{ "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": 0.4450940191745758, "learning_rate": 2.25e-07, "loss": 1.2582, "step": 10 }, { "grad_norm": 0.44087380170822144, "learning_rate": 4.75e-07, "loss": 1.2184, "step": 20 }, { "grad_norm": 0.6192261576652527, "learning_rate": 7.25e-07, "loss": 1.2533, "step": 30 }, { "grad_norm": 0.6090286374092102, "learning_rate": 9.75e-07, "loss": 1.2197, "step": 40 }, { "grad_norm": 0.5392580032348633, "learning_rate": 1.2250000000000001e-06, "loss": 1.2218, "step": 50 }, { "grad_norm": 0.5120265483856201, "learning_rate": 1.475e-06, "loss": 1.2654, "step": 60 }, { "grad_norm": 1.0089325904846191, "learning_rate": 1.7250000000000002e-06, "loss": 1.2511, "step": 70 }, { "grad_norm": 0.851066529750824, "learning_rate": 1.975e-06, "loss": 1.1856, "step": 80 }, { "grad_norm": 0.9756666421890259, "learning_rate": 2.225e-06, "loss": 1.2235, "step": 90 }, { "grad_norm": 0.6284463405609131, "learning_rate": 2.4750000000000004e-06, "loss": 1.2034, "step": 100 }, { "grad_norm": 0.7821374535560608, "learning_rate": 2.725e-06, "loss": 1.2089, "step": 110 }, { "grad_norm": 0.9772488474845886, "learning_rate": 2.975e-06, "loss": 1.1765, "step": 120 }, { "grad_norm": 1.2786946296691895, "learning_rate": 3.225e-06, "loss": 1.1461, "step": 130 }, { "grad_norm": 1.622755765914917, "learning_rate": 3.4750000000000006e-06, "loss": 1.1183, "step": 140 }, { "grad_norm": 1.3053897619247437, "learning_rate": 3.725e-06, "loss": 1.0946, "step": 150 }, { "grad_norm": 1.741074800491333, "learning_rate": 3.975e-06, "loss": 1.1024, "step": 160 }, { "grad_norm": 1.2488824129104614, "learning_rate": 4.225e-06, "loss": 1.0844, "step": 170 }, { "grad_norm": 1.4162659645080566, "learning_rate": 4.475e-06, "loss": 1.0815, "step": 180 }, { "grad_norm": 1.2415138483047485, "learning_rate": 4.7250000000000005e-06, "loss": 1.0942, "step": 190 }, { "grad_norm": 1.3956605195999146, "learning_rate": 4.975000000000001e-06, "loss": 1.0882, "step": 200 }, { "grad_norm": 1.504615306854248, "learning_rate": 5.225e-06, "loss": 1.0724, "step": 210 }, { "grad_norm": 0.9084427952766418, "learning_rate": 5.475e-06, "loss": 1.081, "step": 220 }, { "grad_norm": 1.3696190118789673, "learning_rate": 5.725e-06, "loss": 1.064, "step": 230 }, { "grad_norm": 1.1638978719711304, "learning_rate": 5.975e-06, "loss": 1.0901, "step": 240 }, { "grad_norm": 1.2000818252563477, "learning_rate": 6.2250000000000005e-06, "loss": 1.087, "step": 250 }, { "grad_norm": 1.1980379819869995, "learning_rate": 6.475000000000001e-06, "loss": 1.0699, "step": 260 }, { "grad_norm": 1.162402629852295, "learning_rate": 6.725000000000001e-06, "loss": 1.0603, "step": 270 }, { "grad_norm": 2.6899807453155518, "learning_rate": 6.975000000000001e-06, "loss": 1.0795, "step": 280 }, { "grad_norm": 1.2968990802764893, "learning_rate": 7.2249999999999994e-06, "loss": 1.0708, "step": 290 }, { "grad_norm": 0.9892047643661499, "learning_rate": 7.4750000000000004e-06, "loss": 1.0644, "step": 300 }, { "grad_norm": 1.1268043518066406, "learning_rate": 7.725e-06, "loss": 1.0884, "step": 310 }, { "grad_norm": 1.363742470741272, "learning_rate": 7.975e-06, "loss": 1.062, "step": 320 }, { "grad_norm": 1.147168755531311, "learning_rate": 8.225e-06, "loss": 1.0671, "step": 330 }, { "grad_norm": 1.5073792934417725, "learning_rate": 8.475000000000001e-06, "loss": 1.0549, "step": 340 }, { "grad_norm": 1.5893869400024414, "learning_rate": 8.725e-06, "loss": 1.0651, "step": 350 }, { "grad_norm": 1.5281041860580444, "learning_rate": 8.975e-06, "loss": 1.0594, "step": 360 }, { "grad_norm": 1.6796185970306396, "learning_rate": 9.225e-06, "loss": 1.0415, "step": 370 }, { "grad_norm": 1.5982517004013062, "learning_rate": 9.475e-06, "loss": 1.0074, "step": 380 }, { "grad_norm": 1.5836015939712524, "learning_rate": 9.725000000000001e-06, "loss": 0.9799, "step": 390 }, { "grad_norm": 2.2924182415008545, "learning_rate": 9.975e-06, "loss": 0.9644, "step": 400 }, { "grad_norm": 2.607724905014038, "learning_rate": 1.0225e-05, "loss": 0.9326, "step": 410 }, { "grad_norm": 2.3181707859039307, "learning_rate": 1.0475e-05, "loss": 0.9052, "step": 420 }, { "grad_norm": 2.185060501098633, "learning_rate": 1.0725e-05, "loss": 0.8779, "step": 430 }, { "grad_norm": 2.80810284614563, "learning_rate": 1.0975e-05, "loss": 0.8473, "step": 440 }, { "grad_norm": 2.285569667816162, "learning_rate": 1.1225e-05, "loss": 0.807, "step": 450 }, { "grad_norm": 2.2951853275299072, "learning_rate": 1.1475000000000001e-05, "loss": 0.805, "step": 460 }, { "grad_norm": 2.389371633529663, "learning_rate": 1.1725e-05, "loss": 0.7845, "step": 470 }, { "grad_norm": 2.859194278717041, "learning_rate": 1.1975e-05, "loss": 0.758, "step": 480 }, { "grad_norm": 2.6941275596618652, "learning_rate": 1.2225e-05, "loss": 0.7196, "step": 490 }, { "grad_norm": 3.0113840103149414, "learning_rate": 1.2475e-05, "loss": 0.6641, "step": 500 }, { "grad_norm": 2.973830461502075, "learning_rate": 1.2725000000000001e-05, "loss": 0.6517, "step": 510 }, { "grad_norm": 2.323878526687622, "learning_rate": 1.2975e-05, "loss": 0.6294, "step": 520 }, { "grad_norm": 3.5302443504333496, "learning_rate": 1.3225000000000001e-05, "loss": 0.5944, "step": 530 }, { "grad_norm": 2.633214235305786, "learning_rate": 1.3475000000000002e-05, "loss": 0.5347, "step": 540 }, { "grad_norm": 2.841697931289673, "learning_rate": 1.3725000000000002e-05, "loss": 0.5652, "step": 550 }, { "grad_norm": 3.032012939453125, "learning_rate": 1.3975000000000003e-05, "loss": 0.5397, "step": 560 }, { "grad_norm": 2.320899248123169, "learning_rate": 1.4225e-05, "loss": 0.4787, "step": 570 }, { "grad_norm": 1.7594988346099854, "learning_rate": 1.4475e-05, "loss": 0.4683, "step": 580 }, { "grad_norm": 2.5404696464538574, "learning_rate": 1.4725e-05, "loss": 0.4679, "step": 590 }, { "grad_norm": 2.3531570434570312, "learning_rate": 1.4975e-05, "loss": 0.4562, "step": 600 }, { "grad_norm": 2.6203858852386475, "learning_rate": 1.5225e-05, "loss": 0.4345, "step": 610 }, { "grad_norm": 3.375673532485962, "learning_rate": 1.5475e-05, "loss": 0.4161, "step": 620 }, { "grad_norm": 2.395888566970825, "learning_rate": 1.5725e-05, "loss": 0.3917, "step": 630 }, { "grad_norm": 3.277223587036133, "learning_rate": 1.5975000000000002e-05, "loss": 0.3652, "step": 640 }, { "grad_norm": 3.3627665042877197, "learning_rate": 1.6225e-05, "loss": 0.3392, "step": 650 }, { "grad_norm": 3.633629083633423, "learning_rate": 1.6475e-05, "loss": 0.3523, "step": 660 }, { "grad_norm": 2.9355545043945312, "learning_rate": 1.6725000000000003e-05, "loss": 0.2924, "step": 670 }, { "grad_norm": 3.315392255783081, "learning_rate": 1.6975000000000003e-05, "loss": 0.2902, "step": 680 }, { "grad_norm": 3.646862030029297, "learning_rate": 1.7225e-05, "loss": 0.2769, "step": 690 }, { "grad_norm": 3.7685117721557617, "learning_rate": 1.7475e-05, "loss": 0.2885, "step": 700 }, { "grad_norm": 2.7233712673187256, "learning_rate": 1.7725e-05, "loss": 0.233, "step": 710 }, { "grad_norm": 2.934483289718628, "learning_rate": 1.7975e-05, "loss": 0.2501, "step": 720 }, { "grad_norm": 3.04390025138855, "learning_rate": 1.8225e-05, "loss": 0.2217, "step": 730 }, { "grad_norm": 2.2089200019836426, "learning_rate": 1.8475000000000002e-05, "loss": 0.1961, "step": 740 }, { "grad_norm": 2.970890522003174, "learning_rate": 1.8725e-05, "loss": 0.2236, "step": 750 }, { "grad_norm": 2.791541576385498, "learning_rate": 1.8975e-05, "loss": 0.2339, "step": 760 }, { "grad_norm": 2.395232677459717, "learning_rate": 1.9225e-05, "loss": 0.2349, "step": 770 }, { "grad_norm": 2.212254047393799, "learning_rate": 1.9475000000000002e-05, "loss": 0.2472, "step": 780 }, { "grad_norm": 2.453815221786499, "learning_rate": 1.9725000000000002e-05, "loss": 0.2081, "step": 790 }, { "grad_norm": 2.50999116897583, "learning_rate": 1.9975e-05, "loss": 0.1979, "step": 800 }, { "grad_norm": 2.588348627090454, "learning_rate": 2.0225000000000004e-05, "loss": 0.1965, "step": 810 }, { "grad_norm": 2.9420390129089355, "learning_rate": 2.0475e-05, "loss": 0.2006, "step": 820 }, { "grad_norm": 3.359429121017456, "learning_rate": 2.0725e-05, "loss": 0.218, "step": 830 }, { "grad_norm": 2.7840304374694824, "learning_rate": 2.0975e-05, "loss": 0.2121, "step": 840 }, { "grad_norm": 2.743640899658203, "learning_rate": 2.1225e-05, "loss": 0.1952, "step": 850 }, { "grad_norm": 2.545715093612671, "learning_rate": 2.1475e-05, "loss": 0.1822, "step": 860 }, { "grad_norm": 2.6750504970550537, "learning_rate": 2.1725e-05, "loss": 0.2176, "step": 870 }, { "grad_norm": 2.373985528945923, "learning_rate": 2.1975000000000002e-05, "loss": 0.1845, "step": 880 }, { "grad_norm": 2.3200910091400146, "learning_rate": 2.2225e-05, "loss": 0.1931, "step": 890 }, { "grad_norm": 2.3821654319763184, "learning_rate": 2.2475e-05, "loss": 0.1963, "step": 900 }, { "grad_norm": 2.273237705230713, "learning_rate": 2.2725000000000003e-05, "loss": 0.1847, "step": 910 }, { "grad_norm": 2.0612263679504395, "learning_rate": 2.2975000000000003e-05, "loss": 0.1738, "step": 920 }, { "grad_norm": 2.4503979682922363, "learning_rate": 2.3225000000000002e-05, "loss": 0.1825, "step": 930 }, { "grad_norm": 2.0669591426849365, "learning_rate": 2.3475e-05, "loss": 0.1962, "step": 940 }, { "grad_norm": 2.206240177154541, "learning_rate": 2.3725e-05, "loss": 0.1667, "step": 950 }, { "grad_norm": 2.1718382835388184, "learning_rate": 2.3975e-05, "loss": 0.1733, "step": 960 }, { "grad_norm": 2.2476301193237305, "learning_rate": 2.4225e-05, "loss": 0.2005, "step": 970 }, { "grad_norm": 2.3549509048461914, "learning_rate": 2.4475000000000002e-05, "loss": 0.1665, "step": 980 }, { "grad_norm": 2.682982921600342, "learning_rate": 2.4725e-05, "loss": 0.1548, "step": 990 }, { "grad_norm": 2.2553462982177734, "learning_rate": 2.4975e-05, "loss": 0.1673, "step": 1000 }, { "grad_norm": 1.7266607284545898, "learning_rate": 2.5225e-05, "loss": 0.1681, "step": 1010 }, { "grad_norm": 2.216512680053711, "learning_rate": 2.5475e-05, "loss": 0.1898, "step": 1020 }, { "grad_norm": 2.078916549682617, "learning_rate": 2.5725e-05, "loss": 0.1465, "step": 1030 }, { "grad_norm": 2.1526520252227783, "learning_rate": 2.5974999999999998e-05, "loss": 0.1634, "step": 1040 }, { "grad_norm": 2.2405924797058105, "learning_rate": 2.6225e-05, "loss": 0.1863, "step": 1050 }, { "grad_norm": 1.8892371654510498, "learning_rate": 2.6475e-05, "loss": 0.2024, "step": 1060 }, { "grad_norm": 2.8121836185455322, "learning_rate": 2.6725e-05, "loss": 0.198, "step": 1070 }, { "grad_norm": 2.198469400405884, "learning_rate": 2.6975000000000002e-05, "loss": 0.2062, "step": 1080 }, { "grad_norm": 1.5671281814575195, "learning_rate": 2.7225e-05, "loss": 0.1778, "step": 1090 }, { "grad_norm": 1.6421841382980347, "learning_rate": 2.7475e-05, "loss": 0.144, "step": 1100 }, { "grad_norm": 2.143622636795044, "learning_rate": 2.7725e-05, "loss": 0.2034, "step": 1110 }, { "grad_norm": 2.4689812660217285, "learning_rate": 2.7975000000000002e-05, "loss": 0.1648, "step": 1120 }, { "grad_norm": 1.5174883604049683, "learning_rate": 2.8225e-05, "loss": 0.175, "step": 1130 }, { "grad_norm": 2.3083338737487793, "learning_rate": 2.8475e-05, "loss": 0.1962, "step": 1140 }, { "grad_norm": 2.1703896522521973, "learning_rate": 2.8725e-05, "loss": 0.2037, "step": 1150 }, { "grad_norm": 1.6882590055465698, "learning_rate": 2.8975000000000003e-05, "loss": 0.1721, "step": 1160 }, { "grad_norm": 1.852895736694336, "learning_rate": 2.9225000000000002e-05, "loss": 0.1754, "step": 1170 }, { "grad_norm": 1.5992851257324219, "learning_rate": 2.9475e-05, "loss": 0.134, "step": 1180 }, { "grad_norm": 1.7504045963287354, "learning_rate": 2.9725000000000004e-05, "loss": 0.1707, "step": 1190 }, { "grad_norm": 3.019169807434082, "learning_rate": 2.9975000000000004e-05, "loss": 0.1737, "step": 1200 }, { "grad_norm": 3.2968063354492188, "learning_rate": 3.0225000000000003e-05, "loss": 0.1946, "step": 1210 }, { "grad_norm": 1.7683662176132202, "learning_rate": 3.0475000000000002e-05, "loss": 0.1513, "step": 1220 }, { "grad_norm": 2.0464720726013184, "learning_rate": 3.0725e-05, "loss": 0.1759, "step": 1230 }, { "grad_norm": 1.6531929969787598, "learning_rate": 3.0975e-05, "loss": 0.1756, "step": 1240 }, { "grad_norm": 1.9651556015014648, "learning_rate": 3.122500000000001e-05, "loss": 0.1451, "step": 1250 }, { "grad_norm": 1.829900860786438, "learning_rate": 3.1475e-05, "loss": 0.1683, "step": 1260 }, { "grad_norm": 1.7428900003433228, "learning_rate": 3.1725e-05, "loss": 0.1374, "step": 1270 }, { "grad_norm": 2.0177273750305176, "learning_rate": 3.1975e-05, "loss": 0.1605, "step": 1280 }, { "grad_norm": 1.6010372638702393, "learning_rate": 3.2225e-05, "loss": 0.1565, "step": 1290 }, { "grad_norm": 1.7133370637893677, "learning_rate": 3.2474999999999997e-05, "loss": 0.173, "step": 1300 }, { "grad_norm": 1.40318763256073, "learning_rate": 3.2725e-05, "loss": 0.1496, "step": 1310 }, { "grad_norm": 1.5165878534317017, "learning_rate": 3.2975e-05, "loss": 0.1733, "step": 1320 }, { "grad_norm": 1.3815499544143677, "learning_rate": 3.3225e-05, "loss": 0.1448, "step": 1330 }, { "grad_norm": 1.7382512092590332, "learning_rate": 3.3475e-05, "loss": 0.1433, "step": 1340 }, { "grad_norm": 2.0284533500671387, "learning_rate": 3.3725e-05, "loss": 0.1718, "step": 1350 }, { "grad_norm": 2.068260669708252, "learning_rate": 3.3975e-05, "loss": 0.1561, "step": 1360 }, { "grad_norm": 2.1779673099517822, "learning_rate": 3.4225e-05, "loss": 0.1502, "step": 1370 }, { "grad_norm": 1.6961978673934937, "learning_rate": 3.4475000000000005e-05, "loss": 0.1307, "step": 1380 }, { "grad_norm": 1.3324259519577026, "learning_rate": 3.4725000000000004e-05, "loss": 0.147, "step": 1390 }, { "grad_norm": 1.6298571825027466, "learning_rate": 3.4975e-05, "loss": 0.1386, "step": 1400 }, { "grad_norm": 1.7677408456802368, "learning_rate": 3.5225e-05, "loss": 0.1548, "step": 1410 }, { "grad_norm": 2.0109658241271973, "learning_rate": 3.5475e-05, "loss": 0.1615, "step": 1420 }, { "grad_norm": 1.1270902156829834, "learning_rate": 3.5725e-05, "loss": 0.1309, "step": 1430 }, { "grad_norm": 1.4549148082733154, "learning_rate": 3.5975e-05, "loss": 0.1574, "step": 1440 }, { "grad_norm": 1.6183643341064453, "learning_rate": 3.6225000000000006e-05, "loss": 0.172, "step": 1450 }, { "grad_norm": 1.70325767993927, "learning_rate": 3.6475000000000006e-05, "loss": 0.1623, "step": 1460 }, { "grad_norm": 1.470543384552002, "learning_rate": 3.6725000000000005e-05, "loss": 0.1637, "step": 1470 }, { "grad_norm": 1.1655867099761963, "learning_rate": 3.6975000000000004e-05, "loss": 0.142, "step": 1480 }, { "grad_norm": 1.3313275575637817, "learning_rate": 3.7225000000000004e-05, "loss": 0.1504, "step": 1490 }, { "grad_norm": 1.2178725004196167, "learning_rate": 3.7475e-05, "loss": 0.1465, "step": 1500 }, { "grad_norm": 1.369546890258789, "learning_rate": 3.7725e-05, "loss": 0.1399, "step": 1510 }, { "grad_norm": 1.6173664331436157, "learning_rate": 3.7975e-05, "loss": 0.1496, "step": 1520 }, { "grad_norm": 1.3539457321166992, "learning_rate": 3.8225e-05, "loss": 0.1516, "step": 1530 }, { "grad_norm": 1.610864281654358, "learning_rate": 3.8475e-05, "loss": 0.1362, "step": 1540 }, { "grad_norm": 1.98118257522583, "learning_rate": 3.8725e-05, "loss": 0.1448, "step": 1550 }, { "grad_norm": 1.7497546672821045, "learning_rate": 3.8975e-05, "loss": 0.1656, "step": 1560 }, { "grad_norm": 1.2543483972549438, "learning_rate": 3.9225e-05, "loss": 0.1353, "step": 1570 }, { "grad_norm": 1.4486325979232788, "learning_rate": 3.9475000000000004e-05, "loss": 0.1468, "step": 1580 }, { "grad_norm": 2.025000810623169, "learning_rate": 3.9725e-05, "loss": 0.1535, "step": 1590 }, { "grad_norm": 1.5267282724380493, "learning_rate": 3.9975e-05, "loss": 0.1456, "step": 1600 }, { "grad_norm": 1.0693435668945312, "learning_rate": 4.0225e-05, "loss": 0.1244, "step": 1610 }, { "grad_norm": 1.3195189237594604, "learning_rate": 4.0475e-05, "loss": 0.1375, "step": 1620 }, { "grad_norm": 1.5894659757614136, "learning_rate": 4.0725e-05, "loss": 0.1786, "step": 1630 }, { "grad_norm": 1.338096261024475, "learning_rate": 4.0975e-05, "loss": 0.1373, "step": 1640 }, { "grad_norm": 1.1241178512573242, "learning_rate": 4.1225e-05, "loss": 0.1574, "step": 1650 }, { "grad_norm": 1.3437137603759766, "learning_rate": 4.1475000000000005e-05, "loss": 0.152, "step": 1660 }, { "grad_norm": 1.5859893560409546, "learning_rate": 4.1725000000000005e-05, "loss": 0.138, "step": 1670 }, { "grad_norm": 1.4555431604385376, "learning_rate": 4.1975000000000004e-05, "loss": 0.137, "step": 1680 }, { "grad_norm": 1.2781306505203247, "learning_rate": 4.2225e-05, "loss": 0.1442, "step": 1690 }, { "grad_norm": 1.3589353561401367, "learning_rate": 4.2475e-05, "loss": 0.1512, "step": 1700 }, { "grad_norm": 1.5053144693374634, "learning_rate": 4.2725e-05, "loss": 0.1725, "step": 1710 }, { "grad_norm": 1.4924253225326538, "learning_rate": 4.2975e-05, "loss": 0.1644, "step": 1720 }, { "grad_norm": 1.335911750793457, "learning_rate": 4.322500000000001e-05, "loss": 0.1433, "step": 1730 }, { "grad_norm": 1.1429210901260376, "learning_rate": 4.3475000000000006e-05, "loss": 0.1317, "step": 1740 }, { "grad_norm": 1.7301899194717407, "learning_rate": 4.3725000000000006e-05, "loss": 0.1603, "step": 1750 }, { "grad_norm": 1.7533316612243652, "learning_rate": 4.3975e-05, "loss": 0.1533, "step": 1760 }, { "grad_norm": 1.3811079263687134, "learning_rate": 4.4225e-05, "loss": 0.1428, "step": 1770 }, { "grad_norm": 1.2393150329589844, "learning_rate": 4.4475e-05, "loss": 0.1629, "step": 1780 }, { "grad_norm": 1.6799638271331787, "learning_rate": 4.4725e-05, "loss": 0.136, "step": 1790 }, { "grad_norm": 1.0942517518997192, "learning_rate": 4.4975e-05, "loss": 0.1319, "step": 1800 }, { "grad_norm": 1.0564799308776855, "learning_rate": 4.5225e-05, "loss": 0.1298, "step": 1810 }, { "grad_norm": 1.662104845046997, "learning_rate": 4.5475e-05, "loss": 0.1507, "step": 1820 }, { "grad_norm": 1.593675971031189, "learning_rate": 4.5725e-05, "loss": 0.1408, "step": 1830 }, { "grad_norm": 1.190769076347351, "learning_rate": 4.5975e-05, "loss": 0.1597, "step": 1840 }, { "grad_norm": 1.1505697965621948, "learning_rate": 4.6225e-05, "loss": 0.1715, "step": 1850 }, { "grad_norm": 1.422783374786377, "learning_rate": 4.6475000000000005e-05, "loss": 0.1664, "step": 1860 }, { "grad_norm": 1.096787452697754, "learning_rate": 4.6725000000000004e-05, "loss": 0.1414, "step": 1870 }, { "grad_norm": 1.0750486850738525, "learning_rate": 4.6975000000000003e-05, "loss": 0.1177, "step": 1880 }, { "grad_norm": 1.3206045627593994, "learning_rate": 4.7225e-05, "loss": 0.1477, "step": 1890 }, { "grad_norm": 1.1959607601165771, "learning_rate": 4.7475e-05, "loss": 0.1371, "step": 1900 }, { "grad_norm": 1.5995391607284546, "learning_rate": 4.7725e-05, "loss": 0.1651, "step": 1910 }, { "grad_norm": 1.1774572134017944, "learning_rate": 4.7975e-05, "loss": 0.137, "step": 1920 }, { "grad_norm": 1.0684709548950195, "learning_rate": 4.822500000000001e-05, "loss": 0.1277, "step": 1930 }, { "grad_norm": 1.3912928104400635, "learning_rate": 4.8475000000000006e-05, "loss": 0.1531, "step": 1940 }, { "grad_norm": 1.0053234100341797, "learning_rate": 4.8725000000000005e-05, "loss": 0.1226, "step": 1950 }, { "grad_norm": 1.3087135553359985, "learning_rate": 4.8975000000000005e-05, "loss": 0.1718, "step": 1960 }, { "grad_norm": 1.2536143064498901, "learning_rate": 4.9225000000000004e-05, "loss": 0.1232, "step": 1970 }, { "grad_norm": 1.0806366205215454, "learning_rate": 4.9475e-05, "loss": 0.1502, "step": 1980 }, { "grad_norm": 1.3064851760864258, "learning_rate": 4.9725e-05, "loss": 0.1281, "step": 1990 }, { "grad_norm": 1.3808708190917969, "learning_rate": 4.9975e-05, "loss": 0.1458, "step": 2000 }, { "grad_norm": 1.0702277421951294, "learning_rate": 5.0225e-05, "loss": 0.1778, "step": 2010 }, { "grad_norm": 1.4698598384857178, "learning_rate": 5.047500000000001e-05, "loss": 0.1417, "step": 2020 }, { "grad_norm": 1.3715988397598267, "learning_rate": 5.0725e-05, "loss": 0.126, "step": 2030 }, { "grad_norm": 1.0210734605789185, "learning_rate": 5.0975000000000006e-05, "loss": 0.1476, "step": 2040 }, { "grad_norm": 1.1290373802185059, "learning_rate": 5.1225e-05, "loss": 0.1498, "step": 2050 }, { "grad_norm": 1.2514164447784424, "learning_rate": 5.1475000000000004e-05, "loss": 0.1261, "step": 2060 }, { "grad_norm": 1.29497230052948, "learning_rate": 5.1725000000000004e-05, "loss": 0.1292, "step": 2070 }, { "grad_norm": 1.4093424081802368, "learning_rate": 5.197500000000001e-05, "loss": 0.1368, "step": 2080 }, { "grad_norm": 1.1053262948989868, "learning_rate": 5.2225e-05, "loss": 0.1559, "step": 2090 }, { "grad_norm": 1.1614304780960083, "learning_rate": 5.247500000000001e-05, "loss": 0.1287, "step": 2100 }, { "grad_norm": 1.093902587890625, "learning_rate": 5.2725e-05, "loss": 0.1569, "step": 2110 }, { "grad_norm": 1.0734657049179077, "learning_rate": 5.297500000000001e-05, "loss": 0.1339, "step": 2120 }, { "grad_norm": 0.9110158085823059, "learning_rate": 5.3225e-05, "loss": 0.1292, "step": 2130 }, { "grad_norm": 0.962859570980072, "learning_rate": 5.3475e-05, "loss": 0.1048, "step": 2140 }, { "grad_norm": 0.9994940161705017, "learning_rate": 5.3725000000000005e-05, "loss": 0.1481, "step": 2150 }, { "grad_norm": 0.8495805263519287, "learning_rate": 5.3975e-05, "loss": 0.1238, "step": 2160 }, { "grad_norm": 1.0142875909805298, "learning_rate": 5.4225000000000003e-05, "loss": 0.1373, "step": 2170 }, { "grad_norm": 1.1477816104888916, "learning_rate": 5.4474999999999996e-05, "loss": 0.1356, "step": 2180 }, { "grad_norm": 1.1051725149154663, "learning_rate": 5.4725e-05, "loss": 0.1241, "step": 2190 }, { "grad_norm": 0.8502870798110962, "learning_rate": 5.4975e-05, "loss": 0.1216, "step": 2200 }, { "grad_norm": 0.9266455769538879, "learning_rate": 5.522500000000001e-05, "loss": 0.1366, "step": 2210 }, { "grad_norm": 1.1225684881210327, "learning_rate": 5.5475e-05, "loss": 0.1139, "step": 2220 }, { "grad_norm": 1.085027813911438, "learning_rate": 5.5725000000000006e-05, "loss": 0.1234, "step": 2230 }, { "grad_norm": 1.1744202375411987, "learning_rate": 5.5975e-05, "loss": 0.1219, "step": 2240 }, { "grad_norm": 1.0565526485443115, "learning_rate": 5.6225000000000005e-05, "loss": 0.1503, "step": 2250 }, { "grad_norm": 0.9023357033729553, "learning_rate": 5.6475e-05, "loss": 0.1146, "step": 2260 }, { "grad_norm": 0.8053526282310486, "learning_rate": 5.6725e-05, "loss": 0.1183, "step": 2270 }, { "grad_norm": 0.756077766418457, "learning_rate": 5.6975e-05, "loss": 0.1261, "step": 2280 }, { "grad_norm": 1.2879607677459717, "learning_rate": 5.722500000000001e-05, "loss": 0.1426, "step": 2290 }, { "grad_norm": 1.0077379941940308, "learning_rate": 5.7475e-05, "loss": 0.1156, "step": 2300 }, { "grad_norm": 1.0750725269317627, "learning_rate": 5.772500000000001e-05, "loss": 0.131, "step": 2310 }, { "grad_norm": 1.1108386516571045, "learning_rate": 5.7975e-05, "loss": 0.1129, "step": 2320 }, { "grad_norm": 1.072585105895996, "learning_rate": 5.8225000000000006e-05, "loss": 0.1487, "step": 2330 }, { "grad_norm": 0.887556791305542, "learning_rate": 5.8475000000000005e-05, "loss": 0.1253, "step": 2340 }, { "grad_norm": 0.8978970646858215, "learning_rate": 5.8725000000000004e-05, "loss": 0.1328, "step": 2350 }, { "grad_norm": 1.1019742488861084, "learning_rate": 5.8975000000000004e-05, "loss": 0.1511, "step": 2360 }, { "grad_norm": 0.7421892881393433, "learning_rate": 5.922500000000001e-05, "loss": 0.1407, "step": 2370 }, { "grad_norm": 0.9020371437072754, "learning_rate": 5.9475e-05, "loss": 0.1352, "step": 2380 }, { "grad_norm": 0.8978645205497742, "learning_rate": 5.9724999999999995e-05, "loss": 0.1484, "step": 2390 }, { "grad_norm": 0.7831639647483826, "learning_rate": 5.9975e-05, "loss": 0.1358, "step": 2400 }, { "grad_norm": 0.9385311007499695, "learning_rate": 6.0225e-05, "loss": 0.1271, "step": 2410 }, { "grad_norm": 0.7526076436042786, "learning_rate": 6.0475000000000006e-05, "loss": 0.161, "step": 2420 }, { "grad_norm": 0.8688762784004211, "learning_rate": 6.0725e-05, "loss": 0.1282, "step": 2430 }, { "grad_norm": 1.2067325115203857, "learning_rate": 6.0975000000000005e-05, "loss": 0.1462, "step": 2440 }, { "grad_norm": 1.0242230892181396, "learning_rate": 6.1225e-05, "loss": 0.1216, "step": 2450 }, { "grad_norm": 0.8039762377738953, "learning_rate": 6.1475e-05, "loss": 0.1086, "step": 2460 }, { "grad_norm": 0.9407244920730591, "learning_rate": 6.1725e-05, "loss": 0.1231, "step": 2470 }, { "grad_norm": 0.8426544666290283, "learning_rate": 6.1975e-05, "loss": 0.1339, "step": 2480 }, { "grad_norm": 0.9156101942062378, "learning_rate": 6.2225e-05, "loss": 0.1274, "step": 2490 }, { "grad_norm": 0.9216020703315735, "learning_rate": 6.2475e-05, "loss": 0.118, "step": 2500 }, { "grad_norm": 0.8171583414077759, "learning_rate": 6.2725e-05, "loss": 0.1081, "step": 2510 }, { "grad_norm": 0.908073365688324, "learning_rate": 6.297500000000001e-05, "loss": 0.1214, "step": 2520 }, { "grad_norm": 0.8049786686897278, "learning_rate": 6.3225e-05, "loss": 0.1121, "step": 2530 }, { "grad_norm": 0.980768084526062, "learning_rate": 6.347500000000001e-05, "loss": 0.116, "step": 2540 }, { "grad_norm": 0.7929062843322754, "learning_rate": 6.3725e-05, "loss": 0.1233, "step": 2550 }, { "grad_norm": 1.064815878868103, "learning_rate": 6.397500000000001e-05, "loss": 0.1292, "step": 2560 }, { "grad_norm": 1.0609278678894043, "learning_rate": 6.4225e-05, "loss": 0.1198, "step": 2570 }, { "grad_norm": 0.8754685521125793, "learning_rate": 6.447500000000001e-05, "loss": 0.1232, "step": 2580 }, { "grad_norm": 0.6940460801124573, "learning_rate": 6.4725e-05, "loss": 0.1154, "step": 2590 }, { "grad_norm": 1.0712684392929077, "learning_rate": 6.497500000000001e-05, "loss": 0.1253, "step": 2600 }, { "grad_norm": 0.8746603727340698, "learning_rate": 6.5225e-05, "loss": 0.1324, "step": 2610 }, { "grad_norm": 0.7426412105560303, "learning_rate": 6.5475e-05, "loss": 0.1009, "step": 2620 }, { "grad_norm": 0.8163965940475464, "learning_rate": 6.5725e-05, "loss": 0.1074, "step": 2630 }, { "grad_norm": 0.7868655920028687, "learning_rate": 6.5975e-05, "loss": 0.1339, "step": 2640 }, { "grad_norm": 0.9293838143348694, "learning_rate": 6.6225e-05, "loss": 0.1362, "step": 2650 }, { "grad_norm": 1.0534380674362183, "learning_rate": 6.6475e-05, "loss": 0.1256, "step": 2660 }, { "grad_norm": 0.8612624406814575, "learning_rate": 6.672500000000001e-05, "loss": 0.1453, "step": 2670 }, { "grad_norm": 1.3833494186401367, "learning_rate": 6.6975e-05, "loss": 0.1473, "step": 2680 }, { "grad_norm": 0.7990478873252869, "learning_rate": 6.722500000000001e-05, "loss": 0.1208, "step": 2690 }, { "grad_norm": 1.0897306203842163, "learning_rate": 6.7475e-05, "loss": 0.1089, "step": 2700 }, { "grad_norm": 0.8122950792312622, "learning_rate": 6.7725e-05, "loss": 0.1277, "step": 2710 }, { "grad_norm": 0.591876208782196, "learning_rate": 6.7975e-05, "loss": 0.1346, "step": 2720 }, { "grad_norm": 1.0705283880233765, "learning_rate": 6.8225e-05, "loss": 0.1465, "step": 2730 }, { "grad_norm": 0.7939675450325012, "learning_rate": 6.8475e-05, "loss": 0.1237, "step": 2740 }, { "grad_norm": 1.0062397718429565, "learning_rate": 6.8725e-05, "loss": 0.1356, "step": 2750 }, { "grad_norm": 0.8144045472145081, "learning_rate": 6.8975e-05, "loss": 0.1107, "step": 2760 }, { "grad_norm": 0.846038281917572, "learning_rate": 6.9225e-05, "loss": 0.1198, "step": 2770 }, { "grad_norm": 1.0394920110702515, "learning_rate": 6.9475e-05, "loss": 0.1357, "step": 2780 }, { "grad_norm": 0.8017951846122742, "learning_rate": 6.9725e-05, "loss": 0.1298, "step": 2790 }, { "grad_norm": 0.7662090063095093, "learning_rate": 6.997500000000001e-05, "loss": 0.1242, "step": 2800 }, { "grad_norm": 0.881216287612915, "learning_rate": 7.022500000000001e-05, "loss": 0.152, "step": 2810 }, { "grad_norm": 1.1736364364624023, "learning_rate": 7.0475e-05, "loss": 0.1271, "step": 2820 }, { "grad_norm": 0.8593667149543762, "learning_rate": 7.072500000000001e-05, "loss": 0.116, "step": 2830 }, { "grad_norm": 0.8163681626319885, "learning_rate": 7.0975e-05, "loss": 0.1316, "step": 2840 }, { "grad_norm": 1.0001282691955566, "learning_rate": 7.122500000000001e-05, "loss": 0.1325, "step": 2850 }, { "grad_norm": 0.6653390526771545, "learning_rate": 7.1475e-05, "loss": 0.1171, "step": 2860 }, { "grad_norm": 0.6026877760887146, "learning_rate": 7.172500000000001e-05, "loss": 0.1022, "step": 2870 }, { "grad_norm": 0.771131694316864, "learning_rate": 7.1975e-05, "loss": 0.1057, "step": 2880 }, { "grad_norm": 0.644615888595581, "learning_rate": 7.2225e-05, "loss": 0.1379, "step": 2890 }, { "grad_norm": 0.6703941822052002, "learning_rate": 7.2475e-05, "loss": 0.1287, "step": 2900 }, { "grad_norm": 0.8999572992324829, "learning_rate": 7.272499999999999e-05, "loss": 0.1165, "step": 2910 }, { "grad_norm": 0.845220148563385, "learning_rate": 7.2975e-05, "loss": 0.1216, "step": 2920 }, { "grad_norm": 0.8259590268135071, "learning_rate": 7.3225e-05, "loss": 0.0957, "step": 2930 }, { "grad_norm": 0.8060678839683533, "learning_rate": 7.347500000000001e-05, "loss": 0.104, "step": 2940 }, { "grad_norm": 0.8648831248283386, "learning_rate": 7.3725e-05, "loss": 0.096, "step": 2950 }, { "grad_norm": 0.6988916993141174, "learning_rate": 7.397500000000001e-05, "loss": 0.111, "step": 2960 }, { "grad_norm": 0.8176288604736328, "learning_rate": 7.4225e-05, "loss": 0.1103, "step": 2970 }, { "grad_norm": 0.8183423280715942, "learning_rate": 7.447500000000001e-05, "loss": 0.1207, "step": 2980 }, { "grad_norm": 0.9460701942443848, "learning_rate": 7.4725e-05, "loss": 0.1285, "step": 2990 }, { "grad_norm": 0.6401332020759583, "learning_rate": 7.4975e-05, "loss": 0.1282, "step": 3000 }, { "grad_norm": 0.979728639125824, "learning_rate": 7.5225e-05, "loss": 0.1014, "step": 3010 }, { "grad_norm": 0.838088870048523, "learning_rate": 7.5475e-05, "loss": 0.1108, "step": 3020 }, { "grad_norm": 0.7601118087768555, "learning_rate": 7.5725e-05, "loss": 0.1199, "step": 3030 }, { "grad_norm": 0.9509809017181396, "learning_rate": 7.5975e-05, "loss": 0.1304, "step": 3040 }, { "grad_norm": 0.9746758341789246, "learning_rate": 7.6225e-05, "loss": 0.1324, "step": 3050 }, { "grad_norm": 0.9159213304519653, "learning_rate": 7.6475e-05, "loss": 0.1356, "step": 3060 }, { "grad_norm": 0.9160392880439758, "learning_rate": 7.672500000000001e-05, "loss": 0.1091, "step": 3070 }, { "grad_norm": 0.7965018153190613, "learning_rate": 7.697500000000001e-05, "loss": 0.1248, "step": 3080 }, { "grad_norm": 1.1037979125976562, "learning_rate": 7.722500000000001e-05, "loss": 0.1388, "step": 3090 }, { "grad_norm": 0.8719176054000854, "learning_rate": 7.747500000000001e-05, "loss": 0.1165, "step": 3100 }, { "grad_norm": 0.6693629026412964, "learning_rate": 7.7725e-05, "loss": 0.0984, "step": 3110 }, { "grad_norm": 0.6572700142860413, "learning_rate": 7.797500000000001e-05, "loss": 0.1089, "step": 3120 }, { "grad_norm": 1.0161715745925903, "learning_rate": 7.8225e-05, "loss": 0.1087, "step": 3130 }, { "grad_norm": 0.6955544948577881, "learning_rate": 7.8475e-05, "loss": 0.1198, "step": 3140 }, { "grad_norm": 0.5939022898674011, "learning_rate": 7.8725e-05, "loss": 0.1224, "step": 3150 }, { "grad_norm": 1.0122692584991455, "learning_rate": 7.8975e-05, "loss": 0.1273, "step": 3160 }, { "grad_norm": 0.7815468311309814, "learning_rate": 7.9225e-05, "loss": 0.1265, "step": 3170 }, { "grad_norm": 0.7099046111106873, "learning_rate": 7.9475e-05, "loss": 0.1374, "step": 3180 }, { "grad_norm": 0.8123295307159424, "learning_rate": 7.9725e-05, "loss": 0.1333, "step": 3190 }, { "grad_norm": 0.9019405841827393, "learning_rate": 7.9975e-05, "loss": 0.125, "step": 3200 }, { "grad_norm": 0.56810462474823, "learning_rate": 8.022500000000001e-05, "loss": 0.1127, "step": 3210 }, { "grad_norm": 1.0996744632720947, "learning_rate": 8.0475e-05, "loss": 0.1454, "step": 3220 }, { "grad_norm": 0.7609489560127258, "learning_rate": 8.072500000000001e-05, "loss": 0.1219, "step": 3230 }, { "grad_norm": 0.809726893901825, "learning_rate": 8.0975e-05, "loss": 0.1013, "step": 3240 }, { "grad_norm": 0.7247797846794128, "learning_rate": 8.122500000000001e-05, "loss": 0.121, "step": 3250 }, { "grad_norm": 0.8948407173156738, "learning_rate": 8.1475e-05, "loss": 0.1187, "step": 3260 }, { "grad_norm": 0.9433281421661377, "learning_rate": 8.172500000000001e-05, "loss": 0.1446, "step": 3270 }, { "grad_norm": 0.874411940574646, "learning_rate": 8.1975e-05, "loss": 0.1261, "step": 3280 }, { "grad_norm": 0.630778431892395, "learning_rate": 8.2225e-05, "loss": 0.1127, "step": 3290 }, { "grad_norm": 0.7029404640197754, "learning_rate": 8.2475e-05, "loss": 0.1134, "step": 3300 }, { "grad_norm": 0.7218849062919617, "learning_rate": 8.2725e-05, "loss": 0.1204, "step": 3310 }, { "grad_norm": 0.7443010210990906, "learning_rate": 8.2975e-05, "loss": 0.1414, "step": 3320 }, { "grad_norm": 1.136860966682434, "learning_rate": 8.3225e-05, "loss": 0.1072, "step": 3330 }, { "grad_norm": 0.9608885049819946, "learning_rate": 8.347500000000001e-05, "loss": 0.1218, "step": 3340 }, { "grad_norm": 0.6671708226203918, "learning_rate": 8.3725e-05, "loss": 0.1098, "step": 3350 }, { "grad_norm": 0.8059146404266357, "learning_rate": 8.397500000000001e-05, "loss": 0.1192, "step": 3360 }, { "grad_norm": 0.8038543462753296, "learning_rate": 8.422500000000001e-05, "loss": 0.1198, "step": 3370 }, { "grad_norm": 1.0045106410980225, "learning_rate": 8.447500000000001e-05, "loss": 0.1597, "step": 3380 }, { "grad_norm": 0.9004790186882019, "learning_rate": 8.4725e-05, "loss": 0.1202, "step": 3390 }, { "grad_norm": 0.7346392869949341, "learning_rate": 8.4975e-05, "loss": 0.1215, "step": 3400 }, { "grad_norm": 0.7080278396606445, "learning_rate": 8.5225e-05, "loss": 0.1255, "step": 3410 }, { "grad_norm": 0.7755569815635681, "learning_rate": 8.5475e-05, "loss": 0.1072, "step": 3420 }, { "grad_norm": 0.6630240082740784, "learning_rate": 8.5725e-05, "loss": 0.119, "step": 3430 }, { "grad_norm": 0.7099795937538147, "learning_rate": 8.5975e-05, "loss": 0.123, "step": 3440 }, { "grad_norm": 0.8256115913391113, "learning_rate": 8.6225e-05, "loss": 0.1168, "step": 3450 }, { "grad_norm": 0.6169359683990479, "learning_rate": 8.6475e-05, "loss": 0.1095, "step": 3460 }, { "grad_norm": 0.718407392501831, "learning_rate": 8.672500000000001e-05, "loss": 0.1339, "step": 3470 }, { "grad_norm": 0.840828001499176, "learning_rate": 8.6975e-05, "loss": 0.1189, "step": 3480 }, { "grad_norm": 0.7176337242126465, "learning_rate": 8.7225e-05, "loss": 0.1244, "step": 3490 }, { "grad_norm": 0.661929190158844, "learning_rate": 8.747500000000001e-05, "loss": 0.0932, "step": 3500 }, { "grad_norm": 0.701177716255188, "learning_rate": 8.7725e-05, "loss": 0.0954, "step": 3510 }, { "grad_norm": 0.7534188628196716, "learning_rate": 8.797500000000001e-05, "loss": 0.1073, "step": 3520 }, { "grad_norm": 0.8635209798812866, "learning_rate": 8.8225e-05, "loss": 0.1146, "step": 3530 }, { "grad_norm": 0.8183215260505676, "learning_rate": 8.847500000000001e-05, "loss": 0.1117, "step": 3540 }, { "grad_norm": 0.7120886445045471, "learning_rate": 8.8725e-05, "loss": 0.1116, "step": 3550 }, { "grad_norm": 0.710591733455658, "learning_rate": 8.897500000000001e-05, "loss": 0.0935, "step": 3560 }, { "grad_norm": 0.671069860458374, "learning_rate": 8.9225e-05, "loss": 0.1198, "step": 3570 }, { "grad_norm": 0.6980414390563965, "learning_rate": 8.9475e-05, "loss": 0.1426, "step": 3580 }, { "grad_norm": 0.8670065402984619, "learning_rate": 8.9725e-05, "loss": 0.131, "step": 3590 }, { "grad_norm": 0.6096193790435791, "learning_rate": 8.9975e-05, "loss": 0.1116, "step": 3600 }, { "grad_norm": 0.8352237939834595, "learning_rate": 9.0225e-05, "loss": 0.116, "step": 3610 }, { "grad_norm": 0.613128662109375, "learning_rate": 9.0475e-05, "loss": 0.1096, "step": 3620 }, { "grad_norm": 0.6533567309379578, "learning_rate": 9.072500000000001e-05, "loss": 0.1023, "step": 3630 }, { "grad_norm": 0.5594480037689209, "learning_rate": 9.0975e-05, "loss": 0.1082, "step": 3640 }, { "grad_norm": 0.7117631435394287, "learning_rate": 9.122500000000001e-05, "loss": 0.1177, "step": 3650 }, { "grad_norm": 0.7305066585540771, "learning_rate": 9.1475e-05, "loss": 0.1125, "step": 3660 }, { "grad_norm": 0.7584351897239685, "learning_rate": 9.172500000000001e-05, "loss": 0.1117, "step": 3670 }, { "grad_norm": 0.707732617855072, "learning_rate": 9.1975e-05, "loss": 0.1133, "step": 3680 }, { "grad_norm": 0.7347385883331299, "learning_rate": 9.2225e-05, "loss": 0.1053, "step": 3690 }, { "grad_norm": 0.6765012145042419, "learning_rate": 9.2475e-05, "loss": 0.1104, "step": 3700 }, { "grad_norm": 0.6229556798934937, "learning_rate": 9.2725e-05, "loss": 0.0966, "step": 3710 }, { "grad_norm": 0.6182796955108643, "learning_rate": 9.2975e-05, "loss": 0.1143, "step": 3720 }, { "grad_norm": 0.5133802890777588, "learning_rate": 9.3225e-05, "loss": 0.0955, "step": 3730 }, { "grad_norm": 0.642238199710846, "learning_rate": 9.3475e-05, "loss": 0.1048, "step": 3740 }, { "grad_norm": 0.7992562055587769, "learning_rate": 9.3725e-05, "loss": 0.1165, "step": 3750 }, { "grad_norm": 0.7050136923789978, "learning_rate": 9.397500000000001e-05, "loss": 0.1109, "step": 3760 }, { "grad_norm": 0.7514626979827881, "learning_rate": 9.422500000000001e-05, "loss": 0.1191, "step": 3770 }, { "grad_norm": 0.6311143636703491, "learning_rate": 9.4475e-05, "loss": 0.1102, "step": 3780 }, { "grad_norm": 0.7020630836486816, "learning_rate": 9.472500000000001e-05, "loss": 0.098, "step": 3790 }, { "grad_norm": 0.8160253763198853, "learning_rate": 9.4975e-05, "loss": 0.1342, "step": 3800 }, { "grad_norm": 0.5818893313407898, "learning_rate": 9.522500000000001e-05, "loss": 0.0916, "step": 3810 }, { "grad_norm": 0.7239099740982056, "learning_rate": 9.5475e-05, "loss": 0.1145, "step": 3820 }, { "grad_norm": 0.7238396406173706, "learning_rate": 9.572500000000001e-05, "loss": 0.0916, "step": 3830 }, { "grad_norm": 0.5896762609481812, "learning_rate": 9.5975e-05, "loss": 0.099, "step": 3840 }, { "grad_norm": 0.7232525944709778, "learning_rate": 9.622500000000001e-05, "loss": 0.1184, "step": 3850 }, { "grad_norm": 0.859589159488678, "learning_rate": 9.6475e-05, "loss": 0.1141, "step": 3860 }, { "grad_norm": 0.7538495063781738, "learning_rate": 9.6725e-05, "loss": 0.1187, "step": 3870 }, { "grad_norm": 0.8234988451004028, "learning_rate": 9.6975e-05, "loss": 0.1024, "step": 3880 }, { "grad_norm": 0.7850862145423889, "learning_rate": 9.7225e-05, "loss": 0.1182, "step": 3890 }, { "grad_norm": 0.7575711011886597, "learning_rate": 9.747500000000001e-05, "loss": 0.1095, "step": 3900 }, { "grad_norm": 0.7172124981880188, "learning_rate": 9.7725e-05, "loss": 0.1163, "step": 3910 }, { "grad_norm": 0.6803779602050781, "learning_rate": 9.797500000000001e-05, "loss": 0.1174, "step": 3920 }, { "grad_norm": 0.7377690672874451, "learning_rate": 9.8225e-05, "loss": 0.103, "step": 3930 }, { "grad_norm": 0.6908969879150391, "learning_rate": 9.847500000000001e-05, "loss": 0.1173, "step": 3940 }, { "grad_norm": 0.9406981468200684, "learning_rate": 9.8725e-05, "loss": 0.1152, "step": 3950 }, { "grad_norm": 0.8612977862358093, "learning_rate": 9.897500000000001e-05, "loss": 0.1325, "step": 3960 }, { "grad_norm": 0.6163735389709473, "learning_rate": 9.9225e-05, "loss": 0.1059, "step": 3970 }, { "grad_norm": 1.0496838092803955, "learning_rate": 9.9475e-05, "loss": 0.1165, "step": 3980 }, { "grad_norm": 0.9000345468521118, "learning_rate": 9.9725e-05, "loss": 0.1213, "step": 3990 }, { "grad_norm": 1.189745545387268, "learning_rate": 9.9975e-05, "loss": 0.1195, "step": 4000 }, { "grad_norm": 0.8290432095527649, "learning_rate": 9.999999653982884e-05, "loss": 0.1106, "step": 4010 }, { "grad_norm": 0.6239068508148193, "learning_rate": 9.999998457874392e-05, "loss": 0.1106, "step": 4020 }, { "grad_norm": 0.7563857436180115, "learning_rate": 9.999996407402913e-05, "loss": 0.1183, "step": 4030 }, { "grad_norm": 1.0019500255584717, "learning_rate": 9.999993502568801e-05, "loss": 0.108, "step": 4040 }, { "grad_norm": 0.7458301782608032, "learning_rate": 9.999989743372548e-05, "loss": 0.1148, "step": 4050 }, { "grad_norm": 0.6722531914710999, "learning_rate": 9.999985129814798e-05, "loss": 0.1202, "step": 4060 }, { "grad_norm": 0.6000438928604126, "learning_rate": 9.99997966189634e-05, "loss": 0.1033, "step": 4070 }, { "grad_norm": 0.8481676578521729, "learning_rate": 9.999973339618107e-05, "loss": 0.1068, "step": 4080 }, { "grad_norm": 0.8255953788757324, "learning_rate": 9.999966162981179e-05, "loss": 0.1228, "step": 4090 }, { "grad_norm": 0.5788481831550598, "learning_rate": 9.999958131986784e-05, "loss": 0.1107, "step": 4100 }, { "grad_norm": 0.771466076374054, "learning_rate": 9.999949246636293e-05, "loss": 0.1227, "step": 4110 }, { "grad_norm": 0.7473097443580627, "learning_rate": 9.999939506931224e-05, "loss": 0.1176, "step": 4120 }, { "grad_norm": 0.7939144968986511, "learning_rate": 9.999928912873243e-05, "loss": 0.1105, "step": 4130 }, { "grad_norm": 0.6243991851806641, "learning_rate": 9.999917464464159e-05, "loss": 0.1267, "step": 4140 }, { "grad_norm": 0.6433641314506531, "learning_rate": 9.999905161705929e-05, "loss": 0.1252, "step": 4150 }, { "grad_norm": 0.747241735458374, "learning_rate": 9.999892004600653e-05, "loss": 0.0938, "step": 4160 }, { "grad_norm": 0.6264744997024536, "learning_rate": 9.999877993150581e-05, "loss": 0.0989, "step": 4170 }, { "grad_norm": 0.6145845651626587, "learning_rate": 9.999863127358108e-05, "loss": 0.0983, "step": 4180 }, { "grad_norm": 0.7490112781524658, "learning_rate": 9.999847407225773e-05, "loss": 0.105, "step": 4190 }, { "grad_norm": 0.7576484084129333, "learning_rate": 9.999830832756262e-05, "loss": 0.109, "step": 4200 }, { "grad_norm": 0.8437197804450989, "learning_rate": 9.999813403952407e-05, "loss": 0.1148, "step": 4210 }, { "grad_norm": 0.6621361970901489, "learning_rate": 9.999795120817187e-05, "loss": 0.1156, "step": 4220 }, { "grad_norm": 0.7657190561294556, "learning_rate": 9.999775983353725e-05, "loss": 0.1234, "step": 4230 }, { "grad_norm": 0.6255114674568176, "learning_rate": 9.999755991565292e-05, "loss": 0.1202, "step": 4240 }, { "grad_norm": 0.8662785291671753, "learning_rate": 9.999735145455303e-05, "loss": 0.123, "step": 4250 }, { "grad_norm": 0.6922198534011841, "learning_rate": 9.99971344502732e-05, "loss": 0.1297, "step": 4260 }, { "grad_norm": 0.5280547738075256, "learning_rate": 9.999690890285053e-05, "loss": 0.1099, "step": 4270 }, { "grad_norm": 0.8421913385391235, "learning_rate": 9.999667481232356e-05, "loss": 0.1136, "step": 4280 }, { "grad_norm": 0.73679518699646, "learning_rate": 9.999643217873225e-05, "loss": 0.1071, "step": 4290 }, { "grad_norm": 0.6809839010238647, "learning_rate": 9.999618100211809e-05, "loss": 0.0885, "step": 4300 }, { "grad_norm": 0.6690009236335754, "learning_rate": 9.999592128252402e-05, "loss": 0.1017, "step": 4310 }, { "grad_norm": 0.6766150593757629, "learning_rate": 9.999565301999437e-05, "loss": 0.132, "step": 4320 }, { "grad_norm": 0.8231752514839172, "learning_rate": 9.999537621457502e-05, "loss": 0.1019, "step": 4330 }, { "grad_norm": 0.527861475944519, "learning_rate": 9.999509086631323e-05, "loss": 0.086, "step": 4340 }, { "grad_norm": 1.0153321027755737, "learning_rate": 9.99947969752578e-05, "loss": 0.0832, "step": 4350 }, { "grad_norm": 0.8659564256668091, "learning_rate": 9.999449454145891e-05, "loss": 0.0888, "step": 4360 }, { "grad_norm": 0.6440706253051758, "learning_rate": 9.999418356496827e-05, "loss": 0.102, "step": 4370 }, { "grad_norm": 0.6388813853263855, "learning_rate": 9.999386404583899e-05, "loss": 0.11, "step": 4380 }, { "grad_norm": 0.6528149843215942, "learning_rate": 9.999353598412568e-05, "loss": 0.1024, "step": 4390 }, { "grad_norm": 0.7883272171020508, "learning_rate": 9.999319937988442e-05, "loss": 0.127, "step": 4400 }, { "grad_norm": 0.5105001926422119, "learning_rate": 9.999285423317268e-05, "loss": 0.1121, "step": 4410 }, { "grad_norm": 0.5447722673416138, "learning_rate": 9.999250054404947e-05, "loss": 0.1207, "step": 4420 }, { "grad_norm": 0.7324037551879883, "learning_rate": 9.99921383125752e-05, "loss": 0.1078, "step": 4430 }, { "grad_norm": 0.696150004863739, "learning_rate": 9.99917675388118e-05, "loss": 0.1122, "step": 4440 }, { "grad_norm": 0.5488695502281189, "learning_rate": 9.99913882228226e-05, "loss": 0.1196, "step": 4450 }, { "grad_norm": 0.5755054354667664, "learning_rate": 9.999100036467242e-05, "loss": 0.0991, "step": 4460 }, { "grad_norm": 0.6267532706260681, "learning_rate": 9.999060396442753e-05, "loss": 0.105, "step": 4470 }, { "grad_norm": 0.5897603034973145, "learning_rate": 9.999019902215566e-05, "loss": 0.1046, "step": 4480 }, { "grad_norm": 0.7329015135765076, "learning_rate": 9.998978553792602e-05, "loss": 0.1246, "step": 4490 }, { "grad_norm": 0.5700187683105469, "learning_rate": 9.998936351180926e-05, "loss": 0.1083, "step": 4500 }, { "grad_norm": 0.6982454061508179, "learning_rate": 9.998893294387747e-05, "loss": 0.1358, "step": 4510 }, { "grad_norm": 0.6194480657577515, "learning_rate": 9.998849383420426e-05, "loss": 0.1093, "step": 4520 }, { "grad_norm": 0.4859570562839508, "learning_rate": 9.998804618286465e-05, "loss": 0.1051, "step": 4530 }, { "grad_norm": 0.8402063846588135, "learning_rate": 9.99875899899351e-05, "loss": 0.1228, "step": 4540 }, { "grad_norm": 0.51856529712677, "learning_rate": 9.99871252554936e-05, "loss": 0.114, "step": 4550 }, { "grad_norm": 0.6954576373100281, "learning_rate": 9.998665197961955e-05, "loss": 0.107, "step": 4560 }, { "grad_norm": 0.9151623845100403, "learning_rate": 9.998617016239379e-05, "loss": 0.1146, "step": 4570 }, { "grad_norm": 0.9313297867774963, "learning_rate": 9.998567980389869e-05, "loss": 0.1039, "step": 4580 }, { "grad_norm": 0.6682326793670654, "learning_rate": 9.998518090421802e-05, "loss": 0.1058, "step": 4590 }, { "grad_norm": 0.7727391719818115, "learning_rate": 9.998467346343703e-05, "loss": 0.104, "step": 4600 }, { "grad_norm": 0.7997540235519409, "learning_rate": 9.998415748164243e-05, "loss": 0.1059, "step": 4610 }, { "grad_norm": 0.649812638759613, "learning_rate": 9.998363295892238e-05, "loss": 0.0954, "step": 4620 }, { "grad_norm": 0.7376749515533447, "learning_rate": 9.998309989536652e-05, "loss": 0.1197, "step": 4630 }, { "grad_norm": 0.7037510275840759, "learning_rate": 9.998255829106593e-05, "loss": 0.1205, "step": 4640 }, { "grad_norm": 0.5851507186889648, "learning_rate": 9.998200814611316e-05, "loss": 0.1062, "step": 4650 }, { "grad_norm": 0.6140917539596558, "learning_rate": 9.998144946060219e-05, "loss": 0.1071, "step": 4660 }, { "grad_norm": 0.8016238808631897, "learning_rate": 9.998088223462852e-05, "loss": 0.1024, "step": 4670 }, { "grad_norm": 0.48748600482940674, "learning_rate": 9.998030646828905e-05, "loss": 0.0966, "step": 4680 }, { "grad_norm": 0.5762995481491089, "learning_rate": 9.997972216168217e-05, "loss": 0.1075, "step": 4690 }, { "grad_norm": 0.6107780933380127, "learning_rate": 9.997912931490771e-05, "loss": 0.1112, "step": 4700 }, { "grad_norm": 0.6129292249679565, "learning_rate": 9.9978527928067e-05, "loss": 0.1084, "step": 4710 }, { "grad_norm": 0.5316489934921265, "learning_rate": 9.997791800126277e-05, "loss": 0.1002, "step": 4720 }, { "grad_norm": 0.7813099026679993, "learning_rate": 9.997729953459927e-05, "loss": 0.1152, "step": 4730 }, { "grad_norm": 0.622029185295105, "learning_rate": 9.997667252818214e-05, "loss": 0.1088, "step": 4740 }, { "grad_norm": 0.5464494228363037, "learning_rate": 9.997603698211855e-05, "loss": 0.0964, "step": 4750 }, { "grad_norm": 0.8614035248756409, "learning_rate": 9.99753928965171e-05, "loss": 0.1067, "step": 4760 }, { "grad_norm": 0.7516201138496399, "learning_rate": 9.997474027148781e-05, "loss": 0.1162, "step": 4770 }, { "grad_norm": 0.7478755116462708, "learning_rate": 9.997407910714223e-05, "loss": 0.1287, "step": 4780 }, { "grad_norm": 0.7140412330627441, "learning_rate": 9.997340940359332e-05, "loss": 0.1037, "step": 4790 }, { "grad_norm": 0.8317935466766357, "learning_rate": 9.997273116095552e-05, "loss": 0.113, "step": 4800 }, { "grad_norm": 0.6904240846633911, "learning_rate": 9.997204437934473e-05, "loss": 0.1161, "step": 4810 }, { "grad_norm": 0.9802899360656738, "learning_rate": 9.997134905887829e-05, "loss": 0.1076, "step": 4820 }, { "grad_norm": 0.4730854332447052, "learning_rate": 9.997064519967501e-05, "loss": 0.1059, "step": 4830 }, { "grad_norm": 0.7189297080039978, "learning_rate": 9.996993280185517e-05, "loss": 0.1149, "step": 4840 }, { "grad_norm": 0.8492293357849121, "learning_rate": 9.99692118655405e-05, "loss": 0.0973, "step": 4850 }, { "grad_norm": 0.6474923491477966, "learning_rate": 9.996848239085417e-05, "loss": 0.0933, "step": 4860 }, { "grad_norm": 0.6497602462768555, "learning_rate": 9.996774437792085e-05, "loss": 0.0947, "step": 4870 }, { "grad_norm": 0.5829052925109863, "learning_rate": 9.996699782686664e-05, "loss": 0.1039, "step": 4880 }, { "grad_norm": 0.9635458588600159, "learning_rate": 9.996624273781909e-05, "loss": 0.1299, "step": 4890 }, { "grad_norm": 0.7141345143318176, "learning_rate": 9.996547911090725e-05, "loss": 0.1046, "step": 4900 }, { "grad_norm": 0.6724671721458435, "learning_rate": 9.996470694626157e-05, "loss": 0.1149, "step": 4910 }, { "grad_norm": 0.7165585160255432, "learning_rate": 9.996392624401403e-05, "loss": 0.1111, "step": 4920 }, { "grad_norm": 0.5461900234222412, "learning_rate": 9.996313700429801e-05, "loss": 0.1044, "step": 4930 }, { "grad_norm": 0.6964077949523926, "learning_rate": 9.996233922724836e-05, "loss": 0.0905, "step": 4940 }, { "grad_norm": 0.6053014993667603, "learning_rate": 9.996153291300141e-05, "loss": 0.1033, "step": 4950 }, { "grad_norm": 0.6696963906288147, "learning_rate": 9.996071806169494e-05, "loss": 0.0998, "step": 4960 }, { "grad_norm": 0.5914113521575928, "learning_rate": 9.995989467346817e-05, "loss": 0.1078, "step": 4970 }, { "grad_norm": 0.5314283967018127, "learning_rate": 9.995906274846183e-05, "loss": 0.099, "step": 4980 }, { "grad_norm": 0.5122625231742859, "learning_rate": 9.995822228681803e-05, "loss": 0.0892, "step": 4990 }, { "grad_norm": 0.7496462464332581, "learning_rate": 9.99573732886804e-05, "loss": 0.104, "step": 5000 }, { "grad_norm": 0.9021585583686829, "learning_rate": 9.995651575419402e-05, "loss": 0.1025, "step": 5010 }, { "grad_norm": 0.7104624509811401, "learning_rate": 9.995564968350541e-05, "loss": 0.1033, "step": 5020 }, { "grad_norm": 0.6784711480140686, "learning_rate": 9.995477507676256e-05, "loss": 0.0949, "step": 5030 }, { "grad_norm": 0.532785177230835, "learning_rate": 9.995389193411493e-05, "loss": 0.0931, "step": 5040 }, { "grad_norm": 0.5759444832801819, "learning_rate": 9.995300025571339e-05, "loss": 0.1134, "step": 5050 }, { "grad_norm": 0.5733329057693481, "learning_rate": 9.995210004171034e-05, "loss": 0.0789, "step": 5060 }, { "grad_norm": 0.5392467379570007, "learning_rate": 9.995119129225956e-05, "loss": 0.1218, "step": 5070 }, { "grad_norm": 0.5834360718727112, "learning_rate": 9.995027400751637e-05, "loss": 0.1177, "step": 5080 }, { "grad_norm": 0.5406858921051025, "learning_rate": 9.994934818763751e-05, "loss": 0.0976, "step": 5090 }, { "grad_norm": 0.6180983185768127, "learning_rate": 9.994841383278115e-05, "loss": 0.0983, "step": 5100 }, { "grad_norm": 0.5544050931930542, "learning_rate": 9.994747094310695e-05, "loss": 0.1047, "step": 5110 }, { "grad_norm": 0.6858172416687012, "learning_rate": 9.994651951877604e-05, "loss": 0.1057, "step": 5120 }, { "grad_norm": 0.7371190786361694, "learning_rate": 9.994555955995099e-05, "loss": 0.1109, "step": 5130 }, { "grad_norm": 0.6791462898254395, "learning_rate": 9.994459106679581e-05, "loss": 0.1047, "step": 5140 }, { "grad_norm": 0.6707594990730286, "learning_rate": 9.994361403947603e-05, "loss": 0.101, "step": 5150 }, { "grad_norm": 0.6268891096115112, "learning_rate": 9.994262847815854e-05, "loss": 0.1213, "step": 5160 }, { "grad_norm": 0.6242362856864929, "learning_rate": 9.99416343830118e-05, "loss": 0.1209, "step": 5170 }, { "grad_norm": 0.8102626800537109, "learning_rate": 9.994063175420565e-05, "loss": 0.0972, "step": 5180 }, { "grad_norm": 0.6818050742149353, "learning_rate": 9.99396205919114e-05, "loss": 0.1157, "step": 5190 }, { "grad_norm": 0.7655032873153687, "learning_rate": 9.993860089630185e-05, "loss": 0.1087, "step": 5200 }, { "grad_norm": 0.7191789150238037, "learning_rate": 9.993757266755123e-05, "loss": 0.1093, "step": 5210 }, { "grad_norm": 0.6805552244186401, "learning_rate": 9.993653590583522e-05, "loss": 0.0934, "step": 5220 }, { "grad_norm": 0.6296294927597046, "learning_rate": 9.993549061133102e-05, "loss": 0.0943, "step": 5230 }, { "grad_norm": 0.5743198394775391, "learning_rate": 9.993443678421719e-05, "loss": 0.1012, "step": 5240 }, { "grad_norm": 0.5285204648971558, "learning_rate": 9.993337442467384e-05, "loss": 0.1045, "step": 5250 }, { "grad_norm": 0.8474060297012329, "learning_rate": 9.993230353288248e-05, "loss": 0.0914, "step": 5260 }, { "grad_norm": 0.8477046489715576, "learning_rate": 9.993122410902608e-05, "loss": 0.0893, "step": 5270 }, { "grad_norm": 0.6239337921142578, "learning_rate": 9.993013615328912e-05, "loss": 0.1174, "step": 5280 }, { "grad_norm": 0.5622395873069763, "learning_rate": 9.992903966585747e-05, "loss": 0.1126, "step": 5290 }, { "grad_norm": 0.5645528435707092, "learning_rate": 9.992793464691852e-05, "loss": 0.0928, "step": 5300 }, { "grad_norm": 0.813438892364502, "learning_rate": 9.992682109666105e-05, "loss": 0.0886, "step": 5310 }, { "grad_norm": 0.6114284992218018, "learning_rate": 9.992569901527538e-05, "loss": 0.0915, "step": 5320 }, { "grad_norm": 0.9177957773208618, "learning_rate": 9.99245684029532e-05, "loss": 0.1171, "step": 5330 }, { "grad_norm": 0.7367610931396484, "learning_rate": 9.992342925988774e-05, "loss": 0.1021, "step": 5340 }, { "grad_norm": 0.6662881374359131, "learning_rate": 9.992228158627361e-05, "loss": 0.0883, "step": 5350 }, { "grad_norm": 0.6896058320999146, "learning_rate": 9.992112538230693e-05, "loss": 0.0955, "step": 5360 }, { "grad_norm": 0.7369977831840515, "learning_rate": 9.991996064818527e-05, "loss": 0.0989, "step": 5370 }, { "grad_norm": 0.6927738189697266, "learning_rate": 9.991878738410768e-05, "loss": 0.0977, "step": 5380 }, { "grad_norm": 0.5518178343772888, "learning_rate": 9.991760559027457e-05, "loss": 0.088, "step": 5390 }, { "grad_norm": 0.6252905130386353, "learning_rate": 9.991641526688793e-05, "loss": 0.1002, "step": 5400 }, { "grad_norm": 0.5778160095214844, "learning_rate": 9.991521641415113e-05, "loss": 0.0951, "step": 5410 }, { "grad_norm": 0.7380479574203491, "learning_rate": 9.991400903226904e-05, "loss": 0.1032, "step": 5420 }, { "grad_norm": 1.0188647508621216, "learning_rate": 9.991279312144794e-05, "loss": 0.1021, "step": 5430 }, { "grad_norm": 0.7628278136253357, "learning_rate": 9.991156868189564e-05, "loss": 0.1186, "step": 5440 }, { "grad_norm": 0.6884528398513794, "learning_rate": 9.991033571382131e-05, "loss": 0.0789, "step": 5450 }, { "grad_norm": 0.5618193745613098, "learning_rate": 9.990909421743569e-05, "loss": 0.0893, "step": 5460 }, { "grad_norm": 0.6582702398300171, "learning_rate": 9.990784419295085e-05, "loss": 0.0813, "step": 5470 }, { "grad_norm": 0.7913344502449036, "learning_rate": 9.990658564058044e-05, "loss": 0.096, "step": 5480 }, { "grad_norm": 0.5031030774116516, "learning_rate": 9.990531856053948e-05, "loss": 0.0918, "step": 5490 }, { "grad_norm": 0.6904828548431396, "learning_rate": 9.99040429530445e-05, "loss": 0.0934, "step": 5500 }, { "grad_norm": 0.651679277420044, "learning_rate": 9.990275881831346e-05, "loss": 0.1, "step": 5510 }, { "grad_norm": 0.5958465933799744, "learning_rate": 9.990146615656577e-05, "loss": 0.0842, "step": 5520 }, { "grad_norm": 0.6423187255859375, "learning_rate": 9.990016496802233e-05, "loss": 0.1026, "step": 5530 }, { "grad_norm": 0.6572743058204651, "learning_rate": 9.989885525290548e-05, "loss": 0.0903, "step": 5540 }, { "grad_norm": 0.4688767194747925, "learning_rate": 9.989753701143897e-05, "loss": 0.088, "step": 5550 }, { "grad_norm": 0.7518588900566101, "learning_rate": 9.989621024384812e-05, "loss": 0.1018, "step": 5560 }, { "grad_norm": 0.5295987725257874, "learning_rate": 9.989487495035959e-05, "loss": 0.0911, "step": 5570 }, { "grad_norm": 0.6867901682853699, "learning_rate": 9.989353113120156e-05, "loss": 0.1143, "step": 5580 }, { "grad_norm": 0.6029537320137024, "learning_rate": 9.989217878660366e-05, "loss": 0.0923, "step": 5590 }, { "grad_norm": 0.5682321190834045, "learning_rate": 9.989081791679695e-05, "loss": 0.1163, "step": 5600 }, { "grad_norm": 0.5699365735054016, "learning_rate": 9.988944852201397e-05, "loss": 0.098, "step": 5610 }, { "grad_norm": 0.7397676706314087, "learning_rate": 9.988807060248873e-05, "loss": 0.0863, "step": 5620 }, { "grad_norm": 0.5483518242835999, "learning_rate": 9.988668415845665e-05, "loss": 0.0874, "step": 5630 }, { "grad_norm": 0.7115257978439331, "learning_rate": 9.988528919015466e-05, "loss": 0.1123, "step": 5640 }, { "grad_norm": 0.75655597448349, "learning_rate": 9.988388569782112e-05, "loss": 0.1095, "step": 5650 }, { "grad_norm": 0.7113964557647705, "learning_rate": 9.988247368169583e-05, "loss": 0.0969, "step": 5660 }, { "grad_norm": 0.8868502974510193, "learning_rate": 9.988105314202007e-05, "loss": 0.103, "step": 5670 }, { "grad_norm": 0.5336458683013916, "learning_rate": 9.987962407903659e-05, "loss": 0.1017, "step": 5680 }, { "grad_norm": 0.9902915954589844, "learning_rate": 9.987818649298957e-05, "loss": 0.1067, "step": 5690 }, { "grad_norm": 0.5071224570274353, "learning_rate": 9.987674038412465e-05, "loss": 0.0888, "step": 5700 }, { "grad_norm": 0.5737423300743103, "learning_rate": 9.987528575268891e-05, "loss": 0.0962, "step": 5710 }, { "grad_norm": 0.8637117147445679, "learning_rate": 9.987382259893095e-05, "loss": 0.0962, "step": 5720 }, { "grad_norm": 0.5913756489753723, "learning_rate": 9.987235092310074e-05, "loss": 0.088, "step": 5730 }, { "grad_norm": 0.6296379566192627, "learning_rate": 9.987087072544978e-05, "loss": 0.1171, "step": 5740 }, { "grad_norm": 0.5956076383590698, "learning_rate": 9.9869382006231e-05, "loss": 0.1008, "step": 5750 }, { "grad_norm": 0.48491597175598145, "learning_rate": 9.986788476569875e-05, "loss": 0.0967, "step": 5760 }, { "grad_norm": 0.6099057197570801, "learning_rate": 9.986637900410887e-05, "loss": 0.0919, "step": 5770 }, { "grad_norm": 0.5118264555931091, "learning_rate": 9.986486472171869e-05, "loss": 0.1104, "step": 5780 }, { "grad_norm": 0.6093209981918335, "learning_rate": 9.986334191878692e-05, "loss": 0.093, "step": 5790 }, { "grad_norm": 0.8463785648345947, "learning_rate": 9.986181059557378e-05, "loss": 0.1144, "step": 5800 }, { "grad_norm": 0.5724389553070068, "learning_rate": 9.986027075234094e-05, "loss": 0.1009, "step": 5810 }, { "grad_norm": 0.5834295749664307, "learning_rate": 9.985872238935152e-05, "loss": 0.0887, "step": 5820 }, { "grad_norm": 0.554402768611908, "learning_rate": 9.985716550687008e-05, "loss": 0.0957, "step": 5830 }, { "grad_norm": 0.5125986933708191, "learning_rate": 9.985560010516264e-05, "loss": 0.097, "step": 5840 }, { "grad_norm": 0.6222838759422302, "learning_rate": 9.985402618449668e-05, "loss": 0.0909, "step": 5850 }, { "grad_norm": 0.7341822385787964, "learning_rate": 9.985244374514118e-05, "loss": 0.1154, "step": 5860 }, { "grad_norm": 0.5148488283157349, "learning_rate": 9.985085278736651e-05, "loss": 0.1007, "step": 5870 }, { "grad_norm": 0.6820268034934998, "learning_rate": 9.984925331144452e-05, "loss": 0.1037, "step": 5880 }, { "grad_norm": 0.5910977721214294, "learning_rate": 9.984764531764851e-05, "loss": 0.0935, "step": 5890 }, { "grad_norm": 0.5282042622566223, "learning_rate": 9.984602880625326e-05, "loss": 0.109, "step": 5900 }, { "grad_norm": 0.566143810749054, "learning_rate": 9.9844403777535e-05, "loss": 0.1151, "step": 5910 }, { "grad_norm": 0.8231667876243591, "learning_rate": 9.984277023177135e-05, "loss": 0.0969, "step": 5920 }, { "grad_norm": 0.5480079054832458, "learning_rate": 9.984112816924148e-05, "loss": 0.085, "step": 5930 }, { "grad_norm": 0.7398078441619873, "learning_rate": 9.983947759022596e-05, "loss": 0.1076, "step": 5940 }, { "grad_norm": 0.7255097031593323, "learning_rate": 9.983781849500682e-05, "loss": 0.0958, "step": 5950 }, { "grad_norm": 0.5448275804519653, "learning_rate": 9.98361508838676e-05, "loss": 0.0916, "step": 5960 }, { "grad_norm": 0.7087063193321228, "learning_rate": 9.98344747570932e-05, "loss": 0.0931, "step": 5970 }, { "grad_norm": 0.5673488974571228, "learning_rate": 9.983279011497004e-05, "loss": 0.0977, "step": 5980 }, { "grad_norm": 0.8097339868545532, "learning_rate": 9.983109695778596e-05, "loss": 0.1341, "step": 5990 }, { "grad_norm": 0.5805525183677673, "learning_rate": 9.982939528583032e-05, "loss": 0.1251, "step": 6000 }, { "grad_norm": 0.6439293026924133, "learning_rate": 9.982768509939385e-05, "loss": 0.1084, "step": 6010 }, { "grad_norm": 0.5207499861717224, "learning_rate": 9.982596639876879e-05, "loss": 0.0975, "step": 6020 }, { "grad_norm": 0.5603018999099731, "learning_rate": 9.982423918424881e-05, "loss": 0.0953, "step": 6030 }, { "grad_norm": 0.5426331162452698, "learning_rate": 9.982250345612908e-05, "loss": 0.0945, "step": 6040 }, { "grad_norm": 0.6466575860977173, "learning_rate": 9.982075921470611e-05, "loss": 0.0992, "step": 6050 }, { "grad_norm": 0.5233604311943054, "learning_rate": 9.981900646027802e-05, "loss": 0.094, "step": 6060 }, { "grad_norm": 0.6799477338790894, "learning_rate": 9.981724519314425e-05, "loss": 0.0985, "step": 6070 }, { "grad_norm": 0.6114366054534912, "learning_rate": 9.981547541360581e-05, "loss": 0.0947, "step": 6080 }, { "grad_norm": 0.5623548626899719, "learning_rate": 9.981369712196508e-05, "loss": 0.1032, "step": 6090 }, { "grad_norm": 0.6131870150566101, "learning_rate": 9.981191031852592e-05, "loss": 0.1135, "step": 6100 }, { "grad_norm": 0.4588993191719055, "learning_rate": 9.981011500359362e-05, "loss": 0.0894, "step": 6110 }, { "grad_norm": 0.5274778604507446, "learning_rate": 9.9808311177475e-05, "loss": 0.0955, "step": 6120 }, { "grad_norm": 0.6592124104499817, "learning_rate": 9.980649884047826e-05, "loss": 0.0791, "step": 6130 }, { "grad_norm": 0.6108990907669067, "learning_rate": 9.980467799291307e-05, "loss": 0.0952, "step": 6140 }, { "grad_norm": 0.8480209112167358, "learning_rate": 9.980284863509058e-05, "loss": 0.1112, "step": 6150 }, { "grad_norm": 0.7219366431236267, "learning_rate": 9.980101076732334e-05, "loss": 0.1246, "step": 6160 }, { "grad_norm": 0.7213364243507385, "learning_rate": 9.979916438992544e-05, "loss": 0.1063, "step": 6170 }, { "grad_norm": 0.6228355765342712, "learning_rate": 9.979730950321237e-05, "loss": 0.0787, "step": 6180 }, { "grad_norm": 0.4021826386451721, "learning_rate": 9.979544610750104e-05, "loss": 0.0961, "step": 6190 }, { "grad_norm": 0.5130497217178345, "learning_rate": 9.97935742031099e-05, "loss": 0.0853, "step": 6200 }, { "grad_norm": 0.7956061363220215, "learning_rate": 9.979169379035878e-05, "loss": 0.0984, "step": 6210 }, { "grad_norm": 0.6619382500648499, "learning_rate": 9.978980486956899e-05, "loss": 0.0868, "step": 6220 }, { "grad_norm": 0.627573549747467, "learning_rate": 9.978790744106332e-05, "loss": 0.1021, "step": 6230 }, { "grad_norm": 0.6157965064048767, "learning_rate": 9.978600150516594e-05, "loss": 0.1012, "step": 6240 }, { "grad_norm": 0.650635838508606, "learning_rate": 9.978408706220259e-05, "loss": 0.1103, "step": 6250 }, { "grad_norm": 0.614938497543335, "learning_rate": 9.978216411250032e-05, "loss": 0.0891, "step": 6260 }, { "grad_norm": 0.7166763544082642, "learning_rate": 9.978023265638778e-05, "loss": 0.0933, "step": 6270 }, { "grad_norm": 0.6084437370300293, "learning_rate": 9.977829269419495e-05, "loss": 0.097, "step": 6280 }, { "grad_norm": 0.4914869964122772, "learning_rate": 9.977634422625335e-05, "loss": 0.0933, "step": 6290 }, { "grad_norm": 0.5979200601577759, "learning_rate": 9.97743872528959e-05, "loss": 0.0979, "step": 6300 }, { "grad_norm": 0.6101016402244568, "learning_rate": 9.9772421774457e-05, "loss": 0.0925, "step": 6310 }, { "grad_norm": 0.5937826037406921, "learning_rate": 9.977044779127252e-05, "loss": 0.095, "step": 6320 }, { "grad_norm": 0.613064706325531, "learning_rate": 9.976846530367971e-05, "loss": 0.0947, "step": 6330 }, { "grad_norm": 0.6947755217552185, "learning_rate": 9.976647431201735e-05, "loss": 0.1174, "step": 6340 }, { "grad_norm": 0.6457632184028625, "learning_rate": 9.976447481662568e-05, "loss": 0.0945, "step": 6350 }, { "grad_norm": 0.7148882150650024, "learning_rate": 9.976246681784629e-05, "loss": 0.096, "step": 6360 }, { "grad_norm": 0.752363920211792, "learning_rate": 9.976045031602234e-05, "loss": 0.092, "step": 6370 }, { "grad_norm": 0.49512022733688354, "learning_rate": 9.975842531149837e-05, "loss": 0.0927, "step": 6380 }, { "grad_norm": 0.7812385559082031, "learning_rate": 9.975639180462043e-05, "loss": 0.0977, "step": 6390 }, { "grad_norm": 0.65684974193573, "learning_rate": 9.975434979573596e-05, "loss": 0.0936, "step": 6400 }, { "grad_norm": 0.6646395325660706, "learning_rate": 9.97522992851939e-05, "loss": 0.0825, "step": 6410 }, { "grad_norm": 0.8374908566474915, "learning_rate": 9.975024027334461e-05, "loss": 0.0969, "step": 6420 }, { "grad_norm": 0.3972569406032562, "learning_rate": 9.974817276053993e-05, "loss": 0.0971, "step": 6430 }, { "grad_norm": 0.6531705260276794, "learning_rate": 9.974609674713315e-05, "loss": 0.0871, "step": 6440 }, { "grad_norm": 0.6325659155845642, "learning_rate": 9.9744012233479e-05, "loss": 0.0869, "step": 6450 }, { "grad_norm": 0.5731664299964905, "learning_rate": 9.974191921993366e-05, "loss": 0.0849, "step": 6460 }, { "grad_norm": 0.6209680438041687, "learning_rate": 9.973981770685474e-05, "loss": 0.0954, "step": 6470 }, { "grad_norm": 0.5465537905693054, "learning_rate": 9.97377076946014e-05, "loss": 0.0834, "step": 6480 }, { "grad_norm": 0.6094324588775635, "learning_rate": 9.973558918353412e-05, "loss": 0.1028, "step": 6490 }, { "grad_norm": 0.5681731104850769, "learning_rate": 9.973346217401494e-05, "loss": 0.0833, "step": 6500 }, { "grad_norm": 0.4958348870277405, "learning_rate": 9.973132666640726e-05, "loss": 0.0794, "step": 6510 }, { "grad_norm": 0.6007272601127625, "learning_rate": 9.972918266107602e-05, "loss": 0.0853, "step": 6520 }, { "grad_norm": 0.4193439185619354, "learning_rate": 9.972703015838756e-05, "loss": 0.0865, "step": 6530 }, { "grad_norm": 0.6496378779411316, "learning_rate": 9.97248691587097e-05, "loss": 0.097, "step": 6540 }, { "grad_norm": 0.6994186639785767, "learning_rate": 9.972269966241166e-05, "loss": 0.0888, "step": 6550 }, { "grad_norm": 0.5661810040473938, "learning_rate": 9.972052166986417e-05, "loss": 0.1027, "step": 6560 }, { "grad_norm": 0.5628449320793152, "learning_rate": 9.971833518143938e-05, "loss": 0.0954, "step": 6570 }, { "grad_norm": 0.5862028002738953, "learning_rate": 9.971614019751093e-05, "loss": 0.0787, "step": 6580 }, { "grad_norm": 0.5719397068023682, "learning_rate": 9.971393671845383e-05, "loss": 0.1096, "step": 6590 }, { "grad_norm": 0.6450148820877075, "learning_rate": 9.971172474464464e-05, "loss": 0.0988, "step": 6600 }, { "grad_norm": 0.5321563482284546, "learning_rate": 9.97095042764613e-05, "loss": 0.0833, "step": 6610 }, { "grad_norm": 0.6027038097381592, "learning_rate": 9.970727531428324e-05, "loss": 0.0956, "step": 6620 }, { "grad_norm": 0.5361247062683105, "learning_rate": 9.970503785849132e-05, "loss": 0.0896, "step": 6630 }, { "grad_norm": 0.4923218786716461, "learning_rate": 9.970279190946788e-05, "loss": 0.0766, "step": 6640 }, { "grad_norm": 0.6686718463897705, "learning_rate": 9.970053746759667e-05, "loss": 0.0891, "step": 6650 }, { "grad_norm": 0.7254266738891602, "learning_rate": 9.969827453326292e-05, "loss": 0.0818, "step": 6660 }, { "grad_norm": 0.5273278951644897, "learning_rate": 9.969600310685332e-05, "loss": 0.0754, "step": 6670 }, { "grad_norm": 0.545329749584198, "learning_rate": 9.969372318875596e-05, "loss": 0.0995, "step": 6680 }, { "grad_norm": 0.6293640732765198, "learning_rate": 9.969143477936043e-05, "loss": 0.0961, "step": 6690 }, { "grad_norm": 0.6225606799125671, "learning_rate": 9.968913787905775e-05, "loss": 0.0928, "step": 6700 }, { "grad_norm": 0.516428530216217, "learning_rate": 9.968683248824045e-05, "loss": 0.1046, "step": 6710 }, { "grad_norm": 0.6362929940223694, "learning_rate": 9.968451860730238e-05, "loss": 0.0998, "step": 6720 }, { "grad_norm": 0.5609709620475769, "learning_rate": 9.968219623663896e-05, "loss": 0.0902, "step": 6730 }, { "grad_norm": 0.47521764039993286, "learning_rate": 9.967986537664702e-05, "loss": 0.1141, "step": 6740 }, { "grad_norm": 0.668548047542572, "learning_rate": 9.967752602772483e-05, "loss": 0.0842, "step": 6750 }, { "grad_norm": 0.6331639289855957, "learning_rate": 9.967517819027212e-05, "loss": 0.0991, "step": 6760 }, { "grad_norm": 0.5742775797843933, "learning_rate": 9.967282186469009e-05, "loss": 0.0858, "step": 6770 }, { "grad_norm": 0.5693672895431519, "learning_rate": 9.967045705138135e-05, "loss": 0.0898, "step": 6780 }, { "grad_norm": 0.6180541515350342, "learning_rate": 9.966808375074998e-05, "loss": 0.0864, "step": 6790 }, { "grad_norm": 0.6292174458503723, "learning_rate": 9.966570196320154e-05, "loss": 0.0967, "step": 6800 }, { "grad_norm": 0.6301944851875305, "learning_rate": 9.966331168914299e-05, "loss": 0.0686, "step": 6810 }, { "grad_norm": 0.6306625604629517, "learning_rate": 9.966091292898277e-05, "loss": 0.0932, "step": 6820 }, { "grad_norm": 0.6112961769104004, "learning_rate": 9.965850568313076e-05, "loss": 0.0895, "step": 6830 }, { "grad_norm": 0.8086462616920471, "learning_rate": 9.965608995199827e-05, "loss": 0.1052, "step": 6840 }, { "grad_norm": 0.6185763478279114, "learning_rate": 9.965366573599812e-05, "loss": 0.0892, "step": 6850 }, { "grad_norm": 0.5941610336303711, "learning_rate": 9.965123303554453e-05, "loss": 0.0925, "step": 6860 }, { "grad_norm": 0.7406275272369385, "learning_rate": 9.964879185105317e-05, "loss": 0.1087, "step": 6870 }, { "grad_norm": 0.9753264784812927, "learning_rate": 9.964634218294119e-05, "loss": 0.1027, "step": 6880 }, { "grad_norm": 0.6223640441894531, "learning_rate": 9.964388403162714e-05, "loss": 0.0926, "step": 6890 }, { "grad_norm": 0.8885828852653503, "learning_rate": 9.96414173975311e-05, "loss": 0.0995, "step": 6900 }, { "grad_norm": 0.4605362117290497, "learning_rate": 9.963894228107451e-05, "loss": 0.084, "step": 6910 }, { "grad_norm": 0.5707457661628723, "learning_rate": 9.963645868268032e-05, "loss": 0.1013, "step": 6920 }, { "grad_norm": 0.624945878982544, "learning_rate": 9.963396660277289e-05, "loss": 0.1013, "step": 6930 }, { "grad_norm": 0.6108288168907166, "learning_rate": 9.963146604177807e-05, "loss": 0.0899, "step": 6940 }, { "grad_norm": 0.5419483780860901, "learning_rate": 9.962895700012311e-05, "loss": 0.0825, "step": 6950 }, { "grad_norm": 0.5366808176040649, "learning_rate": 9.962643947823677e-05, "loss": 0.0771, "step": 6960 }, { "grad_norm": 0.6859146356582642, "learning_rate": 9.962391347654921e-05, "loss": 0.0885, "step": 6970 }, { "grad_norm": 0.5920757055282593, "learning_rate": 9.962137899549204e-05, "loss": 0.0966, "step": 6980 }, { "grad_norm": 0.836402416229248, "learning_rate": 9.961883603549835e-05, "loss": 0.0823, "step": 6990 }, { "grad_norm": 0.6983429193496704, "learning_rate": 9.961628459700267e-05, "loss": 0.0912, "step": 7000 }, { "grad_norm": 0.6558199524879456, "learning_rate": 9.961372468044095e-05, "loss": 0.1068, "step": 7010 }, { "grad_norm": 0.5863175392150879, "learning_rate": 9.961115628625062e-05, "loss": 0.0863, "step": 7020 }, { "grad_norm": 0.6710191369056702, "learning_rate": 9.960857941487056e-05, "loss": 0.1068, "step": 7030 }, { "grad_norm": 0.5011793375015259, "learning_rate": 9.960599406674106e-05, "loss": 0.0892, "step": 7040 }, { "grad_norm": 0.645503044128418, "learning_rate": 9.960340024230393e-05, "loss": 0.0891, "step": 7050 }, { "grad_norm": 0.6564745306968689, "learning_rate": 9.960079794200232e-05, "loss": 0.0879, "step": 7060 }, { "grad_norm": 0.5576403141021729, "learning_rate": 9.959818716628096e-05, "loss": 0.0912, "step": 7070 }, { "grad_norm": 0.49751016497612, "learning_rate": 9.95955679155859e-05, "loss": 0.1029, "step": 7080 }, { "grad_norm": 0.8370856642723083, "learning_rate": 9.959294019036472e-05, "loss": 0.0982, "step": 7090 }, { "grad_norm": 0.6316747665405273, "learning_rate": 9.959030399106646e-05, "loss": 0.0798, "step": 7100 }, { "grad_norm": 0.5682113766670227, "learning_rate": 9.958765931814153e-05, "loss": 0.1029, "step": 7110 }, { "grad_norm": 0.5097228288650513, "learning_rate": 9.958500617204184e-05, "loss": 0.0784, "step": 7120 }, { "grad_norm": 0.613924503326416, "learning_rate": 9.958234455322075e-05, "loss": 0.0915, "step": 7130 }, { "grad_norm": 0.7192093729972839, "learning_rate": 9.957967446213308e-05, "loss": 0.1005, "step": 7140 }, { "grad_norm": 0.6318369507789612, "learning_rate": 9.957699589923501e-05, "loss": 0.1107, "step": 7150 }, { "grad_norm": 0.8211934566497803, "learning_rate": 9.957430886498431e-05, "loss": 0.0981, "step": 7160 }, { "grad_norm": 0.6131394505500793, "learning_rate": 9.957161335984008e-05, "loss": 0.0956, "step": 7170 }, { "grad_norm": 0.6803358793258667, "learning_rate": 9.956890938426291e-05, "loss": 0.0865, "step": 7180 }, { "grad_norm": 0.4605611264705658, "learning_rate": 9.956619693871482e-05, "loss": 0.0942, "step": 7190 }, { "grad_norm": 0.768170952796936, "learning_rate": 9.956347602365934e-05, "loss": 0.0808, "step": 7200 }, { "grad_norm": 0.5308643579483032, "learning_rate": 9.956074663956135e-05, "loss": 0.0917, "step": 7210 }, { "grad_norm": 0.772585928440094, "learning_rate": 9.955800878688726e-05, "loss": 0.1032, "step": 7220 }, { "grad_norm": 0.6232753992080688, "learning_rate": 9.955526246610489e-05, "loss": 0.0863, "step": 7230 }, { "grad_norm": 0.7470812201499939, "learning_rate": 9.955250767768349e-05, "loss": 0.0925, "step": 7240 }, { "grad_norm": 0.4979378581047058, "learning_rate": 9.95497444220938e-05, "loss": 0.0938, "step": 7250 }, { "grad_norm": 0.5876529812812805, "learning_rate": 9.954697269980797e-05, "loss": 0.0914, "step": 7260 }, { "grad_norm": 0.7688570618629456, "learning_rate": 9.954419251129962e-05, "loss": 0.0864, "step": 7270 }, { "grad_norm": 0.5376588106155396, "learning_rate": 9.95414038570438e-05, "loss": 0.0898, "step": 7280 }, { "grad_norm": 0.5590142607688904, "learning_rate": 9.953860673751703e-05, "loss": 0.0901, "step": 7290 }, { "grad_norm": 0.5715140104293823, "learning_rate": 9.953580115319725e-05, "loss": 0.082, "step": 7300 }, { "grad_norm": 0.6420608162879944, "learning_rate": 9.953298710456387e-05, "loss": 0.0891, "step": 7310 }, { "grad_norm": 0.5705764889717102, "learning_rate": 9.953016459209771e-05, "loss": 0.0842, "step": 7320 }, { "grad_norm": 0.7505642175674438, "learning_rate": 9.952733361628108e-05, "loss": 0.0945, "step": 7330 }, { "grad_norm": 0.8261575102806091, "learning_rate": 9.952449417759772e-05, "loss": 0.0946, "step": 7340 }, { "grad_norm": 0.6876046061515808, "learning_rate": 9.952164627653279e-05, "loss": 0.0941, "step": 7350 }, { "grad_norm": 0.5205242037773132, "learning_rate": 9.951878991357292e-05, "loss": 0.0869, "step": 7360 }, { "grad_norm": 0.6260062456130981, "learning_rate": 9.951592508920622e-05, "loss": 0.0946, "step": 7370 }, { "grad_norm": 0.45087912678718567, "learning_rate": 9.951305180392219e-05, "loss": 0.0803, "step": 7380 }, { "grad_norm": 0.6674686074256897, "learning_rate": 9.951017005821178e-05, "loss": 0.0951, "step": 7390 }, { "grad_norm": 0.5815925002098083, "learning_rate": 9.95072798525674e-05, "loss": 0.091, "step": 7400 }, { "grad_norm": 0.6603900194168091, "learning_rate": 9.950438118748293e-05, "loss": 0.0955, "step": 7410 }, { "grad_norm": 0.5523955821990967, "learning_rate": 9.950147406345366e-05, "loss": 0.0971, "step": 7420 }, { "grad_norm": 0.48180487751960754, "learning_rate": 9.949855848097635e-05, "loss": 0.1013, "step": 7430 }, { "grad_norm": 0.5516613721847534, "learning_rate": 9.949563444054916e-05, "loss": 0.0941, "step": 7440 }, { "grad_norm": 0.4823560118675232, "learning_rate": 9.949270194267178e-05, "loss": 0.0852, "step": 7450 }, { "grad_norm": 0.5163139700889587, "learning_rate": 9.948976098784526e-05, "loss": 0.0883, "step": 7460 }, { "grad_norm": 0.5708469748497009, "learning_rate": 9.948681157657213e-05, "loss": 0.1171, "step": 7470 }, { "grad_norm": 0.5722930431365967, "learning_rate": 9.948385370935638e-05, "loss": 0.1088, "step": 7480 }, { "grad_norm": 0.6871486306190491, "learning_rate": 9.94808873867034e-05, "loss": 0.0932, "step": 7490 }, { "grad_norm": 0.629265546798706, "learning_rate": 9.947791260912009e-05, "loss": 0.0792, "step": 7500 }, { "grad_norm": 0.6640342473983765, "learning_rate": 9.947492937711474e-05, "loss": 0.0775, "step": 7510 }, { "grad_norm": 0.4979003965854645, "learning_rate": 9.947193769119707e-05, "loss": 0.0766, "step": 7520 }, { "grad_norm": 0.48249080777168274, "learning_rate": 9.946893755187834e-05, "loss": 0.0756, "step": 7530 }, { "grad_norm": 0.5986654162406921, "learning_rate": 9.946592895967115e-05, "loss": 0.084, "step": 7540 }, { "grad_norm": 0.6012784838676453, "learning_rate": 9.94629119150896e-05, "loss": 0.0813, "step": 7550 }, { "grad_norm": 0.5222305655479431, "learning_rate": 9.94598864186492e-05, "loss": 0.0842, "step": 7560 }, { "grad_norm": 0.5613369345664978, "learning_rate": 9.945685247086696e-05, "loss": 0.0922, "step": 7570 }, { "grad_norm": 0.61810302734375, "learning_rate": 9.945381007226129e-05, "loss": 0.0931, "step": 7580 }, { "grad_norm": 0.5824467539787292, "learning_rate": 9.945075922335203e-05, "loss": 0.0897, "step": 7590 }, { "grad_norm": 0.5175787210464478, "learning_rate": 9.944769992466049e-05, "loss": 0.0965, "step": 7600 }, { "grad_norm": 0.4930833578109741, "learning_rate": 9.944463217670945e-05, "loss": 0.0885, "step": 7610 }, { "grad_norm": 0.4884265065193176, "learning_rate": 9.944155598002307e-05, "loss": 0.0775, "step": 7620 }, { "grad_norm": 0.46452802419662476, "learning_rate": 9.943847133512701e-05, "loss": 0.0909, "step": 7630 }, { "grad_norm": 0.7047272324562073, "learning_rate": 9.943537824254834e-05, "loss": 0.0807, "step": 7640 }, { "grad_norm": 0.4784505069255829, "learning_rate": 9.943227670281559e-05, "loss": 0.0851, "step": 7650 }, { "grad_norm": 0.4282028079032898, "learning_rate": 9.942916671645873e-05, "loss": 0.0781, "step": 7660 }, { "grad_norm": 0.5581873059272766, "learning_rate": 9.942604828400916e-05, "loss": 0.0866, "step": 7670 }, { "grad_norm": 0.5395007729530334, "learning_rate": 9.942292140599975e-05, "loss": 0.0926, "step": 7680 }, { "grad_norm": 0.7307352423667908, "learning_rate": 9.94197860829648e-05, "loss": 0.0815, "step": 7690 }, { "grad_norm": 0.7513024806976318, "learning_rate": 9.941664231544004e-05, "loss": 0.0814, "step": 7700 }, { "grad_norm": 0.7062221765518188, "learning_rate": 9.941349010396264e-05, "loss": 0.1072, "step": 7710 }, { "grad_norm": 0.6400302648544312, "learning_rate": 9.941032944907125e-05, "loss": 0.0971, "step": 7720 }, { "grad_norm": 0.721508800983429, "learning_rate": 9.940716035130596e-05, "loss": 0.0949, "step": 7730 }, { "grad_norm": 0.5732572674751282, "learning_rate": 9.940398281120821e-05, "loss": 0.0928, "step": 7740 }, { "grad_norm": 0.49868085980415344, "learning_rate": 9.940079682932102e-05, "loss": 0.0779, "step": 7750 }, { "grad_norm": 0.5582558512687683, "learning_rate": 9.939760240618877e-05, "loss": 0.0947, "step": 7760 }, { "grad_norm": 0.4941428005695343, "learning_rate": 9.939439954235729e-05, "loss": 0.0825, "step": 7770 }, { "grad_norm": 0.4945457875728607, "learning_rate": 9.939118823837387e-05, "loss": 0.0945, "step": 7780 }, { "grad_norm": 0.6239022612571716, "learning_rate": 9.938796849478725e-05, "loss": 0.0858, "step": 7790 }, { "grad_norm": 0.5183038115501404, "learning_rate": 9.938474031214755e-05, "loss": 0.0867, "step": 7800 }, { "grad_norm": 0.6279041767120361, "learning_rate": 9.938150369100643e-05, "loss": 0.0818, "step": 7810 }, { "grad_norm": 0.5758687257766724, "learning_rate": 9.93782586319169e-05, "loss": 0.0782, "step": 7820 }, { "grad_norm": 0.6581218242645264, "learning_rate": 9.937500513543348e-05, "loss": 0.0788, "step": 7830 }, { "grad_norm": 0.7294193506240845, "learning_rate": 9.937174320211207e-05, "loss": 0.074, "step": 7840 }, { "grad_norm": 0.5522366762161255, "learning_rate": 9.936847283251009e-05, "loss": 0.0978, "step": 7850 }, { "grad_norm": 0.6366729736328125, "learning_rate": 9.936519402718632e-05, "loss": 0.0898, "step": 7860 }, { "grad_norm": 0.6705731749534607, "learning_rate": 9.936190678670102e-05, "loss": 0.1082, "step": 7870 }, { "grad_norm": 0.6021522283554077, "learning_rate": 9.935861111161593e-05, "loss": 0.0981, "step": 7880 }, { "grad_norm": 0.584592878818512, "learning_rate": 9.935530700249416e-05, "loss": 0.0751, "step": 7890 }, { "grad_norm": 0.5267229080200195, "learning_rate": 9.935199445990028e-05, "loss": 0.0922, "step": 7900 }, { "grad_norm": 0.6723467707633972, "learning_rate": 9.934867348440033e-05, "loss": 0.0977, "step": 7910 }, { "grad_norm": 0.7331376075744629, "learning_rate": 9.934534407656176e-05, "loss": 0.113, "step": 7920 }, { "grad_norm": 0.5299960970878601, "learning_rate": 9.93420062369535e-05, "loss": 0.0871, "step": 7930 }, { "grad_norm": 0.5351442098617554, "learning_rate": 9.933865996614589e-05, "loss": 0.086, "step": 7940 }, { "grad_norm": 0.9054075479507446, "learning_rate": 9.933530526471068e-05, "loss": 0.0783, "step": 7950 }, { "grad_norm": 0.5845720767974854, "learning_rate": 9.933194213322114e-05, "loss": 0.0847, "step": 7960 }, { "grad_norm": 0.7220547795295715, "learning_rate": 9.932857057225192e-05, "loss": 0.0815, "step": 7970 }, { "grad_norm": 0.5834682583808899, "learning_rate": 9.932519058237912e-05, "loss": 0.09, "step": 7980 }, { "grad_norm": 0.5702254176139832, "learning_rate": 9.932180216418032e-05, "loss": 0.0798, "step": 7990 }, { "grad_norm": 0.549891471862793, "learning_rate": 9.931840531823446e-05, "loss": 0.0764, "step": 8000 }, { "grad_norm": 0.49852505326271057, "learning_rate": 9.9315000045122e-05, "loss": 0.0789, "step": 8010 }, { "grad_norm": 0.41623061895370483, "learning_rate": 9.931158634542481e-05, "loss": 0.0773, "step": 8020 }, { "grad_norm": 0.5035141706466675, "learning_rate": 9.930816421972617e-05, "loss": 0.07, "step": 8030 }, { "grad_norm": 0.5067978501319885, "learning_rate": 9.930473366861086e-05, "loss": 0.0731, "step": 8040 }, { "grad_norm": 0.5973290205001831, "learning_rate": 9.930129469266505e-05, "loss": 0.0903, "step": 8050 }, { "grad_norm": 0.5917403101921082, "learning_rate": 9.929784729247638e-05, "loss": 0.0797, "step": 8060 }, { "grad_norm": 0.6746058464050293, "learning_rate": 9.929439146863389e-05, "loss": 0.0807, "step": 8070 }, { "grad_norm": 0.5355954170227051, "learning_rate": 9.92909272217281e-05, "loss": 0.1032, "step": 8080 }, { "grad_norm": 0.6361847519874573, "learning_rate": 9.928745455235097e-05, "loss": 0.0931, "step": 8090 }, { "grad_norm": 0.7009081244468689, "learning_rate": 9.928397346109588e-05, "loss": 0.0978, "step": 8100 }, { "grad_norm": 0.5605106353759766, "learning_rate": 9.928048394855762e-05, "loss": 0.0942, "step": 8110 }, { "grad_norm": 0.44530001282691956, "learning_rate": 9.92769860153325e-05, "loss": 0.0835, "step": 8120 }, { "grad_norm": 0.6390478610992432, "learning_rate": 9.927347966201819e-05, "loss": 0.1175, "step": 8130 }, { "grad_norm": 0.6196572780609131, "learning_rate": 9.926996488921383e-05, "loss": 0.1052, "step": 8140 }, { "grad_norm": 0.6371376514434814, "learning_rate": 9.926644169752001e-05, "loss": 0.0874, "step": 8150 }, { "grad_norm": 0.5777274370193481, "learning_rate": 9.926291008753875e-05, "loss": 0.0979, "step": 8160 }, { "grad_norm": 0.5090951919555664, "learning_rate": 9.92593700598735e-05, "loss": 0.0924, "step": 8170 }, { "grad_norm": 0.5142790675163269, "learning_rate": 9.925582161512915e-05, "loss": 0.0835, "step": 8180 }, { "grad_norm": 0.5196369290351868, "learning_rate": 9.925226475391205e-05, "loss": 0.0824, "step": 8190 }, { "grad_norm": 0.5893110036849976, "learning_rate": 9.924869947682993e-05, "loss": 0.0915, "step": 8200 }, { "grad_norm": 0.6331132650375366, "learning_rate": 9.924512578449204e-05, "loss": 0.0868, "step": 8210 }, { "grad_norm": 0.5819989442825317, "learning_rate": 9.924154367750901e-05, "loss": 0.0856, "step": 8220 }, { "grad_norm": 0.48844602704048157, "learning_rate": 9.923795315649293e-05, "loss": 0.0822, "step": 8230 }, { "grad_norm": 0.49501869082450867, "learning_rate": 9.92343542220573e-05, "loss": 0.084, "step": 8240 }, { "grad_norm": 0.439561128616333, "learning_rate": 9.92307468748171e-05, "loss": 0.0736, "step": 8250 }, { "grad_norm": 0.5306779742240906, "learning_rate": 9.922713111538873e-05, "loss": 0.0635, "step": 8260 }, { "grad_norm": 0.5263857245445251, "learning_rate": 9.922350694439003e-05, "loss": 0.085, "step": 8270 }, { "grad_norm": 0.5755216479301453, "learning_rate": 9.921987436244024e-05, "loss": 0.0921, "step": 8280 }, { "grad_norm": 0.7336628437042236, "learning_rate": 9.921623337016008e-05, "loss": 0.0946, "step": 8290 }, { "grad_norm": 0.4946325421333313, "learning_rate": 9.921258396817172e-05, "loss": 0.096, "step": 8300 }, { "grad_norm": 0.6157614588737488, "learning_rate": 9.920892615709874e-05, "loss": 0.0936, "step": 8310 }, { "grad_norm": 0.5494729280471802, "learning_rate": 9.920525993756612e-05, "loss": 0.0697, "step": 8320 }, { "grad_norm": 0.4836136996746063, "learning_rate": 9.920158531020036e-05, "loss": 0.0859, "step": 8330 }, { "grad_norm": 0.45351141691207886, "learning_rate": 9.919790227562933e-05, "loss": 0.0708, "step": 8340 }, { "grad_norm": 0.5317928791046143, "learning_rate": 9.919421083448237e-05, "loss": 0.092, "step": 8350 }, { "grad_norm": 0.6345781087875366, "learning_rate": 9.919051098739022e-05, "loss": 0.0777, "step": 8360 }, { "grad_norm": 0.6101707220077515, "learning_rate": 9.918680273498514e-05, "loss": 0.0801, "step": 8370 }, { "grad_norm": 0.5684232115745544, "learning_rate": 9.918308607790072e-05, "loss": 0.0903, "step": 8380 }, { "grad_norm": 0.5066128969192505, "learning_rate": 9.917936101677205e-05, "loss": 0.0771, "step": 8390 }, { "grad_norm": 0.5137205719947815, "learning_rate": 9.917562755223564e-05, "loss": 0.0782, "step": 8400 }, { "grad_norm": 0.7967557907104492, "learning_rate": 9.917188568492944e-05, "loss": 0.0823, "step": 8410 }, { "grad_norm": 0.4700682461261749, "learning_rate": 9.916813541549283e-05, "loss": 0.0809, "step": 8420 }, { "grad_norm": 0.5126561522483826, "learning_rate": 9.916437674456663e-05, "loss": 0.0838, "step": 8430 }, { "grad_norm": 0.45789849758148193, "learning_rate": 9.916060967279308e-05, "loss": 0.0797, "step": 8440 }, { "grad_norm": 0.5851441025733948, "learning_rate": 9.91568342008159e-05, "loss": 0.0922, "step": 8450 }, { "grad_norm": 0.592871904373169, "learning_rate": 9.915305032928019e-05, "loss": 0.0791, "step": 8460 }, { "grad_norm": 0.5705388784408569, "learning_rate": 9.914925805883253e-05, "loss": 0.0729, "step": 8470 }, { "grad_norm": 0.4554780423641205, "learning_rate": 9.914545739012088e-05, "loss": 0.0715, "step": 8480 }, { "grad_norm": 0.6248841881752014, "learning_rate": 9.91416483237947e-05, "loss": 0.0802, "step": 8490 }, { "grad_norm": 0.537782609462738, "learning_rate": 9.913783086050485e-05, "loss": 0.0986, "step": 8500 }, { "grad_norm": 0.5609769821166992, "learning_rate": 9.913400500090364e-05, "loss": 0.0839, "step": 8510 }, { "grad_norm": 0.3873582184314728, "learning_rate": 9.913017074564479e-05, "loss": 0.0768, "step": 8520 }, { "grad_norm": 0.5085590481758118, "learning_rate": 9.912632809538348e-05, "loss": 0.0846, "step": 8530 }, { "grad_norm": 0.5589244961738586, "learning_rate": 9.912247705077629e-05, "loss": 0.0788, "step": 8540 }, { "grad_norm": 0.5696724057197571, "learning_rate": 9.911861761248127e-05, "loss": 0.0825, "step": 8550 }, { "grad_norm": 0.5275394916534424, "learning_rate": 9.91147497811579e-05, "loss": 0.0792, "step": 8560 }, { "grad_norm": 0.7234816551208496, "learning_rate": 9.911087355746709e-05, "loss": 0.1192, "step": 8570 }, { "grad_norm": 0.8538196086883545, "learning_rate": 9.910698894207117e-05, "loss": 0.1013, "step": 8580 }, { "grad_norm": 0.555899441242218, "learning_rate": 9.910309593563392e-05, "loss": 0.0812, "step": 8590 }, { "grad_norm": 0.46745041012763977, "learning_rate": 9.909919453882057e-05, "loss": 0.0678, "step": 8600 }, { "grad_norm": 0.6377336978912354, "learning_rate": 9.90952847522977e-05, "loss": 0.0823, "step": 8610 }, { "grad_norm": 0.6221440434455872, "learning_rate": 9.909136657673346e-05, "loss": 0.0925, "step": 8620 }, { "grad_norm": 0.40222516655921936, "learning_rate": 9.908744001279731e-05, "loss": 0.0713, "step": 8630 }, { "grad_norm": 0.6545020937919617, "learning_rate": 9.90835050611602e-05, "loss": 0.0947, "step": 8640 }, { "grad_norm": 0.603363037109375, "learning_rate": 9.90795617224945e-05, "loss": 0.0897, "step": 8650 }, { "grad_norm": 0.4615740478038788, "learning_rate": 9.907560999747405e-05, "loss": 0.0717, "step": 8660 }, { "grad_norm": 0.6813725233078003, "learning_rate": 9.907164988677408e-05, "loss": 0.0784, "step": 8670 }, { "grad_norm": 0.8261936902999878, "learning_rate": 9.906768139107124e-05, "loss": 0.082, "step": 8680 }, { "grad_norm": 0.506512463092804, "learning_rate": 9.906370451104367e-05, "loss": 0.0913, "step": 8690 }, { "grad_norm": 0.6909865736961365, "learning_rate": 9.905971924737088e-05, "loss": 0.0818, "step": 8700 }, { "grad_norm": 0.5049399137496948, "learning_rate": 9.905572560073387e-05, "loss": 0.0936, "step": 8710 }, { "grad_norm": 0.6611708402633667, "learning_rate": 9.905172357181501e-05, "loss": 0.1074, "step": 8720 }, { "grad_norm": 0.6293368339538574, "learning_rate": 9.904771316129817e-05, "loss": 0.081, "step": 8730 }, { "grad_norm": 0.560131847858429, "learning_rate": 9.904369436986862e-05, "loss": 0.0803, "step": 8740 }, { "grad_norm": 0.5900421142578125, "learning_rate": 9.903966719821303e-05, "loss": 0.1015, "step": 8750 }, { "grad_norm": 0.44425731897354126, "learning_rate": 9.903563164701956e-05, "loss": 0.0766, "step": 8760 }, { "grad_norm": 0.6485220193862915, "learning_rate": 9.903158771697778e-05, "loss": 0.0957, "step": 8770 }, { "grad_norm": 0.5639437437057495, "learning_rate": 9.902753540877867e-05, "loss": 0.0947, "step": 8780 }, { "grad_norm": 0.6572754979133606, "learning_rate": 9.902347472311466e-05, "loss": 0.0899, "step": 8790 }, { "grad_norm": 0.5874217748641968, "learning_rate": 9.901940566067962e-05, "loss": 0.0671, "step": 8800 }, { "grad_norm": 0.6205236315727234, "learning_rate": 9.901532822216883e-05, "loss": 0.0813, "step": 8810 }, { "grad_norm": 0.4463624060153961, "learning_rate": 9.901124240827904e-05, "loss": 0.0684, "step": 8820 }, { "grad_norm": 0.6494914889335632, "learning_rate": 9.900714821970835e-05, "loss": 0.0793, "step": 8830 }, { "grad_norm": 0.6636143326759338, "learning_rate": 9.900304565715641e-05, "loss": 0.0723, "step": 8840 }, { "grad_norm": 0.4206414520740509, "learning_rate": 9.899893472132419e-05, "loss": 0.0688, "step": 8850 }, { "grad_norm": 0.6006529331207275, "learning_rate": 9.899481541291415e-05, "loss": 0.0811, "step": 8860 }, { "grad_norm": 0.8511528968811035, "learning_rate": 9.899068773263016e-05, "loss": 0.0838, "step": 8870 }, { "grad_norm": 0.6526738405227661, "learning_rate": 9.898655168117754e-05, "loss": 0.0925, "step": 8880 }, { "grad_norm": 0.518378734588623, "learning_rate": 9.898240725926302e-05, "loss": 0.0914, "step": 8890 }, { "grad_norm": 0.5009679794311523, "learning_rate": 9.897825446759478e-05, "loss": 0.0784, "step": 8900 }, { "grad_norm": 0.5039425492286682, "learning_rate": 9.897409330688241e-05, "loss": 0.0848, "step": 8910 }, { "grad_norm": 0.7031756639480591, "learning_rate": 9.896992377783692e-05, "loss": 0.0985, "step": 8920 }, { "grad_norm": 0.6705320477485657, "learning_rate": 9.89657458811708e-05, "loss": 0.0728, "step": 8930 }, { "grad_norm": 0.7530456781387329, "learning_rate": 9.896155961759792e-05, "loss": 0.0829, "step": 8940 }, { "grad_norm": 0.6636694073677063, "learning_rate": 9.895736498783361e-05, "loss": 0.0783, "step": 8950 }, { "grad_norm": 0.4695003032684326, "learning_rate": 9.895316199259462e-05, "loss": 0.0776, "step": 8960 }, { "grad_norm": 0.548988938331604, "learning_rate": 9.894895063259909e-05, "loss": 0.0716, "step": 8970 }, { "grad_norm": 0.556376576423645, "learning_rate": 9.894473090856667e-05, "loss": 0.091, "step": 8980 }, { "grad_norm": 0.5494205951690674, "learning_rate": 9.894050282121839e-05, "loss": 0.1017, "step": 8990 }, { "grad_norm": 0.4380757510662079, "learning_rate": 9.893626637127668e-05, "loss": 0.0751, "step": 9000 }, { "grad_norm": 0.5695647597312927, "learning_rate": 9.893202155946546e-05, "loss": 0.0762, "step": 9010 }, { "grad_norm": 0.7548545598983765, "learning_rate": 9.892776838651006e-05, "loss": 0.0802, "step": 9020 }, { "grad_norm": 0.5186774730682373, "learning_rate": 9.892350685313722e-05, "loss": 0.0918, "step": 9030 }, { "grad_norm": 0.4795039892196655, "learning_rate": 9.891923696007513e-05, "loss": 0.0841, "step": 9040 }, { "grad_norm": 0.6422013640403748, "learning_rate": 9.891495870805336e-05, "loss": 0.0752, "step": 9050 }, { "grad_norm": 0.4954337477684021, "learning_rate": 9.891067209780298e-05, "loss": 0.0927, "step": 9060 }, { "grad_norm": 0.642241895198822, "learning_rate": 9.890637713005646e-05, "loss": 0.0649, "step": 9070 }, { "grad_norm": 0.45920902490615845, "learning_rate": 9.890207380554767e-05, "loss": 0.063, "step": 9080 }, { "grad_norm": 0.4753133952617645, "learning_rate": 9.889776212501196e-05, "loss": 0.0812, "step": 9090 }, { "grad_norm": 0.5635771751403809, "learning_rate": 9.889344208918605e-05, "loss": 0.0831, "step": 9100 }, { "grad_norm": 0.5160440802574158, "learning_rate": 9.888911369880812e-05, "loss": 0.0841, "step": 9110 }, { "grad_norm": 0.5459856390953064, "learning_rate": 9.888477695461777e-05, "loss": 0.0857, "step": 9120 }, { "grad_norm": 0.49252229928970337, "learning_rate": 9.888043185735607e-05, "loss": 0.0855, "step": 9130 }, { "grad_norm": 0.5687435269355774, "learning_rate": 9.887607840776542e-05, "loss": 0.0675, "step": 9140 }, { "grad_norm": 0.41442230343818665, "learning_rate": 9.887171660658975e-05, "loss": 0.0959, "step": 9150 }, { "grad_norm": 0.4957631528377533, "learning_rate": 9.886734645457435e-05, "loss": 0.0685, "step": 9160 }, { "grad_norm": 0.6105101704597473, "learning_rate": 9.886296795246597e-05, "loss": 0.0736, "step": 9170 }, { "grad_norm": 0.6506906747817993, "learning_rate": 9.885858110101276e-05, "loss": 0.0811, "step": 9180 }, { "grad_norm": 0.5789777040481567, "learning_rate": 9.885418590096434e-05, "loss": 0.0685, "step": 9190 }, { "grad_norm": 0.5820557475090027, "learning_rate": 9.88497823530717e-05, "loss": 0.0841, "step": 9200 }, { "grad_norm": 0.6317073702812195, "learning_rate": 9.884537045808732e-05, "loss": 0.1042, "step": 9210 }, { "grad_norm": 0.5042415857315063, "learning_rate": 9.884095021676502e-05, "loss": 0.0848, "step": 9220 }, { "grad_norm": 0.7897442579269409, "learning_rate": 9.883652162986017e-05, "loss": 0.0847, "step": 9230 }, { "grad_norm": 0.6954278945922852, "learning_rate": 9.883208469812943e-05, "loss": 0.0852, "step": 9240 }, { "grad_norm": 0.4741733968257904, "learning_rate": 9.882763942233098e-05, "loss": 0.0698, "step": 9250 }, { "grad_norm": 0.4223957657814026, "learning_rate": 9.882318580322441e-05, "loss": 0.0677, "step": 9260 }, { "grad_norm": 0.6547635197639465, "learning_rate": 9.881872384157067e-05, "loss": 0.0859, "step": 9270 }, { "grad_norm": 0.5167057514190674, "learning_rate": 9.881425353813225e-05, "loss": 0.079, "step": 9280 }, { "grad_norm": 0.6863450407981873, "learning_rate": 9.880977489367296e-05, "loss": 0.0852, "step": 9290 }, { "grad_norm": 0.4831887185573578, "learning_rate": 9.88052879089581e-05, "loss": 0.0941, "step": 9300 }, { "grad_norm": 0.43142834305763245, "learning_rate": 9.880079258475434e-05, "loss": 0.0889, "step": 9310 }, { "grad_norm": 0.5420010685920715, "learning_rate": 9.879628892182985e-05, "loss": 0.1016, "step": 9320 }, { "grad_norm": 0.6042842268943787, "learning_rate": 9.879177692095416e-05, "loss": 0.0899, "step": 9330 }, { "grad_norm": 0.436664879322052, "learning_rate": 9.878725658289825e-05, "loss": 0.0852, "step": 9340 }, { "grad_norm": 0.5741012096405029, "learning_rate": 9.878272790843454e-05, "loss": 0.072, "step": 9350 }, { "grad_norm": 0.5866198539733887, "learning_rate": 9.877819089833682e-05, "loss": 0.0725, "step": 9360 }, { "grad_norm": 0.40041622519493103, "learning_rate": 9.877364555338038e-05, "loss": 0.0888, "step": 9370 }, { "grad_norm": 0.5219624638557434, "learning_rate": 9.876909187434186e-05, "loss": 0.0896, "step": 9380 }, { "grad_norm": 0.6842706203460693, "learning_rate": 9.876452986199939e-05, "loss": 0.0783, "step": 9390 }, { "grad_norm": 0.7348324656486511, "learning_rate": 9.875995951713248e-05, "loss": 0.0841, "step": 9400 }, { "grad_norm": 0.5572240948677063, "learning_rate": 9.875538084052207e-05, "loss": 0.0918, "step": 9410 }, { "grad_norm": 0.6661365628242493, "learning_rate": 9.875079383295053e-05, "loss": 0.0833, "step": 9420 }, { "grad_norm": 0.5336458683013916, "learning_rate": 9.874619849520167e-05, "loss": 0.0772, "step": 9430 }, { "grad_norm": 0.5911980867385864, "learning_rate": 9.874159482806069e-05, "loss": 0.0816, "step": 9440 }, { "grad_norm": 0.539929986000061, "learning_rate": 9.873698283231426e-05, "loss": 0.0781, "step": 9450 }, { "grad_norm": 0.5547955632209778, "learning_rate": 9.87323625087504e-05, "loss": 0.0687, "step": 9460 }, { "grad_norm": 0.4650227129459381, "learning_rate": 9.872773385815863e-05, "loss": 0.0986, "step": 9470 }, { "grad_norm": 0.4862208664417267, "learning_rate": 9.872309688132986e-05, "loss": 0.082, "step": 9480 }, { "grad_norm": 0.4302680194377899, "learning_rate": 9.871845157905639e-05, "loss": 0.0738, "step": 9490 }, { "grad_norm": 0.639061450958252, "learning_rate": 9.871379795213201e-05, "loss": 0.0805, "step": 9500 }, { "grad_norm": 0.47629767656326294, "learning_rate": 9.87091360013519e-05, "loss": 0.0793, "step": 9510 }, { "grad_norm": 0.7618424296379089, "learning_rate": 9.870446572751262e-05, "loss": 0.0824, "step": 9520 }, { "grad_norm": 0.4692210257053375, "learning_rate": 9.869978713141224e-05, "loss": 0.0856, "step": 9530 }, { "grad_norm": 0.686406672000885, "learning_rate": 9.869510021385016e-05, "loss": 0.0908, "step": 9540 }, { "grad_norm": 0.478204607963562, "learning_rate": 9.869040497562727e-05, "loss": 0.0921, "step": 9550 }, { "grad_norm": 0.5028038620948792, "learning_rate": 9.868570141754587e-05, "loss": 0.0887, "step": 9560 }, { "grad_norm": 0.4974856674671173, "learning_rate": 9.868098954040965e-05, "loss": 0.0933, "step": 9570 }, { "grad_norm": 0.5777629017829895, "learning_rate": 9.867626934502374e-05, "loss": 0.0934, "step": 9580 }, { "grad_norm": 0.42566266655921936, "learning_rate": 9.86715408321947e-05, "loss": 0.0806, "step": 9590 }, { "grad_norm": 0.581152617931366, "learning_rate": 9.86668040027305e-05, "loss": 0.0729, "step": 9600 }, { "grad_norm": 0.5615130662918091, "learning_rate": 9.866205885744053e-05, "loss": 0.0762, "step": 9610 }, { "grad_norm": 0.5025386810302734, "learning_rate": 9.865730539713563e-05, "loss": 0.0908, "step": 9620 }, { "grad_norm": 0.62486332654953, "learning_rate": 9.8652543622628e-05, "loss": 0.0828, "step": 9630 }, { "grad_norm": 0.6075209379196167, "learning_rate": 9.864777353473132e-05, "loss": 0.0701, "step": 9640 }, { "grad_norm": 0.5043218731880188, "learning_rate": 9.864299513426068e-05, "loss": 0.0772, "step": 9650 }, { "grad_norm": 0.5578736066818237, "learning_rate": 9.863820842203254e-05, "loss": 0.0802, "step": 9660 }, { "grad_norm": 0.6063784956932068, "learning_rate": 9.863341339886483e-05, "loss": 0.0825, "step": 9670 }, { "grad_norm": 0.5920599699020386, "learning_rate": 9.86286100655769e-05, "loss": 0.0663, "step": 9680 }, { "grad_norm": 0.6140075325965881, "learning_rate": 9.862379842298953e-05, "loss": 0.0793, "step": 9690 }, { "grad_norm": 0.4084022343158722, "learning_rate": 9.861897847192485e-05, "loss": 0.0759, "step": 9700 }, { "grad_norm": 0.5866997241973877, "learning_rate": 9.86141502132065e-05, "loss": 0.0756, "step": 9710 }, { "grad_norm": 0.5837543606758118, "learning_rate": 9.860931364765946e-05, "loss": 0.0861, "step": 9720 }, { "grad_norm": 0.6420994997024536, "learning_rate": 9.860446877611021e-05, "loss": 0.0888, "step": 9730 }, { "grad_norm": 0.6025802493095398, "learning_rate": 9.859961559938655e-05, "loss": 0.0753, "step": 9740 }, { "grad_norm": 0.5416555404663086, "learning_rate": 9.85947541183178e-05, "loss": 0.0782, "step": 9750 }, { "grad_norm": 0.6138864755630493, "learning_rate": 9.858988433373463e-05, "loss": 0.0787, "step": 9760 }, { "grad_norm": 0.7460469007492065, "learning_rate": 9.858500624646918e-05, "loss": 0.0913, "step": 9770 }, { "grad_norm": 0.611430823802948, "learning_rate": 9.858011985735497e-05, "loss": 0.0819, "step": 9780 }, { "grad_norm": 0.5472260117530823, "learning_rate": 9.857522516722693e-05, "loss": 0.0695, "step": 9790 }, { "grad_norm": 0.612477719783783, "learning_rate": 9.857032217692145e-05, "loss": 0.0847, "step": 9800 }, { "grad_norm": 0.6730431914329529, "learning_rate": 9.856541088727631e-05, "loss": 0.087, "step": 9810 }, { "grad_norm": 0.4748044013977051, "learning_rate": 9.856049129913072e-05, "loss": 0.0827, "step": 9820 }, { "grad_norm": 0.5210384726524353, "learning_rate": 9.85555634133253e-05, "loss": 0.0774, "step": 9830 }, { "grad_norm": 0.48352348804473877, "learning_rate": 9.855062723070208e-05, "loss": 0.0927, "step": 9840 }, { "grad_norm": 0.6232806444168091, "learning_rate": 9.854568275210454e-05, "loss": 0.094, "step": 9850 }, { "grad_norm": 0.49711430072784424, "learning_rate": 9.854072997837754e-05, "loss": 0.0706, "step": 9860 }, { "grad_norm": 0.719109833240509, "learning_rate": 9.853576891036737e-05, "loss": 0.074, "step": 9870 }, { "grad_norm": 0.8130472898483276, "learning_rate": 9.853079954892177e-05, "loss": 0.0797, "step": 9880 }, { "grad_norm": 0.5468916893005371, "learning_rate": 9.852582189488983e-05, "loss": 0.072, "step": 9890 }, { "grad_norm": 0.5319998264312744, "learning_rate": 9.852083594912212e-05, "loss": 0.0778, "step": 9900 }, { "grad_norm": 0.42753344774246216, "learning_rate": 9.851584171247058e-05, "loss": 0.0765, "step": 9910 }, { "grad_norm": 0.6325960755348206, "learning_rate": 9.851083918578863e-05, "loss": 0.0901, "step": 9920 }, { "grad_norm": 0.7266442179679871, "learning_rate": 9.850582836993103e-05, "loss": 0.0757, "step": 9930 }, { "grad_norm": 0.5723021626472473, "learning_rate": 9.850080926575397e-05, "loss": 0.077, "step": 9940 }, { "grad_norm": 0.5629532337188721, "learning_rate": 9.849578187411515e-05, "loss": 0.0723, "step": 9950 }, { "grad_norm": 0.691513180732727, "learning_rate": 9.849074619587354e-05, "loss": 0.0708, "step": 9960 }, { "grad_norm": 0.5229621529579163, "learning_rate": 9.848570223188964e-05, "loss": 0.0832, "step": 9970 }, { "grad_norm": 0.4520924687385559, "learning_rate": 9.848064998302531e-05, "loss": 0.0734, "step": 9980 }, { "grad_norm": 0.4509672224521637, "learning_rate": 9.847558945014386e-05, "loss": 0.0956, "step": 9990 }, { "grad_norm": 0.6918327212333679, "learning_rate": 9.847052063410996e-05, "loss": 0.0812, "step": 10000 }, { "grad_norm": 0.5079095363616943, "learning_rate": 9.846544353578977e-05, "loss": 0.0924, "step": 10010 }, { "grad_norm": 0.48569029569625854, "learning_rate": 9.846035815605081e-05, "loss": 0.0964, "step": 10020 }, { "grad_norm": 0.48805078864097595, "learning_rate": 9.845526449576204e-05, "loss": 0.0704, "step": 10030 }, { "grad_norm": 0.6332293152809143, "learning_rate": 9.845016255579383e-05, "loss": 0.0943, "step": 10040 }, { "grad_norm": 0.633980393409729, "learning_rate": 9.844505233701794e-05, "loss": 0.0893, "step": 10050 }, { "grad_norm": 0.562027096748352, "learning_rate": 9.843993384030757e-05, "loss": 0.0841, "step": 10060 }, { "grad_norm": 0.4004846513271332, "learning_rate": 9.843480706653737e-05, "loss": 0.0765, "step": 10070 }, { "grad_norm": 0.4759993553161621, "learning_rate": 9.84296720165833e-05, "loss": 0.0889, "step": 10080 }, { "grad_norm": 0.5726919174194336, "learning_rate": 9.842452869132286e-05, "loss": 0.0807, "step": 10090 }, { "grad_norm": 0.5510293841362, "learning_rate": 9.841937709163489e-05, "loss": 0.0769, "step": 10100 }, { "grad_norm": 0.5376730561256409, "learning_rate": 9.841421721839962e-05, "loss": 0.0648, "step": 10110 }, { "grad_norm": 0.5153419971466064, "learning_rate": 9.840904907249879e-05, "loss": 0.075, "step": 10120 }, { "grad_norm": 0.7496724724769592, "learning_rate": 9.840387265481545e-05, "loss": 0.0783, "step": 10130 }, { "grad_norm": 0.535487949848175, "learning_rate": 9.839868796623411e-05, "loss": 0.0706, "step": 10140 }, { "grad_norm": 0.7113622426986694, "learning_rate": 9.839349500764072e-05, "loss": 0.0925, "step": 10150 }, { "grad_norm": 0.988434910774231, "learning_rate": 9.83882937799226e-05, "loss": 0.0773, "step": 10160 }, { "grad_norm": 0.7277330756187439, "learning_rate": 9.838308428396849e-05, "loss": 0.0781, "step": 10170 }, { "grad_norm": 0.5391786098480225, "learning_rate": 9.837786652066854e-05, "loss": 0.0932, "step": 10180 }, { "grad_norm": 0.6520776152610779, "learning_rate": 9.837264049091437e-05, "loss": 0.1012, "step": 10190 }, { "grad_norm": 0.611186146736145, "learning_rate": 9.836740619559893e-05, "loss": 0.0862, "step": 10200 }, { "grad_norm": 0.4251430928707123, "learning_rate": 9.836216363561659e-05, "loss": 0.0913, "step": 10210 }, { "grad_norm": 0.5765807032585144, "learning_rate": 9.835691281186322e-05, "loss": 0.0701, "step": 10220 }, { "grad_norm": 0.5226805806159973, "learning_rate": 9.8351653725236e-05, "loss": 0.0893, "step": 10230 }, { "grad_norm": 0.39136624336242676, "learning_rate": 9.83463863766336e-05, "loss": 0.0712, "step": 10240 }, { "grad_norm": 0.5685550570487976, "learning_rate": 9.834111076695602e-05, "loss": 0.0735, "step": 10250 }, { "grad_norm": 0.7732744216918945, "learning_rate": 9.833582689710477e-05, "loss": 0.0795, "step": 10260 }, { "grad_norm": 0.4934269189834595, "learning_rate": 9.833053476798268e-05, "loss": 0.084, "step": 10270 }, { "grad_norm": 0.5162762403488159, "learning_rate": 9.832523438049404e-05, "loss": 0.0822, "step": 10280 }, { "grad_norm": 0.4971756041049957, "learning_rate": 9.831992573554454e-05, "loss": 0.0783, "step": 10290 }, { "grad_norm": 0.6337680220603943, "learning_rate": 9.831460883404128e-05, "loss": 0.0778, "step": 10300 }, { "grad_norm": 0.6320177316665649, "learning_rate": 9.830928367689278e-05, "loss": 0.0719, "step": 10310 }, { "grad_norm": 0.5558650493621826, "learning_rate": 9.830395026500896e-05, "loss": 0.0865, "step": 10320 }, { "grad_norm": 0.8340535759925842, "learning_rate": 9.829860859930115e-05, "loss": 0.091, "step": 10330 }, { "grad_norm": 0.8027603030204773, "learning_rate": 9.829325868068212e-05, "loss": 0.0862, "step": 10340 }, { "grad_norm": 0.5179392695426941, "learning_rate": 9.8287900510066e-05, "loss": 0.0985, "step": 10350 }, { "grad_norm": 0.4738132655620575, "learning_rate": 9.828253408836834e-05, "loss": 0.0682, "step": 10360 }, { "grad_norm": 0.6378704309463501, "learning_rate": 9.827715941650615e-05, "loss": 0.0908, "step": 10370 }, { "grad_norm": 0.6577363014221191, "learning_rate": 9.82717764953978e-05, "loss": 0.0821, "step": 10380 }, { "grad_norm": 