{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 200000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.8391255140304565, "learning_rate": 6e-08, "loss": 1.2555, "step": 10 }, { "grad_norm": 0.9454184770584106, "learning_rate": 1.2666666666666666e-07, "loss": 1.2698, "step": 20 }, { "grad_norm": 0.8832136392593384, "learning_rate": 1.9333333333333334e-07, "loss": 1.2347, "step": 30 }, { "grad_norm": 0.9063571691513062, "learning_rate": 2.6e-07, "loss": 1.2411, "step": 40 }, { "grad_norm": 0.8986115455627441, "learning_rate": 3.266666666666667e-07, "loss": 1.2657, "step": 50 }, { "grad_norm": 0.960550844669342, "learning_rate": 3.933333333333333e-07, "loss": 1.2646, "step": 60 }, { "grad_norm": 1.0024666786193848, "learning_rate": 4.6e-07, "loss": 1.2509, "step": 70 }, { "grad_norm": 0.7508439421653748, "learning_rate": 5.266666666666667e-07, "loss": 1.2489, "step": 80 }, { "grad_norm": 0.8265761733055115, "learning_rate": 5.933333333333333e-07, "loss": 1.2419, "step": 90 }, { "grad_norm": 1.081796407699585, "learning_rate": 6.6e-07, "loss": 1.221, "step": 100 }, { "grad_norm": 1.0684542655944824, "learning_rate": 7.266666666666668e-07, "loss": 1.1475, "step": 110 }, { "grad_norm": 0.9137612581253052, "learning_rate": 7.933333333333334e-07, "loss": 1.1201, "step": 120 }, { "grad_norm": 1.0358893871307373, "learning_rate": 8.6e-07, "loss": 1.1149, "step": 130 }, { "grad_norm": 0.7965758442878723, "learning_rate": 9.266666666666667e-07, "loss": 1.0842, "step": 140 }, { "grad_norm": 0.7644563913345337, "learning_rate": 9.933333333333335e-07, "loss": 1.1045, "step": 150 }, { "grad_norm": 0.7802261710166931, "learning_rate": 1.06e-06, "loss": 1.1229, "step": 160 }, { "grad_norm": 0.7376846075057983, "learning_rate": 1.1266666666666667e-06, "loss": 1.0993, "step": 170 }, { "grad_norm": 0.8245558738708496, "learning_rate": 1.1933333333333335e-06, "loss": 1.1035, "step": 180 }, { "grad_norm": 0.7680610418319702, "learning_rate": 1.26e-06, "loss": 1.0976, "step": 190 }, { "grad_norm": 0.8455492258071899, "learning_rate": 1.3266666666666667e-06, "loss": 1.1045, "step": 200 }, { "grad_norm": 0.817874014377594, "learning_rate": 1.3933333333333335e-06, "loss": 1.1019, "step": 210 }, { "grad_norm": 2.8411567211151123, "learning_rate": 1.46e-06, "loss": 1.0879, "step": 220 }, { "grad_norm": 0.8665594458580017, "learning_rate": 1.5266666666666667e-06, "loss": 1.1084, "step": 230 }, { "grad_norm": 0.7159661054611206, "learning_rate": 1.5933333333333335e-06, "loss": 1.0909, "step": 240 }, { "grad_norm": 3.745157480239868, "learning_rate": 1.6600000000000002e-06, "loss": 1.0806, "step": 250 }, { "grad_norm": 0.7086023688316345, "learning_rate": 1.7266666666666667e-06, "loss": 1.0743, "step": 260 }, { "grad_norm": 0.9723825454711914, "learning_rate": 1.7933333333333332e-06, "loss": 1.1002, "step": 270 }, { "grad_norm": 0.8491377234458923, "learning_rate": 1.86e-06, "loss": 1.093, "step": 280 }, { "grad_norm": 0.9412674903869629, "learning_rate": 1.9266666666666667e-06, "loss": 1.0992, "step": 290 }, { "grad_norm": 0.7961090207099915, "learning_rate": 1.9933333333333334e-06, "loss": 1.0881, "step": 300 }, { "grad_norm": 1.4506871700286865, "learning_rate": 2.06e-06, "loss": 1.1079, "step": 310 }, { "grad_norm": 0.901292622089386, "learning_rate": 2.126666666666667e-06, "loss": 1.0885, "step": 320 }, { "grad_norm": 0.9550159573554993, "learning_rate": 2.1933333333333332e-06, "loss": 1.0984, "step": 330 }, { "grad_norm": 0.9133328795433044, "learning_rate": 2.26e-06, "loss": 1.0808, "step": 340 }, { "grad_norm": 1.802949070930481, "learning_rate": 2.326666666666667e-06, "loss": 1.0817, "step": 350 }, { "grad_norm": 1.0832804441452026, "learning_rate": 2.3933333333333334e-06, "loss": 1.0699, "step": 360 }, { "grad_norm": 0.7074794173240662, "learning_rate": 2.46e-06, "loss": 1.0998, "step": 370 }, { "grad_norm": 0.9364616274833679, "learning_rate": 2.526666666666667e-06, "loss": 1.0877, "step": 380 }, { "grad_norm": 0.8476104140281677, "learning_rate": 2.593333333333333e-06, "loss": 1.0852, "step": 390 }, { "grad_norm": 0.9374750256538391, "learning_rate": 2.66e-06, "loss": 1.0787, "step": 400 }, { "grad_norm": 0.7408366203308105, "learning_rate": 2.726666666666667e-06, "loss": 1.0827, "step": 410 }, { "grad_norm": 0.8761176466941833, "learning_rate": 2.7933333333333334e-06, "loss": 1.094, "step": 420 }, { "grad_norm": 0.7497237324714661, "learning_rate": 2.86e-06, "loss": 1.0624, "step": 430 }, { "grad_norm": 3.1770246028900146, "learning_rate": 2.926666666666667e-06, "loss": 1.0599, "step": 440 }, { "grad_norm": 0.8125482201576233, "learning_rate": 2.9933333333333336e-06, "loss": 1.0527, "step": 450 }, { "grad_norm": 1.002013921737671, "learning_rate": 3.06e-06, "loss": 1.0551, "step": 460 }, { "grad_norm": 1.6901048421859741, "learning_rate": 3.1266666666666667e-06, "loss": 1.0638, "step": 470 }, { "grad_norm": 0.9401182532310486, "learning_rate": 3.1933333333333334e-06, "loss": 1.0489, "step": 480 }, { "grad_norm": 1.009522795677185, "learning_rate": 3.2599999999999997e-06, "loss": 1.0503, "step": 490 }, { "grad_norm": 1.241262674331665, "learning_rate": 3.326666666666667e-06, "loss": 1.0248, "step": 500 }, { "grad_norm": 1.2674676179885864, "learning_rate": 3.3933333333333336e-06, "loss": 1.015, "step": 510 }, { "grad_norm": 1.679902195930481, "learning_rate": 3.46e-06, "loss": 1.007, "step": 520 }, { "grad_norm": 1.3841948509216309, "learning_rate": 3.526666666666667e-06, "loss": 1.0078, "step": 530 }, { "grad_norm": 1.4035173654556274, "learning_rate": 3.5933333333333334e-06, "loss": 0.9941, "step": 540 }, { "grad_norm": 1.7474493980407715, "learning_rate": 3.66e-06, "loss": 0.9848, "step": 550 }, { "grad_norm": 1.8152002096176147, "learning_rate": 3.7266666666666664e-06, "loss": 0.9645, "step": 560 }, { "grad_norm": 1.571502685546875, "learning_rate": 3.7933333333333336e-06, "loss": 0.9558, "step": 570 }, { "grad_norm": 1.782697319984436, "learning_rate": 3.86e-06, "loss": 0.9147, "step": 580 }, { "grad_norm": 1.8575778007507324, "learning_rate": 3.926666666666667e-06, "loss": 0.9175, "step": 590 }, { "grad_norm": 2.0003371238708496, "learning_rate": 3.993333333333334e-06, "loss": 0.8936, "step": 600 }, { "grad_norm": 2.118455648422241, "learning_rate": 4.06e-06, "loss": 0.8765, "step": 610 }, { "grad_norm": 1.99458646774292, "learning_rate": 4.126666666666667e-06, "loss": 0.8635, "step": 620 }, { "grad_norm": 1.7068768739700317, "learning_rate": 4.1933333333333336e-06, "loss": 0.848, "step": 630 }, { "grad_norm": 1.966342568397522, "learning_rate": 4.26e-06, "loss": 0.821, "step": 640 }, { "grad_norm": 2.45763897895813, "learning_rate": 4.326666666666667e-06, "loss": 0.7915, "step": 650 }, { "grad_norm": 2.332557439804077, "learning_rate": 4.393333333333333e-06, "loss": 0.8148, "step": 660 }, { "grad_norm": 1.8838034868240356, "learning_rate": 4.4600000000000005e-06, "loss": 0.7647, "step": 670 }, { "grad_norm": 2.2900192737579346, "learning_rate": 4.526666666666667e-06, "loss": 0.7666, "step": 680 }, { "grad_norm": 2.258183240890503, "learning_rate": 4.593333333333333e-06, "loss": 0.7611, "step": 690 }, { "grad_norm": 2.8988428115844727, "learning_rate": 4.66e-06, "loss": 0.7547, "step": 700 }, { "grad_norm": 2.752361536026001, "learning_rate": 4.726666666666667e-06, "loss": 0.7076, "step": 710 }, { "grad_norm": 3.302785873413086, "learning_rate": 4.793333333333334e-06, "loss": 0.7183, "step": 720 }, { "grad_norm": 2.552112579345703, "learning_rate": 4.86e-06, "loss": 0.6956, "step": 730 }, { "grad_norm": 2.4396770000457764, "learning_rate": 4.926666666666667e-06, "loss": 0.6635, "step": 740 }, { "grad_norm": 2.5192878246307373, "learning_rate": 4.9933333333333335e-06, "loss": 0.6853, "step": 750 }, { "grad_norm": 2.3035993576049805, "learning_rate": 5.06e-06, "loss": 0.6668, "step": 760 }, { "grad_norm": 2.3620851039886475, "learning_rate": 5.126666666666667e-06, "loss": 0.6576, "step": 770 }, { "grad_norm": 3.15126633644104, "learning_rate": 5.193333333333333e-06, "loss": 0.6798, "step": 780 }, { "grad_norm": 2.9177210330963135, "learning_rate": 5.2600000000000005e-06, "loss": 0.6222, "step": 790 }, { "grad_norm": 2.951901435852051, "learning_rate": 5.326666666666667e-06, "loss": 0.6116, "step": 800 }, { "grad_norm": 3.174729347229004, "learning_rate": 5.393333333333334e-06, "loss": 0.5729, "step": 810 }, { "grad_norm": 2.958993673324585, "learning_rate": 5.46e-06, "loss": 0.5676, "step": 820 }, { "grad_norm": 2.8093504905700684, "learning_rate": 5.5266666666666666e-06, "loss": 0.5688, "step": 830 }, { "grad_norm": 3.409456491470337, "learning_rate": 5.593333333333334e-06, "loss": 0.5479, "step": 840 }, { "grad_norm": 2.024752378463745, "learning_rate": 5.66e-06, "loss": 0.5384, "step": 850 }, { "grad_norm": 3.165050745010376, "learning_rate": 5.726666666666667e-06, "loss": 0.5204, "step": 860 }, { "grad_norm": 2.464144229888916, "learning_rate": 5.793333333333334e-06, "loss": 0.5406, "step": 870 }, { "grad_norm": 4.503970623016357, "learning_rate": 5.86e-06, "loss": 0.5057, "step": 880 }, { "grad_norm": 3.389353036880493, "learning_rate": 5.926666666666667e-06, "loss": 0.5146, "step": 890 }, { "grad_norm": 2.609349489212036, "learning_rate": 5.993333333333333e-06, "loss": 0.493, "step": 900 }, { "grad_norm": 2.6167969703674316, "learning_rate": 6.0600000000000004e-06, "loss": 0.4705, "step": 910 }, { "grad_norm": 3.328962564468384, "learning_rate": 6.126666666666667e-06, "loss": 0.4525, "step": 920 }, { "grad_norm": 3.4176218509674072, "learning_rate": 6.193333333333334e-06, "loss": 0.4662, "step": 930 }, { "grad_norm": 3.5649735927581787, "learning_rate": 6.26e-06, "loss": 0.4645, "step": 940 }, { "grad_norm": 3.1230459213256836, "learning_rate": 6.3266666666666665e-06, "loss": 0.4432, "step": 950 }, { "grad_norm": 3.495453357696533, "learning_rate": 6.393333333333333e-06, "loss": 0.4347, "step": 960 }, { "grad_norm": 4.016031742095947, "learning_rate": 6.460000000000001e-06, "loss": 0.462, "step": 970 }, { "grad_norm": 4.568574905395508, "learning_rate": 6.526666666666667e-06, "loss": 0.4125, "step": 980 }, { "grad_norm": 4.037368297576904, "learning_rate": 6.5933333333333335e-06, "loss": 0.3949, "step": 990 }, { "grad_norm": 3.149322032928467, "learning_rate": 6.660000000000001e-06, "loss": 0.3983, "step": 1000 }, { "grad_norm": 2.9036285877227783, "learning_rate": 6.726666666666667e-06, "loss": 0.369, "step": 1010 }, { "grad_norm": 3.3931548595428467, "learning_rate": 6.793333333333333e-06, "loss": 0.4058, "step": 1020 }, { "grad_norm": 3.034461498260498, "learning_rate": 6.8599999999999995e-06, "loss": 0.3335, "step": 1030 }, { "grad_norm": 3.5072524547576904, "learning_rate": 6.9266666666666675e-06, "loss": 0.3487, "step": 1040 }, { "grad_norm": 2.600381851196289, "learning_rate": 6.993333333333334e-06, "loss": 0.3198, "step": 1050 }, { "grad_norm": 2.875250816345215, "learning_rate": 7.06e-06, "loss": 0.3391, "step": 1060 }, { "grad_norm": 3.1709489822387695, "learning_rate": 7.126666666666667e-06, "loss": 0.3371, "step": 1070 }, { "grad_norm": 3.08388614654541, "learning_rate": 7.193333333333334e-06, "loss": 0.3172, "step": 1080 }, { "grad_norm": 4.453978538513184, "learning_rate": 7.26e-06, "loss": 0.292, "step": 1090 }, { "grad_norm": 4.178189754486084, "learning_rate": 7.326666666666666e-06, "loss": 0.2581, "step": 1100 }, { "grad_norm": 3.6300456523895264, "learning_rate": 7.393333333333334e-06, "loss": 0.3213, "step": 1110 }, { "grad_norm": 2.5389211177825928, "learning_rate": 7.4600000000000006e-06, "loss": 0.2455, "step": 1120 }, { "grad_norm": 3.0071606636047363, "learning_rate": 7.526666666666667e-06, "loss": 0.2673, "step": 1130 }, { "grad_norm": 3.2944495677948, "learning_rate": 7.593333333333334e-06, "loss": 0.2921, "step": 1140 }, { "grad_norm": 3.960489273071289, "learning_rate": 7.660000000000001e-06, "loss": 0.2752, "step": 1150 }, { "grad_norm": 4.6096296310424805, "learning_rate": 7.726666666666667e-06, "loss": 0.2366, "step": 1160 }, { "grad_norm": 3.6024560928344727, "learning_rate": 7.793333333333334e-06, "loss": 0.2334, "step": 1170 }, { "grad_norm": 3.3102385997772217, "learning_rate": 7.860000000000001e-06, "loss": 0.195, "step": 1180 }, { "grad_norm": 3.1402645111083984, "learning_rate": 7.926666666666666e-06, "loss": 0.2165, "step": 1190 }, { "grad_norm": 3.9637863636016846, "learning_rate": 7.993333333333334e-06, "loss": 0.2352, "step": 1200 }, { "grad_norm": 5.875804901123047, "learning_rate": 8.06e-06, "loss": 0.2474, "step": 1210 }, { "grad_norm": 2.5171585083007812, "learning_rate": 8.126666666666668e-06, "loss": 0.2069, "step": 1220 }, { "grad_norm": 3.261903762817383, "learning_rate": 8.193333333333333e-06, "loss": 0.2026, "step": 1230 }, { "grad_norm": 3.044342279434204, "learning_rate": 8.26e-06, "loss": 0.2051, "step": 1240 }, { "grad_norm": 3.5081605911254883, "learning_rate": 8.326666666666668e-06, "loss": 0.1794, "step": 1250 }, { "grad_norm": 2.696167230606079, "learning_rate": 8.393333333333333e-06, "loss": 0.2032, "step": 1260 }, { "grad_norm": 2.699617862701416, "learning_rate": 8.46e-06, "loss": 0.1707, "step": 1270 }, { "grad_norm": 2.889352798461914, "learning_rate": 8.526666666666667e-06, "loss": 0.1864, "step": 1280 }, { "grad_norm": 2.311837911605835, "learning_rate": 8.593333333333335e-06, "loss": 0.1888, "step": 1290 }, { "grad_norm": 2.5595085620880127, "learning_rate": 8.66e-06, "loss": 0.1791, "step": 1300 }, { "grad_norm": 2.209818124771118, "learning_rate": 8.726666666666667e-06, "loss": 0.1747, "step": 1310 }, { "grad_norm": 2.8619556427001953, "learning_rate": 8.793333333333334e-06, "loss": 0.1717, "step": 1320 }, { "grad_norm": 2.2111306190490723, "learning_rate": 8.86e-06, "loss": 0.1561, "step": 1330 }, { "grad_norm": 2.5533363819122314, "learning_rate": 8.926666666666667e-06, "loss": 0.1428, "step": 1340 }, { "grad_norm": 2.0768625736236572, "learning_rate": 8.993333333333334e-06, "loss": 0.1812, "step": 1350 }, { "grad_norm": 4.090967655181885, "learning_rate": 9.06e-06, "loss": 0.1757, "step": 1360 }, { "grad_norm": 3.408297061920166, "learning_rate": 9.126666666666667e-06, "loss": 0.1433, "step": 1370 }, { "grad_norm": 3.0656826496124268, "learning_rate": 9.193333333333334e-06, "loss": 0.1412, "step": 1380 }, { "grad_norm": 4.226680755615234, "learning_rate": 9.260000000000001e-06, "loss": 0.1604, "step": 1390 }, { "grad_norm": 3.056396245956421, "learning_rate": 9.326666666666667e-06, "loss": 0.1426, "step": 1400 }, { "grad_norm": 2.8261098861694336, "learning_rate": 9.393333333333334e-06, "loss": 0.1576, "step": 1410 }, { "grad_norm": 3.396232843399048, "learning_rate": 9.460000000000001e-06, "loss": 0.1594, "step": 1420 }, { "grad_norm": 2.2541089057922363, "learning_rate": 9.526666666666666e-06, "loss": 0.1337, "step": 1430 }, { "grad_norm": 2.5261178016662598, "learning_rate": 9.593333333333334e-06, "loss": 0.1541, "step": 1440 }, { "grad_norm": 2.5692732334136963, "learning_rate": 9.66e-06, "loss": 0.1722, "step": 1450 }, { "grad_norm": 2.1869313716888428, "learning_rate": 9.726666666666668e-06, "loss": 0.1653, "step": 1460 }, { "grad_norm": 2.136422634124756, "learning_rate": 9.793333333333333e-06, "loss": 0.1615, "step": 1470 }, { "grad_norm": 2.4462790489196777, "learning_rate": 9.86e-06, "loss": 0.1395, "step": 1480 }, { "grad_norm": 3.117281675338745, "learning_rate": 9.926666666666668e-06, "loss": 0.1438, "step": 1490 }, { "grad_norm": 3.1215744018554688, "learning_rate": 9.993333333333333e-06, "loss": 0.1444, "step": 1500 }, { "grad_norm": 1.8853398561477661, "learning_rate": 1.006e-05, "loss": 0.1315, "step": 1510 }, { "grad_norm": 2.884345531463623, "learning_rate": 1.0126666666666667e-05, "loss": 0.1431, "step": 1520 }, { "grad_norm": 1.9355385303497314, "learning_rate": 1.0193333333333335e-05, "loss": 0.1596, "step": 1530 }, { "grad_norm": 2.2313246726989746, "learning_rate": 1.026e-05, "loss": 0.141, "step": 1540 }, { "grad_norm": 2.293551445007324, "learning_rate": 1.0326666666666667e-05, "loss": 0.1207, "step": 1550 }, { "grad_norm": 2.4684650897979736, "learning_rate": 1.0393333333333334e-05, "loss": 0.1619, "step": 1560 }, { "grad_norm": 2.6854608058929443, "learning_rate": 1.046e-05, "loss": 0.1217, "step": 1570 }, { "grad_norm": 2.367647647857666, "learning_rate": 1.0526666666666667e-05, "loss": 0.1378, "step": 1580 }, { "grad_norm": 3.1451823711395264, "learning_rate": 1.0593333333333334e-05, "loss": 0.1432, "step": 1590 }, { "grad_norm": 1.7616304159164429, "learning_rate": 1.0660000000000001e-05, "loss": 0.1246, "step": 1600 }, { "grad_norm": 1.3941752910614014, "learning_rate": 1.0726666666666667e-05, "loss": 0.1175, "step": 1610 }, { "grad_norm": 2.164459705352783, "learning_rate": 1.0793333333333334e-05, "loss": 0.1299, "step": 1620 }, { "grad_norm": 2.3120970726013184, "learning_rate": 1.0860000000000001e-05, "loss": 0.1598, "step": 1630 }, { "grad_norm": 1.9093703031539917, "learning_rate": 1.0926666666666667e-05, "loss": 0.1289, "step": 1640 }, { "grad_norm": 2.3545708656311035, "learning_rate": 1.0993333333333334e-05, "loss": 0.1357, "step": 1650 }, { "grad_norm": 2.0453507900238037, "learning_rate": 1.106e-05, "loss": 0.1342, "step": 1660 }, { "grad_norm": 2.161041498184204, "learning_rate": 1.1126666666666668e-05, "loss": 0.1088, "step": 1670 }, { "grad_norm": 2.303750991821289, "learning_rate": 1.1193333333333333e-05, "loss": 0.1267, "step": 1680 }, { "grad_norm": 2.083505868911743, "learning_rate": 1.126e-05, "loss": 0.1194, "step": 1690 }, { "grad_norm": 2.121413469314575, "learning_rate": 1.1326666666666668e-05, "loss": 0.1298, "step": 1700 }, { "grad_norm": 2.6377360820770264, "learning_rate": 1.1393333333333333e-05, "loss": 0.1488, "step": 1710 }, { "grad_norm": 2.5817294120788574, "learning_rate": 1.146e-05, "loss": 0.1301, "step": 1720 }, { "grad_norm": 2.1694142818450928, "learning_rate": 1.1526666666666668e-05, "loss": 0.1261, "step": 1730 }, { "grad_norm": 3.027600049972534, "learning_rate": 1.1593333333333333e-05, "loss": 0.1183, "step": 1740 }, { "grad_norm": 3.41555118560791, "learning_rate": 1.166e-05, "loss": 0.138, "step": 1750 }, { "grad_norm": 2.1148972511291504, "learning_rate": 1.1726666666666667e-05, "loss": 0.1305, "step": 1760 }, { "grad_norm": 2.2071962356567383, "learning_rate": 1.1793333333333334e-05, "loss": 0.1112, "step": 1770 }, { "grad_norm": 2.0614988803863525, "learning_rate": 1.186e-05, "loss": 0.1483, "step": 1780 }, { "grad_norm": 2.966541051864624, "learning_rate": 1.1926666666666667e-05, "loss": 0.1145, "step": 1790 }, { "grad_norm": 2.35766863822937, "learning_rate": 1.1993333333333334e-05, "loss": 0.1208, "step": 1800 }, { "grad_norm": 2.153348445892334, "learning_rate": 1.206e-05, "loss": 0.1123, "step": 1810 }, { "grad_norm": 3.5533463954925537, "learning_rate": 1.2126666666666667e-05, "loss": 0.127, "step": 1820 }, { "grad_norm": 3.1159205436706543, "learning_rate": 1.2193333333333334e-05, "loss": 0.1157, "step": 1830 }, { "grad_norm": 2.3293263912200928, "learning_rate": 1.2260000000000001e-05, "loss": 0.1255, "step": 1840 }, { "grad_norm": 1.8122713565826416, "learning_rate": 1.2326666666666667e-05, "loss": 0.1389, "step": 1850 }, { "grad_norm": 2.332719326019287, "learning_rate": 1.2393333333333334e-05, "loss": 0.1413, "step": 1860 }, { "grad_norm": 2.1529085636138916, "learning_rate": 1.2460000000000001e-05, "loss": 0.1132, "step": 1870 }, { "grad_norm": 1.7745182514190674, "learning_rate": 1.2526666666666666e-05, "loss": 0.0987, "step": 1880 }, { "grad_norm": 2.134572982788086, "learning_rate": 1.2593333333333335e-05, "loss": 0.1172, "step": 1890 }, { "grad_norm": 2.1066529750823975, "learning_rate": 1.2659999999999999e-05, "loss": 0.1075, "step": 1900 }, { "grad_norm": 3.204768419265747, "learning_rate": 1.2726666666666668e-05, "loss": 0.1439, "step": 1910 }, { "grad_norm": 1.4598397016525269, "learning_rate": 1.2793333333333335e-05, "loss": 0.119, "step": 1920 }, { "grad_norm": 1.979782223701477, "learning_rate": 1.286e-05, "loss": 0.1048, "step": 1930 }, { "grad_norm": 2.744701385498047, "learning_rate": 1.2926666666666668e-05, "loss": 0.1195, "step": 1940 }, { "grad_norm": 1.637342929840088, "learning_rate": 1.2993333333333335e-05, "loss": 0.0992, "step": 1950 }, { "grad_norm": 3.2456157207489014, "learning_rate": 1.306e-05, "loss": 0.153, "step": 1960 }, { "grad_norm": 1.9266180992126465, "learning_rate": 1.3126666666666667e-05, "loss": 0.0958, "step": 1970 }, { "grad_norm": 1.8833301067352295, "learning_rate": 1.3193333333333335e-05, "loss": 0.1183, "step": 1980 }, { "grad_norm": 2.0892038345336914, "learning_rate": 1.326e-05, "loss": 0.0984, "step": 1990 }, { "grad_norm": 2.1357359886169434, "learning_rate": 1.3326666666666667e-05, "loss": 0.1133, "step": 2000 }, { "grad_norm": 2.050335645675659, "learning_rate": 1.3393333333333333e-05, "loss": 0.1369, "step": 2010 }, { "grad_norm": 2.5010764598846436, "learning_rate": 1.346e-05, "loss": 0.1118, "step": 2020 }, { "grad_norm": 2.3978700637817383, "learning_rate": 1.3526666666666669e-05, "loss": 0.0888, "step": 2030 }, { "grad_norm": 1.5326893329620361, "learning_rate": 1.3593333333333332e-05, "loss": 0.1226, "step": 2040 }, { "grad_norm": 2.3266408443450928, "learning_rate": 1.3660000000000001e-05, "loss": 0.1121, "step": 2050 }, { "grad_norm": 1.8370636701583862, "learning_rate": 1.3726666666666669e-05, "loss": 0.1019, "step": 2060 }, { "grad_norm": 1.9351521730422974, "learning_rate": 1.3793333333333332e-05, "loss": 0.092, "step": 2070 }, { "grad_norm": 2.2083017826080322, "learning_rate": 1.3860000000000001e-05, "loss": 0.1021, "step": 2080 }, { "grad_norm": 2.5389480590820312, "learning_rate": 1.3926666666666668e-05, "loss": 0.1096, "step": 2090 }, { "grad_norm": 2.3120834827423096, "learning_rate": 1.3993333333333334e-05, "loss": 0.1083, "step": 2100 }, { "grad_norm": 2.0972237586975098, "learning_rate": 1.4060000000000001e-05, "loss": 0.127, "step": 2110 }, { "grad_norm": 1.5987062454223633, "learning_rate": 1.4126666666666668e-05, "loss": 0.1008, "step": 2120 }, { "grad_norm": 1.95419442653656, "learning_rate": 1.4193333333333334e-05, "loss": 0.1016, "step": 2130 }, { "grad_norm": 1.958390235900879, "learning_rate": 1.426e-05, "loss": 0.0779, "step": 2140 }, { "grad_norm": 2.1421940326690674, "learning_rate": 1.4326666666666666e-05, "loss": 0.1141, "step": 2150 }, { "grad_norm": 1.3771287202835083, "learning_rate": 1.4393333333333333e-05, "loss": 0.1031, "step": 2160 }, { "grad_norm": 1.9050300121307373, "learning_rate": 1.4460000000000002e-05, "loss": 0.0996, "step": 2170 }, { "grad_norm": 1.7351733446121216, "learning_rate": 1.4526666666666666e-05, "loss": 0.1166, "step": 2180 }, { "grad_norm": 2.087116241455078, "learning_rate": 1.4593333333333333e-05, "loss": 0.0912, "step": 2190 }, { "grad_norm": 1.4750840663909912, "learning_rate": 1.4660000000000002e-05, "loss": 0.0908, "step": 2200 }, { "grad_norm": 1.7816190719604492, "learning_rate": 1.4726666666666666e-05, "loss": 0.0968, "step": 2210 }, { "grad_norm": 2.759226083755493, "learning_rate": 1.4793333333333335e-05, "loss": 0.0979, "step": 2220 }, { "grad_norm": 2.2915642261505127, "learning_rate": 1.4860000000000002e-05, "loss": 0.1015, "step": 2230 }, { "grad_norm": 1.0579414367675781, "learning_rate": 1.4926666666666667e-05, "loss": 0.0849, "step": 2240 }, { "grad_norm": 1.735168218612671, "learning_rate": 1.4993333333333334e-05, "loss": 0.1122, "step": 2250 }, { "grad_norm": 1.3647804260253906, "learning_rate": 1.5060000000000001e-05, "loss": 0.0837, "step": 2260 }, { "grad_norm": 1.8950369358062744, "learning_rate": 1.5126666666666667e-05, "loss": 0.0892, "step": 2270 }, { "grad_norm": 2.066502571105957, "learning_rate": 1.5193333333333334e-05, "loss": 0.0978, "step": 2280 }, { "grad_norm": 3.0743703842163086, "learning_rate": 1.5260000000000003e-05, "loss": 0.1278, "step": 2290 }, { "grad_norm": 2.0023443698883057, "learning_rate": 1.5326666666666667e-05, "loss": 0.0769, "step": 2300 }, { "grad_norm": 1.6678392887115479, "learning_rate": 1.5393333333333334e-05, "loss": 0.0922, "step": 2310 }, { "grad_norm": 1.2529624700546265, "learning_rate": 1.546e-05, "loss": 0.083, "step": 2320 }, { "grad_norm": 1.6986874341964722, "learning_rate": 1.5526666666666668e-05, "loss": 0.109, "step": 2330 }, { "grad_norm": 1.62709379196167, "learning_rate": 1.5593333333333335e-05, "loss": 0.093, "step": 2340 }, { "grad_norm": 1.5291154384613037, "learning_rate": 1.566e-05, "loss": 0.0966, "step": 2350 }, { "grad_norm": 1.7724781036376953, "learning_rate": 1.5726666666666666e-05, "loss": 0.1114, "step": 2360 }, { "grad_norm": 1.4953116178512573, "learning_rate": 1.5793333333333333e-05, "loss": 0.1092, "step": 2370 }, { "grad_norm": 2.2681994438171387, "learning_rate": 1.586e-05, "loss": 0.1101, "step": 2380 }, { "grad_norm": 1.803451657295227, "learning_rate": 1.5926666666666668e-05, "loss": 0.1127, "step": 2390 }, { "grad_norm": 1.5740747451782227, "learning_rate": 1.5993333333333335e-05, "loss": 0.0977, "step": 2400 }, { "grad_norm": 1.3898863792419434, "learning_rate": 1.606e-05, "loss": 0.0951, "step": 2410 }, { "grad_norm": 1.8877806663513184, "learning_rate": 1.612666666666667e-05, "loss": 0.1329, "step": 2420 }, { "grad_norm": 2.252192258834839, "learning_rate": 1.6193333333333336e-05, "loss": 0.1036, "step": 2430 }, { "grad_norm": 2.0669755935668945, "learning_rate": 1.626e-05, "loss": 0.1041, "step": 2440 }, { "grad_norm": 2.086519479751587, "learning_rate": 1.6326666666666667e-05, "loss": 0.0939, "step": 2450 }, { "grad_norm": 1.5307753086090088, "learning_rate": 1.639333333333333e-05, "loss": 0.0871, "step": 2460 }, { "grad_norm": 1.2660099267959595, "learning_rate": 1.646e-05, "loss": 0.0833, "step": 2470 }, { "grad_norm": 2.0640909671783447, "learning_rate": 1.652666666666667e-05, "loss": 0.1105, "step": 2480 }, { "grad_norm": 1.7546390295028687, "learning_rate": 1.6593333333333333e-05, "loss": 0.0908, "step": 2490 }, { "grad_norm": 1.4260809421539307, "learning_rate": 1.666e-05, "loss": 0.081, "step": 2500 }, { "grad_norm": 1.313043475151062, "learning_rate": 1.6726666666666667e-05, "loss": 0.0835, "step": 2510 }, { "grad_norm": 1.4116817712783813, "learning_rate": 1.6793333333333334e-05, "loss": 0.0892, "step": 2520 }, { "grad_norm": 1.6992125511169434, "learning_rate": 1.686e-05, "loss": 0.0832, "step": 2530 }, { "grad_norm": 5.339611053466797, "learning_rate": 1.692666666666667e-05, "loss": 0.0934, "step": 2540 }, { "grad_norm": 1.428652048110962, "learning_rate": 1.6993333333333332e-05, "loss": 0.0974, "step": 2550 }, { "grad_norm": 1.7005650997161865, "learning_rate": 1.706e-05, "loss": 0.0904, "step": 2560 }, { "grad_norm": 1.493583083152771, "learning_rate": 1.712666666666667e-05, "loss": 0.0897, "step": 2570 }, { "grad_norm": 1.7217246294021606, "learning_rate": 1.7193333333333334e-05, "loss": 0.091, "step": 2580 }, { "grad_norm": 2.1019246578216553, "learning_rate": 1.726e-05, "loss": 0.0847, "step": 2590 }, { "grad_norm": 1.247072458267212, "learning_rate": 1.7326666666666668e-05, "loss": 0.1041, "step": 2600 }, { "grad_norm": 1.6321816444396973, "learning_rate": 1.7393333333333335e-05, "loss": 0.0947, "step": 2610 }, { "grad_norm": 1.5834650993347168, "learning_rate": 1.7460000000000002e-05, "loss": 0.0791, "step": 2620 }, { "grad_norm": 1.6901391744613647, "learning_rate": 1.7526666666666666e-05, "loss": 0.0985, "step": 2630 }, { "grad_norm": 1.8310667276382446, "learning_rate": 1.7593333333333333e-05, "loss": 0.098, "step": 2640 }, { "grad_norm": 1.463097095489502, "learning_rate": 1.766e-05, "loss": 0.0964, "step": 2650 }, { "grad_norm": 1.5365809202194214, "learning_rate": 1.7726666666666667e-05, "loss": 0.0954, "step": 2660 }, { "grad_norm": 1.5144083499908447, "learning_rate": 1.7793333333333335e-05, "loss": 0.1057, "step": 2670 }, { "grad_norm": 2.173612594604492, "learning_rate": 1.7860000000000002e-05, "loss": 0.1109, "step": 2680 }, { "grad_norm": 1.6513830423355103, "learning_rate": 1.7926666666666666e-05, "loss": 0.0876, "step": 2690 }, { "grad_norm": 2.1395273208618164, "learning_rate": 1.7993333333333333e-05, "loss": 0.0844, "step": 2700 }, { "grad_norm": 1.5207836627960205, "learning_rate": 1.8060000000000003e-05, "loss": 0.0874, "step": 2710 }, { "grad_norm": 1.3758955001831055, "learning_rate": 1.8126666666666667e-05, "loss": 0.1151, "step": 2720 }, { "grad_norm": 1.4770007133483887, "learning_rate": 1.8193333333333334e-05, "loss": 0.0902, "step": 2730 }, { "grad_norm": 1.3950353860855103, "learning_rate": 1.826e-05, "loss": 0.0823, "step": 2740 }, { "grad_norm": 1.4987356662750244, "learning_rate": 1.832666666666667e-05, "loss": 0.0855, "step": 2750 }, { "grad_norm": 2.0468995571136475, "learning_rate": 1.8393333333333336e-05, "loss": 0.0828, "step": 2760 }, { "grad_norm": 1.4645957946777344, "learning_rate": 1.846e-05, "loss": 0.0894, "step": 2770 }, { "grad_norm": 1.3471146821975708, "learning_rate": 1.8526666666666667e-05, "loss": 0.0965, "step": 2780 }, { "grad_norm": 1.7715507745742798, "learning_rate": 1.8593333333333334e-05, "loss": 0.089, "step": 2790 }, { "grad_norm": 1.1111129522323608, "learning_rate": 1.866e-05, "loss": 0.0818, "step": 2800 }, { "grad_norm": 1.7058273553848267, "learning_rate": 1.8726666666666668e-05, "loss": 0.1076, "step": 2810 }, { "grad_norm": 2.76073956489563, "learning_rate": 1.8793333333333335e-05, "loss": 0.0923, "step": 2820 }, { "grad_norm": 1.426816701889038, "learning_rate": 1.886e-05, "loss": 0.0807, "step": 2830 }, { "grad_norm": 1.3335570096969604, "learning_rate": 1.8926666666666666e-05, "loss": 0.1035, "step": 2840 }, { "grad_norm": 1.346455693244934, "learning_rate": 1.8993333333333337e-05, "loss": 0.0922, "step": 2850 }, { "grad_norm": 1.5969599485397339, "learning_rate": 1.906e-05, "loss": 0.0838, "step": 2860 }, { "grad_norm": 0.9977166056632996, "learning_rate": 1.9126666666666668e-05, "loss": 0.0752, "step": 2870 }, { "grad_norm": 1.9408197402954102, "learning_rate": 1.9193333333333335e-05, "loss": 0.0743, "step": 2880 }, { "grad_norm": 1.5163705348968506, "learning_rate": 1.9260000000000002e-05, "loss": 0.0887, "step": 2890 }, { "grad_norm": 1.318821668624878, "learning_rate": 1.932666666666667e-05, "loss": 0.0914, "step": 2900 }, { "grad_norm": 2.3384649753570557, "learning_rate": 1.9393333333333336e-05, "loss": 0.0816, "step": 2910 }, { "grad_norm": 1.0777643918991089, "learning_rate": 1.946e-05, "loss": 0.0913, "step": 2920 }, { "grad_norm": 1.6602554321289062, "learning_rate": 1.9526666666666667e-05, "loss": 0.0742, "step": 2930 }, { "grad_norm": 1.0434300899505615, "learning_rate": 1.9593333333333334e-05, "loss": 0.0819, "step": 2940 }, { "grad_norm": 1.6225591897964478, "learning_rate": 1.966e-05, "loss": 0.0798, "step": 2950 }, { "grad_norm": 1.3711835145950317, "learning_rate": 1.972666666666667e-05, "loss": 0.0984, "step": 2960 }, { "grad_norm": 1.5857431888580322, "learning_rate": 1.9793333333333332e-05, "loss": 0.0767, "step": 2970 }, { "grad_norm": 1.3047014474868774, "learning_rate": 1.986e-05, "loss": 0.08, "step": 2980 }, { "grad_norm": 1.2176131010055542, "learning_rate": 1.992666666666667e-05, "loss": 0.0904, "step": 2990 }, { "grad_norm": 0.995647668838501, "learning_rate": 1.9993333333333334e-05, "loss": 0.0918, "step": 3000 }, { "grad_norm": 1.4046549797058105, "learning_rate": 2.006e-05, "loss": 0.0663, "step": 3010 }, { "grad_norm": 1.6117972135543823, "learning_rate": 2.0126666666666668e-05, "loss": 0.0873, "step": 3020 }, { "grad_norm": 1.2952065467834473, "learning_rate": 2.0193333333333332e-05, "loss": 0.0754, "step": 3030 }, { "grad_norm": 1.121705412864685, "learning_rate": 2.0260000000000003e-05, "loss": 0.0884, "step": 3040 }, { "grad_norm": 1.4540568590164185, "learning_rate": 2.032666666666667e-05, "loss": 0.076, "step": 3050 }, { "grad_norm": 1.1719189882278442, "learning_rate": 2.0393333333333333e-05, "loss": 0.083, "step": 3060 }, { "grad_norm": 0.9430192112922668, "learning_rate": 2.046e-05, "loss": 0.0647, "step": 3070 }, { "grad_norm": 1.098494291305542, "learning_rate": 2.0526666666666668e-05, "loss": 0.082, "step": 3080 }, { "grad_norm": 2.3549997806549072, "learning_rate": 2.0593333333333335e-05, "loss": 0.1008, "step": 3090 }, { "grad_norm": 2.056781530380249, "learning_rate": 2.0660000000000002e-05, "loss": 0.08, "step": 3100 }, { "grad_norm": 1.0601413249969482, "learning_rate": 2.0726666666666666e-05, "loss": 0.0717, "step": 3110 }, { "grad_norm": 0.8346056342124939, "learning_rate": 2.0793333333333333e-05, "loss": 0.0731, "step": 3120 }, { "grad_norm": 1.4170947074890137, "learning_rate": 2.086e-05, "loss": 0.0806, "step": 3130 }, { "grad_norm": 1.2079780101776123, "learning_rate": 2.0926666666666667e-05, "loss": 0.0792, "step": 3140 }, { "grad_norm": 1.3463503122329712, "learning_rate": 2.0993333333333334e-05, "loss": 0.081, "step": 3150 }, { "grad_norm": 1.2370054721832275, "learning_rate": 2.106e-05, "loss": 0.0847, "step": 3160 }, { "grad_norm": 1.068710207939148, "learning_rate": 2.1126666666666665e-05, "loss": 0.0786, "step": 3170 }, { "grad_norm": 1.1049692630767822, "learning_rate": 2.1193333333333336e-05, "loss": 0.0877, "step": 3180 }, { "grad_norm": 1.4846868515014648, "learning_rate": 2.1260000000000003e-05, "loss": 0.0946, "step": 3190 }, { "grad_norm": 1.0460684299468994, "learning_rate": 2.1326666666666667e-05, "loss": 0.0987, "step": 3200 }, { "grad_norm": 1.054125428199768, "learning_rate": 2.1393333333333334e-05, "loss": 0.0686, "step": 3210 }, { "grad_norm": 1.5508205890655518, "learning_rate": 2.146e-05, "loss": 0.1014, "step": 3220 }, { "grad_norm": 1.16800856590271, "learning_rate": 2.152666666666667e-05, "loss": 0.097, "step": 3230 }, { "grad_norm": 0.9802716970443726, "learning_rate": 2.1593333333333336e-05, "loss": 0.0813, "step": 3240 }, { "grad_norm": 1.031800627708435, "learning_rate": 2.166e-05, "loss": 0.0913, "step": 3250 }, { "grad_norm": 1.5123769044876099, "learning_rate": 2.1726666666666666e-05, "loss": 0.0817, "step": 3260 }, { "grad_norm": 1.445824146270752, "learning_rate": 2.1793333333333334e-05, "loss": 0.0792, "step": 3270 }, { "grad_norm": 1.5827884674072266, "learning_rate": 2.186e-05, "loss": 0.0667, "step": 3280 }, { "grad_norm": 1.1697243452072144, "learning_rate": 2.1926666666666668e-05, "loss": 0.0691, "step": 3290 }, { "grad_norm": 1.4026011228561401, "learning_rate": 2.1993333333333335e-05, "loss": 0.0713, "step": 3300 }, { "grad_norm": 1.5594357252120972, "learning_rate": 2.206e-05, "loss": 0.0937, "step": 3310 }, { "grad_norm": 0.8742972612380981, "learning_rate": 2.212666666666667e-05, "loss": 0.0966, "step": 3320 }, { "grad_norm": 1.1406413316726685, "learning_rate": 2.2193333333333337e-05, "loss": 0.0843, "step": 3330 }, { "grad_norm": 1.3411115407943726, "learning_rate": 2.226e-05, "loss": 0.0783, "step": 3340 }, { "grad_norm": 1.181656837463379, "learning_rate": 2.2326666666666667e-05, "loss": 0.0756, "step": 3350 }, { "grad_norm": 2.145029306411743, "learning_rate": 2.2393333333333335e-05, "loss": 0.0772, "step": 3360 }, { "grad_norm": 0.8566174507141113, "learning_rate": 2.2460000000000002e-05, "loss": 0.0777, "step": 3370 }, { "grad_norm": 1.0375508069992065, "learning_rate": 2.252666666666667e-05, "loss": 0.1199, "step": 3380 }, { "grad_norm": 1.1641069650650024, "learning_rate": 2.2593333333333336e-05, "loss": 0.0704, "step": 3390 }, { "grad_norm": 1.2732785940170288, "learning_rate": 2.266e-05, "loss": 0.0856, "step": 3400 }, { "grad_norm": 1.2275177240371704, "learning_rate": 2.2726666666666667e-05, "loss": 0.0721, "step": 3410 }, { "grad_norm": 1.1970046758651733, "learning_rate": 2.2793333333333334e-05, "loss": 0.0736, "step": 3420 }, { "grad_norm": 1.1959048509597778, "learning_rate": 2.286e-05, "loss": 0.0769, "step": 3430 }, { "grad_norm": 1.1701743602752686, "learning_rate": 2.292666666666667e-05, "loss": 0.0793, "step": 3440 }, { "grad_norm": 1.5616178512573242, "learning_rate": 2.2993333333333332e-05, "loss": 0.0815, "step": 3450 }, { "grad_norm": 1.0067352056503296, "learning_rate": 2.306e-05, "loss": 0.0667, "step": 3460 }, { "grad_norm": 1.1678297519683838, "learning_rate": 2.312666666666667e-05, "loss": 0.0801, "step": 3470 }, { "grad_norm": 1.147058129310608, "learning_rate": 2.3193333333333334e-05, "loss": 0.0856, "step": 3480 }, { "grad_norm": 1.0712974071502686, "learning_rate": 2.326e-05, "loss": 0.0771, "step": 3490 }, { "grad_norm": 1.069478988647461, "learning_rate": 2.3326666666666668e-05, "loss": 0.0621, "step": 3500 }, { "grad_norm": 1.3175064325332642, "learning_rate": 2.3393333333333335e-05, "loss": 0.0813, "step": 3510 }, { "grad_norm": 1.038460612297058, "learning_rate": 2.3460000000000002e-05, "loss": 0.0781, "step": 3520 }, { "grad_norm": 1.2416340112686157, "learning_rate": 2.352666666666667e-05, "loss": 0.0802, "step": 3530 }, { "grad_norm": 1.2792048454284668, "learning_rate": 2.3593333333333333e-05, "loss": 0.0778, "step": 3540 }, { "grad_norm": 0.861871063709259, "learning_rate": 2.366e-05, "loss": 0.0829, "step": 3550 }, { "grad_norm": 1.053450107574463, "learning_rate": 2.3726666666666668e-05, "loss": 0.0639, "step": 3560 }, { "grad_norm": 1.141169548034668, "learning_rate": 2.3793333333333335e-05, "loss": 0.0821, "step": 3570 }, { "grad_norm": 0.9677315354347229, "learning_rate": 2.3860000000000002e-05, "loss": 0.0903, "step": 3580 }, { "grad_norm": 1.1114718914031982, "learning_rate": 2.3926666666666666e-05, "loss": 0.0828, "step": 3590 }, { "grad_norm": 1.2391595840454102, "learning_rate": 2.3993333333333333e-05, "loss": 0.0814, "step": 3600 }, { "grad_norm": 0.9612681269645691, "learning_rate": 2.4060000000000003e-05, "loss": 0.0719, "step": 3610 }, { "grad_norm": 1.1002568006515503, "learning_rate": 2.4126666666666667e-05, "loss": 0.0678, "step": 3620 }, { "grad_norm": 0.9797812700271606, "learning_rate": 2.4193333333333334e-05, "loss": 0.0623, "step": 3630 }, { "grad_norm": 0.9848971366882324, "learning_rate": 2.426e-05, "loss": 0.0697, "step": 3640 }, { "grad_norm": 1.1300104856491089, "learning_rate": 2.432666666666667e-05, "loss": 0.0816, "step": 3650 }, { "grad_norm": 0.922448456287384, "learning_rate": 2.4393333333333336e-05, "loss": 0.0705, "step": 3660 }, { "grad_norm": 1.5011284351348877, "learning_rate": 2.4460000000000003e-05, "loss": 0.0813, "step": 3670 }, { "grad_norm": 1.284838080406189, "learning_rate": 2.4526666666666667e-05, "loss": 0.0745, "step": 3680 }, { "grad_norm": 1.1874488592147827, "learning_rate": 2.4593333333333334e-05, "loss": 0.0782, "step": 3690 }, { "grad_norm": 0.9749250411987305, "learning_rate": 2.466e-05, "loss": 0.0725, "step": 3700 }, { "grad_norm": 1.5955512523651123, "learning_rate": 2.4726666666666668e-05, "loss": 0.0754, "step": 3710 }, { "grad_norm": 0.9843117594718933, "learning_rate": 2.4793333333333335e-05, "loss": 0.0752, "step": 3720 }, { "grad_norm": 0.9134371280670166, "learning_rate": 2.486e-05, "loss": 0.07, "step": 3730 }, { "grad_norm": 1.1788612604141235, "learning_rate": 2.4926666666666666e-05, "loss": 0.0769, "step": 3740 }, { "grad_norm": 0.9936628341674805, "learning_rate": 2.4993333333333337e-05, "loss": 0.0794, "step": 3750 }, { "grad_norm": 0.994447648525238, "learning_rate": 2.506e-05, "loss": 0.068, "step": 3760 }, { "grad_norm": 1.209518313407898, "learning_rate": 2.512666666666667e-05, "loss": 0.079, "step": 3770 }, { "grad_norm": 1.1208652257919312, "learning_rate": 2.5193333333333335e-05, "loss": 0.0866, "step": 3780 }, { "grad_norm": 0.8419187068939209, "learning_rate": 2.526e-05, "loss": 0.0684, "step": 3790 }, { "grad_norm": 1.0576066970825195, "learning_rate": 2.5326666666666666e-05, "loss": 0.1013, "step": 3800 }, { "grad_norm": 0.8599345684051514, "learning_rate": 2.5393333333333336e-05, "loss": 0.0666, "step": 3810 }, { "grad_norm": 1.3936220407485962, "learning_rate": 2.546e-05, "loss": 0.0872, "step": 3820 }, { "grad_norm": 0.9354259371757507, "learning_rate": 2.5526666666666664e-05, "loss": 0.0658, "step": 3830 }, { "grad_norm": 0.8971160650253296, "learning_rate": 2.5593333333333334e-05, "loss": 0.0641, "step": 3840 }, { "grad_norm": 1.39225172996521, "learning_rate": 2.566e-05, "loss": 0.0818, "step": 3850 }, { "grad_norm": 1.2561806440353394, "learning_rate": 2.5726666666666665e-05, "loss": 0.0703, "step": 3860 }, { "grad_norm": 0.9452486634254456, "learning_rate": 2.5793333333333336e-05, "loss": 0.075, "step": 3870 }, { "grad_norm": 0.9170100092887878, "learning_rate": 2.586e-05, "loss": 0.0663, "step": 3880 }, { "grad_norm": 1.1342521905899048, "learning_rate": 2.5926666666666667e-05, "loss": 0.0726, "step": 3890 }, { "grad_norm": 0.9274302124977112, "learning_rate": 2.5993333333333337e-05, "loss": 0.0647, "step": 3900 }, { "grad_norm": 0.8706011176109314, "learning_rate": 2.606e-05, "loss": 0.0728, "step": 3910 }, { "grad_norm": 1.8271509408950806, "learning_rate": 2.6126666666666665e-05, "loss": 0.0731, "step": 3920 }, { "grad_norm": 0.7823825478553772, "learning_rate": 2.6193333333333336e-05, "loss": 0.0678, "step": 3930 }, { "grad_norm": 0.6146025657653809, "learning_rate": 2.6260000000000003e-05, "loss": 0.0847, "step": 3940 }, { "grad_norm": 0.762973964214325, "learning_rate": 2.6326666666666666e-05, "loss": 0.0767, "step": 3950 }, { "grad_norm": 1.1631197929382324, "learning_rate": 2.6393333333333337e-05, "loss": 0.0776, "step": 3960 }, { "grad_norm": 0.9143610000610352, "learning_rate": 2.646e-05, "loss": 0.0655, "step": 3970 }, { "grad_norm": 0.9174964427947998, "learning_rate": 2.6526666666666668e-05, "loss": 0.0773, "step": 3980 }, { "grad_norm": 1.5294989347457886, "learning_rate": 2.6593333333333335e-05, "loss": 0.0759, "step": 3990 }, { "grad_norm": 1.6611175537109375, "learning_rate": 2.6660000000000002e-05, "loss": 0.0674, "step": 4000 }, { "grad_norm": 1.1287119388580322, "learning_rate": 2.6726666666666666e-05, "loss": 0.0738, "step": 4010 }, { "grad_norm": 0.9787588715553284, "learning_rate": 2.6793333333333337e-05, "loss": 0.0753, "step": 4020 }, { "grad_norm": 0.9541427493095398, "learning_rate": 2.686e-05, "loss": 0.0747, "step": 4030 }, { "grad_norm": 0.9921916127204895, "learning_rate": 2.6926666666666667e-05, "loss": 0.0558, "step": 4040 }, { "grad_norm": 1.7732882499694824, "learning_rate": 2.6993333333333338e-05, "loss": 0.0712, "step": 4050 }, { "grad_norm": 1.0518770217895508, "learning_rate": 2.7060000000000002e-05, "loss": 0.0673, "step": 4060 }, { "grad_norm": 1.112221598625183, "learning_rate": 2.7126666666666666e-05, "loss": 0.062, "step": 4070 }, { "grad_norm": 1.0156141519546509, "learning_rate": 2.7193333333333336e-05, "loss": 0.0739, "step": 4080 }, { "grad_norm": 0.996197521686554, "learning_rate": 2.7260000000000003e-05, "loss": 0.0739, "step": 4090 }, { "grad_norm": 1.0782549381256104, "learning_rate": 2.7326666666666667e-05, "loss": 0.0649, "step": 4100 }, { "grad_norm": 0.9606845378875732, "learning_rate": 2.739333333333333e-05, "loss": 0.075, "step": 4110 }, { "grad_norm": 2.230797052383423, "learning_rate": 2.746e-05, "loss": 0.0828, "step": 4120 }, { "grad_norm": 0.8732755780220032, "learning_rate": 2.752666666666667e-05, "loss": 0.0639, "step": 4130 }, { "grad_norm": 0.6810647249221802, "learning_rate": 2.7593333333333332e-05, "loss": 0.0686, "step": 4140 }, { "grad_norm": 1.2619259357452393, "learning_rate": 2.7660000000000003e-05, "loss": 0.0776, "step": 4150 }, { "grad_norm": 0.7582120895385742, "learning_rate": 2.7726666666666667e-05, "loss": 0.0678, "step": 4160 }, { "grad_norm": 0.7058753371238708, "learning_rate": 2.7793333333333334e-05, "loss": 0.0649, "step": 4170 }, { "grad_norm": 0.9103642106056213, "learning_rate": 2.7860000000000004e-05, "loss": 0.059, "step": 4180 }, { "grad_norm": 0.8977436423301697, "learning_rate": 2.7926666666666668e-05, "loss": 0.0659, "step": 4190 }, { "grad_norm": 0.885173499584198, "learning_rate": 2.7993333333333332e-05, "loss": 0.0658, "step": 4200 }, { "grad_norm": 0.7441266775131226, "learning_rate": 2.8060000000000002e-05, "loss": 0.0727, "step": 4210 }, { "grad_norm": 0.8663538694381714, "learning_rate": 2.8126666666666666e-05, "loss": 0.0841, "step": 4220 }, { "grad_norm": 1.183104395866394, "learning_rate": 2.8193333333333333e-05, "loss": 0.0844, "step": 4230 }, { "grad_norm": 1.2089931964874268, "learning_rate": 2.8260000000000004e-05, "loss": 0.0816, "step": 4240 }, { "grad_norm": 0.799780011177063, "learning_rate": 2.8326666666666668e-05, "loss": 0.0744, "step": 4250 }, { "grad_norm": 0.8159696459770203, "learning_rate": 2.839333333333333e-05, "loss": 0.0718, "step": 4260 }, { "grad_norm": 0.7200425863265991, "learning_rate": 2.8460000000000002e-05, "loss": 0.0773, "step": 4270 }, { "grad_norm": 1.0989195108413696, "learning_rate": 2.852666666666667e-05, "loss": 0.0645, "step": 4280 }, { "grad_norm": 0.8990192413330078, "learning_rate": 2.8593333333333333e-05, "loss": 0.0629, "step": 4290 }, { "grad_norm": 1.078208088874817, "learning_rate": 2.8660000000000003e-05, "loss": 0.0608, "step": 4300 }, { "grad_norm": 0.782849907875061, "learning_rate": 2.8726666666666667e-05, "loss": 0.0646, "step": 4310 }, { "grad_norm": 1.0957661867141724, "learning_rate": 2.8793333333333334e-05, "loss": 0.0902, "step": 4320 }, { "grad_norm": 0.9034069180488586, "learning_rate": 2.8860000000000005e-05, "loss": 0.0627, "step": 4330 }, { "grad_norm": 1.0107156038284302, "learning_rate": 2.892666666666667e-05, "loss": 0.0593, "step": 4340 }, { "grad_norm": 0.6198423504829407, "learning_rate": 2.8993333333333332e-05, "loss": 0.0608, "step": 4350 }, { "grad_norm": 0.8111169934272766, "learning_rate": 2.9060000000000003e-05, "loss": 0.0591, "step": 4360 }, { "grad_norm": 0.7333782911300659, "learning_rate": 2.912666666666667e-05, "loss": 0.0674, "step": 4370 }, { "grad_norm": 0.7249762415885925, "learning_rate": 2.9193333333333334e-05, "loss": 0.0664, "step": 4380 }, { "grad_norm": 0.7170128226280212, "learning_rate": 2.9260000000000004e-05, "loss": 0.0532, "step": 4390 }, { "grad_norm": 0.7519227266311646, "learning_rate": 2.9326666666666668e-05, "loss": 0.076, "step": 4400 }, { "grad_norm": 0.8616819977760315, "learning_rate": 2.9393333333333335e-05, "loss": 0.0726, "step": 4410 }, { "grad_norm": 0.9920185804367065, "learning_rate": 2.946e-05, "loss": 0.082, "step": 4420 }, { "grad_norm": 0.84588223695755, "learning_rate": 2.952666666666667e-05, "loss": 0.0733, "step": 4430 }, { "grad_norm": 1.6959919929504395, "learning_rate": 2.9593333333333333e-05, "loss": 0.0729, "step": 4440 }, { "grad_norm": 0.8465054631233215, "learning_rate": 2.9659999999999997e-05, "loss": 0.0897, "step": 4450 }, { "grad_norm": 0.9899027347564697, "learning_rate": 2.9726666666666668e-05, "loss": 0.0706, "step": 4460 }, { "grad_norm": 0.6400759816169739, "learning_rate": 2.9793333333333335e-05, "loss": 0.067, "step": 4470 }, { "grad_norm": 1.1446712017059326, "learning_rate": 2.986e-05, "loss": 0.0756, "step": 4480 }, { "grad_norm": 0.9955088496208191, "learning_rate": 2.992666666666667e-05, "loss": 0.0697, "step": 4490 }, { "grad_norm": 0.8019505143165588, "learning_rate": 2.9993333333333333e-05, "loss": 0.065, "step": 4500 }, { "grad_norm": 0.6946886777877808, "learning_rate": 3.006e-05, "loss": 0.093, "step": 4510 }, { "grad_norm": 0.7265571355819702, "learning_rate": 3.012666666666667e-05, "loss": 0.0742, "step": 4520 }, { "grad_norm": 0.7359143495559692, "learning_rate": 3.0193333333333335e-05, "loss": 0.064, "step": 4530 }, { "grad_norm": 1.4826843738555908, "learning_rate": 3.0259999999999998e-05, "loss": 0.078, "step": 4540 }, { "grad_norm": 0.7673079371452332, "learning_rate": 3.032666666666667e-05, "loss": 0.0808, "step": 4550 }, { "grad_norm": 0.7406584620475769, "learning_rate": 3.0393333333333336e-05, "loss": 0.0719, "step": 4560 }, { "grad_norm": 0.7279905676841736, "learning_rate": 3.046e-05, "loss": 0.0768, "step": 4570 }, { "grad_norm": 1.5876293182373047, "learning_rate": 3.052666666666667e-05, "loss": 0.077, "step": 4580 }, { "grad_norm": 1.0114260911941528, "learning_rate": 3.059333333333334e-05, "loss": 0.0665, "step": 4590 }, { "grad_norm": 0.9401320815086365, "learning_rate": 3.066e-05, "loss": 0.0672, "step": 4600 }, { "grad_norm": 0.9447844624519348, "learning_rate": 3.072666666666667e-05, "loss": 0.0647, "step": 4610 }, { "grad_norm": 0.8512403964996338, "learning_rate": 3.0793333333333336e-05, "loss": 0.0562, "step": 4620 }, { "grad_norm": 1.314315915107727, "learning_rate": 3.086e-05, "loss": 0.08, "step": 4630 }, { "grad_norm": 0.702836275100708, "learning_rate": 3.092666666666667e-05, "loss": 0.0709, "step": 4640 }, { "grad_norm": 0.7094389796257019, "learning_rate": 3.0993333333333334e-05, "loss": 0.0736, "step": 4650 }, { "grad_norm": 1.0111792087554932, "learning_rate": 3.106e-05, "loss": 0.072, "step": 4660 }, { "grad_norm": 1.1978890895843506, "learning_rate": 3.112666666666667e-05, "loss": 0.0689, "step": 4670 }, { "grad_norm": 0.7638159394264221, "learning_rate": 3.119333333333334e-05, "loss": 0.0666, "step": 4680 }, { "grad_norm": 0.8993260264396667, "learning_rate": 3.126e-05, "loss": 0.086, "step": 4690 }, { "grad_norm": 0.6945252418518066, "learning_rate": 3.132666666666667e-05, "loss": 0.0673, "step": 4700 }, { "grad_norm": 0.8834990859031677, "learning_rate": 3.1393333333333337e-05, "loss": 0.0625, "step": 4710 }, { "grad_norm": 0.7421814799308777, "learning_rate": 3.146e-05, "loss": 0.0578, "step": 4720 }, { "grad_norm": 0.6132028102874756, "learning_rate": 3.1526666666666664e-05, "loss": 0.081, "step": 4730 }, { "grad_norm": 0.7701234221458435, "learning_rate": 3.1593333333333335e-05, "loss": 0.0716, "step": 4740 }, { "grad_norm": 0.6736772060394287, "learning_rate": 3.166e-05, "loss": 0.0605, "step": 4750 }, { "grad_norm": 0.6660627126693726, "learning_rate": 3.172666666666667e-05, "loss": 0.0619, "step": 4760 }, { "grad_norm": 1.2960666418075562, "learning_rate": 3.179333333333333e-05, "loss": 0.0703, "step": 4770 }, { "grad_norm": 1.3012444972991943, "learning_rate": 3.186e-05, "loss": 0.0837, "step": 4780 }, { "grad_norm": 0.8731338381767273, "learning_rate": 3.192666666666667e-05, "loss": 0.0556, "step": 4790 }, { "grad_norm": 0.7405230402946472, "learning_rate": 3.199333333333334e-05, "loss": 0.0602, "step": 4800 }, { "grad_norm": 0.6373447775840759, "learning_rate": 3.206e-05, "loss": 0.0713, "step": 4810 }, { "grad_norm": 0.7714656591415405, "learning_rate": 3.2126666666666665e-05, "loss": 0.0693, "step": 4820 }, { "grad_norm": 0.9021379351615906, "learning_rate": 3.2193333333333336e-05, "loss": 0.0659, "step": 4830 }, { "grad_norm": 0.8354248404502869, "learning_rate": 3.226e-05, "loss": 0.0635, "step": 4840 }, { "grad_norm": 0.9813247919082642, "learning_rate": 3.232666666666666e-05, "loss": 0.067, "step": 4850 }, { "grad_norm": 0.7645657062530518, "learning_rate": 3.2393333333333334e-05, "loss": 0.0549, "step": 4860 }, { "grad_norm": 0.8574753403663635, "learning_rate": 3.2460000000000004e-05, "loss": 0.0647, "step": 4870 }, { "grad_norm": 0.6685387492179871, "learning_rate": 3.252666666666667e-05, "loss": 0.0586, "step": 4880 }, { "grad_norm": 2.051839590072632, "learning_rate": 3.259333333333334e-05, "loss": 0.0796, "step": 4890 }, { "grad_norm": 1.2040079832077026, "learning_rate": 3.266e-05, "loss": 0.0616, "step": 4900 }, { "grad_norm": 0.7284735441207886, "learning_rate": 3.2726666666666666e-05, "loss": 0.0752, "step": 4910 }, { "grad_norm": 0.6873114109039307, "learning_rate": 3.279333333333334e-05, "loss": 0.0694, "step": 4920 }, { "grad_norm": 1.011040449142456, "learning_rate": 3.286e-05, "loss": 0.0698, "step": 4930 }, { "grad_norm": 0.8211449980735779, "learning_rate": 3.2926666666666664e-05, "loss": 0.0662, "step": 4940 }, { "grad_norm": 0.6913883090019226, "learning_rate": 3.2993333333333335e-05, "loss": 0.0703, "step": 4950 }, { "grad_norm": 0.6074178218841553, "learning_rate": 3.3060000000000005e-05, "loss": 0.0588, "step": 4960 }, { "grad_norm": 0.6021671295166016, "learning_rate": 3.312666666666667e-05, "loss": 0.0709, "step": 4970 }, { "grad_norm": 0.7962651252746582, "learning_rate": 3.319333333333334e-05, "loss": 0.0649, "step": 4980 }, { "grad_norm": 0.6869488954544067, "learning_rate": 3.3260000000000003e-05, "loss": 0.0573, "step": 4990 }, { "grad_norm": 0.8237600922584534, "learning_rate": 3.332666666666667e-05, "loss": 0.0804, "step": 5000 }, { "grad_norm": 0.821208119392395, "learning_rate": 3.339333333333334e-05, "loss": 0.0659, "step": 5010 }, { "grad_norm": 0.9800066947937012, "learning_rate": 3.346e-05, "loss": 0.0586, "step": 5020 }, { "grad_norm": 0.8238874077796936, "learning_rate": 3.3526666666666665e-05, "loss": 0.057, "step": 5030 }, { "grad_norm": 0.7166539430618286, "learning_rate": 3.359333333333333e-05, "loss": 0.0698, "step": 5040 }, { "grad_norm": 0.8042401075363159, "learning_rate": 3.366e-05, "loss": 0.0695, "step": 5050 }, { "grad_norm": 0.5746594071388245, "learning_rate": 3.372666666666667e-05, "loss": 0.0488, "step": 5060 }, { "grad_norm": 0.763322114944458, "learning_rate": 3.3793333333333334e-05, "loss": 0.0801, "step": 5070 }, { "grad_norm": 0.5613510608673096, "learning_rate": 3.3860000000000004e-05, "loss": 0.0689, "step": 5080 }, { "grad_norm": 0.8710481524467468, "learning_rate": 3.392666666666667e-05, "loss": 0.0592, "step": 5090 }, { "grad_norm": 0.9161457419395447, "learning_rate": 3.399333333333333e-05, "loss": 0.0615, "step": 5100 }, { "grad_norm": 0.9116977453231812, "learning_rate": 3.406e-05, "loss": 0.0713, "step": 5110 }, { "grad_norm": 0.8748863935470581, "learning_rate": 3.4126666666666666e-05, "loss": 0.0798, "step": 5120 }, { "grad_norm": 0.9572709798812866, "learning_rate": 3.419333333333333e-05, "loss": 0.074, "step": 5130 }, { "grad_norm": 0.694638192653656, "learning_rate": 3.426e-05, "loss": 0.0635, "step": 5140 }, { "grad_norm": 0.7374732494354248, "learning_rate": 3.432666666666667e-05, "loss": 0.0553, "step": 5150 }, { "grad_norm": 0.6361985206604004, "learning_rate": 3.4393333333333335e-05, "loss": 0.0786, "step": 5160 }, { "grad_norm": 0.8946716785430908, "learning_rate": 3.4460000000000005e-05, "loss": 0.0661, "step": 5170 }, { "grad_norm": 0.8116955161094666, "learning_rate": 3.452666666666667e-05, "loss": 0.059, "step": 5180 }, { "grad_norm": 0.8239008784294128, "learning_rate": 3.459333333333333e-05, "loss": 0.087, "step": 5190 }, { "grad_norm": 0.5448088049888611, "learning_rate": 3.4660000000000004e-05, "loss": 0.0591, "step": 5200 }, { "grad_norm": 0.7242969870567322, "learning_rate": 3.472666666666667e-05, "loss": 0.0634, "step": 5210 }, { "grad_norm": 0.6934666037559509, "learning_rate": 3.479333333333333e-05, "loss": 0.0576, "step": 5220 }, { "grad_norm": 0.6051027774810791, "learning_rate": 3.486e-05, "loss": 0.0629, "step": 5230 }, { "grad_norm": 0.5428176522254944, "learning_rate": 3.4926666666666665e-05, "loss": 0.0562, "step": 5240 }, { "grad_norm": 0.7201752066612244, "learning_rate": 3.4993333333333336e-05, "loss": 0.0787, "step": 5250 }, { "grad_norm": 1.1238831281661987, "learning_rate": 3.5060000000000007e-05, "loss": 0.063, "step": 5260 }, { "grad_norm": 0.8912506103515625, "learning_rate": 3.512666666666667e-05, "loss": 0.0662, "step": 5270 }, { "grad_norm": 0.7165629863739014, "learning_rate": 3.5193333333333334e-05, "loss": 0.0669, "step": 5280 }, { "grad_norm": 0.6676174402236938, "learning_rate": 3.5260000000000005e-05, "loss": 0.0654, "step": 5290 }, { "grad_norm": 0.6303908824920654, "learning_rate": 3.532666666666667e-05, "loss": 0.0574, "step": 5300 }, { "grad_norm": 0.7965080142021179, "learning_rate": 3.539333333333333e-05, "loss": 0.0574, "step": 5310 }, { "grad_norm": 0.8078598976135254, "learning_rate": 3.546e-05, "loss": 0.0683, "step": 5320 }, { "grad_norm": 1.207365870475769, "learning_rate": 3.5526666666666666e-05, "loss": 0.074, "step": 5330 }, { "grad_norm": 0.638649046421051, "learning_rate": 3.559333333333334e-05, "loss": 0.0664, "step": 5340 }, { "grad_norm": 0.5407071113586426, "learning_rate": 3.566e-05, "loss": 0.0606, "step": 5350 }, { "grad_norm": 0.5846586227416992, "learning_rate": 3.572666666666667e-05, "loss": 0.0627, "step": 5360 }, { "grad_norm": 0.6921466588973999, "learning_rate": 3.5793333333333335e-05, "loss": 0.0586, "step": 5370 }, { "grad_norm": 0.7357812523841858, "learning_rate": 3.586e-05, "loss": 0.0647, "step": 5380 }, { "grad_norm": 0.7356671690940857, "learning_rate": 3.592666666666667e-05, "loss": 0.0599, "step": 5390 }, { "grad_norm": 0.5571765303611755, "learning_rate": 3.599333333333333e-05, "loss": 0.0612, "step": 5400 }, { "grad_norm": 0.9996488690376282, "learning_rate": 3.606e-05, "loss": 0.0572, "step": 5410 }, { "grad_norm": 0.7226800322532654, "learning_rate": 3.612666666666667e-05, "loss": 0.0587, "step": 5420 }, { "grad_norm": 1.1028459072113037, "learning_rate": 3.619333333333333e-05, "loss": 0.0547, "step": 5430 }, { "grad_norm": 1.4175069332122803, "learning_rate": 3.626e-05, "loss": 0.079, "step": 5440 }, { "grad_norm": 0.5960317254066467, "learning_rate": 3.632666666666667e-05, "loss": 0.0525, "step": 5450 }, { "grad_norm": 0.8025105595588684, "learning_rate": 3.6393333333333336e-05, "loss": 0.0537, "step": 5460 }, { "grad_norm": 0.7392962574958801, "learning_rate": 3.646e-05, "loss": 0.0531, "step": 5470 }, { "grad_norm": 0.5740261077880859, "learning_rate": 3.652666666666667e-05, "loss": 0.0573, "step": 5480 }, { "grad_norm": 0.6097183227539062, "learning_rate": 3.6593333333333334e-05, "loss": 0.0583, "step": 5490 }, { "grad_norm": 0.5852087736129761, "learning_rate": 3.666e-05, "loss": 0.0579, "step": 5500 }, { "grad_norm": 0.646651029586792, "learning_rate": 3.672666666666667e-05, "loss": 0.0695, "step": 5510 }, { "grad_norm": 0.6050301790237427, "learning_rate": 3.679333333333333e-05, "loss": 0.0552, "step": 5520 }, { "grad_norm": 0.6122351288795471, "learning_rate": 3.686e-05, "loss": 0.0598, "step": 5530 }, { "grad_norm": 0.678108811378479, "learning_rate": 3.6926666666666673e-05, "loss": 0.0544, "step": 5540 }, { "grad_norm": 0.5990300178527832, "learning_rate": 3.699333333333334e-05, "loss": 0.0638, "step": 5550 }, { "grad_norm": 0.5679892897605896, "learning_rate": 3.706e-05, "loss": 0.0548, "step": 5560 }, { "grad_norm": 0.8495585322380066, "learning_rate": 3.712666666666667e-05, "loss": 0.0563, "step": 5570 }, { "grad_norm": 1.0542755126953125, "learning_rate": 3.7193333333333335e-05, "loss": 0.0723, "step": 5580 }, { "grad_norm": 0.7522136569023132, "learning_rate": 3.726e-05, "loss": 0.0556, "step": 5590 }, { "grad_norm": 0.490943968296051, "learning_rate": 3.732666666666667e-05, "loss": 0.0631, "step": 5600 }, { "grad_norm": 0.6268877983093262, "learning_rate": 3.739333333333333e-05, "loss": 0.0585, "step": 5610 }, { "grad_norm": 0.5763557553291321, "learning_rate": 3.7460000000000004e-05, "loss": 0.0545, "step": 5620 }, { "grad_norm": 0.684360682964325, "learning_rate": 3.7526666666666674e-05, "loss": 0.0543, "step": 5630 }, { "grad_norm": 0.7144933938980103, "learning_rate": 3.759333333333334e-05, "loss": 0.0549, "step": 5640 }, { "grad_norm": 0.551734209060669, "learning_rate": 3.766e-05, "loss": 0.0756, "step": 5650 }, { "grad_norm": 0.637194037437439, "learning_rate": 3.7726666666666666e-05, "loss": 0.053, "step": 5660 }, { "grad_norm": 0.8360990881919861, "learning_rate": 3.7793333333333336e-05, "loss": 0.0614, "step": 5670 }, { "grad_norm": 0.7797257900238037, "learning_rate": 3.786e-05, "loss": 0.0625, "step": 5680 }, { "grad_norm": 1.0447405576705933, "learning_rate": 3.7926666666666664e-05, "loss": 0.0657, "step": 5690 }, { "grad_norm": 0.4618528187274933, "learning_rate": 3.7993333333333334e-05, "loss": 0.0621, "step": 5700 }, { "grad_norm": 0.5710528492927551, "learning_rate": 3.806e-05, "loss": 0.0631, "step": 5710 }, { "grad_norm": 0.7105844616889954, "learning_rate": 3.812666666666667e-05, "loss": 0.0567, "step": 5720 }, { "grad_norm": 0.6096794605255127, "learning_rate": 3.819333333333334e-05, "loss": 0.0578, "step": 5730 }, { "grad_norm": 0.47971311211586, "learning_rate": 3.826e-05, "loss": 0.0614, "step": 5740 }, { "grad_norm": 0.5534102320671082, "learning_rate": 3.832666666666667e-05, "loss": 0.0502, "step": 5750 }, { "grad_norm": 0.6123351454734802, "learning_rate": 3.839333333333334e-05, "loss": 0.0538, "step": 5760 }, { "grad_norm": 0.8060342073440552, "learning_rate": 3.846e-05, "loss": 0.0613, "step": 5770 }, { "grad_norm": 0.7499107122421265, "learning_rate": 3.8526666666666665e-05, "loss": 0.0663, "step": 5780 }, { "grad_norm": 0.8221561312675476, "learning_rate": 3.8593333333333335e-05, "loss": 0.0691, "step": 5790 }, { "grad_norm": 0.8706967830657959, "learning_rate": 3.866e-05, "loss": 0.0681, "step": 5800 }, { "grad_norm": 0.5787605047225952, "learning_rate": 3.872666666666667e-05, "loss": 0.0652, "step": 5810 }, { "grad_norm": 0.6189667582511902, "learning_rate": 3.879333333333334e-05, "loss": 0.0586, "step": 5820 }, { "grad_norm": 0.6547760963439941, "learning_rate": 3.8860000000000004e-05, "loss": 0.0604, "step": 5830 }, { "grad_norm": 0.8347406983375549, "learning_rate": 3.892666666666667e-05, "loss": 0.0593, "step": 5840 }, { "grad_norm": 0.8371382355690002, "learning_rate": 3.899333333333334e-05, "loss": 0.0578, "step": 5850 }, { "grad_norm": 0.669310450553894, "learning_rate": 3.906e-05, "loss": 0.0689, "step": 5860 }, { "grad_norm": 0.8097068667411804, "learning_rate": 3.9126666666666666e-05, "loss": 0.0644, "step": 5870 }, { "grad_norm": 0.6337735652923584, "learning_rate": 3.9193333333333336e-05, "loss": 0.0702, "step": 5880 }, { "grad_norm": 0.5965064167976379, "learning_rate": 3.926e-05, "loss": 0.0636, "step": 5890 }, { "grad_norm": 0.7111570835113525, "learning_rate": 3.9326666666666664e-05, "loss": 0.0843, "step": 5900 }, { "grad_norm": 0.6558032631874084, "learning_rate": 3.9393333333333335e-05, "loss": 0.0721, "step": 5910 }, { "grad_norm": 0.8811517357826233, "learning_rate": 3.9460000000000005e-05, "loss": 0.058, "step": 5920 }, { "grad_norm": 0.5727807283401489, "learning_rate": 3.952666666666667e-05, "loss": 0.0565, "step": 5930 }, { "grad_norm": 0.7703284621238708, "learning_rate": 3.959333333333334e-05, "loss": 0.0628, "step": 5940 }, { "grad_norm": 0.7855651378631592, "learning_rate": 3.966e-05, "loss": 0.0643, "step": 5950 }, { "grad_norm": 0.6604870557785034, "learning_rate": 3.972666666666667e-05, "loss": 0.0489, "step": 5960 }, { "grad_norm": 1.0329747200012207, "learning_rate": 3.979333333333333e-05, "loss": 0.0626, "step": 5970 }, { "grad_norm": 0.7370186448097229, "learning_rate": 3.986e-05, "loss": 0.0611, "step": 5980 }, { "grad_norm": 0.7186323404312134, "learning_rate": 3.9926666666666665e-05, "loss": 0.0778, "step": 5990 }, { "grad_norm": 0.7756989002227783, "learning_rate": 3.9993333333333336e-05, "loss": 0.0793, "step": 6000 }, { "grad_norm": 0.530129075050354, "learning_rate": 4.0060000000000006e-05, "loss": 0.0598, "step": 6010 }, { "grad_norm": 0.5853173732757568, "learning_rate": 4.012666666666667e-05, "loss": 0.0594, "step": 6020 }, { "grad_norm": 0.7896164655685425, "learning_rate": 4.0193333333333334e-05, "loss": 0.0577, "step": 6030 }, { "grad_norm": 0.7080800533294678, "learning_rate": 4.0260000000000004e-05, "loss": 0.0576, "step": 6040 }, { "grad_norm": 0.7200854420661926, "learning_rate": 4.032666666666667e-05, "loss": 0.0695, "step": 6050 }, { "grad_norm": 0.6230189204216003, "learning_rate": 4.039333333333333e-05, "loss": 0.059, "step": 6060 }, { "grad_norm": 0.7762663960456848, "learning_rate": 4.046e-05, "loss": 0.0592, "step": 6070 }, { "grad_norm": 0.7505147457122803, "learning_rate": 4.0526666666666666e-05, "loss": 0.059, "step": 6080 }, { "grad_norm": 0.6495185494422913, "learning_rate": 4.0593333333333337e-05, "loss": 0.0626, "step": 6090 }, { "grad_norm": 0.5859949588775635, "learning_rate": 4.066e-05, "loss": 0.0566, "step": 6100 }, { "grad_norm": 0.4927872121334076, "learning_rate": 4.072666666666667e-05, "loss": 0.0471, "step": 6110 }, { "grad_norm": 0.6665168404579163, "learning_rate": 4.0793333333333335e-05, "loss": 0.0623, "step": 6120 }, { "grad_norm": 0.5340015292167664, "learning_rate": 4.0860000000000005e-05, "loss": 0.0536, "step": 6130 }, { "grad_norm": 0.6878080368041992, "learning_rate": 4.092666666666667e-05, "loss": 0.0611, "step": 6140 }, { "grad_norm": 0.6374160647392273, "learning_rate": 4.099333333333333e-05, "loss": 0.0804, "step": 6150 }, { "grad_norm": 0.665696918964386, "learning_rate": 4.106e-05, "loss": 0.0817, "step": 6160 }, { "grad_norm": 0.5189060568809509, "learning_rate": 4.112666666666667e-05, "loss": 0.0753, "step": 6170 }, { "grad_norm": 0.49675270915031433, "learning_rate": 4.119333333333333e-05, "loss": 0.0605, "step": 6180 }, { "grad_norm": 0.47174984216690063, "learning_rate": 4.126e-05, "loss": 0.0654, "step": 6190 }, { "grad_norm": 0.4972747266292572, "learning_rate": 4.132666666666667e-05, "loss": 0.0537, "step": 6200 }, { "grad_norm": 1.044490098953247, "learning_rate": 4.1393333333333336e-05, "loss": 0.0638, "step": 6210 }, { "grad_norm": 0.5415423512458801, "learning_rate": 4.1460000000000006e-05, "loss": 0.0499, "step": 6220 }, { "grad_norm": 0.6275346279144287, "learning_rate": 4.152666666666667e-05, "loss": 0.0515, "step": 6230 }, { "grad_norm": 0.6703759431838989, "learning_rate": 4.1593333333333334e-05, "loss": 0.0612, "step": 6240 }, { "grad_norm": 0.629763662815094, "learning_rate": 4.1660000000000004e-05, "loss": 0.0569, "step": 6250 }, { "grad_norm": 0.5016463398933411, "learning_rate": 4.172666666666667e-05, "loss": 0.0442, "step": 6260 }, { "grad_norm": 0.7246597409248352, "learning_rate": 4.179333333333333e-05, "loss": 0.0555, "step": 6270 }, { "grad_norm": 0.49334046244621277, "learning_rate": 4.186e-05, "loss": 0.0482, "step": 6280 }, { "grad_norm": 0.5283438563346863, "learning_rate": 4.192666666666667e-05, "loss": 0.0539, "step": 6290 }, { "grad_norm": 0.6874921917915344, "learning_rate": 4.199333333333334e-05, "loss": 0.0634, "step": 6300 }, { "grad_norm": 0.4755105972290039, "learning_rate": 4.206e-05, "loss": 0.0534, "step": 6310 }, { "grad_norm": 0.5521283149719238, "learning_rate": 4.212666666666667e-05, "loss": 0.0492, "step": 6320 }, { "grad_norm": 0.8532164096832275, "learning_rate": 4.2193333333333335e-05, "loss": 0.0568, "step": 6330 }, { "grad_norm": 0.9092068076133728, "learning_rate": 4.226e-05, "loss": 0.0655, "step": 6340 }, { "grad_norm": 0.7271108031272888, "learning_rate": 4.232666666666667e-05, "loss": 0.0604, "step": 6350 }, { "grad_norm": 0.7300598621368408, "learning_rate": 4.239333333333333e-05, "loss": 0.0624, "step": 6360 }, { "grad_norm": 0.5084381699562073, "learning_rate": 4.246e-05, "loss": 0.0607, "step": 6370 }, { "grad_norm": 0.6179808378219604, "learning_rate": 4.252666666666667e-05, "loss": 0.0528, "step": 6380 }, { "grad_norm": 0.9014716744422913, "learning_rate": 4.259333333333334e-05, "loss": 0.0665, "step": 6390 }, { "grad_norm": 0.6928950548171997, "learning_rate": 4.266e-05, "loss": 0.0646, "step": 6400 }, { "grad_norm": 0.4891488254070282, "learning_rate": 4.272666666666667e-05, "loss": 0.054, "step": 6410 }, { "grad_norm": 0.43803828954696655, "learning_rate": 4.2793333333333336e-05, "loss": 0.0651, "step": 6420 }, { "grad_norm": 0.6117514371871948, "learning_rate": 4.286e-05, "loss": 0.0541, "step": 6430 }, { "grad_norm": 0.611443817615509, "learning_rate": 4.292666666666667e-05, "loss": 0.0582, "step": 6440 }, { "grad_norm": 0.6022903919219971, "learning_rate": 4.2993333333333334e-05, "loss": 0.0493, "step": 6450 }, { "grad_norm": 0.5368929505348206, "learning_rate": 4.306e-05, "loss": 0.0474, "step": 6460 }, { "grad_norm": 0.5600165724754333, "learning_rate": 4.312666666666667e-05, "loss": 0.0524, "step": 6470 }, { "grad_norm": 0.5237003564834595, "learning_rate": 4.319333333333334e-05, "loss": 0.0478, "step": 6480 }, { "grad_norm": 0.9565159678459167, "learning_rate": 4.326e-05, "loss": 0.057, "step": 6490 }, { "grad_norm": 0.4345911145210266, "learning_rate": 4.332666666666667e-05, "loss": 0.0506, "step": 6500 }, { "grad_norm": 0.45902788639068604, "learning_rate": 4.339333333333334e-05, "loss": 0.0479, "step": 6510 }, { "grad_norm": 0.6290907263755798, "learning_rate": 4.346e-05, "loss": 0.0574, "step": 6520 }, { "grad_norm": 0.9923829436302185, "learning_rate": 4.352666666666667e-05, "loss": 0.0557, "step": 6530 }, { "grad_norm": 0.6493626236915588, "learning_rate": 4.3593333333333335e-05, "loss": 0.0724, "step": 6540 }, { "grad_norm": 0.9312186241149902, "learning_rate": 4.366e-05, "loss": 0.0538, "step": 6550 }, { "grad_norm": 0.6737660765647888, "learning_rate": 4.372666666666667e-05, "loss": 0.053, "step": 6560 }, { "grad_norm": 0.5026544332504272, "learning_rate": 4.379333333333333e-05, "loss": 0.0614, "step": 6570 }, { "grad_norm": 0.539628803730011, "learning_rate": 4.3860000000000004e-05, "loss": 0.0494, "step": 6580 }, { "grad_norm": 0.7142332196235657, "learning_rate": 4.3926666666666674e-05, "loss": 0.0653, "step": 6590 }, { "grad_norm": 0.508700430393219, "learning_rate": 4.399333333333334e-05, "loss": 0.065, "step": 6600 }, { "grad_norm": 0.5873987674713135, "learning_rate": 4.406e-05, "loss": 0.0628, "step": 6610 }, { "grad_norm": 0.3860016167163849, "learning_rate": 4.4126666666666665e-05, "loss": 0.0524, "step": 6620 }, { "grad_norm": 0.48712897300720215, "learning_rate": 4.4193333333333336e-05, "loss": 0.0515, "step": 6630 }, { "grad_norm": 0.5502434968948364, "learning_rate": 4.426e-05, "loss": 0.043, "step": 6640 }, { "grad_norm": 0.5834236145019531, "learning_rate": 4.4326666666666664e-05, "loss": 0.0701, "step": 6650 }, { "grad_norm": 0.7394608855247498, "learning_rate": 4.4393333333333334e-05, "loss": 0.0476, "step": 6660 }, { "grad_norm": 0.4474177956581116, "learning_rate": 4.4460000000000005e-05, "loss": 0.052, "step": 6670 }, { "grad_norm": 0.6153122186660767, "learning_rate": 4.452666666666667e-05, "loss": 0.0651, "step": 6680 }, { "grad_norm": 0.6284631490707397, "learning_rate": 4.459333333333334e-05, "loss": 0.0585, "step": 6690 }, { "grad_norm": 0.7914441227912903, "learning_rate": 4.466e-05, "loss": 0.0605, "step": 6700 }, { "grad_norm": 0.5888362526893616, "learning_rate": 4.4726666666666666e-05, "loss": 0.0641, "step": 6710 }, { "grad_norm": 0.5844311714172363, "learning_rate": 4.479333333333334e-05, "loss": 0.0627, "step": 6720 }, { "grad_norm": 0.7058500051498413, "learning_rate": 4.486e-05, "loss": 0.0585, "step": 6730 }, { "grad_norm": 0.5344608426094055, "learning_rate": 4.4926666666666665e-05, "loss": 0.0821, "step": 6740 }, { "grad_norm": 0.5993838906288147, "learning_rate": 4.4993333333333335e-05, "loss": 0.0511, "step": 6750 }, { "grad_norm": 0.3950278162956238, "learning_rate": 4.506e-05, "loss": 0.052, "step": 6760 }, { "grad_norm": 0.574613094329834, "learning_rate": 4.512666666666667e-05, "loss": 0.054, "step": 6770 }, { "grad_norm": 0.6546165347099304, "learning_rate": 4.519333333333334e-05, "loss": 0.0597, "step": 6780 }, { "grad_norm": 0.4548577070236206, "learning_rate": 4.5260000000000004e-05, "loss": 0.0459, "step": 6790 }, { "grad_norm": 0.7027691006660461, "learning_rate": 4.532666666666667e-05, "loss": 0.0739, "step": 6800 }, { "grad_norm": 0.4939044713973999, "learning_rate": 4.539333333333334e-05, "loss": 0.0499, "step": 6810 }, { "grad_norm": 0.7758312225341797, "learning_rate": 4.546e-05, "loss": 0.0645, "step": 6820 }, { "grad_norm": 0.5901247262954712, "learning_rate": 4.5526666666666666e-05, "loss": 0.0515, "step": 6830 }, { "grad_norm": 0.5693337321281433, "learning_rate": 4.5593333333333336e-05, "loss": 0.0646, "step": 6840 }, { "grad_norm": 0.4012148380279541, "learning_rate": 4.566e-05, "loss": 0.0539, "step": 6850 }, { "grad_norm": 0.4982934892177582, "learning_rate": 4.572666666666667e-05, "loss": 0.0474, "step": 6860 }, { "grad_norm": 0.6671026945114136, "learning_rate": 4.579333333333334e-05, "loss": 0.0664, "step": 6870 }, { "grad_norm": 0.39343997836112976, "learning_rate": 4.5860000000000005e-05, "loss": 0.0592, "step": 6880 }, { "grad_norm": 0.4844893515110016, "learning_rate": 4.592666666666667e-05, "loss": 0.0552, "step": 6890 }, { "grad_norm": 0.7803879976272583, "learning_rate": 4.599333333333334e-05, "loss": 0.061, "step": 6900 }, { "grad_norm": 0.5911462306976318, "learning_rate": 4.606e-05, "loss": 0.0547, "step": 6910 }, { "grad_norm": 0.5473074913024902, "learning_rate": 4.612666666666667e-05, "loss": 0.0533, "step": 6920 }, { "grad_norm": 0.6056014895439148, "learning_rate": 4.619333333333333e-05, "loss": 0.0582, "step": 6930 }, { "grad_norm": 0.6032299399375916, "learning_rate": 4.626e-05, "loss": 0.0491, "step": 6940 }, { "grad_norm": 0.6038334965705872, "learning_rate": 4.632666666666667e-05, "loss": 0.0509, "step": 6950 }, { "grad_norm": 0.4764114022254944, "learning_rate": 4.6393333333333335e-05, "loss": 0.0495, "step": 6960 }, { "grad_norm": 0.5333463549613953, "learning_rate": 4.6460000000000006e-05, "loss": 0.051, "step": 6970 }, { "grad_norm": 0.9409000277519226, "learning_rate": 4.652666666666667e-05, "loss": 0.0709, "step": 6980 }, { "grad_norm": 0.6028214693069458, "learning_rate": 4.659333333333333e-05, "loss": 0.0505, "step": 6990 }, { "grad_norm": 0.7944889068603516, "learning_rate": 4.6660000000000004e-05, "loss": 0.0628, "step": 7000 }, { "grad_norm": 0.486703485250473, "learning_rate": 4.672666666666667e-05, "loss": 0.0639, "step": 7010 }, { "grad_norm": 0.6898828744888306, "learning_rate": 4.679333333333333e-05, "loss": 0.0549, "step": 7020 }, { "grad_norm": 0.5503649115562439, "learning_rate": 4.686e-05, "loss": 0.0711, "step": 7030 }, { "grad_norm": 0.6003301739692688, "learning_rate": 4.6926666666666666e-05, "loss": 0.0538, "step": 7040 }, { "grad_norm": 0.6031595468521118, "learning_rate": 4.6993333333333336e-05, "loss": 0.0467, "step": 7050 }, { "grad_norm": 0.5059909820556641, "learning_rate": 4.706000000000001e-05, "loss": 0.0477, "step": 7060 }, { "grad_norm": 0.6075872182846069, "learning_rate": 4.712666666666667e-05, "loss": 0.0557, "step": 7070 }, { "grad_norm": 0.43924379348754883, "learning_rate": 4.7193333333333334e-05, "loss": 0.0655, "step": 7080 }, { "grad_norm": 0.6690900921821594, "learning_rate": 4.7260000000000005e-05, "loss": 0.0545, "step": 7090 }, { "grad_norm": 0.6113554239273071, "learning_rate": 4.732666666666667e-05, "loss": 0.0478, "step": 7100 }, { "grad_norm": 0.7039213180541992, "learning_rate": 4.739333333333333e-05, "loss": 0.0609, "step": 7110 }, { "grad_norm": 0.375674307346344, "learning_rate": 4.746e-05, "loss": 0.0482, "step": 7120 }, { "grad_norm": 0.45301395654678345, "learning_rate": 4.752666666666667e-05, "loss": 0.0496, "step": 7130 }, { "grad_norm": 0.5323056578636169, "learning_rate": 4.759333333333334e-05, "loss": 0.0724, "step": 7140 }, { "grad_norm": 0.5028154850006104, "learning_rate": 4.766000000000001e-05, "loss": 0.0633, "step": 7150 }, { "grad_norm": 1.0472854375839233, "learning_rate": 4.772666666666667e-05, "loss": 0.0558, "step": 7160 }, { "grad_norm": 0.47842827439308167, "learning_rate": 4.7793333333333335e-05, "loss": 0.0611, "step": 7170 }, { "grad_norm": 0.758563220500946, "learning_rate": 4.7860000000000006e-05, "loss": 0.0539, "step": 7180 }, { "grad_norm": 0.4373360276222229, "learning_rate": 4.792666666666667e-05, "loss": 0.0626, "step": 7190 }, { "grad_norm": 0.9097452759742737, "learning_rate": 4.7993333333333333e-05, "loss": 0.0566, "step": 7200 }, { "grad_norm": 0.6082994341850281, "learning_rate": 4.8060000000000004e-05, "loss": 0.0633, "step": 7210 }, { "grad_norm": 0.5146417021751404, "learning_rate": 4.812666666666667e-05, "loss": 0.0534, "step": 7220 }, { "grad_norm": 0.6694885492324829, "learning_rate": 4.819333333333333e-05, "loss": 0.0581, "step": 7230 }, { "grad_norm": 0.620811939239502, "learning_rate": 4.826e-05, "loss": 0.0571, "step": 7240 }, { "grad_norm": 0.4891824424266815, "learning_rate": 4.832666666666667e-05, "loss": 0.0587, "step": 7250 }, { "grad_norm": 0.44405677914619446, "learning_rate": 4.8393333333333336e-05, "loss": 0.0497, "step": 7260 }, { "grad_norm": 0.3847010135650635, "learning_rate": 4.846e-05, "loss": 0.055, "step": 7270 }, { "grad_norm": 0.5437508821487427, "learning_rate": 4.852666666666667e-05, "loss": 0.0523, "step": 7280 }, { "grad_norm": 0.4284757077693939, "learning_rate": 4.8593333333333335e-05, "loss": 0.0542, "step": 7290 }, { "grad_norm": 0.4768996238708496, "learning_rate": 4.866e-05, "loss": 0.0546, "step": 7300 }, { "grad_norm": 0.6228386759757996, "learning_rate": 4.872666666666667e-05, "loss": 0.056, "step": 7310 }, { "grad_norm": 0.5436099171638489, "learning_rate": 4.879333333333333e-05, "loss": 0.0443, "step": 7320 }, { "grad_norm": 0.6532066464424133, "learning_rate": 4.886e-05, "loss": 0.0482, "step": 7330 }, { "grad_norm": 0.4749053120613098, "learning_rate": 4.8926666666666674e-05, "loss": 0.0451, "step": 7340 }, { "grad_norm": 0.5406192541122437, "learning_rate": 4.899333333333334e-05, "loss": 0.0478, "step": 7350 }, { "grad_norm": 0.6736198663711548, "learning_rate": 4.906e-05, "loss": 0.045, "step": 7360 }, { "grad_norm": 0.4369029998779297, "learning_rate": 4.912666666666667e-05, "loss": 0.0509, "step": 7370 }, { "grad_norm": 0.4142475724220276, "learning_rate": 4.9193333333333336e-05, "loss": 0.0556, "step": 7380 }, { "grad_norm": 0.41133758425712585, "learning_rate": 4.926e-05, "loss": 0.0459, "step": 7390 }, { "grad_norm": 0.6196662187576294, "learning_rate": 4.932666666666667e-05, "loss": 0.0557, "step": 7400 }, { "grad_norm": 0.644213855266571, "learning_rate": 4.9393333333333334e-05, "loss": 0.0551, "step": 7410 }, { "grad_norm": 0.5968185067176819, "learning_rate": 4.946e-05, "loss": 0.0672, "step": 7420 }, { "grad_norm": 0.3400327265262604, "learning_rate": 4.952666666666667e-05, "loss": 0.0563, "step": 7430 }, { "grad_norm": 0.48985227942466736, "learning_rate": 4.959333333333334e-05, "loss": 0.0514, "step": 7440 }, { "grad_norm": 0.45586472749710083, "learning_rate": 4.966e-05, "loss": 0.0556, "step": 7450 }, { "grad_norm": 0.44740229845046997, "learning_rate": 4.972666666666667e-05, "loss": 0.0563, "step": 7460 }, { "grad_norm": 0.44795987010002136, "learning_rate": 4.9793333333333337e-05, "loss": 0.0601, "step": 7470 }, { "grad_norm": 0.8394309282302856, "learning_rate": 4.986e-05, "loss": 0.0623, "step": 7480 }, { "grad_norm": 0.5219326615333557, "learning_rate": 4.992666666666667e-05, "loss": 0.0548, "step": 7490 }, { "grad_norm": 0.42233818769454956, "learning_rate": 4.9993333333333335e-05, "loss": 0.052, "step": 7500 }, { "grad_norm": 0.47432994842529297, "learning_rate": 5.0060000000000005e-05, "loss": 0.0498, "step": 7510 }, { "grad_norm": 0.4188641309738159, "learning_rate": 5.012666666666667e-05, "loss": 0.0538, "step": 7520 }, { "grad_norm": 0.4787975549697876, "learning_rate": 5.019333333333334e-05, "loss": 0.0529, "step": 7530 }, { "grad_norm": 0.3708520233631134, "learning_rate": 5.026000000000001e-05, "loss": 0.0587, "step": 7540 }, { "grad_norm": 0.47412174940109253, "learning_rate": 5.032666666666667e-05, "loss": 0.0512, "step": 7550 }, { "grad_norm": 0.37569114565849304, "learning_rate": 5.039333333333334e-05, "loss": 0.0572, "step": 7560 }, { "grad_norm": 0.6000373363494873, "learning_rate": 5.046000000000001e-05, "loss": 0.0662, "step": 7570 }, { "grad_norm": 0.5483166575431824, "learning_rate": 5.0526666666666665e-05, "loss": 0.0636, "step": 7580 }, { "grad_norm": 0.5128931403160095, "learning_rate": 5.0593333333333336e-05, "loss": 0.0587, "step": 7590 }, { "grad_norm": 0.43221381306648254, "learning_rate": 5.0660000000000006e-05, "loss": 0.0777, "step": 7600 }, { "grad_norm": 0.5518417954444885, "learning_rate": 5.072666666666667e-05, "loss": 0.0628, "step": 7610 }, { "grad_norm": 0.4929676055908203, "learning_rate": 5.079333333333334e-05, "loss": 0.063, "step": 7620 }, { "grad_norm": 0.5114158391952515, "learning_rate": 5.0860000000000004e-05, "loss": 0.0589, "step": 7630 }, { "grad_norm": 0.5349626541137695, "learning_rate": 5.092666666666667e-05, "loss": 0.0566, "step": 7640 }, { "grad_norm": 0.4932856857776642, "learning_rate": 5.099333333333334e-05, "loss": 0.047, "step": 7650 }, { "grad_norm": 0.5421043634414673, "learning_rate": 5.106000000000001e-05, "loss": 0.0594, "step": 7660 }, { "grad_norm": 0.5682058334350586, "learning_rate": 5.1126666666666666e-05, "loss": 0.0631, "step": 7670 }, { "grad_norm": 0.4574052393436432, "learning_rate": 5.119333333333334e-05, "loss": 0.0598, "step": 7680 }, { "grad_norm": 0.5155687928199768, "learning_rate": 5.1259999999999994e-05, "loss": 0.0567, "step": 7690 }, { "grad_norm": 0.4017797112464905, "learning_rate": 5.1326666666666664e-05, "loss": 0.0538, "step": 7700 }, { "grad_norm": 0.4335934519767761, "learning_rate": 5.1393333333333335e-05, "loss": 0.0568, "step": 7710 }, { "grad_norm": 0.4052076041698456, "learning_rate": 5.146e-05, "loss": 0.0574, "step": 7720 }, { "grad_norm": 0.3866910934448242, "learning_rate": 5.152666666666667e-05, "loss": 0.0563, "step": 7730 }, { "grad_norm": 0.4716585576534271, "learning_rate": 5.159333333333334e-05, "loss": 0.0573, "step": 7740 }, { "grad_norm": 0.3993101119995117, "learning_rate": 5.166e-05, "loss": 0.0486, "step": 7750 }, { "grad_norm": 0.5190070271492004, "learning_rate": 5.172666666666667e-05, "loss": 0.062, "step": 7760 }, { "grad_norm": 0.521399199962616, "learning_rate": 5.179333333333334e-05, "loss": 0.0551, "step": 7770 }, { "grad_norm": 0.4616048336029053, "learning_rate": 5.1859999999999995e-05, "loss": 0.0549, "step": 7780 }, { "grad_norm": 0.48452556133270264, "learning_rate": 5.1926666666666665e-05, "loss": 0.0571, "step": 7790 }, { "grad_norm": 0.4756247103214264, "learning_rate": 5.1993333333333336e-05, "loss": 0.0534, "step": 7800 }, { "grad_norm": 0.36168503761291504, "learning_rate": 5.206e-05, "loss": 0.0498, "step": 7810 }, { "grad_norm": 0.3573993742465973, "learning_rate": 5.212666666666667e-05, "loss": 0.0502, "step": 7820 }, { "grad_norm": 0.6397037506103516, "learning_rate": 5.219333333333334e-05, "loss": 0.0526, "step": 7830 }, { "grad_norm": 0.44581088423728943, "learning_rate": 5.226e-05, "loss": 0.0471, "step": 7840 }, { "grad_norm": 0.39274877309799194, "learning_rate": 5.232666666666667e-05, "loss": 0.0517, "step": 7850 }, { "grad_norm": 0.46772488951683044, "learning_rate": 5.239333333333334e-05, "loss": 0.0532, "step": 7860 }, { "grad_norm": 0.6599056124687195, "learning_rate": 5.2459999999999996e-05, "loss": 0.0594, "step": 7870 }, { "grad_norm": 0.535521388053894, "learning_rate": 5.2526666666666666e-05, "loss": 0.0515, "step": 7880 }, { "grad_norm": 0.56806880235672, "learning_rate": 5.259333333333334e-05, "loss": 0.0548, "step": 7890 }, { "grad_norm": 0.43246811628341675, "learning_rate": 5.266e-05, "loss": 0.0664, "step": 7900 }, { "grad_norm": 0.4427797794342041, "learning_rate": 5.272666666666667e-05, "loss": 0.0594, "step": 7910 }, { "grad_norm": 0.4746604263782501, "learning_rate": 5.279333333333334e-05, "loss": 0.0773, "step": 7920 }, { "grad_norm": 0.3975016176700592, "learning_rate": 5.286e-05, "loss": 0.0521, "step": 7930 }, { "grad_norm": 0.4433860182762146, "learning_rate": 5.292666666666667e-05, "loss": 0.0528, "step": 7940 }, { "grad_norm": 0.5693032741546631, "learning_rate": 5.299333333333334e-05, "loss": 0.0485, "step": 7950 }, { "grad_norm": 0.4883446991443634, "learning_rate": 5.306e-05, "loss": 0.0477, "step": 7960 }, { "grad_norm": 0.4459402561187744, "learning_rate": 5.312666666666667e-05, "loss": 0.0467, "step": 7970 }, { "grad_norm": 0.41703373193740845, "learning_rate": 5.319333333333334e-05, "loss": 0.0525, "step": 7980 }, { "grad_norm": 0.37204015254974365, "learning_rate": 5.326e-05, "loss": 0.0509, "step": 7990 }, { "grad_norm": 0.5854716300964355, "learning_rate": 5.332666666666667e-05, "loss": 0.0463, "step": 8000 }, { "grad_norm": 0.46919530630111694, "learning_rate": 5.339333333333334e-05, "loss": 0.0538, "step": 8010 }, { "grad_norm": 0.6402229070663452, "learning_rate": 5.346e-05, "loss": 0.0548, "step": 8020 }, { "grad_norm": 0.4047524333000183, "learning_rate": 5.352666666666667e-05, "loss": 0.0537, "step": 8030 }, { "grad_norm": 0.46421679854393005, "learning_rate": 5.359333333333334e-05, "loss": 0.0507, "step": 8040 }, { "grad_norm": 0.44451072812080383, "learning_rate": 5.366e-05, "loss": 0.0486, "step": 8050 }, { "grad_norm": 0.467941552400589, "learning_rate": 5.372666666666667e-05, "loss": 0.049, "step": 8060 }, { "grad_norm": 0.42200058698654175, "learning_rate": 5.379333333333334e-05, "loss": 0.0506, "step": 8070 }, { "grad_norm": 0.6569159030914307, "learning_rate": 5.386e-05, "loss": 0.0653, "step": 8080 }, { "grad_norm": 0.4918549060821533, "learning_rate": 5.3926666666666666e-05, "loss": 0.0517, "step": 8090 }, { "grad_norm": 0.5554744005203247, "learning_rate": 5.399333333333334e-05, "loss": 0.0564, "step": 8100 }, { "grad_norm": 0.6493207812309265, "learning_rate": 5.406e-05, "loss": 0.0728, "step": 8110 }, { "grad_norm": 0.42394495010375977, "learning_rate": 5.412666666666667e-05, "loss": 0.0611, "step": 8120 }, { "grad_norm": 0.6702175140380859, "learning_rate": 5.419333333333334e-05, "loss": 0.0632, "step": 8130 }, { "grad_norm": 0.4567899703979492, "learning_rate": 5.426e-05, "loss": 0.0586, "step": 8140 }, { "grad_norm": 0.5365421175956726, "learning_rate": 5.432666666666667e-05, "loss": 0.047, "step": 8150 }, { "grad_norm": 0.4280093014240265, "learning_rate": 5.439333333333334e-05, "loss": 0.0618, "step": 8160 }, { "grad_norm": 0.3195238709449768, "learning_rate": 5.446e-05, "loss": 0.0501, "step": 8170 }, { "grad_norm": 0.4682216942310333, "learning_rate": 5.452666666666667e-05, "loss": 0.0554, "step": 8180 }, { "grad_norm": 0.3936092257499695, "learning_rate": 5.459333333333334e-05, "loss": 0.0503, "step": 8190 }, { "grad_norm": 0.47407618165016174, "learning_rate": 5.466e-05, "loss": 0.0518, "step": 8200 }, { "grad_norm": 0.4363124966621399, "learning_rate": 5.472666666666667e-05, "loss": 0.0485, "step": 8210 }, { "grad_norm": 0.524634599685669, "learning_rate": 5.479333333333334e-05, "loss": 0.0519, "step": 8220 }, { "grad_norm": 0.5096756815910339, "learning_rate": 5.486e-05, "loss": 0.0448, "step": 8230 }, { "grad_norm": 0.4534046947956085, "learning_rate": 5.492666666666667e-05, "loss": 0.0488, "step": 8240 }, { "grad_norm": 0.4770040810108185, "learning_rate": 5.499333333333334e-05, "loss": 0.0458, "step": 8250 }, { "grad_norm": 0.656970202922821, "learning_rate": 5.506e-05, "loss": 0.0528, "step": 8260 }, { "grad_norm": 0.4216992259025574, "learning_rate": 5.512666666666667e-05, "loss": 0.0433, "step": 8270 }, { "grad_norm": 0.8917325139045715, "learning_rate": 5.519333333333334e-05, "loss": 0.0597, "step": 8280 }, { "grad_norm": 0.5178801417350769, "learning_rate": 5.526e-05, "loss": 0.0458, "step": 8290 }, { "grad_norm": 0.5452189445495605, "learning_rate": 5.5326666666666673e-05, "loss": 0.0598, "step": 8300 }, { "grad_norm": 0.4448328912258148, "learning_rate": 5.5393333333333344e-05, "loss": 0.0528, "step": 8310 }, { "grad_norm": 0.5624200701713562, "learning_rate": 5.546e-05, "loss": 0.0442, "step": 8320 }, { "grad_norm": 0.366788774728775, "learning_rate": 5.552666666666667e-05, "loss": 0.0481, "step": 8330 }, { "grad_norm": 0.3056066036224365, "learning_rate": 5.559333333333333e-05, "loss": 0.0487, "step": 8340 }, { "grad_norm": 0.27744022011756897, "learning_rate": 5.566e-05, "loss": 0.0565, "step": 8350 }, { "grad_norm": 0.381742924451828, "learning_rate": 5.572666666666667e-05, "loss": 0.0414, "step": 8360 }, { "grad_norm": 0.4717567563056946, "learning_rate": 5.579333333333333e-05, "loss": 0.0462, "step": 8370 }, { "grad_norm": 0.4849095940589905, "learning_rate": 5.5860000000000004e-05, "loss": 0.0512, "step": 8380 }, { "grad_norm": 0.3502735197544098, "learning_rate": 5.5926666666666674e-05, "loss": 0.0483, "step": 8390 }, { "grad_norm": 0.420671284198761, "learning_rate": 5.599333333333333e-05, "loss": 0.047, "step": 8400 }, { "grad_norm": 0.36360621452331543, "learning_rate": 5.606e-05, "loss": 0.0458, "step": 8410 }, { "grad_norm": 0.3855123817920685, "learning_rate": 5.612666666666667e-05, "loss": 0.0443, "step": 8420 }, { "grad_norm": 0.5534126162528992, "learning_rate": 5.619333333333333e-05, "loss": 0.0482, "step": 8430 }, { "grad_norm": 0.3384400010108948, "learning_rate": 5.626e-05, "loss": 0.0467, "step": 8440 }, { "grad_norm": 0.4008719325065613, "learning_rate": 5.632666666666667e-05, "loss": 0.0472, "step": 8450 }, { "grad_norm": 0.4824151396751404, "learning_rate": 5.6393333333333334e-05, "loss": 0.0497, "step": 8460 }, { "grad_norm": 0.5830230116844177, "learning_rate": 5.6460000000000005e-05, "loss": 0.0539, "step": 8470 }, { "grad_norm": 0.37582433223724365, "learning_rate": 5.6526666666666675e-05, "loss": 0.0467, "step": 8480 }, { "grad_norm": 0.7263686060905457, "learning_rate": 5.659333333333333e-05, "loss": 0.0495, "step": 8490 }, { "grad_norm": 0.47431549429893494, "learning_rate": 5.666e-05, "loss": 0.0521, "step": 8500 }, { "grad_norm": 0.4045359790325165, "learning_rate": 5.6726666666666674e-05, "loss": 0.0514, "step": 8510 }, { "grad_norm": 0.49129199981689453, "learning_rate": 5.679333333333333e-05, "loss": 0.0503, "step": 8520 }, { "grad_norm": 0.43723195791244507, "learning_rate": 5.686e-05, "loss": 0.0493, "step": 8530 }, { "grad_norm": 0.3957187533378601, "learning_rate": 5.692666666666667e-05, "loss": 0.0488, "step": 8540 }, { "grad_norm": 0.3875593841075897, "learning_rate": 5.699333333333333e-05, "loss": 0.0561, "step": 8550 }, { "grad_norm": 0.3536827862262726, "learning_rate": 5.706e-05, "loss": 0.0585, "step": 8560 }, { "grad_norm": 0.5165871977806091, "learning_rate": 5.712666666666667e-05, "loss": 0.0816, "step": 8570 }, { "grad_norm": 0.49417001008987427, "learning_rate": 5.7193333333333333e-05, "loss": 0.0565, "step": 8580 }, { "grad_norm": 0.4258023202419281, "learning_rate": 5.7260000000000004e-05, "loss": 0.0489, "step": 8590 }, { "grad_norm": 0.3806534707546234, "learning_rate": 5.7326666666666675e-05, "loss": 0.0468, "step": 8600 }, { "grad_norm": 0.5907344818115234, "learning_rate": 5.739333333333333e-05, "loss": 0.0446, "step": 8610 }, { "grad_norm": 0.45540526509284973, "learning_rate": 5.746e-05, "loss": 0.0597, "step": 8620 }, { "grad_norm": 0.5803499817848206, "learning_rate": 5.752666666666667e-05, "loss": 0.0544, "step": 8630 }, { "grad_norm": 0.5524553060531616, "learning_rate": 5.759333333333333e-05, "loss": 0.0466, "step": 8640 }, { "grad_norm": 0.42717868089675903, "learning_rate": 5.766e-05, "loss": 0.0602, "step": 8650 }, { "grad_norm": 0.6276233196258545, "learning_rate": 5.772666666666667e-05, "loss": 0.0512, "step": 8660 }, { "grad_norm": 0.4954438805580139, "learning_rate": 5.7793333333333335e-05, "loss": 0.0518, "step": 8670 }, { "grad_norm": 0.5408304929733276, "learning_rate": 5.7860000000000005e-05, "loss": 0.051, "step": 8680 }, { "grad_norm": 0.5175471305847168, "learning_rate": 5.7926666666666676e-05, "loss": 0.0606, "step": 8690 }, { "grad_norm": 0.5273305773735046, "learning_rate": 5.799333333333333e-05, "loss": 0.052, "step": 8700 }, { "grad_norm": 0.47961896657943726, "learning_rate": 5.806e-05, "loss": 0.0566, "step": 8710 }, { "grad_norm": 0.7129685282707214, "learning_rate": 5.8126666666666674e-05, "loss": 0.0644, "step": 8720 }, { "grad_norm": 0.36906298995018005, "learning_rate": 5.819333333333333e-05, "loss": 0.0506, "step": 8730 }, { "grad_norm": 0.41461482644081116, "learning_rate": 5.826e-05, "loss": 0.0402, "step": 8740 }, { "grad_norm": 0.3666791319847107, "learning_rate": 5.832666666666667e-05, "loss": 0.0526, "step": 8750 }, { "grad_norm": 0.39546456933021545, "learning_rate": 5.8393333333333336e-05, "loss": 0.0501, "step": 8760 }, { "grad_norm": 0.49060606956481934, "learning_rate": 5.8460000000000006e-05, "loss": 0.0489, "step": 8770 }, { "grad_norm": 0.4865029752254486, "learning_rate": 5.852666666666668e-05, "loss": 0.0594, "step": 8780 }, { "grad_norm": 0.3670542538166046, "learning_rate": 5.8593333333333334e-05, "loss": 0.0435, "step": 8790 }, { "grad_norm": 0.39799028635025024, "learning_rate": 5.8660000000000004e-05, "loss": 0.041, "step": 8800 }, { "grad_norm": 0.4557614028453827, "learning_rate": 5.8726666666666675e-05, "loss": 0.0499, "step": 8810 }, { "grad_norm": 0.3501565158367157, "learning_rate": 5.879333333333333e-05, "loss": 0.0492, "step": 8820 }, { "grad_norm": 0.4324937164783478, "learning_rate": 5.886e-05, "loss": 0.0465, "step": 8830 }, { "grad_norm": 0.5099568367004395, "learning_rate": 5.892666666666667e-05, "loss": 0.0475, "step": 8840 }, { "grad_norm": 0.36903810501098633, "learning_rate": 5.8993333333333337e-05, "loss": 0.0395, "step": 8850 }, { "grad_norm": 0.5332229137420654, "learning_rate": 5.906000000000001e-05, "loss": 0.0429, "step": 8860 }, { "grad_norm": 0.6872561573982239, "learning_rate": 5.912666666666668e-05, "loss": 0.049, "step": 8870 }, { "grad_norm": 0.4178507328033447, "learning_rate": 5.9193333333333335e-05, "loss": 0.0534, "step": 8880 }, { "grad_norm": 0.36069661378860474, "learning_rate": 5.9260000000000005e-05, "loss": 0.051, "step": 8890 }, { "grad_norm": 0.4171922504901886, "learning_rate": 5.9326666666666676e-05, "loss": 0.0546, "step": 8900 }, { "grad_norm": 0.41675880551338196, "learning_rate": 5.939333333333333e-05, "loss": 0.0476, "step": 8910 }, { "grad_norm": 0.7640299797058105, "learning_rate": 5.946e-05, "loss": 0.0568, "step": 8920 }, { "grad_norm": 0.5243542790412903, "learning_rate": 5.9526666666666674e-05, "loss": 0.0498, "step": 8930 }, { "grad_norm": 0.4023260772228241, "learning_rate": 5.959333333333334e-05, "loss": 0.0509, "step": 8940 }, { "grad_norm": 0.44034695625305176, "learning_rate": 5.966000000000001e-05, "loss": 0.0464, "step": 8950 }, { "grad_norm": 0.5076577067375183, "learning_rate": 5.9726666666666665e-05, "loss": 0.0499, "step": 8960 }, { "grad_norm": 0.40685954689979553, "learning_rate": 5.9793333333333336e-05, "loss": 0.0474, "step": 8970 }, { "grad_norm": 0.41929033398628235, "learning_rate": 5.9860000000000006e-05, "loss": 0.05, "step": 8980 }, { "grad_norm": 0.5715698599815369, "learning_rate": 5.992666666666666e-05, "loss": 0.0686, "step": 8990 }, { "grad_norm": 0.45178449153900146, "learning_rate": 5.9993333333333334e-05, "loss": 0.0459, "step": 9000 }, { "grad_norm": 0.3437651991844177, "learning_rate": 6.0060000000000004e-05, "loss": 0.0433, "step": 9010 }, { "grad_norm": 0.44896170496940613, "learning_rate": 6.012666666666666e-05, "loss": 0.0512, "step": 9020 }, { "grad_norm": 0.4050815999507904, "learning_rate": 6.019333333333333e-05, "loss": 0.0575, "step": 9030 }, { "grad_norm": 0.5239343643188477, "learning_rate": 6.026e-05, "loss": 0.0535, "step": 9040 }, { "grad_norm": 0.46755269169807434, "learning_rate": 6.0326666666666666e-05, "loss": 0.0506, "step": 9050 }, { "grad_norm": 0.42947953939437866, "learning_rate": 6.039333333333334e-05, "loss": 0.0618, "step": 9060 }, { "grad_norm": 0.49181559681892395, "learning_rate": 6.046000000000001e-05, "loss": 0.0439, "step": 9070 }, { "grad_norm": 0.2990398705005646, "learning_rate": 6.0526666666666664e-05, "loss": 0.0453, "step": 9080 }, { "grad_norm": 0.42388206720352173, "learning_rate": 6.0593333333333335e-05, "loss": 0.0541, "step": 9090 }, { "grad_norm": 0.5191183686256409, "learning_rate": 6.0660000000000005e-05, "loss": 0.0565, "step": 9100 }, { "grad_norm": 0.5549837350845337, "learning_rate": 6.072666666666666e-05, "loss": 0.0497, "step": 9110 }, { "grad_norm": 0.4161820113658905, "learning_rate": 6.079333333333333e-05, "loss": 0.0502, "step": 9120 }, { "grad_norm": 0.4567694067955017, "learning_rate": 6.0860000000000003e-05, "loss": 0.0574, "step": 9130 }, { "grad_norm": 0.3637053072452545, "learning_rate": 6.092666666666667e-05, "loss": 0.0472, "step": 9140 }, { "grad_norm": 0.45002973079681396, "learning_rate": 6.099333333333334e-05, "loss": 0.0542, "step": 9150 }, { "grad_norm": 0.288115531206131, "learning_rate": 6.106e-05, "loss": 0.0454, "step": 9160 }, { "grad_norm": 0.4265120327472687, "learning_rate": 6.112666666666667e-05, "loss": 0.0468, "step": 9170 }, { "grad_norm": 0.46130436658859253, "learning_rate": 6.119333333333333e-05, "loss": 0.0476, "step": 9180 }, { "grad_norm": 0.3245646357536316, "learning_rate": 6.126e-05, "loss": 0.0491, "step": 9190 }, { "grad_norm": 0.3542478382587433, "learning_rate": 6.132666666666667e-05, "loss": 0.0512, "step": 9200 }, { "grad_norm": 0.5123178958892822, "learning_rate": 6.139333333333333e-05, "loss": 0.063, "step": 9210 }, { "grad_norm": 0.4395374059677124, "learning_rate": 6.146000000000001e-05, "loss": 0.054, "step": 9220 }, { "grad_norm": 0.5134905576705933, "learning_rate": 6.152666666666666e-05, "loss": 0.0485, "step": 9230 }, { "grad_norm": 0.40982726216316223, "learning_rate": 6.159333333333334e-05, "loss": 0.0597, "step": 9240 }, { "grad_norm": 0.47157225012779236, "learning_rate": 6.166e-05, "loss": 0.0553, "step": 9250 }, { "grad_norm": 0.29124948382377625, "learning_rate": 6.172666666666667e-05, "loss": 0.0503, "step": 9260 }, { "grad_norm": 0.49929964542388916, "learning_rate": 6.179333333333333e-05, "loss": 0.0459, "step": 9270 }, { "grad_norm": 0.5016180872917175, "learning_rate": 6.186000000000001e-05, "loss": 0.0502, "step": 9280 }, { "grad_norm": 0.6767850518226624, "learning_rate": 6.192666666666667e-05, "loss": 0.0569, "step": 9290 }, { "grad_norm": 0.45311036705970764, "learning_rate": 6.199333333333333e-05, "loss": 0.0591, "step": 9300 }, { "grad_norm": 0.5449201464653015, "learning_rate": 6.206000000000001e-05, "loss": 0.0438, "step": 9310 }, { "grad_norm": 0.41412433981895447, "learning_rate": 6.212666666666666e-05, "loss": 0.0665, "step": 9320 }, { "grad_norm": 0.8090946078300476, "learning_rate": 6.219333333333334e-05, "loss": 0.0479, "step": 9330 }, { "grad_norm": 0.4230554401874542, "learning_rate": 6.226e-05, "loss": 0.0555, "step": 9340 }, { "grad_norm": 0.5115377306938171, "learning_rate": 6.232666666666667e-05, "loss": 0.0577, "step": 9350 }, { "grad_norm": 0.47464096546173096, "learning_rate": 6.239333333333333e-05, "loss": 0.0619, "step": 9360 }, { "grad_norm": 0.33942297101020813, "learning_rate": 6.246000000000001e-05, "loss": 0.0597, "step": 9370 }, { "grad_norm": 0.33970341086387634, "learning_rate": 6.252666666666667e-05, "loss": 0.0502, "step": 9380 }, { "grad_norm": 0.6223675608634949, "learning_rate": 6.259333333333334e-05, "loss": 0.0459, "step": 9390 }, { "grad_norm": 0.4102921187877655, "learning_rate": 6.266000000000001e-05, "loss": 0.0602, "step": 9400 }, { "grad_norm": 0.5330482125282288, "learning_rate": 6.272666666666666e-05, "loss": 0.0579, "step": 9410 }, { "grad_norm": 0.6679390668869019, "learning_rate": 6.279333333333334e-05, "loss": 0.0544, "step": 9420 }, { "grad_norm": 0.3735118806362152, "learning_rate": 6.286e-05, "loss": 0.049, "step": 9430 }, { "grad_norm": 0.4647802412509918, "learning_rate": 6.292666666666667e-05, "loss": 0.046, "step": 9440 }, { "grad_norm": 0.48529568314552307, "learning_rate": 6.299333333333333e-05, "loss": 0.0436, "step": 9450 }, { "grad_norm": 0.436606764793396, "learning_rate": 6.306000000000001e-05, "loss": 0.0567, "step": 9460 }, { "grad_norm": 0.5202628970146179, "learning_rate": 6.312666666666667e-05, "loss": 0.0665, "step": 9470 }, { "grad_norm": 0.39217910170555115, "learning_rate": 6.319333333333334e-05, "loss": 0.0487, "step": 9480 }, { "grad_norm": 0.40251633524894714, "learning_rate": 6.326000000000001e-05, "loss": 0.0536, "step": 9490 }, { "grad_norm": 0.41000691056251526, "learning_rate": 6.332666666666666e-05, "loss": 0.0431, "step": 9500 }, { "grad_norm": 0.41212326288223267, "learning_rate": 6.339333333333334e-05, "loss": 0.0526, "step": 9510 }, { "grad_norm": 0.3948768377304077, "learning_rate": 6.346e-05, "loss": 0.0541, "step": 9520 }, { "grad_norm": 0.40163007378578186, "learning_rate": 6.352666666666667e-05, "loss": 0.0529, "step": 9530 }, { "grad_norm": 0.7326449155807495, "learning_rate": 6.359333333333333e-05, "loss": 0.0573, "step": 9540 }, { "grad_norm": 0.38042405247688293, "learning_rate": 6.366000000000001e-05, "loss": 0.053, "step": 9550 }, { "grad_norm": 0.356251060962677, "learning_rate": 6.372666666666667e-05, "loss": 0.052, "step": 9560 }, { "grad_norm": 0.474716454744339, "learning_rate": 6.379333333333334e-05, "loss": 0.0676, "step": 9570 }, { "grad_norm": 0.3837219476699829, "learning_rate": 6.386e-05, "loss": 0.0561, "step": 9580 }, { "grad_norm": 0.3912246823310852, "learning_rate": 6.392666666666667e-05, "loss": 0.0521, "step": 9590 }, { "grad_norm": 0.434617280960083, "learning_rate": 6.399333333333334e-05, "loss": 0.0438, "step": 9600 }, { "grad_norm": 0.5845884084701538, "learning_rate": 6.405999999999999e-05, "loss": 0.0626, "step": 9610 }, { "grad_norm": 0.3376130163669586, "learning_rate": 6.412666666666667e-05, "loss": 0.0576, "step": 9620 }, { "grad_norm": 0.3797074556350708, "learning_rate": 6.419333333333333e-05, "loss": 0.05, "step": 9630 }, { "grad_norm": 0.35215234756469727, "learning_rate": 6.426e-05, "loss": 0.049, "step": 9640 }, { "grad_norm": 0.3111308515071869, "learning_rate": 6.432666666666666e-05, "loss": 0.0527, "step": 9650 }, { "grad_norm": 0.48507601022720337, "learning_rate": 6.439333333333334e-05, "loss": 0.0506, "step": 9660 }, { "grad_norm": 0.3019435703754425, "learning_rate": 6.446e-05, "loss": 0.0538, "step": 9670 }, { "grad_norm": 0.3913423717021942, "learning_rate": 6.452666666666667e-05, "loss": 0.0456, "step": 9680 }, { "grad_norm": 0.4946564733982086, "learning_rate": 6.459333333333334e-05, "loss": 0.0433, "step": 9690 }, { "grad_norm": 0.28093549609184265, "learning_rate": 6.466e-05, "loss": 0.0431, "step": 9700 }, { "grad_norm": 0.4341411590576172, "learning_rate": 6.472666666666667e-05, "loss": 0.0399, "step": 9710 }, { "grad_norm": 0.4544132649898529, "learning_rate": 6.479333333333334e-05, "loss": 0.045, "step": 9720 }, { "grad_norm": 0.5435458421707153, "learning_rate": 6.486e-05, "loss": 0.0473, "step": 9730 }, { "grad_norm": 0.40976276993751526, "learning_rate": 6.492666666666666e-05, "loss": 0.0422, "step": 9740 }, { "grad_norm": 0.3880355954170227, "learning_rate": 6.499333333333334e-05, "loss": 0.0428, "step": 9750 }, { "grad_norm": 0.3336527943611145, "learning_rate": 6.506e-05, "loss": 0.0531, "step": 9760 }, { "grad_norm": 0.42931029200553894, "learning_rate": 6.512666666666667e-05, "loss": 0.044, "step": 9770 }, { "grad_norm": 0.48509514331817627, "learning_rate": 6.519333333333334e-05, "loss": 0.0429, "step": 9780 }, { "grad_norm": 0.5088478922843933, "learning_rate": 6.526e-05, "loss": 0.0453, "step": 9790 }, { "grad_norm": 0.3981255888938904, "learning_rate": 6.532666666666667e-05, "loss": 0.0388, "step": 9800 }, { "grad_norm": 0.43846380710601807, "learning_rate": 6.539333333333334e-05, "loss": 0.0513, "step": 9810 }, { "grad_norm": 0.4019007682800293, "learning_rate": 6.546e-05, "loss": 0.0492, "step": 9820 }, { "grad_norm": 0.4160029888153076, "learning_rate": 6.552666666666666e-05, "loss": 0.0445, "step": 9830 }, { "grad_norm": 0.5818679928779602, "learning_rate": 6.559333333333334e-05, "loss": 0.0668, "step": 9840 }, { "grad_norm": 0.40179723501205444, "learning_rate": 6.566e-05, "loss": 0.059, "step": 9850 }, { "grad_norm": 0.22493375837802887, "learning_rate": 6.572666666666667e-05, "loss": 0.0449, "step": 9860 }, { "grad_norm": 0.3143998384475708, "learning_rate": 6.579333333333335e-05, "loss": 0.0446, "step": 9870 }, { "grad_norm": 0.5124716758728027, "learning_rate": 6.586e-05, "loss": 0.0471, "step": 9880 }, { "grad_norm": 0.2772192060947418, "learning_rate": 6.592666666666667e-05, "loss": 0.0396, "step": 9890 }, { "grad_norm": 0.4419647753238678, "learning_rate": 6.599333333333334e-05, "loss": 0.0504, "step": 9900 }, { "grad_norm": 0.3195209503173828, "learning_rate": 6.606e-05, "loss": 0.0493, "step": 9910 }, { "grad_norm": 0.4104918837547302, "learning_rate": 6.612666666666666e-05, "loss": 0.0538, "step": 9920 }, { "grad_norm": 0.32307955622673035, "learning_rate": 6.619333333333334e-05, "loss": 0.0465, "step": 9930 }, { "grad_norm": 0.34354475140571594, "learning_rate": 6.626e-05, "loss": 0.051, "step": 9940 }, { "grad_norm": 0.2818838953971863, "learning_rate": 6.632666666666667e-05, "loss": 0.0353, "step": 9950 }, { "grad_norm": 0.42048895359039307, "learning_rate": 6.639333333333335e-05, "loss": 0.0437, "step": 9960 }, { "grad_norm": 0.45537957549095154, "learning_rate": 6.646e-05, "loss": 0.063, "step": 9970 }, { "grad_norm": 0.2652677595615387, "learning_rate": 6.652666666666667e-05, "loss": 0.0466, "step": 9980 }, { "grad_norm": 0.4761482775211334, "learning_rate": 6.659333333333334e-05, "loss": 0.0505, "step": 9990 }, { "grad_norm": 0.618414580821991, "learning_rate": 6.666e-05, "loss": 0.0515, "step": 10000 }, { "grad_norm": 0.3918192982673645, "learning_rate": 6.672666666666667e-05, "loss": 0.0524, "step": 10010 }, { "grad_norm": 0.5777579545974731, "learning_rate": 6.679333333333334e-05, "loss": 0.0592, "step": 10020 }, { "grad_norm": 0.3014216721057892, "learning_rate": 6.686000000000001e-05, "loss": 0.0416, "step": 10030 }, { "grad_norm": 1.040177583694458, "learning_rate": 6.692666666666667e-05, "loss": 0.0502, "step": 10040 }, { "grad_norm": 0.39509275555610657, "learning_rate": 6.699333333333333e-05, "loss": 0.0522, "step": 10050 }, { "grad_norm": 0.44175437092781067, "learning_rate": 6.706e-05, "loss": 0.0552, "step": 10060 }, { "grad_norm": 0.3969416916370392, "learning_rate": 6.712666666666668e-05, "loss": 0.0464, "step": 10070 }, { "grad_norm": 0.31447744369506836, "learning_rate": 6.719333333333334e-05, "loss": 0.0498, "step": 10080 }, { "grad_norm": 0.3433268368244171, "learning_rate": 6.726e-05, "loss": 0.0534, "step": 10090 }, { "grad_norm": 0.39178699254989624, "learning_rate": 6.732666666666667e-05, "loss": 0.0473, "step": 10100 }, { "grad_norm": 0.4743427634239197, "learning_rate": 6.739333333333334e-05, "loss": 0.0445, "step": 10110 }, { "grad_norm": 0.31934818625450134, "learning_rate": 6.746e-05, "loss": 0.0446, "step": 10120 }, { "grad_norm": 0.5977555513381958, "learning_rate": 6.752666666666667e-05, "loss": 0.0513, "step": 10130 }, { "grad_norm": 0.2786189019680023, "learning_rate": 6.759333333333334e-05, "loss": 0.0527, "step": 10140 }, { "grad_norm": 0.4927806556224823, "learning_rate": 6.766e-05, "loss": 0.0531, "step": 10150 }, { "grad_norm": 0.5283465385437012, "learning_rate": 6.772666666666668e-05, "loss": 0.0549, "step": 10160 }, { "grad_norm": 0.33116790652275085, "learning_rate": 6.779333333333334e-05, "loss": 0.0526, "step": 10170 }, { "grad_norm": 0.3450121283531189, "learning_rate": 6.786e-05, "loss": 0.0587, "step": 10180 }, { "grad_norm": 0.5707031488418579, "learning_rate": 6.792666666666667e-05, "loss": 0.0577, "step": 10190 }, { "grad_norm": 0.49753105640411377, "learning_rate": 6.799333333333333e-05, "loss": 0.0539, "step": 10200 }, { "grad_norm": 0.39280709624290466, "learning_rate": 6.806e-05, "loss": 0.0578, "step": 10210 }, { "grad_norm": 0.34533125162124634, "learning_rate": 6.812666666666667e-05, "loss": 0.0472, "step": 10220 }, { "grad_norm": 0.3554990887641907, "learning_rate": 6.819333333333334e-05, "loss": 0.0602, "step": 10230 }, { "grad_norm": 0.3207111656665802, "learning_rate": 6.826e-05, "loss": 0.0537, "step": 10240 }, { "grad_norm": 0.3090677857398987, "learning_rate": 6.832666666666668e-05, "loss": 0.0428, "step": 10250 }, { "grad_norm": 0.4407850503921509, "learning_rate": 6.839333333333333e-05, "loss": 0.0359, "step": 10260 }, { "grad_norm": 0.36921393871307373, "learning_rate": 6.846e-05, "loss": 0.0439, "step": 10270 }, { "grad_norm": 0.42940646409988403, "learning_rate": 6.852666666666667e-05, "loss": 0.0477, "step": 10280 }, { "grad_norm": 0.3604639172554016, "learning_rate": 6.859333333333333e-05, "loss": 0.0582, "step": 10290 }, { "grad_norm": 0.4998013377189636, "learning_rate": 6.866e-05, "loss": 0.048, "step": 10300 }, { "grad_norm": 0.4907115697860718, "learning_rate": 6.872666666666667e-05, "loss": 0.0511, "step": 10310 }, { "grad_norm": 0.39935117959976196, "learning_rate": 6.879333333333334e-05, "loss": 0.0511, "step": 10320 }, { "grad_norm": 0.507604718208313, "learning_rate": 6.886e-05, "loss": 0.0528, "step": 10330 }, { "grad_norm": 0.4092940390110016, "learning_rate": 6.892666666666668e-05, "loss": 0.0562, "step": 10340 }, { "grad_norm": 0.5260224342346191, "learning_rate": 6.899333333333333e-05, "loss": 0.0569, "step": 10350 }, { "grad_norm": 0.27851229906082153, "learning_rate": 6.906e-05, "loss": 0.0474, "step": 10360 }, { "grad_norm": 0.43413612246513367, "learning_rate": 6.912666666666667e-05, "loss": 0.047, "step": 10370 }, { "grad_norm": 0.34468162059783936, "learning_rate": 6.919333333333333e-05, "loss": 0.0572, "step": 10380 }, { "grad_norm": 0.40534648299217224, "learning_rate": 6.926e-05, "loss": 0.0535, "step": 10390 }, { "grad_norm": 0.45371824502944946, "learning_rate": 6.932666666666667e-05, "loss": 0.0563, "step": 10400 }, { "grad_norm": 0.34753134846687317, "learning_rate": 6.939333333333334e-05, "loss": 0.0506, "step": 10410 }, { "grad_norm": 0.27584782242774963, "learning_rate": 6.946e-05, "loss": 0.051, "step": 10420 }, { "grad_norm": 0.4082081913948059, "learning_rate": 6.952666666666668e-05, "loss": 0.0471, "step": 10430 }, { "grad_norm": 0.49820569157600403, "learning_rate": 6.959333333333333e-05, "loss": 0.0436, "step": 10440 }, { "grad_norm": 0.36748766899108887, "learning_rate": 6.966000000000001e-05, "loss": 0.0544, "step": 10450 }, { "grad_norm": 0.49277642369270325, "learning_rate": 6.972666666666667e-05, "loss": 0.0468, "step": 10460 }, { "grad_norm": 0.37839388847351074, "learning_rate": 6.979333333333333e-05, "loss": 0.0458, "step": 10470 }, { "grad_norm": 0.5056620240211487, "learning_rate": 6.986e-05, "loss": 0.0465, "step": 10480 }, { "grad_norm": 0.33400827646255493, "learning_rate": 6.992666666666668e-05, "loss": 0.0487, "step": 10490 }, { "grad_norm": 0.3382699489593506, "learning_rate": 6.999333333333334e-05, "loss": 0.046, "step": 10500 }, { "grad_norm": 0.44116342067718506, "learning_rate": 7.006e-05, "loss": 0.0575, "step": 10510 }, { "grad_norm": 0.3986585736274719, "learning_rate": 7.012666666666667e-05, "loss": 0.0464, "step": 10520 }, { "grad_norm": 0.326126366853714, "learning_rate": 7.019333333333333e-05, "loss": 0.0557, "step": 10530 }, { "grad_norm": 0.3506603240966797, "learning_rate": 7.026000000000001e-05, "loss": 0.0494, "step": 10540 }, { "grad_norm": 0.2351943850517273, "learning_rate": 7.032666666666667e-05, "loss": 0.0417, "step": 10550 }, { "grad_norm": 0.29467546939849854, "learning_rate": 7.039333333333334e-05, "loss": 0.0458, "step": 10560 }, { "grad_norm": 0.3612089157104492, "learning_rate": 7.046e-05, "loss": 0.0482, "step": 10570 }, { "grad_norm": 0.4125777781009674, "learning_rate": 7.052666666666668e-05, "loss": 0.0491, "step": 10580 }, { "grad_norm": 0.5354036092758179, "learning_rate": 7.059333333333333e-05, "loss": 0.0497, "step": 10590 }, { "grad_norm": 0.26005005836486816, "learning_rate": 7.066e-05, "loss": 0.0565, "step": 10600 }, { "grad_norm": 0.5530880689620972, "learning_rate": 7.072666666666667e-05, "loss": 0.0694, "step": 10610 }, { "grad_norm": 0.6266830563545227, "learning_rate": 7.079333333333333e-05, "loss": 0.0689, "step": 10620 }, { "grad_norm": 0.45589688420295715, "learning_rate": 7.086000000000001e-05, "loss": 0.0446, "step": 10630 }, { "grad_norm": 0.4380970597267151, "learning_rate": 7.092666666666667e-05, "loss": 0.058, "step": 10640 }, { "grad_norm": 0.4123903512954712, "learning_rate": 7.099333333333334e-05, "loss": 0.054, "step": 10650 }, { "grad_norm": 0.40881359577178955, "learning_rate": 7.106e-05, "loss": 0.0564, "step": 10660 }, { "grad_norm": 0.45998576283454895, "learning_rate": 7.112666666666668e-05, "loss": 0.0531, "step": 10670 }, { "grad_norm": 0.3810851275920868, "learning_rate": 7.119333333333333e-05, "loss": 0.0487, "step": 10680 }, { "grad_norm": 0.3913674056529999, "learning_rate": 7.126e-05, "loss": 0.0479, "step": 10690 }, { "grad_norm": 0.5356025099754333, "learning_rate": 7.132666666666667e-05, "loss": 0.0717, "step": 10700 }, { "grad_norm": 0.5210488438606262, "learning_rate": 7.139333333333333e-05, "loss": 0.0566, "step": 10710 }, { "grad_norm": 0.31902965903282166, "learning_rate": 7.146000000000001e-05, "loss": 0.0487, "step": 10720 }, { "grad_norm": 0.375192791223526, "learning_rate": 7.152666666666667e-05, "loss": 0.0404, "step": 10730 }, { "grad_norm": 0.4082728922367096, "learning_rate": 7.159333333333334e-05, "loss": 0.0439, "step": 10740 }, { "grad_norm": 0.3301912844181061, "learning_rate": 7.166e-05, "loss": 0.0451, "step": 10750 }, { "grad_norm": 0.41665494441986084, "learning_rate": 7.172666666666668e-05, "loss": 0.0507, "step": 10760 }, { "grad_norm": 0.513468861579895, "learning_rate": 7.179333333333333e-05, "loss": 0.0636, "step": 10770 }, { "grad_norm": 0.4132060110569, "learning_rate": 7.186e-05, "loss": 0.0589, "step": 10780 }, { "grad_norm": 0.5011406540870667, "learning_rate": 7.192666666666667e-05, "loss": 0.0478, "step": 10790 }, { "grad_norm": 0.329355388879776, "learning_rate": 7.199333333333333e-05, "loss": 0.0522, "step": 10800 }, { "grad_norm": 0.43307825922966003, "learning_rate": 7.206000000000001e-05, "loss": 0.0481, "step": 10810 }, { "grad_norm": 0.3120965361595154, "learning_rate": 7.212666666666666e-05, "loss": 0.0499, "step": 10820 }, { "grad_norm": 0.47775205969810486, "learning_rate": 7.219333333333334e-05, "loss": 0.049, "step": 10830 }, { "grad_norm": 0.5353235006332397, "learning_rate": 7.226e-05, "loss": 0.0577, "step": 10840 }, { "grad_norm": 0.4266534745693207, "learning_rate": 7.232666666666667e-05, "loss": 0.0483, "step": 10850 }, { "grad_norm": 0.3446367084980011, "learning_rate": 7.239333333333333e-05, "loss": 0.044, "step": 10860 }, { "grad_norm": 0.4224257171154022, "learning_rate": 7.246000000000001e-05, "loss": 0.0496, "step": 10870 }, { "grad_norm": 0.5006431937217712, "learning_rate": 7.252666666666667e-05, "loss": 0.0582, "step": 10880 }, { "grad_norm": 0.37037166953086853, "learning_rate": 7.259333333333334e-05, "loss": 0.0675, "step": 10890 }, { "grad_norm": 0.4611720144748688, "learning_rate": 7.266000000000001e-05, "loss": 0.0553, "step": 10900 }, { "grad_norm": 0.2916713058948517, "learning_rate": 7.272666666666666e-05, "loss": 0.0482, "step": 10910 }, { "grad_norm": 0.3015526831150055, "learning_rate": 7.279333333333334e-05, "loss": 0.0448, "step": 10920 }, { "grad_norm": 0.3344968557357788, "learning_rate": 7.286e-05, "loss": 0.0596, "step": 10930 }, { "grad_norm": 0.4734781086444855, "learning_rate": 7.292666666666667e-05, "loss": 0.0487, "step": 10940 }, { "grad_norm": 0.4187449514865875, "learning_rate": 7.299333333333333e-05, "loss": 0.0452, "step": 10950 }, { "grad_norm": 0.2941419184207916, "learning_rate": 7.306000000000001e-05, "loss": 0.0414, "step": 10960 }, { "grad_norm": 0.42996135354042053, "learning_rate": 7.312666666666666e-05, "loss": 0.0435, "step": 10970 }, { "grad_norm": 0.378291517496109, "learning_rate": 7.319333333333334e-05, "loss": 0.0455, "step": 10980 }, { "grad_norm": 0.33741551637649536, "learning_rate": 7.326e-05, "loss": 0.0428, "step": 10990 }, { "grad_norm": 0.31537100672721863, "learning_rate": 7.332666666666666e-05, "loss": 0.0419, "step": 11000 }, { "grad_norm": 0.45548760890960693, "learning_rate": 7.339333333333334e-05, "loss": 0.0485, "step": 11010 }, { "grad_norm": 0.2613597512245178, "learning_rate": 7.346e-05, "loss": 0.041, "step": 11020 }, { "grad_norm": 0.6782601475715637, "learning_rate": 7.352666666666667e-05, "loss": 0.0407, "step": 11030 }, { "grad_norm": 0.3929145932197571, "learning_rate": 7.359333333333333e-05, "loss": 0.0434, "step": 11040 }, { "grad_norm": 0.39304783940315247, "learning_rate": 7.366000000000001e-05, "loss": 0.0437, "step": 11050 }, { "grad_norm": 0.4325157105922699, "learning_rate": 7.372666666666666e-05, "loss": 0.0458, "step": 11060 }, { "grad_norm": 0.3069150745868683, "learning_rate": 7.379333333333334e-05, "loss": 0.0397, "step": 11070 }, { "grad_norm": 0.31902065873146057, "learning_rate": 7.386e-05, "loss": 0.0435, "step": 11080 }, { "grad_norm": 0.2939276695251465, "learning_rate": 7.392666666666666e-05, "loss": 0.0446, "step": 11090 }, { "grad_norm": 0.3615293800830841, "learning_rate": 7.399333333333334e-05, "loss": 0.0537, "step": 11100 }, { "grad_norm": 0.4017758369445801, "learning_rate": 7.406e-05, "loss": 0.047, "step": 11110 }, { "grad_norm": 0.26368993520736694, "learning_rate": 7.412666666666667e-05, "loss": 0.041, "step": 11120 }, { "grad_norm": 0.36735936999320984, "learning_rate": 7.419333333333333e-05, "loss": 0.0483, "step": 11130 }, { "grad_norm": 0.34734535217285156, "learning_rate": 7.426000000000001e-05, "loss": 0.0446, "step": 11140 }, { "grad_norm": 0.32815566658973694, "learning_rate": 7.432666666666666e-05, "loss": 0.0378, "step": 11150 }, { "grad_norm": 0.42490872740745544, "learning_rate": 7.439333333333334e-05, "loss": 0.0395, "step": 11160 }, { "grad_norm": 0.4794369339942932, "learning_rate": 7.446e-05, "loss": 0.0395, "step": 11170 }, { "grad_norm": 0.37596356868743896, "learning_rate": 7.452666666666667e-05, "loss": 0.043, "step": 11180 }, { "grad_norm": 0.33033671975135803, "learning_rate": 7.459333333333334e-05, "loss": 0.0416, "step": 11190 }, { "grad_norm": 1.2781473398208618, "learning_rate": 7.466000000000001e-05, "loss": 0.0522, "step": 11200 }, { "grad_norm": 0.317171573638916, "learning_rate": 7.472666666666667e-05, "loss": 0.0551, "step": 11210 }, { "grad_norm": 0.47612670063972473, "learning_rate": 7.479333333333333e-05, "loss": 0.0503, "step": 11220 }, { "grad_norm": 0.6842347979545593, "learning_rate": 7.486000000000001e-05, "loss": 0.075, "step": 11230 }, { "grad_norm": 0.49506914615631104, "learning_rate": 7.492666666666666e-05, "loss": 0.0526, "step": 11240 }, { "grad_norm": 0.37543392181396484, "learning_rate": 7.499333333333334e-05, "loss": 0.0486, "step": 11250 }, { "grad_norm": 0.2920888364315033, "learning_rate": 7.506e-05, "loss": 0.0425, "step": 11260 }, { "grad_norm": 0.2585902810096741, "learning_rate": 7.512666666666667e-05, "loss": 0.0473, "step": 11270 }, { "grad_norm": 0.4542688727378845, "learning_rate": 7.519333333333334e-05, "loss": 0.0555, "step": 11280 }, { "grad_norm": 0.3029760718345642, "learning_rate": 7.526000000000001e-05, "loss": 0.0442, "step": 11290 }, { "grad_norm": 0.46017178893089294, "learning_rate": 7.532666666666667e-05, "loss": 0.046, "step": 11300 }, { "grad_norm": 0.522038996219635, "learning_rate": 7.539333333333334e-05, "loss": 0.0483, "step": 11310 }, { "grad_norm": 0.4054238498210907, "learning_rate": 7.546000000000001e-05, "loss": 0.0491, "step": 11320 }, { "grad_norm": 0.6266618371009827, "learning_rate": 7.552666666666666e-05, "loss": 0.0467, "step": 11330 }, { "grad_norm": 0.3214032053947449, "learning_rate": 7.559333333333334e-05, "loss": 0.0411, "step": 11340 }, { "grad_norm": 0.2372804880142212, "learning_rate": 7.566e-05, "loss": 0.059, "step": 11350 }, { "grad_norm": 0.3049181401729584, "learning_rate": 7.572666666666667e-05, "loss": 0.0396, "step": 11360 }, { "grad_norm": 0.482536256313324, "learning_rate": 7.579333333333335e-05, "loss": 0.0451, "step": 11370 }, { "grad_norm": 0.3592497706413269, "learning_rate": 7.586000000000001e-05, "loss": 0.0526, "step": 11380 }, { "grad_norm": 0.4161100685596466, "learning_rate": 7.592666666666667e-05, "loss": 0.0492, "step": 11390 }, { "grad_norm": 0.36538568139076233, "learning_rate": 7.599333333333334e-05, "loss": 0.0596, "step": 11400 }, { "grad_norm": 0.9115229249000549, "learning_rate": 7.606000000000001e-05, "loss": 0.0583, "step": 11410 }, { "grad_norm": 0.38496676087379456, "learning_rate": 7.612666666666666e-05, "loss": 0.0698, "step": 11420 }, { "grad_norm": 0.36714690923690796, "learning_rate": 7.619333333333334e-05, "loss": 0.0555, "step": 11430 }, { "grad_norm": 0.35133349895477295, "learning_rate": 7.625999999999999e-05, "loss": 0.0559, "step": 11440 }, { "grad_norm": 0.3116353750228882, "learning_rate": 7.632666666666667e-05, "loss": 0.0506, "step": 11450 }, { "grad_norm": 0.4780397117137909, "learning_rate": 7.639333333333333e-05, "loss": 0.0506, "step": 11460 }, { "grad_norm": 0.3923448324203491, "learning_rate": 7.646e-05, "loss": 0.0523, "step": 11470 }, { "grad_norm": 0.34492745995521545, "learning_rate": 7.652666666666667e-05, "loss": 0.0492, "step": 11480 }, { "grad_norm": 0.45867249369621277, "learning_rate": 7.659333333333334e-05, "loss": 0.0527, "step": 11490 }, { "grad_norm": 0.3760645389556885, "learning_rate": 7.666e-05, "loss": 0.0486, "step": 11500 }, { "grad_norm": 0.356158971786499, "learning_rate": 7.672666666666667e-05, "loss": 0.0448, "step": 11510 }, { "grad_norm": 0.3855131268501282, "learning_rate": 7.679333333333334e-05, "loss": 0.0442, "step": 11520 }, { "grad_norm": 0.35286569595336914, "learning_rate": 7.685999999999999e-05, "loss": 0.0434, "step": 11530 }, { "grad_norm": 0.4170685410499573, "learning_rate": 7.692666666666667e-05, "loss": 0.0485, "step": 11540 }, { "grad_norm": 0.4585936963558197, "learning_rate": 7.699333333333333e-05, "loss": 0.0469, "step": 11550 }, { "grad_norm": 0.28729382157325745, "learning_rate": 7.706e-05, "loss": 0.0461, "step": 11560 }, { "grad_norm": 0.34102460741996765, "learning_rate": 7.712666666666667e-05, "loss": 0.0406, "step": 11570 }, { "grad_norm": 0.3709545135498047, "learning_rate": 7.719333333333334e-05, "loss": 0.0437, "step": 11580 }, { "grad_norm": 0.3821510374546051, "learning_rate": 7.726e-05, "loss": 0.0445, "step": 11590 }, { "grad_norm": 0.345838338136673, "learning_rate": 7.732666666666667e-05, "loss": 0.0437, "step": 11600 }, { "grad_norm": 0.3185345530509949, "learning_rate": 7.739333333333334e-05, "loss": 0.0488, "step": 11610 }, { "grad_norm": 0.46388131380081177, "learning_rate": 7.746e-05, "loss": 0.0556, "step": 11620 }, { "grad_norm": 0.42541277408599854, "learning_rate": 7.752666666666667e-05, "loss": 0.06, "step": 11630 }, { "grad_norm": 0.4173830449581146, "learning_rate": 7.759333333333333e-05, "loss": 0.0471, "step": 11640 }, { "grad_norm": 0.44463270902633667, "learning_rate": 7.766e-05, "loss": 0.0526, "step": 11650 }, { "grad_norm": 0.28406450152397156, "learning_rate": 7.772666666666668e-05, "loss": 0.0459, "step": 11660 }, { "grad_norm": 0.26626551151275635, "learning_rate": 7.779333333333334e-05, "loss": 0.0422, "step": 11670 }, { "grad_norm": 0.3441368043422699, "learning_rate": 7.786e-05, "loss": 0.0458, "step": 11680 }, { "grad_norm": 0.4809906780719757, "learning_rate": 7.792666666666667e-05, "loss": 0.0495, "step": 11690 }, { "grad_norm": 0.2899744212627411, "learning_rate": 7.799333333333334e-05, "loss": 0.048, "step": 11700 }, { "grad_norm": 0.3689432740211487, "learning_rate": 7.806e-05, "loss": 0.0468, "step": 11710 }, { "grad_norm": 0.4053790867328644, "learning_rate": 7.812666666666667e-05, "loss": 0.0451, "step": 11720 }, { "grad_norm": 0.43831339478492737, "learning_rate": 7.819333333333334e-05, "loss": 0.0407, "step": 11730 }, { "grad_norm": 0.3520897626876831, "learning_rate": 7.826e-05, "loss": 0.0483, "step": 11740 }, { "grad_norm": 0.39419424533843994, "learning_rate": 7.832666666666668e-05, "loss": 0.0566, "step": 11750 }, { "grad_norm": 0.2687566876411438, "learning_rate": 7.839333333333334e-05, "loss": 0.043, "step": 11760 }, { "grad_norm": 0.386972039937973, "learning_rate": 7.846e-05, "loss": 0.0463, "step": 11770 }, { "grad_norm": 0.38370034098625183, "learning_rate": 7.852666666666667e-05, "loss": 0.0445, "step": 11780 }, { "grad_norm": 0.6130326986312866, "learning_rate": 7.859333333333335e-05, "loss": 0.0385, "step": 11790 }, { "grad_norm": 0.41680780053138733, "learning_rate": 7.866e-05, "loss": 0.0575, "step": 11800 }, { "grad_norm": 0.44040045142173767, "learning_rate": 7.872666666666667e-05, "loss": 0.0555, "step": 11810 }, { "grad_norm": 0.40002086758613586, "learning_rate": 7.879333333333334e-05, "loss": 0.0445, "step": 11820 }, { "grad_norm": 0.43557098507881165, "learning_rate": 7.886e-05, "loss": 0.052, "step": 11830 }, { "grad_norm": 0.40865612030029297, "learning_rate": 7.892666666666666e-05, "loss": 0.0441, "step": 11840 }, { "grad_norm": 0.36259517073631287, "learning_rate": 7.899333333333334e-05, "loss": 0.0471, "step": 11850 }, { "grad_norm": 0.29187336564064026, "learning_rate": 7.906e-05, "loss": 0.0473, "step": 11860 }, { "grad_norm": 0.28003907203674316, "learning_rate": 7.912666666666667e-05, "loss": 0.0411, "step": 11870 }, { "grad_norm": 0.2971418499946594, "learning_rate": 7.919333333333335e-05, "loss": 0.0563, "step": 11880 }, { "grad_norm": 0.4706987142562866, "learning_rate": 7.926e-05, "loss": 0.0454, "step": 11890 }, { "grad_norm": 0.3522617518901825, "learning_rate": 7.932666666666667e-05, "loss": 0.0434, "step": 11900 }, { "grad_norm": 0.3514290153980255, "learning_rate": 7.939333333333334e-05, "loss": 0.0485, "step": 11910 }, { "grad_norm": 0.4116135537624359, "learning_rate": 7.946e-05, "loss": 0.0448, "step": 11920 }, { "grad_norm": 0.2951929569244385, "learning_rate": 7.952666666666667e-05, "loss": 0.0433, "step": 11930 }, { "grad_norm": 0.2872064411640167, "learning_rate": 7.959333333333334e-05, "loss": 0.0453, "step": 11940 }, { "grad_norm": 0.2916097939014435, "learning_rate": 7.966e-05, "loss": 0.0367, "step": 11950 }, { "grad_norm": 0.2614409923553467, "learning_rate": 7.972666666666667e-05, "loss": 0.0378, "step": 11960 }, { "grad_norm": 0.6520981788635254, "learning_rate": 7.979333333333335e-05, "loss": 0.0605, "step": 11970 }, { "grad_norm": 0.4227345287799835, "learning_rate": 7.986e-05, "loss": 0.0518, "step": 11980 }, { "grad_norm": 0.2919676601886749, "learning_rate": 7.992666666666668e-05, "loss": 0.047, "step": 11990 }, { "grad_norm": 0.3635987639427185, "learning_rate": 7.999333333333334e-05, "loss": 0.0539, "step": 12000 }, { "grad_norm": 0.4861406683921814, "learning_rate": 8.006e-05, "loss": 0.048, "step": 12010 }, { "grad_norm": 0.4037449359893799, "learning_rate": 8.012666666666667e-05, "loss": 0.049, "step": 12020 }, { "grad_norm": 0.37279388308525085, "learning_rate": 8.019333333333334e-05, "loss": 0.0462, "step": 12030 }, { "grad_norm": 0.6630434393882751, "learning_rate": 8.026000000000001e-05, "loss": 0.0512, "step": 12040 }, { "grad_norm": 0.2979549467563629, "learning_rate": 8.032666666666667e-05, "loss": 0.0503, "step": 12050 }, { "grad_norm": 0.27219781279563904, "learning_rate": 8.039333333333335e-05, "loss": 0.0495, "step": 12060 }, { "grad_norm": 0.20554843544960022, "learning_rate": 8.046e-05, "loss": 0.0608, "step": 12070 }, { "grad_norm": 0.3564065992832184, "learning_rate": 8.052666666666668e-05, "loss": 0.0465, "step": 12080 }, { "grad_norm": 0.7968202829360962, "learning_rate": 8.059333333333333e-05, "loss": 0.0456, "step": 12090 }, { "grad_norm": 0.4169902205467224, "learning_rate": 8.066e-05, "loss": 0.0533, "step": 12100 }, { "grad_norm": 0.3233066201210022, "learning_rate": 8.072666666666667e-05, "loss": 0.0541, "step": 12110 }, { "grad_norm": 0.5436503887176514, "learning_rate": 8.079333333333333e-05, "loss": 0.0496, "step": 12120 }, { "grad_norm": 0.29901251196861267, "learning_rate": 8.086000000000001e-05, "loss": 0.0521, "step": 12130 }, { "grad_norm": 0.2917186915874481, "learning_rate": 8.092666666666667e-05, "loss": 0.0423, "step": 12140 }, { "grad_norm": 0.27253803610801697, "learning_rate": 8.099333333333334e-05, "loss": 0.0459, "step": 12150 }, { "grad_norm": 0.26578834652900696, "learning_rate": 8.106e-05, "loss": 0.0424, "step": 12160 }, { "grad_norm": 0.24849137663841248, "learning_rate": 8.112666666666668e-05, "loss": 0.0372, "step": 12170 }, { "grad_norm": 0.3932749032974243, "learning_rate": 8.119333333333333e-05, "loss": 0.053, "step": 12180 }, { "grad_norm": 0.44558611512184143, "learning_rate": 8.126e-05, "loss": 0.0451, "step": 12190 }, { "grad_norm": 0.31839385628700256, "learning_rate": 8.132666666666667e-05, "loss": 0.0535, "step": 12200 }, { "grad_norm": 0.533458948135376, "learning_rate": 8.139333333333333e-05, "loss": 0.0609, "step": 12210 }, { "grad_norm": 0.2973533868789673, "learning_rate": 8.146000000000001e-05, "loss": 0.0526, "step": 12220 }, { "grad_norm": 0.33203834295272827, "learning_rate": 8.152666666666667e-05, "loss": 0.0487, "step": 12230 }, { "grad_norm": 0.8942248821258545, "learning_rate": 8.159333333333334e-05, "loss": 0.045, "step": 12240 }, { "grad_norm": 0.4065372347831726, "learning_rate": 8.166e-05, "loss": 0.0655, "step": 12250 }, { "grad_norm": 0.4044954180717468, "learning_rate": 8.172666666666668e-05, "loss": 0.0709, "step": 12260 }, { "grad_norm": 0.29279622435569763, "learning_rate": 8.179333333333333e-05, "loss": 0.0503, "step": 12270 }, { "grad_norm": 0.3729848265647888, "learning_rate": 8.186e-05, "loss": 0.0501, "step": 12280 }, { "grad_norm": 0.33671218156814575, "learning_rate": 8.192666666666667e-05, "loss": 0.0468, "step": 12290 }, { "grad_norm": 0.39106008410453796, "learning_rate": 8.199333333333333e-05, "loss": 0.0381, "step": 12300 }, { "grad_norm": 0.32810062170028687, "learning_rate": 8.206e-05, "loss": 0.0464, "step": 12310 }, { "grad_norm": 0.40377214550971985, "learning_rate": 8.212666666666667e-05, "loss": 0.0485, "step": 12320 }, { "grad_norm": 0.3083757758140564, "learning_rate": 8.219333333333334e-05, "loss": 0.0468, "step": 12330 }, { "grad_norm": 0.27921101450920105, "learning_rate": 8.226e-05, "loss": 0.0546, "step": 12340 }, { "grad_norm": 0.3428201675415039, "learning_rate": 8.232666666666668e-05, "loss": 0.0508, "step": 12350 }, { "grad_norm": 0.32243552803993225, "learning_rate": 8.239333333333333e-05, "loss": 0.0456, "step": 12360 }, { "grad_norm": 0.3227999210357666, "learning_rate": 8.246e-05, "loss": 0.0445, "step": 12370 }, { "grad_norm": 0.2947309911251068, "learning_rate": 8.252666666666667e-05, "loss": 0.048, "step": 12380 }, { "grad_norm": 0.40932586789131165, "learning_rate": 8.259333333333333e-05, "loss": 0.0478, "step": 12390 }, { "grad_norm": 0.33539852499961853, "learning_rate": 8.266e-05, "loss": 0.0491, "step": 12400 }, { "grad_norm": 0.3695005476474762, "learning_rate": 8.272666666666668e-05, "loss": 0.0521, "step": 12410 }, { "grad_norm": 0.2973119914531708, "learning_rate": 8.279333333333334e-05, "loss": 0.0476, "step": 12420 }, { "grad_norm": 0.7236604690551758, "learning_rate": 8.286e-05, "loss": 0.0535, "step": 12430 }, { "grad_norm": 0.3229115903377533, "learning_rate": 8.292666666666668e-05, "loss": 0.0532, "step": 12440 }, { "grad_norm": 0.33220139145851135, "learning_rate": 8.299333333333333e-05, "loss": 0.0445, "step": 12450 }, { "grad_norm": 0.35670217871665955, "learning_rate": 8.306000000000001e-05, "loss": 0.0541, "step": 12460 }, { "grad_norm": 0.2907991111278534, "learning_rate": 8.312666666666667e-05, "loss": 0.0454, "step": 12470 }, { "grad_norm": 0.31332147121429443, "learning_rate": 8.319333333333334e-05, "loss": 0.0343, "step": 12480 }, { "grad_norm": 0.5436424612998962, "learning_rate": 8.326e-05, "loss": 0.0409, "step": 12490 }, { "grad_norm": 0.30149906873703003, "learning_rate": 8.332666666666668e-05, "loss": 0.0473, "step": 12500 }, { "grad_norm": 0.4123278856277466, "learning_rate": 8.339333333333334e-05, "loss": 0.0563, "step": 12510 }, { "grad_norm": 0.3489752411842346, "learning_rate": 8.346e-05, "loss": 0.0428, "step": 12520 }, { "grad_norm": 0.3160141408443451, "learning_rate": 8.352666666666668e-05, "loss": 0.043, "step": 12530 }, { "grad_norm": 0.43875861167907715, "learning_rate": 8.359333333333333e-05, "loss": 0.0513, "step": 12540 }, { "grad_norm": 0.4556746780872345, "learning_rate": 8.366000000000001e-05, "loss": 0.0475, "step": 12550 }, { "grad_norm": 0.4234634339809418, "learning_rate": 8.372666666666667e-05, "loss": 0.0444, "step": 12560 }, { "grad_norm": 0.3534733057022095, "learning_rate": 8.379333333333334e-05, "loss": 0.0468, "step": 12570 }, { "grad_norm": 0.2377411276102066, "learning_rate": 8.386e-05, "loss": 0.0444, "step": 12580 }, { "grad_norm": 0.2516930401325226, "learning_rate": 8.392666666666668e-05, "loss": 0.0496, "step": 12590 }, { "grad_norm": 0.2697164714336395, "learning_rate": 8.399333333333334e-05, "loss": 0.0455, "step": 12600 }, { "grad_norm": 0.2955470681190491, "learning_rate": 8.406e-05, "loss": 0.0361, "step": 12610 }, { "grad_norm": 0.28516876697540283, "learning_rate": 8.412666666666668e-05, "loss": 0.0348, "step": 12620 }, { "grad_norm": 0.35193192958831787, "learning_rate": 8.419333333333333e-05, "loss": 0.0471, "step": 12630 }, { "grad_norm": 0.4469822645187378, "learning_rate": 8.426000000000001e-05, "loss": 0.0642, "step": 12640 }, { "grad_norm": 0.3893265426158905, "learning_rate": 8.432666666666667e-05, "loss": 0.0511, "step": 12650 }, { "grad_norm": 0.3170437216758728, "learning_rate": 8.439333333333334e-05, "loss": 0.0567, "step": 12660 }, { "grad_norm": 0.2677821218967438, "learning_rate": 8.446e-05, "loss": 0.0428, "step": 12670 }, { "grad_norm": 0.3111928403377533, "learning_rate": 8.452666666666668e-05, "loss": 0.056, "step": 12680 }, { "grad_norm": 0.3346620202064514, "learning_rate": 8.459333333333334e-05, "loss": 0.0454, "step": 12690 }, { "grad_norm": 0.34340226650238037, "learning_rate": 8.466e-05, "loss": 0.0518, "step": 12700 }, { "grad_norm": 0.4608018398284912, "learning_rate": 8.472666666666667e-05, "loss": 0.047, "step": 12710 }, { "grad_norm": 0.4004655182361603, "learning_rate": 8.479333333333333e-05, "loss": 0.0508, "step": 12720 }, { "grad_norm": 0.37994471192359924, "learning_rate": 8.486000000000001e-05, "loss": 0.0449, "step": 12730 }, { "grad_norm": 0.31565070152282715, "learning_rate": 8.492666666666666e-05, "loss": 0.0522, "step": 12740 }, { "grad_norm": 0.2661243975162506, "learning_rate": 8.499333333333334e-05, "loss": 0.0449, "step": 12750 }, { "grad_norm": 0.4546412229537964, "learning_rate": 8.506e-05, "loss": 0.0582, "step": 12760 }, { "grad_norm": 0.33549344539642334, "learning_rate": 8.512666666666667e-05, "loss": 0.0568, "step": 12770 }, { "grad_norm": 0.32294514775276184, "learning_rate": 8.519333333333333e-05, "loss": 0.0462, "step": 12780 }, { "grad_norm": 0.30821070075035095, "learning_rate": 8.526000000000001e-05, "loss": 0.0507, "step": 12790 }, { "grad_norm": 0.3600198030471802, "learning_rate": 8.532666666666667e-05, "loss": 0.0441, "step": 12800 }, { "grad_norm": 0.45987173914909363, "learning_rate": 8.539333333333333e-05, "loss": 0.0425, "step": 12810 }, { "grad_norm": 0.38794708251953125, "learning_rate": 8.546000000000001e-05, "loss": 0.0542, "step": 12820 }, { "grad_norm": 0.47257596254348755, "learning_rate": 8.552666666666666e-05, "loss": 0.0483, "step": 12830 }, { "grad_norm": 0.33403462171554565, "learning_rate": 8.559333333333334e-05, "loss": 0.0448, "step": 12840 }, { "grad_norm": 0.3769684135913849, "learning_rate": 8.566e-05, "loss": 0.0412, "step": 12850 }, { "grad_norm": 0.29167792201042175, "learning_rate": 8.572666666666667e-05, "loss": 0.0456, "step": 12860 }, { "grad_norm": 0.397522896528244, "learning_rate": 8.579333333333333e-05, "loss": 0.0457, "step": 12870 }, { "grad_norm": 0.3317742645740509, "learning_rate": 8.586000000000001e-05, "loss": 0.0422, "step": 12880 }, { "grad_norm": 0.36609262228012085, "learning_rate": 8.592666666666667e-05, "loss": 0.0409, "step": 12890 }, { "grad_norm": 0.26024141907691956, "learning_rate": 8.599333333333334e-05, "loss": 0.0471, "step": 12900 }, { "grad_norm": 0.3408243656158447, "learning_rate": 8.606000000000001e-05, "loss": 0.0419, "step": 12910 }, { "grad_norm": 0.31034207344055176, "learning_rate": 8.612666666666666e-05, "loss": 0.0386, "step": 12920 }, { "grad_norm": 0.2627377510070801, "learning_rate": 8.619333333333334e-05, "loss": 0.043, "step": 12930 }, { "grad_norm": 0.3503829538822174, "learning_rate": 8.626e-05, "loss": 0.0444, "step": 12940 }, { "grad_norm": 0.6195757389068604, "learning_rate": 8.632666666666667e-05, "loss": 0.0458, "step": 12950 }, { "grad_norm": 0.3940558433532715, "learning_rate": 8.639333333333333e-05, "loss": 0.0453, "step": 12960 }, { "grad_norm": 0.302742600440979, "learning_rate": 8.646000000000001e-05, "loss": 0.0407, "step": 12970 }, { "grad_norm": 0.30870863795280457, "learning_rate": 8.652666666666667e-05, "loss": 0.0416, "step": 12980 }, { "grad_norm": 0.3325505256652832, "learning_rate": 8.659333333333334e-05, "loss": 0.0427, "step": 12990 }, { "grad_norm": 0.2556358575820923, "learning_rate": 8.666000000000001e-05, "loss": 0.039, "step": 13000 }, { "grad_norm": 0.34519240260124207, "learning_rate": 8.672666666666666e-05, "loss": 0.0592, "step": 13010 }, { "grad_norm": 0.43413734436035156, "learning_rate": 8.679333333333334e-05, "loss": 0.0505, "step": 13020 }, { "grad_norm": 0.33071574568748474, "learning_rate": 8.686e-05, "loss": 0.0571, "step": 13030 }, { "grad_norm": 0.21014545857906342, "learning_rate": 8.692666666666667e-05, "loss": 0.0417, "step": 13040 }, { "grad_norm": 0.34843170642852783, "learning_rate": 8.699333333333333e-05, "loss": 0.0404, "step": 13050 }, { "grad_norm": 0.4310573935508728, "learning_rate": 8.706000000000001e-05, "loss": 0.0471, "step": 13060 }, { "grad_norm": 0.5594879388809204, "learning_rate": 8.712666666666667e-05, "loss": 0.0611, "step": 13070 }, { "grad_norm": 0.6783348321914673, "learning_rate": 8.719333333333334e-05, "loss": 0.0491, "step": 13080 }, { "grad_norm": 0.36615651845932007, "learning_rate": 8.726000000000001e-05, "loss": 0.0583, "step": 13090 }, { "grad_norm": 0.4033343493938446, "learning_rate": 8.732666666666667e-05, "loss": 0.0536, "step": 13100 }, { "grad_norm": 0.42007213830947876, "learning_rate": 8.739333333333334e-05, "loss": 0.0512, "step": 13110 }, { "grad_norm": 0.3000616431236267, "learning_rate": 8.746e-05, "loss": 0.044, "step": 13120 }, { "grad_norm": 0.4344441294670105, "learning_rate": 8.752666666666667e-05, "loss": 0.0517, "step": 13130 }, { "grad_norm": 0.3425014913082123, "learning_rate": 8.759333333333333e-05, "loss": 0.0494, "step": 13140 }, { "grad_norm": 0.4952804148197174, "learning_rate": 8.766000000000001e-05, "loss": 0.0458, "step": 13150 }, { "grad_norm": 0.3435705602169037, "learning_rate": 8.772666666666666e-05, "loss": 0.0375, "step": 13160 }, { "grad_norm": 0.3139479458332062, "learning_rate": 8.779333333333334e-05, "loss": 0.0404, "step": 13170 }, { "grad_norm": 0.23467333614826202, "learning_rate": 8.786e-05, "loss": 0.0439, "step": 13180 }, { "grad_norm": 0.34597036242485046, "learning_rate": 8.792666666666667e-05, "loss": 0.0396, "step": 13190 }, { "grad_norm": 0.4636770188808441, "learning_rate": 8.799333333333334e-05, "loss": 0.0449, "step": 13200 }, { "grad_norm": 0.30059635639190674, "learning_rate": 8.806000000000001e-05, "loss": 0.0388, "step": 13210 }, { "grad_norm": 0.21965917944908142, "learning_rate": 8.812666666666667e-05, "loss": 0.0348, "step": 13220 }, { "grad_norm": 0.32931777834892273, "learning_rate": 8.819333333333333e-05, "loss": 0.0445, "step": 13230 }, { "grad_norm": 0.43035176396369934, "learning_rate": 8.826000000000001e-05, "loss": 0.0645, "step": 13240 }, { "grad_norm": 0.304640531539917, "learning_rate": 8.832666666666666e-05, "loss": 0.0521, "step": 13250 }, { "grad_norm": 0.4269486665725708, "learning_rate": 8.839333333333334e-05, "loss": 0.0489, "step": 13260 }, { "grad_norm": 0.6649566888809204, "learning_rate": 8.846e-05, "loss": 0.0513, "step": 13270 }, { "grad_norm": 0.4619982838630676, "learning_rate": 8.852666666666667e-05, "loss": 0.0464, "step": 13280 }, { "grad_norm": 0.45026645064353943, "learning_rate": 8.859333333333334e-05, "loss": 0.0508, "step": 13290 }, { "grad_norm": 0.48344311118125916, "learning_rate": 8.866000000000001e-05, "loss": 0.0534, "step": 13300 }, { "grad_norm": 0.2680542469024658, "learning_rate": 8.872666666666667e-05, "loss": 0.0482, "step": 13310 }, { "grad_norm": 0.5356824398040771, "learning_rate": 8.879333333333334e-05, "loss": 0.0478, "step": 13320 }, { "grad_norm": 0.2626246213912964, "learning_rate": 8.886e-05, "loss": 0.0373, "step": 13330 }, { "grad_norm": 0.2808692455291748, "learning_rate": 8.892666666666666e-05, "loss": 0.0493, "step": 13340 }, { "grad_norm": 0.3305646479129791, "learning_rate": 8.899333333333334e-05, "loss": 0.0487, "step": 13350 }, { "grad_norm": 0.3702690005302429, "learning_rate": 8.906e-05, "loss": 0.0481, "step": 13360 }, { "grad_norm": 0.37161901593208313, "learning_rate": 8.912666666666667e-05, "loss": 0.0548, "step": 13370 }, { "grad_norm": 0.6498159766197205, "learning_rate": 8.919333333333335e-05, "loss": 0.0548, "step": 13380 }, { "grad_norm": 0.4219115376472473, "learning_rate": 8.926e-05, "loss": 0.0453, "step": 13390 }, { "grad_norm": 0.36034196615219116, "learning_rate": 8.932666666666667e-05, "loss": 0.0477, "step": 13400 }, { "grad_norm": 0.4362104833126068, "learning_rate": 8.939333333333334e-05, "loss": 0.0441, "step": 13410 }, { "grad_norm": 0.2805156707763672, "learning_rate": 8.946e-05, "loss": 0.0511, "step": 13420 }, { "grad_norm": 0.3418152630329132, "learning_rate": 8.952666666666666e-05, "loss": 0.0402, "step": 13430 }, { "grad_norm": 0.4708769619464874, "learning_rate": 8.959333333333334e-05, "loss": 0.0489, "step": 13440 }, { "grad_norm": 0.4634278118610382, "learning_rate": 8.966e-05, "loss": 0.051, "step": 13450 }, { "grad_norm": 0.2802644968032837, "learning_rate": 8.972666666666667e-05, "loss": 0.0644, "step": 13460 }, { "grad_norm": 0.4060828983783722, "learning_rate": 8.979333333333335e-05, "loss": 0.0633, "step": 13470 }, { "grad_norm": 0.30697691440582275, "learning_rate": 8.986e-05, "loss": 0.0555, "step": 13480 }, { "grad_norm": 0.32487764954566956, "learning_rate": 8.992666666666667e-05, "loss": 0.0501, "step": 13490 }, { "grad_norm": 0.32729247212409973, "learning_rate": 8.999333333333334e-05, "loss": 0.0434, "step": 13500 }, { "grad_norm": 0.3441251516342163, "learning_rate": 9.006e-05, "loss": 0.0489, "step": 13510 }, { "grad_norm": 0.37942180037498474, "learning_rate": 9.012666666666667e-05, "loss": 0.0497, "step": 13520 }, { "grad_norm": 0.38147008419036865, "learning_rate": 9.019333333333334e-05, "loss": 0.0505, "step": 13530 }, { "grad_norm": 0.32326796650886536, "learning_rate": 9.026e-05, "loss": 0.0512, "step": 13540 }, { "grad_norm": 0.31086045503616333, "learning_rate": 9.032666666666667e-05, "loss": 0.0416, "step": 13550 }, { "grad_norm": 0.28469958901405334, "learning_rate": 9.039333333333335e-05, "loss": 0.0456, "step": 13560 }, { "grad_norm": 0.23505225777626038, "learning_rate": 9.046e-05, "loss": 0.0381, "step": 13570 }, { "grad_norm": 0.432322233915329, "learning_rate": 9.052666666666668e-05, "loss": 0.0394, "step": 13580 }, { "grad_norm": 0.37779027223587036, "learning_rate": 9.059333333333334e-05, "loss": 0.054, "step": 13590 }, { "grad_norm": 0.3359115719795227, "learning_rate": 9.066e-05, "loss": 0.0541, "step": 13600 }, { "grad_norm": 0.2750062048435211, "learning_rate": 9.072666666666667e-05, "loss": 0.044, "step": 13610 }, { "grad_norm": 0.24061523377895355, "learning_rate": 9.079333333333334e-05, "loss": 0.0551, "step": 13620 }, { "grad_norm": 0.365192174911499, "learning_rate": 9.086e-05, "loss": 0.0394, "step": 13630 }, { "grad_norm": 0.31984418630599976, "learning_rate": 9.092666666666667e-05, "loss": 0.0457, "step": 13640 }, { "grad_norm": 0.31556493043899536, "learning_rate": 9.099333333333334e-05, "loss": 0.042, "step": 13650 }, { "grad_norm": 0.3478049635887146, "learning_rate": 9.106e-05, "loss": 0.0491, "step": 13660 }, { "grad_norm": 0.3928610682487488, "learning_rate": 9.112666666666668e-05, "loss": 0.0435, "step": 13670 }, { "grad_norm": 0.3007183074951172, "learning_rate": 9.119333333333334e-05, "loss": 0.0315, "step": 13680 }, { "grad_norm": 0.33486050367355347, "learning_rate": 9.126e-05, "loss": 0.0393, "step": 13690 }, { "grad_norm": 0.38190433382987976, "learning_rate": 9.132666666666667e-05, "loss": 0.0436, "step": 13700 }, { "grad_norm": 0.3488313555717468, "learning_rate": 9.139333333333334e-05, "loss": 0.0383, "step": 13710 }, { "grad_norm": 0.40111541748046875, "learning_rate": 9.146e-05, "loss": 0.0453, "step": 13720 }, { "grad_norm": 0.41274651885032654, "learning_rate": 9.152666666666667e-05, "loss": 0.043, "step": 13730 }, { "grad_norm": 0.2497245967388153, "learning_rate": 9.159333333333334e-05, "loss": 0.0381, "step": 13740 }, { "grad_norm": 0.2237052470445633, "learning_rate": 9.166e-05, "loss": 0.0364, "step": 13750 }, { "grad_norm": 0.2956802248954773, "learning_rate": 9.172666666666668e-05, "loss": 0.0606, "step": 13760 }, { "grad_norm": 0.23534967005252838, "learning_rate": 9.179333333333334e-05, "loss": 0.041, "step": 13770 }, { "grad_norm": 0.2990895211696625, "learning_rate": 9.186e-05, "loss": 0.0384, "step": 13780 }, { "grad_norm": 0.32121947407722473, "learning_rate": 9.192666666666667e-05, "loss": 0.041, "step": 13790 }, { "grad_norm": 0.2647237479686737, "learning_rate": 9.199333333333335e-05, "loss": 0.0415, "step": 13800 }, { "grad_norm": 0.3466126024723053, "learning_rate": 9.206e-05, "loss": 0.0481, "step": 13810 }, { "grad_norm": 0.4000644385814667, "learning_rate": 9.212666666666667e-05, "loss": 0.0523, "step": 13820 }, { "grad_norm": 0.2989778518676758, "learning_rate": 9.219333333333334e-05, "loss": 0.0543, "step": 13830 }, { "grad_norm": 0.5632543563842773, "learning_rate": 9.226e-05, "loss": 0.0519, "step": 13840 }, { "grad_norm": 0.4615146517753601, "learning_rate": 9.232666666666668e-05, "loss": 0.0442, "step": 13850 }, { "grad_norm": 0.28674477338790894, "learning_rate": 9.239333333333334e-05, "loss": 0.04, "step": 13860 }, { "grad_norm": 0.302936851978302, "learning_rate": 9.246e-05, "loss": 0.0423, "step": 13870 }, { "grad_norm": 0.2339416742324829, "learning_rate": 9.252666666666667e-05, "loss": 0.0383, "step": 13880 }, { "grad_norm": 0.48001423478126526, "learning_rate": 9.259333333333335e-05, "loss": 0.0529, "step": 13890 }, { "grad_norm": 0.30658647418022156, "learning_rate": 9.266e-05, "loss": 0.0341, "step": 13900 }, { "grad_norm": 0.3750864863395691, "learning_rate": 9.272666666666667e-05, "loss": 0.0426, "step": 13910 }, { "grad_norm": 0.4705195128917694, "learning_rate": 9.279333333333334e-05, "loss": 0.0481, "step": 13920 }, { "grad_norm": 0.3970586657524109, "learning_rate": 9.286e-05, "loss": 0.0478, "step": 13930 }, { "grad_norm": 0.33675795793533325, "learning_rate": 9.292666666666668e-05, "loss": 0.0445, "step": 13940 }, { "grad_norm": 0.2710345387458801, "learning_rate": 9.299333333333333e-05, "loss": 0.0425, "step": 13950 }, { "grad_norm": 0.2930898070335388, "learning_rate": 9.306000000000001e-05, "loss": 0.038, "step": 13960 }, { "grad_norm": 0.36458659172058105, "learning_rate": 9.312666666666667e-05, "loss": 0.0634, "step": 13970 }, { "grad_norm": 0.5851505398750305, "learning_rate": 9.319333333333333e-05, "loss": 0.0495, "step": 13980 }, { "grad_norm": 0.3511316478252411, "learning_rate": 9.326e-05, "loss": 0.0429, "step": 13990 }, { "grad_norm": 0.3360661566257477, "learning_rate": 9.332666666666668e-05, "loss": 0.0443, "step": 14000 }, { "grad_norm": 0.3078964352607727, "learning_rate": 9.339333333333334e-05, "loss": 0.0455, "step": 14010 }, { "grad_norm": 0.28956735134124756, "learning_rate": 9.346e-05, "loss": 0.0477, "step": 14020 }, { "grad_norm": 0.3057316839694977, "learning_rate": 9.352666666666667e-05, "loss": 0.0444, "step": 14030 }, { "grad_norm": 0.3384071886539459, "learning_rate": 9.359333333333333e-05, "loss": 0.0446, "step": 14040 }, { "grad_norm": 0.2896701693534851, "learning_rate": 9.366000000000001e-05, "loss": 0.042, "step": 14050 }, { "grad_norm": 0.35228875279426575, "learning_rate": 9.372666666666667e-05, "loss": 0.0442, "step": 14060 }, { "grad_norm": 0.3774707317352295, "learning_rate": 9.379333333333334e-05, "loss": 0.0518, "step": 14070 }, { "grad_norm": 0.682754397392273, "learning_rate": 9.386e-05, "loss": 0.0453, "step": 14080 }, { "grad_norm": 0.302776038646698, "learning_rate": 9.392666666666668e-05, "loss": 0.0408, "step": 14090 }, { "grad_norm": 0.28177186846733093, "learning_rate": 9.399333333333333e-05, "loss": 0.0464, "step": 14100 }, { "grad_norm": 0.2586699426174164, "learning_rate": 9.406e-05, "loss": 0.0492, "step": 14110 }, { "grad_norm": 0.36159875988960266, "learning_rate": 9.412666666666667e-05, "loss": 0.0432, "step": 14120 }, { "grad_norm": 0.4022761881351471, "learning_rate": 9.419333333333333e-05, "loss": 0.0501, "step": 14130 }, { "grad_norm": 0.3591936528682709, "learning_rate": 9.426000000000001e-05, "loss": 0.0443, "step": 14140 }, { "grad_norm": 0.25135552883148193, "learning_rate": 9.432666666666667e-05, "loss": 0.0477, "step": 14150 }, { "grad_norm": 0.4857410788536072, "learning_rate": 9.439333333333334e-05, "loss": 0.0535, "step": 14160 }, { "grad_norm": 0.29195353388786316, "learning_rate": 9.446e-05, "loss": 0.0458, "step": 14170 }, { "grad_norm": 0.5577927827835083, "learning_rate": 9.452666666666668e-05, "loss": 0.0451, "step": 14180 }, { "grad_norm": 0.3900012671947479, "learning_rate": 9.459333333333333e-05, "loss": 0.0557, "step": 14190 }, { "grad_norm": 0.3220136761665344, "learning_rate": 9.466e-05, "loss": 0.0466, "step": 14200 }, { "grad_norm": 0.34323766827583313, "learning_rate": 9.472666666666667e-05, "loss": 0.0375, "step": 14210 }, { "grad_norm": 0.31792646646499634, "learning_rate": 9.479333333333333e-05, "loss": 0.0431, "step": 14220 }, { "grad_norm": 0.29047542810440063, "learning_rate": 9.486000000000001e-05, "loss": 0.0392, "step": 14230 }, { "grad_norm": 0.4559662640094757, "learning_rate": 9.492666666666667e-05, "loss": 0.0416, "step": 14240 }, { "grad_norm": 0.23125441372394562, "learning_rate": 9.499333333333334e-05, "loss": 0.0492, "step": 14250 }, { "grad_norm": 0.22129608690738678, "learning_rate": 9.506e-05, "loss": 0.0435, "step": 14260 }, { "grad_norm": 0.34282585978507996, "learning_rate": 9.512666666666668e-05, "loss": 0.0499, "step": 14270 }, { "grad_norm": 0.3883557617664337, "learning_rate": 9.519333333333333e-05, "loss": 0.0419, "step": 14280 }, { "grad_norm": 0.2920350432395935, "learning_rate": 9.526e-05, "loss": 0.0355, "step": 14290 }, { "grad_norm": 0.31243911385536194, "learning_rate": 9.532666666666667e-05, "loss": 0.0472, "step": 14300 }, { "grad_norm": 0.25533121824264526, "learning_rate": 9.539333333333333e-05, "loss": 0.0338, "step": 14310 }, { "grad_norm": 0.25259193778038025, "learning_rate": 9.546000000000001e-05, "loss": 0.0525, "step": 14320 }, { "grad_norm": 0.22628849744796753, "learning_rate": 9.552666666666667e-05, "loss": 0.0354, "step": 14330 }, { "grad_norm": 0.2942354679107666, "learning_rate": 9.559333333333334e-05, "loss": 0.0377, "step": 14340 }, { "grad_norm": 0.5187931656837463, "learning_rate": 9.566e-05, "loss": 0.0543, "step": 14350 }, { "grad_norm": 0.307882696390152, "learning_rate": 9.572666666666668e-05, "loss": 0.049, "step": 14360 }, { "grad_norm": 0.27464988827705383, "learning_rate": 9.579333333333333e-05, "loss": 0.0519, "step": 14370 }, { "grad_norm": 0.21231116354465485, "learning_rate": 9.586000000000001e-05, "loss": 0.048, "step": 14380 }, { "grad_norm": 0.353622704744339, "learning_rate": 9.592666666666667e-05, "loss": 0.0502, "step": 14390 }, { "grad_norm": 0.2593512237071991, "learning_rate": 9.599333333333333e-05, "loss": 0.0592, "step": 14400 }, { "grad_norm": 0.31772568821907043, "learning_rate": 9.606000000000001e-05, "loss": 0.0455, "step": 14410 }, { "grad_norm": 0.23391170799732208, "learning_rate": 9.612666666666668e-05, "loss": 0.0406, "step": 14420 }, { "grad_norm": 0.3464439809322357, "learning_rate": 9.619333333333334e-05, "loss": 0.045, "step": 14430 }, { "grad_norm": 0.2957136034965515, "learning_rate": 9.626e-05, "loss": 0.038, "step": 14440 }, { "grad_norm": 0.34005850553512573, "learning_rate": 9.632666666666668e-05, "loss": 0.0383, "step": 14450 }, { "grad_norm": 0.21204476058483124, "learning_rate": 9.639333333333333e-05, "loss": 0.0406, "step": 14460 }, { "grad_norm": 0.4148279130458832, "learning_rate": 9.646000000000001e-05, "loss": 0.0484, "step": 14470 }, { "grad_norm": 0.3464168310165405, "learning_rate": 9.652666666666667e-05, "loss": 0.0437, "step": 14480 }, { "grad_norm": 0.3369270861148834, "learning_rate": 9.659333333333334e-05, "loss": 0.0396, "step": 14490 }, { "grad_norm": 0.6012405157089233, "learning_rate": 9.666e-05, "loss": 0.0545, "step": 14500 }, { "grad_norm": 0.324474036693573, "learning_rate": 9.672666666666668e-05, "loss": 0.0544, "step": 14510 }, { "grad_norm": 0.38853365182876587, "learning_rate": 9.679333333333334e-05, "loss": 0.0478, "step": 14520 }, { "grad_norm": 0.5080093145370483, "learning_rate": 9.686e-05, "loss": 0.0652, "step": 14530 }, { "grad_norm": 0.26731589436531067, "learning_rate": 9.692666666666668e-05, "loss": 0.0428, "step": 14540 }, { "grad_norm": 0.3182070553302765, "learning_rate": 9.699333333333333e-05, "loss": 0.0393, "step": 14550 }, { "grad_norm": 0.4099366366863251, "learning_rate": 9.706000000000001e-05, "loss": 0.0365, "step": 14560 }, { "grad_norm": 0.5324279069900513, "learning_rate": 9.712666666666666e-05, "loss": 0.0369, "step": 14570 }, { "grad_norm": 0.28589481115341187, "learning_rate": 9.719333333333334e-05, "loss": 0.0458, "step": 14580 }, { "grad_norm": 0.2694368064403534, "learning_rate": 9.726e-05, "loss": 0.0555, "step": 14590 }, { "grad_norm": 0.40097618103027344, "learning_rate": 9.732666666666666e-05, "loss": 0.0541, "step": 14600 }, { "grad_norm": 0.28370201587677, "learning_rate": 9.739333333333334e-05, "loss": 0.0499, "step": 14610 }, { "grad_norm": 0.2991393208503723, "learning_rate": 9.746e-05, "loss": 0.0416, "step": 14620 }, { "grad_norm": 0.33803170919418335, "learning_rate": 9.752666666666667e-05, "loss": 0.0415, "step": 14630 }, { "grad_norm": 0.3152802586555481, "learning_rate": 9.759333333333333e-05, "loss": 0.0437, "step": 14640 }, { "grad_norm": 0.37789008021354675, "learning_rate": 9.766000000000001e-05, "loss": 0.0426, "step": 14650 }, { "grad_norm": 0.2465018928050995, "learning_rate": 9.772666666666666e-05, "loss": 0.0461, "step": 14660 }, { "grad_norm": 0.4038486182689667, "learning_rate": 9.779333333333334e-05, "loss": 0.0481, "step": 14670 }, { "grad_norm": 0.3209993243217468, "learning_rate": 9.786e-05, "loss": 0.0413, "step": 14680 }, { "grad_norm": 0.26994428038597107, "learning_rate": 9.792666666666667e-05, "loss": 0.0437, "step": 14690 }, { "grad_norm": 0.3100738525390625, "learning_rate": 9.799333333333334e-05, "loss": 0.0444, "step": 14700 }, { "grad_norm": 0.2817119359970093, "learning_rate": 9.806e-05, "loss": 0.0396, "step": 14710 }, { "grad_norm": 0.29911544919013977, "learning_rate": 9.812666666666667e-05, "loss": 0.0443, "step": 14720 }, { "grad_norm": 0.31316128373146057, "learning_rate": 9.819333333333333e-05, "loss": 0.0461, "step": 14730 }, { "grad_norm": 0.2141825258731842, "learning_rate": 9.826000000000001e-05, "loss": 0.0484, "step": 14740 }, { "grad_norm": 0.3191637098789215, "learning_rate": 9.832666666666666e-05, "loss": 0.046, "step": 14750 }, { "grad_norm": 0.25518521666526794, "learning_rate": 9.839333333333334e-05, "loss": 0.0427, "step": 14760 }, { "grad_norm": 0.30536559224128723, "learning_rate": 9.846e-05, "loss": 0.0421, "step": 14770 }, { "grad_norm": 0.2884780168533325, "learning_rate": 9.852666666666667e-05, "loss": 0.047, "step": 14780 }, { "grad_norm": 0.3140869438648224, "learning_rate": 9.859333333333334e-05, "loss": 0.0349, "step": 14790 }, { "grad_norm": 0.3745141923427582, "learning_rate": 9.866000000000001e-05, "loss": 0.0404, "step": 14800 }, { "grad_norm": 0.28711751103401184, "learning_rate": 9.872666666666667e-05, "loss": 0.0345, "step": 14810 }, { "grad_norm": 0.2552744746208191, "learning_rate": 9.879333333333333e-05, "loss": 0.0509, "step": 14820 }, { "grad_norm": 0.2492281049489975, "learning_rate": 9.886000000000001e-05, "loss": 0.0512, "step": 14830 }, { "grad_norm": 0.7613804936408997, "learning_rate": 9.892666666666666e-05, "loss": 0.0717, "step": 14840 }, { "grad_norm": 0.49776187539100647, "learning_rate": 9.899333333333334e-05, "loss": 0.0661, "step": 14850 }, { "grad_norm": 0.33290228247642517, "learning_rate": 9.906e-05, "loss": 0.046, "step": 14860 }, { "grad_norm": 0.23758551478385925, "learning_rate": 9.912666666666667e-05, "loss": 0.0399, "step": 14870 }, { "grad_norm": 0.3027785122394562, "learning_rate": 9.919333333333334e-05, "loss": 0.0406, "step": 14880 }, { "grad_norm": 0.31880998611450195, "learning_rate": 9.926000000000001e-05, "loss": 0.0434, "step": 14890 }, { "grad_norm": 0.3360535204410553, "learning_rate": 9.932666666666667e-05, "loss": 0.0428, "step": 14900 }, { "grad_norm": 0.3087815046310425, "learning_rate": 9.939333333333334e-05, "loss": 0.0435, "step": 14910 }, { "grad_norm": 0.28057000041007996, "learning_rate": 9.946000000000001e-05, "loss": 0.0441, "step": 14920 }, { "grad_norm": 0.3666330873966217, "learning_rate": 9.952666666666666e-05, "loss": 0.0461, "step": 14930 }, { "grad_norm": 0.30293580889701843, "learning_rate": 9.959333333333334e-05, "loss": 0.0468, "step": 14940 }, { "grad_norm": 0.46934422850608826, "learning_rate": 9.966e-05, "loss": 0.0566, "step": 14950 }, { "grad_norm": 0.26078009605407715, "learning_rate": 9.972666666666667e-05, "loss": 0.0454, "step": 14960 }, { "grad_norm": 0.3218095600605011, "learning_rate": 9.979333333333333e-05, "loss": 0.0437, "step": 14970 }, { "grad_norm": 0.27861306071281433, "learning_rate": 9.986000000000001e-05, "loss": 0.0541, "step": 14980 }, { "grad_norm": 0.3864833116531372, "learning_rate": 9.992666666666667e-05, "loss": 0.0462, "step": 14990 }, { "grad_norm": 0.34605708718299866, "learning_rate": 9.999333333333334e-05, "loss": 0.0548, "step": 15000 }, { "grad_norm": 0.37775546312332153, "learning_rate": 9.999999975394338e-05, "loss": 0.044, "step": 15010 }, { "grad_norm": 0.3127991557121277, "learning_rate": 9.999999890337731e-05, "loss": 0.04, "step": 15020 }, { "grad_norm": 0.3761783242225647, "learning_rate": 9.999999744526401e-05, "loss": 0.0421, "step": 15030 }, { "grad_norm": 0.3520197868347168, "learning_rate": 9.999999537960356e-05, "loss": 0.0525, "step": 15040 }, { "grad_norm": 0.26383763551712036, "learning_rate": 9.999999270639596e-05, "loss": 0.0578, "step": 15050 }, { "grad_norm": 0.31270983815193176, "learning_rate": 9.999998942564122e-05, "loss": 0.0439, "step": 15060 }, { "grad_norm": 0.2680501937866211, "learning_rate": 9.99999855373394e-05, "loss": 0.0417, "step": 15070 }, { "grad_norm": 0.36351361870765686, "learning_rate": 9.999998104149056e-05, "loss": 0.0479, "step": 15080 }, { "grad_norm": 0.35188138484954834, "learning_rate": 9.999997593809473e-05, "loss": 0.0452, "step": 15090 }, { "grad_norm": 0.30796200037002563, "learning_rate": 9.999997022715201e-05, "loss": 0.0431, "step": 15100 }, { "grad_norm": 0.27080193161964417, "learning_rate": 9.999996390866243e-05, "loss": 0.0364, "step": 15110 }, { "grad_norm": 0.2570839822292328, "learning_rate": 9.999995698262608e-05, "loss": 0.032, "step": 15120 }, { "grad_norm": 0.31006792187690735, "learning_rate": 9.999994944904306e-05, "loss": 0.0446, "step": 15130 }, { "grad_norm": 0.44584202766418457, "learning_rate": 9.999994130791341e-05, "loss": 0.0416, "step": 15140 }, { "grad_norm": 0.36049649119377136, "learning_rate": 9.99999325592373e-05, "loss": 0.0562, "step": 15150 }, { "grad_norm": 0.2569650411605835, "learning_rate": 9.999992320301478e-05, "loss": 0.0504, "step": 15160 }, { "grad_norm": 0.32674089074134827, "learning_rate": 9.9999913239246e-05, "loss": 0.054, "step": 15170 }, { "grad_norm": 0.4529579281806946, "learning_rate": 9.999990266793106e-05, "loss": 0.0419, "step": 15180 }, { "grad_norm": 0.21856366097927094, "learning_rate": 9.99998914890701e-05, "loss": 0.0402, "step": 15190 }, { "grad_norm": 0.3566026985645294, "learning_rate": 9.999987970266323e-05, "loss": 0.0532, "step": 15200 }, { "grad_norm": 0.24084022641181946, "learning_rate": 9.999986730871064e-05, "loss": 0.0397, "step": 15210 }, { "grad_norm": 0.5531561374664307, "learning_rate": 9.999985430721243e-05, "loss": 0.0455, "step": 15220 }, { "grad_norm": 0.24657200276851654, "learning_rate": 9.999984069816881e-05, "loss": 0.042, "step": 15230 }, { "grad_norm": 0.2800348103046417, "learning_rate": 9.99998264815799e-05, "loss": 0.0371, "step": 15240 }, { "grad_norm": 0.3029840588569641, "learning_rate": 9.99998116574459e-05, "loss": 0.0557, "step": 15250 }, { "grad_norm": 0.3199366629123688, "learning_rate": 9.999979622576698e-05, "loss": 0.0398, "step": 15260 }, { "grad_norm": 0.3078370988368988, "learning_rate": 9.999978018654333e-05, "loss": 0.0451, "step": 15270 }, { "grad_norm": 0.4331963360309601, "learning_rate": 9.999976353977511e-05, "loss": 0.043, "step": 15280 }, { "grad_norm": 0.32901284098625183, "learning_rate": 9.99997462854626e-05, "loss": 0.0532, "step": 15290 }, { "grad_norm": 0.2557967007160187, "learning_rate": 9.999972842360594e-05, "loss": 0.0426, "step": 15300 }, { "grad_norm": 0.23101848363876343, "learning_rate": 9.999970995420537e-05, "loss": 0.0521, "step": 15310 }, { "grad_norm": 0.2581081688404083, "learning_rate": 9.999969087726114e-05, "loss": 0.0388, "step": 15320 }, { "grad_norm": 0.32696184515953064, "learning_rate": 9.999967119277344e-05, "loss": 0.0349, "step": 15330 }, { "grad_norm": 0.23384425044059753, "learning_rate": 9.999965090074254e-05, "loss": 0.0408, "step": 15340 }, { "grad_norm": 0.20590297877788544, "learning_rate": 9.999963000116865e-05, "loss": 0.036, "step": 15350 }, { "grad_norm": 0.19785937666893005, "learning_rate": 9.999960849405208e-05, "loss": 0.0329, "step": 15360 }, { "grad_norm": 0.34093090891838074, "learning_rate": 9.999958637939304e-05, "loss": 0.0396, "step": 15370 }, { "grad_norm": 0.23772752285003662, "learning_rate": 9.999956365719182e-05, "loss": 0.0459, "step": 15380 }, { "grad_norm": 0.4114398658275604, "learning_rate": 9.99995403274487e-05, "loss": 0.05, "step": 15390 }, { "grad_norm": 0.3050225079059601, "learning_rate": 9.999951639016395e-05, "loss": 0.0407, "step": 15400 }, { "grad_norm": 0.36645546555519104, "learning_rate": 9.999949184533788e-05, "loss": 0.0438, "step": 15410 }, { "grad_norm": 0.26522567868232727, "learning_rate": 9.999946669297076e-05, "loss": 0.0394, "step": 15420 }, { "grad_norm": 0.2895286977291107, "learning_rate": 9.999944093306293e-05, "loss": 0.0543, "step": 15430 }, { "grad_norm": 0.2555806040763855, "learning_rate": 9.999941456561465e-05, "loss": 0.0397, "step": 15440 }, { "grad_norm": 0.2372630536556244, "learning_rate": 9.999938759062633e-05, "loss": 0.0427, "step": 15450 }, { "grad_norm": 0.22106917202472687, "learning_rate": 9.999936000809819e-05, "loss": 0.0378, "step": 15460 }, { "grad_norm": 0.39556971192359924, "learning_rate": 9.999933181803066e-05, "loss": 0.0467, "step": 15470 }, { "grad_norm": 0.3685555160045624, "learning_rate": 9.999930302042402e-05, "loss": 0.0495, "step": 15480 }, { "grad_norm": 0.26780229806900024, "learning_rate": 9.999927361527866e-05, "loss": 0.0458, "step": 15490 }, { "grad_norm": 0.2076219618320465, "learning_rate": 9.99992436025949e-05, "loss": 0.0347, "step": 15500 }, { "grad_norm": 0.3104015290737152, "learning_rate": 9.999921298237311e-05, "loss": 0.0442, "step": 15510 }, { "grad_norm": 0.2672710716724396, "learning_rate": 9.999918175461371e-05, "loss": 0.0357, "step": 15520 }, { "grad_norm": 0.2901015281677246, "learning_rate": 9.999914991931702e-05, "loss": 0.0384, "step": 15530 }, { "grad_norm": 0.26111793518066406, "learning_rate": 9.999911747648346e-05, "loss": 0.04, "step": 15540 }, { "grad_norm": 0.38415083289146423, "learning_rate": 9.999908442611341e-05, "loss": 0.0472, "step": 15550 }, { "grad_norm": 0.31036269664764404, "learning_rate": 9.999905076820728e-05, "loss": 0.0401, "step": 15560 }, { "grad_norm": 0.40464305877685547, "learning_rate": 9.999901650276549e-05, "loss": 0.0422, "step": 15570 }, { "grad_norm": 0.35887283086776733, "learning_rate": 9.999898162978843e-05, "loss": 0.0504, "step": 15580 }, { "grad_norm": 0.2992149591445923, "learning_rate": 9.999894614927653e-05, "loss": 0.0441, "step": 15590 }, { "grad_norm": 0.33645546436309814, "learning_rate": 9.999891006123023e-05, "loss": 0.0396, "step": 15600 }, { "grad_norm": 0.28635862469673157, "learning_rate": 9.999887336564998e-05, "loss": 0.0505, "step": 15610 }, { "grad_norm": 0.19122131168842316, "learning_rate": 9.99988360625362e-05, "loss": 0.0394, "step": 15620 }, { "grad_norm": 0.2753975987434387, "learning_rate": 9.999879815188936e-05, "loss": 0.0385, "step": 15630 }, { "grad_norm": 0.2294575572013855, "learning_rate": 9.999875963370991e-05, "loss": 0.0298, "step": 15640 }, { "grad_norm": 0.3972071707248688, "learning_rate": 9.999872050799833e-05, "loss": 0.0348, "step": 15650 }, { "grad_norm": 0.29899394512176514, "learning_rate": 9.999868077475509e-05, "loss": 0.0456, "step": 15660 }, { "grad_norm": 0.41986873745918274, "learning_rate": 9.999864043398066e-05, "loss": 0.0428, "step": 15670 }, { "grad_norm": 0.24362899363040924, "learning_rate": 9.999859948567557e-05, "loss": 0.0345, "step": 15680 }, { "grad_norm": 0.26270779967308044, "learning_rate": 9.999855792984025e-05, "loss": 0.0366, "step": 15690 }, { "grad_norm": 0.38183850049972534, "learning_rate": 9.999851576647528e-05, "loss": 0.0425, "step": 15700 }, { "grad_norm": 0.22870256006717682, "learning_rate": 9.999847299558111e-05, "loss": 0.0417, "step": 15710 }, { "grad_norm": 0.3051847219467163, "learning_rate": 9.999842961715831e-05, "loss": 0.0337, "step": 15720 }, { "grad_norm": 0.25095948576927185, "learning_rate": 9.999838563120737e-05, "loss": 0.047, "step": 15730 }, { "grad_norm": 0.28289875388145447, "learning_rate": 9.999834103772883e-05, "loss": 0.0447, "step": 15740 }, { "grad_norm": 0.3887227773666382, "learning_rate": 9.999829583672327e-05, "loss": 0.0481, "step": 15750 }, { "grad_norm": 0.38378050923347473, "learning_rate": 9.999825002819119e-05, "loss": 0.0519, "step": 15760 }, { "grad_norm": 0.3715156018733978, "learning_rate": 9.999820361213316e-05, "loss": 0.0658, "step": 15770 }, { "grad_norm": 0.26420268416404724, "learning_rate": 9.999815658854976e-05, "loss": 0.0502, "step": 15780 }, { "grad_norm": 0.38660597801208496, "learning_rate": 9.999810895744155e-05, "loss": 0.0561, "step": 15790 }, { "grad_norm": 0.26640015840530396, "learning_rate": 9.99980607188091e-05, "loss": 0.0451, "step": 15800 }, { "grad_norm": 0.3188575208187103, "learning_rate": 9.999801187265301e-05, "loss": 0.0368, "step": 15810 }, { "grad_norm": 0.27696704864501953, "learning_rate": 9.999796241897387e-05, "loss": 0.0384, "step": 15820 }, { "grad_norm": 0.23083822429180145, "learning_rate": 9.99979123577723e-05, "loss": 0.0365, "step": 15830 }, { "grad_norm": 0.26908913254737854, "learning_rate": 9.999786168904886e-05, "loss": 0.0364, "step": 15840 }, { "grad_norm": 0.2942621409893036, "learning_rate": 9.999781041280422e-05, "loss": 0.0326, "step": 15850 }, { "grad_norm": 0.24935857951641083, "learning_rate": 9.999775852903896e-05, "loss": 0.0403, "step": 15860 }, { "grad_norm": 0.3299292027950287, "learning_rate": 9.999770603775373e-05, "loss": 0.0406, "step": 15870 }, { "grad_norm": 0.3235832452774048, "learning_rate": 9.999765293894918e-05, "loss": 0.0491, "step": 15880 }, { "grad_norm": 0.27698129415512085, "learning_rate": 9.999759923262593e-05, "loss": 0.0421, "step": 15890 }, { "grad_norm": 0.32863378524780273, "learning_rate": 9.999754491878464e-05, "loss": 0.0385, "step": 15900 }, { "grad_norm": 0.2829230725765228, "learning_rate": 9.999748999742597e-05, "loss": 0.0413, "step": 15910 }, { "grad_norm": 0.3624000549316406, "learning_rate": 9.999743446855061e-05, "loss": 0.0401, "step": 15920 }, { "grad_norm": 0.3127167522907257, "learning_rate": 9.99973783321592e-05, "loss": 0.0434, "step": 15930 }, { "grad_norm": 0.3064109981060028, "learning_rate": 9.999732158825246e-05, "loss": 0.0446, "step": 15940 }, { "grad_norm": 0.31683605909347534, "learning_rate": 9.999726423683104e-05, "loss": 0.0389, "step": 15950 }, { "grad_norm": 0.25086134672164917, "learning_rate": 9.999720627789566e-05, "loss": 0.0375, "step": 15960 }, { "grad_norm": 0.2557474672794342, "learning_rate": 9.999714771144701e-05, "loss": 0.0414, "step": 15970 }, { "grad_norm": 0.29680126905441284, "learning_rate": 9.99970885374858e-05, "loss": 0.0365, "step": 15980 }, { "grad_norm": 0.2749324142932892, "learning_rate": 9.999702875601277e-05, "loss": 0.0409, "step": 15990 }, { "grad_norm": 0.28331515192985535, "learning_rate": 9.999696836702865e-05, "loss": 0.0385, "step": 16000 }, { "grad_norm": 0.275687575340271, "learning_rate": 9.999690737053416e-05, "loss": 0.0413, "step": 16010 }, { "grad_norm": 0.20789791643619537, "learning_rate": 9.999684576653001e-05, "loss": 0.0323, "step": 16020 }, { "grad_norm": 0.24526657164096832, "learning_rate": 9.999678355501701e-05, "loss": 0.0357, "step": 16030 }, { "grad_norm": 0.2006673365831375, "learning_rate": 9.999672073599586e-05, "loss": 0.039, "step": 16040 }, { "grad_norm": 0.22554539144039154, "learning_rate": 9.999665730946736e-05, "loss": 0.0397, "step": 16050 }, { "grad_norm": 0.2343561202287674, "learning_rate": 9.999659327543228e-05, "loss": 0.0343, "step": 16060 }, { "grad_norm": 0.29732346534729004, "learning_rate": 9.999652863389138e-05, "loss": 0.0399, "step": 16070 }, { "grad_norm": 0.26198408007621765, "learning_rate": 9.999646338484545e-05, "loss": 0.0338, "step": 16080 }, { "grad_norm": 0.24667471647262573, "learning_rate": 9.999639752829528e-05, "loss": 0.0382, "step": 16090 }, { "grad_norm": 0.22436180710792542, "learning_rate": 9.999633106424169e-05, "loss": 0.0322, "step": 16100 }, { "grad_norm": 0.44811302423477173, "learning_rate": 9.999626399268546e-05, "loss": 0.0413, "step": 16110 }, { "grad_norm": 0.3056691884994507, "learning_rate": 9.999619631362742e-05, "loss": 0.0404, "step": 16120 }, { "grad_norm": 0.4003026485443115, "learning_rate": 9.99961280270684e-05, "loss": 0.0395, "step": 16130 }, { "grad_norm": 0.25052469968795776, "learning_rate": 9.99960591330092e-05, "loss": 0.0311, "step": 16140 }, { "grad_norm": 0.21078264713287354, "learning_rate": 9.999598963145069e-05, "loss": 0.0438, "step": 16150 }, { "grad_norm": 0.21413163840770721, "learning_rate": 9.999591952239371e-05, "loss": 0.0372, "step": 16160 }, { "grad_norm": 0.35456612706184387, "learning_rate": 9.99958488058391e-05, "loss": 0.0333, "step": 16170 }, { "grad_norm": 0.2027580589056015, "learning_rate": 9.999577748178773e-05, "loss": 0.0287, "step": 16180 }, { "grad_norm": 0.32975316047668457, "learning_rate": 9.999570555024043e-05, "loss": 0.0338, "step": 16190 }, { "grad_norm": 0.21101340651512146, "learning_rate": 9.999563301119811e-05, "loss": 0.0343, "step": 16200 }, { "grad_norm": 0.2089070975780487, "learning_rate": 9.999555986466166e-05, "loss": 0.0404, "step": 16210 }, { "grad_norm": 0.20011551678180695, "learning_rate": 9.999548611063197e-05, "loss": 0.0277, "step": 16220 }, { "grad_norm": 0.24442210793495178, "learning_rate": 9.99954117491099e-05, "loss": 0.0456, "step": 16230 }, { "grad_norm": 0.43935447931289673, "learning_rate": 9.999533678009639e-05, "loss": 0.0515, "step": 16240 }, { "grad_norm": 0.3771170973777771, "learning_rate": 9.99952612035923e-05, "loss": 0.0424, "step": 16250 }, { "grad_norm": 0.27627551555633545, "learning_rate": 9.999518501959861e-05, "loss": 0.0402, "step": 16260 }, { "grad_norm": 0.19461499154567719, "learning_rate": 9.999510822811621e-05, "loss": 0.0358, "step": 16270 }, { "grad_norm": 0.27661123871803284, "learning_rate": 9.999503082914605e-05, "loss": 0.0457, "step": 16280 }, { "grad_norm": 0.18675652146339417, "learning_rate": 9.999495282268905e-05, "loss": 0.0321, "step": 16290 }, { "grad_norm": 0.2239987999200821, "learning_rate": 9.999487420874618e-05, "loss": 0.0331, "step": 16300 }, { "grad_norm": 0.510857105255127, "learning_rate": 9.999479498731839e-05, "loss": 0.0552, "step": 16310 }, { "grad_norm": 0.2779892086982727, "learning_rate": 9.999471515840662e-05, "loss": 0.0396, "step": 16320 }, { "grad_norm": 0.31209367513656616, "learning_rate": 9.999463472201187e-05, "loss": 0.0489, "step": 16330 }, { "grad_norm": 0.40470901131629944, "learning_rate": 9.999455367813511e-05, "loss": 0.0411, "step": 16340 }, { "grad_norm": 0.289622962474823, "learning_rate": 9.999447202677732e-05, "loss": 0.0381, "step": 16350 }, { "grad_norm": 0.2920103967189789, "learning_rate": 9.999438976793947e-05, "loss": 0.0435, "step": 16360 }, { "grad_norm": 0.3592522144317627, "learning_rate": 9.999430690162261e-05, "loss": 0.0538, "step": 16370 }, { "grad_norm": 0.2698046565055847, "learning_rate": 9.99942234278277e-05, "loss": 0.0487, "step": 16380 }, { "grad_norm": 0.34814441204071045, "learning_rate": 9.999413934655578e-05, "loss": 0.049, "step": 16390 }, { "grad_norm": 0.28780627250671387, "learning_rate": 9.999405465780787e-05, "loss": 0.0413, "step": 16400 }, { "grad_norm": 0.28645089268684387, "learning_rate": 9.999396936158498e-05, "loss": 0.0407, "step": 16410 }, { "grad_norm": 0.2569422721862793, "learning_rate": 9.999388345788819e-05, "loss": 0.0376, "step": 16420 }, { "grad_norm": 0.25943005084991455, "learning_rate": 9.999379694671847e-05, "loss": 0.0389, "step": 16430 }, { "grad_norm": 0.5364888906478882, "learning_rate": 9.999370982807695e-05, "loss": 0.041, "step": 16440 }, { "grad_norm": 0.22712084650993347, "learning_rate": 9.999362210196463e-05, "loss": 0.0348, "step": 16450 }, { "grad_norm": 0.22419314086437225, "learning_rate": 9.99935337683826e-05, "loss": 0.0422, "step": 16460 }, { "grad_norm": 0.2729296386241913, "learning_rate": 9.999344482733197e-05, "loss": 0.0449, "step": 16470 }, { "grad_norm": 0.27146151661872864, "learning_rate": 9.999335527881375e-05, "loss": 0.0451, "step": 16480 }, { "grad_norm": 0.22878769040107727, "learning_rate": 9.999326512282908e-05, "loss": 0.0377, "step": 16490 }, { "grad_norm": 0.3179877996444702, "learning_rate": 9.9993174359379e-05, "loss": 0.0378, "step": 16500 }, { "grad_norm": 0.2965451180934906, "learning_rate": 9.999308298846468e-05, "loss": 0.0417, "step": 16510 }, { "grad_norm": 0.24574986100196838, "learning_rate": 9.999299101008721e-05, "loss": 0.04, "step": 16520 }, { "grad_norm": 0.30706587433815, "learning_rate": 9.999289842424768e-05, "loss": 0.0452, "step": 16530 }, { "grad_norm": 0.27486005425453186, "learning_rate": 9.999280523094724e-05, "loss": 0.0375, "step": 16540 }, { "grad_norm": 0.3510551452636719, "learning_rate": 9.999271143018702e-05, "loss": 0.0382, "step": 16550 }, { "grad_norm": 0.40253689885139465, "learning_rate": 9.999261702196813e-05, "loss": 0.0522, "step": 16560 }, { "grad_norm": 0.5438937544822693, "learning_rate": 9.999252200629176e-05, "loss": 0.0441, "step": 16570 }, { "grad_norm": 0.3076435327529907, "learning_rate": 9.999242638315906e-05, "loss": 0.0549, "step": 16580 }, { "grad_norm": 0.2488473653793335, "learning_rate": 9.999233015257116e-05, "loss": 0.041, "step": 16590 }, { "grad_norm": 0.2344466596841812, "learning_rate": 9.999223331452926e-05, "loss": 0.0348, "step": 16600 }, { "grad_norm": 0.19511371850967407, "learning_rate": 9.999213586903452e-05, "loss": 0.0336, "step": 16610 }, { "grad_norm": 0.3779330551624298, "learning_rate": 9.999203781608813e-05, "loss": 0.0353, "step": 16620 }, { "grad_norm": 0.27275359630584717, "learning_rate": 9.999193915569127e-05, "loss": 0.0365, "step": 16630 }, { "grad_norm": 0.33497050404548645, "learning_rate": 9.999183988784516e-05, "loss": 0.0422, "step": 16640 }, { "grad_norm": 0.18636590242385864, "learning_rate": 9.999174001255099e-05, "loss": 0.0405, "step": 16650 }, { "grad_norm": 0.27898314595222473, "learning_rate": 9.999163952981e-05, "loss": 0.035, "step": 16660 }, { "grad_norm": 0.24622610211372375, "learning_rate": 9.999153843962336e-05, "loss": 0.0519, "step": 16670 }, { "grad_norm": 0.32771316170692444, "learning_rate": 9.999143674199234e-05, "loss": 0.0358, "step": 16680 }, { "grad_norm": 0.3232068121433258, "learning_rate": 9.999133443691817e-05, "loss": 0.0359, "step": 16690 }, { "grad_norm": 0.2704029381275177, "learning_rate": 9.999123152440208e-05, "loss": 0.0388, "step": 16700 }, { "grad_norm": 0.3358702063560486, "learning_rate": 9.999112800444533e-05, "loss": 0.0417, "step": 16710 }, { "grad_norm": 0.35878831148147583, "learning_rate": 9.999102387704917e-05, "loss": 0.0432, "step": 16720 }, { "grad_norm": 0.39414000511169434, "learning_rate": 9.999091914221486e-05, "loss": 0.039, "step": 16730 }, { "grad_norm": 0.40612199902534485, "learning_rate": 9.99908137999437e-05, "loss": 0.0349, "step": 16740 }, { "grad_norm": 0.2795502841472626, "learning_rate": 9.999070785023696e-05, "loss": 0.054, "step": 16750 }, { "grad_norm": 0.2712258994579315, "learning_rate": 9.99906012930959e-05, "loss": 0.0419, "step": 16760 }, { "grad_norm": 0.23037301003932953, "learning_rate": 9.999049412852186e-05, "loss": 0.0415, "step": 16770 }, { "grad_norm": 0.37719377875328064, "learning_rate": 9.99903863565161e-05, "loss": 0.0433, "step": 16780 }, { "grad_norm": 0.2697863280773163, "learning_rate": 9.999027797707994e-05, "loss": 0.0371, "step": 16790 }, { "grad_norm": 0.2654576599597931, "learning_rate": 9.999016899021472e-05, "loss": 0.0373, "step": 16800 }, { "grad_norm": 0.22206997871398926, "learning_rate": 9.999005939592174e-05, "loss": 0.0348, "step": 16810 }, { "grad_norm": 0.20112349092960358, "learning_rate": 9.998994919420234e-05, "loss": 0.0339, "step": 16820 }, { "grad_norm": 0.20635218918323517, "learning_rate": 9.998983838505786e-05, "loss": 0.0383, "step": 16830 }, { "grad_norm": 0.23570024967193604, "learning_rate": 9.998972696848965e-05, "loss": 0.0329, "step": 16840 }, { "grad_norm": 0.22887015342712402, "learning_rate": 9.998961494449906e-05, "loss": 0.0324, "step": 16850 }, { "grad_norm": 0.1920592337846756, "learning_rate": 9.998950231308744e-05, "loss": 0.0371, "step": 16860 }, { "grad_norm": 0.20140165090560913, "learning_rate": 9.998938907425616e-05, "loss": 0.0296, "step": 16870 }, { "grad_norm": 0.22716648876667023, "learning_rate": 9.998927522800663e-05, "loss": 0.0294, "step": 16880 }, { "grad_norm": 0.3521089255809784, "learning_rate": 9.998916077434019e-05, "loss": 0.0472, "step": 16890 }, { "grad_norm": 0.32907795906066895, "learning_rate": 9.998904571325827e-05, "loss": 0.0461, "step": 16900 }, { "grad_norm": 0.27047595381736755, "learning_rate": 9.998893004476224e-05, "loss": 0.0437, "step": 16910 }, { "grad_norm": 0.212936669588089, "learning_rate": 9.99888137688535e-05, "loss": 0.0387, "step": 16920 }, { "grad_norm": 0.2548472583293915, "learning_rate": 9.998869688553348e-05, "loss": 0.0373, "step": 16930 }, { "grad_norm": 0.24382245540618896, "learning_rate": 9.998857939480361e-05, "loss": 0.0429, "step": 16940 }, { "grad_norm": 0.3141598403453827, "learning_rate": 9.998846129666529e-05, "loss": 0.0397, "step": 16950 }, { "grad_norm": 0.2957024872303009, "learning_rate": 9.998834259111996e-05, "loss": 0.0407, "step": 16960 }, { "grad_norm": 0.2915046513080597, "learning_rate": 9.998822327816908e-05, "loss": 0.0377, "step": 16970 }, { "grad_norm": 