learning_rate": 4.481863510703893e-06, "loss": 0.0192, "step": 69680 }, { "grad_norm": 0.1675344556570053, "learning_rate": 4.473314595957523e-06, "loss": 0.0135, "step": 69690 }, { "grad_norm": 0.192201167345047, "learning_rate": 4.464773460476535e-06, "loss": 0.0147, "step": 69700 }, { "grad_norm": 0.2278669774532318, "learning_rate": 4.456240105720372e-06, "loss": 0.0125, "step": 69710 }, { "grad_norm": 0.12696699798107147, "learning_rate": 4.4477145331471405e-06, "loss": 0.0234, "step": 69720 }, { "grad_norm": 0.12672944366931915, "learning_rate": 4.439196744213653e-06, "loss": 0.0117, "step": 69730 }, { "grad_norm": 0.17256270349025726, "learning_rate": 4.430686740375339e-06, "loss": 0.0148, "step": 69740 }, { "grad_norm": 0.16981330513954163, "learning_rate": 4.422184523086342e-06, "loss": 0.0139, "step": 69750 }, { "grad_norm": 0.20206518471240997, "learning_rate": 4.41369009379946e-06, "loss": 0.0141, "step": 69760 }, { "grad_norm": 0.13402654230594635, "learning_rate": 4.405203453966139e-06, "loss": 0.0133, "step": 69770 }, { "grad_norm": 0.20603618025779724, "learning_rate": 4.396724605036539e-06, "loss": 0.0133, "step": 69780 }, { "grad_norm": 0.23482327163219452, "learning_rate": 4.388253548459437e-06, "loss": 0.014, "step": 69790 }, { "grad_norm": 0.25005367398262024, "learning_rate": 4.3797902856823395e-06, "loss": 0.0125, "step": 69800 }, { "grad_norm": 0.30929672718048096, "learning_rate": 4.371334818151357e-06, "loss": 0.0134, "step": 69810 }, { "grad_norm": 0.2067670077085495, "learning_rate": 4.362887147311306e-06, "loss": 0.0145, "step": 69820 }, { "grad_norm": 0.17516390979290009, "learning_rate": 4.354447274605672e-06, "loss": 0.0154, "step": 69830 }, { "grad_norm": 0.2390470802783966, "learning_rate": 4.346015201476572e-06, "loss": 0.0188, "step": 69840 }, { "grad_norm": 0.23179991543293, "learning_rate": 4.337590929364855e-06, "loss": 0.0153, "step": 69850 }, { "grad_norm": 0.18128573894500732, "learning_rate": 4.329174459709973e-06, "loss": 0.0121, "step": 69860 }, { "grad_norm": 0.23757033050060272, "learning_rate": 4.320765793950071e-06, "loss": 0.0152, "step": 69870 }, { "grad_norm": 0.3094055652618408, "learning_rate": 4.312364933521962e-06, "loss": 0.0166, "step": 69880 }, { "grad_norm": 0.16822867095470428, "learning_rate": 4.303971879861129e-06, "loss": 0.0151, "step": 69890 }, { "grad_norm": 0.17880117893218994, "learning_rate": 4.295586634401716e-06, "loss": 0.0172, "step": 69900 }, { "grad_norm": 0.16832235455513, "learning_rate": 4.287209198576536e-06, "loss": 0.0156, "step": 69910 }, { "grad_norm": 0.18892763555049896, "learning_rate": 4.278839573817045e-06, "loss": 0.0165, "step": 69920 }, { "grad_norm": 0.23172757029533386, "learning_rate": 4.270477761553399e-06, "loss": 0.0176, "step": 69930 }, { "grad_norm": 0.20460554957389832, "learning_rate": 4.262123763214393e-06, "loss": 0.0132, "step": 69940 }, { "grad_norm": 0.2232380509376526, "learning_rate": 4.253777580227508e-06, "loss": 0.0177, "step": 69950 }, { "grad_norm": 0.1887289583683014, "learning_rate": 4.2454392140188755e-06, "loss": 0.0145, "step": 69960 }, { "grad_norm": 0.13380181789398193, "learning_rate": 4.2371086660132785e-06, "loss": 0.017, "step": 69970 }, { "grad_norm": 0.1966676563024521, "learning_rate": 4.228785937634205e-06, "loss": 0.0135, "step": 69980 }, { "grad_norm": 0.20285294950008392, "learning_rate": 4.220471030303758e-06, "loss": 0.0118, "step": 69990 }, { "grad_norm": 0.13900548219680786, "learning_rate": 4.212163945442754e-06, "loss": 0.0111, "step": 70000 }, { "grad_norm": 0.22246071696281433, "learning_rate": 4.203864684470621e-06, "loss": 0.0131, "step": 70010 }, { "grad_norm": 0.2091236114501953, "learning_rate": 4.19557324880549e-06, "loss": 0.0148, "step": 70020 }, { "grad_norm": 0.16672788560390472, "learning_rate": 4.187289639864145e-06, "loss": 0.0136, "step": 70030 }, { "grad_norm": 0.13366073369979858, "learning_rate": 4.1790138590619974e-06, "loss": 0.0113, "step": 70040 }, { "grad_norm": 0.22555424273014069, "learning_rate": 4.170745907813195e-06, "loss": 0.0166, "step": 70050 }, { "grad_norm": 0.09773743152618408, "learning_rate": 4.162485787530479e-06, "loss": 0.0134, "step": 70060 }, { "grad_norm": 0.26387086510658264, "learning_rate": 4.154233499625282e-06, "loss": 0.0129, "step": 70070 }, { "grad_norm": 0.209268257021904, "learning_rate": 4.145989045507692e-06, "loss": 0.0143, "step": 70080 }, { "grad_norm": 0.16303743422031403, "learning_rate": 4.1377524265864666e-06, "loss": 0.0175, "step": 70090 }, { "grad_norm": 0.14098148047924042, "learning_rate": 4.1295236442690175e-06, "loss": 0.0129, "step": 70100 }, { "grad_norm": 0.2671727240085602, "learning_rate": 4.121302699961421e-06, "loss": 0.0136, "step": 70110 }, { "grad_norm": 0.29389965534210205, "learning_rate": 4.113089595068403e-06, "loss": 0.0214, "step": 70120 }, { "grad_norm": 0.2106650471687317, "learning_rate": 4.104884330993364e-06, "loss": 0.017, "step": 70130 }, { "grad_norm": 0.16945227980613708, "learning_rate": 4.0966869091383585e-06, "loss": 0.0221, "step": 70140 }, { "grad_norm": 0.18554441630840302, "learning_rate": 4.088497330904101e-06, "loss": 0.0165, "step": 70150 }, { "grad_norm": 0.1624065339565277, "learning_rate": 4.080315597689976e-06, "loss": 0.0141, "step": 70160 }, { "grad_norm": 0.13239434361457825, "learning_rate": 4.072141710893995e-06, "loss": 0.0189, "step": 70170 }, { "grad_norm": 0.2130633443593979, "learning_rate": 4.063975671912879e-06, "loss": 0.0245, "step": 70180 }, { "grad_norm": 0.16576282680034637, "learning_rate": 4.055817482141949e-06, "loss": 0.0121, "step": 70190 }, { "grad_norm": 0.24593158066272736, "learning_rate": 4.047667142975259e-06, "loss": 0.0143, "step": 70200 }, { "grad_norm": 0.21720223128795624, "learning_rate": 4.039524655805443e-06, "loss": 0.0169, "step": 70210 }, { "grad_norm": 0.23624086380004883, "learning_rate": 4.03139002202384e-06, "loss": 0.0125, "step": 70220 }, { "grad_norm": 0.18734104931354523, "learning_rate": 4.023263243020447e-06, "loss": 0.0189, "step": 70230 }, { "grad_norm": 0.15322807431221008, "learning_rate": 4.015144320183884e-06, "loss": 0.0189, "step": 70240 }, { "grad_norm": 0.2057005763053894, "learning_rate": 4.007033254901482e-06, "loss": 0.014, "step": 70250 }, { "grad_norm": 0.33696067333221436, "learning_rate": 3.9989300485591795e-06, "loss": 0.0137, "step": 70260 }, { "grad_norm": 0.18953107297420502, "learning_rate": 3.990834702541601e-06, "loss": 0.0185, "step": 70270 }, { "grad_norm": 0.15627950429916382, "learning_rate": 3.982747218232019e-06, "loss": 0.0115, "step": 70280 }, { "grad_norm": 0.1920037567615509, "learning_rate": 3.97466759701236e-06, "loss": 0.0174, "step": 70290 }, { "grad_norm": 0.22667157649993896, "learning_rate": 3.966595840263226e-06, "loss": 0.0156, "step": 70300 }, { "grad_norm": 0.2262348085641861, "learning_rate": 3.958531949363836e-06, "loss": 0.0163, "step": 70310 }, { "grad_norm": 0.1554517298936844, "learning_rate": 3.950475925692098e-06, "loss": 0.0194, "step": 70320 }, { "grad_norm": 0.23281359672546387, "learning_rate": 3.9424277706245685e-06, "loss": 0.0153, "step": 70330 }, { "grad_norm": 0.12330874800682068, "learning_rate": 3.934387485536451e-06, "loss": 0.0139, "step": 70340 }, { "grad_norm": 0.17378495633602142, "learning_rate": 3.926355071801619e-06, "loss": 0.0125, "step": 70350 }, { "grad_norm": 0.1699250489473343, "learning_rate": 3.9183305307925965e-06, "loss": 0.0134, "step": 70360 }, { "grad_norm": 0.1887662261724472, "learning_rate": 3.910313863880533e-06, "loss": 0.0124, "step": 70370 }, { "grad_norm": 0.229388028383255, "learning_rate": 3.902305072435292e-06, "loss": 0.0115, "step": 70380 }, { "grad_norm": 0.27522799372673035, "learning_rate": 3.894304157825329e-06, "loss": 0.015, "step": 70390 }, { "grad_norm": 0.1988174170255661, "learning_rate": 3.886311121417791e-06, "loss": 0.0113, "step": 70400 }, { "grad_norm": 0.13845616579055786, "learning_rate": 3.878325964578472e-06, "loss": 0.0128, "step": 70410 }, { "grad_norm": 0.13814957439899445, "learning_rate": 3.870348688671815e-06, "loss": 0.0124, "step": 70420 }, { "grad_norm": 0.14745871722698212, "learning_rate": 3.862379295060931e-06, "loss": 0.0109, "step": 70430 }, { "grad_norm": 0.2265882045030594, "learning_rate": 3.854417785107534e-06, "loss": 0.0199, "step": 70440 }, { "grad_norm": 0.20237275958061218, "learning_rate": 3.8464641601720755e-06, "loss": 0.0147, "step": 70450 }, { "grad_norm": 0.19794867932796478, "learning_rate": 3.838518421613579e-06, "loss": 0.0152, "step": 70460 }, { "grad_norm": 0.23370930552482605, "learning_rate": 3.830580570789766e-06, "loss": 0.0174, "step": 70470 }, { "grad_norm": 0.1979142725467682, "learning_rate": 3.822650609057005e-06, "loss": 0.0132, "step": 70480 }, { "grad_norm": 0.19705668091773987, "learning_rate": 3.814728537770285e-06, "loss": 0.0148, "step": 70490 }, { "grad_norm": 0.20497700572013855, "learning_rate": 3.806814358283306e-06, "loss": 0.018, "step": 70500 }, { "grad_norm": 0.2381948083639145, "learning_rate": 3.7989080719483596e-06, "loss": 0.0117, "step": 70510 }, { "grad_norm": 0.16610530018806458, "learning_rate": 3.7910096801164143e-06, "loss": 0.016, "step": 70520 }, { "grad_norm": 0.2691457271575928, "learning_rate": 3.7831191841371016e-06, "loss": 0.0156, "step": 70530 }, { "grad_norm": 0.20788581669330597, "learning_rate": 3.775236585358688e-06, "loss": 0.0159, "step": 70540 }, { "grad_norm": 0.22197580337524414, "learning_rate": 3.7673618851280843e-06, "loss": 0.0166, "step": 70550 }, { "grad_norm": 0.2369464486837387, "learning_rate": 3.759495084790887e-06, "loss": 0.0155, "step": 70560 }, { "grad_norm": 0.2638915181159973, "learning_rate": 3.751636185691282e-06, "loss": 0.0131, "step": 70570 }, { "grad_norm": 0.13008631765842438, "learning_rate": 3.7437851891721607e-06, "loss": 0.0143, "step": 70580 }, { "grad_norm": 0.12012200057506561, "learning_rate": 3.7359420965750404e-06, "loss": 0.0127, "step": 70590 }, { "grad_norm": 0.18002167344093323, "learning_rate": 3.7281069092400922e-06, "loss": 0.0128, "step": 70600 }, { "grad_norm": 0.1984594613313675, "learning_rate": 3.7202796285061348e-06, "loss": 0.0174, "step": 70610 }, { "grad_norm": 0.17252032458782196, "learning_rate": 3.712460255710637e-06, "loss": 0.0182, "step": 70620 }, { "grad_norm": 0.23106169700622559, "learning_rate": 3.704648792189719e-06, "loss": 0.0159, "step": 70630 }, { "grad_norm": 0.22097457945346832, "learning_rate": 3.696845239278124e-06, "loss": 0.0143, "step": 70640 }, { "grad_norm": 0.16184718906879425, "learning_rate": 3.689049598309302e-06, "loss": 0.0165, "step": 70650 }, { "grad_norm": 0.27387481927871704, "learning_rate": 3.681261870615288e-06, "loss": 0.0261, "step": 70660 }, { "grad_norm": 0.21117587387561798, "learning_rate": 3.6734820575268004e-06, "loss": 0.02, "step": 70670 }, { "grad_norm": 0.14151836931705475, "learning_rate": 3.665710160373209e-06, "loss": 0.013, "step": 70680 }, { "grad_norm": 0.14435535669326782, "learning_rate": 3.65794618048248e-06, "loss": 0.0132, "step": 70690 }, { "grad_norm": 0.29066768288612366, "learning_rate": 3.6501901191813125e-06, "loss": 0.0156, "step": 70700 }, { "grad_norm": 0.20010632276535034, "learning_rate": 3.642441977794975e-06, "loss": 0.0158, "step": 70710 }, { "grad_norm": 0.2614581286907196, "learning_rate": 3.634701757647424e-06, "loss": 0.0167, "step": 70720 }, { "grad_norm": 0.21062706410884857, "learning_rate": 3.6269694600612468e-06, "loss": 0.0121, "step": 70730 }, { "grad_norm": 0.22664031386375427, "learning_rate": 3.619245086357681e-06, "loss": 0.0156, "step": 70740 }, { "grad_norm": 0.20521284639835358, "learning_rate": 3.611528637856615e-06, "loss": 0.0153, "step": 70750 }, { "grad_norm": 0.22410349547863007, "learning_rate": 3.6038201158765884e-06, "loss": 0.0127, "step": 70760 }, { "grad_norm": 0.32335200905799866, "learning_rate": 3.5961195217347534e-06, "loss": 0.0252, "step": 70770 }, { "grad_norm": 0.24177834391593933, "learning_rate": 3.588426856746946e-06, "loss": 0.016, "step": 70780 }, { "grad_norm": 0.15618334710597992, "learning_rate": 3.5807421222276316e-06, "loss": 0.0233, "step": 70790 }, { "grad_norm": 0.3388996720314026, "learning_rate": 3.5730653194899156e-06, "loss": 0.0153, "step": 70800 }, { "grad_norm": 0.17816102504730225, "learning_rate": 3.565396449845554e-06, "loss": 0.0153, "step": 70810 }, { "grad_norm": 0.26119181513786316, "learning_rate": 3.557735514604954e-06, "loss": 0.0168, "step": 70820 }, { "grad_norm": 0.11903229355812073, "learning_rate": 3.5500825150771633e-06, "loss": 0.0138, "step": 70830 }, { "grad_norm": 0.14466337859630585, "learning_rate": 3.5424374525698466e-06, "loss": 0.0139, "step": 70840 }, { "grad_norm": 0.20372287929058075, "learning_rate": 3.5348003283893704e-06, "loss": 0.012, "step": 70850 }, { "grad_norm": 0.19334958493709564, "learning_rate": 3.527171143840685e-06, "loss": 0.0171, "step": 70860 }, { "grad_norm": 0.1764417290687561, "learning_rate": 3.519549900227409e-06, "loss": 0.0174, "step": 70870 }, { "grad_norm": 0.1981106400489807, "learning_rate": 3.511936598851828e-06, "loss": 0.0164, "step": 70880 }, { "grad_norm": 0.14487959444522858, "learning_rate": 3.504331241014813e-06, "loss": 0.0152, "step": 70890 }, { "grad_norm": 0.19585704803466797, "learning_rate": 3.4967338280159414e-06, "loss": 0.0129, "step": 70900 }, { "grad_norm": 0.13321858644485474, "learning_rate": 3.4891443611533846e-06, "loss": 0.0183, "step": 70910 }, { "grad_norm": 0.4116405248641968, "learning_rate": 3.481562841723984e-06, "loss": 0.0217, "step": 70920 }, { "grad_norm": 0.17511707544326782, "learning_rate": 3.473989271023215e-06, "loss": 0.0131, "step": 70930 }, { "grad_norm": 0.18935079872608185, "learning_rate": 3.4664236503451854e-06, "loss": 0.0155, "step": 70940 }, { "grad_norm": 0.17945630848407745, "learning_rate": 3.4588659809826674e-06, "loss": 0.0147, "step": 70950 }, { "grad_norm": 0.19072569906711578, "learning_rate": 3.451316264227039e-06, "loss": 0.0233, "step": 70960 }, { "grad_norm": 0.17557252943515778, "learning_rate": 3.4437745013683574e-06, "loss": 