0.6701004505157471, "learning_rate": 9.826638532596308e-05, "loss": 0.0787, "step": 10390 }, { "grad_norm": 0.4846786558628082, "learning_rate": 9.82609859091232e-05, "loss": 0.088, "step": 10400 }, { "grad_norm": 0.5606961846351624, "learning_rate": 9.825557824580076e-05, "loss": 0.0858, "step": 10410 }, { "grad_norm": 0.3989768326282501, "learning_rate": 9.82501623369198e-05, "loss": 0.078, "step": 10420 }, { "grad_norm": 0.8423611521720886, "learning_rate": 9.824473818340574e-05, "loss": 0.0699, "step": 10430 }, { "grad_norm": 0.6343322992324829, "learning_rate": 9.823930578618541e-05, "loss": 0.0722, "step": 10440 }, { "grad_norm": 0.6441065073013306, "learning_rate": 9.823386514618709e-05, "loss": 0.0863, "step": 10450 }, { "grad_norm": 0.576586902141571, "learning_rate": 9.82284162643404e-05, "loss": 0.0725, "step": 10460 }, { "grad_norm": 0.49958524107933044, "learning_rate": 9.822295914157642e-05, "loss": 0.0679, "step": 10470 }, { "grad_norm": 0.412800133228302, "learning_rate": 9.821749377882763e-05, "loss": 0.0699, "step": 10480 }, { "grad_norm": 0.47992992401123047, "learning_rate": 9.821202017702791e-05, "loss": 0.0674, "step": 10490 }, { "grad_norm": 0.5448235869407654, "learning_rate": 9.820653833711253e-05, "loss": 0.0653, "step": 10500 }, { "grad_norm": 0.6277552247047424, "learning_rate": 9.820104826001822e-05, "loss": 0.0821, "step": 10510 }, { "grad_norm": 0.42160260677337646, "learning_rate": 9.819554994668305e-05, "loss": 0.0645, "step": 10520 }, { "grad_norm": 0.5122050642967224, "learning_rate": 9.819004339804654e-05, "loss": 0.0733, "step": 10530 }, { "grad_norm": 0.4826473891735077, "learning_rate": 9.818452861504961e-05, "loss": 0.073, "step": 10540 }, { "grad_norm": 0.5479350686073303, "learning_rate": 9.81790055986346e-05, "loss": 0.0749, "step": 10550 }, { "grad_norm": 0.4311877191066742, "learning_rate": 9.817347434974523e-05, "loss": 0.0649, "step": 10560 }, { "grad_norm": 0.38980942964553833, "learning_rate": 9.816793486932664e-05, "loss": 0.0646, "step": 10570 }, { "grad_norm": 0.4612318277359009, "learning_rate": 9.816238715832538e-05, "loss": 0.0821, "step": 10580 }, { "grad_norm": 0.5589932203292847, "learning_rate": 9.815683121768939e-05, "loss": 0.0755, "step": 10590 }, { "grad_norm": 0.5111701488494873, "learning_rate": 9.815126704836804e-05, "loss": 0.0876, "step": 10600 }, { "grad_norm": 0.505352258682251, "learning_rate": 9.81456946513121e-05, "loss": 0.0975, "step": 10610 }, { "grad_norm": 0.5680928230285645, "learning_rate": 9.814011402747373e-05, "loss": 0.0936, "step": 10620 }, { "grad_norm": 0.6330680251121521, "learning_rate": 9.813452517780651e-05, "loss": 0.0654, "step": 10630 }, { "grad_norm": 0.3565552234649658, "learning_rate": 9.81289281032654e-05, "loss": 0.0805, "step": 10640 }, { "grad_norm": 0.5672487616539001, "learning_rate": 9.812332280480683e-05, "loss": 0.0874, "step": 10650 }, { "grad_norm": 0.42911192774772644, "learning_rate": 9.811770928338854e-05, "loss": 0.0896, "step": 10660 }, { "grad_norm": 0.6882103085517883, "learning_rate": 9.811208753996979e-05, "loss": 0.0917, "step": 10670 }, { "grad_norm": 0.3996044993400574, "learning_rate": 9.810645757551113e-05, "loss": 0.0679, "step": 10680 }, { "grad_norm": 0.4706616997718811, "learning_rate": 9.810081939097459e-05, "loss": 0.0672, "step": 10690 }, { "grad_norm": 0.4916982054710388, "learning_rate": 9.809517298732356e-05, "loss": 0.0986, "step": 10700 }, { "grad_norm": 0.600177526473999, "learning_rate": 9.80895183655229e-05, "loss": 0.0743, "step": 10710 }, { "grad_norm": 0.44162407517433167, "learning_rate": 9.808385552653877e-05, "loss": 0.0714, "step": 10720 }, { "grad_norm": 0.5828859806060791, "learning_rate": 9.807818447133886e-05, "loss": 0.0722, "step": 10730 }, { "grad_norm": 0.4367298483848572, "learning_rate": 9.807250520089215e-05, "loss": 0.0632, "step": 10740 }, { "grad_norm": 0.627566933631897, "learning_rate": 9.806681771616908e-05, "loss": 0.0858, "step": 10750 }, { "grad_norm": 0.5085509419441223, "learning_rate": 9.80611220181415e-05, "loss": 0.0791, "step": 10760 }, { "grad_norm": 0.6521008014678955, "learning_rate": 9.805541810778264e-05, "loss": 0.0783, "step": 10770 }, { "grad_norm": 0.49248215556144714, "learning_rate": 9.804970598606716e-05, "loss": 0.0791, "step": 10780 }, { "grad_norm": 0.6172441840171814, "learning_rate": 9.804398565397106e-05, "loss": 0.0776, "step": 10790 }, { "grad_norm": 0.5373210906982422, "learning_rate": 9.803825711247183e-05, "loss": 0.0834, "step": 10800 }, { "grad_norm": 0.8411969542503357, "learning_rate": 9.803252036254831e-05, "loss": 0.0596, "step": 10810 }, { "grad_norm": 1.7074166536331177, "learning_rate": 9.802677540518076e-05, "loss": 0.0809, "step": 10820 }, { "grad_norm": 0.5723106861114502, "learning_rate": 9.802102224135081e-05, "loss": 0.0852, "step": 10830 }, { "grad_norm": 0.5984023213386536, "learning_rate": 9.801526087204155e-05, "loss": 0.0909, "step": 10840 }, { "grad_norm": 0.6116058826446533, "learning_rate": 9.800949129823743e-05, "loss": 0.0757, "step": 10850 }, { "grad_norm": 0.43418821692466736, "learning_rate": 9.80037135209243e-05, "loss": 0.074, "step": 10860 }, { "grad_norm": 0.7012835144996643, "learning_rate": 9.799792754108946e-05, "loss": 0.0825, "step": 10870 }, { "grad_norm": 0.7359060645103455, "learning_rate": 9.799213335972152e-05, "loss": 0.0979, "step": 10880 }, { "grad_norm": 0.46958959102630615, "learning_rate": 9.798633097781058e-05, "loss": 0.0891, "step": 10890 }, { "grad_norm": 0.5832998156547546, "learning_rate": 9.79805203963481e-05, "loss": 0.0806, "step": 10900 }, { "grad_norm": 0.5285602807998657, "learning_rate": 9.797470161632697e-05, "loss": 0.0824, "step": 10910 }, { "grad_norm": 0.5182202458381653, "learning_rate": 9.796887463874145e-05, "loss": 0.0797, "step": 10920 }, { "grad_norm": 0.38842278718948364, "learning_rate": 9.796303946458718e-05, "loss": 0.0835, "step": 10930 }, { "grad_norm": 0.796804666519165, "learning_rate": 9.795719609486127e-05, "loss": 0.0974, "step": 10940 }, { "grad_norm": 0.5585605502128601, "learning_rate": 9.795134453056219e-05, "loss": 0.0762, "step": 10950 }, { "grad_norm": 0.5850632786750793, "learning_rate": 9.794548477268979e-05, "loss": 0.0661, "step": 10960 }, { "grad_norm": 0.48644909262657166, "learning_rate": 9.793961682224537e-05, "loss": 0.0622, "step": 10970 }, { "grad_norm": 0.552096962928772, "learning_rate": 9.793374068023156e-05, "loss": 0.0808, "step": 10980 }, { "grad_norm": 0.5264179706573486, "learning_rate": 9.792785634765247e-05, "loss": 0.0771, "step": 10990 }, { "grad_norm": 0.5910284519195557, "learning_rate": 9.792196382551357e-05, "loss": 0.0764, "step": 11000 }, { "grad_norm": 0.3499404489994049, "learning_rate": 9.791606311482171e-05, "loss": 0.0698, "step": 11010 }, { "grad_norm": 0.646920919418335, "learning_rate": 9.791015421658518e-05, "loss": 0.0789, "step": 11020 }, { "grad_norm": 0.6192008256912231, "learning_rate": 9.790423713181362e-05, "loss": 0.0816, "step": 11030 }, { "grad_norm": 0.5072930455207825, "learning_rate": 9.789831186151814e-05, "loss": 0.0765, "step": 11040 }, { "grad_norm": 0.6396548748016357, "learning_rate": 9.789237840671118e-05, "loss": 0.0669, "step": 11050 }, { "grad_norm": 0.7279786467552185, "learning_rate": 9.78864367684066e-05, "loss": 0.0893, "step": 11060 }, { "grad_norm": 0.7248376607894897, "learning_rate": 9.788048694761968e-05, "loss": 0.0736, "step": 11070 }, { "grad_norm": 0.45773446559906006, "learning_rate": 9.787452894536709e-05, "loss": 0.0878, "step": 11080 }, { "grad_norm": 0.7290059924125671, "learning_rate": 9.786856276266685e-05, "loss": 0.0656, "step": 11090 }, { "grad_norm": 0.5485090613365173, "learning_rate": 9.786258840053845e-05, "loss": 0.0812, "step": 11100 }, { "grad_norm": 0.4460821747779846, "learning_rate": 9.785660586000273e-05, "loss": 0.0762, "step": 11110 }, { "grad_norm": 0.5007906556129456, "learning_rate": 9.785061514208196e-05, "loss": 0.0639, "step": 11120 }, { "grad_norm": 0.44034698605537415, "learning_rate": 9.784461624779977e-05, "loss": 0.0812, "step": 11130 }, { "grad_norm": 0.3983391523361206, "learning_rate": 9.783860917818123e-05, "loss": 0.0599, "step": 11140 }, { "grad_norm": 0.4222573935985565, "learning_rate": 9.783259393425277e-05, "loss": 0.0622, "step": 11150 }, { "grad_norm": 0.5932636260986328, "learning_rate": 9.782657051704221e-05, "loss": 0.0727, "step": 11160 }, { "grad_norm": 0.6944952011108398, "learning_rate": 9.782053892757883e-05, "loss": 0.0687, "step": 11170 }, { "grad_norm": 0.7155230641365051, "learning_rate": 9.781449916689324e-05, "loss": 0.0921, "step": 11180 }, { "grad_norm": 0.5389136075973511, "learning_rate": 9.780845123601746e-05, "loss": 0.0823, "step": 11190 }, { "grad_norm": 0.49120184779167175, "learning_rate": 9.780239513598492e-05, "loss": 0.0835, "step": 11200 }, { "grad_norm": 0.5790382027626038, "learning_rate": 9.779633086783047e-05, "loss": 0.07, "step": 11210 }, { "grad_norm": 0.46409937739372253, "learning_rate": 9.779025843259031e-05, "loss": 0.0774, "step": 11220 }, { "grad_norm": 0.43792280554771423, "learning_rate": 9.778417783130204e-05, "loss": 0.0851, "step": 11230 }, { "grad_norm": 0.6539749503135681, "learning_rate": 9.777808906500468e-05, "loss": 0.0754, "step": 11240 }, { "grad_norm": 0.6171899437904358, "learning_rate": 9.777199213473862e-05, "loss": 0.0637, "step": 11250 }, { "grad_norm": 0.5753228664398193, "learning_rate": 9.77658870415457e-05, "loss": 0.0611, "step": 11260 }, { "grad_norm": 0.4628838300704956, "learning_rate": 9.775977378646906e-05, "loss": 0.0796, "step": 11270 }, { "grad_norm": 0.6418676376342773, "learning_rate": 9.775365237055331e-05, "loss": 0.0851, "step": 11280 }, { "grad_norm": 0.6440351009368896, "learning_rate": 9.774752279484445e-05, "loss": 0.0743, "step": 11290 }, { "grad_norm": 0.4761498272418976, "learning_rate": 9.774138506038984e-05, "loss": 0.0639, "step": 11300 }, { "grad_norm": 0.6570431590080261, "learning_rate": 9.773523916823826e-05, "loss": 0.0777, "step": 11310 }, { "grad_norm": 0.46304452419281006, "learning_rate": 9.772908511943986e-05, "loss": 0.0688, "step": 11320 }, { "grad_norm": 0.43416905403137207, "learning_rate": 9.77229229150462e-05, "loss": 0.0661, "step": 11330 }, { "grad_norm": 0.4448212683200836, "learning_rate": 9.771675255611024e-05, "loss": 0.0617, "step": 11340 }, { "grad_norm": 0.38071486353874207, "learning_rate": 9.771057404368632e-05, "loss": 0.0859, "step": 11350 }, { "grad_norm": 0.5276381969451904, "learning_rate": 9.770438737883018e-05, "loss": 0.0809, "step": 11360 }, { "grad_norm": 0.5758881568908691, "learning_rate": 9.769819256259898e-05, "loss": 0.067, "step": 11370 }, { "grad_norm": 0.3923940360546112, "learning_rate": 9.769198959605119e-05, "loss": 0.075, "step": 11380 }, { "grad_norm": 0.517054557800293, "learning_rate": 9.768577848024678e-05, "loss": 0.0876, "step": 11390 }, { "grad_norm": 0.4745261073112488, "learning_rate": 9.767955921624702e-05, "loss": 0.0845, "step": 11400 }, { "grad_norm": 0.5826627016067505, "learning_rate": 9.767333180511465e-05, "loss": 0.0761, "step": 11410 }, { "grad_norm": 0.37922611832618713, "learning_rate": 9.766709624791373e-05, "loss": 0.0817, "step": 11420 }, { "grad_norm": 0.53538578748703, "learning_rate": 9.766085254570975e-05, "loss": 0.0727, "step": 11430 }, { "grad_norm": 0.4927736520767212, "learning_rate": 9.76546006995696e-05, "loss": 0.0741, "step": 11440 }, { "grad_norm": 0.666059672832489, "learning_rate": 9.764834071056155e-05, "loss": 0.073, "step": 11450 }, { "grad_norm": 0.5073360800743103, "learning_rate": 9.764207257975526e-05, "loss": 0.0739, "step": 11460 }, { "grad_norm": 0.7291958928108215, "learning_rate": 9.763579630822179e-05, "loss": 0.0697, "step": 11470 }, { "grad_norm": 0.5906637907028198, "learning_rate": 9.762951189703356e-05, "loss": 0.0667, "step": 11480 }, { "grad_norm": 0.5694571137428284, "learning_rate": 9.762321934726442e-05, "loss": 0.0745, "step": 11490 }, { "grad_norm": 0.49660104513168335, "learning_rate": 9.761691865998959e-05, "loss": 0.0656, "step": 11500 }, { "grad_norm": 0.39121827483177185, "learning_rate": 9.76106098362857e-05, "loss": 0.0623, "step": 11510 }, { "grad_norm": 0.39582934975624084, "learning_rate": 9.760429287723072e-05, "loss": 0.0703, "step": 11520 }, { "grad_norm": 0.5451925992965698, "learning_rate": 9.759796778390406e-05, "loss": 0.0778, "step": 11530 }, { "grad_norm": 0.4926786720752716, "learning_rate": 9.759163455738653e-05, "loss": 0.068, "step": 11540 }, { "grad_norm": 0.5379854440689087, "learning_rate": 9.75852931987603e-05, "loss": 0.0752, "step": 11550 }, { "grad_norm": 0.36711734533309937, "learning_rate": 9.757894370910891e-05, "loss": 0.0744, "step": 11560 }, { "grad_norm": 0.5181512832641602, "learning_rate": 9.757258608951733e-05, "loss": 0.0677, "step": 11570 }, { "grad_norm": 0.6176525950431824, "learning_rate": 9.75662203410719e-05, "loss": 0.0839, "step": 11580 }, { "grad_norm": 0.5322485566139221, "learning_rate": 9.755984646486034e-05, "loss": 0.0815, "step": 11590 }, { "grad_norm": 0.5694640874862671, "learning_rate": 9.75534644619718e-05, "loss": 0.0693, "step": 11600 }, { "grad_norm": 0.5337085723876953, "learning_rate": 9.754707433349676e-05, "loss": 0.0927, "step": 11610 }, { "grad_norm": 0.5417370200157166, "learning_rate": 9.754067608052715e-05, "loss": 0.0845, "step": 11620 }, { "grad_norm": 0.4477488100528717, "learning_rate": 9.753426970415622e-05, "loss": 0.1085, "step": 11630 }, { "grad_norm": 0.40797311067581177, "learning_rate": 9.752785520547868e-05, "loss": 0.0724, "step": 11640 }, { "grad_norm": 0.4612121284008026, "learning_rate": 9.752143258559056e-05, "loss": 0.0747, "step": 11650 }, { "grad_norm": 0.4136880338191986, "learning_rate": 9.751500184558933e-05, "loss": 0.0691, "step": 11660 }, { "grad_norm": 0.3307664692401886, "learning_rate": 9.750856298657383e-05, "loss": 0.0682, "step": 11670 }, { "grad_norm": 0.5530723929405212, "learning_rate": 9.750211600964428e-05, "loss": 0.0845, "step": 11680 }, { "grad_norm": 0.590667724609375, "learning_rate": 9.749566091590226e-05, "loss": 0.0711, "step": 11690 }, { "grad_norm": 0.4696106016635895, "learning_rate": 9.748919770645083e-05, "loss": 0.0726, "step": 11700 }, { "grad_norm": 0.37475964426994324, "learning_rate": 9.748272638239432e-05, "loss": 0.0698, "step": 11710 }, { "grad_norm": 0.4326804280281067, "learning_rate": 9.747624694483855e-05, "loss": 0.0727, "step": 11720 }, { "grad_norm": 0.45364564657211304, "learning_rate": 9.746975939489065e-05, "loss": 0.0604, "step": 11730 }, { "grad_norm": 0.4562254250049591, "learning_rate": 9.746326373365918e-05, "loss": 0.0627, "step": 11740 }, { "grad_norm": 0.7155741453170776, "learning_rate": 9.745675996225403e-05, "loss": 0.0737, "step": 11750 }, { "grad_norm": 0.3576071858406067, "learning_rate": 9.745024808178657e-05, "loss": 0.0658, "step": 11760 }, { "grad_norm": 0.7086811661720276, "learning_rate": 9.744372809336947e-05, "loss": 0.0816, "step": 11770 }, { "grad_norm": 0.5903370380401611, "learning_rate": 9.743719999811682e-05, "loss": 0.0649, "step": 11780 }, { "grad_norm": 0.4084456264972687, "learning_rate": 9.743066379714412e-05, "loss": 0.0649, "step": 11790 }, { "grad_norm": 0.4763876497745514, "learning_rate": 9.74241194915682e-05, "loss": 0.0647, "step": 11800 }, { "grad_norm": 0.45195862650871277, "learning_rate": 9.741756708250731e-05, "loss": 0.0814, "step": 11810 }, { "grad_norm": 0.4480753540992737, "learning_rate": 9.741100657108109e-05, "loss": 0.0699, "step": 11820 }, { "grad_norm": 0.5955126285552979, "learning_rate": 9.740443795841054e-05, "loss": 0.0827, "step": 11830 }, { "grad_norm": 0.5698202848434448, "learning_rate": 9.739786124561805e-05, "loss": 0.0719, "step": 11840 }, { "grad_norm": 0.41384971141815186, "learning_rate": 9.73912764338274e-05, "loss": 0.074, "step": 11850 }, { "grad_norm": 0.38490769267082214, "learning_rate": 9.738468352416377e-05, "loss": 0.0794, "step": 11860 }, { "grad_norm": 0.5717015266418457, "learning_rate": 9.737808251775369e-05, "loss": 0.0725, "step": 11870 }, { "grad_norm": 0.3372573256492615, "learning_rate": 9.737147341572512e-05, "loss": 0.0888, "step": 11880 }, { "grad_norm": 0.5079198479652405, "learning_rate": 9.736485621920735e-05, "loss": 0.0738, "step": 11890 }, { "grad_norm": 0.45087894797325134, "learning_rate": 9.735823092933108e-05, "loss": 0.0751, "step": 11900 }, { "grad_norm": 0.5806047916412354, "learning_rate": 9.735159754722838e-05, "loss": 0.0727, "step": 11910 }, { "grad_norm": 0.4418105185031891, "learning_rate": 9.734495607403275e-05, "loss": 0.0678, "step": 11920 }, { "grad_norm": 0.5616791844367981, "learning_rate": 9.733830651087901e-05, "loss": 0.0782, "step": 11930 }, { "grad_norm": 0.46161791682243347, "learning_rate": 9.733164885890338e-05, "loss": 0.0691, "step": 11940 }, { "grad_norm": 0.41420778632164, "learning_rate": 9.732498311924349e-05, "loss": 0.0661, "step": 11950 }, { "grad_norm": 0.5019378662109375, "learning_rate": 9.731830929303833e-05, "loss": 0.0674, "step": 11960 }, { "grad_norm": 0.5741440057754517, "learning_rate": 9.731162738142827e-05, "loss": 0.0755, "step": 11970 }, { "grad_norm": 0.5850145816802979, "learning_rate": 9.730493738555506e-05, "loss": 0.0747, "step": 11980 }, { "grad_norm": 0.48797735571861267, "learning_rate": 9.729823930656186e-05, "loss": 0.0793, "step": 11990 }, { "grad_norm": 0.6049846410751343, "learning_rate": 9.729153314559316e-05, "loss": 0.0827, "step": 12000 }, { "grad_norm": 0.5481655597686768, "learning_rate": 9.728481890379486e-05, "loss": 0.0713, "step": 12010 }, { "grad_norm": 0.5114384293556213, "learning_rate": 9.727809658231428e-05, "loss": 0.063, "step": 12020 }, { "grad_norm": 0.49359166622161865, "learning_rate": 9.727136618230003e-05, "loss": 0.0745, "step": 12030 }, { "grad_norm": 0.6478204131126404, "learning_rate": 9.726462770490219e-05, "loss": 0.073, "step": 12040 }, { "grad_norm": 0.5604483485221863, "learning_rate": 9.725788115127214e-05, "loss": 0.0661, "step": 12050 }, { "grad_norm": 0.3967888355255127, "learning_rate": 9.725112652256274e-05, "loss": 0.0738, "step": 12060 }, { "grad_norm": 0.6299535036087036, "learning_rate": 9.724436381992812e-05, "loss": 0.091, "step": 12070 }, { "grad_norm": 0.6001694202423096, "learning_rate": 9.723759304452387e-05, "loss": 0.0658, "step": 12080 }, { "grad_norm": 0.46143779158592224, "learning_rate": 9.72308141975069e-05, "loss": 0.0672, "step": 12090 }, { "grad_norm": 0.42371809482574463, "learning_rate": 9.722402728003557e-05, "loss": 0.0816, "step": 12100 }, { "grad_norm": 0.557477593421936, "learning_rate": 9.721723229326953e-05, "loss": 0.0723, "step": 12110 }, { "grad_norm": 0.45315366983413696, "learning_rate": 9.721042923836992e-05, "loss": 0.0629, "step": 12120 }, { "grad_norm": 0.4430418610572815, "learning_rate": 9.720361811649914e-05, "loss": 0.0779, "step": 12130 }, { "grad_norm": 0.4417875111103058, "learning_rate": 9.719679892882106e-05, "loss": 0.0674, "step": 12140 }, { "grad_norm": 0.4899342656135559, "learning_rate": 9.718997167650085e-05, "loss": 0.0672, "step": 12150 }, { "grad_norm": 0.6443845629692078, "learning_rate": 9.718313636070515e-05, "loss": 0.0603, "step": 12160 }, { "grad_norm": 0.5194249153137207, "learning_rate": 9.717629298260192e-05, "loss": 0.061, "step": 12170 }, { "grad_norm": 0.6347920894622803, "learning_rate": 9.716944154336047e-05, "loss": 0.0697, "step": 12180 }, { "grad_norm": 0.623352587223053, "learning_rate": 9.716258204415157e-05, "loss": 0.0721, "step": 12190 }, { "grad_norm": 0.39790791273117065, "learning_rate": 9.715571448614728e-05, "loss": 0.0691, "step": 12200 }, { "grad_norm": 0.38249337673187256, "learning_rate": 9.71488388705211e-05, "loss": 0.0938, "step": 12210 }, { "grad_norm": 0.5785043835639954, "learning_rate": 9.714195519844788e-05, "loss": 0.0658, "step": 12220 }, { "grad_norm": 0.47546708583831787, "learning_rate": 9.713506347110386e-05, "loss": 0.0718, "step": 12230 }, { "grad_norm": 0.5457324385643005, "learning_rate": 9.712816368966662e-05, "loss": 0.0768, "step": 12240 }, { "grad_norm": 0.4431699812412262, "learning_rate": 9.712125585531517e-05, "loss": 0.0837, "step": 12250 }, { "grad_norm": 0.6451482176780701, "learning_rate": 9.711433996922988e-05, "loss": 0.0847, "step": 12260 }, { "grad_norm": 0.4200718402862549, "learning_rate": 9.710741603259245e-05, "loss": 0.0743, "step": 12270 }, { "grad_norm": 0.4665679335594177, "learning_rate": 9.710048404658603e-05, "loss": 0.0688, "step": 12280 }, { "grad_norm": 0.4400307238101959, "learning_rate": 9.709354401239508e-05, "loss": 0.0802, "step": 12290 }, { "grad_norm": 0.48775818943977356, "learning_rate": 9.708659593120546e-05, "loss": 0.0704, "step": 12300 }, { "grad_norm": 0.5917724967002869, "learning_rate": 9.707963980420443e-05, "loss": 0.0707, "step": 12310 }, { "grad_norm": 0.6978253126144409, "learning_rate": 9.707267563258058e-05, "loss": 0.0767, "step": 12320 }, { "grad_norm": 0.5554829239845276, "learning_rate": 9.70657034175239e-05, "loss": 0.0691, "step": 12330 }, { "grad_norm": 0.5329478979110718, "learning_rate": 9.705872316022577e-05, "loss": 0.0712, "step": 12340 }, { "grad_norm": 0.6951595544815063, "learning_rate": 9.705173486187891e-05, "loss": 0.0673, "step": 12350 }, { "grad_norm": 0.47553741931915283, "learning_rate": 9.704473852367741e-05, "loss": 0.0849, "step": 12360 }, { "grad_norm": 0.33269721269607544, "learning_rate": 9.70377341468168e-05, "loss": 0.0572, "step": 12370 }, { "grad_norm": 0.4202188551425934, "learning_rate": 9.703072173249389e-05, "loss": 0.0772, "step": 12380 }, { "grad_norm": 0.5906175971031189, "learning_rate": 9.702370128190693e-05, "loss": 0.0761, "step": 12390 }, { "grad_norm": 0.5024974346160889, "learning_rate": 9.701667279625552e-05, "loss": 0.0789, "step": 12400 }, { "grad_norm": 0.6100894212722778, "learning_rate": 9.700963627674065e-05, "loss": 0.0783, "step": 12410 }, { "grad_norm": 0.4065770208835602, "learning_rate": 9.700259172456466e-05, "loss": 0.0557, "step": 12420 }, { "grad_norm": 0.5795661807060242, "learning_rate": 9.699553914093124e-05, "loss": 0.0794, "step": 12430 }, { "grad_norm": 0.49635156989097595, "learning_rate": 9.698847852704553e-05, "loss": 0.0817, "step": 12440 }, { "grad_norm": 0.5384378433227539, "learning_rate": 9.6981409884114e-05, "loss": 0.0697, "step": 12450 }, { "grad_norm": 0.591677188873291, "learning_rate": 9.697433321334443e-05, "loss": 0.0815, "step": 12460 }, { "grad_norm": 0.5954530239105225, "learning_rate": 9.696724851594607e-05, "loss": 0.074, "step": 12470 }, { "grad_norm": 0.4453452229499817, "learning_rate": 9.696015579312952e-05, "loss": 0.0609, "step": 12480 }, { "grad_norm": 0.5399270057678223, "learning_rate": 9.695305504610668e-05, "loss": 0.0735, "step": 12490 }, { "grad_norm": 0.39752113819122314, "learning_rate": 9.694594627609092e-05, "loss": 0.0815, "step": 12500 }, { "grad_norm": 0.5044263601303101, "learning_rate": 9.693882948429691e-05, "loss": 0.082, "step": 12510 }, { "grad_norm": 0.4128181040287018, "learning_rate": 9.693170467194071e-05, "loss": 0.0744, "step": 12520 }, { "grad_norm": 0.37434667348861694, "learning_rate": 9.692457184023977e-05, "loss": 0.0692, "step": 12530 }, { "grad_norm": 0.62531578540802, "learning_rate": 9.691743099041291e-05, "loss": 0.067, "step": 12540 }, { "grad_norm": 0.5074208378791809, "learning_rate": 9.691028212368027e-05, "loss": 0.0739, "step": 12550 }, { "grad_norm": 0.6616808772087097, "learning_rate": 9.690312524126342e-05, "loss": 0.0662, "step": 12560 }, { "grad_norm": 0.4536252021789551, "learning_rate": 9.689596034438527e-05, "loss": 0.065, "step": 12570 }, { "grad_norm": 0.4211670756340027, "learning_rate": 9.688878743427012e-05, "loss": 0.0641, "step": 12580 }, { "grad_norm": 0.46462321281433105, "learning_rate": 9.688160651214359e-05, "loss": 0.0708, "step": 12590 }, { "grad_norm": 0.45246171951293945, "learning_rate": 9.687441757923273e-05, "loss": 0.0735, "step": 12600 }, { "grad_norm": 0.4830716848373413, "learning_rate": 9.68672206367659e-05, "loss": 0.0666, "step": 12610 }, { "grad_norm": 0.4149150550365448, "learning_rate": 9.686001568597291e-05, "loss": 0.0656, "step": 12620 }, { "grad_norm": 0.5499757528305054, "learning_rate": 9.685280272808486e-05, "loss": 0.06, "step": 12630 }, { "grad_norm": 0.5547691583633423, "learning_rate": 9.684558176433424e-05, "loss": 0.0824, "step": 12640 }, { "grad_norm": 0.4891979396343231, "learning_rate": 9.683835279595495e-05, "loss": 0.0676, "step": 12650 }, { "grad_norm": 0.45783451199531555, "learning_rate": 9.683111582418216e-05, "loss": 0.0697, "step": 12660 }, { "grad_norm": 0.45847946405410767, "learning_rate": 9.682387085025254e-05, "loss": 0.0638, "step": 12670 }, { "grad_norm": 0.7029893398284912, "learning_rate": 9.681661787540401e-05, "loss": 0.0777, "step": 12680 }, { "grad_norm": 0.43191900849342346, "learning_rate": 9.680935690087593e-05, "loss": 0.0644, "step": 12690 }, { "grad_norm": 0.43623673915863037, "learning_rate": 9.680208792790901e-05, "loss": 0.0639, "step": 12700 }, { "grad_norm": 0.5437018275260925, "learning_rate": 9.679481095774529e-05, "loss": 0.067, "step": 12710 }, { "grad_norm": 0.5263077616691589, "learning_rate": 9.678752599162822e-05, "loss": 0.072, "step": 12720 }, { "grad_norm": 0.6479591131210327, "learning_rate": 9.678023303080259e-05, "loss": 0.0729, "step": 12730 }, { "grad_norm": 0.5358262658119202, "learning_rate": 9.677293207651459e-05, "loss": 0.0702, "step": 12740 }, { "grad_norm": 0.4779358506202698, "learning_rate": 9.676562313001173e-05, "loss": 0.0642, "step": 12750 }, { "grad_norm": 0.5400911569595337, "learning_rate": 9.675830619254293e-05, "loss": 0.0868, "step": 12760 }, { "grad_norm": 0.6613786816596985, "learning_rate": 9.675098126535843e-05, "loss": 0.0905, "step": 12770 }, { "grad_norm": 0.5532199740409851, "learning_rate": 9.674364834970988e-05, "loss": 0.0845, "step": 12780 }, { "grad_norm": 0.4952218234539032, "learning_rate": 9.673630744685028e-05, "loss": 0.0807, "step": 12790 }, { "grad_norm": 0.5050411820411682, "learning_rate": 9.672895855803397e-05, "loss": 0.0714, "step": 12800 }, { "grad_norm": 0.5447990894317627, "learning_rate": 9.672160168451667e-05, "loss": 0.0713, "step": 12810 }, { "grad_norm": 0.4340939521789551, "learning_rate": 9.671423682755549e-05, "loss": 0.0652, "step": 12820 }, { "grad_norm": 0.6196721196174622, "learning_rate": 9.670686398840888e-05, "loss": 0.0666, "step": 12830 }, { "grad_norm": 0.5377069711685181, "learning_rate": 9.669948316833664e-05, "loss": 0.0551, "step": 12840 }, { "grad_norm": 0.40862929821014404, "learning_rate": 9.669209436859997e-05, "loss": 0.0581, "step": 12850 }, { "grad_norm": 0.5965743064880371, "learning_rate": 9.66846975904614e-05, "loss": 0.0731, "step": 12860 }, { "grad_norm": 0.44210752844810486, "learning_rate": 9.667729283518483e-05, "loss": 0.0578, "step": 12870 }, { "grad_norm": 0.5460032820701599, "learning_rate": 9.666988010403557e-05, "loss": 0.0592, "step": 12880 }, { "grad_norm": 0.6229017376899719, "learning_rate": 9.66624593982802e-05, "loss": 0.0701, "step": 12890 }, { "grad_norm": 0.47698307037353516, "learning_rate": 9.665503071918675e-05, "loss": 0.0797, "step": 12900 }, { "grad_norm": 0.4074243903160095, "learning_rate": 9.664759406802456e-05, "loss": 0.0653, "step": 12910 }, { "grad_norm": 0.4913080334663391, "learning_rate": 9.664014944606437e-05, "loss": 0.0551, "step": 12920 }, { "grad_norm": 0.4545650780200958, "learning_rate": 9.663269685457822e-05, "loss": 0.0609, "step": 12930 }, { "grad_norm": 0.5486201643943787, "learning_rate": 9.662523629483962e-05, "loss": 0.0686, "step": 12940 }, { "grad_norm": 0.8041452765464783, "learning_rate": 9.661776776812333e-05, "loss": 0.0593, "step": 12950 }, { "grad_norm": 0.4873355031013489, "learning_rate": 9.661029127570553e-05, "loss": 0.0528, "step": 12960 }, { "grad_norm": 0.583015501499176, "learning_rate": 9.660280681886373e-05, "loss": 0.0658, "step": 12970 }, { "grad_norm": 0.4372885227203369, "learning_rate": 9.659531439887685e-05, "loss": 0.0531, "step": 12980 }, { "grad_norm": 0.464997261762619, "learning_rate": 9.658781401702511e-05, "loss": 0.0657, "step": 12990 }, { "grad_norm": 0.579846978187561, "learning_rate": 9.658030567459015e-05, "loss": 0.0642, "step": 13000 }, { "grad_norm": 0.5712960362434387, "learning_rate": 9.65727893728549e-05, "loss": 0.0785, "step": 13010 }, { "grad_norm": 0.5287594795227051, "learning_rate": 9.656526511310375e-05, "loss": 0.0785, "step": 13020 }, { "grad_norm": 0.43618282675743103, "learning_rate": 9.655773289662233e-05, "loss": 0.0764, "step": 13030 }, { "grad_norm": 0.706458568572998, "learning_rate": 9.655019272469772e-05, "loss": 0.0677, "step": 13040 }, { "grad_norm": 0.3390490412712097, "learning_rate": 9.654264459861832e-05, "loss": 0.0513, "step": 13050 }, { "grad_norm": 0.5077590346336365, "learning_rate": 9.653508851967391e-05, "loss": 0.0641, "step": 13060 }, { "grad_norm": 0.6085052490234375, "learning_rate": 9.65275244891556e-05, "loss": 0.073, "step": 13070 }, { "grad_norm": 0.4235933721065521, "learning_rate": 9.651995250835591e-05, "loss": 0.0621, "step": 13080 }, { "grad_norm": 0.48851415514945984, "learning_rate": 9.651237257856862e-05, "loss": 0.061, "step": 13090 }, { "grad_norm": 0.36975252628326416, "learning_rate": 9.6504784701089e-05, "loss": 0.0658, "step": 13100 }, { "grad_norm": 0.5245407223701477, "learning_rate": 9.649718887721357e-05, "loss": 0.0607, "step": 13110 }, { "grad_norm": 0.7517634630203247, "learning_rate": 9.648958510824028e-05, "loss": 0.0638, "step": 13120 }, { "grad_norm": 0.5186040997505188, "learning_rate": 9.648197339546837e-05, "loss": 0.0805, "step": 13130 }, { "grad_norm": 0.7235131859779358, "learning_rate": 9.647435374019851e-05, "loss": 0.0858, "step": 13140 }, { "grad_norm": 0.5959060788154602, "learning_rate": 9.646672614373266e-05, "loss": 0.0741, "step": 13150 }, { "grad_norm": 0.38542798161506653, "learning_rate": 9.645909060737418e-05, "loss": 0.057, "step": 13160 }, { "grad_norm": 0.5127326846122742, "learning_rate": 9.645144713242778e-05, "loss": 0.0668, "step": 13170 }, { "grad_norm": 0.43035340309143066, "learning_rate": 9.64437957201995e-05, "loss": 0.0596, "step": 13180 }, { "grad_norm": 0.38521552085876465, "learning_rate": 9.643613637199678e-05, "loss": 0.0547, "step": 13190 }, { "grad_norm": 0.5980406403541565, "learning_rate": 9.642846908912839e-05, "loss": 0.064, "step": 13200 }, { "grad_norm": 0.3953041136264801, "learning_rate": 9.642079387290444e-05, "loss": 0.0615, "step": 13210 }, { "grad_norm": 0.44431382417678833, "learning_rate": 9.641311072463644e-05, "loss": 0.0602, "step": 13220 }, { "grad_norm": 0.6241759657859802, "learning_rate": 9.640541964563722e-05, "loss": 0.0629, "step": 13230 }, { "grad_norm": 0.47209903597831726, "learning_rate": 9.639772063722096e-05, "loss": 0.0811, "step": 13240 }, { "grad_norm": 0.4858880341053009, "learning_rate": 9.639001370070324e-05, "loss": 0.0602, "step": 13250 }, { "grad_norm": 0.5763435363769531, "learning_rate": 9.638229883740095e-05, "loss": 0.0606, "step": 13260 }, { "grad_norm": 0.6294756531715393, "learning_rate": 9.637457604863233e-05, "loss": 0.0624, "step": 13270 }, { "grad_norm": 0.5360835790634155, "learning_rate": 9.636684533571703e-05, "loss": 0.0696, "step": 13280 }, { "grad_norm": 0.4720051884651184, "learning_rate": 9.635910669997599e-05, "loss": 0.062, "step": 13290 }, { "grad_norm": 0.4720982611179352, "learning_rate": 9.635136014273154e-05, "loss": 0.0599, "step": 13300 }, { "grad_norm": 0.4358536899089813, "learning_rate": 9.634360566530735e-05, "loss": 0.068, "step": 13310 }, { "grad_norm": 0.534062385559082, "learning_rate": 9.633584326902845e-05, "loss": 0.0795, "step": 13320 }, { "grad_norm": 0.5178856253623962, "learning_rate": 9.632807295522124e-05, "loss": 0.0595, "step": 13330 }, { "grad_norm": 0.5233091711997986, "learning_rate": 9.632029472521342e-05, "loss": 0.0662, "step": 13340 }, { "grad_norm": 0.4953891634941101, "learning_rate": 9.631250858033409e-05, "loss": 0.0608, "step": 13350 }, { "grad_norm": 0.4273463189601898, "learning_rate": 9.630471452191371e-05, "loss": 0.0586, "step": 13360 }, { "grad_norm": 0.4881856143474579, "learning_rate": 9.629691255128405e-05, "loss": 0.0742, "step": 13370 }, { "grad_norm": 0.5649183988571167, "learning_rate": 9.628910266977825e-05, "loss": 0.0707, "step": 13380 }, { "grad_norm": 0.36476370692253113, "learning_rate": 9.628128487873083e-05, "loss": 0.0742, "step": 13390 }, { "grad_norm": 0.5582756996154785, "learning_rate": 9.627345917947761e-05, "loss": 0.0729, "step": 13400 }, { "grad_norm": 0.558550238609314, "learning_rate": 9.626562557335579e-05, "loss": 0.06, "step": 13410 }, { "grad_norm": 0.3700283467769623, "learning_rate": 9.625778406170393e-05, "loss": 0.0635, "step": 13420 }, { "grad_norm": 0.5086910724639893, "learning_rate": 9.624993464586193e-05, "loss": 0.0617, "step": 13430 }, { "grad_norm": 0.43797725439071655, "learning_rate": 9.624207732717105e-05, "loss": 0.0616, "step": 13440 }, { "grad_norm": 0.6282510161399841, "learning_rate": 9.623421210697386e-05, "loss": 0.076, "step": 13450 }, { "grad_norm": 0.5842523574829102, "learning_rate": 9.622633898661434e-05, "loss": 0.0775, "step": 13460 }, { "grad_norm": 0.5573118925094604, "learning_rate": 9.621845796743778e-05, "loss": 0.0851, "step": 13470 }, { "grad_norm": 0.45400470495224, "learning_rate": 9.621056905079082e-05, "loss": 0.0698, "step": 13480 }, { "grad_norm": 0.45514431595802307, "learning_rate": 9.620267223802149e-05, "loss": 0.0721, "step": 13490 }, { "grad_norm": 0.6166884303092957, "learning_rate": 9.619476753047911e-05, "loss": 0.0568, "step": 13500 }, { "grad_norm": 0.988437294960022, "learning_rate": 9.618685492951438e-05, "loss": 0.0765, "step": 13510 }, { "grad_norm": 0.38977310061454773, "learning_rate": 9.617893443647938e-05, "loss": 0.0866, "step": 13520 }, { "grad_norm": 0.5133600234985352, "learning_rate": 9.617100605272746e-05, "loss": 0.073, "step": 13530 }, { "grad_norm": 0.5089312791824341, "learning_rate": 9.616306977961338e-05, "loss": 0.076, "step": 13540 }, { "grad_norm": 0.43331849575042725, "learning_rate": 9.615512561849326e-05, "loss": 0.0611, "step": 13550 }, { "grad_norm": 0.4403949975967407, "learning_rate": 9.61471735707245e-05, "loss": 0.08, "step": 13560 }, { "grad_norm": 0.4792535603046417, "learning_rate": 9.613921363766592e-05, "loss": 0.0633, "step": 13570 }, { "grad_norm": 0.4347432851791382, "learning_rate": 9.613124582067763e-05, "loss": 0.073, "step": 13580 }, { "grad_norm": 0.6267663240432739, "learning_rate": 9.612327012112112e-05, "loss": 0.0794, "step": 13590 }, { "grad_norm": 0.44754940271377563, "learning_rate": 9.611528654035921e-05, "loss": 0.0926, "step": 13600 }, { "grad_norm": 0.45278435945510864, "learning_rate": 9.610729507975611e-05, "loss": 0.0605, "step": 13610 }, { "grad_norm": 0.4675839841365814, "learning_rate": 9.609929574067731e-05, "loss": 0.0782, "step": 13620 }, { "grad_norm": 0.4977779984474182, "learning_rate": 9.609128852448967e-05, "loss": 0.0659, "step": 13630 }, { "grad_norm": 0.4376055598258972, "learning_rate": 9.608327343256143e-05, "loss": 0.0656, "step": 13640 }, { "grad_norm": 0.48533567786216736, "learning_rate": 9.607525046626216e-05, "loss": 0.0736, "step": 13650 }, { "grad_norm": 0.6227413415908813, "learning_rate": 9.606721962696272e-05, "loss": 0.0733, "step": 13660 }, { "grad_norm": 0.4637225866317749, "learning_rate": 9.60591809160354e-05, "loss": 0.0676, "step": 13670 }, { "grad_norm": 0.5015311241149902, "learning_rate": 9.605113433485378e-05, "loss": 0.0687, "step": 13680 }, { "grad_norm": 0.7189167737960815, "learning_rate": 9.604307988479279e-05, "loss": 0.0615, "step": 13690 }, { "grad_norm": 0.5538801550865173, "learning_rate": 9.603501756722876e-05, "loss": 0.0789, "step": 13700 }, { "grad_norm": 0.3978464901447296, "learning_rate": 9.602694738353927e-05, "loss": 0.0653, "step": 13710 }, { "grad_norm": 0.46740710735321045, "learning_rate": 9.601886933510331e-05, "loss": 0.0548, "step": 13720 }, { "grad_norm": 0.43458986282348633, "learning_rate": 9.60107834233012e-05, "loss": 0.0637, "step": 13730 }, { "grad_norm": 0.5704897046089172, "learning_rate": 9.60026896495146e-05, "loss": 0.059, "step": 13740 }, { "grad_norm": 0.33369454741477966, "learning_rate": 9.599458801512652e-05, "loss": 0.0543, "step": 13750 }, { "grad_norm": 0.5225869417190552, "learning_rate": 9.598647852152129e-05, "loss": 0.085, "step": 13760 }, { "grad_norm": 0.6309137940406799, "learning_rate": 9.597836117008462e-05, "loss": 0.0581, "step": 13770 }, { "grad_norm": 0.4045943021774292, "learning_rate": 9.597023596220356e-05, "loss": 0.0664, "step": 13780 }, { "grad_norm": 0.5805303454399109, "learning_rate": 9.596210289926643e-05, "loss": 0.0646, "step": 13790 }, { "grad_norm": 0.5189526677131653, "learning_rate": 9.5953961982663e-05, "loss": 0.0723, "step": 13800 }, { "grad_norm": 0.5116340517997742, "learning_rate": 9.594581321378431e-05, "loss": 0.0761, "step": 13810 }, { "grad_norm": 0.5062205195426941, "learning_rate": 9.593765659402276e-05, "loss": 0.0659, "step": 13820 }, { "grad_norm": 0.5358405113220215, "learning_rate": 9.59294921247721e-05, "loss": 0.066, "step": 13830 }, { "grad_norm": 0.565453290939331, "learning_rate": 9.59213198074274e-05, "loss": 0.0762, "step": 13840 }, { "grad_norm": 0.6423807144165039, "learning_rate": 9.59131396433851e-05, "loss": 0.0707, "step": 13850 }, { "grad_norm": 0.40845996141433716, "learning_rate": 9.590495163404297e-05, "loss": 0.0688, "step": 13860 }, { "grad_norm": 0.45521050691604614, "learning_rate": 9.589675578080009e-05, "loss": 0.0733, "step": 13870 }, { "grad_norm": 0.4702887237071991, "learning_rate": 9.588855208505694e-05, "loss": 0.0608, "step": 13880 }, { "grad_norm": 0.7235849499702454, "learning_rate": 9.588034054821529e-05, "loss": 0.0765, "step": 13890 }, { "grad_norm": 0.5205695033073425, "learning_rate": 9.587212117167826e-05, "loss": 0.0616, "step": 13900 }, { "grad_norm": 0.6126659512519836, "learning_rate": 9.586389395685033e-05, "loss": 0.0733, "step": 13910 }, { "grad_norm": 0.6445385217666626, "learning_rate": 9.585565890513733e-05, "loss": 0.0745, "step": 13920 }, { "grad_norm": 0.47611165046691895, "learning_rate": 9.584741601794636e-05, "loss": 0.0632, "step": 13930 }, { "grad_norm": 0.4806634783744812, "learning_rate": 9.58391652966859e-05, "loss": 0.0622, "step": 13940 }, { "grad_norm": 0.5038318037986755, "learning_rate": 9.583090674276583e-05, "loss": 0.0683, "step": 13950 }, { "grad_norm": 0.4422120153903961, "learning_rate": 9.582264035759726e-05, "loss": 0.0667, "step": 13960 }, { "grad_norm": 0.6026042103767395, "learning_rate": 9.58143661425927e-05, "loss": 0.094, "step": 13970 }, { "grad_norm": 0.44170942902565, "learning_rate": 9.580608409916601e-05, "loss": 0.0802, "step": 13980 }, { "grad_norm": 0.5370982885360718, "learning_rate": 9.579779422873233e-05, "loss": 0.0691, "step": 13990 }, { "grad_norm": 0.44537118077278137, "learning_rate": 9.578949653270819e-05, "loss": 0.0685, "step": 14000 }, { "grad_norm": 0.4438260495662689, "learning_rate": 9.578119101251144e-05, "loss": 0.0608, "step": 14010 }, { "grad_norm": 0.6541661620140076, "learning_rate": 9.577287766956127e-05, "loss": 0.0741, "step": 14020 }, { "grad_norm": 0.4782533347606659, "learning_rate": 9.57645565052782e-05, "loss": 0.0798, "step": 14030 }, { "grad_norm": 0.5377525687217712, "learning_rate": 9.575622752108407e-05, "loss": 0.0668, "step": 14040 }, { "grad_norm": 0.42665010690689087, "learning_rate": 9.57478907184021e-05, "loss": 0.0506, "step": 14050 }, { "grad_norm": 0.39507919549942017, "learning_rate": 9.573954609865681e-05, "loss": 0.0687, "step": 14060 }, { "grad_norm": 0.4874297082424164, "learning_rate": 9.573119366327408e-05, "loss": 0.0623, "step": 14070 }, { "grad_norm": 0.41108208894729614, "learning_rate": 9.57228334136811e-05, "loss": 0.0644, "step": 14080 }, { "grad_norm": 0.35904911160469055, "learning_rate": 9.571446535130641e-05, "loss": 0.0519, "step": 14090 }, { "grad_norm": 0.4224199950695038, "learning_rate": 9.570608947757988e-05, "loss": 0.0578, "step": 14100 }, { "grad_norm": 0.46534210443496704, "learning_rate": 9.569770579393274e-05, "loss": 0.0712, "step": 14110 }, { "grad_norm": 0.3682231605052948, "learning_rate": 9.56893143017975e-05, "loss": 0.0588, "step": 14120 }, { "grad_norm": 0.5606133341789246, "learning_rate": 9.568091500260806e-05, "loss": 0.0692, "step": 14130 }, { "grad_norm": 0.6674370169639587, "learning_rate": 9.567250789779961e-05, "loss": 0.0684, "step": 14140 }, { "grad_norm": 0.4672556221485138, "learning_rate": 9.566409298880872e-05, "loss": 0.0623, "step": 14150 }, { "grad_norm": 0.5888873934745789, "learning_rate": 9.565567027707326e-05, "loss": 0.0864, "step": 14160 }, { "grad_norm": 0.5354304909706116, "learning_rate": 9.56472397640324e-05, "loss": 0.064, "step": 14170 }, { "grad_norm": 0.39335897564888, "learning_rate": 9.563880145112675e-05, "loss": 0.0585, "step": 14180 }, { "grad_norm": 0.6054362058639526, "learning_rate": 9.563035533979814e-05, "loss": 0.0608, "step": 14190 }, { "grad_norm": 0.5535092949867249, "learning_rate": 9.562190143148981e-05, "loss": 0.0685, "step": 14200 }, { "grad_norm": 0.37813809514045715, "learning_rate": 9.561343972764627e-05, "loss": 0.0696, "step": 14210 }, { "grad_norm": 0.5419427752494812, "learning_rate": 9.560497022971343e-05, "loss": 0.0725, "step": 14220 }, { "grad_norm": 0.4918838441371918, "learning_rate": 9.559649293913847e-05, "loss": 0.0674, "step": 14230 }, { "grad_norm": 0.6257385611534119, "learning_rate": 9.558800785736993e-05, "loss": 0.0772, "step": 14240 }, { "grad_norm": 0.4842541813850403, "learning_rate": 9.557951498585767e-05, "loss": 0.0801, "step": 14250 }, { "grad_norm": 0.580263078212738, "learning_rate": 9.557101432605293e-05, "loss": 0.0664, "step": 14260 }, { "grad_norm": 0.5104680061340332, "learning_rate": 9.556250587940818e-05, "loss": 0.081, "step": 14270 }, { "grad_norm": 0.47315022349357605, "learning_rate": 9.555398964737734e-05, "loss": 0.0669, "step": 14280 }, { "grad_norm": 0.5264695286750793, "learning_rate": 9.554546563141555e-05, "loss": 0.0665, "step": 14290 }, { "grad_norm": 0.394467830657959, "learning_rate": 9.553693383297937e-05, "loss": 0.069, "step": 14300 }, { "grad_norm": 0.4521915018558502, "learning_rate": 9.552839425352663e-05, "loss": 0.0614, "step": 14310 }, { "grad_norm": 1.0034303665161133, "learning_rate": 9.551984689451652e-05, "loss": 0.0798, "step": 14320 }, { "grad_norm": 0.6075925230979919, "learning_rate": 9.551129175740953e-05, "loss": 0.0679, "step": 14330 }, { "grad_norm": 0.43180009722709656, "learning_rate": 9.550272884366754e-05, "loss": 0.0678, "step": 14340 }, { "grad_norm": 0.512881338596344, "learning_rate": 9.549415815475369e-05, "loss": 0.0756, "step": 14350 }, { "grad_norm": 0.3264418840408325, "learning_rate": 9.548557969213247e-05, "loss": 0.0607, "step": 14360 }, { "grad_norm": 0.44842466711997986, "learning_rate": 9.547699345726972e-05, "loss": 0.0601, "step": 14370 }, { "grad_norm": 0.43166783452033997, "learning_rate": 9.546839945163257e-05, "loss": 0.0594, "step": 14380 }, { "grad_norm": 0.48424017429351807, "learning_rate": 9.545979767668953e-05, "loss": 0.0802, "step": 14390 }, { "grad_norm": 0.5272208452224731, "learning_rate": 9.54511881339104e-05, "loss": 0.0644, "step": 14400 }, { "grad_norm": 0.4660505950450897, "learning_rate": 9.54425708247663e-05, "loss": 0.0536, "step": 14410 }, { "grad_norm": 0.4661538004875183, "learning_rate": 9.543394575072972e-05, "loss": 0.059, "step": 14420 }, { "grad_norm": 0.5339792966842651, "learning_rate": 9.542531291327441e-05, "loss": 0.065, "step": 14430 }, { "grad_norm": 0.5154083967208862, "learning_rate": 9.541667231387552e-05, "loss": 0.0572, "step": 14440 }, { "grad_norm": 0.49916815757751465, "learning_rate": 9.540802395400949e-05, "loss": 0.0588, "step": 14450 }, { "grad_norm": 0.41238364577293396, "learning_rate": 9.539936783515406e-05, "loss": 0.0681, "step": 14460 }, { "grad_norm": 0.4791352450847626, "learning_rate": 9.539070395878835e-05, "loss": 0.0684, "step": 14470 }, { "grad_norm": 0.49204710125923157, "learning_rate": 9.538203232639277e-05, "loss": 0.0595, "step": 14480 }, { "grad_norm": 0.59333735704422, "learning_rate": 9.537335293944907e-05, "loss": 0.0561, "step": 14490 }, { "grad_norm": 0.47764351963996887, "learning_rate": 9.536466579944032e-05, "loss": 0.0748, "step": 14500 }, { "grad_norm": 0.4741407036781311, "learning_rate": 9.535597090785091e-05, "loss": 0.0661, "step": 14510 }, { "grad_norm": 0.4551875591278076, "learning_rate": 9.534726826616656e-05, "loss": 0.0615, "step": 14520 }, { "grad_norm": 0.6546453833580017, "learning_rate": 9.53385578758743e-05, "loss": 0.084, "step": 14530 }, { "grad_norm": 0.4893662929534912, "learning_rate": 9.532983973846252e-05, "loss": 0.0541, "step": 14540 }, { "grad_norm": 0.5275176167488098, "learning_rate": 9.53211138554209e-05, "loss": 0.0606, "step": 14550 }, { "grad_norm": 0.4403321444988251, "learning_rate": 9.531238022824047e-05, "loss": 0.0597, "step": 14560 }, { "grad_norm": 0.38642629981040955, "learning_rate": 9.530363885841355e-05, "loss": 0.0545, "step": 14570 }, { "grad_norm": 0.5134483575820923, "learning_rate": 9.52948897474338e-05, "loss": 0.0671, "step": 14580 }, { "grad_norm": 0.48482465744018555, "learning_rate": 9.528613289679622e-05, "loss": 0.0756, "step": 14590 }, { "grad_norm": 0.52732914686203, "learning_rate": 9.52773683079971e-05, "loss": 0.0772, "step": 14600 }, { "grad_norm": 0.417415976524353, "learning_rate": 9.526859598253407e-05, "loss": 0.064, "step": 14610 }, { "grad_norm": 0.5982999205589294, "learning_rate": 9.525981592190609e-05, "loss": 0.0609, "step": 14620 }, { "grad_norm": 0.4215909242630005, "learning_rate": 9.525102812761342e-05, "loss": 0.0584, "step": 14630 }, { "grad_norm": 0.4861172139644623, "learning_rate": 9.524223260115768e-05, "loss": 0.0508, "step": 14640 }, { "grad_norm": 0.44723713397979736, "learning_rate": 9.523342934404175e-05, "loss": 0.0595, "step": 14650 }, { "grad_norm": 0.4293062090873718, "learning_rate": 9.522461835776989e-05, "loss": 0.0648, "step": 14660 }, { "grad_norm": 0.5875836610794067, "learning_rate": 9.521579964384764e-05, "loss": 0.0784, "step": 14670 }, { "grad_norm": 0.43667522072792053, "learning_rate": 9.52069732037819e-05, "loss": 0.063, "step": 14680 }, { "grad_norm": 0.49811217188835144, "learning_rate": 9.519813903908083e-05, "loss": 0.0653, "step": 14690 }, { "grad_norm": 0.41576987504959106, "learning_rate": 9.5189297151254e-05, "loss": 0.0706, "step": 14700 }, { "grad_norm": 0.6417135000228882, "learning_rate": 9.518044754181218e-05, "loss": 0.076, "step": 14710 }, { "grad_norm": 0.525978147983551, "learning_rate": 9.51715902122676e-05, "loss": 0.0786, "step": 14720 }, { "grad_norm": 0.37803342938423157, "learning_rate": 9.516272516413368e-05, "loss": 0.0747, "step": 14730 }, { "grad_norm": 0.3955278694629669, "learning_rate": 9.515385239892525e-05, "loss": 0.0712, "step": 14740 }, { "grad_norm": 0.5650674104690552, "learning_rate": 9.514497191815839e-05, "loss": 0.0686, "step": 14750 }, { "grad_norm": 0.594434916973114, "learning_rate": 9.513608372335055e-05, "loss": 0.0814, "step": 14760 }, { "grad_norm": 0.5388015508651733, "learning_rate": 9.512718781602045e-05, "loss": 0.0594, "step": 14770 }, { "grad_norm": 0.4555971920490265, "learning_rate": 9.511828419768823e-05, "loss": 0.06, "step": 14780 }, { "grad_norm": 0.4923414885997772, "learning_rate": 9.510937286987521e-05, "loss": 0.0527, "step": 14790 }, { "grad_norm": 0.5818886756896973, "learning_rate": 9.510045383410408e-05, "loss": 0.0624, "step": 14800 }, { "grad_norm": 0.6688839197158813, "learning_rate": 9.509152709189892e-05, "loss": 0.0625, "step": 14810 }, { "grad_norm": 0.6270629167556763, "learning_rate": 9.508259264478504e-05, "loss": 0.0626, "step": 14820 }, { "grad_norm": 1.0270594358444214, "learning_rate": 9.507365049428909e-05, "loss": 0.0717, "step": 14830 }, { "grad_norm": 0.31806257367134094, "learning_rate": 9.506470064193902e-05, "loss": 0.0755, "step": 14840 }, { "grad_norm": 0.4596449136734009, "learning_rate": 9.505574308926414e-05, "loss": 0.0831, "step": 14850 }, { "grad_norm": 0.43670493364334106, "learning_rate": 9.504677783779505e-05, "loss": 0.0558, "step": 14860 }, { "grad_norm": 0.38356998562812805, "learning_rate": 9.503780488906365e-05, "loss": 0.0583, "step": 14870 }, { "grad_norm": 0.5455126166343689, "learning_rate": 9.502882424460319e-05, "loss": 0.0664, "step": 14880 }, { "grad_norm": 0.5164798498153687, "learning_rate": 9.501983590594821e-05, "loss": 0.0583, "step": 14890 }, { "grad_norm": 0.46625906229019165, "learning_rate": 9.501083987463455e-05, "loss": 0.0616, "step": 14900 }, { "grad_norm": 0.7084847688674927, "learning_rate": 9.500183615219942e-05, "loss": 0.0637, "step": 14910 }, { "grad_norm": 0.5069482326507568, "learning_rate": 9.49928247401813e-05, "loss": 0.0537, "step": 14920 }, { "grad_norm": 0.5569230914115906, "learning_rate": 9.498380564011997e-05, "loss": 0.0543, "step": 14930 }, { "grad_norm": 0.5198166370391846, "learning_rate": 9.497477885355656e-05, "loss": 0.0636, "step": 14940 }, { "grad_norm": 0.5245279669761658, "learning_rate": 9.496574438203353e-05, "loss": 0.0673, "step": 14950 }, { "grad_norm": 0.5803631544113159, "learning_rate": 9.495670222709459e-05, "loss": 0.0651, "step": 14960 }, { "grad_norm": 0.5716350078582764, "learning_rate": 9.494765239028483e-05, "loss": 0.0598, "step": 14970 }, { "grad_norm": 0.4061238467693329, "learning_rate": 9.493859487315057e-05, "loss": 0.0533, "step": 14980 }, { "grad_norm": 0.5736804604530334, "learning_rate": 9.492952967723953e-05, "loss": 0.0626, "step": 14990 }, { "grad_norm": 0.4662461578845978, "learning_rate": 9.492045680410068e-05, "loss": 0.0624, "step": 15000 }, { "grad_norm": 0.5218215584754944, "learning_rate": 9.491137625528436e-05, "loss": 0.0534, "step": 15010 }, { "grad_norm": 0.41115087270736694, "learning_rate": 9.490228803234215e-05, "loss": 0.0581, "step": 15020 }, { "grad_norm": 0.5672315359115601, "learning_rate": 9.489319213682701e-05, "loss": 0.0628, "step": 15030 }, { "grad_norm": 0.6997116804122925, "learning_rate": 9.488408857029316e-05, "loss": 0.0725, "step": 15040 }, { "grad_norm": 0.6143594980239868, "learning_rate": 9.487497733429616e-05, "loss": 0.0809, "step": 15050 }, { "grad_norm": 0.4845800995826721, "learning_rate": 9.486585843039286e-05, "loss": 0.0713, "step": 15060 }, { "grad_norm": 0.4698023200035095, "learning_rate": 9.485673186014143e-05, "loss": 0.0669, "step": 15070 }, { "grad_norm": 0.3211992383003235, "learning_rate": 9.484759762510137e-05, "loss": 0.0656, "step": 15080 }, { "grad_norm": 0.5876263976097107, "learning_rate": 9.483845572683346e-05, "loss": 0.0724, "step": 15090 }, { "grad_norm": 0.5290430188179016, "learning_rate": 9.48293061668998e-05, "loss": 0.0641, "step": 15100 }, { "grad_norm": 0.43529877066612244, "learning_rate": 9.48201489468638e-05, "loss": 0.0584, "step": 15110 }, { "grad_norm": 0.5213908553123474, "learning_rate": 9.481098406829016e-05, "loss": 0.0592, "step": 15120 }, { "grad_norm": 0.4241611957550049, "learning_rate": 9.480181153274495e-05, "loss": 0.051, "step": 15130 }, { "grad_norm": 0.35669639706611633, "learning_rate": 9.479263134179548e-05, "loss": 0.0491, "step": 15140 }, { "grad_norm": 0.5116244554519653, "learning_rate": 9.478344349701039e-05, "loss": 0.067, "step": 15150 }, { "grad_norm": 0.49487197399139404, "learning_rate": 9.477424799995964e-05, "loss": 0.0659, "step": 15160 }, { "grad_norm": 0.4552171230316162, "learning_rate": 9.476504485221448e-05, "loss": 0.0684, "step": 15170 }, { "grad_norm": 0.5419496893882751, "learning_rate": 9.475583405534748e-05, "loss": 0.0612, "step": 15180 }, { "grad_norm": 0.4193294942378998, "learning_rate": 9.474661561093251e-05, "loss": 0.0594, "step": 15190 }, { "grad_norm": 0.6856876611709595, "learning_rate": 9.473738952054478e-05, "loss": 0.0763, "step": 15200 }, { "grad_norm": 0.41973450779914856, "learning_rate": 9.472815578576073e-05, "loss": 0.0654, "step": 15210 }, { "grad_norm": 0.7359458804130554, "learning_rate": 9.471891440815817e-05, "loss": 0.0723, "step": 15220 }, { "grad_norm": 0.41301658749580383, "learning_rate": 9.470966538931621e-05, "loss": 0.0581, "step": 15230 }, { "grad_norm": 0.5399612188339233, "learning_rate": 9.470040873081525e-05, "loss": 0.0583, "step": 15240 }, { "grad_norm": 0.5048277974128723, "learning_rate": 9.469114443423698e-05, "loss": 0.0772, "step": 15250 }, { "grad_norm": 0.4642925560474396, "learning_rate": 9.468187250116445e-05, "loss": 0.0752, "step": 15260 }, { "grad_norm": 0.3829358220100403, "learning_rate": 9.467259293318197e-05, "loss": 0.0586, "step": 15270 }, { "grad_norm": 0.4113791584968567, "learning_rate": 9.466330573187514e-05, "loss": 0.0562, "step": 15280 }, { "grad_norm": 0.5312934517860413, "learning_rate": 9.46540108988309e-05, "loss": 0.0766, "step": 15290 }, { "grad_norm": 0.458279550075531, "learning_rate": 9.46447084356375e-05, "loss": 0.0686, "step": 15300 }, { "grad_norm": 0.3663201630115509, "learning_rate": 9.463539834388447e-05, "loss": 0.0741, "step": 15310 }, { "grad_norm": 0.49528971314430237, "learning_rate": 9.462608062516263e-05, "loss": 0.0525, "step": 15320 }, { "grad_norm": 0.4656168818473816, "learning_rate": 9.461675528106413e-05, "loss": 0.0529, "step": 15330 }, { "grad_norm": 0.5034117102622986, "learning_rate": 9.460742231318244e-05, "loss": 0.0588, "step": 15340 }, { "grad_norm": 0.4604896903038025, "learning_rate": 9.459808172311229e-05, "loss": 0.0758, "step": 15350 }, { "grad_norm": 0.3366710841655731, "learning_rate": 9.458873351244972e-05, "loss": 0.0575, "step": 15360 }, { "grad_norm": 0.4556906521320343, "learning_rate": 9.457937768279211e-05, "loss": 0.0603, "step": 15370 }, { "grad_norm": 0.3240537941455841, "learning_rate": 9.45700142357381e-05, "loss": 0.0619, "step": 15380 }, { "grad_norm": 0.5949077606201172, "learning_rate": 9.456064317288765e-05, "loss": 0.0638, "step": 15390 }, { "grad_norm": 0.5712973475456238, "learning_rate": 9.455126449584201e-05, "loss": 0.0652, "step": 15400 }, { "grad_norm": 0.3905577063560486, "learning_rate": 9.454187820620375e-05, "loss": 0.0575, "step": 15410 }, { "grad_norm": 0.7605951428413391, "learning_rate": 9.453248430557673e-05, "loss": 0.0806, "step": 15420 }, { "grad_norm": 0.5267527103424072, "learning_rate": 9.452308279556611e-05, "loss": 0.0852, "step": 15430 }, { "grad_norm": 0.39461749792099, "learning_rate": 9.451367367777835e-05, "loss": 0.067, "step": 15440 }, { "grad_norm": 0.45284730195999146, "learning_rate": 9.450425695382122e-05, "loss": 0.0661, "step": 15450 }, { "grad_norm": 0.5368874669075012, "learning_rate": 9.449483262530375e-05, "loss": 0.0568, "step": 15460 }, { "grad_norm": 0.4793208837509155, "learning_rate": 9.448540069383633e-05, "loss": 0.0589, "step": 15470 }, { "grad_norm": 0.6025188565254211, "learning_rate": 9.447596116103061e-05, "loss": 0.0656, "step": 15480 }, { "grad_norm": 0.5593621730804443, "learning_rate": 9.446651402849955e-05, "loss": 0.0712, "step": 15490 }, { "grad_norm": 0.3485308289527893, "learning_rate": 9.44570592978574e-05, "loss": 0.0546, "step": 15500 }, { "grad_norm": 0.43130484223365784, "learning_rate": 9.444759697071972e-05, "loss": 0.0714, "step": 15510 }, { "grad_norm": 0.6281241774559021, "learning_rate": 9.443812704870336e-05, "loss": 0.0548, "step": 15520 }, { "grad_norm": 0.6120501756668091, "learning_rate": 9.442864953342649e-05, "loss": 0.0606, "step": 15530 }, { "grad_norm": 0.45431116223335266, "learning_rate": 9.441916442650852e-05, "loss": 0.0673, "step": 15540 }, { "grad_norm": 0.5413693785667419, "learning_rate": 9.440967172957023e-05, "loss": 0.0655, "step": 15550 }, { "grad_norm": 0.5589629411697388, "learning_rate": 9.440017144423364e-05, "loss": 0.0668, "step": 15560 }, { "grad_norm": 0.5368690490722656, "learning_rate": 9.439066357212209e-05, "loss": 0.0662, "step": 15570 }, { "grad_norm": 0.5925214290618896, "learning_rate": 9.438114811486022e-05, "loss": 0.0746, "step": 15580 }, { "grad_norm": 0.44256147742271423, "learning_rate": 9.4371625074074e-05, "loss": 0.0606, "step": 15590 }, { "grad_norm": 0.42098158597946167, "learning_rate": 9.436209445139059e-05, "loss": 0.0576, "step": 15600 }, { "grad_norm": 0.48540741205215454, "learning_rate": 9.435255624843855e-05, "loss": 0.0688, "step": 15610 }, { "grad_norm": 0.3999684453010559, "learning_rate": 9.43430104668477e-05, "loss": 0.0561, "step": 15620 }, { "grad_norm": 0.3753597140312195, "learning_rate": 9.433345710824914e-05, "loss": 0.0609, "step": 15630 }, { "grad_norm": 0.6851566433906555, "learning_rate": 9.432389617427529e-05, "loss": 0.0649, "step": 15640 }, { "grad_norm": 0.404617041349411, "learning_rate": 9.431432766655984e-05, "loss": 0.0651, "step": 15650 }, { "grad_norm": 0.5926105976104736, "learning_rate": 9.430475158673778e-05, "loss": 0.0694, "step": 15660 }, { "grad_norm": 0.37153372168540955, "learning_rate": 9.429516793644542e-05, "loss": 0.0617, "step": 15670 }, { "grad_norm": 0.37949204444885254, "learning_rate": 9.428557671732034e-05, "loss": 0.0547, "step": 15680 }, { "grad_norm": 0.6770142316818237, "learning_rate": 9.42759779310014e-05, "loss": 0.0682, "step": 15690 }, { "grad_norm": 0.46439191699028015, "learning_rate": 9.426637157912879e-05, "loss": 0.0665, "step": 15700 }, { "grad_norm": 0.46659693121910095, "learning_rate": 9.425675766334397e-05, "loss": 0.0737, "step": 15710 }, { "grad_norm": 0.7174291610717773, "learning_rate": 9.424713618528968e-05, "loss": 0.0641, "step": 15720 }, { "grad_norm": 0.520132839679718, "learning_rate": 9.423750714661e-05, "loss": 0.067, "step": 15730 }, { "grad_norm": 0.6006222367286682, "learning_rate": 9.422787054895022e-05, "loss": 0.064, "step": 15740 }, { "grad_norm": 0.45378631353378296, "learning_rate": 9.4218226393957e-05, "loss": 0.0608, "step": 15750 }, { "grad_norm": 0.5017300844192505, "learning_rate": 9.420857468327828e-05, "loss": 0.0579, "step": 15760 }, { "grad_norm": 0.49484992027282715, "learning_rate": 9.419891541856323e-05, "loss": 0.0551, "step": 15770 }, { "grad_norm": 0.5307866334915161, "learning_rate": 9.41892486014624e-05, "loss": 0.0547, "step": 15780 }, { "grad_norm": 0.5169985890388489, "learning_rate": 9.417957423362756e-05, "loss": 0.0725, "step": 15790 }, { "grad_norm": 0.42131751775741577, "learning_rate": 9.416989231671178e-05, "loss": 0.068, "step": 15800 }, { "grad_norm": 0.4874328076839447, "learning_rate": 9.416020285236946e-05, "loss": 0.0576, "step": 15810 }, { "grad_norm": 0.6196873784065247, "learning_rate": 9.415050584225626e-05, "loss": 0.0714, "step": 15820 }, { "grad_norm": 0.46910202503204346, "learning_rate": 9.414080128802914e-05, "loss": 0.0613, "step": 15830 }, { "grad_norm": 0.2671074867248535, "learning_rate": 9.413108919134632e-05, "loss": 0.0481, "step": 15840 }, { "grad_norm": 0.4911564588546753, "learning_rate": 9.412136955386734e-05, "loss": 0.0587, "step": 15850 }, { "grad_norm": 0.2844705581665039, "learning_rate": 9.411164237725303e-05, "loss": 0.0477, "step": 15860 }, { "grad_norm": 0.5227508544921875, "learning_rate": 9.41019076631655e-05, "loss": 0.0585, "step": 15870 }, { "grad_norm": 0.6734334826469421, "learning_rate": 9.409216541326815e-05, "loss": 0.0645, "step": 15880 }, { "grad_norm": 0.49700090289115906, "learning_rate": 9.408241562922564e-05, "loss": 0.0685, "step": 15890 }, { "grad_norm": 0.3582787811756134, "learning_rate": 9.407265831270395e-05, "loss": 0.0515, "step": 15900 }, { "grad_norm": 0.4953081011772156, "learning_rate": 9.406289346537035e-05, "loss": 0.0612, "step": 15910 }, { "grad_norm": 0.42902714014053345, "learning_rate": 9.405312108889339e-05, "loss": 0.0578, "step": 15920 }, { "grad_norm": 0.42889338731765747, "learning_rate": 9.404334118494288e-05, "loss": 0.0645, "step": 15930 }, { "grad_norm": 0.5315364003181458, "learning_rate": 9.403355375518995e-05, "loss": 0.0602, "step": 15940 }, { "grad_norm": 0.44474688172340393, "learning_rate": 9.4023758801307e-05, "loss": 0.059, "step": 15950 }, { "grad_norm": 0.43232646584510803, "learning_rate": 9.401395632496774e-05, "loss": 0.0525, "step": 15960 }, { "grad_norm": 0.5880248546600342, "learning_rate": 9.400414632784711e-05, "loss": 0.0697, "step": 15970 }, { "grad_norm": 0.46569231152534485, "learning_rate": 9.39943288116214e-05, "loss": 0.0632, "step": 15980 }, { "grad_norm": 0.6337314248085022, "learning_rate": 9.398450377796815e-05, "loss": 0.0715, "step": 15990 }, { "grad_norm": 0.6312068104743958, "learning_rate": 9.397467122856616e-05, "loss": 0.0626, "step": 16000 }, { "grad_norm": 0.6087044477462769, "learning_rate": 9.396483116509558e-05, "loss": 0.0674, "step": 16010 }, { "grad_norm": 0.4938123822212219, "learning_rate": 9.39549835892378e-05, "loss": 0.0573, "step": 16020 }, { "grad_norm": 0.46897849440574646, "learning_rate": 9.39451285026755e-05, "loss": 0.0616, "step": 16030 }, { "grad_norm": 0.4977142810821533, "learning_rate": 9.393526590709262e-05, "loss": 0.0596, "step": 16040 }, { "grad_norm": 0.5178890228271484, "learning_rate": 9.392539580417444e-05, "loss": 0.0622, "step": 16050 }, { "grad_norm": 0.6183578968048096, "learning_rate": 9.391551819560747e-05, "loss": 0.071, "step": 16060 }, { "grad_norm": 0.5286290645599365, "learning_rate": 9.390563308307955e-05, "loss": 0.0675, "step": 16070 }, { "grad_norm": 0.45280539989471436, "learning_rate": 9.389574046827974e-05, "loss": 0.0611, "step": 16080 }, { "grad_norm": 0.6782621145248413, "learning_rate": 9.388584035289845e-05, "loss": 0.0672, "step": 16090 }, { "grad_norm": 0.5177700519561768, "learning_rate": 9.387593273862732e-05, "loss": 0.0707, "step": 16100 }, { "grad_norm": 0.4808092713356018, "learning_rate": 9.386601762715929e-05, "loss": 0.0722, "step": 16110 }, { "grad_norm": 0.48958736658096313, "learning_rate": 9.38560950201886e-05, "loss": 0.0697, "step": 16120 }, { "grad_norm": 0.46728554368019104, "learning_rate": 9.384616491941071e-05, "loss": 0.0579, "step": 16130 }, { "grad_norm": 0.47581276297569275, "learning_rate": 9.383622732652245e-05, "loss": 0.0525, "step": 16140 }, { "grad_norm": 0.4615485668182373, "learning_rate": 9.382628224322187e-05, "loss": 0.0662, "step": 16150 }, { "grad_norm": 0.7231873869895935, "learning_rate": 9.381632967120829e-05, "loss": 0.0728, "step": 16160 }, { "grad_norm": 0.6013010144233704, "learning_rate": 9.380636961218235e-05, "loss": 0.0718, "step": 16170 }, { "grad_norm": 0.34377992153167725, "learning_rate": 9.379640206784597e-05, "loss": 0.0544, "step": 16180 }, { "grad_norm": 0.48000597953796387, "learning_rate": 9.378642703990229e-05, "loss": 0.0644, "step": 16190 }, { "grad_norm": 0.749720573425293, "learning_rate": 9.37764445300558e-05, "loss": 0.0609, "step": 16200 }, { "grad_norm": 0.43950504064559937, "learning_rate": 9.376645454001222e-05, "loss": 0.067, "step": 16210 }, { "grad_norm": 0.360253244638443, "learning_rate": 9.375645707147858e-05, "loss": 0.0623, "step": 16220 }, { "grad_norm": 0.43137967586517334, "learning_rate": 9.374645212616316e-05, "loss": 0.0694, "step": 16230 }, { "grad_norm": 0.5111204385757446, "learning_rate": 9.373643970577555e-05, "loss": 0.0626, "step": 16240 }, { "grad_norm": 0.7051317691802979, "learning_rate": 9.372641981202659e-05, "loss": 0.0638, "step": 16250 }, { "grad_norm": 0.49817606806755066, "learning_rate": 9.37163924466284e-05, "loss": 0.0649, "step": 16260 }, { "grad_norm": 0.5439615845680237, "learning_rate": 9.370635761129438e-05, "loss": 0.0589, "step": 16270 }, { "grad_norm": 0.37601202726364136, "learning_rate": 9.36963153077392e-05, "loss": 0.0628, "step": 16280 }, { "grad_norm": 0.3209352195262909, "learning_rate": 9.368626553767888e-05, "loss": 0.0586, "step": 16290 }, { "grad_norm": 0.4159289598464966, "learning_rate": 9.367620830283057e-05, "loss": 0.044, "step": 16300 }, { "grad_norm": 0.5912683010101318, "learning_rate": 9.366614360491281e-05, "loss": 0.0702, "step": 16310 }, { "grad_norm": 0.3845233917236328, "learning_rate": 9.365607144564539e-05, "loss": 0.046, "step": 16320 }, { "grad_norm": 0.4448559284210205, "learning_rate": 9.364599182674934e-05, "loss": 0.07, "step": 16330 }, { "grad_norm": 0.3673706650733948, "learning_rate": 9.3635904749947e-05, "loss": 0.0561, "step": 16340 }, { "grad_norm": 0.42699679732322693, "learning_rate": 9.362581021696202e-05, "loss": 0.0544, "step": 16350 }, { "grad_norm": 0.42730188369750977, "learning_rate": 9.361570822951921e-05, "loss": 0.0645, "step": 16360 }, { "grad_norm": 0.5626136064529419, "learning_rate": 9.360559878934476e-05, "loss": 0.0648, "step": 16370 }, { "grad_norm": 0.5122339129447937, "learning_rate": 9.359548189816611e-05, "loss": 0.064, "step": 16380 }, { "grad_norm": 0.6057507991790771, "learning_rate": 9.358535755771193e-05, "loss": 0.0722, "step": 16390 }, { "grad_norm": 0.390821248292923, "learning_rate": 9.357522576971221e-05, "loss": 0.0594, "step": 16400 }, { "grad_norm": 0.41848528385162354, "learning_rate": 9.356508653589819e-05, "loss": 0.0565, "step": 16410 }, { "grad_norm": 0.5745152831077576, "learning_rate": 9.355493985800237e-05, "loss": 0.0675, "step": 16420 }, { "grad_norm": 0.4958549439907074, "learning_rate": 9.354478573775857e-05, "loss": 0.0687, "step": 16430 }, { "grad_norm": 0.42994746565818787, "learning_rate": 9.353462417690186e-05, "loss": 0.0587, "step": 16440 }, { "grad_norm": 0.42146992683410645, "learning_rate": 9.352445517716853e-05, "loss": 0.0625, "step": 16450 }, { "grad_norm": 0.5493649840354919, "learning_rate": 9.351427874029621e-05, "loss": 0.0674, "step": 16460 }, { "grad_norm": 0.6189828515052795, "learning_rate": 9.350409486802379e-05, "loss": 0.0626, "step": 16470 }, { "grad_norm": 0.5094215869903564, "learning_rate": 9.349390356209138e-05, "loss": 0.074, "step": 16480 }, { "grad_norm": 0.6508957147598267, "learning_rate": 9.348370482424042e-05, "loss": 0.0719, "step": 16490 }, { "grad_norm": 0.6292356848716736, "learning_rate": 9.347349865621357e-05, "loss": 0.0595, "step": 16500 }, { "grad_norm": 0.538395881652832, "learning_rate": 9.346328505975481e-05, "loss": 0.0594, "step": 16510 }, { "grad_norm": 0.4199070334434509, "learning_rate": 9.345306403660936e-05, "loss": 0.0652, "step": 16520 }, { "grad_norm": 0.5414696335792542, "learning_rate": 9.344283558852371e-05, "loss": 0.0649, "step": 16530 }, { "grad_norm": 0.5517577528953552, "learning_rate": 9.343259971724563e-05, "loss": 0.0667, "step": 16540 }, { "grad_norm": 0.5415768623352051, "learning_rate": 9.342235642452413e-05, "loss": 0.067, "step": 16550 }, { "grad_norm": 0.5814352631568909, "learning_rate": 9.341210571210954e-05, "loss": 0.0835, "step": 16560 }, { "grad_norm": 0.463368684053421, "learning_rate": 9.340184758175338e-05, "loss": 0.0718, "step": 16570 }, { "grad_norm": 0.4682202637195587, "learning_rate": 9.339158203520854e-05, "loss": 0.0699, "step": 16580 }, { "grad_norm": 0.3427947461605072, "learning_rate": 9.338130907422908e-05, "loss": 0.0498, "step": 16590 }, { "grad_norm": 0.3846469819545746, "learning_rate": 9.337102870057037e-05, "loss": 0.0483, "step": 16600 }, { "grad_norm": 0.40555059909820557, "learning_rate": 9.336074091598907e-05, "loss": 0.0552, "step": 16610 }, { "grad_norm": 0.5126828551292419, "learning_rate": 9.335044572224306e-05, "loss": 0.0504, "step": 16620 }, { "grad_norm": 0.5088542103767395, "learning_rate": 9.334014312109151e-05, "loss": 0.0441, "step": 16630 }, { "grad_norm": 0.5559215545654297, "learning_rate": 9.332983311429486e-05, "loss": 0.066, "step": 16640 }, { "grad_norm": 0.5243348479270935, "learning_rate": 9.33195157036148e-05, "loss": 0.0681, "step": 16650 }, { "grad_norm": 0.35514774918556213, "learning_rate": 9.330919089081432e-05, "loss": 0.0522, "step": 16660 }, { "grad_norm": 0.36179983615875244, "learning_rate": 9.32988586776576e-05, "loss": 0.0733, "step": 16670 }, { "grad_norm": 0.6330007314682007, "learning_rate": 9.328851906591016e-05, "loss": 0.0575, "step": 16680 }, { "grad_norm": 0.47376179695129395, "learning_rate": 9.327817205733875e-05, "loss": 0.0656, "step": 16690 }, { "grad_norm": 0.49473145604133606, "learning_rate": 9.326781765371142e-05, "loss": 0.0629, "step": 16700 }, { "grad_norm": 0.5610061287879944, "learning_rate": 9.325745585679741e-05, "loss": 0.0732, "step": 16710 }, { "grad_norm": 0.4582297205924988, "learning_rate": 9.32470866683673e-05, "loss": 0.0719, "step": 16720 }, { "grad_norm": 0.5755264163017273, "learning_rate": 9.323671009019288e-05, "loss": 0.0721, "step": 16730 }, { "grad_norm": 0.4056766629219055, "learning_rate": 9.322632612404725e-05, "loss": 0.057, "step": 16740 }, { "grad_norm": 0.4525549113750458, "learning_rate": 9.321593477170471e-05, "loss": 0.0779, "step": 16750 }, { "grad_norm": 0.4769316017627716, "learning_rate": 9.320553603494088e-05, "loss": 0.0551, "step": 16760 }, { "grad_norm": 0.39762231707572937, "learning_rate": 9.319512991553261e-05, "loss": 0.057, "step": 16770 }, { "grad_norm": 0.40751564502716064, "learning_rate": 9.318471641525803e-05, "loss": 0.0569, "step": 16780 }, { "grad_norm": 0.46939992904663086, "learning_rate": 9.317429553589652e-05, "loss": 0.0524, "step": 16790 }, { "grad_norm": 0.5150743126869202, "learning_rate": 9.316386727922873e-05, "loss": 0.0552, "step": 16800 }, { "grad_norm": 0.5169395804405212, "learning_rate": 9.315343164703656e-05, "loss": 0.0547, "step": 16810 }, { "grad_norm": 0.5098431706428528, "learning_rate": 9.314298864110316e-05, "loss": 0.0568, "step": 16820 }, { "grad_norm": 0.38218918442726135, "learning_rate": 9.313253826321295e-05, "loss": 0.0561, "step": 16830 }, { "grad_norm": 0.36456069350242615, "learning_rate": 9.312208051515165e-05, "loss": 0.0477, "step": 16840 }, { "grad_norm": 0.4187549352645874, "learning_rate": 9.311161539870618e-05, "loss": 0.0522, "step": 16850 }, { "grad_norm": 0.4855102598667145, "learning_rate": 9.310114291566474e-05, "loss": 0.0674, "step": 16860 }, { "grad_norm": 0.5503140091896057, "learning_rate": 9.309066306781679e-05, "loss": 0.0604, "step": 16870 }, { "grad_norm": 0.4200780689716339, "learning_rate": 9.308017585695306e-05, "loss": 0.0544, "step": 16880 }, { "grad_norm": 0.432858407497406, "learning_rate": 9.306968128486552e-05, "loss": 0.0593, "step": 16890 }, { "grad_norm": 0.37399864196777344, "learning_rate": 9.30591793533474e-05, "loss": 0.0601, "step": 16900 }, { "grad_norm": 0.4556344449520111, "learning_rate": 9.304867006419321e-05, "loss": 0.0533, "step": 16910 }, { "grad_norm": 0.5643684267997742, "learning_rate": 9.303815341919868e-05, "loss": 0.0531, "step": 16920 }, { "grad_norm": 0.5165573954582214, "learning_rate": 9.302762942016084e-05, "loss": 0.0591, "step": 16930 }, { "grad_norm": 0.4911012053489685, "learning_rate": 9.301709806887792e-05, "loss": 0.0592, "step": 16940 }, { "grad_norm": 0.47538042068481445, "learning_rate": 9.300655936714948e-05, "loss": 0.0713, "step": 16950 }, { "grad_norm": 0.45840832591056824, "learning_rate": 9.299601331677627e-05, "loss": 0.0651, "step": 16960 }, { "grad_norm": 0.41692250967025757, "learning_rate": 9.298545991956033e-05, "loss": 0.0677, "step": 16970 }, { "grad_norm": 0.556460440158844, "learning_rate": 9.297489917730493e-05, "loss": 0.0591, "step": 16980 }, { "grad_norm": 0.5763121247291565, "learning_rate": 9.296433109181464e-05, "loss": 0.0606, "step": 16990 }, { "grad_norm": 0.6640076637268066, "learning_rate": 9.295375566489523e-05, "loss": 0.0641, "step": 17000 }, { "grad_norm": 0.5270716547966003, "learning_rate": 9.294317289835379e-05, "loss": 0.0795, "step": 17010 }, { "grad_norm": 0.4919300675392151, "learning_rate": 9.293258279399859e-05, "loss": 0.0766, "step": 17020 }, { "grad_norm": 0.5418232679367065, "learning_rate": 9.292198535363919e-05, "loss": 0.0639, "step": 17030 }, { "grad_norm": 0.4816684424877167, "learning_rate": 9.291138057908641e-05, "loss": 0.0552, "step": 17040 }, { "grad_norm": 0.60521399974823, "learning_rate": 9.290076847215234e-05, "loss": 0.0764, "step": 17050 }, { "grad_norm": 0.7299526929855347, "learning_rate": 9.289014903465025e-05, "loss": 0.0848, "step": 17060 }, { "grad_norm": 0.494576096534729, "learning_rate": 9.287952226839475e-05, "loss": 0.0559, "step": 17070 }, { "grad_norm": 0.409984827041626, "learning_rate": 9.286888817520164e-05, "loss": 0.0594, "step": 17080 }, { "grad_norm": 0.3882753551006317, "learning_rate": 9.285824675688803e-05, "loss": 0.0653, "step": 17090 }, { "grad_norm": 0.5608657002449036, "learning_rate": 9.28475980152722e-05, "loss": 0.066, "step": 17100 }, { "grad_norm": 0.5433562994003296, "learning_rate": 9.283694195217379e-05, "loss": 0.0711, "step": 17110 }, { "grad_norm": 0.5897427797317505, "learning_rate": 9.282627856941356e-05, "loss": 0.0596, "step": 17120 }, { "grad_norm": 0.4703240692615509, "learning_rate": 9.281560786881363e-05, "loss": 0.0545, "step": 17130 }, { "grad_norm": 0.4037010073661804, "learning_rate": 9.280492985219733e-05, "loss": 0.0501, "step": 17140 }, { "grad_norm": 0.46372339129447937, "learning_rate": 9.279424452138924e-05, "loss": 0.0579, "step": 17150 }, { "grad_norm": 0.4455404579639435, "learning_rate": 9.278355187821517e-05, "loss": 0.0621, "step": 17160 }, { "grad_norm": 0.5811958312988281, "learning_rate": 9.277285192450224e-05, "loss": 0.0646, "step": 17170 }, { "grad_norm": 0.4524649977684021, "learning_rate": 9.276214466207875e-05, "loss": 0.0561, "step": 17180 }, { "grad_norm": 0.41788434982299805, "learning_rate": 9.275143009277427e-05, "loss": 0.0466, "step": 17190 }, { "grad_norm": 0.4509649872779846, "learning_rate": 9.274070821841964e-05, "loss": 0.0595, "step": 17200 }, { "grad_norm": 0.4497041404247284, "learning_rate": 9.272997904084696e-05, "loss": 0.0517, "step": 17210 }, { "grad_norm": 0.45994478464126587, "learning_rate": 9.271924256188951e-05, "loss": 0.0498, "step": 17220 }, { "grad_norm": 0.41024476289749146, "learning_rate": 9.270849878338189e-05, "loss": 0.0557, "step": 17230 }, { "grad_norm": 0.48551100492477417, "learning_rate": 9.269774770715991e-05, "loss": 0.0552, "step": 17240 }, { "grad_norm": 0.38058122992515564, "learning_rate": 9.268698933506061e-05, "loss": 0.0695, "step": 17250 }, { "grad_norm": 0.35834425687789917, "learning_rate": 9.267622366892235e-05, "loss": 0.0565, "step": 17260 }, { "grad_norm": 0.5193325281143188, "learning_rate": 9.266545071058465e-05, "loss": 0.056, "step": 17270 }, { "grad_norm": 0.4659961462020874, "learning_rate": 9.265467046188833e-05, "loss": 0.051, "step": 17280 }, { "grad_norm": 0.5762758851051331, "learning_rate": 9.264388292467543e-05, "loss": 0.0522, "step": 17290 }, { "grad_norm": 0.604823887348175, "learning_rate": 9.263308810078926e-05, "loss": 0.0574, "step": 17300 }, { "grad_norm": 0.4425051808357239, "learning_rate": 9.262228599207434e-05, "loss": 0.056, "step": 17310 }, { "grad_norm": 0.40980416536331177, "learning_rate": 9.261147660037647e-05, "loss": 0.042, "step": 17320 }, { "grad_norm": 0.3879441022872925, "learning_rate": 9.26006599275427e-05, "loss": 0.0464, "step": 17330 }, { "grad_norm": 0.3866177201271057, "learning_rate": 9.258983597542124e-05, "loss": 0.0498, "step": 17340 }, { "grad_norm": 0.5585817694664001, "learning_rate": 9.257900474586167e-05, "loss": 0.0582, "step": 17350 }, { "grad_norm": 0.413878858089447, "learning_rate": 9.256816624071471e-05, "loss": 0.0655, "step": 17360 }, { "grad_norm": 0.41135358810424805, "learning_rate": 9.25573204618324e-05, "loss": 0.0569, "step": 17370 }, { "grad_norm": 0.9767993092536926, "learning_rate": 9.254646741106796e-05, "loss": 0.0575, "step": 17380 }, { "grad_norm": 0.4020562171936035, "learning_rate": 9.253560709027589e-05, "loss": 0.0534, "step": 17390 }, { "grad_norm": 0.47467413544654846, "learning_rate": 9.252473950131192e-05, "loss": 0.0682, "step": 17400 }, { "grad_norm": 0.4373821020126343, "learning_rate": 9.251386464603302e-05, "loss": 0.0498, "step": 17410 }, { "grad_norm": 0.3401252031326294, "learning_rate": 9.250298252629741e-05, "loss": 0.063, "step": 17420 }, { "grad_norm": 0.4869624674320221, "learning_rate": 9.249209314396454e-05, "loss": 0.058, "step": 17430 }, { "grad_norm": 0.641813337802887, "learning_rate": 9.248119650089513e-05, "loss": 0.074, "step": 17440 }, { "grad_norm": 0.457035630941391, "learning_rate": 9.247029259895108e-05, "loss": 0.0695, "step": 17450 }, { "grad_norm": 0.5567890405654907, "learning_rate": 9.24593814399956e-05, "loss": 0.0596, "step": 17460 }, { "grad_norm": 0.3830871880054474, "learning_rate": 9.244846302589309e-05, "loss": 0.0484, "step": 17470 }, { "grad_norm": 0.46935075521469116, "learning_rate": 9.243753735850923e-05, "loss": 0.0629, "step": 17480 }, { "grad_norm": 0.5049607157707214, "learning_rate": 9.24266044397109e-05, "loss": 0.0625, "step": 17490 }, { "grad_norm": 0.5083919763565063, "learning_rate": 9.241566427136624e-05, "loss": 0.0601, "step": 17500 }, { "grad_norm": 0.3535782992839813, "learning_rate": 9.240471685534463e-05, "loss": 0.06, "step": 17510 }, { "grad_norm": 0.39250829815864563, "learning_rate": 9.239376219351667e-05, "loss": 0.0602, "step": 17520 }, { "grad_norm": 0.5709932446479797, "learning_rate": 9.238280028775425e-05, "loss": 0.0607, "step": 17530 }, { "grad_norm": 0.47858551144599915, "learning_rate": 9.237183113993041e-05, "loss": 0.0674, "step": 17540 }, { "grad_norm": 0.5209916830062866, "learning_rate": 9.236085475191952e-05, "loss": 0.0627, "step": 17550 }, { "grad_norm": 0.48436787724494934, "learning_rate": 9.234987112559709e-05, "loss": 0.0652, "step": 17560 }, { "grad_norm": 0.4484178423881531, "learning_rate": 9.233888026283999e-05, "loss": 0.0705, "step": 17570 }, { "grad_norm": 0.6206410527229309, "learning_rate": 9.232788216552619e-05, "loss": 0.0572, "step": 17580 }, { "grad_norm": 0.46638694405555725, "learning_rate": 9.231687683553502e-05, "loss": 0.0723, "step": 17590 }, { "grad_norm": 0.7819163799285889, "learning_rate": 9.230586427474698e-05, "loss": 0.0716, "step": 17600 }, { "grad_norm": 0.5633795261383057, "learning_rate": 9.229484448504379e-05, "loss": 0.0795, "step": 17610 }, { "grad_norm": 0.4837574064731598, "learning_rate": 9.228381746830843e-05, "loss": 0.0609, "step": 17620 }, { "grad_norm": 0.5766672492027283, "learning_rate": 9.227278322642514e-05, "loss": 0.0539, "step": 17630 }, { "grad_norm": 0.3845198452472687, "learning_rate": 9.226174176127937e-05, "loss": 0.0667, "step": 17640 }, { "grad_norm": 0.3496425747871399, "learning_rate": 9.22506930747578e-05, "loss": 0.0567, "step": 17650 }, { "grad_norm": 0.6723405718803406, "learning_rate": 9.223963716874831e-05, "loss": 0.0765, "step": 17660 }, { "grad_norm": 0.47124478220939636, "learning_rate": 9.222857404514012e-05, "loss": 0.0542, "step": 17670 }, { "grad_norm": 0.5222494006156921, "learning_rate": 9.221750370582355e-05, "loss": 0.0562, "step": 17680 }, { "grad_norm": 0.4098489582538605, "learning_rate": 9.220642615269028e-05, "loss": 0.0529, "step": 17690 }, { "grad_norm": 0.360467791557312, "learning_rate": 9.219534138763311e-05, "loss": 0.0676, "step": 17700 }, { "grad_norm": 0.504114031791687, "learning_rate": 9.218424941254613e-05, "loss": 0.0703, "step": 17710 }, { "grad_norm": 0.4108007252216339, "learning_rate": 9.217315022932468e-05, "loss": 0.0628, "step": 17720 }, { "grad_norm": 0.5085349678993225, "learning_rate": 9.216204383986528e-05, "loss": 0.051, "step": 17730 }, { "grad_norm": 0.3797714114189148, "learning_rate": 9.215093024606574e-05, "loss": 0.0596, "step": 17740 }, { "grad_norm": 0.7043601870536804, "learning_rate": 9.213980944982506e-05, "loss": 0.0763, "step": 17750 }, { "grad_norm": 0.5368098616600037, "learning_rate": 9.212868145304346e-05, "loss": 0.0625, "step": 17760 }, { "grad_norm": 0.49875015020370483, "learning_rate": 9.211754625762241e-05, "loss": 0.0507, "step": 17770 }, { "grad_norm": 0.5001509785652161, "learning_rate": 9.210640386546463e-05, "loss": 0.0632, "step": 17780 }, { "grad_norm": 0.6397436261177063, "learning_rate": 9.209525427847405e-05, "loss": 0.0434, "step": 17790 }, { "grad_norm": 0.45025739073753357, "learning_rate": 9.208409749855583e-05, "loss": 0.0553, "step": 17800 }, { "grad_norm": 0.4356400966644287, "learning_rate": 9.207293352761633e-05, "loss": 0.0588, "step": 17810 }, { "grad_norm": 0.6639442443847656, "learning_rate": 9.206176236756319e-05, "loss": 0.0557, "step": 17820 }, { "grad_norm": 0.40009158849716187, "learning_rate": 9.205058402030525e-05, "loss": 0.0569, "step": 17830 }, { "grad_norm": 0.5621672868728638, "learning_rate": 9.203939848775259e-05, "loss": 0.0581, "step": 17840 }, { "grad_norm": 0.4105580151081085, "learning_rate": 9.202820577181652e-05, "loss": 0.0568, "step": 17850 }, { "grad_norm": 0.3810359835624695, "learning_rate": 9.201700587440953e-05, "loss": 0.0653, "step": 17860 }, { "grad_norm": 0.4541037976741791, "learning_rate": 9.200579879744544e-05, "loss": 0.0649, "step": 17870 }, { "grad_norm": 0.5392724871635437, "learning_rate": 9.199458454283918e-05, "loss": 0.0559, "step": 17880 }, { "grad_norm": 0.39358797669410706, "learning_rate": 9.198336311250697e-05, "loss": 0.0614, "step": 17890 }, { "grad_norm": 0.6187477707862854, "learning_rate": 9.197213450836626e-05, "loss": 0.0479, "step": 17900 }, { "grad_norm": 0.650922954082489, "learning_rate": 9.19608987323357e-05, "loss": 0.0612, "step": 17910 }, { "grad_norm": 0.45110592246055603, "learning_rate": 9.194965578633517e-05, "loss": 0.0534, "step": 17920 }, { "grad_norm": 0.5484747290611267, "learning_rate": 9.193840567228582e-05, "loss": 0.0604, "step": 17930 }, { "grad_norm": 0.45176759362220764, "learning_rate": 9.192714839210994e-05, "loss": 0.0576, "step": 17940 }, { "grad_norm": 0.36819347739219666, "learning_rate": 9.19158839477311e-05, "loss": 0.0589, "step": 17950 }, { "grad_norm": 0.44863811135292053, "learning_rate": 9.190461234107411e-05, "loss": 0.0664, "step": 17960 }, { "grad_norm": 0.6333407163619995, "learning_rate": 9.189333357406496e-05, "loss": 0.0541, "step": 17970 }, { "grad_norm": 0.46482256054878235, "learning_rate": 9.188204764863089e-05, "loss": 0.0636, "step": 17980 }, { "grad_norm": 0.5548675656318665, "learning_rate": 9.187075456670033e-05, "loss": 0.0673, "step": 17990 }, { "grad_norm": 0.31979501247406006, "learning_rate": 9.1859454330203e-05, "loss": 0.0659, "step": 18000 }, { "grad_norm": 0.3850500285625458, "learning_rate": 9.18481469410698e-05, "loss": 0.0652, "step": 18010 }, { "grad_norm": 0.415698766708374, "learning_rate": 9.183683240123281e-05, "loss": 0.0578, "step": 18020 }, { "grad_norm": 0.514049232006073, "learning_rate": 9.182551071262541e-05, "loss": 0.061, "step": 18030 }, { "grad_norm": 0.3664729595184326, "learning_rate": 9.181418187718218e-05, "loss": 0.0587, "step": 18040 }, { "grad_norm": 0.5516924858093262, "learning_rate": 9.180284589683888e-05, "loss": 0.0579, "step": 18050 }, { "grad_norm": 0.543840765953064, "learning_rate": 9.17915027735325e-05, "loss": 0.0657, "step": 18060 }, { "grad_norm": 0.8186569213867188, "learning_rate": 9.178015250920133e-05, "loss": 0.0594, "step": 18070 }, { "grad_norm": 0.45152565836906433, "learning_rate": 9.176879510578477e-05, "loss": 0.0504, "step": 18080 }, { "grad_norm": 0.48463448882102966, "learning_rate": 9.17574305652235e-05, "loss": 0.0608, "step": 18090 }, { "grad_norm": 0.7593916058540344, "learning_rate": 9.174605888945942e-05, "loss": 0.0662, "step": 18100 }, { "grad_norm": 0.4964461326599121, "learning_rate": 9.173468008043564e-05, "loss": 0.0686, "step": 18110 }, { "grad_norm": 0.4187949299812317, "learning_rate": 9.172329414009648e-05, "loss": 0.0584, "step": 18120 }, { "grad_norm": 0.5254566669464111, "learning_rate": 9.171190107038747e-05, "loss": 0.0703, "step": 18130 }, { "grad_norm": 0.5490005612373352, "learning_rate": 9.170050087325541e-05, "loss": 0.0585, "step": 18140 }, { "grad_norm": 0.49209699034690857, "learning_rate": 9.168909355064824e-05, "loss": 0.0545, "step": 18150 }, { "grad_norm": 0.5436842441558838, "learning_rate": 9.167767910451519e-05, "loss": 0.0538, "step": 18160 }, { "grad_norm": 0.3896417021751404, "learning_rate": 9.166625753680669e-05, "loss": 0.0543, "step": 18170 }, { "grad_norm": 0.41175857186317444, "learning_rate": 9.165482884947431e-05, "loss": 0.0581, "step": 18180 }, { "grad_norm": 0.5015326142311096, "learning_rate": 9.164339304447098e-05, "loss": 0.0647, "step": 18190 }, { "grad_norm": 0.3254127502441406, "learning_rate": 9.163195012375072e-05, "loss": 0.0494, "step": 18200 }, { "grad_norm": 0.515649139881134, "learning_rate": 9.16205000892688e-05, "loss": 0.0625, "step": 18210 }, { "grad_norm": 0.51218581199646, "learning_rate": 9.160904294298175e-05, "loss": 0.0592, "step": 18220 }, { "grad_norm": 0.4087151885032654, "learning_rate": 9.159757868684727e-05, "loss": 0.0585, "step": 18230 }, { "grad_norm": 0.6071979999542236, "learning_rate": 9.15861073228243e-05, "loss": 0.0572, "step": 18240 }, { "grad_norm": 0.5947229266166687, "learning_rate": 9.157462885287296e-05, "loss": 0.0728, "step": 18250 }, { "grad_norm": 0.4349576532840729, "learning_rate": 9.156314327895461e-05, "loss": 0.0532, "step": 18260 }, { "grad_norm": 0.4388175904750824, "learning_rate": 9.155165060303185e-05, "loss": 0.0681, "step": 18270 }, { "grad_norm": 0.5263921618461609, "learning_rate": 9.154015082706841e-05, "loss": 0.0566, "step": 18280 }, { "grad_norm": 0.5493752360343933, "learning_rate": 9.152864395302936e-05, "loss": 0.0544, "step": 18290 }, { "grad_norm": 0.5376730561256409, "learning_rate": 9.151712998288085e-05, "loss": 0.0497, "step": 18300 }, { "grad_norm": 0.43168574571609497, "learning_rate": 9.150560891859031e-05, "loss": 0.0637, "step": 18310 }, { "grad_norm": 0.4410978853702545, "learning_rate": 9.14940807621264e-05, "loss": 0.0576, "step": 18320 }, { "grad_norm": 0.4130796790122986, "learning_rate": 9.148254551545894e-05, "loss": 0.0552, "step": 18330 }, { "grad_norm": 0.6159875988960266, "learning_rate": 9.147100318055901e-05, "loss": 0.0535, "step": 18340 }, { "grad_norm": 0.5330118536949158, "learning_rate": 9.145945375939888e-05, "loss": 0.0538, "step": 18350 }, { "grad_norm": 0.3193998634815216, "learning_rate": 9.144789725395203e-05, "loss": 0.0556, "step": 18360 }, { "grad_norm": 0.4712330400943756, "learning_rate": 9.14363336661931e-05, "loss": 0.0541, "step": 18370 }, { "grad_norm": 0.5161160230636597, "learning_rate": 9.142476299809806e-05, "loss": 0.0492, "step": 18380 }, { "grad_norm": 0.3854685425758362, "learning_rate": 9.1413185251644e-05, "loss": 0.0506, "step": 18390 }, { "grad_norm": 0.45059308409690857, "learning_rate": 9.140160042880923e-05, "loss": 0.0646, "step": 18400 }, { "grad_norm": 0.39451950788497925, "learning_rate": 9.139000853157327e-05, "loss": 0.0592, "step": 18410 }, { "grad_norm": 0.35153940320014954, "learning_rate": 9.137840956191688e-05, "loss": 0.0692, "step": 18420 }, { "grad_norm": 0.6013940572738647, "learning_rate": 9.136680352182199e-05, "loss": 0.0559, "step": 18430 }, { "grad_norm": 0.5761241316795349, "learning_rate": 9.135519041327177e-05, "loss": 0.0565, "step": 18440 }, { "grad_norm": 0.49093422293663025, "learning_rate": 9.134357023825058e-05, "loss": 0.062, "step": 18450 }, { "grad_norm": 0.551712155342102, "learning_rate": 9.133194299874398e-05, "loss": 0.055, "step": 18460 }, { "grad_norm": 0.6354609131813049, "learning_rate": 9.132030869673876e-05, "loss": 0.0706, "step": 18470 }, { "grad_norm": 0.5838536024093628, "learning_rate": 9.130866733422288e-05, "loss": 0.0666, "step": 18480 }, { "grad_norm": 0.4493616223335266, "learning_rate": 9.129701891318556e-05, "loss": 0.0508, "step": 18490 }, { "grad_norm": 0.3730444014072418, "learning_rate": 9.128536343561718e-05, "loss": 0.0615, "step": 18500 }, { "grad_norm": 0.6205978393554688, "learning_rate": 9.127370090350934e-05, "loss": 0.0649, "step": 18510 }, { "grad_norm": 0.4474904239177704, "learning_rate": 9.126203131885487e-05, "loss": 0.0599, "step": 18520 }, { "grad_norm": 0.45531249046325684, "learning_rate": 9.125035468364775e-05, "loss": 0.0613, "step": 18530 }, { "grad_norm": 0.4057427942752838, "learning_rate": 9.123867099988322e-05, "loss": 0.0534, "step": 18540 }, { "grad_norm": 0.5632643699645996, "learning_rate": 9.122698026955769e-05, "loss": 0.0599, "step": 18550 }, { "grad_norm": 0.4591205418109894, "learning_rate": 9.12152824946688e-05, "loss": 0.0741, "step": 18560 }, { "grad_norm": 0.4664716422557831, "learning_rate": 9.120357767721538e-05, "loss": 0.0562, "step": 18570 }, { "grad_norm": 0.44739121198654175, "learning_rate": 9.119186581919745e-05, "loss": 0.051, "step": 18580 }, { "grad_norm": 0.45305874943733215, "learning_rate": 9.118014692261624e-05, "loss": 0.052, "step": 18590 }, { "grad_norm": 0.4273267984390259, "learning_rate": 9.116842098947422e-05, "loss": 0.072, "step": 18600 }, { "grad_norm": 0.5495984554290771, "learning_rate": 9.115668802177499e-05, "loss": 0.059, "step": 18610 }, { "grad_norm": 0.35206860303878784, "learning_rate": 9.114494802152342e-05, "loss": 0.0698, "step": 18620 }, { "grad_norm": 0.598809003829956, "learning_rate": 9.113320099072555e-05, "loss": 0.0769, "step": 18630 }, { "grad_norm": 0.43508875370025635, "learning_rate": 9.112144693138864e-05, "loss": 0.0614, "step": 18640 }, { "grad_norm": 0.4241705536842346, "learning_rate": 9.110968584552111e-05, "loss": 0.0577, "step": 18650 }, { "grad_norm": 0.5285400152206421, "learning_rate": 9.109791773513264e-05, "loss": 0.0617, "step": 18660 }, { "grad_norm": 0.6180204749107361, "learning_rate": 9.108614260223403e-05, "loss": 0.0691, "step": 18670 }, { "grad_norm": 0.6225810647010803, "learning_rate": 9.107436044883738e-05, "loss": 0.0572, "step": 18680 }, { "grad_norm": 0.6828646063804626, "learning_rate": 9.10625712769559e-05, "loss": 0.0557, "step": 18690 }, { "grad_norm": 0.4229072332382202, "learning_rate": 9.105077508860406e-05, "loss": 0.0514, "step": 18700 }, { "grad_norm": 0.287636935710907, "learning_rate": 9.103897188579751e-05, "loss": 0.0614, "step": 18710 }, { "grad_norm": 0.43207094073295593, "learning_rate": 9.102716167055308e-05, "loss": 0.0496, "step": 18720 }, { "grad_norm": 0.5314966440200806, "learning_rate": 9.10153444448888e-05, "loss": 0.09, "step": 18730 }, { "grad_norm": 0.38210180401802063, "learning_rate": 9.100352021082393e-05, "loss": 0.0437, "step": 18740 }, { "grad_norm": 0.43727347254753113, "learning_rate": 9.099168897037891e-05, "loss": 0.0385, "step": 18750 }, { "grad_norm": 0.4039049744606018, "learning_rate": 9.097985072557538e-05, "loss": 0.0521, "step": 18760 }, { "grad_norm": 0.4016013741493225, "learning_rate": 9.096800547843615e-05, "loss": 0.0722, "step": 18770 }, { "grad_norm": 0.5875208973884583, "learning_rate": 9.095615323098526e-05, "loss": 0.0674, "step": 18780 }, { "grad_norm": 0.53750079870224, "learning_rate": 9.094429398524795e-05, "loss": 0.0553, "step": 18790 }, { "grad_norm": 0.4094647169113159, "learning_rate": 9.093242774325061e-05, "loss": 0.0503, "step": 18800 }, { "grad_norm": 0.47040629386901855, "learning_rate": 9.092055450702088e-05, "loss": 0.0518, "step": 18810 }, { "grad_norm": 0.40421754121780396, "learning_rate": 9.090867427858756e-05, "loss": 0.0559, "step": 18820 }, { "grad_norm": 0.5514789819717407, "learning_rate": 9.089678705998066e-05, "loss": 0.0598, "step": 18830 }, { "grad_norm": 0.587430477142334, "learning_rate": 9.088489285323139e-05, "loss": 0.0523, "step": 18840 }, { "grad_norm": 0.39395037293434143, "learning_rate": 9.087299166037212e-05, "loss": 0.064, "step": 18850 }, { "grad_norm": 0.5534453392028809, "learning_rate": 9.086108348343647e-05, "loss": 0.063, "step": 18860 }, { "grad_norm": 0.5207448601722717, "learning_rate": 9.08491683244592e-05, "loss": 0.0614, "step": 18870 }, { "grad_norm": 0.3353296220302582, "learning_rate": 9.08372461854763e-05, "loss": 0.0479, "step": 18880 }, { "grad_norm": 0.5603250861167908, "learning_rate": 9.082531706852492e-05, "loss": 0.0535, "step": 18890 }, { "grad_norm": 0.48730912804603577, "learning_rate": 9.081338097564342e-05, "loss": 0.0643, "step": 18900 }, { "grad_norm": 0.403842568397522, "learning_rate": 9.080143790887137e-05, "loss": 0.0503, "step": 18910 }, { "grad_norm": 0.4268682301044464, "learning_rate": 9.07894878702495e-05, "loss": 0.0476, "step": 18920 }, { "grad_norm": 0.5071700811386108, "learning_rate": 9.077753086181974e-05, "loss": 0.0592, "step": 18930 }, { "grad_norm": 0.34661272168159485, "learning_rate": 9.076556688562524e-05, "loss": 0.043, "step": 18940 }, { "grad_norm": 0.44555410742759705, "learning_rate": 9.075359594371029e-05, "loss": 0.0452, "step": 18950 }, { "grad_norm": 0.4511100947856903, "learning_rate": 9.07416180381204e-05, "loss": 0.0565, "step": 18960 }, { "grad_norm": 0.7880027890205383, "learning_rate": 9.072963317090228e-05, "loss": 0.0557, "step": 18970 }, { "grad_norm": 0.450387179851532, "learning_rate": 9.071764134410382e-05, "loss": 0.0548, "step": 18980 }, { "grad_norm": 0.34401094913482666, "learning_rate": 9.070564255977407e-05, "loss": 0.0481, "step": 18990 }, { "grad_norm": 0.6720134019851685, "learning_rate": 9.06936368199633e-05, "loss": 0.0728, "step": 19000 }, { "grad_norm": 0.5057258009910583, "learning_rate": 9.0681624126723e-05, "loss": 0.0569, "step": 19010 }, { "grad_norm": 0.4812263250350952, "learning_rate": 9.066960448210576e-05, "loss": 0.0511, "step": 19020 }, { "grad_norm": 0.4169009029865265, "learning_rate": 9.065757788816543e-05, "loss": 0.0428, "step": 19030 }, { "grad_norm": 0.4436280429363251, "learning_rate": 9.064554434695705e-05, "loss": 0.0673, "step": 19040 }, { "grad_norm": 0.4014548361301422, "learning_rate": 9.063350386053677e-05, "loss": 0.0554, "step": 19050 }, { "grad_norm": 0.5226222276687622, "learning_rate": 9.062145643096202e-05, "loss": 0.0537, "step": 19060 }, { "grad_norm": 0.46936917304992676, "learning_rate": 9.060940206029136e-05, "loss": 0.0585, "step": 19070 }, { "grad_norm": 0.6046851873397827, "learning_rate": 9.059734075058457e-05, "loss": 0.0567, "step": 19080 }, { "grad_norm": 0.3353908061981201, "learning_rate": 9.058527250390257e-05, "loss": 0.0527, "step": 19090 }, { "grad_norm": 0.3800169825553894, "learning_rate": 9.057319732230752e-05, "loss": 0.0608, "step": 19100 }, { "grad_norm": 0.3441491723060608, "learning_rate": 9.056111520786273e-05, "loss": 0.0533, "step": 19110 }, { "grad_norm": 0.41201314330101013, "learning_rate": 9.054902616263268e-05, "loss": 0.0592, "step": 19120 }, { "grad_norm": 0.6169248819351196, "learning_rate": 9.05369301886831e-05, "loss": 0.0654, "step": 19130 }, { "grad_norm": 0.4638742208480835, "learning_rate": 9.052482728808083e-05, "loss": 0.0499, "step": 19140 }, { "grad_norm": 0.40791040658950806, "learning_rate": 9.051271746289391e-05, "loss": 0.0478, "step": 19150 }, { "grad_norm": 0.49566924571990967, "learning_rate": 9.050060071519162e-05, "loss": 0.0612, "step": 19160 }, { "grad_norm": 0.3305149972438812, "learning_rate": 9.048847704704437e-05, "loss": 0.0505, "step": 19170 }, { "grad_norm": 0.39824461936950684, "learning_rate": 9.047634646052376e-05, "loss": 0.0567, "step": 19180 }, { "grad_norm": 0.40828737616539, "learning_rate": 9.046420895770256e-05, "loss": 0.0756, "step": 19190 }, { "grad_norm": 0.5292291045188904, "learning_rate": 9.045206454065473e-05, "loss": 0.0533, "step": 19200 }, { "grad_norm": 0.4687177538871765, "learning_rate": 9.043991321145546e-05, "loss": 0.0491, "step": 19210 }, { "grad_norm": 0.45357105135917664, "learning_rate": 9.042775497218105e-05, "loss": 0.0476, "step": 19220 }, { "grad_norm": 0.36166128516197205, "learning_rate": 9.041558982490901e-05, "loss": 0.052, "step": 19230 }, { "grad_norm": 0.5334241390228271, "learning_rate": 9.040341777171805e-05, "loss": 0.0583, "step": 19240 }, { "grad_norm": 0.46557700634002686, "learning_rate": 9.039123881468802e-05, "loss": 0.0557, "step": 19250 }, { "grad_norm": 0.4598061144351959, "learning_rate": 9.037905295589998e-05, "loss": 0.0531, "step": 19260 }, { "grad_norm": 0.5377718806266785, "learning_rate": 9.036686019743617e-05, "loss": 0.05, "step": 19270 }, { "grad_norm": 0.45645672082901, "learning_rate": 9.035466054137997e-05, "loss": 0.0545, "step": 19280 }, { "grad_norm": 0.5349841117858887, "learning_rate": 9.0342453989816e-05, "loss": 0.0622, "step": 19290 }, { "grad_norm": 0.3744739294052124, "learning_rate": 9.033024054483e-05, "loss": 0.0575, "step": 19300 }, { "grad_norm": 0.43147608637809753, "learning_rate": 9.031802020850894e-05, "loss": 0.0531, "step": 19310 }, { "grad_norm": 0.6959304213523865, "learning_rate": 9.030579298294092e-05, "loss": 0.0669, "step": 19320 }, { "grad_norm": 0.4501650631427765, "learning_rate": 9.029355887021524e-05, "loss": 0.0723, "step": 19330 }, { "grad_norm": 0.5536152720451355, "learning_rate": 9.028131787242238e-05, "loss": 0.0732, "step": 19340 }, { "grad_norm": 0.4347698986530304, "learning_rate": 9.026906999165399e-05, "loss": 0.0603, "step": 19350 }, { "grad_norm": 0.4348585307598114, "learning_rate": 9.025681523000291e-05, "loss": 0.0481, "step": 19360 }, { "grad_norm": 0.44833481311798096, "learning_rate": 9.024455358956315e-05, "loss": 0.0517, "step": 19370 }, { "grad_norm": 0.4339860677719116, "learning_rate": 9.023228507242984e-05, "loss": 0.0607, "step": 19380 }, { "grad_norm": 0.41996753215789795, "learning_rate": 9.022000968069937e-05, "loss": 0.0549, "step": 19390 }, { "grad_norm": 0.46141621470451355, "learning_rate": 9.020772741646928e-05, "loss": 0.0529, "step": 19400 }, { "grad_norm": 0.39703741669654846, "learning_rate": 9.019543828183826e-05, "loss": 0.052, "step": 19410 }, { "grad_norm": 0.4045526683330536, "learning_rate": 9.018314227890616e-05, "loss": 0.0459, "step": 19420 }, { "grad_norm": 0.4385145604610443, "learning_rate": 9.017083940977408e-05, "loss": 0.0743, "step": 19430 }, { "grad_norm": 0.5890378355979919, "learning_rate": 9.015852967654422e-05, "loss": 0.0495, "step": 19440 }, { "grad_norm": 0.42482849955558777, "learning_rate": 9.014621308131996e-05, "loss": 0.0483, "step": 19450 }, { "grad_norm": 0.4680790603160858, "learning_rate": 9.01338896262059e-05, "loss": 0.0558, "step": 19460 }, { "grad_norm": 0.49047791957855225, "learning_rate": 9.012155931330777e-05, "loss": 0.0672, "step": 19470 }, { "grad_norm": 0.4147203266620636, "learning_rate": 9.010922214473246e-05, "loss": 0.0548, "step": 19480 }, { "grad_norm": 0.636111319065094, "learning_rate": 9.009687812258808e-05, "loss": 0.0755, "step": 19490 }, { "grad_norm": 0.5024824142456055, "learning_rate": 9.00845272489839e-05, "loss": 0.057, "step": 19500 }, { "grad_norm": 0.46947798132896423, "learning_rate": 9.007216952603031e-05, "loss": 0.0708, "step": 19510 }, { "grad_norm": 0.40516963601112366, "learning_rate": 9.005980495583894e-05, "loss": 0.0477, "step": 19520 }, { "grad_norm": 0.5301535129547119, "learning_rate": 9.004743354052252e-05, "loss": 0.0546, "step": 19530 }, { "grad_norm": 0.42299872636795044, "learning_rate": 9.003505528219503e-05, "loss": 0.0476, "step": 19540 }, { "grad_norm": 0.5043705701828003, "learning_rate": 9.002267018297154e-05, "loss": 0.0586, "step": 19550 }, { "grad_norm": 0.4844409227371216, "learning_rate": 9.001027824496834e-05, "loss": 0.0647, "step": 19560 }, { "grad_norm": 0.41248220205307007, "learning_rate": 8.999787947030287e-05, "loss": 0.0517, "step": 19570 }, { "grad_norm": 0.5104003548622131, "learning_rate": 8.998547386109376e-05, "loss": 0.065, "step": 19580 }, { "grad_norm": 0.359737366437912, "learning_rate": 8.997306141946073e-05, "loss": 0.0508, "step": 19590 }, { "grad_norm": 0.44135624170303345, "learning_rate": 8.996064214752481e-05, "loss": 0.0696, "step": 19600 }, { "grad_norm": 0.33548271656036377, "learning_rate": 8.994821604740806e-05, "loss": 0.0495, "step": 19610 }, { "grad_norm": 0.4844847321510315, "learning_rate": 8.993578312123377e-05, "loss": 0.0667, "step": 19620 }, { "grad_norm": 0.4781039357185364, "learning_rate": 8.992334337112639e-05, "loss": 0.0493, "step": 19630 }, { "grad_norm": 0.6422054767608643, "learning_rate": 8.991089679921154e-05, "loss": 0.0541, "step": 19640 }, { "grad_norm": 0.5742456316947937, "learning_rate": 8.989844340761599e-05, "loss": 0.0615, "step": 19650 }, { "grad_norm": 0.46189218759536743, "learning_rate": 8.988598319846768e-05, "loss": 0.0492, "step": 19660 }, { "grad_norm": 0.5941416025161743, "learning_rate": 8.987351617389574e-05, "loss": 0.06, "step": 19670 }, { "grad_norm": 0.450020968914032, "learning_rate": 8.98610423360304e-05, "loss": 0.055, "step": 19680 }, { "grad_norm": 0.48521658778190613, "learning_rate": 8.984856168700317e-05, "loss": 0.0532, "step": 19690 }, { "grad_norm": 0.43659278750419617, "learning_rate": 8.983607422894658e-05, "loss": 0.0484, "step": 19700 }, { "grad_norm": 0.46131500601768494, "learning_rate": 8.982357996399442e-05, "loss": 0.0523, "step": 19710 }, { "grad_norm": 0.4139968156814575, "learning_rate": 8.981107889428164e-05, "loss": 0.0573, "step": 19720 }, { "grad_norm": 0.4372800886631012, "learning_rate": 8.979857102194428e-05, "loss": 0.0509, "step": 19730 }, { "grad_norm": 0.42415350675582886, "learning_rate": 8.978605634911968e-05, "loss": 0.0565, "step": 19740 }, { "grad_norm": 0.44390255212783813, "learning_rate": 8.977353487794616e-05, "loss": 0.0469, "step": 19750 }, { "grad_norm": 0.4632333815097809, "learning_rate": 8.976100661056334e-05, "loss": 0.0626, "step": 19760 }, { "grad_norm": 0.4698545038700104, "learning_rate": 8.974847154911197e-05, "loss": 0.0537, "step": 19770 }, { "grad_norm": 0.4559441804885864, "learning_rate": 8.973592969573393e-05, "loss": 0.0565, "step": 19780 }, { "grad_norm": 0.4992777705192566, "learning_rate": 8.972338105257228e-05, "loss": 0.0602, "step": 19790 }, { "grad_norm": 0.3863585293292999, "learning_rate": 8.971082562177125e-05, "loss": 0.0575, "step": 19800 }, { "grad_norm": 0.4998868405818939, "learning_rate": 8.96982634054762e-05, "loss": 0.0537, "step": 19810 }, { "grad_norm": 0.3462291657924652, "learning_rate": 8.96856944058337e-05, "loss": 0.0511, "step": 19820 }, { "grad_norm": 0.3877675533294678, "learning_rate": 8.967311862499144e-05, "loss": 0.0463, "step": 19830 }, { "grad_norm": 0.3140927851200104, "learning_rate": 8.966053606509825e-05, "loss": 0.0451, "step": 19840 }, { "grad_norm": 0.5298641324043274, "learning_rate": 8.964794672830417e-05, "loss": 0.0618, "step": 19850 }, { "grad_norm": 0.492005854845047, "learning_rate": 8.963535061676038e-05, "loss": 0.0556, "step": 19860 }, { "grad_norm": 0.5406186580657959, "learning_rate": 8.962274773261918e-05, "loss": 0.0503, "step": 19870 }, { "grad_norm": 0.41342487931251526, "learning_rate": 8.961013807803409e-05, "loss": 0.0568, "step": 19880 }, { "grad_norm": 0.3772391974925995, "learning_rate": 8.959752165515973e-05, "loss": 0.0562, "step": 19890 }, { "grad_norm": 0.41843220591545105, "learning_rate": 8.958489846615193e-05, "loss": 0.068, "step": 19900 }, { "grad_norm": 0.5172046422958374, "learning_rate": 8.957226851316762e-05, "loss": 0.0622, "step": 19910 }, { "grad_norm": 0.4547855257987976, "learning_rate": 8.955963179836493e-05, "loss": 0.0614, "step": 19920 }, { "grad_norm": 0.36564958095550537, "learning_rate": 8.954698832390312e-05, "loss": 0.0597, "step": 19930 }, { "grad_norm": 0.5796594619750977, "learning_rate": 8.953433809194263e-05, "loss": 0.06, "step": 19940 }, { "grad_norm": 0.41314268112182617, "learning_rate": 8.9521681104645e-05, "loss": 0.0576, "step": 19950 }, { "grad_norm": 0.5212969183921814, "learning_rate": 8.9509017364173e-05, "loss": 0.0555, "step": 19960 }, { "grad_norm": 0.6565716862678528, "learning_rate": 8.949634687269052e-05, "loss": 0.0588, "step": 19970 }, { "grad_norm": 0.5026572346687317, "learning_rate": 8.948366963236259e-05, "loss": 0.0594, "step": 19980 }, { "grad_norm": 0.36079058051109314, "learning_rate": 8.947098564535538e-05, "loss": 0.0476, "step": 19990 }, { "grad_norm": 0.5253843665122986, "learning_rate": 8.945829491383627e-05, "loss": 0.0641, "step": 20000 }, { "grad_norm": 0.5590565204620361, "learning_rate": 8.944559743997374e-05, "loss": 0.0639, "step": 20010 }, { "grad_norm": 0.4926326870918274, "learning_rate": 8.943289322593746e-05, "loss": 0.0478, "step": 20020 }, { "grad_norm": 0.4053904116153717, "learning_rate": 8.942018227389821e-05, "loss": 0.05, "step": 20030 }, { "grad_norm": 0.3728894293308258, "learning_rate": 8.940746458602795e-05, "loss": 0.0644, "step": 20040 }, { "grad_norm": 0.46741995215415955, "learning_rate": 8.939474016449979e-05, "loss": 0.0638, "step": 20050 }, { "grad_norm": 0.4785122871398926, "learning_rate": 8.938200901148799e-05, "loss": 0.0471, "step": 20060 }, { "grad_norm": 0.7007537484169006, "learning_rate": 8.936927112916795e-05, "loss": 0.0659, "step": 20070 }, { "grad_norm": 0.3846506178379059, "learning_rate": 8.935652651971622e-05, "loss": 0.048, "step": 20080 }, { "grad_norm": 0.5250885486602783, "learning_rate": 8.934377518531052e-05, "loss": 0.0591, "step": 20090 }, { "grad_norm": 0.4202224314212799, "learning_rate": 8.933101712812967e-05, "loss": 0.0482, "step": 20100 }, { "grad_norm": 0.5351470112800598, "learning_rate": 8.931825235035374e-05, "loss": 0.06, "step": 20110 }, { "grad_norm": 0.5001919269561768, "learning_rate": 8.930548085416382e-05, "loss": 0.0556, "step": 20120 }, { "grad_norm": 0.3597191870212555, "learning_rate": 8.92927026417422e-05, "loss": 0.0582, "step": 20130 }, { "grad_norm": 0.4585297703742981, "learning_rate": 8.92799177152724e-05, "loss": 0.052, "step": 20140 }, { "grad_norm": 0.4320634603500366, "learning_rate": 8.926712607693895e-05, "loss": 0.058, "step": 20150 }, { "grad_norm": 0.6036260724067688, "learning_rate": 8.925432772892762e-05, "loss": 0.0517, "step": 20160 }, { "grad_norm": 0.49237921833992004, "learning_rate": 8.924152267342529e-05, "loss": 0.0573, "step": 20170 }, { "grad_norm": 0.46246254444122314, "learning_rate": 8.922871091261998e-05, "loss": 0.0523, "step": 20180 }, { "grad_norm": 0.48778852820396423, "learning_rate": 8.92158924487009e-05, "loss": 0.0521, "step": 20190 }, { "grad_norm": 0.3837634027004242, "learning_rate": 8.920306728385834e-05, "loss": 0.05, "step": 20200 }, { "grad_norm": 0.586142897605896, "learning_rate": 8.919023542028379e-05, "loss": 0.0524, "step": 20210 }, { "grad_norm": 0.40260228514671326, "learning_rate": 8.917739686016988e-05, "loss": 0.0448, "step": 20220 }, { "grad_norm": 0.4984338581562042, "learning_rate": 8.916455160571033e-05, "loss": 0.0578, "step": 20230 }, { "grad_norm": 0.5659269690513611, "learning_rate": 8.915169965910008e-05, "loss": 0.0547, "step": 20240 }, { "grad_norm": 0.4210081696510315, "learning_rate": 8.913884102253514e-05, "loss": 0.0566, "step": 20250 }, { "grad_norm": 0.3986816704273224, "learning_rate": 8.912597569821273e-05, "loss": 0.0457, "step": 20260 }, { "grad_norm": 0.4185681641101837, "learning_rate": 8.911310368833118e-05, "loss": 0.0508, "step": 20270 }, { "grad_norm": 0.5127096176147461, "learning_rate": 8.910022499508994e-05, "loss": 0.0448, "step": 20280 }, { "grad_norm": 0.3406956195831299, "learning_rate": 8.908733962068965e-05, "loss": 0.0595, "step": 20290 }, { "grad_norm": 0.45081833004951477, "learning_rate": 8.907444756733207e-05, "loss": 0.0534, "step": 20300 }, { "grad_norm": 0.3560790419578552, "learning_rate": 8.906154883722006e-05, "loss": 0.0604, "step": 20310 }, { "grad_norm": 0.5073303580284119, "learning_rate": 8.904864343255773e-05, "loss": 0.06, "step": 20320 }, { "grad_norm": 0.5100357532501221, "learning_rate": 8.90357313555502e-05, "loss": 0.0493, "step": 20330 }, { "grad_norm": 0.30712050199508667, "learning_rate": 8.90228126084038e-05, "loss": 0.0512, "step": 20340 }, { "grad_norm": 0.4676774740219116, "learning_rate": 8.900988719332601e-05, "loss": 0.0534, "step": 20350 }, { "grad_norm": 0.4481925070285797, "learning_rate": 8.899695511252542e-05, "loss": 0.0562, "step": 20360 }, { "grad_norm": 0.36468949913978577, "learning_rate": 8.898401636821176e-05, "loss": 0.0546, "step": 20370 }, { "grad_norm": 0.4564686715602875, "learning_rate": 