0.21235667169094086, "learning_rate": 9.99881033578141e-05, "loss": 0.0414, "step": 16980 }, { "grad_norm": 0.2880818247795105, "learning_rate": 9.998798283005647e-05, "loss": 0.0403, "step": 16990 }, { "grad_norm": 0.5181629061698914, "learning_rate": 9.998786169489762e-05, "loss": 0.0358, "step": 17000 }, { "grad_norm": 0.2877022325992584, "learning_rate": 9.998773995233908e-05, "loss": 0.0452, "step": 17010 }, { "grad_norm": 0.30543503165245056, "learning_rate": 9.99876176023823e-05, "loss": 0.0407, "step": 17020 }, { "grad_norm": 0.24324218928813934, "learning_rate": 9.998749464502877e-05, "loss": 0.045, "step": 17030 }, { "grad_norm": 0.3136717677116394, "learning_rate": 9.998737108027998e-05, "loss": 0.0434, "step": 17040 }, { "grad_norm": 0.2509809136390686, "learning_rate": 9.998724690813744e-05, "loss": 0.0391, "step": 17050 }, { "grad_norm": 0.5913176536560059, "learning_rate": 9.998712212860264e-05, "loss": 0.0437, "step": 17060 }, { "grad_norm": 0.24342718720436096, "learning_rate": 9.998699674167712e-05, "loss": 0.0326, "step": 17070 }, { "grad_norm": 0.3498716652393341, "learning_rate": 9.998687074736241e-05, "loss": 0.0399, "step": 17080 }, { "grad_norm": 0.44895273447036743, "learning_rate": 9.998674414566001e-05, "loss": 0.0419, "step": 17090 }, { "grad_norm": 0.16791924834251404, "learning_rate": 9.998661693657148e-05, "loss": 0.033, "step": 17100 }, { "grad_norm": 0.3974567949771881, "learning_rate": 9.998648912009835e-05, "loss": 0.0438, "step": 17110 }, { "grad_norm": 0.3118296265602112, "learning_rate": 9.998636069624219e-05, "loss": 0.0318, "step": 17120 }, { "grad_norm": 0.30052509903907776, "learning_rate": 9.998623166500454e-05, "loss": 0.0415, "step": 17130 }, { "grad_norm": 0.1907089650630951, "learning_rate": 9.998610202638699e-05, "loss": 0.0375, "step": 17140 }, { "grad_norm": 0.2665863335132599, "learning_rate": 9.998597178039111e-05, "loss": 0.0397, "step": 17150 }, { "grad_norm": 0.3730449676513672, "learning_rate": 9.998584092701848e-05, "loss": 0.0398, "step": 17160 }, { "grad_norm": 0.2260057032108307, "learning_rate": 9.998570946627067e-05, "loss": 0.0404, "step": 17170 }, { "grad_norm": 0.2994707226753235, "learning_rate": 9.998557739814931e-05, "loss": 0.036, "step": 17180 }, { "grad_norm": 0.3130331337451935, "learning_rate": 9.998544472265598e-05, "loss": 0.0324, "step": 17190 }, { "grad_norm": 0.2779163420200348, "learning_rate": 9.998531143979232e-05, "loss": 0.0381, "step": 17200 }, { "grad_norm": 0.8221421837806702, "learning_rate": 9.998517754955992e-05, "loss": 0.0411, "step": 17210 }, { "grad_norm": 0.33448657393455505, "learning_rate": 9.998504305196042e-05, "loss": 0.047, "step": 17220 }, { "grad_norm": 0.29604580998420715, "learning_rate": 9.998490794699546e-05, "loss": 0.0426, "step": 17230 }, { "grad_norm": 0.3023248016834259, "learning_rate": 9.998477223466667e-05, "loss": 0.0394, "step": 17240 }, { "grad_norm": 0.28459516167640686, "learning_rate": 9.998463591497569e-05, "loss": 0.0559, "step": 17250 }, { "grad_norm": 0.26391127705574036, "learning_rate": 9.998449898792422e-05, "loss": 0.0384, "step": 17260 }, { "grad_norm": 0.18978473544120789, "learning_rate": 9.998436145351386e-05, "loss": 0.0352, "step": 17270 }, { "grad_norm": 0.36309683322906494, "learning_rate": 9.998422331174634e-05, "loss": 0.0329, "step": 17280 }, { "grad_norm": 0.24028083682060242, "learning_rate": 9.99840845626233e-05, "loss": 0.043, "step": 17290 }, { "grad_norm": 0.2894856333732605, "learning_rate": 9.998394520614645e-05, "loss": 0.032, "step": 17300 }, { "grad_norm": 0.26286545395851135, "learning_rate": 9.998380524231747e-05, "loss": 0.0414, "step": 17310 }, { "grad_norm": 0.40233278274536133, "learning_rate": 9.998366467113807e-05, "loss": 0.0387, "step": 17320 }, { "grad_norm": 0.20554274320602417, "learning_rate": 9.998352349260993e-05, "loss": 0.0376, "step": 17330 }, { "grad_norm": 0.3265766203403473, "learning_rate": 9.99833817067348e-05, "loss": 0.0375, "step": 17340 }, { "grad_norm": 0.6233507394790649, "learning_rate": 9.998323931351441e-05, "loss": 0.0366, "step": 17350 }, { "grad_norm": 0.4137099087238312, "learning_rate": 9.998309631295045e-05, "loss": 0.0456, "step": 17360 }, { "grad_norm": 0.27364373207092285, "learning_rate": 9.998295270504468e-05, "loss": 0.04, "step": 17370 }, { "grad_norm": 0.3423931896686554, "learning_rate": 9.998280848979884e-05, "loss": 0.0337, "step": 17380 }, { "grad_norm": 0.2596067190170288, "learning_rate": 9.99826636672147e-05, "loss": 0.0329, "step": 17390 }, { "grad_norm": 0.2552863359451294, "learning_rate": 9.998251823729399e-05, "loss": 0.0422, "step": 17400 }, { "grad_norm": 0.23087099194526672, "learning_rate": 9.998237220003848e-05, "loss": 0.035, "step": 17410 }, { "grad_norm": 0.31257349252700806, "learning_rate": 9.998222555544999e-05, "loss": 0.0382, "step": 17420 }, { "grad_norm": 0.8398702144622803, "learning_rate": 9.998207830353024e-05, "loss": 0.0356, "step": 17430 }, { "grad_norm": 0.4529886245727539, "learning_rate": 9.998193044428106e-05, "loss": 0.0427, "step": 17440 }, { "grad_norm": 0.26900479197502136, "learning_rate": 9.998178197770422e-05, "loss": 0.0416, "step": 17450 }, { "grad_norm": 0.3761301040649414, "learning_rate": 9.998163290380156e-05, "loss": 0.0474, "step": 17460 }, { "grad_norm": 0.309958279132843, "learning_rate": 9.998148322257486e-05, "loss": 0.0397, "step": 17470 }, { "grad_norm": 0.3027549088001251, "learning_rate": 9.998133293402595e-05, "loss": 0.0418, "step": 17480 }, { "grad_norm": 0.4101981818675995, "learning_rate": 9.998118203815665e-05, "loss": 0.0553, "step": 17490 }, { "grad_norm": 0.2549568712711334, "learning_rate": 9.998103053496882e-05, "loss": 0.0452, "step": 17500 }, { "grad_norm": 0.3176654577255249, "learning_rate": 9.998087842446426e-05, "loss": 0.0434, "step": 17510 }, { "grad_norm": 0.24273082613945007, "learning_rate": 9.998072570664483e-05, "loss": 0.0394, "step": 17520 }, { "grad_norm": 0.31688496470451355, "learning_rate": 9.99805723815124e-05, "loss": 0.0327, "step": 17530 }, { "grad_norm": 0.2443731725215912, "learning_rate": 9.998041844906883e-05, "loss": 0.0381, "step": 17540 }, { "grad_norm": 0.2882452607154846, "learning_rate": 9.998026390931598e-05, "loss": 0.0333, "step": 17550 }, { "grad_norm": 0.23267799615859985, "learning_rate": 9.998010876225574e-05, "loss": 0.0478, "step": 17560 }, { "grad_norm": 0.227216437458992, "learning_rate": 9.997995300789e-05, "loss": 0.0379, "step": 17570 }, { "grad_norm": 0.2667263150215149, "learning_rate": 9.997979664622062e-05, "loss": 0.0308, "step": 17580 }, { "grad_norm": 0.28591063618659973, "learning_rate": 9.997963967724952e-05, "loss": 0.0435, "step": 17590 }, { "grad_norm": 0.2033149152994156, "learning_rate": 9.997948210097863e-05, "loss": 0.044, "step": 17600 }, { "grad_norm": 0.3290752172470093, "learning_rate": 9.997932391740983e-05, "loss": 0.0468, "step": 17610 }, { "grad_norm": 0.2047863006591797, "learning_rate": 9.997916512654507e-05, "loss": 0.0303, "step": 17620 }, { "grad_norm": 0.20531970262527466, "learning_rate": 9.997900572838626e-05, "loss": 0.0349, "step": 17630 }, { "grad_norm": 0.23300395905971527, "learning_rate": 9.997884572293534e-05, "loss": 0.0319, "step": 17640 }, { "grad_norm": 0.23998667299747467, "learning_rate": 9.997868511019425e-05, "loss": 0.0358, "step": 17650 }, { "grad_norm": 0.36090222001075745, "learning_rate": 9.997852389016496e-05, "loss": 0.0472, "step": 17660 }, { "grad_norm": 0.20574313402175903, "learning_rate": 9.997836206284941e-05, "loss": 0.0337, "step": 17670 }, { "grad_norm": 0.24145404994487762, "learning_rate": 9.997819962824957e-05, "loss": 0.0421, "step": 17680 }, { "grad_norm": 0.22060349583625793, "learning_rate": 9.997803658636742e-05, "loss": 0.0382, "step": 17690 }, { "grad_norm": 0.16915357112884521, "learning_rate": 9.997787293720495e-05, "loss": 0.0511, "step": 17700 }, { "grad_norm": 0.2803122103214264, "learning_rate": 9.997770868076413e-05, "loss": 0.0383, "step": 17710 }, { "grad_norm": 0.338247686624527, "learning_rate": 9.997754381704696e-05, "loss": 0.0333, "step": 17720 }, { "grad_norm": 0.40659409761428833, "learning_rate": 9.997737834605545e-05, "loss": 0.0377, "step": 17730 }, { "grad_norm": 0.3346559703350067, "learning_rate": 9.99772122677916e-05, "loss": 0.0475, "step": 17740 }, { "grad_norm": 0.4125784933567047, "learning_rate": 9.997704558225743e-05, "loss": 0.0472, "step": 17750 }, { "grad_norm": 0.3219510614871979, "learning_rate": 9.997687828945499e-05, "loss": 0.0395, "step": 17760 }, { "grad_norm": 0.1880550980567932, "learning_rate": 9.997671038938628e-05, "loss": 0.0366, "step": 17770 }, { "grad_norm": 0.29774725437164307, "learning_rate": 9.997654188205334e-05, "loss": 0.0429, "step": 17780 }, { "grad_norm": 0.19602859020233154, "learning_rate": 9.997637276745826e-05, "loss": 0.032, "step": 17790 }, { "grad_norm": 0.20221930742263794, "learning_rate": 9.997620304560304e-05, "loss": 0.0371, "step": 17800 }, { "grad_norm": 0.3842109441757202, "learning_rate": 9.997603271648979e-05, "loss": 0.0395, "step": 17810 }, { "grad_norm": 0.39434748888015747, "learning_rate": 9.997586178012052e-05, "loss": 0.043, "step": 17820 }, { "grad_norm": 0.2694343626499176, "learning_rate": 9.997569023649737e-05, "loss": 0.0368, "step": 17830 }, { "grad_norm": 0.3643442392349243, "learning_rate": 9.997551808562241e-05, "loss": 0.0371, "step": 17840 }, { "grad_norm": 0.26077115535736084, "learning_rate": 9.997534532749771e-05, "loss": 0.0298, "step": 17850 }, { "grad_norm": 0.30015644431114197, "learning_rate": 9.997517196212538e-05, "loss": 0.0387, "step": 17860 }, { "grad_norm": 0.23081903159618378, "learning_rate": 9.997499798950751e-05, "loss": 0.0323, "step": 17870 }, { "grad_norm": 0.37541064620018005, "learning_rate": 9.997482340964625e-05, "loss": 0.0313, "step": 17880 }, { "grad_norm": 0.26715317368507385, "learning_rate": 9.99746482225437e-05, "loss": 0.0357, "step": 17890 }, { "grad_norm": 0.23416520655155182, "learning_rate": 9.997447242820199e-05, "loss": 0.0376, "step": 17900 }, { "grad_norm": 0.39455822110176086, "learning_rate": 9.997429602662325e-05, "loss": 0.0386, "step": 17910 }, { "grad_norm": 0.2220374345779419, "learning_rate": 9.997411901780963e-05, "loss": 0.039, "step": 17920 }, { "grad_norm": 0.22260083258152008, "learning_rate": 9.997394140176328e-05, "loss": 0.0369, "step": 17930 }, { "grad_norm": 0.29394105076789856, "learning_rate": 9.997376317848636e-05, "loss": 0.033, "step": 17940 }, { "grad_norm": 0.2753209173679352, "learning_rate": 9.997358434798103e-05, "loss": 0.043, "step": 17950 }, { "grad_norm": 0.2328256517648697, "learning_rate": 9.997340491024948e-05, "loss": 0.0376, "step": 17960 }, { "grad_norm": 0.1859995722770691, "learning_rate": 9.997322486529387e-05, "loss": 0.0394, "step": 17970 }, { "grad_norm": 0.22094906866550446, "learning_rate": 9.997304421311639e-05, "loss": 0.0444, "step": 17980 }, { "grad_norm": 0.29147180914878845, "learning_rate": 9.997286295371926e-05, "loss": 0.0307, "step": 17990 }, { "grad_norm": 0.17322604358196259, "learning_rate": 9.997268108710464e-05, "loss": 0.046, "step": 18000 }, { "grad_norm": 0.29873186349868774, "learning_rate": 9.997249861327478e-05, "loss": 0.0508, "step": 18010 }, { "grad_norm": 0.2876291573047638, "learning_rate": 9.997231553223186e-05, "loss": 0.0392, "step": 18020 }, { "grad_norm": 0.32941219210624695, "learning_rate": 9.997213184397814e-05, "loss": 0.0455, "step": 18030 }, { "grad_norm": 0.2602584958076477, "learning_rate": 9.997194754851584e-05, "loss": 0.0445, "step": 18040 }, { "grad_norm": 0.24632547795772552, "learning_rate": 9.997176264584717e-05, "loss": 0.0396, "step": 18050 }, { "grad_norm": 0.24125389754772186, "learning_rate": 9.997157713597444e-05, "loss": 0.0351, "step": 18060 }, { "grad_norm": 0.2762829661369324, "learning_rate": 9.997139101889986e-05, "loss": 0.0319, "step": 18070 }, { "grad_norm": 0.21495646238327026, "learning_rate": 9.997120429462567e-05, "loss": 0.0303, "step": 18080 }, { "grad_norm": 0.2935071289539337, "learning_rate": 9.997101696315418e-05, "loss": 0.0413, "step": 18090 }, { "grad_norm": 0.35275423526763916, "learning_rate": 9.997082902448767e-05, "loss": 0.0407, "step": 18100 }, { "grad_norm": 0.2958535850048065, "learning_rate": 9.997064047862839e-05, "loss": 0.0379, "step": 18110 }, { "grad_norm": 0.4805663228034973, "learning_rate": 9.997045132557866e-05, "loss": 0.0353, "step": 18120 }, { "grad_norm": 0.3142434358596802, "learning_rate": 9.997026156534075e-05, "loss": 0.0353, "step": 18130 }, { "grad_norm": 0.1964873969554901, "learning_rate": 9.9970071197917e-05, "loss": 0.031, "step": 18140 }, { "grad_norm": 0.3008022904396057, "learning_rate": 9.996988022330969e-05, "loss": 0.0334, "step": 18150 }, { "grad_norm": 0.17761142551898956, "learning_rate": 9.996968864152116e-05, "loss": 0.0325, "step": 18160 }, { "grad_norm": 0.21833288669586182, "learning_rate": 9.996949645255374e-05, "loss": 0.0355, "step": 18170 }, { "grad_norm": 0.24399596452713013, "learning_rate": 9.996930365640975e-05, "loss": 0.0387, "step": 18180 }, { "grad_norm": 0.19641023874282837, "learning_rate": 9.996911025309155e-05, "loss": 0.0428, "step": 18190 }, { "grad_norm": 0.26769429445266724, "learning_rate": 9.996891624260147e-05, "loss": 0.0357, "step": 18200 }, { "grad_norm": 0.35642510652542114, "learning_rate": 9.996872162494189e-05, "loss": 0.0336, "step": 18210 }, { "grad_norm": 0.21485355496406555, "learning_rate": 9.996852640011516e-05, "loss": 0.0326, "step": 18220 }, { "grad_norm": 0.18273720145225525, "learning_rate": 9.996833056812367e-05, "loss": 0.0384, "step": 18230 }, { "grad_norm": 0.27348366379737854, "learning_rate": 9.996813412896977e-05, "loss": 0.0303, "step": 18240 }, { "grad_norm": 0.2921496629714966, "learning_rate": 9.996793708265586e-05, "loss": 0.0394, "step": 18250 }, { "grad_norm": 0.18968477845191956, "learning_rate": 9.996773942918435e-05, "loss": 0.0312, "step": 18260 }, { "grad_norm": 0.4083889126777649, "learning_rate": 9.996754116855763e-05, "loss": 0.047, "step": 18270 }, { "grad_norm": 0.3228883743286133, "learning_rate": 9.99673423007781e-05, "loss": 0.0362, "step": 18280 }, { "grad_norm": 0.3059958517551422, "learning_rate": 9.996714282584818e-05, "loss": 0.0421, "step": 18290 }, { "grad_norm": 0.23102034628391266, "learning_rate": 9.99669427437703e-05, "loss": 0.0322, "step": 18300 }, { "grad_norm": 0.39039984345436096, "learning_rate": 9.996674205454688e-05, "loss": 0.038, "step": 18310 }, { "grad_norm": 0.3018050789833069, "learning_rate": 9.996654075818039e-05, "loss": 0.0381, "step": 18320 }, { "grad_norm": 0.2956974506378174, "learning_rate": 9.996633885467324e-05, "loss": 0.0343, "step": 18330 }, { "grad_norm": 0.2312692254781723, "learning_rate": 9.996613634402789e-05, "loss": 0.0339, "step": 18340 }, { "grad_norm": 0.19169409573078156, "learning_rate": 9.996593322624682e-05, "loss": 0.033, "step": 18350 }, { "grad_norm": 0.20004048943519592, "learning_rate": 9.996572950133248e-05, "loss": 0.0316, "step": 18360 }, { "grad_norm": 0.24755460023880005, "learning_rate": 9.996552516928735e-05, "loss": 0.0287, "step": 18370 }, { "grad_norm": 0.2615552246570587, "learning_rate": 9.996532023011392e-05, "loss": 0.0368, "step": 18380 }, { "grad_norm": 0.1738905906677246, "learning_rate": 9.996511468381467e-05, "loss": 0.0319, "step": 18390 }, { "grad_norm": 0.22752781212329865, "learning_rate": 9.996490853039211e-05, "loss": 0.0376, "step": 18400 }, { "grad_norm": 0.25343790650367737, "learning_rate": 9.996470176984873e-05, "loss": 0.0332, "step": 18410 }, { "grad_norm": 0.2130117118358612, "learning_rate": 9.996449440218704e-05, "loss": 0.0363, "step": 18420 }, { "grad_norm": 0.2260640561580658, "learning_rate": 9.996428642740958e-05, "loss": 0.0322, "step": 18430 }, { "grad_norm": 0.25990232825279236, "learning_rate": 9.996407784551887e-05, "loss": 0.0284, "step": 18440 }, { "grad_norm": 0.21338871121406555, "learning_rate": 9.996386865651744e-05, "loss": 0.0412, "step": 18450 }, { "grad_norm": 0.3283793330192566, "learning_rate": 9.996365886040782e-05, "loss": 0.0345, "step": 18460 }, { "grad_norm": 0.3733403980731964, "learning_rate": 9.996344845719257e-05, "loss": 0.0359, "step": 18470 }, { "grad_norm": 0.2612029016017914, "learning_rate": 9.996323744687426e-05, "loss": 0.0366, "step": 18480 }, { "grad_norm": 0.27587127685546875, "learning_rate": 9.996302582945543e-05, "loss": 0.0371, "step": 18490 }, { "grad_norm": 0.2108219861984253, "learning_rate": 9.996281360493866e-05, "loss": 0.051, "step": 18500 }, { "grad_norm": 0.25614288449287415, "learning_rate": 9.996260077332656e-05, "loss": 0.0458, "step": 18510 }, { "grad_norm": 0.31614065170288086, "learning_rate": 9.996238733462167e-05, "loss": 0.0333, "step": 18520 }, { "grad_norm": 0.3085174560546875, "learning_rate": 9.99621732888266e-05, "loss": 0.0476, "step": 18530 }, { "grad_norm": 0.17865367233753204, "learning_rate": 9.996195863594396e-05, "loss": 0.0274, "step": 18540 }, { "grad_norm": 0.3078097105026245, "learning_rate": 9.996174337597632e-05, "loss": 0.0419, "step": 18550 }, { "grad_norm": 0.5208872556686401, "learning_rate": 9.996152750892635e-05, "loss": 0.0537, "step": 18560 }, { "grad_norm": 0.38316547870635986, "learning_rate": 9.996131103479667e-05, "loss": 0.0377, "step": 18570 }, { "grad_norm": 0.2856883108615875, "learning_rate": 9.996109395358986e-05, "loss": 0.0402, "step": 18580 }, { "grad_norm": 0.2324475347995758, "learning_rate": 9.99608762653086e-05, "loss": 0.0336, "step": 18590 }, { "grad_norm": 0.43340447545051575, "learning_rate": 9.996065796995552e-05, "loss": 0.043, "step": 18600 }, { "grad_norm": 0.2880116105079651, "learning_rate": 9.996043906753328e-05, "loss": 0.0414, "step": 18610 }, { "grad_norm": 0.22856847941875458, "learning_rate": 9.996021955804453e-05, "loss": 0.0474, "step": 18620 }, { "grad_norm": 0.43750888109207153, "learning_rate": 9.995999944149192e-05, "loss": 0.0622, "step": 18630 }, { "grad_norm": 0.3407144248485565, "learning_rate": 9.995977871787818e-05, "loss": 0.0389, "step": 18640 }, { "grad_norm": 0.2290675938129425, "learning_rate": 9.995955738720595e-05, "loss": 0.0418, "step": 18650 }, { "grad_norm": 0.2489372193813324, "learning_rate": 9.995933544947792e-05, "loss": 0.0395, "step": 18660 }, { "grad_norm": 0.3763177990913391, "learning_rate": 9.995911290469679e-05, "loss": 0.0357, "step": 18670 }, { "grad_norm": 0.2670183777809143, "learning_rate": 9.995888975286529e-05, "loss": 0.0348, "step": 18680 }, { "grad_norm": 0.25180748105049133, "learning_rate": 9.99586659939861e-05, "loss": 0.0325, "step": 18690 }, { "grad_norm": 0.2146281749010086, "learning_rate": 9.995844162806195e-05, "loss": 0.0252, "step": 18700 }, { "grad_norm": 0.17531001567840576, "learning_rate": 9.995821665509558e-05, "loss": 0.0358, "step": 18710 }, { "grad_norm": 0.4102160930633545, "learning_rate": 9.995799107508968e-05, "loss": 0.0322, "step": 18720 }, { "grad_norm": 0.4219489097595215, "learning_rate": 9.995776488804702e-05, "loss": 0.0495, "step": 18730 }, { "grad_norm": 0.27122828364372253, "learning_rate": 9.995753809397037e-05, "loss": 0.0342, "step": 18740 }, { "grad_norm": 0.26698389649391174, "learning_rate": 9.995731069286246e-05, "loss": 0.0289, "step": 18750 }, { "grad_norm": 0.25693637132644653, "learning_rate": 9.995708268472606e-05, "loss": 0.0354, "step": 18760 }, { "grad_norm": 0.18386469781398773, "learning_rate": 9.995685406956392e-05, "loss": 0.045, "step": 18770 }, { "grad_norm": 0.3664935529232025, "learning_rate": 9.995662484737885e-05, "loss": 0.0409, "step": 18780 }, { "grad_norm": 0.34094464778900146, "learning_rate": 9.995639501817362e-05, "loss": 0.038, "step": 18790 }, { "grad_norm": 0.24353289604187012, "learning_rate": 9.995616458195103e-05, "loss": 0.0403, "step": 18800 }, { "grad_norm": 0.2756980061531067, "learning_rate": 9.995593353871386e-05, "loss": 0.0441, "step": 18810 }, { "grad_norm": 0.2568452060222626, "learning_rate": 9.995570188846495e-05, "loss": 0.0374, "step": 18820 }, { "grad_norm": 0.4643152058124542, "learning_rate": 9.995546963120708e-05, "loss": 0.0423, "step": 18830 }, { "grad_norm": 0.30481085181236267, "learning_rate": 9.995523676694309e-05, "loss": 0.0396, "step": 18840 }, { "grad_norm": 0.2632179260253906, "learning_rate": 9.995500329567583e-05, "loss": 0.0448, "step": 18850 }, { "grad_norm": 0.3289913237094879, "learning_rate": 9.995476921740807e-05, "loss": 0.042, "step": 18860 }, { "grad_norm": 0.3524632751941681, "learning_rate": 9.995453453214274e-05, "loss": 0.0345, "step": 18870 }, { "grad_norm": 0.20705099403858185, "learning_rate": 9.995429923988263e-05, "loss": 0.0386, "step": 18880 }, { "grad_norm": 0.24417608976364136, "learning_rate": 9.995406334063062e-05, "loss": 0.0315, "step": 18890 }, { "grad_norm": 0.22519372403621674, "learning_rate": 9.995382683438958e-05, "loss": 0.0443, "step": 18900 }, { "grad_norm": 0.24020890891551971, "learning_rate": 9.995358972116238e-05, "loss": 0.0427, "step": 18910 }, { "grad_norm": 0.2971595525741577, "learning_rate": 9.99533520009519e-05, "loss": 0.0407, "step": 18920 }, { "grad_norm": 0.21066926419734955, "learning_rate": 9.995311367376103e-05, "loss": 0.0388, "step": 18930 }, { "grad_norm": 0.2532672882080078, "learning_rate": 9.995287473959267e-05, "loss": 0.0333, "step": 18940 }, { "grad_norm": 0.24289117753505707, "learning_rate": 9.995263519844971e-05, "loss": 0.0314, "step": 18950 }, { "grad_norm": 0.2918507158756256, "learning_rate": 9.995239505033508e-05, "loss": 0.0382, "step": 18960 }, { "grad_norm": 0.24475368857383728, "learning_rate": 9.995215429525169e-05, "loss": 0.0365, "step": 18970 }, { "grad_norm": 0.2466566562652588, "learning_rate": 9.995191293320244e-05, "loss": 0.0375, "step": 18980 }, { "grad_norm": 0.161740243434906, "learning_rate": 9.99516709641903e-05, "loss": 0.0351, "step": 18990 }, { "grad_norm": 0.28457632660865784, "learning_rate": 9.99514283882182e-05, "loss": 0.0422, "step": 19000 }, { "grad_norm": 0.2224714159965515, "learning_rate": 9.995118520528907e-05, "loss": 0.0409, "step": 19010 }, { "grad_norm": 0.26325786113739014, "learning_rate": 9.995094141540588e-05, "loss": 0.0331, "step": 19020 }, { "grad_norm": 0.32263603806495667, "learning_rate": 9.995069701857161e-05, "loss": 0.0297, "step": 19030 }, { "grad_norm": 0.22751113772392273, "learning_rate": 9.995045201478918e-05, "loss": 0.041, "step": 19040 }, { "grad_norm": 0.261122465133667, "learning_rate": 9.995020640406162e-05, "loss": 0.032, "step": 19050 }, { "grad_norm": 0.2803523540496826, "learning_rate": 9.994996018639189e-05, "loss": 0.0339, "step": 19060 }, { "grad_norm": 0.19491593539714813, "learning_rate": 9.994971336178298e-05, "loss": 0.0287, "step": 19070 }, { "grad_norm": 0.7839483022689819, "learning_rate": 9.994946593023789e-05, "loss": 0.0324, "step": 19080 }, { "grad_norm": 0.2960800528526306, "learning_rate": 9.994921789175962e-05, "loss": 0.0482, "step": 19090 }, { "grad_norm": 0.23607203364372253, "learning_rate": 9.99489692463512e-05, "loss": 0.0487, "step": 19100 }, { "grad_norm": 0.23111426830291748, "learning_rate": 9.994871999401565e-05, "loss": 0.0422, "step": 19110 }, { "grad_norm": 0.24540847539901733, "learning_rate": 9.9948470134756e-05, "loss": 0.0325, "step": 19120 }, { "grad_norm": 0.2526690661907196, "learning_rate": 9.994821966857527e-05, "loss": 0.0419, "step": 19130 }, { "grad_norm": 0.30502328276634216, "learning_rate": 9.99479685954765e-05, "loss": 0.0378, "step": 19140 }, { "grad_norm": 0.19516004621982574, "learning_rate": 9.994771691546277e-05, "loss": 0.032, "step": 19150 }, { "grad_norm": 0.3382168412208557, "learning_rate": 9.99474646285371e-05, "loss": 0.036, "step": 19160 }, { "grad_norm": 0.24775445461273193, "learning_rate": 9.994721173470259e-05, "loss": 0.0349, "step": 19170 }, { "grad_norm": 0.22083206474781036, "learning_rate": 9.994695823396229e-05, "loss": 0.0328, "step": 19180 }, { "grad_norm": 0.27957093715667725, "learning_rate": 9.99467041263193e-05, "loss": 0.0494, "step": 19190 }, { "grad_norm": 0.2181461900472641, "learning_rate": 9.99464494117767e-05, "loss": 0.0355, "step": 19200 }, { "grad_norm": 0.190085306763649, "learning_rate": 9.994619409033759e-05, "loss": 0.0262, "step": 19210 }, { "grad_norm": 0.23165377974510193, "learning_rate": 9.994593816200504e-05, "loss": 0.0306, "step": 19220 }, { "grad_norm": 0.23634999990463257, "learning_rate": 9.994568162678219e-05, "loss": 0.0343, "step": 19230 }, { "grad_norm": 0.40656599402427673, "learning_rate": 9.994542448467216e-05, "loss": 0.0349, "step": 19240 }, { "grad_norm": 0.30754685401916504, "learning_rate": 9.994516673567807e-05, "loss": 0.0339, "step": 19250 }, { "grad_norm": 0.19806133210659027, "learning_rate": 9.994490837980305e-05, "loss": 0.0304, "step": 19260 }, { "grad_norm": 0.26831191778182983, "learning_rate": 9.994464941705022e-05, "loss": 0.0368, "step": 19270 }, { "grad_norm": 0.2444343864917755, "learning_rate": 9.994438984742275e-05, "loss": 0.0323, "step": 19280 }, { "grad_norm": 0.3061121106147766, "learning_rate": 9.99441296709238e-05, "loss": 0.049, "step": 19290 }, { "grad_norm": 0.23699715733528137, "learning_rate": 9.994386888755651e-05, "loss": 0.0367, "step": 19300 }, { "grad_norm": 0.18635773658752441, "learning_rate": 9.994360749732407e-05, "loss": 0.0336, "step": 19310 }, { "grad_norm": 0.30196723341941833, "learning_rate": 9.994334550022964e-05, "loss": 0.0428, "step": 19320 }, { "grad_norm": 0.23351378738880157, "learning_rate": 9.994308289627639e-05, "loss": 0.0495, "step": 19330 }, { "grad_norm": 0.4198554754257202, "learning_rate": 9.994281968546755e-05, "loss": 0.0533, "step": 19340 }, { "grad_norm": 0.2521858811378479, "learning_rate": 9.994255586780629e-05, "loss": 0.0396, "step": 19350 }, { "grad_norm": 0.1393662393093109, "learning_rate": 9.99422914432958e-05, "loss": 0.0332, "step": 19360 }, { "grad_norm": 0.2692260146141052, "learning_rate": 9.994202641193935e-05, "loss": 0.0379, "step": 19370 }, { "grad_norm": 0.2630615234375, "learning_rate": 9.994176077374012e-05, "loss": 0.0484, "step": 19380 }, { "grad_norm": 0.23423713445663452, "learning_rate": 9.994149452870133e-05, "loss": 0.0371, "step": 19390 }, { "grad_norm": 0.2023199498653412, "learning_rate": 9.994122767682623e-05, "loss": 0.0321, "step": 19400 }, { "grad_norm": 0.23473522067070007, "learning_rate": 9.994096021811808e-05, "loss": 0.0348, "step": 19410 }, { "grad_norm": 0.2582721412181854, "learning_rate": 9.994069215258008e-05, "loss": 0.0343, "step": 19420 }, { "grad_norm": 0.3595593273639679, "learning_rate": 9.994042348021556e-05, "loss": 0.0399, "step": 19430 }, { "grad_norm": 0.2235838919878006, "learning_rate": 9.994015420102771e-05, "loss": 0.0311, "step": 19440 }, { "grad_norm": 0.1768053025007248, "learning_rate": 9.993988431501983e-05, "loss": 0.0381, "step": 19450 }, { "grad_norm": 0.14642775058746338, "learning_rate": 9.993961382219523e-05, "loss": 0.0307, "step": 19460 }, { "grad_norm": 0.1919405311346054, "learning_rate": 9.993934272255715e-05, "loss": 0.0329, "step": 19470 }, { "grad_norm": 0.2605191469192505, "learning_rate": 9.993907101610893e-05, "loss": 0.0319, "step": 19480 }, { "grad_norm": 0.2693978250026703, "learning_rate": 9.993879870285383e-05, "loss": 0.0448, "step": 19490 }, { "grad_norm": 0.2397385984659195, "learning_rate": 9.993852578279518e-05, "loss": 0.0364, "step": 19500 }, { "grad_norm": 0.4257088303565979, "learning_rate": 9.993825225593629e-05, "loss": 0.0533, "step": 19510 }, { "grad_norm": 0.24952249228954315, "learning_rate": 9.993797812228048e-05, "loss": 0.0361, "step": 19520 }, { "grad_norm": 0.23527805507183075, "learning_rate": 9.99377033818311e-05, "loss": 0.0378, "step": 19530 }, { "grad_norm": 0.17671188712120056, "learning_rate": 9.993742803459146e-05, "loss": 0.0306, "step": 19540 }, { "grad_norm": 0.3015466630458832, "learning_rate": 9.993715208056493e-05, "loss": 0.0316, "step": 19550 }, { "grad_norm": 0.3114785850048065, "learning_rate": 9.993687551975485e-05, "loss": 0.0396, "step": 19560 }, { "grad_norm": 0.21072383224964142, "learning_rate": 9.99365983521646e-05, "loss": 0.0378, "step": 19570 }, { "grad_norm": 0.3343771994113922, "learning_rate": 9.993632057779752e-05, "loss": 0.0378, "step": 19580 }, { "grad_norm": 0.29904982447624207, "learning_rate": 9.993604219665698e-05, "loss": 0.0408, "step": 19590 }, { "grad_norm": 0.25746890902519226, "learning_rate": 9.99357632087464e-05, "loss": 0.0344, "step": 19600 }, { "grad_norm": 0.3221709132194519, "learning_rate": 9.993548361406914e-05, "loss": 0.0325, "step": 19610 }, { "grad_norm": 0.3404932916164398, "learning_rate": 9.99352034126286e-05, "loss": 0.038, "step": 19620 }, { "grad_norm": 0.3283427059650421, "learning_rate": 9.993492260442822e-05, "loss": 0.0368, "step": 19630 }, { "grad_norm": 0.46678298711776733, "learning_rate": 9.993464118947137e-05, "loss": 0.03, "step": 19640 }, { "grad_norm": 0.24796822667121887, "learning_rate": 9.993435916776149e-05, "loss": 0.0385, "step": 19650 }, { "grad_norm": 0.26212602853775024, "learning_rate": 9.993407653930198e-05, "loss": 0.0414, "step": 19660 }, { "grad_norm": 0.22674569487571716, "learning_rate": 9.993379330409632e-05, "loss": 0.0344, "step": 19670 }, { "grad_norm": 0.23163822293281555, "learning_rate": 9.99335094621479e-05, "loss": 0.0346, "step": 19680 }, { "grad_norm": 0.2802231013774872, "learning_rate": 9.993322501346021e-05, "loss": 0.0378, "step": 19690 }, { "grad_norm": 0.24073991179466248, "learning_rate": 9.993293995803669e-05, "loss": 0.0414, "step": 19700 }, { "grad_norm": 0.19488584995269775, "learning_rate": 9.993265429588079e-05, "loss": 0.0335, "step": 19710 }, { "grad_norm": 0.43240800499916077, "learning_rate": 9.993236802699601e-05, "loss": 0.0353, "step": 19720 }, { "grad_norm": 0.22036844491958618, "learning_rate": 9.993208115138581e-05, "loss": 0.038, "step": 19730 }, { "grad_norm": 0.2028501182794571, "learning_rate": 9.99317936690537e-05, "loss": 0.037, "step": 19740 }, { "grad_norm": 0.2489084005355835, "learning_rate": 9.993150558000313e-05, "loss": 0.0346, "step": 19750 }, { "grad_norm": 0.3899490535259247, "learning_rate": 9.993121688423764e-05, "loss": 0.0381, "step": 19760 }, { "grad_norm": 0.23280535638332367, "learning_rate": 9.99309275817607e-05, "loss": 0.0381, "step": 19770 }, { "grad_norm": 0.2209785431623459, "learning_rate": 9.993063767257586e-05, "loss": 0.036, "step": 19780 }, { "grad_norm": 0.2379789501428604, "learning_rate": 9.993034715668664e-05, "loss": 0.0401, "step": 19790 }, { "grad_norm": 0.1874588429927826, "learning_rate": 9.993005603409655e-05, "loss": 0.0286, "step": 19800 }, { "grad_norm": 0.3338943421840668, "learning_rate": 9.992976430480914e-05, "loss": 0.0318, "step": 19810 }, { "grad_norm": 0.22607070207595825, "learning_rate": 9.992947196882796e-05, "loss": 0.029, "step": 19820 }, { "grad_norm": 0.17988164722919464, "learning_rate": 9.992917902615654e-05, "loss": 0.0286, "step": 19830 }, { "grad_norm": 0.20881834626197815, "learning_rate": 9.992888547679846e-05, "loss": 0.0284, "step": 19840 }, { "grad_norm": 0.5031805634498596, "learning_rate": 9.992859132075728e-05, "loss": 0.0451, "step": 19850 }, { "grad_norm": 0.281378835439682, "learning_rate": 9.992829655803657e-05, "loss": 0.0482, "step": 19860 }, { "grad_norm": 0.2621175944805145, "learning_rate": 9.992800118863992e-05, "loss": 0.0357, "step": 19870 }, { "grad_norm": 0.4028099775314331, "learning_rate": 9.992770521257092e-05, "loss": 0.0517, "step": 19880 }, { "grad_norm": 0.2223953902721405, "learning_rate": 9.992740862983316e-05, "loss": 0.0405, "step": 19890 }, { "grad_norm": 0.38605421781539917, "learning_rate": 9.992711144043024e-05, "loss": 0.0514, "step": 19900 }, { "grad_norm": 0.24585621058940887, "learning_rate": 9.992681364436578e-05, "loss": 0.0349, "step": 19910 }, { "grad_norm": 0.24402301013469696, "learning_rate": 9.992651524164341e-05, "loss": 0.0401, "step": 19920 }, { "grad_norm": 0.23794929683208466, "learning_rate": 9.992621623226672e-05, "loss": 0.0327, "step": 19930 }, { "grad_norm": 0.3519648611545563, "learning_rate": 9.992591661623937e-05, "loss": 0.037, "step": 19940 }, { "grad_norm": 0.6652061343193054, "learning_rate": 9.992561639356502e-05, "loss": 0.0393, "step": 19950 }, { "grad_norm": 0.3273833692073822, "learning_rate": 9.992531556424726e-05, "loss": 0.0412, "step": 19960 }, { "grad_norm": 0.27544599771499634, "learning_rate": 9.992501412828978e-05, "loss": 0.0365, "step": 19970 }, { "grad_norm": 0.3140949010848999, "learning_rate": 9.992471208569626e-05, "loss": 0.039, "step": 19980 }, { "grad_norm": 0.2288181185722351, "learning_rate": 9.992440943647035e-05, "loss": 0.0325, "step": 19990 }, { "grad_norm": 0.22802890837192535, "learning_rate": 9.992410618061572e-05, "loss": 0.0335, "step": 20000 }, { "grad_norm": 0.15210852026939392, "learning_rate": 9.992380231813607e-05, "loss": 0.0261, "step": 20010 }, { "grad_norm": 0.16120602190494537, "learning_rate": 9.992349784903508e-05, "loss": 0.0321, "step": 20020 }, { "grad_norm": 0.19909432530403137, "learning_rate": 9.992319277331646e-05, "loss": 0.0268, "step": 20030 }, { "grad_norm": 0.2415756732225418, "learning_rate": 9.992288709098389e-05, "loss": 0.0479, "step": 20040 }, { "grad_norm": 0.2700631320476532, "learning_rate": 9.992258080204113e-05, "loss": 0.0274, "step": 20050 }, { "grad_norm": 0.5508747100830078, "learning_rate": 9.99222739064919e-05, "loss": 0.0254, "step": 20060 }, { "grad_norm": 0.35128191113471985, "learning_rate": 9.992196640433986e-05, "loss": 0.0403, "step": 20070 }, { "grad_norm": 0.28160005807876587, "learning_rate": 9.992165829558881e-05, "loss": 0.033, "step": 20080 }, { "grad_norm": 0.24137231707572937, "learning_rate": 9.99213495802425e-05, "loss": 0.0377, "step": 20090 }, { "grad_norm": 0.29969853162765503, "learning_rate": 9.992104025830465e-05, "loss": 0.0373, "step": 20100 }, { "grad_norm": 0.24048902094364166, "learning_rate": 9.992073032977903e-05, "loss": 0.0317, "step": 20110 }, { "grad_norm": 0.20205727219581604, "learning_rate": 9.992041979466937e-05, "loss": 0.0323, "step": 20120 }, { "grad_norm": 0.2915516197681427, "learning_rate": 9.992010865297952e-05, "loss": 0.0314, "step": 20130 }, { "grad_norm": 0.21818052232265472, "learning_rate": 9.991979690471319e-05, "loss": 0.0282, "step": 20140 }, { "grad_norm": 0.3054710626602173, "learning_rate": 9.991948454987421e-05, "loss": 0.029, "step": 20150 }, { "grad_norm": 0.19723233580589294, "learning_rate": 9.991917158846635e-05, "loss": 0.0322, "step": 20160 }, { "grad_norm": 0.2929457128047943, "learning_rate": 9.991885802049343e-05, "loss": 0.0306, "step": 20170 }, { "grad_norm": 0.21121464669704437, "learning_rate": 9.991854384595926e-05, "loss": 0.0358, "step": 20180 }, { "grad_norm": 0.170743927359581, "learning_rate": 9.991822906486763e-05, "loss": 0.0318, "step": 20190 }, { "grad_norm": 0.28131574392318726, "learning_rate": 9.991791367722241e-05, "loss": 0.0257, "step": 20200 }, { "grad_norm": 0.15422828495502472, "learning_rate": 9.991759768302739e-05, "loss": 0.0279, "step": 20210 }, { "grad_norm": 0.48766493797302246, "learning_rate": 9.991728108228645e-05, "loss": 0.0453, "step": 20220 }, { "grad_norm": 0.3912550210952759, "learning_rate": 9.99169638750034e-05, "loss": 0.0434, "step": 20230 }, { "grad_norm": 0.24254737794399261, "learning_rate": 9.991664606118212e-05, "loss": 0.0331, "step": 20240 }, { "grad_norm": 0.20615354180335999, "learning_rate": 9.991632764082644e-05, "loss": 0.0339, "step": 20250 }, { "grad_norm": 0.19447076320648193, "learning_rate": 9.991600861394026e-05, "loss": 0.039, "step": 20260 }, { "grad_norm": 0.3379705846309662, "learning_rate": 9.991568898052745e-05, "loss": 0.0367, "step": 20270 }, { "grad_norm": 0.23588021099567413, "learning_rate": 9.99153687405919e-05, "loss": 0.0384, "step": 20280 }, { "grad_norm": 0.2447458803653717, "learning_rate": 9.991504789413748e-05, "loss": 0.048, "step": 20290 }, { "grad_norm": 0.21267230808734894, "learning_rate": 9.99147264411681e-05, "loss": 0.0389, "step": 20300 }, { "grad_norm": 0.23352625966072083, "learning_rate": 9.991440438168767e-05, "loss": 0.0427, "step": 20310 }, { "grad_norm": 0.2137167602777481, "learning_rate": 9.99140817157001e-05, "loss": 0.0336, "step": 20320 }, { "grad_norm": 0.23644663393497467, "learning_rate": 9.991375844320931e-05, "loss": 0.0361, "step": 20330 }, { "grad_norm": 0.27744343876838684, "learning_rate": 9.991343456421923e-05, "loss": 0.0465, "step": 20340 }, { "grad_norm": 0.30836018919944763, "learning_rate": 9.99131100787338e-05, "loss": 0.0359, "step": 20350 }, { "grad_norm": 0.19175921380519867, "learning_rate": 9.991278498675694e-05, "loss": 0.0364, "step": 20360 }, { "grad_norm": 0.7261927723884583, "learning_rate": 9.991245928829262e-05, "loss": 0.0361, "step": 20370 }, { "grad_norm": 0.2901749610900879, "learning_rate": 9.991213298334481e-05, "loss": 0.0456, "step": 20380 }, { "grad_norm": 0.25898632407188416, "learning_rate": 9.991180607191745e-05, "loss": 0.0371, "step": 20390 }, { "grad_norm": 0.20941776037216187, "learning_rate": 9.991147855401453e-05, "loss": 0.0335, "step": 20400 }, { "grad_norm": 0.24803106486797333, "learning_rate": 9.991115042964002e-05, "loss": 0.0271, "step": 20410 }, { "grad_norm": 0.573117733001709, "learning_rate": 9.991082169879788e-05, "loss": 0.0347, "step": 20420 }, { "grad_norm": 0.281367689371109, "learning_rate": 9.991049236149216e-05, "loss": 0.0355, "step": 20430 }, { "grad_norm": 0.23290085792541504, "learning_rate": 9.991016241772683e-05, "loss": 0.0351, "step": 20440 }, { "grad_norm": 0.2076185792684555, "learning_rate": 9.99098318675059e-05, "loss": 0.0319, "step": 20450 }, { "grad_norm": 0.1924331784248352, "learning_rate": 9.99095007108334e-05, "loss": 0.0481, "step": 20460 }, { "grad_norm": 0.33753713965415955, "learning_rate": 9.990916894771334e-05, "loss": 0.0455, "step": 20470 }, { "grad_norm": 0.34328973293304443, "learning_rate": 9.990883657814975e-05, "loss": 0.0369, "step": 20480 }, { "grad_norm": 0.4936140477657318, "learning_rate": 9.990850360214667e-05, "loss": 0.0315, "step": 20490 }, { "grad_norm": 0.3186393082141876, "learning_rate": 9.990817001970816e-05, "loss": 0.034, "step": 20500 }, { "grad_norm": 0.24714329838752747, "learning_rate": 9.990783583083825e-05, "loss": 0.0332, "step": 20510 }, { "grad_norm": 0.3106299638748169, "learning_rate": 9.990750103554102e-05, "loss": 0.0311, "step": 20520 }, { "grad_norm": 0.1784241646528244, "learning_rate": 9.990716563382055e-05, "loss": 0.0369, "step": 20530 }, { "grad_norm": 0.28055626153945923, "learning_rate": 9.990682962568087e-05, "loss": 0.033, "step": 20540 }, { "grad_norm": 0.2505364418029785, "learning_rate": 9.990649301112611e-05, "loss": 0.0368, "step": 20550 }, { "grad_norm": 0.29090213775634766, "learning_rate": 9.990615579016033e-05, "loss": 0.032, "step": 20560 }, { "grad_norm": 0.22272221744060516, "learning_rate": 9.990581796278764e-05, "loss": 0.038, "step": 20570 }, { "grad_norm": 0.3320561945438385, "learning_rate": 9.990547952901215e-05, "loss": 0.0318, "step": 20580 }, { "grad_norm": 0.16809095442295074, "learning_rate": 9.990514048883794e-05, "loss": 0.0323, "step": 20590 }, { "grad_norm": 0.1878856122493744, "learning_rate": 9.990480084226919e-05, "loss": 0.0301, "step": 20600 }, { "grad_norm": 0.23042139410972595, "learning_rate": 9.990446058930996e-05, "loss": 0.0324, "step": 20610 }, { "grad_norm": 0.5025630593299866, "learning_rate": 9.990411972996443e-05, "loss": 0.0279, "step": 20620 }, { "grad_norm": 0.21924059092998505, "learning_rate": 9.990377826423673e-05, "loss": 0.0332, "step": 20630 }, { "grad_norm": 0.20901523530483246, "learning_rate": 9.990343619213101e-05, "loss": 0.0405, "step": 20640 }, { "grad_norm": 0.2789216637611389, "learning_rate": 9.990309351365141e-05, "loss": 0.0388, "step": 20650 }, { "grad_norm": 0.2128421515226364, "learning_rate": 9.990275022880213e-05, "loss": 0.0327, "step": 20660 }, { "grad_norm": 0.24557901918888092, "learning_rate": 9.990240633758731e-05, "loss": 0.03, "step": 20670 }, { "grad_norm": 0.17179369926452637, "learning_rate": 9.990206184001113e-05, "loss": 0.0289, "step": 20680 }, { "grad_norm": 0.19066552817821503, "learning_rate": 9.990171673607777e-05, "loss": 0.0289, "step": 20690 }, { "grad_norm": 0.21557599306106567, "learning_rate": 9.990137102579146e-05, "loss": 0.0305, "step": 20700 }, { "grad_norm": 0.2550622820854187, "learning_rate": 9.990102470915637e-05, "loss": 0.0251, "step": 20710 }, { "grad_norm": 0.20983393490314484, "learning_rate": 9.990067778617673e-05, "loss": 0.0348, "step": 20720 }, { "grad_norm": 0.3302246630191803, "learning_rate": 9.990033025685672e-05, "loss": 0.0419, "step": 20730 }, { "grad_norm": 0.2667137682437897, "learning_rate": 9.98999821212006e-05, "loss": 0.0317, "step": 20740 }, { "grad_norm": 0.19409187138080597, "learning_rate": 9.989963337921256e-05, "loss": 0.0358, "step": 20750 }, { "grad_norm": 0.22343459725379944, "learning_rate": 9.989928403089688e-05, "loss": 0.0365, "step": 20760 }, { "grad_norm": 0.2168780118227005, "learning_rate": 9.989893407625779e-05, "loss": 0.0403, "step": 20770 }, { "grad_norm": 0.22867310047149658, "learning_rate": 9.989858351529953e-05, "loss": 0.0378, "step": 20780 }, { "grad_norm": 0.24042026698589325, "learning_rate": 9.989823234802636e-05, "loss": 0.0311, "step": 20790 }, { "grad_norm": 0.2191496640443802, "learning_rate": 9.989788057444256e-05, "loss": 0.0436, "step": 20800 }, { "grad_norm": 0.2892124056816101, "learning_rate": 9.989752819455241e-05, "loss": 0.0327, "step": 20810 }, { "grad_norm": 0.32017192244529724, "learning_rate": 9.989717520836017e-05, "loss": 0.0427, "step": 20820 }, { "grad_norm": 0.2279079109430313, "learning_rate": 9.989682161587015e-05, "loss": 0.0297, "step": 20830 }, { "grad_norm": 0.298060804605484, "learning_rate": 9.989646741708662e-05, "loss": 0.0402, "step": 20840 }, { "grad_norm": 0.2941182553768158, "learning_rate": 9.989611261201393e-05, "loss": 0.0482, "step": 20850 }, { "grad_norm": 0.355652391910553, "learning_rate": 9.989575720065633e-05, "loss": 0.0477, "step": 20860 }, { "grad_norm": 0.3508433997631073, "learning_rate": 9.98954011830182e-05, "loss": 0.0446, "step": 20870 }, { "grad_norm": 0.1991652250289917, "learning_rate": 9.989504455910381e-05, "loss": 0.035, "step": 20880 }, { "grad_norm": 0.21451812982559204, "learning_rate": 9.989468732891754e-05, "loss": 0.0433, "step": 20890 }, { "grad_norm": 0.20981711149215698, "learning_rate": 9.98943294924637e-05, "loss": 0.0344, "step": 20900 }, { "grad_norm": 0.23911108076572418, "learning_rate": 9.989397104974665e-05, "loss": 0.0355, "step": 20910 }, { "grad_norm": 0.16362364590168, "learning_rate": 9.989361200077074e-05, "loss": 0.034, "step": 20920 }, { "grad_norm": 0.1923627257347107, "learning_rate": 9.989325234554034e-05, "loss": 0.0282, "step": 20930 }, { "grad_norm": 0.4042852520942688, "learning_rate": 9.989289208405983e-05, "loss": 0.0281, "step": 20940 }, { "grad_norm": 0.26127931475639343, "learning_rate": 9.989253121633356e-05, "loss": 0.0349, "step": 20950 }, { "grad_norm": 0.3246753215789795, "learning_rate": 9.989216974236594e-05, "loss": 0.0345, "step": 20960 }, { "grad_norm": 0.14685320854187012, "learning_rate": 9.989180766216134e-05, "loss": 0.0256, "step": 20970 }, { "grad_norm": 0.22064246237277985, "learning_rate": 9.989144497572418e-05, "loss": 0.0239, "step": 20980 }, { "grad_norm": 0.21223600208759308, "learning_rate": 9.989108168305885e-05, "loss": 0.0281, "step": 20990 }, { "grad_norm": 0.272502601146698, "learning_rate": 9.989071778416977e-05, "loss": 0.037, "step": 21000 }, { "grad_norm": 0.2105465978384018, "learning_rate": 9.989035327906138e-05, "loss": 0.0416, "step": 21010 }, { "grad_norm": 0.19391502439975739, "learning_rate": 9.988998816773808e-05, "loss": 0.031, "step": 21020 }, { "grad_norm": 0.23970232903957367, "learning_rate": 9.988962245020433e-05, "loss": 0.0339, "step": 21030 }, { "grad_norm": 0.42621585726737976, "learning_rate": 9.988925612646455e-05, "loss": 0.0378, "step": 21040 }, { "grad_norm": 0.23284180462360382, "learning_rate": 9.988888919652322e-05, "loss": 0.0382, "step": 21050 }, { "grad_norm": 0.2010135054588318, "learning_rate": 9.988852166038477e-05, "loss": 0.0409, "step": 21060 }, { "grad_norm": 0.505618155002594, "learning_rate": 9.988815351805369e-05, "loss": 0.0448, "step": 21070 }, { "grad_norm": 0.19357752799987793, "learning_rate": 9.988778476953443e-05, "loss": 0.0356, "step": 21080 }, { "grad_norm": 0.16525450348854065, "learning_rate": 9.988741541483148e-05, "loss": 0.0307, "step": 21090 }, { "grad_norm": 0.32637691497802734, "learning_rate": 9.988704545394936e-05, "loss": 0.0316, "step": 21100 }, { "grad_norm": 0.2800424098968506, "learning_rate": 9.988667488689252e-05, "loss": 0.0291, "step": 21110 }, { "grad_norm": 0.4006192088127136, "learning_rate": 9.988630371366548e-05, "loss": 0.0304, "step": 21120 }, { "grad_norm": 0.21341563761234283, "learning_rate": 9.988593193427274e-05, "loss": 0.0376, "step": 21130 }, { "grad_norm": 0.2283298224210739, "learning_rate": 9.988555954871884e-05, "loss": 0.0337, "step": 21140 }, { "grad_norm": 0.4413362741470337, "learning_rate": 9.988518655700828e-05, "loss": 0.0335, "step": 21150 }, { "grad_norm": 0.23553135991096497, "learning_rate": 9.988481295914561e-05, "loss": 0.0419, "step": 21160 }, { "grad_norm": 0.22526198625564575, "learning_rate": 9.988443875513535e-05, "loss": 0.0317, "step": 21170 }, { "grad_norm": 0.1802443414926529, "learning_rate": 9.988406394498209e-05, "loss": 0.0297, "step": 21180 }, { "grad_norm": 0.2926875352859497, "learning_rate": 9.988368852869034e-05, "loss": 0.0314, "step": 21190 }, { "grad_norm": 0.19110067188739777, "learning_rate": 9.988331250626467e-05, "loss": 0.0331, "step": 21200 }, { "grad_norm": 0.34326881170272827, "learning_rate": 9.988293587770966e-05, "loss": 0.0318, "step": 21210 }, { "grad_norm": 0.2028147429227829, "learning_rate": 9.988255864302988e-05, "loss": 0.0343, "step": 21220 }, { "grad_norm": 0.18199323117733002, "learning_rate": 9.988218080222992e-05, "loss": 0.0327, "step": 21230 }, { "grad_norm": 0.195076584815979, "learning_rate": 9.988180235531437e-05, "loss": 0.0315, "step": 21240 }, { "grad_norm": 0.21550112962722778, "learning_rate": 9.988142330228781e-05, "loss": 0.0304, "step": 21250 }, { "grad_norm": 0.2124289870262146, "learning_rate": 9.988104364315487e-05, "loss": 0.0306, "step": 21260 }, { "grad_norm": 0.5905463099479675, "learning_rate": 9.988066337792015e-05, "loss": 0.0346, "step": 21270 }, { "grad_norm": 0.24076522886753082, "learning_rate": 9.988028250658829e-05, "loss": 0.0295, "step": 21280 }, { "grad_norm": 0.3344190716743469, "learning_rate": 9.987990102916389e-05, "loss": 0.0352, "step": 21290 }, { "grad_norm": 0.31949853897094727, "learning_rate": 9.98795189456516e-05, "loss": 0.0322, "step": 21300 }, { "grad_norm": 0.2958042025566101, "learning_rate": 9.987913625605606e-05, "loss": 0.0362, "step": 21310 }, { "grad_norm": 0.20966395735740662, "learning_rate": 9.987875296038192e-05, "loss": 0.033, "step": 21320 }, { "grad_norm": 0.26739323139190674, "learning_rate": 9.987836905863384e-05, "loss": 0.0361, "step": 21330 }, { "grad_norm": 0.15789273381233215, "learning_rate": 9.987798455081648e-05, "loss": 0.0288, "step": 21340 }, { "grad_norm": 0.16945087909698486, "learning_rate": 9.987759943693452e-05, "loss": 0.0376, "step": 21350 }, { "grad_norm": 0.2993294596672058, "learning_rate": 9.987721371699264e-05, "loss": 0.0346, "step": 21360 }, { "grad_norm": 0.231096088886261, "learning_rate": 9.987682739099551e-05, "loss": 0.0352, "step": 21370 }, { "grad_norm": 0.2589413821697235, "learning_rate": 9.987644045894784e-05, "loss": 0.0343, "step": 21380 }, { "grad_norm": 0.33683741092681885, "learning_rate": 9.987605292085432e-05, "loss": 0.0329, "step": 21390 }, { "grad_norm": 0.16303113102912903, "learning_rate": 9.987566477671968e-05, "loss": 0.033, "step": 21400 }, { "grad_norm": 0.16436342895030975, "learning_rate": 9.987527602654861e-05, "loss": 0.0338, "step": 21410 }, { "grad_norm": 0.37584638595581055, "learning_rate": 9.987488667034586e-05, "loss": 0.0357, "step": 21420 }, { "grad_norm": 0.2362576425075531, "learning_rate": 9.987449670811612e-05, "loss": 0.0285, "step": 21430 }, { "grad_norm": 0.21836970746517181, "learning_rate": 9.987410613986419e-05, "loss": 0.0341, "step": 21440 }, { "grad_norm": 0.24160699546337128, "learning_rate": 9.987371496559477e-05, "loss": 0.0381, "step": 21450 }, { "grad_norm": 0.19807609915733337, "learning_rate": 9.987332318531262e-05, "loss": 0.0275, "step": 21460 }, { "grad_norm": 0.35837140679359436, "learning_rate": 9.98729307990225e-05, "loss": 0.041, "step": 21470 }, { "grad_norm": 0.3705264627933502, "learning_rate": 9.987253780672919e-05, "loss": 0.0366, "step": 21480 }, { "grad_norm": 0.2807477116584778, "learning_rate": 9.987214420843746e-05, "loss": 0.0385, "step": 21490 }, { "grad_norm": 0.2778357267379761, "learning_rate": 9.987175000415208e-05, "loss": 0.0347, "step": 21500 }, { "grad_norm": 0.23485629260540009, "learning_rate": 9.987135519387787e-05, "loss": 0.0348, "step": 21510 }, { "grad_norm": 0.20627686381340027, "learning_rate": 9.98709597776196e-05, "loss": 0.0336, "step": 21520 }, { "grad_norm": 0.18092237412929535, "learning_rate": 9.987056375538206e-05, "loss": 0.029, "step": 21530 }, { "grad_norm": 0.19293077290058136, "learning_rate": 9.987016712717011e-05, "loss": 0.0299, "step": 21540 }, { "grad_norm": 0.2762642800807953, "learning_rate": 9.986976989298854e-05, "loss": 0.0336, "step": 21550 }, { "grad_norm": 0.2400955706834793, "learning_rate": 9.986937205284216e-05, "loss": 0.0265, "step": 21560 }, { "grad_norm": 0.20106017589569092, "learning_rate": 9.986897360673585e-05, "loss": 0.0352, "step": 21570 }, { "grad_norm": 0.1936720311641693, "learning_rate": 9.986857455467443e-05, "loss": 0.0314, "step": 21580 }, { "grad_norm": 0.20144176483154297, "learning_rate": 9.986817489666273e-05, "loss": 0.0291, "step": 21590 }, { "grad_norm": 0.2196824997663498, "learning_rate": 9.986777463270563e-05, "loss": 0.0279, "step": 21600 }, { "grad_norm": 0.2090831696987152, "learning_rate": 9.986737376280799e-05, "loss": 0.0324, "step": 21610 }, { "grad_norm": 0.1527956873178482, "learning_rate": 9.986697228697467e-05, "loss": 0.0305, "step": 21620 }, { "grad_norm": 0.1986100673675537, "learning_rate": 9.986657020521056e-05, "loss": 0.0294, "step": 21630 }, { "grad_norm": 0.2340007722377777, "learning_rate": 9.986616751752052e-05, "loss": 0.037, "step": 21640 }, { "grad_norm": 0.17314274609088898, "learning_rate": 9.986576422390948e-05, "loss": 0.0298, "step": 21650 }, { "grad_norm": 0.1953875720500946, "learning_rate": 9.986536032438232e-05, "loss": 0.0308, "step": 21660 }, { "grad_norm": 0.20012779533863068, "learning_rate": 9.986495581894395e-05, "loss": 0.0268, "step": 21670 }, { "grad_norm": 0.19840559363365173, "learning_rate": 9.986455070759929e-05, "loss": 0.0301, "step": 21680 }, { "grad_norm": 0.26104220747947693, "learning_rate": 9.986414499035327e-05, "loss": 0.0304, "step": 21690 }, { "grad_norm": 0.151571124792099, "learning_rate": 9.986373866721079e-05, "loss": 0.0298, "step": 21700 }, { "grad_norm": 0.2634979486465454, "learning_rate": 9.986333173817681e-05, "loss": 0.03, "step": 21710 }, { "grad_norm": 0.20080921053886414, "learning_rate": 9.986292420325627e-05, "loss": 0.0415, "step": 21720 }, { "grad_norm": 0.27000555396080017, "learning_rate": 9.986251606245412e-05, "loss": 0.0316, "step": 21730 }, { "grad_norm": 0.33717891573905945, "learning_rate": 9.986210731577534e-05, "loss": 0.0378, "step": 21740 }, { "grad_norm": 0.21066103875637054, "learning_rate": 9.986169796322487e-05, "loss": 0.0271, "step": 21750 }, { "grad_norm": 0.24049100279808044, "learning_rate": 9.986128800480768e-05, "loss": 0.0371, "step": 21760 }, { "grad_norm": 0.28943657875061035, "learning_rate": 9.986087744052876e-05, "loss": 0.0352, "step": 21770 }, { "grad_norm": 0.2420419603586197, "learning_rate": 9.986046627039311e-05, "loss": 0.0499, "step": 21780 }, { "grad_norm": 0.22496327757835388, "learning_rate": 9.986005449440574e-05, "loss": 0.0348, "step": 21790 }, { "grad_norm": 0.30270716547966003, "learning_rate": 9.985964211257162e-05, "loss": 0.0387, "step": 21800 }, { "grad_norm": 0.15627534687519073, "learning_rate": 9.985922912489575e-05, "loss": 0.0294, "step": 21810 }, { "grad_norm": 0.21032090485095978, "learning_rate": 9.98588155313832e-05, "loss": 0.033, "step": 21820 }, { "grad_norm": 0.2978767156600952, "learning_rate": 9.985840133203896e-05, "loss": 0.04, "step": 21830 }, { "grad_norm": 0.32403361797332764, "learning_rate": 9.985798652686807e-05, "loss": 0.0354, "step": 21840 }, { "grad_norm": 0.2573404908180237, "learning_rate": 9.985757111587556e-05, "loss": 0.0322, "step": 21850 }, { "grad_norm": 0.2192285805940628, "learning_rate": 9.985715509906649e-05, "loss": 0.0379, "step": 21860 }, { "grad_norm": 0.18072731792926788, "learning_rate": 9.985673847644592e-05, "loss": 0.0324, "step": 21870 }, { "grad_norm": 0.24881353974342346, "learning_rate": 9.98563212480189e-05, "loss": 0.0241, "step": 21880 }, { "grad_norm": 0.27894383668899536, "learning_rate": 9.985590341379051e-05, "loss": 0.0281, "step": 21890 }, { "grad_norm": 0.23993264138698578, "learning_rate": 9.985548497376582e-05, "loss": 0.0281, "step": 21900 }, { "grad_norm": 0.2711477279663086, "learning_rate": 9.98550659279499e-05, "loss": 0.0333, "step": 21910 }, { "grad_norm": 0.20255550742149353, "learning_rate": 9.985464627634789e-05, "loss": 0.0377, "step": 21920 }, { "grad_norm": 0.2485387921333313, "learning_rate": 9.985422601896484e-05, "loss": 0.0364, "step": 21930 }, { "grad_norm": 0.2487465888261795, "learning_rate": 9.985380515580587e-05, "loss": 0.034, "step": 21940 }, { "grad_norm": 0.21052363514900208, "learning_rate": 9.985338368687609e-05, "loss": 0.0272, "step": 21950 }, { "grad_norm": 0.22090955078601837, "learning_rate": 9.985296161218064e-05, "loss": 0.0282, "step": 21960 }, { "grad_norm": 0.19597980380058289, "learning_rate": 9.985253893172464e-05, "loss": 0.0312, "step": 21970 }, { "grad_norm": 0.19595611095428467, "learning_rate": 9.985211564551322e-05, "loss": 0.0312, "step": 21980 }, { "grad_norm": 0.25524964928627014, "learning_rate": 9.985169175355153e-05, "loss": 0.033, "step": 21990 }, { "grad_norm": 0.2336849719285965, "learning_rate": 9.985126725584471e-05, "loss": 0.0317, "step": 22000 }, { "grad_norm": 0.248225599527359, "learning_rate": 9.985084215239792e-05, "loss": 0.0284, "step": 22010 }, { "grad_norm": 0.24104824662208557, "learning_rate": 9.985041644321635e-05, "loss": 0.0317, "step": 22020 }, { "grad_norm": 0.18231362104415894, "learning_rate": 9.984999012830514e-05, "loss": 0.0273, "step": 22030 }, { "grad_norm": 0.273488849401474, "learning_rate": 9.984956320766949e-05, "loss": 0.0351, "step": 22040 }, { "grad_norm": 0.15119007229804993, "learning_rate": 9.984913568131457e-05, "loss": 0.0344, "step": 22050 }, { "grad_norm": 0.2141983062028885, "learning_rate": 9.98487075492456e-05, "loss": 0.0263, "step": 22060 }, { "grad_norm": 0.24815113842487335, "learning_rate": 9.984827881146777e-05, "loss": 0.0335, "step": 22070 }, { "grad_norm": 0.17627088725566864, "learning_rate": 9.984784946798628e-05, "loss": 0.0329, "step": 22080 }, { "grad_norm": 0.19447848200798035, "learning_rate": 9.984741951880635e-05, "loss": 0.037, "step": 22090 }, { "grad_norm": 0.17229260504245758, "learning_rate": 9.984698896393321e-05, "loss": 0.0292, "step": 22100 }, { "grad_norm": 0.2562866806983948, "learning_rate": 9.984655780337212e-05, "loss": 0.0357, "step": 22110 }, { "grad_norm": 0.307425856590271, "learning_rate": 9.984612603712825e-05, "loss": 0.0417, "step": 22120 }, { "grad_norm": 0.20030765235424042, "learning_rate": 9.98456936652069e-05, "loss": 0.0342, "step": 22130 }, { "grad_norm": 0.2100946605205536, "learning_rate": 9.98452606876133e-05, "loss": 0.0335, "step": 22140 }, { "grad_norm": 0.18903237581253052, "learning_rate": 9.984482710435274e-05, "loss": 0.0268, "step": 22150 }, { "grad_norm": 0.14085471630096436, "learning_rate": 9.984439291543046e-05, "loss": 0.0256, "step": 22160 }, { "grad_norm": 0.32021763920783997, "learning_rate": 9.984395812085175e-05, "loss": 0.0384, "step": 22170 }, { "grad_norm": 0.27052026987075806, "learning_rate": 9.98435227206219e-05, "loss": 0.0313, "step": 22180 }, { "grad_norm": 0.19925644993782043, "learning_rate": 9.984308671474616e-05, "loss": 0.0282, "step": 22190 }, { "grad_norm": 0.23953497409820557, "learning_rate": 9.984265010322987e-05, "loss": 0.0313, "step": 22200 }, { "grad_norm": 0.22156769037246704, "learning_rate": 9.984221288607834e-05, "loss": 0.0262, "step": 22210 }, { "grad_norm": 0.4903034567832947, "learning_rate": 9.984177506329684e-05, "loss": 0.0308, "step": 22220 }, { "grad_norm": 0.44124969840049744, "learning_rate": 9.984133663489072e-05, "loss": 0.0268, "step": 22230 }, { "grad_norm": 0.24512061476707458, "learning_rate": 9.98408976008653e-05, "loss": 0.0321, "step": 22240 }, { "grad_norm": 0.19057337939739227, "learning_rate": 9.984045796122591e-05, "loss": 0.0284, "step": 22250 }, { "grad_norm": 0.2511794865131378, "learning_rate": 9.984001771597792e-05, "loss": 0.0303, "step": 22260 }, { "grad_norm": 0.2723226249217987, "learning_rate": 9.983957686512663e-05, "loss": 0.0294, "step": 22270 }, { "grad_norm": 0.4349762797355652, "learning_rate": 9.983913540867747e-05, "loss": 0.0362, "step": 22280 }, { "grad_norm": 0.28734010457992554, "learning_rate": 9.983869334663571e-05, "loss": 0.0339, "step": 22290 }, { "grad_norm": 0.21122536063194275, "learning_rate": 9.983825067900681e-05, "loss": 0.033, "step": 22300 }, { "grad_norm": 0.14385414123535156, "learning_rate": 9.983780740579608e-05, "loss": 0.0328, "step": 22310 }, { "grad_norm": 0.28535884618759155, "learning_rate": 9.983736352700894e-05, "loss": 0.0322, "step": 22320 }, { "grad_norm": 0.17218557000160217, "learning_rate": 9.983691904265079e-05, "loss": 0.0369, "step": 22330 }, { "grad_norm": 0.1848953664302826, "learning_rate": 9.983647395272701e-05, "loss": 0.0376, "step": 22340 }, { "grad_norm": 0.17273585498332977, "learning_rate": 9.983602825724303e-05, "loss": 0.0297, "step": 22350 }, { "grad_norm": 0.16116581857204437, "learning_rate": 9.983558195620424e-05, "loss": 0.0298, "step": 22360 }, { "grad_norm": 0.22981098294258118, "learning_rate": 9.983513504961608e-05, "loss": 0.0345, "step": 22370 }, { "grad_norm": 0.17891448736190796, "learning_rate": 9.983468753748397e-05, "loss": 0.0343, "step": 22380 }, { "grad_norm": 0.17077529430389404, "learning_rate": 9.983423941981338e-05, "loss": 0.0322, "step": 22390 }, { "grad_norm": 0.2622067332267761, "learning_rate": 9.983379069660971e-05, "loss": 0.0302, "step": 22400 }, { "grad_norm": 0.250853568315506, "learning_rate": 9.983334136787843e-05, "loss": 0.0399, "step": 22410 }, { "grad_norm": 0.2562985420227051, "learning_rate": 9.983289143362503e-05, "loss": 0.0388, "step": 22420 }, { "grad_norm": 0.19202181696891785, "learning_rate": 9.98324408938549e-05, "loss": 0.0353, "step": 22430 }, { "grad_norm": 0.22502025961875916, "learning_rate": 9.98319897485736e-05, "loss": 0.027, "step": 22440 }, { "grad_norm": 0.2944645583629608, "learning_rate": 9.983153799778658e-05, "loss": 0.0282, "step": 22450 }, { "grad_norm": 0.18402154743671417, "learning_rate": 9.98310856414993e-05, "loss": 0.0288, "step": 22460 }, { "grad_norm": 0.23228104412555695, "learning_rate": 9.983063267971729e-05, "loss": 0.031, "step": 22470 }, { "grad_norm": 0.2937447428703308, "learning_rate": 9.983017911244604e-05, "loss": 0.0342, "step": 22480 }, { "grad_norm": 0.28637003898620605, "learning_rate": 9.982972493969108e-05, "loss": 0.03, "step": 22490 }, { "grad_norm": 0.25896093249320984, "learning_rate": 9.982927016145791e-05, "loss": 0.0303, "step": 22500 }, { "grad_norm": 0.23155689239501953, "learning_rate": 9.982881477775206e-05, "loss": 0.0343, "step": 22510 }, { "grad_norm": 0.27736467123031616, "learning_rate": 9.982835878857905e-05, "loss": 0.0374, "step": 22520 }, { "grad_norm": 0.34302884340286255, "learning_rate": 9.982790219394445e-05, "loss": 0.0418, "step": 22530 }, { "grad_norm": 0.2234504222869873, "learning_rate": 9.98274449938538e-05, "loss": 0.0358, "step": 22540 }, { "grad_norm": 0.1912214756011963, "learning_rate": 9.982698718831262e-05, "loss": 0.0437, "step": 22550 }, { "grad_norm": 0.34402987360954285, "learning_rate": 9.982652877732653e-05, "loss": 0.0513, "step": 22560 }, { "grad_norm": 0.20428496599197388, "learning_rate": 9.982606976090107e-05, "loss": 0.0252, "step": 22570 }, { "grad_norm": 0.29350021481513977, "learning_rate": 9.98256101390418e-05, "loss": 0.0329, "step": 22580 }, { "grad_norm": 0.27909135818481445, "learning_rate": 9.982514991175433e-05, "loss": 0.0353, "step": 22590 }, { "grad_norm": 0.2021675407886505, "learning_rate": 9.982468907904426e-05, "loss": 0.0313, "step": 22600 }, { "grad_norm": 0.28154391050338745, "learning_rate": 9.982422764091717e-05, "loss": 0.0296, "step": 22610 }, { "grad_norm": 0.20371876657009125, "learning_rate": 9.982376559737866e-05, "loss": 0.0308, "step": 22620 }, { "grad_norm": 0.2059532254934311, "learning_rate": 9.982330294843436e-05, "loss": 0.0277, "step": 22630 }, { "grad_norm": 0.19940702617168427, "learning_rate": 9.98228396940899e-05, "loss": 0.0349, "step": 22640 }, { "grad_norm": 0.3105680048465729, "learning_rate": 9.982237583435089e-05, "loss": 0.0342, "step": 22650 }, { "grad_norm": 0.38683587312698364, "learning_rate": 9.982191136922296e-05, "loss": 0.0345, "step": 22660 }, { "grad_norm": 0.25428295135498047, "learning_rate": 9.982144629871179e-05, "loss": 0.0299, "step": 22670 }, { "grad_norm": 0.22082604467868805, "learning_rate": 9.982098062282299e-05, "loss": 0.0305, "step": 22680 }, { "grad_norm": 0.40342843532562256, "learning_rate": 9.982051434156225e-05, "loss": 0.0323, "step": 22690 }, { "grad_norm": 0.21492555737495422, "learning_rate": 9.982004745493523e-05, "loss": 0.0266, "step": 22700 }, { "grad_norm": 0.2938617765903473, "learning_rate": 9.981957996294757e-05, "loss": 0.04, "step": 22710 }, { "grad_norm": 0.3680862486362457, "learning_rate": 9.9819111865605e-05, "loss": 0.0341, "step": 22720 }, { "grad_norm": 0.3200622498989105, "learning_rate": 9.981864316291316e-05, "loss": 0.0351, "step": 22730 }, { "grad_norm": 0.3280543088912964, "learning_rate": 9.98181738548778e-05, "loss": 0.0326, "step": 22740 }, { "grad_norm": 0.2847326099872589, "learning_rate": 9.981770394150456e-05, "loss": 0.0376, "step": 22750 }, { "grad_norm": 0.2135103940963745, "learning_rate": 9.98172334227992e-05, "loss": 0.0419, "step": 22760 }, { "grad_norm": 0.26346415281295776, "learning_rate": 9.981676229876741e-05, "loss": 0.0364, "step": 22770 }, { "grad_norm": 0.1487552374601364, "learning_rate": 9.981629056941493e-05, "loss": 0.0281, "step": 22780 }, { "grad_norm": 0.23060546815395355, "learning_rate": 9.981581823474749e-05, "loss": 0.0273, "step": 22790 }, { "grad_norm": 0.26782748103141785, "learning_rate": 9.98153452947708e-05, "loss": 0.0336, "step": 22800 }, { "grad_norm": 0.35119834542274475, "learning_rate": 9.981487174949065e-05, "loss": 0.0343, "step": 22810 }, { "grad_norm": 0.42544320225715637, "learning_rate": 9.981439759891278e-05, "loss": 0.044, "step": 22820 }, { "grad_norm": 0.2105107307434082, "learning_rate": 9.981392284304292e-05, "loss": 0.0449, "step": 22830 }, { "grad_norm": 0.19964812695980072, "learning_rate": 9.98134474818869e-05, "loss": 0.0353, "step": 22840 }, { "grad_norm": 0.3423141837120056, "learning_rate": 9.981297151545044e-05, "loss": 0.0435, "step": 22850 }, { "grad_norm": 0.22474001348018646, "learning_rate": 9.981249494373935e-05, "loss": 0.0388, "step": 22860 }, { "grad_norm": 0.2719127833843231, "learning_rate": 9.981201776675942e-05, "loss": 0.0346, "step": 22870 }, { "grad_norm": 0.29563766717910767, "learning_rate": 9.981153998451645e-05, "loss": 0.0401, "step": 22880 }, { "grad_norm": 0.17985251545906067, "learning_rate": 9.981106159701622e-05, "loss": 0.0294, "step": 22890 }, { "grad_norm": 0.2525409758090973, "learning_rate": 9.981058260426456e-05, "loss": 0.0382, "step": 22900 }, { "grad_norm": 0.2507658004760742, "learning_rate": 9.981010300626732e-05, "loss": 0.0357, "step": 22910 }, { "grad_norm": 0.20616905391216278, "learning_rate": 9.980962280303028e-05, "loss": 0.0369, "step": 22920 }, { "grad_norm": 0.33213576674461365, "learning_rate": 9.980914199455929e-05, "loss": 0.0253, "step": 22930 }, { "grad_norm": 0.2487506866455078, "learning_rate": 9.980866058086021e-05, "loss": 0.025, "step": 22940 }, { "grad_norm": 0.2909976541996002, "learning_rate": 9.980817856193887e-05, "loss": 0.0385, "step": 22950 }, { "grad_norm": 0.21336987614631653, "learning_rate": 9.980769593780115e-05, "loss": 0.0305, "step": 22960 }, { "grad_norm": 0.17971910536289215, "learning_rate": 9.980721270845288e-05, "loss": 0.0271, "step": 22970 }, { "grad_norm": 0.17789678275585175, "learning_rate": 9.980672887389994e-05, "loss": 0.036, "step": 22980 }, { "grad_norm": 0.2519441246986389, "learning_rate": 9.980624443414824e-05, "loss": 0.0347, "step": 22990 }, { "grad_norm": 0.23628434538841248, "learning_rate": 9.980575938920364e-05, "loss": 0.028, "step": 23000 }, { "grad_norm": 0.19812829792499542, "learning_rate": 9.980527373907204e-05, "loss": 0.0304, "step": 23010 }, { "grad_norm": 0.23751425743103027, "learning_rate": 9.980478748375935e-05, "loss": 0.0282, "step": 23020 }, { "grad_norm": 0.2219655066728592, "learning_rate": 9.980430062327144e-05, "loss": 0.026, "step": 23030 }, { "grad_norm": 0.22609089314937592, "learning_rate": 9.980381315761426e-05, "loss": 0.0298, "step": 23040 }, { "grad_norm": 0.2272033393383026, "learning_rate": 9.980332508679375e-05, "loss": 0.0357, "step": 23050 }, { "grad_norm": 0.1884443163871765, "learning_rate": 9.98028364108158e-05, "loss": 0.0298, "step": 23060 }, { "grad_norm": 0.2170446664094925, "learning_rate": 9.980234712968638e-05, "loss": 0.0242, "step": 23070 }, { "grad_norm": 0.24595528841018677, "learning_rate": 9.980185724341142e-05, "loss": 0.0308, "step": 23080 }, { "grad_norm": 0.33708587288856506, "learning_rate": 9.980136675199686e-05, "loss": 0.0306, "step": 23090 }, { "grad_norm": 0.27226465940475464, "learning_rate": 9.980087565544869e-05, "loss": 0.0333, "step": 23100 }, { "grad_norm": 0.21110288798809052, "learning_rate": 9.980038395377284e-05, "loss": 0.036, "step": 23110 }, { "grad_norm": 0.21293410658836365, "learning_rate": 9.979989164697531e-05, "loss": 0.032, "step": 23120 }, { "grad_norm": 0.36423081159591675, "learning_rate": 9.97993987350621e-05, "loss": 0.0324, "step": 23130 }, { "grad_norm": 0.1810254156589508, "learning_rate": 9.979890521803917e-05, "loss": 0.0248, "step": 23140 }, { "grad_norm": 0.2011190801858902, "learning_rate": 9.97984110959125e-05, "loss": 0.049, "step": 23150 }, { "grad_norm": 0.25344884395599365, "learning_rate": 9.979791636868813e-05, "loss": 0.0317, "step": 23160 }, { "grad_norm": 0.29723986983299255, "learning_rate": 9.979742103637206e-05, "loss": 0.029, "step": 23170 }, { "grad_norm": 0.27472084760665894, "learning_rate": 9.979692509897032e-05, "loss": 0.0268, "step": 23180 }, { "grad_norm": 0.23577962815761566, "learning_rate": 9.979642855648893e-05, "loss": 0.0313, "step": 23190 }, { "grad_norm": 0.2632107138633728, "learning_rate": 9.979593140893388e-05, "loss": 0.039, "step": 23200 }, { "grad_norm": 0.2578222155570984, "learning_rate": 9.979543365631128e-05, "loss": 0.0262, "step": 23210 }, { "grad_norm": 0.2060995250940323, "learning_rate": 9.979493529862714e-05, "loss": 0.0344, "step": 23220 }, { "grad_norm": 0.3436312973499298, "learning_rate": 9.979443633588749e-05, "loss": 0.0337, "step": 23230 }, { "grad_norm": 0.19165420532226562, "learning_rate": 9.979393676809847e-05, "loss": 0.0353, "step": 23240 }, { "grad_norm": 0.3317786753177643, "learning_rate": 9.979343659526608e-05, "loss": 0.0366, "step": 23250 }, { "grad_norm": 0.49287861585617065, "learning_rate": 9.979293581739644e-05, "loss": 0.0361, "step": 23260 }, { "grad_norm": 0.25841280817985535, "learning_rate": 9.97924344344956e-05, "loss": 0.0393, "step": 23270 }, { "grad_norm": 0.23481176793575287, "learning_rate": 9.979193244656968e-05, "loss": 0.0358, "step": 23280 }, { "grad_norm": 0.3074663281440735, "learning_rate": 9.979142985362478e-05, "loss": 0.0441, "step": 23290 }, { "grad_norm": 0.2526457905769348, "learning_rate": 9.979092665566697e-05, "loss": 0.039, "step": 23300 }, { "grad_norm": 0.19540998339653015, "learning_rate": 9.97904228527024e-05, "loss": 0.0367, "step": 23310 }, { "grad_norm": 0.3817557990550995, "learning_rate": 9.978991844473719e-05, "loss": 0.0311, "step": 23320 }, { "grad_norm": 0.4851153790950775, "learning_rate": 9.978941343177746e-05, "loss": 0.0381, "step": 23330 }, { "grad_norm": 0.4138534963130951, "learning_rate": 9.978890781382935e-05, "loss": 0.0425, "step": 23340 }, { "grad_norm": 0.2770656943321228, "learning_rate": 9.978840159089899e-05, "loss": 0.0395, "step": 23350 }, { "grad_norm": 0.24679909646511078, "learning_rate": 9.978789476299255e-05, "loss": 0.0335, "step": 23360 }, { "grad_norm": 0.2158975452184677, "learning_rate": 9.97873873301162e-05, "loss": 0.0338, "step": 23370 }, { "grad_norm": 0.2963666021823883, "learning_rate": 9.978687929227606e-05, "loss": 0.0405, "step": 23380 }, { "grad_norm": 0.25057828426361084, "learning_rate": 9.978637064947834e-05, "loss": 0.0422, "step": 23390 }, { "grad_norm": 0.195484921336174, "learning_rate": 9.97858614017292e-05, "loss": 0.0411, "step": 23400 }, { "grad_norm": 0.24206136167049408, "learning_rate": 9.978535154903484e-05, "loss": 0.0436, "step": 23410 }, { "grad_norm": 0.1987524777650833, "learning_rate": 9.978484109140147e-05, "loss": 0.0277, "step": 23420 }, { "grad_norm": 0.16851845383644104, "learning_rate": 9.978433002883527e-05, "loss": 0.0306, "step": 23430 }, { "grad_norm": 0.49006548523902893, "learning_rate": 9.978381836134244e-05, "loss": 0.0293, "step": 23440 }, { "grad_norm": 0.218048095703125, "learning_rate": 9.978330608892923e-05, "loss": 0.0267, "step": 23450 }, { "grad_norm": 0.20195728540420532, "learning_rate": 9.978279321160184e-05, "loss": 0.0322, "step": 23460 }, { "grad_norm": 0.2132238745689392, "learning_rate": 9.978227972936652e-05, "loss": 0.0257, "step": 23470 }, { "grad_norm": 0.1951644867658615, "learning_rate": 9.978176564222948e-05, "loss": 0.0282, "step": 23480 }, { "grad_norm": 0.334697961807251, "learning_rate": 9.978125095019701e-05, "loss": 0.0383, "step": 23490 }, { "grad_norm": 0.4330584406852722, "learning_rate": 9.978073565327533e-05, "loss": 0.046, "step": 23500 }, { "grad_norm": 0.4136599898338318, "learning_rate": 9.978021975147069e-05, "loss": 0.0309, "step": 23510 }, { "grad_norm": 0.22099235653877258, "learning_rate": 9.97797032447894e-05, "loss": 0.055, "step": 23520 }, { "grad_norm": 0.16278882324695587, "learning_rate": 9.977918613323771e-05, "loss": 0.0296, "step": 23530 }, { "grad_norm": 0.20775063335895538, "learning_rate": 9.977866841682192e-05, "loss": 0.0323, "step": 23540 }, { "grad_norm": 0.1690068542957306, "learning_rate": 9.97781500955483e-05, "loss": 0.0242, "step": 23550 }, { "grad_norm": 0.18209385871887207, "learning_rate": 9.977763116942315e-05, "loss": 0.0311, "step": 23560 }, { "grad_norm": 0.23576033115386963, "learning_rate": 9.97771116384528e-05, "loss": 0.0279, "step": 23570 }, { "grad_norm": 0.24708053469657898, "learning_rate": 9.977659150264355e-05, "loss": 0.03, "step": 23580 }, { "grad_norm": 0.35260048508644104, "learning_rate": 9.97760707620017e-05, "loss": 0.0386, "step": 23590 }, { "grad_norm": 0.2066347748041153, "learning_rate": 9.97755494165336e-05, "loss": 0.0341, "step": 23600 }, { "grad_norm": 0.1994493007659912, "learning_rate": 9.977502746624559e-05, "loss": 0.0363, "step": 23610 }, { "grad_norm": 0.16308262944221497, "learning_rate": 9.977450491114398e-05, "loss": 0.0296, "step": 23620 }, { "grad_norm": 0.2801526188850403, "learning_rate": 9.977398175123515e-05, "loss": 0.0304, "step": 23630 }, { "grad_norm": 0.31334707140922546, "learning_rate": 9.977345798652545e-05, "loss": 0.0348, "step": 23640 }, { "grad_norm": 0.3190321624279022, "learning_rate": 9.977293361702125e-05, "loss": 0.0324, "step": 23650 }, { "grad_norm": 0.3063831329345703, "learning_rate": 9.97724086427289e-05, "loss": 0.0405, "step": 23660 }, { "grad_norm": 0.17978321015834808, "learning_rate": 9.97718830636548e-05, "loss": 0.0343, "step": 23670 }, { "grad_norm": 0.2595088481903076, "learning_rate": 9.977135687980531e-05, "loss": 0.0353, "step": 23680 }, { "grad_norm": 0.32254427671432495, "learning_rate": 9.977083009118686e-05, "loss": 0.0325, "step": 23690 }, { "grad_norm": 0.36683234572410583, "learning_rate": 9.977030269780583e-05, "loss": 0.0319, "step": 23700 }, { "grad_norm": 0.27563896775245667, "learning_rate": 9.976977469966862e-05, "loss": 0.0356, "step": 23710 }, { "grad_norm": 0.21067741513252258, "learning_rate": 9.976924609678167e-05, "loss": 0.0286, "step": 23720 }, { "grad_norm": 0.5279686450958252, "learning_rate": 9.976871688915139e-05, "loss": 0.0359, "step": 23730 }, { "grad_norm": 0.22048108279705048, "learning_rate": 9.97681870767842e-05, "loss": 0.0311, "step": 23740 }, { "grad_norm": 0.4407254457473755, "learning_rate": 9.976765665968654e-05, "loss": 0.0307, "step": 23750 }, { "grad_norm": 1.1628741025924683, "learning_rate": 9.97671256378649e-05, "loss": 0.0401, "step": 23760 }, { "grad_norm": 0.22240756452083588, "learning_rate": 9.976659401132566e-05, "loss": 0.0348, "step": 23770 }, { "grad_norm": 0.25544124841690063, "learning_rate": 9.976606178007532e-05, "loss": 0.0342, "step": 23780 }, { "grad_norm": 0.2555626630783081, "learning_rate": 9.976552894412037e-05, "loss": 0.035, "step": 23790 }, { "grad_norm": 0.181696355342865, "learning_rate": 9.976499550346723e-05, "loss": 0.0399, "step": 23800 }, { "grad_norm": 0.15680049359798431, "learning_rate": 9.976446145812242e-05, "loss": 0.0283, "step": 23810 }, { "grad_norm": 0.21564044058322906, "learning_rate": 9.976392680809242e-05, "loss": 0.0325, "step": 23820 }, { "grad_norm": 0.19648706912994385, "learning_rate": 9.976339155338371e-05, "loss": 0.0258, "step": 23830 }, { "grad_norm": 0.22698110342025757, "learning_rate": 9.976285569400281e-05, "loss": 0.0266, "step": 23840 }, { "grad_norm": 0.42763665318489075, "learning_rate": 9.976231922995624e-05, "loss": 0.0259, "step": 23850 }, { "grad_norm": 0.15511901676654816, "learning_rate": 9.97617821612505e-05, "loss": 0.0303, "step": 23860 }, { "grad_norm": 0.2335374355316162, "learning_rate": 9.976124448789214e-05, "loss": 0.0359, "step": 23870 }, { "grad_norm": 0.30330148339271545, "learning_rate": 9.976070620988767e-05, "loss": 0.0278, "step": 23880 }, { "grad_norm": 0.30540359020233154, "learning_rate": 9.976016732724362e-05, "loss": 0.0353, "step": 23890 }, { "grad_norm": 0.2260686308145523, "learning_rate": 9.975962783996657e-05, "loss": 0.0306, "step": 23900 }, { "grad_norm": 0.20873752236366272, "learning_rate": 9.975908774806307e-05, "loss": 0.029, "step": 23910 }, { "grad_norm": 0.28440403938293457, "learning_rate": 9.975854705153965e-05, "loss": 0.0345, "step": 23920 }, { "grad_norm": 0.14420686662197113, "learning_rate": 9.975800575040291e-05, "loss": 0.0285, "step": 23930 }, { "grad_norm": 0.2516545355319977, "learning_rate": 9.975746384465943e-05, "loss": 0.0289, "step": 23940 }, { "grad_norm": 0.15243405103683472, "learning_rate": 9.975692133431579e-05, "loss": 0.0233, "step": 23950 }, { "grad_norm": 0.26020169258117676, "learning_rate": 9.975637821937857e-05, "loss": 0.0354, "step": 23960 }, { "grad_norm": 0.1818939447402954, "learning_rate": 9.975583449985437e-05, "loss": 0.0345, "step": 23970 }, { "grad_norm": 0.2815220355987549, "learning_rate": 9.97552901757498e-05, "loss": 0.0259, "step": 23980 }, { "grad_norm": 0.20720462501049042, "learning_rate": 9.97547452470715e-05, "loss": 0.0281, "step": 23990 }, { "grad_norm": 0.25382572412490845, "learning_rate": 9.975419971382606e-05, "loss": 0.03, "step": 24000 }, { "grad_norm": 0.28709298372268677, "learning_rate": 9.97536535760201e-05, "loss": 0.0288, "step": 24010 }, { "grad_norm": 0.31668511033058167, "learning_rate": 9.97531068336603e-05, "loss": 0.0333, "step": 24020 }, { "grad_norm": 0.18038709461688995, "learning_rate": 9.975255948675326e-05, "loss": 0.0258, "step": 24030 }, { "grad_norm": 0.21669358015060425, "learning_rate": 9.975201153530566e-05, "loss": 0.03, "step": 24040 }, { "grad_norm": 0.23389801383018494, "learning_rate": 9.975146297932414e-05, "loss": 0.0273, "step": 24050 }, { "grad_norm": 0.14209511876106262, "learning_rate": 9.975091381881536e-05, "loss": 0.0322, "step": 24060 }, { "grad_norm": 0.22243139147758484, "learning_rate": 9.975036405378602e-05, "loss": 0.0282, "step": 24070 }, { "grad_norm": 0.22136519849300385, "learning_rate": 9.974981368424278e-05, "loss": 0.0317, "step": 24080 }, { "grad_norm": 0.1645844280719757, "learning_rate": 9.974926271019234e-05, "loss": 0.0233, "step": 24090 }, { "grad_norm": 0.19590628147125244, "learning_rate": 9.974871113164137e-05, "loss": 0.0291, "step": 24100 }, { "grad_norm": 0.16410112380981445, "learning_rate": 9.97481589485966e-05, "loss": 0.0292, "step": 24110 }, { "grad_norm": 0.2060343325138092, "learning_rate": 9.974760616106473e-05, "loss": 0.0225, "step": 24120 }, { "grad_norm": 0.18219032883644104, "learning_rate": 9.974705276905248e-05, "loss": 0.0349, "step": 24130 }, { "grad_norm": 0.1878705620765686, "learning_rate": 9.974649877256656e-05, "loss": 0.0335, "step": 24140 }, { "grad_norm": 0.35776835680007935, "learning_rate": 9.974594417161371e-05, "loss": 0.0281, "step": 24150 }, { "grad_norm": 0.1769046187400818, "learning_rate": 9.974538896620067e-05, "loss": 0.0275, "step": 24160 }, { "grad_norm": 0.3452679216861725, "learning_rate": 9.97448331563342e-05, "loss": 0.0274, "step": 24170 }, { "grad_norm": 0.16507025063037872, "learning_rate": 9.974427674202102e-05, "loss": 0.0231, "step": 24180 }, { "grad_norm": 0.18726296722888947, "learning_rate": 9.974371972326792e-05, "loss": 0.027, "step": 24190 }, { "grad_norm": 0.2792315185070038, "learning_rate": 9.974316210008167e-05, "loss": 0.0308, "step": 24200 }, { "grad_norm": 0.3109993636608124, "learning_rate": 9.974260387246903e-05, "loss": 0.0401, "step": 24210 }, { "grad_norm": 0.19947119057178497, "learning_rate": 9.974204504043679e-05, "loss": 0.0401, "step": 24220 }, { "grad_norm": 0.1733974665403366, "learning_rate": 9.974148560399173e-05, "loss": 0.0386, "step": 24230 }, { "grad_norm": 0.15725083649158478, "learning_rate": 9.974092556314065e-05, "loss": 0.0249, "step": 24240 }, { "grad_norm": 0.2496708184480667, "learning_rate": 9.974036491789038e-05, "loss": 0.0298, "step": 24250 }, { "grad_norm": 0.3415715992450714, "learning_rate": 9.973980366824772e-05, "loss": 0.0309, "step": 24260 }, { "grad_norm": 0.23069874942302704, "learning_rate": 9.973924181421945e-05, "loss": 0.0303, "step": 24270 }, { "grad_norm": 0.26897507905960083, "learning_rate": 9.973867935581245e-05, "loss": 0.0311, "step": 24280 }, { "grad_norm": 0.30491524934768677, "learning_rate": 9.973811629303354e-05, "loss": 0.0328, "step": 24290 }, { "grad_norm": 0.2322584092617035, "learning_rate": 9.973755262588955e-05, "loss": 0.0346, "step": 24300 }, { "grad_norm": 0.2807333767414093, "learning_rate": 9.973698835438733e-05, "loss": 0.0302, "step": 24310 }, { "grad_norm": 0.20768433809280396, "learning_rate": 9.973642347853374e-05, "loss": 0.0263, "step": 24320 }, { "grad_norm": 0.2359740287065506, "learning_rate": 9.973585799833568e-05, "loss": 0.0317, "step": 24330 }, { "grad_norm": 0.24002932012081146, "learning_rate": 9.973529191379993e-05, "loss": 0.0385, "step": 24340 }, { "grad_norm": 0.27748042345046997, "learning_rate": 9.973472522493347e-05, "loss": 0.0353, "step": 24350 }, { "grad_norm": 0.207057386636734, "learning_rate": 9.973415793174313e-05, "loss": 0.0295, "step": 24360 }, { "grad_norm": 0.20381151139736176, "learning_rate": 9.97335900342358e-05, "loss": 0.0333, "step": 24370 }, { "grad_norm": 0.2198643684387207, "learning_rate": 9.973302153241841e-05, "loss": 0.0313, "step": 24380 }, { "grad_norm": 0.24828645586967468, "learning_rate": 9.973245242629786e-05, "loss": 0.0297, "step": 24390 }, { "grad_norm": 0.2501903474330902, "learning_rate": 9.973188271588103e-05, "loss": 0.0315, "step": 24400 }, { "grad_norm": 0.1614692658185959, "learning_rate": 9.973131240117489e-05, "loss": 0.0337, "step": 24410 }, { "grad_norm": 0.1765977442264557, "learning_rate": 9.973074148218636e-05, "loss": 0.0322, "step": 24420 }, { "grad_norm": 0.30095845460891724, "learning_rate": 9.973016995892234e-05, "loss": 0.0327, "step": 24430 }, { "grad_norm": 0.311501145362854, "learning_rate": 9.972959783138981e-05, "loss": 0.0326, "step": 24440 }, { "grad_norm": 0.21733485162258148, "learning_rate": 9.972902509959572e-05, "loss": 0.0334, "step": 24450 }, { "grad_norm": 0.24493400752544403, "learning_rate": 9.972845176354703e-05, "loss": 0.03, "step": 24460 }, { "grad_norm": 0.3559199571609497, "learning_rate": 9.972787782325068e-05, "loss": 0.0375, "step": 24470 }, { "grad_norm": 0.18528375029563904, "learning_rate": 9.97273032787137e-05, "loss": 0.0268, "step": 24480 }, { "grad_norm": 0.3578750193119049, "learning_rate": 9.972672812994299e-05, "loss": 0.0296, "step": 24490 }, { "grad_norm": 0.15406933426856995, "learning_rate": 9.97261523769456e-05, "loss": 0.0247, "step": 24500 }, { "grad_norm": 0.17705687880516052, "learning_rate": 9.972557601972852e-05, "loss": 0.0271, "step": 24510 }, { "grad_norm": 0.26365727186203003, "learning_rate": 9.972499905829875e-05, "loss": 0.0369, "step": 24520 }, { "grad_norm": 0.17473207414150238, "learning_rate": 9.972442149266326e-05, "loss": 0.0317, "step": 24530 }, { "grad_norm": 0.2574896216392517, "learning_rate": 9.972384332282914e-05, "loss": 0.0352, "step": 24540 }, { "grad_norm": 0.22718371450901031, "learning_rate": 9.972326454880334e-05, "loss": 0.0307, "step": 24550 }, { "grad_norm": 0.17925065755844116, "learning_rate": 9.972268517059296e-05, "loss": 0.0291, "step": 24560 }, { "grad_norm": 0.25437629222869873, "learning_rate": 9.9722105188205e-05, "loss": 0.0236, "step": 24570 }, { "grad_norm": 0.2712566554546356, "learning_rate": 9.97215246016465e-05, "loss": 0.03, "step": 24580 }, { "grad_norm": 0.19288380444049835, "learning_rate": 9.972094341092455e-05, "loss": 0.0409, "step": 24590 }, { "grad_norm": 0.21629361808300018, "learning_rate": 9.972036161604618e-05, "loss": 0.0313, "step": 24600 }, { "grad_norm": 0.33479031920433044, "learning_rate": 9.971977921701848e-05, "loss": 0.0387, "step": 24610 }, { "grad_norm": 0.11255202442407608, "learning_rate": 9.971919621384851e-05, "loss": 0.0267, "step": 24620 }, { "grad_norm": 0.3129881024360657, "learning_rate": 9.971861260654338e-05, "loss": 0.0495, "step": 24630 }, { "grad_norm": 0.39097359776496887, "learning_rate": 9.971802839511015e-05, "loss": 0.0355, "step": 24640 }, { "grad_norm": 0.25430208444595337, "learning_rate": 9.971744357955595e-05, "loss": 0.0362, "step": 24650 }, { "grad_norm": 0.2715948820114136, "learning_rate": 9.971685815988786e-05, "loss": 0.0469, "step": 24660 }, { "grad_norm": 0.27203938364982605, "learning_rate": 9.9716272136113e-05, "loss": 0.0321, "step": 24670 }, { "grad_norm": 0.23588308691978455, "learning_rate": 9.971568550823849e-05, "loss": 0.0324, "step": 24680 }, { "grad_norm": 0.19563698768615723, "learning_rate": 9.971509827627147e-05, "loss": 0.0307, "step": 24690 }, { "grad_norm": 0.23706717789173126, "learning_rate": 9.971451044021906e-05, "loss": 0.0303, "step": 24700 }, { "grad_norm": 0.2153143733739853, "learning_rate": 9.971392200008842e-05, "loss": 0.0292, "step": 24710 }, { "grad_norm": 0.21830396354198456, "learning_rate": 9.971333295588667e-05, "loss": 0.0282, "step": 24720 }, { "grad_norm": 0.1929146945476532, "learning_rate": 9.971274330762102e-05, "loss": 0.0236, "step": 24730 }, { "grad_norm": 0.17810463905334473, "learning_rate": 9.971215305529858e-05, "loss": 0.026, "step": 24740 }, { "grad_norm": 0.19012923538684845, "learning_rate": 9.971156219892654e-05, "loss": 0.034, "step": 24750 }, { "grad_norm": 0.2865518629550934, "learning_rate": 9.97109707385121e-05, "loss": 0.0306, "step": 24760 }, { "grad_norm": 0.29181718826293945, "learning_rate": 9.971037867406241e-05, "loss": 0.0371, "step": 24770 }, { "grad_norm": 0.23267006874084473, "learning_rate": 9.97097860055847e-05, "loss": 0.0331, "step": 24780 }, { "grad_norm": 0.2412971705198288, "learning_rate": 9.970919273308615e-05, "loss": 0.0339, "step": 24790 }, { "grad_norm": 0.2598964273929596, "learning_rate": 9.970859885657399e-05, "loss": 0.0314, "step": 24800 }, { "grad_norm": 0.35883432626724243, "learning_rate": 9.970800437605541e-05, "loss": 0.0301, "step": 24810 }, { "grad_norm": 0.18585705757141113, "learning_rate": 9.970740929153763e-05, "loss": 0.0305, "step": 24820 }, { "grad_norm": 0.25867053866386414, "learning_rate": 9.970681360302791e-05, "loss": 0.0356, "step": 24830 }, { "grad_norm": 0.2469307780265808, "learning_rate": 9.970621731053348e-05, "loss": 0.0283, "step": 24840 }, { "grad_norm": 0.23569656908512115, "learning_rate": 9.970562041406158e-05, "loss": 0.0342, "step": 24850 }, { "grad_norm": 0.19905339181423187, "learning_rate": 9.970502291361946e-05, "loss": 0.0365, "step": 24860 }, { "grad_norm": 0.2317948043346405, "learning_rate": 9.970442480921436e-05, "loss": 0.0379, "step": 24870 }, { "grad_norm": 0.17604660987854004, "learning_rate": 9.97038261008536e-05, "loss": 0.0316, "step": 24880 }, { "grad_norm": 0.1747732311487198, "learning_rate": 9.97032267885444e-05, "loss": 0.0343, "step": 24890 }, { "grad_norm": 0.23702971637248993, "learning_rate": 9.970262687229408e-05, "loss": 0.0358, "step": 24900 }, { "grad_norm": 0.23432834446430206, "learning_rate": 9.970202635210992e-05, "loss": 0.0295, "step": 24910 }, { "grad_norm": 0.2576446533203125, "learning_rate": 9.970142522799921e-05, "loss": 0.0265, "step": 24920 }, { "grad_norm": 0.2736454904079437, "learning_rate": 9.970082349996925e-05, "loss": 0.032, "step": 24930 }, { "grad_norm": 0.2891802489757538, "learning_rate": 9.970022116802737e-05, "loss": 0.0347, "step": 24940 }, { "grad_norm": 0.29461634159088135, "learning_rate": 9.969961823218087e-05, "loss": 0.0351, "step": 24950 }, { "grad_norm": 0.31983253359794617, "learning_rate": 9.969901469243708e-05, "loss": 0.0321, "step": 24960 }, { "grad_norm": 0.5907434225082397, "learning_rate": 9.969841054880336e-05, "loss": 0.0334, "step": 24970 }, { "grad_norm": 0.18235591053962708, "learning_rate": 9.9697805801287e-05, "loss": 0.0287, "step": 24980 }, { "grad_norm": 0.29475221037864685, "learning_rate": 9.96972004498954e-05, "loss": 0.047, "step": 24990 }, { "grad_norm": 0.28804174065589905, "learning_rate": 9.969659449463588e-05, "loss": 0.0377, "step": 25000 }, { "grad_norm": 0.2420758157968521, "learning_rate": 9.969598793551582e-05, "loss": 0.0331, "step": 25010 }, { "grad_norm": 0.42263951897621155, "learning_rate": 9.969538077254259e-05, "loss": 0.0332, "step": 25020 }, { "grad_norm": 0.25016042590141296, "learning_rate": 9.969477300572356e-05, "loss": 0.0428, "step": 25030 }, { "grad_norm": 0.2612111270427704, "learning_rate": 9.969416463506612e-05, "loss": 0.0282, "step": 25040 }, { "grad_norm": 0.2684030830860138, "learning_rate": 9.969355566057766e-05, "loss": 0.0331, "step": 25050 }, { "grad_norm": 0.24081169068813324, "learning_rate": 9.969294608226556e-05, "loss": 0.0269, "step": 25060 }, { "grad_norm": 0.1949440985918045, "learning_rate": 9.969233590013727e-05, "loss": 0.0282, "step": 25070 }, { "grad_norm": 0.1937321573495865, "learning_rate": 9.969172511420017e-05, "loss": 0.032, "step": 25080 }, { "grad_norm": 0.2777426242828369, "learning_rate": 9.969111372446171e-05, "loss": 0.0339, "step": 25090 }, { "grad_norm": 0.1670815497636795, "learning_rate": 9.969050173092928e-05, "loss": 0.0306, "step": 25100 }, { "grad_norm": 0.1529490351676941, "learning_rate": 9.968988913361035e-05, "loss": 0.0311, "step": 25110 }, { "grad_norm": 0.2159547507762909, "learning_rate": 9.968927593251235e-05, "loss": 0.0344, "step": 25120 }, { "grad_norm": 0.21838247776031494, "learning_rate": 9.968866212764274e-05, "loss": 0.0427, "step": 25130 }, { "grad_norm": 0.19161415100097656, "learning_rate": 9.968804771900897e-05, "loss": 0.0262, "step": 25140 }, { "grad_norm": 0.16549286246299744, "learning_rate": 9.96874327066185e-05, "loss": 0.0269, "step": 25150 }, { "grad_norm": 0.12523221969604492, "learning_rate": 9.968681709047881e-05, "loss": 0.0262, "step": 25160 }, { "grad_norm": 0.21843403577804565, "learning_rate": 9.968620087059738e-05, "loss": 0.0292, "step": 25170 }, { "grad_norm": 0.39678215980529785, "learning_rate": 9.968558404698171e-05, "loss": 0.027, "step": 25180 }, { "grad_norm": 0.18838849663734436, "learning_rate": 9.968496661963926e-05, "loss": 0.0315, "step": 25190 }, { "grad_norm": 0.14886842668056488, "learning_rate": 9.968434858857758e-05, "loss": 0.0222, "step": 25200 }, { "grad_norm": 0.21801310777664185, "learning_rate": 9.968372995380415e-05, "loss": 0.0332, "step": 25210 }, { "grad_norm": 0.23129388689994812, "learning_rate": 9.968311071532647e-05, "loss": 0.0296, "step": 25220 }, { "grad_norm": 0.3162880837917328, "learning_rate": 9.968249087315212e-05, "loss": 0.03, "step": 25230 }, { "grad_norm": 0.5211525559425354, "learning_rate": 9.968187042728857e-05, "loss": 0.0287, "step": 25240 }, { "grad_norm": 0.28956788778305054, "learning_rate": 9.968124937774339e-05, "loss": 0.0283, "step": 25250 }, { "grad_norm": 0.22613660991191864, "learning_rate": 9.968062772452413e-05, "loss": 0.0301, "step": 25260 }, { "grad_norm": 0.21929007768630981, "learning_rate": 9.968000546763834e-05, "loss": 0.0285, "step": 25270 }, { "grad_norm": 0.19933009147644043, "learning_rate": 9.967938260709357e-05, "loss": 0.0258, "step": 25280 }, { "grad_norm": 0.24087689816951752, "learning_rate": 9.96787591428974e-05, "loss": 0.027, "step": 25290 }, { "grad_norm": 0.24663588404655457, "learning_rate": 9.967813507505741e-05, "loss": 0.0286, "step": 25300 }, { "grad_norm": 0.2613790035247803, "learning_rate": 9.967751040358116e-05, "loss": 0.0295, "step": 25310 }, { "grad_norm": 0.22466248273849487, "learning_rate": 9.967688512847626e-05, "loss": 0.0319, "step": 25320 }, { "grad_norm": 0.1731196641921997, "learning_rate": 9.967625924975031e-05, "loss": 0.0332, "step": 25330 }, { "grad_norm": 0.2536607086658478, "learning_rate": 9.967563276741091e-05, "loss": 0.0335, "step": 25340 }, { "grad_norm": 0.2943989336490631, "learning_rate": 9.967500568146566e-05, "loss": 0.0312, "step": 25350 }, { "grad_norm": 0.21900340914726257, "learning_rate": 9.96743779919222e-05, "loss": 0.0349, "step": 25360 }, { "grad_norm": 0.24665884673595428, "learning_rate": 9.967374969878814e-05, "loss": 0.0331, "step": 25370 }, { "grad_norm": 0.29255402088165283, "learning_rate": 9.967312080207112e-05, "loss": 0.0337, "step": 25380 }, { "grad_norm": 0.19340777397155762, "learning_rate": 9.967249130177878e-05, "loss": 0.0308, "step": 25390 }, { "grad_norm": 0.24039341509342194, "learning_rate": 9.967186119791878e-05, "loss": 0.0333, "step": 25400 }, { "grad_norm": 0.2248096615076065, "learning_rate": 9.967123049049877e-05, "loss": 0.0267, "step": 25410 }, { "grad_norm": 0.2639428377151489, "learning_rate": 9.96705991795264e-05, "loss": 0.0274, "step": 25420 }, { "grad_norm": 0.4002433121204376, "learning_rate": 9.966996726500936e-05, "loss": 0.0308, "step": 25430 }, { "grad_norm": 0.18840976059436798, "learning_rate": 9.966933474695532e-05, "loss": 0.0297, "step": 25440 }, { "grad_norm": 0.17637574672698975, "learning_rate": 9.966870162537199e-05, "loss": 0.0388, "step": 25450 }, { "grad_norm": 0.17958183586597443, "learning_rate": 9.966806790026701e-05, "loss": 0.0237, "step": 25460 }, { "grad_norm": 0.2390131801366806, "learning_rate": 9.966743357164811e-05, "loss": 0.0336, "step": 25470 }, { "grad_norm": 0.24743080139160156, "learning_rate": 9.966679863952302e-05, "loss": 0.0271, "step": 25480 }, { "grad_norm": 0.17080079019069672, "learning_rate": 9.966616310389942e-05, "loss": 0.0333, "step": 25490 }, { "grad_norm": 0.2704302966594696, "learning_rate": 9.966552696478504e-05, "loss": 0.035, "step": 25500 }, { "grad_norm": 0.20743292570114136, "learning_rate": 9.966489022218763e-05, "loss": 0.0303, "step": 25510 }, { "grad_norm": 0.4717515707015991, "learning_rate": 9.96642528761149e-05, "loss": 0.0324, "step": 25520 }, { "grad_norm": 0.21734030544757843, "learning_rate": 9.966361492657461e-05, "loss": 0.0347, "step": 25530 }, { "grad_norm": 0.29356521368026733, "learning_rate": 9.966297637357452e-05, "loss": 0.0404, "step": 25540 }, { "grad_norm": 0.2081044614315033, "learning_rate": 9.966233721712237e-05, "loss": 0.0356, "step": 25550 }, { "grad_norm": 0.22160758078098297, "learning_rate": 9.966169745722593e-05, "loss": 0.0431, "step": 25560 }, { "grad_norm": 0.30821746587753296, "learning_rate": 9.966105709389297e-05, "loss": 0.0351, "step": 25570 }, { "grad_norm": 0.20571079850196838, "learning_rate": 9.966041612713129e-05, "loss": 0.0351, "step": 25580 }, { "grad_norm": 0.2142748087644577, "learning_rate": 9.965977455694867e-05, "loss": 0.0323, "step": 25590 }, { "grad_norm": 0.23291094601154327, "learning_rate": 9.96591323833529e-05, "loss": 0.0253, "step": 25600 }, { "grad_norm": 0.18318335711956024, "learning_rate": 9.965848960635179e-05, "loss": 0.0313, "step": 25610 }, { "grad_norm": 0.21538612246513367, "learning_rate": 9.965784622595313e-05, "loss": 0.0251, "step": 25620 }, { "grad_norm": 0.22922883927822113, "learning_rate": 9.965720224216476e-05, "loss": 0.0236, "step": 25630 }, { "grad_norm": 0.20328639447689056, "learning_rate": 9.96565576549945e-05, "loss": 0.0213, "step": 25640 }, { "grad_norm": 0.2240392565727234, "learning_rate": 9.965591246445018e-05, "loss": 0.0271, "step": 25650 }, { "grad_norm": 0.16973261535167694, "learning_rate": 9.965526667053963e-05, "loss": 0.0243, "step": 25660 }, { "grad_norm": 0.14820460975170135, "learning_rate": 9.965462027327073e-05, "loss": 0.0308, "step": 25670 }, { "grad_norm": 0.17432056367397308, "learning_rate": 9.965397327265132e-05, "loss": 0.0196, "step": 25680 }, { "grad_norm": 0.19204801321029663, "learning_rate": 9.965332566868923e-05, "loss": 0.0256, "step": 25690 }, { "grad_norm": 0.261701762676239, "learning_rate": 9.965267746139235e-05, "loss": 0.0335, "step": 25700 }, { "grad_norm": 0.18944163620471954, "learning_rate": 9.965202865076857e-05, "loss": 0.0292, "step": 25710 }, { "grad_norm": 0.35436907410621643, "learning_rate": 9.965137923682577e-05, "loss": 0.0306, "step": 25720 }, { "grad_norm": 0.2117391973733902, "learning_rate": 9.965072921957181e-05, "loss": 0.0345, "step": 25730 }, { "grad_norm": 0.22670985758304596, "learning_rate": 9.965007859901464e-05, "loss": 0.0318, "step": 25740 }, { "grad_norm": 0.23694993555545807, "learning_rate": 9.964942737516211e-05, "loss": 0.0275, "step": 25750 }, { "grad_norm": 0.2686862349510193, "learning_rate": 9.964877554802217e-05, "loss": 0.025, "step": 25760 }, { "grad_norm": 0.17134955525398254, "learning_rate": 9.964812311760275e-05, "loss": 0.0307, "step": 25770 }, { "grad_norm": 0.18377076089382172, "learning_rate": 9.964747008391173e-05, "loss": 0.0299, "step": 25780 }, { "grad_norm": 0.19515812397003174, "learning_rate": 9.964681644695708e-05, "loss": 0.0272, "step": 25790 }, { "grad_norm": 0.23734919726848602, "learning_rate": 9.964616220674675e-05, "loss": 0.025, "step": 25800 }, { "grad_norm": 0.16969934105873108, "learning_rate": 9.964550736328866e-05, "loss": 0.0238, "step": 25810 }, { "grad_norm": 0.7528077363967896, "learning_rate": 9.964485191659079e-05, "loss": 0.0284, "step": 25820 }, { "grad_norm": 0.13992610573768616, "learning_rate": 9.964419586666109e-05, "loss": 0.032, "step": 25830 }, { "grad_norm": 0.20863208174705505, "learning_rate": 9.964353921350755e-05, "loss": 0.0332, "step": 25840 }, { "grad_norm": 0.19360369443893433, "learning_rate": 9.964288195713813e-05, "loss": 0.0274, "step": 25850 }, { "grad_norm": 0.2076793611049652, "learning_rate": 9.964222409756083e-05, "loss": 0.0246, "step": 25860 }, { "grad_norm": 0.13133355975151062, "learning_rate": 9.964156563478364e-05, "loss": 0.0297, "step": 25870 }, { "grad_norm": 0.2930675446987152, "learning_rate": 9.964090656881455e-05, "loss": 0.0349, "step": 25880 }, { "grad_norm": 0.25719648599624634, "learning_rate": 9.964024689966158e-05, "loss": 0.0312, "step": 25890 }, { "grad_norm": 0.3152405619621277, "learning_rate": 9.963958662733274e-05, "loss": 0.0224, "step": 25900 }, { "grad_norm": 0.40447163581848145, "learning_rate": 9.963892575183604e-05, "loss": 0.036, "step": 25910 }, { "grad_norm": 0.2299501597881317, "learning_rate": 9.963826427317955e-05, "loss": 0.0253, "step": 25920 }, { "grad_norm": 0.1820739358663559, "learning_rate": 9.963760219137127e-05, "loss": 0.0327, "step": 25930 }, { "grad_norm": 0.17887017130851746, "learning_rate": 9.963693950641926e-05, "loss": 0.0303, "step": 25940 }, { "grad_norm": 0.257904052734375, "learning_rate": 9.963627621833157e-05, "loss": 0.0312, "step": 25950 }, { "grad_norm": 0.23013122379779816, "learning_rate": 9.963561232711626e-05, "loss": 0.0284, "step": 25960 }, { "grad_norm": 0.2899428904056549, "learning_rate": 9.963494783278141e-05, "loss": 0.0302, "step": 25970 }, { "grad_norm": 0.23692378401756287, "learning_rate": 9.963428273533506e-05, "loss": 0.0339, "step": 25980 }, { "grad_norm": 0.1964387595653534, "learning_rate": 9.963361703478531e-05, "loss": 0.0297, "step": 25990 }, { "grad_norm": 0.19051550328731537, "learning_rate": 9.963295073114026e-05, "loss": 0.028, "step": 26000 }, { "grad_norm": 0.2666507959365845, "learning_rate": 9.9632283824408e-05, "loss": 0.0353, "step": 26010 }, { "grad_norm": 0.2551060914993286, "learning_rate": 9.963161631459663e-05, "loss": 0.0279, "step": 26020 }, { "grad_norm": 0.2823573052883148, "learning_rate": 9.963094820171426e-05, "loss": 0.029, "step": 26030 }, { "grad_norm": 0.19810214638710022, "learning_rate": 9.963027948576902e-05, "loss": 0.0284, "step": 26040 }, { "grad_norm": 0.31286656856536865, "learning_rate": 9.9629610166769e-05, "loss": 0.0315, "step": 26050 }, { "grad_norm": 0.3688931167125702, "learning_rate": 9.962894024472237e-05, "loss": 0.0268, "step": 26060 }, { "grad_norm": 0.28797832131385803, "learning_rate": 9.962826971963728e-05, "loss": 0.0316, "step": 26070 }, { "grad_norm": 0.5512691140174866, "learning_rate": 9.962759859152184e-05, "loss": 0.0472, "step": 26080 }, { "grad_norm": 0.2343982756137848, "learning_rate": 9.962692686038422e-05, "loss": 0.0369, "step": 26090 }, { "grad_norm": 0.26206517219543457, "learning_rate": 9.962625452623258e-05, "loss": 0.0338, "step": 26100 }, { "grad_norm": 0.16762588918209076, "learning_rate": 9.962558158907509e-05, "loss": 0.0257, "step": 26110 }, { "grad_norm": 0.18098489940166473, "learning_rate": 9.962490804891993e-05, "loss": 0.028, "step": 26120 }, { "grad_norm": 0.21180005371570587, "learning_rate": 9.962423390577529e-05, "loss": 0.0292, "step": 26130 }, { "grad_norm": 0.24630960822105408, "learning_rate": 9.962355915964936e-05, "loss": 0.0325, "step": 26140 }, { "grad_norm": 0.21503575146198273, "learning_rate": 9.962288381055031e-05, "loss": 0.0396, "step": 26150 }, { "grad_norm": 0.3211055099964142, "learning_rate": 9.96222078584864e-05, "loss": 0.0313, "step": 26160 }, { "grad_norm": 0.1881338506937027, "learning_rate": 9.96215313034658e-05, "loss": 0.0365, "step": 26170 }, { "grad_norm": 0.16267584264278412, "learning_rate": 9.962085414549673e-05, "loss": 0.0287, "step": 26180 }, { "grad_norm": 0.5505672097206116, "learning_rate": 9.962017638458746e-05, "loss": 0.0338, "step": 26190 }, { "grad_norm": 0.2723899483680725, "learning_rate": 9.961949802074618e-05, "loss": 0.033, "step": 26200 }, { "grad_norm": 0.2206263691186905, "learning_rate": 9.961881905398117e-05, "loss": 0.0319, "step": 26210 }, { "grad_norm": 0.187630295753479, "learning_rate": 9.961813948430064e-05, "loss": 0.0383, "step": 26220 }, { "grad_norm": 0.23793576657772064, "learning_rate": 9.961745931171287e-05, "loss": 0.0301, "step": 26230 }, { "grad_norm": 0.23425692319869995, "learning_rate": 9.961677853622613e-05, "loss": 0.0258, "step": 26240 }, { "grad_norm": 0.1455235481262207, "learning_rate": 9.961609715784868e-05, "loss": 0.0315, "step": 26250 }, { "grad_norm": 0.3001859486103058, "learning_rate": 9.96154151765888e-05, "loss": 0.0318, "step": 26260 }, { "grad_norm": 0.14365339279174805, "learning_rate": 9.961473259245478e-05, "loss": 0.0294, "step": 26270 }, { "grad_norm": 0.13565056025981903, "learning_rate": 9.961404940545492e-05, "loss": 0.0451, "step": 26280 }, { "grad_norm": 0.24922525882720947, "learning_rate": 9.961336561559753e-05, "loss": 0.0409, "step": 26290 }, { "grad_norm": 0.1734432876110077, "learning_rate": 9.961268122289089e-05, "loss": 0.0368, "step": 26300 }, { "grad_norm": 0.2566494047641754, "learning_rate": 9.961199622734333e-05, "loss": 0.0461, "step": 26310 }, { "grad_norm": 0.25377804040908813, "learning_rate": 9.961131062896318e-05, "loss": 0.0348, "step": 26320 }, { "grad_norm": 0.1786688268184662, "learning_rate": 9.961062442775876e-05, "loss": 0.0329, "step": 26330 }, { "grad_norm": 0.2955072820186615, "learning_rate": 9.960993762373842e-05, "loss": 0.0383, "step": 26340 }, { "grad_norm": 0.1944660246372223, "learning_rate": 9.960925021691049e-05, "loss": 0.0277, "step": 26350 }, { "grad_norm": 0.2272522896528244, "learning_rate": 9.960856220728333e-05, "loss": 0.0344, "step": 26360 }, { "grad_norm": 0.21580086648464203, "learning_rate": 9.960787359486531e-05, "loss": 0.0246, "step": 26370 }, { "grad_norm": 0.2506382465362549, "learning_rate": 9.960718437966479e-05, "loss": 0.0321, "step": 26380 }, { "grad_norm": 0.20567335188388824, "learning_rate": 9.960649456169014e-05, "loss": 0.0371, "step": 26390 }, { "grad_norm": 0.24821065366268158, "learning_rate": 9.960580414094973e-05, "loss": 0.0256, "step": 26400 }, { "grad_norm": 0.6221510171890259, "learning_rate": 9.960511311745198e-05, "loss": 0.0313, "step": 26410 }, { "grad_norm": 0.21784916520118713, "learning_rate": 9.960442149120526e-05, "loss": 0.0319, "step": 26420 }, { "grad_norm": 0.23088382184505463, "learning_rate": 9.9603729262218e-05, "loss": 0.0335, "step": 26430 }, { "grad_norm": 0.23531323671340942, "learning_rate": 9.960303643049858e-05, "loss": 0.0322, "step": 26440 }, { "grad_norm": 0.1853947639465332, "learning_rate": 9.960234299605544e-05, "loss": 0.043, "step": 26450 }, { "grad_norm": 0.1986093670129776, "learning_rate": 9.9601648958897e-05, "loss": 0.0392, "step": 26460 }, { "grad_norm": 0.16659428179264069, "learning_rate": 9.96009543190317e-05, "loss": 0.0289, "step": 26470 }, { "grad_norm": 0.16391342878341675, "learning_rate": 9.960025907646797e-05, "loss": 0.0254, "step": 26480 }, { "grad_norm": 0.18209896981716156, "learning_rate": 9.959956323121425e-05, "loss": 0.0259, "step": 26490 }, { "grad_norm": 0.2014492303133011, "learning_rate": 9.959886678327903e-05, "loss": 0.0305, "step": 26500 }, { "grad_norm": 0.27334463596343994, "learning_rate": 9.959816973267073e-05, "loss": 0.0358, "step": 26510 }, { "grad_norm": 0.2528266906738281, "learning_rate": 9.959747207939785e-05, "loss": 0.0351, "step": 26520 }, { "grad_norm": 0.5195932388305664, "learning_rate": 9.959677382346884e-05, "loss": 0.0279, "step": 26530 }, { "grad_norm": 0.2590564489364624, "learning_rate": 9.959607496489223e-05, "loss": 0.0318, "step": 26540 }, { "grad_norm": 0.24072088301181793, "learning_rate": 9.959537550367646e-05, "loss": 0.0331, "step": 26550 }, { "grad_norm": 0.18889594078063965, "learning_rate": 9.959467543983005e-05, "loss": 0.0255, "step": 26560 }, { "grad_norm": 0.2207963615655899, "learning_rate": 9.959397477336152e-05, "loss": 0.0275, "step": 26570 }, { "grad_norm": 0.24450185894966125, "learning_rate": 9.959327350427937e-05, "loss": 0.03, "step": 26580 }, { "grad_norm": 0.26642197370529175, "learning_rate": 9.959257163259212e-05, "loss": 0.0261, "step": 26590 }, { "grad_norm": 0.20493635535240173, "learning_rate": 9.95918691583083e-05, "loss": 0.0391, "step": 26600 }, { "grad_norm": 0.13584737479686737, "learning_rate": 9.959116608143646e-05, "loss": 0.0244, "step": 26610 }, { "grad_norm": 0.147618368268013, "learning_rate": 9.959046240198511e-05, "loss": 0.0344, "step": 26620 }, { "grad_norm": 0.17988549172878265, "learning_rate": 9.958975811996284e-05, "loss": 0.024, "step": 26630 }, { "grad_norm": 0.18017718195915222, "learning_rate": 9.958905323537817e-05, "loss": 0.0276, "step": 26640 }, { "grad_norm": 0.1810896396636963, "learning_rate": 9.958834774823969e-05, "loss": 0.0262, "step": 26650 }, { "grad_norm": 0.24947801232337952, "learning_rate": 9.958764165855598e-05, "loss": 0.0306, "step": 26660 }, { "grad_norm": 0.21145223081111908, "learning_rate": 9.95869349663356e-05, "loss": 0.0245, "step": 26670 }, { "grad_norm": 0.16651280224323273, "learning_rate": 9.958622767158713e-05, "loss": 0.0318, "step": 26680 }, { "grad_norm": 0.18536575138568878, "learning_rate": 9.958551977431918e-05, "loss": 0.0282, "step": 26690 }, { "grad_norm": 0.2162194401025772, "learning_rate": 9.958481127454036e-05, "loss": 0.0341, "step": 26700 }, { "grad_norm": 0.19743773341178894, "learning_rate": 9.958410217225925e-05, "loss": 0.0278, "step": 26710 }, { "grad_norm": 0.21101833879947662, "learning_rate": 9.95833924674845e-05, "loss": 0.0299, "step": 26720 }, { "grad_norm": 0.21864701807498932, "learning_rate": 9.958268216022471e-05, "loss": 0.0268, "step": 26730 }, { "grad_norm": 0.13908110558986664, "learning_rate": 9.958197125048852e-05, "loss": 0.0291, "step": 26740 }, { "grad_norm": 0.2613542973995209, "learning_rate": 9.958125973828454e-05, "loss": 0.035, "step": 26750 }, { "grad_norm": 0.2573646008968353, "learning_rate": 9.958054762362148e-05, "loss": 0.0339, "step": 26760 }, { "grad_norm": 0.23095689713954926, "learning_rate": 9.957983490650793e-05, "loss": 0.0272, "step": 26770 }, { "grad_norm": 0.285609632730484, "learning_rate": 9.95791215869526e-05, "loss": 0.0308, "step": 26780 }, { "grad_norm": 0.18258929252624512, "learning_rate": 9.957840766496411e-05, "loss": 0.0267, "step": 26790 }, { "grad_norm": 0.21360048651695251, "learning_rate": 9.957769314055117e-05, "loss": 0.0306, "step": 26800 }, { "grad_norm": 0.17877358198165894, "learning_rate": 9.957697801372243e-05, "loss": 0.0269, "step": 26810 }, { "grad_norm": 0.33527034521102905, "learning_rate": 9.957626228448662e-05, "loss": 0.0297, "step": 26820 }, { "grad_norm": 0.4927169978618622, "learning_rate": 9.95755459528524e-05, "loss": 0.0268, "step": 26830 }, { "grad_norm": 0.21187999844551086, "learning_rate": 9.95748290188285e-05, "loss": 0.0339, "step": 26840 }, { "grad_norm": 0.2181297242641449, "learning_rate": 9.957411148242361e-05, "loss": 0.0267, "step": 26850 }, { "grad_norm": 0.24861395359039307, "learning_rate": 9.957339334364648e-05, "loss": 0.0266, "step": 26860 }, { "grad_norm": 0.2635492980480194, "learning_rate": 9.957267460250579e-05, "loss": 0.0281, "step": 26870 }, { "grad_norm": 0.15984797477722168, "learning_rate": 9.957195525901032e-05, "loss": 0.0227, "step": 26880 }, { "grad_norm": 0.323715478181839, "learning_rate": 9.957123531316878e-05, "loss": 0.0249, "step": 26890 }, { "grad_norm": 0.18343880772590637, "learning_rate": 9.957051476498994e-05, "loss": 0.0253, "step": 26900 }, { "grad_norm": 0.36623647809028625, "learning_rate": 9.956979361448253e-05, "loss": 0.0287, "step": 26910 }, { "grad_norm": 0.3018539845943451, "learning_rate": 9.956907186165532e-05, "loss": 0.0329, "step": 26920 }, { "grad_norm": 0.2267402559518814, "learning_rate": 9.95683495065171e-05, "loss": 0.0252, "step": 26930 }, { "grad_norm": 0.2829987406730652, "learning_rate": 9.956762654907664e-05, "loss": 0.0334, "step": 26940 }, { "grad_norm": 0.25262778997421265, "learning_rate": 9.95669029893427e-05, "loss": 0.0335, "step": 26950 }, { "grad_norm": 0.41127148270606995, "learning_rate": 9.956617882732409e-05, "loss": 0.0391, "step": 26960 }, { "grad_norm": 0.1672130525112152, "learning_rate": 9.956545406302962e-05, "loss": 0.0309, "step": 26970 }, { "grad_norm": 0.24096573889255524, "learning_rate": 9.956472869646807e-05, "loss": 0.0296, "step": 26980 }, { "grad_norm": 0.28253042697906494, "learning_rate": 9.956400272764829e-05, "loss": 0.0328, "step": 26990 }, { "grad_norm": 0.19380706548690796, "learning_rate": 9.956327615657907e-05, "loss": 0.0309, "step": 27000 }, { "grad_norm": 0.26185980439186096, "learning_rate": 9.956254898326924e-05, "loss": 0.0325, "step": 27010 }, { "grad_norm": 0.1796237975358963, "learning_rate": 9.956182120772765e-05, "loss": 0.0326, "step": 27020 }, { "grad_norm": 0.43578284978866577, "learning_rate": 9.956109282996315e-05, "loss": 0.0369, "step": 27030 }, { "grad_norm": 0.3517570197582245, "learning_rate": 9.956036384998457e-05, "loss": 0.034, "step": 27040 }, { "grad_norm": 0.2025146335363388, "learning_rate": 9.95596342678008e-05, "loss": 0.0277, "step": 27050 }, { "grad_norm": 0.1799519807100296, "learning_rate": 9.955890408342065e-05, "loss": 0.0325, "step": 27060 }, { "grad_norm": 0.195998877286911, "learning_rate": 9.955817329685303e-05, "loss": 0.0301, "step": 27070 }, { "grad_norm": 0.22590312361717224, "learning_rate": 9.955744190810682e-05, "loss": 0.0312, "step": 27080 }, { "grad_norm": 0.28963878750801086, "learning_rate": 9.95567099171909e-05, "loss": 0.0277, "step": 27090 }, { "grad_norm": 0.18265588581562042, "learning_rate": 9.955597732411417e-05, "loss": 0.0268, "step": 27100 }, { "grad_norm": 0.25555121898651123, "learning_rate": 9.95552441288855e-05, "loss": 0.0411, "step": 27110 }, { "grad_norm": 0.2694869339466095, "learning_rate": 9.955451033151387e-05, "loss": 0.0405, "step": 27120 }, { "grad_norm": 0.2158583551645279, "learning_rate": 9.955377593200813e-05, "loss": 0.0302, "step": 27130 }, { "grad_norm": 0.23024074733257294, "learning_rate": 9.955304093037721e-05, "loss": 0.0321, "step": 27140 }, { "grad_norm": 0.16448748111724854, "learning_rate": 9.955230532663007e-05, "loss": 0.0306, "step": 27150 }, { "grad_norm": 0.40460044145584106, "learning_rate": 9.955156912077565e-05, "loss": 0.0314, "step": 27160 }, { "grad_norm": 0.1906406730413437, "learning_rate": 9.955083231282286e-05, "loss": 0.032, "step": 27170 }, { "grad_norm": 0.23301368951797485, "learning_rate": 9.955009490278068e-05, "loss": 0.0391, "step": 27180 }, { "grad_norm": 0.46520480513572693, "learning_rate": 9.954935689065808e-05, "loss": 0.0307, "step": 27190 }, { "grad_norm": 0.20217734575271606, "learning_rate": 9.954861827646398e-05, "loss": 0.0341, "step": 27200 }, { "grad_norm": 0.5476764440536499, "learning_rate": 9.95478790602074e-05, "loss": 0.0291, "step": 27210 }, { "grad_norm": 0.2652343213558197, "learning_rate": 9.954713924189731e-05, "loss": 0.0327, "step": 27220 }, { "grad_norm": 0.23926743865013123, "learning_rate": 9.954639882154271e-05, "loss": 0.0305, "step": 27230 }, { "grad_norm": 0.14240550994873047, "learning_rate": 9.954565779915257e-05, "loss": 0.0275, "step": 27240 }, { "grad_norm": 0.26561060547828674, "learning_rate": 9.954491617473591e-05, "loss": 0.0393, "step": 27250 }, { "grad_norm": 0.17905494570732117, "learning_rate": 9.954417394830176e-05, "loss": 0.0283, "step": 27260 }, { "grad_norm": 0.2268780618906021, "learning_rate": 9.954343111985908e-05, "loss": 0.0315, "step": 27270 }, { "grad_norm": 0.3726016879081726, "learning_rate": 9.954268768941697e-05, "loss": 0.0322, "step": 27280 }, { "grad_norm": 0.17936688661575317, "learning_rate": 9.954194365698441e-05, "loss": 0.0264, "step": 27290 }, { "grad_norm": 0.20392683148384094, "learning_rate": 9.954119902257047e-05, "loss": 0.0282, "step": 27300 }, { "grad_norm": 0.24225226044654846, "learning_rate": 9.954045378618418e-05, "loss": 0.0315, "step": 27310 }, { "grad_norm": 0.5565541386604309, "learning_rate": 9.95397079478346e-05, "loss": 0.0321, "step": 27320 }, { "grad_norm": 0.19614939391613007, "learning_rate": 9.953896150753082e-05, "loss": 0.0355, "step": 27330 }, { "grad_norm": 0.22319503128528595, "learning_rate": 9.953821446528186e-05, "loss": 0.0268, "step": 27340 }, { "grad_norm": 0.5377812385559082, "learning_rate": 9.953746682109683e-05, "loss": 0.03, "step": 27350 }, { "grad_norm": 0.21779987215995789, "learning_rate": 9.95367185749848e-05, "loss": 0.0289, "step": 27360 }, { "grad_norm": 0.11983184516429901, "learning_rate": 9.953596972695487e-05, "loss": 0.0283, "step": 27370 }, { "grad_norm": 0.1979103535413742, "learning_rate": 9.953522027701614e-05, "loss": 0.0245, "step": 27380 }, { "grad_norm": 0.21991941332817078, "learning_rate": 9.953447022517772e-05, "loss": 0.027, "step": 27390 }, { "grad_norm": 0.2377922087907791, "learning_rate": 9.953371957144872e-05, "loss": 0.0307, "step": 27400 }, { "grad_norm": 0.3603925108909607, "learning_rate": 9.953296831583825e-05, "loss": 0.0266, "step": 27410 }, { "grad_norm": 0.20709404349327087, "learning_rate": 9.953221645835545e-05, "loss": 0.0257, "step": 27420 }, { "grad_norm": 0.2947681248188019, "learning_rate": 9.953146399900945e-05, "loss": 0.0303, "step": 27430 }, { "grad_norm": 0.2895714342594147, "learning_rate": 9.953071093780941e-05, "loss": 0.0323, "step": 27440 }, { "grad_norm": 0.2834731638431549, "learning_rate": 9.952995727476446e-05, "loss": 0.0318, "step": 27450 }, { "grad_norm": 0.22405663132667542, "learning_rate": 9.952920300988378e-05, "loss": 0.0242, "step": 27460 }, { "grad_norm": 0.2173471748828888, "learning_rate": 9.95284481431765e-05, "loss": 0.0383, "step": 27470 }, { "grad_norm": 0.178633913397789, "learning_rate": 9.952769267465182e-05, "loss": 0.0268, "step": 27480 }, { "grad_norm": 0.20013554394245148, "learning_rate": 9.95269366043189e-05, "loss": 0.0275, "step": 27490 }, { "grad_norm": 0.28793227672576904, "learning_rate": 9.952617993218696e-05, "loss": 0.033, "step": 27500 }, { "grad_norm": 0.3108440041542053, "learning_rate": 9.952542265826516e-05, "loss": 0.0304, "step": 27510 }, { "grad_norm": 0.40839967131614685, "learning_rate": 9.952466478256273e-05, "loss": 0.0372, "step": 27520 }, { "grad_norm": 0.2409159392118454, "learning_rate": 9.952390630508884e-05, "loss": 0.0341, "step": 27530 }, { "grad_norm": 0.2772902548313141, "learning_rate": 9.952314722585274e-05, "loss": 0.0395, "step": 27540 }, { "grad_norm": 0.276726633310318, "learning_rate": 9.952238754486365e-05, "loss": 0.042, "step": 27550 }, { "grad_norm": 0.1852642148733139, "learning_rate": 9.952162726213078e-05, "loss": 0.0289, "step": 27560 }, { "grad_norm": 0.21309801936149597, "learning_rate": 9.95208663776634e-05, "loss": 0.0244, "step": 27570 }, { "grad_norm": 0.15263888239860535, "learning_rate": 9.952010489147074e-05, "loss": 0.0273, "step": 27580 }, { "grad_norm": 0.18463419377803802, "learning_rate": 9.951934280356203e-05, "loss": 0.0222, "step": 27590 }, { "grad_norm": 0.22397178411483765, "learning_rate": 9.951858011394656e-05, "loss": 0.0276, "step": 27600 }, { "grad_norm": 0.3187392055988312, "learning_rate": 9.95178168226336e-05, "loss": 0.033, "step": 27610 }, { "grad_norm": 0.24302375316619873, "learning_rate": 9.95170529296324e-05, "loss": 0.0267, "step": 27620 }, { "grad_norm": 0.18139907717704773, "learning_rate": 9.951628843495226e-05, "loss": 0.0319, "step": 27630 }, { "grad_norm": 0.15555860102176666, "learning_rate": 9.951552333860246e-05, "loss": 0.0268, "step": 27640 }, { "grad_norm": 0.286903440952301, "learning_rate": 9.951475764059231e-05, "loss": 0.0298, "step": 27650 }, { "grad_norm": 0.21529823541641235, "learning_rate": 9.95139913409311e-05, "loss": 0.0324, "step": 27660 }, { "grad_norm": 0.2915932834148407, "learning_rate": 9.951322443962815e-05, "loss": 0.0411, "step": 27670 }, { "grad_norm": 0.30586615204811096, "learning_rate": 9.951245693669277e-05, "loss": 0.0377, "step": 27680 }, { "grad_norm": 0.19892573356628418, "learning_rate": 9.951168883213428e-05, "loss": 0.0332, "step": 27690 }, { "grad_norm": 0.40738940238952637, "learning_rate": 9.951092012596204e-05, "loss": 0.0305, "step": 27700 }, { "grad_norm": 0.21770243346691132, "learning_rate": 9.951015081818537e-05, "loss": 0.0297, "step": 27710 }, { "grad_norm": 0.22690561413764954, "learning_rate": 9.950938090881362e-05, "loss": 0.0279, "step": 27720 }, { "grad_norm": 0.19701755046844482, "learning_rate": 9.950861039785614e-05, "loss": 0.0311, "step": 27730 }, { "grad_norm": 0.1908600777387619, "learning_rate": 9.950783928532232e-05, "loss": 0.0305, "step": 27740 }, { "grad_norm": 0.1651744395494461, "learning_rate": 9.950706757122151e-05, "loss": 0.0246, "step": 27750 }, { "grad_norm": 0.18161241710186005, "learning_rate": 9.950629525556306e-05, "loss": 0.0232, "step": 27760 }, { "grad_norm": 0.17592060565948486, "learning_rate": 9.95055223383564e-05, "loss": 0.038, "step": 27770 }, { "grad_norm": 0.27812856435775757, "learning_rate": 9.95047488196109e-05, "loss": 0.0271, "step": 27780 }, { "grad_norm": 0.2806154191493988, "learning_rate": 9.950397469933597e-05, "loss": 0.0281, "step": 27790 }, { "grad_norm": 0.2632733881473541, "learning_rate": 9.950319997754098e-05, "loss": 0.033, "step": 27800 }, { "grad_norm": 0.37412792444229126, "learning_rate": 9.950242465423541e-05, "loss": 0.0394, "step": 27810 }, { "grad_norm": 0.2316305935382843, "learning_rate": 9.950164872942861e-05, "loss": 0.0353, "step": 27820 }, { "grad_norm": 0.20893365144729614, "learning_rate": 9.950087220313005e-05, "loss": 0.03, "step": 27830 }, { "grad_norm": 0.2162056714296341, "learning_rate": 9.950009507534917e-05, "loss": 0.0314, "step": 27840 }, { "grad_norm": 0.23634852468967438, "learning_rate": 9.949931734609538e-05, "loss": 0.0308, "step": 27850 }, { "grad_norm": 0.2961752116680145, "learning_rate": 9.949853901537815e-05, "loss": 0.0302, "step": 27860 }, { "grad_norm": 0.11877293139696121, "learning_rate": 9.949776008320695e-05, "loss": 0.0222, "step": 27870 }, { "grad_norm": 0.253377228975296, "learning_rate": 9.949698054959123e-05, "loss": 0.042, "step": 27880 }, { "grad_norm": 0.21300670504570007, "learning_rate": 9.949620041454046e-05, "loss": 0.0301, "step": 27890 }, { "grad_norm": 0.2817227840423584, "learning_rate": 9.94954196780641e-05, "loss": 0.0287, "step": 27900 }, { "grad_norm": 0.26336297392845154, "learning_rate": 9.94946383401717e-05, "loss": 0.0269, "step": 27910 }, { "grad_norm": 0.23832592368125916, "learning_rate": 9.94938564008727e-05, "loss": 0.0307, "step": 27920 }, { "grad_norm": 0.358860045671463, "learning_rate": 9.94930738601766e-05, "loss": 0.0366, "step": 27930 }, { "grad_norm": 0.3184051811695099, "learning_rate": 9.949229071809293e-05, "loss": 0.0352, "step": 27940 }, { "grad_norm": 0.2483385056257248, "learning_rate": 9.949150697463121e-05, "loss": 0.0275, "step": 27950 }, { "grad_norm": 0.15272732079029083, "learning_rate": 9.949072262980094e-05, "loss": 0.0437, "step": 27960 }, { "grad_norm": 0.20162487030029297, "learning_rate": 9.948993768361166e-05, "loss": 0.028, "step": 27970 }, { "grad_norm": 0.17551667988300323, "learning_rate": 9.948915213607293e-05, "loss": 0.0298, "step": 27980 }, { "grad_norm": 0.32485347986221313, "learning_rate": 9.948836598719425e-05, "loss": 0.0364, "step": 27990 }, { "grad_norm": 0.2084176242351532, "learning_rate": 9.948757923698522e-05, "loss": 0.0373, "step": 28000 }, { "grad_norm": 0.15128812193870544, "learning_rate": 9.948679188545538e-05, "loss": 0.0228, "step": 28010 }, { "grad_norm": 0.3257102966308594, "learning_rate": 9.948600393261427e-05, "loss": 0.0269, "step": 28020 }, { "grad_norm": 0.21823650598526, "learning_rate": 9.948521537847151e-05, "loss": 0.0306, "step": 28030 }, { "grad_norm": 0.23429593443870544, "learning_rate": 9.948442622303667e-05, "loss": 0.0238, "step": 28040 }, { "grad_norm": 0.17050515115261078, "learning_rate": 9.948363646631931e-05, "loss": 0.0262, "step": 28050 }, { "grad_norm": 0.14741702377796173, "learning_rate": 9.948284610832905e-05, "loss": 0.0233, "step": 28060 }, { "grad_norm": 0.28532183170318604, "learning_rate": 9.94820551490755e-05, "loss": 0.0254, "step": 28070 }, { "grad_norm": 0.21833525598049164, "learning_rate": 9.948126358856825e-05, "loss": 0.0293, "step": 28080 }, { "grad_norm": 0.24238398671150208, "learning_rate": 9.948047142681694e-05, "loss": 0.0264, "step": 28090 }, { "grad_norm": 0.2898583710193634, "learning_rate": 9.947967866383117e-05, "loss": 0.0287, "step": 28100 }, { "grad_norm": 0.16829004883766174, "learning_rate": 9.947888529962061e-05, "loss": 0.0269, "step": 28110 }, { "grad_norm": 0.2711838483810425, "learning_rate": 9.947809133419486e-05, "loss": 0.0223, "step": 28120 }, { "grad_norm": 0.19011461734771729, "learning_rate": 9.947729676756359e-05, "loss": 0.029, "step": 28130 }, { "grad_norm": 0.47832930088043213, "learning_rate": 9.947650159973645e-05, "loss": 0.0279, "step": 28140 }, { "grad_norm": 0.35189080238342285, "learning_rate": 9.94757058307231e-05, "loss": 0.035, "step": 28150 }, { "grad_norm": 0.2421049326658249, "learning_rate": 9.947490946053321e-05, "loss": 0.0248, "step": 28160 }, { "grad_norm": 0.315408855676651, "learning_rate": 9.947411248917647e-05, "loss": 0.0312, "step": 28170 }, { "grad_norm": 0.2568835914134979, "learning_rate": 9.947331491666255e-05, "loss": 0.0331, "step": 28180 }, { "grad_norm": 0.21676066517829895, "learning_rate": 9.947251674300113e-05, "loss": 0.047, "step": 28190 }, { "grad_norm": 0.22895117104053497, "learning_rate": 9.947171796820192e-05, "loss": 0.0319, "step": 28200 }, { "grad_norm": 0.14170481264591217, "learning_rate": 9.947091859227464e-05, "loss": 0.0224, "step": 28210 }, { "grad_norm": 0.2548063397407532, "learning_rate": 9.947011861522899e-05, "loss": 0.0319, "step": 28220 }, { "grad_norm": 0.3514321744441986, "learning_rate": 9.946931803707468e-05, "loss": 0.0247, "step": 28230 }, { "grad_norm": 0.2939807176589966, "learning_rate": 9.946851685782147e-05, "loss": 0.0246, "step": 28240 }, { "grad_norm": 0.24071402847766876, "learning_rate": 9.946771507747906e-05, "loss": 0.0304, "step": 28250 }, { "grad_norm": 0.24431291222572327, "learning_rate": 9.946691269605721e-05, "loss": 0.0255, "step": 28260 }, { "grad_norm": 0.26653891801834106, "learning_rate": 9.946610971356567e-05, "loss": 0.0292, "step": 28270 }, { "grad_norm": 0.20588718354701996, "learning_rate": 9.946530613001418e-05, "loss": 0.0291, "step": 28280 }, { "grad_norm": 0.12265927344560623, "learning_rate": 9.946450194541253e-05, "loss": 0.0295, "step": 28290 }, { "grad_norm": 0.18395428359508514, "learning_rate": 9.946369715977047e-05, "loss": 0.027, "step": 28300 }, { "grad_norm": 0.1744460165500641, "learning_rate": 9.94628917730978e-05, "loss": 0.0245, "step": 28310 }, { "grad_norm": 0.14954258501529694, "learning_rate": 9.946208578540428e-05, "loss": 0.0301, "step": 28320 }, { "grad_norm": 0.17918957769870758, "learning_rate": 9.946127919669972e-05, "loss": 0.0296, "step": 28330 }, { "grad_norm": 0.23333993554115295, "learning_rate": 9.946047200699393e-05, "loss": 0.034, "step": 28340 }, { "grad_norm": 0.18335624039173126, "learning_rate": 9.94596642162967e-05, "loss": 0.029, "step": 28350 }, { "grad_norm": 0.29374992847442627, "learning_rate": 9.945885582461785e-05, "loss": 0.0246, "step": 28360 }, { "grad_norm": 0.16732390224933624, "learning_rate": 9.94580468319672e-05, "loss": 0.0275, "step": 28370 }, { "grad_norm": 0.16137827932834625, "learning_rate": 9.94572372383546e-05, "loss": 0.0227, "step": 28380 }, { "grad_norm": 0.1426190584897995, "learning_rate": 9.945642704378986e-05, "loss": 0.0296, "step": 28390 }, { "grad_norm": 0.13246017694473267, "learning_rate": 9.945561624828285e-05, "loss": 0.0272, "step": 28400 }, { "grad_norm": 0.26972997188568115, "learning_rate": 9.94548048518434e-05, "loss": 0.0218, "step": 28410 }, { "grad_norm": 0.22139108180999756, "learning_rate": 9.945399285448137e-05, "loss": 0.0236, "step": 28420 }, { "grad_norm": 0.22000153362751007, "learning_rate": 9.945318025620665e-05, "loss": 0.0335, "step": 28430 }, { "grad_norm": 0.17362123727798462, "learning_rate": 9.945236705702908e-05, "loss": 0.0191, "step": 28440 }, { "grad_norm": 0.18852199614048004, "learning_rate": 9.945155325695858e-05, "loss": 0.0221, "step": 28450 }, { "grad_norm": 0.29119476675987244, "learning_rate": 9.9450738856005e-05, "loss": 0.0252, "step": 28460 }, { "grad_norm": 0.18726304173469543, "learning_rate": 9.944992385417826e-05, "loss": 0.0253, "step": 28470 }, { "grad_norm": 0.26575759053230286, "learning_rate": 9.944910825148826e-05, "loss": 0.0344, "step": 28480 }, { "grad_norm": 0.21194244921207428, "learning_rate": 9.94482920479449e-05, "loss": 0.0274, "step": 28490 }, { "grad_norm": 0.13303551077842712, "learning_rate": 9.944747524355811e-05, "loss": 0.0263, "step": 28500 }, { "grad_norm": 0.25504398345947266, "learning_rate": 9.944665783833782e-05, "loss": 0.0255, "step": 28510 }, { "grad_norm": 0.21084731817245483, "learning_rate": 9.944583983229394e-05, "loss": 0.032, "step": 28520 }, { "grad_norm": 0.2291295975446701, "learning_rate": 9.944502122543641e-05, "loss": 0.0252, "step": 28530 }, { "grad_norm": 0.1766030490398407, "learning_rate": 9.94442020177752e-05, "loss": 0.0235, "step": 28540 }, { "grad_norm": 0.18075278401374817, "learning_rate": 9.944338220932025e-05, "loss": 0.0287, "step": 28550 }, { "grad_norm": 0.15325716137886047, "learning_rate": 9.944256180008154e-05, "loss": 0.031, "step": 28560 }, { "grad_norm": 0.19154870510101318, "learning_rate": 9.9441740790069e-05, "loss": 0.0367, "step": 28570 }, { "grad_norm": 0.13326583802700043, "learning_rate": 9.944091917929264e-05, "loss": 0.0275, "step": 28580 }, { "grad_norm": 0.27342185378074646, "learning_rate": 9.944009696776244e-05, "loss": 0.0311, "step": 28590 }, { "grad_norm": 0.2766611874103546, "learning_rate": 9.943927415548837e-05, "loss": 0.0301, "step": 28600 }, { "grad_norm": 0.7755812406539917, "learning_rate": 9.943845074248044e-05, "loss": 0.0404, "step": 28610 }, { "grad_norm": 0.27738308906555176, "learning_rate": 9.943762672874865e-05, "loss": 0.0395, "step": 28620 }, { "grad_norm": 0.16061116755008698, "learning_rate": 9.943680211430304e-05, "loss": 0.0262, "step": 28630 }, { "grad_norm": 0.1969873458147049, "learning_rate": 9.943597689915361e-05, "loss": 0.0367, "step": 28640 }, { "grad_norm": 0.1660878211259842, "learning_rate": 9.943515108331036e-05, "loss": 0.0234, "step": 28650 }, { "grad_norm": 0.2872418761253357, "learning_rate": 9.943432466678336e-05, "loss": 0.0283, "step": 28660 }, { "grad_norm": 0.2511909008026123, "learning_rate": 9.943349764958266e-05, "loss": 0.0293, "step": 28670 }, { "grad_norm": 0.6418221592903137, "learning_rate": 9.943267003171825e-05, "loss": 0.0314, "step": 28680 }, { "grad_norm": 0.2370155155658722, "learning_rate": 9.943184181320026e-05, "loss": 0.0285, "step": 28690 }, { "grad_norm": 0.1759663224220276, "learning_rate": 9.943101299403871e-05, "loss": 0.0327, "step": 28700 }, { "grad_norm": 0.15555383265018463, "learning_rate": 9.943018357424369e-05, "loss": 0.0234, "step": 28710 }, { "grad_norm": 0.12453809380531311, "learning_rate": 9.942935355382525e-05, "loss": 0.0311, "step": 28720 }, { "grad_norm": 0.21559937298297882, "learning_rate": 9.942852293279351e-05, "loss": 0.0414, "step": 28730 }, { "grad_norm": 0.2160547822713852, "learning_rate": 9.942769171115856e-05, "loss": 0.0295, "step": 28740 }, { "grad_norm": 0.21611474454402924, "learning_rate": 9.942685988893045e-05, "loss": 0.031, "step": 28750 }, { "grad_norm": 0.18645937740802765, "learning_rate": 9.942602746611936e-05, "loss": 0.0355, "step": 28760 }, { "grad_norm": 0.15275821089744568, "learning_rate": 9.942519444273537e-05, "loss": 0.0249, "step": 28770 }, { "grad_norm": 0.22738206386566162, "learning_rate": 9.942436081878859e-05, "loss": 0.027, "step": 28780 }, { "grad_norm": 0.1454688459634781, "learning_rate": 9.942352659428916e-05, "loss": 0.0278, "step": 28790 }, { "grad_norm": 0.21877849102020264, "learning_rate": 9.942269176924722e-05, "loss": 0.0234, "step": 28800 }, { "grad_norm": 0.16726182401180267, "learning_rate": 9.942185634367292e-05, "loss": 0.0193, "step": 28810 }, { "grad_norm": 0.1789451390504837, "learning_rate": 9.942102031757639e-05, "loss": 0.0244, "step": 28820 }, { "grad_norm": 0.19682304561138153, "learning_rate": 9.942018369096781e-05, "loss": 0.0217, "step": 28830 }, { "grad_norm": 0.20118416845798492, "learning_rate": 9.941934646385734e-05, "loss": 0.0249, "step": 28840 }, { "grad_norm": 0.3219028115272522, "learning_rate": 9.941850863625515e-05, "loss": 0.0291, "step": 28850 }, { "grad_norm": 0.17830432951450348, "learning_rate": 9.941767020817142e-05, "loss": 0.0314, "step": 28860 }, { "grad_norm": 0.20276713371276855, "learning_rate": 9.941683117961634e-05, "loss": 0.0283, "step": 28870 }, { "grad_norm": 0.24149788916110992, "learning_rate": 9.941599155060011e-05, "loss": 0.0372, "step": 28880 }, { "grad_norm": 0.21789152920246124, "learning_rate": 9.941515132113292e-05, "loss": 0.0298, "step": 28890 }, { "grad_norm": 0.15443724393844604, "learning_rate": 9.941431049122498e-05, "loss": 0.0271, "step": 28900 }, { "grad_norm": 0.2398681491613388, "learning_rate": 9.94134690608865e-05, "loss": 0.032, "step": 28910 }, { "grad_norm": 0.38438716530799866, "learning_rate": 9.941262703012775e-05, "loss": 0.0265, "step": 28920 }, { "grad_norm": 0.17244504392147064, "learning_rate": 9.941178439895891e-05, "loss": 0.0267, "step": 28930 }, { "grad_norm": 0.27370160818099976, "learning_rate": 9.941094116739025e-05, "loss": 0.0272, "step": 28940 }, { "grad_norm": 0.23775865137577057, "learning_rate": 9.941009733543199e-05, "loss": 0.0268, "step": 28950 }, { "grad_norm": 0.19851525127887726, "learning_rate": 9.940925290309439e-05, "loss": 0.0278, "step": 28960 }, { "grad_norm": 0.1575947403907776, "learning_rate": 9.940840787038773e-05, "loss": 0.029, "step": 28970 }, { "grad_norm": 0.21953152120113373, "learning_rate": 9.940756223732226e-05, "loss": 0.023, "step": 28980 }, { "grad_norm": 0.1852354109287262, "learning_rate": 9.940671600390828e-05, "loss": 0.0234, "step": 28990 }, { "grad_norm": 0.18095064163208008, "learning_rate": 9.940586917015602e-05, "loss": 0.0261, "step": 29000 }, { "grad_norm": 0.24593739211559296, "learning_rate": 9.940502173607582e-05, "loss": 0.0238, "step": 29010 }, { "grad_norm": 0.23376186192035675, "learning_rate": 9.940417370167796e-05, "loss": 0.0299, "step": 29020 }, { "grad_norm": 0.1591847538948059, "learning_rate": 9.940332506697274e-05, "loss": 0.0333, "step": 29030 }, { "grad_norm": 0.24576444923877716, "learning_rate": 9.940247583197047e-05, "loss": 0.0306, "step": 29040 }, { "grad_norm": 0.2598245143890381, "learning_rate": 9.940162599668149e-05, "loss": 0.0236, "step": 29050 }, { "grad_norm": 0.30154818296432495, "learning_rate": 9.94007755611161e-05, "loss": 0.026, "step": 29060 }, { "grad_norm": 0.19810596108436584, "learning_rate": 9.939992452528466e-05, "loss": 0.0236, "step": 29070 }, { "grad_norm": 0.18323861062526703, "learning_rate": 9.939907288919747e-05, "loss": 0.0237, "step": 29080 }, { "grad_norm": 0.19428612291812897, "learning_rate": 9.939822065286493e-05, "loss": 0.0205, "step": 29090 }, { "grad_norm": 0.3694308400154114, "learning_rate": 9.939736781629736e-05, "loss": 0.0273, "step": 29100 }, { "grad_norm": 0.23642027378082275, "learning_rate": 9.939651437950512e-05, "loss": 0.0305, "step": 29110 }, { "grad_norm": 0.2838138937950134, "learning_rate": 9.939566034249861e-05, "loss": 0.0263, "step": 29120 }, { "grad_norm": 0.1392710655927658, "learning_rate": 9.939480570528819e-05, "loss": 0.0265, "step": 29130 }, { "grad_norm": 0.2048589289188385, "learning_rate": 9.939395046788422e-05, "loss": 0.0296, "step": 29140 }, { "grad_norm": 0.23998072743415833, "learning_rate": 9.939309463029713e-05, "loss": 0.0294, "step": 29150 }, { "grad_norm": 0.21595503389835358, "learning_rate": 9.939223819253731e-05, "loss": 0.0287, "step": 29160 }, { "grad_norm": 0.31010422110557556, "learning_rate": 9.939138115461516e-05, "loss": 0.0301, "step": 29170 }, { "grad_norm": 0.2162114977836609, "learning_rate": 9.939052351654111e-05, "loss": 0.0237, "step": 29180 }, { "grad_norm": 0.23975715041160583, "learning_rate": 9.938966527832555e-05, "loss": 0.0332, "step": 29190 }, { "grad_norm": 0.2388041615486145, "learning_rate": 9.938880643997893e-05, "loss": 0.0309, "step": 29200 }, { "grad_norm": 0.21900565922260284, "learning_rate": 9.938794700151168e-05, "loss": 0.0258, "step": 29210 }, { "grad_norm": 0.14010296761989594, "learning_rate": 9.938708696293424e-05, "loss": 0.0443, "step": 29220 }, { "grad_norm": 0.19471849501132965, "learning_rate": 9.938622632425707e-05, "loss": 0.0239, "step": 29230 }, { "grad_norm": 0.214535653591156, "learning_rate": 9.93853650854906e-05, "loss": 0.042, "step": 29240 }, { "grad_norm": 0.17104877531528473, "learning_rate": 9.938450324664535e-05, "loss": 0.0255, "step": 29250 }, { "grad_norm": 0.1565878838300705, "learning_rate": 9.938364080773175e-05, "loss": 0.028, "step": 29260 }, { "grad_norm": 0.18003498017787933, "learning_rate": 9.938277776876028e-05, "loss": 0.0262, "step": 29270 }, { "grad_norm": 0.20607849955558777, "learning_rate": 9.938191412974143e-05, "loss": 0.025, "step": 29280 }, { "grad_norm": 0.22273226082324982, "learning_rate": 9.93810498906857e-05, "loss": 0.0251, "step": 29290 }, { "grad_norm": 0.1959712654352188, "learning_rate": 9.93801850516036e-05, "loss": 0.0254, "step": 29300 }, { "grad_norm": 0.2449394166469574, "learning_rate": 9.937931961250564e-05, "loss": 0.0335, "step": 29310 }, { "grad_norm": 0.19777460396289825, "learning_rate": 9.93784535734023e-05, "loss": 0.0294, "step": 29320 }, { "grad_norm": 0.2055530995130539, "learning_rate": 9.937758693430414e-05, "loss": 0.0256, "step": 29330 }, { "grad_norm": 0.1832256317138672, "learning_rate": 9.937671969522169e-05, "loss": 0.0236, "step": 29340 }, { "grad_norm": 0.30824440717697144, "learning_rate": 9.937585185616546e-05, "loss": 0.0315, "step": 29350 }, { "grad_norm": 0.21196727454662323, "learning_rate": 9.937498341714602e-05, "loss": 0.0275, "step": 29360 }, { "grad_norm": 0.18570266664028168, "learning_rate": 9.93741143781739e-05, "loss": 0.0356, "step": 29370 }, { "grad_norm": 0.17186519503593445, "learning_rate": 9.937324473925969e-05, "loss": 0.025, "step": 29380 }, { "grad_norm": 0.15595966577529907, "learning_rate": 9.937237450041393e-05, "loss": 0.0262, "step": 29390 }, { "grad_norm": 0.24251340329647064, "learning_rate": 9.93715036616472e-05, "loss": 0.0322, "step": 29400 }, { "grad_norm": 0.27649232745170593, "learning_rate": 9.93706322229701e-05, "loss": 0.0226, "step": 29410 }, { "grad_norm": 0.25780320167541504, "learning_rate": 9.936976018439321e-05, "loss": 0.0232, "step": 29420 }, { "grad_norm": 0.19179660081863403, "learning_rate": 9.936888754592711e-05, "loss": 0.0225, "step": 29430 }, { "grad_norm": 0.1815309375524521, "learning_rate": 9.93680143075824e-05, "loss": 0.0241, "step": 29440 }, { "grad_norm": 0.20923210680484772, "learning_rate": 9.936714046936972e-05, "loss": 0.0251, "step": 29450 }, { "grad_norm": 0.7632431983947754, "learning_rate": 9.936626603129969e-05, "loss": 0.03, "step": 29460 }, { "grad_norm": 0.21325723826885223, "learning_rate": 9.936539099338289e-05, "loss": 0.0245, "step": 29470 }, { "grad_norm": 0.20950767397880554, "learning_rate": 9.936451535562999e-05, "loss": 0.0339, "step": 29480 }, { "grad_norm": 0.2830871641635895, "learning_rate": 9.936363911805162e-05, "loss": 0.0249, "step": 29490 }, { "grad_norm": 0.19251392781734467, "learning_rate": 9.936276228065842e-05, "loss": 0.0304, "step": 29500 }, { "grad_norm": 0.3365020751953125, "learning_rate": 9.936188484346107e-05, "loss": 0.0298, "step": 29510 }, { "grad_norm": 0.15309642255306244, "learning_rate": 9.936100680647021e-05, "loss": 0.028, "step": 29520 }, { "grad_norm": 0.13430610299110413, "learning_rate": 9.93601281696965e-05, "loss": 0.0271, "step": 29530 }, { "grad_norm": 0.2055402249097824, "learning_rate": 9.935924893315065e-05, "loss": 0.048, "step": 29540 }, { "grad_norm": 0.20623700320720673, "learning_rate": 9.93583690968433e-05, "loss": 0.0309, "step": 29550 }, { "grad_norm": 0.22932644188404083, "learning_rate": 9.935748866078518e-05, "loss": 0.0284, "step": 29560 }, { "grad_norm": 0.18290828168392181, "learning_rate": 9.935660762498696e-05, "loss": 0.0251, "step": 29570 }, { "grad_norm": 0.2372904121875763, "learning_rate": 9.935572598945937e-05, "loss": 0.0413, "step": 29580 }, { "grad_norm": 0.1752956062555313, "learning_rate": 9.935484375421309e-05, "loss": 0.0345, "step": 29590 }, { "grad_norm": 0.18994267284870148, "learning_rate": 9.935396091925889e-05, "loss": 0.0314, "step": 29600 }, { "grad_norm": 0.15621592104434967, "learning_rate": 9.935307748460744e-05, "loss": 0.0253, "step": 29610 }, { "grad_norm": 0.29557353258132935, "learning_rate": 9.935219345026951e-05, "loss": 0.0346, "step": 29620 }, { "grad_norm": 0.1842295229434967, "learning_rate": 9.935130881625583e-05, "loss": 0.0252, "step": 29630 }, { "grad_norm": 0.36506059765815735, "learning_rate": 9.935042358257715e-05, "loss": 0.028, "step": 29640 }, { "grad_norm": 0.17911303043365479, "learning_rate": 9.934953774924424e-05, "loss": 0.0317, "step": 29650 }, { "grad_norm": 0.35247835516929626, "learning_rate": 9.934865131626784e-05, "loss": 0.041, "step": 29660 }, { "grad_norm": 0.2258862555027008, "learning_rate": 9.934776428365875e-05, "loss": 0.0328, "step": 29670 }, { "grad_norm": 0.22261451184749603, "learning_rate": 9.934687665142773e-05, "loss": 0.0298, "step": 29680 }, { "grad_norm": 0.24500146508216858, "learning_rate": 9.934598841958555e-05, "loss": 0.029, "step": 29690 }, { "grad_norm": 0.18506672978401184, "learning_rate": 9.934509958814304e-05, "loss": 0.0418, "step": 29700 }, { "grad_norm": 0.253597229719162, "learning_rate": 9.934421015711097e-05, "loss": 0.0347, "step": 29710 }, { "grad_norm": 0.1709872931241989, "learning_rate": 9.934332012650017e-05, "loss": 0.0328, "step": 29720 }, { "grad_norm": 0.16831918060779572, "learning_rate": 9.934242949632143e-05, "loss": 0.0387, "step": 29730 }, { "grad_norm": 0.19047656655311584, "learning_rate": 9.934153826658557e-05, "loss": 0.0313, "step": 29740 }, { "grad_norm": 0.14349645376205444, "learning_rate": 9.934064643730345e-05, "loss": 0.0289, "step": 29750 }, { "grad_norm": 0.18180961906909943, "learning_rate": 9.93397540084859e-05, "loss": 0.0291, "step": 29760 }, { "grad_norm": 0.15320877730846405, "learning_rate": 9.933886098014374e-05, "loss": 0.0244, "step": 29770 }, { "grad_norm": 0.16072553396224976, "learning_rate": 9.933796735228785e-05, "loss": 0.0254, "step": 29780 }, { "grad_norm": 0.14037545025348663, "learning_rate": 9.933707312492906e-05, "loss": 0.0248, "step": 29790 }, { "grad_norm": 0.13606321811676025, "learning_rate": 9.933617829807825e-05, "loss": 0.0294, "step": 29800 }, { "grad_norm": 0.1253119260072708, "learning_rate": 9.933528287174629e-05, "loss": 0.0309, "step": 29810 }, { "grad_norm": 2.127514600753784, "learning_rate": 9.933438684594407e-05, "loss": 0.0267, "step": 29820 }, { "grad_norm": 0.32543179392814636, "learning_rate": 9.933349022068248e-05, "loss": 0.0352, "step": 29830 }, { "grad_norm": 0.22549748420715332, "learning_rate": 9.93325929959724e-05, "loss": 0.045, "step": 29840 }, { "grad_norm": 0.31973138451576233, "learning_rate": 9.933169517182473e-05, "loss": 0.0418, "step": 29850 }, { "grad_norm": 0.21994216740131378, "learning_rate": 9.933079674825038e-05, "loss": 0.0321, "step": 29860 }, { "grad_norm": 0.19176870584487915, "learning_rate": 9.93298977252603e-05, "loss": 0.0228, "step": 29870 }, { "grad_norm": 0.21369081735610962, "learning_rate": 9.932899810286535e-05, "loss": 0.0341, "step": 29880 }, { "grad_norm": 0.23197707533836365, "learning_rate": 9.932809788107653e-05, "loss": 0.0275, "step": 29890 }, { "grad_norm": 0.1975989192724228, "learning_rate": 9.932719705990474e-05, "loss": 0.0352, "step": 29900 }, { "grad_norm": 0.272048681974411, "learning_rate": 9.932629563936094e-05, "loss": 0.024, "step": 29910 }, { "grad_norm": 0.19242405891418457, "learning_rate": 9.932539361945606e-05, "loss": 0.0243, "step": 29920 }, { "grad_norm": 0.3802434504032135, "learning_rate": 9.932449100020107e-05, "loss": 0.0322, "step": 29930 }, { "grad_norm": 0.18661721050739288, "learning_rate": 9.932358778160697e-05, "loss": 0.0313, "step": 29940 }, { "grad_norm": 0.3015882670879364, "learning_rate": 9.932268396368469e-05, "loss": 0.027, "step": 29950 }, { "grad_norm": 0.14251479506492615, "learning_rate": 9.932177954644524e-05, "loss": 0.0255, "step": 29960 }, { "grad_norm": 0.17340341210365295, "learning_rate": 9.93208745298996e-05, "loss": 0.0285, "step": 29970 }, { "grad_norm": 0.34806519746780396, "learning_rate": 9.931996891405877e-05, "loss": 0.0306, "step": 29980 }, { "grad_norm": 0.22451487183570862, "learning_rate": 9.931906269893373e-05, "loss": 0.0299, "step": 29990 }, { "grad_norm": 0.24786509573459625, "learning_rate": 9.931815588453555e-05, "loss": 0.0268, "step": 30000 }, { "grad_norm": 0.20172695815563202, "learning_rate": 9.93172484708752e-05, "loss": 0.0295, "step": 30010 }, { "grad_norm": 0.19510135054588318, "learning_rate": 9.93163404579637e-05, "loss": 0.026, "step": 30020 }, { "grad_norm": 0.22817453742027283, "learning_rate": 9.931543184581211e-05, "loss": 0.0278, "step": 30030 }, { "grad_norm": 0.1883191615343094, "learning_rate": 9.931452263443146e-05, "loss": 0.0299, "step": 30040 }, { "grad_norm": 0.22600597143173218, "learning_rate": 9.931361282383282e-05, "loss": 0.0372, "step": 30050 }, { "grad_norm": 0.4707282781600952, "learning_rate": 9.93127024140272e-05, "loss": 0.0318, "step": 30060 }, { "grad_norm": 0.27279171347618103, "learning_rate": 9.931179140502569e-05, "loss": 0.0383, "step": 30070 }, { "grad_norm": 0.2806113660335541, "learning_rate": 9.931087979683937e-05, "loss": 0.035, "step": 30080 }, { "grad_norm": 0.23157575726509094, "learning_rate": 9.930996758947929e-05, "loss": 0.0287, "step": 30090 }, { "grad_norm": 0.22249823808670044, "learning_rate": 9.930905478295655e-05, "loss": 0.0383, "step": 30100 }, { "grad_norm": 0.25782087445259094, "learning_rate": 9.930814137728224e-05, "loss": 0.0373, "step": 30110 }, { "grad_norm": 1.6680406332015991, "learning_rate": 9.930722737246747e-05, "loss": 0.0339, "step": 30120 }, { "grad_norm": 0.35662657022476196, "learning_rate": 9.930631276852331e-05, "loss": 0.0317, "step": 30130 }, { "grad_norm": 0.17768077552318573, "learning_rate": 9.930539756546091e-05, "loss": 0.0331, "step": 30140 }, { "grad_norm": 0.23719029128551483, "learning_rate": 9.930448176329139e-05, "loss": 0.0306, "step": 30150 }, { "grad_norm": 0.19259998202323914, "learning_rate": 9.930356536202585e-05, "loss": 0.0251, "step": 30160 }, { "grad_norm": 0.21001245081424713, "learning_rate": 9.930264836167544e-05, "loss": 0.0265, "step": 30170 }, { "grad_norm": 0.18044033646583557, "learning_rate": 9.93017307622513e-05, "loss": 0.0232, "step": 30180 }, { "grad_norm": 0.19431884586811066, "learning_rate": 9.930081256376461e-05, "loss": 0.0212, "step": 30190 }, { "grad_norm": 0.20853370428085327, "learning_rate": 9.929989376622646e-05, "loss": 0.0254, "step": 30200 }, { "grad_norm": 0.1645205020904541, "learning_rate": 9.929897436964807e-05, "loss": 0.0254, "step": 30210 }, { "grad_norm": 0.1708509624004364, "learning_rate": 9.92980543740406e-05, "loss": 0.0233, "step": 30220 }, { "grad_norm": 0.2426581084728241, "learning_rate": 9.929713377941523e-05, "loss": 0.0287, "step": 30230 }, { "grad_norm": 0.31916654109954834, "learning_rate": 9.929621258578312e-05, "loss": 0.0255, "step": 30240 }, { "grad_norm": 0.2770562767982483, "learning_rate": 9.929529079315551e-05, "loss": 0.0215, "step": 30250 }, { "grad_norm": 0.19910036027431488, "learning_rate": 9.929436840154355e-05, "loss": 0.0238, "step": 30260 }, { "grad_norm": 0.2221745401620865, "learning_rate": 9.929344541095848e-05, "loss": 0.0237, "step": 30270 }, { "grad_norm": 0.27665308117866516, "learning_rate": 9.92925218214115e-05, "loss": 0.0382, "step": 30280 }, { "grad_norm": 0.17673757672309875, "learning_rate": 9.929159763291387e-05, "loss": 0.0318, "step": 30290 }, { "grad_norm": 0.20617257058620453, "learning_rate": 9.929067284547677e-05, "loss": 0.0257, "step": 30300 }, { "grad_norm": 0.2487398386001587, "learning_rate": 9.928974745911146e-05, "loss": 0.0332, "step": 30310 }, { "grad_norm": 0.19075748324394226, "learning_rate": 9.928882147382917e-05, "loss": 0.0313, "step": 30320 }, { "grad_norm": 0.12727507948875427, "learning_rate": 9.928789488964118e-05, "loss": 0.0291, "step": 30330 }, { "grad_norm": 0.17563767731189728, "learning_rate": 9.928696770655872e-05, "loss": 0.0301, "step": 30340 }, { "grad_norm": 0.12486843019723892, "learning_rate": 9.928603992459306e-05, "loss": 0.0226, "step": 30350 }, { "grad_norm": 0.2084939181804657, "learning_rate": 9.928511154375551e-05, "loss": 0.0278, "step": 30360 }, { "grad_norm": 0.1822628378868103, "learning_rate": 9.928418256405729e-05, "loss": 0.0328, "step": 30370 }, { "grad_norm": 0.34592190384864807, "learning_rate": 9.928325298550973e-05, "loss": 0.0298, "step": 30380 }, { "grad_norm": 0.2683582603931427, "learning_rate": 9.928232280812413e-05, "loss": 0.0331, "step": 30390 }, { "grad_norm": 0.3032667338848114, "learning_rate": 9.928139203191176e-05, "loss": 0.0266, "step": 30400 }, { "grad_norm": 0.12079396843910217, "learning_rate": 9.928046065688396e-05, "loss": 0.0339, "step": 30410 }, { "grad_norm": 0.26226431131362915, "learning_rate": 9.927952868305203e-05, "loss": 0.0317, "step": 30420 }, { "grad_norm": 0.33433666825294495, "learning_rate": 9.927859611042729e-05, "loss": 0.0282, "step": 30430 }, { "grad_norm": 0.19046586751937866, "learning_rate": 9.92776629390211e-05, "loss": 0.0225, "step": 30440 }, { "grad_norm": 0.19764184951782227, "learning_rate": 9.927672916884476e-05, "loss": 0.0259, "step": 30450 }, { "grad_norm": 0.1746198534965515, "learning_rate": 9.927579479990964e-05, "loss": 0.0228, "step": 30460 }, { "grad_norm": 0.36391186714172363, "learning_rate": 9.927485983222709e-05, "loss": 0.0255, "step": 30470 }, { "grad_norm": 0.20544131100177765, "learning_rate": 9.927392426580847e-05, "loss": 0.0213, "step": 30480 }, { "grad_norm": 0.19597633183002472, "learning_rate": 9.927298810066514e-05, "loss": 0.0218, "step": 30490 }, { "grad_norm": 0.25822606682777405, "learning_rate": 9.92720513368085e-05, "loss": 0.0264, "step": 30500 }, { "grad_norm": 0.18711356818675995, "learning_rate": 9.927111397424991e-05, "loss": 0.0221, "step": 30510 }, { "grad_norm": 0.17517882585525513, "learning_rate": 9.927017601300077e-05, "loss": 0.0239, "step": 30520 }, { "grad_norm": 0.18886937201023102, "learning_rate": 9.926923745307246e-05, "loss": 0.0286, "step": 30530 }, { "grad_norm": 0.28808170557022095, "learning_rate": 9.926829829447641e-05, "loss": 0.0318, "step": 30540 }, { "grad_norm": 0.23009063303470612, "learning_rate": 9.926735853722401e-05, "loss": 0.029, "step": 30550 }, { "grad_norm": 0.26272231340408325, "learning_rate": 9.926641818132669e-05, "loss": 0.0236, "step": 30560 }, { "grad_norm": 0.20999625325202942, "learning_rate": 9.926547722679586e-05, "loss": 0.0294, "step": 30570 }, { "grad_norm": 0.22451947629451752, "learning_rate": 9.9264535673643e-05, "loss": 0.0263, "step": 30580 }, { "grad_norm": 0.1712537705898285, "learning_rate": 9.92635935218795e-05, "loss": 0.021, "step": 30590 }, { "grad_norm": 0.19448257982730865, "learning_rate": 9.926265077151682e-05, "loss": 0.0284, "step": 30600 }, { "grad_norm": 0.19761182367801666, "learning_rate": 9.926170742256643e-05, "loss": 0.029, "step": 30610 }, { "grad_norm": 0.22034768760204315, "learning_rate": 9.926076347503979e-05, "loss": 0.0346, "step": 30620 }, { "grad_norm": 0.37707698345184326, "learning_rate": 9.925981892894834e-05, "loss": 0.0255, "step": 30630 }, { "grad_norm": 0.2996807098388672, "learning_rate": 9.925887378430361e-05, "loss": 0.0281, "step": 30640 }, { "grad_norm": 0.2679591476917267, "learning_rate": 9.925792804111703e-05, "loss": 0.0365, "step": 30650 }, { "grad_norm": 0.17392806708812714, "learning_rate": 9.925698169940013e-05, "loss": 0.0289, "step": 30660 }, { "grad_norm": 0.17707449197769165, "learning_rate": 9.925603475916439e-05, "loss": 0.0276, "step": 30670 }, { "grad_norm": 0.23923629522323608, "learning_rate": 9.925508722042134e-05, "loss": 0.0235, "step": 30680 }, { "grad_norm": 0.2395300716161728, "learning_rate": 9.925413908318247e-05, "loss": 0.0217, "step": 30690 }, { "grad_norm": 0.1857181191444397, "learning_rate": 9.925319034745928e-05, "loss": 0.0299, "step": 30700 }, { "grad_norm": 0.12877844274044037, "learning_rate": 9.925224101326335e-05, "loss": 0.0254, "step": 30710 }, { "grad_norm": 0.27490854263305664, "learning_rate": 9.925129108060618e-05, "loss": 0.0228, "step": 30720 }, { "grad_norm": 0.15685102343559265, "learning_rate": 9.925034054949933e-05, "loss": 0.0323, "step": 30730 }, { "grad_norm": 0.2927572429180145, "learning_rate": 9.924938941995434e-05, "loss": 0.0298, "step": 30740 }, { "grad_norm": 0.2208448052406311, "learning_rate": 9.924843769198276e-05, "loss": 0.023, "step": 30750 }, { "grad_norm": 0.18096210062503815, "learning_rate": 9.924748536559618e-05, "loss": 0.0338, "step": 30760 }, { "grad_norm": 0.17887498438358307, "learning_rate": 9.924653244080614e-05, "loss": 0.0264, "step": 30770 }, { "grad_norm": 0.1915551722049713, "learning_rate": 9.924557891762425e-05, "loss": 0.0256, "step": 30780 }, { "grad_norm": 0.2813945710659027, "learning_rate": 9.924462479606207e-05, "loss": 0.0257, "step": 30790 }, { "grad_norm": 0.4488559663295746, "learning_rate": 9.92436700761312e-05, "loss": 0.0313, "step": 30800 }, { "grad_norm": 0.17775286734104156, "learning_rate": 9.924271475784325e-05, "loss": 0.0259, "step": 30810 }, { "grad_norm": 0.18632060289382935, "learning_rate": 9.924175884120983e-05, "loss": 0.027, "step": 30820 }, { "grad_norm": 0.23100808262825012, "learning_rate": 9.924080232624254e-05, "loss": 0.0278, "step": 30830 }, { "grad_norm": 0.25537109375, "learning_rate": 9.923984521295302e-05, "loss": 0.0293, "step": 30840 }, { "grad_norm": 0.21396389603614807, "learning_rate": 9.923888750135289e-05, "loss": 0.0293, "step": 30850 }, { "grad_norm": 0.1722903698682785, "learning_rate": 9.923792919145377e-05, "loss": 0.0228, "step": 30860 }, { "grad_norm": 0.17879271507263184, "learning_rate": 9.923697028326733e-05, "loss": 0.0222, "step": 30870 }, { "grad_norm": 0.18651634454727173, "learning_rate": 9.923601077680523e-05, "loss": 0.0273, "step": 30880 }, { "grad_norm": 0.16314364969730377, "learning_rate": 9.923505067207909e-05, "loss": 0.0296, "step": 30890 }, { "grad_norm": 0.21471872925758362, "learning_rate": 9.92340899691006e-05, "loss": 0.0391, "step": 30900 }, { "grad_norm": 0.20475420355796814, "learning_rate": 9.923312866788145e-05, "loss": 0.0347, "step": 30910 }, { "grad_norm": 0.22230404615402222, "learning_rate": 9.923216676843329e-05, "loss": 0.0298, "step": 30920 }, { "grad_norm": 0.1458771675825119, "learning_rate": 9.923120427076782e-05, "loss": 0.0221, "step": 30930 }, { "grad_norm": 0.18933475017547607, "learning_rate": 9.923024117489674e-05, "loss": 0.0263, "step": 30940 }, { "grad_norm": 0.15033909678459167, "learning_rate": 9.922927748083175e-05, "loss": 0.0326, "step": 30950 }, { "grad_norm": 0.1594727635383606, "learning_rate": 9.922831318858457e-05, "loss": 0.0251, "step": 30960 }, { "grad_norm": 0.27370795607566833, "learning_rate": 9.922734829816688e-05, "loss": 0.0281, "step": 30970 }, { "grad_norm": 0.2623414099216461, "learning_rate": 9.922638280959045e-05, "loss": 0.0265, "step": 30980 }, { "grad_norm": 0.259732186794281, "learning_rate": 9.922541672286699e-05, "loss": 0.0284, "step": 30990 }, { "grad_norm": 0.304641991853714, "learning_rate": 9.922445003800823e-05, "loss": 0.0289, "step": 31000 }, { "grad_norm": 0.23132699728012085, "learning_rate": 9.922348275502593e-05, "loss": 0.0337, "step": 31010 }, { "grad_norm": 0.18196909129619598, "learning_rate": 9.922251487393184e-05, "loss": 0.0228, "step": 31020 }, { "grad_norm": 0.2511710524559021, "learning_rate": 9.922154639473771e-05, "loss": 0.0282, "step": 31030 }, { "grad_norm": 0.288780152797699, "learning_rate": 9.922057731745535e-05, "loss": 0.0277, "step": 31040 }, { "grad_norm": 0.1342831254005432, "learning_rate": 9.921960764209648e-05, "loss": 0.0277, "step": 31050 }, { "grad_norm": 0.17221124470233917, "learning_rate": 9.92186373686729e-05, "loss": 0.0318, "step": 31060 }, { "grad_norm": 0.22486203908920288, "learning_rate": 9.921766649719642e-05, "loss": 0.0274, "step": 31070 }, { "grad_norm": 0.2122229039669037, "learning_rate": 9.921669502767882e-05, "loss": 0.0299, "step": 31080 }, { "grad_norm": 0.14941759407520294, "learning_rate": 9.921572296013191e-05, "loss": 0.0319, "step": 31090 }, { "grad_norm": 0.2243933379650116, "learning_rate": 9.921475029456749e-05, "loss": 0.0261, "step": 31100 }, { "grad_norm": 0.2545417547225952, "learning_rate": 9.92137770309974e-05, "loss": 0.0259, "step": 31110 }, { "grad_norm": 0.2882518768310547, "learning_rate": 9.921280316943344e-05, "loss": 0.0242, "step": 31120 }, { "grad_norm": 0.19531482458114624, "learning_rate": 9.921182870988745e-05, "loss": 0.0271, "step": 31130 }, { "grad_norm": 0.19370697438716888, "learning_rate": 9.92108536523713e-05, "loss": 0.0274, "step": 31140 }, { "grad_norm": 0.30882030725479126, "learning_rate": 9.92098779968968e-05, "loss": 0.0275, "step": 31150 }, { "grad_norm": 0.2238057553768158, "learning_rate": 9.920890174347583e-05, "loss": 0.0314, "step": 31160 }, { "grad_norm": 0.2495976686477661, "learning_rate": 9.920792489212023e-05, "loss": 0.0219, "step": 31170 }, { "grad_norm": 0.26732543110847473, "learning_rate": 9.920694744284189e-05, "loss": 0.0304, "step": 31180 }, { "grad_norm": 0.14430443942546844, "learning_rate": 9.920596939565268e-05, "loss": 0.0249, "step": 31190 }, { "grad_norm": 0.4510362446308136, "learning_rate": 9.920499075056448e-05, "loss": 0.0281, "step": 31200 }, { "grad_norm": 0.2916736602783203, "learning_rate": 9.92040115075892e-05, "loss": 0.0249, "step": 31210 }, { "grad_norm": 0.17639848589897156, "learning_rate": 9.920303166673872e-05, "loss": 0.0292, "step": 31220 }, { "grad_norm": 0.2353617548942566, "learning_rate": 9.920205122802494e-05, "loss": 0.044, "step": 31230 }, { "grad_norm": 0.15743035078048706, "learning_rate": 9.920107019145978e-05, "loss": 0.0367, "step": 31240 }, { "grad_norm": 0.2245972603559494, "learning_rate": 9.920008855705519e-05, "loss": 0.0341, "step": 31250 }, { "grad_norm": 0.22997905313968658, "learning_rate": 9.919910632482303e-05, "loss": 0.0357, "step": 31260 }, { "grad_norm": 0.20433756709098816, "learning_rate": 9.919812349477531e-05, "loss": 0.0247, "step": 31270 }, { "grad_norm": 0.29098695516586304, "learning_rate": 9.919714006692393e-05, "loss": 0.0336, "step": 31280 }, { "grad_norm": 0.21640637516975403, "learning_rate": 9.919615604128084e-05, "loss": 0.0297, "step": 31290 }, { "grad_norm": 0.29164397716522217, "learning_rate": 9.919517141785802e-05, "loss": 0.0297, "step": 31300 }, { "grad_norm": 0.18779394030570984, "learning_rate": 9.91941861966674e-05, "loss": 0.0352, "step": 31310 }, { "grad_norm": 0.18821464478969574, "learning_rate": 9.919320037772098e-05, "loss": 0.0345, "step": 31320 }, { "grad_norm": 0.15799665451049805, "learning_rate": 9.919221396103073e-05, "loss": 0.0231, "step": 31330 }, { "grad_norm": 0.19392739236354828, "learning_rate": 9.919122694660863e-05, "loss": 0.027, "step": 31340 }, { "grad_norm": 0.2792776823043823, "learning_rate": 9.919023933446669e-05, "loss": 0.0247, "step": 31350 }, { "grad_norm": 0.1832922101020813, "learning_rate": 9.918925112461688e-05, "loss": 0.0328, "step": 31360 }, { "grad_norm": 0.3004947602748871, "learning_rate": 9.918826231707126e-05, "loss": 0.0337, "step": 31370 }, { "grad_norm": 0.2552206814289093, "learning_rate": 9.918727291184178e-05, "loss": 0.0306, "step": 31380 }, { "grad_norm": 0.23371170461177826, "learning_rate": 9.918628290894049e-05, "loss": 0.0267, "step": 31390 }, { "grad_norm": 0.15668915212154388, "learning_rate": 9.918529230837945e-05, "loss": 0.0239, "step": 31400 }, { "grad_norm": 0.18267714977264404, "learning_rate": 9.918430111017066e-05, "loss": 0.022, "step": 31410 }, { "grad_norm": 0.15367397665977478, "learning_rate": 9.918330931432616e-05, "loss": 0.0271, "step": 31420 }, { "grad_norm": 0.23553310334682465, "learning_rate": 9.918231692085803e-05, "loss": 0.0271, "step": 31430 }, { "grad_norm": 0.217375710606575, "learning_rate": 9.918132392977832e-05, "loss": 0.0275, "step": 31440 }, { "grad_norm": 0.2555864155292511, "learning_rate": 9.918033034109908e-05, "loss": 0.0231, "step": 31450 }, { "grad_norm": 0.19001002609729767, "learning_rate": 9.917933615483238e-05, "loss": 0.0281, "step": 31460 }, { "grad_norm": 0.12170929461717606, "learning_rate": 9.917834137099032e-05, "loss": 0.0237, "step": 31470 }, { "grad_norm": 0.12708380818367004, "learning_rate": 9.9177345989585e-05, "loss": 0.0261, "step": 31480 }, { "grad_norm": 0.16280488669872284, "learning_rate": 9.917635001062847e-05, "loss": 0.0269, "step": 31490 }, { "grad_norm": 0.5775290131568909, "learning_rate": 9.917535343413287e-05, "loss": 0.0303, "step": 31500 }, { "grad_norm": 0.24653004109859467, "learning_rate": 9.917435626011029e-05, "loss": 0.0304, "step": 31510 }, { "grad_norm": 0.2308112233877182, "learning_rate": 9.917335848857286e-05, "loss": 0.0282, "step": 31520 }, { "grad_norm": 0.2265971153974533, "learning_rate": 9.917236011953268e-05, "loss": 0.0286, "step": 31530 }, { "grad_norm": 0.15788307785987854, "learning_rate": 9.917136115300192e-05, "loss": 0.0264, "step": 31540 }, { "grad_norm": 0.18172761797904968, "learning_rate": 9.917036158899269e-05, "loss": 0.0261, "step": 31550 }, { "grad_norm": 0.20888212323188782, "learning_rate": 9.916936142751713e-05, "loss": 0.0264, "step": 31560 }, { "grad_norm": 0.26808667182922363, "learning_rate": 9.916836066858743e-05, "loss": 0.0264, "step": 31570 }, { "grad_norm": 0.1786404848098755, "learning_rate": 9.91673593122157e-05, "loss": 0.0333, "step": 31580 }, { "grad_norm": 0.23652039468288422, "learning_rate": 9.916635735841415e-05, "loss": 0.0267, "step": 31590 }, { "grad_norm": 0.18775145709514618, "learning_rate": 9.916535480719493e-05, "loss": 0.0296, "step": 31600 }, { "grad_norm": 0.2732885181903839, "learning_rate": 9.916435165857023e-05, "loss": 0.0278, "step": 31610 }, { "grad_norm": 0.31044888496398926, "learning_rate": 9.916334791255224e-05, "loss": 0.0239, "step": 31620 }, { "grad_norm": 0.13907742500305176, "learning_rate": 9.916234356915315e-05, "loss": 0.0203, "step": 31630 }, { "grad_norm": 0.14270517230033875, "learning_rate": 9.916133862838517e-05, "loss": 0.0272, "step": 31640 }, { "grad_norm": 0.2338506430387497, "learning_rate": 9.916033309026051e-05, "loss": 0.0265, "step": 31650 }, { "grad_norm": 0.1836884319782257, "learning_rate": 9.91593269547914e-05, "loss": 0.0288, "step": 31660 }, { "grad_norm": 0.12652526795864105, "learning_rate": 9.915832022199003e-05, "loss": 0.0236, "step": 31670 }, { "grad_norm": 0.1685173511505127, "learning_rate": 9.915731289186866e-05, "loss": 0.0299, "step": 31680 }, { "grad_norm": 0.21033044159412384, "learning_rate": 9.915630496443955e-05, "loss": 0.0289, "step": 31690 }, { "grad_norm": 0.14011424779891968, "learning_rate": 9.91552964397149e-05, "loss": 0.0248, "step": 31700 }, { "grad_norm": 0.2421000748872757, "learning_rate": 9.9154287317707e-05, "loss": 0.0262, "step": 31710 }, { "grad_norm": 0.23370616137981415, "learning_rate": 9.91532775984281e-05, "loss": 0.0255, "step": 31720 }, { "grad_norm": 0.18893282115459442, "learning_rate": 9.915226728189047e-05, "loss": 0.0264, "step": 31730 }, { "grad_norm": 0.21343939006328583, "learning_rate": 9.915125636810638e-05, "loss": 0.0287, "step": 31740 }, { "grad_norm": 0.20113110542297363, "learning_rate": 9.915024485708812e-05, "loss": 0.0249, "step": 31750 }, { "grad_norm": 0.22152765095233917, "learning_rate": 9.914923274884798e-05, "loss": 0.0271, "step": 31760 }, { "grad_norm": 0.3326776623725891, "learning_rate": 9.914822004339825e-05, "loss": 0.0362, "step": 31770 }, { "grad_norm": 0.227987140417099, "learning_rate": 9.914720674075125e-05, "loss": 0.0304, "step": 31780 }, { "grad_norm": 0.2473517656326294, "learning_rate": 9.914619284091929e-05, "loss": 0.0383, "step": 31790 }, { "grad_norm": 0.18295642733573914, "learning_rate": 9.914517834391468e-05, "loss": 0.023, "step": 31800 }, { "grad_norm": 0.2232523411512375, "learning_rate": 9.914416324974975e-05, "loss": 0.0374, "step": 31810 }, { "grad_norm": 0.15108966827392578, "learning_rate": 9.914314755843683e-05, "loss": 0.027, "step": 31820 }, { "grad_norm": 0.18140225112438202, "learning_rate": 9.914213126998828e-05, "loss": 0.0344, "step": 31830 }, { "grad_norm": 0.27953046560287476, "learning_rate": 9.914111438441644e-05, "loss": 0.028, "step": 31840 }, { "grad_norm": 0.29717081785202026, "learning_rate": 9.914009690173364e-05, "loss": 0.0316, "step": 31850 }, { "grad_norm": 0.19237321615219116, "learning_rate": 9.91390788219523e-05, "loss": 0.0289, "step": 31860 }, { "grad_norm": 0.1088433712720871, "learning_rate": 9.913806014508474e-05, "loss": 0.0325, "step": 31870 }, { "grad_norm": 0.24778640270233154, "learning_rate": 9.913704087114335e-05, "loss": 0.0378, "step": 31880 }, { "grad_norm": 0.232936292886734, "learning_rate": 9.913602100014053e-05, "loss": 0.0379, "step": 31890 }, { "grad_norm": 0.1892034411430359, "learning_rate": 9.913500053208866e-05, "loss": 0.0267, "step": 31900 }, { "grad_norm": 0.16247424483299255, "learning_rate": 9.913397946700014e-05, "loss": 0.0334, "step": 31910 }, { "grad_norm": 0.1923115998506546, "learning_rate": 9.913295780488738e-05, "loss": 0.0362, "step": 31920 }, { "grad_norm": 0.29627710580825806, "learning_rate": 9.91319355457628e-05, "loss": 0.0253, "step": 31930 }, { "grad_norm": 0.13987348973751068, "learning_rate": 9.913091268963881e-05, "loss": 0.0235, "step": 31940 }, { "grad_norm": 0.2211812436580658, "learning_rate": 9.912988923652784e-05, "loss": 0.0337, "step": 31950 }, { "grad_norm": 0.20939025282859802, "learning_rate": 9.912886518644234e-05, "loss": 0.0309, "step": 31960 }, { "grad_norm": 0.23893098533153534, "learning_rate": 9.912784053939473e-05, "loss": 0.0261, "step": 31970 }, { "grad_norm": 0.2176782637834549, "learning_rate": 9.912681529539748e-05, "loss": 0.0266, "step": 31980 }, { "grad_norm": 0.2267174869775772, "learning_rate": 9.912578945446305e-05, "loss": 0.0313, "step": 31990 }, { "grad_norm": 0.21495211124420166, "learning_rate": 9.912476301660387e-05, "loss": 0.0207, "step": 32000 }, { "grad_norm": 0.22513438761234283, "learning_rate": 9.912373598183246e-05, "loss": 0.0316, "step": 32010 }, { "grad_norm": 0.2764797508716583, "learning_rate": 9.912270835016126e-05, "loss": 0.0328, "step": 32020 }, { "grad_norm": 0.2180551290512085, "learning_rate": 9.912168012160278e-05, "loss": 0.0328, "step": 32030 }, { "grad_norm": 0.26220986247062683, "learning_rate": 9.912065129616952e-05, "loss": 0.0307, "step": 32040 }, { "grad_norm": 0.24943552911281586, "learning_rate": 9.911962187387397e-05, "loss": 0.0293, "step": 32050 }, { "grad_norm": 0.23556815087795258, "learning_rate": 9.911859185472862e-05, "loss": 0.0249, "step": 32060 }, { "grad_norm": 0.20537197589874268, "learning_rate": 9.911756123874602e-05, "loss": 0.0253, "step": 32070 }, { "grad_norm": 0.19050712883472443, "learning_rate": 9.911653002593867e-05, "loss": 0.0206, "step": 32080 }, { "grad_norm": 0.16769810020923615, "learning_rate": 9.91154982163191e-05, "loss": 0.0339, "step": 32090 }, { "grad_norm": 0.25168970227241516, "learning_rate": 9.911446580989988e-05, "loss": 0.0266, "step": 32100 }, { "grad_norm": 0.16934305429458618, "learning_rate": 9.911343280669349e-05, "loss": 0.0245, "step": 32110 }, { "grad_norm": 0.14896444976329803, "learning_rate": 9.911239920671254e-05, "loss": 0.0233, "step": 32120 }, { "grad_norm": 0.2543586790561676, "learning_rate": 9.911136500996956e-05, "loss": 0.0243, "step": 32130 }, { "grad_norm": 0.18116368353366852, "learning_rate": 9.911033021647715e-05, "loss": 0.0249, "step": 32140 }, { "grad_norm": 0.2654705345630646, "learning_rate": 9.910929482624784e-05, "loss": 0.0216, "step": 32150 }, { "grad_norm": 0.21597802639007568, "learning_rate": 9.910825883929424e-05, "loss": 0.0219, "step": 32160 }, { "grad_norm": 0.13640500605106354, "learning_rate": 9.910722225562891e-05, "loss": 0.0175, "step": 32170 }, { "grad_norm": 0.177837073802948, "learning_rate": 9.910618507526448e-05, "loss": 0.024, "step": 32180 }, { "grad_norm": 0.1161685585975647, "learning_rate": 9.910514729821354e-05, "loss": 0.0211, "step": 32190 }, { "grad_norm": 0.18777583539485931, "learning_rate": 9.91041089244887e-05, "loss": 0.0274, "step": 32200 }, { "grad_norm": 0.25004562735557556, "learning_rate": 9.910306995410257e-05, "loss": 0.0316, "step": 32210 }, { "grad_norm": 0.1825799196958542, "learning_rate": 9.910203038706778e-05, "loss": 0.026, "step": 32220 }, { "grad_norm": 0.20259864628314972, "learning_rate": 9.910099022339696e-05, "loss": 0.0291, "step": 32230 }, { "grad_norm": 0.2422925978899002, "learning_rate": 9.909994946310275e-05, "loss": 0.0281, "step": 32240 }, { "grad_norm": 0.2994549572467804, "learning_rate": 9.909890810619779e-05, "loss": 0.0255, "step": 32250 }, { "grad_norm": 0.21356917917728424, "learning_rate": 9.909786615269474e-05, "loss": 0.0234, "step": 32260 }, { "grad_norm": 0.1970149725675583, "learning_rate": 9.909682360260625e-05, "loss": 0.0203, "step": 32270 }, { "grad_norm": 0.163401797413826, "learning_rate": 9.909578045594503e-05, "loss": 0.0333, "step": 32280 }, { "grad_norm": 0.19061064720153809, "learning_rate": 9.909473671272371e-05, "loss": 0.0262, "step": 32290 }, { "grad_norm": 0.16657108068466187, "learning_rate": 9.909369237295498e-05, "loss": 0.0295, "step": 32300 }, { "grad_norm": 0.1962669938802719, "learning_rate": 9.909264743665152e-05, "loss": 0.0206, "step": 32310 }, { "grad_norm": 0.19188398122787476, "learning_rate": 9.909160190382607e-05, "loss": 0.0253, "step": 32320 }, { "grad_norm": 0.22203366458415985, "learning_rate": 9.90905557744913e-05, "loss": 0.0296, "step": 32330 }, { "grad_norm": 0.16445651650428772, "learning_rate": 9.908950904865993e-05, "loss": 0.0258, "step": 32340 }, { "grad_norm": 0.23813042044639587, "learning_rate": 9.908846172634467e-05, "loss": 0.0239, "step": 32350 }, { "grad_norm": 0.1683516502380371, "learning_rate": 9.908741380755826e-05, "loss": 0.0251, "step": 32360 }, { "grad_norm": 0.2617533802986145, "learning_rate": 9.908636529231341e-05, "loss": 0.0287, "step": 32370 }, { "grad_norm": 0.16846904158592224, "learning_rate": 9.90853161806229e-05, "loss": 0.0261, "step": 32380 }, { "grad_norm": 0.6391136050224304, "learning_rate": 9.908426647249942e-05, "loss": 0.0266, "step": 32390 }, { "grad_norm": 0.21438190340995789, "learning_rate": 9.90832161679558e-05, "loss": 0.0354, "step": 32400 }, { "grad_norm": 0.21412278711795807, "learning_rate": 9.908216526700474e-05, "loss": 0.0239, "step": 32410 }, { "grad_norm": 0.15394553542137146, "learning_rate": 9.908111376965903e-05, "loss": 0.0296, "step": 32420 }, { "grad_norm": 0.1992381066083908, "learning_rate": 9.908006167593145e-05, "loss": 0.0256, "step": 32430 }, { "grad_norm": 0.2834840714931488, "learning_rate": 9.907900898583477e-05, "loss": 0.0263, "step": 32440 }, { "grad_norm": 0.19082380831241608, "learning_rate": 9.90779556993818e-05, "loss": 0.0261, "step": 32450 }, { "grad_norm": 0.18008118867874146, "learning_rate": 9.907690181658534e-05, "loss": 0.0246, "step": 32460 }, { "grad_norm": 0.27245715260505676, "learning_rate": 9.907584733745818e-05, "loss": 0.0232, "step": 32470 }, { "grad_norm": 0.19522233307361603, "learning_rate": 9.907479226201315e-05, "loss": 0.0209, "step": 32480 }, { "grad_norm": 0.2639898955821991, "learning_rate": 9.907373659026305e-05, "loss": 0.0219, "step": 32490 }, { "grad_norm": 0.21185877919197083, "learning_rate": 9.907268032222071e-05, "loss": 0.0221, "step": 32500 }, { "grad_norm": 0.31363460421562195, "learning_rate": 9.907162345789898e-05, "loss": 0.0258, "step": 32510 }, { "grad_norm": 0.1603327840566635, "learning_rate": 9.907056599731069e-05, "loss": 0.0333, "step": 32520 }, { "grad_norm": 0.17804007232189178, "learning_rate": 9.90695079404687e-05, "loss": 0.0317, "step": 32530 }, { "grad_norm": 0.3328550159931183, "learning_rate": 9.906844928738585e-05, "loss": 0.0334, "step": 32540 }, { "grad_norm": 0.27538952231407166, "learning_rate": 9.906739003807503e-05, "loss": 0.0287, "step": 32550 }, { "grad_norm": 0.24738378822803497, "learning_rate": 9.906633019254907e-05, "loss": 0.0273, "step": 32560 }, { "grad_norm": 0.31178227066993713, "learning_rate": 9.906526975082088e-05, "loss": 0.0241, "step": 32570 }, { "grad_norm": 0.3439294099807739, "learning_rate": 9.906420871290335e-05, "loss": 0.0293, "step": 32580 }, { "grad_norm": 0.15370222926139832, "learning_rate": 9.906314707880935e-05, "loss": 0.0227, "step": 32590 }, { "grad_norm": 0.23794305324554443, "learning_rate": 9.906208484855179e-05, "loss": 0.0304, "step": 32600 }, { "grad_norm": 0.24782998859882355, "learning_rate": 9.906102202214358e-05, "loss": 0.031, "step": 32610 }, { "grad_norm": 0.18545718491077423, "learning_rate": 9.905995859959762e-05, "loss": 0.026, "step": 32620 }, { "grad_norm": 0.17556382715702057, "learning_rate": 9.905889458092686e-05, "loss": 0.0272, "step": 32630 }, { "grad_norm": 0.7091342210769653, "learning_rate": 9.90578299661442e-05, "loss": 0.0347, "step": 32640 }, { "grad_norm": 0.21679261326789856, "learning_rate": 9.90567647552626e-05, "loss": 0.0311, "step": 32650 }, { "grad_norm": 0.21110785007476807, "learning_rate": 9.905569894829497e-05, "loss": 0.0276, "step": 32660 }, { "grad_norm": 0.2320442646741867, "learning_rate": 9.905463254525428e-05, "loss": 0.0301, "step": 32670 }, { "grad_norm": 0.24985185265541077, "learning_rate": 9.905356554615353e-05, "loss": 0.0324, "step": 32680 }, { "grad_norm": 0.2943550646305084, "learning_rate": 9.905249795100561e-05, "loss": 0.0264, "step": 32690 }, { "grad_norm": 0.2523258924484253, "learning_rate": 9.905142975982354e-05, "loss": 0.0274, "step": 32700 }, { "grad_norm": 0.46625247597694397, "learning_rate": 9.905036097262027e-05, "loss": 0.0304, "step": 32710 }, { "grad_norm": 0.1520586758852005, "learning_rate": 9.904929158940881e-05, "loss": 0.0308, "step": 32720 }, { "grad_norm": 0.13780158758163452, "learning_rate": 9.904822161020216e-05, "loss": 0.0239, "step": 32730 }, { "grad_norm": 0.20599910616874695, "learning_rate": 9.904715103501329e-05, "loss": 0.025, "step": 32740 }, { "grad_norm": 0.15927699208259583, "learning_rate": 9.904607986385523e-05, "loss": 0.0234, "step": 32750 }, { "grad_norm": 0.16458754241466522, "learning_rate": 9.9045008096741e-05, "loss": 0.0218, "step": 32760 }, { "grad_norm": 0.1537729948759079, "learning_rate": 9.90439357336836e-05, "loss": 0.0202, "step": 32770 }, { "grad_norm": 0.19798478484153748, "learning_rate": 9.904286277469609e-05, "loss": 0.0302, "step": 32780 }, { "grad_norm": 0.15863969922065735, "learning_rate": 9.90417892197915e-05, "loss": 0.0255, "step": 32790 }, { "grad_norm": 0.19900207221508026, "learning_rate": 9.904071506898285e-05, "loss": 0.0256, "step": 32800 }, { "grad_norm": 0.14725668728351593, "learning_rate": 9.903964032228323e-05, "loss": 0.0216, "step": 32810 }, { "grad_norm": 0.18537048995494843, "learning_rate": 9.903856497970568e-05, "loss": 0.0228, "step": 32820 }, { "grad_norm": 0.19808845221996307, "learning_rate": 9.903748904126325e-05, "loss": 0.0297, "step": 32830 }, { "grad_norm": 0.2563878297805786, "learning_rate": 9.903641250696902e-05, "loss": 0.0287, "step": 32840 }, { "grad_norm": 0.22384078800678253, "learning_rate": 9.903533537683609e-05, "loss": 0.0332, "step": 32850 }, { "grad_norm": 0.18375249207019806, "learning_rate": 9.903425765087755e-05, "loss": 0.0279, "step": 32860 }, { "grad_norm": 0.20987258851528168, "learning_rate": 9.903317932910647e-05, "loss": 0.0197, "step": 32870 }, { "grad_norm": 0.16086441278457642, "learning_rate": 9.903210041153596e-05, "loss": 0.0236, "step": 32880 }, { "grad_norm": 0.16750800609588623, "learning_rate": 9.903102089817915e-05, "loss": 0.0256, "step": 32890 }, { "grad_norm": 0.1732023060321808, "learning_rate": 9.902994078904915e-05, "loss": 0.0303, "step": 32900 }, { "grad_norm": 0.22124354541301727, "learning_rate": 9.902886008415906e-05, "loss": 0.025, "step": 32910 }, { "grad_norm": 0.4188512861728668, "learning_rate": 9.902777878352203e-05, "loss": 0.0252, "step": 32920 }, { "grad_norm": 0.1757868081331253, "learning_rate": 9.90266968871512e-05, "loss": 0.0221, "step": 32930 }, { "grad_norm": 0.48307502269744873, "learning_rate": 9.902561439505973e-05, "loss": 0.025, "step": 32940 }, { "grad_norm": 0.36682555079460144, "learning_rate": 9.902453130726075e-05, "loss": 0.0263, "step": 32950 }, { "grad_norm": 0.15648937225341797, "learning_rate": 9.902344762376742e-05, "loss": 0.0349, "step": 32960 }, { "grad_norm": 1.2678589820861816, "learning_rate": 9.902236334459291e-05, "loss": 0.0357, "step": 32970 }, { "grad_norm": 0.36047664284706116, "learning_rate": 9.90212784697504e-05, "loss": 0.0296, "step": 32980 }, { "grad_norm": 0.1834372580051422, "learning_rate": 9.902019299925309e-05, "loss": 0.0234, "step": 32990 }, { "grad_norm": 0.16996213793754578, "learning_rate": 9.901910693311415e-05, "loss": 0.0225, "step": 33000 }, { "grad_norm": 0.2951415181159973, "learning_rate": 9.901802027134677e-05, "loss": 0.0325, "step": 33010 }, { "grad_norm": 0.23564238846302032, "learning_rate": 9.901693301396417e-05, "loss": 0.0325, "step": 33020 }, { "grad_norm": 0.19363509118556976, "learning_rate": 9.901584516097954e-05, "loss": 0.0283, "step": 33030 }, { "grad_norm": 0.2216198742389679, "learning_rate": 9.901475671240613e-05, "loss": 0.0255, "step": 33040 }, { "grad_norm": 0.4009194076061249, "learning_rate": 9.901366766825714e-05, "loss": 0.0288, "step": 33050 }, { "grad_norm": 0.19804522395133972, "learning_rate": 9.901257802854582e-05, "loss": 0.0278, "step": 33060 }, { "grad_norm": 0.19789472222328186, "learning_rate": 9.90114877932854e-05, "loss": 0.027, "step": 33070 }, { "grad_norm": 0.3032293915748596, "learning_rate": 9.90103969624891e-05, "loss": 0.0351, "step": 33080 }, { "grad_norm": 0.2202845960855484, "learning_rate": 9.900930553617024e-05, "loss": 0.0223, "step": 33090 }, { "grad_norm": 0.14381006360054016, "learning_rate": 9.900821351434203e-05, "loss": 0.0294, "step": 33100 }, { "grad_norm": 0.22400712966918945, "learning_rate": 9.900712089701776e-05, "loss": 0.0264, "step": 33110 }, { "grad_norm": 0.18780793249607086, "learning_rate": 9.90060276842107e-05, "loss": 0.0312, "step": 33120 }, { "grad_norm": 0.20108450949192047, "learning_rate": 9.900493387593412e-05, "loss": 0.0299, "step": 33130 }, { "grad_norm": 0.21075919270515442, "learning_rate": 9.900383947220133e-05, "loss": 0.0333, "step": 33140 }, { "grad_norm": 0.13154296576976776, "learning_rate": 9.900274447302564e-05, "loss": 0.0269, "step": 33150 }, { "grad_norm": 0.23503577709197998, "learning_rate": 9.900164887842032e-05, "loss": 0.0246, "step": 33160 }, { "grad_norm": 0.17968037724494934, "learning_rate": 9.900055268839872e-05, "loss": 0.0259, "step": 33170 }, { "grad_norm": 0.27323049306869507, "learning_rate": 9.899945590297412e-05, "loss": 0.0295, "step": 33180 }, { "grad_norm": 0.11543268710374832, "learning_rate": 9.899835852215988e-05, "loss": 0.0251, "step": 33190 }, { "grad_norm": 0.18954049050807953, "learning_rate": 9.899726054596932e-05, "loss": 0.024, "step": 33200 }, { "grad_norm": 0.2299930900335312, "learning_rate": 9.899616197441579e-05, "loss": 0.0264, "step": 33210 }, { "grad_norm": 0.19761134684085846, "learning_rate": 9.89950628075126e-05, "loss": 0.0281, "step": 33220 }, { "grad_norm": 0.20294193923473358, "learning_rate": 9.899396304527317e-05, "loss": 0.0466, "step": 33230 }, { "grad_norm": 0.20392099022865295, "learning_rate": 9.899286268771084e-05, "loss": 0.0319, "step": 33240 }, { "grad_norm": 0.24197345972061157, "learning_rate": 9.899176173483894e-05, "loss": 0.0384, "step": 33250 }, { "grad_norm": 0.24732615053653717, "learning_rate": 9.899066018667089e-05, "loss": 0.029, "step": 33260 }, { "grad_norm": 0.24120581150054932, "learning_rate": 9.898955804322007e-05, "loss": 0.023, "step": 33270 }, { "grad_norm": 0.19891029596328735, "learning_rate": 9.898845530449987e-05, "loss": 0.0276, "step": 33280 }, { "grad_norm": 0.20564836263656616, "learning_rate": 9.898735197052367e-05, "loss": 0.0361, "step": 33290 }, { "grad_norm": 0.2486206293106079, "learning_rate": 9.898624804130489e-05, "loss": 0.026, "step": 33300 }, { "grad_norm": 0.3374587595462799, "learning_rate": 9.898514351685697e-05, "loss": 0.0269, "step": 33310 }, { "grad_norm": 0.14702856540679932, "learning_rate": 9.898403839719329e-05, "loss": 0.0262, "step": 33320 }, { "grad_norm": 0.20423904061317444, "learning_rate": 9.898293268232731e-05, "loss": 0.0277, "step": 33330 }, { "grad_norm": 0.5449733138084412, "learning_rate": 9.898182637227243e-05, "loss": 0.0308, "step": 33340 }, { "grad_norm": 0.18198361992835999, "learning_rate": 9.898071946704211e-05, "loss": 0.0285, "step": 33350 }, { "grad_norm": 0.1434844434261322, "learning_rate": 9.897961196664982e-05, "loss": 0.025, "step": 33360 }, { "grad_norm": 0.20079968869686127, "learning_rate": 9.8978503871109e-05, "loss": 0.0294, "step": 33370 }, { "grad_norm": 0.1999945193529129, "learning_rate": 9.89773951804331e-05, "loss": 0.0349, "step": 33380 }, { "grad_norm": 0.21376124024391174, "learning_rate": 9.897628589463561e-05, "loss": 0.0337, "step": 33390 }, { "grad_norm": 0.17458148300647736, "learning_rate": 9.897517601373002e-05, "loss": 0.0332, "step": 33400 }, { "grad_norm": 0.18888989090919495, "learning_rate": 9.897406553772979e-05, "loss": 0.0233, "step": 33410 }, { "grad_norm": 0.20759087800979614, "learning_rate": 9.897295446664843e-05, "loss": 0.0306, "step": 33420 }, { "grad_norm": 0.23208563029766083, "learning_rate": 9.897184280049943e-05, "loss": 0.0259, "step": 33430 }, { "grad_norm": 0.19128061830997467, "learning_rate": 9.89707305392963e-05, "loss": 0.0255, "step": 33440 }, { "grad_norm": 0.17950274050235748, "learning_rate": 9.896961768305255e-05, "loss": 0.0262, "step": 33450 }, { "grad_norm": 0.26823028922080994, "learning_rate": 9.896850423178173e-05, "loss": 0.032, "step": 33460 }, { "grad_norm": 