0.0141, "step": 70970 }, { "grad_norm": 0.24884027242660522, "learning_rate": 3.4362406936952916e-06, "loss": 0.0144, "step": 70980 }, { "grad_norm": 0.2671458125114441, "learning_rate": 3.4287148424951786e-06, "loss": 0.0143, "step": 70990 }, { "grad_norm": 0.20417793095111847, "learning_rate": 3.4211969490539617e-06, "loss": 0.0122, "step": 71000 }, { "grad_norm": 0.13838601112365723, "learning_rate": 3.4136870146562584e-06, "loss": 0.0128, "step": 71010 }, { "grad_norm": 0.1655319631099701, "learning_rate": 3.406185040585308e-06, "loss": 0.0139, "step": 71020 }, { "grad_norm": 0.25213783979415894, "learning_rate": 3.3986910281229966e-06, "loss": 0.0169, "step": 71030 }, { "grad_norm": 0.21223336458206177, "learning_rate": 3.391204978549828e-06, "loss": 0.0166, "step": 71040 }, { "grad_norm": 0.21042731404304504, "learning_rate": 3.3837268931449785e-06, "loss": 0.0198, "step": 71050 }, { "grad_norm": 0.1967599093914032, "learning_rate": 3.376256773186248e-06, "loss": 0.0193, "step": 71060 }, { "grad_norm": 0.18222004175186157, "learning_rate": 3.3687946199500664e-06, "loss": 0.0149, "step": 71070 }, { "grad_norm": 0.2149314284324646, "learning_rate": 3.3613404347115295e-06, "loss": 0.0127, "step": 71080 }, { "grad_norm": 0.15537893772125244, "learning_rate": 3.35389421874433e-06, "loss": 0.01, "step": 71090 }, { "grad_norm": 0.16181957721710205, "learning_rate": 3.3464559733208446e-06, "loss": 0.0139, "step": 71100 }, { "grad_norm": 0.25170445442199707, "learning_rate": 3.3390256997120505e-06, "loss": 0.0193, "step": 71110 }, { "grad_norm": 0.2694161534309387, "learning_rate": 3.331603399187583e-06, "loss": 0.0166, "step": 71120 }, { "grad_norm": 0.2116469293832779, "learning_rate": 3.324189073015721e-06, "loss": 0.0177, "step": 71130 }, { "grad_norm": 0.19516336917877197, "learning_rate": 3.3167827224633465e-06, "loss": 0.0143, "step": 71140 }, { "grad_norm": 0.20587287843227386, "learning_rate": 3.309384348796024e-06, "loss": 0.0116, "step": 71150 }, { "grad_norm": 0.2054608017206192, "learning_rate": 3.301993953277921e-06, "loss": 0.0143, "step": 71160 }, { "grad_norm": 0.23553888499736786, "learning_rate": 3.29461153717186e-06, "loss": 0.0123, "step": 71170 }, { "grad_norm": 0.23167546093463898, "learning_rate": 3.287237101739293e-06, "loss": 0.0198, "step": 71180 }, { "grad_norm": 0.18143106997013092, "learning_rate": 3.2798706482403075e-06, "loss": 0.0192, "step": 71190 }, { "grad_norm": 0.20097433030605316, "learning_rate": 3.2725121779336285e-06, "loss": 0.0165, "step": 71200 }, { "grad_norm": 0.23293018341064453, "learning_rate": 3.265161692076618e-06, "loss": 0.0178, "step": 71210 }, { "grad_norm": 0.19517825543880463, "learning_rate": 3.257819191925282e-06, "loss": 0.0131, "step": 71220 }, { "grad_norm": 0.2322435975074768, "learning_rate": 3.2504846787342392e-06, "loss": 0.0131, "step": 71230 }, { "grad_norm": 0.1572752743959427, "learning_rate": 3.243158153756759e-06, "loss": 0.014, "step": 71240 }, { "grad_norm": 0.2614861726760864, "learning_rate": 3.2358396182447514e-06, "loss": 0.0134, "step": 71250 }, { "grad_norm": 0.1858888864517212, "learning_rate": 3.2285290734487496e-06, "loss": 0.0135, "step": 71260 }, { "grad_norm": 0.22307029366493225, "learning_rate": 3.221226520617926e-06, "loss": 0.013, "step": 71270 }, { "grad_norm": 0.23442737758159637, "learning_rate": 3.213931961000094e-06, "loss": 0.0188, "step": 71280 }, { "grad_norm": 0.21840065717697144, "learning_rate": 3.2066453958416733e-06, "loss": 0.0135, "step": 71290 }, { "grad_norm": 0.2347346842288971, "learning_rate": 3.199366826387773e-06, "loss": 0.0161, "step": 71300 }, { "grad_norm": 0.16775599122047424, "learning_rate": 3.192096253882071e-06, "loss": 0.0161, "step": 71310 }, { "grad_norm": 0.1499798446893692, "learning_rate": 3.1848336795669177e-06, "loss": 0.0158, "step": 71320 }, { "grad_norm": 0.1812485307455063, "learning_rate": 3.1775791046833035e-06, "loss": 0.0168, "step": 71330 }, { "grad_norm": 0.10328243672847748, "learning_rate": 3.170332530470804e-06, "loss": 0.0137, "step": 71340 }, { "grad_norm": 0.19126151502132416, "learning_rate": 3.1630939581677012e-06, "loss": 0.0168, "step": 71350 }, { "grad_norm": 0.15491706132888794, "learning_rate": 3.155863389010838e-06, "loss": 0.0154, "step": 71360 }, { "grad_norm": 0.17584259808063507, "learning_rate": 3.1486408242357323e-06, "loss": 0.0188, "step": 71370 }, { "grad_norm": 0.2951336205005646, "learning_rate": 3.1414262650765248e-06, "loss": 0.032, "step": 71380 }, { "grad_norm": 0.20790080726146698, "learning_rate": 3.134219712765979e-06, "loss": 0.0159, "step": 71390 }, { "grad_norm": 0.2775455117225647, "learning_rate": 3.1270211685355044e-06, "loss": 0.0206, "step": 71400 }, { "grad_norm": 0.1835619956254959, "learning_rate": 3.1198306336151338e-06, "loss": 0.0156, "step": 71410 }, { "grad_norm": 0.15668074786663055, "learning_rate": 3.11264810923354e-06, "loss": 0.0139, "step": 71420 }, { "grad_norm": 0.16076533496379852, "learning_rate": 3.105473596618008e-06, "loss": 0.0109, "step": 71430 }, { "grad_norm": 0.2710533142089844, "learning_rate": 3.0983070969944683e-06, "loss": 0.0139, "step": 71440 }, { "grad_norm": 0.1207670345902443, "learning_rate": 3.09114861158748e-06, "loss": 0.0143, "step": 71450 }, { "grad_norm": 0.1734674721956253, "learning_rate": 3.0839981416202314e-06, "loss": 0.015, "step": 71460 }, { "grad_norm": 0.2743757963180542, "learning_rate": 3.0768556883145507e-06, "loss": 0.0136, "step": 71470 }, { "grad_norm": 0.16187426447868347, "learning_rate": 3.0697212528908834e-06, "loss": 0.0118, "step": 71480 }, { "grad_norm": 0.1508360356092453, "learning_rate": 3.0625948365683e-06, "loss": 0.0156, "step": 71490 }, { "grad_norm": 0.17330147325992584, "learning_rate": 3.055476440564531e-06, "loss": 0.0164, "step": 71500 }, { "grad_norm": 0.15719790756702423, "learning_rate": 3.0483660660958924e-06, "loss": 0.0152, "step": 71510 }, { "grad_norm": 0.2347610592842102, "learning_rate": 3.0412637143773624e-06, "loss": 0.0214, "step": 71520 }, { "grad_norm": 0.2186623066663742, "learning_rate": 3.034169386622554e-06, "loss": 0.0157, "step": 71530 }, { "grad_norm": 0.18786394596099854, "learning_rate": 3.027083084043658e-06, "loss": 0.0142, "step": 71540 }, { "grad_norm": 0.21954385936260223, "learning_rate": 3.0200048078515676e-06, "loss": 0.0141, "step": 71550 }, { "grad_norm": 0.1975552886724472, "learning_rate": 3.012934559255742e-06, "loss": 0.0156, "step": 71560 }, { "grad_norm": 0.19407515227794647, "learning_rate": 3.0058723394642994e-06, "loss": 0.0161, "step": 71570 }, { "grad_norm": 0.18677286803722382, "learning_rate": 2.998818149683985e-06, "loss": 0.0197, "step": 71580 }, { "grad_norm": 0.2623957395553589, "learning_rate": 2.9917719911201627e-06, "loss": 0.0142, "step": 71590 }, { "grad_norm": 0.1495334953069687, "learning_rate": 2.9847338649768352e-06, "loss": 0.012, "step": 71600 }, { "grad_norm": 0.1802535355091095, "learning_rate": 2.977703772456608e-06, "loss": 0.0141, "step": 71610 }, { "grad_norm": 0.1763339340686798, "learning_rate": 2.9706817147607535e-06, "loss": 0.014, "step": 71620 }, { "grad_norm": 0.16406996548175812, "learning_rate": 2.9636676930891393e-06, "loss": 0.0158, "step": 71630 }, { "grad_norm": 0.16390760242938995, "learning_rate": 2.9566617086402625e-06, "loss": 0.0147, "step": 71640 }, { "grad_norm": 0.19054235517978668, "learning_rate": 2.9496637626112655e-06, "loss": 0.0162, "step": 71650 }, { "grad_norm": 0.1364361047744751, "learning_rate": 2.9426738561979027e-06, "loss": 0.0138, "step": 71660 }, { "grad_norm": 0.19372251629829407, "learning_rate": 2.9356919905945524e-06, "loss": 0.0271, "step": 71670 }, { "grad_norm": 0.2391776740550995, "learning_rate": 2.928718166994243e-06, "loss": 0.0176, "step": 71680 }, { "grad_norm": 0.2560413181781769, "learning_rate": 2.9217523865885833e-06, "loss": 0.0181, "step": 71690 }, { "grad_norm": 0.12147305905818939, "learning_rate": 2.9147946505678537e-06, "loss": 0.0136, "step": 71700 }, { "grad_norm": 0.24730579555034637, "learning_rate": 2.9078449601209313e-06, "loss": 0.0132, "step": 71710 }, { "grad_norm": 0.1855059117078781, "learning_rate": 2.900903316435333e-06, "loss": 0.0171, "step": 71720 }, { "grad_norm": 0.2696044445037842, "learning_rate": 2.8939697206971983e-06, "loss": 0.0171, "step": 71730 }, { "grad_norm": 0.18339809775352478, "learning_rate": 2.8870441740912746e-06, "loss": 0.0157, "step": 71740 }, { "grad_norm": 0.17962169647216797, "learning_rate": 2.8801266778009762e-06, "loss": 0.0166, "step": 71750 }, { "grad_norm": 0.19567345082759857, "learning_rate": 2.8732172330082798e-06, "loss": 0.0176, "step": 71760 }, { "grad_norm": 0.14533568918704987, "learning_rate": 2.8663158408938517e-06, "loss": 0.014, "step": 71770 }, { "grad_norm": 0.17034663259983063, "learning_rate": 2.859422502636938e-06, "loss": 0.0137, "step": 71780 }, { "grad_norm": 0.14134705066680908, "learning_rate": 2.8525372194154076e-06, "loss": 0.0136, "step": 71790 }, { "grad_norm": 0.10403470695018768, "learning_rate": 2.8456599924057913e-06, "loss": 0.012, "step": 71800 }, { "grad_norm": 0.18866559863090515, "learning_rate": 2.8387908227831995e-06, "loss": 0.0166, "step": 71810 }, { "grad_norm": 0.16512969136238098, "learning_rate": 2.831929711721404e-06, "loss": 0.0136, "step": 71820 }, { "grad_norm": 0.16629526019096375, "learning_rate": 2.825076660392767e-06, "loss": 0.0135, "step": 71830 }, { "grad_norm": 0.20153911411762238, "learning_rate": 2.8182316699682908e-06, "loss": 0.0166, "step": 71840 }, { "grad_norm": 0.11454286426305771, "learning_rate": 2.811394741617601e-06, "loss": 0.0163, "step": 71850 }, { "grad_norm": 0.2396967113018036, "learning_rate": 2.8045658765089356e-06, "loss": 0.0167, "step": 71860 }, { "grad_norm": 0.20726172626018524, "learning_rate": 2.7977450758091605e-06, "loss": 0.0258, "step": 71870 }, { "grad_norm": 0.15205663442611694, "learning_rate": 2.790932340683783e-06, "loss": 0.0201, "step": 71880 }, { "grad_norm": 0.18059347569942474, "learning_rate": 2.7841276722968823e-06, "loss": 0.0136, "step": 71890 }, { "grad_norm": 0.21170176565647125, "learning_rate": 2.7773310718112067e-06, "loss": 0.0147, "step": 71900 }, { "grad_norm": 0.17334575951099396, "learning_rate": 2.7705425403881102e-06, "loss": 0.0138, "step": 71910 }, { "grad_norm": 0.11323227733373642, "learning_rate": 2.763762079187565e-06, "loss": 0.0121, "step": 71920 }, { "grad_norm": 0.12531393766403198, "learning_rate": 2.756989689368178e-06, "loss": 0.0189, "step": 71930 }, { "grad_norm": 0.27500128746032715, "learning_rate": 2.750225372087134e-06, "loss": 0.0167, "step": 71940 }, { "grad_norm": 0.20126594603061676, "learning_rate": 2.7434691285003033e-06, "loss": 0.0125, "step": 71950 }, { "grad_norm": 0.17949402332305908, "learning_rate": 2.7367209597621178e-06, "loss": 0.0172, "step": 71960 }, { "grad_norm": 0.17419248819351196, "learning_rate": 2.729980867025683e-06, "loss": 0.0169, "step": 71970 }, { "grad_norm": 0.2607504427433014, "learning_rate": 2.7232488514426724e-06, "loss": 0.0132, "step": 71980 }, { "grad_norm": 0.24923120439052582, "learning_rate": 2.7165249141634097e-06, "loss": 0.0129, "step": 71990 }, { "grad_norm": 0.20055529475212097, "learning_rate": 2.709809056336837e-06, "loss": 0.0225, "step": 72000 }, { "grad_norm": 0.2219591736793518, "learning_rate": 2.703101279110498e-06, "loss": 0.0157, "step": 72010 }, { "grad_norm": 0.22241997718811035, "learning_rate": 2.696401583630587e-06, "loss": 0.0182, "step": 72020 }, { "grad_norm": 0.19516924023628235, "learning_rate": 2.689709971041887e-06, "loss": 0.0128, "step": 72030 }, { "grad_norm": 0.18966518342494965, "learning_rate": 2.6830264424878126e-06, "loss": 0.0162, "step": 72040 }, { "grad_norm": 0.21403084695339203, "learning_rate": 2.676350999110394e-06, "loss": 0.0255, "step": 72050 }, { "grad_norm": 0.17245249450206757, "learning_rate": 2.6696836420502856e-06, "loss": 0.0217, "step": 72060 }, { "grad_norm": 0.188980370759964, "learning_rate": 2.663024372446765e-06, "loss": 0.0117, "step": 72070 }, { "grad_norm": 0.1616469919681549, "learning_rate": 2.6563731914377055e-06, "loss": 0.0148, "step": 72080 }, { "grad_norm": 0.2155972421169281, "learning_rate": 2.6497301001596087e-06, "loss": 0.02, "step": 72090 }, { "grad_norm": 0.2235182523727417, "learning_rate": 2.643095099747611e-06, "loss": 0.0181, "step": 72100 }, { "grad_norm": 0.14167483150959015, "learning_rate": 2.636468191335445e-06, "loss": 0.0159, "step": 72110 }, { "grad_norm": 0.15372052788734436, "learning_rate": 2.629849376055471e-06, "loss": 0.0148, "step": 72120 }, { "grad_norm": 0.1287245899438858, "learning_rate": 2.6232386550386678e-06, "loss": 0.017, "step": 72130 }, { "grad_norm": 0.17103327810764313, "learning_rate": 2.6166360294146097e-06, "loss": 0.0144, "step": 72140 }, { "grad_norm": 0.16959957778453827, "learning_rate": 2.6100415003115275e-06, "loss": 0.0162, "step": 72150 }, { "grad_norm": 0.22624000906944275, "learning_rate": 2.603455068856225e-06, "loss": 0.0159, "step": 72160 }, { "grad_norm": 0.20036695897579193, "learning_rate": 2.5968767361741584e-06, "loss": 0.0148, "step": 72170 }, { "grad_norm": 0.198612779378891, "learning_rate": 2.590306503389378e-06, "loss": 0.0197, "step": 72180 }, { "grad_norm": 0.1500081866979599, "learning_rate": 2.583744371624558e-06, "loss": 0.0237, "step": 72190 }, { "grad_norm": 0.30091798305511475, "learning_rate": 2.57719034200099e-06, "loss": 0.0133, "step": 72200 }, { "grad_norm": 0.2565554976463318, "learning_rate": 2.570644415638568e-06, "loss": 0.0145, "step": 72210 }, { "grad_norm": 0.1306001991033554, "learning_rate": 2.5641065936558295e-06, "loss": 0.019, "step": 72220 }, { "grad_norm": 0.16050314903259277, "learning_rate": 2.557576877169898e-06, "loss": 0.021, "step": 72230 }, { "grad_norm": 0.20194213092327118, "learning_rate": 2.5510552672965205e-06, "loss": 0.015, "step": 72240 }, { "grad_norm": 0.14573383331298828, "learning_rate": 2.544541765150077e-06, "loss": 0.0143, "step": 72250 }, { "grad_norm": 0.18872910737991333, "learning_rate": 2.5380363718435165e-06, "loss": 