8.897107096259593e-05, "loss": 0.0584, "step": 20380 }, { "grad_norm": 0.3883002996444702, "learning_rate": 8.895811889788994e-05, "loss": 0.0508, "step": 20390 }, { "grad_norm": 0.3712468147277832, "learning_rate": 8.894516017630692e-05, "loss": 0.0475, "step": 20400 }, { "grad_norm": 0.3943104147911072, "learning_rate": 8.893219480006118e-05, "loss": 0.0516, "step": 20410 }, { "grad_norm": 0.5235199928283691, "learning_rate": 8.891922277136817e-05, "loss": 0.0592, "step": 20420 }, { "grad_norm": 0.44092461466789246, "learning_rate": 8.890624409244441e-05, "loss": 0.0576, "step": 20430 }, { "grad_norm": 0.42935457825660706, "learning_rate": 8.889325876550763e-05, "loss": 0.0566, "step": 20440 }, { "grad_norm": 0.4995921552181244, "learning_rate": 8.888026679277666e-05, "loss": 0.0508, "step": 20450 }, { "grad_norm": 0.38707348704338074, "learning_rate": 8.886726817647147e-05, "loss": 0.0832, "step": 20460 }, { "grad_norm": 0.5162052512168884, "learning_rate": 8.885426291881319e-05, "loss": 0.0691, "step": 20470 }, { "grad_norm": 0.5147560834884644, "learning_rate": 8.884125102202401e-05, "loss": 0.0571, "step": 20480 }, { "grad_norm": 0.4277162551879883, "learning_rate": 8.882823248832736e-05, "loss": 0.0473, "step": 20490 }, { "grad_norm": 0.5090664029121399, "learning_rate": 8.881520731994772e-05, "loss": 0.0538, "step": 20500 }, { "grad_norm": 0.2959591746330261, "learning_rate": 8.880217551911077e-05, "loss": 0.0532, "step": 20510 }, { "grad_norm": 0.36124369502067566, "learning_rate": 8.878913708804323e-05, "loss": 0.0521, "step": 20520 }, { "grad_norm": 0.48629796504974365, "learning_rate": 8.877609202897308e-05, "loss": 0.0526, "step": 20530 }, { "grad_norm": 0.41337716579437256, "learning_rate": 8.876304034412933e-05, "loss": 0.0491, "step": 20540 }, { "grad_norm": 0.4370241165161133, "learning_rate": 8.874998203574214e-05, "loss": 0.0554, "step": 20550 }, { "grad_norm": 0.6514738202095032, "learning_rate": 8.873691710604284e-05, "loss": 0.0451, "step": 20560 }, { "grad_norm": 0.5570005178451538, "learning_rate": 8.872384555726387e-05, "loss": 0.0548, "step": 20570 }, { "grad_norm": 0.6530277729034424, "learning_rate": 8.871076739163878e-05, "loss": 0.0537, "step": 20580 }, { "grad_norm": 0.4667433798313141, "learning_rate": 8.86976826114023e-05, "loss": 0.0553, "step": 20590 }, { "grad_norm": 0.4540743827819824, "learning_rate": 8.868459121879023e-05, "loss": 0.0582, "step": 20600 }, { "grad_norm": 0.4316296875476837, "learning_rate": 8.867149321603956e-05, "loss": 0.0655, "step": 20610 }, { "grad_norm": 0.5392968058586121, "learning_rate": 8.865838860538835e-05, "loss": 0.0568, "step": 20620 }, { "grad_norm": 0.5097460746765137, "learning_rate": 8.864527738907585e-05, "loss": 0.0493, "step": 20630 }, { "grad_norm": 0.4226332902908325, "learning_rate": 8.863215956934239e-05, "loss": 0.0532, "step": 20640 }, { "grad_norm": 0.44279083609580994, "learning_rate": 8.861903514842947e-05, "loss": 0.0418, "step": 20650 }, { "grad_norm": 0.36200571060180664, "learning_rate": 8.860590412857966e-05, "loss": 0.0468, "step": 20660 }, { "grad_norm": 0.5019418001174927, "learning_rate": 8.85927665120367e-05, "loss": 0.0581, "step": 20670 }, { "grad_norm": 0.5741456747055054, "learning_rate": 8.857962230104546e-05, "loss": 0.0595, "step": 20680 }, { "grad_norm": 0.4767618775367737, "learning_rate": 8.856647149785193e-05, "loss": 0.0458, "step": 20690 }, { "grad_norm": 0.44423574209213257, "learning_rate": 8.855331410470322e-05, "loss": 0.0549, "step": 20700 }, { "grad_norm": 0.41790130734443665, "learning_rate": 8.854015012384756e-05, "loss": 0.0514, "step": 20710 }, { "grad_norm": 0.46500420570373535, "learning_rate": 8.852697955753433e-05, "loss": 0.0539, "step": 20720 }, { "grad_norm": 0.41479381918907166, "learning_rate": 8.851380240801399e-05, "loss": 0.0595, "step": 20730 }, { "grad_norm": 0.35073891282081604, "learning_rate": 8.850061867753818e-05, "loss": 0.0359, "step": 20740 }, { "grad_norm": 0.4755604863166809, "learning_rate": 8.848742836835963e-05, "loss": 0.0575, "step": 20750 }, { "grad_norm": 0.47525009512901306, "learning_rate": 8.847423148273221e-05, "loss": 0.0542, "step": 20760 }, { "grad_norm": 0.4804207980632782, "learning_rate": 8.846102802291092e-05, "loss": 0.0683, "step": 20770 }, { "grad_norm": 0.46845096349716187, "learning_rate": 8.844781799115183e-05, "loss": 0.0582, "step": 20780 }, { "grad_norm": 0.5818317532539368, "learning_rate": 8.84346013897122e-05, "loss": 0.0434, "step": 20790 }, { "grad_norm": 0.4564650356769562, "learning_rate": 8.842137822085038e-05, "loss": 0.0636, "step": 20800 }, { "grad_norm": 0.5493792295455933, "learning_rate": 8.840814848682585e-05, "loss": 0.046, "step": 20810 }, { "grad_norm": 0.35649174451828003, "learning_rate": 8.83949121898992e-05, "loss": 0.0619, "step": 20820 }, { "grad_norm": 0.37773987650871277, "learning_rate": 8.838166933233217e-05, "loss": 0.0483, "step": 20830 }, { "grad_norm": 0.28669315576553345, "learning_rate": 8.83684199163876e-05, "loss": 0.0531, "step": 20840 }, { "grad_norm": 0.3376960754394531, "learning_rate": 8.835516394432943e-05, "loss": 0.0519, "step": 20850 }, { "grad_norm": 0.5586808919906616, "learning_rate": 8.834190141842276e-05, "loss": 0.057, "step": 20860 }, { "grad_norm": 0.40092453360557556, "learning_rate": 8.83286323409338e-05, "loss": 0.047, "step": 20870 }, { "grad_norm": 0.4524976313114166, "learning_rate": 8.831535671412986e-05, "loss": 0.0566, "step": 20880 }, { "grad_norm": 0.5465089082717896, "learning_rate": 8.830207454027938e-05, "loss": 0.0561, "step": 20890 }, { "grad_norm": 0.4086674749851227, "learning_rate": 8.828878582165192e-05, "loss": 0.0568, "step": 20900 }, { "grad_norm": 0.4508064389228821, "learning_rate": 8.827549056051818e-05, "loss": 0.0501, "step": 20910 }, { "grad_norm": 0.4247058033943176, "learning_rate": 8.826218875914993e-05, "loss": 0.0549, "step": 20920 }, { "grad_norm": 0.40854746103286743, "learning_rate": 8.82488804198201e-05, "loss": 0.0478, "step": 20930 }, { "grad_norm": 0.29377391934394836, "learning_rate": 8.82355655448027e-05, "loss": 0.0477, "step": 20940 }, { "grad_norm": 0.5301110744476318, "learning_rate": 8.822224413637293e-05, "loss": 0.0544, "step": 20950 }, { "grad_norm": 0.4532994031906128, "learning_rate": 8.820891619680697e-05, "loss": 0.0459, "step": 20960 }, { "grad_norm": 0.47871634364128113, "learning_rate": 8.819558172838227e-05, "loss": 0.0524, "step": 20970 }, { "grad_norm": 0.39979612827301025, "learning_rate": 8.818224073337731e-05, "loss": 0.0465, "step": 20980 }, { "grad_norm": 0.39805856347084045, "learning_rate": 8.816889321407169e-05, "loss": 0.0499, "step": 20990 }, { "grad_norm": 0.5621553659439087, "learning_rate": 8.815553917274615e-05, "loss": 0.0528, "step": 21000 }, { "grad_norm": 0.38345280289649963, "learning_rate": 8.81421786116825e-05, "loss": 0.0655, "step": 21010 }, { "grad_norm": 0.5616830587387085, "learning_rate": 8.812881153316373e-05, "loss": 0.0562, "step": 21020 }, { "grad_norm": 0.4901368021965027, "learning_rate": 8.81154379394739e-05, "loss": 0.0629, "step": 21030 }, { "grad_norm": 0.4471442401409149, "learning_rate": 8.810205783289818e-05, "loss": 0.0574, "step": 21040 }, { "grad_norm": 0.5397526621818542, "learning_rate": 8.808867121572286e-05, "loss": 0.0508, "step": 21050 }, { "grad_norm": 0.4995998740196228, "learning_rate": 8.807527809023537e-05, "loss": 0.0682, "step": 21060 }, { "grad_norm": 0.39249515533447266, "learning_rate": 8.80618784587242e-05, "loss": 0.0463, "step": 21070 }, { "grad_norm": 0.42725643515586853, "learning_rate": 8.804847232347902e-05, "loss": 0.0567, "step": 21080 }, { "grad_norm": 0.48477914929389954, "learning_rate": 8.803505968679054e-05, "loss": 0.0586, "step": 21090 }, { "grad_norm": 0.5196827054023743, "learning_rate": 8.802164055095061e-05, "loss": 0.0664, "step": 21100 }, { "grad_norm": 0.4387155771255493, "learning_rate": 8.80082149182522e-05, "loss": 0.0494, "step": 21110 }, { "grad_norm": 0.38872143626213074, "learning_rate": 8.79947827909894e-05, "loss": 0.0474, "step": 21120 }, { "grad_norm": 0.3758730888366699, "learning_rate": 8.798134417145738e-05, "loss": 0.0554, "step": 21130 }, { "grad_norm": 0.46240735054016113, "learning_rate": 8.796789906195243e-05, "loss": 0.0529, "step": 21140 }, { "grad_norm": 0.43077167868614197, "learning_rate": 8.795444746477195e-05, "loss": 0.0566, "step": 21150 }, { "grad_norm": 0.44085776805877686, "learning_rate": 8.794098938221446e-05, "loss": 0.052, "step": 21160 }, { "grad_norm": 0.43823814392089844, "learning_rate": 8.792752481657957e-05, "loss": 0.0541, "step": 21170 }, { "grad_norm": 0.45917147397994995, "learning_rate": 8.791405377016802e-05, "loss": 0.0562, "step": 21180 }, { "grad_norm": 0.6941176056861877, "learning_rate": 8.790057624528163e-05, "loss": 0.0568, "step": 21190 }, { "grad_norm": 0.45972663164138794, "learning_rate": 8.788709224422333e-05, "loss": 0.0539, "step": 21200 }, { "grad_norm": 0.6137548685073853, "learning_rate": 8.787360176929717e-05, "loss": 0.049, "step": 21210 }, { "grad_norm": 0.555791437625885, "learning_rate": 8.786010482280834e-05, "loss": 0.0598, "step": 21220 }, { "grad_norm": 0.3788408041000366, "learning_rate": 8.784660140706306e-05, "loss": 0.0579, "step": 21230 }, { "grad_norm": 0.5504308938980103, "learning_rate": 8.783309152436872e-05, "loss": 0.0461, "step": 21240 }, { "grad_norm": 0.48645034432411194, "learning_rate": 8.781957517703375e-05, "loss": 0.0451, "step": 21250 }, { "grad_norm": 0.43467676639556885, "learning_rate": 8.780605236736776e-05, "loss": 0.0576, "step": 21260 }, { "grad_norm": 0.38826411962509155, "learning_rate": 8.779252309768142e-05, "loss": 0.0589, "step": 21270 }, { "grad_norm": 0.5017406344413757, "learning_rate": 8.777898737028652e-05, "loss": 0.0578, "step": 21280 }, { "grad_norm": 0.5313907861709595, "learning_rate": 8.776544518749591e-05, "loss": 0.0584, "step": 21290 }, { "grad_norm": 0.3906121253967285, "learning_rate": 8.775189655162364e-05, "loss": 0.0442, "step": 21300 }, { "grad_norm": 0.40902194380760193, "learning_rate": 8.773834146498474e-05, "loss": 0.0462, "step": 21310 }, { "grad_norm": 0.3543802499771118, "learning_rate": 8.772477992989545e-05, "loss": 0.0453, "step": 21320 }, { "grad_norm": 0.5134013891220093, "learning_rate": 8.771121194867304e-05, "loss": 0.0639, "step": 21330 }, { "grad_norm": 0.8063485622406006, "learning_rate": 8.769763752363589e-05, "loss": 0.0564, "step": 21340 }, { "grad_norm": 0.38001424074172974, "learning_rate": 8.768405665710352e-05, "loss": 0.0525, "step": 21350 }, { "grad_norm": 0.4208486080169678, "learning_rate": 8.767046935139655e-05, "loss": 0.0556, "step": 21360 }, { "grad_norm": 0.33756738901138306, "learning_rate": 8.765687560883666e-05, "loss": 0.0494, "step": 21370 }, { "grad_norm": 0.8404547572135925, "learning_rate": 8.764327543174664e-05, "loss": 0.0499, "step": 21380 }, { "grad_norm": 0.5076411366462708, "learning_rate": 8.762966882245038e-05, "loss": 0.0462, "step": 21390 }, { "grad_norm": 0.573948085308075, "learning_rate": 8.761605578327291e-05, "loss": 0.065, "step": 21400 }, { "grad_norm": 0.37743836641311646, "learning_rate": 8.76024363165403e-05, "loss": 0.0696, "step": 21410 }, { "grad_norm": 0.4418126344680786, "learning_rate": 8.758881042457976e-05, "loss": 0.0529, "step": 21420 }, { "grad_norm": 0.44787874817848206, "learning_rate": 8.757517810971957e-05, "loss": 0.0497, "step": 21430 }, { "grad_norm": 0.4451925456523895, "learning_rate": 8.756153937428913e-05, "loss": 0.0553, "step": 21440 }, { "grad_norm": 0.43845120072364807, "learning_rate": 8.754789422061889e-05, "loss": 0.0644, "step": 21450 }, { "grad_norm": 0.35740360617637634, "learning_rate": 8.753424265104052e-05, "loss": 0.046, "step": 21460 }, { "grad_norm": 0.48515474796295166, "learning_rate": 8.752058466788659e-05, "loss": 0.0643, "step": 21470 }, { "grad_norm": 0.5395092964172363, "learning_rate": 8.750692027349097e-05, "loss": 0.0489, "step": 21480 }, { "grad_norm": 0.45201873779296875, "learning_rate": 8.749324947018847e-05, "loss": 0.0535, "step": 21490 }, { "grad_norm": 0.40997186303138733, "learning_rate": 8.747957226031507e-05, "loss": 0.0488, "step": 21500 }, { "grad_norm": 0.4272725582122803, "learning_rate": 8.746588864620787e-05, "loss": 0.0572, "step": 21510 }, { "grad_norm": 0.44307199120521545, "learning_rate": 8.745219863020498e-05, "loss": 0.0552, "step": 21520 }, { "grad_norm": 0.4387497901916504, "learning_rate": 8.743850221464564e-05, "loss": 0.0467, "step": 21530 }, { "grad_norm": 0.4224478006362915, "learning_rate": 8.742479940187026e-05, "loss": 0.0524, "step": 21540 }, { "grad_norm": 0.5681551694869995, "learning_rate": 8.74110901942202e-05, "loss": 0.0568, "step": 21550 }, { "grad_norm": 0.4821337163448334, "learning_rate": 8.739737459403803e-05, "loss": 0.0515, "step": 21560 }, { "grad_norm": 0.364027202129364, "learning_rate": 8.738365260366737e-05, "loss": 0.0523, "step": 21570 }, { "grad_norm": 0.46532973647117615, "learning_rate": 8.736992422545292e-05, "loss": 0.042, "step": 21580 }, { "grad_norm": 0.43760064244270325, "learning_rate": 8.73561894617405e-05, "loss": 0.0657, "step": 21590 }, { "grad_norm": 0.5554463863372803, "learning_rate": 8.734244831487697e-05, "loss": 0.054, "step": 21600 }, { "grad_norm": 0.46088099479675293, "learning_rate": 8.732870078721035e-05, "loss": 0.0616, "step": 21610 }, { "grad_norm": 0.5587250590324402, "learning_rate": 8.731494688108972e-05, "loss": 0.0542, "step": 21620 }, { "grad_norm": 0.3841747045516968, "learning_rate": 8.730118659886523e-05, "loss": 0.0586, "step": 21630 }, { "grad_norm": 0.5069071054458618, "learning_rate": 8.728741994288814e-05, "loss": 0.0563, "step": 21640 }, { "grad_norm": 0.3884247839450836, "learning_rate": 8.727364691551079e-05, "loss": 0.0491, "step": 21650 }, { "grad_norm": 0.3729124367237091, "learning_rate": 8.725986751908661e-05, "loss": 0.0618, "step": 21660 }, { "grad_norm": 0.31090983748435974, "learning_rate": 8.724608175597016e-05, "loss": 0.0404, "step": 21670 }, { "grad_norm": 0.3855820596218109, "learning_rate": 8.723228962851699e-05, "loss": 0.0497, "step": 21680 }, { "grad_norm": 0.8020986914634705, "learning_rate": 8.721849113908385e-05, "loss": 0.0521, "step": 21690 }, { "grad_norm": 0.4392431080341339, "learning_rate": 8.720468629002848e-05, "loss": 0.0467, "step": 21700 }, { "grad_norm": 0.5336581468582153, "learning_rate": 8.719087508370978e-05, "loss": 0.0482, "step": 21710 }, { "grad_norm": 0.38103821873664856, "learning_rate": 8.717705752248772e-05, "loss": 0.0547, "step": 21720 }, { "grad_norm": 0.4525810480117798, "learning_rate": 8.71632336087233e-05, "loss": 0.0428, "step": 21730 }, { "grad_norm": 0.7424191832542419, "learning_rate": 8.71494033447787e-05, "loss": 0.0514, "step": 21740 }, { "grad_norm": 0.3965155780315399, "learning_rate": 8.713556673301708e-05, "loss": 0.0444, "step": 21750 }, { "grad_norm": 0.44887009263038635, "learning_rate": 8.712172377580278e-05, "loss": 0.0638, "step": 21760 }, { "grad_norm": 0.521520733833313, "learning_rate": 8.710787447550114e-05, "loss": 0.0611, "step": 21770 }, { "grad_norm": 0.46409907937049866, "learning_rate": 8.70940188344787e-05, "loss": 0.0707, "step": 21780 }, { "grad_norm": 0.7458663582801819, "learning_rate": 8.708015685510293e-05, "loss": 0.0552, "step": 21790 }, { "grad_norm": 0.45353731513023376, "learning_rate": 8.706628853974252e-05, "loss": 0.0581, "step": 21800 }, { "grad_norm": 0.45424261689186096, "learning_rate": 8.705241389076715e-05, "loss": 0.0494, "step": 21810 }, { "grad_norm": 0.3586316406726837, "learning_rate": 8.703853291054764e-05, "loss": 0.0545, "step": 21820 }, { "grad_norm": 0.5026910901069641, "learning_rate": 8.702464560145587e-05, "loss": 0.0605, "step": 21830 }, { "grad_norm": 0.41456353664398193, "learning_rate": 8.701075196586476e-05, "loss": 0.0485, "step": 21840 }, { "grad_norm": 0.499266117811203, "learning_rate": 8.699685200614842e-05, "loss": 0.0497, "step": 21850 }, { "grad_norm": 0.44008544087409973, "learning_rate": 8.698294572468193e-05, "loss": 0.0578, "step": 21860 }, { "grad_norm": 0.47472989559173584, "learning_rate": 8.696903312384148e-05, "loss": 0.0563, "step": 21870 }, { "grad_norm": 0.4489458203315735, "learning_rate": 8.695511420600439e-05, "loss": 0.055, "step": 21880 }, { "grad_norm": 0.34716200828552246, "learning_rate": 8.694118897354901e-05, "loss": 0.0534, "step": 21890 }, { "grad_norm": 0.559211254119873, "learning_rate": 8.692725742885478e-05, "loss": 0.0498, "step": 21900 }, { "grad_norm": 0.4432108998298645, "learning_rate": 8.691331957430221e-05, "loss": 0.0482, "step": 21910 }, { "grad_norm": 0.42511093616485596, "learning_rate": 8.68993754122729e-05, "loss": 0.0525, "step": 21920 }, { "grad_norm": 0.4352816045284271, "learning_rate": 8.688542494514955e-05, "loss": 0.0487, "step": 21930 }, { "grad_norm": 0.533475935459137, "learning_rate": 8.68714681753159e-05, "loss": 0.0547, "step": 21940 }, { "grad_norm": 0.4349527657032013, "learning_rate": 8.685750510515676e-05, "loss": 0.0457, "step": 21950 }, { "grad_norm": 0.3692726492881775, "learning_rate": 8.684353573705805e-05, "loss": 0.053, "step": 21960 }, { "grad_norm": 0.36711031198501587, "learning_rate": 8.682956007340677e-05, "loss": 0.044, "step": 21970 }, { "grad_norm": 0.48376601934432983, "learning_rate": 8.681557811659095e-05, "loss": 0.0474, "step": 21980 }, { "grad_norm": 0.6279628872871399, "learning_rate": 8.680158986899974e-05, "loss": 0.0494, "step": 21990 }, { "grad_norm": 0.5155056118965149, "learning_rate": 8.678759533302335e-05, "loss": 0.0493, "step": 22000 }, { "grad_norm": 0.3613743484020233, "learning_rate": 8.677359451105308e-05, "loss": 0.0384, "step": 22010 }, { "grad_norm": 0.285760760307312, "learning_rate": 8.675958740548123e-05, "loss": 0.0495, "step": 22020 }, { "grad_norm": 0.34992706775665283, "learning_rate": 8.674557401870129e-05, "loss": 0.0557, "step": 22030 }, { "grad_norm": 0.5860740542411804, "learning_rate": 8.673155435310775e-05, "loss": 0.0587, "step": 22040 }, { "grad_norm": 0.6662628054618835, "learning_rate": 8.671752841109617e-05, "loss": 0.0474, "step": 22050 }, { "grad_norm": 0.4101827144622803, "learning_rate": 8.670349619506321e-05, "loss": 0.0468, "step": 22060 }, { "grad_norm": 0.4729050397872925, "learning_rate": 8.66894577074066e-05, "loss": 0.0482, "step": 22070 }, { "grad_norm": 0.4787956178188324, "learning_rate": 8.667541295052513e-05, "loss": 0.0465, "step": 22080 }, { "grad_norm": 0.47539156675338745, "learning_rate": 8.666136192681865e-05, "loss": 0.0579, "step": 22090 }, { "grad_norm": 0.40650102496147156, "learning_rate": 8.664730463868811e-05, "loss": 0.0489, "step": 22100 }, { "grad_norm": 0.5631855130195618, "learning_rate": 8.663324108853552e-05, "loss": 0.0591, "step": 22110 }, { "grad_norm": 0.4036981463432312, "learning_rate": 8.661917127876395e-05, "loss": 0.0729, "step": 22120 }, { "grad_norm": 0.40373694896698, "learning_rate": 8.660509521177754e-05, "loss": 0.0585, "step": 22130 }, { "grad_norm": 0.40905147790908813, "learning_rate": 8.65910128899815e-05, "loss": 0.0559, "step": 22140 }, { "grad_norm": 0.38197022676467896, "learning_rate": 8.657692431578214e-05, "loss": 0.0475, "step": 22150 }, { "grad_norm": 0.3609628975391388, "learning_rate": 8.656282949158679e-05, "loss": 0.0405, "step": 22160 }, { "grad_norm": 0.3604646325111389, "learning_rate": 8.654872841980388e-05, "loss": 0.0406, "step": 22170 }, { "grad_norm": 0.4301302134990692, "learning_rate": 8.653462110284289e-05, "loss": 0.0435, "step": 22180 }, { "grad_norm": 0.4584335386753082, "learning_rate": 8.652050754311437e-05, "loss": 0.0452, "step": 22190 }, { "grad_norm": 0.6037063598632812, "learning_rate": 8.650638774302995e-05, "loss": 0.0532, "step": 22200 }, { "grad_norm": 0.5808549523353577, "learning_rate": 8.649226170500233e-05, "loss": 0.0541, "step": 22210 }, { "grad_norm": 0.32520952820777893, "learning_rate": 8.647812943144524e-05, "loss": 0.0445, "step": 22220 }, { "grad_norm": 0.4288790225982666, "learning_rate": 8.646399092477351e-05, "loss": 0.0476, "step": 22230 }, { "grad_norm": 0.7607295513153076, "learning_rate": 8.644984618740301e-05, "loss": 0.0577, "step": 22240 }, { "grad_norm": 0.5306458473205566, "learning_rate": 8.643569522175073e-05, "loss": 0.0483, "step": 22250 }, { "grad_norm": 0.4816506803035736, "learning_rate": 8.642153803023463e-05, "loss": 0.0537, "step": 22260 }, { "grad_norm": 0.3226570785045624, "learning_rate": 8.640737461527383e-05, "loss": 0.0502, "step": 22270 }, { "grad_norm": 0.4744247496128082, "learning_rate": 8.639320497928845e-05, "loss": 0.0556, "step": 22280 }, { "grad_norm": 0.44228395819664, "learning_rate": 8.637902912469969e-05, "loss": 0.0484, "step": 22290 }, { "grad_norm": 0.47194135189056396, "learning_rate": 8.636484705392982e-05, "loss": 0.0477, "step": 22300 }, { "grad_norm": 0.47519201040267944, "learning_rate": 8.635065876940216e-05, "loss": 0.0494, "step": 22310 }, { "grad_norm": 0.483062207698822, "learning_rate": 8.633646427354112e-05, "loss": 0.0604, "step": 22320 }, { "grad_norm": 0.3433179557323456, "learning_rate": 8.632226356877213e-05, "loss": 0.0475, "step": 22330 }, { "grad_norm": 0.3718262314796448, "learning_rate": 8.630805665752173e-05, "loss": 0.0543, "step": 22340 }, { "grad_norm": 0.40039849281311035, "learning_rate": 8.629384354221748e-05, "loss": 0.0453, "step": 22350 }, { "grad_norm": 0.340813010931015, "learning_rate": 8.627962422528797e-05, "loss": 0.0408, "step": 22360 }, { "grad_norm": 0.3320048153400421, "learning_rate": 8.626539870916296e-05, "loss": 0.0594, "step": 22370 }, { "grad_norm": 0.48468974232673645, "learning_rate": 8.625116699627317e-05, "loss": 0.0527, "step": 22380 }, { "grad_norm": 0.35203322768211365, "learning_rate": 8.623692908905041e-05, "loss": 0.0405, "step": 22390 }, { "grad_norm": 0.6247178316116333, "learning_rate": 8.622268498992755e-05, "loss": 0.0628, "step": 22400 }, { "grad_norm": 0.4459642171859741, "learning_rate": 8.620843470133851e-05, "loss": 0.079, "step": 22410 }, { "grad_norm": 0.49852392077445984, "learning_rate": 8.619417822571829e-05, "loss": 0.0608, "step": 22420 }, { "grad_norm": 0.3869307041168213, "learning_rate": 8.617991556550292e-05, "loss": 0.0404, "step": 22430 }, { "grad_norm": 0.3526206314563751, "learning_rate": 8.616564672312952e-05, "loss": 0.0478, "step": 22440 }, { "grad_norm": 0.455959677696228, "learning_rate": 8.61513717010362e-05, "loss": 0.0586, "step": 22450 }, { "grad_norm": 0.45288988947868347, "learning_rate": 8.613709050166221e-05, "loss": 0.0566, "step": 22460 }, { "grad_norm": 0.5640770196914673, "learning_rate": 8.61228031274478e-05, "loss": 0.0524, "step": 22470 }, { "grad_norm": 0.4279681146144867, "learning_rate": 8.610850958083431e-05, "loss": 0.0493, "step": 22480 }, { "grad_norm": 0.560596227645874, "learning_rate": 8.609420986426409e-05, "loss": 0.0532, "step": 22490 }, { "grad_norm": 0.4606693983078003, "learning_rate": 8.60799039801806e-05, "loss": 0.0531, "step": 22500 }, { "grad_norm": 0.6375902891159058, "learning_rate": 8.606559193102828e-05, "loss": 0.0471, "step": 22510 }, { "grad_norm": 0.5994689464569092, "learning_rate": 8.605127371925273e-05, "loss": 0.0507, "step": 22520 }, { "grad_norm": 0.5448409914970398, "learning_rate": 8.603694934730047e-05, "loss": 0.0588, "step": 22530 }, { "grad_norm": 0.3382795453071594, "learning_rate": 8.602261881761919e-05, "loss": 0.0375, "step": 22540 }, { "grad_norm": 0.2883553206920624, "learning_rate": 8.600828213265759e-05, "loss": 0.0545, "step": 22550 }, { "grad_norm": 0.6256393194198608, "learning_rate": 8.599393929486539e-05, "loss": 0.0605, "step": 22560 }, { "grad_norm": 0.4675459563732147, "learning_rate": 8.59795903066934e-05, "loss": 0.0558, "step": 22570 }, { "grad_norm": 0.33796554803848267, "learning_rate": 8.596523517059347e-05, "loss": 0.0548, "step": 22580 }, { "grad_norm": 0.4892895817756653, "learning_rate": 8.59508738890185e-05, "loss": 0.048, "step": 22590 }, { "grad_norm": 0.42994576692581177, "learning_rate": 8.593650646442246e-05, "loss": 0.0479, "step": 22600 }, { "grad_norm": 0.46275219321250916, "learning_rate": 8.59221328992603e-05, "loss": 0.05, "step": 22610 }, { "grad_norm": 0.3954418897628784, "learning_rate": 8.590775319598813e-05, "loss": 0.0519, "step": 22620 }, { "grad_norm": 0.48400798439979553, "learning_rate": 8.589336735706301e-05, "loss": 0.0526, "step": 22630 }, { "grad_norm": 0.7124236226081848, "learning_rate": 8.587897538494307e-05, "loss": 0.0659, "step": 22640 }, { "grad_norm": 0.495524525642395, "learning_rate": 8.586457728208756e-05, "loss": 0.0573, "step": 22650 }, { "grad_norm": 0.48316246271133423, "learning_rate": 8.585017305095667e-05, "loss": 0.0546, "step": 22660 }, { "grad_norm": 0.4687352180480957, "learning_rate": 8.583576269401173e-05, "loss": 0.0496, "step": 22670 }, { "grad_norm": 0.5510452389717102, "learning_rate": 8.582134621371504e-05, "loss": 0.0502, "step": 22680 }, { "grad_norm": 0.3590925931930542, "learning_rate": 8.580692361253e-05, "loss": 0.0516, "step": 22690 }, { "grad_norm": 0.49225538969039917, "learning_rate": 8.579249489292104e-05, "loss": 0.0585, "step": 22700 }, { "grad_norm": 0.4374981224536896, "learning_rate": 8.577806005735363e-05, "loss": 0.0599, "step": 22710 }, { "grad_norm": 0.5747857689857483, "learning_rate": 8.576361910829429e-05, "loss": 0.0434, "step": 22720 }, { "grad_norm": 0.29110002517700195, "learning_rate": 8.574917204821057e-05, "loss": 0.0507, "step": 22730 }, { "grad_norm": 0.4106258749961853, "learning_rate": 8.57347188795711e-05, "loss": 0.0512, "step": 22740 }, { "grad_norm": 0.43361008167266846, "learning_rate": 8.572025960484551e-05, "loss": 0.062, "step": 22750 }, { "grad_norm": 0.5163834095001221, "learning_rate": 8.57057942265045e-05, "loss": 0.0629, "step": 22760 }, { "grad_norm": 0.34153658151626587, "learning_rate": 8.569132274701984e-05, "loss": 0.0529, "step": 22770 }, { "grad_norm": 0.5209330320358276, "learning_rate": 8.567684516886427e-05, "loss": 0.0469, "step": 22780 }, { "grad_norm": 0.664365291595459, "learning_rate": 8.56623614945116e-05, "loss": 0.055, "step": 22790 }, { "grad_norm": 0.5741286873817444, "learning_rate": 8.564787172643675e-05, "loss": 0.0498, "step": 22800 }, { "grad_norm": 0.4954053461551666, "learning_rate": 8.563337586711559e-05, "loss": 0.0579, "step": 22810 }, { "grad_norm": 0.37914296984672546, "learning_rate": 8.561887391902506e-05, "loss": 0.0604, "step": 22820 }, { "grad_norm": 0.4596537947654724, "learning_rate": 8.560436588464316e-05, "loss": 0.0493, "step": 22830 }, { "grad_norm": 0.5361856818199158, "learning_rate": 8.55898517664489e-05, "loss": 0.0591, "step": 22840 }, { "grad_norm": 0.5432749390602112, "learning_rate": 8.557533156692236e-05, "loss": 0.0693, "step": 22850 }, { "grad_norm": 0.31084713339805603, "learning_rate": 8.556080528854467e-05, "loss": 0.0483, "step": 22860 }, { "grad_norm": 0.3743155598640442, "learning_rate": 8.554627293379791e-05, "loss": 0.0498, "step": 22870 }, { "grad_norm": 0.40296313166618347, "learning_rate": 8.553173450516531e-05, "loss": 0.0541, "step": 22880 }, { "grad_norm": 0.34807705879211426, "learning_rate": 8.551719000513108e-05, "loss": 0.0458, "step": 22890 }, { "grad_norm": 0.5391793251037598, "learning_rate": 8.550263943618049e-05, "loss": 0.0654, "step": 22900 }, { "grad_norm": 0.46227332949638367, "learning_rate": 8.54880828007998e-05, "loss": 0.0506, "step": 22910 }, { "grad_norm": 0.49464938044548035, "learning_rate": 8.547352010147637e-05, "loss": 0.0617, "step": 22920 }, { "grad_norm": 0.4663451313972473, "learning_rate": 8.545895134069855e-05, "loss": 0.0476, "step": 22930 }, { "grad_norm": 0.580549418926239, "learning_rate": 8.544437652095576e-05, "loss": 0.0466, "step": 22940 }, { "grad_norm": 0.46395164728164673, "learning_rate": 8.542979564473843e-05, "loss": 0.066, "step": 22950 }, { "grad_norm": 0.48183175921440125, "learning_rate": 8.541520871453802e-05, "loss": 0.0459, "step": 22960 }, { "grad_norm": 0.5095179677009583, "learning_rate": 8.540061573284705e-05, "loss": 0.0492, "step": 22970 }, { "grad_norm": 0.35135531425476074, "learning_rate": 8.538601670215906e-05, "loss": 0.0551, "step": 22980 }, { "grad_norm": 0.5307966470718384, "learning_rate": 8.537141162496864e-05, "loss": 0.0581, "step": 22990 }, { "grad_norm": 0.4106384813785553, "learning_rate": 8.535680050377137e-05, "loss": 0.0448, "step": 23000 }, { "grad_norm": 0.35857251286506653, "learning_rate": 8.534218334106391e-05, "loss": 0.0429, "step": 23010 }, { "grad_norm": 0.39644157886505127, "learning_rate": 8.532756013934393e-05, "loss": 0.0474, "step": 23020 }, { "grad_norm": 0.4158598482608795, "learning_rate": 8.531293090111012e-05, "loss": 0.0475, "step": 23030 }, { "grad_norm": 0.4620705544948578, "learning_rate": 8.529829562886225e-05, "loss": 0.0433, "step": 23040 }, { "grad_norm": 0.5216135382652283, "learning_rate": 8.528365432510105e-05, "loss": 0.0574, "step": 23050 }, { "grad_norm": 0.3612344264984131, "learning_rate": 8.526900699232833e-05, "loss": 0.0476, "step": 23060 }, { "grad_norm": 0.4659462571144104, "learning_rate": 8.525435363304695e-05, "loss": 0.0488, "step": 23070 }, { "grad_norm": 0.4333738684654236, "learning_rate": 8.523969424976072e-05, "loss": 0.0459, "step": 23080 }, { "grad_norm": 0.4262881875038147, "learning_rate": 8.522502884497457e-05, "loss": 0.0468, "step": 23090 }, { "grad_norm": 0.5457329154014587, "learning_rate": 8.521035742119437e-05, "loss": 0.0607, "step": 23100 }, { "grad_norm": 0.3518655002117157, "learning_rate": 8.519567998092712e-05, "loss": 0.0553, "step": 23110 }, { "grad_norm": 0.36079010367393494, "learning_rate": 8.518099652668075e-05, "loss": 0.0516, "step": 23120 }, { "grad_norm": 0.48090752959251404, "learning_rate": 8.516630706096429e-05, "loss": 0.0524, "step": 23130 }, { "grad_norm": 0.38600054383277893, "learning_rate": 8.515161158628773e-05, "loss": 0.0423, "step": 23140 }, { "grad_norm": 0.4414597749710083, "learning_rate": 8.513691010516216e-05, "loss": 0.0694, "step": 23150 }, { "grad_norm": 0.4506916403770447, "learning_rate": 8.512220262009966e-05, "loss": 0.0421, "step": 23160 }, { "grad_norm": 0.3041972219944, "learning_rate": 8.510748913361332e-05, "loss": 0.0459, "step": 23170 }, { "grad_norm": 0.4122397005558014, "learning_rate": 8.509276964821726e-05, "loss": 0.04, "step": 23180 }, { "grad_norm": 0.536444365978241, "learning_rate": 8.507804416642669e-05, "loss": 0.048, "step": 23190 }, { "grad_norm": 0.48712196946144104, "learning_rate": 8.506331269075774e-05, "loss": 0.0495, "step": 23200 }, { "grad_norm": 0.3564129173755646, "learning_rate": 8.504857522372765e-05, "loss": 0.05, "step": 23210 }, { "grad_norm": 0.4789790213108063, "learning_rate": 8.503383176785461e-05, "loss": 0.0494, "step": 23220 }, { "grad_norm": 0.46215900778770447, "learning_rate": 8.501908232565792e-05, "loss": 0.0498, "step": 23230 }, { "grad_norm": 0.540328860282898, "learning_rate": 8.50043268996578e-05, "loss": 0.0543, "step": 23240 }, { "grad_norm": 0.33141645789146423, "learning_rate": 8.498956549237562e-05, "loss": 0.0483, "step": 23250 }, { "grad_norm": 0.41273871064186096, "learning_rate": 8.497479810633366e-05, "loss": 0.0453, "step": 23260 }, { "grad_norm": 0.558935821056366, "learning_rate": 8.496002474405525e-05, "loss": 0.0539, "step": 23270 }, { "grad_norm": 0.5964717268943787, "learning_rate": 8.494524540806478e-05, "loss": 0.0475, "step": 23280 }, { "grad_norm": 0.358440101146698, "learning_rate": 8.493046010088761e-05, "loss": 0.0546, "step": 23290 }, { "grad_norm": 0.5056034922599792, "learning_rate": 8.491566882505018e-05, "loss": 0.0489, "step": 23300 }, { "grad_norm": 0.36103934049606323, "learning_rate": 8.490087158307988e-05, "loss": 0.0579, "step": 23310 }, { "grad_norm": 0.7054301500320435, "learning_rate": 8.488606837750518e-05, "loss": 0.0569, "step": 23320 }, { "grad_norm": 0.410176545381546, "learning_rate": 8.487125921085552e-05, "loss": 0.0459, "step": 23330 }, { "grad_norm": 0.5531030297279358, "learning_rate": 8.485644408566141e-05, "loss": 0.0564, "step": 23340 }, { "grad_norm": 0.37952739000320435, "learning_rate": 8.484162300445431e-05, "loss": 0.0487, "step": 23350 }, { "grad_norm": 0.41995376348495483, "learning_rate": 8.482679596976676e-05, "loss": 0.0555, "step": 23360 }, { "grad_norm": 0.45300421118736267, "learning_rate": 8.48119629841323e-05, "loss": 0.0595, "step": 23370 }, { "grad_norm": 0.4196922481060028, "learning_rate": 8.479712405008547e-05, "loss": 0.0556, "step": 23380 }, { "grad_norm": 0.36663123965263367, "learning_rate": 8.478227917016184e-05, "loss": 0.0559, "step": 23390 }, { "grad_norm": 0.48244887590408325, "learning_rate": 8.476742834689801e-05, "loss": 0.0544, "step": 23400 }, { "grad_norm": 0.4099421799182892, "learning_rate": 8.475257158283157e-05, "loss": 0.0615, "step": 23410 }, { "grad_norm": 0.44680696725845337, "learning_rate": 8.473770888050112e-05, "loss": 0.0481, "step": 23420 }, { "grad_norm": 0.4064391255378723, "learning_rate": 8.47228402424463e-05, "loss": 0.0502, "step": 23430 }, { "grad_norm": 0.4354308843612671, "learning_rate": 8.470796567120775e-05, "loss": 0.0411, "step": 23440 }, { "grad_norm": 0.4861031770706177, "learning_rate": 8.469308516932714e-05, "loss": 0.0556, "step": 23450 }, { "grad_norm": 0.5815814137458801, "learning_rate": 8.467819873934714e-05, "loss": 0.0492, "step": 23460 }, { "grad_norm": 0.4036702811717987, "learning_rate": 8.466330638381143e-05, "loss": 0.0492, "step": 23470 }, { "grad_norm": 0.4382713735103607, "learning_rate": 8.464840810526469e-05, "loss": 0.0458, "step": 23480 }, { "grad_norm": 0.2949933707714081, "learning_rate": 8.463350390625264e-05, "loss": 0.0495, "step": 23490 }, { "grad_norm": 0.5588585734367371, "learning_rate": 8.4618593789322e-05, "loss": 0.0627, "step": 23500 }, { "grad_norm": 0.41471731662750244, "learning_rate": 8.46036777570205e-05, "loss": 0.0524, "step": 23510 }, { "grad_norm": 0.4987942576408386, "learning_rate": 8.458875581189688e-05, "loss": 0.0702, "step": 23520 }, { "grad_norm": 0.3322915732860565, "learning_rate": 8.457382795650092e-05, "loss": 0.04, "step": 23530 }, { "grad_norm": 0.37850338220596313, "learning_rate": 8.455889419338335e-05, "loss": 0.059, "step": 23540 }, { "grad_norm": 0.40220367908477783, "learning_rate": 8.454395452509593e-05, "loss": 0.0459, "step": 23550 }, { "grad_norm": 0.5295088887214661, "learning_rate": 8.452900895419146e-05, "loss": 0.0493, "step": 23560 }, { "grad_norm": 0.33309680223464966, "learning_rate": 8.451405748322376e-05, "loss": 0.0435, "step": 23570 }, { "grad_norm": 0.3055424690246582, "learning_rate": 8.449910011474759e-05, "loss": 0.0458, "step": 23580 }, { "grad_norm": 0.32977503538131714, "learning_rate": 8.448413685131876e-05, "loss": 0.0548, "step": 23590 }, { "grad_norm": 0.4284144639968872, "learning_rate": 8.446916769549407e-05, "loss": 0.0443, "step": 23600 }, { "grad_norm": 0.31887951493263245, "learning_rate": 8.445419264983136e-05, "loss": 0.0496, "step": 23610 }, { "grad_norm": 0.34254926443099976, "learning_rate": 8.443921171688947e-05, "loss": 0.0461, "step": 23620 }, { "grad_norm": 0.36541661620140076, "learning_rate": 8.442422489922819e-05, "loss": 0.0404, "step": 23630 }, { "grad_norm": 0.6077966094017029, "learning_rate": 8.440923219940838e-05, "loss": 0.0576, "step": 23640 }, { "grad_norm": 0.38305121660232544, "learning_rate": 8.439423361999189e-05, "loss": 0.0458, "step": 23650 }, { "grad_norm": 0.38476434350013733, "learning_rate": 8.437922916354155e-05, "loss": 0.0475, "step": 23660 }, { "grad_norm": 0.22639119625091553, "learning_rate": 8.436421883262123e-05, "loss": 0.0477, "step": 23670 }, { "grad_norm": 0.40658390522003174, "learning_rate": 8.434920262979577e-05, "loss": 0.0601, "step": 23680 }, { "grad_norm": 0.3563940227031708, "learning_rate": 8.433418055763104e-05, "loss": 0.0496, "step": 23690 }, { "grad_norm": 0.4303280711174011, "learning_rate": 8.431915261869389e-05, "loss": 0.0547, "step": 23700 }, { "grad_norm": 0.42283326387405396, "learning_rate": 8.43041188155522e-05, "loss": 0.0459, "step": 23710 }, { "grad_norm": 0.36225050687789917, "learning_rate": 8.428907915077481e-05, "loss": 0.0415, "step": 23720 }, { "grad_norm": 0.5139141082763672, "learning_rate": 8.42740336269316e-05, "loss": 0.0568, "step": 23730 }, { "grad_norm": 0.42160648107528687, "learning_rate": 8.425898224659345e-05, "loss": 0.0456, "step": 23740 }, { "grad_norm": 0.2502971887588501, "learning_rate": 8.42439250123322e-05, "loss": 0.0493, "step": 23750 }, { "grad_norm": 0.33914682269096375, "learning_rate": 8.422886192672076e-05, "loss": 0.0501, "step": 23760 }, { "grad_norm": 0.4706593453884125, "learning_rate": 8.421379299233297e-05, "loss": 0.042, "step": 23770 }, { "grad_norm": 0.470940500497818, "learning_rate": 8.419871821174371e-05, "loss": 0.0462, "step": 23780 }, { "grad_norm": 0.3453812599182129, "learning_rate": 8.418363758752884e-05, "loss": 0.0458, "step": 23790 }, { "grad_norm": 0.2176009565591812, "learning_rate": 8.416855112226523e-05, "loss": 0.0484, "step": 23800 }, { "grad_norm": 0.4840913712978363, "learning_rate": 8.415345881853075e-05, "loss": 0.0409, "step": 23810 }, { "grad_norm": 0.39830946922302246, "learning_rate": 8.413836067890426e-05, "loss": 0.0469, "step": 23820 }, { "grad_norm": 0.348396360874176, "learning_rate": 8.41232567059656e-05, "loss": 0.0468, "step": 23830 }, { "grad_norm": 0.3854120969772339, "learning_rate": 8.410814690229565e-05, "loss": 0.047, "step": 23840 }, { "grad_norm": 0.3633671998977661, "learning_rate": 8.409303127047626e-05, "loss": 0.0462, "step": 23850 }, { "grad_norm": 0.35305607318878174, "learning_rate": 8.407790981309028e-05, "loss": 0.0506, "step": 23860 }, { "grad_norm": 0.4514060914516449, "learning_rate": 8.406278253272153e-05, "loss": 0.0634, "step": 23870 }, { "grad_norm": 0.40742847323417664, "learning_rate": 8.404764943195487e-05, "loss": 0.045, "step": 23880 }, { "grad_norm": 0.3671453893184662, "learning_rate": 8.403251051337613e-05, "loss": 0.0535, "step": 23890 }, { "grad_norm": 0.475466787815094, "learning_rate": 8.401736577957214e-05, "loss": 0.0422, "step": 23900 }, { "grad_norm": 0.33608943223953247, "learning_rate": 8.40022152331307e-05, "loss": 0.0396, "step": 23910 }, { "grad_norm": 0.35720738768577576, "learning_rate": 8.398705887664064e-05, "loss": 0.0545, "step": 23920 }, { "grad_norm": 0.31931811571121216, "learning_rate": 8.397189671269177e-05, "loss": 0.0444, "step": 23930 }, { "grad_norm": 0.4043586552143097, "learning_rate": 8.395672874387488e-05, "loss": 0.037, "step": 23940 }, { "grad_norm": 0.3734058737754822, "learning_rate": 8.394155497278177e-05, "loss": 0.0365, "step": 23950 }, { "grad_norm": 0.49873828887939453, "learning_rate": 8.392637540200523e-05, "loss": 0.0709, "step": 23960 }, { "grad_norm": 0.4295731782913208, "learning_rate": 8.391119003413902e-05, "loss": 0.0457, "step": 23970 }, { "grad_norm": 0.4566115438938141, "learning_rate": 8.38959988717779e-05, "loss": 0.0434, "step": 23980 }, { "grad_norm": 0.5295789837837219, "learning_rate": 8.388080191751764e-05, "loss": 0.053, "step": 23990 }, { "grad_norm": 0.5269126296043396, "learning_rate": 8.386559917395496e-05, "loss": 0.0499, "step": 24000 }, { "grad_norm": 0.3259131610393524, "learning_rate": 8.385039064368761e-05, "loss": 0.0442, "step": 24010 }, { "grad_norm": 0.37729114294052124, "learning_rate": 8.383517632931431e-05, "loss": 0.0484, "step": 24020 }, { "grad_norm": 0.5026257038116455, "learning_rate": 8.381995623343477e-05, "loss": 0.0501, "step": 24030 }, { "grad_norm": 0.4094681739807129, "learning_rate": 8.380473035864968e-05, "loss": 0.0473, "step": 24040 }, { "grad_norm": 0.2939695715904236, "learning_rate": 8.378949870756076e-05, "loss": 0.0477, "step": 24050 }, { "grad_norm": 0.30134087800979614, "learning_rate": 8.377426128277063e-05, "loss": 0.0519, "step": 24060 }, { "grad_norm": 0.48567888140678406, "learning_rate": 8.375901808688298e-05, "loss": 0.0398, "step": 24070 }, { "grad_norm": 0.5182593464851379, "learning_rate": 8.374376912250246e-05, "loss": 0.0571, "step": 24080 }, { "grad_norm": 0.41566208004951477, "learning_rate": 8.372851439223468e-05, "loss": 0.0399, "step": 24090 }, { "grad_norm": 0.540439248085022, "learning_rate": 8.371325389868627e-05, "loss": 0.0471, "step": 24100 }, { "grad_norm": 0.4422050416469574, "learning_rate": 8.369798764446482e-05, "loss": 0.0442, "step": 24110 }, { "grad_norm": 0.3885682225227356, "learning_rate": 8.368271563217893e-05, "loss": 0.0429, "step": 24120 }, { "grad_norm": 0.45577317476272583, "learning_rate": 8.366743786443817e-05, "loss": 0.0557, "step": 24130 }, { "grad_norm": 0.6896984577178955, "learning_rate": 8.365215434385309e-05, "loss": 0.0546, "step": 24140 }, { "grad_norm": 0.5462573170661926, "learning_rate": 8.36368650730352e-05, "loss": 0.0488, "step": 24150 }, { "grad_norm": 0.39378973841667175, "learning_rate": 8.362157005459705e-05, "loss": 0.0532, "step": 24160 }, { "grad_norm": 0.4427284002304077, "learning_rate": 8.360626929115213e-05, "loss": 0.0493, "step": 24170 }, { "grad_norm": 0.37257203459739685, "learning_rate": 8.359096278531492e-05, "loss": 0.0505, "step": 24180 }, { "grad_norm": 0.627628743648529, "learning_rate": 8.357565053970088e-05, "loss": 0.0426, "step": 24190 }, { "grad_norm": 0.28777486085891724, "learning_rate": 8.356033255692647e-05, "loss": 0.044, "step": 24200 }, { "grad_norm": 0.5800065398216248, "learning_rate": 8.354500883960911e-05, "loss": 0.0457, "step": 24210 }, { "grad_norm": 0.3962002694606781, "learning_rate": 8.352967939036717e-05, "loss": 0.0536, "step": 24220 }, { "grad_norm": 0.41006967425346375, "learning_rate": 8.35143442118201e-05, "loss": 0.0528, "step": 24230 }, { "grad_norm": 0.6366256475448608, "learning_rate": 8.349900330658819e-05, "loss": 0.0468, "step": 24240 }, { "grad_norm": 0.3938700258731842, "learning_rate": 8.348365667729284e-05, "loss": 0.0489, "step": 24250 }, { "grad_norm": 0.5254015922546387, "learning_rate": 8.346830432655633e-05, "loss": 0.047, "step": 24260 }, { "grad_norm": 0.46156588196754456, "learning_rate": 8.345294625700195e-05, "loss": 0.0496, "step": 24270 }, { "grad_norm": 0.48163121938705444, "learning_rate": 8.343758247125402e-05, "loss": 0.0483, "step": 24280 }, { "grad_norm": 0.5404894351959229, "learning_rate": 8.342221297193776e-05, "loss": 0.0618, "step": 24290 }, { "grad_norm": 0.3751140534877777, "learning_rate": 8.34068377616794e-05, "loss": 0.0568, "step": 24300 }, { "grad_norm": 0.5133981108665466, "learning_rate": 8.339145684310615e-05, "loss": 0.0487, "step": 24310 }, { "grad_norm": 0.47378435730934143, "learning_rate": 8.337607021884618e-05, "loss": 0.0491, "step": 24320 }, { "grad_norm": 0.3508508801460266, "learning_rate": 8.336067789152867e-05, "loss": 0.0396, "step": 24330 }, { "grad_norm": 0.3530959486961365, "learning_rate": 8.334527986378369e-05, "loss": 0.046, "step": 24340 }, { "grad_norm": 0.5199484825134277, "learning_rate": 8.332987613824239e-05, "loss": 0.0526, "step": 24350 }, { "grad_norm": 0.4348183870315552, "learning_rate": 8.331446671753685e-05, "loss": 0.0446, "step": 24360 }, { "grad_norm": 0.34695136547088623, "learning_rate": 8.329905160430007e-05, "loss": 0.0451, "step": 24370 }, { "grad_norm": 0.39719101786613464, "learning_rate": 8.328363080116611e-05, "loss": 0.0408, "step": 24380 }, { "grad_norm": 0.3913695514202118, "learning_rate": 8.326820431076997e-05, "loss": 0.0387, "step": 24390 }, { "grad_norm": 0.512397289276123, "learning_rate": 8.325277213574759e-05, "loss": 0.045, "step": 24400 }, { "grad_norm": 0.46561938524246216, "learning_rate": 8.32373342787359e-05, "loss": 0.0448, "step": 24410 }, { "grad_norm": 0.42540279030799866, "learning_rate": 8.322189074237285e-05, "loss": 0.0564, "step": 24420 }, { "grad_norm": 0.43543514609336853, "learning_rate": 8.32064415292973e-05, "loss": 0.0468, "step": 24430 }, { "grad_norm": 0.38446030020713806, "learning_rate": 8.319098664214907e-05, "loss": 0.0337, "step": 24440 }, { "grad_norm": 0.32658636569976807, "learning_rate": 8.3175526083569e-05, "loss": 0.0501, "step": 24450 }, { "grad_norm": 0.36135804653167725, "learning_rate": 8.316005985619889e-05, "loss": 0.047, "step": 24460 }, { "grad_norm": 0.3456731140613556, "learning_rate": 8.314458796268147e-05, "loss": 0.0476, "step": 24470 }, { "grad_norm": 0.4068079888820648, "learning_rate": 8.312911040566047e-05, "loss": 0.0449, "step": 24480 }, { "grad_norm": 0.430989146232605, "learning_rate": 8.31136271877806e-05, "loss": 0.0515, "step": 24490 }, { "grad_norm": 0.3801964223384857, "learning_rate": 8.309813831168748e-05, "loss": 0.0537, "step": 24500 }, { "grad_norm": 0.37932440638542175, "learning_rate": 8.308264378002777e-05, "loss": 0.0497, "step": 24510 }, { "grad_norm": 0.3852376937866211, "learning_rate": 8.306714359544906e-05, "loss": 0.0523, "step": 24520 }, { "grad_norm": 0.5260618925094604, "learning_rate": 8.30516377605999e-05, "loss": 0.0427, "step": 24530 }, { "grad_norm": 0.4279966652393341, "learning_rate": 8.30361262781298e-05, "loss": 0.05, "step": 24540 }, { "grad_norm": 0.3698895573616028, "learning_rate": 8.302060915068924e-05, "loss": 0.0506, "step": 24550 }, { "grad_norm": 0.5673385858535767, "learning_rate": 8.300508638092972e-05, "loss": 0.0497, "step": 24560 }, { "grad_norm": 0.3773633539676666, "learning_rate": 8.298955797150361e-05, "loss": 0.0509, "step": 24570 }, { "grad_norm": 0.5438192486763, "learning_rate": 8.297402392506433e-05, "loss": 0.0563, "step": 24580 }, { "grad_norm": 0.3852373957633972, "learning_rate": 8.295848424426617e-05, "loss": 0.0459, "step": 24590 }, { "grad_norm": 0.38028600811958313, "learning_rate": 8.29429389317645e-05, "loss": 0.0486, "step": 24600 }, { "grad_norm": 0.3514958322048187, "learning_rate": 8.292738799021556e-05, "loss": 0.0517, "step": 24610 }, { "grad_norm": 0.2954168915748596, "learning_rate": 8.291183142227656e-05, "loss": 0.0428, "step": 24620 }, { "grad_norm": 0.30488717555999756, "learning_rate": 8.289626923060572e-05, "loss": 0.0573, "step": 24630 }, { "grad_norm": 1.06304132938385, "learning_rate": 8.288070141786218e-05, "loss": 0.0504, "step": 24640 }, { "grad_norm": 0.34164974093437195, "learning_rate": 8.286512798670605e-05, "loss": 0.048, "step": 24650 }, { "grad_norm": 0.4985242784023285, "learning_rate": 8.284954893979842e-05, "loss": 0.0583, "step": 24660 }, { "grad_norm": 0.39339524507522583, "learning_rate": 8.283396427980131e-05, "loss": 0.0587, "step": 24670 }, { "grad_norm": 0.5141417384147644, "learning_rate": 8.281837400937771e-05, "loss": 0.0495, "step": 24680 }, { "grad_norm": 0.6060546040534973, "learning_rate": 8.28027781311916e-05, "loss": 0.0542, "step": 24690 }, { "grad_norm": 0.46980521082878113, "learning_rate": 8.278717664790785e-05, "loss": 0.0505, "step": 24700 }, { "grad_norm": 0.4949112832546234, "learning_rate": 8.277156956219234e-05, "loss": 0.0452, "step": 24710 }, { "grad_norm": 0.36799126863479614, "learning_rate": 8.275595687671189e-05, "loss": 0.0435, "step": 24720 }, { "grad_norm": 0.5219479203224182, "learning_rate": 8.27403385941343e-05, "loss": 0.0572, "step": 24730 }, { "grad_norm": 0.4548843801021576, "learning_rate": 8.272471471712828e-05, "loss": 0.0506, "step": 24740 }, { "grad_norm": 0.39961427450180054, "learning_rate": 8.270908524836355e-05, "loss": 0.0552, "step": 24750 }, { "grad_norm": 0.43050920963287354, "learning_rate": 8.269345019051074e-05, "loss": 0.0601, "step": 24760 }, { "grad_norm": 0.3754954934120178, "learning_rate": 8.267780954624147e-05, "loss": 0.0518, "step": 24770 }, { "grad_norm": 0.47540751099586487, "learning_rate": 8.266216331822827e-05, "loss": 0.0523, "step": 24780 }, { "grad_norm": 0.5640543699264526, "learning_rate": 8.264651150914469e-05, "loss": 0.0529, "step": 24790 }, { "grad_norm": 0.4872850477695465, "learning_rate": 8.263085412166517e-05, "loss": 0.0455, "step": 24800 }, { "grad_norm": 0.620668351650238, "learning_rate": 8.261519115846514e-05, "loss": 0.0525, "step": 24810 }, { "grad_norm": 0.31495505571365356, "learning_rate": 8.259952262222096e-05, "loss": 0.0423, "step": 24820 }, { "grad_norm": 0.42957234382629395, "learning_rate": 8.258384851560997e-05, "loss": 0.0507, "step": 24830 }, { "grad_norm": 0.5037599205970764, "learning_rate": 8.256816884131044e-05, "loss": 0.0542, "step": 24840 }, { "grad_norm": 0.43055859208106995, "learning_rate": 8.255248360200159e-05, "loss": 0.0497, "step": 24850 }, { "grad_norm": 0.6806164383888245, "learning_rate": 8.253679280036359e-05, "loss": 0.0507, "step": 24860 }, { "grad_norm": 0.4897133409976959, "learning_rate": 8.252109643907762e-05, "loss": 0.0615, "step": 24870 }, { "grad_norm": 0.6419857740402222, "learning_rate": 8.250539452082569e-05, "loss": 0.0489, "step": 24880 }, { "grad_norm": 0.5865186452865601, "learning_rate": 8.248968704829087e-05, "loss": 0.0517, "step": 24890 }, { "grad_norm": 0.5037316679954529, "learning_rate": 8.247397402415714e-05, "loss": 0.0554, "step": 24900 }, { "grad_norm": 0.45652955770492554, "learning_rate": 8.24582554511094e-05, "loss": 0.0452, "step": 24910 }, { "grad_norm": 0.4009056091308594, "learning_rate": 8.244253133183355e-05, "loss": 0.0439, "step": 24920 }, { "grad_norm": 0.5714954137802124, "learning_rate": 8.24268016690164e-05, "loss": 0.0445, "step": 24930 }, { "grad_norm": 0.39059123396873474, "learning_rate": 8.241106646534571e-05, "loss": 0.0484, "step": 24940 }, { "grad_norm": 0.4554324150085449, "learning_rate": 8.23953257235102e-05, "loss": 0.0525, "step": 24950 }, { "grad_norm": 0.44571220874786377, "learning_rate": 8.237957944619956e-05, "loss": 0.0433, "step": 24960 }, { "grad_norm": 0.25087082386016846, "learning_rate": 8.236382763610437e-05, "loss": 0.0437, "step": 24970 }, { "grad_norm": 0.49358901381492615, "learning_rate": 8.234807029591619e-05, "loss": 0.0435, "step": 24980 }, { "grad_norm": 0.44058021903038025, "learning_rate": 8.233230742832752e-05, "loss": 0.0582, "step": 24990 }, { "grad_norm": 0.5569478273391724, "learning_rate": 8.231653903603178e-05, "loss": 0.0459, "step": 25000 }, { "grad_norm": 0.2827131450176239, "learning_rate": 8.23007651217234e-05, "loss": 0.0544, "step": 25010 }, { "grad_norm": 0.4651375114917755, "learning_rate": 8.228498568809769e-05, "loss": 0.0453, "step": 25020 }, { "grad_norm": 0.6681604981422424, "learning_rate": 8.22692007378509e-05, "loss": 0.0582, "step": 25030 }, { "grad_norm": 0.4850638806819916, "learning_rate": 8.225341027368028e-05, "loss": 0.0444, "step": 25040 }, { "grad_norm": 0.3866121470928192, "learning_rate": 8.223761429828399e-05, "loss": 0.0486, "step": 25050 }, { "grad_norm": 0.45750194787979126, "learning_rate": 8.22218128143611e-05, "loss": 0.0526, "step": 25060 }, { "grad_norm": 0.42381367087364197, "learning_rate": 8.220600582461166e-05, "loss": 0.0423, "step": 25070 }, { "grad_norm": 0.40443527698516846, "learning_rate": 8.219019333173668e-05, "loss": 0.0432, "step": 25080 }, { "grad_norm": 0.5716631412506104, "learning_rate": 8.217437533843805e-05, "loss": 0.0503, "step": 25090 }, { "grad_norm": 0.3301677703857422, "learning_rate": 8.215855184741867e-05, "loss": 0.0429, "step": 25100 }, { "grad_norm": 0.4313983619213104, "learning_rate": 8.21427228613823e-05, "loss": 0.048, "step": 25110 }, { "grad_norm": 0.38469398021698, "learning_rate": 8.21268883830337e-05, "loss": 0.0484, "step": 25120 }, { "grad_norm": 0.44785594940185547, "learning_rate": 8.211104841507855e-05, "loss": 0.058, "step": 25130 }, { "grad_norm": 0.34178683161735535, "learning_rate": 8.209520296022346e-05, "loss": 0.0505, "step": 25140 }, { "grad_norm": 0.360141396522522, "learning_rate": 8.207935202117599e-05, "loss": 0.0562, "step": 25150 }, { "grad_norm": 0.5647066235542297, "learning_rate": 8.206349560064463e-05, "loss": 0.0494, "step": 25160 }, { "grad_norm": 0.4054309129714966, "learning_rate": 8.204763370133881e-05, "loss": 0.0397, "step": 25170 }, { "grad_norm": 0.4335468113422394, "learning_rate": 8.203176632596892e-05, "loss": 0.046, "step": 25180 }, { "grad_norm": 0.3878146708011627, "learning_rate": 8.20158934772462e-05, "loss": 0.049, "step": 25190 }, { "grad_norm": 0.3845997452735901, "learning_rate": 8.200001515788294e-05, "loss": 0.0386, "step": 25200 }, { "grad_norm": 0.5029948949813843, "learning_rate": 8.198413137059228e-05, "loss": 0.0453, "step": 25210 }, { "grad_norm": 0.3875768482685089, "learning_rate": 8.196824211808835e-05, "loss": 0.0498, "step": 25220 }, { "grad_norm": 0.502512514591217, "learning_rate": 8.195234740308617e-05, "loss": 0.0529, "step": 25230 }, { "grad_norm": 0.4401469826698303, "learning_rate": 8.193644722830171e-05, "loss": 0.0395, "step": 25240 }, { "grad_norm": 0.3817487955093384, "learning_rate": 8.19205415964519e-05, "loss": 0.0405, "step": 25250 }, { "grad_norm": 0.5988786816596985, "learning_rate": 8.190463051025456e-05, "loss": 0.0406, "step": 25260 }, { "grad_norm": 0.5604453682899475, "learning_rate": 8.188871397242843e-05, "loss": 0.0489, "step": 25270 }, { "grad_norm": 0.5411672592163086, "learning_rate": 8.187279198569326e-05, "loss": 0.0464, "step": 25280 }, { "grad_norm": 0.45368534326553345, "learning_rate": 8.185686455276966e-05, "loss": 0.0427, "step": 25290 }, { "grad_norm": 0.3871101438999176, "learning_rate": 8.184093167637921e-05, "loss": 0.0478, "step": 25300 }, { "grad_norm": 0.43814054131507874, "learning_rate": 8.182499335924437e-05, "loss": 0.0478, "step": 25310 }, { "grad_norm": 0.4196620285511017, "learning_rate": 8.18090496040886e-05, "loss": 0.0375, "step": 25320 }, { "grad_norm": 0.40837162733078003, "learning_rate": 8.179310041363621e-05, "loss": 0.0686, "step": 25330 }, { "grad_norm": 0.39051440358161926, "learning_rate": 8.17771457906125e-05, "loss": 0.0419, "step": 25340 }, { "grad_norm": 0.44718071818351746, "learning_rate": 8.176118573774371e-05, "loss": 0.0451, "step": 25350 }, { "grad_norm": 0.8493516445159912, "learning_rate": 8.174522025775692e-05, "loss": 0.0461, "step": 25360 }, { "grad_norm": 0.5139433741569519, "learning_rate": 8.172924935338022e-05, "loss": 0.0545, "step": 25370 }, { "grad_norm": 0.5038511753082275, "learning_rate": 8.171327302734262e-05, "loss": 0.0397, "step": 25380 }, { "grad_norm": 0.3681439757347107, "learning_rate": 8.169729128237401e-05, "loss": 0.0486, "step": 25390 }, { "grad_norm": 0.3447295129299164, "learning_rate": 8.168130412120525e-05, "loss": 0.048, "step": 25400 }, { "grad_norm": 0.3705219030380249, "learning_rate": 8.16653115465681e-05, "loss": 0.0432, "step": 25410 }, { "grad_norm": 0.5379815101623535, "learning_rate": 8.164931356119526e-05, "loss": 0.0418, "step": 25420 }, { "grad_norm": 0.6502675414085388, "learning_rate": 8.163331016782032e-05, "loss": 0.0575, "step": 25430 }, { "grad_norm": 0.3978792726993561, "learning_rate": 8.161730136917785e-05, "loss": 0.0489, "step": 25440 }, { "grad_norm": 0.35746994614601135, "learning_rate": 8.160128716800333e-05, "loss": 0.0558, "step": 25450 }, { "grad_norm": 0.3354397714138031, "learning_rate": 8.158526756703313e-05, "loss": 0.0338, "step": 25460 }, { "grad_norm": 0.35958346724510193, "learning_rate": 8.156924256900455e-05, "loss": 0.053, "step": 25470 }, { "grad_norm": 0.27091360092163086, "learning_rate": 8.155321217665584e-05, "loss": 0.0447, "step": 25480 }, { "grad_norm": 0.42457887530326843, "learning_rate": 8.153717639272614e-05, "loss": 0.0425, "step": 25490 }, { "grad_norm": 0.37470540404319763, "learning_rate": 8.152113521995555e-05, "loss": 0.058, "step": 25500 }, { "grad_norm": 0.4143825173377991, "learning_rate": 8.150508866108505e-05, "loss": 0.0483, "step": 25510 }, { "grad_norm": 0.4094310998916626, "learning_rate": 8.148903671885657e-05, "loss": 0.0422, "step": 25520 }, { "grad_norm": 0.5108510255813599, "learning_rate": 8.147297939601292e-05, "loss": 0.0499, "step": 25530 }, { "grad_norm": 0.6066151261329651, "learning_rate": 8.145691669529792e-05, "loss": 0.0476, "step": 25540 }, { "grad_norm": 0.38957223296165466, "learning_rate": 8.144084861945618e-05, "loss": 0.0497, "step": 25550 }, { "grad_norm": 0.4676397740840912, "learning_rate": 8.142477517123333e-05, "loss": 0.052, "step": 25560 }, { "grad_norm": 0.5551285147666931, "learning_rate": 8.140869635337586e-05, "loss": 0.0486, "step": 25570 }, { "grad_norm": 0.43246838450431824, "learning_rate": 8.139261216863123e-05, "loss": 0.0444, "step": 25580 }, { "grad_norm": 0.48109981417655945, "learning_rate": 8.137652261974776e-05, "loss": 0.0507, "step": 25590 }, { "grad_norm": 0.41367676854133606, "learning_rate": 8.136042770947472e-05, "loss": 0.048, "step": 25600 }, { "grad_norm": 0.28403791785240173, "learning_rate": 8.134432744056228e-05, "loss": 0.0458, "step": 25610 }, { "grad_norm": 0.40481242537498474, "learning_rate": 8.132822181576158e-05, "loss": 0.0448, "step": 25620 }, { "grad_norm": 0.4369506239891052, "learning_rate": 8.131211083782459e-05, "loss": 0.0557, "step": 25630 }, { "grad_norm": 0.4882025122642517, "learning_rate": 8.129599450950424e-05, "loss": 0.0381, "step": 25640 }, { "grad_norm": 0.3580963909626007, "learning_rate": 8.127987283355438e-05, "loss": 0.0478, "step": 25650 }, { "grad_norm": 0.4470657706260681, "learning_rate": 8.126374581272976e-05, "loss": 0.0337, "step": 25660 }, { "grad_norm": 0.459058552980423, "learning_rate": 8.124761344978605e-05, "loss": 0.0635, "step": 25670 }, { "grad_norm": 0.377674400806427, "learning_rate": 8.12314757474798e-05, "loss": 0.0394, "step": 25680 }, { "grad_norm": 0.33713680505752563, "learning_rate": 8.121533270856856e-05, "loss": 0.0374, "step": 25690 }, { "grad_norm": 0.3725307285785675, "learning_rate": 8.119918433581069e-05, "loss": 0.0445, "step": 25700 }, { "grad_norm": 0.4173773229122162, "learning_rate": 8.118303063196551e-05, "loss": 0.0438, "step": 25710 }, { "grad_norm": 0.6104341745376587, "learning_rate": 8.116687159979326e-05, "loss": 0.0423, "step": 25720 }, { "grad_norm": 0.41081270575523376, "learning_rate": 8.115070724205508e-05, "loss": 0.0471, "step": 25730 }, { "grad_norm": 0.4820351004600525, "learning_rate": 8.113453756151296e-05, "loss": 0.043, "step": 25740 }, { "grad_norm": 0.4038068652153015, "learning_rate": 8.111836256092995e-05, "loss": 0.0479, "step": 25750 }, { "grad_norm": 0.5856612920761108, "learning_rate": 8.110218224306985e-05, "loss": 0.0459, "step": 25760 }, { "grad_norm": 0.42268696427345276, "learning_rate": 8.108599661069745e-05, "loss": 0.0406, "step": 25770 }, { "grad_norm": 0.38916847109794617, "learning_rate": 8.106980566657845e-05, "loss": 0.0408, "step": 25780 }, { "grad_norm": 0.3926314115524292, "learning_rate": 8.10536094134794e-05, "loss": 0.038, "step": 25790 }, { "grad_norm": 0.5435163378715515, "learning_rate": 8.103740785416783e-05, "loss": 0.05, "step": 25800 }, { "grad_norm": 0.5051057934761047, "learning_rate": 8.102120099141212e-05, "loss": 0.0408, "step": 25810 }, { "grad_norm": 0.423565149307251, "learning_rate": 8.100498882798163e-05, "loss": 0.0406, "step": 25820 }, { "grad_norm": 0.4550929665565491, "learning_rate": 8.09887713666465e-05, "loss": 0.0438, "step": 25830 }, { "grad_norm": 0.39598438143730164, "learning_rate": 8.09725486101779e-05, "loss": 0.0574, "step": 25840 }, { "grad_norm": 0.7139393091201782, "learning_rate": 8.095632056134784e-05, "loss": 0.0473, "step": 25850 }, { "grad_norm": 0.5383309125900269, "learning_rate": 8.094008722292925e-05, "loss": 0.044, "step": 25860 }, { "grad_norm": 0.4167187511920929, "learning_rate": 8.092384859769598e-05, "loss": 0.0479, "step": 25870 }, { "grad_norm": 0.45420074462890625, "learning_rate": 8.090760468842275e-05, "loss": 0.0655, "step": 25880 }, { "grad_norm": 0.28888052701950073, "learning_rate": 8.089135549788521e-05, "loss": 0.0344, "step": 25890 }, { "grad_norm": 0.3517484664916992, "learning_rate": 8.087510102885987e-05, "loss": 0.0382, "step": 25900 }, { "grad_norm": 0.4310140311717987, "learning_rate": 8.085884128412422e-05, "loss": 0.0449, "step": 25910 }, { "grad_norm": 0.39146503806114197, "learning_rate": 8.084257626645659e-05, "loss": 0.0409, "step": 25920 }, { "grad_norm": 0.3644979000091553, "learning_rate": 8.08263059786362e-05, "loss": 0.0395, "step": 25930 }, { "grad_norm": 0.4251171350479126, "learning_rate": 8.081003042344325e-05, "loss": 0.0375, "step": 25940 }, { "grad_norm": 0.4605148732662201, "learning_rate": 8.079374960365872e-05, "loss": 0.053, "step": 25950 }, { "grad_norm": 0.4424394369125366, "learning_rate": 8.077746352206463e-05, "loss": 0.053, "step": 25960 }, { "grad_norm": 0.536100447177887, "learning_rate": 8.076117218144377e-05, "loss": 0.0448, "step": 25970 }, { "grad_norm": 0.5396823883056641, "learning_rate": 8.074487558457991e-05, "loss": 0.0512, "step": 25980 }, { "grad_norm": 0.45761817693710327, "learning_rate": 8.072857373425768e-05, "loss": 0.0534, "step": 25990 }, { "grad_norm": 0.43525195121765137, "learning_rate": 8.071226663326264e-05, "loss": 0.0425, "step": 26000 }, { "grad_norm": 0.4290257692337036, "learning_rate": 8.069595428438121e-05, "loss": 0.0583, "step": 26010 }, { "grad_norm": 0.40310433506965637, "learning_rate": 8.067963669040072e-05, "loss": 0.0419, "step": 26020 }, { "grad_norm": 0.46301379799842834, "learning_rate": 8.066331385410942e-05, "loss": 0.042, "step": 26030 }, { "grad_norm": 0.42742404341697693, "learning_rate": 8.064698577829641e-05, "loss": 0.0561, "step": 26040 }, { "grad_norm": 0.4036644399166107, "learning_rate": 8.063065246575175e-05, "loss": 0.0462, "step": 26050 }, { "grad_norm": 0.3788716495037079, "learning_rate": 8.061431391926631e-05, "loss": 0.037, "step": 26060 }, { "grad_norm": 0.3608112633228302, "learning_rate": 8.059797014163195e-05, "loss": 0.0509, "step": 26070 }, { "grad_norm": 0.4825901687145233, "learning_rate": 8.058162113564133e-05, "loss": 0.0557, "step": 26080 }, { "grad_norm": 0.3902919292449951, "learning_rate": 8.056526690408806e-05, "loss": 0.0501, "step": 26090 }, { "grad_norm": 0.5085869431495667, "learning_rate": 8.054890744976666e-05, "loss": 0.0492, "step": 26100 }, { "grad_norm": 0.363294780254364, "learning_rate": 8.053254277547248e-05, "loss": 0.0347, "step": 26110 }, { "grad_norm": 0.32483458518981934, "learning_rate": 8.051617288400182e-05, "loss": 0.0434, "step": 26120 }, { "grad_norm": 0.39804351329803467, "learning_rate": 8.049979777815182e-05, "loss": 0.0419, "step": 26130 }, { "grad_norm": 0.396444171667099, "learning_rate": 8.048341746072054e-05, "loss": 0.0417, "step": 26140 }, { "grad_norm": 0.38465115427970886, "learning_rate": 8.046703193450696e-05, "loss": 0.0458, "step": 26150 }, { "grad_norm": 0.33512189984321594, "learning_rate": 8.04506412023109e-05, "loss": 0.0413, "step": 26160 }, { "grad_norm": 0.3910392224788666, "learning_rate": 8.043424526693306e-05, "loss": 0.0417, "step": 26170 }, { "grad_norm": 0.5151392221450806, "learning_rate": 8.04178441311751e-05, "loss": 0.0404, "step": 26180 }, { "grad_norm": 0.42713961005210876, "learning_rate": 8.04014377978395e-05, "loss": 0.0469, "step": 26190 }, { "grad_norm": 0.37398669123649597, "learning_rate": 8.038502626972967e-05, "loss": 0.0465, "step": 26200 }, { "grad_norm": 0.4412079155445099, "learning_rate": 8.036860954964989e-05, "loss": 0.0479, "step": 26210 }, { "grad_norm": 0.40356582403182983, "learning_rate": 8.035218764040531e-05, "loss": 0.0463, "step": 26220 }, { "grad_norm": 0.47113037109375, "learning_rate": 8.033576054480199e-05, "loss": 0.0485, "step": 26230 }, { "grad_norm": 0.43603894114494324, "learning_rate": 8.031932826564688e-05, "loss": 0.0475, "step": 26240 }, { "grad_norm": 0.33874931931495667, "learning_rate": 8.030289080574782e-05, "loss": 0.0471, "step": 26250 }, { "grad_norm": 0.37915584444999695, "learning_rate": 8.028644816791349e-05, "loss": 0.0516, "step": 26260 }, { "grad_norm": 0.4469153881072998, "learning_rate": 8.027000035495351e-05, "loss": 0.0413, "step": 26270 }, { "grad_norm": 0.41634368896484375, "learning_rate": 8.025354736967836e-05, "loss": 0.0568, "step": 26280 }, { "grad_norm": 0.5540003180503845, "learning_rate": 8.023708921489941e-05, "loss": 0.0581, "step": 26290 }, { "grad_norm": 0.7871448397636414, "learning_rate": 8.022062589342887e-05, "loss": 0.055, "step": 26300 }, { "grad_norm": 0.4024234116077423, "learning_rate": 8.020415740807993e-05, "loss": 0.0618, "step": 26310 }, { "grad_norm": 0.5368990898132324, "learning_rate": 8.018768376166656e-05, "loss": 0.0469, "step": 26320 }, { "grad_norm": 0.35016658902168274, "learning_rate": 8.017120495700368e-05, "loss": 0.0422, "step": 26330 }, { "grad_norm": 0.48583459854125977, "learning_rate": 8.015472099690704e-05, "loss": 0.0523, "step": 26340 }, { "grad_norm": 0.3609102964401245, "learning_rate": 8.013823188419332e-05, "loss": 0.0423, "step": 26350 }, { "grad_norm": 0.37580108642578125, "learning_rate": 8.012173762168006e-05, "loss": 0.0461, "step": 26360 }, { "grad_norm": 0.30131447315216064, "learning_rate": 8.010523821218567e-05, "loss": 0.048, "step": 26370 }, { "grad_norm": 0.47466719150543213, "learning_rate": 8.008873365852945e-05, "loss": 0.0468, "step": 26380 }, { "grad_norm": 0.3798046112060547, "learning_rate": 8.007222396353157e-05, "loss": 0.0497, "step": 26390 }, { "grad_norm": 0.5164165496826172, "learning_rate": 8.00557091300131e-05, "loss": 0.0469, "step": 26400 }, { "grad_norm": 0.3422185778617859, "learning_rate": 8.003918916079597e-05, "loss": 0.0449, "step": 26410 }, { "grad_norm": 0.41634225845336914, "learning_rate": 8.002266405870298e-05, "loss": 0.0576, "step": 26420 }, { "grad_norm": 0.42755600810050964, "learning_rate": 8.000613382655782e-05, "loss": 0.0565, "step": 26430 }, { "grad_norm": 0.39578157663345337, "learning_rate": 7.998959846718505e-05, "loss": 0.0494, "step": 26440 }, { "grad_norm": 0.6261708736419678, "learning_rate": 7.997305798341012e-05, "loss": 0.052, "step": 26450 }, { "grad_norm": 0.41147488355636597, "learning_rate": 7.995651237805937e-05, "loss": 0.05, "step": 26460 }, { "grad_norm": 0.41052567958831787, "learning_rate": 7.993996165395996e-05, "loss": 0.0496, "step": 26470 }, { "grad_norm": 0.379311740398407, "learning_rate": 7.992340581393996e-05, "loss": 0.0441, "step": 26480 }, { "grad_norm": 0.5278164148330688, "learning_rate": 7.990684486082831e-05, "loss": 0.0461, "step": 26490 }, { "grad_norm": 0.48617106676101685, "learning_rate": 7.989027879745482e-05, "loss": 0.0486, "step": 26500 }, { "grad_norm": 0.4790695905685425, "learning_rate": 7.98737076266502e-05, "loss": 0.0542, "step": 26510 }, { "grad_norm": 0.23466433584690094, "learning_rate": 7.985713135124598e-05, "loss": 0.0478, "step": 26520 }, { "grad_norm": 0.38506028056144714, "learning_rate": 7.98405499740746e-05, "loss": 0.0452, "step": 26530 }, { "grad_norm": 0.39572441577911377, "learning_rate": 7.98239634979694e-05, "loss": 0.0431, "step": 26540 }, { "grad_norm": 0.4801928997039795, "learning_rate": 7.98073719257645e-05, "loss": 0.0485, "step": 26550 }, { "grad_norm": 0.7209150195121765, "learning_rate": 7.979077526029499e-05, "loss": 0.0556, "step": 26560 }, { "grad_norm": 0.3666312098503113, "learning_rate": 7.977417350439675e-05, "loss": 0.0445, "step": 26570 }, { "grad_norm": 0.3036925792694092, "learning_rate": 7.97575666609066e-05, "loss": 0.0372, "step": 26580 }, { "grad_norm": 0.3763616979122162, "learning_rate": 7.974095473266216e-05, "loss": 0.0495, "step": 26590 }, { "grad_norm": 0.31869474053382874, "learning_rate": 7.972433772250198e-05, "loss": 0.0412, "step": 26600 }, { "grad_norm": 0.31520429253578186, "learning_rate": 7.970771563326544e-05, "loss": 0.0396, "step": 26610 }, { "grad_norm": 0.3742499053478241, "learning_rate": 7.96910884677928e-05, "loss": 0.0422, "step": 26620 }, { "grad_norm": 0.4091358780860901, "learning_rate": 7.967445622892523e-05, "loss": 0.0495, "step": 26630 }, { "grad_norm": 0.5234507322311401, "learning_rate": 7.965781891950465e-05, "loss": 0.0446, "step": 26640 }, { "grad_norm": 0.357318252325058, "learning_rate": 7.964117654237397e-05, "loss": 0.0401, "step": 26650 }, { "grad_norm": 0.4837140142917633, "learning_rate": 7.962452910037692e-05, "loss": 0.0483, "step": 26660 }, { "grad_norm": 0.4623947739601135, "learning_rate": 7.96078765963581e-05, "loss": 0.043, "step": 26670 }, { "grad_norm": 0.7017301321029663, "learning_rate": 7.95912190331629e-05, "loss": 0.0498, "step": 26680 }, { "grad_norm": 0.3607621192932129, "learning_rate": 7.957455641363772e-05, "loss": 0.0471, "step": 26690 }, { "grad_norm": 0.49568697810173035, "learning_rate": 7.955788874062968e-05, "loss": 0.0606, "step": 26700 }, { "grad_norm": 0.4426247179508209, "learning_rate": 7.95412160169869e-05, "loss": 0.0497, "step": 26710 }, { "grad_norm": 0.537030816078186, "learning_rate": 7.952453824555824e-05, "loss": 0.0388, "step": 26720 }, { "grad_norm": 0.44080302119255066, "learning_rate": 7.95078554291935e-05, "loss": 0.0395, "step": 26730 }, { "grad_norm": 0.3640194535255432, "learning_rate": 7.94911675707433e-05, "loss": 0.0379, "step": 26740 }, { "grad_norm": 0.5912141799926758, "learning_rate": 7.947447467305915e-05, "loss": 0.0582, "step": 26750 }, { "grad_norm": 0.3525256812572479, "learning_rate": 7.94577767389934e-05, "loss": 0.0421, "step": 26760 }, { "grad_norm": 0.35247814655303955, "learning_rate": 7.944107377139928e-05, "loss": 0.0438, "step": 26770 }, { "grad_norm": 0.417721688747406, "learning_rate": 7.942436577313088e-05, "loss": 0.0403, "step": 26780 }, { "grad_norm": 0.352960467338562, "learning_rate": 7.940765274704312e-05, "loss": 0.043, "step": 26790 }, { "grad_norm": 0.30500948429107666, "learning_rate": 7.939093469599181e-05, "loss": 0.0412, "step": 26800 }, { "grad_norm": 0.30363479256629944, "learning_rate": 7.93742116228336e-05, "loss": 0.0429, "step": 26810 }, { "grad_norm": 0.5037378668785095, "learning_rate": 7.935748353042602e-05, "loss": 0.0456, "step": 26820 }, { "grad_norm": 0.39375630021095276, "learning_rate": 7.934075042162744e-05, "loss": 0.05, "step": 26830 }, { "grad_norm": 0.47409117221832275, "learning_rate": 7.932401229929705e-05, "loss": 0.048, "step": 26840 }, { "grad_norm": 0.39227500557899475, "learning_rate": 7.9307269166295e-05, "loss": 0.0453, "step": 26850 }, { "grad_norm": 0.27102893590927124, "learning_rate": 7.92905210254822e-05, "loss": 0.0393, "step": 26860 }, { "grad_norm": 0.3455027639865875, "learning_rate": 7.927376787972045e-05, "loss": 0.0395, "step": 26870 }, { "grad_norm": 0.42572110891342163, "learning_rate": 7.92570097318724e-05, "loss": 0.0419, "step": 26880 }, { "grad_norm": 0.7139022350311279, "learning_rate": 7.924024658480158e-05, "loss": 0.0453, "step": 26890 }, { "grad_norm": 0.43869560956954956, "learning_rate": 7.922347844137233e-05, "loss": 0.041, "step": 26900 }, { "grad_norm": 0.4645061194896698, "learning_rate": 7.92067053044499e-05, "loss": 0.0459, "step": 26910 }, { "grad_norm": 0.5659179091453552, "learning_rate": 7.918992717690031e-05, "loss": 0.0518, "step": 26920 }, { "grad_norm": 0.4340420961380005, "learning_rate": 7.917314406159053e-05, "loss": 0.0475, "step": 26930 }, { "grad_norm": 0.5612347722053528, "learning_rate": 7.915635596138832e-05, "loss": 0.064, "step": 26940 }, { "grad_norm": 0.49918094277381897, "learning_rate": 7.913956287916228e-05, "loss": 0.0459, "step": 26950 }, { "grad_norm": 0.29210710525512695, "learning_rate": 7.912276481778193e-05, "loss": 0.0486, "step": 26960 }, { "grad_norm": 0.41984429955482483, "learning_rate": 7.910596178011759e-05, "loss": 0.058, "step": 26970 }, { "grad_norm": 0.3360656797885895, "learning_rate": 7.908915376904043e-05, "loss": 0.0469, "step": 26980 }, { "grad_norm": 0.5108562111854553, "learning_rate": 7.907234078742247e-05, "loss": 0.0445, "step": 26990 }, { "grad_norm": 0.39503517746925354, "learning_rate": 7.90555228381366e-05, "loss": 0.0388, "step": 27000 }, { "grad_norm": 0.4687086045742035, "learning_rate": 7.903869992405656e-05, "loss": 0.0448, "step": 27010 }, { "grad_norm": 0.34978848695755005, "learning_rate": 7.902187204805691e-05, "loss": 0.0418, "step": 27020 }, { "grad_norm": 0.33104902505874634, "learning_rate": 7.900503921301308e-05, "loss": 0.0404, "step": 27030 }, { "grad_norm": 0.4178422689437866, "learning_rate": 7.898820142180133e-05, "loss": 0.0528, "step": 27040 }, { "grad_norm": 0.5167988538742065, "learning_rate": 7.897135867729879e-05, "loss": 0.0408, "step": 27050 }, { "grad_norm": 0.41559213399887085, "learning_rate": 7.89545109823834e-05, "loss": 0.0385, "step": 27060 }, { "grad_norm": 0.47454944252967834, "learning_rate": 7.8937658339934e-05, "loss": 0.0463, "step": 27070 }, { "grad_norm": 0.47810450196266174, "learning_rate": 7.892080075283026e-05, "loss": 0.0481, "step": 27080 }, { "grad_norm": 0.3611404299736023, "learning_rate": 7.890393822395263e-05, "loss": 0.0379, "step": 27090 }, { "grad_norm": 0.31826549768447876, "learning_rate": 7.88870707561825e-05, "loss": 0.0407, "step": 27100 }, { "grad_norm": 0.5238865613937378, "learning_rate": 7.887019835240203e-05, "loss": 0.0506, "step": 27110 }, { "grad_norm": 0.5151017904281616, "learning_rate": 7.885332101549427e-05, "loss": 0.0454, "step": 27120 }, { "grad_norm": 0.43680137395858765, "learning_rate": 7.883643874834308e-05, "loss": 0.0429, "step": 27130 }, { "grad_norm": 0.2818663716316223, "learning_rate": 7.881955155383321e-05, "loss": 0.0575, "step": 27140 }, { "grad_norm": 0.4736890494823456, "learning_rate": 7.880265943485017e-05, "loss": 0.0416, "step": 27150 }, { "grad_norm": 0.32807430624961853, "learning_rate": 7.878576239428038e-05, "loss": 0.0442, "step": 27160 }, { "grad_norm": 0.45609134435653687, "learning_rate": 7.87688604350111e-05, "loss": 0.0499, "step": 27170 }, { "grad_norm": 0.5715551972389221, "learning_rate": 7.875195355993042e-05, "loss": 0.0578, "step": 27180 }, { "grad_norm": 0.33086544275283813, "learning_rate": 7.873504177192724e-05, "loss": 0.045, "step": 27190 }, { "grad_norm": 0.4773387014865875, "learning_rate": 7.87181250738913e-05, "loss": 0.0532, "step": 27200 }, { "grad_norm": 0.499703586101532, "learning_rate": 7.870120346871324e-05, "loss": 0.0458, "step": 27210 }, { "grad_norm": 0.4935605823993683, "learning_rate": 7.86842769592845e-05, "loss": 0.0419, "step": 27220 }, { "grad_norm": 0.3810887634754181, "learning_rate": 7.866734554849732e-05, "loss": 0.0433, "step": 27230 }, { "grad_norm": 0.3956353962421417, "learning_rate": 7.865040923924486e-05, "loss": 0.0398, "step": 27240 }, { "grad_norm": 0.48770761489868164, "learning_rate": 7.863346803442104e-05, "loss": 0.0506, "step": 27250 }, { "grad_norm": 0.44107210636138916, "learning_rate": 7.861652193692067e-05, "loss": 0.0366, "step": 27260 }, { "grad_norm": 0.4999356269836426, "learning_rate": 7.859957094963937e-05, "loss": 0.0418, "step": 27270 }, { "grad_norm": 0.3214062750339508, "learning_rate": 7.858261507547357e-05, "loss": 0.0422, "step": 27280 }, { "grad_norm": 0.46260735392570496, "learning_rate": 7.856565431732061e-05, "loss": 0.039, "step": 27290 }, { "grad_norm": 0.28223928809165955, "learning_rate": 7.854868867807859e-05, "loss": 0.0341, "step": 27300 }, { "grad_norm": 0.49436402320861816, "learning_rate": 7.85317181606465e-05, "loss": 0.0447, "step": 27310 }, { "grad_norm": 0.3358601927757263, "learning_rate": 7.85147427679241e-05, "loss": 0.0359, "step": 27320 }, { "grad_norm": 0.47689616680145264, "learning_rate": 7.849776250281205e-05, "loss": 0.034, "step": 27330 }, { "grad_norm": 0.39507389068603516, "learning_rate": 7.84807773682118e-05, "loss": 0.0446, "step": 27340 }, { "grad_norm": 0.3781898319721222, "learning_rate": 7.846378736702565e-05, "loss": 0.0485, "step": 27350 }, { "grad_norm": 0.424154669046402, "learning_rate": 7.844679250215671e-05, "loss": 0.0511, "step": 27360 }, { "grad_norm": 0.37052780389785767, "learning_rate": 7.842979277650898e-05, "loss": 0.036, "step": 27370 }, { "grad_norm": 0.454923152923584, "learning_rate": 7.84127881929872e-05, "loss": 0.0426, "step": 27380 }, { "grad_norm": 0.4761653542518616, "learning_rate": 7.839577875449704e-05, "loss": 0.0463, "step": 27390 }, { "grad_norm": 0.5268871188163757, "learning_rate": 7.837876446394489e-05, "loss": 0.0377, "step": 27400 }, { "grad_norm": 0.4460377097129822, "learning_rate": 7.836174532423805e-05, "loss": 0.0485, "step": 27410 }, { "grad_norm": 0.49883103370666504, "learning_rate": 7.834472133828466e-05, "loss": 0.0501, "step": 27420 }, { "grad_norm": 0.5192834138870239, "learning_rate": 7.832769250899359e-05, "loss": 0.0446, "step": 27430 }, { "grad_norm": 0.48081856966018677, "learning_rate": 7.831065883927464e-05, "loss": 0.0513, "step": 27440 }, { "grad_norm": 0.3459348678588867, "learning_rate": 7.829362033203841e-05, "loss": 0.0464, "step": 27450 }, { "grad_norm": 0.4125630855560303, "learning_rate": 7.827657699019628e-05, "loss": 0.0436, "step": 27460 }, { "grad_norm": 0.4546297490596771, "learning_rate": 7.825952881666052e-05, "loss": 0.0627, "step": 27470 }, { "grad_norm": 0.477810800075531, "learning_rate": 7.824247581434418e-05, "loss": 0.0421, "step": 27480 }, { "grad_norm": 0.5552513599395752, "learning_rate": 7.822541798616116e-05, "loss": 0.0402, "step": 27490 }, { "grad_norm": 0.5413710474967957, "learning_rate": 7.820835533502617e-05, "loss": 0.053, "step": 27500 }, { "grad_norm": 0.5146156549453735, "learning_rate": 7.819128786385475e-05, "loss": 0.0615, "step": 27510 }, { "grad_norm": 0.3798138499259949, "learning_rate": 7.817421557556329e-05, "loss": 0.052, "step": 27520 }, { "grad_norm": 0.4280366599559784, "learning_rate": 7.815713847306893e-05, "loss": 0.0547, "step": 27530 }, { "grad_norm": 0.3730621635913849, "learning_rate": 7.81400565592897e-05, "loss": 0.0425, "step": 27540 }, { "grad_norm": 0.38260984420776367, "learning_rate": 7.812296983714444e-05, "loss": 0.055, "step": 27550 }, { "grad_norm": 0.4322867691516876, "learning_rate": 7.810587830955281e-05, "loss": 0.0476, "step": 27560 }, { "grad_norm": 0.4551068842411041, "learning_rate": 7.808878197943528e-05, "loss": 0.0488, "step": 27570 }, { "grad_norm": 0.4620976448059082, "learning_rate": 7.807168084971312e-05, "loss": 0.0444, "step": 27580 }, { "grad_norm": 0.4433280825614929, "learning_rate": 7.805457492330849e-05, "loss": 0.0396, "step": 27590 }, { "grad_norm": 0.4507081210613251, "learning_rate": 7.803746420314428e-05, "loss": 0.043, "step": 27600 }, { "grad_norm": 0.4889121651649475, "learning_rate": 7.802034869214428e-05, "loss": 0.0513, "step": 27610 }, { "grad_norm": 0.5119116902351379, "learning_rate": 7.800322839323303e-05, "loss": 0.0476, "step": 27620 }, { "grad_norm": 0.43861156702041626, "learning_rate": 7.798610330933593e-05, "loss": 0.0456, "step": 27630 }, { "grad_norm": 0.2986944913864136, "learning_rate": 7.796897344337922e-05, "loss": 0.0476, "step": 27640 }, { "grad_norm": 0.3240344226360321, "learning_rate": 7.795183879828989e-05, "loss": 0.0548, "step": 27650 }, { "grad_norm": 0.4590821862220764, "learning_rate": 7.793469937699579e-05, "loss": 0.0554, "step": 27660 }, { "grad_norm": 0.373725026845932, "learning_rate": 7.791755518242558e-05, "loss": 0.0516, "step": 27670 }, { "grad_norm": 0.45902717113494873, "learning_rate": 7.790040621750876e-05, "loss": 0.0512, "step": 27680 }, { "grad_norm": 0.34957045316696167, "learning_rate": 7.788325248517558e-05, "loss": 0.0373, "step": 27690 }, { "grad_norm": 0.44577762484550476, "learning_rate": 7.786609398835715e-05, "loss": 0.042, "step": 27700 }, { "grad_norm": 0.4954073131084442, "learning_rate": 7.784893072998541e-05, "loss": 0.0433, "step": 27710 }, { "grad_norm": 0.2585253119468689, "learning_rate": 7.783176271299306e-05, "loss": 0.0442, "step": 27720 }, { "grad_norm": 0.32164427638053894, "learning_rate": 7.781458994031368e-05, "loss": 0.0353, "step": 27730 }, { "grad_norm": 0.510894775390625, "learning_rate": 7.779741241488161e-05, "loss": 0.0442, "step": 27740 }, { "grad_norm": 0.4764552116394043, "learning_rate": 7.7780230139632e-05, "loss": 0.0399, "step": 27750 }, { "grad_norm": 0.579217791557312, "learning_rate": 7.776304311750087e-05, "loss": 0.0515, "step": 27760 }, { "grad_norm": 0.3633778989315033, "learning_rate": 7.7745851351425e-05, "loss": 0.052, "step": 27770 }, { "grad_norm": 0.4713967740535736, "learning_rate": 7.772865484434197e-05, "loss": 0.0493, "step": 27780 }, { "grad_norm": 0.5026031732559204, "learning_rate": 7.77114535991902e-05, "loss": 0.0505, "step": 27790 }, { "grad_norm": 0.4487420320510864, "learning_rate": 7.769424761890893e-05, "loss": 0.0461, "step": 27800 }, { "grad_norm": 0.4328596293926239, "learning_rate": 7.767703690643817e-05, "loss": 0.053, "step": 27810 }, { "grad_norm": 0.3475213050842285, "learning_rate": 7.76598214647188e-05, "loss": 0.0369, "step": 27820 }, { "grad_norm": 0.42082926630973816, "learning_rate": 7.764260129669241e-05, "loss": 0.0439, "step": 27830 }, { "grad_norm": 0.3965909779071808, "learning_rate": 7.76253764053015e-05, "loss": 0.0452, "step": 27840 }, { "grad_norm": 0.4676735997200012, "learning_rate": 7.760814679348932e-05, "loss": 0.0399, "step": 27850 }, { "grad_norm": 0.3639766275882721, "learning_rate": 7.759091246419992e-05, "loss": 0.047, "step": 27860 }, { "grad_norm": 0.4722185730934143, "learning_rate": 7.75736734203782e-05, "loss": 0.0413, "step": 27870 }, { "grad_norm": 0.5937740206718445, "learning_rate": 7.755642966496985e-05, "loss": 0.0637, "step": 27880 }, { "grad_norm": 0.35626691579818726, "learning_rate": 7.753918120092132e-05, "loss": 0.0416, "step": 27890 }, { "grad_norm": 0.5127270817756653, "learning_rate": 7.752192803117993e-05, "loss": 0.0363, "step": 27900 }, { "grad_norm": 0.390949010848999, "learning_rate": 7.750467015869377e-05, "loss": 0.0425, "step": 27910 }, { "grad_norm": 0.3148009181022644, "learning_rate": 7.748740758641174e-05, "loss": 0.0489, "step": 27920 }, { "grad_norm": 0.32980942726135254, "learning_rate": 7.74701403172835e-05, "loss": 0.045, "step": 27930 }, { "grad_norm": 0.35863739252090454, "learning_rate": 7.745286835425962e-05, "loss": 0.0391, "step": 27940 }, { "grad_norm": 0.5843835473060608, "learning_rate": 7.743559170029138e-05, "loss": 0.0424, "step": 27950 }, { "grad_norm": 0.34523847699165344, "learning_rate": 7.741831035833087e-05, "loss": 0.0557, "step": 27960 }, { "grad_norm": 0.34790685772895813, "learning_rate": 7.740102433133102e-05, "loss": 0.0385, "step": 27970 }, { "grad_norm": 0.3015878200531006, "learning_rate": 7.738373362224553e-05, "loss": 0.0394, "step": 27980 }, { "grad_norm": 0.44172513484954834, "learning_rate": 7.73664382340289e-05, "loss": 0.0468, "step": 27990 }, { "grad_norm": 0.4458095133304596, "learning_rate": 7.734913816963647e-05, "loss": 0.0498, "step": 28000 }, { "grad_norm": 0.29722076654434204, "learning_rate": 7.73318334320243e-05, "loss": 0.0391, "step": 28010 }, { "grad_norm": 0.40787166357040405, "learning_rate": 7.731452402414934e-05, "loss": 0.0478, "step": 28020 }, { "grad_norm": 0.43499210476875305, "learning_rate": 7.729720994896928e-05, "loss": 0.0474, "step": 28030 }, { "grad_norm": 0.4702828824520111, "learning_rate": 7.727989120944262e-05, "loss": 0.047, "step": 28040 }, { "grad_norm": 0.3148435354232788, "learning_rate": 7.726256780852865e-05, "loss": 0.0424, "step": 28050 }, { "grad_norm": 0.3461373746395111, "learning_rate": 7.724523974918749e-05, "loss": 0.0432, "step": 28060 }, { "grad_norm": 0.3589492440223694, "learning_rate": 7.722790703438002e-05, "loss": 0.0451, "step": 28070 }, { "grad_norm": 0.311428427696228, "learning_rate": 7.72105696670679e-05, "loss": 0.0428, "step": 28080 }, { "grad_norm": 0.41638755798339844, "learning_rate": 7.719322765021364e-05, "loss": 0.0399, "step": 28090 }, { "grad_norm": 0.416213721036911, "learning_rate": 7.717588098678051e-05, "loss": 0.0443, "step": 28100 }, { "grad_norm": 0.48878321051597595, "learning_rate": 7.715852967973258e-05, "loss": 0.0474, "step": 28110 }, { "grad_norm": 0.37340739369392395, "learning_rate": 7.714117373203474e-05, "loss": 0.0474, "step": 28120 }, { "grad_norm": 0.38572293519973755, "learning_rate": 7.712381314665259e-05, "loss": 0.0398, "step": 28130 }, { "grad_norm": 0.3583850562572479, "learning_rate": 7.710644792655261e-05, "loss": 0.0338, "step": 28140 }, { "grad_norm": 0.6079774498939514, "learning_rate": 7.708907807470207e-05, "loss": 0.0573, "step": 28150 }, { "grad_norm": 0.4076559245586395, "learning_rate": 7.707170359406896e-05, "loss": 0.0418, "step": 28160 }, { "grad_norm": 0.4651680588722229, "learning_rate": 7.705432448762213e-05, "loss": 0.0387, "step": 28170 }, { "grad_norm": 0.4288778603076935, "learning_rate": 7.703694075833117e-05, "loss": 0.0517, "step": 28180 }, { "grad_norm": 0.4495231807231903, "learning_rate": 7.70195524091665e-05, "loss": 0.0539, "step": 28190 }, { "grad_norm": 0.4604308307170868, "learning_rate": 7.70021594430993e-05, "loss": 0.0442, "step": 28200 }, { "grad_norm": 0.4307192862033844, "learning_rate": 7.698476186310157e-05, "loss": 0.0422, "step": 28210 }, { "grad_norm": 0.3972584307193756, "learning_rate": 7.696735967214608e-05, "loss": 0.0466, "step": 28220 }, { "grad_norm": 0.45331886410713196, "learning_rate": 7.694995287320636e-05, "loss": 0.0389, "step": 28230 }, { "grad_norm": 0.34738919138908386, "learning_rate": 7.693254146925679e-05, "loss": 0.0375, "step": 28240 }, { "grad_norm": 0.4016478955745697, "learning_rate": 7.691512546327251e-05, "loss": 0.05, "step": 28250 }, { "grad_norm": 0.39721426367759705, "learning_rate": 7.689770485822939e-05, "loss": 0.0466, "step": 28260 }, { "grad_norm": 0.34066513180732727, "learning_rate": 7.688027965710416e-05, "loss": 0.0417, "step": 28270 }, { "grad_norm": 0.35613444447517395, "learning_rate": 7.686284986287433e-05, "loss": 0.0479, "step": 28280 }, { "grad_norm": 0.350129097700119, "learning_rate": 7.684541547851817e-05, "loss": 0.0382, "step": 28290 }, { "grad_norm": 0.37263208627700806, "learning_rate": 7.68279765070147e-05, "loss": 0.0436, "step": 28300 }, { "grad_norm": 0.5027300119400024, "learning_rate": 7.68105329513438e-05, "loss": 0.0408, "step": 28310 }, { "grad_norm": 0.3468482792377472, "learning_rate": 7.67930848144861e-05, "loss": 0.0424, "step": 28320 }, { "grad_norm": 0.47429078817367554, "learning_rate": 7.6775632099423e-05, "loss": 0.0556, "step": 28330 }, { "grad_norm": 0.40438589453697205, "learning_rate": 7.675817480913667e-05, "loss": 0.0427, "step": 28340 }, { "grad_norm": 0.5119240880012512, "learning_rate": 7.674071294661011e-05, "loss": 0.0447, "step": 28350 }, { "grad_norm": 0.31705808639526367, "learning_rate": 7.672324651482707e-05, "loss": 0.0424, "step": 28360 }, { "grad_norm": 0.38536685705184937, "learning_rate": 7.670577551677209e-05, "loss": 0.0501, "step": 28370 }, { "grad_norm": 0.44820505380630493, "learning_rate": 7.668829995543047e-05, "loss": 0.0446, "step": 28380 }, { "grad_norm": 0.38282179832458496, "learning_rate": 7.667081983378832e-05, "loss": 0.0445, "step": 28390 }, { "grad_norm": 0.4524764120578766, "learning_rate": 7.66533351548325e-05, "loss": 0.0423, "step": 28400 }, { "grad_norm": 0.46708056330680847, "learning_rate": 7.663584592155069e-05, "loss": 0.0456, "step": 28410 }, { "grad_norm": 0.43326497077941895, "learning_rate": 7.661835213693129e-05, "loss": 0.0474, "step": 28420 }, { "grad_norm": 0.5510104298591614, "learning_rate": 7.660085380396353e-05, "loss": 0.0468, "step": 28430 }, { "grad_norm": 0.39175179600715637, "learning_rate": 7.658335092563738e-05, "loss": 0.0385, "step": 28440 }, { "grad_norm": 0.41563528776168823, "learning_rate": 7.656584350494362e-05, "loss": 0.0493, "step": 28450 }, { "grad_norm": 0.339163601398468, "learning_rate": 7.654833154487378e-05, "loss": 0.0431, "step": 28460 }, { "grad_norm": 0.5913695693016052, "learning_rate": 7.653081504842017e-05, "loss": 0.0399, "step": 28470 }, { "grad_norm": 0.4570125937461853, "learning_rate": 7.65132940185759e-05, "loss": 0.0534, "step": 28480 }, { "grad_norm": 0.5596985816955566, "learning_rate": 7.649576845833481e-05, "loss": 0.0374, "step": 28490 }, { "grad_norm": 0.3680844306945801, "learning_rate": 7.647823837069156e-05, "loss": 0.0404, "step": 28500 }, { "grad_norm": 0.5000932216644287, "learning_rate": 7.646070375864156e-05, "loss": 0.0417, "step": 28510 }, { "grad_norm": 0.31219708919525146, "learning_rate": 7.644316462518099e-05, "loss": 0.0471, "step": 28520 }, { "grad_norm": 0.4481915831565857, "learning_rate": 7.642562097330679e-05, "loss": 0.0448, "step": 28530 }, { "grad_norm": 0.7291484475135803, "learning_rate": 7.640807280601671e-05, "loss": 0.0449, "step": 28540 }, { "grad_norm": 0.4364989697933197, "learning_rate": 7.639052012630927e-05, "loss": 0.0493, "step": 28550 }, { "grad_norm": 0.4711017310619354, "learning_rate": 7.63729629371837e-05, "loss": 0.0517, "step": 28560 }, { "grad_norm": 0.46408024430274963, "learning_rate": 7.635540124164009e-05, "loss": 0.0583, "step": 28570 }, { "grad_norm": 0.42161688208580017, "learning_rate": 7.633783504267922e-05, "loss": 0.0491, "step": 28580 }, { "grad_norm": 0.39740127325057983, "learning_rate": 7.632026434330269e-05, "loss": 0.0488, "step": 28590 }, { "grad_norm": 0.5081238746643066, "learning_rate": 7.630268914651282e-05, "loss": 0.0408, "step": 28600 }, { "grad_norm": 0.4836149215698242, "learning_rate": 7.628510945531278e-05, "loss": 0.0597, "step": 28610 }, { "grad_norm": 0.41092050075531006, "learning_rate": 7.626752527270641e-05, "loss": 0.0456, "step": 28620 }, { "grad_norm": 0.39010173082351685, "learning_rate": 7.62499366016984e-05, "loss": 0.0402, "step": 28630 }, { "grad_norm": 0.5057068467140198, "learning_rate": 7.623234344529416e-05, "loss": 0.0469, "step": 28640 }, { "grad_norm": 0.2672922611236572, "learning_rate": 7.62147458064999e-05, "loss": 0.0412, "step": 28650 }, { "grad_norm": 0.41379430890083313, "learning_rate": 7.619714368832254e-05, "loss": 0.0456, "step": 28660 }, { "grad_norm": 0.40410298109054565, "learning_rate": 7.61795370937698e-05, "loss": 0.0385, "step": 28670 }, { "grad_norm": 0.31789764761924744, "learning_rate": 7.61619260258502e-05, "loss": 0.0434, "step": 28680 }, { "grad_norm": 0.36931654810905457, "learning_rate": 7.614431048757298e-05, "loss": 0.0493, "step": 28690 }, { "grad_norm": 0.2628128230571747, "learning_rate": 7.612669048194814e-05, "loss": 0.0382, "step": 28700 }, { "grad_norm": 0.4047898054122925, "learning_rate": 7.610906601198646e-05, "loss": 0.0389, "step": 28710 }, { "grad_norm": 0.4483082592487335, "learning_rate": 7.60914370806995e-05, "loss": 0.0554, "step": 28720 }, { "grad_norm": 0.38309434056282043, "learning_rate": 7.607380369109953e-05, "loss": 0.0608, "step": 28730 }, { "grad_norm": 0.38015392422676086, "learning_rate": 7.605616584619961e-05, "loss": 0.0387, "step": 28740 }, { "grad_norm": 0.5318105816841125, "learning_rate": 7.603852354901362e-05, "loss": 0.0484, "step": 28750 }, { "grad_norm": 0.2930002212524414, "learning_rate": 7.602087680255609e-05, "loss": 0.0574, "step": 28760 }, { "grad_norm": 0.37042900919914246, "learning_rate": 7.600322560984238e-05, "loss": 0.0386, "step": 28770 }, { "grad_norm": 0.3850346505641937, "learning_rate": 7.598556997388863e-05, "loss": 0.0449, "step": 28780 }, { "grad_norm": 0.4368472099304199, "learning_rate": 7.596790989771166e-05, "loss": 0.0471, "step": 28790 }, { "grad_norm": 0.466219425201416, "learning_rate": 7.595024538432914e-05, "loss": 0.0384, "step": 28800 }, { "grad_norm": 0.28564608097076416, "learning_rate": 7.59325764367594e-05, "loss": 0.0418, "step": 28810 }, { "grad_norm": 0.40907353162765503, "learning_rate": 7.59149030580216e-05, "loss": 0.0418, "step": 28820 }, { "grad_norm": 0.4329425096511841, "learning_rate": 7.589722525113562e-05, "loss": 0.0377, "step": 28830 }, { "grad_norm": 0.39153966307640076, "learning_rate": 7.587954301912216e-05, "loss": 0.0387, "step": 28840 }, { "grad_norm": 0.5195205211639404, "learning_rate": 7.586185636500263e-05, "loss": 0.0475, "step": 28850 }, { "grad_norm": 0.47791793942451477, "learning_rate": 7.584416529179914e-05, "loss": 0.0481, "step": 28860 }, { "grad_norm": 0.3337494432926178, "learning_rate": 7.582646980253465e-05, "loss": 0.0401, "step": 28870 }, { "grad_norm": 0.33628904819488525, "learning_rate": 7.580876990023282e-05, "loss": 0.0537, "step": 28880 }, { "grad_norm": 0.41553327441215515, "learning_rate": 7.579106558791809e-05, "loss": 0.0436, "step": 28890 }, { "grad_norm": 0.37479618191719055, "learning_rate": 7.577335686861565e-05, "loss": 0.0391, "step": 28900 }, { "grad_norm": 0.5151191353797913, "learning_rate": 7.575564374535141e-05, "loss": 0.0488, "step": 28910 }, { "grad_norm": 0.2885504364967346, "learning_rate": 7.573792622115207e-05, "loss": 0.0335, "step": 28920 }, { "grad_norm": 0.418960303068161, "learning_rate": 7.572020429904507e-05, "loss": 0.0394, "step": 28930 }, { "grad_norm": 0.3938234746456146, "learning_rate": 7.570247798205861e-05, "loss": 0.0412, "step": 28940 }, { "grad_norm": 0.3304603397846222, "learning_rate": 7.568474727322164e-05, "loss": 0.0484, "step": 28950 }, { "grad_norm": 0.4467475414276123, "learning_rate": 7.566701217556384e-05, "loss": 0.0424, "step": 28960 }, { "grad_norm": 0.47939130663871765, "learning_rate": 7.564927269211564e-05, "loss": 0.0444, "step": 28970 }, { "grad_norm": 0.6184108257293701, "learning_rate": 7.563152882590824e-05, "loss": 0.0446, "step": 28980 }, { "grad_norm": 0.4698997139930725, "learning_rate": 7.56137805799736e-05, "loss": 0.0471, "step": 28990 }, { "grad_norm": 0.4578916132450104, "learning_rate": 7.559602795734439e-05, "loss": 0.0441, "step": 29000 }, { "grad_norm": 0.5417869687080383, "learning_rate": 7.557827096105408e-05, "loss": 0.0423, "step": 29010 }, { "grad_norm": 0.33436140418052673, "learning_rate": 7.55605095941368e-05, "loss": 0.0412, "step": 29020 }, { "grad_norm": 0.40566954016685486, "learning_rate": 7.554274385962752e-05, "loss": 0.0402, "step": 29030 }, { "grad_norm": 0.39276936650276184, "learning_rate": 7.552497376056191e-05, "loss": 0.0401, "step": 29040 }, { "grad_norm": 0.7140429019927979, "learning_rate": 7.550719929997639e-05, "loss": 0.04, "step": 29050 }, { "grad_norm": 0.411138117313385, "learning_rate": 7.548942048090813e-05, "loss": 0.0412, "step": 29060 }, { "grad_norm": 0.41358813643455505, "learning_rate": 7.547163730639506e-05, "loss": 0.0446, "step": 29070 }, { "grad_norm": 0.48150455951690674, "learning_rate": 7.545384977947583e-05, "loss": 0.0426, "step": 29080 }, { "grad_norm": 0.47399353981018066, "learning_rate": 7.543605790318981e-05, "loss": 0.0418, "step": 29090 }, { "grad_norm": 0.5302382111549377, "learning_rate": 7.54182616805772e-05, "loss": 0.0391, "step": 29100 }, { "grad_norm": 0.29800596833229065, "learning_rate": 7.540046111467885e-05, "loss": 0.0353, "step": 29110 }, { "grad_norm": 0.33767029643058777, "learning_rate": 7.53826562085364e-05, "loss": 0.0423, "step": 29120 }, { "grad_norm": 0.40497392416000366, "learning_rate": 7.536484696519221e-05, "loss": 0.0353, "step": 29130 }, { "grad_norm": 0.43178486824035645, "learning_rate": 7.534703338768942e-05, "loss": 0.0434, "step": 29140 }, { "grad_norm": 0.47246184945106506, "learning_rate": 7.532921547907185e-05, "loss": 0.0434, "step": 29150 }, { "grad_norm": 0.4136105477809906, "learning_rate": 7.531139324238412e-05, "loss": 0.0435, "step": 29160 }, { "grad_norm": 0.5905318260192871, "learning_rate": 7.529356668067157e-05, "loss": 0.0595, "step": 29170 }, { "grad_norm": 0.5507924556732178, "learning_rate": 7.527573579698023e-05, "loss": 0.0501, "step": 29180 }, { "grad_norm": 0.42048102617263794, "learning_rate": 7.525790059435693e-05, "loss": 0.0405, "step": 29190 }, { "grad_norm": 0.4282647669315338, "learning_rate": 7.524006107584926e-05, "loss": 0.0406, "step": 29200 }, { "grad_norm": 0.44731149077415466, "learning_rate": 7.522221724450544e-05, "loss": 0.0364, "step": 29210 }, { "grad_norm": 0.5257572531700134, "learning_rate": 7.520436910337451e-05, "loss": 0.0557, "step": 29220 }, { "grad_norm": 0.21800148487091064, "learning_rate": 7.518651665550627e-05, "loss": 0.0391, "step": 29230 }, { "grad_norm": 0.7859484553337097, "learning_rate": 7.516865990395117e-05, "loss": 0.0561, "step": 29240 }, { "grad_norm": 0.3381710350513458, "learning_rate": 7.515079885176047e-05, "loss": 0.0366, "step": 29250 }, { "grad_norm": 0.2738135755062103, "learning_rate": 7.513293350198612e-05, "loss": 0.0392, "step": 29260 }, { "grad_norm": 0.37422290444374084, "learning_rate": 7.511506385768081e-05, "loss": 0.0468, "step": 29270 }, { "grad_norm": 0.49286949634552, "learning_rate": 7.509718992189801e-05, "loss": 0.0468, "step": 29280 }, { "grad_norm": 0.4258803129196167, "learning_rate": 7.507931169769182e-05, "loss": 0.041, "step": 29290 }, { "grad_norm": 0.47925320267677307, "learning_rate": 7.506142918811722e-05, "loss": 0.0409, "step": 29300 }, { "grad_norm": 0.4313034117221832, "learning_rate": 7.504354239622978e-05, "loss": 0.0504, "step": 29310 }, { "grad_norm": 0.4023558497428894, "learning_rate": 7.50256513250859e-05, "loss": 0.0453, "step": 29320 }, { "grad_norm": 0.4786669909954071, "learning_rate": 7.500775597774265e-05, "loss": 0.0422, "step": 29330 }, { "grad_norm": 0.45799699425697327, "learning_rate": 7.498985635725788e-05, "loss": 0.0384, "step": 29340 }, { "grad_norm": 0.3890485465526581, "learning_rate": 7.497195246669012e-05, "loss": 0.0445, "step": 29350 }, { "grad_norm": 0.3738795220851898, "learning_rate": 7.495404430909868e-05, "loss": 0.0436, "step": 29360 }, { "grad_norm": 0.4278547167778015, "learning_rate": 7.493613188754356e-05, "loss": 0.0432, "step": 29370 }, { "grad_norm": 0.36558759212493896, "learning_rate": 7.49182152050855e-05, "loss": 0.0392, "step": 29380 }, { "grad_norm": 0.4894861876964569, "learning_rate": 7.490029426478598e-05, "loss": 0.0384, "step": 29390 }, { "grad_norm": 0.46995994448661804, "learning_rate": 7.488236906970719e-05, "loss": 0.041, "step": 29400 }, { "grad_norm": 0.561512291431427, "learning_rate": 7.486443962291207e-05, "loss": 0.0458, "step": 29410 }, { "grad_norm": 0.42478033900260925, "learning_rate": 7.484650592746424e-05, "loss": 0.0368, "step": 29420 }, { "grad_norm": 0.3682441711425781, "learning_rate": 7.482856798642811e-05, "loss": 0.045, "step": 29430 }, { "grad_norm": 0.29069435596466064, "learning_rate": 7.481062580286878e-05, "loss": 0.0399, "step": 29440 }, { "grad_norm": 0.2815939784049988, "learning_rate": 7.479267937985208e-05, "loss": 0.0376, "step": 29450 }, { "grad_norm": 0.2865943908691406, "learning_rate": 7.477472872044456e-05, "loss": 0.0349, "step": 29460 }, { "grad_norm": 0.6199798583984375, "learning_rate": 7.475677382771347e-05, "loss": 0.0454, "step": 29470 }, { "grad_norm": 0.41849595308303833, "learning_rate": 7.473881470472683e-05, "loss": 0.0479, "step": 29480 }, { "grad_norm": 0.49376022815704346, "learning_rate": 7.472085135455337e-05, "loss": 0.0463, "step": 29490 }, { "grad_norm": 0.3007417619228363, "learning_rate": 7.470288378026256e-05, "loss": 0.0492, "step": 29500 }, { "grad_norm": 0.446940153837204, "learning_rate": 7.468491198492451e-05, "loss": 0.0415, "step": 29510 }, { "grad_norm": 0.4486255347728729, "learning_rate": 7.466693597161013e-05, "loss": 0.0388, "step": 29520 }, { "grad_norm": 0.32036536931991577, "learning_rate": 7.464895574339104e-05, "loss": 0.0379, "step": 29530 }, { "grad_norm": 0.344046950340271, "learning_rate": 7.463097130333958e-05, "loss": 0.0604, "step": 29540 }, { "grad_norm": 0.3066578507423401, "learning_rate": 7.461298265452876e-05, "loss": 0.0439, "step": 29550 }, { "grad_norm": 0.36957094073295593, "learning_rate": 7.459498980003239e-05, "loss": 0.0344, "step": 29560 }, { "grad_norm": 0.544036865234375, "learning_rate": 7.457699274292493e-05, "loss": 0.0462, "step": 29570 }, { "grad_norm": 0.3101796507835388, "learning_rate": 7.455899148628159e-05, "loss": 0.0466, "step": 29580 }, { "grad_norm": 0.35787472128868103, "learning_rate": 7.45409860331783e-05, "loss": 0.048, "step": 29590 }, { "grad_norm": 0.4696907103061676, "learning_rate": 7.452297638669169e-05, "loss": 0.0378, "step": 29600 }, { "grad_norm": 0.5409725904464722, "learning_rate": 7.450496254989911e-05, "loss": 0.0353, "step": 29610 }, { "grad_norm": 0.3701360821723938, "learning_rate": 7.448694452587866e-05, "loss": 0.047, "step": 29620 }, { "grad_norm": 0.42457422614097595, "learning_rate": 7.44689223177091e-05, "loss": 0.0547, "step": 29630 }, { "grad_norm": 0.40415263175964355, "learning_rate": 7.445089592846994e-05, "loss": 0.0418, "step": 29640 }, { "grad_norm": 0.3571165204048157, "learning_rate": 7.443286536124141e-05, "loss": 0.0435, "step": 29650 }, { "grad_norm": 0.4294906258583069, "learning_rate": 7.441483061910443e-05, "loss": 0.0486, "step": 29660 }, { "grad_norm": 0.5263157486915588, "learning_rate": 7.439679170514064e-05, "loss": 0.0481, "step": 29670 }, { "grad_norm": 0.2646380066871643, "learning_rate": 7.43787486224324e-05, "loss": 0.0443, "step": 29680 }, { "grad_norm": 0.5563288927078247, "learning_rate": 7.436070137406276e-05, "loss": 0.0406, "step": 29690 }, { "grad_norm": 0.45317402482032776, "learning_rate": 7.434264996311556e-05, "loss": 0.0389, "step": 29700 }, { "grad_norm": 0.5261595249176025, "learning_rate": 7.432459439267525e-05, "loss": 0.041, "step": 29710 }, { "grad_norm": 0.30971238017082214, "learning_rate": 7.430653466582701e-05, "loss": 0.0401, "step": 29720 }, { "grad_norm": 0.32310864329338074, "learning_rate": 7.42884707856568e-05, "loss": 0.0492, "step": 29730 }, { "grad_norm": 0.4130401313304901, "learning_rate": 7.427040275525122e-05, "loss": 0.0412, "step": 29740 }, { "grad_norm": 0.4479958713054657, "learning_rate": 7.42523305776976e-05, "loss": 0.0465, "step": 29750 }, { "grad_norm": 0.34378984570503235, "learning_rate": 7.4234254256084e-05, "loss": 0.0389, "step": 29760 }, { "grad_norm": 0.3603690266609192, "learning_rate": 7.421617379349915e-05, "loss": 0.0513, "step": 29770 }, { "grad_norm": 0.26406630873680115, "learning_rate": 7.41980891930325e-05, "loss": 0.0385, "step": 29780 }, { "grad_norm": 0.28655001521110535, "learning_rate": 7.418000045777425e-05, "loss": 0.0325, "step": 29790 }, { "grad_norm": 0.44574856758117676, "learning_rate": 7.416190759081523e-05, "loss": 0.0439, "step": 29800 }, { "grad_norm": 0.4105384349822998, "learning_rate": 7.414381059524704e-05, "loss": 0.0424, "step": 29810 }, { "grad_norm": 0.4575483500957489, "learning_rate": 7.412570947416195e-05, "loss": 0.0401, "step": 29820 }, { "grad_norm": 0.5189444422721863, "learning_rate": 7.410760423065295e-05, "loss": 0.053, "step": 29830 }, { "grad_norm": 0.3665393590927124, "learning_rate": 7.408949486781372e-05, "loss": 0.0519, "step": 29840 }, { "grad_norm": 0.3982275128364563, "learning_rate": 7.407138138873868e-05, "loss": 0.0439, "step": 29850 }, { "grad_norm": 0.4676406979560852, "learning_rate": 7.405326379652292e-05, "loss": 0.0449, "step": 29860 }, { "grad_norm": 0.3771146535873413, "learning_rate": 7.403514209426222e-05, "loss": 0.0425, "step": 29870 }, { "grad_norm": 0.48667412996292114, "learning_rate": 7.40170162850531e-05, "loss": 0.047, "step": 29880 }, { "grad_norm": 0.4435019791126251, "learning_rate": 7.399888637199278e-05, "loss": 0.0447, "step": 29890 }, { "grad_norm": 0.42442426085472107, "learning_rate": 7.398075235817914e-05, "loss": 0.0505, "step": 29900 }, { "grad_norm": 0.402201384305954, "learning_rate": 7.39626142467108e-05, "loss": 0.0468, "step": 29910 }, { "grad_norm": 0.4259071946144104, "learning_rate": 7.394447204068706e-05, "loss": 0.0479, "step": 29920 }, { "grad_norm": 0.38278481364250183, "learning_rate": 7.392632574320793e-05, "loss": 0.0513, "step": 29930 }, { "grad_norm": 0.4275813102722168, "learning_rate": 7.390817535737411e-05, "loss": 0.0431, "step": 29940 }, { "grad_norm": 0.4088529944419861, "learning_rate": 7.389002088628703e-05, "loss": 0.0418, "step": 29950 }, { "grad_norm": 0.3976821303367615, "learning_rate": 7.387186233304877e-05, "loss": 0.047, "step": 29960 }, { "grad_norm": 0.4346614181995392, "learning_rate": 7.385369970076212e-05, "loss": 0.0418, "step": 29970 }, { "grad_norm": 0.3375232219696045, "learning_rate": 7.38355329925306e-05, "loss": 0.0501, "step": 29980 }, { "grad_norm": 0.3910380005836487, "learning_rate": 7.381736221145838e-05, "loss": 0.0454, "step": 29990 }, { "grad_norm": 0.40550222992897034, "learning_rate": 7.37991873606504e-05, "loss": 0.0335, "step": 30000 }, { "grad_norm": 0.47976362705230713, "learning_rate": 7.378100844321218e-05, "loss": 0.0468, "step": 30010 }, { "grad_norm": 0.37926793098449707, "learning_rate": 7.376282546225004e-05, "loss": 0.0432, "step": 30020 }, { "grad_norm": 0.5288286805152893, "learning_rate": 7.374463842087094e-05, "loss": 0.038, "step": 30030 }, { "grad_norm": 0.5246590971946716, "learning_rate": 7.372644732218254e-05, "loss": 0.0443, "step": 30040 }, { "grad_norm": 0.44598865509033203, "learning_rate": 7.370825216929322e-05, "loss": 0.051, "step": 30050 }, { "grad_norm": 0.33865514397621155, "learning_rate": 7.369005296531205e-05, "loss": 0.0495, "step": 30060 }, { "grad_norm": 0.3408973813056946, "learning_rate": 7.367184971334873e-05, "loss": 0.0435, "step": 30070 }, { "grad_norm": 0.4439693093299866, "learning_rate": 7.365364241651371e-05, "loss": 0.0398, "step": 30080 }, { "grad_norm": 0.49853119254112244, "learning_rate": 7.363543107791815e-05, "loss": 0.0374, "step": 30090 }, { "grad_norm": 0.4731699526309967, "learning_rate": 7.361721570067384e-05, "loss": 0.0412, "step": 30100 }, { "grad_norm": 0.40416210889816284, "learning_rate": 7.359899628789331e-05, "loss": 0.0497, "step": 30110 }, { "grad_norm": 0.6756773591041565, "learning_rate": 7.358077284268974e-05, "loss": 0.05, "step": 30120 }, { "grad_norm": 0.4344578683376312, "learning_rate": 7.356254536817702e-05, "loss": 0.0391, "step": 30130 }, { "grad_norm": 0.4256182014942169, "learning_rate": 7.354431386746973e-05, "loss": 0.0421, "step": 30140 }, { "grad_norm": 0.39064323902130127, "learning_rate": 7.352607834368316e-05, "loss": 0.0356, "step": 30150 }, { "grad_norm": 0.4514285922050476, "learning_rate": 7.350783879993324e-05, "loss": 0.04, "step": 30160 }, { "grad_norm": 0.3504771292209625, "learning_rate": 7.348959523933658e-05, "loss": 0.0394, "step": 30170 }, { "grad_norm": 0.47570109367370605, "learning_rate": 7.347134766501057e-05, "loss": 0.0403, "step": 30180 }, { "grad_norm": 0.4697554111480713, "learning_rate": 7.345309608007315e-05, "loss": 0.0369, "step": 30190 }, { "grad_norm": 0.3663078546524048, "learning_rate": 7.343484048764308e-05, "loss": 0.0461, "step": 30200 }, { "grad_norm": 0.33584296703338623, "learning_rate": 7.341658089083972e-05, "loss": 0.0452, "step": 30210 }, { "grad_norm": 0.44357797503471375, "learning_rate": 7.339831729278313e-05, "loss": 0.0376, "step": 30220 }, { "grad_norm": 0.40831324458122253, "learning_rate": 7.338004969659404e-05, "loss": 0.0386, "step": 30230 }, { "grad_norm": 0.48238956928253174, "learning_rate": 7.336177810539391e-05, "loss": 0.0495, "step": 30240 }, { "grad_norm": 0.4675774574279785, "learning_rate": 7.334350252230485e-05, "loss": 0.0372, "step": 30250 }, { "grad_norm": 0.4099578857421875, "learning_rate": 7.332522295044965e-05, "loss": 0.0354, "step": 30260 }, { "grad_norm": 0.45457756519317627, "learning_rate": 7.33069393929518e-05, "loss": 0.0445, "step": 30270 }, { "grad_norm": 0.5578832030296326, "learning_rate": 7.328865185293545e-05, "loss": 0.0545, "step": 30280 }, { "grad_norm": 0.4127418100833893, "learning_rate": 7.327036033352546e-05, "loss": 0.0455, "step": 30290 }, { "grad_norm": 0.5001524090766907, "learning_rate": 7.325206483784733e-05, "loss": 0.0465, "step": 30300 }, { "grad_norm": 0.3567120432853699, "learning_rate": 7.323376536902724e-05, "loss": 0.0555, "step": 30310 }, { "grad_norm": 0.514005720615387, "learning_rate": 7.321546193019213e-05, "loss": 0.0453, "step": 30320 }, { "grad_norm": 0.5101946592330933, "learning_rate": 7.31971545244695e-05, "loss": 0.0459, "step": 30330 }, { "grad_norm": 0.3014160394668579, "learning_rate": 7.31788431549876e-05, "loss": 0.0464, "step": 30340 }, { "grad_norm": 0.49192893505096436, "learning_rate": 7.316052782487534e-05, "loss": 0.0409, "step": 30350 }, { "grad_norm": 0.36506855487823486, "learning_rate": 7.314220853726234e-05, "loss": 0.0493, "step": 