0.18914546072483063, "learning_rate": 9.896739018549735e-05, "loss": 0.0221, "step": 33470 }, { "grad_norm": 0.14705105125904083, "learning_rate": 9.896627554421292e-05, "loss": 0.0262, "step": 33480 }, { "grad_norm": 0.30732959508895874, "learning_rate": 9.896516030794204e-05, "loss": 0.0264, "step": 33490 }, { "grad_norm": 0.23049677908420563, "learning_rate": 9.896404447669823e-05, "loss": 0.0309, "step": 33500 }, { "grad_norm": 0.24913299083709717, "learning_rate": 9.896292805049503e-05, "loss": 0.0314, "step": 33510 }, { "grad_norm": 0.1606157273054123, "learning_rate": 9.896181102934603e-05, "loss": 0.0181, "step": 33520 }, { "grad_norm": 0.36979594826698303, "learning_rate": 9.896069341326481e-05, "loss": 0.0302, "step": 33530 }, { "grad_norm": 0.28987160325050354, "learning_rate": 9.895957520226493e-05, "loss": 0.0219, "step": 33540 }, { "grad_norm": 0.23600251972675323, "learning_rate": 9.895845639635999e-05, "loss": 0.0332, "step": 33550 }, { "grad_norm": 0.19980113208293915, "learning_rate": 9.895733699556357e-05, "loss": 0.0269, "step": 33560 }, { "grad_norm": 0.15476632118225098, "learning_rate": 9.895621699988928e-05, "loss": 0.0394, "step": 33570 }, { "grad_norm": 0.4452591836452484, "learning_rate": 9.895509640935075e-05, "loss": 0.0233, "step": 33580 }, { "grad_norm": 0.16627287864685059, "learning_rate": 9.895397522396158e-05, "loss": 0.0266, "step": 33590 }, { "grad_norm": 0.156823992729187, "learning_rate": 9.895285344373537e-05, "loss": 0.0279, "step": 33600 }, { "grad_norm": 0.17893393337726593, "learning_rate": 9.895173106868578e-05, "loss": 0.0231, "step": 33610 }, { "grad_norm": 0.1742222011089325, "learning_rate": 9.895060809882644e-05, "loss": 0.0237, "step": 33620 }, { "grad_norm": 0.19076955318450928, "learning_rate": 9.894948453417098e-05, "loss": 0.0295, "step": 33630 }, { "grad_norm": 0.16476847231388092, "learning_rate": 9.89483603747331e-05, "loss": 0.0254, "step": 33640 }, { "grad_norm": 0.21617722511291504, "learning_rate": 9.894723562052641e-05, "loss": 0.0225, "step": 33650 }, { "grad_norm": 0.16408464312553406, "learning_rate": 9.894611027156458e-05, "loss": 0.021, "step": 33660 }, { "grad_norm": 0.20873408019542694, "learning_rate": 9.894498432786132e-05, "loss": 0.0261, "step": 33670 }, { "grad_norm": 0.1408521980047226, "learning_rate": 9.894385778943027e-05, "loss": 0.0249, "step": 33680 }, { "grad_norm": 0.20120809972286224, "learning_rate": 9.894273065628514e-05, "loss": 0.0246, "step": 33690 }, { "grad_norm": 0.17631304264068604, "learning_rate": 9.894160292843964e-05, "loss": 0.0264, "step": 33700 }, { "grad_norm": 0.11010181903839111, "learning_rate": 9.894047460590744e-05, "loss": 0.0272, "step": 33710 }, { "grad_norm": 0.1941552311182022, "learning_rate": 9.893934568870227e-05, "loss": 0.0276, "step": 33720 }, { "grad_norm": 0.22937844693660736, "learning_rate": 9.893821617683784e-05, "loss": 0.0282, "step": 33730 }, { "grad_norm": 0.17288316786289215, "learning_rate": 9.893708607032787e-05, "loss": 0.0245, "step": 33740 }, { "grad_norm": 0.13360752165317535, "learning_rate": 9.893595536918612e-05, "loss": 0.0264, "step": 33750 }, { "grad_norm": 0.1309540718793869, "learning_rate": 9.893482407342632e-05, "loss": 0.0248, "step": 33760 }, { "grad_norm": 0.15300574898719788, "learning_rate": 9.893369218306217e-05, "loss": 0.0285, "step": 33770 }, { "grad_norm": 0.1252441704273224, "learning_rate": 9.893255969810748e-05, "loss": 0.0223, "step": 33780 }, { "grad_norm": 0.16831675171852112, "learning_rate": 9.893142661857597e-05, "loss": 0.0201, "step": 33790 }, { "grad_norm": 0.24531303346157074, "learning_rate": 9.893029294448145e-05, "loss": 0.0288, "step": 33800 }, { "grad_norm": 0.3417972922325134, "learning_rate": 9.892915867583767e-05, "loss": 0.0332, "step": 33810 }, { "grad_norm": 0.363031268119812, "learning_rate": 9.89280238126584e-05, "loss": 0.0402, "step": 33820 }, { "grad_norm": 0.24525557458400726, "learning_rate": 9.892688835495747e-05, "loss": 0.029, "step": 33830 }, { "grad_norm": 0.21376247704029083, "learning_rate": 9.892575230274863e-05, "loss": 0.0269, "step": 33840 }, { "grad_norm": 0.17818856239318848, "learning_rate": 9.892461565604572e-05, "loss": 0.0272, "step": 33850 }, { "grad_norm": 0.20164771378040314, "learning_rate": 9.892347841486252e-05, "loss": 0.022, "step": 33860 }, { "grad_norm": 0.1744028776884079, "learning_rate": 9.892234057921288e-05, "loss": 0.0252, "step": 33870 }, { "grad_norm": 0.3489358425140381, "learning_rate": 9.89212021491106e-05, "loss": 0.0246, "step": 33880 }, { "grad_norm": 0.1594022661447525, "learning_rate": 9.892006312456954e-05, "loss": 0.028, "step": 33890 }, { "grad_norm": 0.13317972421646118, "learning_rate": 9.891892350560352e-05, "loss": 0.0226, "step": 33900 }, { "grad_norm": 0.2056473344564438, "learning_rate": 9.891778329222638e-05, "loss": 0.0272, "step": 33910 }, { "grad_norm": 0.3124181628227234, "learning_rate": 9.8916642484452e-05, "loss": 0.0462, "step": 33920 }, { "grad_norm": 0.30036357045173645, "learning_rate": 9.891550108229423e-05, "loss": 0.0227, "step": 33930 }, { "grad_norm": 0.582510769367218, "learning_rate": 9.891435908576692e-05, "loss": 0.0284, "step": 33940 }, { "grad_norm": 0.2161092609167099, "learning_rate": 9.891321649488398e-05, "loss": 0.0465, "step": 33950 }, { "grad_norm": 0.19592691957950592, "learning_rate": 9.891207330965928e-05, "loss": 0.0309, "step": 33960 }, { "grad_norm": 0.15165744721889496, "learning_rate": 9.89109295301067e-05, "loss": 0.0323, "step": 33970 }, { "grad_norm": 0.1957562118768692, "learning_rate": 9.890978515624017e-05, "loss": 0.0272, "step": 33980 }, { "grad_norm": 0.1638174057006836, "learning_rate": 9.890864018807354e-05, "loss": 0.0256, "step": 33990 }, { "grad_norm": 0.2270812839269638, "learning_rate": 9.890749462562077e-05, "loss": 0.0285, "step": 34000 }, { "grad_norm": 0.2719383239746094, "learning_rate": 9.890634846889577e-05, "loss": 0.0308, "step": 34010 }, { "grad_norm": 0.2839253544807434, "learning_rate": 9.890520171791245e-05, "loss": 0.0251, "step": 34020 }, { "grad_norm": 0.3601446747779846, "learning_rate": 9.890405437268475e-05, "loss": 0.029, "step": 34030 }, { "grad_norm": 0.20254404842853546, "learning_rate": 9.890290643322661e-05, "loss": 0.024, "step": 34040 }, { "grad_norm": 0.17207354307174683, "learning_rate": 9.8901757899552e-05, "loss": 0.0233, "step": 34050 }, { "grad_norm": 0.22913990914821625, "learning_rate": 9.890060877167487e-05, "loss": 0.0246, "step": 34060 }, { "grad_norm": 0.15593020617961884, "learning_rate": 9.889945904960916e-05, "loss": 0.021, "step": 34070 }, { "grad_norm": 0.1745046228170395, "learning_rate": 9.889830873336885e-05, "loss": 0.0189, "step": 34080 }, { "grad_norm": 0.14994336664676666, "learning_rate": 9.889715782296794e-05, "loss": 0.0195, "step": 34090 }, { "grad_norm": 0.23522332310676575, "learning_rate": 9.889600631842038e-05, "loss": 0.0214, "step": 34100 }, { "grad_norm": 0.24778224527835846, "learning_rate": 9.889485421974017e-05, "loss": 0.025, "step": 34110 }, { "grad_norm": 0.22241786122322083, "learning_rate": 9.889370152694133e-05, "loss": 0.0205, "step": 34120 }, { "grad_norm": 0.3407926857471466, "learning_rate": 9.889254824003785e-05, "loss": 0.0259, "step": 34130 }, { "grad_norm": 0.20256167650222778, "learning_rate": 9.889139435904375e-05, "loss": 0.0244, "step": 34140 }, { "grad_norm": 0.22276969254016876, "learning_rate": 9.889023988397303e-05, "loss": 0.024, "step": 34150 }, { "grad_norm": 0.18078915774822235, "learning_rate": 9.888908481483976e-05, "loss": 0.0265, "step": 34160 }, { "grad_norm": 0.27192211151123047, "learning_rate": 9.888792915165795e-05, "loss": 0.0329, "step": 34170 }, { "grad_norm": 0.23091620206832886, "learning_rate": 9.888677289444163e-05, "loss": 0.0243, "step": 34180 }, { "grad_norm": 0.16178615391254425, "learning_rate": 9.888561604320486e-05, "loss": 0.0233, "step": 34190 }, { "grad_norm": 0.23880699276924133, "learning_rate": 9.888445859796171e-05, "loss": 0.0256, "step": 34200 }, { "grad_norm": 0.14929920434951782, "learning_rate": 9.888330055872623e-05, "loss": 0.0238, "step": 34210 }, { "grad_norm": 0.21624407172203064, "learning_rate": 9.888214192551249e-05, "loss": 0.0264, "step": 34220 }, { "grad_norm": 0.20147554576396942, "learning_rate": 9.888098269833457e-05, "loss": 0.0252, "step": 34230 }, { "grad_norm": 0.16681689023971558, "learning_rate": 9.887982287720657e-05, "loss": 0.0203, "step": 34240 }, { "grad_norm": 0.1664818674325943, "learning_rate": 9.887866246214257e-05, "loss": 0.0289, "step": 34250 }, { "grad_norm": 0.1955612450838089, "learning_rate": 9.887750145315668e-05, "loss": 0.0254, "step": 34260 }, { "grad_norm": 0.29103437066078186, "learning_rate": 9.887633985026298e-05, "loss": 0.0248, "step": 34270 }, { "grad_norm": 0.20240943133831024, "learning_rate": 9.887517765347562e-05, "loss": 0.0231, "step": 34280 }, { "grad_norm": 0.16988322138786316, "learning_rate": 9.88740148628087e-05, "loss": 0.0275, "step": 34290 }, { "grad_norm": 0.201512411236763, "learning_rate": 9.887285147827634e-05, "loss": 0.0285, "step": 34300 }, { "grad_norm": 0.2178766131401062, "learning_rate": 9.88716874998927e-05, "loss": 0.034, "step": 34310 }, { "grad_norm": 0.4295260012149811, "learning_rate": 9.887052292767193e-05, "loss": 0.027, "step": 34320 }, { "grad_norm": 0.22426478564739227, "learning_rate": 9.886935776162816e-05, "loss": 0.0232, "step": 34330 }, { "grad_norm": 0.4261420965194702, "learning_rate": 9.886819200177554e-05, "loss": 0.031, "step": 34340 }, { "grad_norm": 0.19818077981472015, "learning_rate": 9.886702564812825e-05, "loss": 0.0281, "step": 34350 }, { "grad_norm": 0.19772152602672577, "learning_rate": 9.886585870070047e-05, "loss": 0.0272, "step": 34360 }, { "grad_norm": 0.2342953085899353, "learning_rate": 9.886469115950636e-05, "loss": 0.0247, "step": 34370 }, { "grad_norm": 0.22066231071949005, "learning_rate": 9.886352302456012e-05, "loss": 0.0263, "step": 34380 }, { "grad_norm": 0.17373241484165192, "learning_rate": 9.886235429587594e-05, "loss": 0.0247, "step": 34390 }, { "grad_norm": 0.43105098605155945, "learning_rate": 9.886118497346803e-05, "loss": 0.0227, "step": 34400 }, { "grad_norm": 0.2472200244665146, "learning_rate": 9.886001505735059e-05, "loss": 0.0286, "step": 34410 }, { "grad_norm": 0.18485133349895477, "learning_rate": 9.885884454753781e-05, "loss": 0.0248, "step": 34420 }, { "grad_norm": 0.1541675180196762, "learning_rate": 9.885767344404396e-05, "loss": 0.0276, "step": 34430 }, { "grad_norm": 0.18494561314582825, "learning_rate": 9.885650174688326e-05, "loss": 0.0281, "step": 34440 }, { "grad_norm": 0.14305149018764496, "learning_rate": 9.885532945606992e-05, "loss": 0.0267, "step": 34450 }, { "grad_norm": 0.18156039714813232, "learning_rate": 9.885415657161818e-05, "loss": 0.0228, "step": 34460 }, { "grad_norm": 0.15300337970256805, "learning_rate": 9.885298309354234e-05, "loss": 0.027, "step": 34470 }, { "grad_norm": 0.25505512952804565, "learning_rate": 9.885180902185662e-05, "loss": 0.0296, "step": 34480 }, { "grad_norm": 0.19196604192256927, "learning_rate": 9.885063435657532e-05, "loss": 0.0309, "step": 34490 }, { "grad_norm": 0.15130290389060974, "learning_rate": 9.884945909771266e-05, "loss": 0.0318, "step": 34500 }, { "grad_norm": 0.1832459717988968, "learning_rate": 9.884828324528295e-05, "loss": 0.0299, "step": 34510 }, { "grad_norm": 0.3006736636161804, "learning_rate": 9.88471067993005e-05, "loss": 0.0322, "step": 34520 }, { "grad_norm": 0.18628141283988953, "learning_rate": 9.884592975977956e-05, "loss": 0.0278, "step": 34530 }, { "grad_norm": 0.1944764405488968, "learning_rate": 9.884475212673447e-05, "loss": 0.027, "step": 34540 }, { "grad_norm": 0.13926352560520172, "learning_rate": 9.884357390017954e-05, "loss": 0.0203, "step": 34550 }, { "grad_norm": 0.22560317814350128, "learning_rate": 9.884239508012904e-05, "loss": 0.0212, "step": 34560 }, { "grad_norm": 0.19855475425720215, "learning_rate": 9.884121566659733e-05, "loss": 0.0252, "step": 34570 }, { "grad_norm": 0.1553129106760025, "learning_rate": 9.884003565959874e-05, "loss": 0.0252, "step": 34580 }, { "grad_norm": 0.15335988998413086, "learning_rate": 9.883885505914761e-05, "loss": 0.0271, "step": 34590 }, { "grad_norm": 0.24903087317943573, "learning_rate": 9.883767386525829e-05, "loss": 0.0338, "step": 34600 }, { "grad_norm": 0.2259790599346161, "learning_rate": 9.88364920779451e-05, "loss": 0.029, "step": 34610 }, { "grad_norm": 0.23239466547966003, "learning_rate": 9.883530969722243e-05, "loss": 0.0251, "step": 34620 }, { "grad_norm": 0.21097488701343536, "learning_rate": 9.883412672310463e-05, "loss": 0.0272, "step": 34630 }, { "grad_norm": 0.23114946484565735, "learning_rate": 9.88329431556061e-05, "loss": 0.0206, "step": 34640 }, { "grad_norm": 0.26605895161628723, "learning_rate": 9.883175899474119e-05, "loss": 0.0303, "step": 34650 }, { "grad_norm": 0.16765418648719788, "learning_rate": 9.883057424052433e-05, "loss": 0.0263, "step": 34660 }, { "grad_norm": 0.16972005367279053, "learning_rate": 9.882938889296987e-05, "loss": 0.0196, "step": 34670 }, { "grad_norm": 0.14902903139591217, "learning_rate": 9.882820295209223e-05, "loss": 0.0254, "step": 34680 }, { "grad_norm": 0.18503470718860626, "learning_rate": 9.882701641790582e-05, "loss": 0.0253, "step": 34690 }, { "grad_norm": 0.24438261985778809, "learning_rate": 9.882582929042507e-05, "loss": 0.0246, "step": 34700 }, { "grad_norm": 0.31350022554397583, "learning_rate": 9.88246415696644e-05, "loss": 0.0289, "step": 34710 }, { "grad_norm": 0.15912985801696777, "learning_rate": 9.882345325563822e-05, "loss": 0.0418, "step": 34720 }, { "grad_norm": 0.14490348100662231, "learning_rate": 9.882226434836099e-05, "loss": 0.0212, "step": 34730 }, { "grad_norm": 0.14923155307769775, "learning_rate": 9.882107484784715e-05, "loss": 0.0207, "step": 34740 }, { "grad_norm": 0.21207980811595917, "learning_rate": 9.881988475411116e-05, "loss": 0.025, "step": 34750 }, { "grad_norm": 0.194719135761261, "learning_rate": 9.881869406716748e-05, "loss": 0.0282, "step": 34760 }, { "grad_norm": 0.16513755917549133, "learning_rate": 9.881750278703057e-05, "loss": 0.0193, "step": 34770 }, { "grad_norm": 0.2055402398109436, "learning_rate": 9.881631091371491e-05, "loss": 0.0339, "step": 34780 }, { "grad_norm": 0.20579391717910767, "learning_rate": 9.881511844723499e-05, "loss": 0.0198, "step": 34790 }, { "grad_norm": 0.2213021069765091, "learning_rate": 9.881392538760529e-05, "loss": 0.0208, "step": 34800 }, { "grad_norm": 0.1965363621711731, "learning_rate": 9.88127317348403e-05, "loss": 0.0212, "step": 34810 }, { "grad_norm": 0.16131238639354706, "learning_rate": 9.881153748895455e-05, "loss": 0.0199, "step": 34820 }, { "grad_norm": 0.14845451712608337, "learning_rate": 9.881034264996252e-05, "loss": 0.0265, "step": 34830 }, { "grad_norm": 0.24337756633758545, "learning_rate": 9.880914721787872e-05, "loss": 0.0232, "step": 34840 }, { "grad_norm": 0.19681431353092194, "learning_rate": 9.880795119271773e-05, "loss": 0.0288, "step": 34850 }, { "grad_norm": 0.2800789475440979, "learning_rate": 9.880675457449404e-05, "loss": 0.028, "step": 34860 }, { "grad_norm": 0.212031751871109, "learning_rate": 9.880555736322221e-05, "loss": 0.0268, "step": 34870 }, { "grad_norm": 0.1524788737297058, "learning_rate": 9.880435955891676e-05, "loss": 0.0186, "step": 34880 }, { "grad_norm": 0.22178973257541656, "learning_rate": 9.880316116159229e-05, "loss": 0.0254, "step": 34890 }, { "grad_norm": 0.16344408690929413, "learning_rate": 9.880196217126332e-05, "loss": 0.0187, "step": 34900 }, { "grad_norm": 0.39213037490844727, "learning_rate": 9.880076258794441e-05, "loss": 0.0325, "step": 34910 }, { "grad_norm": 0.1956106424331665, "learning_rate": 9.87995624116502e-05, "loss": 0.0356, "step": 34920 }, { "grad_norm": 0.3528823256492615, "learning_rate": 9.87983616423952e-05, "loss": 0.0219, "step": 34930 }, { "grad_norm": 0.1820535510778427, "learning_rate": 9.879716028019405e-05, "loss": 0.0224, "step": 34940 }, { "grad_norm": 0.21144239604473114, "learning_rate": 9.879595832506132e-05, "loss": 0.0239, "step": 34950 }, { "grad_norm": 0.18143177032470703, "learning_rate": 9.879475577701163e-05, "loss": 0.0282, "step": 34960 }, { "grad_norm": 0.17896400392055511, "learning_rate": 9.879355263605958e-05, "loss": 0.0244, "step": 34970 }, { "grad_norm": 0.24293942749500275, "learning_rate": 9.879234890221982e-05, "loss": 0.032, "step": 34980 }, { "grad_norm": 0.17285561561584473, "learning_rate": 9.879114457550693e-05, "loss": 0.0297, "step": 34990 }, { "grad_norm": 0.17937086522579193, "learning_rate": 9.878993965593557e-05, "loss": 0.0269, "step": 35000 }, { "grad_norm": 0.15155017375946045, "learning_rate": 9.878873414352039e-05, "loss": 0.0297, "step": 35010 }, { "grad_norm": 0.38082581758499146, "learning_rate": 9.878752803827602e-05, "loss": 0.0257, "step": 35020 }, { "grad_norm": 0.3976885676383972, "learning_rate": 9.878632134021712e-05, "loss": 0.0287, "step": 35030 }, { "grad_norm": 0.13584232330322266, "learning_rate": 9.878511404935835e-05, "loss": 0.0288, "step": 35040 }, { "grad_norm": 0.1678888201713562, "learning_rate": 9.878390616571438e-05, "loss": 0.0264, "step": 35050 }, { "grad_norm": 0.12470997124910355, "learning_rate": 9.87826976892999e-05, "loss": 0.0237, "step": 35060 }, { "grad_norm": 0.15398812294006348, "learning_rate": 9.878148862012957e-05, "loss": 0.0305, "step": 35070 }, { "grad_norm": 0.22727103531360626, "learning_rate": 9.878027895821812e-05, "loss": 0.0257, "step": 35080 }, { "grad_norm": 0.18442212045192719, "learning_rate": 9.877906870358022e-05, "loss": 0.0321, "step": 35090 }, { "grad_norm": 0.15800118446350098, "learning_rate": 9.877785785623056e-05, "loss": 0.0363, "step": 35100 }, { "grad_norm": 0.19611555337905884, "learning_rate": 9.877664641618389e-05, "loss": 0.0243, "step": 35110 }, { "grad_norm": 0.28530049324035645, "learning_rate": 9.877543438345491e-05, "loss": 0.0244, "step": 35120 }, { "grad_norm": 0.2135738581418991, "learning_rate": 9.877422175805835e-05, "loss": 0.0256, "step": 35130 }, { "grad_norm": 0.2354225069284439, "learning_rate": 9.877300854000893e-05, "loss": 0.033, "step": 35140 }, { "grad_norm": 0.17328236997127533, "learning_rate": 9.877179472932143e-05, "loss": 0.0248, "step": 35150 }, { "grad_norm": 0.2849290668964386, "learning_rate": 9.877058032601056e-05, "loss": 0.0295, "step": 35160 }, { "grad_norm": 0.34750574827194214, "learning_rate": 9.876936533009112e-05, "loss": 0.0249, "step": 35170 }, { "grad_norm": 0.2843712568283081, "learning_rate": 9.876814974157781e-05, "loss": 0.0234, "step": 35180 }, { "grad_norm": 0.17119267582893372, "learning_rate": 9.876693356048545e-05, "loss": 0.026, "step": 35190 }, { "grad_norm": 0.2136966586112976, "learning_rate": 9.87657167868288e-05, "loss": 0.0278, "step": 35200 }, { "grad_norm": 0.15230077505111694, "learning_rate": 9.876449942062266e-05, "loss": 0.0397, "step": 35210 }, { "grad_norm": 0.2579224407672882, "learning_rate": 9.876328146188181e-05, "loss": 0.0252, "step": 35220 }, { "grad_norm": 0.1565501093864441, "learning_rate": 9.876206291062104e-05, "loss": 0.0281, "step": 35230 }, { "grad_norm": 0.2245185524225235, "learning_rate": 9.876084376685517e-05, "loss": 0.027, "step": 35240 }, { "grad_norm": 0.14660412073135376, "learning_rate": 9.8759624030599e-05, "loss": 0.0289, "step": 35250 }, { "grad_norm": 0.2146032750606537, "learning_rate": 9.875840370186738e-05, "loss": 0.02, "step": 35260 }, { "grad_norm": 0.21075789630413055, "learning_rate": 9.87571827806751e-05, "loss": 0.0245, "step": 35270 }, { "grad_norm": 0.2544015944004059, "learning_rate": 9.875596126703704e-05, "loss": 0.0269, "step": 35280 }, { "grad_norm": 0.2904157042503357, "learning_rate": 9.8754739160968e-05, "loss": 0.0299, "step": 35290 }, { "grad_norm": 0.1804998219013214, "learning_rate": 9.875351646248284e-05, "loss": 0.0296, "step": 35300 }, { "grad_norm": 0.17498324811458588, "learning_rate": 9.875229317159645e-05, "loss": 0.0224, "step": 35310 }, { "grad_norm": 0.1398436278104782, "learning_rate": 9.875106928832364e-05, "loss": 0.019, "step": 35320 }, { "grad_norm": 0.36724600195884705, "learning_rate": 9.874984481267933e-05, "loss": 0.0242, "step": 35330 }, { "grad_norm": 0.1744910627603531, "learning_rate": 9.874861974467836e-05, "loss": 0.0238, "step": 35340 }, { "grad_norm": 0.21006959676742554, "learning_rate": 9.874739408433565e-05, "loss": 0.0195, "step": 35350 }, { "grad_norm": 0.16617180407047272, "learning_rate": 9.874616783166607e-05, "loss": 0.0282, "step": 35360 }, { "grad_norm": 0.20123545825481415, "learning_rate": 9.874494098668452e-05, "loss": 0.0178, "step": 35370 }, { "grad_norm": 0.43449684977531433, "learning_rate": 9.874371354940592e-05, "loss": 0.0278, "step": 35380 }, { "grad_norm": 0.24443113803863525, "learning_rate": 9.874248551984519e-05, "loss": 0.031, "step": 35390 }, { "grad_norm": 0.21869643032550812, "learning_rate": 9.874125689801723e-05, "loss": 0.0335, "step": 35400 }, { "grad_norm": 0.1909332424402237, "learning_rate": 9.874002768393698e-05, "loss": 0.0317, "step": 35410 }, { "grad_norm": 0.29487502574920654, "learning_rate": 9.873879787761938e-05, "loss": 0.0284, "step": 35420 }, { "grad_norm": 0.23115146160125732, "learning_rate": 9.873756747907935e-05, "loss": 0.027, "step": 35430 }, { "grad_norm": 0.2307697981595993, "learning_rate": 9.873633648833187e-05, "loss": 0.0302, "step": 35440 }, { "grad_norm": 0.17338436841964722, "learning_rate": 9.873510490539189e-05, "loss": 0.0209, "step": 35450 }, { "grad_norm": 0.17397530376911163, "learning_rate": 9.873387273027438e-05, "loss": 0.0227, "step": 35460 }, { "grad_norm": 0.27361810207366943, "learning_rate": 9.873263996299429e-05, "loss": 0.021, "step": 35470 }, { "grad_norm": 0.15822891891002655, "learning_rate": 9.873140660356663e-05, "loss": 0.0206, "step": 35480 }, { "grad_norm": 0.09825539588928223, "learning_rate": 9.873017265200635e-05, "loss": 0.032, "step": 35490 }, { "grad_norm": 0.19733384251594543, "learning_rate": 9.872893810832847e-05, "loss": 0.0249, "step": 35500 }, { "grad_norm": 0.19269420206546783, "learning_rate": 9.872770297254799e-05, "loss": 0.0289, "step": 35510 }, { "grad_norm": 0.265450656414032, "learning_rate": 9.872646724467991e-05, "loss": 0.0308, "step": 35520 }, { "grad_norm": 0.20387864112854004, "learning_rate": 9.872523092473924e-05, "loss": 0.025, "step": 35530 }, { "grad_norm": 0.21783946454524994, "learning_rate": 9.872399401274101e-05, "loss": 0.0281, "step": 35540 }, { "grad_norm": 0.24725717306137085, "learning_rate": 9.872275650870027e-05, "loss": 0.0207, "step": 35550 }, { "grad_norm": 0.21694809198379517, "learning_rate": 9.872151841263201e-05, "loss": 0.0238, "step": 35560 }, { "grad_norm": 0.21318703889846802, "learning_rate": 9.872027972455132e-05, "loss": 0.0224, "step": 35570 }, { "grad_norm": 0.19202584028244019, "learning_rate": 9.871904044447322e-05, "loss": 0.02, "step": 35580 }, { "grad_norm": 0.2750649154186249, "learning_rate": 9.87178005724128e-05, "loss": 0.0268, "step": 35590 }, { "grad_norm": 0.20296350121498108, "learning_rate": 9.871656010838508e-05, "loss": 0.036, "step": 35600 }, { "grad_norm": 0.15496017038822174, "learning_rate": 9.871531905240517e-05, "loss": 0.0209, "step": 35610 }, { "grad_norm": 0.23731815814971924, "learning_rate": 9.871407740448815e-05, "loss": 0.0266, "step": 35620 }, { "grad_norm": 0.11684069782495499, "learning_rate": 9.871283516464908e-05, "loss": 0.0303, "step": 35630 }, { "grad_norm": 0.18345187604427338, "learning_rate": 9.871159233290307e-05, "loss": 0.0216, "step": 35640 }, { "grad_norm": 0.16450902819633484, "learning_rate": 9.871034890926523e-05, "loss": 0.0187, "step": 35650 }, { "grad_norm": 0.1690567433834076, "learning_rate": 9.870910489375065e-05, "loss": 0.0264, "step": 35660 }, { "grad_norm": 0.18966618180274963, "learning_rate": 9.870786028637446e-05, "loss": 0.0309, "step": 35670 }, { "grad_norm": 0.17181812226772308, "learning_rate": 9.870661508715178e-05, "loss": 0.0217, "step": 35680 }, { "grad_norm": 0.17707736790180206, "learning_rate": 9.870536929609776e-05, "loss": 0.0256, "step": 35690 }, { "grad_norm": 0.2259531170129776, "learning_rate": 9.87041229132275e-05, "loss": 0.0224, "step": 35700 }, { "grad_norm": 0.33588486909866333, "learning_rate": 9.870287593855616e-05, "loss": 0.0219, "step": 35710 }, { "grad_norm": 0.17308278381824493, "learning_rate": 9.870162837209889e-05, "loss": 0.0345, "step": 35720 }, { "grad_norm": 0.19117774069309235, "learning_rate": 9.870038021387086e-05, "loss": 0.0249, "step": 35730 }, { "grad_norm": 0.15341266989707947, "learning_rate": 9.869913146388723e-05, "loss": 0.0179, "step": 35740 }, { "grad_norm": 0.18431507050991058, "learning_rate": 9.869788212216318e-05, "loss": 0.0226, "step": 35750 }, { "grad_norm": 0.21741463243961334, "learning_rate": 9.869663218871387e-05, "loss": 0.0237, "step": 35760 }, { "grad_norm": 0.15051665902137756, "learning_rate": 9.869538166355451e-05, "loss": 0.0278, "step": 35770 }, { "grad_norm": 0.1757694035768509, "learning_rate": 9.869413054670028e-05, "loss": 0.0284, "step": 35780 }, { "grad_norm": 0.1750781238079071, "learning_rate": 9.869287883816641e-05, "loss": 0.0203, "step": 35790 }, { "grad_norm": 0.1499384492635727, "learning_rate": 9.869162653796807e-05, "loss": 0.0309, "step": 35800 }, { "grad_norm": 0.11925660818815231, "learning_rate": 9.869037364612048e-05, "loss": 0.0287, "step": 35810 }, { "grad_norm": 0.39939507842063904, "learning_rate": 9.86891201626389e-05, "loss": 0.022, "step": 35820 }, { "grad_norm": 0.1670980304479599, "learning_rate": 9.868786608753853e-05, "loss": 0.0273, "step": 35830 }, { "grad_norm": 0.15325947105884552, "learning_rate": 9.868661142083464e-05, "loss": 0.027, "step": 35840 }, { "grad_norm": 0.16408489644527435, "learning_rate": 9.868535616254243e-05, "loss": 0.0258, "step": 35850 }, { "grad_norm": 0.22807323932647705, "learning_rate": 9.868410031267718e-05, "loss": 0.0242, "step": 35860 }, { "grad_norm": 0.3251553177833557, "learning_rate": 9.868284387125415e-05, "loss": 0.0297, "step": 35870 }, { "grad_norm": 0.28291887044906616, "learning_rate": 9.868158683828862e-05, "loss": 0.0272, "step": 35880 }, { "grad_norm": 0.21168094873428345, "learning_rate": 9.868032921379582e-05, "loss": 0.0251, "step": 35890 }, { "grad_norm": 0.20257283747196198, "learning_rate": 9.867907099779106e-05, "loss": 0.0267, "step": 35900 }, { "grad_norm": 0.24375349283218384, "learning_rate": 9.867781219028964e-05, "loss": 0.0396, "step": 35910 }, { "grad_norm": 0.14169001579284668, "learning_rate": 9.867655279130683e-05, "loss": 0.0298, "step": 35920 }, { "grad_norm": 0.17476654052734375, "learning_rate": 9.867529280085795e-05, "loss": 0.029, "step": 35930 }, { "grad_norm": 0.2566589117050171, "learning_rate": 9.86740322189583e-05, "loss": 0.0258, "step": 35940 }, { "grad_norm": 0.14725810289382935, "learning_rate": 9.867277104562323e-05, "loss": 0.0253, "step": 35950 }, { "grad_norm": 0.16486133635044098, "learning_rate": 9.867150928086801e-05, "loss": 0.0267, "step": 35960 }, { "grad_norm": 0.3098093867301941, "learning_rate": 9.8670246924708e-05, "loss": 0.0283, "step": 35970 }, { "grad_norm": 0.22435274720191956, "learning_rate": 9.866898397715855e-05, "loss": 0.0249, "step": 35980 }, { "grad_norm": 0.2093319296836853, "learning_rate": 9.866772043823498e-05, "loss": 0.0323, "step": 35990 }, { "grad_norm": 0.2013726532459259, "learning_rate": 9.866645630795267e-05, "loss": 0.0292, "step": 36000 }, { "grad_norm": 0.265740305185318, "learning_rate": 9.866519158632697e-05, "loss": 0.0221, "step": 36010 }, { "grad_norm": 0.24003979563713074, "learning_rate": 9.866392627337322e-05, "loss": 0.0229, "step": 36020 }, { "grad_norm": 0.16110791265964508, "learning_rate": 9.866266036910683e-05, "loss": 0.0291, "step": 36030 }, { "grad_norm": 0.21177704632282257, "learning_rate": 9.866139387354318e-05, "loss": 0.0228, "step": 36040 }, { "grad_norm": 0.18066170811653137, "learning_rate": 9.866012678669764e-05, "loss": 0.03, "step": 36050 }, { "grad_norm": 0.18438012897968292, "learning_rate": 9.865885910858564e-05, "loss": 0.0246, "step": 36060 }, { "grad_norm": 0.26709306240081787, "learning_rate": 9.865759083922253e-05, "loss": 0.0255, "step": 36070 }, { "grad_norm": 0.1764245182275772, "learning_rate": 9.865632197862377e-05, "loss": 0.029, "step": 36080 }, { "grad_norm": 0.17691120505332947, "learning_rate": 9.865505252680476e-05, "loss": 0.0232, "step": 36090 }, { "grad_norm": 0.2747384011745453, "learning_rate": 9.865378248378091e-05, "loss": 0.029, "step": 36100 }, { "grad_norm": 0.22641588747501373, "learning_rate": 9.865251184956767e-05, "loss": 0.0233, "step": 36110 }, { "grad_norm": 0.30007457733154297, "learning_rate": 9.865124062418048e-05, "loss": 0.0241, "step": 36120 }, { "grad_norm": 0.17215481400489807, "learning_rate": 9.864996880763478e-05, "loss": 0.0182, "step": 36130 }, { "grad_norm": 0.40000078082084656, "learning_rate": 9.864869639994603e-05, "loss": 0.0308, "step": 36140 }, { "grad_norm": 0.17197778820991516, "learning_rate": 9.864742340112969e-05, "loss": 0.0234, "step": 36150 }, { "grad_norm": 0.17119231820106506, "learning_rate": 9.864614981120122e-05, "loss": 0.0263, "step": 36160 }, { "grad_norm": 0.18069013953208923, "learning_rate": 9.864487563017609e-05, "loss": 0.0227, "step": 36170 }, { "grad_norm": 0.258205384016037, "learning_rate": 9.864360085806981e-05, "loss": 0.0238, "step": 36180 }, { "grad_norm": 0.21849484741687775, "learning_rate": 9.864232549489782e-05, "loss": 0.0332, "step": 36190 }, { "grad_norm": 0.15818238258361816, "learning_rate": 9.864104954067567e-05, "loss": 0.0202, "step": 36200 }, { "grad_norm": 0.29022741317749023, "learning_rate": 9.863977299541887e-05, "loss": 0.0307, "step": 36210 }, { "grad_norm": 0.48034173250198364, "learning_rate": 9.863849585914286e-05, "loss": 0.0276, "step": 36220 }, { "grad_norm": 0.1770801544189453, "learning_rate": 9.863721813186324e-05, "loss": 0.0264, "step": 36230 }, { "grad_norm": 0.16114212572574615, "learning_rate": 9.863593981359548e-05, "loss": 0.0302, "step": 36240 }, { "grad_norm": 0.2025817632675171, "learning_rate": 9.863466090435513e-05, "loss": 0.0242, "step": 36250 }, { "grad_norm": 0.1487598419189453, "learning_rate": 9.863338140415774e-05, "loss": 0.0283, "step": 36260 }, { "grad_norm": 0.17969851195812225, "learning_rate": 9.863210131301883e-05, "loss": 0.0239, "step": 36270 }, { "grad_norm": 0.2641802132129669, "learning_rate": 9.863082063095399e-05, "loss": 0.0238, "step": 36280 }, { "grad_norm": 0.17167522013187408, "learning_rate": 9.862953935797876e-05, "loss": 0.0342, "step": 36290 }, { "grad_norm": 0.1386445015668869, "learning_rate": 9.862825749410871e-05, "loss": 0.0242, "step": 36300 }, { "grad_norm": 0.18878917396068573, "learning_rate": 9.862697503935942e-05, "loss": 0.0248, "step": 36310 }, { "grad_norm": 0.1542813926935196, "learning_rate": 9.862569199374648e-05, "loss": 0.0183, "step": 36320 }, { "grad_norm": 0.2053453028202057, "learning_rate": 9.862440835728547e-05, "loss": 0.0231, "step": 36330 }, { "grad_norm": 0.20149551331996918, "learning_rate": 9.862312412999198e-05, "loss": 0.0242, "step": 36340 }, { "grad_norm": 0.17509602010250092, "learning_rate": 9.862183931188162e-05, "loss": 0.0363, "step": 36350 }, { "grad_norm": 0.25701528787612915, "learning_rate": 9.862055390297003e-05, "loss": 0.0291, "step": 36360 }, { "grad_norm": 0.19883432984352112, "learning_rate": 9.861926790327278e-05, "loss": 0.0304, "step": 36370 }, { "grad_norm": 0.2553974390029907, "learning_rate": 9.861798131280553e-05, "loss": 0.0318, "step": 36380 }, { "grad_norm": 0.23676887154579163, "learning_rate": 9.861669413158391e-05, "loss": 0.0233, "step": 36390 }, { "grad_norm": 0.1454206258058548, "learning_rate": 9.861540635962354e-05, "loss": 0.0244, "step": 36400 }, { "grad_norm": 0.1419392079114914, "learning_rate": 9.861411799694009e-05, "loss": 0.029, "step": 36410 }, { "grad_norm": 0.15431448817253113, "learning_rate": 9.861282904354922e-05, "loss": 0.022, "step": 36420 }, { "grad_norm": 0.2259203940629959, "learning_rate": 9.861153949946655e-05, "loss": 0.0238, "step": 36430 }, { "grad_norm": 0.6131718754768372, "learning_rate": 9.86102493647078e-05, "loss": 0.0208, "step": 36440 }, { "grad_norm": 0.17172934114933014, "learning_rate": 9.860895863928863e-05, "loss": 0.0256, "step": 36450 }, { "grad_norm": 0.1740218549966812, "learning_rate": 9.86076673232247e-05, "loss": 0.0281, "step": 36460 }, { "grad_norm": 0.1596238911151886, "learning_rate": 9.860637541653172e-05, "loss": 0.0284, "step": 36470 }, { "grad_norm": 0.18743322789669037, "learning_rate": 9.860508291922541e-05, "loss": 0.0267, "step": 36480 }, { "grad_norm": 0.1927119940519333, "learning_rate": 9.860378983132143e-05, "loss": 0.0283, "step": 36490 }, { "grad_norm": 0.2109488844871521, "learning_rate": 9.860249615283552e-05, "loss": 0.0204, "step": 36500 }, { "grad_norm": 0.43746691942214966, "learning_rate": 9.860120188378339e-05, "loss": 0.0217, "step": 36510 }, { "grad_norm": 0.10112501680850983, "learning_rate": 9.859990702418077e-05, "loss": 0.0205, "step": 36520 }, { "grad_norm": 0.48664191365242004, "learning_rate": 9.85986115740434e-05, "loss": 0.0244, "step": 36530 }, { "grad_norm": 0.2916712760925293, "learning_rate": 9.859731553338701e-05, "loss": 0.0225, "step": 36540 }, { "grad_norm": 0.1966078132390976, "learning_rate": 9.859601890222735e-05, "loss": 0.0191, "step": 36550 }, { "grad_norm": 0.20493361353874207, "learning_rate": 9.859472168058018e-05, "loss": 0.0248, "step": 36560 }, { "grad_norm": 0.19388653337955475, "learning_rate": 9.859342386846127e-05, "loss": 0.0228, "step": 36570 }, { "grad_norm": 0.2215440720319748, "learning_rate": 9.859212546588636e-05, "loss": 0.0285, "step": 36580 }, { "grad_norm": 0.20838917791843414, "learning_rate": 9.859082647287126e-05, "loss": 0.0273, "step": 36590 }, { "grad_norm": 0.1537734419107437, "learning_rate": 9.858952688943174e-05, "loss": 0.0227, "step": 36600 }, { "grad_norm": 0.14507007598876953, "learning_rate": 9.85882267155836e-05, "loss": 0.0275, "step": 36610 }, { "grad_norm": 0.2185254842042923, "learning_rate": 9.858692595134263e-05, "loss": 0.0272, "step": 36620 }, { "grad_norm": 0.13625946640968323, "learning_rate": 9.858562459672462e-05, "loss": 0.0292, "step": 36630 }, { "grad_norm": 0.2054070085287094, "learning_rate": 9.858432265174541e-05, "loss": 0.028, "step": 36640 }, { "grad_norm": 0.21468466520309448, "learning_rate": 9.85830201164208e-05, "loss": 0.0284, "step": 36650 }, { "grad_norm": 0.3551252782344818, "learning_rate": 9.858171699076664e-05, "loss": 0.0251, "step": 36660 }, { "grad_norm": 0.22923903167247772, "learning_rate": 9.858041327479875e-05, "loss": 0.0291, "step": 36670 }, { "grad_norm": 0.1868377923965454, "learning_rate": 9.857910896853295e-05, "loss": 0.0235, "step": 36680 }, { "grad_norm": 0.22978636622428894, "learning_rate": 9.857780407198512e-05, "loss": 0.0258, "step": 36690 }, { "grad_norm": 0.2494896501302719, "learning_rate": 9.85764985851711e-05, "loss": 0.0264, "step": 36700 }, { "grad_norm": 0.22344441711902618, "learning_rate": 9.857519250810678e-05, "loss": 0.0303, "step": 36710 }, { "grad_norm": 0.2503279745578766, "learning_rate": 9.8573885840808e-05, "loss": 0.0275, "step": 36720 }, { "grad_norm": 0.3634963035583496, "learning_rate": 9.857257858329063e-05, "loss": 0.025, "step": 36730 }, { "grad_norm": 0.21074503660202026, "learning_rate": 9.857127073557057e-05, "loss": 0.0326, "step": 36740 }, { "grad_norm": 0.1903180480003357, "learning_rate": 9.856996229766374e-05, "loss": 0.0248, "step": 36750 }, { "grad_norm": 0.14817595481872559, "learning_rate": 9.856865326958599e-05, "loss": 0.0301, "step": 36760 }, { "grad_norm": 0.20054425299167633, "learning_rate": 9.856734365135324e-05, "loss": 0.0325, "step": 36770 }, { "grad_norm": 0.21647754311561584, "learning_rate": 9.856603344298143e-05, "loss": 0.0358, "step": 36780 }, { "grad_norm": 0.20794112980365753, "learning_rate": 9.856472264448645e-05, "loss": 0.0292, "step": 36790 }, { "grad_norm": 0.4610183537006378, "learning_rate": 9.856341125588423e-05, "loss": 0.0294, "step": 36800 }, { "grad_norm": 0.1995154768228531, "learning_rate": 9.856209927719071e-05, "loss": 0.0319, "step": 36810 }, { "grad_norm": 0.20014449954032898, "learning_rate": 9.856078670842186e-05, "loss": 0.0268, "step": 36820 }, { "grad_norm": 0.2757686972618103, "learning_rate": 9.855947354959357e-05, "loss": 0.0232, "step": 36830 }, { "grad_norm": 0.2272876650094986, "learning_rate": 9.855815980072185e-05, "loss": 0.0262, "step": 36840 }, { "grad_norm": 0.19440753757953644, "learning_rate": 9.855684546182262e-05, "loss": 0.0312, "step": 36850 }, { "grad_norm": 0.22740493714809418, "learning_rate": 9.85555305329119e-05, "loss": 0.0393, "step": 36860 }, { "grad_norm": 0.21273034811019897, "learning_rate": 9.855421501400562e-05, "loss": 0.0335, "step": 36870 }, { "grad_norm": 0.22178032994270325, "learning_rate": 9.855289890511979e-05, "loss": 0.0369, "step": 36880 }, { "grad_norm": 0.2233651727437973, "learning_rate": 9.85515822062704e-05, "loss": 0.0239, "step": 36890 }, { "grad_norm": 0.2980891466140747, "learning_rate": 9.855026491747344e-05, "loss": 0.0285, "step": 36900 }, { "grad_norm": 0.16910181939601898, "learning_rate": 9.854894703874493e-05, "loss": 0.0262, "step": 36910 }, { "grad_norm": 0.2797057032585144, "learning_rate": 9.854762857010087e-05, "loss": 0.0267, "step": 36920 }, { "grad_norm": 0.23106957972049713, "learning_rate": 9.85463095115573e-05, "loss": 0.0217, "step": 36930 }, { "grad_norm": 0.3053126335144043, "learning_rate": 9.854498986313021e-05, "loss": 0.0304, "step": 36940 }, { "grad_norm": 0.2832980155944824, "learning_rate": 9.854366962483568e-05, "loss": 0.0269, "step": 36950 }, { "grad_norm": 0.24474775791168213, "learning_rate": 9.854234879668971e-05, "loss": 0.0225, "step": 36960 }, { "grad_norm": 0.16512371599674225, "learning_rate": 9.854102737870838e-05, "loss": 0.028, "step": 36970 }, { "grad_norm": 0.21740418672561646, "learning_rate": 9.853970537090775e-05, "loss": 0.0315, "step": 36980 }, { "grad_norm": 0.1806318610906601, "learning_rate": 9.853838277330385e-05, "loss": 0.0256, "step": 36990 }, { "grad_norm": 0.20269785821437836, "learning_rate": 9.853705958591279e-05, "loss": 0.0323, "step": 37000 }, { "grad_norm": 0.8657807111740112, "learning_rate": 9.853573580875062e-05, "loss": 0.0282, "step": 37010 }, { "grad_norm": 0.14651739597320557, "learning_rate": 9.853441144183345e-05, "loss": 0.0255, "step": 37020 }, { "grad_norm": 0.12377981841564178, "learning_rate": 9.853308648517732e-05, "loss": 0.0273, "step": 37030 }, { "grad_norm": 0.23940929770469666, "learning_rate": 9.85317609387984e-05, "loss": 0.0311, "step": 37040 }, { "grad_norm": 0.14689260721206665, "learning_rate": 9.853043480271276e-05, "loss": 0.0274, "step": 37050 }, { "grad_norm": 0.5281751751899719, "learning_rate": 9.85291080769365e-05, "loss": 0.0269, "step": 37060 }, { "grad_norm": 0.17475567758083344, "learning_rate": 9.852778076148576e-05, "loss": 0.0287, "step": 37070 }, { "grad_norm": 0.163800448179245, "learning_rate": 9.852645285637668e-05, "loss": 0.0304, "step": 37080 }, { "grad_norm": 0.23849263787269592, "learning_rate": 9.852512436162538e-05, "loss": 0.0358, "step": 37090 }, { "grad_norm": 0.3985513746738434, "learning_rate": 9.8523795277248e-05, "loss": 0.0223, "step": 37100 }, { "grad_norm": 0.25832051038742065, "learning_rate": 9.852246560326069e-05, "loss": 0.026, "step": 37110 }, { "grad_norm": 0.1510312259197235, "learning_rate": 9.852113533967961e-05, "loss": 0.0252, "step": 37120 }, { "grad_norm": 0.17413625121116638, "learning_rate": 9.851980448652092e-05, "loss": 0.0334, "step": 37130 }, { "grad_norm": 0.24051493406295776, "learning_rate": 9.85184730438008e-05, "loss": 0.0271, "step": 37140 }, { "grad_norm": 0.14385543763637543, "learning_rate": 9.851714101153542e-05, "loss": 0.0272, "step": 37150 }, { "grad_norm": 0.14875982701778412, "learning_rate": 9.851580838974097e-05, "loss": 0.0274, "step": 37160 }, { "grad_norm": 0.16868697106838226, "learning_rate": 9.851447517843366e-05, "loss": 0.0244, "step": 37170 }, { "grad_norm": 0.2356998175382614, "learning_rate": 9.851314137762965e-05, "loss": 0.035, "step": 37180 }, { "grad_norm": 0.17496781051158905, "learning_rate": 9.851180698734517e-05, "loss": 0.0203, "step": 37190 }, { "grad_norm": 0.2715550363063812, "learning_rate": 9.851047200759645e-05, "loss": 0.0265, "step": 37200 }, { "grad_norm": 0.14696745574474335, "learning_rate": 9.850913643839967e-05, "loss": 0.0242, "step": 37210 }, { "grad_norm": 0.15242516994476318, "learning_rate": 9.850780027977109e-05, "loss": 0.0258, "step": 37220 }, { "grad_norm": 0.2614109218120575, "learning_rate": 9.850646353172694e-05, "loss": 0.0284, "step": 37230 }, { "grad_norm": 0.45449739694595337, "learning_rate": 9.850512619428346e-05, "loss": 0.031, "step": 37240 }, { "grad_norm": 0.4326365292072296, "learning_rate": 9.85037882674569e-05, "loss": 0.0274, "step": 37250 }, { "grad_norm": 0.19980603456497192, "learning_rate": 9.850244975126352e-05, "loss": 0.0269, "step": 37260 }, { "grad_norm": 0.19015660881996155, "learning_rate": 9.850111064571958e-05, "loss": 0.03, "step": 37270 }, { "grad_norm": 0.16166065633296967, "learning_rate": 9.849977095084135e-05, "loss": 0.0261, "step": 37280 }, { "grad_norm": 0.18083316087722778, "learning_rate": 9.849843066664512e-05, "loss": 0.0214, "step": 37290 }, { "grad_norm": 0.23018778860569, "learning_rate": 9.849708979314716e-05, "loss": 0.0318, "step": 37300 }, { "grad_norm": 0.24369868636131287, "learning_rate": 9.849574833036377e-05, "loss": 0.0314, "step": 37310 }, { "grad_norm": 0.20493128895759583, "learning_rate": 9.849440627831126e-05, "loss": 0.0286, "step": 37320 }, { "grad_norm": 0.23143096268177032, "learning_rate": 9.849306363700591e-05, "loss": 0.0251, "step": 37330 }, { "grad_norm": 0.22769521176815033, "learning_rate": 9.849172040646406e-05, "loss": 0.0238, "step": 37340 }, { "grad_norm": 0.2654460668563843, "learning_rate": 9.849037658670202e-05, "loss": 0.0234, "step": 37350 }, { "grad_norm": 0.1688818484544754, "learning_rate": 9.848903217773613e-05, "loss": 0.0278, "step": 37360 }, { "grad_norm": 0.14169715344905853, "learning_rate": 9.848768717958272e-05, "loss": 0.0271, "step": 37370 }, { "grad_norm": 0.1618756353855133, "learning_rate": 9.848634159225812e-05, "loss": 0.0237, "step": 37380 }, { "grad_norm": 0.1481938660144806, "learning_rate": 9.848499541577871e-05, "loss": 0.019, "step": 37390 }, { "grad_norm": 0.4207388460636139, "learning_rate": 9.84836486501608e-05, "loss": 0.026, "step": 37400 }, { "grad_norm": 0.1816495656967163, "learning_rate": 9.848230129542079e-05, "loss": 0.0254, "step": 37410 }, { "grad_norm": 0.13423433899879456, "learning_rate": 9.848095335157506e-05, "loss": 0.025, "step": 37420 }, { "grad_norm": 0.3501146733760834, "learning_rate": 9.847960481863997e-05, "loss": 0.0192, "step": 37430 }, { "grad_norm": 0.1637686789035797, "learning_rate": 9.847825569663189e-05, "loss": 0.0299, "step": 37440 }, { "grad_norm": 0.2669409513473511, "learning_rate": 9.847690598556726e-05, "loss": 0.0295, "step": 37450 }, { "grad_norm": 0.23511555790901184, "learning_rate": 9.847555568546243e-05, "loss": 0.0236, "step": 37460 }, { "grad_norm": 0.21044094860553741, "learning_rate": 9.847420479633384e-05, "loss": 0.0373, "step": 37470 }, { "grad_norm": 0.20751629769802094, "learning_rate": 9.847285331819788e-05, "loss": 0.029, "step": 37480 }, { "grad_norm": 0.18789900839328766, "learning_rate": 9.8471501251071e-05, "loss": 0.0284, "step": 37490 }, { "grad_norm": 0.20276401937007904, "learning_rate": 9.847014859496963e-05, "loss": 0.0303, "step": 37500 }, { "grad_norm": 0.4713238775730133, "learning_rate": 9.846879534991018e-05, "loss": 0.0241, "step": 37510 }, { "grad_norm": 0.22898828983306885, "learning_rate": 9.84674415159091e-05, "loss": 0.0318, "step": 37520 }, { "grad_norm": 0.2267298847436905, "learning_rate": 9.846608709298284e-05, "loss": 0.0272, "step": 37530 }, { "grad_norm": 0.18161068856716156, "learning_rate": 9.846473208114786e-05, "loss": 0.026, "step": 37540 }, { "grad_norm": 0.3087984621524811, "learning_rate": 9.846337648042063e-05, "loss": 0.0232, "step": 37550 }, { "grad_norm": 0.3044237494468689, "learning_rate": 9.846202029081763e-05, "loss": 0.0323, "step": 37560 }, { "grad_norm": 0.22130398452281952, "learning_rate": 9.846066351235532e-05, "loss": 0.0242, "step": 37570 }, { "grad_norm": 0.18378452956676483, "learning_rate": 9.845930614505019e-05, "loss": 0.0266, "step": 37580 }, { "grad_norm": 0.3346919119358063, "learning_rate": 9.845794818891875e-05, "loss": 0.0278, "step": 37590 }, { "grad_norm": 0.24523743987083435, "learning_rate": 9.845658964397748e-05, "loss": 0.0293, "step": 37600 }, { "grad_norm": 0.23635397851467133, "learning_rate": 9.84552305102429e-05, "loss": 0.025, "step": 37610 }, { "grad_norm": 0.16322055459022522, "learning_rate": 9.84538707877315e-05, "loss": 0.0283, "step": 37620 }, { "grad_norm": 0.24642585217952728, "learning_rate": 9.845251047645983e-05, "loss": 0.022, "step": 37630 }, { "grad_norm": 0.1827327460050583, "learning_rate": 9.845114957644442e-05, "loss": 0.0226, "step": 37640 }, { "grad_norm": 0.1805797964334488, "learning_rate": 9.84497880877018e-05, "loss": 0.026, "step": 37650 }, { "grad_norm": 0.17899341881275177, "learning_rate": 9.844842601024851e-05, "loss": 0.0316, "step": 37660 }, { "grad_norm": 0.11968603730201721, "learning_rate": 9.844706334410108e-05, "loss": 0.0236, "step": 37670 }, { "grad_norm": 0.25395429134368896, "learning_rate": 9.844570008927611e-05, "loss": 0.0267, "step": 37680 }, { "grad_norm": 0.13454420864582062, "learning_rate": 9.844433624579014e-05, "loss": 0.0223, "step": 37690 }, { "grad_norm": 0.24304060637950897, "learning_rate": 9.844297181365974e-05, "loss": 0.0255, "step": 37700 }, { "grad_norm": 0.24760229885578156, "learning_rate": 9.84416067929015e-05, "loss": 0.0225, "step": 37710 }, { "grad_norm": 0.28151270747184753, "learning_rate": 9.8440241183532e-05, "loss": 0.0271, "step": 37720 }, { "grad_norm": 0.20046484470367432, "learning_rate": 9.843887498556783e-05, "loss": 0.0215, "step": 37730 }, { "grad_norm": 0.2109517753124237, "learning_rate": 9.84375081990256e-05, "loss": 0.0213, "step": 37740 }, { "grad_norm": 0.19905130565166473, "learning_rate": 9.843614082392192e-05, "loss": 0.0265, "step": 37750 }, { "grad_norm": 0.19490860402584076, "learning_rate": 9.843477286027338e-05, "loss": 0.0356, "step": 37760 }, { "grad_norm": 0.22670979797840118, "learning_rate": 9.843340430809663e-05, "loss": 0.0269, "step": 37770 }, { "grad_norm": 0.2215762734413147, "learning_rate": 9.843203516740829e-05, "loss": 0.0241, "step": 37780 }, { "grad_norm": 0.29847225546836853, "learning_rate": 9.8430665438225e-05, "loss": 0.0271, "step": 37790 }, { "grad_norm": 0.1764344424009323, "learning_rate": 9.842929512056338e-05, "loss": 0.027, "step": 37800 }, { "grad_norm": 0.13721714913845062, "learning_rate": 9.842792421444011e-05, "loss": 0.0315, "step": 37810 }, { "grad_norm": 0.1947089284658432, "learning_rate": 9.842655271987185e-05, "loss": 0.0248, "step": 37820 }, { "grad_norm": 0.3593345284461975, "learning_rate": 9.842518063687524e-05, "loss": 0.0222, "step": 37830 }, { "grad_norm": 0.24918286502361298, "learning_rate": 9.842380796546697e-05, "loss": 0.0338, "step": 37840 }, { "grad_norm": 0.21414335072040558, "learning_rate": 9.84224347056637e-05, "loss": 0.0323, "step": 37850 }, { "grad_norm": 0.20683737099170685, "learning_rate": 9.842106085748215e-05, "loss": 0.0256, "step": 37860 }, { "grad_norm": 0.11033275723457336, "learning_rate": 9.841968642093898e-05, "loss": 0.0249, "step": 37870 }, { "grad_norm": 0.2525642216205597, "learning_rate": 9.84183113960509e-05, "loss": 0.0297, "step": 37880 }, { "grad_norm": 0.21744738519191742, "learning_rate": 9.841693578283463e-05, "loss": 0.028, "step": 37890 }, { "grad_norm": 0.17819656431674957, "learning_rate": 9.841555958130688e-05, "loss": 0.025, "step": 37900 }, { "grad_norm": 0.16668008267879486, "learning_rate": 9.841418279148435e-05, "loss": 0.025, "step": 37910 }, { "grad_norm": 0.15883177518844604, "learning_rate": 9.841280541338382e-05, "loss": 0.0285, "step": 37920 }, { "grad_norm": 0.2859416902065277, "learning_rate": 9.841142744702197e-05, "loss": 0.0279, "step": 37930 }, { "grad_norm": 0.483280211687088, "learning_rate": 9.841004889241559e-05, "loss": 0.0264, "step": 37940 }, { "grad_norm": 0.16767853498458862, "learning_rate": 9.840866974958138e-05, "loss": 0.0278, "step": 37950 }, { "grad_norm": 0.16931021213531494, "learning_rate": 9.840729001853615e-05, "loss": 0.0235, "step": 37960 }, { "grad_norm": 0.18009844422340393, "learning_rate": 9.840590969929664e-05, "loss": 0.0251, "step": 37970 }, { "grad_norm": 0.13159213960170746, "learning_rate": 9.840452879187962e-05, "loss": 0.0168, "step": 37980 }, { "grad_norm": 0.15368826687335968, "learning_rate": 9.840314729630186e-05, "loss": 0.0285, "step": 37990 }, { "grad_norm": 0.19547516107559204, "learning_rate": 9.840176521258017e-05, "loss": 0.0392, "step": 38000 }, { "grad_norm": 0.23376943171024323, "learning_rate": 9.840038254073134e-05, "loss": 0.0315, "step": 38010 }, { "grad_norm": 0.19040626287460327, "learning_rate": 9.839899928077216e-05, "loss": 0.0204, "step": 38020 }, { "grad_norm": 0.13725240528583527, "learning_rate": 9.839761543271946e-05, "loss": 0.0231, "step": 38030 }, { "grad_norm": 0.15387876331806183, "learning_rate": 9.839623099659002e-05, "loss": 0.024, "step": 38040 }, { "grad_norm": 0.17442786693572998, "learning_rate": 9.839484597240067e-05, "loss": 0.0325, "step": 38050 }, { "grad_norm": 0.17558549344539642, "learning_rate": 9.839346036016827e-05, "loss": 0.0301, "step": 38060 }, { "grad_norm": 0.22310402989387512, "learning_rate": 9.839207415990963e-05, "loss": 0.0191, "step": 38070 }, { "grad_norm": 0.24513216316699982, "learning_rate": 9.83906873716416e-05, "loss": 0.0338, "step": 38080 }, { "grad_norm": 0.18691812455654144, "learning_rate": 9.838929999538102e-05, "loss": 0.0236, "step": 38090 }, { "grad_norm": 0.1993422955274582, "learning_rate": 9.838791203114477e-05, "loss": 0.0283, "step": 38100 }, { "grad_norm": 0.2463599443435669, "learning_rate": 9.83865234789497e-05, "loss": 0.0267, "step": 38110 }, { "grad_norm": 0.18747000396251678, "learning_rate": 9.838513433881268e-05, "loss": 0.0254, "step": 38120 }, { "grad_norm": 0.1744031459093094, "learning_rate": 9.838374461075059e-05, "loss": 0.0236, "step": 38130 }, { "grad_norm": 0.2276834100484848, "learning_rate": 9.838235429478033e-05, "loss": 0.026, "step": 38140 }, { "grad_norm": 0.17684420943260193, "learning_rate": 9.838096339091877e-05, "loss": 0.0225, "step": 38150 }, { "grad_norm": 0.17597059905529022, "learning_rate": 9.837957189918285e-05, "loss": 0.0263, "step": 38160 }, { "grad_norm": 0.25848329067230225, "learning_rate": 9.837817981958945e-05, "loss": 0.0273, "step": 38170 }, { "grad_norm": 0.15437474846839905, "learning_rate": 9.837678715215548e-05, "loss": 0.0297, "step": 38180 }, { "grad_norm": 0.19962477684020996, "learning_rate": 9.837539389689787e-05, "loss": 0.0354, "step": 38190 }, { "grad_norm": 0.2210417091846466, "learning_rate": 9.837400005383354e-05, "loss": 0.0257, "step": 38200 }, { "grad_norm": 0.4547201991081238, "learning_rate": 9.837260562297946e-05, "loss": 0.0293, "step": 38210 }, { "grad_norm": 0.27448368072509766, "learning_rate": 9.837121060435254e-05, "loss": 0.019, "step": 38220 }, { "grad_norm": 0.20642857253551483, "learning_rate": 9.836981499796974e-05, "loss": 0.0255, "step": 38230 }, { "grad_norm": 0.17268265783786774, "learning_rate": 9.836841880384802e-05, "loss": 0.0253, "step": 38240 }, { "grad_norm": 0.25348779559135437, "learning_rate": 9.836702202200434e-05, "loss": 0.0298, "step": 38250 }, { "grad_norm": 0.19738708436489105, "learning_rate": 9.83656246524557e-05, "loss": 0.023, "step": 38260 }, { "grad_norm": 0.20905064046382904, "learning_rate": 9.836422669521902e-05, "loss": 0.0297, "step": 38270 }, { "grad_norm": 0.40503990650177, "learning_rate": 9.836282815031134e-05, "loss": 0.0193, "step": 38280 }, { "grad_norm": 0.4149630665779114, "learning_rate": 9.836142901774963e-05, "loss": 0.0235, "step": 38290 }, { "grad_norm": 0.19732144474983215, "learning_rate": 9.83600292975509e-05, "loss": 0.0289, "step": 38300 }, { "grad_norm": 0.21318550407886505, "learning_rate": 9.835862898973216e-05, "loss": 0.0244, "step": 38310 }, { "grad_norm": 0.2005588561296463, "learning_rate": 9.83572280943104e-05, "loss": 0.0263, "step": 38320 }, { "grad_norm": 0.19899383187294006, "learning_rate": 9.83558266113027e-05, "loss": 0.0273, "step": 38330 }, { "grad_norm": 0.12326163798570633, "learning_rate": 9.8354424540726e-05, "loss": 0.0244, "step": 38340 }, { "grad_norm": 0.1304539442062378, "learning_rate": 9.835302188259743e-05, "loss": 0.0223, "step": 38350 }, { "grad_norm": 0.26998358964920044, "learning_rate": 9.835161863693398e-05, "loss": 0.021, "step": 38360 }, { "grad_norm": 0.15944725275039673, "learning_rate": 9.83502148037527e-05, "loss": 0.0216, "step": 38370 }, { "grad_norm": 0.1453494429588318, "learning_rate": 9.834881038307068e-05, "loss": 0.0297, "step": 38380 }, { "grad_norm": 0.15838050842285156, "learning_rate": 9.834740537490495e-05, "loss": 0.0205, "step": 38390 }, { "grad_norm": 0.22569021582603455, "learning_rate": 9.83459997792726e-05, "loss": 0.0233, "step": 38400 }, { "grad_norm": 0.19674859941005707, "learning_rate": 9.834459359619071e-05, "loss": 0.0231, "step": 38410 }, { "grad_norm": 0.2366604208946228, "learning_rate": 9.834318682567635e-05, "loss": 0.0336, "step": 38420 }, { "grad_norm": 0.1848134547472, "learning_rate": 9.834177946774665e-05, "loss": 0.0243, "step": 38430 }, { "grad_norm": 0.21667249500751495, "learning_rate": 9.834037152241868e-05, "loss": 0.03, "step": 38440 }, { "grad_norm": 0.2881753146648407, "learning_rate": 9.833896298970955e-05, "loss": 0.0319, "step": 38450 }, { "grad_norm": 0.17301273345947266, "learning_rate": 9.833755386963637e-05, "loss": 0.036, "step": 38460 }, { "grad_norm": 0.26825013756752014, "learning_rate": 9.833614416221628e-05, "loss": 0.0332, "step": 38470 }, { "grad_norm": 0.24325798451900482, "learning_rate": 9.833473386746641e-05, "loss": 0.0276, "step": 38480 }, { "grad_norm": 0.16307193040847778, "learning_rate": 9.833332298540388e-05, "loss": 0.0308, "step": 38490 }, { "grad_norm": 0.19065265357494354, "learning_rate": 9.833191151604583e-05, "loss": 0.0258, "step": 38500 }, { "grad_norm": 0.17470555007457733, "learning_rate": 9.833049945940945e-05, "loss": 0.03, "step": 38510 }, { "grad_norm": 0.195336252450943, "learning_rate": 9.832908681551185e-05, "loss": 0.0271, "step": 38520 }, { "grad_norm": 0.19664344191551208, "learning_rate": 9.832767358437023e-05, "loss": 0.0258, "step": 38530 }, { "grad_norm": 0.1873660534620285, "learning_rate": 9.832625976600173e-05, "loss": 0.0229, "step": 38540 }, { "grad_norm": 0.22643564641475677, "learning_rate": 9.832484536042355e-05, "loss": 0.0263, "step": 38550 }, { "grad_norm": 0.10527341067790985, "learning_rate": 9.832343036765288e-05, "loss": 0.0264, "step": 38560 }, { "grad_norm": 0.161809504032135, "learning_rate": 9.83220147877069e-05, "loss": 0.0304, "step": 38570 }, { "grad_norm": 0.23899804055690765, "learning_rate": 9.832059862060282e-05, "loss": 0.0316, "step": 38580 }, { "grad_norm": 0.1480158120393753, "learning_rate": 9.831918186635785e-05, "loss": 0.0277, "step": 38590 }, { "grad_norm": 0.29110151529312134, "learning_rate": 9.831776452498919e-05, "loss": 0.033, "step": 38600 }, { "grad_norm": 0.22981759905815125, "learning_rate": 9.831634659651407e-05, "loss": 0.025, "step": 38610 }, { "grad_norm": 0.15900534391403198, "learning_rate": 9.831492808094973e-05, "loss": 0.0215, "step": 38620 }, { "grad_norm": 0.511364221572876, "learning_rate": 9.83135089783134e-05, "loss": 0.0256, "step": 38630 }, { "grad_norm": 0.22440193593502045, "learning_rate": 9.83120892886223e-05, "loss": 0.0232, "step": 38640 }, { "grad_norm": 0.36793699860572815, "learning_rate": 9.831066901189372e-05, "loss": 0.0257, "step": 38650 }, { "grad_norm": 0.14327611029148102, "learning_rate": 9.830924814814488e-05, "loss": 0.0237, "step": 38660 }, { "grad_norm": 0.20176063477993011, "learning_rate": 9.830782669739308e-05, "loss": 0.0229, "step": 38670 }, { "grad_norm": 0.17480118572711945, "learning_rate": 9.830640465965557e-05, "loss": 0.026, "step": 38680 }, { "grad_norm": 0.15391644835472107, "learning_rate": 9.830498203494964e-05, "loss": 0.0244, "step": 38690 }, { "grad_norm": 0.1917368471622467, "learning_rate": 9.830355882329257e-05, "loss": 0.0229, "step": 38700 }, { "grad_norm": 0.16741743683815002, "learning_rate": 9.830213502470165e-05, "loss": 0.0313, "step": 38710 }, { "grad_norm": 0.25133538246154785, "learning_rate": 9.830071063919418e-05, "loss": 0.0233, "step": 38720 }, { "grad_norm": 0.17540967464447021, "learning_rate": 9.829928566678748e-05, "loss": 0.0198, "step": 38730 }, { "grad_norm": 0.18736113607883453, "learning_rate": 9.829786010749885e-05, "loss": 0.0224, "step": 38740 }, { "grad_norm": 0.18350893259048462, "learning_rate": 9.829643396134563e-05, "loss": 0.0212, "step": 38750 }, { "grad_norm": 0.1774912029504776, "learning_rate": 9.829500722834513e-05, "loss": 0.0201, "step": 38760 }, { "grad_norm": 0.38366907835006714, "learning_rate": 9.829357990851468e-05, "loss": 0.022, "step": 38770 }, { "grad_norm": 0.20968672633171082, "learning_rate": 9.829215200187167e-05, "loss": 0.0217, "step": 38780 }, { "grad_norm": 0.28494760394096375, "learning_rate": 9.82907235084334e-05, "loss": 0.022, "step": 38790 }, { "grad_norm": 0.2155420482158661, "learning_rate": 9.828929442821725e-05, "loss": 0.0196, "step": 38800 }, { "grad_norm": 0.2919505536556244, "learning_rate": 9.828786476124056e-05, "loss": 0.024, "step": 38810 }, { "grad_norm": 0.13144557178020477, "learning_rate": 9.828643450752076e-05, "loss": 0.0296, "step": 38820 }, { "grad_norm": 0.1996070146560669, "learning_rate": 9.828500366707516e-05, "loss": 0.0249, "step": 38830 }, { "grad_norm": 0.17623434960842133, "learning_rate": 9.828357223992119e-05, "loss": 0.0207, "step": 38840 }, { "grad_norm": 0.17392756044864655, "learning_rate": 9.828214022607623e-05, "loss": 0.0343, "step": 38850 }, { "grad_norm": 0.30372411012649536, "learning_rate": 9.828070762555767e-05, "loss": 0.0209, "step": 38860 }, { "grad_norm": 0.22360098361968994, "learning_rate": 9.827927443838293e-05, "loss": 0.022, "step": 38870 }, { "grad_norm": 0.23658333718776703, "learning_rate": 9.827784066456943e-05, "loss": 0.0288, "step": 38880 }, { "grad_norm": 0.18096016347408295, "learning_rate": 9.827640630413459e-05, "loss": 0.0219, "step": 38890 }, { "grad_norm": 0.22428104281425476, "learning_rate": 9.82749713570958e-05, "loss": 0.0268, "step": 38900 }, { "grad_norm": 0.23745307326316833, "learning_rate": 9.827353582347057e-05, "loss": 0.027, "step": 38910 }, { "grad_norm": 0.24163250625133514, "learning_rate": 9.827209970327629e-05, "loss": 0.0287, "step": 38920 }, { "grad_norm": 0.22126580774784088, "learning_rate": 9.82706629965304e-05, "loss": 0.0197, "step": 38930 }, { "grad_norm": 0.18714584410190582, "learning_rate": 9.82692257032504e-05, "loss": 0.0213, "step": 38940 }, { "grad_norm": 0.2173675298690796, "learning_rate": 9.826778782345374e-05, "loss": 0.0228, "step": 38950 }, { "grad_norm": 0.4319329261779785, "learning_rate": 9.826634935715788e-05, "loss": 0.0245, "step": 38960 }, { "grad_norm": 0.33221158385276794, "learning_rate": 9.82649103043803e-05, "loss": 0.0253, "step": 38970 }, { "grad_norm": 0.3520304560661316, "learning_rate": 9.826347066513847e-05, "loss": 0.0266, "step": 38980 }, { "grad_norm": 0.19553330540657043, "learning_rate": 9.826203043944994e-05, "loss": 0.0217, "step": 38990 }, { "grad_norm": 0.1419878453016281, "learning_rate": 9.826058962733216e-05, "loss": 0.0206, "step": 39000 }, { "grad_norm": 0.12710869312286377, "learning_rate": 9.825914822880264e-05, "loss": 0.0221, "step": 39010 }, { "grad_norm": 0.17326897382736206, "learning_rate": 9.825770624387891e-05, "loss": 0.0234, "step": 39020 }, { "grad_norm": 0.21165227890014648, "learning_rate": 9.82562636725785e-05, "loss": 0.0243, "step": 39030 }, { "grad_norm": 0.37467893958091736, "learning_rate": 9.82548205149189e-05, "loss": 0.0281, "step": 39040 }, { "grad_norm": 0.2407497614622116, "learning_rate": 9.825337677091768e-05, "loss": 0.0216, "step": 39050 }, { "grad_norm": 0.1339772492647171, "learning_rate": 9.825193244059238e-05, "loss": 0.0243, "step": 39060 }, { "grad_norm": 0.2295427918434143, "learning_rate": 9.825048752396055e-05, "loss": 0.0238, "step": 39070 }, { "grad_norm": 0.1641652137041092, "learning_rate": 9.824904202103972e-05, "loss": 0.0226, "step": 39080 }, { "grad_norm": 0.12873004376888275, "learning_rate": 9.824759593184749e-05, "loss": 0.0226, "step": 39090 }, { "grad_norm": 0.17020004987716675, "learning_rate": 9.824614925640141e-05, "loss": 0.0262, "step": 39100 }, { "grad_norm": 0.18328534066677094, "learning_rate": 9.824470199471907e-05, "loss": 0.0234, "step": 39110 }, { "grad_norm": 0.20036081969738007, "learning_rate": 9.824325414681805e-05, "loss": 0.0236, "step": 39120 }, { "grad_norm": 0.3638070225715637, "learning_rate": 9.824180571271594e-05, "loss": 0.0236, "step": 39130 }, { "grad_norm": 0.2579513192176819, "learning_rate": 9.824035669243035e-05, "loss": 0.0241, "step": 39140 }, { "grad_norm": 0.26567497849464417, "learning_rate": 9.823890708597887e-05, "loss": 0.0336, "step": 39150 }, { "grad_norm": 0.16402189433574677, "learning_rate": 9.823745689337912e-05, "loss": 0.0261, "step": 39160 }, { "grad_norm": 0.4351866841316223, "learning_rate": 9.823600611464874e-05, "loss": 0.0259, "step": 39170 }, { "grad_norm": 0.09302376210689545, "learning_rate": 9.823455474980532e-05, "loss": 0.0217, "step": 39180 }, { "grad_norm": 0.2385748326778412, "learning_rate": 9.823310279886654e-05, "loss": 0.0274, "step": 39190 }, { "grad_norm": 0.16888642311096191, "learning_rate": 9.823165026185001e-05, "loss": 0.0296, "step": 39200 }, { "grad_norm": 0.19828376173973083, "learning_rate": 9.823019713877339e-05, "loss": 0.0285, "step": 39210 }, { "grad_norm": 0.18828970193862915, "learning_rate": 9.822874342965435e-05, "loss": 0.0285, "step": 39220 }, { "grad_norm": 0.20659871399402618, "learning_rate": 9.822728913451053e-05, "loss": 0.0273, "step": 39230 }, { "grad_norm": 0.17136994004249573, "learning_rate": 9.822583425335963e-05, "loss": 0.0227, "step": 39240 }, { "grad_norm": 0.13161268830299377, "learning_rate": 9.822437878621931e-05, "loss": 0.0196, "step": 39250 }, { "grad_norm": 0.14868281781673431, "learning_rate": 9.822292273310723e-05, "loss": 0.0219, "step": 39260 }, { "grad_norm": 0.6854614019393921, "learning_rate": 9.822146609404112e-05, "loss": 0.0207, "step": 39270 }, { "grad_norm": 0.5809521675109863, "learning_rate": 9.822000886903869e-05, "loss": 0.0318, "step": 39280 }, { "grad_norm": 0.2714976966381073, "learning_rate": 9.821855105811761e-05, "loss": 0.0408, "step": 39290 }, { "grad_norm": 0.24788235127925873, "learning_rate": 9.82170926612956e-05, "loss": 0.0328, "step": 39300 }, { "grad_norm": 0.19388705492019653, "learning_rate": 9.82156336785904e-05, "loss": 0.0321, "step": 39310 }, { "grad_norm": 0.20978309214115143, "learning_rate": 9.821417411001971e-05, "loss": 0.0281, "step": 39320 }, { "grad_norm": 0.1968778371810913, "learning_rate": 9.82127139556013e-05, "loss": 0.0331, "step": 39330 }, { "grad_norm": 0.18645820021629333, "learning_rate": 9.82112532153529e-05, "loss": 0.0264, "step": 39340 }, { "grad_norm": 0.2060585618019104, "learning_rate": 9.820979188929225e-05, "loss": 0.0285, "step": 39350 }, { "grad_norm": 0.19337938725948334, "learning_rate": 9.82083299774371e-05, "loss": 0.0312, "step": 39360 }, { "grad_norm": 0.21223539113998413, "learning_rate": 9.820686747980525e-05, "loss": 0.0261, "step": 39370 }, { "grad_norm": 0.1437368094921112, "learning_rate": 9.820540439641441e-05, "loss": 0.0258, "step": 39380 }, { "grad_norm": 0.23206593096256256, "learning_rate": 9.820394072728243e-05, "loss": 0.025, "step": 39390 }, { "grad_norm": 0.21960191428661346, "learning_rate": 9.820247647242703e-05, "loss": 0.0303, "step": 39400 }, { "grad_norm": 0.17646656930446625, "learning_rate": 9.820101163186603e-05, "loss": 0.0397, "step": 39410 }, { "grad_norm": 0.16563713550567627, "learning_rate": 9.819954620561725e-05, "loss": 0.0183, "step": 39420 }, { "grad_norm": 0.271501362323761, "learning_rate": 9.819808019369847e-05, "loss": 0.021, "step": 39430 }, { "grad_norm": 0.24185702204704285, "learning_rate": 9.81966135961275e-05, "loss": 0.0288, "step": 39440 }, { "grad_norm": 0.1763223260641098, "learning_rate": 9.819514641292217e-05, "loss": 0.0207, "step": 39450 }, { "grad_norm": 0.1974698305130005, "learning_rate": 9.819367864410032e-05, "loss": 0.0294, "step": 39460 }, { "grad_norm": 0.16448479890823364, "learning_rate": 9.819221028967975e-05, "loss": 0.0237, "step": 39470 }, { "grad_norm": 0.11824100464582443, "learning_rate": 9.819074134967834e-05, "loss": 0.0328, "step": 39480 }, { "grad_norm": 0.22471944987773895, "learning_rate": 9.818927182411392e-05, "loss": 0.0357, "step": 39490 }, { "grad_norm": 0.23171253502368927, "learning_rate": 9.818780171300435e-05, "loss": 0.0241, "step": 39500 }, { "grad_norm": 0.1959281712770462, "learning_rate": 9.81863310163675e-05, "loss": 0.0224, "step": 39510 }, { "grad_norm": 0.2372509092092514, "learning_rate": 9.818485973422121e-05, "loss": 0.024, "step": 39520 }, { "grad_norm": 0.20096555352210999, "learning_rate": 9.818338786658339e-05, "loss": 0.0255, "step": 39530 }, { "grad_norm": 0.22452640533447266, "learning_rate": 9.818191541347193e-05, "loss": 0.026, "step": 39540 }, { "grad_norm": 0.1939859390258789, "learning_rate": 9.818044237490469e-05, "loss": 0.0309, "step": 39550 }, { "grad_norm": 0.20656414330005646, "learning_rate": 9.817896875089959e-05, "loss": 0.0273, "step": 39560 }, { "grad_norm": 0.32180750370025635, "learning_rate": 9.817749454147454e-05, "loss": 0.0274, "step": 39570 }, { "grad_norm": 0.3105609714984894, "learning_rate": 9.817601974664744e-05, "loss": 0.0282, "step": 39580 }, { "grad_norm": 0.17684578895568848, "learning_rate": 9.81745443664362e-05, "loss": 0.0265, "step": 39590 }, { "grad_norm": 0.27145424485206604, "learning_rate": 9.817306840085877e-05, "loss": 0.0261, "step": 39600 }, { "grad_norm": 0.2560100257396698, "learning_rate": 9.817159184993308e-05, "loss": 0.0239, "step": 39610 }, { "grad_norm": 0.17861822247505188, "learning_rate": 9.817011471367706e-05, "loss": 0.0204, "step": 39620 }, { "grad_norm": 0.15029464662075043, "learning_rate": 9.816863699210868e-05, "loss": 0.0269, "step": 39630 }, { "grad_norm": 0.20908279716968536, "learning_rate": 9.816715868524586e-05, "loss": 0.031, "step": 39640 }, { "grad_norm": 0.24880437552928925, "learning_rate": 9.81656797931066e-05, "loss": 0.0373, "step": 39650 }, { "grad_norm": 0.145856112241745, "learning_rate": 9.816420031570885e-05, "loss": 0.0236, "step": 39660 }, { "grad_norm": 0.24638885259628296, "learning_rate": 9.81627202530706e-05, "loss": 0.0252, "step": 39670 }, { "grad_norm": 0.4197080731391907, "learning_rate": 9.816123960520982e-05, "loss": 0.0202, "step": 39680 }, { "grad_norm": 0.44410082697868347, "learning_rate": 9.815975837214451e-05, "loss": 0.0296, "step": 39690 }, { "grad_norm": 0.12877851724624634, "learning_rate": 9.815827655389267e-05, "loss": 0.0288, "step": 39700 }, { "grad_norm": 0.16181743144989014, "learning_rate": 9.815679415047228e-05, "loss": 0.0257, "step": 39710 }, { "grad_norm": 0.18147464096546173, "learning_rate": 9.815531116190139e-05, "loss": 0.0223, "step": 39720 }, { "grad_norm": 0.2453669309616089, "learning_rate": 9.8153827588198e-05, "loss": 0.0291, "step": 39730 }, { "grad_norm": 0.22932671010494232, "learning_rate": 9.815234342938014e-05, "loss": 0.0222, "step": 39740 }, { "grad_norm": 0.2648828625679016, "learning_rate": 9.815085868546585e-05, "loss": 0.0265, "step": 39750 }, { "grad_norm": 0.1842188835144043, "learning_rate": 9.814937335647316e-05, "loss": 0.0247, "step": 39760 }, { "grad_norm": 0.3413131833076477, "learning_rate": 9.814788744242012e-05, "loss": 0.0334, "step": 39770 }, { "grad_norm": 0.12374551594257355, "learning_rate": 9.814640094332478e-05, "loss": 0.0201, "step": 39780 }, { "grad_norm": 0.14302340149879456, "learning_rate": 9.814491385920523e-05, "loss": 0.025, "step": 39790 }, { "grad_norm": 0.3678019344806671, "learning_rate": 9.814342619007951e-05, "loss": 0.0282, "step": 39800 }, { "grad_norm": 0.2095627337694168, "learning_rate": 9.81419379359657e-05, "loss": 0.0271, "step": 39810 }, { "grad_norm": 0.18824677169322968, "learning_rate": 9.814044909688191e-05, "loss": 0.0312, "step": 39820 }, { "grad_norm": 0.19556939601898193, "learning_rate": 9.813895967284621e-05, "loss": 0.0206, "step": 39830 }, { "grad_norm": 0.22012993693351746, "learning_rate": 9.813746966387669e-05, "loss": 0.0256, "step": 39840 }, { "grad_norm": 0.16440871357917786, "learning_rate": 9.813597906999146e-05, "loss": 0.0187, "step": 39850 }, { "grad_norm": 0.2495690882205963, "learning_rate": 9.813448789120864e-05, "loss": 0.0236, "step": 39860 }, { "grad_norm": 0.23512199521064758, "learning_rate": 9.813299612754634e-05, "loss": 0.0292, "step": 39870 }, { "grad_norm": 1.055298089981079, "learning_rate": 9.81315037790227e-05, "loss": 0.0306, "step": 39880 }, { "grad_norm": 0.1447969377040863, "learning_rate": 9.813001084565585e-05, "loss": 0.0243, "step": 39890 }, { "grad_norm": 0.1510382443666458, "learning_rate": 9.812851732746391e-05, "loss": 0.028, "step": 39900 }, { "grad_norm": 0.12621557712554932, "learning_rate": 9.812702322446505e-05, "loss": 0.0369, "step": 39910 }, { "grad_norm": 0.22410738468170166, "learning_rate": 9.812552853667742e-05, "loss": 0.0276, "step": 39920 }, { "grad_norm": 0.17486600577831268, "learning_rate": 9.812403326411919e-05, "loss": 0.0295, "step": 39930 }, { "grad_norm": 0.18218190968036652, "learning_rate": 9.812253740680851e-05, "loss": 0.0227, "step": 39940 }, { "grad_norm": 0.7208329439163208, "learning_rate": 9.812104096476358e-05, "loss": 0.0265, "step": 39950 }, { "grad_norm": 0.18134962022304535, "learning_rate": 9.811954393800255e-05, "loss": 0.0264, "step": 39960 }, { "grad_norm": 0.17236030101776123, "learning_rate": 9.811804632654362e-05, "loss": 0.0242, "step": 39970 }, { "grad_norm": 0.17639170587062836, "learning_rate": 9.811654813040502e-05, "loss": 0.0287, "step": 39980 }, { "grad_norm": 0.1288754791021347, "learning_rate": 9.811504934960492e-05, "loss": 0.0305, "step": 39990 }, { "grad_norm": 0.23152604699134827, "learning_rate": 9.811354998416154e-05, "loss": 0.0307, "step": 40000 }, { "grad_norm": 0.20830702781677246, "learning_rate": 9.81120500340931e-05, "loss": 0.0225, "step": 40010 }, { "grad_norm": 0.16744889318943024, "learning_rate": 9.811054949941783e-05, "loss": 0.0244, "step": 40020 }, { "grad_norm": 0.21403849124908447, "learning_rate": 9.810904838015397e-05, "loss": 0.0262, "step": 40030 }, { "grad_norm": 0.14978885650634766, "learning_rate": 9.810754667631973e-05, "loss": 0.0309, "step": 40040 }, { "grad_norm": 0.21913795173168182, "learning_rate": 9.810604438793339e-05, "loss": 0.0248, "step": 40050 }, { "grad_norm": 0.13925327360630035, "learning_rate": 9.810454151501319e-05, "loss": 0.0219, "step": 40060 }, { "grad_norm": 0.22570393979549408, "learning_rate": 9.810303805757739e-05, "loss": 0.0248, "step": 40070 }, { "grad_norm": 0.1947871446609497, "learning_rate": 9.810153401564426e-05, "loss": 0.0275, "step": 40080 }, { "grad_norm": 0.33381491899490356, "learning_rate": 9.810002938923208e-05, "loss": 0.0294, "step": 40090 }, { "grad_norm": 0.20763857662677765, "learning_rate": 9.809852417835912e-05, "loss": 0.0215, "step": 40100 }, { "grad_norm": 0.13471956551074982, "learning_rate": 9.809701838304368e-05, "loss": 0.0242, "step": 40110 }, { "grad_norm": 0.15060672163963318, "learning_rate": 9.809551200330407e-05, "loss": 0.0232, "step": 40120 }, { "grad_norm": 0.1939065158367157, "learning_rate": 9.809400503915857e-05, "loss": 0.0261, "step": 40130 }, { "grad_norm": 0.24369874596595764, "learning_rate": 9.80924974906255e-05, "loss": 0.0348, "step": 40140 }, { "grad_norm": 0.25953561067581177, "learning_rate": 9.809098935772318e-05, "loss": 0.0281, "step": 40150 }, { "grad_norm": 0.24184007942676544, "learning_rate": 9.808948064046993e-05, "loss": 0.0251, "step": 40160 }, { "grad_norm": 0.1470346301794052, "learning_rate": 9.80879713388841e-05, "loss": 0.0213, "step": 40170 }, { "grad_norm": 0.13892005383968353, "learning_rate": 9.808646145298399e-05, "loss": 0.0255, "step": 40180 }, { "grad_norm": 0.20675399899482727, "learning_rate": 9.808495098278799e-05, "loss": 0.0261, "step": 40190 }, { "grad_norm": 0.10756202042102814, "learning_rate": 9.808343992831443e-05, "loss": 0.0219, "step": 40200 }, { "grad_norm": 0.1697818487882614, "learning_rate": 9.808192828958169e-05, "loss": 0.0242, "step": 40210 }, { "grad_norm": 0.21793942153453827, "learning_rate": 9.808041606660811e-05, "loss": 0.0266, "step": 40220 }, { "grad_norm": 0.21191325783729553, "learning_rate": 9.80789032594121e-05, "loss": 0.02, "step": 40230 }, { "grad_norm": 0.19913871586322784, "learning_rate": 9.807738986801198e-05, "loss": 0.0295, "step": 40240 }, { "grad_norm": 0.14767198264598846, "learning_rate": 9.807587589242621e-05, "loss": 0.0219, "step": 40250 }, { "grad_norm": 0.4632834196090698, "learning_rate": 9.807436133267316e-05, "loss": 0.0245, "step": 40260 }, { "grad_norm": 0.18021950125694275, "learning_rate": 9.807284618877121e-05, "loss": 0.0286, "step": 40270 }, { "grad_norm": 0.21571137011051178, "learning_rate": 9.807133046073881e-05, "loss": 0.0279, "step": 40280 }, { "grad_norm": 0.13684432208538055, "learning_rate": 9.806981414859435e-05, "loss": 0.026, "step": 40290 }, { "grad_norm": 0.1911536008119583, "learning_rate": 9.806829725235627e-05, "loss": 0.0257, "step": 40300 }, { "grad_norm": 0.35926589369773865, "learning_rate": 9.806677977204298e-05, "loss": 0.027, "step": 40310 }, { "grad_norm": 0.1412472277879715, "learning_rate": 9.806526170767294e-05, "loss": 0.0215, "step": 40320 }, { "grad_norm": 0.2265111654996872, "learning_rate": 9.80637430592646e-05, "loss": 0.0291, "step": 40330 }, { "grad_norm": 0.1666867882013321, "learning_rate": 9.806222382683639e-05, "loss": 0.0188, "step": 40340 }, { "grad_norm": 0.24767448008060455, "learning_rate": 9.806070401040678e-05, "loss": 0.0206, "step": 40350 }, { "grad_norm": 0.28206393122673035, "learning_rate": 9.805918360999424e-05, "loss": 0.02, "step": 40360 }, { "grad_norm": 0.13390594720840454, "learning_rate": 9.805766262561726e-05, "loss": 0.0192, "step": 40370 }, { "grad_norm": 0.6958501935005188, "learning_rate": 9.805614105729429e-05, "loss": 0.0257, "step": 40380 }, { "grad_norm": 0.1906990259885788, "learning_rate": 9.805461890504384e-05, "loss": 0.0211, "step": 40390 }, { "grad_norm": 0.14138905704021454, "learning_rate": 9.805309616888438e-05, "loss": 0.0287, "step": 40400 }, { "grad_norm": 0.19155314564704895, "learning_rate": 9.805157284883446e-05, "loss": 0.0298, "step": 40410 }, { "grad_norm": 0.2594584822654724, "learning_rate": 9.805004894491255e-05, "loss": 0.0243, "step": 40420 }, { "grad_norm": 0.18801435828208923, "learning_rate": 9.804852445713716e-05, "loss": 0.0301, "step": 40430 }, { "grad_norm": 0.2230319380760193, "learning_rate": 9.804699938552687e-05, "loss": 0.0217, "step": 40440 }, { "grad_norm": 0.15301696956157684, "learning_rate": 9.804547373010015e-05, "loss": 0.0177, "step": 40450 }, { "grad_norm": 0.4937712848186493, "learning_rate": 9.804394749087555e-05, "loss": 0.023, "step": 40460 }, { "grad_norm": 0.17811095714569092, "learning_rate": 9.804242066787167e-05, "loss": 0.0271, "step": 40470 }, { "grad_norm": 0.188156396150589, "learning_rate": 9.804089326110697e-05, "loss": 0.028, "step": 40480 }, { "grad_norm": 0.11063109338283539, "learning_rate": 9.803936527060009e-05, "loss": 0.0249, "step": 40490 }, { "grad_norm": 0.3205508291721344, "learning_rate": 9.803783669636955e-05, "loss": 0.021, "step": 40500 }, { "grad_norm": 0.19815081357955933, "learning_rate": 9.803630753843395e-05, "loss": 0.0224, "step": 40510 }, { "grad_norm": 0.14952786266803741, "learning_rate": 9.803477779681185e-05, "loss": 0.0228, "step": 40520 }, { "grad_norm": 0.138950914144516, "learning_rate": 9.803324747152185e-05, "loss": 0.0265, "step": 40530 }, { "grad_norm": 0.1379811018705368, "learning_rate": 9.803171656258254e-05, "loss": 0.0169, "step": 40540 }, { "grad_norm": 0.23595380783081055, "learning_rate": 9.803018507001253e-05, "loss": 0.0264, "step": 40550 }, { "grad_norm": 0.1662423461675644, "learning_rate": 9.802865299383043e-05, "loss": 0.0235, "step": 40560 }, { "grad_norm": 0.18425217270851135, "learning_rate": 9.802712033405483e-05, "loss": 0.0239, "step": 40570 }, { "grad_norm": 0.16626226902008057, "learning_rate": 9.802558709070439e-05, "loss": 0.0248, "step": 40580 }, { "grad_norm": 0.6554023027420044, "learning_rate": 9.802405326379772e-05, "loss": 0.0354, "step": 40590 }, { "grad_norm": 0.14709316194057465, "learning_rate": 9.802251885335345e-05, "loss": 0.0332, "step": 40600 }, { "grad_norm": 0.21069520711898804, "learning_rate": 9.802098385939025e-05, "loss": 0.0288, "step": 40610 }, { "grad_norm": 0.22799831628799438, "learning_rate": 9.801944828192675e-05, "loss": 0.0202, "step": 40620 }, { "grad_norm": 0.20838004350662231, "learning_rate": 9.801791212098162e-05, "loss": 0.0213, "step": 40630 }, { "grad_norm": 0.19432930648326874, "learning_rate": 9.801637537657351e-05, "loss": 0.0244, "step": 40640 }, { "grad_norm": 0.11879009753465652, "learning_rate": 9.801483804872112e-05, "loss": 0.0185, "step": 40650 }, { "grad_norm": 0.20449449121952057, "learning_rate": 9.80133001374431e-05, "loss": 0.0294, "step": 40660 }, { "grad_norm": 0.17145444452762604, "learning_rate": 9.801176164275816e-05, "loss": 0.0274, "step": 40670 }, { "grad_norm": 0.18156275153160095, "learning_rate": 9.801022256468498e-05, "loss": 0.0218, "step": 40680 }, { "grad_norm": 0.21730779111385345, "learning_rate": 9.800868290324227e-05, "loss": 0.0219, "step": 40690 }, { "grad_norm": 0.13165198266506195, "learning_rate": 9.800714265844872e-05, "loss": 0.0225, "step": 40700 }, { "grad_norm": 0.14015153050422668, "learning_rate": 9.800560183032308e-05, "loss": 0.0272, "step": 40710 }, { "grad_norm": 0.22614741325378418, "learning_rate": 9.800406041888405e-05, "loss": 0.0208, "step": 40720 }, { "grad_norm": 0.18851837515830994, "learning_rate": 9.800251842415036e-05, "loss": 0.0207, "step": 40730 }, { "grad_norm": 0.161324605345726, "learning_rate": 9.800097584614075e-05, "loss": 0.0177, "step": 40740 }, { "grad_norm": 0.17542965710163116, "learning_rate": 9.799943268487397e-05, "loss": 0.0196, "step": 40750 }, { "grad_norm": 0.1914799064397812, "learning_rate": 9.799788894036875e-05, "loss": 0.0223, "step": 40760 }, { "grad_norm": 0.21647782623767853, "learning_rate": 9.799634461264387e-05, "loss": 0.0227, "step": 40770 }, { "grad_norm": 0.3087926208972931, "learning_rate": 9.799479970171808e-05, "loss": 0.0208, "step": 40780 }, { "grad_norm": 0.19420871138572693, "learning_rate": 9.799325420761016e-05, "loss": 0.0253, "step": 40790 }, { "grad_norm": 0.23535484075546265, "learning_rate": 9.799170813033889e-05, "loss": 0.0251, "step": 40800 }, { "grad_norm": 0.32242321968078613, "learning_rate": 9.799016146992306e-05, "loss": 0.0274, "step": 40810 }, { "grad_norm": 0.15206032991409302, "learning_rate": 9.798861422638147e-05, "loss": 0.0196, "step": 40820 }, { "grad_norm": 0.1635991483926773, "learning_rate": 9.798706639973288e-05, "loss": 0.0185, "step": 40830 }, { "grad_norm": 0.23825910687446594, "learning_rate": 9.798551798999613e-05, "loss": 0.022, "step": 40840 }, { "grad_norm": 0.2606455385684967, "learning_rate": 9.798396899719005e-05, "loss": 0.0227, "step": 40850 }, { "grad_norm": 0.12156105786561966, "learning_rate": 9.798241942133342e-05, "loss": 0.0184, "step": 40860 }, { "grad_norm": 0.21800200641155243, "learning_rate": 9.798086926244511e-05, "loss": 0.0244, "step": 40870 }, { "grad_norm": 0.20969337224960327, "learning_rate": 9.797931852054393e-05, "loss": 0.0245, "step": 40880 }, { "grad_norm": 0.1876085251569748, "learning_rate": 9.797776719564873e-05, "loss": 0.0231, "step": 40890 }, { "grad_norm": 0.11898674815893173, "learning_rate": 9.797621528777837e-05, "loss": 0.0263, "step": 40900 }, { "grad_norm": 0.240255668759346, "learning_rate": 9.797466279695168e-05, "loss": 0.0259, "step": 40910 }, { "grad_norm": 0.18436680734157562, "learning_rate": 9.797310972318755e-05, "loss": 0.0216, "step": 40920 }, { "grad_norm": 0.1968485563993454, "learning_rate": 9.797155606650483e-05, "loss": 0.0153, "step": 40930 }, { "grad_norm": 0.1480994075536728, "learning_rate": 9.797000182692243e-05, "loss": 0.0226, "step": 40940 }, { "grad_norm": 0.32888516783714294, "learning_rate": 9.796844700445921e-05, "loss": 0.0166, "step": 40950 }, { "grad_norm": 0.19368597865104675, "learning_rate": 9.796689159913406e-05, "loss": 0.021, "step": 40960 }, { "grad_norm": 0.28532665967941284, "learning_rate": 9.79653356109659e-05, "loss": 0.0192, "step": 40970 }, { "grad_norm": 0.131425142288208, "learning_rate": 9.796377903997363e-05, "loss": 0.0227, "step": 40980 }, { "grad_norm": 0.13920973241329193, "learning_rate": 9.796222188617615e-05, "loss": 0.0243, "step": 40990 }, { "grad_norm": 5.16571044921875, "learning_rate": 9.796066414959238e-05, "loss": 0.0304, "step": 41000 }, { "grad_norm": 0.17673282325267792, "learning_rate": 9.795910583024129e-05, "loss": 0.0235, "step": 41010 }, { "grad_norm": 0.27788758277893066, "learning_rate": 9.795754692814176e-05, "loss": 0.0238, "step": 41020 }, { "grad_norm": 0.3534852862358093, "learning_rate": 9.795598744331275e-05, "loss": 0.0363, "step": 41030 }, { "grad_norm": 0.21812044084072113, "learning_rate": 9.795442737577323e-05, "loss": 0.0284, "step": 41040 }, { "grad_norm": 0.13425102829933167, "learning_rate": 9.795286672554213e-05, "loss": 0.0295, "step": 41050 }, { "grad_norm": 0.12370077520608902, "learning_rate": 9.795130549263843e-05, "loss": 0.0206, "step": 41060 }, { "grad_norm": 0.2872648537158966, "learning_rate": 9.794974367708109e-05, "loss": 0.0237, "step": 41070 }, { "grad_norm": 0.27989575266838074, "learning_rate": 9.79481812788891e-05, "loss": 0.0202, "step": 41080 }, { "grad_norm": 0.18970036506652832, "learning_rate": 9.794661829808142e-05, "loss": 0.022, "step": 41090 }, { "grad_norm": 0.18735262751579285, "learning_rate": 9.794505473467708e-05, "loss": 0.021, "step": 41100 }, { "grad_norm": 0.16243429481983185, "learning_rate": 9.794349058869506e-05, "loss": 0.0214, "step": 41110 }, { "grad_norm": 0.17026397585868835, "learning_rate": 9.794192586015433e-05, "loss": 0.0266, "step": 41120 }, { "grad_norm": 0.28182101249694824, "learning_rate": 9.794036054907396e-05, "loss": 0.0294, "step": 41130 }, { "grad_norm": 0.2712608873844147, "learning_rate": 9.793879465547296e-05, "loss": 0.0277, "step": 41140 }, { "grad_norm": 0.19670172035694122, "learning_rate": 9.793722817937032e-05, "loss": 0.0281, "step": 41150 }, { "grad_norm": 0.16718564927577972, "learning_rate": 9.793566112078509e-05, "loss": 0.0223, "step": 41160 }, { "grad_norm": 0.3090154826641083, "learning_rate": 9.793409347973635e-05, "loss": 0.025, "step": 41170 }, { "grad_norm": 0.14188309013843536, "learning_rate": 9.79325252562431e-05, "loss": 0.024, "step": 41180 }, { "grad_norm": 0.22806189954280853, "learning_rate": 9.793095645032441e-05, "loss": 0.0195, "step": 41190 }, { "grad_norm": 0.20609910786151886, "learning_rate": 9.792938706199934e-05, "loss": 0.0233, "step": 41200 }, { "grad_norm": 0.20174121856689453, "learning_rate": 9.792781709128698e-05, "loss": 0.0225, "step": 41210 }, { "grad_norm": 0.23112569749355316, "learning_rate": 9.792624653820638e-05, "loss": 0.0223, "step": 41220 }, { "grad_norm": 0.2709551453590393, "learning_rate": 9.792467540277663e-05, "loss": 0.0309, "step": 41230 }, { "grad_norm": 0.18224088847637177, "learning_rate": 9.792310368501683e-05, "loss": 0.0219, "step": 41240 }, { "grad_norm": 0.20412950217723846, "learning_rate": 9.792153138494608e-05, "loss": 0.0271, "step": 41250 }, { "grad_norm": 0.17778009176254272, "learning_rate": 9.791995850258348e-05, "loss": 0.0273, "step": 41260 }, { "grad_norm": 0.18196648359298706, "learning_rate": 9.791838503794812e-05, "loss": 0.0271, "step": 41270 }, { "grad_norm": 0.19256600737571716, "learning_rate": 9.791681099105915e-05, "loss": 0.025, "step": 41280 }, { "grad_norm": 0.22778652608394623, "learning_rate": 9.791523636193571e-05, "loss": 0.0236, "step": 41290 }, { "grad_norm": 0.19928587973117828, "learning_rate": 9.791366115059689e-05, "loss": 0.0307, "step": 41300 }, { "grad_norm": 0.20747271180152893, "learning_rate": 9.791208535706184e-05, "loss": 0.024, "step": 41310 }, { "grad_norm": 0.17693379521369934, "learning_rate": 9.791050898134972e-05, "loss": 0.0186, "step": 41320 }, { "grad_norm": 0.2214629203081131, "learning_rate": 9.790893202347969e-05, "loss": 0.0254, "step": 41330 }, { "grad_norm": 0.301718145608902, "learning_rate": 9.79073544834709e-05, "loss": 0.0228, "step": 41340 }, { "grad_norm": 0.21780581772327423, "learning_rate": 9.79057763613425e-05, "loss": 0.0262, "step": 41350 }, { "grad_norm": 0.1717461347579956, "learning_rate": 9.790419765711372e-05, "loss": 0.0224, "step": 41360 }, { "grad_norm": 0.6013464331626892, "learning_rate": 9.790261837080368e-05, "loss": 0.0276, "step": 41370 }, { "grad_norm": 0.1930408924818039, "learning_rate": 9.790103850243161e-05, "loss": 0.025, "step": 41380 }, { "grad_norm": 0.2011304795742035, "learning_rate": 9.789945805201672e-05, "loss": 0.0201, "step": 41390 }, { "grad_norm": 0.2744230628013611, "learning_rate": 9.789787701957815e-05, "loss": 0.0219, "step": 41400 }, { "grad_norm": 0.1219785213470459, "learning_rate": 9.789629540513518e-05, "loss": 0.021, "step": 41410 }, { "grad_norm": 0.12317019701004028, "learning_rate": 9.789471320870699e-05, "loss": 0.023, "step": 41420 }, { "grad_norm": 0.10806842893362045, "learning_rate": 9.789313043031281e-05, "loss": 0.0164, "step": 41430 }, { "grad_norm": 0.12254869192838669, "learning_rate": 9.789154706997187e-05, "loss": 0.0155, "step": 41440 }, { "grad_norm": 0.17746883630752563, "learning_rate": 9.788996312770344e-05, "loss": 0.0252, "step": 41450 }, { "grad_norm": 0.17592866718769073, "learning_rate": 9.788837860352673e-05, "loss": 0.0259, "step": 41460 }, { "grad_norm": 0.2167874127626419, "learning_rate": 9.788679349746101e-05, "loss": 0.0191, "step": 41470 }, { "grad_norm": 0.45500287413597107, "learning_rate": 9.788520780952553e-05, "loss": 0.0284, "step": 41480 }, { "grad_norm": 0.21434222161769867, "learning_rate": 9.788362153973955e-05, "loss": 0.0262, "step": 41490 }, { "grad_norm": 0.6877976059913635, "learning_rate": 9.788203468812238e-05, "loss": 0.0269, "step": 41500 }, { "grad_norm": 0.38587263226509094, "learning_rate": 9.788044725469327e-05, "loss": 0.0302, "step": 41510 }, { "grad_norm": 0.27699753642082214, "learning_rate": 9.787885923947153e-05, "loss": 0.0317, "step": 41520 }, { "grad_norm": 0.1884605586528778, "learning_rate": 9.787727064247643e-05, "loss": 0.0251, "step": 41530 }, { "grad_norm": 0.1756252646446228, "learning_rate": 9.787568146372729e-05, "loss": 0.0268, "step": 41540 }, { "grad_norm": 0.16477005183696747, "learning_rate": 9.787409170324345e-05, "loss": 0.0237, "step": 41550 }, { "grad_norm": 0.25449663400650024, "learning_rate": 9.787250136104415e-05, "loss": 0.0241, "step": 41560 }, { "grad_norm": 0.25922033190727234, "learning_rate": 9.787091043714879e-05, "loss": 0.0297, "step": 41570 }, { "grad_norm": 0.19735413789749146, "learning_rate": 9.786931893157666e-05, "loss": 0.0284, "step": 41580 }, { "grad_norm": 0.1708632856607437, "learning_rate": 9.786772684434711e-05, "loss": 0.0242, "step": 41590 }, { "grad_norm": 0.1850290149450302, "learning_rate": 9.786613417547947e-05, "loss": 0.029, "step": 41600 }, { "grad_norm": 0.21160629391670227, "learning_rate": 9.786454092499314e-05, "loss": 0.0314, "step": 41610 }, { "grad_norm": 0.2936936914920807, "learning_rate": 9.786294709290741e-05, "loss": 0.0295, "step": 41620 }, { "grad_norm": 0.3347053527832031, "learning_rate": 9.78613526792417e-05, "loss": 0.0305, "step": 41630 }, { "grad_norm": 0.2044273465871811, "learning_rate": 9.785975768401537e-05, "loss": 0.0214, "step": 41640 }, { "grad_norm": 0.14680302143096924, "learning_rate": 9.785816210724779e-05, "loss": 0.0241, "step": 41650 }, { "grad_norm": 0.1619403213262558, "learning_rate": 9.785656594895836e-05, "loss": 0.0255, "step": 41660 }, { "grad_norm": 0.16230648756027222, "learning_rate": 9.785496920916646e-05, "loss": 0.0252, "step": 41670 }, { "grad_norm": 0.21792282164096832, "learning_rate": 9.785337188789151e-05, "loss": 0.0259, "step": 41680 }, { "grad_norm": 0.13160455226898193, "learning_rate": 9.78517739851529e-05, "loss": 0.0283, "step": 41690 }, { "grad_norm": 0.25251033902168274, "learning_rate": 9.785017550097006e-05, "loss": 0.0286, "step": 41700 }, { "grad_norm": 0.14436836540699005, "learning_rate": 9.784857643536243e-05, "loss": 0.0247, "step": 41710 }, { "grad_norm": 0.23208457231521606, "learning_rate": 9.78469767883494e-05, "loss": 0.0248, "step": 41720 }, { "grad_norm": 0.23579053580760956, "learning_rate": 9.784537655995043e-05, "loss": 0.0266, "step": 41730 }, { "grad_norm": 0.20660363137722015, "learning_rate": 9.784377575018496e-05, "loss": 0.0189, "step": 41740 }, { "grad_norm": 0.2307589054107666, "learning_rate": 9.784217435907244e-05, "loss": 0.0276, "step": 41750 }, { "grad_norm": 0.136100634932518, "learning_rate": 9.784057238663234e-05, "loss": 0.0243, "step": 41760 }, { "grad_norm": 0.31086164712905884, "learning_rate": 9.783896983288412e-05, "loss": 0.0261, "step": 41770 }, { "grad_norm": 0.1560056060552597, "learning_rate": 9.783736669784725e-05, "loss": 0.0195, "step": 41780 }, { "grad_norm": 0.20955219864845276, "learning_rate": 9.783576298154119e-05, "loss": 0.0261, "step": 41790 }, { "grad_norm": 0.19039228558540344, "learning_rate": 9.783415868398545e-05, "loss": 0.0239, "step": 41800 }, { "grad_norm": 0.15277385711669922, "learning_rate": 9.783255380519953e-05, "loss": 0.0251, "step": 41810 }, { "grad_norm": 0.1134861558675766, "learning_rate": 9.783094834520291e-05, "loss": 0.0186, "step": 41820 }, { "grad_norm": 0.20594370365142822, "learning_rate": 9.782934230401512e-05, "loss": 0.0182, "step": 41830 }, { "grad_norm": 0.18101942539215088, "learning_rate": 9.782773568165565e-05, "loss": 0.0301, "step": 41840 }, { "grad_norm": 0.1803562045097351, "learning_rate": 9.782612847814404e-05, "loss": 0.0237, "step": 41850 }, { "grad_norm": 0.1567118912935257, "learning_rate": 9.782452069349981e-05, "loss": 0.0333, "step": 41860 }, { "grad_norm": 0.13512718677520752, "learning_rate": 9.78229123277425e-05, "loss": 0.0225, "step": 41870 }, { "grad_norm": 0.1294996291399002, "learning_rate": 9.782130338089165e-05, "loss": 0.0297, "step": 41880 }, { "grad_norm": 0.1343621015548706, "learning_rate": 9.781969385296683e-05, "loss": 0.0217, "step": 41890 }, { "grad_norm": 0.12385815382003784, "learning_rate": 9.781808374398757e-05, "loss": 0.024, "step": 41900 }, { "grad_norm": 0.23526118695735931, "learning_rate": 9.781647305397343e-05, "loss": 0.0227, "step": 41910 }, { "grad_norm": 0.22313588857650757, "learning_rate": 9.781486178294402e-05, "loss": 0.018, "step": 41920 }, { "grad_norm": 0.1789008527994156, "learning_rate": 9.781324993091889e-05, "loss": 0.0215, "step": 41930 }, { "grad_norm": 0.23759299516677856, "learning_rate": 9.781163749791762e-05, "loss": 0.0239, "step": 41940 }, { "grad_norm": 0.1869460642337799, "learning_rate": 9.781002448395982e-05, "loss": 0.0264, "step": 41950 }, { "grad_norm": 0.217560812830925, "learning_rate": 9.780841088906508e-05, "loss": 0.0243, "step": 41960 }, { "grad_norm": 0.16673427820205688, "learning_rate": 9.7806796713253e-05, "loss": 0.0252, "step": 41970 }, { "grad_norm": 0.18753407895565033, "learning_rate": 9.780518195654322e-05, "loss": 0.0236, "step": 41980 }, { "grad_norm": 0.26328667998313904, "learning_rate": 9.780356661895534e-05, "loss": 0.0263, "step": 41990 }, { "grad_norm": 0.2502020299434662, "learning_rate": 9.780195070050898e-05, "loss": 0.0204, "step": 42000 }, { "grad_norm": 0.28394120931625366, "learning_rate": 9.78003342012238e-05, "loss": 0.0267, "step": 42010 }, { "grad_norm": 0.1698048859834671, "learning_rate": 9.779871712111942e-05, "loss": 0.0222, "step": 42020 }, { "grad_norm": 0.4230375289916992, "learning_rate": 9.77970994602155e-05, "loss": 0.0294, "step": 42030 }, { "grad_norm": 0.13659250736236572, "learning_rate": 9.779548121853169e-05, "loss": 0.039, "step": 42040 }, { "grad_norm": 0.2320864051580429, "learning_rate": 9.779386239608765e-05, "loss": 0.0274, "step": 42050 }, { "grad_norm": 0.21275371313095093, "learning_rate": 9.779224299290308e-05, "loss": 0.0265, "step": 42060 }, { "grad_norm": 0.1400972455739975, "learning_rate": 9.779062300899762e-05, "loss": 0.0221, "step": 42070 }, { "grad_norm": 0.5534012317657471, "learning_rate": 9.778900244439097e-05, "loss": 0.0215, "step": 42080 }, { "grad_norm": 0.19043008983135223, "learning_rate": 9.778738129910281e-05, "loss": 0.0244, "step": 42090 }, { "grad_norm": 0.1978493481874466, "learning_rate": 9.778575957315287e-05, "loss": 0.0206, "step": 42100 }, { "grad_norm": 0.13310272991657257, "learning_rate": 9.778413726656081e-05, "loss": 0.0221, "step": 42110 }, { "grad_norm": 0.18553927540779114, "learning_rate": 9.778251437934637e-05, "loss": 0.0233, "step": 42120 }, { "grad_norm": 0.14933042228221893, "learning_rate": 9.778089091152929e-05, "loss": 0.0247, "step": 42130 }, { "grad_norm": 0.2592235803604126, "learning_rate": 9.777926686312926e-05, "loss": 0.0242, "step": 42140 }, { "grad_norm": 0.18282543122768402, "learning_rate": 9.777764223416602e-05, "loss": 0.0254, "step": 42150 }, { "grad_norm": 0.20122969150543213, "learning_rate": 9.777601702465933e-05, "loss": 0.0234, "step": 42160 }, { "grad_norm": 0.2771543562412262, "learning_rate": 9.77743912346289e-05, "loss": 0.0239, "step": 42170 }, { "grad_norm": 0.14755934476852417, "learning_rate": 9.777276486409454e-05, "loss": 0.0193, "step": 42180 }, { "grad_norm": 0.2983400821685791, "learning_rate": 9.777113791307598e-05, "loss": 0.0241, "step": 42190 }, { "grad_norm": 0.23348736763000488, "learning_rate": 9.776951038159296e-05, "loss": 0.029, "step": 42200 }, { "grad_norm": 0.13498230278491974, "learning_rate": 9.776788226966531e-05, "loss": 0.0287, "step": 42210 }, { "grad_norm": 0.2652571201324463, "learning_rate": 9.776625357731279e-05, "loss": 0.0211, "step": 42220 }, { "grad_norm": 0.13238124549388885, "learning_rate": 9.776462430455519e-05, "loss": 0.029, "step": 42230 }, { "grad_norm": 0.19131837785243988, "learning_rate": 9.776299445141231e-05, "loss": 0.0266, "step": 42240 }, { "grad_norm": 0.15835003554821014, "learning_rate": 9.776136401790393e-05, "loss": 0.018, "step": 42250 }, { "grad_norm": 0.2197747677564621, "learning_rate": 9.77597330040499e-05, "loss": 0.0286, "step": 42260 }, { "grad_norm": 0.17009606957435608, "learning_rate": 9.775810140987003e-05, "loss": 0.0208, "step": 42270 }, { "grad_norm": 0.17848539352416992, "learning_rate": 9.775646923538411e-05, "loss": 0.0257, "step": 42280 }, { "grad_norm": 0.30258578062057495, "learning_rate": 9.775483648061203e-05, "loss": 0.0204, "step": 42290 }, { "grad_norm": 0.1494055837392807, "learning_rate": 9.775320314557358e-05, "loss": 0.0192, "step": 42300 }, { "grad_norm": 0.2306053638458252, "learning_rate": 9.775156923028863e-05, "loss": 0.0224, "step": 42310 }, { "grad_norm": 0.2936207056045532, "learning_rate": 9.774993473477702e-05, "loss": 0.0253, "step": 42320 }, { "grad_norm": 0.15868030488491058, "learning_rate": 9.774829965905863e-05, "loss": 0.0191, "step": 42330 }, { "grad_norm": 0.2991556227207184, "learning_rate": 9.774666400315333e-05, "loss": 0.029, "step": 42340 }, { "grad_norm": 0.17225655913352966, "learning_rate": 9.774502776708096e-05, "loss": 0.026, "step": 42350 }, { "grad_norm": 0.16714370250701904, "learning_rate": 9.774339095086144e-05, "loss": 0.0253, "step": 42360 }, { "grad_norm": 0.21493153274059296, "learning_rate": 9.774175355451463e-05, "loss": 0.0228, "step": 42370 }, { "grad_norm": 0.12641023099422455, "learning_rate": 9.774011557806044e-05, "loss": 0.0231, "step": 42380 }, { "grad_norm": 0.15960289537906647, "learning_rate": 9.773847702151877e-05, "loss": 0.0244, "step": 42390 }, { "grad_norm": 0.21843107044696808, "learning_rate": 9.773683788490955e-05, "loss": 0.0274, "step": 42400 }, { "grad_norm": 0.14923901855945587, "learning_rate": 9.773519816825265e-05, "loss": 0.0176, "step": 42410 }, { "grad_norm": 0.18139517307281494, "learning_rate": 9.773355787156804e-05, "loss": 0.0222, "step": 42420 }, { "grad_norm": 0.13363568484783173, "learning_rate": 9.773191699487564e-05, "loss": 0.0211, "step": 42430 }, { "grad_norm": 0.18449363112449646, "learning_rate": 9.773027553819537e-05, "loss": 0.0189, "step": 42440 }, { "grad_norm": 0.1925106644630432, "learning_rate": 9.77286335015472e-05, "loss": 0.0293, "step": 42450 }, { "grad_norm": 0.14494866132736206, "learning_rate": 9.772699088495106e-05, "loss": 0.0246, "step": 42460 }, { "grad_norm": 0.19543634355068207, "learning_rate": 9.772534768842692e-05, "loss": 0.0199, "step": 42470 }, { "grad_norm": 0.3167090117931366, "learning_rate": 9.772370391199473e-05, "loss": 0.0269, "step": 42480 }, { "grad_norm": 0.37124574184417725, "learning_rate": 9.77220595556745e-05, "loss": 0.0266, "step": 42490 }, { "grad_norm": 0.1440996378660202, "learning_rate": 9.772041461948618e-05, "loss": 0.026, "step": 42500 }, { "grad_norm": 0.17457543313503265, "learning_rate": 9.771876910344977e-05, "loss": 0.0273, "step": 42510 }, { "grad_norm": 0.14075009524822235, "learning_rate": 9.771712300758526e-05, "loss": 0.0233, "step": 42520 }, { "grad_norm": 0.16822940111160278, "learning_rate": 9.771547633191266e-05, "loss": 0.0196, "step": 42530 }, { "grad_norm": 0.17696253955364227, "learning_rate": 9.771382907645197e-05, "loss": 0.0236, "step": 42540 }, { "grad_norm": 0.13932566344738007, "learning_rate": 9.77121812412232e-05, "loss": 0.0202, "step": 42550 }, { "grad_norm": 0.3089773654937744, "learning_rate": 9.771053282624638e-05, "loss": 0.0292, "step": 42560 }, { "grad_norm": 0.2504238188266754, "learning_rate": 9.770888383154156e-05, "loss": 0.0286, "step": 42570 }, { "grad_norm": 0.24876365065574646, "learning_rate": 9.770723425712873e-05, "loss": 0.0283, "step": 42580 }, { "grad_norm": 0.274507611989975, "learning_rate": 9.770558410302796e-05, "loss": 0.0312, "step": 42590 }, { "grad_norm": 0.24674323201179504, "learning_rate": 9.770393336925932e-05, "loss": 0.0251, "step": 42600 }, { "grad_norm": 0.16604332625865936, "learning_rate": 9.770228205584284e-05, "loss": 0.026, "step": 42610 }, { "grad_norm": 0.22431707382202148, "learning_rate": 9.77006301627986e-05, "loss": 0.0318, "step": 42620 }, { "grad_norm": 0.2419140636920929, "learning_rate": 9.769897769014667e-05, "loss": 0.0266, "step": 42630 }, { "grad_norm": 0.27942416071891785, "learning_rate": 9.76973246379071e-05, "loss": 0.0263, "step": 42640 }, { "grad_norm": 0.20183588564395905, "learning_rate": 9.769567100610003e-05, "loss": 0.024, "step": 42650 }, { "grad_norm": 0.2746678292751312, "learning_rate": 9.769401679474553e-05, "loss": 0.0261, "step": 42660 }, { "grad_norm": 0.19162681698799133, "learning_rate": 9.769236200386367e-05, "loss": 0.0208, "step": 42670 }, { "grad_norm": 0.22972020506858826, "learning_rate": 9.769070663347459e-05, "loss": 0.025, "step": 42680 }, { "grad_norm": 0.2573641538619995, "learning_rate": 9.768905068359841e-05, "loss": 0.026, "step": 42690 }, { "grad_norm": 0.44595006108283997, "learning_rate": 9.768739415425523e-05, "loss": 0.0229, "step": 42700 }, { "grad_norm": 0.14594466984272003, "learning_rate": 9.768573704546517e-05, "loss": 0.0234, "step": 42710 }, { "grad_norm": 0.30076900124549866, "learning_rate": 9.768407935724842e-05, "loss": 0.0385, "step": 42720 }, { "grad_norm": 0.21260389685630798, "learning_rate": 9.768242108962506e-05, "loss": 0.0251, "step": 42730 }, { "grad_norm": 0.13713890314102173, "learning_rate": 9.768076224261526e-05, "loss": 0.0216, "step": 42740 }, { "grad_norm": 0.26562944054603577, "learning_rate": 9.767910281623919e-05, "loss": 0.0297, "step": 42750 }, { "grad_norm": 0.19992834329605103, "learning_rate": 9.767744281051701e-05, "loss": 0.0249, "step": 42760 }, { "grad_norm": 0.3146915137767792, "learning_rate": 9.767578222546888e-05, "loss": 0.0225, "step": 42770 }, { "grad_norm": 0.43582186102867126, "learning_rate": 9.767412106111496e-05, "loss": 0.0302, "step": 42780 }, { "grad_norm": 0.30907243490219116, "learning_rate": 9.767245931747549e-05, "loss": 0.0259, "step": 42790 }, { "grad_norm": 0.16228905320167542, "learning_rate": 9.76707969945706e-05, "loss": 0.0173, "step": 42800 }, { "grad_norm": 0.15801124274730682, "learning_rate": 9.766913409242053e-05, "loss": 0.0273, "step": 42810 }, { "grad_norm": 0.25429776310920715, "learning_rate": 9.766747061104548e-05, "loss": 0.022, "step": 42820 }, { "grad_norm": 0.2483978122472763, "learning_rate": 9.766580655046566e-05, "loss": 0.0244, "step": 42830 }, { "grad_norm": 0.23268072307109833, "learning_rate": 9.766414191070128e-05, "loss": 0.0344, "step": 42840 }, { "grad_norm": 0.18248224258422852, "learning_rate": 9.766247669177258e-05, "loss": 0.0257, "step": 42850 }, { "grad_norm": 0.264546662569046, "learning_rate": 9.766081089369979e-05, "loss": 0.0206, "step": 42860 }, { "grad_norm": 0.15780766308307648, "learning_rate": 9.765914451650313e-05, "loss": 0.0163, "step": 42870 }, { "grad_norm": 0.19859343767166138, "learning_rate": 9.765747756020287e-05, "loss": 0.0208, "step": 42880 }, { "grad_norm": 0.2583305537700653, "learning_rate": 9.765581002481927e-05, "loss": 0.0198, "step": 42890 }, { "grad_norm": 0.13411685824394226, "learning_rate": 9.765414191037257e-05, "loss": 0.0249, "step": 42900 }, { "grad_norm": 0.5148282647132874, "learning_rate": 9.765247321688307e-05, "loss": 0.0336, "step": 42910 }, { "grad_norm": 0.2661021649837494, "learning_rate": 9.765080394437102e-05, "loss": 0.0269, "step": 42920 }, { "grad_norm": 0.140353724360466, "learning_rate": 9.764913409285672e-05, "loss": 0.0238, "step": 42930 }, { "grad_norm": 0.3382452428340912, "learning_rate": 9.764746366236044e-05, "loss": 0.026, "step": 42940 }, { "grad_norm": 0.20342901349067688, "learning_rate": 9.764579265290251e-05, "loss": 0.0232, "step": 42950 }, { "grad_norm": 0.42393791675567627, "learning_rate": 9.76441210645032e-05, "loss": 0.0225, "step": 42960 }, { "grad_norm": 0.20341388881206512, "learning_rate": 9.764244889718284e-05, "loss": 0.0183, "step": 42970 }, { "grad_norm": 0.3862929940223694, "learning_rate": 9.764077615096173e-05, "loss": 0.0247, "step": 42980 }, { "grad_norm": 0.18402832746505737, "learning_rate": 9.763910282586022e-05, "loss": 0.0261, "step": 42990 }, { "grad_norm": 0.17791013419628143, "learning_rate": 9.763742892189865e-05, "loss": 0.0264, "step": 43000 }, { "grad_norm": 0.3118821978569031, "learning_rate": 9.763575443909732e-05, "loss": 0.0259, "step": 43010 }, { "grad_norm": 0.18205046653747559, "learning_rate": 9.763407937747659e-05, "loss": 0.0181, "step": 43020 }, { "grad_norm": 0.12463074177503586, "learning_rate": 9.763240373705684e-05, "loss": 0.0299, "step": 43030 }, { "grad_norm": 0.2868044376373291, "learning_rate": 9.76307275178584e-05, "loss": 0.0215, "step": 43040 }, { "grad_norm": 0.2769322097301483, "learning_rate": 9.762905071990165e-05, "loss": 0.0224, "step": 43050 }, { "grad_norm": 0.1747397929430008, "learning_rate": 9.762737334320697e-05, "loss": 0.0227, "step": 43060 }, { "grad_norm": 0.17777031660079956, "learning_rate": 9.762569538779474e-05, "loss": 0.0234, "step": 43070 }, { "grad_norm": 0.15465037524700165, "learning_rate": 9.762401685368534e-05, "loss": 0.0294, "step": 43080 }, { "grad_norm": 0.2268814891576767, "learning_rate": 9.762233774089917e-05, "loss": 0.0225, "step": 43090 }, { "grad_norm": 0.19001853466033936, "learning_rate": 9.762065804945663e-05, "loss": 0.0232, "step": 43100 }, { "grad_norm": 0.2236187756061554, "learning_rate": 9.761897777937813e-05, "loss": 0.0193, "step": 43110 }, { "grad_norm": 0.21565018594264984, "learning_rate": 9.761729693068411e-05, "loss": 0.0284, "step": 43120 }, { "grad_norm": 0.13993485271930695, "learning_rate": 9.761561550339496e-05, "loss": 0.0232, "step": 43130 }, { "grad_norm": 0.24963003396987915, "learning_rate": 9.761393349753114e-05, "loss": 0.0222, "step": 43140 }, { "grad_norm": 0.20626461505889893, "learning_rate": 9.761225091311305e-05, "loss": 0.0182, "step": 43150 }, { "grad_norm": 0.20593106746673584, "learning_rate": 9.761056775016117e-05, "loss": 0.0381, "step": 43160 }, { "grad_norm": 0.22585003077983856, "learning_rate": 9.760888400869595e-05, "loss": 0.0317, "step": 43170 }, { "grad_norm": 0.24773013591766357, "learning_rate": 9.760719968873781e-05, "loss": 0.0299, "step": 43180 }, { "grad_norm": 0.18358835577964783, "learning_rate": 9.760551479030728e-05, "loss": 0.0223, "step": 43190 }, { "grad_norm": 0.13191425800323486, "learning_rate": 9.760382931342477e-05, "loss": 0.0221, "step": 43200 }, { "grad_norm": 0.22229178249835968, "learning_rate": 9.76021432581108e-05, "loss": 0.0234, "step": 43210 }, { "grad_norm": 0.1451011598110199, "learning_rate": 9.760045662438585e-05, "loss": 0.0175, "step": 43220 }, { "grad_norm": 0.17187334597110748, "learning_rate": 9.75987694122704e-05, "loss": 0.0259, "step": 43230 }, { "grad_norm": 0.20301884412765503, "learning_rate": 9.759708162178497e-05, "loss": 0.0256, "step": 43240 }, { "grad_norm": 0.16510109603405, "learning_rate": 9.759539325295005e-05, "loss": 0.0275, "step": 43250 }, { "grad_norm": 0.154915452003479, "learning_rate": 9.759370430578618e-05, "loss": 0.0228, "step": 43260 }, { "grad_norm": 0.1520489901304245, "learning_rate": 9.759201478031385e-05, "loss": 0.0191, "step": 43270 }, { "grad_norm": 0.1469689905643463, "learning_rate": 9.75903246765536e-05, "loss": 0.0184, "step": 43280 }, { "grad_norm": 0.1989915668964386, "learning_rate": 9.7588633994526e-05, "loss": 0.0219, "step": 43290 }, { "grad_norm": 0.1850268691778183, "learning_rate": 9.758694273425155e-05, "loss": 0.0198, "step": 43300 }, { "grad_norm": 0.19258733093738556, "learning_rate": 9.758525089575081e-05, "loss": 0.0262, "step": 43310 }, { "grad_norm": 0.41623586416244507, "learning_rate": 9.758355847904435e-05, "loss": 0.026, "step": 43320 }, { "grad_norm": 0.21603119373321533, "learning_rate": 9.758186548415273e-05, "loss": 0.0215, "step": 43330 }, { "grad_norm": 0.25332897901535034, "learning_rate": 9.758017191109652e-05, "loss": 0.0269, "step": 43340 }, { "grad_norm": 0.42182326316833496, "learning_rate": 9.757847775989631e-05, "loss": 0.0246, "step": 43350 }, { "grad_norm": 0.1665167212486267, "learning_rate": 9.757678303057265e-05, "loss": 0.0199, "step": 43360 }, { "grad_norm": 0.1954297423362732, "learning_rate": 9.757508772314616e-05, "loss": 0.0235, "step": 43370 }, { "grad_norm": 0.18543122708797455, "learning_rate": 9.757339183763745e-05, "loss": 0.0267, "step": 43380 }, { "grad_norm": 0.19932769238948822, "learning_rate": 9.757169537406709e-05, "loss": 0.0228, "step": 43390 }, { "grad_norm": 0.15163609385490417, "learning_rate": 9.756999833245575e-05, "loss": 0.0196, "step": 43400 }, { "grad_norm": 0.1904682219028473, "learning_rate": 9.756830071282398e-05, "loss": 0.0222, "step": 43410 }, { "grad_norm": 0.1445799469947815, "learning_rate": 9.756660251519248e-05, "loss": 0.0234, "step": 43420 }, { "grad_norm": 0.14596112072467804, "learning_rate": 9.756490373958183e-05, "loss": 0.0199, "step": 43430 }, { "grad_norm": 0.5205649137496948, "learning_rate": 9.756320438601269e-05, "loss": 0.029, "step": 43440 }, { "grad_norm": 0.12961067259311676, "learning_rate": 9.75615044545057e-05, "loss": 0.0256, "step": 43450 }, { "grad_norm": 0.32232916355133057, "learning_rate": 9.755980394508155e-05, "loss": 0.0263, "step": 43460 }, { "grad_norm": 0.2058486044406891, "learning_rate": 9.755810285776086e-05, "loss": 0.021, "step": 43470 }, { "grad_norm": 0.15767887234687805, "learning_rate": 9.755640119256432e-05, "loss": 0.0248, "step": 43480 }, { "grad_norm": 0.15420648455619812, "learning_rate": 9.755469894951261e-05, "loss": 0.0185, "step": 43490 }, { "grad_norm": 0.20935694873332977, "learning_rate": 9.75529961286264e-05, "loss": 0.0215, "step": 43500 }, { "grad_norm": 0.23165686428546906, "learning_rate": 9.75512927299264e-05, "loss": 0.0283, "step": 43510 }, { "grad_norm": 0.1703108549118042, "learning_rate": 9.754958875343329e-05, "loss": 0.0207, "step": 43520 }, { "grad_norm": 0.19823868572711945, "learning_rate": 9.754788419916779e-05, "loss": 0.0265, "step": 43530 }, { "grad_norm": 0.15950937569141388, "learning_rate": 9.75461790671506e-05, "loss": 0.0199, "step": 43540 }, { "grad_norm": 0.20309269428253174, "learning_rate": 9.754447335740244e-05, "loss": 0.0247, "step": 43550 }, { "grad_norm": 0.18365070223808289, "learning_rate": 9.754276706994406e-05, "loss": 0.0304, "step": 43560 }, { "grad_norm": 0.19389770925045013, "learning_rate": 9.754106020479614e-05, "loss": 0.0304, "step": 43570 }, { "grad_norm": 0.14058297872543335, "learning_rate": 9.753935276197947e-05, "loss": 0.02, "step": 43580 }, { "grad_norm": 0.15125541388988495, "learning_rate": 9.753764474151477e-05, "loss": 0.0185, "step": 43590 }, { "grad_norm": 0.1590174436569214, "learning_rate": 9.753593614342282e-05, "loss": 0.0222, "step": 43600 }, { "grad_norm": 0.11614079773426056, "learning_rate": 9.753422696772435e-05, "loss": 0.024, "step": 43610 }, { "grad_norm": 0.19757500290870667, "learning_rate": 9.753251721444014e-05, "loss": 0.0233, "step": 43620 }, { "grad_norm": 0.24081772565841675, "learning_rate": 9.753080688359099e-05, "loss": 0.0216, "step": 43630 }, { "grad_norm": 0.15701937675476074, "learning_rate": 9.752909597519763e-05, "loss": 0.0191, "step": 43640 }, { "grad_norm": 0.1575314998626709, "learning_rate": 9.75273844892809e-05, "loss": 0.0225, "step": 43650 }, { "grad_norm": 0.12454992532730103, "learning_rate": 9.752567242586157e-05, "loss": 0.0272, "step": 43660 }, { "grad_norm": 0.17690409719944, "learning_rate": 9.752395978496045e-05, "loss": 0.0211, "step": 43670 }, { "grad_norm": 0.1586727648973465, "learning_rate": 9.752224656659833e-05, "loss": 0.0228, "step": 43680 }, { "grad_norm": 0.15533220767974854, "learning_rate": 9.752053277079607e-05, "loss": 0.0231, "step": 43690 }, { "grad_norm": 0.2643463611602783, "learning_rate": 9.751881839757446e-05, "loss": 0.024, "step": 43700 }, { "grad_norm": 0.2160692662000656, "learning_rate": 9.751710344695435e-05, "loss": 0.023, "step": 43710 }, { "grad_norm": 0.3252705931663513, "learning_rate": 9.751538791895656e-05, "loss": 0.022, "step": 43720 }, { "grad_norm": 0.16016580164432526, "learning_rate": 9.751367181360195e-05, "loss": 0.0171, "step": 43730 }, { "grad_norm": 0.20026683807373047, "learning_rate": 9.751195513091136e-05, "loss": 0.0239, "step": 43740 }, { "grad_norm": 0.1406589299440384, "learning_rate": 9.751023787090568e-05, "loss": 0.0242, "step": 43750 }, { "grad_norm": 0.12570807337760925, "learning_rate": 9.750852003360573e-05, "loss": 0.0227, "step": 43760 }, { "grad_norm": 0.3039027452468872, "learning_rate": 9.750680161903241e-05, "loss": 0.0231, "step": 43770 }, { "grad_norm": 0.3347807824611664, "learning_rate": 9.750508262720659e-05, "loss": 0.0213, "step": 43780 }, { "grad_norm": 0.1953873336315155, "learning_rate": 9.750336305814917e-05, "loss": 0.0192, "step": 43790 }, { "grad_norm": 0.21738189458847046, "learning_rate": 9.750164291188104e-05, "loss": 0.0209, "step": 43800 }, { "grad_norm": 0.21164070069789886, "learning_rate": 9.74999221884231e-05, "loss": 0.0278, "step": 43810 }, { "grad_norm": 0.15350155532360077, "learning_rate": 9.749820088779625e-05, "loss": 0.0277, "step": 43820 }, { "grad_norm": 0.1481991559267044, "learning_rate": 9.749647901002142e-05, "loss": 0.0264, "step": 43830 }, { "grad_norm": 0.22316990792751312, "learning_rate": 9.749475655511952e-05, "loss": 0.0229, "step": 43840 }, { "grad_norm": 0.1979137361049652, "learning_rate": 9.74930335231115e-05, "loss": 0.0199, "step": 43850 }, { "grad_norm": 0.23368073999881744, "learning_rate": 9.749130991401827e-05, "loss": 0.0242, "step": 43860 }, { "grad_norm": 0.44399338960647583, "learning_rate": 9.748958572786079e-05, "loss": 0.0311, "step": 43870 }, { "grad_norm": 0.1382211595773697, "learning_rate": 9.748786096466001e-05, "loss": 0.0214, "step": 43880 }, { "grad_norm": 0.2703968584537506, "learning_rate": 9.748613562443687e-05, "loss": 0.0299, "step": 43890 }, { "grad_norm": 0.19947059452533722, "learning_rate": 9.748440970721236e-05, "loss": 0.0341, "step": 43900 }, { "grad_norm": 0.1966526210308075, "learning_rate": 9.748268321300742e-05, "loss": 0.0355, "step": 43910 }, { "grad_norm": 0.19093966484069824, "learning_rate": 9.748095614184308e-05, "loss": 0.0252, "step": 43920 }, { "grad_norm": 0.2629360556602478, "learning_rate": 9.747922849374027e-05, "loss": 0.0308, "step": 43930 }, { "grad_norm": 0.19721652567386627, "learning_rate": 9.747750026872e-05, "loss": 0.0299, "step": 43940 }, { "grad_norm": 0.3106619119644165, "learning_rate": 9.747577146680329e-05, "loss": 0.0301, "step": 43950 }, { "grad_norm": 0.36090531945228577, "learning_rate": 9.747404208801115e-05, "loss": 0.0272, "step": 43960 }, { "grad_norm": 0.16584648191928864, "learning_rate": 9.747231213236454e-05, "loss": 0.0253, "step": 43970 }, { "grad_norm": 0.22707296907901764, "learning_rate": 9.747058159988453e-05, "loss": 0.0272, "step": 43980 }, { "grad_norm": 0.21237295866012573, "learning_rate": 9.746885049059213e-05, "loss": 0.0369, "step": 43990 }, { "grad_norm": 0.278348833322525, "learning_rate": 9.746711880450838e-05, "loss": 0.0309, "step": 44000 }, { "grad_norm": 0.17343156039714813, "learning_rate": 9.746538654165433e-05, "loss": 0.0265, "step": 44010 }, { "grad_norm": 0.17158465087413788, "learning_rate": 9.746365370205101e-05, "loss": 0.0252, "step": 44020 }, { "grad_norm": 0.2183270901441574, "learning_rate": 9.74619202857195e-05, "loss": 0.0228, "step": 44030 }, { "grad_norm": 0.5595855116844177, "learning_rate": 9.746018629268083e-05, "loss": 0.0216, "step": 44040 }, { "grad_norm": 0.28633150458335876, "learning_rate": 9.745845172295609e-05, "loss": 0.022, "step": 44050 }, { "grad_norm": 0.18481981754302979, "learning_rate": 9.745671657656634e-05, "loss": 0.0207, "step": 44060 }, { "grad_norm": 0.13160273432731628, "learning_rate": 9.74549808535327e-05, "loss": 0.0184, "step": 44070 }, { "grad_norm": 0.1533658653497696, "learning_rate": 9.745324455387623e-05, "loss": 0.0233, "step": 44080 }, { "grad_norm": 0.19794471561908722, "learning_rate": 9.745150767761804e-05, "loss": 0.0259, "step": 44090 }, { "grad_norm": 0.17831704020500183, "learning_rate": 9.744977022477921e-05, "loss": 0.0269, "step": 44100 }, { "grad_norm": 0.1637008935213089, "learning_rate": 9.74480321953809e-05, "loss": 0.0183, "step": 44110 }, { "grad_norm": 0.19279783964157104, "learning_rate": 9.744629358944418e-05, "loss": 0.0215, "step": 44120 }, { "grad_norm": 0.13478349149227142, "learning_rate": 9.74445544069902e-05, "loss": 0.0236, "step": 44130 }, { "grad_norm": 0.16068536043167114, "learning_rate": 9.74428146480401e-05, "loss": 0.0222, "step": 44140 }, { "grad_norm": 0.2909959554672241, "learning_rate": 9.7441074312615e-05, "loss": 0.0248, "step": 44150 }, { "grad_norm": 0.20763671398162842, "learning_rate": 9.743933340073606e-05, "loss": 0.0229, "step": 44160 }, { "grad_norm": 0.19053812325000763, "learning_rate": 9.743759191242443e-05, "loss": 0.0173, "step": 44170 }, { "grad_norm": 0.16231349110603333, "learning_rate": 9.743584984770125e-05, "loss": 0.0194, "step": 44180 }, { "grad_norm": 0.1977962851524353, "learning_rate": 9.743410720658773e-05, "loss": 0.0277, "step": 44190 }, { "grad_norm": 0.21500107645988464, "learning_rate": 9.743236398910501e-05, "loss": 0.0301, "step": 44200 }, { "grad_norm": 0.17880839109420776, "learning_rate": 9.74306201952743e-05, "loss": 0.0166, "step": 44210 }, { "grad_norm": 0.18692351877689362, "learning_rate": 9.742887582511675e-05, "loss": 0.015, "step": 44220 }, { "grad_norm": 0.12577159702777863, "learning_rate": 9.742713087865358e-05, "loss": 0.0147, "step": 44230 }, { "grad_norm": 0.16710451245307922, "learning_rate": 9.742538535590601e-05, "loss": 0.0219, "step": 44240 }, { "grad_norm": 0.2270856499671936, "learning_rate": 9.742363925689521e-05, "loss": 0.0228, "step": 44250 }, { "grad_norm": 0.19031286239624023, "learning_rate": 9.742189258164243e-05, "loss": 0.0267, "step": 44260 }, { "grad_norm": 0.5400604605674744, "learning_rate": 9.742014533016888e-05, "loss": 0.0257, "step": 44270 }, { "grad_norm": 0.22265072166919708, "learning_rate": 9.74183975024958e-05, "loss": 0.0251, "step": 44280 }, { "grad_norm": 0.1999969482421875, "learning_rate": 9.741664909864441e-05, "loss": 0.0294, "step": 44290 }, { "grad_norm": 0.15361334383487701, "learning_rate": 9.741490011863597e-05, "loss": 0.0302, "step": 44300 }, { "grad_norm": 0.14020362496376038, "learning_rate": 9.741315056249171e-05, "loss": 0.0208, "step": 44310 }, { "grad_norm": 0.3621240556240082, "learning_rate": 9.741140043023294e-05, "loss": 0.0228, "step": 44320 }, { "grad_norm": 0.23215673863887787, "learning_rate": 9.740964972188086e-05, "loss": 0.0204, "step": 44330 }, { "grad_norm": 0.1254643052816391, "learning_rate": 9.74078984374568e-05, "loss": 0.0231, "step": 44340 }, { "grad_norm": 0.17315727472305298, "learning_rate": 9.740614657698198e-05, "loss": 0.0219, "step": 44350 }, { "grad_norm": 0.1301051676273346, "learning_rate": 9.740439414047773e-05, "loss": 0.0182, "step": 44360 }, { "grad_norm": 0.13966673612594604, "learning_rate": 9.740264112796536e-05, "loss": 0.0218, "step": 44370 }, { "grad_norm": 0.21165518462657928, "learning_rate": 9.740088753946612e-05, "loss": 0.0214, "step": 44380 }, { "grad_norm": 0.23020310699939728, "learning_rate": 9.739913337500134e-05, "loss": 0.0236, "step": 44390 }, { "grad_norm": 0.18518157303333282, "learning_rate": 9.739737863459235e-05, "loss": 0.0241, "step": 44400 }, { "grad_norm": 0.3153916597366333, "learning_rate": 9.739562331826045e-05, "loss": 0.0235, "step": 44410 }, { "grad_norm": 0.18346895277500153, "learning_rate": 9.739386742602698e-05, "loss": 0.0278, "step": 44420 }, { "grad_norm": 0.16091398894786835, "learning_rate": 9.739211095791329e-05, "loss": 0.0219, "step": 44430 }, { "grad_norm": 0.18234458565711975, "learning_rate": 9.739035391394069e-05, "loss": 0.0228, "step": 44440 }, { "grad_norm": 0.22080351412296295, "learning_rate": 9.738859629413055e-05, "loss": 0.0251, "step": 44450 }, { "grad_norm": 0.24285343289375305, "learning_rate": 9.738683809850422e-05, "loss": 0.0239, "step": 44460 }, { "grad_norm": 0.1950758695602417, "learning_rate": 9.738507932708307e-05, "loss": 0.0226, "step": 44470 }, { "grad_norm": 0.17344334721565247, "learning_rate": 9.738331997988845e-05, "loss": 0.0217, "step": 44480 }, { "grad_norm": 0.11377441883087158, "learning_rate": 9.738156005694179e-05, "loss": 0.018, "step": 44490 }, { "grad_norm": 0.131331205368042, "learning_rate": 9.737979955826442e-05, "loss": 0.0213, "step": 44500 }, { "grad_norm": 0.11593620479106903, "learning_rate": 9.737803848387775e-05, "loss": 0.0209, "step": 44510 }, { "grad_norm": 0.1170743852853775, "learning_rate": 9.737627683380317e-05, "loss": 0.0241, "step": 44520 }, { "grad_norm": 0.22284217178821564, "learning_rate": 9.73745146080621e-05, "loss": 0.0276, "step": 44530 }, { "grad_norm": 0.18295474350452423, "learning_rate": 9.737275180667597e-05, "loss": 0.0281, "step": 44540 }, { "grad_norm": 0.26284438371658325, "learning_rate": 9.737098842966617e-05, "loss": 0.0285, "step": 44550 }, { "grad_norm": 0.22869884967803955, "learning_rate": 9.736922447705413e-05, "loss": 0.028, "step": 44560 }, { "grad_norm": 0.2493734508752823, "learning_rate": 9.736745994886127e-05, "loss": 0.029, "step": 44570 }, { "grad_norm": 0.14836324751377106, "learning_rate": 9.736569484510908e-05, "loss": 0.0436, "step": 44580 }, { "grad_norm": 0.26622632145881653, "learning_rate": 9.736392916581896e-05, "loss": 0.0218, "step": 44590 }, { "grad_norm": 0.2005293369293213, "learning_rate": 9.736216291101239e-05, "loss": 0.0216, "step": 44600 }, { "grad_norm": 0.19595777988433838, "learning_rate": 9.736039608071081e-05, "loss": 0.0242, "step": 44610 }, { "grad_norm": 0.15987391769886017, "learning_rate": 9.735862867493572e-05, "loss": 0.0206, "step": 44620 }, { "grad_norm": 0.28811562061309814, "learning_rate": 9.735686069370857e-05, "loss": 0.0237, "step": 44630 }, { "grad_norm": 0.19153033196926117, "learning_rate": 9.735509213705086e-05, "loss": 0.0312, "step": 44640 }, { "grad_norm": 0.4743078947067261, "learning_rate": 9.735332300498405e-05, "loss": 0.0313, "step": 44650 }, { "grad_norm": 0.21918202936649323, "learning_rate": 9.735155329752966e-05, "loss": 0.0271, "step": 44660 }, { "grad_norm": 0.25687864422798157, "learning_rate": 9.734978301470919e-05, "loss": 0.0345, "step": 44670 }, { "grad_norm": 0.20599780976772308, "learning_rate": 9.734801215654417e-05, "loss": 0.0224, "step": 44680 }, { "grad_norm": 0.15698863565921783, "learning_rate": 9.734624072305607e-05, "loss": 0.024, "step": 44690 }, { "grad_norm": 0.26131007075309753, "learning_rate": 9.734446871426645e-05, "loss": 0.0232, "step": 44700 }, { "grad_norm": 0.28071850538253784, "learning_rate": 9.734269613019687e-05, "loss": 0.0275, "step": 44710 }, { "grad_norm": 0.13004006445407867, "learning_rate": 9.734092297086879e-05, "loss": 0.0248, "step": 44720 }, { "grad_norm": 0.16518515348434448, "learning_rate": 9.733914923630381e-05, "loss": 0.0305, "step": 44730 }, { "grad_norm": 0.2799348533153534, "learning_rate": 9.733737492652348e-05, "loss": 0.0324, "step": 44740 }, { "grad_norm": 0.17557376623153687, "learning_rate": 9.733560004154934e-05, "loss": 0.0314, "step": 44750 }, { "grad_norm": 0.18698808550834656, "learning_rate": 9.7333824581403e-05, "loss": 0.0282, "step": 44760 }, { "grad_norm": 0.16857405006885529, "learning_rate": 9.733204854610596e-05, "loss": 0.0295, "step": 44770 }, { "grad_norm": 0.1286671757698059, "learning_rate": 9.733027193567986e-05, "loss": 0.0261, "step": 44780 }, { "grad_norm": 0.13071489334106445, "learning_rate": 9.732849475014625e-05, "loss": 0.0272, "step": 44790 }, { "grad_norm": 0.6004272103309631, "learning_rate": 9.732671698952676e-05, "loss": 0.0239, "step": 44800 }, { "grad_norm": 0.37060031294822693, "learning_rate": 9.732493865384299e-05, "loss": 0.0286, "step": 44810 }, { "grad_norm": 0.1689678132534027, "learning_rate": 9.73231597431165e-05, "loss": 0.0275, "step": 44820 }, { "grad_norm": 0.30965304374694824, "learning_rate": 9.732138025736895e-05, "loss": 0.0287, "step": 44830 }, { "grad_norm": 0.23957636952400208, "learning_rate": 9.731960019662196e-05, "loss": 0.0261, "step": 44840 }, { "grad_norm": 0.14126206934452057, "learning_rate": 9.731781956089714e-05, "loss": 0.0198, "step": 44850 }, { "grad_norm": 0.2674007713794708, "learning_rate": 9.731603835021614e-05, "loss": 0.0234, "step": 44860 }, { "grad_norm": 0.24383163452148438, "learning_rate": 9.73142565646006e-05, "loss": 0.0248, "step": 44870 }, { "grad_norm": 0.21105335652828217, "learning_rate": 9.731247420407219e-05, "loss": 0.0206, "step": 44880 }, { "grad_norm": 0.19709710776805878, "learning_rate": 9.731069126865254e-05, "loss": 0.0255, "step": 44890 }, { "grad_norm": 0.3460436165332794, "learning_rate": 9.73089077583633e-05, "loss": 0.0234, "step": 44900 }, { "grad_norm": 0.23790442943572998, "learning_rate": 9.730712367322618e-05, "loss": 0.0256, "step": 44910 }, { "grad_norm": 0.2646668255329132, "learning_rate": 9.730533901326285e-05, "loss": 0.0296, "step": 44920 }, { "grad_norm": 0.1938677877187729, "learning_rate": 9.730355377849496e-05, "loss": 0.0178, "step": 44930 }, { "grad_norm": 0.1853252351284027, "learning_rate": 9.730176796894425e-05, "loss": 0.0191, "step": 44940 }, { "grad_norm": 0.3966864049434662, "learning_rate": 9.72999815846324e-05, "loss": 0.0237, "step": 44950 }, { "grad_norm": 0.16341839730739594, "learning_rate": 9.72981946255811e-05, "loss": 0.0185, "step": 44960 }, { "grad_norm": 0.16529123485088348, "learning_rate": 9.72964070918121e-05, "loss": 0.0299, "step": 44970 }, { "grad_norm": 0.41830313205718994, "learning_rate": 9.729461898334709e-05, "loss": 0.0228, "step": 44980 }, { "grad_norm": 0.5974642634391785, "learning_rate": 9.72928303002078e-05, "loss": 0.0236, "step": 44990 }, { "grad_norm": 0.3301568627357483, "learning_rate": 9.729104104241598e-05, "loss": 0.0316, "step": 45000 }, { "grad_norm": 0.1434749960899353, "learning_rate": 9.728925120999334e-05, "loss": 0.02, "step": 45010 }, { "grad_norm": 0.20131535828113556, "learning_rate": 9.728746080296167e-05, "loss": 0.0285, "step": 45020 }, { "grad_norm": 0.2447214275598526, "learning_rate": 9.728566982134271e-05, "loss": 0.0265, "step": 45030 }, { "grad_norm": 0.17009957134723663, "learning_rate": 9.728387826515819e-05, "loss": 0.0206, "step": 45040 }, { "grad_norm": 0.18247805535793304, "learning_rate": 9.728208613442993e-05, "loss": 0.0277, "step": 45050 }, { "grad_norm": 0.20148904621601105, "learning_rate": 9.728029342917967e-05, "loss": 0.0229, "step": 45060 }, { "grad_norm": 0.16266359388828278, "learning_rate": 9.727850014942921e-05, "loss": 0.0248, "step": 45070 }, { "grad_norm": 0.29473215341567993, "learning_rate": 9.727670629520034e-05, "loss": 0.0227, "step": 45080 }, { "grad_norm": 0.20496948063373566, "learning_rate": 9.727491186651484e-05, "loss": 0.0212, "step": 45090 }, { "grad_norm": 0.140741765499115, "learning_rate": 9.727311686339453e-05, "loss": 0.029, "step": 45100 }, { "grad_norm": 0.17062611877918243, "learning_rate": 9.727132128586122e-05, "loss": 0.0266, "step": 45110 }, { "grad_norm": 0.21891528367996216, "learning_rate": 9.726952513393672e-05, "loss": 0.0241, "step": 45120 }, { "grad_norm": 0.3189612925052643, "learning_rate": 9.726772840764287e-05, "loss": 0.0226, "step": 45130 }, { "grad_norm": 0.12032736092805862, "learning_rate": 9.726593110700148e-05, "loss": 0.019, "step": 45140 }, { "grad_norm": 0.18264487385749817, "learning_rate": 9.72641332320344e-05, "loss": 0.0196, "step": 45150 }, { "grad_norm": 0.15916331112384796, "learning_rate": 9.726233478276347e-05, "loss": 0.0219, "step": 45160 }, { "grad_norm": 0.16636447608470917, "learning_rate": 9.726053575921057e-05, "loss": 0.021, "step": 45170 }, { "grad_norm": 0.1406891942024231, "learning_rate": 9.725873616139753e-05, "loss": 0.0195, "step": 45180 }, { "grad_norm": 0.16384932398796082, "learning_rate": 9.725693598934622e-05, "loss": 0.027, "step": 45190 }, { "grad_norm": 0.1269030123949051, "learning_rate": 9.725513524307853e-05, "loss": 0.0306, "step": 45200 }, { "grad_norm": 0.22128351032733917, "learning_rate": 9.725333392261634e-05, "loss": 0.0198, "step": 45210 }, { "grad_norm": 0.6938214302062988, "learning_rate": 9.725153202798149e-05, "loss": 0.0248, "step": 45220 }, { "grad_norm": 0.9573599100112915, "learning_rate": 9.724972955919595e-05, "loss": 0.0217, "step": 45230 }, { "grad_norm": 0.38742485642433167, "learning_rate": 9.724792651628156e-05, "loss": 0.033, "step": 45240 }, { "grad_norm": 0.5257445573806763, "learning_rate": 9.724612289926028e-05, "loss": 0.0266, "step": 45250 }, { "grad_norm": 0.17713342607021332, "learning_rate": 9.724431870815398e-05, "loss": 0.027, "step": 45260 }, { "grad_norm": 0.32261204719543457, "learning_rate": 9.724251394298461e-05, "loss": 0.0264, "step": 45270 }, { "grad_norm": 0.28015029430389404, "learning_rate": 9.724070860377408e-05, "loss": 0.0217, "step": 45280 }, { "grad_norm": 0.13406801223754883, "learning_rate": 9.723890269054434e-05, "loss": 0.0208, "step": 45290 }, { "grad_norm": 0.2001907229423523, "learning_rate": 9.723709620331733e-05, "loss": 0.0253, "step": 45300 }, { "grad_norm": 0.23700973391532898, "learning_rate": 9.723528914211502e-05, "loss": 0.0253, "step": 45310 }, { "grad_norm": 0.1459961235523224, "learning_rate": 9.723348150695934e-05, "loss": 0.0249, "step": 45320 }, { "grad_norm": 0.258344829082489, "learning_rate": 9.723167329787227e-05, "loss": 0.0224, "step": 45330 }, { "grad_norm": 0.2638021409511566, "learning_rate": 9.722986451487574e-05, "loss": 0.0228, "step": 45340 }, { "grad_norm": 0.34432095289230347, "learning_rate": 9.72280551579918e-05, "loss": 0.0293, "step": 45350 }, { "grad_norm": 0.27094200253486633, "learning_rate": 9.72262452272424e-05, "loss": 0.0356, "step": 45360 }, { "grad_norm": 0.18737535178661346, "learning_rate": 9.722443472264951e-05, "loss": 0.0244, "step": 45370 }, { "grad_norm": 0.2907542586326599, "learning_rate": 9.722262364423517e-05, "loss": 0.0242, "step": 45380 }, { "grad_norm": 0.2169872224330902, "learning_rate": 9.722081199202136e-05, "loss": 0.0261, "step": 45390 }, { "grad_norm": 0.18991327285766602, "learning_rate": 9.72189997660301e-05, "loss": 0.0197, "step": 45400 }, { "grad_norm": 0.3283742368221283, "learning_rate": 9.721718696628341e-05, "loss": 0.0242, "step": 45410 }, { "grad_norm": 0.13373735547065735, "learning_rate": 9.721537359280331e-05, "loss": 0.0264, "step": 45420 }, { "grad_norm": 0.2923954129219055, "learning_rate": 9.721355964561185e-05, "loss": 0.0229, "step": 45430 }, { "grad_norm": 0.2078963667154312, "learning_rate": 9.721174512473107e-05, "loss": 0.023, "step": 45440 }, { "grad_norm": 0.203223317861557, "learning_rate": 9.7209930030183e-05, "loss": 0.0225, "step": 45450 }, { "grad_norm": 0.21231915056705475, "learning_rate": 9.720811436198971e-05, "loss": 0.031, "step": 45460 }, { "grad_norm": 0.19216611981391907, "learning_rate": 9.720629812017327e-05, "loss": 0.0297, "step": 45470 }, { "grad_norm": 0.12009035795927048, "learning_rate": 9.720448130475572e-05, "loss": 0.0244, "step": 45480 }, { "grad_norm": 0.2175753265619278, "learning_rate": 9.720266391575916e-05, "loss": 0.0253, "step": 45490 }, { "grad_norm": 0.1978592723608017, "learning_rate": 9.720084595320567e-05, "loss": 0.0249, "step": 45500 }, { "grad_norm": 0.18588824570178986, "learning_rate": 9.719902741711735e-05, "loss": 0.0253, "step": 45510 }, { "grad_norm": 0.5525831580162048, "learning_rate": 9.719720830751627e-05, "loss": 0.0298, "step": 45520 }, { "grad_norm": 0.1562202423810959, "learning_rate": 9.719538862442455e-05, "loss": 0.0246, "step": 45530 }, { "grad_norm": 0.20517681539058685, "learning_rate": 9.719356836786429e-05, "loss": 0.0236, "step": 45540 }, { "grad_norm": 0.1503298282623291, "learning_rate": 9.719174753785761e-05, "loss": 0.0216, "step": 45550 }, { "grad_norm": 0.21795156598091125, "learning_rate": 9.718992613442666e-05, "loss": 0.0276, "step": 45560 }, { "grad_norm": 0.13756772875785828, "learning_rate": 9.718810415759354e-05, "loss": 0.0284, "step": 45570 }, { "grad_norm": 0.2528526186943054, "learning_rate": 9.718628160738041e-05, "loss": 0.0258, "step": 45580 }, { "grad_norm": 0.14781434834003448, "learning_rate": 9.71844584838094e-05, "loss": 0.0233, "step": 45590 }, { "grad_norm": 0.15224504470825195, "learning_rate": 9.718263478690268e-05, "loss": 0.022, "step": 45600 }, { "grad_norm": 0.14957991242408752, "learning_rate": 9.718081051668239e-05, "loss": 0.02, "step": 45610 }, { "grad_norm": 0.14006368815898895, "learning_rate": 9.717898567317071e-05, "loss": 0.0263, "step": 45620 }, { "grad_norm": 0.21021419763565063, "learning_rate": 9.71771602563898e-05, "loss": 0.0178, "step": 45630 }, { "grad_norm": 0.26104941964149475, "learning_rate": 9.717533426636187e-05, "loss": 0.0252, "step": 45640 }, { "grad_norm": 0.1652638465166092, "learning_rate": 9.717350770310907e-05, "loss": 0.0194, "step": 45650 }, { "grad_norm": 0.11665856093168259, "learning_rate": 9.717168056665362e-05, "loss": 0.019, "step": 45660 }, { "grad_norm": 0.22139333188533783, "learning_rate": 9.716985285701772e-05, "loss": 0.0268, "step": 45670 }, { "grad_norm": 0.186898335814476, "learning_rate": 9.716802457422356e-05, "loss": 0.0258, "step": 45680 }, { "grad_norm": 0.2603028118610382, "learning_rate": 9.716619571829338e-05, "loss": 0.0217, "step": 45690 }, { "grad_norm": 0.11918987333774567, "learning_rate": 9.716436628924937e-05, "loss": 0.0246, "step": 45700 }, { "grad_norm": 0.18852533400058746, "learning_rate": 9.71625362871138e-05, "loss": 0.029, "step": 45710 }, { "grad_norm": 0.16423076391220093, "learning_rate": 9.716070571190887e-05, "loss": 0.0281, "step": 45720 }, { "grad_norm": 0.23553837835788727, "learning_rate": 9.715887456365684e-05, "loss": 0.0235, "step": 45730 }, { "grad_norm": 0.279990553855896, "learning_rate": 9.715704284237997e-05, "loss": 0.0249, "step": 45740 }, { "grad_norm": 0.1269577592611313, "learning_rate": 9.715521054810051e-05, "loss": 0.0209, "step": 45750 }, { "grad_norm": 0.1729193776845932, "learning_rate": 9.71533776808407e-05, "loss": 0.0227, "step": 45760 }, { "grad_norm": 0.17117124795913696, "learning_rate": 9.715154424062284e-05, "loss": 0.0228, "step": 45770 }, { "grad_norm": 0.1308424174785614, "learning_rate": 9.71497102274692e-05, "loss": 0.0245, "step": 45780 }, { "grad_norm": 0.19954189658164978, "learning_rate": 9.714787564140206e-05, "loss": 0.0225, "step": 45790 }, { "grad_norm": 0.20761363208293915, "learning_rate": 9.714604048244373e-05, "loss": 0.0194, "step": 45800 }, { "grad_norm": 0.2594711482524872, "learning_rate": 9.714420475061648e-05, "loss": 0.0232, "step": 45810 }, { "grad_norm": 0.20753861963748932, "learning_rate": 9.714236844594263e-05, "loss": 0.0258, "step": 45820 }, { "grad_norm": 0.5157989859580994, "learning_rate": 9.71405315684445e-05, "loss": 0.025, "step": 45830 }, { "grad_norm": 0.4077084958553314, "learning_rate": 9.713869411814441e-05, "loss": 0.028, "step": 45840 }, { "grad_norm": 0.2069886177778244, "learning_rate": 9.713685609506467e-05, "loss": 0.0255, "step": 45850 }, { "grad_norm": 0.23580411076545715, "learning_rate": 9.713501749922762e-05, "loss": 0.0238, "step": 45860 }, { "grad_norm": 0.2686854600906372, "learning_rate": 9.713317833065561e-05, "loss": 0.0258, "step": 45870 }, { "grad_norm": 0.1997908502817154, "learning_rate": 9.7131338589371e-05, "loss": 0.0231, "step": 45880 }, { "grad_norm": 0.21201090514659882, "learning_rate": 9.712949827539611e-05, "loss": 0.0259, "step": 45890 }, { "grad_norm": 0.24144089221954346, "learning_rate": 9.712765738875332e-05, "loss": 0.0279, "step": 45900 }, { "grad_norm": 0.20767289400100708, "learning_rate": 9.712581592946499e-05, "loss": 0.0224, "step": 45910 }, { "grad_norm": 0.19918610155582428, "learning_rate": 9.71239738975535e-05, "loss": 0.021, "step": 45920 }, { "grad_norm": 0.14193296432495117, "learning_rate": 9.712213129304124e-05, "loss": 0.0192, "step": 45930 }, { "grad_norm": 0.3597370982170105, "learning_rate": 9.712028811595059e-05, "loss": 0.021, "step": 45940 }, { "grad_norm": 0.2668195962905884, "learning_rate": 9.711844436630397e-05, "loss": 0.0243, "step": 45950 }, { "grad_norm": 0.12343829870223999, "learning_rate": 9.711660004412375e-05, "loss": 0.0237, "step": 45960 }, { "grad_norm": 0.21587851643562317, "learning_rate": 9.711475514943236e-05, "loss": 0.0273, "step": 45970 }, { "grad_norm": 0.17498454451560974, "learning_rate": 9.711290968225219e-05, "loss": 0.0215, "step": 45980 }, { "grad_norm": 0.2546606659889221, "learning_rate": 9.711106364260571e-05, "loss": 0.0377, "step": 45990 }, { "grad_norm": 0.18368618190288544, "learning_rate": 9.710921703051533e-05, "loss": 0.0237, "step": 46000 }, { "grad_norm": 0.16954277455806732, "learning_rate": 9.710736984600346e-05, "loss": 0.022, "step": 46010 }, { "grad_norm": 0.1352665275335312, "learning_rate": 9.710552208909258e-05, "loss": 0.0236, "step": 46020 }, { "grad_norm": 0.22116215527057648, "learning_rate": 9.710367375980513e-05, "loss": 0.0219, "step": 46030 }, { "grad_norm": 0.18128246068954468, "learning_rate": 9.710182485816358e-05, "loss": 0.0209, "step": 46040 }, { "grad_norm": 0.1201145201921463, "learning_rate": 9.709997538419038e-05, "loss": 0.0239, "step": 46050 }, { "grad_norm": 0.1910262256860733, "learning_rate": 9.7098125337908e-05, "loss": 0.0202, "step": 46060 }, { "grad_norm": 0.16829390823841095, "learning_rate": 9.709627471933896e-05, "loss": 0.0305, "step": 46070 }, { "grad_norm": 0.16860446333885193, "learning_rate": 9.709442352850569e-05, "loss": 0.0241, "step": 46080 }, { "grad_norm": 0.462974488735199, "learning_rate": 9.709257176543071e-05, "loss": 0.0316, "step": 46090 }, { "grad_norm": 0.21597778797149658, "learning_rate": 9.709071943013652e-05, "loss": 0.026, "step": 46100 }, { "grad_norm": 0.16440147161483765, "learning_rate": 9.708886652264563e-05, "loss": 0.0299, "step": 46110 }, { "grad_norm": 0.11920679360628128, "learning_rate": 9.708701304298056e-05, "loss": 0.026, "step": 46120 }, { "grad_norm": 0.2194056510925293, "learning_rate": 9.708515899116382e-05, "loss": 0.031, "step": 46130 }, { "grad_norm": 0.14728842675685883, "learning_rate": 9.708330436721794e-05, "loss": 0.0325, "step": 46140 }, { "grad_norm": 0.19069956243038177, "learning_rate": 9.708144917116546e-05, "loss": 0.0229, "step": 46150 }, { "grad_norm": 0.6177862286567688, "learning_rate": 9.707959340302893e-05, "loss": 0.025, "step": 46160 }, { "grad_norm": 0.36197149753570557, "learning_rate": 9.707773706283087e-05, "loss": 0.0279, "step": 46170 }, { "grad_norm": 0.25554201006889343, "learning_rate": 9.707588015059386e-05, "loss": 0.0289, "step": 46180 }, { "grad_norm": 0.25815197825431824, "learning_rate": 9.707402266634047e-05, "loss": 0.0257, "step": 46190 }, { "grad_norm": 0.11082825809717178, "learning_rate": 9.707216461009325e-05, "loss": 0.022, "step": 46200 }, { "grad_norm": 0.2406529039144516, "learning_rate": 9.70703059818748e-05, "loss": 0.0217, "step": 46210 }, { "grad_norm": 0.2217433750629425, "learning_rate": 9.706844678170767e-05, "loss": 0.022, "step": 46220 }, { "grad_norm": 0.19455143809318542, "learning_rate": 9.706658700961448e-05, "loss": 0.0225, "step": 46230 }, { "grad_norm": 0.12953336536884308, "learning_rate": 9.706472666561783e-05, "loss": 0.0213, "step": 46240 }, { "grad_norm": 0.17609043419361115, "learning_rate": 9.70628657497403e-05, "loss": 0.0213, "step": 46250 }, { "grad_norm": 0.18834200501441956, "learning_rate": 9.706100426200452e-05, "loss": 0.0259, "step": 46260 }, { "grad_norm": 0.14537470042705536, "learning_rate": 9.705914220243313e-05, "loss": 0.0208, "step": 46270 }, { "grad_norm": 0.12073009461164474, "learning_rate": 9.705727957104871e-05, "loss": 0.0197, "step": 46280 }, { "grad_norm": 0.11527795344591141, "learning_rate": 9.70554163678739e-05, "loss": 0.0196, "step": 46290 }, { "grad_norm": 0.1806868314743042, "learning_rate": 9.705355259293137e-05, "loss": 0.0213, "step": 46300 }, { "grad_norm": 0.2601151168346405, "learning_rate": 9.705168824624376e-05, "loss": 0.0236, "step": 46310 }, { "grad_norm": 0.15621328353881836, "learning_rate": 9.70498233278337e-05, "loss": 0.0207, "step": 46320 }, { "grad_norm": 0.17337670922279358, "learning_rate": 9.704795783772386e-05, "loss": 0.0233, "step": 46330 }, { "grad_norm": 0.22098247706890106, "learning_rate": 9.704609177593691e-05, "loss": 0.029, "step": 46340 }, { "grad_norm": 0.18032507598400116, "learning_rate": 9.704422514249554e-05, "loss": 0.0189, "step": 46350 }, { "grad_norm": 0.42208051681518555, "learning_rate": 9.70423579374224e-05, "loss": 0.0272, "step": 46360 }, { "grad_norm": 0.12500865757465363, "learning_rate": 9.704049016074022e-05, "loss": 0.0194, "step": 46370 }, { "grad_norm": 0.16347543895244598, "learning_rate": 9.703862181247166e-05, "loss": 0.0202, "step": 46380 }, { "grad_norm": 0.18477533757686615, "learning_rate": 9.703675289263942e-05, "loss": 0.0258, "step": 46390 }, { "grad_norm": 0.15038925409317017, "learning_rate": 9.703488340126623e-05, "loss": 0.0206, "step": 46400 }, { "grad_norm": 0.23208262026309967, "learning_rate": 9.70330133383748e-05, "loss": 0.018, "step": 46410 }, { "grad_norm": 0.27547258138656616, "learning_rate": 9.703114270398785e-05, "loss": 0.0218, "step": 46420 }, { "grad_norm": 0.1995430737733841, "learning_rate": 9.702927149812812e-05, "loss": 0.0291, "step": 46430 }, { "grad_norm": 0.2363438904285431, "learning_rate": 9.702739972081834e-05, "loss": 0.0229, "step": 46440 }, { "grad_norm": 0.305415540933609, "learning_rate": 9.702552737208124e-05, "loss": 0.0256, "step": 46450 }, { "grad_norm": 0.18723136186599731, "learning_rate": 9.702365445193958e-05, "loss": 0.0231, "step": 46460 }, { "grad_norm": 0.1927066296339035, "learning_rate": 9.702178096041612e-05, "loss": 0.0222, "step": 46470 }, { "grad_norm": 0.15125739574432373, "learning_rate": 9.701990689753365e-05, "loss": 0.0213, "step": 46480 }, { "grad_norm": 0.20001009106636047, "learning_rate": 9.701803226331489e-05, "loss": 0.0216, "step": 46490 }, { "grad_norm": 0.11899100244045258, "learning_rate": 9.701615705778267e-05, "loss": 0.0235, "step": 46500 }, { "grad_norm": 0.31326544284820557, "learning_rate": 9.701428128095972e-05, "loss": 0.0266, "step": 46510 }, { "grad_norm": 0.1518823355436325, "learning_rate": 9.701240493286888e-05, "loss": 0.0179, "step": 46520 }, { "grad_norm": 0.19351544976234436, "learning_rate": 9.701052801353294e-05, "loss": 0.0296, "step": 46530 }, { "grad_norm": 0.18977360427379608, "learning_rate": 9.70086505229747e-05, "loss": 0.022, "step": 46540 }, { "grad_norm": 0.20499223470687866, "learning_rate": 9.700677246121695e-05, "loss": 0.0279, "step": 46550 }, { "grad_norm": 0.1399441957473755, "learning_rate": 9.700489382828256e-05, "loss": 0.0237, "step": 46560 }, { "grad_norm": 0.1715782880783081, "learning_rate": 9.70030146241943e-05, "loss": 0.0191, "step": 46570 }, { "grad_norm": 0.1717081367969513, "learning_rate": 9.700113484897506e-05, "loss": 0.022, "step": 46580 }, { "grad_norm": 0.18906354904174805, "learning_rate": 9.699925450264765e-05, "loss": 0.0192, "step": 46590 }, { "grad_norm": 0.1618446409702301, "learning_rate": 9.699737358523492e-05, "loss": 0.0148, "step": 46600 }, { "grad_norm": 0.18646828830242157, "learning_rate": 9.699549209675973e-05, "loss": 0.0212, "step": 46610 }, { "grad_norm": 0.4523940980434418, "learning_rate": 9.699361003724495e-05, "loss": 0.0221, "step": 46620 }, { "grad_norm": 0.2839326858520508, "learning_rate": 9.699172740671342e-05, "loss": 0.0307, "step": 46630 }, { "grad_norm": 0.20237192511558533, "learning_rate": 9.698984420518804e-05, "loss": 0.0203, "step": 46640 }, { "grad_norm": 0.22545315325260162, "learning_rate": 9.69879604326917e-05, "loss": 0.0238, "step": 46650 }, { "grad_norm": 0.18205423653125763, "learning_rate": 9.698607608924726e-05, "loss": 0.0237, "step": 46660 }, { "grad_norm": 0.2364015430212021, "learning_rate": 9.698419117487762e-05, "loss": 0.0249, "step": 46670 }, { "grad_norm": 0.1343216449022293, "learning_rate": 9.698230568960574e-05, "loss": 0.0211, "step": 46680 }, { "grad_norm": 0.1262458860874176, "learning_rate": 9.698041963345446e-05, "loss": 0.0232, "step": 46690 }, { "grad_norm": 0.15084141492843628, "learning_rate": 9.697853300644672e-05, "loss": 0.0236, "step": 46700 }, { "grad_norm": 0.2489035427570343, "learning_rate": 9.697664580860546e-05, "loss": 0.0196, "step": 46710 }, { "grad_norm": 0.11881116032600403, "learning_rate": 9.69747580399536e-05, "loss": 0.0187, "step": 46720 }, { "grad_norm": 0.12398957461118698, "learning_rate": 9.697286970051407e-05, "loss": 0.0229, "step": 46730 }, { "grad_norm": 0.11471900343894958, "learning_rate": 9.697098079030982e-05, "loss": 0.0212, "step": 46740 }, { "grad_norm": 0.7192392945289612, "learning_rate": 9.696909130936382e-05, "loss": 0.0217, "step": 46750 }, { "grad_norm": 0.20219282805919647, "learning_rate": 9.6967201257699e-05, "loss": 0.0195, "step": 46760 }, { "grad_norm": 1.1209934949874878, "learning_rate": 9.696531063533835e-05, "loss": 0.0298, "step": 46770 }, { "grad_norm": 0.21437443792819977, "learning_rate": 9.696341944230484e-05, "loss": 0.0232, "step": 46780 }, { "grad_norm": 0.1542252153158188, "learning_rate": 9.696152767862143e-05, "loss": 0.0199, "step": 46790 }, { "grad_norm": 0.31845325231552124, "learning_rate": 9.695963534431113e-05, "loss": 0.0265, "step": 46800 }, { "grad_norm": 0.1750032752752304, "learning_rate": 9.695774243939691e-05, "loss": 0.0313, "step": 46810 }, { "grad_norm": 0.12921658158302307, "learning_rate": 9.695584896390179e-05, "loss": 0.0221, "step": 46820 }, { "grad_norm": 0.17623358964920044, "learning_rate": 9.695395491784878e-05, "loss": 0.0243, "step": 46830 }, { "grad_norm": 0.18108633160591125, "learning_rate": 9.695206030126087e-05, "loss": 0.0204, "step": 46840 }, { "grad_norm": 0.1318129152059555, "learning_rate": 9.695016511416111e-05, "loss": 0.018, "step": 46850 }, { "grad_norm": 0.11815506964921951, "learning_rate": 9.694826935657252e-05, "loss": 0.0216, "step": 46860 }, { "grad_norm": 0.19764511287212372, "learning_rate": 9.694637302851813e-05, "loss": 0.018, "step": 46870 }, { "grad_norm": 0.3054027855396271, "learning_rate": 9.694447613002096e-05, "loss": 0.0205, "step": 46880 }, { "grad_norm": 0.15446914732456207, "learning_rate": 9.69425786611041e-05, "loss": 0.0201, "step": 46890 }, { "grad_norm": 0.17236942052841187, "learning_rate": 9.694068062179058e-05, "loss": 0.0252, "step": 46900 }, { "grad_norm": 0.20533786714076996, "learning_rate": 9.69387820121035e-05, "loss": 0.0197, "step": 46910 }, { "grad_norm": 0.23400349915027618, "learning_rate": 9.693688283206586e-05, "loss": 0.019, "step": 46920 }, { "grad_norm": 0.12003859877586365, "learning_rate": 9.693498308170079e-05, "loss": 0.0223, "step": 46930 }, { "grad_norm": 0.22724053263664246, "learning_rate": 9.693308276103137e-05, "loss": 0.0262, "step": 46940 }, { "grad_norm": 0.2125234454870224, "learning_rate": 9.693118187008068e-05, "loss": 0.021, "step": 46950 }, { "grad_norm": 0.15399079024791718, "learning_rate": 9.692928040887183e-05, "loss": 0.0207, "step": 46960 }, { "grad_norm": 0.2207496613264084, "learning_rate": 9.692737837742789e-05, "loss": 0.0214, "step": 46970 }, { "grad_norm": 0.18155322968959808, "learning_rate": 9.692547577577202e-05, "loss": 0.0248, "step": 46980 }, { "grad_norm": 0.1708664745092392, "learning_rate": 9.69235726039273e-05, "loss": 0.022, "step": 46990 }, { "grad_norm": 0.18492308259010315, "learning_rate": 9.692166886191686e-05, "loss": 0.0226, "step": 47000 }, { "grad_norm": 0.213248610496521, "learning_rate": 9.691976454976387e-05, "loss": 0.0255, "step": 47010 }, { "grad_norm": 0.15993544459342957, "learning_rate": 9.691785966749143e-05, "loss": 0.0205, "step": 47020 }, { "grad_norm": 0.16748873889446259, "learning_rate": 9.691595421512269e-05, "loss": 0.0233, "step": 47030 }, { "grad_norm": 0.16313831508159637, "learning_rate": 9.691404819268082e-05, "loss": 0.0248, "step": 47040 }, { "grad_norm": 0.14500394463539124, "learning_rate": 9.691214160018895e-05, "loss": 0.0224, "step": 47050 }, { "grad_norm": 0.16441063582897186, "learning_rate": 9.691023443767029e-05, "loss": 0.0197, "step": 47060 }, { "grad_norm": 0.10915835201740265, "learning_rate": 9.690832670514798e-05, "loss": 0.0215, "step": 47070 }, { "grad_norm": 0.2049749195575714, "learning_rate": 9.690641840264522e-05, "loss": 0.0234, "step": 47080 }, { "grad_norm": 0.27416205406188965, "learning_rate": 9.690450953018518e-05, "loss": 0.0196, "step": 47090 }, { "grad_norm": 0.1391197144985199, "learning_rate": 9.690260008779106e-05, "loss": 0.0221, "step": 47100 }, { "grad_norm": 0.16651508212089539, "learning_rate": 9.690069007548608e-05, "loss": 0.0201, "step": 47110 }, { "grad_norm": 0.25063496828079224, "learning_rate": 9.689877949329342e-05, "loss": 0.0186, "step": 47120 }, { "grad_norm": 0.3640158176422119, "learning_rate": 9.689686834123632e-05, "loss": 0.026, "step": 47130 }, { "grad_norm": 0.2815108895301819, "learning_rate": 9.6894956619338e-05, "loss": 0.0262, "step": 47140 }, { "grad_norm": 0.4364365339279175, "learning_rate": 9.689304432762167e-05, "loss": 0.0165, "step": 47150 }, { "grad_norm": 0.2047806680202484, "learning_rate": 9.689113146611057e-05, "loss": 0.0219, "step": 47160 }, { "grad_norm": 0.21106968820095062, "learning_rate": 9.688921803482796e-05, "loss": 0.0215, "step": 47170 }, { "grad_norm": 0.22209204733371735, "learning_rate": 9.688730403379708e-05, "loss": 0.0237, "step": 47180 }, { "grad_norm": 0.17883749306201935, "learning_rate": 9.688538946304118e-05, "loss": 0.0228, "step": 47190 }, { "grad_norm": 0.13092751801013947, "learning_rate": 9.688347432258353e-05, "loss": 0.0219, "step": 47200 }, { "grad_norm": 0.14580821990966797, "learning_rate": 9.68815586124474e-05, "loss": 0.025, "step": 47210 }, { "grad_norm": 0.22203093767166138, "learning_rate": 9.687964233265608e-05, "loss": 0.0262, "step": 47220 }, { "grad_norm": 0.18475402891635895, "learning_rate": 9.687772548323284e-05, "loss": 0.0239, "step": 47230 }, { "grad_norm": 0.2131156474351883, "learning_rate": 9.687580806420098e-05, "loss": 0.0259, "step": 47240 }, { "grad_norm": 0.17060324549674988, "learning_rate": 9.687389007558379e-05, "loss": 0.0237, "step": 47250 }, { "grad_norm": 0.2682816982269287, "learning_rate": 9.687197151740457e-05, "loss": 0.0284, "step": 47260 }, { "grad_norm": 0.1510842740535736, "learning_rate": 9.687005238968665e-05, "loss": 0.0225, "step": 47270 }, { "grad_norm": 0.2591942846775055, "learning_rate": 9.686813269245333e-05, "loss": 0.0196, "step": 47280 }, { "grad_norm": 0.29480838775634766, "learning_rate": 9.686621242572797e-05, "loss": 0.0259, "step": 47290 }, { "grad_norm": 0.1636887937784195, "learning_rate": 9.686429158953388e-05, "loss": 0.0205, "step": 47300 }, { "grad_norm": 0.20974884927272797, "learning_rate": 9.686237018389438e-05, "loss": 0.0221, "step": 47310 }, { "grad_norm": 0.16723692417144775, "learning_rate": 9.686044820883285e-05, "loss": 0.0181, "step": 47320 }, { "grad_norm": 0.32593995332717896, "learning_rate": 9.685852566437261e-05, "loss": 0.0241, "step": 47330 }, { "grad_norm": 0.2542063295841217, "learning_rate": 9.685660255053706e-05, "loss": 0.0167, "step": 47340 }, { "grad_norm": 0.1979479342699051, "learning_rate": 9.685467886734956e-05, "loss": 0.0231, "step": 47350 }, { "grad_norm": 0.2192223072052002, "learning_rate": 9.685275461483346e-05, "loss": 0.0261, "step": 47360 }, { "grad_norm": 0.19728706777095795, "learning_rate": 9.685082979301214e-05, "loss": 0.0311, "step": 47370 }, { "grad_norm": 0.930519700050354, "learning_rate": 9.684890440190903e-05, "loss": 0.0271, "step": 47380 }, { "grad_norm": 0.47734934091567993, "learning_rate": 9.684697844154747e-05, "loss": 0.0225, "step": 47390 }, { "grad_norm": 0.1849936693906784, "learning_rate": 9.684505191195092e-05, "loss": 0.0217, "step": 47400 }, { "grad_norm": 0.25158873200416565, "learning_rate": 9.684312481314275e-05, "loss": 0.0189, "step": 47410 }, { "grad_norm": 0.209320530295372, "learning_rate": 9.684119714514638e-05, "loss": 0.032, "step": 47420 }, { "grad_norm": 0.11541245877742767, "learning_rate": 9.683926890798525e-05, "loss": 0.0218, "step": 47430 }, { "grad_norm": 0.3497641682624817, "learning_rate": 9.683734010168276e-05, "loss": 0.0322, "step": 47440 }, { "grad_norm": 0.2539689242839813, "learning_rate": 9.683541072626239e-05, "loss": 0.0241, "step": 47450 }, { "grad_norm": 0.14398528635501862, "learning_rate": 9.683348078174754e-05, "loss": 0.0267, "step": 47460 }, { "grad_norm": 0.2044420838356018, "learning_rate": 9.683155026816168e-05, "loss": 0.027, "step": 47470 }, { "grad_norm": 0.2622843086719513, "learning_rate": 9.682961918552828e-05, "loss": 0.0219, "step": 47480 }, { "grad_norm": 0.1190154030919075, "learning_rate": 9.682768753387079e-05, "loss": 0.02, "step": 47490 }, { "grad_norm": 0.19325146079063416, "learning_rate": 9.682575531321269e-05, "loss": 0.0196, "step": 47500 }, { "grad_norm": 0.1638941764831543, "learning_rate": 9.682382252357745e-05, "loss": 0.0212, "step": 47510 }, { "grad_norm": 0.20980007946491241, "learning_rate": 9.682188916498855e-05, "loss": 0.0206, "step": 47520 }, { "grad_norm": 0.29737406969070435, "learning_rate": 9.681995523746948e-05, "loss": 0.021, "step": 47530 }, { "grad_norm": 0.5044664144515991, "learning_rate": 9.681802074104376e-05, "loss": 0.0192, "step": 47540 }, { "grad_norm": 0.14337176084518433, "learning_rate": 9.681608567573487e-05, "loss": 0.025, "step": 47550 }, { "grad_norm": 0.2545539140701294, "learning_rate": 9.681415004156638e-05, "loss": 0.0256, "step": 47560 }, { "grad_norm": 0.10550950467586517, "learning_rate": 9.681221383856173e-05, "loss": 0.0211, "step": 47570 }, { "grad_norm": 0.3156871199607849, "learning_rate": 9.68102770667445e-05, "loss": 0.0315, "step": 47580 }, { "grad_norm": 0.24541029334068298, "learning_rate": 9.680833972613818e-05, "loss": 0.0287, "step": 47590 }, { "grad_norm": 0.274027019739151, "learning_rate": 9.680640181676636e-05, "loss": 0.038, "step": 47600 }, { "grad_norm": 0.1967511624097824, "learning_rate": 9.680446333865258e-05, "loss": 0.0237, "step": 47610 }, { "grad_norm": 0.1748926341533661, "learning_rate": 9.680252429182035e-05, "loss": 0.0215, "step": 47620 }, { "grad_norm": 0.16809217631816864, "learning_rate": 9.680058467629328e-05, "loss": 0.0255, "step": 47630 }, { "grad_norm": 0.2553067207336426, "learning_rate": 9.679864449209491e-05, "loss": 0.0221, "step": 47640 }, { "grad_norm": 0.14264638721942902, "learning_rate": 9.679670373924883e-05, "loss": 0.02, "step": 47650 }, { "grad_norm": 0.2533211410045624, "learning_rate": 9.67947624177786e-05, "loss": 0.0209, "step": 47660 }, { "grad_norm": 0.15773624181747437, "learning_rate": 9.679282052770785e-05, "loss": 0.0217, "step": 47670 }, { "grad_norm": 0.19775797426700592, "learning_rate": 9.679087806906014e-05, "loss": 0.0186, "step": 47680 }, { "grad_norm": 0.12040494382381439, "learning_rate": 9.678893504185908e-05, "loss": 0.0179, "step": 47690 }, { "grad_norm": 0.2945191562175751, "learning_rate": 9.678699144612828e-05, "loss": 0.0246, "step": 47700 }, { "grad_norm": 0.1700924187898636, "learning_rate": 9.678504728189138e-05, "loss": 0.0247, "step": 47710 }, { "grad_norm": 0.2508852779865265, "learning_rate": 9.678310254917197e-05, "loss": 0.0344, "step": 47720 }, { "grad_norm": 0.2224244922399521, "learning_rate": 9.678115724799369e-05, "loss": 0.0247, "step": 47730 }, { "grad_norm": 0.18506446480751038, "learning_rate": 9.677921137838018e-05, "loss": 0.0239, "step": 47740 }, { "grad_norm": 0.1288209855556488, "learning_rate": 9.677726494035508e-05, "loss": 0.019, "step": 47750 }, { "grad_norm": 0.16182437539100647, "learning_rate": 9.677531793394205e-05, "loss": 0.0235, "step": 47760 }, { "grad_norm": 0.16141381859779358, "learning_rate": 9.677337035916477e-05, "loss": 0.0217, "step": 47770 }, { "grad_norm": 0.2610599100589752, "learning_rate": 9.677142221604685e-05, "loss": 0.0265, "step": 47780 }, { "grad_norm": 0.2428954392671585, "learning_rate": 9.676947350461199e-05, "loss": 0.0276, "step": 47790 }, { "grad_norm": 0.12710322439670563, "learning_rate": 9.676752422488388e-05, "loss": 0.0224, "step": 47800 }, { "grad_norm": 0.21181683242321014, "learning_rate": 9.67655743768862e-05, "loss": 0.0261, "step": 47810 }, { "grad_norm": 1.222923755645752, "learning_rate": 9.676362396064263e-05, "loss": 0.028, "step": 47820 }, { "grad_norm": 0.19481439888477325, "learning_rate": 9.676167297617686e-05, "loss": 0.0277, "step": 47830 }, { "grad_norm": 0.2126171737909317, "learning_rate": 9.675972142351264e-05, "loss": 0.0236, "step": 47840 }, { "grad_norm": 0.28796789050102234, "learning_rate": 9.675776930267364e-05, "loss": 0.042, "step": 47850 }, { "grad_norm": 0.17017912864685059, "learning_rate": 9.67558166136836e-05, "loss": 0.0195, "step": 47860 }, { "grad_norm": 0.3443196415901184, "learning_rate": 9.675386335656624e-05, "loss": 0.0207, "step": 47870 }, { "grad_norm": 0.13404789566993713, "learning_rate": 9.67519095313453e-05, "loss": 0.0277, "step": 47880 }, { "grad_norm": 0.17375442385673523, "learning_rate": 9.674995513804452e-05, "loss": 0.0244, "step": 47890 }, { "grad_norm": 0.15364786982536316, "learning_rate": 9.674800017668765e-05, "loss": 0.0294, "step": 47900 }, { "grad_norm": 0.27125582098960876, "learning_rate": 9.674604464729844e-05, "loss": 0.0235, "step": 47910 }, { "grad_norm": 0.3583747446537018, "learning_rate": 9.674408854990064e-05, "loss": 0.0239, "step": 47920 }, { "grad_norm": 0.17502501606941223, "learning_rate": 9.674213188451804e-05, "loss": 0.0229, "step": 47930 }, { "grad_norm": 0.2442699372768402, "learning_rate": 9.67401746511744e-05, "loss": 0.0236, "step": 47940 }, { "grad_norm": 0.22353503108024597, "learning_rate": 9.673821684989352e-05, "loss": 0.0263, "step": 47950 }, { "grad_norm": 0.16423001885414124, "learning_rate": 9.673625848069916e-05, "loss": 0.0292, "step": 47960 }, { "grad_norm": 0.14740177989006042, "learning_rate": 9.673429954361514e-05, "loss": 0.019, "step": 47970 }, { "grad_norm": 0.23910079896450043, "learning_rate": 9.673234003866526e-05, "loss": 0.023, "step": 47980 }, { "grad_norm": 0.2621716856956482, "learning_rate": 9.673037996587333e-05, "loss": 0.0252, "step": 47990 }, { "grad_norm": 0.14542730152606964, "learning_rate": 9.672841932526315e-05, "loss": 0.0159, "step": 48000 }, { "grad_norm": 0.11816121637821198, "learning_rate": 9.672645811685857e-05, "loss": 0.0214, "step": 48010 }, { "grad_norm": 0.21334294974803925, "learning_rate": 9.672449634068339e-05, "loss": 0.0286, "step": 48020 }, { "grad_norm": 0.27380767464637756, "learning_rate": 9.672253399676148e-05, "loss": 0.0238, "step": 48030 }, { "grad_norm": 0.252300888299942, "learning_rate": 9.672057108511665e-05, "loss": 0.022, "step": 48040 }, { "grad_norm": 0.16435667872428894, "learning_rate": 9.671860760577279e-05, "loss": 0.0227, "step": 48050 }, { "grad_norm": 0.33187335729599, "learning_rate": 9.671664355875373e-05, "loss": 0.0297, "step": 48060 }, { "grad_norm": 0.1986348181962967, "learning_rate": 9.671467894408335e-05, "loss": 0.0327, "step": 48070 }, { "grad_norm": 0.26390981674194336, "learning_rate": 9.671271376178551e-05, "loss": 0.0286, "step": 48080 }, { "grad_norm": 0.26517191529273987, "learning_rate": 9.671074801188408e-05, "loss": 0.0258, "step": 48090 }, { "grad_norm": 0.1702200174331665, "learning_rate": 9.670878169440297e-05, "loss": 0.0232, "step": 48100 }, { "grad_norm": 0.15918965637683868, "learning_rate": 9.670681480936606e-05, "loss": 0.0229, "step": 48110 }, { "grad_norm": 0.21467402577400208, "learning_rate": 9.670484735679725e-05, "loss": 0.0324, "step": 48120 }, { "grad_norm": 0.21548987925052643, "learning_rate": 9.670287933672046e-05, "loss": 0.0245, "step": 48130 }, { "grad_norm": 0.16264984011650085, "learning_rate": 9.670091074915959e-05, "loss": 0.0293, "step": 48140 }, { "grad_norm": 0.6368612051010132, "learning_rate": 9.669894159413854e-05, "loss": 0.0241, "step": 48150 }, { "grad_norm": 0.16433005034923553, "learning_rate": 9.669697187168127e-05, "loss": 0.0229, "step": 48160 }, { "grad_norm": 0.25497591495513916, "learning_rate": 9.669500158181172e-05, "loss": 0.0244, "step": 48170 }, { "grad_norm": 0.22200153768062592, "learning_rate": 9.66930307245538e-05, "loss": 0.0221, "step": 48180 }, { "grad_norm": 0.14658041298389435, "learning_rate": 9.669105929993147e-05, "loss": 0.0222, "step": 48190 }, { "grad_norm": 0.20280183851718903, "learning_rate": 9.668908730796868e-05, "loss": 0.0216, "step": 48200 }, { "grad_norm": 0.18243007361888885, "learning_rate": 9.668711474868941e-05, "loss": 0.0255, "step": 48210 }, { "grad_norm": 0.18032154440879822, "learning_rate": 9.668514162211761e-05, "loss": 0.0298, "step": 48220 }, { "grad_norm": 0.17964698374271393, "learning_rate": 9.668316792827725e-05, "loss": 0.0222, "step": 48230 }, { "grad_norm": 0.29285427927970886, "learning_rate": 9.668119366719235e-05, "loss": 0.0264, "step": 48240 }, { "grad_norm": 0.1848965883255005, "learning_rate": 9.667921883888685e-05, "loss": 0.0216, "step": 48250 }, { "grad_norm": 0.17625507712364197, "learning_rate": 9.667724344338478e-05, "loss": 0.0194, "step": 48260 }, { "grad_norm": 0.14870668947696686, "learning_rate": 9.667526748071011e-05, "loss": 0.0196, "step": 48270 }, { "grad_norm": 0.2545824944972992, "learning_rate": 9.66732909508869e-05, "loss": 0.024, "step": 48280 }, { "grad_norm": 0.1684781014919281, "learning_rate": 9.667131385393913e-05, "loss": 0.0182, "step": 48290 }, { "grad_norm": 0.1690780222415924, "learning_rate": 9.666933618989082e-05, "loss": 0.0204, "step": 48300 }, { "grad_norm": 0.21652065217494965, "learning_rate": 9.666735795876602e-05, "loss": 0.0199, "step": 48310 }, { "grad_norm": 0.12544316053390503, "learning_rate": 9.666537916058876e-05, "loss": 0.0216, "step": 48320 }, { "grad_norm": 0.30799734592437744, "learning_rate": 9.666339979538308e-05, "loss": 0.0216, "step": 48330 }, { "grad_norm": 0.19243845343589783, "learning_rate": 9.666141986317305e-05, "loss": 0.0207, "step": 48340 }, { "grad_norm": 0.20615480840206146, "learning_rate": 9.66594393639827e-05, "loss": 0.0245, "step": 48350 }, { "grad_norm": 0.12133743613958359, "learning_rate": 9.66574582978361e-05, "loss": 0.018, "step": 48360 }, { "grad_norm": 0.12967248260974884, "learning_rate": 9.665547666475734e-05, "loss": 0.0227, "step": 48370 }, { "grad_norm": 0.11818668246269226, "learning_rate": 9.66534944647705e-05, "loss": 0.0309, "step": 48380 }, { "grad_norm": 0.15278424322605133, "learning_rate": 9.665151169789964e-05, "loss": 0.0263, "step": 48390 }, { "grad_norm": 0.11473697423934937, "learning_rate": 9.664952836416887e-05, "loss": 0.0217, "step": 48400 }, { "grad_norm": 0.14562197029590607, "learning_rate": 9.664754446360227e-05, "loss": 0.0174, "step": 48410 }, { "grad_norm": 0.11301328241825104, "learning_rate": 9.664555999622397e-05, "loss": 0.0194, "step": 48420 }, { "grad_norm": 0.17945514619350433, "learning_rate": 9.66435749620581e-05, "loss": 0.0256, "step": 48430 }, { "grad_norm": 0.1796969324350357, "learning_rate": 9.664158936112873e-05, "loss": 0.0281, "step": 48440 }, { "grad_norm": 0.19695881009101868, "learning_rate": 9.663960319346002e-05, "loss": 0.0194, "step": 48450 }, { "grad_norm": 0.23140013217926025, "learning_rate": 9.663761645907609e-05, "loss": 0.025, "step": 48460 }, { "grad_norm": 0.24746911227703094, "learning_rate": 9.66356291580011e-05, "loss": 0.0211, "step": 48470 }, { "grad_norm": 0.25120559334754944, "learning_rate": 9.663364129025919e-05, "loss": 0.0261, "step": 48480 }, { "grad_norm": 0.17945541441440582, "learning_rate": 9.663165285587448e-05, "loss": 0.0188, "step": 48490 }, { "grad_norm": 0.17505981028079987, "learning_rate": 9.662966385487118e-05, "loss": 0.0224, "step": 48500 }, { "grad_norm": 0.13871978223323822, "learning_rate": 9.662767428727343e-05, "loss": 0.0186, "step": 48510 }, { "grad_norm": 0.2037312090396881, "learning_rate": 9.662568415310543e-05, "loss": 0.0207, "step": 48520 }, { "grad_norm": 0.1675492376089096, "learning_rate": 9.662369345239132e-05, "loss": 0.0187, "step": 48530 }, { "grad_norm": 0.10611089318990707, "learning_rate": 9.662170218515533e-05, "loss": 0.0199, "step": 48540 }, { "grad_norm": 0.18058234453201294, "learning_rate": 9.661971035142164e-05, "loss": 0.0231, "step": 48550 }, { "grad_norm": 0.0945432186126709, "learning_rate": 9.661771795121444e-05, "loss": 0.0211, "step": 48560 }, { "grad_norm": 0.20017682015895844, "learning_rate": 9.661572498455798e-05, "loss": 0.0256, "step": 48570 }, { "grad_norm": 0.18965116143226624, "learning_rate": 9.661373145147643e-05, "loss": 0.022, "step": 48580 }, { "grad_norm": 0.1655527651309967, "learning_rate": 9.661173735199402e-05, "loss": 0.0198, "step": 48590 }, { "grad_norm": 0.24387435615062714, "learning_rate": 9.660974268613501e-05, "loss": 0.0195, "step": 48600 }, { "grad_norm": 0.159906268119812, "learning_rate": 9.660774745392363e-05, "loss": 0.0461, "step": 48610 }, { "grad_norm": 0.28673821687698364, "learning_rate": 9.66057516553841e-05, "loss": 0.022, "step": 48620 }, { "grad_norm": 0.19147394597530365, "learning_rate": 9.660375529054068e-05, "loss": 0.0195, "step": 48630 }, { "grad_norm": 0.2045178860425949, "learning_rate": 9.660175835941763e-05, "loss": 0.0219, "step": 48640 }, { "grad_norm": 0.1721971184015274, "learning_rate": 9.659976086203923e-05, "loss": 0.0239, "step": 48650 }, { "grad_norm": 0.23713675141334534, "learning_rate": 9.659776279842972e-05, "loss": 0.0244, "step": 48660 }, { "grad_norm": 0.1524079144001007, "learning_rate": 9.659576416861342e-05, "loss": 0.0247, "step": 48670 }, { "grad_norm": 0.15582816302776337, "learning_rate": 9.659376497261457e-05, "loss": 0.0215, "step": 48680 }, { "grad_norm": 0.15175674855709076, "learning_rate": 9.65917652104575e-05, "loss": 0.0221, "step": 48690 }, { "grad_norm": 0.14095857739448547, "learning_rate": 9.658976488216648e-05, "loss": 0.019, "step": 48700 }, { "grad_norm": 0.33610299229621887, "learning_rate": 9.658776398776583e-05, "loss": 0.0286, "step": 48710 }, { "grad_norm": 0.1496945470571518, "learning_rate": 9.658576252727987e-05, "loss": 0.0259, "step": 48720 }, { "grad_norm": 0.25515446066856384, "learning_rate": 9.65837605007329e-05, "loss": 0.0271, "step": 48730 }, { "grad_norm": 0.15911589562892914, "learning_rate": 9.658175790814927e-05, "loss": 0.0254, "step": 48740 }, { "grad_norm": 0.21764302253723145, "learning_rate": 9.65797547495533e-05, "loss": 0.0187, "step": 48750 }, { "grad_norm": 0.24260905385017395, "learning_rate": 9.657775102496933e-05, "loss": 0.0239, "step": 48760 }, { "grad_norm": 0.16800148785114288, "learning_rate": 9.65757467344217e-05, "loss": 0.0188, "step": 48770 }, { "grad_norm": 0.12130135297775269, "learning_rate": 9.657374187793477e-05, "loss": 0.0224, "step": 48780 }, { "grad_norm": 0.11607987433671951, "learning_rate": 9.65717364555329e-05, "loss": 0.0184, "step": 48790 }, { "grad_norm": 0.18539832532405853, "learning_rate": 9.656973046724048e-05, "loss": 0.018, "step": 48800 }, { "grad_norm": 0.22152814269065857, "learning_rate": 9.656772391308186e-05, "loss": 0.0228, "step": 48810 }, { "grad_norm": 0.22538088262081146, "learning_rate": 9.656571679308141e-05, "loss": 0.0215, "step": 48820 }, { "grad_norm": 0.2169121354818344, "learning_rate": 9.656370910726355e-05, "loss": 0.0311, "step": 48830 }, { "grad_norm": 0.2515248954296112, "learning_rate": 9.656170085565269e-05, "loss": 0.0303, "step": 48840 }, { "grad_norm": 0.16201071441173553, "learning_rate": 9.655969203827316e-05, "loss": 0.0209, "step": 48850 }, { "grad_norm": 0.14379151165485382, "learning_rate": 9.655768265514941e-05, "loss": 0.0204, "step": 48860 }, { "grad_norm": 0.18070903420448303, "learning_rate": 9.655567270630589e-05, "loss": 0.0245, "step": 48870 }, { "grad_norm": 0.11627386510372162, "learning_rate": 9.655366219176696e-05, "loss": 0.0273, "step": 48880 }, { "grad_norm": 0.1227191686630249, "learning_rate": 9.65516511115571e-05, "loss": 0.0223, "step": 48890 }, { "grad_norm": 0.14290352165699005, "learning_rate": 9.654963946570073e-05, "loss": 0.0249, "step": 48900 }, { "grad_norm": 0.16371451318264008, "learning_rate": 9.654762725422229e-05, "loss": 0.0303, "step": 48910 }, { "grad_norm": 0.2096206545829773, "learning_rate": 9.654561447714621e-05, "loss": 0.0208, "step": 48920 }, { "grad_norm": 0.1485656499862671, "learning_rate": 9.654360113449698e-05, "loss": 0.0204, "step": 48930 }, { "grad_norm": 0.3316325843334198, "learning_rate": 9.654158722629906e-05, "loss": 0.0208, "step": 48940 }, { "grad_norm": 0.2683253586292267, "learning_rate": 9.65395727525769e-05, "loss": 0.0221, "step": 48950 }, { "grad_norm": 0.14613573253154755, "learning_rate": 9.653755771335499e-05, "loss": 0.0195, "step": 48960 }, { "grad_norm": 0.14627785980701447, "learning_rate": 9.653554210865783e-05, "loss": 0.0187, "step": 48970 }, { "grad_norm": 0.17904408276081085, "learning_rate": 9.653352593850988e-05, "loss": 0.0255, "step": 48980 }, { "grad_norm": 0.16522718966007233, "learning_rate": 9.653150920293568e-05, "loss": 0.0201, "step": 48990 }, { "grad_norm": 0.2525902986526489, "learning_rate": 9.652949190195967e-05, "loss": 0.0206, "step": 49000 }, { "grad_norm": 0.25058916211128235, "learning_rate": 9.652747403560644e-05, "loss": 0.0284, "step": 49010 }, { "grad_norm": 0.16597004234790802, "learning_rate": 9.652545560390045e-05, "loss": 0.0212, "step": 49020 }, { "grad_norm": 0.24705520272254944, "learning_rate": 9.652343660686626e-05, "loss": 0.0298, "step": 49030 }, { "grad_norm": 0.17701955139636993, "learning_rate": 9.652141704452837e-05, "loss": 0.025, "step": 49040 }, { "grad_norm": 0.16741959750652313, "learning_rate": 9.651939691691137e-05, "loss": 0.0233, "step": 49050 }, { "grad_norm": 0.14212502539157867, "learning_rate": 9.651737622403974e-05, "loss": 0.0206, "step": 49060 }, { "grad_norm": 0.22429299354553223, "learning_rate": 9.651535496593809e-05, "loss": 0.0266, "step": 49070 }, { "grad_norm": 0.19496704638004303, "learning_rate": 9.651333314263096e-05, "loss": 0.0208, "step": 49080 }, { "grad_norm": 0.12966108322143555, "learning_rate": 9.651131075414292e-05, "loss": 0.0249, "step": 49090 }, { "grad_norm": 0.11775051057338715, "learning_rate": 9.650928780049853e-05, "loss": 0.0203, "step": 49100 }, { "grad_norm": 0.12156084924936295, "learning_rate": 9.650726428172239e-05, "loss": 0.0258, "step": 49110 }, { "grad_norm": 0.17664286494255066, "learning_rate": 9.650524019783906e-05, "loss": 0.0244, "step": 49120 }, { "grad_norm": 0.132612407207489, "learning_rate": 9.650321554887318e-05, "loss": 0.0216, "step": 49130 }, { "grad_norm": 0.1578652262687683, "learning_rate": 9.65011903348493e-05, "loss": 0.0193, "step": 49140 }, { "grad_norm": 0.14811652898788452, "learning_rate": 9.649916455579206e-05, "loss": 0.0185, "step": 49150 }, { "grad_norm": 0.1808883249759674, "learning_rate": 9.649713821172608e-05, "loss": 0.0184, "step": 49160 }, { "grad_norm": 0.27010104060173035, "learning_rate": 9.649511130267597e-05, "loss": 0.0222, "step": 49170 }, { "grad_norm": 0.1808987259864807, "learning_rate": 9.649308382866635e-05, "loss": 0.0236, "step": 49180 }, { "grad_norm": 0.1569320112466812, "learning_rate": 9.649105578972186e-05, "loss": 0.0169, "step": 49190 }, { "grad_norm": 0.1491139829158783, "learning_rate": 9.648902718586717e-05, "loss": 0.0205, "step": 49200 }, { "grad_norm": 0.6423057913780212, "learning_rate": 9.648699801712688e-05, "loss": 0.0211, "step": 49210 }, { "grad_norm": 0.19155465066432953, "learning_rate": 9.648496828352569e-05, "loss": 0.0269, "step": 49220 }, { "grad_norm": 0.17304691672325134, "learning_rate": 9.648293798508823e-05, "loss": 0.0164, "step": 49230 }, { "grad_norm": 0.1900937706232071, "learning_rate": 9.648090712183919e-05, "loss": 0.0234, "step": 49240 }, { "grad_norm": 0.14317531883716583, "learning_rate": 9.647887569380326e-05, "loss": 0.0306, "step": 49250 }, { "grad_norm": 0.15004011988639832, "learning_rate": 9.64768437010051e-05, "loss": 0.035, "step": 49260 }, { "grad_norm": 0.239028200507164, "learning_rate": 9.647481114346941e-05, "loss": 0.0265, "step": 49270 }, { "grad_norm": 0.19954814016819, "learning_rate": 9.647277802122087e-05, "loss": 0.0241, "step": 49280 }, { "grad_norm": 0.137613445520401, "learning_rate": 9.647074433428421e-05, "loss": 0.0216, "step": 49290 }, { "grad_norm": 0.18229609727859497, "learning_rate": 9.646871008268412e-05, "loss": 0.023, "step": 49300 }, { "grad_norm": 0.25158125162124634, "learning_rate": 9.646667526644535e-05, "loss": 0.025, "step": 49310 }, { "grad_norm": 0.1460191160440445, "learning_rate": 9.64646398855926e-05, "loss": 0.0188, "step": 49320 }, { "grad_norm": 0.2599422335624695, "learning_rate": 9.646260394015059e-05, "loss": 0.0233, "step": 49330 }, { "grad_norm": 0.19988080859184265, "learning_rate": 9.646056743014407e-05, "loss": 0.0252, "step": 49340 }, { "grad_norm": 0.1341286599636078, "learning_rate": 9.645853035559781e-05, "loss": 0.0188, "step": 49350 }, { "grad_norm": 0.15503087639808655, "learning_rate": 9.645649271653654e-05, "loss": 0.0218, "step": 49360 }, { "grad_norm": 0.10053510218858719, "learning_rate": 9.645445451298501e-05, "loss": 0.019, "step": 49370 }, { "grad_norm": 0.16941498219966888, "learning_rate": 9.6452415744968e-05, "loss": 0.0201, "step": 49380 }, { "grad_norm": 0.5429403185844421, "learning_rate": 9.645037641251029e-05, "loss": 0.0253, "step": 49390 }, { "grad_norm": 0.18723252415657043, "learning_rate": 9.644833651563664e-05, "loss": 0.0254, "step": 49400 }, { "grad_norm": 0.16412915289402008, "learning_rate": 9.644629605437186e-05, "loss": 0.0251, "step": 49410 }, { "grad_norm": 0.09025458246469498, "learning_rate": 9.64442550287407e-05, "loss": 0.0183, "step": 49420 }, { "grad_norm": 0.2810123562812805, "learning_rate": 9.644221343876801e-05, "loss": 0.0206, "step": 49430 }, { "grad_norm": 0.16756975650787354, "learning_rate": 9.644017128447859e-05, "loss": 0.0195, "step": 49440 }, { "grad_norm": 0.17037494480609894, "learning_rate": 9.643812856589723e-05, "loss": 0.0177, "step": 49450 }, { "grad_norm": 0.20716221630573273, "learning_rate": 9.643608528304877e-05, "loss": 0.0231, "step": 49460 }, { "grad_norm": 0.23111870884895325, "learning_rate": 9.643404143595801e-05, "loss": 0.0198, "step": 49470 }, { "grad_norm": 0.10385734587907791, "learning_rate": 9.643199702464982e-05, "loss": 0.0167, "step": 49480 }, { "grad_norm": 0.1432560682296753, "learning_rate": 9.642995204914902e-05, "loss": 0.027, "step": 49490 }, { "grad_norm": 0.22547847032546997, "learning_rate": 9.642790650948048e-05, "loss": 0.0354, "step": 49500 }, { "grad_norm": 0.16606880724430084, "learning_rate": 9.642586040566903e-05, "loss": 0.0241, "step": 49510 }, { "grad_norm": 0.16526305675506592, "learning_rate": 9.642381373773955e-05, "loss": 0.023, "step": 49520 }, { "grad_norm": 0.12792368233203888, "learning_rate": 9.642176650571691e-05, "loss": 0.0232, "step": 49530 }, { "grad_norm": 0.16615146398544312, "learning_rate": 9.641971870962597e-05, "loss": 0.0266, "step": 49540 }, { "grad_norm": 0.17741607129573822, "learning_rate": 9.641767034949161e-05, "loss": 0.0269, "step": 49550 }, { "grad_norm": 0.22345001995563507, "learning_rate": 9.641562142533874e-05, "loss": 0.0184, "step": 49560 }, { "grad_norm": 0.17345431447029114, "learning_rate": 9.641357193719223e-05, "loss": 0.0252, "step": 49570 }, { "grad_norm": 0.1605241447687149, "learning_rate": 9.6411521885077e-05, "loss": 0.0207, "step": 49580 }, { "grad_norm": 0.49986615777015686, "learning_rate": 9.640947126901798e-05, "loss": 0.0186, "step": 49590 }, { "grad_norm": 0.18981996178627014, "learning_rate": 9.640742008904005e-05, "loss": 0.0268, "step": 49600 }, { "grad_norm": 0.32836636900901794, "learning_rate": 9.640536834516815e-05, "loss": 0.0251, "step": 49610 }, { "grad_norm": 0.49431702494621277, "learning_rate": 9.640331603742723e-05, "loss": 0.0268, "step": 49620 }, { "grad_norm": 0.3393435478210449, "learning_rate": 9.640126316584219e-05, "loss": 0.0207, "step": 49630 }, { "grad_norm": 0.2245357781648636, "learning_rate": 9.639920973043801e-05, "loss": 0.0265, "step": 49640 }, { "grad_norm": 0.1724606156349182, "learning_rate": 9.639715573123961e-05, "loss": 0.0249, "step": 49650 }, { "grad_norm": 0.2633366584777832, "learning_rate": 9.639510116827196e-05, "loss": 0.0312, "step": 49660 }, { "grad_norm": 0.23780936002731323, "learning_rate": 9.639304604156003e-05, "loss": 0.0299, "step": 49670 }, { "grad_norm": 0.22076818346977234, "learning_rate": 9.63909903511288e-05, "loss": 0.0263, "step": 49680 }, { "grad_norm": 0.2652040719985962, "learning_rate": 9.638893409700321e-05, "loss": 0.0212, "step": 49690 }, { "grad_norm": 0.1753925085067749, "learning_rate": 9.638687727920828e-05, "loss": 0.0203, "step": 49700 }, { "grad_norm": 0.14669251441955566, "learning_rate": 9.638481989776901e-05, "loss": 0.0248, "step": 49710 }, { "grad_norm": 0.18198241293430328, "learning_rate": 9.638276195271037e-05, "loss": 0.0242, "step": 49720 }, { "grad_norm": 0.1902315765619278, "learning_rate": 9.638070344405738e-05, "loss": 0.0259, "step": 49730 }, { "grad_norm": 0.19612905383110046, "learning_rate": 9.637864437183507e-05, "loss": 0.0222, "step": 49740 }, { "grad_norm": 0.18454469740390778, "learning_rate": 9.637658473606843e-05, "loss": 0.0224, "step": 49750 }, { "grad_norm": 0.15553481876850128, "learning_rate": 9.637452453678251e-05, "loss": 0.0206, "step": 49760 }, { "grad_norm": 0.23424352705478668, "learning_rate": 9.63724637740023e-05, "loss": 0.0234, "step": 49770 }, { "grad_norm": 0.19774316251277924, "learning_rate": 9.637040244775291e-05, "loss": 0.0214, "step": 49780 }, { "grad_norm": 0.17533527314662933, "learning_rate": 9.636834055805933e-05, "loss": 0.025, "step": 49790 }, { "grad_norm": 0.18910503387451172, "learning_rate": 9.636627810494663e-05, "loss": 0.0319, "step": 49800 }, { "grad_norm": 0.16544261574745178, "learning_rate": 9.636421508843988e-05, "loss": 0.0218, "step": 49810 }, { "grad_norm": 0.17333906888961792, "learning_rate": 9.636215150856414e-05, "loss": 0.0307, "step": 49820 }, { "grad_norm": 0.14147600531578064, "learning_rate": 9.636008736534448e-05, "loss": 0.0231, "step": 49830 }, { "grad_norm": 0.28082114458084106, "learning_rate": 9.6358022658806e-05, "loss": 0.0262, "step": 49840 }, { "grad_norm": 0.2818644344806671, "learning_rate": 9.635595738897377e-05, "loss": 0.0201, "step": 49850 }, { "grad_norm": 0.3620566427707672, "learning_rate": 9.635389155587289e-05, "loss": 0.0264, "step": 49860 }, { "grad_norm": 0.1801050454378128, "learning_rate": 9.635182515952847e-05, "loss": 0.0242, "step": 49870 }, { "grad_norm": 0.24350431561470032, "learning_rate": 9.63497581999656e-05, "loss": 0.0263, "step": 49880 }, { "grad_norm": 0.27126431465148926, "learning_rate": 9.634769067720941e-05, "loss": 0.0298, "step": 49890 }, { "grad_norm": 0.17144286632537842, "learning_rate": 9.634562259128502e-05, "loss": 0.022, "step": 49900 }, { "grad_norm": 0.19947458803653717, "learning_rate": 9.634355394221756e-05, "loss": 0.0164, "step": 49910 }, { "grad_norm": 0.14677675068378448, "learning_rate": 9.634148473003216e-05, "loss": 0.0223, "step": 49920 }, { "grad_norm": 0.1713489145040512, "learning_rate": 9.633941495475396e-05, "loss": 0.0194, "step": 49930 }, { "grad_norm": 0.14918017387390137, "learning_rate": 9.633734461640813e-05, "loss": 0.0241, "step": 49940 }, { "grad_norm": 0.20348505675792694, "learning_rate": 