0.0161, "step": 72260 }, { "grad_norm": 0.20647065341472626, "learning_rate": 2.5315390884884714e-06, "loss": 0.0129, "step": 72270 }, { "grad_norm": 0.15579372644424438, "learning_rate": 2.5250499161951214e-06, "loss": 0.0133, "step": 72280 }, { "grad_norm": 0.173071026802063, "learning_rate": 2.518568856072301e-06, "loss": 0.0115, "step": 72290 }, { "grad_norm": 0.17650634050369263, "learning_rate": 2.512095909227441e-06, "loss": 0.0147, "step": 72300 }, { "grad_norm": 0.20630834996700287, "learning_rate": 2.50563107676659e-06, "loss": 0.0184, "step": 72310 }, { "grad_norm": 0.18023577332496643, "learning_rate": 2.4991743597944152e-06, "loss": 0.0137, "step": 72320 }, { "grad_norm": 0.22784708440303802, "learning_rate": 2.492725759414205e-06, "loss": 0.0176, "step": 72330 }, { "grad_norm": 0.163030743598938, "learning_rate": 2.4862852767278234e-06, "loss": 0.0093, "step": 72340 }, { "grad_norm": 0.17595165967941284, "learning_rate": 2.479852912835784e-06, "loss": 0.0094, "step": 72350 }, { "grad_norm": 0.19372142851352692, "learning_rate": 2.4734286688372075e-06, "loss": 0.0152, "step": 72360 }, { "grad_norm": 0.17134931683540344, "learning_rate": 2.467012545829811e-06, "loss": 0.0194, "step": 72370 }, { "grad_norm": 0.2184877097606659, "learning_rate": 2.460604544909945e-06, "loss": 0.013, "step": 72380 }, { "grad_norm": 0.2489520162343979, "learning_rate": 2.454204667172555e-06, "loss": 0.0123, "step": 72390 }, { "grad_norm": 0.19796337187290192, "learning_rate": 2.447812913711217e-06, "loss": 0.0137, "step": 72400 }, { "grad_norm": 0.23398953676223755, "learning_rate": 2.44142928561808e-06, "loss": 0.0172, "step": 72410 }, { "grad_norm": 0.2296549379825592, "learning_rate": 2.435053783983965e-06, "loss": 0.0147, "step": 72420 }, { "grad_norm": 0.18301351368427277, "learning_rate": 2.4286864098982453e-06, "loss": 0.0143, "step": 72430 }, { "grad_norm": 0.11551548540592194, "learning_rate": 2.4223271644489397e-06, "loss": 0.0169, "step": 72440 }, { "grad_norm": 0.1800559163093567, "learning_rate": 2.415976048722679e-06, "loss": 0.0149, "step": 72450 }, { "grad_norm": 0.16202540695667267, "learning_rate": 2.4096330638046673e-06, "loss": 0.0125, "step": 72460 }, { "grad_norm": 0.22726985812187195, "learning_rate": 2.4032982107787816e-06, "loss": 0.015, "step": 72470 }, { "grad_norm": 0.1643519550561905, "learning_rate": 2.396971490727451e-06, "loss": 0.0133, "step": 72480 }, { "grad_norm": 0.15808038413524628, "learning_rate": 2.3906529047317493e-06, "loss": 0.012, "step": 72490 }, { "grad_norm": 0.17203649878501892, "learning_rate": 2.3843424538713465e-06, "loss": 0.0117, "step": 72500 }, { "grad_norm": 0.17822560667991638, "learning_rate": 2.3780401392245298e-06, "loss": 0.0162, "step": 72510 }, { "grad_norm": 0.2170555740594864, "learning_rate": 2.3717459618681883e-06, "loss": 0.0143, "step": 72520 }, { "grad_norm": 0.199043408036232, "learning_rate": 2.3654599228778274e-06, "loss": 0.0145, "step": 72530 }, { "grad_norm": 0.1623600274324417, "learning_rate": 2.3591820233275607e-06, "loss": 0.0154, "step": 72540 }, { "grad_norm": 0.10465438663959503, "learning_rate": 2.3529122642901024e-06, "loss": 0.0138, "step": 72550 }, { "grad_norm": 0.23407548666000366, "learning_rate": 2.3466506468367845e-06, "loss": 0.0225, "step": 72560 }, { "grad_norm": 0.15438416600227356, "learning_rate": 2.3403971720375506e-06, "loss": 0.0145, "step": 72570 }, { "grad_norm": 0.2117127925157547, "learning_rate": 2.334151840960952e-06, "loss": 0.0139, "step": 72580 }, { "grad_norm": 0.1877017766237259, "learning_rate": 2.3279146546741347e-06, "loss": 0.0128, "step": 72590 }, { "grad_norm": 0.1761779934167862, "learning_rate": 2.32168561424288e-06, "loss": 0.0168, "step": 72600 }, { "grad_norm": 0.1866685450077057, "learning_rate": 2.3154647207315307e-06, "loss": 0.0182, "step": 72610 }, { "grad_norm": 0.22503136098384857, "learning_rate": 2.309251975203103e-06, "loss": 0.015, "step": 72620 }, { "grad_norm": 0.1737843155860901, "learning_rate": 2.303047378719159e-06, "loss": 0.017, "step": 72630 }, { "grad_norm": 0.21497780084609985, "learning_rate": 2.2968509323399e-06, "loss": 0.02, "step": 72640 }, { "grad_norm": 0.13242971897125244, "learning_rate": 2.290662637124147e-06, "loss": 0.0109, "step": 72650 }, { "grad_norm": 0.13967789709568024, "learning_rate": 2.2844824941292807e-06, "loss": 0.0121, "step": 72660 }, { "grad_norm": 0.23105409741401672, "learning_rate": 2.2783105044113406e-06, "loss": 0.0189, "step": 72670 }, { "grad_norm": 0.2617062032222748, "learning_rate": 2.272146669024944e-06, "loss": 0.0137, "step": 72680 }, { "grad_norm": 0.1489855796098709, "learning_rate": 2.26599098902332e-06, "loss": 0.0143, "step": 72690 }, { "grad_norm": 0.23340749740600586, "learning_rate": 2.2598434654583113e-06, "loss": 0.0143, "step": 72700 }, { "grad_norm": 0.15143480896949768, "learning_rate": 2.253704099380355e-06, "loss": 0.0201, "step": 72710 }, { "grad_norm": 0.2682841718196869, "learning_rate": 2.247572891838512e-06, "loss": 0.0164, "step": 72720 }, { "grad_norm": 0.32895714044570923, "learning_rate": 2.241449843880422e-06, "loss": 0.018, "step": 72730 }, { "grad_norm": 0.13845570385456085, "learning_rate": 2.235334956552354e-06, "loss": 0.0222, "step": 72740 }, { "grad_norm": 0.13972093164920807, "learning_rate": 2.2292282308991775e-06, "loss": 0.0169, "step": 72750 }, { "grad_norm": 0.2006336897611618, "learning_rate": 2.223129667964363e-06, "loss": 0.0158, "step": 72760 }, { "grad_norm": 0.1890943944454193, "learning_rate": 2.2170392687899834e-06, "loss": 0.0133, "step": 72770 }, { "grad_norm": 0.15460076928138733, "learning_rate": 2.210957034416733e-06, "loss": 0.0126, "step": 72780 }, { "grad_norm": 0.15644796192646027, "learning_rate": 2.2048829658838867e-06, "loss": 0.0138, "step": 72790 }, { "grad_norm": 0.17969000339508057, "learning_rate": 2.1988170642293525e-06, "loss": 0.0136, "step": 72800 }, { "grad_norm": 0.10479383170604706, "learning_rate": 2.1927593304896075e-06, "loss": 0.011, "step": 72810 }, { "grad_norm": 0.22810229659080505, "learning_rate": 2.1867097656997626e-06, "loss": 0.0162, "step": 72820 }, { "grad_norm": 0.17224936187267303, "learning_rate": 2.180668370893524e-06, "loss": 0.0138, "step": 72830 }, { "grad_norm": 0.2606518566608429, "learning_rate": 2.174635147103199e-06, "loss": 0.0152, "step": 72840 }, { "grad_norm": 0.24067272245883942, "learning_rate": 2.1686100953597075e-06, "loss": 0.0138, "step": 72850 }, { "grad_norm": 0.20023593306541443, "learning_rate": 2.1625932166925433e-06, "loss": 0.0132, "step": 72860 }, { "grad_norm": 0.4227897524833679, "learning_rate": 2.1565845121298556e-06, "loss": 0.0138, "step": 72870 }, { "grad_norm": 0.21825313568115234, "learning_rate": 2.150583982698351e-06, "loss": 0.016, "step": 72880 }, { "grad_norm": 0.12305549532175064, "learning_rate": 2.144591629423359e-06, "loss": 0.0109, "step": 72890 }, { "grad_norm": 0.2793918251991272, "learning_rate": 2.138607453328817e-06, "loss": 0.015, "step": 72900 }, { "grad_norm": 0.11482802033424377, "learning_rate": 2.132631455437234e-06, "loss": 0.0147, "step": 72910 }, { "grad_norm": 0.14415885508060455, "learning_rate": 2.1266636367697714e-06, "loss": 0.0218, "step": 72920 }, { "grad_norm": 0.22731979191303253, "learning_rate": 2.120703998346152e-06, "loss": 0.0132, "step": 72930 }, { "grad_norm": 0.15632617473602295, "learning_rate": 2.1147525411847114e-06, "loss": 0.0132, "step": 72940 }, { "grad_norm": 0.17207731306552887, "learning_rate": 2.108809266302403e-06, "loss": 0.0128, "step": 72950 }, { "grad_norm": 0.2582785487174988, "learning_rate": 2.102874174714764e-06, "loss": 0.0165, "step": 72960 }, { "grad_norm": 0.24760234355926514, "learning_rate": 2.0969472674359393e-06, "loss": 0.0175, "step": 72970 }, { "grad_norm": 0.17151190340518951, "learning_rate": 2.091028545478668e-06, "loss": 0.0179, "step": 72980 }, { "grad_norm": 0.1453692764043808, "learning_rate": 2.08511800985432e-06, "loss": 0.0131, "step": 72990 }, { "grad_norm": 0.15178050100803375, "learning_rate": 2.07921566157282e-06, "loss": 0.0217, "step": 73000 }, { "grad_norm": 0.19669723510742188, "learning_rate": 2.073321501642728e-06, "loss": 0.013, "step": 73010 }, { "grad_norm": 0.13703325390815735, "learning_rate": 2.0674355310711937e-06, "loss": 0.0163, "step": 73020 }, { "grad_norm": 0.28120774030685425, "learning_rate": 2.0615577508639682e-06, "loss": 0.0111, "step": 73030 }, { "grad_norm": 0.133737251162529, "learning_rate": 2.055688162025399e-06, "loss": 0.0122, "step": 73040 }, { "grad_norm": 0.22410473227500916, "learning_rate": 2.0498267655584546e-06, "loss": 0.0168, "step": 73050 }, { "grad_norm": 0.26939159631729126, "learning_rate": 2.0439735624646626e-06, "loss": 0.0171, "step": 73060 }, { "grad_norm": 0.1896185427904129, "learning_rate": 2.0381285537442008e-06, "loss": 0.0134, "step": 73070 }, { "grad_norm": 0.18164576590061188, "learning_rate": 2.032291740395803e-06, "loss": 0.0134, "step": 73080 }, { "grad_norm": 0.22278448939323425, "learning_rate": 2.0264631234168276e-06, "loss": 0.0153, "step": 73090 }, { "grad_norm": 0.27690020203590393, "learning_rate": 2.0206427038032326e-06, "loss": 0.0157, "step": 73100 }, { "grad_norm": 0.17560917139053345, "learning_rate": 2.0148304825495457e-06, "loss": 0.0106, "step": 73110 }, { "grad_norm": 0.26257750391960144, "learning_rate": 2.0090264606489496e-06, "loss": 0.0133, "step": 73120 }, { "grad_norm": 0.18808528780937195, "learning_rate": 2.003230639093162e-06, "loss": 0.014, "step": 73130 }, { "grad_norm": 0.20861485600471497, "learning_rate": 1.9974430188725525e-06, "loss": 0.0141, "step": 73140 }, { "grad_norm": 0.2306770384311676, "learning_rate": 1.9916636009760513e-06, "loss": 0.0179, "step": 73150 }, { "grad_norm": 0.10240110009908676, "learning_rate": 1.985892386391208e-06, "loss": 0.0183, "step": 73160 }, { "grad_norm": 0.15157395601272583, "learning_rate": 1.980129376104173e-06, "loss": 0.0135, "step": 73170 }, { "grad_norm": 0.12511557340621948, "learning_rate": 1.9743745710996796e-06, "loss": 0.0177, "step": 73180 }, { "grad_norm": 0.13190431892871857, "learning_rate": 1.9686279723610757e-06, "loss": 0.0151, "step": 73190 }, { "grad_norm": 0.2099877893924713, "learning_rate": 1.962889580870281e-06, "loss": 0.0194, "step": 73200 }, { "grad_norm": 0.24251559376716614, "learning_rate": 1.9571593976078385e-06, "loss": 0.0134, "step": 73210 }, { "grad_norm": 0.2155933976173401, "learning_rate": 1.9514374235528876e-06, "loss": 0.0141, "step": 73220 }, { "grad_norm": 0.19190172851085663, "learning_rate": 1.9457236596831408e-06, "loss": 0.0107, "step": 73230 }, { "grad_norm": 0.23024237155914307, "learning_rate": 1.940018106974939e-06, "loss": 0.019, "step": 73240 }, { "grad_norm": 0.15058520436286926, "learning_rate": 1.9343207664032025e-06, "loss": 0.0134, "step": 73250 }, { "grad_norm": 0.145660400390625, "learning_rate": 1.928631638941436e-06, "loss": 0.0119, "step": 73260 }, { "grad_norm": 0.1772662103176117, "learning_rate": 1.922950725561784e-06, "loss": 0.0124, "step": 73270 }, { "grad_norm": 0.14876000583171844, "learning_rate": 1.9172780272349312e-06, "loss": 0.0108, "step": 73280 }, { "grad_norm": 0.1274648904800415, "learning_rate": 1.911613544930202e-06, "loss": 0.0128, "step": 73290 }, { "grad_norm": 0.18595659732818604, "learning_rate": 1.9059572796155e-06, "loss": 0.0146, "step": 73300 }, { "grad_norm": 0.21794258058071136, "learning_rate": 1.9003092322573192e-06, "loss": 0.0101, "step": 73310 }, { "grad_norm": 0.15746383368968964, "learning_rate": 1.8946694038207647e-06, "loss": 0.0182, "step": 73320 }, { "grad_norm": 0.19891424477100372, "learning_rate": 1.8890377952695215e-06, "loss": 0.0119, "step": 73330 }, { "grad_norm": 0.18517832458019257, "learning_rate": 1.8834144075658865e-06, "loss": 0.0138, "step": 73340 }, { "grad_norm": 0.20160344243049622, "learning_rate": 1.8777992416707357e-06, "loss": 0.0177, "step": 73350 }, { "grad_norm": 0.312420129776001, "learning_rate": 1.8721922985435458e-06, "loss": 0.0128, "step": 73360 }, { "grad_norm": 0.16657906770706177, "learning_rate": 1.8665935791424006e-06, "loss": 0.0125, "step": 73370 }, { "grad_norm": 0.1962577998638153, "learning_rate": 1.8610030844239512e-06, "loss": 0.0113, "step": 73380 }, { "grad_norm": 0.12548452615737915, "learning_rate": 1.8554208153434838e-06, "loss": 0.0117, "step": 73390 }, { "grad_norm": 0.16495299339294434, "learning_rate": 1.8498467728548296e-06, "loss": 0.0095, "step": 73400 }, { "grad_norm": 0.24877898395061493, "learning_rate": 1.8442809579104547e-06, "loss": 0.0145, "step": 73410 }, { "grad_norm": 0.18932801485061646, "learning_rate": 1.8387233714614039e-06, "loss": 0.0162, "step": 73420 }, { "grad_norm": 0.13007546961307526, "learning_rate": 1.8331740144573174e-06, "loss": 0.0204, "step": 73430 }, { "grad_norm": 0.17826424539089203, "learning_rate": 1.8276328878464255e-06, "loss": 0.0112, "step": 73440 }, { "grad_norm": 0.27464863657951355, "learning_rate": 1.82209999257556e-06, "loss": 0.0164, "step": 73450 }, { "grad_norm": 0.16648563742637634, "learning_rate": 1.8165753295901311e-06, "loss": 0.0134, "step": 73460 }, { "grad_norm": 0.16867108643054962, "learning_rate": 1.8110588998341616e-06, "loss": 0.014, "step": 73470 }, { "grad_norm": 0.1927386075258255, "learning_rate": 1.805550704250253e-06, "loss": 0.0217, "step": 73480 }, { "grad_norm": 0.19445662200450897, "learning_rate": 1.8000507437796077e-06, "loss": 0.0118, "step": 73490 }, { "grad_norm": 0.21753619611263275, "learning_rate": 1.7945590193620242e-06, "loss": 0.0152, "step": 73500 }, { "grad_norm": 0.14922019839286804, "learning_rate": 1.7890755319358742e-06, "loss": 0.0125, "step": 73510 }, { "grad_norm": 0.19792526960372925, "learning_rate": 1.7836002824381525e-06, "loss": 0.0177, "step": 73520 }, { "grad_norm": 0.20634007453918457, "learning_rate": 1.7781332718044165e-06, "loss": 0.0109, "step": 73530 }, { "grad_norm": 0.2645297050476074, "learning_rate": 1.772674500968835e-06, "loss": 0.0149, "step": 73540 }, { "grad_norm": 0.20380084216594696, "learning_rate": 1.7672239708641624e-06, "loss": 