30360 }, { "grad_norm": 0.5237746238708496, "learning_rate": 7.312388529527884e-05, "loss": 0.0461, "step": 30370 }, { "grad_norm": 0.40740737318992615, "learning_rate": 7.310555810205577e-05, "loss": 0.0454, "step": 30380 }, { "grad_norm": 0.5397483706474304, "learning_rate": 7.308722696072476e-05, "loss": 0.0521, "step": 30390 }, { "grad_norm": 0.4655739367008209, "learning_rate": 7.306889187441811e-05, "loss": 0.0439, "step": 30400 }, { "grad_norm": 0.35746869444847107, "learning_rate": 7.305055284626876e-05, "loss": 0.0331, "step": 30410 }, { "grad_norm": 0.3780776858329773, "learning_rate": 7.303220987941037e-05, "loss": 0.0476, "step": 30420 }, { "grad_norm": 0.27663907408714294, "learning_rate": 7.301386297697726e-05, "loss": 0.0436, "step": 30430 }, { "grad_norm": 0.45792970061302185, "learning_rate": 7.299551214210438e-05, "loss": 0.0448, "step": 30440 }, { "grad_norm": 0.36742132902145386, "learning_rate": 7.297715737792738e-05, "loss": 0.0399, "step": 30450 }, { "grad_norm": 0.4272221028804779, "learning_rate": 7.295879868758265e-05, "loss": 0.0376, "step": 30460 }, { "grad_norm": 0.4652102291584015, "learning_rate": 7.294043607420713e-05, "loss": 0.0394, "step": 30470 }, { "grad_norm": 0.4037248194217682, "learning_rate": 7.292206954093852e-05, "loss": 0.0365, "step": 30480 }, { "grad_norm": 0.46494677662849426, "learning_rate": 7.290369909091515e-05, "loss": 0.0419, "step": 30490 }, { "grad_norm": 0.5768139362335205, "learning_rate": 7.2885324727276e-05, "loss": 0.0438, "step": 30500 }, { "grad_norm": 0.35653984546661377, "learning_rate": 7.286694645316076e-05, "loss": 0.0443, "step": 30510 }, { "grad_norm": 0.40811166167259216, "learning_rate": 7.284856427170982e-05, "loss": 0.0424, "step": 30520 }, { "grad_norm": 0.3047514259815216, "learning_rate": 7.283017818606414e-05, "loss": 0.0445, "step": 30530 }, { "grad_norm": 0.41844692826271057, "learning_rate": 7.28117881993654e-05, "loss": 0.0453, "step": 30540 }, { "grad_norm": 0.44804468750953674, "learning_rate": 7.279339431475598e-05, "loss": 0.0413, "step": 30550 }, { "grad_norm": 0.2817738950252533, "learning_rate": 7.277499653537887e-05, "loss": 0.0411, "step": 30560 }, { "grad_norm": 0.5230221152305603, "learning_rate": 7.275659486437776e-05, "loss": 0.0464, "step": 30570 }, { "grad_norm": 0.38444992899894714, "learning_rate": 7.273818930489695e-05, "loss": 0.0458, "step": 30580 }, { "grad_norm": 0.4862200617790222, "learning_rate": 7.271977986008151e-05, "loss": 0.0427, "step": 30590 }, { "grad_norm": 0.4340429902076721, "learning_rate": 7.270136653307705e-05, "loss": 0.0383, "step": 30600 }, { "grad_norm": 0.34505927562713623, "learning_rate": 7.268294932702994e-05, "loss": 0.0397, "step": 30610 }, { "grad_norm": 0.3501929044723511, "learning_rate": 7.266452824508719e-05, "loss": 0.039, "step": 30620 }, { "grad_norm": 0.44004660844802856, "learning_rate": 7.264610329039643e-05, "loss": 0.0386, "step": 30630 }, { "grad_norm": 0.3207496404647827, "learning_rate": 7.262767446610599e-05, "loss": 0.0347, "step": 30640 }, { "grad_norm": 0.681224524974823, "learning_rate": 7.260924177536485e-05, "loss": 0.0457, "step": 30650 }, { "grad_norm": 0.5444517731666565, "learning_rate": 7.259080522132265e-05, "loss": 0.042, "step": 30660 }, { "grad_norm": 0.4015241265296936, "learning_rate": 7.257236480712972e-05, "loss": 0.0341, "step": 30670 }, { "grad_norm": 0.4494757354259491, "learning_rate": 7.255392053593697e-05, "loss": 0.038, "step": 30680 }, { "grad_norm": 0.4832625985145569, "learning_rate": 7.253547241089607e-05, "loss": 0.0368, "step": 30690 }, { "grad_norm": 0.3271884024143219, "learning_rate": 7.251702043515927e-05, "loss": 0.0369, "step": 30700 }, { "grad_norm": 0.5940749049186707, "learning_rate": 7.249856461187952e-05, "loss": 0.0399, "step": 30710 }, { "grad_norm": 0.4275551736354828, "learning_rate": 7.248010494421042e-05, "loss": 0.0461, "step": 30720 }, { "grad_norm": 0.3808910548686981, "learning_rate": 7.246164143530622e-05, "loss": 0.0626, "step": 30730 }, { "grad_norm": 0.4156351387500763, "learning_rate": 7.244317408832181e-05, "loss": 0.0422, "step": 30740 }, { "grad_norm": 0.5681973099708557, "learning_rate": 7.242470290641279e-05, "loss": 0.0386, "step": 30750 }, { "grad_norm": 0.603255569934845, "learning_rate": 7.240622789273536e-05, "loss": 0.046, "step": 30760 }, { "grad_norm": 0.4638485908508301, "learning_rate": 7.238774905044638e-05, "loss": 0.0501, "step": 30770 }, { "grad_norm": 0.33746057748794556, "learning_rate": 7.236926638270341e-05, "loss": 0.0469, "step": 30780 }, { "grad_norm": 0.38814085721969604, "learning_rate": 7.23507798926646e-05, "loss": 0.0403, "step": 30790 }, { "grad_norm": 0.4586164951324463, "learning_rate": 7.23322895834888e-05, "loss": 0.0482, "step": 30800 }, { "grad_norm": 0.37817254662513733, "learning_rate": 7.231379545833552e-05, "loss": 0.0448, "step": 30810 }, { "grad_norm": 0.3487911820411682, "learning_rate": 7.229529752036487e-05, "loss": 0.0381, "step": 30820 }, { "grad_norm": 0.4301575720310211, "learning_rate": 7.227679577273765e-05, "loss": 0.0441, "step": 30830 }, { "grad_norm": 0.46005579829216003, "learning_rate": 7.225829021861529e-05, "loss": 0.0507, "step": 30840 }, { "grad_norm": 0.32842177152633667, "learning_rate": 7.223978086115992e-05, "loss": 0.0369, "step": 30850 }, { "grad_norm": 0.404079794883728, "learning_rate": 7.222126770353425e-05, "loss": 0.0445, "step": 30860 }, { "grad_norm": 0.5745561122894287, "learning_rate": 7.22027507489017e-05, "loss": 0.0401, "step": 30870 }, { "grad_norm": 0.3542007505893707, "learning_rate": 7.218423000042627e-05, "loss": 0.0375, "step": 30880 }, { "grad_norm": 0.43655306100845337, "learning_rate": 7.216570546127268e-05, "loss": 0.044, "step": 30890 }, { "grad_norm": 0.3379457890987396, "learning_rate": 7.214717713460626e-05, "loss": 0.0482, "step": 30900 }, { "grad_norm": 0.3675794005393982, "learning_rate": 7.2128645023593e-05, "loss": 0.0439, "step": 30910 }, { "grad_norm": 0.4131506383419037, "learning_rate": 7.211010913139951e-05, "loss": 0.0473, "step": 30920 }, { "grad_norm": 0.43342694640159607, "learning_rate": 7.209156946119308e-05, "loss": 0.0396, "step": 30930 }, { "grad_norm": 0.4935196340084076, "learning_rate": 7.207302601614166e-05, "loss": 0.0388, "step": 30940 }, { "grad_norm": 0.4415241479873657, "learning_rate": 7.205447879941378e-05, "loss": 0.0381, "step": 30950 }, { "grad_norm": 0.4136928617954254, "learning_rate": 7.203592781417866e-05, "loss": 0.0437, "step": 30960 }, { "grad_norm": 0.49105513095855713, "learning_rate": 7.201737306360617e-05, "loss": 0.0411, "step": 30970 }, { "grad_norm": 0.40522173047065735, "learning_rate": 7.19988145508668e-05, "loss": 0.0426, "step": 30980 }, { "grad_norm": 0.3477037847042084, "learning_rate": 7.198025227913168e-05, "loss": 0.037, "step": 30990 }, { "grad_norm": 0.4982511103153229, "learning_rate": 7.196168625157261e-05, "loss": 0.0425, "step": 31000 }, { "grad_norm": 0.43229344487190247, "learning_rate": 7.194311647136201e-05, "loss": 0.0451, "step": 31010 }, { "grad_norm": 0.5047707557678223, "learning_rate": 7.192454294167297e-05, "loss": 0.0393, "step": 31020 }, { "grad_norm": 0.5917832255363464, "learning_rate": 7.190596566567917e-05, "loss": 0.0384, "step": 31030 }, { "grad_norm": 0.5325729846954346, "learning_rate": 7.188738464655496e-05, "loss": 0.0426, "step": 31040 }, { "grad_norm": 0.3671446144580841, "learning_rate": 7.186879988747533e-05, "loss": 0.0303, "step": 31050 }, { "grad_norm": 0.3757465183734894, "learning_rate": 7.185021139161592e-05, "loss": 0.036, "step": 31060 }, { "grad_norm": 0.3629359304904938, "learning_rate": 7.1831619162153e-05, "loss": 0.0427, "step": 31070 }, { "grad_norm": 0.39293980598449707, "learning_rate": 7.181302320226345e-05, "loss": 0.0372, "step": 31080 }, { "grad_norm": 0.3480146825313568, "learning_rate": 7.179442351512482e-05, "loss": 0.0398, "step": 31090 }, { "grad_norm": 0.4259927570819855, "learning_rate": 7.177582010391528e-05, "loss": 0.0357, "step": 31100 }, { "grad_norm": 0.451811820268631, "learning_rate": 7.175721297181366e-05, "loss": 0.0452, "step": 31110 }, { "grad_norm": 0.4229372441768646, "learning_rate": 7.173860212199942e-05, "loss": 0.043, "step": 31120 }, { "grad_norm": 0.5617655515670776, "learning_rate": 7.171998755765263e-05, "loss": 0.041, "step": 31130 }, { "grad_norm": 0.38777777552604675, "learning_rate": 7.170136928195398e-05, "loss": 0.0371, "step": 31140 }, { "grad_norm": 0.30196335911750793, "learning_rate": 7.168274729808489e-05, "loss": 0.0329, "step": 31150 }, { "grad_norm": 0.33689141273498535, "learning_rate": 7.166412160922728e-05, "loss": 0.0335, "step": 31160 }, { "grad_norm": 0.48054245114326477, "learning_rate": 7.164549221856382e-05, "loss": 0.0344, "step": 31170 }, { "grad_norm": 0.35339996218681335, "learning_rate": 7.162685912927775e-05, "loss": 0.0504, "step": 31180 }, { "grad_norm": 0.4079221189022064, "learning_rate": 7.160822234455294e-05, "loss": 0.036, "step": 31190 }, { "grad_norm": 0.3007989525794983, "learning_rate": 7.158958186757391e-05, "loss": 0.0357, "step": 31200 }, { "grad_norm": 0.3146354556083679, "learning_rate": 7.157093770152582e-05, "loss": 0.0306, "step": 31210 }, { "grad_norm": 0.43023058772087097, "learning_rate": 7.155228984959446e-05, "loss": 0.033, "step": 31220 }, { "grad_norm": 0.26073306798934937, "learning_rate": 7.153363831496621e-05, "loss": 0.0395, "step": 31230 }, { "grad_norm": 0.39791133999824524, "learning_rate": 7.151498310082811e-05, "loss": 0.0494, "step": 31240 }, { "grad_norm": 0.3682907223701477, "learning_rate": 7.149632421036784e-05, "loss": 0.0422, "step": 31250 }, { "grad_norm": 0.47510573267936707, "learning_rate": 7.147766164677369e-05, "loss": 0.0694, "step": 31260 }, { "grad_norm": 0.3546130061149597, "learning_rate": 7.145899541323459e-05, "loss": 0.0344, "step": 31270 }, { "grad_norm": 0.29650992155075073, "learning_rate": 7.144032551294007e-05, "loss": 0.0431, "step": 31280 }, { "grad_norm": 0.6103703379631042, "learning_rate": 7.14216519490803e-05, "loss": 0.0416, "step": 31290 }, { "grad_norm": 0.5398656725883484, "learning_rate": 7.140297472484609e-05, "loss": 0.0504, "step": 31300 }, { "grad_norm": 0.4620329737663269, "learning_rate": 7.138429384342891e-05, "loss": 0.0615, "step": 31310 }, { "grad_norm": 0.3572627902030945, "learning_rate": 7.136560930802074e-05, "loss": 0.044, "step": 31320 }, { "grad_norm": 0.4210347533226013, "learning_rate": 7.134692112181431e-05, "loss": 0.036, "step": 31330 }, { "grad_norm": 0.38510382175445557, "learning_rate": 7.13282292880029e-05, "loss": 0.041, "step": 31340 }, { "grad_norm": 0.4436907470226288, "learning_rate": 7.130953380978043e-05, "loss": 0.0385, "step": 31350 }, { "grad_norm": 0.4937743544578552, "learning_rate": 7.129083469034144e-05, "loss": 0.0496, "step": 31360 }, { "grad_norm": 0.43058809638023376, "learning_rate": 7.127213193288112e-05, "loss": 0.0418, "step": 31370 }, { "grad_norm": 0.36356860399246216, "learning_rate": 7.125342554059522e-05, "loss": 0.0379, "step": 31380 }, { "grad_norm": 0.40259668231010437, "learning_rate": 7.12347155166802e-05, "loss": 0.0403, "step": 31390 }, { "grad_norm": 0.3975936770439148, "learning_rate": 7.121600186433306e-05, "loss": 0.04, "step": 31400 }, { "grad_norm": 0.32033345103263855, "learning_rate": 7.119728458675148e-05, "loss": 0.0377, "step": 31410 }, { "grad_norm": 0.3372529149055481, "learning_rate": 7.117856368713369e-05, "loss": 0.0417, "step": 31420 }, { "grad_norm": 0.38300973176956177, "learning_rate": 7.115983916867861e-05, "loss": 0.0393, "step": 31430 }, { "grad_norm": 0.8841261267662048, "learning_rate": 7.114111103458574e-05, "loss": 0.042, "step": 31440 }, { "grad_norm": 0.4645504951477051, "learning_rate": 7.11223792880552e-05, "loss": 0.0443, "step": 31450 }, { "grad_norm": 0.34838414192199707, "learning_rate": 7.110364393228773e-05, "loss": 0.0447, "step": 31460 }, { "grad_norm": 0.351236492395401, "learning_rate": 7.108490497048471e-05, "loss": 0.038, "step": 31470 }, { "grad_norm": 0.40096738934516907, "learning_rate": 7.10661624058481e-05, "loss": 0.0496, "step": 31480 }, { "grad_norm": 0.4101569950580597, "learning_rate": 7.10474162415805e-05, "loss": 0.0445, "step": 31490 }, { "grad_norm": 0.3454856276512146, "learning_rate": 7.102866648088511e-05, "loss": 0.0496, "step": 31500 }, { "grad_norm": 0.48830628395080566, "learning_rate": 7.100991312696576e-05, "loss": 0.0362, "step": 31510 }, { "grad_norm": 0.4029625654220581, "learning_rate": 7.099115618302686e-05, "loss": 0.0364, "step": 31520 }, { "grad_norm": 0.2848798930644989, "learning_rate": 7.097239565227349e-05, "loss": 0.0445, "step": 31530 }, { "grad_norm": 0.3355681896209717, "learning_rate": 7.09536315379113e-05, "loss": 0.031, "step": 31540 }, { "grad_norm": 0.33386659622192383, "learning_rate": 7.093486384314656e-05, "loss": 0.0378, "step": 31550 }, { "grad_norm": 0.2955378293991089, "learning_rate": 7.091609257118616e-05, "loss": 0.0383, "step": 31560 }, { "grad_norm": 0.48387640714645386, "learning_rate": 7.08973177252376e-05, "loss": 0.0447, "step": 31570 }, { "grad_norm": 0.3009887635707855, "learning_rate": 7.087853930850898e-05, "loss": 0.0365, "step": 31580 }, { "grad_norm": 0.3585432767868042, "learning_rate": 7.085975732420903e-05, "loss": 0.0416, "step": 31590 }, { "grad_norm": 0.36207282543182373, "learning_rate": 7.084097177554706e-05, "loss": 0.0323, "step": 31600 }, { "grad_norm": 0.49038073420524597, "learning_rate": 7.082218266573301e-05, "loss": 0.0477, "step": 31610 }, { "grad_norm": 0.4637976288795471, "learning_rate": 7.080338999797743e-05, "loss": 0.0461, "step": 31620 }, { "grad_norm": 0.3642628788948059, "learning_rate": 7.07845937754915e-05, "loss": 0.0374, "step": 31630 }, { "grad_norm": 0.446112722158432, "learning_rate": 7.076579400148693e-05, "loss": 0.0402, "step": 31640 }, { "grad_norm": 0.446257621049881, "learning_rate": 7.074699067917611e-05, "loss": 0.0447, "step": 31650 }, { "grad_norm": 0.5344151258468628, "learning_rate": 7.072818381177201e-05, "loss": 0.0394, "step": 31660 }, { "grad_norm": 0.6001952290534973, "learning_rate": 7.070937340248823e-05, "loss": 0.0453, "step": 31670 }, { "grad_norm": 0.3829023540019989, "learning_rate": 7.069055945453893e-05, "loss": 0.0422, "step": 31680 }, { "grad_norm": 0.744693398475647, "learning_rate": 7.067174197113892e-05, "loss": 0.0366, "step": 31690 }, { "grad_norm": 0.34783926606178284, "learning_rate": 7.065292095550355e-05, "loss": 0.0379, "step": 31700 }, { "grad_norm": 0.36916735768318176, "learning_rate": 7.063409641084887e-05, "loss": 0.0416, "step": 31710 }, { "grad_norm": 0.4012898802757263, "learning_rate": 7.061526834039145e-05, "loss": 0.0495, "step": 31720 }, { "grad_norm": 0.468414843082428, "learning_rate": 7.05964367473485e-05, "loss": 0.0349, "step": 31730 }, { "grad_norm": 0.3304891288280487, "learning_rate": 7.057760163493783e-05, "loss": 0.0452, "step": 31740 }, { "grad_norm": 0.3379177451133728, "learning_rate": 7.055876300637783e-05, "loss": 0.0367, "step": 31750 }, { "grad_norm": 0.33040881156921387, "learning_rate": 7.053992086488753e-05, "loss": 0.0391, "step": 31760 }, { "grad_norm": 0.36912938952445984, "learning_rate": 7.052107521368651e-05, "loss": 0.0455, "step": 31770 }, { "grad_norm": 0.3898616433143616, "learning_rate": 7.0502226055995e-05, "loss": 0.0453, "step": 31780 }, { "grad_norm": 0.3277739882469177, "learning_rate": 7.048337339503379e-05, "loss": 0.0467, "step": 31790 }, { "grad_norm": 0.40001624822616577, "learning_rate": 7.046451723402427e-05, "loss": 0.0364, "step": 31800 }, { "grad_norm": 0.4591866135597229, "learning_rate": 7.044565757618848e-05, "loss": 0.0442, "step": 31810 }, { "grad_norm": 0.3437979817390442, "learning_rate": 7.042679442474899e-05, "loss": 0.0404, "step": 31820 }, { "grad_norm": 0.4000855088233948, "learning_rate": 7.040792778292902e-05, "loss": 0.0499, "step": 31830 }, { "grad_norm": 0.28989094495773315, "learning_rate": 7.038905765395234e-05, "loss": 0.0488, "step": 31840 }, { "grad_norm": 0.507162868976593, "learning_rate": 7.037018404104334e-05, "loss": 0.0498, "step": 31850 }, { "grad_norm": 0.32225358486175537, "learning_rate": 7.035130694742702e-05, "loss": 0.0346, "step": 31860 }, { "grad_norm": 0.44038259983062744, "learning_rate": 7.033242637632897e-05, "loss": 0.0477, "step": 31870 }, { "grad_norm": 0.4679635763168335, "learning_rate": 7.031354233097534e-05, "loss": 0.04, "step": 31880 }, { "grad_norm": 0.3553755581378937, "learning_rate": 7.029465481459289e-05, "loss": 0.0463, "step": 31890 }, { "grad_norm": 0.3323988616466522, "learning_rate": 7.027576383040898e-05, "loss": 0.0372, "step": 31900 }, { "grad_norm": 0.38754987716674805, "learning_rate": 7.025686938165159e-05, "loss": 0.0488, "step": 31910 }, { "grad_norm": 0.45454302430152893, "learning_rate": 7.023797147154924e-05, "loss": 0.0539, "step": 31920 }, { "grad_norm": 0.30358654260635376, "learning_rate": 7.021907010333111e-05, "loss": 0.0353, "step": 31930 }, { "grad_norm": 0.35454973578453064, "learning_rate": 7.020016528022685e-05, "loss": 0.0379, "step": 31940 }, { "grad_norm": 0.4319273829460144, "learning_rate": 7.018125700546683e-05, "loss": 0.046, "step": 31950 }, { "grad_norm": 0.3772251605987549, "learning_rate": 7.016234528228196e-05, "loss": 0.0429, "step": 31960 }, { "grad_norm": 0.4213131368160248, "learning_rate": 7.014343011390372e-05, "loss": 0.0378, "step": 31970 }, { "grad_norm": 0.34635287523269653, "learning_rate": 7.01245115035642e-05, "loss": 0.0323, "step": 31980 }, { "grad_norm": 0.3615938127040863, "learning_rate": 7.010558945449606e-05, "loss": 0.0534, "step": 31990 }, { "grad_norm": 0.48499760031700134, "learning_rate": 7.008666396993258e-05, "loss": 0.0324, "step": 32000 }, { "grad_norm": 0.43836769461631775, "learning_rate": 7.006773505310759e-05, "loss": 0.0487, "step": 32010 }, { "grad_norm": 0.536651611328125, "learning_rate": 7.004880270725553e-05, "loss": 0.0387, "step": 32020 }, { "grad_norm": 0.3943322002887726, "learning_rate": 7.002986693561144e-05, "loss": 0.0397, "step": 32030 }, { "grad_norm": 0.37889766693115234, "learning_rate": 7.001092774141089e-05, "loss": 0.0405, "step": 32040 }, { "grad_norm": 0.29185086488723755, "learning_rate": 6.999198512789009e-05, "loss": 0.0381, "step": 32050 }, { "grad_norm": 0.46487709879875183, "learning_rate": 6.997303909828584e-05, "loss": 0.0508, "step": 32060 }, { "grad_norm": 0.5402994155883789, "learning_rate": 6.995408965583544e-05, "loss": 0.042, "step": 32070 }, { "grad_norm": 0.6545975804328918, "learning_rate": 6.993513680377688e-05, "loss": 0.0315, "step": 32080 }, { "grad_norm": 0.43128910660743713, "learning_rate": 6.991618054534868e-05, "loss": 0.0447, "step": 32090 }, { "grad_norm": 0.35669568181037903, "learning_rate": 6.989722088378991e-05, "loss": 0.0358, "step": 32100 }, { "grad_norm": 0.36619994044303894, "learning_rate": 6.987825782234027e-05, "loss": 0.0332, "step": 32110 }, { "grad_norm": 0.3046262264251709, "learning_rate": 6.985929136424006e-05, "loss": 0.0348, "step": 32120 }, { "grad_norm": 0.34300366044044495, "learning_rate": 6.984032151273012e-05, "loss": 0.0413, "step": 32130 }, { "grad_norm": 0.41316330432891846, "learning_rate": 6.982134827105186e-05, "loss": 0.0394, "step": 32140 }, { "grad_norm": 0.6207972764968872, "learning_rate": 6.980237164244729e-05, "loss": 0.042, "step": 32150 }, { "grad_norm": 0.40209102630615234, "learning_rate": 6.9783391630159e-05, "loss": 0.0323, "step": 32160 }, { "grad_norm": 0.3755539059638977, "learning_rate": 6.976440823743015e-05, "loss": 0.0472, "step": 32170 }, { "grad_norm": 0.26783624291419983, "learning_rate": 6.974542146750451e-05, "loss": 0.0329, "step": 32180 }, { "grad_norm": 0.3729046881198883, "learning_rate": 6.972643132362637e-05, "loss": 0.0298, "step": 32190 }, { "grad_norm": 0.20057173073291779, "learning_rate": 6.970743780904064e-05, "loss": 0.0412, "step": 32200 }, { "grad_norm": 0.4116031229496002, "learning_rate": 6.968844092699277e-05, "loss": 0.0449, "step": 32210 }, { "grad_norm": 0.3799420893192291, "learning_rate": 6.966944068072883e-05, "loss": 0.0427, "step": 32220 }, { "grad_norm": 0.33766499161720276, "learning_rate": 6.965043707349545e-05, "loss": 0.0331, "step": 32230 }, { "grad_norm": 0.3838205933570862, "learning_rate": 6.963143010853982e-05, "loss": 0.0465, "step": 32240 }, { "grad_norm": 0.4788450300693512, "learning_rate": 6.961241978910971e-05, "loss": 0.0412, "step": 32250 }, { "grad_norm": 0.40327686071395874, "learning_rate": 6.959340611845344e-05, "loss": 0.0359, "step": 32260 }, { "grad_norm": 0.37200453877449036, "learning_rate": 6.957438909981995e-05, "loss": 0.0336, "step": 32270 }, { "grad_norm": 0.2540236711502075, "learning_rate": 6.955536873645872e-05, "loss": 0.031, "step": 32280 }, { "grad_norm": 0.35806843638420105, "learning_rate": 6.953634503161982e-05, "loss": 0.0364, "step": 32290 }, { "grad_norm": 0.3504107892513275, "learning_rate": 6.951731798855387e-05, "loss": 0.0389, "step": 32300 }, { "grad_norm": 0.3934171199798584, "learning_rate": 6.949828761051208e-05, "loss": 0.0458, "step": 32310 }, { "grad_norm": 0.43916380405426025, "learning_rate": 6.947925390074622e-05, "loss": 0.0374, "step": 32320 }, { "grad_norm": 0.26283368468284607, "learning_rate": 6.946021686250863e-05, "loss": 0.0307, "step": 32330 }, { "grad_norm": 0.5183061361312866, "learning_rate": 6.94411764990522e-05, "loss": 0.0383, "step": 32340 }, { "grad_norm": 0.2932320237159729, "learning_rate": 6.942213281363044e-05, "loss": 0.0341, "step": 32350 }, { "grad_norm": 0.34521809220314026, "learning_rate": 6.940308580949737e-05, "loss": 0.0368, "step": 32360 }, { "grad_norm": 0.3675234019756317, "learning_rate": 6.93840354899076e-05, "loss": 0.0346, "step": 32370 }, { "grad_norm": 0.40976637601852417, "learning_rate": 6.936498185811633e-05, "loss": 0.0349, "step": 32380 }, { "grad_norm": 0.302168071269989, "learning_rate": 6.93459249173793e-05, "loss": 0.0273, "step": 32390 }, { "grad_norm": 0.3673899471759796, "learning_rate": 6.932686467095279e-05, "loss": 0.037, "step": 32400 }, { "grad_norm": 0.4129010736942291, "learning_rate": 6.930780112209373e-05, "loss": 0.0378, "step": 32410 }, { "grad_norm": 0.5242124795913696, "learning_rate": 6.92887342740595e-05, "loss": 0.049, "step": 32420 }, { "grad_norm": 0.3563799262046814, "learning_rate": 6.926966413010816e-05, "loss": 0.0455, "step": 32430 }, { "grad_norm": 0.2793363332748413, "learning_rate": 6.925059069349824e-05, "loss": 0.0294, "step": 32440 }, { "grad_norm": 0.2895384430885315, "learning_rate": 6.923151396748886e-05, "loss": 0.038, "step": 32450 }, { "grad_norm": 0.35022875666618347, "learning_rate": 6.921243395533974e-05, "loss": 0.0315, "step": 32460 }, { "grad_norm": 0.4206695556640625, "learning_rate": 6.919335066031109e-05, "loss": 0.0426, "step": 32470 }, { "grad_norm": 0.33705809712409973, "learning_rate": 6.917426408566379e-05, "loss": 0.0404, "step": 32480 }, { "grad_norm": 0.3934139013290405, "learning_rate": 6.915517423465916e-05, "loss": 0.034, "step": 32490 }, { "grad_norm": 0.3682957589626312, "learning_rate": 6.913608111055914e-05, "loss": 0.0378, "step": 32500 }, { "grad_norm": 0.2605642080307007, "learning_rate": 6.911698471662623e-05, "loss": 0.0312, "step": 32510 }, { "grad_norm": 0.4312247335910797, "learning_rate": 6.90978850561235e-05, "loss": 0.0326, "step": 32520 }, { "grad_norm": 0.40867120027542114, "learning_rate": 6.907878213231454e-05, "loss": 0.0395, "step": 32530 }, { "grad_norm": 0.5755842328071594, "learning_rate": 6.90596759484635e-05, "loss": 0.0408, "step": 32540 }, { "grad_norm": 0.5299112796783447, "learning_rate": 6.904056650783514e-05, "loss": 0.0426, "step": 32550 }, { "grad_norm": 0.4954319894313812, "learning_rate": 6.902145381369471e-05, "loss": 0.042, "step": 32560 }, { "grad_norm": 0.3221684992313385, "learning_rate": 6.900233786930808e-05, "loss": 0.042, "step": 32570 }, { "grad_norm": 0.3908902108669281, "learning_rate": 6.898321867794161e-05, "loss": 0.0398, "step": 32580 }, { "grad_norm": 0.3976384997367859, "learning_rate": 6.896409624286226e-05, "loss": 0.0434, "step": 32590 }, { "grad_norm": 0.3929549753665924, "learning_rate": 6.894497056733754e-05, "loss": 0.0418, "step": 32600 }, { "grad_norm": 0.595285952091217, "learning_rate": 6.89258416546355e-05, "loss": 0.041, "step": 32610 }, { "grad_norm": 0.9377960562705994, "learning_rate": 6.890670950802474e-05, "loss": 0.0509, "step": 32620 }, { "grad_norm": 0.4481419026851654, "learning_rate": 6.88875741307744e-05, "loss": 0.0383, "step": 32630 }, { "grad_norm": 0.2400379478931427, "learning_rate": 6.886843552615425e-05, "loss": 0.0338, "step": 32640 }, { "grad_norm": 0.36525169014930725, "learning_rate": 6.884929369743451e-05, "loss": 0.0365, "step": 32650 }, { "grad_norm": 0.5631470084190369, "learning_rate": 6.8830148647886e-05, "loss": 0.0353, "step": 32660 }, { "grad_norm": 0.3960780203342438, "learning_rate": 6.88110003807801e-05, "loss": 0.0474, "step": 32670 }, { "grad_norm": 0.5119132399559021, "learning_rate": 6.87918488993887e-05, "loss": 0.0401, "step": 32680 }, { "grad_norm": 0.5331695079803467, "learning_rate": 6.877269420698431e-05, "loss": 0.0424, "step": 32690 }, { "grad_norm": 0.44734707474708557, "learning_rate": 6.875353630683989e-05, "loss": 0.0396, "step": 32700 }, { "grad_norm": 0.48872703313827515, "learning_rate": 6.873437520222905e-05, "loss": 0.0447, "step": 32710 }, { "grad_norm": 0.3115462064743042, "learning_rate": 6.871521089642585e-05, "loss": 0.0347, "step": 32720 }, { "grad_norm": 0.4114500880241394, "learning_rate": 6.869604339270498e-05, "loss": 0.0346, "step": 32730 }, { "grad_norm": 0.3607260584831238, "learning_rate": 6.867687269434164e-05, "loss": 0.0395, "step": 32740 }, { "grad_norm": 0.30710601806640625, "learning_rate": 6.865769880461156e-05, "loss": 0.0414, "step": 32750 }, { "grad_norm": 0.4306182265281677, "learning_rate": 6.863852172679104e-05, "loss": 0.0398, "step": 32760 }, { "grad_norm": 0.4888444244861603, "learning_rate": 6.861934146415693e-05, "loss": 0.0414, "step": 32770 }, { "grad_norm": 0.4001436233520508, "learning_rate": 6.86001580199866e-05, "loss": 0.0479, "step": 32780 }, { "grad_norm": 0.4132622480392456, "learning_rate": 6.858097139755798e-05, "loss": 0.0326, "step": 32790 }, { "grad_norm": 0.42267149686813354, "learning_rate": 6.856178160014955e-05, "loss": 0.0396, "step": 32800 }, { "grad_norm": 0.36007869243621826, "learning_rate": 6.85425886310403e-05, "loss": 0.0351, "step": 32810 }, { "grad_norm": 0.3556577265262604, "learning_rate": 6.852339249350979e-05, "loss": 0.037, "step": 32820 }, { "grad_norm": 0.3963170051574707, "learning_rate": 6.850419319083812e-05, "loss": 0.0465, "step": 32830 }, { "grad_norm": 0.32879915833473206, "learning_rate": 6.848499072630592e-05, "loss": 0.0394, "step": 32840 }, { "grad_norm": 0.4209091067314148, "learning_rate": 6.846578510319439e-05, "loss": 0.0487, "step": 32850 }, { "grad_norm": 0.4774651527404785, "learning_rate": 6.844657632478519e-05, "loss": 0.0459, "step": 32860 }, { "grad_norm": 0.3886943459510803, "learning_rate": 6.842736439436063e-05, "loss": 0.033, "step": 32870 }, { "grad_norm": 0.37979814410209656, "learning_rate": 6.84081493152035e-05, "loss": 0.0396, "step": 32880 }, { "grad_norm": 0.3586367070674896, "learning_rate": 6.83889310905971e-05, "loss": 0.0339, "step": 32890 }, { "grad_norm": 0.3588692545890808, "learning_rate": 6.836970972382533e-05, "loss": 0.045, "step": 32900 }, { "grad_norm": 0.4594184160232544, "learning_rate": 6.835048521817257e-05, "loss": 0.041, "step": 32910 }, { "grad_norm": 0.35118985176086426, "learning_rate": 6.833125757692379e-05, "loss": 0.034, "step": 32920 }, { "grad_norm": 0.3825680613517761, "learning_rate": 6.831202680336441e-05, "loss": 0.034, "step": 32930 }, { "grad_norm": 0.37648820877075195, "learning_rate": 6.829279290078052e-05, "loss": 0.0361, "step": 32940 }, { "grad_norm": 0.2329156994819641, "learning_rate": 6.827355587245863e-05, "loss": 0.0414, "step": 32950 }, { "grad_norm": 0.2180642932653427, "learning_rate": 6.82543157216858e-05, "loss": 0.0385, "step": 32960 }, { "grad_norm": 0.35115474462509155, "learning_rate": 6.823507245174969e-05, "loss": 0.0381, "step": 32970 }, { "grad_norm": 0.465416818857193, "learning_rate": 6.821582606593841e-05, "loss": 0.0452, "step": 32980 }, { "grad_norm": 0.4111056923866272, "learning_rate": 6.81965765675407e-05, "loss": 0.0391, "step": 32990 }, { "grad_norm": 0.34232667088508606, "learning_rate": 6.81773239598457e-05, "loss": 0.0392, "step": 33000 }, { "grad_norm": 0.5909772515296936, "learning_rate": 6.815806824614319e-05, "loss": 0.0478, "step": 33010 }, { "grad_norm": 0.2825952470302582, "learning_rate": 6.813880942972343e-05, "loss": 0.0343, "step": 33020 }, { "grad_norm": 0.7614328265190125, "learning_rate": 6.811954751387726e-05, "loss": 0.0485, "step": 33030 }, { "grad_norm": 0.400722861289978, "learning_rate": 6.810028250189598e-05, "loss": 0.0357, "step": 33040 }, { "grad_norm": 0.4312315881252289, "learning_rate": 6.808101439707147e-05, "loss": 0.0539, "step": 33050 }, { "grad_norm": 0.4050709307193756, "learning_rate": 6.806174320269609e-05, "loss": 0.0439, "step": 33060 }, { "grad_norm": 0.4019336402416229, "learning_rate": 6.804246892206281e-05, "loss": 0.0397, "step": 33070 }, { "grad_norm": 0.504482090473175, "learning_rate": 6.802319155846506e-05, "loss": 0.0518, "step": 33080 }, { "grad_norm": 0.40308699011802673, "learning_rate": 6.800391111519679e-05, "loss": 0.037, "step": 33090 }, { "grad_norm": 0.2884618639945984, "learning_rate": 6.798462759555253e-05, "loss": 0.0422, "step": 33100 }, { "grad_norm": 0.2595341205596924, "learning_rate": 6.79653410028273e-05, "loss": 0.0357, "step": 33110 }, { "grad_norm": 0.2444145232439041, "learning_rate": 6.794605134031663e-05, "loss": 0.0381, "step": 33120 }, { "grad_norm": 0.2624179720878601, "learning_rate": 6.792675861131661e-05, "loss": 0.0445, "step": 33130 }, { "grad_norm": 0.4926702380180359, "learning_rate": 6.790746281912386e-05, "loss": 0.0451, "step": 33140 }, { "grad_norm": 0.30962106585502625, "learning_rate": 6.788816396703546e-05, "loss": 0.029, "step": 33150 }, { "grad_norm": 0.35440969467163086, "learning_rate": 6.78688620583491e-05, "loss": 0.0329, "step": 33160 }, { "grad_norm": 0.37411192059516907, "learning_rate": 6.784955709636292e-05, "loss": 0.038, "step": 33170 }, { "grad_norm": 0.45193302631378174, "learning_rate": 6.783024908437564e-05, "loss": 0.0406, "step": 33180 }, { "grad_norm": 0.41094768047332764, "learning_rate": 6.781093802568641e-05, "loss": 0.0395, "step": 33190 }, { "grad_norm": 0.5264074802398682, "learning_rate": 6.779162392359504e-05, "loss": 0.0363, "step": 33200 }, { "grad_norm": 0.3752690255641937, "learning_rate": 6.777230678140172e-05, "loss": 0.0411, "step": 33210 }, { "grad_norm": 0.3730120360851288, "learning_rate": 6.775298660240726e-05, "loss": 0.0413, "step": 33220 }, { "grad_norm": 0.6878198385238647, "learning_rate": 6.773366338991292e-05, "loss": 0.0541, "step": 33230 }, { "grad_norm": 0.2956201136112213, "learning_rate": 6.771433714722052e-05, "loss": 0.0364, "step": 33240 }, { "grad_norm": 0.3647640347480774, "learning_rate": 6.769500787763239e-05, "loss": 0.0433, "step": 33250 }, { "grad_norm": 0.2854848802089691, "learning_rate": 6.76756755844514e-05, "loss": 0.0423, "step": 33260 }, { "grad_norm": 0.5137770175933838, "learning_rate": 6.765634027098087e-05, "loss": 0.0403, "step": 33270 }, { "grad_norm": 0.47284138202667236, "learning_rate": 6.763700194052468e-05, "loss": 0.0464, "step": 33280 }, { "grad_norm": 0.4006390869617462, "learning_rate": 6.761766059638723e-05, "loss": 0.0423, "step": 33290 }, { "grad_norm": 0.4227416217327118, "learning_rate": 6.759831624187345e-05, "loss": 0.0378, "step": 33300 }, { "grad_norm": 0.5203801393508911, "learning_rate": 6.757896888028871e-05, "loss": 0.0463, "step": 33310 }, { "grad_norm": 0.33563244342803955, "learning_rate": 6.7559618514939e-05, "loss": 0.0365, "step": 33320 }, { "grad_norm": 0.3646700084209442, "learning_rate": 6.754026514913073e-05, "loss": 0.0348, "step": 33330 }, { "grad_norm": 0.4883088171482086, "learning_rate": 6.752090878617087e-05, "loss": 0.0426, "step": 33340 }, { "grad_norm": 0.3432185649871826, "learning_rate": 6.75015494293669e-05, "loss": 0.029, "step": 33350 }, { "grad_norm": 0.3299180567264557, "learning_rate": 6.74821870820268e-05, "loss": 0.0356, "step": 33360 }, { "grad_norm": 0.37787944078445435, "learning_rate": 6.746282174745907e-05, "loss": 0.051, "step": 33370 }, { "grad_norm": 0.49886471033096313, "learning_rate": 6.744345342897271e-05, "loss": 0.0389, "step": 33380 }, { "grad_norm": 0.3779803514480591, "learning_rate": 6.742408212987724e-05, "loss": 0.0387, "step": 33390 }, { "grad_norm": 0.32226818799972534, "learning_rate": 6.740470785348269e-05, "loss": 0.0337, "step": 33400 }, { "grad_norm": 0.3943879306316376, "learning_rate": 6.738533060309958e-05, "loss": 0.0405, "step": 33410 }, { "grad_norm": 0.4512251019477844, "learning_rate": 6.736595038203894e-05, "loss": 0.0362, "step": 33420 }, { "grad_norm": 0.44267091155052185, "learning_rate": 6.734656719361236e-05, "loss": 0.0386, "step": 33430 }, { "grad_norm": 0.4320632517337799, "learning_rate": 6.732718104113189e-05, "loss": 0.046, "step": 33440 }, { "grad_norm": 0.3634827733039856, "learning_rate": 6.730779192791006e-05, "loss": 0.0307, "step": 33450 }, { "grad_norm": 0.8471387028694153, "learning_rate": 6.728839985725997e-05, "loss": 0.0446, "step": 33460 }, { "grad_norm": 0.567206621170044, "learning_rate": 6.726900483249517e-05, "loss": 0.0384, "step": 33470 }, { "grad_norm": 0.513873279094696, "learning_rate": 6.724960685692976e-05, "loss": 0.0368, "step": 33480 }, { "grad_norm": 0.3585256338119507, "learning_rate": 6.723020593387833e-05, "loss": 0.0334, "step": 33490 }, { "grad_norm": 0.3888189196586609, "learning_rate": 6.721080206665593e-05, "loss": 0.0411, "step": 33500 }, { "grad_norm": 0.2861941456794739, "learning_rate": 6.719139525857819e-05, "loss": 0.0433, "step": 33510 }, { "grad_norm": 0.3835413157939911, "learning_rate": 6.717198551296117e-05, "loss": 0.0405, "step": 33520 }, { "grad_norm": 0.38669919967651367, "learning_rate": 6.715257283312148e-05, "loss": 0.0407, "step": 33530 }, { "grad_norm": 0.4684915840625763, "learning_rate": 6.713315722237623e-05, "loss": 0.0431, "step": 33540 }, { "grad_norm": 0.3777918815612793, "learning_rate": 6.7113738684043e-05, "loss": 0.0358, "step": 33550 }, { "grad_norm": 0.3251170516014099, "learning_rate": 6.709431722143989e-05, "loss": 0.038, "step": 33560 }, { "grad_norm": 0.2144797295331955, "learning_rate": 6.70748928378855e-05, "loss": 0.0436, "step": 33570 }, { "grad_norm": 0.2992017865180969, "learning_rate": 6.705546553669891e-05, "loss": 0.0284, "step": 33580 }, { "grad_norm": 0.2746258080005646, "learning_rate": 6.703603532119974e-05, "loss": 0.0303, "step": 33590 }, { "grad_norm": 0.30368486046791077, "learning_rate": 6.701660219470808e-05, "loss": 0.0342, "step": 33600 }, { "grad_norm": 0.6311549544334412, "learning_rate": 6.69971661605445e-05, "loss": 0.0344, "step": 33610 }, { "grad_norm": 0.6042923331260681, "learning_rate": 6.697772722203008e-05, "loss": 0.0508, "step": 33620 }, { "grad_norm": 0.3406333029270172, "learning_rate": 6.695828538248643e-05, "loss": 0.041, "step": 33630 }, { "grad_norm": 0.32678723335266113, "learning_rate": 6.693884064523563e-05, "loss": 0.0401, "step": 33640 }, { "grad_norm": 0.36572566628456116, "learning_rate": 6.691939301360023e-05, "loss": 0.0361, "step": 33650 }, { "grad_norm": 0.4532778561115265, "learning_rate": 6.689994249090333e-05, "loss": 0.0369, "step": 33660 }, { "grad_norm": 0.3912407457828522, "learning_rate": 6.688048908046845e-05, "loss": 0.0439, "step": 33670 }, { "grad_norm": 0.4124993681907654, "learning_rate": 6.686103278561969e-05, "loss": 0.0295, "step": 33680 }, { "grad_norm": 0.4446219503879547, "learning_rate": 6.684157360968156e-05, "loss": 0.0353, "step": 33690 }, { "grad_norm": 0.31090930104255676, "learning_rate": 6.682211155597911e-05, "loss": 0.0299, "step": 33700 }, { "grad_norm": 0.5725067257881165, "learning_rate": 6.680264662783789e-05, "loss": 0.0413, "step": 33710 }, { "grad_norm": 0.37552133202552795, "learning_rate": 6.678317882858391e-05, "loss": 0.0433, "step": 33720 }, { "grad_norm": 0.4560695290565491, "learning_rate": 6.67637081615437e-05, "loss": 0.036, "step": 33730 }, { "grad_norm": 0.36853694915771484, "learning_rate": 6.674423463004427e-05, "loss": 0.0334, "step": 33740 }, { "grad_norm": 0.33930400013923645, "learning_rate": 6.672475823741308e-05, "loss": 0.028, "step": 33750 }, { "grad_norm": 0.34757018089294434, "learning_rate": 6.670527898697811e-05, "loss": 0.0395, "step": 33760 }, { "grad_norm": 0.4691735506057739, "learning_rate": 6.668579688206788e-05, "loss": 0.0396, "step": 33770 }, { "grad_norm": 0.5250791907310486, "learning_rate": 6.666631192601131e-05, "loss": 0.0378, "step": 33780 }, { "grad_norm": 0.4911656975746155, "learning_rate": 6.664682412213785e-05, "loss": 0.0477, "step": 33790 }, { "grad_norm": 0.3275296092033386, "learning_rate": 6.662733347377745e-05, "loss": 0.0362, "step": 33800 }, { "grad_norm": 0.18999150395393372, "learning_rate": 6.660783998426051e-05, "loss": 0.0323, "step": 33810 }, { "grad_norm": 0.4257615804672241, "learning_rate": 6.658834365691794e-05, "loss": 0.0361, "step": 33820 }, { "grad_norm": 0.5055348873138428, "learning_rate": 6.656884449508115e-05, "loss": 0.0338, "step": 33830 }, { "grad_norm": 0.21479637920856476, "learning_rate": 6.654934250208198e-05, "loss": 0.0308, "step": 33840 }, { "grad_norm": 0.3460366129875183, "learning_rate": 6.65298376812528e-05, "loss": 0.0374, "step": 33850 }, { "grad_norm": 0.2472110390663147, "learning_rate": 6.651033003592646e-05, "loss": 0.0356, "step": 33860 }, { "grad_norm": 0.3648896813392639, "learning_rate": 6.649081956943626e-05, "loss": 0.0371, "step": 33870 }, { "grad_norm": 0.284819632768631, "learning_rate": 6.647130628511604e-05, "loss": 0.0333, "step": 33880 }, { "grad_norm": 0.5000724196434021, "learning_rate": 6.645179018630005e-05, "loss": 0.0416, "step": 33890 }, { "grad_norm": 0.31987762451171875, "learning_rate": 6.643227127632309e-05, "loss": 0.0384, "step": 33900 }, { "grad_norm": 0.4132862091064453, "learning_rate": 6.641274955852038e-05, "loss": 0.041, "step": 33910 }, { "grad_norm": 0.5504515171051025, "learning_rate": 6.639322503622768e-05, "loss": 0.0609, "step": 33920 }, { "grad_norm": 0.43151524662971497, "learning_rate": 6.637369771278116e-05, "loss": 0.0395, "step": 33930 }, { "grad_norm": 0.3699433207511902, "learning_rate": 6.635416759151751e-05, "loss": 0.0317, "step": 33940 }, { "grad_norm": 0.3463514745235443, "learning_rate": 6.633463467577394e-05, "loss": 0.0532, "step": 33950 }, { "grad_norm": 0.36056873202323914, "learning_rate": 6.631509896888803e-05, "loss": 0.0401, "step": 33960 }, { "grad_norm": 0.3177317678928375, "learning_rate": 6.629556047419794e-05, "loss": 0.0282, "step": 33970 }, { "grad_norm": 0.37254202365875244, "learning_rate": 6.627601919504223e-05, "loss": 0.0316, "step": 33980 }, { "grad_norm": 0.35195648670196533, "learning_rate": 6.625647513476001e-05, "loss": 0.0322, "step": 33990 }, { "grad_norm": 0.35731130838394165, "learning_rate": 6.62369282966908e-05, "loss": 0.0412, "step": 34000 }, { "grad_norm": 0.21701447665691376, "learning_rate": 6.621737868417464e-05, "loss": 0.0361, "step": 34010 }, { "grad_norm": 0.32204151153564453, "learning_rate": 6.619782630055198e-05, "loss": 0.0364, "step": 34020 }, { "grad_norm": 0.37381646037101746, "learning_rate": 6.617827114916382e-05, "loss": 0.0391, "step": 34030 }, { "grad_norm": 0.39921051263809204, "learning_rate": 6.615871323335161e-05, "loss": 0.0409, "step": 34040 }, { "grad_norm": 0.3320764899253845, "learning_rate": 6.613915255645725e-05, "loss": 0.0384, "step": 34050 }, { "grad_norm": 0.38453254103660583, "learning_rate": 6.611958912182312e-05, "loss": 0.0307, "step": 34060 }, { "grad_norm": 0.34379321336746216, "learning_rate": 6.610002293279207e-05, "loss": 0.036, "step": 34070 }, { "grad_norm": 0.4223739504814148, "learning_rate": 6.608045399270746e-05, "loss": 0.0369, "step": 34080 }, { "grad_norm": 0.4078851342201233, "learning_rate": 6.606088230491304e-05, "loss": 0.0343, "step": 34090 }, { "grad_norm": 0.28105252981185913, "learning_rate": 6.604130787275312e-05, "loss": 0.0375, "step": 34100 }, { "grad_norm": 0.3192872405052185, "learning_rate": 6.602173069957242e-05, "loss": 0.0285, "step": 34110 }, { "grad_norm": 0.5211154222488403, "learning_rate": 6.600215078871612e-05, "loss": 0.0372, "step": 34120 }, { "grad_norm": 0.2929498553276062, "learning_rate": 6.598256814352992e-05, "loss": 0.0356, "step": 34130 }, { "grad_norm": 0.43139761686325073, "learning_rate": 6.596298276735995e-05, "loss": 0.0348, "step": 34140 }, { "grad_norm": 0.33123779296875, "learning_rate": 6.594339466355282e-05, "loss": 0.0359, "step": 34150 }, { "grad_norm": 0.29233935475349426, "learning_rate": 6.592380383545558e-05, "loss": 0.0323, "step": 34160 }, { "grad_norm": 0.3399288058280945, "learning_rate": 6.590421028641577e-05, "loss": 0.0445, "step": 34170 }, { "grad_norm": 0.40626513957977295, "learning_rate": 6.588461401978143e-05, "loss": 0.0385, "step": 34180 }, { "grad_norm": 0.33965402841567993, "learning_rate": 6.586501503890099e-05, "loss": 0.0343, "step": 34190 }, { "grad_norm": 0.36839428544044495, "learning_rate": 6.584541334712338e-05, "loss": 0.0382, "step": 34200 }, { "grad_norm": 0.30545103549957275, "learning_rate": 6.582580894779802e-05, "loss": 0.0381, "step": 34210 }, { "grad_norm": 0.5703664422035217, "learning_rate": 6.580620184427473e-05, "loss": 0.0388, "step": 34220 }, { "grad_norm": 0.29508256912231445, "learning_rate": 6.578659203990385e-05, "loss": 0.0376, "step": 34230 }, { "grad_norm": 0.5771275758743286, "learning_rate": 6.576697953803615e-05, "loss": 0.0369, "step": 34240 }, { "grad_norm": 0.41343826055526733, "learning_rate": 6.57473643420229e-05, "loss": 0.0388, "step": 34250 }, { "grad_norm": 0.5478432774543762, "learning_rate": 6.572774645521574e-05, "loss": 0.0405, "step": 34260 }, { "grad_norm": 0.5832951068878174, "learning_rate": 6.570812588096688e-05, "loss": 0.0431, "step": 34270 }, { "grad_norm": 0.3402160406112671, "learning_rate": 6.568850262262893e-05, "loss": 0.0402, "step": 34280 }, { "grad_norm": 0.3041023015975952, "learning_rate": 6.566887668355497e-05, "loss": 0.0384, "step": 34290 }, { "grad_norm": 0.38678252696990967, "learning_rate": 6.564924806709851e-05, "loss": 0.0335, "step": 34300 }, { "grad_norm": 0.47502410411834717, "learning_rate": 6.562961677661359e-05, "loss": 0.0419, "step": 34310 }, { "grad_norm": 0.42285215854644775, "learning_rate": 6.56099828154546e-05, "loss": 0.0358, "step": 34320 }, { "grad_norm": 0.3204478919506073, "learning_rate": 6.55903461869765e-05, "loss": 0.0274, "step": 34330 }, { "grad_norm": 0.3085521161556244, "learning_rate": 6.557070689453465e-05, "loss": 0.0402, "step": 34340 }, { "grad_norm": 0.5577387809753418, "learning_rate": 6.555106494148482e-05, "loss": 0.0376, "step": 34350 }, { "grad_norm": 0.37162303924560547, "learning_rate": 6.553142033118333e-05, "loss": 0.0332, "step": 34360 }, { "grad_norm": 0.37374448776245117, "learning_rate": 6.551177306698688e-05, "loss": 0.0429, "step": 34370 }, { "grad_norm": 0.7196159958839417, "learning_rate": 6.549212315225267e-05, "loss": 0.0389, "step": 34380 }, { "grad_norm": 0.42073458433151245, "learning_rate": 6.547247059033833e-05, "loss": 0.0416, "step": 34390 }, { "grad_norm": 0.49200519919395447, "learning_rate": 6.545281538460193e-05, "loss": 0.0429, "step": 34400 }, { "grad_norm": 0.397610068321228, "learning_rate": 6.543315753840202e-05, "loss": 0.0394, "step": 34410 }, { "grad_norm": 0.6493840217590332, "learning_rate": 6.541349705509758e-05, "loss": 0.0408, "step": 34420 }, { "grad_norm": 0.20979374647140503, "learning_rate": 6.539383393804805e-05, "loss": 0.0382, "step": 34430 }, { "grad_norm": 0.3522494435310364, "learning_rate": 6.537416819061333e-05, "loss": 0.0411, "step": 34440 }, { "grad_norm": 0.6068335175514221, "learning_rate": 6.535449981615375e-05, "loss": 0.048, "step": 34450 }, { "grad_norm": 0.3088289499282837, "learning_rate": 6.53348288180301e-05, "loss": 0.0323, "step": 34460 }, { "grad_norm": 0.3144272267818451, "learning_rate": 6.531515519960361e-05, "loss": 0.0452, "step": 34470 }, { "grad_norm": 0.4019072949886322, "learning_rate": 6.529547896423597e-05, "loss": 0.0381, "step": 34480 }, { "grad_norm": 0.3125740587711334, "learning_rate": 6.52758001152893e-05, "loss": 0.0367, "step": 34490 }, { "grad_norm": 0.4198952317237854, "learning_rate": 6.525611865612618e-05, "loss": 0.0383, "step": 34500 }, { "grad_norm": 0.5582078695297241, "learning_rate": 6.523643459010966e-05, "loss": 0.0413, "step": 34510 }, { "grad_norm": 0.5038368105888367, "learning_rate": 6.521674792060317e-05, "loss": 0.044, "step": 34520 }, { "grad_norm": 0.4454152584075928, "learning_rate": 6.519705865097063e-05, "loss": 0.0499, "step": 34530 }, { "grad_norm": 0.38599255681037903, "learning_rate": 6.517736678457641e-05, "loss": 0.038, "step": 34540 }, { "grad_norm": 0.3994236886501312, "learning_rate": 6.515767232478534e-05, "loss": 0.0405, "step": 34550 }, { "grad_norm": 0.35826507210731506, "learning_rate": 6.51379752749626e-05, "loss": 0.0342, "step": 34560 }, { "grad_norm": 0.40966084599494934, "learning_rate": 6.511827563847393e-05, "loss": 0.0332, "step": 34570 }, { "grad_norm": 0.39551323652267456, "learning_rate": 6.509857341868542e-05, "loss": 0.038, "step": 34580 }, { "grad_norm": 0.46885743737220764, "learning_rate": 6.507886861896367e-05, "loss": 0.0365, "step": 34590 }, { "grad_norm": 0.3425697088241577, "learning_rate": 6.505916124267567e-05, "loss": 0.0439, "step": 34600 }, { "grad_norm": 0.30364811420440674, "learning_rate": 6.503945129318891e-05, "loss": 0.0346, "step": 34610 }, { "grad_norm": 0.4715520441532135, "learning_rate": 6.501973877387122e-05, "loss": 0.0317, "step": 34620 }, { "grad_norm": 0.39070382714271545, "learning_rate": 6.500002368809098e-05, "loss": 0.0406, "step": 34630 }, { "grad_norm": 0.3887290358543396, "learning_rate": 6.498030603921694e-05, "loss": 0.0408, "step": 34640 }, { "grad_norm": 0.3500918745994568, "learning_rate": 6.496058583061832e-05, "loss": 0.0405, "step": 34650 }, { "grad_norm": 0.33851519227027893, "learning_rate": 6.494086306566475e-05, "loss": 0.0266, "step": 34660 }, { "grad_norm": 0.3245106041431427, "learning_rate": 6.492113774772632e-05, "loss": 0.0371, "step": 34670 }, { "grad_norm": 0.6567056775093079, "learning_rate": 6.490140988017354e-05, "loss": 0.0461, "step": 34680 }, { "grad_norm": 0.47380468249320984, "learning_rate": 6.488167946637736e-05, "loss": 0.0381, "step": 34690 }, { "grad_norm": 0.4901616871356964, "learning_rate": 6.486194650970915e-05, "loss": 0.0455, "step": 34700 }, { "grad_norm": 0.38001763820648193, "learning_rate": 6.48422110135408e-05, "loss": 0.0426, "step": 34710 }, { "grad_norm": 0.6832957863807678, "learning_rate": 6.482247298124451e-05, "loss": 0.053, "step": 34720 }, { "grad_norm": 0.6069344282150269, "learning_rate": 6.480273241619297e-05, "loss": 0.0407, "step": 34730 }, { "grad_norm": 0.5209649205207825, "learning_rate": 6.478298932175933e-05, "loss": 0.0403, "step": 34740 }, { "grad_norm": 0.3501022458076477, "learning_rate": 6.476324370131712e-05, "loss": 0.0372, "step": 34750 }, { "grad_norm": 0.4172563850879669, "learning_rate": 6.474349555824036e-05, "loss": 0.0481, "step": 34760 }, { "grad_norm": 0.31946706771850586, "learning_rate": 6.472374489590342e-05, "loss": 0.0312, "step": 34770 }, { "grad_norm": 0.6548653244972229, "learning_rate": 6.470399171768118e-05, "loss": 0.0515, "step": 34780 }, { "grad_norm": 0.4406540095806122, "learning_rate": 6.468423602694891e-05, "loss": 0.0404, "step": 34790 }, { "grad_norm": 0.4711112976074219, "learning_rate": 6.466447782708232e-05, "loss": 0.0319, "step": 34800 }, { "grad_norm": 0.43919456005096436, "learning_rate": 6.464471712145754e-05, "loss": 0.0328, "step": 34810 }, { "grad_norm": 0.3592548072338104, "learning_rate": 6.462495391345114e-05, "loss": 0.0388, "step": 34820 }, { "grad_norm": 0.450865238904953, "learning_rate": 6.46051882064401e-05, "loss": 0.0314, "step": 34830 }, { "grad_norm": 0.3589326739311218, "learning_rate": 6.458542000380186e-05, "loss": 0.0406, "step": 34840 }, { "grad_norm": 0.4366418421268463, "learning_rate": 6.456564930891424e-05, "loss": 0.0369, "step": 34850 }, { "grad_norm": 0.44246944785118103, "learning_rate": 6.454587612515555e-05, "loss": 0.0455, "step": 34860 }, { "grad_norm": 0.3754095733165741, "learning_rate": 6.452610045590444e-05, "loss": 0.0441, "step": 34870 }, { "grad_norm": 0.4754062592983246, "learning_rate": 6.450632230454005e-05, "loss": 0.0324, "step": 34880 }, { "grad_norm": 0.38519468903541565, "learning_rate": 6.448654167444195e-05, "loss": 0.0514, "step": 34890 }, { "grad_norm": 0.5636332035064697, "learning_rate": 6.446675856899005e-05, "loss": 0.0386, "step": 34900 }, { "grad_norm": 0.34603559970855713, "learning_rate": 6.444697299156481e-05, "loss": 0.0386, "step": 34910 }, { "grad_norm": 0.4023629426956177, "learning_rate": 6.442718494554701e-05, "loss": 0.0401, "step": 34920 }, { "grad_norm": 0.39890149235725403, "learning_rate": 6.440739443431787e-05, "loss": 0.0375, "step": 34930 }, { "grad_norm": 0.2724074721336365, "learning_rate": 6.438760146125906e-05, "loss": 0.0338, "step": 34940 }, { "grad_norm": 0.33071574568748474, "learning_rate": 6.436780602975267e-05, "loss": 0.0429, "step": 34950 }, { "grad_norm": 0.3755177855491638, "learning_rate": 6.43480081431812e-05, "loss": 0.0396, "step": 34960 }, { "grad_norm": 0.3844226598739624, "learning_rate": 6.432820780492756e-05, "loss": 0.0372, "step": 34970 }, { "grad_norm": 0.3669443428516388, "learning_rate": 6.430840501837506e-05, "loss": 0.0477, "step": 34980 }, { "grad_norm": 0.45258766412734985, "learning_rate": 6.428859978690748e-05, "loss": 0.0389, "step": 34990 }, { "grad_norm": 0.3314303755760193, "learning_rate": 6.426879211390901e-05, "loss": 0.0413, "step": 35000 }, { "grad_norm": 0.40403878688812256, "learning_rate": 6.424898200276422e-05, "loss": 0.0422, "step": 35010 }, { "grad_norm": 0.29323244094848633, "learning_rate": 6.42291694568581e-05, "loss": 0.032, "step": 35020 }, { "grad_norm": 0.41675013303756714, "learning_rate": 6.42093544795761e-05, "loss": 0.0449, "step": 35030 }, { "grad_norm": 0.3310436010360718, "learning_rate": 6.418953707430403e-05, "loss": 0.0404, "step": 35040 }, { "grad_norm": 0.35592949390411377, "learning_rate": 6.416971724442819e-05, "loss": 0.0353, "step": 35050 }, { "grad_norm": 0.28991109132766724, "learning_rate": 6.414989499333519e-05, "loss": 0.037, "step": 35060 }, { "grad_norm": 0.3437623381614685, "learning_rate": 6.413007032441214e-05, "loss": 0.0368, "step": 35070 }, { "grad_norm": 0.4759017527103424, "learning_rate": 6.411024324104653e-05, "loss": 0.0405, "step": 35080 }, { "grad_norm": 0.3921799063682556, "learning_rate": 6.409041374662628e-05, "loss": 0.0372, "step": 35090 }, { "grad_norm": 0.4851076900959015, "learning_rate": 6.407058184453967e-05, "loss": 0.048, "step": 35100 }, { "grad_norm": 0.3489908277988434, "learning_rate": 6.405074753817548e-05, "loss": 0.0353, "step": 35110 }, { "grad_norm": 0.30503153800964355, "learning_rate": 6.40309108309228e-05, "loss": 0.0376, "step": 35120 }, { "grad_norm": 0.3622809648513794, "learning_rate": 6.401107172617122e-05, "loss": 0.041, "step": 35130 }, { "grad_norm": 0.4075831174850464, "learning_rate": 6.399123022731068e-05, "loss": 0.0548, "step": 35140 }, { "grad_norm": 0.3633197546005249, "learning_rate": 6.397138633773157e-05, "loss": 0.0341, "step": 35150 }, { "grad_norm": 0.48059701919555664, "learning_rate": 6.395154006082463e-05, "loss": 0.0369, "step": 35160 }, { "grad_norm": 0.497844398021698, "learning_rate": 6.393169139998109e-05, "loss": 0.0375, "step": 35170 }, { "grad_norm": 0.4444634020328522, "learning_rate": 6.39118403585925e-05, "loss": 0.0416, "step": 35180 }, { "grad_norm": 0.3399771451950073, "learning_rate": 6.38919869400509e-05, "loss": 0.0348, "step": 35190 }, { "grad_norm": 0.33281493186950684, "learning_rate": 6.387213114774865e-05, "loss": 0.0391, "step": 35200 }, { "grad_norm": 0.3687630891799927, "learning_rate": 6.385227298507863e-05, "loss": 0.0567, "step": 35210 }, { "grad_norm": 0.30961695313453674, "learning_rate": 6.3832412455434e-05, "loss": 0.0334, "step": 35220 }, { "grad_norm": 0.5626817345619202, "learning_rate": 6.381254956220841e-05, "loss": 0.035, "step": 35230 }, { "grad_norm": 0.28321143984794617, "learning_rate": 6.379268430879586e-05, "loss": 0.0332, "step": 35240 }, { "grad_norm": 0.30797556042671204, "learning_rate": 6.37728166985908e-05, "loss": 0.034, "step": 35250 }, { "grad_norm": 0.3171619772911072, "learning_rate": 6.375294673498804e-05, "loss": 0.0327, "step": 35260 }, { "grad_norm": 0.3481070399284363, "learning_rate": 6.373307442138284e-05, "loss": 0.0377, "step": 35270 }, { "grad_norm": 0.35120150446891785, "learning_rate": 6.371319976117081e-05, "loss": 0.0483, "step": 35280 }, { "grad_norm": 0.4448549747467041, "learning_rate": 6.3693322757748e-05, "loss": 0.0325, "step": 35290 }, { "grad_norm": 0.4103727638721466, "learning_rate": 6.367344341451086e-05, "loss": 0.0369, "step": 35300 }, { "grad_norm": 0.39537060260772705, "learning_rate": 6.36535617348562e-05, "loss": 0.0379, "step": 35310 }, { "grad_norm": 0.3944995701313019, "learning_rate": 6.363367772218128e-05, "loss": 0.0381, "step": 35320 }, { "grad_norm": 0.3646262586116791, "learning_rate": 6.36137913798837e-05, "loss": 0.0411, "step": 35330 }, { "grad_norm": 0.5677449107170105, "learning_rate": 6.359390271136151e-05, "loss": 0.0317, "step": 35340 }, { "grad_norm": 0.3648473620414734, "learning_rate": 6.357401172001314e-05, "loss": 0.0355, "step": 35350 }, { "grad_norm": 0.4677514135837555, "learning_rate": 6.355411840923742e-05, "loss": 0.0379, "step": 35360 }, { "grad_norm": 0.3500714898109436, "learning_rate": 6.353422278243358e-05, "loss": 0.0353, "step": 35370 }, { "grad_norm": 0.4827217161655426, "learning_rate": 6.351432484300121e-05, "loss": 0.0391, "step": 35380 }, { "grad_norm": 0.35738423466682434, "learning_rate": 6.349442459434036e-05, "loss": 0.0405, "step": 35390 }, { "grad_norm": 0.33223074674606323, "learning_rate": 6.34745220398514e-05, "loss": 0.0407, "step": 35400 }, { "grad_norm": 0.5362094044685364, "learning_rate": 6.345461718293518e-05, "loss": 0.0356, "step": 35410 }, { "grad_norm": 0.28643086552619934, "learning_rate": 6.343471002699286e-05, "loss": 0.045, "step": 35420 }, { "grad_norm": 0.4061466455459595, "learning_rate": 6.341480057542602e-05, "loss": 0.0333, "step": 35430 }, { "grad_norm": 0.3121757507324219, "learning_rate": 6.339488883163667e-05, "loss": 0.0407, "step": 35440 }, { "grad_norm": 0.28400859236717224, "learning_rate": 6.337497479902716e-05, "loss": 0.0415, "step": 35450 }, { "grad_norm": 0.5047480463981628, "learning_rate": 6.335505848100027e-05, "loss": 0.0426, "step": 35460 }, { "grad_norm": 0.35529249906539917, "learning_rate": 6.333513988095915e-05, "loss": 0.0282, "step": 35470 }, { "grad_norm": 0.5476996898651123, "learning_rate": 6.331521900230735e-05, "loss": 0.0389, "step": 35480 }, { "grad_norm": 0.24185854196548462, "learning_rate": 6.329529584844878e-05, "loss": 0.0324, "step": 35490 }, { "grad_norm": 0.5223464369773865, "learning_rate": 6.327537042278777e-05, "loss": 0.0365, "step": 35500 }, { "grad_norm": 0.5229699015617371, "learning_rate": 6.325544272872905e-05, "loss": 0.0359, "step": 35510 }, { "grad_norm": 0.3571482002735138, "learning_rate": 6.323551276967771e-05, "loss": 0.0417, "step": 35520 }, { "grad_norm": 0.4344388544559479, "learning_rate": 6.321558054903922e-05, "loss": 0.0362, "step": 35530 }, { "grad_norm": 0.3921179473400116, "learning_rate": 6.319564607021947e-05, "loss": 0.0388, "step": 35540 }, { "grad_norm": 0.4875636696815491, "learning_rate": 6.31757093366247e-05, "loss": 0.0378, "step": 35550 }, { "grad_norm": 0.4373462498188019, "learning_rate": 6.315577035166154e-05, "loss": 0.0374, "step": 35560 }, { "grad_norm": 0.5196254849433899, "learning_rate": 6.313582911873708e-05, "loss": 0.0321, "step": 35570 }, { "grad_norm": 0.32568633556365967, "learning_rate": 6.311588564125865e-05, "loss": 0.0325, "step": 35580 }, { "grad_norm": 0.603614091873169, "learning_rate": 6.30959399226341e-05, "loss": 0.0425, "step": 35590 }, { "grad_norm": 0.4316583275794983, "learning_rate": 6.30759919662716e-05, "loss": 0.0505, "step": 35600 }, { "grad_norm": 0.3452112078666687, "learning_rate": 6.30560417755797e-05, "loss": 0.0384, "step": 35610 }, { "grad_norm": 0.34142038226127625, "learning_rate": 6.303608935396735e-05, "loss": 0.0283, "step": 35620 }, { "grad_norm": 0.26578548550605774, "learning_rate": 6.301613470484386e-05, "loss": 0.0322, "step": 35630 }, { "grad_norm": 0.41974422335624695, "learning_rate": 6.299617783161893e-05, "loss": 0.0415, "step": 35640 }, { "grad_norm": 0.47707894444465637, "learning_rate": 6.297621873770266e-05, "loss": 0.0349, "step": 35650 }, { "grad_norm": 0.2786535620689392, "learning_rate": 6.29562574265055e-05, "loss": 0.039, "step": 35660 }, { "grad_norm": 0.29815831780433655, "learning_rate": 6.293629390143834e-05, "loss": 0.0359, "step": 35670 }, { "grad_norm": 0.4060559570789337, "learning_rate": 6.291632816591232e-05, "loss": 0.0306, "step": 35680 }, { "grad_norm": 0.3382952809333801, "learning_rate": 6.28963602233391e-05, "loss": 0.0358, "step": 35690 }, { "grad_norm": 0.43966737389564514, "learning_rate": 6.287639007713062e-05, "loss": 0.0357, "step": 35700 }, { "grad_norm": 0.3756888806819916, "learning_rate": 6.285641773069926e-05, "loss": 0.0368, "step": 35710 }, { "grad_norm": 0.4137192666530609, "learning_rate": 6.283644318745773e-05, "loss": 0.0292, "step": 35720 }, { "grad_norm": 0.3881758451461792, "learning_rate": 6.281646645081912e-05, "loss": 0.039, "step": 35730 }, { "grad_norm": 0.2927871644496918, "learning_rate": 6.279648752419693e-05, "loss": 0.0313, "step": 35740 }, { "grad_norm": 0.4057323932647705, "learning_rate": 6.2776506411005e-05, "loss": 0.0412, "step": 35750 }, { "grad_norm": 0.3673946261405945, "learning_rate": 6.275652311465758e-05, "loss": 0.0311, "step": 35760 }, { "grad_norm": 0.37688612937927246, "learning_rate": 6.273653763856926e-05, "loss": 0.0347, "step": 35770 }, { "grad_norm": 0.2833835184574127, "learning_rate": 6.271654998615501e-05, "loss": 0.036, "step": 35780 }, { "grad_norm": 0.3989521265029907, "learning_rate": 6.269656016083013e-05, "loss": 0.0295, "step": 35790 }, { "grad_norm": 0.4099116027355194, "learning_rate": 6.267656816601038e-05, "loss": 0.0397, "step": 35800 }, { "grad_norm": 0.40596336126327515, "learning_rate": 6.265657400511185e-05, "loss": 0.0346, "step": 35810 }, { "grad_norm": 0.3771127164363861, "learning_rate": 6.263657768155098e-05, "loss": 0.041, "step": 35820 }, { "grad_norm": 0.48786336183547974, "learning_rate": 6.261657919874457e-05, "loss": 0.0389, "step": 35830 }, { "grad_norm": 0.48240622878074646, "learning_rate": 6.259657856010986e-05, "loss": 0.0449, "step": 35840 }, { "grad_norm": 0.3845537602901459, "learning_rate": 6.257657576906439e-05, "loss": 0.038, "step": 35850 }, { "grad_norm": 0.4042571187019348, "learning_rate": 6.255657082902609e-05, "loss": 0.0312, "step": 35860 }, { "grad_norm": 0.33472493290901184, "learning_rate": 6.253656374341325e-05, "loss": 0.0276, "step": 35870 }, { "grad_norm": 0.36705952882766724, "learning_rate": 6.251655451564457e-05, "loss": 0.0343, "step": 35880 }, { "grad_norm": 0.5362883806228638, "learning_rate": 6.249654314913902e-05, "loss": 0.0354, "step": 35890 }, { "grad_norm": 0.28195691108703613, "learning_rate": 6.247652964731604e-05, "loss": 0.0341, "step": 35900 }, { "grad_norm": 0.6672795414924622, "learning_rate": 6.245651401359537e-05, "loss": 0.0514, "step": 35910 }, { "grad_norm": 0.444439172744751, "learning_rate": 6.243649625139715e-05, "loss": 0.0425, "step": 35920 }, { "grad_norm": 1.2211183309555054, "learning_rate": 6.241647636414185e-05, "loss": 0.0432, "step": 35930 }, { "grad_norm": 0.39579910039901733, "learning_rate": 6.239645435525034e-05, "loss": 0.0474, "step": 35940 }, { "grad_norm": 0.2714027464389801, "learning_rate": 6.237643022814381e-05, "loss": 0.0353, "step": 35950 }, { "grad_norm": 0.579781711101532, "learning_rate": 6.235640398624386e-05, "loss": 0.0331, "step": 35960 }, { "grad_norm": 0.2608727216720581, "learning_rate": 6.233637563297243e-05, "loss": 0.0444, "step": 35970 }, { "grad_norm": 0.4880481958389282, "learning_rate": 6.23163451717518e-05, "loss": 0.0408, "step": 35980 }, { "grad_norm": 0.3649269640445709, "learning_rate": 6.229631260600463e-05, "loss": 0.0365, "step": 35990 }, { "grad_norm": 0.3889285922050476, "learning_rate": 6.227627793915392e-05, "loss": 0.0437, "step": 36000 }, { "grad_norm": 0.39232540130615234, "learning_rate": 6.225624117462309e-05, "loss": 0.0334, "step": 36010 }, { "grad_norm": 0.3335677683353424, "learning_rate": 6.223620231583586e-05, "loss": 0.0418, "step": 36020 }, { "grad_norm": 0.40111568570137024, "learning_rate": 6.221616136621629e-05, "loss": 0.0385, "step": 36030 }, { "grad_norm": 0.3052195608615875, "learning_rate": 6.219611832918887e-05, "loss": 0.0387, "step": 36040 }, { "grad_norm": 0.4272320866584778, "learning_rate": 6.217607320817838e-05, "loss": 0.0322, "step": 36050 }, { "grad_norm": 0.40741270780563354, "learning_rate": 6.215602600661001e-05, "loss": 0.0414, "step": 36060 }, { "grad_norm": 0.40532803535461426, "learning_rate": 6.213597672790925e-05, "loss": 0.0329, "step": 36070 }, { "grad_norm": 0.2976875603199005, "learning_rate": 6.2115925375502e-05, "loss": 0.0332, "step": 36080 }, { "grad_norm": 0.422887921333313, "learning_rate": 6.209587195281447e-05, "loss": 0.0365, "step": 36090 }, { "grad_norm": 0.43043947219848633, "learning_rate": 6.207581646327324e-05, "loss": 0.0463, "step": 36100 }, { "grad_norm": 0.48788148164749146, "learning_rate": 6.205575891030526e-05, "loss": 0.0289, "step": 36110 }, { "grad_norm": 0.27430322766304016, "learning_rate": 6.203569929733781e-05, "loss": 0.0252, "step": 36120 }, { "grad_norm": 0.32908257842063904, "learning_rate": 6.201563762779852e-05, "loss": 0.0372, "step": 36130 }, { "grad_norm": 0.29282817244529724, "learning_rate": 6.199557390511538e-05, "loss": 0.0321, "step": 36140 }, { "grad_norm": 0.531650185585022, "learning_rate": 6.197550813271675e-05, "loss": 0.0313, "step": 36150 }, { "grad_norm": 0.3082340359687805, "learning_rate": 6.195544031403131e-05, "loss": 0.0363, "step": 36160 }, { "grad_norm": 0.35790106654167175, "learning_rate": 6.19353704524881e-05, "loss": 0.0342, "step": 36170 }, { "grad_norm": 0.4625588655471802, "learning_rate": 6.191529855151652e-05, "loss": 0.0313, "step": 36180 }, { "grad_norm": 0.3197106420993805, "learning_rate": 6.189522461454629e-05, "loss": 0.0429, "step": 36190 }, { "grad_norm": 0.32730594277381897, "learning_rate": 6.187514864500752e-05, "loss": 0.0329, "step": 36200 }, { "grad_norm": 0.5106143951416016, "learning_rate": 6.185507064633062e-05, "loss": 0.0399, "step": 36210 }, { "grad_norm": 0.4383701682090759, "learning_rate": 6.18349906219464e-05, "loss": 0.0417, "step": 36220 }, { "grad_norm": 0.35359200835227966, "learning_rate": 6.181490857528596e-05, "loss": 0.0361, "step": 36230 }, { "grad_norm": 0.3639635741710663, "learning_rate": 6.179482450978077e-05, "loss": 0.0303, "step": 36240 }, { "grad_norm": 0.306465208530426, "learning_rate": 6.177473842886269e-05, "loss": 0.0321, "step": 36250 }, { "grad_norm": 0.29082006216049194, "learning_rate": 6.175465033596382e-05, "loss": 0.036, "step": 36260 }, { "grad_norm": 0.2939744293689728, "learning_rate": 6.173456023451671e-05, "loss": 0.0325, "step": 36270 }, { "grad_norm": 0.48726046085357666, "learning_rate": 6.171446812795422e-05, "loss": 0.0394, "step": 36280 }, { "grad_norm": 0.45166388154029846, "learning_rate": 6.169437401970949e-05, "loss": 0.0465, "step": 36290 }, { "grad_norm": 0.35112908482551575, "learning_rate": 6.16742779132161e-05, "loss": 0.0325, "step": 36300 }, { "grad_norm": 0.4489472210407257, "learning_rate": 6.165417981190789e-05, "loss": 0.0344, "step": 36310 }, { "grad_norm": 0.338824599981308, "learning_rate": 6.16340797192191e-05, "loss": 0.0359, "step": 36320 }, { "grad_norm": 0.4880026876926422, "learning_rate": 6.161397763858427e-05, "loss": 0.0372, "step": 36330 }, { "grad_norm": 0.5236928462982178, "learning_rate": 6.159387357343834e-05, "loss": 0.0347, "step": 36340 }, { "grad_norm": 0.45207658410072327, "learning_rate": 6.157376752721648e-05, "loss": 0.0404, "step": 36350 }, { "grad_norm": 0.2848813831806183, "learning_rate": 6.155365950335428e-05, "loss": 0.0331, "step": 36360 }, { "grad_norm": 0.4415797293186188, "learning_rate": 6.153354950528768e-05, "loss": 0.0337, "step": 36370 }, { "grad_norm": 0.3469238877296448, "learning_rate": 6.151343753645293e-05, "loss": 0.0422, "step": 36380 }, { "grad_norm": 0.24297533929347992, "learning_rate": 6.149332360028657e-05, "loss": 0.0407, "step": 36390 }, { "grad_norm": 0.31344854831695557, "learning_rate": 6.147320770022555e-05, "loss": 0.0407, "step": 36400 }, { "grad_norm": 0.4143078327178955, "learning_rate": 6.145308983970715e-05, "loss": 0.0367, "step": 36410 }, { "grad_norm": 0.326040655374527, "learning_rate": 6.143297002216892e-05, "loss": 0.0351, "step": 36420 }, { "grad_norm": 0.37306979298591614, "learning_rate": 6.141284825104882e-05, "loss": 0.0417, "step": 36430 }, { "grad_norm": 0.4729524552822113, "learning_rate": 6.13927245297851e-05, "loss": 0.0389, "step": 36440 }, { "grad_norm": 0.39007464051246643, "learning_rate": 6.137259886181633e-05, "loss": 0.0346, "step": 36450 }, { "grad_norm": 0.37077537178993225, "learning_rate": 6.135247125058145e-05, "loss": 0.0348, "step": 36460 }, { "grad_norm": 0.26947569847106934, "learning_rate": 6.133234169951974e-05, "loss": 0.0283, "step": 36470 }, { "grad_norm": 0.4365943372249603, "learning_rate": 6.131221021207078e-05, "loss": 0.0355, "step": 36480 }, { "grad_norm": 0.3343089520931244, "learning_rate": 6.129207679167448e-05, "loss": 0.0327, "step": 36490 }, { "grad_norm": 0.3519025444984436, "learning_rate": 6.127194144177109e-05, "loss": 0.0335, "step": 36500 }, { "grad_norm": 0.3507915139198303, "learning_rate": 6.125180416580118e-05, "loss": 0.0348, "step": 36510 }, { "grad_norm": 0.38621631264686584, "learning_rate": 6.123166496720571e-05, "loss": 0.0346, "step": 36520 }, { "grad_norm": 0.4058786630630493, "learning_rate": 6.121152384942588e-05, "loss": 0.032, "step": 36530 }, { "grad_norm": 0.5091175436973572, "learning_rate": 6.119138081590324e-05, "loss": 0.0318, "step": 36540 }, { "grad_norm": 0.40260282158851624, "learning_rate": 6.117123587007971e-05, "loss": 0.0386, "step": 36550 }, { "grad_norm": 0.31768307089805603, "learning_rate": 6.11510890153975e-05, "loss": 0.0342, "step": 36560 }, { "grad_norm": 0.3447524309158325, "learning_rate": 6.113094025529916e-05, "loss": 0.0297, "step": 36570 }, { "grad_norm": 0.2836441397666931, "learning_rate": 6.111078959322757e-05, "loss": 0.048, "step": 36580 }, { "grad_norm": 0.5325050354003906, "learning_rate": 6.109063703262592e-05, "loss": 0.0468, "step": 36590 }, { "grad_norm": 0.31799256801605225, "learning_rate": 6.107048257693772e-05, "loss": 0.0369, "step": 36600 }, { "grad_norm": 0.30139613151550293, "learning_rate": 6.105032622960683e-05, "loss": 0.0325, "step": 36610 }, { "grad_norm": 0.2876310646533966, "learning_rate": 6.103016799407743e-05, "loss": 0.032, "step": 36620 }, { "grad_norm": 0.35843607783317566, "learning_rate": 6.1010007873793984e-05, "loss": 0.0347, "step": 36630 }, { "grad_norm": 0.28402096033096313, "learning_rate": 6.098984587220131e-05, "loss": 0.0298, "step": 36640 }, { "grad_norm": 0.2828271985054016, "learning_rate": 6.096968199274456e-05, "loss": 0.0344, "step": 36650 }, { "grad_norm": 0.5122531652450562, "learning_rate": 6.0949516238869166e-05, "loss": 0.0328, "step": 36660 }, { "grad_norm": 0.40571314096450806, "learning_rate": 6.092934861402092e-05, "loss": 0.043, "step": 36670 }, { "grad_norm": 0.565025269985199, "learning_rate": 6.0909179121645924e-05, "loss": 0.0433, "step": 36680 }, { "grad_norm": 0.46976935863494873, "learning_rate": 6.0889007765190576e-05, "loss": 0.0442, "step": 36690 }, { "grad_norm": 0.32577764987945557, "learning_rate": 6.086883454810162e-05, "loss": 0.0353, "step": 36700 }, { "grad_norm": 0.4125698208808899, "learning_rate": 6.0848659473826084e-05, "loss": 0.041, "step": 36710 }, { "grad_norm": 0.406112939119339, "learning_rate": 6.082848254581138e-05, "loss": 0.0382, "step": 36720 }, { "grad_norm": 0.3836565315723419, "learning_rate": 6.080830376750517e-05, "loss": 0.0335, "step": 36730 }, { "grad_norm": 0.28200462460517883, "learning_rate": 6.0788123142355445e-05, "loss": 0.0287, "step": 36740 }, { "grad_norm": 0.47268205881118774, "learning_rate": 6.076794067381052e-05, "loss": 0.0328, "step": 36750 }, { "grad_norm": 0.34113985300064087, "learning_rate": 6.074775636531905e-05, "loss": 0.0333, "step": 36760 }, { "grad_norm": 0.3659083843231201, "learning_rate": 6.072757022032997e-05, "loss": 0.0364, "step": 36770 }, { "grad_norm": 0.5205027461051941, "learning_rate": 6.070738224229253e-05, "loss": 0.0454, "step": 36780 }, { "grad_norm": 0.399254709482193, "learning_rate": 6.0687192434656314e-05, "loss": 0.0412, "step": 36790 }, { "grad_norm": 0.45459261536598206, "learning_rate": 6.066700080087121e-05, "loss": 0.0357, "step": 36800 }, { "grad_norm": 0.3716212213039398, "learning_rate": 6.0646807344387424e-05, "loss": 0.0347, "step": 36810 }, { "grad_norm": 0.3535866439342499, "learning_rate": 6.062661206865543e-05, "loss": 0.0442, "step": 36820 }, { "grad_norm": 0.3806038796901703, "learning_rate": 6.06064149771261e-05, "loss": 0.0385, "step": 36830 }, { "grad_norm": 0.4044092297554016, "learning_rate": 6.058621607325051e-05, "loss": 0.0357, "step": 36840 }, { "grad_norm": 0.5028988122940063, "learning_rate": 6.056601536048014e-05, "loss": 0.0412, "step": 36850 }, { "grad_norm": 0.35525014996528625, "learning_rate": 6.0545812842266725e-05, "loss": 0.0455, "step": 36860 }, { "grad_norm": 0.32606595754623413, "learning_rate": 6.052560852206232e-05, "loss": 0.0353, "step": 36870 }, { "grad_norm": 0.30734363198280334, "learning_rate": 6.05054024033193e-05, "loss": 0.0444, "step": 36880 }, { "grad_norm": 0.32525548338890076, "learning_rate": 6.048519448949032e-05, "loss": 0.029, "step": 36890 }, { "grad_norm": 0.34712010622024536, "learning_rate": 6.046498478402839e-05, "loss": 0.0435, "step": 36900 }, { "grad_norm": 0.4840649366378784, "learning_rate": 6.044477329038677e-05, "loss": 0.0364, "step": 36910 }, { "grad_norm": 0.46293437480926514, "learning_rate": 6.042456001201906e-05, "loss": 0.0354, "step": 36920 }, { "grad_norm": 0.3264986276626587, "learning_rate": 6.040434495237917e-05, "loss": 0.0372, "step": 36930 }, { "grad_norm": 0.5145646929740906, "learning_rate": 6.0384128114921256e-05, "loss": 0.0399, "step": 36940 }, { "grad_norm": 0.4018043577671051, "learning_rate": 6.036390950309987e-05, "loss": 0.0266, "step": 36950 }, { "grad_norm": 0.4914610683917999, "learning_rate": 6.0343689120369805e-05, "loss": 0.0383, "step": 36960 }, { "grad_norm": 0.3439989686012268, "learning_rate": 6.032346697018616e-05, "loss": 0.0299, "step": 36970 }, { "grad_norm": 0.3800242841243744, "learning_rate": 6.0303243056004375e-05, "loss": 0.0489, "step": 36980 }, { "grad_norm": 0.44122523069381714, "learning_rate": 6.0283017381280136e-05, "loss": 0.035, "step": 36990 }, { "grad_norm": 0.3673299252986908, "learning_rate": 6.026278994946948e-05, "loss": 0.0317, "step": 37000 }, { "grad_norm": 0.3615025579929352, "learning_rate": 6.02425607640287e-05, "loss": 0.0395, "step": 37010 }, { "grad_norm": 0.3284735381603241, "learning_rate": 6.022232982841441e-05, "loss": 0.0353, "step": 37020 }, { "grad_norm": 0.3466147482395172, "learning_rate": 6.020209714608355e-05, "loss": 0.0319, "step": 37030 }, { "grad_norm": 0.37756043672561646, "learning_rate": 6.018186272049331e-05, "loss": 0.0343, "step": 37040 }, { "grad_norm": 0.5276240706443787, "learning_rate": 6.01616265551012e-05, "loss": 0.0384, "step": 37050 }, { "grad_norm": 0.449457049369812, "learning_rate": 6.014138865336503e-05, "loss": 0.0258, "step": 37060 }, { "grad_norm": 0.3823310434818268, "learning_rate": 6.0121149018742905e-05, "loss": 0.0332, "step": 37070 }, { "grad_norm": 0.3212636411190033, "learning_rate": 6.010090765469325e-05, "loss": 0.0338, "step": 37080 }, { "grad_norm": 0.38887155055999756, "learning_rate": 6.008066456467473e-05, "loss": 0.0306, "step": 37090 }, { "grad_norm": 0.2948232889175415, "learning_rate": 6.0060419752146335e-05, "loss": 0.0325, "step": 37100 }, { "grad_norm": 0.44101855158805847, "learning_rate": 6.0040173220567353e-05, "loss": 0.0368, "step": 37110 }, { "grad_norm": 0.43235114216804504, "learning_rate": 6.001992497339737e-05, "loss": 0.0368, "step": 37120 }, { "grad_norm": 0.25954368710517883, "learning_rate": 5.999967501409626e-05, "loss": 0.0384, "step": 37130 }, { "grad_norm": 0.2813403904438019, "learning_rate": 5.997942334612418e-05, "loss": 0.0301, "step": 37140 }, { "grad_norm": 0.3531177341938019, "learning_rate": 5.995916997294158e-05, "loss": 0.0295, "step": 37150 }, { "grad_norm": 0.2897784411907196, "learning_rate": 5.9938914898009214e-05, "loss": 0.034, "step": 37160 }, { "grad_norm": 0.3726854920387268, "learning_rate": 5.991865812478813e-05, "loss": 0.0267, "step": 37170 }, { "grad_norm": 0.3780357241630554, "learning_rate": 5.989839965673964e-05, "loss": 0.0522, "step": 37180 }, { "grad_norm": 0.2522846758365631, "learning_rate": 5.987813949732539e-05, "loss": 0.0385, "step": 37190 }, { "grad_norm": 0.3299083709716797, "learning_rate": 5.9857877650007255e-05, "loss": 0.0339, "step": 37200 }, { "grad_norm": 0.30805835127830505, "learning_rate": 5.983761411824744e-05, "loss": 0.0279, "step": 37210 }, { "grad_norm": 0.31886669993400574, "learning_rate": 5.981734890550844e-05, "loss": 0.0341, "step": 37220 }, { "grad_norm": 0.20326931774616241, "learning_rate": 5.979708201525301e-05, "loss": 0.0344, "step": 37230 }, { "grad_norm": 0.4259614944458008, "learning_rate": 5.977681345094422e-05, "loss": 0.0389, "step": 37240 }, { "grad_norm": 0.5232101678848267, "learning_rate": 5.97565432160454e-05, "loss": 0.0293, "step": 37250 }, { "grad_norm": 0.2600761651992798, "learning_rate": 5.9736271314020186e-05, "loss": 0.0312, "step": 37260 }, { "grad_norm": 0.273074209690094, "learning_rate": 5.971599774833251e-05, "loss": 0.0313, "step": 37270 }, { "grad_norm": 0.35099634528160095, "learning_rate": 5.9695722522446525e-05, "loss": 0.0359, "step": 37280 }, { "grad_norm": 0.37224069237709045, "learning_rate": 5.9675445639826765e-05, "loss": 0.0331, "step": 37290 }, { "grad_norm": 0.5737484693527222, "learning_rate": 5.965516710393796e-05, "loss": 0.0331, "step": 37300 }, { "grad_norm": 0.4264421761035919, "learning_rate": 5.963488691824516e-05, "loss": 0.0456, "step": 37310 }, { "grad_norm": 0.4370018243789673, "learning_rate": 5.96146050862137e-05, "loss": 0.0432, "step": 37320 }, { "grad_norm": 0.5500684976577759, "learning_rate": 5.959432161130919e-05, "loss": 0.0368, "step": 37330 }, { "grad_norm": 0.4254685640335083, "learning_rate": 5.9574036496997545e-05, "loss": 0.0403, "step": 37340 }, { "grad_norm": 0.4495697617530823, "learning_rate": 5.955374974674488e-05, "loss": 0.0355, "step": 37350 }, { "grad_norm": 0.40841034054756165, "learning_rate": 5.9533461364017696e-05, "loss": 0.0336, "step": 37360 }, { "grad_norm": 0.3478676974773407, "learning_rate": 5.9513171352282716e-05, "loss": 0.035, "step": 37370 }, { "grad_norm": 0.4024178087711334, "learning_rate": 5.949287971500692e-05, "loss": 0.0301, "step": 37380 }, { "grad_norm": 0.2943178713321686, "learning_rate": 5.947258645565762e-05, "loss": 0.0276, "step": 37390 }, { "grad_norm": 0.5668132305145264, "learning_rate": 5.945229157770237e-05, "loss": 0.0332, "step": 37400 }, { "grad_norm": 0.43798181414604187, "learning_rate": 5.9431995084609006e-05, "loss": 0.04, "step": 37410 }, { "grad_norm": 0.33436107635498047, "learning_rate": 5.941169697984564e-05, "loss": 0.0284, "step": 37420 }, { "grad_norm": 0.4660167098045349, "learning_rate": 5.9391397266880675e-05, "loss": 0.0413, "step": 37430 }, { "grad_norm": 0.36666661500930786, "learning_rate": 5.937109594918279e-05, "loss": 0.0492, "step": 37440 }, { "grad_norm": 0.29375991225242615, "learning_rate": 5.9350793030220884e-05, "loss": 0.0509, "step": 37450 }, { "grad_norm": 0.37668654322624207, "learning_rate": 5.933048851346421e-05, "loss": 0.0328, "step": 37460 }, { "grad_norm": 0.3114556670188904, "learning_rate": 5.931018240238224e-05, "loss": 0.0492, "step": 37470 }, { "grad_norm": 0.4046109914779663, "learning_rate": 5.928987470044471e-05, "loss": 0.0295, "step": 37480 }, { "grad_norm": 0.35506877303123474, "learning_rate": 5.9269565411121695e-05, "loss": 0.0371, "step": 37490 }, { "grad_norm": 0.4148198366165161, "learning_rate": 5.924925453788347e-05, "loss": 0.0353, "step": 37500 }, { "grad_norm": 0.33605024218559265, "learning_rate": 5.92289420842006e-05, "loss": 0.0409, "step": 37510 }, { "grad_norm": 0.47036877274513245, "learning_rate": 5.9208628053543945e-05, "loss": 0.0485, "step": 37520 }, { "grad_norm": 0.4281199276447296, "learning_rate": 5.918831244938462e-05, "loss": 0.0307, "step": 37530 }, { "grad_norm": 0.4572729468345642, "learning_rate": 5.9167995275194e-05, "loss": 0.0415, "step": 37540 }, { "grad_norm": 0.22301062941551208, "learning_rate": 5.914767653444373e-05, "loss": 0.0346, "step": 37550 }, { "grad_norm": 0.4953862130641937, "learning_rate": 5.912735623060572e-05, "loss": 0.0376, "step": 37560 }, { "grad_norm": 0.3272532522678375, "learning_rate": 5.910703436715217e-05, "loss": 0.034, "step": 37570 }, { "grad_norm": 0.3066650927066803, "learning_rate": 5.908671094755552e-05, "loss": 0.0329, "step": 37580 }, { "grad_norm": 0.34000179171562195, "learning_rate": 5.906638597528851e-05, "loss": 0.032, "step": 37590 }, { "grad_norm": 0.34833550453186035, "learning_rate": 5.9046059453824076e-05, "loss": 0.032, "step": 37600 }, { "grad_norm": 0.4455796778202057, "learning_rate": 5.9025731386635505e-05, "loss": 0.0351, "step": 37610 }, { "grad_norm": 0.3011871874332428, "learning_rate": 5.900540177719629e-05, "loss": 0.0308, "step": 37620 }, { "grad_norm": 0.46187740564346313, "learning_rate": 5.898507062898021e-05, "loss": 0.0332, "step": 37630 }, { "grad_norm": 0.29264554381370544, "learning_rate": 5.8964737945461316e-05, "loss": 0.0394, "step": 37640 }, { "grad_norm": 0.25104820728302, "learning_rate": 5.8944403730113885e-05, "loss": 0.0357, "step": 37650 }, { "grad_norm": 0.3815864324569702, "learning_rate": 5.892406798641248e-05, "loss": 0.0403, "step": 37660 }, { "grad_norm": 0.4424167573451996, "learning_rate": 5.890373071783193e-05, "loss": 0.0318, "step": 37670 }, { "grad_norm": 0.3425232470035553, "learning_rate": 5.888339192784732e-05, "loss": 0.0325, "step": 37680 }, { "grad_norm": 0.3386748135089874, "learning_rate": 5.8863051619934003e-05, "loss": 0.0369, "step": 37690 }, { "grad_norm": 0.4019509553909302, "learning_rate": 5.8842709797567554e-05, "loss": 0.0269, "step": 37700 }, { "grad_norm": 0.38067322969436646, "learning_rate": 5.8822366464223855e-05, "loss": 0.0316, "step": 37710 }, { "grad_norm": 0.4196341037750244, "learning_rate": 5.880202162337901e-05, "loss": 0.0488, "step": 37720 }, { "grad_norm": 0.33140844106674194, "learning_rate": 5.8781675278509405e-05, "loss": 0.0316, "step": 37730 }, { "grad_norm": 0.3429969251155853, "learning_rate": 5.8761327433091696e-05, "loss": 0.0312, "step": 37740 }, { "grad_norm": 0.4022177755832672, "learning_rate": 5.874097809060275e-05, "loss": 0.044, "step": 37750 }, { "grad_norm": 0.6057506203651428, "learning_rate": 5.87206272545197e-05, "loss": 0.0544, "step": 37760 }, { "grad_norm": 0.3500456213951111, "learning_rate": 5.8700274928319955e-05, "loss": 0.0425, "step": 37770 }, { "grad_norm": 0.44542205333709717, "learning_rate": 5.867992111548118e-05, "loss": 0.0366, "step": 37780 }, { "grad_norm": 0.3334999978542328, "learning_rate": 5.865956581948131e-05, "loss": 0.0344, "step": 37790 }, { "grad_norm": 0.2599007189273834, "learning_rate": 5.863920904379845e-05, "loss": 0.0311, "step": 37800 }, { "grad_norm": 0.4934954345226288, "learning_rate": 5.8618850791911064e-05, "loss": 0.041, "step": 37810 }, { "grad_norm": 0.406710147857666, "learning_rate": 5.859849106729779e-05, "loss": 0.0367, "step": 37820 }, { "grad_norm": 0.9250608682632446, "learning_rate": 5.857812987343758e-05, "loss": 0.0372, "step": 37830 }, { "grad_norm": 0.38338229060173035, "learning_rate": 5.855776721380957e-05, "loss": 0.04, "step": 37840 }, { "grad_norm": 0.32672131061553955, "learning_rate": 5.8537403091893217e-05, "loss": 0.0312, "step": 37850 }, { "grad_norm": 0.45459383726119995, "learning_rate": 5.851703751116816e-05, "loss": 0.0401, "step": 37860 }, { "grad_norm": 0.33790865540504456, "learning_rate": 5.8496670475114336e-05, "loss": 0.029, "step": 37870 }, { "grad_norm": 0.3048189580440521, "learning_rate": 5.84763019872119e-05, "loss": 0.0442, "step": 37880 }, { "grad_norm": 0.3137475848197937, "learning_rate": 5.845593205094131e-05, "loss": 0.034, "step": 37890 }, { "grad_norm": 0.36551931500434875, "learning_rate": 5.843556066978318e-05, "loss": 0.0365, "step": 37900 }, { "grad_norm": 0.3311881124973297, "learning_rate": 5.8415187847218455e-05, "loss": 0.0308, "step": 37910 }, { "grad_norm": 0.572693407535553, "learning_rate": 5.839481358672827e-05, "loss": 0.0329, "step": 37920 }, { "grad_norm": 0.3020082712173462, "learning_rate": 5.837443789179407e-05, "loss": 0.0413, "step": 37930 }, { "grad_norm": 0.31314516067504883, "learning_rate": 5.8354060765897445e-05, "loss": 0.0366, "step": 37940 }, { "grad_norm": 0.33544352650642395, "learning_rate": 5.8333682212520334e-05, "loss": 0.0443, "step": 37950 }, { "grad_norm": 0.48461735248565674, "learning_rate": 5.831330223514486e-05, "loss": 0.0409, "step": 37960 }, { "grad_norm": 0.30472299456596375, "learning_rate": 5.8292920837253396e-05, "loss": 0.0331, "step": 37970 }, { "grad_norm": 0.341264933347702, "learning_rate": 5.827253802232857e-05, "loss": 0.0348, "step": 37980 }, { "grad_norm": 0.44931885600090027, "learning_rate": 5.825215379385327e-05, "loss": 0.0326, "step": 37990 }, { "grad_norm": 0.3341763913631439, "learning_rate": 5.823176815531057e-05, "loss": 0.0452, "step": 38000 }, { "grad_norm": 0.25852033495903015, "learning_rate": 5.8211381110183826e-05, "loss": 0.0349, "step": 38010 }, { "grad_norm": 0.4641020596027374, "learning_rate": 5.8190992661956645e-05, "loss": 0.0264, "step": 38020 }, { "grad_norm": 0.2823946177959442, "learning_rate": 5.817060281411284e-05, "loss": 0.0325, "step": 38030 }, { "grad_norm": 0.3134869933128357, "learning_rate": 5.815021157013647e-05, "loss": 0.0305, "step": 38040 }, { "grad_norm": 0.3871088922023773, "learning_rate": 5.8129818933511856e-05, "loss": 0.0423, "step": 38050 }, { "grad_norm": 0.4373658001422882, "learning_rate": 5.8109424907723544e-05, "loss": 0.0361, "step": 38060 }, { "grad_norm": 0.3827882409095764, "learning_rate": 5.80890294962563e-05, "loss": 0.0297, "step": 38070 }, { "grad_norm": 0.45191138982772827, "learning_rate": 5.806863270259515e-05, "loss": 0.0418, "step": 38080 }, { "grad_norm": 1.1160391569137573, "learning_rate": 5.804823453022536e-05, "loss": 0.0368, "step": 38090 }, { "grad_norm": 0.3892521858215332, "learning_rate": 5.80278349826324e-05, "loss": 0.0358, "step": 38100 }, { "grad_norm": 0.37652283906936646, "learning_rate": 5.8007434063302014e-05, "loss": 0.0321, "step": 38110 }, { "grad_norm": 0.3921974301338196, "learning_rate": 5.7987031775720136e-05, "loss": 0.0316, "step": 38120 }, { "grad_norm": 0.3308081030845642, "learning_rate": 5.7966628123372976e-05, "loss": 0.044, "step": 38130 }, { "grad_norm": 0.521110475063324, "learning_rate": 5.7946223109746956e-05, "loss": 0.0462, "step": 38140 }, { "grad_norm": 0.5852583050727844, "learning_rate": 5.7925816738328754e-05, "loss": 0.0365, "step": 38150 }, { "grad_norm": 0.3907223045825958, "learning_rate": 5.790540901260521e-05, "loss": 0.0357, "step": 38160 }, { "grad_norm": 0.3514499366283417, "learning_rate": 5.788499993606351e-05, "loss": 0.032, "step": 38170 }, { "grad_norm": 0.2966075539588928, "learning_rate": 5.786458951219096e-05, "loss": 0.0337, "step": 38180 }, { "grad_norm": 0.3246433436870575, "learning_rate": 5.784417774447517e-05, "loss": 0.0392, "step": 38190 }, { "grad_norm": 0.3052370846271515, "learning_rate": 5.782376463640393e-05, "loss": 0.0357, "step": 38200 }, { "grad_norm": 0.5010147094726562, "learning_rate": 5.780335019146531e-05, "loss": 0.0382, "step": 38210 }, { "grad_norm": 0.2566237151622772, "learning_rate": 5.778293441314755e-05, "loss": 0.0321, "step": 38220 }, { "grad_norm": 0.38808444142341614, "learning_rate": 5.776251730493917e-05, "loss": 0.0415, "step": 38230 }, { "grad_norm": 0.36486777663230896, "learning_rate": 5.774209887032887e-05, "loss": 0.0381, "step": 38240 }, { "grad_norm": 0.24173875153064728, "learning_rate": 5.772167911280565e-05, "loss": 0.0311, "step": 38250 }, { "grad_norm": 0.49535050988197327, "learning_rate": 5.770125803585864e-05, "loss": 0.0277, "step": 38260 }, { "grad_norm": 0.307878315448761, "learning_rate": 5.768083564297726e-05, "loss": 0.0387, "step": 38270 }, { "grad_norm": 0.4669781029224396, "learning_rate": 5.766041193765114e-05, "loss": 0.0334, "step": 38280 }, { "grad_norm": 0.4509091079235077, "learning_rate": 5.763998692337015e-05, "loss": 0.0292, "step": 38290 }, { "grad_norm": 0.3590562343597412, "learning_rate": 5.761956060362433e-05, "loss": 0.0391, "step": 38300 }, { "grad_norm": 0.28099197149276733, "learning_rate": 5.7599132981904e-05, "loss": 0.0369, "step": 38310 }, { "grad_norm": 0.5022881031036377, "learning_rate": 5.75787040616997e-05, "loss": 0.0457, "step": 38320 }, { "grad_norm": 0.3895176947116852, "learning_rate": 5.755827384650212e-05, "loss": 0.0433, "step": 38330 }, { "grad_norm": 0.3657049834728241, "learning_rate": 5.753784233980228e-05, "loss": 0.0374, "step": 38340 }, { "grad_norm": 0.31966909766197205, "learning_rate": 5.751740954509135e-05, "loss": 0.0261, "step": 38350 }, { "grad_norm": 0.26563572883605957, "learning_rate": 5.7496975465860715e-05, "loss": 0.028, "step": 38360 }, { "grad_norm": 0.24607670307159424, "learning_rate": 5.747654010560202e-05, "loss": 0.033, "step": 38370 }, { "grad_norm": 0.2658368647098541, "learning_rate": 5.7456103467807097e-05, "loss": 0.0381, "step": 38380 }, { "grad_norm": 0.33906450867652893, "learning_rate": 5.7435665555968046e-05, "loss": 0.0313, "step": 38390 }, { "grad_norm": 0.3559724688529968, "learning_rate": 5.74152263735771e-05, "loss": 0.0361, "step": 38400 }, { "grad_norm": 0.32126328349113464, "learning_rate": 5.739478592412677e-05, "loss": 0.0336, "step": 38410 }, { "grad_norm": 0.6394398808479309, "learning_rate": 5.7374344211109766e-05, "loss": 0.0444, "step": 38420 }, { "grad_norm": 0.34873342514038086, "learning_rate": 5.735390123801904e-05, "loss": 0.0386, "step": 38430 }, { "grad_norm": 0.3450757563114166, "learning_rate": 5.7333457008347704e-05, "loss": 0.0413, "step": 38440 }, { "grad_norm": 0.33579644560813904, "learning_rate": 5.7313011525589156e-05, "loss": 0.0348, "step": 38450 }, { "grad_norm": 0.2784176170825958, "learning_rate": 5.729256479323694e-05, "loss": 0.0395, "step": 38460 }, { "grad_norm": 0.40680423378944397, "learning_rate": 5.727211681478485e-05, "loss": 0.0376, "step": 38470 }, { "grad_norm": 0.36438149213790894, "learning_rate": 5.7251667593726886e-05, "loss": 0.039, "step": 38480 }, { "grad_norm": 0.29555341601371765, "learning_rate": 5.723121713355728e-05, "loss": 0.0432, "step": 38490 }, { "grad_norm": 0.4728519916534424, "learning_rate": 5.721076543777044e-05, "loss": 0.038, "step": 38500 }, { "grad_norm": 0.560916006565094, "learning_rate": 5.7190312509860986e-05, "loss": 0.0391, "step": 38510 }, { "grad_norm": 0.29848289489746094, "learning_rate": 5.716985835332379e-05, "loss": 0.0385, "step": 38520 }, { "grad_norm": 0.3625255227088928, "learning_rate": 5.714940297165389e-05, "loss": 0.0318, "step": 38530 }, { "grad_norm": 0.48234280943870544, "learning_rate": 5.712894636834656e-05, "loss": 0.0322, "step": 38540 }, { "grad_norm": 0.4888463318347931, "learning_rate": 5.7108488546897276e-05, "loss": 0.0331, "step": 38550 }, { "grad_norm": 0.2641024887561798, "learning_rate": 5.708802951080172e-05, "loss": 0.0288, "step": 38560 }, { "grad_norm": 0.3097371757030487, "learning_rate": 5.706756926355576e-05, "loss": 0.0359, "step": 38570 }, { "grad_norm": 0.31698814034461975, "learning_rate": 5.704710780865554e-05, "loss": 0.0347, "step": 38580 }, { "grad_norm": 0.3566191494464874, "learning_rate": 5.7026645149597325e-05, "loss": 0.0312, "step": 38590 }, { "grad_norm": 0.3561563789844513, "learning_rate": 5.700618128987764e-05, "loss": 0.0388, "step": 38600 }, { "grad_norm": 0.31961455941200256, "learning_rate": 5.698571623299317e-05, "loss": 0.0339, "step": 38610 }, { "grad_norm": 0.3248055875301361, "learning_rate": 5.696524998244086e-05, "loss": 0.0304, "step": 38620 }, { "grad_norm": 0.34267672896385193, "learning_rate": 5.6944782541717836e-05, "loss": 0.0454, "step": 38630 }, { "grad_norm": 0.39169564843177795, "learning_rate": 5.69243139143214e-05, "loss": 0.0274, "step": 38640 }, { "grad_norm": 0.2430994063615799, "learning_rate": 5.6903844103749125e-05, "loss": 0.0364, "step": 38650 }, { "grad_norm": 0.3879821300506592, "learning_rate": 5.688337311349869e-05, "loss": 0.0316, "step": 38660 }, { "grad_norm": 0.3355487287044525, "learning_rate": 5.6862900947068074e-05, "loss": 0.0308, "step": 38670 }, { "grad_norm": 0.27314111590385437, "learning_rate": 5.6842427607955374e-05, "loss": 0.04, "step": 38680 }, { "grad_norm": 0.25838354229927063, "learning_rate": 5.682195309965893e-05, "loss": 0.0246, "step": 38690 }, { "grad_norm": 0.3175004720687866, "learning_rate": 5.6801477425677294e-05, "loss": 0.0356, "step": 38700 }, { "grad_norm": 0.3562730848789215, "learning_rate": 5.678100058950917e-05, "loss": 0.0403, "step": 38710 }, { "grad_norm": 0.42183470726013184, "learning_rate": 5.676052259465352e-05, "loss": 0.0346, "step": 38720 }, { "grad_norm": 0.4138879179954529, "learning_rate": 5.674004344460945e-05, "loss": 0.0322, "step": 38730 }, { "grad_norm": 0.4832630753517151, "learning_rate": 5.6719563142876295e-05, "loss": 0.0383, "step": 38740 }, { "grad_norm": 0.3980237543582916, "learning_rate": 5.669908169295359e-05, "loss": 0.0356, "step": 38750 }, { "grad_norm": 0.4157692492008209, "learning_rate": 5.667859909834105e-05, "loss": 0.0356, "step": 38760 }, { "grad_norm": 0.39653271436691284, "learning_rate": 5.6658115362538565e-05, "loss": 0.036, "step": 38770 }, { "grad_norm": 0.3311317265033722, "learning_rate": 5.663763048904628e-05, "loss": 0.0335, "step": 38780 }, { "grad_norm": 0.4338502585887909, "learning_rate": 5.661714448136447e-05, "loss": 0.027, "step": 38790 }, { "grad_norm": 0.35068562626838684, "learning_rate": 5.659665734299366e-05, "loss": 0.0376, "step": 38800 }, { "grad_norm": 0.4155128598213196, "learning_rate": 5.6576169077434516e-05, "loss": 0.0414, "step": 38810 }, { "grad_norm": 0.2580190598964691, "learning_rate": 5.6555679688187944e-05, "loss": 0.0329, "step": 38820 }, { "grad_norm": 0.5968492031097412, "learning_rate": 5.6535189178755e-05, "loss": 0.0381, "step": 38830 }, { "grad_norm": 0.4104914367198944, "learning_rate": 5.651469755263698e-05, "loss": 0.0398, "step": 38840 }, { "grad_norm": 0.20434586703777313, "learning_rate": 5.6494204813335316e-05, "loss": 0.0395, "step": 38850 }, { "grad_norm": 0.3531784415245056, "learning_rate": 5.647371096435168e-05, "loss": 0.0272, "step": 38860 }, { "grad_norm": 0.3077331781387329, "learning_rate": 5.645321600918788e-05, "loss": 0.0281, "step": 38870 }, { "grad_norm": 0.4081142544746399, "learning_rate": 5.643271995134597e-05, "loss": 0.0369, "step": 38880 }, { "grad_norm": 0.33844342827796936, "learning_rate": 5.641222279432814e-05, "loss": 0.0343, "step": 38890 }, { "grad_norm": 0.2949662506580353, "learning_rate": 5.6391724541636834e-05, "loss": 0.0329, "step": 38900 }, { "grad_norm": 0.3442774713039398, "learning_rate": 5.6371225196774605e-05, "loss": 0.0338, "step": 38910 }, { "grad_norm": 0.30480754375457764, "learning_rate": 5.635072476324423e-05, "loss": 0.04, "step": 38920 }, { "grad_norm": 0.4972060024738312, "learning_rate": 5.63302232445487e-05, "loss": 0.0314, "step": 38930 }, { "grad_norm": 0.3695453405380249, "learning_rate": 5.6309720644191144e-05, "loss": 0.0348, "step": 38940 }, { "grad_norm": 0.43388232588768005, "learning_rate": 5.628921696567491e-05, "loss": 0.028, "step": 38950 }, { "grad_norm": 0.4569739103317261, "learning_rate": 5.62687122125035e-05, "loss": 0.0312, "step": 38960 }, { "grad_norm": 0.4501401484012604, "learning_rate": 5.624820638818062e-05, "loss": 0.0301, "step": 38970 }, { "grad_norm": 0.3402929902076721, "learning_rate": 5.6227699496210164e-05, "loss": 0.0373, "step": 38980 }, { "grad_norm": 0.39102447032928467, "learning_rate": 5.6207191540096195e-05, "loss": 0.0355, "step": 38990 }, { "grad_norm": 0.24098463356494904, "learning_rate": 5.618668252334296e-05, "loss": 0.0337, "step": 39000 }, { "grad_norm": 0.32502031326293945, "learning_rate": 5.616617244945488e-05, "loss": 0.029, "step": 39010 }, { "grad_norm": 0.2808416485786438, "learning_rate": 5.614566132193656e-05, "loss": 0.0274, "step": 39020 }, { "grad_norm": 0.32804474234580994, "learning_rate": 5.612514914429282e-05, "loss": 0.0327, "step": 39030 }, { "grad_norm": 0.2961147129535675, "learning_rate": 5.610463592002863e-05, "loss": 0.033, "step": 39040 }, { "grad_norm": 0.49307066202163696, "learning_rate": 5.608412165264909e-05, "loss": 0.0383, "step": 39050 }, { "grad_norm": 0.35804012417793274, "learning_rate": 5.606360634565959e-05, "loss": 0.0276, "step": 39060 }, { "grad_norm": 0.5558376908302307, "learning_rate": 5.604309000256559e-05, "loss": 0.0326, "step": 39070 }, { "grad_norm": 0.4027444124221802, "learning_rate": 5.6022572626872785e-05, "loss": 0.0351, "step": 39080 }, { "grad_norm": 0.42859938740730286, "learning_rate": 5.600205422208704e-05, "loss": 0.0423, "step": 39090 }, { "grad_norm": 0.3585583567619324, "learning_rate": 5.59815347917144e-05, "loss": 0.0325, "step": 39100 }, { "grad_norm": 0.575927734375, "learning_rate": 5.596101433926103e-05, "loss": 0.0365, "step": 39110 }, { "grad_norm": 0.4469166398048401, "learning_rate": 5.5940492868233364e-05, "loss": 0.0273, "step": 39120 }, { "grad_norm": 0.3751949965953827, "learning_rate": 5.591997038213793e-05, "loss": 0.028, "step": 39130 }, { "grad_norm": 0.4823097586631775, "learning_rate": 5.5899446884481475e-05, "loss": 0.028, "step": 39140 }, { "grad_norm": 0.2544409930706024, "learning_rate": 5.5878922378770906e-05, "loss": 0.0487, "step": 39150 }, { "grad_norm": 0.4004693925380707, "learning_rate": 5.5858396868513285e-05, "loss": 0.0343, "step": 39160 }, { "grad_norm": 0.44355666637420654, "learning_rate": 5.583787035721586e-05, "loss": 0.0288, "step": 39170 }, { "grad_norm": 0.33117392659187317, "learning_rate": 5.581734284838606e-05, "loss": 0.0319, "step": 39180 }, { "grad_norm": 0.3255082964897156, "learning_rate": 5.579681434553147e-05, "loss": 0.0343, "step": 39190 }, { "grad_norm": 0.3549135625362396, "learning_rate": 5.5776284852159854e-05, "loss": 0.0328, "step": 39200 }, { "grad_norm": 0.393718421459198, "learning_rate": 5.575575437177913e-05, "loss": 0.0346, "step": 39210 }, { "grad_norm": 0.3549710512161255, "learning_rate": 5.573522290789742e-05, "loss": 0.0282, "step": 39220 }, { "grad_norm": 0.32251280546188354, "learning_rate": 5.571469046402298e-05, "loss": 0.0332, "step": 39230 }, { "grad_norm": 0.35402852296829224, "learning_rate": 5.5694157043664205e-05, "loss": 0.0329, "step": 39240 }, { "grad_norm": 0.31402072310447693, "learning_rate": 5.567362265032975e-05, "loss": 0.0289, "step": 39250 }, { "grad_norm": 0.3393152356147766, "learning_rate": 5.565308728752836e-05, "loss": 0.032, "step": 39260 }, { "grad_norm": 0.3029457628726959, "learning_rate": 5.5632550958768956e-05, "loss": 0.0278, "step": 39270 }, { "grad_norm": 0.3583279252052307, "learning_rate": 5.5612013667560636e-05, "loss": 0.0314, "step": 39280 }, { "grad_norm": 0.6250771880149841, "learning_rate": 5.5591475417412676e-05, "loss": 0.0568, "step": 39290 }, { "grad_norm": 0.4581860899925232, "learning_rate": 5.557093621183451e-05, "loss": 0.0295, "step": 39300 }, { "grad_norm": 0.2959786057472229, "learning_rate": 5.55503960543357e-05, "loss": 0.0304, "step": 39310 }, { "grad_norm": 0.2520996332168579, "learning_rate": 5.552985494842601e-05, "loss": 0.031, "step": 39320 }, { "grad_norm": 0.6801066994667053, "learning_rate": 5.550931289761534e-05, "loss": 0.0355, "step": 39330 }, { "grad_norm": 0.3976683020591736, "learning_rate": 5.548876990541378e-05, "loss": 0.0394, "step": 39340 }, { "grad_norm": 0.3473523259162903, "learning_rate": 5.5468225975331564e-05, "loss": 0.0419, "step": 39350 }, { "grad_norm": 0.34637531638145447, "learning_rate": 5.544768111087909e-05, "loss": 0.0337, "step": 39360 }, { "grad_norm": 0.35795900225639343, "learning_rate": 5.54271353155669e-05, "loss": 0.0392, "step": 39370 }, { "grad_norm": 0.5974076986312866, "learning_rate": 5.5406588592905715e-05, "loss": 0.0347, "step": 39380 }, { "grad_norm": 0.2745848596096039, "learning_rate": 5.5386040946406416e-05, "loss": 0.0382, "step": 39390 }, { "grad_norm": 0.46329909563064575, "learning_rate": 5.536549237958004e-05, "loss": 0.0474, "step": 39400 }, { "grad_norm": 0.4301801323890686, "learning_rate": 5.5344942895937744e-05, "loss": 0.0506, "step": 39410 }, { "grad_norm": 0.4473715126514435, "learning_rate": 5.5324392498990904e-05, "loss": 0.0306, "step": 39420 }, { "grad_norm": 0.3832473158836365, "learning_rate": 5.5303841192251e-05, "loss": 0.0345, "step": 39430 }, { "grad_norm": 0.3106936514377594, "learning_rate": 5.52832889792297e-05, "loss": 0.035, "step": 39440 }, { "grad_norm": 0.48765772581100464, "learning_rate": 5.526273586343881e-05, "loss": 0.0296, "step": 39450 }, { "grad_norm": 0.45415329933166504, "learning_rate": 5.5242181848390306e-05, "loss": 0.0344, "step": 39460 }, { "grad_norm": 0.44659391045570374, "learning_rate": 5.5221626937596285e-05, "loss": 0.026, "step": 39470 }, { "grad_norm": 0.5110980272293091, "learning_rate": 5.520107113456903e-05, "loss": 0.0359, "step": 39480 }, { "grad_norm": 0.2972685992717743, "learning_rate": 5.5180514442820974e-05, "loss": 0.0366, "step": 39490 }, { "grad_norm": 0.31319355964660645, "learning_rate": 5.515995686586469e-05, "loss": 0.0272, "step": 39500 }, { "grad_norm": 0.4555009603500366, "learning_rate": 5.5139398407212916e-05, "loss": 0.0266, "step": 39510 }, { "grad_norm": 0.3372078835964203, "learning_rate": 5.511883907037849e-05, "loss": 0.0328, "step": 39520 }, { "grad_norm": 0.46201959252357483, "learning_rate": 5.509827885887449e-05, "loss": 0.0375, "step": 39530 }, { "grad_norm": 0.33458030223846436, "learning_rate": 5.507771777621406e-05, "loss": 0.0366, "step": 39540 }, { "grad_norm": 0.40395587682724, "learning_rate": 5.505715582591052e-05, "loss": 0.0402, "step": 39550 }, { "grad_norm": 0.5155742764472961, "learning_rate": 5.50365930114774e-05, "loss": 0.0355, "step": 39560 }, { "grad_norm": 0.2790762186050415, "learning_rate": 5.5016029336428255e-05, "loss": 0.0288, "step": 39570 }, { "grad_norm": 0.33211714029312134, "learning_rate": 5.49954648042769e-05, "loss": 0.0374, "step": 39580 }, { "grad_norm": 0.3199204206466675, "learning_rate": 5.4974899418537226e-05, "loss": 0.0432, "step": 39590 }, { "grad_norm": 0.7683839201927185, "learning_rate": 5.4954333182723316e-05, "loss": 0.0371, "step": 39600 }, { "grad_norm": 0.372227281332016, "learning_rate": 5.493376610034937e-05, "loss": 0.0294, "step": 39610 }, { "grad_norm": 0.47379106283187866, "learning_rate": 5.4913198174929735e-05, "loss": 0.0356, "step": 39620 }, { "grad_norm": 0.31305059790611267, "learning_rate": 5.489262940997889e-05, "loss": 0.0345, "step": 39630 }, { "grad_norm": 0.46162012219429016, "learning_rate": 5.487205980901151e-05, "loss": 0.0494, "step": 39640 }, { "grad_norm": 0.4948415756225586, "learning_rate": 5.485148937554234e-05, "loss": 0.055, "step": 39650 }, { "grad_norm": 0.3610188364982605, "learning_rate": 5.483091811308635e-05, "loss": 0.0327, "step": 39660 }, { "grad_norm": 0.32787370681762695, "learning_rate": 5.4810346025158564e-05, "loss": 0.0343, "step": 39670 }, { "grad_norm": 0.4288881719112396, "learning_rate": 5.478977311527421e-05, "loss": 0.0323, "step": 39680 }, { "grad_norm": 0.3038208782672882, "learning_rate": 5.476919938694863e-05, "loss": 0.032, "step": 39690 }, { "grad_norm": 0.347931444644928, "learning_rate": 5.474862484369733e-05, "loss": 0.0248, "step": 39700 }, { "grad_norm": 0.5637123584747314, "learning_rate": 5.472804948903589e-05, "loss": 0.0306, "step": 39710 }, { "grad_norm": 0.27155598998069763, "learning_rate": 5.470747332648013e-05, "loss": 0.0293, "step": 39720 }, { "grad_norm": 0.29969242215156555, "learning_rate": 5.468689635954591e-05, "loss": 0.0487, "step": 39730 }, { "grad_norm": 0.28792354464530945, "learning_rate": 5.46663185917493e-05, "loss": 0.0345, "step": 39740 }, { "grad_norm": 0.4444187581539154, "learning_rate": 5.464574002660645e-05, "loss": 0.0417, "step": 39750 }, { "grad_norm": 0.190573051571846, "learning_rate": 5.4625160667633724e-05, "loss": 0.0321, "step": 39760 }, { "grad_norm": 0.3050972819328308, "learning_rate": 5.4604580518347505e-05, "loss": 0.0354, "step": 39770 }, { "grad_norm": 0.3596509099006653, "learning_rate": 5.458399958226442e-05, "loss": 0.0316, "step": 39780 }, { "grad_norm": 0.268395334482193, "learning_rate": 5.456341786290119e-05, "loss": 0.0369, "step": 39790 }, { "grad_norm": 0.394582599401474, "learning_rate": 5.454283536377465e-05, "loss": 0.0396, "step": 39800 }, { "grad_norm": 0.26822030544281006, "learning_rate": 5.452225208840179e-05, "loss": 0.0374, "step": 39810 }, { "grad_norm": 0.28889280557632446, "learning_rate": 5.450166804029975e-05, "loss": 0.0366, "step": 39820 }, { "grad_norm": 0.5459140539169312, "learning_rate": 5.448108322298574e-05, "loss": 0.028, "step": 39830 }, { "grad_norm": 0.48777130246162415, "learning_rate": 5.446049763997717e-05, "loss": 0.0327, "step": 39840 }, { "grad_norm": 0.3077956736087799, "learning_rate": 5.4439911294791546e-05, "loss": 0.0315, "step": 39850 }, { "grad_norm": 0.37226828932762146, "learning_rate": 5.441932419094652e-05, "loss": 0.0334, "step": 39860 }, { "grad_norm": 0.2944977581501007, "learning_rate": 5.439873633195985e-05, "loss": 0.0326, "step": 39870 }, { "grad_norm": 0.41627341508865356, "learning_rate": 5.437814772134947e-05, "loss": 0.0389, "step": 39880 }, { "grad_norm": 0.37051114439964294, "learning_rate": 5.4357558362633366e-05, "loss": 0.0365, "step": 39890 }, { "grad_norm": 0.33301234245300293, "learning_rate": 5.4336968259329726e-05, "loss": 0.0249, "step": 39900 }, { "grad_norm": 0.29857876896858215, "learning_rate": 5.431637741495681e-05, "loss": 0.0345, "step": 39910 }, { "grad_norm": 0.47048357129096985, "learning_rate": 5.429578583303307e-05, "loss": 0.0347, "step": 39920 }, { "grad_norm": 0.4043743312358856, "learning_rate": 5.427519351707701e-05, "loss": 0.0342, "step": 39930 }, { "grad_norm": 0.4101112484931946, "learning_rate": 5.42546004706073e-05, "loss": 0.0354, "step": 39940 }, { "grad_norm": 0.34230080246925354, "learning_rate": 5.4234006697142735e-05, "loss": 0.0339, "step": 39950 }, { "grad_norm": 0.25008824467658997, "learning_rate": 5.421341220020224e-05, "loss": 0.038, "step": 39960 }, { "grad_norm": 0.24710741639137268, "learning_rate": 5.419281698330482e-05, "loss": 0.0365, "step": 39970 }, { "grad_norm": 0.31625640392303467, "learning_rate": 5.4172221049969665e-05, "loss": 0.0352, "step": 39980 }, { "grad_norm": 0.38397061824798584, "learning_rate": 5.415162440371604e-05, "loss": 0.0336, "step": 39990 }, { "grad_norm": 0.42559394240379333, "learning_rate": 5.413102704806334e-05, "loss": 0.0388, "step": 40000 }, { "grad_norm": 0.32440367341041565, "learning_rate": 5.41104289865311e-05, "loss": 0.0321, "step": 40010 }, { "grad_norm": 0.25092101097106934, "learning_rate": 5.408983022263898e-05, "loss": 0.0309, "step": 40020 }, { "grad_norm": 0.3658304512500763, "learning_rate": 5.406923075990673e-05, "loss": 0.0361, "step": 40030 }, { "grad_norm": 0.2360789179801941, "learning_rate": 5.404863060185423e-05, "loss": 0.0311, "step": 40040 }, { "grad_norm": 0.46211543679237366, "learning_rate": 5.40280297520015e-05, "loss": 0.0362, "step": 40050 }, { "grad_norm": 0.39636850357055664, "learning_rate": 5.400742821386865e-05, "loss": 0.0323, "step": 40060 }, { "grad_norm": 0.49552711844444275, "learning_rate": 5.398682599097592e-05, "loss": 0.0397, "step": 40070 }, { "grad_norm": 0.39113280177116394, "learning_rate": 5.396622308684367e-05, "loss": 0.0305, "step": 40080 }, { "grad_norm": 0.40337926149368286, "learning_rate": 5.394561950499236e-05, "loss": 0.0373, "step": 40090 }, { "grad_norm": 0.3733592927455902, "learning_rate": 5.39250152489426e-05, "loss": 0.039, "step": 40100 }, { "grad_norm": 0.2307194322347641, "learning_rate": 5.390441032221507e-05, "loss": 0.0276, "step": 40110 }, { "grad_norm": 0.3508322238922119, "learning_rate": 5.388380472833062e-05, "loss": 0.0341, "step": 40120 }, { "grad_norm": 0.2835425138473511, "learning_rate": 5.386319847081016e-05, "loss": 0.036, "step": 40130 }, { "grad_norm": 0.513185441493988, "learning_rate": 5.384259155317473e-05, "loss": 0.0375, "step": 40140 }, { "grad_norm": 0.4157317280769348, "learning_rate": 5.38219839789455e-05, "loss": 0.0372, "step": 40150 }, { "grad_norm": 0.2871297299861908, "learning_rate": 5.380137575164376e-05, "loss": 0.0356, "step": 40160 }, { "grad_norm": 0.3593544065952301, "learning_rate": 5.378076687479085e-05, "loss": 0.0323, "step": 40170 }, { "grad_norm": 0.33883100748062134, "learning_rate": 5.3760157351908305e-05, "loss": 0.0311, "step": 40180 }, { "grad_norm": 0.36147210001945496, "learning_rate": 5.373954718651768e-05, "loss": 0.0341, "step": 40190 }, { "grad_norm": 0.2935299277305603, "learning_rate": 5.371893638214074e-05, "loss": 0.0274, "step": 40200 }, { "grad_norm": 0.44394201040267944, "learning_rate": 5.369832494229927e-05, "loss": 0.0356, "step": 40210 }, { "grad_norm": 0.4162580370903015, "learning_rate": 5.367771287051524e-05, "loss": 0.0336, "step": 40220 }, { "grad_norm": 0.2708948254585266, "learning_rate": 5.365710017031065e-05, "loss": 0.0297, "step": 40230 }, { "grad_norm": 0.3621467053890228, "learning_rate": 5.3636486845207654e-05, "loss": 0.0421, "step": 40240 }, { "grad_norm": 0.28860655426979065, "learning_rate": 5.3615872898728526e-05, "loss": 0.0239, "step": 40250 }, { "grad_norm": 0.4583348035812378, "learning_rate": 5.3595258334395614e-05, "loss": 0.0323, "step": 40260 }, { "grad_norm": 0.3452882468700409, "learning_rate": 5.3574643155731396e-05, "loss": 0.0408, "step": 40270 }, { "grad_norm": 0.29361438751220703, "learning_rate": 5.3554027366258405e-05, "loss": 0.0255, "step": 40280 }, { "grad_norm": 0.3710920214653015, "learning_rate": 5.3533410969499355e-05, "loss": 0.0369, "step": 40290 }, { "grad_norm": 0.32650572061538696, "learning_rate": 5.3512793968977e-05, "loss": 0.0356, "step": 40300 }, { "grad_norm": 0.47587499022483826, "learning_rate": 5.349217636821423e-05, "loss": 0.0452, "step": 40310 }, { "grad_norm": 0.40414804220199585, "learning_rate": 5.347155817073405e-05, "loss": 0.0261, "step": 40320 }, { "grad_norm": 0.4035249650478363, "learning_rate": 5.34509393800595e-05, "loss": 