9.63352737150198e-05, "loss": 0.0207, "step": 49950 }, { "grad_norm": 0.24415741860866547, "learning_rate": 9.633320225061417e-05, "loss": 0.0196, "step": 49960 }, { "grad_norm": 0.11083124577999115, "learning_rate": 9.633113022321636e-05, "loss": 0.0209, "step": 49970 }, { "grad_norm": 0.698153555393219, "learning_rate": 9.632905763285158e-05, "loss": 0.0224, "step": 49980 }, { "grad_norm": 0.20359623432159424, "learning_rate": 9.632698447954503e-05, "loss": 0.0171, "step": 49990 }, { "grad_norm": 0.14136509597301483, "learning_rate": 9.632491076332185e-05, "loss": 0.0221, "step": 50000 }, { "grad_norm": 0.24557633697986603, "learning_rate": 9.632283648420728e-05, "loss": 0.0293, "step": 50010 }, { "grad_norm": 0.23273824155330658, "learning_rate": 9.632076164222651e-05, "loss": 0.0239, "step": 50020 }, { "grad_norm": 0.17402978241443634, "learning_rate": 9.631868623740475e-05, "loss": 0.0193, "step": 50030 }, { "grad_norm": 0.23738503456115723, "learning_rate": 9.631661026976722e-05, "loss": 0.0304, "step": 50040 }, { "grad_norm": 0.8935754895210266, "learning_rate": 9.631453373933914e-05, "loss": 0.0274, "step": 50050 }, { "grad_norm": 0.5712735652923584, "learning_rate": 9.631245664614576e-05, "loss": 0.0284, "step": 50060 }, { "grad_norm": 0.27340832352638245, "learning_rate": 9.631037899021229e-05, "loss": 0.02, "step": 50070 }, { "grad_norm": 0.21451140940189362, "learning_rate": 9.630830077156401e-05, "loss": 0.0259, "step": 50080 }, { "grad_norm": 0.19348956644535065, "learning_rate": 9.630622199022614e-05, "loss": 0.0267, "step": 50090 }, { "grad_norm": 0.5696625113487244, "learning_rate": 9.630414264622397e-05, "loss": 0.0229, "step": 50100 }, { "grad_norm": 0.16923843324184418, "learning_rate": 9.630206273958272e-05, "loss": 0.03, "step": 50110 }, { "grad_norm": 0.12020424753427505, "learning_rate": 9.62999822703277e-05, "loss": 0.0258, "step": 50120 }, { "grad_norm": 0.18191543221473694, "learning_rate": 9.62979012384842e-05, "loss": 0.0243, "step": 50130 }, { "grad_norm": 0.15356439352035522, "learning_rate": 9.629581964407746e-05, "loss": 0.0203, "step": 50140 }, { "grad_norm": 0.19493094086647034, "learning_rate": 9.629373748713281e-05, "loss": 0.0232, "step": 50150 }, { "grad_norm": 0.32374438643455505, "learning_rate": 9.629165476767555e-05, "loss": 0.0282, "step": 50160 }, { "grad_norm": 1.2233036756515503, "learning_rate": 9.628957148573098e-05, "loss": 0.026, "step": 50170 }, { "grad_norm": 0.1634882390499115, "learning_rate": 9.628748764132441e-05, "loss": 0.0269, "step": 50180 }, { "grad_norm": 0.19720537960529327, "learning_rate": 9.628540323448114e-05, "loss": 0.0263, "step": 50190 }, { "grad_norm": 0.36027610301971436, "learning_rate": 9.628331826522654e-05, "loss": 0.0293, "step": 50200 }, { "grad_norm": 0.1922156661748886, "learning_rate": 9.628123273358594e-05, "loss": 0.0297, "step": 50210 }, { "grad_norm": 0.3253571689128876, "learning_rate": 9.627914663958463e-05, "loss": 0.0259, "step": 50220 }, { "grad_norm": 0.17344863712787628, "learning_rate": 9.627705998324802e-05, "loss": 0.0163, "step": 50230 }, { "grad_norm": 0.10805697739124298, "learning_rate": 9.627497276460142e-05, "loss": 0.0179, "step": 50240 }, { "grad_norm": 0.25145822763442993, "learning_rate": 9.627288498367022e-05, "loss": 0.022, "step": 50250 }, { "grad_norm": 0.24060970544815063, "learning_rate": 9.627079664047979e-05, "loss": 0.0249, "step": 50260 }, { "grad_norm": 0.1928597092628479, "learning_rate": 9.626870773505548e-05, "loss": 0.0262, "step": 50270 }, { "grad_norm": 0.19728805124759674, "learning_rate": 9.62666182674227e-05, "loss": 0.0278, "step": 50280 }, { "grad_norm": 0.20355631411075592, "learning_rate": 9.626452823760681e-05, "loss": 0.0262, "step": 50290 }, { "grad_norm": 0.17656658589839935, "learning_rate": 9.626243764563321e-05, "loss": 0.0294, "step": 50300 }, { "grad_norm": 0.21697674691677094, "learning_rate": 9.626034649152733e-05, "loss": 0.0254, "step": 50310 }, { "grad_norm": 0.2851104438304901, "learning_rate": 9.625825477531456e-05, "loss": 0.0292, "step": 50320 }, { "grad_norm": 0.15828350186347961, "learning_rate": 9.625616249702032e-05, "loss": 0.018, "step": 50330 }, { "grad_norm": 0.23366957902908325, "learning_rate": 9.625406965667003e-05, "loss": 0.0235, "step": 50340 }, { "grad_norm": 0.2611834406852722, "learning_rate": 9.625197625428913e-05, "loss": 0.0195, "step": 50350 }, { "grad_norm": 0.2507435977458954, "learning_rate": 9.624988228990305e-05, "loss": 0.0209, "step": 50360 }, { "grad_norm": 0.9251725673675537, "learning_rate": 9.624778776353722e-05, "loss": 0.0357, "step": 50370 }, { "grad_norm": 0.2243080735206604, "learning_rate": 9.624569267521712e-05, "loss": 0.0257, "step": 50380 }, { "grad_norm": 0.23170264065265656, "learning_rate": 9.624359702496819e-05, "loss": 0.0242, "step": 50390 }, { "grad_norm": 0.19608382880687714, "learning_rate": 9.624150081281589e-05, "loss": 0.0241, "step": 50400 }, { "grad_norm": 0.22916841506958008, "learning_rate": 9.623940403878569e-05, "loss": 0.028, "step": 50410 }, { "grad_norm": 0.19929738342761993, "learning_rate": 9.623730670290307e-05, "loss": 0.0297, "step": 50420 }, { "grad_norm": 0.18701507151126862, "learning_rate": 9.623520880519352e-05, "loss": 0.0246, "step": 50430 }, { "grad_norm": 0.14070633053779602, "learning_rate": 9.623311034568254e-05, "loss": 0.0267, "step": 50440 }, { "grad_norm": 0.22446663677692413, "learning_rate": 9.623101132439561e-05, "loss": 0.0278, "step": 50450 }, { "grad_norm": 0.3598111569881439, "learning_rate": 9.622891174135825e-05, "loss": 0.0225, "step": 50460 }, { "grad_norm": 0.21271881461143494, "learning_rate": 9.622681159659597e-05, "loss": 0.0187, "step": 50470 }, { "grad_norm": 0.13131959736347198, "learning_rate": 9.622471089013427e-05, "loss": 0.0209, "step": 50480 }, { "grad_norm": 0.12616606056690216, "learning_rate": 9.622260962199869e-05, "loss": 0.0258, "step": 50490 }, { "grad_norm": 0.18036194145679474, "learning_rate": 9.622050779221477e-05, "loss": 0.034, "step": 50500 }, { "grad_norm": 0.1746194064617157, "learning_rate": 9.621840540080805e-05, "loss": 0.0244, "step": 50510 }, { "grad_norm": 0.1574813723564148, "learning_rate": 9.621630244780404e-05, "loss": 0.026, "step": 50520 }, { "grad_norm": 0.1423124223947525, "learning_rate": 9.621419893322834e-05, "loss": 0.0211, "step": 50530 }, { "grad_norm": 0.17312076687812805, "learning_rate": 9.621209485710647e-05, "loss": 0.0198, "step": 50540 }, { "grad_norm": 0.1717585325241089, "learning_rate": 9.620999021946404e-05, "loss": 0.0228, "step": 50550 }, { "grad_norm": 0.16940680146217346, "learning_rate": 9.620788502032658e-05, "loss": 0.0237, "step": 50560 }, { "grad_norm": 0.14852184057235718, "learning_rate": 9.62057792597197e-05, "loss": 0.0209, "step": 50570 }, { "grad_norm": 0.15035013854503632, "learning_rate": 9.620367293766898e-05, "loss": 0.02, "step": 50580 }, { "grad_norm": 0.22593729197978973, "learning_rate": 9.620156605420001e-05, "loss": 0.0217, "step": 50590 }, { "grad_norm": 0.15653616189956665, "learning_rate": 9.619945860933839e-05, "loss": 0.0233, "step": 50600 }, { "grad_norm": 0.1552274078130722, "learning_rate": 9.619735060310972e-05, "loss": 0.0227, "step": 50610 }, { "grad_norm": 0.17028556764125824, "learning_rate": 9.619524203553962e-05, "loss": 0.0265, "step": 50620 }, { "grad_norm": 0.22617952525615692, "learning_rate": 9.619313290665373e-05, "loss": 0.0298, "step": 50630 }, { "grad_norm": 0.1572096198797226, "learning_rate": 9.619102321647765e-05, "loss": 0.0253, "step": 50640 }, { "grad_norm": 0.23723861575126648, "learning_rate": 9.618891296503702e-05, "loss": 0.0225, "step": 50650 }, { "grad_norm": 0.1750580221414566, "learning_rate": 9.618680215235751e-05, "loss": 0.0227, "step": 50660 }, { "grad_norm": 0.423617422580719, "learning_rate": 9.618469077846473e-05, "loss": 0.0278, "step": 50670 }, { "grad_norm": 0.44009527564048767, "learning_rate": 9.618257884338435e-05, "loss": 0.0271, "step": 50680 }, { "grad_norm": 0.1810288280248642, "learning_rate": 9.618046634714205e-05, "loss": 0.0198, "step": 50690 }, { "grad_norm": 0.18128429353237152, "learning_rate": 9.617835328976345e-05, "loss": 0.0291, "step": 50700 }, { "grad_norm": 0.221732035279274, "learning_rate": 9.617623967127429e-05, "loss": 0.0259, "step": 50710 }, { "grad_norm": 0.1748523712158203, "learning_rate": 9.61741254917002e-05, "loss": 0.0222, "step": 50720 }, { "grad_norm": 0.1525876522064209, "learning_rate": 9.61720107510669e-05, "loss": 0.0201, "step": 50730 }, { "grad_norm": 0.17870482802391052, "learning_rate": 9.616989544940009e-05, "loss": 0.0299, "step": 50740 }, { "grad_norm": 0.17562039196491241, "learning_rate": 9.616777958672543e-05, "loss": 0.0253, "step": 50750 }, { "grad_norm": 0.24065792560577393, "learning_rate": 9.616566316306867e-05, "loss": 0.0193, "step": 50760 }, { "grad_norm": 0.19016601145267487, "learning_rate": 9.616354617845552e-05, "loss": 0.0258, "step": 50770 }, { "grad_norm": 0.22791796922683716, "learning_rate": 9.616142863291169e-05, "loss": 0.026, "step": 50780 }, { "grad_norm": 0.32600080966949463, "learning_rate": 9.615931052646293e-05, "loss": 0.0193, "step": 50790 }, { "grad_norm": 0.13855433464050293, "learning_rate": 9.615719185913496e-05, "loss": 0.0315, "step": 50800 }, { "grad_norm": 0.1346912831068039, "learning_rate": 9.615507263095353e-05, "loss": 0.022, "step": 50810 }, { "grad_norm": 0.8167099952697754, "learning_rate": 9.61529528419444e-05, "loss": 0.0275, "step": 50820 }, { "grad_norm": 0.12674081325531006, "learning_rate": 9.615083249213329e-05, "loss": 0.022, "step": 50830 }, { "grad_norm": 0.13827316462993622, "learning_rate": 9.614871158154601e-05, "loss": 0.0177, "step": 50840 }, { "grad_norm": 0.11687008291482925, "learning_rate": 9.61465901102083e-05, "loss": 0.0227, "step": 50850 }, { "grad_norm": 0.16885431110858917, "learning_rate": 9.614446807814597e-05, "loss": 0.0236, "step": 50860 }, { "grad_norm": 0.28092965483665466, "learning_rate": 9.614234548538477e-05, "loss": 0.0258, "step": 50870 }, { "grad_norm": 0.23027513921260834, "learning_rate": 9.614022233195051e-05, "loss": 0.0195, "step": 50880 }, { "grad_norm": 0.28795966506004333, "learning_rate": 9.613809861786899e-05, "loss": 0.021, "step": 50890 }, { "grad_norm": 0.21679599583148956, "learning_rate": 9.613597434316602e-05, "loss": 0.0239, "step": 50900 }, { "grad_norm": 0.13214844465255737, "learning_rate": 9.613384950786738e-05, "loss": 0.0217, "step": 50910 }, { "grad_norm": 0.2405553013086319, "learning_rate": 9.613172411199891e-05, "loss": 0.0256, "step": 50920 }, { "grad_norm": 0.15735703706741333, "learning_rate": 9.612959815558645e-05, "loss": 0.0219, "step": 50930 }, { "grad_norm": 0.19058088958263397, "learning_rate": 9.612747163865582e-05, "loss": 0.0272, "step": 50940 }, { "grad_norm": 0.17811018228530884, "learning_rate": 9.612534456123285e-05, "loss": 0.026, "step": 50950 }, { "grad_norm": 0.33199986815452576, "learning_rate": 9.612321692334341e-05, "loss": 0.0264, "step": 50960 }, { "grad_norm": 0.1440800428390503, "learning_rate": 9.612108872501331e-05, "loss": 0.0206, "step": 50970 }, { "grad_norm": 0.2897842526435852, "learning_rate": 9.611895996626845e-05, "loss": 0.026, "step": 50980 }, { "grad_norm": 0.17159676551818848, "learning_rate": 9.611683064713469e-05, "loss": 0.0268, "step": 50990 }, { "grad_norm": 0.18836164474487305, "learning_rate": 9.611470076763788e-05, "loss": 0.0248, "step": 51000 }, { "grad_norm": 0.17175233364105225, "learning_rate": 9.611257032780394e-05, "loss": 0.0255, "step": 51010 }, { "grad_norm": 0.11719001084566116, "learning_rate": 9.611043932765872e-05, "loss": 0.0155, "step": 51020 }, { "grad_norm": 0.12784911692142487, "learning_rate": 9.610830776722812e-05, "loss": 0.0281, "step": 51030 }, { "grad_norm": 0.2615887522697449, "learning_rate": 9.610617564653805e-05, "loss": 0.0273, "step": 51040 }, { "grad_norm": 0.26351112127304077, "learning_rate": 9.610404296561441e-05, "loss": 0.023, "step": 51050 }, { "grad_norm": 0.13729998469352722, "learning_rate": 9.610190972448314e-05, "loss": 0.0259, "step": 51060 }, { "grad_norm": 0.1379062384366989, "learning_rate": 9.609977592317012e-05, "loss": 0.0184, "step": 51070 }, { "grad_norm": 0.1262621432542801, "learning_rate": 9.609764156170132e-05, "loss": 0.0191, "step": 51080 }, { "grad_norm": 0.20028412342071533, "learning_rate": 9.609550664010263e-05, "loss": 0.0321, "step": 51090 }, { "grad_norm": 0.30733445286750793, "learning_rate": 9.60933711584e-05, "loss": 0.035, "step": 51100 }, { "grad_norm": 0.6093963384628296, "learning_rate": 9.609123511661941e-05, "loss": 0.0256, "step": 51110 }, { "grad_norm": 0.18455694615840912, "learning_rate": 9.60890985147868e-05, "loss": 0.0303, "step": 51120 }, { "grad_norm": 0.19507762789726257, "learning_rate": 9.608696135292812e-05, "loss": 0.0229, "step": 51130 }, { "grad_norm": 0.15098519623279572, "learning_rate": 9.608482363106935e-05, "loss": 0.0202, "step": 51140 }, { "grad_norm": 0.19355809688568115, "learning_rate": 9.608268534923646e-05, "loss": 0.0289, "step": 51150 }, { "grad_norm": 0.1599779576063156, "learning_rate": 9.608054650745543e-05, "loss": 0.0321, "step": 51160 }, { "grad_norm": 0.18864408135414124, "learning_rate": 9.607840710575226e-05, "loss": 0.0288, "step": 51170 }, { "grad_norm": 0.11426537483930588, "learning_rate": 9.607626714415294e-05, "loss": 0.0232, "step": 51180 }, { "grad_norm": 0.13176417350769043, "learning_rate": 9.607412662268346e-05, "loss": 0.0202, "step": 51190 }, { "grad_norm": 0.12550713121891022, "learning_rate": 9.607198554136985e-05, "loss": 0.0248, "step": 51200 }, { "grad_norm": 0.18105781078338623, "learning_rate": 9.606984390023812e-05, "loss": 0.0298, "step": 51210 }, { "grad_norm": 0.1668563038110733, "learning_rate": 9.606770169931427e-05, "loss": 0.0235, "step": 51220 }, { "grad_norm": 0.13574963808059692, "learning_rate": 9.606555893862437e-05, "loss": 0.0184, "step": 51230 }, { "grad_norm": 0.19101062417030334, "learning_rate": 9.606341561819443e-05, "loss": 0.0235, "step": 51240 }, { "grad_norm": 0.1768464297056198, "learning_rate": 9.606127173805048e-05, "loss": 0.0172, "step": 51250 }, { "grad_norm": 0.1112203299999237, "learning_rate": 9.605912729821862e-05, "loss": 0.0167, "step": 51260 }, { "grad_norm": 0.24761012196540833, "learning_rate": 9.605698229872487e-05, "loss": 0.0229, "step": 51270 }, { "grad_norm": 0.26162189245224, "learning_rate": 9.605483673959528e-05, "loss": 0.0168, "step": 51280 }, { "grad_norm": 0.1736016422510147, "learning_rate": 9.605269062085596e-05, "loss": 0.0183, "step": 51290 }, { "grad_norm": 0.18167127668857574, "learning_rate": 9.605054394253295e-05, "loss": 0.0269, "step": 51300 }, { "grad_norm": 0.19312606751918793, "learning_rate": 9.604839670465236e-05, "loss": 0.0228, "step": 51310 }, { "grad_norm": 0.1926419734954834, "learning_rate": 9.604624890724027e-05, "loss": 0.0233, "step": 51320 }, { "grad_norm": 0.17562489211559296, "learning_rate": 9.60441005503228e-05, "loss": 0.0251, "step": 51330 }, { "grad_norm": 0.12542420625686646, "learning_rate": 9.604195163392602e-05, "loss": 0.0281, "step": 51340 }, { "grad_norm": 0.16575711965560913, "learning_rate": 9.603980215807607e-05, "loss": 0.0208, "step": 51350 }, { "grad_norm": 0.16410379111766815, "learning_rate": 9.603765212279904e-05, "loss": 0.0215, "step": 51360 }, { "grad_norm": 0.1364687830209732, "learning_rate": 9.603550152812106e-05, "loss": 0.0178, "step": 51370 }, { "grad_norm": 0.13969585299491882, "learning_rate": 9.60333503740683e-05, "loss": 0.0193, "step": 51380 }, { "grad_norm": 0.09794287383556366, "learning_rate": 9.603119866066686e-05, "loss": 0.0223, "step": 51390 }, { "grad_norm": 0.1277105063199997, "learning_rate": 9.602904638794288e-05, "loss": 0.0146, "step": 51400 }, { "grad_norm": 0.09387316554784775, "learning_rate": 9.602689355592254e-05, "loss": 0.0207, "step": 51410 }, { "grad_norm": 0.24960047006607056, "learning_rate": 9.602474016463199e-05, "loss": 0.026, "step": 51420 }, { "grad_norm": 0.6207131743431091, "learning_rate": 9.602258621409739e-05, "loss": 0.0303, "step": 51430 }, { "grad_norm": 0.1561940461397171, "learning_rate": 9.602043170434491e-05, "loss": 0.0251, "step": 51440 }, { "grad_norm": 0.2453654408454895, "learning_rate": 9.601827663540075e-05, "loss": 0.0315, "step": 51450 }, { "grad_norm": 0.17551226913928986, "learning_rate": 9.601612100729106e-05, "loss": 0.0292, "step": 51460 }, { "grad_norm": 0.15342384576797485, "learning_rate": 9.601396482004206e-05, "loss": 0.0248, "step": 51470 }, { "grad_norm": 0.15623041987419128, "learning_rate": 9.601180807367995e-05, "loss": 0.0317, "step": 51480 }, { "grad_norm": 0.16342347860336304, "learning_rate": 9.600965076823092e-05, "loss": 0.0216, "step": 51490 }, { "grad_norm": 0.1302560716867447, "learning_rate": 9.600749290372119e-05, "loss": 0.0276, "step": 51500 }, { "grad_norm": 0.2232041209936142, "learning_rate": 9.6005334480177e-05, "loss": 0.0311, "step": 51510 }, { "grad_norm": 0.21113498508930206, "learning_rate": 9.600317549762454e-05, "loss": 0.0255, "step": 51520 }, { "grad_norm": 0.23354461789131165, "learning_rate": 9.600101595609008e-05, "loss": 0.0272, "step": 51530 }, { "grad_norm": 0.17042165994644165, "learning_rate": 9.599885585559985e-05, "loss": 0.0239, "step": 51540 }, { "grad_norm": 0.1458822637796402, "learning_rate": 9.599669519618006e-05, "loss": 0.0251, "step": 51550 }, { "grad_norm": 0.12307567149400711, "learning_rate": 9.599453397785702e-05, "loss": 0.0208, "step": 51560 }, { "grad_norm": 0.14984329044818878, "learning_rate": 9.599237220065696e-05, "loss": 0.0275, "step": 51570 }, { "grad_norm": 0.14379648864269257, "learning_rate": 9.599020986460616e-05, "loss": 0.0209, "step": 51580 }, { "grad_norm": 0.1285437047481537, "learning_rate": 9.598804696973087e-05, "loss": 0.0235, "step": 51590 }, { "grad_norm": 0.15973123908042908, "learning_rate": 9.598588351605742e-05, "loss": 0.0195, "step": 51600 }, { "grad_norm": 0.1288333684206009, "learning_rate": 9.598371950361205e-05, "loss": 0.0197, "step": 51610 }, { "grad_norm": 0.12680982053279877, "learning_rate": 9.598155493242106e-05, "loss": 0.025, "step": 51620 }, { "grad_norm": 0.22625471651554108, "learning_rate": 9.597938980251078e-05, "loss": 0.0219, "step": 51630 }, { "grad_norm": 0.15297992527484894, "learning_rate": 9.597722411390749e-05, "loss": 0.0201, "step": 51640 }, { "grad_norm": 0.16097544133663177, "learning_rate": 9.597505786663754e-05, "loss": 0.0225, "step": 51650 }, { "grad_norm": 0.12535123527050018, "learning_rate": 9.597289106072719e-05, "loss": 0.0201, "step": 51660 }, { "grad_norm": 0.21993917226791382, "learning_rate": 9.597072369620284e-05, "loss": 0.0207, "step": 51670 }, { "grad_norm": 0.16853995621204376, "learning_rate": 9.596855577309078e-05, "loss": 0.0215, "step": 51680 }, { "grad_norm": 0.13384240865707397, "learning_rate": 9.596638729141738e-05, "loss": 0.0214, "step": 51690 }, { "grad_norm": 0.14204765856266022, "learning_rate": 9.596421825120895e-05, "loss": 0.0175, "step": 51700 }, { "grad_norm": 0.20479059219360352, "learning_rate": 9.59620486524919e-05, "loss": 0.0186, "step": 51710 }, { "grad_norm": 0.6455451846122742, "learning_rate": 9.595987849529255e-05, "loss": 0.0249, "step": 51720 }, { "grad_norm": 0.19882452487945557, "learning_rate": 9.595770777963728e-05, "loss": 0.0192, "step": 51730 }, { "grad_norm": 0.303392618894577, "learning_rate": 9.595553650555248e-05, "loss": 0.0236, "step": 51740 }, { "grad_norm": 0.12333758920431137, "learning_rate": 9.595336467306451e-05, "loss": 0.0182, "step": 51750 }, { "grad_norm": 0.4118896722793579, "learning_rate": 9.595119228219978e-05, "loss": 0.0177, "step": 51760 }, { "grad_norm": 0.15253297984600067, "learning_rate": 9.594901933298468e-05, "loss": 0.0138, "step": 51770 }, { "grad_norm": 0.16416676342487335, "learning_rate": 9.594684582544559e-05, "loss": 0.0183, "step": 51780 }, { "grad_norm": 0.1024470403790474, "learning_rate": 9.594467175960897e-05, "loss": 0.0182, "step": 51790 }, { "grad_norm": 0.13730929791927338, "learning_rate": 9.59424971355012e-05, "loss": 0.0199, "step": 51800 }, { "grad_norm": 0.6105456948280334, "learning_rate": 9.594032195314872e-05, "loss": 0.0239, "step": 51810 }, { "grad_norm": 0.20550909638404846, "learning_rate": 9.593814621257794e-05, "loss": 0.0214, "step": 51820 }, { "grad_norm": 0.20350097119808197, "learning_rate": 9.593596991381531e-05, "loss": 0.0205, "step": 51830 }, { "grad_norm": 0.15437579154968262, "learning_rate": 9.593379305688728e-05, "loss": 0.0267, "step": 51840 }, { "grad_norm": 0.16283674538135529, "learning_rate": 9.593161564182029e-05, "loss": 0.0198, "step": 51850 }, { "grad_norm": 0.14050467312335968, "learning_rate": 9.592943766864083e-05, "loss": 0.0261, "step": 51860 }, { "grad_norm": 0.24719104170799255, "learning_rate": 9.59272591373753e-05, "loss": 0.0248, "step": 51870 }, { "grad_norm": 0.14228148758411407, "learning_rate": 9.592508004805023e-05, "loss": 0.0194, "step": 51880 }, { "grad_norm": 0.1527470201253891, "learning_rate": 9.592290040069207e-05, "loss": 0.0284, "step": 51890 }, { "grad_norm": 0.2453998327255249, "learning_rate": 9.592072019532731e-05, "loss": 0.0216, "step": 51900 }, { "grad_norm": 0.1977088451385498, "learning_rate": 9.591853943198245e-05, "loss": 0.0189, "step": 51910 }, { "grad_norm": 0.17552392184734344, "learning_rate": 9.5916358110684e-05, "loss": 0.0228, "step": 51920 }, { "grad_norm": 0.16520142555236816, "learning_rate": 9.591417623145843e-05, "loss": 0.0217, "step": 51930 }, { "grad_norm": 0.11356237530708313, "learning_rate": 9.591199379433225e-05, "loss": 0.0179, "step": 51940 }, { "grad_norm": 0.1375478357076645, "learning_rate": 9.590981079933202e-05, "loss": 0.0228, "step": 51950 }, { "grad_norm": 0.25073525309562683, "learning_rate": 9.590762724648423e-05, "loss": 0.0258, "step": 51960 }, { "grad_norm": 0.2756420373916626, "learning_rate": 9.590544313581545e-05, "loss": 0.0217, "step": 51970 }, { "grad_norm": 0.5368892550468445, "learning_rate": 9.590325846735216e-05, "loss": 0.0213, "step": 51980 }, { "grad_norm": 0.14352738857269287, "learning_rate": 9.590107324112095e-05, "loss": 0.0267, "step": 51990 }, { "grad_norm": 0.278921514749527, "learning_rate": 9.589888745714837e-05, "loss": 0.0246, "step": 52000 }, { "grad_norm": 0.26569533348083496, "learning_rate": 9.589670111546097e-05, "loss": 0.0328, "step": 52010 }, { "grad_norm": 0.1809491366147995, "learning_rate": 9.589451421608531e-05, "loss": 0.0237, "step": 52020 }, { "grad_norm": 0.1490488201379776, "learning_rate": 9.589232675904798e-05, "loss": 0.0226, "step": 52030 }, { "grad_norm": 0.17421691119670868, "learning_rate": 9.589013874437555e-05, "loss": 0.0357, "step": 52040 }, { "grad_norm": 0.24746179580688477, "learning_rate": 9.58879501720946e-05, "loss": 0.0202, "step": 52050 }, { "grad_norm": 0.13901405036449432, "learning_rate": 9.588576104223174e-05, "loss": 0.0262, "step": 52060 }, { "grad_norm": 0.18739312887191772, "learning_rate": 9.588357135481355e-05, "loss": 0.0285, "step": 52070 }, { "grad_norm": 0.1609112024307251, "learning_rate": 9.588138110986665e-05, "loss": 0.0249, "step": 52080 }, { "grad_norm": 0.1895420402288437, "learning_rate": 9.587919030741766e-05, "loss": 0.0221, "step": 52090 }, { "grad_norm": 0.16174732148647308, "learning_rate": 9.587699894749317e-05, "loss": 0.022, "step": 52100 }, { "grad_norm": 0.13602979481220245, "learning_rate": 9.587480703011984e-05, "loss": 0.0218, "step": 52110 }, { "grad_norm": 0.16598305106163025, "learning_rate": 9.587261455532429e-05, "loss": 0.021, "step": 52120 }, { "grad_norm": 0.12709835171699524, "learning_rate": 9.587042152313317e-05, "loss": 0.0287, "step": 52130 }, { "grad_norm": 0.1249295249581337, "learning_rate": 9.586822793357311e-05, "loss": 0.0289, "step": 52140 }, { "grad_norm": 0.1480100154876709, "learning_rate": 9.586603378667077e-05, "loss": 0.0211, "step": 52150 }, { "grad_norm": 0.13039077818393707, "learning_rate": 9.586383908245283e-05, "loss": 0.02, "step": 52160 }, { "grad_norm": 0.12839965522289276, "learning_rate": 9.586164382094593e-05, "loss": 0.0222, "step": 52170 }, { "grad_norm": 0.18994268774986267, "learning_rate": 9.585944800217677e-05, "loss": 0.0275, "step": 52180 }, { "grad_norm": 0.1231975331902504, "learning_rate": 9.5857251626172e-05, "loss": 0.0198, "step": 52190 }, { "grad_norm": 0.1639077514410019, "learning_rate": 9.585505469295833e-05, "loss": 0.0271, "step": 52200 }, { "grad_norm": 0.1907871812582016, "learning_rate": 9.585285720256246e-05, "loss": 0.0237, "step": 52210 }, { "grad_norm": 0.13619638979434967, "learning_rate": 9.585065915501108e-05, "loss": 0.0226, "step": 52220 }, { "grad_norm": 0.13876454532146454, "learning_rate": 9.58484605503309e-05, "loss": 0.0242, "step": 52230 }, { "grad_norm": 0.17057953774929047, "learning_rate": 9.584626138854864e-05, "loss": 0.0229, "step": 52240 }, { "grad_norm": 0.21851807832717896, "learning_rate": 9.584406166969101e-05, "loss": 0.0185, "step": 52250 }, { "grad_norm": 0.13701306283473969, "learning_rate": 9.584186139378474e-05, "loss": 0.0193, "step": 52260 }, { "grad_norm": 0.17090164124965668, "learning_rate": 9.58396605608566e-05, "loss": 0.0204, "step": 52270 }, { "grad_norm": 0.14854498207569122, "learning_rate": 9.583745917093327e-05, "loss": 0.0201, "step": 52280 }, { "grad_norm": 0.14555490016937256, "learning_rate": 9.583525722404157e-05, "loss": 0.0179, "step": 52290 }, { "grad_norm": 0.1588381975889206, "learning_rate": 9.58330547202082e-05, "loss": 0.02, "step": 52300 }, { "grad_norm": 0.2213679552078247, "learning_rate": 9.583085165945992e-05, "loss": 0.0214, "step": 52310 }, { "grad_norm": 0.1469346433877945, "learning_rate": 9.582864804182356e-05, "loss": 0.0188, "step": 52320 }, { "grad_norm": 0.17396607995033264, "learning_rate": 9.582644386732584e-05, "loss": 0.0239, "step": 52330 }, { "grad_norm": 0.19389735162258148, "learning_rate": 9.582423913599357e-05, "loss": 0.0209, "step": 52340 }, { "grad_norm": 0.12109671533107758, "learning_rate": 9.582203384785351e-05, "loss": 0.0205, "step": 52350 }, { "grad_norm": 0.13150325417518616, "learning_rate": 9.581982800293248e-05, "loss": 0.0221, "step": 52360 }, { "grad_norm": 0.12072500586509705, "learning_rate": 9.581762160125729e-05, "loss": 0.0215, "step": 52370 }, { "grad_norm": 0.2749835252761841, "learning_rate": 9.581541464285472e-05, "loss": 0.0236, "step": 52380 }, { "grad_norm": 0.19504199922084808, "learning_rate": 9.581320712775161e-05, "loss": 0.0224, "step": 52390 }, { "grad_norm": 0.18331684172153473, "learning_rate": 9.58109990559748e-05, "loss": 0.0262, "step": 52400 }, { "grad_norm": 0.2218855321407318, "learning_rate": 9.580879042755109e-05, "loss": 0.0221, "step": 52410 }, { "grad_norm": 0.22283107042312622, "learning_rate": 9.580658124250731e-05, "loss": 0.0247, "step": 52420 }, { "grad_norm": 0.16617797315120697, "learning_rate": 9.580437150087033e-05, "loss": 0.0225, "step": 52430 }, { "grad_norm": 0.4236678183078766, "learning_rate": 9.580216120266698e-05, "loss": 0.0237, "step": 52440 }, { "grad_norm": 0.1767648309469223, "learning_rate": 9.579995034792414e-05, "loss": 0.0184, "step": 52450 }, { "grad_norm": 0.11585726588964462, "learning_rate": 9.579773893666865e-05, "loss": 0.0279, "step": 52460 }, { "grad_norm": 0.1449018269777298, "learning_rate": 9.579552696892741e-05, "loss": 0.0245, "step": 52470 }, { "grad_norm": 0.1912185400724411, "learning_rate": 9.579331444472725e-05, "loss": 0.0231, "step": 52480 }, { "grad_norm": 0.13555096089839935, "learning_rate": 9.579110136409511e-05, "loss": 0.0324, "step": 52490 }, { "grad_norm": 0.15025152266025543, "learning_rate": 9.578888772705785e-05, "loss": 0.0183, "step": 52500 }, { "grad_norm": 0.18931354582309723, "learning_rate": 9.578667353364237e-05, "loss": 0.0219, "step": 52510 }, { "grad_norm": 0.16375015676021576, "learning_rate": 9.578445878387556e-05, "loss": 0.0188, "step": 52520 }, { "grad_norm": 0.09756000339984894, "learning_rate": 9.578224347778437e-05, "loss": 0.0218, "step": 52530 }, { "grad_norm": 0.17818541824817657, "learning_rate": 9.578002761539568e-05, "loss": 0.0291, "step": 52540 }, { "grad_norm": 0.11508288234472275, "learning_rate": 9.577781119673645e-05, "loss": 0.0186, "step": 52550 }, { "grad_norm": 0.14693762362003326, "learning_rate": 9.577559422183357e-05, "loss": 0.0247, "step": 52560 }, { "grad_norm": 0.21995727717876434, "learning_rate": 9.577337669071403e-05, "loss": 0.0269, "step": 52570 }, { "grad_norm": 0.12143514305353165, "learning_rate": 9.577115860340473e-05, "loss": 0.0174, "step": 52580 }, { "grad_norm": 0.16407372057437897, "learning_rate": 9.576893995993264e-05, "loss": 0.0284, "step": 52590 }, { "grad_norm": 0.22484086453914642, "learning_rate": 9.57667207603247e-05, "loss": 0.0241, "step": 52600 }, { "grad_norm": 0.29752621054649353, "learning_rate": 9.57645010046079e-05, "loss": 0.0227, "step": 52610 }, { "grad_norm": 0.25210291147232056, "learning_rate": 9.576228069280921e-05, "loss": 0.0209, "step": 52620 }, { "grad_norm": 0.1359328180551529, "learning_rate": 9.576005982495562e-05, "loss": 0.0188, "step": 52630 }, { "grad_norm": 0.1061217337846756, "learning_rate": 9.575783840107407e-05, "loss": 0.0201, "step": 52640 }, { "grad_norm": 0.5698025226593018, "learning_rate": 9.575561642119159e-05, "loss": 0.0285, "step": 52650 }, { "grad_norm": 0.1467549204826355, "learning_rate": 9.575339388533517e-05, "loss": 0.0191, "step": 52660 }, { "grad_norm": 0.18788523972034454, "learning_rate": 9.575117079353182e-05, "loss": 0.0183, "step": 52670 }, { "grad_norm": 0.24871498346328735, "learning_rate": 9.574894714580854e-05, "loss": 0.0172, "step": 52680 }, { "grad_norm": 0.25415557622909546, "learning_rate": 9.574672294219237e-05, "loss": 0.0216, "step": 52690 }, { "grad_norm": 0.17503201961517334, "learning_rate": 9.574449818271032e-05, "loss": 0.0198, "step": 52700 }, { "grad_norm": 0.2092614322900772, "learning_rate": 9.574227286738942e-05, "loss": 0.023, "step": 52710 }, { "grad_norm": 0.3039608597755432, "learning_rate": 9.574004699625671e-05, "loss": 0.0193, "step": 52720 }, { "grad_norm": 0.26769787073135376, "learning_rate": 9.573782056933927e-05, "loss": 0.0281, "step": 52730 }, { "grad_norm": 0.2170180380344391, "learning_rate": 9.57355935866641e-05, "loss": 0.022, "step": 52740 }, { "grad_norm": 0.24328283965587616, "learning_rate": 9.573336604825831e-05, "loss": 0.0276, "step": 52750 }, { "grad_norm": 0.11322557926177979, "learning_rate": 9.573113795414891e-05, "loss": 0.0207, "step": 52760 }, { "grad_norm": 0.2925291657447815, "learning_rate": 9.572890930436305e-05, "loss": 0.0264, "step": 52770 }, { "grad_norm": 0.17171041667461395, "learning_rate": 9.572668009892773e-05, "loss": 0.0264, "step": 52780 }, { "grad_norm": 0.1967329978942871, "learning_rate": 9.57244503378701e-05, "loss": 0.0381, "step": 52790 }, { "grad_norm": 0.18543076515197754, "learning_rate": 9.572222002121722e-05, "loss": 0.0291, "step": 52800 }, { "grad_norm": 0.19927822053432465, "learning_rate": 9.571998914899619e-05, "loss": 0.0232, "step": 52810 }, { "grad_norm": 0.14600785076618195, "learning_rate": 9.571775772123412e-05, "loss": 0.0225, "step": 52820 }, { "grad_norm": 0.24162614345550537, "learning_rate": 9.571552573795815e-05, "loss": 0.0253, "step": 52830 }, { "grad_norm": 0.19620691239833832, "learning_rate": 9.571329319919535e-05, "loss": 0.0245, "step": 52840 }, { "grad_norm": 0.5164036750793457, "learning_rate": 9.571106010497291e-05, "loss": 0.0239, "step": 52850 }, { "grad_norm": 0.2220112532377243, "learning_rate": 9.57088264553179e-05, "loss": 0.024, "step": 52860 }, { "grad_norm": 0.1869691014289856, "learning_rate": 9.570659225025751e-05, "loss": 0.0213, "step": 52870 }, { "grad_norm": 0.15817508101463318, "learning_rate": 9.570435748981888e-05, "loss": 0.0225, "step": 52880 }, { "grad_norm": 0.15739411115646362, "learning_rate": 9.570212217402913e-05, "loss": 0.0215, "step": 52890 }, { "grad_norm": 0.17288368940353394, "learning_rate": 9.569988630291545e-05, "loss": 0.0205, "step": 52900 }, { "grad_norm": 0.19536037743091583, "learning_rate": 9.569764987650502e-05, "loss": 0.0216, "step": 52910 }, { "grad_norm": 0.16434365510940552, "learning_rate": 9.569541289482498e-05, "loss": 0.0182, "step": 52920 }, { "grad_norm": 0.23631079494953156, "learning_rate": 9.569317535790255e-05, "loss": 0.0246, "step": 52930 }, { "grad_norm": 0.32728907465934753, "learning_rate": 9.569093726576488e-05, "loss": 0.0357, "step": 52940 }, { "grad_norm": 0.15597118437290192, "learning_rate": 9.568869861843919e-05, "loss": 0.0359, "step": 52950 }, { "grad_norm": 0.12908190488815308, "learning_rate": 9.568645941595267e-05, "loss": 0.0196, "step": 52960 }, { "grad_norm": 0.24907556176185608, "learning_rate": 9.568421965833254e-05, "loss": 0.029, "step": 52970 }, { "grad_norm": 0.2600168287754059, "learning_rate": 9.568197934560602e-05, "loss": 0.0267, "step": 52980 }, { "grad_norm": 0.12698233127593994, "learning_rate": 9.567973847780029e-05, "loss": 0.0211, "step": 52990 }, { "grad_norm": 0.17573031783103943, "learning_rate": 9.567749705494263e-05, "loss": 0.0239, "step": 53000 }, { "grad_norm": 0.29710325598716736, "learning_rate": 9.567525507706024e-05, "loss": 0.0237, "step": 53010 }, { "grad_norm": 0.2625836133956909, "learning_rate": 9.567301254418038e-05, "loss": 0.0224, "step": 53020 }, { "grad_norm": 0.1879122406244278, "learning_rate": 9.567076945633031e-05, "loss": 0.0197, "step": 53030 }, { "grad_norm": 0.1320965439081192, "learning_rate": 9.566852581353726e-05, "loss": 0.0256, "step": 53040 }, { "grad_norm": 0.20240077376365662, "learning_rate": 9.56662816158285e-05, "loss": 0.0216, "step": 53050 }, { "grad_norm": 0.17547230422496796, "learning_rate": 9.566403686323131e-05, "loss": 0.0209, "step": 53060 }, { "grad_norm": 0.15484201908111572, "learning_rate": 9.566179155577294e-05, "loss": 0.0259, "step": 53070 }, { "grad_norm": 0.17791958153247833, "learning_rate": 9.565954569348072e-05, "loss": 0.0253, "step": 53080 }, { "grad_norm": 0.1602264642715454, "learning_rate": 9.565729927638188e-05, "loss": 0.0225, "step": 53090 }, { "grad_norm": 0.17595969140529633, "learning_rate": 9.565505230450376e-05, "loss": 0.0218, "step": 53100 }, { "grad_norm": 0.13196446001529694, "learning_rate": 9.565280477787367e-05, "loss": 0.0183, "step": 53110 }, { "grad_norm": 0.3487047851085663, "learning_rate": 9.565055669651886e-05, "loss": 0.0279, "step": 53120 }, { "grad_norm": 0.20550929009914398, "learning_rate": 9.56483080604667e-05, "loss": 0.022, "step": 53130 }, { "grad_norm": 0.1468190848827362, "learning_rate": 9.564605886974451e-05, "loss": 0.0146, "step": 53140 }, { "grad_norm": 0.3908796012401581, "learning_rate": 9.56438091243796e-05, "loss": 0.028, "step": 53150 }, { "grad_norm": 0.14196760952472687, "learning_rate": 9.564155882439931e-05, "loss": 0.0217, "step": 53160 }, { "grad_norm": 0.18074369430541992, "learning_rate": 9.563930796983098e-05, "loss": 0.0259, "step": 53170 }, { "grad_norm": 0.23261883854866028, "learning_rate": 9.563705656070198e-05, "loss": 0.0274, "step": 53180 }, { "grad_norm": 0.1848904937505722, "learning_rate": 9.563480459703966e-05, "loss": 0.0241, "step": 53190 }, { "grad_norm": 0.13174580037593842, "learning_rate": 9.563255207887136e-05, "loss": 0.02, "step": 53200 }, { "grad_norm": 0.17943301796913147, "learning_rate": 9.563029900622449e-05, "loss": 0.0217, "step": 53210 }, { "grad_norm": 0.15203197300434113, "learning_rate": 9.562804537912639e-05, "loss": 0.0155, "step": 53220 }, { "grad_norm": 0.2134803980588913, "learning_rate": 9.562579119760447e-05, "loss": 0.0204, "step": 53230 }, { "grad_norm": 0.14125873148441315, "learning_rate": 9.56235364616861e-05, "loss": 0.02, "step": 53240 }, { "grad_norm": 0.22097691893577576, "learning_rate": 9.562128117139868e-05, "loss": 0.0243, "step": 53250 }, { "grad_norm": 0.11154001206159592, "learning_rate": 9.561902532676963e-05, "loss": 0.0164, "step": 53260 }, { "grad_norm": 0.3078298270702362, "learning_rate": 9.561676892782633e-05, "loss": 0.0245, "step": 53270 }, { "grad_norm": 0.1688402146100998, "learning_rate": 9.561451197459625e-05, "loss": 0.0239, "step": 53280 }, { "grad_norm": 0.21570271253585815, "learning_rate": 9.561225446710676e-05, "loss": 0.0188, "step": 53290 }, { "grad_norm": 0.393510639667511, "learning_rate": 9.560999640538532e-05, "loss": 0.019, "step": 53300 }, { "grad_norm": 0.17751306295394897, "learning_rate": 9.560773778945938e-05, "loss": 0.0245, "step": 53310 }, { "grad_norm": 0.2615048885345459, "learning_rate": 9.560547861935634e-05, "loss": 0.0205, "step": 53320 }, { "grad_norm": 0.11859329044818878, "learning_rate": 9.560321889510368e-05, "loss": 0.0174, "step": 53330 }, { "grad_norm": 0.40433359146118164, "learning_rate": 9.560095861672885e-05, "loss": 0.0307, "step": 53340 }, { "grad_norm": 0.12563548982143402, "learning_rate": 9.559869778425933e-05, "loss": 0.0266, "step": 53350 }, { "grad_norm": 0.21371175348758698, "learning_rate": 9.559643639772258e-05, "loss": 0.0216, "step": 53360 }, { "grad_norm": 0.20641443133354187, "learning_rate": 9.559417445714607e-05, "loss": 0.0226, "step": 53370 }, { "grad_norm": 0.12924686074256897, "learning_rate": 9.55919119625573e-05, "loss": 0.0207, "step": 53380 }, { "grad_norm": 0.23324531316757202, "learning_rate": 9.558964891398374e-05, "loss": 0.0237, "step": 53390 }, { "grad_norm": 0.11684541404247284, "learning_rate": 9.55873853114529e-05, "loss": 0.024, "step": 53400 }, { "grad_norm": 0.13088379800319672, "learning_rate": 9.55851211549923e-05, "loss": 0.0232, "step": 53410 }, { "grad_norm": 0.2732868194580078, "learning_rate": 9.558285644462943e-05, "loss": 0.0195, "step": 53420 }, { "grad_norm": 0.23410005867481232, "learning_rate": 9.558059118039183e-05, "loss": 0.0173, "step": 53430 }, { "grad_norm": 0.15407849848270416, "learning_rate": 9.557832536230701e-05, "loss": 0.0208, "step": 53440 }, { "grad_norm": 0.11658912152051926, "learning_rate": 9.55760589904025e-05, "loss": 0.0182, "step": 53450 }, { "grad_norm": 0.1688896119594574, "learning_rate": 9.557379206470583e-05, "loss": 0.0228, "step": 53460 }, { "grad_norm": 0.170802041888237, "learning_rate": 9.557152458524457e-05, "loss": 0.0161, "step": 53470 }, { "grad_norm": 0.2948214113712311, "learning_rate": 9.556925655204626e-05, "loss": 0.0229, "step": 53480 }, { "grad_norm": 0.1797502189874649, "learning_rate": 9.556698796513845e-05, "loss": 0.0277, "step": 53490 }, { "grad_norm": 0.14420954883098602, "learning_rate": 9.556471882454872e-05, "loss": 0.0172, "step": 53500 }, { "grad_norm": 0.16520756483078003, "learning_rate": 9.556244913030465e-05, "loss": 0.0277, "step": 53510 }, { "grad_norm": 0.14242789149284363, "learning_rate": 9.556017888243378e-05, "loss": 0.0245, "step": 53520 }, { "grad_norm": 0.2844971716403961, "learning_rate": 9.555790808096374e-05, "loss": 0.028, "step": 53530 }, { "grad_norm": 0.15845368802547455, "learning_rate": 9.55556367259221e-05, "loss": 0.0186, "step": 53540 }, { "grad_norm": 0.17479132115840912, "learning_rate": 9.555336481733644e-05, "loss": 0.0221, "step": 53550 }, { "grad_norm": 0.21270878612995148, "learning_rate": 9.55510923552344e-05, "loss": 0.019, "step": 53560 }, { "grad_norm": 0.2403697669506073, "learning_rate": 9.554881933964359e-05, "loss": 0.0262, "step": 53570 }, { "grad_norm": 0.1610264629125595, "learning_rate": 9.554654577059163e-05, "loss": 0.0199, "step": 53580 }, { "grad_norm": 0.12562455236911774, "learning_rate": 9.554427164810611e-05, "loss": 0.0185, "step": 53590 }, { "grad_norm": 0.16482345759868622, "learning_rate": 9.55419969722147e-05, "loss": 0.0224, "step": 53600 }, { "grad_norm": 0.20920875668525696, "learning_rate": 9.553972174294503e-05, "loss": 0.0233, "step": 53610 }, { "grad_norm": 0.23299150168895721, "learning_rate": 9.553744596032474e-05, "loss": 0.0179, "step": 53620 }, { "grad_norm": 0.16583923995494843, "learning_rate": 9.553516962438149e-05, "loss": 0.0194, "step": 53630 }, { "grad_norm": 0.1953882873058319, "learning_rate": 9.553289273514294e-05, "loss": 0.0236, "step": 53640 }, { "grad_norm": 0.145137757062912, "learning_rate": 9.553061529263673e-05, "loss": 0.0245, "step": 53650 }, { "grad_norm": 0.18266573548316956, "learning_rate": 9.55283372968906e-05, "loss": 0.028, "step": 53660 }, { "grad_norm": 0.2215011864900589, "learning_rate": 9.552605874793214e-05, "loss": 0.0243, "step": 53670 }, { "grad_norm": 0.17101643979549408, "learning_rate": 9.552377964578912e-05, "loss": 0.0227, "step": 53680 }, { "grad_norm": 0.16484840214252472, "learning_rate": 9.552149999048917e-05, "loss": 0.0218, "step": 53690 }, { "grad_norm": 0.37657344341278076, "learning_rate": 9.551921978206001e-05, "loss": 0.0215, "step": 53700 }, { "grad_norm": 0.17959003150463104, "learning_rate": 9.551693902052938e-05, "loss": 0.0222, "step": 53710 }, { "grad_norm": 0.13187134265899658, "learning_rate": 9.551465770592495e-05, "loss": 0.0216, "step": 53720 }, { "grad_norm": 0.14972110092639923, "learning_rate": 9.551237583827446e-05, "loss": 0.0251, "step": 53730 }, { "grad_norm": 0.1709645539522171, "learning_rate": 9.551009341760564e-05, "loss": 0.0216, "step": 53740 }, { "grad_norm": 0.1647544801235199, "learning_rate": 9.550781044394621e-05, "loss": 0.0274, "step": 53750 }, { "grad_norm": 0.14006631076335907, "learning_rate": 9.55055269173239e-05, "loss": 0.0288, "step": 53760 }, { "grad_norm": 0.16938921809196472, "learning_rate": 9.55032428377665e-05, "loss": 0.0212, "step": 53770 }, { "grad_norm": 0.16688872873783112, "learning_rate": 9.550095820530172e-05, "loss": 0.0239, "step": 53780 }, { "grad_norm": 0.1232624426484108, "learning_rate": 9.549867301995734e-05, "loss": 0.0204, "step": 53790 }, { "grad_norm": 0.18057087063789368, "learning_rate": 9.549638728176114e-05, "loss": 0.0201, "step": 53800 }, { "grad_norm": 0.15994055569171906, "learning_rate": 9.549410099074085e-05, "loss": 0.0169, "step": 53810 }, { "grad_norm": 0.19746729731559753, "learning_rate": 9.549181414692431e-05, "loss": 0.0201, "step": 53820 }, { "grad_norm": 0.18742892146110535, "learning_rate": 9.548952675033926e-05, "loss": 0.0218, "step": 53830 }, { "grad_norm": 0.29670965671539307, "learning_rate": 9.548723880101351e-05, "loss": 0.0314, "step": 53840 }, { "grad_norm": 0.15110516548156738, "learning_rate": 9.548495029897486e-05, "loss": 0.0202, "step": 53850 }, { "grad_norm": 0.1515321284532547, "learning_rate": 9.548266124425113e-05, "loss": 0.0207, "step": 53860 }, { "grad_norm": 0.26452600955963135, "learning_rate": 9.548037163687011e-05, "loss": 0.0261, "step": 53870 }, { "grad_norm": 0.14428356289863586, "learning_rate": 9.547808147685965e-05, "loss": 0.0261, "step": 53880 }, { "grad_norm": 0.4002090394496918, "learning_rate": 9.547579076424756e-05, "loss": 0.0267, "step": 53890 }, { "grad_norm": 0.17572259902954102, "learning_rate": 9.547349949906166e-05, "loss": 0.0176, "step": 53900 }, { "grad_norm": 0.20790015161037445, "learning_rate": 9.547120768132982e-05, "loss": 0.0276, "step": 53910 }, { "grad_norm": 0.19524338841438293, "learning_rate": 9.546891531107986e-05, "loss": 0.0204, "step": 53920 }, { "grad_norm": 0.19306068122386932, "learning_rate": 9.546662238833967e-05, "loss": 0.0285, "step": 53930 }, { "grad_norm": 0.319286972284317, "learning_rate": 9.546432891313708e-05, "loss": 0.0283, "step": 53940 }, { "grad_norm": 0.26252973079681396, "learning_rate": 9.546203488549997e-05, "loss": 0.0229, "step": 53950 }, { "grad_norm": 0.13377529382705688, "learning_rate": 9.545974030545621e-05, "loss": 0.0248, "step": 53960 }, { "grad_norm": 0.13166163861751556, "learning_rate": 9.545744517303368e-05, "loss": 0.0195, "step": 53970 }, { "grad_norm": 0.226660817861557, "learning_rate": 9.545514948826028e-05, "loss": 0.0195, "step": 53980 }, { "grad_norm": 0.0871511846780777, "learning_rate": 9.54528532511639e-05, "loss": 0.0261, "step": 53990 }, { "grad_norm": 0.16291627287864685, "learning_rate": 9.545055646177243e-05, "loss": 0.0159, "step": 54000 }, { "grad_norm": 0.14971943199634552, "learning_rate": 9.544825912011379e-05, "loss": 0.0197, "step": 54010 }, { "grad_norm": 0.11742836236953735, "learning_rate": 9.54459612262159e-05, "loss": 0.0208, "step": 54020 }, { "grad_norm": 0.14414477348327637, "learning_rate": 9.544366278010666e-05, "loss": 0.0324, "step": 54030 }, { "grad_norm": 0.31921425461769104, "learning_rate": 9.544136378181401e-05, "loss": 0.0253, "step": 54040 }, { "grad_norm": 0.3102840781211853, "learning_rate": 9.543906423136589e-05, "loss": 0.0208, "step": 54050 }, { "grad_norm": 0.14285236597061157, "learning_rate": 9.543676412879024e-05, "loss": 0.0219, "step": 54060 }, { "grad_norm": 0.4287378191947937, "learning_rate": 9.543446347411501e-05, "loss": 0.0249, "step": 54070 }, { "grad_norm": 0.2012711763381958, "learning_rate": 9.543216226736815e-05, "loss": 0.0196, "step": 54080 }, { "grad_norm": 0.13582633435726166, "learning_rate": 9.542986050857761e-05, "loss": 0.0217, "step": 54090 }, { "grad_norm": 0.13693973422050476, "learning_rate": 9.542755819777139e-05, "loss": 0.0228, "step": 54100 }, { "grad_norm": 0.14845360815525055, "learning_rate": 9.542525533497743e-05, "loss": 0.0222, "step": 54110 }, { "grad_norm": 0.28931257128715515, "learning_rate": 9.542295192022374e-05, "loss": 0.0223, "step": 54120 }, { "grad_norm": 0.14523129165172577, "learning_rate": 9.542064795353831e-05, "loss": 0.0256, "step": 54130 }, { "grad_norm": 0.17477145791053772, "learning_rate": 9.541834343494911e-05, "loss": 0.0257, "step": 54140 }, { "grad_norm": 0.15296486020088196, "learning_rate": 9.541603836448417e-05, "loss": 0.0258, "step": 54150 }, { "grad_norm": 0.3249233067035675, "learning_rate": 9.541373274217145e-05, "loss": 0.0342, "step": 54160 }, { "grad_norm": 0.16070368885993958, "learning_rate": 9.541142656803903e-05, "loss": 0.0197, "step": 54170 }, { "grad_norm": 0.16426439583301544, "learning_rate": 9.540911984211489e-05, "loss": 0.0163, "step": 54180 }, { "grad_norm": 0.801121711730957, "learning_rate": 9.540681256442707e-05, "loss": 0.0243, "step": 54190 }, { "grad_norm": 0.17712299525737762, "learning_rate": 9.54045047350036e-05, "loss": 0.0271, "step": 54200 }, { "grad_norm": 0.18985165655612946, "learning_rate": 9.540219635387253e-05, "loss": 0.0277, "step": 54210 }, { "grad_norm": 0.2955690026283264, "learning_rate": 9.539988742106191e-05, "loss": 0.0251, "step": 54220 }, { "grad_norm": 0.1825176179409027, "learning_rate": 9.53975779365998e-05, "loss": 0.0163, "step": 54230 }, { "grad_norm": 0.6607115864753723, "learning_rate": 9.539526790051426e-05, "loss": 0.0335, "step": 54240 }, { "grad_norm": 0.151728555560112, "learning_rate": 9.539295731283336e-05, "loss": 0.0263, "step": 54250 }, { "grad_norm": 0.2282545417547226, "learning_rate": 9.539064617358515e-05, "loss": 0.024, "step": 54260 }, { "grad_norm": 0.16339676082134247, "learning_rate": 9.538833448279773e-05, "loss": 0.0243, "step": 54270 }, { "grad_norm": 0.17956022918224335, "learning_rate": 9.538602224049922e-05, "loss": 0.0197, "step": 54280 }, { "grad_norm": 0.2641856372356415, "learning_rate": 9.538370944671766e-05, "loss": 0.0324, "step": 54290 }, { "grad_norm": 0.15782485902309418, "learning_rate": 9.53813961014812e-05, "loss": 0.0251, "step": 54300 }, { "grad_norm": 0.19448868930339813, "learning_rate": 9.537908220481791e-05, "loss": 0.0199, "step": 54310 }, { "grad_norm": 0.1478617936372757, "learning_rate": 9.537676775675595e-05, "loss": 0.0214, "step": 54320 }, { "grad_norm": 0.15841402113437653, "learning_rate": 9.537445275732341e-05, "loss": 0.0225, "step": 54330 }, { "grad_norm": 0.22803381085395813, "learning_rate": 9.537213720654843e-05, "loss": 0.0181, "step": 54340 }, { "grad_norm": 0.11191420257091522, "learning_rate": 9.536982110445914e-05, "loss": 0.0173, "step": 54350 }, { "grad_norm": 0.16038423776626587, "learning_rate": 9.536750445108367e-05, "loss": 0.0218, "step": 54360 }, { "grad_norm": 0.0982515811920166, "learning_rate": 9.536518724645022e-05, "loss": 0.0235, "step": 54370 }, { "grad_norm": 0.1702672839164734, "learning_rate": 9.53628694905869e-05, "loss": 0.0275, "step": 54380 }, { "grad_norm": 0.15211418271064758, "learning_rate": 9.536055118352187e-05, "loss": 0.0198, "step": 54390 }, { "grad_norm": 0.12887197732925415, "learning_rate": 9.535823232528334e-05, "loss": 0.0264, "step": 54400 }, { "grad_norm": 0.12175850570201874, "learning_rate": 9.535591291589945e-05, "loss": 0.0262, "step": 54410 }, { "grad_norm": 0.14123177528381348, "learning_rate": 9.53535929553984e-05, "loss": 0.0232, "step": 54420 }, { "grad_norm": 0.15839079022407532, "learning_rate": 9.535127244380837e-05, "loss": 0.0261, "step": 54430 }, { "grad_norm": 0.2723773121833801, "learning_rate": 9.534895138115755e-05, "loss": 0.0303, "step": 54440 }, { "grad_norm": 0.23145262897014618, "learning_rate": 9.534662976747418e-05, "loss": 0.02, "step": 54450 }, { "grad_norm": 0.2662648558616638, "learning_rate": 9.534430760278644e-05, "loss": 0.0253, "step": 54460 }, { "grad_norm": 0.18350590765476227, "learning_rate": 9.534198488712253e-05, "loss": 0.0228, "step": 54470 }, { "grad_norm": 0.19228753447532654, "learning_rate": 9.533966162051071e-05, "loss": 0.0203, "step": 54480 }, { "grad_norm": 0.1801132708787918, "learning_rate": 9.533733780297919e-05, "loss": 0.0229, "step": 54490 }, { "grad_norm": 0.16668181121349335, "learning_rate": 9.53350134345562e-05, "loss": 0.0237, "step": 54500 }, { "grad_norm": 0.20740535855293274, "learning_rate": 9.533268851527001e-05, "loss": 0.0195, "step": 54510 }, { "grad_norm": 0.21917547285556793, "learning_rate": 9.533036304514884e-05, "loss": 0.0215, "step": 54520 }, { "grad_norm": 0.21729417145252228, "learning_rate": 9.532803702422098e-05, "loss": 0.023, "step": 54530 }, { "grad_norm": 0.2200617492198944, "learning_rate": 9.532571045251465e-05, "loss": 0.0174, "step": 54540 }, { "grad_norm": 0.4029034376144409, "learning_rate": 9.532338333005815e-05, "loss": 0.0375, "step": 54550 }, { "grad_norm": 0.2562166750431061, "learning_rate": 9.532105565687976e-05, "loss": 0.0255, "step": 54560 }, { "grad_norm": 0.14897876977920532, "learning_rate": 9.531872743300774e-05, "loss": 0.0252, "step": 54570 }, { "grad_norm": 0.2340649962425232, "learning_rate": 9.531639865847039e-05, "loss": 0.0281, "step": 54580 }, { "grad_norm": 0.1629336178302765, "learning_rate": 9.531406933329602e-05, "loss": 0.0175, "step": 54590 }, { "grad_norm": 0.24574537575244904, "learning_rate": 9.531173945751293e-05, "loss": 0.0213, "step": 54600 }, { "grad_norm": 0.23879775404930115, "learning_rate": 9.530940903114942e-05, "loss": 0.0313, "step": 54610 }, { "grad_norm": 0.16846798360347748, "learning_rate": 9.530707805423381e-05, "loss": 0.0269, "step": 54620 }, { "grad_norm": 0.17483486235141754, "learning_rate": 9.530474652679444e-05, "loss": 0.0271, "step": 54630 }, { "grad_norm": 0.15248309075832367, "learning_rate": 9.530241444885961e-05, "loss": 0.0181, "step": 54640 }, { "grad_norm": 0.3024568557739258, "learning_rate": 9.530008182045766e-05, "loss": 0.0209, "step": 54650 }, { "grad_norm": 0.382691353559494, "learning_rate": 9.529774864161695e-05, "loss": 0.0208, "step": 54660 }, { "grad_norm": 0.22140221297740936, "learning_rate": 9.529541491236585e-05, "loss": 0.0224, "step": 54670 }, { "grad_norm": 0.2502489387989044, "learning_rate": 9.529308063273265e-05, "loss": 0.0208, "step": 54680 }, { "grad_norm": 0.20310364663600922, "learning_rate": 9.529074580274579e-05, "loss": 0.0222, "step": 54690 }, { "grad_norm": 0.12389063090085983, "learning_rate": 9.52884104224336e-05, "loss": 0.0211, "step": 54700 }, { "grad_norm": 0.24905556440353394, "learning_rate": 9.528607449182447e-05, "loss": 0.0271, "step": 54710 }, { "grad_norm": 0.13056279718875885, "learning_rate": 9.528373801094676e-05, "loss": 0.0236, "step": 54720 }, { "grad_norm": 0.17265550792217255, "learning_rate": 9.528140097982889e-05, "loss": 0.0203, "step": 54730 }, { "grad_norm": 0.3338611423969269, "learning_rate": 9.527906339849925e-05, "loss": 0.0203, "step": 54740 }, { "grad_norm": 0.10388899594545364, "learning_rate": 9.527672526698624e-05, "loss": 0.0198, "step": 54750 }, { "grad_norm": 0.13517120480537415, "learning_rate": 9.527438658531826e-05, "loss": 0.0234, "step": 54760 }, { "grad_norm": 0.20531821250915527, "learning_rate": 9.527204735352374e-05, "loss": 0.0224, "step": 54770 }, { "grad_norm": 0.19996289908885956, "learning_rate": 9.52697075716311e-05, "loss": 0.0213, "step": 54780 }, { "grad_norm": 0.2656513452529907, "learning_rate": 9.526736723966878e-05, "loss": 0.0184, "step": 54790 }, { "grad_norm": 0.18549983203411102, "learning_rate": 9.52650263576652e-05, "loss": 0.0185, "step": 54800 }, { "grad_norm": 0.1583249270915985, "learning_rate": 9.52626849256488e-05, "loss": 0.0197, "step": 54810 }, { "grad_norm": 0.25800031423568726, "learning_rate": 9.526034294364807e-05, "loss": 0.0261, "step": 54820 }, { "grad_norm": 0.18411238491535187, "learning_rate": 9.525800041169142e-05, "loss": 0.0372, "step": 54830 }, { "grad_norm": 0.21413925290107727, "learning_rate": 9.525565732980735e-05, "loss": 0.0246, "step": 54840 }, { "grad_norm": 0.22724543511867523, "learning_rate": 9.525331369802431e-05, "loss": 0.0229, "step": 54850 }, { "grad_norm": 0.13698434829711914, "learning_rate": 9.525096951637078e-05, "loss": 0.0167, "step": 54860 }, { "grad_norm": 0.152090385556221, "learning_rate": 9.524862478487524e-05, "loss": 0.0174, "step": 54870 }, { "grad_norm": 0.22576215863227844, "learning_rate": 9.52462795035662e-05, "loss": 0.025, "step": 54880 }, { "grad_norm": 0.14085854589939117, "learning_rate": 9.524393367247212e-05, "loss": 0.0253, "step": 54890 }, { "grad_norm": 0.20133036375045776, "learning_rate": 9.524158729162155e-05, "loss": 0.0272, "step": 54900 }, { "grad_norm": 0.27656084299087524, "learning_rate": 9.523924036104297e-05, "loss": 0.0197, "step": 54910 }, { "grad_norm": 0.18724478781223297, "learning_rate": 9.523689288076491e-05, "loss": 0.022, "step": 54920 }, { "grad_norm": 0.34532055258750916, "learning_rate": 9.52345448508159e-05, "loss": 0.0187, "step": 54930 }, { "grad_norm": 0.1775708943605423, "learning_rate": 9.523219627122445e-05, "loss": 0.0241, "step": 54940 }, { "grad_norm": 0.24799954891204834, "learning_rate": 9.522984714201911e-05, "loss": 0.0228, "step": 54950 }, { "grad_norm": 0.504073977470398, "learning_rate": 9.522749746322844e-05, "loss": 0.0245, "step": 54960 }, { "grad_norm": 0.14577624201774597, "learning_rate": 9.522514723488095e-05, "loss": 0.0256, "step": 54970 }, { "grad_norm": 0.1693023145198822, "learning_rate": 9.522279645700522e-05, "loss": 0.0179, "step": 54980 }, { "grad_norm": 0.16872228682041168, "learning_rate": 9.522044512962984e-05, "loss": 0.0257, "step": 54990 }, { "grad_norm": 0.1711934357881546, "learning_rate": 9.521809325278335e-05, "loss": 0.0217, "step": 55000 }, { "grad_norm": 0.17557241022586823, "learning_rate": 9.521574082649431e-05, "loss": 0.0174, "step": 55010 }, { "grad_norm": 0.20511402189731598, "learning_rate": 9.521338785079134e-05, "loss": 0.0222, "step": 55020 }, { "grad_norm": 0.28323137760162354, "learning_rate": 9.521103432570304e-05, "loss": 0.0276, "step": 55030 }, { "grad_norm": 0.3217543363571167, "learning_rate": 9.520868025125798e-05, "loss": 0.0212, "step": 55040 }, { "grad_norm": 0.11359425634145737, "learning_rate": 9.520632562748474e-05, "loss": 0.0196, "step": 55050 }, { "grad_norm": 0.13836586475372314, "learning_rate": 9.5203970454412e-05, "loss": 0.0239, "step": 55060 }, { "grad_norm": 0.17081807553768158, "learning_rate": 9.520161473206832e-05, "loss": 0.0259, "step": 55070 }, { "grad_norm": 0.1925978660583496, "learning_rate": 9.519925846048234e-05, "loss": 0.0198, "step": 55080 }, { "grad_norm": 0.19105105102062225, "learning_rate": 9.519690163968272e-05, "loss": 0.0201, "step": 55090 }, { "grad_norm": 0.1366904228925705, "learning_rate": 9.519454426969803e-05, "loss": 0.0207, "step": 55100 }, { "grad_norm": 0.16317810118198395, "learning_rate": 9.5192186350557e-05, "loss": 0.025, "step": 55110 }, { "grad_norm": 0.23300881683826447, "learning_rate": 9.518982788228821e-05, "loss": 0.0141, "step": 55120 }, { "grad_norm": 0.18046267330646515, "learning_rate": 9.518746886492035e-05, "loss": 0.0217, "step": 55130 }, { "grad_norm": 0.1762377917766571, "learning_rate": 9.518510929848208e-05, "loss": 0.0212, "step": 55140 }, { "grad_norm": 0.17074497044086456, "learning_rate": 9.518274918300208e-05, "loss": 0.0236, "step": 55150 }, { "grad_norm": 0.15195152163505554, "learning_rate": 9.5180388518509e-05, "loss": 0.0214, "step": 55160 }, { "grad_norm": 0.19934678077697754, "learning_rate": 9.517802730503156e-05, "loss": 0.0244, "step": 55170 }, { "grad_norm": 0.18733805418014526, "learning_rate": 9.517566554259842e-05, "loss": 0.0256, "step": 55180 }, { "grad_norm": 0.28058215975761414, "learning_rate": 9.51733032312383e-05, "loss": 0.0192, "step": 55190 }, { "grad_norm": 0.1797550916671753, "learning_rate": 9.51709403709799e-05, "loss": 0.024, "step": 55200 }, { "grad_norm": 0.1936953216791153, "learning_rate": 9.51685769618519e-05, "loss": 0.0288, "step": 55210 }, { "grad_norm": 0.14560839533805847, "learning_rate": 9.516621300388306e-05, "loss": 0.0248, "step": 55220 }, { "grad_norm": 0.3245117962360382, "learning_rate": 9.51638484971021e-05, "loss": 0.0257, "step": 55230 }, { "grad_norm": 0.19003790616989136, "learning_rate": 9.516148344153773e-05, "loss": 0.0232, "step": 55240 }, { "grad_norm": 0.1701788753271103, "learning_rate": 9.515911783721868e-05, "loss": 0.0195, "step": 55250 }, { "grad_norm": 0.10324309021234512, "learning_rate": 9.515675168417372e-05, "loss": 0.0194, "step": 55260 }, { "grad_norm": 0.24445442855358124, "learning_rate": 9.515438498243161e-05, "loss": 0.0224, "step": 55270 }, { "grad_norm": 0.16352418065071106, "learning_rate": 9.515201773202106e-05, "loss": 0.0239, "step": 55280 }, { "grad_norm": 0.13578423857688904, "learning_rate": 9.514964993297089e-05, "loss": 0.0266, "step": 55290 }, { "grad_norm": 0.16160237789154053, "learning_rate": 9.514728158530983e-05, "loss": 0.0238, "step": 55300 }, { "grad_norm": 0.16939575970172882, "learning_rate": 9.51449126890667e-05, "loss": 0.0229, "step": 55310 }, { "grad_norm": 0.1550995409488678, "learning_rate": 9.514254324427022e-05, "loss": 0.0196, "step": 55320 }, { "grad_norm": 0.1914553940296173, "learning_rate": 9.514017325094923e-05, "loss": 0.019, "step": 55330 }, { "grad_norm": 0.1857004463672638, "learning_rate": 9.513780270913253e-05, "loss": 0.0218, "step": 55340 }, { "grad_norm": 0.17173899710178375, "learning_rate": 9.51354316188489e-05, "loss": 0.0217, "step": 55350 }, { "grad_norm": 0.2778022289276123, "learning_rate": 9.513305998012716e-05, "loss": 0.0242, "step": 55360 }, { "grad_norm": 0.27716439962387085, "learning_rate": 9.513068779299612e-05, "loss": 0.0197, "step": 55370 }, { "grad_norm": 0.17755639553070068, "learning_rate": 9.512831505748463e-05, "loss": 0.0238, "step": 55380 }, { "grad_norm": 0.11935122311115265, "learning_rate": 9.512594177362148e-05, "loss": 0.0212, "step": 55390 }, { "grad_norm": 0.156917005777359, "learning_rate": 9.512356794143556e-05, "loss": 0.0265, "step": 55400 }, { "grad_norm": 0.1727336049079895, "learning_rate": 9.512119356095567e-05, "loss": 0.0272, "step": 55410 }, { "grad_norm": 0.16766758263111115, "learning_rate": 9.511881863221068e-05, "loss": 0.0285, "step": 55420 }, { "grad_norm": 0.2439567595720291, "learning_rate": 9.511644315522945e-05, "loss": 0.0194, "step": 55430 }, { "grad_norm": 0.16689571738243103, "learning_rate": 9.511406713004084e-05, "loss": 0.025, "step": 55440 }, { "grad_norm": 0.19549749791622162, "learning_rate": 9.511169055667371e-05, "loss": 0.0199, "step": 55450 }, { "grad_norm": 0.22324606776237488, "learning_rate": 9.510931343515694e-05, "loss": 0.0223, "step": 55460 }, { "grad_norm": 0.1651865839958191, "learning_rate": 9.510693576551943e-05, "loss": 0.0241, "step": 55470 }, { "grad_norm": 0.1552126258611679, "learning_rate": 9.510455754779006e-05, "loss": 0.0203, "step": 55480 }, { "grad_norm": 0.1477281004190445, "learning_rate": 9.510217878199774e-05, "loss": 0.0204, "step": 55490 }, { "grad_norm": 0.08105068653821945, "learning_rate": 9.509979946817136e-05, "loss": 0.0198, "step": 55500 }, { "grad_norm": 0.17146094143390656, "learning_rate": 9.509741960633981e-05, "loss": 0.023, "step": 55510 }, { "grad_norm": 0.10740794241428375, "learning_rate": 9.509503919653207e-05, "loss": 0.0194, "step": 55520 }, { "grad_norm": 0.11154953390359879, "learning_rate": 9.5092658238777e-05, "loss": 0.02, "step": 55530 }, { "grad_norm": 0.15545839071273804, "learning_rate": 9.509027673310358e-05, "loss": 0.0155, "step": 55540 }, { "grad_norm": 0.2263738363981247, "learning_rate": 9.508789467954072e-05, "loss": 0.0182, "step": 55550 }, { "grad_norm": 0.37820786237716675, "learning_rate": 9.508551207811735e-05, "loss": 0.0254, "step": 55560 }, { "grad_norm": 0.19002537429332733, "learning_rate": 9.508312892886247e-05, "loss": 0.0222, "step": 55570 }, { "grad_norm": 0.18388530611991882, "learning_rate": 9.5080745231805e-05, "loss": 0.0222, "step": 55580 }, { "grad_norm": 0.07650210708379745, "learning_rate": 9.507836098697389e-05, "loss": 0.016, "step": 55590 }, { "grad_norm": 0.20056597888469696, "learning_rate": 9.507597619439816e-05, "loss": 0.0304, "step": 55600 }, { "grad_norm": 0.20585404336452484, "learning_rate": 9.507359085410675e-05, "loss": 0.0231, "step": 55610 }, { "grad_norm": 0.238165482878685, "learning_rate": 9.507120496612865e-05, "loss": 0.0191, "step": 55620 }, { "grad_norm": 0.10978572070598602, "learning_rate": 9.506881853049285e-05, "loss": 0.0177, "step": 55630 }, { "grad_norm": 0.16689848899841309, "learning_rate": 9.506643154722837e-05, "loss": 0.0176, "step": 55640 }, { "grad_norm": 0.14653052389621735, "learning_rate": 9.50640440163642e-05, "loss": 0.024, "step": 55650 }, { "grad_norm": 0.4586203396320343, "learning_rate": 9.506165593792933e-05, "loss": 0.0226, "step": 55660 }, { "grad_norm": 0.1622745841741562, "learning_rate": 9.50592673119528e-05, "loss": 0.0185, "step": 55670 }, { "grad_norm": 0.21035771071910858, "learning_rate": 9.505687813846363e-05, "loss": 0.0286, "step": 55680 }, { "grad_norm": 0.14128097891807556, "learning_rate": 9.505448841749085e-05, "loss": 0.0216, "step": 55690 }, { "grad_norm": 0.1892474740743637, "learning_rate": 9.505209814906351e-05, "loss": 0.022, "step": 55700 }, { "grad_norm": 0.1305409073829651, "learning_rate": 9.504970733321062e-05, "loss": 0.018, "step": 55710 }, { "grad_norm": 0.20081274211406708, "learning_rate": 9.504731596996125e-05, "loss": 0.0226, "step": 55720 }, { "grad_norm": 0.2472854107618332, "learning_rate": 9.504492405934448e-05, "loss": 0.0217, "step": 55730 }, { "grad_norm": 0.17364059388637543, "learning_rate": 9.504253160138934e-05, "loss": 0.025, "step": 55740 }, { "grad_norm": 0.34268924593925476, "learning_rate": 9.504013859612493e-05, "loss": 0.0237, "step": 55750 }, { "grad_norm": 0.15731081366539001, "learning_rate": 9.503774504358029e-05, "loss": 0.0222, "step": 55760 }, { "grad_norm": 0.22190891206264496, "learning_rate": 9.503535094378453e-05, "loss": 0.03, "step": 55770 }, { "grad_norm": 0.15726251900196075, "learning_rate": 9.503295629676674e-05, "loss": 0.0224, "step": 55780 }, { "grad_norm": 0.23662978410720825, "learning_rate": 9.503056110255601e-05, "loss": 0.0243, "step": 55790 }, { "grad_norm": 0.1526069939136505, "learning_rate": 9.502816536118147e-05, "loss": 0.0224, "step": 55800 }, { "grad_norm": 0.4267808794975281, "learning_rate": 9.502576907267218e-05, "loss": 0.0227, "step": 55810 }, { "grad_norm": 0.17200496792793274, "learning_rate": 9.502337223705729e-05, "loss": 0.0201, "step": 55820 }, { "grad_norm": 0.1440892219543457, "learning_rate": 9.502097485436591e-05, "loss": 0.0245, "step": 55830 }, { "grad_norm": 0.15222764015197754, "learning_rate": 9.501857692462719e-05, "loss": 0.0259, "step": 55840 }, { "grad_norm": 0.22836169600486755, "learning_rate": 9.501617844787025e-05, "loss": 0.023, "step": 55850 }, { "grad_norm": 0.18197061121463776, "learning_rate": 9.501377942412425e-05, "loss": 0.0161, "step": 55860 }, { "grad_norm": 0.1269969791173935, "learning_rate": 9.501137985341832e-05, "loss": 0.0191, "step": 55870 }, { "grad_norm": 0.1829550713300705, "learning_rate": 9.500897973578162e-05, "loss": 0.0228, "step": 55880 }, { "grad_norm": 0.1598139852285385, "learning_rate": 9.500657907124334e-05, "loss": 0.0235, "step": 55890 }, { "grad_norm": 0.45234963297843933, "learning_rate": 9.500417785983262e-05, "loss": 0.0213, "step": 55900 }, { "grad_norm": 0.1484004557132721, "learning_rate": 9.500177610157865e-05, "loss": 0.0211, "step": 55910 }, { "grad_norm": 0.5753706693649292, "learning_rate": 9.499937379651062e-05, "loss": 0.0237, "step": 55920 }, { "grad_norm": 0.1360699087381363, "learning_rate": 9.499697094465769e-05, "loss": 0.0184, "step": 55930 }, { "grad_norm": 0.1463056355714798, "learning_rate": 9.499456754604909e-05, "loss": 0.0237, "step": 55940 }, { "grad_norm": 0.1727968007326126, "learning_rate": 9.499216360071402e-05, "loss": 0.0182, "step": 55950 }, { "grad_norm": 0.08715378493070602, "learning_rate": 9.498975910868168e-05, "loss": 0.0198, "step": 55960 }, { "grad_norm": 0.14317403733730316, "learning_rate": 9.498735406998128e-05, "loss": 0.0268, "step": 55970 }, { "grad_norm": 0.19145165383815765, "learning_rate": 9.498494848464206e-05, "loss": 0.0316, "step": 55980 }, { "grad_norm": 0.14796386659145355, "learning_rate": 9.498254235269325e-05, "loss": 0.0294, "step": 55990 }, { "grad_norm": 0.36102116107940674, "learning_rate": 9.498013567416407e-05, "loss": 0.0192, "step": 56000 }, { "grad_norm": 0.16985847055912018, "learning_rate": 9.497772844908376e-05, "loss": 0.0182, "step": 56010 }, { "grad_norm": 0.19442398846149445, "learning_rate": 9.497532067748162e-05, "loss": 0.0296, "step": 56020 }, { "grad_norm": 0.20638321340084076, "learning_rate": 9.497291235938685e-05, "loss": 0.0237, "step": 56030 }, { "grad_norm": 0.1439342498779297, "learning_rate": 9.497050349482872e-05, "loss": 0.025, "step": 56040 }, { "grad_norm": 0.20275916159152985, "learning_rate": 9.496809408383654e-05, "loss": 0.0199, "step": 56050 }, { "grad_norm": 0.1212225928902626, "learning_rate": 9.496568412643954e-05, "loss": 0.029, "step": 56060 }, { "grad_norm": 0.2998092770576477, "learning_rate": 9.496327362266702e-05, "loss": 0.0288, "step": 56070 }, { "grad_norm": 0.17766812443733215, "learning_rate": 9.496086257254829e-05, "loss": 0.0277, "step": 56080 }, { "grad_norm": 0.16287101805210114, "learning_rate": 9.495845097611262e-05, "loss": 0.0234, "step": 56090 }, { "grad_norm": 0.14330101013183594, "learning_rate": 9.495603883338932e-05, "loss": 0.0215, "step": 56100 }, { "grad_norm": 0.14677737653255463, "learning_rate": 9.495362614440771e-05, "loss": 0.0223, "step": 56110 }, { "grad_norm": 0.17892609536647797, "learning_rate": 9.49512129091971e-05, "loss": 0.0202, "step": 56120 }, { "grad_norm": 0.14371779561042786, "learning_rate": 9.49487991277868e-05, "loss": 0.0265, "step": 56130 }, { "grad_norm": 0.3616880774497986, "learning_rate": 9.494638480020617e-05, "loss": 0.0303, "step": 56140 }, { "grad_norm": 0.2174626886844635, "learning_rate": 9.494396992648451e-05, "loss": 0.0203, "step": 56150 }, { "grad_norm": 0.2662329375743866, "learning_rate": 9.494155450665119e-05, "loss": 0.0212, "step": 56160 }, { "grad_norm": 0.14116430282592773, "learning_rate": 9.493913854073554e-05, "loss": 0.0221, "step": 56170 }, { "grad_norm": 0.2141978144645691, "learning_rate": 9.493672202876696e-05, "loss": 0.0221, "step": 56180 }, { "grad_norm": 0.3365090787410736, "learning_rate": 9.493430497077476e-05, "loss": 0.0196, "step": 56190 }, { "grad_norm": 0.16496501863002777, "learning_rate": 9.493188736678831e-05, "loss": 0.0233, "step": 56200 }, { "grad_norm": 0.24575039744377136, "learning_rate": 9.492946921683703e-05, "loss": 0.022, "step": 56210 }, { "grad_norm": 0.15547789633274078, "learning_rate": 9.492705052095028e-05, "loss": 0.022, "step": 56220 }, { "grad_norm": 0.3491836488246918, "learning_rate": 9.492463127915742e-05, "loss": 0.0217, "step": 56230 }, { "grad_norm": 0.2170579880475998, "learning_rate": 9.49222114914879e-05, "loss": 0.022, "step": 56240 }, { "grad_norm": 0.22190290689468384, "learning_rate": 9.491979115797109e-05, "loss": 0.0207, "step": 56250 }, { "grad_norm": 0.2652110159397125, "learning_rate": 9.49173702786364e-05, "loss": 0.0204, "step": 56260 }, { "grad_norm": 0.15638260543346405, "learning_rate": 9.491494885351328e-05, "loss": 0.0213, "step": 56270 }, { "grad_norm": 0.23136550188064575, "learning_rate": 9.49125268826311e-05, "loss": 0.0215, "step": 56280 }, { "grad_norm": 0.23091600835323334, "learning_rate": 9.491010436601931e-05, "loss": 0.0223, "step": 56290 }, { "grad_norm": 0.1430177092552185, "learning_rate": 9.490768130370736e-05, "loss": 0.0169, "step": 56300 }, { "grad_norm": 0.14279857277870178, "learning_rate": 9.490525769572468e-05, "loss": 0.0239, "step": 56310 }, { "grad_norm": 0.14751732349395752, "learning_rate": 9.490283354210072e-05, "loss": 0.0222, "step": 56320 }, { "grad_norm": 0.10411057621240616, "learning_rate": 9.490040884286493e-05, "loss": 0.0229, "step": 56330 }, { "grad_norm": 0.2127363383769989, "learning_rate": 9.48979835980468e-05, "loss": 0.024, "step": 56340 }, { "grad_norm": 0.23376984894275665, "learning_rate": 9.489555780767576e-05, "loss": 0.0258, "step": 56350 }, { "grad_norm": 0.2594870328903198, "learning_rate": 9.489313147178131e-05, "loss": 0.0229, "step": 56360 }, { "grad_norm": 0.15950971841812134, "learning_rate": 9.489070459039293e-05, "loss": 0.0211, "step": 56370 }, { "grad_norm": 0.11938376724720001, "learning_rate": 9.48882771635401e-05, "loss": 0.0208, "step": 56380 }, { "grad_norm": 0.4521246552467346, "learning_rate": 9.488584919125232e-05, "loss": 0.0295, "step": 56390 }, { "grad_norm": 0.23595358431339264, "learning_rate": 9.48834206735591e-05, "loss": 0.0282, "step": 56400 }, { "grad_norm": 0.11010845750570297, "learning_rate": 9.488099161048994e-05, "loss": 0.0192, "step": 56410 }, { "grad_norm": 0.3173739016056061, "learning_rate": 9.487856200207436e-05, "loss": 0.0203, "step": 56420 }, { "grad_norm": 0.1152467280626297, "learning_rate": 9.487613184834187e-05, "loss": 0.0146, "step": 56430 }, { "grad_norm": 0.12818197906017303, "learning_rate": 9.487370114932202e-05, "loss": 0.0219, "step": 56440 }, { "grad_norm": 0.2073737382888794, "learning_rate": 9.487126990504431e-05, "loss": 0.0183, "step": 56450 }, { "grad_norm": 0.3226705491542816, "learning_rate": 9.486883811553833e-05, "loss": 0.0161, "step": 56460 }, { "grad_norm": 0.16251003742218018, "learning_rate": 9.486640578083359e-05, "loss": 0.0234, "step": 56470 }, { "grad_norm": 0.1684807986021042, "learning_rate": 9.486397290095968e-05, "loss": 0.0199, "step": 56480 }, { "grad_norm": 0.2682415246963501, "learning_rate": 9.486153947594612e-05, "loss": 0.023, "step": 56490 }, { "grad_norm": 0.1586795151233673, "learning_rate": 9.48591055058225e-05, "loss": 0.0274, "step": 56500 }, { "grad_norm": 0.13028812408447266, "learning_rate": 9.485667099061839e-05, "loss": 0.0237, "step": 56510 }, { "grad_norm": 0.1321800947189331, "learning_rate": 9.485423593036339e-05, "loss": 0.0224, "step": 56520 }, { "grad_norm": 0.13424859941005707, "learning_rate": 9.485180032508708e-05, "loss": 0.0167, "step": 56530 }, { "grad_norm": 0.36111700534820557, "learning_rate": 9.484936417481902e-05, "loss": 0.032, "step": 56540 }, { "grad_norm": 0.26380741596221924, "learning_rate": 9.484692747958885e-05, "loss": 0.0191, "step": 56550 }, { "grad_norm": 0.1783716082572937, "learning_rate": 9.484449023942619e-05, "loss": 0.024, "step": 56560 }, { "grad_norm": 0.3483045697212219, "learning_rate": 9.484205245436062e-05, "loss": 0.0246, "step": 56570 }, { "grad_norm": 0.19677361845970154, "learning_rate": 9.483961412442177e-05, "loss": 0.0269, "step": 56580 }, { "grad_norm": 0.32033103704452515, "learning_rate": 9.483717524963926e-05, "loss": 0.0199, "step": 56590 }, { "grad_norm": 0.1258082240819931, "learning_rate": 9.483473583004276e-05, "loss": 0.0252, "step": 56600 }, { "grad_norm": 0.12163286656141281, "learning_rate": 9.483229586566189e-05, "loss": 0.0238, "step": 56610 }, { "grad_norm": 0.134754940867424, "learning_rate": 9.482985535652629e-05, "loss": 0.0232, "step": 56620 }, { "grad_norm": 0.2758459746837616, "learning_rate": 9.482741430266561e-05, "loss": 0.025, "step": 56630 }, { "grad_norm": 0.147297665476799, "learning_rate": 9.482497270410953e-05, "loss": 0.018, "step": 56640 }, { "grad_norm": 0.18707391619682312, "learning_rate": 9.482253056088773e-05, "loss": 0.0195, "step": 56650 }, { "grad_norm": 0.567268967628479, "learning_rate": 9.482008787302983e-05, "loss": 0.0271, "step": 56660 }, { "grad_norm": 0.3077690899372101, "learning_rate": 9.481764464056556e-05, "loss": 0.0265, "step": 56670 }, { "grad_norm": 0.15699779987335205, "learning_rate": 9.48152008635246e-05, "loss": 0.0204, "step": 56680 }, { "grad_norm": 0.21985726058483124, "learning_rate": 9.481275654193661e-05, "loss": 0.0155, "step": 56690 }, { "grad_norm": 0.12813109159469604, "learning_rate": 9.481031167583135e-05, "loss": 0.0231, "step": 56700 }, { "grad_norm": 0.13363471627235413, "learning_rate": 9.480786626523849e-05, "loss": 0.0198, "step": 56710 }, { "grad_norm": 0.1635543555021286, "learning_rate": 9.480542031018775e-05, "loss": 0.0257, "step": 56720 }, { "grad_norm": 1.2423114776611328, "learning_rate": 9.480297381070883e-05, "loss": 0.0213, "step": 56730 }, { "grad_norm": 0.143046572804451, "learning_rate": 9.48005267668315e-05, "loss": 0.0207, "step": 56740 }, { "grad_norm": 0.2108885794878006, "learning_rate": 9.479807917858545e-05, "loss": 0.0284, "step": 56750 }, { "grad_norm": 0.11356852203607559, "learning_rate": 9.479563104600046e-05, "loss": 0.0176, "step": 56760 }, { "grad_norm": 0.14659540355205536, "learning_rate": 9.479318236910625e-05, "loss": 0.0196, "step": 56770 }, { "grad_norm": 0.1582697629928589, "learning_rate": 9.479073314793259e-05, "loss": 0.0164, "step": 56780 }, { "grad_norm": 0.1382920742034912, "learning_rate": 9.478828338250922e-05, "loss": 0.025, "step": 56790 }, { "grad_norm": 0.15188877284526825, "learning_rate": 9.478583307286592e-05, "loss": 0.0166, "step": 56800 }, { "grad_norm": 0.20942094922065735, "learning_rate": 9.478338221903248e-05, "loss": 0.0285, "step": 56810 }, { "grad_norm": 0.10517831146717072, "learning_rate": 9.478093082103865e-05, "loss": 0.0231, "step": 56820 }, { "grad_norm": 0.12408769875764847, "learning_rate": 9.477847887891423e-05, "loss": 0.025, "step": 56830 }, { "grad_norm": 0.22591717541217804, "learning_rate": 9.477602639268902e-05, "loss": 0.0266, "step": 56840 }, { "grad_norm": 0.2243831306695938, "learning_rate": 9.477357336239281e-05, "loss": 0.0258, "step": 56850 }, { "grad_norm": 0.2229357361793518, "learning_rate": 9.477111978805542e-05, "loss": 0.0254, "step": 56860 }, { "grad_norm": 0.16153623163700104, "learning_rate": 9.476866566970664e-05, "loss": 0.0257, "step": 56870 }, { "grad_norm": 0.17813332378864288, "learning_rate": 9.47662110073763e-05, "loss": 0.024, "step": 56880 }, { "grad_norm": 0.288816899061203, "learning_rate": 9.476375580109425e-05, "loss": 0.0299, "step": 56890 }, { "grad_norm": 0.13889168202877045, "learning_rate": 9.476130005089029e-05, "loss": 0.0213, "step": 56900 }, { "grad_norm": 0.3008243441581726, "learning_rate": 9.475884375679427e-05, "loss": 0.0248, "step": 56910 }, { "grad_norm": 0.19444790482521057, "learning_rate": 9.475638691883604e-05, "loss": 0.0192, "step": 56920 }, { "grad_norm": 0.24508552253246307, "learning_rate": 9.475392953704546e-05, "loss": 0.0251, "step": 56930 }, { "grad_norm": 0.17630016803741455, "learning_rate": 9.475147161145237e-05, "loss": 0.0242, "step": 56940 }, { "grad_norm": 0.17144820094108582, "learning_rate": 9.474901314208666e-05, "loss": 0.0231, "step": 56950 }, { "grad_norm": 0.20384356379508972, "learning_rate": 9.474655412897818e-05, "loss": 0.0216, "step": 56960 }, { "grad_norm": 0.15624558925628662, "learning_rate": 9.474409457215683e-05, "loss": 0.0275, "step": 56970 }, { "grad_norm": 0.1737464815378189, "learning_rate": 9.474163447165247e-05, "loss": 0.0177, "step": 56980 }, { "grad_norm": 0.28679361939430237, "learning_rate": 9.473917382749502e-05, "loss": 0.0241, "step": 56990 }, { "grad_norm": 0.39722520112991333, "learning_rate": 9.473671263971438e-05, "loss": 0.0204, "step": 57000 }, { "grad_norm": 0.22576047480106354, "learning_rate": 9.473425090834041e-05, "loss": 0.0321, "step": 57010 }, { "grad_norm": 0.23343078792095184, "learning_rate": 9.473178863340309e-05, "loss": 0.0234, "step": 57020 }, { "grad_norm": 0.22149105370044708, "learning_rate": 9.472932581493228e-05, "loss": 0.0193, "step": 57030 }, { "grad_norm": 0.22823596000671387, "learning_rate": 9.472686245295793e-05, "loss": 0.0224, "step": 57040 }, { "grad_norm": 0.16189081966876984, "learning_rate": 9.472439854750999e-05, "loss": 0.0214, "step": 57050 }, { "grad_norm": 0.13123898208141327, "learning_rate": 9.472193409861836e-05, "loss": 0.0177, "step": 57060 }, { "grad_norm": 0.11836782842874527, "learning_rate": 9.471946910631302e-05, "loss": 0.0178, "step": 57070 }, { "grad_norm": 0.2939164340496063, "learning_rate": 9.47170035706239e-05, "loss": 0.0218, "step": 57080 }, { "grad_norm": 0.15555734932422638, "learning_rate": 9.471453749158096e-05, "loss": 0.0183, "step": 57090 }, { "grad_norm": 0.1646527349948883, "learning_rate": 9.47120708692142e-05, "loss": 0.0193, "step": 57100 }, { "grad_norm": 0.18358059227466583, "learning_rate": 9.470960370355356e-05, "loss": 0.0198, "step": 57110 }, { "grad_norm": 0.18150222301483154, "learning_rate": 9.4707135994629e-05, "loss": 0.0205, "step": 57120 }, { "grad_norm": 0.16614916920661926, "learning_rate": 9.470466774247054e-05, "loss": 0.0176, "step": 57130 }, { "grad_norm": 0.25256332755088806, "learning_rate": 9.470219894710814e-05, "loss": 0.017, "step": 57140 }, { "grad_norm": 0.37886643409729004, "learning_rate": 9.469972960857182e-05, "loss": 0.0192, "step": 57150 }, { "grad_norm": 0.12603959441184998, "learning_rate": 9.469725972689161e-05, "loss": 0.0223, "step": 57160 }, { "grad_norm": 0.1584560126066208, "learning_rate": 9.469478930209748e-05, "loss": 0.0239, "step": 57170 }, { "grad_norm": 0.26657363772392273, "learning_rate": 9.469231833421945e-05, "loss": 0.0242, "step": 57180 }, { "grad_norm": 0.5088979601860046, "learning_rate": 9.468984682328756e-05, "loss": 0.0267, "step": 57190 }, { "grad_norm": 0.15187276899814606, "learning_rate": 9.468737476933184e-05, "loss": 0.0229, "step": 57200 }, { "grad_norm": 0.14588557183742523, "learning_rate": 9.468490217238232e-05, "loss": 0.0169, "step": 57210 }, { "grad_norm": 0.3017658591270447, "learning_rate": 9.468242903246907e-05, "loss": 0.0158, "step": 57220 }, { "grad_norm": 0.1309737116098404, "learning_rate": 9.46799553496221e-05, "loss": 0.0198, "step": 57230 }, { "grad_norm": 0.1597120463848114, "learning_rate": 9.46774811238715e-05, "loss": 0.0241, "step": 57240 }, { "grad_norm": 0.14158789813518524, "learning_rate": 9.467500635524731e-05, "loss": 0.0243, "step": 57250 }, { "grad_norm": 0.19656915962696075, "learning_rate": 9.46725310437796e-05, "loss": 0.0222, "step": 57260 }, { "grad_norm": 0.24167600274085999, "learning_rate": 9.467005518949848e-05, "loss": 0.0238, "step": 57270 }, { "grad_norm": 0.16176916658878326, "learning_rate": 9.466757879243402e-05, "loss": 0.0181, "step": 57280 }, { "grad_norm": 0.15425562858581543, "learning_rate": 9.46651018526163e-05, "loss": 0.021, "step": 57290 }, { "grad_norm": 0.17949937283992767, "learning_rate": 9.46626243700754e-05, "loss": 0.023, "step": 57300 }, { "grad_norm": 0.24608725309371948, "learning_rate": 9.466014634484147e-05, "loss": 0.0175, "step": 57310 }, { "grad_norm": 0.14271822571754456, "learning_rate": 9.465766777694458e-05, "loss": 0.0207, "step": 57320 }, { "grad_norm": 0.278682678937912, "learning_rate": 9.465518866641487e-05, "loss": 0.021, "step": 57330 }, { "grad_norm": 0.14869044721126556, "learning_rate": 9.465270901328247e-05, "loss": 0.0201, "step": 57340 }, { "grad_norm": 0.10751646757125854, "learning_rate": 9.465022881757749e-05, "loss": 0.017, "step": 57350 }, { "grad_norm": 0.126175194978714, "learning_rate": 9.464774807933005e-05, "loss": 0.0253, "step": 57360 }, { "grad_norm": 0.5950605273246765, "learning_rate": 9.464526679857034e-05, "loss": 0.0277, "step": 57370 }, { "grad_norm": 0.16286341845989227, "learning_rate": 9.464278497532848e-05, "loss": 0.0213, "step": 57380 }, { "grad_norm": 0.13630884885787964, "learning_rate": 9.464030260963462e-05, "loss": 0.0238, "step": 57390 }, { "grad_norm": 0.21849507093429565, "learning_rate": 9.463781970151895e-05, "loss": 0.0219, "step": 57400 }, { "grad_norm": 0.17697204649448395, "learning_rate": 9.463533625101162e-05, "loss": 0.0239, "step": 57410 }, { "grad_norm": 0.27265533804893494, "learning_rate": 9.463285225814283e-05, "loss": 0.0207, "step": 57420 }, { "grad_norm": 0.15285874903202057, "learning_rate": 9.463036772294272e-05, "loss": 0.0199, "step": 57430 }, { "grad_norm": 0.22120872139930725, "learning_rate": 9.462788264544152e-05, "loss": 0.0273, "step": 57440 }, { "grad_norm": 0.2145271748304367, "learning_rate": 9.462539702566939e-05, "loss": 0.0254, "step": 57450 }, { "grad_norm": 0.15316054224967957, "learning_rate": 9.462291086365657e-05, "loss": 0.0235, "step": 57460 }, { "grad_norm": 0.2198425978422165, "learning_rate": 9.462042415943324e-05, "loss": 0.0179, "step": 57470 }, { "grad_norm": 0.17647604644298553, "learning_rate": 9.461793691302965e-05, "loss": 0.0202, "step": 57480 }, { "grad_norm": 0.08717237412929535, "learning_rate": 9.461544912447597e-05, "loss": 0.0172, "step": 57490 }, { "grad_norm": 0.10363325476646423, "learning_rate": 9.461296079380249e-05, "loss": 0.0213, "step": 57500 }, { "grad_norm": 0.42988938093185425, "learning_rate": 9.461047192103939e-05, "loss": 0.0232, "step": 57510 }, { "grad_norm": 0.1490163505077362, "learning_rate": 9.460798250621694e-05, "loss": 0.0167, "step": 57520 }, { "grad_norm": 0.3250468969345093, "learning_rate": 9.46054925493654e-05, "loss": 0.0256, "step": 57530 }, { "grad_norm": 0.17088888585567474, "learning_rate": 9.4603002050515e-05, "loss": 0.0214, "step": 57540 }, { "grad_norm": 0.12579989433288574, "learning_rate": 9.460051100969602e-05, "loss": 0.0255, "step": 57550 }, { "grad_norm": 0.5342118740081787, "learning_rate": 9.459801942693871e-05, "loss": 0.0246, "step": 57560 }, { "grad_norm": 0.18338918685913086, "learning_rate": 9.459552730227336e-05, "loss": 0.0186, "step": 57570 }, { "grad_norm": 0.4200848639011383, "learning_rate": 9.459303463573026e-05, "loss": 0.0211, "step": 57580 }, { "grad_norm": 0.1532762348651886, "learning_rate": 9.459054142733968e-05, "loss": 0.023, "step": 57590 }, { "grad_norm": 0.2521043121814728, "learning_rate": 9.458804767713192e-05, "loss": 0.0218, "step": 57600 }, { "grad_norm": 0.24202954769134521, "learning_rate": 9.458555338513728e-05, "loss": 0.0194, "step": 57610 }, { "grad_norm": 0.15501223504543304, "learning_rate": 9.458305855138607e-05, "loss": 0.0235, "step": 57620 }, { "grad_norm": 0.47122520208358765, "learning_rate": 9.458056317590862e-05, "loss": 0.0282, "step": 57630 }, { "grad_norm": 0.10585086792707443, "learning_rate": 9.457806725873522e-05, "loss": 0.0178, "step": 57640 }, { "grad_norm": 0.2046978771686554, "learning_rate": 9.457557079989622e-05, "loss": 0.0234, "step": 57650 }, { "grad_norm": 0.20720775425434113, "learning_rate": 9.457307379942195e-05, "loss": 0.0241, "step": 57660 }, { "grad_norm": 0.24915367364883423, "learning_rate": 9.457057625734273e-05, "loss": 0.0258, "step": 57670 }, { "grad_norm": 0.2672445774078369, "learning_rate": 9.456807817368895e-05, "loss": 0.0206, "step": 57680 }, { "grad_norm": 0.13396063446998596, "learning_rate": 9.456557954849094e-05, "loss": 0.021, "step": 57690 }, { "grad_norm": 0.13507325947284698, "learning_rate": 9.456308038177906e-05, "loss": 0.0235, "step": 57700 }, { "grad_norm": 0.11952374875545502, "learning_rate": 9.456058067358367e-05, "loss": 0.0175, "step": 57710 }, { "grad_norm": 0.1662580668926239, "learning_rate": 9.455808042393516e-05, "loss": 0.0211, "step": 57720 }, { "grad_norm": 0.24034321308135986, "learning_rate": 9.45555796328639e-05, "loss": 0.0245, "step": 57730 }, { "grad_norm": 0.21317891776561737, "learning_rate": 9.455307830040028e-05, "loss": 0.0196, "step": 57740 }, { "grad_norm": 0.14732405543327332, "learning_rate": 9.455057642657469e-05, "loss": 0.0275, "step": 57750 }, { "grad_norm": 0.24029114842414856, "learning_rate": 9.454807401141754e-05, "loss": 0.023, "step": 57760 }, { "grad_norm": 0.30015382170677185, "learning_rate": 9.454557105495923e-05, "loss": 0.0203, "step": 57770 }, { "grad_norm": 0.22659535706043243, "learning_rate": 9.454306755723017e-05, "loss": 0.0174, "step": 57780 }, { "grad_norm": 0.17148980498313904, "learning_rate": 9.45405635182608e-05, "loss": 0.0197, "step": 57790 }, { "grad_norm": 0.22618763148784637, "learning_rate": 9.45380589380815e-05, "loss": 0.0374, "step": 57800 }, { "grad_norm": 0.46060848236083984, "learning_rate": 9.453555381672275e-05, "loss": 0.022, "step": 57810 }, { "grad_norm": 0.17866605520248413, "learning_rate": 9.453304815421497e-05, "loss": 0.0244, "step": 57820 }, { "grad_norm": 0.1629134565591812, "learning_rate": 9.453054195058862e-05, "loss": 0.0198, "step": 57830 }, { "grad_norm": 0.21204431354999542, "learning_rate": 9.452803520587412e-05, "loss": 0.0206, "step": 57840 }, { "grad_norm": 0.13450978696346283, "learning_rate": 9.452552792010197e-05, "loss": 0.0321, "step": 57850 }, { "grad_norm": 0.1402466595172882, "learning_rate": 9.452302009330261e-05, "loss": 0.0252, "step": 57860 }, { "grad_norm": 0.1429671049118042, "learning_rate": 9.452051172550652e-05, "loss": 0.0228, "step": 57870 }, { "grad_norm": 0.24490751326084137, "learning_rate": 9.451800281674417e-05, "loss": 0.0194, "step": 57880 }, { "grad_norm": 0.15328794717788696, "learning_rate": 9.451549336704606e-05, "loss": 0.0241, "step": 57890 }, { "grad_norm": 0.5248415470123291, "learning_rate": 9.451298337644267e-05, "loss": 0.0232, "step": 57900 }, { "grad_norm": 0.10932085663080215, "learning_rate": 9.451047284496452e-05, "loss": 0.018, "step": 57910 }, { "grad_norm": 0.19166362285614014, "learning_rate": 9.450796177264209e-05, "loss": 0.0349, "step": 57920 }, { "grad_norm": 0.2877453565597534, "learning_rate": 9.450545015950591e-05, "loss": 0.0251, "step": 57930 }, { "grad_norm": 0.21883350610733032, "learning_rate": 9.450293800558648e-05, "loss": 0.0222, "step": 57940 }, { "grad_norm": 0.14945246279239655, "learning_rate": 9.450042531091433e-05, "loss": 0.0217, "step": 57950 }, { "grad_norm": 0.17444217205047607, "learning_rate": 9.449791207552001e-05, "loss": 0.0214, "step": 57960 }, { "grad_norm": 0.14301545917987823, "learning_rate": 9.449539829943404e-05, "loss": 0.019, "step": 57970 }, { "grad_norm": 0.13625454902648926, "learning_rate": 9.449288398268697e-05, "loss": 0.0176, "step": 57980 }, { "grad_norm": 0.11439467966556549, "learning_rate": 9.449036912530935e-05, "loss": 0.0251, "step": 57990 }, { "grad_norm": 0.11692271381616592, "learning_rate": 9.448785372733174e-05, "loss": 0.0169, "step": 58000 }, { "grad_norm": 0.23106259107589722, "learning_rate": 9.448533778878471e-05, "loss": 0.0252, "step": 58010 }, { "grad_norm": 0.11810916662216187, "learning_rate": 9.448282130969882e-05, "loss": 0.022, "step": 58020 }, { "grad_norm": 0.19893896579742432, "learning_rate": 9.448030429010466e-05, "loss": 0.0226, "step": 58030 }, { "grad_norm": 0.14741674065589905, "learning_rate": 9.44777867300328e-05, "loss": 0.0196, "step": 58040 }, { "grad_norm": 0.07172039896249771, "learning_rate": 9.447526862951384e-05, "loss": 0.019, "step": 58050 }, { "grad_norm": 0.177054300904274, "learning_rate": 9.447274998857836e-05, "loss": 0.0172, "step": 58060 }, { "grad_norm": 0.3865766227245331, "learning_rate": 9.447023080725698e-05, "loss": 0.0212, "step": 58070 }, { "grad_norm": 0.24830561876296997, "learning_rate": 9.446771108558033e-05, "loss": 0.0227, "step": 58080 }, { "grad_norm": 0.15898212790489197, "learning_rate": 9.446519082357898e-05, "loss": 0.0251, "step": 58090 }, { "grad_norm": 0.15484824776649475, "learning_rate": 9.44626700212836e-05, "loss": 0.0282, "step": 58100 }, { "grad_norm": 0.13320809602737427, "learning_rate": 9.44601486787248e-05, "loss": 0.0206, "step": 58110 }, { "grad_norm": 0.21006423234939575, "learning_rate": 9.445762679593321e-05, "loss": 0.0242, "step": 58120 }, { "grad_norm": 0.18989089131355286, "learning_rate": 9.445510437293948e-05, "loss": 0.0182, "step": 58130 }, { "grad_norm": 0.45061424374580383, "learning_rate": 9.445258140977426e-05, "loss": 0.0238, "step": 58140 }, { "grad_norm": 0.17060616612434387, "learning_rate": 9.445005790646819e-05, "loss": 0.021, "step": 58150 }, { "grad_norm": 0.36159899830818176, "learning_rate": 9.444753386305196e-05, "loss": 0.0264, "step": 58160 }, { "grad_norm": 0.2441108524799347, "learning_rate": 9.444500927955624e-05, "loss": 0.0265, "step": 58170 }, { "grad_norm": 0.2789059579372406, "learning_rate": 9.444248415601168e-05, "loss": 0.0256, "step": 58180 }, { "grad_norm": 0.192763090133667, "learning_rate": 9.443995849244899e-05, "loss": 0.0216, "step": 58190 }, { "grad_norm": 0.177365243434906, "learning_rate": 9.443743228889883e-05, "loss": 0.0265, "step": 58200 }, { "grad_norm": 0.2016001045703888, "learning_rate": 9.443490554539191e-05, "loss": 0.0172, "step": 58210 }, { "grad_norm": 0.19334259629249573, "learning_rate": 9.443237826195897e-05, "loss": 0.0208, "step": 58220 }, { "grad_norm": 0.20014964044094086, "learning_rate": 9.442985043863064e-05, "loss": 0.0241, "step": 58230 }, { "grad_norm": 0.5035596489906311, "learning_rate": 9.44273220754377e-05, "loss": 0.0263, "step": 58240 }, { "grad_norm": 0.19095036387443542, "learning_rate": 9.442479317241084e-05, "loss": 0.0215, "step": 58250 }, { "grad_norm": 0.1827237457036972, "learning_rate": 9.44222637295808e-05, "loss": 0.0251, "step": 58260 }, { "grad_norm": 0.12878483533859253, "learning_rate": 9.441973374697833e-05, "loss": 0.0235, "step": 58270 }, { "grad_norm": 0.5453417897224426, "learning_rate": 9.441720322463414e-05, "loss": 0.0196, "step": 58280 }, { "grad_norm": 0.1953306347131729, "learning_rate": 9.4414672162579e-05, "loss": 0.0223, "step": 58290 }, { "grad_norm": 0.10234871506690979, "learning_rate": 9.441214056084365e-05, "loss": 0.0176, "step": 58300 }, { "grad_norm": 0.2838640511035919, "learning_rate": 9.440960841945887e-05, "loss": 0.0219, "step": 58310 }, { "grad_norm": 0.2446010559797287, "learning_rate": 9.440707573845542e-05, "loss": 0.0198, "step": 58320 }, { "grad_norm": 0.26250210404396057, "learning_rate": 9.440454251786406e-05, "loss": 0.0205, "step": 58330 }, { "grad_norm": 0.19003887474536896, "learning_rate": 9.44020087577156e-05, "loss": 0.0161, "step": 58340 }, { "grad_norm": 0.16064271330833435, "learning_rate": 9.439947445804081e-05, "loss": 0.0162, "step": 58350 }, { "grad_norm": 0.13422240316867828, "learning_rate": 9.439693961887046e-05, "loss": 0.0165, "step": 58360 }, { "grad_norm": 0.2009613811969757, "learning_rate": 9.43944042402354e-05, "loss": 0.0229, "step": 58370 }, { "grad_norm": 0.463421493768692, "learning_rate": 9.43918683221664e-05, "loss": 0.0257, "step": 58380 }, { "grad_norm": 0.15837545692920685, "learning_rate": 9.43893318646943e-05, "loss": 0.0249, "step": 58390 }, { "grad_norm": 0.1869589388370514, "learning_rate": 9.43867948678499e-05, "loss": 0.0185, "step": 58400 }, { "grad_norm": 0.15547870099544525, "learning_rate": 9.438425733166404e-05, "loss": 0.0227, "step": 58410 }, { "grad_norm": 0.2595999240875244, "learning_rate": 9.438171925616754e-05, "loss": 0.0302, "step": 58420 }, { "grad_norm": 0.1722559779882431, "learning_rate": 9.437918064139126e-05, "loss": 0.023, "step": 58430 }, { "grad_norm": 0.11849753558635712, "learning_rate": 9.437664148736603e-05, "loss": 0.0246, "step": 58440 }, { "grad_norm": 0.13695038855075836, "learning_rate": 9.437410179412269e-05, "loss": 0.0211, "step": 58450 }, { "grad_norm": 0.20395752787590027, "learning_rate": 9.437156156169213e-05, "loss": 0.0197, "step": 58460 }, { "grad_norm": 0.13046430051326752, "learning_rate": 9.436902079010521e-05, "loss": 0.0252, "step": 58470 }, { "grad_norm": 0.14277389645576477, "learning_rate": 9.436647947939279e-05, "loss": 0.0248, "step": 58480 }, { "grad_norm": 0.18138858675956726, "learning_rate": 9.436393762958575e-05, "loss": 0.0312, "step": 58490 }, { "grad_norm": 0.24824665486812592, "learning_rate": 9.4361395240715e-05, "loss": 0.024, "step": 58500 }, { "grad_norm": 0.24472688138484955, "learning_rate": 9.435885231281139e-05, "loss": 0.0261, "step": 58510 }, { "grad_norm": 0.19049149751663208, "learning_rate": 9.435630884590586e-05, "loss": 0.0246, "step": 58520 }, { "grad_norm": 0.3186527490615845, "learning_rate": 9.435376484002928e-05, "loss": 0.0224, "step": 58530 }, { "grad_norm": 0.11159194260835648, "learning_rate": 9.435122029521259e-05, "loss": 0.0211, "step": 58540 }, { "grad_norm": 0.20553554594516754, "learning_rate": 9.43486752114867e-05, "loss": 0.0221, "step": 58550 }, { "grad_norm": 0.38343578577041626, "learning_rate": 9.434612958888253e-05, "loss": 0.0212, "step": 58560 }, { "grad_norm": 0.5232278108596802, "learning_rate": 9.434358342743103e-05, "loss": 0.0175, "step": 58570 }, { "grad_norm": 0.19886375963687897, "learning_rate": 9.434103672716311e-05, "loss": 0.0214, "step": 58580 }, { "grad_norm": 0.25102725625038147, "learning_rate": 9.433848948810973e-05, "loss": 0.0206, "step": 58590 }, { "grad_norm": 0.13439346849918365, "learning_rate": 9.433594171030184e-05, "loss": 0.0266, "step": 58600 }, { "grad_norm": 0.16675856709480286, "learning_rate": 9.43333933937704e-05, "loss": 0.0236, "step": 58610 }, { "grad_norm": 0.17430755496025085, "learning_rate": 9.433084453854638e-05, "loss": 0.0229, "step": 58620 }, { "grad_norm": 0.18935184180736542, "learning_rate": 9.432829514466074e-05, "loss": 0.0194, "step": 58630 }, { "grad_norm": 0.16649945080280304, "learning_rate": 9.432574521214445e-05, "loss": 0.0178, "step": 58640 }, { "grad_norm": 0.1588820517063141, "learning_rate": 9.432319474102852e-05, "loss": 0.0141, "step": 58650 }, { "grad_norm": 0.08818025887012482, "learning_rate": 9.432064373134392e-05, "loss": 0.0173, "step": 58660 }, { "grad_norm": 0.10921107232570648, "learning_rate": 9.431809218312166e-05, "loss": 0.0151, "step": 58670 }, { "grad_norm": 0.11662967503070831, "learning_rate": 9.431554009639272e-05, "loss": 0.0202, "step": 58680 }, { "grad_norm": 1.2389881610870361, "learning_rate": 9.431298747118814e-05, "loss": 0.0216, "step": 58690 }, { "grad_norm": 0.13768713176250458, "learning_rate": 9.431043430753892e-05, "loss": 0.0171, "step": 58700 }, { "grad_norm": 0.13095033168792725, "learning_rate": 9.430788060547609e-05, "loss": 0.0263, "step": 58710 }, { "grad_norm": 0.1619904637336731, "learning_rate": 9.430532636503068e-05, "loss": 0.0209, "step": 58720 }, { "grad_norm": 0.17405961453914642, "learning_rate": 9.43027715862337e-05, "loss": 0.0231, "step": 58730 }, { "grad_norm": 0.19077889621257782, "learning_rate": 9.430021626911625e-05, "loss": 0.0247, "step": 58740 }, { "grad_norm": 1.0446726083755493, "learning_rate": 9.429766041370934e-05, "loss": 0.0277, "step": 58750 }, { "grad_norm": 0.2487526834011078, "learning_rate": 9.429510402004401e-05, "loss": 0.0252, "step": 58760 }, { "grad_norm": 0.8065372109413147, "learning_rate": 9.429254708815136e-05, "loss": 0.022, "step": 58770 }, { "grad_norm": 0.08365184813737869, "learning_rate": 9.428998961806244e-05, "loss": 0.024, "step": 58780 }, { "grad_norm": 0.14920054376125336, "learning_rate": 9.428743160980833e-05, "loss": 0.0173, "step": 58790 }, { "grad_norm": 0.14725936949253082, "learning_rate": 9.428487306342011e-05, "loss": 0.0209, "step": 58800 }, { "grad_norm": 0.38438254594802856, "learning_rate": 9.428231397892888e-05, "loss": 0.0444, "step": 58810 }, { "grad_norm": 0.20683659613132477, "learning_rate": 9.427975435636571e-05, "loss": 0.0334, "step": 58820 }, { "grad_norm": 0.13044017553329468, "learning_rate": 9.427719419576173e-05, "loss": 0.0288, "step": 58830 }, { "grad_norm": 0.1365404576063156, "learning_rate": 9.427463349714804e-05, "loss": 0.0328, "step": 58840 }, { "grad_norm": 0.5061400532722473, "learning_rate": 9.427207226055574e-05, "loss": 0.0178, "step": 58850 }, { "grad_norm": 0.16901510953903198, "learning_rate": 9.426951048601595e-05, "loss": 0.0229, "step": 58860 }, { "grad_norm": 0.18649180233478546, "learning_rate": 9.426694817355983e-05, "loss": 0.0258, "step": 58870 }, { "grad_norm": 0.24148935079574585, "learning_rate": 9.42643853232185e-05, "loss": 0.0256, "step": 58880 }, { "grad_norm": 0.10263705253601074, "learning_rate": 9.426182193502309e-05, "loss": 0.0173, "step": 58890 }, { "grad_norm": 0.18575415015220642, "learning_rate": 9.425925800900475e-05, "loss": 0.0214, "step": 58900 }, { "grad_norm": 0.2827199697494507, "learning_rate": 9.425669354519464e-05, "loss": 0.0318, "step": 58910 }, { "grad_norm": 0.12047106772661209, "learning_rate": 9.42541285436239e-05, "loss": 0.0199, "step": 58920 }, { "grad_norm": 0.20427967607975006, "learning_rate": 9.425156300432373e-05, "loss": 0.0263, "step": 58930 }, { "grad_norm": 0.17002150416374207, "learning_rate": 9.424899692732529e-05, "loss": 0.0225, "step": 58940 }, { "grad_norm": 0.17311207950115204, "learning_rate": 9.424643031265976e-05, "loss": 0.0279, "step": 58950 }, { "grad_norm": 0.24242189526557922, "learning_rate": 9.424386316035833e-05, "loss": 0.0326, "step": 58960 }, { "grad_norm": 0.18227368593215942, "learning_rate": 9.424129547045218e-05, "loss": 0.0215, "step": 58970 }, { "grad_norm": 0.19022125005722046, "learning_rate": 9.423872724297252e-05, "loss": 0.0248, "step": 58980 }, { "grad_norm": 0.20222872495651245, "learning_rate": 9.423615847795057e-05, "loss": 0.0222, "step": 58990 }, { "grad_norm": 0.17493639886379242, "learning_rate": 9.423358917541751e-05, "loss": 0.0179, "step": 59000 }, { "grad_norm": 0.16891391575336456, "learning_rate": 9.423101933540459e-05, "loss": 0.0249, "step": 59010 }, { "grad_norm": 0.4962776005268097, "learning_rate": 9.422844895794301e-05, "loss": 0.0269, "step": 59020 }, { "grad_norm": 0.3907283842563629, "learning_rate": 9.422587804306404e-05, "loss": 0.0328, "step": 59030 }, { "grad_norm": 0.17766882479190826, "learning_rate": 9.422330659079887e-05, "loss": 0.0256, "step": 59040 }, { "grad_norm": 0.1669415682554245, "learning_rate": 9.42207346011788e-05, "loss": 0.0245, "step": 59050 }, { "grad_norm": 0.3362230062484741, "learning_rate": 9.421816207423503e-05, "loss": 0.0244, "step": 59060 }, { "grad_norm": 0.17196768522262573, "learning_rate": 9.421558900999885e-05, "loss": 0.0221, "step": 59070 }, { "grad_norm": 0.14063377678394318, "learning_rate": 9.421301540850152e-05, "loss": 0.0232, "step": 59080 }, { "grad_norm": 0.20921918749809265, "learning_rate": 9.42104412697743e-05, "loss": 0.0247, "step": 59090 }, { "grad_norm": 0.2330334633588791, "learning_rate": 9.420786659384848e-05, "loss": 0.0264, "step": 59100 }, { "grad_norm": 0.14591601490974426, "learning_rate": 9.420529138075535e-05, "loss": 0.0238, "step": 59110 }, { "grad_norm": 0.1976509839296341, "learning_rate": 9.420271563052618e-05, "loss": 0.021, "step": 59120 }, { "grad_norm": 0.1494802087545395, "learning_rate": 9.42001393431923e-05, "loss": 0.0199, "step": 59130 }, { "grad_norm": 0.1338866502046585, "learning_rate": 9.419756251878498e-05, "loss": 0.0161, "step": 59140 }, { "grad_norm": 0.2023003101348877, "learning_rate": 9.419498515733554e-05, "loss": 0.0206, "step": 59150 }, { "grad_norm": 0.1561623215675354, "learning_rate": 9.41924072588753e-05, "loss": 0.0216, "step": 59160 }, { "grad_norm": 0.20756468176841736, "learning_rate": 9.418982882343561e-05, "loss": 0.0199, "step": 59170 }, { "grad_norm": 0.16514047980308533, "learning_rate": 9.418724985104777e-05, "loss": 0.0166, "step": 59180 }, { "grad_norm": 0.11821231991052628, "learning_rate": 9.418467034174312e-05, "loss": 0.02, "step": 59190 }, { "grad_norm": 0.13706323504447937, "learning_rate": 9.4182090295553e-05, "loss": 0.0172, "step": 59200 }, { "grad_norm": 0.1734621375799179, "learning_rate": 9.417950971250878e-05, "loss": 0.0217, "step": 59210 }, { "grad_norm": 0.35045647621154785, "learning_rate": 9.417692859264181e-05, "loss": 0.0217, "step": 59220 }, { "grad_norm": 0.13074831664562225, "learning_rate": 9.417434693598344e-05, "loss": 0.0206, "step": 59230 }, { "grad_norm": 0.2754362225532532, "learning_rate": 9.417176474256504e-05, "loss": 0.0217, "step": 59240 }, { "grad_norm": 0.1377735733985901, "learning_rate": 9.4169182012418e-05, "loss": 0.0211, "step": 59250 }, { "grad_norm": 0.1215018779039383, "learning_rate": 9.41665987455737e-05, "loss": 0.024, "step": 59260 }, { "grad_norm": 0.29376471042633057, "learning_rate": 9.416401494206353e-05, "loss": 0.0207, "step": 59270 }, { "grad_norm": 0.18685968220233917, "learning_rate": 9.416143060191886e-05, "loss": 0.0162, "step": 59280 }, { "grad_norm": 0.7492620944976807, "learning_rate": 9.415884572517113e-05, "loss": 0.022, "step": 59290 }, { "grad_norm": 0.18364164233207703, "learning_rate": 9.415626031185172e-05, "loss": 0.0231, "step": 59300 }, { "grad_norm": 0.2627500593662262, "learning_rate": 9.415367436199207e-05, "loss": 0.0265, "step": 59310 }, { "grad_norm": 0.2636328637599945, "learning_rate": 9.41510878756236e-05, "loss": 0.0179, "step": 59320 }, { "grad_norm": 0.14005479216575623, "learning_rate": 9.414850085277769e-05, "loss": 0.0193, "step": 59330 }, { "grad_norm": 0.10104070603847504, "learning_rate": 9.414591329348583e-05, "loss": 0.0174, "step": 59340 }, { "grad_norm": 0.28983038663864136, "learning_rate": 9.414332519777944e-05, "loss": 0.0178, "step": 59350 }, { "grad_norm": 0.16557657718658447, "learning_rate": 9.414073656568998e-05, "loss": 0.0274, "step": 59360 }, { "grad_norm": 0.1579710990190506, "learning_rate": 9.413814739724889e-05, "loss": 0.0168, "step": 59370 }, { "grad_norm": 0.14636006951332092, "learning_rate": 9.413555769248763e-05, "loss": 0.0235, "step": 59380 }, { "grad_norm": 0.15290172398090363, "learning_rate": 9.413296745143769e-05, "loss": 0.0195, "step": 59390 }, { "grad_norm": 0.15079399943351746, "learning_rate": 9.413037667413051e-05, "loss": 0.0219, "step": 59400 }, { "grad_norm": 0.19219070672988892, "learning_rate": 9.412778536059759e-05, "loss": 0.0218, "step": 59410 }, { "grad_norm": 0.12655486166477203, "learning_rate": 9.412519351087041e-05, "loss": 0.0198, "step": 59420 }, { "grad_norm": 0.4680507183074951, "learning_rate": 9.412260112498049e-05, "loss": 0.0232, "step": 59430 }, { "grad_norm": 0.2681507170200348, "learning_rate": 9.412000820295928e-05, "loss": 0.0268, "step": 59440 }, { "grad_norm": 0.27132460474967957, "learning_rate": 9.411741474483833e-05, "loss": 0.0273, "step": 59450 }, { "grad_norm": 0.1434796154499054, "learning_rate": 9.411482075064913e-05, "loss": 0.0186, "step": 59460 }, { "grad_norm": 0.33101919293403625, "learning_rate": 9.411222622042322e-05, "loss": 0.0212, "step": 59470 }, { "grad_norm": 0.1040656641125679, "learning_rate": 9.410963115419209e-05, "loss": 0.0157, "step": 59480 }, { "grad_norm": 0.13632626831531525, "learning_rate": 9.410703555198731e-05, "loss": 0.0203, "step": 59490 }, { "grad_norm": 0.1255250722169876, "learning_rate": 9.410443941384039e-05, "loss": 0.018, "step": 59500 }, { "grad_norm": 0.13089531660079956, "learning_rate": 9.41018427397829e-05, "loss": 0.0179, "step": 59510 }, { "grad_norm": 0.12966451048851013, "learning_rate": 9.40992455298464e-05, "loss": 0.0168, "step": 59520 }, { "grad_norm": 0.15907128155231476, "learning_rate": 9.40966477840624e-05, "loss": 0.0199, "step": 59530 }, { "grad_norm": 0.15620769560337067, "learning_rate": 9.409404950246251e-05, "loss": 0.0193, "step": 59540 }, { "grad_norm": 0.1834365725517273, "learning_rate": 9.409145068507829e-05, "loss": 0.0254, "step": 59550 }, { "grad_norm": 0.19664829969406128, "learning_rate": 9.408885133194131e-05, "loss": 0.0183, "step": 59560 }, { "grad_norm": 0.1724589765071869, "learning_rate": 9.408625144308316e-05, "loss": 0.0287, "step": 59570 }, { "grad_norm": 0.18439437448978424, "learning_rate": 9.408365101853543e-05, "loss": 0.0239, "step": 59580 }, { "grad_norm": 0.10949227213859558, "learning_rate": 9.408105005832972e-05, "loss": 0.0222, "step": 59590 }, { "grad_norm": 0.3731769621372223, "learning_rate": 9.407844856249763e-05, "loss": 0.0213, "step": 59600 }, { "grad_norm": 0.13941334187984467, "learning_rate": 9.407584653107078e-05, "loss": 0.0197, "step": 59610 }, { "grad_norm": 0.1610056608915329, "learning_rate": 9.407324396408077e-05, "loss": 0.0224, "step": 59620 }, { "grad_norm": 0.1199769526720047, "learning_rate": 9.407064086155924e-05, "loss": 0.0166, "step": 59630 }, { "grad_norm": 0.1513604074716568, "learning_rate": 9.40680372235378e-05, "loss": 0.0158, "step": 59640 }, { "grad_norm": 0.30741971731185913, "learning_rate": 9.406543305004812e-05, "loss": 0.0226, "step": 59650 }, { "grad_norm": 0.2551557123661041, "learning_rate": 9.406282834112182e-05, "loss": 0.0239, "step": 59660 }, { "grad_norm": 0.1495988517999649, "learning_rate": 9.406022309679054e-05, "loss": 0.026, "step": 59670 }, { "grad_norm": 0.2343447357416153, "learning_rate": 9.405761731708597e-05, "loss": 0.0291, "step": 59680 }, { "grad_norm": 0.18819659948349, "learning_rate": 9.405501100203974e-05, "loss": 0.0234, "step": 59690 }, { "grad_norm": 0.21247701346874237, "learning_rate": 9.405240415168353e-05, "loss": 0.0239, "step": 59700 }, { "grad_norm": 0.2209630161523819, "learning_rate": 9.404979676604902e-05, "loss": 0.0242, "step": 59710 }, { "grad_norm": 0.5013740062713623, "learning_rate": 9.40471888451679e-05, "loss": 0.0268, "step": 59720 }, { "grad_norm": 0.25634774565696716, "learning_rate": 9.404458038907183e-05, "loss": 0.0301, "step": 59730 }, { "grad_norm": 0.34032705426216125, "learning_rate": 9.404197139779253e-05, "loss": 0.025, "step": 59740 }, { "grad_norm": 0.2033694088459015, "learning_rate": 9.403936187136169e-05, "loss": 0.0212, "step": 59750 }, { "grad_norm": 0.18783989548683167, "learning_rate": 9.403675180981104e-05, "loss": 0.021, "step": 59760 }, { "grad_norm": 0.7486284971237183, "learning_rate": 9.403414121317226e-05, "loss": 0.0292, "step": 59770 }, { "grad_norm": 0.2725285291671753, "learning_rate": 9.403153008147709e-05, "loss": 0.0226, "step": 59780 }, { "grad_norm": 0.17903591692447662, "learning_rate": 9.402891841475728e-05, "loss": 0.0222, "step": 59790 }, { "grad_norm": 0.11313623189926147, "learning_rate": 9.402630621304452e-05, "loss": 0.0191, "step": 59800 }, { "grad_norm": 0.173056960105896, "learning_rate": 9.402369347637057e-05, "loss": 0.0212, "step": 59810 }, { "grad_norm": 0.2392863780260086, "learning_rate": 9.402108020476718e-05, "loss": 0.0251, "step": 59820 }, { "grad_norm": 0.3701034486293793, "learning_rate": 9.401846639826612e-05, "loss": 0.0252, "step": 59830 }, { "grad_norm": 0.14028629660606384, "learning_rate": 9.401585205689912e-05, "loss": 0.0201, "step": 59840 }, { "grad_norm": 0.12482137978076935, "learning_rate": 9.401323718069796e-05, "loss": 0.0169, "step": 59850 }, { "grad_norm": 0.18045203387737274, "learning_rate": 9.401062176969442e-05, "loss": 0.0218, "step": 59860 }, { "grad_norm": 0.1628827005624771, "learning_rate": 9.400800582392027e-05, "loss": 0.0194, "step": 59870 }, { "grad_norm": 0.6342527270317078, "learning_rate": 9.400538934340728e-05, "loss": 0.0239, "step": 59880 }, { "grad_norm": 0.15795940160751343, "learning_rate": 9.400277232818728e-05, "loss": 0.0227, "step": 59890 }, { "grad_norm": 0.29711851477622986, "learning_rate": 9.400015477829205e-05, "loss": 0.0251, "step": 59900 }, { "grad_norm": 0.14319099485874176, "learning_rate": 9.399753669375339e-05, "loss": 0.0186, "step": 59910 }, { "grad_norm": 0.10618603229522705, "learning_rate": 9.399491807460312e-05, "loss": 0.0205, "step": 59920 }, { "grad_norm": 0.35981741547584534, "learning_rate": 9.399229892087306e-05, "loss": 0.0301, "step": 59930 }, { "grad_norm": 0.1867755651473999, "learning_rate": 9.398967923259503e-05, "loss": 0.0151, "step": 59940 }, { "grad_norm": 0.17178118228912354, "learning_rate": 9.398705900980088e-05, "loss": 0.0237, "step": 59950 }, { "grad_norm": 0.2745596170425415, "learning_rate": 9.398443825252241e-05, "loss": 0.0177, "step": 59960 }, { "grad_norm": 0.2583482265472412, "learning_rate": 9.398181696079149e-05, "loss": 0.022, "step": 59970 }, { "grad_norm": 0.13339994847774506, "learning_rate": 9.397919513463998e-05, "loss": 0.0176, "step": 59980 }, { "grad_norm": 0.2341676652431488, "learning_rate": 9.397657277409971e-05, "loss": 0.0221, "step": 59990 }, { "grad_norm": 0.20734967291355133, "learning_rate": 9.397394987920257e-05, "loss": 0.0213, "step": 60000 }, { "grad_norm": 0.21061457693576813, "learning_rate": 9.397132644998042e-05, "loss": 0.019, "step": 60010 }, { "grad_norm": 0.23399481177330017, "learning_rate": 9.396870248646515e-05, "loss": 0.0171, "step": 60020 }, { "grad_norm": 0.2653924524784088, "learning_rate": 9.396607798868861e-05, "loss": 0.0219, "step": 60030 }, { "grad_norm": 0.23816139996051788, "learning_rate": 9.396345295668272e-05, "loss": 0.0267, "step": 60040 }, { "grad_norm": 0.12171458452939987, "learning_rate": 9.396082739047937e-05, "loss": 0.0247, "step": 60050 }, { "grad_norm": 0.15132594108581543, "learning_rate": 9.395820129011046e-05, "loss": 0.0183, "step": 60060 }, { "grad_norm": 0.1965581774711609, "learning_rate": 9.39555746556079e-05, "loss": 0.0225, "step": 60070 }, { "grad_norm": 0.17135761678218842, "learning_rate": 9.395294748700361e-05, "loss": 0.0178, "step": 60080 }, { "grad_norm": 0.20032382011413574, "learning_rate": 9.395031978432952e-05, "loss": 0.0209, "step": 60090 }, { "grad_norm": 0.15462960302829742, "learning_rate": 9.394769154761754e-05, "loss": 0.0325, "step": 60100 }, { "grad_norm": 0.15038233995437622, "learning_rate": 9.39450627768996e-05, "loss": 0.0257, "step": 60110 }, { "grad_norm": 0.16026484966278076, "learning_rate": 9.394243347220766e-05, "loss": 0.0195, "step": 60120 }, { "grad_norm": 0.1278619021177292, "learning_rate": 9.393980363357368e-05, "loss": 0.0161, "step": 60130 }, { "grad_norm": 0.14391878247261047, "learning_rate": 9.393717326102959e-05, "loss": 0.0217, "step": 60140 }, { "grad_norm": 0.12982302904129028, "learning_rate": 9.393454235460736e-05, "loss": 0.0202, "step": 60150 }, { "grad_norm": 0.2455216497182846, "learning_rate": 9.393191091433896e-05, "loss": 0.0333, "step": 60160 }, { "grad_norm": 0.1688448041677475, "learning_rate": 9.392927894025637e-05, "loss": 0.0219, "step": 60170 }, { "grad_norm": 0.10852355509996414, "learning_rate": 9.392664643239155e-05, "loss": 0.027, "step": 60180 }, { "grad_norm": 0.1375061273574829, "learning_rate": 9.392401339077652e-05, "loss": 0.0214, "step": 60190 }, { "grad_norm": 0.18643838167190552, "learning_rate": 9.392137981544325e-05, "loss": 0.0214, "step": 60200 }, { "grad_norm": 0.14305385947227478, "learning_rate": 9.391874570642375e-05, "loss": 0.026, "step": 60210 }, { "grad_norm": 0.23122647404670715, "learning_rate": 9.391611106375002e-05, "loss": 0.0186, "step": 60220 }, { "grad_norm": 0.16870243847370148, "learning_rate": 9.391347588745407e-05, "loss": 0.0282, "step": 60230 }, { "grad_norm": 0.15674445033073425, "learning_rate": 9.391084017756794e-05, "loss": 0.0172, "step": 60240 }, { "grad_norm": 0.21362748742103577, "learning_rate": 9.390820393412363e-05, "loss": 0.0239, "step": 60250 }, { "grad_norm": 0.2694045603275299, "learning_rate": 9.39055671571532e-05, "loss": 0.0222, "step": 60260 }, { "grad_norm": 0.21708323061466217, "learning_rate": 9.390292984668867e-05, "loss": 0.0245, "step": 60270 }, { "grad_norm": 0.29791560769081116, "learning_rate": 9.390029200276208e-05, "loss": 0.0206, "step": 60280 }, { "grad_norm": 0.16200800240039825, "learning_rate": 9.389765362540551e-05, "loss": 0.0179, "step": 60290 }, { "grad_norm": 0.17658624053001404, "learning_rate": 9.389501471465099e-05, "loss": 0.0171, "step": 60300 }, { "grad_norm": 0.9826251864433289, "learning_rate": 9.38923752705306e-05, "loss": 0.0255, "step": 60310 }, { "grad_norm": 0.4896315932273865, "learning_rate": 9.388973529307641e-05, "loss": 0.0239, "step": 60320 }, { "grad_norm": 0.16031847894191742, "learning_rate": 9.388709478232049e-05, "loss": 0.0219, "step": 60330 }, { "grad_norm": 0.1525508314371109, "learning_rate": 9.388445373829495e-05, "loss": 0.019, "step": 60340 }, { "grad_norm": 0.10994391143321991, "learning_rate": 9.388181216103185e-05, "loss": 0.0254, "step": 60350 }, { "grad_norm": 0.1168757826089859, "learning_rate": 9.38791700505633e-05, "loss": 0.0163, "step": 60360 }, { "grad_norm": 0.08203878998756409, "learning_rate": 9.38765274069214e-05, "loss": 0.0199, "step": 60370 }, { "grad_norm": 0.21114923059940338, "learning_rate": 9.38738842301383e-05, "loss": 0.0187, "step": 60380 }, { "grad_norm": 0.14655302464962006, "learning_rate": 9.387124052024605e-05, "loss": 0.0185, "step": 60390 }, { "grad_norm": 0.1733410358428955, "learning_rate": 9.386859627727682e-05, "loss": 0.0203, "step": 60400 }, { "grad_norm": 0.10718060284852982, "learning_rate": 9.386595150126272e-05, "loss": 0.0189, "step": 60410 }, { "grad_norm": 0.19767946004867554, "learning_rate": 9.386330619223589e-05, "loss": 0.0176, "step": 60420 }, { "grad_norm": 0.39346688985824585, "learning_rate": 9.386066035022848e-05, "loss": 0.0266, "step": 60430 }, { "grad_norm": 0.3401245176792145, "learning_rate": 9.385801397527263e-05, "loss": 0.0239, "step": 60440 }, { "grad_norm": 0.21630972623825073, "learning_rate": 9.385536706740051e-05, "loss": 0.0256, "step": 60450 }, { "grad_norm": 0.18209627270698547, "learning_rate": 9.385271962664427e-05, "loss": 0.0303, "step": 60460 }, { "grad_norm": 0.2996249496936798, "learning_rate": 9.385007165303609e-05, "loss": 0.0202, "step": 60470 }, { "grad_norm": 0.1318511962890625, "learning_rate": 9.384742314660812e-05, "loss": 0.0152, "step": 60480 }, { "grad_norm": 0.17925536632537842, "learning_rate": 9.384477410739257e-05, "loss": 0.0298, "step": 60490 }, { "grad_norm": 0.157028928399086, "learning_rate": 9.384212453542162e-05, "loss": 0.0212, "step": 60500 }, { "grad_norm": 0.23618443310260773, "learning_rate": 9.383947443072747e-05, "loss": 0.0154, "step": 60510 }, { "grad_norm": 0.11160893738269806, "learning_rate": 9.383682379334231e-05, "loss": 0.0204, "step": 60520 }, { "grad_norm": 0.17920036613941193, "learning_rate": 9.383417262329834e-05, "loss": 0.0221, "step": 60530 }, { "grad_norm": 0.15063731372356415, "learning_rate": 9.38315209206278e-05, "loss": 0.0163, "step": 60540 }, { "grad_norm": 0.25131145119667053, "learning_rate": 9.38288686853629e-05, "loss": 0.0201, "step": 60550 }, { "grad_norm": 0.11561650782823563, "learning_rate": 9.382621591753584e-05, "loss": 0.0211, "step": 60560 }, { "grad_norm": 0.12462697923183441, "learning_rate": 9.38235626171789e-05, "loss": 0.0169, "step": 60570 }, { "grad_norm": 0.14655931293964386, "learning_rate": 9.38209087843243e-05, "loss": 0.0205, "step": 60580 }, { "grad_norm": 0.15797752141952515, "learning_rate": 9.381825441900427e-05, "loss": 0.0235, "step": 60590 }, { "grad_norm": 0.13218176364898682, "learning_rate": 9.381559952125108e-05, "loss": 0.0201, "step": 60600 }, { "grad_norm": 0.15486450493335724, "learning_rate": 9.3812944091097e-05, "loss": 0.0262, "step": 60610 }, { "grad_norm": 0.14759887754917145, "learning_rate": 9.381028812857426e-05, "loss": 0.0153, "step": 60620 }, { "grad_norm": 0.1550043225288391, "learning_rate": 9.380763163371517e-05, "loss": 0.0178, "step": 60630 }, { "grad_norm": 0.4344606399536133, "learning_rate": 9.3804974606552e-05, "loss": 0.0216, "step": 60640 }, { "grad_norm": 0.12947843968868256, "learning_rate": 9.380231704711703e-05, "loss": 0.0207, "step": 60650 }, { "grad_norm": 0.21167217195034027, "learning_rate": 9.379965895544254e-05, "loss": 0.0185, "step": 60660 }, { "grad_norm": 0.27498701214790344, "learning_rate": 9.379700033156085e-05, "loss": 0.0215, "step": 60670 }, { "grad_norm": 0.14906394481658936, "learning_rate": 9.379434117550425e-05, "loss": 0.017, "step": 60680 }, { "grad_norm": 0.15341368317604065, "learning_rate": 9.379168148730506e-05, "loss": 0.0212, "step": 60690 }, { "grad_norm": 0.11947844177484512, "learning_rate": 9.378902126699558e-05, "loss": 0.0211, "step": 60700 }, { "grad_norm": 0.15498213469982147, "learning_rate": 9.378636051460816e-05, "loss": 0.0205, "step": 60710 }, { "grad_norm": 0.14609390497207642, "learning_rate": 9.378369923017511e-05, "loss": 0.0204, "step": 60720 }, { "grad_norm": 0.17713558673858643, "learning_rate": 9.378103741372879e-05, "loss": 0.0187, "step": 60730 }, { "grad_norm": 0.22378985583782196, "learning_rate": 9.377837506530151e-05, "loss": 0.0184, "step": 60740 }, { "grad_norm": 0.11856130510568619, "learning_rate": 9.377571218492567e-05, "loss": 0.028, "step": 60750 }, { "grad_norm": 0.10689736902713776, "learning_rate": 9.377304877263358e-05, "loss": 0.0224, "step": 60760 }, { "grad_norm": 0.18650078773498535, "learning_rate": 9.37703848284576e-05, "loss": 0.023, "step": 60770 }, { "grad_norm": 0.2266194075345993, "learning_rate": 9.376772035243013e-05, "loss": 0.0271, "step": 60780 }, { "grad_norm": 0.2098860740661621, "learning_rate": 9.376505534458355e-05, "loss": 0.0186, "step": 60790 }, { "grad_norm": 0.16325484216213226, "learning_rate": 9.37623898049502e-05, "loss": 0.0191, "step": 60800 }, { "grad_norm": 0.14681001007556915, "learning_rate": 9.375972373356252e-05, "loss": 0.027, "step": 60810 }, { "grad_norm": 0.15618003904819489, "learning_rate": 9.375705713045287e-05, "loss": 0.0166, "step": 60820 }, { "grad_norm": 0.12179335951805115, "learning_rate": 9.375438999565366e-05, "loss": 0.0219, "step": 60830 }, { "grad_norm": 0.14963406324386597, "learning_rate": 9.375172232919731e-05, "loss": 0.0223, "step": 60840 }, { "grad_norm": 0.18326331675052643, "learning_rate": 9.374905413111622e-05, "loss": 0.0243, "step": 60850 }, { "grad_norm": 0.1677376627922058, "learning_rate": 9.374638540144281e-05, "loss": 0.0199, "step": 60860 }, { "grad_norm": 0.20885829627513885, "learning_rate": 9.374371614020951e-05, "loss": 0.0175, "step": 60870 }, { "grad_norm": 0.22729864716529846, "learning_rate": 9.374104634744878e-05, "loss": 0.024, "step": 60880 }, { "grad_norm": 0.19127972424030304, "learning_rate": 9.373837602319301e-05, "loss": 0.019, "step": 60890 }, { "grad_norm": 0.13320349156856537, "learning_rate": 9.37357051674747e-05, "loss": 0.0134, "step": 60900 }, { "grad_norm": 0.22961148619651794, "learning_rate": 9.373303378032628e-05, "loss": 0.0178, "step": 60910 }, { "grad_norm": 0.12531602382659912, "learning_rate": 9.373036186178019e-05, "loss": 0.0193, "step": 60920 }, { "grad_norm": 0.22061854600906372, "learning_rate": 9.372768941186891e-05, "loss": 0.0303, "step": 60930 }, { "grad_norm": 0.10218196362257004, "learning_rate": 9.372501643062494e-05, "loss": 0.0209, "step": 60940 }, { "grad_norm": 0.2123701274394989, "learning_rate": 9.372234291808073e-05, "loss": 0.0179, "step": 60950 }, { "grad_norm": 0.18562361598014832, "learning_rate": 9.371966887426877e-05, "loss": 0.0179, "step": 60960 }, { "grad_norm": 0.17448267340660095, "learning_rate": 9.371699429922157e-05, "loss": 0.0192, "step": 60970 }, { "grad_norm": 0.18152642250061035, "learning_rate": 9.371431919297158e-05, "loss": 0.0271, "step": 60980 }, { "grad_norm": 0.15533128380775452, "learning_rate": 9.371164355555135e-05, "loss": 0.0243, "step": 60990 }, { "grad_norm": 0.16296204924583435, "learning_rate": 9.370896738699339e-05, "loss": 0.023, "step": 61000 }, { "grad_norm": 0.36942723393440247, "learning_rate": 9.370629068733022e-05, "loss": 0.0283, "step": 61010 }, { "grad_norm": 0.11749851703643799, "learning_rate": 9.370361345659433e-05, "loss": 0.0244, "step": 61020 }, { "grad_norm": 0.18790194392204285, "learning_rate": 9.370093569481829e-05, "loss": 0.0205, "step": 61030 }, { "grad_norm": 0.23033598065376282, "learning_rate": 9.369825740203462e-05, "loss": 0.0257, "step": 61040 }, { "grad_norm": 0.14468219876289368, "learning_rate": 9.369557857827584e-05, "loss": 0.016, "step": 61050 }, { "grad_norm": 0.25677376985549927, "learning_rate": 9.369289922357455e-05, "loss": 0.032, "step": 61060 }, { "grad_norm": 0.18085452914237976, "learning_rate": 9.369021933796327e-05, "loss": 0.0153, "step": 61070 }, { "grad_norm": 0.2424948364496231, "learning_rate": 9.368753892147458e-05, "loss": 0.0167, "step": 61080 }, { "grad_norm": 0.20923247933387756, "learning_rate": 9.368485797414105e-05, "loss": 0.0172, "step": 61090 }, { "grad_norm": 0.13916891813278198, "learning_rate": 9.368217649599524e-05, "loss": 0.0166, "step": 61100 }, { "grad_norm": 0.10056197643280029, "learning_rate": 9.367949448706974e-05, "loss": 0.0196, "step": 61110 }, { "grad_norm": 0.21853706240653992, "learning_rate": 9.367681194739715e-05, "loss": 0.0177, "step": 61120 }, { "grad_norm": 0.2187555879354477, "learning_rate": 9.367412887701005e-05, "loss": 0.0219, "step": 61130 }, { "grad_norm": 0.1729695349931717, "learning_rate": 9.367144527594106e-05, "loss": 0.0275, "step": 61140 }, { "grad_norm": 0.10589683055877686, "learning_rate": 9.366876114422278e-05, "loss": 0.0183, "step": 61150 }, { "grad_norm": 0.15353268384933472, "learning_rate": 9.36660764818878e-05, "loss": 0.0272, "step": 61160 }, { "grad_norm": 0.1853160262107849, "learning_rate": 9.366339128896878e-05, "loss": 0.019, "step": 61170 }, { "grad_norm": 0.1367763727903366, "learning_rate": 9.366070556549833e-05, "loss": 0.0214, "step": 61180 }, { "grad_norm": 0.17852665483951569, "learning_rate": 9.365801931150909e-05, "loss": 0.0247, "step": 61190 }, { "grad_norm": 0.2508077025413513, "learning_rate": 9.365533252703368e-05, "loss": 0.0201, "step": 61200 }, { "grad_norm": 0.547030508518219, "learning_rate": 9.365264521210478e-05, "loss": 0.0244, "step": 61210 }, { "grad_norm": 0.12140636146068573, "learning_rate": 9.3649957366755e-05, "loss": 0.018, "step": 61220 }, { "grad_norm": 0.12603223323822021, "learning_rate": 9.364726899101703e-05, "loss": 0.0273, "step": 61230 }, { "grad_norm": 0.18873964250087738, "learning_rate": 9.364458008492356e-05, "loss": 0.0258, "step": 61240 }, { "grad_norm": 0.16940023005008698, "learning_rate": 9.364189064850721e-05, "loss": 0.0184, "step": 61250 }, { "grad_norm": 0.27075767517089844, "learning_rate": 9.36392006818007e-05, "loss": 0.0225, "step": 61260 }, { "grad_norm": 0.1856098622083664, "learning_rate": 9.363651018483667e-05, "loss": 0.0243, "step": 61270 }, { "grad_norm": 0.16156744956970215, "learning_rate": 9.363381915764788e-05, "loss": 0.0214, "step": 61280 }, { "grad_norm": 0.19157704710960388, "learning_rate": 9.363112760026698e-05, "loss": 0.0269, "step": 61290 }, { "grad_norm": 0.17035415768623352, "learning_rate": 9.362843551272666e-05, "loss": 0.0207, "step": 61300 }, { "grad_norm": 0.15616488456726074, "learning_rate": 9.362574289505968e-05, "loss": 0.0161, "step": 61310 }, { "grad_norm": 0.13201864063739777, "learning_rate": 9.362304974729874e-05, "loss": 0.0174, "step": 61320 }, { "grad_norm": 0.21249452233314514, "learning_rate": 9.362035606947654e-05, "loss": 0.0264, "step": 61330 }, { "grad_norm": 0.15558761358261108, "learning_rate": 9.361766186162584e-05, "loss": 0.0214, "step": 61340 }, { "grad_norm": 0.3202587962150574, "learning_rate": 9.361496712377935e-05, "loss": 0.0238, "step": 61350 }, { "grad_norm": 0.27501529455184937, "learning_rate": 9.361227185596985e-05, "loss": 0.0189, "step": 61360 }, { "grad_norm": 0.20231279730796814, "learning_rate": 9.360957605823007e-05, "loss": 0.0164, "step": 61370 }, { "grad_norm": 0.2965502440929413, "learning_rate": 9.360687973059275e-05, "loss": 0.0184, "step": 61380 }, { "grad_norm": 0.1748504638671875, "learning_rate": 9.360418287309067e-05, "loss": 0.0224, "step": 61390 }, { "grad_norm": 0.10651702433824539, "learning_rate": 9.360148548575662e-05, "loss": 0.0225, "step": 61400 }, { "grad_norm": 0.13080285489559174, "learning_rate": 9.359878756862333e-05, "loss": 0.0247, "step": 61410 }, { "grad_norm": 0.20323757827281952, "learning_rate": 9.359608912172362e-05, "loss": 0.026, "step": 61420 }, { "grad_norm": 0.36020660400390625, "learning_rate": 9.359339014509028e-05, "loss": 0.0189, "step": 61430 }, { "grad_norm": 0.18665358424186707, "learning_rate": 9.359069063875607e-05, "loss": 0.021, "step": 61440 }, { "grad_norm": 0.19331936538219452, "learning_rate": 9.358799060275382e-05, "loss": 0.0168, "step": 61450 }, { "grad_norm": 0.24016034603118896, "learning_rate": 9.358529003711632e-05, "loss": 0.0147, "step": 61460 }, { "grad_norm": 0.12496428936719894, "learning_rate": 9.35825889418764e-05, "loss": 0.0219, "step": 61470 }, { "grad_norm": 0.15506283938884735, "learning_rate": 9.357988731706689e-05, "loss": 0.0233, "step": 61480 }, { "grad_norm": 0.26667940616607666, "learning_rate": 9.357718516272058e-05, "loss": 0.0224, "step": 61490 }, { "grad_norm": 0.3416602313518524, "learning_rate": 9.357448247887036e-05, "loss": 0.0271, "step": 61500 }, { "grad_norm": 0.3102073073387146, "learning_rate": 9.3571779265549e-05, "loss": 0.0271, "step": 61510 }, { "grad_norm": 0.15513242781162262, "learning_rate": 9.35690755227894e-05, "loss": 0.0203, "step": 61520 }, { "grad_norm": 0.12454777956008911, "learning_rate": 9.356637125062439e-05, "loss": 0.0163, "step": 61530 }, { "grad_norm": 0.7295777797698975, "learning_rate": 9.356366644908685e-05, "loss": 0.021, "step": 61540 }, { "grad_norm": 0.13021427392959595, "learning_rate": 9.356096111820963e-05, "loss": 0.0261, "step": 61550 }, { "grad_norm": 0.23797635734081268, "learning_rate": 9.355825525802559e-05, "loss": 0.0312, "step": 61560 }, { "grad_norm": 0.14037536084651947, "learning_rate": 9.355554886856762e-05, "loss": 0.0155, "step": 61570 }, { "grad_norm": 0.1392320841550827, "learning_rate": 9.355284194986862e-05, "loss": 0.0165, "step": 61580 }, { "grad_norm": 0.15784576535224915, "learning_rate": 9.355013450196146e-05, "loss": 0.0237, "step": 61590 }, { "grad_norm": 0.11001148074865341, "learning_rate": 9.354742652487907e-05, "loss": 0.0233, "step": 61600 }, { "grad_norm": 0.13116121292114258, "learning_rate": 9.35447180186543e-05, "loss": 0.0224, "step": 61610 }, { "grad_norm": 0.2469598650932312, "learning_rate": 9.354200898332012e-05, "loss": 0.02, "step": 61620 }, { "grad_norm": 0.24142779409885406, "learning_rate": 9.35392994189094e-05, "loss": 0.0178, "step": 61630 }, { "grad_norm": 0.1658746749162674, "learning_rate": 9.35365893254551e-05, "loss": 0.0191, "step": 61640 }, { "grad_norm": 0.35795748233795166, "learning_rate": 9.353387870299013e-05, "loss": 0.0292, "step": 61650 }, { "grad_norm": 0.1747218668460846, "learning_rate": 9.353116755154742e-05, "loss": 0.0316, "step": 61660 }, { "grad_norm": 0.35136109590530396, "learning_rate": 9.352845587115994e-05, "loss": 0.0284, "step": 61670 }, { "grad_norm": 0.1588790863752365, "learning_rate": 9.352574366186062e-05, "loss": 0.0202, "step": 61680 }, { "grad_norm": 0.14408065378665924, "learning_rate": 9.352303092368242e-05, "loss": 0.0192, "step": 61690 }, { "grad_norm": 0.24268169701099396, "learning_rate": 9.35203176566583e-05, "loss": 0.0239, "step": 61700 }, { "grad_norm": 0.2120664268732071, "learning_rate": 9.351760386082123e-05, "loss": 0.0233, "step": 61710 }, { "grad_norm": 0.1713356375694275, "learning_rate": 9.351488953620418e-05, "loss": 0.0222, "step": 61720 }, { "grad_norm": 0.2590939700603485, "learning_rate": 9.351217468284014e-05, "loss": 0.0186, "step": 61730 }, { "grad_norm": 0.19559483230113983, "learning_rate": 9.35094593007621e-05, "loss": 0.0214, "step": 61740 }, { "grad_norm": 0.13780397176742554, "learning_rate": 9.350674339000303e-05, "loss": 0.0187, "step": 61750 }, { "grad_norm": 0.25266724824905396, "learning_rate": 9.350402695059598e-05, "loss": 0.0243, "step": 61760 }, { "grad_norm": 0.1336468756198883, "learning_rate": 9.350130998257391e-05, "loss": 0.0224, "step": 61770 }, { "grad_norm": 0.11774249374866486, "learning_rate": 9.349859248596986e-05, "loss": 0.0249, "step": 61780 }, { "grad_norm": 0.11637789011001587, "learning_rate": 9.349587446081684e-05, "loss": 0.0157, "step": 61790 }, { "grad_norm": 0.15499503910541534, "learning_rate": 9.349315590714786e-05, "loss": 0.0164, "step": 61800 }, { "grad_norm": 0.15407875180244446, "learning_rate": 9.3490436824996e-05, "loss": 0.0193, "step": 61810 }, { "grad_norm": 0.15553061664104462, "learning_rate": 9.348771721439426e-05, "loss": 0.0232, "step": 61820 }, { "grad_norm": 0.15615998208522797, "learning_rate": 9.34849970753757e-05, "loss": 0.0196, "step": 61830 }, { "grad_norm": 0.14440429210662842, "learning_rate": 9.348227640797336e-05, "loss": 0.0207, "step": 61840 }, { "grad_norm": 0.13959819078445435, "learning_rate": 9.347955521222032e-05, "loss": 0.0176, "step": 61850 }, { "grad_norm": 0.1899237483739853, "learning_rate": 9.347683348814962e-05, "loss": 0.0189, "step": 61860 }, { "grad_norm": 0.29086992144584656, "learning_rate": 9.347411123579435e-05, "loss": 0.0197, "step": 61870 }, { "grad_norm": 0.20842811465263367, "learning_rate": 9.347138845518757e-05, "loss": 0.0198, "step": 61880 }, { "grad_norm": 0.37081101536750793, "learning_rate": 9.34686651463624e-05, "loss": 0.0207, "step": 61890 }, { "grad_norm": 0.13196580111980438, "learning_rate": 9.346594130935189e-05, "loss": 0.0193, "step": 61900 }, { "grad_norm": 0.16572241485118866, "learning_rate": 9.346321694418916e-05, "loss": 0.017, "step": 61910 }, { "grad_norm": 0.16536691784858704, "learning_rate": 9.34604920509073e-05, "loss": 0.0201, "step": 61920 }, { "grad_norm": 0.12751837074756622, "learning_rate": 9.345776662953944e-05, "loss": 0.0238, "step": 61930 }, { "grad_norm": 0.17839454114437103, "learning_rate": 9.345504068011868e-05, "loss": 0.0213, "step": 61940 }, { "grad_norm": 0.13940292596817017, "learning_rate": 9.345231420267814e-05, "loss": 0.0227, "step": 61950 }, { "grad_norm": 0.20797094702720642, "learning_rate": 9.344958719725098e-05, "loss": 0.0229, "step": 61960 }, { "grad_norm": 0.203548401594162, "learning_rate": 9.34468596638703e-05, "loss": 0.0205, "step": 61970 }, { "grad_norm": 0.18151570856571198, "learning_rate": 9.344413160256924e-05, "loss": 0.0221, "step": 61980 }, { "grad_norm": 0.20909148454666138, "learning_rate": 9.344140301338098e-05, "loss": 0.0272, "step": 61990 }, { "grad_norm": 0.1219138577580452, "learning_rate": 9.343867389633864e-05, "loss": 0.0232, "step": 62000 }, { "grad_norm": 0.1328514963388443, "learning_rate": 9.343594425147542e-05, "loss": 0.02, "step": 62010 }, { "grad_norm": 0.21541909873485565, "learning_rate": 9.343321407882446e-05, "loss": 0.0177, "step": 62020 }, { "grad_norm": 0.13754212856292725, "learning_rate": 9.343048337841893e-05, "loss": 0.0204, "step": 62030 }, { "grad_norm": 0.17160190641880035, "learning_rate": 9.342775215029204e-05, "loss": 0.0246, "step": 62040 }, { "grad_norm": 0.22737689316272736, "learning_rate": 9.342502039447695e-05, "loss": 0.0251, "step": 62050 }, { "grad_norm": 0.13884444534778595, "learning_rate": 9.342228811100685e-05, "loss": 0.0174, "step": 62060 }, { "grad_norm": 0.2206239551305771, "learning_rate": 9.341955529991497e-05, "loss": 0.0193, "step": 62070 }, { "grad_norm": 0.8783884048461914, "learning_rate": 9.34168219612345e-05, "loss": 0.0192, "step": 62080 }, { "grad_norm": 0.17852534353733063, "learning_rate": 9.341408809499863e-05, "loss": 0.0268, "step": 62090 }, { "grad_norm": 0.1303386241197586, "learning_rate": 9.341135370124062e-05, "loss": 0.0366, "step": 62100 }, { "grad_norm": 0.1323457807302475, "learning_rate": 9.340861877999367e-05, "loss": 0.0169, "step": 62110 }, { "grad_norm": 0.15279459953308105, "learning_rate": 9.340588333129101e-05, "loss": 0.0147, "step": 62120 }, { "grad_norm": 0.2530720829963684, "learning_rate": 9.340314735516588e-05, "loss": 0.0202, "step": 62130 }, { "grad_norm": 0.1621725708246231, "learning_rate": 9.340041085165156e-05, "loss": 0.0184, "step": 62140 }, { "grad_norm": 0.17152459919452667, "learning_rate": 9.339767382078125e-05, "loss": 0.0242, "step": 62150 }, { "grad_norm": 0.1364840269088745, "learning_rate": 9.339493626258824e-05, "loss": 0.0295, "step": 62160 }, { "grad_norm": 0.250622421503067, "learning_rate": 9.339219817710576e-05, "loss": 0.0211, "step": 62170 }, { "grad_norm": 0.1431104689836502, "learning_rate": 9.338945956436712e-05, "loss": 0.0227, "step": 62180 }, { "grad_norm": 0.1226002424955368, "learning_rate": 9.338672042440559e-05, "loss": 0.0169, "step": 62190 }, { "grad_norm": 0.17851710319519043, "learning_rate": 9.338398075725442e-05, "loss": 0.0279, "step": 62200 }, { "grad_norm": 0.1293424665927887, "learning_rate": 9.338124056294694e-05, "loss": 0.0205, "step": 62210 }, { "grad_norm": 0.40175876021385193, "learning_rate": 9.337849984151642e-05, "loss": 0.0221, "step": 62220 }, { "grad_norm": 0.2069086879491806, "learning_rate": 9.337575859299617e-05, "loss": 0.0204, "step": 62230 }, { "grad_norm": 0.11795026808977127, "learning_rate": 9.33730168174195e-05, "loss": 0.0162, "step": 62240 }, { "grad_norm": 0.20145642757415771, "learning_rate": 9.337027451481973e-05, "loss": 0.0188, "step": 62250 }, { "grad_norm": 0.08604521304368973, "learning_rate": 9.336753168523017e-05, "loss": 0.02, "step": 62260 }, { "grad_norm": 0.2077191323041916, "learning_rate": 9.336478832868416e-05, "loss": 0.0195, "step": 62270 }, { "grad_norm": 0.20965106785297394, "learning_rate": 9.336204444521502e-05, "loss": 0.0189, "step": 62280 }, { "grad_norm": 0.1716184914112091, "learning_rate": 9.335930003485612e-05, "loss": 0.0209, "step": 62290 }, { "grad_norm": 0.2383241504430771, "learning_rate": 9.335655509764076e-05, "loss": 0.0185, "step": 62300 }, { "grad_norm": 0.19415737688541412, "learning_rate": 9.335380963360233e-05, "loss": 0.02, "step": 62310 }, { "grad_norm": 0.38801491260528564, "learning_rate": 9.335106364277419e-05, "loss": 0.0199, "step": 62320 }, { "grad_norm": 0.21905535459518433, "learning_rate": 9.334831712518969e-05, "loss": 0.0218, "step": 62330 }, { "grad_norm": 0.16637758910655975, "learning_rate": 9.334557008088221e-05, "loss": 0.0193, "step": 62340 }, { "grad_norm": 0.17174337804317474, "learning_rate": 9.334282250988511e-05, "loss": 0.0223, "step": 62350 }, { "grad_norm": 0.1890353411436081, "learning_rate": 9.334007441223183e-05, "loss": 0.0205, "step": 62360 }, { "grad_norm": 0.13944785296916962, "learning_rate": 9.333732578795571e-05, "loss": 0.0226, "step": 62370 }, { "grad_norm": 0.2274729460477829, "learning_rate": 9.333457663709015e-05, "loss": 0.0209, "step": 62380 }, { "grad_norm": 0.21561740338802338, "learning_rate": 9.333182695966858e-05, "loss": 0.0214, "step": 62390 }, { "grad_norm": 0.19806161522865295, "learning_rate": 9.33290767557244e-05, "loss": 0.0207, "step": 62400 }, { "grad_norm": 0.21657425165176392, "learning_rate": 9.332632602529103e-05, "loss": 0.0165, "step": 62410 }, { "grad_norm": 0.3404656946659088, "learning_rate": 9.33235747684019e-05, "loss": 0.0202, "step": 62420 }, { "grad_norm": 0.16078133881092072, "learning_rate": 9.332082298509042e-05, "loss": 0.0197, "step": 62430 }, { "grad_norm": 0.12419886887073517, "learning_rate": 9.331807067539005e-05, "loss": 0.0249, "step": 62440 }, { "grad_norm": 0.4333553910255432, "learning_rate": 9.331531783933421e-05, "loss": 0.0168, "step": 62450 }, { "grad_norm": 0.14159737527370453, "learning_rate": 9.331256447695635e-05, "loss": 0.0173, "step": 62460 }, { "grad_norm": 0.08406936377286911, "learning_rate": 9.330981058828996e-05, "loss": 0.0214, "step": 62470 }, { "grad_norm": 0.2483842819929123, "learning_rate": 9.330705617336848e-05, "loss": 0.0216, "step": 62480 }, { "grad_norm": 0.12684044241905212, "learning_rate": 9.330430123222537e-05, "loss": 0.0207, "step": 62490 }, { "grad_norm": 0.1548205018043518, "learning_rate": 9.330154576489411e-05, "loss": 0.0185, "step": 62500 }, { "grad_norm": 0.19373752176761627, "learning_rate": 9.32987897714082e-05, "loss": 0.0181, "step": 62510 }, { "grad_norm": 0.1385062336921692, "learning_rate": 9.32960332518011e-05, "loss": 0.0188, "step": 62520 }, { "grad_norm": 0.21835869550704956, "learning_rate": 9.329327620610633e-05, "loss": 0.0188, "step": 62530 }, { "grad_norm": 0.16309335827827454, "learning_rate": 9.329051863435738e-05, "loss": 0.0215, "step": 62540 }, { "grad_norm": 0.5834730863571167, "learning_rate": 9.328776053658774e-05, "loss": 0.0169, "step": 62550 }, { "grad_norm": 0.22917625308036804, "learning_rate": 9.328500191283096e-05, "loss": 0.0228, "step": 62560 }, { "grad_norm": 0.15020304918289185, "learning_rate": 9.328224276312052e-05, "loss": 0.025, "step": 62570 }, { "grad_norm": 0.1449255794286728, "learning_rate": 9.327948308748997e-05, "loss": 0.0372, "step": 62580 }, { "grad_norm": 0.42353153228759766, "learning_rate": 9.327672288597285e-05, "loss": 0.0249, "step": 62590 }, { "grad_norm": 0.14972253143787384, "learning_rate": 9.327396215860267e-05, "loss": 0.0216, "step": 62600 }, { "grad_norm": 0.16118845343589783, "learning_rate": 9.327120090541299e-05, "loss": 0.0204, "step": 62610 }, { "grad_norm": 0.1371067613363266, "learning_rate": 9.326843912643738e-05, "loss": 0.0195, "step": 62620 }, { "grad_norm": 0.18446394801139832, "learning_rate": 9.326567682170936e-05, "loss": 0.0229, "step": 62630 }, { "grad_norm": 0.19755807518959045, "learning_rate": 9.326291399126255e-05, "loss": 0.0348, "step": 62640 }, { "grad_norm": 0.14324335753917694, "learning_rate": 9.326015063513047e-05, "loss": 0.0192, "step": 62650 }, { "grad_norm": 0.11469327658414841, "learning_rate": 9.32573867533467e-05, "loss": 0.0191, "step": 62660 }, { "grad_norm": 0.1506369560956955, "learning_rate": 9.325462234594486e-05, "loss": 0.0217, "step": 62670 }, { "grad_norm": 0.16629086434841156, "learning_rate": 9.325185741295852e-05, "loss": 0.0214, "step": 62680 }, { "grad_norm": 0.21744374930858612, "learning_rate": 9.324909195442126e-05, "loss": 0.0262, "step": 62690 }, { "grad_norm": 0.15955069661140442, "learning_rate": 9.324632597036671e-05, "loss": 0.0185, "step": 62700 }, { "grad_norm": 0.3357553482055664, "learning_rate": 9.324355946082848e-05, "loss": 0.0243, "step": 62710 }, { "grad_norm": 0.1813650131225586, "learning_rate": 9.324079242584016e-05, "loss": 0.0254, "step": 62720 }, { "grad_norm": 0.14101368188858032, "learning_rate": 9.323802486543538e-05, "loss": 0.021, "step": 62730 }, { "grad_norm": 0.11360295861959457, "learning_rate": 9.323525677964778e-05, "loss": 0.0327, "step": 62740 }, { "grad_norm": 0.3033709228038788, "learning_rate": 9.3232488168511e-05, "loss": 0.0243, "step": 62750 }, { "grad_norm": 0.12812329828739166, "learning_rate": 9.322971903205866e-05, "loss": 0.0293, "step": 62760 }, { "grad_norm": 0.16309258341789246, "learning_rate": 9.322694937032441e-05, "loss": 0.0217, "step": 62770 }, { "grad_norm": 0.2125190794467926, "learning_rate": 9.322417918334192e-05, "loss": 0.0262, "step": 62780 }, { "grad_norm": 0.11300618946552277, "learning_rate": 9.322140847114486e-05, "loss": 0.0239, "step": 62790 }, { "grad_norm": 0.2716982066631317, "learning_rate": 9.321863723376685e-05, "loss": 0.0207, "step": 62800 }, { "grad_norm": 0.15599745512008667, "learning_rate": 9.321586547124161e-05, "loss": 0.0207, "step": 62810 }, { "grad_norm": 0.1694059669971466, "learning_rate": 9.32130931836028e-05, "loss": 0.0204, "step": 62820 }, { "grad_norm": 0.10993550717830658, "learning_rate": 9.32103203708841e-05, "loss": 0.0221, "step": 62830 }, { "grad_norm": 0.5510085821151733, "learning_rate": 9.320754703311922e-05, "loss": 0.0204, "step": 62840 }, { "grad_norm": 0.17324821650981903, "learning_rate": 9.320477317034185e-05, "loss": 0.0189, "step": 62850 }, { "grad_norm": 0.20174917578697205, "learning_rate": 9.320199878258568e-05, "loss": 0.0232, "step": 62860 }, { "grad_norm": 0.19622749090194702, "learning_rate": 9.319922386988445e-05, "loss": 0.0191, "step": 62870 }, { "grad_norm": 0.20342671871185303, "learning_rate": 9.319644843227186e-05, "loss": 0.0222, "step": 62880 }, { "grad_norm": 0.16398075222969055, "learning_rate": 9.319367246978163e-05, "loss": 0.0213, "step": 62890 }, { "grad_norm": 0.2218909114599228, "learning_rate": 9.31908959824475e-05, "loss": 0.019, "step": 62900 }, { "grad_norm": 0.2074245661497116, "learning_rate": 9.318811897030322e-05, "loss": 0.0235, "step": 62910 }, { "grad_norm": 0.1825033277273178, "learning_rate": 9.318534143338249e-05, "loss": 0.0238, "step": 62920 }, { "grad_norm": 0.21306295692920685, "learning_rate": 9.318256337171911e-05, "loss": 0.0179, "step": 62930 }, { "grad_norm": 0.21292872726917267, "learning_rate": 9.317978478534681e-05, "loss": 0.0175, "step": 62940 }, { "grad_norm": 0.27339184284210205, "learning_rate": 9.317700567429935e-05, "loss": 0.0215, "step": 62950 }, { "grad_norm": 0.19410763680934906, "learning_rate": 9.31742260386105e-05, "loss": 0.0214, "step": 62960 }, { "grad_norm": 0.19313837587833405, "learning_rate": 9.317144587831405e-05, "loss": 0.0284, "step": 62970 }, { "grad_norm": 0.16513735055923462, "learning_rate": 9.316866519344378e-05, "loss": 0.0183, "step": 62980 }, { "grad_norm": 0.21731068193912506, "learning_rate": 9.316588398403345e-05, "loss": 0.0208, "step": 62990 }, { "grad_norm": 0.17580558359622955, "learning_rate": 9.316310225011689e-05, "loss": 0.0195, "step": 63000 }, { "grad_norm": 0.16113677620887756, "learning_rate": 9.316031999172788e-05, "loss": 0.0263, "step": 63010 }, { "grad_norm": 0.11726745218038559, "learning_rate": 9.315753720890024e-05, "loss": 0.0221, "step": 63020 }, { "grad_norm": 0.20257417857646942, "learning_rate": 9.315475390166777e-05, "loss": 0.0254, "step": 63030 }, { "grad_norm": 0.1457224190235138, "learning_rate": 9.315197007006429e-05, "loss": 0.016, "step": 63040 }, { "grad_norm": 0.1797008365392685, "learning_rate": 9.314918571412363e-05, "loss": 0.02, "step": 63050 }, { "grad_norm": 0.14988131821155548, "learning_rate": 9.314640083387963e-05, "loss": 0.022, "step": 63060 }, { "grad_norm": 0.21073885262012482, "learning_rate": 9.314361542936613e-05, "loss": 0.0186, "step": 63070 }, { "grad_norm": 0.16774767637252808, "learning_rate": 9.314082950061695e-05, "loss": 0.0168, "step": 63080 }, { "grad_norm": 0.1932477205991745, "learning_rate": 9.313804304766598e-05, "loss": 0.0183, "step": 63090 }, { "grad_norm": 0.18374407291412354, "learning_rate": 9.313525607054705e-05, "loss": 0.0197, "step": 63100 }, { "grad_norm": 0.1359470784664154, "learning_rate": 9.313246856929403e-05, "loss": 0.0166, "step": 63110 }, { "grad_norm": 0.17342999577522278, "learning_rate": 9.312968054394078e-05, "loss": 0.0244, "step": 63120 }, { "grad_norm": 0.15479759871959686, "learning_rate": 9.31268919945212e-05, "loss": 0.0258, "step": 63130 }, { "grad_norm": 0.12435384094715118, "learning_rate": 9.312410292106917e-05, "loss": 0.0211, "step": 63140 }, { "grad_norm": 0.16272544860839844, "learning_rate": 9.312131332361857e-05, "loss": 0.0212, "step": 63150 }, { "grad_norm": 0.7450464963912964, "learning_rate": 9.311852320220331e-05, "loss": 0.0242, "step": 63160 }, { "grad_norm": 0.9339380860328674, "learning_rate": 9.311573255685726e-05, "loss": 0.0218, "step": 63170 }, { "grad_norm": 0.18539805710315704, "learning_rate": 9.311294138761437e-05, "loss": 0.0217, "step": 63180 }, { "grad_norm": 0.1144350916147232, "learning_rate": 9.311014969450853e-05, "loss": 0.0236, "step": 63190 }, { "grad_norm": 0.4745129942893982, "learning_rate": 9.310735747757367e-05, "loss": 0.021, "step": 63200 }, { "grad_norm": 0.18660441040992737, "learning_rate": 9.310456473684371e-05, "loss": 0.0174, "step": 63210 }, { "grad_norm": 0.13959118723869324, "learning_rate": 9.31017714723526e-05, "loss": 0.0193, "step": 63220 }, { "grad_norm": 0.17649759352207184, "learning_rate": 9.309897768413428e-05, "loss": 0.0215, "step": 63230 }, { "grad_norm": 0.19587473571300507, "learning_rate": 9.309618337222267e-05, "loss": 0.0225, "step": 63240 }, { "grad_norm": 0.18112631142139435, "learning_rate": 9.309338853665174e-05, "loss": 0.019, "step": 63250 }, { "grad_norm": 0.1668669432401657, "learning_rate": 9.309059317745547e-05, "loss": 0.029, "step": 63260 }, { "grad_norm": 0.14508618414402008, "learning_rate": 9.308779729466781e-05, "loss": 0.0186, "step": 63270 }, { "grad_norm": 0.18696577847003937, "learning_rate": 9.308500088832272e-05, "loss": 0.0204, "step": 63280 }, { "grad_norm": 0.1722160130739212, "learning_rate": 9.30822039584542e-05, "loss": 0.0194, "step": 63290 }, { "grad_norm": 0.19197241961956024, "learning_rate": 9.307940650509621e-05, "loss": 0.0176, "step": 63300 }, { "grad_norm": 0.12075423449277878, "learning_rate": 9.307660852828276e-05, "loss": 0.0179, "step": 63310 }, { "grad_norm": 0.14233635365962982, "learning_rate": 9.307381002804785e-05, "loss": 0.0274, "step": 63320 }, { "grad_norm": 0.23272809386253357, "learning_rate": 9.307101100442548e-05, "loss": 0.0246, "step": 63330 }, { "grad_norm": 0.21392524242401123, "learning_rate": 9.306821145744967e-05, "loss": 0.0188, "step": 63340 }, { "grad_norm": 0.16131509840488434, "learning_rate": 9.306541138715441e-05, "loss": 0.0246, "step": 63350 }, { "grad_norm": 0.16119827330112457, "learning_rate": 9.306261079357374e-05, "loss": 0.0233, "step": 63360 }, { "grad_norm": 0.21960556507110596, "learning_rate": 9.305980967674171e-05, "loss": 0.0353, "step": 63370 }, { "grad_norm": 0.22180306911468506, "learning_rate": 9.305700803669232e-05, "loss": 0.0242, "step": 63380 }, { "grad_norm": 0.19928321242332458, "learning_rate": 9.305420587345966e-05, "loss": 0.0226, "step": 63390 }, { "grad_norm": 0.1667373776435852, "learning_rate": 9.305140318707772e-05, "loss": 0.0212, "step": 63400 }, { "grad_norm": 0.18360228836536407, "learning_rate": 9.304859997758059e-05, "loss": 0.0207, "step": 63410 }, { "grad_norm": 0.23429572582244873, "learning_rate": 9.304579624500232e-05, "loss": 0.0277, "step": 63420 }, { "grad_norm": 0.14450466632843018, "learning_rate": 9.304299198937699e-05, "loss": 0.0235, "step": 63430 }, { "grad_norm": 0.17324042320251465, "learning_rate": 9.304018721073867e-05, "loss": 0.0237, "step": 63440 }, { "grad_norm": 0.14829304814338684, "learning_rate": 9.303738190912143e-05, "loss": 0.0213, "step": 63450 }, { "grad_norm": 0.15745386481285095, "learning_rate": 9.303457608455939e-05, "loss": 0.0203, "step": 63460 }, { "grad_norm": 0.20554746687412262, "learning_rate": 9.30317697370866e-05, "loss": 0.018, "step": 63470 }, { "grad_norm": 0.21138381958007812, "learning_rate": 9.302896286673717e-05, "loss": 0.0194, "step": 63480 }, { "grad_norm": 0.13348576426506042, "learning_rate": 9.302615547354523e-05, "loss": 0.0225, "step": 63490 }, { "grad_norm": 0.2330571860074997, "learning_rate": 9.302334755754487e-05, "loss": 0.024, "step": 63500 }, { "grad_norm": 0.1502498984336853, "learning_rate": 9.302053911877022e-05, "loss": 0.0205, "step": 63510 }, { "grad_norm": 0.15024656057357788, "learning_rate": 9.301773015725541e-05, "loss": 0.0266, "step": 63520 }, { "grad_norm": 0.1576843559741974, "learning_rate": 9.301492067303454e-05, "loss": 0.0289, "step": 63530 }, { "grad_norm": 0.12860648334026337, "learning_rate": 9.301211066614179e-05, "loss": 0.0141, "step": 63540 }, { "grad_norm": 0.15620508790016174, "learning_rate": 9.300930013661129e-05, "loss": 0.0202, "step": 63550 }, { "grad_norm": 0.24182990193367004, "learning_rate": 9.300648908447718e-05, "loss": 0.018, "step": 63560 }, { "grad_norm": 0.15203164517879486, "learning_rate": 9.300367750977362e-05, "loss": 0.022, "step": 63570 }, { "grad_norm": 0.16462908685207367, "learning_rate": 9.300086541253476e-05, "loss": 0.0196, "step": 63580 }, { "grad_norm": 0.14658989012241364, "learning_rate": 9.299805279279481e-05, "loss": 0.0251, "step": 63590 }, { "grad_norm": 0.17543061077594757, "learning_rate": 9.299523965058791e-05, "loss": 0.0204, "step": 63600 }, { "grad_norm": 0.2127636969089508, "learning_rate": 9.299242598594826e-05, "loss": 0.0268, "step": 63610 }, { "grad_norm": 0.16542308032512665, "learning_rate": 9.298961179891004e-05, "loss": 0.0212, "step": 63620 }, { "grad_norm": 0.11957207322120667, "learning_rate": 9.298679708950744e-05, "loss": 0.0271, "step": 63630 }, { "grad_norm": 0.09149568527936935, "learning_rate": 9.298398185777468e-05, "loss": 0.0174, "step": 63640 }, { "grad_norm": 0.09447304904460907, "learning_rate": 9.298116610374595e-05, "loss": 0.0217, "step": 63650 }, { "grad_norm": 0.1834770143032074, "learning_rate": 9.297834982745549e-05, "loss": 0.0226, "step": 63660 }, { "grad_norm": 0.21530017256736755, "learning_rate": 9.297553302893747e-05, "loss": 0.0239, "step": 63670 }, { "grad_norm": 0.15774005651474, "learning_rate": 9.297271570822617e-05, "loss": 0.0258, "step": 63680 }, { "grad_norm": 0.39667659997940063, "learning_rate": 9.296989786535579e-05, "loss": 0.0308, "step": 63690 }, { "grad_norm": 0.2577515244483948, "learning_rate": 9.296707950036059e-05, "loss": 0.0283, "step": 63700 }, { "grad_norm": 0.19091787934303284, "learning_rate": 9.296426061327479e-05, "loss": 0.0282, "step": 63710 }, { "grad_norm": 0.1489354968070984, "learning_rate": 9.296144120413267e-05, "loss": 0.0226, "step": 63720 }, { "grad_norm": 0.18605554103851318, "learning_rate": 9.295862127296848e-05, "loss": 0.0235, "step": 63730 }, { "grad_norm": 0.11812455952167511, "learning_rate": 9.295580081981647e-05, "loss": 0.0255, "step": 63740 }, { "grad_norm": 0.20942124724388123, "learning_rate": 9.295297984471094e-05, "loss": 0.025, "step": 63750 }, { "grad_norm": 0.22298894822597504, "learning_rate": 9.295015834768614e-05, "loss": 0.0207, "step": 63760 }, { "grad_norm": 0.14333075284957886, "learning_rate": 9.294733632877637e-05, "loss": 0.0216, "step": 63770 }, { "grad_norm": 0.13212350010871887, "learning_rate": 9.294451378801591e-05, "loss": 0.0226, "step": 63780 }, { "grad_norm": 0.14461566507816315, "learning_rate": 9.294169072543906e-05, "loss": 0.0193, "step": 63790 }, { "grad_norm": 0.21419692039489746, "learning_rate": 9.293886714108013e-05, "loss": 0.0212, "step": 63800 }, { "grad_norm": 0.16428042948246002, "learning_rate": 9.293604303497343e-05, "loss": 0.0186, "step": 63810 }, { "grad_norm": 0.18526266515254974, "learning_rate": 9.293321840715326e-05, "loss": 0.0265, "step": 63820 }, { "grad_norm": 0.11815878003835678, "learning_rate": 9.293039325765393e-05, "loss": 0.0256, "step": 63830 }, { "grad_norm": 0.14911293983459473, "learning_rate": 9.292756758650983e-05, "loss": 0.0247, "step": 63840 }, { "grad_norm": 0.11794400215148926, "learning_rate": 9.292474139375522e-05, "loss": 0.0204, "step": 63850 }, { "grad_norm": 0.7548872232437134, "learning_rate": 9.292191467942449e-05, "loss": 0.0173, "step": 63860 }, { "grad_norm": 0.13045282661914825, "learning_rate": 9.291908744355196e-05, "loss": 0.0199, "step": 63870 }, { "grad_norm": 0.14282022416591644, "learning_rate": 9.291625968617201e-05, "loss": 0.0174, "step": 63880 }, { "grad_norm": 0.12344055622816086, "learning_rate": 9.291343140731899e-05, "loss": 0.0218, "step": 63890 }, { "grad_norm": 0.16016152501106262, "learning_rate": 9.291060260702724e-05, "loss": 0.0183, "step": 63900 }, { "grad_norm": 0.3472888767719269, "learning_rate": 9.290777328533115e-05, "loss": 0.0169, "step": 63910 }, { "grad_norm": 0.21572108566761017, "learning_rate": 9.290494344226512e-05, "loss": 0.0237, "step": 63920 }, { "grad_norm": 0.20892350375652313, "learning_rate": 9.290211307786352e-05, "loss": 0.0269, "step": 63930 }, { "grad_norm": 0.3154829144477844, "learning_rate": 9.289928219216073e-05, "loss": 0.0208, "step": 63940 }, { "grad_norm": 0.1414186805486679, "learning_rate": 9.289645078519117e-05, "loss": 0.0199, "step": 63950 }, { "grad_norm": 0.16011974215507507, "learning_rate": 9.289361885698922e-05, "loss": 0.0226, "step": 63960 }, { "grad_norm": 0.6321157813072205, "learning_rate": 9.28907864075893e-05, "loss": 0.02, "step": 63970 }, { "grad_norm": 0.13584716618061066, "learning_rate": 9.288795343702584e-05, "loss": 0.0216, "step": 63980 }, { "grad_norm": 0.12188518792390823, "learning_rate": 9.288511994533325e-05, "loss": 0.0162, "step": 63990 }, { "grad_norm": 0.20719721913337708, "learning_rate": 9.288228593254597e-05, "loss": 0.0195, "step": 64000 }, { "grad_norm": 0.38929125666618347, "learning_rate": 9.287945139869843e-05, "loss": 0.0218, "step": 64010 }, { "grad_norm": 0.10668779164552689, "learning_rate": 9.287661634382505e-05, "loss": 0.0169, "step": 64020 }, { "grad_norm": 0.27550989389419556, "learning_rate": 9.287378076796031e-05, "loss": 0.0218, "step": 64030 }, { "grad_norm": 0.1429418921470642, "learning_rate": 9.287094467113865e-05, "loss": 0.0182, "step": 64040 }, { "grad_norm": 0.11046314984560013, "learning_rate": 9.286810805339454e-05, "loss": 0.0189, "step": 64050 }, { "grad_norm": 0.22733831405639648, "learning_rate": 9.286527091476245e-05, "loss": 0.0218, "step": 64060 }, { "grad_norm": 0.2562029957771301, "learning_rate": 9.286243325527684e-05, "loss": 0.0242, "step": 64070 }, { "grad_norm": 0.270234614610672, "learning_rate": 9.28595950749722e-05, "loss": 0.0241, "step": 64080 }, { "grad_norm": 0.12132029980421066, "learning_rate": 9.285675637388301e-05, "loss": 0.0187, "step": 64090 }, { "grad_norm": 0.2612956166267395, "learning_rate": 9.285391715204375e-05, "loss": 0.0193, "step": 64100 }, { "grad_norm": 0.13146522641181946, "learning_rate": 9.285107740948896e-05, "loss": 0.02, "step": 64110 }, { "grad_norm": 0.11045555025339127, "learning_rate": 9.28482371462531e-05, "loss": 0.0161, "step": 64120 }, { "grad_norm": 0.13045310974121094, "learning_rate": 9.284539636237072e-05, "loss": 0.0225, "step": 64130 }, { "grad_norm": 0.12679851055145264, "learning_rate": 9.284255505787632e-05, "loss": 0.0175, "step": 64140 }, { "grad_norm": 0.46481046080589294, "learning_rate": 9.283971323280442e-05, "loss": 0.0262, "step": 64150 }, { "grad_norm": 0.20060789585113525, "learning_rate": 9.283687088718957e-05, "loss": 0.0168, "step": 64160 }, { "grad_norm": 0.22966468334197998, "learning_rate": 9.283402802106629e-05, "loss": 0.0224, "step": 64170 }, { "grad_norm": 0.13395074009895325, "learning_rate": 9.283118463446911e-05, "loss": 0.0201, "step": 64180 }, { "grad_norm": 0.14356070756912231, "learning_rate": 9.282834072743262e-05, "loss": 0.02, "step": 64190 }, { "grad_norm": 0.24326162040233612, "learning_rate": 9.282549629999134e-05, "loss": 0.0239, "step": 64200 }, { "grad_norm": 0.14486613869667053, "learning_rate": 9.282265135217985e-05, "loss": 0.0175, "step": 64210 }, { "grad_norm": 0.11966494470834732, "learning_rate": 9.281980588403272e-05, "loss": 0.0201, "step": 64220 }, { "grad_norm": 0.11972657591104507, "learning_rate": 9.281695989558451e-05, "loss": 0.0222, "step": 64230 }, { "grad_norm": 0.16158241033554077, "learning_rate": 9.281411338686982e-05, "loss": 0.0214, "step": 64240 }, { "grad_norm": 0.20183415710926056, "learning_rate": 9.281126635792322e-05, "loss": 0.0168, "step": 64250 }, { "grad_norm": 0.1536305844783783, "learning_rate": 9.280841880877932e-05, "loss": 0.0195, "step": 64260 }, { "grad_norm": 0.1533913016319275, "learning_rate": 9.280557073947271e-05, "loss": 0.0141, "step": 64270 }, { "grad_norm": 0.16931374371051788, "learning_rate": 9.280272215003801e-05, "loss": 0.0228, "step": 64280 }, { "grad_norm": 0.20900215208530426, "learning_rate": 9.279987304050982e-05, "loss": 0.0198, "step": 64290 }, { "grad_norm": 0.11161047220230103, "learning_rate": 9.279702341092274e-05, "loss": 0.0203, "step": 64300 }, { "grad_norm": 0.1422426998615265, "learning_rate": 9.279417326131143e-05, "loss": 0.019, "step": 64310 }, { "grad_norm": 0.11470961570739746, "learning_rate": 9.279132259171053e-05, "loss": 0.0198, "step": 64320 }, { "grad_norm": 0.15971039235591888, "learning_rate": 9.278847140215464e-05, "loss": 0.0217, "step": 64330 }, { "grad_norm": 0.2246352732181549, "learning_rate": 9.278561969267842e-05, "loss": 0.0236, "step": 64340 }, { "grad_norm": 0.14760613441467285, "learning_rate": 9.278276746331654e-05, "loss": 0.0183, "step": 64350 }, { "grad_norm": 0.14112001657485962, "learning_rate": 9.277991471410363e-05, "loss": 0.0217, "step": 64360 }, { "grad_norm": 0.10626363009214401, "learning_rate": 9.27770614450744e-05, "loss": 0.024, "step": 64370 }, { "grad_norm": 0.29425883293151855, "learning_rate": 9.277420765626343e-05, "loss": 0.0264, "step": 64380 }, { "grad_norm": 0.585120439529419, "learning_rate": 9.277135334770548e-05, "loss": 0.0198, "step": 64390 }, { "grad_norm": 0.12518192827701569, "learning_rate": 9.276849851943521e-05, "loss": 0.0198, "step": 64400 }, { "grad_norm": 0.10816238820552826, "learning_rate": 9.27656431714873e-05, "loss": 0.0169, "step": 64410 }, { "grad_norm": 0.09980780631303787, "learning_rate": 9.276278730389642e-05, "loss": 0.0322, "step": 64420 }, { "grad_norm": 0.20064722001552582, "learning_rate": 9.275993091669734e-05, "loss": 0.0209, "step": 64430 }, { "grad_norm": 0.24474990367889404, "learning_rate": 9.27570740099247e-05, "loss": 0.0193, "step": 64440 }, { "grad_norm": 0.12117928266525269, "learning_rate": 9.275421658361325e-05, "loss": 0.0132, "step": 64450 }, { "grad_norm": 0.24454385042190552, "learning_rate": 9.27513586377977e-05, "loss": 0.0172, "step": 64460 }, { "grad_norm": 0.11088403314352036, "learning_rate": 9.274850017251278e-05, "loss": 0.016, "step": 64470 }, { "grad_norm": 0.164296492934227, "learning_rate": 9.274564118779323e-05, "loss": 0.0204, "step": 64480 }, { "grad_norm": 0.20806556940078735, "learning_rate": 9.274278168367378e-05, "loss": 0.0265, "step": 64490 }, { "grad_norm": 0.2834213078022003, "learning_rate": 9.273992166018916e-05, "loss": 0.0204, "step": 64500 }, { "grad_norm": 0.1824965626001358, "learning_rate": 9.273706111737415e-05, "loss": 0.0188, "step": 64510 }, { "grad_norm": 0.12501810491085052, "learning_rate": 9.27342000552635e-05, "loss": 0.0209, "step": 64520 }, { "grad_norm": 0.22884882986545563, "learning_rate": 9.273133847389197e-05, "loss": 0.0194, "step": 64530 }, { "grad_norm": 0.2102479189634323, "learning_rate": 9.272847637329431e-05, "loss": 0.0189, "step": 64540 }, { "grad_norm": 0.1509949266910553, "learning_rate": 9.272561375350535e-05, "loss": 0.0209, "step": 64550 }, { "grad_norm": 0.145485982298851, "learning_rate": 9.272275061455984e-05, "loss": 0.0208, "step": 64560 }, { "grad_norm": 0.13645465672016144, "learning_rate": 9.271988695649256e-05, "loss": 0.0217, "step": 64570 }, { "grad_norm": 0.09282032400369644, "learning_rate": 9.271702277933833e-05, "loss": 0.0196, "step": 64580 }, { "grad_norm": 0.114710234105587, "learning_rate": 9.271415808313193e-05, "loss": 0.0243, "step": 64590 }, { "grad_norm": 0.7324726581573486, "learning_rate": 9.271129286790819e-05, "loss": 0.0195, "step": 64600 }, { "grad_norm": 0.16438640654087067, "learning_rate": 9.270842713370191e-05, "loss": 0.0223, "step": 64610 }, { "grad_norm": 0.1312442123889923, "learning_rate": 9.270556088054791e-05, "loss": 0.0271, "step": 64620 }, { "grad_norm": 0.3630571663379669, "learning_rate": 9.270269410848105e-05, "loss": 0.0261, "step": 64630 }, { "grad_norm": 0.21007800102233887, "learning_rate": 9.269982681753612e-05, "loss": 0.0332, "step": 64640 }, { "grad_norm": 0.21184083819389343, "learning_rate": 9.269695900774798e-05, "loss": 0.0287, "step": 64650 }, { "grad_norm": 0.17654983699321747, "learning_rate": 9.269409067915147e-05, "loss": 0.0255, "step": 64660 }, { "grad_norm": 0.1639050394296646, "learning_rate": 9.269122183178146e-05, "loss": 0.0189, "step": 64670 }, { "grad_norm": 0.16371266543865204, "learning_rate": 9.26883524656728e-05, "loss": 0.026, "step": 64680 }, { "grad_norm": 0.11488479375839233, "learning_rate": 9.268548258086033e-05, "loss": 0.0247, "step": 64690 }, { "grad_norm": 0.25518664717674255, "learning_rate": 9.268261217737897e-05, "loss": 0.022, "step": 64700 }, { "grad_norm": 0.1984778493642807, "learning_rate": 9.267974125526356e-05, "loss": 0.0232, "step": 64710 }, { "grad_norm": 0.20584337413311005, "learning_rate": 9.2676869814549e-05, "loss": 0.0284, "step": 64720 }, { "grad_norm": 0.1607016921043396, "learning_rate": 9.267399785527017e-05, "loss": 0.0131, "step": 64730 }, { "grad_norm": 0.14602933824062347, "learning_rate": 9.2671125377462e-05, "loss": 0.0217, "step": 64740 }, { "grad_norm": 0.22676846385002136, "learning_rate": 9.266825238115935e-05, "loss": 0.0164, "step": 64750 }, { "grad_norm": 0.11790921539068222, "learning_rate": 9.266537886639714e-05, "loss": 0.0204, "step": 64760 }, { "grad_norm": 0.13948288559913635, "learning_rate": 9.266250483321031e-05, "loss": 0.0202, "step": 64770 }, { "grad_norm": 0.18556608259677887, "learning_rate": 9.265963028163376e-05, "loss": 0.0243, "step": 64780 }, { "grad_norm": 0.15293766558170319, "learning_rate": 9.265675521170243e-05, "loss": 0.03, "step": 64790 }, { "grad_norm": 0.13334394991397858, "learning_rate": 9.265387962345125e-05, "loss": 0.0238, "step": 64800 }, { "grad_norm": 0.1384868174791336, "learning_rate": 9.265100351691516e-05, "loss": 0.017, "step": 64810 }, { "grad_norm": 0.13681092858314514, "learning_rate": 9.26481268921291e-05, "loss": 0.0141, "step": 64820 }, { "grad_norm": 0.15037761628627777, "learning_rate": 9.264524974912803e-05, "loss": 0.0243, "step": 64830 }, { "grad_norm": 0.16074730455875397, "learning_rate": 9.264237208794692e-05, "loss": 0.0173, "step": 64840 }, { "grad_norm": 0.1685955673456192, "learning_rate": 9.263949390862073e-05, "loss": 0.0155, "step": 64850 }, { "grad_norm": 0.15147413313388824, "learning_rate": 9.263661521118442e-05, "loss": 0.0188, "step": 64860 }, { "grad_norm": 0.16502924263477325, "learning_rate": 9.263373599567298e-05, "loss": 0.024, "step": 64870 }, { "grad_norm": 0.13591915369033813, "learning_rate": 9.26308562621214e-05, "loss": 0.0205, "step": 64880 }, { "grad_norm": 0.19555190205574036, "learning_rate": 9.262797601056468e-05, "loss": 0.0224, "step": 64890 }, { "grad_norm": 0.22342099249362946, "learning_rate": 9.262509524103778e-05, "loss": 0.0199, "step": 64900 }, { "grad_norm": 0.19027194380760193, "learning_rate": 9.262221395357574e-05, "loss": 0.0244, "step": 64910 }, { "grad_norm": 0.27328285574913025, "learning_rate": 9.261933214821357e-05, "loss": 0.024, "step": 64920 }, { "grad_norm": 0.16776950657367706, "learning_rate": 9.261644982498626e-05, "loss": 0.0238, "step": 64930 }, { "grad_norm": 0.11218835413455963, "learning_rate": 9.261356698392884e-05, "loss": 0.0226, "step": 64940 }, { "grad_norm": 0.28268173336982727, "learning_rate": 9.261068362507636e-05, "loss": 0.0235, "step": 64950 }, { "grad_norm": 0.31345853209495544, "learning_rate": 9.260779974846383e-05, "loss": 0.0206, "step": 64960 }, { "grad_norm": 0.14685723185539246, "learning_rate": 9.260491535412631e-05, "loss": 0.0172, "step": 64970 }, { "grad_norm": 0.11392773687839508, "learning_rate": 9.260203044209885e-05, "loss": 0.022, "step": 64980 }, { "grad_norm": 0.17138782143592834, "learning_rate": 9.259914501241652e-05, "loss": 0.0181, "step": 64990 }, { "grad_norm": 0.19587118923664093, "learning_rate": 9.259625906511431e-05, "loss": 0.0181, "step": 65000 }, { "grad_norm": 0.1523810178041458, "learning_rate": 9.259337260022736e-05, "loss": 0.0197, "step": 65010 }, { "grad_norm": 0.1419934779405594, "learning_rate": 9.259048561779071e-05, "loss": 0.0226, "step": 65020 }, { "grad_norm": 0.31038644909858704, "learning_rate": 9.258759811783946e-05, "loss": 0.022, "step": 65030 }, { "grad_norm": 0.25732266902923584, "learning_rate": 9.258471010040868e-05, "loss": 0.0204, "step": 65040 }, { "grad_norm": 0.5302042961120605, "learning_rate": 9.258182156553346e-05, "loss": 0.0199, "step": 65050 }, { "grad_norm": 0.792522668838501, "learning_rate": 9.257893251324892e-05, "loss": 0.0179, "step": 65060 }, { "grad_norm": 0.17073900997638702, "learning_rate": 9.257604294359012e-05, "loss": 0.0209, "step": 65070 }, { "grad_norm": 0.1509384959936142, "learning_rate": 9.257315285659222e-05, "loss": 0.0223, "step": 65080 }, { "grad_norm": 0.15984384715557098, "learning_rate": 9.257026225229031e-05, "loss": 0.0213, "step": 65090 }, { "grad_norm": 0.15445232391357422, "learning_rate": 9.256737113071951e-05, "loss": 0.0146, "step": 65100 }, { "grad_norm": 0.11154574900865555, "learning_rate": 9.256447949191498e-05, "loss": 0.0263, "step": 65110 }, { "grad_norm": 0.5440018773078918, "learning_rate": 9.256158733591184e-05, "loss": 0.0272, "step": 65120 }, { "grad_norm": 0.4936562478542328, "learning_rate": 9.255869466274522e-05, "loss": 0.0254, "step": 65130 }, { "grad_norm": 0.14771054685115814, "learning_rate": 9.255580147245027e-05, "loss": 0.0277, "step": 65140 }, { "grad_norm": 0.14658872783184052, "learning_rate": 9.255290776506215e-05, "loss": 0.0198, "step": 65150 }, { "grad_norm": 0.16985151171684265, "learning_rate": 9.255001354061604e-05, "loss": 0.0184, "step": 65160 }, { "grad_norm": 0.14127175509929657, "learning_rate": 9.25471187991471e-05, "loss": 0.0225, "step": 65170 }, { "grad_norm": 0.15371282398700714, "learning_rate": 9.254422354069048e-05, "loss": 0.0281, "step": 65180 }, { "grad_norm": 0.20337526500225067, "learning_rate": 9.254132776528137e-05, "loss": 0.0245, "step": 65190 }, { "grad_norm": 0.14913305640220642, "learning_rate": 9.253843147295497e-05, "loss": 0.0184, "step": 65200 }, { "grad_norm": 0.13107368350028992, "learning_rate": 9.253553466374646e-05, "loss": 0.0307, "step": 65210 }, { "grad_norm": 0.3055514395236969, "learning_rate": 9.253263733769105e-05, "loss": 0.0279, "step": 65220 }, { "grad_norm": 0.13023720681667328, "learning_rate": 9.252973949482393e-05, "loss": 0.0233, "step": 65230 }, { "grad_norm": 0.16284388303756714, "learning_rate": 9.252684113518033e-05, "loss": 0.0251, "step": 65240 }, { "grad_norm": 0.17733028531074524, "learning_rate": 9.252394225879545e-05, "loss": 0.0237, "step": 65250 }, { "grad_norm": 0.16394218802452087, "learning_rate": 9.252104286570453e-05, "loss": 0.0209, "step": 65260 }, { "grad_norm": 0.22866061329841614, "learning_rate": 9.251814295594279e-05, "loss": 0.0183, "step": 65270 }, { "grad_norm": 0.2342258244752884, "learning_rate": 9.251524252954547e-05, "loss": 0.0186, "step": 65280 }, { "grad_norm": 0.1244584321975708, "learning_rate": 9.25123415865478e-05, "loss": 0.0191, "step": 65290 }, { "grad_norm": 0.1334535926580429, "learning_rate": 9.250944012698506e-05, "loss": 0.0204, "step": 65300 }, { "grad_norm": 0.1714443862438202, "learning_rate": 9.250653815089247e-05, "loss": 0.0207, "step": 65310 }, { "grad_norm": 0.39111801981925964, "learning_rate": 9.250363565830532e-05, "loss": 0.0224, "step": 65320 }, { "grad_norm": 0.18899589776992798, "learning_rate": 9.250073264925887e-05, "loss": 0.0255, "step": 65330 }, { "grad_norm": 0.1737273633480072, "learning_rate": 9.249782912378838e-05, "loss": 0.0236, "step": 65340 }, { "grad_norm": 0.13423658907413483, "learning_rate": 9.249492508192914e-05, "loss": 0.0192, "step": 65350 }, { "grad_norm": 0.16384002566337585, "learning_rate": 9.249202052371646e-05, "loss": 0.019, "step": 65360 }, { "grad_norm": 0.11696290224790573, "learning_rate": 9.248911544918559e-05, "loss": 0.0186, "step": 65370 }, { "grad_norm": 0.11915040761232376, "learning_rate": 9.248620985837186e-05, "loss": 0.0164, "step": 65380 }, { "grad_norm": 0.10117678344249725, "learning_rate": 9.248330375131055e-05, "loss": 0.0229, "step": 65390 }, { "grad_norm": 0.30902886390686035, "learning_rate": 9.248039712803701e-05, "loss": 0.0188, "step": 65400 }, { "grad_norm": 0.21987447142601013, "learning_rate": 9.247748998858653e-05, "loss": 0.0234, "step": 65410 }, { "grad_norm": 0.11162744462490082, "learning_rate": 9.247458233299444e-05, "loss": 0.0171, "step": 65420 }, { "grad_norm": 0.1806253045797348, "learning_rate": 9.247167416129607e-05, "loss": 0.0164, "step": 65430 }, { "grad_norm": 0.19703178107738495, "learning_rate": 9.246876547352676e-05, "loss": 0.0167, "step": 65440 }, { "grad_norm": 0.2902010977268219, "learning_rate": 9.246585626972186e-05, "loss": 0.0177, "step": 65450 }, { "grad_norm": 0.32410016655921936, "learning_rate": 9.24629465499167e-05, "loss": 0.0268, "step": 65460 }, { "grad_norm": 0.13184531033039093, "learning_rate": 9.246003631414666e-05, "loss": 0.0156, "step": 65470 }, { "grad_norm": 0.13681428134441376, "learning_rate": 9.245712556244708e-05, "loss": 0.0195, "step": 65480 }, { "grad_norm": 0.12400940805673599, "learning_rate": 9.245421429485335e-05, "loss": 0.0212, "step": 65490 }, { "grad_norm": 0.1786620318889618, "learning_rate": 9.245130251140082e-05, "loss": 0.0211, "step": 65500 }, { "grad_norm": 0.10520234704017639, "learning_rate": 9.244839021212488e-05, "loss": 0.0185, "step": 65510 }, { "grad_norm": 0.18256312608718872, "learning_rate": 9.244547739706093e-05, "loss": 0.0261, "step": 65520 }, { "grad_norm": 0.21656781435012817, "learning_rate": 9.244256406624436e-05, "loss": 0.0199, "step": 65530 }, { "grad_norm": 0.12406191229820251, "learning_rate": 9.243965021971057e-05, "loss": 0.0163, "step": 65540 }, { "grad_norm": 0.11863255500793457, "learning_rate": 9.243673585749494e-05, "loss": 0.0197, "step": 65550 }, { "grad_norm": 0.1618729680776596, "learning_rate": 9.243382097963291e-05, "loss": 0.0183, "step": 65560 }, { "grad_norm": 0.1521953046321869, "learning_rate": 9.243090558615989e-05, "loss": 0.0146, "step": 65570 }, { "grad_norm": 0.15663784742355347, "learning_rate": 9.24279896771113e-05, "loss": 0.0188, "step": 65580 }, { "grad_norm": 0.11179233342409134, "learning_rate": 9.242507325252259e-05, "loss": 0.0184, "step": 65590 }, { "grad_norm": 2.2396764755249023, "learning_rate": 9.242215631242919e-05, "loss": 0.0226, "step": 65600 }, { "grad_norm": 0.09376486390829086, "learning_rate": 9.241923885686652e-05, "loss": 0.0233, "step": 65610 }, { "grad_norm": 0.1466195285320282, "learning_rate": 9.241632088587004e-05, "loss": 0.0198, "step": 65620 }, { "grad_norm": 0.2184535413980484, "learning_rate": 9.241340239947523e-05, "loss": 0.0291, "step": 65630 }, { "grad_norm": 0.189660906791687, "learning_rate": 9.241048339771754e-05, "loss": 0.0202, "step": 65640 }, { "grad_norm": 0.12690997123718262, "learning_rate": 9.240756388063242e-05, "loss": 0.0206, "step": 65650 }, { "grad_norm": 0.147138312458992, "learning_rate": 9.240464384825538e-05, "loss": 0.0194, "step": 65660 }, { "grad_norm": 0.12958398461341858, "learning_rate": 9.240172330062186e-05, "loss": 0.0192, "step": 65670 }, { "grad_norm": 0.19942139089107513, "learning_rate": 9.239880223776738e-05, "loss": 0.0164, "step": 65680 }, { "grad_norm": 0.2807976305484772, "learning_rate": 9.239588065972742e-05, "loss": 0.0199, "step": 65690 }, { "grad_norm": 0.1679793745279312, "learning_rate": 9.239295856653747e-05, "loss": 0.0194, "step": 65700 }, { "grad_norm": 0.12148886919021606, "learning_rate": 9.239003595823306e-05, "loss": 0.0158, "step": 65710 }, { "grad_norm": 0.17723993957042694, "learning_rate": 9.238711283484968e-05, "loss": 0.0191, "step": 65720 }, { "grad_norm": 0.08858657628297806, "learning_rate": 9.238418919642287e-05, "loss": 0.017, "step": 65730 }, { "grad_norm": 0.18424077332019806, "learning_rate": 9.238126504298814e-05, "loss": 0.0156, "step": 65740 }, { "grad_norm": 0.1677166074514389, "learning_rate": 9.237834037458102e-05, "loss": 0.0166, "step": 65750 }, { "grad_norm": 0.22318850457668304, "learning_rate": 9.237541519123707e-05, "loss": 0.0172, "step": 65760 }, { "grad_norm": 0.21002519130706787, "learning_rate": 9.23724894929918e-05, "loss": 0.0244, "step": 65770 }, { "grad_norm": 0.2429230809211731, "learning_rate": 9.236956327988078e-05, "loss": 0.028, "step": 65780 }, { "grad_norm": 0.14425648748874664, "learning_rate": 9.236663655193955e-05, "loss": 0.0231, "step": 65790 }, { "grad_norm": 0.21182627975940704, "learning_rate": 9.236370930920371e-05, "loss": 0.0269, "step": 65800 }, { "grad_norm": 0.20612333714962006, "learning_rate": 9.23607815517088e-05, "loss": 0.0268, "step": 65810 }, { "grad_norm": 0.36795634031295776, "learning_rate": 9.235785327949038e-05, "loss": 0.019, "step": 65820 }, { "grad_norm": 0.5225317478179932, "learning_rate": 9.235492449258407e-05, "loss": 0.0171, "step": 65830 }, { "grad_norm": 0.22587476670742035, "learning_rate": 9.235199519102542e-05, "loss": 0.0256, "step": 65840 }, { "grad_norm": 0.20510722696781158, "learning_rate": 9.234906537485006e-05, "loss": 0.0274, "step": 65850 }, { "grad_norm": 0.2734074592590332, "learning_rate": 9.234613504409356e-05, "loss": 0.028, "step": 65860 }, { "grad_norm": 0.12146077305078506, "learning_rate": 9.234320419879154e-05, "loss": 0.026, "step": 65870 }, { "grad_norm": 0.16701924800872803, "learning_rate": 9.234027283897961e-05, "loss": 0.0276, "step": 65880 }, { "grad_norm": 0.29317644238471985, "learning_rate": 9.233734096469339e-05, "loss": 0.0237, "step": 65890 }, { "grad_norm": 0.2005164474248886, "learning_rate": 9.23344085759685e-05, "loss": 0.0327, "step": 65900 }, { "grad_norm": 0.4777703881263733, "learning_rate": 9.233147567284057e-05, "loss": 0.0299, "step": 65910 }, { "grad_norm": 0.1776566356420517, "learning_rate": 9.232854225534527e-05, "loss": 0.028, "step": 65920 }, { "grad_norm": 0.28235316276550293, "learning_rate": 9.23256083235182e-05, "loss": 0.0224, "step": 65930 }, { "grad_norm": 0.30621108412742615, "learning_rate": 9.232267387739502e-05, "loss": 0.0277, "step": 65940 }, { "grad_norm": 0.12418196350336075, "learning_rate": 9.23197389170114e-05, "loss": 0.021, "step": 65950 }, { "grad_norm": 0.20763926208019257, "learning_rate": 9.231680344240298e-05, "loss": 0.027, "step": 65960 }, { "grad_norm": 0.17280909419059753, "learning_rate": 9.231386745360546e-05, "loss": 0.0226, "step": 65970 }, { "grad_norm": 0.15956005454063416, "learning_rate": 9.231093095065451e-05, "loss": 0.0184, "step": 65980 }, { "grad_norm": 0.15829342603683472, "learning_rate": 9.23079939335858e-05, "loss": 0.03, "step": 65990 }, { "grad_norm": 0.16088765859603882, "learning_rate": 9.2305056402435e-05, "loss": 0.0221, "step": 66000 }, { "grad_norm": 0.13487274944782257, "learning_rate": 9.230211835723783e-05, "loss": 0.0152, "step": 66010 }, { "grad_norm": 0.10196973383426666, "learning_rate": 9.229917979802997e-05, "loss": 0.0156, "step": 66020 }, { "grad_norm": 0.13441382348537445, "learning_rate": 9.229624072484716e-05, "loss": 0.0199, "step": 66030 }, { "grad_norm": 0.23600199818611145, "learning_rate": 9.229330113772507e-05, "loss": 0.0273, "step": 66040 }, { "grad_norm": 0.16314589977264404, "learning_rate": 9.229036103669946e-05, "loss": 0.0157, "step": 66050 }, { "grad_norm": 0.14265766739845276, "learning_rate": 9.228742042180603e-05, "loss": 0.0167, "step": 66060 }, { "grad_norm": 0.10525742173194885, "learning_rate": 9.228447929308051e-05, "loss": 0.0176, "step": 66070 }, { "grad_norm": 0.14423754811286926, "learning_rate": 9.228153765055864e-05, "loss": 0.0197, "step": 66080 }, { "grad_norm": 0.11990070343017578, "learning_rate": 9.227859549427618e-05, "loss": 0.0168, "step": 66090 }, { "grad_norm": 0.11638092249631882, "learning_rate": 9.227565282426887e-05, "loss": 0.0155, "step": 66100 }, { "grad_norm": 0.1424451619386673, "learning_rate": 9.227270964057245e-05, "loss": 0.0195, "step": 66110 }, { "grad_norm": 0.1742822527885437, "learning_rate": 9.22697659432227e-05, "loss": 0.0184, "step": 66120 }, { "grad_norm": 0.15519319474697113, "learning_rate": 9.226682173225537e-05, "loss": 0.0232, "step": 66130 }, { "grad_norm": 0.15087208151817322, "learning_rate": 9.226387700770626e-05, "loss": 0.0256, "step": 66140 }, { "grad_norm": 0.1248890832066536, "learning_rate": 9.226093176961115e-05, "loss": 0.0158, "step": 66150 }, { "grad_norm": 0.1579570174217224, "learning_rate": 9.225798601800582e-05, "loss": 0.0219, "step": 66160 }, { "grad_norm": 0.3337019979953766, "learning_rate": 9.225503975292604e-05, "loss": 0.0247, "step": 66170 }, { "grad_norm": 0.16101770102977753, "learning_rate": 9.225209297440764e-05, "loss": 0.0191, "step": 66180 }, { "grad_norm": 0.09927334636449814, "learning_rate": 9.224914568248644e-05, "loss": 0.0196, "step": 66190 }, { "grad_norm": 0.3312061131000519, "learning_rate": 9.22461978771982e-05, "loss": 0.0155, "step": 66200 }, { "grad_norm": 0.2355961799621582, "learning_rate": 9.224324955857879e-05, "loss": 0.0179, "step": 66210 }, { "grad_norm": 0.1142604649066925, "learning_rate": 9.224030072666401e-05, "loss": 0.0163, "step": 66220 }, { "grad_norm": 0.20999504625797272, "learning_rate": 9.223735138148968e-05, "loss": 0.0213, "step": 66230 }, { "grad_norm": 0.18371808528900146, "learning_rate": 9.223440152309166e-05, "loss": 0.0205, "step": 66240 }, { "grad_norm": 0.13350136578083038, "learning_rate": 9.22314511515058e-05, "loss": 0.0211, "step": 66250 }, { "grad_norm": 0.3032406270503998, "learning_rate": 9.222850026676791e-05, "loss": 0.0176, "step": 66260 }, { "grad_norm": 0.12685522437095642, "learning_rate": 9.222554886891389e-05, "loss": 0.0219, "step": 66270 }, { "grad_norm": 0.15151241421699524, "learning_rate": 9.222259695797958e-05, "loss": 0.0221, "step": 66280 }, { "grad_norm": 0.17292356491088867, "learning_rate": 9.221964453400084e-05, "loss": 0.0206, "step": 66290 }, { "grad_norm": 0.09757264703512192, "learning_rate": 9.221669159701358e-05, "loss": 0.0187, "step": 66300 }, { "grad_norm": 0.3368794620037079, "learning_rate": 9.221373814705365e-05, "loss": 0.0178, "step": 66310 }, { "grad_norm": 0.21748515963554382, "learning_rate": 9.221078418415693e-05, "loss": 0.0192, "step": 66320 }, { "grad_norm": 0.20248118042945862, "learning_rate": 9.220782970835934e-05, "loss": 0.0209, "step": 66330 }, { "grad_norm": 0.1502389758825302, "learning_rate": 9.220487471969677e-05, "loss": 0.0213, "step": 66340 }, { "grad_norm": 0.12099188566207886, "learning_rate": 9.220191921820511e-05, "loss": 0.0161, "step": 66350 }, { "grad_norm": 0.1845773607492447, "learning_rate": 9.219896320392031e-05, "loss": 0.0195, "step": 66360 }, { "grad_norm": 0.1425829827785492, "learning_rate": 9.219600667687824e-05, "loss": 0.0167, "step": 66370 }, { "grad_norm": 0.2083635777235031, "learning_rate": 9.219304963711487e-05, "loss": 0.0271, "step": 66380 }, { "grad_norm": 0.16767080128192902, "learning_rate": 9.21900920846661e-05, "loss": 0.022, "step": 66390 }, { "grad_norm": 0.14135263860225677, "learning_rate": 9.218713401956788e-05, "loss": 0.0197, "step": 66400 }, { "grad_norm": 0.3637532591819763, "learning_rate": 9.218417544185615e-05, "loss": 0.0231, "step": 66410 }, { "grad_norm": 0.14327573776245117, "learning_rate": 9.218121635156686e-05, "loss": 0.0265, "step": 66420 }, { "grad_norm": 0.14641515910625458, "learning_rate": 9.217825674873597e-05, "loss": 0.0258, "step": 66430 }, { "grad_norm": 0.1583191603422165, "learning_rate": 9.217529663339942e-05, "loss": 0.0328, "step": 66440 }, { "grad_norm": 0.12381580471992493, "learning_rate": 9.21723360055932e-05, "loss": 0.0227, "step": 66450 }, { "grad_norm": 0.2914864122867584, "learning_rate": 9.216937486535331e-05, "loss": 0.0291, "step": 66460 }, { "grad_norm": 0.25730499625205994, "learning_rate": 9.216641321271567e-05, "loss": 0.0253, "step": 66470 }, { "grad_norm": 0.14431723952293396, "learning_rate": 9.216345104771631e-05, "loss": 0.02, "step": 66480 }, { "grad_norm": 0.3373565673828125, "learning_rate": 9.21604883703912e-05, "loss": 0.0232, "step": 66490 }, { "grad_norm": 0.27025672793388367, "learning_rate": 9.215752518077635e-05, "loss": 0.0242, "step": 66500 }, { "grad_norm": 0.1766459345817566, "learning_rate": 9.215456147890779e-05, "loss": 0.0212, "step": 66510 }, { "grad_norm": 0.12671121954917908, "learning_rate": 9.215159726482148e-05, "loss": 0.0209, "step": 66520 }, { "grad_norm": 0.11077595502138138, "learning_rate": 9.214863253855348e-05, "loss": 0.0201, "step": 66530 }, { "grad_norm": 0.42793893814086914, "learning_rate": 9.214566730013981e-05, "loss": 0.0215, "step": 66540 }, { "grad_norm": 0.3488708734512329, "learning_rate": 9.214270154961647e-05, "loss": 0.0235, "step": 66550 }, { "grad_norm": 0.37369412183761597, "learning_rate": 9.213973528701952e-05, "loss": 0.0218, "step": 66560 }, { "grad_norm": 0.17046615481376648, "learning_rate": 9.213676851238501e-05, "loss": 0.0211, "step": 66570 }, { "grad_norm": 0.1620938926935196, "learning_rate": 9.213380122574898e-05, "loss": 0.0261, "step": 66580 }, { "grad_norm": 0.30036628246307373, "learning_rate": 9.213083342714747e-05, "loss": 0.0248, "step": 66590 }, { "grad_norm": 0.18106955289840698, "learning_rate": 9.212786511661656e-05, "loss": 0.0204, "step": 66600 }, { "grad_norm": 0.18643821775913239, "learning_rate": 9.212489629419232e-05, "loss": 0.026, "step": 66610 }, { "grad_norm": 0.1872742921113968, "learning_rate": 9.212192695991082e-05, "loss": 0.0276, "step": 66620 }, { "grad_norm": 0.12506048381328583, "learning_rate": 9.211895711380812e-05, "loss": 0.0261, "step": 66630 }, { "grad_norm": 0.1443672627210617, "learning_rate": 9.211598675592034e-05, "loss": 0.0208, "step": 66640 }, { "grad_norm": 0.2015349268913269, "learning_rate": 9.211301588628354e-05, "loss": 0.0245, "step": 66650 }, { "grad_norm": 0.22236761450767517, "learning_rate": 9.211004450493386e-05, "loss": 0.022, "step": 66660 }, { "grad_norm": 0.20374096930027008, "learning_rate": 9.210707261190736e-05, "loss": 0.0152, "step": 66670 }, { "grad_norm": 0.18800674378871918, "learning_rate": 9.210410020724018e-05, "loss": 0.0242, "step": 66680 }, { "grad_norm": 0.26418885588645935, "learning_rate": 9.210112729096843e-05, "loss": 0.0239, "step": 66690 }, { "grad_norm": 0.18524955213069916, "learning_rate": 9.209815386312824e-05, "loss": 0.0203, "step": 66700 }, { "grad_norm": 0.09300102293491364, "learning_rate": 9.209517992375572e-05, "loss": 0.0168, "step": 66710 }, { "grad_norm": 0.09627605974674225, "learning_rate": 9.209220547288704e-05, "loss": 0.0162, "step": 66720 }, { "grad_norm": 0.12227671593427658, "learning_rate": 9.20892305105583e-05, "loss": 0.0238, "step": 66730 }, { "grad_norm": 0.1710820496082306, "learning_rate": 9.208625503680568e-05, "loss": 0.0199, "step": 66740 }, { "grad_norm": 0.16145436465740204, "learning_rate": 9.208327905166533e-05, "loss": 0.0198, "step": 66750 }, { "grad_norm": 0.28459829092025757, "learning_rate": 9.208030255517339e-05, "loss": 0.021, "step": 66760 }, { "grad_norm": 0.14626972377300262, "learning_rate": 9.207732554736606e-05, "loss": 0.0217, "step": 66770 }, { "grad_norm": 0.17574578523635864, "learning_rate": 9.20743480282795e-05, "loss": 0.0239, "step": 66780 }, { "grad_norm": 0.21964962780475616, "learning_rate": 9.207136999794987e-05, "loss": 0.022, "step": 66790 }, { "grad_norm": 0.1410832554101944, "learning_rate": 9.206839145641339e-05, "loss": 0.0271, "step": 66800 }, { "grad_norm": 0.12674209475517273, "learning_rate": 9.20654124037062e-05, "loss": 0.0189, "step": 66810 }, { "grad_norm": 0.0947103500366211, "learning_rate": 9.206243283986458e-05, "loss": 0.0186, "step": 66820 }, { "grad_norm": 0.13075129687786102, "learning_rate": 9.205945276492465e-05, "loss": 0.0214, "step": 66830 }, { "grad_norm": 0.11716633290052414, "learning_rate": 9.205647217892268e-05, "loss": 0.0154, "step": 66840 }, { "grad_norm": 0.09978114068508148, "learning_rate": 9.205349108189486e-05, "loss": 0.0171, "step": 66850 }, { "grad_norm": 0.1329030990600586, "learning_rate": 9.205050947387741e-05, "loss": 0.0209, "step": 66860 }, { "grad_norm": 0.17961782217025757, "learning_rate": 9.204752735490658e-05, "loss": 0.0332, "step": 66870 }, { "grad_norm": 0.12800166010856628, "learning_rate": 9.204454472501859e-05, "loss": 0.0168, "step": 66880 }, { "grad_norm": 0.14806106686592102, "learning_rate": 9.204156158424968e-05, "loss": 0.0197, "step": 66890 }, { "grad_norm": 0.12282098084688187, "learning_rate": 9.20385779326361e-05, "loss": 0.0219, "step": 66900 }, { "grad_norm": 0.17244021594524384, "learning_rate": 9.203559377021412e-05, "loss": 0.0241, "step": 66910 }, { "grad_norm": 0.1652834564447403, "learning_rate": 9.203260909701999e-05, "loss": 0.0259, "step": 66920 }, { "grad_norm": 0.27797845005989075, "learning_rate": 9.202962391308996e-05, "loss": 0.0213, "step": 66930 }, { "grad_norm": 0.17541980743408203, "learning_rate": 9.202663821846032e-05, "loss": 0.0257, "step": 66940 }, { "grad_norm": 0.15025940537452698, "learning_rate": 9.202365201316734e-05, "loss": 0.0182, "step": 66950 }, { "grad_norm": 0.3298403024673462, "learning_rate": 9.202066529724734e-05, "loss": 0.0222, "step": 66960 }, { "grad_norm": 0.19619321823120117, "learning_rate": 9.201767807073656e-05, "loss": 0.0204, "step": 66970 }, { "grad_norm": 0.14090394973754883, "learning_rate": 9.201469033367132e-05, "loss": 0.0179, "step": 66980 }, { "grad_norm": 0.16858559846878052, "learning_rate": 9.201170208608791e-05, "loss": 0.0201, "step": 66990 }, { "grad_norm": 0.1787874847650528, "learning_rate": 9.200871332802267e-05, "loss": 0.0195, "step": 67000 }, { "grad_norm": 0.1636018455028534, "learning_rate": 9.200572405951192e-05, "loss": 0.0176, "step": 67010 }, { "grad_norm": 0.17043903470039368, "learning_rate": 9.200273428059193e-05, "loss": 0.0245, "step": 67020 }, { "grad_norm": 0.15163129568099976, "learning_rate": 9.199974399129907e-05, "loss": 0.0227, "step": 67030 }, { "grad_norm": 0.09027677029371262, "learning_rate": 9.199675319166967e-05, "loss": 0.0159, "step": 67040 }, { "grad_norm": 0.34973347187042236, "learning_rate": 9.199376188174007e-05, "loss": 0.029, "step": 67050 }, { "grad_norm": 0.20339122414588928, "learning_rate": 9.199077006154661e-05, "loss": 0.0164, "step": 67060 }, { "grad_norm": 0.13230912387371063, "learning_rate": 9.198777773112566e-05, "loss": 0.0198, "step": 67070 }, { "grad_norm": 0.1417582482099533, "learning_rate": 9.198478489051354e-05, "loss": 0.0171, "step": 67080 }, { "grad_norm": 0.11135435104370117, "learning_rate": 9.198179153974667e-05, "loss": 0.016, "step": 67090 }, { "grad_norm": 0.18774597346782684, "learning_rate": 9.197879767886138e-05, "loss": 0.023, "step": 67100 }, { "grad_norm": 0.10472778230905533, "learning_rate": 9.197580330789407e-05, "loss": 0.0192, "step": 67110 }, { "grad_norm": 0.17578649520874023, "learning_rate": 9.197280842688112e-05, "loss": 0.0198, "step": 67120 }, { "grad_norm": 0.1700524240732193, "learning_rate": 9.196981303585892e-05, "loss": 0.0168, "step": 67130 }, { "grad_norm": 0.10466880351305008, "learning_rate": 9.196681713486386e-05, "loss": 0.0204, "step": 67140 }, { "grad_norm": 0.1483103483915329, "learning_rate": 9.196382072393235e-05, "loss": 0.0167, "step": 67150 }, { "grad_norm": 0.2170431762933731, "learning_rate": 9.196082380310079e-05, "loss": 0.0232, "step": 67160 }, { "grad_norm": 0.13462825119495392, "learning_rate": 9.195782637240562e-05, "loss": 0.014, "step": 67170 }, { "grad_norm": 0.15459099411964417, "learning_rate": 9.195482843188323e-05, "loss": 0.0231, "step": 67180 }, { "grad_norm": 0.12969572842121124, "learning_rate": 9.195182998157007e-05, "loss": 0.0201, "step": 67190 }, { "grad_norm": 0.09750332683324814, "learning_rate": 9.194883102150256e-05, "loss": 0.0248, "step": 67200 }, { "grad_norm": 0.8889363408088684, "learning_rate": 9.194583155171716e-05, "loss": 0.0263, "step": 67210 }, { "grad_norm": 0.25187546014785767, "learning_rate": 9.19428315722503e-05, "loss": 0.0283, "step": 67220 }, { "grad_norm": 0.2571432292461395, "learning_rate": 9.193983108313841e-05, "loss": 0.0372, "step": 67230 }, { "grad_norm": 0.11456728726625443, "learning_rate": 9.193683008441801e-05, "loss": 0.0228, "step": 67240 }, { "grad_norm": 0.22082196176052094, "learning_rate": 9.193382857612552e-05, "loss": 0.0233, "step": 67250 }, { "grad_norm": 0.1493406742811203, "learning_rate": 9.193082655829742e-05, "loss": 0.0265, "step": 67260 }, { "grad_norm": 0.18172499537467957, "learning_rate": 9.192782403097018e-05, "loss": 0.0187, "step": 67270 }, { "grad_norm": 0.15862271189689636, "learning_rate": 9.19248209941803e-05, "loss": 0.0157, "step": 67280 }, { "grad_norm": 0.1109466627240181, "learning_rate": 9.192181744796425e-05, "loss": 0.0166, "step": 67290 }, { "grad_norm": 0.28080305457115173, "learning_rate": 9.191881339235855e-05, "loss": 0.0228, "step": 67300 }, { "grad_norm": 0.2317284792661667, "learning_rate": 9.191580882739967e-05, "loss": 0.0273, "step": 67310 }, { "grad_norm": 0.1842600554227829, "learning_rate": 9.191280375312417e-05, "loss": 0.0216, "step": 67320 }, { "grad_norm": 0.11311449110507965, "learning_rate": 9.19097981695685e-05, "loss": 0.0299, "step": 67330 }, { "grad_norm": 0.22181029617786407, "learning_rate": 9.190679207676922e-05, "loss": 0.0234, "step": 67340 }, { "grad_norm": 0.2832247316837311, "learning_rate": 9.190378547476286e-05, "loss": 0.0178, "step": 67350 }, { "grad_norm": 0.14721286296844482, "learning_rate": 9.190077836358594e-05, "loss": 0.0216, "step": 67360 }, { "grad_norm": 0.11009693890810013, "learning_rate": 9.189777074327501e-05, "loss": 0.0241, "step": 67370 }, { "grad_norm": 0.14942242205142975, "learning_rate": 9.189476261386659e-05, "loss": 0.0238, "step": 67380 }, { "grad_norm": 0.14752130210399628, "learning_rate": 9.189175397539727e-05, "loss": 0.0157, "step": 67390 }, { "grad_norm": 0.17278121411800385, "learning_rate": 9.188874482790358e-05, "loss": 0.0201, "step": 67400 }, { "grad_norm": 0.1812010109424591, "learning_rate": 9.188573517142208e-05, "loss": 0.0247, "step": 67410 }, { "grad_norm": 0.18731097877025604, "learning_rate": 9.188272500598935e-05, "loss": 0.0185, "step": 67420 }, { "grad_norm": 0.16099436581134796, "learning_rate": 9.187971433164198e-05, "loss": 0.0161, "step": 67430 }, { "grad_norm": 0.13191887736320496, "learning_rate": 9.187670314841653e-05, "loss": 0.0151, "step": 67440 }, { "grad_norm": 0.1102467030286789, "learning_rate": 9.187369145634958e-05, "loss": 0.0159, "step": 67450 }, { "grad_norm": 0.1321619302034378, "learning_rate": 9.187067925547779e-05, "loss": 0.0225, "step": 67460 }, { "grad_norm": 0.17482754588127136, "learning_rate": 9.186766654583767e-05, "loss": 0.0236, "step": 67470 }, { "grad_norm": 0.13020852208137512, "learning_rate": 9.18646533274659e-05, "loss": 0.0222, "step": 67480 }, { "grad_norm": 0.6313052177429199, "learning_rate": 9.186163960039906e-05, "loss": 0.0208, "step": 67490 }, { "grad_norm": 0.5463593006134033, "learning_rate": 9.185862536467377e-05, "loss": 0.0183, "step": 67500 }, { "grad_norm": 0.21654576063156128, "learning_rate": 9.185561062032666e-05, "loss": 0.0306, "step": 67510 }, { "grad_norm": 0.2148670107126236, "learning_rate": 9.185259536739437e-05, "loss": 0.0262, "step": 67520 }, { "grad_norm": 0.1759098917245865, "learning_rate": 9.184957960591351e-05, "loss": 0.0202, "step": 67530 }, { "grad_norm": 0.2366378754377365, "learning_rate": 9.184656333592076e-05, "loss": 0.0223, "step": 67540 }, { "grad_norm": 0.18366484344005585, "learning_rate": 9.184354655745277e-05, "loss": 0.0204, "step": 67550 }, { "grad_norm": 0.0912274569272995, "learning_rate": 9.184052927054616e-05, "loss": 0.0199, "step": 67560 }, { "grad_norm": 0.17385613918304443, "learning_rate": 9.183751147523762e-05, "loss": 0.022, "step": 67570 }, { "grad_norm": 0.2482956498861313, "learning_rate": 9.183449317156382e-05, "loss": 0.0256, "step": 67580 }, { "grad_norm": 0.1313188076019287, "learning_rate": 9.183147435956145e-05, "loss": 0.0248, "step": 67590 }, { "grad_norm": 0.14632540941238403, "learning_rate": 9.182845503926713e-05, "loss": 0.0202, "step": 67600 }, { "grad_norm": 0.20074579119682312, "learning_rate": 9.182543521071763e-05, "loss": 0.0172, "step": 67610 }, { "grad_norm": 0.29757505655288696, "learning_rate": 9.182241487394958e-05, "loss": 0.0247, "step": 67620 }, { "grad_norm": 0.17561721801757812, "learning_rate": 9.181939402899969e-05, "loss": 0.0265, "step": 67630 }, { "grad_norm": 0.12029027193784714, "learning_rate": 9.18163726759047e-05, "loss": 0.0239, "step": 67640 }, { "grad_norm": 0.11540521681308746, "learning_rate": 9.181335081470129e-05, "loss": 0.0227, "step": 67650 }, { "grad_norm": 0.10798604041337967, "learning_rate": 9.181032844542619e-05, "loss": 0.021, "step": 67660 }, { "grad_norm": 0.22709177434444427, "learning_rate": 9.180730556811612e-05, "loss": 0.0254, "step": 67670 }, { "grad_norm": 0.11308005452156067, "learning_rate": 9.180428218280783e-05, "loss": 0.0195, "step": 67680 }, { "grad_norm": 0.1326722651720047, "learning_rate": 9.180125828953803e-05, "loss": 0.0154, "step": 67690 }, { "grad_norm": 0.2096606194972992, "learning_rate": 9.179823388834347e-05, "loss": 0.0183, "step": 67700 }, { "grad_norm": 0.15340512990951538, "learning_rate": 9.179520897926092e-05, "loss": 0.0157, "step": 67710 }, { "grad_norm": 0.16219522058963776, "learning_rate": 9.17921835623271e-05, "loss": 0.0184, "step": 67720 }, { "grad_norm": 0.16091755032539368, "learning_rate": 9.178915763757881e-05, "loss": 0.0189, "step": 67730 }, { "grad_norm": 0.14598321914672852, "learning_rate": 9.17861312050528e-05, "loss": 0.0252, "step": 67740 }, { "grad_norm": 0.19639037549495697, "learning_rate": 9.178310426478582e-05, "loss": 0.0205, "step": 67750 }, { "grad_norm": 0.13041169941425323, "learning_rate": 9.17800768168147e-05, "loss": 0.0171, "step": 67760 }, { "grad_norm": 0.1125759556889534, "learning_rate": 9.177704886117619e-05, "loss": 0.0206, "step": 67770 }, { "grad_norm": 0.1158660352230072, "learning_rate": 9.177402039790708e-05, "loss": 0.0218, "step": 67780 }, { "grad_norm": 1.1267551183700562, "learning_rate": 9.177099142704419e-05, "loss": 0.0285, "step": 67790 }, { "grad_norm": 0.1645468920469284, "learning_rate": 9.176796194862433e-05, "loss": 0.0201, "step": 67800 }, { "grad_norm": 0.14569443464279175, "learning_rate": 9.176493196268427e-05, "loss": 0.0213, "step": 67810 }, { "grad_norm": 0.27782562375068665, "learning_rate": 9.176190146926087e-05, "loss": 0.0259, "step": 67820 }, { "grad_norm": 0.19059991836547852, "learning_rate": 9.175887046839093e-05, "loss": 0.018, "step": 67830 }, { "grad_norm": 0.18176235258579254, "learning_rate": 9.175583896011131e-05, "loss": 0.017, "step": 67840 }, { "grad_norm": 0.16220463812351227, "learning_rate": 9.17528069444588e-05, "loss": 0.0272, "step": 67850 }, { "grad_norm": 0.3431028127670288, "learning_rate": 9.174977442147027e-05, "loss": 0.0198, "step": 67860 }, { "grad_norm": 0.17483294010162354, "learning_rate": 9.174674139118256e-05, "loss": 0.0216, "step": 67870 }, { "grad_norm": 0.1603832095861435, "learning_rate": 9.174370785363253e-05, "loss": 0.0208, "step": 67880 }, { "grad_norm": 0.1427345722913742, "learning_rate": 9.174067380885703e-05, "loss": 0.0216, "step": 67890 }, { "grad_norm": 0.1766911894083023, "learning_rate": 9.173763925689294e-05, "loss": 0.016, "step": 67900 }, { "grad_norm": 0.1248970553278923, "learning_rate": 9.173460419777714e-05, "loss": 0.0177, "step": 67910 }, { "grad_norm": 0.09752269834280014, "learning_rate": 9.173156863154646e-05, "loss": 0.017, "step": 67920 }, { "grad_norm": 0.09385088831186295, "learning_rate": 9.172853255823785e-05, "loss": 0.0163, "step": 67930 }, { "grad_norm": 0.17045439779758453, "learning_rate": 9.172549597788815e-05, "loss": 0.0202, "step": 67940 }, { "grad_norm": 0.1931534707546234, "learning_rate": 9.17224588905343e-05, "loss": 0.0252, "step": 67950 }, { "grad_norm": 0.2421119660139084, "learning_rate": 9.171942129621317e-05, "loss": 0.0162, "step": 67960 }, { "grad_norm": 0.43636104464530945, "learning_rate": 9.171638319496167e-05, "loss": 0.0187, "step": 67970 }, { "grad_norm": 0.12154753506183624, "learning_rate": 9.171334458681676e-05, "loss": 0.0172, "step": 67980 }, { "grad_norm": 0.1913508027791977, "learning_rate": 9.17103054718153e-05, "loss": 0.0293, "step": 67990 }, { "grad_norm": 0.17791225016117096, "learning_rate": 9.170726584999425e-05, "loss": 0.0244, "step": 68000 }, { "grad_norm": 0.15466539561748505, "learning_rate": 9.170422572139055e-05, "loss": 0.0214, "step": 68010 }, { "grad_norm": 0.19000151753425598, "learning_rate": 9.170118508604115e-05, "loss": 0.0168, "step": 68020 }, { "grad_norm": 0.09997840225696564, "learning_rate": 9.169814394398296e-05, "loss": 0.0172, "step": 68030 }, { "grad_norm": 0.21858441829681396, "learning_rate": 9.169510229525294e-05, "loss": 0.02, "step": 68040 }, { "grad_norm": 0.2114977240562439, "learning_rate": 9.169206013988807e-05, "loss": 0.0237, "step": 68050 }, { "grad_norm": 0.11201635003089905, "learning_rate": 9.168901747792531e-05, "loss": 0.0175, "step": 68060 }, { "grad_norm": 0.13828857243061066, "learning_rate": 9.168597430940162e-05, "loss": 0.0207, "step": 68070 }, { "grad_norm": 0.14207789301872253, "learning_rate": 9.1682930634354e-05, "loss": 0.0154, "step": 68080 }, { "grad_norm": 0.12694571912288666, "learning_rate": 9.16798864528194e-05, "loss": 0.0164, "step": 68090 }, { "grad_norm": 0.14256051182746887, "learning_rate": 9.167684176483483e-05, "loss": 0.0259, "step": 68100 }, { "grad_norm": 0.36627963185310364, "learning_rate": 9.167379657043728e-05, "loss": 0.0228, "step": 68110 }, { "grad_norm": 0.18978707492351532, "learning_rate": 9.167075086966376e-05, "loss": 0.0182, "step": 68120 }, { "grad_norm": 0.10285597294569016, "learning_rate": 9.166770466255128e-05, "loss": 0.0221, "step": 68130 }, { "grad_norm": 0.19323459267616272, "learning_rate": 9.166465794913685e-05, "loss": 0.0261, "step": 68140 }, { "grad_norm": 0.18463799357414246, "learning_rate": 9.166161072945747e-05, "loss": 0.0246, "step": 68150 }, { "grad_norm": 0.1949373483657837, "learning_rate": 9.16585630035502e-05, "loss": 0.022, "step": 68160 }, { "grad_norm": 0.21945656836032867, "learning_rate": 9.165551477145204e-05, "loss": 0.0271, "step": 68170 }, { "grad_norm": 0.09969337284564972, "learning_rate": 9.165246603320006e-05, "loss": 0.0201, "step": 68180 }, { "grad_norm": 0.1726091057062149, "learning_rate": 9.16494167888313e-05, "loss": 0.0223, "step": 68190 }, { "grad_norm": 0.20119506120681763, "learning_rate": 9.164636703838278e-05, "loss": 0.023, "step": 68200 }, { "grad_norm": 0.25293922424316406, "learning_rate": 9.16433167818916e-05, "loss": 0.0242, "step": 68210 }, { "grad_norm": 0.18542975187301636, "learning_rate": 9.16402660193948e-05, "loss": 0.0218, "step": 68220 }, { "grad_norm": 0.12202634662389755, "learning_rate": 9.163721475092945e-05, "loss": 0.0219, "step": 68230 }, { "grad_norm": 0.11040211468935013, "learning_rate": 9.163416297653262e-05, "loss": 0.0186, "step": 68240 }, { "grad_norm": 0.15414437651634216, "learning_rate": 9.16311106962414e-05, "loss": 0.0203, "step": 68250 }, { "grad_norm": 0.15473616123199463, "learning_rate": 9.16280579100929e-05, "loss": 0.0242, "step": 68260 }, { "grad_norm": 0.18646977841854095, "learning_rate": 9.162500461812418e-05, "loss": 0.0222, "step": 68270 }, { "grad_norm": 0.20937250554561615, "learning_rate": 9.162195082037234e-05, "loss": 0.0285, "step": 68280 }, { "grad_norm": 0.18720227479934692, "learning_rate": 9.161889651687452e-05, "loss": 0.0253, "step": 68290 }, { "grad_norm": 0.2149318903684616, "learning_rate": 9.16158417076678e-05, "loss": 0.0181, "step": 68300 }, { "grad_norm": 0.1833752989768982, "learning_rate": 9.161278639278932e-05, "loss": 0.0209, "step": 68310 }, { "grad_norm": 0.18082790076732635, "learning_rate": 9.160973057227619e-05, "loss": 0.0163, "step": 68320 }, { "grad_norm": 0.2968289852142334, "learning_rate": 9.160667424616555e-05, "loss": 0.0237, "step": 68330 }, { "grad_norm": 0.17682470381259918, "learning_rate": 9.160361741449453e-05, "loss": 0.0183, "step": 68340 }, { "grad_norm": 0.18268544971942902, "learning_rate": 9.16005600773003e-05, "loss": 0.0164, "step": 68350 }, { "grad_norm": 0.1282242387533188, "learning_rate": 9.159750223461996e-05, "loss": 0.0133, "step": 68360 }, { "grad_norm": 0.13066713511943817, "learning_rate": 9.15944438864907e-05, "loss": 0.0186, "step": 68370 }, { "grad_norm": 0.22345110774040222, "learning_rate": 9.159138503294968e-05, "loss": 0.0239, "step": 68380 }, { "grad_norm": 0.10718050599098206, "learning_rate": 9.158832567403406e-05, "loss": 0.0204, "step": 68390 }, { "grad_norm": 0.10850881785154343, "learning_rate": 9.158526580978102e-05, "loss": 0.0209, "step": 68400 }, { "grad_norm": 0.24128560721874237, "learning_rate": 9.158220544022773e-05, "loss": 0.0161, "step": 68410 }, { "grad_norm": 0.13009250164031982, "learning_rate": 9.157914456541138e-05, "loss": 0.0178, "step": 68420 }, { "grad_norm": 0.1314106434583664, "learning_rate": 9.157608318536919e-05, "loss": 0.0199, "step": 68430 }, { "grad_norm": 0.204018235206604, "learning_rate": 9.157302130013831e-05, "loss": 0.0226, "step": 68440 }, { "grad_norm": 0.11065395921468735, "learning_rate": 9.156995890975599e-05, "loss": 0.0196, "step": 68450 }, { "grad_norm": 0.24988238513469696, "learning_rate": 9.156689601425942e-05, "loss": 0.0165, "step": 68460 }, { "grad_norm": 0.16792897880077362, "learning_rate": 9.15638326136858e-05, "loss": 0.0253, "step": 68470 }, { "grad_norm": 0.10283096134662628, "learning_rate": 9.156076870807239e-05, "loss": 0.0205, "step": 68480 }, { "grad_norm": 0.09701838344335556, "learning_rate": 9.155770429745638e-05, "loss": 0.0249, "step": 68490 }, { "grad_norm": 0.19534409046173096, "learning_rate": 9.155463938187505e-05, "loss": 0.0167, "step": 68500 }, { "grad_norm": 0.35336828231811523, "learning_rate": 9.15515739613656e-05, "loss": 0.0191, "step": 68510 }, { "grad_norm": 0.24376098811626434, "learning_rate": 9.154850803596529e-05, "loss": 0.021, "step": 68520 }, { "grad_norm": 0.16710910201072693, "learning_rate": 9.15454416057114e-05, "loss": 0.0162, "step": 68530 }, { "grad_norm": 0.15846474468708038, "learning_rate": 9.154237467064115e-05, "loss": 0.0348, "step": 68540 }, { "grad_norm": 0.1828782558441162, "learning_rate": 9.153930723079184e-05, "loss": 0.0317, "step": 68550 }, { "grad_norm": 0.42287519574165344, "learning_rate": 9.153623928620072e-05, "loss": 0.0195, "step": 68560 }, { "grad_norm": 0.12568846344947815, "learning_rate": 9.153317083690507e-05, "loss": 0.0203, "step": 68570 }, { "grad_norm": 0.20101559162139893, "learning_rate": 9.153010188294219e-05, "loss": 0.0176, "step": 68580 }, { "grad_norm": 0.15164902806282043, "learning_rate": 9.152703242434936e-05, "loss": 0.0252, "step": 68590 }, { "grad_norm": 0.37038859724998474, "learning_rate": 9.152396246116387e-05, "loss": 0.0196, "step": 68600 }, { "grad_norm": 0.20466557145118713, "learning_rate": 9.152089199342304e-05, "loss": 0.0274, "step": 68610 }, { "grad_norm": 0.22567644715309143, "learning_rate": 9.151782102116417e-05, "loss": 0.0198, "step": 68620 }, { "grad_norm": 0.16235531866550446, "learning_rate": 9.151474954442457e-05, "loss": 0.0168, "step": 68630 }, { "grad_norm": 0.3403942883014679, "learning_rate": 9.151167756324157e-05, "loss": 0.0202, "step": 68640 }, { "grad_norm": 0.22586314380168915, "learning_rate": 9.150860507765248e-05, "loss": 0.0227, "step": 68650 }, { "grad_norm": 0.12801121175289154, "learning_rate": 9.150553208769466e-05, "loss": 0.0204, "step": 68660 }, { "grad_norm": 0.167381152510643, "learning_rate": 9.150245859340544e-05, "loss": 0.0167, "step": 68670 }, { "grad_norm": 0.19274817407131195, "learning_rate": 9.149938459482218e-05, "loss": 0.0221, "step": 68680 }, { "grad_norm": 0.20827682316303253, "learning_rate": 9.149631009198218e-05, "loss": 0.0222, "step": 68690 }, { "grad_norm": 0.16786760091781616, "learning_rate": 9.149323508492286e-05, "loss": 0.0269, "step": 68700 }, { "grad_norm": 0.19545161724090576, "learning_rate": 9.149015957368155e-05, "loss": 0.0168, "step": 68710 }, { "grad_norm": 0.17266379296779633, "learning_rate": 9.148708355829562e-05, "loss": 0.0176, "step": 68720 }, { "grad_norm": 0.13534019887447357, "learning_rate": 9.148400703880246e-05, "loss": 0.0201, "step": 68730 }, { "grad_norm": 0.12389761209487915, "learning_rate": 9.148093001523945e-05, "loss": 0.0225, "step": 68740 }, { "grad_norm": 0.4796209931373596, "learning_rate": 9.147785248764397e-05, "loss": 0.0261, "step": 68750 }, { "grad_norm": 0.23014041781425476, "learning_rate": 9.147477445605343e-05, "loss": 0.0195, "step": 68760 }, { "grad_norm": 0.2004476636648178, "learning_rate": 9.147169592050522e-05, "loss": 0.0223, "step": 68770 }, { "grad_norm": 0.3709953725337982, "learning_rate": 9.146861688103674e-05, "loss": 0.0236, "step": 68780 }, { "grad_norm": 0.12221518903970718, "learning_rate": 9.146553733768542e-05, "loss": 0.0219, "step": 68790 }, { "grad_norm": 0.14898648858070374, "learning_rate": 9.146245729048865e-05, "loss": 0.0117, "step": 68800 }, { "grad_norm": 0.09712889045476913, "learning_rate": 9.145937673948391e-05, "loss": 0.0246, "step": 68810 }, { "grad_norm": 0.13346725702285767, "learning_rate": 9.145629568470857e-05, "loss": 0.0216, "step": 68820 }, { "grad_norm": 0.14594903588294983, "learning_rate": 9.145321412620011e-05, "loss": 0.025, "step": 68830 }, { "grad_norm": 0.11471019685268402, "learning_rate": 9.145013206399597e-05, "loss": 0.0212, "step": 68840 }, { "grad_norm": 0.152347132563591, "learning_rate": 9.144704949813357e-05, "loss": 0.0196, "step": 68850 }, { "grad_norm": 0.20221900939941406, "learning_rate": 9.14439664286504e-05, "loss": 0.0218, "step": 68860 }, { "grad_norm": 0.18948855996131897, "learning_rate": 9.144088285558389e-05, "loss": 0.0217, "step": 68870 }, { "grad_norm": 0.26431727409362793, "learning_rate": 9.143779877897155e-05, "loss": 0.0177, "step": 68880 }, { "grad_norm": 0.28399112820625305, "learning_rate": 9.143471419885082e-05, "loss": 0.0271, "step": 68890 }, { "grad_norm": 0.13031522929668427, "learning_rate": 9.14316291152592e-05, "loss": 0.018, "step": 68900 }, { "grad_norm": 0.16027940809726715, "learning_rate": 9.142854352823416e-05, "loss": 0.0235, "step": 68910 }, { "grad_norm": 0.18183623254299164, "learning_rate": 9.14254574378132e-05, "loss": 0.0218, "step": 68920 }, { "grad_norm": 0.1976381093263626, "learning_rate": 9.142237084403382e-05, "loss": 0.0189, "step": 68930 }, { "grad_norm": 0.2091815322637558, "learning_rate": 9.141928374693353e-05, "loss": 0.0228, "step": 68940 }, { "grad_norm": 0.20595292747020721, "learning_rate": 9.141619614654984e-05, "loss": 0.0229, "step": 68950 }, { "grad_norm": 0.2619512677192688, "learning_rate": 9.141310804292025e-05, "loss": 0.0196, "step": 68960 }, { "grad_norm": 0.2594659626483917, "learning_rate": 9.14100194360823e-05, "loss": 0.0371, "step": 68970 }, { "grad_norm": 0.14121849834918976, "learning_rate": 9.140693032607353e-05, "loss": 0.019, "step": 68980 }, { "grad_norm": 0.24860051274299622, "learning_rate": 9.140384071293145e-05, "loss": 0.0238, "step": 68990 }, { "grad_norm": 0.11379802227020264, "learning_rate": 9.140075059669362e-05, "loss": 0.0247, "step": 69000 }, { "grad_norm": 0.16391032934188843, "learning_rate": 9.139765997739758e-05, "loss": 0.0184, "step": 69010 }, { "grad_norm": 0.6283875107765198, "learning_rate": 9.139456885508086e-05, "loss": 0.0222, "step": 69020 }, { "grad_norm": 0.14522424340248108, "learning_rate": 9.139147722978108e-05, "loss": 0.0162, "step": 69030 }, { "grad_norm": 0.40466076135635376, "learning_rate": 9.138838510153576e-05, "loss": 0.0208, "step": 69040 }, { "grad_norm": 0.3124370276927948, "learning_rate": 9.138529247038247e-05, "loss": 0.0258, "step": 69050 }, { "grad_norm": 0.17028068006038666, "learning_rate": 9.138219933635883e-05, "loss": 0.0211, "step": 69060 }, { "grad_norm": 0.13437101244926453, "learning_rate": 9.13791056995024e-05, "loss": 0.0219, "step": 69070 }, { "grad_norm": 0.28901711106300354, "learning_rate": 9.137601155985073e-05, "loss": 0.0226, "step": 69080 }, { "grad_norm": 0.2270914912223816, "learning_rate": 9.137291691744147e-05, "loss": 0.0234, "step": 69090 }, { "grad_norm": 0.09442440420389175, "learning_rate": 9.136982177231221e-05, "loss": 0.018, "step": 69100 }, { "grad_norm": 0.21041260659694672, "learning_rate": 9.136672612450057e-05, "loss": 0.0197, "step": 69110 }, { "grad_norm": 0.5034667253494263, "learning_rate": 9.136362997404413e-05, "loss": 0.0233, "step": 69120 }, { "grad_norm": 0.14140628278255463, "learning_rate": 9.136053332098054e-05, "loss": 0.0231, "step": 69130 }, { "grad_norm": 0.1831071823835373, "learning_rate": 9.135743616534742e-05, "loss": 0.0256, "step": 69140 }, { "grad_norm": 0.1120910570025444, "learning_rate": 9.13543385071824e-05, "loss": 0.0211, "step": 69150 }, { "grad_norm": 0.25620877742767334, "learning_rate": 9.135124034652313e-05, "loss": 0.0259, "step": 69160 }, { "grad_norm": 0.1368495374917984, "learning_rate": 9.134814168340724e-05, "loss": 0.0219, "step": 69170 }, { "grad_norm": 0.40961530804634094, "learning_rate": 9.134504251787239e-05, "loss": 0.0268, "step": 69180 }, { "grad_norm": 0.18552365899085999, "learning_rate": 9.134194284995624e-05, "loss": 0.0254, "step": 69190 }, { "grad_norm": 0.17332492768764496, "learning_rate": 9.133884267969646e-05, "loss": 0.019, "step": 69200 }, { "grad_norm": 0.14797765016555786, "learning_rate": 9.133574200713068e-05, "loss": 0.0193, "step": 69210 }, { "grad_norm": 0.15218272805213928, "learning_rate": 9.133264083229665e-05, "loss": 0.0183, "step": 69220 }, { "grad_norm": 0.11140969395637512, "learning_rate": 9.132953915523198e-05, "loss": 0.016, "step": 69230 }, { "grad_norm": 0.13671565055847168, "learning_rate": 9.132643697597439e-05, "loss": 0.0278, "step": 69240 }, { "grad_norm": 0.08881943672895432, "learning_rate": 9.132333429456158e-05, "loss": 0.0191, "step": 69250 }, { "grad_norm": 0.11296487599611282, "learning_rate": 9.132023111103125e-05, "loss": 0.0163, "step": 69260 }, { "grad_norm": 0.1097300574183464, "learning_rate": 9.131712742542108e-05, "loss": 0.0172, "step": 69270 }, { "grad_norm": 0.14206480979919434, "learning_rate": 9.13140232377688e-05, "loss": 0.0155, "step": 69280 }, { "grad_norm": 0.12313711643218994, "learning_rate": 9.131091854811216e-05, "loss": 0.0135, "step": 69290 }, { "grad_norm": 0.10569386184215546, "learning_rate": 9.130781335648883e-05, "loss": 0.0195, "step": 69300 }, { "grad_norm": 0.18129457533359528, "learning_rate": 9.130470766293657e-05, "loss": 0.0195, "step": 69310 }, { "grad_norm": 0.13372328877449036, "learning_rate": 9.130160146749312e-05, "loss": 0.0193, "step": 69320 }, { "grad_norm": 0.1687345653772354, "learning_rate": 9.129849477019621e-05, "loss": 0.0219, "step": 69330 }, { "grad_norm": 0.14242029190063477, "learning_rate": 9.12953875710836e-05, "loss": 0.0218, "step": 69340 }, { "grad_norm": 0.3001305162906647, "learning_rate": 9.129227987019304e-05, "loss": 0.0198, "step": 69350 }, { "grad_norm": 0.11659961938858032, "learning_rate": 9.12891716675623e-05, "loss": 0.0155, "step": 69360 }, { "grad_norm": 0.3550879955291748, "learning_rate": 9.128606296322915e-05, "loss": 0.0262, "step": 69370 }, { "grad_norm": 0.23685438930988312, "learning_rate": 9.128295375723133e-05, "loss": 0.0218, "step": 69380 }, { "grad_norm": 0.3162209093570709, "learning_rate": 9.127984404960666e-05, "loss": 0.0234, "step": 69390 }, { "grad_norm": 0.19070900976657867, "learning_rate": 9.12767338403929e-05, "loss": 0.0178, "step": 69400 }, { "grad_norm": 0.14080728590488434, "learning_rate": 9.127362312962786e-05, "loss": 0.0195, "step": 69410 }, { "grad_norm": 0.2946014702320099, "learning_rate": 9.127051191734933e-05, "loss": 0.0258, "step": 69420 }, { "grad_norm": 0.1983547806739807, "learning_rate": 9.126740020359512e-05, "loss": 0.0214, "step": 69430 }, { "grad_norm": 0.20860448479652405, "learning_rate": 9.126428798840302e-05, "loss": 0.0143, "step": 69440 }, { "grad_norm": 0.2029859870672226, "learning_rate": 9.126117527181086e-05, "loss": 0.0165, "step": 69450 }, { "grad_norm": 0.12468814849853516, "learning_rate": 9.125806205385648e-05, "loss": 0.0216, "step": 69460 }, { "grad_norm": 0.2026088386774063, "learning_rate": 9.125494833457769e-05, "loss": 0.0203, "step": 69470 }, { "grad_norm": 0.1846483051776886, "learning_rate": 9.125183411401233e-05, "loss": 0.0222, "step": 69480 }, { "grad_norm": 0.11536048352718353, "learning_rate": 9.124871939219821e-05, "loss": 0.0179, "step": 69490 }, { "grad_norm": 0.17499208450317383, "learning_rate": 9.124560416917323e-05, "loss": 0.0196, "step": 69500 }, { "grad_norm": 0.1498529613018036, "learning_rate": 9.12424884449752e-05, "loss": 0.0192, "step": 69510 }, { "grad_norm": 0.11415967345237732, "learning_rate": 9.1239372219642e-05, "loss": 0.0197, "step": 69520 }, { "grad_norm": 1.9315659999847412, "learning_rate": 9.123625549321151e-05, "loss": 0.0227, "step": 69530 }, { "grad_norm": 0.15235739946365356, "learning_rate": 9.123313826572156e-05, "loss": 0.0208, "step": 69540 }, { "grad_norm": 0.2592880427837372, "learning_rate": 9.123002053721005e-05, "loss": 0.0203, "step": 69550 }, { "grad_norm": 0.23010766506195068, "learning_rate": 9.122690230771488e-05, "loss": 0.027, "step": 69560 }, { "grad_norm": 0.1374623030424118, "learning_rate": 9.122378357727393e-05, "loss": 0.0194, "step": 69570 }, { "grad_norm": 0.24814869463443756, "learning_rate": 9.122066434592506e-05, "loss": 0.0188, "step": 69580 }, { "grad_norm": 0.14844316244125366, "learning_rate": 9.12175446137062e-05, "loss": 0.0176, "step": 69590 }, { "grad_norm": 0.2311951220035553, "learning_rate": 9.121442438065528e-05, "loss": 0.0204, "step": 69600 }, { "grad_norm": 0.21101120114326477, "learning_rate": 9.121130364681016e-05, "loss": 0.0163, "step": 69610 }, { "grad_norm": 0.1258840411901474, "learning_rate": 9.120818241220882e-05, "loss": 0.0198, "step": 69620 }, { "grad_norm": 0.3965049982070923, "learning_rate": 9.120506067688914e-05, "loss": 0.029, "step": 69630 }, { "grad_norm": 0.12317975610494614, "learning_rate": 9.120193844088908e-05, "loss": 0.0165, "step": 69640 }, { "grad_norm": 0.20457324385643005, "learning_rate": 9.119881570424656e-05, "loss": 0.0187, "step": 69650 }, { "grad_norm": 0.2546406090259552, "learning_rate": 9.119569246699952e-05, "loss": 0.0216, "step": 69660 }, { "grad_norm": 0.1839415431022644, "learning_rate": 9.119256872918594e-05, "loss": 0.0217, "step": 69670 }, { "grad_norm": 0.16365206241607666, "learning_rate": 9.118944449084377e-05, "loss": 0.025, "step": 69680 }, { "grad_norm": 0.2699916958808899, "learning_rate": 9.118631975201092e-05, "loss": 0.021, "step": 69690 }, { "grad_norm": 0.41497552394866943, "learning_rate": 9.118319451272544e-05, "loss": 0.0229, "step": 69700 }, { "grad_norm": 0.13679660856723785, "learning_rate": 9.118006877302524e-05, "loss": 0.0165, "step": 69710 }, { "grad_norm": 0.20054395496845245, "learning_rate": 9.117694253294833e-05, "loss": 0.0268, "step": 69720 }, { "grad_norm": 0.11581552028656006, "learning_rate": 9.117381579253268e-05, "loss": 0.0203, "step": 69730 }, { "grad_norm": 0.17281436920166016, "learning_rate": 9.11706885518163e-05, "loss": 0.0198, "step": 69740 }, { "grad_norm": 0.12105272710323334, "learning_rate": 9.11675608108372e-05, "loss": 0.0248, "step": 69750 }, { "grad_norm": 0.1476944386959076, "learning_rate": 9.116443256963333e-05, "loss": 0.0222, "step": 69760 }, { "grad_norm": 0.09959197044372559, "learning_rate": 9.116130382824277e-05, "loss": 0.0189, "step": 69770 }, { "grad_norm": 0.2415970414876938, "learning_rate": 9.11581745867035e-05, "loss": 0.0221, "step": 69780 }, { "grad_norm": 0.17906765639781952, "learning_rate": 9.115504484505355e-05, "loss": 0.0178, "step": 69790 }, { "grad_norm": 0.1310759037733078, "learning_rate": 9.115191460333095e-05, "loss": 0.0185, "step": 69800 }, { "grad_norm": 0.1329449862241745, "learning_rate": 9.114878386157372e-05, "loss": 0.0194, "step": 69810 }, { "grad_norm": 0.18201687932014465, "learning_rate": 9.114565261981992e-05, "loss": 0.0174, "step": 69820 }, { "grad_norm": 0.23300759494304657, "learning_rate": 9.114252087810759e-05, "loss": 0.0202, "step": 69830 }, { "grad_norm": 0.17119774222373962, "learning_rate": 9.11393886364748e-05, "loss": 0.0263, "step": 69840 }, { "grad_norm": 0.20972028374671936, "learning_rate": 9.113625589495958e-05, "loss": 0.0288, "step": 69850 }, { "grad_norm": 0.2410273253917694, "learning_rate": 9.113312265360003e-05, "loss": 0.0231, "step": 69860 }, { "grad_norm": 0.15020549297332764, "learning_rate": 9.112998891243419e-05, "loss": 0.0214, "step": 69870 }, { "grad_norm": 0.2977784276008606, "learning_rate": 9.112685467150015e-05, "loss": 0.0259, "step": 69880 }, { "grad_norm": 0.1377464383840561, "learning_rate": 9.1123719930836e-05, "loss": 0.0189, "step": 69890 }, { "grad_norm": 0.2068646103143692, "learning_rate": 9.112058469047982e-05, "loss": 0.0228, "step": 69900 }, { "grad_norm": 0.2648332417011261, "learning_rate": 9.111744895046972e-05, "loss": 0.0171, "step": 69910 }, { "grad_norm": 0.12379566580057144, "learning_rate": 9.11143127108438e-05, "loss": 0.027, "step": 69920 }, { "grad_norm": 0.2469799816608429, "learning_rate": 9.111117597164015e-05, "loss": 0.016, "step": 69930 }, { "grad_norm": 0.12172913551330566, "learning_rate": 9.11080387328969e-05, "loss": 0.0184, "step": 69940 }, { "grad_norm": 0.09665952622890472, "learning_rate": 9.110490099465216e-05, "loss": 0.0167, "step": 69950 }, { "grad_norm": 0.09132298082113266, "learning_rate": 9.11017627569441e-05, "loss": 0.0191, "step": 69960 }, { "grad_norm": 0.14052365720272064, "learning_rate": 9.109862401981077e-05, "loss": 0.0167, "step": 69970 }, { "grad_norm": 0.38374197483062744, "learning_rate": 9.109548478329037e-05, "loss": 0.024, "step": 69980 }, { "grad_norm": 0.15516449511051178, "learning_rate": 9.109234504742103e-05, "loss": 0.0198, "step": 69990 }, { "grad_norm": 0.095101498067379, "learning_rate": 9.10892048122409e-05, "loss": 0.0144, "step": 70000 }, { "grad_norm": 0.17052797973155975, "learning_rate": 9.108606407778815e-05, "loss": 0.0219, "step": 70010 }, { "grad_norm": 0.11891573667526245, "learning_rate": 9.108292284410091e-05, "loss": 0.0205, "step": 70020 }, { "grad_norm": 0.18384812772274017, "learning_rate": 9.107978111121737e-05, "loss": 0.0193, "step": 70030 }, { "grad_norm": 0.07306130230426788, "learning_rate": 9.107663887917571e-05, "loss": 0.0148, "step": 70040 }, { "grad_norm": 0.15693749487400055, "learning_rate": 9.10734961480141e-05, "loss": 0.0196, "step": 70050 }, { "grad_norm": 0.17927604913711548, "learning_rate": 9.107035291777075e-05, "loss": 0.0195, "step": 70060 }, { "grad_norm": 0.137942835688591, "learning_rate": 9.106720918848381e-05, "loss": 0.0194, "step": 70070 }, { "grad_norm": 0.11953535676002502, "learning_rate": 9.106406496019152e-05, "loss": 0.0137, "step": 70080 }, { "grad_norm": 0.14731262624263763, "learning_rate": 9.106092023293206e-05, "loss": 0.0222, "step": 70090 }, { "grad_norm": 0.12906426191329956, "learning_rate": 9.105777500674366e-05, "loss": 0.0238, "step": 70100 }, { "grad_norm": 0.22223380208015442, "learning_rate": 9.105462928166452e-05, "loss": 0.0181, "step": 70110 }, { "grad_norm": 0.2068043351173401, "learning_rate": 9.10514830577329e-05, "loss": 0.0184, "step": 70120 }, { "grad_norm": 0.1502375453710556, "learning_rate": 9.104833633498697e-05, "loss": 0.016, "step": 70130 }, { "grad_norm": 0.2818756401538849, "learning_rate": 9.104518911346501e-05, "loss": 0.0276, "step": 70140 }, { "grad_norm": 0.1598767340183258, "learning_rate": 9.104204139320524e-05, "loss": 0.0179, "step": 70150 }, { "grad_norm": 0.17142026126384735, "learning_rate": 9.103889317424593e-05, "loss": 0.0191, "step": 70160 }, { "grad_norm": 0.10859014838933945, "learning_rate": 9.103574445662532e-05, "loss": 0.0247, "step": 70170 }, { "grad_norm": 0.12020556628704071, "learning_rate": 9.103259524038166e-05, "loss": 0.0193, "step": 70180 }, { "grad_norm": 0.1395062953233719, "learning_rate": 9.102944552555323e-05, "loss": 0.0183, "step": 70190 }, { "grad_norm": 0.25389325618743896, "learning_rate": 9.10262953121783e-05, "loss": 0.0226, "step": 70200 }, { "grad_norm": 0.09926868230104446, "learning_rate": 9.102314460029516e-05, "loss": 0.0214, "step": 70210 }, { "grad_norm": 0.13493570685386658, "learning_rate": 9.101999338994207e-05, "loss": 0.0163, "step": 70220 }, { "grad_norm": 0.1257234811782837, "learning_rate": 9.101684168115732e-05, "loss": 0.0204, "step": 70230 }, { "grad_norm": 0.24236753582954407, "learning_rate": 9.101368947397923e-05, "loss": 0.0228, "step": 70240 }, { "grad_norm": 0.19736725091934204, "learning_rate": 9.101053676844608e-05, "loss": 0.0215, "step": 70250 }, { "grad_norm": 0.19104790687561035, "learning_rate": 9.10073835645962e-05, "loss": 0.0203, "step": 70260 }, { "grad_norm": 0.18274523317813873, "learning_rate": 9.100422986246788e-05, "loss": 0.0221, "step": 70270 }, { "grad_norm": 0.1600748598575592, "learning_rate": 9.100107566209946e-05, "loss": 0.0137, "step": 70280 }, { "grad_norm": 0.20736274123191833, "learning_rate": 9.099792096352926e-05, "loss": 0.0258, "step": 70290 }, { "grad_norm": 0.1865999847650528, "learning_rate": 9.099476576679563e-05, "loss": 0.026, "step": 70300 }, { "grad_norm": 0.23180927336215973, "learning_rate": 9.099161007193687e-05, "loss": 0.0248, "step": 70310 }, { "grad_norm": 0.13167113065719604, "learning_rate": 9.098845387899133e-05, "loss": 0.0202, "step": 70320 }, { "grad_norm": 0.19615383446216583, "learning_rate": 9.09852971879974e-05, "loss": 0.0205, "step": 70330 }, { "grad_norm": 0.15002533793449402, "learning_rate": 9.09821399989934e-05, "loss": 0.0207, "step": 70340 }, { "grad_norm": 0.09374526143074036, "learning_rate": 9.09789823120177e-05, "loss": 0.0156, "step": 70350 }, { "grad_norm": 0.11704961210489273, "learning_rate": 9.09758241271087e-05, "loss": 0.014, "step": 70360 }, { "grad_norm": 0.2995206415653229, "learning_rate": 9.097266544430472e-05, "loss": 0.0203, "step": 70370 }, { "grad_norm": 0.2819269597530365, "learning_rate": 9.096950626364418e-05, "loss": 0.0194, "step": 70380 }, { "grad_norm": 0.14023588597774506, "learning_rate": 9.096634658516547e-05, "loss": 0.0188, "step": 70390 }, { "grad_norm": 0.26842957735061646, "learning_rate": 9.096318640890695e-05, "loss": 0.0188, "step": 70400 }, { "grad_norm": 0.09432550519704819, "learning_rate": 9.096002573490707e-05, "loss": 0.0175, "step": 70410 }, { "grad_norm": 0.1276616007089615, "learning_rate": 9.095686456320418e-05, "loss": 0.0239, "step": 70420 }, { "grad_norm": 0.17792007327079773, "learning_rate": 9.095370289383672e-05, "loss": 0.0168, "step": 70430 }, { "grad_norm": 0.15289057791233063, "learning_rate": 9.09505407268431e-05, "loss": 0.0282, "step": 70440 }, { "grad_norm": 0.13080476224422455, "learning_rate": 9.094737806226176e-05, "loss": 0.0198, "step": 70450 }, { "grad_norm": 0.2920226752758026, "learning_rate": 9.09442149001311e-05, "loss": 0.0184, "step": 70460 }, { "grad_norm": 0.2166767567396164, "learning_rate": 9.094105124048958e-05, "loss": 0.0218, "step": 70470 }, { "grad_norm": 0.19544664025306702, "learning_rate": 9.093788708337563e-05, "loss": 0.0287, "step": 70480 }, { "grad_norm": 0.1515793651342392, "learning_rate": 9.09347224288277e-05, "loss": 0.0198, "step": 70490 }, { "grad_norm": 0.4520549774169922, "learning_rate": 9.093155727688425e-05, "loss": 0.0233, "step": 70500 }, { "grad_norm": 0.30869388580322266, "learning_rate": 9.092839162758372e-05, "loss": 0.024, "step": 70510 }, { "grad_norm": 0.2134932279586792, "learning_rate": 9.09252254809646e-05, "loss": 0.029, "step": 70520 }, { "grad_norm": 0.12170910835266113, "learning_rate": 9.092205883706535e-05, "loss": 0.0203, "step": 70530 }, { "grad_norm": 0.18442466855049133, "learning_rate": 9.091889169592444e-05, "loss": 0.0209, "step": 70540 }, { "grad_norm": 0.3014955520629883, "learning_rate": 9.091572405758038e-05, "loss": 0.0189, "step": 70550 }, { "grad_norm": 0.32859909534454346, "learning_rate": 9.091255592207161e-05, "loss": 0.0214, "step": 70560 }, { "grad_norm": 0.16466498374938965, "learning_rate": 9.090938728943669e-05, "loss": 0.0211, "step": 70570 }, { "grad_norm": 0.2637108862400055, "learning_rate": 9.090621815971408e-05, "loss": 0.0227, "step": 70580 }, { "grad_norm": 0.1673242300748825, "learning_rate": 9.09030485329423e-05, "loss": 0.0243, "step": 70590 }, { "grad_norm": 0.15906989574432373, "learning_rate": 9.089987840915985e-05, "loss": 0.0226, "step": 70600 }, { "grad_norm": 0.2201230227947235, "learning_rate": 9.089670778840527e-05, "loss": 0.0249, "step": 70610 }, { "grad_norm": 0.19487929344177246, "learning_rate": 9.089353667071707e-05, "loss": 0.0221, "step": 70620 }, { "grad_norm": 0.11126261949539185, "learning_rate": 9.089036505613379e-05, "loss": 0.0186, "step": 70630 }, { "grad_norm": 0.14134004712104797, "learning_rate": 9.088719294469397e-05, "loss": 0.0186, "step": 70640 }, { "grad_norm": 0.18044057488441467, "learning_rate": 9.088402033643615e-05, "loss": 0.0315, "step": 70650 }, { "grad_norm": 0.2556300759315491, "learning_rate": 9.088084723139889e-05, "loss": 0.0265, "step": 70660 }, { "grad_norm": 0.19778551161289215, "learning_rate": 9.087767362962074e-05, "loss": 0.029, "step": 70670 }, { "grad_norm": 0.16326557099819183, "learning_rate": 9.087449953114025e-05, "loss": 0.0216, "step": 70680 }, { "grad_norm": 0.18595588207244873, "learning_rate": 9.0871324935996e-05, "loss": 0.018, "step": 70690 }, { "grad_norm": 0.21420297026634216, "learning_rate": 9.086814984422657e-05, "loss": 0.0187, "step": 70700 }, { "grad_norm": 0.21146875619888306, "learning_rate": 9.086497425587053e-05, "loss": 0.0279, "step": 70710 }, { "grad_norm": 0.1314854621887207, "learning_rate": 9.086179817096648e-05, "loss": 0.0168, "step": 70720 }, { "grad_norm": 0.10495135933160782, "learning_rate": 9.085862158955299e-05, "loss": 0.0154, "step": 70730 }, { "grad_norm": 0.14745120704174042, "learning_rate": 9.085544451166867e-05, "loss": 0.0215, "step": 70740 }, { "grad_norm": 0.1987554132938385, "learning_rate": 9.085226693735211e-05, "loss": 0.026, "step": 70750 }, { "grad_norm": 0.18341943621635437, "learning_rate": 9.084908886664196e-05, "loss": 0.0204, "step": 70760 }, { "grad_norm": 0.2698677182197571, "learning_rate": 9.084591029957681e-05, "loss": 0.024, "step": 70770 }, { "grad_norm": 0.13592281937599182, "learning_rate": 9.084273123619529e-05, "loss": 0.0157, "step": 70780 }, { "grad_norm": 0.11062426120042801, "learning_rate": 9.0839551676536e-05, "loss": 0.0272, "step": 70790 }, { "grad_norm": 0.22133119404315948, "learning_rate": 9.083637162063762e-05, "loss": 0.0185, "step": 70800 }, { "grad_norm": 0.14846830070018768, "learning_rate": 9.083319106853876e-05, "loss": 0.0165, "step": 70810 }, { "grad_norm": 0.1666121780872345, "learning_rate": 9.083001002027807e-05, "loss": 0.0233, "step": 70820 }, { "grad_norm": 0.1731671392917633, "learning_rate": 9.082682847589423e-05, "loss": 0.0166, "step": 70830 }, { "grad_norm": 0.16483573615550995, "learning_rate": 9.082364643542585e-05, "loss": 0.0189, "step": 70840 }, { "grad_norm": 0.12551148235797882, "learning_rate": 9.082046389891164e-05, "loss": 0.0194, "step": 70850 }, { "grad_norm": 0.12112446874380112, "learning_rate": 9.081728086639025e-05, "loss": 0.0206, "step": 70860 }, { "grad_norm": 0.15730565786361694, "learning_rate": 9.081409733790036e-05, "loss": 0.0162, "step": 70870 }, { "grad_norm": 0.18187960982322693, "learning_rate": 9.081091331348064e-05, "loss": 0.0205, "step": 70880 }, { "grad_norm": 0.14453434944152832, "learning_rate": 9.080772879316982e-05, "loss": 0.0197, "step": 70890 }, { "grad_norm": 0.20714709162712097, "learning_rate": 9.080454377700653e-05, "loss": 0.0175, "step": 70900 }, { "grad_norm": 0.2096042037010193, "learning_rate": 9.080135826502953e-05, "loss": 0.0256, "step": 70910 }, { "grad_norm": 0.1685941368341446, "learning_rate": 9.079817225727749e-05, "loss": 0.0268, "step": 70920 }, { "grad_norm": 0.14604447782039642, "learning_rate": 9.079498575378915e-05, "loss": 0.0148, "step": 70930 }, { "grad_norm": 0.1507808119058609, "learning_rate": 9.079179875460322e-05, "loss": 0.021, "step": 70940 }, { "grad_norm": 0.14134198427200317, "learning_rate": 9.07886112597584e-05, "loss": 0.0186, "step": 70950 }, { "grad_norm": 0.16983766853809357, "learning_rate": 9.078542326929347e-05, "loss": 0.0244, "step": 70960 }, { "grad_norm": 0.15164323151111603, "learning_rate": 9.078223478324712e-05, "loss": 0.0206, "step": 70970 }, { "grad_norm": 0.1665324568748474, "learning_rate": 9.077904580165811e-05, "loss": 0.0215, "step": 70980 }, { "grad_norm": 0.2929956614971161, "learning_rate": 9.07758563245652e-05, "loss": 0.0199, "step": 70990 }, { "grad_norm": 0.3541264533996582, "learning_rate": 9.077266635200715e-05, "loss": 0.0213, "step": 71000 }, { "grad_norm": 0.35350197553634644, "learning_rate": 9.076947588402268e-05, "loss": 0.0225, "step": 71010 }, { "grad_norm": 0.18802443146705627, "learning_rate": 9.076628492065061e-05, "loss": 0.0277, "step": 71020 }, { "grad_norm": 0.23688316345214844, "learning_rate": 9.076309346192967e-05, "loss": 0.0202, "step": 71030 }, { "grad_norm": 0.22723177075386047, "learning_rate": 9.075990150789866e-05, "loss": 0.0222, "step": 71040 }, { "grad_norm": 0.36206871271133423, "learning_rate": 9.075670905859637e-05, "loss": 0.0316, "step": 71050 }, { "grad_norm": 0.1725560873746872, "learning_rate": 9.07535161140616e-05, "loss": 0.0262, "step": 71060 }, { "grad_norm": 0.6224416494369507, "learning_rate": 9.07503226743331e-05, "loss": 0.0322, "step": 71070 }, { "grad_norm": 0.22290770709514618, "learning_rate": 9.074712873944971e-05, "loss": 0.0234, "step": 71080 }, { "grad_norm": 0.1301247775554657, "learning_rate": 9.074393430945025e-05, "loss": 0.0193, "step": 71090 }, { "grad_norm": 0.11501097679138184, "learning_rate": 9.074073938437351e-05, "loss": 0.0187, "step": 71100 }, { "grad_norm": 0.26428836584091187, "learning_rate": 9.073754396425833e-05, "loss": 0.0243, "step": 71110 }, { "grad_norm": 0.21663455665111542, "learning_rate": 9.073434804914352e-05, "loss": 0.0187, "step": 71120 }, { "grad_norm": 0.15146203339099884, "learning_rate": 9.073115163906793e-05, "loss": 0.0259, "step": 71130 }, { "grad_norm": 0.3045581579208374, "learning_rate": 9.072795473407038e-05, "loss": 0.0201, "step": 71140 }, { "grad_norm": 0.13672389090061188, "learning_rate": 9.072475733418973e-05, "loss": 0.0156, "step": 71150 }, { "grad_norm": 0.20266111195087433, "learning_rate": 9.072155943946483e-05, "loss": 0.0142, "step": 71160 }, { "grad_norm": 0.19760563969612122, "learning_rate": 9.071836104993453e-05, "loss": 0.0197, "step": 71170 }, { "grad_norm": 0.19284069538116455, "learning_rate": 9.07151621656377e-05, "loss": 0.0274, "step": 71180 }, { "grad_norm": 0.1464940458536148, "learning_rate": 9.071196278661322e-05, "loss": 0.0304, "step": 71190 }, { "grad_norm": 0.2037191390991211, "learning_rate": 9.070876291289994e-05, "loss": 0.0178, "step": 71200 }, { "grad_norm": 0.21111376583576202, "learning_rate": 9.070556254453676e-05, "loss": 0.0226, "step": 71210 }, { "grad_norm": 0.18384328484535217, "learning_rate": 9.070236168156256e-05, "loss": 0.021, "step": 71220 }, { "grad_norm": 0.0992087796330452, "learning_rate": 9.069916032401624e-05, "loss": 0.0171, "step": 71230 }, { "grad_norm": 0.1016169860959053, "learning_rate": 9.069595847193669e-05, "loss": 0.0151, "step": 71240 }, { "grad_norm": 0.11457736790180206, "learning_rate": 9.069275612536283e-05, "loss": 0.0221, "step": 71250 }, { "grad_norm": 0.14693957567214966, "learning_rate": 9.068955328433355e-05, "loss": 0.0175, "step": 71260 }, { "grad_norm": 0.15539909899234772, "learning_rate": 9.06863499488878e-05, "loss": 0.0167, "step": 71270 }, { "grad_norm": 0.12135223299264908, "learning_rate": 9.068314611906447e-05, "loss": 0.0194, "step": 71280 }, { "grad_norm": 0.3453020751476288, "learning_rate": 9.067994179490251e-05, "loss": 0.0181, "step": 71290 }, { "grad_norm": 0.09483451396226883, "learning_rate": 9.067673697644084e-05, "loss": 0.0221, "step": 71300 }, { "grad_norm": 0.1759386509656906, "learning_rate": 9.067353166371841e-05, "loss": 0.0186, "step": 71310 }, { "grad_norm": 0.1917649358510971, "learning_rate": 9.067032585677418e-05, "loss": 0.0211, "step": 71320 }, { "grad_norm": 0.15215224027633667, "learning_rate": 9.066711955564708e-05, "loss": 0.0225, "step": 71330 }, { "grad_norm": 0.1583637148141861, "learning_rate": 9.06639127603761e-05, "loss": 0.0242, "step": 71340 }, { "grad_norm": 0.2505873441696167, "learning_rate": 9.066070547100016e-05, "loss": 0.0236, "step": 71350 }, { "grad_norm": 0.13942816853523254, "learning_rate": 9.065749768755826e-05, "loss": 0.0284, "step": 71360 }, { "grad_norm": 0.25028714537620544, "learning_rate": 9.06542894100894e-05, "loss": 0.0259, "step": 71370 }, { "grad_norm": 0.6730759143829346, "learning_rate": 9.065108063863253e-05, "loss": 0.0337, "step": 71380 }, { "grad_norm": 0.18025878071784973, "learning_rate": 9.064787137322664e-05, "loss": 0.0203, "step": 71390 }, { "grad_norm": 0.1994324028491974, "learning_rate": 9.064466161391075e-05, "loss": 0.025, "step": 71400 }, { "grad_norm": 0.18336354196071625, "learning_rate": 9.064145136072384e-05, "loss": 0.0268, "step": 71410 }, { "grad_norm": 0.16034600138664246, "learning_rate": 9.063824061370491e-05, "loss": 0.0209, "step": 71420 }, { "grad_norm": 0.14058174192905426, "learning_rate": 9.063502937289301e-05, "loss": 0.017, "step": 71430 }, { "grad_norm": 0.21352358162403107, "learning_rate": 9.063181763832713e-05, "loss": 0.0191, "step": 71440 }, { "grad_norm": 0.2086855173110962, "learning_rate": 9.06286054100463e-05, "loss": 0.0209, "step": 71450 }, { "grad_norm": 0.1086350753903389, "learning_rate": 9.062539268808955e-05, "loss": 0.019, "step": 71460 }, { "grad_norm": 0.7647484540939331, "learning_rate": 9.062217947249592e-05, "loss": 0.0246, "step": 71470 }, { "grad_norm": 0.5302586555480957, "learning_rate": 9.061896576330447e-05, "loss": 0.0326, "step": 71480 }, { "grad_norm": 0.15179970860481262, "learning_rate": 9.061575156055421e-05, "loss": 0.0281, "step": 71490 }, { "grad_norm": 0.12104017287492752, "learning_rate": 9.061253686428426e-05, "loss": 0.0229, "step": 71500 }, { "grad_norm": 0.10902940481901169, "learning_rate": 9.060932167453362e-05, "loss": 0.0157, "step": 71510 }, { "grad_norm": 0.28360515832901, "learning_rate": 9.060610599134137e-05, "loss": 0.0252, "step": 71520 }, { "grad_norm": 0.15857550501823425, "learning_rate": 9.060288981474661e-05, "loss": 0.0173, "step": 71530 }, { "grad_norm": 0.09604915231466293, "learning_rate": 9.059967314478838e-05, "loss": 0.0234, "step": 71540 }, { "grad_norm": 0.2287217229604721, "learning_rate": 9.059645598150582e-05, "loss": 0.0138, "step": 71550 }, { "grad_norm": 0.17206455767154694, "learning_rate": 9.059323832493798e-05, "loss": 0.0269, "step": 71560 }, { "grad_norm": 0.2943798899650574, "learning_rate": 9.059002017512397e-05, "loss": 0.0214, "step": 71570 }, { "grad_norm": 0.1829109489917755, "learning_rate": 9.058680153210289e-05, "loss": 0.0221, "step": 71580 }, { "grad_norm": 0.15648634731769562, "learning_rate": 9.058358239591384e-05, "loss": 0.0183, "step": 71590 }, { "grad_norm": 0.1742997020483017, "learning_rate": 9.058036276659595e-05, "loss": 0.0213, "step": 71600 }, { "grad_norm": 0.12733514606952667, "learning_rate": 9.057714264418836e-05, "loss": 0.0172, "step": 71610 }, { "grad_norm": 0.16436780989170074, "learning_rate": 9.057392202873014e-05, "loss": 0.0183, "step": 71620 }, { "grad_norm": 0.0804988369345665, "learning_rate": 9.057070092026048e-05, "loss": 0.0169, "step": 71630 }, { "grad_norm": 0.18785427510738373, "learning_rate": 9.056747931881851e-05, "loss": 0.0186, "step": 71640 }, { "grad_norm": 0.1498895287513733, "learning_rate": 9.056425722444335e-05, "loss": 0.0131, "step": 71650 }, { "grad_norm": 0.16940250992774963, "learning_rate": 9.056103463717415e-05, "loss": 0.0173, "step": 71660 }, { "grad_norm": 0.19398123025894165, "learning_rate": 9.055781155705011e-05, "loss": 0.0356, "step": 71670 }, { "grad_norm": 0.14291046559810638, "learning_rate": 9.055458798411035e-05, "loss": 0.0238, "step": 71680 }, { "grad_norm": 0.27226048707962036, "learning_rate": 9.055136391839407e-05, "loss": 0.0231, "step": 71690 }, { "grad_norm": 0.48672056198120117, "learning_rate": 9.054813935994042e-05, "loss": 0.0199, "step": 71700 }, { "grad_norm": 0.3219093680381775, "learning_rate": 9.05449143087886e-05, "loss": 0.0228, "step": 71710 }, { "grad_norm": 0.188123881816864, "learning_rate": 9.054168876497778e-05, "loss": 0.0217, "step": 71720 }, { "grad_norm": 0.2966061532497406, "learning_rate": 9.053846272854718e-05, "loss": 0.0264, "step": 71730 }, { "grad_norm": 0.18122777342796326, "learning_rate": 9.053523619953596e-05, "loss": 0.02, "step": 71740 }, { "grad_norm": 0.19566549360752106, "learning_rate": 9.053200917798337e-05, "loss": 0.0224, "step": 71750 }, { "grad_norm": 0.15726915001869202, "learning_rate": 9.05287816639286e-05, "loss": 0.0284, "step": 71760 }, { "grad_norm": 0.1713976413011551, "learning_rate": 9.052555365741086e-05, "loss": 0.0243, "step": 71770 }, { "grad_norm": 0.14083746075630188, "learning_rate": 9.052232515846936e-05, "loss": 0.0181, "step": 71780 }, { "grad_norm": 0.21805988252162933, "learning_rate": 9.051909616714339e-05, "loss": 0.0215, "step": 71790 }, { "grad_norm": 0.11382462829351425, "learning_rate": 9.051586668347211e-05, "loss": 0.012, "step": 71800 }, { "grad_norm": 0.2700977027416229, "learning_rate": 9.051263670749482e-05, "loss": 0.0254, "step": 71810 }, { "grad_norm": 0.12731310725212097, "learning_rate": 9.050940623925073e-05, "loss": 0.0188, "step": 71820 }, { "grad_norm": 0.23579251766204834, "learning_rate": 9.050617527877911e-05, "loss": 0.021, "step": 71830 }, { "grad_norm": 0.17184795439243317, "learning_rate": 9.050294382611924e-05, "loss": 0.022, "step": 71840 }, { "grad_norm": 0.232150599360466, "learning_rate": 9.049971188131032e-05, "loss": 0.0203, "step": 71850 }, { "grad_norm": 0.5053168535232544, "learning_rate": 9.04964794443917e-05, "loss": 0.0263, "step": 71860 }, { "grad_norm": 0.2123265415430069, "learning_rate": 9.04932465154026e-05, "loss": 0.0286, "step": 71870 }, { "grad_norm": 0.14762535691261292, "learning_rate": 9.049001309438232e-05, "loss": 0.0252, "step": 71880 }, { "grad_norm": 0.22117452323436737, "learning_rate": 9.048677918137017e-05, "loss": 0.0273, "step": 71890 }, { "grad_norm": 0.1861911565065384, "learning_rate": 9.04835447764054e-05, "loss": 0.0263, "step": 71900 }, { "grad_norm": 0.13479487597942352, "learning_rate": 9.048030987952735e-05, "loss": 0.025, "step": 71910 }, { "grad_norm": 0.1638960987329483, "learning_rate": 9.047707449077533e-05, "loss": 0.0165, "step": 71920 }, { "grad_norm": 0.10264160484075546, "learning_rate": 9.047383861018862e-05, "loss": 0.0191, "step": 71930 }, { "grad_norm": 0.1313420534133911, "learning_rate": 9.047060223780657e-05, "loss": 0.0214, "step": 71940 }, { "grad_norm": 0.28527262806892395, "learning_rate": 9.046736537366848e-05, "loss": 0.0183, "step": 71950 }, { "grad_norm": 0.2559489607810974, "learning_rate": 9.04641280178137e-05, "loss": 0.0209, "step": 71960 }, { "grad_norm": 0.1886758804321289, "learning_rate": 9.046089017028156e-05, "loss": 0.0248, "step": 71970 }, { "grad_norm": 0.1164906844496727, "learning_rate": 9.04576518311114e-05, "loss": 0.0163, "step": 71980 }, { "grad_norm": 0.13493230938911438, "learning_rate": 9.045441300034257e-05, "loss": 0.014, "step": 71990 }, { "grad_norm": 0.3575981855392456, "learning_rate": 9.045117367801444e-05, "loss": 0.032, "step": 72000 }, { "grad_norm": 0.20616424083709717, "learning_rate": 9.044793386416635e-05, "loss": 0.0236, "step": 72010 }, { "grad_norm": 0.1473337560892105, "learning_rate": 9.044469355883767e-05, "loss": 0.0219, "step": 72020 }, { "grad_norm": 0.3189336955547333, "learning_rate": 9.044145276206778e-05, "loss": 0.0215, "step": 72030 }, { "grad_norm": 0.24828091263771057, "learning_rate": 9.043821147389606e-05, "loss": 0.0184, "step": 72040 }, { "grad_norm": 0.28233790397644043, "learning_rate": 9.043496969436188e-05, "loss": 0.0335, "step": 72050 }, { "grad_norm": 0.1875705122947693, "learning_rate": 9.043172742350465e-05, "loss": 0.0193, "step": 72060 }, { "grad_norm": 0.11329081654548645, "learning_rate": 9.042848466136376e-05, "loss": 0.0164, "step": 72070 }, { "grad_norm": 0.14497873187065125, "learning_rate": 9.04252414079786e-05, "loss": 0.024, "step": 72080 }, { "grad_norm": 0.12119746208190918, "learning_rate": 9.04219976633886e-05, "loss": 0.0261, "step": 72090 }, { "grad_norm": 0.1573183387517929, "learning_rate": 9.041875342763315e-05, "loss": 0.0206, "step": 72100 }, { "grad_norm": 0.31777653098106384, "learning_rate": 9.041550870075171e-05, "loss": 0.0239, "step": 72110 }, { "grad_norm": 0.15284863114356995, "learning_rate": 9.041226348278365e-05, "loss": 0.0284, "step": 72120 }, { "grad_norm": 0.27266937494277954, "learning_rate": 9.040901777376844e-05, "loss": 0.0203, "step": 72130 }, { "grad_norm": 0.15810534358024597, "learning_rate": 9.040577157374553e-05, "loss": 0.0211, "step": 72140 }, { "grad_norm": 0.17367838323116302, "learning_rate": 9.040252488275433e-05, "loss": 0.0277, "step": 72150 }, { "grad_norm": 0.2329663634300232, "learning_rate": 9.039927770083432e-05, "loss": 0.022, "step": 72160 }, { "grad_norm": 0.23388554155826569, "learning_rate": 9.039603002802493e-05, "loss": 0.0245, "step": 72170 }, { "grad_norm": 0.2153134047985077, "learning_rate": 9.039278186436563e-05, "loss": 0.0213, "step": 72180 }, { "grad_norm": 0.22098056972026825, "learning_rate": 9.038953320989592e-05, "loss": 0.0325, "step": 72190 }, { "grad_norm": 0.14055976271629333, "learning_rate": 9.038628406465522e-05, "loss": 0.0196, "step": 72200 }, { "grad_norm": 0.1639450639486313, "learning_rate": 9.038303442868304e-05, "loss": 0.0254, "step": 72210 }, { "grad_norm": 0.14619943499565125, "learning_rate": 9.037978430201888e-05, "loss": 0.0323, "step": 72220 }, { "grad_norm": 0.15255776047706604, "learning_rate": 9.03765336847022e-05, "loss": 0.0196, "step": 72230 }, { "grad_norm": 0.15412786602973938, "learning_rate": 9.037328257677252e-05, "loss": 0.0183, "step": 72240 }, { "grad_norm": 0.11859593540430069, "learning_rate": 9.037003097826935e-05, "loss": 0.0209, "step": 72250 }, { "grad_norm": 0.15542449057102203, "learning_rate": 9.036677888923217e-05, "loss": 0.0207, "step": 72260 }, { "grad_norm": 0.08419610559940338, "learning_rate": 9.036352630970052e-05, "loss": 0.0166, "step": 72270 }, { "grad_norm": 0.2763098478317261, "learning_rate": 9.036027323971392e-05, "loss": 0.0211, "step": 72280 }, { "grad_norm": 0.1092124730348587, "learning_rate": 9.035701967931188e-05, "loss": 0.0176, "step": 72290 }, { "grad_norm": 0.11104096472263336, "learning_rate": 9.035376562853397e-05, "loss": 0.0181, "step": 72300 }, { "grad_norm": 0.16512873768806458, "learning_rate": 9.035051108741969e-05, "loss": 0.0178, "step": 72310 }, { "grad_norm": 0.14778105914592743, "learning_rate": 9.03472560560086e-05, "loss": 0.0206, "step": 72320 }, { "grad_norm": 0.2713487148284912, "learning_rate": 9.034400053434027e-05, "loss": 0.0185, "step": 72330 }, { "grad_norm": 0.1494089961051941, "learning_rate": 9.034074452245423e-05, "loss": 0.0169, "step": 72340 }, { "grad_norm": 0.12051840126514435, "learning_rate": 9.033748802039006e-05, "loss": 0.0189, "step": 72350 }, { "grad_norm": 0.23215420544147491, "learning_rate": 9.033423102818734e-05, "loss": 0.021, "step": 72360 }, { "grad_norm": 0.1967337727546692, "learning_rate": 9.033097354588561e-05, "loss": 0.0247, "step": 72370 }, { "grad_norm": 0.168965145945549, "learning_rate": 9.032771557352448e-05, "loss": 0.0179, "step": 72380 }, { "grad_norm": 0.16530314087867737, "learning_rate": 9.032445711114353e-05, "loss": 0.0169, "step": 72390 }, { "grad_norm": 0.19048264622688293, "learning_rate": 9.032119815878236e-05, "loss": 0.0182, "step": 72400 }, { "grad_norm": 0.22453007102012634, "learning_rate": 9.031793871648056e-05, "loss": 0.0205, "step": 72410 }, { "grad_norm": 0.33184781670570374, "learning_rate": 9.031467878427773e-05, "loss": 0.0192, "step": 72420 }, { "grad_norm": 0.3056976795196533, "learning_rate": 9.03114183622135e-05, "loss": 0.0254, "step": 72430 }, { "grad_norm": 0.17552055418491364, "learning_rate": 9.030815745032746e-05, "loss": 0.0191, "step": 72440 }, { "grad_norm": 0.1399172693490982, "learning_rate": 9.030489604865927e-05, "loss": 0.0171, "step": 72450 }, { "grad_norm": 0.15956124663352966, "learning_rate": 9.030163415724853e-05, "loss": 0.0203, "step": 72460 }, { "grad_norm": 0.16624954342842102, "learning_rate": 9.02983717761349e-05, "loss": 0.0225, "step": 72470 }, { "grad_norm": 0.22912293672561646, "learning_rate": 9.029510890535797e-05, "loss": 0.0172, "step": 72480 }, { "grad_norm": 0.11105113476514816, "learning_rate": 9.029184554495745e-05, "loss": 0.0181, "step": 72490 }, { "grad_norm": 0.13126935064792633, "learning_rate": 9.028858169497295e-05, "loss": 0.0199, "step": 72500 }, { "grad_norm": 0.1929067224264145, "learning_rate": 9.028531735544415e-05, "loss": 0.0225, "step": 72510 }, { "grad_norm": 0.13165093958377838, "learning_rate": 9.028205252641071e-05, "loss": 0.019, "step": 72520 }, { "grad_norm": 0.13709188997745514, "learning_rate": 9.027878720791229e-05, "loss": 0.0254, "step": 72530 }, { "grad_norm": 0.1638507843017578, "learning_rate": 9.02755213999886e-05, "loss": 0.0207, "step": 72540 }, { "grad_norm": 0.10159008204936981, "learning_rate": 9.027225510267928e-05, "loss": 0.0197, "step": 72550 }, { "grad_norm": 0.1745610535144806, "learning_rate": 9.026898831602404e-05, "loss": 0.0254, "step": 72560 }, { "grad_norm": 0.18338222801685333, "learning_rate": 9.026572104006257e-05, "loss": 0.0178, "step": 72570 }, { "grad_norm": 0.20291075110435486, "learning_rate": 9.026245327483456e-05, "loss": 0.024, "step": 72580 }, { "grad_norm": 0.13618822395801544, "learning_rate": 9.025918502037975e-05, "loss": 0.0228, "step": 72590 }, { "grad_norm": 0.2477627843618393, "learning_rate": 9.025591627673782e-05, "loss": 0.0219, "step": 72600 }, { "grad_norm": 0.13115814328193665, "learning_rate": 9.025264704394851e-05, "loss": 0.0207, "step": 72610 }, { "grad_norm": 0.2507253587245941, "learning_rate": 9.024937732205152e-05, "loss": 0.0166, "step": 72620 }, { "grad_norm": 0.25275716185569763, "learning_rate": 9.024610711108659e-05, "loss": 0.0252, "step": 72630 }, { "grad_norm": 0.1804663985967636, "learning_rate": 9.024283641109348e-05, "loss": 0.0188, "step": 72640 }, { "grad_norm": 0.32539257407188416, "learning_rate": 9.02395652221119e-05, "loss": 0.0239, "step": 72650 }, { "grad_norm": 0.2192137986421585, "learning_rate": 9.023629354418162e-05, "loss": 0.0201, "step": 72660 }, { "grad_norm": 0.1494574099779129, "learning_rate": 9.023302137734238e-05, "loss": 0.0185, "step": 72670 }, { "grad_norm": 0.1621675342321396, "learning_rate": 9.022974872163392e-05, "loss": 0.0189, "step": 72680 }, { "grad_norm": 0.14949347078800201, "learning_rate": 9.022647557709604e-05, "loss": 0.0182, "step": 72690 }, { "grad_norm": 0.21921919286251068, "learning_rate": 9.022320194376853e-05, "loss": 0.016, "step": 72700 }, { "grad_norm": 0.1726532131433487, "learning_rate": 9.021992782169112e-05, "loss": 0.0231, "step": 72710 }, { "grad_norm": 0.19702598452568054, "learning_rate": 9.021665321090361e-05, "loss": 0.0189, "step": 72720 }, { "grad_norm": 0.1770191788673401, "learning_rate": 9.02133781114458e-05, "loss": 0.0169, "step": 72730 }, { "grad_norm": 0.14676684141159058, "learning_rate": 9.021010252335747e-05, "loss": 0.0326, "step": 72740 }, { "grad_norm": 0.1571701020002365, "learning_rate": 9.020682644667844e-05, "loss": 0.0195, "step": 72750 }, { "grad_norm": 0.3086611330509186, "learning_rate": 9.02035498814485e-05, "loss": 0.0245, "step": 72760 }, { "grad_norm": 0.24365660548210144, "learning_rate": 9.020027282770747e-05, "loss": 0.0216, "step": 72770 }, { "grad_norm": 0.09736054390668869, "learning_rate": 9.019699528549518e-05, "loss": 0.014, "step": 72780 }, { "grad_norm": 0.08948101103305817, "learning_rate": 9.019371725485144e-05, "loss": 0.0175, "step": 72790 }, { "grad_norm": 0.20444171130657196, "learning_rate": 9.019043873581607e-05, "loss": 0.0212, "step": 72800 }, { "grad_norm": 0.2199610471725464, "learning_rate": 9.018715972842895e-05, "loss": 0.02, "step": 72810 }, { "grad_norm": 0.24626705050468445, "learning_rate": 9.018388023272987e-05, "loss": 0.0174, "step": 72820 }, { "grad_norm": 0.1056843176484108, "learning_rate": 9.018060024875872e-05, "loss": 0.0196, "step": 72830 }, { "grad_norm": 0.26235249638557434, "learning_rate": 9.017731977655535e-05, "loss": 0.0145, "step": 72840 }, { "grad_norm": 0.09824561327695847, "learning_rate": 9.017403881615959e-05, "loss": 0.0177, "step": 72850 }, { "grad_norm": 0.13291075825691223, "learning_rate": 9.017075736761133e-05, "loss": 0.0195, "step": 72860 }, { "grad_norm": 0.15658415853977203, "learning_rate": 9.016747543095046e-05, "loss": 0.0178, "step": 72870 }, { "grad_norm": 0.2789921760559082, "learning_rate": 9.016419300621683e-05, "loss": 0.0205, "step": 72880 }, { "grad_norm": 0.23999793827533722, "learning_rate": 9.016091009345033e-05, "loss": 0.0185, "step": 72890 }, { "grad_norm": 0.22204646468162537, "learning_rate": 9.015762669269085e-05, "loss": 0.0174, "step": 72900 }, { "grad_norm": 0.2190602868795395, "learning_rate": 9.01543428039783e-05, "loss": 0.0228, "step": 72910 }, { "grad_norm": 0.14175759255886078, "learning_rate": 9.015105842735257e-05, "loss": 0.0269, "step": 72920 }, { "grad_norm": 0.3048720359802246, "learning_rate": 9.014777356285356e-05, "loss": 0.0225, "step": 72930 }, { "grad_norm": 0.7716957330703735, "learning_rate": 9.01444882105212e-05, "loss": 0.0211, "step": 72940 }, { "grad_norm": 0.23034319281578064, "learning_rate": 9.01412023703954e-05, "loss": 0.0222, "step": 72950 }, { "grad_norm": 0.2082025110721588, "learning_rate": 9.013791604251611e-05, "loss": 0.0204, "step": 72960 }, { "grad_norm": 0.17998412251472473, "learning_rate": 9.013462922692324e-05, "loss": 0.0154, "step": 72970 }, { "grad_norm": 0.27069607377052307, "learning_rate": 9.013134192365673e-05, "loss": 0.019, "step": 72980 }, { "grad_norm": 0.1268302947282791, "learning_rate": 9.01280541327565e-05, "loss": 0.0211, "step": 72990 }, { "grad_norm": 0.14786218106746674, "learning_rate": 9.012476585426257e-05, "loss": 0.0166, "step": 73000 }, { "grad_norm": 0.37538477778434753, "learning_rate": 9.012147708821483e-05, "loss": 0.0187, "step": 73010 }, { "grad_norm": 0.2563977539539337, "learning_rate": 9.011818783465326e-05, "loss": 0.0232, "step": 73020 }, { "grad_norm": 0.3935903012752533, "learning_rate": 9.011489809361782e-05, "loss": 0.0305, "step": 73030 }, { "grad_norm": 0.16307558119297028, "learning_rate": 9.011160786514852e-05, "loss": 0.0209, "step": 73040 }, { "grad_norm": 0.15785133838653564, "learning_rate": 9.01083171492853e-05, "loss": 0.0204, "step": 73050 }, { "grad_norm": 0.10411522537469864, "learning_rate": 9.010502594606815e-05, "loss": 0.0202, "step": 73060 }, { "grad_norm": 0.32157132029533386, "learning_rate": 9.010173425553708e-05, "loss": 0.0179, "step": 73070 }, { "grad_norm": 0.1650697886943817, "learning_rate": 9.009844207773206e-05, "loss": 0.0184, "step": 73080 }, { "grad_norm": 0.14385321736335754, "learning_rate": 9.009514941269313e-05, "loss": 0.0204, "step": 73090 }, { "grad_norm": 0.13424919545650482, "learning_rate": 9.009185626046026e-05, "loss": 0.0197, "step": 73100 }, { "grad_norm": 0.25808200240135193, "learning_rate": 9.008856262107349e-05, "loss": 0.022, "step": 73110 }, { "grad_norm": 0.10565876960754395, "learning_rate": 9.008526849457284e-05, "loss": 0.0183, "step": 73120 }, { "grad_norm": 0.33611103892326355, "learning_rate": 9.008197388099833e-05, "loss": 0.0192, "step": 73130 }, { "grad_norm": 0.1461099535226822, "learning_rate": 9.007867878038998e-05, "loss": 0.027, "step": 73140 }, { "grad_norm": 0.32433295249938965, "learning_rate": 9.007538319278785e-05, "loss": 0.0232, "step": 73150 }, { "grad_norm": 0.12326456606388092, "learning_rate": 9.007208711823198e-05, "loss": 0.023, "step": 73160 }, { "grad_norm": 0.12787719070911407, "learning_rate": 9.006879055676243e-05, "loss": 0.0228, "step": 73170 }, { "grad_norm": 0.49072718620300293, "learning_rate": 9.006549350841923e-05, "loss": 0.0211, "step": 73180 }, { "grad_norm": 0.19560150802135468, "learning_rate": 9.006219597324245e-05, "loss": 0.0236, "step": 73190 }, { "grad_norm": 0.14383302628993988, "learning_rate": 9.005889795127219e-05, "loss": 0.0219, "step": 73200 }, { "grad_norm": 0.5957331657409668, "learning_rate": 9.005559944254846e-05, "loss": 0.0196, "step": 73210 }, { "grad_norm": 0.1592661440372467, "learning_rate": 9.00523004471114e-05, "loss": 0.0174, "step": 73220 }, { "grad_norm": 0.15189075469970703, "learning_rate": 9.004900096500107e-05, "loss": 0.0177, "step": 73230 }, { "grad_norm": 0.18400362133979797, "learning_rate": 9.004570099625758e-05, "loss": 0.0287, "step": 73240 }, { "grad_norm": 0.24812036752700806, "learning_rate": 9.004240054092099e-05, "loss": 0.0243, "step": 73250 }, { "grad_norm": 0.09156054258346558, "learning_rate": 9.003909959903144e-05, "loss": 0.0141, "step": 73260 }, { "grad_norm": 0.15166033804416656, "learning_rate": 9.003579817062901e-05, "loss": 0.0179, "step": 73270 }, { "grad_norm": 0.12120076268911362, "learning_rate": 9.003249625575385e-05, "loss": 0.0194, "step": 73280 }, { "grad_norm": 0.14212141931056976, "learning_rate": 9.002919385444607e-05, "loss": 0.0225, "step": 73290 }, { "grad_norm": 0.13309481739997864, "learning_rate": 9.002589096674578e-05, "loss": 0.0174, "step": 73300 }, { "grad_norm": 0.15952162444591522, "learning_rate": 9.002258759269312e-05, "loss": 0.0137, "step": 73310 }, { "grad_norm": 0.11756313592195511, "learning_rate": 9.001928373232825e-05, "loss": 0.0254, "step": 73320 }, { "grad_norm": 0.43339335918426514, "learning_rate": 9.001597938569128e-05, "loss": 0.0217, "step": 73330 }, { "grad_norm": 0.11257100850343704, "learning_rate": 9.001267455282239e-05, "loss": 0.017, "step": 73340 }, { "grad_norm": 0.1366475224494934, "learning_rate": 9.000936923376171e-05, "loss": 0.0194, "step": 73350 }, { "grad_norm": 0.1540013700723648, "learning_rate": 9.000606342854942e-05, "loss": 0.0188, "step": 73360 }, { "grad_norm": 0.13235358893871307, "learning_rate": 9.00027571372257e-05, "loss": 0.0136, "step": 73370 }, { "grad_norm": 0.1541438102722168, "learning_rate": 8.999945035983072e-05, "loss": 0.0231, "step": 73380 }, { "grad_norm": 0.1408441811800003, "learning_rate": 8.999614309640464e-05, "loss": 0.0165, "step": 73390 }, { "grad_norm": 0.0995313748717308, "learning_rate": 8.999283534698766e-05, "loss": 0.014, "step": 73400 }, { "grad_norm": 0.10506021231412888, "learning_rate": 8.998952711161996e-05, "loss": 0.0222, "step": 73410 }, { "grad_norm": 0.16267426311969757, "learning_rate": 8.998621839034177e-05, "loss": 0.018, "step": 73420 }, { "grad_norm": 0.17194023728370667, "learning_rate": 8.998290918319327e-05, "loss": 0.0206, "step": 73430 }, { "grad_norm": 0.18218234181404114, "learning_rate": 8.997959949021466e-05, "loss": 0.022, "step": 73440 }, { "grad_norm": 0.18543586134910583, "learning_rate": 8.997628931144619e-05, "loss": 0.014, "step": 73450 }, { "grad_norm": 0.5335498452186584, "learning_rate": 8.997297864692803e-05, "loss": 0.0179, "step": 73460 }, { "grad_norm": 0.3273671865463257, "learning_rate": 8.996966749670046e-05, "loss": 0.018, "step": 73470 }, { "grad_norm": 0.4084474444389343, "learning_rate": 8.996635586080369e-05, "loss": 0.0239, "step": 73480 }, { "grad_norm": 0.16270872950553894, "learning_rate": 8.996304373927795e-05, "loss": 0.0221, "step": 73490 }, { "grad_norm": 0.1743171215057373, "learning_rate": 8.995973113216352e-05, "loss": 0.0226, "step": 73500 }, { "grad_norm": 0.11441297084093094, "learning_rate": 8.995641803950062e-05, "loss": 0.0218, "step": 73510 }, { "grad_norm": 0.34897279739379883, "learning_rate": 8.99531044613295e-05, "loss": 0.0211, "step": 73520 }, { "grad_norm": 0.16778691112995148, "learning_rate": 8.994979039769046e-05, "loss": 0.0188, "step": 73530 }, { "grad_norm": 0.12329418957233429, "learning_rate": 8.994647584862374e-05, "loss": 0.0178, "step": 73540 }, { "grad_norm": 0.1427796334028244, "learning_rate": 8.994316081416961e-05, "loss": 0.0187, "step": 73550 }, { "grad_norm": 0.13326090574264526, "learning_rate": 8.993984529436839e-05, "loss": 0.019, "step": 73560 }, { "grad_norm": 0.23530544340610504, "learning_rate": 8.99365292892603e-05, "loss": 0.0262, "step": 73570 }, { "grad_norm": 0.12355607748031616, "learning_rate": 8.993321279888571e-05, "loss": 0.0178, "step": 73580 }, { "grad_norm": 0.08610320091247559, "learning_rate": 8.992989582328485e-05, "loss": 0.0122, "step": 73590 }, { "grad_norm": 0.22152729332447052, "learning_rate": 8.992657836249809e-05, "loss": 0.0183, "step": 73600 }, { "grad_norm": 0.11542396247386932, "learning_rate": 8.992326041656568e-05, "loss": 0.0284, "step": 73610 }, { "grad_norm": 0.08927411586046219, "learning_rate": 8.991994198552798e-05, "loss": 0.0162, "step": 73620 }, { "grad_norm": 0.3498496115207672, "learning_rate": 8.991662306942529e-05, "loss": 0.0256, "step": 73630 }, { "grad_norm": 0.11503594368696213, "learning_rate": 8.991330366829794e-05, "loss": 0.0152, "step": 73640 }, { "grad_norm": 0.13062356412410736, "learning_rate": 8.990998378218627e-05, "loss": 0.023, "step": 73650 }, { "grad_norm": 0.14406824111938477, "learning_rate": 8.990666341113061e-05, "loss": 0.0178, "step": 73660 }, { "grad_norm": 0.13315026462078094, "learning_rate": 8.99033425551713e-05, "loss": 0.0207, "step": 73670 }, { "grad_norm": 0.32624804973602295, "learning_rate": 8.990002121434874e-05, "loss": 0.0237, "step": 73680 }, { "grad_norm": 0.22857670485973358, "learning_rate": 8.989669938870321e-05, "loss": 0.0207, "step": 73690 }, { "grad_norm": 0.13236349821090698, "learning_rate": 8.989337707827513e-05, "loss": 0.0204, "step": 73700 }, { "grad_norm": 0.19657665491104126, "learning_rate": 8.989005428310483e-05, "loss": 0.0304, "step": 73710 }, { "grad_norm": 0.18780039250850677, "learning_rate": 8.988673100323274e-05, "loss": 0.017, "step": 73720 }, { "grad_norm": 0.10431686043739319, "learning_rate": 8.988340723869922e-05, "loss": 0.0176, "step": 73730 }, { "grad_norm": 0.2070523500442505, "learning_rate": 8.988008298954461e-05, "loss": 0.0148, "step": 73740 }, { "grad_norm": 0.12412182986736298, "learning_rate": 8.987675825580936e-05, "loss": 0.0172, "step": 73750 }, { "grad_norm": 0.2731969952583313, "learning_rate": 8.987343303753384e-05, "loss": 0.0189, "step": 73760 }, { "grad_norm": 0.20346741378307343, "learning_rate": 8.987010733475846e-05, "loss": 0.0206, "step": 73770 }, { "grad_norm": 0.1116880476474762, "learning_rate": 8.986678114752363e-05, "loss": 0.0196, "step": 73780 }, { "grad_norm": 0.14035116136074066, "learning_rate": 8.986345447586978e-05, "loss": 0.019, "step": 73790 }, { "grad_norm": 0.23005661368370056, "learning_rate": 8.98601273198373e-05, "loss": 0.0152, "step": 73800 }, { "grad_norm": 0.23180603981018066, "learning_rate": 8.985679967946666e-05, "loss": 0.0162, "step": 73810 }, { "grad_norm": 0.14472249150276184, "learning_rate": 8.985347155479827e-05, "loss": 0.0178, "step": 73820 }, { "grad_norm": 0.12294841557741165, "learning_rate": 8.985014294587258e-05, "loss": 0.0195, "step": 73830 }, { "grad_norm": 0.18051911890506744, "learning_rate": 8.984681385273001e-05, "loss": 0.0209, "step": 73840 }, { "grad_norm": 0.09758193790912628, "learning_rate": 8.984348427541103e-05, "loss": 0.0211, "step": 73850 }, { "grad_norm": 0.12535583972930908, "learning_rate": 8.98401542139561e-05, "loss": 0.0148, "step": 73860 }, { "grad_norm": 0.1243865117430687, "learning_rate": 8.98368236684057e-05, "loss": 0.0166, "step": 73870 }, { "grad_norm": 0.1011604443192482, "learning_rate": 8.983349263880026e-05, "loss": 0.0174, "step": 73880 }, { "grad_norm": 0.1659475415945053, "learning_rate": 8.983016112518028e-05, "loss": 0.0171, "step": 73890 }, { "grad_norm": 0.14127172529697418, "learning_rate": 8.982682912758624e-05, "loss": 0.0202, "step": 73900 }, { "grad_norm": 0.12770891189575195, "learning_rate": 8.982349664605863e-05, "loss": 0.0171, "step": 73910 }, { "grad_norm": 0.14764313399791718, "learning_rate": 8.982016368063792e-05, "loss": 0.0186, "step": 73920 }, { "grad_norm": 0.11348937451839447, "learning_rate": 8.981683023136462e-05, "loss": 0.0186, "step": 73930 }, { "grad_norm": 0.7608084082603455, "learning_rate": 8.981349629827926e-05, "loss": 0.0214, "step": 73940 }, { "grad_norm": 0.1600913256406784, "learning_rate": 8.981016188142233e-05, "loss": 0.0227, "step": 73950 }, { "grad_norm": 0.1918720155954361, "learning_rate": 8.980682698083433e-05, "loss": 0.0164, "step": 73960 }, { "grad_norm": 0.2608889043331146, "learning_rate": 8.980349159655581e-05, "loss": 0.018, "step": 73970 }, { "grad_norm": 0.34331417083740234, "learning_rate": 8.980015572862727e-05, "loss": 0.0211, "step": 73980 }, { "grad_norm": 0.14802061021327972, "learning_rate": 8.979681937708928e-05, "loss": 0.0293, "step": 73990 }, { "grad_norm": 0.31447634100914, "learning_rate": 8.979348254198235e-05, "loss": 0.0186, "step": 74000 }, { "grad_norm": 0.1460695117712021, "learning_rate": 8.979014522334704e-05, "loss": 0.0269, "step": 74010 }, { "grad_norm": 0.13899417221546173, "learning_rate": 8.97868074212239e-05, "loss": 0.025, "step": 74020 }, { "grad_norm": 0.16715222597122192, "learning_rate": 8.978346913565348e-05, "loss": 0.0186, "step": 74030 }, { "grad_norm": 0.15203724801540375, "learning_rate": 8.978013036667634e-05, "loss": 0.0345, "step": 74040 }, { "grad_norm": 0.22417926788330078, "learning_rate": 8.977679111433308e-05, "loss": 0.0229, "step": 74050 }, { "grad_norm": 0.16147685050964355, "learning_rate": 8.977345137866422e-05, "loss": 0.0291, "step": 74060 }, { "grad_norm": 0.22557786107063293, "learning_rate": 8.97701111597104e-05, "loss": 0.0239, "step": 74070 }, { "grad_norm": 0.09899292886257172, "learning_rate": 8.976677045751217e-05, "loss": 0.0215, "step": 74080 }, { "grad_norm": 0.4023134708404541, "learning_rate": 8.976342927211014e-05, "loss": 0.0266, "step": 74090 }, { "grad_norm": 0.24910427629947662, "learning_rate": 8.976008760354488e-05, "loss": 0.0216, "step": 74100 }, { "grad_norm": 0.21447445452213287, "learning_rate": 8.975674545185703e-05, "loss": 0.0189, "step": 74110 }, { "grad_norm": 0.12933993339538574, "learning_rate": 8.97534028170872e-05, "loss": 0.0188, "step": 74120 }, { "grad_norm": 0.23193801939487457, "learning_rate": 8.975005969927599e-05, "loss": 0.0276, "step": 74130 }, { "grad_norm": 0.17575308680534363, "learning_rate": 8.974671609846402e-05, "loss": 0.0286, "step": 74140 }, { "grad_norm": 0.14858537912368774, "learning_rate": 8.97433720146919e-05, "loss": 0.0214, "step": 74150 }, { "grad_norm": 0.16544461250305176, "learning_rate": 8.974002744800033e-05, "loss": 0.0218, "step": 74160 }, { "grad_norm": 0.12826091051101685, "learning_rate": 8.973668239842986e-05, "loss": 0.0201, "step": 74170 }, { "grad_norm": 0.13641737401485443, "learning_rate": 8.97333368660212e-05, "loss": 0.0224, "step": 74180 }, { "grad_norm": 0.16138412058353424, "learning_rate": 8.972999085081498e-05, "loss": 0.0205, "step": 74190 }, { "grad_norm": 0.1213146522641182, "learning_rate": 8.972664435285188e-05, "loss": 0.0224, "step": 74200 }, { "grad_norm": 0.17287173867225647, "learning_rate": 8.972329737217251e-05, "loss": 0.0264, "step": 74210 }, { "grad_norm": 0.1436336785554886, "learning_rate": 8.971994990881759e-05, "loss": 0.0172, "step": 74220 }, { "grad_norm": 0.16506631672382355, "learning_rate": 8.971660196282779e-05, "loss": 0.0225, "step": 74230 }, { "grad_norm": 0.11703915148973465, "learning_rate": 8.971325353424374e-05, "loss": 0.0173, "step": 74240 }, { "grad_norm": 0.14809006452560425, "learning_rate": 8.97099046231062e-05, "loss": 0.0165, "step": 74250 }, { "grad_norm": 0.29614076018333435, "learning_rate": 8.970655522945581e-05, "loss": 0.0227, "step": 74260 }, { "grad_norm": 0.19815140962600708, "learning_rate": 8.970320535333327e-05, "loss": 0.0197, "step": 74270 }, { "grad_norm": 0.17012162506580353, "learning_rate": 8.969985499477933e-05, "loss": 0.0274, "step": 74280 }, { "grad_norm": 0.1889636367559433, "learning_rate": 8.969650415383466e-05, "loss": 0.0214, "step": 74290 }, { "grad_norm": 0.12939120829105377, "learning_rate": 8.969315283053997e-05, "loss": 0.0207, "step": 74300 }, { "grad_norm": 0.2371685653924942, "learning_rate": 8.9689801024936e-05, "loss": 0.0247, "step": 74310 }, { "grad_norm": 0.21248534321784973, "learning_rate": 8.96864487370635e-05, "loss": 0.0181, "step": 74320 }, { "grad_norm": 0.1425555944442749, "learning_rate": 8.968309596696315e-05, "loss": 0.0161, "step": 74330 }, { "grad_norm": 0.16326729953289032, "learning_rate": 8.967974271467572e-05, "loss": 0.0216, "step": 74340 }, { "grad_norm": 0.16415323317050934, "learning_rate": 8.967638898024195e-05, "loss": 0.0224, "step": 74350 }, { "grad_norm": 0.08796561509370804, "learning_rate": 8.967303476370261e-05, "loss": 0.0194, "step": 74360 }, { "grad_norm": 0.19767539203166962, "learning_rate": 8.966968006509844e-05, "loss": 0.0242, "step": 74370 }, { "grad_norm": 0.12053772807121277, "learning_rate": 8.966632488447018e-05, "loss": 0.0189, "step": 74380 }, { "grad_norm": 0.1213236153125763, "learning_rate": 8.966296922185864e-05, "loss": 0.0208, "step": 74390 }, { "grad_norm": 0.15652404725551605, "learning_rate": 8.965961307730457e-05, "loss": 0.0211, "step": 74400 }, { "grad_norm": 0.1349484771490097, "learning_rate": 8.965625645084877e-05, "loss": 0.0176, "step": 74410 }, { "grad_norm": 0.29658353328704834, "learning_rate": 8.965289934253198e-05, "loss": 0.0215, "step": 74420 }, { "grad_norm": 0.21929717063903809, "learning_rate": 8.964954175239506e-05, "loss": 0.0224, "step": 74430 }, { "grad_norm": 0.2038898915052414, "learning_rate": 8.964618368047875e-05, "loss": 0.023, "step": 74440 }, { "grad_norm": 0.1442195326089859, "learning_rate": 8.964282512682387e-05, "loss": 0.0182, "step": 74450 }, { "grad_norm": 0.09541638940572739, "learning_rate": 8.963946609147126e-05, "loss": 0.0155, "step": 74460 }, { "grad_norm": 0.10310603678226471, "learning_rate": 8.963610657446169e-05, "loss": 0.0182, "step": 74470 }, { "grad_norm": 0.3470711410045624, "learning_rate": 8.963274657583602e-05, "loss": 0.0208, "step": 74480 }, { "grad_norm": 0.12744839489459991, "learning_rate": 8.962938609563504e-05, "loss": 0.019, "step": 74490 }, { "grad_norm": 0.2131136804819107, "learning_rate": 8.962602513389961e-05, "loss": 0.0198, "step": 74500 }, { "grad_norm": 0.17688393592834473, "learning_rate": 8.962266369067058e-05, "loss": 0.0199, "step": 74510 }, { "grad_norm": 0.265405535697937, "learning_rate": 8.961930176598877e-05, "loss": 0.016, "step": 74520 }, { "grad_norm": 0.27155953645706177, "learning_rate": 8.961593935989502e-05, "loss": 0.0194, "step": 74530 }, { "grad_norm": 0.10139691829681396, "learning_rate": 8.961257647243021e-05, "loss": 0.0186, "step": 74540 }, { "grad_norm": 0.23326846957206726, "learning_rate": 8.960921310363519e-05, "loss": 0.0199, "step": 74550 }, { "grad_norm": 0.09523646533489227, "learning_rate": 8.960584925355085e-05, "loss": 0.0193, "step": 74560 }, { "grad_norm": 0.12855452299118042, "learning_rate": 8.960248492221804e-05, "loss": 0.0208, "step": 74570 }, { "grad_norm": 0.24821066856384277, "learning_rate": 8.959912010967764e-05, "loss": 0.0243, "step": 74580 }, { "grad_norm": 0.12003308534622192, "learning_rate": 8.959575481597055e-05, "loss": 0.0188, "step": 74590 }, { "grad_norm": 0.11416016519069672, "learning_rate": 8.959238904113765e-05, "loss": 0.0161, "step": 74600 }, { "grad_norm": 0.20462048053741455, "learning_rate": 8.958902278521984e-05, "loss": 0.0164, "step": 74610 }, { "grad_norm": 0.1944141834974289, "learning_rate": 8.958565604825803e-05, "loss": 0.0172, "step": 74620 }, { "grad_norm": 0.19442899525165558, "learning_rate": 8.958228883029312e-05, "loss": 0.0182, "step": 74630 }, { "grad_norm": 0.17050951719284058, "learning_rate": 8.957892113136604e-05, "loss": 0.0204, "step": 74640 }, { "grad_norm": 0.14598172903060913, "learning_rate": 8.95755529515177e-05, "loss": 0.0193, "step": 74650 }, { "grad_norm": 0.20789791643619537, "learning_rate": 8.9572184290789e-05, "loss": 0.0188, "step": 74660 }, { "grad_norm": 0.212441086769104, "learning_rate": 8.956881514922092e-05, "loss": 0.0165, "step": 74670 }, { "grad_norm": 0.14648789167404175, "learning_rate": 8.956544552685437e-05, "loss": 0.0301, "step": 74680 }, { "grad_norm": 0.17863532900810242, "learning_rate": 8.95620754237303e-05, "loss": 0.0209, "step": 74690 }, { "grad_norm": 0.1575019508600235, "learning_rate": 8.955870483988966e-05, "loss": 0.0236, "step": 74700 }, { "grad_norm": 0.17777711153030396, "learning_rate": 8.955533377537342e-05, "loss": 0.0164, "step": 74710 }, { "grad_norm": 0.3445785343647003, "learning_rate": 8.955196223022251e-05, "loss": 0.0203, "step": 74720 }, { "grad_norm": 0.17677436769008636, "learning_rate": 8.954859020447793e-05, "loss": 0.0225, "step": 74730 }, { "grad_norm": 0.11552947759628296, "learning_rate": 8.954521769818063e-05, "loss": 0.0196, "step": 74740 }, { "grad_norm": 0.1918059140443802, "learning_rate": 8.954184471137161e-05, "loss": 0.0227, "step": 74750 }, { "grad_norm": 0.21780437231063843, "learning_rate": 8.953847124409183e-05, "loss": 0.0212, "step": 74760 }, { "grad_norm": 0.1503055840730667, "learning_rate": 8.95350972963823e-05, "loss": 0.021, "step": 74770 }, { "grad_norm": 0.1515863984823227, "learning_rate": 8.953172286828401e-05, "loss": 0.0161, "step": 74780 }, { "grad_norm": 0.12524566054344177, "learning_rate": 8.952834795983796e-05, "loss": 0.025, "step": 74790 }, { "grad_norm": 0.12619167566299438, "learning_rate": 8.952497257108517e-05, "loss": 0.0213, "step": 74800 }, { "grad_norm": 0.14199867844581604, "learning_rate": 8.952159670206664e-05, "loss": 0.0255, "step": 74810 }, { "grad_norm": 0.24205195903778076, "learning_rate": 8.95182203528234e-05, "loss": 0.021, "step": 74820 }, { "grad_norm": 0.15254397690296173, "learning_rate": 8.951484352339645e-05, "loss": 0.0183, "step": 74830 }, { "grad_norm": 0.15867246687412262, "learning_rate": 8.951146621382687e-05, "loss": 0.0199, "step": 74840 }, { "grad_norm": 0.2037430703639984, "learning_rate": 8.950808842415565e-05, "loss": 0.0165, "step": 74850 }, { "grad_norm": 0.12442663311958313, "learning_rate": 8.950471015442386e-05, "loss": 0.019, "step": 74860 }, { "grad_norm": 0.25646573305130005, "learning_rate": 8.950133140467255e-05, "loss": 0.019, "step": 74870 }, { "grad_norm": 0.09775739908218384, "learning_rate": 8.949795217494276e-05, "loss": 0.0136, "step": 74880 }, { "grad_norm": 0.16820064187049866, "learning_rate": 8.949457246527558e-05, "loss": 0.0191, "step": 74890 }, { "grad_norm": 0.4983777403831482, "learning_rate": 8.949119227571204e-05, "loss": 0.017, "step": 74900 }, { "grad_norm": 0.15155954658985138, "learning_rate": 8.948781160629322e-05, "loss": 0.0172, "step": 74910 }, { "grad_norm": 0.1142888143658638, "learning_rate": 8.948443045706021e-05, "loss": 0.0137, "step": 74920 }, { "grad_norm": 0.16950055956840515, "learning_rate": 8.948104882805409e-05, "loss": 0.0157, "step": 74930 }, { "grad_norm": 0.13917067646980286, "learning_rate": 8.947766671931596e-05, "loss": 0.0205, "step": 74940 }, { "grad_norm": 0.11226215213537216, "learning_rate": 8.94742841308869e-05, "loss": 0.0238, "step": 74950 }, { "grad_norm": 0.1283176690340042, "learning_rate": 8.947090106280802e-05, "loss": 0.0186, "step": 74960 }, { "grad_norm": 0.11479760706424713, "learning_rate": 8.946751751512041e-05, "loss": 0.0228, "step": 74970 }, { "grad_norm": 0.13373349606990814, "learning_rate": 8.946413348786522e-05, "loss": 0.0229, "step": 74980 }, { "grad_norm": 0.13678178191184998, "learning_rate": 8.946074898108355e-05, "loss": 0.0164, "step": 74990 }, { "grad_norm": 0.12223684787750244, "learning_rate": 8.94573639948165e-05, "loss": 0.0177, "step": 75000 }, { "grad_norm": 0.14402808248996735, "learning_rate": 8.945397852910524e-05, "loss": 0.0145, "step": 75010 }, { "grad_norm": 0.15694940090179443, "learning_rate": 8.945059258399089e-05, "loss": 0.0186, "step": 75020 }, { "grad_norm": 0.17855167388916016, "learning_rate": 8.944720615951459e-05, "loss": 0.0205, "step": 75030 }, { "grad_norm": 0.1818806231021881, "learning_rate": 8.944381925571748e-05, "loss": 0.0197, "step": 75040 }, { "grad_norm": 0.13745039701461792, "learning_rate": 8.944043187264073e-05, "loss": 0.0263, "step": 75050 }, { "grad_norm": 0.1242576390504837, "learning_rate": 8.943704401032551e-05, "loss": 0.0155, "step": 75060 }, { "grad_norm": 0.21789413690567017, "learning_rate": 8.943365566881295e-05, "loss": 0.0225, "step": 75070 }, { "grad_norm": 0.11374340206384659, "learning_rate": 8.943026684814426e-05, "loss": 0.0213, "step": 75080 }, { "grad_norm": 0.14034335315227509, "learning_rate": 8.942687754836059e-05, "loss": 0.0217, "step": 75090 }, { "grad_norm": 0.12775802612304688, "learning_rate": 8.942348776950313e-05, "loss": 0.0186, "step": 75100 }, { "grad_norm": 0.23681074380874634, "learning_rate": 8.942009751161308e-05, "loss": 0.0198, "step": 75110 }, { "grad_norm": 0.236179381608963, "learning_rate": 8.941670677473162e-05, "loss": 0.0186, "step": 75120 }, { "grad_norm": 0.22973480820655823, "learning_rate": 8.941331555889995e-05, "loss": 0.0186, "step": 75130 }, { "grad_norm": 0.47938138246536255, "learning_rate": 8.94099238641593e-05, "loss": 0.027, "step": 75140 }, { "grad_norm": 0.15538658201694489, "learning_rate": 8.940653169055087e-05, "loss": 0.0204, "step": 75150 }, { "grad_norm": 0.12228372693061829, "learning_rate": 8.940313903811586e-05, "loss": 0.0196, "step": 75160 }, { "grad_norm": 0.3051622211933136, "learning_rate": 8.939974590689551e-05, "loss": 0.025, "step": 75170 }, { "grad_norm": 0.14569365978240967, "learning_rate": 8.939635229693104e-05, "loss": 0.0162, "step": 75180 }, { "grad_norm": 0.11566819995641708, "learning_rate": 8.93929582082637e-05, "loss": 0.0218, "step": 75190 }, { "grad_norm": 0.19729740917682648, "learning_rate": 8.938956364093474e-05, "loss": 0.0186, "step": 75200 }, { "grad_norm": 0.20092225074768066, "learning_rate": 8.938616859498536e-05, "loss": 0.0265, "step": 75210 }, { "grad_norm": 0.1595376878976822, "learning_rate": 8.938277307045687e-05, "loss": 0.0201, "step": 75220 }, { "grad_norm": 0.10546022653579712, "learning_rate": 8.937937706739051e-05, "loss": 0.0179, "step": 75230 }, { "grad_norm": 0.11799153685569763, "learning_rate": 8.937598058582751e-05, "loss": 0.016, "step": 75240 }, { "grad_norm": 0.15604494512081146, "learning_rate": 8.937258362580918e-05, "loss": 0.0178, "step": 75250 }, { "grad_norm": 0.16501791775226593, "learning_rate": 8.93691861873768e-05, "loss": 0.0222, "step": 75260 }, { "grad_norm": 0.11031803488731384, "learning_rate": 8.936578827057163e-05, "loss": 0.0312, "step": 75270 }, { "grad_norm": 0.1336682140827179, "learning_rate": 8.936238987543496e-05, "loss": 0.0238, "step": 75280 }, { "grad_norm": 0.14987584948539734, "learning_rate": 8.935899100200808e-05, "loss": 0.0234, "step": 75290 }, { "grad_norm": 0.1500117927789688, "learning_rate": 8.935559165033231e-05, "loss": 0.0167, "step": 75300 }, { "grad_norm": 0.17794464528560638, "learning_rate": 8.935219182044894e-05, "loss": 0.0214, "step": 75310 }, { "grad_norm": 0.11392414569854736, "learning_rate": 8.93487915123993e-05, "loss": 0.0238, "step": 75320 }, { "grad_norm": 0.10956080257892609, "learning_rate": 8.934539072622467e-05, "loss": 0.0199, "step": 75330 }, { "grad_norm": 0.13721461594104767, "learning_rate": 8.93419894619664e-05, "loss": 0.0252, "step": 75340 }, { "grad_norm": 0.15268836915493011, "learning_rate": 8.933858771966581e-05, "loss": 0.0169, "step": 75350 }, { "grad_norm": 0.22866566479206085, "learning_rate": 8.933518549936427e-05, "loss": 0.0171, "step": 75360 }, { "grad_norm": 0.19796212017536163, "learning_rate": 8.933178280110305e-05, "loss": 0.0192, "step": 75370 }, { "grad_norm": 0.1767316311597824, "learning_rate": 8.932837962492355e-05, "loss": 0.0215, "step": 75380 }, { "grad_norm": 0.3012024760246277, "learning_rate": 8.932497597086708e-05, "loss": 0.0209, "step": 75390 }, { "grad_norm": 0.868140459060669, "learning_rate": 8.932157183897505e-05, "loss": 0.0252, "step": 75400 }, { "grad_norm": 0.1706843078136444, "learning_rate": 8.931816722928878e-05, "loss": 0.0203, "step": 75410 }, { "grad_norm": 0.35442131757736206, "learning_rate": 8.931476214184968e-05, "loss": 0.0198, "step": 75420 }, { "grad_norm": 0.17262768745422363, "learning_rate": 8.931135657669909e-05, "loss": 0.0178, "step": 75430 }, { "grad_norm": 0.1547788828611374, "learning_rate": 8.930795053387839e-05, "loss": 0.0166, "step": 75440 }, { "grad_norm": 0.1965954452753067, "learning_rate": 8.930454401342899e-05, "loss": 0.0201, "step": 75450 }, { "grad_norm": 0.10053568333387375, "learning_rate": 8.930113701539227e-05, "loss": 0.0145, "step": 75460 }, { "grad_norm": 0.25148189067840576, "learning_rate": 8.929772953980962e-05, "loss": 0.0226, "step": 75470 }, { "grad_norm": 0.4252585470676422, "learning_rate": 8.929432158672245e-05, "loss": 0.0228, "step": 75480 }, { "grad_norm": 0.11967423558235168, "learning_rate": 8.929091315617219e-05, "loss": 0.0177, "step": 75490 }, { "grad_norm": 0.17126503586769104, "learning_rate": 8.928750424820023e-05, "loss": 0.0162, "step": 75500 }, { "grad_norm": 0.2972796559333801, "learning_rate": 8.928409486284799e-05, "loss": 0.0198, "step": 75510 }, { "grad_norm": 0.10566522181034088, "learning_rate": 8.928068500015691e-05, "loss": 0.0213, "step": 75520 }, { "grad_norm": 0.20805813372135162, "learning_rate": 8.927727466016843e-05, "loss": 0.0213, "step": 75530 }, { "grad_norm": 0.11966188251972198, "learning_rate": 8.927386384292397e-05, "loss": 0.0142, "step": 75540 }, { "grad_norm": 0.13784773647785187, "learning_rate": 8.9270452548465e-05, "loss": 0.019, "step": 75550 }, { "grad_norm": 0.10305169969797134, "learning_rate": 8.926704077683294e-05, "loss": 0.0164, "step": 75560 }, { "grad_norm": 0.3908858895301819, "learning_rate": 8.926362852806926e-05, "loss": 0.0187, "step": 75570 }, { "grad_norm": 0.1805647909641266, "learning_rate": 8.926021580221542e-05, "loss": 0.0135, "step": 75580 }, { "grad_norm": 0.2822469472885132, "learning_rate": 8.925680259931291e-05, "loss": 0.0234, "step": 75590 }, { "grad_norm": 0.11786238104104996, "learning_rate": 8.925338891940316e-05, "loss": 0.0183, "step": 75600 }, { "grad_norm": 0.1423412412405014, "learning_rate": 8.924997476252769e-05, "loss": 0.0181, "step": 75610 }, { "grad_norm": 0.30858415365219116, "learning_rate": 8.924656012872797e-05, "loss": 0.0219, "step": 75620 }, { "grad_norm": 0.13650068640708923, "learning_rate": 8.924314501804548e-05, "loss": 0.0143, "step": 75630 }, { "grad_norm": 0.16394557058811188, "learning_rate": 8.923972943052171e-05, "loss": 0.0184, "step": 75640 }, { "grad_norm": 0.227508082985878, "learning_rate": 8.92363133661982e-05, "loss": 0.0235, "step": 75650 }, { "grad_norm": 0.12696999311447144, "learning_rate": 8.923289682511644e-05, "loss": 0.0184, "step": 75660 }, { "grad_norm": 0.13336768746376038, "learning_rate": 8.922947980731792e-05, "loss": 0.0192, "step": 75670 }, { "grad_norm": 0.12311000376939774, "learning_rate": 8.92260623128442e-05, "loss": 0.0182, "step": 75680 }, { "grad_norm": 0.14865756034851074, "learning_rate": 8.92226443417368e-05, "loss": 0.0223, "step": 75690 }, { "grad_norm": 0.35479816794395447, "learning_rate": 8.921922589403721e-05, "loss": 0.0219, "step": 75700 }, { "grad_norm": 0.13061675429344177, "learning_rate": 8.921580696978702e-05, "loss": 0.0188, "step": 75710 }, { "grad_norm": 0.13020770251750946, "learning_rate": 8.921238756902773e-05, "loss": 0.0218, "step": 75720 }, { "grad_norm": 0.2419147789478302, "learning_rate": 8.920896769180092e-05, "loss": 0.019, "step": 75730 }, { "grad_norm": 0.13368254899978638, "learning_rate": 8.920554733814812e-05, "loss": 0.0182, "step": 75740 }, { "grad_norm": 0.13895171880722046, "learning_rate": 8.920212650811093e-05, "loss": 0.0165, "step": 75750 }, { "grad_norm": 0.10449593514204025, "learning_rate": 8.919870520173085e-05, "loss": 0.0206, "step": 75760 }, { "grad_norm": 0.7613905668258667, "learning_rate": 8.919528341904953e-05, "loss": 0.0221, "step": 75770 }, { "grad_norm": 0.134146049618721, "learning_rate": 8.91918611601085e-05, "loss": 0.0209, "step": 75780 }, { "grad_norm": 0.12169140577316284, "learning_rate": 8.918843842494935e-05, "loss": 0.0183, "step": 75790 }, { "grad_norm": 0.25764158368110657, "learning_rate": 8.918501521361366e-05, "loss": 0.0172, "step": 75800 }, { "grad_norm": 0.09848044067621231, "learning_rate": 8.918159152614306e-05, "loss": 0.0183, "step": 75810 }, { "grad_norm": 0.6797671318054199, "learning_rate": 8.917816736257911e-05, "loss": 0.023, "step": 75820 }, { "grad_norm": 0.18155142664909363, "learning_rate": 8.917474272296345e-05, "loss": 0.0279, "step": 75830 }, { "grad_norm": 0.18840138614177704, "learning_rate": 8.917131760733768e-05, "loss": 0.016, "step": 75840 }, { "grad_norm": 0.25046297907829285, "learning_rate": 8.916789201574341e-05, "loss": 0.0199, "step": 75850 }, { "grad_norm": 0.21847832202911377, "learning_rate": 8.916446594822228e-05, "loss": 0.0232, "step": 75860 }, { "grad_norm": 0.14907090365886688, "learning_rate": 8.91610394048159e-05, "loss": 0.0209, "step": 75870 }, { "grad_norm": 0.08800502866506577, "learning_rate": 8.915761238556592e-05, "loss": 0.0176, "step": 75880 }, { "grad_norm": 0.11144750565290451, "learning_rate": 8.9154184890514e-05, "loss": 0.0195, "step": 75890 }, { "grad_norm": 0.14691981673240662, "learning_rate": 8.915075691970173e-05, "loss": 0.0147, "step": 75900 }, { "grad_norm": 0.16924019157886505, "learning_rate": 8.914732847317081e-05, "loss": 0.0171, "step": 75910 }, { "grad_norm": 0.24813495576381683, "learning_rate": 8.91438995509629e-05, "loss": 0.0173, "step": 75920 }, { "grad_norm": 0.18851524591445923, "learning_rate": 8.914047015311966e-05, "loss": 0.0174, "step": 75930 }, { "grad_norm": 0.14360256493091583, "learning_rate": 8.913704027968273e-05, "loss": 0.023, "step": 75940 }, { "grad_norm": 0.08885979652404785, "learning_rate": 8.913360993069382e-05, "loss": 0.0146, "step": 75950 }, { "grad_norm": 0.20019903779029846, "learning_rate": 8.91301791061946e-05, "loss": 0.0164, "step": 75960 }, { "grad_norm": 0.20721450448036194, "learning_rate": 8.912674780622676e-05, "loss": 0.019, "step": 75970 }, { "grad_norm": 0.17336510121822357, "learning_rate": 8.912331603083199e-05, "loss": 0.0174, "step": 75980 }, { "grad_norm": 0.19796648621559143, "learning_rate": 8.911988378005199e-05, "loss": 0.0183, "step": 75990 }, { "grad_norm": 0.08157771080732346, "learning_rate": 8.911645105392848e-05, "loss": 0.0191, "step": 76000 }, { "grad_norm": 0.19218109548091888, "learning_rate": 8.911301785250314e-05, "loss": 0.0182, "step": 76010 }, { "grad_norm": 0.18506178259849548, "learning_rate": 8.910958417581772e-05, "loss": 0.0207, "step": 76020 }, { "grad_norm": 0.1444539874792099, "learning_rate": 8.910615002391392e-05, "loss": 0.0188, "step": 76030 }, { "grad_norm": 0.13900673389434814, "learning_rate": 8.910271539683346e-05, "loss": 0.0231, "step": 76040 }, { "grad_norm": 0.24862608313560486, "learning_rate": 8.90992802946181e-05, "loss": 0.018, "step": 76050 }, { "grad_norm": 0.15587744116783142, "learning_rate": 8.909584471730957e-05, "loss": 0.0189, "step": 76060 }, { "grad_norm": 0.12458018213510513, "learning_rate": 8.909240866494961e-05, "loss": 0.0143, "step": 76070 }, { "grad_norm": 0.15238730609416962, "learning_rate": 8.908897213757999e-05, "loss": 0.0198, "step": 76080 }, { "grad_norm": 0.22840581834316254, "learning_rate": 8.908553513524245e-05, "loss": 0.0203, "step": 76090 }, { "grad_norm": 0.12674210965633392, "learning_rate": 8.908209765797873e-05, "loss": 0.0172, "step": 76100 }, { "grad_norm": 0.25460299849510193, "learning_rate": 8.907865970583064e-05, "loss": 0.0154, "step": 76110 }, { "grad_norm": 0.15511676669120789, "learning_rate": 8.907522127883995e-05, "loss": 0.0236, "step": 76120 }, { "grad_norm": 0.08452187478542328, "learning_rate": 8.907178237704842e-05, "loss": 0.0141, "step": 76130 }, { "grad_norm": 0.08810712397098541, "learning_rate": 8.906834300049785e-05, "loss": 0.0209, "step": 76140 }, { "grad_norm": 0.19894075393676758, "learning_rate": 8.906490314923e-05, "loss": 0.0172, "step": 76150 }, { "grad_norm": 0.10419493913650513, "learning_rate": 8.906146282328671e-05, "loss": 0.0155, "step": 76160 }, { "grad_norm": 0.2713780403137207, "learning_rate": 8.905802202270977e-05, "loss": 0.0222, "step": 76170 }, { "grad_norm": 0.1353204846382141, "learning_rate": 8.905458074754098e-05, "loss": 0.0177, "step": 76180 }, { "grad_norm": 0.17637507617473602, "learning_rate": 8.905113899782218e-05, "loss": 0.0214, "step": 76190 }, { "grad_norm": 0.1295398473739624, "learning_rate": 8.904769677359515e-05, "loss": 0.0164, "step": 76200 }, { "grad_norm": 0.16103821992874146, "learning_rate": 8.904425407490174e-05, "loss": 0.0183, "step": 76210 }, { "grad_norm": 0.11149554699659348, "learning_rate": 8.904081090178378e-05, "loss": 0.0194, "step": 76220 }, { "grad_norm": 0.1653389185667038, "learning_rate": 8.90373672542831e-05, "loss": 0.0234, "step": 76230 }, { "grad_norm": 0.09265837073326111, "learning_rate": 8.903392313244157e-05, "loss": 0.0189, "step": 76240 }, { "grad_norm": 0.14043937623500824, "learning_rate": 8.9030478536301e-05, "loss": 0.0169, "step": 76250 }, { "grad_norm": 0.14048811793327332, "learning_rate": 8.902703346590327e-05, "loss": 0.0177, "step": 76260 }, { "grad_norm": 0.2085329294204712, "learning_rate": 8.902358792129024e-05, "loss": 0.0146, "step": 76270 }, { "grad_norm": 0.1312103122472763, "learning_rate": 8.902014190250377e-05, "loss": 0.0191, "step": 76280 }, { "grad_norm": 0.23731650412082672, "learning_rate": 8.901669540958574e-05, "loss": 0.0194, "step": 76290 }, { "grad_norm": 0.19520387053489685, "learning_rate": 8.901324844257802e-05, "loss": 0.0143, "step": 76300 }, { "grad_norm": 0.1339445263147354, "learning_rate": 8.90098010015225e-05, "loss": 0.0211, "step": 76310 }, { "grad_norm": 0.19084572792053223, "learning_rate": 8.900635308646106e-05, "loss": 0.0162, "step": 76320 }, { "grad_norm": 0.13359901309013367, "learning_rate": 8.900290469743559e-05, "loss": 0.0137, "step": 76330 }, { "grad_norm": 0.15272891521453857, "learning_rate": 8.899945583448803e-05, "loss": 0.0189, "step": 76340 }, { "grad_norm": 0.1686047911643982, "learning_rate": 8.899600649766024e-05, "loss": 0.0261, "step": 76350 }, { "grad_norm": 0.13232247531414032, "learning_rate": 8.899255668699415e-05, "loss": 0.018, "step": 76360 }, { "grad_norm": 0.20565493404865265, "learning_rate": 8.898910640253171e-05, "loss": 0.016, "step": 76370 }, { "grad_norm": 0.2233867347240448, "learning_rate": 8.898565564431478e-05, "loss": 0.0142, "step": 76380 }, { "grad_norm": 0.15187609195709229, "learning_rate": 8.898220441238534e-05, "loss": 0.0187, "step": 76390 }, { "grad_norm": 0.13576383888721466, "learning_rate": 8.897875270678532e-05, "loss": 0.019, "step": 76400 }, { "grad_norm": 0.18422773480415344, "learning_rate": 8.897530052755665e-05, "loss": 0.018, "step": 76410 }, { "grad_norm": 0.11637714505195618, "learning_rate": 8.897184787474127e-05, "loss": 0.0168, "step": 76420 }, { "grad_norm": 0.1740257292985916, "learning_rate": 8.896839474838115e-05, "loss": 0.0208, "step": 76430 }, { "grad_norm": 0.11835543811321259, "learning_rate": 8.896494114851823e-05, "loss": 0.02, "step": 76440 }, { "grad_norm": 0.1255938857793808, "learning_rate": 8.89614870751945e-05, "loss": 0.0214, "step": 76450 }, { "grad_norm": 0.19202616810798645, "learning_rate": 8.895803252845192e-05, "loss": 0.0245, "step": 76460 }, { "grad_norm": 0.11094332486391068, "learning_rate": 8.895457750833244e-05, "loss": 0.0174, "step": 76470 }, { "grad_norm": 0.09077750146389008, "learning_rate": 8.895112201487808e-05, "loss": 0.0173, "step": 76480 }, { "grad_norm": 0.10075956583023071, "learning_rate": 8.89476660481308e-05, "loss": 0.0258, "step": 76490 }, { "grad_norm": 0.11717704683542252, "learning_rate": 8.894420960813263e-05, "loss": 0.0179, "step": 76500 }, { "grad_norm": 0.23397959768772125, "learning_rate": 8.894075269492552e-05, "loss": 0.0226, "step": 76510 }, { "grad_norm": 0.1995154619216919, "learning_rate": 8.89372953085515e-05, "loss": 0.0176, "step": 76520 }, { "grad_norm": 0.19606292247772217, "learning_rate": 8.893383744905259e-05, "loss": 0.0183, "step": 76530 }, { "grad_norm": 0.22417648136615753, "learning_rate": 8.893037911647079e-05, "loss": 0.0231, "step": 76540 }, { "grad_norm": 0.5906581878662109, "learning_rate": 8.892692031084814e-05, "loss": 0.0231, "step": 76550 }, { "grad_norm": 0.14188739657402039, "learning_rate": 8.892346103222664e-05, "loss": 0.0196, "step": 76560 }, { "grad_norm": 0.13952191174030304, "learning_rate": 8.892000128064834e-05, "loss": 0.0262, "step": 76570 }, { "grad_norm": 0.17691951990127563, "learning_rate": 8.891654105615528e-05, "loss": 0.0197, "step": 76580 }, { "grad_norm": 0.19097009301185608, "learning_rate": 8.891308035878951e-05, "loss": 0.022, "step": 76590 }, { "grad_norm": 0.16431452333927155, "learning_rate": 8.890961918859306e-05, "loss": 0.0186, "step": 76600 }, { "grad_norm": 0.2195304036140442, "learning_rate": 8.890615754560801e-05, "loss": 0.0226, "step": 76610 }, { "grad_norm": 0.16719217598438263, "learning_rate": 8.890269542987642e-05, "loss": 0.0145, "step": 76620 }, { "grad_norm": 0.22192589938640594, "learning_rate": 8.889923284144033e-05, "loss": 0.0185, "step": 76630 }, { "grad_norm": 0.10575196146965027, "learning_rate": 8.889576978034184e-05, "loss": 0.0161, "step": 76640 }, { "grad_norm": 0.14084911346435547, "learning_rate": 8.889230624662303e-05, "loss": 0.0233, "step": 76650 }, { "grad_norm": 0.10504938662052155, "learning_rate": 8.888884224032598e-05, "loss": 0.0209, "step": 76660 }, { "grad_norm": 0.19483859837055206, "learning_rate": 8.888537776149277e-05, "loss": 0.0213, "step": 76670 }, { "grad_norm": 0.2795093059539795, "learning_rate": 8.888191281016551e-05, "loss": 0.0244, "step": 76680 }, { "grad_norm": 0.1398412436246872, "learning_rate": 8.88784473863863e-05, "loss": 0.0187, "step": 76690 }, { "grad_norm": 0.129315584897995, "learning_rate": 8.887498149019725e-05, "loss": 0.0201, "step": 76700 }, { "grad_norm": 0.11079253256320953, "learning_rate": 8.887151512164048e-05, "loss": 0.0156, "step": 76710 }, { "grad_norm": 0.16691644489765167, "learning_rate": 8.886804828075807e-05, "loss": 0.0193, "step": 76720 }, { "grad_norm": 0.16893137991428375, "learning_rate": 8.88645809675922e-05, "loss": 0.0216, "step": 76730 }, { "grad_norm": 0.1594657003879547, "learning_rate": 8.886111318218498e-05, "loss": 0.0201, "step": 76740 }, { "grad_norm": 0.15168385207653046, "learning_rate": 8.885764492457854e-05, "loss": 0.0177, "step": 76750 }, { "grad_norm": 0.18793882429599762, "learning_rate": 8.8854176194815e-05, "loss": 0.0193, "step": 76760 }, { "grad_norm": 0.218228280544281, "learning_rate": 8.885070699293656e-05, "loss": 0.0237, "step": 76770 }, { "grad_norm": 0.13703280687332153, "learning_rate": 8.884723731898536e-05, "loss": 0.0256, "step": 76780 }, { "grad_norm": 0.22054484486579895, "learning_rate": 8.884376717300354e-05, "loss": 0.019, "step": 76790 }, { "grad_norm": 0.17922180891036987, "learning_rate": 8.884029655503325e-05, "loss": 0.0188, "step": 76800 }, { "grad_norm": 0.15911166369915009, "learning_rate": 8.883682546511671e-05, "loss": 0.0205, "step": 76810 }, { "grad_norm": 0.27170997858047485, "learning_rate": 8.883335390329607e-05, "loss": 0.0216, "step": 76820 }, { "grad_norm": 0.1905769258737564, "learning_rate": 8.88298818696135e-05, "loss": 0.0224, "step": 76830 }, { "grad_norm": 0.12388460338115692, "learning_rate": 8.882640936411122e-05, "loss": 0.0152, "step": 76840 }, { "grad_norm": 0.2203727662563324, "learning_rate": 8.882293638683141e-05, "loss": 0.0231, "step": 76850 }, { "grad_norm": 0.13303247094154358, "learning_rate": 8.881946293781626e-05, "loss": 0.0199, "step": 76860 }, { "grad_norm": 0.28475677967071533, "learning_rate": 8.881598901710798e-05, "loss": 0.0186, "step": 76870 }, { "grad_norm": 0.09620364010334015, "learning_rate": 8.88125146247488e-05, "loss": 0.0216, "step": 76880 }, { "grad_norm": 0.12538820505142212, "learning_rate": 8.88090397607809e-05, "loss": 0.023, "step": 76890 }, { "grad_norm": 0.19351479411125183, "learning_rate": 8.880556442524655e-05, "loss": 0.0173, "step": 76900 }, { "grad_norm": 0.24573896825313568, "learning_rate": 8.880208861818792e-05, "loss": 0.0171, "step": 76910 }, { "grad_norm": 0.22310595214366913, "learning_rate": 8.879861233964732e-05, "loss": 0.0205, "step": 76920 }, { "grad_norm": 0.14416827261447906, "learning_rate": 8.879513558966693e-05, "loss": 0.0134, "step": 76930 }, { "grad_norm": 0.08125440776348114, "learning_rate": 8.8791658368289e-05, "loss": 0.021, "step": 76940 }, { "grad_norm": 0.15023063123226166, "learning_rate": 8.87881806755558e-05, "loss": 0.0177, "step": 76950 }, { "grad_norm": 0.16877271234989166, "learning_rate": 8.878470251150959e-05, "loss": 0.0218, "step": 76960 }, { "grad_norm": 0.28441882133483887, "learning_rate": 8.878122387619264e-05, "loss": 0.0238, "step": 76970 }, { "grad_norm": 0.2690139412879944, "learning_rate": 8.877774476964718e-05, "loss": 0.0214, "step": 76980 }, { "grad_norm": 0.1267445981502533, "learning_rate": 8.877426519191552e-05, "loss": 0.0176, "step": 76990 }, { "grad_norm": 0.2608370780944824, "learning_rate": 8.877078514303993e-05, "loss": 0.0195, "step": 77000 }, { "grad_norm": 0.21813765168190002, "learning_rate": 8.876730462306268e-05, "loss": 0.0187, "step": 77010 }, { "grad_norm": 0.20786936581134796, "learning_rate": 8.87638236320261e-05, "loss": 0.022, "step": 77020 }, { "grad_norm": 0.23943714797496796, "learning_rate": 8.876034216997245e-05, "loss": 0.0222, "step": 77030 }, { "grad_norm": 0.23822259902954102, "learning_rate": 8.875686023694404e-05, "loss": 0.0186, "step": 77040 }, { "grad_norm": 0.14173264801502228, "learning_rate": 8.875337783298321e-05, "loss": 0.0175, "step": 77050 }, { "grad_norm": 0.21822786331176758, "learning_rate": 8.874989495813223e-05, "loss": 0.0187, "step": 77060 }, { "grad_norm": 0.18296614289283752, "learning_rate": 8.874641161243345e-05, "loss": 0.0235, "step": 77070 }, { "grad_norm": 0.21819919347763062, "learning_rate": 8.874292779592918e-05, "loss": 0.0178, "step": 77080 }, { "grad_norm": 0.1473543345928192, "learning_rate": 8.873944350866175e-05, "loss": 0.0183, "step": 77090 }, { "grad_norm": 0.11387893557548523, "learning_rate": 8.873595875067352e-05, "loss": 0.0217, "step": 77100 }, { "grad_norm": 0.42332619428634644, "learning_rate": 8.87324735220068e-05, "loss": 0.027, "step": 77110 }, { "grad_norm": 0.248310387134552, "learning_rate": 8.872898782270397e-05, "loss": 0.0242, "step": 77120 }, { "grad_norm": 0.14179348945617676, "learning_rate": 8.872550165280738e-05, "loss": 0.0208, "step": 77130 }, { "grad_norm": 0.16219022870063782, "learning_rate": 8.872201501235937e-05, "loss": 0.0165, "step": 77140 }, { "grad_norm": 0.11120424419641495, "learning_rate": 8.871852790140233e-05, "loss": 0.0179, "step": 77150 }, { "grad_norm": 0.2435186803340912, "learning_rate": 8.871504031997862e-05, "loss": 0.02, "step": 77160 }, { "grad_norm": 0.10227540880441666, "learning_rate": 8.87115522681306e-05, "loss": 0.0177, "step": 77170 }, { "grad_norm": 0.2409019023180008, "learning_rate": 8.870806374590068e-05, "loss": 0.0223, "step": 77180 }, { "grad_norm": 0.3043392300605774, "learning_rate": 8.870457475333123e-05, "loss": 0.0205, "step": 77190 }, { "grad_norm": 0.17156578600406647, "learning_rate": 8.870108529046466e-05, "loss": 0.0173, "step": 77200 }, { "grad_norm": 0.13268519937992096, "learning_rate": 8.869759535734336e-05, "loss": 0.0264, "step": 77210 }, { "grad_norm": 0.10698169469833374, "learning_rate": 8.869410495400976e-05, "loss": 0.0156, "step": 77220 }, { "grad_norm": 0.1764843612909317, "learning_rate": 8.869061408050623e-05, "loss": 0.0242, "step": 77230 }, { "grad_norm": 0.16287273168563843, "learning_rate": 8.868712273687521e-05, "loss": 0.0292, "step": 77240 }, { "grad_norm": 0.13100233674049377, "learning_rate": 8.868363092315914e-05, "loss": 0.0187, "step": 77250 }, { "grad_norm": 0.1312282681465149, "learning_rate": 8.868013863940041e-05, "loss": 0.0222, "step": 77260 }, { "grad_norm": 0.11389719694852829, "learning_rate": 8.867664588564148e-05, "loss": 0.017, "step": 77270 }, { "grad_norm": 0.17961573600769043, "learning_rate": 8.867315266192479e-05, "loss": 0.0187, "step": 77280 }, { "grad_norm": 0.1608220934867859, "learning_rate": 8.866965896829278e-05, "loss": 0.0157, "step": 77290 }, { "grad_norm": 0.15822456777095795, "learning_rate": 8.866616480478791e-05, "loss": 0.0199, "step": 77300 }, { "grad_norm": 0.1446225941181183, "learning_rate": 8.866267017145262e-05, "loss": 0.0202, "step": 77310 }, { "grad_norm": 0.1763424426317215, "learning_rate": 8.865917506832939e-05, "loss": 0.0175, "step": 77320 }, { "grad_norm": 0.20786583423614502, "learning_rate": 8.865567949546068e-05, "loss": 0.0202, "step": 77330 }, { "grad_norm": 0.17489945888519287, "learning_rate": 8.865218345288897e-05, "loss": 0.0195, "step": 77340 }, { "grad_norm": 0.1660374402999878, "learning_rate": 8.864868694065674e-05, "loss": 0.0213, "step": 77350 }, { "grad_norm": 0.14519262313842773, "learning_rate": 8.864518995880646e-05, "loss": 0.0225, "step": 77360 }, { "grad_norm": 0.11913421005010605, "learning_rate": 8.864169250738064e-05, "loss": 0.0214, "step": 77370 }, { "grad_norm": 0.19508115947246552, "learning_rate": 8.863819458642179e-05, "loss": 0.0198, "step": 77380 }, { "grad_norm": 0.3594183325767517, "learning_rate": 8.863469619597238e-05, "loss": 0.0209, "step": 77390 }, { "grad_norm": 0.17811399698257446, "learning_rate": 8.863119733607493e-05, "loss": 0.0241, "step": 77400 }, { "grad_norm": 1.1624436378479004, "learning_rate": 8.862769800677197e-05, "loss": 0.0164, "step": 77410 }, { "grad_norm": 0.21403120458126068, "learning_rate": 8.8624198208106e-05, "loss": 0.0231, "step": 77420 }, { "grad_norm": 0.22279223799705505, "learning_rate": 8.862069794011956e-05, "loss": 0.0184, "step": 77430 }, { "grad_norm": 0.37257543206214905, "learning_rate": 8.861719720285517e-05, "loss": 0.0293, "step": 77440 }, { "grad_norm": 0.15319150686264038, "learning_rate": 8.861369599635538e-05, "loss": 0.0232, "step": 77450 }, { "grad_norm": 0.13655340671539307, "learning_rate": 8.861019432066272e-05, "loss": 0.0241, "step": 77460 }, { "grad_norm": 0.17055019736289978, "learning_rate": 8.860669217581975e-05, "loss": 0.02, "step": 77470 }, { "grad_norm": 0.16750866174697876, "learning_rate": 8.860318956186901e-05, "loss": 0.0196, "step": 77480 }, { "grad_norm": 0.10934819281101227, "learning_rate": 8.859968647885309e-05, "loss": 0.0138, "step": 77490 }, { "grad_norm": 0.18181827664375305, "learning_rate": 8.859618292681451e-05, "loss": 0.0223, "step": 77500 }, { "grad_norm": 0.12319652736186981, "learning_rate": 8.859267890579587e-05, "loss": 0.0196, "step": 77510 }, { "grad_norm": 0.1916998326778412, "learning_rate": 8.858917441583976e-05, "loss": 0.0198, "step": 77520 }, { "grad_norm": 0.1398032009601593, "learning_rate": 8.858566945698874e-05, "loss": 0.0225, "step": 77530 }, { "grad_norm": 0.17058366537094116, "learning_rate": 8.858216402928541e-05, "loss": 0.0205, "step": 77540 }, { "grad_norm": 0.11107078194618225, "learning_rate": 8.857865813277235e-05, "loss": 0.0229, "step": 77550 }, { "grad_norm": 0.287963330745697, "learning_rate": 8.857515176749217e-05, "loss": 0.0196, "step": 77560 }, { "grad_norm": 0.24820758402347565, "learning_rate": 8.857164493348749e-05, "loss": 0.0141, "step": 77570 }, { "grad_norm": 0.11627227813005447, "learning_rate": 8.856813763080089e-05, "loss": 0.018, "step": 77580 }, { "grad_norm": 0.11381760239601135, "learning_rate": 8.8564629859475e-05, "loss": 0.019, "step": 77590 }, { "grad_norm": 0.18173415958881378, "learning_rate": 8.856112161955245e-05, "loss": 0.0176, "step": 77600 }, { "grad_norm": 0.21383512020111084, "learning_rate": 8.855761291107587e-05, "loss": 0.0206, "step": 77610 }, { "grad_norm": 0.20214328169822693, "learning_rate": 8.85541037340879e-05, "loss": 0.0176, "step": 77620 }, { "grad_norm": 0.12291538715362549, "learning_rate": 8.855059408863114e-05, "loss": 0.0236, "step": 77630 }, { "grad_norm": 0.8196642398834229, "learning_rate": 8.854708397474827e-05, "loss": 0.0231, "step": 77640 }, { "grad_norm": 0.14809368550777435, "learning_rate": 8.854357339248196e-05, "loss": 0.0163, "step": 77650 }, { "grad_norm": 0.12438271194696426, "learning_rate": 8.854006234187482e-05, "loss": 0.0213, "step": 77660 }, { "grad_norm": 0.1896716207265854, "learning_rate": 8.853655082296955e-05, "loss": 0.0156, "step": 77670 }, { "grad_norm": 0.34655603766441345, "learning_rate": 8.85330388358088e-05, "loss": 0.0245, "step": 77680 }, { "grad_norm": 0.3489059507846832, "learning_rate": 8.852952638043523e-05, "loss": 0.0233, "step": 77690 }, { "grad_norm": 0.35052478313446045, "learning_rate": 8.852601345689155e-05, "loss": 0.0195, "step": 77700 }, { "grad_norm": 0.19185343384742737, "learning_rate": 8.852250006522042e-05, "loss": 0.0214, "step": 77710 }, { "grad_norm": 0.235655277967453, "learning_rate": 8.851898620546456e-05, "loss": 0.0234, "step": 77720 }, { "grad_norm": 0.1264650523662567, "learning_rate": 8.851547187766665e-05, "loss": 0.0186, "step": 77730 }, { "grad_norm": 0.2675684690475464, "learning_rate": 8.851195708186938e-05, "loss": 0.0197, "step": 77740 }, { "grad_norm": 0.2602648437023163, "learning_rate": 8.850844181811547e-05, "loss": 0.0223, "step": 77750 }, { "grad_norm": 0.1625760793685913, "learning_rate": 8.850492608644763e-05, "loss": 0.018, "step": 77760 }, { "grad_norm": 0.12106417119503021, "learning_rate": 8.85014098869086e-05, "loss": 0.0222, "step": 77770 }, { "grad_norm": 0.13440777361392975, "learning_rate": 8.849789321954109e-05, "loss": 0.02, "step": 77780 }, { "grad_norm": 0.1885388344526291, "learning_rate": 8.849437608438782e-05, "loss": 0.0215, "step": 77790 }, { "grad_norm": 0.2341417372226715, "learning_rate": 8.849085848149152e-05, "loss": 0.02, "step": 77800 }, { "grad_norm": 0.21551604568958282, "learning_rate": 8.848734041089497e-05, "loss": 0.0269, "step": 77810 }, { "grad_norm": 0.12277509272098541, "learning_rate": 8.84838218726409e-05, "loss": 0.0159, "step": 77820 }, { "grad_norm": 0.1301765739917755, "learning_rate": 8.848030286677205e-05, "loss": 0.0177, "step": 77830 }, { "grad_norm": 0.18791481852531433, "learning_rate": 8.847678339333119e-05, "loss": 0.0234, "step": 77840 }, { "grad_norm": 0.4014081060886383, "learning_rate": 8.847326345236106e-05, "loss": 0.0207, "step": 77850 }, { "grad_norm": 0.2521534562110901, "learning_rate": 8.846974304390449e-05, "loss": 0.0232, "step": 77860 }, { "grad_norm": 0.18408317863941193, "learning_rate": 8.84662221680042e-05, "loss": 0.0226, "step": 77870 }, { "grad_norm": 0.16986003518104553, "learning_rate": 8.8462700824703e-05, "loss": 0.0253, "step": 77880 }, { "grad_norm": 0.11342357099056244, "learning_rate": 8.845917901404368e-05, "loss": 0.0213, "step": 77890 }, { "grad_norm": 0.2591092586517334, "learning_rate": 8.8455656736069e-05, "loss": 0.0316, "step": 77900 }, { "grad_norm": 0.13066667318344116, "learning_rate": 8.84521339908218e-05, "loss": 0.0258, "step": 77910 }, { "grad_norm": 0.18083599209785461, "learning_rate": 8.844861077834486e-05, "loss": 0.0228, "step": 77920 }, { "grad_norm": 0.2091190069913864, "learning_rate": 8.8445087098681e-05, "loss": 0.0374, "step": 77930 }, { "grad_norm": 0.1305617392063141, "learning_rate": 8.844156295187303e-05, "loss": 0.0326, "step": 77940 }, { "grad_norm": 0.3411625623703003, "learning_rate": 8.843803833796378e-05, "loss": 0.0284, "step": 77950 }, { "grad_norm": 0.2326718419790268, "learning_rate": 8.843451325699608e-05, "loss": 0.0279, "step": 77960 }, { "grad_norm": 0.20408286154270172, "learning_rate": 8.843098770901274e-05, "loss": 0.0198, "step": 77970 }, { "grad_norm": 0.3343762755393982, "learning_rate": 8.842746169405663e-05, "loss": 0.0285, "step": 77980 }, { "grad_norm": 0.15044082701206207, "learning_rate": 8.842393521217057e-05, "loss": 0.0209, "step": 77990 }, { "grad_norm": 0.5651640295982361, "learning_rate": 8.842040826339741e-05, "loss": 0.0292, "step": 78000 }, { "grad_norm": 0.15949517488479614, "learning_rate": 8.841688084778001e-05, "loss": 0.0306, "step": 78010 }, { "grad_norm": 0.1625763177871704, "learning_rate": 8.841335296536127e-05, "loss": 0.0236, "step": 78020 }, { "grad_norm": 0.10999567806720734, "learning_rate": 8.8409824616184e-05, "loss": 0.0219, "step": 78030 }, { "grad_norm": 0.11818453669548035, "learning_rate": 8.840629580029109e-05, "loss": 0.0264, "step": 78040 }, { "grad_norm": 0.15832415223121643, "learning_rate": 8.840276651772544e-05, "loss": 0.0227, "step": 78050 }, { "grad_norm": 0.21363545954227448, "learning_rate": 8.839923676852992e-05, "loss": 0.0196, "step": 78060 }, { "grad_norm": 0.11411435902118683, "learning_rate": 8.83957065527474e-05, "loss": 0.023, "step": 78070 }, { "grad_norm": 0.10364216566085815, "learning_rate": 8.839217587042081e-05, "loss": 0.0137, "step": 78080 }, { "grad_norm": 0.21183758974075317, "learning_rate": 8.838864472159303e-05, "loss": 0.0215, "step": 78090 }, { "grad_norm": 0.14293456077575684, "learning_rate": 8.838511310630697e-05, "loss": 0.0218, "step": 78100 }, { "grad_norm": 0.17964904010295868, "learning_rate": 8.838158102460554e-05, "loss": 0.0266, "step": 78110 }, { "grad_norm": 0.12492601573467255, "learning_rate": 8.837804847653168e-05, "loss": 0.0193, "step": 78120 }, { "grad_norm": 0.16391293704509735, "learning_rate": 8.837451546212827e-05, "loss": 0.021, "step": 78130 }, { "grad_norm": 0.18658527731895447, "learning_rate": 8.83709819814383e-05, "loss": 0.0263, "step": 78140 }, { "grad_norm": 0.1386057585477829, "learning_rate": 8.836744803450464e-05, "loss": 0.0229, "step": 78150 }, { "grad_norm": 0.18714965879917145, "learning_rate": 8.836391362137027e-05, "loss": 0.0211, "step": 78160 }, { "grad_norm": 0.14686670899391174, "learning_rate": 8.836037874207813e-05, "loss": 0.022, "step": 78170 }, { "grad_norm": 0.27302321791648865, "learning_rate": 8.835684339667117e-05, "loss": 0.0196, "step": 78180 }, { "grad_norm": 0.19287988543510437, "learning_rate": 8.835330758519236e-05, "loss": 0.0378, "step": 78190 }, { "grad_norm": 0.2308647781610489, "learning_rate": 8.834977130768462e-05, "loss": 0.0233, "step": 78200 }, { "grad_norm": 0.11378063261508942, "learning_rate": 8.834623456419097e-05, "loss": 0.0194, "step": 78210 }, { "grad_norm": 0.37821832299232483, "learning_rate": 8.834269735475438e-05, "loss": 0.0223, "step": 78220 }, { "grad_norm": 0.14275339245796204, "learning_rate": 8.83391596794178e-05, "loss": 0.0169, "step": 78230 }, { "grad_norm": 0.1425836682319641, "learning_rate": 8.833562153822422e-05, "loss": 0.0165, "step": 78240 }, { "grad_norm": 0.22552628815174103, "learning_rate": 8.833208293121667e-05, "loss": 0.018, "step": 78250 }, { "grad_norm": 0.11782576143741608, "learning_rate": 8.83285438584381e-05, "loss": 0.0188, "step": 78260 }, { "grad_norm": 0.16148850321769714, "learning_rate": 8.832500431993153e-05, "loss": 0.0235, "step": 78270 }, { "grad_norm": 0.28530073165893555, "learning_rate": 8.832146431573999e-05, "loss": 0.021, "step": 78280 }, { "grad_norm": 0.10953468829393387, "learning_rate": 8.831792384590647e-05, "loss": 0.0218, "step": 78290 }, { "grad_norm": 0.18805745244026184, "learning_rate": 8.831438291047401e-05, "loss": 0.0298, "step": 78300 }, { "grad_norm": 0.1607741266489029, "learning_rate": 8.83108415094856e-05, "loss": 0.0228, "step": 78310 }, { "grad_norm": 0.12420926243066788, "learning_rate": 8.83072996429843e-05, "loss": 0.0204, "step": 78320 }, { "grad_norm": 0.1488858461380005, "learning_rate": 8.830375731101315e-05, "loss": 0.0167, "step": 78330 }, { "grad_norm": 0.11530506610870361, "learning_rate": 8.830021451361518e-05, "loss": 0.0202, "step": 78340 }, { "grad_norm": 0.15080146491527557, "learning_rate": 8.829667125083344e-05, "loss": 0.0196, "step": 78350 }, { "grad_norm": 0.23738302290439606, "learning_rate": 8.8293127522711e-05, "loss": 0.0252, "step": 78360 }, { "grad_norm": 0.10198121517896652, "learning_rate": 8.828958332929088e-05, "loss": 0.0157, "step": 78370 }, { "grad_norm": 0.12714751064777374, "learning_rate": 8.828603867061619e-05, "loss": 0.0157, "step": 78380 }, { "grad_norm": 0.2026783674955368, "learning_rate": 8.828249354672997e-05, "loss": 0.021, "step": 78390 }, { "grad_norm": 0.16140888631343842, "learning_rate": 8.827894795767533e-05, "loss": 0.0212, "step": 78400 }, { "grad_norm": 0.1419668346643448, "learning_rate": 8.82754019034953e-05, "loss": 0.0187, "step": 78410 }, { "grad_norm": 0.13390229642391205, "learning_rate": 8.827185538423302e-05, "loss": 0.018, "step": 78420 }, { "grad_norm": 0.13080275058746338, "learning_rate": 8.826830839993157e-05, "loss": 0.0143, "step": 78430 }, { "grad_norm": 0.11915198713541031, "learning_rate": 8.826476095063403e-05, "loss": 0.0175, "step": 78440 }, { "grad_norm": 0.2012500911951065, "learning_rate": 8.826121303638351e-05, "loss": 0.0163, "step": 78450 }, { "grad_norm": 0.12652158737182617, "learning_rate": 8.825766465722313e-05, "loss": 0.0189, "step": 78460 }, { "grad_norm": 0.14570288360118866, "learning_rate": 8.825411581319601e-05, "loss": 0.0182, "step": 78470 }, { "grad_norm": 0.13067542016506195, "learning_rate": 8.825056650434527e-05, "loss": 0.0203, "step": 78480 }, { "grad_norm": 0.22136062383651733, "learning_rate": 8.824701673071402e-05, "loss": 0.0265, "step": 78490 }, { "grad_norm": 0.1476549655199051, "learning_rate": 8.82434664923454e-05, "loss": 0.0171, "step": 78500 }, { "grad_norm": 0.12297544628381729, "learning_rate": 8.823991578928256e-05, "loss": 0.0209, "step": 78510 }, { "grad_norm": 0.20357933640480042, "learning_rate": 8.823636462156866e-05, "loss": 0.0204, "step": 78520 }, { "grad_norm": 0.1704193353652954, "learning_rate": 8.823281298924681e-05, "loss": 0.017, "step": 78530 }, { "grad_norm": 0.1645328253507614, "learning_rate": 8.822926089236019e-05, "loss": 0.0172, "step": 78540 }, { "grad_norm": 0.09288160502910614, "learning_rate": 8.822570833095193e-05, "loss": 0.0193, "step": 78550 }, { "grad_norm": 0.24958311021327972, "learning_rate": 8.822215530506526e-05, "loss": 0.0179, "step": 78560 }, { "grad_norm": 0.12148735672235489, "learning_rate": 8.82186018147433e-05, "loss": 0.0248, "step": 78570 }, { "grad_norm": 0.1139117032289505, "learning_rate": 8.821504786002925e-05, "loss": 0.0163, "step": 78580 }, { "grad_norm": 0.11363617330789566, "learning_rate": 8.821149344096628e-05, "loss": 0.0185, "step": 78590 }, { "grad_norm": 0.186635822057724, "learning_rate": 8.820793855759758e-05, "loss": 0.0147, "step": 78600 }, { "grad_norm": 0.10300447046756744, "learning_rate": 8.820438320996637e-05, "loss": 0.0163, "step": 78610 }, { "grad_norm": 0.11862882226705551, "learning_rate": 8.820082739811582e-05, "loss": 0.016, "step": 78620 }, { "grad_norm": 0.17280213534832, "learning_rate": 8.819727112208916e-05, "loss": 0.0177, "step": 78630 }, { "grad_norm": 0.20490117371082306, "learning_rate": 8.819371438192958e-05, "loss": 0.0157, "step": 78640 }, { "grad_norm": 0.28216755390167236, "learning_rate": 8.819015717768032e-05, "loss": 0.0236, "step": 78650 }, { "grad_norm": 0.12161198258399963, "learning_rate": 8.818659950938457e-05, "loss": 0.0209, "step": 78660 }, { "grad_norm": 0.16807539761066437, "learning_rate": 8.81830413770856e-05, "loss": 0.0204, "step": 78670 }, { "grad_norm": 0.11078451573848724, "learning_rate": 8.817948278082663e-05, "loss": 0.0134, "step": 78680 }, { "grad_norm": 0.21769213676452637, "learning_rate": 8.817592372065089e-05, "loss": 0.0178, "step": 78690 }, { "grad_norm": 0.1150120198726654, "learning_rate": 8.817236419660164e-05, "loss": 0.0166, "step": 78700 }, { "grad_norm": 0.27325645089149475, "learning_rate": 8.816880420872212e-05, "loss": 0.0163, "step": 78710 }, { "grad_norm": 0.10486552864313126, "learning_rate": 8.816524375705559e-05, "loss": 0.0183, "step": 78720 }, { "grad_norm": 0.11024220287799835, "learning_rate": 8.816168284164531e-05, "loss": 0.0189, "step": 78730 }, { "grad_norm": 0.6234546303749084, "learning_rate": 8.815812146253456e-05, "loss": 0.0187, "step": 78740 }, { "grad_norm": 0.1735508143901825, "learning_rate": 8.815455961976661e-05, "loss": 0.0212, "step": 78750 }, { "grad_norm": 0.1400803178548813, "learning_rate": 8.815099731338472e-05, "loss": 0.0189, "step": 78760 }, { "grad_norm": 0.13573862612247467, "learning_rate": 8.814743454343221e-05, "loss": 0.0341, "step": 78770 }, { "grad_norm": 0.13801996409893036, "learning_rate": 8.814387130995234e-05, "loss": 0.0237, "step": 78780 }, { "grad_norm": 0.17923399806022644, "learning_rate": 8.814030761298843e-05, "loss": 0.0227, "step": 78790 }, { "grad_norm": 0.21810433268547058, "learning_rate": 8.813674345258378e-05, "loss": 0.035, "step": 78800 }, { "grad_norm": 0.2291831225156784, "learning_rate": 8.813317882878168e-05, "loss": 0.0225, "step": 78810 }, { "grad_norm": 0.15631598234176636, "learning_rate": 8.812961374162546e-05, "loss": 0.0195, "step": 78820 }, { "grad_norm": 0.14416810870170593, "learning_rate": 8.812604819115842e-05, "loss": 0.02, "step": 78830 }, { "grad_norm": 0.15395916998386383, "learning_rate": 8.812248217742389e-05, "loss": 0.0223, "step": 78840 }, { "grad_norm": 0.1461600512266159, "learning_rate": 8.811891570046523e-05, "loss": 0.0214, "step": 78850 }, { "grad_norm": 0.12312854826450348, "learning_rate": 8.811534876032575e-05, "loss": 0.0172, "step": 78860 }, { "grad_norm": 0.1854751706123352, "learning_rate": 8.811178135704879e-05, "loss": 0.0228, "step": 78870 }, { "grad_norm": 0.16618534922599792, "learning_rate": 8.81082134906777e-05, "loss": 0.0223, "step": 78880 }, { "grad_norm": 0.13707929849624634, "learning_rate": 8.810464516125582e-05, "loss": 0.0184, "step": 78890 }, { "grad_norm": 0.13117451965808868, "learning_rate": 8.810107636882656e-05, "loss": 0.0258, "step": 78900 }, { "grad_norm": 0.15586444735527039, "learning_rate": 8.809750711343323e-05, "loss": 0.017, "step": 78910 }, { "grad_norm": 0.1374531388282776, "learning_rate": 8.809393739511922e-05, "loss": 0.0188, "step": 78920 }, { "grad_norm": 0.10654664784669876, "learning_rate": 8.80903672139279e-05, "loss": 0.0305, "step": 78930 }, { "grad_norm": 0.12583401799201965, "learning_rate": 8.808679656990266e-05, "loss": 0.0197, "step": 78940 }, { "grad_norm": 0.08060403913259506, "learning_rate": 8.80832254630869e-05, "loss": 0.0184, "step": 78950 }, { "grad_norm": 0.14950579404830933, "learning_rate": 8.807965389352397e-05, "loss": 0.017, "step": 78960 }, { "grad_norm": 0.21834878623485565, "learning_rate": 8.807608186125729e-05, "loss": 0.0203, "step": 78970 }, { "grad_norm": 0.18467721343040466, "learning_rate": 8.807250936633026e-05, "loss": 0.0192, "step": 78980 }, { "grad_norm": 0.11763587594032288, "learning_rate": 8.806893640878632e-05, "loss": 0.0183, "step": 78990 }, { "grad_norm": 0.24020838737487793, "learning_rate": 8.806536298866886e-05, "loss": 0.0132, "step": 79000 }, { "grad_norm": 0.23603218793869019, "learning_rate": 8.806178910602127e-05, "loss": 0.0193, "step": 79010 }, { "grad_norm": 0.18559850752353668, "learning_rate": 8.805821476088704e-05, "loss": 0.0218, "step": 79020 }, { "grad_norm": 0.17004428803920746, "learning_rate": 8.805463995330955e-05, "loss": 0.0196, "step": 79030 }, { "grad_norm": 0.11927387863397598, "learning_rate": 8.805106468333225e-05, "loss": 0.0162, "step": 79040 }, { "grad_norm": 0.14514687657356262, "learning_rate": 8.804748895099859e-05, "loss": 0.0204, "step": 79050 }, { "grad_norm": 0.2175983339548111, "learning_rate": 8.804391275635202e-05, "loss": 0.0206, "step": 79060 }, { "grad_norm": 0.172864630818367, "learning_rate": 8.8040336099436e-05, "loss": 0.0207, "step": 79070 }, { "grad_norm": 0.16503706574440002, "learning_rate": 8.803675898029397e-05, "loss": 0.0199, "step": 79080 }, { "grad_norm": 0.17072825133800507, "learning_rate": 8.803318139896942e-05, "loss": 0.0197, "step": 79090 }, { "grad_norm": 0.09540047496557236, "learning_rate": 8.802960335550578e-05, "loss": 0.026, "step": 79100 }, { "grad_norm": 0.11494240164756775, "learning_rate": 8.802602484994658e-05, "loss": 0.0341, "step": 79110 }, { "grad_norm": 0.14283181726932526, "learning_rate": 8.802244588233526e-05, "loss": 0.0202, "step": 79120 }, { "grad_norm": 0.17590069770812988, "learning_rate": 8.801886645271533e-05, "loss": 0.0167, "step": 79130 }, { "grad_norm": 0.1418672353029251, "learning_rate": 8.801528656113027e-05, "loss": 0.0171, "step": 79140 }, { "grad_norm": 0.1275354027748108, "learning_rate": 8.801170620762358e-05, "loss": 0.03, "step": 79150 }, { "grad_norm": 0.20523490011692047, "learning_rate": 8.800812539223879e-05, "loss": 0.0185, "step": 79160 }, { "grad_norm": 0.19333355128765106, "learning_rate": 8.800454411501938e-05, "loss": 0.0173, "step": 79170 }, { "grad_norm": 0.12196309864521027, "learning_rate": 8.800096237600888e-05, "loss": 0.0202, "step": 79180 }, { "grad_norm": 0.3565683364868164, "learning_rate": 8.799738017525081e-05, "loss": 0.0241, "step": 79190 }, { "grad_norm": 0.11819224059581757, "learning_rate": 8.799379751278869e-05, "loss": 0.017, "step": 79200 }, { "grad_norm": 0.1984013319015503, "learning_rate": 8.799021438866606e-05, "loss": 0.0233, "step": 79210 }, { "grad_norm": 0.22092875838279724, "learning_rate": 8.798663080292646e-05, "loss": 0.0149, "step": 79220 }, { "grad_norm": 0.09507443010807037, "learning_rate": 8.798304675561342e-05, "loss": 0.0255, "step": 79230 }, { "grad_norm": 0.13299371302127838, "learning_rate": 8.797946224677052e-05, "loss": 0.018, "step": 79240 }, { "grad_norm": 0.1259029656648636, "learning_rate": 8.797587727644128e-05, "loss": 0.0203, "step": 79250 }, { "grad_norm": 0.11379542946815491, "learning_rate": 8.797229184466927e-05, "loss": 0.0239, "step": 79260 }, { "grad_norm": 0.1290396749973297, "learning_rate": 8.796870595149806e-05, "loss": 0.0238, "step": 79270 }, { "grad_norm": 0.13821052014827728, "learning_rate": 8.796511959697124e-05, "loss": 0.0194, "step": 79280 }, { "grad_norm": 0.11870741099119186, "learning_rate": 8.796153278113237e-05, "loss": 0.0269, "step": 79290 }, { "grad_norm": 0.13653996586799622, "learning_rate": 8.795794550402503e-05, "loss": 0.0205, "step": 79300 }, { "grad_norm": 0.15993015468120575, "learning_rate": 8.795435776569282e-05, "loss": 0.0197, "step": 79310 }, { "grad_norm": 0.1222812607884407, "learning_rate": 8.795076956617932e-05, "loss": 0.0171, "step": 79320 }, { "grad_norm": 0.08800695091485977, "learning_rate": 8.794718090552815e-05, "loss": 0.0261, "step": 79330 }, { "grad_norm": 0.10401789098978043, "learning_rate": 8.794359178378289e-05, "loss": 0.0155, "step": 79340 }, { "grad_norm": 0.13681955635547638, "learning_rate": 8.794000220098717e-05, "loss": 0.0287, "step": 79350 }, { "grad_norm": 0.16060487926006317, "learning_rate": 8.79364121571846e-05, "loss": 0.016, "step": 79360 }, { "grad_norm": 0.10686279088258743, "learning_rate": 8.79328216524188e-05, "loss": 0.016, "step": 79370 }, { "grad_norm": 0.2413162738084793, "learning_rate": 8.792923068673342e-05, "loss": 0.0186, "step": 79380 }, { "grad_norm": 0.19116094708442688, "learning_rate": 8.792563926017207e-05, "loss": 0.0208, "step": 79390 }, { "grad_norm": 0.16564369201660156, "learning_rate": 8.792204737277838e-05, "loss": 0.0274, "step": 79400 }, { "grad_norm": 0.15628571808338165, "learning_rate": 8.791845502459604e-05, "loss": 0.0203, "step": 79410 }, { "grad_norm": 0.15349596738815308, "learning_rate": 8.791486221566865e-05, "loss": 0.0219, "step": 79420 }, { "grad_norm": 0.1630636304616928, "learning_rate": 8.791126894603987e-05, "loss": 0.0194, "step": 79430 }, { "grad_norm": 0.1958470195531845, "learning_rate": 8.790767521575341e-05, "loss": 0.0177, "step": 79440 }, { "grad_norm": 0.22595594823360443, "learning_rate": 8.790408102485287e-05, "loss": 0.026, "step": 79450 }, { "grad_norm": 0.11068154871463776, "learning_rate": 8.7900486373382e-05, "loss": 0.0183, "step": 79460 }, { "grad_norm": 0.17827251553535461, "learning_rate": 8.789689126138441e-05, "loss": 0.023, "step": 79470 }, { "grad_norm": 0.23697251081466675, "learning_rate": 8.789329568890382e-05, "loss": 0.0184, "step": 79480 }, { "grad_norm": 0.12607666850090027, "learning_rate": 8.78896996559839e-05, "loss": 0.012, "step": 79490 }, { "grad_norm": 0.07987885922193527, "learning_rate": 8.788610316266837e-05, "loss": 0.0176, "step": 79500 }, { "grad_norm": 0.14870810508728027, "learning_rate": 8.788250620900091e-05, "loss": 0.0148, "step": 79510 }, { "grad_norm": 0.1363552063703537, "learning_rate": 8.787890879502522e-05, "loss": 0.0146, "step": 79520 }, { "grad_norm": 0.20916485786437988, "learning_rate": 8.787531092078504e-05, "loss": 0.0184, "step": 79530 }, { "grad_norm": 0.1353681981563568, "learning_rate": 8.787171258632407e-05, "loss": 0.0164, "step": 79540 }, { "grad_norm": 0.21358519792556763, "learning_rate": 8.786811379168604e-05, "loss": 0.0187, "step": 79550 }, { "grad_norm": 0.11838192492723465, "learning_rate": 8.786451453691466e-05, "loss": 0.0173, "step": 79560 }, { "grad_norm": 0.10974660515785217, "learning_rate": 8.78609148220537e-05, "loss": 0.0153, "step": 79570 }, { "grad_norm": 0.134652242064476, "learning_rate": 8.785731464714686e-05, "loss": 0.0174, "step": 79580 }, { "grad_norm": 0.25465741753578186, "learning_rate": 8.78537140122379e-05, "loss": 0.0227, "step": 79590 }, { "grad_norm": 0.13387762010097504, "learning_rate": 8.785011291737058e-05, "loss": 0.0235, "step": 79600 }, { "grad_norm": 0.11057952791452408, "learning_rate": 8.784651136258865e-05, "loss": 0.0211, "step": 79610 }, { "grad_norm": 0.1496470421552658, "learning_rate": 8.784290934793589e-05, "loss": 0.0198, "step": 79620 }, { "grad_norm": 0.20921693742275238, "learning_rate": 8.783930687345603e-05, "loss": 0.0264, "step": 79630 }, { "grad_norm": 0.12847945094108582, "learning_rate": 8.783570393919288e-05, "loss": 0.0154, "step": 79640 }, { "grad_norm": 0.14291219413280487, "learning_rate": 8.78321005451902e-05, "loss": 0.0186, "step": 79650 }, { "grad_norm": 0.20421160757541656, "learning_rate": 8.782849669149178e-05, "loss": 0.0147, "step": 79660 }, { "grad_norm": 0.14673738181591034, "learning_rate": 8.78248923781414e-05, "loss": 0.0174, "step": 79670 }, { "grad_norm": 0.14742526412010193, "learning_rate": 8.782128760518287e-05, "loss": 0.0199, "step": 79680 }, { "grad_norm": 0.19534295797348022, "learning_rate": 8.781768237266e-05, "loss": 0.0215, "step": 79690 }, { "grad_norm": 0.16375300288200378, "learning_rate": 8.781407668061656e-05, "loss": 0.0199, "step": 79700 }, { "grad_norm": 0.19376377761363983, "learning_rate": 8.781047052909641e-05, "loss": 0.0227, "step": 79710 }, { "grad_norm": 0.10660707950592041, "learning_rate": 8.780686391814333e-05, "loss": 0.0252, "step": 79720 }, { "grad_norm": 0.11860376596450806, "learning_rate": 8.780325684780115e-05, "loss": 0.0227, "step": 79730 }, { "grad_norm": 0.2528459429740906, "learning_rate": 8.779964931811373e-05, "loss": 0.0251, "step": 79740 }, { "grad_norm": 0.10553175956010818, "learning_rate": 8.779604132912487e-05, "loss": 0.0199, "step": 79750 }, { "grad_norm": 0.1777782440185547, "learning_rate": 8.779243288087842e-05, "loss": 0.019, "step": 79760 }, { "grad_norm": 0.3218020498752594, "learning_rate": 8.778882397341823e-05, "loss": 0.0291, "step": 79770 }, { "grad_norm": 0.13072794675827026, "learning_rate": 8.778521460678815e-05, "loss": 0.0169, "step": 79780 }, { "grad_norm": 0.10100986808538437, "learning_rate": 8.778160478103204e-05, "loss": 0.0214, "step": 79790 }, { "grad_norm": 0.14209410548210144, "learning_rate": 8.777799449619375e-05, "loss": 0.0201, "step": 79800 }, { "grad_norm": 0.0748608186841011, "learning_rate": 8.777438375231717e-05, "loss": 0.0184, "step": 79810 }, { "grad_norm": 0.15306606888771057, "learning_rate": 8.777077254944616e-05, "loss": 0.0214, "step": 79820 }, { "grad_norm": 0.11593560129404068, "learning_rate": 8.77671608876246e-05, "loss": 0.0229, "step": 79830 }, { "grad_norm": 0.1369006335735321, "learning_rate": 8.776354876689638e-05, "loss": 0.0209, "step": 79840 }, { "grad_norm": 0.15279707312583923, "learning_rate": 8.775993618730538e-05, "loss": 0.0242, "step": 79850 }, { "grad_norm": 0.1362902969121933, "learning_rate": 8.775632314889552e-05, "loss": 0.0186, "step": 79860 }, { "grad_norm": 0.13184764981269836, "learning_rate": 8.775270965171067e-05, "loss": 0.0214, "step": 79870 }, { "grad_norm": 0.1819269061088562, "learning_rate": 8.774909569579476e-05, "loss": 0.0208, "step": 79880 }, { "grad_norm": 0.10267554968595505, "learning_rate": 8.774548128119169e-05, "loss": 0.0189, "step": 79890 }, { "grad_norm": 0.2209523618221283, "learning_rate": 8.77418664079454e-05, "loss": 0.0147, "step": 79900 }, { "grad_norm": 0.22104059159755707, "learning_rate": 8.773825107609979e-05, "loss": 0.0188, "step": 79910 }, { "grad_norm": 0.13552111387252808, "learning_rate": 8.773463528569879e-05, "loss": 0.0146, "step": 79920 }, { "grad_norm": 0.23075444996356964, "learning_rate": 8.773101903678635e-05, "loss": 0.0241, "step": 79930 }, { "grad_norm": 0.11851087212562561, "learning_rate": 8.77274023294064e-05, "loss": 0.0192, "step": 79940 }, { "grad_norm": 0.11711614578962326, "learning_rate": 8.77237851636029e-05, "loss": 0.0173, "step": 79950 }, { "grad_norm": 0.17933906614780426, "learning_rate": 8.772016753941979e-05, "loss": 0.0247, "step": 79960 }, { "grad_norm": 0.28945717215538025, "learning_rate": 8.771654945690104e-05, "loss": 0.0202, "step": 79970 }, { "grad_norm": 0.11622192710638046, "learning_rate": 8.771293091609058e-05, "loss": 0.0151, "step": 79980 }, { "grad_norm": 0.17026115953922272, "learning_rate": 8.770931191703243e-05, "loss": 0.0191, "step": 79990 }, { "grad_norm": 0.16347156465053558, "learning_rate": 8.770569245977051e-05, "loss": 0.0137, "step": 80000 }, { "grad_norm": 0.32337114214897156, "learning_rate": 8.770207254434886e-05, "loss": 0.022, "step": 80010 }, { "grad_norm": 0.14508908987045288, "learning_rate": 8.769845217081141e-05, "loss": 0.0139, "step": 80020 }, { "grad_norm": 0.17333251237869263, "learning_rate": 8.769483133920218e-05, "loss": 0.0164, "step": 80030 }, { "grad_norm": 0.13743312656879425, "learning_rate": 8.769121004956516e-05, "loss": 0.016, "step": 80040 }, { "grad_norm": 0.17007459700107574, "learning_rate": 8.768758830194434e-05, "loss": 0.0126, "step": 80050 }, { "grad_norm": 0.13561789691448212, "learning_rate": 8.768396609638374e-05, "loss": 0.0171, "step": 80060 }, { "grad_norm": 0.18723554909229279, "learning_rate": 8.768034343292737e-05, "loss": 0.0207, "step": 80070 }, { "grad_norm": 0.09243085235357285, "learning_rate": 8.767672031161926e-05, "loss": 0.0117, "step": 80080 }, { "grad_norm": 0.22806155681610107, "learning_rate": 8.767309673250344e-05, "loss": 0.0255, "step": 80090 }, { "grad_norm": 0.13570988178253174, "learning_rate": 8.76694726956239e-05, "loss": 0.0217, "step": 80100 }, { "grad_norm": 0.14752499759197235, "learning_rate": 8.76658482010247e-05, "loss": 0.0233, "step": 80110 }, { "grad_norm": 0.30323827266693115, "learning_rate": 8.766222324874989e-05, "loss": 0.03, "step": 80120 }, { "grad_norm": 0.17957338690757751, "learning_rate": 8.765859783884351e-05, "loss": 0.0234, "step": 80130 }, { "grad_norm": 0.19146054983139038, "learning_rate": 8.765497197134962e-05, "loss": 0.0175, "step": 80140 }, { "grad_norm": 0.09749428182840347, "learning_rate": 8.765134564631225e-05, "loss": 0.0207, "step": 80150 }, { "grad_norm": 0.15694911777973175, "learning_rate": 8.76477188637755e-05, "loss": 0.0162, "step": 80160 }, { "grad_norm": 0.2625349760055542, "learning_rate": 8.764409162378342e-05, "loss": 0.0211, "step": 80170 }, { "grad_norm": 0.16655628383159637, "learning_rate": 8.764046392638008e-05, "loss": 0.0282, "step": 80180 }, { "grad_norm": 0.09683870524168015, "learning_rate": 8.763683577160955e-05, "loss": 0.0196, "step": 80190 }, { "grad_norm": 0.21532830595970154, "learning_rate": 8.763320715951595e-05, "loss": 0.0176, "step": 80200 }, { "grad_norm": 0.14625072479248047, "learning_rate": 8.762957809014334e-05, "loss": 0.0159, "step": 80210 }, { "grad_norm": 0.12199926376342773, "learning_rate": 8.762594856353583e-05, "loss": 0.0231, "step": 80220 }, { "grad_norm": 0.09798053652048111, "learning_rate": 8.762231857973752e-05, "loss": 0.016, "step": 80230 }, { "grad_norm": 0.11409603804349899, "learning_rate": 8.761868813879253e-05, "loss": 0.0168, "step": 80240 }, { "grad_norm": 0.31218886375427246, "learning_rate": 8.761505724074495e-05, "loss": 0.022, "step": 80250 }, { "grad_norm": 0.14831462502479553, "learning_rate": 8.761142588563891e-05, "loss": 0.0174, "step": 80260 }, { "grad_norm": 0.19994701445102692, "learning_rate": 8.760779407351853e-05, "loss": 0.0175, "step": 80270 }, { "grad_norm": 0.16507172584533691, "learning_rate": 8.760416180442795e-05, "loss": 0.0194, "step": 80280 }, { "grad_norm": 0.14340370893478394, "learning_rate": 8.760052907841129e-05, "loss": 0.0194, "step": 80290 }, { "grad_norm": 0.10609007626771927, "learning_rate": 8.759689589551272e-05, "loss": 0.0167, "step": 80300 }, { "grad_norm": 0.23379695415496826, "learning_rate": 8.759326225577635e-05, "loss": 0.0201, "step": 80310 }, { "grad_norm": 0.12663239240646362, "learning_rate": 8.758962815924636e-05, "loss": 0.0189, "step": 80320 }, { "grad_norm": 0.2257198691368103, "learning_rate": 8.75859936059669e-05, "loss": 0.0179, "step": 80330 }, { "grad_norm": 0.10090809315443039, "learning_rate": 8.758235859598212e-05, "loss": 0.0194, "step": 80340 }, { "grad_norm": 0.1299905776977539, "learning_rate": 8.757872312933622e-05, "loss": 0.016, "step": 80350 }, { "grad_norm": 0.12386221438646317, "learning_rate": 8.757508720607332e-05, "loss": 0.0269, "step": 80360 }, { "grad_norm": 0.17354917526245117, "learning_rate": 8.757145082623766e-05, "loss": 0.018, "step": 80370 }, { "grad_norm": 0.14096124470233917, "learning_rate": 8.75678139898734e-05, "loss": 0.0219, "step": 80380 }, { "grad_norm": 0.22179225087165833, "learning_rate": 8.756417669702472e-05, "loss": 0.0142, "step": 80390 }, { "grad_norm": 0.1491132527589798, "learning_rate": 8.756053894773583e-05, "loss": 0.0172, "step": 80400 }, { "grad_norm": 0.41380536556243896, "learning_rate": 8.755690074205092e-05, "loss": 0.0276, "step": 80410 }, { "grad_norm": 0.1630486100912094, "learning_rate": 8.755326208001422e-05, "loss": 0.0196, "step": 80420 }, { "grad_norm": 0.178922638297081, "learning_rate": 8.754962296166992e-05, "loss": 0.0169, "step": 80430 }, { "grad_norm": 0.13169224560260773, "learning_rate": 8.754598338706224e-05, "loss": 0.0298, "step": 80440 }, { "grad_norm": 0.12610724568367004, "learning_rate": 8.754234335623544e-05, "loss": 0.0196, "step": 80450 }, { "grad_norm": 0.13529430329799652, "learning_rate": 8.75387028692337e-05, "loss": 0.0174, "step": 80460 }, { "grad_norm": 0.15068501234054565, "learning_rate": 8.753506192610129e-05, "loss": 0.0221, "step": 80470 }, { "grad_norm": 0.20455028116703033, "learning_rate": 8.753142052688243e-05, "loss": 0.0237, "step": 80480 }, { "grad_norm": 0.11738040298223495, "learning_rate": 8.752777867162137e-05, "loss": 0.0179, "step": 80490 }, { "grad_norm": 0.15080757439136505, "learning_rate": 8.752413636036237e-05, "loss": 0.0199, "step": 80500 }, { "grad_norm": 0.13861514627933502, "learning_rate": 8.752049359314969e-05, "loss": 0.0202, "step": 80510 }, { "grad_norm": 0.12271542847156525, "learning_rate": 8.751685037002759e-05, "loss": 0.0207, "step": 80520 }, { "grad_norm": 0.1814255565404892, "learning_rate": 8.751320669104034e-05, "loss": 0.0243, "step": 80530 }, { "grad_norm": 0.18618662655353546, "learning_rate": 8.75095625562322e-05, "loss": 0.017, "step": 80540 }, { "grad_norm": 0.21520893275737762, "learning_rate": 8.750591796564746e-05, "loss": 0.0217, "step": 80550 }, { "grad_norm": 0.14847876131534576, "learning_rate": 8.750227291933042e-05, "loss": 0.0177, "step": 80560 }, { "grad_norm": 0.343173623085022, "learning_rate": 8.749862741732532e-05, "loss": 0.0246, "step": 80570 }, { "grad_norm": 0.09965117275714874, "learning_rate": 8.749498145967654e-05, "loss": 0.0162, "step": 80580 }, { "grad_norm": 0.2000083476305008, "learning_rate": 8.74913350464283e-05, "loss": 0.0284, "step": 80590 }, { "grad_norm": 0.21220338344573975, "learning_rate": 8.748768817762496e-05, "loss": 0.0202, "step": 80600 }, { "grad_norm": 0.15131619572639465, "learning_rate": 8.748404085331081e-05, "loss": 0.0145, "step": 80610 }, { "grad_norm": 0.24133452773094177, "learning_rate": 8.748039307353017e-05, "loss": 0.0269, "step": 80620 }, { "grad_norm": 0.16884461045265198, "learning_rate": 8.747674483832737e-05, "loss": 0.0242, "step": 80630 }, { "grad_norm": 0.1882205754518509, "learning_rate": 8.747309614774673e-05, "loss": 0.0195, "step": 80640 }, { "grad_norm": 0.18006132543087006, "learning_rate": 8.746944700183259e-05, "loss": 0.0218, "step": 80650 }, { "grad_norm": 0.15769606828689575, "learning_rate": 8.74657974006293e-05, "loss": 0.025, "step": 80660 }, { "grad_norm": 0.11604741215705872, "learning_rate": 8.746214734418119e-05, "loss": 0.0207, "step": 80670 }, { "grad_norm": 0.14856122434139252, "learning_rate": 8.745849683253262e-05, "loss": 0.018, "step": 80680 }, { "grad_norm": 0.20900051295757294, "learning_rate": 8.745484586572796e-05, "loss": 0.0192, "step": 80690 }, { "grad_norm": 0.16536352038383484, "learning_rate": 8.745119444381154e-05, "loss": 0.026, "step": 80700 }, { "grad_norm": 0.14854486286640167, "learning_rate": 8.744754256682776e-05, "loss": 0.0247, "step": 80710 }, { "grad_norm": 0.2668551504611969, "learning_rate": 8.744389023482098e-05, "loss": 0.0201, "step": 80720 }, { "grad_norm": 0.09763775020837784, "learning_rate": 8.744023744783558e-05, "loss": 0.0137, "step": 80730 }, { "grad_norm": 0.23432183265686035, "learning_rate": 8.743658420591595e-05, "loss": 0.0184, "step": 80740 }, { "grad_norm": 0.21921809017658234, "learning_rate": 8.743293050910647e-05, "loss": 0.0188, "step": 80750 }, { "grad_norm": 0.1693814992904663, "learning_rate": 8.742927635745155e-05, "loss": 0.0138, "step": 80760 }, { "grad_norm": 0.09442543238401413, "learning_rate": 8.742562175099558e-05, "loss": 0.0166, "step": 80770 }, { "grad_norm": 0.16578996181488037, "learning_rate": 8.742196668978297e-05, "loss": 0.0156, "step": 80780 }, { "grad_norm": 0.1484367698431015, "learning_rate": 8.741831117385812e-05, "loss": 0.0325, "step": 80790 }, { "grad_norm": 0.13815292716026306, "learning_rate": 8.741465520326548e-05, "loss": 0.0171, "step": 80800 }, { "grad_norm": 0.2379537969827652, "learning_rate": 8.741099877804943e-05, "loss": 0.0224, "step": 80810 }, { "grad_norm": 0.09775146096944809, "learning_rate": 8.740734189825443e-05, "loss": 0.0205, "step": 80820 }, { "grad_norm": 0.10448029637336731, "learning_rate": 8.740368456392493e-05, "loss": 0.0175, "step": 80830 }, { "grad_norm": 0.23587666451931, "learning_rate": 8.740002677510532e-05, "loss": 0.0179, "step": 80840 }, { "grad_norm": 0.20436188578605652, "learning_rate": 8.739636853184008e-05, "loss": 0.0206, "step": 80850 }, { "grad_norm": 0.17873214185237885, "learning_rate": 8.739270983417366e-05, "loss": 0.0208, "step": 80860 }, { "grad_norm": 0.13046081364154816, "learning_rate": 8.738905068215052e-05, "loss": 0.017, "step": 80870 }, { "grad_norm": 0.1657794713973999, "learning_rate": 8.73853910758151e-05, "loss": 0.0177, "step": 80880 }, { "grad_norm": 0.17441442608833313, "learning_rate": 8.738173101521189e-05, "loss": 0.0217, "step": 80890 }, { "grad_norm": 0.16678424179553986, "learning_rate": 8.737807050038534e-05, "loss": 0.0185, "step": 80900 }, { "grad_norm": 0.19549813866615295, "learning_rate": 8.737440953137997e-05, "loss": 0.0187, "step": 80910 }, { "grad_norm": 0.41860681772232056, "learning_rate": 8.737074810824019e-05, "loss": 0.0191, "step": 80920 }, { "grad_norm": 0.18212562799453735, "learning_rate": 8.736708623101058e-05, "loss": 0.0205, "step": 80930 }, { "grad_norm": 0.27725696563720703, "learning_rate": 8.736342389973557e-05, "loss": 0.0198, "step": 80940 }, { "grad_norm": 0.1829034984111786, "learning_rate": 8.73597611144597e-05, "loss": 0.0154, "step": 80950 }, { "grad_norm": 0.10886410623788834, "learning_rate": 8.735609787522745e-05, "loss": 0.0182, "step": 80960 }, { "grad_norm": 0.20909753441810608, "learning_rate": 8.735243418208333e-05, "loss": 0.0201, "step": 80970 }, { "grad_norm": 0.10176228731870651, "learning_rate": 8.734877003507188e-05, "loss": 0.0197, "step": 80980 }, { "grad_norm": 0.12124767899513245, "learning_rate": 8.73451054342376e-05, "loss": 0.024, "step": 80990 }, { "grad_norm": 0.12080924212932587, "learning_rate": 8.734144037962504e-05, "loss": 0.02, "step": 81000 }, { "grad_norm": 0.15114250779151917, "learning_rate": 8.733777487127872e-05, "loss": 0.0237, "step": 81010 }, { "grad_norm": 0.10680188983678818, "learning_rate": 8.733410890924318e-05, "loss": 0.0196, "step": 81020 }, { "grad_norm": 0.39515089988708496, "learning_rate": 8.733044249356298e-05, "loss": 0.0237, "step": 81030 }, { "grad_norm": 0.21908321976661682, "learning_rate": 8.732677562428264e-05, "loss": 0.0223, "step": 81040 }, { "grad_norm": 0.3284992277622223, "learning_rate": 8.732310830144675e-05, "loss": 0.0272, "step": 81050 }, { "grad_norm": 0.12242887169122696, "learning_rate": 8.731944052509985e-05, "loss": 0.0204, "step": 81060 }, { "grad_norm": 0.14116130769252777, "learning_rate": 8.731577229528652e-05, "loss": 0.022, "step": 81070 }, { "grad_norm": 0.13572289049625397, "learning_rate": 8.73121036120513e-05, "loss": 0.0146, "step": 81080 }, { "grad_norm": 0.11619623005390167, "learning_rate": 8.730843447543882e-05, "loss": 0.0182, "step": 81090 }, { "grad_norm": 0.11266595870256424, "learning_rate": 8.730476488549362e-05, "loss": 0.0179, "step": 81100 }, { "grad_norm": 0.14739374816417694, "learning_rate": 8.73010948422603e-05, "loss": 0.0153, "step": 81110 }, { "grad_norm": 0.21366599202156067, "learning_rate": 8.729742434578348e-05, "loss": 0.0183, "step": 81120 }, { "grad_norm": 0.15182778239250183, "learning_rate": 8.729375339610773e-05, "loss": 0.0228, "step": 81130 }, { "grad_norm": 0.20157364010810852, "learning_rate": 8.729008199327766e-05, "loss": 0.0217, "step": 81140 }, { "grad_norm": 0.22238077223300934, "learning_rate": 8.728641013733789e-05, "loss": 0.019, "step": 81150 }, { "grad_norm": 0.11717714369297028, "learning_rate": 8.728273782833303e-05, "loss": 0.0175, "step": 81160 }, { "grad_norm": 0.13322091102600098, "learning_rate": 8.727906506630769e-05, "loss": 0.0171, "step": 81170 }, { "grad_norm": 0.17766395211219788, "learning_rate": 8.727539185130654e-05, "loss": 0.0176, "step": 81180 }, { "grad_norm": 0.1392512321472168, "learning_rate": 8.727171818337417e-05, "loss": 0.021, "step": 81190 }, { "grad_norm": 0.2023412436246872, "learning_rate": 8.726804406255525e-05, "loss": 0.0253, "step": 81200 }, { "grad_norm": 0.15031546354293823, "learning_rate": 8.726436948889438e-05, "loss": 0.0187, "step": 81210 }, { "grad_norm": 0.14349403977394104, "learning_rate": 8.726069446243625e-05, "loss": 0.0182, "step": 81220 }, { "grad_norm": 0.16665248572826385, "learning_rate": 8.725701898322551e-05, "loss": 0.0221, "step": 81230 }, { "grad_norm": 0.15434236824512482, "learning_rate": 8.725334305130682e-05, "loss": 0.0219, "step": 81240 }, { "grad_norm": 0.22456075251102448, "learning_rate": 8.724966666672481e-05, "loss": 0.0179, "step": 81250 }, { "grad_norm": 0.1307644098997116, "learning_rate": 8.724598982952421e-05, "loss": 0.0206, "step": 81260 }, { "grad_norm": 0.2603188157081604, "learning_rate": 8.724231253974964e-05, "loss": 0.0161, "step": 81270 }, { "grad_norm": 0.19982606172561646, "learning_rate": 8.723863479744583e-05, "loss": 0.0191, "step": 81280 }, { "grad_norm": 0.39452943205833435, "learning_rate": 8.723495660265745e-05, "loss": 0.0165, "step": 81290 }, { "grad_norm": 0.17014861106872559, "learning_rate": 8.723127795542919e-05, "loss": 0.0178, "step": 81300 }, { "grad_norm": 0.08468419313430786, "learning_rate": 8.722759885580573e-05, "loss": 0.0166, "step": 81310 }, { "grad_norm": 0.16091527044773102, "learning_rate": 8.722391930383182e-05, "loss": 0.021, "step": 81320 }, { "grad_norm": 0.25299811363220215, "learning_rate": 8.722023929955213e-05, "loss": 0.0235, "step": 81330 }, { "grad_norm": 0.11084416508674622, "learning_rate": 8.72165588430114e-05, "loss": 0.0252, "step": 81340 }, { "grad_norm": 0.18174540996551514, "learning_rate": 8.721287793425434e-05, "loss": 0.0271, "step": 81350 }, { "grad_norm": 0.25446227192878723, "learning_rate": 8.720919657332566e-05, "loss": 0.0246, "step": 81360 }, { "grad_norm": 0.13380277156829834, "learning_rate": 8.720551476027015e-05, "loss": 0.0169, "step": 81370 }, { "grad_norm": 0.21744213998317719, "learning_rate": 8.720183249513248e-05, "loss": 0.0174, "step": 81380 }, { "grad_norm": 0.1576084941625595, "learning_rate": 8.719814977795741e-05, "loss": 0.0188, "step": 81390 }, { "grad_norm": 0.12891645729541779, "learning_rate": 8.719446660878973e-05, "loss": 0.0179, "step": 81400 }, { "grad_norm": 0.1721636950969696, "learning_rate": 8.719078298767415e-05, "loss": 0.0241, "step": 81410 }, { "grad_norm": 0.14012788236141205, "learning_rate": 8.718709891465544e-05, "loss": 0.021, "step": 81420 }, { "grad_norm": 0.18927247822284698, "learning_rate": 8.718341438977837e-05, "loss": 0.0199, "step": 81430 }, { "grad_norm": 0.17266294360160828, "learning_rate": 8.71797294130877e-05, "loss": 0.0259, "step": 81440 }, { "grad_norm": 0.12078187614679337, "learning_rate": 8.717604398462822e-05, "loss": 0.0175, "step": 81450 }, { "grad_norm": 0.12232168763875961, "learning_rate": 8.717235810444471e-05, "loss": 0.017, "step": 81460 }, { "grad_norm": 0.08749318867921829, "learning_rate": 8.716867177258194e-05, "loss": 0.0177, "step": 81470 }, { "grad_norm": 0.10855832695960999, "learning_rate": 8.716498498908473e-05, "loss": 0.0193, "step": 81480 }, { "grad_norm": 0.15710684657096863, "learning_rate": 8.716129775399786e-05, "loss": 0.0204, "step": 81490 }, { "grad_norm": 0.11417689919471741, "learning_rate": 8.715761006736614e-05, "loss": 0.0166, "step": 81500 }, { "grad_norm": 0.24688801169395447, "learning_rate": 8.715392192923437e-05, "loss": 0.0176, "step": 81510 }, { "grad_norm": 0.1651269793510437, "learning_rate": 8.715023333964736e-05, "loss": 0.015, "step": 81520 }, { "grad_norm": 0.18361949920654297, "learning_rate": 8.714654429864994e-05, "loss": 0.0151, "step": 81530 }, { "grad_norm": 0.19857415556907654, "learning_rate": 8.714285480628694e-05, "loss": 0.0266, "step": 81540 }, { "grad_norm": 0.12673431634902954, "learning_rate": 8.713916486260319e-05, "loss": 0.0173, "step": 81550 }, { "grad_norm": 0.21366427838802338, "learning_rate": 8.713547446764352e-05, "loss": 0.0197, "step": 81560 }, { "grad_norm": 0.1273445188999176, "learning_rate": 8.713178362145278e-05, "loss": 0.0277, "step": 81570 }, { "grad_norm": 0.10623952001333237, "learning_rate": 8.712809232407579e-05, "loss": 0.019, "step": 81580 }, { "grad_norm": 0.18706770241260529, "learning_rate": 8.712440057555744e-05, "loss": 0.0164, "step": 81590 }, { "grad_norm": 0.15656349062919617, "learning_rate": 8.712070837594257e-05, "loss": 0.0233, "step": 81600 }, { "grad_norm": 0.13537850975990295, "learning_rate": 8.711701572527605e-05, "loss": 0.0191, "step": 81610 }, { "grad_norm": 0.12097591906785965, "learning_rate": 8.711332262360274e-05, "loss": 0.0175, "step": 81620 }, { "grad_norm": 0.21633782982826233, "learning_rate": 8.710962907096751e-05, "loss": 0.0187, "step": 81630 }, { "grad_norm": 0.13106070458889008, "learning_rate": 8.710593506741526e-05, "loss": 0.0252, "step": 81640 }, { "grad_norm": 0.16213028132915497, "learning_rate": 8.710224061299086e-05, "loss": 0.0143, "step": 81650 }, { "grad_norm": 0.1314125657081604, "learning_rate": 8.709854570773921e-05, "loss": 0.0205, "step": 81660 }, { "grad_norm": 0.196740061044693, "learning_rate": 8.70948503517052e-05, "loss": 0.0168, "step": 81670 }, { "grad_norm": 0.15121223032474518, "learning_rate": 8.709115454493375e-05, "loss": 0.0172, "step": 81680 }, { "grad_norm": 0.15599586069583893, "learning_rate": 8.708745828746973e-05, "loss": 0.0256, "step": 81690 }, { "grad_norm": 0.10007275640964508, "learning_rate": 8.708376157935808e-05, "loss": 0.0201, "step": 81700 }, { "grad_norm": 0.1183745265007019, "learning_rate": 8.708006442064372e-05, "loss": 0.0211, "step": 81710 }, { "grad_norm": 0.10775541514158249, "learning_rate": 8.707636681137156e-05, "loss": 0.0176, "step": 81720 }, { "grad_norm": 0.1476835459470749, "learning_rate": 8.707266875158655e-05, "loss": 0.0212, "step": 81730 }, { "grad_norm": 0.22472067177295685, "learning_rate": 8.70689702413336e-05, "loss": 0.0213, "step": 81740 }, { "grad_norm": 0.2089468091726303, "learning_rate": 8.706527128065767e-05, "loss": 0.0232, "step": 81750 }, { "grad_norm": 0.13237681984901428, "learning_rate": 8.706157186960371e-05, "loss": 0.0177, "step": 81760 }, { "grad_norm": 0.1084391251206398, "learning_rate": 8.705787200821664e-05, "loss": 0.0203, "step": 81770 }, { "grad_norm": 0.13316486775875092, "learning_rate": 8.705417169654145e-05, "loss": 0.0178, "step": 81780 }, { "grad_norm": 0.13485708832740784, "learning_rate": 8.705047093462307e-05, "loss": 0.0181, "step": 81790 }, { "grad_norm": 0.19243772327899933, "learning_rate": 8.704676972250651e-05, "loss": 0.0168, "step": 81800 }, { "grad_norm": 0.14383430778980255, "learning_rate": 8.70430680602367e-05, "loss": 0.0178, "step": 81810 }, { "grad_norm": 0.18405534327030182, "learning_rate": 8.703936594785865e-05, "loss": 0.0204, "step": 81820 }, { "grad_norm": 0.18311022222042084, "learning_rate": 8.703566338541733e-05, "loss": 0.0158, "step": 81830 }, { "grad_norm": 0.0846015065908432, "learning_rate": 8.703196037295773e-05, "loss": 0.0128, "step": 81840 }, { "grad_norm": 0.13743428885936737, "learning_rate": 8.702825691052486e-05, "loss": 0.0244, "step": 81850 }, { "grad_norm": 0.08369320631027222, "learning_rate": 8.70245529981637e-05, "loss": 0.0164, "step": 81860 }, { "grad_norm": 0.21162758767604828, "learning_rate": 8.702084863591926e-05, "loss": 0.0235, "step": 81870 }, { "grad_norm": 0.5259585380554199, "learning_rate": 8.701714382383655e-05, "loss": 0.0255, "step": 81880 }, { "grad_norm": 0.3497006297111511, "learning_rate": 8.701343856196061e-05, "loss": 0.0224, "step": 81890 }, { "grad_norm": 0.13370665907859802, "learning_rate": 8.700973285033643e-05, "loss": 0.0214, "step": 81900 }, { "grad_norm": 0.27912020683288574, "learning_rate": 8.700602668900906e-05, "loss": 0.0183, "step": 81910 }, { "grad_norm": 0.27159449458122253, "learning_rate": 8.700232007802351e-05, "loss": 0.0242, "step": 81920 }, { "grad_norm": 0.31299903988838196, "learning_rate": 8.699861301742486e-05, "loss": 0.0204, "step": 81930 }, { "grad_norm": 0.19203124940395355, "learning_rate": 8.699490550725811e-05, "loss": 0.0155, "step": 81940 }, { "grad_norm": 0.26858481764793396, "learning_rate": 8.699119754756835e-05, "loss": 0.0183, "step": 81950 }, { "grad_norm": 0.15350237488746643, "learning_rate": 8.69874891384006e-05, "loss": 0.0159, "step": 81960 }, { "grad_norm": 0.22038160264492035, "learning_rate": 8.698378027979994e-05, "loss": 0.0232, "step": 81970 }, { "grad_norm": 0.4659968912601471, "learning_rate": 8.698007097181142e-05, "loss": 0.0243, "step": 81980 }, { "grad_norm": 0.3272516429424286, "learning_rate": 8.697636121448014e-05, "loss": 0.0229, "step": 81990 }, { "grad_norm": 0.10058533400297165, "learning_rate": 8.697265100785115e-05, "loss": 0.0162, "step": 82000 }, { "grad_norm": 0.25317060947418213, "learning_rate": 8.696894035196953e-05, "loss": 0.0165, "step": 82010 }, { "grad_norm": 0.2526092231273651, "learning_rate": 8.696522924688041e-05, "loss": 0.0293, "step": 82020 }, { "grad_norm": 0.216456338763237, "learning_rate": 8.696151769262883e-05, "loss": 0.017, "step": 82030 }, { "grad_norm": 0.12163818627595901, "learning_rate": 8.695780568925992e-05, "loss": 0.0162, "step": 82040 }, { "grad_norm": 0.09902268648147583, "learning_rate": 8.695409323681877e-05, "loss": 0.0156, "step": 82050 }, { "grad_norm": 0.14601801335811615, "learning_rate": 8.695038033535052e-05, "loss": 0.0151, "step": 82060 }, { "grad_norm": 0.19758230447769165, "learning_rate": 8.694666698490024e-05, "loss": 0.021, "step": 82070 }, { "grad_norm": 0.22545163333415985, "learning_rate": 8.694295318551309e-05, "loss": 0.0214, "step": 82080 }, { "grad_norm": 0.3140813708305359, "learning_rate": 8.693923893723416e-05, "loss": 0.0207, "step": 82090 }, { "grad_norm": 0.1940816193819046, "learning_rate": 8.69355242401086e-05, "loss": 0.017, "step": 82100 }, { "grad_norm": 0.1159023717045784, "learning_rate": 8.693180909418156e-05, "loss": 0.0165, "step": 82110 }, { "grad_norm": 0.16275157034397125, "learning_rate": 8.692809349949816e-05, "loss": 0.0178, "step": 82120 }, { "grad_norm": 0.1100875586271286, "learning_rate": 8.692437745610357e-05, "loss": 0.0276, "step": 82130 }, { "grad_norm": 0.15483668446540833, "learning_rate": 8.692066096404293e-05, "loss": 0.0145, "step": 82140 }, { "grad_norm": 0.17381888628005981, "learning_rate": 8.691694402336139e-05, "loss": 0.0192, "step": 82150 }, { "grad_norm": 0.13953882455825806, "learning_rate": 8.691322663410413e-05, "loss": 0.0196, "step": 82160 }, { "grad_norm": 0.17176616191864014, "learning_rate": 8.690950879631632e-05, "loss": 0.0227, "step": 82170 }, { "grad_norm": 0.1348821371793747, "learning_rate": 8.690579051004312e-05, "loss": 0.0185, "step": 82180 }, { "grad_norm": 0.15928979218006134, "learning_rate": 8.690207177532971e-05, "loss": 0.0194, "step": 82190 }, { "grad_norm": 0.12716034054756165, "learning_rate": 8.689835259222132e-05, "loss": 0.019, "step": 82200 }, { "grad_norm": 0.13497544825077057, "learning_rate": 8.689463296076308e-05, "loss": 0.0143, "step": 82210 }, { "grad_norm": 0.1743541806936264, "learning_rate": 8.689091288100025e-05, "loss": 0.0211, "step": 82220 }, { "grad_norm": 0.17960499227046967, "learning_rate": 8.688719235297796e-05, "loss": 0.0172, "step": 82230 }, { "grad_norm": 0.44281816482543945, "learning_rate": 8.688347137674146e-05, "loss": 0.0218, "step": 82240 }, { "grad_norm": 0.17078739404678345, "learning_rate": 8.687974995233597e-05, "loss": 0.0167, "step": 82250 }, { "grad_norm": 0.21665632724761963, "learning_rate": 8.68760280798067e-05, "loss": 0.0177, "step": 82260 }, { "grad_norm": 0.19784556329250336, "learning_rate": 8.687230575919886e-05, "loss": 0.0304, "step": 82270 }, { "grad_norm": 0.1165923997759819, "learning_rate": 8.68685829905577e-05, "loss": 0.0207, "step": 82280 }, { "grad_norm": 0.15652261674404144, "learning_rate": 8.686485977392844e-05, "loss": 0.0278, "step": 82290 }, { "grad_norm": 0.1403314173221588, "learning_rate": 8.686113610935634e-05, "loss": 0.0201, "step": 82300 }, { "grad_norm": 0.13108119368553162, "learning_rate": 8.685741199688662e-05, "loss": 0.0156, "step": 82310 }, { "grad_norm": 0.21109123528003693, "learning_rate": 8.685368743656456e-05, "loss": 0.0197, "step": 82320 }, { "grad_norm": 0.1243269294500351, "learning_rate": 8.684996242843539e-05, "loss": 0.0154, "step": 82330 }, { "grad_norm": 0.6588522791862488, "learning_rate": 8.684623697254439e-05, "loss": 0.021, "step": 82340 }, { "grad_norm": 0.21213498711585999, "learning_rate": 8.684251106893681e-05, "loss": 0.0159, "step": 82350 }, { "grad_norm": 0.1445446014404297, "learning_rate": 8.683878471765796e-05, "loss": 0.0154, "step": 82360 }, { "grad_norm": 0.22844360768795013, "learning_rate": 8.683505791875308e-05, "loss": 0.018, "step": 82370 }, { "grad_norm": 0.14712680876255035, "learning_rate": 8.683133067226748e-05, "loss": 0.0224, "step": 82380 }, { "grad_norm": 0.1002931147813797, "learning_rate": 8.682760297824641e-05, "loss": 0.0142, "step": 82390 }, { "grad_norm": 0.17452673614025116, "learning_rate": 8.682387483673522e-05, "loss": 0.028, "step": 82400 }, { "grad_norm": 0.22429445385932922, "learning_rate": 8.682014624777917e-05, "loss": 0.0214, "step": 82410 }, { "grad_norm": 0.18247170746326447, "learning_rate": 8.681641721142358e-05, "loss": 0.0218, "step": 82420 }, { "grad_norm": 0.16794034838676453, "learning_rate": 8.681268772771376e-05, "loss": 0.0215, "step": 82430 }, { "grad_norm": 0.139957457780838, "learning_rate": 8.680895779669502e-05, "loss": 0.0198, "step": 82440 }, { "grad_norm": 0.1855982393026352, "learning_rate": 8.680522741841271e-05, "loss": 0.0179, "step": 82450 }, { "grad_norm": 0.2071235328912735, "learning_rate": 8.680149659291212e-05, "loss": 0.0215, "step": 82460 }, { "grad_norm": 0.2766380310058594, "learning_rate": 8.679776532023861e-05, "loss": 0.025, "step": 82470 }, { "grad_norm": 0.1784510463476181, "learning_rate": 8.679403360043749e-05, "loss": 0.0214, "step": 82480 }, { "grad_norm": 0.12156347185373306, "learning_rate": 8.679030143355416e-05, "loss": 0.0182, "step": 82490 }, { "grad_norm": 0.13015711307525635, "learning_rate": 8.67865688196339e-05, "loss": 0.0186, "step": 82500 }, { "grad_norm": 0.16132064163684845, "learning_rate": 8.67828357587221e-05, "loss": 0.0165, "step": 82510 }, { "grad_norm": 0.10745260864496231, "learning_rate": 8.677910225086412e-05, "loss": 0.0191, "step": 82520 }, { "grad_norm": 0.16915103793144226, "learning_rate": 8.677536829610533e-05, "loss": 0.0188, "step": 82530 }, { "grad_norm": 0.17153553664684296, "learning_rate": 8.677163389449108e-05, "loss": 0.0151, "step": 82540 }, { "grad_norm": 0.1603875309228897, "learning_rate": 8.676789904606676e-05, "loss": 0.0202, "step": 82550 }, { "grad_norm": 0.09739939123392105, "learning_rate": 8.676416375087776e-05, "loss": 0.0194, "step": 82560 }, { "grad_norm": 0.16613604128360748, "learning_rate": 8.676042800896947e-05, "loss": 0.0173, "step": 82570 }, { "grad_norm": 0.13691769540309906, "learning_rate": 8.675669182038725e-05, "loss": 0.0207, "step": 82580 }, { "grad_norm": 0.23648636043071747, "learning_rate": 8.675295518517655e-05, "loss": 0.0233, "step": 82590 }, { "grad_norm": 0.17224079370498657, "learning_rate": 8.674921810338272e-05, "loss": 0.0205, "step": 82600 }, { "grad_norm": 0.23361767828464508, "learning_rate": 8.674548057505119e-05, "loss": 0.0288, "step": 82610 }, { "grad_norm": 0.19463993608951569, "learning_rate": 8.674174260022738e-05, "loss": 0.022, "step": 82620 }, { "grad_norm": 0.11932065337896347, "learning_rate": 8.673800417895672e-05, "loss": 0.0232, "step": 82630 }, { "grad_norm": 0.23595274984836578, "learning_rate": 8.67342653112846e-05, "loss": 0.0267, "step": 82640 }, { "grad_norm": 0.15575741231441498, "learning_rate": 8.673052599725651e-05, "loss": 0.015, "step": 82650 }, { "grad_norm": 0.1927988976240158, "learning_rate": 8.672678623691783e-05, "loss": 0.0177, "step": 82660 }, { "grad_norm": 0.14608536660671234, "learning_rate": 8.672304603031402e-05, "loss": 0.023, "step": 82670 }, { "grad_norm": 0.11371046304702759, "learning_rate": 8.671930537749053e-05, "loss": 0.0191, "step": 82680 }, { "grad_norm": 0.11372414231300354, "learning_rate": 8.67155642784928e-05, "loss": 0.0147, "step": 82690 }, { "grad_norm": 0.13683182001113892, "learning_rate": 8.671182273336632e-05, "loss": 0.015, "step": 82700 }, { "grad_norm": 0.503807008266449, "learning_rate": 8.67080807421565e-05, "loss": 0.0149, "step": 82710 }, { "grad_norm": 0.11585737019777298, "learning_rate": 8.670433830490887e-05, "loss": 0.0188, "step": 82720 }, { "grad_norm": 0.3003867268562317, "learning_rate": 8.670059542166887e-05, "loss": 0.0194, "step": 82730 }, { "grad_norm": 0.2691928446292877, "learning_rate": 8.669685209248198e-05, "loss": 0.0262, "step": 82740 }, { "grad_norm": 0.14864157140254974, "learning_rate": 8.669310831739369e-05, "loss": 0.0165, "step": 82750 }, { "grad_norm": 0.15028922259807587, "learning_rate": 8.66893640964495e-05, "loss": 0.024, "step": 82760 }, { "grad_norm": 0.19945958256721497, "learning_rate": 8.668561942969488e-05, "loss": 0.0181, "step": 82770 }, { "grad_norm": 0.1205434799194336, "learning_rate": 8.668187431717535e-05, "loss": 0.0167, "step": 82780 }, { "grad_norm": 0.10313469916582108, "learning_rate": 8.667812875893641e-05, "loss": 0.0158, "step": 82790 }, { "grad_norm": 0.11651510000228882, "learning_rate": 8.66743827550236e-05, "loss": 0.0179, "step": 82800 }, { "grad_norm": 0.12367431074380875, "learning_rate": 8.66706363054824e-05, "loss": 0.0135, "step": 82810 }, { "grad_norm": 0.0856219008564949, "learning_rate": 8.666688941035832e-05, "loss": 0.0116, "step": 82820 }, { "grad_norm": 0.08769617229700089, "learning_rate": 8.666314206969694e-05, "loss": 0.0145, "step": 82830 }, { "grad_norm": 0.1540956199169159, "learning_rate": 8.665939428354377e-05, "loss": 0.0208, "step": 82840 }, { "grad_norm": 0.16014495491981506, "learning_rate": 8.665564605194434e-05, "loss": 0.0224, "step": 82850 }, { "grad_norm": 0.7842066884040833, "learning_rate": 8.66518973749442e-05, "loss": 0.0231, "step": 82860 }, { "grad_norm": 0.2683362066745758, "learning_rate": 8.664814825258892e-05, "loss": 0.0239, "step": 82870 }, { "grad_norm": 0.22216571867465973, "learning_rate": 8.664439868492402e-05, "loss": 0.0212, "step": 82880 }, { "grad_norm": 0.23730286955833435, "learning_rate": 8.664064867199508e-05, "loss": 0.0192, "step": 82890 }, { "grad_norm": 