0.0139, "step": 73550 }, { "grad_norm": 0.1888246238231659, "learning_rate": 1.7617816824217371e-06, "loss": 0.012, "step": 73560 }, { "grad_norm": 0.1581467092037201, "learning_rate": 1.7563476365715148e-06, "loss": 0.0221, "step": 73570 }, { "grad_norm": 0.20025111734867096, "learning_rate": 1.750921834242003e-06, "loss": 0.017, "step": 73580 }, { "grad_norm": 0.20054922997951508, "learning_rate": 1.7455042763603436e-06, "loss": 0.0156, "step": 73590 }, { "grad_norm": 0.19619837403297424, "learning_rate": 1.7400949638522345e-06, "loss": 0.0112, "step": 73600 }, { "grad_norm": 0.1916642189025879, "learning_rate": 1.734693897641987e-06, "loss": 0.027, "step": 73610 }, { "grad_norm": 0.2220882624387741, "learning_rate": 1.7293010786524955e-06, "loss": 0.0123, "step": 73620 }, { "grad_norm": 0.2039487212896347, "learning_rate": 1.72391650780524e-06, "loss": 0.0182, "step": 73630 }, { "grad_norm": 0.2021266222000122, "learning_rate": 1.718540186020301e-06, "loss": 0.0117, "step": 73640 }, { "grad_norm": 0.18993626534938812, "learning_rate": 1.7131721142163437e-06, "loss": 0.0127, "step": 73650 }, { "grad_norm": 0.2105688601732254, "learning_rate": 1.7078122933106233e-06, "loss": 0.0161, "step": 73660 }, { "grad_norm": 0.1284400373697281, "learning_rate": 1.7024607242189905e-06, "loss": 0.0169, "step": 73670 }, { "grad_norm": 0.17800946533679962, "learning_rate": 1.6971174078558749e-06, "loss": 0.0129, "step": 73680 }, { "grad_norm": 0.2526850998401642, "learning_rate": 1.6917823451343073e-06, "loss": 0.0158, "step": 73690 }, { "grad_norm": 0.23747381567955017, "learning_rate": 1.6864555369659142e-06, "loss": 0.0158, "step": 73700 }, { "grad_norm": 0.19996124505996704, "learning_rate": 1.681136984260878e-06, "loss": 0.0178, "step": 73710 }, { "grad_norm": 0.19134801626205444, "learning_rate": 1.6758266879280172e-06, "loss": 0.0143, "step": 73720 }, { "grad_norm": 0.25111693143844604, "learning_rate": 1.6705246488746996e-06, "loss": 0.0122, "step": 73730 }, { "grad_norm": 0.10752279311418533, "learning_rate": 1.6652308680069062e-06, "loss": 0.0165, "step": 73740 }, { "grad_norm": 0.19091244041919708, "learning_rate": 1.6599453462292081e-06, "loss": 0.0168, "step": 73750 }, { "grad_norm": 0.24687620997428894, "learning_rate": 1.6546680844447326e-06, "loss": 0.0184, "step": 73760 }, { "grad_norm": 0.188956156373024, "learning_rate": 1.6493990835552475e-06, "loss": 0.0164, "step": 73770 }, { "grad_norm": 0.1693211942911148, "learning_rate": 1.6441383444610547e-06, "loss": 0.0129, "step": 73780 }, { "grad_norm": 0.154872328042984, "learning_rate": 1.6388858680610908e-06, "loss": 0.0168, "step": 73790 }, { "grad_norm": 0.21842816472053528, "learning_rate": 1.6336416552528543e-06, "loss": 0.013, "step": 73800 }, { "grad_norm": 0.09163112193346024, "learning_rate": 1.6284057069324288e-06, "loss": 0.0172, "step": 73810 }, { "grad_norm": 0.16145730018615723, "learning_rate": 1.623178023994504e-06, "loss": 0.0176, "step": 73820 }, { "grad_norm": 0.14418044686317444, "learning_rate": 1.6179586073323483e-06, "loss": 0.0161, "step": 73830 }, { "grad_norm": 0.2569654881954193, "learning_rate": 1.612747457837821e-06, "loss": 0.0156, "step": 73840 }, { "grad_norm": 0.20956645905971527, "learning_rate": 1.607544576401354e-06, "loss": 0.0129, "step": 73850 }, { "grad_norm": 0.15060093998908997, "learning_rate": 1.6023499639119754e-06, "loss": 0.0147, "step": 73860 }, { "grad_norm": 0.1729051172733307, "learning_rate": 1.5971636212573138e-06, "loss": 0.0133, "step": 73870 }, { "grad_norm": 0.21890628337860107, "learning_rate": 1.5919855493235714e-06, "loss": 0.0146, "step": 73880 }, { "grad_norm": 0.16534599661827087, "learning_rate": 1.586815748995535e-06, "loss": 0.0122, "step": 73890 }, { "grad_norm": 0.1740911304950714, "learning_rate": 1.5816542211565866e-06, "loss": 0.0126, "step": 73900 }, { "grad_norm": 0.18667013943195343, "learning_rate": 1.5765009666886765e-06, "loss": 0.0136, "step": 73910 }, { "grad_norm": 0.23245827853679657, "learning_rate": 1.5713559864723781e-06, "loss": 0.0153, "step": 73920 }, { "grad_norm": 0.23923544585704803, "learning_rate": 1.5662192813868104e-06, "loss": 0.0142, "step": 73930 }, { "grad_norm": 0.14897876977920532, "learning_rate": 1.561090852309699e-06, "loss": 0.0181, "step": 73940 }, { "grad_norm": 0.11885964870452881, "learning_rate": 1.5559707001173651e-06, "loss": 0.0126, "step": 73950 }, { "grad_norm": 0.1420379877090454, "learning_rate": 1.5508588256846757e-06, "loss": 0.0164, "step": 73960 }, { "grad_norm": 0.13126727938652039, "learning_rate": 1.5457552298851319e-06, "loss": 0.0124, "step": 73970 }, { "grad_norm": 0.1850547194480896, "learning_rate": 1.5406599135907918e-06, "loss": 0.0151, "step": 73980 }, { "grad_norm": 0.22592370212078094, "learning_rate": 1.5355728776723088e-06, "loss": 0.0166, "step": 73990 }, { "grad_norm": 0.21021641790866852, "learning_rate": 1.5304941229989155e-06, "loss": 0.0142, "step": 74000 }, { "grad_norm": 0.17725858092308044, "learning_rate": 1.5254236504384345e-06, "loss": 0.0149, "step": 74010 }, { "grad_norm": 0.17272500693798065, "learning_rate": 1.5203614608572726e-06, "loss": 0.0144, "step": 74020 }, { "grad_norm": 0.1163245141506195, "learning_rate": 1.5153075551204044e-06, "loss": 0.0117, "step": 74030 }, { "grad_norm": 0.22375933825969696, "learning_rate": 1.5102619340914225e-06, "loss": 0.0259, "step": 74040 }, { "grad_norm": 0.20993904769420624, "learning_rate": 1.505224598632482e-06, "loss": 0.014, "step": 74050 }, { "grad_norm": 0.17742043733596802, "learning_rate": 1.5001955496043164e-06, "loss": 0.0164, "step": 74060 }, { "grad_norm": 0.16263343393802643, "learning_rate": 1.4951747878662604e-06, "loss": 0.0132, "step": 74070 }, { "grad_norm": 0.14324437081813812, "learning_rate": 1.4901623142762221e-06, "loss": 0.0135, "step": 74080 }, { "grad_norm": 0.15371602773666382, "learning_rate": 1.4851581296907001e-06, "loss": 0.0205, "step": 74090 }, { "grad_norm": 0.1723650097846985, "learning_rate": 1.4801622349647714e-06, "loss": 0.014, "step": 74100 }, { "grad_norm": 0.24492627382278442, "learning_rate": 1.4751746309520976e-06, "loss": 0.0151, "step": 74110 }, { "grad_norm": 0.14741814136505127, "learning_rate": 1.4701953185049132e-06, "loss": 0.0208, "step": 74120 }, { "grad_norm": 0.1534925252199173, "learning_rate": 1.4652242984740661e-06, "loss": 0.014, "step": 74130 }, { "grad_norm": 0.14310364425182343, "learning_rate": 1.4602615717089484e-06, "loss": 0.0208, "step": 74140 }, { "grad_norm": 0.2102544754743576, "learning_rate": 1.4553071390575657e-06, "loss": 0.0151, "step": 74150 }, { "grad_norm": 0.23133182525634766, "learning_rate": 1.450361001366496e-06, "loss": 0.0208, "step": 74160 }, { "grad_norm": 0.18554934859275818, "learning_rate": 1.4454231594809021e-06, "loss": 0.0144, "step": 74170 }, { "grad_norm": 0.6704948544502258, "learning_rate": 1.4404936142445036e-06, "loss": 0.0158, "step": 74180 }, { "grad_norm": 0.1734171360731125, "learning_rate": 1.4355723664996546e-06, "loss": 0.0121, "step": 74190 }, { "grad_norm": 0.257373183965683, "learning_rate": 1.4306594170872433e-06, "loss": 0.0138, "step": 74200 }, { "grad_norm": 0.18099191784858704, "learning_rate": 1.4257547668467598e-06, "loss": 0.02, "step": 74210 }, { "grad_norm": 0.15381082892417908, "learning_rate": 1.4208584166162886e-06, "loss": 0.0151, "step": 74220 }, { "grad_norm": 0.2320646494626999, "learning_rate": 1.4159703672324553e-06, "loss": 0.0168, "step": 74230 }, { "grad_norm": 0.22926421463489532, "learning_rate": 1.4110906195305252e-06, "loss": 0.0134, "step": 74240 }, { "grad_norm": 0.17560027539730072, "learning_rate": 1.406219174344292e-06, "loss": 0.0141, "step": 74250 }, { "grad_norm": 0.19459597766399384, "learning_rate": 1.401356032506157e-06, "loss": 0.0136, "step": 74260 }, { "grad_norm": 0.1899181306362152, "learning_rate": 1.3965011948471051e-06, "loss": 0.0171, "step": 74270 }, { "grad_norm": 0.2910880744457245, "learning_rate": 1.3916546621966842e-06, "loss": 0.0151, "step": 74280 }, { "grad_norm": 0.24603092670440674, "learning_rate": 1.3868164353830482e-06, "loss": 0.0118, "step": 74290 }, { "grad_norm": 0.1799238622188568, "learning_rate": 1.3819865152329082e-06, "loss": 0.0109, "step": 74300 }, { "grad_norm": 0.18545971810817719, "learning_rate": 1.377164902571565e-06, "loss": 0.0189, "step": 74310 }, { "grad_norm": 0.3746405839920044, "learning_rate": 1.372351598222904e-06, "loss": 0.0145, "step": 74320 }, { "grad_norm": 0.17125870287418365, "learning_rate": 1.3675466030093843e-06, "loss": 0.0138, "step": 74330 }, { "grad_norm": 0.18418416380882263, "learning_rate": 1.3627499177520486e-06, "loss": 0.0134, "step": 74340 }, { "grad_norm": 0.16887617111206055, "learning_rate": 1.3579615432705196e-06, "loss": 0.0171, "step": 74350 }, { "grad_norm": 0.15975256264209747, "learning_rate": 1.3531814803829978e-06, "loss": 0.0166, "step": 74360 }, { "grad_norm": 0.11150114238262177, "learning_rate": 1.3484097299062803e-06, "loss": 0.0156, "step": 74370 }, { "grad_norm": 0.1256740838289261, "learning_rate": 1.343646292655698e-06, "loss": 0.0134, "step": 74380 }, { "grad_norm": 0.15785782039165497, "learning_rate": 1.3388911694452277e-06, "loss": 0.0154, "step": 74390 }, { "grad_norm": 0.27851444482803345, "learning_rate": 1.3341443610873638e-06, "loss": 0.0181, "step": 74400 }, { "grad_norm": 0.2261459082365036, "learning_rate": 1.329405868393213e-06, "loss": 0.0139, "step": 74410 }, { "grad_norm": 0.13928712904453278, "learning_rate": 1.3246756921724613e-06, "loss": 0.019, "step": 74420 }, { "grad_norm": 0.24095675349235535, "learning_rate": 1.3199538332333506e-06, "loss": 0.0138, "step": 74430 }, { "grad_norm": 0.14573363959789276, "learning_rate": 1.3152402923827411e-06, "loss": 0.0166, "step": 74440 }, { "grad_norm": 0.2515307664871216, "learning_rate": 1.3105350704260277e-06, "loss": 0.0121, "step": 74450 }, { "grad_norm": 0.2031824290752411, "learning_rate": 1.305838168167206e-06, "loss": 0.0133, "step": 74460 }, { "grad_norm": 0.1328979879617691, "learning_rate": 1.301149586408862e-06, "loss": 0.0119, "step": 74470 }, { "grad_norm": 0.20532077550888062, "learning_rate": 1.2964693259521322e-06, "loss": 0.017, "step": 74480 }, { "grad_norm": 0.20037086308002472, "learning_rate": 1.2917973875967548e-06, "loss": 0.0119, "step": 74490 }, { "grad_norm": 0.27232441306114197, "learning_rate": 1.2871337721410249e-06, "loss": 0.0202, "step": 74500 }, { "grad_norm": 0.20071463286876678, "learning_rate": 1.2824784803818379e-06, "loss": 0.0164, "step": 74510 }, { "grad_norm": 0.19175799190998077, "learning_rate": 1.2778315131146467e-06, "loss": 0.0134, "step": 74520 }, { "grad_norm": 0.19229544699192047, "learning_rate": 1.2731928711334994e-06, "loss": 0.0134, "step": 74530 }, { "grad_norm": 0.20056916773319244, "learning_rate": 1.2685625552310122e-06, "loss": 0.0123, "step": 74540 }, { "grad_norm": 0.14661698043346405, "learning_rate": 1.263940566198374e-06, "loss": 0.0135, "step": 74550 }, { "grad_norm": 0.12299211323261261, "learning_rate": 1.259326904825353e-06, "loss": 0.0144, "step": 74560 }, { "grad_norm": 0.23790079355239868, "learning_rate": 1.2547215719003137e-06, "loss": 0.0167, "step": 74570 }, { "grad_norm": 0.12818394601345062, "learning_rate": 1.2501245682101703e-06, "loss": 0.0119, "step": 74580 }, { "grad_norm": 0.17391787469387054, "learning_rate": 1.2455358945404171e-06, "loss": 0.0143, "step": 74590 }, { "grad_norm": 0.3397931754589081, "learning_rate": 1.2409555516751493e-06, "loss": 0.0138, "step": 74600 }, { "grad_norm": 0.19669441878795624, "learning_rate": 1.2363835403970125e-06, "loss": 0.0145, "step": 74610 }, { "grad_norm": 0.14108695089817047, "learning_rate": 1.2318198614872489e-06, "loss": 0.0147, "step": 74620 }, { "grad_norm": 0.24300235509872437, "learning_rate": 1.2272645157256457e-06, "loss": 0.0117, "step": 74630 }, { "grad_norm": 0.17220750451087952, "learning_rate": 1.222717503890608e-06, "loss": 0.018, "step": 74640 }, { "grad_norm": 0.24332045018672943, "learning_rate": 1.218178826759081e-06, "loss": 0.0187, "step": 74650 }, { "grad_norm": 0.19691260159015656, "learning_rate": 1.213648485106611e-06, "loss": 0.0154, "step": 74660 }, { "grad_norm": 0.13825008273124695, "learning_rate": 1.2091264797073066e-06, "loss": 0.0218, "step": 74670 }, { "grad_norm": 0.27668774127960205, "learning_rate": 1.2046128113338494e-06, "loss": 0.0289, "step": 74680 }, { "grad_norm": 0.14797088503837585, "learning_rate": 1.200107480757512e-06, "loss": 0.0204, "step": 74690 }, { "grad_norm": 0.19074024260044098, "learning_rate": 1.1956104887481168e-06, "loss": 0.0211, "step": 74700 }, { "grad_norm": 0.2526986002922058, "learning_rate": 1.191121836074094e-06, "loss": 0.0125, "step": 74710 }, { "grad_norm": 0.19193010032176971, "learning_rate": 1.1866415235024186e-06, "loss": 0.0128, "step": 74720 }, { "grad_norm": 0.1675553172826767, "learning_rate": 1.182169551798662e-06, "loss": 0.0181, "step": 74730 }, { "grad_norm": 0.19767604768276215, "learning_rate": 1.177705921726957e-06, "loss": 0.013, "step": 74740 }, { "grad_norm": 0.13592709600925446, "learning_rate": 1.173250634050016e-06, "loss": 0.0139, "step": 74750 }, { "grad_norm": 0.17498165369033813, "learning_rate": 1.1688036895291354e-06, "loss": 0.0137, "step": 74760 }, { "grad_norm": 0.15938393771648407, "learning_rate": 1.1643650889241574e-06, "loss": 0.0133, "step": 74770 }, { "grad_norm": 0.20703691244125366, "learning_rate": 1.1599348329935311e-06, "loss": 0.0159, "step": 74780 }, { "grad_norm": 0.21196548640727997, "learning_rate": 1.1555129224942673e-06, "loss": 0.0233, "step": 74790 }, { "grad_norm": 0.17057764530181885, "learning_rate": 1.1510993581819396e-06, "loss": 0.0158, "step": 74800 }, { "grad_norm": 0.1887022852897644, "learning_rate": 1.1466941408107112e-06, "loss": 0.0191, "step": 74810 }, { "grad_norm": 0.23428666591644287, "learning_rate": 1.1422972711333247e-06, "loss": 0.0147, "step": 74820 }, { "grad_norm": 0.2478952407836914, "learning_rate": 1.1379087499010565e-06, "loss": 0.0138, "step": 74830 }, { "grad_norm": 0.2444295585155487, "learning_rate": 