0.0331, "step": 40330 }, { "grad_norm": 0.4030267000198364, "learning_rate": 5.343031999971381e-05, "loss": 0.0306, "step": 40340 }, { "grad_norm": 0.3209514617919922, "learning_rate": 5.340970003322026e-05, "loss": 0.0251, "step": 40350 }, { "grad_norm": 0.3062622845172882, "learning_rate": 5.33890794841022e-05, "loss": 0.0313, "step": 40360 }, { "grad_norm": 0.33886274695396423, "learning_rate": 5.336845835588318e-05, "loss": 0.0294, "step": 40370 }, { "grad_norm": 0.44274193048477173, "learning_rate": 5.334783665208674e-05, "loss": 0.0328, "step": 40380 }, { "grad_norm": 0.31782424449920654, "learning_rate": 5.332721437623657e-05, "loss": 0.0279, "step": 40390 }, { "grad_norm": 0.34605881571769714, "learning_rate": 5.3306591531856464e-05, "loss": 0.0328, "step": 40400 }, { "grad_norm": 0.3329976797103882, "learning_rate": 5.3285968122470295e-05, "loss": 0.0277, "step": 40410 }, { "grad_norm": 0.25420260429382324, "learning_rate": 5.326534415160207e-05, "loss": 0.0325, "step": 40420 }, { "grad_norm": 0.31132110953330994, "learning_rate": 5.324471962277582e-05, "loss": 0.0407, "step": 40430 }, { "grad_norm": 0.4074718952178955, "learning_rate": 5.3224094539515746e-05, "loss": 0.0296, "step": 40440 }, { "grad_norm": 0.2506604790687561, "learning_rate": 5.3203468905346076e-05, "loss": 0.0272, "step": 40450 }, { "grad_norm": 0.2314375638961792, "learning_rate": 5.3182842723791195e-05, "loss": 0.034, "step": 40460 }, { "grad_norm": 0.40872102975845337, "learning_rate": 5.316221599837554e-05, "loss": 0.0338, "step": 40470 }, { "grad_norm": 0.3896690309047699, "learning_rate": 5.314158873262366e-05, "loss": 0.0409, "step": 40480 }, { "grad_norm": 0.3094088137149811, "learning_rate": 5.312096093006018e-05, "loss": 0.0265, "step": 40490 }, { "grad_norm": 0.4217425286769867, "learning_rate": 5.3100332594209846e-05, "loss": 0.0387, "step": 40500 }, { "grad_norm": 0.3481653332710266, "learning_rate": 5.3079703728597454e-05, "loss": 0.0304, "step": 40510 }, { "grad_norm": 0.4337088167667389, "learning_rate": 5.305907433674794e-05, "loss": 0.0313, "step": 40520 }, { "grad_norm": 0.4024776220321655, "learning_rate": 5.3038444422186284e-05, "loss": 0.0371, "step": 40530 }, { "grad_norm": 0.40352287888526917, "learning_rate": 5.301781398843757e-05, "loss": 0.0282, "step": 40540 }, { "grad_norm": 0.3167709708213806, "learning_rate": 5.299718303902699e-05, "loss": 0.036, "step": 40550 }, { "grad_norm": 0.23803944885730743, "learning_rate": 5.297655157747979e-05, "loss": 0.0341, "step": 40560 }, { "grad_norm": 0.4412538409233093, "learning_rate": 5.295591960732136e-05, "loss": 0.0288, "step": 40570 }, { "grad_norm": 0.336606502532959, "learning_rate": 5.293528713207708e-05, "loss": 0.0358, "step": 40580 }, { "grad_norm": 0.28561165928840637, "learning_rate": 5.291465415527253e-05, "loss": 0.035, "step": 40590 }, { "grad_norm": 0.3779652416706085, "learning_rate": 5.2894020680433296e-05, "loss": 0.0321, "step": 40600 }, { "grad_norm": 0.29106974601745605, "learning_rate": 5.287338671108507e-05, "loss": 0.0291, "step": 40610 }, { "grad_norm": 0.4304145872592926, "learning_rate": 5.285275225075367e-05, "loss": 0.0294, "step": 40620 }, { "grad_norm": 0.29002389311790466, "learning_rate": 5.283211730296492e-05, "loss": 0.0273, "step": 40630 }, { "grad_norm": 0.6916865110397339, "learning_rate": 5.281148187124477e-05, "loss": 0.0324, "step": 40640 }, { "grad_norm": 0.3646462559700012, "learning_rate": 5.279084595911927e-05, "loss": 0.0267, "step": 40650 }, { "grad_norm": 0.44628003239631653, "learning_rate": 5.277020957011453e-05, "loss": 0.036, "step": 40660 }, { "grad_norm": 0.33903804421424866, "learning_rate": 5.274957270775673e-05, "loss": 0.0421, "step": 40670 }, { "grad_norm": 0.327956885099411, "learning_rate": 5.2728935375572164e-05, "loss": 0.0315, "step": 40680 }, { "grad_norm": 0.22537413239479065, "learning_rate": 5.2708297577087176e-05, "loss": 0.0333, "step": 40690 }, { "grad_norm": 0.38990819454193115, "learning_rate": 5.2687659315828196e-05, "loss": 0.0363, "step": 40700 }, { "grad_norm": 0.3678536117076874, "learning_rate": 5.266702059532175e-05, "loss": 0.0309, "step": 40710 }, { "grad_norm": 0.3845175802707672, "learning_rate": 5.264638141909444e-05, "loss": 0.0354, "step": 40720 }, { "grad_norm": 0.34861767292022705, "learning_rate": 5.2625741790672925e-05, "loss": 0.0354, "step": 40730 }, { "grad_norm": 0.3970898687839508, "learning_rate": 5.2605101713583935e-05, "loss": 0.0285, "step": 40740 }, { "grad_norm": 0.4022856652736664, "learning_rate": 5.2584461191354315e-05, "loss": 0.0362, "step": 40750 }, { "grad_norm": 0.28629350662231445, "learning_rate": 5.2563820227510964e-05, "loss": 0.0307, "step": 40760 }, { "grad_norm": 0.3054412007331848, "learning_rate": 5.2543178825580864e-05, "loss": 0.0297, "step": 40770 }, { "grad_norm": 0.5257007479667664, "learning_rate": 5.2522536989091054e-05, "loss": 0.0397, "step": 40780 }, { "grad_norm": 0.38647326827049255, "learning_rate": 5.2501894721568655e-05, "loss": 0.0333, "step": 40790 }, { "grad_norm": 0.42603710293769836, "learning_rate": 5.248125202654089e-05, "loss": 0.035, "step": 40800 }, { "grad_norm": 0.35268065333366394, "learning_rate": 5.246060890753501e-05, "loss": 0.0483, "step": 40810 }, { "grad_norm": 0.34098201990127563, "learning_rate": 5.243996536807837e-05, "loss": 0.0328, "step": 40820 }, { "grad_norm": 0.3044204115867615, "learning_rate": 5.2419321411698384e-05, "loss": 0.029, "step": 40830 }, { "grad_norm": 0.31709110736846924, "learning_rate": 5.239867704192253e-05, "loss": 0.0374, "step": 40840 }, { "grad_norm": 0.3698665499687195, "learning_rate": 5.237803226227838e-05, "loss": 0.0298, "step": 40850 }, { "grad_norm": 0.31223657727241516, "learning_rate": 5.235738707629354e-05, "loss": 0.0293, "step": 40860 }, { "grad_norm": 0.3431258499622345, "learning_rate": 5.233674148749575e-05, "loss": 0.0373, "step": 40870 }, { "grad_norm": 0.37363216280937195, "learning_rate": 5.231609549941272e-05, "loss": 0.0401, "step": 40880 }, { "grad_norm": 0.3352501392364502, "learning_rate": 5.229544911557233e-05, "loss": 0.0366, "step": 40890 }, { "grad_norm": 0.4041000306606293, "learning_rate": 5.227480233950246e-05, "loss": 0.0335, "step": 40900 }, { "grad_norm": 0.2631635069847107, "learning_rate": 5.22541551747311e-05, "loss": 0.0297, "step": 40910 }, { "grad_norm": 0.44112369418144226, "learning_rate": 5.223350762478626e-05, "loss": 0.0286, "step": 40920 }, { "grad_norm": 0.23085014522075653, "learning_rate": 5.221285969319606e-05, "loss": 0.0268, "step": 40930 }, { "grad_norm": 0.383521169424057, "learning_rate": 5.219221138348865e-05, "loss": 0.0304, "step": 40940 }, { "grad_norm": 0.3235667943954468, "learning_rate": 5.217156269919228e-05, "loss": 0.0341, "step": 40950 }, { "grad_norm": 0.23965641856193542, "learning_rate": 5.215091364383523e-05, "loss": 0.0291, "step": 40960 }, { "grad_norm": 0.4235501289367676, "learning_rate": 5.213026422094588e-05, "loss": 0.0345, "step": 40970 }, { "grad_norm": 0.3758448362350464, "learning_rate": 5.210961443405262e-05, "loss": 0.033, "step": 40980 }, { "grad_norm": 0.5488965511322021, "learning_rate": 5.208896428668396e-05, "loss": 0.0351, "step": 40990 }, { "grad_norm": 0.4004513621330261, "learning_rate": 5.206831378236845e-05, "loss": 0.036, "step": 41000 }, { "grad_norm": 0.3324032425880432, "learning_rate": 5.2047662924634666e-05, "loss": 0.0327, "step": 41010 }, { "grad_norm": 0.34414395689964294, "learning_rate": 5.20270117170113e-05, "loss": 0.0257, "step": 41020 }, { "grad_norm": 0.44849878549575806, "learning_rate": 5.200636016302707e-05, "loss": 0.0452, "step": 41030 }, { "grad_norm": 0.4394295811653137, "learning_rate": 5.1985708266210754e-05, "loss": 0.0358, "step": 41040 }, { "grad_norm": 0.2606488764286041, "learning_rate": 5.1965056030091206e-05, "loss": 0.039, "step": 41050 }, { "grad_norm": 0.5829083919525146, "learning_rate": 5.194440345819732e-05, "loss": 0.0298, "step": 41060 }, { "grad_norm": 0.3351556062698364, "learning_rate": 5.1923750554058084e-05, "loss": 0.0401, "step": 41070 }, { "grad_norm": 0.30308717489242554, "learning_rate": 5.1903097321202476e-05, "loss": 0.0261, "step": 41080 }, { "grad_norm": 0.39833560585975647, "learning_rate": 5.18824437631596e-05, "loss": 0.0414, "step": 41090 }, { "grad_norm": 0.46274831891059875, "learning_rate": 5.186178988345856e-05, "loss": 0.0358, "step": 41100 }, { "grad_norm": 0.3554205894470215, "learning_rate": 5.184113568562855e-05, "loss": 0.0333, "step": 41110 }, { "grad_norm": 0.3294537663459778, "learning_rate": 5.18204811731988e-05, "loss": 0.0398, "step": 41120 }, { "grad_norm": 0.4767654538154602, "learning_rate": 5.179982634969862e-05, "loss": 0.0337, "step": 41130 }, { "grad_norm": 0.5526541471481323, "learning_rate": 5.177917121865734e-05, "loss": 0.0396, "step": 41140 }, { "grad_norm": 0.2573021650314331, "learning_rate": 5.1758515783604346e-05, "loss": 0.0329, "step": 41150 }, { "grad_norm": 0.4857744574546814, "learning_rate": 5.173786004806911e-05, "loss": 0.0409, "step": 41160 }, { "grad_norm": 0.5529017448425293, "learning_rate": 5.1717204015581135e-05, "loss": 0.0354, "step": 41170 }, { "grad_norm": 0.3936319947242737, "learning_rate": 5.169654768966994e-05, "loss": 0.0367, "step": 41180 }, { "grad_norm": 0.400039941072464, "learning_rate": 5.1675891073865156e-05, "loss": 0.0315, "step": 41190 }, { "grad_norm": 0.33957305550575256, "learning_rate": 5.1655234171696424e-05, "loss": 0.0313, "step": 41200 }, { "grad_norm": 0.34067827463150024, "learning_rate": 5.163457698669343e-05, "loss": 0.0291, "step": 41210 }, { "grad_norm": 0.3413448929786682, "learning_rate": 5.1613919522385936e-05, "loss": 0.0333, "step": 41220 }, { "grad_norm": 0.39364176988601685, "learning_rate": 5.1593261782303746e-05, "loss": 0.0462, "step": 41230 }, { "grad_norm": 0.30149465799331665, "learning_rate": 5.157260376997669e-05, "loss": 0.0296, "step": 41240 }, { "grad_norm": 0.3095955550670624, "learning_rate": 5.155194548893466e-05, "loss": 0.035, "step": 41250 }, { "grad_norm": 0.29856380820274353, "learning_rate": 5.153128694270759e-05, "loss": 0.0424, "step": 41260 }, { "grad_norm": 0.34721702337265015, "learning_rate": 5.151062813482548e-05, "loss": 0.0371, "step": 41270 }, { "grad_norm": 0.30342164635658264, "learning_rate": 5.148996906881832e-05, "loss": 0.0283, "step": 41280 }, { "grad_norm": 0.4157499372959137, "learning_rate": 5.1469309748216196e-05, "loss": 0.0321, "step": 41290 }, { "grad_norm": 0.30712890625, "learning_rate": 5.144865017654923e-05, "loss": 0.028, "step": 41300 }, { "grad_norm": 0.39397817850112915, "learning_rate": 5.1427990357347556e-05, "loss": 0.0364, "step": 41310 }, { "grad_norm": 0.3472864329814911, "learning_rate": 5.140733029414139e-05, "loss": 0.0303, "step": 41320 }, { "grad_norm": 0.5018752813339233, "learning_rate": 5.138666999046098e-05, "loss": 0.03, "step": 41330 }, { "grad_norm": 0.2818186283111572, "learning_rate": 5.136600944983658e-05, "loss": 0.0281, "step": 41340 }, { "grad_norm": 0.2621479034423828, "learning_rate": 5.134534867579853e-05, "loss": 0.0305, "step": 41350 }, { "grad_norm": 0.3209240138530731, "learning_rate": 5.1324687671877184e-05, "loss": 0.0263, "step": 41360 }, { "grad_norm": 0.4048687219619751, "learning_rate": 5.130402644160296e-05, "loss": 0.0392, "step": 41370 }, { "grad_norm": 0.5465109348297119, "learning_rate": 5.128336498850628e-05, "loss": 0.0346, "step": 41380 }, { "grad_norm": 0.4248104691505432, "learning_rate": 5.126270331611761e-05, "loss": 0.0309, "step": 41390 }, { "grad_norm": 0.43099498748779297, "learning_rate": 5.124204142796748e-05, "loss": 0.029, "step": 41400 }, { "grad_norm": 0.37076878547668457, "learning_rate": 5.122137932758644e-05, "loss": 0.0336, "step": 41410 }, { "grad_norm": 0.21821095049381256, "learning_rate": 5.1200717018505074e-05, "loss": 0.0237, "step": 41420 }, { "grad_norm": 0.4648541510105133, "learning_rate": 5.118005450425403e-05, "loss": 0.0284, "step": 41430 }, { "grad_norm": 0.37507784366607666, "learning_rate": 5.115939178836392e-05, "loss": 0.025, "step": 41440 }, { "grad_norm": 0.37754422426223755, "learning_rate": 5.113872887436547e-05, "loss": 0.0355, "step": 41450 }, { "grad_norm": 0.3954295217990875, "learning_rate": 5.11180657657894e-05, "loss": 0.0358, "step": 41460 }, { "grad_norm": 0.43327900767326355, "learning_rate": 5.109740246616649e-05, "loss": 0.0226, "step": 41470 }, { "grad_norm": 0.5328859090805054, "learning_rate": 5.10767389790275e-05, "loss": 0.0405, "step": 41480 }, { "grad_norm": 0.34289664030075073, "learning_rate": 5.1056075307903265e-05, "loss": 0.0327, "step": 41490 }, { "grad_norm": 0.3258005678653717, "learning_rate": 5.1035411456324634e-05, "loss": 0.0282, "step": 41500 }, { "grad_norm": 0.46244558691978455, "learning_rate": 5.1014747427822504e-05, "loss": 0.0342, "step": 41510 }, { "grad_norm": 0.40351203083992004, "learning_rate": 5.0994083225927804e-05, "loss": 0.0331, "step": 41520 }, { "grad_norm": 0.5674296617507935, "learning_rate": 5.0973418854171475e-05, "loss": 0.0373, "step": 41530 }, { "grad_norm": 0.3989226520061493, "learning_rate": 5.095275431608447e-05, "loss": 0.0247, "step": 41540 }, { "grad_norm": 0.39137911796569824, "learning_rate": 5.093208961519782e-05, "loss": 0.0333, "step": 41550 }, { "grad_norm": 0.31255626678466797, "learning_rate": 5.091142475504255e-05, "loss": 0.0252, "step": 41560 }, { "grad_norm": 0.5126262903213501, "learning_rate": 5.089075973914971e-05, "loss": 0.039, "step": 41570 }, { "grad_norm": 0.5030915141105652, "learning_rate": 5.08700945710504e-05, "loss": 0.0463, "step": 41580 }, { "grad_norm": 0.32791972160339355, "learning_rate": 5.0849429254275714e-05, "loss": 0.0303, "step": 41590 }, { "grad_norm": 0.3259525001049042, "learning_rate": 5.0828763792356804e-05, "loss": 0.0307, "step": 41600 }, { "grad_norm": 0.3968481719493866, "learning_rate": 5.080809818882483e-05, "loss": 0.0368, "step": 41610 }, { "grad_norm": 0.381883442401886, "learning_rate": 5.078743244721097e-05, "loss": 0.0312, "step": 41620 }, { "grad_norm": 0.26773130893707275, "learning_rate": 5.0766766571046455e-05, "loss": 0.0357, "step": 41630 }, { "grad_norm": 0.370211124420166, "learning_rate": 5.07461005638625e-05, "loss": 0.0279, "step": 41640 }, { "grad_norm": 0.2786586582660675, "learning_rate": 5.072543442919037e-05, "loss": 0.0328, "step": 41650 }, { "grad_norm": 0.4656504690647125, "learning_rate": 5.070476817056132e-05, "loss": 0.0453, "step": 41660 }, { "grad_norm": 0.36501526832580566, "learning_rate": 5.068410179150668e-05, "loss": 0.0276, "step": 41670 }, { "grad_norm": 0.402473509311676, "learning_rate": 5.066343529555775e-05, "loss": 0.0284, "step": 41680 }, { "grad_norm": 0.548746645450592, "learning_rate": 5.064276868624587e-05, "loss": 0.0341, "step": 41690 }, { "grad_norm": 0.36138710379600525, "learning_rate": 5.0622101967102396e-05, "loss": 0.0261, "step": 41700 }, { "grad_norm": 0.25848057866096497, "learning_rate": 5.0601435141658714e-05, "loss": 0.0261, "step": 41710 }, { "grad_norm": 0.39386776089668274, "learning_rate": 5.058076821344622e-05, "loss": 0.0348, "step": 41720 }, { "grad_norm": 0.484031081199646, "learning_rate": 5.056010118599632e-05, "loss": 0.0455, "step": 41730 }, { "grad_norm": 0.5002301931381226, "learning_rate": 5.053943406284044e-05, "loss": 0.03, "step": 41740 }, { "grad_norm": 0.32107409834861755, "learning_rate": 5.051876684751006e-05, "loss": 0.0331, "step": 41750 }, { "grad_norm": 0.205850288271904, "learning_rate": 5.0498099543536584e-05, "loss": 0.0332, "step": 41760 }, { "grad_norm": 0.3794197738170624, "learning_rate": 5.047743215445152e-05, "loss": 0.0345, "step": 41770 }, { "grad_norm": 0.4589959681034088, "learning_rate": 5.045676468378637e-05, "loss": 0.0327, "step": 41780 }, { "grad_norm": 0.3471080958843231, "learning_rate": 5.0436097135072626e-05, "loss": 0.0323, "step": 41790 }, { "grad_norm": 0.3467681109905243, "learning_rate": 5.041542951184181e-05, "loss": 0.0298, "step": 41800 }, { "grad_norm": 0.27109476923942566, "learning_rate": 5.039476181762545e-05, "loss": 0.0323, "step": 41810 }, { "grad_norm": 0.20763027667999268, "learning_rate": 5.037409405595508e-05, "loss": 0.0252, "step": 41820 }, { "grad_norm": 0.5280638337135315, "learning_rate": 5.035342623036229e-05, "loss": 0.0319, "step": 41830 }, { "grad_norm": 0.31527766585350037, "learning_rate": 5.033275834437862e-05, "loss": 0.0364, "step": 41840 }, { "grad_norm": 0.5104402899742126, "learning_rate": 5.031209040153564e-05, "loss": 0.03, "step": 41850 }, { "grad_norm": 0.48180466890335083, "learning_rate": 5.0291422405364955e-05, "loss": 0.0387, "step": 41860 }, { "grad_norm": 0.297556072473526, "learning_rate": 5.0270754359398133e-05, "loss": 0.028, "step": 41870 }, { "grad_norm": 0.3269113600254059, "learning_rate": 5.025008626716682e-05, "loss": 0.0241, "step": 41880 }, { "grad_norm": 0.46850860118865967, "learning_rate": 5.0229418132202585e-05, "loss": 0.0239, "step": 41890 }, { "grad_norm": 0.31217166781425476, "learning_rate": 5.020874995803707e-05, "loss": 0.0459, "step": 41900 }, { "grad_norm": 0.25615599751472473, "learning_rate": 5.01880817482019e-05, "loss": 0.0329, "step": 41910 }, { "grad_norm": 0.31960317492485046, "learning_rate": 5.01674135062287e-05, "loss": 0.0269, "step": 41920 }, { "grad_norm": 0.2970135509967804, "learning_rate": 5.0146745235649115e-05, "loss": 0.0334, "step": 41930 }, { "grad_norm": 0.4688030183315277, "learning_rate": 5.0126076939994795e-05, "loss": 0.0264, "step": 41940 }, { "grad_norm": 0.35683882236480713, "learning_rate": 5.010540862279736e-05, "loss": 0.0344, "step": 41950 }, { "grad_norm": 0.651207685470581, "learning_rate": 5.008474028758846e-05, "loss": 0.0336, "step": 41960 }, { "grad_norm": 0.40133291482925415, "learning_rate": 5.0064071937899784e-05, "loss": 0.0344, "step": 41970 }, { "grad_norm": 0.4137856066226959, "learning_rate": 5.004340357726296e-05, "loss": 0.0268, "step": 41980 }, { "grad_norm": 0.4227973520755768, "learning_rate": 5.002273520920965e-05, "loss": 0.0376, "step": 41990 }, { "grad_norm": 0.4176860749721527, "learning_rate": 5.000206683727151e-05, "loss": 0.0318, "step": 42000 }, { "grad_norm": 0.27504992485046387, "learning_rate": 4.998139846498021e-05, "loss": 0.0261, "step": 42010 }, { "grad_norm": 0.4829476475715637, "learning_rate": 4.9960730095867405e-05, "loss": 0.0364, "step": 42020 }, { "grad_norm": 0.4612080156803131, "learning_rate": 4.9940061733464755e-05, "loss": 0.0396, "step": 42030 }, { "grad_norm": 0.25831007957458496, "learning_rate": 4.991939338130392e-05, "loss": 0.0388, "step": 42040 }, { "grad_norm": 0.4830611050128937, "learning_rate": 4.989872504291653e-05, "loss": 0.0485, "step": 42050 }, { "grad_norm": 0.41900011897087097, "learning_rate": 4.9878056721834273e-05, "loss": 0.0329, "step": 42060 }, { "grad_norm": 0.3239488899707794, "learning_rate": 4.98573884215888e-05, "loss": 0.0323, "step": 42070 }, { "grad_norm": 0.43904829025268555, "learning_rate": 4.9836720145711715e-05, "loss": 0.036, "step": 42080 }, { "grad_norm": 0.3384913504123688, "learning_rate": 4.9816051897734725e-05, "loss": 0.0403, "step": 42090 }, { "grad_norm": 0.3274458944797516, "learning_rate": 4.979538368118941e-05, "loss": 0.0375, "step": 42100 }, { "grad_norm": 0.2799476683139801, "learning_rate": 4.9774715499607446e-05, "loss": 0.0314, "step": 42110 }, { "grad_norm": 0.3454388380050659, "learning_rate": 4.9754047356520416e-05, "loss": 0.0363, "step": 42120 }, { "grad_norm": 0.3241339325904846, "learning_rate": 4.973337925545999e-05, "loss": 0.0261, "step": 42130 }, { "grad_norm": 0.2997104227542877, "learning_rate": 4.9712711199957746e-05, "loss": 0.0318, "step": 42140 }, { "grad_norm": 0.46832942962646484, "learning_rate": 4.969204319354531e-05, "loss": 0.0355, "step": 42150 }, { "grad_norm": 0.40216192603111267, "learning_rate": 4.9671375239754267e-05, "loss": 0.0298, "step": 42160 }, { "grad_norm": 0.3147958219051361, "learning_rate": 4.9650707342116195e-05, "loss": 0.0322, "step": 42170 }, { "grad_norm": 0.23310299217700958, "learning_rate": 4.96300395041627e-05, "loss": 0.0312, "step": 42180 }, { "grad_norm": 0.29818516969680786, "learning_rate": 4.960937172942532e-05, "loss": 0.0333, "step": 42190 }, { "grad_norm": 0.3416692912578583, "learning_rate": 4.958870402143565e-05, "loss": 0.0336, "step": 42200 }, { "grad_norm": 0.29268452525138855, "learning_rate": 4.9568036383725186e-05, "loss": 0.0319, "step": 42210 }, { "grad_norm": 0.346894770860672, "learning_rate": 4.9547368819825516e-05, "loss": 0.03, "step": 42220 }, { "grad_norm": 0.38979852199554443, "learning_rate": 4.952670133326812e-05, "loss": 0.0355, "step": 42230 }, { "grad_norm": 0.3516300618648529, "learning_rate": 4.950603392758453e-05, "loss": 0.0309, "step": 42240 }, { "grad_norm": 0.2729610800743103, "learning_rate": 4.948536660630621e-05, "loss": 0.0285, "step": 42250 }, { "grad_norm": 0.40726250410079956, "learning_rate": 4.9464699372964685e-05, "loss": 0.0358, "step": 42260 }, { "grad_norm": 0.4173330068588257, "learning_rate": 4.9444032231091395e-05, "loss": 0.0343, "step": 42270 }, { "grad_norm": 0.3864860534667969, "learning_rate": 4.9423365184217776e-05, "loss": 0.031, "step": 42280 }, { "grad_norm": 0.314602255821228, "learning_rate": 4.940269823587529e-05, "loss": 0.0268, "step": 42290 }, { "grad_norm": 0.60073322057724, "learning_rate": 4.938203138959533e-05, "loss": 0.0244, "step": 42300 }, { "grad_norm": 0.35842472314834595, "learning_rate": 4.9361364648909325e-05, "loss": 0.0318, "step": 42310 }, { "grad_norm": 0.2871144711971283, "learning_rate": 4.9340698017348605e-05, "loss": 0.0319, "step": 42320 }, { "grad_norm": 0.4162743389606476, "learning_rate": 4.932003149844458e-05, "loss": 0.0284, "step": 42330 }, { "grad_norm": 0.4025638699531555, "learning_rate": 4.929936509572857e-05, "loss": 0.0386, "step": 42340 }, { "grad_norm": 0.38585153222084045, "learning_rate": 4.927869881273191e-05, "loss": 0.0339, "step": 42350 }, { "grad_norm": 0.39535582065582275, "learning_rate": 4.9258032652985894e-05, "loss": 0.0339, "step": 42360 }, { "grad_norm": 0.5023564696311951, "learning_rate": 4.9237366620021786e-05, "loss": 0.0372, "step": 42370 }, { "grad_norm": 0.24948765337467194, "learning_rate": 4.921670071737089e-05, "loss": 0.0273, "step": 42380 }, { "grad_norm": 0.3465684652328491, "learning_rate": 4.91960349485644e-05, "loss": 0.0332, "step": 42390 }, { "grad_norm": 0.3356480300426483, "learning_rate": 4.9175369317133566e-05, "loss": 0.0313, "step": 42400 }, { "grad_norm": 0.3142395615577698, "learning_rate": 4.915470382660953e-05, "loss": 0.0281, "step": 42410 }, { "grad_norm": 0.47453635931015015, "learning_rate": 4.9134038480523524e-05, "loss": 0.03, "step": 42420 }, { "grad_norm": 0.4620782136917114, "learning_rate": 4.911337328240664e-05, "loss": 0.0304, "step": 42430 }, { "grad_norm": 0.22747714817523956, "learning_rate": 4.909270823579003e-05, "loss": 0.0285, "step": 42440 }, { "grad_norm": 0.4283902049064636, "learning_rate": 4.907204334420476e-05, "loss": 0.0263, "step": 42450 }, { "grad_norm": 0.32358747720718384, "learning_rate": 4.9051378611181884e-05, "loss": 0.0299, "step": 42460 }, { "grad_norm": 0.2474699765443802, "learning_rate": 4.9030714040252486e-05, "loss": 0.027, "step": 42470 }, { "grad_norm": 0.33574190735816956, "learning_rate": 4.901004963494752e-05, "loss": 0.0211, "step": 42480 }, { "grad_norm": 0.3689906895160675, "learning_rate": 4.898938539879802e-05, "loss": 0.0379, "step": 42490 }, { "grad_norm": 0.31631484627723694, "learning_rate": 4.8968721335334904e-05, "loss": 0.0293, "step": 42500 }, { "grad_norm": 0.2645129859447479, "learning_rate": 4.894805744808912e-05, "loss": 0.032, "step": 42510 }, { "grad_norm": 0.3321966528892517, "learning_rate": 4.8927393740591524e-05, "loss": 0.0323, "step": 42520 }, { "grad_norm": 0.3675123453140259, "learning_rate": 4.890673021637302e-05, "loss": 0.0273, "step": 42530 }, { "grad_norm": 0.34183269739151, "learning_rate": 4.888606687896442e-05, "loss": 0.028, "step": 42540 }, { "grad_norm": 0.3554548919200897, "learning_rate": 4.886540373189652e-05, "loss": 0.0283, "step": 42550 }, { "grad_norm": 0.36948662996292114, "learning_rate": 4.88447407787001e-05, "loss": 0.0313, "step": 42560 }, { "grad_norm": 0.34462225437164307, "learning_rate": 4.8824078022905864e-05, "loss": 0.0261, "step": 42570 }, { "grad_norm": 0.31316235661506653, "learning_rate": 4.880341546804455e-05, "loss": 0.0243, "step": 42580 }, { "grad_norm": 0.28070446848869324, "learning_rate": 4.878275311764679e-05, "loss": 0.0339, "step": 42590 }, { "grad_norm": 0.31092119216918945, "learning_rate": 4.8762090975243254e-05, "loss": 0.0305, "step": 42600 }, { "grad_norm": 0.304592102766037, "learning_rate": 4.874142904436448e-05, "loss": 0.0245, "step": 42610 }, { "grad_norm": 0.35226333141326904, "learning_rate": 4.8720767328541094e-05, "loss": 0.0396, "step": 42620 }, { "grad_norm": 0.367553174495697, "learning_rate": 4.870010583130358e-05, "loss": 0.0281, "step": 42630 }, { "grad_norm": 0.32996004819869995, "learning_rate": 4.867944455618241e-05, "loss": 0.023, "step": 42640 }, { "grad_norm": 0.30065515637397766, "learning_rate": 4.865878350670807e-05, "loss": 0.0315, "step": 42650 }, { "grad_norm": 0.4498574137687683, "learning_rate": 4.8638122686410914e-05, "loss": 0.035, "step": 42660 }, { "grad_norm": 0.29181671142578125, "learning_rate": 4.861746209882137e-05, "loss": 0.0282, "step": 42670 }, { "grad_norm": 0.2711544930934906, "learning_rate": 4.859680174746972e-05, "loss": 0.0305, "step": 42680 }, { "grad_norm": 0.49228760600090027, "learning_rate": 4.857614163588629e-05, "loss": 0.0449, "step": 42690 }, { "grad_norm": 0.27456575632095337, "learning_rate": 4.855548176760131e-05, "loss": 0.0291, "step": 42700 }, { "grad_norm": 0.35649439692497253, "learning_rate": 4.8534822146144986e-05, "loss": 0.0347, "step": 42710 }, { "grad_norm": 0.3977842628955841, "learning_rate": 4.851416277504749e-05, "loss": 0.0406, "step": 42720 }, { "grad_norm": 0.467220664024353, "learning_rate": 4.8493503657838923e-05, "loss": 0.0261, "step": 42730 }, { "grad_norm": 0.529390811920166, "learning_rate": 4.84728447980494e-05, "loss": 0.0286, "step": 42740 }, { "grad_norm": 0.30157357454299927, "learning_rate": 4.8452186199208914e-05, "loss": 0.0301, "step": 42750 }, { "grad_norm": 0.34539321064949036, "learning_rate": 4.843152786484749e-05, "loss": 0.0311, "step": 42760 }, { "grad_norm": 0.312845379114151, "learning_rate": 4.8410869798495035e-05, "loss": 0.0275, "step": 42770 }, { "grad_norm": 0.22783046960830688, "learning_rate": 4.8390212003681486e-05, "loss": 0.0305, "step": 42780 }, { "grad_norm": 0.3995252847671509, "learning_rate": 4.836955448393667e-05, "loss": 0.032, "step": 42790 }, { "grad_norm": 0.24257080256938934, "learning_rate": 4.8348897242790406e-05, "loss": 0.029, "step": 42800 }, { "grad_norm": 0.4086291491985321, "learning_rate": 4.832824028377243e-05, "loss": 0.0262, "step": 42810 }, { "grad_norm": 0.49687305092811584, "learning_rate": 4.830758361041249e-05, "loss": 0.0351, "step": 42820 }, { "grad_norm": 0.3811231255531311, "learning_rate": 4.828692722624022e-05, "loss": 0.0322, "step": 42830 }, { "grad_norm": 0.26573121547698975, "learning_rate": 4.826627113478522e-05, "loss": 0.0472, "step": 42840 }, { "grad_norm": 0.2095414102077484, "learning_rate": 4.824561533957708e-05, "loss": 0.0284, "step": 42850 }, { "grad_norm": 0.28287604451179504, "learning_rate": 4.822495984414527e-05, "loss": 0.0268, "step": 42860 }, { "grad_norm": 0.5098594427108765, "learning_rate": 4.8204304652019304e-05, "loss": 0.0284, "step": 42870 }, { "grad_norm": 0.32250744104385376, "learning_rate": 4.8183649766728537e-05, "loss": 0.0314, "step": 42880 }, { "grad_norm": 0.45759713649749756, "learning_rate": 4.816299519180237e-05, "loss": 0.0312, "step": 42890 }, { "grad_norm": 0.2994266152381897, "learning_rate": 4.8142340930770066e-05, "loss": 0.0301, "step": 42900 }, { "grad_norm": 0.22737164795398712, "learning_rate": 4.81216869871609e-05, "loss": 0.0265, "step": 42910 }, { "grad_norm": 0.3273414075374603, "learning_rate": 4.8101033364504064e-05, "loss": 0.0299, "step": 42920 }, { "grad_norm": 0.26257601380348206, "learning_rate": 4.808038006632866e-05, "loss": 0.0403, "step": 42930 }, { "grad_norm": 0.31164345145225525, "learning_rate": 4.805972709616383e-05, "loss": 0.0311, "step": 42940 }, { "grad_norm": 0.2804299294948578, "learning_rate": 4.8039074457538556e-05, "loss": 0.024, "step": 42950 }, { "grad_norm": 0.3141792416572571, "learning_rate": 4.801842215398184e-05, "loss": 0.0356, "step": 42960 }, { "grad_norm": 0.32419243454933167, "learning_rate": 4.799777018902256e-05, "loss": 0.033, "step": 42970 }, { "grad_norm": 0.5266632437705994, "learning_rate": 4.797711856618961e-05, "loss": 0.0314, "step": 42980 }, { "grad_norm": 0.42790621519088745, "learning_rate": 4.7956467289011765e-05, "loss": 0.036, "step": 42990 }, { "grad_norm": 0.359892338514328, "learning_rate": 4.793581636101778e-05, "loss": 0.0319, "step": 43000 }, { "grad_norm": 0.36867064237594604, "learning_rate": 4.7915165785736326e-05, "loss": 0.0326, "step": 43010 }, { "grad_norm": 0.3469797372817993, "learning_rate": 4.789451556669599e-05, "loss": 0.0347, "step": 43020 }, { "grad_norm": 0.19132274389266968, "learning_rate": 4.7873865707425384e-05, "loss": 0.0344, "step": 43030 }, { "grad_norm": 0.3031531572341919, "learning_rate": 4.7853216211452975e-05, "loss": 0.0264, "step": 43040 }, { "grad_norm": 0.4323164224624634, "learning_rate": 4.78325670823072e-05, "loss": 0.0319, "step": 43050 }, { "grad_norm": 0.31795331835746765, "learning_rate": 4.781191832351641e-05, "loss": 0.0348, "step": 43060 }, { "grad_norm": 0.3684547245502472, "learning_rate": 4.7791269938608965e-05, "loss": 0.0344, "step": 43070 }, { "grad_norm": 0.21201521158218384, "learning_rate": 4.777062193111305e-05, "loss": 0.0294, "step": 43080 }, { "grad_norm": 0.3199866712093353, "learning_rate": 4.77499743045569e-05, "loss": 0.0338, "step": 43090 }, { "grad_norm": 0.42544668912887573, "learning_rate": 4.7729327062468596e-05, "loss": 0.0328, "step": 43100 }, { "grad_norm": 0.3143492341041565, "learning_rate": 4.770868020837619e-05, "loss": 0.0301, "step": 43110 }, { "grad_norm": 0.3532503843307495, "learning_rate": 4.768803374580768e-05, "loss": 0.0286, "step": 43120 }, { "grad_norm": 0.32179155945777893, "learning_rate": 4.766738767829094e-05, "loss": 0.0312, "step": 43130 }, { "grad_norm": 0.3529799282550812, "learning_rate": 4.764674200935388e-05, "loss": 0.0327, "step": 43140 }, { "grad_norm": 0.3089449107646942, "learning_rate": 4.762609674252424e-05, "loss": 0.0296, "step": 43150 }, { "grad_norm": 0.2607254683971405, "learning_rate": 4.760545188132974e-05, "loss": 0.0415, "step": 43160 }, { "grad_norm": 0.2579675614833832, "learning_rate": 4.758480742929801e-05, "loss": 0.0315, "step": 43170 }, { "grad_norm": 0.27393436431884766, "learning_rate": 4.756416338995664e-05, "loss": 0.0337, "step": 43180 }, { "grad_norm": 0.41071248054504395, "learning_rate": 4.7543519766833145e-05, "loss": 0.0246, "step": 43190 }, { "grad_norm": 0.24924494326114655, "learning_rate": 4.752287656345492e-05, "loss": 0.0285, "step": 43200 }, { "grad_norm": 0.30867546796798706, "learning_rate": 4.7502233783349355e-05, "loss": 0.0283, "step": 43210 }, { "grad_norm": 0.2834223806858063, "learning_rate": 4.7481591430043694e-05, "loss": 0.0217, "step": 43220 }, { "grad_norm": 0.3317147195339203, "learning_rate": 4.746094950706521e-05, "loss": 0.0333, "step": 43230 }, { "grad_norm": 0.34103620052337646, "learning_rate": 4.744030801794099e-05, "loss": 0.0308, "step": 43240 }, { "grad_norm": 0.5284645557403564, "learning_rate": 4.741966696619813e-05, "loss": 0.04, "step": 43250 }, { "grad_norm": 0.23997469246387482, "learning_rate": 4.739902635536361e-05, "loss": 0.0381, "step": 43260 }, { "grad_norm": 0.44415283203125, "learning_rate": 4.737838618896436e-05, "loss": 0.0321, "step": 43270 }, { "grad_norm": 0.473682165145874, "learning_rate": 4.7357746470527203e-05, "loss": 0.0278, "step": 43280 }, { "grad_norm": 0.402076780796051, "learning_rate": 4.73371072035789e-05, "loss": 0.0345, "step": 43290 }, { "grad_norm": 0.41719886660575867, "learning_rate": 4.731646839164616e-05, "loss": 0.0336, "step": 43300 }, { "grad_norm": 0.3728691041469574, "learning_rate": 4.729583003825556e-05, "loss": 0.0311, "step": 43310 }, { "grad_norm": 0.2655532658100128, "learning_rate": 4.727519214693367e-05, "loss": 0.0314, "step": 43320 }, { "grad_norm": 0.2840285003185272, "learning_rate": 4.725455472120689e-05, "loss": 0.0283, "step": 43330 }, { "grad_norm": 0.3321065306663513, "learning_rate": 4.723391776460164e-05, "loss": 0.0344, "step": 43340 }, { "grad_norm": 0.2986536920070648, "learning_rate": 4.7213281280644186e-05, "loss": 0.0333, "step": 43350 }, { "grad_norm": 0.3998212218284607, "learning_rate": 4.719264527286075e-05, "loss": 0.0293, "step": 43360 }, { "grad_norm": 0.29899802803993225, "learning_rate": 4.717200974477744e-05, "loss": 0.0317, "step": 43370 }, { "grad_norm": 0.3086966276168823, "learning_rate": 4.715137469992034e-05, "loss": 0.0311, "step": 43380 }, { "grad_norm": 0.3094511926174164, "learning_rate": 4.713074014181539e-05, "loss": 0.0268, "step": 43390 }, { "grad_norm": 0.26301026344299316, "learning_rate": 4.711010607398846e-05, "loss": 0.0289, "step": 43400 }, { "grad_norm": 0.29617050290107727, "learning_rate": 4.708947249996538e-05, "loss": 0.0244, "step": 43410 }, { "grad_norm": 0.2533190846443176, "learning_rate": 4.706883942327183e-05, "loss": 0.0278, "step": 43420 }, { "grad_norm": 0.33852341771125793, "learning_rate": 4.704820684743347e-05, "loss": 0.0308, "step": 43430 }, { "grad_norm": 0.38952985405921936, "learning_rate": 4.702757477597581e-05, "loss": 0.0328, "step": 43440 }, { "grad_norm": 0.36711540818214417, "learning_rate": 4.700694321242435e-05, "loss": 0.0301, "step": 43450 }, { "grad_norm": 0.4794553816318512, "learning_rate": 4.698631216030441e-05, "loss": 0.0316, "step": 43460 }, { "grad_norm": 0.246877983212471, "learning_rate": 4.6965681623141314e-05, "loss": 0.0277, "step": 43470 }, { "grad_norm": 0.26811039447784424, "learning_rate": 4.694505160446024e-05, "loss": 0.0273, "step": 43480 }, { "grad_norm": 0.39690732955932617, "learning_rate": 4.692442210778628e-05, "loss": 0.0312, "step": 43490 }, { "grad_norm": 0.4066978693008423, "learning_rate": 4.690379313664448e-05, "loss": 0.0313, "step": 43500 }, { "grad_norm": 0.3977679908275604, "learning_rate": 4.688316469455973e-05, "loss": 0.03, "step": 43510 }, { "grad_norm": 0.429800808429718, "learning_rate": 4.6862536785056914e-05, "loss": 0.028, "step": 43520 }, { "grad_norm": 0.325452983379364, "learning_rate": 4.684190941166073e-05, "loss": 0.0298, "step": 43530 }, { "grad_norm": 0.3146474063396454, "learning_rate": 4.682128257789587e-05, "loss": 0.0339, "step": 43540 }, { "grad_norm": 0.4067109227180481, "learning_rate": 4.680065628728687e-05, "loss": 0.0244, "step": 43550 }, { "grad_norm": 0.44523561000823975, "learning_rate": 4.678003054335822e-05, "loss": 0.0337, "step": 43560 }, { "grad_norm": 0.38259828090667725, "learning_rate": 4.675940534963428e-05, "loss": 0.0328, "step": 43570 }, { "grad_norm": 0.3576885461807251, "learning_rate": 4.673878070963931e-05, "loss": 0.0353, "step": 43580 }, { "grad_norm": 0.25862160325050354, "learning_rate": 4.671815662689756e-05, "loss": 0.0242, "step": 43590 }, { "grad_norm": 0.28888532519340515, "learning_rate": 4.669753310493306e-05, "loss": 0.0262, "step": 43600 }, { "grad_norm": 0.29775071144104004, "learning_rate": 4.6676910147269845e-05, "loss": 0.0336, "step": 43610 }, { "grad_norm": 0.34701117873191833, "learning_rate": 4.665628775743177e-05, "loss": 0.0298, "step": 43620 }, { "grad_norm": 0.45836082100868225, "learning_rate": 4.66356659389427e-05, "loss": 0.0323, "step": 43630 }, { "grad_norm": 0.39052045345306396, "learning_rate": 4.661504469532629e-05, "loss": 0.0307, "step": 43640 }, { "grad_norm": 0.3413746654987335, "learning_rate": 4.659442403010617e-05, "loss": 0.0239, "step": 43650 }, { "grad_norm": 0.36120572686195374, "learning_rate": 4.6573803946805845e-05, "loss": 0.0241, "step": 43660 }, { "grad_norm": 0.28421247005462646, "learning_rate": 4.6553184448948695e-05, "loss": 0.0272, "step": 43670 }, { "grad_norm": 0.31744396686553955, "learning_rate": 4.653256554005807e-05, "loss": 0.0301, "step": 43680 }, { "grad_norm": 0.23078548908233643, "learning_rate": 4.6511947223657145e-05, "loss": 0.0376, "step": 43690 }, { "grad_norm": 0.41549772024154663, "learning_rate": 4.649132950326906e-05, "loss": 0.032, "step": 43700 }, { "grad_norm": 0.3592272400856018, "learning_rate": 4.647071238241679e-05, "loss": 0.0316, "step": 43710 }, { "grad_norm": 0.4071977734565735, "learning_rate": 4.6450095864623264e-05, "loss": 0.0303, "step": 43720 }, { "grad_norm": 0.3066325783729553, "learning_rate": 4.642947995341123e-05, "loss": 0.027, "step": 43730 }, { "grad_norm": 0.3934672474861145, "learning_rate": 4.640886465230345e-05, "loss": 0.0307, "step": 43740 }, { "grad_norm": 0.35294386744499207, "learning_rate": 4.6388249964822485e-05, "loss": 0.0298, "step": 43750 }, { "grad_norm": 0.40634578466415405, "learning_rate": 4.63676358944908e-05, "loss": 0.029, "step": 43760 }, { "grad_norm": 0.46155089139938354, "learning_rate": 4.634702244483081e-05, "loss": 0.0344, "step": 43770 }, { "grad_norm": 0.3351749777793884, "learning_rate": 4.632640961936475e-05, "loss": 0.0328, "step": 43780 }, { "grad_norm": 0.23412510752677917, "learning_rate": 4.6305797421614835e-05, "loss": 0.027, "step": 43790 }, { "grad_norm": 0.3169061839580536, "learning_rate": 4.6285185855103105e-05, "loss": 0.0291, "step": 43800 }, { "grad_norm": 0.4708224833011627, "learning_rate": 4.626457492335151e-05, "loss": 0.0325, "step": 43810 }, { "grad_norm": 0.2562912106513977, "learning_rate": 4.624396462988188e-05, "loss": 0.0294, "step": 43820 }, { "grad_norm": 0.3107246160507202, "learning_rate": 4.6223354978215985e-05, "loss": 0.0249, "step": 43830 }, { "grad_norm": 0.38064417243003845, "learning_rate": 4.620274597187544e-05, "loss": 0.0291, "step": 43840 }, { "grad_norm": 0.4322049617767334, "learning_rate": 4.6182137614381736e-05, "loss": 0.0345, "step": 43850 }, { "grad_norm": 0.3832939863204956, "learning_rate": 4.616152990925631e-05, "loss": 0.027, "step": 43860 }, { "grad_norm": 0.3607197403907776, "learning_rate": 4.614092286002041e-05, "loss": 0.0301, "step": 43870 }, { "grad_norm": 0.2992905378341675, "learning_rate": 4.612031647019528e-05, "loss": 0.0345, "step": 43880 }, { "grad_norm": 0.3189133107662201, "learning_rate": 4.609971074330193e-05, "loss": 0.0298, "step": 43890 }, { "grad_norm": 0.4591258764266968, "learning_rate": 4.607910568286136e-05, "loss": 0.0341, "step": 43900 }, { "grad_norm": 0.5630350112915039, "learning_rate": 4.605850129239438e-05, "loss": 0.0334, "step": 43910 }, { "grad_norm": 0.46470361948013306, "learning_rate": 4.603789757542174e-05, "loss": 0.0317, "step": 43920 }, { "grad_norm": 0.6303235292434692, "learning_rate": 4.601729453546404e-05, "loss": 0.0329, "step": 43930 }, { "grad_norm": 0.4693615436553955, "learning_rate": 4.599669217604177e-05, "loss": 0.0358, "step": 43940 }, { "grad_norm": 0.34029698371887207, "learning_rate": 4.597609050067532e-05, "loss": 0.0322, "step": 43950 }, { "grad_norm": 0.3746715486049652, "learning_rate": 4.595548951288495e-05, "loss": 0.0398, "step": 43960 }, { "grad_norm": 0.2783588469028473, "learning_rate": 4.593488921619081e-05, "loss": 0.0345, "step": 43970 }, { "grad_norm": 0.33709338307380676, "learning_rate": 4.591428961411289e-05, "loss": 0.037, "step": 43980 }, { "grad_norm": 0.39509332180023193, "learning_rate": 4.589369071017117e-05, "loss": 0.0437, "step": 43990 }, { "grad_norm": 0.29201117157936096, "learning_rate": 4.587309250788538e-05, "loss": 0.033, "step": 44000 }, { "grad_norm": 0.339519739151001, "learning_rate": 4.585249501077522e-05, "loss": 0.0264, "step": 44010 }, { "grad_norm": 0.35114914178848267, "learning_rate": 4.583189822236021e-05, "loss": 0.0392, "step": 44020 }, { "grad_norm": 0.3247678875923157, "learning_rate": 4.5811302146159816e-05, "loss": 0.0255, "step": 44030 }, { "grad_norm": 0.5587801337242126, "learning_rate": 4.579070678569332e-05, "loss": 0.0363, "step": 44040 }, { "grad_norm": 0.28977951407432556, "learning_rate": 4.5770112144479904e-05, "loss": 0.0285, "step": 44050 }, { "grad_norm": 0.3360966145992279, "learning_rate": 4.5749518226038645e-05, "loss": 0.0271, "step": 44060 }, { "grad_norm": 0.3738487958908081, "learning_rate": 4.572892503388845e-05, "loss": 0.0291, "step": 44070 }, { "grad_norm": 0.27127528190612793, "learning_rate": 4.570833257154817e-05, "loss": 0.0306, "step": 44080 }, { "grad_norm": 0.3101758360862732, "learning_rate": 4.568774084253646e-05, "loss": 0.0371, "step": 44090 }, { "grad_norm": 0.26998063921928406, "learning_rate": 4.566714985037191e-05, "loss": 0.0376, "step": 44100 }, { "grad_norm": 0.41428491473197937, "learning_rate": 4.564655959857295e-05, "loss": 0.0295, "step": 44110 }, { "grad_norm": 0.4177580773830414, "learning_rate": 4.5625970090657893e-05, "loss": 0.0251, "step": 44120 }, { "grad_norm": 0.30799052119255066, "learning_rate": 4.560538133014491e-05, "loss": 0.0303, "step": 44130 }, { "grad_norm": 0.3159548342227936, "learning_rate": 4.5584793320552055e-05, "loss": 0.0266, "step": 44140 }, { "grad_norm": 0.25899359583854675, "learning_rate": 4.556420606539728e-05, "loss": 0.0244, "step": 44150 }, { "grad_norm": 0.2900778353214264, "learning_rate": 4.554361956819836e-05, "loss": 0.0282, "step": 44160 }, { "grad_norm": 0.316010057926178, "learning_rate": 4.552303383247299e-05, "loss": 0.0255, "step": 44170 }, { "grad_norm": 0.45146888494491577, "learning_rate": 4.5502448861738664e-05, "loss": 0.031, "step": 44180 }, { "grad_norm": 0.37937578558921814, "learning_rate": 4.548186465951284e-05, "loss": 0.0359, "step": 44190 }, { "grad_norm": 0.32697227597236633, "learning_rate": 4.546128122931277e-05, "loss": 0.0269, "step": 44200 }, { "grad_norm": 0.26032644510269165, "learning_rate": 4.544069857465561e-05, "loss": 0.0281, "step": 44210 }, { "grad_norm": 0.3088161051273346, "learning_rate": 4.542011669905837e-05, "loss": 0.0265, "step": 44220 }, { "grad_norm": 0.3462545573711395, "learning_rate": 4.53995356060379e-05, "loss": 0.0281, "step": 44230 }, { "grad_norm": 0.23767723143100739, "learning_rate": 4.5378955299110994e-05, "loss": 0.0244, "step": 44240 }, { "grad_norm": 0.40068626403808594, "learning_rate": 4.5358375781794233e-05, "loss": 0.0318, "step": 44250 }, { "grad_norm": 0.4110613763332367, "learning_rate": 4.533779705760411e-05, "loss": 0.0346, "step": 44260 }, { "grad_norm": 0.4433673322200775, "learning_rate": 4.5317219130056934e-05, "loss": 0.0302, "step": 44270 }, { "grad_norm": 0.3999027609825134, "learning_rate": 4.5296642002668964e-05, "loss": 0.0365, "step": 44280 }, { "grad_norm": 0.409607470035553, "learning_rate": 4.5276065678956216e-05, "loss": 0.0409, "step": 44290 }, { "grad_norm": 0.45006421208381653, "learning_rate": 4.525549016243466e-05, "loss": 0.0328, "step": 44300 }, { "grad_norm": 0.29215723276138306, "learning_rate": 4.523491545662008e-05, "loss": 0.0286, "step": 44310 }, { "grad_norm": 0.3758581876754761, "learning_rate": 4.5214341565028104e-05, "loss": 0.0346, "step": 44320 }, { "grad_norm": 0.3764615058898926, "learning_rate": 4.519376849117428e-05, "loss": 0.0289, "step": 44330 }, { "grad_norm": 0.24630562961101532, "learning_rate": 4.517319623857395e-05, "loss": 0.0308, "step": 44340 }, { "grad_norm": 0.3227881193161011, "learning_rate": 4.5152624810742394e-05, "loss": 0.0272, "step": 44350 }, { "grad_norm": 0.2236466407775879, "learning_rate": 4.5132054211194664e-05, "loss": 0.0265, "step": 44360 }, { "grad_norm": 0.7889732122421265, "learning_rate": 4.511148444344574e-05, "loss": 0.029, "step": 44370 }, { "grad_norm": 0.6693759560585022, "learning_rate": 4.509091551101041e-05, "loss": 0.0361, "step": 44380 }, { "grad_norm": 0.37376877665519714, "learning_rate": 4.507034741740338e-05, "loss": 0.0356, "step": 44390 }, { "grad_norm": 0.3151438236236572, "learning_rate": 4.5049780166139145e-05, "loss": 0.0316, "step": 44400 }, { "grad_norm": 0.3571125864982605, "learning_rate": 4.5029213760732075e-05, "loss": 0.0406, "step": 44410 }, { "grad_norm": 0.3136291801929474, "learning_rate": 4.5008648204696434e-05, "loss": 0.0306, "step": 44420 }, { "grad_norm": 0.184987410902977, "learning_rate": 4.4988083501546284e-05, "loss": 0.029, "step": 44430 }, { "grad_norm": 0.33879607915878296, "learning_rate": 4.4967519654795606e-05, "loss": 0.0284, "step": 44440 }, { "grad_norm": 0.3321819305419922, "learning_rate": 4.494695666795816e-05, "loss": 0.029, "step": 44450 }, { "grad_norm": 0.40852198004722595, "learning_rate": 4.4926394544547644e-05, "loss": 0.0312, "step": 44460 }, { "grad_norm": 0.3939702808856964, "learning_rate": 4.490583328807752e-05, "loss": 0.0362, "step": 44470 }, { "grad_norm": 0.3631855845451355, "learning_rate": 4.488527290206117e-05, "loss": 0.0323, "step": 44480 }, { "grad_norm": 0.39643850922584534, "learning_rate": 4.48647133900118e-05, "loss": 0.027, "step": 44490 }, { "grad_norm": 0.3378858268260956, "learning_rate": 4.484415475544244e-05, "loss": 0.0329, "step": 44500 }, { "grad_norm": 0.24001479148864746, "learning_rate": 4.4823597001866035e-05, "loss": 0.0277, "step": 44510 }, { "grad_norm": 0.2695952355861664, "learning_rate": 4.480304013279532e-05, "loss": 0.0259, "step": 44520 }, { "grad_norm": 0.3520267605781555, "learning_rate": 4.478248415174292e-05, "loss": 0.0255, "step": 44530 }, { "grad_norm": 0.41005662083625793, "learning_rate": 4.476192906222126e-05, "loss": 0.0347, "step": 44540 }, { "grad_norm": 0.36574727296829224, "learning_rate": 4.474137486774268e-05, "loss": 0.0391, "step": 44550 }, { "grad_norm": 0.35628485679626465, "learning_rate": 4.4720821571819296e-05, "loss": 0.0301, "step": 44560 }, { "grad_norm": 0.2863302230834961, "learning_rate": 4.470026917796314e-05, "loss": 0.0372, "step": 44570 }, { "grad_norm": 0.4162857234477997, "learning_rate": 4.4679717689686005e-05, "loss": 0.0447, "step": 44580 }, { "grad_norm": 0.2699683606624603, "learning_rate": 4.465916711049963e-05, "loss": 0.0239, "step": 44590 }, { "grad_norm": 0.2631834149360657, "learning_rate": 4.4638617443915524e-05, "loss": 0.0298, "step": 44600 }, { "grad_norm": 0.4112457036972046, "learning_rate": 4.4618068693445055e-05, "loss": 0.0298, "step": 44610 }, { "grad_norm": 0.6126990914344788, "learning_rate": 4.459752086259946e-05, "loss": 0.0291, "step": 44620 }, { "grad_norm": 0.46070268750190735, "learning_rate": 4.457697395488977e-05, "loss": 0.0364, "step": 44630 }, { "grad_norm": 0.2998032569885254, "learning_rate": 4.455642797382693e-05, "loss": 0.037, "step": 44640 }, { "grad_norm": 0.29408764839172363, "learning_rate": 4.4535882922921643e-05, "loss": 0.0303, "step": 44650 }, { "grad_norm": 0.3873385190963745, "learning_rate": 4.451533880568455e-05, "loss": 0.0274, "step": 44660 }, { "grad_norm": 0.3315960168838501, "learning_rate": 4.449479562562603e-05, "loss": 0.0294, "step": 44670 }, { "grad_norm": 0.46396955847740173, "learning_rate": 4.447425338625639e-05, "loss": 0.0308, "step": 44680 }, { "grad_norm": 0.31022965908050537, "learning_rate": 4.4453712091085705e-05, "loss": 0.029, "step": 44690 }, { "grad_norm": 0.3764539062976837, "learning_rate": 4.443317174362392e-05, "loss": 0.0407, "step": 44700 }, { "grad_norm": 0.3823290467262268, "learning_rate": 4.441263234738085e-05, "loss": 0.0239, "step": 44710 }, { "grad_norm": 0.35292109847068787, "learning_rate": 4.4392093905866086e-05, "loss": 0.0214, "step": 44720 }, { "grad_norm": 0.4144594669342041, "learning_rate": 4.437155642258911e-05, "loss": 0.0263, "step": 44730 }, { "grad_norm": 0.3738105595111847, "learning_rate": 4.4351019901059177e-05, "loss": 0.0316, "step": 44740 }, { "grad_norm": 0.47486019134521484, "learning_rate": 4.433048434478547e-05, "loss": 0.0476, "step": 44750 }, { "grad_norm": 0.30835792422294617, "learning_rate": 4.4309949757276906e-05, "loss": 0.028, "step": 44760 }, { "grad_norm": 0.30882027745246887, "learning_rate": 4.428941614204233e-05, "loss": 0.039, "step": 44770 }, { "grad_norm": 0.3194412887096405, "learning_rate": 4.426888350259034e-05, "loss": 0.0332, "step": 44780 }, { "grad_norm": 0.42464759945869446, "learning_rate": 4.424835184242939e-05, "loss": 0.0385, "step": 44790 }, { "grad_norm": 0.28651073575019836, "learning_rate": 4.422782116506784e-05, "loss": 0.0266, "step": 44800 }, { "grad_norm": 0.6686350107192993, "learning_rate": 4.4207291474013754e-05, "loss": 0.0353, "step": 44810 }, { "grad_norm": 0.320340096950531, "learning_rate": 4.418676277277514e-05, "loss": 0.0414, "step": 44820 }, { "grad_norm": 0.3566051721572876, "learning_rate": 4.416623506485976e-05, "loss": 0.0373, "step": 44830 }, { "grad_norm": 0.49599340558052063, "learning_rate": 4.414570835377527e-05, "loss": 0.0301, "step": 44840 }, { "grad_norm": 0.27649328112602234, "learning_rate": 4.412518264302909e-05, "loss": 0.0285, "step": 44850 }, { "grad_norm": 0.3683408498764038, "learning_rate": 4.4104657936128535e-05, "loss": 0.0294, "step": 44860 }, { "grad_norm": 0.3675743639469147, "learning_rate": 4.408413423658071e-05, "loss": 0.0328, "step": 44870 }, { "grad_norm": 0.6120147109031677, "learning_rate": 4.406361154789253e-05, "loss": 0.0337, "step": 44880 }, { "grad_norm": 0.4042995274066925, "learning_rate": 4.4043089873570776e-05, "loss": 0.031, "step": 44890 }, { "grad_norm": 0.4094383418560028, "learning_rate": 4.4022569217122025e-05, "loss": 0.029, "step": 44900 }, { "grad_norm": 0.3323797583580017, "learning_rate": 4.400204958205274e-05, "loss": 0.0303, "step": 44910 }, { "grad_norm": 0.406749427318573, "learning_rate": 4.3981530971869125e-05, "loss": 0.0451, "step": 44920 }, { "grad_norm": 0.3369084596633911, "learning_rate": 4.396101339007727e-05, "loss": 0.028, "step": 44930 }, { "grad_norm": 0.2718517780303955, "learning_rate": 4.394049684018304e-05, "loss": 0.0251, "step": 44940 }, { "grad_norm": 0.35381415486335754, "learning_rate": 4.3919981325692186e-05, "loss": 0.035, "step": 44950 }, { "grad_norm": 0.22077645361423492, "learning_rate": 4.389946685011024e-05, "loss": 0.0243, "step": 44960 }, { "grad_norm": 0.2879001498222351, "learning_rate": 4.387895341694255e-05, "loss": 0.0341, "step": 44970 }, { "grad_norm": 0.302605003118515, "learning_rate": 4.3858441029694324e-05, "loss": 0.0277, "step": 44980 }, { "grad_norm": 0.24279852211475372, "learning_rate": 4.3837929691870527e-05, "loss": 0.0253, "step": 44990 }, { "grad_norm": 0.327980101108551, "learning_rate": 4.381741940697604e-05, "loss": 0.0302, "step": 45000 }, { "grad_norm": 0.3610619604587555, "learning_rate": 4.379691017851547e-05, "loss": 0.0197, "step": 45010 }, { "grad_norm": 0.3314509987831116, "learning_rate": 4.3776402009993304e-05, "loss": 0.0307, "step": 45020 }, { "grad_norm": 0.5035527944564819, "learning_rate": 4.3755894904913794e-05, "loss": 0.0284, "step": 45030 }, { "grad_norm": 0.2706180810928345, "learning_rate": 4.373538886678109e-05, "loss": 0.0257, "step": 45040 }, { "grad_norm": 0.32730183005332947, "learning_rate": 4.371488389909909e-05, "loss": 0.0283, "step": 45050 }, { "grad_norm": 0.34904539585113525, "learning_rate": 4.3694380005371515e-05, "loss": 0.0319, "step": 45060 }, { "grad_norm": 0.2653549015522003, "learning_rate": 4.367387718910196e-05, "loss": 0.0253, "step": 45070 }, { "grad_norm": 0.364631325006485, "learning_rate": 4.3653375453793764e-05, "loss": 0.0281, "step": 45080 }, { "grad_norm": 0.5003122687339783, "learning_rate": 4.3632874802950136e-05, "loss": 0.0258, "step": 45090 }, { "grad_norm": 0.2718318998813629, "learning_rate": 4.3612375240074034e-05, "loss": 0.0283, "step": 45100 }, { "grad_norm": 0.4925832748413086, "learning_rate": 4.3591876768668325e-05, "loss": 0.0398, "step": 45110 }, { "grad_norm": 0.30227065086364746, "learning_rate": 4.3571379392235605e-05, "loss": 0.0248, "step": 45120 }, { "grad_norm": 0.3861779272556305, "learning_rate": 4.3550883114278335e-05, "loss": 0.0242, "step": 45130 }, { "grad_norm": 0.31615981459617615, "learning_rate": 4.353038793829876e-05, "loss": 0.0297, "step": 45140 }, { "grad_norm": 0.3863179683685303, "learning_rate": 4.350989386779891e-05, "loss": 0.0296, "step": 45150 }, { "grad_norm": 0.27000075578689575, "learning_rate": 4.3489400906280724e-05, "loss": 0.0391, "step": 45160 }, { "grad_norm": 0.3686857521533966, "learning_rate": 4.3468909057245845e-05, "loss": 0.0311, "step": 45170 }, { "grad_norm": 0.31387951970100403, "learning_rate": 4.3448418324195794e-05, "loss": 0.0293, "step": 45180 }, { "grad_norm": 0.37125131487846375, "learning_rate": 4.342792871063184e-05, "loss": 0.0287, "step": 45190 }, { "grad_norm": 0.28709468245506287, "learning_rate": 4.3407440220055145e-05, "loss": 0.0311, "step": 45200 }, { "grad_norm": 0.501924455165863, "learning_rate": 4.33869528559666e-05, "loss": 0.0242, "step": 45210 }, { "grad_norm": 0.3207210600376129, "learning_rate": 4.336646662186696e-05, "loss": 0.0261, "step": 45220 }, { "grad_norm": 0.3297499418258667, "learning_rate": 4.334598152125672e-05, "loss": 0.0305, "step": 45230 }, { "grad_norm": 0.2802707254886627, "learning_rate": 4.3325497557636276e-05, "loss": 0.0322, "step": 45240 }, { "grad_norm": 0.3110077679157257, "learning_rate": 4.330501473450574e-05, "loss": 0.0321, "step": 45250 }, { "grad_norm": 0.35029181838035583, "learning_rate": 4.328453305536507e-05, "loss": 0.0242, "step": 45260 }, { "grad_norm": 0.34738463163375854, "learning_rate": 4.326405252371404e-05, "loss": 0.0252, "step": 45270 }, { "grad_norm": 0.2725614905357361, "learning_rate": 4.324357314305221e-05, "loss": 0.0231, "step": 45280 }, { "grad_norm": 0.2557404637336731, "learning_rate": 4.3223094916878945e-05, "loss": 0.0286, "step": 45290 }, { "grad_norm": 0.21650470793247223, "learning_rate": 4.320261784869338e-05, "loss": 0.0379, "step": 45300 }, { "grad_norm": 0.4166056513786316, "learning_rate": 4.318214194199455e-05, "loss": 0.035, "step": 45310 }, { "grad_norm": 0.2693397104740143, "learning_rate": 4.316166720028118e-05, "loss": 0.0245, "step": 45320 }, { "grad_norm": 0.3534841239452362, "learning_rate": 4.3141193627051864e-05, "loss": 0.0293, "step": 45330 }, { "grad_norm": 0.3848060667514801, "learning_rate": 4.312072122580496e-05, "loss": 0.0267, "step": 45340 }, { "grad_norm": 0.2929958999156952, "learning_rate": 4.3100250000038646e-05, "loss": 0.0314, "step": 45350 }, { "grad_norm": 0.46057596802711487, "learning_rate": 4.307977995325091e-05, "loss": 0.0388, "step": 45360 }, { "grad_norm": 0.3490596413612366, "learning_rate": 4.30593110889395e-05, "loss": 0.0255, "step": 45370 }, { "grad_norm": 0.4606248438358307, "learning_rate": 4.3038843410602016e-05, "loss": 0.025, "step": 45380 }, { "grad_norm": 0.3019281029701233, "learning_rate": 4.3018376921735774e-05, "loss": 0.0302, "step": 45390 }, { "grad_norm": 0.7659102082252502, "learning_rate": 4.299791162583799e-05, "loss": 0.0304, "step": 45400 }, { "grad_norm": 0.24316167831420898, "learning_rate": 4.29774475264056e-05, "loss": 0.0274, "step": 45410 }, { "grad_norm": 0.4225425720214844, "learning_rate": 4.2956984626935365e-05, "loss": 0.0195, "step": 45420 }, { "grad_norm": 0.27620622515678406, "learning_rate": 4.293652293092383e-05, "loss": 0.0257, "step": 45430 }, { "grad_norm": 0.3868531882762909, "learning_rate": 4.2916062441867324e-05, "loss": 0.0319, "step": 45440 }, { "grad_norm": 0.3492678105831146, "learning_rate": 4.289560316326201e-05, "loss": 0.0354, "step": 45450 }, { "grad_norm": 0.46856218576431274, "learning_rate": 4.28751450986038e-05, "loss": 0.0373, "step": 45460 }, { "grad_norm": 0.2853761911392212, "learning_rate": 4.2854688251388444e-05, "loss": 0.0303, "step": 45470 }, { "grad_norm": 0.2857886850833893, "learning_rate": 4.2834232625111425e-05, "loss": 0.0286, "step": 45480 }, { "grad_norm": 0.3352734446525574, "learning_rate": 4.2813778223268086e-05, "loss": 0.0288, "step": 45490 }, { "grad_norm": 0.9156073331832886, "learning_rate": 4.2793325049353477e-05, "loss": 0.0347, "step": 45500 }, { "grad_norm": 0.45847028493881226, "learning_rate": 4.2772873106862535e-05, "loss": 0.0277, "step": 45510 }, { "grad_norm": 0.3628597855567932, "learning_rate": 4.275242239928993e-05, "loss": 0.0305, "step": 45520 }, { "grad_norm": 0.23042216897010803, "learning_rate": 4.273197293013009e-05, "loss": 0.0204, "step": 45530 }, { "grad_norm": 0.3741197884082794, "learning_rate": 4.271152470287731e-05, "loss": 0.0272, "step": 45540 }, { "grad_norm": 0.24756334722042084, "learning_rate": 4.2691077721025594e-05, "loss": 0.0327, "step": 45550 }, { "grad_norm": 0.284393846988678, "learning_rate": 4.267063198806883e-05, "loss": 0.0253, "step": 45560 }, { "grad_norm": 0.3155111074447632, "learning_rate": 4.2650187507500574e-05, "loss": 0.0282, "step": 45570 }, { "grad_norm": 0.3159348666667938, "learning_rate": 4.2629744282814275e-05, "loss": 0.029, "step": 45580 }, { "grad_norm": 0.3273067772388458, "learning_rate": 4.2609302317503074e-05, "loss": 0.024, "step": 45590 }, { "grad_norm": 0.44133371114730835, "learning_rate": 4.258886161505999e-05, "loss": 0.0277, "step": 45600 }, { "grad_norm": 0.2727375626564026, "learning_rate": 4.2568422178977775e-05, "loss": 0.0256, "step": 45610 }, { "grad_norm": 0.2798321843147278, "learning_rate": 4.254798401274894e-05, "loss": 0.0367, "step": 45620 }, { "grad_norm": 0.29585862159729004, "learning_rate": 4.252754711986583e-05, "loss": 0.0276, "step": 45630 }, { "grad_norm": 0.25946691632270813, "learning_rate": 4.250711150382052e-05, "loss": 0.0247, "step": 45640 }, { "grad_norm": 0.3596782982349396, "learning_rate": 4.248667716810495e-05, "loss": 0.031, "step": 45650 }, { "grad_norm": 0.5850317478179932, "learning_rate": 4.246624411621074e-05, "loss": 0.0289, "step": 45660 }, { "grad_norm": 0.3476848900318146, "learning_rate": 4.244581235162938e-05, "loss": 0.0252, "step": 45670 }, { "grad_norm": 0.24298374354839325, "learning_rate": 4.2425381877852075e-05, "loss": 0.0318, "step": 45680 }, { "grad_norm": 0.33447957038879395, "learning_rate": 4.2404952698369856e-05, "loss": 0.028, "step": 45690 }, { "grad_norm": 0.2508474886417389, "learning_rate": 4.238452481667349e-05, "loss": 0.0308, "step": 45700 }, { "grad_norm": 0.33008629083633423, "learning_rate": 4.2364098236253526e-05, "loss": 0.032, "step": 45710 }, { "grad_norm": 0.3515710234642029, "learning_rate": 4.2343672960600356e-05, "loss": 0.0289, "step": 45720 }, { "grad_norm": 0.29946228861808777, "learning_rate": 4.232324899320406e-05, "loss": 0.024, "step": 45730 }, { "grad_norm": 0.2685279846191406, "learning_rate": 4.230282633755457e-05, "loss": 0.0292, "step": 45740 }, { "grad_norm": 0.2833418548107147, "learning_rate": 4.2282404997141515e-05, "loss": 0.0319, "step": 45750 }, { "grad_norm": 0.23205982148647308, "learning_rate": 4.2261984975454397e-05, "loss": 0.0316, "step": 45760 }, { "grad_norm": 0.37563344836235046, "learning_rate": 4.224156627598239e-05, "loss": 0.0281, "step": 45770 }, { "grad_norm": 0.2673056125640869, "learning_rate": 4.222114890221453e-05, "loss": 0.0327, "step": 45780 }, { "grad_norm": 0.32874155044555664, "learning_rate": 4.2200732857639546e-05, "loss": 0.035, "step": 45790 }, { "grad_norm": 0.2751951515674591, "learning_rate": 4.2180318145746035e-05, "loss": 0.0255, "step": 45800 }, { "grad_norm": 0.36420610547065735, "learning_rate": 4.215990477002227e-05, "loss": 0.0313, "step": 45810 }, { "grad_norm": 0.24560418725013733, "learning_rate": 4.2139492733956356e-05, "loss": 0.0259, "step": 45820 }, { "grad_norm": 0.40422773361206055, "learning_rate": 4.211908204103615e-05, "loss": 0.0322, "step": 45830 }, { "grad_norm": 0.3296053409576416, "learning_rate": 4.2098672694749265e-05, "loss": 0.0424, "step": 45840 }, { "grad_norm": 0.24360348284244537, "learning_rate": 4.2078264698583133e-05, "loss": 0.038, "step": 45850 }, { "grad_norm": 0.25787559151649475, "learning_rate": 4.205785805602488e-05, "loss": 0.0294, "step": 45860 }, { "grad_norm": 0.30873680114746094, "learning_rate": 4.203745277056149e-05, "loss": 0.03, "step": 45870 }, { "grad_norm": 0.31389951705932617, "learning_rate": 4.201704884567964e-05, "loss": 0.0277, "step": 45880 }, { "grad_norm": 0.6120323538780212, "learning_rate": 4.1996646284865816e-05, "loss": 0.0412, "step": 45890 }, { "grad_norm": 0.30092552304267883, "learning_rate": 4.197624509160625e-05, "loss": 0.0283, "step": 45900 }, { "grad_norm": 0.2592684030532837, "learning_rate": 4.195584526938692e-05, "loss": 0.0251, "step": 45910 }, { "grad_norm": 0.284191370010376, "learning_rate": 4.193544682169365e-05, "loss": 0.0293, "step": 45920 }, { "grad_norm": 0.4397244155406952, "learning_rate": 4.1915049752011946e-05, "loss": 0.0315, "step": 45930 }, { "grad_norm": 0.24338684976100922, "learning_rate": 4.189465406382712e-05, "loss": 0.022, "step": 45940 }, { "grad_norm": 0.3510124087333679, "learning_rate": 4.187425976062422e-05, "loss": 0.0258, "step": 45950 }, { "grad_norm": 0.32802167534828186, "learning_rate": 4.18538668458881e-05, "loss": 0.0255, "step": 45960 }, { "grad_norm": 0.27483391761779785, "learning_rate": 4.183347532310333e-05, "loss": 0.0256, "step": 45970 }, { "grad_norm": 0.296114444732666, "learning_rate": 4.181308519575429e-05, "loss": 0.0321, "step": 45980 }, { "grad_norm": 0.38025960326194763, "learning_rate": 4.179269646732507e-05, "loss": 0.0394, "step": 45990 }, { "grad_norm": 0.2743109464645386, "learning_rate": 4.177230914129954e-05, "loss": 0.0293, "step": 46000 }, { "grad_norm": 0.4368826448917389, "learning_rate": 4.175192322116136e-05, "loss": 0.0268, "step": 46010 }, { "grad_norm": 0.2875835597515106, "learning_rate": 4.173153871039391e-05, "loss": 0.0224, "step": 46020 }, { "grad_norm": 0.4854128062725067, "learning_rate": 4.171115561248036e-05, "loss": 0.0306, "step": 46030 }, { "grad_norm": 0.2737292945384979, "learning_rate": 4.16907739309036e-05, "loss": 0.0259, "step": 46040 }, { "grad_norm": 0.36984649300575256, "learning_rate": 4.167039366914633e-05, "loss": 0.0212, "step": 46050 }, { "grad_norm": 0.32584148645401, "learning_rate": 4.165001483069096e-05, "loss": 0.0342, "step": 46060 }, { "grad_norm": 0.26898810267448425, "learning_rate": 4.16296374190197e-05, "loss": 0.0319, "step": 46070 }, { "grad_norm": 0.3225986063480377, "learning_rate": 4.1609261437614464e-05, "loss": 0.0254, "step": 46080 }, { "grad_norm": 0.48525270819664, "learning_rate": 4.158888688995696e-05, "loss": 0.0318, "step": 46090 }, { "grad_norm": 0.4150524139404297, "learning_rate": 4.1568513779528645e-05, "loss": 0.0285, "step": 46100 }, { "grad_norm": 0.33763498067855835, "learning_rate": 4.1548142109810704e-05, "loss": 0.0351, "step": 46110 }, { "grad_norm": 0.30947190523147583, "learning_rate": 4.152777188428414e-05, "loss": 0.0246, "step": 46120 }, { "grad_norm": 0.41474688053131104, "learning_rate": 4.1507403106429646e-05, "loss": 0.0293, "step": 46130 }, { "grad_norm": 0.35814759135246277, "learning_rate": 4.148703577972768e-05, "loss": 0.0215, "step": 46140 }, { "grad_norm": 0.3415694236755371, "learning_rate": 4.146666990765846e-05, "loss": 0.0259, "step": 46150 }, { "grad_norm": 0.35139939188957214, "learning_rate": 4.1446305493701986e-05, "loss": 0.0341, "step": 46160 }, { "grad_norm": 0.38946273922920227, "learning_rate": 4.142594254133796e-05, "loss": 0.0262, "step": 46170 }, { "grad_norm": 0.3094348907470703, "learning_rate": 4.1405581054045835e-05, "loss": 0.03, "step": 46180 }, { "grad_norm": 0.28157123923301697, "learning_rate": 4.1385221035304864e-05, "loss": 0.027, "step": 46190 }, { "grad_norm": 0.3186221122741699, "learning_rate": 4.136486248859398e-05, "loss": 0.0299, "step": 46200 }, { "grad_norm": 0.29909831285476685, "learning_rate": 4.1344505417391955e-05, "loss": 0.033, "step": 46210 }, { "grad_norm": 0.36959779262542725, "learning_rate": 4.132414982517721e-05, "loss": 0.031, "step": 46220 }, { "grad_norm": 0.2284223735332489, "learning_rate": 4.130379571542798e-05, "loss": 0.0265, "step": 46230 }, { "grad_norm": 0.35877859592437744, "learning_rate": 4.12834430916222e-05, "loss": 0.0315, "step": 46240 }, { "grad_norm": 0.43133795261383057, "learning_rate": 4.126309195723763e-05, "loss": 0.0288, "step": 46250 }, { "grad_norm": 0.263556569814682, "learning_rate": 4.124274231575168e-05, "loss": 0.0277, "step": 46260 }, { "grad_norm": 0.23777438700199127, "learning_rate": 4.122239417064154e-05, "loss": 0.0219, "step": 46270 }, { "grad_norm": 0.3847959339618683, "learning_rate": 4.1202047525384184e-05, "loss": 0.0234, "step": 46280 }, { "grad_norm": 0.30906933546066284, "learning_rate": 4.118170238345627e-05, "loss": 0.0243, "step": 46290 }, { "grad_norm": 0.4214986264705658, "learning_rate": 4.1161358748334256e-05, "loss": 0.0264, "step": 46300 }, { "grad_norm": 0.33431997895240784, "learning_rate": 4.1141016623494266e-05, "loss": 0.0295, "step": 46310 }, { "grad_norm": 0.3144509792327881, "learning_rate": 4.112067601241227e-05, "loss": 0.0252, "step": 46320 }, { "grad_norm": 0.27923259139060974, "learning_rate": 4.110033691856387e-05, "loss": 0.0229, "step": 46330 }, { "grad_norm": 0.3841714560985565, "learning_rate": 4.107999934542451e-05, "loss": 0.0338, "step": 46340 }, { "grad_norm": 0.34178176522254944, "learning_rate": 4.105966329646928e-05, "loss": 0.026, "step": 46350 }, { "grad_norm": 0.32275858521461487, "learning_rate": 4.103932877517308e-05, "loss": 0.0273, "step": 46360 }, { "grad_norm": 0.2491082400083542, "learning_rate": 4.101899578501052e-05, "loss": 0.0212, "step": 46370 }, { "grad_norm": 0.36870336532592773, "learning_rate": 4.099866432945595e-05, "loss": 0.0282, "step": 46380 }, { "grad_norm": 0.42784643173217773, "learning_rate": 4.097833441198346e-05, "loss": 0.0308, "step": 46390 }, { "grad_norm": 0.27821481227874756, "learning_rate": 4.095800603606685e-05, "loss": 0.0258, "step": 46400 }, { "grad_norm": 0.31716880202293396, "learning_rate": 4.093767920517975e-05, "loss": 0.0328, "step": 46410 }, { "grad_norm": 0.4541681110858917, "learning_rate": 4.091735392279539e-05, "loss": 0.0332, "step": 46420 }, { "grad_norm": 0.258026659488678, "learning_rate": 4.089703019238685e-05, "loss": 0.021, "step": 46430 }, { "grad_norm": 0.26864394545555115, "learning_rate": 4.0876708017426866e-05, "loss": 0.0249, "step": 46440 }, { "grad_norm": 0.45862677693367004, "learning_rate": 4.085638740138798e-05, "loss": 0.0242, "step": 46450 }, { "grad_norm": 0.28972887992858887, "learning_rate": 4.08360683477424e-05, "loss": 0.0266, "step": 46460 }, { "grad_norm": 0.3989526629447937, "learning_rate": 4.0815750859962085e-05, "loss": 0.0317, "step": 46470 }, { "grad_norm": 0.30475062131881714, "learning_rate": 4.079543494151879e-05, "loss": 0.0301, "step": 46480 }, { "grad_norm": 0.29639920592308044, "learning_rate": 4.07751205958839e-05, "loss": 0.0242, "step": 46490 }, { "grad_norm": 0.24377121031284332, "learning_rate": 4.0754807826528615e-05, "loss": 0.0214, "step": 46500 }, { "grad_norm": 0.4106605052947998, "learning_rate": 4.073449663692379e-05, "loss": 0.031, "step": 46510 }, { "grad_norm": 0.38424625992774963, "learning_rate": 4.07141870305401e-05, "loss": 0.0192, "step": 46520 }, { "grad_norm": 0.37500080466270447, "learning_rate": 4.0693879010847866e-05, "loss": 0.0431, "step": 46530 }, { "grad_norm": 0.41823050379753113, "learning_rate": 4.06735725813172e-05, "loss": 0.0284, "step": 46540 }, { "grad_norm": 0.3947899341583252, "learning_rate": 4.0653267745417903e-05, "loss": 0.0212, "step": 46550 }, { "grad_norm": 0.3958407938480377, "learning_rate": 4.063296450661949e-05, "loss": 0.0227, "step": 46560 }, { "grad_norm": 0.4290772080421448, "learning_rate": 4.061266286839128e-05, "loss": 0.031, "step": 46570 }, { "grad_norm": 0.3525260388851166, "learning_rate": 4.0592362834202225e-05, "loss": 0.0279, "step": 46580 }, { "grad_norm": 0.4869534373283386, "learning_rate": 4.057206440752107e-05, "loss": 0.0206, "step": 46590 }, { "grad_norm": 0.27462661266326904, "learning_rate": 4.0551767591816245e-05, "loss": 0.0237, "step": 46600 }, { "grad_norm": 0.33154675364494324, "learning_rate": 4.0531472390555935e-05, "loss": 0.0312, "step": 46610 }, { "grad_norm": 0.41557303071022034, "learning_rate": 4.051117880720802e-05, "loss": 0.0258, "step": 46620 }, { "grad_norm": 0.22937262058258057, "learning_rate": 4.049088684524015e-05, "loss": 0.0211, "step": 46630 }, { "grad_norm": 0.4086650311946869, "learning_rate": 4.0470596508119636e-05, "loss": 0.0246, "step": 46640 }, { "grad_norm": 0.4825632870197296, "learning_rate": 4.0450307799313524e-05, "loss": 0.03, "step": 46650 }, { "grad_norm": 0.39524537324905396, "learning_rate": 4.0430020722288656e-05, "loss": 0.0309, "step": 46660 }, { "grad_norm": 0.5192285180091858, "learning_rate": 4.040973528051148e-05, "loss": 0.0263, "step": 46670 }, { "grad_norm": 0.224782794713974, "learning_rate": 4.038945147744827e-05, "loss": 0.0298, "step": 46680 }, { "grad_norm": 0.20038239657878876, "learning_rate": 4.0369169316564945e-05, "loss": 0.0221, "step": 46690 }, { "grad_norm": 0.3316596746444702, "learning_rate": 4.034888880132718e-05, "loss": 0.0238, "step": 46700 }, { "grad_norm": 0.6273772716522217, "learning_rate": 4.032860993520035e-05, "loss": 0.0283, "step": 46710 }, { "grad_norm": 0.27125978469848633, "learning_rate": 4.030833272164959e-05, "loss": 0.0262, "step": 46720 }, { "grad_norm": 0.32504764199256897, "learning_rate": 4.0288057164139705e-05, "loss": 0.0343, "step": 46730 }, { "grad_norm": 0.21121715009212494, "learning_rate": 4.0267783266135205e-05, "loss": 0.0228, "step": 46740 }, { "grad_norm": 0.22555620968341827, "learning_rate": 4.024751103110039e-05, "loss": 0.0288, "step": 46750 }, { "grad_norm": 0.30918562412261963, "learning_rate": 4.0227240462499176e-05, "loss": 0.0322, "step": 46760 }, { "grad_norm": 0.25579550862312317, "learning_rate": 4.020697156379531e-05, "loss": 0.0317, "step": 46770 }, { "grad_norm": 0.39754733443260193, "learning_rate": 4.018670433845215e-05, "loss": 0.0253, "step": 46780 }, { "grad_norm": 0.3568943738937378, "learning_rate": 4.016643878993284e-05, "loss": 0.0232, "step": 46790 }, { "grad_norm": 0.3298322260379791, "learning_rate": 4.014617492170017e-05, "loss": 0.0301, "step": 46800 }, { "grad_norm": 0.25930166244506836, "learning_rate": 4.0125912737216726e-05, "loss": 0.0384, "step": 46810 }, { "grad_norm": 0.34290194511413574, "learning_rate": 4.0105652239944735e-05, "loss": 0.0309, "step": 46820 }, { "grad_norm": 0.39021599292755127, "learning_rate": 4.0085393433346144e-05, "loss": 0.0344, "step": 46830 }, { "grad_norm": 0.5418732762336731, "learning_rate": 4.006513632088268e-05, "loss": 0.0233, "step": 46840 }, { "grad_norm": 0.27108490467071533, "learning_rate": 4.004488090601567e-05, "loss": 0.0234, "step": 46850 }, { "grad_norm": 0.23385530710220337, "learning_rate": 4.002462719220626e-05, "loss": 0.0237, "step": 46860 }, { "grad_norm": 0.3319096267223358, "learning_rate": 4.000437518291522e-05, "loss": 0.0299, "step": 46870 }, { "grad_norm": 0.37542420625686646, "learning_rate": 3.9984124881603094e-05, "loss": 0.0276, "step": 46880 }, { "grad_norm": 0.24409034848213196, "learning_rate": 3.9963876291730086e-05, "loss": 0.0222, "step": 46890 }, { "grad_norm": 0.29179805517196655, "learning_rate": 3.994362941675614e-05, "loss": 0.0209, "step": 46900 }, { "grad_norm": 0.3258097171783447, "learning_rate": 3.992338426014088e-05, "loss": 0.0255, "step": 46910 }, { "grad_norm": 0.28899523615837097, "learning_rate": 3.9903140825343636e-05, "loss": 0.0332, "step": 46920 }, { "grad_norm": 0.330291748046875, "learning_rate": 3.98828991158235e-05, "loss": 0.0216, "step": 46930 }, { "grad_norm": 0.36605870723724365, "learning_rate": 3.9862659135039185e-05, "loss": 0.0335, "step": 46940 }, { "grad_norm": 0.419974684715271, "learning_rate": 3.984242088644918e-05, "loss": 0.0275, "step": 46950 }, { "grad_norm": 0.23833385109901428, "learning_rate": 3.9822184373511615e-05, "loss": 0.0226, "step": 46960 }, { "grad_norm": 0.4253326654434204, "learning_rate": 3.98019495996844e-05, "loss": 0.0266, "step": 46970 }, { "grad_norm": 0.2744506895542145, "learning_rate": 3.978171656842507e-05, "loss": 0.0296, "step": 46980 }, { "grad_norm": 0.31338950991630554, "learning_rate": 3.976148528319091e-05, "loss": 0.0237, "step": 46990 }, { "grad_norm": 0.36722299456596375, "learning_rate": 3.974125574743888e-05, "loss": 0.0308, "step": 47000 }, { "grad_norm": 0.42060181498527527, "learning_rate": 3.9721027964625686e-05, "loss": 0.0348, "step": 47010 }, { "grad_norm": 0.3033559322357178, "learning_rate": 3.9700801938207676e-05, "loss": 0.027, "step": 47020 }, { "grad_norm": 0.36764824390411377, "learning_rate": 3.9680577671640916e-05, "loss": 0.0295, "step": 47030 }, { "grad_norm": 0.3145565688610077, "learning_rate": 3.966035516838121e-05, "loss": 0.0354, "step": 47040 }, { "grad_norm": 0.3488847315311432, "learning_rate": 3.9640134431884014e-05, "loss": 0.0213, "step": 47050 }, { "grad_norm": 0.3172609508037567, "learning_rate": 3.961991546560451e-05, "loss": 0.0195, "step": 47060 }, { "grad_norm": 0.34454983472824097, "learning_rate": 3.959969827299753e-05, "loss": 0.0267, "step": 47070 }, { "grad_norm": 0.38989946246147156, "learning_rate": 3.9579482857517684e-05, "loss": 0.0289, "step": 47080 }, { "grad_norm": 0.30323898792266846, "learning_rate": 3.955926922261921e-05, "loss": 0.0263, "step": 47090 }, { "grad_norm": 0.23959006369113922, "learning_rate": 3.9539057371756084e-05, "loss": 0.0274, "step": 47100 }, { "grad_norm": 0.3564640283584595, "learning_rate": 3.951884730838195e-05, "loss": 0.0206, "step": 47110 }, { "grad_norm": 0.26611486077308655, "learning_rate": 3.949863903595012e-05, "loss": 0.0213, "step": 47120 }, { "grad_norm": 0.348861426115036, "learning_rate": 3.947843255791369e-05, "loss": 0.0327, "step": 47130 }, { "grad_norm": 0.3498615026473999, "learning_rate": 3.9458227877725364e-05, "loss": 0.0363, "step": 47140 }, { "grad_norm": 0.34426623582839966, "learning_rate": 3.943802499883758e-05, "loss": 0.0302, "step": 47150 }, { "grad_norm": 0.373139351606369, "learning_rate": 3.9417823924702437e-05, "loss": 0.0333, "step": 47160 }, { "grad_norm": 0.3966515362262726, "learning_rate": 3.939762465877178e-05, "loss": 0.0318, "step": 47170 }, { "grad_norm": 0.3381877541542053, "learning_rate": 3.937742720449708e-05, "loss": 0.0317, "step": 47180 }, { "grad_norm": 0.44791021943092346, "learning_rate": 3.9357231565329563e-05, "loss": 0.0244, "step": 47190 }, { "grad_norm": 0.23999552428722382, "learning_rate": 3.933703774472008e-05, "loss": 0.0212, "step": 47200 }, { "grad_norm": 0.45552489161491394, "learning_rate": 3.93168457461192e-05, "loss": 0.0271, "step": 47210 }, { "grad_norm": 0.42440029978752136, "learning_rate": 3.9296655572977216e-05, "loss": 0.0272, "step": 47220 }, { "grad_norm": 0.3283584713935852, "learning_rate": 3.927646722874404e-05, "loss": 0.034, "step": 47230 }, { "grad_norm": 0.3073466122150421, "learning_rate": 3.925628071686934e-05, "loss": 0.0276, "step": 47240 }, { "grad_norm": 0.22431063652038574, "learning_rate": 3.9236096040802415e-05, "loss": 0.0249, "step": 47250 }, { "grad_norm": 0.4970540702342987, "learning_rate": 3.9215913203992294e-05, "loss": 0.0413, "step": 47260 }, { "grad_norm": 0.21515503525733948, "learning_rate": 3.9195732209887645e-05, "loss": 0.0261, "step": 47270 }, { "grad_norm": 0.2536461055278778, "learning_rate": 3.9175553061936875e-05, "loss": 0.0301, "step": 47280 }, { "grad_norm": 0.21427392959594727, "learning_rate": 3.9155375763588045e-05, "loss": 0.0326, "step": 47290 }, { "grad_norm": 0.30812907218933105, "learning_rate": 3.913520031828889e-05, "loss": 0.0283, "step": 47300 }, { "grad_norm": 0.281379371881485, "learning_rate": 3.911502672948685e-05, "loss": 0.0298, "step": 47310 }, { "grad_norm": 0.30420878529548645, "learning_rate": 3.9094855000629014e-05, "loss": 0.0252, "step": 47320 }, { "grad_norm": 0.4571322500705719, "learning_rate": 3.907468513516223e-05, "loss": 0.0245, "step": 47330 }, { "grad_norm": 0.3477184474468231, "learning_rate": 3.905451713653294e-05, "loss": 0.028, "step": 47340 }, { "grad_norm": 0.305111289024353, "learning_rate": 3.903435100818731e-05, "loss": 0.027, "step": 47350 }, { "grad_norm": 0.405333548784256, "learning_rate": 3.901418675357117e-05, "loss": 0.0304, "step": 47360 }, { "grad_norm": 0.2990652024745941, "learning_rate": 3.8994024376130075e-05, "loss": 0.0308, "step": 47370 }, { "grad_norm": 0.32655060291290283, "learning_rate": 3.8973863879309194e-05, "loss": 0.0317, "step": 47380 }, { "grad_norm": 0.2402873933315277, "learning_rate": 3.8953705266553394e-05, "loss": 0.0221, "step": 47390 }, { "grad_norm": 0.4085114300251007, "learning_rate": 3.893354854130727e-05, "loss": 0.0314, "step": 47400 }, { "grad_norm": 0.3525473475456238, "learning_rate": 3.8913393707015006e-05, "loss": 0.0236, "step": 47410 }, { "grad_norm": 0.2956635653972626, "learning_rate": 3.889324076712056e-05, "loss": 0.0332, "step": 47420 }, { "grad_norm": 0.3424193263053894, "learning_rate": 3.8873089725067476e-05, "loss": 0.0252, "step": 47430 }, { "grad_norm": 0.2640986144542694, "learning_rate": 3.885294058429905e-05, "loss": 0.0348, "step": 47440 }, { "grad_norm": 0.34575602412223816, "learning_rate": 3.8832793348258206e-05, "loss": 0.0274, "step": 47450 }, { "grad_norm": 0.3607881963253021, "learning_rate": 3.881264802038756e-05, "loss": 0.0399, "step": 47460 }, { "grad_norm": 0.43171486258506775, "learning_rate": 3.879250460412939e-05, "loss": 0.0328, "step": 47470 }, { "grad_norm": 0.2921178340911865, "learning_rate": 3.8772363102925644e-05, "loss": 0.0304, "step": 47480 }, { "grad_norm": 0.346824049949646, "learning_rate": 3.875222352021798e-05, "loss": 0.0218, "step": 47490 }, { "grad_norm": 0.26934435963630676, "learning_rate": 3.8732085859447686e-05, "loss": 0.0271, "step": 47500 }, { "grad_norm": 0.28378841280937195, "learning_rate": 3.871195012405575e-05, "loss": 0.0268, "step": 47510 }, { "grad_norm": 0.3921985924243927, "learning_rate": 3.869181631748278e-05, "loss": 0.0233, "step": 47520 }, { "grad_norm": 0.31834471225738525, "learning_rate": 3.867168444316915e-05, "loss": 0.0246, "step": 47530 }, { "grad_norm": 0.5433662533760071, "learning_rate": 3.865155450455481e-05, "loss": 0.0271, "step": 47540 }, { "grad_norm": 0.3042515814304352, "learning_rate": 3.8631426505079426e-05, "loss": 0.0253, "step": 47550 }, { "grad_norm": 0.43740323185920715, "learning_rate": 3.8611300448182304e-05, "loss": 0.0391, "step": 47560 }, { "grad_norm": 0.3607327342033386, "learning_rate": 3.859117633730248e-05, "loss": 0.0309, "step": 47570 }, { "grad_norm": 0.38317564129829407, "learning_rate": 3.857105417587858e-05, "loss": 0.031, "step": 47580 }, { "grad_norm": 0.27325567603111267, "learning_rate": 3.855093396734894e-05, "loss": 0.0224, "step": 47590 }, { "grad_norm": 0.3012843132019043, "learning_rate": 3.8530815715151545e-05, "loss": 0.0406, "step": 47600 }, { "grad_norm": 0.2768773138523102, "learning_rate": 3.851069942272405e-05, "loss": 0.0232, "step": 47610 }, { "grad_norm": 0.21261286735534668, "learning_rate": 3.849058509350382e-05, "loss": 0.0293, "step": 47620 }, { "grad_norm": 0.29008227586746216, "learning_rate": 3.8470472730927783e-05, "loss": 0.0358, "step": 47630 }, { "grad_norm": 0.41885578632354736, "learning_rate": 3.845036233843264e-05, "loss": 0.0273, "step": 47640 }, { "grad_norm": 0.38187122344970703, "learning_rate": 3.843025391945469e-05, "loss": 0.0271, "step": 47650 }, { "grad_norm": 0.34153181314468384, "learning_rate": 3.841014747742992e-05, "loss": 0.0292, "step": 47660 }, { "grad_norm": 0.26196396350860596, "learning_rate": 3.839004301579397e-05, "loss": 0.0264, "step": 47670 }, { "grad_norm": 0.248494952917099, "learning_rate": 3.8369940537982097e-05, "loss": 0.0297, "step": 47680 }, { "grad_norm": 0.3299236595630646, "learning_rate": 3.834984004742933e-05, "loss": 0.0234, "step": 47690 }, { "grad_norm": 0.5721471905708313, "learning_rate": 3.832974154757026e-05, "loss": 0.0257, "step": 47700 }, { "grad_norm": 0.2526814937591553, "learning_rate": 3.830964504183919e-05, "loss": 0.0264, "step": 47710 }, { "grad_norm": 0.3205870985984802, "learning_rate": 3.828955053367003e-05, "loss": 0.0271, "step": 47720 }, { "grad_norm": 0.3079828917980194, "learning_rate": 3.826945802649642e-05, "loss": 0.0258, "step": 47730 }, { "grad_norm": 0.2052823007106781, "learning_rate": 3.824936752375159e-05, "loss": 0.0347, "step": 47740 }, { "grad_norm": 0.31250855326652527, "learning_rate": 3.822927902886848e-05, "loss": 0.0266, "step": 47750 }, { "grad_norm": 0.3118168115615845, "learning_rate": 3.8209192545279653e-05, "loss": 0.0302, "step": 47760 }, { "grad_norm": 0.40889427065849304, "learning_rate": 3.8189108076417326e-05, "loss": 0.0271, "step": 47770 }, { "grad_norm": 0.43541350960731506, "learning_rate": 3.816902562571342e-05, "loss": 0.0397, "step": 47780 }, { "grad_norm": 0.8749641180038452, "learning_rate": 3.814894519659944e-05, "loss": 0.0252, "step": 47790 }, { "grad_norm": 0.1979866474866867, "learning_rate": 3.812886679250661e-05, "loss": 0.0237, "step": 47800 }, { "grad_norm": 0.16955268383026123, "learning_rate": 3.810879041686575e-05, "loss": 0.0193, "step": 47810 }, { "grad_norm": 0.4665316939353943, "learning_rate": 3.808871607310741e-05, "loss": 0.0337, "step": 47820 }, { "grad_norm": 0.370583713054657, "learning_rate": 3.806864376466169e-05, "loss": 0.0301, "step": 47830 }, { "grad_norm": 0.26180943846702576, "learning_rate": 3.804857349495845e-05, "loss": 0.028, "step": 47840 }, { "grad_norm": 0.3363507091999054, "learning_rate": 3.802850526742713e-05, "loss": 0.0444, "step": 47850 }, { "grad_norm": 0.27222996950149536, "learning_rate": 3.800843908549683e-05, "loss": 0.0252, "step": 47860 }, { "grad_norm": 0.249431774020195, "learning_rate": 3.7988374952596325e-05, "loss": 0.0223, "step": 47870 }, { "grad_norm": 0.38254109025001526, "learning_rate": 3.7968312872154e-05, "loss": 0.024, "step": 47880 }, { "grad_norm": 0.36510151624679565, "learning_rate": 3.7948252847597965e-05, "loss": 0.0277, "step": 47890 }, { "grad_norm": 0.3064848482608795, "learning_rate": 3.7928194882355885e-05, "loss": 0.0278, "step": 47900 }, { "grad_norm": 0.477247953414917, "learning_rate": 3.790813897985515e-05, "loss": 0.0262, "step": 47910 }, { "grad_norm": 0.4618920087814331, "learning_rate": 3.7888085143522726e-05, "loss": 0.0342, "step": 47920 }, { "grad_norm": 0.3441004455089569, "learning_rate": 3.7868033376785314e-05, "loss": 0.0295, "step": 47930 }, { "grad_norm": 0.35796597599983215, "learning_rate": 3.784798368306919e-05, "loss": 0.0218, "step": 47940 }, { "grad_norm": 0.28671151399612427, "learning_rate": 3.782793606580029e-05, "loss": 0.0229, "step": 47950 }, { "grad_norm": 0.2510596215724945, "learning_rate": 3.7807890528404205e-05, "loss": 0.027, "step": 47960 }, { "grad_norm": 0.18872316181659698, "learning_rate": 3.778784707430616e-05, "loss": 0.0264, "step": 47970 }, { "grad_norm": 0.31311601400375366, "learning_rate": 3.776780570693107e-05, "loss": 0.0195, "step": 47980 }, { "grad_norm": 0.3833637237548828, "learning_rate": 3.774776642970342e-05, "loss": 0.0301, "step": 47990 }, { "grad_norm": 0.3310626447200775, "learning_rate": 3.77277292460474e-05, "loss": 0.025, "step": 48000 }, { "grad_norm": 0.4517248868942261, "learning_rate": 3.770769415938678e-05, "loss": 0.0256, "step": 48010 }, { "grad_norm": 0.33753255009651184, "learning_rate": 3.768766117314506e-05, "loss": 0.0312, "step": 48020 }, { "grad_norm": 0.35591578483581543, "learning_rate": 3.76676302907453e-05, "loss": 0.0289, "step": 48030 }, { "grad_norm": 0.6968933343887329, "learning_rate": 3.764760151561021e-05, "loss": 0.0302, "step": 48040 }, { "grad_norm": 0.2952387034893036, "learning_rate": 3.76275748511622e-05, "loss": 0.0252, "step": 48050 }, { "grad_norm": 0.24918031692504883, "learning_rate": 3.7607550300823255e-05, "loss": 0.0284, "step": 48060 }, { "grad_norm": 0.35528281331062317, "learning_rate": 3.7587527868015044e-05, "loss": 0.0394, "step": 48070 }, { "grad_norm": 0.4030306041240692, "learning_rate": 3.756750755615881e-05, "loss": 0.0338, "step": 48080 }, { "grad_norm": 0.2728749215602875, "learning_rate": 3.754748936867553e-05, "loss": 0.0272, "step": 48090 }, { "grad_norm": 0.34170830249786377, "learning_rate": 3.752747330898573e-05, "loss": 0.0284, "step": 48100 }, { "grad_norm": 0.2780967354774475, "learning_rate": 3.750745938050962e-05, "loss": 0.0303, "step": 48110 }, { "grad_norm": 0.28423693776130676, "learning_rate": 3.7487447586667025e-05, "loss": 0.0259, "step": 48120 }, { "grad_norm": 0.26810377836227417, "learning_rate": 3.7467437930877426e-05, "loss": 0.0229, "step": 48130 }, { "grad_norm": 0.36930570006370544, "learning_rate": 3.744743041655992e-05, "loss": 0.023, "step": 48140 }, { "grad_norm": 0.3534263074398041, "learning_rate": 3.742742504713324e-05, "loss": 0.0231, "step": 48150 }, { "grad_norm": 0.2702306807041168, "learning_rate": 3.740742182601576e-05, "loss": 0.0281, "step": 48160 }, { "grad_norm": 0.27330487966537476, "learning_rate": 3.7387420756625464e-05, "loss": 0.0277, "step": 48170 }, { "grad_norm": 0.3555448353290558, "learning_rate": 3.736742184238002e-05, "loss": 0.0254, "step": 48180 }, { "grad_norm": 0.3923838436603546, "learning_rate": 3.7347425086696684e-05, "loss": 0.0251, "step": 48190 }, { "grad_norm": 0.5096080899238586, "learning_rate": 3.732743049299235e-05, "loss": 0.0287, "step": 48200 }, { "grad_norm": 0.180882066488266, "learning_rate": 3.730743806468354e-05, "loss": 0.0183, "step": 48210 }, { "grad_norm": 0.2831237018108368, "learning_rate": 3.7287447805186436e-05, "loss": 0.0216, "step": 48220 }, { "grad_norm": 0.5957297682762146, "learning_rate": 3.726745971791682e-05, "loss": 0.0352, "step": 48230 }, { "grad_norm": 0.4360227882862091, "learning_rate": 3.724747380629008e-05, "loss": 0.0398, "step": 48240 }, { "grad_norm": 0.2771076261997223, "learning_rate": 3.72274900737213e-05, "loss": 0.0293, "step": 48250 }, { "grad_norm": 0.42921578884124756, "learning_rate": 3.7207508523625123e-05, "loss": 0.0321, "step": 48260 }, { "grad_norm": 0.2478051632642746, "learning_rate": 3.718752915941588e-05, "loss": 0.0218, "step": 48270 }, { "grad_norm": 0.3792790174484253, "learning_rate": 3.7167551984507464e-05, "loss": 0.0299, "step": 48280 }, { "grad_norm": 0.39694175124168396, "learning_rate": 3.714757700231346e-05, "loss": 0.0279, "step": 48290 }, { "grad_norm": 0.37563571333885193, "learning_rate": 3.712760421624703e-05, "loss": 0.0239, "step": 48300 }, { "grad_norm": 0.7385362386703491, "learning_rate": 3.710763362972099e-05, "loss": 0.0444, "step": 48310 }, { "grad_norm": 0.3450705409049988, "learning_rate": 3.708766524614774e-05, "loss": 0.0286, "step": 48320 }, { "grad_norm": 0.17763298749923706, "learning_rate": 3.7067699068939335e-05, "loss": 0.0184, "step": 48330 }, { "grad_norm": 0.2574061453342438, "learning_rate": 3.704773510150748e-05, "loss": 0.0257, "step": 48340 }, { "grad_norm": 0.3873981833457947, "learning_rate": 3.702777334726344e-05, "loss": 0.0284, "step": 48350 }, { "grad_norm": 0.3265822231769562, "learning_rate": 3.7007813809618164e-05, "loss": 0.0213, "step": 48360 }, { "grad_norm": 0.2928823232650757, "learning_rate": 3.6987856491982145e-05, "loss": 0.029, "step": 48370 }, { "grad_norm": 0.27526354789733887, "learning_rate": 3.696790139776558e-05, "loss": 0.0285, "step": 48380 }, { "grad_norm": 0.2588725686073303, "learning_rate": 3.6947948530378235e-05, "loss": 0.0293, "step": 48390 }, { "grad_norm": 0.2434828132390976, "learning_rate": 3.6927997893229516e-05, "loss": 0.0293, "step": 48400 }, { "grad_norm": 0.24149556457996368, "learning_rate": 3.690804948972845e-05, "loss": 0.0266, "step": 48410 }, { "grad_norm": 0.20311303436756134, "learning_rate": 3.688810332328363e-05, "loss": 0.0279, "step": 48420 }, { "grad_norm": 0.30417826771736145, "learning_rate": 3.686815939730336e-05, "loss": 0.0255, "step": 48430 }, { "grad_norm": 0.24473892152309418, "learning_rate": 3.684821771519548e-05, "loss": 0.0276, "step": 48440 }, { "grad_norm": 0.3889055550098419, "learning_rate": 3.68282782803675e-05, "loss": 0.0264, "step": 48450 }, { "grad_norm": 0.3055255115032196, "learning_rate": 3.6808341096226504e-05, "loss": 0.0366, "step": 48460 }, { "grad_norm": 0.5767792463302612, "learning_rate": 3.678840616617924e-05, "loss": 0.0314, "step": 48470 }, { "grad_norm": 0.33239516615867615, "learning_rate": 3.6768473493632e-05, "loss": 0.0318, "step": 48480 }, { "grad_norm": 0.4097796678543091, "learning_rate": 3.6748543081990783e-05, "loss": 0.0251, "step": 48490 }, { "grad_norm": 0.2401820719242096, "learning_rate": 3.672861493466112e-05, "loss": 0.0262, "step": 48500 }, { "grad_norm": 0.23431788384914398, "learning_rate": 3.670868905504818e-05, "loss": 0.0248, "step": 48510 }, { "grad_norm": 0.3045526146888733, "learning_rate": 3.6688765446556784e-05, "loss": 0.027, "step": 48520 }, { "grad_norm": 0.26186665892601013, "learning_rate": 3.6668844112591276e-05, "loss": 0.0286, "step": 48530 }, { "grad_norm": 0.3095360994338989, "learning_rate": 3.664892505655573e-05, "loss": 0.0265, "step": 48540 }, { "grad_norm": 0.28217095136642456, "learning_rate": 3.662900828185373e-05, "loss": 0.0296, "step": 48550 }, { "grad_norm": 0.39844948053359985, "learning_rate": 3.6609093791888516e-05, "loss": 0.0241, "step": 48560 }, { "grad_norm": 0.27866366505622864, "learning_rate": 3.658918159006292e-05, "loss": 0.0328, "step": 48570 }, { "grad_norm": 0.24276021122932434, "learning_rate": 3.656927167977942e-05, "loss": 0.0271, "step": 48580 }, { "grad_norm": 0.4324358105659485, "learning_rate": 3.654936406444006e-05, "loss": 0.0306, "step": 48590 }, { "grad_norm": 0.23542632162570953, "learning_rate": 3.65294587474465e-05, "loss": 0.0254, "step": 48600 }, { "grad_norm": 0.25339555740356445, "learning_rate": 3.650955573220002e-05, "loss": 0.0602, "step": 48610 }, { "grad_norm": 0.41186144948005676, "learning_rate": 3.648965502210149e-05, "loss": 0.0251, "step": 48620 }, { "grad_norm": 0.3726568818092346, "learning_rate": 3.646975662055142e-05, "loss": 0.0269, "step": 48630 }, { "grad_norm": 0.3261260986328125, "learning_rate": 3.644986053094987e-05, "loss": 0.0237, "step": 48640 }, { "grad_norm": 0.4244026243686676, "learning_rate": 3.642996675669659e-05, "loss": 0.033, "step": 48650 }, { "grad_norm": 0.24892058968544006, "learning_rate": 3.641007530119083e-05, "loss": 0.0255, "step": 48660 }, { "grad_norm": 0.3211333155632019, "learning_rate": 3.639018616783153e-05, "loss": 0.0326, "step": 48670 }, { "grad_norm": 0.30337846279144287, "learning_rate": 3.637029936001719e-05, "loss": 0.0243, "step": 48680 }, { "grad_norm": 0.5095124244689941, "learning_rate": 3.6350414881145886e-05, "loss": 0.0302, "step": 48690 }, { "grad_norm": 0.3756624460220337, "learning_rate": 3.6330532734615386e-05, "loss": 0.0275, "step": 48700 }, { "grad_norm": 0.29709500074386597, "learning_rate": 3.6310652923822975e-05, "loss": 0.0332, "step": 48710 }, { "grad_norm": 0.3572749197483063, "learning_rate": 3.629077545216558e-05, "loss": 0.0285, "step": 48720 }, { "grad_norm": 0.34954050183296204, "learning_rate": 3.627090032303969e-05, "loss": 0.0338, "step": 48730 }, { "grad_norm": 0.22612638771533966, "learning_rate": 3.625102753984146e-05, "loss": 0.0343, "step": 48740 }, { "grad_norm": 0.31442973017692566, "learning_rate": 3.623115710596659e-05, "loss": 0.026, "step": 48750 }, { "grad_norm": 0.41363242268562317, "learning_rate": 3.6211289024810395e-05, "loss": 0.0319, "step": 48760 }, { "grad_norm": 0.1862097680568695, "learning_rate": 3.619142329976777e-05, "loss": 0.0234, "step": 48770 }, { "grad_norm": 0.2645684778690338, "learning_rate": 3.6171559934233247e-05, "loss": 0.0212, "step": 48780 }, { "grad_norm": 0.7760612368583679, "learning_rate": 3.615169893160093e-05, "loss": 0.0247, "step": 48790 }, { "grad_norm": 0.31857210397720337, "learning_rate": 3.61318402952645e-05, "loss": 0.028, "step": 48800 }, { "grad_norm": 0.3821542263031006, "learning_rate": 3.6111984028617285e-05, "loss": 0.0298, "step": 48810 }, { "grad_norm": 0.2986445426940918, "learning_rate": 3.6092130135052134e-05, "loss": 0.0275, "step": 48820 }, { "grad_norm": 0.3555185794830322, "learning_rate": 3.6072278617961584e-05, "loss": 0.0328, "step": 48830 }, { "grad_norm": 0.445713073015213, "learning_rate": 3.605242948073767e-05, "loss": 0.0327, "step": 48840 }, { "grad_norm": 0.29893696308135986, "learning_rate": 3.603258272677212e-05, "loss": 0.0232, "step": 48850 }, { "grad_norm": 0.2778802812099457, "learning_rate": 3.601273835945616e-05, "loss": 0.0331, "step": 48860 }, { "grad_norm": 0.31948214769363403, "learning_rate": 3.5992896382180664e-05, "loss": 0.0273, "step": 48870 }, { "grad_norm": 0.27572813630104065, "learning_rate": 3.597305679833609e-05, "loss": 0.0284, "step": 48880 }, { "grad_norm": 0.25365787744522095, "learning_rate": 3.595321961131245e-05, "loss": 0.027, "step": 48890 }, { "grad_norm": 0.3935295343399048, "learning_rate": 3.593338482449942e-05, "loss": 0.0358, "step": 48900 }, { "grad_norm": 0.345140278339386, "learning_rate": 3.591355244128618e-05, "loss": 0.0382, "step": 48910 }, { "grad_norm": 0.35479605197906494, "learning_rate": 3.589372246506158e-05, "loss": 0.0334, "step": 48920 }, { "grad_norm": 0.2389886975288391, "learning_rate": 3.5873894899213984e-05, "loss": 0.0247, "step": 48930 }, { "grad_norm": 0.40510091185569763, "learning_rate": 3.5854069747131416e-05, "loss": 0.028, "step": 48940 }, { "grad_norm": 0.38404253125190735, "learning_rate": 3.583424701220143e-05, "loss": 0.0331, "step": 48950 }, { "grad_norm": 0.299967497587204, "learning_rate": 3.581442669781121e-05, "loss": 0.0279, "step": 48960 }, { "grad_norm": 0.4345623552799225, "learning_rate": 3.579460880734749e-05, "loss": 0.0258, "step": 48970 }, { "grad_norm": 0.273252010345459, "learning_rate": 3.577479334419657e-05, "loss": 0.0255, "step": 48980 }, { "grad_norm": 0.32056093215942383, "learning_rate": 3.575498031174444e-05, "loss": 0.0257, "step": 48990 }, { "grad_norm": 0.33701300621032715, "learning_rate": 3.573516971337657e-05, "loss": 0.0276, "step": 49000 }, { "grad_norm": 0.4986162483692169, "learning_rate": 3.5715361552478046e-05, "loss": 0.0349, "step": 49010 }, { "grad_norm": 0.3549339175224304, "learning_rate": 3.5695555832433536e-05, "loss": 0.0252, "step": 49020 }, { "grad_norm": 0.3589768707752228, "learning_rate": 3.5675752556627325e-05, "loss": 0.0267, "step": 49030 }, { "grad_norm": 0.41881513595581055, "learning_rate": 3.565595172844322e-05, "loss": 0.0293, "step": 49040 }, { "grad_norm": 0.2876962423324585, "learning_rate": 3.5636153351264666e-05, "loss": 0.0294, "step": 49050 }, { "grad_norm": 0.31953728199005127, "learning_rate": 3.5616357428474655e-05, "loss": 0.0262, "step": 49060 }, { "grad_norm": 0.283311128616333, "learning_rate": 3.559656396345575e-05, "loss": 0.0291, "step": 49070 }, { "grad_norm": 0.23594075441360474, "learning_rate": 3.5576772959590146e-05, "loss": 0.0335, "step": 49080 }, { "grad_norm": 0.16277754306793213, "learning_rate": 3.5556984420259545e-05, "loss": 0.0213, "step": 49090 }, { "grad_norm": 0.23995932936668396, "learning_rate": 3.5537198348845305e-05, "loss": 0.0251, "step": 49100 }, { "grad_norm": 0.6183784008026123, "learning_rate": 3.551741474872831e-05, "loss": 0.0267, "step": 49110 }, { "grad_norm": 0.3761577010154724, "learning_rate": 3.549763362328903e-05, "loss": 0.033, "step": 49120 }, { "grad_norm": 0.31772541999816895, "learning_rate": 3.5477854975907515e-05, "loss": 0.027, "step": 49130 }, { "grad_norm": 0.2822343111038208, "learning_rate": 3.5458078809963416e-05, "loss": 0.0226, "step": 49140 }, { "grad_norm": 0.3037961721420288, "learning_rate": 3.543830512883594e-05, "loss": 0.0209, "step": 49150 }, { "grad_norm": 0.2763158082962036, "learning_rate": 3.5418533935903824e-05, "loss": 0.0238, "step": 49160 }, { "grad_norm": 0.5290076732635498, "learning_rate": 3.539876523454547e-05, "loss": 0.0254, "step": 49170 }, { "grad_norm": 0.26926249265670776, "learning_rate": 3.537899902813878e-05, "loss": 0.0203, "step": 49180 }, { "grad_norm": 0.25066766142845154, "learning_rate": 3.5359235320061293e-05, "loss": 0.0269, "step": 49190 }, { "grad_norm": 0.23783767223358154, "learning_rate": 3.533947411369003e-05, "loss": 0.0182, "step": 49200 }, { "grad_norm": 0.3175586760044098, "learning_rate": 3.53197154124017e-05, "loss": 0.0278, "step": 49210 }, { "grad_norm": 0.4865645170211792, "learning_rate": 3.52999592195725e-05, "loss": 0.0296, "step": 49220 }, { "grad_norm": 0.3146733343601227, "learning_rate": 3.5280205538578224e-05, "loss": 0.0254, "step": 49230 }, { "grad_norm": 0.4101314842700958, "learning_rate": 3.5260454372794236e-05, "loss": 0.0258, "step": 49240 }, { "grad_norm": 0.2625490427017212, "learning_rate": 3.524070572559545e-05, "loss": 0.0263, "step": 49250 }, { "grad_norm": 0.3616931140422821, "learning_rate": 3.5220959600356395e-05, "loss": 0.0257, "step": 49260 }, { "grad_norm": 0.3260040879249573, "learning_rate": 3.5201216000451145e-05, "loss": 0.0212, "step": 49270 }, { "grad_norm": 0.31991249322891235, "learning_rate": 3.5181474929253335e-05, "loss": 0.0291, "step": 49280 }, { "grad_norm": 0.3162200450897217, "learning_rate": 3.516173639013615e-05, "loss": 0.0272, "step": 49290 }, { "grad_norm": 0.22162793576717377, "learning_rate": 3.5142000386472406e-05, "loss": 0.025, "step": 49300 }, { "grad_norm": 0.4024030566215515, "learning_rate": 3.5122266921634427e-05, "loss": 0.0233, "step": 49310 }, { "grad_norm": 0.37364888191223145, "learning_rate": 3.510253599899413e-05, "loss": 0.0258, "step": 49320 }, { "grad_norm": 0.33314090967178345, "learning_rate": 3.5082807621922965e-05, "loss": 0.0194, "step": 49330 }, { "grad_norm": 0.4344021677970886, "learning_rate": 3.506308179379201e-05, "loss": 0.0305, "step": 49340 }, { "grad_norm": 0.20588696002960205, "learning_rate": 3.5043358517971844e-05, "loss": 0.0252, "step": 49350 }, { "grad_norm": 0.22794710099697113, "learning_rate": 3.502363779783264e-05, "loss": 0.025, "step": 49360 }, { "grad_norm": 0.2991485893726349, "learning_rate": 3.500391963674415e-05, "loss": 0.0218, "step": 49370 }, { "grad_norm": 0.2771572768688202, "learning_rate": 3.4984204038075615e-05, "loss": 0.0247, "step": 49380 }, { "grad_norm": 0.260506272315979, "learning_rate": 3.496449100519595e-05, "loss": 0.0273, "step": 49390 }, { "grad_norm": 0.24649791419506073, "learning_rate": 3.494478054147354e-05, "loss": 0.0252, "step": 49400 }, { "grad_norm": 0.36289408802986145, "learning_rate": 3.4925072650276395e-05, "loss": 0.0228, "step": 49410 }, { "grad_norm": 0.198319211602211, "learning_rate": 3.4905367334972016e-05, "loss": 0.0245, "step": 49420 }, { "grad_norm": 0.3551221489906311, "learning_rate": 3.488566459892752e-05, "loss": 0.023, "step": 49430 }, { "grad_norm": 0.18811491131782532, "learning_rate": 3.4865964445509585e-05, "loss": 0.0227, "step": 49440 }, { "grad_norm": 0.3681928515434265, "learning_rate": 3.484626687808438e-05, "loss": 0.0264, "step": 49450 }, { "grad_norm": 0.33895793557167053, "learning_rate": 3.4826571900017735e-05, "loss": 0.0221, "step": 49460 }, { "grad_norm": 0.30816778540611267, "learning_rate": 3.480687951467495e-05, "loss": 0.0217, "step": 49470 }, { "grad_norm": 0.25892359018325806, "learning_rate": 3.4787189725420925e-05, "loss": 0.0232, "step": 49480 }, { "grad_norm": 0.4810996651649475, "learning_rate": 3.4767502535620086e-05, "loss": 0.0245, "step": 49490 }, { "grad_norm": 0.24145042896270752, "learning_rate": 3.474781794863648e-05, "loss": 0.0365, "step": 49500 }, { "grad_norm": 0.2602117955684662, "learning_rate": 3.472813596783363e-05, "loss": 0.0308, "step": 49510 }, { "grad_norm": 0.3349647521972656, "learning_rate": 3.470845659657466e-05, "loss": 0.0287, "step": 49520 }, { "grad_norm": 0.34917911887168884, "learning_rate": 3.468877983822223e-05, "loss": 0.0255, "step": 49530 }, { "grad_norm": 0.3150339126586914, "learning_rate": 3.466910569613855e-05, "loss": 0.0211, "step": 49540 }, { "grad_norm": 0.24278654158115387, "learning_rate": 3.464943417368542e-05, "loss": 0.0275, "step": 49550 }, { "grad_norm": 0.3070736527442932, "learning_rate": 3.462976527422415e-05, "loss": 0.022, "step": 49560 }, { "grad_norm": 0.30681857466697693, "learning_rate": 3.461009900111562e-05, "loss": 0.0243, "step": 49570 }, { "grad_norm": 0.376520037651062, "learning_rate": 3.459043535772023e-05, "loss": 0.0271, "step": 49580 }, { "grad_norm": 0.3576035499572754, "learning_rate": 3.4570774347398014e-05, "loss": 0.0306, "step": 49590 }, { "grad_norm": 0.5138664245605469, "learning_rate": 3.4551115973508454e-05, "loss": 0.031, "step": 49600 }, { "grad_norm": 0.340304970741272, "learning_rate": 3.453146023941066e-05, "loss": 0.0284, "step": 49610 }, { "grad_norm": 0.30012384057044983, "learning_rate": 3.451180714846325e-05, "loss": 0.0279, "step": 49620 }, { "grad_norm": 0.3097527027130127, "learning_rate": 3.449215670402438e-05, "loss": 0.0279, "step": 49630 }, { "grad_norm": 0.36897972226142883, "learning_rate": 3.447250890945181e-05, "loss": 0.0285, "step": 49640 }, { "grad_norm": 0.434308797121048, "learning_rate": 3.4452863768102754e-05, "loss": 0.0221, "step": 49650 }, { "grad_norm": 0.3169851005077362, "learning_rate": 3.443322128333409e-05, "loss": 0.0276, "step": 49660 }, { "grad_norm": 0.4857611060142517, "learning_rate": 3.441358145850215e-05, "loss": 0.0272, "step": 49670 }, { "grad_norm": 0.2929283678531647, "learning_rate": 3.439394429696286e-05, "loss": 0.0264, "step": 49680 }, { "grad_norm": 0.4518865942955017, "learning_rate": 3.4374309802071644e-05, "loss": 0.0264, "step": 49690 }, { "grad_norm": 0.332719087600708, "learning_rate": 3.435467797718353e-05, "loss": 0.0289, "step": 49700 }, { "grad_norm": 0.3585015833377838, "learning_rate": 3.433504882565306e-05, "loss": 0.0235, "step": 49710 }, { "grad_norm": 0.3094712495803833, "learning_rate": 3.43154223508343e-05, "loss": 0.0356, "step": 49720 }, { "grad_norm": 0.1649327427148819, "learning_rate": 3.429579855608089e-05, "loss": 0.0237, "step": 49730 }, { "grad_norm": 0.2854905426502228, "learning_rate": 3.427617744474597e-05, "loss": 0.0224, "step": 49740 }, { "grad_norm": 0.22864983975887299, "learning_rate": 3.425655902018231e-05, "loss": 0.0257, "step": 49750 }, { "grad_norm": 0.26656049489974976, "learning_rate": 3.423694328574211e-05, "loss": 0.0213, "step": 49760 }, { "grad_norm": 0.4121078848838806, "learning_rate": 3.42173302447772e-05, "loss": 0.0238, "step": 49770 }, { "grad_norm": 0.4185843765735626, "learning_rate": 3.419771990063886e-05, "loss": 0.0216, "step": 49780 }, { "grad_norm": 0.2720124423503876, "learning_rate": 3.417811225667803e-05, "loss": 0.023, "step": 49790 }, { "grad_norm": 0.38392525911331177, "learning_rate": 3.415850731624508e-05, "loss": 0.0274, "step": 49800 }, { "grad_norm": 0.38339507579803467, "learning_rate": 3.4138905082689945e-05, "loss": 0.0277, "step": 49810 }, { "grad_norm": 0.28760889172554016, "learning_rate": 3.4119305559362145e-05, "loss": 0.0283, "step": 49820 }, { "grad_norm": 0.22358930110931396, "learning_rate": 3.4099708749610684e-05, "loss": 0.023, "step": 49830 }, { "grad_norm": 0.2896823287010193, "learning_rate": 3.408011465678413e-05, "loss": 0.0288, "step": 49840 }, { "grad_norm": 0.3969699442386627, "learning_rate": 3.406052328423055e-05, "loss": 0.0271, "step": 49850 }, { "grad_norm": 0.28768089413642883, "learning_rate": 3.4040934635297615e-05, "loss": 0.023, "step": 49860 }, { "grad_norm": 0.278839111328125, "learning_rate": 3.4021348713332466e-05, "loss": 0.0218, "step": 49870 }, { "grad_norm": 0.35419201850891113, "learning_rate": 3.4001765521681807e-05, "loss": 0.0213, "step": 49880 }, { "grad_norm": 0.3862415850162506, "learning_rate": 3.398218506369188e-05, "loss": 0.0346, "step": 49890 }, { "grad_norm": 0.24813127517700195, "learning_rate": 3.3962607342708404e-05, "loss": 0.0259, "step": 49900 }, { "grad_norm": 0.4187919795513153, "learning_rate": 3.394303236207673e-05, "loss": 0.0265, "step": 49910 }, { "grad_norm": 0.24565443396568298, "learning_rate": 3.392346012514166e-05, "loss": 0.0284, "step": 49920 }, { "grad_norm": 0.3413770794868469, "learning_rate": 3.390389063524757e-05, "loss": 0.0226, "step": 49930 }, { "grad_norm": 0.2971169054508209, "learning_rate": 3.3884323895738324e-05, "loss": 0.0221, "step": 49940 }, { "grad_norm": 0.4326755404472351, "learning_rate": 3.386475990995738e-05, "loss": 0.0336, "step": 49950 }, { "grad_norm": 0.33177387714385986, "learning_rate": 3.384519868124765e-05, "loss": 0.022, "step": 49960 }, { "grad_norm": 0.3745568096637726, "learning_rate": 3.3825640212951645e-05, "loss": 0.0201, "step": 49970 }, { "grad_norm": 0.2791585624217987, "learning_rate": 3.380608450841134e-05, "loss": 0.0292, "step": 49980 }, { "grad_norm": 0.3435535728931427, "learning_rate": 3.3786531570968305e-05, "loss": 0.0254, "step": 49990 }, { "grad_norm": 0.21959127485752106, "learning_rate": 3.3766981403963584e-05, "loss": 0.0271, "step": 50000 }, { "grad_norm": 0.25093546509742737, "learning_rate": 3.374743401073775e-05, "loss": 0.025, "step": 50010 }, { "grad_norm": 0.37400591373443604, "learning_rate": 3.372788939463095e-05, "loss": 0.0216, "step": 50020 }, { "grad_norm": 0.3301298916339874, "learning_rate": 3.370834755898281e-05, "loss": 0.0244, "step": 50030 }, { "grad_norm": 0.2826608419418335, "learning_rate": 3.3688808507132493e-05, "loss": 0.0271, "step": 50040 }, { "grad_norm": 0.33513709902763367, "learning_rate": 3.3669272242418685e-05, "loss": 0.0326, "step": 50050 }, { "grad_norm": 0.5148831009864807, "learning_rate": 3.364973876817961e-05, "loss": 0.0239, "step": 50060 }, { "grad_norm": 0.3292032480239868, "learning_rate": 3.363020808775299e-05, "loss": 0.0218, "step": 50070 }, { "grad_norm": 0.32644057273864746, "learning_rate": 3.361068020447611e-05, "loss": 0.0235, "step": 50080 }, { "grad_norm": 0.3030967712402344, "learning_rate": 3.3591155121685724e-05, "loss": 0.0208, "step": 50090 }, { "grad_norm": 0.4295433461666107, "learning_rate": 3.357163284271814e-05, "loss": 0.0263, "step": 50100 }, { "grad_norm": 0.26266199350357056, "learning_rate": 3.355211337090919e-05, "loss": 0.0168, "step": 50110 }, { "grad_norm": 0.1998760998249054, "learning_rate": 3.353259670959421e-05, "loss": 0.0246, "step": 50120 }, { "grad_norm": 0.23060263693332672, "learning_rate": 3.351308286210808e-05, "loss": 0.025, "step": 50130 }, { "grad_norm": 0.30335235595703125, "learning_rate": 3.3493571831785156e-05, "loss": 0.0261, "step": 50140 }, { "grad_norm": 0.267001211643219, "learning_rate": 3.347406362195936e-05, "loss": 0.0202, "step": 50150 }, { "grad_norm": 0.31448209285736084, "learning_rate": 3.345455823596411e-05, "loss": 0.0227, "step": 50160 }, { "grad_norm": 0.3294947147369385, "learning_rate": 3.3435055677132346e-05, "loss": 0.0202, "step": 50170 }, { "grad_norm": 0.3953411877155304, "learning_rate": 3.3415555948796505e-05, "loss": 0.0238, "step": 50180 }, { "grad_norm": 0.49433740973472595, "learning_rate": 3.3396059054288556e-05, "loss": 0.0342, "step": 50190 }, { "grad_norm": 0.23263001441955566, "learning_rate": 3.3376564996940015e-05, "loss": 0.0338, "step": 50200 }, { "grad_norm": 0.23162798583507538, "learning_rate": 3.3357073780081836e-05, "loss": 0.0266, "step": 50210 }, { "grad_norm": 0.2715785503387451, "learning_rate": 3.333758540704459e-05, "loss": 0.0271, "step": 50220 }, { "grad_norm": 0.2550732493400574, "learning_rate": 3.331809988115827e-05, "loss": 0.022, "step": 50230 }, { "grad_norm": 0.36645248532295227, "learning_rate": 3.329861720575244e-05, "loss": 0.0233, "step": 50240 }, { "grad_norm": 0.2483275979757309, "learning_rate": 3.3279137384156126e-05, "loss": 0.0209, "step": 50250 }, { "grad_norm": 0.355980783700943, "learning_rate": 3.3259660419697934e-05, "loss": 0.0339, "step": 50260 }, { "grad_norm": 0.20308317244052887, "learning_rate": 