0.2301274985074997, "learning_rate": 8.663689821384766e-05, "loss": 0.0174, "step": 82900 }, { "grad_norm": 0.13779757916927338, "learning_rate": 8.663314731052734e-05, "loss": 0.0175, "step": 82910 }, { "grad_norm": 0.29542773962020874, "learning_rate": 8.662939596207968e-05, "loss": 0.0141, "step": 82920 }, { "grad_norm": 0.12826299667358398, "learning_rate": 8.66256441685503e-05, "loss": 0.0167, "step": 82930 }, { "grad_norm": 0.1418542116880417, "learning_rate": 8.662189192998474e-05, "loss": 0.0194, "step": 82940 }, { "grad_norm": 0.1469375640153885, "learning_rate": 8.661813924642863e-05, "loss": 0.0206, "step": 82950 }, { "grad_norm": 0.1986682265996933, "learning_rate": 8.661438611792756e-05, "loss": 0.0139, "step": 82960 }, { "grad_norm": 0.16859233379364014, "learning_rate": 8.661063254452712e-05, "loss": 0.0163, "step": 82970 }, { "grad_norm": 0.10305655747652054, "learning_rate": 8.660687852627292e-05, "loss": 0.0161, "step": 82980 }, { "grad_norm": 0.18922404944896698, "learning_rate": 8.660312406321059e-05, "loss": 0.0183, "step": 82990 }, { "grad_norm": 0.11408740282058716, "learning_rate": 8.659936915538576e-05, "loss": 0.0136, "step": 83000 }, { "grad_norm": 0.12840795516967773, "learning_rate": 8.659561380284402e-05, "loss": 0.0163, "step": 83010 }, { "grad_norm": 0.13791470229625702, "learning_rate": 8.659185800563102e-05, "loss": 0.0169, "step": 83020 }, { "grad_norm": 0.13417896628379822, "learning_rate": 8.658810176379241e-05, "loss": 0.019, "step": 83030 }, { "grad_norm": 0.09837450087070465, "learning_rate": 8.65843450773738e-05, "loss": 0.0172, "step": 83040 }, { "grad_norm": 0.13021036982536316, "learning_rate": 8.658058794642088e-05, "loss": 0.017, "step": 83050 }, { "grad_norm": 0.13195669651031494, "learning_rate": 8.657683037097926e-05, "loss": 0.0145, "step": 83060 }, { "grad_norm": 0.23582284152507782, "learning_rate": 8.657307235109462e-05, "loss": 0.019, "step": 83070 }, { "grad_norm": 0.12664778530597687, "learning_rate": 8.656931388681262e-05, "loss": 0.0148, "step": 83080 }, { "grad_norm": 0.1672115921974182, "learning_rate": 8.656555497817893e-05, "loss": 0.0171, "step": 83090 }, { "grad_norm": 0.1749221831560135, "learning_rate": 8.656179562523922e-05, "loss": 0.0167, "step": 83100 }, { "grad_norm": 0.17063701152801514, "learning_rate": 8.655803582803919e-05, "loss": 0.0223, "step": 83110 }, { "grad_norm": 0.12181560695171356, "learning_rate": 8.65542755866245e-05, "loss": 0.0233, "step": 83120 }, { "grad_norm": 0.14805956184864044, "learning_rate": 8.655051490104085e-05, "loss": 0.0181, "step": 83130 }, { "grad_norm": 0.12491903454065323, "learning_rate": 8.654675377133392e-05, "loss": 0.0194, "step": 83140 }, { "grad_norm": 0.12348896265029907, "learning_rate": 8.654299219754945e-05, "loss": 0.0216, "step": 83150 }, { "grad_norm": 0.19763463735580444, "learning_rate": 8.65392301797331e-05, "loss": 0.0185, "step": 83160 }, { "grad_norm": 0.2119596302509308, "learning_rate": 8.653546771793061e-05, "loss": 0.029, "step": 83170 }, { "grad_norm": 0.15630000829696655, "learning_rate": 8.653170481218769e-05, "loss": 0.0169, "step": 83180 }, { "grad_norm": 0.12063287198543549, "learning_rate": 8.652794146255008e-05, "loss": 0.0167, "step": 83190 }, { "grad_norm": 0.16346201300621033, "learning_rate": 8.652417766906348e-05, "loss": 0.0142, "step": 83200 }, { "grad_norm": 0.13207955658435822, "learning_rate": 8.652041343177364e-05, "loss": 0.0186, "step": 83210 }, { "grad_norm": 0.10427170991897583, "learning_rate": 8.651664875072629e-05, "loss": 0.0172, "step": 83220 }, { "grad_norm": 0.2093636840581894, "learning_rate": 8.651288362596718e-05, "loss": 0.0237, "step": 83230 }, { "grad_norm": 0.18491794168949127, "learning_rate": 8.650911805754208e-05, "loss": 0.0232, "step": 83240 }, { "grad_norm": 0.2291373610496521, "learning_rate": 8.65053520454967e-05, "loss": 0.0209, "step": 83250 }, { "grad_norm": 0.16053299605846405, "learning_rate": 8.650158558987685e-05, "loss": 0.0213, "step": 83260 }, { "grad_norm": 0.12720715999603271, "learning_rate": 8.649781869072826e-05, "loss": 0.0139, "step": 83270 }, { "grad_norm": 0.31772997975349426, "learning_rate": 8.64940513480967e-05, "loss": 0.0222, "step": 83280 }, { "grad_norm": 0.21079590916633606, "learning_rate": 8.649028356202798e-05, "loss": 0.0184, "step": 83290 }, { "grad_norm": 0.15127089619636536, "learning_rate": 8.648651533256785e-05, "loss": 0.0186, "step": 83300 }, { "grad_norm": 0.11978566646575928, "learning_rate": 8.648274665976211e-05, "loss": 0.017, "step": 83310 }, { "grad_norm": 0.13683801889419556, "learning_rate": 8.647897754365656e-05, "loss": 0.0227, "step": 83320 }, { "grad_norm": 0.15810683369636536, "learning_rate": 8.647520798429698e-05, "loss": 0.0208, "step": 83330 }, { "grad_norm": 0.13360141217708588, "learning_rate": 8.647143798172919e-05, "loss": 0.0206, "step": 83340 }, { "grad_norm": 0.2796359360218048, "learning_rate": 8.6467667535999e-05, "loss": 0.021, "step": 83350 }, { "grad_norm": 0.15817168354988098, "learning_rate": 8.646389664715223e-05, "loss": 0.0222, "step": 83360 }, { "grad_norm": 0.16381309926509857, "learning_rate": 8.646012531523467e-05, "loss": 0.0201, "step": 83370 }, { "grad_norm": 0.1640457659959793, "learning_rate": 8.645635354029216e-05, "loss": 0.0205, "step": 83380 }, { "grad_norm": 0.1426449567079544, "learning_rate": 8.645258132237054e-05, "loss": 0.0201, "step": 83390 }, { "grad_norm": 0.18939338624477386, "learning_rate": 8.644880866151564e-05, "loss": 0.018, "step": 83400 }, { "grad_norm": 0.11764074116945267, "learning_rate": 8.64450355577733e-05, "loss": 0.0218, "step": 83410 }, { "grad_norm": 0.11628838628530502, "learning_rate": 8.644126201118936e-05, "loss": 0.0164, "step": 83420 }, { "grad_norm": 0.07742521911859512, "learning_rate": 8.64374880218097e-05, "loss": 0.025, "step": 83430 }, { "grad_norm": 0.12888936698436737, "learning_rate": 8.643371358968015e-05, "loss": 0.0203, "step": 83440 }, { "grad_norm": 0.2651544511318207, "learning_rate": 8.642993871484657e-05, "loss": 0.0193, "step": 83450 }, { "grad_norm": 0.19688503444194794, "learning_rate": 8.642616339735485e-05, "loss": 0.0183, "step": 83460 }, { "grad_norm": 0.17621518671512604, "learning_rate": 8.642238763725085e-05, "loss": 0.0207, "step": 83470 }, { "grad_norm": 0.17249725759029388, "learning_rate": 8.641861143458045e-05, "loss": 0.0153, "step": 83480 }, { "grad_norm": 0.1674879491329193, "learning_rate": 8.641483478938954e-05, "loss": 0.0138, "step": 83490 }, { "grad_norm": 0.2470581829547882, "learning_rate": 8.6411057701724e-05, "loss": 0.0145, "step": 83500 }, { "grad_norm": 0.14875122904777527, "learning_rate": 8.640728017162974e-05, "loss": 0.021, "step": 83510 }, { "grad_norm": 0.1737530380487442, "learning_rate": 8.640350219915266e-05, "loss": 0.0188, "step": 83520 }, { "grad_norm": 0.11583848297595978, "learning_rate": 8.639972378433863e-05, "loss": 0.0174, "step": 83530 }, { "grad_norm": 0.10841894894838333, "learning_rate": 8.63959449272336e-05, "loss": 0.0157, "step": 83540 }, { "grad_norm": 0.10913580656051636, "learning_rate": 8.639216562788349e-05, "loss": 0.0208, "step": 83550 }, { "grad_norm": 0.179605171084404, "learning_rate": 8.638838588633419e-05, "loss": 0.0229, "step": 83560 }, { "grad_norm": 0.21845801174640656, "learning_rate": 8.638460570263165e-05, "loss": 0.0211, "step": 83570 }, { "grad_norm": 0.1590053141117096, "learning_rate": 8.63808250768218e-05, "loss": 0.019, "step": 83580 }, { "grad_norm": 0.1951061338186264, "learning_rate": 8.637704400895058e-05, "loss": 0.023, "step": 83590 }, { "grad_norm": 0.06304959207773209, "learning_rate": 8.637326249906392e-05, "loss": 0.018, "step": 83600 }, { "grad_norm": 0.15012064576148987, "learning_rate": 8.636948054720779e-05, "loss": 0.0211, "step": 83610 }, { "grad_norm": 0.19807466864585876, "learning_rate": 8.636569815342813e-05, "loss": 0.0222, "step": 83620 }, { "grad_norm": 0.35663825273513794, "learning_rate": 8.636191531777089e-05, "loss": 0.0194, "step": 83630 }, { "grad_norm": 0.15508891642093658, "learning_rate": 8.635813204028207e-05, "loss": 0.0286, "step": 83640 }, { "grad_norm": 0.13083086907863617, "learning_rate": 8.63543483210076e-05, "loss": 0.0158, "step": 83650 }, { "grad_norm": 0.24724304676055908, "learning_rate": 8.635056415999347e-05, "loss": 0.0208, "step": 83660 }, { "grad_norm": 0.1218617707490921, "learning_rate": 8.634677955728567e-05, "loss": 0.0213, "step": 83670 }, { "grad_norm": 0.1324426829814911, "learning_rate": 8.63429945129302e-05, "loss": 0.0188, "step": 83680 }, { "grad_norm": 0.14324049651622772, "learning_rate": 8.6339209026973e-05, "loss": 0.0184, "step": 83690 }, { "grad_norm": 0.16277366876602173, "learning_rate": 8.633542309946014e-05, "loss": 0.0199, "step": 83700 }, { "grad_norm": 0.3155321180820465, "learning_rate": 8.633163673043756e-05, "loss": 0.0175, "step": 83710 }, { "grad_norm": 0.16027534008026123, "learning_rate": 8.632784991995129e-05, "loss": 0.0223, "step": 83720 }, { "grad_norm": 0.13555467128753662, "learning_rate": 8.632406266804733e-05, "loss": 0.0151, "step": 83730 }, { "grad_norm": 0.15373900532722473, "learning_rate": 8.632027497477174e-05, "loss": 0.0153, "step": 83740 }, { "grad_norm": 0.3745821714401245, "learning_rate": 8.63164868401705e-05, "loss": 0.0201, "step": 83750 }, { "grad_norm": 0.13084687292575836, "learning_rate": 8.631269826428967e-05, "loss": 0.0172, "step": 83760 }, { "grad_norm": 0.14084461331367493, "learning_rate": 8.630890924717527e-05, "loss": 0.0222, "step": 83770 }, { "grad_norm": 0.1792386919260025, "learning_rate": 8.630511978887334e-05, "loss": 0.0178, "step": 83780 }, { "grad_norm": 0.16590140759944916, "learning_rate": 8.630132988942991e-05, "loss": 0.0224, "step": 83790 }, { "grad_norm": 0.09503622353076935, "learning_rate": 8.629753954889107e-05, "loss": 0.0229, "step": 83800 }, { "grad_norm": 0.11478069424629211, "learning_rate": 8.629374876730284e-05, "loss": 0.0159, "step": 83810 }, { "grad_norm": 0.20247109234333038, "learning_rate": 8.62899575447113e-05, "loss": 0.0191, "step": 83820 }, { "grad_norm": 0.23488153517246246, "learning_rate": 8.628616588116252e-05, "loss": 0.0195, "step": 83830 }, { "grad_norm": 0.10882233083248138, "learning_rate": 8.628237377670257e-05, "loss": 0.0177, "step": 83840 }, { "grad_norm": 0.08490093797445297, "learning_rate": 8.627858123137748e-05, "loss": 0.0157, "step": 83850 }, { "grad_norm": 0.16249775886535645, "learning_rate": 8.627478824523342e-05, "loss": 0.0198, "step": 83860 }, { "grad_norm": 0.16184790432453156, "learning_rate": 8.627099481831641e-05, "loss": 0.0195, "step": 83870 }, { "grad_norm": 0.20900577306747437, "learning_rate": 8.626720095067257e-05, "loss": 0.0112, "step": 83880 }, { "grad_norm": 0.1539406031370163, "learning_rate": 8.626340664234802e-05, "loss": 0.0223, "step": 83890 }, { "grad_norm": 0.2035655975341797, "learning_rate": 8.625961189338882e-05, "loss": 0.017, "step": 83900 }, { "grad_norm": 0.15576864778995514, "learning_rate": 8.62558167038411e-05, "loss": 0.0154, "step": 83910 }, { "grad_norm": 0.09917964786291122, "learning_rate": 8.6252021073751e-05, "loss": 0.0174, "step": 83920 }, { "grad_norm": 0.12418081611394882, "learning_rate": 8.624822500316458e-05, "loss": 0.0172, "step": 83930 }, { "grad_norm": 0.13367293775081635, "learning_rate": 8.624442849212802e-05, "loss": 0.0147, "step": 83940 }, { "grad_norm": 0.22904735803604126, "learning_rate": 8.624063154068744e-05, "loss": 0.0244, "step": 83950 }, { "grad_norm": 0.1693611592054367, "learning_rate": 8.623683414888895e-05, "loss": 0.0181, "step": 83960 }, { "grad_norm": 0.12688441574573517, "learning_rate": 8.623303631677873e-05, "loss": 0.0216, "step": 83970 }, { "grad_norm": 0.16442714631557465, "learning_rate": 8.62292380444029e-05, "loss": 0.0307, "step": 83980 }, { "grad_norm": 0.11879345774650574, "learning_rate": 8.622543933180762e-05, "loss": 0.0199, "step": 83990 }, { "grad_norm": 0.1854509562253952, "learning_rate": 8.622164017903907e-05, "loss": 0.0195, "step": 84000 }, { "grad_norm": 0.15860065817832947, "learning_rate": 8.621784058614337e-05, "loss": 0.0163, "step": 84010 }, { "grad_norm": 0.10735062509775162, "learning_rate": 8.621404055316673e-05, "loss": 0.0165, "step": 84020 }, { "grad_norm": 0.08875743299722672, "learning_rate": 8.621024008015528e-05, "loss": 0.015, "step": 84030 }, { "grad_norm": 0.19113264977931976, "learning_rate": 8.620643916715525e-05, "loss": 0.0176, "step": 84040 }, { "grad_norm": 0.18741118907928467, "learning_rate": 8.620263781421278e-05, "loss": 0.0191, "step": 84050 }, { "grad_norm": 0.14654722809791565, "learning_rate": 8.619883602137408e-05, "loss": 0.0157, "step": 84060 }, { "grad_norm": 0.10113100707530975, "learning_rate": 8.619503378868536e-05, "loss": 0.0179, "step": 84070 }, { "grad_norm": 0.12468304485082626, "learning_rate": 8.619123111619281e-05, "loss": 0.0145, "step": 84080 }, { "grad_norm": 0.18377551436424255, "learning_rate": 8.618742800394261e-05, "loss": 0.0276, "step": 84090 }, { "grad_norm": 0.34478163719177246, "learning_rate": 8.618362445198101e-05, "loss": 0.0224, "step": 84100 }, { "grad_norm": 0.14172333478927612, "learning_rate": 8.61798204603542e-05, "loss": 0.0257, "step": 84110 }, { "grad_norm": 0.584160566329956, "learning_rate": 8.617601602910842e-05, "loss": 0.0236, "step": 84120 }, { "grad_norm": 0.17230482399463654, "learning_rate": 8.617221115828988e-05, "loss": 0.0177, "step": 84130 }, { "grad_norm": 0.15785478055477142, "learning_rate": 8.616840584794483e-05, "loss": 0.0181, "step": 84140 }, { "grad_norm": 0.2800338864326477, "learning_rate": 8.61646000981195e-05, "loss": 0.0159, "step": 84150 }, { "grad_norm": 0.08631586283445358, "learning_rate": 8.616079390886014e-05, "loss": 0.0164, "step": 84160 }, { "grad_norm": 0.23142889142036438, "learning_rate": 8.615698728021298e-05, "loss": 0.0225, "step": 84170 }, { "grad_norm": 0.1457299292087555, "learning_rate": 8.61531802122243e-05, "loss": 0.0147, "step": 84180 }, { "grad_norm": 0.22883345186710358, "learning_rate": 8.614937270494032e-05, "loss": 0.0198, "step": 84190 }, { "grad_norm": 0.16919773817062378, "learning_rate": 8.614556475840735e-05, "loss": 0.0193, "step": 84200 }, { "grad_norm": 0.3732273280620575, "learning_rate": 8.614175637267165e-05, "loss": 0.0239, "step": 84210 }, { "grad_norm": 0.1367814838886261, "learning_rate": 8.613794754777946e-05, "loss": 0.0194, "step": 84220 }, { "grad_norm": 0.4253174662590027, "learning_rate": 8.613413828377711e-05, "loss": 0.0186, "step": 84230 }, { "grad_norm": 0.20078974962234497, "learning_rate": 8.613032858071086e-05, "loss": 0.0267, "step": 84240 }, { "grad_norm": 0.1256297081708908, "learning_rate": 8.6126518438627e-05, "loss": 0.0195, "step": 84250 }, { "grad_norm": 0.22614185512065887, "learning_rate": 8.612270785757183e-05, "loss": 0.0211, "step": 84260 }, { "grad_norm": 0.2436436116695404, "learning_rate": 8.611889683759166e-05, "loss": 0.0219, "step": 84270 }, { "grad_norm": 0.12949572503566742, "learning_rate": 8.61150853787328e-05, "loss": 0.0167, "step": 84280 }, { "grad_norm": 0.2731499969959259, "learning_rate": 8.611127348104153e-05, "loss": 0.0245, "step": 84290 }, { "grad_norm": 0.15182970464229584, "learning_rate": 8.61074611445642e-05, "loss": 0.0218, "step": 84300 }, { "grad_norm": 0.2184743881225586, "learning_rate": 8.610364836934714e-05, "loss": 0.0233, "step": 84310 }, { "grad_norm": 0.20008140802383423, "learning_rate": 8.609983515543667e-05, "loss": 0.0183, "step": 84320 }, { "grad_norm": 0.20326656103134155, "learning_rate": 8.60960215028791e-05, "loss": 0.0209, "step": 84330 }, { "grad_norm": 0.26566073298454285, "learning_rate": 8.609220741172079e-05, "loss": 0.0247, "step": 84340 }, { "grad_norm": 0.1434270143508911, "learning_rate": 8.60883928820081e-05, "loss": 0.0201, "step": 84350 }, { "grad_norm": 0.07257530838251114, "learning_rate": 8.608457791378734e-05, "loss": 0.0171, "step": 84360 }, { "grad_norm": 0.1668376326560974, "learning_rate": 8.60807625071049e-05, "loss": 0.0219, "step": 84370 }, { "grad_norm": 0.13986538350582123, "learning_rate": 8.607694666200715e-05, "loss": 0.0151, "step": 84380 }, { "grad_norm": 0.13956737518310547, "learning_rate": 8.60731303785404e-05, "loss": 0.0147, "step": 84390 }, { "grad_norm": 0.17333193123340607, "learning_rate": 8.606931365675108e-05, "loss": 0.0174, "step": 84400 }, { "grad_norm": 0.1406414955854416, "learning_rate": 8.606549649668553e-05, "loss": 0.0181, "step": 84410 }, { "grad_norm": 0.21178318560123444, "learning_rate": 8.606167889839016e-05, "loss": 0.0161, "step": 84420 }, { "grad_norm": 0.12435048818588257, "learning_rate": 8.605786086191132e-05, "loss": 0.014, "step": 84430 }, { "grad_norm": 0.139927476644516, "learning_rate": 8.605404238729544e-05, "loss": 0.0183, "step": 84440 }, { "grad_norm": 0.1458626091480255, "learning_rate": 8.605022347458891e-05, "loss": 0.022, "step": 84450 }, { "grad_norm": 0.14539241790771484, "learning_rate": 8.604640412383814e-05, "loss": 0.0164, "step": 84460 }, { "grad_norm": 0.16118544340133667, "learning_rate": 8.60425843350895e-05, "loss": 0.0162, "step": 84470 }, { "grad_norm": 0.1720018833875656, "learning_rate": 8.603876410838944e-05, "loss": 0.0205, "step": 84480 }, { "grad_norm": 0.19315862655639648, "learning_rate": 8.603494344378438e-05, "loss": 0.017, "step": 84490 }, { "grad_norm": 0.18511056900024414, "learning_rate": 8.603112234132073e-05, "loss": 0.0253, "step": 84500 }, { "grad_norm": 0.21867221593856812, "learning_rate": 8.60273008010449e-05, "loss": 0.0248, "step": 84510 }, { "grad_norm": 0.12186747044324875, "learning_rate": 8.60234788230034e-05, "loss": 0.0164, "step": 84520 }, { "grad_norm": 0.11442321538925171, "learning_rate": 8.601965640724257e-05, "loss": 0.0201, "step": 84530 }, { "grad_norm": 0.4834730327129364, "learning_rate": 8.601583355380894e-05, "loss": 0.0188, "step": 84540 }, { "grad_norm": 0.25598132610321045, "learning_rate": 8.60120102627489e-05, "loss": 0.026, "step": 84550 }, { "grad_norm": 0.12611812353134155, "learning_rate": 8.600818653410895e-05, "loss": 0.0211, "step": 84560 }, { "grad_norm": 0.2156783640384674, "learning_rate": 8.600436236793554e-05, "loss": 0.0211, "step": 84570 }, { "grad_norm": 0.1707216501235962, "learning_rate": 8.600053776427513e-05, "loss": 0.0145, "step": 84580 }, { "grad_norm": 0.1879657655954361, "learning_rate": 8.59967127231742e-05, "loss": 0.016, "step": 84590 }, { "grad_norm": 0.12252150475978851, "learning_rate": 8.599288724467921e-05, "loss": 0.0178, "step": 84600 }, { "grad_norm": 0.13689488172531128, "learning_rate": 8.598906132883667e-05, "loss": 0.0159, "step": 84610 }, { "grad_norm": 0.11037079244852066, "learning_rate": 8.598523497569305e-05, "loss": 0.015, "step": 84620 }, { "grad_norm": 0.1691151261329651, "learning_rate": 8.598140818529485e-05, "loss": 0.0172, "step": 84630 }, { "grad_norm": 0.12656714022159576, "learning_rate": 8.597758095768858e-05, "loss": 0.0198, "step": 84640 }, { "grad_norm": 0.1757490038871765, "learning_rate": 8.597375329292073e-05, "loss": 0.016, "step": 84650 }, { "grad_norm": 0.17726989090442657, "learning_rate": 8.59699251910378e-05, "loss": 0.018, "step": 84660 }, { "grad_norm": 0.16062197089195251, "learning_rate": 8.596609665208631e-05, "loss": 0.0184, "step": 84670 }, { "grad_norm": 0.14137978851795197, "learning_rate": 8.59622676761128e-05, "loss": 0.0144, "step": 84680 }, { "grad_norm": 0.22554920613765717, "learning_rate": 8.595843826316379e-05, "loss": 0.0177, "step": 84690 }, { "grad_norm": 0.1940527707338333, "learning_rate": 8.59546084132858e-05, "loss": 0.0217, "step": 84700 }, { "grad_norm": 0.1707262247800827, "learning_rate": 8.595077812652536e-05, "loss": 0.0166, "step": 84710 }, { "grad_norm": 0.9624587297439575, "learning_rate": 8.594694740292905e-05, "loss": 0.0179, "step": 84720 }, { "grad_norm": 0.7116419076919556, "learning_rate": 8.594311624254337e-05, "loss": 0.0245, "step": 84730 }, { "grad_norm": 0.12921583652496338, "learning_rate": 8.59392846454149e-05, "loss": 0.0217, "step": 84740 }, { "grad_norm": 0.20292022824287415, "learning_rate": 8.593545261159017e-05, "loss": 0.0161, "step": 84750 }, { "grad_norm": 0.15740960836410522, "learning_rate": 8.593162014111577e-05, "loss": 0.0234, "step": 84760 }, { "grad_norm": 0.20605343580245972, "learning_rate": 8.592778723403827e-05, "loss": 0.0206, "step": 84770 }, { "grad_norm": 0.3438887894153595, "learning_rate": 8.592395389040424e-05, "loss": 0.0205, "step": 84780 }, { "grad_norm": 0.13925650715827942, "learning_rate": 8.592012011026023e-05, "loss": 0.0182, "step": 84790 }, { "grad_norm": 0.2511052191257477, "learning_rate": 8.591628589365286e-05, "loss": 0.0257, "step": 84800 }, { "grad_norm": 0.3139905035495758, "learning_rate": 8.59124512406287e-05, "loss": 0.0199, "step": 84810 }, { "grad_norm": 0.23666012287139893, "learning_rate": 8.590861615123435e-05, "loss": 0.0182, "step": 84820 }, { "grad_norm": 0.20700131356716156, "learning_rate": 8.590478062551643e-05, "loss": 0.0223, "step": 84830 }, { "grad_norm": 0.21008582413196564, "learning_rate": 8.59009446635215e-05, "loss": 0.0138, "step": 84840 }, { "grad_norm": 0.14136308431625366, "learning_rate": 8.589710826529622e-05, "loss": 0.0212, "step": 84850 }, { "grad_norm": 0.20430971682071686, "learning_rate": 8.589327143088715e-05, "loss": 0.0209, "step": 84860 }, { "grad_norm": 0.17750394344329834, "learning_rate": 8.588943416034097e-05, "loss": 0.0219, "step": 84870 }, { "grad_norm": 0.42059382796287537, "learning_rate": 8.588559645370428e-05, "loss": 0.0169, "step": 84880 }, { "grad_norm": 0.08431315422058105, "learning_rate": 8.58817583110237e-05, "loss": 0.0214, "step": 84890 }, { "grad_norm": 0.22353392839431763, "learning_rate": 8.587791973234588e-05, "loss": 0.0177, "step": 84900 }, { "grad_norm": 0.15800926089286804, "learning_rate": 8.587408071771745e-05, "loss": 0.0208, "step": 84910 }, { "grad_norm": 0.17860527336597443, "learning_rate": 8.58702412671851e-05, "loss": 0.0208, "step": 84920 }, { "grad_norm": 0.15129686892032623, "learning_rate": 8.586640138079543e-05, "loss": 0.0204, "step": 84930 }, { "grad_norm": 0.16337889432907104, "learning_rate": 8.586256105859512e-05, "loss": 0.0186, "step": 84940 }, { "grad_norm": 0.1474427431821823, "learning_rate": 8.585872030063082e-05, "loss": 0.0146, "step": 84950 }, { "grad_norm": 0.1878012716770172, "learning_rate": 8.585487910694925e-05, "loss": 0.016, "step": 84960 }, { "grad_norm": 0.2529858648777008, "learning_rate": 8.585103747759702e-05, "loss": 0.0191, "step": 84970 }, { "grad_norm": 0.1433602273464203, "learning_rate": 8.584719541262082e-05, "loss": 0.0153, "step": 84980 }, { "grad_norm": 0.11393795162439346, "learning_rate": 8.584335291206739e-05, "loss": 0.0189, "step": 84990 }, { "grad_norm": 0.14287856221199036, "learning_rate": 8.583950997598336e-05, "loss": 0.0224, "step": 85000 }, { "grad_norm": 0.1006377711892128, "learning_rate": 8.583566660441545e-05, "loss": 0.0144, "step": 85010 }, { "grad_norm": 0.11585632711648941, "learning_rate": 8.583182279741035e-05, "loss": 0.0197, "step": 85020 }, { "grad_norm": 0.10394825041294098, "learning_rate": 8.582797855501476e-05, "loss": 0.0242, "step": 85030 }, { "grad_norm": 0.14548945426940918, "learning_rate": 8.582413387727542e-05, "loss": 0.0202, "step": 85040 }, { "grad_norm": 0.16027851402759552, "learning_rate": 8.582028876423904e-05, "loss": 0.025, "step": 85050 }, { "grad_norm": 0.2931574285030365, "learning_rate": 8.581644321595231e-05, "loss": 0.028, "step": 85060 }, { "grad_norm": 0.13202834129333496, "learning_rate": 8.5812597232462e-05, "loss": 0.0176, "step": 85070 }, { "grad_norm": 0.15208964049816132, "learning_rate": 8.580875081381481e-05, "loss": 0.0225, "step": 85080 }, { "grad_norm": 0.15112091600894928, "learning_rate": 8.580490396005746e-05, "loss": 0.0266, "step": 85090 }, { "grad_norm": 0.1691749393939972, "learning_rate": 8.580105667123677e-05, "loss": 0.0237, "step": 85100 }, { "grad_norm": 0.1906755417585373, "learning_rate": 8.579720894739943e-05, "loss": 0.0209, "step": 85110 }, { "grad_norm": 0.15182258188724518, "learning_rate": 8.579336078859217e-05, "loss": 0.0221, "step": 85120 }, { "grad_norm": 0.2561766803264618, "learning_rate": 8.578951219486182e-05, "loss": 0.0192, "step": 85130 }, { "grad_norm": 0.16788920760154724, "learning_rate": 8.578566316625507e-05, "loss": 0.0168, "step": 85140 }, { "grad_norm": 0.10922590643167496, "learning_rate": 8.578181370281875e-05, "loss": 0.014, "step": 85150 }, { "grad_norm": 0.10766538232564926, "learning_rate": 8.577796380459961e-05, "loss": 0.0196, "step": 85160 }, { "grad_norm": 0.2714080214500427, "learning_rate": 8.577411347164442e-05, "loss": 0.0265, "step": 85170 }, { "grad_norm": 0.47738221287727356, "learning_rate": 8.577026270399997e-05, "loss": 0.0241, "step": 85180 }, { "grad_norm": 0.2209782600402832, "learning_rate": 8.576641150171307e-05, "loss": 0.0208, "step": 85190 }, { "grad_norm": 0.17051473259925842, "learning_rate": 8.57625598648305e-05, "loss": 0.0202, "step": 85200 }, { "grad_norm": 0.3758584260940552, "learning_rate": 8.575870779339905e-05, "loss": 0.0161, "step": 85210 }, { "grad_norm": 0.28570255637168884, "learning_rate": 8.575485528746556e-05, "loss": 0.022, "step": 85220 }, { "grad_norm": 0.1773059368133545, "learning_rate": 8.57510023470768e-05, "loss": 0.0196, "step": 85230 }, { "grad_norm": 0.1490158885717392, "learning_rate": 8.574714897227963e-05, "loss": 0.0199, "step": 85240 }, { "grad_norm": 0.1472170650959015, "learning_rate": 8.574329516312081e-05, "loss": 0.0217, "step": 85250 }, { "grad_norm": 0.23719123005867004, "learning_rate": 8.573944091964723e-05, "loss": 0.0205, "step": 85260 }, { "grad_norm": 0.18721963465213776, "learning_rate": 8.573558624190569e-05, "loss": 0.0248, "step": 85270 }, { "grad_norm": 0.4171038568019867, "learning_rate": 8.573173112994306e-05, "loss": 0.0208, "step": 85280 }, { "grad_norm": 0.14486895501613617, "learning_rate": 8.572787558380614e-05, "loss": 0.0183, "step": 85290 }, { "grad_norm": 0.11706584692001343, "learning_rate": 8.572401960354182e-05, "loss": 0.0209, "step": 85300 }, { "grad_norm": 0.5834584832191467, "learning_rate": 8.572016318919691e-05, "loss": 0.019, "step": 85310 }, { "grad_norm": 0.12419681996107101, "learning_rate": 8.57163063408183e-05, "loss": 0.0206, "step": 85320 }, { "grad_norm": 0.09245019406080246, "learning_rate": 8.571244905845284e-05, "loss": 0.0168, "step": 85330 }, { "grad_norm": 0.14121893048286438, "learning_rate": 8.570859134214741e-05, "loss": 0.0162, "step": 85340 }, { "grad_norm": 0.150288388133049, "learning_rate": 8.570473319194889e-05, "loss": 0.0159, "step": 85350 }, { "grad_norm": 0.1473989635705948, "learning_rate": 8.570087460790415e-05, "loss": 0.019, "step": 85360 }, { "grad_norm": 0.10294623672962189, "learning_rate": 8.569701559006008e-05, "loss": 0.0198, "step": 85370 }, { "grad_norm": 0.19074393808841705, "learning_rate": 8.569315613846354e-05, "loss": 0.0196, "step": 85380 }, { "grad_norm": 0.2472086250782013, "learning_rate": 8.568929625316146e-05, "loss": 0.0214, "step": 85390 }, { "grad_norm": 0.1330564022064209, "learning_rate": 8.568543593420076e-05, "loss": 0.0185, "step": 85400 }, { "grad_norm": 0.11868121474981308, "learning_rate": 8.568157518162829e-05, "loss": 0.0138, "step": 85410 }, { "grad_norm": 0.16580674052238464, "learning_rate": 8.5677713995491e-05, "loss": 0.0224, "step": 85420 }, { "grad_norm": 0.15252059698104858, "learning_rate": 8.567385237583581e-05, "loss": 0.028, "step": 85430 }, { "grad_norm": 0.16106894612312317, "learning_rate": 8.566999032270964e-05, "loss": 0.0158, "step": 85440 }, { "grad_norm": 0.18290485441684723, "learning_rate": 8.56661278361594e-05, "loss": 0.0189, "step": 85450 }, { "grad_norm": 0.19267317652702332, "learning_rate": 8.566226491623203e-05, "loss": 0.0194, "step": 85460 }, { "grad_norm": 0.1344577968120575, "learning_rate": 8.565840156297447e-05, "loss": 0.0174, "step": 85470 }, { "grad_norm": 0.15379378199577332, "learning_rate": 8.565453777643366e-05, "loss": 0.0237, "step": 85480 }, { "grad_norm": 0.12197834998369217, "learning_rate": 8.565067355665656e-05, "loss": 0.0162, "step": 85490 }, { "grad_norm": 0.1416729837656021, "learning_rate": 8.564680890369011e-05, "loss": 0.0189, "step": 85500 }, { "grad_norm": 0.1286294013261795, "learning_rate": 8.564294381758128e-05, "loss": 0.0205, "step": 85510 }, { "grad_norm": 0.08142464607954025, "learning_rate": 8.563907829837703e-05, "loss": 0.0184, "step": 85520 }, { "grad_norm": 0.18027649819850922, "learning_rate": 8.563521234612433e-05, "loss": 0.0176, "step": 85530 }, { "grad_norm": 0.1095258817076683, "learning_rate": 8.563134596087015e-05, "loss": 0.0225, "step": 85540 }, { "grad_norm": 0.11856258660554886, "learning_rate": 8.562747914266149e-05, "loss": 0.0192, "step": 85550 }, { "grad_norm": 0.3653431832790375, "learning_rate": 8.56236118915453e-05, "loss": 0.0186, "step": 85560 }, { "grad_norm": 0.20780028402805328, "learning_rate": 8.56197442075686e-05, "loss": 0.0173, "step": 85570 }, { "grad_norm": 0.1184854805469513, "learning_rate": 8.561587609077837e-05, "loss": 0.0191, "step": 85580 }, { "grad_norm": 0.11279736459255219, "learning_rate": 8.561200754122163e-05, "loss": 0.0141, "step": 85590 }, { "grad_norm": 0.13907329738140106, "learning_rate": 8.560813855894539e-05, "loss": 0.0229, "step": 85600 }, { "grad_norm": 0.09275732934474945, "learning_rate": 8.560426914399662e-05, "loss": 0.014, "step": 85610 }, { "grad_norm": 0.13428722321987152, "learning_rate": 8.560039929642236e-05, "loss": 0.0186, "step": 85620 }, { "grad_norm": 0.12575238943099976, "learning_rate": 8.559652901626966e-05, "loss": 0.0185, "step": 85630 }, { "grad_norm": 0.1564721316099167, "learning_rate": 8.559265830358552e-05, "loss": 0.0192, "step": 85640 }, { "grad_norm": 0.16217537224292755, "learning_rate": 8.558878715841696e-05, "loss": 0.0172, "step": 85650 }, { "grad_norm": 0.14063793420791626, "learning_rate": 8.558491558081104e-05, "loss": 0.0203, "step": 85660 }, { "grad_norm": 0.18987438082695007, "learning_rate": 8.55810435708148e-05, "loss": 0.0152, "step": 85670 }, { "grad_norm": 0.11467534303665161, "learning_rate": 8.557717112847529e-05, "loss": 0.0165, "step": 85680 }, { "grad_norm": 0.19200503826141357, "learning_rate": 8.557329825383956e-05, "loss": 0.0209, "step": 85690 }, { "grad_norm": 0.1551825851202011, "learning_rate": 8.556942494695467e-05, "loss": 0.0185, "step": 85700 }, { "grad_norm": 0.11811771988868713, "learning_rate": 8.556555120786769e-05, "loss": 0.0161, "step": 85710 }, { "grad_norm": 0.185810387134552, "learning_rate": 8.556167703662566e-05, "loss": 0.0161, "step": 85720 }, { "grad_norm": 0.20239587128162384, "learning_rate": 8.555780243327569e-05, "loss": 0.0171, "step": 85730 }, { "grad_norm": 0.1474744826555252, "learning_rate": 8.555392739786485e-05, "loss": 0.0217, "step": 85740 }, { "grad_norm": 0.16659782826900482, "learning_rate": 8.555005193044023e-05, "loss": 0.0218, "step": 85750 }, { "grad_norm": 0.19895483553409576, "learning_rate": 8.55461760310489e-05, "loss": 0.0162, "step": 85760 }, { "grad_norm": 0.2776299715042114, "learning_rate": 8.5542299699738e-05, "loss": 0.0272, "step": 85770 }, { "grad_norm": 0.10462170094251633, "learning_rate": 8.553842293655459e-05, "loss": 0.018, "step": 85780 }, { "grad_norm": 0.14189711213111877, "learning_rate": 8.553454574154577e-05, "loss": 0.018, "step": 85790 }, { "grad_norm": 0.140191450715065, "learning_rate": 8.553066811475867e-05, "loss": 0.0244, "step": 85800 }, { "grad_norm": 0.21461516618728638, "learning_rate": 8.552679005624041e-05, "loss": 0.0175, "step": 85810 }, { "grad_norm": 0.14680713415145874, "learning_rate": 8.55229115660381e-05, "loss": 0.0215, "step": 85820 }, { "grad_norm": 0.13004542887210846, "learning_rate": 8.55190326441989e-05, "loss": 0.022, "step": 85830 }, { "grad_norm": 0.22822469472885132, "learning_rate": 8.551515329076991e-05, "loss": 0.0188, "step": 85840 }, { "grad_norm": 0.14168845117092133, "learning_rate": 8.551127350579827e-05, "loss": 0.0192, "step": 85850 }, { "grad_norm": 0.18986175954341888, "learning_rate": 8.550739328933112e-05, "loss": 0.0162, "step": 85860 }, { "grad_norm": 0.18275386095046997, "learning_rate": 8.550351264141563e-05, "loss": 0.0216, "step": 85870 }, { "grad_norm": 0.1014118492603302, "learning_rate": 8.549963156209894e-05, "loss": 0.0155, "step": 85880 }, { "grad_norm": 0.18962004780769348, "learning_rate": 8.54957500514282e-05, "loss": 0.0211, "step": 85890 }, { "grad_norm": 0.1503937840461731, "learning_rate": 8.549186810945059e-05, "loss": 0.0149, "step": 85900 }, { "grad_norm": 0.21005305647850037, "learning_rate": 8.548798573621327e-05, "loss": 0.023, "step": 85910 }, { "grad_norm": 0.21149133145809174, "learning_rate": 8.548410293176341e-05, "loss": 0.0174, "step": 85920 }, { "grad_norm": 0.1188218891620636, "learning_rate": 8.548021969614821e-05, "loss": 0.0213, "step": 85930 }, { "grad_norm": 0.14857439696788788, "learning_rate": 8.547633602941484e-05, "loss": 0.0177, "step": 85940 }, { "grad_norm": 0.15005263686180115, "learning_rate": 8.547245193161049e-05, "loss": 0.0205, "step": 85950 }, { "grad_norm": 0.2398514598608017, "learning_rate": 8.546856740278235e-05, "loss": 0.0223, "step": 85960 }, { "grad_norm": 0.1530696004629135, "learning_rate": 8.546468244297765e-05, "loss": 0.0193, "step": 85970 }, { "grad_norm": 0.20353637635707855, "learning_rate": 8.546079705224354e-05, "loss": 0.0165, "step": 85980 }, { "grad_norm": 0.1436242014169693, "learning_rate": 8.54569112306273e-05, "loss": 0.0153, "step": 85990 }, { "grad_norm": 0.14490966498851776, "learning_rate": 8.545302497817609e-05, "loss": 0.0127, "step": 86000 }, { "grad_norm": 0.27436500787734985, "learning_rate": 8.544913829493718e-05, "loss": 0.02, "step": 86010 }, { "grad_norm": 0.07889626920223236, "learning_rate": 8.544525118095774e-05, "loss": 0.019, "step": 86020 }, { "grad_norm": 0.24176274240016937, "learning_rate": 8.544136363628503e-05, "loss": 0.0164, "step": 86030 }, { "grad_norm": 0.12219464033842087, "learning_rate": 8.54374756609663e-05, "loss": 0.0138, "step": 86040 }, { "grad_norm": 0.15436160564422607, "learning_rate": 8.54335872550488e-05, "loss": 0.023, "step": 86050 }, { "grad_norm": 0.20623086392879486, "learning_rate": 8.542969841857972e-05, "loss": 0.0208, "step": 86060 }, { "grad_norm": 0.1060752272605896, "learning_rate": 8.542580915160638e-05, "loss": 0.0157, "step": 86070 }, { "grad_norm": 0.16448809206485748, "learning_rate": 8.542191945417601e-05, "loss": 0.0143, "step": 86080 }, { "grad_norm": 0.17562928795814514, "learning_rate": 8.541802932633588e-05, "loss": 0.0144, "step": 86090 }, { "grad_norm": 0.151505708694458, "learning_rate": 8.541413876813325e-05, "loss": 0.0182, "step": 86100 }, { "grad_norm": 0.14705978333950043, "learning_rate": 8.54102477796154e-05, "loss": 0.0197, "step": 86110 }, { "grad_norm": 0.09199996292591095, "learning_rate": 8.540635636082959e-05, "loss": 0.0199, "step": 86120 }, { "grad_norm": 0.2302589863538742, "learning_rate": 8.540246451182314e-05, "loss": 0.0193, "step": 86130 }, { "grad_norm": 0.12910762429237366, "learning_rate": 8.539857223264331e-05, "loss": 0.0244, "step": 86140 }, { "grad_norm": 0.12767985463142395, "learning_rate": 8.53946795233374e-05, "loss": 0.018, "step": 86150 }, { "grad_norm": 0.19136624038219452, "learning_rate": 8.539078638395273e-05, "loss": 0.0173, "step": 86160 }, { "grad_norm": 0.32220080494880676, "learning_rate": 8.538689281453657e-05, "loss": 0.028, "step": 86170 }, { "grad_norm": 0.1304560750722885, "learning_rate": 8.538299881513628e-05, "loss": 0.0178, "step": 86180 }, { "grad_norm": 0.11798734962940216, "learning_rate": 8.537910438579913e-05, "loss": 0.0215, "step": 86190 }, { "grad_norm": 0.2745780646800995, "learning_rate": 8.537520952657246e-05, "loss": 0.0272, "step": 86200 }, { "grad_norm": 0.13110560178756714, "learning_rate": 8.537131423750359e-05, "loss": 0.0193, "step": 86210 }, { "grad_norm": 0.23589108884334564, "learning_rate": 8.536741851863988e-05, "loss": 0.0213, "step": 86220 }, { "grad_norm": 0.08768255263566971, "learning_rate": 8.536352237002861e-05, "loss": 0.0136, "step": 86230 }, { "grad_norm": 0.12273043394088745, "learning_rate": 8.535962579171717e-05, "loss": 0.0244, "step": 86240 }, { "grad_norm": 0.25923439860343933, "learning_rate": 8.535572878375288e-05, "loss": 0.0229, "step": 86250 }, { "grad_norm": 0.17612536251544952, "learning_rate": 8.535183134618311e-05, "loss": 0.0201, "step": 86260 }, { "grad_norm": 0.16580095887184143, "learning_rate": 8.534793347905522e-05, "loss": 0.0165, "step": 86270 }, { "grad_norm": 0.23487825691699982, "learning_rate": 8.534403518241657e-05, "loss": 0.0195, "step": 86280 }, { "grad_norm": 0.23181571066379547, "learning_rate": 8.534013645631452e-05, "loss": 0.0281, "step": 86290 }, { "grad_norm": 0.1346694827079773, "learning_rate": 8.533623730079643e-05, "loss": 0.016, "step": 86300 }, { "grad_norm": 0.09621018916368484, "learning_rate": 8.533233771590971e-05, "loss": 0.0192, "step": 86310 }, { "grad_norm": 0.18912915885448456, "learning_rate": 8.532843770170171e-05, "loss": 0.0176, "step": 86320 }, { "grad_norm": 0.17969880998134613, "learning_rate": 8.532453725821984e-05, "loss": 0.0164, "step": 86330 }, { "grad_norm": 0.1445922553539276, "learning_rate": 8.532063638551151e-05, "loss": 0.0223, "step": 86340 }, { "grad_norm": 0.14550776779651642, "learning_rate": 8.531673508362408e-05, "loss": 0.0203, "step": 86350 }, { "grad_norm": 0.1729935258626938, "learning_rate": 8.531283335260498e-05, "loss": 0.0187, "step": 86360 }, { "grad_norm": 0.15464414656162262, "learning_rate": 8.530893119250161e-05, "loss": 0.0191, "step": 86370 }, { "grad_norm": 0.09565561264753342, "learning_rate": 8.53050286033614e-05, "loss": 0.0148, "step": 86380 }, { "grad_norm": 0.14360889792442322, "learning_rate": 8.530112558523175e-05, "loss": 0.0213, "step": 86390 }, { "grad_norm": 0.1324932873249054, "learning_rate": 8.529722213816009e-05, "loss": 0.0154, "step": 86400 }, { "grad_norm": 0.17087964713573456, "learning_rate": 8.529331826219387e-05, "loss": 0.023, "step": 86410 }, { "grad_norm": 0.07977771759033203, "learning_rate": 8.52894139573805e-05, "loss": 0.0163, "step": 86420 }, { "grad_norm": 0.08757725358009338, "learning_rate": 8.528550922376744e-05, "loss": 0.0169, "step": 86430 }, { "grad_norm": 0.1631375402212143, "learning_rate": 8.528160406140211e-05, "loss": 0.0218, "step": 86440 }, { "grad_norm": 0.3439091145992279, "learning_rate": 8.5277698470332e-05, "loss": 0.0246, "step": 86450 }, { "grad_norm": 0.22502782940864563, "learning_rate": 8.527379245060453e-05, "loss": 0.0201, "step": 86460 }, { "grad_norm": 0.1907387524843216, "learning_rate": 8.52698860022672e-05, "loss": 0.0192, "step": 86470 }, { "grad_norm": 0.2705223262310028, "learning_rate": 8.526597912536743e-05, "loss": 0.0213, "step": 86480 }, { "grad_norm": 0.22149129211902618, "learning_rate": 8.526207181995272e-05, "loss": 0.0319, "step": 86490 }, { "grad_norm": 1.0283572673797607, "learning_rate": 8.525816408607055e-05, "loss": 0.0192, "step": 86500 }, { "grad_norm": 0.1442197859287262, "learning_rate": 8.52542559237684e-05, "loss": 0.0209, "step": 86510 }, { "grad_norm": 0.15927357971668243, "learning_rate": 8.525034733309376e-05, "loss": 0.0156, "step": 86520 }, { "grad_norm": 0.37299206852912903, "learning_rate": 8.524643831409412e-05, "loss": 0.0222, "step": 86530 }, { "grad_norm": 0.11455301195383072, "learning_rate": 8.524252886681697e-05, "loss": 0.0168, "step": 86540 }, { "grad_norm": 0.09789784252643585, "learning_rate": 8.52386189913098e-05, "loss": 0.0169, "step": 86550 }, { "grad_norm": 0.23013095557689667, "learning_rate": 8.523470868762016e-05, "loss": 0.0224, "step": 86560 }, { "grad_norm": 0.17920607328414917, "learning_rate": 8.523079795579554e-05, "loss": 0.0191, "step": 86570 }, { "grad_norm": 0.1169576421380043, "learning_rate": 8.522688679588346e-05, "loss": 0.0154, "step": 86580 }, { "grad_norm": 0.15860117971897125, "learning_rate": 8.522297520793144e-05, "loss": 0.0202, "step": 86590 }, { "grad_norm": 0.12012401968240738, "learning_rate": 8.521906319198703e-05, "loss": 0.019, "step": 86600 }, { "grad_norm": 0.1759967803955078, "learning_rate": 8.521515074809772e-05, "loss": 0.021, "step": 86610 }, { "grad_norm": 0.10252005606889725, "learning_rate": 8.521123787631112e-05, "loss": 0.0159, "step": 86620 }, { "grad_norm": 0.1391231268644333, "learning_rate": 8.520732457667471e-05, "loss": 0.0189, "step": 86630 }, { "grad_norm": 0.219136044383049, "learning_rate": 8.520341084923607e-05, "loss": 0.0222, "step": 86640 }, { "grad_norm": 0.14039330184459686, "learning_rate": 8.519949669404274e-05, "loss": 0.0233, "step": 86650 }, { "grad_norm": 0.19743123650550842, "learning_rate": 8.519558211114232e-05, "loss": 0.018, "step": 86660 }, { "grad_norm": 0.10561313480138779, "learning_rate": 8.519166710058231e-05, "loss": 0.0232, "step": 86670 }, { "grad_norm": 0.14802010357379913, "learning_rate": 8.518775166241034e-05, "loss": 0.0209, "step": 86680 }, { "grad_norm": 0.308513343334198, "learning_rate": 8.518383579667396e-05, "loss": 0.0209, "step": 86690 }, { "grad_norm": 0.20373687148094177, "learning_rate": 8.517991950342077e-05, "loss": 0.0188, "step": 86700 }, { "grad_norm": 0.48889273405075073, "learning_rate": 8.517600278269833e-05, "loss": 0.0168, "step": 86710 }, { "grad_norm": 0.15614809095859528, "learning_rate": 8.517208563455423e-05, "loss": 0.0198, "step": 86720 }, { "grad_norm": 0.12984798848628998, "learning_rate": 8.516816805903611e-05, "loss": 0.016, "step": 86730 }, { "grad_norm": 0.17288394272327423, "learning_rate": 8.516425005619153e-05, "loss": 0.0218, "step": 86740 }, { "grad_norm": 0.1545385867357254, "learning_rate": 8.516033162606811e-05, "loss": 0.0187, "step": 86750 }, { "grad_norm": 0.1869743913412094, "learning_rate": 8.515641276871346e-05, "loss": 0.0158, "step": 86760 }, { "grad_norm": 0.12384099513292313, "learning_rate": 8.515249348417522e-05, "loss": 0.017, "step": 86770 }, { "grad_norm": 0.24403543770313263, "learning_rate": 8.514857377250098e-05, "loss": 0.0194, "step": 86780 }, { "grad_norm": 0.21866779029369354, "learning_rate": 8.514465363373839e-05, "loss": 0.0191, "step": 86790 }, { "grad_norm": 0.18036366999149323, "learning_rate": 8.514073306793506e-05, "loss": 0.0172, "step": 86800 }, { "grad_norm": 0.2265336960554123, "learning_rate": 8.513681207513867e-05, "loss": 0.0247, "step": 86810 }, { "grad_norm": 0.1636040210723877, "learning_rate": 8.513289065539683e-05, "loss": 0.0194, "step": 86820 }, { "grad_norm": 0.44784656167030334, "learning_rate": 8.512896880875718e-05, "loss": 0.0216, "step": 86830 }, { "grad_norm": 0.09421535581350327, "learning_rate": 8.51250465352674e-05, "loss": 0.0207, "step": 86840 }, { "grad_norm": 0.20626653730869293, "learning_rate": 8.512112383497515e-05, "loss": 0.0179, "step": 86850 }, { "grad_norm": 0.11811219900846481, "learning_rate": 8.511720070792808e-05, "loss": 0.0152, "step": 86860 }, { "grad_norm": 0.1234465017914772, "learning_rate": 8.511327715417388e-05, "loss": 0.0177, "step": 86870 }, { "grad_norm": 0.164466992020607, "learning_rate": 8.510935317376019e-05, "loss": 0.0187, "step": 86880 }, { "grad_norm": 0.24075108766555786, "learning_rate": 8.51054287667347e-05, "loss": 0.016, "step": 86890 }, { "grad_norm": 0.16341155767440796, "learning_rate": 8.510150393314513e-05, "loss": 0.0221, "step": 86900 }, { "grad_norm": 0.41767171025276184, "learning_rate": 8.509757867303914e-05, "loss": 0.0148, "step": 86910 }, { "grad_norm": 0.2676306962966919, "learning_rate": 8.509365298646444e-05, "loss": 0.0165, "step": 86920 }, { "grad_norm": 0.14072591066360474, "learning_rate": 8.50897268734687e-05, "loss": 0.0199, "step": 86930 }, { "grad_norm": 0.2010946422815323, "learning_rate": 8.508580033409967e-05, "loss": 0.0215, "step": 86940 }, { "grad_norm": 0.6737737655639648, "learning_rate": 8.508187336840502e-05, "loss": 0.0163, "step": 86950 }, { "grad_norm": 0.1970141977071762, "learning_rate": 8.50779459764325e-05, "loss": 0.0175, "step": 86960 }, { "grad_norm": 0.32510998845100403, "learning_rate": 8.50740181582298e-05, "loss": 0.0258, "step": 86970 }, { "grad_norm": 0.180264413356781, "learning_rate": 8.507008991384469e-05, "loss": 0.02, "step": 86980 }, { "grad_norm": 0.13181929290294647, "learning_rate": 8.506616124332485e-05, "loss": 0.0171, "step": 86990 }, { "grad_norm": 0.11215022206306458, "learning_rate": 8.506223214671807e-05, "loss": 0.0204, "step": 87000 }, { "grad_norm": 0.1267370730638504, "learning_rate": 8.505830262407204e-05, "loss": 0.0229, "step": 87010 }, { "grad_norm": 0.1474151611328125, "learning_rate": 8.505437267543454e-05, "loss": 0.0172, "step": 87020 }, { "grad_norm": 0.13144727051258087, "learning_rate": 8.505044230085333e-05, "loss": 0.018, "step": 87030 }, { "grad_norm": 0.27035245299339294, "learning_rate": 8.504651150037614e-05, "loss": 0.013, "step": 87040 }, { "grad_norm": 0.21722614765167236, "learning_rate": 8.504258027405074e-05, "loss": 0.0211, "step": 87050 }, { "grad_norm": 0.30585619807243347, "learning_rate": 8.503864862192491e-05, "loss": 0.0223, "step": 87060 }, { "grad_norm": 0.16095489263534546, "learning_rate": 8.503471654404643e-05, "loss": 0.0223, "step": 87070 }, { "grad_norm": 0.11812064051628113, "learning_rate": 8.503078404046305e-05, "loss": 0.0168, "step": 87080 }, { "grad_norm": 0.2261762171983719, "learning_rate": 8.502685111122258e-05, "loss": 0.0149, "step": 87090 }, { "grad_norm": 0.1995089203119278, "learning_rate": 8.502291775637278e-05, "loss": 0.0191, "step": 87100 }, { "grad_norm": 0.16264867782592773, "learning_rate": 8.501898397596148e-05, "loss": 0.0192, "step": 87110 }, { "grad_norm": 0.14592866599559784, "learning_rate": 8.501504977003645e-05, "loss": 0.022, "step": 87120 }, { "grad_norm": 0.2370399683713913, "learning_rate": 8.501111513864551e-05, "loss": 0.0173, "step": 87130 }, { "grad_norm": 0.24912221729755402, "learning_rate": 8.500718008183648e-05, "loss": 0.021, "step": 87140 }, { "grad_norm": 0.165052592754364, "learning_rate": 8.500324459965714e-05, "loss": 0.0196, "step": 87150 }, { "grad_norm": 0.19585053622722626, "learning_rate": 8.499930869215534e-05, "loss": 0.0186, "step": 87160 }, { "grad_norm": 0.1780564785003662, "learning_rate": 8.499537235937888e-05, "loss": 0.0181, "step": 87170 }, { "grad_norm": 0.22417952120304108, "learning_rate": 8.499143560137561e-05, "loss": 0.0181, "step": 87180 }, { "grad_norm": 0.16197161376476288, "learning_rate": 8.498749841819337e-05, "loss": 0.0187, "step": 87190 }, { "grad_norm": 0.18943274021148682, "learning_rate": 8.498356080987999e-05, "loss": 0.024, "step": 87200 }, { "grad_norm": 0.12711471319198608, "learning_rate": 8.49796227764833e-05, "loss": 0.0202, "step": 87210 }, { "grad_norm": 0.19112063944339752, "learning_rate": 8.497568431805119e-05, "loss": 0.0221, "step": 87220 }, { "grad_norm": 0.25847843289375305, "learning_rate": 8.497174543463147e-05, "loss": 0.019, "step": 87230 }, { "grad_norm": 0.23920105397701263, "learning_rate": 8.496780612627205e-05, "loss": 0.0282, "step": 87240 }, { "grad_norm": 0.4337892234325409, "learning_rate": 8.496386639302076e-05, "loss": 0.0262, "step": 87250 }, { "grad_norm": 0.1191248893737793, "learning_rate": 8.495992623492547e-05, "loss": 0.019, "step": 87260 }, { "grad_norm": 0.33472540974617004, "learning_rate": 8.495598565203407e-05, "loss": 0.017, "step": 87270 }, { "grad_norm": 0.1907440423965454, "learning_rate": 8.495204464439446e-05, "loss": 0.016, "step": 87280 }, { "grad_norm": 0.1970759928226471, "learning_rate": 8.494810321205449e-05, "loss": 0.0199, "step": 87290 }, { "grad_norm": 0.14743219316005707, "learning_rate": 8.494416135506208e-05, "loss": 0.016, "step": 87300 }, { "grad_norm": 0.1327456384897232, "learning_rate": 8.49402190734651e-05, "loss": 0.021, "step": 87310 }, { "grad_norm": 0.11265194416046143, "learning_rate": 8.493627636731147e-05, "loss": 0.018, "step": 87320 }, { "grad_norm": 0.14247067272663116, "learning_rate": 8.493233323664911e-05, "loss": 0.0223, "step": 87330 }, { "grad_norm": 0.15089218318462372, "learning_rate": 8.492838968152592e-05, "loss": 0.0269, "step": 87340 }, { "grad_norm": 0.1011275202035904, "learning_rate": 8.492444570198982e-05, "loss": 0.0229, "step": 87350 }, { "grad_norm": 0.20688538253307343, "learning_rate": 8.492050129808871e-05, "loss": 0.0229, "step": 87360 }, { "grad_norm": 0.16782240569591522, "learning_rate": 8.491655646987056e-05, "loss": 0.0195, "step": 87370 }, { "grad_norm": 0.12068171054124832, "learning_rate": 8.491261121738328e-05, "loss": 0.0208, "step": 87380 }, { "grad_norm": 0.13813355565071106, "learning_rate": 8.49086655406748e-05, "loss": 0.0156, "step": 87390 }, { "grad_norm": 0.13529512286186218, "learning_rate": 8.490471943979308e-05, "loss": 0.02, "step": 87400 }, { "grad_norm": 0.22666385769844055, "learning_rate": 8.490077291478606e-05, "loss": 0.0172, "step": 87410 }, { "grad_norm": 0.2976256310939789, "learning_rate": 8.489682596570171e-05, "loss": 0.0139, "step": 87420 }, { "grad_norm": 0.09219065308570862, "learning_rate": 8.489287859258796e-05, "loss": 0.0131, "step": 87430 }, { "grad_norm": 0.22932328283786774, "learning_rate": 8.488893079549281e-05, "loss": 0.0173, "step": 87440 }, { "grad_norm": 0.19808459281921387, "learning_rate": 8.48849825744642e-05, "loss": 0.019, "step": 87450 }, { "grad_norm": 0.18745198845863342, "learning_rate": 8.488103392955012e-05, "loss": 0.016, "step": 87460 }, { "grad_norm": 0.37773397564888, "learning_rate": 8.487708486079853e-05, "loss": 0.0207, "step": 87470 }, { "grad_norm": 0.23234547674655914, "learning_rate": 8.487313536825744e-05, "loss": 0.0165, "step": 87480 }, { "grad_norm": 0.1516256332397461, "learning_rate": 8.486918545197483e-05, "loss": 0.0187, "step": 87490 }, { "grad_norm": 0.20180118083953857, "learning_rate": 8.486523511199869e-05, "loss": 0.0213, "step": 87500 }, { "grad_norm": 0.22844156622886658, "learning_rate": 8.486128434837704e-05, "loss": 0.0225, "step": 87510 }, { "grad_norm": 0.3824211359024048, "learning_rate": 8.485733316115785e-05, "loss": 0.0166, "step": 87520 }, { "grad_norm": 0.11813582479953766, "learning_rate": 8.485338155038915e-05, "loss": 0.0289, "step": 87530 }, { "grad_norm": 0.263761967420578, "learning_rate": 8.484942951611896e-05, "loss": 0.0205, "step": 87540 }, { "grad_norm": 0.14867788553237915, "learning_rate": 8.484547705839532e-05, "loss": 0.0163, "step": 87550 }, { "grad_norm": 0.10615356266498566, "learning_rate": 8.48415241772662e-05, "loss": 0.024, "step": 87560 }, { "grad_norm": 0.06936527788639069, "learning_rate": 8.483757087277969e-05, "loss": 0.019, "step": 87570 }, { "grad_norm": 0.1965254843235016, "learning_rate": 8.483361714498379e-05, "loss": 0.0133, "step": 87580 }, { "grad_norm": 0.18530257046222687, "learning_rate": 8.482966299392656e-05, "loss": 0.0214, "step": 87590 }, { "grad_norm": 0.131022110581398, "learning_rate": 8.482570841965604e-05, "loss": 0.0158, "step": 87600 }, { "grad_norm": 0.16187775135040283, "learning_rate": 8.482175342222029e-05, "loss": 0.0198, "step": 87610 }, { "grad_norm": 0.1603076308965683, "learning_rate": 8.481779800166735e-05, "loss": 0.0224, "step": 87620 }, { "grad_norm": 0.14265157282352448, "learning_rate": 8.481384215804529e-05, "loss": 0.0193, "step": 87630 }, { "grad_norm": 0.22288239002227783, "learning_rate": 8.480988589140217e-05, "loss": 0.0185, "step": 87640 }, { "grad_norm": 0.08840083330869675, "learning_rate": 8.480592920178608e-05, "loss": 0.0159, "step": 87650 }, { "grad_norm": 0.1339513212442398, "learning_rate": 8.48019720892451e-05, "loss": 0.0175, "step": 87660 }, { "grad_norm": 0.1933240443468094, "learning_rate": 8.479801455382727e-05, "loss": 0.0226, "step": 87670 }, { "grad_norm": 0.17434509098529816, "learning_rate": 8.479405659558072e-05, "loss": 0.0209, "step": 87680 }, { "grad_norm": 0.20548292994499207, "learning_rate": 8.479009821455355e-05, "loss": 0.0201, "step": 87690 }, { "grad_norm": 0.15570208430290222, "learning_rate": 8.478613941079383e-05, "loss": 0.0174, "step": 87700 }, { "grad_norm": 0.25243258476257324, "learning_rate": 8.478218018434968e-05, "loss": 0.0224, "step": 87710 }, { "grad_norm": 0.3376392722129822, "learning_rate": 8.477822053526917e-05, "loss": 0.0176, "step": 87720 }, { "grad_norm": 0.23235781490802765, "learning_rate": 8.477426046360046e-05, "loss": 0.0192, "step": 87730 }, { "grad_norm": 0.14692102372646332, "learning_rate": 8.477029996939167e-05, "loss": 0.0175, "step": 87740 }, { "grad_norm": 0.10280723124742508, "learning_rate": 8.476633905269089e-05, "loss": 0.0131, "step": 87750 }, { "grad_norm": 0.14084771275520325, "learning_rate": 8.476237771354627e-05, "loss": 0.0163, "step": 87760 }, { "grad_norm": 0.12794162333011627, "learning_rate": 8.475841595200593e-05, "loss": 0.0229, "step": 87770 }, { "grad_norm": 0.1805116832256317, "learning_rate": 8.475445376811803e-05, "loss": 0.0271, "step": 87780 }, { "grad_norm": 0.12528659403324127, "learning_rate": 8.475049116193071e-05, "loss": 0.0211, "step": 87790 }, { "grad_norm": 0.25546860694885254, "learning_rate": 8.47465281334921e-05, "loss": 0.0221, "step": 87800 }, { "grad_norm": 0.37105345726013184, "learning_rate": 8.474256468285037e-05, "loss": 0.0166, "step": 87810 }, { "grad_norm": 0.1977127194404602, "learning_rate": 8.473860081005369e-05, "loss": 0.0214, "step": 87820 }, { "grad_norm": 0.1733889877796173, "learning_rate": 8.473463651515019e-05, "loss": 0.0159, "step": 87830 }, { "grad_norm": 0.21427960693836212, "learning_rate": 8.473067179818808e-05, "loss": 0.0153, "step": 87840 }, { "grad_norm": 0.1539335399866104, "learning_rate": 8.47267066592155e-05, "loss": 0.0174, "step": 87850 }, { "grad_norm": 0.16066059470176697, "learning_rate": 8.472274109828065e-05, "loss": 0.0175, "step": 87860 }, { "grad_norm": 0.12115050107240677, "learning_rate": 8.471877511543172e-05, "loss": 0.0224, "step": 87870 }, { "grad_norm": 0.3044911026954651, "learning_rate": 8.471480871071689e-05, "loss": 0.0189, "step": 87880 }, { "grad_norm": 0.1719946712255478, "learning_rate": 8.471084188418436e-05, "loss": 0.0199, "step": 87890 }, { "grad_norm": 0.20026032626628876, "learning_rate": 8.470687463588231e-05, "loss": 0.0183, "step": 87900 }, { "grad_norm": 0.20430368185043335, "learning_rate": 8.470290696585899e-05, "loss": 0.0316, "step": 87910 }, { "grad_norm": 0.1712537705898285, "learning_rate": 8.469893887416259e-05, "loss": 0.0148, "step": 87920 }, { "grad_norm": 0.1648942232131958, "learning_rate": 8.469497036084128e-05, "loss": 0.0177, "step": 87930 }, { "grad_norm": 0.1291321963071823, "learning_rate": 8.469100142594337e-05, "loss": 0.0173, "step": 87940 }, { "grad_norm": 0.15888254344463348, "learning_rate": 8.4687032069517e-05, "loss": 0.0205, "step": 87950 }, { "grad_norm": 0.234336718916893, "learning_rate": 8.468306229161047e-05, "loss": 0.02, "step": 87960 }, { "grad_norm": 0.09850534796714783, "learning_rate": 8.467909209227197e-05, "loss": 0.0175, "step": 87970 }, { "grad_norm": 0.0936402752995491, "learning_rate": 8.467512147154976e-05, "loss": 0.0194, "step": 87980 }, { "grad_norm": 0.21173883974552155, "learning_rate": 8.46711504294921e-05, "loss": 0.0184, "step": 87990 }, { "grad_norm": 0.11985787004232407, "learning_rate": 8.46671789661472e-05, "loss": 0.0158, "step": 88000 }, { "grad_norm": 0.2094273418188095, "learning_rate": 8.466320708156336e-05, "loss": 0.0167, "step": 88010 }, { "grad_norm": 0.14315059781074524, "learning_rate": 8.465923477578883e-05, "loss": 0.0156, "step": 88020 }, { "grad_norm": 0.0987689271569252, "learning_rate": 8.465526204887187e-05, "loss": 0.0156, "step": 88030 }, { "grad_norm": 0.08081338554620743, "learning_rate": 8.465128890086075e-05, "loss": 0.0142, "step": 88040 }, { "grad_norm": 0.27820247411727905, "learning_rate": 8.464731533180376e-05, "loss": 0.0178, "step": 88050 }, { "grad_norm": 0.14896078407764435, "learning_rate": 8.464334134174919e-05, "loss": 0.0177, "step": 88060 }, { "grad_norm": 0.19959111511707306, "learning_rate": 8.46393669307453e-05, "loss": 0.0253, "step": 88070 }, { "grad_norm": 0.170808807015419, "learning_rate": 8.46353920988404e-05, "loss": 0.0182, "step": 88080 }, { "grad_norm": 0.27782657742500305, "learning_rate": 8.463141684608278e-05, "loss": 0.0254, "step": 88090 }, { "grad_norm": 0.38489654660224915, "learning_rate": 8.462744117252074e-05, "loss": 0.0231, "step": 88100 }, { "grad_norm": 0.13187678158283234, "learning_rate": 8.46234650782026e-05, "loss": 0.0214, "step": 88110 }, { "grad_norm": 0.20770715177059174, "learning_rate": 8.461948856317669e-05, "loss": 0.027, "step": 88120 }, { "grad_norm": 0.14239583909511566, "learning_rate": 8.461551162749128e-05, "loss": 0.0207, "step": 88130 }, { "grad_norm": 0.24413564801216125, "learning_rate": 8.461153427119474e-05, "loss": 0.0251, "step": 88140 }, { "grad_norm": 0.22925329208374023, "learning_rate": 8.460755649433538e-05, "loss": 0.0177, "step": 88150 }, { "grad_norm": 0.25420090556144714, "learning_rate": 8.460357829696152e-05, "loss": 0.0226, "step": 88160 }, { "grad_norm": 0.19449710845947266, "learning_rate": 8.459959967912153e-05, "loss": 0.0196, "step": 88170 }, { "grad_norm": 0.2851214110851288, "learning_rate": 8.459562064086372e-05, "loss": 0.0216, "step": 88180 }, { "grad_norm": 0.5726491212844849, "learning_rate": 8.459164118223646e-05, "loss": 0.0238, "step": 88190 }, { "grad_norm": 0.1615966409444809, "learning_rate": 8.45876613032881e-05, "loss": 0.0261, "step": 88200 }, { "grad_norm": 0.14969401061534882, "learning_rate": 8.458368100406701e-05, "loss": 0.0279, "step": 88210 }, { "grad_norm": 0.14546065032482147, "learning_rate": 8.457970028462153e-05, "loss": 0.0211, "step": 88220 }, { "grad_norm": 0.15028046071529388, "learning_rate": 8.457571914500004e-05, "loss": 0.0208, "step": 88230 }, { "grad_norm": 0.08959818631410599, "learning_rate": 8.457173758525095e-05, "loss": 0.0189, "step": 88240 }, { "grad_norm": 0.12925323843955994, "learning_rate": 8.456775560542256e-05, "loss": 0.0204, "step": 88250 }, { "grad_norm": 0.15544450283050537, "learning_rate": 8.456377320556332e-05, "loss": 0.0168, "step": 88260 }, { "grad_norm": 0.11391211301088333, "learning_rate": 8.455979038572162e-05, "loss": 0.0183, "step": 88270 }, { "grad_norm": 0.2764202654361725, "learning_rate": 8.455580714594582e-05, "loss": 0.0179, "step": 88280 }, { "grad_norm": 0.3214711546897888, "learning_rate": 8.455182348628434e-05, "loss": 0.0199, "step": 88290 }, { "grad_norm": 1.5891354084014893, "learning_rate": 8.454783940678559e-05, "loss": 0.0238, "step": 88300 }, { "grad_norm": 0.18603132665157318, "learning_rate": 8.454385490749795e-05, "loss": 0.024, "step": 88310 }, { "grad_norm": 0.12229344993829727, "learning_rate": 8.453986998846988e-05, "loss": 0.0234, "step": 88320 }, { "grad_norm": 0.1170460432767868, "learning_rate": 8.453588464974978e-05, "loss": 0.0182, "step": 88330 }, { "grad_norm": 0.12900973856449127, "learning_rate": 8.453189889138607e-05, "loss": 0.0231, "step": 88340 }, { "grad_norm": 0.25372058153152466, "learning_rate": 8.452791271342717e-05, "loss": 0.015, "step": 88350 }, { "grad_norm": 0.1587338149547577, "learning_rate": 8.452392611592153e-05, "loss": 0.0224, "step": 88360 }, { "grad_norm": 0.19360747933387756, "learning_rate": 8.451993909891762e-05, "loss": 0.0189, "step": 88370 }, { "grad_norm": 0.12001793086528778, "learning_rate": 8.451595166246384e-05, "loss": 0.0211, "step": 88380 }, { "grad_norm": 0.14662711322307587, "learning_rate": 8.451196380660865e-05, "loss": 0.0191, "step": 88390 }, { "grad_norm": 0.21534599363803864, "learning_rate": 8.450797553140053e-05, "loss": 0.0207, "step": 88400 }, { "grad_norm": 0.25919994711875916, "learning_rate": 8.450398683688792e-05, "loss": 0.0171, "step": 88410 }, { "grad_norm": 0.24938800930976868, "learning_rate": 8.44999977231193e-05, "loss": 0.019, "step": 88420 }, { "grad_norm": 0.23954586684703827, "learning_rate": 8.449600819014312e-05, "loss": 0.02, "step": 88430 }, { "grad_norm": 0.1379183530807495, "learning_rate": 8.449201823800787e-05, "loss": 0.0153, "step": 88440 }, { "grad_norm": 0.3993411064147949, "learning_rate": 8.448802786676205e-05, "loss": 0.0222, "step": 88450 }, { "grad_norm": 0.25900155305862427, "learning_rate": 8.448403707645412e-05, "loss": 0.0252, "step": 88460 }, { "grad_norm": 0.08991159498691559, "learning_rate": 8.448004586713258e-05, "loss": 0.0193, "step": 88470 }, { "grad_norm": 0.2728562653064728, "learning_rate": 8.447605423884593e-05, "loss": 0.0225, "step": 88480 }, { "grad_norm": 0.2187659740447998, "learning_rate": 8.447206219164268e-05, "loss": 0.0207, "step": 88490 }, { "grad_norm": 0.15922361612319946, "learning_rate": 8.446806972557131e-05, "loss": 0.0146, "step": 88500 }, { "grad_norm": 0.22189994156360626, "learning_rate": 8.446407684068037e-05, "loss": 0.0229, "step": 88510 }, { "grad_norm": 0.17761993408203125, "learning_rate": 8.446008353701833e-05, "loss": 0.0146, "step": 88520 }, { "grad_norm": 0.18417009711265564, "learning_rate": 8.445608981463375e-05, "loss": 0.017, "step": 88530 }, { "grad_norm": 0.11814706772565842, "learning_rate": 8.445209567357515e-05, "loss": 0.0172, "step": 88540 }, { "grad_norm": 0.14219491183757782, "learning_rate": 8.444810111389108e-05, "loss": 0.0188, "step": 88550 }, { "grad_norm": 0.1727515459060669, "learning_rate": 8.444410613563004e-05, "loss": 0.0162, "step": 88560 }, { "grad_norm": 0.2406172901391983, "learning_rate": 8.444011073884057e-05, "loss": 0.021, "step": 88570 }, { "grad_norm": 0.15215753018856049, "learning_rate": 8.443611492357126e-05, "loss": 0.0147, "step": 88580 }, { "grad_norm": 0.12118595093488693, "learning_rate": 8.443211868987063e-05, "loss": 0.0168, "step": 88590 }, { "grad_norm": 0.12409740686416626, "learning_rate": 8.442812203778726e-05, "loss": 0.0209, "step": 88600 }, { "grad_norm": 0.3777909278869629, "learning_rate": 8.44241249673697e-05, "loss": 0.0172, "step": 88610 }, { "grad_norm": 0.09821128845214844, "learning_rate": 8.44201274786665e-05, "loss": 0.0137, "step": 88620 }, { "grad_norm": 0.5036303997039795, "learning_rate": 8.441612957172626e-05, "loss": 0.0183, "step": 88630 }, { "grad_norm": 0.10331942141056061, "learning_rate": 8.441213124659757e-05, "loss": 0.0245, "step": 88640 }, { "grad_norm": 0.16552212834358215, "learning_rate": 8.4408132503329e-05, "loss": 0.0238, "step": 88650 }, { "grad_norm": 0.0946429893374443, "learning_rate": 8.44041333419691e-05, "loss": 0.0186, "step": 88660 }, { "grad_norm": 0.1031133309006691, "learning_rate": 8.440013376256653e-05, "loss": 0.0201, "step": 88670 }, { "grad_norm": 0.1979539841413498, "learning_rate": 8.439613376516984e-05, "loss": 0.0227, "step": 88680 }, { "grad_norm": 0.1395532488822937, "learning_rate": 8.439213334982765e-05, "loss": 0.0203, "step": 88690 }, { "grad_norm": 0.2731119394302368, "learning_rate": 8.438813251658857e-05, "loss": 0.019, "step": 88700 }, { "grad_norm": 0.12639577686786652, "learning_rate": 8.438413126550119e-05, "loss": 0.019, "step": 88710 }, { "grad_norm": 0.20535413920879364, "learning_rate": 8.438012959661419e-05, "loss": 0.018, "step": 88720 }, { "grad_norm": 0.09152714163064957, "learning_rate": 8.437612750997614e-05, "loss": 0.019, "step": 88730 }, { "grad_norm": 0.1998382955789566, "learning_rate": 8.437212500563567e-05, "loss": 0.021, "step": 88740 }, { "grad_norm": 0.18067564070224762, "learning_rate": 8.436812208364144e-05, "loss": 0.0156, "step": 88750 }, { "grad_norm": 0.26670071482658386, "learning_rate": 8.436411874404209e-05, "loss": 0.0179, "step": 88760 }, { "grad_norm": 0.11464416235685349, "learning_rate": 8.436011498688624e-05, "loss": 0.015, "step": 88770 }, { "grad_norm": 0.1539650708436966, "learning_rate": 8.435611081222255e-05, "loss": 0.0174, "step": 88780 }, { "grad_norm": 0.2309517115354538, "learning_rate": 8.435210622009967e-05, "loss": 0.0262, "step": 88790 }, { "grad_norm": 0.10722005367279053, "learning_rate": 8.434810121056627e-05, "loss": 0.019, "step": 88800 }, { "grad_norm": 0.24668829143047333, "learning_rate": 8.434409578367102e-05, "loss": 0.0284, "step": 88810 }, { "grad_norm": 0.2858259975910187, "learning_rate": 8.434008993946256e-05, "loss": 0.0218, "step": 88820 }, { "grad_norm": 0.1858581304550171, "learning_rate": 8.433608367798961e-05, "loss": 0.0199, "step": 88830 }, { "grad_norm": 0.18471620976924896, "learning_rate": 8.433207699930081e-05, "loss": 0.02, "step": 88840 }, { "grad_norm": 0.10993298888206482, "learning_rate": 8.432806990344486e-05, "loss": 0.0153, "step": 88850 }, { "grad_norm": 0.14266109466552734, "learning_rate": 8.432406239047046e-05, "loss": 0.0186, "step": 88860 }, { "grad_norm": 0.2678760886192322, "learning_rate": 8.432005446042627e-05, "loss": 0.018, "step": 88870 }, { "grad_norm": 0.30504724383354187, "learning_rate": 8.431604611336105e-05, "loss": 0.0199, "step": 88880 }, { "grad_norm": 0.11737262457609177, "learning_rate": 8.431203734932345e-05, "loss": 0.0184, "step": 88890 }, { "grad_norm": 0.12863832712173462, "learning_rate": 8.430802816836221e-05, "loss": 0.0185, "step": 88900 }, { "grad_norm": 0.14601534605026245, "learning_rate": 8.430401857052601e-05, "loss": 0.0221, "step": 88910 }, { "grad_norm": 0.17130471765995026, "learning_rate": 8.430000855586362e-05, "loss": 0.0189, "step": 88920 }, { "grad_norm": 0.12095329165458679, "learning_rate": 8.429599812442372e-05, "loss": 0.0137, "step": 88930 }, { "grad_norm": 0.15669138729572296, "learning_rate": 8.429198727625509e-05, "loss": 0.02, "step": 88940 }, { "grad_norm": 0.09177438169717789, "learning_rate": 8.428797601140642e-05, "loss": 0.0178, "step": 88950 }, { "grad_norm": 0.1535039097070694, "learning_rate": 8.428396432992648e-05, "loss": 0.0147, "step": 88960 }, { "grad_norm": 0.20066313445568085, "learning_rate": 8.4279952231864e-05, "loss": 0.0208, "step": 88970 }, { "grad_norm": 0.10873587429523468, "learning_rate": 8.427593971726772e-05, "loss": 0.0164, "step": 88980 }, { "grad_norm": 0.11081498116254807, "learning_rate": 8.427192678618643e-05, "loss": 0.0197, "step": 88990 }, { "grad_norm": 0.17844395339488983, "learning_rate": 8.426791343866887e-05, "loss": 0.0204, "step": 89000 }, { "grad_norm": 0.20230531692504883, "learning_rate": 8.426389967476381e-05, "loss": 0.0189, "step": 89010 }, { "grad_norm": 0.09826493263244629, "learning_rate": 8.425988549452e-05, "loss": 0.0195, "step": 89020 }, { "grad_norm": 0.17810621857643127, "learning_rate": 8.425587089798624e-05, "loss": 0.0229, "step": 89030 }, { "grad_norm": 0.5166541934013367, "learning_rate": 8.425185588521131e-05, "loss": 0.0202, "step": 89040 }, { "grad_norm": 0.16314394772052765, "learning_rate": 8.424784045624398e-05, "loss": 0.0213, "step": 89050 }, { "grad_norm": 0.21402430534362793, "learning_rate": 8.424382461113307e-05, "loss": 0.0205, "step": 89060 }, { "grad_norm": 0.18240328133106232, "learning_rate": 8.423980834992736e-05, "loss": 0.0169, "step": 89070 }, { "grad_norm": 0.21327023208141327, "learning_rate": 8.423579167267564e-05, "loss": 0.0203, "step": 89080 }, { "grad_norm": 0.09514178335666656, "learning_rate": 8.423177457942674e-05, "loss": 0.0218, "step": 89090 }, { "grad_norm": 0.3769887387752533, "learning_rate": 8.422775707022946e-05, "loss": 0.0221, "step": 89100 }, { "grad_norm": 0.6149035096168518, "learning_rate": 8.42237391451326e-05, "loss": 0.0235, "step": 89110 }, { "grad_norm": 0.2591797113418579, "learning_rate": 8.4219720804185e-05, "loss": 0.0177, "step": 89120 }, { "grad_norm": 0.17553958296775818, "learning_rate": 8.421570204743549e-05, "loss": 0.0169, "step": 89130 }, { "grad_norm": 0.07911518216133118, "learning_rate": 8.421168287493288e-05, "loss": 0.0122, "step": 89140 }, { "grad_norm": 0.8804277181625366, "learning_rate": 8.420766328672604e-05, "loss": 0.0224, "step": 89150 }, { "grad_norm": 0.1253734529018402, "learning_rate": 8.420364328286379e-05, "loss": 0.0169, "step": 89160 }, { "grad_norm": 0.13400310277938843, "learning_rate": 8.419962286339497e-05, "loss": 0.0198, "step": 89170 }, { "grad_norm": 0.14560769498348236, "learning_rate": 8.419560202836844e-05, "loss": 0.0205, "step": 89180 }, { "grad_norm": 0.08064743876457214, "learning_rate": 8.419158077783309e-05, "loss": 0.0182, "step": 89190 }, { "grad_norm": 0.12821006774902344, "learning_rate": 8.418755911183773e-05, "loss": 0.0198, "step": 89200 }, { "grad_norm": 0.12839598953723907, "learning_rate": 8.418353703043125e-05, "loss": 0.021, "step": 89210 }, { "grad_norm": 0.12104007601737976, "learning_rate": 8.417951453366251e-05, "loss": 0.0183, "step": 89220 }, { "grad_norm": 0.2828879654407501, "learning_rate": 8.41754916215804e-05, "loss": 0.019, "step": 89230 }, { "grad_norm": 0.14513465762138367, "learning_rate": 8.417146829423383e-05, "loss": 0.0197, "step": 89240 }, { "grad_norm": 0.7328037619590759, "learning_rate": 8.416744455167162e-05, "loss": 0.0212, "step": 89250 }, { "grad_norm": 0.18380141258239746, "learning_rate": 8.416342039394272e-05, "loss": 0.0171, "step": 89260 }, { "grad_norm": 0.1619201898574829, "learning_rate": 8.4159395821096e-05, "loss": 0.0199, "step": 89270 }, { "grad_norm": 0.13613344728946686, "learning_rate": 8.415537083318038e-05, "loss": 0.0197, "step": 89280 }, { "grad_norm": 0.4361744523048401, "learning_rate": 8.415134543024474e-05, "loss": 0.0161, "step": 89290 }, { "grad_norm": 0.23945952951908112, "learning_rate": 8.4147319612338e-05, "loss": 0.0207, "step": 89300 }, { "grad_norm": 0.17193447053432465, "learning_rate": 8.41432933795091e-05, "loss": 0.0173, "step": 89310 }, { "grad_norm": 0.18828940391540527, "learning_rate": 8.413926673180696e-05, "loss": 0.0176, "step": 89320 }, { "grad_norm": 0.2150416374206543, "learning_rate": 8.413523966928048e-05, "loss": 0.0151, "step": 89330 }, { "grad_norm": 0.20476464927196503, "learning_rate": 8.413121219197861e-05, "loss": 0.0216, "step": 89340 }, { "grad_norm": 0.14131541550159454, "learning_rate": 8.41271842999503e-05, "loss": 0.0194, "step": 89350 }, { "grad_norm": 0.14342360198497772, "learning_rate": 8.412315599324448e-05, "loss": 0.0222, "step": 89360 }, { "grad_norm": 0.3087693452835083, "learning_rate": 8.411912727191006e-05, "loss": 0.022, "step": 89370 }, { "grad_norm": 0.19175660610198975, "learning_rate": 8.411509813599608e-05, "loss": 0.0184, "step": 89380 }, { "grad_norm": 0.18351098895072937, "learning_rate": 8.411106858555142e-05, "loss": 0.0155, "step": 89390 }, { "grad_norm": 0.06971293687820435, "learning_rate": 8.410703862062508e-05, "loss": 0.015, "step": 89400 }, { "grad_norm": 0.13747824728488922, "learning_rate": 8.4103008241266e-05, "loss": 0.0168, "step": 89410 }, { "grad_norm": 0.14185963571071625, "learning_rate": 8.40989774475232e-05, "loss": 0.0172, "step": 89420 }, { "grad_norm": 0.5413790941238403, "learning_rate": 8.40949462394456e-05, "loss": 0.0213, "step": 89430 }, { "grad_norm": 0.13808515667915344, "learning_rate": 8.409091461708224e-05, "loss": 0.0148, "step": 89440 }, { "grad_norm": 0.16861236095428467, "learning_rate": 8.408688258048207e-05, "loss": 0.0176, "step": 89450 }, { "grad_norm": 0.2356429547071457, "learning_rate": 8.408285012969408e-05, "loss": 0.0219, "step": 89460 }, { "grad_norm": 0.33702245354652405, "learning_rate": 8.40788172647673e-05, "loss": 0.0179, "step": 89470 }, { "grad_norm": 0.20586326718330383, "learning_rate": 8.40747839857507e-05, "loss": 0.0189, "step": 89480 }, { "grad_norm": 0.175283282995224, "learning_rate": 8.407075029269332e-05, "loss": 0.0197, "step": 89490 }, { "grad_norm": 0.15141485631465912, "learning_rate": 8.406671618564413e-05, "loss": 0.0199, "step": 89500 }, { "grad_norm": 0.1900210827589035, "learning_rate": 8.406268166465218e-05, "loss": 0.0196, "step": 89510 }, { "grad_norm": 0.14726729691028595, "learning_rate": 8.405864672976652e-05, "loss": 0.0177, "step": 89520 }, { "grad_norm": 0.17093582451343536, "learning_rate": 8.405461138103612e-05, "loss": 0.0226, "step": 89530 }, { "grad_norm": 0.11503078043460846, "learning_rate": 8.405057561851005e-05, "loss": 0.0148, "step": 89540 }, { "grad_norm": 0.2594788372516632, "learning_rate": 8.40465394422373e-05, "loss": 0.0222, "step": 89550 }, { "grad_norm": 0.37004512548446655, "learning_rate": 8.4042502852267e-05, "loss": 0.0236, "step": 89560 }, { "grad_norm": 0.25336289405822754, "learning_rate": 8.403846584864812e-05, "loss": 0.0224, "step": 89570 }, { "grad_norm": 0.1660294234752655, "learning_rate": 8.403442843142974e-05, "loss": 0.0183, "step": 89580 }, { "grad_norm": 0.11203552782535553, "learning_rate": 8.403039060066094e-05, "loss": 0.0235, "step": 89590 }, { "grad_norm": 0.12944673001766205, "learning_rate": 8.402635235639075e-05, "loss": 0.0163, "step": 89600 }, { "grad_norm": 0.24691303074359894, "learning_rate": 8.402231369866827e-05, "loss": 0.0212, "step": 89610 }, { "grad_norm": 0.16444973647594452, "learning_rate": 8.401827462754253e-05, "loss": 0.0182, "step": 89620 }, { "grad_norm": 0.13859343528747559, "learning_rate": 8.401423514306265e-05, "loss": 0.0145, "step": 89630 }, { "grad_norm": 0.25654780864715576, "learning_rate": 8.401019524527769e-05, "loss": 0.0232, "step": 89640 }, { "grad_norm": 0.19463218748569489, "learning_rate": 8.400615493423675e-05, "loss": 0.0164, "step": 89650 }, { "grad_norm": 0.23036406934261322, "learning_rate": 8.400211420998893e-05, "loss": 0.0262, "step": 89660 }, { "grad_norm": 0.12884673476219177, "learning_rate": 8.399807307258332e-05, "loss": 0.0142, "step": 89670 }, { "grad_norm": 0.12063475698232651, "learning_rate": 8.399403152206901e-05, "loss": 0.0199, "step": 89680 }, { "grad_norm": 0.10051393508911133, "learning_rate": 8.398998955849513e-05, "loss": 0.0148, "step": 89690 }, { "grad_norm": 0.12102481722831726, "learning_rate": 8.398594718191079e-05, "loss": 0.0212, "step": 89700 }, { "grad_norm": 0.10781954973936081, "learning_rate": 8.398190439236508e-05, "loss": 0.0246, "step": 89710 }, { "grad_norm": 0.11118197441101074, "learning_rate": 8.397786118990717e-05, "loss": 0.0182, "step": 89720 }, { "grad_norm": 0.1480540931224823, "learning_rate": 8.397381757458616e-05, "loss": 0.0168, "step": 89730 }, { "grad_norm": 0.20613816380500793, "learning_rate": 8.396977354645119e-05, "loss": 0.0101, "step": 89740 }, { "grad_norm": 0.0776311382651329, "learning_rate": 8.396572910555138e-05, "loss": 0.0126, "step": 89750 }, { "grad_norm": 0.11314938217401505, "learning_rate": 8.396168425193592e-05, "loss": 0.0182, "step": 89760 }, { "grad_norm": 0.19884945452213287, "learning_rate": 8.395763898565392e-05, "loss": 0.0201, "step": 89770 }, { "grad_norm": 0.18763065338134766, "learning_rate": 8.395359330675454e-05, "loss": 0.0314, "step": 89780 }, { "grad_norm": 0.22545212507247925, "learning_rate": 8.394954721528696e-05, "loss": 0.0217, "step": 89790 }, { "grad_norm": 0.1425563097000122, "learning_rate": 8.394550071130031e-05, "loss": 0.0209, "step": 89800 }, { "grad_norm": 0.25114208459854126, "learning_rate": 8.39414537948438e-05, "loss": 0.0215, "step": 89810 }, { "grad_norm": 0.12887956202030182, "learning_rate": 8.393740646596655e-05, "loss": 0.0193, "step": 89820 }, { "grad_norm": 0.1281188577413559, "learning_rate": 8.393335872471777e-05, "loss": 0.0178, "step": 89830 }, { "grad_norm": 0.17030930519104004, "learning_rate": 8.392931057114667e-05, "loss": 0.0207, "step": 89840 }, { "grad_norm": 0.2153603583574295, "learning_rate": 8.39252620053024e-05, "loss": 0.0205, "step": 89850 }, { "grad_norm": 0.1333637237548828, "learning_rate": 8.392121302723417e-05, "loss": 0.018, "step": 89860 }, { "grad_norm": 0.10162222385406494, "learning_rate": 8.391716363699115e-05, "loss": 0.0193, "step": 89870 }, { "grad_norm": 0.16544277966022491, "learning_rate": 8.391311383462259e-05, "loss": 0.0179, "step": 89880 }, { "grad_norm": 0.23665574193000793, "learning_rate": 8.390906362017767e-05, "loss": 0.0181, "step": 89890 }, { "grad_norm": 0.41605696082115173, "learning_rate": 8.390501299370563e-05, "loss": 0.0204, "step": 89900 }, { "grad_norm": 0.29670068621635437, "learning_rate": 8.390096195525565e-05, "loss": 0.018, "step": 89910 }, { "grad_norm": 0.20016728341579437, "learning_rate": 8.389691050487699e-05, "loss": 0.0186, "step": 89920 }, { "grad_norm": 0.2573634684085846, "learning_rate": 8.389285864261887e-05, "loss": 0.0184, "step": 89930 }, { "grad_norm": 0.2443331778049469, "learning_rate": 8.38888063685305e-05, "loss": 0.017, "step": 89940 }, { "grad_norm": 0.2382807582616806, "learning_rate": 8.388475368266113e-05, "loss": 0.0206, "step": 89950 }, { "grad_norm": 0.21565520763397217, "learning_rate": 8.388070058506001e-05, "loss": 0.0165, "step": 89960 }, { "grad_norm": 0.18003007769584656, "learning_rate": 8.38766470757764e-05, "loss": 0.0178, "step": 89970 }, { "grad_norm": 0.19079242646694183, "learning_rate": 8.387259315485954e-05, "loss": 0.0254, "step": 89980 }, { "grad_norm": 0.13347719609737396, "learning_rate": 8.38685388223587e-05, "loss": 0.0188, "step": 89990 }, { "grad_norm": 0.1445087492465973, "learning_rate": 8.386448407832313e-05, "loss": 0.0173, "step": 90000 }, { "grad_norm": 0.28316986560821533, "learning_rate": 8.38604289228021e-05, "loss": 0.0241, "step": 90010 }, { "grad_norm": 0.3677741587162018, "learning_rate": 8.38563733558449e-05, "loss": 0.0242, "step": 90020 }, { "grad_norm": 0.16876743733882904, "learning_rate": 8.385231737750078e-05, "loss": 0.0155, "step": 90030 }, { "grad_norm": 0.13788707554340363, "learning_rate": 8.384826098781906e-05, "loss": 0.0172, "step": 90040 }, { "grad_norm": 0.16698801517486572, "learning_rate": 8.384420418684901e-05, "loss": 0.0173, "step": 90050 }, { "grad_norm": 0.15397484600543976, "learning_rate": 8.384014697463992e-05, "loss": 0.0168, "step": 90060 }, { "grad_norm": 0.11902610957622528, "learning_rate": 8.383608935124109e-05, "loss": 0.0212, "step": 90070 }, { "grad_norm": 0.3347267210483551, "learning_rate": 8.383203131670183e-05, "loss": 0.0184, "step": 90080 }, { "grad_norm": 0.13749326765537262, "learning_rate": 8.382797287107145e-05, "loss": 0.0191, "step": 90090 }, { "grad_norm": 0.2429104596376419, "learning_rate": 8.382391401439927e-05, "loss": 0.0196, "step": 90100 }, { "grad_norm": 0.3046579360961914, "learning_rate": 8.381985474673458e-05, "loss": 0.019, "step": 90110 }, { "grad_norm": 0.15065079927444458, "learning_rate": 8.381579506812673e-05, "loss": 0.0141, "step": 90120 }, { "grad_norm": 0.18446677923202515, "learning_rate": 8.381173497862504e-05, "loss": 0.0127, "step": 90130 }, { "grad_norm": 0.1288757473230362, "learning_rate": 8.380767447827886e-05, "loss": 0.0193, "step": 90140 }, { "grad_norm": 0.17700108885765076, "learning_rate": 8.380361356713751e-05, "loss": 0.0176, "step": 90150 }, { "grad_norm": 0.41187378764152527, "learning_rate": 8.379955224525032e-05, "loss": 0.0213, "step": 90160 }, { "grad_norm": 0.3553341031074524, "learning_rate": 8.379549051266669e-05, "loss": 0.0219, "step": 90170 }, { "grad_norm": 0.12561151385307312, "learning_rate": 8.379142836943592e-05, "loss": 0.0169, "step": 90180 }, { "grad_norm": 0.10023611783981323, "learning_rate": 8.37873658156074e-05, "loss": 0.0203, "step": 90190 }, { "grad_norm": 0.11256921291351318, "learning_rate": 8.378330285123048e-05, "loss": 0.0213, "step": 90200 }, { "grad_norm": 0.1986224502325058, "learning_rate": 8.377923947635453e-05, "loss": 0.0177, "step": 90210 }, { "grad_norm": 0.15555141866207123, "learning_rate": 8.377517569102894e-05, "loss": 0.017, "step": 90220 }, { "grad_norm": 0.12255090475082397, "learning_rate": 8.377111149530306e-05, "loss": 0.0198, "step": 90230 }, { "grad_norm": 0.17699562013149261, "learning_rate": 8.376704688922629e-05, "loss": 0.0209, "step": 90240 }, { "grad_norm": 0.21253502368927002, "learning_rate": 8.376298187284803e-05, "loss": 0.0227, "step": 90250 }, { "grad_norm": 0.1149151623249054, "learning_rate": 8.375891644621767e-05, "loss": 0.0132, "step": 90260 }, { "grad_norm": 0.16909319162368774, "learning_rate": 8.375485060938459e-05, "loss": 0.0203, "step": 90270 }, { "grad_norm": 0.11112091690301895, "learning_rate": 8.375078436239821e-05, "loss": 0.026, "step": 90280 }, { "grad_norm": 0.1842193901538849, "learning_rate": 8.374671770530793e-05, "loss": 0.0158, "step": 90290 }, { "grad_norm": 0.11605577170848846, "learning_rate": 8.374265063816316e-05, "loss": 0.0194, "step": 90300 }, { "grad_norm": 0.1389468014240265, "learning_rate": 8.373858316101335e-05, "loss": 0.0219, "step": 90310 }, { "grad_norm": 0.3774714469909668, "learning_rate": 8.373451527390787e-05, "loss": 0.0186, "step": 90320 }, { "grad_norm": 0.11204340308904648, "learning_rate": 8.373044697689621e-05, "loss": 0.02, "step": 90330 }, { "grad_norm": 0.24131371080875397, "learning_rate": 8.372637827002776e-05, "loss": 0.0194, "step": 90340 }, { "grad_norm": 0.18308766186237335, "learning_rate": 8.372230915335195e-05, "loss": 0.0135, "step": 90350 }, { "grad_norm": 0.30302536487579346, "learning_rate": 8.371823962691827e-05, "loss": 0.0197, "step": 90360 }, { "grad_norm": 0.20939882099628448, "learning_rate": 8.371416969077614e-05, "loss": 0.0213, "step": 90370 }, { "grad_norm": 0.16496002674102783, "learning_rate": 8.371009934497502e-05, "loss": 0.0192, "step": 90380 }, { "grad_norm": 0.18474923074245453, "learning_rate": 8.370602858956434e-05, "loss": 0.0199, "step": 90390 }, { "grad_norm": 0.1413397639989853, "learning_rate": 8.370195742459362e-05, "loss": 0.0184, "step": 90400 }, { "grad_norm": 0.09002116322517395, "learning_rate": 8.36978858501123e-05, "loss": 0.0185, "step": 90410 }, { "grad_norm": 0.14449375867843628, "learning_rate": 8.369381386616981e-05, "loss": 0.0144, "step": 90420 }, { "grad_norm": 0.12134625017642975, "learning_rate": 8.36897414728157e-05, "loss": 0.0184, "step": 90430 }, { "grad_norm": 0.1661798506975174, "learning_rate": 8.368566867009942e-05, "loss": 0.0196, "step": 90440 }, { "grad_norm": 0.15166138112545013, "learning_rate": 8.368159545807047e-05, "loss": 0.0166, "step": 90450 }, { "grad_norm": 0.17521582543849945, "learning_rate": 8.36775218367783e-05, "loss": 0.0227, "step": 90460 }, { "grad_norm": 0.11137286573648453, "learning_rate": 8.367344780627246e-05, "loss": 0.019, "step": 90470 }, { "grad_norm": 0.19825193285942078, "learning_rate": 8.366937336660246e-05, "loss": 0.0207, "step": 90480 }, { "grad_norm": 0.09739566594362259, "learning_rate": 8.366529851781776e-05, "loss": 0.0169, "step": 90490 }, { "grad_norm": 0.15910884737968445, "learning_rate": 8.36612232599679e-05, "loss": 0.0191, "step": 90500 }, { "grad_norm": 0.10739949345588684, "learning_rate": 8.36571475931024e-05, "loss": 0.0179, "step": 90510 }, { "grad_norm": 0.09716257452964783, "learning_rate": 8.365307151727078e-05, "loss": 0.0163, "step": 90520 }, { "grad_norm": 0.10623311251401901, "learning_rate": 8.364899503252257e-05, "loss": 0.021, "step": 90530 }, { "grad_norm": 0.10999783873558044, "learning_rate": 8.36449181389073e-05, "loss": 0.0187, "step": 90540 }, { "grad_norm": 0.11687827110290527, "learning_rate": 8.36408408364745e-05, "loss": 0.0153, "step": 90550 }, { "grad_norm": 0.08815228193998337, "learning_rate": 8.363676312527372e-05, "loss": 0.0126, "step": 90560 }, { "grad_norm": 0.2950849235057831, "learning_rate": 8.363268500535452e-05, "loss": 0.0251, "step": 90570 }, { "grad_norm": 0.18447203934192657, "learning_rate": 8.362860647676644e-05, "loss": 0.0159, "step": 90580 }, { "grad_norm": 0.17431262135505676, "learning_rate": 8.362452753955904e-05, "loss": 0.0188, "step": 90590 }, { "grad_norm": 0.19478119909763336, "learning_rate": 8.362044819378188e-05, "loss": 0.0193, "step": 90600 }, { "grad_norm": 0.10881534218788147, "learning_rate": 8.361636843948454e-05, "loss": 0.0142, "step": 90610 }, { "grad_norm": 0.10337185114622116, "learning_rate": 8.361228827671657e-05, "loss": 0.0171, "step": 90620 }, { "grad_norm": 0.09211532771587372, "learning_rate": 8.360820770552757e-05, "loss": 0.0171, "step": 90630 }, { "grad_norm": 0.06385701149702072, "learning_rate": 8.360412672596712e-05, "loss": 0.0151, "step": 90640 }, { "grad_norm": 0.10174893587827682, "learning_rate": 8.36000453380848e-05, "loss": 0.0222, "step": 90650 }, { "grad_norm": 0.21797171235084534, "learning_rate": 8.35959635419302e-05, "loss": 0.0157, "step": 90660 }, { "grad_norm": 0.0928405299782753, "learning_rate": 8.359188133755294e-05, "loss": 0.014, "step": 90670 }, { "grad_norm": 0.1540917307138443, "learning_rate": 8.358779872500259e-05, "loss": 0.013, "step": 90680 }, { "grad_norm": 0.09010709822177887, "learning_rate": 8.358371570432877e-05, "loss": 0.0182, "step": 90690 }, { "grad_norm": 0.0836794376373291, "learning_rate": 8.35796322755811e-05, "loss": 0.0191, "step": 90700 }, { "grad_norm": 0.179361030459404, "learning_rate": 8.357554843880919e-05, "loss": 0.0287, "step": 90710 }, { "grad_norm": 0.17399464547634125, "learning_rate": 8.357146419406268e-05, "loss": 0.0197, "step": 90720 }, { "grad_norm": 0.15352512896060944, "learning_rate": 8.356737954139118e-05, "loss": 0.019, "step": 90730 }, { "grad_norm": 0.29434022307395935, "learning_rate": 8.356329448084432e-05, "loss": 0.0166, "step": 90740 }, { "grad_norm": 0.2322928011417389, "learning_rate": 8.355920901247174e-05, "loss": 0.0215, "step": 90750 }, { "grad_norm": 0.2475721389055252, "learning_rate": 8.355512313632309e-05, "loss": 0.0191, "step": 90760 }, { "grad_norm": 0.16282612085342407, "learning_rate": 8.355103685244801e-05, "loss": 0.0241, "step": 90770 }, { "grad_norm": 0.6852753758430481, "learning_rate": 8.354695016089616e-05, "loss": 0.0194, "step": 90780 }, { "grad_norm": 0.08959652483463287, "learning_rate": 8.35428630617172e-05, "loss": 0.0217, "step": 90790 }, { "grad_norm": 0.2096175253391266, "learning_rate": 8.353877555496076e-05, "loss": 0.0196, "step": 90800 }, { "grad_norm": 0.18410778045654297, "learning_rate": 8.353468764067654e-05, "loss": 0.021, "step": 90810 }, { "grad_norm": 0.26640743017196655, "learning_rate": 8.353059931891422e-05, "loss": 0.0197, "step": 90820 }, { "grad_norm": 0.14282608032226562, "learning_rate": 8.352651058972344e-05, "loss": 0.0152, "step": 90830 }, { "grad_norm": 0.1281682401895523, "learning_rate": 8.352242145315392e-05, "loss": 0.0172, "step": 90840 }, { "grad_norm": 0.09216340631246567, "learning_rate": 8.351833190925531e-05, "loss": 0.0183, "step": 90850 }, { "grad_norm": 0.13575264811515808, "learning_rate": 8.351424195807733e-05, "loss": 0.016, "step": 90860 }, { "grad_norm": 0.12954387068748474, "learning_rate": 8.351015159966966e-05, "loss": 0.0214, "step": 90870 }, { "grad_norm": 0.15862682461738586, "learning_rate": 8.350606083408202e-05, "loss": 0.0184, "step": 90880 }, { "grad_norm": 0.10995998233556747, "learning_rate": 8.35019696613641e-05, "loss": 0.02, "step": 90890 }, { "grad_norm": 0.13636037707328796, "learning_rate": 8.349787808156562e-05, "loss": 0.0173, "step": 90900 }, { "grad_norm": 0.11560717970132828, "learning_rate": 8.34937860947363e-05, "loss": 0.0145, "step": 90910 }, { "grad_norm": 0.27704277634620667, "learning_rate": 8.348969370092584e-05, "loss": 0.0235, "step": 90920 }, { "grad_norm": 0.13928622007369995, "learning_rate": 8.348560090018399e-05, "loss": 0.0181, "step": 90930 }, { "grad_norm": 0.12076496332883835, "learning_rate": 8.348150769256047e-05, "loss": 0.0206, "step": 90940 }, { "grad_norm": 0.36961641907691956, "learning_rate": 8.347741407810503e-05, "loss": 0.0205, "step": 90950 }, { "grad_norm": 0.1433633714914322, "learning_rate": 8.34733200568674e-05, "loss": 0.0194, "step": 90960 }, { "grad_norm": 0.20285457372665405, "learning_rate": 8.346922562889732e-05, "loss": 0.0188, "step": 90970 }, { "grad_norm": 0.21693548560142517, "learning_rate": 8.346513079424455e-05, "loss": 0.0216, "step": 90980 }, { "grad_norm": 0.14343591034412384, "learning_rate": 8.346103555295883e-05, "loss": 0.0152, "step": 90990 }, { "grad_norm": 0.1624753177165985, "learning_rate": 8.345693990508996e-05, "loss": 0.0168, "step": 91000 }, { "grad_norm": 0.25050055980682373, "learning_rate": 8.345284385068768e-05, "loss": 0.0204, "step": 91010 }, { "grad_norm": 0.09320200979709625, "learning_rate": 8.344874738980174e-05, "loss": 0.0132, "step": 91020 }, { "grad_norm": 0.11452575027942657, "learning_rate": 8.344465052248196e-05, "loss": 0.0197, "step": 91030 }, { "grad_norm": 0.16386885941028595, "learning_rate": 8.34405532487781e-05, "loss": 0.0225, "step": 91040 }, { "grad_norm": 0.37412309646606445, "learning_rate": 8.343645556873994e-05, "loss": 0.0175, "step": 91050 }, { "grad_norm": 0.10067670047283173, "learning_rate": 8.343235748241727e-05, "loss": 0.0139, "step": 91060 }, { "grad_norm": 0.20234465599060059, "learning_rate": 8.342825898985989e-05, "loss": 0.0231, "step": 91070 }, { "grad_norm": 0.14757677912712097, "learning_rate": 8.34241600911176e-05, "loss": 0.0172, "step": 91080 }, { "grad_norm": 0.10951801389455795, "learning_rate": 8.342006078624023e-05, "loss": 0.013, "step": 91090 }, { "grad_norm": 0.12165353447198868, "learning_rate": 8.341596107527753e-05, "loss": 0.0173, "step": 91100 }, { "grad_norm": 0.24651078879833221, "learning_rate": 8.341186095827939e-05, "loss": 0.0177, "step": 91110 }, { "grad_norm": 0.1412951648235321, "learning_rate": 8.340776043529557e-05, "loss": 0.0231, "step": 91120 }, { "grad_norm": 0.12929347157478333, "learning_rate": 8.340365950637592e-05, "loss": 0.0202, "step": 91130 }, { "grad_norm": 0.14628328382968903, "learning_rate": 8.339955817157028e-05, "loss": 0.0196, "step": 91140 }, { "grad_norm": 0.1805093139410019, "learning_rate": 8.339545643092848e-05, "loss": 0.0166, "step": 91150 }, { "grad_norm": 0.13908421993255615, "learning_rate": 8.339135428450034e-05, "loss": 0.018, "step": 91160 }, { "grad_norm": 0.1433599591255188, "learning_rate": 8.338725173233571e-05, "loss": 0.0144, "step": 91170 }, { "grad_norm": 0.2751673758029938, "learning_rate": 8.338314877448447e-05, "loss": 0.0178, "step": 91180 }, { "grad_norm": 0.33662018179893494, "learning_rate": 8.337904541099643e-05, "loss": 0.0172, "step": 91190 }, { "grad_norm": 0.20810912549495697, "learning_rate": 8.337494164192149e-05, "loss": 0.0174, "step": 91200 }, { "grad_norm": 0.2092868685722351, "learning_rate": 8.33708374673095e-05, "loss": 0.0156, "step": 91210 }, { "grad_norm": 0.18334361910820007, "learning_rate": 8.336673288721031e-05, "loss": 0.0182, "step": 91220 }, { "grad_norm": 0.1377646028995514, "learning_rate": 8.336262790167383e-05, "loss": 0.0174, "step": 91230 }, { "grad_norm": 0.1824389398097992, "learning_rate": 8.335852251074992e-05, "loss": 0.0223, "step": 91240 }, { "grad_norm": 0.16482049226760864, "learning_rate": 8.335441671448845e-05, "loss": 0.0168, "step": 91250 }, { "grad_norm": 0.12393384426832199, "learning_rate": 8.335031051293933e-05, "loss": 0.0173, "step": 91260 }, { "grad_norm": 0.14614900946617126, "learning_rate": 8.334620390615246e-05, "loss": 0.0156, "step": 91270 }, { "grad_norm": 0.13315635919570923, "learning_rate": 8.334209689417773e-05, "loss": 0.0208, "step": 91280 }, { "grad_norm": 0.07577251642942429, "learning_rate": 8.333798947706504e-05, "loss": 0.0166, "step": 91290 }, { "grad_norm": 0.120452381670475, "learning_rate": 8.333388165486429e-05, "loss": 0.0172, "step": 91300 }, { "grad_norm": 0.17065951228141785, "learning_rate": 8.332977342762541e-05, "loss": 0.0164, "step": 91310 }, { "grad_norm": 0.1781887263059616, "learning_rate": 8.332566479539833e-05, "loss": 0.0184, "step": 91320 }, { "grad_norm": 0.11659174412488937, "learning_rate": 8.332155575823294e-05, "loss": 0.0201, "step": 91330 }, { "grad_norm": 0.15608443319797516, "learning_rate": 8.33174463161792e-05, "loss": 0.0186, "step": 91340 }, { "grad_norm": 0.28427743911743164, "learning_rate": 8.331333646928701e-05, "loss": 0.0186, "step": 91350 }, { "grad_norm": 0.11274076253175735, "learning_rate": 8.330922621760635e-05, "loss": 0.018, "step": 91360 }, { "grad_norm": 0.14581230282783508, "learning_rate": 8.330511556118713e-05, "loss": 0.0178, "step": 91370 }, { "grad_norm": 0.12821730971336365, "learning_rate": 8.330100450007931e-05, "loss": 0.0202, "step": 91380 }, { "grad_norm": 0.13613496720790863, "learning_rate": 8.329689303433285e-05, "loss": 0.0154, "step": 91390 }, { "grad_norm": 0.09717558324337006, "learning_rate": 8.32927811639977e-05, "loss": 0.0151, "step": 91400 }, { "grad_norm": 0.12419096380472183, "learning_rate": 8.328866888912384e-05, "loss": 0.0205, "step": 91410 }, { "grad_norm": 0.1538405865430832, "learning_rate": 8.32845562097612e-05, "loss": 0.0189, "step": 91420 }, { "grad_norm": 0.351915568113327, "learning_rate": 8.32804431259598e-05, "loss": 0.03, "step": 91430 }, { "grad_norm": 0.25253868103027344, "learning_rate": 8.327632963776956e-05, "loss": 0.0158, "step": 91440 }, { "grad_norm": 0.10491067916154861, "learning_rate": 8.327221574524053e-05, "loss": 0.0163, "step": 91450 }, { "grad_norm": 0.14112578332424164, "learning_rate": 8.326810144842265e-05, "loss": 0.0213, "step": 91460 }, { "grad_norm": 0.19974161684513092, "learning_rate": 8.326398674736592e-05, "loss": 0.0148, "step": 91470 }, { "grad_norm": 0.180871844291687, "learning_rate": 8.325987164212036e-05, "loss": 0.0213, "step": 91480 }, { "grad_norm": 0.28006288409233093, "learning_rate": 8.325575613273595e-05, "loss": 0.017, "step": 91490 }, { "grad_norm": 0.12847700715065002, "learning_rate": 8.32516402192627e-05, "loss": 0.0188, "step": 91500 }, { "grad_norm": 0.11553016304969788, "learning_rate": 8.324752390175063e-05, "loss": 0.0215, "step": 91510 }, { "grad_norm": 0.1361130028963089, "learning_rate": 8.324340718024977e-05, "loss": 0.018, "step": 91520 }, { "grad_norm": 0.10228218138217926, "learning_rate": 8.32392900548101e-05, "loss": 0.0163, "step": 91530 }, { "grad_norm": 0.20605547726154327, "learning_rate": 8.323517252548169e-05, "loss": 0.0137, "step": 91540 }, { "grad_norm": 0.1384933590888977, "learning_rate": 8.323105459231455e-05, "loss": 0.0199, "step": 91550 }, { "grad_norm": 0.1527438461780548, "learning_rate": 8.322693625535872e-05, "loss": 0.0214, "step": 91560 }, { "grad_norm": 0.14212076365947723, "learning_rate": 8.322281751466424e-05, "loss": 0.0147, "step": 91570 }, { "grad_norm": 0.13611586391925812, "learning_rate": 8.321869837028116e-05, "loss": 0.0226, "step": 91580 }, { "grad_norm": 0.1944645643234253, "learning_rate": 8.321457882225952e-05, "loss": 0.0153, "step": 91590 }, { "grad_norm": 0.09535910934209824, "learning_rate": 8.32104588706494e-05, "loss": 0.0206, "step": 91600 }, { "grad_norm": 0.16181285679340363, "learning_rate": 8.320633851550086e-05, "loss": 0.0209, "step": 91610 }, { "grad_norm": 0.18131853640079498, "learning_rate": 8.320221775686394e-05, "loss": 0.022, "step": 91620 }, { "grad_norm": 0.21014158427715302, "learning_rate": 8.319809659478873e-05, "loss": 0.0228, "step": 91630 }, { "grad_norm": 0.1654411256313324, "learning_rate": 8.31939750293253e-05, "loss": 0.0161, "step": 91640 }, { "grad_norm": 0.23625987768173218, "learning_rate": 8.318985306052374e-05, "loss": 0.0208, "step": 91650 }, { "grad_norm": 0.09834383428096771, "learning_rate": 8.318573068843413e-05, "loss": 0.0209, "step": 91660 }, { "grad_norm": 0.14850912988185883, "learning_rate": 8.318160791310653e-05, "loss": 0.0193, "step": 91670 }, { "grad_norm": 0.13745474815368652, "learning_rate": 8.317748473459111e-05, "loss": 0.0221, "step": 91680 }, { "grad_norm": 0.12213033437728882, "learning_rate": 8.31733611529379e-05, "loss": 0.0183, "step": 91690 }, { "grad_norm": 0.20872929692268372, "learning_rate": 8.316923716819704e-05, "loss": 0.0205, "step": 91700 }, { "grad_norm": 0.14761994779109955, "learning_rate": 8.316511278041862e-05, "loss": 0.0187, "step": 91710 }, { "grad_norm": 0.23287145793437958, "learning_rate": 8.316098798965279e-05, "loss": 0.0197, "step": 91720 }, { "grad_norm": 0.14019981026649475, "learning_rate": 8.315686279594964e-05, "loss": 0.0167, "step": 91730 }, { "grad_norm": 0.11335264891386032, "learning_rate": 8.315273719935928e-05, "loss": 0.015, "step": 91740 }, { "grad_norm": 0.1340082585811615, "learning_rate": 8.314861119993186e-05, "loss": 0.0152, "step": 91750 }, { "grad_norm": 0.2446603626012802, "learning_rate": 8.314448479771754e-05, "loss": 0.0199, "step": 91760 }, { "grad_norm": 0.1395225077867508, "learning_rate": 8.314035799276643e-05, "loss": 0.0145, "step": 91770 }, { "grad_norm": 0.14029912650585175, "learning_rate": 8.313623078512869e-05, "loss": 0.0195, "step": 91780 }, { "grad_norm": 0.16125153005123138, "learning_rate": 8.313210317485443e-05, "loss": 0.0226, "step": 91790 }, { "grad_norm": 0.13039393723011017, "learning_rate": 8.312797516199387e-05, "loss": 0.0155, "step": 91800 }, { "grad_norm": 0.11725015193223953, "learning_rate": 8.31238467465971e-05, "loss": 0.0172, "step": 91810 }, { "grad_norm": 0.16506531834602356, "learning_rate": 8.311971792871435e-05, "loss": 0.018, "step": 91820 }, { "grad_norm": 0.09546354413032532, "learning_rate": 8.311558870839573e-05, "loss": 0.0168, "step": 91830 }, { "grad_norm": 0.1801043152809143, "learning_rate": 8.311145908569147e-05, "loss": 0.0151, "step": 91840 }, { "grad_norm": 0.0966993048787117, "learning_rate": 8.310732906065168e-05, "loss": 0.0147, "step": 91850 }, { "grad_norm": 0.1629258692264557, "learning_rate": 8.310319863332662e-05, "loss": 0.018, "step": 91860 }, { "grad_norm": 0.1270003616809845, "learning_rate": 8.309906780376642e-05, "loss": 0.0176, "step": 91870 }, { "grad_norm": 0.12787339091300964, "learning_rate": 8.309493657202131e-05, "loss": 0.0158, "step": 91880 }, { "grad_norm": 0.3491702675819397, "learning_rate": 8.309080493814147e-05, "loss": 0.0238, "step": 91890 }, { "grad_norm": 0.20175960659980774, "learning_rate": 8.30866729021771e-05, "loss": 0.0193, "step": 91900 }, { "grad_norm": 0.12447132915258408, "learning_rate": 8.308254046417843e-05, "loss": 0.0154, "step": 91910 }, { "grad_norm": 0.11629170179367065, "learning_rate": 8.307840762419563e-05, "loss": 0.0186, "step": 91920 }, { "grad_norm": 0.1341092735528946, "learning_rate": 8.307427438227897e-05, "loss": 0.0171, "step": 91930 }, { "grad_norm": 0.11540468782186508, "learning_rate": 8.307014073847864e-05, "loss": 0.0163, "step": 91940 }, { "grad_norm": 0.16859270632266998, "learning_rate": 8.306600669284488e-05, "loss": 0.0241, "step": 91950 }, { "grad_norm": 0.2804662585258484, "learning_rate": 8.306187224542791e-05, "loss": 0.0214, "step": 91960 }, { "grad_norm": 0.24767932295799255, "learning_rate": 8.3057737396278e-05, "loss": 0.0169, "step": 91970 }, { "grad_norm": 0.23225687444210052, "learning_rate": 8.305360214544535e-05, "loss": 0.0201, "step": 91980 }, { "grad_norm": 0.13550223410129547, "learning_rate": 8.304946649298023e-05, "loss": 0.0214, "step": 91990 }, { "grad_norm": 0.2020355612039566, "learning_rate": 8.304533043893288e-05, "loss": 0.0187, "step": 92000 }, { "grad_norm": 0.2523353695869446, "learning_rate": 8.304119398335357e-05, "loss": 0.0182, "step": 92010 }, { "grad_norm": 0.1508183777332306, "learning_rate": 8.303705712629257e-05, "loss": 0.0177, "step": 92020 }, { "grad_norm": 0.19904448091983795, "learning_rate": 8.303291986780014e-05, "loss": 0.0211, "step": 92030 }, { "grad_norm": 0.14862778782844543, "learning_rate": 8.302878220792651e-05, "loss": 0.0188, "step": 92040 }, { "grad_norm": 0.12651078402996063, "learning_rate": 8.302464414672202e-05, "loss": 0.0137, "step": 92050 }, { "grad_norm": 0.25817960500717163, "learning_rate": 8.302050568423692e-05, "loss": 0.0189, "step": 92060 }, { "grad_norm": 0.16327399015426636, "learning_rate": 8.301636682052149e-05, "loss": 0.0252, "step": 92070 }, { "grad_norm": 0.2194640040397644, "learning_rate": 8.301222755562603e-05, "loss": 0.0187, "step": 92080 }, { "grad_norm": 0.1466691493988037, "learning_rate": 8.300808788960083e-05, "loss": 0.0192, "step": 92090 }, { "grad_norm": 0.13426917791366577, "learning_rate": 8.300394782249621e-05, "loss": 0.0216, "step": 92100 }, { "grad_norm": 0.18812616169452667, "learning_rate": 8.299980735436248e-05, "loss": 0.0174, "step": 92110 }, { "grad_norm": 0.08564213663339615, "learning_rate": 8.299566648524991e-05, "loss": 0.0179, "step": 92120 }, { "grad_norm": 0.08528655767440796, "learning_rate": 8.299152521520884e-05, "loss": 0.0128, "step": 92130 }, { "grad_norm": 0.20418094098567963, "learning_rate": 8.29873835442896e-05, "loss": 0.0187, "step": 92140 }, { "grad_norm": 0.1009918823838234, "learning_rate": 8.29832414725425e-05, "loss": 0.0179, "step": 92150 }, { "grad_norm": 0.09338368475437164, "learning_rate": 8.297909900001787e-05, "loss": 0.0189, "step": 92160 }, { "grad_norm": 0.12968909740447998, "learning_rate": 8.297495612676604e-05, "loss": 0.0191, "step": 92170 }, { "grad_norm": 0.12777452170848846, "learning_rate": 8.297081285283738e-05, "loss": 0.0212, "step": 92180 }, { "grad_norm": 0.19034312665462494, "learning_rate": 8.296666917828222e-05, "loss": 0.0291, "step": 92190 }, { "grad_norm": 0.18501651287078857, "learning_rate": 8.296252510315089e-05, "loss": 0.0183, "step": 92200 }, { "grad_norm": 0.9882037043571472, "learning_rate": 8.295838062749374e-05, "loss": 0.019, "step": 92210 }, { "grad_norm": 0.13697083294391632, "learning_rate": 8.295423575136117e-05, "loss": 0.0215, "step": 92220 }, { "grad_norm": 0.24932445585727692, "learning_rate": 8.295009047480354e-05, "loss": 0.0235, "step": 92230 }, { "grad_norm": 0.11100450903177261, "learning_rate": 8.294594479787116e-05, "loss": 0.0194, "step": 92240 }, { "grad_norm": 0.1261095404624939, "learning_rate": 8.294179872061447e-05, "loss": 0.02, "step": 92250 }, { "grad_norm": 0.14203181862831116, "learning_rate": 8.293765224308381e-05, "loss": 0.0153, "step": 92260 }, { "grad_norm": 0.08420731872320175, "learning_rate": 8.29335053653296e-05, "loss": 0.0231, "step": 92270 }, { "grad_norm": 0.24673332273960114, "learning_rate": 8.292935808740218e-05, "loss": 0.0174, "step": 92280 }, { "grad_norm": 0.1481637805700302, "learning_rate": 8.292521040935197e-05, "loss": 0.0159, "step": 92290 }, { "grad_norm": 0.12501388788223267, "learning_rate": 8.292106233122939e-05, "loss": 0.0176, "step": 92300 }, { "grad_norm": 0.21391278505325317, "learning_rate": 8.29169138530848e-05, "loss": 0.0273, "step": 92310 }, { "grad_norm": 0.21393166482448578, "learning_rate": 8.291276497496863e-05, "loss": 0.017, "step": 92320 }, { "grad_norm": 0.11444293707609177, "learning_rate": 8.290861569693129e-05, "loss": 0.014, "step": 92330 }, { "grad_norm": 0.15624457597732544, "learning_rate": 8.29044660190232e-05, "loss": 0.0138, "step": 92340 }, { "grad_norm": 0.1252920776605606, "learning_rate": 8.29003159412948e-05, "loss": 0.0163, "step": 92350 }, { "grad_norm": 0.13485665619373322, "learning_rate": 8.289616546379646e-05, "loss": 0.0181, "step": 92360 }, { "grad_norm": 0.17515797913074493, "learning_rate": 8.289201458657868e-05, "loss": 0.0206, "step": 92370 }, { "grad_norm": 0.1324046105146408, "learning_rate": 8.288786330969185e-05, "loss": 0.0154, "step": 92380 }, { "grad_norm": 0.13153399527072906, "learning_rate": 8.288371163318645e-05, "loss": 0.0173, "step": 92390 }, { "grad_norm": 0.17314910888671875, "learning_rate": 8.287955955711289e-05, "loss": 0.0193, "step": 92400 }, { "grad_norm": 0.3048243820667267, "learning_rate": 8.287540708152164e-05, "loss": 0.0225, "step": 92410 }, { "grad_norm": 0.3284614384174347, "learning_rate": 8.287125420646316e-05, "loss": 0.0193, "step": 92420 }, { "grad_norm": 0.14531169831752777, "learning_rate": 8.286710093198791e-05, "loss": 0.0177, "step": 92430 }, { "grad_norm": 0.2945728302001953, "learning_rate": 8.286294725814635e-05, "loss": 0.0198, "step": 92440 }, { "grad_norm": 0.14399296045303345, "learning_rate": 8.285879318498893e-05, "loss": 0.0216, "step": 92450 }, { "grad_norm": 0.10339146852493286, "learning_rate": 8.285463871256618e-05, "loss": 0.0175, "step": 92460 }, { "grad_norm": 0.7809372544288635, "learning_rate": 8.285048384092853e-05, "loss": 0.0178, "step": 92470 }, { "grad_norm": 0.1867409199476242, "learning_rate": 8.28463285701265e-05, "loss": 0.0245, "step": 92480 }, { "grad_norm": 0.13827913999557495, "learning_rate": 8.284217290021055e-05, "loss": 0.0198, "step": 92490 }, { "grad_norm": 0.13683080673217773, "learning_rate": 8.28380168312312e-05, "loss": 0.0174, "step": 92500 }, { "grad_norm": 0.09924260526895523, "learning_rate": 8.283386036323894e-05, "loss": 0.0136, "step": 92510 }, { "grad_norm": 0.2561062276363373, "learning_rate": 8.282970349628427e-05, "loss": 0.0167, "step": 92520 }, { "grad_norm": 0.24182580411434174, "learning_rate": 8.282554623041773e-05, "loss": 0.0146, "step": 92530 }, { "grad_norm": 0.21344022452831268, "learning_rate": 8.282138856568978e-05, "loss": 0.017, "step": 92540 }, { "grad_norm": 0.1042039692401886, "learning_rate": 8.2817230502151e-05, "loss": 0.0191, "step": 92550 }, { "grad_norm": 0.11954181641340256, "learning_rate": 8.281307203985186e-05, "loss": 0.0169, "step": 92560 }, { "grad_norm": 0.2037450522184372, "learning_rate": 8.280891317884291e-05, "loss": 0.0194, "step": 92570 }, { "grad_norm": 0.1270485669374466, "learning_rate": 8.28047539191747e-05, "loss": 0.017, "step": 92580 }, { "grad_norm": 0.6618634462356567, "learning_rate": 8.280059426089777e-05, "loss": 0.0244, "step": 92590 }, { "grad_norm": 0.1499238759279251, "learning_rate": 8.279643420406265e-05, "loss": 0.012, "step": 92600 }, { "grad_norm": 0.09281831979751587, "learning_rate": 8.279227374871988e-05, "loss": 0.0222, "step": 92610 }, { "grad_norm": 0.09644486755132675, "learning_rate": 8.278811289492001e-05, "loss": 0.0188, "step": 92620 }, { "grad_norm": 0.13760128617286682, "learning_rate": 8.278395164271362e-05, "loss": 0.0159, "step": 92630 }, { "grad_norm": 0.17574022710323334, "learning_rate": 8.277978999215127e-05, "loss": 0.0194, "step": 92640 }, { "grad_norm": 0.12542182207107544, "learning_rate": 8.277562794328352e-05, "loss": 0.0236, "step": 92650 }, { "grad_norm": 0.15238535404205322, "learning_rate": 8.277146549616094e-05, "loss": 0.0181, "step": 92660 }, { "grad_norm": 0.13523639738559723, "learning_rate": 8.276730265083412e-05, "loss": 0.0201, "step": 92670 }, { "grad_norm": 0.16366936266422272, "learning_rate": 8.276313940735363e-05, "loss": 0.0165, "step": 92680 }, { "grad_norm": 0.19551247358322144, "learning_rate": 8.275897576577007e-05, "loss": 0.0131, "step": 92690 }, { "grad_norm": 0.11592742055654526, "learning_rate": 8.275481172613402e-05, "loss": 0.0118, "step": 92700 }, { "grad_norm": 0.11911523342132568, "learning_rate": 8.275064728849608e-05, "loss": 0.0313, "step": 92710 }, { "grad_norm": 0.11684094369411469, "learning_rate": 8.274648245290686e-05, "loss": 0.0158, "step": 92720 }, { "grad_norm": 0.1676412969827652, "learning_rate": 8.274231721941695e-05, "loss": 0.0202, "step": 92730 }, { "grad_norm": 0.2563565671443939, "learning_rate": 8.273815158807699e-05, "loss": 0.0169, "step": 92740 }, { "grad_norm": 0.14557653665542603, "learning_rate": 8.273398555893756e-05, "loss": 0.0161, "step": 92750 }, { "grad_norm": 0.18203341960906982, "learning_rate": 8.272981913204931e-05, "loss": 0.0228, "step": 92760 }, { "grad_norm": 0.2818710505962372, "learning_rate": 8.272565230746284e-05, "loss": 0.0212, "step": 92770 }, { "grad_norm": 0.1929074078798294, "learning_rate": 8.272148508522882e-05, "loss": 0.0187, "step": 92780 }, { "grad_norm": 0.16987361013889313, "learning_rate": 8.271731746539784e-05, "loss": 0.0184, "step": 92790 }, { "grad_norm": 0.3561716079711914, "learning_rate": 8.271314944802057e-05, "loss": 0.0189, "step": 92800 }, { "grad_norm": 0.1394377499818802, "learning_rate": 8.270898103314764e-05, "loss": 0.0168, "step": 92810 }, { "grad_norm": 0.19831959903240204, "learning_rate": 8.270481222082974e-05, "loss": 0.0195, "step": 92820 }, { "grad_norm": 0.19020316004753113, "learning_rate": 8.270064301111746e-05, "loss": 0.0174, "step": 92830 }, { "grad_norm": 0.20822454988956451, "learning_rate": 8.26964734040615e-05, "loss": 0.0231, "step": 92840 }, { "grad_norm": 0.13295577466487885, "learning_rate": 8.269230339971253e-05, "loss": 0.0209, "step": 92850 }, { "grad_norm": 0.11281085759401321, "learning_rate": 8.268813299812119e-05, "loss": 0.0163, "step": 92860 }, { "grad_norm": 0.1905255913734436, "learning_rate": 8.268396219933819e-05, "loss": 0.0201, "step": 92870 }, { "grad_norm": 0.1479039192199707, "learning_rate": 8.267979100341418e-05, "loss": 0.0199, "step": 92880 }, { "grad_norm": 0.24443235993385315, "learning_rate": 8.267561941039985e-05, "loss": 0.0249, "step": 92890 }, { "grad_norm": 0.13036295771598816, "learning_rate": 8.26714474203459e-05, "loss": 0.0201, "step": 92900 }, { "grad_norm": 0.19913391768932343, "learning_rate": 8.266727503330302e-05, "loss": 0.0216, "step": 92910 }, { "grad_norm": 0.44493716955184937, "learning_rate": 8.26631022493219e-05, "loss": 0.0196, "step": 92920 }, { "grad_norm": 0.1406840980052948, "learning_rate": 8.265892906845325e-05, "loss": 0.0232, "step": 92930 }, { "grad_norm": 0.13258405029773712, "learning_rate": 8.265475549074777e-05, "loss": 0.0171, "step": 92940 }, { "grad_norm": 0.10534311085939407, "learning_rate": 8.265058151625617e-05, "loss": 0.0166, "step": 92950 }, { "grad_norm": 0.2002236545085907, "learning_rate": 8.26464071450292e-05, "loss": 0.0229, "step": 92960 }, { "grad_norm": 0.13218975067138672, "learning_rate": 8.264223237711756e-05, "loss": 0.0221, "step": 92970 }, { "grad_norm": 0.1302231103181839, "learning_rate": 8.263805721257198e-05, "loss": 0.0199, "step": 92980 }, { "grad_norm": 0.571414589881897, "learning_rate": 8.263388165144317e-05, "loss": 0.0157, "step": 92990 }, { "grad_norm": 0.24020439386367798, "learning_rate": 8.26297056937819e-05, "loss": 0.0213, "step": 93000 }, { "grad_norm": 0.12260981649160385, "learning_rate": 8.26255293396389e-05, "loss": 0.0142, "step": 93010 }, { "grad_norm": 0.1577349156141281, "learning_rate": 8.26213525890649e-05, "loss": 0.0166, "step": 93020 }, { "grad_norm": 0.13649646937847137, "learning_rate": 8.261717544211067e-05, "loss": 0.0219, "step": 93030 }, { "grad_norm": 0.17444904148578644, "learning_rate": 8.261299789882697e-05, "loss": 0.0158, "step": 93040 }, { "grad_norm": 0.1653943806886673, "learning_rate": 8.260881995926455e-05, "loss": 0.0164, "step": 93050 }, { "grad_norm": 0.32914599776268005, "learning_rate": 8.260464162347418e-05, "loss": 0.0193, "step": 93060 }, { "grad_norm": 0.1507880538702011, "learning_rate": 8.260046289150662e-05, "loss": 0.0192, "step": 93070 }, { "grad_norm": 0.16440193355083466, "learning_rate": 8.259628376341269e-05, "loss": 0.0172, "step": 93080 }, { "grad_norm": 0.1455906629562378, "learning_rate": 8.25921042392431e-05, "loss": 0.0152, "step": 93090 }, { "grad_norm": 0.10241580754518509, "learning_rate": 8.25879243190487e-05, "loss": 0.0196, "step": 93100 }, { "grad_norm": 0.11204005777835846, "learning_rate": 8.258374400288022e-05, "loss": 0.0136, "step": 93110 }, { "grad_norm": 0.13468395173549652, "learning_rate": 8.257956329078851e-05, "loss": 0.0179, "step": 93120 }, { "grad_norm": 0.5039673447608948, "learning_rate": 8.257538218282433e-05, "loss": 0.017, "step": 93130 }, { "grad_norm": 0.15334805846214294, "learning_rate": 8.257120067903852e-05, "loss": 0.018, "step": 93140 }, { "grad_norm": 0.095217764377594, "learning_rate": 8.256701877948185e-05, "loss": 0.0187, "step": 93150 }, { "grad_norm": 0.1947375386953354, "learning_rate": 8.256283648420516e-05, "loss": 0.0207, "step": 93160 }, { "grad_norm": 0.12464521080255508, "learning_rate": 8.255865379325927e-05, "loss": 0.0188, "step": 93170 }, { "grad_norm": 0.21679723262786865, "learning_rate": 8.2554470706695e-05, "loss": 0.0194, "step": 93180 }, { "grad_norm": 0.13505978882312775, "learning_rate": 8.255028722456314e-05, "loss": 0.0169, "step": 93190 }, { "grad_norm": 0.11302081495523453, "learning_rate": 8.25461033469146e-05, "loss": 0.0131, "step": 93200 }, { "grad_norm": 0.18334145843982697, "learning_rate": 8.254191907380016e-05, "loss": 0.0204, "step": 93210 }, { "grad_norm": 0.10505029559135437, "learning_rate": 8.253773440527066e-05, "loss": 0.0135, "step": 93220 }, { "grad_norm": 0.13767236471176147, "learning_rate": 8.2533549341377e-05, "loss": 0.0183, "step": 93230 }, { "grad_norm": 0.12395305186510086, "learning_rate": 8.252936388216998e-05, "loss": 0.0157, "step": 93240 }, { "grad_norm": 0.10559185594320297, "learning_rate": 8.252517802770048e-05, "loss": 0.0203, "step": 93250 }, { "grad_norm": 0.14090381562709808, "learning_rate": 8.252099177801935e-05, "loss": 0.0179, "step": 93260 }, { "grad_norm": 0.1521911472082138, "learning_rate": 8.251680513317746e-05, "loss": 0.0203, "step": 93270 }, { "grad_norm": 0.14240053296089172, "learning_rate": 8.25126180932257e-05, "loss": 0.022, "step": 93280 }, { "grad_norm": 0.215646430850029, "learning_rate": 8.250843065821494e-05, "loss": 0.0172, "step": 93290 }, { "grad_norm": 0.23158583045005798, "learning_rate": 8.250424282819605e-05, "loss": 0.0194, "step": 93300 }, { "grad_norm": 0.15366636216640472, "learning_rate": 8.25000546032199e-05, "loss": 0.0168, "step": 93310 }, { "grad_norm": 0.14553610980510712, "learning_rate": 8.249586598333742e-05, "loss": 0.0178, "step": 93320 }, { "grad_norm": 0.16951356828212738, "learning_rate": 8.24916769685995e-05, "loss": 0.0208, "step": 93330 }, { "grad_norm": 0.1430576741695404, "learning_rate": 8.2487487559057e-05, "loss": 0.0175, "step": 93340 }, { "grad_norm": 0.1831742227077484, "learning_rate": 8.248329775476087e-05, "loss": 0.0203, "step": 93350 }, { "grad_norm": 0.2007676661014557, "learning_rate": 8.2479107555762e-05, "loss": 0.0159, "step": 93360 }, { "grad_norm": 0.19361139833927155, "learning_rate": 8.247491696211129e-05, "loss": 0.0183, "step": 93370 }, { "grad_norm": 0.12189988046884537, "learning_rate": 8.24707259738597e-05, "loss": 0.013, "step": 93380 }, { "grad_norm": 0.257569283246994, "learning_rate": 8.246653459105813e-05, "loss": 0.0153, "step": 93390 }, { "grad_norm": 0.1990320384502411, "learning_rate": 8.246234281375751e-05, "loss": 0.0204, "step": 93400 }, { "grad_norm": 0.15951502323150635, "learning_rate": 8.245815064200877e-05, "loss": 0.0214, "step": 93410 }, { "grad_norm": 0.11533740907907486, "learning_rate": 8.245395807586284e-05, "loss": 0.0227, "step": 93420 }, { "grad_norm": 0.3766331672668457, "learning_rate": 8.24497651153707e-05, "loss": 0.0174, "step": 93430 }, { "grad_norm": 0.19390343129634857, "learning_rate": 8.244557176058327e-05, "loss": 0.0203, "step": 93440 }, { "grad_norm": 0.3811660408973694, "learning_rate": 8.244137801155151e-05, "loss": 0.0203, "step": 93450 }, { "grad_norm": 0.13566280901432037, "learning_rate": 8.243718386832638e-05, "loss": 0.0224, "step": 93460 }, { "grad_norm": 0.23952583968639374, "learning_rate": 8.243298933095883e-05, "loss": 0.0208, "step": 93470 }, { "grad_norm": 0.13634029030799866, "learning_rate": 8.242879439949984e-05, "loss": 0.0176, "step": 93480 }, { "grad_norm": 0.07747291773557663, "learning_rate": 8.242459907400039e-05, "loss": 0.0163, "step": 93490 }, { "grad_norm": 0.14017586410045624, "learning_rate": 8.242040335451144e-05, "loss": 0.0221, "step": 93500 }, { "grad_norm": 0.1252204328775406, "learning_rate": 8.241620724108399e-05, "loss": 0.0156, "step": 93510 }, { "grad_norm": 0.11731743812561035, "learning_rate": 8.2412010733769e-05, "loss": 0.0221, "step": 93520 }, { "grad_norm": 0.24504956603050232, "learning_rate": 8.240781383261749e-05, "loss": 0.0227, "step": 93530 }, { "grad_norm": 0.39986804127693176, "learning_rate": 8.240361653768042e-05, "loss": 0.0183, "step": 93540 }, { "grad_norm": 0.2268902212381363, "learning_rate": 8.239941884900884e-05, "loss": 0.02, "step": 93550 }, { "grad_norm": 0.1718139797449112, "learning_rate": 8.239522076665374e-05, "loss": 0.0261, "step": 93560 }, { "grad_norm": 0.27259689569473267, "learning_rate": 8.23910222906661e-05, "loss": 0.0239, "step": 93570 }, { "grad_norm": 0.2103809267282486, "learning_rate": 8.238682342109695e-05, "loss": 0.0193, "step": 93580 }, { "grad_norm": 0.10791011154651642, "learning_rate": 8.238262415799734e-05, "loss": 0.0172, "step": 93590 }, { "grad_norm": 0.1260710209608078, "learning_rate": 8.237842450141826e-05, "loss": 0.0178, "step": 93600 }, { "grad_norm": 0.14093734323978424, "learning_rate": 8.237422445141075e-05, "loss": 0.0235, "step": 93610 }, { "grad_norm": 0.18511824309825897, "learning_rate": 8.237002400802586e-05, "loss": 0.017, "step": 93620 }, { "grad_norm": 0.11262056231498718, "learning_rate": 8.23658231713146e-05, "loss": 0.016, "step": 93630 }, { "grad_norm": 0.1845521777868271, "learning_rate": 8.236162194132803e-05, "loss": 0.0177, "step": 93640 }, { "grad_norm": 0.2361632138490677, "learning_rate": 8.235742031811722e-05, "loss": 0.0159, "step": 93650 }, { "grad_norm": 0.34118837118148804, "learning_rate": 8.235321830173318e-05, "loss": 0.0162, "step": 93660 }, { "grad_norm": 0.14221912622451782, "learning_rate": 8.234901589222701e-05, "loss": 0.0115, "step": 93670 }, { "grad_norm": 0.1405843049287796, "learning_rate": 8.234481308964974e-05, "loss": 0.017, "step": 93680 }, { "grad_norm": 0.1322680115699768, "learning_rate": 8.234060989405246e-05, "loss": 0.0182, "step": 93690 }, { "grad_norm": 0.1502198725938797, "learning_rate": 8.233640630548622e-05, "loss": 0.0153, "step": 93700 }, { "grad_norm": 0.2549576759338379, "learning_rate": 8.233220232400214e-05, "loss": 0.0126, "step": 93710 }, { "grad_norm": 0.073760025203228, "learning_rate": 8.232799794965125e-05, "loss": 0.0171, "step": 93720 }, { "grad_norm": 0.14327804744243622, "learning_rate": 8.232379318248466e-05, "loss": 0.0147, "step": 93730 }, { "grad_norm": 0.10500077158212662, "learning_rate": 8.231958802255348e-05, "loss": 0.0165, "step": 93740 }, { "grad_norm": 0.11528622359037399, "learning_rate": 8.23153824699088e-05, "loss": 0.0166, "step": 93750 }, { "grad_norm": 0.19957812130451202, "learning_rate": 8.231117652460169e-05, "loss": 0.0182, "step": 93760 }, { "grad_norm": 0.10259387642145157, "learning_rate": 8.230697018668329e-05, "loss": 0.0176, "step": 93770 }, { "grad_norm": 0.12851771712303162, "learning_rate": 8.23027634562047e-05, "loss": 0.0198, "step": 93780 }, { "grad_norm": 0.1549951583147049, "learning_rate": 8.229855633321702e-05, "loss": 0.0155, "step": 93790 }, { "grad_norm": 0.1739400327205658, "learning_rate": 8.22943488177714e-05, "loss": 0.0191, "step": 93800 }, { "grad_norm": 0.0865427628159523, "learning_rate": 8.229014090991896e-05, "loss": 0.0175, "step": 93810 }, { "grad_norm": 0.1280565708875656, "learning_rate": 8.22859326097108e-05, "loss": 0.0221, "step": 93820 }, { "grad_norm": 0.14816801249980927, "learning_rate": 8.228172391719809e-05, "loss": 0.0147, "step": 93830 }, { "grad_norm": 0.1361987441778183, "learning_rate": 8.227751483243195e-05, "loss": 0.013, "step": 93840 }, { "grad_norm": 0.19688406586647034, "learning_rate": 8.227330535546353e-05, "loss": 0.0171, "step": 93850 }, { "grad_norm": 0.12589134275913239, "learning_rate": 8.226909548634398e-05, "loss": 0.0138, "step": 93860 }, { "grad_norm": 0.1870233118534088, "learning_rate": 8.226488522512446e-05, "loss": 0.0211, "step": 93870 }, { "grad_norm": 0.1959398239850998, "learning_rate": 8.226067457185611e-05, "loss": 0.0203, "step": 93880 }, { "grad_norm": 0.22412672638893127, "learning_rate": 8.225646352659011e-05, "loss": 0.0166, "step": 93890 }, { "grad_norm": 0.21721014380455017, "learning_rate": 8.225225208937761e-05, "loss": 0.0202, "step": 93900 }, { "grad_norm": 0.1306368112564087, "learning_rate": 8.224804026026981e-05, "loss": 0.0161, "step": 93910 }, { "grad_norm": 0.13037075102329254, "learning_rate": 8.224382803931788e-05, "loss": 0.0203, "step": 93920 }, { "grad_norm": 0.11214586347341537, "learning_rate": 8.223961542657297e-05, "loss": 0.0196, "step": 93930 }, { "grad_norm": 0.11522168666124344, "learning_rate": 8.223540242208631e-05, "loss": 0.0122, "step": 93940 }, { "grad_norm": 0.2438928782939911, "learning_rate": 8.223118902590908e-05, "loss": 0.0229, "step": 93950 }, { "grad_norm": 0.1485014706850052, "learning_rate": 8.222697523809248e-05, "loss": 0.0175, "step": 93960 }, { "grad_norm": 0.13575682044029236, "learning_rate": 8.22227610586877e-05, "loss": 0.022, "step": 93970 }, { "grad_norm": 0.2776731252670288, "learning_rate": 8.221854648774592e-05, "loss": 0.0229, "step": 93980 }, { "grad_norm": 0.11678196489810944, "learning_rate": 8.221433152531841e-05, "loss": 0.0157, "step": 93990 }, { "grad_norm": 0.10430468618869781, "learning_rate": 8.221011617145635e-05, "loss": 0.0169, "step": 94000 }, { "grad_norm": 0.3170519769191742, "learning_rate": 8.220590042621097e-05, "loss": 0.0182, "step": 94010 }, { "grad_norm": 0.13713791966438293, "learning_rate": 8.220168428963348e-05, "loss": 0.0184, "step": 94020 }, { "grad_norm": 0.17165954411029816, "learning_rate": 8.219746776177513e-05, "loss": 0.015, "step": 94030 }, { "grad_norm": 0.09905490279197693, "learning_rate": 8.219325084268715e-05, "loss": 0.0148, "step": 94040 }, { "grad_norm": 0.5000123977661133, "learning_rate": 8.218903353242077e-05, "loss": 0.0238, "step": 94050 }, { "grad_norm": 0.18172964453697205, "learning_rate": 8.218481583102726e-05, "loss": 0.0176, "step": 94060 }, { "grad_norm": 0.15095840394496918, "learning_rate": 8.218059773855782e-05, "loss": 0.0236, "step": 94070 }, { "grad_norm": 0.12405566871166229, "learning_rate": 8.217637925506376e-05, "loss": 0.015, "step": 94080 }, { "grad_norm": 0.12302054464817047, "learning_rate": 8.21721603805963e-05, "loss": 0.0143, "step": 94090 }, { "grad_norm": 0.14251339435577393, "learning_rate": 8.216794111520671e-05, "loss": 0.0176, "step": 94100 }, { "grad_norm": 0.24696169793605804, "learning_rate": 8.216372145894627e-05, "loss": 0.0204, "step": 94110 }, { "grad_norm": 0.20410974323749542, "learning_rate": 8.215950141186626e-05, "loss": 0.02, "step": 94120 }, { "grad_norm": 0.3280445337295532, "learning_rate": 8.215528097401793e-05, "loss": 0.0182, "step": 94130 }, { "grad_norm": 0.1253565102815628, "learning_rate": 8.215106014545258e-05, "loss": 0.0122, "step": 94140 }, { "grad_norm": 0.12024349719285965, "learning_rate": 8.21468389262215e-05, "loss": 0.0197, "step": 94150 }, { "grad_norm": 0.2264428436756134, "learning_rate": 8.214261731637596e-05, "loss": 0.0236, "step": 94160 }, { "grad_norm": 0.246934175491333, "learning_rate": 8.213839531596728e-05, "loss": 0.019, "step": 94170 }, { "grad_norm": 0.18491888046264648, "learning_rate": 8.213417292504674e-05, "loss": 0.0221, "step": 94180 }, { "grad_norm": 0.17105160653591156, "learning_rate": 8.212995014366569e-05, "loss": 0.0162, "step": 94190 }, { "grad_norm": 0.12866812944412231, "learning_rate": 8.212572697187539e-05, "loss": 0.0232, "step": 94200 }, { "grad_norm": 0.17391037940979004, "learning_rate": 8.212150340972717e-05, "loss": 0.0186, "step": 94210 }, { "grad_norm": 0.13011963665485382, "learning_rate": 8.211727945727237e-05, "loss": 0.0178, "step": 94220 }, { "grad_norm": 0.09219653904438019, "learning_rate": 8.211305511456229e-05, "loss": 0.0195, "step": 94230 }, { "grad_norm": 0.2589029371738434, "learning_rate": 8.210883038164828e-05, "loss": 0.0222, "step": 94240 }, { "grad_norm": 0.34236201643943787, "learning_rate": 8.210460525858166e-05, "loss": 0.0187, "step": 94250 }, { "grad_norm": 0.10638836771249771, "learning_rate": 8.210037974541377e-05, "loss": 0.0158, "step": 94260 }, { "grad_norm": 0.1614103615283966, "learning_rate": 8.209615384219596e-05, "loss": 0.0148, "step": 94270 }, { "grad_norm": 0.14858511090278625, "learning_rate": 8.209192754897958e-05, "loss": 0.0151, "step": 94280 }, { "grad_norm": 0.45986101031303406, "learning_rate": 8.2087700865816e-05, "loss": 0.0172, "step": 94290 }, { "grad_norm": 0.21414920687675476, "learning_rate": 8.208347379275654e-05, "loss": 0.024, "step": 94300 }, { "grad_norm": 0.20775428414344788, "learning_rate": 8.207924632985258e-05, "loss": 0.0165, "step": 94310 }, { "grad_norm": 0.11600753664970398, "learning_rate": 8.207501847715549e-05, "loss": 0.0122, "step": 94320 }, { "grad_norm": 0.2994750440120697, "learning_rate": 8.207079023471664e-05, "loss": 0.0192, "step": 94330 }, { "grad_norm": 0.15017017722129822, "learning_rate": 8.206656160258742e-05, "loss": 0.0193, "step": 94340 }, { "grad_norm": 0.11026088893413544, "learning_rate": 8.206233258081918e-05, "loss": 0.0169, "step": 94350 }, { "grad_norm": 0.16155675053596497, "learning_rate": 8.205810316946334e-05, "loss": 0.0211, "step": 94360 }, { "grad_norm": 0.20813898742198944, "learning_rate": 8.205387336857128e-05, "loss": 0.0179, "step": 94370 }, { "grad_norm": 0.23429255187511444, "learning_rate": 8.204964317819439e-05, "loss": 0.0184, "step": 94380 }, { "grad_norm": 0.09335491061210632, "learning_rate": 8.204541259838408e-05, "loss": 0.0154, "step": 94390 }, { "grad_norm": 0.19091811776161194, "learning_rate": 8.204118162919174e-05, "loss": 0.0141, "step": 94400 }, { "grad_norm": 0.14759421348571777, "learning_rate": 8.20369502706688e-05, "loss": 0.0189, "step": 94410 }, { "grad_norm": 0.09496694058179855, "learning_rate": 8.203271852286667e-05, "loss": 0.0186, "step": 94420 }, { "grad_norm": 0.1886414736509323, "learning_rate": 8.202848638583675e-05, "loss": 0.0164, "step": 94430 }, { "grad_norm": 0.17363294959068298, "learning_rate": 8.20242538596305e-05, "loss": 0.0178, "step": 94440 }, { "grad_norm": 0.11884216964244843, "learning_rate": 8.202002094429933e-05, "loss": 0.0174, "step": 94450 }, { "grad_norm": 0.18938829004764557, "learning_rate": 8.201578763989466e-05, "loss": 0.0183, "step": 94460 }, { "grad_norm": 0.49065932631492615, "learning_rate": 8.201155394646796e-05, "loss": 0.0166, "step": 94470 }, { "grad_norm": 0.11439521610736847, "learning_rate": 8.200731986407064e-05, "loss": 0.0149, "step": 94480 }, { "grad_norm": 0.18914532661437988, "learning_rate": 8.200308539275417e-05, "loss": 0.0207, "step": 94490 }, { "grad_norm": 0.17030665278434753, "learning_rate": 8.199885053256999e-05, "loss": 0.0175, "step": 94500 }, { "grad_norm": 0.22363494336605072, "learning_rate": 8.199461528356958e-05, "loss": 0.0177, "step": 94510 }, { "grad_norm": 0.17294102907180786, "learning_rate": 8.199037964580436e-05, "loss": 0.0193, "step": 94520 }, { "grad_norm": 0.1452796310186386, "learning_rate": 8.198614361932583e-05, "loss": 0.0154, "step": 94530 }, { "grad_norm": 0.15229150652885437, "learning_rate": 8.198190720418549e-05, "loss": 0.0214, "step": 94540 }, { "grad_norm": 0.1519497036933899, "learning_rate": 8.197767040043474e-05, "loss": 0.0217, "step": 94550 }, { "grad_norm": 0.10960084944963455, "learning_rate": 8.197343320812512e-05, "loss": 0.0249, "step": 94560 }, { "grad_norm": 0.2227003425359726, "learning_rate": 8.19691956273081e-05, "loss": 0.0186, "step": 94570 }, { "grad_norm": 0.1811753809452057, "learning_rate": 8.196495765803515e-05, "loss": 0.0163, "step": 94580 }, { "grad_norm": 0.15727443993091583, "learning_rate": 8.19607193003578e-05, "loss": 0.0196, "step": 94590 }, { "grad_norm": 0.10358268767595291, "learning_rate": 8.195648055432755e-05, "loss": 0.0199, "step": 94600 }, { "grad_norm": 0.15039785206317902, "learning_rate": 8.195224141999587e-05, "loss": 0.0181, "step": 94610 }, { "grad_norm": 0.17952430248260498, "learning_rate": 8.19480018974143e-05, "loss": 0.0283, "step": 94620 }, { "grad_norm": 0.17617151141166687, "learning_rate": 8.194376198663434e-05, "loss": 0.0198, "step": 94630 }, { "grad_norm": 0.12803936004638672, "learning_rate": 8.19395216877075e-05, "loss": 0.0173, "step": 94640 }, { "grad_norm": 0.18269655108451843, "learning_rate": 8.193528100068532e-05, "loss": 0.0171, "step": 94650 }, { "grad_norm": 0.12709330022335052, "learning_rate": 8.193103992561933e-05, "loss": 0.0205, "step": 94660 }, { "grad_norm": 0.11794568598270416, "learning_rate": 8.192679846256106e-05, "loss": 0.0156, "step": 94670 }, { "grad_norm": 0.1531306654214859, "learning_rate": 8.192255661156202e-05, "loss": 0.0152, "step": 94680 }, { "grad_norm": 0.07091566175222397, "learning_rate": 8.19183143726738e-05, "loss": 0.0177, "step": 94690 }, { "grad_norm": 0.1189584732055664, "learning_rate": 8.191407174594794e-05, "loss": 0.022, "step": 94700 }, { "grad_norm": 0.13609763979911804, "learning_rate": 8.190982873143594e-05, "loss": 0.0172, "step": 94710 }, { "grad_norm": 0.2868569791316986, "learning_rate": 8.190558532918941e-05, "loss": 0.0187, "step": 94720 }, { "grad_norm": 0.16114406287670135, "learning_rate": 8.19013415392599e-05, "loss": 0.0185, "step": 94730 }, { "grad_norm": 0.12013813108205795, "learning_rate": 8.189709736169898e-05, "loss": 0.0179, "step": 94740 }, { "grad_norm": 0.14270484447479248, "learning_rate": 8.189285279655818e-05, "loss": 0.0182, "step": 94750 }, { "grad_norm": 0.14969255030155182, "learning_rate": 8.188860784388913e-05, "loss": 0.0165, "step": 94760 }, { "grad_norm": 0.192628413438797, "learning_rate": 8.188436250374339e-05, "loss": 0.0217, "step": 94770 }, { "grad_norm": 0.21614983677864075, "learning_rate": 8.188011677617251e-05, "loss": 0.019, "step": 94780 }, { "grad_norm": 0.1950734555721283, "learning_rate": 8.187587066122814e-05, "loss": 0.0177, "step": 94790 }, { "grad_norm": 0.12657827138900757, "learning_rate": 8.187162415896184e-05, "loss": 0.0149, "step": 94800 }, { "grad_norm": 0.16070681810379028, "learning_rate": 8.186737726942522e-05, "loss": 0.0164, "step": 94810 }, { "grad_norm": 0.12516677379608154, "learning_rate": 8.186312999266987e-05, "loss": 0.0187, "step": 94820 }, { "grad_norm": 0.09764726459980011, "learning_rate": 8.18588823287474e-05, "loss": 0.0127, "step": 94830 }, { "grad_norm": 0.11489815264940262, "learning_rate": 8.185463427770944e-05, "loss": 0.0176, "step": 94840 }, { "grad_norm": 0.12811896204948425, "learning_rate": 8.18503858396076e-05, "loss": 0.0189, "step": 94850 }, { "grad_norm": 0.12301433831453323, "learning_rate": 8.184613701449349e-05, "loss": 0.0169, "step": 94860 }, { "grad_norm": 0.20848219096660614, "learning_rate": 8.184188780241876e-05, "loss": 0.0207, "step": 94870 }, { "grad_norm": 0.1520564705133438, "learning_rate": 8.183763820343501e-05, "loss": 0.0162, "step": 94880 }, { "grad_norm": 0.11959702521562576, "learning_rate": 8.183338821759392e-05, "loss": 0.0432, "step": 94890 }, { "grad_norm": 0.18850177526474, "learning_rate": 8.18291378449471e-05, "loss": 0.0154, "step": 94900 }, { "grad_norm": 0.13653959333896637, "learning_rate": 8.18248870855462e-05, "loss": 0.0226, "step": 94910 }, { "grad_norm": 0.11029843986034393, "learning_rate": 8.182063593944287e-05, "loss": 0.0156, "step": 94920 }, { "grad_norm": 0.10919865220785141, "learning_rate": 8.181638440668879e-05, "loss": 0.0276, "step": 94930 }, { "grad_norm": 0.11732134222984314, "learning_rate": 8.181213248733556e-05, "loss": 0.0156, "step": 94940 }, { "grad_norm": 0.18385756015777588, "learning_rate": 8.180788018143492e-05, "loss": 0.0232, "step": 94950 }, { "grad_norm": 0.1635207086801529, "learning_rate": 8.180362748903849e-05, "loss": 0.0248, "step": 94960 }, { "grad_norm": 0.06988440454006195, "learning_rate": 8.179937441019798e-05, "loss": 0.0177, "step": 94970 }, { "grad_norm": 0.148172065615654, "learning_rate": 8.179512094496502e-05, "loss": 0.0177, "step": 94980 }, { "grad_norm": 0.13199666142463684, "learning_rate": 8.179086709339132e-05, "loss": 0.0183, "step": 94990 }, { "grad_norm": 0.1453758180141449, "learning_rate": 8.178661285552858e-05, "loss": 0.0233, "step": 95000 }, { "grad_norm": 0.15252310037612915, "learning_rate": 8.178235823142849e-05, "loss": 0.0201, "step": 95010 }, { "grad_norm": 0.11872158199548721, "learning_rate": 8.177810322114274e-05, "loss": 0.0136, "step": 95020 }, { "grad_norm": 0.16964609920978546, "learning_rate": 8.177384782472302e-05, "loss": 0.0238, "step": 95030 }, { "grad_norm": 0.15847735106945038, "learning_rate": 8.176959204222106e-05, "loss": 0.0174, "step": 95040 }, { "grad_norm": 0.14419962465763092, "learning_rate": 8.176533587368854e-05, "loss": 0.0184, "step": 95050 }, { "grad_norm": 0.12758436799049377, "learning_rate": 8.176107931917721e-05, "loss": 0.0276, "step": 95060 }, { "grad_norm": 0.14215153455734253, "learning_rate": 8.175682237873878e-05, "loss": 0.0182, "step": 95070 }, { "grad_norm": 0.1044856533408165, "learning_rate": 8.175256505242497e-05, "loss": 0.0127, "step": 95080 }, { "grad_norm": 0.09649238735437393, "learning_rate": 8.174830734028753e-05, "loss": 0.015, "step": 95090 }, { "grad_norm": 0.10553701967000961, "learning_rate": 8.174404924237817e-05, "loss": 0.0166, "step": 95100 }, { "grad_norm": 0.4476751685142517, "learning_rate": 8.173979075874864e-05, "loss": 0.0168, "step": 95110 }, { "grad_norm": 0.1322060078382492, "learning_rate": 8.173553188945069e-05, "loss": 0.0188, "step": 95120 }, { "grad_norm": 0.23362553119659424, "learning_rate": 8.173127263453606e-05, "loss": 0.0164, "step": 95130 }, { "grad_norm": 0.19459135830402374, "learning_rate": 8.172701299405651e-05, "loss": 0.0201, "step": 95140 }, { "grad_norm": 0.14207522571086884, "learning_rate": 8.17227529680638e-05, "loss": 0.0192, "step": 95150 }, { "grad_norm": 0.13891322910785675, "learning_rate": 8.171849255660968e-05, "loss": 0.0174, "step": 95160 }, { "grad_norm": 0.15856759250164032, "learning_rate": 8.171423175974594e-05, "loss": 0.0199, "step": 95170 }, { "grad_norm": 0.24481616914272308, "learning_rate": 8.170997057752434e-05, "loss": 0.024, "step": 95180 }, { "grad_norm": 0.12694521248340607, "learning_rate": 8.170570900999665e-05, "loss": 0.0189, "step": 95190 }, { "grad_norm": 0.869465708732605, "learning_rate": 8.170144705721465e-05, "loss": 0.0197, "step": 95200 }, { "grad_norm": 0.12513066828250885, "learning_rate": 8.169718471923016e-05, "loss": 0.0197, "step": 95210 }, { "grad_norm": 0.215819850564003, "learning_rate": 8.169292199609494e-05, "loss": 0.0194, "step": 95220 }, { "grad_norm": 0.2272675782442093, "learning_rate": 8.16886588878608e-05, "loss": 0.0265, "step": 95230 }, { "grad_norm": 0.2897655963897705, "learning_rate": 8.168439539457953e-05, "loss": 0.0219, "step": 95240 }, { "grad_norm": 0.1557639241218567, "learning_rate": 8.168013151630294e-05, "loss": 0.023, "step": 95250 }, { "grad_norm": 0.18321868777275085, "learning_rate": 8.167586725308283e-05, "loss": 0.0208, "step": 95260 }, { "grad_norm": 0.122048020362854, "learning_rate": 8.167160260497105e-05, "loss": 0.0169, "step": 95270 }, { "grad_norm": 0.21661080420017242, "learning_rate": 8.166733757201937e-05, "loss": 0.0198, "step": 95280 }, { "grad_norm": 0.6177728176116943, "learning_rate": 8.166307215427967e-05, "loss": 0.0196, "step": 95290 }, { "grad_norm": 0.14045952260494232, "learning_rate": 8.165880635180372e-05, "loss": 0.0179, "step": 95300 }, { "grad_norm": 0.07520271837711334, "learning_rate": 8.16545401646434e-05, "loss": 0.0181, "step": 95310 }, { "grad_norm": 0.12577515840530396, "learning_rate": 8.165027359285052e-05, "loss": 0.018, "step": 95320 }, { "grad_norm": 0.12912119925022125, "learning_rate": 8.164600663647693e-05, "loss": 0.0183, "step": 95330 }, { "grad_norm": 0.10769156366586685, "learning_rate": 8.164173929557449e-05, "loss": 0.0119, "step": 95340 }, { "grad_norm": 0.22424812614917755, "learning_rate": 8.163747157019503e-05, "loss": 0.0212, "step": 95350 }, { "grad_norm": 0.4706481099128723, "learning_rate": 8.163320346039044e-05, "loss": 0.0226, "step": 95360 }, { "grad_norm": 0.562933087348938, "learning_rate": 8.162893496621255e-05, "loss": 0.0328, "step": 95370 }, { "grad_norm": 0.33467209339141846, "learning_rate": 8.162466608771324e-05, "loss": 0.0235, "step": 95380 }, { "grad_norm": 0.20356416702270508, "learning_rate": 8.162039682494437e-05, "loss": 0.0178, "step": 95390 }, { "grad_norm": 0.14689601957798004, "learning_rate": 8.161612717795783e-05, "loss": 0.0161, "step": 95400 }, { "grad_norm": 0.2599480152130127, "learning_rate": 8.161185714680551e-05, "loss": 0.0208, "step": 95410 }, { "grad_norm": 0.23770327866077423, "learning_rate": 8.160758673153926e-05, "loss": 0.0138, "step": 95420 }, { "grad_norm": 0.2011980563402176, "learning_rate": 8.160331593221099e-05, "loss": 0.0244, "step": 95430 }, { "grad_norm": 0.1512584090232849, "learning_rate": 8.159904474887259e-05, "loss": 0.0187, "step": 95440 }, { "grad_norm": 0.19129647314548492, "learning_rate": 8.159477318157598e-05, "loss": 0.0161, "step": 95450 }, { "grad_norm": 0.18081875145435333, "learning_rate": 8.159050123037301e-05, "loss": 0.0204, "step": 95460 }, { "grad_norm": 0.15559889376163483, "learning_rate": 8.158622889531567e-05, "loss": 0.0211, "step": 95470 }, { "grad_norm": 0.13487690687179565, "learning_rate": 8.158195617645579e-05, "loss": 0.0225, "step": 95480 }, { "grad_norm": 0.09887703508138657, "learning_rate": 8.157768307384535e-05, "loss": 0.0164, "step": 95490 }, { "grad_norm": 0.14445872604846954, "learning_rate": 8.157340958753624e-05, "loss": 0.0152, "step": 95500 }, { "grad_norm": 0.10896944999694824, "learning_rate": 8.156913571758037e-05, "loss": 0.0153, "step": 95510 }, { "grad_norm": 0.19655188918113708, "learning_rate": 8.156486146402972e-05, "loss": 0.0288, "step": 95520 }, { "grad_norm": 0.1397816687822342, "learning_rate": 8.156058682693619e-05, "loss": 0.0156, "step": 95530 }, { "grad_norm": 0.10134811699390411, "learning_rate": 8.155631180635172e-05, "loss": 0.0241, "step": 95540 }, { "grad_norm": 0.09904795140028, "learning_rate": 8.155203640232829e-05, "loss": 0.0131, "step": 95550 }, { "grad_norm": 0.2251509577035904, "learning_rate": 8.154776061491782e-05, "loss": 0.0218, "step": 95560 }, { "grad_norm": 0.15709303319454193, "learning_rate": 8.154348444417227e-05, "loss": 0.0158, "step": 95570 }, { "grad_norm": 0.1134856715798378, "learning_rate": 8.15392078901436e-05, "loss": 0.0178, "step": 95580 }, { "grad_norm": 0.12996742129325867, "learning_rate": 8.153493095288379e-05, "loss": 0.0155, "step": 95590 }, { "grad_norm": 0.1529322862625122, "learning_rate": 8.153065363244479e-05, "loss": 0.0187, "step": 95600 }, { "grad_norm": 0.11820680648088455, "learning_rate": 8.152637592887857e-05, "loss": 0.0187, "step": 95610 }, { "grad_norm": 0.10506044328212738, "learning_rate": 8.152209784223712e-05, "loss": 0.0174, "step": 95620 }, { "grad_norm": 0.129425510764122, "learning_rate": 8.151781937257241e-05, "loss": 0.0164, "step": 95630 }, { "grad_norm": 0.18368403613567352, "learning_rate": 8.151354051993644e-05, "loss": 0.0256, "step": 95640 }, { "grad_norm": 0.2846182882785797, "learning_rate": 8.150926128438122e-05, "loss": 0.0251, "step": 95650 }, { "grad_norm": 0.1118854358792305, "learning_rate": 8.150498166595871e-05, "loss": 0.0188, "step": 95660 }, { "grad_norm": 0.17802117764949799, "learning_rate": 8.150070166472093e-05, "loss": 0.0223, "step": 95670 }, { "grad_norm": 0.12641192972660065, "learning_rate": 8.149642128071988e-05, "loss": 0.0234, "step": 95680 }, { "grad_norm": 0.12782734632492065, "learning_rate": 8.149214051400757e-05, "loss": 0.0211, "step": 95690 }, { "grad_norm": 0.2507172226905823, "learning_rate": 8.148785936463603e-05, "loss": 0.0225, "step": 95700 }, { "grad_norm": 0.17769187688827515, "learning_rate": 8.148357783265727e-05, "loss": 0.0154, "step": 95710 }, { "grad_norm": 0.3437817394733429, "learning_rate": 8.147929591812331e-05, "loss": 0.019, "step": 95720 }, { "grad_norm": 0.1122889369726181, "learning_rate": 8.147501362108617e-05, "loss": 0.0211, "step": 95730 }, { "grad_norm": 0.1441066414117813, "learning_rate": 8.147073094159792e-05, "loss": 0.0182, "step": 95740 }, { "grad_norm": 0.14658457040786743, "learning_rate": 8.146644787971056e-05, "loss": 0.0208, "step": 95750 }, { "grad_norm": 0.12670792639255524, "learning_rate": 8.146216443547615e-05, "loss": 0.017, "step": 95760 }, { "grad_norm": 0.1361253410577774, "learning_rate": 8.145788060894674e-05, "loss": 0.0161, "step": 95770 }, { "grad_norm": 0.1786518096923828, "learning_rate": 8.145359640017438e-05, "loss": 0.0256, "step": 95780 }, { "grad_norm": 0.12570281326770782, "learning_rate": 8.144931180921114e-05, "loss": 0.0142, "step": 95790 }, { "grad_norm": 0.14426061511039734, "learning_rate": 8.144502683610904e-05, "loss": 0.0144, "step": 95800 }, { "grad_norm": 0.19467435777187347, "learning_rate": 8.144074148092021e-05, "loss": 0.0157, "step": 95810 }, { "grad_norm": 0.10782832652330399, "learning_rate": 8.143645574369664e-05, "loss": 0.0161, "step": 95820 }, { "grad_norm": 0.1254698932170868, "learning_rate": 8.14321696244905e-05, "loss": 0.0185, "step": 95830 }, { "grad_norm": 0.15792347490787506, "learning_rate": 8.142788312335379e-05, "loss": 0.0183, "step": 95840 }, { "grad_norm": 0.14198513329029083, "learning_rate": 8.142359624033865e-05, "loss": 0.0177, "step": 95850 }, { "grad_norm": 0.16562354564666748, "learning_rate": 8.141930897549713e-05, "loss": 0.0205, "step": 95860 }, { "grad_norm": 0.10648253560066223, "learning_rate": 8.141502132888132e-05, "loss": 0.0164, "step": 95870 }, { "grad_norm": 0.12076761573553085, "learning_rate": 8.141073330054338e-05, "loss": 0.0174, "step": 95880 }, { "grad_norm": 0.11391458660364151, "learning_rate": 8.140644489053535e-05, "loss": 0.018, "step": 95890 }, { "grad_norm": 0.25579845905303955, "learning_rate": 8.140215609890938e-05, "loss": 0.0172, "step": 95900 }, { "grad_norm": 0.14023146033287048, "learning_rate": 8.139786692571755e-05, "loss": 0.0238, "step": 95910 }, { "grad_norm": 0.15038752555847168, "learning_rate": 8.139357737101201e-05, "loss": 0.0197, "step": 95920 }, { "grad_norm": 0.12544028460979462, "learning_rate": 8.138928743484484e-05, "loss": 0.0213, "step": 95930 }, { "grad_norm": 0.27694636583328247, "learning_rate": 8.138499711726821e-05, "loss": 0.018, "step": 95940 }, { "grad_norm": 0.1789402961730957, "learning_rate": 8.138070641833421e-05, "loss": 0.0211, "step": 95950 }, { "grad_norm": 0.16945357620716095, "learning_rate": 8.137641533809504e-05, "loss": 0.0219, "step": 95960 }, { "grad_norm": 0.26406487822532654, "learning_rate": 8.137212387660276e-05, "loss": 0.0147, "step": 95970 }, { "grad_norm": 0.2376059740781784, "learning_rate": 8.136783203390958e-05, "loss": 0.0212, "step": 95980 }, { "grad_norm": 0.10912193357944489, "learning_rate": 8.13635398100676e-05, "loss": 0.0131, "step": 95990 }, { "grad_norm": 0.12514571845531464, "learning_rate": 8.135924720512903e-05, "loss": 0.0219, "step": 96000 }, { "grad_norm": 0.16858866810798645, "learning_rate": 8.135495421914597e-05, "loss": 0.0182, "step": 96010 }, { "grad_norm": 0.10860247164964676, "learning_rate": 8.135066085217062e-05, "loss": 0.0209, "step": 96020 }, { "grad_norm": 0.18270371854305267, "learning_rate": 8.134636710425513e-05, "loss": 0.0204, "step": 96030 }, { "grad_norm": 0.13828100264072418, "learning_rate": 8.134207297545171e-05, "loss": 0.0167, "step": 96040 }, { "grad_norm": 0.19092495739459991, "learning_rate": 8.13377784658125e-05, "loss": 0.0164, "step": 96050 }, { "grad_norm": 0.13798893988132477, "learning_rate": 8.133348357538971e-05, "loss": 0.0172, "step": 96060 }, { "grad_norm": 0.6720852851867676, "learning_rate": 8.132918830423548e-05, "loss": 0.0215, "step": 96070 }, { "grad_norm": 0.13105152547359467, "learning_rate": 8.132489265240205e-05, "loss": 0.0245, "step": 96080 }, { "grad_norm": 0.1303482949733734, "learning_rate": 8.13205966199416e-05, "loss": 0.0193, "step": 96090 }, { "grad_norm": 0.15079358220100403, "learning_rate": 8.131630020690632e-05, "loss": 0.0251, "step": 96100 }, { "grad_norm": 0.2079942673444748, "learning_rate": 8.131200341334843e-05, "loss": 0.0151, "step": 96110 }, { "grad_norm": 0.17061394453048706, "learning_rate": 8.130770623932013e-05, "loss": 0.0177, "step": 96120 }, { "grad_norm": 0.14054132997989655, "learning_rate": 8.130340868487365e-05, "loss": 0.0213, "step": 96130 }, { "grad_norm": 0.13000629842281342, "learning_rate": 8.12991107500612e-05, "loss": 0.0194, "step": 96140 }, { "grad_norm": 0.145872101187706, "learning_rate": 8.1294812434935e-05, "loss": 0.0197, "step": 96150 }, { "grad_norm": 0.33534640073776245, "learning_rate": 8.129051373954728e-05, "loss": 0.0185, "step": 96160 }, { "grad_norm": 0.15766827762126923, "learning_rate": 8.128621466395028e-05, "loss": 0.0136, "step": 96170 }, { "grad_norm": 0.1656128615140915, "learning_rate": 8.128191520819621e-05, "loss": 0.021, "step": 96180 }, { "grad_norm": 0.11082527041435242, "learning_rate": 8.127761537233736e-05, "loss": 0.018, "step": 96190 }, { "grad_norm": 0.13319867849349976, "learning_rate": 8.127331515642595e-05, "loss": 0.0142, "step": 96200 }, { "grad_norm": 0.12117575109004974, "learning_rate": 8.126901456051421e-05, "loss": 0.0188, "step": 96210 }, { "grad_norm": 0.20722368359565735, "learning_rate": 8.126471358465445e-05, "loss": 0.0216, "step": 96220 }, { "grad_norm": 0.17845870554447174, "learning_rate": 8.126041222889889e-05, "loss": 0.0158, "step": 96230 }, { "grad_norm": 0.21637271344661713, "learning_rate": 8.125611049329981e-05, "loss": 0.0216, "step": 96240 }, { "grad_norm": 0.14999155700206757, "learning_rate": 8.125180837790946e-05, "loss": 0.0167, "step": 96250 }, { "grad_norm": 0.15042343735694885, "learning_rate": 8.124750588278017e-05, "loss": 0.0167, "step": 96260 }, { "grad_norm": 0.22023040056228638, "learning_rate": 8.124320300796414e-05, "loss": 0.0171, "step": 96270 }, { "grad_norm": 0.23825475573539734, "learning_rate": 8.12388997535137e-05, "loss": 0.0184, "step": 96280 }, { "grad_norm": 0.13372893631458282, "learning_rate": 8.123459611948116e-05, "loss": 0.0161, "step": 96290 }, { "grad_norm": 0.11273714154958725, "learning_rate": 8.123029210591876e-05, "loss": 0.0245, "step": 96300 }, { "grad_norm": 0.30382728576660156, "learning_rate": 8.122598771287883e-05, "loss": 0.02, "step": 96310 }, { "grad_norm": 0.4504086971282959, "learning_rate": 8.122168294041368e-05, "loss": 0.0229, "step": 96320 }, { "grad_norm": 0.21465961635112762, "learning_rate": 8.12173777885756e-05, "loss": 0.0191, "step": 96330 }, { "grad_norm": 0.20488370954990387, "learning_rate": 8.12130722574169e-05, "loss": 0.021, "step": 96340 }, { "grad_norm": 0.17655037343502045, "learning_rate": 8.120876634698991e-05, "loss": 0.0161, "step": 96350 }, { "grad_norm": 0.1606602668762207, "learning_rate": 8.120446005734694e-05, "loss": 0.0148, "step": 96360 }, { "grad_norm": 0.1751766949892044, "learning_rate": 8.12001533885403e-05, "loss": 0.0192, "step": 96370 }, { "grad_norm": 0.14213450253009796, "learning_rate": 8.119584634062237e-05, "loss": 0.0182, "step": 96380 }, { "grad_norm": 0.18653328716754913, "learning_rate": 8.119153891364544e-05, "loss": 0.0193, "step": 96390 }, { "grad_norm": 0.14091989398002625, "learning_rate": 8.118723110766185e-05, "loss": 0.0277, "step": 96400 }, { "grad_norm": 0.1285490244626999, "learning_rate": 8.118292292272399e-05, "loss": 0.0184, "step": 96410 }, { "grad_norm": 0.1900555044412613, "learning_rate": 8.117861435888413e-05, "loss": 0.0182, "step": 96420 }, { "grad_norm": 0.1280701905488968, "learning_rate": 8.11743054161947e-05, "loss": 0.0146, "step": 96430 }, { "grad_norm": 0.14234353601932526, "learning_rate": 8.116999609470801e-05, "loss": 0.0187, "step": 96440 }, { "grad_norm": 0.0931018814444542, "learning_rate": 8.116568639447644e-05, "loss": 0.0158, "step": 96450 }, { "grad_norm": 0.22025330364704132, "learning_rate": 8.116137631555236e-05, "loss": 0.0252, "step": 96460 }, { "grad_norm": 0.216285839676857, "learning_rate": 8.115706585798814e-05, "loss": 0.0247, "step": 96470 }, { "grad_norm": 0.07556726783514023, "learning_rate": 8.115275502183615e-05, "loss": 0.0141, "step": 96480 }, { "grad_norm": 0.22334614396095276, "learning_rate": 8.114844380714876e-05, "loss": 0.0179, "step": 96490 }, { "grad_norm": 0.18527643382549286, "learning_rate": 8.114413221397838e-05, "loss": 0.0165, "step": 96500 }, { "grad_norm": 0.14930778741836548, "learning_rate": 8.11398202423774e-05, "loss": 0.0205, "step": 96510 }, { "grad_norm": 0.1697632074356079, "learning_rate": 8.113550789239818e-05, "loss": 0.0213, "step": 96520 }, { "grad_norm": 0.20606549084186554, "learning_rate": 8.113119516409316e-05, "loss": 0.0186, "step": 96530 }, { "grad_norm": 0.1957187056541443, "learning_rate": 8.112688205751472e-05, "loss": 0.0178, "step": 96540 }, { "grad_norm": 0.18083107471466064, "learning_rate": 8.112256857271528e-05, "loss": 0.0156, "step": 96550 }, { "grad_norm": 0.08820375055074692, "learning_rate": 8.111825470974724e-05, "loss": 0.0124, "step": 96560 }, { "grad_norm": 0.22666871547698975, "learning_rate": 8.111394046866302e-05, "loss": 0.0187, "step": 96570 }, { "grad_norm": 0.0903477817773819, "learning_rate": 8.110962584951506e-05, "loss": 0.0176, "step": 96580 }, { "grad_norm": 0.11029616743326187, "learning_rate": 8.110531085235578e-05, "loss": 0.0196, "step": 96590 }, { "grad_norm": 0.21310503780841827, "learning_rate": 8.110099547723757e-05, "loss": 0.0211, "step": 96600 }, { "grad_norm": 0.10636355727910995, "learning_rate": 8.109667972421293e-05, "loss": 0.0169, "step": 96610 }, { "grad_norm": 0.4896810054779053, "learning_rate": 8.109236359333426e-05, "loss": 0.0204, "step": 96620 }, { "grad_norm": 0.33426737785339355, "learning_rate": 8.108804708465403e-05, "loss": 0.0226, "step": 96630 }, { "grad_norm": 0.1419135183095932, "learning_rate": 8.108373019822466e-05, "loss": 0.0183, "step": 96640 }, { "grad_norm": 0.1338171660900116, "learning_rate": 8.107941293409863e-05, "loss": 0.0127, "step": 96650 }, { "grad_norm": 0.1259165108203888, "learning_rate": 8.107509529232838e-05, "loss": 0.0215, "step": 96660 }, { "grad_norm": 0.3215402364730835, "learning_rate": 8.107077727296639e-05, "loss": 0.02, "step": 96670 }, { "grad_norm": 0.20613233745098114, "learning_rate": 8.106645887606512e-05, "loss": 0.0175, "step": 96680 }, { "grad_norm": 0.1517503261566162, "learning_rate": 8.106214010167704e-05, "loss": 0.0174, "step": 96690 }, { "grad_norm": 0.13730089366436005, "learning_rate": 8.105782094985464e-05, "loss": 0.015, "step": 96700 }, { "grad_norm": 0.09430649876594543, "learning_rate": 8.105350142065037e-05, "loss": 0.0158, "step": 96710 }, { "grad_norm": 0.1284157782793045, "learning_rate": 8.104918151411676e-05, "loss": 0.0141, "step": 96720 }, { "grad_norm": 0.12947829067707062, "learning_rate": 8.104486123030627e-05, "loss": 0.0202, "step": 96730 }, { "grad_norm": 0.13153837621212006, "learning_rate": 8.104054056927139e-05, "loss": 0.019, "step": 96740 }, { "grad_norm": 0.24417512118816376, "learning_rate": 8.103621953106466e-05, "loss": 0.0201, "step": 96750 }, { "grad_norm": 0.12422245740890503, "learning_rate": 8.103189811573855e-05, "loss": 0.0205, "step": 96760 }, { "grad_norm": 0.1443338692188263, "learning_rate": 8.102757632334558e-05, "loss": 0.017, "step": 96770 }, { "grad_norm": 0.18403081595897675, "learning_rate": 8.102325415393826e-05, "loss": 0.0148, "step": 96780 }, { "grad_norm": 0.12574408948421478, "learning_rate": 8.101893160756913e-05, "loss": 0.0197, "step": 96790 }, { "grad_norm": 0.23214375972747803, "learning_rate": 8.101460868429067e-05, "loss": 0.0168, "step": 96800 }, { "grad_norm": 0.17249299585819244, "learning_rate": 8.101028538415543e-05, "loss": 0.021, "step": 96810 }, { "grad_norm": 0.13565368950366974, "learning_rate": 8.100596170721597e-05, "loss": 0.0177, "step": 96820 }, { "grad_norm": 0.27702057361602783, "learning_rate": 8.100163765352478e-05, "loss": 0.019, "step": 96830 }, { "grad_norm": 0.1982119381427765, "learning_rate": 8.099731322313442e-05, "loss": 0.0218, "step": 96840 }, { "grad_norm": 0.16321496665477753, "learning_rate": 8.099298841609745e-05, "loss": 0.0199, "step": 96850 }, { "grad_norm": 0.25793105363845825, "learning_rate": 8.098866323246642e-05, "loss": 0.0178, "step": 96860 }, { "grad_norm": 0.10511690378189087, "learning_rate": 8.098433767229384e-05, "loss": 0.0216, "step": 96870 }, { "grad_norm": 0.16232740879058838, "learning_rate": 8.098001173563232e-05, "loss": 0.0164, "step": 96880 }, { "grad_norm": 0.22925826907157898, "learning_rate": 8.097568542253441e-05, "loss": 0.0168, "step": 96890 }, { "grad_norm": 0.19603213667869568, "learning_rate": 8.097135873305267e-05, "loss": 0.0219, "step": 96900 }, { "grad_norm": 0.11708672344684601, "learning_rate": 8.096703166723968e-05, "loss": 0.0166, "step": 96910 }, { "grad_norm": 0.13306482136249542, "learning_rate": 8.096270422514802e-05, "loss": 0.0177, "step": 96920 }, { "grad_norm": 0.1529102474451065, "learning_rate": 8.095837640683029e-05, "loss": 0.0155, "step": 96930 }, { "grad_norm": 0.09778742492198944, "learning_rate": 8.095404821233903e-05, "loss": 0.013, "step": 96940 }, { "grad_norm": 0.13274718821048737, "learning_rate": 8.094971964172687e-05, "loss": 0.0218, "step": 96950 }, { "grad_norm": 0.16123667359352112, "learning_rate": 8.094539069504639e-05, "loss": 0.0196, "step": 96960 }, { "grad_norm": 0.1072932779788971, "learning_rate": 8.094106137235019e-05, "loss": 0.0124, "step": 96970 }, { "grad_norm": 0.2268509566783905, "learning_rate": 8.093673167369089e-05, "loss": 0.02, "step": 96980 }, { "grad_norm": 0.17200084030628204, "learning_rate": 8.093240159912108e-05, "loss": 0.0217, "step": 96990 }, { "grad_norm": 0.15354928374290466, "learning_rate": 8.09280711486934e-05, "loss": 0.0161, "step": 97000 }, { "grad_norm": 0.18037815392017365, "learning_rate": 8.092374032246045e-05, "loss": 0.019, "step": 97010 }, { "grad_norm": 0.1064763143658638, "learning_rate": 8.091940912047486e-05, "loss": 0.0151, "step": 97020 }, { "grad_norm": 0.0993042066693306, "learning_rate": 8.091507754278927e-05, "loss": 0.0156, "step": 97030 }, { "grad_norm": 0.15817907452583313, "learning_rate": 8.091074558945626e-05, "loss": 0.0197, "step": 97040 }, { "grad_norm": 0.14023564755916595, "learning_rate": 8.090641326052854e-05, "loss": 0.02, "step": 97050 }, { "grad_norm": 0.2619136869907379, "learning_rate": 8.09020805560587e-05, "loss": 0.0223, "step": 97060 }, { "grad_norm": 0.13801896572113037, "learning_rate": 8.089774747609942e-05, "loss": 0.0243, "step": 97070 }, { "grad_norm": 0.22562275826931, "learning_rate": 8.089341402070331e-05, "loss": 0.0187, "step": 97080 }, { "grad_norm": 0.16920626163482666, "learning_rate": 8.088908018992309e-05, "loss": 0.0171, "step": 97090 }, { "grad_norm": 0.2124997079372406, "learning_rate": 8.088474598381135e-05, "loss": 0.0245, "step": 97100 }, { "grad_norm": 0.15277281403541565, "learning_rate": 8.088041140242081e-05, "loss": 0.0207, "step": 97110 }, { "grad_norm": 0.07763136923313141, "learning_rate": 8.08760764458041e-05, "loss": 0.023, "step": 97120 }, { "grad_norm": 0.16214619576931, "learning_rate": 8.087174111401393e-05, "loss": 0.0167, "step": 97130 }, { "grad_norm": 0.8244383931159973, "learning_rate": 8.086740540710293e-05, "loss": 0.0217, "step": 97140 }, { "grad_norm": 0.11407586187124252, "learning_rate": 8.086306932512382e-05, "loss": 0.0223, "step": 97150 }, { "grad_norm": 0.5547046661376953, "learning_rate": 8.085873286812928e-05, "loss": 0.0239, "step": 97160 }, { "grad_norm": 0.1583438664674759, "learning_rate": 8.085439603617199e-05, "loss": 0.0237, "step": 97170 }, { "grad_norm": 0.22644923627376556, "learning_rate": 8.085005882930468e-05, "loss": 0.0272, "step": 97180 }, { "grad_norm": 0.13070769608020782, "learning_rate": 8.084572124758001e-05, "loss": 0.0221, "step": 97190 }, { "grad_norm": 0.5149794816970825, "learning_rate": 8.084138329105071e-05, "loss": 0.0214, "step": 97200 }, { "grad_norm": 0.12838128209114075, "learning_rate": 8.08370449597695e-05, "loss": 0.0186, "step": 97210 }, { "grad_norm": 0.2599090337753296, "learning_rate": 8.083270625378905e-05, "loss": 0.0188, "step": 97220 }, { "grad_norm": 0.13784721493721008, "learning_rate": 8.08283671731621e-05, "loss": 0.0155, "step": 97230 }, { "grad_norm": 0.159348264336586, "learning_rate": 8.082402771794141e-05, "loss": 0.0117, "step": 97240 }, { "grad_norm": 0.21797187626361847, "learning_rate": 8.081968788817967e-05, "loss": 0.0192, "step": 97250 }, { "grad_norm": 0.21510349214076996, "learning_rate": 8.081534768392963e-05, "loss": 0.0192, "step": 97260 }, { "grad_norm": 0.12129464000463486, "learning_rate": 8.081100710524401e-05, "loss": 0.0223, "step": 97270 }, { "grad_norm": 0.11982426047325134, "learning_rate": 8.080666615217557e-05, "loss": 0.0208, "step": 97280 }, { "grad_norm": 0.09958916157484055, "learning_rate": 8.080232482477705e-05, "loss": 0.0172, "step": 97290 }, { "grad_norm": 0.18892687559127808, "learning_rate": 8.07979831231012e-05, "loss": 0.0198, "step": 97300 }, { "grad_norm": 0.16537824273109436, "learning_rate": 8.079364104720078e-05, "loss": 0.0192, "step": 97310 }, { "grad_norm": 0.19953115284442902, "learning_rate": 8.078929859712855e-05, "loss": 0.0132, "step": 97320 }, { "grad_norm": 0.16419360041618347, "learning_rate": 8.078495577293727e-05, "loss": 0.0188, "step": 97330 }, { "grad_norm": 0.14047455787658691, "learning_rate": 8.07806125746797e-05, "loss": 0.0181, "step": 97340 }, { "grad_norm": 0.09814010560512543, "learning_rate": 8.077626900240865e-05, "loss": 0.0183, "step": 97350 }, { "grad_norm": 0.1311388909816742, "learning_rate": 8.077192505617684e-05, "loss": 0.0146, "step": 97360 }, { "grad_norm": 0.10581032931804657, "learning_rate": 8.076758073603712e-05, "loss": 0.0226, "step": 97370 }, { "grad_norm": 0.15611866116523743, "learning_rate": 8.076323604204222e-05, "loss": 0.0233, "step": 97380 }, { "grad_norm": 0.1104649007320404, "learning_rate": 8.075889097424496e-05, "loss": 0.0223, "step": 97390 }, { "grad_norm": 0.10902673006057739, "learning_rate": 8.075454553269815e-05, "loss": 0.0234, "step": 97400 }, { "grad_norm": 0.13500535488128662, "learning_rate": 8.075019971745457e-05, "loss": 0.0171, "step": 97410 }, { "grad_norm": 0.08566364645957947, "learning_rate": 8.074585352856701e-05, "loss": 0.0157, "step": 97420 }, { "grad_norm": 0.14429587125778198, "learning_rate": 8.074150696608832e-05, "loss": 0.0161, "step": 97430 }, { "grad_norm": 0.16792349517345428, "learning_rate": 8.07371600300713e-05, "loss": 0.0179, "step": 97440 }, { "grad_norm": 0.1807679384946823, "learning_rate": 8.073281272056874e-05, "loss": 0.0237, "step": 97450 }, { "grad_norm": 0.1461266279220581, "learning_rate": 8.072846503763351e-05, "loss": 0.012, "step": 97460 }, { "grad_norm": 0.2135184407234192, "learning_rate": 8.072411698131841e-05, "loss": 0.0184, "step": 97470 }, { "grad_norm": 0.12630446255207062, "learning_rate": 8.071976855167629e-05, "loss": 0.0124, "step": 97480 }, { "grad_norm": 0.11888469010591507, "learning_rate": 8.071541974875996e-05, "loss": 0.0197, "step": 97490 }, { "grad_norm": 0.1138504222035408, "learning_rate": 8.071107057262231e-05, "loss": 0.0157, "step": 97500 }, { "grad_norm": 0.1331159770488739, "learning_rate": 8.070672102331614e-05, "loss": 0.0193, "step": 97510 }, { "grad_norm": 0.13079597055912018, "learning_rate": 8.07023711008943e-05, "loss": 0.0182, "step": 97520 }, { "grad_norm": 0.17639383673667908, "learning_rate": 8.069802080540969e-05, "loss": 0.0148, "step": 97530 }, { "grad_norm": 0.13342903554439545, "learning_rate": 8.069367013691512e-05, "loss": 0.0208, "step": 97540 }, { "grad_norm": 0.123087078332901, "learning_rate": 8.068931909546349e-05, "loss": 0.0284, "step": 97550 }, { "grad_norm": 0.1962965726852417, "learning_rate": 8.068496768110766e-05, "loss": 0.0198, "step": 97560 }, { "grad_norm": 0.14763015508651733, "learning_rate": 8.06806158939005e-05, "loss": 0.0163, "step": 97570 }, { "grad_norm": 0.1493232101202011, "learning_rate": 8.067626373389488e-05, "loss": 0.0191, "step": 97580 }, { "grad_norm": 0.15312960743904114, "learning_rate": 8.06719112011437e-05, "loss": 0.0143, "step": 97590 }, { "grad_norm": 0.2625870704650879, "learning_rate": 8.066755829569985e-05, "loss": 0.0188, "step": 97600 }, { "grad_norm": 0.11752720922231674, "learning_rate": 8.066320501761619e-05, "loss": 0.0165, "step": 97610 }, { "grad_norm": 0.1488388329744339, "learning_rate": 8.065885136694566e-05, "loss": 0.0214, "step": 97620 }, { "grad_norm": 0.1589730978012085, "learning_rate": 8.065449734374113e-05, "loss": 0.0194, "step": 97630 }, { "grad_norm": 0.09361426532268524, "learning_rate": 8.065014294805551e-05, "loss": 0.0154, "step": 97640 }, { "grad_norm": 0.1488959789276123, "learning_rate": 8.064578817994172e-05, "loss": 0.0139, "step": 97650 }, { "grad_norm": 0.27134373784065247, "learning_rate": 8.064143303945266e-05, "loss": 0.021, "step": 97660 }, { "grad_norm": 0.2979338467121124, "learning_rate": 8.063707752664127e-05, "loss": 0.0157, "step": 97670 }, { "grad_norm": 0.18158908188343048, "learning_rate": 8.063272164156046e-05, "loss": 0.0225, "step": 97680 }, { "grad_norm": 0.16795988380908966, "learning_rate": 8.062836538426316e-05, "loss": 0.0255, "step": 97690 }, { "grad_norm": 0.1954992264509201, "learning_rate": 8.06240087548023e-05, "loss": 0.0191, "step": 97700 }, { "grad_norm": 0.4050682783126831, "learning_rate": 8.061965175323083e-05, "loss": 0.0186, "step": 97710 }, { "grad_norm": 0.24578998982906342, "learning_rate": 8.061529437960167e-05, "loss": 0.0184, "step": 97720 }, { "grad_norm": 0.11105625331401825, "learning_rate": 8.061093663396778e-05, "loss": 0.0156, "step": 97730 }, { "grad_norm": 0.12825869023799896, "learning_rate": 8.060657851638209e-05, "loss": 0.0235, "step": 97740 }, { "grad_norm": 0.19487576186656952, "learning_rate": 8.06022200268976e-05, "loss": 0.0174, "step": 97750 }, { "grad_norm": 0.14907220005989075, "learning_rate": 8.059786116556724e-05, "loss": 0.0167, "step": 97760 }, { "grad_norm": 0.1141790971159935, "learning_rate": 8.059350193244396e-05, "loss": 0.0213, "step": 97770 }, { "grad_norm": 0.15561282634735107, "learning_rate": 8.058914232758078e-05, "loss": 0.0195, "step": 97780 }, { "grad_norm": 0.17648904025554657, "learning_rate": 8.05847823510306e-05, "loss": 0.0245, "step": 97790 }, { "grad_norm": 0.22649133205413818, "learning_rate": 8.058042200284645e-05, "loss": 0.0219, "step": 97800 }, { "grad_norm": 0.13618893921375275, "learning_rate": 8.057606128308132e-05, "loss": 0.0192, "step": 97810 }, { "grad_norm": 0.13046500086784363, "learning_rate": 8.057170019178817e-05, "loss": 0.0203, "step": 97820 }, { "grad_norm": 0.1426497846841812, "learning_rate": 8.056733872901997e-05, "loss": 0.0244, "step": 97830 }, { "grad_norm": 0.14926843345165253, "learning_rate": 8.056297689482978e-05, "loss": 0.0197, "step": 97840 }, { "grad_norm": 0.13952326774597168, "learning_rate": 8.055861468927055e-05, "loss": 0.0222, "step": 97850 }, { "grad_norm": 0.14173577725887299, "learning_rate": 8.055425211239529e-05, "loss": 0.018, "step": 97860 }, { "grad_norm": 0.23906102776527405, "learning_rate": 8.054988916425702e-05, "loss": 0.0192, "step": 97870 }, { "grad_norm": 0.12016920000314713, "learning_rate": 8.054552584490878e-05, "loss": 0.0188, "step": 97880 }, { "grad_norm": 0.10826601088047028, "learning_rate": 8.054116215440354e-05, "loss": 0.0153, "step": 97890 }, { "grad_norm": 0.2996395528316498, "learning_rate": 8.053679809279434e-05, "loss": 0.0231, "step": 97900 }, { "grad_norm": 0.1332225501537323, "learning_rate": 8.053243366013422e-05, "loss": 0.0152, "step": 97910 }, { "grad_norm": 0.16236236691474915, "learning_rate": 8.052806885647621e-05, "loss": 0.0238, "step": 97920 }, { "grad_norm": 0.13176946341991425, "learning_rate": 8.052370368187332e-05, "loss": 0.0136, "step": 97930 }, { "grad_norm": 0.14784054458141327, "learning_rate": 8.051933813637863e-05, "loss": 0.0229, "step": 97940 }, { "grad_norm": 0.12069683521986008, "learning_rate": 8.051497222004516e-05, "loss": 0.0163, "step": 97950 }, { "grad_norm": 0.13115864992141724, "learning_rate": 8.051060593292596e-05, "loss": 0.0214, "step": 97960 }, { "grad_norm": 0.10487959533929825, "learning_rate": 8.05062392750741e-05, "loss": 0.0167, "step": 97970 }, { "grad_norm": 0.6972189545631409, "learning_rate": 8.050187224654263e-05, "loss": 0.0157, "step": 97980 }, { "grad_norm": 0.16164803504943848, "learning_rate": 8.049750484738461e-05, "loss": 0.0152, "step": 97990 }, { "grad_norm": 0.22960099577903748, "learning_rate": 8.049313707765312e-05, "loss": 0.0217, "step": 98000 }, { "grad_norm": 0.1343831568956375, "learning_rate": 8.048876893740121e-05, "loss": 0.0164, "step": 98010 }, { "grad_norm": 0.12947292625904083, "learning_rate": 8.048440042668198e-05, "loss": 0.0194, "step": 98020 }, { "grad_norm": 0.159746453166008, "learning_rate": 8.048003154554851e-05, "loss": 0.0223, "step": 98030 }, { "grad_norm": 0.13818547129631042, "learning_rate": 8.047566229405386e-05, "loss": 0.0139, "step": 98040 }, { "grad_norm": 0.1463712900876999, "learning_rate": 8.047129267225115e-05, "loss": 0.0229, "step": 98050 }, { "grad_norm": 0.2162671983242035, "learning_rate": 8.046692268019345e-05, "loss": 0.0148, "step": 98060 }, { "grad_norm": 0.22427338361740112, "learning_rate": 8.046255231793389e-05, "loss": 0.0219, "step": 98070 }, { "grad_norm": 0.2702845633029938, "learning_rate": 8.045818158552555e-05, "loss": 0.0228, "step": 98080 }, { "grad_norm": 0.10472891479730606, "learning_rate": 8.045381048302157e-05, "loss": 0.0185, "step": 98090 }, { "grad_norm": 0.15815967321395874, "learning_rate": 8.044943901047502e-05, "loss": 0.0164, "step": 98100 }, { "grad_norm": 0.18131043016910553, "learning_rate": 8.044506716793904e-05, "loss": 0.019, "step": 98110 }, { "grad_norm": 0.19996120035648346, "learning_rate": 8.044069495546674e-05, "loss": 0.022, "step": 98120 }, { "grad_norm": 0.11566619575023651, "learning_rate": 8.043632237311127e-05, "loss": 0.017, "step": 98130 }, { "grad_norm": 0.19937950372695923, "learning_rate": 8.043194942092574e-05, "loss": 0.0235, "step": 98140 }, { "grad_norm": 0.15632979571819305, "learning_rate": 8.042757609896327e-05, "loss": 0.0233, "step": 98150 }, { "grad_norm": 0.11557856947183609, "learning_rate": 8.042320240727705e-05, "loss": 0.0163, "step": 98160 }, { "grad_norm": 0.15625110268592834, "learning_rate": 8.04188283459202e-05, "loss": 0.0165, "step": 98170 }, { "grad_norm": 0.4806530773639679, "learning_rate": 8.041445391494584e-05, "loss": 0.0221, "step": 98180 }, { "grad_norm": 0.15518319606781006, "learning_rate": 8.041007911440718e-05, "loss": 0.0164, "step": 98190 }, { "grad_norm": 0.08471672236919403, "learning_rate": 8.040570394435733e-05, "loss": 0.0174, "step": 98200 }, { "grad_norm": 0.252485454082489, "learning_rate": 8.040132840484946e-05, "loss": 0.0215, "step": 98210 }, { "grad_norm": 0.10618630051612854, "learning_rate": 8.039695249593676e-05, "loss": 0.0206, "step": 98220 }, { "grad_norm": 0.1965673565864563, "learning_rate": 8.039257621767238e-05, "loss": 0.021, "step": 98230 }, { "grad_norm": 0.12462431192398071, "learning_rate": 8.038819957010952e-05, "loss": 0.014, "step": 98240 }, { "grad_norm": 0.19745299220085144, "learning_rate": 8.038382255330133e-05, "loss": 0.021, "step": 98250 }, { "grad_norm": 0.15800808370113373, "learning_rate": 8.037944516730101e-05, "loss": 0.0215, "step": 98260 }, { "grad_norm": 0.15196014940738678, "learning_rate": 8.037506741216174e-05, "loss": 0.022, "step": 98270 }, { "grad_norm": 0.08755394071340561, "learning_rate": 8.037068928793674e-05, "loss": 0.0165, "step": 98280 }, { "grad_norm": 0.13928644359111786, "learning_rate": 8.036631079467917e-05, "loss": 0.0245, "step": 98290 }, { "grad_norm": 0.2074417769908905, "learning_rate": 8.036193193244228e-05, "loss": 0.0158, "step": 98300 }, { "grad_norm": 0.26832717657089233, "learning_rate": 8.035755270127922e-05, "loss": 0.0206, "step": 98310 }, { "grad_norm": 0.15156452357769012, "learning_rate": 8.035317310124327e-05, "loss": 0.0142, "step": 98320 }, { "grad_norm": 0.11797501146793365, "learning_rate": 8.034879313238758e-05, "loss": 0.0239, "step": 98330 }, { "grad_norm": 0.48422950506210327, "learning_rate": 8.03444127947654e-05, "loss": 0.0166, "step": 98340 }, { "grad_norm": 0.10066299140453339, "learning_rate": 8.034003208842997e-05, "loss": 0.0156, "step": 98350 }, { "grad_norm": 0.1403294950723648, "learning_rate": 8.033565101343448e-05, "loss": 0.0185, "step": 98360 }, { "grad_norm": 0.3053474426269531, "learning_rate": 8.033126956983221e-05, "loss": 0.0208, "step": 98370 }, { "grad_norm": 0.1794678419828415, "learning_rate": 8.032688775767637e-05, "loss": 0.0126, "step": 98380 }, { "grad_norm": 0.1853281408548355, "learning_rate": 8.032250557702021e-05, "loss": 0.0288, "step": 98390 }, { "grad_norm": 0.17210976779460907, "learning_rate": 8.0318123027917e-05, "loss": 0.0244, "step": 98400 }, { "grad_norm": 0.09383139759302139, "learning_rate": 8.031374011041993e-05, "loss": 0.0193, "step": 98410 }, { "grad_norm": 0.22450226545333862, "learning_rate": 8.030935682458231e-05, "loss": 0.0213, "step": 98420 }, { "grad_norm": 0.31728869676589966, "learning_rate": 8.03049731704574e-05, "loss": 0.0235, "step": 98430 }, { "grad_norm": 0.14705538749694824, "learning_rate": 8.030058914809844e-05, "loss": 0.0197, "step": 98440 }, { "grad_norm": 0.1615728884935379, "learning_rate": 8.029620475755872e-05, "loss": 0.0171, "step": 98450 }, { "grad_norm": 0.13935282826423645, "learning_rate": 8.029181999889152e-05, "loss": 0.0195, "step": 98460 }, { "grad_norm": 0.13385626673698425, "learning_rate": 8.028743487215009e-05, "loss": 0.0196, "step": 98470 }, { "grad_norm": 0.23225627839565277, "learning_rate": 8.028304937738773e-05, "loss": 0.0216, "step": 98480 }, { "grad_norm": 0.1309186965227127, "learning_rate": 8.027866351465775e-05, "loss": 0.0179, "step": 98490 }, { "grad_norm": 0.11179378628730774, "learning_rate": 8.02742772840134e-05, "loss": 0.0177, "step": 98500 }, { "grad_norm": 0.12231124192476273, "learning_rate": 8.026989068550801e-05, "loss": 0.0169, "step": 98510 }, { "grad_norm": 0.17088279128074646, "learning_rate": 8.026550371919485e-05, "loss": 0.0222, "step": 98520 }, { "grad_norm": 0.10347051173448563, "learning_rate": 8.026111638512725e-05, "loss": 0.014, "step": 98530 }, { "grad_norm": 0.14637310802936554, "learning_rate": 8.025672868335851e-05, "loss": 0.0155, "step": 98540 }, { "grad_norm": 0.10006166249513626, "learning_rate": 8.025234061394198e-05, "loss": 0.0168, "step": 98550 }, { "grad_norm": 0.22659306228160858, "learning_rate": 8.024795217693094e-05, "loss": 0.0171, "step": 98560 }, { "grad_norm": 0.11486983299255371, "learning_rate": 8.024356337237868e-05, "loss": 0.0149, "step": 98570 }, { "grad_norm": 0.0971333310008049, "learning_rate": 8.023917420033862e-05, "loss": 0.0144, "step": 98580 }, { "grad_norm": 0.9670413732528687, "learning_rate": 8.0234784660864e-05, "loss": 0.0179, "step": 98590 }, { "grad_norm": 0.20567724108695984, "learning_rate": 8.023039475400825e-05, "loss": 0.0146, "step": 98600 }, { "grad_norm": 0.126217782497406, "learning_rate": 8.022600447982462e-05, "loss": 0.0184, "step": 98610 }, { "grad_norm": 0.1013953909277916, "learning_rate": 8.022161383836652e-05, "loss": 0.0213, "step": 98620 }, { "grad_norm": 0.12524646520614624, "learning_rate": 8.021722282968727e-05, "loss": 0.0151, "step": 98630 }, { "grad_norm": 0.14982694387435913, "learning_rate": 8.021283145384022e-05, "loss": 0.0135, "step": 98640 }, { "grad_norm": 0.19153814017772675, "learning_rate": 8.020843971087877e-05, "loss": 0.0144, "step": 98650 }, { "grad_norm": 0.2644265592098236, "learning_rate": 8.020404760085622e-05, "loss": 0.0205, "step": 98660 }, { "grad_norm": 0.10079871863126755, "learning_rate": 8.0199655123826e-05, "loss": 0.0148, "step": 98670 }, { "grad_norm": 0.21547675132751465, "learning_rate": 8.019526227984147e-05, "loss": 0.0184, "step": 98680 }, { "grad_norm": 0.1859365999698639, "learning_rate": 8.019086906895596e-05, "loss": 0.0203, "step": 98690 }, { "grad_norm": 0.10058873146772385, "learning_rate": 8.018647549122291e-05, "loss": 0.0145, "step": 98700 }, { "grad_norm": 0.16549114882946014, "learning_rate": 8.018208154669568e-05, "loss": 0.0215, "step": 98710 }, { "grad_norm": 0.14121100306510925, "learning_rate": 8.017768723542765e-05, "loss": 0.0144, "step": 98720 }, { "grad_norm": 0.3454425632953644, "learning_rate": 8.017329255747223e-05, "loss": 0.0224, "step": 98730 }, { "grad_norm": 0.30473318696022034, "learning_rate": 8.016889751288283e-05, "loss": 0.0212, "step": 98740 }, { "grad_norm": 0.11664406210184097, "learning_rate": 8.016450210171283e-05, "loss": 0.0148, "step": 98750 }, { "grad_norm": 0.40504762530326843, "learning_rate": 8.016010632401565e-05, "loss": 0.0161, "step": 98760 }, { "grad_norm": 0.1676805168390274, "learning_rate": 8.015571017984471e-05, "loss": 0.0211, "step": 98770 }, { "grad_norm": 0.15211008489131927, "learning_rate": 8.015131366925341e-05, "loss": 0.0158, "step": 98780 }, { "grad_norm": 0.18248295783996582, "learning_rate": 8.014691679229518e-05, "loss": 0.0172, "step": 98790 }, { "grad_norm": 0.15490899980068207, "learning_rate": 8.014251954902347e-05, "loss": 0.016, "step": 98800 }, { "grad_norm": 0.16616623103618622, "learning_rate": 8.013812193949166e-05, "loss": 0.0218, "step": 98810 }, { "grad_norm": 0.09072157740592957, "learning_rate": 8.013372396375323e-05, "loss": 0.0194, "step": 98820 }, { "grad_norm": 0.12755918502807617, "learning_rate": 8.012932562186159e-05, "loss": 0.0196, "step": 98830 }, { "grad_norm": 0.08832785487174988, "learning_rate": 8.012492691387021e-05, "loss": 0.0185, "step": 98840 }, { "grad_norm": 0.10161464661359787, "learning_rate": 8.012052783983252e-05, "loss": 0.0227, "step": 98850 }, { "grad_norm": 0.13145561516284943, "learning_rate": 8.011612839980198e-05, "loss": 0.0175, "step": 98860 }, { "grad_norm": 0.14017991721630096, "learning_rate": 8.011172859383204e-05, "loss": 0.0119, "step": 98870 }, { "grad_norm": 0.11545822769403458, "learning_rate": 8.010732842197616e-05, "loss": 0.0122, "step": 98880 }, { "grad_norm": 0.19020622968673706, "learning_rate": 8.010292788428784e-05, "loss": 0.0179, "step": 98890 }, { "grad_norm": 0.1651829034090042, "learning_rate": 8.00985269808205e-05, "loss": 0.0175, "step": 98900 }, { "grad_norm": 0.12990856170654297, "learning_rate": 8.009412571162765e-05, "loss": 0.0214, "step": 98910 }, { "grad_norm": 0.1210581362247467, "learning_rate": 8.008972407676277e-05, "loss": 0.0228, "step": 98920 }, { "grad_norm": 0.27595579624176025, "learning_rate": 8.008532207627932e-05, "loss": 0.0171, "step": 98930 }, { "grad_norm": 0.1978050023317337, "learning_rate": 8.008091971023079e-05, "loss": 0.021, "step": 98940 }, { "grad_norm": 0.28462883830070496, "learning_rate": 8.007651697867069e-05, "loss": 0.0179, "step": 98950 }, { "grad_norm": 0.11936690658330917, "learning_rate": 8.007211388165252e-05, "loss": 0.0169, "step": 98960 }, { "grad_norm": 0.1813027709722519, "learning_rate": 8.006771041922978e-05, "loss": 0.0174, "step": 98970 }, { "grad_norm": 0.12699632346630096, "learning_rate": 8.006330659145595e-05, "loss": 0.0167, "step": 98980 }, { "grad_norm": 0.6222018003463745, "learning_rate": 8.005890239838458e-05, "loss": 0.02, "step": 98990 }, { "grad_norm": 0.12840916216373444, "learning_rate": 8.005449784006917e-05, "loss": 0.0167, "step": 99000 }, { "grad_norm": 0.12235007435083389, "learning_rate": 8.005009291656322e-05, "loss": 0.0175, "step": 99010 }, { "grad_norm": 0.1680077761411667, "learning_rate": 8.004568762792026e-05, "loss": 0.0171, "step": 99020 }, { "grad_norm": 0.08633022755384445, "learning_rate": 8.004128197419385e-05, "loss": 0.0117, "step": 99030 }, { "grad_norm": 0.10182247310876846, "learning_rate": 8.003687595543749e-05, "loss": 0.0172, "step": 99040 }, { "grad_norm": 0.11976797878742218, "learning_rate": 8.003246957170473e-05, "loss": 0.0189, "step": 99050 }, { "grad_norm": 0.08907808363437653, "learning_rate": 8.00280628230491e-05, "loss": 0.0166, "step": 99060 }, { "grad_norm": 0.08562018722295761, "learning_rate": 8.002365570952417e-05, "loss": 0.0135, "step": 99070 }, { "grad_norm": 0.14014281332492828, "learning_rate": 8.001924823118346e-05, "loss": 0.0205, "step": 99080 }, { "grad_norm": 0.18483303487300873, "learning_rate": 8.001484038808056e-05, "loss": 0.0167, "step": 99090 }, { "grad_norm": 0.11914694309234619, "learning_rate": 8.0010432180269e-05, "loss": 0.0163, "step": 99100 }, { "grad_norm": 0.13883763551712036, "learning_rate": 8.000602360780234e-05, "loss": 0.0249, "step": 99110 }, { "grad_norm": 0.1982998251914978, "learning_rate": 8.000161467073419e-05, "loss": 0.0224, "step": 99120 }, { "grad_norm": 0.1703428477048874, "learning_rate": 7.999720536911807e-05, "loss": 0.0148, "step": 99130 }, { "grad_norm": 0.09104295819997787, "learning_rate": 7.999279570300761e-05, "loss": 0.0148, "step": 99140 }, { "grad_norm": 0.1657022088766098, "learning_rate": 7.998838567245634e-05, "loss": 0.022, "step": 99150 }, { "grad_norm": 0.14443598687648773, "learning_rate": 7.998397527751788e-05, "loss": 0.0135, "step": 99160 }, { "grad_norm": 0.228066086769104, "learning_rate": 7.997956451824582e-05, "loss": 0.0155, "step": 99170 }, { "grad_norm": 0.1492381989955902, "learning_rate": 7.997515339469372e-05, "loss": 0.0217, "step": 99180 }, { "grad_norm": 0.1073029562830925, "learning_rate": 7.997074190691522e-05, "loss": 0.0148, "step": 99190 }, { "grad_norm": 0.32443714141845703, "learning_rate": 7.996633005496392e-05, "loss": 0.0167, "step": 99200 }, { "grad_norm": 0.11336670815944672, "learning_rate": 7.99619178388934e-05, "loss": 0.0151, "step": 99210 }, { "grad_norm": 0.16650506854057312, "learning_rate": 7.995750525875731e-05, "loss": 0.0154, "step": 99220 }, { "grad_norm": 0.08367903530597687, "learning_rate": 7.995309231460923e-05, "loss": 0.018, "step": 99230 }, { "grad_norm": 0.1729411482810974, "learning_rate": 7.99486790065028e-05, "loss": 0.0221, "step": 99240 }, { "grad_norm": 0.11518768221139908, "learning_rate": 7.994426533449165e-05, "loss": 0.0149, "step": 99250 }, { "grad_norm": 0.06985260546207428, "learning_rate": 7.99398512986294e-05, "loss": 0.018, "step": 99260 }, { "grad_norm": 0.11569979786872864, "learning_rate": 7.993543689896968e-05, "loss": 0.0141, "step": 99270 }, { "grad_norm": 0.08340764045715332, "learning_rate": 7.993102213556616e-05, "loss": 0.0166, "step": 99280 }, { "grad_norm": 0.09697617590427399, "learning_rate": 7.992660700847244e-05, "loss": 0.0164, "step": 99290 }, { "grad_norm": 0.1302482634782791, "learning_rate": 7.99221915177422e-05, "loss": 0.0183, "step": 99300 }, { "grad_norm": 0.09398990124464035, "learning_rate": 7.991777566342909e-05, "loss": 0.022, "step": 99310 }, { "grad_norm": 0.12418503314256668, "learning_rate": 7.991335944558673e-05, "loss": 0.0264, "step": 99320 }, { "grad_norm": 0.3831760287284851, "learning_rate": 7.990894286426883e-05, "loss": 0.0227, "step": 99330 }, { "grad_norm": 0.1773456186056137, "learning_rate": 7.990452591952902e-05, "loss": 0.0151, "step": 99340 }, { "grad_norm": 0.3012937605381012, "learning_rate": 7.990010861142101e-05, "loss": 0.0239, "step": 99350 }, { "grad_norm": 0.12175504118204117, "learning_rate": 7.989569093999843e-05, "loss": 0.0157, "step": 99360 }, { "grad_norm": 0.1515110433101654, "learning_rate": 7.989127290531499e-05, "loss": 0.0237, "step": 99370 }, { "grad_norm": 