1.1335285778638127e-06, "loss": 0.0111, "step": 74840 }, { "grad_norm": 0.2217155247926712, "learning_rate": 1.1291567557700333e-06, "loss": 0.0129, "step": 74850 }, { "grad_norm": 0.2769966423511505, "learning_rate": 1.124793284366743e-06, "loss": 0.0155, "step": 74860 }, { "grad_norm": 0.10575027018785477, "learning_rate": 1.1204381643995455e-06, "loss": 0.012, "step": 74870 }, { "grad_norm": 0.14254692196846008, "learning_rate": 1.116091396612595e-06, "loss": 0.0136, "step": 74880 }, { "grad_norm": 0.16040897369384766, "learning_rate": 1.1117529817486538e-06, "loss": 0.0133, "step": 74890 }, { "grad_norm": 0.15149353444576263, "learning_rate": 1.107422920549034e-06, "loss": 0.0123, "step": 74900 }, { "grad_norm": 0.21346794068813324, "learning_rate": 1.1031012137536157e-06, "loss": 0.0146, "step": 74910 }, { "grad_norm": 0.1675005853176117, "learning_rate": 1.0987878621008695e-06, "loss": 0.0117, "step": 74920 }, { "grad_norm": 0.18458311259746552, "learning_rate": 1.0944828663278284e-06, "loss": 0.0125, "step": 74930 }, { "grad_norm": 0.1727297157049179, "learning_rate": 1.0901862271700925e-06, "loss": 0.0174, "step": 74940 }, { "grad_norm": 0.516139805316925, "learning_rate": 1.0858979453618467e-06, "loss": 0.0141, "step": 74950 }, { "grad_norm": 0.2533590793609619, "learning_rate": 1.0816180216358441e-06, "loss": 0.0163, "step": 74960 }, { "grad_norm": 0.14327841997146606, "learning_rate": 1.0773464567233937e-06, "loss": 0.0134, "step": 74970 }, { "grad_norm": 0.12213815748691559, "learning_rate": 1.0730832513543953e-06, "loss": 0.0126, "step": 74980 }, { "grad_norm": 0.2151070088148117, "learning_rate": 1.0688284062573218e-06, "loss": 0.0178, "step": 74990 }, { "grad_norm": 0.2037702202796936, "learning_rate": 1.0645819221591969e-06, "loss": 0.0115, "step": 75000 }, { "grad_norm": 0.2624107003211975, "learning_rate": 1.0603437997856347e-06, "loss": 0.0159, "step": 75010 }, { "grad_norm": 0.2487114667892456, "learning_rate": 1.0561140398608228e-06, "loss": 0.0139, "step": 75020 }, { "grad_norm": 0.1917773336172104, "learning_rate": 1.0518926431074937e-06, "loss": 0.0129, "step": 75030 }, { "grad_norm": 0.1486605852842331, "learning_rate": 1.0476796102469877e-06, "loss": 0.0136, "step": 75040 }, { "grad_norm": 0.18594203889369965, "learning_rate": 1.0434749419991786e-06, "loss": 0.015, "step": 75050 }, { "grad_norm": 0.1555250734090805, "learning_rate": 1.0392786390825415e-06, "loss": 0.0146, "step": 75060 }, { "grad_norm": 0.23739475011825562, "learning_rate": 1.0350907022141087e-06, "loss": 0.0147, "step": 75070 }, { "grad_norm": 0.20887288451194763, "learning_rate": 1.0309111321094744e-06, "loss": 0.0129, "step": 75080 }, { "grad_norm": 0.17318305373191833, "learning_rate": 1.026739929482834e-06, "loss": 0.0106, "step": 75090 }, { "grad_norm": 0.25626322627067566, "learning_rate": 1.022577095046906e-06, "loss": 0.0171, "step": 75100 }, { "grad_norm": 0.15921877324581146, "learning_rate": 1.0184226295130217e-06, "loss": 0.0118, "step": 75110 }, { "grad_norm": 0.296112984418869, "learning_rate": 1.0142765335910575e-06, "loss": 0.0179, "step": 75120 }, { "grad_norm": 0.2553103566169739, "learning_rate": 1.0101388079894746e-06, "loss": 0.0127, "step": 75130 }, { "grad_norm": 0.12417350709438324, "learning_rate": 1.0060094534152908e-06, "loss": 0.0169, "step": 75140 }, { "grad_norm": 0.20148542523384094, "learning_rate": 1.0018884705741028e-06, "loss": 0.0124, "step": 75150 }, { "grad_norm": 0.2344363033771515, "learning_rate": 9.977758601700805e-07, "loss": 0.0206, "step": 75160 }, { "grad_norm": 0.18126127123832703, "learning_rate": 9.93671622905945e-07, "loss": 0.025, "step": 75170 }, { "grad_norm": 0.185630664229393, "learning_rate": 9.895757594830024e-07, "loss": 0.014, "step": 75180 }, { "grad_norm": 0.11715544760227203, "learning_rate": 9.854882706011204e-07, "loss": 0.0114, "step": 75190 }, { "grad_norm": 0.2059541493654251, "learning_rate": 9.814091569587513e-07, "loss": 0.0142, "step": 75200 }, { "grad_norm": 0.16823448240756989, "learning_rate": 9.77338419252888e-07, "loss": 0.0213, "step": 75210 }, { "grad_norm": 0.18572328984737396, "learning_rate": 9.732760581791289e-07, "loss": 0.0148, "step": 75220 }, { "grad_norm": 0.19254781305789948, "learning_rate": 9.692220744315972e-07, "loss": 0.0173, "step": 75230 }, { "grad_norm": 0.15267214179039001, "learning_rate": 9.651764687030162e-07, "loss": 0.0128, "step": 75240 }, { "grad_norm": 0.13687871396541595, "learning_rate": 9.611392416846776e-07, "loss": 0.0135, "step": 75250 }, { "grad_norm": 0.2447763979434967, "learning_rate": 9.571103940664238e-07, "loss": 0.0122, "step": 75260 }, { "grad_norm": 0.2511659562587738, "learning_rate": 9.530899265366822e-07, "loss": 0.0152, "step": 75270 }, { "grad_norm": 0.14038950204849243, "learning_rate": 9.490778397824307e-07, "loss": 0.0124, "step": 75280 }, { "grad_norm": 0.2506433129310608, "learning_rate": 9.450741344892322e-07, "loss": 0.0141, "step": 75290 }, { "grad_norm": 0.2143273800611496, "learning_rate": 9.410788113412116e-07, "loss": 0.0156, "step": 75300 }, { "grad_norm": 0.2197890430688858, "learning_rate": 9.370918710210563e-07, "loss": 0.012, "step": 75310 }, { "grad_norm": 0.15197142958641052, "learning_rate": 9.331133142100268e-07, "loss": 0.0155, "step": 75320 }, { "grad_norm": 0.1972273737192154, "learning_rate": 9.291431415879459e-07, "loss": 0.0114, "step": 75330 }, { "grad_norm": 0.22575023770332336, "learning_rate": 9.251813538332155e-07, "loss": 0.0173, "step": 75340 }, { "grad_norm": 0.1328144520521164, "learning_rate": 9.212279516227884e-07, "loss": 0.011, "step": 75350 }, { "grad_norm": 0.28746241331100464, "learning_rate": 9.172829356322021e-07, "loss": 0.0108, "step": 75360 }, { "grad_norm": 0.15024353563785553, "learning_rate": 9.133463065355452e-07, "loss": 0.0164, "step": 75370 }, { "grad_norm": 0.19312895834445953, "learning_rate": 9.094180650054796e-07, "loss": 0.0163, "step": 75380 }, { "grad_norm": 0.22307054698467255, "learning_rate": 9.054982117132404e-07, "loss": 0.0115, "step": 75390 }, { "grad_norm": 0.2536543607711792, "learning_rate": 9.015867473286143e-07, "loss": 0.0168, "step": 75400 }, { "grad_norm": 0.24964623153209686, "learning_rate": 8.976836725199778e-07, "loss": 0.0139, "step": 75410 }, { "grad_norm": 0.2981503903865814, "learning_rate": 8.937889879542416e-07, "loss": 0.0156, "step": 75420 }, { "grad_norm": 0.2838948965072632, "learning_rate": 8.899026942969068e-07, "loss": 0.0143, "step": 75430 }, { "grad_norm": 0.21361343562602997, "learning_rate": 8.860247922120424e-07, "loss": 0.0153, "step": 75440 }, { "grad_norm": 0.2723942697048187, "learning_rate": 8.821552823622737e-07, "loss": 0.0132, "step": 75450 }, { "grad_norm": 0.09831958264112473, "learning_rate": 8.782941654087884e-07, "loss": 0.0102, "step": 75460 }, { "grad_norm": 0.2093483805656433, "learning_rate": 8.744414420113478e-07, "loss": 0.0167, "step": 75470 }, { "grad_norm": 0.15781211853027344, "learning_rate": 8.705971128282753e-07, "loss": 0.0124, "step": 75480 }, { "grad_norm": 0.1918620765209198, "learning_rate": 8.66761178516473e-07, "loss": 0.0117, "step": 75490 }, { "grad_norm": 0.18286630511283875, "learning_rate": 8.629336397313781e-07, "loss": 0.0149, "step": 75500 }, { "grad_norm": 0.1707284301519394, "learning_rate": 8.591144971270282e-07, "loss": 0.0167, "step": 75510 }, { "grad_norm": 0.20900402963161469, "learning_rate": 8.553037513560069e-07, "loss": 0.0175, "step": 75520 }, { "grad_norm": 0.20491093397140503, "learning_rate": 8.515014030694546e-07, "loss": 0.0148, "step": 75530 }, { "grad_norm": 0.29289644956588745, "learning_rate": 8.47707452917107e-07, "loss": 0.0132, "step": 75540 }, { "grad_norm": 0.20077067613601685, "learning_rate": 8.439219015472343e-07, "loss": 0.0156, "step": 75550 }, { "grad_norm": 0.14547142386436462, "learning_rate": 8.401447496066861e-07, "loss": 0.0124, "step": 75560 }, { "grad_norm": 0.19977442920207977, "learning_rate": 8.363759977408792e-07, "loss": 0.0114, "step": 75570 }, { "grad_norm": 0.12986977398395538, "learning_rate": 8.326156465937817e-07, "loss": 0.0142, "step": 75580 }, { "grad_norm": 0.2762098014354706, "learning_rate": 8.28863696807941e-07, "loss": 0.0179, "step": 75590 }, { "grad_norm": 0.18853110074996948, "learning_rate": 8.251201490244609e-07, "loss": 0.0152, "step": 75600 }, { "grad_norm": 0.17909149825572968, "learning_rate": 8.213850038830129e-07, "loss": 0.0139, "step": 75610 }, { "grad_norm": 0.17092646658420563, "learning_rate": 8.176582620218309e-07, "loss": 0.0163, "step": 75620 }, { "grad_norm": 0.19078293442726135, "learning_rate": 8.139399240777057e-07, "loss": 0.0111, "step": 75630 }, { "grad_norm": 0.23775961995124817, "learning_rate": 8.102299906860122e-07, "loss": 0.015, "step": 75640 }, { "grad_norm": 0.2695343494415283, "learning_rate": 8.065284624806657e-07, "loss": 0.0205, "step": 75650 }, { "grad_norm": 0.12917952239513397, "learning_rate": 8.028353400941601e-07, "loss": 0.0127, "step": 75660 }, { "grad_norm": 0.27792805433273315, "learning_rate": 7.991506241575575e-07, "loss": 0.0167, "step": 75670 }, { "grad_norm": 0.22434720396995544, "learning_rate": 7.954743153004541e-07, "loss": 0.0163, "step": 75680 }, { "grad_norm": 0.13905984163284302, "learning_rate": 7.918064141510528e-07, "loss": 0.0166, "step": 75690 }, { "grad_norm": 0.23653678596019745, "learning_rate": 7.881469213360859e-07, "loss": 0.0207, "step": 75700 }, { "grad_norm": 0.19587865471839905, "learning_rate": 7.844958374808642e-07, "loss": 0.011, "step": 75710 }, { "grad_norm": 0.1352699100971222, "learning_rate": 7.808531632092608e-07, "loss": 0.0121, "step": 75720 }, { "grad_norm": 0.15409669280052185, "learning_rate": 7.772188991436946e-07, "loss": 0.0171, "step": 75730 }, { "grad_norm": 0.2202983945608139, "learning_rate": 7.735930459051799e-07, "loss": 0.0151, "step": 75740 }, { "grad_norm": 0.22011783719062805, "learning_rate": 7.699756041132655e-07, "loss": 0.0155, "step": 75750 }, { "grad_norm": 0.17695513367652893, "learning_rate": 7.663665743860793e-07, "loss": 0.0144, "step": 75760 }, { "grad_norm": 0.15033289790153503, "learning_rate": 7.627659573403001e-07, "loss": 0.0139, "step": 75770 }, { "grad_norm": 0.2400313913822174, "learning_rate": 7.591737535911802e-07, "loss": 0.0162, "step": 75780 }, { "grad_norm": 0.12249476462602615, "learning_rate": 7.555899637525288e-07, "loss": 0.0151, "step": 75790 }, { "grad_norm": 0.22658462822437286, "learning_rate": 7.520145884367058e-07, "loss": 0.0132, "step": 75800 }, { "grad_norm": 0.1847149282693863, "learning_rate": 7.484476282546615e-07, "loss": 0.015, "step": 75810 }, { "grad_norm": 0.20045246183872223, "learning_rate": 7.448890838158806e-07, "loss": 0.0182, "step": 75820 }, { "grad_norm": 0.18535853922367096, "learning_rate": 7.413389557284267e-07, "loss": 0.0149, "step": 75830 }, { "grad_norm": 0.162751242518425, "learning_rate": 7.377972445989201e-07, "loss": 0.0149, "step": 75840 }, { "grad_norm": 0.17066790163516998, "learning_rate": 7.342639510325377e-07, "loss": 0.0184, "step": 75850 }, { "grad_norm": 0.2552420198917389, "learning_rate": 7.307390756330246e-07, "loss": 0.0141, "step": 75860 }, { "grad_norm": 0.2309439480304718, "learning_rate": 7.272226190026876e-07, "loss": 0.0111, "step": 75870 }, { "grad_norm": 0.143477201461792, "learning_rate": 7.237145817423907e-07, "loss": 0.0116, "step": 75880 }, { "grad_norm": 0.17531682550907135, "learning_rate": 7.202149644515654e-07, "loss": 0.0142, "step": 75890 }, { "grad_norm": 0.20469553768634796, "learning_rate": 7.167237677281946e-07, "loss": 0.0148, "step": 75900 }, { "grad_norm": 0.15228725969791412, "learning_rate": 7.132409921688288e-07, "loss": 0.0128, "step": 75910 }, { "grad_norm": 0.2043333351612091, "learning_rate": 7.09766638368592e-07, "loss": 0.0118, "step": 75920 }, { "grad_norm": 0.18702025711536407, "learning_rate": 7.063007069211313e-07, "loss": 0.0104, "step": 75930 }, { "grad_norm": 0.2791317105293274, "learning_rate": 7.028431984187067e-07, "loss": 0.0139, "step": 75940 }, { "grad_norm": 0.0945250615477562, "learning_rate": 6.993941134520898e-07, "loss": 0.0103, "step": 75950 }, { "grad_norm": 0.21118150651454926, "learning_rate": 6.959534526106537e-07, "loss": 0.015, "step": 75960 }, { "grad_norm": 0.20258572697639465, "learning_rate": 6.925212164822948e-07, "loss": 0.0139, "step": 75970 }, { "grad_norm": 0.32912909984588623, "learning_rate": 6.890974056535049e-07, "loss": 0.0142, "step": 75980 }, { "grad_norm": 0.20271676778793335, "learning_rate": 6.856820207093106e-07, "loss": 0.012, "step": 75990 }, { "grad_norm": 0.1778566539287567, "learning_rate": 6.822750622333063e-07, "loss": 0.0106, "step": 76000 }, { "grad_norm": 0.2205493152141571, "learning_rate": 6.788765308076539e-07, "loss": 0.0168, "step": 76010 }, { "grad_norm": 0.2406304031610489, "learning_rate": 6.754864270130668e-07, "loss": 0.0125, "step": 76020 }, { "grad_norm": 0.16855959594249725, "learning_rate": 6.721047514288203e-07, "loss": 0.0133, "step": 76030 }, { "grad_norm": 0.21889007091522217, "learning_rate": 6.687315046327469e-07, "loss": 0.0137, "step": 76040 }, { "grad_norm": 0.2477891892194748, "learning_rate": 6.65366687201252e-07, "loss": 0.015, "step": 76050 }, { "grad_norm": 0.13380149006843567, "learning_rate": 6.620102997092814e-07, "loss": 0.0164, "step": 76060 }, { "grad_norm": 0.23681172728538513, "learning_rate": 6.586623427303596e-07, "loss": 0.0106, "step": 76070 }, { "grad_norm": 0.18225964903831482, "learning_rate": 6.553228168365455e-07, "loss": 0.0181, "step": 76080 }, { "grad_norm": 0.2067812830209732, "learning_rate": 6.519917225984884e-07, "loss": 0.0175, "step": 76090 }, { "grad_norm": 0.22957982122898102, "learning_rate": 6.486690605853718e-07, "loss": 0.0176, "step": 76100 }, { "grad_norm": 0.20428550243377686, "learning_rate": 6.453548313649527e-07, "loss": 0.0128, "step": 76110 }, { "grad_norm": 0.27696341276168823, "learning_rate": 6.420490355035446e-07, "loss": 0.018, "step": 76120 }, { "grad_norm": 0.16716328263282776, "learning_rate": 6.387516735660071e-07, "loss": 0.0151, "step": 