3.3240186315705926e-05, "loss": 0.0281, "step": 50270 }, { "grad_norm": 0.2679484784603119, "learning_rate": 3.322071507550769e-05, "loss": 0.0305, "step": 50280 }, { "grad_norm": 0.20234735310077667, "learning_rate": 3.320124670243033e-05, "loss": 0.0228, "step": 50290 }, { "grad_norm": 0.23163025081157684, "learning_rate": 3.318178119980045e-05, "loss": 0.027, "step": 50300 }, { "grad_norm": 0.35030749440193176, "learning_rate": 3.316231857094421e-05, "loss": 0.0259, "step": 50310 }, { "grad_norm": 0.3367919623851776, "learning_rate": 3.31428588191872e-05, "loss": 0.0299, "step": 50320 }, { "grad_norm": 0.4054240584373474, "learning_rate": 3.312340194785458e-05, "loss": 0.0282, "step": 50330 }, { "grad_norm": 0.42928779125213623, "learning_rate": 3.3103947960270975e-05, "loss": 0.0202, "step": 50340 }, { "grad_norm": 0.34202542901039124, "learning_rate": 3.308449685976058e-05, "loss": 0.0186, "step": 50350 }, { "grad_norm": 0.30640363693237305, "learning_rate": 3.3065048649647024e-05, "loss": 0.02, "step": 50360 }, { "grad_norm": 0.5441875457763672, "learning_rate": 3.304560333325348e-05, "loss": 0.0232, "step": 50370 }, { "grad_norm": 0.2938997149467468, "learning_rate": 3.3026160913902624e-05, "loss": 0.0295, "step": 50380 }, { "grad_norm": 0.4960513114929199, "learning_rate": 3.300672139491662e-05, "loss": 0.0325, "step": 50390 }, { "grad_norm": 0.34651944041252136, "learning_rate": 3.298728477961717e-05, "loss": 0.025, "step": 50400 }, { "grad_norm": 0.5061760544776917, "learning_rate": 3.296785107132544e-05, "loss": 0.0345, "step": 50410 }, { "grad_norm": 0.23326709866523743, "learning_rate": 3.2948420273362156e-05, "loss": 0.018, "step": 50420 }, { "grad_norm": 0.23118725419044495, "learning_rate": 3.292899238904747e-05, "loss": 0.0279, "step": 50430 }, { "grad_norm": 0.2389550656080246, "learning_rate": 3.29095674217011e-05, "loss": 0.0227, "step": 50440 }, { "grad_norm": 0.23699712753295898, "learning_rate": 3.289014537464224e-05, "loss": 0.0276, "step": 50450 }, { "grad_norm": 0.18176297843456268, "learning_rate": 3.287072625118955e-05, "loss": 0.0279, "step": 50460 }, { "grad_norm": 0.3962687849998474, "learning_rate": 3.285131005466129e-05, "loss": 0.0252, "step": 50470 }, { "grad_norm": 0.2693261206150055, "learning_rate": 3.2831896788375105e-05, "loss": 0.0255, "step": 50480 }, { "grad_norm": 0.18039552867412567, "learning_rate": 3.281248645564822e-05, "loss": 0.0224, "step": 50490 }, { "grad_norm": 0.16846193373203278, "learning_rate": 3.2793079059797306e-05, "loss": 0.0269, "step": 50500 }, { "grad_norm": 0.27198049426078796, "learning_rate": 3.277367460413859e-05, "loss": 0.0289, "step": 50510 }, { "grad_norm": 0.27695348858833313, "learning_rate": 3.275427309198773e-05, "loss": 0.0288, "step": 50520 }, { "grad_norm": 0.21892808377742767, "learning_rate": 3.273487452665993e-05, "loss": 0.0242, "step": 50530 }, { "grad_norm": 0.3050314784049988, "learning_rate": 3.271547891146986e-05, "loss": 0.0285, "step": 50540 }, { "grad_norm": 0.4413679838180542, "learning_rate": 3.269608624973173e-05, "loss": 0.0265, "step": 50550 }, { "grad_norm": 0.345356285572052, "learning_rate": 3.26766965447592e-05, "loss": 0.0295, "step": 50560 }, { "grad_norm": 0.30858656764030457, "learning_rate": 3.2657309799865424e-05, "loss": 0.03, "step": 50570 }, { "grad_norm": 0.2274487465620041, "learning_rate": 3.2637926018363084e-05, "loss": 0.0192, "step": 50580 }, { "grad_norm": 0.3890484571456909, "learning_rate": 3.261854520356432e-05, "loss": 0.026, "step": 50590 }, { "grad_norm": 0.26368284225463867, "learning_rate": 3.2599167358780816e-05, "loss": 0.0172, "step": 50600 }, { "grad_norm": 0.2600559890270233, "learning_rate": 3.2579792487323676e-05, "loss": 0.0192, "step": 50610 }, { "grad_norm": 0.24455949664115906, "learning_rate": 3.256042059250358e-05, "loss": 0.0273, "step": 50620 }, { "grad_norm": 0.4139123260974884, "learning_rate": 3.254105167763062e-05, "loss": 0.0254, "step": 50630 }, { "grad_norm": 0.2511120140552521, "learning_rate": 3.252168574601443e-05, "loss": 0.027, "step": 50640 }, { "grad_norm": 0.35864487290382385, "learning_rate": 3.2502322800964115e-05, "loss": 0.0243, "step": 50650 }, { "grad_norm": 0.2923842668533325, "learning_rate": 3.248296284578825e-05, "loss": 0.0256, "step": 50660 }, { "grad_norm": 0.3651517331600189, "learning_rate": 3.246360588379497e-05, "loss": 0.0273, "step": 50670 }, { "grad_norm": 0.3201633393764496, "learning_rate": 3.24442519182918e-05, "loss": 0.0324, "step": 50680 }, { "grad_norm": 0.37326523661613464, "learning_rate": 3.2424900952585845e-05, "loss": 0.0254, "step": 50690 }, { "grad_norm": 0.2943326234817505, "learning_rate": 3.240555298998362e-05, "loss": 0.0224, "step": 50700 }, { "grad_norm": 0.37276241183280945, "learning_rate": 3.2386208033791194e-05, "loss": 0.0261, "step": 50710 }, { "grad_norm": 0.2648792266845703, "learning_rate": 3.2366866087314064e-05, "loss": 0.0231, "step": 50720 }, { "grad_norm": 0.2290656417608261, "learning_rate": 3.234752715385727e-05, "loss": 0.0238, "step": 50730 }, { "grad_norm": 0.3871876299381256, "learning_rate": 3.232819123672529e-05, "loss": 0.0302, "step": 50740 }, { "grad_norm": 0.3669782280921936, "learning_rate": 3.2308858339222076e-05, "loss": 0.03, "step": 50750 }, { "grad_norm": 0.6942676901817322, "learning_rate": 3.2289528464651144e-05, "loss": 0.0301, "step": 50760 }, { "grad_norm": 0.20081032812595367, "learning_rate": 3.22702016163154e-05, "loss": 0.0259, "step": 50770 }, { "grad_norm": 0.2774629592895508, "learning_rate": 3.2250877797517295e-05, "loss": 0.0234, "step": 50780 }, { "grad_norm": 0.30354738235473633, "learning_rate": 3.223155701155872e-05, "loss": 0.0291, "step": 50790 }, { "grad_norm": 0.21463216841220856, "learning_rate": 3.2212239261741105e-05, "loss": 0.0314, "step": 50800 }, { "grad_norm": 0.19533725082874298, "learning_rate": 3.219292455136528e-05, "loss": 0.024, "step": 50810 }, { "grad_norm": 0.23532265424728394, "learning_rate": 3.217361288373165e-05, "loss": 0.0195, "step": 50820 }, { "grad_norm": 0.2371896654367447, "learning_rate": 3.215430426214002e-05, "loss": 0.0224, "step": 50830 }, { "grad_norm": 0.35506629943847656, "learning_rate": 3.2134998689889705e-05, "loss": 0.0224, "step": 50840 }, { "grad_norm": 0.40315911173820496, "learning_rate": 3.211569617027952e-05, "loss": 0.0262, "step": 50850 }, { "grad_norm": 0.2797197997570038, "learning_rate": 3.209639670660769e-05, "loss": 0.0301, "step": 50860 }, { "grad_norm": 0.5469263195991516, "learning_rate": 3.207710030217202e-05, "loss": 0.027, "step": 50870 }, { "grad_norm": 0.7107163667678833, "learning_rate": 3.205780696026972e-05, "loss": 0.0248, "step": 50880 }, { "grad_norm": 0.16454780101776123, "learning_rate": 3.203851668419749e-05, "loss": 0.0212, "step": 50890 }, { "grad_norm": 0.36681482195854187, "learning_rate": 3.201922947725149e-05, "loss": 0.0277, "step": 50900 }, { "grad_norm": 0.2280040681362152, "learning_rate": 3.199994534272742e-05, "loss": 0.0209, "step": 50910 }, { "grad_norm": 0.3131628632545471, "learning_rate": 3.1980664283920394e-05, "loss": 0.0283, "step": 50920 }, { "grad_norm": 0.3224947154521942, "learning_rate": 3.196138630412499e-05, "loss": 0.0211, "step": 50930 }, { "grad_norm": 0.29143068194389343, "learning_rate": 3.194211140663532e-05, "loss": 0.0197, "step": 50940 }, { "grad_norm": 0.33406540751457214, "learning_rate": 3.1922839594744914e-05, "loss": 0.0316, "step": 50950 }, { "grad_norm": 0.36044901609420776, "learning_rate": 3.190357087174683e-05, "loss": 0.0245, "step": 50960 }, { "grad_norm": 0.26192015409469604, "learning_rate": 3.188430524093353e-05, "loss": 0.0188, "step": 50970 }, { "grad_norm": 0.30859580636024475, "learning_rate": 3.1865042705597014e-05, "loss": 0.0248, "step": 50980 }, { "grad_norm": 0.3662097454071045, "learning_rate": 3.184578326902868e-05, "loss": 0.0229, "step": 50990 }, { "grad_norm": 0.19327598810195923, "learning_rate": 3.182652693451949e-05, "loss": 0.0231, "step": 51000 }, { "grad_norm": 0.40095254778862, "learning_rate": 3.180727370535981e-05, "loss": 0.026, "step": 51010 }, { "grad_norm": 0.2933151125907898, "learning_rate": 3.178802358483945e-05, "loss": 0.0266, "step": 51020 }, { "grad_norm": 0.30166634917259216, "learning_rate": 3.17687765762478e-05, "loss": 0.0281, "step": 51030 }, { "grad_norm": 0.20946720242500305, "learning_rate": 3.174953268287357e-05, "loss": 0.0236, "step": 51040 }, { "grad_norm": 0.3717834949493408, "learning_rate": 3.173029190800508e-05, "loss": 0.0275, "step": 51050 }, { "grad_norm": 0.30768951773643494, "learning_rate": 3.171105425493e-05, "loss": 0.0329, "step": 51060 }, { "grad_norm": 0.26485252380371094, "learning_rate": 3.169181972693557e-05, "loss": 0.0183, "step": 51070 }, { "grad_norm": 0.3020639419555664, "learning_rate": 3.1672588327308404e-05, "loss": 0.0247, "step": 51080 }, { "grad_norm": 0.3747830092906952, "learning_rate": 3.1653360059334646e-05, "loss": 0.0292, "step": 51090 }, { "grad_norm": 0.4342251420021057, "learning_rate": 3.163413492629985e-05, "loss": 0.0426, "step": 51100 }, { "grad_norm": 0.39700430631637573, "learning_rate": 3.161491293148912e-05, "loss": 0.0242, "step": 51110 }, { "grad_norm": 0.2944357693195343, "learning_rate": 3.1595694078186925e-05, "loss": 0.023, "step": 51120 }, { "grad_norm": 0.28931933641433716, "learning_rate": 3.157647836967724e-05, "loss": 0.0208, "step": 51130 }, { "grad_norm": 0.2667836844921112, "learning_rate": 3.155726580924354e-05, "loss": 0.0225, "step": 51140 }, { "grad_norm": 0.33883583545684814, "learning_rate": 3.1538056400168676e-05, "loss": 0.0306, "step": 51150 }, { "grad_norm": 0.24215860664844513, "learning_rate": 3.151885014573506e-05, "loss": 0.0285, "step": 51160 }, { "grad_norm": 0.3267059028148651, "learning_rate": 3.149964704922448e-05, "loss": 0.0323, "step": 51170 }, { "grad_norm": 0.2874744236469269, "learning_rate": 3.148044711391825e-05, "loss": 0.0228, "step": 51180 }, { "grad_norm": 0.35206228494644165, "learning_rate": 3.146125034309707e-05, "loss": 0.0226, "step": 51190 }, { "grad_norm": 0.299912691116333, "learning_rate": 3.1442056740041195e-05, "loss": 0.0285, "step": 51200 }, { "grad_norm": 0.30968862771987915, "learning_rate": 3.142286630803026e-05, "loss": 0.0351, "step": 51210 }, { "grad_norm": 0.306423544883728, "learning_rate": 3.140367905034337e-05, "loss": 0.0255, "step": 51220 }, { "grad_norm": 0.25317710638046265, "learning_rate": 3.138449497025914e-05, "loss": 0.0268, "step": 51230 }, { "grad_norm": 0.3399195671081543, "learning_rate": 3.136531407105557e-05, "loss": 0.0306, "step": 51240 }, { "grad_norm": 0.28050386905670166, "learning_rate": 3.1346136356010184e-05, "loss": 0.0287, "step": 51250 }, { "grad_norm": 0.3001210391521454, "learning_rate": 3.132696182839988e-05, "loss": 0.0231, "step": 51260 }, { "grad_norm": 0.28606095910072327, "learning_rate": 3.130779049150111e-05, "loss": 0.03, "step": 51270 }, { "grad_norm": 0.36911851167678833, "learning_rate": 3.128862234858971e-05, "loss": 0.0282, "step": 51280 }, { "grad_norm": 0.3191661834716797, "learning_rate": 3.1269457402941005e-05, "loss": 0.0191, "step": 51290 }, { "grad_norm": 0.29449957609176636, "learning_rate": 3.125029565782974e-05, "loss": 0.027, "step": 51300 }, { "grad_norm": 0.3265678882598877, "learning_rate": 3.123113711653013e-05, "loss": 0.0234, "step": 51310 }, { "grad_norm": 0.2619001269340515, "learning_rate": 3.121198178231587e-05, "loss": 0.0332, "step": 51320 }, { "grad_norm": 0.38458916544914246, "learning_rate": 3.119282965846006e-05, "loss": 0.0254, "step": 51330 }, { "grad_norm": 0.2823044955730438, "learning_rate": 3.1173680748235303e-05, "loss": 0.0266, "step": 51340 }, { "grad_norm": 0.3092055916786194, "learning_rate": 3.115453505491358e-05, "loss": 0.0258, "step": 51350 }, { "grad_norm": 0.3586612045764923, "learning_rate": 3.1135392581766404e-05, "loss": 0.0301, "step": 51360 }, { "grad_norm": 0.29508668184280396, "learning_rate": 3.111625333206467e-05, "loss": 0.0236, "step": 51370 }, { "grad_norm": 0.29654720425605774, "learning_rate": 3.109711730907881e-05, "loss": 0.017, "step": 51380 }, { "grad_norm": 0.32274237275123596, "learning_rate": 3.107798451607859e-05, "loss": 0.0255, "step": 51390 }, { "grad_norm": 0.3162539005279541, "learning_rate": 3.105885495633328e-05, "loss": 0.0235, "step": 51400 }, { "grad_norm": 0.8905379176139832, "learning_rate": 3.103972863311163e-05, "loss": 0.0247, "step": 51410 }, { "grad_norm": 0.25954434275627136, "learning_rate": 3.102060554968178e-05, "loss": 0.0234, "step": 51420 }, { "grad_norm": 0.21405765414237976, "learning_rate": 3.100148570931137e-05, "loss": 0.027, "step": 51430 }, { "grad_norm": 0.35789549350738525, "learning_rate": 3.098236911526744e-05, "loss": 0.0343, "step": 51440 }, { "grad_norm": 0.2885074317455292, "learning_rate": 3.0963255770816495e-05, "loss": 0.0205, "step": 51450 }, { "grad_norm": 0.3668120205402374, "learning_rate": 3.094414567922447e-05, "loss": 0.0259, "step": 51460 }, { "grad_norm": 0.3344574272632599, "learning_rate": 3.092503884375678e-05, "loss": 0.0302, "step": 51470 }, { "grad_norm": 0.3995610177516937, "learning_rate": 3.0905935267678255e-05, "loss": 0.0256, "step": 51480 }, { "grad_norm": 0.2986537218093872, "learning_rate": 3.0886834954253166e-05, "loss": 0.0292, "step": 51490 }, { "grad_norm": 0.40734538435935974, "learning_rate": 3.086773790674524e-05, "loss": 0.027, "step": 51500 }, { "grad_norm": 0.3239482641220093, "learning_rate": 3.0848644128417617e-05, "loss": 0.0246, "step": 51510 }, { "grad_norm": 0.27720770239830017, "learning_rate": 3.082955362253294e-05, "loss": 0.0278, "step": 51520 }, { "grad_norm": 0.38650941848754883, "learning_rate": 3.081046639235323e-05, "loss": 0.0272, "step": 51530 }, { "grad_norm": 0.45878833532333374, "learning_rate": 3.0791382441139986e-05, "loss": 0.0233, "step": 51540 }, { "grad_norm": 0.3345717191696167, "learning_rate": 3.077230177215411e-05, "loss": 0.0318, "step": 51550 }, { "grad_norm": 0.1762552708387375, "learning_rate": 3.0753224388656e-05, "loss": 0.0287, "step": 51560 }, { "grad_norm": 0.23643143475055695, "learning_rate": 3.073415029390544e-05, "loss": 0.028, "step": 51570 }, { "grad_norm": 0.365103155374527, "learning_rate": 3.071507949116166e-05, "loss": 0.028, "step": 51580 }, { "grad_norm": 0.45404475927352905, "learning_rate": 3.069601198368337e-05, "loss": 0.0194, "step": 51590 }, { "grad_norm": 0.3484460711479187, "learning_rate": 3.067694777472864e-05, "loss": 0.0204, "step": 51600 }, { "grad_norm": 0.3020403981208801, "learning_rate": 3.065788686755508e-05, "loss": 0.0251, "step": 51610 }, { "grad_norm": 0.21090389788150787, "learning_rate": 3.063882926541961e-05, "loss": 0.0217, "step": 51620 }, { "grad_norm": 0.41773509979248047, "learning_rate": 3.061977497157872e-05, "loss": 0.0257, "step": 51630 }, { "grad_norm": 0.3476683497428894, "learning_rate": 3.0600723989288224e-05, "loss": 0.0233, "step": 51640 }, { "grad_norm": 0.4818059802055359, "learning_rate": 3.058167632180343e-05, "loss": 0.0283, "step": 51650 }, { "grad_norm": 0.32794585824012756, "learning_rate": 3.0562631972379074e-05, "loss": 0.0238, "step": 51660 }, { "grad_norm": 0.28508028388023376, "learning_rate": 3.0543590944269276e-05, "loss": 0.0237, "step": 51670 }, { "grad_norm": 0.3558342158794403, "learning_rate": 3.052455324072766e-05, "loss": 0.0307, "step": 51680 }, { "grad_norm": 0.22332178056240082, "learning_rate": 3.0505518865007244e-05, "loss": 0.0273, "step": 51690 }, { "grad_norm": 0.35799238085746765, "learning_rate": 3.048648782036048e-05, "loss": 0.0222, "step": 51700 }, { "grad_norm": 0.3346271812915802, "learning_rate": 3.0467460110039236e-05, "loss": 0.0303, "step": 51710 }, { "grad_norm": 0.36253660917282104, "learning_rate": 3.0448435737294866e-05, "loss": 0.0269, "step": 51720 }, { "grad_norm": 0.28429579734802246, "learning_rate": 3.042941470537808e-05, "loss": 0.0215, "step": 51730 }, { "grad_norm": 0.25449061393737793, "learning_rate": 3.0410397017539072e-05, "loss": 0.0187, "step": 51740 }, { "grad_norm": 0.24428433179855347, "learning_rate": 3.039138267702742e-05, "loss": 0.021, "step": 51750 }, { "grad_norm": 0.3855448067188263, "learning_rate": 3.0372371687092193e-05, "loss": 0.0238, "step": 51760 }, { "grad_norm": 0.4271518588066101, "learning_rate": 3.0353364050981823e-05, "loss": 0.0221, "step": 51770 }, { "grad_norm": 0.25449642539024353, "learning_rate": 3.033435977194418e-05, "loss": 0.03, "step": 51780 }, { "grad_norm": 0.20450280606746674, "learning_rate": 3.0315358853226618e-05, "loss": 0.0245, "step": 51790 }, { "grad_norm": 0.30296626687049866, "learning_rate": 3.0296361298075844e-05, "loss": 0.0198, "step": 51800 }, { "grad_norm": 0.39622846245765686, "learning_rate": 3.027736710973803e-05, "loss": 0.0287, "step": 51810 }, { "grad_norm": 0.29608774185180664, "learning_rate": 3.025837629145875e-05, "loss": 0.0244, "step": 51820 }, { "grad_norm": 0.3185681402683258, "learning_rate": 3.0239388846483048e-05, "loss": 0.0207, "step": 51830 }, { "grad_norm": 0.2862210273742676, "learning_rate": 3.022040477805532e-05, "loss": 0.0263, "step": 51840 }, { "grad_norm": 0.33624890446662903, "learning_rate": 3.020142408941946e-05, "loss": 0.0231, "step": 51850 }, { "grad_norm": 0.5812607407569885, "learning_rate": 3.018244678381873e-05, "loss": 0.0283, "step": 51860 }, { "grad_norm": 0.3722003996372223, "learning_rate": 3.0163472864495812e-05, "loss": 0.0321, "step": 51870 }, { "grad_norm": 0.2220609188079834, "learning_rate": 3.014450233469287e-05, "loss": 0.0322, "step": 51880 }, { "grad_norm": 0.2603941857814789, "learning_rate": 3.0125535197651422e-05, "loss": 0.0333, "step": 51890 }, { "grad_norm": 0.3027763366699219, "learning_rate": 3.0106571456612448e-05, "loss": 0.024, "step": 51900 }, { "grad_norm": 0.35621267557144165, "learning_rate": 3.0087611114816305e-05, "loss": 0.037, "step": 51910 }, { "grad_norm": 0.3090595304965973, "learning_rate": 3.006865417550284e-05, "loss": 0.0205, "step": 51920 }, { "grad_norm": 0.3929535746574402, "learning_rate": 3.0049700641911242e-05, "loss": 0.0275, "step": 51930 }, { "grad_norm": 0.24850758910179138, "learning_rate": 3.0030750517280175e-05, "loss": 0.021, "step": 51940 }, { "grad_norm": 0.26654765009880066, "learning_rate": 3.0011803804847682e-05, "loss": 0.0271, "step": 51950 }, { "grad_norm": 0.39797526597976685, "learning_rate": 2.9992860507851224e-05, "loss": 0.0244, "step": 51960 }, { "grad_norm": 0.26643696427345276, "learning_rate": 2.9973920629527723e-05, "loss": 0.0197, "step": 51970 }, { "grad_norm": 0.2788718044757843, "learning_rate": 2.9954984173113453e-05, "loss": 0.0223, "step": 51980 }, { "grad_norm": 0.36844271421432495, "learning_rate": 2.993605114184418e-05, "loss": 0.0258, "step": 51990 }, { "grad_norm": 0.37844520807266235, "learning_rate": 2.9917121538955005e-05, "loss": 0.0341, "step": 52000 }, { "grad_norm": 0.3061673939228058, "learning_rate": 2.98981953676805e-05, "loss": 0.0309, "step": 52010 }, { "grad_norm": 0.2156819850206375, "learning_rate": 2.9879272631254594e-05, "loss": 0.0228, "step": 52020 }, { "grad_norm": 0.24019598960876465, "learning_rate": 2.986035333291073e-05, "loss": 0.0233, "step": 52030 }, { "grad_norm": 0.383996844291687, "learning_rate": 2.9841437475881646e-05, "loss": 0.0323, "step": 52040 }, { "grad_norm": 0.3431647717952728, "learning_rate": 2.9822525063399555e-05, "loss": 0.0264, "step": 52050 }, { "grad_norm": 0.40707576274871826, "learning_rate": 2.9803616098696087e-05, "loss": 0.0263, "step": 52060 }, { "grad_norm": 0.3042140603065491, "learning_rate": 2.9784710585002228e-05, "loss": 0.0256, "step": 52070 }, { "grad_norm": 0.36454686522483826, "learning_rate": 2.9765808525548467e-05, "loss": 0.019, "step": 52080 }, { "grad_norm": 0.3128316104412079, "learning_rate": 2.974690992356461e-05, "loss": 0.023, "step": 52090 }, { "grad_norm": 0.36517828702926636, "learning_rate": 2.9728014782279934e-05, "loss": 0.0281, "step": 52100 }, { "grad_norm": 0.2925448715686798, "learning_rate": 2.970912310492307e-05, "loss": 0.0233, "step": 52110 }, { "grad_norm": 0.3358093500137329, "learning_rate": 2.9690234894722134e-05, "loss": 0.0232, "step": 52120 }, { "grad_norm": 0.2687634825706482, "learning_rate": 2.9671350154904577e-05, "loss": 0.0318, "step": 52130 }, { "grad_norm": 0.4317832887172699, "learning_rate": 2.9652468888697282e-05, "loss": 0.03, "step": 52140 }, { "grad_norm": 0.30355700850486755, "learning_rate": 2.9633591099326562e-05, "loss": 0.0216, "step": 52150 }, { "grad_norm": 0.5006200075149536, "learning_rate": 2.9614716790018072e-05, "loss": 0.0216, "step": 52160 }, { "grad_norm": 0.27551934123039246, "learning_rate": 2.959584596399697e-05, "loss": 0.0193, "step": 52170 }, { "grad_norm": 0.4804236888885498, "learning_rate": 2.9576978624487717e-05, "loss": 0.0273, "step": 52180 }, { "grad_norm": 0.34847697615623474, "learning_rate": 2.9558114774714263e-05, "loss": 0.0257, "step": 52190 }, { "grad_norm": 0.28248462080955505, "learning_rate": 2.9539254417899897e-05, "loss": 0.0279, "step": 52200 }, { "grad_norm": 0.19296737015247345, "learning_rate": 2.9520397557267365e-05, "loss": 0.0192, "step": 52210 }, { "grad_norm": 0.28153732419013977, "learning_rate": 2.9501544196038765e-05, "loss": 0.0272, "step": 52220 }, { "grad_norm": 0.1586153656244278, "learning_rate": 2.9482694337435614e-05, "loss": 0.0285, "step": 52230 }, { "grad_norm": 0.4198175370693207, "learning_rate": 2.946384798467887e-05, "loss": 0.0248, "step": 52240 }, { "grad_norm": 0.27829137444496155, "learning_rate": 2.9445005140988824e-05, "loss": 0.0179, "step": 52250 }, { "grad_norm": 0.2644846737384796, "learning_rate": 2.942616580958524e-05, "loss": 0.0239, "step": 52260 }, { "grad_norm": 0.49999621510505676, "learning_rate": 2.9407329993687193e-05, "loss": 0.0248, "step": 52270 }, { "grad_norm": 0.36890918016433716, "learning_rate": 2.938849769651326e-05, "loss": 0.0226, "step": 52280 }, { "grad_norm": 0.40515363216400146, "learning_rate": 2.9369668921281345e-05, "loss": 0.0303, "step": 52290 }, { "grad_norm": 0.23839867115020752, "learning_rate": 2.9350843671208773e-05, "loss": 0.0221, "step": 52300 }, { "grad_norm": 0.2996829152107239, "learning_rate": 2.9332021949512245e-05, "loss": 0.0205, "step": 52310 }, { "grad_norm": 0.4422568380832672, "learning_rate": 2.9313203759407908e-05, "loss": 0.0236, "step": 52320 }, { "grad_norm": 0.3495320677757263, "learning_rate": 2.929438910411127e-05, "loss": 0.0331, "step": 52330 }, { "grad_norm": 0.4965597987174988, "learning_rate": 2.927557798683722e-05, "loss": 0.03, "step": 52340 }, { "grad_norm": 0.2518079876899719, "learning_rate": 2.9256770410800095e-05, "loss": 0.0178, "step": 52350 }, { "grad_norm": 0.2155594676733017, "learning_rate": 2.9237966379213554e-05, "loss": 0.0223, "step": 52360 }, { "grad_norm": 0.46421509981155396, "learning_rate": 2.9219165895290736e-05, "loss": 0.0237, "step": 52370 }, { "grad_norm": 0.30798912048339844, "learning_rate": 2.9200368962244083e-05, "loss": 0.0266, "step": 52380 }, { "grad_norm": 0.30625101923942566, "learning_rate": 2.9181575583285536e-05, "loss": 0.0228, "step": 52390 }, { "grad_norm": 0.24434007704257965, "learning_rate": 2.916278576162632e-05, "loss": 0.0228, "step": 52400 }, { "grad_norm": 0.4407804012298584, "learning_rate": 2.9143999500477116e-05, "loss": 0.0225, "step": 52410 }, { "grad_norm": 0.4842066168785095, "learning_rate": 2.9125216803048004e-05, "loss": 0.0289, "step": 52420 }, { "grad_norm": 0.40567222237586975, "learning_rate": 2.9106437672548403e-05, "loss": 0.0197, "step": 52430 }, { "grad_norm": 0.2153712511062622, "learning_rate": 2.9087662112187154e-05, "loss": 0.0252, "step": 52440 }, { "grad_norm": 0.2225436270236969, "learning_rate": 2.9068890125172498e-05, "loss": 0.0224, "step": 52450 }, { "grad_norm": 0.21249625086784363, "learning_rate": 2.9050121714712058e-05, "loss": 0.0339, "step": 52460 }, { "grad_norm": 0.400099515914917, "learning_rate": 2.9031356884012835e-05, "loss": 0.0259, "step": 52470 }, { "grad_norm": 0.3665972352027893, "learning_rate": 2.9012595636281247e-05, "loss": 0.0316, "step": 52480 }, { "grad_norm": 0.21159762144088745, "learning_rate": 2.8993837974723016e-05, "loss": 0.043, "step": 52490 }, { "grad_norm": 0.27634838223457336, "learning_rate": 2.8975083902543394e-05, "loss": 0.0278, "step": 52500 }, { "grad_norm": 0.3598772883415222, "learning_rate": 2.895633342294688e-05, "loss": 0.0256, "step": 52510 }, { "grad_norm": 0.2327413558959961, "learning_rate": 2.893758653913744e-05, "loss": 0.0251, "step": 52520 }, { "grad_norm": 0.24484652280807495, "learning_rate": 2.8918843254318423e-05, "loss": 0.0259, "step": 52530 }, { "grad_norm": 0.39972028136253357, "learning_rate": 2.8900103571692483e-05, "loss": 0.029, "step": 52540 }, { "grad_norm": 0.3089250326156616, "learning_rate": 2.8881367494461796e-05, "loss": 0.0184, "step": 52550 }, { "grad_norm": 0.2617608308792114, "learning_rate": 2.886263502582779e-05, "loss": 0.0299, "step": 52560 }, { "grad_norm": 0.3432762622833252, "learning_rate": 2.8843906168991353e-05, "loss": 0.0235, "step": 52570 }, { "grad_norm": 0.3372119665145874, "learning_rate": 2.8825180927152728e-05, "loss": 0.0252, "step": 52580 }, { "grad_norm": 0.3419502079486847, "learning_rate": 2.8806459303511547e-05, "loss": 0.0241, "step": 52590 }, { "grad_norm": 0.2779596447944641, "learning_rate": 2.8787741301266852e-05, "loss": 0.0197, "step": 52600 }, { "grad_norm": 0.39942219853401184, "learning_rate": 2.876902692361697e-05, "loss": 0.0218, "step": 52610 }, { "grad_norm": 0.3288646638393402, "learning_rate": 2.875031617375976e-05, "loss": 0.0233, "step": 52620 }, { "grad_norm": 0.17817586660385132, "learning_rate": 2.8731609054892318e-05, "loss": 0.0208, "step": 52630 }, { "grad_norm": 0.2645319700241089, "learning_rate": 2.871290557021119e-05, "loss": 0.0212, "step": 52640 }, { "grad_norm": 0.3114839494228363, "learning_rate": 2.8694205722912298e-05, "loss": 0.0256, "step": 52650 }, { "grad_norm": 0.24931707978248596, "learning_rate": 2.8675509516190936e-05, "loss": 0.029, "step": 52660 }, { "grad_norm": 0.2680509388446808, "learning_rate": 2.8656816953241772e-05, "loss": 0.0314, "step": 52670 }, { "grad_norm": 0.38285061717033386, "learning_rate": 2.863812803725887e-05, "loss": 0.0217, "step": 52680 }, { "grad_norm": 0.21170096099376678, "learning_rate": 2.8619442771435623e-05, "loss": 0.0189, "step": 52690 }, { "grad_norm": 0.37788650393486023, "learning_rate": 2.860076115896484e-05, "loss": 0.032, "step": 52700 }, { "grad_norm": 0.21235306560993195, "learning_rate": 2.8582083203038712e-05, "loss": 0.0225, "step": 52710 }, { "grad_norm": 0.24607722461223602, "learning_rate": 2.8563408906848778e-05, "loss": 0.0203, "step": 52720 }, { "grad_norm": 0.2873680591583252, "learning_rate": 2.8544738273585993e-05, "loss": 0.0212, "step": 52730 }, { "grad_norm": 0.36729034781455994, "learning_rate": 2.8526071306440595e-05, "loss": 0.0208, "step": 52740 }, { "grad_norm": 0.2918832004070282, "learning_rate": 2.850740800860233e-05, "loss": 0.0203, "step": 52750 }, { "grad_norm": 0.2400875836610794, "learning_rate": 2.8488748383260178e-05, "loss": 0.0194, "step": 52760 }, { "grad_norm": 0.30080196261405945, "learning_rate": 2.8470092433602635e-05, "loss": 0.0297, "step": 52770 }, { "grad_norm": 0.27125653624534607, "learning_rate": 2.845144016281745e-05, "loss": 0.0194, "step": 52780 }, { "grad_norm": 0.3035341203212738, "learning_rate": 2.8432791574091744e-05, "loss": 0.0373, "step": 52790 }, { "grad_norm": 0.29336094856262207, "learning_rate": 2.8414146670612134e-05, "loss": 0.0307, "step": 52800 }, { "grad_norm": 0.29192039370536804, "learning_rate": 2.8395505455564446e-05, "loss": 0.0308, "step": 52810 }, { "grad_norm": 0.21312473714351654, "learning_rate": 2.837686793213403e-05, "loss": 0.0245, "step": 52820 }, { "grad_norm": 0.3775694668292999, "learning_rate": 2.8358234103505477e-05, "loss": 0.0273, "step": 52830 }, { "grad_norm": 0.4189864695072174, "learning_rate": 2.8339603972862806e-05, "loss": 0.024, "step": 52840 }, { "grad_norm": 0.19062112271785736, "learning_rate": 2.83209775433894e-05, "loss": 0.0213, "step": 52850 }, { "grad_norm": 0.3356653153896332, "learning_rate": 2.8302354818268013e-05, "loss": 0.0288, "step": 52860 }, { "grad_norm": 0.3440225124359131, "learning_rate": 2.8283735800680754e-05, "loss": 0.0252, "step": 52870 }, { "grad_norm": 0.3143397569656372, "learning_rate": 2.826512049380913e-05, "loss": 0.0247, "step": 52880 }, { "grad_norm": 0.2694108486175537, "learning_rate": 2.8246508900833935e-05, "loss": 0.0252, "step": 52890 }, { "grad_norm": 0.24041803181171417, "learning_rate": 2.822790102493541e-05, "loss": 0.031, "step": 52900 }, { "grad_norm": 0.2482491135597229, "learning_rate": 2.820929686929314e-05, "loss": 0.0241, "step": 52910 }, { "grad_norm": 0.2784014940261841, "learning_rate": 2.819069643708605e-05, "loss": 0.022, "step": 52920 }, { "grad_norm": 0.3883322477340698, "learning_rate": 2.8172099731492484e-05, "loss": 0.0303, "step": 52930 }, { "grad_norm": 0.34418371319770813, "learning_rate": 2.815350675569003e-05, "loss": 0.0307, "step": 52940 }, { "grad_norm": 0.22146297991275787, "learning_rate": 2.8134917512855807e-05, "loss": 0.0312, "step": 52950 }, { "grad_norm": 0.23058006167411804, "learning_rate": 2.8116332006166134e-05, "loss": 0.0217, "step": 52960 }, { "grad_norm": 0.3124954402446747, "learning_rate": 2.809775023879685e-05, "loss": 0.0345, "step": 52970 }, { "grad_norm": 0.29354727268218994, "learning_rate": 2.807917221392299e-05, "loss": 0.027, "step": 52980 }, { "grad_norm": 0.4020622968673706, "learning_rate": 2.8060597934719067e-05, "loss": 0.0197, "step": 52990 }, { "grad_norm": 0.29411911964416504, "learning_rate": 2.8042027404358927e-05, "loss": 0.0251, "step": 53000 }, { "grad_norm": 0.3036443591117859, "learning_rate": 2.802346062601572e-05, "loss": 0.0249, "step": 53010 }, { "grad_norm": 0.4576999545097351, "learning_rate": 2.800489760286207e-05, "loss": 0.0354, "step": 53020 }, { "grad_norm": 0.32983455061912537, "learning_rate": 2.7986338338069834e-05, "loss": 0.0272, "step": 53030 }, { "grad_norm": 0.3469272553920746, "learning_rate": 2.7967782834810297e-05, "loss": 0.0316, "step": 53040 }, { "grad_norm": 0.2876230478286743, "learning_rate": 2.794923109625409e-05, "loss": 0.0223, "step": 53050 }, { "grad_norm": 0.32874596118927, "learning_rate": 2.7930683125571205e-05, "loss": 0.0234, "step": 53060 }, { "grad_norm": 0.289628267288208, "learning_rate": 2.7912138925930997e-05, "loss": 0.0257, "step": 53070 }, { "grad_norm": 0.29048579931259155, "learning_rate": 2.7893598500502117e-05, "loss": 0.039, "step": 53080 }, { "grad_norm": 0.3445075750350952, "learning_rate": 2.7875061852452644e-05, "loss": 0.028, "step": 53090 }, { "grad_norm": 0.4908199608325958, "learning_rate": 2.7856528984949982e-05, "loss": 0.0314, "step": 53100 }, { "grad_norm": 0.3453595042228699, "learning_rate": 2.7837999901160888e-05, "loss": 0.0243, "step": 53110 }, { "grad_norm": 0.2663866877555847, "learning_rate": 2.7819474604251484e-05, "loss": 0.0245, "step": 53120 }, { "grad_norm": 0.5190099477767944, "learning_rate": 2.780095309738725e-05, "loss": 0.0307, "step": 53130 }, { "grad_norm": 0.45961204171180725, "learning_rate": 2.778243538373294e-05, "loss": 0.0231, "step": 53140 }, { "grad_norm": 0.2953331470489502, "learning_rate": 2.7763921466452826e-05, "loss": 0.0245, "step": 53150 }, { "grad_norm": 0.33652982115745544, "learning_rate": 2.7745411348710336e-05, "loss": 0.0333, "step": 53160 }, { "grad_norm": 0.4361628293991089, "learning_rate": 2.7726905033668395e-05, "loss": 0.0326, "step": 53170 }, { "grad_norm": 0.29762375354766846, "learning_rate": 2.7708402524489214e-05, "loss": 0.0281, "step": 53180 }, { "grad_norm": 0.3487311005592346, "learning_rate": 2.7689903824334364e-05, "loss": 0.0248, "step": 53190 }, { "grad_norm": 0.2883744537830353, "learning_rate": 2.7671408936364785e-05, "loss": 0.0213, "step": 53200 }, { "grad_norm": 0.2355210930109024, "learning_rate": 2.765291786374069e-05, "loss": 0.0252, "step": 53210 }, { "grad_norm": 0.2813021242618561, "learning_rate": 2.7634430609621775e-05, "loss": 0.0224, "step": 53220 }, { "grad_norm": 0.3331087827682495, "learning_rate": 2.7615947177166956e-05, "loss": 0.0243, "step": 53230 }, { "grad_norm": 0.21233096718788147, "learning_rate": 2.7597467569534553e-05, "loss": 0.019, "step": 53240 }, { "grad_norm": 0.3462173044681549, "learning_rate": 2.757899178988226e-05, "loss": 0.023, "step": 53250 }, { "grad_norm": 0.31688061356544495, "learning_rate": 2.7560519841367005e-05, "loss": 0.0259, "step": 53260 }, { "grad_norm": 0.22880108654499054, "learning_rate": 2.7542051727145237e-05, "loss": 0.0193, "step": 53270 }, { "grad_norm": 0.3332504630088806, "learning_rate": 2.7523587450372578e-05, "loss": 0.0345, "step": 53280 }, { "grad_norm": 0.2575741112232208, "learning_rate": 2.750512701420409e-05, "loss": 0.026, "step": 53290 }, { "grad_norm": 0.24954985082149506, "learning_rate": 2.7486670421794158e-05, "loss": 0.0223, "step": 53300 }, { "grad_norm": 0.28849583864212036, "learning_rate": 2.7468217676296515e-05, "loss": 0.0362, "step": 53310 }, { "grad_norm": 0.4609951972961426, "learning_rate": 2.7449768780864226e-05, "loss": 0.0289, "step": 53320 }, { "grad_norm": 0.41714367270469666, "learning_rate": 2.7431323738649724e-05, "loss": 0.023, "step": 53330 }, { "grad_norm": 0.4740573465824127, "learning_rate": 2.7412882552804713e-05, "loss": 0.0297, "step": 53340 }, { "grad_norm": 0.3334474265575409, "learning_rate": 2.739444522648032e-05, "loss": 0.0238, "step": 53350 }, { "grad_norm": 0.24970175325870514, "learning_rate": 2.737601176282697e-05, "loss": 0.0284, "step": 53360 }, { "grad_norm": 0.34119901061058044, "learning_rate": 2.735758216499445e-05, "loss": 0.0227, "step": 53370 }, { "grad_norm": 0.26635605096817017, "learning_rate": 2.7339156436131864e-05, "loss": 0.0224, "step": 53380 }, { "grad_norm": 0.2787517309188843, "learning_rate": 2.7320734579387663e-05, "loss": 0.0239, "step": 53390 }, { "grad_norm": 0.22011294960975647, "learning_rate": 2.730231659790966e-05, "loss": 0.0201, "step": 53400 }, { "grad_norm": 0.34633898735046387, "learning_rate": 2.728390249484494e-05, "loss": 0.0323, "step": 53410 }, { "grad_norm": 0.2581776976585388, "learning_rate": 2.7265492273340032e-05, "loss": 0.0207, "step": 53420 }, { "grad_norm": 0.23166297376155853, "learning_rate": 2.7247085936540678e-05, "loss": 0.0235, "step": 53430 }, { "grad_norm": 0.3570344150066376, "learning_rate": 2.722868348759205e-05, "loss": 0.0269, "step": 53440 }, { "grad_norm": 0.22821763157844543, "learning_rate": 2.7210284929638635e-05, "loss": 0.0256, "step": 53450 }, { "grad_norm": 0.18973323702812195, "learning_rate": 2.7191890265824183e-05, "loss": 0.0222, "step": 53460 }, { "grad_norm": 0.389775812625885, "learning_rate": 2.7173499499291926e-05, "loss": 0.0185, "step": 53470 }, { "grad_norm": 0.4295373558998108, "learning_rate": 2.7155112633184277e-05, "loss": 0.0226, "step": 53480 }, { "grad_norm": 0.3414923846721649, "learning_rate": 2.713672967064307e-05, "loss": 0.0286, "step": 53490 }, { "grad_norm": 0.2171580046415329, "learning_rate": 2.711835061480945e-05, "loss": 0.0161, "step": 53500 }, { "grad_norm": 0.40684977173805237, "learning_rate": 2.7099975468823896e-05, "loss": 0.0311, "step": 53510 }, { "grad_norm": 0.35880419611930847, "learning_rate": 2.708160423582622e-05, "loss": 0.0224, "step": 53520 }, { "grad_norm": 0.46342384815216064, "learning_rate": 2.706323691895557e-05, "loss": 0.0309, "step": 53530 }, { "grad_norm": 0.37124186754226685, "learning_rate": 2.70448735213504e-05, "loss": 0.0249, "step": 53540 }, { "grad_norm": 0.41657572984695435, "learning_rate": 2.702651404614852e-05, "loss": 0.0249, "step": 53550 }, { "grad_norm": 0.25058919191360474, "learning_rate": 2.7008158496487056e-05, "loss": 0.02, "step": 53560 }, { "grad_norm": 0.32840797305107117, "learning_rate": 2.6989806875502487e-05, "loss": 0.0242, "step": 53570 }, { "grad_norm": 0.3313806653022766, "learning_rate": 2.6971459186330584e-05, "loss": 0.0205, "step": 53580 }, { "grad_norm": 0.2775733172893524, "learning_rate": 2.695311543210648e-05, "loss": 0.0236, "step": 53590 }, { "grad_norm": 0.3339666724205017, "learning_rate": 2.693477561596463e-05, "loss": 0.026, "step": 53600 }, { "grad_norm": 0.2791959345340729, "learning_rate": 2.6916439741038756e-05, "loss": 0.0242, "step": 53610 }, { "grad_norm": 0.32684096693992615, "learning_rate": 2.689810781046203e-05, "loss": 0.0262, "step": 53620 }, { "grad_norm": 0.268166720867157, "learning_rate": 2.6879779827366823e-05, "loss": 0.0216, "step": 53630 }, { "grad_norm": 0.3017682433128357, "learning_rate": 2.6861455794884904e-05, "loss": 0.0268, "step": 53640 }, { "grad_norm": 0.3631342649459839, "learning_rate": 2.6843135716147373e-05, "loss": 0.0238, "step": 53650 }, { "grad_norm": 0.37403249740600586, "learning_rate": 2.6824819594284556e-05, "loss": 0.0258, "step": 53660 }, { "grad_norm": 0.30155065655708313, "learning_rate": 2.6806507432426275e-05, "loss": 0.0267, "step": 53670 }, { "grad_norm": 0.28944095969200134, "learning_rate": 2.6788199233701512e-05, "loss": 0.0264, "step": 53680 }, { "grad_norm": 0.22934460639953613, "learning_rate": 2.6769895001238652e-05, "loss": 0.0285, "step": 53690 }, { "grad_norm": 0.540972113609314, "learning_rate": 2.67515947381654e-05, "loss": 0.0302, "step": 53700 }, { "grad_norm": 0.17839857935905457, "learning_rate": 2.6733298447608745e-05, "loss": 0.022, "step": 53710 }, { "grad_norm": 0.2728847861289978, "learning_rate": 2.6715006132695074e-05, "loss": 0.0274, "step": 53720 }, { "grad_norm": 0.30567529797554016, "learning_rate": 2.6696717796549985e-05, "loss": 0.034, "step": 53730 }, { "grad_norm": 0.3320259153842926, "learning_rate": 2.667843344229848e-05, "loss": 0.022, "step": 53740 }, { "grad_norm": 0.34681662917137146, "learning_rate": 2.6660153073064846e-05, "loss": 0.0285, "step": 53750 }, { "grad_norm": 0.23253397643566132, "learning_rate": 2.664187669197271e-05, "loss": 0.0333, "step": 53760 }, { "grad_norm": 0.43984708189964294, "learning_rate": 2.6623604302145005e-05, "loss": 0.0278, "step": 53770 }, { "grad_norm": 0.4159981310367584, "learning_rate": 2.6605335906703975e-05, "loss": 0.0267, "step": 53780 }, { "grad_norm": 0.3289508819580078, "learning_rate": 2.65870715087712e-05, "loss": 0.0297, "step": 53790 }, { "grad_norm": 0.3296215236186981, "learning_rate": 2.6568811111467573e-05, "loss": 0.0217, "step": 53800 }, { "grad_norm": 0.26820263266563416, "learning_rate": 2.6550554717913258e-05, "loss": 0.0259, "step": 53810 }, { "grad_norm": 0.27703067660331726, "learning_rate": 2.6532302331227805e-05, "loss": 0.024, "step": 53820 }, { "grad_norm": 0.43665507435798645, "learning_rate": 2.651405395453004e-05, "loss": 0.0218, "step": 53830 }, { "grad_norm": 0.30200597643852234, "learning_rate": 2.6495809590938115e-05, "loss": 0.0273, "step": 53840 }, { "grad_norm": 0.37804752588272095, "learning_rate": 2.64775692435695e-05, "loss": 0.0317, "step": 53850 }, { "grad_norm": 0.2303745150566101, "learning_rate": 2.6459332915540928e-05, "loss": 0.0229, "step": 53860 }, { "grad_norm": 0.272320955991745, "learning_rate": 2.644110060996856e-05, "loss": 0.0314, "step": 53870 }, { "grad_norm": 0.2430582195520401, "learning_rate": 2.642287232996774e-05, "loss": 0.0218, "step": 53880 }, { "grad_norm": 0.29738062620162964, "learning_rate": 2.6404648078653205e-05, "loss": 0.0156, "step": 53890 }, { "grad_norm": 0.23809902369976044, "learning_rate": 2.6386427859139002e-05, "loss": 0.0217, "step": 53900 }, { "grad_norm": 0.44520917534828186, "learning_rate": 2.63682116745384e-05, "loss": 0.0272, "step": 53910 }, { "grad_norm": 0.30659836530685425, "learning_rate": 2.6349999527964138e-05, "loss": 0.0195, "step": 53920 }, { "grad_norm": 0.3484768569469452, "learning_rate": 2.63317914225281e-05, "loss": 0.032, "step": 53930 }, { "grad_norm": 0.3173659145832062, "learning_rate": 2.6313587361341585e-05, "loss": 0.0329, "step": 53940 }, { "grad_norm": 0.3930099606513977, "learning_rate": 2.6295387347515165e-05, "loss": 0.0319, "step": 53950 }, { "grad_norm": 0.32893264293670654, "learning_rate": 2.6277191384158727e-05, "loss": 0.0324, "step": 53960 }, { "grad_norm": 0.30910536646842957, "learning_rate": 2.6258999474381457e-05, "loss": 0.0272, "step": 53970 }, { "grad_norm": 0.2135336697101593, "learning_rate": 2.624081162129186e-05, "loss": 0.0244, "step": 53980 }, { "grad_norm": 0.40014466643333435, "learning_rate": 2.6222627827997765e-05, "loss": 0.0397, "step": 53990 }, { "grad_norm": 0.3856484293937683, "learning_rate": 2.6204448097606236e-05, "loss": 0.0214, "step": 54000 }, { "grad_norm": 0.3281245827674866, "learning_rate": 2.6186272433223726e-05, "loss": 0.0269, "step": 54010 }, { "grad_norm": 0.18676826357841492, "learning_rate": 2.6168100837955943e-05, "loss": 0.0237, "step": 54020 }, { "grad_norm": 0.20927083492279053, "learning_rate": 2.6149933314907926e-05, "loss": 0.033, "step": 54030 }, { "grad_norm": 0.31708192825317383, "learning_rate": 2.6131769867184e-05, "loss": 0.0287, "step": 54040 }, { "grad_norm": 0.30046549439430237, "learning_rate": 2.611361049788783e-05, "loss": 0.0207, "step": 54050 }, { "grad_norm": 0.3099637031555176, "learning_rate": 2.6095455210122292e-05, "loss": 0.0209, "step": 54060 }, { "grad_norm": 0.37716495990753174, "learning_rate": 2.6077304006989712e-05, "loss": 0.0195, "step": 54070 }, { "grad_norm": 0.4090610444545746, "learning_rate": 2.6059156891591562e-05, "loss": 0.02, "step": 54080 }, { "grad_norm": 0.2612040936946869, "learning_rate": 2.6041013867028718e-05, "loss": 0.0207, "step": 54090 }, { "grad_norm": 0.27948135137557983, "learning_rate": 2.6022874936401347e-05, "loss": 0.0201, "step": 54100 }, { "grad_norm": 0.21305499970912933, "learning_rate": 2.6004740102808832e-05, "loss": 0.0236, "step": 54110 }, { "grad_norm": 0.2691386640071869, "learning_rate": 2.598660936935e-05, "loss": 0.0171, "step": 54120 }, { "grad_norm": 0.2883831858634949, "learning_rate": 2.5968482739122845e-05, "loss": 0.0246, "step": 54130 }, { "grad_norm": 0.32539039850234985, "learning_rate": 2.595036021522472e-05, "loss": 0.0231, "step": 54140 }, { "grad_norm": 0.2762129008769989, "learning_rate": 2.5932241800752278e-05, "loss": 0.026, "step": 54150 }, { "grad_norm": 0.36512938141822815, "learning_rate": 2.5914127498801453e-05, "loss": 0.0354, "step": 54160 }, { "grad_norm": 0.34465980529785156, "learning_rate": 2.5896017312467497e-05, "loss": 0.0228, "step": 54170 }, { "grad_norm": 0.3933189809322357, "learning_rate": 2.587791124484493e-05, "loss": 0.0217, "step": 54180 }, { "grad_norm": 0.21913717687129974, "learning_rate": 2.5859809299027615e-05, "loss": 0.0234, "step": 54190 }, { "grad_norm": 0.3152351677417755, "learning_rate": 2.5841711478108632e-05, "loss": 0.0309, "step": 54200 }, { "grad_norm": 0.4453723132610321, "learning_rate": 2.582361778518043e-05, "loss": 0.0251, "step": 54210 }, { "grad_norm": 0.294483482837677, "learning_rate": 2.580552822333472e-05, "loss": 0.0255, "step": 54220 }, { "grad_norm": 0.3024880588054657, "learning_rate": 2.578744279566252e-05, "loss": 0.0275, "step": 54230 }, { "grad_norm": 0.30783510208129883, "learning_rate": 2.576936150525413e-05, "loss": 0.0317, "step": 54240 }, { "grad_norm": 0.4839842915534973, "learning_rate": 2.5751284355199168e-05, "loss": 0.0211, "step": 54250 }, { "grad_norm": 0.3042627274990082, "learning_rate": 2.573321134858646e-05, "loss": 0.0202, "step": 54260 }, { "grad_norm": 0.3438180983066559, "learning_rate": 2.5715142488504286e-05, "loss": 0.0202, "step": 54270 }, { "grad_norm": 0.44046151638031006, "learning_rate": 2.5697077778040042e-05, "loss": 0.0245, "step": 54280 }, { "grad_norm": 0.3039562404155731, "learning_rate": 2.5679017220280522e-05, "loss": 0.0385, "step": 54290 }, { "grad_norm": 0.2988620400428772, "learning_rate": 2.5660960818311796e-05, "loss": 0.0233, "step": 54300 }, { "grad_norm": 0.2688893973827362, "learning_rate": 2.564290857521915e-05, "loss": 0.0249, "step": 54310 }, { "grad_norm": 0.22328446805477142, "learning_rate": 2.5624860494087298e-05, "loss": 0.0223, "step": 54320 }, { "grad_norm": 0.31782275438308716, "learning_rate": 2.5606816578000115e-05, "loss": 0.0331, "step": 54330 }, { "grad_norm": 0.3276013433933258, "learning_rate": 2.558877683004082e-05, "loss": 0.0257, "step": 54340 }, { "grad_norm": 0.5056112408638, "learning_rate": 2.557074125329192e-05, "loss": 0.0217, "step": 54350 }, { "grad_norm": 0.3428358733654022, "learning_rate": 2.5552709850835195e-05, "loss": 0.0201, "step": 54360 }, { "grad_norm": 0.29870229959487915, "learning_rate": 2.5534682625751738e-05, "loss": 0.0187, "step": 54370 }, { "grad_norm": 0.3860834836959839, "learning_rate": 2.551665958112186e-05, "loss": 0.0299, "step": 54380 }, { "grad_norm": 0.20898230373859406, "learning_rate": 2.549864072002527e-05, "loss": 0.0187, "step": 54390 }, { "grad_norm": 0.3990287482738495, "learning_rate": 2.5480626045540858e-05, "loss": 0.025, "step": 54400 }, { "grad_norm": 0.2577800452709198, "learning_rate": 2.546261556074684e-05, "loss": 0.0234, "step": 54410 }, { "grad_norm": 0.27131187915802, "learning_rate": 2.5444609268720726e-05, "loss": 0.0199, "step": 54420 }, { "grad_norm": 0.16975881159305573, "learning_rate": 2.5426607172539297e-05, "loss": 0.0205, "step": 54430 }, { "grad_norm": 0.22171719372272491, "learning_rate": 2.5408609275278617e-05, "loss": 0.0324, "step": 54440 }, { "grad_norm": 0.19715794920921326, "learning_rate": 2.5390615580014055e-05, "loss": 0.0248, "step": 54450 }, { "grad_norm": 0.4098506271839142, "learning_rate": 2.5372626089820207e-05, "loss": 0.0316, "step": 54460 }, { "grad_norm": 0.4589376449584961, "learning_rate": 2.5354640807770997e-05, "loss": 0.0199, "step": 54470 }, { "grad_norm": 0.2542993426322937, "learning_rate": 2.5336659736939622e-05, "loss": 0.0238, "step": 54480 }, { "grad_norm": 0.23600585758686066, "learning_rate": 2.5318682880398554e-05, "loss": 0.0274, "step": 54490 }, { "grad_norm": 0.35270097851753235, "learning_rate": 2.530071024121956e-05, "loss": 0.0312, "step": 54500 }, { "grad_norm": 0.4033114016056061, "learning_rate": 2.5282741822473627e-05, "loss": 0.021, "step": 54510 }, { "grad_norm": 0.346083402633667, "learning_rate": 2.526477762723114e-05, "loss": 0.0254, "step": 54520 }, { "grad_norm": 0.3133794665336609, "learning_rate": 2.5246817658561618e-05, "loss": 0.0254, "step": 54530 }, { "grad_norm": 0.16599059104919434, "learning_rate": 2.5228861919533965e-05, "loss": 0.0252, "step": 54540 }, { "grad_norm": 0.3362375497817993, "learning_rate": 2.5210910413216326e-05, "loss": 0.0385, "step": 54550 }, { "grad_norm": 0.28486645221710205, "learning_rate": 2.5192963142676086e-05, "loss": 0.0151, "step": 54560 }, { "grad_norm": 0.3358989655971527, "learning_rate": 2.517502011098001e-05, "loss": 0.0184, "step": 54570 }, { "grad_norm": 0.20360451936721802, "learning_rate": 2.5157081321193987e-05, "loss": 0.0218, "step": 54580 }, { "grad_norm": 0.18593434989452362, "learning_rate": 2.5139146776383356e-05, "loss": 0.0214, "step": 54590 }, { "grad_norm": 0.256961852312088, "learning_rate": 2.5121216479612575e-05, "loss": 0.0205, "step": 54600 }, { "grad_norm": 0.3703812062740326, "learning_rate": 2.510329043394546e-05, "loss": 0.0287, "step": 54610 }, { "grad_norm": 0.26919499039649963, "learning_rate": 2.508536864244508e-05, "loss": 0.0283, "step": 54620 }, { "grad_norm": 0.294779509305954, "learning_rate": 2.5067451108173778e-05, "loss": 0.0216, "step": 54630 }, { "grad_norm": 0.3101426661014557, "learning_rate": 2.5049537834193204e-05, "loss": 0.0236, "step": 54640 }, { "grad_norm": 0.17306870222091675, "learning_rate": 2.5031628823564194e-05, "loss": 0.0219, "step": 54650 }, { "grad_norm": 0.47087833285331726, "learning_rate": 2.5013724079346933e-05, "loss": 0.0237, "step": 54660 }, { "grad_norm": 0.36338382959365845, "learning_rate": 2.4995823604600854e-05, "loss": 0.021, "step": 54670 }, { "grad_norm": 0.24506855010986328, "learning_rate": 2.497792740238465e-05, "loss": 0.0228, "step": 54680 }, { "grad_norm": 0.397160142660141, "learning_rate": 2.49600354757563e-05, "loss": 0.0318, "step": 54690 }, { "grad_norm": 0.32242727279663086, "learning_rate": 2.494214782777306e-05, "loss": 0.0211, "step": 54700 }, { "grad_norm": 0.4954585134983063, "learning_rate": 2.4924264461491386e-05, "loss": 0.0317, "step": 54710 }, { "grad_norm": 0.38180816173553467, "learning_rate": 2.4906385379967133e-05, "loss": 0.0255, "step": 54720 }, { "grad_norm": 0.33700069785118103, "learning_rate": 2.4888510586255285e-05, "loss": 0.023, "step": 54730 }, { "grad_norm": 0.22100889682769775, "learning_rate": 2.487064008341018e-05, "loss": 0.0277, "step": 54740 }, { "grad_norm": 0.22929534316062927, "learning_rate": 2.4852773874485407e-05, "loss": 0.0215, "step": 54750 }, { "grad_norm": 0.4414694309234619, "learning_rate": 2.483491196253377e-05, "loss": 0.0235, "step": 54760 }, { "grad_norm": 0.3750428855419159, "learning_rate": 2.4817054350607443e-05, "loss": 0.0328, "step": 54770 }, { "grad_norm": 0.358952134847641, "learning_rate": 2.4799201041757742e-05, "loss": 0.0207, "step": 54780 }, { "grad_norm": 0.3494131863117218, "learning_rate": 2.4781352039035373e-05, "loss": 0.026, "step": 54790 }, { "grad_norm": 0.5089514255523682, "learning_rate": 2.4763507345490194e-05, "loss": 0.0186, "step": 54800 }, { "grad_norm": 0.3743666112422943, "learning_rate": 2.4745666964171386e-05, "loss": 0.0265, "step": 54810 }, { "grad_norm": 0.3816669285297394, "learning_rate": 2.4727830898127387e-05, "loss": 0.0251, "step": 54820 }, { "grad_norm": 0.2786911725997925, "learning_rate": 2.4709999150405895e-05, "loss": 0.0328, "step": 54830 }, { "grad_norm": 0.5176382064819336, "learning_rate": 2.469217172405388e-05, "loss": 0.0287, "step": 54840 }, { "grad_norm": 0.285060316324234, "learning_rate": 2.4674348622117527e-05, "loss": 0.019, "step": 54850 }, { "grad_norm": 0.24023178219795227, "learning_rate": 2.465652984764234e-05, "loss": 0.0179, "step": 54860 }, { "grad_norm": 0.21297460794448853, "learning_rate": 2.4638715403673056e-05, "loss": 0.0226, "step": 54870 }, { "grad_norm": 0.3817112445831299, "learning_rate": 2.4620905293253676e-05, "loss": 0.0256, "step": 54880 }, { "grad_norm": 0.29228174686431885, "learning_rate": 2.4603099519427463e-05, "loss": 0.0298, "step": 54890 }, { "grad_norm": 0.3097103536128998, "learning_rate": 2.458529808523695e-05, "loss": 0.0207, "step": 54900 }, { "grad_norm": 0.2594025731086731, "learning_rate": 2.456750099372387e-05, "loss": 0.017, "step": 54910 }, { "grad_norm": 0.29824623465538025, "learning_rate": 2.4549708247929327e-05, "loss": 0.0356, "step": 54920 }, { "grad_norm": 0.3012634515762329, "learning_rate": 2.4531919850893554e-05, "loss": 0.019, "step": 54930 }, { "grad_norm": 0.29338952898979187, "learning_rate": 2.4514135805656125e-05, "loss": 0.0195, "step": 54940 }, { "grad_norm": 0.3642998933792114, "learning_rate": 2.449635611525587e-05, "loss": 0.0287, "step": 54950 }, { "grad_norm": 0.24189959466457367, "learning_rate": 2.447858078273079e-05, "loss": 0.0276, "step": 54960 }, { "grad_norm": 0.35199764370918274, "learning_rate": 2.4460809811118275e-05, "loss": 0.0359, "step": 54970 }, { "grad_norm": 0.32756784558296204, "learning_rate": 2.444304320345483e-05, "loss": 0.0237, "step": 54980 }, { "grad_norm": 0.3508939743041992, "learning_rate": 2.4425280962776354e-05, "loss": 0.0239, "step": 54990 }, { "grad_norm": 0.31909939646720886, "learning_rate": 2.4407523092117875e-05, "loss": 0.0284, "step": 55000 }, { "grad_norm": 0.273524671792984, "learning_rate": 2.438976959451374e-05, "loss": 0.0222, "step": 55010 }, { "grad_norm": 0.3265354037284851, "learning_rate": 2.4372020472997565e-05, "loss": 0.0219, "step": 55020 }, { "grad_norm": 0.36307594180107117, "learning_rate": 2.4354275730602122e-05, "loss": 0.0221, "step": 55030 }, { "grad_norm": 0.2974824607372284, "learning_rate": 2.433653537035958e-05, "loss": 0.0223, "step": 55040 }, { "grad_norm": 0.283564031124115, "learning_rate": 2.431879939530123e-05, "loss": 0.0225, "step": 55050 }, { "grad_norm": 0.35972797870635986, "learning_rate": 2.4301067808457684e-05, "loss": 0.0219, "step": 55060 }, { "grad_norm": 0.22108906507492065, "learning_rate": 2.428334061285878e-05, "loss": 0.0234, "step": 55070 }, { "grad_norm": 0.3239505887031555, "learning_rate": 2.426561781153361e-05, "loss": 0.0237, "step": 55080 }, { "grad_norm": 0.2863493263721466, "learning_rate": 2.424789940751052e-05, "loss": 0.0221, "step": 55090 }, { "grad_norm": 0.23387333750724792, "learning_rate": 2.423018540381712e-05, "loss": 0.0273, "step": 55100 }, { "grad_norm": 0.15616841614246368, "learning_rate": 2.421247580348021e-05, "loss": 0.0216, "step": 55110 }, { "grad_norm": 0.3499460518360138, "learning_rate": 2.4194770609525897e-05, "loss": 0.0214, "step": 55120 }, { "grad_norm": 0.46779197454452515, "learning_rate": 2.417706982497951e-05, "loss": 0.0264, "step": 55130 }, { "grad_norm": 0.3366173803806305, "learning_rate": 2.415937345286563e-05, "loss": 0.0274, "step": 55140 }, { "grad_norm": 0.3609934151172638, "learning_rate": 2.4141681496208087e-05, "loss": 0.0259, "step": 55150 }, { "grad_norm": 0.4127473831176758, "learning_rate": 2.4123993958029946e-05, "loss": 0.0232, "step": 55160 }, { "grad_norm": 0.41559967398643494, "learning_rate": 2.4106310841353548e-05, "loss": 0.0191, "step": 55170 }, { "grad_norm": 0.3666224479675293, "learning_rate": 2.4088632149200398e-05, "loss": 0.0324, "step": 55180 }, { "grad_norm": 0.26800045371055603, "learning_rate": 2.4070957884591367e-05, "loss": 0.024, "step": 55190 }, { "grad_norm": 0.44207650423049927, "learning_rate": 2.4053288050546464e-05, "loss": 0.0293, "step": 55200 }, { "grad_norm": 0.3058120310306549, "learning_rate": 2.403562265008498e-05, "loss": 0.0239, "step": 55210 }, { "grad_norm": 0.30617186427116394, "learning_rate": 2.4017961686225483e-05, "loss": 0.0278, "step": 55220 }, { "grad_norm": 0.2782208025455475, "learning_rate": 2.400030516198568e-05, "loss": 0.0303, "step": 55230 }, { "grad_norm": 0.3713842034339905, "learning_rate": 2.3982653080382673e-05, "loss": 0.0273, "step": 55240 }, { "grad_norm": 0.26162266731262207, "learning_rate": 2.396500544443266e-05, "loss": 0.0237, "step": 55250 }, { "grad_norm": 0.35499778389930725, "learning_rate": 2.3947362257151156e-05, "loss": 0.0181, "step": 55260 }, { "grad_norm": 0.4189026951789856, "learning_rate": 2.39297235215529e-05, "loss": 0.0293, "step": 55270 }, { "grad_norm": 0.4208619296550751, "learning_rate": 2.3912089240651873e-05, "loss": 0.0275, "step": 55280 }, { "grad_norm": 0.1841774582862854, "learning_rate": 2.389445941746129e-05, "loss": 0.0285, "step": 55290 }, { "grad_norm": 0.4057818055152893, "learning_rate": 2.3876834054993625e-05, "loss": 0.0226, "step": 55300 }, { "grad_norm": 0.2016608715057373, "learning_rate": 2.3859213156260522e-05, "loss": 0.0201, "step": 55310 }, { "grad_norm": 0.29518619179725647, "learning_rate": 2.3841596724272948e-05, "loss": 0.0284, "step": 55320 }, { "grad_norm": 0.21220546960830688, "learning_rate": 2.382398476204106e-05, "loss": 0.0233, "step": 55330 }, { "grad_norm": 0.307155966758728, "learning_rate": 2.3806377272574254e-05, "loss": 0.0205, "step": 55340 }, { "grad_norm": 0.23078207671642303, "learning_rate": 2.3788774258881174e-05, "loss": 0.0196, "step": 55350 }, { "grad_norm": 0.3346405625343323, "learning_rate": 2.37711757239697e-05, "loss": 0.0224, "step": 55360 }, { "grad_norm": 0.35565048456192017, "learning_rate": 2.3753581670846954e-05, "loss": 0.0221, "step": 55370 }, { "grad_norm": 0.2055657058954239, "learning_rate": 2.3735992102519216e-05, "loss": 0.0261, "step": 55380 }, { "grad_norm": 0.36557552218437195, "learning_rate": 2.371840702199215e-05, "loss": 0.0278, "step": 55390 }, { "grad_norm": 0.3258399963378906, "learning_rate": 2.3700826432270494e-05, "loss": 0.0234, "step": 55400 }, { "grad_norm": 0.3447779715061188, "learning_rate": 2.3683250336358326e-05, "loss": 0.0305, "step": 55410 }, { "grad_norm": 0.31462520360946655, "learning_rate": 2.3665678737258923e-05, "loss": 0.0248, "step": 55420 }, { "grad_norm": 0.17797422409057617, "learning_rate": 2.3648111637974745e-05, "loss": 0.019, "step": 55430 }, { "grad_norm": 0.30751192569732666, "learning_rate": 2.36305490415076e-05, "loss": 0.0278, "step": 55440 }, { "grad_norm": 0.4164366126060486, "learning_rate": 2.3612990950858392e-05, "loss": 0.0229, "step": 55450 }, { "grad_norm": 0.2966763377189636, "learning_rate": 2.359543736902735e-05, "loss": 0.0195, "step": 55460 }, { "grad_norm": 0.4340825080871582, "learning_rate": 2.3577888299013896e-05, "loss": 0.0238, "step": 55470 }, { "grad_norm": 0.42433425784111023, "learning_rate": 2.3560343743816683e-05, "loss": 0.0221, "step": 55480 }, { "grad_norm": 0.44824567437171936, "learning_rate": 2.354280370643362e-05, "loss": 0.0272, "step": 55490 }, { "grad_norm": 0.24854205548763275, "learning_rate": 2.3525268189861777e-05, "loss": 0.0193, "step": 55500 }, { "grad_norm": 0.4631250202655792, "learning_rate": 2.3507737197097513e-05, "loss": 0.0239, "step": 55510 }, { "grad_norm": 0.41668766736984253, "learning_rate": 2.34902107311364e-05, "loss": 0.0242, "step": 55520 }, { "grad_norm": 0.3691382110118866, "learning_rate": 2.3472688794973225e-05, "loss": 0.0238, "step": 55530 }, { "grad_norm": 0.29261845350265503, "learning_rate": 2.3455171391602016e-05, "loss": 0.0249, "step": 55540 }, { "grad_norm": 0.2568424344062805, "learning_rate": 2.343765852401601e-05, "loss": 0.0195, "step": 55550 }, { "grad_norm": 0.24563376605510712, "learning_rate": 2.342015019520768e-05, "loss": 0.0278, "step": 55560 }, { "grad_norm": 0.2599451243877411, "learning_rate": 2.3402646408168742e-05, "loss": 0.0182, "step": 55570 }, { "grad_norm": 0.37964266538619995, "learning_rate": 2.3385147165890074e-05, "loss": 0.0283, "step": 55580 }, { "grad_norm": 0.37746065855026245, "learning_rate": 2.336765247136184e-05, "loss": 0.023, "step": 55590 }, { "grad_norm": 0.4569949209690094, "learning_rate": 2.33501623275734e-05, "loss": 0.0283, "step": 55600 }, { "grad_norm": 0.4287833571434021, "learning_rate": 2.333267673751334e-05, "loss": 0.0234, "step": 55610 }, { "grad_norm": 0.36946552991867065, "learning_rate": 2.33151957041695e-05, "loss": 0.0205, "step": 55620 }, { "grad_norm": 0.35988250374794006, "learning_rate": 2.329771923052884e-05, "loss": 0.0248, "step": 55630 }, { "grad_norm": 0.31940457224845886, "learning_rate": 2.3280247319577697e-05, "loss": 0.0198, "step": 55640 }, { "grad_norm": 0.29069966077804565, "learning_rate": 2.3262779974301473e-05, "loss": 0.0253, "step": 55650 }, { "grad_norm": 0.3263130486011505, "learning_rate": 2.3245317197684895e-05, "loss": 0.0206, "step": 55660 }, { "grad_norm": 0.5518240332603455, "learning_rate": 2.322785899271188e-05, "loss": 0.0232, "step": 55670 }, { "grad_norm": 0.25666457414627075, "learning_rate": 2.3210405362365507e-05, "loss": 0.0321, "step": 55680 }, { "grad_norm": 0.3782290518283844, "learning_rate": 2.31929563096282e-05, "loss": 0.0207, "step": 55690 }, { "grad_norm": 0.29506781697273254, "learning_rate": 2.317551183748146e-05, "loss": 0.0231, "step": 55700 }, { "grad_norm": 0.3963664472103119, "learning_rate": 2.3158071948906103e-05, "loss": 0.0248, "step": 55710 }, { "grad_norm": 0.20945261418819427, "learning_rate": 2.314063664688212e-05, "loss": 0.0186, "step": 55720 }, { "grad_norm": 0.22956696152687073, "learning_rate": 2.3123205934388725e-05, "loss": 0.019, "step": 55730 }, { "grad_norm": 0.5678451657295227, "learning_rate": 2.310577981440436e-05, "loss": 0.0315, "step": 55740 }, { "grad_norm": 0.49305692315101624, "learning_rate": 2.3088358289906665e-05, "loss": 0.0182, "step": 55750 }, { "grad_norm": 0.27783018350601196, "learning_rate": 2.307094136387252e-05, "loss": 0.0222, "step": 55760 }, { "grad_norm": 0.4221341907978058, "learning_rate": 2.305352903927796e-05, "loss": 0.0278, "step": 55770 }, { "grad_norm": 0.2625111937522888, "learning_rate": 2.303612131909831e-05, "loss": 0.018, "step": 55780 }, { "grad_norm": 0.335907518863678, "learning_rate": 2.3018718206308054e-05, "loss": 0.0259, "step": 55790 }, { "grad_norm": 0.32554060220718384, "learning_rate": 2.3001319703880925e-05, "loss": 0.0229, "step": 55800 }, { "grad_norm": 0.32102513313293457, "learning_rate": 2.2983925814789835e-05, "loss": 0.0169, "step": 55810 }, { "grad_norm": 0.3021968603134155, "learning_rate": 2.2966536542006957e-05, "loss": 0.0215, "step": 55820 }, { "grad_norm": 0.22512921690940857, "learning_rate": 2.294915188850358e-05, "loss": 0.0215, "step": 55830 }, { "grad_norm": 0.23697100579738617, "learning_rate": 2.2931771857250333e-05, "loss": 0.0233, "step": 55840 }, { "grad_norm": 0.17112700641155243, "learning_rate": 2.2914396451216946e-05, "loss": 0.0207, "step": 55850 }, { "grad_norm": 0.36649081110954285, "learning_rate": 2.2897025673372412e-05, "loss": 0.0254, "step": 55860 }, { "grad_norm": 0.23194816708564758, "learning_rate": 2.287965952668494e-05, "loss": 0.0242, "step": 55870 }, { "grad_norm": 0.299024760723114, "learning_rate": 2.2862298014121873e-05, "loss": 0.022, "step": 55880 }, { "grad_norm": 0.2820923924446106, "learning_rate": 2.28449411386499e-05, "loss": 0.0217, "step": 55890 }, { "grad_norm": 0.13981524109840393, "learning_rate": 2.282758890323477e-05, "loss": 0.0211, "step": 55900 }, { "grad_norm": 0.19113393127918243, "learning_rate": 2.2810241310841528e-05, "loss": 0.0219, "step": 55910 }, { "grad_norm": 0.30529993772506714, "learning_rate": 2.2792898364434412e-05, "loss": 0.0202, "step": 55920 }, { "grad_norm": 0.23212642967700958, "learning_rate": 2.2775560066976843e-05, "loss": 0.0198, "step": 55930 }, { "grad_norm": 0.24764789640903473, "learning_rate": 2.275822642143147e-05, "loss": 0.0248, "step": 55940 }, { "grad_norm": 0.39991992712020874, "learning_rate": 2.274089743076014e-05, "loss": 0.0228, "step": 55950 }, { "grad_norm": 0.46878623962402344, "learning_rate": 2.2723573097923917e-05, "loss": 0.0211, "step": 55960 }, { "grad_norm": 0.17193514108657837, "learning_rate": 2.2706253425883024e-05, "loss": 0.0157, "step": 55970 }, { "grad_norm": 0.3429236114025116, "learning_rate": 2.2688938417596933e-05, "loss": 0.0222, "step": 55980 }, { "grad_norm": 0.344957172870636, "learning_rate": 2.2671628076024305e-05, "loss": 0.022, "step": 55990 }, { "grad_norm": 0.3153017461299896, "learning_rate": 2.2654322404123008e-05, "loss": 0.0219, "step": 56000 }, { "grad_norm": 0.22225669026374817, "learning_rate": 2.2637021404850105e-05, "loss": 0.0173, "step": 56010 }, { "grad_norm": 0.3424198031425476, "learning_rate": 2.2619725081161885e-05, "loss": 0.0271, "step": 56020 }, { "grad_norm": 0.26831620931625366, "learning_rate": 2.260243343601376e-05, "loss": 0.0253, "step": 56030 }, { "grad_norm": 0.1964840590953827, "learning_rate": 2.2585146472360473e-05, "loss": 0.0253, "step": 56040 }, { "grad_norm": 0.29906997084617615, "learning_rate": 2.2567864193155834e-05, "loss": 0.0205, "step": 56050 }, { "grad_norm": 0.32447367906570435, "learning_rate": 2.255058660135294e-05, "loss": 0.0239, "step": 56060 }, { "grad_norm": 0.47774460911750793, "learning_rate": 2.253331369990407e-05, "loss": 0.0241, "step": 56070 }, { "grad_norm": 0.24751469492912292, "learning_rate": 2.251604549176063e-05, "loss": 0.0214, "step": 56080 }, { "grad_norm": 0.28469225764274597, "learning_rate": 2.249878197987337e-05, "loss": 0.0202, "step": 56090 }, { "grad_norm": 0.2647620737552643, "learning_rate": 2.2481523167192087e-05, "loss": 0.0197, "step": 56100 }, { "grad_norm": 0.2158212810754776, "learning_rate": 2.2464269056665867e-05, "loss": 0.0158, "step": 56110 }, { "grad_norm": 0.28841400146484375, "learning_rate": 2.2447019651242958e-05, "loss": 0.0254, "step": 56120 }, { "grad_norm": 0.23906251788139343, "learning_rate": 2.2429774953870818e-05, "loss": 0.0159, "step": 56130 }, { "grad_norm": 0.3460851311683655, "learning_rate": 2.241253496749611e-05, "loss": 0.0291, "step": 56140 }, { "grad_norm": 0.3481374680995941, "learning_rate": 2.2395299695064614e-05, "loss": 0.0202, "step": 56150 }, { "grad_norm": 0.34786975383758545, "learning_rate": 2.237806913952145e-05, "loss": 0.023, "step": 56160 }, { "grad_norm": 0.2953888177871704, "learning_rate": 2.2360843303810798e-05, "loss": 0.0221, "step": 56170 }, { "grad_norm": 0.23976999521255493, "learning_rate": 2.2343622190876084e-05, "loss": 0.02, "step": 56180 }, { "grad_norm": 0.20700716972351074, "learning_rate": 2.2326405803659935e-05, "loss": 0.0205, "step": 56190 }, { "grad_norm": 0.4012430012226105, "learning_rate": 2.230919414510416e-05, "loss": 0.0231, "step": 56200 }, { "grad_norm": 0.3771389424800873, "learning_rate": 2.229198721814976e-05, "loss": 0.0241, "step": 56210 }, { "grad_norm": 0.3154171407222748, "learning_rate": 2.2274785025736948e-05, "loss": 0.018, "step": 56220 }, { "grad_norm": 0.404267281293869, "learning_rate": 2.225758757080507e-05, "loss": 0.0257, "step": 56230 }, { "grad_norm": 0.39975231885910034, "learning_rate": 2.2240394856292723e-05, "loss": 0.024, "step": 56240 }, { "grad_norm": 0.2848794162273407, "learning_rate": 2.2223206885137664e-05, "loss": 0.0205, "step": 56250 }, { "grad_norm": 0.34914255142211914, "learning_rate": 2.2206023660276853e-05, "loss": 0.0255, "step": 56260 }, { "grad_norm": 0.29669517278671265, "learning_rate": 2.218884518464645e-05, "loss": 0.0243, "step": 56270 }, { "grad_norm": 0.3968856632709503, "learning_rate": 2.2171671461181732e-05, "loss": 0.0256, "step": 56280 }, { "grad_norm": 0.31083354353904724, "learning_rate": 2.2154502492817292e-05, "loss": 0.0263, "step": 56290 }, { "grad_norm": 0.29391971230506897, "learning_rate": 2.2137338282486782e-05, "loss": 0.0172, "step": 56300 }, { "grad_norm": 0.29517483711242676, "learning_rate": 2.2120178833123113e-05, "loss": 0.0196, "step": 56310 }, { "grad_norm": 0.2552933990955353, "learning_rate": 2.210302414765838e-05, "loss": 0.022, "step": 56320 }, { "grad_norm": 0.3205808401107788, "learning_rate": 2.2085874229023808e-05, "loss": 0.0198, "step": 56330 }, { "grad_norm": 0.2592824697494507, "learning_rate": 2.2068729080149907e-05, "loss": 0.0199, "step": 56340 }, { "grad_norm": 0.44875621795654297, "learning_rate": 2.205158870396625e-05, "loss": 0.0262, "step": 56350 }, { "grad_norm": 0.3522105813026428, "learning_rate": 2.2034453103401732e-05, "loss": 0.0221, "step": 56360 }, { "grad_norm": 0.24995191395282745, "learning_rate": 2.20173222813843e-05, "loss": 0.0183, "step": 56370 }, { "grad_norm": 0.3075889050960541, "learning_rate": 2.200019624084116e-05, "loss": 0.0307, "step": 56380 }, { "grad_norm": 0.23854783177375793, "learning_rate": 2.1983074984698687e-05, "loss": 0.0205, "step": 56390 }, { "grad_norm": 0.4170220196247101, "learning_rate": 2.1965958515882433e-05, "loss": 0.0287, "step": 56400 }, { "grad_norm": 0.22897812724113464, "learning_rate": 2.1948846837317162e-05, "loss": 0.0213, "step": 56410 }, { "grad_norm": 0.31618934869766235, "learning_rate": 2.1931739951926738e-05, "loss": 0.0192, "step": 56420 }, { "grad_norm": 0.31426921486854553, "learning_rate": 2.191463786263429e-05, "loss": 0.0179, "step": 56430 }, { "grad_norm": 0.43586209416389465, "learning_rate": 2.1897540572362095e-05, "loss": 0.0229, "step": 56440 }, { "grad_norm": 0.1801619678735733, "learning_rate": 2.1880448084031614e-05, "loss": 0.0159, "step": 56450 }, { "grad_norm": 0.5262104272842407, "learning_rate": 2.1863360400563482e-05, "loss": 0.023, "step": 56460 }, { "grad_norm": 0.27167370915412903, "learning_rate": 2.184627752487754e-05, "loss": 0.021, "step": 56470 }, { "grad_norm": 0.3831111788749695, "learning_rate": 2.1829199459892725e-05, "loss": 0.0214, "step": 56480 }, { "grad_norm": 0.4051913917064667, "learning_rate": 2.1812126208527282e-05, "loss": 0.0263, "step": 56490 }, { "grad_norm": 0.26557987928390503, "learning_rate": 2.179505777369852e-05, "loss": 0.0258, "step": 56500 }, { "grad_norm": 0.3643815219402313, "learning_rate": 2.1777994158322974e-05, "loss": 0.0256, "step": 56510 }, { "grad_norm": 0.2627946734428406, "learning_rate": 2.176093536531637e-05, "loss": 0.0223, "step": 56520 }, { "grad_norm": 0.29401320219039917, "learning_rate": 2.1743881397593537e-05, "loss": 0.0212, "step": 56530 }, { "grad_norm": 0.37636685371398926, "learning_rate": 2.1726832258068595e-05, "loss": 0.0332, "step": 56540 }, { "grad_norm": 0.3442032039165497, "learning_rate": 2.170978794965472e-05, "loss": 0.0253, "step": 56550 }, { "grad_norm": 0.18255014717578888, "learning_rate": 2.169274847526438e-05, "loss": 0.026, "step": 56560 }, { "grad_norm": 0.28852155804634094, "learning_rate": 2.1675713837809103e-05, "loss": 0.0274, "step": 56570 }, { "grad_norm": 0.2847847640514374, "learning_rate": 2.1658684040199655e-05, "loss": 0.0239, "step": 56580 }, { "grad_norm": 0.30314403772354126, "learning_rate": 2.1641659085345974e-05, "loss": 0.0206, "step": 56590 }, { "grad_norm": 0.23791533708572388, "learning_rate": 2.1624638976157154e-05, "loss": 0.0225, "step": 56600 }, { "grad_norm": 0.2805851697921753, "learning_rate": 2.1607623715541476e-05, "loss": 0.0281, "step": 56610 }, { "grad_norm": 0.2855519652366638, "learning_rate": 2.159061330640636e-05, "loss": 0.0155, "step": 56620 }, { "grad_norm": 0.29758140444755554, "learning_rate": 2.1573607751658425e-05, "loss": 0.0189, "step": 56630 }, { "grad_norm": 0.19272153079509735, "learning_rate": 2.155660705420347e-05, "loss": 0.0255, "step": 56640 }, { "grad_norm": 0.21266090869903564, "learning_rate": 2.153961121694644e-05, "loss": 0.0201, "step": 56650 }, { "grad_norm": 0.26098933815956116, "learning_rate": 2.152262024279145e-05, "loss": 0.0317, "step": 56660 }, { "grad_norm": 0.366353303194046, "learning_rate": 2.150563413464183e-05, "loss": 0.0237, "step": 56670 }, { "grad_norm": 0.3456684947013855, "learning_rate": 2.1488652895399974e-05, "loss": 0.0215, "step": 56680 }, { "grad_norm": 0.2654033601284027, "learning_rate": 2.1471676527967587e-05, "loss": 0.0227, "step": 56690 }, { "grad_norm": 0.23497986793518066, "learning_rate": 2.1454705035245414e-05, "loss": 0.0255, "step": 56700 }, { "grad_norm": 0.1575268805027008, "learning_rate": 2.143773842013343e-05, "loss": 0.0176, "step": 56710 }, { "grad_norm": 0.35835835337638855, "learning_rate": 2.1420776685530796e-05, "loss": 0.0217, "step": 56720 }, { "grad_norm": 0.2828618884086609, "learning_rate": 2.1403819834335742e-05, "loss": 0.0166, "step": 56730 }, { "grad_norm": 0.21723966300487518, "learning_rate": 2.1386867869445808e-05, "loss": 0.0235, "step": 56740 }, { "grad_norm": 0.17597007751464844, "learning_rate": 2.1369920793757548e-05, "loss": 0.026, "step": 56750 }, { "grad_norm": 0.31248170137405396, "learning_rate": 2.1352978610166828e-05, "loss": 0.0164, "step": 56760 }, { "grad_norm": 0.3105241060256958, "learning_rate": 2.1336041321568546e-05, "loss": 0.0233, "step": 56770 }, { "grad_norm": 0.31039687991142273, "learning_rate": 2.131910893085684e-05, "loss": 0.0176, "step": 56780 }, { "grad_norm": 0.22022172808647156, "learning_rate": 2.1302181440925012e-05, "loss": 0.026, "step": 56790 }, { "grad_norm": 0.20961272716522217, "learning_rate": 2.128525885466546e-05, "loss": 0.0208, "step": 56800 }, { "grad_norm": 0.3032037019729614, "learning_rate": 2.1268341174969847e-05, "loss": 0.0253, "step": 56810 }, { "grad_norm": 0.29290759563446045, "learning_rate": 2.1251428404728894e-05, "loss": 0.0196, "step": 56820 }, { "grad_norm": 0.3110068440437317, "learning_rate": 2.123452054683256e-05, "loss": 0.0178, "step": 56830 }, { "grad_norm": 0.3245558440685272, "learning_rate": 2.1217617604169927e-05, "loss": 0.0261, "step": 56840 }, { "grad_norm": 0.37380310893058777, "learning_rate": 2.120071957962924e-05, "loss": 0.0254, "step": 56850 }, { "grad_norm": 0.2790997624397278, "learning_rate": 2.1183826476097917e-05, "loss": 0.0321, "step": 56860 }, { "grad_norm": 0.19330282509326935, "learning_rate": 2.116693829646254e-05, "loss": 0.021, "step": 56870 }, { "grad_norm": 0.424356073141098, "learning_rate": 2.1150055043608806e-05, "loss": 0.0171, "step": 56880 }, { "grad_norm": 0.41104352474212646, "learning_rate": 2.113317672042162e-05, "loss": 0.0256, "step": 56890 }, { "grad_norm": 0.1953037679195404, "learning_rate": 2.1116303329785025e-05, "loss": 0.0189, "step": 56900 }, { "grad_norm": 0.26571935415267944, "learning_rate": 2.109943487458222e-05, "loss": 0.0301, "step": 56910 }, { "grad_norm": 0.28622934222221375, "learning_rate": 2.1082571357695574e-05, "loss": 0.024, "step": 56920 }, { "grad_norm": 0.2416459619998932, "learning_rate": 2.1065712782006557e-05, "loss": 0.0204, "step": 56930 }, { "grad_norm": 0.3584193289279938, "learning_rate": 2.104885915039591e-05, "loss": 0.0175, "step": 56940 }, { "grad_norm": 0.3928532004356384, "learning_rate": 2.103201046574338e-05, "loss": 0.0203, "step": 56950 }, { "grad_norm": 0.28162020444869995, "learning_rate": 2.1015166730928022e-05, "loss": 0.0156, "step": 56960 }, { "grad_norm": 0.3259463906288147, "learning_rate": 2.0998327948827912e-05, "loss": 0.0232, "step": 56970 }, { "grad_norm": 0.28199124336242676, "learning_rate": 2.0981494122320363e-05, "loss": 0.025, "step": 56980 }, { "grad_norm": 0.41963905096054077, "learning_rate": 2.0964665254281822e-05, "loss": 0.026, "step": 56990 }, { "grad_norm": 0.27449676394462585, "learning_rate": 2.094784134758784e-05, "loss": 0.0209, "step": 57000 }, { "grad_norm": 0.3741905987262726, "learning_rate": 2.0931022405113226e-05, "loss": 0.0336, "step": 57010 }, { "grad_norm": 0.25115644931793213, "learning_rate": 2.091420842973183e-05, "loss": 0.0283, "step": 57020 }, { "grad_norm": 0.24279505014419556, "learning_rate": 2.0897399424316715e-05, "loss": 0.0183, "step": 57030 }, { "grad_norm": 0.35183945298194885, "learning_rate": 2.0880595391740078e-05, "loss": 0.0254, "step": 57040 }, { "grad_norm": 0.3415166735649109, "learning_rate": 2.0863796334873277e-05, "loss": 0.0225, "step": 57050 }, { "grad_norm": 0.4345661401748657, "learning_rate": 2.08470022565868e-05, "loss": 0.0215, "step": 57060 }, { "grad_norm": 0.52607661485672, "learning_rate": 2.0830213159750317e-05, "loss": 0.0215, "step": 57070 }, { "grad_norm": 0.24323801696300507, "learning_rate": 2.0813429047232596e-05, "loss": 0.0235, "step": 57080 }, { "grad_norm": 0.19347058236598969, "learning_rate": 2.0796649921901594e-05, "loss": 0.0211, "step": 57090 }, { "grad_norm": 0.2045186311006546, "learning_rate": 2.077987578662441e-05, "loss": 0.0179, "step": 57100 }, { "grad_norm": 0.36097902059555054, "learning_rate": 2.0763106644267277e-05, "loss": 0.0274, "step": 57110 }, { "grad_norm": 0.20957723259925842, "learning_rate": 2.0746342497695607e-05, "loss": 0.0164, "step": 57120 }, { "grad_norm": 0.21839548647403717, "learning_rate": 2.0729583349773886e-05, "loss": 0.0144, "step": 57130 }, { "grad_norm": 0.3571532070636749, "learning_rate": 2.0712829203365853e-05, "loss": 0.0261, "step": 57140 }, { "grad_norm": 0.18482862412929535, "learning_rate": 2.0696080061334267e-05, "loss": 0.021, "step": 57150 }, { "grad_norm": 0.430922269821167, "learning_rate": 2.067933592654117e-05, "loss": 0.0176, "step": 57160 }, { "grad_norm": 0.4053666591644287, "learning_rate": 2.066259680184763e-05, "loss": 0.0219, "step": 57170 }, { "grad_norm": 0.2877019941806793, "learning_rate": 2.0645862690113908e-05, "loss": 0.0188, "step": 57180 }, { "grad_norm": 0.2487383633852005, "learning_rate": 2.0629133594199436e-05, "loss": 0.0224, "step": 57190 }, { "grad_norm": 0.3049262762069702, "learning_rate": 2.0612409516962704e-05, "loss": 0.0192, "step": 57200 }, { "grad_norm": 0.32275763154029846, "learning_rate": 2.0595690461261467e-05, "loss": 0.0178, "step": 57210 }, { "grad_norm": 0.4132617712020874, "learning_rate": 2.0578976429952503e-05, "loss": 0.0181, "step": 57220 }, { "grad_norm": 0.29966720938682556, "learning_rate": 2.0562267425891802e-05, "loss": 0.0208, "step": 57230 }, { "grad_norm": 0.24946527183055878, "learning_rate": 2.0545563451934467e-05, "loss": 0.0239, "step": 57240 }, { "grad_norm": 0.3286827802658081, "learning_rate": 2.0528864510934764e-05, "loss": 0.0246, "step": 57250 }, { "grad_norm": 0.2968653738498688, "learning_rate": 2.0512170605746096e-05, "loss": 0.0306, "step": 57260 }, { "grad_norm": 0.2873244881629944, "learning_rate": 2.0495481739220963e-05, "loss": 0.0259, "step": 57270 }, { "grad_norm": 0.3017657399177551, "learning_rate": 2.0478797914211045e-05, "loss": 0.0228, "step": 57280 }, { "grad_norm": 0.26137372851371765, "learning_rate": 2.046211913356716e-05, "loss": 0.026, "step": 57290 }, { "grad_norm": 0.29495769739151, "learning_rate": 2.0445445400139247e-05, "loss": 0.0258, "step": 57300 }, { "grad_norm": 0.5149932503700256, "learning_rate": 2.0428776716776405e-05, "loss": 0.0234, "step": 57310 }, { "grad_norm": 0.3049994707107544, "learning_rate": 2.0412113086326856e-05, "loss": 0.0209, "step": 57320 }, { "grad_norm": 0.15287767350673676, "learning_rate": 2.0395454511637918e-05, "loss": 0.0188, "step": 57330 }, { "grad_norm": 0.3037005066871643, "learning_rate": 2.037880099555616e-05, "loss": 0.0182, "step": 57340 }, { "grad_norm": 0.3531648814678192, "learning_rate": 2.0362152540927144e-05, "loss": 0.0187, "step": 57350 }, { "grad_norm": 0.32145771384239197, "learning_rate": 2.0345509150595666e-05, "loss": 0.0331, "step": 57360 }, { "grad_norm": 0.2802131772041321, "learning_rate": 2.0328870827405617e-05, "loss": 0.0275, "step": 57370 }, { "grad_norm": 0.2974739074707031, "learning_rate": 2.0312237574200043e-05, "loss": 0.0168, "step": 57380 }, { "grad_norm": 0.33147716522216797, "learning_rate": 2.029560939382112e-05, "loss": 0.02, "step": 57390 }, { "grad_norm": 0.28253716230392456, "learning_rate": 2.0278986289110097e-05, "loss": 0.0213, "step": 57400 }, { "grad_norm": 0.3398585021495819, "learning_rate": 2.0262368262907484e-05, "loss": 0.0211, "step": 57410 }, { "grad_norm": 0.3354627192020416, "learning_rate": 2.024575531805279e-05, "loss": 0.0231, "step": 57420 }, { "grad_norm": 0.3147572875022888, "learning_rate": 2.0229147457384735e-05, "loss": 0.0183, "step": 57430 }, { "grad_norm": 0.283926397562027, "learning_rate": 2.0212544683741157e-05, "loss": 0.0307, "step": 57440 }, { "grad_norm": 0.46512365341186523, "learning_rate": 2.0195946999958976e-05, "loss": 0.0285, "step": 57450 }, { "grad_norm": 0.27352413535118103, "learning_rate": 2.017935440887434e-05, "loss": 0.0235, "step": 57460 }, { "grad_norm": 0.29461538791656494, "learning_rate": 2.0162766913322423e-05, "loss": 0.0251, "step": 57470 }, { "grad_norm": 0.36668506264686584, "learning_rate": 2.0146184516137588e-05, "loss": 0.0269, "step": 57480 }, { "grad_norm": 0.2316720336675644, "learning_rate": 2.012960722015332e-05, "loss": 0.0208, "step": 57490 }, { "grad_norm": 0.4254511594772339, "learning_rate": 2.0113035028202214e-05, "loss": 0.0262, "step": 57500 }, { "grad_norm": 0.2505313456058502, "learning_rate": 2.009646794311602e-05, "loss": 0.0281, "step": 57510 }, { "grad_norm": 0.1672341525554657, "learning_rate": 2.007990596772559e-05, "loss": 0.0201, "step": 57520 }, { "grad_norm": 0.29957258701324463, "learning_rate": 2.0063349104860923e-05, "loss": 0.0172, "step": 57530 }, { "grad_norm": 0.39855775237083435, "learning_rate": 2.0046797357351116e-05, "loss": 0.029, "step": 57540 }, { "grad_norm": 0.2557062804698944, "learning_rate": 2.0030250728024412e-05, "loss": 0.0174, "step": 57550 }, { "grad_norm": 0.16135762631893158, "learning_rate": 2.001370921970819e-05, "loss": 0.0179, "step": 57560 }, { "grad_norm": 0.2611357867717743, "learning_rate": 1.9997172835228932e-05, "loss": 0.0206, "step": 57570 }, { "grad_norm": 0.36550673842430115, "learning_rate": 1.9980641577412262e-05, "loss": 0.0207, "step": 57580 }, { "grad_norm": 0.35077086091041565, "learning_rate": 1.9964115449082925e-05, "loss": 0.0227, "step": 57590 }, { "grad_norm": 0.26178598403930664, "learning_rate": 1.9947594453064742e-05, "loss": 0.0294, "step": 57600 }, { "grad_norm": 0.26609519124031067, "learning_rate": 1.9931078592180774e-05, "loss": 0.0204, "step": 57610 }, { "grad_norm": 0.2887312173843384, "learning_rate": 1.9914567869253065e-05, "loss": 0.022, "step": 57620 }, { "grad_norm": 0.20247486233711243, "learning_rate": 1.989806228710287e-05, "loss": 0.0294, "step": 57630 }, { "grad_norm": 0.32846102118492126, "learning_rate": 1.9881561848550555e-05, "loss": 0.0208, "step": 57640 }, { "grad_norm": 0.505169689655304, "learning_rate": 1.9865066556415544e-05, "loss": 0.0275, "step": 57650 }, { "grad_norm": 0.25845301151275635, "learning_rate": 1.98485764135165e-05, "loss": 0.0221, "step": 57660 }, { "grad_norm": 0.18564863502979279, "learning_rate": 1.983209142267109e-05, "loss": 0.0225, "step": 57670 }, { "grad_norm": 0.253224641084671, "learning_rate": 1.9815611586696165e-05, "loss": 0.0193, "step": 57680 }, { "grad_norm": 0.278642863035202, "learning_rate": 1.9799136908407667e-05, "loss": 0.0215, "step": 57690 }, { "grad_norm": 0.22567547857761383, "learning_rate": 1.9782667390620678e-05, "loss": 0.0198, "step": 57700 }, { "grad_norm": 0.22856497764587402, "learning_rate": 1.976620303614939e-05, "loss": 0.0178, "step": 57710 }, { "grad_norm": 0.25470080971717834, "learning_rate": 1.9749743847807108e-05, "loss": 0.0189, "step": 57720 }, { "grad_norm": 0.3185634911060333, "learning_rate": 1.9733289828406272e-05, "loss": 0.0184, "step": 57730 }, { "grad_norm": 0.5230122208595276, "learning_rate": 1.9716840980758382e-05, "loss": 0.023, "step": 57740 }, { "grad_norm": 0.1686767041683197, "learning_rate": 1.9700397307674134e-05, "loss": 0.0181, "step": 57750 }, { "grad_norm": 0.28113800287246704, "learning_rate": 1.968395881196328e-05, "loss": 0.0173, "step": 57760 }, { "grad_norm": 0.36277133226394653, "learning_rate": 1.966752549643473e-05, "loss": 0.0234, "step": 57770 }, { "grad_norm": 0.3488534986972809, "learning_rate": 1.965109736389647e-05, "loss": 0.0211, "step": 57780 }, { "grad_norm": 0.4279229938983917, "learning_rate": 1.9634674417155645e-05, "loss": 0.0198, "step": 57790 }, { "grad_norm": 0.1430191993713379, "learning_rate": 1.9618256659018434e-05, "loss": 0.0374, "step": 57800 }, { "grad_norm": 0.44127583503723145, "learning_rate": 1.9601844092290257e-05, "loss": 0.0251, "step": 57810 }, { "grad_norm": 0.33973419666290283, "learning_rate": 1.9585436719775512e-05, "loss": 0.0177, "step": 57820 }, { "grad_norm": 0.5528987646102905, "learning_rate": 1.9569034544277793e-05, "loss": 0.024, "step": 57830 }, { "grad_norm": 0.26428577303886414, "learning_rate": 1.9552637568599798e-05, "loss": 0.0195, "step": 57840 }, { "grad_norm": 0.4048834443092346, "learning_rate": 1.953624579554327e-05, "loss": 0.0325, "step": 57850 }, { "grad_norm": 0.30402863025665283, "learning_rate": 1.951985922790918e-05, "loss": 0.0307, "step": 57860 }, { "grad_norm": 0.2942570745944977, "learning_rate": 1.9503477868497505e-05, "loss": 0.0159, "step": 57870 }, { "grad_norm": 0.25328850746154785, "learning_rate": 1.9487101720107375e-05, "loss": 0.0237, "step": 57880 }, { "grad_norm": 0.30256059765815735, "learning_rate": 1.9470730785537032e-05, "loss": 0.021, "step": 57890 }, { "grad_norm": 0.1917562037706375, "learning_rate": 1.9454365067583823e-05, "loss": 0.0266, "step": 57900 }, { "grad_norm": 0.28015944361686707, "learning_rate": 1.9438004569044215e-05, "loss": 0.0164, "step": 57910 }, { "grad_norm": 0.2825837731361389, "learning_rate": 1.9421649292713724e-05, "loss": 0.0393, "step": 57920 }, { "grad_norm": 0.2612088918685913, "learning_rate": 1.9405299241387076e-05, "loss": 0.0242, "step": 57930 }, { "grad_norm": 0.2694719731807709, "learning_rate": 1.9388954417858007e-05, "loss": 0.0199, "step": 57940 }, { "grad_norm": 0.34062817692756653, "learning_rate": 1.9372614824919417e-05, "loss": 0.0194, "step": 57950 }, { "grad_norm": 0.31675171852111816, "learning_rate": 1.9356280465363284e-05, "loss": 0.0253, "step": 57960 }, { "grad_norm": 0.3331143260002136, "learning_rate": 1.9339951341980723e-05, "loss": 0.0222, "step": 57970 }, { "grad_norm": 0.2985215187072754, "learning_rate": 1.9323627457561916e-05, "loss": 0.0176, "step": 57980 }, { "grad_norm": 0.22570806741714478, "learning_rate": 1.9307308814896198e-05, "loss": 0.0137, "step": 57990 }, { "grad_norm": 0.17391188442707062, "learning_rate": 1.9290995416771935e-05, "loss": 0.0174, "step": 58000 }, { "grad_norm": 0.27499645948410034, "learning_rate": 1.9274687265976665e-05, "loss": 0.0197, "step": 58010 }, { "grad_norm": 0.2863370478153229, "learning_rate": 1.9258384365297e-05, "loss": 0.0233, "step": 58020 }, { "grad_norm": 0.3096730411052704, "learning_rate": 1.924208671751866e-05, "loss": 0.0235, "step": 58030 }, { "grad_norm": 0.26988711953163147, "learning_rate": 1.9225794325426492e-05, "loss": 0.019, "step": 58040 }, { "grad_norm": 0.2181294709444046, "learning_rate": 1.920950719180436e-05, "loss": 0.0193, "step": 58050 }, { "grad_norm": 0.24836844205856323, "learning_rate": 1.919322531943536e-05, "loss": 0.0177, "step": 58060 }, { "grad_norm": 0.19770623743534088, "learning_rate": 1.917694871110157e-05, "loss": 0.0205, "step": 58070 }, { "grad_norm": 0.29465651512145996, "learning_rate": 1.9160677369584234e-05, "loss": 0.0182, "step": 58080 }, { "grad_norm": 0.3612777888774872, "learning_rate": 1.9144411297663694e-05, "loss": 0.0191, "step": 58090 }, { "grad_norm": 0.21093812584877014, "learning_rate": 1.9128150498119328e-05, "loss": 0.0249, "step": 58100 }, { "grad_norm": 0.2149544507265091, "learning_rate": 1.9111894973729726e-05, "loss": 0.0184, "step": 58110 }, { "grad_norm": 0.4684544801712036, "learning_rate": 1.9095644727272454e-05, "loss": 0.0254, "step": 58120 }, { "grad_norm": 0.31226032972335815, "learning_rate": 1.907939976152429e-05, "loss": 0.0236, "step": 58130 }, { "grad_norm": 0.3752997815608978, "learning_rate": 1.906316007926101e-05, "loss": 0.0204, "step": 58140 }, { "grad_norm": 0.2561136484146118, "learning_rate": 1.904692568325755e-05, "loss": 0.0218, "step": 58150 }, { "grad_norm": 0.3005928099155426, "learning_rate": 1.9030696576287925e-05, "loss": 0.0216, "step": 58160 }, { "grad_norm": 0.5644886493682861, "learning_rate": 1.9014472761125242e-05, "loss": 0.0194, "step": 58170 }, { "grad_norm": 0.2487514615058899, "learning_rate": 1.899825424054172e-05, "loss": 0.0242, "step": 58180 }, { "grad_norm": 0.2910962700843811, "learning_rate": 1.898204101730863e-05, "loss": 0.0246, "step": 58190 }, { "grad_norm": 0.28947821259498596, "learning_rate": 1.8965833094196394e-05, "loss": 0.0305, "step": 58200 }, { "grad_norm": 0.3168828785419464, "learning_rate": 1.8949630473974495e-05, "loss": 0.0175, "step": 58210 }, { "grad_norm": 0.2659476399421692, "learning_rate": 1.8933433159411517e-05, "loss": 0.0245, "step": 58220 }, { "grad_norm": 0.2760142982006073, "learning_rate": 1.891724115327514e-05, "loss": 0.0187, "step": 58230 }, { "grad_norm": 0.1653367578983307, "learning_rate": 1.8901054458332152e-05, "loss": 0.0197, "step": 58240 }, { "grad_norm": 0.2835707664489746, "learning_rate": 1.8884873077348364e-05, "loss": 0.0209, "step": 58250 }, { "grad_norm": 0.2536466717720032, "learning_rate": 1.88686970130888e-05, "loss": 0.0267, "step": 58260 }, { "grad_norm": 0.30019840598106384, "learning_rate": 1.8852526268317456e-05, "loss": 0.0231, "step": 58270 }, { "grad_norm": 0.4299471080303192, "learning_rate": 1.883636084579749e-05, "loss": 0.0181, "step": 58280 }, { "grad_norm": 0.31996551156044006, "learning_rate": 1.8820200748291145e-05, "loss": 0.0213, "step": 58290 }, { "grad_norm": 0.20225322246551514, "learning_rate": 1.8804045978559686e-05, "loss": 0.0197, "step": 58300 }, { "grad_norm": 0.2809683084487915, "learning_rate": 1.8787896539363594e-05, "loss": 0.0209, "step": 58310 }, { "grad_norm": 0.2242305725812912, "learning_rate": 1.8771752433462298e-05, "loss": 0.0194, "step": 58320 }, { "grad_norm": 0.24328762292861938, "learning_rate": 1.8755613663614447e-05, "loss": 0.0196, "step": 58330 }, { "grad_norm": 0.3787776529788971, "learning_rate": 1.873948023257767e-05, "loss": 0.0199, "step": 58340 }, { "grad_norm": 0.33590370416641235, "learning_rate": 1.8723352143108747e-05, "loss": 0.022, "step": 58350 }, { "grad_norm": 0.2628783881664276, "learning_rate": 1.870722939796352e-05, "loss": 0.0231, "step": 58360 }, { "grad_norm": 0.3354879319667816, "learning_rate": 1.869111199989693e-05, "loss": 0.0197, "step": 58370 }, { "grad_norm": 0.28604310750961304, "learning_rate": 1.8674999951663018e-05, "loss": 0.0232, "step": 58380 }, { "grad_norm": 0.17156030237674713, "learning_rate": 1.8658893256014857e-05, "loss": 0.0229, "step": 58390 }, { "grad_norm": 0.23305220901966095, "learning_rate": 1.8642791915704655e-05, "loss": 0.0176, "step": 58400 }, { "grad_norm": 0.3110596537590027, "learning_rate": 1.8626695933483697e-05, "loss": 0.0182, "step": 58410 }, { "grad_norm": 0.24070808291435242, "learning_rate": 1.8610605312102335e-05, "loss": 0.0184, "step": 58420 }, { "grad_norm": 0.24091607332229614, "learning_rate": 1.8594520054310032e-05, "loss": 0.0257, "step": 58430 }, { "grad_norm": 0.3645646870136261, "learning_rate": 1.8578440162855326e-05, "loss": 0.0232, "step": 58440 }, { "grad_norm": 0.2607828974723816, "learning_rate": 1.8562365640485774e-05, "loss": 0.0229, "step": 58450 }, { "grad_norm": 0.5417611598968506, "learning_rate": 1.854629648994815e-05, "loss": 0.0156, "step": 58460 }, { "grad_norm": 0.25531813502311707, "learning_rate": 1.8530232713988183e-05, "loss": 0.022, "step": 58470 }, { "grad_norm": 0.36827588081359863, "learning_rate": 1.851417431535074e-05, "loss": 0.0192, "step": 58480 }, { "grad_norm": 0.2515951693058014, "learning_rate": 1.8498121296779785e-05, "loss": 0.035, "step": 58490 }, { "grad_norm": 0.3167261481285095, "learning_rate": 1.848207366101829e-05, "loss": 0.0271, "step": 58500 }, { "grad_norm": 0.1804349571466446, "learning_rate": 1.8466031410808422e-05, "loss": 0.0198, "step": 58510 }, { "grad_norm": 0.33989307284355164, "learning_rate": 1.844999454889129e-05, "loss": 0.0257, "step": 58520 }, { "grad_norm": 0.4209316670894623, "learning_rate": 1.8433963078007237e-05, "loss": 0.0291, "step": 58530 }, { "grad_norm": 0.20833340287208557, "learning_rate": 1.841793700089554e-05, "loss": 0.0179, "step": 58540 }, { "grad_norm": 0.30719825625419617, "learning_rate": 1.8401916320294644e-05, "loss": 0.023, "step": 58550 }, { "grad_norm": 0.33554187417030334, "learning_rate": 1.838590103894205e-05, "loss": 0.0192, "step": 58560 }, { "grad_norm": 0.3154371678829193, "learning_rate": 1.8369891159574292e-05, "loss": 0.0175, "step": 58570 }, { "grad_norm": 0.34178513288497925, "learning_rate": 1.835388668492708e-05, "loss": 0.023, "step": 58580 }, { "grad_norm": 0.19275197386741638, "learning_rate": 1.8337887617735095e-05, "loss": 0.021, "step": 58590 }, { "grad_norm": 0.27302786707878113, "learning_rate": 1.8321893960732157e-05, "loss": 0.0192, "step": 58600 }, { "grad_norm": 0.3545147478580475, "learning_rate": 1.8305905716651138e-05, "loss": 0.0208, "step": 58610 }, { "grad_norm": 0.1546659767627716, "learning_rate": 1.8289922888224e-05, "loss": 0.0187, "step": 58620 }, { "grad_norm": 0.295867919921875, "learning_rate": 1.8273945478181765e-05, "loss": 0.0226, "step": 58630 }, { "grad_norm": 0.34105798602104187, "learning_rate": 1.8257973489254558e-05, "loss": 0.0176, "step": 58640 }, { "grad_norm": 0.2987159490585327, "learning_rate": 1.824200692417152e-05, "loss": 0.0231, "step": 58650 }, { "grad_norm": 0.2869669198989868, "learning_rate": 1.8226045785660912e-05, "loss": 0.0201, "step": 58660 }, { "grad_norm": 0.30663028359413147, "learning_rate": 1.821009007645006e-05, "loss": 0.0191, "step": 58670 }, { "grad_norm": 0.2440437525510788, "learning_rate": 1.8194139799265357e-05, "loss": 0.0184, "step": 58680 }, { "grad_norm": 0.22362618148326874, "learning_rate": 1.8178194956832295e-05, "loss": 0.021, "step": 58690 }, { "grad_norm": 0.19105127453804016, "learning_rate": 1.8162255551875346e-05, "loss": 0.0179, "step": 58700 }, { "grad_norm": 0.25659045577049255, "learning_rate": 1.8146321587118194e-05, "loss": 0.0258, "step": 58710 }, { "grad_norm": 0.21189473569393158, "learning_rate": 1.8130393065283448e-05, "loss": 0.014, "step": 58720 }, { "grad_norm": 0.16681233048439026, "learning_rate": 1.8114469989092925e-05, "loss": 0.0184, "step": 58730 }, { "grad_norm": 0.371390163898468, "learning_rate": 1.8098552361267397e-05, "loss": 0.0176, "step": 58740 }, { "grad_norm": 0.29977089166641235, "learning_rate": 1.8082640184526763e-05, "loss": 0.0168, "step": 58750 }, { "grad_norm": 0.2396034449338913, "learning_rate": 1.806673346158999e-05, "loss": 0.0249, "step": 58760 }, { "grad_norm": 0.2501909136772156, "learning_rate": 1.8050832195175067e-05, "loss": 0.023, "step": 58770 }, { "grad_norm": 0.28126060962677, "learning_rate": 1.8034936387999136e-05, "loss": 0.0261, "step": 58780 }, { "grad_norm": 0.2555795907974243, "learning_rate": 1.8019046042778315e-05, "loss": 0.0211, "step": 58790 }, { "grad_norm": 0.32808274030685425, "learning_rate": 1.800316116222785e-05, "loss": 0.0223, "step": 58800 }, { "grad_norm": 0.3188146948814392, "learning_rate": 1.7987281749062018e-05, "loss": 0.0452, "step": 58810 }, { "grad_norm": 0.15736830234527588, "learning_rate": 1.7971407805994195e-05, "loss": 0.0277, "step": 58820 }, { "grad_norm": 0.21687810122966766, "learning_rate": 1.7955539335736787e-05, "loss": 0.0231, "step": 58830 }, { "grad_norm": 0.2863287329673767, "learning_rate": 1.7939676341001304e-05, "loss": 0.0248, "step": 58840 }, { "grad_norm": 0.6847284436225891, "learning_rate": 1.7923818824498275e-05, "loss": 0.0235, "step": 58850 }, { "grad_norm": 0.3846658766269684, "learning_rate": 1.7907966788937315e-05, "loss": 0.0273, "step": 58860 }, { "grad_norm": 0.32900819182395935, "learning_rate": 1.7892120237027116e-05, "loss": 0.0259, "step": 58870 }, { "grad_norm": 0.20981557667255402, "learning_rate": 1.7876279171475413e-05, "loss": 0.0262, "step": 58880 }, { "grad_norm": 0.3101307451725006, "learning_rate": 1.7860443594989028e-05, "loss": 0.0203, "step": 58890 }, { "grad_norm": 0.4481273293495178, "learning_rate": 1.784461351027379e-05, "loss": 0.0242, "step": 58900 }, { "grad_norm": 0.27202850580215454, "learning_rate": 1.7828788920034677e-05, "loss": 0.0213, "step": 58910 }, { "grad_norm": 0.40736642479896545, "learning_rate": 1.7812969826975623e-05, "loss": 0.0243, "step": 58920 }, { "grad_norm": 0.3068479597568512, "learning_rate": 1.7797156233799738e-05, "loss": 0.0224, "step": 58930 }, { "grad_norm": 0.2802692949771881, "learning_rate": 1.7781348143209094e-05, "loss": 0.0214, "step": 58940 }, { "grad_norm": 0.341427206993103, "learning_rate": 1.7765545557904862e-05, "loss": 0.0357, "step": 58950 }, { "grad_norm": 0.39739853143692017, "learning_rate": 1.7749748480587302e-05, "loss": 0.0213, "step": 58960 }, { "grad_norm": 0.28341370820999146, "learning_rate": 1.773395691395564e-05, "loss": 0.0181, "step": 58970 }, { "grad_norm": 0.18426816165447235, "learning_rate": 1.7718170860708305e-05, "loss": 0.0171, "step": 58980 }, { "grad_norm": 0.30186599493026733, "learning_rate": 1.770239032354264e-05, "loss": 0.0173, "step": 58990 }, { "grad_norm": 0.26354342699050903, "learning_rate": 1.7686615305155124e-05, "loss": 0.0281, "step": 59000 }, { "grad_norm": 0.2312670648097992, "learning_rate": 1.767084580824128e-05, "loss": 0.0171, "step": 59010 }, { "grad_norm": 0.4882020950317383, "learning_rate": 1.765508183549569e-05, "loss": 0.0276, "step": 59020 }, { "grad_norm": 0.33677905797958374, "learning_rate": 1.763932338961199e-05, "loss": 0.0226, "step": 59030 }, { "grad_norm": 0.18000556528568268, "learning_rate": 1.762357047328284e-05, "loss": 0.0215, "step": 59040 }, { "grad_norm": 0.2982247769832611, "learning_rate": 1.7607823089199997e-05, "loss": 0.029, "step": 59050 }, { "grad_norm": 0.36436623334884644, "learning_rate": 1.7592081240054265e-05, "loss": 0.0232, "step": 59060 }, { "grad_norm": 0.6280304193496704, "learning_rate": 1.7576344928535488e-05, "loss": 0.0281, "step": 59070 }, { "grad_norm": 0.2035650610923767, "learning_rate": 1.7560614157332573e-05, "loss": 0.0234, "step": 59080 }, { "grad_norm": 0.26392605900764465, "learning_rate": 1.7544888929133495e-05, "loss": 0.0228, "step": 59090 }, { "grad_norm": 0.23022553324699402, "learning_rate": 1.752916924662522e-05, "loss": 0.022, "step": 59100 }, { "grad_norm": 0.27736374735832214, "learning_rate": 1.751345511249387e-05, "loss": 0.0192, "step": 59110 }, { "grad_norm": 0.2737398147583008, "learning_rate": 1.749774652942452e-05, "loss": 0.0175, "step": 59120 }, { "grad_norm": 0.3769804537296295, "learning_rate": 1.748204350010135e-05, "loss": 0.0245, "step": 59130 }, { "grad_norm": 0.2865353524684906, "learning_rate": 1.7466346027207574e-05, "loss": 0.0172, "step": 59140 }, { "grad_norm": 0.30757707357406616, "learning_rate": 1.745065411342547e-05, "loss": 0.0187, "step": 59150 }, { "grad_norm": 0.2020430564880371, "learning_rate": 1.7434967761436366e-05, "loss": 0.0203, "step": 59160 }, { "grad_norm": 0.397312194108963, "learning_rate": 1.741928697392058e-05, "loss": 0.0192, "step": 59170 }, { "grad_norm": 0.34766989946365356, "learning_rate": 1.7403611753557597e-05, "loss": 0.017, "step": 59180 }, { "grad_norm": 0.3412821590900421, "learning_rate": 1.738794210302584e-05, "loss": 0.0156, "step": 59190 }, { "grad_norm": 0.2734825313091278, "learning_rate": 1.7372278025002837e-05, "loss": 0.0144, "step": 59200 }, { "grad_norm": 0.39011022448539734, "learning_rate": 1.7356619522165164e-05, "loss": 0.0222, "step": 59210 }, { "grad_norm": 0.21646751463413239, "learning_rate": 1.7340966597188378e-05, "loss": 0.018, "step": 59220 }, { "grad_norm": 0.2389940768480301, "learning_rate": 1.732531925274722e-05, "loss": 0.0204, "step": 59230 }, { "grad_norm": 0.18410712480545044, "learning_rate": 1.7309677491515318e-05, "loss": 0.022, "step": 59240 }, { "grad_norm": 0.3322753608226776, "learning_rate": 1.7294041316165455e-05, "loss": 0.0199, "step": 59250 }, { "grad_norm": 0.4487570524215698, "learning_rate": 1.727841072936942e-05, "loss": 0.0223, "step": 59260 }, { "grad_norm": 0.31425943970680237, "learning_rate": 1.7262785733798058e-05, "loss": 0.0171, "step": 59270 }, { "grad_norm": 0.3750581443309784, "learning_rate": 1.7247166332121246e-05, "loss": 0.0255, "step": 59280 }, { "grad_norm": 0.22540251910686493, "learning_rate": 1.7231552527007933e-05, "loss": 0.0178, "step": 59290 }, { "grad_norm": 0.2091735601425171, "learning_rate": 1.721594432112606e-05, "loss": 0.0204, "step": 59300 }, { "grad_norm": 0.2739276885986328, "learning_rate": 1.7200341717142655e-05, "loss": 0.0214, "step": 59310 }, { "grad_norm": 0.3469107449054718, "learning_rate": 1.718474471772377e-05, "loss": 0.0216, "step": 59320 }, { "grad_norm": 0.1900692880153656, "learning_rate": 1.7169153325534528e-05, "loss": 0.0192, "step": 59330 }, { "grad_norm": 0.15820133686065674, "learning_rate": 1.7153567543239045e-05, "loss": 0.0215, "step": 59340 }, { "grad_norm": 0.3750211298465729, "learning_rate": 1.7137987373500525e-05, "loss": 0.0195, "step": 59350 }, { "grad_norm": 0.23142312467098236, "learning_rate": 1.7122412818981198e-05, "loss": 0.0189, "step": 59360 }, { "grad_norm": 0.3928084671497345, "learning_rate": 1.7106843882342283e-05, "loss": 0.0196, "step": 59370 }, { "grad_norm": 0.14056217670440674, "learning_rate": 1.709128056624415e-05, "loss": 0.0211, "step": 59380 }, { "grad_norm": 0.2614624500274658, "learning_rate": 1.70757228733461e-05, "loss": 0.0184, "step": 59390 }, { "grad_norm": 0.28094327449798584, "learning_rate": 1.706017080630653e-05, "loss": 0.0189, "step": 59400 }, { "grad_norm": 0.29060274362564087, "learning_rate": 1.7044624367782873e-05, "loss": 0.0238, "step": 59410 }, { "grad_norm": 0.2828405499458313, "learning_rate": 1.7029083560431553e-05, "loss": 0.0203, "step": 59420 }, { "grad_norm": 0.3408413231372833, "learning_rate": 1.701354838690813e-05, "loss": 0.0245, "step": 59430 }, { "grad_norm": 0.35390812158584595, "learning_rate": 1.6998018849867087e-05, "loss": 0.0202, "step": 59440 }, { "grad_norm": 0.2056458592414856, "learning_rate": 1.698249495196202e-05, "loss": 0.0202, "step": 59450 }, { "grad_norm": 0.3054809868335724, "learning_rate": 1.6966976695845528e-05, "loss": 0.0229, "step": 59460 }, { "grad_norm": 0.25205305218696594, "learning_rate": 1.6951464084169268e-05, "loss": 0.0154, "step": 59470 }, { "grad_norm": 0.2547605335712433, "learning_rate": 1.6935957119583916e-05, "loss": 0.0256, "step": 59480 }, { "grad_norm": 0.2943670153617859, "learning_rate": 1.6920455804739205e-05, "loss": 0.0211, "step": 59490 }, { "grad_norm": 0.2732965350151062, "learning_rate": 1.6904960142283855e-05, "loss": 0.0229, "step": 59500 }, { "grad_norm": 0.19863247871398926, "learning_rate": 1.6889470134865666e-05, "loss": 0.0236, "step": 59510 }, { "grad_norm": 0.30263030529022217, "learning_rate": 1.687398578513145e-05, "loss": 0.0208, "step": 59520 }, { "grad_norm": 0.20426487922668457, "learning_rate": 1.6858507095727066e-05, "loss": 0.0235, "step": 59530 }, { "grad_norm": 0.46620649099349976, "learning_rate": 1.6843034069297403e-05, "loss": 0.0231, "step": 59540 }, { "grad_norm": 0.3596016466617584, "learning_rate": 1.682756670848637e-05, "loss": 0.0321, "step": 59550 }, { "grad_norm": 0.2711925506591797, "learning_rate": 1.6812105015936937e-05, "loss": 0.0168, "step": 59560 }, { "grad_norm": 0.2548338770866394, "learning_rate": 1.6796648994291027e-05, "loss": 0.0311, "step": 59570 }, { "grad_norm": 0.2341727912425995, "learning_rate": 1.678119864618973e-05, "loss": 0.0142, "step": 59580 }, { "grad_norm": 0.27378344535827637, "learning_rate": 1.6765753974273023e-05, "loss": 0.016, "step": 59590 }, { "grad_norm": 0.17680244147777557, "learning_rate": 1.675031498118001e-05, "loss": 0.0162, "step": 59600 }, { "grad_norm": 0.3404827117919922, "learning_rate": 1.6734881669548804e-05, "loss": 0.0178, "step": 59610 }, { "grad_norm": 0.251682311296463, "learning_rate": 1.671945404201648e-05, "loss": 0.0232, "step": 59620 }, { "grad_norm": 0.27301421761512756, "learning_rate": 1.670403210121927e-05, "loss": 0.0162, "step": 59630 }, { "grad_norm": 0.342434287071228, "learning_rate": 1.6688615849792312e-05, "loss": 0.0203, "step": 59640 }, { "grad_norm": 0.3787449896335602, "learning_rate": 1.6673205290369832e-05, "loss": 0.0219, "step": 59650 }, { "grad_norm": 0.3444303274154663, "learning_rate": 1.6657800425585086e-05, "loss": 0.0242, "step": 59660 }, { "grad_norm": 0.31721439957618713, "learning_rate": 1.6642401258070327e-05, "loss": 0.0278, "step": 59670 }, { "grad_norm": 0.34661608934402466, "learning_rate": 1.6627007790456884e-05, "loss": 0.0255, "step": 59680 }, { "grad_norm": 0.3474047780036926, "learning_rate": 1.6611620025375037e-05, "loss": 0.0183, "step": 59690 }, { "grad_norm": 0.2647140324115753, "learning_rate": 1.6596237965454154e-05, "loss": 0.0268, "step": 59700 }, { "grad_norm": 0.308707058429718, "learning_rate": 1.6580861613322607e-05, "loss": 0.0243, "step": 59710 }, { "grad_norm": 0.3666343688964844, "learning_rate": 1.6565490971607796e-05, "loss": 0.0161, "step": 59720 }, { "grad_norm": 0.268449991941452, "learning_rate": 1.655012604293614e-05, "loss": 0.0233, "step": 59730 }, { "grad_norm": 0.24363653361797333, "learning_rate": 1.653476682993309e-05, "loss": 0.0267, "step": 59740 }, { "grad_norm": 0.231988787651062, "learning_rate": 1.651941333522311e-05, "loss": 0.0169, "step": 59750 }, { "grad_norm": 0.3329761326313019, "learning_rate": 1.6504065561429715e-05, "loss": 0.0231, "step": 59760 }, { "grad_norm": 0.2903544008731842, "learning_rate": 1.6488723511175385e-05, "loss": 0.0204, "step": 59770 }, { "grad_norm": 0.37508219480514526, "learning_rate": 1.6473387187081668e-05, "loss": 0.0205, "step": 59780 }, { "grad_norm": 0.20030957460403442, "learning_rate": 1.6458056591769123e-05, "loss": 0.0164, "step": 59790 }, { "grad_norm": 0.238486647605896, "learning_rate": 1.6442731727857335e-05, "loss": 0.0204, "step": 59800 }, { "grad_norm": 0.23407071828842163, "learning_rate": 1.642741259796492e-05, "loss": 0.0157, "step": 59810 }, { "grad_norm": 0.283070832490921, "learning_rate": 1.641209920470944e-05, "loss": 0.0187, "step": 59820 }, { "grad_norm": 0.3608572781085968, "learning_rate": 1.639679155070762e-05, "loss": 0.0178, "step": 59830 }, { "grad_norm": 0.25371137261390686, "learning_rate": 1.6381489638575048e-05, "loss": 0.0172, "step": 59840 }, { "grad_norm": 0.18943819403648376, "learning_rate": 1.636619347092643e-05, "loss": 0.0215, "step": 59850 }, { "grad_norm": 0.42305681109428406, "learning_rate": 1.6350903050375476e-05, "loss": 0.0204, "step": 59860 }, { "grad_norm": 0.3673294186592102, "learning_rate": 1.6335618379534856e-05, "loss": 0.0212, "step": 59870 }, { "grad_norm": 0.3500216007232666, "learning_rate": 1.6320339461016364e-05, "loss": 0.0165, "step": 59880 }, { "grad_norm": 0.27804479002952576, "learning_rate": 1.6305066297430687e-05, "loss": 0.0183, "step": 59890 }, { "grad_norm": 0.24956077337265015, "learning_rate": 1.6289798891387654e-05, "loss": 0.0222, "step": 59900 }, { "grad_norm": 0.2469254434108734, "learning_rate": 1.6274537245495995e-05, "loss": 0.014, "step": 59910 }, { "grad_norm": 0.23672904074192047, "learning_rate": 1.6259281362363527e-05, "loss": 0.0151, "step": 59920 }, { "grad_norm": 0.29395851492881775, "learning_rate": 1.6244031244597068e-05, "loss": 0.0292, "step": 59930 }, { "grad_norm": 0.20795059204101562, "learning_rate": 1.622878689480244e-05, "loss": 0.0184, "step": 59940 }, { "grad_norm": 0.2747776508331299, "learning_rate": 1.6213548315584498e-05, "loss": 0.026, "step": 59950 }, { "grad_norm": 0.2780425548553467, "learning_rate": 1.6198315509547074e-05, "loss": 0.0251, "step": 59960 }, { "grad_norm": 0.23409897089004517, "learning_rate": 1.6183088479293056e-05, "loss": 0.016, "step": 59970 }, { "grad_norm": 0.31653839349746704, "learning_rate": 1.6167867227424316e-05, "loss": 0.023, "step": 59980 }, { "grad_norm": 0.4283687174320221, "learning_rate": 1.615265175654176e-05, "loss": 0.0213, "step": 59990 }, { "grad_norm": 0.529841423034668, "learning_rate": 1.613744206924529e-05, "loss": 0.0181, "step": 60000 }, { "grad_norm": 0.3972591161727905, "learning_rate": 1.6122238168133845e-05, "loss": 0.0261, "step": 60010 }, { "grad_norm": 0.24324531853199005, "learning_rate": 1.6107040055805305e-05, "loss": 0.0225, "step": 60020 }, { "grad_norm": 0.4514460265636444, "learning_rate": 1.6091847734856673e-05, "loss": 0.0208, "step": 60030 }, { "grad_norm": 0.39886465668678284, "learning_rate": 1.6076661207883865e-05, "loss": 0.0224, "step": 60040 }, { "grad_norm": 0.17009204626083374, "learning_rate": 1.6061480477481848e-05, "loss": 0.0205, "step": 60050 }, { "grad_norm": 0.4742911458015442, "learning_rate": 1.604630554624461e-05, "loss": 0.0182, "step": 60060 }, { "grad_norm": 0.426878422498703, "learning_rate": 1.603113641676509e-05, "loss": 0.024, "step": 60070 }, { "grad_norm": 0.2888212203979492, "learning_rate": 1.6015973091635338e-05, "loss": 0.0218, "step": 60080 }, { "grad_norm": 0.18626682460308075, "learning_rate": 1.6000815573446288e-05, "loss": 0.0191, "step": 60090 }, { "grad_norm": 0.32362937927246094, "learning_rate": 1.5985663864788002e-05, "loss": 0.032, "step": 60100 }, { "grad_norm": 0.23750391602516174, "learning_rate": 1.597051796824946e-05, "loss": 0.0231, "step": 60110 }, { "grad_norm": 0.23535408079624176, "learning_rate": 1.5955377886418682e-05, "loss": 0.021, "step": 60120 }, { "grad_norm": 0.20664705336093903, "learning_rate": 1.5940243621882704e-05, "loss": 0.0204, "step": 60130 }, { "grad_norm": 0.24548956751823425, "learning_rate": 1.5925115177227555e-05, "loss": 0.0234, "step": 60140 }, { "grad_norm": 0.24359826743602753, "learning_rate": 1.5909992555038288e-05, "loss": 0.0189, "step": 60150 }, { "grad_norm": 0.21553683280944824, "learning_rate": 1.5894875757898912e-05, "loss": 0.026, "step": 60160 }, { "grad_norm": 0.32424312829971313, "learning_rate": 1.58797647883925e-05, "loss": 0.0196, "step": 60170 }, { "grad_norm": 0.30657243728637695, "learning_rate": 1.586465964910109e-05, "loss": 0.0283, "step": 60180 }, { "grad_norm": 0.22513018548488617, "learning_rate": 1.5849560342605734e-05, "loss": 0.0185, "step": 60190 }, { "grad_norm": 0.4132100045681, "learning_rate": 1.583446687148651e-05, "loss": 0.0203, "step": 60200 }, { "grad_norm": 0.31296300888061523, "learning_rate": 1.581937923832248e-05, "loss": 0.0199, "step": 60210 }, { "grad_norm": 0.25085997581481934, "learning_rate": 1.5804297445691663e-05, "loss": 0.0239, "step": 60220 }, { "grad_norm": 0.31605803966522217, "learning_rate": 1.578922149617119e-05, "loss": 0.0226, "step": 60230 }, { "grad_norm": 0.3124222755432129, "learning_rate": 1.5774151392337084e-05, "loss": 0.0194, "step": 60240 }, { "grad_norm": 0.3366437256336212, "learning_rate": 1.575908713676442e-05, "loss": 0.0197, "step": 60250 }, { "grad_norm": 0.32334017753601074, "learning_rate": 1.5744028732027293e-05, "loss": 0.0249, "step": 60260 }, { "grad_norm": 0.41469690203666687, "learning_rate": 1.5728976180698723e-05, "loss": 0.0231, "step": 60270 }, { "grad_norm": 0.2932194173336029, "learning_rate": 1.5713929485350836e-05, "loss": 0.0208, "step": 60280 }, { "grad_norm": 0.19520729780197144, "learning_rate": 1.569888864855465e-05, "loss": 0.0225, "step": 60290 }, { "grad_norm": 0.2321937531232834, "learning_rate": 1.568385367288028e-05, "loss": 0.0193, "step": 60300 }, { "grad_norm": 0.2469947189092636, "learning_rate": 1.5668824560896755e-05, "loss": 0.0229, "step": 60310 }, { "grad_norm": 0.2632278800010681, "learning_rate": 1.5653801315172156e-05, "loss": 0.023, "step": 60320 }, { "grad_norm": 0.3895263373851776, "learning_rate": 1.5638783938273554e-05, "loss": 0.0215, "step": 60330 }, { "grad_norm": 0.32098016142845154, "learning_rate": 1.5623772432766966e-05, "loss": 0.0272, "step": 60340 }, { "grad_norm": 0.19487157464027405, "learning_rate": 1.5608766801217507e-05, "loss": 0.0185, "step": 60350 }, { "grad_norm": 0.20037570595741272, "learning_rate": 1.5593767046189184e-05, "loss": 0.0192, "step": 60360 }, { "grad_norm": 0.32801005244255066, "learning_rate": 1.5578773170245064e-05, "loss": 0.0197, "step": 60370 }, { "grad_norm": 0.40780240297317505, "learning_rate": 1.5563785175947182e-05, "loss": 0.0275, "step": 60380 }, { "grad_norm": 0.34272462129592896, "learning_rate": 1.5548803065856582e-05, "loss": 0.0153, "step": 60390 }, { "grad_norm": 0.35997065901756287, "learning_rate": 1.55338268425333e-05, "loss": 0.0207, "step": 60400 }, { "grad_norm": 0.275964617729187, "learning_rate": 1.5518856508536373e-05, "loss": 0.0199, "step": 60410 }, { "grad_norm": 0.34932851791381836, "learning_rate": 1.5503892066423786e-05, "loss": 0.0209, "step": 60420 }, { "grad_norm": 0.6056084036827087, "learning_rate": 1.548893351875258e-05, "loss": 0.0351, "step": 60430 }, { "grad_norm": 0.3325495719909668, "learning_rate": 1.547398086807875e-05, "loss": 0.019, "step": 60440 }, { "grad_norm": 0.40989765524864197, "learning_rate": 1.5459034116957305e-05, "loss": 0.0291, "step": 60450 }, { "grad_norm": 0.34178513288497925, "learning_rate": 1.544409326794225e-05, "loss": 0.0322, "step": 60460 }, { "grad_norm": 0.24079445004463196, "learning_rate": 1.542915832358651e-05, "loss": 0.0188, "step": 60470 }, { "grad_norm": 0.3115508258342743, "learning_rate": 1.541422928644214e-05, "loss": 0.0192, "step": 60480 }, { "grad_norm": 0.36171168088912964, "learning_rate": 1.5399306159060024e-05, "loss": 0.0298, "step": 60490 }, { "grad_norm": 0.211768239736557, "learning_rate": 1.5384388943990185e-05, "loss": 0.0185, "step": 60500 }, { "grad_norm": 0.3235252797603607, "learning_rate": 1.5369477643781526e-05, "loss": 0.0182, "step": 60510 }, { "grad_norm": 0.25707098841667175, "learning_rate": 1.5354572260981985e-05, "loss": 0.0203, "step": 60520 }, { "grad_norm": 0.3489812910556793, "learning_rate": 1.533967279813851e-05, "loss": 0.015, "step": 60530 }, { "grad_norm": 0.3214423656463623, "learning_rate": 1.5324779257796956e-05, "loss": 0.0236, "step": 60540 }, { "grad_norm": 0.2574880123138428, "learning_rate": 1.5309891642502293e-05, "loss": 0.0212, "step": 60550 }, { "grad_norm": 0.21651431918144226, "learning_rate": 1.5295009954798357e-05, "loss": 0.0165, "step": 60560 }, { "grad_norm": 0.27896013855934143, "learning_rate": 1.5280134197228036e-05, "loss": 0.0184, "step": 60570 }, { "grad_norm": 0.14877212047576904, "learning_rate": 1.526526437233319e-05, "loss": 0.0203, "step": 60580 }, { "grad_norm": 0.25013309717178345, "learning_rate": 1.525040048265466e-05, "loss": 0.0202, "step": 60590 }, { "grad_norm": 0.3419767916202545, "learning_rate": 1.5235542530732288e-05, "loss": 0.0155, "step": 60600 }, { "grad_norm": 0.2092740386724472, "learning_rate": 1.5220690519104901e-05, "loss": 0.0243, "step": 60610 }, { "grad_norm": 0.2718680500984192, "learning_rate": 1.5205844450310275e-05, "loss": 0.0173, "step": 60620 }, { "grad_norm": 0.30090054869651794, "learning_rate": 1.519100432688521e-05, "loss": 0.0195, "step": 60630 }, { "grad_norm": 0.22703371942043304, "learning_rate": 1.5176170151365476e-05, "loss": 0.0208, "step": 60640 }, { "grad_norm": 0.2998499870300293, "learning_rate": 1.5161341926285833e-05, "loss": 0.0171, "step": 60650 }, { "grad_norm": 0.26570942997932434, "learning_rate": 1.5146519654180025e-05, "loss": 0.0157, "step": 60660 }, { "grad_norm": 0.3303144574165344, "learning_rate": 1.5131703337580739e-05, "loss": 0.0237, "step": 60670 }, { "grad_norm": 0.29576441645622253, "learning_rate": 1.5116892979019731e-05, "loss": 0.0226, "step": 60680 }, { "grad_norm": 0.2650686502456665, "learning_rate": 1.5102088581027623e-05, "loss": 0.019, "step": 60690 }, { "grad_norm": 0.2671697735786438, "learning_rate": 1.5087290146134154e-05, "loss": 0.0176, "step": 60700 }, { "grad_norm": 0.24695968627929688, "learning_rate": 1.5072497676867915e-05, "loss": 0.0195, "step": 60710 }, { "grad_norm": 0.24164678156375885, "learning_rate": 1.505771117575655e-05, "loss": 0.0224, "step": 60720 }, { "grad_norm": 0.4772138297557831, "learning_rate": 1.5042930645326691e-05, "loss": 0.0155, "step": 60730 }, { "grad_norm": 0.3020775616168976, "learning_rate": 1.5028156088103879e-05, "loss": 0.0216, "step": 60740 }, { "grad_norm": 0.5038131475448608, "learning_rate": 1.5013387506612735e-05, "loss": 0.0283, "step": 60750 }, { "grad_norm": 0.3085716962814331, "learning_rate": 1.4998624903376767e-05, "loss": 0.0246, "step": 60760 }, { "grad_norm": 0.4661196172237396, "learning_rate": 1.4983868280918517e-05, "loss": 0.0207, "step": 60770 }, { "grad_norm": 0.2511206269264221, "learning_rate": 1.4969117641759478e-05, "loss": 0.0215, "step": 60780 }, { "grad_norm": 0.2137136310338974, "learning_rate": 1.495437298842014e-05, "loss": 0.0204, "step": 60790 }, { "grad_norm": 0.2845293879508972, "learning_rate": 1.4939634323419976e-05, "loss": 0.0194, "step": 60800 }, { "grad_norm": 0.33646586537361145, "learning_rate": 1.4924901649277384e-05, "loss": 0.0197, "step": 60810 }, { "grad_norm": 0.2088509351015091, "learning_rate": 1.4910174968509793e-05, "loss": 0.0205, "step": 60820 }, { "grad_norm": 0.31605610251426697, "learning_rate": 1.4895454283633598e-05, "loss": 0.0223, "step": 60830 }, { "grad_norm": 0.38763830065727234, "learning_rate": 1.488073959716415e-05, "loss": 0.0217, "step": 60840 }, { "grad_norm": 0.2539733946323395, "learning_rate": 1.4866030911615791e-05, "loss": 0.0201, "step": 60850 }, { "grad_norm": 0.24240653216838837, "learning_rate": 1.4851328229501849e-05, "loss": 0.0165, "step": 60860 }, { "grad_norm": 0.1753007024526596, "learning_rate": 1.4836631553334562e-05, "loss": 0.0203, "step": 60870 }, { "grad_norm": 0.27777716517448425, "learning_rate": 1.4821940885625251e-05, "loss": 0.0244, "step": 60880 }, { "grad_norm": 0.3875623345375061, "learning_rate": 1.4807256228884109e-05, "loss": 0.0211, "step": 60890 }, { "grad_norm": 0.19105537235736847, "learning_rate": 1.4792577585620353e-05, "loss": 0.0184, "step": 60900 }, { "grad_norm": 0.2118639051914215, "learning_rate": 1.4777904958342164e-05, "loss": 0.0229, "step": 60910 }, { "grad_norm": 0.2829149663448334, "learning_rate": 1.4763238349556691e-05, "loss": 0.0169, "step": 60920 }, { "grad_norm": 0.2584630250930786, "learning_rate": 1.4748577761770072e-05, "loss": 0.0236, "step": 60930 }, { "grad_norm": 0.2795957028865814, "learning_rate": 1.4733923197487354e-05, "loss": 0.018, "step": 60940 }, { "grad_norm": 0.33879727125167847, "learning_rate": 1.4719274659212662e-05, "loss": 0.0219, "step": 60950 }, { "grad_norm": 0.256955087184906, "learning_rate": 1.4704632149448983e-05, "loss": 0.0208, "step": 60960 }, { "grad_norm": 0.45019447803497314, "learning_rate": 1.4689995670698343e-05, "loss": 0.0325, "step": 60970 }, { "grad_norm": 0.2583770751953125, "learning_rate": 1.4675365225461728e-05, "loss": 0.0255, "step": 60980 }, { "grad_norm": 0.3959498107433319, "learning_rate": 1.4660740816239032e-05, "loss": 0.0164, "step": 60990 }, { "grad_norm": 0.3301544487476349, "learning_rate": 1.464612244552923e-05, "loss": 0.0223, "step": 61000 }, { "grad_norm": 0.28206735849380493, "learning_rate": 1.4631510115830161e-05, "loss": 0.0219, "step": 61010 }, { "grad_norm": 0.27643904089927673, "learning_rate": 1.4616903829638679e-05, "loss": 0.0193, "step": 61020 }, { "grad_norm": 0.2647942006587982, "learning_rate": 1.46023035894506e-05, "loss": 0.0199, "step": 61030 }, { "grad_norm": 0.41089898347854614, "learning_rate": 1.4587709397760713e-05, "loss": 0.0226, "step": 61040 }, { "grad_norm": 0.19728679955005646, "learning_rate": 1.4573121257062755e-05, "loss": 0.018, "step": 61050 }, { "grad_norm": 0.42692428827285767, "learning_rate": 1.4558539169849472e-05, "loss": 0.0264, "step": 61060 }, { "grad_norm": 0.4070594012737274, "learning_rate": 1.4543963138612499e-05, "loss": 0.0181, "step": 61070 }, { "grad_norm": 0.28886446356773376, "learning_rate": 1.4529393165842498e-05, "loss": 0.0204, "step": 61080 }, { "grad_norm": 0.2422354817390442, "learning_rate": 1.4514829254029084e-05, "loss": 0.0187, "step": 61090 }, { "grad_norm": 0.15784859657287598, "learning_rate": 1.450027140566083e-05, "loss": 0.0163, "step": 61100 }, { "grad_norm": 0.2621479034423828, "learning_rate": 1.4485719623225268e-05, "loss": 0.0176, "step": 61110 }, { "grad_norm": 0.3277587294578552, "learning_rate": 1.4471173909208912e-05, "loss": 0.0179, "step": 61120 }, { "grad_norm": 0.16368508338928223, "learning_rate": 1.4456634266097236e-05, "loss": 0.0196, "step": 61130 }, { "grad_norm": 0.2251151204109192, "learning_rate": 1.444210069637461e-05, "loss": 0.0301, "step": 61140 }, { "grad_norm": 0.27025550603866577, "learning_rate": 1.4427573202524502e-05, "loss": 0.0146, "step": 61150 }, { "grad_norm": 0.2339688092470169, "learning_rate": 1.4413051787029208e-05, "loss": 0.0254, "step": 61160 }, { "grad_norm": 0.18319563567638397, "learning_rate": 1.439853645237006e-05, "loss": 0.0198, "step": 61170 }, { "grad_norm": 0.25139573216438293, "learning_rate": 1.4384027201027344e-05, "loss": 0.0151, "step": 61180 }, { "grad_norm": 0.3238712549209595, "learning_rate": 1.4369524035480253e-05, "loss": 0.0206, "step": 61190 }, { "grad_norm": 0.33875083923339844, "learning_rate": 1.4355026958207035e-05, "loss": 0.0187, "step": 61200 }, { "grad_norm": 0.414673775434494, "learning_rate": 1.434053597168481e-05, "loss": 0.0212, "step": 61210 }, { "grad_norm": 0.47968727350234985, "learning_rate": 1.432605107838969e-05, "loss": 0.0198, "step": 61220 }, { "grad_norm": 0.3881264626979828, "learning_rate": 1.4311572280796759e-05, "loss": 0.0263, "step": 61230 }, { "grad_norm": 0.21089105308055878, "learning_rate": 1.4297099581380047e-05, "loss": 0.0243, "step": 61240 }, { "grad_norm": 0.4897814691066742, "learning_rate": 1.4282632982612538e-05, "loss": 0.021, "step": 61250 }, { "grad_norm": 0.33498913049697876, "learning_rate": 1.4268172486966197e-05, "loss": 0.0274, "step": 61260 }, { "grad_norm": 0.2637805640697479, "learning_rate": 1.4253718096911894e-05, "loss": 0.0207, "step": 61270 }, { "grad_norm": 0.3506273329257965, "learning_rate": 1.4239269814919504e-05, "loss": 0.0259, "step": 61280 }, { "grad_norm": 0.2227724939584732, "learning_rate": 1.4224827643457843e-05, "loss": 0.0287, "step": 61290 }, { "grad_norm": 0.30351370573043823, "learning_rate": 1.4210391584994686e-05, "loss": 0.0218, "step": 61300 }, { "grad_norm": 0.3850429058074951, "learning_rate": 1.4195961641996763e-05, "loss": 0.0201, "step": 61310 }, { "grad_norm": 0.31205910444259644, "learning_rate": 1.4181537816929751e-05, "loss": 0.0186, "step": 61320 }, { "grad_norm": 0.17420844733715057, "learning_rate": 1.4167120112258303e-05, "loss": 0.0225, "step": 61330 }, { "grad_norm": 0.227693110704422, "learning_rate": 1.4152708530445963e-05, "loss": 0.0156, "step": 61340 }, { "grad_norm": 0.25732603669166565, "learning_rate": 1.4138303073955344e-05, "loss": 0.0162, "step": 61350 }, { "grad_norm": 0.333673357963562, "learning_rate": 1.4123903745247897e-05, "loss": 0.021, "step": 61360 }, { "grad_norm": 0.12081195414066315, "learning_rate": 1.4109510546784082e-05, "loss": 0.0136, "step": 61370 }, { "grad_norm": 0.3707043528556824, "learning_rate": 1.4095123481023325e-05, "loss": 0.0264, "step": 61380 }, { "grad_norm": 0.43116408586502075, "learning_rate": 1.4080742550423936e-05, "loss": 0.0209, "step": 61390 }, { "grad_norm": 0.21123681962490082, "learning_rate": 1.4066367757443278e-05, "loss": 0.0168, "step": 61400 }, { "grad_norm": 0.3109770715236664, "learning_rate": 1.4051999104537572e-05, "loss": 0.0167, "step": 61410 }, { "grad_norm": 0.25085628032684326, "learning_rate": 1.4037636594162034e-05, "loss": 0.0197, "step": 61420 }, { "grad_norm": 0.2129286825656891, "learning_rate": 1.402328022877083e-05, "loss": 0.016, "step": 61430 }, { "grad_norm": 0.16565483808517456, "learning_rate": 1.4008930010817073e-05, "loss": 0.0231, "step": 61440 }, { "grad_norm": 0.29159224033355713, "learning_rate": 1.3994585942752831e-05, "loss": 0.016, "step": 61450 }, { "grad_norm": 0.28189486265182495, "learning_rate": 1.398024802702909e-05, "loss": 0.015, "step": 61460 }, { "grad_norm": 0.5144730806350708, "learning_rate": 1.3965916266095819e-05, "loss": 0.0242, "step": 61470 }, { "grad_norm": 0.29524603486061096, "learning_rate": 1.395159066240192e-05, "loss": 0.0196, "step": 61480 }, { "grad_norm": 0.2247762680053711, "learning_rate": 1.3937271218395248e-05, "loss": 0.0183, "step": 61490 }, { "grad_norm": 0.3494146466255188, "learning_rate": 1.3922957936522613e-05, "loss": 0.021, "step": 61500 }, { "grad_norm": 0.2974020540714264, "learning_rate": 1.3908650819229758e-05, "loss": 0.0189, "step": 61510 }, { "grad_norm": 0.1952522099018097, "learning_rate": 1.389434986896137e-05, "loss": 0.0165, "step": 61520 }, { "grad_norm": 0.32072529196739197, "learning_rate": 1.388005508816112e-05, "loss": 0.0138, "step": 61530 }, { "grad_norm": 0.34806886315345764, "learning_rate": 1.3865766479271557e-05, "loss": 0.0259, "step": 61540 }, { "grad_norm": 0.27006596326828003, "learning_rate": 1.3851484044734225e-05, "loss": 0.0308, "step": 61550 }, { "grad_norm": 0.35000503063201904, "learning_rate": 1.3837207786989608e-05, "loss": 0.0277, "step": 61560 }, { "grad_norm": 0.23498770594596863, "learning_rate": 1.382293770847713e-05, "loss": 0.0176, "step": 61570 }, { "grad_norm": 0.29493042826652527, "learning_rate": 1.3808673811635159e-05, "loss": 0.0188, "step": 61580 }, { "grad_norm": 0.327679306268692, "learning_rate": 1.3794416098900975e-05, "loss": 0.0238, "step": 61590 }, { "grad_norm": 0.17700690031051636, "learning_rate": 1.3780164572710886e-05, "loss": 0.0239, "step": 61600 }, { "grad_norm": 0.23372164368629456, "learning_rate": 1.3765919235500035e-05, "loss": 0.0218, "step": 61610 }, { "grad_norm": 0.18567848205566406, "learning_rate": 1.375168008970259e-05, "loss": 0.0186, "step": 61620 }, { "grad_norm": 0.19352108240127563, "learning_rate": 1.3737447137751635e-05, "loss": 0.0122, "step": 61630 }, { "grad_norm": 0.432381808757782, "learning_rate": 1.372322038207915e-05, "loss": 0.0194, "step": 61640 }, { "grad_norm": 0.2907665967941284, "learning_rate": 1.3708999825116159e-05, "loss": 0.0278, "step": 61650 }, { "grad_norm": 0.26802605390548706, "learning_rate": 1.3694785469292526e-05, "loss": 0.0241, "step": 61660 }, { "grad_norm": 0.28381791710853577, "learning_rate": 1.3680577317037101e-05, "loss": 0.0163, "step": 61670 }, { "grad_norm": 0.3916725516319275, "learning_rate": 1.3666375370777678e-05, "loss": 0.0212, "step": 61680 }, { "grad_norm": 0.32815414667129517, "learning_rate": 1.3652179632940981e-05, "loss": 0.0147, "step": 61690 }, { "grad_norm": 0.20158226788043976, "learning_rate": 1.3637990105952664e-05, "loss": 0.017, "step": 61700 }, { "grad_norm": 0.32365068793296814, "learning_rate": 1.3623806792237337e-05, "loss": 0.0217, "step": 61710 }, { "grad_norm": 0.20837923884391785, "learning_rate": 1.3609629694218551e-05, "loss": 0.0219, "step": 61720 }, { "grad_norm": 0.3631752133369446, "learning_rate": 1.3595458814318762e-05, "loss": 0.0218, "step": 61730 }, { "grad_norm": 0.21215415000915527, "learning_rate": 1.3581294154959389e-05, "loss": 0.0234, "step": 61740 }, { "grad_norm": 0.4584795832633972, "learning_rate": 1.3567135718560792e-05, "loss": 0.0196, "step": 61750 }, { "grad_norm": 0.20396658778190613, "learning_rate": 1.3552983507542261e-05, "loss": 0.0152, "step": 61760 }, { "grad_norm": 0.24841871857643127, "learning_rate": 1.3538837524322023e-05, "loss": 0.0213, "step": 61770 }, { "grad_norm": 0.15184786915779114, "learning_rate": 1.3524697771317251e-05, "loss": 0.0269, "step": 61780 }, { "grad_norm": 0.3083540201187134, "learning_rate": 1.3510564250943997e-05, "loss": 0.0184, "step": 61790 }, { "grad_norm": 0.34644535183906555, "learning_rate": 1.3496436965617353e-05, "loss": 0.0161, "step": 61800 }, { "grad_norm": 0.21438336372375488, "learning_rate": 1.3482315917751243e-05, "loss": 0.0209, "step": 61810 }, { "grad_norm": 0.16134028136730194, "learning_rate": 1.3468201109758582e-05, "loss": 0.0138, "step": 61820 }, { "grad_norm": 0.17511029541492462, "learning_rate": 1.3454092544051222e-05, "loss": 0.019, "step": 61830 }, { "grad_norm": 0.2948000431060791, "learning_rate": 1.3439990223039878e-05, "loss": 0.0192, "step": 61840 }, { "grad_norm": 0.17633184790611267, "learning_rate": 1.3425894149134321e-05, "loss": 0.015, "step": 61850 }, { "grad_norm": 0.43525683879852295, "learning_rate": 1.3411804324743132e-05, "loss": 0.0207, "step": 61860 }, { "grad_norm": 0.2749800682067871, "learning_rate": 1.3397720752273896e-05, "loss": 0.0215, "step": 61870 }, { "grad_norm": 0.18325327336788177, "learning_rate": 1.3383643434133108e-05, "loss": 0.0163, "step": 61880 }, { "grad_norm": 0.22337502241134644, "learning_rate": 1.3369572372726196e-05, "loss": 0.0206, "step": 61890 }, { "grad_norm": 0.18321593105793, "learning_rate": 1.3355507570457525e-05, "loss": 0.0148, "step": 61900 }, { "grad_norm": 0.2060653269290924, "learning_rate": 1.334144902973038e-05, "loss": 0.0156, "step": 61910 }, { "grad_norm": 0.34454652667045593, "learning_rate": 1.3327396752946997e-05, "loss": 0.0205, "step": 61920 }, { "grad_norm": 0.2462986260652542, "learning_rate": 1.3313350742508496e-05, "loss": 0.0226, "step": 61930 }, { "grad_norm": 0.2049323469400406, "learning_rate": 1.3299311000814973e-05, "loss": 0.0229, "step": 61940 }, { "grad_norm": 0.28706255555152893, "learning_rate": 1.3285277530265432e-05, "loss": 0.0223, "step": 61950 }, { "grad_norm": 0.4255671799182892, "learning_rate": 1.3271250333257813e-05, "loss": 0.0203, "step": 61960 }, { "grad_norm": 0.16137182712554932, "learning_rate": 1.3257229412188977e-05, "loss": 0.0158, "step": 61970 }, { "grad_norm": 0.2918176054954529, "learning_rate": 1.3243214769454726e-05, "loss": 0.0194, "step": 61980 }, { "grad_norm": 0.2721744775772095, "learning_rate": 1.3229206407449751e-05, "loss": 0.0152, "step": 61990 }, { "grad_norm": 0.2756923735141754, "learning_rate": 1.3215204328567742e-05, "loss": 0.0235, "step": 62000 }, { "grad_norm": 0.16579246520996094, "learning_rate": 1.3201208535201232e-05, "loss": 0.0196, "step": 62010 }, { "grad_norm": 0.21916936337947845, "learning_rate": 1.3187219029741737e-05, "loss": 0.0182, "step": 62020 }, { "grad_norm": 0.3246799111366272, "learning_rate": 1.3173235814579693e-05, "loss": 0.0177, "step": 62030 }, { "grad_norm": 0.2988986372947693, "learning_rate": 1.3159258892104398e-05, "loss": 0.0192, "step": 62040 }, { "grad_norm": 0.3275321125984192, "learning_rate": 1.3145288264704198e-05, "loss": 0.0216, "step": 62050 }, { "grad_norm": 0.16598093509674072, "learning_rate": 1.3131323934766237e-05, "loss": 0.0159, "step": 62060 }, { "grad_norm": 0.12136055529117584, "learning_rate": 1.311736590467666e-05, "loss": 0.0151, "step": 62070 }, { "grad_norm": 0.34016671776771545, "learning_rate": 1.3103414176820505e-05, "loss": 0.0172, "step": 62080 }, { "grad_norm": 0.21881568431854248, "learning_rate": 1.308946875358174e-05, "loss": 0.0291, "step": 62090 }, { "grad_norm": 0.25339794158935547, "learning_rate": 1.3075529637343276e-05, "loss": 0.0284, "step": 62100 }, { "grad_norm": 0.3760688304901123, "learning_rate": 1.3061596830486883e-05, "loss": 0.0247, "step": 62110 }, { "grad_norm": 0.33204925060272217, "learning_rate": 1.3047670335393353e-05, "loss": 0.0134, "step": 62120 }, { "grad_norm": 0.3689684569835663, "learning_rate": 1.3033750154442298e-05, "loss": 0.026, "step": 62130 }, { "grad_norm": 0.21385376155376434, "learning_rate": 1.3019836290012316e-05, "loss": 0.0165, "step": 62140 }, { "grad_norm": 0.24001668393611908, "learning_rate": 1.30059287444809e-05, "loss": 0.0182, "step": 62150 }, { "grad_norm": 0.35326963663101196, "learning_rate": 1.299202752022447e-05, "loss": 0.0269, "step": 62160 }, { "grad_norm": 0.23559562861919403, "learning_rate": 1.2978132619618371e-05, "loss": 0.0151, "step": 62170 }, { "grad_norm": 0.20107626914978027, "learning_rate": 1.2964244045036866e-05, "loss": 0.0214, "step": 62180 }, { "grad_norm": 0.32279691100120544, "learning_rate": 1.295036179885311e-05, "loss": 0.0153, "step": 62190 }, { "grad_norm": 0.29575830698013306, "learning_rate": 1.293648588343922e-05, "loss": 0.0234, "step": 62200 }, { "grad_norm": 0.3091807961463928, "learning_rate": 1.2922616301166196e-05, "loss": 0.0181, "step": 62210 }, { "grad_norm": 0.40431052446365356, "learning_rate": 1.2908753054403976e-05, "loss": 0.0146, "step": 62220 }, { "grad_norm": 0.30849435925483704, "learning_rate": 1.2894896145521429e-05, "loss": 0.0157, "step": 62230 }, { "grad_norm": 0.21533513069152832, "learning_rate": 1.2881045576886275e-05, "loss": 0.0177, "step": 62240 }, { "grad_norm": 0.24260319769382477, "learning_rate": 1.2867201350865254e-05, "loss": 0.0211, "step": 62250 }, { "grad_norm": 0.18628692626953125, "learning_rate": 1.2853363469823914e-05, "loss": 0.0167, "step": 62260 }, { "grad_norm": 0.23098231852054596, "learning_rate": 1.283953193612682e-05, "loss": 0.0137, "step": 62270 }, { "grad_norm": 0.27426639199256897, "learning_rate": 1.2825706752137372e-05, "loss": 0.0166, "step": 62280 }, { "grad_norm": 0.32070663571357727, "learning_rate": 1.2811887920217896e-05, "loss": 0.0176, "step": 62290 }, { "grad_norm": 0.2253694236278534, "learning_rate": 1.2798075442729707e-05, "loss": 0.0163, "step": 62300 }, { "grad_norm": 0.51468825340271, "learning_rate": 1.2784269322032922e-05, "loss": 0.0167, "step": 62310 }, { "grad_norm": 0.23071914911270142, "learning_rate": 1.2770469560486686e-05, "loss": 0.021, "step": 62320 }, { "grad_norm": 0.1683356910943985, "learning_rate": 1.2756676160448956e-05, "loss": 0.02, "step": 62330 }, { "grad_norm": 0.39637380838394165, "learning_rate": 1.2742889124276663e-05, "loss": 0.0199, "step": 62340 }, { "grad_norm": 0.22396671772003174, "learning_rate": 1.2729108454325639e-05, "loss": 0.0225, "step": 62350 }, { "grad_norm": 0.2546790540218353, "learning_rate": 1.2715334152950614e-05, "loss": 0.0189, "step": 62360 }, { "grad_norm": 0.173140287399292, "learning_rate": 1.2701566222505246e-05, "loss": 0.0183, "step": 62370 }, { "grad_norm": 0.3764305114746094, "learning_rate": 1.2687804665342107e-05, "loss": 0.0242, "step": 62380 }, { "grad_norm": 0.26717498898506165, "learning_rate": 1.267404948381265e-05, "loss": 0.0168, "step": 62390 }, { "grad_norm": 0.34117573499679565, "learning_rate": 1.2660300680267267e-05, "loss": 0.0162, "step": 62400 }, { "grad_norm": 0.38885024189949036, "learning_rate": 1.2646558257055257e-05, "loss": 0.0177, "step": 62410 }, { "grad_norm": 0.17004762589931488, "learning_rate": 1.263282221652482e-05, "loss": 0.0155, "step": 62420 }, { "grad_norm": 0.17125175893306732, "learning_rate": 1.2619092561023088e-05, "loss": 0.0171, "step": 62430 }, { "grad_norm": 0.2839864194393158, "learning_rate": 1.2605369292896036e-05, "loss": 0.0201, "step": 62440 }, { "grad_norm": 0.2864833474159241, "learning_rate": 1.2591652414488658e-05, "loss": 0.0206, "step": 62450 }, { "grad_norm": 0.2534332573413849, "learning_rate": 1.2577941928144732e-05, "loss": 0.0212, "step": 62460 }, { "grad_norm": 0.2588215172290802, "learning_rate": 1.256423783620706e-05, "loss": 0.024, "step": 62470 }, { "grad_norm": 0.555641770362854, "learning_rate": 1.2550540141017264e-05, "loss": 0.0192, "step": 62480 }, { "grad_norm": 0.1959281861782074, "learning_rate": 1.253684884491591e-05, "loss": 0.0194, "step": 62490 }, { "grad_norm": 0.24757391214370728, "learning_rate": 1.2523163950242483e-05, "loss": 0.017, "step": 62500 }, { "grad_norm": 0.19201558828353882, "learning_rate": 1.2509485459335313e-05, "loss": 0.019, "step": 62510 }, { "grad_norm": 0.355339914560318, "learning_rate": 1.2495813374531739e-05, "loss": 0.0223, "step": 62520 }, { "grad_norm": 0.3359787166118622, "learning_rate": 1.2482147698167907e-05, "loss": 0.0216, "step": 62530 }, { "grad_norm": 0.2929850220680237, "learning_rate": 1.2468488432578911e-05, "loss": 0.0218, "step": 62540 }, { "grad_norm": 0.2437979131937027, "learning_rate": 1.2454835580098761e-05, "loss": 0.0175, "step": 62550 }, { "grad_norm": 0.30613330006599426, "learning_rate": 1.2441189143060338e-05, "loss": 0.0213, "step": 62560 }, { "grad_norm": 0.23462149500846863, "learning_rate": 1.242754912379548e-05, "loss": 0.0159, "step": 62570 }, { "grad_norm": 0.29817038774490356, "learning_rate": 1.2413915524634844e-05, "loss": 0.0322, "step": 62580 }, { "grad_norm": 0.38462093472480774, "learning_rate": 1.2400288347908073e-05, "loss": 0.0225, "step": 62590 }, { "grad_norm": 0.3016878366470337, "learning_rate": 1.2386667595943663e-05, "loss": 0.018, "step": 62600 }, { "grad_norm": 0.2787228524684906, "learning_rate": 1.2373053271069035e-05, "loss": 0.0173, "step": 62610 }, { "grad_norm": 0.33242619037628174, "learning_rate": 1.2359445375610501e-05, "loss": 0.0225, "step": 62620 }, { "grad_norm": 0.1671520322561264, "learning_rate": 1.2345843911893301e-05, "loss": 0.0154, "step": 62630 }, { "grad_norm": 0.196891650557518, "learning_rate": 1.2332248882241498e-05, "loss": 0.0294, "step": 62640 }, { "grad_norm": 0.37178364396095276, "learning_rate": 1.2318660288978178e-05, "loss": 0.0219, "step": 62650 }, { "grad_norm": 0.19133926928043365, "learning_rate": 1.2305078134425213e-05, "loss": 0.0188, "step": 62660 }, { "grad_norm": 0.14918752014636993, "learning_rate": 1.2291502420903434e-05, "loss": 0.0169, "step": 62670 }, { "grad_norm": 0.2722042500972748, "learning_rate": 1.2277933150732567e-05, "loss": 0.0211, "step": 62680 }, { "grad_norm": 0.11589618027210236, "learning_rate": 1.2264370326231216e-05, "loss": 0.0163, "step": 62690 }, { "grad_norm": 0.2663857936859131, "learning_rate": 1.2250813949716927e-05, "loss": 0.0166, "step": 62700 }, { "grad_norm": 0.1716974973678589, "learning_rate": 1.2237264023506063e-05, "loss": 0.0177, "step": 62710 }, { "grad_norm": 0.3115706443786621, "learning_rate": 1.2223720549913987e-05, "loss": 0.0176, "step": 62720 }, { "grad_norm": 0.38018232583999634, "learning_rate": 1.221018353125487e-05, "loss": 0.0176, "step": 62730 }, { "grad_norm": 0.23443397879600525, "learning_rate": 1.2196652969841837e-05, "loss": 0.0232, "step": 62740 }, { "grad_norm": 0.34514182806015015, "learning_rate": 1.2183128867986904e-05, "loss": 0.0244, "step": 62750 }, { "grad_norm": 0.3663700520992279, "learning_rate": 1.2169611228000927e-05, "loss": 0.0193, "step": 62760 }, { "grad_norm": 0.24950002133846283, "learning_rate": 1.2156100052193752e-05, "loss": 0.0202, "step": 62770 }, { "grad_norm": 0.28103166818618774, "learning_rate": 1.214259534287403e-05, "loss": 0.0188, "step": 62780 }, { "grad_norm": 0.18211022019386292, "learning_rate": 1.2129097102349363e-05, "loss": 0.0156, "step": 62790 }, { "grad_norm": 0.2644881010055542, "learning_rate": 1.2115605332926227e-05, "loss": 0.0212, "step": 62800 }, { "grad_norm": 0.18657556176185608, "learning_rate": 1.2102120036910003e-05, "loss": 0.0256, "step": 62810 }, { "grad_norm": 0.3096879720687866, "learning_rate": 1.2088641216604956e-05, "loss": 0.0208, "step": 62820 }, { "grad_norm": 0.21339134871959686, "learning_rate": 1.2075168874314264e-05, "loss": 0.0156, "step": 62830 }, { "grad_norm": 0.33260706067085266, "learning_rate": 1.2061703012339942e-05, "loss": 0.0149, "step": 62840 }, { "grad_norm": 0.4215319752693176, "learning_rate": 1.204824363298297e-05, "loss": 0.0243, "step": 62850 }, { "grad_norm": 0.35204610228538513, "learning_rate": 1.2034790738543173e-05, "loss": 0.0213, "step": 62860 }, { "grad_norm": 0.19096893072128296, "learning_rate": 1.2021344331319284e-05, "loss": 0.0175, "step": 62870 }, { "grad_norm": 0.1783086210489273, "learning_rate": 1.200790441360894e-05, "loss": 0.0196, "step": 62880 }, { "grad_norm": 0.2450135201215744, "learning_rate": 1.199447098770864e-05, "loss": 0.0174, "step": 62890 }, { "grad_norm": 0.2715039551258087, "learning_rate": 1.198104405591381e-05, "loss": 0.0193, "step": 62900 }, { "grad_norm": 0.2283925861120224, "learning_rate": 1.1967623620518697e-05, "loss": 0.0132, "step": 62910 }, { "grad_norm": 0.33262085914611816, "learning_rate": 1.1954209683816554e-05, "loss": 0.0198, "step": 62920 }, { "grad_norm": 0.1550428867340088, "learning_rate": 1.1940802248099402e-05, "loss": 0.02, "step": 62930 }, { "grad_norm": 0.3489465117454529, "learning_rate": 1.1927401315658232e-05, "loss": 0.0217, "step": 62940 }, { "grad_norm": 0.34899118542671204, "learning_rate": 1.1914006888782898e-05, "loss": 0.0233, "step": 62950 }, { "grad_norm": 0.3107294738292694, "learning_rate": 1.1900618969762106e-05, "loss": 0.0184, "step": 62960 }, { "grad_norm": 0.20435158908367157, "learning_rate": 1.188723756088354e-05, "loss": 0.0173, "step": 62970 }, { "grad_norm": 0.2786814570426941, "learning_rate": 1.1873862664433672e-05, "loss": 0.0172, "step": 62980 }, { "grad_norm": 0.343618243932724, "learning_rate": 1.1860494282697927e-05, "loss": 0.0197, "step": 62990 }, { "grad_norm": 0.318092942237854, "learning_rate": 1.1847132417960588e-05, "loss": 0.0195, "step": 63000 }, { "grad_norm": 0.3419012427330017, "learning_rate": 1.1833777072504832e-05, "loss": 0.021, "step": 63010 }, { "grad_norm": 0.19338369369506836, "learning_rate": 1.1820428248612731e-05, "loss": 0.0218, "step": 63020 }, { "grad_norm": 0.3547983169555664, "learning_rate": 1.1807085948565245e-05, "loss": 0.0205, "step": 63030 }, { "grad_norm": 0.49589425325393677, "learning_rate": 1.1793750174642172e-05, "loss": 0.0198, "step": 63040 }, { "grad_norm": 0.2677927613258362, "learning_rate": 1.1780420929122254e-05, "loss": 0.0207, "step": 63050 }, { "grad_norm": 0.3992229700088501, "learning_rate": 1.176709821428309e-05, "loss": 0.0208, "step": 63060 }, { "grad_norm": 0.30208003520965576, "learning_rate": 1.1753782032401173e-05, "loss": 0.0238, "step": 63070 }, { "grad_norm": 0.3343414068222046, "learning_rate": 1.1740472385751866e-05, "loss": 0.0226, "step": 63080 }, { "grad_norm": 0.29902562499046326, "learning_rate": 1.172716927660943e-05, "loss": 0.016, "step": 63090 }, { "grad_norm": 0.3758874237537384, "learning_rate": 1.1713872707247015e-05, "loss": 0.024, "step": 63100 }, { "grad_norm": 0.3667500615119934, "learning_rate": 1.1700582679936595e-05, "loss": 0.0164, "step": 63110 }, { "grad_norm": 0.2603797912597656, "learning_rate": 1.1687299196949136e-05, "loss": 0.0277, "step": 63120 }, { "grad_norm": 1.2017054557800293, "learning_rate": 1.1674022260554374e-05, "loss": 0.0199, "step": 63130 }, { "grad_norm": 0.32638660073280334, "learning_rate": 1.1660751873020987e-05, "loss": 0.0145, "step": 63140 }, { "grad_norm": 0.35532087087631226, "learning_rate": 1.1647488036616538e-05, "loss": 0.0188, "step": 63150 }, { "grad_norm": 0.19821672141551971, "learning_rate": 1.1634230753607417e-05, "loss": 0.0231, "step": 63160 }, { "grad_norm": 0.35529887676239014, "learning_rate": 1.1620980026258982e-05, "loss": 0.0182, "step": 63170 }, { "grad_norm": 0.34488052129745483, "learning_rate": 1.1607735856835373e-05, "loss": 0.0173, "step": 63180 }, { "grad_norm": 0.4032881259918213, "learning_rate": 1.1594498247599677e-05, "loss": 0.0245, "step": 63190 }, { "grad_norm": 0.23782877624034882, "learning_rate": 1.158126720081384e-05, "loss": 0.0239, "step": 63200 }, { "grad_norm": 0.42191755771636963, "learning_rate": 1.156804271873868e-05, "loss": 0.0209, "step": 63210 }, { "grad_norm": 0.25876227021217346, "learning_rate": 1.1554824803633924e-05, "loss": 0.0225, "step": 63220 }, { "grad_norm": 0.23846375942230225, "learning_rate": 1.154161345775811e-05, "loss": 0.0156, "step": 63230 }, { "grad_norm": 0.14725498855113983, "learning_rate": 1.1528408683368724e-05, "loss": 0.0191, "step": 63240 }, { "grad_norm": 0.36384645104408264, "learning_rate": 1.1515210482722088e-05, "loss": 0.0242, "step": 63250 }, { "grad_norm": 0.23471485078334808, "learning_rate": 1.150201885807342e-05, "loss": 0.0245, "step": 63260 }, { "grad_norm": 0.24047011137008667, "learning_rate": 1.1488833811676807e-05, "loss": 0.0151, "step": 63270 }, { "grad_norm": 0.33001795411109924, "learning_rate": 1.1475655345785213e-05, "loss": 0.019, "step": 63280 }, { "grad_norm": 0.27108779549598694, "learning_rate": 1.1462483462650481e-05, "loss": 0.0171, "step": 63290 }, { "grad_norm": 0.3482268750667572, "learning_rate": 1.1449318164523331e-05, "loss": 0.02, "step": 63300 }, { "grad_norm": 0.46310633420944214, "learning_rate": 1.1436159453653334e-05, "loss": 0.0178, "step": 63310 }, { "grad_norm": 0.4010215699672699, "learning_rate": 1.1423007332288955e-05, "loss": 0.0263, "step": 63320 }, { "grad_norm": 0.3157312870025635, "learning_rate": 1.1409861802677546e-05, "loss": 0.0219, "step": 63330 }, { "grad_norm": 0.24264831840991974, "learning_rate": 1.1396722867065313e-05, "loss": 0.0157, "step": 63340 }, { "grad_norm": 0.4245418310165405, "learning_rate": 1.1383590527697352e-05, "loss": 0.0218, "step": 63350 }, { "grad_norm": 0.3328368663787842, "learning_rate": 1.137046478681758e-05, "loss": 0.0153, "step": 63360 }, { "grad_norm": 0.36298415064811707, "learning_rate": 1.1357345646668888e-05, "loss": 0.0332, "step": 63370 }, { "grad_norm": 0.16707417368888855, "learning_rate": 1.1344233109492924e-05, "loss": 0.0192, "step": 63380 }, { "grad_norm": 0.7434853911399841, "learning_rate": 1.1331127177530292e-05, "loss": 0.0217, "step": 63390 }, { "grad_norm": 0.21766601502895355, "learning_rate": 1.1318027853020441e-05, "loss": 0.0159, "step": 63400 }, { "grad_norm": 0.21429267525672913, "learning_rate": 1.1304935138201645e-05, "loss": 0.0185, "step": 63410 }, { "grad_norm": 0.36526086926460266, "learning_rate": 1.1291849035311153e-05, "loss": 0.0195, "step": 63420 }, { "grad_norm": 0.285646915435791, "learning_rate": 1.1278769546584972e-05, "loss": 0.0207, "step": 63430 }, { "grad_norm": 0.31568440794944763, "learning_rate": 1.1265696674258052e-05, "loss": 0.0151, "step": 63440 }, { "grad_norm": 0.28135988116264343, "learning_rate": 1.1252630420564186e-05, "loss": 0.0176, "step": 63450 }, { "grad_norm": 0.31719106435775757, "learning_rate": 1.1239570787736036e-05, "loss": 0.0175, "step": 63460 }, { "grad_norm": 0.2248547524213791, "learning_rate": 1.1226517778005135e-05, "loss": 0.017, "step": 63470 }, { "grad_norm": 0.26939910650253296, "learning_rate": 1.1213471393601893e-05, "loss": 0.0157, "step": 63480 }, { "grad_norm": 0.33226311206817627, "learning_rate": 1.1200431636755587e-05, "loss": 0.0267, "step": 63490 }, { "grad_norm": 0.33916565775871277, "learning_rate": 1.1187398509694336e-05, "loss": 0.0261, "step": 63500 }, { "grad_norm": 0.33884304761886597, "learning_rate": 1.1174372014645146e-05, "loss": 0.0207, "step": 63510 }, { "grad_norm": 0.20106667280197144, "learning_rate": 1.1161352153833899e-05, "loss": 0.0293, "step": 63520 }, { "grad_norm": 0.24830682575702667, "learning_rate": 1.1148338929485325e-05, "loss": 0.0198, "step": 63530 }, { "grad_norm": 0.39422282576560974, "learning_rate": 1.113533234382304e-05, "loss": 0.0159, "step": 63540 }, { "grad_norm": 0.2793957591056824, "learning_rate": 1.1122332399069513e-05, "loss": 0.0178, "step": 63550 }, { "grad_norm": 0.2981221675872803, "learning_rate": 1.1109339097446047e-05, "loss": 0.0199, "step": 63560 }, { "grad_norm": 0.41505658626556396, "learning_rate": 1.1096352441172897e-05, "loss": 0.029, "step": 63570 }, { "grad_norm": 0.3159962296485901, "learning_rate": 1.1083372432469086e-05, "loss": 0.0168, "step": 63580 }, { "grad_norm": 0.21698330342769623, "learning_rate": 1.107039907355255e-05, "loss": 0.0233, "step": 63590 }, { "grad_norm": 0.2930280268192291, "learning_rate": 1.1057432366640103e-05, "loss": 0.0167, "step": 63600 }, { "grad_norm": 0.4770091474056244, "learning_rate": 1.1044472313947352e-05, "loss": 0.0252, "step": 63610 }, { "grad_norm": 0.45692241191864014, "learning_rate": 1.1031518917688877e-05, "loss": 0.0209, "step": 63620 }, { "grad_norm": 0.19518035650253296, "learning_rate": 1.1018572180078007e-05, "loss": 0.027, "step": 63630 }, { "grad_norm": 0.34773123264312744, "learning_rate": 1.1005632103327018e-05, "loss": 0.0158, "step": 63640 }, { "grad_norm": 0.32426488399505615, "learning_rate": 1.0992698689646996e-05, "loss": 0.02, "step": 63650 }, { "grad_norm": 0.2368328720331192, "learning_rate": 1.0979771941247919e-05, "loss": 0.0214, "step": 63660 }, { "grad_norm": 0.2839478850364685, "learning_rate": 1.0966851860338611e-05, "loss": 0.017, "step": 63670 }, { "grad_norm": 0.18291883170604706, "learning_rate": 1.0953938449126766e-05, "loss": 0.021, "step": 63680 }, { "grad_norm": 0.42377111315727234, "learning_rate": 1.0941031709818933e-05, "loss": 0.0319, "step": 63690 }, { "grad_norm": 0.287616491317749, "learning_rate": 1.0928131644620509e-05, "loss": 0.0297, "step": 63700 }, { "grad_norm": 0.5021877884864807, "learning_rate": 1.0915238255735766e-05, "loss": 0.0217, "step": 63710 }, { "grad_norm": 0.17046687006950378, "learning_rate": 1.0902351545367833e-05, "loss": 0.017, "step": 63720 }, { "grad_norm": 0.36717700958251953, "learning_rate": 1.0889471515718702e-05, "loss": 0.0193, "step": 63730 }, { "grad_norm": 0.26448675990104675, "learning_rate": 1.087659816898921e-05, "loss": 0.0137, "step": 63740 }, { "grad_norm": 0.3811897933483124, "learning_rate": 1.0863731507379082e-05, "loss": 0.0183, "step": 63750 }, { "grad_norm": 0.389693021774292, "learning_rate": 1.0850871533086827e-05, "loss": 0.0241, "step": 63760 }, { "grad_norm": 0.20819103717803955, "learning_rate": 1.0838018248309927e-05, "loss": 0.0201, "step": 63770 }, { "grad_norm": 0.3447091281414032, "learning_rate": 1.0825171655244615e-05, "loss": 0.022, "step": 63780 }, { "grad_norm": 0.27559342980384827, "learning_rate": 1.0812331756086025e-05, "loss": 0.0223, "step": 63790 }, { "grad_norm": 0.2756491005420685, "learning_rate": 1.079949855302817e-05, "loss": 0.0185, "step": 63800 }, { "grad_norm": 0.40948206186294556, "learning_rate": 1.0786672048263852e-05, "loss": 0.0239, "step": 63810 }, { "grad_norm": 0.34257009625434875, "learning_rate": 1.0773852243984817e-05, "loss": 0.0221, "step": 63820 }, { "grad_norm": 0.28587207198143005, "learning_rate": 1.0761039142381586e-05, "loss": 0.0222, "step": 63830 }, { "grad_norm": 0.33928874135017395, "learning_rate": 1.0748232745643577e-05, "loss": 0.019, "step": 63840 }, { "grad_norm": 0.2653891146183014, "learning_rate": 1.0735433055959055e-05, "loss": 0.0252, "step": 63850 }, { "grad_norm": 0.37557852268218994, "learning_rate": 1.0722640075515133e-05, "loss": 0.0184, "step": 63860 }, { "grad_norm": 0.23571135103702545, "learning_rate": 1.0709853806497795e-05, "loss": 0.0169, "step": 63870 }, { "grad_norm": 0.26315221190452576, "learning_rate": 1.0697074251091831e-05, "loss": 0.0184, "step": 63880 }, { "grad_norm": 0.32368189096450806, "learning_rate": 1.0684301411480962e-05, "loss": 0.0172, "step": 63890 }, { "grad_norm": 0.25781822204589844, "learning_rate": 1.067153528984769e-05, "loss": 0.0154, "step": 63900 }, { "grad_norm": 0.2554537057876587, "learning_rate": 1.0658775888373395e-05, "loss": 0.0186, "step": 63910 }, { "grad_norm": 0.2521118223667145, "learning_rate": 1.0646023209238314e-05, "loss": 0.0266, "step": 63920 }, { "grad_norm": 0.3893734812736511, "learning_rate": 1.0633277254621537e-05, "loss": 0.0194, "step": 63930 }, { "grad_norm": 0.11686151474714279, "learning_rate": 1.0620538026700994e-05, "loss": 0.0149, "step": 63940 }, { "grad_norm": 0.28626129031181335, "learning_rate": 1.0607805527653486e-05, "loss": 0.0227, "step": 63950 }, { "grad_norm": 0.3087295889854431, "learning_rate": 1.059507975965462e-05, "loss": 0.0183, "step": 63960 }, { "grad_norm": 0.21177101135253906, "learning_rate": 1.0582360724878898e-05, "loss": 0.0151, "step": 63970 }, { "grad_norm": 0.3145795464515686, "learning_rate": 1.0569648425499651e-05, "loss": 0.0216, "step": 63980 }, { "grad_norm": 0.22305364906787872, "learning_rate": 1.0556942863689063e-05, "loss": 0.0166, "step": 63990 }, { "grad_norm": 0.308268666267395, "learning_rate": 1.054424404161819e-05, "loss": 0.0198, "step": 64000 }, { "grad_norm": 0.22635050117969513, "learning_rate": 1.053155196145686e-05, "loss": 0.0184, "step": 64010 }, { "grad_norm": 0.3533676862716675, "learning_rate": 1.0518866625373863e-05, "loss": 0.0168, "step": 64020 }, { "grad_norm": 0.2282438427209854, "learning_rate": 1.0506188035536735e-05, "loss": 0.0196, "step": 64030 }, { "grad_norm": 0.3428679406642914, "learning_rate": 1.0493516194111919e-05, "loss": 0.016, "step": 64040 }, { "grad_norm": 0.4412238299846649, "learning_rate": 1.0480851103264688e-05, "loss": 0.02, "step": 64050 }, { "grad_norm": 0.14694438874721527, "learning_rate": 1.046819276515913e-05, "loss": 0.0169, "step": 64060 }, { "grad_norm": 0.29241809248924255, "learning_rate": 1.0455541181958256e-05, "loss": 0.0165, "step": 64070 }, { "grad_norm": 0.18179337680339813, "learning_rate": 1.0442896355823822e-05, "loss": 0.0193, "step": 64080 }, { "grad_norm": 0.37616369128227234, "learning_rate": 1.0430258288916539e-05, "loss": 0.0172, "step": 64090 }, { "grad_norm": 0.19932381808757782, "learning_rate": 1.0417626983395868e-05, "loss": 0.0157, "step": 64100 }, { "grad_norm": 0.21545374393463135, "learning_rate": 1.0405002441420165e-05, "loss": 0.0174, "step": 64110 }, { "grad_norm": 0.36375948786735535, "learning_rate": 1.039238466514662e-05, "loss": 0.0206, "step": 64120 }, { "grad_norm": 0.23700354993343353, "learning_rate": 1.0379773656731262e-05, "loss": 0.0243, "step": 64130 }, { "grad_norm": 0.26611679792404175, "learning_rate": 1.0367169418328986e-05, "loss": 0.0159, "step": 64140 }, { "grad_norm": 0.3484284281730652, "learning_rate": 1.0354571952093484e-05, "loss": 0.0203, "step": 64150 }, { "grad_norm": 0.24897494912147522, "learning_rate": 1.034198126017733e-05, "loss": 0.0173, "step": 64160 }, { "grad_norm": 0.2654881775379181, "learning_rate": 1.0329397344731928e-05, "loss": 0.0177, "step": 64170 }, { "grad_norm": 0.31858596205711365, "learning_rate": 1.0316820207907524e-05, "loss": 0.0219, "step": 64180 }, { "grad_norm": 0.27442270517349243, "learning_rate": 1.0304249851853215e-05, "loss": 0.0224, "step": 64190 }, { "grad_norm": 0.3312547206878662, "learning_rate": 1.0291686278716933e-05, "loss": 0.02, "step": 64200 }, { "grad_norm": 0.5779716372489929, "learning_rate": 1.0279129490645418e-05, "loss": 0.0196, "step": 64210 }, { "grad_norm": 0.2640024721622467, "learning_rate": 1.0266579489784328e-05, "loss": 0.0157, "step": 64220 }, { "grad_norm": 0.2973218858242035, "learning_rate": 1.0254036278278084e-05, "loss": 0.0222, "step": 64230 }, { "grad_norm": 0.21729709208011627, "learning_rate": 1.0241499858269982e-05, "loss": 0.0133, "step": 64240 }, { "grad_norm": 0.2561582028865814, "learning_rate": 1.0228970231902169e-05, "loss": 0.0153, "step": 64250 }, { "grad_norm": 0.2968396246433258, "learning_rate": 1.0216447401315582e-05, "loss": 0.0203, "step": 64260 }, { "grad_norm": 0.362815260887146, "learning_rate": 1.020393136865007e-05, "loss": 0.0157, "step": 64270 }, { "grad_norm": 0.3795025944709778, "learning_rate": 1.0191422136044242e-05, "loss": 0.0265, "step": 64280 }, { "grad_norm": 0.22314117848873138, "learning_rate": 1.017891970563563e-05, "loss": 0.018, "step": 64290 }, { "grad_norm": 0.24423915147781372, "learning_rate": 1.0166424079560516e-05, "loss": 0.0257, "step": 64300 }, { "grad_norm": 0.23398879170417786, "learning_rate": 1.0153935259954078e-05, "loss": 0.0145, "step": 64310 }, { "grad_norm": 0.22607266902923584, "learning_rate": 1.0141453248950311e-05, "loss": 0.0196, "step": 64320 }, { "grad_norm": 0.3653962314128876, "learning_rate": 1.0128978048682054e-05, "loss": 0.0217, "step": 64330 }, { "grad_norm": 0.3469941020011902, "learning_rate": 1.0116509661280982e-05, "loss": 0.022, "step": 64340 }, { "grad_norm": 0.11473428457975388, "learning_rate": 1.0104048088877576e-05, "loss": 0.0164, "step": 64350 }, { "grad_norm": 0.1568479835987091, "learning_rate": 1.0091593333601201e-05, "loss": 0.0183, "step": 64360 }, { "grad_norm": 0.27976658940315247, "learning_rate": 1.0079145397580031e-05, "loss": 0.0159, "step": 64370 }, { "grad_norm": 0.3734692633152008, "learning_rate": 1.006670428294107e-05, "loss": 0.0309, "step": 64380 }, { "grad_norm": 0.20443077385425568, "learning_rate": 1.0054269991810166e-05, "loss": 0.0159, "step": 64390 }, { "grad_norm": 0.2780006527900696, "learning_rate": 1.0041842526312024e-05, "loss": 0.018, "step": 64400 }, { "grad_norm": 0.20832614600658417, "learning_rate": 1.0029421888570101e-05, "loss": 0.017, "step": 64410 }, { "grad_norm": 0.19643425941467285, "learning_rate": 1.0017008080706813e-05, "loss": 0.0267, "step": 64420 }, { "grad_norm": 0.2429143190383911, "learning_rate": 1.0004601104843287e-05, "loss": 0.0185, "step": 64430 }, { "grad_norm": 0.26336219906806946, "learning_rate": 9.992200963099562e-06, "loss": 0.0227, "step": 64440 }, { "grad_norm": 0.3578222692012787, "learning_rate": 9.979807657594486e-06, "loss": 0.0165, "step": 64450 }, { "grad_norm": 0.3792129158973694, "learning_rate": 9.967421190445703e-06, "loss": 0.0193, "step": 64460 }, { "grad_norm": 0.25594159960746765, "learning_rate": 9.955041563769769e-06, "loss": 0.0189, "step": 64470 }, { "grad_norm": 0.192132830619812, "learning_rate": 9.942668779681974e-06, "loss": 0.0232, "step": 64480 }, { "grad_norm": 0.2245355248451233, "learning_rate": 9.930302840296541e-06, "loss": 0.0142, "step": 64490 }, { "grad_norm": 0.141597718000412, "learning_rate": 9.917943747726426e-06, "loss": 0.0159, "step": 64500 }, { "grad_norm": 0.2409479171037674, "learning_rate": 9.905591504083484e-06, "loss": 0.0188, "step": 64510 }, { "grad_norm": 0.3558749854564667, "learning_rate": 9.893246111478382e-06, "loss": 0.0146, "step": 64520 }, { "grad_norm": 0.36257439851760864, "learning_rate": 9.88090757202057e-06, "loss": 0.0165, "step": 64530 }, { "grad_norm": 0.174039825797081, "learning_rate": 9.868575887818421e-06, "loss": 0.023, "step": 64540 }, { "grad_norm": 0.17252086102962494, "learning_rate": 9.856251060979044e-06, "loss": 0.0172, "step": 64550 }, { "grad_norm": 0.32776308059692383, "learning_rate": 9.843933093608426e-06, "loss": 0.0267, "step": 64560 }, { "grad_norm": 0.3521707355976105, "learning_rate": 9.831621987811368e-06, "loss": 0.0267, "step": 64570 }, { "grad_norm": 0.2348358929157257, "learning_rate": 9.819317745691509e-06, "loss": 0.0164, "step": 64580 }, { "grad_norm": 0.3796221911907196, "learning_rate": 9.8070203693513e-06, "loss": 0.0177, "step": 64590 }, { "grad_norm": 0.3245092034339905, "learning_rate": 9.794729860892048e-06, "loss": 0.0194, "step": 64600 }, { "grad_norm": 0.30645042657852173, "learning_rate": 9.782446222413827e-06, "loss": 0.0248, "step": 64610 }, { "grad_norm": 0.2846423089504242, "learning_rate": 9.770169456015598e-06, "loss": 0.0209, "step": 64620 }, { "grad_norm": 0.3159308135509491, "learning_rate": 9.75789956379512e-06, "loss": 0.021, "step": 64630 }, { "grad_norm": 0.2460821270942688, "learning_rate": 9.74563654784898e-06, "loss": 0.0183, "step": 64640 }, { "grad_norm": 0.2050757110118866, "learning_rate": 9.733380410272596e-06, "loss": 0.013, "step": 64650 }, { "grad_norm": 0.3419475257396698, "learning_rate": 9.721131153160207e-06, "loss": 0.0159, "step": 64660 }, { "grad_norm": 0.2955856919288635, "learning_rate": 9.708888778604886e-06, "loss": 0.021, "step": 64670 }, { "grad_norm": 0.30405306816101074, "learning_rate": 9.696653288698488e-06, "loss": 0.0194, "step": 64680 }, { "grad_norm": 0.2971068024635315, "learning_rate": 9.684424685531763e-06, "loss": 0.0243, "step": 64690 }, { "grad_norm": 0.27921199798583984, "learning_rate": 9.672202971194216e-06, "loss": 0.0224, "step": 64700 }, { "grad_norm": 0.185531347990036, "learning_rate": 9.659988147774213e-06, "loss": 0.023, "step": 64710 }, { "grad_norm": 0.21496616303920746, "learning_rate": 9.647780217358942e-06, "loss": 0.0201, "step": 64720 }, { "grad_norm": 0.15386682748794556, "learning_rate": 9.635579182034376e-06, "loss": 0.0183, "step": 64730 }, { "grad_norm": 0.3780650496482849, "learning_rate": 9.623385043885386e-06, "loss": 0.0266, "step": 64740 }, { "grad_norm": 0.26705193519592285, "learning_rate": 9.61119780499557e-06, "loss": 0.0122, "step": 64750 }, { "grad_norm": 0.30307498574256897, "learning_rate": 9.599017467447418e-06, "loss": 0.0171, "step": 64760 }, { "grad_norm": 0.2470817118883133, "learning_rate": 9.586844033322206e-06, "loss": 0.02, "step": 64770 }, { "grad_norm": 0.3225655257701874, "learning_rate": 9.574677504700052e-06, "loss": 0.0214, "step": 64780 }, { "grad_norm": 0.21617595851421356, "learning_rate": 9.562517883659877e-06, "loss": 0.0257, "step": 64790 }, { "grad_norm": 0.25503072142601013, "learning_rate": 9.55036517227944e-06, "loss": 0.0173, "step": 64800 }, { "grad_norm": 0.20697706937789917, "learning_rate": 9.538219372635282e-06, "loss": 0.0191, "step": 64810 }, { "grad_norm": 0.18100784718990326, "learning_rate": 9.526080486802802e-06, "loss": 0.0147, "step": 64820 }, { "grad_norm": 0.26491791009902954, "learning_rate": 9.513948516856203e-06, "loss": 0.0187, "step": 64830 }, { "grad_norm": 0.21749496459960938, "learning_rate": 9.501823464868504e-06, "loss": 0.0139, "step": 64840 }, { "grad_norm": 0.3500630855560303, "learning_rate": 9.489705332911547e-06, "loss": 0.0184, "step": 64850 }, { "grad_norm": 0.3162156045436859, "learning_rate": 9.477594123055994e-06, "loss": 0.0206, "step": 64860 }, { "grad_norm": 0.20083928108215332, "learning_rate": 9.465489837371321e-06, "loss": 0.0251, "step": 64870 }, { "grad_norm": 0.3826524317264557, "learning_rate": 9.453392477925794e-06, "loss": 0.0188, "step": 64880 }, { "grad_norm": 0.22969578206539154, "learning_rate": 9.441302046786566e-06, "loss": 0.0172, "step": 64890 }, { "grad_norm": 0.33501943945884705, "learning_rate": 9.429218546019519e-06, "loss": 0.0174, "step": 64900 }, { "grad_norm": 0.30152320861816406, "learning_rate": 9.41714197768941e-06, "loss": 0.0256, "step": 64910 }, { "grad_norm": 0.29791900515556335, "learning_rate": 9.405072343859805e-06, "loss": 0.0196, "step": 64920 }, { "grad_norm": 0.3095923662185669, "learning_rate": 9.393009646593043e-06, "loss": 0.0139, "step": 64930 }, { "grad_norm": 0.24485254287719727, "learning_rate": 9.38095388795035e-06, "loss": 0.0207, "step": 64940 }, { "grad_norm": 0.3078668713569641, "learning_rate": 9.36890506999169e-06, "loss": 0.0202, "step": 64950 }, { "grad_norm": 0.20309370756149292, "learning_rate": 9.356863194775894e-06, "loss": 0.0231, "step": 64960 }, { "grad_norm": 0.34675729274749756, "learning_rate": 9.344828264360583e-06, "loss": 0.0222, "step": 64970 }, { "grad_norm": 0.24313172698020935, "learning_rate": 9.332800280802201e-06, "loss": 0.0226, "step": 64980 }, { "grad_norm": 0.23148970305919647, "learning_rate": 9.32077924615602e-06, "loss": 0.0147, "step": 64990 }, { "grad_norm": 0.27491965889930725, "learning_rate": 9.308765162476063e-06, "loss": 0.0159, "step": 65000 }, { "grad_norm": 0.26843422651290894, "learning_rate": 9.296758031815239e-06, "loss": 0.0219, "step": 65010 }, { "grad_norm": 0.29656675457954407, "learning_rate": 9.284757856225229e-06, "loss": 0.0142, "step": 65020 }, { "grad_norm": 0.20335957407951355, "learning_rate": 9.272764637756538e-06, "loss": 0.0184, "step": 65030 }, { "grad_norm": 0.2589907944202423, "learning_rate": 9.260778378458479e-06, "loss": 0.0219, "step": 65040 }, { "grad_norm": 0.13399432599544525, "learning_rate": 9.248799080379172e-06, "loss": 0.0196, "step": 65050 }, { "grad_norm": 0.24345239996910095, "learning_rate": 9.236826745565558e-06, "loss": 0.0196, "step": 65060 }, { "grad_norm": 0.26818975806236267, "learning_rate": 9.224861376063388e-06, "loss": 0.0177, "step": 65070 }, { "grad_norm": 0.16023363173007965, "learning_rate": 9.212902973917192e-06, "loss": 0.014, "step": 65080 }, { "grad_norm": 0.23116451501846313, "learning_rate": 9.20095154117035e-06, "loss": 0.0139, "step": 65090 }, { "grad_norm": 0.1852494776248932, "learning_rate": 9.189007079865036e-06, "loss": 0.0171, "step": 65100 }, { "grad_norm": 0.22316795587539673, "learning_rate": 9.177069592042226e-06, "loss": 0.0178, "step": 65110 }, { "grad_norm": 0.3743717670440674, "learning_rate": 9.165139079741724e-06, "loss": 0.0245, "step": 65120 }, { "grad_norm": 0.31380313634872437, "learning_rate": 9.153215545002098e-06, "loss": 0.0287, "step": 65130 }, { "grad_norm": 0.2088330090045929, "learning_rate": 9.141298989860798e-06, "loss": 0.026, "step": 65140 }, { "grad_norm": 0.2843937873840332, "learning_rate": 9.129389416353994e-06, "loss": 0.0174, "step": 65150 }, { "grad_norm": 0.36626070737838745, "learning_rate": 9.11748682651673e-06, "loss": 0.0169, "step": 65160 }, { "grad_norm": 0.3580969572067261, "learning_rate": 9.105591222382837e-06, "loss": 0.0192, "step": 65170 }, { "grad_norm": 0.1832444816827774, "learning_rate": 9.093702605984915e-06, "loss": 0.0144, "step": 65180 }, { "grad_norm": 0.5055596828460693, "learning_rate": 9.081820979354455e-06, "loss": 0.0206, "step": 65190 }, { "grad_norm": 0.23820778727531433, "learning_rate": 9.069946344521663e-06, "loss": 0.0176, "step": 65200 }, { "grad_norm": 0.26829594373703003, "learning_rate": 9.058078703515598e-06, "loss": 0.0196, "step": 65210 }, { "grad_norm": 0.2490549236536026, "learning_rate": 9.046218058364125e-06, "loss": 0.0225, "step": 65220 }, { "grad_norm": 0.3737298846244812, "learning_rate": 9.034364411093893e-06, "loss": 0.015, "step": 65230 }, { "grad_norm": 0.42336148023605347, "learning_rate": 9.022517763730371e-06, "loss": 0.0208, "step": 65240 }, { "grad_norm": 0.19154047966003418, "learning_rate": 9.010678118297827e-06, "loss": 0.0213, "step": 65250 }, { "grad_norm": 0.41150757670402527, "learning_rate": 8.998845476819345e-06, "loss": 0.0176, "step": 65260 }, { "grad_norm": 0.2615169882774353, "learning_rate": 8.987019841316773e-06, "loss": 0.0193, "step": 65270 }, { "grad_norm": 0.33814263343811035, "learning_rate": 8.975201213810802e-06, "loss": 0.0148, "step": 65280 }, { "grad_norm": 0.2876981496810913, "learning_rate": 8.963389596320915e-06, "loss": 0.0175, "step": 65290 }, { "grad_norm": 0.35616323351860046, "learning_rate": 8.951584990865391e-06, "loss": 0.0192, "step": 65300 }, { "grad_norm": 0.34323224425315857, "learning_rate": 8.939787399461319e-06, "loss": 0.0247, "step": 65310 }, { "grad_norm": 0.39513951539993286, "learning_rate": 8.927996824124591e-06, "loss": 0.0188, "step": 65320 }, { "grad_norm": 0.36850786209106445, "learning_rate": 8.916213266869854e-06, "loss": 0.0246, "step": 65330 }, { "grad_norm": 0.24228574335575104, "learning_rate": 8.904436729710658e-06, "loss": 0.0161, "step": 65340 }, { "grad_norm": 0.33034199476242065, "learning_rate": 8.892667214659245e-06, "loss": 0.0133, "step": 65350 }, { "grad_norm": 0.3372928500175476, "learning_rate": 8.880904723726713e-06, "loss": 0.0203, "step": 65360 }, { "grad_norm": 0.26531338691711426, "learning_rate": 8.869149258922971e-06, "loss": 0.0138, "step": 65370 }, { "grad_norm": 0.22367653250694275, "learning_rate": 8.857400822256662e-06, "loss": 0.0165, "step": 65380 }, { "grad_norm": 0.29545271396636963, "learning_rate": 8.845659415735324e-06, "loss": 0.0227, "step": 65390 }, { "grad_norm": 0.2013123333454132, "learning_rate": 8.83392504136521e-06, "loss": 0.0187, "step": 65400 }, { "grad_norm": 0.1679873764514923, "learning_rate": 8.822197701151407e-06, "loss": 0.0228, "step": 65410 }, { "grad_norm": 0.3146106004714966, "learning_rate": 8.810477397097804e-06, "loss": 0.0184, "step": 65420 }, { "grad_norm": 0.1478455513715744, "learning_rate": 8.798764131207077e-06, "loss": 0.02, "step": 65430 }, { "grad_norm": 0.15573816001415253, "learning_rate": 8.787057905480706e-06, "loss": 0.0157, "step": 65440 }, { "grad_norm": 0.2053576558828354, "learning_rate": 8.775358721918958e-06, "loss": 0.0165, "step": 65450 }, { "grad_norm": 0.1888885498046875, "learning_rate": 8.763666582520923e-06, "loss": 0.0166, "step": 65460 }, { "grad_norm": 0.33539608120918274, "learning_rate": 8.751981489284445e-06, "loss": 0.0179, "step": 65470 }, { "grad_norm": 0.31354576349258423, "learning_rate": 8.740303444206188e-06, "loss": 0.0201, "step": 65480 }, { "grad_norm": 0.4018164873123169, "learning_rate": 8.72863244928162e-06, "loss": 0.0147, "step": 65490 }, { "grad_norm": 0.28370046615600586, "learning_rate": 8.716968506504991e-06, "loss": 0.0244, "step": 65500 }, { "grad_norm": 0.31355705857276917, "learning_rate": 8.70531161786935e-06, "loss": 0.0145, "step": 65510 }, { "grad_norm": 0.3347127139568329, "learning_rate": 8.693661785366558e-06, "loss": 0.0288, "step": 65520 }, { "grad_norm": 0.29530468583106995, "learning_rate": 8.682019010987208e-06, "loss": 0.0286, "step": 65530 }, { "grad_norm": 0.34150075912475586, "learning_rate": 8.670383296720786e-06, "loss": 0.0157, "step": 65540 }, { "grad_norm": 0.27830207347869873, "learning_rate": 8.658754644555478e-06, "loss": 0.0178, "step": 65550 }, { "grad_norm": 0.19586063921451569, "learning_rate": 8.647133056478313e-06, "loss": 0.0193, "step": 65560 }, { "grad_norm": 0.2856370210647583, "learning_rate": 8.635518534475123e-06, "loss": 0.0215, "step": 65570 }, { "grad_norm": 0.17406922578811646, "learning_rate": 8.623911080530467e-06, "loss": 0.0156, "step": 65580 }, { "grad_norm": 0.26524218916893005, "learning_rate": 8.61231069662779e-06, "loss": 0.0156, "step": 65590 }, { "grad_norm": 0.39637282490730286, "learning_rate": 8.600717384749252e-06, "loss": 0.0172, "step": 65600 }, { "grad_norm": 0.2788766920566559, "learning_rate": 8.58913114687584e-06, "loss": 0.0156, "step": 65610 }, { "grad_norm": 0.36985984444618225, "learning_rate": 8.57755198498732e-06, "loss": 0.0191, "step": 65620 }, { "grad_norm": 0.2526796758174896, "learning_rate": 8.565979901062265e-06, "loss": 0.023, "step": 65630 }, { "grad_norm": 0.5041442513465881, "learning_rate": 8.554414897078033e-06, "loss": 0.0185, "step": 65640 }, { "grad_norm": 0.2800056040287018, "learning_rate": 8.542856975010727e-06, "loss": 0.0166, "step": 65650 }, { "grad_norm": 0.21415668725967407, "learning_rate": 8.531306136835337e-06, "loss": 0.0136, "step": 65660 }, { "grad_norm": 0.20592908561229706, "learning_rate": 8.519762384525548e-06, "loss": 0.021, "step": 65670 }, { "grad_norm": 0.41969770193099976, "learning_rate": 8.508225720053875e-06, "loss": 0.0191, "step": 65680 }, { "grad_norm": 0.2612782418727875, "learning_rate": 8.496696145391625e-06, "loss": 0.0173, "step": 65690 }, { "grad_norm": 0.14345994591712952, "learning_rate": 8.485173662508889e-06, "loss": 0.0134, "step": 65700 }, { "grad_norm": 0.32275593280792236, "learning_rate": 8.473658273374536e-06, "loss": 0.0147, "step": 65710 }, { "grad_norm": 0.26084643602371216, "learning_rate": 8.462149979956253e-06, "loss": 0.0108, "step": 65720 }, { "grad_norm": 0.1792377382516861, "learning_rate": 8.450648784220461e-06, "loss": 0.0176, "step": 65730 }, { "grad_norm": 0.26809027791023254, "learning_rate": 8.439154688132417e-06, "loss": 0.017, "step": 65740 }, { "grad_norm": 0.35825395584106445, "learning_rate": 8.427667693656143e-06, "loss": 0.0164, "step": 65750 }, { "grad_norm": 0.27489200234413147, "learning_rate": 8.416187802754454e-06, "loss": 0.0179, "step": 65760 }, { "grad_norm": 0.1901341825723648, "learning_rate": 8.404715017388965e-06, "loss": 0.0149, "step": 65770 }, { "grad_norm": 0.43554428219795227, "learning_rate": 8.393249339520015e-06, "loss": 0.0238, "step": 65780 }, { "grad_norm": 0.3228898048400879, "learning_rate": 8.381790771106834e-06, "loss": 0.0157, "step": 65790 }, { "grad_norm": 0.4450990557670593, "learning_rate": 8.370339314107339e-06, "loss": 0.0133, "step": 65800 }, { "grad_norm": 0.268544465303421, "learning_rate": 8.358894970478281e-06, "loss": 0.0151, "step": 65810 }, { "grad_norm": 0.2287135273218155, "learning_rate": 8.347457742175196e-06, "loss": 0.0168, "step": 65820 }, { "grad_norm": 0.40986108779907227, "learning_rate": 8.33602763115236e-06, "loss": 0.0155, "step": 65830 }, { "grad_norm": 0.4323256313800812, "learning_rate": 8.324604639362916e-06, "loss": 0.0281, "step": 65840 }, { "grad_norm": 0.37251368165016174, "learning_rate": 8.31318876875869e-06, "loss": 0.0171, "step": 65850 }, { "grad_norm": 0.27116385102272034, "learning_rate": 8.30178002129039e-06, "loss": 0.0171, "step": 65860 }, { "grad_norm": 0.455375999212265, "learning_rate": 8.290378398907423e-06, "loss": 0.022, "step": 65870 }, { "grad_norm": 0.2689781188964844, "learning_rate": 8.278983903558029e-06, "loss": 0.0267, "step": 65880 }, { "grad_norm": 0.30658456683158875, "learning_rate": 8.26759653718921e-06, "loss": 0.0203, "step": 65890 }, { "grad_norm": 0.2484823316335678, "learning_rate": 8.25621630174676e-06, "loss": 0.0228, "step": 65900 }, { "grad_norm": 0.28515806794166565, "learning_rate": 8.244843199175261e-06, "loss": 0.0229, "step": 65910 }, { "grad_norm": 0.252741277217865, "learning_rate": 8.233477231418046e-06, "loss": 0.0136, "step": 65920 }, { "grad_norm": 0.29033148288726807, "learning_rate": 8.22211840041725e-06, "loss": 0.0222, "step": 65930 }, { "grad_norm": 0.3716675043106079, "learning_rate": 8.210766708113792e-06, "loss": 0.0222, "step": 65940 }, { "grad_norm": 0.5560818910598755, "learning_rate": 8.199422156447367e-06, "loss": 0.0195, "step": 65950 }, { "grad_norm": 0.35184672474861145, "learning_rate": 8.188084747356451e-06, "loss": 0.022, "step": 65960 }, { "grad_norm": 0.19898709654808044, "learning_rate": 8.176754482778299e-06, "loss": 0.0177, "step": 65970 }, { "grad_norm": 0.2536811828613281, "learning_rate": 8.165431364648918e-06, "loss": 0.0159, "step": 65980 }, { "grad_norm": 0.21092502772808075, "learning_rate": 8.154115394903162e-06, "loss": 0.0182, "step": 65990 }, { "grad_norm": 0.32898131012916565, "learning_rate": 8.142806575474582e-06, "loss": 0.0159, "step": 66000 }, { "grad_norm": 0.19782309234142303, "learning_rate": 8.131504908295562e-06, "loss": 0.0133, "step": 66010 }, { "grad_norm": 0.34597957134246826, "learning_rate": 8.120210395297262e-06, "loss": 0.014, "step": 66020 }, { "grad_norm": 0.22260722517967224, "learning_rate": 8.108923038409565e-06, "loss": 0.0199, "step": 66030 }, { "grad_norm": 0.382503867149353, "learning_rate": 8.097642839561226e-06, "loss": 0.0176, "step": 66040 }, { "grad_norm": 0.24120327830314636, "learning_rate": 8.08636980067966e-06, "loss": 0.0212, "step": 66050 }, { "grad_norm": 0.19460220634937286, "learning_rate": 8.075103923691186e-06, "loss": 0.0118, "step": 66060 }, { "grad_norm": 0.1613473743200302, "learning_rate": 8.063845210520793e-06, "loss": 0.0157, "step": 66070 }, { "grad_norm": 0.1503162831068039, "learning_rate": 8.052593663092295e-06, "loss": 0.018, "step": 66080 }, { "grad_norm": 0.4755060076713562, "learning_rate": 8.04134928332827e-06, "loss": 0.0196, "step": 66090 }, { "grad_norm": 0.2452109009027481, "learning_rate": 8.030112073150086e-06, "loss": 0.0228, "step": 66100 }, { "grad_norm": 0.1826932579278946, "learning_rate": 8.018882034477881e-06, "loss": 0.014, "step": 66110 }, { "grad_norm": 0.2891702651977539, "learning_rate": 8.007659169230541e-06, "loss": 0.0151, "step": 66120 }, { "grad_norm": 0.34917888045310974, "learning_rate": 7.996443479325755e-06, "loss": 0.017, "step": 66130 }, { "grad_norm": 0.20431679487228394, "learning_rate": 7.985234966679977e-06, "loss": 0.0239, "step": 66140 }, { "grad_norm": 0.2706136405467987, "learning_rate": 7.974033633208438e-06, "loss": 0.018, "step": 66150 }, { "grad_norm": 0.39012905955314636, "learning_rate": 7.962839480825135e-06, "loss": 0.0146, "step": 66160 }, { "grad_norm": 0.3677935302257538, "learning_rate": 7.951652511442858e-06, "loss": 0.0227, "step": 66170 }, { "grad_norm": 0.24604925513267517, "learning_rate": 7.940472726973125e-06, "loss": 0.0185, "step": 66180 }, { "grad_norm": 0.2990710735321045, "learning_rate": 7.929300129326289e-06, "loss": 0.0155, "step": 66190 }, { "grad_norm": 0.337103009223938, "learning_rate": 7.91813472041142e-06, "loss": 0.0187, "step": 66200 }, { "grad_norm": 0.16332131624221802, "learning_rate": 7.906976502136376e-06, "loss": 0.0162, "step": 66210 }, { "grad_norm": 0.28095319867134094, "learning_rate": 7.89582547640782e-06, "loss": 0.0221, "step": 66220 }, { "grad_norm": 0.23301169276237488, "learning_rate": 7.884681645131115e-06, "loss": 0.0187, "step": 66230 }, { "grad_norm": 0.4074094891548157, "learning_rate": 7.87354501021048e-06, "loss": 0.0153, "step": 66240 }, { "grad_norm": 0.21285021305084229, "learning_rate": 7.86241557354882e-06, "loss": 0.0196, "step": 66250 }, { "grad_norm": 0.1633581966161728, "learning_rate": 7.8512933370479e-06, "loss": 0.017, "step": 66260 }, { "grad_norm": 0.18135008215904236, "learning_rate": 7.840178302608158e-06, "loss": 0.017, "step": 66270 }, { "grad_norm": 0.2652268409729004, "learning_rate": 7.82907047212888e-06, "loss": 0.0212, "step": 66280 }, { "grad_norm": 0.35307368636131287, "learning_rate": 7.81796984750809e-06, "loss": 0.0237, "step": 66290 }, { "grad_norm": 0.3220979571342468, "learning_rate": 7.806876430642546e-06, "loss": 0.0198, "step": 66300 }, { "grad_norm": 0.287623792886734, "learning_rate": 7.795790223427862e-06, "loss": 0.0201, "step": 66310 }, { "grad_norm": 0.3251935839653015, "learning_rate": 7.784711227758324e-06, "loss": 0.0232, "step": 66320 }, { "grad_norm": 0.21859946846961975, "learning_rate": 7.773639445527053e-06, "loss": 0.0237, "step": 66330 }, { "grad_norm": 0.43205636739730835, "learning_rate": 7.762574878625905e-06, "loss": 0.017, "step": 66340 }, { "grad_norm": 0.20430874824523926, "learning_rate": 7.751517528945513e-06, "loss": 0.0152, "step": 66350 }, { "grad_norm": 0.19660310447216034, "learning_rate": 7.740467398375278e-06, "loss": 0.017, "step": 66360 }, { "grad_norm": 0.3326385021209717, "learning_rate": 7.729424488803378e-06, "loss": 0.0176, "step": 66370 }, { "grad_norm": 0.3418580889701843, "learning_rate": 7.71838880211671e-06, "loss": 0.0277, "step": 66380 }, { "grad_norm": 0.29534515738487244, "learning_rate": 7.707360340200997e-06, "loss": 0.0152, "step": 66390 }, { "grad_norm": 0.2962932288646698, "learning_rate": 7.696339104940697e-06, "loss": 0.0169, "step": 66400 }, { "grad_norm": 0.19334262609481812, "learning_rate": 7.685325098219038e-06, "loss": 0.0199, "step": 66410 }, { "grad_norm": 0.19254961609840393, "learning_rate": 7.674318321918017e-06, "loss": 0.0172, "step": 66420 }, { "grad_norm": 0.2509831488132477, "learning_rate": 7.663318777918366e-06, "loss": 0.0211, "step": 66430 }, { "grad_norm": 0.25793880224227905, "learning_rate": 7.652326468099647e-06, "loss": 0.0207, "step": 66440 }, { "grad_norm": 0.3860264718532562, "learning_rate": 7.641341394340096e-06, "loss": 0.0158, "step": 66450 }, { "grad_norm": 0.3742409348487854, "learning_rate": 7.630363558516818e-06, "loss": 0.0157, "step": 66460 }, { "grad_norm": 0.33537188172340393, "learning_rate": 7.619392962505578e-06, "loss": 0.0214, "step": 66470 }, { "grad_norm": 0.26500722765922546, "learning_rate": 7.6084296081809725e-06, "loss": 0.0166, "step": 66480 }, { "grad_norm": 0.396365225315094, "learning_rate": 7.597473497416347e-06, "loss": 0.0216, "step": 66490 }, { "grad_norm": 0.17855334281921387, "learning_rate": 7.586524632083764e-06, "loss": 0.0153, "step": 66500 }, { "grad_norm": 0.25437918305397034, "learning_rate": 7.5755830140541326e-06, "loss": 0.013, "step": 66510 }, { "grad_norm": 0.34088465571403503, "learning_rate": 7.5646486451970475e-06, "loss": 0.0155, "step": 66520 }, { "grad_norm": 0.23490780591964722, "learning_rate": 7.553721527380897e-06, "loss": 0.0172, "step": 66530 }, { "grad_norm": 0.48940742015838623, "learning_rate": 7.542801662472826e-06, "loss": 0.0192, "step": 66540 }, { "grad_norm": 0.28602349758148193, "learning_rate": 7.5318890523387475e-06, "loss": 0.0185, "step": 66550 }, { "grad_norm": 0.3161574602127075, "learning_rate": 7.52098369884332e-06, "loss": 0.017, "step": 66560 }, { "grad_norm": 0.2299226075410843, "learning_rate": 7.510085603849992e-06, "loss": 0.0187, "step": 66570 }, { "grad_norm": 0.1982109099626541, "learning_rate": 7.499194769220918e-06, "loss": 0.0161, "step": 66580 }, { "grad_norm": 0.27637916803359985, "learning_rate": 7.48831119681706e-06, "loss": 0.0253, "step": 66590 }, { "grad_norm": 0.34547796845436096, "learning_rate": 7.477434888498119e-06, "loss": 0.0157, "step": 66600 }, { "grad_norm": 0.19285301864147186, "learning_rate": 7.466565846122564e-06, "loss": 0.021, "step": 66610 }, { "grad_norm": 0.3339967429637909, "learning_rate": 7.455704071547626e-06, "loss": 0.02, "step": 66620 }, { "grad_norm": 0.30385252833366394, "learning_rate": 7.444849566629247e-06, "loss": 0.0147, "step": 66630 }, { "grad_norm": 0.2791314423084259, "learning_rate": 7.434002333222212e-06, "loss": 0.0154, "step": 66640 }, { "grad_norm": 0.34136220812797546, "learning_rate": 7.423162373179976e-06, "loss": 0.0182, "step": 66650 }, { "grad_norm": 0.33067354559898376, "learning_rate": 7.412329688354835e-06, "loss": 0.017, "step": 66660 }, { "grad_norm": 0.2790120542049408, "learning_rate": 7.40150428059776e-06, "loss": 0.0135, "step": 66670 }, { "grad_norm": 0.3191775977611542, "learning_rate": 7.3906861517585354e-06, "loss": 0.0188, "step": 66680 }, { "grad_norm": 0.292698472738266, "learning_rate": 7.37987530368569e-06, "loss": 0.0198, "step": 66690 }, { "grad_norm": 0.2603072226047516, "learning_rate": 7.369071738226474e-06, "loss": 0.0165, "step": 66700 }, { "grad_norm": 0.18647131323814392, "learning_rate": 7.358275457226954e-06, "loss": 0.0171, "step": 66710 }, { "grad_norm": 0.25793904066085815, "learning_rate": 7.347486462531899e-06, "loss": 0.0175, "step": 66720 }, { "grad_norm": 0.22644685208797455, "learning_rate": 7.336704755984858e-06, "loss": 0.0162, "step": 66730 }, { "grad_norm": 0.4178645610809326, "learning_rate": 7.325930339428133e-06, "loss": 0.0174, "step": 66740 }, { "grad_norm": 0.19845318794250488, "learning_rate": 7.315163214702769e-06, "loss": 0.0155, "step": 66750 }, { "grad_norm": 0.40182554721832275, "learning_rate": 7.304403383648595e-06, "loss": 0.0223, "step": 66760 }, { "grad_norm": 0.21405944228172302, "learning_rate": 7.293650848104139e-06, "loss": 0.0187, "step": 66770 }, { "grad_norm": 0.22588381171226501, "learning_rate": 7.2829056099067374e-06, "loss": 0.019, "step": 66780 }, { "grad_norm": 0.4851626753807068, "learning_rate": 7.2721676708924494e-06, "loss": 0.0175, "step": 66790 }, { "grad_norm": 0.4914233386516571, "learning_rate": 7.261437032896096e-06, "loss": 0.0278, "step": 66800 }, { "grad_norm": 0.32321202754974365, "learning_rate": 7.250713697751255e-06, "loss": 0.0183, "step": 66810 }, { "grad_norm": 0.20515772700309753, "learning_rate": 7.239997667290255e-06, "loss": 0.0178, "step": 66820 }, { "grad_norm": 0.16346700489521027, "learning_rate": 7.2292889433441425e-06, "loss": 0.0131, "step": 66830 }, { "grad_norm": 0.2833842635154724, "learning_rate": 7.218587527742793e-06, "loss": 0.0146, "step": 66840 }, { "grad_norm": 0.34526363015174866, "learning_rate": 7.207893422314749e-06, "loss": 0.0158, "step": 66850 }, { "grad_norm": 0.24543172121047974, "learning_rate": 7.1972066288873545e-06, "loss": 0.0166, "step": 66860 }, { "grad_norm": 0.3244672119617462, "learning_rate": 7.186527149286687e-06, "loss": 0.0323, "step": 66870 }, { "grad_norm": 0.2726391851902008, "learning_rate": 7.175854985337576e-06, "loss": 0.0223, "step": 66880 }, { "grad_norm": 0.29091763496398926, "learning_rate": 7.16519013886362e-06, "loss": 0.0188, "step": 66890 }, { "grad_norm": 0.39322173595428467, "learning_rate": 7.154532611687109e-06, "loss": 0.0139, "step": 66900 }, { "grad_norm": 0.2721700966358185, "learning_rate": 7.143882405629177e-06, "loss": 0.017, "step": 66910 }, { "grad_norm": 0.23246461153030396, "learning_rate": 7.133239522509605e-06, "loss": 0.0143, "step": 66920 }, { "grad_norm": 0.14841075241565704, "learning_rate": 7.1226039641469956e-06, "loss": 0.0156, "step": 66930 }, { "grad_norm": 0.24871134757995605, "learning_rate": 7.111975732358678e-06, "loss": 0.0174, "step": 66940 }, { "grad_norm": 0.2068396806716919, "learning_rate": 7.101354828960693e-06, "loss": 0.0145, "step": 66950 }, { "grad_norm": 0.2555117905139923, "learning_rate": 7.090741255767918e-06, "loss": 0.0172, "step": 66960 }, { "grad_norm": 0.23634833097457886, "learning_rate": 7.080135014593875e-06, "loss": 0.0162, "step": 66970 }, { "grad_norm": 0.3035007119178772, "learning_rate": 7.069536107250896e-06, "loss": 0.0159, "step": 66980 }, { "grad_norm": 0.23112230002880096, "learning_rate": 7.058944535550049e-06, "loss": 0.0131, "step": 66990 }, { "grad_norm": 0.20252372324466705, "learning_rate": 7.048360301301138e-06, "loss": 0.0193, "step": 67000 }, { "grad_norm": 0.4055778682231903, "learning_rate": 7.03778340631272e-06, "loss": 0.0185, "step": 67010 }, { "grad_norm": 0.3230940103530884, "learning_rate": 7.0272138523921e-06, "loss": 0.0206, "step": 67020 }, { "grad_norm": 0.21719548106193542, "learning_rate": 7.016651641345334e-06, "loss": 0.0201, "step": 67030 }, { "grad_norm": 0.483332097530365, "learning_rate": 7.0060967749772e-06, "loss": 0.0176, "step": 67040 }, { "grad_norm": 0.49945464730262756, "learning_rate": 6.995549255091238e-06, "loss": 0.0283, "step": 67050 }, { "grad_norm": 0.22309137880802155, "learning_rate": 6.98500908348973e-06, "loss": 0.0212, "step": 67060 }, { "grad_norm": 0.42098942399024963, "learning_rate": 6.974476261973711e-06, "loss": 0.0181, "step": 67070 }, { "grad_norm": 0.25128239393234253, "learning_rate": 6.963950792342949e-06, "loss": 0.0147, "step": 67080 }, { "grad_norm": 0.250665545463562, "learning_rate": 6.9534326763959715e-06, "loss": 0.0136, "step": 67090 }, { "grad_norm": 0.33456841111183167, "learning_rate": 6.942921915929995e-06, "loss": 0.0173, "step": 67100 }, { "grad_norm": 0.19744068384170532, "learning_rate": 6.9324185127410735e-06, "loss": 0.0194, "step": 67110 }, { "grad_norm": 0.3200302720069885, "learning_rate": 6.921922468623921e-06, "loss": 0.0235, "step": 67120 }, { "grad_norm": 0.2575736939907074, "learning_rate": 6.911433785372023e-06, "loss": 0.0149, "step": 67130 }, { "grad_norm": 0.29189538955688477, "learning_rate": 6.900952464777632e-06, "loss": 0.0236, "step": 67140 }, { "grad_norm": 0.35558417439460754, "learning_rate": 6.8904785086316815e-06, "loss": 0.0175, "step": 67150 }, { "grad_norm": 0.42279186844825745, "learning_rate": 6.880011918723927e-06, "loss": 0.0147, "step": 67160 }, { "grad_norm": 0.2996011972427368, "learning_rate": 6.86955269684279e-06, "loss": 0.0152, "step": 67170 }, { "grad_norm": 0.3696849048137665, "learning_rate": 6.859100844775473e-06, "loss": 0.0274, "step": 67180 }, { "grad_norm": 0.16658860445022583, "learning_rate": 6.84865636430792e-06, "loss": 0.0176, "step": 67190 }, { "grad_norm": 0.23680561780929565, "learning_rate": 6.838219257224804e-06, "loss": 0.0142, "step": 67200 }, { "grad_norm": 0.25993508100509644, "learning_rate": 6.827789525309536e-06, "loss": 0.0187, "step": 67210 }, { "grad_norm": 0.3781513571739197, "learning_rate": 6.81736717034428e-06, "loss": 0.0248, "step": 67220 }, { "grad_norm": 0.19632352888584137, "learning_rate": 6.806952194109933e-06, "loss": 0.0303, "step": 67230 }, { "grad_norm": 0.2751877009868622, "learning_rate": 6.796544598386112e-06, "loss": 0.0129, "step": 67240 }, { "grad_norm": 0.25141796469688416, "learning_rate": 6.786144384951204e-06, "loss": 0.0205, "step": 67250 }, { "grad_norm": 0.2857585847377777, "learning_rate": 6.775751555582322e-06, "loss": 0.0176, "step": 67260 }, { "grad_norm": 0.3227040767669678, "learning_rate": 6.76536611205531e-06, "loss": 0.016, "step": 67270 }, { "grad_norm": 0.29964834451675415, "learning_rate": 6.754988056144762e-06, "loss": 0.0225, "step": 67280 }, { "grad_norm": 0.2852201461791992, "learning_rate": 6.744617389624014e-06, "loss": 0.0145, "step": 67290 }, { "grad_norm": 0.4911403954029083, "learning_rate": 6.734254114265087e-06, "loss": 0.0185, "step": 67300 }, { "grad_norm": 0.2796526253223419, "learning_rate": 6.723898231838843e-06, "loss": 0.0164, "step": 67310 }, { "grad_norm": 0.22691696882247925, "learning_rate": 6.713549744114766e-06, "loss": 0.0205, "step": 67320 }, { "grad_norm": 0.15688872337341309, "learning_rate": 6.703208652861159e-06, "loss": 0.017, "step": 67330 }, { "grad_norm": 0.28824129700660706, "learning_rate": 6.69287495984503e-06, "loss": 0.0188, "step": 67340 }, { "grad_norm": 0.31471681594848633, "learning_rate": 6.6825486668321e-06, "loss": 0.0189, "step": 67350 }, { "grad_norm": 0.21114438772201538, "learning_rate": 6.672229775586886e-06, "loss": 0.0187, "step": 67360 }, { "grad_norm": 0.24986793100833893, "learning_rate": 6.661918287872576e-06, "loss": 0.013, "step": 67370 }, { "grad_norm": 0.19476410746574402, "learning_rate": 6.6516142054511346e-06, "loss": 0.0206, "step": 67380 }, { "grad_norm": 0.31058821082115173, "learning_rate": 6.641317530083241e-06, "loss": 0.0161, "step": 67390 }, { "grad_norm": 0.2688857913017273, "learning_rate": 6.631028263528322e-06, "loss": 0.0171, "step": 67400 }, { "grad_norm": 0.29743078351020813, "learning_rate": 6.620746407544537e-06, "loss": 0.0132, "step": 67410 }, { "grad_norm": 0.30320581793785095, "learning_rate": 6.610471963888742e-06, "loss": 0.0171, "step": 67420 }, { "grad_norm": 0.2723594605922699, "learning_rate": 6.600204934316606e-06, "loss": 0.0147, "step": 67430 }, { "grad_norm": 0.16245558857917786, "learning_rate": 6.589945320582452e-06, "loss": 0.0172, "step": 67440 }, { "grad_norm": 0.4187379777431488, "learning_rate": 6.579693124439374e-06, "loss": 0.0169, "step": 67450 }, { "grad_norm": 0.23763501644134521, "learning_rate": 6.56944834763919e-06, "loss": 0.0154, "step": 67460 }, { "grad_norm": 0.29059815406799316, "learning_rate": 6.5592109919324575e-06, "loss": 0.0165, "step": 67470 }, { "grad_norm": 0.17512068152427673, "learning_rate": 6.54898105906846e-06, "loss": 0.0162, "step": 67480 }, { "grad_norm": 0.25123175978660583, "learning_rate": 6.5387585507952155e-06, "loss": 0.0254, "step": 67490 }, { "grad_norm": 0.24352921545505524, "learning_rate": 6.528543468859461e-06, "loss": 0.0161, "step": 67500 }, { "grad_norm": 0.22048461437225342, "learning_rate": 6.518335815006682e-06, "loss": 0.0197, "step": 67510 }, { "grad_norm": 0.31290140748023987, "learning_rate": 6.5081355909810845e-06, "loss": 0.0179, "step": 67520 }, { "grad_norm": 0.26356446743011475, "learning_rate": 6.497942798525608e-06, "loss": 0.0234, "step": 67530 }, { "grad_norm": 0.3060029149055481, "learning_rate": 6.487757439381936e-06, "loss": 0.0187, "step": 67540 }, { "grad_norm": 0.39420729875564575, "learning_rate": 6.477579515290433e-06, "loss": 0.0194, "step": 67550 }, { "grad_norm": 0.27613434195518494, "learning_rate": 6.467409027990273e-06, "loss": 0.023, "step": 67560 }, { "grad_norm": 0.2790251672267914, "learning_rate": 6.457245979219279e-06, "loss": 0.0214, "step": 67570 }, { "grad_norm": 0.2561887800693512, "learning_rate": 6.447090370714054e-06, "loss": 0.0266, "step": 67580 }, { "grad_norm": 0.24998106062412262, "learning_rate": 6.436942204209917e-06, "loss": 0.022, "step": 67590 }, { "grad_norm": 0.31208887696266174, "learning_rate": 6.4268014814408804e-06, "loss": 0.0261, "step": 67600 }, { "grad_norm": 0.31367310881614685, "learning_rate": 6.41666820413977e-06, "loss": 0.0168, "step": 67610 }, { "grad_norm": 0.51338791847229, "learning_rate": 6.406542374038033e-06, "loss": 0.0247, "step": 67620 }, { "grad_norm": 0.2690032720565796, "learning_rate": 6.396423992865935e-06, "loss": 0.0141, "step": 67630 }, { "grad_norm": 0.2926654517650604, "learning_rate": 6.386313062352412e-06, "loss": 0.0176, "step": 67640 }, { "grad_norm": 0.17292046546936035, "learning_rate": 6.376209584225152e-06, "loss": 0.0158, "step": 67650 }, { "grad_norm": 0.19820775091648102, "learning_rate": 6.366113560210557e-06, "loss": 0.018, "step": 67660 }, { "grad_norm": 0.5218705534934998, "learning_rate": 6.356024992033766e-06, "loss": 0.0257, "step": 67670 }, { "grad_norm": 0.31342631578445435, "learning_rate": 6.345943881418648e-06, "loss": 0.015, "step": 67680 }, { "grad_norm": 0.19568917155265808, "learning_rate": 6.33587023008777e-06, "loss": 0.0156, "step": 67690 }, { "grad_norm": 0.1867767870426178, "learning_rate": 6.32580403976245e-06, "loss": 0.0128, "step": 67700 }, { "grad_norm": 0.33270299434661865, "learning_rate": 6.31574531216273e-06, "loss": 0.0177, "step": 67710 }, { "grad_norm": 0.3962385058403015, "learning_rate": 6.305694049007371e-06, "loss": 0.0217, "step": 67720 }, { "grad_norm": 0.3139978349208832, "learning_rate": 6.2956502520138575e-06, "loss": 0.0153, "step": 67730 }, { "grad_norm": 0.41832679510116577, "learning_rate": 6.285613922898409e-06, "loss": 0.0145, "step": 67740 }, { "grad_norm": 0.3317209780216217, "learning_rate": 6.275585063375927e-06, "loss": 0.0219, "step": 67750 }, { "grad_norm": 0.1854781210422516, "learning_rate": 6.265563675160113e-06, "loss": 0.0189, "step": 67760 }, { "grad_norm": 0.2861252427101135, "learning_rate": 6.25554975996332e-06, "loss": 0.0159, "step": 67770 }, { "grad_norm": 0.21247419714927673, "learning_rate": 6.245543319496661e-06, "loss": 0.0126, "step": 67780 }, { "grad_norm": 0.29178524017333984, "learning_rate": 6.2355443554699685e-06, "loss": 0.0187, "step": 67790 }, { "grad_norm": 0.19852764904499054, "learning_rate": 6.225552869591766e-06, "loss": 0.0167, "step": 67800 }, { "grad_norm": 0.30834800004959106, "learning_rate": 6.215568863569365e-06, "loss": 0.0189, "step": 67810 }, { "grad_norm": 0.4130944609642029, "learning_rate": 6.205592339108712e-06, "loss": 0.0207, "step": 67820 }, { "grad_norm": 0.3055512607097626, "learning_rate": 6.195623297914577e-06, "loss": 0.0158, "step": 67830 }, { "grad_norm": 0.32580628991127014, "learning_rate": 6.185661741690357e-06, "loss": 0.0155, "step": 67840 }, { "grad_norm": 0.21498608589172363, "learning_rate": 6.1757076721382145e-06, "loss": 0.0228, "step": 67850 }, { "grad_norm": 0.14873267710208893, "learning_rate": 6.165761090959038e-06, "loss": 0.019, "step": 67860 }, { "grad_norm": 0.27704477310180664, "learning_rate": 6.155821999852424e-06, "loss": 0.022, "step": 67870 }, { "grad_norm": 0.19468730688095093, "learning_rate": 6.145890400516696e-06, "loss": 0.0201, "step": 67880 }, { "grad_norm": 0.19019600749015808, "learning_rate": 6.1359662946488816e-06, "loss": 0.0156, "step": 67890 }, { "grad_norm": 0.3089979588985443, "learning_rate": 6.12604968394474e-06, "loss": 0.0143, "step": 67900 }, { "grad_norm": 0.2714986503124237, "learning_rate": 6.11614057009875e-06, "loss": 0.0199, "step": 67910 }, { "grad_norm": 0.2137863039970398, "learning_rate": 6.106238954804111e-06, "loss": 0.0188, "step": 67920 }, { "grad_norm": 0.42602354288101196, "learning_rate": 6.096344839752738e-06, "loss": 0.0175, "step": 67930 }, { "grad_norm": 0.30376434326171875, "learning_rate": 6.086458226635278e-06, "loss": 0.0196, "step": 67940 }, { "grad_norm": 0.3705054819583893, "learning_rate": 6.076579117141046e-06, "loss": 0.0278, "step": 67950 }, { "grad_norm": 0.38083845376968384, "learning_rate": 6.066707512958153e-06, "loss": 0.0179, "step": 67960 }, { "grad_norm": 0.26084575057029724, "learning_rate": 6.056843415773361e-06, "loss": 0.0214, "step": 67970 }, { "grad_norm": 0.28417664766311646, "learning_rate": 6.0469868272721776e-06, "loss": 0.0222, "step": 67980 }, { "grad_norm": 0.22969551384449005, "learning_rate": 6.037137749138844e-06, "loss": 0.0223, "step": 67990 }, { "grad_norm": 0.34683647751808167, "learning_rate": 6.027296183056252e-06, "loss": 0.0205, "step": 68000 }, { "grad_norm": 0.3060990571975708, "learning_rate": 6.017462130706114e-06, "loss": 0.0143, "step": 68010 }, { "grad_norm": 0.31004202365875244, "learning_rate": 6.0076355937687516e-06, "loss": 0.0217, "step": 68020 }, { "grad_norm": 0.272550493478775, "learning_rate": 5.9978165739232925e-06, "loss": 0.0169, "step": 68030 }, { "grad_norm": 0.1631144881248474, "learning_rate": 5.988005072847508e-06, "loss": 0.0173, "step": 68040 }, { "grad_norm": 0.3346308767795563, "learning_rate": 5.978201092217928e-06, "loss": 0.0262, "step": 68050 }, { "grad_norm": 0.1913563311100006, "learning_rate": 5.9684046337097895e-06, "loss": 0.0137, "step": 68060 }, { "grad_norm": 0.33774715662002563, "learning_rate": 5.958615698997016e-06, "loss": 0.0189, "step": 68070 }, { "grad_norm": 0.2821984589099884, "learning_rate": 5.948834289752303e-06, "loss": 0.0194, "step": 68080 }, { "grad_norm": 0.29699987173080444, "learning_rate": 5.939060407646996e-06, "loss": 0.0215, "step": 68090 }, { "grad_norm": 0.2254890352487564, "learning_rate": 5.929294054351198e-06, "loss": 0.0194, "step": 68100 }, { "grad_norm": 0.23413844406604767, "learning_rate": 5.919535231533707e-06, "loss": 0.015, "step": 68110 }, { "grad_norm": 0.308102011680603, "learning_rate": 5.909783940862046e-06, "loss": 0.0154, "step": 68120 }, { "grad_norm": 0.26450398564338684, "learning_rate": 5.900040184002436e-06, "loss": 0.017, "step": 68130 }, { "grad_norm": 0.384491503238678, "learning_rate": 5.890303962619831e-06, "loss": 0.0143, "step": 68140 }, { "grad_norm": 0.41739794611930847, "learning_rate": 5.88057527837787e-06, "loss": 0.0186, "step": 68150 }, { "grad_norm": 0.2930373251438141, "learning_rate": 5.870854132938924e-06, "loss": 0.0204, "step": 68160 }, { "grad_norm": 0.21905028820037842, "learning_rate": 5.861140527964071e-06, "loss": 0.0195, "step": 68170 }, { "grad_norm": 0.12574687600135803, "learning_rate": 5.851434465113098e-06, "loss": 0.0115, "step": 68180 }, { "grad_norm": 0.2692309617996216, "learning_rate": 5.841735946044524e-06, "loss": 0.018, "step": 68190 }, { "grad_norm": 0.20733648538589478, "learning_rate": 5.832044972415523e-06, "loss": 0.0246, "step": 68200 }, { "grad_norm": 0.20538201928138733, "learning_rate": 5.8223615458820655e-06, "loss": 0.0137, "step": 68210 }, { "grad_norm": 0.24081391096115112, "learning_rate": 5.812685668098733e-06, "loss": 0.0166, "step": 68220 }, { "grad_norm": 0.22385862469673157, "learning_rate": 5.803017340718913e-06, "loss": 0.02, "step": 68230 }, { "grad_norm": 0.18785631656646729, "learning_rate": 5.793356565394636e-06, "loss": 0.0143, "step": 68240 }, { "grad_norm": 0.27489930391311646, "learning_rate": 5.783703343776658e-06, "loss": 0.0174, "step": 68250 }, { "grad_norm": 0.36867403984069824, "learning_rate": 5.774057677514477e-06, "loss": 0.0144, "step": 68260 }, { "grad_norm": 0.25998610258102417, "learning_rate": 5.764419568256235e-06, "loss": 0.0211, "step": 68270 }, { "grad_norm": 0.41402706503868103, "learning_rate": 5.754789017648865e-06, "loss": 0.0326, "step": 68280 }, { "grad_norm": 0.16562554240226746, "learning_rate": 5.745166027337934e-06, "loss": 0.0196, "step": 68290 }, { "grad_norm": 0.16938737034797668, "learning_rate": 5.735550598967754e-06, "loss": 0.0169, "step": 68300 }, { "grad_norm": 0.31064578890800476, "learning_rate": 5.725942734181339e-06, "loss": 0.0209, "step": 68310 }, { "grad_norm": 0.1830388903617859, "learning_rate": 5.7163424346204146e-06, "loss": 0.0195, "step": 68320 }, { "grad_norm": 0.3114524185657501, "learning_rate": 5.706749701925407e-06, "loss": 0.0148, "step": 68330 }, { "grad_norm": 0.1827474683523178, "learning_rate": 5.697164537735461e-06, "loss": 0.0152, "step": 68340 }, { "grad_norm": 0.24457408487796783, "learning_rate": 5.687586943688406e-06, "loss": 0.0132, "step": 68350 }, { "grad_norm": 0.20751871168613434, "learning_rate": 5.678016921420792e-06, "loss": 0.0175, "step": 68360 }, { "grad_norm": 0.2967385947704315, "learning_rate": 5.668454472567875e-06, "loss": 0.0254, "step": 68370 }, { "grad_norm": 0.3675763010978699, "learning_rate": 5.658899598763617e-06, "loss": 0.0232, "step": 68380 }, { "grad_norm": 0.1661239117383957, "learning_rate": 5.6493523016407035e-06, "loss": 0.0128, "step": 68390 }, { "grad_norm": 0.22988566756248474, "learning_rate": 5.6398125828304615e-06, "loss": 0.0184, "step": 68400 }, { "grad_norm": 0.45879554748535156, "learning_rate": 5.630280443963015e-06, "loss": 0.0174, "step": 68410 }, { "grad_norm": 0.16975529491901398, "learning_rate": 5.620755886667112e-06, "loss": 0.0157, "step": 68420 }, { "grad_norm": 0.30261021852493286, "learning_rate": 5.611238912570271e-06, "loss": 0.0164, "step": 68430 }, { "grad_norm": 0.5858074426651001, "learning_rate": 5.601729523298649e-06, "loss": 0.0276, "step": 68440 }, { "grad_norm": 0.2761446535587311, "learning_rate": 5.592227720477161e-06, "loss": 0.0165, "step": 68450 }, { "grad_norm": 0.4448928236961365, "learning_rate": 5.582733505729415e-06, "loss": 0.019, "step": 68460 }, { "grad_norm": 0.45716235041618347, "learning_rate": 5.573246880677668e-06, "loss": 0.0207, "step": 68470 }, { "grad_norm": 0.17110252380371094, "learning_rate": 5.56376784694298e-06, "loss": 0.0175, "step": 68480 }, { "grad_norm": 0.28888794779777527, "learning_rate": 5.554296406145027e-06, "loss": 0.0203, "step": 68490 }, { "grad_norm": 0.3874979615211487, "learning_rate": 5.544832559902219e-06, "loss": 0.0155, "step": 68500 }, { "grad_norm": 0.27956464886665344, "learning_rate": 5.53537630983168e-06, "loss": 0.0217, "step": 68510 }, { "grad_norm": 0.21805919706821442, "learning_rate": 5.5259276575492125e-06, "loss": 0.0203, "step": 68520 }, { "grad_norm": 0.2612959146499634, "learning_rate": 5.516486604669357e-06, "loss": 0.0261, "step": 68530 }, { "grad_norm": 0.3806387484073639, "learning_rate": 5.507053152805303e-06, "loss": 0.0283, "step": 68540 }, { "grad_norm": 0.1972789317369461, "learning_rate": 5.497627303568975e-06, "loss": 0.0193, "step": 68550 }, { "grad_norm": 0.34743213653564453, "learning_rate": 5.488209058571003e-06, "loss": 0.0157, "step": 68560 }, { "grad_norm": 0.23689478635787964, "learning_rate": 5.4787984194207054e-06, "loss": 0.0198, "step": 68570 }, { "grad_norm": 0.31967028975486755, "learning_rate": 5.469395387726095e-06, "loss": 0.0201, "step": 68580 }, { "grad_norm": 0.28803256154060364, "learning_rate": 5.45999996509391e-06, "loss": 0.0185, "step": 68590 }, { "grad_norm": 0.20510593056678772, "learning_rate": 5.450612153129536e-06, "loss": 0.0206, "step": 68600 }, { "grad_norm": 0.1943170428276062, "learning_rate": 5.4412319534371426e-06, "loss": 0.0191, "step": 68610 }, { "grad_norm": 0.22420942783355713, "learning_rate": 5.431859367619513e-06, "loss": 0.0147, "step": 68620 }, { "grad_norm": 0.29630374908447266, "learning_rate": 5.422494397278172e-06, "loss": 0.0181, "step": 68630 }, { "grad_norm": 0.42743560671806335, "learning_rate": 5.413137044013344e-06, "loss": 0.0205, "step": 68640 }, { "grad_norm": 0.2059350460767746, "learning_rate": 5.403787309423941e-06, "loss": 0.0159, "step": 68650 }, { "grad_norm": 0.26236921548843384, "learning_rate": 5.3944451951075835e-06, "loss": 0.0139, "step": 68660 }, { "grad_norm": 0.20754413306713104, "learning_rate": 5.385110702660562e-06, "loss": 0.0169, "step": 68670 }, { "grad_norm": 0.27880173921585083, "learning_rate": 5.375783833677922e-06, "loss": 0.0207, "step": 68680 }, { "grad_norm": 0.2741541266441345, "learning_rate": 5.36646458975334e-06, "loss": 0.0195, "step": 68690 }, { "grad_norm": 0.31925514340400696, "learning_rate": 5.3571529724792294e-06, "loss": 0.0135, "step": 68700 }, { "grad_norm": 0.39892876148223877, "learning_rate": 5.347848983446702e-06, "loss": 0.0206, "step": 68710 }, { "grad_norm": 0.27213379740715027, "learning_rate": 5.3385526242455185e-06, "loss": 0.0184, "step": 68720 }, { "grad_norm": 0.21433289349079132, "learning_rate": 5.329263896464226e-06, "loss": 0.0195, "step": 68730 }, { "grad_norm": 0.28176602721214294, "learning_rate": 5.3199828016899715e-06, "loss": 0.0169, "step": 68740 }, { "grad_norm": 0.277564138174057, "learning_rate": 5.310709341508657e-06, "loss": 0.0202, "step": 68750 }, { "grad_norm": 0.2325631082057953, "learning_rate": 5.301443517504861e-06, "loss": 0.0171, "step": 68760 }, { "grad_norm": 0.3249674141407013, "learning_rate": 5.292185331261862e-06, "loss": 0.013, "step": 68770 }, { "grad_norm": 0.1732049286365509, "learning_rate": 5.28293478436162e-06, "loss": 0.0165, "step": 68780 }, { "grad_norm": 0.14128656685352325, "learning_rate": 5.273691878384829e-06, "loss": 0.0196, "step": 68790 }, { "grad_norm": 0.2754075527191162, "learning_rate": 5.264456614910812e-06, "loss": 0.0221, "step": 68800 }, { "grad_norm": 0.3410904109477997, "learning_rate": 5.255228995517647e-06, "loss": 0.0232, "step": 68810 }, { "grad_norm": 0.22352352738380432, "learning_rate": 5.246009021782067e-06, "loss": 0.0156, "step": 68820 }, { "grad_norm": 0.18727941811084747, "learning_rate": 5.2367966952795225e-06, "loss": 0.0222, "step": 68830 }, { "grad_norm": 0.20998118817806244, "learning_rate": 5.2275920175841485e-06, "loss": 0.0161, "step": 68840 }, { "grad_norm": 0.22774161398410797, "learning_rate": 5.218394990268771e-06, "loss": 0.0169, "step": 68850 }, { "grad_norm": 0.36241012811660767, "learning_rate": 5.209205614904916e-06, "loss": 0.0226, "step": 68860 }, { "grad_norm": 0.3165237009525299, "learning_rate": 5.200023893062777e-06, "loss": 0.0187, "step": 68870 }, { "grad_norm": 0.16882987320423126, "learning_rate": 5.190849826311289e-06, "loss": 0.0144, "step": 68880 }, { "grad_norm": 0.26028382778167725, "learning_rate": 5.181683416218025e-06, "loss": 0.021, "step": 68890 }, { "grad_norm": 0.5330731272697449, "learning_rate": 5.172524664349276e-06, "loss": 0.0165, "step": 68900 }, { "grad_norm": 0.17689095437526703, "learning_rate": 5.1633735722700416e-06, "loss": 0.0176, "step": 68910 }, { "grad_norm": 0.14698752760887146, "learning_rate": 5.154230141543958e-06, "loss": 0.0211, "step": 68920 }, { "grad_norm": 0.24639302492141724, "learning_rate": 5.145094373733433e-06, "loss": 0.0147, "step": 68930 }, { "grad_norm": 0.5930618047714233, "learning_rate": 5.135966270399478e-06, "loss": 0.0145, "step": 68940 }, { "grad_norm": 0.3669230043888092, "learning_rate": 5.126845833101857e-06, "loss": 0.0248, "step": 68950 }, { "grad_norm": 0.20007026195526123, "learning_rate": 5.1177330633989994e-06, "loss": 0.0175, "step": 68960 }, { "grad_norm": 0.3630664050579071, "learning_rate": 5.108627962848033e-06, "loss": 0.0314, "step": 68970 }, { "grad_norm": 0.22677290439605713, "learning_rate": 5.099530533004759e-06, "loss": 0.0182, "step": 68980 }, { "grad_norm": 0.22971396148204803, "learning_rate": 5.090440775423699e-06, "loss": 0.0142, "step": 68990 }, { "grad_norm": 0.3047954738140106, "learning_rate": 5.081358691658022e-06, "loss": 0.0196, "step": 69000 }, { "grad_norm": 0.3582388758659363, "learning_rate": 5.072284283259621e-06, "loss": 0.0203, "step": 69010 }, { "grad_norm": 0.22815018892288208, "learning_rate": 5.063217551779053e-06, "loss": 0.0171, "step": 69020 }, { "grad_norm": 0.3737517297267914, "learning_rate": 5.054158498765588e-06, "loss": 0.0181, "step": 69030 }, { "grad_norm": 0.45812997221946716, "learning_rate": 5.0451071257671625e-06, "loss": 0.0191, "step": 69040 }, { "grad_norm": 0.5866071581840515, "learning_rate": 5.036063434330407e-06, "loss": 0.0209, "step": 69050 }, { "grad_norm": 0.138636976480484, "learning_rate": 5.027027426000652e-06, "loss": 0.0145, "step": 69060 }, { "grad_norm": 0.10362141579389572, "learning_rate": 5.017999102321886e-06, "loss": 0.0163, "step": 69070 }, { "grad_norm": 0.38070547580718994, "learning_rate": 5.0089784648368224e-06, "loss": 0.0218, "step": 69080 }, { "grad_norm": 0.1586374044418335, "learning_rate": 4.999965515086829e-06, "loss": 0.0139, "step": 69090 }, { "grad_norm": 0.21951964497566223, "learning_rate": 4.990960254611976e-06, "loss": 0.0172, "step": 69100 }, { "grad_norm": 0.43629971146583557, "learning_rate": 4.9819626849510194e-06, "loss": 0.0209, "step": 69110 }, { "grad_norm": 0.1788860708475113, "learning_rate": 4.97297280764138e-06, "loss": 0.0252, "step": 69120 }, { "grad_norm": 0.19339168071746826, "learning_rate": 4.963990624219211e-06, "loss": 0.018, "step": 69130 }, { "grad_norm": 0.23548655211925507, "learning_rate": 4.955016136219298e-06, "loss": 0.0182, "step": 69140 }, { "grad_norm": 0.26141873002052307, "learning_rate": 4.946049345175146e-06, "loss": 0.0153, "step": 69150 }, { "grad_norm": 0.29206404089927673, "learning_rate": 4.937090252618937e-06, "loss": 0.0191, "step": 69160 }, { "grad_norm": 0.2374485731124878, "learning_rate": 4.928138860081521e-06, "loss": 0.0285, "step": 69170 }, { "grad_norm": 0.24601969122886658, "learning_rate": 4.919195169092472e-06, "loss": 0.0224, "step": 69180 }, { "grad_norm": 0.15306095778942108, "learning_rate": 4.910259181179994e-06, "loss": 0.0193, "step": 69190 }, { "grad_norm": 0.4943068027496338, "learning_rate": 4.901330897871015e-06, "loss": 0.0193, "step": 69200 }, { "grad_norm": 0.15804460644721985, "learning_rate": 4.8924103206911375e-06, "loss": 0.0144, "step": 69210 }, { "grad_norm": 0.2863945960998535, "learning_rate": 4.883497451164637e-06, "loss": 0.0224, "step": 69220 }, { "grad_norm": 0.22344139218330383, "learning_rate": 4.874592290814484e-06, "loss": 0.0199, "step": 69230 }, { "grad_norm": 0.4314550459384918, "learning_rate": 4.865694841162327e-06, "loss": 0.0257, "step": 69240 }, { "grad_norm": 0.1343001425266266, "learning_rate": 4.856805103728496e-06, "loss": 0.0157, "step": 69250 }, { "grad_norm": 0.2294521927833557, "learning_rate": 4.847923080032007e-06, "loss": 0.0134, "step": 69260 }, { "grad_norm": 0.2124212086200714, "learning_rate": 4.8390487715905486e-06, "loss": 0.0157, "step": 69270 }, { "grad_norm": 0.20462334156036377, "learning_rate": 4.8301821799205e-06, "loss": 0.0146, "step": 69280 }, { "grad_norm": 0.1649322360754013, "learning_rate": 4.821323306536918e-06, "loss": 0.0131, "step": 69290 }, { "grad_norm": 0.34682217240333557, "learning_rate": 4.8124721529535455e-06, "loss": 0.0177, "step": 69300 }, { "grad_norm": 0.38282066583633423, "learning_rate": 4.803628720682807e-06, "loss": 0.0233, "step": 69310 }, { "grad_norm": 0.2752598524093628, "learning_rate": 4.794793011235776e-06, "loss": 0.0204, "step": 69320 }, { "grad_norm": 0.38266900181770325, "learning_rate": 4.78596502612228e-06, "loss": 0.0246, "step": 69330 }, { "grad_norm": 0.5628576278686523, "learning_rate": 4.777144766850738e-06, "loss": 0.0214, "step": 69340 }, { "grad_norm": 0.1673825979232788, "learning_rate": 4.768332234928313e-06, "loss": 0.0195, "step": 69350 }, { "grad_norm": 0.45170992612838745, "learning_rate": 4.759527431860828e-06, "loss": 0.0165, "step": 69360 }, { "grad_norm": 0.32876095175743103, "learning_rate": 4.750730359152755e-06, "loss": 0.0201, "step": 69370 }, { "grad_norm": 0.18130017817020416, "learning_rate": 4.741941018307311e-06, "loss": 0.014, "step": 69380 }, { "grad_norm": 0.21842089295387268, "learning_rate": 4.733159410826321e-06, "loss": 0.0167, "step": 69390 }, { "grad_norm": 0.2875071167945862, "learning_rate": 4.724385538210357e-06, "loss": 0.0168, "step": 69400 }, { "grad_norm": 0.35922524333000183, "learning_rate": 4.715619401958599e-06, "loss": 0.0187, "step": 69410 }, { "grad_norm": 0.25368478894233704, "learning_rate": 4.706861003568958e-06, "loss": 0.0197, "step": 69420 }, { "grad_norm": 0.18651089072227478, "learning_rate": 4.698110344538003e-06, "loss": 0.0177, "step": 69430 }, { "grad_norm": 0.23749421536922455, "learning_rate": 4.689367426360975e-06, "loss": 0.0147, "step": 69440 }, { "grad_norm": 0.3133898973464966, "learning_rate": 4.680632250531819e-06, "loss": 0.017, "step": 69450 }, { "grad_norm": 0.2114371508359909, "learning_rate": 4.671904818543115e-06, "loss": 0.0152, "step": 69460 }, { "grad_norm": 0.2540093660354614, "learning_rate": 4.66318513188615e-06, "loss": 0.0142, "step": 69470 }, { "grad_norm": 0.1516668200492859, "learning_rate": 4.654473192050884e-06, "loss": 0.0167, "step": 69480 }, { "grad_norm": 0.15820255875587463, "learning_rate": 4.64576900052594e-06, "loss": 0.0153, "step": 69490 }, { "grad_norm": 0.5399831533432007, "learning_rate": 4.637072558798638e-06, "loss": 0.0138, "step": 69500 }, { "grad_norm": 0.26563140749931335, "learning_rate": 4.628383868354969e-06, "loss": 0.0174, "step": 69510 }, { "grad_norm": 0.3753955662250519, "learning_rate": 4.6197029306795595e-06, "loss": 0.0144, "step": 69520 }, { "grad_norm": 0.18631890416145325, "learning_rate": 4.611029747255779e-06, "loss": 0.0168, "step": 69530 }, { "grad_norm": 0.2682463526725769, "learning_rate": 4.6023643195656164e-06, "loss": 0.0176, "step": 69540 }, { "grad_norm": 0.19003310799598694, "learning_rate": 4.593706649089768e-06, "loss": 0.0105, "step": 69550 }, { "grad_norm": 0.23039956390857697, "learning_rate": 4.585056737307597e-06, "loss": 0.0143, "step": 69560 }, { "grad_norm": 0.32472726702690125, "learning_rate": 4.576414585697103e-06, "loss": 0.0129, "step": 69570 }, { "grad_norm": 0.4147892892360687, "learning_rate": 4.567780195735044e-06, "loss": 0.0169, "step": 69580 }, { "grad_norm": 0.2068846970796585, "learning_rate": 4.559153568896757e-06, "loss": 0.0131, "step": 69590 }, { "grad_norm": 0.48933833837509155, "learning_rate": 4.550534706656329e-06, "loss": 0.0173, "step": 69600 }, { "grad_norm": 0.29751262068748474, "learning_rate": 4.541923610486465e-06, "loss": 0.0174, "step": 69610 }, { "grad_norm": 0.4001093804836273, "learning_rate": 4.533320281858578e-06, "loss": 0.0166, "step": 69620 }, { "grad_norm": 0.234378844499588, "learning_rate": 4.52472472224274e-06, "loss": 0.0175, "step": 69630 }, { "grad_norm": 0.37466174364089966, "learning_rate": 4.51613693310769e-06, "loss": 0.0182, "step": 69640 }, { "grad_norm": 0.2424078732728958, "learning_rate": 4.507556915920868e-06, "loss": 0.0171, "step": 69650 }, { "grad_norm": 0.2017093002796173, "learning_rate": 4.498984672148332e-06, "loss": 0.0182, "step": 69660 }, { "grad_norm": 0.21784457564353943, "learning_rate": 4.490420203254864e-06, "loss": 0.0216, "step": 69670 }, { "grad_norm": 0.4437245726585388, "learning_rate": 4.481863510703893e-06, "loss": 0.0204, "step": 69680 }, { "grad_norm": 0.15142002701759338, "learning_rate": 4.473314595957523e-06, "loss": 0.0211, "step": 69690 }, { "grad_norm": 0.3484920263290405, "learning_rate": 4.464773460476535e-06, "loss": 0.0153, "step": 69700 }, { "grad_norm": 0.2089306116104126, "learning_rate": 4.456240105720372e-06, "loss": 0.0157, "step": 69710 }, { "grad_norm": 0.1724369376897812, "learning_rate": 4.4477145331471405e-06, "loss": 0.0265, "step": 69720 }, { "grad_norm": 0.20513761043548584, "learning_rate": 4.439196744213653e-06, "loss": 0.0142, "step": 69730 }, { "grad_norm": 0.2843133509159088, "learning_rate": 4.430686740375339e-06, "loss": 0.0176, "step": 69740 }, { "grad_norm": 0.25695371627807617, "learning_rate": 4.422184523086342e-06, "loss": 0.0164, "step": 69750 }, { "grad_norm": 0.1742795705795288, "learning_rate": 4.41369009379946e-06, "loss": 0.0116, "step": 69760 }, { "grad_norm": 0.2520606517791748, "learning_rate": 4.405203453966139e-06, "loss": 0.016, "step": 69770 }, { "grad_norm": 0.3163359761238098, "learning_rate": 4.396724605036539e-06, "loss": 0.0187, "step": 69780 }, { "grad_norm": 0.3126355707645416, "learning_rate": 4.388253548459437e-06, "loss": 0.0193, "step": 69790 }, { "grad_norm": 0.34828826785087585, "learning_rate": 4.3797902856823395e-06, "loss": 0.0164, "step": 69800 }, { "grad_norm": 0.16275396943092346, "learning_rate": 4.371334818151357e-06, "loss": 0.0131, "step": 69810 }, { "grad_norm": 0.39477530121803284, "learning_rate": 4.362887147311306e-06, "loss": 0.0158, "step": 69820 }, { "grad_norm": 0.1831512451171875, "learning_rate": 4.354447274605672e-06, "loss": 0.0138, "step": 69830 }, { "grad_norm": 0.33027008175849915, "learning_rate": 4.346015201476572e-06, "loss": 0.0259, "step": 69840 }, { "grad_norm": 0.21050769090652466, "learning_rate": 4.337590929364855e-06, "loss": 0.0148, "step": 69850 }, { "grad_norm": 0.15845876932144165, "learning_rate": 4.329174459709973e-06, "loss": 0.0162, "step": 69860 }, { "grad_norm": 0.23586958646774292, "learning_rate": 4.320765793950071e-06, "loss": 0.0177, "step": 69870 }, { "grad_norm": 0.32878732681274414, "learning_rate": 4.312364933521962e-06, "loss": 0.0206, "step": 69880 }, { "grad_norm": 0.309943825006485, "learning_rate": 4.303971879861129e-06, "loss": 0.0166, "step": 69890 }, { "grad_norm": 0.3803987503051758, "learning_rate": 4.295586634401716e-06, "loss": 0.0203, "step": 69900 }, { "grad_norm": 0.30482780933380127, "learning_rate": 4.287209198576536e-06, "loss": 0.0149, "step": 69910 }, { "grad_norm": 0.5861318707466125, "learning_rate": 4.278839573817045e-06, "loss": 0.0224, "step": 69920 }, { "grad_norm": 0.36344826221466064, "learning_rate": 4.270477761553399e-06, "loss": 0.0173, "step": 69930 }, { "grad_norm": 0.39129960536956787, "learning_rate": 4.262123763214393e-06, "loss": 0.0121, "step": 69940 }, { "grad_norm": 0.21858350932598114, "learning_rate": 4.253777580227508e-06, "loss": 0.0182, "step": 69950 }, { "grad_norm": 0.2878131568431854, "learning_rate": 4.2454392140188755e-06, "loss": 0.0159, "step": 69960 }, { "grad_norm": 0.26318320631980896, "learning_rate": 4.2371086660132785e-06, "loss": 0.0169, "step": 69970 }, { "grad_norm": 0.3388229012489319, "learning_rate": 4.228785937634205e-06, "loss": 0.0135, "step": 69980 }, { "grad_norm": 0.3741840422153473, "learning_rate": 4.220471030303758e-06, "loss": 0.014, "step": 69990 }, { "grad_norm": 0.24332080781459808, "learning_rate": 4.212163945442754e-06, "loss": 0.0157, "step": 70000 }, { "grad_norm": 0.29786932468414307, "learning_rate": 4.203864684470621e-06, "loss": 0.016, "step": 70010 }, { "grad_norm": 0.37726134061813354, "learning_rate": 4.19557324880549e-06, "loss": 0.021, "step": 70020 }, { "grad_norm": 0.24612653255462646, "learning_rate": 4.187289639864145e-06, "loss": 0.0162, "step": 70030 }, { "grad_norm": 0.17790257930755615, "learning_rate": 4.1790138590619974e-06, "loss": 0.0115, "step": 70040 }, { "grad_norm": 0.3713638186454773, "learning_rate": 4.170745907813195e-06, "loss": 0.0208, "step": 70050 }, { "grad_norm": 0.31482651829719543, "learning_rate": 4.162485787530479e-06, "loss": 0.0153, "step": 70060 }, { "grad_norm": 0.37939533591270447, "learning_rate": 4.154233499625282e-06, "loss": 0.0164, "step": 70070 }, { "grad_norm": 0.20425370335578918, "learning_rate": 4.145989045507692e-06, "loss": 0.0164, "step": 70080 }, { "grad_norm": 0.2579162120819092, "learning_rate": 4.1377524265864666e-06, "loss": 