0.27068036794662476, "learning_rate": 7.988685450742437e-05, "loss": 0.0189, "step": 99380 }, { "grad_norm": 0.10732448846101761, "learning_rate": 7.988243574638024e-05, "loss": 0.0162, "step": 99390 }, { "grad_norm": 0.11446521431207657, "learning_rate": 7.987801662223631e-05, "loss": 0.016, "step": 99400 }, { "grad_norm": 0.15023723244667053, "learning_rate": 7.987359713504626e-05, "loss": 0.0153, "step": 99410 }, { "grad_norm": 0.11166605353355408, "learning_rate": 7.986917728486378e-05, "loss": 0.0173, "step": 99420 }, { "grad_norm": 0.10131090134382248, "learning_rate": 7.986475707174261e-05, "loss": 0.0167, "step": 99430 }, { "grad_norm": 0.16012156009674072, "learning_rate": 7.986033649573645e-05, "loss": 0.0222, "step": 99440 }, { "grad_norm": 0.12255629897117615, "learning_rate": 7.985591555689901e-05, "loss": 0.0144, "step": 99450 }, { "grad_norm": 0.15319153666496277, "learning_rate": 7.9851494255284e-05, "loss": 0.0196, "step": 99460 }, { "grad_norm": 0.11750449240207672, "learning_rate": 7.984707259094517e-05, "loss": 0.0178, "step": 99470 }, { "grad_norm": 0.12063965201377869, "learning_rate": 7.98426505639362e-05, "loss": 0.017, "step": 99480 }, { "grad_norm": 0.17858441174030304, "learning_rate": 7.983822817431088e-05, "loss": 0.0158, "step": 99490 }, { "grad_norm": 0.17098961770534515, "learning_rate": 7.983380542212288e-05, "loss": 0.0189, "step": 99500 }, { "grad_norm": 0.14976821839809418, "learning_rate": 7.9829382307426e-05, "loss": 0.0134, "step": 99510 }, { "grad_norm": 0.12187675386667252, "learning_rate": 7.982495883027395e-05, "loss": 0.0149, "step": 99520 }, { "grad_norm": 0.11460771411657333, "learning_rate": 7.982053499072051e-05, "loss": 0.025, "step": 99530 }, { "grad_norm": 0.10683952271938324, "learning_rate": 7.98161107888194e-05, "loss": 0.0156, "step": 99540 }, { "grad_norm": 0.39833810925483704, "learning_rate": 7.981168622462439e-05, "loss": 0.0158, "step": 99550 }, { "grad_norm": 0.08847232908010483, "learning_rate": 7.980726129818925e-05, "loss": 0.016, "step": 99560 }, { "grad_norm": 0.11916650086641312, "learning_rate": 7.980283600956774e-05, "loss": 0.0163, "step": 99570 }, { "grad_norm": 0.15991094708442688, "learning_rate": 7.979841035881363e-05, "loss": 0.0204, "step": 99580 }, { "grad_norm": 0.12970243394374847, "learning_rate": 7.979398434598072e-05, "loss": 0.0156, "step": 99590 }, { "grad_norm": 0.10653506219387054, "learning_rate": 7.978955797112275e-05, "loss": 0.0186, "step": 99600 }, { "grad_norm": 0.1176295056939125, "learning_rate": 7.978513123429353e-05, "loss": 0.0171, "step": 99610 }, { "grad_norm": 0.12721946835517883, "learning_rate": 7.978070413554684e-05, "loss": 0.0154, "step": 99620 }, { "grad_norm": 0.12763439118862152, "learning_rate": 7.977627667493648e-05, "loss": 0.0187, "step": 99630 }, { "grad_norm": 0.14366555213928223, "learning_rate": 7.977184885251625e-05, "loss": 0.0174, "step": 99640 }, { "grad_norm": 0.22549305856227875, "learning_rate": 7.976742066833995e-05, "loss": 0.0185, "step": 99650 }, { "grad_norm": 0.08950894325971603, "learning_rate": 7.976299212246136e-05, "loss": 0.0208, "step": 99660 }, { "grad_norm": 0.15552698075771332, "learning_rate": 7.975856321493434e-05, "loss": 0.0207, "step": 99670 }, { "grad_norm": 0.21764306724071503, "learning_rate": 7.975413394581268e-05, "loss": 0.02, "step": 99680 }, { "grad_norm": 0.13226646184921265, "learning_rate": 7.974970431515019e-05, "loss": 0.0209, "step": 99690 }, { "grad_norm": 0.17773988842964172, "learning_rate": 7.97452743230007e-05, "loss": 0.0173, "step": 99700 }, { "grad_norm": 0.17719370126724243, "learning_rate": 7.974084396941805e-05, "loss": 0.0222, "step": 99710 }, { "grad_norm": 0.08795289695262909, "learning_rate": 7.973641325445607e-05, "loss": 0.0163, "step": 99720 }, { "grad_norm": 0.08543747663497925, "learning_rate": 7.973198217816858e-05, "loss": 0.0175, "step": 99730 }, { "grad_norm": 0.11223660409450531, "learning_rate": 7.972755074060945e-05, "loss": 0.0188, "step": 99740 }, { "grad_norm": 0.12971551716327667, "learning_rate": 7.97231189418325e-05, "loss": 0.0171, "step": 99750 }, { "grad_norm": 0.09227702021598816, "learning_rate": 7.971868678189161e-05, "loss": 0.0211, "step": 99760 }, { "grad_norm": 0.1814587116241455, "learning_rate": 7.971425426084059e-05, "loss": 0.0194, "step": 99770 }, { "grad_norm": 0.1046004593372345, "learning_rate": 7.970982137873333e-05, "loss": 0.0186, "step": 99780 }, { "grad_norm": 0.1169830709695816, "learning_rate": 7.970538813562371e-05, "loss": 0.0203, "step": 99790 }, { "grad_norm": 0.1130852922797203, "learning_rate": 7.970095453156556e-05, "loss": 0.0171, "step": 99800 }, { "grad_norm": 0.23239581286907196, "learning_rate": 7.969652056661277e-05, "loss": 0.0152, "step": 99810 }, { "grad_norm": 0.15478581190109253, "learning_rate": 7.969208624081924e-05, "loss": 0.0175, "step": 99820 }, { "grad_norm": 0.3393002152442932, "learning_rate": 7.96876515542388e-05, "loss": 0.0174, "step": 99830 }, { "grad_norm": 0.15772892534732819, "learning_rate": 7.968321650692536e-05, "loss": 0.0152, "step": 99840 }, { "grad_norm": 0.16176529228687286, "learning_rate": 7.967878109893283e-05, "loss": 0.017, "step": 99850 }, { "grad_norm": 0.20333388447761536, "learning_rate": 7.96743453303151e-05, "loss": 0.0208, "step": 99860 }, { "grad_norm": 0.26767808198928833, "learning_rate": 7.966990920112603e-05, "loss": 0.0377, "step": 99870 }, { "grad_norm": 0.1818295419216156, "learning_rate": 7.966547271141957e-05, "loss": 0.0166, "step": 99880 }, { "grad_norm": 0.18386943638324738, "learning_rate": 7.966103586124959e-05, "loss": 0.0204, "step": 99890 }, { "grad_norm": 0.20455166697502136, "learning_rate": 7.965659865067004e-05, "loss": 0.0198, "step": 99900 }, { "grad_norm": 0.21562500298023224, "learning_rate": 7.965216107973481e-05, "loss": 0.0173, "step": 99910 }, { "grad_norm": 0.10978995263576508, "learning_rate": 7.964772314849782e-05, "loss": 0.0218, "step": 99920 }, { "grad_norm": 0.2944597601890564, "learning_rate": 7.9643284857013e-05, "loss": 0.016, "step": 99930 }, { "grad_norm": 0.09164228290319443, "learning_rate": 7.963884620533428e-05, "loss": 0.0163, "step": 99940 }, { "grad_norm": 0.17068105936050415, "learning_rate": 7.96344071935156e-05, "loss": 0.013, "step": 99950 }, { "grad_norm": 0.1446647346019745, "learning_rate": 7.962996782161089e-05, "loss": 0.0222, "step": 99960 }, { "grad_norm": 0.15170669555664062, "learning_rate": 7.962552808967412e-05, "loss": 0.0187, "step": 99970 }, { "grad_norm": 0.11035847663879395, "learning_rate": 7.962108799775917e-05, "loss": 0.0167, "step": 99980 }, { "grad_norm": 0.1661766916513443, "learning_rate": 7.961664754592006e-05, "loss": 0.0168, "step": 99990 }, { "grad_norm": 0.09763991832733154, "learning_rate": 7.961220673421072e-05, "loss": 0.0155, "step": 100000 }, { "grad_norm": 0.15743838250637054, "learning_rate": 7.960776556268512e-05, "loss": 0.0306, "step": 100010 }, { "grad_norm": 0.09636972099542618, "learning_rate": 7.960332403139719e-05, "loss": 0.0167, "step": 100020 }, { "grad_norm": 0.1352142095565796, "learning_rate": 7.959888214040093e-05, "loss": 0.0162, "step": 100030 }, { "grad_norm": 0.07965526729822159, "learning_rate": 7.95944398897503e-05, "loss": 0.0181, "step": 100040 }, { "grad_norm": 0.13584285974502563, "learning_rate": 7.95899972794993e-05, "loss": 0.0168, "step": 100050 }, { "grad_norm": 0.1981627643108368, "learning_rate": 7.958555430970189e-05, "loss": 0.014, "step": 100060 }, { "grad_norm": 0.16584427654743195, "learning_rate": 7.958111098041206e-05, "loss": 0.0129, "step": 100070 }, { "grad_norm": 0.12379296869039536, "learning_rate": 7.957666729168381e-05, "loss": 0.017, "step": 100080 }, { "grad_norm": 0.08584072440862656, "learning_rate": 7.957222324357111e-05, "loss": 0.0108, "step": 100090 }, { "grad_norm": 0.10978443920612335, "learning_rate": 7.956777883612799e-05, "loss": 0.0231, "step": 100100 }, { "grad_norm": 0.09486430138349533, "learning_rate": 7.956333406940843e-05, "loss": 0.0191, "step": 100110 }, { "grad_norm": 0.09123032540082932, "learning_rate": 7.955888894346646e-05, "loss": 0.0129, "step": 100120 }, { "grad_norm": 0.38057848811149597, "learning_rate": 7.955444345835607e-05, "loss": 0.0224, "step": 100130 }, { "grad_norm": 0.17510288953781128, "learning_rate": 7.954999761413128e-05, "loss": 0.0143, "step": 100140 }, { "grad_norm": 0.08279234170913696, "learning_rate": 7.954555141084614e-05, "loss": 0.0188, "step": 100150 }, { "grad_norm": 0.1335991472005844, "learning_rate": 7.954110484855465e-05, "loss": 0.0163, "step": 100160 }, { "grad_norm": 0.19532130658626556, "learning_rate": 7.953665792731084e-05, "loss": 0.0157, "step": 100170 }, { "grad_norm": 0.12587237358093262, "learning_rate": 7.953221064716876e-05, "loss": 0.0225, "step": 100180 }, { "grad_norm": 0.15259695053100586, "learning_rate": 7.952776300818243e-05, "loss": 0.0178, "step": 100190 }, { "grad_norm": 0.1274176388978958, "learning_rate": 7.95233150104059e-05, "loss": 0.0211, "step": 100200 }, { "grad_norm": 0.12081122398376465, "learning_rate": 7.951886665389322e-05, "loss": 0.0179, "step": 100210 }, { "grad_norm": 0.15176251530647278, "learning_rate": 7.951441793869843e-05, "loss": 0.0168, "step": 100220 }, { "grad_norm": 0.13284198939800262, "learning_rate": 7.95099688648756e-05, "loss": 0.0172, "step": 100230 }, { "grad_norm": 0.2185387760400772, "learning_rate": 7.950551943247879e-05, "loss": 0.022, "step": 100240 }, { "grad_norm": 0.2326173186302185, "learning_rate": 7.950106964156206e-05, "loss": 0.0152, "step": 100250 }, { "grad_norm": 0.369457483291626, "learning_rate": 7.949661949217948e-05, "loss": 0.0193, "step": 100260 }, { "grad_norm": 0.1432572901248932, "learning_rate": 7.949216898438512e-05, "loss": 0.018, "step": 100270 }, { "grad_norm": 0.09683030843734741, "learning_rate": 7.948771811823307e-05, "loss": 0.0179, "step": 100280 }, { "grad_norm": 0.1788514405488968, "learning_rate": 7.94832668937774e-05, "loss": 0.0177, "step": 100290 }, { "grad_norm": 0.18459418416023254, "learning_rate": 7.94788153110722e-05, "loss": 0.0212, "step": 100300 }, { "grad_norm": 0.11231325566768646, "learning_rate": 7.947436337017155e-05, "loss": 0.0237, "step": 100310 }, { "grad_norm": 0.46814480423927307, "learning_rate": 7.946991107112957e-05, "loss": 0.0179, "step": 100320 }, { "grad_norm": 0.23668649792671204, "learning_rate": 7.946545841400035e-05, "loss": 0.0181, "step": 100330 }, { "grad_norm": 0.1760917752981186, "learning_rate": 7.946100539883798e-05, "loss": 0.0194, "step": 100340 }, { "grad_norm": 0.11301375925540924, "learning_rate": 7.945655202569659e-05, "loss": 0.0158, "step": 100350 }, { "grad_norm": 0.13207146525382996, "learning_rate": 7.945209829463027e-05, "loss": 0.0165, "step": 100360 }, { "grad_norm": 0.13863392174243927, "learning_rate": 7.944764420569314e-05, "loss": 0.0156, "step": 100370 }, { "grad_norm": 0.2229968160390854, "learning_rate": 7.944318975893936e-05, "loss": 0.0208, "step": 100380 }, { "grad_norm": 0.1364854872226715, "learning_rate": 7.9438734954423e-05, "loss": 0.0174, "step": 100390 }, { "grad_norm": 0.10287787765264511, "learning_rate": 7.943427979219822e-05, "loss": 0.0133, "step": 100400 }, { "grad_norm": 0.31138697266578674, "learning_rate": 7.942982427231916e-05, "loss": 0.0168, "step": 100410 }, { "grad_norm": 0.15338599681854248, "learning_rate": 7.942536839483995e-05, "loss": 0.0202, "step": 100420 }, { "grad_norm": 0.22308017313480377, "learning_rate": 7.942091215981473e-05, "loss": 0.0208, "step": 100430 }, { "grad_norm": 0.24338868260383606, "learning_rate": 7.941645556729763e-05, "loss": 0.027, "step": 100440 }, { "grad_norm": 0.10259867459535599, "learning_rate": 7.941199861734285e-05, "loss": 0.0245, "step": 100450 }, { "grad_norm": 0.15937639772891998, "learning_rate": 7.940754131000449e-05, "loss": 0.0208, "step": 100460 }, { "grad_norm": 0.16222624480724335, "learning_rate": 7.940308364533675e-05, "loss": 0.0189, "step": 100470 }, { "grad_norm": 0.17233888804912567, "learning_rate": 7.939862562339378e-05, "loss": 0.0163, "step": 100480 }, { "grad_norm": 0.10705933719873428, "learning_rate": 7.939416724422976e-05, "loss": 0.0137, "step": 100490 }, { "grad_norm": 0.18215402960777283, "learning_rate": 7.938970850789886e-05, "loss": 0.0178, "step": 100500 }, { "grad_norm": 0.1072247326374054, "learning_rate": 7.938524941445522e-05, "loss": 0.0183, "step": 100510 }, { "grad_norm": 0.07163470983505249, "learning_rate": 7.938078996395309e-05, "loss": 0.0148, "step": 100520 }, { "grad_norm": 0.1251349002122879, "learning_rate": 7.937633015644661e-05, "loss": 0.0163, "step": 100530 }, { "grad_norm": 0.09862309694290161, "learning_rate": 7.937186999198997e-05, "loss": 0.0141, "step": 100540 }, { "grad_norm": 0.227819561958313, "learning_rate": 7.936740947063738e-05, "loss": 0.0174, "step": 100550 }, { "grad_norm": 0.14418743550777435, "learning_rate": 7.936294859244307e-05, "loss": 0.0217, "step": 100560 }, { "grad_norm": 0.16676828265190125, "learning_rate": 7.935848735746119e-05, "loss": 0.0142, "step": 100570 }, { "grad_norm": 0.1568422168493271, "learning_rate": 7.935402576574597e-05, "loss": 0.0224, "step": 100580 }, { "grad_norm": 0.22172848880290985, "learning_rate": 7.934956381735161e-05, "loss": 0.0144, "step": 100590 }, { "grad_norm": 0.4844871759414673, "learning_rate": 7.934510151233236e-05, "loss": 0.0171, "step": 100600 }, { "grad_norm": 0.35857436060905457, "learning_rate": 7.93406388507424e-05, "loss": 0.0189, "step": 100610 }, { "grad_norm": 0.20946311950683594, "learning_rate": 7.9336175832636e-05, "loss": 0.0177, "step": 100620 }, { "grad_norm": 0.23504440486431122, "learning_rate": 7.933171245806734e-05, "loss": 0.0243, "step": 100630 }, { "grad_norm": 0.27656036615371704, "learning_rate": 7.93272487270907e-05, "loss": 0.0256, "step": 100640 }, { "grad_norm": 0.1761731058359146, "learning_rate": 7.932278463976028e-05, "loss": 0.0201, "step": 100650 }, { "grad_norm": 0.10401330143213272, "learning_rate": 7.931832019613035e-05, "loss": 0.0247, "step": 100660 }, { "grad_norm": 0.14449089765548706, "learning_rate": 7.931385539625515e-05, "loss": 0.0155, "step": 100670 }, { "grad_norm": 0.08767588436603546, "learning_rate": 7.930939024018893e-05, "loss": 0.0186, "step": 100680 }, { "grad_norm": 0.11480472981929779, "learning_rate": 7.930492472798594e-05, "loss": 0.0131, "step": 100690 }, { "grad_norm": 0.15817996859550476, "learning_rate": 7.930045885970047e-05, "loss": 0.0135, "step": 100700 }, { "grad_norm": 0.2051147222518921, "learning_rate": 7.929599263538673e-05, "loss": 0.0159, "step": 100710 }, { "grad_norm": 0.19005891680717468, "learning_rate": 7.929152605509904e-05, "loss": 0.0246, "step": 100720 }, { "grad_norm": 0.2655092477798462, "learning_rate": 7.928705911889164e-05, "loss": 0.0134, "step": 100730 }, { "grad_norm": 0.1457352489233017, "learning_rate": 7.928259182681883e-05, "loss": 0.0184, "step": 100740 }, { "grad_norm": 0.12528251111507416, "learning_rate": 7.927812417893487e-05, "loss": 0.0205, "step": 100750 }, { "grad_norm": 0.13568872213363647, "learning_rate": 7.927365617529407e-05, "loss": 0.0161, "step": 100760 }, { "grad_norm": 0.13998298346996307, "learning_rate": 7.92691878159507e-05, "loss": 0.0193, "step": 100770 }, { "grad_norm": 0.22233738005161285, "learning_rate": 7.926471910095907e-05, "loss": 0.0158, "step": 100780 }, { "grad_norm": 0.13702258467674255, "learning_rate": 7.926025003037348e-05, "loss": 0.0187, "step": 100790 }, { "grad_norm": 0.1283722072839737, "learning_rate": 7.925578060424819e-05, "loss": 0.0211, "step": 100800 }, { "grad_norm": 0.1531277745962143, "learning_rate": 7.925131082263758e-05, "loss": 0.0176, "step": 100810 }, { "grad_norm": 0.11789628863334656, "learning_rate": 7.924684068559592e-05, "loss": 0.0136, "step": 100820 }, { "grad_norm": 0.11862687021493912, "learning_rate": 7.924237019317753e-05, "loss": 0.0179, "step": 100830 }, { "grad_norm": 0.15761445462703705, "learning_rate": 7.923789934543673e-05, "loss": 0.0165, "step": 100840 }, { "grad_norm": 0.1241445392370224, "learning_rate": 7.923342814242785e-05, "loss": 0.0202, "step": 100850 }, { "grad_norm": 0.11860969662666321, "learning_rate": 7.92289565842052e-05, "loss": 0.0127, "step": 100860 }, { "grad_norm": 0.079643614590168, "learning_rate": 7.922448467082315e-05, "loss": 0.0118, "step": 100870 }, { "grad_norm": 0.09204116463661194, "learning_rate": 7.922001240233601e-05, "loss": 0.0153, "step": 100880 }, { "grad_norm": 0.10618190467357635, "learning_rate": 7.921553977879815e-05, "loss": 0.0155, "step": 100890 }, { "grad_norm": 0.2669060528278351, "learning_rate": 7.921106680026387e-05, "loss": 0.0182, "step": 100900 }, { "grad_norm": 0.12080114334821701, "learning_rate": 7.920659346678755e-05, "loss": 0.0164, "step": 100910 }, { "grad_norm": 0.08409317582845688, "learning_rate": 7.920211977842357e-05, "loss": 0.0115, "step": 100920 }, { "grad_norm": 0.2637283205986023, "learning_rate": 7.919764573522624e-05, "loss": 0.0187, "step": 100930 }, { "grad_norm": 0.15138019621372223, "learning_rate": 7.919317133724996e-05, "loss": 0.0174, "step": 100940 }, { "grad_norm": 0.07521034777164459, "learning_rate": 7.918869658454908e-05, "loss": 0.0151, "step": 100950 }, { "grad_norm": 0.14510409533977509, "learning_rate": 7.918422147717798e-05, "loss": 0.014, "step": 100960 }, { "grad_norm": 0.183885395526886, "learning_rate": 7.917974601519102e-05, "loss": 0.0186, "step": 100970 }, { "grad_norm": 0.2890075147151947, "learning_rate": 7.917527019864262e-05, "loss": 0.017, "step": 100980 }, { "grad_norm": 0.26498571038246155, "learning_rate": 7.917079402758713e-05, "loss": 0.0182, "step": 100990 }, { "grad_norm": 0.13781902194023132, "learning_rate": 7.916631750207894e-05, "loss": 0.0179, "step": 101000 }, { "grad_norm": 0.1515299677848816, "learning_rate": 7.916184062217247e-05, "loss": 0.0205, "step": 101010 }, { "grad_norm": 1.2981518507003784, "learning_rate": 7.91573633879221e-05, "loss": 0.0216, "step": 101020 }, { "grad_norm": 0.20019061863422394, "learning_rate": 7.915288579938224e-05, "loss": 0.022, "step": 101030 }, { "grad_norm": 0.10042297840118408, "learning_rate": 7.914840785660729e-05, "loss": 0.0157, "step": 101040 }, { "grad_norm": 0.14880253374576569, "learning_rate": 7.914392955965165e-05, "loss": 0.021, "step": 101050 }, { "grad_norm": 0.18886636197566986, "learning_rate": 7.913945090856976e-05, "loss": 0.0236, "step": 101060 }, { "grad_norm": 0.177069291472435, "learning_rate": 7.913497190341603e-05, "loss": 0.0186, "step": 101070 }, { "grad_norm": 0.10675152391195297, "learning_rate": 7.913049254424489e-05, "loss": 0.0168, "step": 101080 }, { "grad_norm": 0.11811511218547821, "learning_rate": 7.912601283111077e-05, "loss": 0.0184, "step": 101090 }, { "grad_norm": 0.09732223302125931, "learning_rate": 7.912153276406807e-05, "loss": 0.0167, "step": 101100 }, { "grad_norm": 0.18930687010288239, "learning_rate": 7.911705234317126e-05, "loss": 0.0196, "step": 101110 }, { "grad_norm": 0.368979811668396, "learning_rate": 7.911257156847476e-05, "loss": 0.0211, "step": 101120 }, { "grad_norm": 0.15582352876663208, "learning_rate": 7.910809044003304e-05, "loss": 0.0176, "step": 101130 }, { "grad_norm": 0.12969675660133362, "learning_rate": 7.910360895790055e-05, "loss": 0.0189, "step": 101140 }, { "grad_norm": 0.19296376407146454, "learning_rate": 7.909912712213172e-05, "loss": 0.0211, "step": 101150 }, { "grad_norm": 0.2178696244955063, "learning_rate": 7.909464493278103e-05, "loss": 0.0193, "step": 101160 }, { "grad_norm": 0.18104784190654755, "learning_rate": 7.909016238990292e-05, "loss": 0.0172, "step": 101170 }, { "grad_norm": 0.11601541936397552, "learning_rate": 7.90856794935519e-05, "loss": 0.016, "step": 101180 }, { "grad_norm": 0.16225595772266388, "learning_rate": 7.908119624378238e-05, "loss": 0.0209, "step": 101190 }, { "grad_norm": 0.17521719634532928, "learning_rate": 7.907671264064888e-05, "loss": 0.0186, "step": 101200 }, { "grad_norm": 0.21932893991470337, "learning_rate": 7.907222868420587e-05, "loss": 0.0162, "step": 101210 }, { "grad_norm": 0.13087217509746552, "learning_rate": 7.906774437450784e-05, "loss": 0.0189, "step": 101220 }, { "grad_norm": 0.17266151309013367, "learning_rate": 7.906325971160927e-05, "loss": 0.0205, "step": 101230 }, { "grad_norm": 0.1205463781952858, "learning_rate": 7.905877469556464e-05, "loss": 0.0183, "step": 101240 }, { "grad_norm": 0.322070449590683, "learning_rate": 7.905428932642848e-05, "loss": 0.0216, "step": 101250 }, { "grad_norm": 0.1042952761054039, "learning_rate": 7.904980360425525e-05, "loss": 0.0247, "step": 101260 }, { "grad_norm": 0.1539156287908554, "learning_rate": 7.904531752909949e-05, "loss": 0.019, "step": 101270 }, { "grad_norm": 0.13689786195755005, "learning_rate": 7.90408311010157e-05, "loss": 0.0142, "step": 101280 }, { "grad_norm": 0.09752862900495529, "learning_rate": 7.90363443200584e-05, "loss": 0.0206, "step": 101290 }, { "grad_norm": 0.12995468080043793, "learning_rate": 7.90318571862821e-05, "loss": 0.0249, "step": 101300 }, { "grad_norm": 0.30985119938850403, "learning_rate": 7.902736969974132e-05, "loss": 0.0226, "step": 101310 }, { "grad_norm": 0.10571135580539703, "learning_rate": 7.902288186049059e-05, "loss": 0.0172, "step": 101320 }, { "grad_norm": 0.2566550672054291, "learning_rate": 7.901839366858443e-05, "loss": 0.0197, "step": 101330 }, { "grad_norm": 0.26292791962623596, "learning_rate": 7.901390512407741e-05, "loss": 0.0193, "step": 101340 }, { "grad_norm": 0.12777303159236908, "learning_rate": 7.900941622702404e-05, "loss": 0.022, "step": 101350 }, { "grad_norm": 0.14594772458076477, "learning_rate": 7.900492697747887e-05, "loss": 0.0162, "step": 101360 }, { "grad_norm": 0.10494711995124817, "learning_rate": 7.900043737549647e-05, "loss": 0.0172, "step": 101370 }, { "grad_norm": 0.1526685357093811, "learning_rate": 7.899594742113136e-05, "loss": 0.0161, "step": 101380 }, { "grad_norm": 0.23052699863910675, "learning_rate": 7.899145711443812e-05, "loss": 0.0186, "step": 101390 }, { "grad_norm": 0.13611425459384918, "learning_rate": 7.89869664554713e-05, "loss": 0.0227, "step": 101400 }, { "grad_norm": 0.1195727065205574, "learning_rate": 7.898247544428547e-05, "loss": 0.0222, "step": 101410 }, { "grad_norm": 0.12299350649118423, "learning_rate": 7.89779840809352e-05, "loss": 0.0153, "step": 101420 }, { "grad_norm": 0.08857397735118866, "learning_rate": 7.897349236547506e-05, "loss": 0.018, "step": 101430 }, { "grad_norm": 0.20221711695194244, "learning_rate": 7.896900029795961e-05, "loss": 0.0115, "step": 101440 }, { "grad_norm": 0.2783205509185791, "learning_rate": 7.89645078784435e-05, "loss": 0.0159, "step": 101450 }, { "grad_norm": 0.12799391150474548, "learning_rate": 7.896001510698125e-05, "loss": 0.0153, "step": 101460 }, { "grad_norm": 0.3696100115776062, "learning_rate": 7.895552198362748e-05, "loss": 0.0256, "step": 101470 }, { "grad_norm": 0.11109910905361176, "learning_rate": 7.895102850843676e-05, "loss": 0.0212, "step": 101480 }, { "grad_norm": 0.18490619957447052, "learning_rate": 7.894653468146374e-05, "loss": 0.0271, "step": 101490 }, { "grad_norm": 0.1680108606815338, "learning_rate": 7.894204050276297e-05, "loss": 0.0178, "step": 101500 }, { "grad_norm": 0.1328989565372467, "learning_rate": 7.893754597238908e-05, "loss": 0.0193, "step": 101510 }, { "grad_norm": 0.25366878509521484, "learning_rate": 7.893305109039671e-05, "loss": 0.0219, "step": 101520 }, { "grad_norm": 0.19696299731731415, "learning_rate": 7.892855585684043e-05, "loss": 0.0195, "step": 101530 }, { "grad_norm": 0.1449262499809265, "learning_rate": 7.892406027177488e-05, "loss": 0.017, "step": 101540 }, { "grad_norm": 0.16757184267044067, "learning_rate": 7.891956433525471e-05, "loss": 0.0197, "step": 101550 }, { "grad_norm": 0.20184428989887238, "learning_rate": 7.89150680473345e-05, "loss": 0.0242, "step": 101560 }, { "grad_norm": 0.08268193900585175, "learning_rate": 7.891057140806895e-05, "loss": 0.0158, "step": 101570 }, { "grad_norm": 0.18154732882976532, "learning_rate": 7.890607441751263e-05, "loss": 0.0156, "step": 101580 }, { "grad_norm": 0.07887069880962372, "learning_rate": 7.890157707572023e-05, "loss": 0.0208, "step": 101590 }, { "grad_norm": 0.23692874610424042, "learning_rate": 7.889707938274636e-05, "loss": 0.0234, "step": 101600 }, { "grad_norm": 0.12014973908662796, "learning_rate": 7.889258133864572e-05, "loss": 0.0187, "step": 101610 }, { "grad_norm": 0.40506690740585327, "learning_rate": 7.888808294347291e-05, "loss": 0.021, "step": 101620 }, { "grad_norm": 0.16877536475658417, "learning_rate": 7.888358419728263e-05, "loss": 0.0198, "step": 101630 }, { "grad_norm": 0.12465576082468033, "learning_rate": 7.887908510012954e-05, "loss": 0.0185, "step": 101640 }, { "grad_norm": 0.15606622397899628, "learning_rate": 7.88745856520683e-05, "loss": 0.0206, "step": 101650 }, { "grad_norm": 0.16735631227493286, "learning_rate": 7.887008585315357e-05, "loss": 0.0179, "step": 101660 }, { "grad_norm": 0.2450966238975525, "learning_rate": 7.886558570344005e-05, "loss": 0.0209, "step": 101670 }, { "grad_norm": 0.13831394910812378, "learning_rate": 7.886108520298241e-05, "loss": 0.0154, "step": 101680 }, { "grad_norm": 0.13798809051513672, "learning_rate": 7.885658435183532e-05, "loss": 0.0184, "step": 101690 }, { "grad_norm": 0.25033295154571533, "learning_rate": 7.88520831500535e-05, "loss": 0.0218, "step": 101700 }, { "grad_norm": 0.35381537675857544, "learning_rate": 7.884758159769164e-05, "loss": 0.0153, "step": 101710 }, { "grad_norm": 0.21073806285858154, "learning_rate": 7.884307969480439e-05, "loss": 0.0175, "step": 101720 }, { "grad_norm": 0.16575250029563904, "learning_rate": 7.883857744144651e-05, "loss": 0.0161, "step": 101730 }, { "grad_norm": 0.11727634072303772, "learning_rate": 7.883407483767269e-05, "loss": 0.0223, "step": 101740 }, { "grad_norm": 0.1328471601009369, "learning_rate": 7.882957188353764e-05, "loss": 0.0239, "step": 101750 }, { "grad_norm": 0.3053939938545227, "learning_rate": 7.882506857909606e-05, "loss": 0.0177, "step": 101760 }, { "grad_norm": 0.1032266691327095, "learning_rate": 7.882056492440269e-05, "loss": 0.015, "step": 101770 }, { "grad_norm": 0.20016102492809296, "learning_rate": 7.881606091951223e-05, "loss": 0.0261, "step": 101780 }, { "grad_norm": 0.16717012226581573, "learning_rate": 7.881155656447944e-05, "loss": 0.0233, "step": 101790 }, { "grad_norm": 0.17440511286258698, "learning_rate": 7.880705185935904e-05, "loss": 0.021, "step": 101800 }, { "grad_norm": 0.1440117061138153, "learning_rate": 7.880254680420574e-05, "loss": 0.0168, "step": 101810 }, { "grad_norm": 0.20650969445705414, "learning_rate": 7.879804139907431e-05, "loss": 0.0193, "step": 101820 }, { "grad_norm": 0.12432236969470978, "learning_rate": 7.87935356440195e-05, "loss": 0.0185, "step": 101830 }, { "grad_norm": 0.13419467210769653, "learning_rate": 7.878902953909604e-05, "loss": 0.0162, "step": 101840 }, { "grad_norm": 0.3064473271369934, "learning_rate": 7.878452308435867e-05, "loss": 0.0181, "step": 101850 }, { "grad_norm": 0.17625929415225983, "learning_rate": 7.87800162798622e-05, "loss": 0.0175, "step": 101860 }, { "grad_norm": 0.18863660097122192, "learning_rate": 7.877550912566134e-05, "loss": 0.0158, "step": 101870 }, { "grad_norm": 0.16624605655670166, "learning_rate": 7.877100162181088e-05, "loss": 0.0159, "step": 101880 }, { "grad_norm": 0.12713715434074402, "learning_rate": 7.876649376836558e-05, "loss": 0.0199, "step": 101890 }, { "grad_norm": 0.12212806195020676, "learning_rate": 7.876198556538023e-05, "loss": 0.0176, "step": 101900 }, { "grad_norm": 0.20613506436347961, "learning_rate": 7.875747701290959e-05, "loss": 0.0186, "step": 101910 }, { "grad_norm": 0.14474734663963318, "learning_rate": 7.875296811100845e-05, "loss": 0.0171, "step": 101920 }, { "grad_norm": 0.14939875900745392, "learning_rate": 7.874845885973162e-05, "loss": 0.0127, "step": 101930 }, { "grad_norm": 0.12272340059280396, "learning_rate": 7.874394925913385e-05, "loss": 0.0124, "step": 101940 }, { "grad_norm": 0.1283230483531952, "learning_rate": 7.873943930926997e-05, "loss": 0.0215, "step": 101950 }, { "grad_norm": 0.20193220674991608, "learning_rate": 7.873492901019476e-05, "loss": 0.014, "step": 101960 }, { "grad_norm": 0.13624656200408936, "learning_rate": 7.873041836196305e-05, "loss": 0.0189, "step": 101970 }, { "grad_norm": 0.10325963795185089, "learning_rate": 7.872590736462962e-05, "loss": 0.0258, "step": 101980 }, { "grad_norm": 0.13819079101085663, "learning_rate": 7.872139601824929e-05, "loss": 0.0194, "step": 101990 }, { "grad_norm": 0.11765462160110474, "learning_rate": 7.87168843228769e-05, "loss": 0.0164, "step": 102000 }, { "grad_norm": 0.2285917103290558, "learning_rate": 7.871237227856721e-05, "loss": 0.0253, "step": 102010 }, { "grad_norm": 0.17326891422271729, "learning_rate": 7.870785988537512e-05, "loss": 0.0179, "step": 102020 }, { "grad_norm": 0.13143324851989746, "learning_rate": 7.870334714335544e-05, "loss": 0.0226, "step": 102030 }, { "grad_norm": 0.12920498847961426, "learning_rate": 7.869883405256295e-05, "loss": 0.0207, "step": 102040 }, { "grad_norm": 0.20597811043262482, "learning_rate": 7.869432061305255e-05, "loss": 0.0146, "step": 102050 }, { "grad_norm": 0.18797622621059418, "learning_rate": 7.868980682487907e-05, "loss": 0.0227, "step": 102060 }, { "grad_norm": 0.16648411750793457, "learning_rate": 7.868529268809735e-05, "loss": 0.0171, "step": 102070 }, { "grad_norm": 0.19784438610076904, "learning_rate": 7.868077820276223e-05, "loss": 0.0194, "step": 102080 }, { "grad_norm": 0.2540125250816345, "learning_rate": 7.867626336892858e-05, "loss": 0.0184, "step": 102090 }, { "grad_norm": 0.18554873764514923, "learning_rate": 7.867174818665126e-05, "loss": 0.0208, "step": 102100 }, { "grad_norm": 0.22962918877601624, "learning_rate": 7.866723265598512e-05, "loss": 0.0193, "step": 102110 }, { "grad_norm": 0.10345469415187836, "learning_rate": 7.866271677698504e-05, "loss": 0.0148, "step": 102120 }, { "grad_norm": 0.14113272726535797, "learning_rate": 7.86582005497059e-05, "loss": 0.0189, "step": 102130 }, { "grad_norm": 0.19979234039783478, "learning_rate": 7.865368397420256e-05, "loss": 0.0206, "step": 102140 }, { "grad_norm": 0.1897195726633072, "learning_rate": 7.864916705052989e-05, "loss": 0.0159, "step": 102150 }, { "grad_norm": 0.1348043531179428, "learning_rate": 7.864464977874282e-05, "loss": 0.0182, "step": 102160 }, { "grad_norm": 0.12706181406974792, "learning_rate": 7.864013215889618e-05, "loss": 0.0148, "step": 102170 }, { "grad_norm": 0.17207397520542145, "learning_rate": 7.863561419104491e-05, "loss": 0.0211, "step": 102180 }, { "grad_norm": 0.20223194360733032, "learning_rate": 7.863109587524389e-05, "loss": 0.0174, "step": 102190 }, { "grad_norm": 0.1186680942773819, "learning_rate": 7.862657721154802e-05, "loss": 0.0211, "step": 102200 }, { "grad_norm": 0.1232265904545784, "learning_rate": 7.862205820001221e-05, "loss": 0.0179, "step": 102210 }, { "grad_norm": 0.15830443799495697, "learning_rate": 7.861753884069137e-05, "loss": 0.016, "step": 102220 }, { "grad_norm": 0.10431817919015884, "learning_rate": 7.861301913364042e-05, "loss": 0.0206, "step": 102230 }, { "grad_norm": 0.16895832121372223, "learning_rate": 7.860849907891425e-05, "loss": 0.0191, "step": 102240 }, { "grad_norm": 0.15458382666110992, "learning_rate": 7.860397867656784e-05, "loss": 0.0141, "step": 102250 }, { "grad_norm": 0.08560466021299362, "learning_rate": 7.859945792665605e-05, "loss": 0.0166, "step": 102260 }, { "grad_norm": 0.11686054617166519, "learning_rate": 7.859493682923387e-05, "loss": 0.0199, "step": 102270 }, { "grad_norm": 0.11547117680311203, "learning_rate": 7.859041538435619e-05, "loss": 0.0148, "step": 102280 }, { "grad_norm": 0.13500761985778809, "learning_rate": 7.858589359207797e-05, "loss": 0.0165, "step": 102290 }, { "grad_norm": 0.15670634806156158, "learning_rate": 7.858137145245416e-05, "loss": 0.0174, "step": 102300 }, { "grad_norm": 0.2400086522102356, "learning_rate": 7.85768489655397e-05, "loss": 0.0162, "step": 102310 }, { "grad_norm": 0.1617688536643982, "learning_rate": 7.857232613138955e-05, "loss": 0.0166, "step": 102320 }, { "grad_norm": 0.09109608083963394, "learning_rate": 7.856780295005865e-05, "loss": 0.0148, "step": 102330 }, { "grad_norm": 0.19615153968334198, "learning_rate": 7.856327942160197e-05, "loss": 0.0139, "step": 102340 }, { "grad_norm": 0.14445464313030243, "learning_rate": 7.85587555460745e-05, "loss": 0.0126, "step": 102350 }, { "grad_norm": 0.20515188574790955, "learning_rate": 7.855423132353117e-05, "loss": 0.024, "step": 102360 }, { "grad_norm": 0.09163273125886917, "learning_rate": 7.854970675402696e-05, "loss": 0.0169, "step": 102370 }, { "grad_norm": 0.17462605237960815, "learning_rate": 7.854518183761688e-05, "loss": 0.0149, "step": 102380 }, { "grad_norm": 0.17182041704654694, "learning_rate": 7.854065657435588e-05, "loss": 0.0161, "step": 102390 }, { "grad_norm": 0.17660994827747345, "learning_rate": 7.853613096429894e-05, "loss": 0.0202, "step": 102400 }, { "grad_norm": 0.3830774128437042, "learning_rate": 7.853160500750107e-05, "loss": 0.0156, "step": 102410 }, { "grad_norm": 0.2259218990802765, "learning_rate": 7.852707870401728e-05, "loss": 0.0215, "step": 102420 }, { "grad_norm": 0.13692204654216766, "learning_rate": 7.852255205390253e-05, "loss": 0.0193, "step": 102430 }, { "grad_norm": 0.14731605350971222, "learning_rate": 7.851802505721186e-05, "loss": 0.0184, "step": 102440 }, { "grad_norm": 0.11520332098007202, "learning_rate": 7.851349771400025e-05, "loss": 0.0194, "step": 102450 }, { "grad_norm": 0.17766141891479492, "learning_rate": 7.850897002432274e-05, "loss": 0.018, "step": 102460 }, { "grad_norm": 0.14280295372009277, "learning_rate": 7.850444198823431e-05, "loss": 0.0165, "step": 102470 }, { "grad_norm": 0.1315905749797821, "learning_rate": 7.849991360578999e-05, "loss": 0.016, "step": 102480 }, { "grad_norm": 0.14488275349140167, "learning_rate": 7.849538487704483e-05, "loss": 0.0148, "step": 102490 }, { "grad_norm": 0.1507508009672165, "learning_rate": 7.849085580205382e-05, "loss": 0.0154, "step": 102500 }, { "grad_norm": 0.07915201783180237, "learning_rate": 7.848632638087202e-05, "loss": 0.0224, "step": 102510 }, { "grad_norm": 0.22546058893203735, "learning_rate": 7.848179661355446e-05, "loss": 0.0183, "step": 102520 }, { "grad_norm": 0.2681324779987335, "learning_rate": 7.847726650015619e-05, "loss": 0.0221, "step": 102530 }, { "grad_norm": 0.2925804853439331, "learning_rate": 7.847273604073222e-05, "loss": 0.0158, "step": 102540 }, { "grad_norm": 0.10490187257528305, "learning_rate": 7.846820523533764e-05, "loss": 0.0152, "step": 102550 }, { "grad_norm": 0.09941636025905609, "learning_rate": 7.846367408402749e-05, "loss": 0.0193, "step": 102560 }, { "grad_norm": 0.08645401149988174, "learning_rate": 7.845914258685682e-05, "loss": 0.0163, "step": 102570 }, { "grad_norm": 0.10605521500110626, "learning_rate": 7.84546107438807e-05, "loss": 0.0162, "step": 102580 }, { "grad_norm": 0.1067020520567894, "learning_rate": 7.84500785551542e-05, "loss": 0.0178, "step": 102590 }, { "grad_norm": 0.13995839655399323, "learning_rate": 7.844554602073238e-05, "loss": 0.0224, "step": 102600 }, { "grad_norm": 0.0790572464466095, "learning_rate": 7.844101314067032e-05, "loss": 0.019, "step": 102610 }, { "grad_norm": 0.15070699155330658, "learning_rate": 7.843647991502309e-05, "loss": 0.0206, "step": 102620 }, { "grad_norm": 0.1429881453514099, "learning_rate": 7.843194634384578e-05, "loss": 0.0164, "step": 102630 }, { "grad_norm": 0.12465198338031769, "learning_rate": 7.84274124271935e-05, "loss": 0.0209, "step": 102640 }, { "grad_norm": 0.15471462905406952, "learning_rate": 7.84228781651213e-05, "loss": 0.0175, "step": 102650 }, { "grad_norm": 0.14374902844429016, "learning_rate": 7.841834355768429e-05, "loss": 0.0147, "step": 102660 }, { "grad_norm": 0.09519828855991364, "learning_rate": 7.841380860493759e-05, "loss": 0.0176, "step": 102670 }, { "grad_norm": 0.11877135187387466, "learning_rate": 7.840927330693629e-05, "loss": 0.0163, "step": 102680 }, { "grad_norm": 0.22122299671173096, "learning_rate": 7.840473766373548e-05, "loss": 0.0166, "step": 102690 }, { "grad_norm": 0.20166882872581482, "learning_rate": 7.840020167539028e-05, "loss": 0.0253, "step": 102700 }, { "grad_norm": 0.27076858282089233, "learning_rate": 7.839566534195584e-05, "loss": 0.0233, "step": 102710 }, { "grad_norm": 0.22620631754398346, "learning_rate": 7.839112866348724e-05, "loss": 0.0172, "step": 102720 }, { "grad_norm": 0.11739300936460495, "learning_rate": 7.838659164003963e-05, "loss": 0.0141, "step": 102730 }, { "grad_norm": 0.1938081830739975, "learning_rate": 7.838205427166814e-05, "loss": 0.0184, "step": 102740 }, { "grad_norm": 0.13666464388370514, "learning_rate": 7.837751655842788e-05, "loss": 0.021, "step": 102750 }, { "grad_norm": 0.14169180393218994, "learning_rate": 7.837297850037399e-05, "loss": 0.0187, "step": 102760 }, { "grad_norm": 0.10548200458288193, "learning_rate": 7.836844009756164e-05, "loss": 0.014, "step": 102770 }, { "grad_norm": 0.19143027067184448, "learning_rate": 7.836390135004596e-05, "loss": 0.0165, "step": 102780 }, { "grad_norm": 0.12238288670778275, "learning_rate": 7.83593622578821e-05, "loss": 0.0233, "step": 102790 }, { "grad_norm": 0.17182676494121552, "learning_rate": 7.83548228211252e-05, "loss": 0.024, "step": 102800 }, { "grad_norm": 0.12322485446929932, "learning_rate": 7.835028303983043e-05, "loss": 0.017, "step": 102810 }, { "grad_norm": 0.14263461530208588, "learning_rate": 7.834574291405297e-05, "loss": 0.017, "step": 102820 }, { "grad_norm": 0.46288633346557617, "learning_rate": 7.834120244384795e-05, "loss": 0.0211, "step": 102830 }, { "grad_norm": 0.5006077289581299, "learning_rate": 7.833666162927058e-05, "loss": 0.0221, "step": 102840 }, { "grad_norm": 0.2285572588443756, "learning_rate": 7.833212047037602e-05, "loss": 0.0192, "step": 102850 }, { "grad_norm": 0.1649094820022583, "learning_rate": 7.832757896721943e-05, "loss": 0.0151, "step": 102860 }, { "grad_norm": 0.12604020535945892, "learning_rate": 7.832303711985602e-05, "loss": 0.0169, "step": 102870 }, { "grad_norm": 0.6517438888549805, "learning_rate": 7.831849492834096e-05, "loss": 0.0229, "step": 102880 }, { "grad_norm": 0.14209552109241486, "learning_rate": 7.831395239272945e-05, "loss": 0.0139, "step": 102890 }, { "grad_norm": 0.18109482526779175, "learning_rate": 7.830940951307669e-05, "loss": 0.0188, "step": 102900 }, { "grad_norm": 0.12756583094596863, "learning_rate": 7.830486628943788e-05, "loss": 0.0176, "step": 102910 }, { "grad_norm": 0.15389129519462585, "learning_rate": 7.830032272186822e-05, "loss": 0.0213, "step": 102920 }, { "grad_norm": 0.22155176103115082, "learning_rate": 7.829577881042291e-05, "loss": 0.0168, "step": 102930 }, { "grad_norm": 0.18537771701812744, "learning_rate": 7.829123455515718e-05, "loss": 0.0196, "step": 102940 }, { "grad_norm": 0.11283069849014282, "learning_rate": 7.828668995612622e-05, "loss": 0.0196, "step": 102950 }, { "grad_norm": 0.24809622764587402, "learning_rate": 7.82821450133853e-05, "loss": 0.0158, "step": 102960 }, { "grad_norm": 0.1258021891117096, "learning_rate": 7.82775997269896e-05, "loss": 0.0146, "step": 102970 }, { "grad_norm": 0.12356715649366379, "learning_rate": 7.827305409699437e-05, "loss": 0.0141, "step": 102980 }, { "grad_norm": 0.12147783488035202, "learning_rate": 7.826850812345483e-05, "loss": 0.0154, "step": 102990 }, { "grad_norm": 0.32028573751449585, "learning_rate": 7.826396180642623e-05, "loss": 0.0187, "step": 103000 }, { "grad_norm": 0.25941962003707886, "learning_rate": 7.825941514596381e-05, "loss": 0.021, "step": 103010 }, { "grad_norm": 0.2012079507112503, "learning_rate": 7.825486814212282e-05, "loss": 0.0133, "step": 103020 }, { "grad_norm": 0.14849461615085602, "learning_rate": 7.825032079495849e-05, "loss": 0.0177, "step": 103030 }, { "grad_norm": 0.12978756427764893, "learning_rate": 7.82457731045261e-05, "loss": 0.0186, "step": 103040 }, { "grad_norm": 0.19856826961040497, "learning_rate": 7.82412250708809e-05, "loss": 0.0174, "step": 103050 }, { "grad_norm": 0.14008596539497375, "learning_rate": 7.823667669407815e-05, "loss": 0.0186, "step": 103060 }, { "grad_norm": 0.19773374497890472, "learning_rate": 7.823212797417311e-05, "loss": 0.0153, "step": 103070 }, { "grad_norm": 0.11837449669837952, "learning_rate": 7.822757891122107e-05, "loss": 0.0193, "step": 103080 }, { "grad_norm": 0.1618051528930664, "learning_rate": 7.822302950527729e-05, "loss": 0.0176, "step": 103090 }, { "grad_norm": 0.13644041121006012, "learning_rate": 7.821847975639705e-05, "loss": 0.0149, "step": 103100 }, { "grad_norm": 0.12039975076913834, "learning_rate": 7.821392966463564e-05, "loss": 0.0204, "step": 103110 }, { "grad_norm": 0.13433478772640228, "learning_rate": 7.820937923004835e-05, "loss": 0.022, "step": 103120 }, { "grad_norm": 0.1814330667257309, "learning_rate": 7.820482845269047e-05, "loss": 0.0211, "step": 103130 }, { "grad_norm": 0.1249200627207756, "learning_rate": 7.82002773326173e-05, "loss": 0.0219, "step": 103140 }, { "grad_norm": 0.10327727347612381, "learning_rate": 7.819572586988413e-05, "loss": 0.0167, "step": 103150 }, { "grad_norm": 0.1282658725976944, "learning_rate": 7.819117406454626e-05, "loss": 0.018, "step": 103160 }, { "grad_norm": 0.1787179708480835, "learning_rate": 7.818662191665901e-05, "loss": 0.0182, "step": 103170 }, { "grad_norm": 0.18852649629116058, "learning_rate": 7.81820694262777e-05, "loss": 0.0133, "step": 103180 }, { "grad_norm": 0.1069706380367279, "learning_rate": 7.817751659345765e-05, "loss": 0.0131, "step": 103190 }, { "grad_norm": 0.17754285037517548, "learning_rate": 7.817296341825413e-05, "loss": 0.0156, "step": 103200 }, { "grad_norm": 0.22447633743286133, "learning_rate": 7.816840990072254e-05, "loss": 0.0159, "step": 103210 }, { "grad_norm": 0.2427491694688797, "learning_rate": 7.816385604091815e-05, "loss": 0.0166, "step": 103220 }, { "grad_norm": 0.11530563980340958, "learning_rate": 7.815930183889635e-05, "loss": 0.0164, "step": 103230 }, { "grad_norm": 0.13702233135700226, "learning_rate": 7.815474729471241e-05, "loss": 0.0149, "step": 103240 }, { "grad_norm": 0.13602541387081146, "learning_rate": 7.815019240842174e-05, "loss": 0.0193, "step": 103250 }, { "grad_norm": 0.14656485617160797, "learning_rate": 7.814563718007963e-05, "loss": 0.0198, "step": 103260 }, { "grad_norm": 0.22442476451396942, "learning_rate": 7.814108160974148e-05, "loss": 0.0175, "step": 103270 }, { "grad_norm": 0.20632915198802948, "learning_rate": 7.81365256974626e-05, "loss": 0.0184, "step": 103280 }, { "grad_norm": 0.11974942684173584, "learning_rate": 7.813196944329836e-05, "loss": 0.0154, "step": 103290 }, { "grad_norm": 0.1473257839679718, "learning_rate": 7.812741284730413e-05, "loss": 0.017, "step": 103300 }, { "grad_norm": 0.12383320927619934, "learning_rate": 7.81228559095353e-05, "loss": 0.0175, "step": 103310 }, { "grad_norm": 0.18743635714054108, "learning_rate": 7.81182986300472e-05, "loss": 0.0218, "step": 103320 }, { "grad_norm": 0.38590002059936523, "learning_rate": 7.811374100889524e-05, "loss": 0.0186, "step": 103330 }, { "grad_norm": 0.33226457238197327, "learning_rate": 7.810918304613478e-05, "loss": 0.0245, "step": 103340 }, { "grad_norm": 0.11836285144090652, "learning_rate": 7.810462474182119e-05, "loss": 0.019, "step": 103350 }, { "grad_norm": 0.20783311128616333, "learning_rate": 7.810006609600988e-05, "loss": 0.0184, "step": 103360 }, { "grad_norm": 0.09213384240865707, "learning_rate": 7.809550710875624e-05, "loss": 0.0197, "step": 103370 }, { "grad_norm": 0.18604832887649536, "learning_rate": 7.809094778011565e-05, "loss": 0.0208, "step": 103380 }, { "grad_norm": 0.4065151512622833, "learning_rate": 7.808638811014354e-05, "loss": 0.0136, "step": 103390 }, { "grad_norm": 0.2260139137506485, "learning_rate": 7.80818280988953e-05, "loss": 0.0165, "step": 103400 }, { "grad_norm": 0.15535792708396912, "learning_rate": 7.807726774642631e-05, "loss": 0.0144, "step": 103410 }, { "grad_norm": 0.14375044405460358, "learning_rate": 7.807270705279204e-05, "loss": 0.023, "step": 103420 }, { "grad_norm": 0.14493240416049957, "learning_rate": 7.806814601804784e-05, "loss": 0.0192, "step": 103430 }, { "grad_norm": 0.09519188106060028, "learning_rate": 7.806358464224919e-05, "loss": 0.0216, "step": 103440 }, { "grad_norm": 0.09549640864133835, "learning_rate": 7.805902292545147e-05, "loss": 0.0113, "step": 103450 }, { "grad_norm": 0.09839591383934021, "learning_rate": 7.805446086771016e-05, "loss": 0.019, "step": 103460 }, { "grad_norm": 0.1625232994556427, "learning_rate": 7.804989846908063e-05, "loss": 0.0282, "step": 103470 }, { "grad_norm": 0.418364018201828, "learning_rate": 7.804533572961836e-05, "loss": 0.0193, "step": 103480 }, { "grad_norm": 0.08829813450574875, "learning_rate": 7.80407726493788e-05, "loss": 0.0161, "step": 103490 }, { "grad_norm": 0.13082186877727509, "learning_rate": 7.803620922841736e-05, "loss": 0.014, "step": 103500 }, { "grad_norm": 0.13479627668857574, "learning_rate": 7.80316454667895e-05, "loss": 0.0255, "step": 103510 }, { "grad_norm": 0.14107170701026917, "learning_rate": 7.802708136455071e-05, "loss": 0.0168, "step": 103520 }, { "grad_norm": 0.5771377086639404, "learning_rate": 7.802251692175638e-05, "loss": 0.0208, "step": 103530 }, { "grad_norm": 0.34045034646987915, "learning_rate": 7.801795213846204e-05, "loss": 0.0186, "step": 103540 }, { "grad_norm": 0.12423829734325409, "learning_rate": 7.801338701472311e-05, "loss": 0.018, "step": 103550 }, { "grad_norm": 0.12590596079826355, "learning_rate": 7.800882155059511e-05, "loss": 0.0181, "step": 103560 }, { "grad_norm": 0.1417698711156845, "learning_rate": 7.800425574613345e-05, "loss": 0.0168, "step": 103570 }, { "grad_norm": 0.14051257073879242, "learning_rate": 7.799968960139366e-05, "loss": 0.0195, "step": 103580 }, { "grad_norm": 0.18919523060321808, "learning_rate": 7.79951231164312e-05, "loss": 0.0228, "step": 103590 }, { "grad_norm": 0.1234852746129036, "learning_rate": 7.799055629130156e-05, "loss": 0.0194, "step": 103600 }, { "grad_norm": 0.15936289727687836, "learning_rate": 7.798598912606024e-05, "loss": 0.0201, "step": 103610 }, { "grad_norm": 0.35662490129470825, "learning_rate": 7.798142162076273e-05, "loss": 0.0147, "step": 103620 }, { "grad_norm": 0.3471509516239166, "learning_rate": 7.797685377546451e-05, "loss": 0.026, "step": 103630 }, { "grad_norm": 0.14881153404712677, "learning_rate": 7.797228559022113e-05, "loss": 0.0185, "step": 103640 }, { "grad_norm": 0.24232590198516846, "learning_rate": 7.796771706508805e-05, "loss": 0.0176, "step": 103650 }, { "grad_norm": 0.12164522707462311, "learning_rate": 7.796314820012081e-05, "loss": 0.0161, "step": 103660 }, { "grad_norm": 0.14370839297771454, "learning_rate": 7.795857899537492e-05, "loss": 0.0165, "step": 103670 }, { "grad_norm": 0.1788703203201294, "learning_rate": 7.79540094509059e-05, "loss": 0.015, "step": 103680 }, { "grad_norm": 0.23545509576797485, "learning_rate": 7.794943956676925e-05, "loss": 0.0164, "step": 103690 }, { "grad_norm": 0.13627521693706512, "learning_rate": 7.794486934302054e-05, "loss": 0.0156, "step": 103700 }, { "grad_norm": 0.19972920417785645, "learning_rate": 7.794029877971529e-05, "loss": 0.0177, "step": 103710 }, { "grad_norm": 0.15776975452899933, "learning_rate": 7.793572787690902e-05, "loss": 0.0193, "step": 103720 }, { "grad_norm": 0.08591251820325851, "learning_rate": 7.793115663465727e-05, "loss": 0.016, "step": 103730 }, { "grad_norm": 0.10547355562448502, "learning_rate": 7.792658505301561e-05, "loss": 0.012, "step": 103740 }, { "grad_norm": 0.17094270884990692, "learning_rate": 7.792201313203957e-05, "loss": 0.0139, "step": 103750 }, { "grad_norm": 0.1578175127506256, "learning_rate": 7.791744087178471e-05, "loss": 0.017, "step": 103760 }, { "grad_norm": 0.11269477009773254, "learning_rate": 7.791286827230657e-05, "loss": 0.0174, "step": 103770 }, { "grad_norm": 0.25314798951148987, "learning_rate": 7.790829533366075e-05, "loss": 0.0201, "step": 103780 }, { "grad_norm": 0.12094336748123169, "learning_rate": 7.790372205590277e-05, "loss": 0.0168, "step": 103790 }, { "grad_norm": 0.1989850550889969, "learning_rate": 7.789914843908824e-05, "loss": 0.0188, "step": 103800 }, { "grad_norm": 0.10376306623220444, "learning_rate": 7.78945744832727e-05, "loss": 0.0133, "step": 103810 }, { "grad_norm": 0.12253729999065399, "learning_rate": 7.789000018851176e-05, "loss": 0.0197, "step": 103820 }, { "grad_norm": 0.13434456288814545, "learning_rate": 7.788542555486098e-05, "loss": 0.0152, "step": 103830 }, { "grad_norm": 0.3152373731136322, "learning_rate": 7.788085058237594e-05, "loss": 0.0195, "step": 103840 }, { "grad_norm": 0.19987575709819794, "learning_rate": 7.787627527111226e-05, "loss": 0.0174, "step": 103850 }, { "grad_norm": 0.36112669110298157, "learning_rate": 7.78716996211255e-05, "loss": 0.0252, "step": 103860 }, { "grad_norm": 0.17007175087928772, "learning_rate": 7.786712363247129e-05, "loss": 0.0168, "step": 103870 }, { "grad_norm": 0.15178851783275604, "learning_rate": 7.786254730520519e-05, "loss": 0.0214, "step": 103880 }, { "grad_norm": 0.3219456970691681, "learning_rate": 7.785797063938284e-05, "loss": 0.0197, "step": 103890 }, { "grad_norm": 0.23905250430107117, "learning_rate": 7.785339363505987e-05, "loss": 0.0141, "step": 103900 }, { "grad_norm": 0.09358254820108414, "learning_rate": 7.784881629229185e-05, "loss": 0.0176, "step": 103910 }, { "grad_norm": 0.11147378385066986, "learning_rate": 7.784423861113443e-05, "loss": 0.0186, "step": 103920 }, { "grad_norm": 0.24924877285957336, "learning_rate": 7.78396605916432e-05, "loss": 0.0207, "step": 103930 }, { "grad_norm": 0.12356427311897278, "learning_rate": 7.783508223387383e-05, "loss": 0.0195, "step": 103940 }, { "grad_norm": 0.1675875037908554, "learning_rate": 7.783050353788193e-05, "loss": 0.0168, "step": 103950 }, { "grad_norm": 0.13026368618011475, "learning_rate": 7.782592450372312e-05, "loss": 0.0152, "step": 103960 }, { "grad_norm": 0.12379653751850128, "learning_rate": 7.782134513145307e-05, "loss": 0.0264, "step": 103970 }, { "grad_norm": 0.22121800482273102, "learning_rate": 7.781676542112739e-05, "loss": 0.0212, "step": 103980 }, { "grad_norm": 0.155947744846344, "learning_rate": 7.781218537280178e-05, "loss": 0.0206, "step": 103990 }, { "grad_norm": 0.09951136261224747, "learning_rate": 7.780760498653183e-05, "loss": 0.0164, "step": 104000 }, { "grad_norm": 0.27771154046058655, "learning_rate": 7.780302426237323e-05, "loss": 0.0198, "step": 104010 }, { "grad_norm": 0.23942264914512634, "learning_rate": 7.779844320038163e-05, "loss": 0.0186, "step": 104020 }, { "grad_norm": 0.10297129303216934, "learning_rate": 7.779386180061272e-05, "loss": 0.0159, "step": 104030 }, { "grad_norm": 0.08676628768444061, "learning_rate": 7.778928006312212e-05, "loss": 0.018, "step": 104040 }, { "grad_norm": 0.1096409410238266, "learning_rate": 7.778469798796555e-05, "loss": 0.0174, "step": 104050 }, { "grad_norm": 0.08763369172811508, "learning_rate": 7.778011557519867e-05, "loss": 0.0156, "step": 104060 }, { "grad_norm": 0.2418375164270401, "learning_rate": 7.777553282487714e-05, "loss": 0.0161, "step": 104070 }, { "grad_norm": 0.147224560379982, "learning_rate": 7.777094973705668e-05, "loss": 0.0184, "step": 104080 }, { "grad_norm": 0.10494812577962875, "learning_rate": 7.776636631179292e-05, "loss": 0.0168, "step": 104090 }, { "grad_norm": 0.19892948865890503, "learning_rate": 7.776178254914163e-05, "loss": 0.0238, "step": 104100 }, { "grad_norm": 0.12840677797794342, "learning_rate": 7.775719844915848e-05, "loss": 0.02, "step": 104110 }, { "grad_norm": 0.193912535905838, "learning_rate": 7.775261401189914e-05, "loss": 0.0116, "step": 104120 }, { "grad_norm": 0.17968223989009857, "learning_rate": 7.774802923741934e-05, "loss": 0.0175, "step": 104130 }, { "grad_norm": 0.11036614328622818, "learning_rate": 7.774344412577479e-05, "loss": 0.0148, "step": 104140 }, { "grad_norm": 0.08787889033555984, "learning_rate": 7.77388586770212e-05, "loss": 0.0127, "step": 104150 }, { "grad_norm": 0.0976497232913971, "learning_rate": 7.773427289121429e-05, "loss": 0.0179, "step": 104160 }, { "grad_norm": 0.12711219489574432, "learning_rate": 7.772968676840977e-05, "loss": 0.0153, "step": 104170 }, { "grad_norm": 0.07330921292304993, "learning_rate": 7.772510030866339e-05, "loss": 0.0134, "step": 104180 }, { "grad_norm": 0.25921133160591125, "learning_rate": 7.772051351203085e-05, "loss": 0.02, "step": 104190 }, { "grad_norm": 0.26645520329475403, "learning_rate": 7.771592637856789e-05, "loss": 0.02, "step": 104200 }, { "grad_norm": 0.09087251126766205, "learning_rate": 7.771133890833026e-05, "loss": 0.0128, "step": 104210 }, { "grad_norm": 0.11441554129123688, "learning_rate": 7.770675110137371e-05, "loss": 0.0155, "step": 104220 }, { "grad_norm": 0.07501764595508575, "learning_rate": 7.770216295775398e-05, "loss": 0.0148, "step": 104230 }, { "grad_norm": 0.29339975118637085, "learning_rate": 7.76975744775268e-05, "loss": 0.0278, "step": 104240 }, { "grad_norm": 0.21388676762580872, "learning_rate": 7.769298566074794e-05, "loss": 0.0192, "step": 104250 }, { "grad_norm": 0.18356327712535858, "learning_rate": 7.768839650747317e-05, "loss": 0.0165, "step": 104260 }, { "grad_norm": 0.11305078864097595, "learning_rate": 7.768380701775822e-05, "loss": 0.0147, "step": 104270 }, { "grad_norm": 0.24345338344573975, "learning_rate": 7.767921719165889e-05, "loss": 0.0215, "step": 104280 }, { "grad_norm": 0.17070835828781128, "learning_rate": 7.767462702923095e-05, "loss": 0.0171, "step": 104290 }, { "grad_norm": 0.20780934393405914, "learning_rate": 7.767003653053015e-05, "loss": 0.0147, "step": 104300 }, { "grad_norm": 0.18592874705791473, "learning_rate": 7.766544569561228e-05, "loss": 0.0206, "step": 104310 }, { "grad_norm": 0.1793803870677948, "learning_rate": 7.766085452453312e-05, "loss": 0.0168, "step": 104320 }, { "grad_norm": 0.2331584393978119, "learning_rate": 7.765626301734848e-05, "loss": 0.0136, "step": 104330 }, { "grad_norm": 0.16067686676979065, "learning_rate": 7.765167117411412e-05, "loss": 0.0148, "step": 104340 }, { "grad_norm": 0.1159551590681076, "learning_rate": 7.764707899488584e-05, "loss": 0.0153, "step": 104350 }, { "grad_norm": 0.17438462376594543, "learning_rate": 7.764248647971945e-05, "loss": 0.0216, "step": 104360 }, { "grad_norm": 0.27399182319641113, "learning_rate": 7.763789362867075e-05, "loss": 0.0168, "step": 104370 }, { "grad_norm": 0.12241092324256897, "learning_rate": 7.763330044179557e-05, "loss": 0.0166, "step": 104380 }, { "grad_norm": 0.4721856713294983, "learning_rate": 7.762870691914968e-05, "loss": 0.0225, "step": 104390 }, { "grad_norm": 0.13608834147453308, "learning_rate": 7.762411306078892e-05, "loss": 0.0188, "step": 104400 }, { "grad_norm": 0.11423696577548981, "learning_rate": 7.761951886676908e-05, "loss": 0.0172, "step": 104410 }, { "grad_norm": 0.12285684049129486, "learning_rate": 7.761492433714604e-05, "loss": 0.0212, "step": 104420 }, { "grad_norm": 0.12593019008636475, "learning_rate": 7.761032947197558e-05, "loss": 0.0125, "step": 104430 }, { "grad_norm": 0.07327904552221298, "learning_rate": 7.760573427131356e-05, "loss": 0.0142, "step": 104440 }, { "grad_norm": 0.19983629882335663, "learning_rate": 7.760113873521581e-05, "loss": 0.0142, "step": 104450 }, { "grad_norm": 0.2560577094554901, "learning_rate": 7.759654286373814e-05, "loss": 0.017, "step": 104460 }, { "grad_norm": 0.10961107909679413, "learning_rate": 7.759194665693646e-05, "loss": 0.0198, "step": 104470 }, { "grad_norm": 0.08523673564195633, "learning_rate": 7.758735011486653e-05, "loss": 0.0206, "step": 104480 }, { "grad_norm": 0.119829460978508, "learning_rate": 7.758275323758429e-05, "loss": 0.0208, "step": 104490 }, { "grad_norm": 0.11532720923423767, "learning_rate": 7.757815602514552e-05, "loss": 0.0192, "step": 104500 }, { "grad_norm": 0.09602422267198563, "learning_rate": 7.757355847760613e-05, "loss": 0.0138, "step": 104510 }, { "grad_norm": 0.08408825844526291, "learning_rate": 7.756896059502197e-05, "loss": 0.0127, "step": 104520 }, { "grad_norm": 0.123358815908432, "learning_rate": 7.756436237744891e-05, "loss": 0.014, "step": 104530 }, { "grad_norm": 0.08195691555738449, "learning_rate": 7.755976382494284e-05, "loss": 0.02, "step": 104540 }, { "grad_norm": 0.10744653642177582, "learning_rate": 7.75551649375596e-05, "loss": 0.0186, "step": 104550 }, { "grad_norm": 0.08779431134462357, "learning_rate": 7.755056571535508e-05, "loss": 0.0141, "step": 104560 }, { "grad_norm": 0.3258765935897827, "learning_rate": 7.754596615838518e-05, "loss": 0.0196, "step": 104570 }, { "grad_norm": 0.15410883724689484, "learning_rate": 7.754136626670579e-05, "loss": 0.0154, "step": 104580 }, { "grad_norm": 0.15720847249031067, "learning_rate": 7.753676604037282e-05, "loss": 0.0143, "step": 104590 }, { "grad_norm": 0.0850246250629425, "learning_rate": 7.753216547944211e-05, "loss": 0.0167, "step": 104600 }, { "grad_norm": 0.23256276547908783, "learning_rate": 7.75275645839696e-05, "loss": 0.0254, "step": 104610 }, { "grad_norm": 0.16222594678401947, "learning_rate": 7.75229633540112e-05, "loss": 0.0157, "step": 104620 }, { "grad_norm": 0.11890454590320587, "learning_rate": 7.75183617896228e-05, "loss": 0.0163, "step": 104630 }, { "grad_norm": 0.15668885409832, "learning_rate": 7.751375989086032e-05, "loss": 0.0158, "step": 104640 }, { "grad_norm": 0.20324064791202545, "learning_rate": 7.75091576577797e-05, "loss": 0.0166, "step": 104650 }, { "grad_norm": 0.12584681808948517, "learning_rate": 7.750455509043681e-05, "loss": 0.0147, "step": 104660 }, { "grad_norm": 0.2608245313167572, "learning_rate": 7.749995218888763e-05, "loss": 0.0193, "step": 104670 }, { "grad_norm": 0.3601495623588562, "learning_rate": 7.749534895318806e-05, "loss": 0.0238, "step": 104680 }, { "grad_norm": 0.11724632978439331, "learning_rate": 7.749074538339405e-05, "loss": 0.0271, "step": 104690 }, { "grad_norm": 0.21039889752864838, "learning_rate": 7.748614147956153e-05, "loss": 0.0199, "step": 104700 }, { "grad_norm": 0.5617552995681763, "learning_rate": 7.748153724174642e-05, "loss": 0.0228, "step": 104710 }, { "grad_norm": 0.22839581966400146, "learning_rate": 7.747693267000469e-05, "loss": 0.019, "step": 104720 }, { "grad_norm": 0.1353600025177002, "learning_rate": 7.74723277643923e-05, "loss": 0.0186, "step": 104730 }, { "grad_norm": 0.091698557138443, "learning_rate": 7.746772252496519e-05, "loss": 0.0142, "step": 104740 }, { "grad_norm": 0.19686104357242584, "learning_rate": 7.74631169517793e-05, "loss": 0.0188, "step": 104750 }, { "grad_norm": 0.10832618921995163, "learning_rate": 7.745851104489061e-05, "loss": 0.0171, "step": 104760 }, { "grad_norm": 0.10679630190134048, "learning_rate": 7.74539048043551e-05, "loss": 0.0185, "step": 104770 }, { "grad_norm": 0.23703131079673767, "learning_rate": 7.744929823022872e-05, "loss": 0.0196, "step": 104780 }, { "grad_norm": 0.22114719450473785, "learning_rate": 7.744469132256746e-05, "loss": 0.0145, "step": 104790 }, { "grad_norm": 0.23811091482639313, "learning_rate": 7.744008408142726e-05, "loss": 0.0186, "step": 104800 }, { "grad_norm": 0.32732394337654114, "learning_rate": 7.743547650686415e-05, "loss": 0.0242, "step": 104810 }, { "grad_norm": 0.1963001936674118, "learning_rate": 7.74308685989341e-05, "loss": 0.0168, "step": 104820 }, { "grad_norm": 0.1102307066321373, "learning_rate": 7.742626035769309e-05, "loss": 0.0266, "step": 104830 }, { "grad_norm": 0.21267887949943542, "learning_rate": 7.742165178319712e-05, "loss": 0.0188, "step": 104840 }, { "grad_norm": 0.10238654911518097, "learning_rate": 7.74170428755022e-05, "loss": 0.0158, "step": 104850 }, { "grad_norm": 0.278246134519577, "learning_rate": 7.741243363466433e-05, "loss": 0.0167, "step": 104860 }, { "grad_norm": 0.11496302485466003, "learning_rate": 7.74078240607395e-05, "loss": 0.0165, "step": 104870 }, { "grad_norm": 0.11745373159646988, "learning_rate": 7.740321415378373e-05, "loss": 0.0167, "step": 104880 }, { "grad_norm": 0.14282697439193726, "learning_rate": 7.739860391385303e-05, "loss": 0.014, "step": 104890 }, { "grad_norm": 0.11410645395517349, "learning_rate": 7.739399334100342e-05, "loss": 0.0151, "step": 104900 }, { "grad_norm": 0.18588672578334808, "learning_rate": 7.738938243529094e-05, "loss": 0.0228, "step": 104910 }, { "grad_norm": 0.10502102226018906, "learning_rate": 7.73847711967716e-05, "loss": 0.0152, "step": 104920 }, { "grad_norm": 0.0845157578587532, "learning_rate": 7.738015962550142e-05, "loss": 0.0153, "step": 104930 }, { "grad_norm": 0.11181261390447617, "learning_rate": 7.737554772153647e-05, "loss": 0.0117, "step": 104940 }, { "grad_norm": 0.2505328357219696, "learning_rate": 7.737093548493276e-05, "loss": 0.0236, "step": 104950 }, { "grad_norm": 0.0896485298871994, "learning_rate": 7.736632291574634e-05, "loss": 0.0146, "step": 104960 }, { "grad_norm": 0.09959739446640015, "learning_rate": 7.736171001403326e-05, "loss": 0.013, "step": 104970 }, { "grad_norm": 0.1343715339899063, "learning_rate": 7.735709677984957e-05, "loss": 0.021, "step": 104980 }, { "grad_norm": 0.22038909792900085, "learning_rate": 7.735248321325133e-05, "loss": 0.0194, "step": 104990 }, { "grad_norm": 0.10114336013793945, "learning_rate": 7.734786931429457e-05, "loss": 0.0186, "step": 105000 }, { "grad_norm": 0.08341526985168457, "learning_rate": 7.73432550830354e-05, "loss": 0.0175, "step": 105010 }, { "grad_norm": 0.1945524662733078, "learning_rate": 7.733864051952987e-05, "loss": 0.0185, "step": 105020 }, { "grad_norm": 0.162000373005867, "learning_rate": 7.733402562383403e-05, "loss": 0.0124, "step": 105030 }, { "grad_norm": 0.1992443948984146, "learning_rate": 7.732941039600398e-05, "loss": 0.0179, "step": 105040 }, { "grad_norm": 0.0812079980969429, "learning_rate": 7.732479483609578e-05, "loss": 0.0166, "step": 105050 }, { "grad_norm": 0.2695067524909973, "learning_rate": 7.732017894416554e-05, "loss": 0.0184, "step": 105060 }, { "grad_norm": 0.10660355538129807, "learning_rate": 7.731556272026932e-05, "loss": 0.0157, "step": 105070 }, { "grad_norm": 0.1562405675649643, "learning_rate": 7.731094616446323e-05, "loss": 0.0149, "step": 105080 }, { "grad_norm": 0.0980735793709755, "learning_rate": 7.730632927680334e-05, "loss": 0.0173, "step": 105090 }, { "grad_norm": 0.08703682571649551, "learning_rate": 7.730171205734578e-05, "loss": 0.0201, "step": 105100 }, { "grad_norm": 0.12459119409322739, "learning_rate": 7.729709450614666e-05, "loss": 0.0152, "step": 105110 }, { "grad_norm": 0.14089299738407135, "learning_rate": 7.729247662326205e-05, "loss": 0.0179, "step": 105120 }, { "grad_norm": 0.18195967376232147, "learning_rate": 7.728785840874808e-05, "loss": 0.023, "step": 105130 }, { "grad_norm": 0.20916911959648132, "learning_rate": 7.728323986266086e-05, "loss": 0.0246, "step": 105140 }, { "grad_norm": 0.09646084904670715, "learning_rate": 7.727862098505654e-05, "loss": 0.0175, "step": 105150 }, { "grad_norm": 0.47257372736930847, "learning_rate": 7.727400177599118e-05, "loss": 0.0195, "step": 105160 }, { "grad_norm": 0.08601514995098114, "learning_rate": 7.726938223552098e-05, "loss": 0.0173, "step": 105170 }, { "grad_norm": 0.1217152327299118, "learning_rate": 7.726476236370203e-05, "loss": 0.0165, "step": 105180 }, { "grad_norm": 0.1592848151922226, "learning_rate": 7.726014216059048e-05, "loss": 0.0132, "step": 105190 }, { "grad_norm": 0.11275668442249298, "learning_rate": 7.725552162624248e-05, "loss": 0.014, "step": 105200 }, { "grad_norm": 1.7806651592254639, "learning_rate": 7.725090076071413e-05, "loss": 0.0247, "step": 105210 }, { "grad_norm": 0.1599230319261551, "learning_rate": 7.724627956406162e-05, "loss": 0.0231, "step": 105220 }, { "grad_norm": 0.4038817584514618, "learning_rate": 7.724165803634108e-05, "loss": 0.0232, "step": 105230 }, { "grad_norm": 0.2958866357803345, "learning_rate": 7.723703617760869e-05, "loss": 0.0147, "step": 105240 }, { "grad_norm": 0.10469330102205276, "learning_rate": 7.723241398792057e-05, "loss": 0.0199, "step": 105250 }, { "grad_norm": 0.09891588240861893, "learning_rate": 7.722779146733293e-05, "loss": 0.0135, "step": 105260 }, { "grad_norm": 0.179185152053833, "learning_rate": 7.722316861590191e-05, "loss": 0.0129, "step": 105270 }, { "grad_norm": 0.11271728575229645, "learning_rate": 7.721854543368367e-05, "loss": 0.0177, "step": 105280 }, { "grad_norm": 0.13856734335422516, "learning_rate": 7.721392192073444e-05, "loss": 0.0144, "step": 105290 }, { "grad_norm": 0.15751641988754272, "learning_rate": 7.720929807711034e-05, "loss": 0.0173, "step": 105300 }, { "grad_norm": 0.28263169527053833, "learning_rate": 7.720467390286758e-05, "loss": 0.0178, "step": 105310 }, { "grad_norm": 0.12403440475463867, "learning_rate": 7.720004939806236e-05, "loss": 0.0174, "step": 105320 }, { "grad_norm": 0.13608664274215698, "learning_rate": 7.719542456275084e-05, "loss": 0.0165, "step": 105330 }, { "grad_norm": 0.21286484599113464, "learning_rate": 7.719079939698925e-05, "loss": 0.0224, "step": 105340 }, { "grad_norm": 0.10014811158180237, "learning_rate": 7.718617390083376e-05, "loss": 0.0147, "step": 105350 }, { "grad_norm": 0.11639697104692459, "learning_rate": 7.718154807434061e-05, "loss": 0.0183, "step": 105360 }, { "grad_norm": 0.11132273077964783, "learning_rate": 7.717692191756596e-05, "loss": 0.0163, "step": 105370 }, { "grad_norm": 0.13682599365711212, "learning_rate": 7.717229543056608e-05, "loss": 0.0225, "step": 105380 }, { "grad_norm": 0.18466752767562866, "learning_rate": 7.716766861339714e-05, "loss": 0.0222, "step": 105390 }, { "grad_norm": 0.09999493509531021, "learning_rate": 7.716304146611536e-05, "loss": 0.0164, "step": 105400 }, { "grad_norm": 0.1093432679772377, "learning_rate": 7.715841398877701e-05, "loss": 0.0182, "step": 105410 }, { "grad_norm": 0.12693443894386292, "learning_rate": 7.715378618143827e-05, "loss": 0.0244, "step": 105420 }, { "grad_norm": 0.14700625836849213, "learning_rate": 7.714915804415542e-05, "loss": 0.021, "step": 105430 }, { "grad_norm": 0.12990839779376984, "learning_rate": 7.714452957698462e-05, "loss": 0.0194, "step": 105440 }, { "grad_norm": 0.24790407717227936, "learning_rate": 7.713990077998219e-05, "loss": 0.0186, "step": 105450 }, { "grad_norm": 0.19432558119297028, "learning_rate": 7.713527165320432e-05, "loss": 0.0212, "step": 105460 }, { "grad_norm": 0.15705189108848572, "learning_rate": 7.713064219670729e-05, "loss": 0.017, "step": 105470 }, { "grad_norm": 0.2711915373802185, "learning_rate": 7.712601241054734e-05, "loss": 0.0119, "step": 105480 }, { "grad_norm": 0.1577998399734497, "learning_rate": 7.712138229478075e-05, "loss": 0.025, "step": 105490 }, { "grad_norm": 0.14023415744304657, "learning_rate": 7.711675184946373e-05, "loss": 0.0207, "step": 105500 }, { "grad_norm": 0.10530346632003784, "learning_rate": 7.711212107465257e-05, "loss": 0.0217, "step": 105510 }, { "grad_norm": 0.1267337054014206, "learning_rate": 7.710748997040356e-05, "loss": 0.0164, "step": 105520 }, { "grad_norm": 0.09461536258459091, "learning_rate": 7.710285853677295e-05, "loss": 0.0193, "step": 105530 }, { "grad_norm": 0.0681988000869751, "learning_rate": 7.709822677381702e-05, "loss": 0.0159, "step": 105540 }, { "grad_norm": 0.11408218741416931, "learning_rate": 7.709359468159203e-05, "loss": 0.0123, "step": 105550 }, { "grad_norm": 0.11781199276447296, "learning_rate": 7.70889622601543e-05, "loss": 0.0244, "step": 105560 }, { "grad_norm": 0.0940670594573021, "learning_rate": 7.708432950956011e-05, "loss": 0.0145, "step": 105570 }, { "grad_norm": 0.14641045033931732, "learning_rate": 7.707969642986572e-05, "loss": 0.0172, "step": 105580 }, { "grad_norm": 0.12166480720043182, "learning_rate": 7.707506302112746e-05, "loss": 0.0136, "step": 105590 }, { "grad_norm": 0.14123763144016266, "learning_rate": 7.707042928340164e-05, "loss": 0.0214, "step": 105600 }, { "grad_norm": 0.13272783160209656, "learning_rate": 7.706579521674453e-05, "loss": 0.0146, "step": 105610 }, { "grad_norm": 0.28022560477256775, "learning_rate": 7.706116082121245e-05, "loss": 0.0168, "step": 105620 }, { "grad_norm": 0.133108451962471, "learning_rate": 7.705652609686173e-05, "loss": 0.0243, "step": 105630 }, { "grad_norm": 0.23729307949543, "learning_rate": 7.705189104374866e-05, "loss": 0.0179, "step": 105640 }, { "grad_norm": 0.1314147561788559, "learning_rate": 7.704725566192958e-05, "loss": 0.0151, "step": 105650 }, { "grad_norm": 0.1176702231168747, "learning_rate": 7.70426199514608e-05, "loss": 0.0143, "step": 105660 }, { "grad_norm": 0.20700426399707794, "learning_rate": 7.703798391239866e-05, "loss": 0.0168, "step": 105670 }, { "grad_norm": 0.18372295796871185, "learning_rate": 7.703334754479949e-05, "loss": 0.0154, "step": 105680 }, { "grad_norm": 0.08364325761795044, "learning_rate": 7.70287108487196e-05, "loss": 0.02, "step": 105690 }, { "grad_norm": 0.2323814332485199, "learning_rate": 7.702407382421539e-05, "loss": 0.0243, "step": 105700 }, { "grad_norm": 0.1254739761352539, "learning_rate": 7.701943647134315e-05, "loss": 0.015, "step": 105710 }, { "grad_norm": 0.08369084447622299, "learning_rate": 7.701479879015925e-05, "loss": 0.0158, "step": 105720 }, { "grad_norm": 0.10495400428771973, "learning_rate": 7.701016078072005e-05, "loss": 0.0172, "step": 105730 }, { "grad_norm": 0.1141093522310257, "learning_rate": 7.700552244308188e-05, "loss": 0.0171, "step": 105740 }, { "grad_norm": 0.1380188912153244, "learning_rate": 7.700088377730111e-05, "loss": 0.0176, "step": 105750 }, { "grad_norm": 0.19172349572181702, "learning_rate": 7.699624478343413e-05, "loss": 0.0127, "step": 105760 }, { "grad_norm": 0.13064908981323242, "learning_rate": 7.699160546153727e-05, "loss": 0.0208, "step": 105770 }, { "grad_norm": 0.1219458132982254, "learning_rate": 7.698696581166691e-05, "loss": 0.0163, "step": 105780 }, { "grad_norm": 0.23296011984348297, "learning_rate": 7.698232583387945e-05, "loss": 0.0156, "step": 105790 }, { "grad_norm": 0.23176319897174835, "learning_rate": 7.697768552823126e-05, "loss": 0.0171, "step": 105800 }, { "grad_norm": 0.13941878080368042, "learning_rate": 7.69730448947787e-05, "loss": 0.0168, "step": 105810 }, { "grad_norm": 0.6339561939239502, "learning_rate": 7.69684039335782e-05, "loss": 0.0156, "step": 105820 }, { "grad_norm": 0.13709427416324615, "learning_rate": 7.696376264468612e-05, "loss": 0.0166, "step": 105830 }, { "grad_norm": 0.16681058704853058, "learning_rate": 7.695912102815886e-05, "loss": 0.0198, "step": 105840 }, { "grad_norm": 0.10074210911989212, "learning_rate": 7.695447908405283e-05, "loss": 0.0148, "step": 105850 }, { "grad_norm": 0.1844741404056549, "learning_rate": 7.694983681242442e-05, "loss": 0.0158, "step": 105860 }, { "grad_norm": 0.09526991844177246, "learning_rate": 7.694519421333007e-05, "loss": 0.0138, "step": 105870 }, { "grad_norm": 0.11275079101324081, "learning_rate": 7.694055128682616e-05, "loss": 0.0125, "step": 105880 }, { "grad_norm": 0.09171044081449509, "learning_rate": 7.693590803296912e-05, "loss": 0.015, "step": 105890 }, { "grad_norm": 0.17169731855392456, "learning_rate": 7.693126445181536e-05, "loss": 0.0123, "step": 105900 }, { "grad_norm": 0.10047747939825058, "learning_rate": 7.692662054342131e-05, "loss": 0.0182, "step": 105910 }, { "grad_norm": 0.1386168748140335, "learning_rate": 7.692197630784339e-05, "loss": 0.0183, "step": 105920 }, { "grad_norm": 0.09051622450351715, "learning_rate": 7.691733174513806e-05, "loss": 0.0136, "step": 105930 }, { "grad_norm": 0.23016105592250824, "learning_rate": 7.691268685536172e-05, "loss": 0.016, "step": 105940 }, { "grad_norm": 0.14379453659057617, "learning_rate": 7.690804163857082e-05, "loss": 0.0139, "step": 105950 }, { "grad_norm": 0.12283437699079514, "learning_rate": 7.690339609482182e-05, "loss": 0.0125, "step": 105960 }, { "grad_norm": 0.32695430517196655, "learning_rate": 7.689875022417115e-05, "loss": 0.0195, "step": 105970 }, { "grad_norm": 0.14377185702323914, "learning_rate": 7.689410402667527e-05, "loss": 0.0206, "step": 105980 }, { "grad_norm": 0.20369835197925568, "learning_rate": 7.688945750239064e-05, "loss": 0.0147, "step": 105990 }, { "grad_norm": 0.18304334580898285, "learning_rate": 7.688481065137371e-05, "loss": 0.0149, "step": 106000 }, { "grad_norm": 0.17872999608516693, "learning_rate": 7.688016347368094e-05, "loss": 0.0151, "step": 106010 }, { "grad_norm": 0.1984097957611084, "learning_rate": 7.687551596936882e-05, "loss": 0.0232, "step": 106020 }, { "grad_norm": 0.2472076117992401, "learning_rate": 7.687086813849378e-05, "loss": 0.0179, "step": 106030 }, { "grad_norm": 0.21092073619365692, "learning_rate": 7.686621998111234e-05, "loss": 0.02, "step": 106040 }, { "grad_norm": 0.11435452103614807, "learning_rate": 7.686157149728096e-05, "loss": 0.0253, "step": 106050 }, { "grad_norm": 0.1854221075773239, "learning_rate": 7.685692268705614e-05, "loss": 0.0232, "step": 106060 }, { "grad_norm": 0.12412990629673004, "learning_rate": 7.685227355049434e-05, "loss": 0.0169, "step": 106070 }, { "grad_norm": 0.21674948930740356, "learning_rate": 7.684762408765205e-05, "loss": 0.0186, "step": 106080 }, { "grad_norm": 0.14855429530143738, "learning_rate": 7.684297429858579e-05, "loss": 0.0232, "step": 106090 }, { "grad_norm": 0.1442665308713913, "learning_rate": 7.683832418335204e-05, "loss": 0.0175, "step": 106100 }, { "grad_norm": 0.2521336078643799, "learning_rate": 7.683367374200732e-05, "loss": 0.0197, "step": 106110 }, { "grad_norm": 0.11161479353904724, "learning_rate": 7.682902297460812e-05, "loss": 0.0146, "step": 106120 }, { "grad_norm": 0.2606000602245331, "learning_rate": 7.682437188121097e-05, "loss": 0.0194, "step": 106130 }, { "grad_norm": 0.15075939893722534, "learning_rate": 7.681972046187238e-05, "loss": 0.0137, "step": 106140 }, { "grad_norm": 0.19602800905704498, "learning_rate": 7.681506871664886e-05, "loss": 0.0145, "step": 106150 }, { "grad_norm": 0.10040726512670517, "learning_rate": 7.681041664559693e-05, "loss": 0.0196, "step": 106160 }, { "grad_norm": 0.25766521692276, "learning_rate": 7.680576424877312e-05, "loss": 0.0149, "step": 106170 }, { "grad_norm": 0.08395516127347946, "learning_rate": 7.680111152623399e-05, "loss": 0.0115, "step": 106180 }, { "grad_norm": 0.20060645043849945, "learning_rate": 7.679645847803603e-05, "loss": 0.0151, "step": 106190 }, { "grad_norm": 0.20430538058280945, "learning_rate": 7.67918051042358e-05, "loss": 0.0204, "step": 106200 }, { "grad_norm": 0.08602791279554367, "learning_rate": 7.678715140488985e-05, "loss": 0.0133, "step": 106210 }, { "grad_norm": 0.45408007502555847, "learning_rate": 7.678249738005472e-05, "loss": 0.0173, "step": 106220 }, { "grad_norm": 0.20254668593406677, "learning_rate": 7.677784302978697e-05, "loss": 0.0171, "step": 106230 }, { "grad_norm": 0.29227977991104126, "learning_rate": 7.67731883541431e-05, "loss": 0.0176, "step": 106240 }, { "grad_norm": 0.7101343870162964, "learning_rate": 7.676853335317977e-05, "loss": 0.0274, "step": 106250 }, { "grad_norm": 0.11912170052528381, "learning_rate": 7.676387802695345e-05, "loss": 0.0191, "step": 106260 }, { "grad_norm": 0.129439577460289, "learning_rate": 7.675922237552075e-05, "loss": 0.0194, "step": 106270 }, { "grad_norm": 0.17951981723308563, "learning_rate": 7.675456639893823e-05, "loss": 0.0146, "step": 106280 }, { "grad_norm": 0.23564930260181427, "learning_rate": 7.674991009726249e-05, "loss": 0.0167, "step": 106290 }, { "grad_norm": 0.1685769259929657, "learning_rate": 7.674525347055006e-05, "loss": 0.0209, "step": 106300 }, { "grad_norm": 0.0997268334031105, "learning_rate": 7.674059651885757e-05, "loss": 0.0171, "step": 106310 }, { "grad_norm": 0.09044469147920609, "learning_rate": 7.673593924224158e-05, "loss": 0.0144, "step": 106320 }, { "grad_norm": 0.11702409386634827, "learning_rate": 7.673128164075867e-05, "loss": 0.0188, "step": 106330 }, { "grad_norm": 0.42378631234169006, "learning_rate": 7.672662371446546e-05, "loss": 0.0184, "step": 106340 }, { "grad_norm": 0.1840747892856598, "learning_rate": 7.672196546341853e-05, "loss": 0.0204, "step": 106350 }, { "grad_norm": 0.20691294968128204, "learning_rate": 7.671730688767451e-05, "loss": 0.0258, "step": 106360 }, { "grad_norm": 0.5172538757324219, "learning_rate": 7.671264798728997e-05, "loss": 0.0199, "step": 106370 }, { "grad_norm": 0.09401796013116837, "learning_rate": 7.670798876232154e-05, "loss": 0.0172, "step": 106380 }, { "grad_norm": 0.1550881564617157, "learning_rate": 7.670332921282582e-05, "loss": 0.0143, "step": 106390 }, { "grad_norm": 0.18487513065338135, "learning_rate": 7.669866933885945e-05, "loss": 0.0229, "step": 106400 }, { "grad_norm": 0.09249845892190933, "learning_rate": 7.669400914047904e-05, "loss": 0.0231, "step": 106410 }, { "grad_norm": 0.10437507927417755, "learning_rate": 7.668934861774121e-05, "loss": 0.0168, "step": 106420 }, { "grad_norm": 0.16862371563911438, "learning_rate": 7.668468777070259e-05, "loss": 0.0173, "step": 106430 }, { "grad_norm": 0.2565438747406006, "learning_rate": 7.668002659941983e-05, "loss": 0.0173, "step": 106440 }, { "grad_norm": 0.14437897503376007, "learning_rate": 7.667536510394956e-05, "loss": 0.015, "step": 106450 }, { "grad_norm": 0.15046226978302002, "learning_rate": 7.66707032843484e-05, "loss": 0.0135, "step": 106460 }, { "grad_norm": 0.2790289521217346, "learning_rate": 7.666604114067304e-05, "loss": 0.021, "step": 106470 }, { "grad_norm": 0.2577068507671356, "learning_rate": 7.666137867298009e-05, "loss": 0.0187, "step": 106480 }, { "grad_norm": 0.06392280012369156, "learning_rate": 7.66567158813262e-05, "loss": 0.0096, "step": 106490 }, { "grad_norm": 0.19278377294540405, "learning_rate": 7.665205276576807e-05, "loss": 0.0122, "step": 106500 }, { "grad_norm": 0.1402694135904312, "learning_rate": 7.664738932636231e-05, "loss": 0.019, "step": 106510 }, { "grad_norm": 0.17314288020133972, "learning_rate": 7.664272556316564e-05, "loss": 0.0286, "step": 106520 }, { "grad_norm": 0.12517036497592926, "learning_rate": 7.663806147623467e-05, "loss": 0.0166, "step": 106530 }, { "grad_norm": 0.15337103605270386, "learning_rate": 7.663339706562612e-05, "loss": 0.0171, "step": 106540 }, { "grad_norm": 0.2851823568344116, "learning_rate": 7.662873233139664e-05, "loss": 0.0188, "step": 106550 }, { "grad_norm": 0.20968908071517944, "learning_rate": 7.662406727360294e-05, "loss": 0.03, "step": 106560 }, { "grad_norm": 0.5388702154159546, "learning_rate": 7.661940189230166e-05, "loss": 0.0191, "step": 106570 }, { "grad_norm": 0.20228223502635956, "learning_rate": 7.661473618754953e-05, "loss": 0.024, "step": 106580 }, { "grad_norm": 0.198480486869812, "learning_rate": 7.661007015940323e-05, "loss": 0.0187, "step": 106590 }, { "grad_norm": 0.28049877285957336, "learning_rate": 7.660540380791942e-05, "loss": 0.0222, "step": 106600 }, { "grad_norm": 0.4505532681941986, "learning_rate": 7.660073713315488e-05, "loss": 0.0217, "step": 106610 }, { "grad_norm": 0.12800081074237823, "learning_rate": 7.659607013516624e-05, "loss": 0.0183, "step": 106620 }, { "grad_norm": 0.17044292390346527, "learning_rate": 7.659140281401026e-05, "loss": 0.0243, "step": 106630 }, { "grad_norm": 0.17180365324020386, "learning_rate": 7.658673516974362e-05, "loss": 0.0172, "step": 106640 }, { "grad_norm": 0.2172529697418213, "learning_rate": 7.658206720242302e-05, "loss": 0.0123, "step": 106650 }, { "grad_norm": 0.16045354306697845, "learning_rate": 7.657739891210523e-05, "loss": 0.0176, "step": 106660 }, { "grad_norm": 0.10304742306470871, "learning_rate": 7.657273029884695e-05, "loss": 0.0149, "step": 106670 }, { "grad_norm": 0.25045451521873474, "learning_rate": 7.656806136270492e-05, "loss": 0.0238, "step": 106680 }, { "grad_norm": 0.3491806089878082, "learning_rate": 7.656339210373584e-05, "loss": 0.0216, "step": 106690 }, { "grad_norm": 0.17720940709114075, "learning_rate": 7.655872252199646e-05, "loss": 0.0135, "step": 106700 }, { "grad_norm": 0.15258486568927765, "learning_rate": 7.655405261754354e-05, "loss": 0.017, "step": 106710 }, { "grad_norm": 0.38765600323677063, "learning_rate": 7.654938239043381e-05, "loss": 0.0254, "step": 106720 }, { "grad_norm": 0.33588337898254395, "learning_rate": 7.654471184072401e-05, "loss": 0.0285, "step": 106730 }, { "grad_norm": 0.13208669424057007, "learning_rate": 7.65400409684709e-05, "loss": 0.0118, "step": 106740 }, { "grad_norm": 0.1936904639005661, "learning_rate": 7.653536977373123e-05, "loss": 0.017, "step": 106750 }, { "grad_norm": 0.28689971566200256, "learning_rate": 7.653069825656178e-05, "loss": 0.0194, "step": 106760 }, { "grad_norm": 0.16024444997310638, "learning_rate": 7.65260264170193e-05, "loss": 0.0163, "step": 106770 }, { "grad_norm": 0.1292455941438675, "learning_rate": 7.652135425516053e-05, "loss": 0.0229, "step": 106780 }, { "grad_norm": 0.12681065499782562, "learning_rate": 7.651668177104227e-05, "loss": 0.0197, "step": 106790 }, { "grad_norm": 0.1645423024892807, "learning_rate": 7.65120089647213e-05, "loss": 0.0231, "step": 106800 }, { "grad_norm": 0.13385340571403503, "learning_rate": 7.650733583625439e-05, "loss": 0.0171, "step": 106810 }, { "grad_norm": 0.14354665577411652, "learning_rate": 7.650266238569831e-05, "loss": 0.0212, "step": 106820 }, { "grad_norm": 0.17437416315078735, "learning_rate": 7.649798861310988e-05, "loss": 0.0209, "step": 106830 }, { "grad_norm": 0.16197744011878967, "learning_rate": 7.649331451854587e-05, "loss": 0.0258, "step": 106840 }, { "grad_norm": 0.11902859807014465, "learning_rate": 7.648864010206305e-05, "loss": 0.0174, "step": 106850 }, { "grad_norm": 0.12752902507781982, "learning_rate": 7.648396536371826e-05, "loss": 0.0158, "step": 106860 }, { "grad_norm": 0.14099375903606415, "learning_rate": 7.647929030356829e-05, "loss": 0.0165, "step": 106870 }, { "grad_norm": 0.11737842857837677, "learning_rate": 7.647461492166993e-05, "loss": 0.0209, "step": 106880 }, { "grad_norm": 0.12330179661512375, "learning_rate": 7.646993921808001e-05, "loss": 0.0219, "step": 106890 }, { "grad_norm": 0.09742366522550583, "learning_rate": 7.646526319285533e-05, "loss": 0.0229, "step": 106900 }, { "grad_norm": 0.12550325691699982, "learning_rate": 7.646058684605273e-05, "loss": 0.0154, "step": 106910 }, { "grad_norm": 0.15193220973014832, "learning_rate": 7.6455910177729e-05, "loss": 0.0212, "step": 106920 }, { "grad_norm": 0.09434651583433151, "learning_rate": 7.645123318794098e-05, "loss": 0.0187, "step": 106930 }, { "grad_norm": 0.10209333151578903, "learning_rate": 7.644655587674552e-05, "loss": 0.019, "step": 106940 }, { "grad_norm": 0.1440494954586029, "learning_rate": 7.644187824419943e-05, "loss": 0.0193, "step": 106950 }, { "grad_norm": 0.11347822844982147, "learning_rate": 7.643720029035956e-05, "loss": 0.0177, "step": 106960 }, { "grad_norm": 0.11514081060886383, "learning_rate": 7.643252201528273e-05, "loss": 0.0202, "step": 106970 }, { "grad_norm": 0.09792875498533249, "learning_rate": 7.64278434190258e-05, "loss": 0.0196, "step": 106980 }, { "grad_norm": 0.2650446593761444, "learning_rate": 7.642316450164563e-05, "loss": 0.0142, "step": 106990 }, { "grad_norm": 0.11590029299259186, "learning_rate": 7.641848526319908e-05, "loss": 0.0153, "step": 107000 }, { "grad_norm": 0.13953538239002228, "learning_rate": 7.641380570374296e-05, "loss": 0.0218, "step": 107010 }, { "grad_norm": 0.20700229704380035, "learning_rate": 7.640912582333418e-05, "loss": 0.0149, "step": 107020 }, { "grad_norm": 0.12222190201282501, "learning_rate": 7.640444562202959e-05, "loss": 0.0117, "step": 107030 }, { "grad_norm": 0.11210251599550247, "learning_rate": 7.639976509988605e-05, "loss": 0.0194, "step": 107040 }, { "grad_norm": 0.09612302482128143, "learning_rate": 7.639508425696042e-05, "loss": 0.0181, "step": 107050 }, { "grad_norm": 0.1535581350326538, "learning_rate": 7.639040309330963e-05, "loss": 0.0168, "step": 107060 }, { "grad_norm": 0.1948564052581787, "learning_rate": 7.63857216089905e-05, "loss": 0.0169, "step": 107070 }, { "grad_norm": 0.11534301936626434, "learning_rate": 7.638103980405995e-05, "loss": 0.0172, "step": 107080 }, { "grad_norm": 0.30978620052337646, "learning_rate": 7.637635767857486e-05, "loss": 0.0126, "step": 107090 }, { "grad_norm": 0.16589519381523132, "learning_rate": 7.637167523259211e-05, "loss": 0.0142, "step": 107100 }, { "grad_norm": 0.1905984878540039, "learning_rate": 7.636699246616862e-05, "loss": 0.0198, "step": 107110 }, { "grad_norm": 0.15148602426052094, "learning_rate": 7.636230937936128e-05, "loss": 0.016, "step": 107120 }, { "grad_norm": 0.20681369304656982, "learning_rate": 7.635762597222698e-05, "loss": 0.0144, "step": 107130 }, { "grad_norm": 0.21090075373649597, "learning_rate": 7.635294224482263e-05, "loss": 0.0185, "step": 107140 }, { "grad_norm": 0.11411796510219574, "learning_rate": 7.634825819720517e-05, "loss": 0.0179, "step": 107150 }, { "grad_norm": 0.05407348647713661, "learning_rate": 7.634357382943149e-05, "loss": 0.0134, "step": 107160 }, { "grad_norm": 0.135623037815094, "learning_rate": 7.63388891415585e-05, "loss": 0.0124, "step": 107170 }, { "grad_norm": 0.2554948627948761, "learning_rate": 7.633420413364317e-05, "loss": 0.0186, "step": 107180 }, { "grad_norm": 0.28016623854637146, "learning_rate": 7.632951880574237e-05, "loss": 0.0187, "step": 107190 }, { "grad_norm": 0.08972613513469696, "learning_rate": 7.632483315791306e-05, "loss": 0.0158, "step": 107200 }, { "grad_norm": 0.11560898274183273, "learning_rate": 7.632014719021217e-05, "loss": 0.0265, "step": 107210 }, { "grad_norm": 0.12627634406089783, "learning_rate": 7.631546090269665e-05, "loss": 0.0169, "step": 107220 }, { "grad_norm": 0.20697487890720367, "learning_rate": 7.631077429542344e-05, "loss": 0.0159, "step": 107230 }, { "grad_norm": 0.9649046659469604, "learning_rate": 7.630608736844947e-05, "loss": 0.021, "step": 107240 }, { "grad_norm": 0.20993688702583313, "learning_rate": 7.63014001218317e-05, "loss": 0.0147, "step": 107250 }, { "grad_norm": 0.13451017439365387, "learning_rate": 7.62967125556271e-05, "loss": 0.0147, "step": 107260 }, { "grad_norm": 0.17963099479675293, "learning_rate": 7.62920246698926e-05, "loss": 0.0189, "step": 107270 }, { "grad_norm": 0.1394205540418625, "learning_rate": 7.628733646468517e-05, "loss": 0.0256, "step": 107280 }, { "grad_norm": 0.1564304083585739, "learning_rate": 7.628264794006182e-05, "loss": 0.0212, "step": 107290 }, { "grad_norm": 0.19642338156700134, "learning_rate": 7.627795909607944e-05, "loss": 0.0141, "step": 107300 }, { "grad_norm": 0.2774362862110138, "learning_rate": 7.627326993279506e-05, "loss": 0.0226, "step": 107310 }, { "grad_norm": 0.1722831279039383, "learning_rate": 7.626858045026566e-05, "loss": 0.0175, "step": 107320 }, { "grad_norm": 0.1592138707637787, "learning_rate": 7.626389064854818e-05, "loss": 0.0167, "step": 107330 }, { "grad_norm": 0.13853751122951508, "learning_rate": 7.625920052769966e-05, "loss": 0.0198, "step": 107340 }, { "grad_norm": 0.11415641754865646, "learning_rate": 7.625451008777705e-05, "loss": 0.0155, "step": 107350 }, { "grad_norm": 0.10772140324115753, "learning_rate": 7.624981932883734e-05, "loss": 0.0121, "step": 107360 }, { "grad_norm": 0.2826029658317566, "learning_rate": 7.624512825093755e-05, "loss": 0.0142, "step": 107370 }, { "grad_norm": 0.07432051748037338, "learning_rate": 7.624043685413467e-05, "loss": 0.014, "step": 107380 }, { "grad_norm": 0.12954898178577423, "learning_rate": 7.623574513848572e-05, "loss": 0.0217, "step": 107390 }, { "grad_norm": 0.15303561091423035, "learning_rate": 7.623105310404768e-05, "loss": 0.0156, "step": 107400 }, { "grad_norm": 0.16093067824840546, "learning_rate": 7.622636075087759e-05, "loss": 0.018, "step": 107410 }, { "grad_norm": 0.16060350835323334, "learning_rate": 7.622166807903243e-05, "loss": 0.0108, "step": 107420 }, { "grad_norm": 0.10749603062868118, "learning_rate": 7.621697508856929e-05, "loss": 0.0136, "step": 107430 }, { "grad_norm": 0.18939842283725739, "learning_rate": 7.621228177954513e-05, "loss": 0.0179, "step": 107440 }, { "grad_norm": 0.2490067332983017, "learning_rate": 7.620758815201698e-05, "loss": 0.0168, "step": 107450 }, { "grad_norm": 0.24725638329982758, "learning_rate": 7.620289420604192e-05, "loss": 0.0181, "step": 107460 }, { "grad_norm": 0.1313173472881317, "learning_rate": 7.619819994167692e-05, "loss": 0.023, "step": 107470 }, { "grad_norm": 0.08952437341213226, "learning_rate": 7.619350535897909e-05, "loss": 0.0157, "step": 107480 }, { "grad_norm": 0.11493358761072159, "learning_rate": 7.61888104580054e-05, "loss": 0.0133, "step": 107490 }, { "grad_norm": 0.2589699327945709, "learning_rate": 7.618411523881297e-05, "loss": 0.0183, "step": 107500 }, { "grad_norm": 0.0875622034072876, "learning_rate": 7.617941970145881e-05, "loss": 0.0175, "step": 107510 }, { "grad_norm": 0.26788264513015747, "learning_rate": 7.617472384599999e-05, "loss": 0.0197, "step": 107520 }, { "grad_norm": 0.16343183815479279, "learning_rate": 7.617002767249355e-05, "loss": 0.0213, "step": 107530 }, { "grad_norm": 0.22929129004478455, "learning_rate": 7.616533118099657e-05, "loss": 0.0193, "step": 107540 }, { "grad_norm": 0.1349170058965683, "learning_rate": 7.61606343715661e-05, "loss": 0.0141, "step": 107550 }, { "grad_norm": 0.10940438508987427, "learning_rate": 7.615593724425925e-05, "loss": 0.015, "step": 107560 }, { "grad_norm": 0.09842505306005478, "learning_rate": 7.615123979913305e-05, "loss": 0.0124, "step": 107570 }, { "grad_norm": 0.1636841893196106, "learning_rate": 7.614654203624458e-05, "loss": 0.0164, "step": 107580 }, { "grad_norm": 0.1261211335659027, "learning_rate": 7.614184395565096e-05, "loss": 0.0171, "step": 107590 }, { "grad_norm": 0.1541985422372818, "learning_rate": 7.613714555740924e-05, "loss": 0.0186, "step": 107600 }, { "grad_norm": 0.1566108763217926, "learning_rate": 7.613244684157654e-05, "loss": 0.0157, "step": 107610 }, { "grad_norm": 0.1350100189447403, "learning_rate": 7.612774780820994e-05, "loss": 0.0151, "step": 107620 }, { "grad_norm": 0.23344025015830994, "learning_rate": 7.612304845736651e-05, "loss": 0.0211, "step": 107630 }, { "grad_norm": 0.22586070001125336, "learning_rate": 7.61183487891034e-05, "loss": 0.0172, "step": 107640 }, { "grad_norm": 0.10768099129199982, "learning_rate": 7.61136488034777e-05, "loss": 0.0113, "step": 107650 }, { "grad_norm": 0.15952295064926147, "learning_rate": 7.610894850054648e-05, "loss": 0.0122, "step": 107660 }, { "grad_norm": 0.08543584495782852, "learning_rate": 7.61042478803669e-05, "loss": 0.0182, "step": 107670 }, { "grad_norm": 0.08342879265546799, "learning_rate": 7.609954694299607e-05, "loss": 0.0118, "step": 107680 }, { "grad_norm": 0.13240313529968262, "learning_rate": 7.60948456884911e-05, "loss": 0.0156, "step": 107690 }, { "grad_norm": 0.22897347807884216, "learning_rate": 7.60901441169091e-05, "loss": 0.0191, "step": 107700 }, { "grad_norm": 0.17539383471012115, "learning_rate": 7.608544222830724e-05, "loss": 0.0223, "step": 107710 }, { "grad_norm": 0.24567340314388275, "learning_rate": 7.608074002274262e-05, "loss": 0.0168, "step": 107720 }, { "grad_norm": 0.19102545082569122, "learning_rate": 7.607603750027237e-05, "loss": 0.0189, "step": 107730 }, { "grad_norm": 0.26396673917770386, "learning_rate": 7.607133466095365e-05, "loss": 0.0183, "step": 107740 }, { "grad_norm": 0.32809579372406006, "learning_rate": 7.606663150484362e-05, "loss": 0.0211, "step": 107750 }, { "grad_norm": 0.18254995346069336, "learning_rate": 7.606192803199938e-05, "loss": 0.0236, "step": 107760 }, { "grad_norm": 0.29563209414482117, "learning_rate": 7.605722424247812e-05, "loss": 0.0219, "step": 107770 }, { "grad_norm": 0.25814756751060486, "learning_rate": 7.605252013633699e-05, "loss": 0.0157, "step": 107780 }, { "grad_norm": 0.3213356137275696, "learning_rate": 7.604781571363313e-05, "loss": 0.0187, "step": 107790 }, { "grad_norm": 0.1262350231409073, "learning_rate": 7.604311097442372e-05, "loss": 0.0161, "step": 107800 }, { "grad_norm": 0.1612406224012375, "learning_rate": 7.603840591876592e-05, "loss": 0.0175, "step": 107810 }, { "grad_norm": 0.13378268480300903, "learning_rate": 7.60337005467169e-05, "loss": 0.0177, "step": 107820 }, { "grad_norm": 0.1750924438238144, "learning_rate": 7.602899485833384e-05, "loss": 0.0159, "step": 107830 }, { "grad_norm": 0.12822777032852173, "learning_rate": 7.602428885367395e-05, "loss": 0.0185, "step": 107840 }, { "grad_norm": 0.17263633012771606, "learning_rate": 7.601958253279435e-05, "loss": 0.0148, "step": 107850 }, { "grad_norm": 0.1438101828098297, "learning_rate": 7.601487589575224e-05, "loss": 0.0171, "step": 107860 }, { "grad_norm": 0.34251588582992554, "learning_rate": 7.601016894260485e-05, "loss": 0.0198, "step": 107870 }, { "grad_norm": 0.1343938112258911, "learning_rate": 7.600546167340934e-05, "loss": 0.0173, "step": 107880 }, { "grad_norm": 0.24856464564800262, "learning_rate": 7.600075408822291e-05, "loss": 0.0216, "step": 107890 }, { "grad_norm": 0.10448692739009857, "learning_rate": 7.599604618710279e-05, "loss": 0.013, "step": 107900 }, { "grad_norm": 0.15134842693805695, "learning_rate": 7.599133797010616e-05, "loss": 0.015, "step": 107910 }, { "grad_norm": 0.36833348870277405, "learning_rate": 7.598662943729021e-05, "loss": 0.0237, "step": 107920 }, { "grad_norm": 0.1293671578168869, "learning_rate": 7.598192058871221e-05, "loss": 0.0177, "step": 107930 }, { "grad_norm": 0.14948372542858124, "learning_rate": 7.597721142442932e-05, "loss": 0.0328, "step": 107940 }, { "grad_norm": 0.08499033004045486, "learning_rate": 7.597250194449878e-05, "loss": 0.0139, "step": 107950 }, { "grad_norm": 0.08611196279525757, "learning_rate": 7.596779214897784e-05, "loss": 0.015, "step": 107960 }, { "grad_norm": 0.11005068570375443, "learning_rate": 7.59630820379237e-05, "loss": 0.0156, "step": 107970 }, { "grad_norm": 0.2540181875228882, "learning_rate": 7.59583716113936e-05, "loss": 0.0178, "step": 107980 }, { "grad_norm": 0.22638510167598724, "learning_rate": 7.595366086944478e-05, "loss": 0.0141, "step": 107990 }, { "grad_norm": 0.1622212827205658, "learning_rate": 7.594894981213447e-05, "loss": 0.0134, "step": 108000 }, { "grad_norm": 0.17045468091964722, "learning_rate": 7.594423843951992e-05, "loss": 0.0177, "step": 108010 }, { "grad_norm": 0.17743298411369324, "learning_rate": 7.593952675165837e-05, "loss": 0.0152, "step": 108020 }, { "grad_norm": 0.212377667427063, "learning_rate": 7.593481474860709e-05, "loss": 0.0181, "step": 108030 }, { "grad_norm": 0.15954181551933289, "learning_rate": 7.593010243042333e-05, "loss": 0.0204, "step": 108040 }, { "grad_norm": 0.20381413400173187, "learning_rate": 7.592538979716434e-05, "loss": 0.0131, "step": 108050 }, { "grad_norm": 0.1452367603778839, "learning_rate": 7.592067684888737e-05, "loss": 0.0147, "step": 108060 }, { "grad_norm": 0.41490650177001953, "learning_rate": 7.591596358564971e-05, "loss": 0.0187, "step": 108070 }, { "grad_norm": 0.10725663602352142, "learning_rate": 7.591125000750864e-05, "loss": 0.0166, "step": 108080 }, { "grad_norm": 0.16287347674369812, "learning_rate": 7.59065361145214e-05, "loss": 0.0125, "step": 108090 }, { "grad_norm": 0.11254508048295975, "learning_rate": 7.59018219067453e-05, "loss": 0.0165, "step": 108100 }, { "grad_norm": 0.5250164866447449, "learning_rate": 7.58971073842376e-05, "loss": 0.0203, "step": 108110 }, { "grad_norm": 0.14959372580051422, "learning_rate": 7.58923925470556e-05, "loss": 0.015, "step": 108120 }, { "grad_norm": 0.09065669029951096, "learning_rate": 7.588767739525659e-05, "loss": 0.0181, "step": 108130 }, { "grad_norm": 0.11030968278646469, "learning_rate": 7.588296192889784e-05, "loss": 0.0142, "step": 108140 }, { "grad_norm": 0.24027007818222046, "learning_rate": 7.587824614803667e-05, "loss": 0.0135, "step": 108150 }, { "grad_norm": 0.13976185023784637, "learning_rate": 7.58735300527304e-05, "loss": 0.0161, "step": 108160 }, { "grad_norm": 0.2675187885761261, "learning_rate": 7.586881364303629e-05, "loss": 0.0229, "step": 108170 }, { "grad_norm": 0.27384287118911743, "learning_rate": 7.586409691901167e-05, "loss": 0.0216, "step": 108180 }, { "grad_norm": 0.13105015456676483, "learning_rate": 7.585937988071386e-05, "loss": 0.0179, "step": 108190 }, { "grad_norm": 0.0984480232000351, "learning_rate": 7.585466252820014e-05, "loss": 0.0235, "step": 108200 }, { "grad_norm": 0.25989362597465515, "learning_rate": 7.584994486152788e-05, "loss": 0.0162, "step": 108210 }, { "grad_norm": 0.06989555060863495, "learning_rate": 7.58452268807544e-05, "loss": 0.0182, "step": 108220 }, { "grad_norm": 0.1033216193318367, "learning_rate": 7.584050858593699e-05, "loss": 0.0168, "step": 108230 }, { "grad_norm": 0.13052304089069366, "learning_rate": 7.5835789977133e-05, "loss": 0.0154, "step": 108240 }, { "grad_norm": 0.1236162930727005, "learning_rate": 7.583107105439979e-05, "loss": 0.0141, "step": 108250 }, { "grad_norm": 0.21003201603889465, "learning_rate": 7.582635181779467e-05, "loss": 0.0224, "step": 108260 }, { "grad_norm": 0.08894944190979004, "learning_rate": 7.582163226737499e-05, "loss": 0.0139, "step": 108270 }, { "grad_norm": 0.1268407702445984, "learning_rate": 7.58169124031981e-05, "loss": 0.0201, "step": 108280 }, { "grad_norm": 0.16781479120254517, "learning_rate": 7.581219222532134e-05, "loss": 0.0188, "step": 108290 }, { "grad_norm": 0.11206068098545074, "learning_rate": 7.580747173380209e-05, "loss": 0.0168, "step": 108300 }, { "grad_norm": 0.1362084001302719, "learning_rate": 7.580275092869766e-05, "loss": 0.015, "step": 108310 }, { "grad_norm": 0.0990123599767685, "learning_rate": 7.579802981006548e-05, "loss": 0.0144, "step": 108320 }, { "grad_norm": 0.08407631516456604, "learning_rate": 7.579330837796287e-05, "loss": 0.0199, "step": 108330 }, { "grad_norm": 0.06416098773479462, "learning_rate": 7.57885866324472e-05, "loss": 0.0122, "step": 108340 }, { "grad_norm": 0.07008426636457443, "learning_rate": 7.578386457357587e-05, "loss": 0.0296, "step": 108350 }, { "grad_norm": 0.2640101909637451, "learning_rate": 7.577914220140623e-05, "loss": 0.0142, "step": 108360 }, { "grad_norm": 0.15436986088752747, "learning_rate": 7.577441951599567e-05, "loss": 0.0195, "step": 108370 }, { "grad_norm": 0.2615632116794586, "learning_rate": 7.576969651740157e-05, "loss": 0.022, "step": 108380 }, { "grad_norm": 0.277328222990036, "learning_rate": 7.576497320568135e-05, "loss": 0.0207, "step": 108390 }, { "grad_norm": 0.11972983181476593, "learning_rate": 7.576024958089236e-05, "loss": 0.0145, "step": 108400 }, { "grad_norm": 0.22435536980628967, "learning_rate": 7.575552564309202e-05, "loss": 0.0245, "step": 108410 }, { "grad_norm": 0.1630164533853531, "learning_rate": 7.575080139233774e-05, "loss": 0.0167, "step": 108420 }, { "grad_norm": 0.11027640849351883, "learning_rate": 7.57460768286869e-05, "loss": 0.013, "step": 108430 }, { "grad_norm": 0.13211113214492798, "learning_rate": 7.574135195219694e-05, "loss": 0.0126, "step": 108440 }, { "grad_norm": 0.15094298124313354, "learning_rate": 7.57366267629252e-05, "loss": 0.013, "step": 108450 }, { "grad_norm": 0.11697249859571457, "learning_rate": 7.573190126092919e-05, "loss": 0.0132, "step": 108460 }, { "grad_norm": 0.1339435875415802, "learning_rate": 7.572717544626627e-05, "loss": 0.0218, "step": 108470 }, { "grad_norm": 0.1333601474761963, "learning_rate": 7.572244931899388e-05, "loss": 0.0114, "step": 108480 }, { "grad_norm": 0.13770785927772522, "learning_rate": 7.571772287916945e-05, "loss": 0.0168, "step": 108490 }, { "grad_norm": 0.3656400144100189, "learning_rate": 7.57129961268504e-05, "loss": 0.0192, "step": 108500 }, { "grad_norm": 0.20982739329338074, "learning_rate": 7.570826906209417e-05, "loss": 0.0227, "step": 108510 }, { "grad_norm": 0.19211922585964203, "learning_rate": 7.570354168495818e-05, "loss": 0.0184, "step": 108520 }, { "grad_norm": 0.2684617340564728, "learning_rate": 7.569881399549992e-05, "loss": 0.0203, "step": 108530 }, { "grad_norm": 0.2014337182044983, "learning_rate": 7.569408599377678e-05, "loss": 0.0145, "step": 108540 }, { "grad_norm": 0.11764080077409744, "learning_rate": 7.568935767984624e-05, "loss": 0.0145, "step": 108550 }, { "grad_norm": 0.11647230386734009, "learning_rate": 7.568462905376577e-05, "loss": 0.0186, "step": 108560 }, { "grad_norm": 0.16512541472911835, "learning_rate": 7.567990011559278e-05, "loss": 0.0147, "step": 108570 }, { "grad_norm": 0.17389586567878723, "learning_rate": 7.567517086538478e-05, "loss": 0.0153, "step": 108580 }, { "grad_norm": 0.14836134016513824, "learning_rate": 7.56704413031992e-05, "loss": 0.0195, "step": 108590 }, { "grad_norm": 0.08988198637962341, "learning_rate": 7.566571142909353e-05, "loss": 0.015, "step": 108600 }, { "grad_norm": 0.10060248523950577, "learning_rate": 7.566098124312524e-05, "loss": 0.0118, "step": 108610 }, { "grad_norm": 0.14420869946479797, "learning_rate": 7.565625074535179e-05, "loss": 0.016, "step": 108620 }, { "grad_norm": 0.07581894099712372, "learning_rate": 7.565151993583066e-05, "loss": 0.0141, "step": 108630 }, { "grad_norm": 0.15820908546447754, "learning_rate": 7.564678881461935e-05, "loss": 0.0151, "step": 108640 }, { "grad_norm": 0.16362519562244415, "learning_rate": 7.564205738177535e-05, "loss": 0.0158, "step": 108650 }, { "grad_norm": 0.1181502640247345, "learning_rate": 7.563732563735613e-05, "loss": 0.0136, "step": 108660 }, { "grad_norm": 0.28022709488868713, "learning_rate": 7.56325935814192e-05, "loss": 0.0186, "step": 108670 }, { "grad_norm": 0.11497798562049866, "learning_rate": 7.562786121402208e-05, "loss": 0.0169, "step": 108680 }, { "grad_norm": 0.2824694514274597, "learning_rate": 7.562312853522222e-05, "loss": 0.0216, "step": 108690 }, { "grad_norm": 0.16507147252559662, "learning_rate": 7.561839554507717e-05, "loss": 0.0208, "step": 108700 }, { "grad_norm": 0.20162783563137054, "learning_rate": 7.561366224364443e-05, "loss": 0.0197, "step": 108710 }, { "grad_norm": 0.30132126808166504, "learning_rate": 7.56089286309815e-05, "loss": 0.0225, "step": 108720 }, { "grad_norm": 0.12876035273075104, "learning_rate": 7.56041947071459e-05, "loss": 0.0187, "step": 108730 }, { "grad_norm": 0.09357351064682007, "learning_rate": 7.559946047219518e-05, "loss": 0.0177, "step": 108740 }, { "grad_norm": 0.2899240553379059, "learning_rate": 7.559472592618683e-05, "loss": 0.0255, "step": 108750 }, { "grad_norm": 0.1408970206975937, "learning_rate": 7.558999106917841e-05, "loss": 0.0136, "step": 108760 }, { "grad_norm": 0.3174547553062439, "learning_rate": 7.558525590122743e-05, "loss": 0.0236, "step": 108770 }, { "grad_norm": 0.17856265604496002, "learning_rate": 7.558052042239144e-05, "loss": 0.016, "step": 108780 }, { "grad_norm": 0.37626221776008606, "learning_rate": 7.557578463272797e-05, "loss": 0.0218, "step": 108790 }, { "grad_norm": 0.13024966418743134, "learning_rate": 7.557104853229457e-05, "loss": 0.0151, "step": 108800 }, { "grad_norm": 0.1116037517786026, "learning_rate": 7.556631212114878e-05, "loss": 0.016, "step": 108810 }, { "grad_norm": 0.17734025418758392, "learning_rate": 7.556157539934817e-05, "loss": 0.0145, "step": 108820 }, { "grad_norm": 0.10442502796649933, "learning_rate": 7.555683836695028e-05, "loss": 0.0125, "step": 108830 }, { "grad_norm": 0.46070441603660583, "learning_rate": 7.555210102401267e-05, "loss": 0.0197, "step": 108840 }, { "grad_norm": 0.17853862047195435, "learning_rate": 7.554736337059292e-05, "loss": 0.0147, "step": 108850 }, { "grad_norm": 0.10374338924884796, "learning_rate": 7.554262540674857e-05, "loss": 0.0161, "step": 108860 }, { "grad_norm": 0.1404549777507782, "learning_rate": 7.55378871325372e-05, "loss": 0.0169, "step": 108870 }, { "grad_norm": 0.13607308268547058, "learning_rate": 7.553314854801641e-05, "loss": 0.0204, "step": 108880 }, { "grad_norm": 0.13486652076244354, "learning_rate": 7.552840965324375e-05, "loss": 0.0165, "step": 108890 }, { "grad_norm": 0.11108959466218948, "learning_rate": 7.552367044827681e-05, "loss": 0.0156, "step": 108900 }, { "grad_norm": 0.23758429288864136, "learning_rate": 7.551893093317318e-05, "loss": 0.0177, "step": 108910 }, { "grad_norm": 0.08904172480106354, "learning_rate": 7.551419110799043e-05, "loss": 0.0167, "step": 108920 }, { "grad_norm": 0.08250151574611664, "learning_rate": 7.550945097278617e-05, "loss": 0.0136, "step": 108930 }, { "grad_norm": 0.14900889992713928, "learning_rate": 7.550471052761801e-05, "loss": 0.0171, "step": 108940 }, { "grad_norm": 0.10315093398094177, "learning_rate": 7.549996977254352e-05, "loss": 0.0126, "step": 108950 }, { "grad_norm": 0.24243032932281494, "learning_rate": 7.549522870762034e-05, "loss": 0.0135, "step": 108960 }, { "grad_norm": 0.1411363035440445, "learning_rate": 7.549048733290604e-05, "loss": 0.0123, "step": 108970 }, { "grad_norm": 0.13530181348323822, "learning_rate": 7.548574564845827e-05, "loss": 0.0215, "step": 108980 }, { "grad_norm": 0.18182741105556488, "learning_rate": 7.548100365433462e-05, "loss": 0.0228, "step": 108990 }, { "grad_norm": 0.18073445558547974, "learning_rate": 7.547626135059272e-05, "loss": 0.0208, "step": 109000 }, { "grad_norm": 0.155302032828331, "learning_rate": 7.547151873729019e-05, "loss": 0.0168, "step": 109010 }, { "grad_norm": 0.0919308066368103, "learning_rate": 7.546677581448466e-05, "loss": 0.0196, "step": 109020 }, { "grad_norm": 0.20741824805736542, "learning_rate": 7.546203258223375e-05, "loss": 0.019, "step": 109030 }, { "grad_norm": 0.07796276360750198, "learning_rate": 7.545728904059511e-05, "loss": 0.0161, "step": 109040 }, { "grad_norm": 0.12922623753547668, "learning_rate": 7.545254518962638e-05, "loss": 0.0131, "step": 109050 }, { "grad_norm": 0.14956142008304596, "learning_rate": 7.54478010293852e-05, "loss": 0.0165, "step": 109060 }, { "grad_norm": 0.11354786157608032, "learning_rate": 7.544305655992919e-05, "loss": 0.018, "step": 109070 }, { "grad_norm": 0.1038462221622467, "learning_rate": 7.543831178131603e-05, "loss": 0.0151, "step": 109080 }, { "grad_norm": 0.12164083123207092, "learning_rate": 7.543356669360335e-05, "loss": 0.0146, "step": 109090 }, { "grad_norm": 0.12016307562589645, "learning_rate": 7.542882129684885e-05, "loss": 0.0188, "step": 109100 }, { "grad_norm": 0.13816148042678833, "learning_rate": 7.542407559111013e-05, "loss": 0.0117, "step": 109110 }, { "grad_norm": 0.1189880222082138, "learning_rate": 7.541932957644492e-05, "loss": 0.0173, "step": 109120 }, { "grad_norm": 0.16198313236236572, "learning_rate": 7.541458325291085e-05, "loss": 0.019, "step": 109130 }, { "grad_norm": 0.11533212661743164, "learning_rate": 7.540983662056557e-05, "loss": 0.0168, "step": 109140 }, { "grad_norm": 0.12404297292232513, "learning_rate": 7.540508967946682e-05, "loss": 0.0171, "step": 109150 }, { "grad_norm": 0.14428743720054626, "learning_rate": 7.540034242967223e-05, "loss": 0.0142, "step": 109160 }, { "grad_norm": 0.10606022924184799, "learning_rate": 7.539559487123949e-05, "loss": 0.0173, "step": 109170 }, { "grad_norm": 0.08641829341650009, "learning_rate": 7.53908470042263e-05, "loss": 0.0177, "step": 109180 }, { "grad_norm": 0.1501002013683319, "learning_rate": 7.538609882869035e-05, "loss": 0.019, "step": 109190 }, { "grad_norm": 0.23827481269836426, "learning_rate": 7.538135034468931e-05, "loss": 0.0255, "step": 109200 }, { "grad_norm": 0.11429520696401596, "learning_rate": 7.537660155228092e-05, "loss": 0.0143, "step": 109210 }, { "grad_norm": 0.10777326673269272, "learning_rate": 7.537185245152286e-05, "loss": 0.0159, "step": 109220 }, { "grad_norm": 0.1507164090871811, "learning_rate": 7.536710304247283e-05, "loss": 0.0149, "step": 109230 }, { "grad_norm": 0.20503193140029907, "learning_rate": 7.536235332518855e-05, "loss": 0.0228, "step": 109240 }, { "grad_norm": 0.09116800129413605, "learning_rate": 7.535760329972772e-05, "loss": 0.0131, "step": 109250 }, { "grad_norm": 0.12581022083759308, "learning_rate": 7.535285296614809e-05, "loss": 0.018, "step": 109260 }, { "grad_norm": 0.20579153299331665, "learning_rate": 7.534810232450736e-05, "loss": 0.0181, "step": 109270 }, { "grad_norm": 0.14805303514003754, "learning_rate": 7.534335137486323e-05, "loss": 0.0192, "step": 109280 }, { "grad_norm": 0.15308578312397003, "learning_rate": 7.533860011727348e-05, "loss": 0.0162, "step": 109290 }, { "grad_norm": 0.10082697123289108, "learning_rate": 7.533384855179581e-05, "loss": 0.0172, "step": 109300 }, { "grad_norm": 0.18517477810382843, "learning_rate": 7.532909667848795e-05, "loss": 0.0179, "step": 109310 }, { "grad_norm": 0.11873334646224976, "learning_rate": 7.532434449740765e-05, "loss": 0.0164, "step": 109320 }, { "grad_norm": 0.21991227567195892, "learning_rate": 7.531959200861267e-05, "loss": 0.0198, "step": 109330 }, { "grad_norm": 0.14007607102394104, "learning_rate": 7.531483921216072e-05, "loss": 0.014, "step": 109340 }, { "grad_norm": 0.15058059990406036, "learning_rate": 7.531008610810958e-05, "loss": 0.0175, "step": 109350 }, { "grad_norm": 0.1255333125591278, "learning_rate": 7.530533269651701e-05, "loss": 0.0154, "step": 109360 }, { "grad_norm": 0.159871906042099, "learning_rate": 7.530057897744074e-05, "loss": 0.0169, "step": 109370 }, { "grad_norm": 0.17940151691436768, "learning_rate": 7.529582495093854e-05, "loss": 0.0177, "step": 109380 }, { "grad_norm": 0.10407131165266037, "learning_rate": 7.529107061706822e-05, "loss": 0.0158, "step": 109390 }, { "grad_norm": 0.14886663854122162, "learning_rate": 7.528631597588746e-05, "loss": 0.0186, "step": 109400 }, { "grad_norm": 0.1272914558649063, "learning_rate": 7.528156102745413e-05, "loss": 0.0144, "step": 109410 }, { "grad_norm": 0.20467637479305267, "learning_rate": 7.527680577182595e-05, "loss": 0.0181, "step": 109420 }, { "grad_norm": 0.10677749663591385, "learning_rate": 7.527205020906071e-05, "loss": 0.0174, "step": 109430 }, { "grad_norm": 0.23438633978366852, "learning_rate": 7.52672943392162e-05, "loss": 0.0235, "step": 109440 }, { "grad_norm": 0.16958466172218323, "learning_rate": 7.526253816235023e-05, "loss": 0.0144, "step": 109450 }, { "grad_norm": 0.1415642946958542, "learning_rate": 7.525778167852054e-05, "loss": 0.0158, "step": 109460 }, { "grad_norm": 0.39702484011650085, "learning_rate": 7.525302488778497e-05, "loss": 0.0239, "step": 109470 }, { "grad_norm": 0.12466470152139664, "learning_rate": 7.52482677902013e-05, "loss": 0.02, "step": 109480 }, { "grad_norm": 0.14763711392879486, "learning_rate": 7.524351038582735e-05, "loss": 0.0187, "step": 109490 }, { "grad_norm": 0.28101104497909546, "learning_rate": 7.523875267472089e-05, "loss": 0.0209, "step": 109500 }, { "grad_norm": 0.14997760951519012, "learning_rate": 7.523399465693977e-05, "loss": 0.016, "step": 109510 }, { "grad_norm": 0.11450081318616867, "learning_rate": 7.52292363325418e-05, "loss": 0.0188, "step": 109520 }, { "grad_norm": 0.13075660169124603, "learning_rate": 7.522447770158477e-05, "loss": 0.0174, "step": 109530 }, { "grad_norm": 0.1584513634443283, "learning_rate": 7.521971876412653e-05, "loss": 0.02, "step": 109540 }, { "grad_norm": 0.12590104341506958, "learning_rate": 7.521495952022489e-05, "loss": 0.0159, "step": 109550 }, { "grad_norm": 0.1047654077410698, "learning_rate": 7.52101999699377e-05, "loss": 0.0129, "step": 109560 }, { "grad_norm": 0.11246182024478912, "learning_rate": 7.520544011332276e-05, "loss": 0.0201, "step": 109570 }, { "grad_norm": 0.10554236173629761, "learning_rate": 7.520067995043794e-05, "loss": 0.0191, "step": 109580 }, { "grad_norm": 0.2349259853363037, "learning_rate": 7.519591948134104e-05, "loss": 0.0264, "step": 109590 }, { "grad_norm": 0.20909149944782257, "learning_rate": 7.519115870608996e-05, "loss": 0.0268, "step": 109600 }, { "grad_norm": 0.19549447298049927, "learning_rate": 7.51863976247425e-05, "loss": 0.0185, "step": 109610 }, { "grad_norm": 0.13118433952331543, "learning_rate": 7.518163623735654e-05, "loss": 0.0159, "step": 109620 }, { "grad_norm": 0.16289551556110382, "learning_rate": 7.517687454398992e-05, "loss": 0.0221, "step": 109630 }, { "grad_norm": 0.5509017705917358, "learning_rate": 7.517211254470049e-05, "loss": 0.0183, "step": 109640 }, { "grad_norm": 0.23863635957241058, "learning_rate": 7.516735023954614e-05, "loss": 0.0173, "step": 109650 }, { "grad_norm": 0.0781233012676239, "learning_rate": 7.516258762858472e-05, "loss": 0.0184, "step": 109660 }, { "grad_norm": 0.12322953343391418, "learning_rate": 7.51578247118741e-05, "loss": 0.0153, "step": 109670 }, { "grad_norm": 0.3836939036846161, "learning_rate": 7.515306148947215e-05, "loss": 0.0236, "step": 109680 }, { "grad_norm": 0.1201063022017479, "learning_rate": 7.514829796143677e-05, "loss": 0.0183, "step": 109690 }, { "grad_norm": 0.08596941083669662, "learning_rate": 7.514353412782583e-05, "loss": 0.0164, "step": 109700 }, { "grad_norm": 0.2687908113002777, "learning_rate": 7.513876998869719e-05, "loss": 0.0148, "step": 109710 }, { "grad_norm": 0.11554422229528427, "learning_rate": 7.513400554410877e-05, "loss": 0.0173, "step": 109720 }, { "grad_norm": 0.2027084082365036, "learning_rate": 7.512924079411846e-05, "loss": 0.0199, "step": 109730 }, { "grad_norm": 0.15890611708164215, "learning_rate": 7.512447573878414e-05, "loss": 0.0178, "step": 109740 }, { "grad_norm": 0.13706010580062866, "learning_rate": 7.51197103781637e-05, "loss": 0.0169, "step": 109750 }, { "grad_norm": 0.09650546312332153, "learning_rate": 7.511494471231509e-05, "loss": 0.0183, "step": 109760 }, { "grad_norm": 0.28060007095336914, "learning_rate": 7.511017874129618e-05, "loss": 0.0215, "step": 109770 }, { "grad_norm": 0.20489338040351868, "learning_rate": 7.51054124651649e-05, "loss": 0.0158, "step": 109780 }, { "grad_norm": 0.141753688454628, "learning_rate": 7.510064588397914e-05, "loss": 0.0168, "step": 109790 }, { "grad_norm": 0.18532301485538483, "learning_rate": 7.509587899779684e-05, "loss": 0.0143, "step": 109800 }, { "grad_norm": 0.11019591242074966, "learning_rate": 7.509111180667591e-05, "loss": 0.0149, "step": 109810 }, { "grad_norm": 0.25724029541015625, "learning_rate": 7.50863443106743e-05, "loss": 0.0283, "step": 109820 }, { "grad_norm": 0.11763332784175873, "learning_rate": 7.50815765098499e-05, "loss": 0.0288, "step": 109830 }, { "grad_norm": 0.16917352378368378, "learning_rate": 7.507680840426067e-05, "loss": 0.017, "step": 109840 }, { "grad_norm": 0.15298736095428467, "learning_rate": 7.507203999396455e-05, "loss": 0.0161, "step": 109850 }, { "grad_norm": 0.0977083370089531, "learning_rate": 7.506727127901946e-05, "loss": 0.0137, "step": 109860 }, { "grad_norm": 0.10359272360801697, "learning_rate": 7.506250225948336e-05, "loss": 0.0156, "step": 109870 }, { "grad_norm": 0.33852696418762207, "learning_rate": 7.505773293541421e-05, "loss": 0.0227, "step": 109880 }, { "grad_norm": 0.17028839886188507, "learning_rate": 7.505296330686993e-05, "loss": 0.0162, "step": 109890 }, { "grad_norm": 0.3396780788898468, "learning_rate": 7.50481933739085e-05, "loss": 0.0179, "step": 109900 }, { "grad_norm": 0.1336807757616043, "learning_rate": 7.504342313658785e-05, "loss": 0.0204, "step": 109910 }, { "grad_norm": 0.10578586161136627, "learning_rate": 7.503865259496599e-05, "loss": 0.0157, "step": 109920 }, { "grad_norm": 0.13161540031433105, "learning_rate": 7.503388174910084e-05, "loss": 0.0112, "step": 109930 }, { "grad_norm": 0.12167531251907349, "learning_rate": 7.50291105990504e-05, "loss": 0.0219, "step": 109940 }, { "grad_norm": 0.1162378191947937, "learning_rate": 7.502433914487265e-05, "loss": 0.0189, "step": 109950 }, { "grad_norm": 0.19164864718914032, "learning_rate": 7.501956738662554e-05, "loss": 0.0218, "step": 109960 }, { "grad_norm": 0.10284529626369476, "learning_rate": 7.501479532436706e-05, "loss": 0.0185, "step": 109970 }, { "grad_norm": 0.14876127243041992, "learning_rate": 7.501002295815521e-05, "loss": 0.0182, "step": 109980 }, { "grad_norm": 0.13589373230934143, "learning_rate": 7.500525028804794e-05, "loss": 0.0144, "step": 109990 }, { "grad_norm": 0.21504248678684235, "learning_rate": 7.500047731410329e-05, "loss": 0.02, "step": 110000 }, { "grad_norm": 0.13808168470859528, "learning_rate": 7.499570403637924e-05, "loss": 0.0149, "step": 110010 }, { "grad_norm": 0.19682610034942627, "learning_rate": 7.499093045493379e-05, "loss": 0.0198, "step": 110020 }, { "grad_norm": 0.13297313451766968, "learning_rate": 7.498615656982493e-05, "loss": 0.0165, "step": 110030 }, { "grad_norm": 0.24408985674381256, "learning_rate": 7.498138238111068e-05, "loss": 0.0148, "step": 110040 }, { "grad_norm": 0.18066610395908356, "learning_rate": 7.497660788884903e-05, "loss": 0.0166, "step": 110050 }, { "grad_norm": 0.12505239248275757, "learning_rate": 7.497183309309804e-05, "loss": 0.0196, "step": 110060 }, { "grad_norm": 0.21822558343410492, "learning_rate": 7.496705799391569e-05, "loss": 0.0217, "step": 110070 }, { "grad_norm": 0.1320957988500595, "learning_rate": 7.496228259136001e-05, "loss": 0.0181, "step": 110080 }, { "grad_norm": 0.09978280961513519, "learning_rate": 7.495750688548902e-05, "loss": 0.0142, "step": 110090 }, { "grad_norm": 0.1490941196680069, "learning_rate": 7.495273087636077e-05, "loss": 0.0208, "step": 110100 }, { "grad_norm": 0.14316140115261078, "learning_rate": 7.494795456403329e-05, "loss": 0.0222, "step": 110110 }, { "grad_norm": 0.164677232503891, "learning_rate": 7.494317794856459e-05, "loss": 0.0133, "step": 110120 }, { "grad_norm": 0.3508874177932739, "learning_rate": 7.493840103001273e-05, "loss": 0.0201, "step": 110130 }, { "grad_norm": 0.1300192028284073, "learning_rate": 7.493362380843576e-05, "loss": 0.0165, "step": 110140 }, { "grad_norm": 0.3026164770126343, "learning_rate": 7.49288462838917e-05, "loss": 0.0174, "step": 110150 }, { "grad_norm": 0.16722045838832855, "learning_rate": 7.492406845643864e-05, "loss": 0.0122, "step": 110160 }, { "grad_norm": 0.057560697197914124, "learning_rate": 7.491929032613461e-05, "loss": 0.0108, "step": 110170 }, { "grad_norm": 0.08009479194879532, "learning_rate": 7.491451189303769e-05, "loss": 0.0153, "step": 110180 }, { "grad_norm": 0.1526009887456894, "learning_rate": 7.49097331572059e-05, "loss": 0.0175, "step": 110190 }, { "grad_norm": 0.09706718474626541, "learning_rate": 7.490495411869735e-05, "loss": 0.0152, "step": 110200 }, { "grad_norm": 0.10385295748710632, "learning_rate": 7.490017477757009e-05, "loss": 0.0147, "step": 110210 }, { "grad_norm": 0.11700908839702606, "learning_rate": 7.489539513388219e-05, "loss": 0.014, "step": 110220 }, { "grad_norm": 0.3727584779262543, "learning_rate": 7.489061518769173e-05, "loss": 0.0187, "step": 110230 }, { "grad_norm": 0.09116989374160767, "learning_rate": 7.488583493905682e-05, "loss": 0.0154, "step": 110240 }, { "grad_norm": 0.12004461884498596, "learning_rate": 7.48810543880355e-05, "loss": 0.0184, "step": 110250 }, { "grad_norm": 0.15486370027065277, "learning_rate": 7.487627353468589e-05, "loss": 0.0193, "step": 110260 }, { "grad_norm": 0.09690110385417938, "learning_rate": 7.487149237906606e-05, "loss": 0.0187, "step": 110270 }, { "grad_norm": 0.12570108473300934, "learning_rate": 7.486671092123412e-05, "loss": 0.0203, "step": 110280 }, { "grad_norm": 0.08636609464883804, "learning_rate": 7.486192916124817e-05, "loss": 0.0167, "step": 110290 }, { "grad_norm": 0.10803794860839844, "learning_rate": 7.48571470991663e-05, "loss": 0.0143, "step": 110300 }, { "grad_norm": 0.1283240020275116, "learning_rate": 7.485236473504662e-05, "loss": 0.0163, "step": 110310 }, { "grad_norm": 0.14876314997673035, "learning_rate": 7.484758206894724e-05, "loss": 0.0154, "step": 110320 }, { "grad_norm": 0.14005956053733826, "learning_rate": 7.484279910092628e-05, "loss": 0.02, "step": 110330 }, { "grad_norm": 0.22058984637260437, "learning_rate": 7.483801583104186e-05, "loss": 0.0204, "step": 110340 }, { "grad_norm": 0.1947595626115799, "learning_rate": 7.48332322593521e-05, "loss": 0.0137, "step": 110350 }, { "grad_norm": 0.09210599213838577, "learning_rate": 7.482844838591512e-05, "loss": 0.0105, "step": 110360 }, { "grad_norm": 0.4236738681793213, "learning_rate": 7.482366421078903e-05, "loss": 0.0121, "step": 110370 }, { "grad_norm": 0.1382744461297989, "learning_rate": 7.481887973403202e-05, "loss": 0.0126, "step": 110380 }, { "grad_norm": 0.10027588903903961, "learning_rate": 7.481409495570215e-05, "loss": 0.0142, "step": 110390 }, { "grad_norm": 0.12513229250907898, "learning_rate": 7.480930987585762e-05, "loss": 0.0166, "step": 110400 }, { "grad_norm": 0.18393264710903168, "learning_rate": 7.480452449455655e-05, "loss": 0.0165, "step": 110410 }, { "grad_norm": 0.15763594210147858, "learning_rate": 7.479973881185708e-05, "loss": 0.0124, "step": 110420 }, { "grad_norm": 0.10309659689664841, "learning_rate": 7.47949528278174e-05, "loss": 0.0144, "step": 110430 }, { "grad_norm": 0.1085144579410553, "learning_rate": 7.479016654249558e-05, "loss": 0.0196, "step": 110440 }, { "grad_norm": 0.10682875663042068, "learning_rate": 7.478537995594988e-05, "loss": 0.0161, "step": 110450 }, { "grad_norm": 0.13649627566337585, "learning_rate": 7.478059306823838e-05, "loss": 0.0184, "step": 110460 }, { "grad_norm": 0.09698404371738434, "learning_rate": 7.47758058794193e-05, "loss": 0.0126, "step": 110470 }, { "grad_norm": 0.11740433424711227, "learning_rate": 7.477101838955077e-05, "loss": 0.0167, "step": 110480 }, { "grad_norm": 0.14703838527202606, "learning_rate": 7.476623059869098e-05, "loss": 0.0128, "step": 110490 }, { "grad_norm": 0.1416357457637787, "learning_rate": 7.47614425068981e-05, "loss": 0.0144, "step": 110500 }, { "grad_norm": 0.06477468460798264, "learning_rate": 7.475665411423032e-05, "loss": 0.0129, "step": 110510 }, { "grad_norm": 0.18335795402526855, "learning_rate": 7.475186542074584e-05, "loss": 0.0242, "step": 110520 }, { "grad_norm": 0.18565633893013, "learning_rate": 7.474707642650279e-05, "loss": 0.0193, "step": 110530 }, { "grad_norm": 0.13068954646587372, "learning_rate": 7.474228713155942e-05, "loss": 0.0175, "step": 110540 }, { "grad_norm": 0.16061972081661224, "learning_rate": 7.473749753597389e-05, "loss": 0.0199, "step": 110550 }, { "grad_norm": 0.2906905710697174, "learning_rate": 7.473270763980441e-05, "loss": 0.0184, "step": 110560 }, { "grad_norm": 0.37995433807373047, "learning_rate": 7.472791744310919e-05, "loss": 0.0194, "step": 110570 }, { "grad_norm": 0.1174277514219284, "learning_rate": 7.472312694594641e-05, "loss": 0.0209, "step": 110580 }, { "grad_norm": 0.11085907369852066, "learning_rate": 7.471833614837431e-05, "loss": 0.0165, "step": 110590 }, { "grad_norm": 0.1686118096113205, "learning_rate": 7.471354505045108e-05, "loss": 0.0278, "step": 110600 }, { "grad_norm": 0.1016639918088913, "learning_rate": 7.470875365223496e-05, "loss": 0.0188, "step": 110610 }, { "grad_norm": 0.34392014145851135, "learning_rate": 7.470396195378413e-05, "loss": 0.0172, "step": 110620 }, { "grad_norm": 0.132076695561409, "learning_rate": 7.469916995515684e-05, "loss": 0.0194, "step": 110630 }, { "grad_norm": 0.1542704552412033, "learning_rate": 7.469437765641132e-05, "loss": 0.0139, "step": 110640 }, { "grad_norm": 0.15259599685668945, "learning_rate": 7.468958505760581e-05, "loss": 0.0154, "step": 110650 }, { "grad_norm": 0.133221834897995, "learning_rate": 7.468479215879852e-05, "loss": 0.0183, "step": 110660 }, { "grad_norm": 0.17467525601387024, "learning_rate": 7.467999896004769e-05, "loss": 0.0161, "step": 110670 }, { "grad_norm": 0.15952900052070618, "learning_rate": 7.46752054614116e-05, "loss": 0.017, "step": 110680 }, { "grad_norm": 0.29118505120277405, "learning_rate": 7.467041166294844e-05, "loss": 0.018, "step": 110690 }, { "grad_norm": 0.2402573674917221, "learning_rate": 7.466561756471648e-05, "loss": 0.0176, "step": 110700 }, { "grad_norm": 0.08182262629270554, "learning_rate": 7.4660823166774e-05, "loss": 0.0155, "step": 110710 }, { "grad_norm": 0.16502398252487183, "learning_rate": 7.465602846917923e-05, "loss": 0.0178, "step": 110720 }, { "grad_norm": 0.12483272701501846, "learning_rate": 7.465123347199041e-05, "loss": 0.0218, "step": 110730 }, { "grad_norm": 0.475464791059494, "learning_rate": 7.464643817526586e-05, "loss": 0.0169, "step": 110740 }, { "grad_norm": 0.2700277268886566, "learning_rate": 7.46416425790638e-05, "loss": 0.0167, "step": 110750 }, { "grad_norm": 0.37796148657798767, "learning_rate": 7.463684668344254e-05, "loss": 0.018, "step": 110760 }, { "grad_norm": 0.11847800016403198, "learning_rate": 7.463205048846031e-05, "loss": 0.0187, "step": 110770 }, { "grad_norm": 0.17710544168949127, "learning_rate": 7.462725399417541e-05, "loss": 0.0175, "step": 110780 }, { "grad_norm": 0.10947313904762268, "learning_rate": 7.462245720064612e-05, "loss": 0.0174, "step": 110790 }, { "grad_norm": 0.12241894006729126, "learning_rate": 7.461766010793073e-05, "loss": 0.0164, "step": 110800 }, { "grad_norm": 0.11571957916021347, "learning_rate": 7.461286271608754e-05, "loss": 0.0169, "step": 110810 }, { "grad_norm": 0.3193072974681854, "learning_rate": 7.460806502517481e-05, "loss": 0.0207, "step": 110820 }, { "grad_norm": 0.09173373132944107, "learning_rate": 7.460326703525086e-05, "loss": 0.0163, "step": 110830 }, { "grad_norm": 0.15595954656600952, "learning_rate": 7.4598468746374e-05, "loss": 0.0157, "step": 110840 }, { "grad_norm": 0.2525619864463806, "learning_rate": 7.459367015860252e-05, "loss": 0.0171, "step": 110850 }, { "grad_norm": 0.2546268701553345, "learning_rate": 7.458887127199471e-05, "loss": 0.0199, "step": 110860 }, { "grad_norm": 0.137197807431221, "learning_rate": 7.458407208660892e-05, "loss": 0.0174, "step": 110870 }, { "grad_norm": 0.09263118356466293, "learning_rate": 7.457927260250342e-05, "loss": 0.0131, "step": 110880 }, { "grad_norm": 0.17201077938079834, "learning_rate": 7.457447281973656e-05, "loss": 0.0203, "step": 110890 }, { "grad_norm": 0.24595007300376892, "learning_rate": 7.456967273836665e-05, "loss": 0.0186, "step": 110900 }, { "grad_norm": 0.2578316032886505, "learning_rate": 7.456487235845204e-05, "loss": 0.0159, "step": 110910 }, { "grad_norm": 0.16652347147464752, "learning_rate": 7.4560071680051e-05, "loss": 0.0209, "step": 110920 }, { "grad_norm": 0.1414967030286789, "learning_rate": 7.455527070322193e-05, "loss": 0.0163, "step": 110930 }, { "grad_norm": 0.14236368238925934, "learning_rate": 7.455046942802313e-05, "loss": 0.0191, "step": 110940 }, { "grad_norm": 0.12080882489681244, "learning_rate": 7.454566785451294e-05, "loss": 0.0189, "step": 110950 }, { "grad_norm": 0.12409380078315735, "learning_rate": 7.454086598274972e-05, "loss": 0.02, "step": 110960 }, { "grad_norm": 0.17627574503421783, "learning_rate": 7.45360638127918e-05, "loss": 0.0155, "step": 110970 }, { "grad_norm": 0.17624689638614655, "learning_rate": 7.453126134469755e-05, "loss": 0.0169, "step": 110980 }, { "grad_norm": 0.20576328039169312, "learning_rate": 7.45264585785253e-05, "loss": 0.0178, "step": 110990 }, { "grad_norm": 0.16647358238697052, "learning_rate": 7.452165551433343e-05, "loss": 0.013, "step": 111000 }, { "grad_norm": 0.17567864060401917, "learning_rate": 7.45168521521803e-05, "loss": 0.0147, "step": 111010 }, { "grad_norm": 0.20423991978168488, "learning_rate": 7.451204849212425e-05, "loss": 0.0167, "step": 111020 }, { "grad_norm": 0.17464901506900787, "learning_rate": 7.450724453422368e-05, "loss": 0.0173, "step": 111030 }, { "grad_norm": 0.0761180967092514, "learning_rate": 7.450244027853695e-05, "loss": 0.0153, "step": 111040 }, { "grad_norm": 0.2863599359989166, "learning_rate": 7.449763572512242e-05, "loss": 0.0182, "step": 111050 }, { "grad_norm": 0.12403354048728943, "learning_rate": 7.449283087403851e-05, "loss": 0.0186, "step": 111060 }, { "grad_norm": 0.28175732493400574, "learning_rate": 7.448802572534356e-05, "loss": 0.0156, "step": 111070 }, { "grad_norm": 0.1737304925918579, "learning_rate": 7.4483220279096e-05, "loss": 0.0199, "step": 111080 }, { "grad_norm": 0.31233593821525574, "learning_rate": 7.447841453535418e-05, "loss": 0.0213, "step": 111090 }, { "grad_norm": 0.10768727213144302, "learning_rate": 7.447360849417652e-05, "loss": 0.0172, "step": 111100 }, { "grad_norm": 0.3823252022266388, "learning_rate": 7.44688021556214e-05, "loss": 0.0184, "step": 111110 }, { "grad_norm": 0.10551947355270386, "learning_rate": 7.446399551974723e-05, "loss": 0.0196, "step": 111120 }, { "grad_norm": 0.23144352436065674, "learning_rate": 7.445918858661244e-05, "loss": 0.0138, "step": 111130 }, { "grad_norm": 0.16982993483543396, "learning_rate": 7.445438135627539e-05, "loss": 0.0195, "step": 111140 }, { "grad_norm": 0.0993686243891716, "learning_rate": 7.444957382879454e-05, "loss": 0.0159, "step": 111150 }, { "grad_norm": 0.13119696080684662, "learning_rate": 7.444476600422828e-05, "loss": 0.0128, "step": 111160 }, { "grad_norm": 0.12865538895130157, "learning_rate": 7.4439957882635e-05, "loss": 0.0184, "step": 111170 }, { "grad_norm": 0.15999998152256012, "learning_rate": 7.44351494640732e-05, "loss": 0.018, "step": 111180 }, { "grad_norm": 0.13388952612876892, "learning_rate": 7.443034074860124e-05, "loss": 0.0187, "step": 111190 }, { "grad_norm": 0.09541312605142593, "learning_rate": 7.442553173627757e-05, "loss": 0.0127, "step": 111200 }, { "grad_norm": 0.16018876433372498, "learning_rate": 7.442072242716064e-05, "loss": 0.0175, "step": 111210 }, { "grad_norm": 0.36773866415023804, "learning_rate": 7.441591282130888e-05, "loss": 0.0236, "step": 111220 }, { "grad_norm": 0.2244611382484436, "learning_rate": 7.441110291878072e-05, "loss": 0.0205, "step": 111230 }, { "grad_norm": 0.1291535496711731, "learning_rate": 7.44062927196346e-05, "loss": 0.0184, "step": 111240 }, { "grad_norm": 0.1582288295030594, "learning_rate": 7.4401482223929e-05, "loss": 0.0151, "step": 111250 }, { "grad_norm": 0.17391307651996613, "learning_rate": 7.439667143172232e-05, "loss": 0.0195, "step": 111260 }, { "grad_norm": 0.20194624364376068, "learning_rate": 7.439186034307308e-05, "loss": 0.0223, "step": 111270 }, { "grad_norm": 0.18975955247879028, "learning_rate": 7.43870489580397e-05, "loss": 0.016, "step": 111280 }, { "grad_norm": 0.39168012142181396, "learning_rate": 7.438223727668065e-05, "loss": 0.0199, "step": 111290 }, { "grad_norm": 0.09598863869905472, "learning_rate": 7.437742529905438e-05, "loss": 0.0152, "step": 111300 }, { "grad_norm": 0.23735007643699646, "learning_rate": 7.437261302521941e-05, "loss": 0.0176, "step": 111310 }, { "grad_norm": 0.2004585564136505, "learning_rate": 7.436780045523416e-05, "loss": 0.0175, "step": 111320 }, { "grad_norm": 0.2146221399307251, "learning_rate": 7.436298758915712e-05, "loss": 0.0224, "step": 111330 }, { "grad_norm": 0.301120400428772, "learning_rate": 7.435817442704679e-05, "loss": 0.0225, "step": 111340 }, { "grad_norm": 0.1648356169462204, "learning_rate": 7.435336096896164e-05, "loss": 0.0242, "step": 111350 }, { "grad_norm": 0.16327208280563354, "learning_rate": 7.434854721496017e-05, "loss": 0.0272, "step": 111360 }, { "grad_norm": 0.06783407926559448, "learning_rate": 7.434373316510083e-05, "loss": 0.0181, "step": 111370 }, { "grad_norm": 0.0857158675789833, "learning_rate": 7.433891881944218e-05, "loss": 0.017, "step": 111380 }, { "grad_norm": 0.17704619467258453, "learning_rate": 7.433410417804268e-05, "loss": 0.0139, "step": 111390 }, { "grad_norm": 0.15099070966243744, "learning_rate": 7.432928924096083e-05, "loss": 0.0191, "step": 111400 }, { "grad_norm": 0.23897875845432281, "learning_rate": 7.432447400825515e-05, "loss": 0.0177, "step": 111410 }, { "grad_norm": 0.06813423335552216, "learning_rate": 7.431965847998414e-05, "loss": 0.0166, "step": 111420 }, { "grad_norm": 0.0820167064666748, "learning_rate": 7.431484265620633e-05, "loss": 0.017, "step": 111430 }, { "grad_norm": 0.0861833468079567, "learning_rate": 7.431002653698021e-05, "loss": 0.0215, "step": 111440 }, { "grad_norm": 0.10403557121753693, "learning_rate": 7.430521012236433e-05, "loss": 0.0153, "step": 111450 }, { "grad_norm": 0.15253247320652008, "learning_rate": 7.430039341241719e-05, "loss": 0.0169, "step": 111460 }, { "grad_norm": 0.11313679069280624, "learning_rate": 7.429557640719733e-05, "loss": 0.0134, "step": 111470 }, { "grad_norm": 0.13148590922355652, "learning_rate": 7.429075910676328e-05, "loss": 0.0204, "step": 111480 }, { "grad_norm": 0.11064166575670242, "learning_rate": 7.428594151117357e-05, "loss": 0.014, "step": 111490 }, { "grad_norm": 0.1269715130329132, "learning_rate": 7.428112362048675e-05, "loss": 0.0165, "step": 111500 }, { "grad_norm": 0.13533900678157806, "learning_rate": 7.427630543476134e-05, "loss": 0.016, "step": 111510 }, { "grad_norm": 0.36951664090156555, "learning_rate": 7.42714869540559e-05, "loss": 0.0223, "step": 111520 }, { "grad_norm": 0.09627004712820053, "learning_rate": 7.426666817842898e-05, "loss": 0.0164, "step": 111530 }, { "grad_norm": 0.13745197653770447, "learning_rate": 7.426184910793912e-05, "loss": 0.0163, "step": 111540 }, { "grad_norm": 0.2630833089351654, "learning_rate": 7.42570297426449e-05, "loss": 0.0192, "step": 111550 }, { "grad_norm": 0.10697620362043381, "learning_rate": 7.425221008260487e-05, "loss": 0.0179, "step": 111560 }, { "grad_norm": 0.6140393018722534, "learning_rate": 7.424739012787758e-05, "loss": 0.0203, "step": 111570 }, { "grad_norm": 0.15674591064453125, "learning_rate": 7.424256987852161e-05, "loss": 0.0144, "step": 111580 }, { "grad_norm": 0.1548345685005188, "learning_rate": 7.423774933459551e-05, "loss": 0.0176, "step": 111590 }, { "grad_norm": 0.17810606956481934, "learning_rate": 7.423292849615789e-05, "loss": 0.0173, "step": 111600 }, { "grad_norm": 0.0701943039894104, "learning_rate": 7.42281073632673e-05, "loss": 0.0135, "step": 111610 }, { "grad_norm": 0.15286517143249512, "learning_rate": 7.422328593598234e-05, "loss": 0.0186, "step": 111620 }, { "grad_norm": 0.16539326310157776, "learning_rate": 7.421846421436158e-05, "loss": 0.0194, "step": 111630 }, { "grad_norm": 0.1213562861084938, "learning_rate": 7.421364219846361e-05, "loss": 0.0146, "step": 111640 }, { "grad_norm": 0.20678137242794037, "learning_rate": 7.420881988834704e-05, "loss": 0.0215, "step": 111650 }, { "grad_norm": 0.16787666082382202, "learning_rate": 7.420399728407044e-05, "loss": 0.0179, "step": 111660 }, { "grad_norm": 0.1700584590435028, "learning_rate": 7.419917438569242e-05, "loss": 0.0156, "step": 111670 }, { "grad_norm": 0.18170607089996338, "learning_rate": 7.419435119327158e-05, "loss": 0.0144, "step": 111680 }, { "grad_norm": 0.13982026278972626, "learning_rate": 7.418952770686653e-05, "loss": 0.0161, "step": 111690 }, { "grad_norm": 0.17830534279346466, "learning_rate": 7.41847039265359e-05, "loss": 0.0141, "step": 111700 }, { "grad_norm": 0.16968940198421478, "learning_rate": 7.417987985233828e-05, "loss": 0.0221, "step": 111710 }, { "grad_norm": 0.08906492590904236, "learning_rate": 7.417505548433227e-05, "loss": 0.0194, "step": 111720 }, { "grad_norm": 0.21106453239917755, "learning_rate": 7.417023082257652e-05, "loss": 0.0197, "step": 111730 }, { "grad_norm": 0.13091665506362915, "learning_rate": 7.416540586712964e-05, "loss": 0.0133, "step": 111740 }, { "grad_norm": 0.17538872361183167, "learning_rate": 7.416058061805027e-05, "loss": 0.0152, "step": 111750 }, { "grad_norm": 0.23557057976722717, "learning_rate": 7.415575507539702e-05, "loss": 0.0165, "step": 111760 }, { "grad_norm": 0.1936734914779663, "learning_rate": 7.415092923922856e-05, "loss": 0.021, "step": 111770 }, { "grad_norm": 0.15572381019592285, "learning_rate": 7.414610310960349e-05, "loss": 0.0173, "step": 111780 }, { "grad_norm": 0.0958598405122757, "learning_rate": 7.414127668658049e-05, "loss": 0.0164, "step": 111790 }, { "grad_norm": 0.12045552581548691, "learning_rate": 7.413644997021819e-05, "loss": 0.0195, "step": 111800 }, { "grad_norm": 0.21529710292816162, "learning_rate": 7.41316229605752e-05, "loss": 0.015, "step": 111810 }, { "grad_norm": 0.09256593883037567, "learning_rate": 7.412679565771022e-05, "loss": 0.0196, "step": 111820 }, { "grad_norm": 0.2063179612159729, "learning_rate": 7.41219680616819e-05, "loss": 0.0121, "step": 111830 }, { "grad_norm": 0.22785519063472748, "learning_rate": 7.41171401725489e-05, "loss": 0.0161, "step": 111840 }, { "grad_norm": 0.2190387099981308, "learning_rate": 7.411231199036986e-05, "loss": 0.0188, "step": 111850 }, { "grad_norm": 0.09634178876876831, "learning_rate": 7.410748351520346e-05, "loss": 0.0192, "step": 111860 }, { "grad_norm": 0.39214205741882324, "learning_rate": 7.41026547471084e-05, "loss": 0.017, "step": 111870 }, { "grad_norm": 0.08723193407058716, "learning_rate": 7.409782568614331e-05, "loss": 0.0183, "step": 111880 }, { "grad_norm": 0.19995397329330444, "learning_rate": 7.409299633236689e-05, "loss": 0.0183, "step": 111890 }, { "grad_norm": 0.08796519786119461, "learning_rate": 7.408816668583781e-05, "loss": 0.0156, "step": 111900 }, { "grad_norm": 0.14867070317268372, "learning_rate": 7.408333674661477e-05, "loss": 0.0202, "step": 111910 }, { "grad_norm": 0.182554692029953, "learning_rate": 7.407850651475645e-05, "loss": 0.0215, "step": 111920 }, { "grad_norm": 0.2682846784591675, "learning_rate": 7.407367599032153e-05, "loss": 0.018, "step": 111930 }, { "grad_norm": 0.15720167756080627, "learning_rate": 7.406884517336873e-05, "loss": 0.0165, "step": 111940 }, { "grad_norm": 0.15921859443187714, "learning_rate": 7.406401406395673e-05, "loss": 0.0191, "step": 111950 }, { "grad_norm": 0.11743852496147156, "learning_rate": 7.405918266214426e-05, "loss": 0.0141, "step": 111960 }, { "grad_norm": 0.1298726499080658, "learning_rate": 7.405435096798999e-05, "loss": 0.0161, "step": 111970 }, { "grad_norm": 0.11912448704242706, "learning_rate": 7.404951898155263e-05, "loss": 0.015, "step": 111980 }, { "grad_norm": 0.1266707479953766, "learning_rate": 7.404468670289092e-05, "loss": 0.0161, "step": 111990 }, { "grad_norm": 0.14409855008125305, "learning_rate": 7.403985413206357e-05, "loss": 0.0119, "step": 112000 }, { "grad_norm": 0.09600888937711716, "learning_rate": 7.403502126912929e-05, "loss": 0.0161, "step": 112010 }, { "grad_norm": 0.1619083285331726, "learning_rate": 7.40301881141468e-05, "loss": 0.0179, "step": 112020 }, { "grad_norm": 0.15499989688396454, "learning_rate": 7.402535466717484e-05, "loss": 0.0153, "step": 112030 }, { "grad_norm": 0.18977390229701996, "learning_rate": 7.402052092827214e-05, "loss": 0.0148, "step": 112040 }, { "grad_norm": 0.2411951869726181, "learning_rate": 7.401568689749745e-05, "loss": 0.0184, "step": 112050 }, { "grad_norm": 0.11457444727420807, "learning_rate": 7.401085257490947e-05, "loss": 0.0179, "step": 112060 }, { "grad_norm": 0.18349535763263702, "learning_rate": 7.400601796056696e-05, "loss": 0.0152, "step": 112070 }, { "grad_norm": 0.2257140874862671, "learning_rate": 7.400118305452867e-05, "loss": 0.0208, "step": 112080 }, { "grad_norm": 0.12082207947969437, "learning_rate": 7.399634785685335e-05, "loss": 0.0207, "step": 112090 }, { "grad_norm": 0.2520082890987396, "learning_rate": 7.399151236759974e-05, "loss": 0.0176, "step": 112100 }, { "grad_norm": 0.22945840656757355, "learning_rate": 7.39866765868266e-05, "loss": 0.0158, "step": 112110 }, { "grad_norm": 0.10740393400192261, "learning_rate": 7.398184051459271e-05, "loss": 0.0184, "step": 112120 }, { "grad_norm": 0.10277896374464035, "learning_rate": 7.397700415095679e-05, "loss": 0.0113, "step": 112130 }, { "grad_norm": 0.09353625029325485, "learning_rate": 7.397216749597765e-05, "loss": 0.0203, "step": 112140 }, { "grad_norm": 0.9499643445014954, "learning_rate": 7.396733054971403e-05, "loss": 0.0169, "step": 112150 }, { "grad_norm": 0.24002166092395782, "learning_rate": 7.396249331222472e-05, "loss": 0.0168, "step": 112160 }, { "grad_norm": 0.1079939529299736, "learning_rate": 7.395765578356849e-05, "loss": 0.0196, "step": 112170 }, { "grad_norm": 0.1487198919057846, "learning_rate": 7.395281796380414e-05, "loss": 0.0161, "step": 112180 }, { "grad_norm": 0.12946149706840515, "learning_rate": 7.39479798529904e-05, "loss": 0.0151, "step": 112190 }, { "grad_norm": 0.18685296177864075, "learning_rate": 7.394314145118613e-05, "loss": 0.0195, "step": 112200 }, { "grad_norm": 0.09588474035263062, "learning_rate": 7.393830275845007e-05, "loss": 0.0183, "step": 112210 }, { "grad_norm": 0.2950880527496338, "learning_rate": 7.393346377484103e-05, "loss": 0.0214, "step": 112220 }, { "grad_norm": 0.13039179146289825, "learning_rate": 7.392862450041781e-05, "loss": 0.0143, "step": 112230 }, { "grad_norm": 0.40370988845825195, "learning_rate": 7.39237849352392e-05, "loss": 0.0157, "step": 112240 }, { "grad_norm": 0.2635745704174042, "learning_rate": 7.391894507936403e-05, "loss": 0.0149, "step": 112250 }, { "grad_norm": 0.08613885939121246, "learning_rate": 7.391410493285109e-05, "loss": 0.0184, "step": 112260 }, { "grad_norm": 0.1864638328552246, "learning_rate": 7.39092644957592e-05, "loss": 0.0188, "step": 112270 }, { "grad_norm": 0.10678773373365402, "learning_rate": 7.390442376814716e-05, "loss": 0.0222, "step": 112280 }, { "grad_norm": 0.12814097106456757, "learning_rate": 7.389958275007381e-05, "loss": 0.0181, "step": 112290 }, { "grad_norm": 0.13265150785446167, "learning_rate": 7.389474144159796e-05, "loss": 0.0148, "step": 112300 }, { "grad_norm": 0.09798348695039749, "learning_rate": 7.388989984277845e-05, "loss": 0.0147, "step": 112310 }, { "grad_norm": 0.18312868475914001, "learning_rate": 7.388505795367409e-05, "loss": 0.0169, "step": 112320 }, { "grad_norm": 0.2893839478492737, "learning_rate": 7.388021577434374e-05, "loss": 0.0154, "step": 112330 }, { "grad_norm": 0.16051772236824036, "learning_rate": 7.38753733048462e-05, "loss": 0.0135, "step": 112340 }, { "grad_norm": 0.21048356592655182, "learning_rate": 7.387053054524035e-05, "loss": 0.0128, "step": 112350 }, { "grad_norm": 0.19476629793643951, "learning_rate": 7.3865687495585e-05, "loss": 0.0209, "step": 112360 }, { "grad_norm": 0.3626050651073456, "learning_rate": 7.386084415593904e-05, "loss": 0.0224, "step": 112370 }, { "grad_norm": 0.23689492046833038, "learning_rate": 7.385600052636127e-05, "loss": 0.0227, "step": 112380 }, { "grad_norm": 0.25683119893074036, "learning_rate": 7.385115660691058e-05, "loss": 0.013, "step": 112390 }, { "grad_norm": 0.16342709958553314, "learning_rate": 7.384631239764581e-05, "loss": 0.0159, "step": 112400 }, { "grad_norm": 0.17818285524845123, "learning_rate": 7.384146789862585e-05, "loss": 0.0254, "step": 112410 }, { "grad_norm": 0.26905879378318787, "learning_rate": 7.383662310990952e-05, "loss": 0.0149, "step": 112420 }, { "grad_norm": 0.11321479082107544, "learning_rate": 7.383177803155573e-05, "loss": 0.0174, "step": 112430 }, { "grad_norm": 0.08412790298461914, "learning_rate": 7.382693266362334e-05, "loss": 0.0127, "step": 112440 }, { "grad_norm": 0.18681637942790985, "learning_rate": 7.382208700617121e-05, "loss": 0.0156, "step": 112450 }, { "grad_norm": 0.11120330542325974, "learning_rate": 7.381724105925824e-05, "loss": 0.0183, "step": 112460 }, { "grad_norm": 0.1393079310655594, "learning_rate": 7.381239482294329e-05, "loss": 0.0218, "step": 112470 }, { "grad_norm": 0.09837935864925385, "learning_rate": 7.380754829728528e-05, "loss": 0.0151, "step": 112480 }, { "grad_norm": 0.16409917175769806, "learning_rate": 7.380270148234306e-05, "loss": 0.0152, "step": 112490 }, { "grad_norm": 0.155416339635849, "learning_rate": 7.379785437817556e-05, "loss": 0.0154, "step": 112500 }, { "grad_norm": 0.09968237578868866, "learning_rate": 7.379300698484164e-05, "loss": 0.0182, "step": 112510 }, { "grad_norm": 0.17705316841602325, "learning_rate": 7.378815930240024e-05, "loss": 0.0201, "step": 112520 }, { "grad_norm": 0.15052735805511475, "learning_rate": 7.378331133091024e-05, "loss": 0.0182, "step": 112530 }, { "grad_norm": 0.08493935316801071, "learning_rate": 7.377846307043055e-05, "loss": 0.0193, "step": 112540 }, { "grad_norm": 0.10349863022565842, "learning_rate": 7.377361452102009e-05, "loss": 0.0221, "step": 112550 }, { "grad_norm": 0.08244285732507706, "learning_rate": 7.376876568273776e-05, "loss": 0.019, "step": 112560 }, { "grad_norm": 0.10085524618625641, "learning_rate": 7.376391655564249e-05, "loss": 0.0202, "step": 112570 }, { "grad_norm": 0.08798374980688095, "learning_rate": 7.375906713979318e-05, "loss": 0.018, "step": 112580 }, { "grad_norm": 0.16165754199028015, "learning_rate": 7.37542174352488e-05, "loss": 0.0205, "step": 112590 }, { "grad_norm": 0.11119119822978973, "learning_rate": 7.374936744206824e-05, "loss": 0.0167, "step": 112600 }, { "grad_norm": 0.16566678881645203, "learning_rate": 7.374451716031043e-05, "loss": 0.019, "step": 112610 }, { "grad_norm": 0.12696391344070435, "learning_rate": 7.373966659003433e-05, "loss": 0.0148, "step": 112620 }, { "grad_norm": 0.1260656714439392, "learning_rate": 7.373481573129886e-05, "loss": 0.0185, "step": 112630 }, { "grad_norm": 0.1618443876504898, "learning_rate": 7.372996458416298e-05, "loss": 0.0135, "step": 112640 }, { "grad_norm": 0.13022392988204956, "learning_rate": 7.372511314868561e-05, "loss": 0.0175, "step": 112650 }, { "grad_norm": 0.1582420915365219, "learning_rate": 7.372026142492573e-05, "loss": 0.0192, "step": 112660 }, { "grad_norm": 0.14090678095817566, "learning_rate": 7.371540941294226e-05, "loss": 0.0163, "step": 112670 }, { "grad_norm": 0.19488783180713654, "learning_rate": 7.37105571127942e-05, "loss": 0.0195, "step": 112680 }, { "grad_norm": 0.13312307000160217, "learning_rate": 7.370570452454047e-05, "loss": 0.0124, "step": 112690 }, { "grad_norm": 0.16603338718414307, "learning_rate": 7.370085164824004e-05, "loss": 0.0176, "step": 112700 }, { "grad_norm": 0.22600877285003662, "learning_rate": 7.36959984839519e-05, "loss": 0.0209, "step": 112710 }, { "grad_norm": 0.2132808119058609, "learning_rate": 7.3691145031735e-05, "loss": 0.019, "step": 112720 }, { "grad_norm": 1.217806100845337, "learning_rate": 7.368629129164833e-05, "loss": 0.0179, "step": 112730 }, { "grad_norm": 0.5351449847221375, "learning_rate": 7.368143726375083e-05, "loss": 0.0204, "step": 112740 }, { "grad_norm": 0.1506086140871048, "learning_rate": 7.367658294810154e-05, "loss": 0.015, "step": 112750 }, { "grad_norm": 0.08412507176399231, "learning_rate": 7.367172834475939e-05, "loss": 0.0175, "step": 112760 }, { "grad_norm": 0.9021131992340088, "learning_rate": 7.36668734537834e-05, "loss": 0.0233, "step": 112770 }, { "grad_norm": 0.18968087434768677, "learning_rate": 7.366201827523256e-05, "loss": 0.0213, "step": 112780 }, { "grad_norm": 0.10209179669618607, "learning_rate": 7.365716280916584e-05, "loss": 0.0172, "step": 112790 }, { "grad_norm": 0.16797484457492828, "learning_rate": 7.365230705564227e-05, "loss": 0.0153, "step": 112800 }, { "grad_norm": 0.1341474950313568, "learning_rate": 7.364745101472083e-05, "loss": 0.0188, "step": 112810 }, { "grad_norm": 0.6242405772209167, "learning_rate": 7.364259468646053e-05, "loss": 0.0187, "step": 112820 }, { "grad_norm": 0.13073976337909698, "learning_rate": 7.363773807092038e-05, "loss": 0.016, "step": 112830 }, { "grad_norm": 0.3351243734359741, "learning_rate": 7.363288116815941e-05, "loss": 0.0232, "step": 112840 }, { "grad_norm": 0.29231882095336914, "learning_rate": 7.362802397823663e-05, "loss": 0.0191, "step": 112850 }, { "grad_norm": 0.27204686403274536, "learning_rate": 7.362316650121102e-05, "loss": 0.0214, "step": 112860 }, { "grad_norm": 0.1663995236158371, "learning_rate": 7.361830873714165e-05, "loss": 0.0162, "step": 112870 }, { "grad_norm": 0.11488457024097443, "learning_rate": 7.361345068608752e-05, "loss": 0.0183, "step": 112880 }, { "grad_norm": 0.1957646608352661, "learning_rate": 7.360859234810767e-05, "loss": 0.0162, "step": 112890 }, { "grad_norm": 0.20145468413829803, "learning_rate": 7.360373372326114e-05, "loss": 0.0187, "step": 112900 }, { "grad_norm": 0.17563527822494507, "learning_rate": 7.359887481160697e-05, "loss": 0.0212, "step": 112910 }, { "grad_norm": 0.21295033395290375, "learning_rate": 7.359401561320417e-05, "loss": 0.0168, "step": 112920 }, { "grad_norm": 0.15989285707473755, "learning_rate": 7.358915612811181e-05, "loss": 0.0134, "step": 112930 }, { "grad_norm": 0.2600860893726349, "learning_rate": 7.358429635638894e-05, "loss": 0.0181, "step": 112940 }, { "grad_norm": 0.1291910707950592, "learning_rate": 7.357943629809458e-05, "loss": 0.0179, "step": 112950 }, { "grad_norm": 0.12428493797779083, "learning_rate": 7.357457595328783e-05, "loss": 0.0172, "step": 112960 }, { "grad_norm": 0.1949966996908188, "learning_rate": 7.356971532202771e-05, "loss": 0.0213, "step": 112970 }, { "grad_norm": 0.14818443357944489, "learning_rate": 7.35648544043733e-05, "loss": 0.0206, "step": 112980 }, { "grad_norm": 0.1793886125087738, "learning_rate": 7.355999320038366e-05, "loss": 0.0214, "step": 112990 }, { "grad_norm": 0.15436699986457825, "learning_rate": 7.355513171011787e-05, "loss": 0.0154, "step": 113000 }, { "grad_norm": 0.17365585267543793, "learning_rate": 7.355026993363496e-05, "loss": 0.0168, "step": 113010 }, { "grad_norm": 0.145403191447258, "learning_rate": 7.354540787099406e-05, "loss": 0.0149, "step": 113020 }, { "grad_norm": 0.15369023382663727, "learning_rate": 7.354054552225422e-05, "loss": 0.0117, "step": 113030 }, { "grad_norm": 0.2527059018611908, "learning_rate": 7.353568288747452e-05, "loss": 0.0171, "step": 113040 }, { "grad_norm": 0.10674075037240982, "learning_rate": 7.353081996671404e-05, "loss": 0.0169, "step": 113050 }, { "grad_norm": 0.2694079875946045, "learning_rate": 7.352595676003192e-05, "loss": 0.0193, "step": 113060 }, { "grad_norm": 0.19933642446994781, "learning_rate": 7.352109326748717e-05, "loss": 0.019, "step": 113070 }, { "grad_norm": 0.16259852051734924, "learning_rate": 7.351622948913896e-05, "loss": 0.0146, "step": 113080 }, { "grad_norm": 0.13854560256004333, "learning_rate": 7.351136542504634e-05, "loss": 0.0148, "step": 113090 }, { "grad_norm": 0.1820245087146759, "learning_rate": 7.350650107526845e-05, "loss": 0.0227, "step": 113100 }, { "grad_norm": 0.27020272612571716, "learning_rate": 7.350163643986435e-05, "loss": 0.018, "step": 113110 }, { "grad_norm": 0.3690086603164673, "learning_rate": 7.349677151889321e-05, "loss": 0.0161, "step": 113120 }, { "grad_norm": 0.1669238954782486, "learning_rate": 7.349190631241408e-05, "loss": 0.0227, "step": 113130 }, { "grad_norm": 0.09931308776140213, "learning_rate": 7.348704082048613e-05, "loss": 0.0143, "step": 113140 }, { "grad_norm": 0.0691194161772728, "learning_rate": 7.348217504316845e-05, "loss": 0.0115, "step": 113150 }, { "grad_norm": 0.22550015151500702, "learning_rate": 7.347730898052018e-05, "loss": 0.0218, "step": 113160 }, { "grad_norm": 0.0984552800655365, "learning_rate": 7.347244263260043e-05, "loss": 0.0131, "step": 113170 }, { "grad_norm": 0.19948147237300873, "learning_rate": 7.346757599946834e-05, "loss": 0.0121, "step": 113180 }, { "grad_norm": 0.09638497233390808, "learning_rate": 7.346270908118306e-05, "loss": 0.012, "step": 113190 }, { "grad_norm": 0.18212591111660004, "learning_rate": 7.34578418778037e-05, "loss": 0.0132, "step": 113200 }, { "grad_norm": 0.10373075306415558, "learning_rate": 7.345297438938942e-05, "loss": 0.0168, "step": 113210 }, { "grad_norm": 0.07999027520418167, "learning_rate": 7.344810661599935e-05, "loss": 0.0166, "step": 113220 }, { "grad_norm": 0.09815216064453125, "learning_rate": 7.344323855769265e-05, "loss": 0.0224, "step": 113230 }, { "grad_norm": 0.38003310561180115, "learning_rate": 7.343837021452846e-05, "loss": 0.0245, "step": 113240 }, { "grad_norm": 0.1613585352897644, "learning_rate": 7.343350158656596e-05, "loss": 0.0219, "step": 113250 }, { "grad_norm": 0.14908964931964874, "learning_rate": 7.342863267386428e-05, "loss": 0.0151, "step": 113260 }, { "grad_norm": 0.17104530334472656, "learning_rate": 7.342376347648259e-05, "loss": 0.0204, "step": 113270 }, { "grad_norm": 0.28817734122276306, "learning_rate": 7.341889399448006e-05, "loss": 0.0161, "step": 113280 }, { "grad_norm": 0.11434182524681091, "learning_rate": 7.341402422791586e-05, "loss": 0.0156, "step": 113290 }, { "grad_norm": 0.2878049910068512, "learning_rate": 7.340915417684915e-05, "loss": 0.0283, "step": 113300 }, { "grad_norm": 0.12995591759681702, "learning_rate": 7.340428384133912e-05, "loss": 0.0169, "step": 113310 }, { "grad_norm": 0.08014947921037674, "learning_rate": 7.339941322144494e-05, "loss": 0.0154, "step": 113320 }, { "grad_norm": 0.08027800917625427, "learning_rate": 7.339454231722582e-05, "loss": 0.0152, "step": 113330 }, { "grad_norm": 0.07613403350114822, "learning_rate": 7.338967112874089e-05, "loss": 0.0191, "step": 113340 }, { "grad_norm": 0.17466047406196594, "learning_rate": 7.33847996560494e-05, "loss": 0.0156, "step": 113350 }, { "grad_norm": 0.1313277631998062, "learning_rate": 7.337992789921049e-05, "loss": 0.0164, "step": 113360 }, { "grad_norm": 0.12677648663520813, "learning_rate": 7.33750558582834e-05, "loss": 0.0224, "step": 113370 }, { "grad_norm": 0.09653767198324203, "learning_rate": 7.33701835333273e-05, "loss": 0.0136, "step": 113380 }, { "grad_norm": 0.09992081671953201, "learning_rate": 7.336531092440142e-05, "loss": 0.0146, "step": 113390 }, { "grad_norm": 0.128747820854187, "learning_rate": 7.336043803156493e-05, "loss": 0.0172, "step": 113400 }, { "grad_norm": 0.14188018441200256, "learning_rate": 7.335556485487708e-05, "loss": 0.0176, "step": 113410 }, { "grad_norm": 0.20546182990074158, "learning_rate": 7.335069139439706e-05, "loss": 0.0158, "step": 113420 }, { "grad_norm": 0.36554425954818726, "learning_rate": 7.334581765018409e-05, "loss": 0.0209, "step": 113430 }, { "grad_norm": 0.14170925319194794, "learning_rate": 7.33409436222974e-05, "loss": 0.0153, "step": 113440 }, { "grad_norm": 0.1552387773990631, "learning_rate": 7.33360693107962e-05, "loss": 0.0158, "step": 113450 }, { "grad_norm": 0.1450924426317215, "learning_rate": 7.333119471573974e-05, "loss": 0.0185, "step": 113460 }, { "grad_norm": 0.11120006442070007, "learning_rate": 7.332631983718722e-05, "loss": 0.0142, "step": 113470 }, { "grad_norm": 0.12921883165836334, "learning_rate": 7.33214446751979e-05, "loss": 0.0188, "step": 113480 }, { "grad_norm": 0.13427166640758514, "learning_rate": 7.3316569229831e-05, "loss": 0.0213, "step": 113490 }, { "grad_norm": 0.13318558037281036, "learning_rate": 7.331169350114577e-05, "loss": 0.0194, "step": 113500 }, { "grad_norm": 0.3401375114917755, "learning_rate": 7.330681748920146e-05, "loss": 0.0234, "step": 113510 }, { "grad_norm": 0.19508101046085358, "learning_rate": 7.330194119405731e-05, "loss": 0.0163, "step": 113520 }, { "grad_norm": 0.14255137741565704, "learning_rate": 7.329706461577257e-05, "loss": 0.0174, "step": 113530 }, { "grad_norm": 0.21225926280021667, "learning_rate": 7.32921877544065e-05, "loss": 0.019, "step": 113540 }, { "grad_norm": 0.09260585904121399, "learning_rate": 7.328731061001837e-05, "loss": 0.0153, "step": 113550 }, { "grad_norm": 0.16546925902366638, "learning_rate": 7.328243318266741e-05, "loss": 0.0158, "step": 113560 }, { "grad_norm": 0.12960408627986908, "learning_rate": 7.327755547241291e-05, "loss": 0.0196, "step": 113570 }, { "grad_norm": 0.1416008621454239, "learning_rate": 7.327267747931414e-05, "loss": 0.0179, "step": 113580 }, { "grad_norm": 0.13662128150463104, "learning_rate": 7.326779920343035e-05, "loss": 0.02, "step": 113590 }, { "grad_norm": 0.14332698285579681, "learning_rate": 7.326292064482085e-05, "loss": 0.0179, "step": 113600 }, { "grad_norm": 0.1565283238887787, "learning_rate": 7.32580418035449e-05, "loss": 0.0185, "step": 113610 }, { "grad_norm": 0.14583159983158112, "learning_rate": 7.325316267966177e-05, "loss": 0.017, "step": 113620 }, { "grad_norm": 0.1851668804883957, "learning_rate": 7.324828327323076e-05, "loss": 0.0209, "step": 113630 }, { "grad_norm": 0.15765193104743958, "learning_rate": 7.324340358431117e-05, "loss": 0.0155, "step": 113640 }, { "grad_norm": 0.328311949968338, "learning_rate": 7.323852361296227e-05, "loss": 0.0169, "step": 113650 }, { "grad_norm": 1.3736125230789185, "learning_rate": 7.323364335924337e-05, "loss": 0.026, "step": 113660 }, { "grad_norm": 0.14413221180438995, "learning_rate": 7.322876282321379e-05, "loss": 0.019, "step": 113670 }, { "grad_norm": 0.14449156820774078, "learning_rate": 7.322388200493277e-05, "loss": 0.0147, "step": 113680 }, { "grad_norm": 0.16176360845565796, "learning_rate": 7.321900090445967e-05, "loss": 0.0222, "step": 113690 }, { "grad_norm": 0.10098208487033844, "learning_rate": 7.321411952185379e-05, "loss": 0.0118, "step": 113700 }, { "grad_norm": 0.3449155390262604, "learning_rate": 7.320923785717445e-05, "loss": 0.0181, "step": 113710 }, { "grad_norm": 0.09277437627315521, "learning_rate": 7.320435591048094e-05, "loss": 0.0197, "step": 113720 }, { "grad_norm": 0.3965664505958557, "learning_rate": 7.319947368183262e-05, "loss": 0.022, "step": 113730 }, { "grad_norm": 0.19491945207118988, "learning_rate": 7.319459117128877e-05, "loss": 0.0159, "step": 113740 }, { "grad_norm": 0.1592499017715454, "learning_rate": 7.318970837890875e-05, "loss": 0.0236, "step": 113750 }, { "grad_norm": 0.1777333766222, "learning_rate": 7.318482530475186e-05, "loss": 0.0215, "step": 113760 }, { "grad_norm": 0.2477274090051651, "learning_rate": 7.317994194887747e-05, "loss": 0.0203, "step": 113770 }, { "grad_norm": 0.1141253411769867, "learning_rate": 7.317505831134489e-05, "loss": 0.0158, "step": 113780 }, { "grad_norm": 0.09525101631879807, "learning_rate": 7.317017439221347e-05, "loss": 0.0183, "step": 113790 }, { "grad_norm": 0.20345531404018402, "learning_rate": 7.316529019154257e-05, "loss": 0.0197, "step": 113800 }, { "grad_norm": 0.09725865721702576, "learning_rate": 7.316040570939151e-05, "loss": 0.013, "step": 113810 }, { "grad_norm": 0.11468294262886047, "learning_rate": 7.315552094581966e-05, "loss": 0.016, "step": 113820 }, { "grad_norm": 0.09856114536523819, "learning_rate": 7.315063590088637e-05, "loss": 0.0124, "step": 113830 }, { "grad_norm": 0.08163686096668243, "learning_rate": 7.314575057465098e-05, "loss": 0.0211, "step": 113840 }, { "grad_norm": 0.375539630651474, "learning_rate": 7.314086496717289e-05, "loss": 0.0238, "step": 113850 }, { "grad_norm": 0.17051784694194794, "learning_rate": 7.313597907851142e-05, "loss": 0.0158, "step": 113860 }, { "grad_norm": 0.08539558202028275, "learning_rate": 7.313109290872598e-05, "loss": 0.015, "step": 113870 }, { "grad_norm": 0.2049523890018463, "learning_rate": 7.31262064578759e-05, "loss": 0.0158, "step": 113880 }, { "grad_norm": 0.1294829547405243, "learning_rate": 7.312131972602058e-05, "loss": 0.0143, "step": 113890 }, { "grad_norm": 0.2087751179933548, "learning_rate": 7.311643271321942e-05, "loss": 0.0172, "step": 113900 }, { "grad_norm": 0.12071582674980164, "learning_rate": 7.311154541953176e-05, "loss": 0.0183, "step": 113910 }, { "grad_norm": 0.15353235602378845, "learning_rate": 7.3106657845017e-05, "loss": 0.0204, "step": 113920 }, { "grad_norm": 0.1645790934562683, "learning_rate": 7.310176998973453e-05, "loss": 0.0173, "step": 113930 }, { "grad_norm": 0.14355744421482086, "learning_rate": 7.309688185374376e-05, "loss": 0.0124, "step": 113940 }, { "grad_norm": 0.11345502734184265, "learning_rate": 7.309199343710405e-05, "loss": 0.0178, "step": 113950 }, { "grad_norm": 0.16325294971466064, "learning_rate": 7.308710473987483e-05, "loss": 0.024, "step": 113960 }, { "grad_norm": 0.13642095029354095, "learning_rate": 7.308221576211548e-05, "loss": 0.0163, "step": 113970 }, { "grad_norm": 0.29913461208343506, "learning_rate": 7.307732650388542e-05, "loss": 0.0176, "step": 113980 }, { "grad_norm": 0.15908144414424896, "learning_rate": 7.307243696524405e-05, "loss": 0.0193, "step": 113990 }, { "grad_norm": 0.45315325260162354, "learning_rate": 7.306754714625078e-05, "loss": 0.0167, "step": 114000 }, { "grad_norm": 0.18781761825084686, "learning_rate": 7.306265704696504e-05, "loss": 0.0217, "step": 114010 }, { "grad_norm": 0.09775309264659882, "learning_rate": 7.305776666744625e-05, "loss": 0.0186, "step": 114020 }, { "grad_norm": 0.25799834728240967, "learning_rate": 7.305287600775382e-05, "loss": 0.019, "step": 114030 }, { "grad_norm": 0.17212048172950745, "learning_rate": 7.304798506794718e-05, "loss": 0.015, "step": 114040 }, { "grad_norm": 0.35747650265693665, "learning_rate": 7.304309384808577e-05, "loss": 0.0213, "step": 114050 }, { "grad_norm": 0.1353636533021927, "learning_rate": 7.303820234822901e-05, "loss": 0.0194, "step": 114060 }, { "grad_norm": 0.5275862812995911, "learning_rate": 7.303331056843631e-05, "loss": 0.0166, "step": 114070 }, { "grad_norm": 0.11090400069952011, "learning_rate": 7.302841850876718e-05, "loss": 0.0158, "step": 114080 }, { "grad_norm": 0.15374095737934113, "learning_rate": 7.302352616928099e-05, "loss": 0.0234, "step": 114090 }, { "grad_norm": 0.19834305346012115, "learning_rate": 7.301863355003725e-05, "loss": 0.0141, "step": 114100 }, { "grad_norm": 0.08101361989974976, "learning_rate": 7.301374065109535e-05, "loss": 0.0154, "step": 114110 }, { "grad_norm": 0.10369081050157547, "learning_rate": 7.30088474725148e-05, "loss": 0.0138, "step": 114120 }, { "grad_norm": 0.13206638395786285, "learning_rate": 7.300395401435501e-05, "loss": 0.019, "step": 114130 }, { "grad_norm": 0.1695287674665451, "learning_rate": 7.299906027667547e-05, "loss": 0.0193, "step": 114140 }, { "grad_norm": 0.15806083381175995, "learning_rate": 7.299416625953564e-05, "loss": 0.0167, "step": 114150 }, { "grad_norm": 0.1284857541322708, "learning_rate": 7.298927196299496e-05, "loss": 0.0189, "step": 114160 }, { "grad_norm": 0.38486340641975403, "learning_rate": 7.298437738711292e-05, "loss": 0.021, "step": 114170 }, { "grad_norm": 0.14347527921199799, "learning_rate": 7.297948253194899e-05, "loss": 0.0151, "step": 114180 }, { "grad_norm": 0.12499044090509415, "learning_rate": 7.297458739756267e-05, "loss": 0.0197, "step": 114190 }, { "grad_norm": 0.15311425924301147, "learning_rate": 7.296969198401342e-05, "loss": 0.0159, "step": 114200 }, { "grad_norm": 0.11215373128652573, "learning_rate": 7.296479629136073e-05, "loss": 0.016, "step": 114210 }, { "grad_norm": 0.14327701926231384, "learning_rate": 7.295990031966407e-05, "loss": 0.0166, "step": 114220 }, { "grad_norm": 0.21368475258350372, "learning_rate": 7.295500406898295e-05, "loss": 0.0185, "step": 114230 }, { "grad_norm": 0.1782139241695404, "learning_rate": 7.295010753937686e-05, "loss": 0.0194, "step": 114240 }, { "grad_norm": 0.1468857377767563, "learning_rate": 7.294521073090527e-05, "loss": 0.0187, "step": 114250 }, { "grad_norm": 0.10098199546337128, "learning_rate": 7.294031364362773e-05, "loss": 0.0163, "step": 114260 }, { "grad_norm": 0.09534882009029388, "learning_rate": 7.29354162776037e-05, "loss": 0.0131, "step": 114270 }, { "grad_norm": 0.11065957695245743, "learning_rate": 7.293051863289274e-05, "loss": 0.0206, "step": 114280 }, { "grad_norm": 0.20084647834300995, "learning_rate": 7.29256207095543e-05, "loss": 0.0195, "step": 114290 }, { "grad_norm": 0.2156662493944168, "learning_rate": 7.292072250764795e-05, "loss": 0.0216, "step": 114300 }, { "grad_norm": 0.10638631135225296, "learning_rate": 7.291582402723316e-05, "loss": 0.0152, "step": 114310 }, { "grad_norm": 0.11368037760257721, "learning_rate": 7.291092526836947e-05, "loss": 0.0163, "step": 114320 }, { "grad_norm": 0.12049780786037445, "learning_rate": 7.290602623111642e-05, "loss": 0.0123, "step": 114330 }, { "grad_norm": 0.14221683144569397, "learning_rate": 7.290112691553351e-05, "loss": 0.0171, "step": 114340 }, { "grad_norm": 0.12274623662233353, "learning_rate": 7.28962273216803e-05, "loss": 0.0165, "step": 114350 }, { "grad_norm": 0.11045438051223755, "learning_rate": 7.28913274496163e-05, "loss": 0.0186, "step": 114360 }, { "grad_norm": 0.15149365365505219, "learning_rate": 7.288642729940106e-05, "loss": 0.0157, "step": 114370 }, { "grad_norm": 0.15343402326107025, "learning_rate": 7.288152687109413e-05, "loss": 0.0222, "step": 114380 }, { "grad_norm": 0.2254808247089386, "learning_rate": 7.287662616475504e-05, "loss": 0.0168, "step": 114390 }, { "grad_norm": 0.13677307963371277, "learning_rate": 7.287172518044334e-05, "loss": 0.023, "step": 114400 }, { "grad_norm": 0.13605688512325287, "learning_rate": 7.286682391821858e-05, "loss": 0.0181, "step": 114410 }, { "grad_norm": 0.11106447130441666, "learning_rate": 7.286192237814033e-05, "loss": 0.0194, "step": 114420 }, { "grad_norm": 0.23249302804470062, "learning_rate": 7.285702056026813e-05, "loss": 0.0178, "step": 114430 }, { "grad_norm": 0.1345415860414505, "learning_rate": 7.285211846466154e-05, "loss": 0.0184, "step": 114440 }, { "grad_norm": 0.1768706738948822, "learning_rate": 7.284721609138014e-05, "loss": 0.0319, "step": 114450 }, { "grad_norm": 0.2931731641292572, "learning_rate": 7.284231344048351e-05, "loss": 0.0196, "step": 114460 }, { "grad_norm": 0.11484426259994507, "learning_rate": 7.28374105120312e-05, "loss": 0.0197, "step": 114470 }, { "grad_norm": 0.10407652705907822, "learning_rate": 7.283250730608277e-05, "loss": 0.0196, "step": 114480 }, { "grad_norm": 0.44323012232780457, "learning_rate": 7.282760382269783e-05, "loss": 0.0177, "step": 114490 }, { "grad_norm": 0.10672406107187271, "learning_rate": 7.282270006193595e-05, "loss": 0.0192, "step": 114500 }, { "grad_norm": 0.30496180057525635, "learning_rate": 7.281779602385673e-05, "loss": 0.024, "step": 114510 }, { "grad_norm": 0.11610753834247589, "learning_rate": 7.281289170851972e-05, "loss": 0.0146, "step": 114520 }, { "grad_norm": 0.10606703162193298, "learning_rate": 7.280798711598456e-05, "loss": 0.0241, "step": 114530 }, { "grad_norm": 0.13359564542770386, "learning_rate": 7.280308224631081e-05, "loss": 0.0199, "step": 114540 }, { "grad_norm": 0.137953519821167, "learning_rate": 7.279817709955809e-05, "loss": 0.0192, "step": 114550 }, { "grad_norm": 0.10175827890634537, "learning_rate": 7.279327167578599e-05, "loss": 0.0204, "step": 114560 }, { "grad_norm": 0.15635615587234497, "learning_rate": 7.27883659750541e-05, "loss": 0.0183, "step": 114570 }, { "grad_norm": 0.11658396571874619, "learning_rate": 7.278345999742208e-05, "loss": 0.0164, "step": 114580 }, { "grad_norm": 0.22569338977336884, "learning_rate": 7.277855374294948e-05, "loss": 0.0166, "step": 114590 }, { "grad_norm": 0.2599743902683258, "learning_rate": 7.277364721169597e-05, "loss": 0.0127, "step": 114600 }, { "grad_norm": 0.22007250785827637, "learning_rate": 7.276874040372113e-05, "loss": 0.016, "step": 114610 }, { "grad_norm": 0.08492617309093475, "learning_rate": 7.27638333190846e-05, "loss": 0.0148, "step": 114620 }, { "grad_norm": 0.15408658981323242, "learning_rate": 7.275892595784601e-05, "loss": 0.0134, "step": 114630 }, { "grad_norm": 0.1520669311285019, "learning_rate": 7.275401832006496e-05, "loss": 0.0156, "step": 114640 }, { "grad_norm": 0.12750831246376038, "learning_rate": 7.274911040580114e-05, "loss": 0.0162, "step": 114650 }, { "grad_norm": 0.10875941812992096, "learning_rate": 7.274420221511411e-05, "loss": 0.0164, "step": 114660 }, { "grad_norm": 0.13365137577056885, "learning_rate": 7.273929374806358e-05, "loss": 0.0164, "step": 114670 }, { "grad_norm": 0.20953518152236938, "learning_rate": 7.273438500470914e-05, "loss": 0.0171, "step": 114680 }, { "grad_norm": 0.08980228751897812, "learning_rate": 7.272947598511047e-05, "loss": 0.0098, "step": 114690 }, { "grad_norm": 0.10090334713459015, "learning_rate": 7.272456668932721e-05, "loss": 0.0188, "step": 114700 }, { "grad_norm": 0.15718668699264526, "learning_rate": 7.2719657117419e-05, "loss": 0.0152, "step": 114710 }, { "grad_norm": 0.14650432765483856, "learning_rate": 7.271474726944551e-05, "loss": 0.0154, "step": 114720 }, { "grad_norm": 0.11487013101577759, "learning_rate": 7.270983714546639e-05, "loss": 0.0165, "step": 114730 }, { "grad_norm": 0.09293618053197861, "learning_rate": 7.270492674554131e-05, "loss": 0.0142, "step": 114740 }, { "grad_norm": 0.08843095600605011, "learning_rate": 7.270001606972993e-05, "loss": 0.0139, "step": 114750 }, { "grad_norm": 0.08318822085857391, "learning_rate": 7.269510511809192e-05, "loss": 0.0149, "step": 114760 }, { "grad_norm": 0.13613979518413544, "learning_rate": 7.269019389068696e-05, "loss": 0.0119, "step": 114770 }, { "grad_norm": 0.19498206675052643, "learning_rate": 7.268528238757472e-05, "loss": 0.0182, "step": 114780 }, { "grad_norm": 0.08661253750324249, "learning_rate": 7.268037060881489e-05, "loss": 0.0123, "step": 114790 }, { "grad_norm": 0.07868605852127075, "learning_rate": 7.267545855446713e-05, "loss": 0.0139, "step": 114800 }, { "grad_norm": 0.07229478657245636, "learning_rate": 7.267054622459115e-05, "loss": 0.0151, "step": 114810 }, { "grad_norm": 0.14301499724388123, "learning_rate": 7.266563361924663e-05, "loss": 0.0183, "step": 114820 }, { "grad_norm": 0.09944784641265869, "learning_rate": 7.266072073849325e-05, "loss": 0.0101, "step": 114830 }, { "grad_norm": 0.13798698782920837, "learning_rate": 7.26558075823907e-05, "loss": 0.0184, "step": 114840 }, { "grad_norm": 0.13216082751750946, "learning_rate": 7.265089415099874e-05, "loss": 0.015, "step": 114850 }, { "grad_norm": 0.10228993743658066, "learning_rate": 7.2645980444377e-05, "loss": 0.0196, "step": 114860 }, { "grad_norm": 0.14304809272289276, "learning_rate": 7.264106646258523e-05, "loss": 0.0172, "step": 114870 }, { "grad_norm": 0.18658307194709778, "learning_rate": 7.263615220568313e-05, "loss": 0.0197, "step": 114880 }, { "grad_norm": 0.08850408345460892, "learning_rate": 7.263123767373039e-05, "loss": 0.0111, "step": 114890 }, { "grad_norm": 0.10343757271766663, "learning_rate": 7.262632286678675e-05, "loss": 0.0162, "step": 114900 }, { "grad_norm": 0.5761141777038574, "learning_rate": 7.262140778491194e-05, "loss": 0.0133, "step": 114910 }, { "grad_norm": 0.19815954566001892, "learning_rate": 7.261649242816566e-05, "loss": 0.0146, "step": 114920 }, { "grad_norm": 0.08719426393508911, "learning_rate": 7.261157679660763e-05, "loss": 0.0175, "step": 114930 }, { "grad_norm": 0.152347132563591, "learning_rate": 7.260666089029761e-05, "loss": 0.0134, "step": 114940 }, { "grad_norm": 0.23798716068267822, "learning_rate": 7.260174470929531e-05, "loss": 0.0139, "step": 114950 }, { "grad_norm": 0.12287045270204544, "learning_rate": 7.259682825366046e-05, "loss": 0.0165, "step": 114960 }, { "grad_norm": 0.07720214128494263, "learning_rate": 7.259191152345284e-05, "loss": 0.0131, "step": 114970 }, { "grad_norm": 0.09084118902683258, "learning_rate": 7.258699451873214e-05, "loss": 0.0126, "step": 114980 }, { "grad_norm": 0.06675361096858978, "learning_rate": 7.258207723955814e-05, "loss": 0.0138, "step": 114990 }, { "grad_norm": 0.10837846994400024, "learning_rate": 7.257715968599057e-05, "loss": 0.0185, "step": 115000 }, { "grad_norm": 0.23004192113876343, "learning_rate": 7.25722418580892e-05, "loss": 0.0191, "step": 115010 }, { "grad_norm": 0.07689334452152252, "learning_rate": 7.25673237559138e-05, "loss": 0.0226, "step": 115020 }, { "grad_norm": 0.18524619936943054, "learning_rate": 7.256240537952408e-05, "loss": 0.0182, "step": 115030 }, { "grad_norm": 0.1037968099117279, "learning_rate": 7.255748672897985e-05, "loss": 0.0157, "step": 115040 }, { "grad_norm": 0.1995110958814621, "learning_rate": 7.255256780434085e-05, "loss": 0.0223, "step": 115050 }, { "grad_norm": 0.4425911009311676, "learning_rate": 7.254764860566686e-05, "loss": 0.0142, "step": 115060 }, { "grad_norm": 0.10795250535011292, "learning_rate": 7.254272913301765e-05, "loss": 0.0185, "step": 115070 }, { "grad_norm": 0.0996507927775383, "learning_rate": 7.2537809386453e-05, "loss": 0.0174, "step": 115080 }, { "grad_norm": 0.1405395269393921, "learning_rate": 7.253288936603269e-05, "loss": 0.0203, "step": 115090 }, { "grad_norm": 0.136460542678833, "learning_rate": 7.252796907181649e-05, "loss": 0.0184, "step": 115100 }, { "grad_norm": 0.1615627557039261, "learning_rate": 7.25230485038642e-05, "loss": 0.0138, "step": 115110 }, { "grad_norm": 0.13553903996944427, "learning_rate": 7.25181276622356e-05, "loss": 0.0142, "step": 115120 }, { "grad_norm": 0.12715256214141846, "learning_rate": 7.25132065469905e-05, "loss": 0.0132, "step": 115130 }, { "grad_norm": 0.08327940106391907, "learning_rate": 7.250828515818868e-05, "loss": 0.0143, "step": 115140 }, { "grad_norm": 0.19561627507209778, "learning_rate": 7.250336349588994e-05, "loss": 0.0117, "step": 115150 }, { "grad_norm": 0.12282148003578186, "learning_rate": 7.249844156015408e-05, "loss": 0.0168, "step": 115160 }, { "grad_norm": 0.10220607370138168, "learning_rate": 7.249351935104094e-05, "loss": 0.0197, "step": 115170 }, { "grad_norm": 0.13873052597045898, "learning_rate": 7.248859686861027e-05, "loss": 0.0139, "step": 115180 }, { "grad_norm": 0.18004447221755981, "learning_rate": 7.248367411292194e-05, "loss": 0.0228, "step": 115190 }, { "grad_norm": 0.07956725358963013, "learning_rate": 7.247875108403572e-05, "loss": 0.016, "step": 115200 }, { "grad_norm": 0.2617070972919464, "learning_rate": 7.247382778201147e-05, "loss": 0.0174, "step": 115210 }, { "grad_norm": 0.07343779504299164, "learning_rate": 7.246890420690898e-05, "loss": 0.0116, "step": 115220 }, { "grad_norm": 0.07000383734703064, "learning_rate": 7.246398035878811e-05, "loss": 0.0159, "step": 115230 }, { "grad_norm": 0.16089986264705658, "learning_rate": 7.245905623770864e-05, "loss": 0.0119, "step": 115240 }, { "grad_norm": 0.2827131450176239, "learning_rate": 7.245413184373045e-05, "loss": 0.0141, "step": 115250 }, { "grad_norm": 0.1660739630460739, "learning_rate": 7.244920717691336e-05, "loss": 0.0126, "step": 115260 }, { "grad_norm": 0.2177361398935318, "learning_rate": 7.24442822373172e-05, "loss": 0.0165, "step": 115270 }, { "grad_norm": 0.08855637162923813, "learning_rate": 7.243935702500182e-05, "loss": 0.0169, "step": 115280 }, { "grad_norm": 0.16610054671764374, "learning_rate": 7.243443154002707e-05, "loss": 0.0115, "step": 115290 }, { "grad_norm": 0.1255740225315094, "learning_rate": 7.242950578245279e-05, "loss": 0.0181, "step": 115300 }, { "grad_norm": 0.10440448671579361, "learning_rate": 7.242457975233885e-05, "loss": 0.017, "step": 115310 }, { "grad_norm": 0.08223900198936462, "learning_rate": 7.241965344974508e-05, "loss": 0.0159, "step": 115320 }, { "grad_norm": 0.09953030943870544, "learning_rate": 7.241472687473136e-05, "loss": 0.012, "step": 115330 }, { "grad_norm": 0.11322721093893051, "learning_rate": 7.240980002735753e-05, "loss": 0.0151, "step": 115340 }, { "grad_norm": 0.2312570959329605, "learning_rate": 7.24048729076835e-05, "loss": 0.0182, "step": 115350 }, { "grad_norm": 0.1396428644657135, "learning_rate": 7.23999455157691e-05, "loss": 0.019, "step": 115360 }, { "grad_norm": 0.11395929753780365, "learning_rate": 7.23950178516742e-05, "loss": 0.0132, "step": 115370 }, { "grad_norm": 0.10337395966053009, "learning_rate": 7.23900899154587e-05, "loss": 0.0212, "step": 115380 }, { "grad_norm": 0.18086779117584229, "learning_rate": 7.238516170718245e-05, "loss": 0.0154, "step": 115390 }, { "grad_norm": 0.4563775956630707, "learning_rate": 7.238023322690537e-05, "loss": 0.0137, "step": 115400 }, { "grad_norm": 0.09539207071065903, "learning_rate": 7.237530447468732e-05, "loss": 0.0135, "step": 115410 }, { "grad_norm": 0.1280994415283203, "learning_rate": 7.237037545058821e-05, "loss": 0.017, "step": 115420 }, { "grad_norm": 0.07174121588468552, "learning_rate": 7.236544615466789e-05, "loss": 0.017, "step": 115430 }, { "grad_norm": 0.11838575452566147, "learning_rate": 7.236051658698629e-05, "loss": 0.0147, "step": 115440 }, { "grad_norm": 0.19978244602680206, "learning_rate": 7.235558674760331e-05, "loss": 0.0207, "step": 115450 }, { "grad_norm": 0.16380982100963593, "learning_rate": 7.235065663657883e-05, "loss": 0.0145, "step": 115460 }, { "grad_norm": 0.1506451666355133, "learning_rate": 7.234572625397279e-05, "loss": 0.0226, "step": 115470 }, { "grad_norm": 0.174968883395195, "learning_rate": 7.234079559984505e-05, "loss": 0.0156, "step": 115480 }, { "grad_norm": 0.12395468354225159, "learning_rate": 7.233586467425558e-05, "loss": 0.0153, "step": 115490 }, { "grad_norm": 0.1821771264076233, "learning_rate": 7.233093347726426e-05, "loss": 0.0154, "step": 115500 }, { "grad_norm": 0.5159241557121277, "learning_rate": 7.232600200893101e-05, "loss": 0.0208, "step": 115510 }, { "grad_norm": 0.1332574337720871, "learning_rate": 7.232107026931576e-05, "loss": 0.0141, "step": 115520 }, { "grad_norm": 0.11318811774253845, "learning_rate": 7.231613825847841e-05, "loss": 0.0159, "step": 115530 }, { "grad_norm": 0.18039646744728088, "learning_rate": 7.231120597647893e-05, "loss": 0.0195, "step": 115540 }, { "grad_norm": 0.1902354210615158, "learning_rate": 7.230627342337724e-05, "loss": 0.0146, "step": 115550 }, { "grad_norm": 0.15075404942035675, "learning_rate": 7.230134059923325e-05, "loss": 0.0164, "step": 115560 }, { "grad_norm": 0.23674871027469635, "learning_rate": 7.229640750410692e-05, "loss": 0.0182, "step": 115570 }, { "grad_norm": 0.14511924982070923, "learning_rate": 7.229147413805819e-05, "loss": 0.0175, "step": 115580 }, { "grad_norm": 0.1098276823759079, "learning_rate": 7.2286540501147e-05, "loss": 0.0135, "step": 115590 }, { "grad_norm": 0.09451724588871002, "learning_rate": 7.22816065934333e-05, "loss": 0.0173, "step": 115600 }, { "grad_norm": 0.07471822947263718, "learning_rate": 7.227667241497705e-05, "loss": 0.0108, "step": 115610 }, { "grad_norm": 0.2584015429019928, "learning_rate": 7.227173796583819e-05, "loss": 0.0169, "step": 115620 }, { "grad_norm": 0.2884345054626465, "learning_rate": 7.226680324607668e-05, "loss": 0.0203, "step": 115630 }, { "grad_norm": 0.10962551832199097, "learning_rate": 7.22618682557525e-05, "loss": 0.0145, "step": 115640 }, { "grad_norm": 0.12394752353429794, "learning_rate": 7.22569329949256e-05, "loss": 0.0218, "step": 115650 }, { "grad_norm": 0.12432648986577988, "learning_rate": 7.225199746365594e-05, "loss": 0.0173, "step": 115660 }, { "grad_norm": 0.14577825367450714, "learning_rate": 7.224706166200352e-05, "loss": 0.0196, "step": 115670 }, { "grad_norm": 0.16389013826847076, "learning_rate": 7.224212559002829e-05, "loss": 0.0209, "step": 115680 }, { "grad_norm": 0.6422470211982727, "learning_rate": 7.223718924779023e-05, "loss": 0.0156, "step": 115690 }, { "grad_norm": 0.21695342659950256, "learning_rate": 7.223225263534932e-05, "loss": 