76130 }, { "grad_norm": 0.11331204324960709, "learning_rate": 6.35462746115778e-07, "loss": 0.0112, "step": 76140 }, { "grad_norm": 0.33433613181114197, "learning_rate": 6.321822537148358e-07, "loss": 0.0131, "step": 76150 }, { "grad_norm": 0.12756019830703735, "learning_rate": 6.289101969237432e-07, "loss": 0.012, "step": 76160 }, { "grad_norm": 0.24397650361061096, "learning_rate": 6.256465763015918e-07, "loss": 0.016, "step": 76170 }, { "grad_norm": 0.14571219682693481, "learning_rate": 6.22391392406052e-07, "loss": 0.0122, "step": 76180 }, { "grad_norm": 0.11938133835792542, "learning_rate": 6.191446457933403e-07, "loss": 0.0097, "step": 76190 }, { "grad_norm": 0.1814296841621399, "learning_rate": 6.159063370182406e-07, "loss": 0.0125, "step": 76200 }, { "grad_norm": 0.12624286115169525, "learning_rate": 6.126764666340879e-07, "loss": 0.0102, "step": 76210 }, { "grad_norm": 0.09685009717941284, "learning_rate": 6.094550351927852e-07, "loss": 0.0114, "step": 76220 }, { "grad_norm": 0.19203825294971466, "learning_rate": 6.062420432447757e-07, "loss": 0.0129, "step": 76230 }, { "grad_norm": 0.1895035207271576, "learning_rate": 6.030374913390813e-07, "loss": 0.0141, "step": 76240 }, { "grad_norm": 0.291908323764801, "learning_rate": 5.998413800232694e-07, "loss": 0.0151, "step": 76250 }, { "grad_norm": 0.34833088517189026, "learning_rate": 5.966537098434755e-07, "loss": 0.0143, "step": 76260 }, { "grad_norm": 0.11058124899864197, "learning_rate": 5.934744813443694e-07, "loss": 0.0114, "step": 76270 }, { "grad_norm": 0.19888100028038025, "learning_rate": 5.903036950692054e-07, "loss": 0.0179, "step": 76280 }, { "grad_norm": 0.15523773431777954, "learning_rate": 5.871413515597835e-07, "loss": 0.0148, "step": 76290 }, { "grad_norm": 0.19443152844905853, "learning_rate": 5.839874513564547e-07, "loss": 0.0143, "step": 76300 }, { "grad_norm": 0.16937756538391113, "learning_rate": 5.808419949981436e-07, "loss": 0.0129, "step": 76310 }, { "grad_norm": 0.14082536101341248, "learning_rate": 5.777049830223257e-07, "loss": 0.0153, "step": 76320 }, { "grad_norm": 0.1043321043252945, "learning_rate": 5.745764159650114e-07, "loss": 0.0107, "step": 76330 }, { "grad_norm": 0.1922563761472702, "learning_rate": 5.71456294360806e-07, "loss": 0.013, "step": 76340 }, { "grad_norm": 0.15605133771896362, "learning_rate": 5.683446187428443e-07, "loss": 0.0161, "step": 76350 }, { "grad_norm": 0.31273001432418823, "learning_rate": 5.652413896428288e-07, "loss": 0.0149, "step": 76360 }, { "grad_norm": 0.19150866568088531, "learning_rate": 5.621466075910131e-07, "loss": 0.0144, "step": 76370 }, { "grad_norm": 0.1969740390777588, "learning_rate": 5.590602731162187e-07, "loss": 0.0128, "step": 76380 }, { "grad_norm": 0.18495792150497437, "learning_rate": 5.559823867458125e-07, "loss": 0.0119, "step": 76390 }, { "grad_norm": 0.1510903388261795, "learning_rate": 5.529129490057128e-07, "loss": 0.0206, "step": 76400 }, { "grad_norm": 0.2017884999513626, "learning_rate": 5.498519604204167e-07, "loss": 0.0144, "step": 76410 }, { "grad_norm": 0.27894189953804016, "learning_rate": 5.467994215129501e-07, "loss": 0.014, "step": 76420 }, { "grad_norm": 0.14630535244941711, "learning_rate": 5.437553328049183e-07, "loss": 0.0145, "step": 76430 }, { "grad_norm": 0.22858455777168274, "learning_rate": 5.407196948164661e-07, "loss": 0.0118, "step": 76440 }, { "grad_norm": 0.158273845911026, "learning_rate": 5.376925080663009e-07, "loss": 0.0126, "step": 76450 }, { "grad_norm": 0.19560620188713074, "learning_rate": 5.346737730716977e-07, "loss": 0.0146, "step": 76460 }, { "grad_norm": 0.20264290273189545, "learning_rate": 5.316634903484607e-07, "loss": 0.0121, "step": 76470 }, { "grad_norm": 0.19205258786678314, "learning_rate": 5.286616604109728e-07, "loss": 0.0116, "step": 76480 }, { "grad_norm": 0.10124899446964264, "learning_rate": 5.256682837721627e-07, "loss": 0.0169, "step": 76490 }, { "grad_norm": 0.16168345510959625, "learning_rate": 5.226833609435156e-07, "loss": 0.0138, "step": 76500 }, { "grad_norm": 0.16839589178562164, "learning_rate": 5.19706892435079e-07, "loss": 0.0141, "step": 76510 }, { "grad_norm": 0.22307650744915009, "learning_rate": 5.167388787554406e-07, "loss": 0.016, "step": 76520 }, { "grad_norm": 0.201924666762352, "learning_rate": 5.137793204117614e-07, "loss": 0.0144, "step": 76530 }, { "grad_norm": 0.16817377507686615, "learning_rate": 5.108282179097479e-07, "loss": 0.0157, "step": 76540 }, { "grad_norm": 0.17769208550453186, "learning_rate": 5.078855717536523e-07, "loss": 0.0123, "step": 76550 }, { "grad_norm": 0.13163866102695465, "learning_rate": 5.049513824463059e-07, "loss": 0.011, "step": 76560 }, { "grad_norm": 0.1524469554424286, "learning_rate": 5.020256504890742e-07, "loss": 0.0125, "step": 76570 }, { "grad_norm": 0.14015451073646545, "learning_rate": 4.991083763818849e-07, "loss": 0.0165, "step": 76580 }, { "grad_norm": 0.2071596086025238, "learning_rate": 4.961995606232228e-07, "loss": 0.0189, "step": 76590 }, { "grad_norm": 0.2134992778301239, "learning_rate": 4.932992037101236e-07, "loss": 0.014, "step": 76600 }, { "grad_norm": 0.18909825384616852, "learning_rate": 4.904073061381798e-07, "loss": 0.0176, "step": 76610 }, { "grad_norm": 0.1675223857164383, "learning_rate": 4.875238684015349e-07, "loss": 0.0148, "step": 76620 }, { "grad_norm": 0.2537861168384552, "learning_rate": 4.846488909928948e-07, "loss": 0.0157, "step": 76630 }, { "grad_norm": 0.23022295534610748, "learning_rate": 4.81782374403511e-07, "loss": 0.0179, "step": 76640 }, { "grad_norm": 0.13331331312656403, "learning_rate": 4.789243191231918e-07, "loss": 0.0135, "step": 76650 }, { "grad_norm": 0.20694707334041595, "learning_rate": 4.76074725640302e-07, "loss": 0.0186, "step": 76660 }, { "grad_norm": 0.23060767352581024, "learning_rate": 4.732335944417632e-07, "loss": 0.0213, "step": 76670 }, { "grad_norm": 0.2400713562965393, "learning_rate": 4.704009260130371e-07, "loss": 0.0164, "step": 76680 }, { "grad_norm": 0.27851349115371704, "learning_rate": 4.675767208381587e-07, "loss": 0.0152, "step": 76690 }, { "grad_norm": 0.49214744567871094, "learning_rate": 4.647609793997032e-07, "loss": 0.0141, "step": 76700 }, { "grad_norm": 0.19379249215126038, "learning_rate": 4.619537021788023e-07, "loss": 0.0154, "step": 76710 }, { "grad_norm": 0.26377391815185547, "learning_rate": 4.5915488965515005e-07, "loss": 0.0177, "step": 76720 }, { "grad_norm": 0.16342739760875702, "learning_rate": 4.563645423069807e-07, "loss": 0.0133, "step": 76730 }, { "grad_norm": 0.269386887550354, "learning_rate": 4.535826606110849e-07, "loss": 0.0144, "step": 76740 }, { "grad_norm": 0.21636052429676056, "learning_rate": 4.508092450428214e-07, "loss": 0.0154, "step": 76750 }, { "grad_norm": 0.13035589456558228, "learning_rate": 4.480442960760778e-07, "loss": 0.0177, "step": 76760 }, { "grad_norm": 0.11426281929016113, "learning_rate": 4.4528781418332053e-07, "loss": 0.013, "step": 76770 }, { "grad_norm": 0.2217555195093155, "learning_rate": 4.4253979983555073e-07, "loss": 0.0219, "step": 76780 }, { "grad_norm": 0.21361273527145386, "learning_rate": 4.3980025350233154e-07, "loss": 0.0131, "step": 76790 }, { "grad_norm": 0.17812687158584595, "learning_rate": 4.370691756517664e-07, "loss": 0.0148, "step": 76800 }, { "grad_norm": 0.1960245817899704, "learning_rate": 4.3434656675053753e-07, "loss": 0.0153, "step": 76810 }, { "grad_norm": 0.230648010969162, "learning_rate": 4.3163242726385613e-07, "loss": 0.0119, "step": 76820 }, { "grad_norm": 0.13354752957820892, "learning_rate": 4.289267576554956e-07, "loss": 0.0153, "step": 76830 }, { "grad_norm": 0.1913883090019226, "learning_rate": 4.262295583877807e-07, "loss": 0.0104, "step": 76840 }, { "grad_norm": 0.1476486325263977, "learning_rate": 4.235408299215815e-07, "loss": 0.0149, "step": 76850 }, { "grad_norm": 0.19400018453598022, "learning_rate": 4.2086057271634173e-07, "loss": 0.0157, "step": 76860 }, { "grad_norm": 0.12669110298156738, "learning_rate": 4.181887872300394e-07, "loss": 0.015, "step": 76870 }, { "grad_norm": 0.1687619537115097, "learning_rate": 4.155254739192038e-07, "loss": 0.0126, "step": 76880 }, { "grad_norm": 0.1905943751335144, "learning_rate": 4.128706332389265e-07, "loss": 0.0222, "step": 76890 }, { "grad_norm": 0.33339831233024597, "learning_rate": 4.1022426564284453e-07, "loss": 0.0168, "step": 76900 }, { "grad_norm": 0.11916547268629074, "learning_rate": 4.075863715831574e-07, "loss": 0.0177, "step": 76910 }, { "grad_norm": 0.21439334750175476, "learning_rate": 4.0495695151059887e-07, "loss": 0.0199, "step": 76920 }, { "grad_norm": 0.2074434608221054, "learning_rate": 4.023360058744763e-07, "loss": 0.0123, "step": 76930 }, { "grad_norm": 0.15608476102352142, "learning_rate": 3.997235351226314e-07, "loss": 0.0119, "step": 76940 }, { "grad_norm": 0.16876398026943207, "learning_rate": 3.971195397014571e-07, "loss": 0.0159, "step": 76950 }, { "grad_norm": 0.23014581203460693, "learning_rate": 3.9452402005591417e-07, "loss": 0.0182, "step": 76960 }, { "grad_norm": 0.24335221946239471, "learning_rate": 3.9193697662950334e-07, "loss": 0.0169, "step": 76970 }, { "grad_norm": 0.27691319584846497, "learning_rate": 3.893584098642822e-07, "loss": 0.0188, "step": 76980 }, { "grad_norm": 0.19667017459869385, "learning_rate": 3.867883202008538e-07, "loss": 0.0196, "step": 76990 }, { "grad_norm": 0.18240253627300262, "learning_rate": 3.842267080783779e-07, "loss": 0.0149, "step": 77000 }, { "grad_norm": 0.15589307248592377, "learning_rate": 3.816735739345656e-07, "loss": 0.0142, "step": 77010 }, { "grad_norm": 0.1501428186893463, "learning_rate": 3.791289182056679e-07, "loss": 0.0142, "step": 77020 }, { "grad_norm": 0.14639811217784882, "learning_rate": 3.7659274132650913e-07, "loss": 0.0156, "step": 77030 }, { "grad_norm": 0.18757638335227966, "learning_rate": 3.7406504373044826e-07, "loss": 0.0145, "step": 77040 }, { "grad_norm": 0.15699951350688934, "learning_rate": 3.7154582584939533e-07, "loss": 0.0166, "step": 77050 }, { "grad_norm": 0.21118232607841492, "learning_rate": 3.690350881138227e-07, "loss": 0.0153, "step": 77060 }, { "grad_norm": 0.21001999080181122, "learning_rate": 3.665328309527427e-07, "loss": 0.0151, "step": 77070 }, { "grad_norm": 0.21354378759860992, "learning_rate": 3.640390547937245e-07, "loss": 0.0158, "step": 77080 }, { "grad_norm": 0.2255016416311264, "learning_rate": 3.615537600628882e-07, "loss": 0.0153, "step": 77090 }, { "grad_norm": 0.2292691320180893, "learning_rate": 3.5907694718489427e-07, "loss": 0.0171, "step": 77100 }, { "grad_norm": 0.1441882997751236, "learning_rate": 3.566086165829707e-07, "loss": 0.0141, "step": 77110 }, { "grad_norm": 0.22087526321411133, "learning_rate": 3.5414876867888024e-07, "loss": 0.0145, "step": 77120 }, { "grad_norm": 0.13605594635009766, "learning_rate": 3.5169740389295326e-07, "loss": 0.0183, "step": 77130 }, { "grad_norm": 0.2171921730041504, "learning_rate": 3.4925452264405466e-07, "loss": 0.013, "step": 77140 }, { "grad_norm": 0.27731040120124817, "learning_rate": 3.468201253496062e-07, "loss": 0.0127, "step": 77150 }, { "grad_norm": 0.22846972942352295, "learning_rate": 3.443942124255861e-07, "loss": 0.0174, "step": 77160 }, { "grad_norm": 0.18672485649585724, "learning_rate": 3.4197678428650183e-07, "loss": 0.0171, "step": 77170 }, { "grad_norm": 0.16339221596717834, "learning_rate": 3.3956784134544504e-07, "loss": 0.0145, "step": 77180 }, { "grad_norm": 0.23665863275527954, "learning_rate": 3.3716738401402547e-07, "loss": 0.0111, "step": 77190 }, { "grad_norm": 0.18846125900745392, "learning_rate": 3.3477541270241495e-07, "loss": 0.0149, "step": 77200 }, { "grad_norm": 0.17953023314476013, "learning_rate": 3.3239192781934215e-07, "loss": 0.0141, "step": 77210 }, { "grad_norm": 0.14799754321575165, "learning_rate": 3.3001692977207563e-07, "loss": 0.0167, "step": 77220 }, { "grad_norm": 0.14621534943580627, "learning_rate": 3.276504189664409e-07, "loss": 0.014, "step": 77230 }, { "grad_norm": 0.20015348494052887, "learning_rate": 3.252923958068088e-07, "loss": 0.0142, "step": 77240 }, { "grad_norm": 0.17003154754638672, "learning_rate": 3.2294286069609046e-07, "loss": 0.0165, "step": 77250 }, { "grad_norm": 0.1322064995765686, "learning_rate": 3.2060181403577583e-07, "loss": 0.0135, "step": 77260 }, { "grad_norm": 0.20174942910671234, "learning_rate": 3.1826925622587823e-07, "loss": 0.0136, "step": 77270 }, { "grad_norm": 0.20180022716522217, "learning_rate": 3.159451876649622e-07, "loss": 0.0146, "step": 77280 }, { "grad_norm": 0.2276095449924469, "learning_rate": 3.1362960875015444e-07, "loss": 0.0109, "step": 77290 }, { "grad_norm": 0.25278958678245544, "learning_rate": 3.1132251987711637e-07, "loss": 0.0202, "step": 77300 }, { "grad_norm": 0.2916601300239563, "learning_rate": 3.0902392144007695e-07, "loss": 0.0172, "step": 77310 }, { "grad_norm": 0.1610771268606186, "learning_rate": 3.0673381383179436e-07, "loss": 0.016, "step": 77320 }, { "grad_norm": 0.15354447066783905, "learning_rate": 3.0445219744358875e-07, "loss": 0.0147, "step": 77330 }, { "grad_norm": 0.15908676385879517, "learning_rate": 3.021790726653262e-07, "loss": 0.0198, "step": 77340 }, { "grad_norm": 0.2809450328350067, "learning_rate": 2.9991443988542366e-07, "loss": 0.0178, "step": 77350 }, { "grad_norm": 0.2356766015291214, "learning_rate": 2.976582994908439e-07, "loss": 0.0132, "step": 77360 }, { "grad_norm": 0.2100374698638916, "learning_rate": 2.954106518671007e-07, "loss": 0.0138, "step": 77370 }, { "grad_norm": 0.22884170711040497, "learning_rate": 2.9317149739825356e-07, "loss": 0.0135, "step": 77380 }, { "grad_norm": 0.2364913523197174, "learning_rate": 2.909408364669075e-07, "loss": 0.0183, "step": 77390 }, { "grad_norm": 0.2803739011287689, "learning_rate": 2.887186694542299e-07, "loss": 0.0223, "step": 77400 }, { "grad_norm": 0.19122318923473358, "learning_rate": 2.865049967399225e-07, "loss": 0.0121, "step": 77410 }, { "grad_norm": 0.15684203803539276, "learning_rate": 2.842998187022439e-07, "loss": 0.0123, "step": 77420 }, { "grad_norm": 