0.0198, "step": 70090 }, { "grad_norm": 0.15014062821865082, "learning_rate": 4.1295236442690175e-06, "loss": 0.0133, "step": 70100 }, { "grad_norm": 0.4289488196372986, "learning_rate": 4.121302699961421e-06, "loss": 0.0171, "step": 70110 }, { "grad_norm": 0.2165112942457199, "learning_rate": 4.113089595068403e-06, "loss": 0.0231, "step": 70120 }, { "grad_norm": 0.3001466393470764, "learning_rate": 4.104884330993364e-06, "loss": 0.0179, "step": 70130 }, { "grad_norm": 0.31149646639823914, "learning_rate": 4.0966869091383585e-06, "loss": 0.0245, "step": 70140 }, { "grad_norm": 0.28897127509117126, "learning_rate": 4.088497330904101e-06, "loss": 0.0177, "step": 70150 }, { "grad_norm": 0.15043869614601135, "learning_rate": 4.080315597689976e-06, "loss": 0.0153, "step": 70160 }, { "grad_norm": 0.19238430261611938, "learning_rate": 4.072141710893995e-06, "loss": 0.023, "step": 70170 }, { "grad_norm": 0.17455731332302094, "learning_rate": 4.063975671912879e-06, "loss": 0.0287, "step": 70180 }, { "grad_norm": 0.3003777861595154, "learning_rate": 4.055817482141949e-06, "loss": 0.0121, "step": 70190 }, { "grad_norm": 0.2526649236679077, "learning_rate": 4.047667142975259e-06, "loss": 0.0141, "step": 70200 }, { "grad_norm": 0.23188172280788422, "learning_rate": 4.039524655805443e-06, "loss": 0.0198, "step": 70210 }, { "grad_norm": 0.3458250164985657, "learning_rate": 4.03139002202384e-06, "loss": 0.012, "step": 70220 }, { "grad_norm": 0.26944753527641296, "learning_rate": 4.023263243020447e-06, "loss": 0.0197, "step": 70230 }, { "grad_norm": 0.39851710200309753, "learning_rate": 4.015144320183884e-06, "loss": 0.0221, "step": 70240 }, { "grad_norm": 0.18492019176483154, "learning_rate": 4.007033254901482e-06, "loss": 0.0157, "step": 70250 }, { "grad_norm": 0.2644301950931549, "learning_rate": 3.9989300485591795e-06, "loss": 0.0227, "step": 70260 }, { "grad_norm": 0.656489908695221, "learning_rate": 3.990834702541601e-06, "loss": 0.0232, "step": 70270 }, { "grad_norm": 0.2799689769744873, "learning_rate": 3.982747218232019e-06, "loss": 0.016, "step": 70280 }, { "grad_norm": 0.24427762627601624, "learning_rate": 3.97466759701236e-06, "loss": 0.0188, "step": 70290 }, { "grad_norm": 0.4636442959308624, "learning_rate": 3.966595840263226e-06, "loss": 0.0179, "step": 70300 }, { "grad_norm": 0.3598453402519226, "learning_rate": 3.958531949363836e-06, "loss": 0.0145, "step": 70310 }, { "grad_norm": 0.2181224226951599, "learning_rate": 3.950475925692098e-06, "loss": 0.0208, "step": 70320 }, { "grad_norm": 0.21912650763988495, "learning_rate": 3.9424277706245685e-06, "loss": 0.0151, "step": 70330 }, { "grad_norm": 0.15490442514419556, "learning_rate": 3.934387485536451e-06, "loss": 0.0142, "step": 70340 }, { "grad_norm": 0.19498999416828156, "learning_rate": 3.926355071801619e-06, "loss": 0.0167, "step": 70350 }, { "grad_norm": 0.3364621698856354, "learning_rate": 3.9183305307925965e-06, "loss": 0.0174, "step": 70360 }, { "grad_norm": 0.2469261735677719, "learning_rate": 3.910313863880533e-06, "loss": 0.014, "step": 70370 }, { "grad_norm": 0.1831824630498886, "learning_rate": 3.902305072435292e-06, "loss": 0.0129, "step": 70380 }, { "grad_norm": 0.2251203954219818, "learning_rate": 3.894304157825329e-06, "loss": 0.0151, "step": 70390 }, { "grad_norm": 0.2677046060562134, "learning_rate": 3.886311121417791e-06, "loss": 0.0143, "step": 70400 }, { "grad_norm": 0.30971524119377136, "learning_rate": 3.878325964578472e-06, "loss": 0.0177, "step": 70410 }, { "grad_norm": 0.34124013781547546, "learning_rate": 3.870348688671815e-06, "loss": 0.0118, "step": 70420 }, { "grad_norm": 0.37855613231658936, "learning_rate": 3.862379295060931e-06, "loss": 0.015, "step": 70430 }, { "grad_norm": 0.1807985156774521, "learning_rate": 3.854417785107534e-06, "loss": 0.0213, "step": 70440 }, { "grad_norm": 0.4885012209415436, "learning_rate": 3.8464641601720755e-06, "loss": 0.0168, "step": 70450 }, { "grad_norm": 0.24610042572021484, "learning_rate": 3.838518421613579e-06, "loss": 0.0176, "step": 70460 }, { "grad_norm": 0.2952112555503845, "learning_rate": 3.830580570789766e-06, "loss": 0.0165, "step": 70470 }, { "grad_norm": 0.18081779778003693, "learning_rate": 3.822650609057005e-06, "loss": 0.0143, "step": 70480 }, { "grad_norm": 0.24742747843265533, "learning_rate": 3.814728537770285e-06, "loss": 0.0203, "step": 70490 }, { "grad_norm": 0.3334287405014038, "learning_rate": 3.806814358283306e-06, "loss": 0.0182, "step": 70500 }, { "grad_norm": 0.41645580530166626, "learning_rate": 3.7989080719483596e-06, "loss": 0.0154, "step": 70510 }, { "grad_norm": 0.19751837849617004, "learning_rate": 3.7910096801164143e-06, "loss": 0.0164, "step": 70520 }, { "grad_norm": 0.2751290202140808, "learning_rate": 3.7831191841371016e-06, "loss": 0.0169, "step": 70530 }, { "grad_norm": 0.32044529914855957, "learning_rate": 3.775236585358688e-06, "loss": 0.0203, "step": 70540 }, { "grad_norm": 0.3079614043235779, "learning_rate": 3.7673618851280843e-06, "loss": 0.0201, "step": 70550 }, { "grad_norm": 0.44855862855911255, "learning_rate": 3.759495084790887e-06, "loss": 0.0188, "step": 70560 }, { "grad_norm": 0.2681485116481781, "learning_rate": 3.751636185691282e-06, "loss": 0.0189, "step": 70570 }, { "grad_norm": 0.17210569977760315, "learning_rate": 3.7437851891721607e-06, "loss": 0.0154, "step": 70580 }, { "grad_norm": 0.38253921270370483, "learning_rate": 3.7359420965750404e-06, "loss": 0.0171, "step": 70590 }, { "grad_norm": 0.22839754819869995, "learning_rate": 3.7281069092400922e-06, "loss": 0.0144, "step": 70600 }, { "grad_norm": 0.31952229142189026, "learning_rate": 3.7202796285061348e-06, "loss": 0.0152, "step": 70610 }, { "grad_norm": 0.24044311046600342, "learning_rate": 3.712460255710637e-06, "loss": 0.0204, "step": 70620 }, { "grad_norm": 0.2352113127708435, "learning_rate": 3.704648792189719e-06, "loss": 0.0196, "step": 70630 }, { "grad_norm": 0.29383063316345215, "learning_rate": 3.696845239278124e-06, "loss": 0.0204, "step": 70640 }, { "grad_norm": 0.26826995611190796, "learning_rate": 3.689049598309302e-06, "loss": 0.0191, "step": 70650 }, { "grad_norm": 0.3852885067462921, "learning_rate": 3.681261870615288e-06, "loss": 0.0285, "step": 70660 }, { "grad_norm": 0.2604270279407501, "learning_rate": 3.6734820575268004e-06, "loss": 0.0252, "step": 70670 }, { "grad_norm": 0.21957604587078094, "learning_rate": 3.665710160373209e-06, "loss": 0.0166, "step": 70680 }, { "grad_norm": 0.32798463106155396, "learning_rate": 3.65794618048248e-06, "loss": 0.0169, "step": 70690 }, { "grad_norm": 0.22642111778259277, "learning_rate": 3.6501901191813125e-06, "loss": 0.0132, "step": 70700 }, { "grad_norm": 0.24967174232006073, "learning_rate": 3.642441977794975e-06, "loss": 0.0162, "step": 70710 }, { "grad_norm": 0.29597553610801697, "learning_rate": 3.634701757647424e-06, "loss": 0.0178, "step": 70720 }, { "grad_norm": 0.21153436601161957, "learning_rate": 3.6269694600612468e-06, "loss": 0.0138, "step": 70730 }, { "grad_norm": 0.2119734138250351, "learning_rate": 3.619245086357681e-06, "loss": 0.0149, "step": 70740 }, { "grad_norm": 0.4213539659976959, "learning_rate": 3.611528637856615e-06, "loss": 0.0179, "step": 70750 }, { "grad_norm": 0.36106255650520325, "learning_rate": 3.6038201158765884e-06, "loss": 0.0162, "step": 70760 }, { "grad_norm": 0.5167421102523804, "learning_rate": 3.5961195217347534e-06, "loss": 0.0309, "step": 70770 }, { "grad_norm": 0.2321782261133194, "learning_rate": 3.588426856746946e-06, "loss": 0.0185, "step": 70780 }, { "grad_norm": 0.1715077906847, "learning_rate": 3.5807421222276316e-06, "loss": 0.0218, "step": 70790 }, { "grad_norm": 0.3279902935028076, "learning_rate": 3.5730653194899156e-06, "loss": 0.0183, "step": 70800 }, { "grad_norm": 0.27388736605644226, "learning_rate": 3.565396449845554e-06, "loss": 0.0171, "step": 70810 }, { "grad_norm": 0.46950042247772217, "learning_rate": 3.557735514604954e-06, "loss": 0.0222, "step": 70820 }, { "grad_norm": 0.08886988461017609, "learning_rate": 3.5500825150771633e-06, "loss": 0.0137, "step": 70830 }, { "grad_norm": 0.14148686826229095, "learning_rate": 3.5424374525698466e-06, "loss": 0.0135, "step": 70840 }, { "grad_norm": 0.19595898687839508, "learning_rate": 3.5348003283893704e-06, "loss": 0.0145, "step": 70850 }, { "grad_norm": 0.17052504420280457, "learning_rate": 3.527171143840685e-06, "loss": 0.0136, "step": 70860 }, { "grad_norm": 0.2151554375886917, "learning_rate": 3.519549900227409e-06, "loss": 0.0212, "step": 70870 }, { "grad_norm": 0.41591939330101013, "learning_rate": 3.511936598851828e-06, "loss": 0.0167, "step": 70880 }, { "grad_norm": 0.19160869717597961, "learning_rate": 3.504331241014813e-06, "loss": 0.0151, "step": 70890 }, { "grad_norm": 0.24404388666152954, "learning_rate": 3.4967338280159414e-06, "loss": 0.014, "step": 70900 }, { "grad_norm": 0.4005235731601715, "learning_rate": 3.4891443611533846e-06, "loss": 0.0197, "step": 70910 }, { "grad_norm": 0.291115403175354, "learning_rate": 3.481562841723984e-06, "loss": 0.0236, "step": 70920 }, { "grad_norm": 0.2580549120903015, "learning_rate": 3.473989271023215e-06, "loss": 0.0147, "step": 70930 }, { "grad_norm": 0.2719067633152008, "learning_rate": 3.4664236503451854e-06, "loss": 0.0178, "step": 70940 }, { "grad_norm": 0.20386257767677307, "learning_rate": 3.4588659809826674e-06, "loss": 0.0159, "step": 70950 }, { "grad_norm": 0.27045121788978577, "learning_rate": 3.451316264227039e-06, "loss": 0.0259, "step": 70960 }, { "grad_norm": 0.14919060468673706, "learning_rate": 3.4437745013683574e-06, "loss": 0.0161, "step": 70970 }, { "grad_norm": 0.2448524832725525, "learning_rate": 3.4362406936952916e-06, "loss": 0.0137, "step": 70980 }, { "grad_norm": 0.3171704411506653, "learning_rate": 3.4287148424951786e-06, "loss": 0.0182, "step": 70990 }, { "grad_norm": 0.30382242798805237, "learning_rate": 3.4211969490539617e-06, "loss": 0.0153, "step": 71000 }, { "grad_norm": 0.4180153012275696, "learning_rate": 3.4136870146562584e-06, "loss": 0.0155, "step": 71010 }, { "grad_norm": 0.29634225368499756, "learning_rate": 3.406185040585308e-06, "loss": 0.0164, "step": 71020 }, { "grad_norm": 0.31404247879981995, "learning_rate": 3.3986910281229966e-06, "loss": 0.019, "step": 71030 }, { "grad_norm": 0.3851989805698395, "learning_rate": 3.391204978549828e-06, "loss": 0.0204, "step": 71040 }, { "grad_norm": 0.2625066339969635, "learning_rate": 3.3837268931449785e-06, "loss": 0.0266, "step": 71050 }, { "grad_norm": 0.1855875551700592, "learning_rate": 3.376256773186248e-06, "loss": 0.0228, "step": 71060 }, { "grad_norm": 0.2453867793083191, "learning_rate": 3.3687946199500664e-06, "loss": 0.0164, "step": 71070 }, { "grad_norm": 0.23301634192466736, "learning_rate": 3.3613404347115295e-06, "loss": 0.0155, "step": 71080 }, { "grad_norm": 0.1833430528640747, "learning_rate": 3.35389421874433e-06, "loss": 0.0134, "step": 71090 }, { "grad_norm": 0.13246743381023407, "learning_rate": 3.3464559733208446e-06, "loss": 0.0154, "step": 71100 }, { "grad_norm": 0.2937585115432739, "learning_rate": 3.3390256997120505e-06, "loss": 0.0204, "step": 71110 }, { "grad_norm": 0.2830008268356323, "learning_rate": 3.331603399187583e-06, "loss": 0.0201, "step": 71120 }, { "grad_norm": 0.25914162397384644, "learning_rate": 3.324189073015721e-06, "loss": 0.0177, "step": 71130 }, { "grad_norm": 0.17751160264015198, "learning_rate": 3.3167827224633465e-06, "loss": 0.0161, "step": 71140 }, { "grad_norm": 0.178306445479393, "learning_rate": 3.309384348796024e-06, "loss": 0.0137, "step": 71150 }, { "grad_norm": 0.29042428731918335, "learning_rate": 3.301993953277921e-06, "loss": 0.0157, "step": 71160 }, { "grad_norm": 0.30823296308517456, "learning_rate": 3.29461153717186e-06, "loss": 0.0142, "step": 71170 }, { "grad_norm": 0.22588150203227997, "learning_rate": 3.287237101739293e-06, "loss": 0.0214, "step": 71180 }, { "grad_norm": 0.37267839908599854, "learning_rate": 3.2798706482403075e-06, "loss": 0.0215, "step": 71190 }, { "grad_norm": 0.3202255368232727, "learning_rate": 3.2725121779336285e-06, "loss": 0.0211, "step": 71200 }, { "grad_norm": 0.335488885641098, "learning_rate": 3.265161692076618e-06, "loss": 0.0207, "step": 71210 }, { "grad_norm": 0.1904083788394928, "learning_rate": 3.257819191925282e-06, "loss": 0.0131, "step": 71220 }, { "grad_norm": 0.2457454949617386, "learning_rate": 3.2504846787342392e-06, "loss": 0.0168, "step": 71230 }, { "grad_norm": 0.19810323417186737, "learning_rate": 3.243158153756759e-06, "loss": 0.0148, "step": 71240 }, { "grad_norm": 0.19986845552921295, "learning_rate": 3.2358396182447514e-06, "loss": 0.0146, "step": 71250 }, { "grad_norm": 0.2842155992984772, "learning_rate": 3.2285290734487496e-06, "loss": 0.013, "step": 71260 }, { "grad_norm": 0.2840595245361328, "learning_rate": 3.221226520617926e-06, "loss": 0.0186, "step": 71270 }, { "grad_norm": 0.1970677524805069, "learning_rate": 3.213931961000094e-06, "loss": 0.0232, "step": 71280 }, { "grad_norm": 0.47148674726486206, "learning_rate": 3.2066453958416733e-06, "loss": 0.014, "step": 71290 }, { "grad_norm": 0.21147266030311584, "learning_rate": 3.199366826387773e-06, "loss": 0.0173, "step": 71300 }, { "grad_norm": 0.258396178483963, "learning_rate": 3.192096253882071e-06, "loss": 0.0166, "step": 71310 }, { "grad_norm": 0.30201220512390137, "learning_rate": 3.1848336795669177e-06, "loss": 0.0162, "step": 71320 }, { "grad_norm": 0.31256839632987976, "learning_rate": 3.1775791046833035e-06, "loss": 0.0197, "step": 71330 }, { "grad_norm": 0.09702008217573166, "learning_rate": 3.170332530470804e-06, "loss": 0.018, "step": 71340 }, { "grad_norm": 0.33992552757263184, "learning_rate": 3.1630939581677012e-06, "loss": 0.02, "step": 71350 }, { "grad_norm": 0.33712249994277954, "learning_rate": 3.155863389010838e-06, "loss": 0.0189, "step": 71360 }, { "grad_norm": 0.2564527988433838, "learning_rate": 3.1486408242357323e-06, "loss": 0.0214, "step": 71370 }, { "grad_norm": 0.4821546673774719, "learning_rate": 3.1414262650765248e-06, "loss": 0.0369, "step": 71380 }, { "grad_norm": 0.3280326724052429, "learning_rate": 3.134219712765979e-06, "loss": 0.0172, "step": 71390 }, { "grad_norm": 0.643653392791748, "learning_rate": 3.1270211685355044e-06, "loss": 0.0203, "step": 71400 }, { "grad_norm": 0.1648021638393402, "learning_rate": 3.1198306336151338e-06, "loss": 0.0198, "step": 71410 }, { "grad_norm": 0.2777416706085205, "learning_rate": 3.11264810923354e-06, "loss": 0.018, "step": 71420 }, { "grad_norm": 0.1521771252155304, "learning_rate": 3.105473596618008e-06, "loss": 0.0128, "step": 71430 }, { "grad_norm": 0.3307570219039917, "learning_rate": 3.0983070969944683e-06, "loss": 0.0186, "step": 71440 }, { "grad_norm": 0.36263951659202576, "learning_rate": 3.09114861158748e-06, "loss": 0.0152, "step": 71450 }, { "grad_norm": 0.19639275968074799, "learning_rate": 3.0839981416202314e-06, "loss": 0.0176, "step": 71460 }, { "grad_norm": 0.29072147607803345, "learning_rate": 3.0768556883145507e-06, "loss": 0.0118, "step": 71470 }, { "grad_norm": 0.3347247242927551, "learning_rate": 3.0697212528908834e-06, "loss": 0.0146, "step": 71480 }, { "grad_norm": 0.21707841753959656, "learning_rate": 3.0625948365683e-06, "loss": 0.0153, "step": 71490 }, { "grad_norm": 0.2980087995529175, "learning_rate": 3.055476440564531e-06, "loss": 0.0183, "step": 71500 }, { "grad_norm": 0.2911650538444519, "learning_rate": 3.0483660660958924e-06, "loss": 0.0169, "step": 71510 }, { "grad_norm": 0.28493833541870117, "learning_rate": 3.0412637143773624e-06, "loss": 0.024, "step": 71520 }, { "grad_norm": 0.32811304926872253, "learning_rate": 3.034169386622554e-06, "loss": 0.0177, "step": 71530 }, { "grad_norm": 0.3046318292617798, "learning_rate": 3.027083084043658e-06, "loss": 0.0148, "step": 71540 }, { "grad_norm": 0.33000192046165466, "learning_rate": 3.0200048078515676e-06, "loss": 0.0168, "step": 71550 }, { "grad_norm": 0.4950631260871887, "learning_rate": 3.012934559255742e-06, "loss": 0.0144, "step": 71560 }, { "grad_norm": 0.27559253573417664, "learning_rate": 3.0058723394642994e-06, "loss": 0.0223, "step": 71570 }, { "grad_norm": 0.4530442953109741, "learning_rate": 2.998818149683985e-06, "loss": 0.0236, "step": 71580 }, { "grad_norm": 0.2738150358200073, "learning_rate": 2.9917719911201627e-06, "loss": 0.0193, "step": 71590 }, { "grad_norm": 0.30130162835121155, "learning_rate": 2.9847338649768352e-06, "loss": 0.0174, "step": 71600 }, { "grad_norm": 0.2773135006427765, "learning_rate": 2.977703772456608e-06, "loss": 0.0129, "step": 71610 }, { "grad_norm": 0.25101983547210693, "learning_rate": 2.9706817147607535e-06, "loss": 0.0143, "step": 71620 }, { "grad_norm": 0.2794308662414551, "learning_rate": 2.9636676930891393e-06, "loss": 0.0173, "step": 71630 }, { "grad_norm": 0.16775283217430115, "learning_rate": 2.9566617086402625e-06, "loss": 0.0142, "step": 71640 }, { "grad_norm": 0.2193041890859604, "learning_rate": 2.9496637626112655e-06, "loss": 0.0148, "step": 71650 }, { "grad_norm": 0.287283331155777, "learning_rate": 2.9426738561979027e-06, "loss": 0.0154, "step": 71660 }, { "grad_norm": 0.27365636825561523, "learning_rate": 2.9356919905945524e-06, "loss": 0.0316, "step": 71670 }, { "grad_norm": 0.2716982662677765, "learning_rate": 2.928718166994243e-06, "loss": 0.0166, "step": 71680 }, { "grad_norm": 0.3540399670600891, "learning_rate": 2.9217523865885833e-06, "loss": 0.0219, "step": 71690 }, { "grad_norm": 0.3091144859790802, "learning_rate": 2.9147946505678537e-06, "loss": 0.0179, "step": 71700 }, { "grad_norm": 0.20154418051242828, "learning_rate": 2.9078449601209313e-06, "loss": 0.0131, "step": 71710 }, { "grad_norm": 0.3097274899482727, "learning_rate": 2.900903316435333e-06, "loss": 0.0182, "step": 71720 }, { "grad_norm": 0.3835768699645996, "learning_rate": 2.8939697206971983e-06, "loss": 0.0224, "step": 71730 }, { "grad_norm": 0.442935973405838, "learning_rate": 2.8870441740912746e-06, "loss": 0.0182, "step": 71740 }, { "grad_norm": 0.3432784378528595, "learning_rate": 2.8801266778009762e-06, "loss": 0.0159, "step": 71750 }, { "grad_norm": 0.24721525609493256, "learning_rate": 2.8732172330082798e-06, "loss": 0.0196, "step": 71760 }, { "grad_norm": 0.2362542599439621, "learning_rate": 2.8663158408938517e-06, "loss": 0.016, "step": 71770 }, { "grad_norm": 0.2771233916282654, "learning_rate": 2.859422502636938e-06, "loss": 0.0173, "step": 71780 }, { "grad_norm": 0.2550264596939087, "learning_rate": 2.8525372194154076e-06, "loss": 0.0166, "step": 71790 }, { "grad_norm": 0.24166733026504517, "learning_rate": 2.8456599924057913e-06, "loss": 0.015, "step": 71800 }, { "grad_norm": 0.27374398708343506, "learning_rate": 2.8387908227831995e-06, "loss": 0.0197, "step": 71810 }, { "grad_norm": 0.25194716453552246, "learning_rate": 2.831929711721404e-06, "loss": 0.0179, "step": 71820 }, { "grad_norm": 0.17826175689697266, "learning_rate": 2.825076660392767e-06, "loss": 0.0152, "step": 71830 }, { "grad_norm": 0.260496586561203, "learning_rate": 2.8182316699682908e-06, "loss": 0.021, "step": 71840 }, { "grad_norm": 0.23308326303958893, "learning_rate": 2.811394741617601e-06, "loss": 0.0211, "step": 71850 }, { "grad_norm": 0.4791697561740875, "learning_rate": 2.8045658765089356e-06, "loss": 0.0261, "step": 71860 }, { "grad_norm": 0.33935636281967163, "learning_rate": 2.7977450758091605e-06, "loss": 0.029, "step": 71870 }, { "grad_norm": 0.22907324135303497, "learning_rate": 2.790932340683783e-06, "loss": 0.0187, "step": 71880 }, { "grad_norm": 0.23337262868881226, "learning_rate": 2.7841276722968823e-06, "loss": 0.0183, "step": 71890 }, { "grad_norm": 0.2986617982387543, "learning_rate": 2.7773310718112067e-06, "loss": 0.0188, "step": 71900 }, { "grad_norm": 0.34343868494033813, "learning_rate": 2.7705425403881102e-06, "loss": 0.0212, "step": 71910 }, { "grad_norm": 0.1464129090309143, "learning_rate": 2.763762079187565e-06, "loss": 0.0134, "step": 71920 }, { "grad_norm": 0.2502942979335785, "learning_rate": 2.756989689368178e-06, "loss": 0.0191, "step": 71930 }, { "grad_norm": 0.1710071861743927, "learning_rate": 2.750225372087134e-06, "loss": 0.017, "step": 71940 }, { "grad_norm": 0.2854923605918884, "learning_rate": 2.7434691285003033e-06, "loss": 0.0145, "step": 71950 }, { "grad_norm": 0.3251780569553375, "learning_rate": 2.7367209597621178e-06, "loss": 0.0164, "step": 71960 }, { "grad_norm": 0.25510865449905396, "learning_rate": 2.729980867025683e-06, "loss": 0.0175, "step": 71970 }, { "grad_norm": 0.2706867456436157, "learning_rate": 2.7232488514426724e-06, "loss": 0.0172, "step": 71980 }, { "grad_norm": 0.3245527744293213, "learning_rate": 2.7165249141634097e-06, "loss": 0.0182, "step": 71990 }, { "grad_norm": 0.3057931065559387, "learning_rate": 2.709809056336837e-06, "loss": 0.0236, "step": 72000 }, { "grad_norm": 0.21209022402763367, "learning_rate": 2.703101279110498e-06, "loss": 0.0176, "step": 72010 }, { "grad_norm": 0.5233945250511169, "learning_rate": 2.696401583630587e-06, "loss": 0.0222, "step": 72020 }, { "grad_norm": 0.40770450234413147, "learning_rate": 2.689709971041887e-06, "loss": 0.0157, "step": 72030 }, { "grad_norm": 0.3470188081264496, "learning_rate": 2.6830264424878126e-06, "loss": 0.0197, "step": 72040 }, { "grad_norm": 0.2427050918340683, "learning_rate": 2.676350999110394e-06, "loss": 0.03, "step": 72050 }, { "grad_norm": 0.16337525844573975, "learning_rate": 2.6696836420502856e-06, "loss": 0.022, "step": 72060 }, { "grad_norm": 0.2592671513557434, "learning_rate": 2.663024372446765e-06, "loss": 0.017, "step": 72070 }, { "grad_norm": 0.41292545199394226, "learning_rate": 2.6563731914377055e-06, "loss": 0.0214, "step": 72080 }, { "grad_norm": 0.24360167980194092, "learning_rate": 2.6497301001596087e-06, "loss": 0.021, "step": 72090 }, { "grad_norm": 0.3662203848361969, "learning_rate": 2.643095099747611e-06, "loss": 0.0254, "step": 72100 }, { "grad_norm": 0.18710191547870636, "learning_rate": 2.636468191335445e-06, "loss": 0.0151, "step": 72110 }, { "grad_norm": 0.2889966070652008, "learning_rate": 2.629849376055471e-06, "loss": 0.0181, "step": 72120 }, { "grad_norm": 0.18799012899398804, "learning_rate": 2.6232386550386678e-06, "loss": 0.0172, "step": 72130 }, { "grad_norm": 0.22835159301757812, "learning_rate": 2.6166360294146097e-06, "loss": 0.0165, "step": 72140 }, { "grad_norm": 0.23034121096134186, "learning_rate": 2.6100415003115275e-06, "loss": 0.0189, "step": 72150 }, { "grad_norm": 0.21052750945091248, "learning_rate": 2.603455068856225e-06, "loss": 0.02, "step": 72160 }, { "grad_norm": 0.1912398338317871, "learning_rate": 2.5968767361741584e-06, "loss": 0.0179, "step": 72170 }, { "grad_norm": 0.21913348138332367, "learning_rate": 2.590306503389378e-06, "loss": 0.0197, "step": 72180 }, { "grad_norm": 0.38790199160575867, "learning_rate": 2.583744371624558e-06, "loss": 0.0271, "step": 72190 }, { "grad_norm": 0.3676619529724121, "learning_rate": 2.57719034200099e-06, "loss": 0.014, "step": 72200 }, { "grad_norm": 0.2634883224964142, "learning_rate": 2.570644415638568e-06, "loss": 0.0168, "step": 72210 }, { "grad_norm": 0.10976766794919968, "learning_rate": 2.5641065936558295e-06, "loss": 0.0227, "step": 72220 }, { "grad_norm": 0.49691784381866455, "learning_rate": 2.557576877169898e-06, "loss": 0.0231, "step": 72230 }, { "grad_norm": 0.22716248035430908, "learning_rate": 2.5510552672965205e-06, "loss": 0.0141, "step": 72240 }, { "grad_norm": 0.22920207679271698, "learning_rate": 2.544541765150077e-06, "loss": 0.0156, "step": 72250 }, { "grad_norm": 0.22515305876731873, "learning_rate": 2.5380363718435165e-06, "loss": 0.0236, "step": 72260 }, { "grad_norm": 0.18740214407444, "learning_rate": 2.5315390884884714e-06, "loss": 0.0149, "step": 72270 }, { "grad_norm": 0.3407123386859894, "learning_rate": 2.5250499161951214e-06, "loss": 0.0165, "step": 72280 }, { "grad_norm": 0.31501880288124084, "learning_rate": 2.518568856072301e-06, "loss": 0.0123, "step": 72290 }, { "grad_norm": 0.33399710059165955, "learning_rate": 2.512095909227441e-06, "loss": 0.0183, "step": 72300 }, { "grad_norm": 0.24394822120666504, "learning_rate": 2.50563107676659e-06, "loss": 0.0238, "step": 72310 }, { "grad_norm": 0.21698759496212006, "learning_rate": 2.4991743597944152e-06, "loss": 0.0189, "step": 72320 }, { "grad_norm": 0.17845414578914642, "learning_rate": 2.492725759414205e-06, "loss": 0.018, "step": 72330 }, { "grad_norm": 0.2515830993652344, "learning_rate": 2.4862852767278234e-06, "loss": 0.0122, "step": 72340 }, { "grad_norm": 0.2488960474729538, "learning_rate": 2.479852912835784e-06, "loss": 0.0118, "step": 72350 }, { "grad_norm": 0.282126247882843, "learning_rate": 2.4734286688372075e-06, "loss": 0.0203, "step": 72360 }, { "grad_norm": 0.3240048885345459, "learning_rate": 2.467012545829811e-06, "loss": 0.0267, "step": 72370 }, { "grad_norm": 0.21924278140068054, "learning_rate": 2.460604544909945e-06, "loss": 0.0149, "step": 72380 }, { "grad_norm": 0.2806157171726227, "learning_rate": 2.454204667172555e-06, "loss": 0.0112, "step": 72390 }, { "grad_norm": 0.3669748604297638, "learning_rate": 2.447812913711217e-06, "loss": 0.0184, "step": 72400 }, { "grad_norm": 0.16330480575561523, "learning_rate": 2.44142928561808e-06, "loss": 0.0197, "step": 72410 }, { "grad_norm": 0.23446010053157806, "learning_rate": 2.435053783983965e-06, "loss": 0.0161, "step": 72420 }, { "grad_norm": 0.13973186910152435, "learning_rate": 2.4286864098982453e-06, "loss": 0.0164, "step": 72430 }, { "grad_norm": 0.24064160883426666, "learning_rate": 2.4223271644489397e-06, "loss": 0.0169, "step": 72440 }, { "grad_norm": 0.20304720103740692, "learning_rate": 2.415976048722679e-06, "loss": 0.0153, "step": 72450 }, { "grad_norm": 0.36176344752311707, "learning_rate": 2.4096330638046673e-06, "loss": 0.0167, "step": 72460 }, { "grad_norm": 0.27894431352615356, "learning_rate": 2.4032982107787816e-06, "loss": 0.0175, "step": 72470 }, { "grad_norm": 0.29595375061035156, "learning_rate": 2.396971490727451e-06, "loss": 0.022, "step": 72480 }, { "grad_norm": 0.22464069724082947, "learning_rate": 2.3906529047317493e-06, "loss": 0.0136, "step": 72490 }, { "grad_norm": 0.4765695035457611, "learning_rate": 2.3843424538713465e-06, "loss": 0.0151, "step": 72500 }, { "grad_norm": 0.5188554525375366, "learning_rate": 2.3780401392245298e-06, "loss": 0.0148, "step": 72510 }, { "grad_norm": 0.2364806979894638, "learning_rate": 2.3717459618681883e-06, "loss": 0.0149, "step": 72520 }, { "grad_norm": 0.22118470072746277, "learning_rate": 2.3654599228778274e-06, "loss": 0.0171, "step": 72530 }, { "grad_norm": 0.2620807886123657, "learning_rate": 2.3591820233275607e-06, "loss": 0.0208, "step": 72540 }, { "grad_norm": 0.1253894865512848, "learning_rate": 2.3529122642901024e-06, "loss": 0.0183, "step": 72550 }, { "grad_norm": 0.2267734706401825, "learning_rate": 2.3466506468367845e-06, "loss": 0.028, "step": 72560 }, { "grad_norm": 0.15995393693447113, "learning_rate": 2.3403971720375506e-06, "loss": 0.0207, "step": 72570 }, { "grad_norm": 0.23862825334072113, "learning_rate": 2.334151840960952e-06, "loss": 0.0134, "step": 72580 }, { "grad_norm": 0.2660394310951233, "learning_rate": 2.3279146546741347e-06, "loss": 0.0114, "step": 72590 }, { "grad_norm": 0.25209125876426697, "learning_rate": 2.32168561424288e-06, "loss": 0.0136, "step": 72600 }, { "grad_norm": 0.3624818027019501, "learning_rate": 2.3154647207315307e-06, "loss": 0.0196, "step": 72610 }, { "grad_norm": 0.21598127484321594, "learning_rate": 2.309251975203103e-06, "loss": 0.018, "step": 72620 }, { "grad_norm": 0.20467069745063782, "learning_rate": 2.303047378719159e-06, "loss": 0.0189, "step": 72630 }, { "grad_norm": 0.33889710903167725, "learning_rate": 2.2968509323399e-06, "loss": 0.0217, "step": 72640 }, { "grad_norm": 0.2736404836177826, "learning_rate": 2.290662637124147e-06, "loss": 0.0148, "step": 72650 }, { "grad_norm": 0.2519579529762268, "learning_rate": 2.2844824941292807e-06, "loss": 0.0172, "step": 72660 }, { "grad_norm": 0.43128281831741333, "learning_rate": 2.2783105044113406e-06, "loss": 0.0194, "step": 72670 }, { "grad_norm": 0.3151150643825531, "learning_rate": 2.272146669024944e-06, "loss": 0.0167, "step": 72680 }, { "grad_norm": 0.38095277547836304, "learning_rate": 2.26599098902332e-06, "loss": 0.0142, "step": 72690 }, { "grad_norm": 0.2629021406173706, "learning_rate": 2.2598434654583113e-06, "loss": 0.0188, "step": 72700 }, { "grad_norm": 0.2783558964729309, "learning_rate": 2.253704099380355e-06, "loss": 0.0231, "step": 72710 }, { "grad_norm": 0.43773069977760315, "learning_rate": 2.247572891838512e-06, "loss": 0.0178, "step": 72720 }, { "grad_norm": 0.2517934739589691, "learning_rate": 2.241449843880422e-06, "loss": 0.0169, "step": 72730 }, { "grad_norm": 0.16339127719402313, "learning_rate": 2.235334956552354e-06, "loss": 0.0191, "step": 72740 }, { "grad_norm": 0.15311437845230103, "learning_rate": 2.2292282308991775e-06, "loss": 0.0171, "step": 72750 }, { "grad_norm": 0.18680985271930695, "learning_rate": 2.223129667964363e-06, "loss": 0.0173, "step": 72760 }, { "grad_norm": 0.22509463131427765, "learning_rate": 2.2170392687899834e-06, "loss": 0.0118, "step": 72770 }, { "grad_norm": 0.21402055025100708, "learning_rate": 2.210957034416733e-06, "loss": 0.0145, "step": 72780 }, { "grad_norm": 0.24028131365776062, "learning_rate": 2.2048829658838867e-06, "loss": 0.0156, "step": 72790 }, { "grad_norm": 0.12826576828956604, "learning_rate": 2.1988170642293525e-06, "loss": 0.0138, "step": 72800 }, { "grad_norm": 0.18545742332935333, "learning_rate": 2.1927593304896075e-06, "loss": 0.0105, "step": 72810 }, { "grad_norm": 0.12065503746271133, "learning_rate": 2.1867097656997626e-06, "loss": 0.0143, "step": 72820 }, { "grad_norm": 0.24397489428520203, "learning_rate": 2.180668370893524e-06, "loss": 0.0155, "step": 72830 }, { "grad_norm": 0.2553078532218933, "learning_rate": 2.174635147103199e-06, "loss": 0.0167, "step": 72840 }, { "grad_norm": 0.2528461813926697, "learning_rate": 2.1686100953597075e-06, "loss": 0.0136, "step": 72850 }, { "grad_norm": 0.38919320702552795, "learning_rate": 2.1625932166925433e-06, "loss": 0.0165, "step": 72860 }, { "grad_norm": 0.3338817059993744, "learning_rate": 2.1565845121298556e-06, "loss": 0.0154, "step": 72870 }, { "grad_norm": 0.17862585186958313, "learning_rate": 2.150583982698351e-06, "loss": 0.0157, "step": 72880 }, { "grad_norm": 0.2511981725692749, "learning_rate": 2.144591629423359e-06, "loss": 0.0153, "step": 72890 }, { "grad_norm": 0.2755897641181946, "learning_rate": 2.138607453328817e-06, "loss": 0.0164, "step": 72900 }, { "grad_norm": 0.357610285282135, "learning_rate": 2.132631455437234e-06, "loss": 0.0182, "step": 72910 }, { "grad_norm": 0.23352554440498352, "learning_rate": 2.1266636367697714e-06, "loss": 0.0226, "step": 72920 }, { "grad_norm": 0.28155115246772766, "learning_rate": 2.120703998346152e-06, "loss": 0.0178, "step": 72930 }, { "grad_norm": 0.25142353773117065, "learning_rate": 2.1147525411847114e-06, "loss": 0.0173, "step": 72940 }, { "grad_norm": 0.23720256984233856, "learning_rate": 2.108809266302403e-06, "loss": 0.0125, "step": 72950 }, { "grad_norm": 0.3367057144641876, "learning_rate": 2.102874174714764e-06, "loss": 0.0234, "step": 72960 }, { "grad_norm": 0.30192720890045166, "learning_rate": 2.0969472674359393e-06, "loss": 0.02, "step": 72970 }, { "grad_norm": 0.20766152441501617, "learning_rate": 2.091028545478668e-06, "loss": 0.0159, "step": 72980 }, { "grad_norm": 0.27351662516593933, "learning_rate": 2.08511800985432e-06, "loss": 0.0153, "step": 72990 }, { "grad_norm": 0.147443488240242, "learning_rate": 2.07921566157282e-06, "loss": 0.0246, "step": 73000 }, { "grad_norm": 0.2687933146953583, "learning_rate": 2.073321501642728e-06, "loss": 0.0132, "step": 73010 }, { "grad_norm": 0.3107171356678009, "learning_rate": 2.0674355310711937e-06, "loss": 0.0191, "step": 73020 }, { "grad_norm": 0.29746976494789124, "learning_rate": 2.0615577508639682e-06, "loss": 0.0112, "step": 73030 }, { "grad_norm": 0.2861424684524536, "learning_rate": 2.055688162025399e-06, "loss": 0.0135, "step": 73040 }, { "grad_norm": 0.1894461214542389, "learning_rate": 2.0498267655584546e-06, "loss": 0.0191, "step": 73050 }, { "grad_norm": 0.18923939764499664, "learning_rate": 2.0439735624646626e-06, "loss": 0.0171, "step": 73060 }, { "grad_norm": 0.2639177143573761, "learning_rate": 2.0381285537442008e-06, "loss": 0.014, "step": 73070 }, { "grad_norm": 0.4832114279270172, "learning_rate": 2.032291740395803e-06, "loss": 0.0184, "step": 73080 }, { "grad_norm": 0.2511865794658661, "learning_rate": 2.0264631234168276e-06, "loss": 0.0141, "step": 73090 }, { "grad_norm": 0.3563922047615051, "learning_rate": 2.0206427038032326e-06, "loss": 0.0176, "step": 73100 }, { "grad_norm": 0.19391530752182007, "learning_rate": 2.0148304825495457e-06, "loss": 0.0115, "step": 73110 }, { "grad_norm": 0.20993687212467194, "learning_rate": 2.0090264606489496e-06, "loss": 0.0106, "step": 73120 }, { "grad_norm": 0.18562117218971252, "learning_rate": 2.003230639093162e-06, "loss": 0.0153, "step": 73130 }, { "grad_norm": 0.17003336548805237, "learning_rate": 1.9974430188725525e-06, "loss": 0.0187, "step": 73140 }, { "grad_norm": 0.16986966133117676, "learning_rate": 1.9916636009760513e-06, "loss": 0.017, "step": 73150 }, { "grad_norm": 0.1498444825410843, "learning_rate": 1.985892386391208e-06, "loss": 0.0175, "step": 73160 }, { "grad_norm": 0.26718541979789734, "learning_rate": 1.980129376104173e-06, "loss": 0.0208, "step": 73170 }, { "grad_norm": 0.08319185674190521, "learning_rate": 1.9743745710996796e-06, "loss": 0.0186, "step": 73180 }, { "grad_norm": 0.1567346155643463, "learning_rate": 1.9686279723610757e-06, "loss": 0.0166, "step": 73190 }, { "grad_norm": 0.3242879807949066, "learning_rate": 1.962889580870281e-06, "loss": 0.019, "step": 73200 }, { "grad_norm": 0.37962645292282104, "learning_rate": 1.9571593976078385e-06, "loss": 0.0123, "step": 73210 }, { "grad_norm": 0.2138078659772873, "learning_rate": 1.9514374235528876e-06, "loss": 0.0191, "step": 73220 }, { "grad_norm": 0.22992214560508728, "learning_rate": 1.9457236596831408e-06, "loss": 0.012, "step": 73230 }, { "grad_norm": 0.14340414106845856, "learning_rate": 1.940018106974939e-06, "loss": 0.0203, "step": 73240 }, { "grad_norm": 0.37047964334487915, "learning_rate": 1.9343207664032025e-06, "loss": 0.0189, "step": 73250 }, { "grad_norm": 0.3293004631996155, "learning_rate": 1.928631638941436e-06, "loss": 0.0157, "step": 73260 }, { "grad_norm": 0.2268308848142624, "learning_rate": 1.922950725561784e-06, "loss": 0.0165, "step": 73270 }, { "grad_norm": 0.2585739195346832, "learning_rate": 1.9172780272349312e-06, "loss": 0.0099, "step": 73280 }, { "grad_norm": 0.17893105745315552, "learning_rate": 1.911613544930202e-06, "loss": 0.0153, "step": 73290 }, { "grad_norm": 0.24026194214820862, "learning_rate": 1.9059572796155e-06, "loss": 0.0195, "step": 73300 }, { "grad_norm": 0.28968411684036255, "learning_rate": 1.9003092322573192e-06, "loss": 0.0142, "step": 73310 }, { "grad_norm": 0.33698320388793945, "learning_rate": 1.8946694038207647e-06, "loss": 0.0272, "step": 73320 }, { "grad_norm": 0.3355383276939392, "learning_rate": 1.8890377952695215e-06, "loss": 0.0172, "step": 73330 }, { "grad_norm": 0.3166024684906006, "learning_rate": 1.8834144075658865e-06, "loss": 0.0152, "step": 73340 }, { "grad_norm": 0.23236754536628723, "learning_rate": 1.8777992416707357e-06, "loss": 0.017, "step": 73350 }, { "grad_norm": 0.2614282965660095, "learning_rate": 1.8721922985435458e-06, "loss": 0.0142, "step": 73360 }, { "grad_norm": 0.4499240517616272, "learning_rate": 1.8665935791424006e-06, "loss": 0.0214, "step": 73370 }, { "grad_norm": 0.37824520468711853, "learning_rate": 1.8610030844239512e-06, "loss": 0.0153, "step": 73380 }, { "grad_norm": 0.25680699944496155, "learning_rate": 1.8554208153434838e-06, "loss": 0.0149, "step": 73390 }, { "grad_norm": 0.2092752605676651, "learning_rate": 1.8498467728548296e-06, "loss": 0.0122, "step": 73400 }, { "grad_norm": 0.19935479760169983, "learning_rate": 1.8442809579104547e-06, "loss": 0.0141, "step": 73410 }, { "grad_norm": 0.40892544388771057, "learning_rate": 1.8387233714614039e-06, "loss": 0.0169, "step": 73420 }, { "grad_norm": 0.2561829388141632, "learning_rate": 1.8331740144573174e-06, "loss": 0.0237, "step": 73430 }, { "grad_norm": 0.24955882132053375, "learning_rate": 1.8276328878464255e-06, "loss": 0.0138, "step": 73440 }, { "grad_norm": 0.3151242733001709, "learning_rate": 1.82209999257556e-06, "loss": 0.017, "step": 73450 }, { "grad_norm": 0.27106475830078125, "learning_rate": 1.8165753295901311e-06, "loss": 0.0157, "step": 73460 }, { "grad_norm": 0.09437248855829239, "learning_rate": 1.8110588998341616e-06, "loss": 0.02, "step": 73470 }, { "grad_norm": 0.2279033064842224, "learning_rate": 1.805550704250253e-06, "loss": 0.0288, "step": 73480 }, { "grad_norm": 0.23905006051063538, "learning_rate": 1.8000507437796077e-06, "loss": 0.0136, "step": 73490 }, { "grad_norm": 0.29412296414375305, "learning_rate": 1.7945590193620242e-06, "loss": 0.0174, "step": 73500 }, { "grad_norm": 0.39220622181892395, "learning_rate": 1.7890755319358742e-06, "loss": 0.0172, "step": 73510 }, { "grad_norm": 0.5038588643074036, "learning_rate": 1.7836002824381525e-06, "loss": 0.0202, "step": 73520 }, { "grad_norm": 0.13794757425785065, "learning_rate": 1.7781332718044165e-06, "loss": 0.0136, "step": 73530 }, { "grad_norm": 0.27359384298324585, "learning_rate": 1.772674500968835e-06, "loss": 0.018, "step": 73540 }, { "grad_norm": 0.4652453064918518, "learning_rate": 1.7672239708641624e-06, "loss": 0.019, "step": 73550 }, { "grad_norm": 0.15625448524951935, "learning_rate": 1.7617816824217371e-06, "loss": 0.0143, "step": 73560 }, { "grad_norm": 0.28106778860092163, "learning_rate": 1.7563476365715148e-06, "loss": 0.0221, "step": 73570 }, { "grad_norm": 0.2016599029302597, "learning_rate": 1.750921834242003e-06, "loss": 0.0261, "step": 73580 }, { "grad_norm": 0.2865433096885681, "learning_rate": 1.7455042763603436e-06, "loss": 0.0158, "step": 73590 }, { "grad_norm": 0.20564696192741394, "learning_rate": 1.7400949638522345e-06, "loss": 0.0134, "step": 73600 }, { "grad_norm": 0.2610589563846588, "learning_rate": 1.734693897641987e-06, "loss": 0.0255, "step": 73610 }, { "grad_norm": 0.17091819643974304, "learning_rate": 1.7293010786524955e-06, "loss": 0.0142, "step": 73620 }, { "grad_norm": 0.35135167837142944, "learning_rate": 1.72391650780524e-06, "loss": 0.0211, "step": 73630 }, { "grad_norm": 0.2801767587661743, "learning_rate": 1.718540186020301e-06, "loss": 0.0148, "step": 73640 }, { "grad_norm": 0.3394782543182373, "learning_rate": 1.7131721142163437e-06, "loss": 0.0177, "step": 73650 }, { "grad_norm": 0.32041066884994507, "learning_rate": 1.7078122933106233e-06, "loss": 0.0232, "step": 73660 }, { "grad_norm": 0.13467110693454742, "learning_rate": 1.7024607242189905e-06, "loss": 0.0169, "step": 73670 }, { "grad_norm": 0.2333473116159439, "learning_rate": 1.6971174078558749e-06, "loss": 0.014, "step": 73680 }, { "grad_norm": 0.3161637485027313, "learning_rate": 1.6917823451343073e-06, "loss": 0.0203, "step": 73690 }, { "grad_norm": 0.26079466938972473, "learning_rate": 1.6864555369659142e-06, "loss": 0.0149, "step": 73700 }, { "grad_norm": 0.18917424976825714, "learning_rate": 1.681136984260878e-06, "loss": 0.0183, "step": 73710 }, { "grad_norm": 0.2919810712337494, "learning_rate": 1.6758266879280172e-06, "loss": 0.0128, "step": 73720 }, { "grad_norm": 0.3360748589038849, "learning_rate": 1.6705246488746996e-06, "loss": 0.0133, "step": 73730 }, { "grad_norm": 0.28747987747192383, "learning_rate": 1.6652308680069062e-06, "loss": 0.017, "step": 73740 }, { "grad_norm": 0.32468074560165405, "learning_rate": 1.6599453462292081e-06, "loss": 0.0179, "step": 73750 }, { "grad_norm": 0.41640031337738037, "learning_rate": 1.6546680844447326e-06, "loss": 0.0201, "step": 73760 }, { "grad_norm": 0.2482677400112152, "learning_rate": 1.6493990835552475e-06, "loss": 0.0182, "step": 73770 }, { "grad_norm": 0.2922252118587494, "learning_rate": 1.6441383444610547e-06, "loss": 0.0139, "step": 73780 }, { "grad_norm": 0.20350664854049683, "learning_rate": 1.6388858680610908e-06, "loss": 0.015, "step": 73790 }, { "grad_norm": 0.2548910677433014, "learning_rate": 1.6336416552528543e-06, "loss": 0.0133, "step": 73800 }, { "grad_norm": 0.335909366607666, "learning_rate": 1.6284057069324288e-06, "loss": 0.0216, "step": 73810 }, { "grad_norm": 0.16419415175914764, "learning_rate": 1.623178023994504e-06, "loss": 0.0232, "step": 73820 }, { "grad_norm": 0.2521587610244751, "learning_rate": 1.6179586073323483e-06, "loss": 0.0194, "step": 73830 }, { "grad_norm": 0.34593018889427185, "learning_rate": 1.612747457837821e-06, "loss": 0.016, "step": 73840 }, { "grad_norm": 0.16966718435287476, "learning_rate": 1.607544576401354e-06, "loss": 0.0151, "step": 73850 }, { "grad_norm": 0.255148321390152, "learning_rate": 1.6023499639119754e-06, "loss": 0.0171, "step": 73860 }, { "grad_norm": 0.3558322787284851, "learning_rate": 1.5971636212573138e-06, "loss": 0.0184, "step": 73870 }, { "grad_norm": 0.22441990673542023, "learning_rate": 1.5919855493235714e-06, "loss": 0.0193, "step": 73880 }, { "grad_norm": 0.25361669063568115, "learning_rate": 1.586815748995535e-06, "loss": 0.0104, "step": 73890 }, { "grad_norm": 0.12696990370750427, "learning_rate": 1.5816542211565866e-06, "loss": 0.0129, "step": 73900 }, { "grad_norm": 0.20505847036838531, "learning_rate": 1.5765009666886765e-06, "loss": 0.0143, "step": 73910 }, { "grad_norm": 0.25741177797317505, "learning_rate": 1.5713559864723781e-06, "loss": 0.0187, "step": 73920 }, { "grad_norm": 0.34506821632385254, "learning_rate": 1.5662192813868104e-06, "loss": 0.0145, "step": 73930 }, { "grad_norm": 0.28876954317092896, "learning_rate": 1.561090852309699e-06, "loss": 0.0214, "step": 73940 }, { "grad_norm": 0.16023015975952148, "learning_rate": 1.5559707001173651e-06, "loss": 0.0177, "step": 73950 }, { "grad_norm": 0.373868465423584, "learning_rate": 1.5508588256846757e-06, "loss": 0.0182, "step": 73960 }, { "grad_norm": 0.3266337513923645, "learning_rate": 1.5457552298851319e-06, "loss": 0.0145, "step": 73970 }, { "grad_norm": 0.45144832134246826, "learning_rate": 1.5406599135907918e-06, "loss": 0.0192, "step": 73980 }, { "grad_norm": 0.5188606381416321, "learning_rate": 1.5355728776723088e-06, "loss": 0.0198, "step": 73990 }, { "grad_norm": 0.14277468621730804, "learning_rate": 1.5304941229989155e-06, "loss": 0.0177, "step": 74000 }, { "grad_norm": 0.2652300000190735, "learning_rate": 1.5254236504384345e-06, "loss": 0.016, "step": 74010 }, { "grad_norm": 0.2667950391769409, "learning_rate": 1.5203614608572726e-06, "loss": 0.0212, "step": 74020 }, { "grad_norm": 0.1589064747095108, "learning_rate": 1.5153075551204044e-06, "loss": 0.0119, "step": 74030 }, { "grad_norm": 0.20669302344322205, "learning_rate": 1.5102619340914225e-06, "loss": 0.0269, "step": 74040 }, { "grad_norm": 0.37927836179733276, "learning_rate": 1.505224598632482e-06, "loss": 0.0142, "step": 74050 }, { "grad_norm": 0.1485144942998886, "learning_rate": 1.5001955496043164e-06, "loss": 0.021, "step": 74060 }, { "grad_norm": 0.28174281120300293, "learning_rate": 1.4951747878662604e-06, "loss": 0.0115, "step": 74070 }, { "grad_norm": 0.14754147827625275, "learning_rate": 1.4901623142762221e-06, "loss": 0.0129, "step": 74080 }, { "grad_norm": 0.2144499123096466, "learning_rate": 1.4851581296907001e-06, "loss": 0.0182, "step": 74090 }, { "grad_norm": 0.4644833505153656, "learning_rate": 1.4801622349647714e-06, "loss": 0.0147, "step": 74100 }, { "grad_norm": 0.242391437292099, "learning_rate": 1.4751746309520976e-06, "loss": 0.017, "step": 74110 }, { "grad_norm": 0.2461199015378952, "learning_rate": 1.4701953185049132e-06, "loss": 0.022, "step": 74120 }, { "grad_norm": 0.2321084439754486, "learning_rate": 1.4652242984740661e-06, "loss": 0.0193, "step": 74130 }, { "grad_norm": 0.22165800631046295, "learning_rate": 1.4602615717089484e-06, "loss": 0.0252, "step": 74140 }, { "grad_norm": 0.3481142520904541, "learning_rate": 1.4553071390575657e-06, "loss": 0.0171, "step": 74150 }, { "grad_norm": 0.20924973487854004, "learning_rate": 1.450361001366496e-06, "loss": 0.0259, "step": 74160 }, { "grad_norm": 0.19371570646762848, "learning_rate": 1.4454231594809021e-06, "loss": 0.0162, "step": 74170 }, { "grad_norm": 0.29783347249031067, "learning_rate": 1.4404936142445036e-06, "loss": 0.0163, "step": 74180 }, { "grad_norm": 0.1993081271648407, "learning_rate": 1.4355723664996546e-06, "loss": 0.019, "step": 74190 }, { "grad_norm": 0.2862643301486969, "learning_rate": 1.4306594170872433e-06, "loss": 0.0132, "step": 74200 }, { "grad_norm": 0.3175918459892273, "learning_rate": 1.4257547668467598e-06, "loss": 0.0223, "step": 74210 }, { "grad_norm": 0.2613670229911804, "learning_rate": 1.4208584166162886e-06, "loss": 0.021, "step": 74220 }, { "grad_norm": 0.31412824988365173, "learning_rate": 1.4159703672324553e-06, "loss": 0.019, "step": 74230 }, { "grad_norm": 0.26353520154953003, "learning_rate": 1.4110906195305252e-06, "loss": 0.0172, "step": 74240 }, { "grad_norm": 0.3736203610897064, "learning_rate": 1.406219174344292e-06, "loss": 0.0175, "step": 74250 }, { "grad_norm": 0.6303986310958862, "learning_rate": 1.401356032506157e-06, "loss": 0.0185, "step": 74260 }, { "grad_norm": 0.3467719256877899, "learning_rate": 1.3965011948471051e-06, "loss": 0.0175, "step": 74270 }, { "grad_norm": 0.222660094499588, "learning_rate": 1.3916546621966842e-06, "loss": 0.0147, "step": 74280 }, { "grad_norm": 0.34091687202453613, "learning_rate": 1.3868164353830482e-06, "loss": 0.014, "step": 74290 }, { "grad_norm": 0.2180185317993164, "learning_rate": 1.3819865152329082e-06, "loss": 0.0122, "step": 74300 }, { "grad_norm": 0.271043062210083, "learning_rate": 1.377164902571565e-06, "loss": 0.0233, "step": 74310 }, { "grad_norm": 0.22579990327358246, "learning_rate": 1.372351598222904e-06, "loss": 0.0192, "step": 74320 }, { "grad_norm": 0.18950936198234558, "learning_rate": 1.3675466030093843e-06, "loss": 0.0143, "step": 74330 }, { "grad_norm": 0.3328295648097992, "learning_rate": 1.3627499177520486e-06, "loss": 0.02, "step": 74340 }, { "grad_norm": 0.19908316433429718, "learning_rate": 1.3579615432705196e-06, "loss": 0.0175, "step": 74350 }, { "grad_norm": 0.21751493215560913, "learning_rate": 1.3531814803829978e-06, "loss": 0.0166, "step": 74360 }, { "grad_norm": 0.2118358016014099, "learning_rate": 1.3484097299062803e-06, "loss": 0.0198, "step": 74370 }, { "grad_norm": 0.09452024847269058, "learning_rate": 1.343646292655698e-06, "loss": 0.0202, "step": 74380 }, { "grad_norm": 0.17775462567806244, "learning_rate": 1.3388911694452277e-06, "loss": 0.0163, "step": 74390 }, { "grad_norm": 0.3584577143192291, "learning_rate": 1.3341443610873638e-06, "loss": 0.022, "step": 74400 }, { "grad_norm": 0.31466007232666016, "learning_rate": 1.329405868393213e-06, "loss": 0.0172, "step": 74410 }, { "grad_norm": 0.23121586441993713, "learning_rate": 1.3246756921724613e-06, "loss": 0.0208, "step": 74420 }, { "grad_norm": 0.2279001921415329, "learning_rate": 1.3199538332333506e-06, "loss": 0.016, "step": 74430 }, { "grad_norm": 0.21735011041164398, "learning_rate": 1.3152402923827411e-06, "loss": 0.0213, "step": 74440 }, { "grad_norm": 0.3188868761062622, "learning_rate": 1.3105350704260277e-06, "loss": 0.016, "step": 74450 }, { "grad_norm": 0.4430491030216217, "learning_rate": 1.305838168167206e-06, "loss": 0.0128, "step": 74460 }, { "grad_norm": 0.24608056247234344, "learning_rate": 1.301149586408862e-06, "loss": 0.0113, "step": 74470 }, { "grad_norm": 0.14512313902378082, "learning_rate": 1.2964693259521322e-06, "loss": 0.016, "step": 74480 }, { "grad_norm": 0.18027208745479584, "learning_rate": 1.2917973875967548e-06, "loss": 0.0146, "step": 74490 }, { "grad_norm": 0.29449743032455444, "learning_rate": 1.2871337721410249e-06, "loss": 0.0166, "step": 74500 }, { "grad_norm": 0.18431058526039124, "learning_rate": 1.2824784803818379e-06, "loss": 0.0178, "step": 74510 }, { "grad_norm": 0.2272864133119583, "learning_rate": 1.2778315131146467e-06, "loss": 0.0133, "step": 74520 }, { "grad_norm": 0.3247848153114319, "learning_rate": 1.2731928711334994e-06, "loss": 0.0135, "step": 74530 }, { "grad_norm": 0.20612043142318726, "learning_rate": 1.2685625552310122e-06, "loss": 0.0118, "step": 74540 }, { "grad_norm": 0.2431199997663498, "learning_rate": 1.263940566198374e-06, "loss": 0.0169, "step": 74550 }, { "grad_norm": 0.16931454837322235, "learning_rate": 1.259326904825353e-06, "loss": 0.0153, "step": 74560 }, { "grad_norm": 0.3899374008178711, "learning_rate": 1.2547215719003137e-06, "loss": 0.0193, "step": 74570 }, { "grad_norm": 0.2092130333185196, "learning_rate": 1.2501245682101703e-06, "loss": 0.0123, "step": 74580 }, { "grad_norm": 0.2323491871356964, "learning_rate": 1.2455358945404171e-06, "loss": 0.0143, "step": 74590 }, { "grad_norm": 0.3248220384120941, "learning_rate": 1.2409555516751493e-06, "loss": 0.0134, "step": 74600 }, { "grad_norm": 0.2492532730102539, "learning_rate": 1.2363835403970125e-06, "loss": 0.0157, "step": 74610 }, { "grad_norm": 0.3665224611759186, "learning_rate": 1.2318198614872489e-06, "loss": 0.0183, "step": 74620 }, { "grad_norm": 0.2243870049715042, "learning_rate": 1.2272645157256457e-06, "loss": 0.0138, "step": 74630 }, { "grad_norm": 0.3772449493408203, "learning_rate": 1.222717503890608e-06, "loss": 0.0213, "step": 74640 }, { "grad_norm": 0.3342970013618469, "learning_rate": 1.218178826759081e-06, "loss": 0.0234, "step": 74650 }, { "grad_norm": 0.8135418891906738, "learning_rate": 1.213648485106611e-06, "loss": 0.0205, "step": 74660 }, { "grad_norm": 0.21722184121608734, "learning_rate": 1.2091264797073066e-06, "loss": 0.0205, "step": 74670 }, { "grad_norm": 0.3357754945755005, "learning_rate": 1.2046128113338494e-06, "loss": 0.0281, "step": 74680 }, { "grad_norm": 0.3014982044696808, "learning_rate": 1.200107480757512e-06, "loss": 0.0178, "step": 74690 }, { "grad_norm": 0.25341904163360596, "learning_rate": 1.1956104887481168e-06, "loss": 0.0244, "step": 74700 }, { "grad_norm": 0.45632079243659973, "learning_rate": 1.191121836074094e-06, "loss": 0.0151, "step": 74710 }, { "grad_norm": 0.3011428713798523, "learning_rate": 1.1866415235024186e-06, "loss": 0.0125, "step": 74720 }, { "grad_norm": 0.24122372269630432, "learning_rate": 1.182169551798662e-06, "loss": 0.0178, "step": 74730 }, { "grad_norm": 0.19819340109825134, "learning_rate": 1.177705921726957e-06, "loss": 0.0135, "step": 74740 }, { "grad_norm": 0.17992965877056122, "learning_rate": 1.173250634050016e-06, "loss": 0.0152, "step": 74750 }, { "grad_norm": 0.23030106723308563, "learning_rate": 1.1688036895291354e-06, "loss": 0.017, "step": 74760 }, { "grad_norm": 0.32030174136161804, "learning_rate": 1.1643650889241574e-06, "loss": 0.0161, "step": 74770 }, { "grad_norm": 0.35747796297073364, "learning_rate": 1.1599348329935311e-06, "loss": 0.0189, "step": 74780 }, { "grad_norm": 0.31607887148857117, "learning_rate": 1.1555129224942673e-06, "loss": 0.026, "step": 74790 }, { "grad_norm": 0.17677772045135498, "learning_rate": 1.1510993581819396e-06, "loss": 0.0157, "step": 74800 }, { "grad_norm": 0.3267647922039032, "learning_rate": 1.1466941408107112e-06, "loss": 0.0248, "step": 74810 }, { "grad_norm": 0.2311938852071762, "learning_rate": 1.1422972711333247e-06, "loss": 0.0159, "step": 74820 }, { "grad_norm": 0.18698450922966003, "learning_rate": 1.1379087499010565e-06, "loss": 0.0138, "step": 74830 }, { "grad_norm": 0.31728941202163696, "learning_rate": 1.1335285778638127e-06, "loss": 0.013, "step": 74840 }, { "grad_norm": 0.3259546160697937, "learning_rate": 1.1291567557700333e-06, "loss": 0.0197, "step": 74850 }, { "grad_norm": 0.4181780517101288, "learning_rate": 1.124793284366743e-06, "loss": 0.0168, "step": 74860 }, { "grad_norm": 0.2685205936431885, "learning_rate": 1.1204381643995455e-06, "loss": 0.0152, "step": 74870 }, { "grad_norm": 0.327216237783432, "learning_rate": 1.116091396612595e-06, "loss": 0.0176, "step": 74880 }, { "grad_norm": 0.26948612928390503, "learning_rate": 1.1117529817486538e-06, "loss": 0.0172, "step": 74890 }, { "grad_norm": 0.26150527596473694, "learning_rate": 1.107422920549034e-06, "loss": 0.0142, "step": 74900 }, { "grad_norm": 0.394995778799057, "learning_rate": 1.1031012137536157e-06, "loss": 0.0159, "step": 74910 }, { "grad_norm": 0.17466354370117188, "learning_rate": 1.0987878621008695e-06, "loss": 0.0168, "step": 74920 }, { "grad_norm": 0.2878662049770355, "learning_rate": 1.0944828663278284e-06, "loss": 0.0168, "step": 74930 }, { "grad_norm": 0.1857266128063202, "learning_rate": 1.0901862271700925e-06, "loss": 0.0165, "step": 74940 }, { "grad_norm": 0.2934063673019409, "learning_rate": 1.0858979453618467e-06, "loss": 0.0175, "step": 74950 }, { "grad_norm": 0.2657327353954315, "learning_rate": 1.0816180216358441e-06, "loss": 0.0162, "step": 74960 }, { "grad_norm": 0.3661000728607178, "learning_rate": 1.0773464567233937e-06, "loss": 0.0131, "step": 74970 }, { "grad_norm": 0.22015345096588135, "learning_rate": 1.0730832513543953e-06, "loss": 0.0152, "step": 74980 }, { "grad_norm": 0.28739145398139954, "learning_rate": 1.0688284062573218e-06, "loss": 0.0197, "step": 74990 }, { "grad_norm": 0.42014259099960327, "learning_rate": 1.0645819221591969e-06, "loss": 0.0177, "step": 75000 }, { "grad_norm": 0.3721542954444885, "learning_rate": 1.0603437997856347e-06, "loss": 0.0139, "step": 75010 }, { "grad_norm": 0.2299734503030777, "learning_rate": 1.0561140398608228e-06, "loss": 0.0141, "step": 75020 }, { "grad_norm": 0.22276221215724945, "learning_rate": 1.0518926431074937e-06, "loss": 0.012, "step": 75030 }, { "grad_norm": 0.2072758674621582, "learning_rate": 1.0476796102469877e-06, "loss": 0.0176, "step": 75040 }, { "grad_norm": 0.32618623971939087, "learning_rate": 1.0434749419991786e-06, "loss": 0.0173, "step": 75050 }, { "grad_norm": 0.22084848582744598, "learning_rate": 1.0392786390825415e-06, "loss": 0.015, "step": 75060 }, { "grad_norm": 0.259073406457901, "learning_rate": 1.0350907022141087e-06, "loss": 0.0178, "step": 75070 }, { "grad_norm": 0.23548974096775055, "learning_rate": 1.0309111321094744e-06, "loss": 0.0169, "step": 75080 }, { "grad_norm": 0.1582847535610199, "learning_rate": 1.026739929482834e-06, "loss": 0.0133, "step": 75090 }, { "grad_norm": 0.2947421371936798, "learning_rate": 1.022577095046906e-06, "loss": 0.0167, "step": 75100 }, { "grad_norm": 0.2531931698322296, "learning_rate": 1.0184226295130217e-06, "loss": 0.0187, "step": 75110 }, { "grad_norm": 0.3536534607410431, "learning_rate": 1.0142765335910575e-06, "loss": 0.0211, "step": 75120 }, { "grad_norm": 0.4057387411594391, "learning_rate": 1.0101388079894746e-06, "loss": 0.0147, "step": 75130 }, { "grad_norm": 0.16837972402572632, "learning_rate": 1.0060094534152908e-06, "loss": 0.0195, "step": 75140 }, { "grad_norm": 0.1827598661184311, "learning_rate": 1.0018884705741028e-06, "loss": 0.0119, "step": 75150 }, { "grad_norm": 0.3523397445678711, "learning_rate": 9.977758601700805e-07, "loss": 0.0239, "step": 75160 }, { "grad_norm": 0.41920700669288635, "learning_rate": 9.93671622905945e-07, "loss": 0.0233, "step": 75170 }, { "grad_norm": 0.27231961488723755, "learning_rate": 9.895757594830024e-07, "loss": 0.0146, "step": 75180 }, { "grad_norm": 0.23829209804534912, "learning_rate": 9.854882706011204e-07, "loss": 0.0149, "step": 75190 }, { "grad_norm": 0.35631710290908813, "learning_rate": 9.814091569587513e-07, "loss": 0.0159, "step": 75200 }, { "grad_norm": 0.4934217929840088, "learning_rate": 9.77338419252888e-07, "loss": 0.0201, "step": 75210 }, { "grad_norm": 0.31354933977127075, "learning_rate": 9.732760581791289e-07, "loss": 0.0165, "step": 75220 }, { "grad_norm": 0.23668745160102844, "learning_rate": 9.692220744315972e-07, "loss": 0.0181, "step": 75230 }, { "grad_norm": 0.2575167417526245, "learning_rate": 9.651764687030162e-07, "loss": 0.0166, "step": 75240 }, { "grad_norm": 0.19897741079330444, "learning_rate": 9.611392416846776e-07, "loss": 0.0135, "step": 75250 }, { "grad_norm": 0.3197329044342041, "learning_rate": 9.571103940664238e-07, "loss": 0.0151, "step": 75260 }, { "grad_norm": 0.17256148159503937, "learning_rate": 9.530899265366822e-07, "loss": 0.0194, "step": 75270 }, { "grad_norm": 0.3757726550102234, "learning_rate": 9.490778397824307e-07, "loss": 0.0181, "step": 75280 }, { "grad_norm": 0.2471575289964676, "learning_rate": 9.450741344892322e-07, "loss": 0.0127, "step": 75290 }, { "grad_norm": 0.35893458127975464, "learning_rate": 9.410788113412116e-07, "loss": 0.0161, "step": 75300 }, { "grad_norm": 0.2143394500017166, "learning_rate": 9.370918710210563e-07, "loss": 0.016, "step": 75310 }, { "grad_norm": 0.2829333245754242, "learning_rate": 9.331133142100268e-07, "loss": 0.0148, "step": 75320 }, { "grad_norm": 0.3761897087097168, "learning_rate": 9.291431415879459e-07, "loss": 0.0113, "step": 75330 }, { "grad_norm": 0.2911415100097656, "learning_rate": 9.251813538332155e-07, "loss": 0.0209, "step": 75340 }, { "grad_norm": 0.3291282057762146, "learning_rate": 9.212279516227884e-07, "loss": 0.0147, "step": 75350 }, { "grad_norm": 0.2725762128829956, "learning_rate": 9.172829356322021e-07, "loss": 0.0145, "step": 75360 }, { "grad_norm": 0.3759704530239105, "learning_rate": 9.133463065355452e-07, "loss": 0.0132, "step": 75370 }, { "grad_norm": 0.3718796372413635, "learning_rate": 9.094180650054796e-07, "loss": 0.021, "step": 75380 }, { "grad_norm": 0.25605034828186035, "learning_rate": 9.054982117132404e-07, "loss": 0.0143, "step": 75390 }, { "grad_norm": 0.3222743272781372, "learning_rate": 9.015867473286143e-07, "loss": 0.0168, "step": 75400 }, { "grad_norm": 0.30815303325653076, "learning_rate": 8.976836725199778e-07, "loss": 0.0136, "step": 75410 }, { "grad_norm": 0.3292016386985779, "learning_rate": 8.937889879542416e-07, "loss": 0.0167, "step": 75420 }, { "grad_norm": 0.3525734841823578, "learning_rate": 8.899026942969068e-07, "loss": 0.0161, "step": 75430 }, { "grad_norm": 0.21404092013835907, "learning_rate": 8.860247922120424e-07, "loss": 0.0139, "step": 75440 }, { "grad_norm": 0.30228081345558167, "learning_rate": 8.821552823622737e-07, "loss": 0.0143, "step": 75450 }, { "grad_norm": 0.2527017295360565, "learning_rate": 8.782941654087884e-07, "loss": 0.0153, "step": 75460 }, { "grad_norm": 0.4433108866214752, "learning_rate": 8.744414420113478e-07, "loss": 0.0188, "step": 75470 }, { "grad_norm": 0.20550476014614105, "learning_rate": 8.705971128282753e-07, "loss": 0.0149, "step": 75480 }, { "grad_norm": 0.13431373238563538, "learning_rate": 8.66761178516473e-07, "loss": 0.0163, "step": 75490 }, { "grad_norm": 0.3052317202091217, "learning_rate": 8.629336397313781e-07, "loss": 0.0182, "step": 75500 }, { "grad_norm": 0.2322329878807068, "learning_rate": 8.591144971270282e-07, "loss": 0.0176, "step": 75510 }, { "grad_norm": 0.2095039188861847, "learning_rate": 8.553037513560069e-07, "loss": 0.0221, "step": 75520 }, { "grad_norm": 0.3774680495262146, "learning_rate": 8.515014030694546e-07, "loss": 0.0136, "step": 75530 }, { "grad_norm": 0.2058408558368683, "learning_rate": 8.47707452917107e-07, "loss": 0.0153, "step": 75540 }, { "grad_norm": 0.24535085260868073, "learning_rate": 8.439219015472343e-07, "loss": 0.0151, "step": 75550 }, { "grad_norm": 0.34855377674102783, "learning_rate": 8.401447496066861e-07, "loss": 0.0126, "step": 75560 }, { "grad_norm": 0.2406008392572403, "learning_rate": 8.363759977408792e-07, "loss": 0.0128, "step": 75570 }, { "grad_norm": 0.3295781910419464, "learning_rate": 8.326156465937817e-07, "loss": 0.0149, "step": 75580 }, { "grad_norm": 0.4567453861236572, "learning_rate": 8.28863696807941e-07, "loss": 0.0238, "step": 75590 }, { "grad_norm": 0.22357162833213806, "learning_rate": 8.251201490244609e-07, "loss": 0.0199, "step": 75600 }, { "grad_norm": 0.183163583278656, "learning_rate": 8.213850038830129e-07, "loss": 0.0135, "step": 75610 }, { "grad_norm": 0.22776281833648682, "learning_rate": 8.176582620218309e-07, "loss": 0.0139, "step": 75620 }, { "grad_norm": 0.2517589330673218, "learning_rate": 8.139399240777057e-07, "loss": 0.015, "step": 75630 }, { "grad_norm": 0.25687599182128906, "learning_rate": 8.102299906860122e-07, "loss": 0.0197, "step": 75640 }, { "grad_norm": 0.4153926968574524, "learning_rate": 8.065284624806657e-07, "loss": 0.0206, "step": 75650 }, { "grad_norm": 0.28078293800354004, "learning_rate": 8.028353400941601e-07, "loss": 0.0181, "step": 75660 }, { "grad_norm": 0.18945732712745667, "learning_rate": 7.991506241575575e-07, "loss": 0.0197, "step": 75670 }, { "grad_norm": 0.31661489605903625, "learning_rate": 7.954743153004541e-07, "loss": 0.0182, "step": 75680 }, { "grad_norm": 0.19791170954704285, "learning_rate": 7.918064141510528e-07, "loss": 0.0193, "step": 75690 }, { "grad_norm": 0.1400516927242279, "learning_rate": 7.881469213360859e-07, "loss": 0.019, "step": 75700 }, { "grad_norm": 0.319828599691391, "learning_rate": 7.844958374808642e-07, "loss": 0.0155, "step": 75710 }, { "grad_norm": 0.27905037999153137, "learning_rate": 7.808531632092608e-07, "loss": 0.0153, "step": 75720 }, { "grad_norm": 0.1984499841928482, "learning_rate": 7.772188991436946e-07, "loss": 0.0171, "step": 75730 }, { "grad_norm": 0.2976208031177521, "learning_rate": 7.735930459051799e-07, "loss": 0.0178, "step": 75740 }, { "grad_norm": 0.3360615372657776, "learning_rate": 7.699756041132655e-07, "loss": 0.02, "step": 75750 }, { "grad_norm": 0.3140636384487152, "learning_rate": 7.663665743860793e-07, "loss": 0.0166, "step": 75760 }, { "grad_norm": 0.24132388830184937, "learning_rate": 7.627659573403001e-07, "loss": 0.0162, "step": 75770 }, { "grad_norm": 0.2746235430240631, "learning_rate": 7.591737535911802e-07, "loss": 0.0184, "step": 75780 }, { "grad_norm": 0.20629508793354034, "learning_rate": 7.555899637525288e-07, "loss": 0.0145, "step": 75790 }, { "grad_norm": 0.2481718361377716, "learning_rate": 7.520145884367058e-07, "loss": 0.0145, "step": 75800 }, { "grad_norm": 0.22529201209545135, "learning_rate": 7.484476282546615e-07, "loss": 0.0161, "step": 75810 }, { "grad_norm": 0.3748706877231598, "learning_rate": 7.448890838158806e-07, "loss": 0.0223, "step": 75820 }, { "grad_norm": 0.2709726095199585, "learning_rate": 7.413389557284267e-07, "loss": 0.0166, "step": 75830 }, { "grad_norm": 0.21940456330776215, "learning_rate": 7.377972445989201e-07, "loss": 0.0168, "step": 75840 }, { "grad_norm": 0.20877082645893097, "learning_rate": 7.342639510325377e-07, "loss": 0.0164, "step": 75850 }, { "grad_norm": 0.32221537828445435, "learning_rate": 7.307390756330246e-07, "loss": 0.0165, "step": 75860 }, { "grad_norm": 0.2943877577781677, "learning_rate": 7.272226190026876e-07, "loss": 0.0122, "step": 75870 }, { "grad_norm": 0.1858738511800766, "learning_rate": 7.237145817423907e-07, "loss": 0.0146, "step": 75880 }, { "grad_norm": 0.27274829149246216, "learning_rate": 7.202149644515654e-07, "loss": 0.0181, "step": 75890 }, { "grad_norm": 0.2488672286272049, "learning_rate": 7.167237677281946e-07, "loss": 0.0149, "step": 75900 }, { "grad_norm": 0.34685152769088745, "learning_rate": 7.132409921688288e-07, "loss": 0.0188, "step": 75910 }, { "grad_norm": 0.23840011656284332, "learning_rate": 7.09766638368592e-07, "loss": 0.0166, "step": 75920 }, { "grad_norm": 0.30559730529785156, "learning_rate": 7.063007069211313e-07, "loss": 0.0182, "step": 75930 }, { "grad_norm": 0.39743050932884216, "learning_rate": 7.028431984187067e-07, "loss": 0.0184, "step": 75940 }, { "grad_norm": 0.11658677458763123, "learning_rate": 6.993941134520898e-07, "loss": 0.0126, "step": 75950 }, { "grad_norm": 0.2853301763534546, "learning_rate": 6.959534526106537e-07, "loss": 0.0141, "step": 75960 }, { "grad_norm": 0.25134021043777466, "learning_rate": 6.925212164822948e-07, "loss": 0.0145, "step": 75970 }, { "grad_norm": 0.27998048067092896, "learning_rate": 6.890974056535049e-07, "loss": 0.0196, "step": 75980 }, { "grad_norm": 0.1717183142900467, "learning_rate": 6.856820207093106e-07, "loss": 0.0129, "step": 75990 }, { "grad_norm": 0.19000621140003204, "learning_rate": 6.822750622333063e-07, "loss": 0.0118, "step": 76000 }, { "grad_norm": 0.3280876874923706, "learning_rate": 6.788765308076539e-07, "loss": 0.0192, "step": 76010 }, { "grad_norm": 0.23177973926067352, "learning_rate": 6.754864270130668e-07, "loss": 0.0141, "step": 76020 }, { "grad_norm": 0.19575068354606628, "learning_rate": 6.721047514288203e-07, "loss": 0.0127, "step": 76030 }, { "grad_norm": 0.3143620789051056, "learning_rate": 6.687315046327469e-07, "loss": 0.0187, "step": 76040 }, { "grad_norm": 0.3403315842151642, "learning_rate": 6.65366687201252e-07, "loss": 0.0181, "step": 76050 }, { "grad_norm": 0.25502684712409973, "learning_rate": 6.620102997092814e-07, "loss": 0.0204, "step": 76060 }, { "grad_norm": 0.20599456131458282, "learning_rate": 6.586623427303596e-07, "loss": 0.0136, "step": 76070 }, { "grad_norm": 0.17807850241661072, "learning_rate": 6.553228168365455e-07, "loss": 0.017, "step": 76080 }, { "grad_norm": 0.46401289105415344, "learning_rate": 6.519917225984884e-07, "loss": 0.0238, "step": 76090 }, { "grad_norm": 0.2392270565032959, "learning_rate": 6.486690605853718e-07, "loss": 0.0155, "step": 76100 }, { "grad_norm": 0.36718636751174927, "learning_rate": 6.453548313649527e-07, "loss": 0.0141, "step": 76110 }, { "grad_norm": 0.5506377220153809, "learning_rate": 6.420490355035446e-07, "loss": 0.0195, "step": 76120 }, { "grad_norm": 0.33930760622024536, "learning_rate": 6.387516735660071e-07, "loss": 0.0152, "step": 76130 }, { "grad_norm": 0.1799507737159729, "learning_rate": 6.35462746115778e-07, "loss": 0.0112, "step": 76140 }, { "grad_norm": 0.26490408182144165, "learning_rate": 6.321822537148358e-07, "loss": 0.0175, "step": 76150 }, { "grad_norm": 0.12658587098121643, "learning_rate": 6.289101969237432e-07, "loss": 0.0132, "step": 76160 }, { "grad_norm": 0.3798108994960785, "learning_rate": 6.256465763015918e-07, "loss": 0.0173, "step": 76170 }, { "grad_norm": 0.29381099343299866, "learning_rate": 6.22391392406052e-07, "loss": 0.0133, "step": 76180 }, { "grad_norm": 0.25563645362854004, "learning_rate": 6.191446457933403e-07, "loss": 0.0145, "step": 76190 }, { "grad_norm": 0.24054144322872162, "learning_rate": 6.159063370182406e-07, "loss": 0.0156, "step": 76200 }, { "grad_norm": 0.3091430962085724, "learning_rate": 6.126764666340879e-07, "loss": 0.0138, "step": 76210 }, { "grad_norm": 0.13137421011924744, "learning_rate": 6.094550351927852e-07, "loss": 0.0138, "step": 76220 }, { "grad_norm": 0.2672315835952759, "learning_rate": 6.062420432447757e-07, "loss": 0.0148, "step": 76230 }, { "grad_norm": 0.18669234216213226, "learning_rate": 6.030374913390813e-07, "loss": 0.0153, "step": 76240 }, { "grad_norm": 0.37048161029815674, "learning_rate": 5.998413800232694e-07, "loss": 0.0181, "step": 76250 }, { "grad_norm": 0.2524470090866089, "learning_rate": 5.966537098434755e-07, "loss": 0.0143, "step": 76260 }, { "grad_norm": 0.21797993779182434, "learning_rate": 5.934744813443694e-07, "loss": 0.0141, "step": 76270 }, { "grad_norm": 0.25089365243911743, "learning_rate": 5.903036950692054e-07, "loss": 0.0177, "step": 76280 }, { "grad_norm": 0.15635566413402557, "learning_rate": 5.871413515597835e-07, "loss": 0.0227, "step": 76290 }, { "grad_norm": 0.3274947702884674, "learning_rate": 5.839874513564547e-07, "loss": 0.0111, "step": 76300 }, { "grad_norm": 0.2299400120973587, "learning_rate": 5.808419949981436e-07, "loss": 0.0139, "step": 76310 }, { "grad_norm": 0.11260773986577988, "learning_rate": 5.777049830223257e-07, "loss": 0.0161, "step": 76320 }, { "grad_norm": 0.3005184531211853, "learning_rate": 5.745764159650114e-07, "loss": 0.0126, "step": 76330 }, { "grad_norm": 0.3906448483467102, "learning_rate": 5.71456294360806e-07, "loss": 0.0178, "step": 76340 }, { "grad_norm": 0.2497321218252182, "learning_rate": 5.683446187428443e-07, "loss": 0.0192, "step": 76350 }, { "grad_norm": 0.3024030029773712, "learning_rate": 5.652413896428288e-07, "loss": 0.0162, "step": 76360 }, { "grad_norm": 0.30556440353393555, "learning_rate": 5.621466075910131e-07, "loss": 0.0149, "step": 76370 }, { "grad_norm": 0.25267964601516724, "learning_rate": 5.590602731162187e-07, "loss": 0.0133, "step": 76380 }, { "grad_norm": 0.35232236981391907, "learning_rate": 5.559823867458125e-07, "loss": 0.0129, "step": 76390 }, { "grad_norm": 0.4015028178691864, "learning_rate": 5.529129490057128e-07, "loss": 0.019, "step": 76400 }, { "grad_norm": 0.24854357540607452, "learning_rate": 5.498519604204167e-07, "loss": 0.0157, "step": 76410 }, { "grad_norm": 0.26942822337150574, "learning_rate": 5.467994215129501e-07, "loss": 0.0157, "step": 76420 }, { "grad_norm": 0.19270557165145874, "learning_rate": 5.437553328049183e-07, "loss": 0.0134, "step": 76430 }, { "grad_norm": 0.46513932943344116, "learning_rate": 5.407196948164661e-07, "loss": 0.0182, "step": 76440 }, { "grad_norm": 0.32704102993011475, "learning_rate": 5.376925080663009e-07, "loss": 0.016, "step": 76450 }, { "grad_norm": 0.28045347332954407, "learning_rate": 5.346737730716977e-07, "loss": 0.0207, "step": 76460 }, { "grad_norm": 0.18827006220817566, "learning_rate": 5.316634903484607e-07, "loss": 0.0102, "step": 76470 }, { "grad_norm": 0.34063756465911865, "learning_rate": 5.286616604109728e-07, "loss": 0.0154, "step": 76480 }, { "grad_norm": 0.24963273108005524, "learning_rate": 5.256682837721627e-07, "loss": 0.0194, "step": 76490 }, { "grad_norm": 0.25144222378730774, "learning_rate": 5.226833609435156e-07, "loss": 0.0136, "step": 76500 }, { "grad_norm": 0.2530090808868408, "learning_rate": 5.19706892435079e-07, "loss": 0.0175, "step": 76510 }, { "grad_norm": 0.35905593633651733, "learning_rate": 5.167388787554406e-07, "loss": 0.0185, "step": 76520 }, { "grad_norm": 0.24829287827014923, "learning_rate": 5.137793204117614e-07, "loss": 0.0174, "step": 76530 }, { "grad_norm": 0.3561054766178131, "learning_rate": 5.108282179097479e-07, "loss": 0.019, "step": 76540 }, { "grad_norm": 0.21828527748584747, "learning_rate": 5.078855717536523e-07, "loss": 0.0138, "step": 76550 }, { "grad_norm": 0.1436656266450882, "learning_rate": 5.049513824463059e-07, "loss": 0.0127, "step": 76560 }, { "grad_norm": 0.19559413194656372, "learning_rate": 5.020256504890742e-07, "loss": 0.0165, "step": 76570 }, { "grad_norm": 0.28466981649398804, "learning_rate": 4.991083763818849e-07, "loss": 0.02, "step": 76580 }, { "grad_norm": 0.2687489688396454, "learning_rate": 4.961995606232228e-07, "loss": 0.019, "step": 76590 }, { "grad_norm": 0.273258239030838, "learning_rate": 4.932992037101236e-07, "loss": 0.0163, "step": 76600 }, { "grad_norm": 0.377768337726593, "learning_rate": 4.904073061381798e-07, "loss": 0.0212, "step": 76610 }, { "grad_norm": 0.2212773859500885, "learning_rate": 4.875238684015349e-07, "loss": 0.0212, "step": 76620 }, { "grad_norm": 0.2706802189350128, "learning_rate": 4.846488909928948e-07, "loss": 0.0121, "step": 76630 }, { "grad_norm": 0.2647736370563507, "learning_rate": 4.81782374403511e-07, "loss": 0.018, "step": 76640 }, { "grad_norm": 0.16322419047355652, "learning_rate": 4.789243191231918e-07, "loss": 0.0149, "step": 76650 }, { "grad_norm": 0.34443485736846924, "learning_rate": 4.76074725640302e-07, "loss": 0.0187, "step": 76660 }, { "grad_norm": 0.34078270196914673, "learning_rate": 4.732335944417632e-07, "loss": 0.0255, "step": 76670 }, { "grad_norm": 0.24710245430469513, "learning_rate": 4.704009260130371e-07, "loss": 0.0179, "step": 76680 }, { "grad_norm": 0.2658557593822479, "learning_rate": 4.675767208381587e-07, "loss": 0.0183, "step": 76690 }, { "grad_norm": 0.22234968841075897, "learning_rate": 4.647609793997032e-07, "loss": 0.0158, "step": 76700 }, { "grad_norm": 0.16884267330169678, "learning_rate": 4.619537021788023e-07, "loss": 0.014, "step": 76710 }, { "grad_norm": 0.25486132502555847, "learning_rate": 4.5915488965515005e-07, "loss": 0.0156, "step": 76720 }, { "grad_norm": 0.37946128845214844, "learning_rate": 4.563645423069807e-07, "loss": 0.0147, "step": 76730 }, { "grad_norm": 0.4301549196243286, "learning_rate": 4.535826606110849e-07, "loss": 0.0151, "step": 76740 }, { "grad_norm": 0.29296889901161194, "learning_rate": 4.508092450428214e-07, "loss": 0.0189, "step": 76750 }, { "grad_norm": 0.3045935034751892, "learning_rate": 4.480442960760778e-07, "loss": 0.0192, "step": 76760 }, { "grad_norm": 0.21611055731773376, "learning_rate": 4.4528781418332053e-07, "loss": 0.014, "step": 76770 }, { "grad_norm": 0.3632727265357971, "learning_rate": 4.4253979983555073e-07, "loss": 0.024, "step": 76780 }, { "grad_norm": 0.16629862785339355, "learning_rate": 4.3980025350233154e-07, "loss": 0.0143, "step": 76790 }, { "grad_norm": 0.20204974710941315, "learning_rate": 4.370691756517664e-07, "loss": 0.0166, "step": 76800 }, { "grad_norm": 0.2004856914281845, "learning_rate": 4.3434656675053753e-07, "loss": 0.0165, "step": 76810 }, { "grad_norm": 0.31915515661239624, "learning_rate": 4.3163242726385613e-07, "loss": 0.0154, "step": 76820 }, { "grad_norm": 0.24851632118225098, "learning_rate": 4.289267576554956e-07, "loss": 0.0167, "step": 76830 }, { "grad_norm": 0.4350730776786804, "learning_rate": 4.262295583877807e-07, "loss": 0.0121, "step": 76840 }, { "grad_norm": 0.19043520092964172, "learning_rate": 4.235408299215815e-07, "loss": 0.0177, "step": 76850 }, { "grad_norm": 0.19846202433109283, "learning_rate": 4.2086057271634173e-07, "loss": 0.0127, "step": 76860 }, { "grad_norm": 0.28609851002693176, "learning_rate": 4.181887872300394e-07, "loss": 0.0166, "step": 76870 }, { "grad_norm": 0.20883117616176605, "learning_rate": 4.155254739192038e-07, "loss": 0.0145, "step": 76880 }, { "grad_norm": 0.39155638217926025, "learning_rate": 4.128706332389265e-07, "loss": 0.0242, "step": 76890 }, { "grad_norm": 0.39839881658554077, "learning_rate": 4.1022426564284453e-07, "loss": 0.0169, "step": 76900 }, { "grad_norm": 0.18140406906604767, "learning_rate": 4.075863715831574e-07, "loss": 0.0147, "step": 76910 }, { "grad_norm": 0.2428084909915924, "learning_rate": 4.0495695151059887e-07, "loss": 0.0208, "step": 76920 }, { "grad_norm": 0.31221044063568115, "learning_rate": 4.023360058744763e-07, "loss": 0.0149, "step": 76930 }, { "grad_norm": 0.2692212462425232, "learning_rate": 3.997235351226314e-07, "loss": 0.0104, "step": 76940 }, { "grad_norm": 0.25555211305618286, "learning_rate": 3.971195397014571e-07, "loss": 0.0168, "step": 76950 }, { "grad_norm": 0.28321439027786255, "learning_rate": 3.9452402005591417e-07, "loss": 0.0234, "step": 76960 }, { "grad_norm": 0.28022220730781555, "learning_rate": 3.9193697662950334e-07, "loss": 0.0127, "step": 76970 }, { "grad_norm": 0.4411988854408264, "learning_rate": 3.893584098642822e-07, "loss": 0.0227, "step": 76980 }, { "grad_norm": 0.21648627519607544, "learning_rate": 3.867883202008538e-07, "loss": 0.0174, "step": 76990 }, { "grad_norm": 0.22150172293186188, "learning_rate": 3.842267080783779e-07, "loss": 0.0142, "step": 77000 }, { "grad_norm": 0.2716163396835327, "learning_rate": 3.816735739345656e-07, "loss": 0.0174, "step": 77010 }, { "grad_norm": 0.25917088985443115, "learning_rate": 3.791289182056679e-07, "loss": 0.0195, "step": 77020 }, { "grad_norm": 0.2765424847602844, "learning_rate": 3.7659274132650913e-07, "loss": 0.0162, "step": 77030 }, { "grad_norm": 0.21196307241916656, "learning_rate": 3.7406504373044826e-07, "loss": 0.0124, "step": 77040 }, { "grad_norm": 0.25736063718795776, "learning_rate": 3.7154582584939533e-07, "loss": 0.0139, "step": 77050 }, { "grad_norm": 0.3718065619468689, "learning_rate": 3.690350881138227e-07, "loss": 0.018, "step": 77060 }, { "grad_norm": 0.22260797023773193, "learning_rate": 3.665328309527427e-07, "loss": 0.0174, "step": 77070 }, { "grad_norm": 0.23375073075294495, "learning_rate": 3.640390547937245e-07, "loss": 0.0159, "step": 77080 }, { "grad_norm": 0.2874743640422821, "learning_rate": 3.615537600628882e-07, "loss": 0.0171, "step": 77090 }, { "grad_norm": 0.2929687798023224, "learning_rate": 3.5907694718489427e-07, "loss": 0.017, "step": 77100 }, { "grad_norm": 0.2758352756500244, "learning_rate": 3.566086165829707e-07, "loss": 0.0229, "step": 77110 }, { "grad_norm": 0.38113662600517273, "learning_rate": 3.5414876867888024e-07, "loss": 0.0197, "step": 77120 }, { "grad_norm": 0.3090789020061493, "learning_rate": 3.5169740389295326e-07, "loss": 0.0211, "step": 77130 }, { "grad_norm": 0.1782923936843872, "learning_rate": 3.4925452264405466e-07, "loss": 0.0131, "step": 77140 }, { "grad_norm": 0.43642857670783997, "learning_rate": 3.468201253496062e-07, "loss": 0.0156, "step": 77150 }, { "grad_norm": 0.5345141887664795, "learning_rate": 3.443942124255861e-07, "loss": 0.0206, "step": 77160 }, { "grad_norm": 0.23599395155906677, "learning_rate": 3.4197678428650183e-07, "loss": 0.0216, "step": 77170 }, { "grad_norm": 0.22111700475215912, "learning_rate": 3.3956784134544504e-07, "loss": 0.0168, "step": 77180 }, { "grad_norm": 0.2788463532924652, "learning_rate": 3.3716738401402547e-07, "loss": 0.013, "step": 77190 }, { "grad_norm": 0.35275593400001526, "learning_rate": 3.3477541270241495e-07, "loss": 0.014, "step": 77200 }, { "grad_norm": 0.12376118451356888, "learning_rate": 3.3239192781934215e-07, "loss": 0.0173, "step": 77210 }, { "grad_norm": 0.24044251441955566, "learning_rate": 3.3001692977207563e-07, "loss": 0.0157, "step": 77220 }, { "grad_norm": 0.2702999413013458, "learning_rate": 3.276504189664409e-07, "loss": 0.0149, "step": 77230 }, { "grad_norm": 0.20947889983654022, "learning_rate": 3.252923958068088e-07, "loss": 0.0165, "step": 77240 }, { "grad_norm": 0.3931386172771454, "learning_rate": 3.2294286069609046e-07, "loss": 0.0153, "step": 77250 }, { "grad_norm": 0.19729971885681152, "learning_rate": 3.2060181403577583e-07, "loss": 0.0151, "step": 77260 }, { "grad_norm": 0.26287412643432617, "learning_rate": 3.1826925622587823e-07, "loss": 0.0159, "step": 77270 }, { "grad_norm": 0.35244107246398926, "learning_rate": 3.159451876649622e-07, "loss": 0.0172, "step": 77280 }, { "grad_norm": 0.26130756735801697, "learning_rate": 3.1362960875015444e-07, "loss": 0.0133, "step": 77290 }, { "grad_norm": 0.3550121486186981, "learning_rate": 3.1132251987711637e-07, "loss": 0.0193, "step": 77300 }, { "grad_norm": 0.2369992434978485, "learning_rate": 3.0902392144007695e-07, "loss": 0.0166, "step": 77310 }, { "grad_norm": 0.443534255027771, "learning_rate": 3.0673381383179436e-07, "loss": 0.0181, "step": 77320 }, { "grad_norm": 0.28939229249954224, "learning_rate": 3.0445219744358875e-07, "loss": 0.0153, "step": 77330 }, { "grad_norm": 0.15095765888690948, "learning_rate": 3.021790726653262e-07, "loss": 0.0204, "step": 77340 }, { "grad_norm": 0.287004292011261, "learning_rate": 2.9991443988542366e-07, "loss": 0.0187, "step": 77350 }, { "grad_norm": 0.27939870953559875, "learning_rate": 2.976582994908439e-07, "loss": 0.0142, "step": 77360 }, { "grad_norm": 0.3096957504749298, "learning_rate": 2.954106518671007e-07, "loss": 0.0175, "step": 77370 }, { "grad_norm": 0.2765882909297943, "learning_rate": 2.9317149739825356e-07, "loss": 0.0157, "step": 77380 }, { "grad_norm": 0.4280487298965454, "learning_rate": 2.909408364669075e-07, "loss": 0.0202, "step": 77390 }, { "grad_norm": 0.5177282094955444, "learning_rate": 2.887186694542299e-07, "loss": 0.0184, "step": 77400 }, { "grad_norm": 0.20315325260162354, "learning_rate": 2.865049967399225e-07, "loss": 0.0181, "step": 77410 }, { "grad_norm": 0.2256656140089035, "learning_rate": 2.842998187022439e-07, "loss": 0.0139, "step": 77420 }, { "grad_norm": 0.254751980304718, "learning_rate": 2.8210313571800374e-07, "loss": 0.0173, "step": 77430 }, { "grad_norm": 0.2850491404533386, "learning_rate": 2.7991494816255184e-07, "loss": 0.0154, "step": 77440 }, { "grad_norm": 0.21048282086849213, "learning_rate": 2.777352564097779e-07, "loss": 0.0132, "step": 77450 }, { "grad_norm": 0.2509320378303528, "learning_rate": 2.755640608321508e-07, "loss": 0.0158, "step": 77460 }, { "grad_norm": 0.2603122591972351, "learning_rate": 2.73401361800657e-07, "loss": 0.0143, "step": 77470 }, { "grad_norm": 0.19211599230766296, "learning_rate": 2.7124715968484537e-07, "loss": 0.0146, "step": 77480 }, { "grad_norm": 0.2079826444387436, "learning_rate": 2.691014548528104e-07, "loss": 0.0154, "step": 77490 }, { "grad_norm": 0.1542251855134964, "learning_rate": 2.669642476711864e-07, "loss": 0.0158, "step": 77500 }, { "grad_norm": 0.31034788489341736, "learning_rate": 2.648355385051815e-07, "loss": 0.0194, "step": 77510 }, { "grad_norm": 0.2935277819633484, "learning_rate": 2.627153277185157e-07, "loss": 0.0172, "step": 77520 }, { "grad_norm": 0.2852359712123871, "learning_rate": 2.606036156734881e-07, "loss": 0.0141, "step": 77530 }, { "grad_norm": 0.3080524802207947, "learning_rate": 2.5850040273092127e-07, "loss": 0.0143, "step": 77540 }, { "grad_norm": 0.3259744644165039, "learning_rate": 2.5640568925020536e-07, "loss": 0.023, "step": 77550 }, { "grad_norm": 0.19200468063354492, "learning_rate": 2.5431947558927083e-07, "loss": 0.0187, "step": 77560 }, { "grad_norm": 0.2204982042312622, "learning_rate": 2.522417621045825e-07, "loss": 0.0135, "step": 77570 }, { "grad_norm": 0.2786376178264618, "learning_rate": 2.501725491511786e-07, "loss": 0.0212, "step": 77580 }, { "grad_norm": 0.18107649683952332, "learning_rate": 2.481118370826263e-07, "loss": 0.0157, "step": 77590 }, { "grad_norm": 0.284617155790329, "learning_rate": 2.4605962625104393e-07, "loss": 0.0155, "step": 77600 }, { "grad_norm": 0.1805233508348465, "learning_rate": 2.440159170070955e-07, "loss": 0.0214, "step": 77610 }, { "grad_norm": 0.20095035433769226, "learning_rate": 2.41980709699996e-07, "loss": 0.0141, "step": 77620 }, { "grad_norm": 0.1926993876695633, "learning_rate": 2.3995400467751174e-07, "loss": 0.0156, "step": 77630 }, { "grad_norm": 0.3915431797504425, "learning_rate": 2.3793580228594902e-07, "loss": 0.0189, "step": 77640 }, { "grad_norm": 0.20613445341587067, "learning_rate": 2.3592610287015982e-07, "loss": 0.0156, "step": 77650 }, { "grad_norm": 0.277484655380249, "learning_rate": 2.3392490677354718e-07, "loss": 0.0138, "step": 77660 }, { "grad_norm": 0.4352057874202728, "learning_rate": 2.3193221433806533e-07, "loss": 0.0139, "step": 77670 }, { "grad_norm": 0.22563199698925018, "learning_rate": 2.299480259042086e-07, "loss": 0.0158, "step": 77680 }, { "grad_norm": 0.2479342520236969, "learning_rate": 2.2797234181102244e-07, "loss": 0.0159, "step": 77690 }, { "grad_norm": 0.20624586939811707, "learning_rate": 2.2600516239609238e-07, "loss": 0.0149, "step": 77700 }, { "grad_norm": 0.13791461288928986, "learning_rate": 2.240464879955606e-07, "loss": 0.0216, "step": 77710 }, { "grad_norm": 0.346546471118927, "learning_rate": 2.2209631894410387e-07, "loss": 0.0165, "step": 77720 }, { "grad_norm": 0.2117190957069397, "learning_rate": 2.2015465557496117e-07, "loss": 0.0137, "step": 77730 }, { "grad_norm": 0.27261531352996826, "learning_rate": 2.1822149821990602e-07, "loss": 0.0156, "step": 77740 }, { "grad_norm": 0.47465264797210693, "learning_rate": 2.1629684720926303e-07, "loss": 0.0288, "step": 77750 }, { "grad_norm": 0.2938443124294281, "learning_rate": 2.1438070287189693e-07, "loss": 0.0132, "step": 77760 }, { "grad_norm": 0.25465601682662964, "learning_rate": 2.1247306553523472e-07, "loss": 0.0162, "step": 77770 }, { "grad_norm": 0.18240879476070404, "learning_rate": 2.105739355252323e-07, "loss": 0.0175, "step": 77780 }, { "grad_norm": 0.15363231301307678, "learning_rate": 2.0868331316639678e-07, "loss": 0.0128, "step": 77790 }, { "grad_norm": 0.3468427360057831, "learning_rate": 2.0680119878179193e-07, "loss": 0.0175, "step": 77800 }, { "grad_norm": 0.34422698616981506, "learning_rate": 2.0492759269301054e-07, "loss": 0.0167, "step": 77810 }, { "grad_norm": 0.8500049710273743, "learning_rate": 2.0306249522021315e-07, "loss": 0.0143, "step": 77820 }, { "grad_norm": 0.20869925618171692, "learning_rate": 2.0120590668207816e-07, "loss": 0.0133, "step": 77830 }, { "grad_norm": 0.4869578182697296, "learning_rate": 1.993578273958574e-07, "loss": 0.0176, "step": 77840 }, { "grad_norm": 0.32194581627845764, "learning_rate": 1.975182576773371e-07, "loss": 0.0143, "step": 77850 }, { "grad_norm": 0.43903848528862, "learning_rate": 1.9568719784083812e-07, "loss": 0.022, "step": 77860 }, { "grad_norm": 0.3968982696533203, "learning_rate": 1.9386464819924898e-07, "loss": 0.0174, "step": 77870 }, { "grad_norm": 0.27052003145217896, "learning_rate": 1.9205060906399285e-07, "loss": 0.0159, "step": 77880 }, { "grad_norm": 0.21354390680789948, "learning_rate": 1.9024508074503845e-07, "loss": 0.0183, "step": 77890 }, { "grad_norm": 0.21551725268363953, "learning_rate": 1.8844806355089452e-07, "loss": 0.0298, "step": 77900 }, { "grad_norm": 0.2629096806049347, "learning_rate": 1.8665955778863208e-07, "loss": 0.0212, "step": 77910 }, { "grad_norm": 0.22283923625946045, "learning_rate": 1.848795637638512e-07, "loss": 0.0159, "step": 77920 }, { "grad_norm": 0.3450230360031128, "learning_rate": 1.831080817807085e-07, "loss": 0.0201, "step": 77930 }, { "grad_norm": 0.18622490763664246, "learning_rate": 1.813451121418952e-07, "loss": 0.0241, "step": 77940 }, { "grad_norm": 0.30593299865722656, "learning_rate": 1.795906551486648e-07, "loss": 0.014, "step": 77950 }, { "grad_norm": 0.23724348843097687, "learning_rate": 1.7784471110079414e-07, "loss": 0.0171, "step": 77960 }, { "grad_norm": 0.35293474793434143, "learning_rate": 1.7610728029662793e-07, "loss": 0.0146, "step": 77970 }, { "grad_norm": 0.15981093049049377, "learning_rate": 1.7437836303303979e-07, "loss": 0.0111, "step": 77980 }, { "grad_norm": 0.28355279564857483, "learning_rate": 1.726579596054545e-07, "loss": 0.0113, "step": 77990 }, { "grad_norm": 0.3320918083190918, "learning_rate": 1.7094607030784803e-07, "loss": 0.0179, "step": 78000 }, { "grad_norm": 0.12550583481788635, "learning_rate": 1.6924269543272532e-07, "loss": 0.0152, "step": 78010 }, { "grad_norm": 0.24380731582641602, "learning_rate": 1.6754783527115348e-07, "loss": 0.0153, "step": 78020 }, { "grad_norm": 0.18839764595031738, "learning_rate": 1.6586149011273422e-07, "loss": 0.0174, "step": 78030 }, { "grad_norm": 0.2571549117565155, "learning_rate": 1.6418366024562038e-07, "loss": 0.0234, "step": 78040 }, { "grad_norm": 0.25661516189575195, "learning_rate": 1.6251434595651037e-07, "loss": 0.0162, "step": 78050 }, { "grad_norm": 0.39847269654273987, "learning_rate": 1.608535475306372e-07, "loss": 0.0187, "step": 78060 }, { "grad_norm": 0.18357741832733154, "learning_rate": 1.5920126525179048e-07, "loss": 0.0167, "step": 78070 }, { "grad_norm": 0.2503470778465271, "learning_rate": 1.5755749940229435e-07, "loss": 0.0125, "step": 78080 }, { "grad_norm": 0.22793696820735931, "learning_rate": 1.559222502630353e-07, "loss": 0.0192, "step": 78090 }, { "grad_norm": 0.32307684421539307, "learning_rate": 1.5429551811341757e-07, "loss": 0.018, "step": 78100 }, { "grad_norm": 0.21485541760921478, "learning_rate": 1.5267730323141882e-07, "loss": 0.0205, "step": 78110 }, { "grad_norm": 0.3166072964668274, "learning_rate": 1.5106760589353454e-07, "loss": 0.0151, "step": 78120 }, { "grad_norm": 0.1671709418296814, "learning_rate": 1.4946642637483355e-07, "loss": 0.0136, "step": 78130 }, { "grad_norm": 0.3579616844654083, "learning_rate": 1.4787376494889703e-07, "loss": 0.0156, "step": 78140 }, { "grad_norm": 0.2661765217781067, "learning_rate": 1.4628962188787955e-07, "loss": 0.0161, "step": 78150 }, { "grad_norm": 0.27878233790397644, "learning_rate": 1.4471399746247006e-07, "loss": 0.0168, "step": 78160 }, { "grad_norm": 0.1367245316505432, "learning_rate": 1.4314689194188103e-07, "loss": 0.0175, "step": 78170 }, { "grad_norm": 0.5070884823799133, "learning_rate": 1.4158830559390933e-07, "loss": 0.0176, "step": 78180 }, { "grad_norm": 0.38138964772224426, "learning_rate": 1.4003823868486422e-07, "loss": 0.0296, "step": 78190 }, { "grad_norm": 0.25343433022499084, "learning_rate": 1.3849669147960598e-07, "loss": 0.0182, "step": 78200 }, { "grad_norm": 0.24139636754989624, "learning_rate": 1.3696366424155726e-07, "loss": 0.0143, "step": 78210 }, { "grad_norm": 0.42917400598526, "learning_rate": 1.35439157232653e-07, "loss": 0.0221, "step": 78220 }, { "grad_norm": 0.2538667619228363, "learning_rate": 1.3392317071340144e-07, "loss": 0.0163, "step": 78230 }, { "grad_norm": 0.3658050298690796, "learning_rate": 1.3241570494283984e-07, "loss": 0.0185, "step": 78240 }, { "grad_norm": 0.29131701588630676, "learning_rate": 1.3091676017855103e-07, "loss": 0.0178, "step": 78250 }, { "grad_norm": 0.3004339933395386, "learning_rate": 1.2942633667666904e-07, "loss": 0.0161, "step": 78260 }, { "grad_norm": 0.14439038932323456, "learning_rate": 1.2794443469185679e-07, "loss": 0.0205, "step": 78270 }, { "grad_norm": 0.3283950388431549, "learning_rate": 1.2647105447734508e-07, "loss": 0.014, "step": 78280 }, { "grad_norm": 0.31332018971443176, "learning_rate": 1.2500619628488254e-07, "loss": 0.0157, "step": 78290 }, { "grad_norm": 0.284039169549942, "learning_rate": 1.2354986036477446e-07, "loss": 0.0198, "step": 78300 }, { "grad_norm": 0.22202149033546448, "learning_rate": 1.221020469658718e-07, "loss": 0.0162, "step": 78310 }, { "grad_norm": 0.37749409675598145, "learning_rate": 1.2066275633556556e-07, "loss": 0.0172, "step": 78320 }, { "grad_norm": 0.43868765234947205, "learning_rate": 1.192319887197979e-07, "loss": 0.0142, "step": 78330 }, { "grad_norm": 0.17441372573375702, "learning_rate": 1.1780974436303438e-07, "loss": 0.0154, "step": 78340 }, { "grad_norm": 0.24356530606746674, "learning_rate": 1.1639602350830836e-07, "loss": 0.0187, "step": 78350 }, { "grad_norm": 0.24341589212417603, "learning_rate": 1.1499082639718217e-07, "loss": 0.0175, "step": 78360 }, { "grad_norm": 0.11876644194126129, "learning_rate": 1.1359415326976374e-07, "loss": 0.0128, "step": 78370 }, { "grad_norm": 0.2540845274925232, "learning_rate": 1.1220600436470663e-07, "loss": 0.0149, "step": 78380 }, { "grad_norm": 0.32082420587539673, "learning_rate": 1.1082637991920996e-07, "loss": 0.0202, "step": 78390 }, { "grad_norm": 0.37230199575424194, "learning_rate": 1.0945528016901851e-07, "loss": 0.0189, "step": 78400 }, { "grad_norm": 0.3092486262321472, "learning_rate": 1.0809270534840599e-07, "loss": 0.0176, "step": 78410 }, { "grad_norm": 0.3064783215522766, "learning_rate": 1.067386556902028e-07, "loss": 0.0156, "step": 78420 }, { "grad_norm": 0.15185588598251343, "learning_rate": 1.05393131425785e-07, "loss": 0.017, "step": 78430 }, { "grad_norm": 0.33403661847114563, "learning_rate": 1.0405613278505199e-07, "loss": 0.0143, "step": 78440 }, { "grad_norm": 0.30232149362564087, "learning_rate": 1.027276599964766e-07, "loss": 0.0194, "step": 78450 }, { "grad_norm": 0.2710351347923279, "learning_rate": 1.0140771328704391e-07, "loss": 0.0128, "step": 78460 }, { "grad_norm": 0.2613557279109955, "learning_rate": 1.0009629288231237e-07, "loss": 0.0179, "step": 78470 }, { "grad_norm": 0.21320785582065582, "learning_rate": 9.879339900635276e-08, "loss": 0.0182, "step": 78480 }, { "grad_norm": 0.3656618297100067, "learning_rate": 9.74990318817981e-08, "loss": 0.0212, "step": 78490 }, { "grad_norm": 0.3020022213459015, "learning_rate": 9.621319172982701e-08, "loss": 0.0149, "step": 78500 }, { "grad_norm": 0.3308510482311249, "learning_rate": 9.493587877015264e-08, "loss": 0.0166, "step": 78510 }, { "grad_norm": 0.36319831013679504, "learning_rate": 9.366709322102263e-08, "loss": 0.0145, "step": 78520 }, { "grad_norm": 0.2736055552959442, "learning_rate": 9.240683529924688e-08, "loss": 0.0157, "step": 78530 }, { "grad_norm": 0.17726659774780273, "learning_rate": 9.115510522016979e-08, "loss": 0.0151, "step": 78540 }, { "grad_norm": 0.20998205244541168, "learning_rate": 8.991190319767583e-08, "loss": 0.0182, "step": 78550 }, { "grad_norm": 0.2693849503993988, "learning_rate": 8.867722944419509e-08, "loss": 0.0142, "step": 78560 }, { "grad_norm": 0.23500272631645203, "learning_rate": 8.745108417069214e-08, "loss": 0.0142, "step": 78570 }, { "grad_norm": 0.23338429629802704, "learning_rate": 8.623346758669381e-08, "loss": 0.0126, "step": 78580 }, { "grad_norm": 0.3336118757724762, "learning_rate": 8.502437990025037e-08, "loss": 0.0131, "step": 78590 }, { "grad_norm": 0.24007245898246765, "learning_rate": 8.382382131795764e-08, "loss": 0.0152, "step": 78600 }, { "grad_norm": 0.24468807876110077, "learning_rate": 8.263179204496818e-08, "loss": 0.0198, "step": 78610 }, { "grad_norm": 0.18533474206924438, "learning_rate": 8.144829228496354e-08, "loss": 0.0141, "step": 78620 }, { "grad_norm": 0.3368026316165924, "learning_rate": 8.027332224016526e-08, "loss": 0.0168, "step": 78630 }, { "grad_norm": 0.27811747789382935, "learning_rate": 7.910688211135164e-08, "loss": 0.0196, "step": 78640 }, { "grad_norm": 0.2869374454021454, "learning_rate": 7.794897209783546e-08, "loss": 0.0236, "step": 78650 }, { "grad_norm": 0.2579083740711212, "learning_rate": 7.679959239746403e-08, "loss": 0.0186, "step": 78660 }, { "grad_norm": 0.11860818415880203, "learning_rate": 7.565874320664135e-08, "loss": 0.0163, "step": 78670 }, { "grad_norm": 0.29304468631744385, "learning_rate": 7.45264247203059e-08, "loss": 0.0189, "step": 78680 }, { "grad_norm": 0.3190048038959503, "learning_rate": 7.340263713194184e-08, "loss": 0.0134, "step": 78690 }, { "grad_norm": 0.4130377471446991, "learning_rate": 7.228738063356777e-08, "loss": 0.0168, "step": 78700 }, { "grad_norm": 0.39145976305007935, "learning_rate": 7.118065541575903e-08, "loss": 0.0185, "step": 78710 }, { "grad_norm": 0.21945279836654663, "learning_rate": 7.008246166761435e-08, "loss": 0.0099, "step": 78720 }, { "grad_norm": 0.33454516530036926, "learning_rate": 6.899279957679472e-08, "loss": 0.0149, "step": 78730 }, { "grad_norm": 0.248804971575737, "learning_rate": 6.791166932949011e-08, "loss": 0.015, "step": 78740 }, { "grad_norm": 0.27070099115371704, "learning_rate": 6.683907111043608e-08, "loss": 0.0174, "step": 78750 }, { "grad_norm": 0.1655091792345047, "learning_rate": 6.577500510290824e-08, "loss": 0.0187, "step": 78760 }, { "grad_norm": 0.41220203042030334, "learning_rate": 6.47194714887278e-08, "loss": 0.0298, "step": 78770 }, { "grad_norm": 0.28600451350212097, "learning_rate": 6.367247044825608e-08, "loss": 0.0146, "step": 78780 }, { "grad_norm": 0.22714115679264069, "learning_rate": 6.263400216039994e-08, "loss": 0.014, "step": 78790 }, { "grad_norm": 0.21311427652835846, "learning_rate": 6.160406680260078e-08, "loss": 0.0292, "step": 78800 }, { "grad_norm": 0.19378693401813507, "learning_rate": 6.058266455084561e-08, "loss": 0.0144, "step": 78810 }, { "grad_norm": 0.25057053565979004, "learning_rate": 5.956979557967257e-08, "loss": 0.0152, "step": 78820 }, { "grad_norm": 0.2294262796640396, "learning_rate": 5.856546006214325e-08, "loss": 0.0181, "step": 78830 }, { "grad_norm": 0.2305862158536911, "learning_rate": 5.7569658169881466e-08, "loss": 0.0149, "step": 78840 }, { "grad_norm": 0.4441611170768738, "learning_rate": 5.658239007303445e-08, "loss": 0.0232, "step": 78850 }, { "grad_norm": 0.26543158292770386, "learning_rate": 5.560365594030059e-08, "loss": 0.0191, "step": 78860 }, { "grad_norm": 0.16279909014701843, "learning_rate": 5.463345593891833e-08, "loss": 0.0112, "step": 78870 }, { "grad_norm": 0.1721169650554657, "learning_rate": 5.367179023467173e-08, "loss": 0.0192, "step": 78880 }, { "grad_norm": 0.19600653648376465, "learning_rate": 5.271865899187378e-08, "loss": 0.0146, "step": 78890 }, { "grad_norm": 0.275739848613739, "learning_rate": 5.177406237340532e-08, "loss": 0.0271, "step": 78900 }, { "grad_norm": 0.3039376735687256, "learning_rate": 5.083800054065946e-08, "loss": 0.0112, "step": 78910 }, { "grad_norm": 0.3249381184577942, "learning_rate": 4.991047365358603e-08, "loss": 0.0145, "step": 78920 }, { "grad_norm": 0.4045347571372986, "learning_rate": 4.899148187067493e-08, "loss": 0.0251, "step": 78930 }, { "grad_norm": 0.2154603898525238, "learning_rate": 4.808102534895609e-08, "loss": 0.0123, "step": 78940 }, { "grad_norm": 0.31295686960220337, "learning_rate": 4.717910424400507e-08, "loss": 0.017, "step": 78950 }, { "grad_norm": 0.395875483751297, "learning_rate": 4.628571870993193e-08, "loss": 0.0188, "step": 78960 }, { "grad_norm": 0.45450150966644287, "learning_rate": 4.5400868899392324e-08, "loss": 0.0168, "step": 78970 }, { "grad_norm": 0.35663849115371704, "learning_rate": 4.4524554963576436e-08, "loss": 0.0195, "step": 78980 }, { "grad_norm": 0.3525203466415405, "learning_rate": 4.36567770522367e-08, "loss": 0.0192, "step": 78990 }, { "grad_norm": 0.2036030888557434, "learning_rate": 4.279753531364339e-08, "loss": 0.0133, "step": 79000 }, { "grad_norm": 0.359333336353302, "learning_rate": 4.1946829894617955e-08, "loss": 0.021, "step": 79010 }, { "grad_norm": 0.2897184193134308, "learning_rate": 4.11046609405219e-08, "loss": 0.0235, "step": 79020 }, { "grad_norm": 0.4385201930999756, "learning_rate": 4.027102859526233e-08, "loss": 0.0133, "step": 79030 }, { "grad_norm": 0.4745353162288666, "learning_rate": 3.9445933001280856e-08, "loss": 0.0183, "step": 79040 }, { "grad_norm": 0.2191520482301712, "learning_rate": 3.8629374299564704e-08, "loss": 0.0197, "step": 79050 }, { "grad_norm": 0.2420509159564972, "learning_rate": 3.78213526296467e-08, "loss": 0.0171, "step": 79060 }, { "grad_norm": 0.17806808650493622, "learning_rate": 3.702186812958308e-08, "loss": 0.0155, "step": 79070 }, { "grad_norm": 0.2639656960964203, "learning_rate": 3.623092093599789e-08, "loss": 0.015, "step": 79080 }, { "grad_norm": 0.4329395890235901, "learning_rate": 3.544851118403303e-08, "loss": 0.0151, "step": 79090 }, { "grad_norm": 0.337544709444046, "learning_rate": 3.46746390073871e-08, "loss": 0.0158, "step": 79100 }, { "grad_norm": 0.36387062072753906, "learning_rate": 3.390930453829322e-08, "loss": 0.023, "step": 79110 }, { "grad_norm": 0.20477195084095, "learning_rate": 3.3152507907519005e-08, "loss": 0.0133, "step": 79120 }, { "grad_norm": 0.19305500388145447, "learning_rate": 3.240424924438323e-08, "loss": 0.0147, "step": 79130 }, { "grad_norm": 0.2818741500377655, "learning_rate": 3.1664528676750295e-08, "loss": 0.0188, "step": 79140 }, { "grad_norm": 0.2868594229221344, "learning_rate": 3.093334633100797e-08, "loss": 0.0251, "step": 79150 }, { "grad_norm": 0.3421196937561035, "learning_rate": 3.021070233210077e-08, "loss": 0.0158, "step": 79160 }, { "grad_norm": 0.2370985895395279, "learning_rate": 2.9496596803507692e-08, "loss": 0.0134, "step": 79170 }, { "grad_norm": 0.17559681832790375, "learning_rate": 2.879102986725335e-08, "loss": 0.0142, "step": 79180 }, { "grad_norm": 0.16303713619709015, "learning_rate": 2.809400164389131e-08, "loss": 0.0148, "step": 79190 }, { "grad_norm": 0.20198673009872437, "learning_rate": 2.740551225253185e-08, "loss": 0.018, "step": 79200 }, { "grad_norm": 0.23700354993343353, "learning_rate": 2.6725561810819754e-08, "loss": 0.0232, "step": 79210 }, { "grad_norm": 0.37945985794067383, "learning_rate": 2.605415043493431e-08, "loss": 0.0163, "step": 79220 }, { "grad_norm": 0.3052528202533722, "learning_rate": 2.5391278239605965e-08, "loss": 0.0192, "step": 79230 }, { "grad_norm": 0.3495563268661499, "learning_rate": 2.4736945338094118e-08, "loss": 0.0165, "step": 79240 }, { "grad_norm": 0.151637002825737, "learning_rate": 2.4091151842214887e-08, "loss": 0.0126, "step": 79250 }, { "grad_norm": 0.29322871565818787, "learning_rate": 2.3453897862318884e-08, "loss": 0.0141, "step": 79260 }, { "grad_norm": 0.14648579061031342, "learning_rate": 2.2825183507285686e-08, "loss": 0.0155, "step": 79270 }, { "grad_norm": 0.11905628442764282, "learning_rate": 2.220500888455157e-08, "loss": 0.0185, "step": 79280 }, { "grad_norm": 0.24021610617637634, "learning_rate": 2.1593374100081777e-08, "loss": 0.0189, "step": 79290 }, { "grad_norm": 0.3142549693584442, "learning_rate": 2.09902792583927e-08, "loss": 0.0219, "step": 79300 }, { "grad_norm": 0.26157069206237793, "learning_rate": 2.0395724462540788e-08, "loss": 0.014, "step": 79310 }, { "grad_norm": 0.11289995908737183, "learning_rate": 1.9809709814111453e-08, "loss": 0.0109, "step": 79320 }, { "grad_norm": 0.4755651354789734, "learning_rate": 1.923223541324126e-08, "loss": 0.0216, "step": 79330 }, { "grad_norm": 0.27815142273902893, "learning_rate": 1.8663301358606833e-08, "loss": 0.0195, "step": 79340 }, { "grad_norm": 0.35885873436927795, "learning_rate": 1.8102907747419295e-08, "loss": 0.0208, "step": 79350 }, { "grad_norm": 0.259957492351532, "learning_rate": 1.7551054675435385e-08, "loss": 0.0164, "step": 79360 }, { "grad_norm": 0.27860522270202637, "learning_rate": 1.7007742236957447e-08, "loss": 0.0201, "step": 79370 }, { "grad_norm": 0.25463083386421204, "learning_rate": 1.647297052481678e-08, "loss": 0.0159, "step": 79380 }, { "grad_norm": 0.28993967175483704, "learning_rate": 1.5946739630390285e-08, "loss": 0.0206, "step": 79390 }, { "grad_norm": 0.18433062732219696, "learning_rate": 1.5429049643606032e-08, "loss": 0.0205, "step": 79400 }, { "grad_norm": 0.14734071493148804, "learning_rate": 1.4919900652909935e-08, "loss": 0.0211, "step": 79410 }, { "grad_norm": 0.3399134874343872, "learning_rate": 1.4419292745310175e-08, "loss": 0.025, "step": 79420 }, { "grad_norm": 0.3279474973678589, "learning_rate": 1.3927226006343885e-08, "loss": 0.0171, "step": 79430 }, { "grad_norm": 0.3929234445095062, "learning_rate": 1.3443700520093805e-08, "loss": 0.0218, "step": 79440 }, { "grad_norm": 0.18214507400989532, "learning_rate": 1.296871636917718e-08, "loss": 0.0156, "step": 79450 }, { "grad_norm": 0.29972562193870544, "learning_rate": 1.2502273634762419e-08, "loss": 0.0151, "step": 79460 }, { "grad_norm": 0.3762259781360626, "learning_rate": 1.2044372396546876e-08, "loss": 0.0174, "step": 79470 }, { "grad_norm": 0.3213222324848175, "learning_rate": 1.1595012732773524e-08, "loss": 0.017, "step": 79480 }, { "grad_norm": 0.35603067278862, "learning_rate": 1.1154194720225386e-08, "loss": 0.0154, "step": 79490 }, { "grad_norm": 0.3489323854446411, "learning_rate": 1.0721918434231093e-08, "loss": 0.0158, "step": 79500 }, { "grad_norm": 0.2940753102302551, "learning_rate": 1.0298183948648232e-08, "loss": 0.0162, "step": 79510 }, { "grad_norm": 0.2967495024204254, "learning_rate": 9.88299133588555e-09, "loss": 0.0149, "step": 79520 }, { "grad_norm": 0.23131661117076874, "learning_rate": 9.476340666891847e-09, "loss": 0.0242, "step": 79530 }, { "grad_norm": 0.36024805903434753, "learning_rate": 9.078232011139332e-09, "loss": 0.0146, "step": 79540 }, { "grad_norm": 0.30330371856689453, "learning_rate": 8.68866543666802e-09, "loss": 0.0177, "step": 79550 }, { "grad_norm": 0.1875361055135727, "learning_rate": 8.307641010035782e-09, "loss": 0.0152, "step": 79560 }, { "grad_norm": 0.21520745754241943, "learning_rate": 7.93515879635165e-09, "loss": 0.018, "step": 79570 }, { "grad_norm": 0.33318889141082764, "learning_rate": 7.571218859264706e-09, "loss": 0.0175, "step": 79580 }, { "grad_norm": 0.1623622179031372, "learning_rate": 7.215821260958544e-09, "loss": 0.0186, "step": 79590 }, { "grad_norm": 0.24799098074436188, "learning_rate": 6.868966062162363e-09, "loss": 0.0152, "step": 79600 }, { "grad_norm": 0.09300418198108673, "learning_rate": 6.530653322145419e-09, "loss": 0.0129, "step": 79610 }, { "grad_norm": 0.1711205691099167, "learning_rate": 6.200883098717025e-09, "loss": 0.0124, "step": 79620 }, { "grad_norm": 0.45822903513908386, "learning_rate": 5.879655448226551e-09, "loss": 0.0183, "step": 79630 }, { "grad_norm": 0.2673857808113098, "learning_rate": 5.566970425557872e-09, "loss": 0.0138, "step": 79640 }, { "grad_norm": 0.3454464077949524, "learning_rate": 5.26282808414047e-09, "loss": 0.0163, "step": 79650 }, { "grad_norm": 0.2574422061443329, "learning_rate": 4.967228475949437e-09, "loss": 0.0173, "step": 79660 }, { "grad_norm": 0.3686855137348175, "learning_rate": 4.680171651494369e-09, "loss": 0.0165, "step": 79670 }, { "grad_norm": 0.326215922832489, "learning_rate": 4.40165765981937e-09, "loss": 0.017, "step": 79680 }, { "grad_norm": 0.25465157628059387, "learning_rate": 4.131686548519698e-09, "loss": 0.0129, "step": 79690 }, { "grad_norm": 0.34889984130859375, "learning_rate": 3.8702583637251214e-09, "loss": 0.0138, "step": 79700 }, { "grad_norm": 0.1383337825536728, "learning_rate": 3.617373150105463e-09, "loss": 0.015, "step": 79710 }, { "grad_norm": 0.27374425530433655, "learning_rate": 3.3730309508706036e-09, "loss": 0.015, "step": 79720 }, { "grad_norm": 0.3593941926956177, "learning_rate": 3.137231807781582e-09, "loss": 0.013, "step": 79730 }, { "grad_norm": 0.19385282695293427, "learning_rate": 2.9099757611172894e-09, "loss": 0.0158, "step": 79740 }, { "grad_norm": 0.2657916843891144, "learning_rate": 2.6912628497133275e-09, "loss": 0.0158, "step": 79750 }, { "grad_norm": 0.3763248920440674, "learning_rate": 2.481093110945354e-09, "loss": 0.0204, "step": 79760 }, { "grad_norm": 0.16394703090190887, "learning_rate": 2.2794665807235327e-09, "loss": 0.0158, "step": 79770 }, { "grad_norm": 0.2664320468902588, "learning_rate": 2.0863832934980843e-09, "loss": 0.0143, "step": 79780 }, { "grad_norm": 0.3117581903934479, "learning_rate": 1.901843282264837e-09, "loss": 0.0193, "step": 79790 }, { "grad_norm": 0.1736520230770111, "learning_rate": 1.7258465785541245e-09, "loss": 0.0183, "step": 79800 }, { "grad_norm": 0.21627281606197357, "learning_rate": 1.5583932124418887e-09, "loss": 0.0181, "step": 79810 }, { "grad_norm": 0.3930692970752716, "learning_rate": 1.3994832125441282e-09, "loss": 0.0165, "step": 79820 }, { "grad_norm": 0.37306955456733704, "learning_rate": 1.2491166060057958e-09, "loss": 0.016, "step": 79830 }, { "grad_norm": 0.2986108362674713, "learning_rate": 1.1072934185230032e-09, "loss": 0.0157, "step": 79840 }, { "grad_norm": 0.4161982536315918, "learning_rate": 9.740136743319195e-10, "loss": 0.0175, "step": 79850 }, { "grad_norm": 0.19975723326206207, "learning_rate": 8.492773962087696e-10, "loss": 0.014, "step": 79860 }, { "grad_norm": 0.3159901797771454, "learning_rate": 7.330846054587337e-10, "loss": 0.0223, "step": 79870 }, { "grad_norm": 0.3383888006210327, "learning_rate": 6.254353219492526e-10, "loss": 0.0184, "step": 79880 }, { "grad_norm": 0.3497539460659027, "learning_rate": 5.263295640600685e-10, "loss": 0.013, "step": 79890 }, { "grad_norm": 0.15753450989723206, "learning_rate": 4.357673487387359e-10, "loss": 0.0119, "step": 79900 }, { "grad_norm": 0.3235551118850708, "learning_rate": 3.537486914506616e-10, "loss": 0.0136, "step": 79910 }, { "grad_norm": 0.2363913208246231, "learning_rate": 2.8027360621241117e-10, "loss": 0.0179, "step": 79920 }, { "grad_norm": 0.29086434841156006, "learning_rate": 2.153421055806071e-10, "loss": 0.0172, "step": 79930 }, { "grad_norm": 0.24057385325431824, "learning_rate": 1.5895420064637733e-10, "loss": 0.0159, "step": 79940 }, { "grad_norm": 0.21922197937965393, "learning_rate": 1.1110990105200891e-10, "loss": 0.0151, "step": 79950 }, { "grad_norm": 0.507436215877533, "learning_rate": 7.180921496874326e-11, "loss": 0.0176, "step": 79960 }, { "grad_norm": 0.3493092656135559, "learning_rate": 4.10521491134297e-11, "loss": 0.0146, "step": 79970 }, { "grad_norm": 0.19335876405239105, "learning_rate": 1.883870874297422e-11, "loss": 0.0132, "step": 79980 }, { "grad_norm": 0.4042668640613556, "learning_rate": 5.1688976432373584e-12, "loss": 0.0152, "step": 79990 }, { "grad_norm": 0.29723086953163147, "learning_rate": 4.271816234080461e-14, "loss": 0.0141, "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 }