0.21667243540287018, "learning_rate": 2.8210313571800374e-07, "loss": 0.0142, "step": 77430 }, { "grad_norm": 0.2944920063018799, "learning_rate": 2.7991494816255184e-07, "loss": 0.0112, "step": 77440 }, { "grad_norm": 0.20366932451725006, "learning_rate": 2.777352564097779e-07, "loss": 0.0143, "step": 77450 }, { "grad_norm": 0.2687089145183563, "learning_rate": 2.755640608321508e-07, "loss": 0.0133, "step": 77460 }, { "grad_norm": 0.2139131873846054, "learning_rate": 2.73401361800657e-07, "loss": 0.0126, "step": 77470 }, { "grad_norm": 0.21399863064289093, "learning_rate": 2.7124715968484537e-07, "loss": 0.0144, "step": 77480 }, { "grad_norm": 0.1714714914560318, "learning_rate": 2.691014548528104e-07, "loss": 0.0145, "step": 77490 }, { "grad_norm": 0.13884063065052032, "learning_rate": 2.669642476711864e-07, "loss": 0.0155, "step": 77500 }, { "grad_norm": 0.1537598967552185, "learning_rate": 2.648355385051815e-07, "loss": 0.0135, "step": 77510 }, { "grad_norm": 0.14457564055919647, "learning_rate": 2.627153277185157e-07, "loss": 0.0165, "step": 77520 }, { "grad_norm": 0.21941019594669342, "learning_rate": 2.606036156734881e-07, "loss": 0.0142, "step": 77530 }, { "grad_norm": 0.13154351711273193, "learning_rate": 2.5850040273092127e-07, "loss": 0.013, "step": 77540 }, { "grad_norm": 0.14333270490169525, "learning_rate": 2.5640568925020536e-07, "loss": 0.0163, "step": 77550 }, { "grad_norm": 0.2343636006116867, "learning_rate": 2.5431947558927083e-07, "loss": 0.0167, "step": 77560 }, { "grad_norm": 0.18776537477970123, "learning_rate": 2.522417621045825e-07, "loss": 0.0121, "step": 77570 }, { "grad_norm": 0.2161119431257248, "learning_rate": 2.501725491511786e-07, "loss": 0.0183, "step": 77580 }, { "grad_norm": 0.1407102346420288, "learning_rate": 2.481118370826263e-07, "loss": 0.0121, "step": 77590 }, { "grad_norm": 0.22327330708503723, "learning_rate": 2.4605962625104393e-07, "loss": 0.0157, "step": 77600 }, { "grad_norm": 0.137800931930542, "learning_rate": 2.440159170070955e-07, "loss": 0.0176, "step": 77610 }, { "grad_norm": 0.2163296788930893, "learning_rate": 2.41980709699996e-07, "loss": 0.0114, "step": 77620 }, { "grad_norm": 0.13224780559539795, "learning_rate": 2.3995400467751174e-07, "loss": 0.0137, "step": 77630 }, { "grad_norm": 0.15524250268936157, "learning_rate": 2.3793580228594902e-07, "loss": 0.0199, "step": 77640 }, { "grad_norm": 0.2041645050048828, "learning_rate": 2.3592610287015982e-07, "loss": 0.0117, "step": 77650 }, { "grad_norm": 0.1859617680311203, "learning_rate": 2.3392490677354718e-07, "loss": 0.0159, "step": 77660 }, { "grad_norm": 0.19646181166172028, "learning_rate": 2.3193221433806533e-07, "loss": 0.0116, "step": 77670 }, { "grad_norm": 0.169602170586586, "learning_rate": 2.299480259042086e-07, "loss": 0.0123, "step": 77680 }, { "grad_norm": 0.130095437169075, "learning_rate": 2.2797234181102244e-07, "loss": 0.0126, "step": 77690 }, { "grad_norm": 0.14980515837669373, "learning_rate": 2.2600516239609238e-07, "loss": 0.0114, "step": 77700 }, { "grad_norm": 0.18740186095237732, "learning_rate": 2.240464879955606e-07, "loss": 0.0149, "step": 77710 }, { "grad_norm": 0.20584289729595184, "learning_rate": 2.2209631894410387e-07, "loss": 0.0131, "step": 77720 }, { "grad_norm": 0.14810970425605774, "learning_rate": 2.2015465557496117e-07, "loss": 0.0123, "step": 77730 }, { "grad_norm": 0.2694249153137207, "learning_rate": 2.1822149821990602e-07, "loss": 0.0155, "step": 77740 }, { "grad_norm": 0.25783348083496094, "learning_rate": 2.1629684720926303e-07, "loss": 0.0196, "step": 77750 }, { "grad_norm": 0.15208427608013153, "learning_rate": 2.1438070287189693e-07, "loss": 0.0127, "step": 77760 }, { "grad_norm": 0.14087678492069244, "learning_rate": 2.1247306553523472e-07, "loss": 0.0146, "step": 77770 }, { "grad_norm": 0.2308276742696762, "learning_rate": 2.105739355252323e-07, "loss": 0.0151, "step": 77780 }, { "grad_norm": 0.1932353973388672, "learning_rate": 2.0868331316639678e-07, "loss": 0.0123, "step": 77790 }, { "grad_norm": 0.18507029116153717, "learning_rate": 2.0680119878179193e-07, "loss": 0.0141, "step": 77800 }, { "grad_norm": 0.23711824417114258, "learning_rate": 2.0492759269301054e-07, "loss": 0.0145, "step": 77810 }, { "grad_norm": 0.09780074656009674, "learning_rate": 2.0306249522021315e-07, "loss": 0.0114, "step": 77820 }, { "grad_norm": 0.16710536181926727, "learning_rate": 2.0120590668207816e-07, "loss": 0.0119, "step": 77830 }, { "grad_norm": 0.18042944371700287, "learning_rate": 1.993578273958574e-07, "loss": 0.0138, "step": 77840 }, { "grad_norm": 0.19282443821430206, "learning_rate": 1.975182576773371e-07, "loss": 0.0138, "step": 77850 }, { "grad_norm": 0.3305833339691162, "learning_rate": 1.9568719784083812e-07, "loss": 0.0172, "step": 77860 }, { "grad_norm": 0.17320361733436584, "learning_rate": 1.9386464819924898e-07, "loss": 0.0157, "step": 77870 }, { "grad_norm": 0.1360599845647812, "learning_rate": 1.9205060906399285e-07, "loss": 0.0132, "step": 77880 }, { "grad_norm": 0.11311756074428558, "learning_rate": 1.9024508074503845e-07, "loss": 0.0136, "step": 77890 }, { "grad_norm": 0.16034549474716187, "learning_rate": 1.8844806355089452e-07, "loss": 0.0184, "step": 77900 }, { "grad_norm": 0.17850074172019958, "learning_rate": 1.8665955778863208e-07, "loss": 0.0227, "step": 77910 }, { "grad_norm": 0.15755577385425568, "learning_rate": 1.848795637638512e-07, "loss": 0.0119, "step": 77920 }, { "grad_norm": 0.2812889516353607, "learning_rate": 1.831080817807085e-07, "loss": 0.0176, "step": 77930 }, { "grad_norm": 0.13100206851959229, "learning_rate": 1.813451121418952e-07, "loss": 0.0218, "step": 77940 }, { "grad_norm": 0.19759997725486755, "learning_rate": 1.795906551486648e-07, "loss": 0.012, "step": 77950 }, { "grad_norm": 0.1845732182264328, "learning_rate": 1.7784471110079414e-07, "loss": 0.0164, "step": 77960 }, { "grad_norm": 0.17984920740127563, "learning_rate": 1.7610728029662793e-07, "loss": 0.0139, "step": 77970 }, { "grad_norm": 0.17873239517211914, "learning_rate": 1.7437836303303979e-07, "loss": 0.0128, "step": 77980 }, { "grad_norm": 0.16387732326984406, "learning_rate": 1.726579596054545e-07, "loss": 0.0098, "step": 77990 }, { "grad_norm": 0.15601494908332825, "learning_rate": 1.7094607030784803e-07, "loss": 0.0165, "step": 78000 }, { "grad_norm": 0.2155209630727768, "learning_rate": 1.6924269543272532e-07, "loss": 0.0157, "step": 78010 }, { "grad_norm": 0.16371136903762817, "learning_rate": 1.6754783527115348e-07, "loss": 0.0117, "step": 78020 }, { "grad_norm": 0.12595896422863007, "learning_rate": 1.6586149011273422e-07, "loss": 0.0123, "step": 78030 }, { "grad_norm": 0.1945309191942215, "learning_rate": 1.6418366024562038e-07, "loss": 0.0209, "step": 78040 }, { "grad_norm": 0.1453551948070526, "learning_rate": 1.6251434595651037e-07, "loss": 0.0153, "step": 78050 }, { "grad_norm": 0.17956160008907318, "learning_rate": 1.608535475306372e-07, "loss": 0.0137, "step": 78060 }, { "grad_norm": 0.1324070543050766, "learning_rate": 1.5920126525179048e-07, "loss": 0.0118, "step": 78070 }, { "grad_norm": 0.20686212182044983, "learning_rate": 1.5755749940229435e-07, "loss": 0.0138, "step": 78080 }, { "grad_norm": 0.1732499748468399, "learning_rate": 1.559222502630353e-07, "loss": 0.0158, "step": 78090 }, { "grad_norm": 0.17685885727405548, "learning_rate": 1.5429551811341757e-07, "loss": 0.0198, "step": 78100 }, { "grad_norm": 0.19587501883506775, "learning_rate": 1.5267730323141882e-07, "loss": 0.0156, "step": 78110 }, { "grad_norm": 0.23709332942962646, "learning_rate": 1.5106760589353454e-07, "loss": 0.0154, "step": 78120 }, { "grad_norm": 0.1959703117609024, "learning_rate": 1.4946642637483355e-07, "loss": 0.0123, "step": 78130 }, { "grad_norm": 0.2680871784687042, "learning_rate": 1.4787376494889703e-07, "loss": 0.0138, "step": 78140 }, { "grad_norm": 0.2477889209985733, "learning_rate": 1.4628962188787955e-07, "loss": 0.0121, "step": 78150 }, { "grad_norm": 0.1789921224117279, "learning_rate": 1.4471399746247006e-07, "loss": 0.0154, "step": 78160 }, { "grad_norm": 0.19868946075439453, "learning_rate": 1.4314689194188103e-07, "loss": 0.0128, "step": 78170 }, { "grad_norm": 0.23223158717155457, "learning_rate": 1.4158830559390933e-07, "loss": 0.0159, "step": 78180 }, { "grad_norm": 0.14802543818950653, "learning_rate": 1.4003823868486422e-07, "loss": 0.0235, "step": 78190 }, { "grad_norm": 0.14324553310871124, "learning_rate": 1.3849669147960598e-07, "loss": 0.0166, "step": 78200 }, { "grad_norm": 0.16911105811595917, "learning_rate": 1.3696366424155726e-07, "loss": 0.0116, "step": 78210 }, { "grad_norm": 0.29792267084121704, "learning_rate": 1.35439157232653e-07, "loss": 0.0211, "step": 78220 }, { "grad_norm": 0.1929832547903061, "learning_rate": 1.3392317071340144e-07, "loss": 0.0153, "step": 78230 }, { "grad_norm": 0.17528273165225983, "learning_rate": 1.3241570494283984e-07, "loss": 0.0143, "step": 78240 }, { "grad_norm": 0.31502240896224976, "learning_rate": 1.3091676017855103e-07, "loss": 0.0185, "step": 78250 }, { "grad_norm": 0.22799505293369293, "learning_rate": 1.2942633667666904e-07, "loss": 0.0153, "step": 78260 }, { "grad_norm": 0.15679185092449188, "learning_rate": 1.2794443469185679e-07, "loss": 0.0217, "step": 78270 }, { "grad_norm": 0.17307090759277344, "learning_rate": 1.2647105447734508e-07, "loss": 0.0152, "step": 78280 }, { "grad_norm": 0.1825002133846283, "learning_rate": 1.2500619628488254e-07, "loss": 0.0129, "step": 78290 }, { "grad_norm": 0.22032791376113892, "learning_rate": 1.2354986036477446e-07, "loss": 0.0204, "step": 78300 }, { "grad_norm": 0.16959461569786072, "learning_rate": 1.221020469658718e-07, "loss": 0.0167, "step": 78310 }, { "grad_norm": 0.24631640315055847, "learning_rate": 1.2066275633556556e-07, "loss": 0.0143, "step": 78320 }, { "grad_norm": 0.2941567301750183, "learning_rate": 1.192319887197979e-07, "loss": 0.014, "step": 78330 }, { "grad_norm": 0.13750174641609192, "learning_rate": 1.1780974436303438e-07, "loss": 0.0118, "step": 78340 }, { "grad_norm": 0.14362382888793945, "learning_rate": 1.1639602350830836e-07, "loss": 0.0148, "step": 78350 }, { "grad_norm": 0.1775663197040558, "learning_rate": 1.1499082639718217e-07, "loss": 0.0147, "step": 78360 }, { "grad_norm": 0.13904261589050293, "learning_rate": 1.1359415326976374e-07, "loss": 0.0116, "step": 78370 }, { "grad_norm": 0.23845545947551727, "learning_rate": 1.1220600436470663e-07, "loss": 0.0129, "step": 78380 }, { "grad_norm": 0.18127410113811493, "learning_rate": 1.1082637991920996e-07, "loss": 0.0146, "step": 78390 }, { "grad_norm": 0.17168448865413666, "learning_rate": 1.0945528016901851e-07, "loss": 0.0148, "step": 78400 }, { "grad_norm": 0.18047872185707092, "learning_rate": 1.0809270534840599e-07, "loss": 0.017, "step": 78410 }, { "grad_norm": 0.18502730131149292, "learning_rate": 1.067386556902028e-07, "loss": 0.0108, "step": 78420 }, { "grad_norm": 0.13203799724578857, "learning_rate": 1.05393131425785e-07, "loss": 0.0104, "step": 78430 }, { "grad_norm": 0.23555707931518555, "learning_rate": 1.0405613278505199e-07, "loss": 0.0133, "step": 78440 }, { "grad_norm": 0.18394756317138672, "learning_rate": 1.027276599964766e-07, "loss": 0.0141, "step": 78450 }, { "grad_norm": 0.22531965374946594, "learning_rate": 1.0140771328704391e-07, "loss": 0.0115, "step": 78460 }, { "grad_norm": 0.2078036367893219, "learning_rate": 1.0009629288231237e-07, "loss": 0.0147, "step": 78470 }, { "grad_norm": 0.1859203428030014, "learning_rate": 9.879339900635276e-08, "loss": 0.0177, "step": 78480 }, { "grad_norm": 0.23926205933094025, "learning_rate": 9.74990318817981e-08, "loss": 0.02, "step": 78490 }, { "grad_norm": 0.40219366550445557, "learning_rate": 9.621319172982701e-08, "loss": 0.0144, "step": 78500 }, { "grad_norm": 0.18446670472621918, "learning_rate": 9.493587877015264e-08, "loss": 0.0122, "step": 78510 }, { "grad_norm": 0.2121935784816742, "learning_rate": 9.366709322102263e-08, "loss": 0.012, "step": 78520 }, { "grad_norm": 0.3277290165424347, "learning_rate": 9.240683529924688e-08, "loss": 0.0186, "step": 78530 }, { "grad_norm": 0.1800345778465271, "learning_rate": 9.115510522016979e-08, "loss": 0.0155, "step": 78540 }, { "grad_norm": 0.19231298565864563, "learning_rate": 8.991190319767583e-08, "loss": 0.0111, "step": 78550 }, { "grad_norm": 0.178465336561203, "learning_rate": 8.867722944419509e-08, "loss": 0.0125, "step": 78560 }, { "grad_norm": 0.16166259348392487, "learning_rate": 8.745108417069214e-08, "loss": 0.0109, "step": 78570 }, { "grad_norm": 0.29305729269981384, "learning_rate": 8.623346758669381e-08, "loss": 0.0119, "step": 78580 }, { "grad_norm": 0.16770422458648682, "learning_rate": 8.502437990025037e-08, "loss": 0.0122, "step": 78590 }, { "grad_norm": 0.13167116045951843, "learning_rate": 8.382382131795764e-08, "loss": 0.0148, "step": 78600 }, { "grad_norm": 0.2006356120109558, "learning_rate": 8.263179204496818e-08, "loss": 0.0179, "step": 78610 }, { "grad_norm": 0.1795479953289032, "learning_rate": 8.144829228496354e-08, "loss": 0.0132, "step": 78620 }, { "grad_norm": 0.24552172422409058, "learning_rate": 8.027332224016526e-08, "loss": 0.0172, "step": 78630 }, { "grad_norm": 0.21773572266101837, "learning_rate": 7.910688211135164e-08, "loss": 0.0179, "step": 78640 }, { "grad_norm": 0.18644998967647552, "learning_rate": 7.794897209783546e-08, "loss": 0.0171, "step": 78650 }, { "grad_norm": 0.13371604681015015, "learning_rate": 7.679959239746403e-08, "loss": 0.0176, "step": 78660 }, { "grad_norm": 0.1656017154455185, "learning_rate": 7.565874320664135e-08, "loss": 0.0146, "step": 78670 }, { "grad_norm": 0.29039010405540466, "learning_rate": 7.45264247203059e-08, "loss": 0.0154, "step": 78680 }, { "grad_norm": 0.13097520172595978, "learning_rate": 7.340263713194184e-08, "loss": 0.013, "step": 78690 }, { "grad_norm": 0.22464238107204437, "learning_rate": 7.228738063356777e-08, "loss": 0.014, "step": 78700 }, { "grad_norm": 0.3660430908203125, "learning_rate": 7.118065541575903e-08, "loss": 0.0187, "step": 78710 }, { "grad_norm": 0.12468378990888596, "learning_rate": 7.008246166761435e-08, "loss": 0.0102, "step": 78720 }, { "grad_norm": 0.18792657554149628, "learning_rate": 6.899279957679472e-08, "loss": 0.0116, "step": 78730 }, { "grad_norm": 0.18666906654834747, "learning_rate": 6.791166932949011e-08, "loss": 0.0129, "step": 78740 }, { "grad_norm": 0.23257632553577423, "learning_rate": 6.683907111043608e-08, "loss": 0.0161, "step": 78750 }, { "grad_norm": 0.12924715876579285, "learning_rate": 6.577500510290824e-08, "loss": 0.0122, "step": 78760 }, { "grad_norm": 0.2485431283712387, "learning_rate": 6.47194714887278e-08, "loss": 0.0275, "step": 78770 }, { "grad_norm": 0.1173091009259224, "learning_rate": 6.367247044825608e-08, "loss": 0.0126, "step": 78780 }, { "grad_norm": 0.21557097136974335, "learning_rate": 6.263400216039994e-08, "loss": 0.0134, "step": 78790 }, { "grad_norm": 0.25773385167121887, "learning_rate": 6.160406680260078e-08, "loss": 0.0255, "step": 78800 }, { "grad_norm": 0.1948331892490387, "learning_rate": 6.058266455084561e-08, "loss": 0.0132, "step": 78810 }, { "grad_norm": 0.1567980945110321, "learning_rate": 5.956979557967257e-08, "loss": 0.0129, "step": 78820 }, { "grad_norm": 0.19054371118545532, "learning_rate": 5.856546006214325e-08, "loss": 0.0179, "step": 78830 }, { "grad_norm": 0.21197044849395752, "learning_rate": 5.7569658169881466e-08, "loss": 0.0139, "step": 78840 }, { "grad_norm": 0.20595525205135345, "learning_rate": 5.658239007303445e-08, "loss": 0.0149, "step": 78850 }, { "grad_norm": 0.19704090058803558, "learning_rate": 5.560365594030059e-08, "loss": 0.016, "step": 78860 }, { "grad_norm": 0.16607724130153656, "learning_rate": 5.463345593891833e-08, "loss": 0.0106, "step": 78870 }, { "grad_norm": 0.1470111459493637, "learning_rate": 5.367179023467173e-08, "loss": 0.0159, "step": 78880 }, { "grad_norm": 0.21238945424556732, "learning_rate": 5.271865899187378e-08, "loss": 0.012, "step": 78890 }, { "grad_norm": 0.21883513033390045, "learning_rate": 5.177406237340532e-08, "loss": 0.0259, "step": 78900 }, { "grad_norm": 0.20732827484607697, "learning_rate": 5.083800054065946e-08, "loss": 0.0118, "step": 78910 }, { "grad_norm": 0.19117364287376404, "learning_rate": 4.991047365358603e-08, "loss": 0.0114, "step": 78920 }, { "grad_norm": 0.3054584562778473, "learning_rate": 4.899148187067493e-08, "loss": 0.0231, "step": 78930 }, { "grad_norm": 0.17675362527370453, "learning_rate": 4.808102534895609e-08, "loss": 0.0102, "step": 78940 }, { "grad_norm": 0.13150092959403992, "learning_rate": 4.717910424400507e-08, "loss": 0.0148, "step": 78950 }, { "grad_norm": 0.2924870550632477, "learning_rate": 4.628571870993193e-08, "loss": 0.0163, "step": 78960 }, { "grad_norm": 0.21503248810768127, "learning_rate": 4.5400868899392324e-08, "loss": 0.016, "step": 78970 }, { "grad_norm": 0.22843025624752045, "learning_rate": 4.4524554963576436e-08, "loss": 0.0196, "step": 78980 }, { "grad_norm": 0.22715651988983154, "learning_rate": 4.36567770522367e-08, "loss": 0.0133, "step": 78990 }, { "grad_norm": 0.16938482224941254, "learning_rate": 4.279753531364339e-08, "loss": 0.0137, "step": 79000 }, { "grad_norm": 0.23865710198879242, "learning_rate": 4.1946829894617955e-08, "loss": 0.0173, "step": 79010 }, { "grad_norm": 0.12523958086967468, "learning_rate": 4.11046609405219e-08, "loss": 0.017, "step": 79020 }, { "grad_norm": 0.22308237850666046, "learning_rate": 4.027102859526233e-08, "loss": 0.0089, "step": 79030 }, { "grad_norm": 0.2093782275915146, "learning_rate": 3.9445933001280856e-08, "loss": 0.0149, "step": 79040 }, { "grad_norm": 0.3110027015209198, "learning_rate": 3.8629374299564704e-08, "loss": 0.0163, "step": 79050 }, { "grad_norm": 0.16559571027755737, "learning_rate": 3.78213526296467e-08, "loss": 0.0167, "step": 79060 }, { "grad_norm": 0.20169158279895782, "learning_rate": 3.702186812958308e-08, "loss": 0.0137, "step": 79070 }, { "grad_norm": 0.19568517804145813, "learning_rate": 3.623092093599789e-08, "loss": 0.0166, "step": 79080 }, { "grad_norm": 0.3121170699596405, "learning_rate": 3.544851118403303e-08, "loss": 0.0154, "step": 79090 }, { "grad_norm": 0.24434489011764526, "learning_rate": 3.46746390073871e-08, "loss": 0.0173, "step": 79100 }, { "grad_norm": 0.22364266216754913, "learning_rate": 3.390930453829322e-08, "loss": 0.0215, "step": 79110 }, { "grad_norm": 0.1835254281759262, "learning_rate": 3.3152507907519005e-08, "loss": 0.0121, "step": 79120 }, { "grad_norm": 0.17056041955947876, "learning_rate": 3.240424924438323e-08, "loss": 0.0146, "step": 79130 }, { "grad_norm": 0.225466787815094, "learning_rate": 3.1664528676750295e-08, "loss": 0.0166, "step": 79140 }, { "grad_norm": 0.20340967178344727, "learning_rate": 3.093334633100797e-08, "loss": 0.0221, "step": 79150 }, { "grad_norm": 0.17623580992221832, "learning_rate": 3.021070233210077e-08, "loss": 0.0158, "step": 79160 }, { "grad_norm": 0.1899210810661316, "learning_rate": 2.9496596803507692e-08, "loss": 0.0158, "step": 79170 }, { "grad_norm": 0.1914484053850174, "learning_rate": 2.879102986725335e-08, "loss": 0.0137, "step": 79180 }, { "grad_norm": 0.15402290225028992, "learning_rate": 2.809400164389131e-08, "loss": 0.0159, "step": 79190 }, { "grad_norm": 0.20451417565345764, "learning_rate": 2.740551225253185e-08, "loss": 0.0136, "step": 79200 }, { "grad_norm": 0.22042855620384216, "learning_rate": 2.6725561810819754e-08, "loss": 0.018, "step": 79210 }, { "grad_norm": 0.19773021340370178, "learning_rate": 2.605415043493431e-08, "loss": 0.0147, "step": 79220 }, { "grad_norm": 0.21318332850933075, "learning_rate": 2.5391278239605965e-08, "loss": 0.0126, "step": 79230 }, { "grad_norm": 0.15619277954101562, "learning_rate": 2.4736945338094118e-08, "loss": 0.0129, "step": 79240 }, { "grad_norm": 0.2585088610649109, "learning_rate": 2.4091151842214887e-08, "loss": 0.0154, "step": 79250 }, { "grad_norm": 0.08653997629880905, "learning_rate": 2.3453897862318884e-08, "loss": 0.0136, "step": 79260 }, { "grad_norm": 0.19904324412345886, "learning_rate": 2.2825183507285686e-08, "loss": 0.0126, "step": 79270 }, { "grad_norm": 0.2249118685722351, "learning_rate": 2.220500888455157e-08, "loss": 0.0178, "step": 79280 }, { "grad_norm": 0.17372162640094757, "learning_rate": 2.1593374100081777e-08, "loss": 0.0183, "step": 79290 }, { "grad_norm": 0.18741710484027863, "learning_rate": 2.09902792583927e-08, "loss": 0.0189, "step": 79300 }, { "grad_norm": 0.2804661691188812, "learning_rate": 2.0395724462540788e-08, "loss": 0.0131, "step": 79310 }, { "grad_norm": 0.14535295963287354, "learning_rate": 1.9809709814111453e-08, "loss": 0.0113, "step": 79320 }, { "grad_norm": 0.20388491451740265, "learning_rate": 1.923223541324126e-08, "loss": 0.0196, "step": 79330 }, { "grad_norm": 0.17909055948257446, "learning_rate": 1.8663301358606833e-08, "loss": 0.0166, "step": 79340 }, { "grad_norm": 0.10782293975353241, "learning_rate": 1.8102907747419295e-08, "loss": 0.0203, "step": 79350 }, { "grad_norm": 0.15001101791858673, "learning_rate": 1.7551054675435385e-08, "loss": 0.0137, "step": 79360 }, { "grad_norm": 0.12692448496818542, "learning_rate": 1.7007742236957447e-08, "loss": 0.016, "step": 79370 }, { "grad_norm": 0.178747296333313, "learning_rate": 1.647297052481678e-08, "loss": 0.0159, "step": 79380 }, { "grad_norm": 0.17696543037891388, "learning_rate": 1.5946739630390285e-08, "loss": 0.0187, "step": 79390 }, { "grad_norm": 0.22492267191410065, "learning_rate": 1.5429049643606032e-08, "loss": 0.0191, "step": 79400 }, { "grad_norm": 0.1467408984899521, "learning_rate": 1.4919900652909935e-08, "loss": 0.0161, "step": 79410 }, { "grad_norm": 0.32177165150642395, "learning_rate": 1.4419292745310175e-08, "loss": 0.0167, "step": 79420 }, { "grad_norm": 0.17044265568256378, "learning_rate": 1.3927226006343885e-08, "loss": 0.0115, "step": 79430 }, { "grad_norm": 0.29245659708976746, "learning_rate": 1.3443700520093805e-08, "loss": 0.0178, "step": 79440 }, { "grad_norm": 0.21409034729003906, "learning_rate": 1.296871636917718e-08, "loss": 0.0135, "step": 79450 }, { "grad_norm": 0.22234798967838287, "learning_rate": 1.2502273634762419e-08, "loss": 0.0136, "step": 79460 }, { "grad_norm": 0.22372975945472717, "learning_rate": 1.2044372396546876e-08, "loss": 0.0165, "step": 79470 }, { "grad_norm": 0.3085148334503174, "learning_rate": 1.1595012732773524e-08, "loss": 0.0161, "step": 79480 }, { "grad_norm": 0.11778078973293304, "learning_rate": 1.1154194720225386e-08, "loss": 0.0154, "step": 79490 }, { "grad_norm": 0.18802817165851593, "learning_rate": 1.0721918434231093e-08, "loss": 0.0126, "step": 79500 }, { "grad_norm": 0.17850367724895477, "learning_rate": 1.0298183948648232e-08, "loss": 0.0154, "step": 79510 }, { "grad_norm": 0.3674936592578888, "learning_rate": 9.88299133588555e-09, "loss": 0.0131, "step": 79520 }, { "grad_norm": 0.26183846592903137, "learning_rate": 9.476340666891847e-09, "loss": 0.0178, "step": 79530 }, { "grad_norm": 0.14535024762153625, "learning_rate": 9.078232011139332e-09, "loss": 0.0135, "step": 79540 }, { "grad_norm": 0.2088516503572464, "learning_rate": 8.68866543666802e-09, "loss": 0.0162, "step": 79550 }, { "grad_norm": 0.13776333630084991, "learning_rate": 8.307641010035782e-09, "loss": 0.0143, "step": 79560 }, { "grad_norm": 0.15282179415225983, "learning_rate": 7.93515879635165e-09, "loss": 0.0154, "step": 79570 }, { "grad_norm": 0.16962096095085144, "learning_rate": 7.571218859264706e-09, "loss": 0.0153, "step": 79580 }, { "grad_norm": 0.14501310884952545, "learning_rate": 7.215821260958544e-09, "loss": 0.0191, "step": 79590 }, { "grad_norm": 0.1652631014585495, "learning_rate": 6.868966062162363e-09, "loss": 0.0128, "step": 79600 }, { "grad_norm": 0.14761726558208466, "learning_rate": 6.530653322145419e-09, "loss": 0.0116, "step": 79610 }, { "grad_norm": 0.11335444450378418, "learning_rate": 6.200883098717025e-09, "loss": 0.0109, "step": 79620 }, { "grad_norm": 0.3068511486053467, "learning_rate": 5.879655448226551e-09, "loss": 0.0207, "step": 79630 }, { "grad_norm": 0.19211210310459137, "learning_rate": 5.566970425557872e-09, "loss": 0.0119, "step": 79640 }, { "grad_norm": 0.28761041164398193, "learning_rate": 5.26282808414047e-09, "loss": 0.0114, "step": 79650 }, { "grad_norm": 0.28413650393486023, "learning_rate": 4.967228475949437e-09, "loss": 0.0147, "step": 79660 }, { "grad_norm": 0.16927024722099304, "learning_rate": 4.680171651494369e-09, "loss": 0.0128, "step": 79670 }, { "grad_norm": 0.21613860130310059, "learning_rate": 4.40165765981937e-09, "loss": 0.0165, "step": 79680 }, { "grad_norm": 0.17522601783275604, "learning_rate": 4.131686548519698e-09, "loss": 0.0103, "step": 79690 }, { "grad_norm": 0.22446301579475403, "learning_rate": 3.8702583637251214e-09, "loss": 0.0141, "step": 79700 }, { "grad_norm": 0.11455366015434265, "learning_rate": 3.617373150105463e-09, "loss": 0.0148, "step": 79710 }, { "grad_norm": 0.17791719734668732, "learning_rate": 3.3730309508706036e-09, "loss": 0.0131, "step": 79720 }, { "grad_norm": 0.24952183663845062, "learning_rate": 3.137231807781582e-09, "loss": 0.0137, "step": 79730 }, { "grad_norm": 0.13823123276233673, "learning_rate": 2.9099757611172894e-09, "loss": 0.0147, "step": 79740 }, { "grad_norm": 0.19392143189907074, "learning_rate": 2.6912628497133275e-09, "loss": 0.0172, "step": 79750 }, { "grad_norm": 0.22048434615135193, "learning_rate": 2.481093110945354e-09, "loss": 0.0178, "step": 79760 }, { "grad_norm": 0.1696510910987854, "learning_rate": 2.2794665807235327e-09, "loss": 0.0144, "step": 79770 }, { "grad_norm": 0.11885488033294678, "learning_rate": 2.0863832934980843e-09, "loss": 0.014, "step": 79780 }, { "grad_norm": 0.23664042353630066, "learning_rate": 1.901843282264837e-09, "loss": 0.0167, "step": 79790 }, { "grad_norm": 0.2020101696252823, "learning_rate": 1.7258465785541245e-09, "loss": 0.0184, "step": 79800 }, { "grad_norm": 0.15120121836662292, "learning_rate": 1.5583932124418887e-09, "loss": 0.0121, "step": 79810 }, { "grad_norm": 0.15841339528560638, "learning_rate": 1.3994832125441282e-09, "loss": 0.0126, "step": 79820 }, { "grad_norm": 0.23369570076465607, "learning_rate": 1.2491166060057958e-09, "loss": 0.0155, "step": 79830 }, { "grad_norm": 0.2355811893939972, "learning_rate": 1.1072934185230032e-09, "loss": 0.0137, "step": 79840 }, { "grad_norm": 0.10279503464698792, "learning_rate": 9.740136743319195e-10, "loss": 0.0154, "step": 79850 }, { "grad_norm": 0.17167577147483826, "learning_rate": 8.492773962087696e-10, "loss": 0.0136, "step": 79860 }, { "grad_norm": 0.22057273983955383, "learning_rate": 7.330846054587337e-10, "loss": 0.0179, "step": 79870 }, { "grad_norm": 0.16024574637413025, "learning_rate": 6.254353219492526e-10, "loss": 0.014, "step": 79880 }, { "grad_norm": 0.21115514636039734, "learning_rate": 5.263295640600685e-10, "loss": 0.0129, "step": 79890 }, { "grad_norm": 0.1399703174829483, "learning_rate": 4.357673487387359e-10, "loss": 0.0137, "step": 79900 }, { "grad_norm": 0.22455549240112305, "learning_rate": 3.537486914506616e-10, "loss": 0.0161, "step": 79910 }, { "grad_norm": 0.13639865815639496, "learning_rate": 2.8027360621241117e-10, "loss": 0.0128, "step": 79920 }, { "grad_norm": 0.17588289082050323, "learning_rate": 2.153421055806071e-10, "loss": 0.0144, "step": 79930 }, { "grad_norm": 0.17796531319618225, "learning_rate": 1.5895420064637733e-10, "loss": 0.0158, "step": 79940 }, { "grad_norm": 0.19844171404838562, "learning_rate": 1.1110990105200891e-10, "loss": 0.0143, "step": 79950 }, { "grad_norm": 0.193439781665802, "learning_rate": 7.180921496874326e-11, "loss": 0.0161, "step": 79960 }, { "grad_norm": 0.22714529931545258, "learning_rate": 4.10521491134297e-11, "loss": 0.0119, "step": 79970 }, { "grad_norm": 0.20715954899787903, "learning_rate": 1.883870874297422e-11, "loss": 0.0119, "step": 79980 }, { "grad_norm": 0.18060506880283356, "learning_rate": 5.1688976432373584e-12, "loss": 0.0133, "step": 79990 }, { "grad_norm": 0.2432006299495697, "learning_rate": 4.271816234080461e-14, "loss": 0.0145, "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 }