0.0226, "step": 115700 }, { "grad_norm": 0.16692358255386353, "learning_rate": 7.222731575276555e-05, "loss": 0.0149, "step": 115710 }, { "grad_norm": 0.18039163947105408, "learning_rate": 7.222237860009892e-05, "loss": 0.0199, "step": 115720 }, { "grad_norm": 0.17350263893604279, "learning_rate": 7.22174411774094e-05, "loss": 0.0125, "step": 115730 }, { "grad_norm": 0.07796238362789154, "learning_rate": 7.221250348475698e-05, "loss": 0.0145, "step": 115740 }, { "grad_norm": 0.11908775568008423, "learning_rate": 7.22075655222017e-05, "loss": 0.0166, "step": 115750 }, { "grad_norm": 0.0853220522403717, "learning_rate": 7.22026272898035e-05, "loss": 0.0148, "step": 115760 }, { "grad_norm": 0.19553017616271973, "learning_rate": 7.219768878762244e-05, "loss": 0.0186, "step": 115770 }, { "grad_norm": 0.10350681841373444, "learning_rate": 7.219275001571849e-05, "loss": 0.0146, "step": 115780 }, { "grad_norm": 0.14604607224464417, "learning_rate": 7.218781097415168e-05, "loss": 0.0209, "step": 115790 }, { "grad_norm": 0.12323441356420517, "learning_rate": 7.218287166298201e-05, "loss": 0.0182, "step": 115800 }, { "grad_norm": 0.07064726203680038, "learning_rate": 7.217793208226952e-05, "loss": 0.0138, "step": 115810 }, { "grad_norm": 0.15715277194976807, "learning_rate": 7.21729922320742e-05, "loss": 0.017, "step": 115820 }, { "grad_norm": 0.10870640724897385, "learning_rate": 7.216805211245609e-05, "loss": 0.0147, "step": 115830 }, { "grad_norm": 0.10561583191156387, "learning_rate": 7.216311172347525e-05, "loss": 0.018, "step": 115840 }, { "grad_norm": 0.17974965274333954, "learning_rate": 7.215817106519164e-05, "loss": 0.0167, "step": 115850 }, { "grad_norm": 0.07154151052236557, "learning_rate": 7.215323013766533e-05, "loss": 0.016, "step": 115860 }, { "grad_norm": 0.28569430112838745, "learning_rate": 7.214828894095638e-05, "loss": 0.0284, "step": 115870 }, { "grad_norm": 0.16628773510456085, "learning_rate": 7.21433474751248e-05, "loss": 0.0187, "step": 115880 }, { "grad_norm": 0.1455024778842926, "learning_rate": 7.213840574023063e-05, "loss": 0.0166, "step": 115890 }, { "grad_norm": 0.07966883480548859, "learning_rate": 7.213346373633393e-05, "loss": 0.0158, "step": 115900 }, { "grad_norm": 0.15522097051143646, "learning_rate": 7.212852146349475e-05, "loss": 0.022, "step": 115910 }, { "grad_norm": 0.14064869284629822, "learning_rate": 7.212357892177315e-05, "loss": 0.0138, "step": 115920 }, { "grad_norm": 0.20775361359119415, "learning_rate": 7.211863611122918e-05, "loss": 0.0184, "step": 115930 }, { "grad_norm": 0.12238532304763794, "learning_rate": 7.211369303192288e-05, "loss": 0.0142, "step": 115940 }, { "grad_norm": 0.2698793411254883, "learning_rate": 7.210874968391435e-05, "loss": 0.021, "step": 115950 }, { "grad_norm": 0.14530228078365326, "learning_rate": 7.210380606726362e-05, "loss": 0.0166, "step": 115960 }, { "grad_norm": 0.10417116433382034, "learning_rate": 7.20988621820308e-05, "loss": 0.0192, "step": 115970 }, { "grad_norm": 0.1452053338289261, "learning_rate": 7.209391802827592e-05, "loss": 0.0216, "step": 115980 }, { "grad_norm": 0.18723328411579132, "learning_rate": 7.208897360605909e-05, "loss": 0.0209, "step": 115990 }, { "grad_norm": 0.11595012247562408, "learning_rate": 7.208402891544038e-05, "loss": 0.0172, "step": 116000 }, { "grad_norm": 0.12508709728717804, "learning_rate": 7.207908395647983e-05, "loss": 0.0209, "step": 116010 }, { "grad_norm": 0.21848775446414948, "learning_rate": 7.20741387292376e-05, "loss": 0.0171, "step": 116020 }, { "grad_norm": 0.07811488211154938, "learning_rate": 7.206919323377373e-05, "loss": 0.0159, "step": 116030 }, { "grad_norm": 0.10713225603103638, "learning_rate": 7.206424747014833e-05, "loss": 0.0159, "step": 116040 }, { "grad_norm": 0.1631222516298294, "learning_rate": 7.205930143842147e-05, "loss": 0.0175, "step": 116050 }, { "grad_norm": 0.14922796189785004, "learning_rate": 7.20543551386533e-05, "loss": 0.0153, "step": 116060 }, { "grad_norm": 0.09739647060632706, "learning_rate": 7.204940857090389e-05, "loss": 0.0182, "step": 116070 }, { "grad_norm": 0.18201254308223724, "learning_rate": 7.204446173523333e-05, "loss": 0.0164, "step": 116080 }, { "grad_norm": 0.10634583234786987, "learning_rate": 7.203951463170176e-05, "loss": 0.0136, "step": 116090 }, { "grad_norm": 0.06821738183498383, "learning_rate": 7.203456726036927e-05, "loss": 0.0121, "step": 116100 }, { "grad_norm": 0.1038001999258995, "learning_rate": 7.2029619621296e-05, "loss": 0.0143, "step": 116110 }, { "grad_norm": 0.08139454573392868, "learning_rate": 7.202467171454202e-05, "loss": 0.0174, "step": 116120 }, { "grad_norm": 0.11326460540294647, "learning_rate": 7.201972354016752e-05, "loss": 0.0165, "step": 116130 }, { "grad_norm": 0.11250697076320648, "learning_rate": 7.201477509823255e-05, "loss": 0.0155, "step": 116140 }, { "grad_norm": 0.08154527842998505, "learning_rate": 7.200982638879733e-05, "loss": 0.0236, "step": 116150 }, { "grad_norm": 0.1877274364233017, "learning_rate": 7.20048774119219e-05, "loss": 0.0184, "step": 116160 }, { "grad_norm": 0.16536065936088562, "learning_rate": 7.199992816766644e-05, "loss": 0.0212, "step": 116170 }, { "grad_norm": 0.117851622402668, "learning_rate": 7.199497865609108e-05, "loss": 0.0186, "step": 116180 }, { "grad_norm": 0.07904289662837982, "learning_rate": 7.199002887725597e-05, "loss": 0.0119, "step": 116190 }, { "grad_norm": 0.10263115912675858, "learning_rate": 7.198507883122124e-05, "loss": 0.0157, "step": 116200 }, { "grad_norm": 0.14021621644496918, "learning_rate": 7.198012851804705e-05, "loss": 0.0156, "step": 116210 }, { "grad_norm": 0.17597231268882751, "learning_rate": 7.197517793779354e-05, "loss": 0.0175, "step": 116220 }, { "grad_norm": 0.10112421959638596, "learning_rate": 7.197022709052088e-05, "loss": 0.0123, "step": 116230 }, { "grad_norm": 0.07766782492399216, "learning_rate": 7.19652759762892e-05, "loss": 0.0157, "step": 116240 }, { "grad_norm": 0.10997356474399567, "learning_rate": 7.19603245951587e-05, "loss": 0.0167, "step": 116250 }, { "grad_norm": 0.15016606450080872, "learning_rate": 7.19553729471895e-05, "loss": 0.0141, "step": 116260 }, { "grad_norm": 0.10586763173341751, "learning_rate": 7.195042103244181e-05, "loss": 0.0157, "step": 116270 }, { "grad_norm": 0.25433778762817383, "learning_rate": 7.194546885097576e-05, "loss": 0.0137, "step": 116280 }, { "grad_norm": 0.13898000121116638, "learning_rate": 7.194051640285157e-05, "loss": 0.0211, "step": 116290 }, { "grad_norm": 0.11174916476011276, "learning_rate": 7.193556368812936e-05, "loss": 0.0187, "step": 116300 }, { "grad_norm": 0.30627042055130005, "learning_rate": 7.193061070686936e-05, "loss": 0.0127, "step": 116310 }, { "grad_norm": 0.10040690004825592, "learning_rate": 7.192565745913173e-05, "loss": 0.0156, "step": 116320 }, { "grad_norm": 0.12525247037410736, "learning_rate": 7.192070394497667e-05, "loss": 0.0147, "step": 116330 }, { "grad_norm": 0.2219940423965454, "learning_rate": 7.191575016446435e-05, "loss": 0.0234, "step": 116340 }, { "grad_norm": 0.24987922608852386, "learning_rate": 7.191079611765496e-05, "loss": 0.0152, "step": 116350 }, { "grad_norm": 0.10193899273872375, "learning_rate": 7.190584180460874e-05, "loss": 0.0166, "step": 116360 }, { "grad_norm": 0.2011168897151947, "learning_rate": 7.190088722538582e-05, "loss": 0.0191, "step": 116370 }, { "grad_norm": 0.14042136073112488, "learning_rate": 7.189593238004648e-05, "loss": 0.0135, "step": 116380 }, { "grad_norm": 0.2227778136730194, "learning_rate": 7.189097726865088e-05, "loss": 0.0143, "step": 116390 }, { "grad_norm": 0.2877017557621002, "learning_rate": 7.188602189125923e-05, "loss": 0.0199, "step": 116400 }, { "grad_norm": 0.2808566391468048, "learning_rate": 7.188106624793175e-05, "loss": 0.0211, "step": 116410 }, { "grad_norm": 0.16177572309970856, "learning_rate": 7.187611033872865e-05, "loss": 0.0295, "step": 116420 }, { "grad_norm": 0.14794398844242096, "learning_rate": 7.187115416371018e-05, "loss": 0.0196, "step": 116430 }, { "grad_norm": 0.13963964581489563, "learning_rate": 7.18661977229365e-05, "loss": 0.0166, "step": 116440 }, { "grad_norm": 0.20901648700237274, "learning_rate": 7.18612410164679e-05, "loss": 0.0185, "step": 116450 }, { "grad_norm": 0.130106121301651, "learning_rate": 7.185628404436458e-05, "loss": 0.0237, "step": 116460 }, { "grad_norm": 0.23921911418437958, "learning_rate": 7.185132680668676e-05, "loss": 0.0193, "step": 116470 }, { "grad_norm": 0.1185656413435936, "learning_rate": 7.184636930349472e-05, "loss": 0.015, "step": 116480 }, { "grad_norm": 0.18583551049232483, "learning_rate": 7.184141153484862e-05, "loss": 0.0134, "step": 116490 }, { "grad_norm": 0.07099153846502304, "learning_rate": 7.183645350080877e-05, "loss": 0.0144, "step": 116500 }, { "grad_norm": 0.24399547278881073, "learning_rate": 7.18314952014354e-05, "loss": 0.0155, "step": 116510 }, { "grad_norm": 0.06672599166631699, "learning_rate": 7.182653663678873e-05, "loss": 0.0168, "step": 116520 }, { "grad_norm": 0.1640509068965912, "learning_rate": 7.182157780692905e-05, "loss": 0.0208, "step": 116530 }, { "grad_norm": 0.10890483111143112, "learning_rate": 7.18166187119166e-05, "loss": 0.0169, "step": 116540 }, { "grad_norm": 0.15996935963630676, "learning_rate": 7.181165935181163e-05, "loss": 0.0192, "step": 116550 }, { "grad_norm": 0.1681789755821228, "learning_rate": 7.18066997266744e-05, "loss": 0.014, "step": 116560 }, { "grad_norm": 0.1825590431690216, "learning_rate": 7.180173983656518e-05, "loss": 0.0215, "step": 116570 }, { "grad_norm": 0.1276136189699173, "learning_rate": 7.179677968154424e-05, "loss": 0.0168, "step": 116580 }, { "grad_norm": 0.2695171535015106, "learning_rate": 7.179181926167184e-05, "loss": 0.0218, "step": 116590 }, { "grad_norm": 0.14806663990020752, "learning_rate": 7.178685857700827e-05, "loss": 0.0138, "step": 116600 }, { "grad_norm": 0.1139407828450203, "learning_rate": 7.178189762761378e-05, "loss": 0.023, "step": 116610 }, { "grad_norm": 0.13737502694129944, "learning_rate": 7.177693641354868e-05, "loss": 0.0189, "step": 116620 }, { "grad_norm": 0.16587094962596893, "learning_rate": 7.177197493487326e-05, "loss": 0.0224, "step": 116630 }, { "grad_norm": 0.16882000863552094, "learning_rate": 7.176701319164777e-05, "loss": 0.0178, "step": 116640 }, { "grad_norm": 0.21410802006721497, "learning_rate": 7.176205118393251e-05, "loss": 0.0186, "step": 116650 }, { "grad_norm": 0.1599307805299759, "learning_rate": 7.17570889117878e-05, "loss": 0.0156, "step": 116660 }, { "grad_norm": 0.1164630576968193, "learning_rate": 7.17521263752739e-05, "loss": 0.0166, "step": 116670 }, { "grad_norm": 0.2208763062953949, "learning_rate": 7.174716357445115e-05, "loss": 0.0203, "step": 116680 }, { "grad_norm": 0.19023890793323517, "learning_rate": 7.17422005093798e-05, "loss": 0.021, "step": 116690 }, { "grad_norm": 0.12603972852230072, "learning_rate": 7.173723718012022e-05, "loss": 0.0179, "step": 116700 }, { "grad_norm": 0.15777698159217834, "learning_rate": 7.173227358673267e-05, "loss": 0.0198, "step": 116710 }, { "grad_norm": 0.1799030303955078, "learning_rate": 7.172730972927748e-05, "loss": 0.0156, "step": 116720 }, { "grad_norm": 0.1224314421415329, "learning_rate": 7.172234560781495e-05, "loss": 0.0176, "step": 116730 }, { "grad_norm": 0.08821552246809006, "learning_rate": 7.171738122240543e-05, "loss": 0.0173, "step": 116740 }, { "grad_norm": 0.16739973425865173, "learning_rate": 7.171241657310921e-05, "loss": 0.019, "step": 116750 }, { "grad_norm": 0.10654225200414658, "learning_rate": 7.170745165998663e-05, "loss": 0.0158, "step": 116760 }, { "grad_norm": 0.12271880358457565, "learning_rate": 7.170248648309801e-05, "loss": 0.0132, "step": 116770 }, { "grad_norm": 0.12016583234071732, "learning_rate": 7.16975210425037e-05, "loss": 0.0193, "step": 116780 }, { "grad_norm": 0.18194036185741425, "learning_rate": 7.169255533826403e-05, "loss": 0.0155, "step": 116790 }, { "grad_norm": 0.10260002315044403, "learning_rate": 7.168758937043933e-05, "loss": 0.0146, "step": 116800 }, { "grad_norm": 0.15607339143753052, "learning_rate": 7.168262313908993e-05, "loss": 0.0151, "step": 116810 }, { "grad_norm": 0.14983493089675903, "learning_rate": 7.16776566442762e-05, "loss": 0.014, "step": 116820 }, { "grad_norm": 0.09439606219530106, "learning_rate": 7.167268988605845e-05, "loss": 0.0144, "step": 116830 }, { "grad_norm": 0.2652810513973236, "learning_rate": 7.166772286449708e-05, "loss": 0.0125, "step": 116840 }, { "grad_norm": 0.16555155813694, "learning_rate": 7.166275557965241e-05, "loss": 0.0202, "step": 116850 }, { "grad_norm": 0.1673024743795395, "learning_rate": 7.16577880315848e-05, "loss": 0.02, "step": 116860 }, { "grad_norm": 0.16709618270397186, "learning_rate": 7.165282022035462e-05, "loss": 0.016, "step": 116870 }, { "grad_norm": 0.1839124709367752, "learning_rate": 7.164785214602226e-05, "loss": 0.0223, "step": 116880 }, { "grad_norm": 0.09521247446537018, "learning_rate": 7.164288380864803e-05, "loss": 0.014, "step": 116890 }, { "grad_norm": 0.10447660088539124, "learning_rate": 7.163791520829232e-05, "loss": 0.0184, "step": 116900 }, { "grad_norm": 0.11864666640758514, "learning_rate": 7.163294634501553e-05, "loss": 0.0178, "step": 116910 }, { "grad_norm": 0.29373571276664734, "learning_rate": 7.162797721887801e-05, "loss": 0.0144, "step": 116920 }, { "grad_norm": 0.11007582396268845, "learning_rate": 7.162300782994015e-05, "loss": 0.0119, "step": 116930 }, { "grad_norm": 0.18449361622333527, "learning_rate": 7.161803817826231e-05, "loss": 0.0166, "step": 116940 }, { "grad_norm": 0.14307467639446259, "learning_rate": 7.161306826390491e-05, "loss": 0.0162, "step": 116950 }, { "grad_norm": 0.12491422146558762, "learning_rate": 7.160809808692835e-05, "loss": 0.0133, "step": 116960 }, { "grad_norm": 0.1277530938386917, "learning_rate": 7.160312764739295e-05, "loss": 0.0157, "step": 116970 }, { "grad_norm": 0.13768121600151062, "learning_rate": 7.159815694535919e-05, "loss": 0.0145, "step": 116980 }, { "grad_norm": 0.11429761350154877, "learning_rate": 7.15931859808874e-05, "loss": 0.0148, "step": 116990 }, { "grad_norm": 0.21418173611164093, "learning_rate": 7.158821475403803e-05, "loss": 0.015, "step": 117000 }, { "grad_norm": 0.08701959997415543, "learning_rate": 7.158324326487147e-05, "loss": 0.0194, "step": 117010 }, { "grad_norm": 0.11732081323862076, "learning_rate": 7.157827151344813e-05, "loss": 0.0206, "step": 117020 }, { "grad_norm": 0.16095012426376343, "learning_rate": 7.15732994998284e-05, "loss": 0.0169, "step": 117030 }, { "grad_norm": 0.14569982886314392, "learning_rate": 7.156832722407271e-05, "loss": 0.02, "step": 117040 }, { "grad_norm": 0.20055201649665833, "learning_rate": 7.156335468624151e-05, "loss": 0.0132, "step": 117050 }, { "grad_norm": 0.12041648477315903, "learning_rate": 7.155838188639517e-05, "loss": 0.0197, "step": 117060 }, { "grad_norm": 0.15355676412582397, "learning_rate": 7.155340882459414e-05, "loss": 0.0154, "step": 117070 }, { "grad_norm": 0.11446376889944077, "learning_rate": 7.154843550089885e-05, "loss": 0.0165, "step": 117080 }, { "grad_norm": 0.1728115826845169, "learning_rate": 7.154346191536971e-05, "loss": 0.0158, "step": 117090 }, { "grad_norm": 0.08797873556613922, "learning_rate": 7.153848806806718e-05, "loss": 0.0183, "step": 117100 }, { "grad_norm": 0.08239037543535233, "learning_rate": 7.153351395905168e-05, "loss": 0.0163, "step": 117110 }, { "grad_norm": 0.15891461074352264, "learning_rate": 7.152853958838366e-05, "loss": 0.0145, "step": 117120 }, { "grad_norm": 0.08291279524564743, "learning_rate": 7.152356495612354e-05, "loss": 0.011, "step": 117130 }, { "grad_norm": 0.11569324135780334, "learning_rate": 7.151859006233179e-05, "loss": 0.0171, "step": 117140 }, { "grad_norm": 0.26725688576698303, "learning_rate": 7.151361490706886e-05, "loss": 0.016, "step": 117150 }, { "grad_norm": 0.2730071544647217, "learning_rate": 7.15086394903952e-05, "loss": 0.0179, "step": 117160 }, { "grad_norm": 0.24130317568778992, "learning_rate": 7.150366381237125e-05, "loss": 0.0224, "step": 117170 }, { "grad_norm": 0.14588335156440735, "learning_rate": 7.149868787305749e-05, "loss": 0.0181, "step": 117180 }, { "grad_norm": 0.12836897373199463, "learning_rate": 7.149371167251438e-05, "loss": 0.0166, "step": 117190 }, { "grad_norm": 0.09495070576667786, "learning_rate": 7.148873521080237e-05, "loss": 0.0227, "step": 117200 }, { "grad_norm": 0.10843612998723984, "learning_rate": 7.148375848798194e-05, "loss": 0.0133, "step": 117210 }, { "grad_norm": 0.20784935355186462, "learning_rate": 7.147878150411356e-05, "loss": 0.0236, "step": 117220 }, { "grad_norm": 0.21647830307483673, "learning_rate": 7.147380425925771e-05, "loss": 0.0176, "step": 117230 }, { "grad_norm": 0.1316125988960266, "learning_rate": 7.146882675347486e-05, "loss": 0.0162, "step": 117240 }, { "grad_norm": 0.2413434386253357, "learning_rate": 7.14638489868255e-05, "loss": 0.0138, "step": 117250 }, { "grad_norm": 0.14727294445037842, "learning_rate": 7.14588709593701e-05, "loss": 0.0164, "step": 117260 }, { "grad_norm": 0.19007694721221924, "learning_rate": 7.145389267116917e-05, "loss": 0.0148, "step": 117270 }, { "grad_norm": 0.3818257451057434, "learning_rate": 7.144891412228318e-05, "loss": 0.0194, "step": 117280 }, { "grad_norm": 0.21761251986026764, "learning_rate": 7.144393531277264e-05, "loss": 0.0214, "step": 117290 }, { "grad_norm": 0.18024952709674835, "learning_rate": 7.143895624269802e-05, "loss": 0.0168, "step": 117300 }, { "grad_norm": 0.1158817857503891, "learning_rate": 7.143397691211986e-05, "loss": 0.0185, "step": 117310 }, { "grad_norm": 0.15041451156139374, "learning_rate": 7.142899732109864e-05, "loss": 0.0171, "step": 117320 }, { "grad_norm": 0.13201016187667847, "learning_rate": 7.142401746969487e-05, "loss": 0.0158, "step": 117330 }, { "grad_norm": 0.21731846034526825, "learning_rate": 7.141903735796904e-05, "loss": 0.0157, "step": 117340 }, { "grad_norm": 0.17379000782966614, "learning_rate": 7.141405698598172e-05, "loss": 0.0139, "step": 117350 }, { "grad_norm": 0.2097267508506775, "learning_rate": 7.140907635379336e-05, "loss": 0.015, "step": 117360 }, { "grad_norm": 0.11331567913293839, "learning_rate": 7.140409546146452e-05, "loss": 0.015, "step": 117370 }, { "grad_norm": 0.16222168505191803, "learning_rate": 7.139911430905571e-05, "loss": 0.02, "step": 117380 }, { "grad_norm": 0.17642474174499512, "learning_rate": 7.139413289662745e-05, "loss": 0.0188, "step": 117390 }, { "grad_norm": 0.6169338226318359, "learning_rate": 7.138915122424029e-05, "loss": 0.0211, "step": 117400 }, { "grad_norm": 0.10308096557855606, "learning_rate": 7.138416929195474e-05, "loss": 0.0165, "step": 117410 }, { "grad_norm": 0.07533916085958481, "learning_rate": 7.137918709983134e-05, "loss": 0.0126, "step": 117420 }, { "grad_norm": 0.10183040052652359, "learning_rate": 7.137420464793063e-05, "loss": 0.0125, "step": 117430 }, { "grad_norm": 0.2936237156391144, "learning_rate": 7.136922193631317e-05, "loss": 0.0169, "step": 117440 }, { "grad_norm": 0.10621175915002823, "learning_rate": 7.136423896503948e-05, "loss": 0.0137, "step": 117450 }, { "grad_norm": 0.10895542800426483, "learning_rate": 7.13592557341701e-05, "loss": 0.0142, "step": 117460 }, { "grad_norm": 0.21144579350948334, "learning_rate": 7.135427224376561e-05, "loss": 0.0177, "step": 117470 }, { "grad_norm": 0.1759093552827835, "learning_rate": 7.134928849388655e-05, "loss": 0.0135, "step": 117480 }, { "grad_norm": 0.12184105813503265, "learning_rate": 7.134430448459348e-05, "loss": 0.0178, "step": 117490 }, { "grad_norm": 0.1092265322804451, "learning_rate": 7.133932021594694e-05, "loss": 0.0212, "step": 117500 }, { "grad_norm": 0.1373601257801056, "learning_rate": 7.133433568800752e-05, "loss": 0.0187, "step": 117510 }, { "grad_norm": 0.11286690831184387, "learning_rate": 7.13293509008358e-05, "loss": 0.0186, "step": 117520 }, { "grad_norm": 0.27003782987594604, "learning_rate": 7.132436585449231e-05, "loss": 0.0153, "step": 117530 }, { "grad_norm": 0.17908000946044922, "learning_rate": 7.131938054903763e-05, "loss": 0.022, "step": 117540 }, { "grad_norm": 0.1323666125535965, "learning_rate": 7.131439498453237e-05, "loss": 0.0136, "step": 117550 }, { "grad_norm": 0.20933538675308228, "learning_rate": 7.130940916103707e-05, "loss": 0.0125, "step": 117560 }, { "grad_norm": 0.20285607874393463, "learning_rate": 7.130442307861235e-05, "loss": 0.0159, "step": 117570 }, { "grad_norm": 0.1592434048652649, "learning_rate": 7.129943673731876e-05, "loss": 0.0143, "step": 117580 }, { "grad_norm": 0.1971443146467209, "learning_rate": 7.12944501372169e-05, "loss": 0.018, "step": 117590 }, { "grad_norm": 0.13226185739040375, "learning_rate": 7.128946327836738e-05, "loss": 0.0203, "step": 117600 }, { "grad_norm": 0.2833586633205414, "learning_rate": 7.128447616083077e-05, "loss": 0.018, "step": 117610 }, { "grad_norm": 0.12803690135478973, "learning_rate": 7.127948878466768e-05, "loss": 0.0159, "step": 117620 }, { "grad_norm": 0.13482216000556946, "learning_rate": 7.127450114993871e-05, "loss": 0.0171, "step": 117630 }, { "grad_norm": 0.09394072741270065, "learning_rate": 7.126951325670447e-05, "loss": 0.0137, "step": 117640 }, { "grad_norm": 0.18082332611083984, "learning_rate": 7.126452510502554e-05, "loss": 0.0214, "step": 117650 }, { "grad_norm": 0.14631889760494232, "learning_rate": 7.125953669496259e-05, "loss": 0.0148, "step": 117660 }, { "grad_norm": 0.11715500801801682, "learning_rate": 7.125454802657617e-05, "loss": 0.0165, "step": 117670 }, { "grad_norm": 0.08775252848863602, "learning_rate": 7.124955909992693e-05, "loss": 0.0157, "step": 117680 }, { "grad_norm": 0.17622575163841248, "learning_rate": 7.124456991507548e-05, "loss": 0.0204, "step": 117690 }, { "grad_norm": 0.10048894584178925, "learning_rate": 7.123958047208246e-05, "loss": 0.0178, "step": 117700 }, { "grad_norm": 0.1574784219264984, "learning_rate": 7.123459077100848e-05, "loss": 0.0163, "step": 117710 }, { "grad_norm": 0.12243463099002838, "learning_rate": 7.122960081191417e-05, "loss": 0.0191, "step": 117720 }, { "grad_norm": 0.12519928812980652, "learning_rate": 7.122461059486016e-05, "loss": 0.0149, "step": 117730 }, { "grad_norm": 0.11528525501489639, "learning_rate": 7.12196201199071e-05, "loss": 0.0132, "step": 117740 }, { "grad_norm": 0.1527291089296341, "learning_rate": 7.121462938711561e-05, "loss": 0.0189, "step": 117750 }, { "grad_norm": 0.124269500374794, "learning_rate": 7.120963839654635e-05, "loss": 0.0127, "step": 117760 }, { "grad_norm": 0.6006835699081421, "learning_rate": 7.120464714825998e-05, "loss": 0.0183, "step": 117770 }, { "grad_norm": 0.16114740073680878, "learning_rate": 7.11996556423171e-05, "loss": 0.0183, "step": 117780 }, { "grad_norm": 0.18106922507286072, "learning_rate": 7.119466387877839e-05, "loss": 0.016, "step": 117790 }, { "grad_norm": 0.12747807800769806, "learning_rate": 7.11896718577045e-05, "loss": 0.018, "step": 117800 }, { "grad_norm": 0.18502801656723022, "learning_rate": 7.11846795791561e-05, "loss": 0.0152, "step": 117810 }, { "grad_norm": 0.18538770079612732, "learning_rate": 7.117968704319384e-05, "loss": 0.0196, "step": 117820 }, { "grad_norm": 0.14671234786510468, "learning_rate": 7.117469424987839e-05, "loss": 0.0154, "step": 117830 }, { "grad_norm": 0.1539025604724884, "learning_rate": 7.116970119927041e-05, "loss": 0.0194, "step": 117840 }, { "grad_norm": 0.21047016978263855, "learning_rate": 7.116470789143057e-05, "loss": 0.0167, "step": 117850 }, { "grad_norm": 0.4505155086517334, "learning_rate": 7.115971432641955e-05, "loss": 0.0229, "step": 117860 }, { "grad_norm": 0.21080762147903442, "learning_rate": 7.115472050429802e-05, "loss": 0.0142, "step": 117870 }, { "grad_norm": 0.2738512456417084, "learning_rate": 7.114972642512665e-05, "loss": 0.0134, "step": 117880 }, { "grad_norm": 0.3597167432308197, "learning_rate": 7.114473208896616e-05, "loss": 0.0226, "step": 117890 }, { "grad_norm": 0.08855769783258438, "learning_rate": 7.113973749587719e-05, "loss": 0.0183, "step": 117900 }, { "grad_norm": 0.19408787786960602, "learning_rate": 7.113474264592045e-05, "loss": 0.022, "step": 117910 }, { "grad_norm": 0.13335716724395752, "learning_rate": 7.112974753915665e-05, "loss": 0.0152, "step": 117920 }, { "grad_norm": 0.18553897738456726, "learning_rate": 7.112475217564647e-05, "loss": 0.0172, "step": 117930 }, { "grad_norm": 0.11151681840419769, "learning_rate": 7.111975655545059e-05, "loss": 0.0175, "step": 117940 }, { "grad_norm": 0.12097867578268051, "learning_rate": 7.111476067862973e-05, "loss": 0.0176, "step": 117950 }, { "grad_norm": 0.16466554999351501, "learning_rate": 7.110976454524459e-05, "loss": 0.0119, "step": 117960 }, { "grad_norm": 0.08866379410028458, "learning_rate": 7.110476815535588e-05, "loss": 0.0144, "step": 117970 }, { "grad_norm": 0.16096527874469757, "learning_rate": 7.109977150902433e-05, "loss": 0.0152, "step": 117980 }, { "grad_norm": 0.16818684339523315, "learning_rate": 7.10947746063106e-05, "loss": 0.0182, "step": 117990 }, { "grad_norm": 0.15268543362617493, "learning_rate": 7.108977744727547e-05, "loss": 0.0161, "step": 118000 }, { "grad_norm": 0.30089107155799866, "learning_rate": 7.108478003197963e-05, "loss": 0.0175, "step": 118010 }, { "grad_norm": 0.10786403715610504, "learning_rate": 7.10797823604838e-05, "loss": 0.0201, "step": 118020 }, { "grad_norm": 0.06616287678480148, "learning_rate": 7.107478443284869e-05, "loss": 0.0132, "step": 118030 }, { "grad_norm": 0.15114492177963257, "learning_rate": 7.106978624913509e-05, "loss": 0.0209, "step": 118040 }, { "grad_norm": 0.100642129778862, "learning_rate": 7.106478780940366e-05, "loss": 0.0159, "step": 118050 }, { "grad_norm": 0.07668694853782654, "learning_rate": 7.10597891137152e-05, "loss": 0.0144, "step": 118060 }, { "grad_norm": 0.17848047614097595, "learning_rate": 7.105479016213039e-05, "loss": 0.0177, "step": 118070 }, { "grad_norm": 0.14976398646831512, "learning_rate": 7.104979095471003e-05, "loss": 0.0184, "step": 118080 }, { "grad_norm": 0.15860126912593842, "learning_rate": 7.10447914915148e-05, "loss": 0.0192, "step": 118090 }, { "grad_norm": 0.27245914936065674, "learning_rate": 7.103979177260551e-05, "loss": 0.0184, "step": 118100 }, { "grad_norm": 0.4874148368835449, "learning_rate": 7.103479179804287e-05, "loss": 0.0202, "step": 118110 }, { "grad_norm": 0.12930920720100403, "learning_rate": 7.102979156788767e-05, "loss": 0.0137, "step": 118120 }, { "grad_norm": 0.1816425323486328, "learning_rate": 7.102479108220063e-05, "loss": 0.0136, "step": 118130 }, { "grad_norm": 0.12967252731323242, "learning_rate": 7.101979034104254e-05, "loss": 0.0124, "step": 118140 }, { "grad_norm": 0.09843152016401291, "learning_rate": 7.101478934447414e-05, "loss": 0.0151, "step": 118150 }, { "grad_norm": 0.14346326887607574, "learning_rate": 7.100978809255622e-05, "loss": 0.0173, "step": 118160 }, { "grad_norm": 0.3569229245185852, "learning_rate": 7.100478658534952e-05, "loss": 0.0196, "step": 118170 }, { "grad_norm": 0.09080960601568222, "learning_rate": 7.099978482291484e-05, "loss": 0.018, "step": 118180 }, { "grad_norm": 0.09383329749107361, "learning_rate": 7.099478280531295e-05, "loss": 0.0179, "step": 118190 }, { "grad_norm": 0.08665081858634949, "learning_rate": 7.098978053260464e-05, "loss": 0.0151, "step": 118200 }, { "grad_norm": 0.17914444208145142, "learning_rate": 7.098477800485066e-05, "loss": 0.0183, "step": 118210 }, { "grad_norm": 0.10245011001825333, "learning_rate": 7.09797752221118e-05, "loss": 0.0188, "step": 118220 }, { "grad_norm": 0.1608692854642868, "learning_rate": 7.09747721844489e-05, "loss": 0.017, "step": 118230 }, { "grad_norm": 0.12714046239852905, "learning_rate": 7.096976889192268e-05, "loss": 0.0181, "step": 118240 }, { "grad_norm": 0.4146439731121063, "learning_rate": 7.0964765344594e-05, "loss": 0.0197, "step": 118250 }, { "grad_norm": 0.10699351131916046, "learning_rate": 7.095976154252362e-05, "loss": 0.0143, "step": 118260 }, { "grad_norm": 0.232599139213562, "learning_rate": 7.095475748577233e-05, "loss": 0.0125, "step": 118270 }, { "grad_norm": 0.190468430519104, "learning_rate": 7.094975317440097e-05, "loss": 0.0148, "step": 118280 }, { "grad_norm": 0.12621568143367767, "learning_rate": 7.094474860847032e-05, "loss": 0.0157, "step": 118290 }, { "grad_norm": 0.2163465917110443, "learning_rate": 7.09397437880412e-05, "loss": 0.0232, "step": 118300 }, { "grad_norm": 0.21061116456985474, "learning_rate": 7.093473871317442e-05, "loss": 0.0149, "step": 118310 }, { "grad_norm": 0.18498674035072327, "learning_rate": 7.092973338393081e-05, "loss": 0.017, "step": 118320 }, { "grad_norm": 0.08151570707559586, "learning_rate": 7.092472780037117e-05, "loss": 0.0154, "step": 118330 }, { "grad_norm": 0.29802411794662476, "learning_rate": 7.091972196255632e-05, "loss": 0.0217, "step": 118340 }, { "grad_norm": 0.13599851727485657, "learning_rate": 7.091471587054712e-05, "loss": 0.0136, "step": 118350 }, { "grad_norm": 0.21981315314769745, "learning_rate": 7.090970952440435e-05, "loss": 0.0139, "step": 118360 }, { "grad_norm": 0.2983214259147644, "learning_rate": 7.090470292418889e-05, "loss": 0.0182, "step": 118370 }, { "grad_norm": 0.12432720512151718, "learning_rate": 7.089969606996154e-05, "loss": 0.0196, "step": 118380 }, { "grad_norm": 0.1460530161857605, "learning_rate": 7.089468896178316e-05, "loss": 0.0161, "step": 118390 }, { "grad_norm": 0.12523004412651062, "learning_rate": 7.088968159971457e-05, "loss": 0.0185, "step": 118400 }, { "grad_norm": 0.1458287388086319, "learning_rate": 7.088467398381663e-05, "loss": 0.0161, "step": 118410 }, { "grad_norm": 0.39678263664245605, "learning_rate": 7.087966611415019e-05, "loss": 0.0152, "step": 118420 }, { "grad_norm": 0.21878470480442047, "learning_rate": 7.087465799077608e-05, "loss": 0.0135, "step": 118430 }, { "grad_norm": 0.2321922928094864, "learning_rate": 7.086964961375518e-05, "loss": 0.0163, "step": 118440 }, { "grad_norm": 0.09046734124422073, "learning_rate": 7.086464098314832e-05, "loss": 0.0125, "step": 118450 }, { "grad_norm": 0.13410647213459015, "learning_rate": 7.085963209901638e-05, "loss": 0.0115, "step": 118460 }, { "grad_norm": 0.2132820188999176, "learning_rate": 7.08546229614202e-05, "loss": 0.02, "step": 118470 }, { "grad_norm": 0.1766316443681717, "learning_rate": 7.084961357042068e-05, "loss": 0.0145, "step": 118480 }, { "grad_norm": 0.24261344969272614, "learning_rate": 7.084460392607868e-05, "loss": 0.0184, "step": 118490 }, { "grad_norm": 0.11742761731147766, "learning_rate": 7.083959402845504e-05, "loss": 0.0185, "step": 118500 }, { "grad_norm": 0.11475104093551636, "learning_rate": 7.083458387761065e-05, "loss": 0.0212, "step": 118510 }, { "grad_norm": 0.1183866634964943, "learning_rate": 7.082957347360642e-05, "loss": 0.0179, "step": 118520 }, { "grad_norm": 0.09929608553647995, "learning_rate": 7.08245628165032e-05, "loss": 0.0157, "step": 118530 }, { "grad_norm": 0.09844493120908737, "learning_rate": 7.081955190636185e-05, "loss": 0.0148, "step": 118540 }, { "grad_norm": 0.09715308248996735, "learning_rate": 7.081454074324332e-05, "loss": 0.0203, "step": 118550 }, { "grad_norm": 0.09644633531570435, "learning_rate": 7.080952932720845e-05, "loss": 0.013, "step": 118560 }, { "grad_norm": 0.1462705433368683, "learning_rate": 7.080451765831817e-05, "loss": 0.0174, "step": 118570 }, { "grad_norm": 0.1972828209400177, "learning_rate": 7.079950573663336e-05, "loss": 0.0175, "step": 118580 }, { "grad_norm": 0.10880438983440399, "learning_rate": 7.079449356221489e-05, "loss": 0.0174, "step": 118590 }, { "grad_norm": 0.21418815851211548, "learning_rate": 7.078948113512372e-05, "loss": 0.016, "step": 118600 }, { "grad_norm": 0.14365817606449127, "learning_rate": 7.07844684554207e-05, "loss": 0.0165, "step": 118610 }, { "grad_norm": 0.14089149236679077, "learning_rate": 7.077945552316678e-05, "loss": 0.0137, "step": 118620 }, { "grad_norm": 0.1820652037858963, "learning_rate": 7.077444233842285e-05, "loss": 0.0148, "step": 118630 }, { "grad_norm": 0.5525891184806824, "learning_rate": 7.076942890124983e-05, "loss": 0.0183, "step": 118640 }, { "grad_norm": 0.15976868569850922, "learning_rate": 7.076441521170865e-05, "loss": 0.0183, "step": 118650 }, { "grad_norm": 0.14437851309776306, "learning_rate": 7.07594012698602e-05, "loss": 0.0194, "step": 118660 }, { "grad_norm": 0.12975884974002838, "learning_rate": 7.075438707576544e-05, "loss": 0.0132, "step": 118670 }, { "grad_norm": 0.15377408266067505, "learning_rate": 7.074937262948527e-05, "loss": 0.0143, "step": 118680 }, { "grad_norm": 0.13361714780330658, "learning_rate": 7.074435793108065e-05, "loss": 0.0195, "step": 118690 }, { "grad_norm": 0.4927814304828644, "learning_rate": 7.073934298061247e-05, "loss": 0.0149, "step": 118700 }, { "grad_norm": 0.09993527084589005, "learning_rate": 7.073432777814172e-05, "loss": 0.0142, "step": 118710 }, { "grad_norm": 0.13645538687705994, "learning_rate": 7.072931232372929e-05, "loss": 0.0138, "step": 118720 }, { "grad_norm": 0.11504853516817093, "learning_rate": 7.072429661743616e-05, "loss": 0.0142, "step": 118730 }, { "grad_norm": 0.12084610760211945, "learning_rate": 7.071928065932324e-05, "loss": 0.0179, "step": 118740 }, { "grad_norm": 0.16697898507118225, "learning_rate": 7.071426444945151e-05, "loss": 0.0154, "step": 118750 }, { "grad_norm": 0.26799967885017395, "learning_rate": 7.070924798788191e-05, "loss": 0.0145, "step": 118760 }, { "grad_norm": 0.12545856833457947, "learning_rate": 7.070423127467539e-05, "loss": 0.0158, "step": 118770 }, { "grad_norm": 0.057007696479558945, "learning_rate": 7.069921430989292e-05, "loss": 0.0157, "step": 118780 }, { "grad_norm": 0.09770768135786057, "learning_rate": 7.069419709359544e-05, "loss": 0.013, "step": 118790 }, { "grad_norm": 0.1275063306093216, "learning_rate": 7.068917962584394e-05, "loss": 0.0154, "step": 118800 }, { "grad_norm": 0.1287049949169159, "learning_rate": 7.068416190669937e-05, "loss": 0.0188, "step": 118810 }, { "grad_norm": 0.12295689433813095, "learning_rate": 7.067914393622269e-05, "loss": 0.0165, "step": 118820 }, { "grad_norm": 0.21853315830230713, "learning_rate": 7.06741257144749e-05, "loss": 0.0171, "step": 118830 }, { "grad_norm": 0.085940420627594, "learning_rate": 7.066910724151696e-05, "loss": 0.0169, "step": 118840 }, { "grad_norm": 0.15910519659519196, "learning_rate": 7.066408851740986e-05, "loss": 0.0166, "step": 118850 }, { "grad_norm": 0.11111671477556229, "learning_rate": 7.065906954221455e-05, "loss": 0.0181, "step": 118860 }, { "grad_norm": 0.2015143632888794, "learning_rate": 7.065405031599207e-05, "loss": 0.016, "step": 118870 }, { "grad_norm": 0.10185688734054565, "learning_rate": 7.064903083880335e-05, "loss": 0.0182, "step": 118880 }, { "grad_norm": 0.07767615467309952, "learning_rate": 7.064401111070944e-05, "loss": 0.0192, "step": 118890 }, { "grad_norm": 0.07182929664850235, "learning_rate": 7.063899113177128e-05, "loss": 0.0208, "step": 118900 }, { "grad_norm": 0.0762908011674881, "learning_rate": 7.06339709020499e-05, "loss": 0.0157, "step": 118910 }, { "grad_norm": 0.144718736410141, "learning_rate": 7.06289504216063e-05, "loss": 0.0162, "step": 118920 }, { "grad_norm": 0.14914025366306305, "learning_rate": 7.062392969050145e-05, "loss": 0.0164, "step": 118930 }, { "grad_norm": 0.095128133893013, "learning_rate": 7.061890870879642e-05, "loss": 0.0253, "step": 118940 }, { "grad_norm": 0.18130482733249664, "learning_rate": 7.061388747655214e-05, "loss": 0.0148, "step": 118950 }, { "grad_norm": 0.20923112332820892, "learning_rate": 7.060886599382969e-05, "loss": 0.0169, "step": 118960 }, { "grad_norm": 0.2572091817855835, "learning_rate": 7.060384426069006e-05, "loss": 0.0191, "step": 118970 }, { "grad_norm": 0.22661058604717255, "learning_rate": 7.059882227719427e-05, "loss": 0.0141, "step": 118980 }, { "grad_norm": 0.12339272350072861, "learning_rate": 7.059380004340333e-05, "loss": 0.0143, "step": 118990 }, { "grad_norm": 0.14835257828235626, "learning_rate": 7.058877755937828e-05, "loss": 0.0133, "step": 119000 }, { "grad_norm": 0.2246703952550888, "learning_rate": 7.058375482518014e-05, "loss": 0.0156, "step": 119010 }, { "grad_norm": 0.13371695578098297, "learning_rate": 7.057873184086994e-05, "loss": 0.0185, "step": 119020 }, { "grad_norm": 0.08217060565948486, "learning_rate": 7.057370860650873e-05, "loss": 0.0105, "step": 119030 }, { "grad_norm": 0.5478920340538025, "learning_rate": 7.056868512215754e-05, "loss": 0.017, "step": 119040 }, { "grad_norm": 0.11786506325006485, "learning_rate": 7.05636613878774e-05, "loss": 0.0157, "step": 119050 }, { "grad_norm": 0.18814624845981598, "learning_rate": 7.055863740372936e-05, "loss": 0.0185, "step": 119060 }, { "grad_norm": 0.0942763239145279, "learning_rate": 7.055361316977446e-05, "loss": 0.0162, "step": 119070 }, { "grad_norm": 0.17486947774887085, "learning_rate": 7.054858868607377e-05, "loss": 0.0203, "step": 119080 }, { "grad_norm": 0.09701469540596008, "learning_rate": 7.05435639526883e-05, "loss": 0.0138, "step": 119090 }, { "grad_norm": 0.09167743474245071, "learning_rate": 7.053853896967914e-05, "loss": 0.0125, "step": 119100 }, { "grad_norm": 0.1516866385936737, "learning_rate": 7.053351373710735e-05, "loss": 0.0193, "step": 119110 }, { "grad_norm": 0.17500221729278564, "learning_rate": 7.052848825503398e-05, "loss": 0.0203, "step": 119120 }, { "grad_norm": 0.13466043770313263, "learning_rate": 7.052346252352008e-05, "loss": 0.0157, "step": 119130 }, { "grad_norm": 0.17816303670406342, "learning_rate": 7.051843654262676e-05, "loss": 0.0188, "step": 119140 }, { "grad_norm": 0.1334196776151657, "learning_rate": 7.051341031241505e-05, "loss": 0.0133, "step": 119150 }, { "grad_norm": 0.12716202437877655, "learning_rate": 7.050838383294603e-05, "loss": 0.0176, "step": 119160 }, { "grad_norm": 0.15856267511844635, "learning_rate": 7.050335710428078e-05, "loss": 0.0194, "step": 119170 }, { "grad_norm": 0.13743196427822113, "learning_rate": 7.04983301264804e-05, "loss": 0.0199, "step": 119180 }, { "grad_norm": 0.10809537023305893, "learning_rate": 7.049330289960594e-05, "loss": 0.014, "step": 119190 }, { "grad_norm": 0.1358654648065567, "learning_rate": 7.048827542371851e-05, "loss": 0.0153, "step": 119200 }, { "grad_norm": 0.22238224744796753, "learning_rate": 7.048324769887919e-05, "loss": 0.0153, "step": 119210 }, { "grad_norm": 0.08931659162044525, "learning_rate": 7.047821972514907e-05, "loss": 0.0166, "step": 119220 }, { "grad_norm": 0.2507650554180145, "learning_rate": 7.047319150258925e-05, "loss": 0.0175, "step": 119230 }, { "grad_norm": 0.14684350788593292, "learning_rate": 7.04681630312608e-05, "loss": 0.0132, "step": 119240 }, { "grad_norm": 0.11629582941532135, "learning_rate": 7.046313431122487e-05, "loss": 0.0186, "step": 119250 }, { "grad_norm": 0.11735551804304123, "learning_rate": 7.045810534254252e-05, "loss": 0.0178, "step": 119260 }, { "grad_norm": 0.11412794142961502, "learning_rate": 7.045307612527489e-05, "loss": 0.0143, "step": 119270 }, { "grad_norm": 0.20440775156021118, "learning_rate": 7.044804665948305e-05, "loss": 0.015, "step": 119280 }, { "grad_norm": 0.16193819046020508, "learning_rate": 7.044301694522817e-05, "loss": 0.014, "step": 119290 }, { "grad_norm": 0.1267552375793457, "learning_rate": 7.043798698257132e-05, "loss": 0.0127, "step": 119300 }, { "grad_norm": 0.17843304574489594, "learning_rate": 7.043295677157362e-05, "loss": 0.0186, "step": 119310 }, { "grad_norm": 0.10051511973142624, "learning_rate": 7.04279263122962e-05, "loss": 0.0165, "step": 119320 }, { "grad_norm": 0.15183483064174652, "learning_rate": 7.042289560480023e-05, "loss": 0.0163, "step": 119330 }, { "grad_norm": 0.15391303598880768, "learning_rate": 7.041786464914674e-05, "loss": 0.0306, "step": 119340 }, { "grad_norm": 0.1479000300168991, "learning_rate": 7.041283344539696e-05, "loss": 0.0164, "step": 119350 }, { "grad_norm": 0.19322219491004944, "learning_rate": 7.040780199361195e-05, "loss": 0.015, "step": 119360 }, { "grad_norm": 0.1392347812652588, "learning_rate": 7.04027702938529e-05, "loss": 0.0174, "step": 119370 }, { "grad_norm": 0.4002395272254944, "learning_rate": 7.039773834618092e-05, "loss": 0.0191, "step": 119380 }, { "grad_norm": 0.11185751110315323, "learning_rate": 7.039270615065715e-05, "loss": 0.0161, "step": 119390 }, { "grad_norm": 0.1259412318468094, "learning_rate": 7.038767370734276e-05, "loss": 0.0139, "step": 119400 }, { "grad_norm": 0.2638702988624573, "learning_rate": 7.038264101629888e-05, "loss": 0.0184, "step": 119410 }, { "grad_norm": 0.16079187393188477, "learning_rate": 7.037760807758666e-05, "loss": 0.0125, "step": 119420 }, { "grad_norm": 0.1354466676712036, "learning_rate": 7.037257489126728e-05, "loss": 0.0174, "step": 119430 }, { "grad_norm": 0.18980209529399872, "learning_rate": 7.036754145740186e-05, "loss": 0.0202, "step": 119440 }, { "grad_norm": 0.21204911172389984, "learning_rate": 7.03625077760516e-05, "loss": 0.0196, "step": 119450 }, { "grad_norm": 0.11006921529769897, "learning_rate": 7.035747384727764e-05, "loss": 0.0174, "step": 119460 }, { "grad_norm": 0.0969458743929863, "learning_rate": 7.035243967114114e-05, "loss": 0.0163, "step": 119470 }, { "grad_norm": 0.13469435274600983, "learning_rate": 7.03474052477033e-05, "loss": 0.0143, "step": 119480 }, { "grad_norm": 0.15168145298957825, "learning_rate": 7.034237057702525e-05, "loss": 0.0191, "step": 119490 }, { "grad_norm": 0.10476918518543243, "learning_rate": 7.033733565916821e-05, "loss": 0.0182, "step": 119500 }, { "grad_norm": 0.09214913100004196, "learning_rate": 7.033230049419333e-05, "loss": 0.0162, "step": 119510 }, { "grad_norm": 0.13959629833698273, "learning_rate": 7.032726508216182e-05, "loss": 0.018, "step": 119520 }, { "grad_norm": 0.1627308577299118, "learning_rate": 7.032222942313482e-05, "loss": 0.0142, "step": 119530 }, { "grad_norm": 0.14888979494571686, "learning_rate": 7.031719351717357e-05, "loss": 0.017, "step": 119540 }, { "grad_norm": 0.14174577593803406, "learning_rate": 7.031215736433922e-05, "loss": 0.0157, "step": 119550 }, { "grad_norm": 0.13500717282295227, "learning_rate": 7.030712096469299e-05, "loss": 0.0136, "step": 119560 }, { "grad_norm": 0.22905850410461426, "learning_rate": 7.030208431829607e-05, "loss": 0.0174, "step": 119570 }, { "grad_norm": 0.09290327876806259, "learning_rate": 7.029704742520966e-05, "loss": 0.016, "step": 119580 }, { "grad_norm": 0.09536261856555939, "learning_rate": 7.029201028549493e-05, "loss": 0.0188, "step": 119590 }, { "grad_norm": 0.23204471170902252, "learning_rate": 7.028697289921316e-05, "loss": 0.023, "step": 119600 }, { "grad_norm": 0.10723324120044708, "learning_rate": 7.028193526642548e-05, "loss": 0.0199, "step": 119610 }, { "grad_norm": 0.1141846776008606, "learning_rate": 7.027689738719315e-05, "loss": 0.0153, "step": 119620 }, { "grad_norm": 0.18506188690662384, "learning_rate": 7.027185926157738e-05, "loss": 0.0143, "step": 119630 }, { "grad_norm": 0.18674969673156738, "learning_rate": 7.026682088963938e-05, "loss": 0.02, "step": 119640 }, { "grad_norm": 0.11904570460319519, "learning_rate": 7.026178227144036e-05, "loss": 0.0154, "step": 119650 }, { "grad_norm": 0.1143159493803978, "learning_rate": 7.025674340704156e-05, "loss": 0.0132, "step": 119660 }, { "grad_norm": 0.13319022953510284, "learning_rate": 7.025170429650419e-05, "loss": 0.0159, "step": 119670 }, { "grad_norm": 0.1673375368118286, "learning_rate": 7.02466649398895e-05, "loss": 0.019, "step": 119680 }, { "grad_norm": 0.11745933443307877, "learning_rate": 7.024162533725873e-05, "loss": 0.018, "step": 119690 }, { "grad_norm": 0.2616868317127228, "learning_rate": 7.023658548867308e-05, "loss": 0.0187, "step": 119700 }, { "grad_norm": 0.16325658559799194, "learning_rate": 7.023154539419384e-05, "loss": 0.0198, "step": 119710 }, { "grad_norm": 0.140799418091774, "learning_rate": 7.02265050538822e-05, "loss": 0.0212, "step": 119720 }, { "grad_norm": 0.13648301362991333, "learning_rate": 7.022146446779942e-05, "loss": 0.0161, "step": 119730 }, { "grad_norm": 0.1644069105386734, "learning_rate": 7.021642363600676e-05, "loss": 0.0172, "step": 119740 }, { "grad_norm": 0.15894679725170135, "learning_rate": 7.021138255856547e-05, "loss": 0.0123, "step": 119750 }, { "grad_norm": 0.17598913609981537, "learning_rate": 7.02063412355368e-05, "loss": 0.0158, "step": 119760 }, { "grad_norm": 0.2594430148601532, "learning_rate": 7.020129966698198e-05, "loss": 0.0174, "step": 119770 }, { "grad_norm": 0.24962632358074188, "learning_rate": 7.019625785296233e-05, "loss": 0.013, "step": 119780 }, { "grad_norm": 0.17447826266288757, "learning_rate": 7.019121579353906e-05, "loss": 0.0141, "step": 119790 }, { "grad_norm": 0.1189214363694191, "learning_rate": 7.018617348877346e-05, "loss": 0.0186, "step": 119800 }, { "grad_norm": 0.7171754837036133, "learning_rate": 7.01811309387268e-05, "loss": 0.0164, "step": 119810 }, { "grad_norm": 0.11184637993574142, "learning_rate": 7.017608814346035e-05, "loss": 0.0135, "step": 119820 }, { "grad_norm": 0.19455017149448395, "learning_rate": 7.017104510303536e-05, "loss": 0.0188, "step": 119830 }, { "grad_norm": 0.2296992540359497, "learning_rate": 7.016600181751313e-05, "loss": 0.0125, "step": 119840 }, { "grad_norm": 0.1148526668548584, "learning_rate": 7.016095828695496e-05, "loss": 0.0148, "step": 119850 }, { "grad_norm": 0.1669379323720932, "learning_rate": 7.01559145114221e-05, "loss": 0.0159, "step": 119860 }, { "grad_norm": 0.2536166310310364, "learning_rate": 7.015087049097584e-05, "loss": 0.0167, "step": 119870 }, { "grad_norm": 0.15485073626041412, "learning_rate": 7.014582622567747e-05, "loss": 0.016, "step": 119880 }, { "grad_norm": 0.11067573726177216, "learning_rate": 7.014078171558831e-05, "loss": 0.018, "step": 119890 }, { "grad_norm": 0.1518079936504364, "learning_rate": 7.013573696076964e-05, "loss": 0.0189, "step": 119900 }, { "grad_norm": 0.7447783350944519, "learning_rate": 7.013069196128274e-05, "loss": 0.0216, "step": 119910 }, { "grad_norm": 0.09296485781669617, "learning_rate": 7.012564671718893e-05, "loss": 0.0148, "step": 119920 }, { "grad_norm": 0.10631918907165527, "learning_rate": 7.01206012285495e-05, "loss": 0.0204, "step": 119930 }, { "grad_norm": 0.1189119815826416, "learning_rate": 7.011555549542578e-05, "loss": 0.0147, "step": 119940 }, { "grad_norm": 0.13859957456588745, "learning_rate": 7.011050951787908e-05, "loss": 0.0136, "step": 119950 }, { "grad_norm": 0.18613122403621674, "learning_rate": 7.010546329597069e-05, "loss": 0.0251, "step": 119960 }, { "grad_norm": 0.16482537984848022, "learning_rate": 7.010041682976193e-05, "loss": 0.0213, "step": 119970 }, { "grad_norm": 0.10501295328140259, "learning_rate": 7.009537011931413e-05, "loss": 0.0166, "step": 119980 }, { "grad_norm": 0.3162990212440491, "learning_rate": 7.009032316468861e-05, "loss": 0.021, "step": 119990 }, { "grad_norm": 0.1391402781009674, "learning_rate": 7.00852759659467e-05, "loss": 0.019, "step": 120000 }, { "grad_norm": 0.08242897689342499, "learning_rate": 7.008022852314974e-05, "loss": 0.0207, "step": 120010 }, { "grad_norm": 0.13311389088630676, "learning_rate": 7.007518083635903e-05, "loss": 0.026, "step": 120020 }, { "grad_norm": 0.09112399816513062, "learning_rate": 7.00701329056359e-05, "loss": 0.0194, "step": 120030 }, { "grad_norm": 0.4473228454589844, "learning_rate": 7.006508473104174e-05, "loss": 0.0184, "step": 120040 }, { "grad_norm": 0.08733563870191574, "learning_rate": 7.006003631263784e-05, "loss": 0.0186, "step": 120050 }, { "grad_norm": 0.1699923872947693, "learning_rate": 7.005498765048554e-05, "loss": 0.0154, "step": 120060 }, { "grad_norm": 0.2054908275604248, "learning_rate": 7.004993874464623e-05, "loss": 0.0156, "step": 120070 }, { "grad_norm": 0.1344425082206726, "learning_rate": 7.004488959518123e-05, "loss": 0.0155, "step": 120080 }, { "grad_norm": 0.1573779582977295, "learning_rate": 7.003984020215188e-05, "loss": 0.0177, "step": 120090 }, { "grad_norm": 0.15273398160934448, "learning_rate": 7.003479056561957e-05, "loss": 0.0153, "step": 120100 }, { "grad_norm": 0.08266624063253403, "learning_rate": 7.002974068564563e-05, "loss": 0.0163, "step": 120110 }, { "grad_norm": 0.2482222616672516, "learning_rate": 7.002469056229144e-05, "loss": 0.0146, "step": 120120 }, { "grad_norm": 0.14563976228237152, "learning_rate": 7.001964019561836e-05, "loss": 0.0219, "step": 120130 }, { "grad_norm": 0.15062512457370758, "learning_rate": 7.001458958568772e-05, "loss": 0.0206, "step": 120140 }, { "grad_norm": 0.09548745304346085, "learning_rate": 7.000953873256094e-05, "loss": 0.0182, "step": 120150 }, { "grad_norm": 0.1405588686466217, "learning_rate": 7.000448763629936e-05, "loss": 0.0148, "step": 120160 }, { "grad_norm": 0.10422684252262115, "learning_rate": 6.999943629696437e-05, "loss": 0.0129, "step": 120170 }, { "grad_norm": 0.13178719580173492, "learning_rate": 6.999438471461735e-05, "loss": 0.0165, "step": 120180 }, { "grad_norm": 0.12800918519496918, "learning_rate": 6.998933288931968e-05, "loss": 0.0188, "step": 120190 }, { "grad_norm": 0.10144089162349701, "learning_rate": 6.998428082113273e-05, "loss": 0.0133, "step": 120200 }, { "grad_norm": 0.16327816247940063, "learning_rate": 6.997922851011792e-05, "loss": 0.0185, "step": 120210 }, { "grad_norm": 0.09562528878450394, "learning_rate": 6.997417595633661e-05, "loss": 0.0165, "step": 120220 }, { "grad_norm": 0.10287202149629593, "learning_rate": 6.99691231598502e-05, "loss": 0.0209, "step": 120230 }, { "grad_norm": 0.17717455327510834, "learning_rate": 6.99640701207201e-05, "loss": 0.0149, "step": 120240 }, { "grad_norm": 0.1414141207933426, "learning_rate": 6.99590168390077e-05, "loss": 0.0157, "step": 120250 }, { "grad_norm": 0.15470367670059204, "learning_rate": 6.995396331477437e-05, "loss": 0.0155, "step": 120260 }, { "grad_norm": 0.15196721255779266, "learning_rate": 6.994890954808157e-05, "loss": 0.0187, "step": 120270 }, { "grad_norm": 0.144920215010643, "learning_rate": 6.994385553899069e-05, "loss": 0.0145, "step": 120280 }, { "grad_norm": 0.1840793937444687, "learning_rate": 6.993880128756314e-05, "loss": 0.0173, "step": 120290 }, { "grad_norm": 0.37822917103767395, "learning_rate": 6.993374679386031e-05, "loss": 0.0179, "step": 120300 }, { "grad_norm": 0.09363466501235962, "learning_rate": 6.992869205794365e-05, "loss": 0.0163, "step": 120310 }, { "grad_norm": 0.26904305815696716, "learning_rate": 6.992363707987455e-05, "loss": 0.0146, "step": 120320 }, { "grad_norm": 0.11889996379613876, "learning_rate": 6.991858185971447e-05, "loss": 0.0146, "step": 120330 }, { "grad_norm": 0.12832990288734436, "learning_rate": 6.991352639752479e-05, "loss": 0.0229, "step": 120340 }, { "grad_norm": 0.10125644505023956, "learning_rate": 6.990847069336698e-05, "loss": 0.0155, "step": 120350 }, { "grad_norm": 0.10211411118507385, "learning_rate": 6.990341474730244e-05, "loss": 0.0113, "step": 120360 }, { "grad_norm": 0.09303683787584305, "learning_rate": 6.989835855939263e-05, "loss": 0.017, "step": 120370 }, { "grad_norm": 0.13458169996738434, "learning_rate": 6.989330212969898e-05, "loss": 0.0151, "step": 120380 }, { "grad_norm": 0.11991986632347107, "learning_rate": 6.988824545828292e-05, "loss": 0.0179, "step": 120390 }, { "grad_norm": 0.11830010265111923, "learning_rate": 6.98831885452059e-05, "loss": 0.0196, "step": 120400 }, { "grad_norm": 0.12479987740516663, "learning_rate": 6.987813139052937e-05, "loss": 0.0159, "step": 120410 }, { "grad_norm": 0.13849449157714844, "learning_rate": 6.987307399431477e-05, "loss": 0.0137, "step": 120420 }, { "grad_norm": 0.10167132318019867, "learning_rate": 6.986801635662356e-05, "loss": 0.0179, "step": 120430 }, { "grad_norm": 0.09448028355836868, "learning_rate": 6.986295847751719e-05, "loss": 0.0121, "step": 120440 }, { "grad_norm": 0.14144913852214813, "learning_rate": 6.985790035705713e-05, "loss": 0.0226, "step": 120450 }, { "grad_norm": 0.21246567368507385, "learning_rate": 6.985284199530481e-05, "loss": 0.0211, "step": 120460 }, { "grad_norm": 0.09249133616685867, "learning_rate": 6.984778339232174e-05, "loss": 0.0142, "step": 120470 }, { "grad_norm": 0.09361909329891205, "learning_rate": 6.984272454816936e-05, "loss": 0.017, "step": 120480 }, { "grad_norm": 0.21310006082057953, "learning_rate": 6.983766546290913e-05, "loss": 0.0209, "step": 120490 }, { "grad_norm": 0.15156994760036469, "learning_rate": 6.983260613660254e-05, "loss": 0.0212, "step": 120500 }, { "grad_norm": 0.10235602408647537, "learning_rate": 6.982754656931106e-05, "loss": 0.0195, "step": 120510 }, { "grad_norm": 0.21144361793994904, "learning_rate": 6.982248676109619e-05, "loss": 0.0196, "step": 120520 }, { "grad_norm": 0.0755293071269989, "learning_rate": 6.981742671201936e-05, "loss": 0.0111, "step": 120530 }, { "grad_norm": 0.17341166734695435, "learning_rate": 6.98123664221421e-05, "loss": 0.0155, "step": 120540 }, { "grad_norm": 0.09497971832752228, "learning_rate": 6.980730589152586e-05, "loss": 0.0219, "step": 120550 }, { "grad_norm": 0.11494549363851547, "learning_rate": 6.980224512023218e-05, "loss": 0.0142, "step": 120560 }, { "grad_norm": 0.18196070194244385, "learning_rate": 6.979718410832251e-05, "loss": 0.0138, "step": 120570 }, { "grad_norm": 0.10139264166355133, "learning_rate": 6.979212285585837e-05, "loss": 0.0195, "step": 120580 }, { "grad_norm": 0.2552188038825989, "learning_rate": 6.978706136290125e-05, "loss": 0.02, "step": 120590 }, { "grad_norm": 0.15551190078258514, "learning_rate": 6.978199962951264e-05, "loss": 0.0156, "step": 120600 }, { "grad_norm": 0.1948012262582779, "learning_rate": 6.977693765575407e-05, "loss": 0.0162, "step": 120610 }, { "grad_norm": 0.09508196264505386, "learning_rate": 6.977187544168703e-05, "loss": 0.0152, "step": 120620 }, { "grad_norm": 0.1434468924999237, "learning_rate": 6.976681298737304e-05, "loss": 0.02, "step": 120630 }, { "grad_norm": 0.10465627908706665, "learning_rate": 6.976175029287359e-05, "loss": 0.0161, "step": 120640 }, { "grad_norm": 0.14621439576148987, "learning_rate": 6.975668735825025e-05, "loss": 0.0164, "step": 120650 }, { "grad_norm": 1.4821815490722656, "learning_rate": 6.975162418356448e-05, "loss": 0.0197, "step": 120660 }, { "grad_norm": 0.13498972356319427, "learning_rate": 6.974656076887783e-05, "loss": 0.0164, "step": 120670 }, { "grad_norm": 0.23019874095916748, "learning_rate": 6.974149711425184e-05, "loss": 0.0198, "step": 120680 }, { "grad_norm": 0.21685335040092468, "learning_rate": 6.9736433219748e-05, "loss": 0.026, "step": 120690 }, { "grad_norm": 0.1673615425825119, "learning_rate": 6.973136908542788e-05, "loss": 0.02, "step": 120700 }, { "grad_norm": 0.15315291285514832, "learning_rate": 6.972630471135297e-05, "loss": 0.0184, "step": 120710 }, { "grad_norm": 0.2522125244140625, "learning_rate": 6.972124009758486e-05, "loss": 0.0191, "step": 120720 }, { "grad_norm": 0.20588496327400208, "learning_rate": 6.971617524418504e-05, "loss": 0.0178, "step": 120730 }, { "grad_norm": 0.5465681552886963, "learning_rate": 6.971111015121509e-05, "loss": 0.0191, "step": 120740 }, { "grad_norm": 0.22228385508060455, "learning_rate": 6.970604481873654e-05, "loss": 0.0155, "step": 120750 }, { "grad_norm": 0.1442815065383911, "learning_rate": 6.970097924681095e-05, "loss": 0.0161, "step": 120760 }, { "grad_norm": 0.13059945404529572, "learning_rate": 6.969591343549985e-05, "loss": 0.0179, "step": 120770 }, { "grad_norm": 0.12137098610401154, "learning_rate": 6.96908473848648e-05, "loss": 0.0121, "step": 120780 }, { "grad_norm": 0.10984982550144196, "learning_rate": 6.968578109496737e-05, "loss": 0.0164, "step": 120790 }, { "grad_norm": 0.13714420795440674, "learning_rate": 6.96807145658691e-05, "loss": 0.0206, "step": 120800 }, { "grad_norm": 0.08787361532449722, "learning_rate": 6.967564779763159e-05, "loss": 0.0157, "step": 120810 }, { "grad_norm": 0.19531606137752533, "learning_rate": 6.967058079031637e-05, "loss": 0.0155, "step": 120820 }, { "grad_norm": 0.17047905921936035, "learning_rate": 6.966551354398502e-05, "loss": 0.0175, "step": 120830 }, { "grad_norm": 0.1914534568786621, "learning_rate": 6.966044605869913e-05, "loss": 0.0118, "step": 120840 }, { "grad_norm": 0.11253592371940613, "learning_rate": 6.965537833452024e-05, "loss": 0.0142, "step": 120850 }, { "grad_norm": 0.1452282965183258, "learning_rate": 6.965031037150996e-05, "loss": 0.0203, "step": 120860 }, { "grad_norm": 0.0802844911813736, "learning_rate": 6.964524216972983e-05, "loss": 0.011, "step": 120870 }, { "grad_norm": 0.1864681988954544, "learning_rate": 6.964017372924147e-05, "loss": 0.0243, "step": 120880 }, { "grad_norm": 0.1020885482430458, "learning_rate": 6.963510505010646e-05, "loss": 0.0211, "step": 120890 }, { "grad_norm": 0.12659956514835358, "learning_rate": 6.963003613238639e-05, "loss": 0.0182, "step": 120900 }, { "grad_norm": 0.17160077393054962, "learning_rate": 6.962496697614283e-05, "loss": 0.0141, "step": 120910 }, { "grad_norm": 0.1425762176513672, "learning_rate": 6.961989758143741e-05, "loss": 0.0177, "step": 120920 }, { "grad_norm": 0.08576346933841705, "learning_rate": 6.961482794833171e-05, "loss": 0.0157, "step": 120930 }, { "grad_norm": 0.09994746744632721, "learning_rate": 6.960975807688732e-05, "loss": 0.0197, "step": 120940 }, { "grad_norm": 0.18871183693408966, "learning_rate": 6.960468796716585e-05, "loss": 0.0108, "step": 120950 }, { "grad_norm": 0.17459198832511902, "learning_rate": 6.959961761922894e-05, "loss": 0.0147, "step": 120960 }, { "grad_norm": 0.09617377817630768, "learning_rate": 6.959454703313815e-05, "loss": 0.0192, "step": 120970 }, { "grad_norm": 0.2442551851272583, "learning_rate": 6.958947620895511e-05, "loss": 0.0132, "step": 120980 }, { "grad_norm": 0.08472845703363419, "learning_rate": 6.958440514674144e-05, "loss": 0.0118, "step": 120990 }, { "grad_norm": 0.07569492608308792, "learning_rate": 6.957933384655877e-05, "loss": 0.0175, "step": 121000 }, { "grad_norm": 0.07835878431797028, "learning_rate": 6.957426230846868e-05, "loss": 0.0116, "step": 121010 }, { "grad_norm": 0.11622819304466248, "learning_rate": 6.956919053253286e-05, "loss": 0.0179, "step": 121020 }, { "grad_norm": 0.2373303472995758, "learning_rate": 6.956411851881287e-05, "loss": 0.0198, "step": 121030 }, { "grad_norm": 0.07023061066865921, "learning_rate": 6.955904626737038e-05, "loss": 0.0211, "step": 121040 }, { "grad_norm": 0.149479478597641, "learning_rate": 6.9553973778267e-05, "loss": 0.0142, "step": 121050 }, { "grad_norm": 0.1929446905851364, "learning_rate": 6.95489010515644e-05, "loss": 0.0222, "step": 121060 }, { "grad_norm": 0.12416782975196838, "learning_rate": 6.954382808732418e-05, "loss": 0.02, "step": 121070 }, { "grad_norm": 0.21791523694992065, "learning_rate": 6.953875488560799e-05, "loss": 0.0177, "step": 121080 }, { "grad_norm": 0.12864996492862701, "learning_rate": 6.953368144647749e-05, "loss": 0.0148, "step": 121090 }, { "grad_norm": 0.13496766984462738, "learning_rate": 6.95286077699943e-05, "loss": 0.0204, "step": 121100 }, { "grad_norm": 0.2048039585351944, "learning_rate": 6.952353385622011e-05, "loss": 0.0188, "step": 121110 }, { "grad_norm": 0.08111954480409622, "learning_rate": 6.951845970521654e-05, "loss": 0.0154, "step": 121120 }, { "grad_norm": 0.20865483582019806, "learning_rate": 6.951338531704526e-05, "loss": 0.0156, "step": 121130 }, { "grad_norm": 0.13773754239082336, "learning_rate": 6.950831069176793e-05, "loss": 0.0145, "step": 121140 }, { "grad_norm": 0.09003570675849915, "learning_rate": 6.95032358294462e-05, "loss": 0.0153, "step": 121150 }, { "grad_norm": 0.2562515139579773, "learning_rate": 6.949816073014174e-05, "loss": 0.0155, "step": 121160 }, { "grad_norm": 0.13514287769794464, "learning_rate": 6.949308539391622e-05, "loss": 0.0149, "step": 121170 }, { "grad_norm": 0.13982900977134705, "learning_rate": 6.94880098208313e-05, "loss": 0.0143, "step": 121180 }, { "grad_norm": 0.1535017490386963, "learning_rate": 6.948293401094866e-05, "loss": 0.0161, "step": 121190 }, { "grad_norm": 0.14544177055358887, "learning_rate": 6.947785796432999e-05, "loss": 0.0139, "step": 121200 }, { "grad_norm": 0.10636109113693237, "learning_rate": 6.947278168103694e-05, "loss": 0.0176, "step": 121210 }, { "grad_norm": 0.14284953474998474, "learning_rate": 6.946770516113122e-05, "loss": 0.0199, "step": 121220 }, { "grad_norm": 0.1371656060218811, "learning_rate": 6.94626284046745e-05, "loss": 0.0219, "step": 121230 }, { "grad_norm": 0.4668113589286804, "learning_rate": 6.945755141172848e-05, "loss": 0.0193, "step": 121240 }, { "grad_norm": 0.0783737450838089, "learning_rate": 6.945247418235482e-05, "loss": 0.017, "step": 121250 }, { "grad_norm": 0.3202158510684967, "learning_rate": 6.944739671661523e-05, "loss": 0.0209, "step": 121260 }, { "grad_norm": 0.1705755591392517, "learning_rate": 6.944231901457144e-05, "loss": 0.0177, "step": 121270 }, { "grad_norm": 0.16003791987895966, "learning_rate": 6.943724107628508e-05, "loss": 0.0128, "step": 121280 }, { "grad_norm": 0.3019654452800751, "learning_rate": 6.943216290181791e-05, "loss": 0.015, "step": 121290 }, { "grad_norm": 0.13630744814872742, "learning_rate": 6.94270844912316e-05, "loss": 0.0207, "step": 121300 }, { "grad_norm": 0.24310365319252014, "learning_rate": 6.942200584458788e-05, "loss": 0.0164, "step": 121310 }, { "grad_norm": 0.2528828978538513, "learning_rate": 6.941692696194846e-05, "loss": 0.0196, "step": 121320 }, { "grad_norm": 0.10683897137641907, "learning_rate": 6.941184784337502e-05, "loss": 0.0204, "step": 121330 }, { "grad_norm": 0.1668747365474701, "learning_rate": 6.940676848892933e-05, "loss": 0.0158, "step": 121340 }, { "grad_norm": 0.13418595492839813, "learning_rate": 6.940168889867306e-05, "loss": 0.0157, "step": 121350 }, { "grad_norm": 0.15947994589805603, "learning_rate": 6.939660907266796e-05, "loss": 0.0163, "step": 121360 }, { "grad_norm": 0.17886830866336823, "learning_rate": 6.939152901097575e-05, "loss": 0.0175, "step": 121370 }, { "grad_norm": 0.08542542159557343, "learning_rate": 6.938644871365814e-05, "loss": 0.0161, "step": 121380 }, { "grad_norm": 0.06202961876988411, "learning_rate": 6.938136818077687e-05, "loss": 0.0144, "step": 121390 }, { "grad_norm": 0.09902871400117874, "learning_rate": 6.93762874123937e-05, "loss": 0.0171, "step": 121400 }, { "grad_norm": 0.11438601464033127, "learning_rate": 6.937120640857034e-05, "loss": 0.012, "step": 121410 }, { "grad_norm": 0.07252556830644608, "learning_rate": 6.936612516936852e-05, "loss": 0.0149, "step": 121420 }, { "grad_norm": 0.14293664693832397, "learning_rate": 6.936104369485e-05, "loss": 0.0132, "step": 121430 }, { "grad_norm": 0.14523206651210785, "learning_rate": 6.935596198507652e-05, "loss": 0.0178, "step": 121440 }, { "grad_norm": 0.1901342272758484, "learning_rate": 6.935088004010983e-05, "loss": 0.0151, "step": 121450 }, { "grad_norm": 0.14237643778324127, "learning_rate": 6.934579786001167e-05, "loss": 0.0218, "step": 121460 }, { "grad_norm": 0.11401377618312836, "learning_rate": 6.93407154448438e-05, "loss": 0.0144, "step": 121470 }, { "grad_norm": 0.20621301233768463, "learning_rate": 6.9335632794668e-05, "loss": 0.0195, "step": 121480 }, { "grad_norm": 0.20702393352985382, "learning_rate": 6.933054990954597e-05, "loss": 0.0224, "step": 121490 }, { "grad_norm": 0.13935822248458862, "learning_rate": 6.932546678953953e-05, "loss": 0.0168, "step": 121500 }, { "grad_norm": 0.21625418961048126, "learning_rate": 6.932038343471041e-05, "loss": 0.0224, "step": 121510 }, { "grad_norm": 0.17579999566078186, "learning_rate": 6.931529984512041e-05, "loss": 0.0186, "step": 121520 }, { "grad_norm": 0.15038739144802094, "learning_rate": 6.931021602083125e-05, "loss": 0.0199, "step": 121530 }, { "grad_norm": 0.23369938135147095, "learning_rate": 6.930513196190477e-05, "loss": 0.0148, "step": 121540 }, { "grad_norm": 0.12391198426485062, "learning_rate": 6.930004766840268e-05, "loss": 0.0252, "step": 121550 }, { "grad_norm": 0.1468130499124527, "learning_rate": 6.92949631403868e-05, "loss": 0.0169, "step": 121560 }, { "grad_norm": 0.10816393792629242, "learning_rate": 6.928987837791891e-05, "loss": 0.0221, "step": 121570 }, { "grad_norm": 0.14622187614440918, "learning_rate": 6.928479338106077e-05, "loss": 0.0162, "step": 121580 }, { "grad_norm": 0.12386857718229294, "learning_rate": 6.927970814987419e-05, "loss": 0.0187, "step": 121590 }, { "grad_norm": 0.15327903628349304, "learning_rate": 6.927462268442093e-05, "loss": 0.0139, "step": 121600 }, { "grad_norm": 0.39043933153152466, "learning_rate": 6.926953698476284e-05, "loss": 0.0169, "step": 121610 }, { "grad_norm": 0.17007014155387878, "learning_rate": 6.926445105096167e-05, "loss": 0.0169, "step": 121620 }, { "grad_norm": 0.13270975649356842, "learning_rate": 6.925936488307923e-05, "loss": 0.0183, "step": 121630 }, { "grad_norm": 0.11530657857656479, "learning_rate": 6.925427848117731e-05, "loss": 0.0161, "step": 121640 }, { "grad_norm": 0.209110826253891, "learning_rate": 6.924919184531774e-05, "loss": 0.0161, "step": 121650 }, { "grad_norm": 0.1950828731060028, "learning_rate": 6.92441049755623e-05, "loss": 0.0159, "step": 121660 }, { "grad_norm": 0.17217831313610077, "learning_rate": 6.923901787197284e-05, "loss": 0.0214, "step": 121670 }, { "grad_norm": 0.16160394251346588, "learning_rate": 6.923393053461114e-05, "loss": 0.0171, "step": 121680 }, { "grad_norm": 0.1652863770723343, "learning_rate": 6.922884296353901e-05, "loss": 0.0173, "step": 121690 }, { "grad_norm": 0.15597176551818848, "learning_rate": 6.922375515881829e-05, "loss": 0.0105, "step": 121700 }, { "grad_norm": 0.08428250253200531, "learning_rate": 6.921866712051081e-05, "loss": 0.0194, "step": 121710 }, { "grad_norm": 0.0700620785355568, "learning_rate": 6.921357884867836e-05, "loss": 0.0128, "step": 121720 }, { "grad_norm": 0.13836292922496796, "learning_rate": 6.92084903433828e-05, "loss": 0.018, "step": 121730 }, { "grad_norm": 0.12057065963745117, "learning_rate": 6.920340160468593e-05, "loss": 0.0179, "step": 121740 }, { "grad_norm": 0.23449374735355377, "learning_rate": 6.919831263264962e-05, "loss": 0.0254, "step": 121750 }, { "grad_norm": 0.14728587865829468, "learning_rate": 6.919322342733567e-05, "loss": 0.0131, "step": 121760 }, { "grad_norm": 0.12008197605609894, "learning_rate": 6.918813398880595e-05, "loss": 0.0214, "step": 121770 }, { "grad_norm": 0.1921464055776596, "learning_rate": 6.918304431712227e-05, "loss": 0.024, "step": 121780 }, { "grad_norm": 0.10597430169582367, "learning_rate": 6.91779544123465e-05, "loss": 0.0124, "step": 121790 }, { "grad_norm": 0.3320605158805847, "learning_rate": 6.917286427454047e-05, "loss": 0.017, "step": 121800 }, { "grad_norm": 0.287436306476593, "learning_rate": 6.916777390376603e-05, "loss": 0.0178, "step": 121810 }, { "grad_norm": 0.22017982602119446, "learning_rate": 6.916268330008507e-05, "loss": 0.0145, "step": 121820 }, { "grad_norm": 0.13178092241287231, "learning_rate": 6.915759246355937e-05, "loss": 0.0184, "step": 121830 }, { "grad_norm": 0.16608691215515137, "learning_rate": 6.915250139425087e-05, "loss": 0.0223, "step": 121840 }, { "grad_norm": 0.13833022117614746, "learning_rate": 6.914741009222138e-05, "loss": 0.0244, "step": 121850 }, { "grad_norm": 0.16713595390319824, "learning_rate": 6.914231855753278e-05, "loss": 0.016, "step": 121860 }, { "grad_norm": 0.36711248755455017, "learning_rate": 6.913722679024693e-05, "loss": 0.0196, "step": 121870 }, { "grad_norm": 0.15109600126743317, "learning_rate": 6.913213479042572e-05, "loss": 0.0165, "step": 121880 }, { "grad_norm": 0.1523696333169937, "learning_rate": 6.9127042558131e-05, "loss": 0.0173, "step": 121890 }, { "grad_norm": 0.4980040490627289, "learning_rate": 6.912195009342466e-05, "loss": 0.0179, "step": 121900 }, { "grad_norm": 0.12037355452775955, "learning_rate": 6.911685739636858e-05, "loss": 0.0212, "step": 121910 }, { "grad_norm": 0.10817331075668335, "learning_rate": 6.91117644670246e-05, "loss": 0.0152, "step": 121920 }, { "grad_norm": 0.2022840976715088, "learning_rate": 6.910667130545468e-05, "loss": 0.0205, "step": 121930 }, { "grad_norm": 0.143341064453125, "learning_rate": 6.910157791172062e-05, "loss": 0.0167, "step": 121940 }, { "grad_norm": 0.31499436497688293, "learning_rate": 6.909648428588439e-05, "loss": 0.0143, "step": 121950 }, { "grad_norm": 0.27026692032814026, "learning_rate": 6.909139042800783e-05, "loss": 0.0198, "step": 121960 }, { "grad_norm": 0.10903982073068619, "learning_rate": 6.908629633815285e-05, "loss": 0.0134, "step": 121970 }, { "grad_norm": 0.1447744071483612, "learning_rate": 6.908120201638136e-05, "loss": 0.0193, "step": 121980 }, { "grad_norm": 0.054081209003925323, "learning_rate": 6.907610746275523e-05, "loss": 0.0159, "step": 121990 }, { "grad_norm": 0.20463868975639343, "learning_rate": 6.907101267733641e-05, "loss": 0.0131, "step": 122000 }, { "grad_norm": 0.10384928435087204, "learning_rate": 6.906591766018675e-05, "loss": 0.0241, "step": 122010 }, { "grad_norm": 0.12396865338087082, "learning_rate": 6.906082241136822e-05, "loss": 0.0136, "step": 122020 }, { "grad_norm": 0.173477441072464, "learning_rate": 6.905572693094268e-05, "loss": 0.0163, "step": 122030 }, { "grad_norm": 0.2543976902961731, "learning_rate": 6.905063121897207e-05, "loss": 0.0134, "step": 122040 }, { "grad_norm": 0.05930526554584503, "learning_rate": 6.904553527551831e-05, "loss": 0.0126, "step": 122050 }, { "grad_norm": 0.07424304634332657, "learning_rate": 6.90404391006433e-05, "loss": 0.0149, "step": 122060 }, { "grad_norm": 0.22513476014137268, "learning_rate": 6.903534269440902e-05, "loss": 0.0113, "step": 122070 }, { "grad_norm": 0.19882482290267944, "learning_rate": 6.90302460568773e-05, "loss": 0.0144, "step": 122080 }, { "grad_norm": 0.08651745319366455, "learning_rate": 6.902514918811015e-05, "loss": 0.0189, "step": 122090 }, { "grad_norm": 0.377521276473999, "learning_rate": 6.902005208816946e-05, "loss": 0.0217, "step": 122100 }, { "grad_norm": 0.21729470789432526, "learning_rate": 6.90149547571172e-05, "loss": 0.0163, "step": 122110 }, { "grad_norm": 0.20541195571422577, "learning_rate": 6.900985719501528e-05, "loss": 0.0187, "step": 122120 }, { "grad_norm": 0.18868578970432281, "learning_rate": 6.900475940192564e-05, "loss": 0.0142, "step": 122130 }, { "grad_norm": 0.7144052982330322, "learning_rate": 6.899966137791022e-05, "loss": 0.0183, "step": 122140 }, { "grad_norm": 0.103177510201931, "learning_rate": 6.8994563123031e-05, "loss": 0.0126, "step": 122150 }, { "grad_norm": 0.10739067941904068, "learning_rate": 6.89894646373499e-05, "loss": 0.0195, "step": 122160 }, { "grad_norm": 0.10791165381669998, "learning_rate": 6.898436592092885e-05, "loss": 0.014, "step": 122170 }, { "grad_norm": 0.22430667281150818, "learning_rate": 6.897926697382986e-05, "loss": 0.0132, "step": 122180 }, { "grad_norm": 0.1991119384765625, "learning_rate": 6.897416779611485e-05, "loss": 0.0176, "step": 122190 }, { "grad_norm": 0.08128982037305832, "learning_rate": 6.896906838784577e-05, "loss": 0.0168, "step": 122200 }, { "grad_norm": 0.14453287422657013, "learning_rate": 6.896396874908462e-05, "loss": 0.0153, "step": 122210 }, { "grad_norm": 0.0898350179195404, "learning_rate": 6.895886887989334e-05, "loss": 0.0126, "step": 122220 }, { "grad_norm": 0.13476237654685974, "learning_rate": 6.89537687803339e-05, "loss": 0.0137, "step": 122230 }, { "grad_norm": 0.24121986329555511, "learning_rate": 6.894866845046825e-05, "loss": 0.0167, "step": 122240 }, { "grad_norm": 0.23987890779972076, "learning_rate": 6.894356789035843e-05, "loss": 0.0143, "step": 122250 }, { "grad_norm": 1.0649800300598145, "learning_rate": 6.893846710006635e-05, "loss": 0.0184, "step": 122260 }, { "grad_norm": 0.14784519374370575, "learning_rate": 6.893336607965402e-05, "loss": 0.0198, "step": 122270 }, { "grad_norm": 0.1536182016134262, "learning_rate": 6.892826482918342e-05, "loss": 0.0175, "step": 122280 }, { "grad_norm": 0.23378567397594452, "learning_rate": 6.892316334871652e-05, "loss": 0.023, "step": 122290 }, { "grad_norm": 0.23469476401805878, "learning_rate": 6.891806163831532e-05, "loss": 0.0246, "step": 122300 }, { "grad_norm": 0.14894816279411316, "learning_rate": 6.891295969804181e-05, "loss": 0.0181, "step": 122310 }, { "grad_norm": 0.11855567991733551, "learning_rate": 6.8907857527958e-05, "loss": 0.0166, "step": 122320 }, { "grad_norm": 0.1425321400165558, "learning_rate": 6.890275512812584e-05, "loss": 0.0198, "step": 122330 }, { "grad_norm": 0.16891489923000336, "learning_rate": 6.889765249860738e-05, "loss": 0.031, "step": 122340 }, { "grad_norm": 0.11058282852172852, "learning_rate": 6.889254963946459e-05, "loss": 0.0142, "step": 122350 }, { "grad_norm": 0.0839417427778244, "learning_rate": 6.88874465507595e-05, "loss": 0.0185, "step": 122360 }, { "grad_norm": 0.15244536101818085, "learning_rate": 6.888234323255408e-05, "loss": 0.0185, "step": 122370 }, { "grad_norm": 0.06773164868354797, "learning_rate": 6.887723968491037e-05, "loss": 0.0129, "step": 122380 }, { "grad_norm": 0.21921668946743011, "learning_rate": 6.887213590789038e-05, "loss": 0.015, "step": 122390 }, { "grad_norm": 0.2667587995529175, "learning_rate": 6.886703190155612e-05, "loss": 0.0206, "step": 122400 }, { "grad_norm": 0.23043379187583923, "learning_rate": 6.88619276659696e-05, "loss": 0.018, "step": 122410 }, { "grad_norm": 0.47617360949516296, "learning_rate": 6.885682320119286e-05, "loss": 0.0214, "step": 122420 }, { "grad_norm": 0.13839249312877655, "learning_rate": 6.88517185072879e-05, "loss": 0.016, "step": 122430 }, { "grad_norm": 0.1172814890742302, "learning_rate": 6.884661358431678e-05, "loss": 0.0128, "step": 122440 }, { "grad_norm": 0.11286313831806183, "learning_rate": 6.884150843234149e-05, "loss": 0.0163, "step": 122450 }, { "grad_norm": 0.09257850795984268, "learning_rate": 6.883640305142411e-05, "loss": 0.0138, "step": 122460 }, { "grad_norm": 0.10059206932783127, "learning_rate": 6.883129744162662e-05, "loss": 0.0135, "step": 122470 }, { "grad_norm": 0.08035404980182648, "learning_rate": 6.882619160301112e-05, "loss": 0.012, "step": 122480 }, { "grad_norm": 0.08095159381628036, "learning_rate": 6.882108553563959e-05, "loss": 0.0129, "step": 122490 }, { "grad_norm": 0.10757260769605637, "learning_rate": 6.88159792395741e-05, "loss": 0.017, "step": 122500 }, { "grad_norm": 0.04936559125781059, "learning_rate": 6.881087271487671e-05, "loss": 0.0192, "step": 122510 }, { "grad_norm": 0.07157205045223236, "learning_rate": 6.880576596160945e-05, "loss": 0.0154, "step": 122520 }, { "grad_norm": 0.2670047879219055, "learning_rate": 6.880065897983438e-05, "loss": 0.0143, "step": 122530 }, { "grad_norm": 0.05620913207530975, "learning_rate": 6.879555176961355e-05, "loss": 0.0134, "step": 122540 }, { "grad_norm": 0.08540663868188858, "learning_rate": 6.879044433100904e-05, "loss": 0.0127, "step": 122550 }, { "grad_norm": 0.05013076961040497, "learning_rate": 6.878533666408286e-05, "loss": 0.0129, "step": 122560 }, { "grad_norm": 0.08455322682857513, "learning_rate": 6.878022876889711e-05, "loss": 0.0204, "step": 122570 }, { "grad_norm": 0.08985572308301926, "learning_rate": 6.877512064551386e-05, "loss": 0.0128, "step": 122580 }, { "grad_norm": 0.09000681340694427, "learning_rate": 6.877001229399516e-05, "loss": 0.0123, "step": 122590 }, { "grad_norm": 0.07612228393554688, "learning_rate": 6.876490371440308e-05, "loss": 0.0184, "step": 122600 }, { "grad_norm": 0.09028179943561554, "learning_rate": 6.875979490679972e-05, "loss": 0.0126, "step": 122610 }, { "grad_norm": 0.16270066797733307, "learning_rate": 6.875468587124714e-05, "loss": 0.0156, "step": 122620 }, { "grad_norm": 0.10950859636068344, "learning_rate": 6.874957660780741e-05, "loss": 0.0219, "step": 122630 }, { "grad_norm": 0.1409030258655548, "learning_rate": 6.874446711654263e-05, "loss": 0.0169, "step": 122640 }, { "grad_norm": 0.09704620391130447, "learning_rate": 6.873935739751486e-05, "loss": 0.0156, "step": 122650 }, { "grad_norm": 0.1075415387749672, "learning_rate": 6.873424745078623e-05, "loss": 0.0139, "step": 122660 }, { "grad_norm": 0.14216956496238708, "learning_rate": 6.872913727641877e-05, "loss": 0.0144, "step": 122670 }, { "grad_norm": 0.22259755432605743, "learning_rate": 6.872402687447464e-05, "loss": 0.0167, "step": 122680 }, { "grad_norm": 0.2475111335515976, "learning_rate": 6.87189162450159e-05, "loss": 0.0151, "step": 122690 }, { "grad_norm": 0.06768914312124252, "learning_rate": 6.871380538810464e-05, "loss": 0.0141, "step": 122700 }, { "grad_norm": 0.11299591511487961, "learning_rate": 6.8708694303803e-05, "loss": 0.0147, "step": 122710 }, { "grad_norm": 0.1878587305545807, "learning_rate": 6.870358299217303e-05, "loss": 0.016, "step": 122720 }, { "grad_norm": 0.24481993913650513, "learning_rate": 6.869847145327689e-05, "loss": 0.0185, "step": 122730 }, { "grad_norm": 0.19023579359054565, "learning_rate": 6.869335968717666e-05, "loss": 0.021, "step": 122740 }, { "grad_norm": 0.09326884150505066, "learning_rate": 6.868824769393446e-05, "loss": 0.014, "step": 122750 }, { "grad_norm": 0.11348670721054077, "learning_rate": 6.868313547361241e-05, "loss": 0.0155, "step": 122760 }, { "grad_norm": 0.19771957397460938, "learning_rate": 6.867802302627262e-05, "loss": 0.0147, "step": 122770 }, { "grad_norm": 0.16852033138275146, "learning_rate": 6.867291035197723e-05, "loss": 0.0183, "step": 122780 }, { "grad_norm": 0.49445274472236633, "learning_rate": 6.866779745078832e-05, "loss": 0.0163, "step": 122790 }, { "grad_norm": 0.10743232071399689, "learning_rate": 6.866268432276806e-05, "loss": 0.0156, "step": 122800 }, { "grad_norm": 0.1830534189939499, "learning_rate": 6.865757096797856e-05, "loss": 0.0182, "step": 122810 }, { "grad_norm": 0.18554550409317017, "learning_rate": 6.865245738648197e-05, "loss": 0.0191, "step": 122820 }, { "grad_norm": 0.1098044142127037, "learning_rate": 6.86473435783404e-05, "loss": 0.0219, "step": 122830 }, { "grad_norm": 0.4922669231891632, "learning_rate": 6.8642229543616e-05, "loss": 0.0216, "step": 122840 }, { "grad_norm": 0.20628902316093445, "learning_rate": 6.863711528237093e-05, "loss": 0.0225, "step": 122850 }, { "grad_norm": 0.11713939160108566, "learning_rate": 6.86320007946673e-05, "loss": 0.0165, "step": 122860 }, { "grad_norm": 0.20476630330085754, "learning_rate": 6.862688608056727e-05, "loss": 0.02, "step": 122870 }, { "grad_norm": 0.12954646348953247, "learning_rate": 6.862177114013297e-05, "loss": 0.0131, "step": 122880 }, { "grad_norm": 0.13151347637176514, "learning_rate": 6.861665597342659e-05, "loss": 0.017, "step": 122890 }, { "grad_norm": 0.2080363631248474, "learning_rate": 6.861154058051024e-05, "loss": 0.0142, "step": 122900 }, { "grad_norm": 0.12146700173616409, "learning_rate": 6.860642496144612e-05, "loss": 0.0171, "step": 122910 }, { "grad_norm": 0.09441909193992615, "learning_rate": 6.860130911629636e-05, "loss": 0.0167, "step": 122920 }, { "grad_norm": 0.13174128532409668, "learning_rate": 6.859619304512315e-05, "loss": 0.0151, "step": 122930 }, { "grad_norm": 0.2537311017513275, "learning_rate": 6.859107674798863e-05, "loss": 0.0238, "step": 122940 }, { "grad_norm": 0.3820193409919739, "learning_rate": 6.858596022495496e-05, "loss": 0.0177, "step": 122950 }, { "grad_norm": 0.14736023545265198, "learning_rate": 6.858084347608433e-05, "loss": 0.0164, "step": 122960 }, { "grad_norm": 0.15175914764404297, "learning_rate": 6.857572650143891e-05, "loss": 0.0138, "step": 122970 }, { "grad_norm": 0.19255565106868744, "learning_rate": 6.857060930108088e-05, "loss": 0.0173, "step": 122980 }, { "grad_norm": 0.15195395052433014, "learning_rate": 6.85654918750724e-05, "loss": 0.0154, "step": 122990 }, { "grad_norm": 0.25494951009750366, "learning_rate": 6.856037422347567e-05, "loss": 0.0146, "step": 123000 }, { "grad_norm": 0.10672994703054428, "learning_rate": 6.855525634635289e-05, "loss": 0.0157, "step": 123010 }, { "grad_norm": 0.08526728302240372, "learning_rate": 6.85501382437662e-05, "loss": 0.0119, "step": 123020 }, { "grad_norm": 0.22962838411331177, "learning_rate": 6.854501991577784e-05, "loss": 0.0116, "step": 123030 }, { "grad_norm": 0.09088412672281265, "learning_rate": 6.853990136244996e-05, "loss": 0.0116, "step": 123040 }, { "grad_norm": 0.10163691639900208, "learning_rate": 6.853478258384476e-05, "loss": 0.0148, "step": 123050 }, { "grad_norm": 0.10751554369926453, "learning_rate": 6.852966358002449e-05, "loss": 0.0159, "step": 123060 }, { "grad_norm": 0.13499020040035248, "learning_rate": 6.852454435105129e-05, "loss": 0.0209, "step": 123070 }, { "grad_norm": 0.13250179588794708, "learning_rate": 6.851942489698737e-05, "loss": 0.0254, "step": 123080 }, { "grad_norm": 0.14185355603694916, "learning_rate": 6.851430521789499e-05, "loss": 0.014, "step": 123090 }, { "grad_norm": 0.12142810970544815, "learning_rate": 6.850918531383629e-05, "loss": 0.0152, "step": 123100 }, { "grad_norm": 0.19451291859149933, "learning_rate": 6.850406518487354e-05, "loss": 0.0162, "step": 123110 }, { "grad_norm": 0.10024917870759964, "learning_rate": 6.84989448310689e-05, "loss": 0.0161, "step": 123120 }, { "grad_norm": 0.7409954071044922, "learning_rate": 6.849382425248464e-05, "loss": 0.0149, "step": 123130 }, { "grad_norm": 0.354757696390152, "learning_rate": 6.848870344918292e-05, "loss": 0.0162, "step": 123140 }, { "grad_norm": 0.14736877381801605, "learning_rate": 6.848358242122604e-05, "loss": 0.0238, "step": 123150 }, { "grad_norm": 0.0908743292093277, "learning_rate": 6.847846116867614e-05, "loss": 0.0185, "step": 123160 }, { "grad_norm": 0.1323871612548828, "learning_rate": 6.847333969159551e-05, "loss": 0.0174, "step": 123170 }, { "grad_norm": 0.10062111169099808, "learning_rate": 6.846821799004637e-05, "loss": 0.0151, "step": 123180 }, { "grad_norm": 0.07163155823945999, "learning_rate": 6.846309606409094e-05, "loss": 0.0146, "step": 123190 }, { "grad_norm": 0.0915946364402771, "learning_rate": 6.845797391379144e-05, "loss": 0.0149, "step": 123200 }, { "grad_norm": 0.21562060713768005, "learning_rate": 6.845285153921014e-05, "loss": 0.0189, "step": 123210 }, { "grad_norm": 0.09388565272092819, "learning_rate": 6.844772894040929e-05, "loss": 0.0141, "step": 123220 }, { "grad_norm": 0.14179667830467224, "learning_rate": 6.84426061174511e-05, "loss": 0.0194, "step": 123230 }, { "grad_norm": 0.08377237617969513, "learning_rate": 6.843748307039783e-05, "loss": 0.0137, "step": 123240 }, { "grad_norm": 0.11052984744310379, "learning_rate": 6.843235979931175e-05, "loss": 0.0178, "step": 123250 }, { "grad_norm": 0.11481724679470062, "learning_rate": 6.84272363042551e-05, "loss": 0.0135, "step": 123260 }, { "grad_norm": 0.10803012549877167, "learning_rate": 6.842211258529009e-05, "loss": 0.0164, "step": 123270 }, { "grad_norm": 0.16899332404136658, "learning_rate": 6.841698864247906e-05, "loss": 0.0162, "step": 123280 }, { "grad_norm": 0.11442717909812927, "learning_rate": 6.841186447588421e-05, "loss": 0.0226, "step": 123290 }, { "grad_norm": 0.1392749547958374, "learning_rate": 6.840674008556782e-05, "loss": 0.0125, "step": 123300 }, { "grad_norm": 0.10525579750537872, "learning_rate": 6.840161547159216e-05, "loss": 0.0144, "step": 123310 }, { "grad_norm": 0.18130549788475037, "learning_rate": 6.839649063401951e-05, "loss": 0.0136, "step": 123320 }, { "grad_norm": 0.19258305430412292, "learning_rate": 6.839136557291214e-05, "loss": 0.0163, "step": 123330 }, { "grad_norm": 0.19452370703220367, "learning_rate": 6.838624028833229e-05, "loss": 0.033, "step": 123340 }, { "grad_norm": 0.12396053969860077, "learning_rate": 6.838111478034228e-05, "loss": 0.0153, "step": 123350 }, { "grad_norm": 0.10151291638612747, "learning_rate": 6.837598904900434e-05, "loss": 0.0147, "step": 123360 }, { "grad_norm": 0.14633125066757202, "learning_rate": 6.837086309438081e-05, "loss": 0.0262, "step": 123370 }, { "grad_norm": 0.12432382255792618, "learning_rate": 6.836573691653394e-05, "loss": 0.0162, "step": 123380 }, { "grad_norm": 0.7487927079200745, "learning_rate": 6.836061051552602e-05, "loss": 0.0192, "step": 123390 }, { "grad_norm": 0.07667200267314911, "learning_rate": 6.835548389141936e-05, "loss": 0.0187, "step": 123400 }, { "grad_norm": 0.11646035313606262, "learning_rate": 6.835035704427623e-05, "loss": 0.0202, "step": 123410 }, { "grad_norm": 0.29170453548431396, "learning_rate": 6.834522997415895e-05, "loss": 0.015, "step": 123420 }, { "grad_norm": 0.13644526898860931, "learning_rate": 6.834010268112979e-05, "loss": 0.0153, "step": 123430 }, { "grad_norm": 0.11105386167764664, "learning_rate": 6.833497516525108e-05, "loss": 0.0155, "step": 123440 }, { "grad_norm": 0.2869279384613037, "learning_rate": 6.83298474265851e-05, "loss": 0.0135, "step": 123450 }, { "grad_norm": 0.38397473096847534, "learning_rate": 6.832471946519416e-05, "loss": 0.0139, "step": 123460 }, { "grad_norm": 0.17487451434135437, "learning_rate": 6.83195912811406e-05, "loss": 0.0256, "step": 123470 }, { "grad_norm": 0.13794319331645966, "learning_rate": 6.83144628744867e-05, "loss": 0.0156, "step": 123480 }, { "grad_norm": 0.2961544394493103, "learning_rate": 6.830933424529477e-05, "loss": 0.0193, "step": 123490 }, { "grad_norm": 0.112458735704422, "learning_rate": 6.830420539362716e-05, "loss": 0.0145, "step": 123500 }, { "grad_norm": 0.17111578583717346, "learning_rate": 6.829907631954617e-05, "loss": 0.0169, "step": 123510 }, { "grad_norm": 0.10168692469596863, "learning_rate": 6.829394702311411e-05, "loss": 0.0177, "step": 123520 }, { "grad_norm": 0.16278886795043945, "learning_rate": 6.828881750439334e-05, "loss": 0.0164, "step": 123530 }, { "grad_norm": 0.09917530417442322, "learning_rate": 6.828368776344615e-05, "loss": 0.0144, "step": 123540 }, { "grad_norm": 0.15892884135246277, "learning_rate": 6.827855780033492e-05, "loss": 0.0181, "step": 123550 }, { "grad_norm": 0.098898746073246, "learning_rate": 6.827342761512193e-05, "loss": 0.0179, "step": 123560 }, { "grad_norm": 0.19404806196689606, "learning_rate": 6.826829720786955e-05, "loss": 0.0159, "step": 123570 }, { "grad_norm": 0.24741564691066742, "learning_rate": 6.826316657864012e-05, "loss": 0.0143, "step": 123580 }, { "grad_norm": 0.07520540058612823, "learning_rate": 6.825803572749596e-05, "loss": 0.0151, "step": 123590 }, { "grad_norm": 0.1007646694779396, "learning_rate": 6.825290465449943e-05, "loss": 0.0157, "step": 123600 }, { "grad_norm": 0.15815073251724243, "learning_rate": 6.824777335971287e-05, "loss": 0.0114, "step": 123610 }, { "grad_norm": 0.0681375041604042, "learning_rate": 6.824264184319865e-05, "loss": 0.0117, "step": 123620 }, { "grad_norm": 0.21451836824417114, "learning_rate": 6.823751010501909e-05, "loss": 0.0239, "step": 123630 }, { "grad_norm": 0.11489243060350418, "learning_rate": 6.823237814523658e-05, "loss": 0.0197, "step": 123640 }, { "grad_norm": 0.28890252113342285, "learning_rate": 6.822724596391344e-05, "loss": 0.0154, "step": 123650 }, { "grad_norm": 0.09997189044952393, "learning_rate": 6.822211356111208e-05, "loss": 0.0165, "step": 123660 }, { "grad_norm": 0.15390649437904358, "learning_rate": 6.821698093689481e-05, "loss": 0.0142, "step": 123670 }, { "grad_norm": 0.1367359161376953, "learning_rate": 6.821184809132404e-05, "loss": 0.0154, "step": 123680 }, { "grad_norm": 0.061941783875226974, "learning_rate": 6.820671502446211e-05, "loss": 0.014, "step": 123690 }, { "grad_norm": 0.1873438060283661, "learning_rate": 6.820158173637142e-05, "loss": 0.0105, "step": 123700 }, { "grad_norm": 0.3078621029853821, "learning_rate": 6.81964482271143e-05, "loss": 0.0114, "step": 123710 }, { "grad_norm": 0.07386250048875809, "learning_rate": 6.819131449675317e-05, "loss": 0.0161, "step": 123720 }, { "grad_norm": 0.10820338875055313, "learning_rate": 6.818618054535039e-05, "loss": 0.0161, "step": 123730 }, { "grad_norm": 0.20519734919071198, "learning_rate": 6.818104637296835e-05, "loss": 0.0175, "step": 123740 }, { "grad_norm": 0.19293589890003204, "learning_rate": 6.817591197966941e-05, "loss": 0.0147, "step": 123750 }, { "grad_norm": 0.15554079413414001, "learning_rate": 6.8170777365516e-05, "loss": 0.0162, "step": 123760 }, { "grad_norm": 0.19753073155879974, "learning_rate": 6.816564253057048e-05, "loss": 0.012, "step": 123770 }, { "grad_norm": 0.09525210410356522, "learning_rate": 6.816050747489526e-05, "loss": 0.0302, "step": 123780 }, { "grad_norm": 0.1436033546924591, "learning_rate": 6.815537219855273e-05, "loss": 0.0188, "step": 123790 }, { "grad_norm": 0.13527099788188934, "learning_rate": 6.815023670160529e-05, "loss": 0.0195, "step": 123800 }, { "grad_norm": 0.16171200573444366, "learning_rate": 6.814510098411532e-05, "loss": 0.013, "step": 123810 }, { "grad_norm": 0.13726231455802917, "learning_rate": 6.813996504614524e-05, "loss": 0.016, "step": 123820 }, { "grad_norm": 0.28269800543785095, "learning_rate": 6.813482888775746e-05, "loss": 0.0161, "step": 123830 }, { "grad_norm": 0.16787685453891754, "learning_rate": 6.81296925090144e-05, "loss": 0.0146, "step": 123840 }, { "grad_norm": 0.10519900172948837, "learning_rate": 6.812455590997845e-05, "loss": 0.0131, "step": 123850 }, { "grad_norm": 0.18381810188293457, "learning_rate": 6.811941909071205e-05, "loss": 0.0134, "step": 123860 }, { "grad_norm": 0.1475072056055069, "learning_rate": 6.811428205127756e-05, "loss": 0.0134, "step": 123870 }, { "grad_norm": 0.12420589476823807, "learning_rate": 6.810914479173749e-05, "loss": 0.0174, "step": 123880 }, { "grad_norm": 0.126713827252388, "learning_rate": 6.810400731215419e-05, "loss": 0.0154, "step": 123890 }, { "grad_norm": 0.08010613173246384, "learning_rate": 6.80988696125901e-05, "loss": 0.0154, "step": 123900 }, { "grad_norm": 0.16770051419734955, "learning_rate": 6.809373169310766e-05, "loss": 0.0116, "step": 123910 }, { "grad_norm": 0.1640745997428894, "learning_rate": 6.80885935537693e-05, "loss": 0.0158, "step": 123920 }, { "grad_norm": 0.1505524218082428, "learning_rate": 6.808345519463744e-05, "loss": 0.0274, "step": 123930 }, { "grad_norm": 0.10025294125080109, "learning_rate": 6.807831661577455e-05, "loss": 0.0155, "step": 123940 }, { "grad_norm": 0.13404275476932526, "learning_rate": 6.807317781724302e-05, "loss": 0.0242, "step": 123950 }, { "grad_norm": 0.08179458975791931, "learning_rate": 6.806803879910533e-05, "loss": 0.0155, "step": 123960 }, { "grad_norm": 0.14732882380485535, "learning_rate": 6.80628995614239e-05, "loss": 0.0182, "step": 123970 }, { "grad_norm": 0.4982583820819855, "learning_rate": 6.805776010426119e-05, "loss": 0.0126, "step": 123980 }, { "grad_norm": 0.20873726904392242, "learning_rate": 6.805262042767966e-05, "loss": 0.0222, "step": 123990 }, { "grad_norm": 0.1631188541650772, "learning_rate": 6.804748053174172e-05, "loss": 0.0166, "step": 124000 }, { "grad_norm": 0.0969904214143753, "learning_rate": 6.804234041650988e-05, "loss": 0.0151, "step": 124010 }, { "grad_norm": 0.24297058582305908, "learning_rate": 6.803720008204655e-05, "loss": 0.0164, "step": 124020 }, { "grad_norm": 0.15139177441596985, "learning_rate": 6.803205952841421e-05, "loss": 0.0126, "step": 124030 }, { "grad_norm": 0.1096731573343277, "learning_rate": 6.802691875567533e-05, "loss": 0.0148, "step": 124040 }, { "grad_norm": 0.21766282618045807, "learning_rate": 6.802177776389236e-05, "loss": 0.016, "step": 124050 }, { "grad_norm": 0.17383670806884766, "learning_rate": 6.801663655312778e-05, "loss": 0.0195, "step": 124060 }, { "grad_norm": 0.16243012249469757, "learning_rate": 6.801149512344404e-05, "loss": 0.0141, "step": 124070 }, { "grad_norm": 0.1232949048280716, "learning_rate": 6.800635347490366e-05, "loss": 0.0153, "step": 124080 }, { "grad_norm": 0.14078707993030548, "learning_rate": 6.800121160756906e-05, "loss": 0.0157, "step": 124090 }, { "grad_norm": 0.10943101346492767, "learning_rate": 6.799606952150275e-05, "loss": 0.016, "step": 124100 }, { "grad_norm": 0.09947886317968369, "learning_rate": 6.79909272167672e-05, "loss": 0.0171, "step": 124110 }, { "grad_norm": 0.06667786836624146, "learning_rate": 6.798578469342491e-05, "loss": 0.0123, "step": 124120 }, { "grad_norm": 0.12634281814098358, "learning_rate": 6.798064195153834e-05, "loss": 0.0135, "step": 124130 }, { "grad_norm": 0.1216403916478157, "learning_rate": 6.797549899117e-05, "loss": 0.0117, "step": 124140 }, { "grad_norm": 0.1201629787683487, "learning_rate": 6.797035581238238e-05, "loss": 0.0181, "step": 124150 }, { "grad_norm": 0.08744537830352783, "learning_rate": 6.796521241523795e-05, "loss": 0.0271, "step": 124160 }, { "grad_norm": 0.10269135981798172, "learning_rate": 6.796006879979927e-05, "loss": 0.0151, "step": 124170 }, { "grad_norm": 0.13951142132282257, "learning_rate": 6.795492496612874e-05, "loss": 0.0187, "step": 124180 }, { "grad_norm": 0.16713038086891174, "learning_rate": 6.794978091428896e-05, "loss": 0.0179, "step": 124190 }, { "grad_norm": 0.154548779129982, "learning_rate": 6.794463664434237e-05, "loss": 0.0202, "step": 124200 }, { "grad_norm": 0.13223402202129364, "learning_rate": 6.793949215635152e-05, "loss": 0.0149, "step": 124210 }, { "grad_norm": 0.11536553502082825, "learning_rate": 6.793434745037891e-05, "loss": 0.0153, "step": 124220 }, { "grad_norm": 0.09398847818374634, "learning_rate": 6.792920252648703e-05, "loss": 0.0106, "step": 124230 }, { "grad_norm": 0.30070093274116516, "learning_rate": 6.792405738473841e-05, "loss": 0.0198, "step": 124240 }, { "grad_norm": 0.17899090051651, "learning_rate": 6.791891202519558e-05, "loss": 0.0178, "step": 124250 }, { "grad_norm": 0.09915383160114288, "learning_rate": 6.791376644792105e-05, "loss": 0.0157, "step": 124260 }, { "grad_norm": 0.6662577390670776, "learning_rate": 6.790862065297732e-05, "loss": 0.0146, "step": 124270 }, { "grad_norm": 0.2121484875679016, "learning_rate": 6.790347464042698e-05, "loss": 0.019, "step": 124280 }, { "grad_norm": 0.11772079765796661, "learning_rate": 6.789832841033249e-05, "loss": 0.0176, "step": 124290 }, { "grad_norm": 0.18580634891986847, "learning_rate": 6.789318196275642e-05, "loss": 0.0133, "step": 124300 }, { "grad_norm": 0.1503037065267563, "learning_rate": 6.788803529776131e-05, "loss": 0.0136, "step": 124310 }, { "grad_norm": 0.12274935096502304, "learning_rate": 6.788288841540966e-05, "loss": 0.0198, "step": 124320 }, { "grad_norm": 0.48906657099723816, "learning_rate": 6.787774131576404e-05, "loss": 0.0185, "step": 124330 }, { "grad_norm": 0.1829380989074707, "learning_rate": 6.787259399888697e-05, "loss": 0.0199, "step": 124340 }, { "grad_norm": 0.12231551855802536, "learning_rate": 6.786744646484103e-05, "loss": 0.0134, "step": 124350 }, { "grad_norm": 0.7060552835464478, "learning_rate": 6.786229871368873e-05, "loss": 0.0152, "step": 124360 }, { "grad_norm": 0.7178144454956055, "learning_rate": 6.785715074549265e-05, "loss": 0.022, "step": 124370 }, { "grad_norm": 0.13656629621982574, "learning_rate": 6.785200256031534e-05, "loss": 0.0173, "step": 124380 }, { "grad_norm": 0.17596589028835297, "learning_rate": 6.784685415821931e-05, "loss": 0.013, "step": 124390 }, { "grad_norm": 0.604681134223938, "learning_rate": 6.784170553926716e-05, "loss": 0.0227, "step": 124400 }, { "grad_norm": 0.2538834810256958, "learning_rate": 6.783655670352145e-05, "loss": 0.0242, "step": 124410 }, { "grad_norm": 0.18514534831047058, "learning_rate": 6.783140765104474e-05, "loss": 0.0181, "step": 124420 }, { "grad_norm": 1.07912278175354, "learning_rate": 6.782625838189957e-05, "loss": 0.023, "step": 124430 }, { "grad_norm": 0.10969606041908264, "learning_rate": 6.782110889614857e-05, "loss": 0.0142, "step": 124440 }, { "grad_norm": 0.23308417201042175, "learning_rate": 6.781595919385424e-05, "loss": 0.0193, "step": 124450 }, { "grad_norm": 0.18105225265026093, "learning_rate": 6.781080927507919e-05, "loss": 0.0226, "step": 124460 }, { "grad_norm": 0.12386298179626465, "learning_rate": 6.780565913988599e-05, "loss": 0.0141, "step": 124470 }, { "grad_norm": 0.10745733231306076, "learning_rate": 6.780050878833722e-05, "loss": 0.017, "step": 124480 }, { "grad_norm": 0.10744275897741318, "learning_rate": 6.779535822049547e-05, "loss": 0.0158, "step": 124490 }, { "grad_norm": 0.10957951098680496, "learning_rate": 6.77902074364233e-05, "loss": 0.019, "step": 124500 }, { "grad_norm": 0.16342726349830627, "learning_rate": 6.778505643618334e-05, "loss": 0.021, "step": 124510 }, { "grad_norm": 0.1136152371764183, "learning_rate": 6.777990521983812e-05, "loss": 0.0127, "step": 124520 }, { "grad_norm": 0.12191509455442429, "learning_rate": 6.77747537874503e-05, "loss": 0.0143, "step": 124530 }, { "grad_norm": 0.10921311378479004, "learning_rate": 6.77696021390824e-05, "loss": 0.0203, "step": 124540 }, { "grad_norm": 0.10567808151245117, "learning_rate": 6.776445027479707e-05, "loss": 0.0144, "step": 124550 }, { "grad_norm": 0.053389620035886765, "learning_rate": 6.775929819465691e-05, "loss": 0.0172, "step": 124560 }, { "grad_norm": 0.12286850810050964, "learning_rate": 6.775414589872451e-05, "loss": 0.0164, "step": 124570 }, { "grad_norm": 0.10432354360818863, "learning_rate": 6.774899338706247e-05, "loss": 0.0184, "step": 124580 }, { "grad_norm": 0.10945514589548111, "learning_rate": 6.77438406597334e-05, "loss": 0.0194, "step": 124590 }, { "grad_norm": 0.15111863613128662, "learning_rate": 6.773868771679991e-05, "loss": 0.0195, "step": 124600 }, { "grad_norm": 0.2893924117088318, "learning_rate": 6.77335345583246e-05, "loss": 0.0232, "step": 124610 }, { "grad_norm": 0.13410478830337524, "learning_rate": 6.772838118437013e-05, "loss": 0.0146, "step": 124620 }, { "grad_norm": 0.1363779455423355, "learning_rate": 6.77232275949991e-05, "loss": 0.0214, "step": 124630 }, { "grad_norm": 0.13484805822372437, "learning_rate": 6.771807379027408e-05, "loss": 0.0146, "step": 124640 }, { "grad_norm": 0.19635629653930664, "learning_rate": 6.771291977025778e-05, "loss": 0.0146, "step": 124650 }, { "grad_norm": 0.10001106560230255, "learning_rate": 6.770776553501275e-05, "loss": 0.0139, "step": 124660 }, { "grad_norm": 0.10410687327384949, "learning_rate": 6.770261108460166e-05, "loss": 0.0181, "step": 124670 }, { "grad_norm": 0.16959987580776215, "learning_rate": 6.769745641908712e-05, "loss": 0.018, "step": 124680 }, { "grad_norm": 0.06374785304069519, "learning_rate": 6.769230153853179e-05, "loss": 0.0206, "step": 124690 }, { "grad_norm": 0.14866770803928375, "learning_rate": 6.768714644299829e-05, "loss": 0.0118, "step": 124700 }, { "grad_norm": 0.12855514883995056, "learning_rate": 6.768199113254925e-05, "loss": 0.0145, "step": 124710 }, { "grad_norm": 0.5349376797676086, "learning_rate": 6.767683560724733e-05, "loss": 0.0156, "step": 124720 }, { "grad_norm": 0.4023537039756775, "learning_rate": 6.767167986715516e-05, "loss": 0.015, "step": 124730 }, { "grad_norm": 0.2642700970172882, "learning_rate": 6.766652391233542e-05, "loss": 0.0186, "step": 124740 }, { "grad_norm": 0.11648296564817429, "learning_rate": 6.766136774285071e-05, "loss": 0.0163, "step": 124750 }, { "grad_norm": 0.07859934121370316, "learning_rate": 6.765621135876372e-05, "loss": 0.0179, "step": 124760 }, { "grad_norm": 0.17737887799739838, "learning_rate": 6.765105476013709e-05, "loss": 0.0183, "step": 124770 }, { "grad_norm": 0.7108476758003235, "learning_rate": 6.764589794703348e-05, "loss": 0.0191, "step": 124780 }, { "grad_norm": 0.12573301792144775, "learning_rate": 6.764074091951554e-05, "loss": 0.0217, "step": 124790 }, { "grad_norm": 0.3810305893421173, "learning_rate": 6.763558367764595e-05, "loss": 0.0134, "step": 124800 }, { "grad_norm": 0.5176571607589722, "learning_rate": 6.763042622148736e-05, "loss": 0.0157, "step": 124810 }, { "grad_norm": 0.11290959268808365, "learning_rate": 6.762526855110245e-05, "loss": 0.0109, "step": 124820 }, { "grad_norm": 0.09161484986543655, "learning_rate": 6.762011066655389e-05, "loss": 0.0126, "step": 124830 }, { "grad_norm": 0.08571804314851761, "learning_rate": 6.761495256790435e-05, "loss": 0.0154, "step": 124840 }, { "grad_norm": 0.27216649055480957, "learning_rate": 6.76097942552165e-05, "loss": 0.0158, "step": 124850 }, { "grad_norm": 0.09928879886865616, "learning_rate": 6.760463572855302e-05, "loss": 0.0139, "step": 124860 }, { "grad_norm": 0.14742527902126312, "learning_rate": 6.759947698797661e-05, "loss": 0.0135, "step": 124870 }, { "grad_norm": 0.26343804597854614, "learning_rate": 6.759431803354992e-05, "loss": 0.0138, "step": 124880 }, { "grad_norm": 0.13491952419281006, "learning_rate": 6.758915886533566e-05, "loss": 0.0141, "step": 124890 }, { "grad_norm": 0.12778988480567932, "learning_rate": 6.758399948339653e-05, "loss": 0.0175, "step": 124900 }, { "grad_norm": 0.18331095576286316, "learning_rate": 6.757883988779517e-05, "loss": 0.0138, "step": 124910 }, { "grad_norm": 0.08557624369859695, "learning_rate": 6.757368007859434e-05, "loss": 0.0147, "step": 124920 }, { "grad_norm": 0.787407398223877, "learning_rate": 6.756852005585668e-05, "loss": 0.016, "step": 124930 }, { "grad_norm": 0.2402029037475586, "learning_rate": 6.756335981964492e-05, "loss": 0.0123, "step": 124940 }, { "grad_norm": 0.1504012495279312, "learning_rate": 6.755819937002177e-05, "loss": 0.0161, "step": 124950 }, { "grad_norm": 0.15719375014305115, "learning_rate": 6.75530387070499e-05, "loss": 0.0207, "step": 124960 }, { "grad_norm": 0.12363569438457489, "learning_rate": 6.754787783079206e-05, "loss": 0.0145, "step": 124970 }, { "grad_norm": 0.2731616795063019, "learning_rate": 6.754271674131091e-05, "loss": 0.0128, "step": 124980 }, { "grad_norm": 0.09497208893299103, "learning_rate": 6.753755543866922e-05, "loss": 0.0143, "step": 124990 }, { "grad_norm": 0.15536193549633026, "learning_rate": 6.753239392292965e-05, "loss": 0.0185, "step": 125000 }, { "grad_norm": 0.1392546445131302, "learning_rate": 6.752723219415494e-05, "loss": 0.0158, "step": 125010 }, { "grad_norm": 0.13758236169815063, "learning_rate": 6.752207025240782e-05, "loss": 0.0238, "step": 125020 }, { "grad_norm": 0.10312077403068542, "learning_rate": 6.7516908097751e-05, "loss": 0.0181, "step": 125030 }, { "grad_norm": 0.10544416308403015, "learning_rate": 6.75117457302472e-05, "loss": 0.017, "step": 125040 }, { "grad_norm": 0.09721069037914276, "learning_rate": 6.750658314995916e-05, "loss": 0.0156, "step": 125050 }, { "grad_norm": 0.05754028633236885, "learning_rate": 6.750142035694962e-05, "loss": 0.0168, "step": 125060 }, { "grad_norm": 0.18469981849193573, "learning_rate": 6.74962573512813e-05, "loss": 0.0191, "step": 125070 }, { "grad_norm": 0.22404111921787262, "learning_rate": 6.749109413301691e-05, "loss": 0.0154, "step": 125080 }, { "grad_norm": 0.15307419002056122, "learning_rate": 6.748593070221923e-05, "loss": 0.0161, "step": 125090 }, { "grad_norm": 0.1659415066242218, "learning_rate": 6.748076705895097e-05, "loss": 0.0198, "step": 125100 }, { "grad_norm": 0.15561701357364655, "learning_rate": 6.74756032032749e-05, "loss": 0.0146, "step": 125110 }, { "grad_norm": 0.3263052701950073, "learning_rate": 6.747043913525373e-05, "loss": 0.0178, "step": 125120 }, { "grad_norm": 0.1859137862920761, "learning_rate": 6.746527485495025e-05, "loss": 0.0159, "step": 125130 }, { "grad_norm": 0.13774189352989197, "learning_rate": 6.746011036242717e-05, "loss": 0.0169, "step": 125140 }, { "grad_norm": 0.11370722204446793, "learning_rate": 6.745494565774728e-05, "loss": 0.018, "step": 125150 }, { "grad_norm": 0.14758992195129395, "learning_rate": 6.744978074097332e-05, "loss": 0.0186, "step": 125160 }, { "grad_norm": 0.14124907553195953, "learning_rate": 6.744461561216805e-05, "loss": 0.0148, "step": 125170 }, { "grad_norm": 0.09514889866113663, "learning_rate": 6.743945027139423e-05, "loss": 0.0207, "step": 125180 }, { "grad_norm": 0.1385117471218109, "learning_rate": 6.74342847187146e-05, "loss": 0.0168, "step": 125190 }, { "grad_norm": 0.08040860295295715, "learning_rate": 6.742911895419198e-05, "loss": 0.0186, "step": 125200 }, { "grad_norm": 0.15876352787017822, "learning_rate": 6.74239529778891e-05, "loss": 0.0182, "step": 125210 }, { "grad_norm": 0.41121169924736023, "learning_rate": 6.741878678986873e-05, "loss": 0.0203, "step": 125220 }, { "grad_norm": 0.11099644750356674, "learning_rate": 6.741362039019367e-05, "loss": 0.0159, "step": 125230 }, { "grad_norm": 0.1195443645119667, "learning_rate": 6.740845377892667e-05, "loss": 0.0183, "step": 125240 }, { "grad_norm": 0.2021516114473343, "learning_rate": 6.740328695613053e-05, "loss": 0.023, "step": 125250 }, { "grad_norm": 0.4441560208797455, "learning_rate": 6.7398119921868e-05, "loss": 0.0151, "step": 125260 }, { "grad_norm": 0.21811027824878693, "learning_rate": 6.739295267620192e-05, "loss": 0.0199, "step": 125270 }, { "grad_norm": 0.2502376139163971, "learning_rate": 6.738778521919501e-05, "loss": 0.0183, "step": 125280 }, { "grad_norm": 0.20574861764907837, "learning_rate": 6.738261755091012e-05, "loss": 0.0168, "step": 125290 }, { "grad_norm": 0.17205221951007843, "learning_rate": 6.737744967141e-05, "loss": 0.0154, "step": 125300 }, { "grad_norm": 0.13473129272460938, "learning_rate": 6.737228158075747e-05, "loss": 0.016, "step": 125310 }, { "grad_norm": 0.12541627883911133, "learning_rate": 6.73671132790153e-05, "loss": 0.0169, "step": 125320 }, { "grad_norm": 0.09091860800981522, "learning_rate": 6.736194476624632e-05, "loss": 0.0119, "step": 125330 }, { "grad_norm": 0.10957096517086029, "learning_rate": 6.735677604251332e-05, "loss": 0.014, "step": 125340 }, { "grad_norm": 0.18763838708400726, "learning_rate": 6.735160710787909e-05, "loss": 0.0224, "step": 125350 }, { "grad_norm": 0.3250468373298645, "learning_rate": 6.734643796240647e-05, "loss": 0.0156, "step": 125360 }, { "grad_norm": 0.08027035742998123, "learning_rate": 6.734126860615824e-05, "loss": 0.0099, "step": 125370 }, { "grad_norm": 0.08725203573703766, "learning_rate": 6.733609903919722e-05, "loss": 0.0125, "step": 125380 }, { "grad_norm": 0.1230132058262825, "learning_rate": 6.733092926158623e-05, "loss": 0.0145, "step": 125390 }, { "grad_norm": 0.11959145963191986, "learning_rate": 6.732575927338808e-05, "loss": 0.0152, "step": 125400 }, { "grad_norm": 0.1781194806098938, "learning_rate": 6.73205890746656e-05, "loss": 0.0151, "step": 125410 }, { "grad_norm": 0.11583657562732697, "learning_rate": 6.731541866548164e-05, "loss": 0.0169, "step": 125420 }, { "grad_norm": 0.15419548749923706, "learning_rate": 6.731024804589896e-05, "loss": 0.0189, "step": 125430 }, { "grad_norm": 0.1253412812948227, "learning_rate": 6.730507721598043e-05, "loss": 0.013, "step": 125440 }, { "grad_norm": 0.15960419178009033, "learning_rate": 6.729990617578889e-05, "loss": 0.0159, "step": 125450 }, { "grad_norm": 0.11948049068450928, "learning_rate": 6.729473492538713e-05, "loss": 0.0136, "step": 125460 }, { "grad_norm": 0.15456971526145935, "learning_rate": 6.728956346483802e-05, "loss": 0.0216, "step": 125470 }, { "grad_norm": 0.1285395622253418, "learning_rate": 6.72843917942044e-05, "loss": 0.018, "step": 125480 }, { "grad_norm": 0.15347343683242798, "learning_rate": 6.727921991354909e-05, "loss": 0.0145, "step": 125490 }, { "grad_norm": 0.09169043600559235, "learning_rate": 6.727404782293494e-05, "loss": 0.0131, "step": 125500 }, { "grad_norm": 0.15887506306171417, "learning_rate": 6.726887552242482e-05, "loss": 0.019, "step": 125510 }, { "grad_norm": 0.27363771200180054, "learning_rate": 6.726370301208154e-05, "loss": 0.0134, "step": 125520 }, { "grad_norm": 0.1786680966615677, "learning_rate": 6.725853029196797e-05, "loss": 0.0166, "step": 125530 }, { "grad_norm": 0.12209447473287582, "learning_rate": 6.725335736214697e-05, "loss": 0.0138, "step": 125540 }, { "grad_norm": 0.08105980604887009, "learning_rate": 6.724818422268137e-05, "loss": 0.0157, "step": 125550 }, { "grad_norm": 0.16186250746250153, "learning_rate": 6.724301087363407e-05, "loss": 0.0168, "step": 125560 }, { "grad_norm": 0.10275445878505707, "learning_rate": 6.72378373150679e-05, "loss": 0.0154, "step": 125570 }, { "grad_norm": 0.15620256960391998, "learning_rate": 6.723266354704574e-05, "loss": 0.0159, "step": 125580 }, { "grad_norm": 0.1343785673379898, "learning_rate": 6.722748956963043e-05, "loss": 0.019, "step": 125590 }, { "grad_norm": 0.13763287663459778, "learning_rate": 6.722231538288486e-05, "loss": 0.0159, "step": 125600 }, { "grad_norm": 0.24638032913208008, "learning_rate": 6.721714098687191e-05, "loss": 0.0322, "step": 125610 }, { "grad_norm": 0.09479015320539474, "learning_rate": 6.721196638165443e-05, "loss": 0.0184, "step": 125620 }, { "grad_norm": 0.14454977214336395, "learning_rate": 6.720679156729532e-05, "loss": 0.0177, "step": 125630 }, { "grad_norm": 0.13424070179462433, "learning_rate": 6.720161654385743e-05, "loss": 0.0217, "step": 125640 }, { "grad_norm": 0.11995252221822739, "learning_rate": 6.719644131140367e-05, "loss": 0.0149, "step": 125650 }, { "grad_norm": 0.21989142894744873, "learning_rate": 6.71912658699969e-05, "loss": 0.0191, "step": 125660 }, { "grad_norm": 0.11288296431303024, "learning_rate": 6.718609021970003e-05, "loss": 0.0186, "step": 125670 }, { "grad_norm": 0.1987304389476776, "learning_rate": 6.718091436057593e-05, "loss": 0.0157, "step": 125680 }, { "grad_norm": 0.08670078217983246, "learning_rate": 6.717573829268749e-05, "loss": 0.0115, "step": 125690 }, { "grad_norm": 0.1195513904094696, "learning_rate": 6.717056201609764e-05, "loss": 0.0174, "step": 125700 }, { "grad_norm": 0.12141544371843338, "learning_rate": 6.716538553086923e-05, "loss": 0.0159, "step": 125710 }, { "grad_norm": 0.30625417828559875, "learning_rate": 6.716020883706518e-05, "loss": 0.0137, "step": 125720 }, { "grad_norm": 0.342945396900177, "learning_rate": 6.715503193474839e-05, "loss": 0.0181, "step": 125730 }, { "grad_norm": 0.3682471811771393, "learning_rate": 6.714985482398179e-05, "loss": 0.0124, "step": 125740 }, { "grad_norm": 0.12301842868328094, "learning_rate": 6.714467750482823e-05, "loss": 0.0109, "step": 125750 }, { "grad_norm": 0.2702128291130066, "learning_rate": 6.713949997735067e-05, "loss": 0.0181, "step": 125760 }, { "grad_norm": 0.20127326250076294, "learning_rate": 6.713432224161199e-05, "loss": 0.0294, "step": 125770 }, { "grad_norm": 0.1288071721792221, "learning_rate": 6.712914429767511e-05, "loss": 0.0133, "step": 125780 }, { "grad_norm": 0.1748887449502945, "learning_rate": 6.712396614560297e-05, "loss": 0.0199, "step": 125790 }, { "grad_norm": 1.0539538860321045, "learning_rate": 6.711878778545848e-05, "loss": 0.0157, "step": 125800 }, { "grad_norm": 0.17897291481494904, "learning_rate": 6.711360921730454e-05, "loss": 0.0157, "step": 125810 }, { "grad_norm": 0.43303215503692627, "learning_rate": 6.710843044120409e-05, "loss": 0.0221, "step": 125820 }, { "grad_norm": 0.09012322127819061, "learning_rate": 6.710325145722007e-05, "loss": 0.0214, "step": 125830 }, { "grad_norm": 0.08512530475854874, "learning_rate": 6.70980722654154e-05, "loss": 0.0101, "step": 125840 }, { "grad_norm": 0.15818752348423004, "learning_rate": 6.709289286585299e-05, "loss": 0.0162, "step": 125850 }, { "grad_norm": 0.15545612573623657, "learning_rate": 6.708771325859581e-05, "loss": 0.0162, "step": 125860 }, { "grad_norm": 0.2799729108810425, "learning_rate": 6.708253344370676e-05, "loss": 0.0191, "step": 125870 }, { "grad_norm": 0.3332309126853943, "learning_rate": 6.707735342124882e-05, "loss": 0.0182, "step": 125880 }, { "grad_norm": 0.12774845957756042, "learning_rate": 6.707217319128489e-05, "loss": 0.0213, "step": 125890 }, { "grad_norm": 0.08404707163572311, "learning_rate": 6.706699275387796e-05, "loss": 0.0122, "step": 125900 }, { "grad_norm": 0.1719086617231369, "learning_rate": 6.706181210909095e-05, "loss": 0.0147, "step": 125910 }, { "grad_norm": 0.10852813720703125, "learning_rate": 6.705663125698682e-05, "loss": 0.0167, "step": 125920 }, { "grad_norm": 0.1768878996372223, "learning_rate": 6.705145019762849e-05, "loss": 0.0146, "step": 125930 }, { "grad_norm": 0.23663602769374847, "learning_rate": 6.704626893107897e-05, "loss": 0.0142, "step": 125940 }, { "grad_norm": 0.12213914096355438, "learning_rate": 6.704108745740117e-05, "loss": 0.0175, "step": 125950 }, { "grad_norm": 0.22413992881774902, "learning_rate": 6.703590577665808e-05, "loss": 0.0173, "step": 125960 }, { "grad_norm": 0.1967398077249527, "learning_rate": 6.703072388891263e-05, "loss": 0.013, "step": 125970 }, { "grad_norm": 0.20994025468826294, "learning_rate": 6.702554179422783e-05, "loss": 0.0179, "step": 125980 }, { "grad_norm": 0.19796112179756165, "learning_rate": 6.702035949266658e-05, "loss": 0.019, "step": 125990 }, { "grad_norm": 0.1143549308180809, "learning_rate": 6.701517698429194e-05, "loss": 0.0168, "step": 126000 }, { "grad_norm": 0.1390397548675537, "learning_rate": 6.70099942691668e-05, "loss": 0.0136, "step": 126010 }, { "grad_norm": 0.09080854058265686, "learning_rate": 6.700481134735421e-05, "loss": 0.0164, "step": 126020 }, { "grad_norm": 0.13522768020629883, "learning_rate": 6.699962821891706e-05, "loss": 0.0156, "step": 126030 }, { "grad_norm": 0.4783197045326233, "learning_rate": 6.699444488391842e-05, "loss": 0.0134, "step": 126040 }, { "grad_norm": 0.2627684473991394, "learning_rate": 6.698926134242119e-05, "loss": 0.0204, "step": 126050 }, { "grad_norm": 0.10516802221536636, "learning_rate": 6.698407759448843e-05, "loss": 0.019, "step": 126060 }, { "grad_norm": 0.18129278719425201, "learning_rate": 6.697889364018306e-05, "loss": 0.0152, "step": 126070 }, { "grad_norm": 0.12604258954524994, "learning_rate": 6.697370947956812e-05, "loss": 0.0177, "step": 126080 }, { "grad_norm": 0.1585696041584015, "learning_rate": 6.696852511270659e-05, "loss": 0.0156, "step": 126090 }, { "grad_norm": 0.21236243844032288, "learning_rate": 6.696334053966144e-05, "loss": 0.0154, "step": 126100 }, { "grad_norm": 0.08091053366661072, "learning_rate": 6.69581557604957e-05, "loss": 0.0122, "step": 126110 }, { "grad_norm": 0.18396632373332977, "learning_rate": 6.695297077527236e-05, "loss": 0.0181, "step": 126120 }, { "grad_norm": 0.08280232548713684, "learning_rate": 6.694778558405441e-05, "loss": 0.0178, "step": 126130 }, { "grad_norm": 0.1513419896364212, "learning_rate": 6.694260018690485e-05, "loss": 0.0152, "step": 126140 }, { "grad_norm": 0.10158863663673401, "learning_rate": 6.693741458388674e-05, "loss": 0.0178, "step": 126150 }, { "grad_norm": 0.11349830776453018, "learning_rate": 6.693222877506302e-05, "loss": 0.0166, "step": 126160 }, { "grad_norm": 0.16770587861537933, "learning_rate": 6.692704276049674e-05, "loss": 0.0164, "step": 126170 }, { "grad_norm": 0.074150450527668, "learning_rate": 6.692185654025092e-05, "loss": 0.0154, "step": 126180 }, { "grad_norm": 0.11482592672109604, "learning_rate": 6.691667011438855e-05, "loss": 0.0158, "step": 126190 }, { "grad_norm": 0.1510077863931656, "learning_rate": 6.691148348297268e-05, "loss": 0.0172, "step": 126200 }, { "grad_norm": 0.18900713324546814, "learning_rate": 6.69062966460663e-05, "loss": 0.0215, "step": 126210 }, { "grad_norm": 0.09147045016288757, "learning_rate": 6.690110960373246e-05, "loss": 0.0132, "step": 126220 }, { "grad_norm": 0.11045711487531662, "learning_rate": 6.689592235603419e-05, "loss": 0.0122, "step": 126230 }, { "grad_norm": 0.12866823375225067, "learning_rate": 6.68907349030345e-05, "loss": 0.0187, "step": 126240 }, { "grad_norm": 0.23891128599643707, "learning_rate": 6.688554724479646e-05, "loss": 0.0153, "step": 126250 }, { "grad_norm": 0.19184540212154388, "learning_rate": 6.688035938138303e-05, "loss": 0.0184, "step": 126260 }, { "grad_norm": 0.15792913734912872, "learning_rate": 6.687517131285735e-05, "loss": 0.0152, "step": 126270 }, { "grad_norm": 0.29357004165649414, "learning_rate": 6.686998303928237e-05, "loss": 0.0152, "step": 126280 }, { "grad_norm": 0.11773869395256042, "learning_rate": 6.686479456072118e-05, "loss": 0.0188, "step": 126290 }, { "grad_norm": 0.16696268320083618, "learning_rate": 6.685960587723681e-05, "loss": 0.0132, "step": 126300 }, { "grad_norm": 0.09147301316261292, "learning_rate": 6.685441698889232e-05, "loss": 0.0143, "step": 126310 }, { "grad_norm": 0.16931407153606415, "learning_rate": 6.684922789575074e-05, "loss": 0.0117, "step": 126320 }, { "grad_norm": 0.14397376775741577, "learning_rate": 6.684403859787514e-05, "loss": 0.0146, "step": 126330 }, { "grad_norm": 0.1730654239654541, "learning_rate": 6.683884909532857e-05, "loss": 0.012, "step": 126340 }, { "grad_norm": 0.1779273897409439, "learning_rate": 6.683365938817406e-05, "loss": 0.0137, "step": 126350 }, { "grad_norm": 0.0628683939576149, "learning_rate": 6.682846947647472e-05, "loss": 0.0175, "step": 126360 }, { "grad_norm": 0.11440417915582657, "learning_rate": 6.682327936029357e-05, "loss": 0.0179, "step": 126370 }, { "grad_norm": 0.21467934548854828, "learning_rate": 6.68180890396937e-05, "loss": 0.019, "step": 126380 }, { "grad_norm": 0.27879074215888977, "learning_rate": 6.681289851473818e-05, "loss": 0.0181, "step": 126390 }, { "grad_norm": 0.40062662959098816, "learning_rate": 6.680770778549006e-05, "loss": 0.019, "step": 126400 }, { "grad_norm": 0.2272789180278778, "learning_rate": 6.680251685201242e-05, "loss": 0.0139, "step": 126410 }, { "grad_norm": 0.09986750781536102, "learning_rate": 6.679732571436834e-05, "loss": 0.0217, "step": 126420 }, { "grad_norm": 0.06530109792947769, "learning_rate": 6.679213437262088e-05, "loss": 0.0174, "step": 126430 }, { "grad_norm": 0.11805307120084763, "learning_rate": 6.678694282683315e-05, "loss": 0.0155, "step": 126440 }, { "grad_norm": 0.1415034383535385, "learning_rate": 6.678175107706821e-05, "loss": 0.0125, "step": 126450 }, { "grad_norm": 0.27501410245895386, "learning_rate": 6.677655912338913e-05, "loss": 0.0118, "step": 126460 }, { "grad_norm": 0.254902720451355, "learning_rate": 6.677136696585906e-05, "loss": 0.022, "step": 126470 }, { "grad_norm": 0.11279398202896118, "learning_rate": 6.676617460454102e-05, "loss": 0.0167, "step": 126480 }, { "grad_norm": 0.1205369085073471, "learning_rate": 6.676098203949812e-05, "loss": 0.0175, "step": 126490 }, { "grad_norm": 0.10455075651407242, "learning_rate": 6.675578927079348e-05, "loss": 0.0207, "step": 126500 }, { "grad_norm": 0.11265379190444946, "learning_rate": 6.675059629849017e-05, "loss": 0.015, "step": 126510 }, { "grad_norm": 0.4730168282985687, "learning_rate": 6.674540312265131e-05, "loss": 0.0157, "step": 126520 }, { "grad_norm": 0.15530677139759064, "learning_rate": 6.674020974333998e-05, "loss": 0.0221, "step": 126530 }, { "grad_norm": 0.15648771822452545, "learning_rate": 6.67350161606193e-05, "loss": 0.0244, "step": 126540 }, { "grad_norm": 0.1809929758310318, "learning_rate": 6.672982237455237e-05, "loss": 0.0182, "step": 126550 }, { "grad_norm": 0.15369898080825806, "learning_rate": 6.672462838520232e-05, "loss": 0.0158, "step": 126560 }, { "grad_norm": 0.14665460586547852, "learning_rate": 6.671943419263223e-05, "loss": 0.0171, "step": 126570 }, { "grad_norm": 0.08052816241979599, "learning_rate": 6.671423979690523e-05, "loss": 0.0187, "step": 126580 }, { "grad_norm": 0.15435461699962616, "learning_rate": 6.670904519808443e-05, "loss": 0.0131, "step": 126590 }, { "grad_norm": 0.1339273899793625, "learning_rate": 6.670385039623295e-05, "loss": 0.0161, "step": 126600 }, { "grad_norm": 0.09367808699607849, "learning_rate": 6.669865539141394e-05, "loss": 0.0171, "step": 126610 }, { "grad_norm": 0.11258795857429504, "learning_rate": 6.669346018369049e-05, "loss": 0.0161, "step": 126620 }, { "grad_norm": 0.24232268333435059, "learning_rate": 6.668826477312574e-05, "loss": 0.0149, "step": 126630 }, { "grad_norm": 0.11226176470518112, "learning_rate": 6.668306915978281e-05, "loss": 0.0194, "step": 126640 }, { "grad_norm": 0.21018704771995544, "learning_rate": 6.667787334372484e-05, "loss": 0.017, "step": 126650 }, { "grad_norm": 0.11197890341281891, "learning_rate": 6.667267732501496e-05, "loss": 0.0175, "step": 126660 }, { "grad_norm": 0.120551697909832, "learning_rate": 6.666748110371632e-05, "loss": 0.0134, "step": 126670 }, { "grad_norm": 0.0953846350312233, "learning_rate": 6.666228467989203e-05, "loss": 0.0204, "step": 126680 }, { "grad_norm": 0.11773362755775452, "learning_rate": 6.665708805360527e-05, "loss": 0.0155, "step": 126690 }, { "grad_norm": 0.14107158780097961, "learning_rate": 6.665189122491915e-05, "loss": 0.0201, "step": 126700 }, { "grad_norm": 0.1003764420747757, "learning_rate": 6.664669419389684e-05, "loss": 0.0186, "step": 126710 }, { "grad_norm": 0.07170370221138, "learning_rate": 6.664149696060146e-05, "loss": 0.0137, "step": 126720 }, { "grad_norm": 0.065617635846138, "learning_rate": 6.663629952509619e-05, "loss": 0.0166, "step": 126730 }, { "grad_norm": 0.11776774376630783, "learning_rate": 6.663110188744417e-05, "loss": 0.0158, "step": 126740 }, { "grad_norm": 0.3341875970363617, "learning_rate": 6.662590404770857e-05, "loss": 0.0224, "step": 126750 }, { "grad_norm": 0.07739418745040894, "learning_rate": 6.662070600595252e-05, "loss": 0.0186, "step": 126760 }, { "grad_norm": 0.11389309167861938, "learning_rate": 6.661550776223922e-05, "loss": 0.0217, "step": 126770 }, { "grad_norm": 0.14857245981693268, "learning_rate": 6.66103093166318e-05, "loss": 0.0198, "step": 126780 }, { "grad_norm": 0.1461065113544464, "learning_rate": 6.660511066919344e-05, "loss": 0.0175, "step": 126790 }, { "grad_norm": 0.14834927022457123, "learning_rate": 6.659991181998733e-05, "loss": 0.0134, "step": 126800 }, { "grad_norm": 0.1041678786277771, "learning_rate": 6.659471276907658e-05, "loss": 0.0176, "step": 126810 }, { "grad_norm": 0.31116539239883423, "learning_rate": 6.658951351652444e-05, "loss": 0.0212, "step": 126820 }, { "grad_norm": 0.1625915914773941, "learning_rate": 6.658431406239402e-05, "loss": 0.0179, "step": 126830 }, { "grad_norm": 0.11829911917448044, "learning_rate": 6.657911440674855e-05, "loss": 0.0224, "step": 126840 }, { "grad_norm": 0.1984969973564148, "learning_rate": 6.657391454965117e-05, "loss": 0.0204, "step": 126850 }, { "grad_norm": 0.1339404284954071, "learning_rate": 6.656871449116508e-05, "loss": 0.0151, "step": 126860 }, { "grad_norm": 0.0749623253941536, "learning_rate": 6.656351423135346e-05, "loss": 0.0185, "step": 126870 }, { "grad_norm": 0.07895589619874954, "learning_rate": 6.655831377027952e-05, "loss": 0.0131, "step": 126880 }, { "grad_norm": 0.20472155511379242, "learning_rate": 6.655311310800643e-05, "loss": 0.0183, "step": 126890 }, { "grad_norm": 0.445466011762619, "learning_rate": 6.654791224459737e-05, "loss": 0.015, "step": 126900 }, { "grad_norm": 0.10982632637023926, "learning_rate": 6.654271118011557e-05, "loss": 0.0108, "step": 126910 }, { "grad_norm": 0.10697837918996811, "learning_rate": 6.65375099146242e-05, "loss": 0.012, "step": 126920 }, { "grad_norm": 0.20740674436092377, "learning_rate": 6.653230844818647e-05, "loss": 0.016, "step": 126930 }, { "grad_norm": 0.2417367696762085, "learning_rate": 6.652710678086558e-05, "loss": 0.0183, "step": 126940 }, { "grad_norm": 0.18068768084049225, "learning_rate": 6.652190491272476e-05, "loss": 0.0175, "step": 126950 }, { "grad_norm": 0.08873621374368668, "learning_rate": 6.651670284382718e-05, "loss": 0.0138, "step": 126960 }, { "grad_norm": 0.08938863128423691, "learning_rate": 6.651150057423607e-05, "loss": 0.0155, "step": 126970 }, { "grad_norm": 0.16163606941699982, "learning_rate": 6.650629810401465e-05, "loss": 0.0134, "step": 126980 }, { "grad_norm": 0.2011035531759262, "learning_rate": 6.65010954332261e-05, "loss": 0.0164, "step": 126990 }, { "grad_norm": 0.06466765701770782, "learning_rate": 6.649589256193367e-05, "loss": 0.0161, "step": 127000 }, { "grad_norm": 0.1829829066991806, "learning_rate": 6.649068949020058e-05, "loss": 0.0214, "step": 127010 }, { "grad_norm": 0.2610629200935364, "learning_rate": 6.648548621809003e-05, "loss": 0.0167, "step": 127020 }, { "grad_norm": 0.1919516623020172, "learning_rate": 6.648028274566526e-05, "loss": 0.0237, "step": 127030 }, { "grad_norm": 0.15588359534740448, "learning_rate": 6.647507907298949e-05, "loss": 0.0141, "step": 127040 }, { "grad_norm": 0.12917034327983856, "learning_rate": 6.646987520012597e-05, "loss": 0.0189, "step": 127050 }, { "grad_norm": 0.0862571969628334, "learning_rate": 6.64646711271379e-05, "loss": 0.0132, "step": 127060 }, { "grad_norm": 0.10968383401632309, "learning_rate": 6.645946685408856e-05, "loss": 0.0196, "step": 127070 }, { "grad_norm": 0.06438973546028137, "learning_rate": 6.645426238104113e-05, "loss": 0.0134, "step": 127080 }, { "grad_norm": 0.3558560013771057, "learning_rate": 6.644905770805889e-05, "loss": 0.0139, "step": 127090 }, { "grad_norm": 0.10234864801168442, "learning_rate": 6.644385283520504e-05, "loss": 0.0175, "step": 127100 }, { "grad_norm": 0.1326010525226593, "learning_rate": 6.64386477625429e-05, "loss": 0.018, "step": 127110 }, { "grad_norm": 0.231787770986557, "learning_rate": 6.643344249013562e-05, "loss": 0.0203, "step": 127120 }, { "grad_norm": 0.09280460327863693, "learning_rate": 6.642823701804653e-05, "loss": 0.0122, "step": 127130 }, { "grad_norm": 0.09226807206869125, "learning_rate": 6.642303134633885e-05, "loss": 0.0139, "step": 127140 }, { "grad_norm": 0.1781940758228302, "learning_rate": 6.641782547507581e-05, "loss": 0.0103, "step": 127150 }, { "grad_norm": 0.08850450068712234, "learning_rate": 6.641261940432071e-05, "loss": 0.0133, "step": 127160 }, { "grad_norm": 0.12158173322677612, "learning_rate": 6.640741313413678e-05, "loss": 0.0141, "step": 127170 }, { "grad_norm": 0.17593945562839508, "learning_rate": 6.640220666458729e-05, "loss": 0.0137, "step": 127180 }, { "grad_norm": 0.2507002651691437, "learning_rate": 6.63969999957355e-05, "loss": 0.0193, "step": 127190 }, { "grad_norm": 0.15726427733898163, "learning_rate": 6.639179312764469e-05, "loss": 0.0194, "step": 127200 }, { "grad_norm": 0.10140062868595123, "learning_rate": 6.638658606037811e-05, "loss": 0.0158, "step": 127210 }, { "grad_norm": 0.11536353081464767, "learning_rate": 6.638137879399903e-05, "loss": 0.0105, "step": 127220 }, { "grad_norm": 0.16581648588180542, "learning_rate": 6.637617132857075e-05, "loss": 0.0107, "step": 127230 }, { "grad_norm": 0.33000439405441284, "learning_rate": 6.63709636641565e-05, "loss": 0.0144, "step": 127240 }, { "grad_norm": 0.10989876091480255, "learning_rate": 6.636575580081961e-05, "loss": 0.0107, "step": 127250 }, { "grad_norm": 0.19743570685386658, "learning_rate": 6.636054773862332e-05, "loss": 0.0157, "step": 127260 }, { "grad_norm": 0.6477600932121277, "learning_rate": 6.635533947763093e-05, "loss": 0.0168, "step": 127270 }, { "grad_norm": 0.15166041254997253, "learning_rate": 6.635013101790573e-05, "loss": 0.0166, "step": 127280 }, { "grad_norm": 0.1376771479845047, "learning_rate": 6.6344922359511e-05, "loss": 0.0178, "step": 127290 }, { "grad_norm": 0.13551326096057892, "learning_rate": 6.633971350251002e-05, "loss": 0.0163, "step": 127300 }, { "grad_norm": 0.10863205790519714, "learning_rate": 6.63345044469661e-05, "loss": 0.0183, "step": 127310 }, { "grad_norm": 0.09542673826217651, "learning_rate": 6.632929519294253e-05, "loss": 0.0126, "step": 127320 }, { "grad_norm": 0.08585277944803238, "learning_rate": 6.632408574050261e-05, "loss": 0.0156, "step": 127330 }, { "grad_norm": 0.1508304923772812, "learning_rate": 6.631887608970963e-05, "loss": 0.0173, "step": 127340 }, { "grad_norm": 0.0930749848484993, "learning_rate": 6.631366624062689e-05, "loss": 0.0197, "step": 127350 }, { "grad_norm": 0.12046589702367783, "learning_rate": 6.630845619331773e-05, "loss": 0.0127, "step": 127360 }, { "grad_norm": 0.13914909958839417, "learning_rate": 6.63032459478454e-05, "loss": 0.0164, "step": 127370 }, { "grad_norm": 0.10293874889612198, "learning_rate": 6.629803550427326e-05, "loss": 0.0196, "step": 127380 }, { "grad_norm": 0.12352783977985382, "learning_rate": 6.629282486266458e-05, "loss": 0.0151, "step": 127390 }, { "grad_norm": 0.09111316502094269, "learning_rate": 6.628761402308273e-05, "loss": 0.0149, "step": 127400 }, { "grad_norm": 0.21024347841739655, "learning_rate": 6.628240298559096e-05, "loss": 0.0156, "step": 127410 }, { "grad_norm": 0.35531020164489746, "learning_rate": 6.627719175025261e-05, "loss": 0.0183, "step": 127420 }, { "grad_norm": 0.11693157255649567, "learning_rate": 6.627198031713105e-05, "loss": 0.018, "step": 127430 }, { "grad_norm": 0.25404322147369385, "learning_rate": 6.626676868628954e-05, "loss": 0.0165, "step": 127440 }, { "grad_norm": 0.12952721118927002, "learning_rate": 6.626155685779143e-05, "loss": 0.017, "step": 127450 }, { "grad_norm": 0.13273751735687256, "learning_rate": 6.625634483170007e-05, "loss": 0.0179, "step": 127460 }, { "grad_norm": 0.21801359951496124, "learning_rate": 6.625113260807875e-05, "loss": 0.0193, "step": 127470 }, { "grad_norm": 0.14154687523841858, "learning_rate": 6.624592018699084e-05, "loss": 0.0185, "step": 127480 }, { "grad_norm": 0.15298600494861603, "learning_rate": 6.624070756849964e-05, "loss": 0.0193, "step": 127490 }, { "grad_norm": 0.06994598358869553, "learning_rate": 6.623549475266855e-05, "loss": 0.0089, "step": 127500 }, { "grad_norm": 0.09575553238391876, "learning_rate": 6.623028173956083e-05, "loss": 0.0188, "step": 127510 }, { "grad_norm": 0.15033385157585144, "learning_rate": 6.622506852923988e-05, "loss": 0.0191, "step": 127520 }, { "grad_norm": 0.1393628716468811, "learning_rate": 6.621985512176905e-05, "loss": 0.0136, "step": 127530 }, { "grad_norm": 0.09037426859140396, "learning_rate": 6.621464151721164e-05, "loss": 0.0176, "step": 127540 }, { "grad_norm": 0.1167544275522232, "learning_rate": 6.620942771563104e-05, "loss": 0.0174, "step": 127550 }, { "grad_norm": 0.1773804873228073, "learning_rate": 6.620421371709055e-05, "loss": 0.0211, "step": 127560 }, { "grad_norm": 0.13592667877674103, "learning_rate": 6.61989995216536e-05, "loss": 0.0157, "step": 127570 }, { "grad_norm": 0.6475173234939575, "learning_rate": 6.619378512938351e-05, "loss": 0.0158, "step": 127580 }, { "grad_norm": 0.5590659976005554, "learning_rate": 6.618857054034365e-05, "loss": 0.0171, "step": 127590 }, { "grad_norm": 0.14078116416931152, "learning_rate": 6.618335575459734e-05, "loss": 0.0191, "step": 127600 }, { "grad_norm": 0.07419757544994354, "learning_rate": 6.617814077220801e-05, "loss": 0.0173, "step": 127610 }, { "grad_norm": 0.14698979258537292, "learning_rate": 6.617292559323899e-05, "loss": 0.0095, "step": 127620 }, { "grad_norm": 0.22452741861343384, "learning_rate": 6.616771021775363e-05, "loss": 0.0207, "step": 127630 }, { "grad_norm": 0.13265632092952728, "learning_rate": 6.616249464581535e-05, "loss": 0.0158, "step": 127640 }, { "grad_norm": 0.2577575147151947, "learning_rate": 6.615727887748749e-05, "loss": 0.0124, "step": 127650 }, { "grad_norm": 0.12299443781375885, "learning_rate": 6.615206291283344e-05, "loss": 0.0208, "step": 127660 }, { "grad_norm": 0.14054147899150848, "learning_rate": 6.614684675191658e-05, "loss": 0.0143, "step": 127670 }, { "grad_norm": 0.11715493351221085, "learning_rate": 6.614163039480027e-05, "loss": 0.0147, "step": 127680 }, { "grad_norm": 0.09691444784402847, "learning_rate": 6.613641384154794e-05, "loss": 0.0135, "step": 127690 }, { "grad_norm": 0.3103664517402649, "learning_rate": 6.613119709222292e-05, "loss": 0.0131, "step": 127700 }, { "grad_norm": 0.8819142580032349, "learning_rate": 6.612598014688864e-05, "loss": 0.0193, "step": 127710 }, { "grad_norm": 0.1198350042104721, "learning_rate": 6.612076300560847e-05, "loss": 0.0167, "step": 127720 }, { "grad_norm": 0.24869178235530853, "learning_rate": 6.611554566844581e-05, "loss": 0.0161, "step": 127730 }, { "grad_norm": 0.7298271656036377, "learning_rate": 6.611032813546405e-05, "loss": 0.0123, "step": 127740 }, { "grad_norm": 0.17989864945411682, "learning_rate": 6.61051104067266e-05, "loss": 0.0191, "step": 127750 }, { "grad_norm": 0.19785140454769135, "learning_rate": 6.609989248229684e-05, "loss": 0.0172, "step": 127760 }, { "grad_norm": 0.10150475054979324, "learning_rate": 6.609467436223819e-05, "loss": 0.016, "step": 127770 }, { "grad_norm": 0.5774552822113037, "learning_rate": 6.608945604661407e-05, "loss": 0.0207, "step": 127780 }, { "grad_norm": 0.1653977781534195, "learning_rate": 6.608423753548785e-05, "loss": 0.0158, "step": 127790 }, { "grad_norm": 0.20874907076358795, "learning_rate": 6.607901882892296e-05, "loss": 0.0167, "step": 127800 }, { "grad_norm": 0.20048265159130096, "learning_rate": 6.607379992698281e-05, "loss": 0.025, "step": 127810 }, { "grad_norm": 0.4826362431049347, "learning_rate": 6.606858082973082e-05, "loss": 0.0199, "step": 127820 }, { "grad_norm": 0.13760346174240112, "learning_rate": 6.606336153723038e-05, "loss": 0.0211, "step": 127830 }, { "grad_norm": 0.1719818115234375, "learning_rate": 6.605814204954496e-05, "loss": 0.0144, "step": 127840 }, { "grad_norm": 0.08849412202835083, "learning_rate": 6.605292236673794e-05, "loss": 0.0173, "step": 127850 }, { "grad_norm": 0.12457884848117828, "learning_rate": 6.604770248887275e-05, "loss": 0.0165, "step": 127860 }, { "grad_norm": 0.150710791349411, "learning_rate": 6.604248241601285e-05, "loss": 0.0162, "step": 127870 }, { "grad_norm": 0.07972145825624466, "learning_rate": 6.60372621482216e-05, "loss": 0.0154, "step": 127880 }, { "grad_norm": 0.09626784175634384, "learning_rate": 6.603204168556251e-05, "loss": 0.0189, "step": 127890 }, { "grad_norm": 0.15506912767887115, "learning_rate": 6.602682102809896e-05, "loss": 0.0179, "step": 127900 }, { "grad_norm": 0.09345240890979767, "learning_rate": 6.60216001758944e-05, "loss": 0.021, "step": 127910 }, { "grad_norm": 0.17702388763427734, "learning_rate": 6.601637912901227e-05, "loss": 0.0149, "step": 127920 }, { "grad_norm": 0.1101200059056282, "learning_rate": 6.601115788751602e-05, "loss": 0.0134, "step": 127930 }, { "grad_norm": 0.1468552201986313, "learning_rate": 6.600593645146909e-05, "loss": 0.0196, "step": 127940 }, { "grad_norm": 0.2669890820980072, "learning_rate": 6.600071482093491e-05, "loss": 0.0195, "step": 127950 }, { "grad_norm": 0.16489766538143158, "learning_rate": 6.599549299597695e-05, "loss": 0.0142, "step": 127960 }, { "grad_norm": 0.1473282277584076, "learning_rate": 6.599027097665863e-05, "loss": 0.0149, "step": 127970 }, { "grad_norm": 0.0848415344953537, "learning_rate": 6.598504876304343e-05, "loss": 0.0153, "step": 127980 }, { "grad_norm": 0.12049847096204758, "learning_rate": 6.59798263551948e-05, "loss": 0.0155, "step": 127990 }, { "grad_norm": 0.16752775013446808, "learning_rate": 6.597460375317619e-05, "loss": 0.0192, "step": 128000 }, { "grad_norm": 0.10838823020458221, "learning_rate": 6.596938095705105e-05, "loss": 0.0163, "step": 128010 }, { "grad_norm": 0.22209717333316803, "learning_rate": 6.596415796688287e-05, "loss": 0.0213, "step": 128020 }, { "grad_norm": 0.2293900102376938, "learning_rate": 6.59589347827351e-05, "loss": 0.0139, "step": 128030 }, { "grad_norm": 0.08608727157115936, "learning_rate": 6.595371140467117e-05, "loss": 0.0146, "step": 128040 }, { "grad_norm": 0.08841552585363388, "learning_rate": 6.594848783275462e-05, "loss": 0.0128, "step": 128050 }, { "grad_norm": 0.32298070192337036, "learning_rate": 6.594326406704887e-05, "loss": 0.0187, "step": 128060 }, { "grad_norm": 0.11436548084020615, "learning_rate": 6.593804010761741e-05, "loss": 0.0191, "step": 128070 }, { "grad_norm": 0.13324971497058868, "learning_rate": 6.593281595452372e-05, "loss": 0.0165, "step": 128080 }, { "grad_norm": 0.2200845181941986, "learning_rate": 6.592759160783127e-05, "loss": 0.0182, "step": 128090 }, { "grad_norm": 0.14573049545288086, "learning_rate": 6.592236706760354e-05, "loss": 0.0153, "step": 128100 }, { "grad_norm": 0.18387198448181152, "learning_rate": 6.591714233390402e-05, "loss": 0.0141, "step": 128110 }, { "grad_norm": 0.2514013350009918, "learning_rate": 6.591191740679619e-05, "loss": 0.0199, "step": 128120 }, { "grad_norm": 0.17527969181537628, "learning_rate": 6.590669228634354e-05, "loss": 0.0164, "step": 128130 }, { "grad_norm": 0.07546654343605042, "learning_rate": 6.590146697260957e-05, "loss": 0.0139, "step": 128140 }, { "grad_norm": 0.12835285067558289, "learning_rate": 6.589624146565775e-05, "loss": 0.0113, "step": 128150 }, { "grad_norm": 0.1427449733018875, "learning_rate": 6.58910157655516e-05, "loss": 0.0193, "step": 128160 }, { "grad_norm": 0.10134599357843399, "learning_rate": 6.588578987235459e-05, "loss": 0.0116, "step": 128170 }, { "grad_norm": 0.17504984140396118, "learning_rate": 6.588056378613023e-05, "loss": 0.0163, "step": 128180 }, { "grad_norm": 0.0948554128408432, "learning_rate": 6.587533750694206e-05, "loss": 0.017, "step": 128190 }, { "grad_norm": 0.15210704505443573, "learning_rate": 6.587011103485352e-05, "loss": 0.016, "step": 128200 }, { "grad_norm": 0.1945606768131256, "learning_rate": 6.586488436992816e-05, "loss": 0.0155, "step": 128210 }, { "grad_norm": 0.112053781747818, "learning_rate": 6.585965751222947e-05, "loss": 0.0136, "step": 128220 }, { "grad_norm": 0.09360717236995697, "learning_rate": 6.585443046182097e-05, "loss": 0.0199, "step": 128230 }, { "grad_norm": 0.18232011795043945, "learning_rate": 6.584920321876617e-05, "loss": 0.0209, "step": 128240 }, { "grad_norm": 0.13053956627845764, "learning_rate": 6.584397578312859e-05, "loss": 0.0133, "step": 128250 }, { "grad_norm": 0.09951980412006378, "learning_rate": 6.583874815497175e-05, "loss": 0.0109, "step": 128260 }, { "grad_norm": 0.10390853136777878, "learning_rate": 6.583352033435914e-05, "loss": 0.0136, "step": 128270 }, { "grad_norm": 0.1378580778837204, "learning_rate": 6.582829232135433e-05, "loss": 0.0115, "step": 128280 }, { "grad_norm": 0.2458488792181015, "learning_rate": 6.582306411602081e-05, "loss": 0.0148, "step": 128290 }, { "grad_norm": 0.17058438062667847, "learning_rate": 6.581783571842216e-05, "loss": 0.0146, "step": 128300 }, { "grad_norm": 0.13501699268817902, "learning_rate": 6.581260712862183e-05, "loss": 0.02, "step": 128310 }, { "grad_norm": 0.11125078052282333, "learning_rate": 6.580737834668341e-05, "loss": 0.0128, "step": 128320 }, { "grad_norm": 0.12487677484750748, "learning_rate": 6.58021493726704e-05, "loss": 0.0112, "step": 128330 }, { "grad_norm": 0.13193079829216003, "learning_rate": 6.579692020664639e-05, "loss": 0.0178, "step": 128340 }, { "grad_norm": 0.10937055945396423, "learning_rate": 6.579169084867486e-05, "loss": 0.0212, "step": 128350 }, { "grad_norm": 0.11776671558618546, "learning_rate": 6.578646129881938e-05, "loss": 0.0116, "step": 128360 }, { "grad_norm": 0.0730077400803566, "learning_rate": 6.57812315571435e-05, "loss": 0.0142, "step": 128370 }, { "grad_norm": 0.10111316293478012, "learning_rate": 6.577600162371074e-05, "loss": 0.0134, "step": 128380 }, { "grad_norm": 0.08978912234306335, "learning_rate": 6.577077149858468e-05, "loss": 0.0191, "step": 128390 }, { "grad_norm": 0.20155826210975647, "learning_rate": 6.576554118182885e-05, "loss": 0.0171, "step": 128400 }, { "grad_norm": 0.11794821172952652, "learning_rate": 6.576031067350681e-05, "loss": 0.0159, "step": 128410 }, { "grad_norm": 0.164389505982399, "learning_rate": 6.575507997368212e-05, "loss": 0.0175, "step": 128420 }, { "grad_norm": 0.11527350544929504, "learning_rate": 6.574984908241833e-05, "loss": 0.0204, "step": 128430 }, { "grad_norm": 0.1808052510023117, "learning_rate": 6.5744617999779e-05, "loss": 0.015, "step": 128440 }, { "grad_norm": 0.10655559599399567, "learning_rate": 6.573938672582769e-05, "loss": 0.0141, "step": 128450 }, { "grad_norm": 0.12214472144842148, "learning_rate": 6.573415526062798e-05, "loss": 0.018, "step": 128460 }, { "grad_norm": 0.22869746387004852, "learning_rate": 6.572892360424342e-05, "loss": 0.0188, "step": 128470 }, { "grad_norm": 0.1123848482966423, "learning_rate": 6.57236917567376e-05, "loss": 0.0181, "step": 128480 }, { "grad_norm": 0.08838623762130737, "learning_rate": 6.571845971817407e-05, "loss": 0.0143, "step": 128490 }, { "grad_norm": 0.2760597765445709, "learning_rate": 6.571322748861643e-05, "loss": 0.0177, "step": 128500 }, { "grad_norm": 0.09651604294776917, "learning_rate": 6.570799506812824e-05, "loss": 0.0125, "step": 128510 }, { "grad_norm": 0.2479773312807083, "learning_rate": 6.570276245677305e-05, "loss": 0.0162, "step": 128520 }, { "grad_norm": 0.26482611894607544, "learning_rate": 6.569752965461448e-05, "loss": 0.0169, "step": 128530 }, { "grad_norm": 0.1313815861940384, "learning_rate": 6.569229666171613e-05, "loss": 0.0125, "step": 128540 }, { "grad_norm": 0.07375622540712357, "learning_rate": 6.568706347814154e-05, "loss": 0.0112, "step": 128550 }, { "grad_norm": 0.09558121860027313, "learning_rate": 6.56818301039543e-05, "loss": 0.013, "step": 128560 }, { "grad_norm": 0.07694180309772491, "learning_rate": 6.567659653921805e-05, "loss": 0.0127, "step": 128570 }, { "grad_norm": 0.20022279024124146, "learning_rate": 6.567136278399634e-05, "loss": 0.0167, "step": 128580 }, { "grad_norm": 0.10925307124853134, "learning_rate": 6.566612883835276e-05, "loss": 0.0143, "step": 128590 }, { "grad_norm": 0.10853027552366257, "learning_rate": 6.566089470235095e-05, "loss": 0.0153, "step": 128600 }, { "grad_norm": 0.238721564412117, "learning_rate": 6.565566037605447e-05, "loss": 0.0167, "step": 128610 }, { "grad_norm": 0.18741339445114136, "learning_rate": 6.565042585952693e-05, "loss": 0.0153, "step": 128620 }, { "grad_norm": 0.12806639075279236, "learning_rate": 6.564519115283194e-05, "loss": 0.0168, "step": 128630 }, { "grad_norm": 0.19208601117134094, "learning_rate": 6.563995625603312e-05, "loss": 0.0153, "step": 128640 }, { "grad_norm": 0.10060447454452515, "learning_rate": 6.563472116919405e-05, "loss": 0.0267, "step": 128650 }, { "grad_norm": 0.15252725780010223, "learning_rate": 6.562948589237837e-05, "loss": 0.0159, "step": 128660 }, { "grad_norm": 0.13861919939517975, "learning_rate": 6.562425042564967e-05, "loss": 0.0188, "step": 128670 }, { "grad_norm": 0.22537361085414886, "learning_rate": 6.561901476907157e-05, "loss": 0.0269, "step": 128680 }, { "grad_norm": 0.14743074774742126, "learning_rate": 6.56137789227077e-05, "loss": 0.0165, "step": 128690 }, { "grad_norm": 0.09151103347539902, "learning_rate": 6.560854288662167e-05, "loss": 0.0152, "step": 128700 }, { "grad_norm": 0.1492435485124588, "learning_rate": 6.560330666087711e-05, "loss": 0.012, "step": 128710 }, { "grad_norm": 0.13651320338249207, "learning_rate": 6.559807024553765e-05, "loss": 0.0203, "step": 128720 }, { "grad_norm": 0.081081323325634, "learning_rate": 6.55928336406669e-05, "loss": 0.0161, "step": 128730 }, { "grad_norm": 0.21630994975566864, "learning_rate": 6.558759684632851e-05, "loss": 0.0161, "step": 128740 }, { "grad_norm": 0.15820816159248352, "learning_rate": 6.558235986258608e-05, "loss": 0.0107, "step": 128750 }, { "grad_norm": 0.10268840938806534, "learning_rate": 6.557712268950329e-05, "loss": 0.0142, "step": 128760 }, { "grad_norm": 0.1204322800040245, "learning_rate": 6.557188532714374e-05, "loss": 0.0153, "step": 128770 }, { "grad_norm": 0.1842982918024063, "learning_rate": 6.556664777557109e-05, "loss": 0.0149, "step": 128780 }, { "grad_norm": 0.09457171708345413, "learning_rate": 6.556141003484895e-05, "loss": 0.0176, "step": 128790 }, { "grad_norm": 0.20739266276359558, "learning_rate": 6.5556172105041e-05, "loss": 0.0178, "step": 128800 }, { "grad_norm": 0.09583453088998795, "learning_rate": 6.555093398621088e-05, "loss": 0.0218, "step": 128810 }, { "grad_norm": 0.0868617445230484, "learning_rate": 6.554569567842221e-05, "loss": 0.014, "step": 128820 }, { "grad_norm": 0.11728081107139587, "learning_rate": 6.554045718173867e-05, "loss": 0.0123, "step": 128830 }, { "grad_norm": 0.08219458907842636, "learning_rate": 6.553521849622391e-05, "loss": 0.0121, "step": 128840 }, { "grad_norm": 0.10434267669916153, "learning_rate": 6.552997962194157e-05, "loss": 0.0124, "step": 128850 }, { "grad_norm": 0.09612059593200684, "learning_rate": 6.552474055895531e-05, "loss": 0.0133, "step": 128860 }, { "grad_norm": 0.14681461453437805, "learning_rate": 6.55195013073288e-05, "loss": 0.0163, "step": 128870 }, { "grad_norm": 0.16618861258029938, "learning_rate": 6.55142618671257e-05, "loss": 0.02, "step": 128880 }, { "grad_norm": 0.12434549629688263, "learning_rate": 6.550902223840967e-05, "loss": 0.0201, "step": 128890 }, { "grad_norm": 0.14873787760734558, "learning_rate": 6.550378242124437e-05, "loss": 0.0197, "step": 128900 }, { "grad_norm": 0.8165755271911621, "learning_rate": 6.549854241569347e-05, "loss": 0.0172, "step": 128910 }, { "grad_norm": 0.22134320437908173, "learning_rate": 6.549330222182067e-05, "loss": 0.0148, "step": 128920 }, { "grad_norm": 0.14013558626174927, "learning_rate": 6.54880618396896e-05, "loss": 0.018, "step": 128930 }, { "grad_norm": 0.08992417901754379, "learning_rate": 6.548282126936397e-05, "loss": 0.0192, "step": 128940 }, { "grad_norm": 0.11168788373470306, "learning_rate": 6.547758051090742e-05, "loss": 0.0156, "step": 128950 }, { "grad_norm": 0.18301886320114136, "learning_rate": 6.547233956438368e-05, "loss": 0.0175, "step": 128960 }, { "grad_norm": 0.10823408514261246, "learning_rate": 6.546709842985639e-05, "loss": 0.0217, "step": 128970 }, { "grad_norm": 0.10495366156101227, "learning_rate": 6.546185710738926e-05, "loss": 0.0179, "step": 128980 }, { "grad_norm": 0.10086328536272049, "learning_rate": 6.545661559704596e-05, "loss": 0.0191, "step": 128990 }, { "grad_norm": 0.11743805557489395, "learning_rate": 6.545137389889018e-05, "loss": 0.0224, "step": 129000 }, { "grad_norm": 0.08888774365186691, "learning_rate": 6.544613201298564e-05, "loss": 0.0207, "step": 129010 }, { "grad_norm": 0.1297236979007721, "learning_rate": 6.544088993939599e-05, "loss": 0.0146, "step": 129020 }, { "grad_norm": 0.14309756457805634, "learning_rate": 6.543564767818497e-05, "loss": 0.0157, "step": 129030 }, { "grad_norm": 0.14121301472187042, "learning_rate": 6.543040522941624e-05, "loss": 0.0172, "step": 129040 }, { "grad_norm": 0.16899904608726501, "learning_rate": 6.542516259315352e-05, "loss": 0.0166, "step": 129050 }, { "grad_norm": 0.22404775023460388, "learning_rate": 6.541991976946052e-05, "loss": 0.0187, "step": 129060 }, { "grad_norm": 0.22558361291885376, "learning_rate": 6.541467675840093e-05, "loss": 0.0193, "step": 129070 }, { "grad_norm": 0.1648510843515396, "learning_rate": 6.540943356003847e-05, "loss": 0.0205, "step": 129080 }, { "grad_norm": 0.22144410014152527, "learning_rate": 6.540419017443683e-05, "loss": 0.0137, "step": 129090 }, { "grad_norm": 0.1951313465833664, "learning_rate": 6.539894660165975e-05, "loss": 0.0244, "step": 129100 }, { "grad_norm": 0.4451216459274292, "learning_rate": 6.539370284177092e-05, "loss": 0.0182, "step": 129110 }, { "grad_norm": 1.0835119485855103, "learning_rate": 6.538845889483407e-05, "loss": 0.0201, "step": 129120 }, { "grad_norm": 0.15366914868354797, "learning_rate": 6.538321476091292e-05, "loss": 0.0202, "step": 129130 }, { "grad_norm": 0.12172413617372513, "learning_rate": 6.537797044007118e-05, "loss": 0.0216, "step": 129140 }, { "grad_norm": 0.21132320165634155, "learning_rate": 6.537272593237257e-05, "loss": 0.017, "step": 129150 }, { "grad_norm": 0.23082895576953888, "learning_rate": 6.536748123788084e-05, "loss": 0.0202, "step": 129160 }, { "grad_norm": 0.1081860363483429, "learning_rate": 6.536223635665971e-05, "loss": 0.0186, "step": 129170 }, { "grad_norm": 0.16708330810070038, "learning_rate": 6.535699128877288e-05, "loss": 0.0173, "step": 129180 }, { "grad_norm": 0.25835415720939636, "learning_rate": 6.535174603428413e-05, "loss": 0.0156, "step": 129190 }, { "grad_norm": 0.5477100610733032, "learning_rate": 6.534650059325715e-05, "loss": 0.0171, "step": 129200 }, { "grad_norm": 0.1460331231355667, "learning_rate": 6.534125496575573e-05, "loss": 0.017, "step": 129210 }, { "grad_norm": 0.21607714891433716, "learning_rate": 6.533600915184356e-05, "loss": 0.018, "step": 129220 }, { "grad_norm": 0.23942500352859497, "learning_rate": 6.533076315158438e-05, "loss": 0.0218, "step": 129230 }, { "grad_norm": 0.09644079208374023, "learning_rate": 6.532551696504198e-05, "loss": 0.0162, "step": 129240 }, { "grad_norm": 0.07372292876243591, "learning_rate": 6.532027059228006e-05, "loss": 0.0125, "step": 129250 }, { "grad_norm": 0.11418547481298447, "learning_rate": 6.53150240333624e-05, "loss": 0.0152, "step": 129260 }, { "grad_norm": 0.10968806594610214, "learning_rate": 6.530977728835274e-05, "loss": 0.0137, "step": 129270 }, { "grad_norm": 0.09867976605892181, "learning_rate": 6.530453035731482e-05, "loss": 0.0126, "step": 129280 }, { "grad_norm": 0.3727087378501892, "learning_rate": 6.529928324031242e-05, "loss": 0.0142, "step": 129290 }, { "grad_norm": 0.19683144986629486, "learning_rate": 6.529403593740929e-05, "loss": 0.015, "step": 129300 }, { "grad_norm": 0.13568083941936493, "learning_rate": 6.528878844866916e-05, "loss": 0.0244, "step": 129310 }, { "grad_norm": 0.10741677135229111, "learning_rate": 6.528354077415582e-05, "loss": 0.0174, "step": 129320 }, { "grad_norm": 0.08114264160394669, "learning_rate": 6.527829291393304e-05, "loss": 0.0148, "step": 129330 }, { "grad_norm": 0.2095142900943756, "learning_rate": 6.527304486806456e-05, "loss": 0.017, "step": 129340 }, { "grad_norm": 0.15718238055706024, "learning_rate": 6.526779663661419e-05, "loss": 0.0119, "step": 129350 }, { "grad_norm": 0.1224118322134018, "learning_rate": 6.526254821964565e-05, "loss": 0.0143, "step": 129360 }, { "grad_norm": 0.1707104742527008, "learning_rate": 6.525729961722276e-05, "loss": 0.013, "step": 129370 }, { "grad_norm": 0.10285097360610962, "learning_rate": 6.525205082940925e-05, "loss": 0.0144, "step": 129380 }, { "grad_norm": 0.08347497880458832, "learning_rate": 6.524680185626894e-05, "loss": 0.0181, "step": 129390 }, { "grad_norm": 0.3158474862575531, "learning_rate": 6.52415526978656e-05, "loss": 0.0178, "step": 129400 }, { "grad_norm": 0.10640103369951248, "learning_rate": 6.523630335426298e-05, "loss": 0.0173, "step": 129410 }, { "grad_norm": 0.09020546823740005, "learning_rate": 6.523105382552491e-05, "loss": 0.014, "step": 129420 }, { "grad_norm": 0.1300293654203415, "learning_rate": 6.522580411171515e-05, "loss": 0.0132, "step": 129430 }, { "grad_norm": 0.4700896441936493, "learning_rate": 6.522055421289749e-05, "loss": 0.014, "step": 129440 }, { "grad_norm": 0.11266522854566574, "learning_rate": 6.521530412913573e-05, "loss": 0.0177, "step": 129450 }, { "grad_norm": 0.11573623865842819, "learning_rate": 6.521005386049366e-05, "loss": 0.0138, "step": 129460 }, { "grad_norm": 0.11048050969839096, "learning_rate": 6.520480340703508e-05, "loss": 0.014, "step": 129470 }, { "grad_norm": 0.11991442739963531, "learning_rate": 6.519955276882375e-05, "loss": 0.0159, "step": 129480 }, { "grad_norm": 0.12000016868114471, "learning_rate": 6.519430194592355e-05, "loss": 0.0181, "step": 129490 }, { "grad_norm": 0.16520151495933533, "learning_rate": 6.51890509383982e-05, "loss": 0.013, "step": 129500 }, { "grad_norm": 0.11842586100101471, "learning_rate": 6.518379974631156e-05, "loss": 0.0139, "step": 129510 }, { "grad_norm": 0.12300523370504379, "learning_rate": 6.51785483697274e-05, "loss": 0.0203, "step": 129520 }, { "grad_norm": 0.11616457998752594, "learning_rate": 6.517329680870956e-05, "loss": 0.0188, "step": 129530 }, { "grad_norm": 0.12540502846240997, "learning_rate": 6.516804506332183e-05, "loss": 0.0115, "step": 129540 }, { "grad_norm": 0.27105993032455444, "learning_rate": 6.516279313362804e-05, "loss": 0.013, "step": 129550 }, { "grad_norm": 0.09838273376226425, "learning_rate": 6.515754101969199e-05, "loss": 0.0169, "step": 129560 }, { "grad_norm": 0.10039836913347244, "learning_rate": 6.515228872157752e-05, "loss": 0.0233, "step": 129570 }, { "grad_norm": 0.20240096747875214, "learning_rate": 6.514703623934841e-05, "loss": 0.0178, "step": 129580 }, { "grad_norm": 0.06889315694570541, "learning_rate": 6.514178357306852e-05, "loss": 0.0156, "step": 129590 }, { "grad_norm": 0.4551399052143097, "learning_rate": 6.513653072280166e-05, "loss": 0.0217, "step": 129600 }, { "grad_norm": 0.3030335307121277, "learning_rate": 6.513127768861169e-05, "loss": 0.0103, "step": 129610 }, { "grad_norm": 0.11436525732278824, "learning_rate": 6.512602447056238e-05, "loss": 0.0162, "step": 129620 }, { "grad_norm": 0.12658748030662537, "learning_rate": 6.51207710687176e-05, "loss": 0.0142, "step": 129630 }, { "grad_norm": 0.13149812817573547, "learning_rate": 6.511551748314117e-05, "loss": 0.0137, "step": 129640 }, { "grad_norm": 0.17500872910022736, "learning_rate": 6.511026371389693e-05, "loss": 0.0224, "step": 129650 }, { "grad_norm": 0.5709513425827026, "learning_rate": 6.510500976104872e-05, "loss": 0.015, "step": 129660 }, { "grad_norm": 0.11004496365785599, "learning_rate": 6.50997556246604e-05, "loss": 0.0152, "step": 129670 }, { "grad_norm": 0.18647605180740356, "learning_rate": 6.509450130479577e-05, "loss": 0.0176, "step": 129680 }, { "grad_norm": 0.15313267707824707, "learning_rate": 6.50892468015187e-05, "loss": 0.0124, "step": 129690 }, { "grad_norm": 0.10094373673200607, "learning_rate": 6.508399211489303e-05, "loss": 0.0208, "step": 129700 }, { "grad_norm": 0.09655105322599411, "learning_rate": 6.507873724498264e-05, "loss": 0.0173, "step": 129710 }, { "grad_norm": 0.07802291959524155, "learning_rate": 6.507348219185134e-05, "loss": 0.0123, "step": 129720 }, { "grad_norm": 0.096431665122509, "learning_rate": 6.506822695556298e-05, "loss": 0.0147, "step": 129730 }, { "grad_norm": 0.14374902844429016, "learning_rate": 6.506297153618147e-05, "loss": 0.0198, "step": 129740 }, { "grad_norm": 0.0720893144607544, "learning_rate": 6.505771593377061e-05, "loss": 0.0105, "step": 129750 }, { "grad_norm": 0.1081995964050293, "learning_rate": 6.50524601483943e-05, "loss": 0.0152, "step": 129760 }, { "grad_norm": 0.15248465538024902, "learning_rate": 6.504720418011637e-05, "loss": 0.0115, "step": 129770 }, { "grad_norm": 0.22139674425125122, "learning_rate": 6.504194802900072e-05, "loss": 0.0148, "step": 129780 }, { "grad_norm": 0.11116588860750198, "learning_rate": 6.50366916951112e-05, "loss": 0.0155, "step": 129790 }, { "grad_norm": 0.19496607780456543, "learning_rate": 6.503143517851167e-05, "loss": 0.0156, "step": 129800 }, { "grad_norm": 0.12407686561346054, "learning_rate": 6.5026178479266e-05, "loss": 0.015, "step": 129810 }, { "grad_norm": 0.09452451765537262, "learning_rate": 6.502092159743809e-05, "loss": 0.015, "step": 129820 }, { "grad_norm": 0.2738456130027771, "learning_rate": 6.50156645330918e-05, "loss": 0.0153, "step": 129830 }, { "grad_norm": 0.16031138598918915, "learning_rate": 6.5010407286291e-05, "loss": 0.0095, "step": 129840 }, { "grad_norm": 0.07581064105033875, "learning_rate": 6.50051498570996e-05, "loss": 0.0083, "step": 129850 }, { "grad_norm": 0.11193782091140747, "learning_rate": 6.499989224558143e-05, "loss": 0.0255, "step": 129860 }, { "grad_norm": 0.12504692375659943, "learning_rate": 6.499463445180043e-05, "loss": 0.0173, "step": 129870 }, { "grad_norm": 0.09161633253097534, "learning_rate": 6.498937647582046e-05, "loss": 0.0137, "step": 129880 }, { "grad_norm": 0.48370152711868286, "learning_rate": 6.49841183177054e-05, "loss": 0.0143, "step": 129890 }, { "grad_norm": 0.1265556514263153, "learning_rate": 6.497885997751918e-05, "loss": 0.0148, "step": 129900 }, { "grad_norm": 0.2371896207332611, "learning_rate": 6.497360145532565e-05, "loss": 0.016, "step": 129910 }, { "grad_norm": 0.11913386732339859, "learning_rate": 6.496834275118872e-05, "loss": 0.0166, "step": 129920 }, { "grad_norm": 0.08216465264558792, "learning_rate": 6.496308386517232e-05, "loss": 0.0168, "step": 129930 }, { "grad_norm": 0.14852796494960785, "learning_rate": 6.49578247973403e-05, "loss": 0.0148, "step": 129940 }, { "grad_norm": 0.15685540437698364, "learning_rate": 6.49525655477566e-05, "loss": 0.0185, "step": 129950 }, { "grad_norm": 0.22362232208251953, "learning_rate": 6.494730611648509e-05, "loss": 0.0211, "step": 129960 }, { "grad_norm": 0.15588268637657166, "learning_rate": 6.494204650358973e-05, "loss": 0.0187, "step": 129970 }, { "grad_norm": 0.18250985443592072, "learning_rate": 6.493678670913436e-05, "loss": 0.0181, "step": 129980 }, { "grad_norm": 0.13689568638801575, "learning_rate": 6.493152673318296e-05, "loss": 0.0153, "step": 129990 }, { "grad_norm": 0.11749911308288574, "learning_rate": 6.49262665757994e-05, "loss": 0.0149, "step": 130000 }, { "grad_norm": 0.15219582617282867, "learning_rate": 6.49210062370476e-05, "loss": 0.0159, "step": 130010 }, { "grad_norm": 0.10730582475662231, "learning_rate": 6.491574571699149e-05, "loss": 0.0162, "step": 130020 }, { "grad_norm": 0.0955018550157547, "learning_rate": 6.4910485015695e-05, "loss": 0.0165, "step": 130030 }, { "grad_norm": 0.2044663280248642, "learning_rate": 6.490522413322205e-05, "loss": 0.015, "step": 130040 }, { "grad_norm": 0.31176871061325073, "learning_rate": 6.489996306963652e-05, "loss": 0.0187, "step": 130050 }, { "grad_norm": 0.1453000009059906, "learning_rate": 6.489470182500239e-05, "loss": 0.017, "step": 130060 }, { "grad_norm": 0.06824369728565216, "learning_rate": 6.488944039938355e-05, "loss": 0.0126, "step": 130070 }, { "grad_norm": 0.5323032736778259, "learning_rate": 6.488417879284399e-05, "loss": 0.0164, "step": 130080 }, { "grad_norm": 0.0786951407790184, "learning_rate": 6.487891700544757e-05, "loss": 0.0215, "step": 130090 }, { "grad_norm": 0.10720494389533997, "learning_rate": 6.487365503725827e-05, "loss": 0.0183, "step": 130100 }, { "grad_norm": 0.14460287988185883, "learning_rate": 6.486839288834003e-05, "loss": 0.0188, "step": 130110 }, { "grad_norm": 0.27500995993614197, "learning_rate": 6.486313055875676e-05, "loss": 0.0162, "step": 130120 }, { "grad_norm": 0.14454562962055206, "learning_rate": 6.485786804857244e-05, "loss": 0.0143, "step": 130130 }, { "grad_norm": 0.0706041231751442, "learning_rate": 6.485260535785099e-05, "loss": 0.0156, "step": 130140 }, { "grad_norm": 0.12973137199878693, "learning_rate": 6.484734248665634e-05, "loss": 0.0161, "step": 130150 }, { "grad_norm": 0.12600670754909515, "learning_rate": 6.484207943505249e-05, "loss": 0.0192, "step": 130160 }, { "grad_norm": 0.08627194166183472, "learning_rate": 6.483681620310334e-05, "loss": 0.0204, "step": 130170 }, { "grad_norm": 0.3103761672973633, "learning_rate": 6.483155279087287e-05, "loss": 0.0168, "step": 130180 }, { "grad_norm": 0.14810428023338318, "learning_rate": 6.482628919842503e-05, "loss": 0.0134, "step": 130190 }, { "grad_norm": 0.13005205988883972, "learning_rate": 6.48210254258238e-05, "loss": 0.0158, "step": 130200 }, { "grad_norm": 0.1565612554550171, "learning_rate": 6.481576147313308e-05, "loss": 0.0151, "step": 130210 }, { "grad_norm": 0.4676104784011841, "learning_rate": 6.48104973404169e-05, "loss": 0.0161, "step": 130220 }, { "grad_norm": 0.1300172209739685, "learning_rate": 6.480523302773918e-05, "loss": 0.02, "step": 130230 }, { "grad_norm": 0.17876030504703522, "learning_rate": 6.47999685351639e-05, "loss": 0.0183, "step": 130240 }, { "grad_norm": 0.15202033519744873, "learning_rate": 6.479470386275503e-05, "loss": 0.0179, "step": 130250 }, { "grad_norm": 0.24113374948501587, "learning_rate": 6.478943901057654e-05, "loss": 0.0181, "step": 130260 }, { "grad_norm": 0.1445942372083664, "learning_rate": 6.478417397869241e-05, "loss": 0.0141, "step": 130270 }, { "grad_norm": 0.1840343028306961, "learning_rate": 6.47789087671666e-05, "loss": 0.0136, "step": 130280 }, { "grad_norm": 0.10260491818189621, "learning_rate": 6.477364337606309e-05, "loss": 0.0165, "step": 130290 }, { "grad_norm": 0.085086390376091, "learning_rate": 6.476837780544588e-05, "loss": 0.017, "step": 130300 }, { "grad_norm": 0.11351461708545685, "learning_rate": 6.476311205537892e-05, "loss": 0.0159, "step": 130310 }, { "grad_norm": 0.1307610124349594, "learning_rate": 6.475784612592622e-05, "loss": 0.0218, "step": 130320 }, { "grad_norm": 0.161237433552742, "learning_rate": 6.475258001715175e-05, "loss": 0.023, "step": 130330 }, { "grad_norm": 0.11109907180070877, "learning_rate": 6.47473137291195e-05, "loss": 0.0142, "step": 130340 }, { "grad_norm": 0.08300323784351349, "learning_rate": 6.474204726189348e-05, "loss": 0.0132, "step": 130350 }, { "grad_norm": 0.07452141493558884, "learning_rate": 6.473678061553765e-05, "loss": 0.0139, "step": 130360 }, { "grad_norm": 0.19793756306171417, "learning_rate": 6.473151379011603e-05, "loss": 0.0247, "step": 130370 }, { "grad_norm": 0.2026573270559311, "learning_rate": 6.47262467856926e-05, "loss": 0.0171, "step": 130380 }, { "grad_norm": 0.11159732192754745, "learning_rate": 6.472097960233139e-05, "loss": 0.0163, "step": 130390 }, { "grad_norm": 0.09872892498970032, "learning_rate": 6.471571224009636e-05, "loss": 0.0145, "step": 130400 }, { "grad_norm": 0.1606181114912033, "learning_rate": 6.471044469905153e-05, "loss": 0.0152, "step": 130410 }, { "grad_norm": 0.0849386528134346, "learning_rate": 6.470517697926093e-05, "loss": 0.013, "step": 130420 }, { "grad_norm": 0.11498191207647324, "learning_rate": 6.469990908078854e-05, "loss": 0.0138, "step": 130430 }, { "grad_norm": 0.1838086098432541, "learning_rate": 6.469464100369835e-05, "loss": 0.0157, "step": 130440 }, { "grad_norm": 0.06546059250831604, "learning_rate": 6.468937274805442e-05, "loss": 0.0155, "step": 130450 }, { "grad_norm": 0.06593454629182816, "learning_rate": 6.468410431392072e-05, "loss": 0.0183, "step": 130460 }, { "grad_norm": 0.0902574211359024, "learning_rate": 6.467883570136132e-05, "loss": 0.013, "step": 130470 }, { "grad_norm": 0.16538818180561066, "learning_rate": 6.467356691044019e-05, "loss": 0.0141, "step": 130480 }, { "grad_norm": 0.1226019635796547, "learning_rate": 6.466829794122136e-05, "loss": 0.019, "step": 130490 }, { "grad_norm": 0.07337793707847595, "learning_rate": 6.466302879376887e-05, "loss": 0.0133, "step": 130500 }, { "grad_norm": 0.09019973874092102, "learning_rate": 6.465775946814674e-05, "loss": 0.0161, "step": 130510 }, { "grad_norm": 0.10920966416597366, "learning_rate": 6.465248996441899e-05, "loss": 0.0169, "step": 130520 }, { "grad_norm": 0.06290100514888763, "learning_rate": 6.464722028264966e-05, "loss": 0.012, "step": 130530 }, { "grad_norm": 0.11881177872419357, "learning_rate": 6.464195042290277e-05, "loss": 0.0177, "step": 130540 }, { "grad_norm": 0.35240235924720764, "learning_rate": 6.463668038524234e-05, "loss": 0.0148, "step": 130550 }, { "grad_norm": 0.11204811185598373, "learning_rate": 6.463141016973245e-05, "loss": 0.0228, "step": 130560 }, { "grad_norm": 0.10185608267784119, "learning_rate": 6.46261397764371e-05, "loss": 0.0151, "step": 130570 }, { "grad_norm": 0.21421383321285248, "learning_rate": 6.462086920542034e-05, "loss": 0.0174, "step": 130580 }, { "grad_norm": 0.15976464748382568, "learning_rate": 6.461559845674621e-05, "loss": 0.0108, "step": 130590 }, { "grad_norm": 0.23458978533744812, "learning_rate": 6.461032753047878e-05, "loss": 0.021, "step": 130600 }, { "grad_norm": 0.5308149456977844, "learning_rate": 6.460505642668206e-05, "loss": 0.0218, "step": 130610 }, { "grad_norm": 0.12336402386426926, "learning_rate": 6.459978514542011e-05, "loss": 0.0179, "step": 130620 }, { "grad_norm": 0.07444779574871063, "learning_rate": 6.4594513686757e-05, "loss": 0.0149, "step": 130630 }, { "grad_norm": 0.14421452581882477, "learning_rate": 6.458924205075676e-05, "loss": 0.0126, "step": 130640 }, { "grad_norm": 0.11314789205789566, "learning_rate": 6.458397023748346e-05, "loss": 0.02, "step": 130650 }, { "grad_norm": 0.11479543894529343, "learning_rate": 6.457869824700116e-05, "loss": 0.0166, "step": 130660 }, { "grad_norm": 0.11287058889865875, "learning_rate": 6.45734260793739e-05, "loss": 0.0118, "step": 130670 }, { "grad_norm": 0.1150592714548111, "learning_rate": 6.456815373466575e-05, "loss": 0.0184, "step": 130680 }, { "grad_norm": 0.10776633769273758, "learning_rate": 6.456288121294079e-05, "loss": 0.017, "step": 130690 }, { "grad_norm": 0.16918674111366272, "learning_rate": 6.455760851426307e-05, "loss": 0.0145, "step": 130700 }, { "grad_norm": 0.10199128836393356, "learning_rate": 6.455233563869665e-05, "loss": 0.0137, "step": 130710 }, { "grad_norm": 0.08615708351135254, "learning_rate": 6.454706258630564e-05, "loss": 0.0135, "step": 130720 }, { "grad_norm": 0.0844048336148262, "learning_rate": 6.454178935715405e-05, "loss": 0.0167, "step": 130730 }, { "grad_norm": 0.6537627577781677, "learning_rate": 6.4536515951306e-05, "loss": 0.0179, "step": 130740 }, { "grad_norm": 0.15112446248531342, "learning_rate": 6.453124236882557e-05, "loss": 0.0134, "step": 130750 }, { "grad_norm": 0.15573205053806305, "learning_rate": 6.452596860977681e-05, "loss": 0.0199, "step": 130760 }, { "grad_norm": 0.1220470443367958, "learning_rate": 6.452069467422384e-05, "loss": 0.012, "step": 130770 }, { "grad_norm": 0.05454432964324951, "learning_rate": 6.451542056223069e-05, "loss": 0.0177, "step": 130780 }, { "grad_norm": 0.11563192307949066, "learning_rate": 6.451014627386149e-05, "loss": 0.018, "step": 130790 }, { "grad_norm": 0.1264733374118805, "learning_rate": 6.45048718091803e-05, "loss": 0.0135, "step": 130800 }, { "grad_norm": 0.10845544934272766, "learning_rate": 6.449959716825124e-05, "loss": 0.0147, "step": 130810 }, { "grad_norm": 0.12958092987537384, "learning_rate": 6.449432235113839e-05, "loss": 0.0131, "step": 130820 }, { "grad_norm": 0.24519145488739014, "learning_rate": 6.448904735790582e-05, "loss": 0.0134, "step": 130830 }, { "grad_norm": 0.18358495831489563, "learning_rate": 6.448377218861765e-05, "loss": 0.0155, "step": 130840 }, { "grad_norm": 0.13775834441184998, "learning_rate": 6.447849684333797e-05, "loss": 0.0168, "step": 130850 }, { "grad_norm": 0.4135674834251404, "learning_rate": 6.44732213221309e-05, "loss": 0.0241, "step": 130860 }, { "grad_norm": 0.2114391326904297, "learning_rate": 6.44679456250605e-05, "loss": 0.0135, "step": 130870 }, { "grad_norm": 0.12237229198217392, "learning_rate": 6.446266975219093e-05, "loss": 0.0183, "step": 130880 }, { "grad_norm": 0.19298849999904633, "learning_rate": 6.445739370358625e-05, "loss": 0.0144, "step": 130890 }, { "grad_norm": 0.25239482522010803, "learning_rate": 6.44521174793106e-05, "loss": 0.0147, "step": 130900 }, { "grad_norm": 0.11437389254570007, "learning_rate": 6.444684107942805e-05, "loss": 0.012, "step": 130910 }, { "grad_norm": 0.2908707559108734, "learning_rate": 6.444156450400276e-05, "loss": 0.0162, "step": 130920 }, { "grad_norm": 0.15024983882904053, "learning_rate": 6.443628775309883e-05, "loss": 0.0201, "step": 130930 }, { "grad_norm": 0.11721563339233398, "learning_rate": 6.443101082678036e-05, "loss": 0.0186, "step": 130940 }, { "grad_norm": 0.13579921424388885, "learning_rate": 6.44257337251115e-05, "loss": 0.0151, "step": 130950 }, { "grad_norm": 0.13867966830730438, "learning_rate": 6.442045644815634e-05, "loss": 0.0151, "step": 130960 }, { "grad_norm": 0.1003974974155426, "learning_rate": 6.441517899597904e-05, "loss": 0.0161, "step": 130970 }, { "grad_norm": 0.09250376373529434, "learning_rate": 6.440990136864368e-05, "loss": 0.0179, "step": 130980 }, { "grad_norm": 0.254675030708313, "learning_rate": 6.440462356621441e-05, "loss": 0.0147, "step": 130990 }, { "grad_norm": 0.9934236407279968, "learning_rate": 6.439934558875537e-05, "loss": 0.0162, "step": 131000 }, { "grad_norm": 0.5884169340133667, "learning_rate": 6.439406743633071e-05, "loss": 0.0135, "step": 131010 }, { "grad_norm": 0.08073776215314865, "learning_rate": 6.438878910900452e-05, "loss": 0.0169, "step": 131020 }, { "grad_norm": 0.40555131435394287, "learning_rate": 6.438351060684096e-05, "loss": 0.0191, "step": 131030 }, { "grad_norm": 0.3024228513240814, "learning_rate": 6.437823192990415e-05, "loss": 0.0149, "step": 131040 }, { "grad_norm": 0.26492634415626526, "learning_rate": 6.437295307825828e-05, "loss": 0.0184, "step": 131050 }, { "grad_norm": 0.08600039035081863, "learning_rate": 6.436767405196743e-05, "loss": 0.0219, "step": 131060 }, { "grad_norm": 0.2703291177749634, "learning_rate": 6.43623948510958e-05, "loss": 0.0133, "step": 131070 }, { "grad_norm": 0.11877399682998657, "learning_rate": 6.435711547570748e-05, "loss": 0.0219, "step": 131080 }, { "grad_norm": 0.40356287360191345, "learning_rate": 6.435183592586669e-05, "loss": 0.0204, "step": 131090 }, { "grad_norm": 0.12720973789691925, "learning_rate": 6.43465562016375e-05, "loss": 0.0146, "step": 131100 }, { "grad_norm": 0.11718939244747162, "learning_rate": 6.434127630308416e-05, "loss": 0.0149, "step": 131110 }, { "grad_norm": 0.09346279501914978, "learning_rate": 6.433599623027072e-05, "loss": 0.0207, "step": 131120 }, { "grad_norm": 0.21943189203739166, "learning_rate": 6.433071598326142e-05, "loss": 0.0165, "step": 131130 }, { "grad_norm": 0.11784207075834274, "learning_rate": 6.432543556212038e-05, "loss": 0.0149, "step": 131140 }, { "grad_norm": 0.28033727407455444, "learning_rate": 6.432015496691179e-05, "loss": 0.0175, "step": 131150 }, { "grad_norm": 0.11154861748218536, "learning_rate": 6.431487419769978e-05, "loss": 0.0183, "step": 131160 }, { "grad_norm": 0.10523923486471176, "learning_rate": 6.430959325454852e-05, "loss": 0.0126, "step": 131170 }, { "grad_norm": 0.07645434886217117, "learning_rate": 6.430431213752222e-05, "loss": 0.0135, "step": 131180 }, { "grad_norm": 0.13703610002994537, "learning_rate": 6.4299030846685e-05, "loss": 0.0135, "step": 131190 }, { "grad_norm": 0.1329515427350998, "learning_rate": 6.429374938210108e-05, "loss": 0.0154, "step": 131200 }, { "grad_norm": 0.15675684809684753, "learning_rate": 6.42884677438346e-05, "loss": 0.0166, "step": 131210 }, { "grad_norm": 0.09454845637083054, "learning_rate": 6.428318593194974e-05, "loss": 0.0179, "step": 131220 }, { "grad_norm": 0.19543807208538055, "learning_rate": 6.427790394651067e-05, "loss": 0.0161, "step": 131230 }, { "grad_norm": 0.08630012720823288, "learning_rate": 6.42726217875816e-05, "loss": 0.0191, "step": 131240 }, { "grad_norm": 0.19043795764446259, "learning_rate": 6.426733945522672e-05, "loss": 0.0218, "step": 131250 }, { "grad_norm": 0.1256001740694046, "learning_rate": 6.426205694951016e-05, "loss": 0.0174, "step": 131260 }, { "grad_norm": 0.09619442373514175, "learning_rate": 6.425677427049616e-05, "loss": 0.0106, "step": 131270 }, { "grad_norm": 0.27371925115585327, "learning_rate": 6.425149141824889e-05, "loss": 0.0153, "step": 131280 }, { "grad_norm": 0.18685385584831238, "learning_rate": 6.424620839283255e-05, "loss": 0.0139, "step": 131290 }, { "grad_norm": 0.13029181957244873, "learning_rate": 6.424092519431132e-05, "loss": 0.0187, "step": 131300 }, { "grad_norm": 0.15059775114059448, "learning_rate": 6.423564182274942e-05, "loss": 0.0192, "step": 131310 }, { "grad_norm": 0.07952716946601868, "learning_rate": 6.423035827821101e-05, "loss": 0.011, "step": 131320 }, { "grad_norm": 0.0933355838060379, "learning_rate": 6.42250745607603e-05, "loss": 0.0115, "step": 131330 }, { "grad_norm": 0.4342159032821655, "learning_rate": 6.421979067046153e-05, "loss": 0.0155, "step": 131340 }, { "grad_norm": 0.14789694547653198, "learning_rate": 6.421450660737887e-05, "loss": 0.0149, "step": 131350 }, { "grad_norm": 0.18441584706306458, "learning_rate": 6.420922237157654e-05, "loss": 0.0135, "step": 131360 }, { "grad_norm": 0.18831227719783783, "learning_rate": 6.420393796311874e-05, "loss": 0.0215, "step": 131370 }, { "grad_norm": 0.2342681586742401, "learning_rate": 6.419865338206968e-05, "loss": 0.0234, "step": 131380 }, { "grad_norm": 0.5413998365402222, "learning_rate": 6.419336862849356e-05, "loss": 0.022, "step": 131390 }, { "grad_norm": 0.15398955345153809, "learning_rate": 6.418808370245462e-05, "loss": 0.0239, "step": 131400 }, { "grad_norm": 0.15202300250530243, "learning_rate": 6.418279860401708e-05, "loss": 0.022, "step": 131410 }, { "grad_norm": 0.18438270688056946, "learning_rate": 6.417751333324513e-05, "loss": 0.0173, "step": 131420 }, { "grad_norm": 0.17047779262065887, "learning_rate": 6.417222789020303e-05, "loss": 0.0202, "step": 131430 }, { "grad_norm": 0.12168695032596588, "learning_rate": 6.416694227495496e-05, "loss": 0.0166, "step": 131440 }, { "grad_norm": 0.08750417083501816, "learning_rate": 6.416165648756517e-05, "loss": 0.0103, "step": 131450 }, { "grad_norm": 0.11941908299922943, "learning_rate": 6.415637052809788e-05, "loss": 0.0186, "step": 131460 }, { "grad_norm": 0.12741637229919434, "learning_rate": 6.415108439661733e-05, "loss": 0.0155, "step": 131470 }, { "grad_norm": 0.06799318641424179, "learning_rate": 6.414579809318775e-05, "loss": 0.0134, "step": 131480 }, { "grad_norm": 0.11927909404039383, "learning_rate": 6.414051161787334e-05, "loss": 0.0207, "step": 131490 }, { "grad_norm": 0.22441932559013367, "learning_rate": 6.413522497073839e-05, "loss": 0.0138, "step": 131500 }, { "grad_norm": 0.14781078696250916, "learning_rate": 6.41299381518471e-05, "loss": 0.0172, "step": 131510 }, { "grad_norm": 0.12275878340005875, "learning_rate": 6.412465116126371e-05, "loss": 0.0166, "step": 131520 }, { "grad_norm": 0.09963709115982056, "learning_rate": 6.41193639990525e-05, "loss": 0.0153, "step": 131530 }, { "grad_norm": 0.22121040523052216, "learning_rate": 6.411407666527767e-05, "loss": 0.0175, "step": 131540 }, { "grad_norm": 0.09094469994306564, "learning_rate": 6.410878916000348e-05, "loss": 0.0189, "step": 131550 }, { "grad_norm": 0.09442993998527527, "learning_rate": 6.410350148329419e-05, "loss": 0.0163, "step": 131560 }, { "grad_norm": 0.24365828931331635, "learning_rate": 6.409821363521403e-05, "loss": 0.0154, "step": 131570 }, { "grad_norm": 0.1143803596496582, "learning_rate": 6.409292561582727e-05, "loss": 0.0143, "step": 131580 }, { "grad_norm": 0.149146169424057, "learning_rate": 6.408763742519815e-05, "loss": 0.0177, "step": 131590 }, { "grad_norm": 0.08555668592453003, "learning_rate": 6.408234906339094e-05, "loss": 0.0155, "step": 131600 }, { "grad_norm": 0.137079656124115, "learning_rate": 6.40770605304699e-05, "loss": 0.0126, "step": 131610 }, { "grad_norm": 0.31778135895729065, "learning_rate": 6.407177182649927e-05, "loss": 0.0187, "step": 131620 }, { "grad_norm": 0.12260105460882187, "learning_rate": 6.406648295154333e-05, "loss": 0.0145, "step": 131630 }, { "grad_norm": 0.0828658789396286, "learning_rate": 6.406119390566637e-05, "loss": 0.017, "step": 131640 }, { "grad_norm": 0.1396913379430771, "learning_rate": 6.405590468893258e-05, "loss": 0.0221, "step": 131650 }, { "grad_norm": 0.16601203382015228, "learning_rate": 6.40506153014063e-05, "loss": 0.0152, "step": 131660 }, { "grad_norm": 0.13951052725315094, "learning_rate": 6.404532574315179e-05, "loss": 0.0194, "step": 131670 }, { "grad_norm": 0.16954800486564636, "learning_rate": 6.404003601423329e-05, "loss": 0.0134, "step": 131680 }, { "grad_norm": 0.27610698342323303, "learning_rate": 6.40347461147151e-05, "loss": 0.0171, "step": 131690 }, { "grad_norm": 0.08388873934745789, "learning_rate": 6.40294560446615e-05, "loss": 0.0162, "step": 131700 }, { "grad_norm": 0.17275522649288177, "learning_rate": 6.402416580413676e-05, "loss": 0.0125, "step": 131710 }, { "grad_norm": 0.10676868259906769, "learning_rate": 6.401887539320517e-05, "loss": 0.0182, "step": 131720 }, { "grad_norm": 0.09625512361526489, "learning_rate": 6.4013584811931e-05, "loss": 0.0188, "step": 131730 }, { "grad_norm": 0.15359729528427124, "learning_rate": 6.400829406037856e-05, "loss": 0.0196, "step": 131740 }, { "grad_norm": 0.13201968371868134, "learning_rate": 6.40030031386121e-05, "loss": 0.013, "step": 131750 }, { "grad_norm": 0.12242525815963745, "learning_rate": 6.399771204669594e-05, "loss": 0.0185, "step": 131760 }, { "grad_norm": 0.14594796299934387, "learning_rate": 6.399242078469437e-05, "loss": 0.0201, "step": 131770 }, { "grad_norm": 0.11571234464645386, "learning_rate": 6.398712935267166e-05, "loss": 0.0153, "step": 131780 }, { "grad_norm": 0.1853780448436737, "learning_rate": 6.398183775069214e-05, "loss": 0.0225, "step": 131790 }, { "grad_norm": 0.11108104884624481, "learning_rate": 6.397654597882008e-05, "loss": 0.0153, "step": 131800 }, { "grad_norm": 0.23244497179985046, "learning_rate": 6.397125403711979e-05, "loss": 0.017, "step": 131810 }, { "grad_norm": 0.12894895672798157, "learning_rate": 6.396596192565557e-05, "loss": 0.0188, "step": 131820 }, { "grad_norm": 0.08769233524799347, "learning_rate": 6.396066964449173e-05, "loss": 0.0135, "step": 131830 }, { "grad_norm": 0.08140871673822403, "learning_rate": 6.395537719369257e-05, "loss": 0.012, "step": 131840 }, { "grad_norm": 0.16055089235305786, "learning_rate": 6.395008457332241e-05, "loss": 0.0206, "step": 131850 }, { "grad_norm": 0.15289326012134552, "learning_rate": 6.394479178344554e-05, "loss": 0.0158, "step": 131860 }, { "grad_norm": 0.10557471215724945, "learning_rate": 6.393949882412629e-05, "loss": 0.0184, "step": 131870 }, { "grad_norm": 0.20393121242523193, "learning_rate": 6.393420569542897e-05, "loss": 0.0245, "step": 131880 }, { "grad_norm": 0.09388589859008789, "learning_rate": 6.39289123974179e-05, "loss": 0.0164, "step": 131890 }, { "grad_norm": 0.1621299684047699, "learning_rate": 6.392361893015737e-05, "loss": 0.013, "step": 131900 }, { "grad_norm": 0.15150801837444305, "learning_rate": 6.391832529371176e-05, "loss": 0.0162, "step": 131910 }, { "grad_norm": 0.15530650317668915, "learning_rate": 6.391303148814534e-05, "loss": 0.0165, "step": 131920 }, { "grad_norm": 0.16215041279792786, "learning_rate": 6.390773751352243e-05, "loss": 0.014, "step": 131930 }, { "grad_norm": 0.23337198793888092, "learning_rate": 6.390244336990741e-05, "loss": 0.01, "step": 131940 }, { "grad_norm": 0.08619735389947891, "learning_rate": 6.389714905736456e-05, "loss": 0.0155, "step": 131950 }, { "grad_norm": 0.12977714836597443, "learning_rate": 6.389185457595823e-05, "loss": 0.0166, "step": 131960 }, { "grad_norm": 0.06803252547979355, "learning_rate": 6.388655992575275e-05, "loss": 0.0135, "step": 131970 }, { "grad_norm": 0.2612632215023041, "learning_rate": 6.388126510681247e-05, "loss": 0.0185, "step": 131980 }, { "grad_norm": 0.11962917447090149, "learning_rate": 6.38759701192017e-05, "loss": 0.0152, "step": 131990 }, { "grad_norm": 0.1607922464609146, "learning_rate": 6.387067496298479e-05, "loss": 0.0162, "step": 132000 }, { "grad_norm": 0.09071134775876999, "learning_rate": 6.386537963822607e-05, "loss": 0.0126, "step": 132010 }, { "grad_norm": 0.2777434289455414, "learning_rate": 6.386008414498991e-05, "loss": 0.0163, "step": 132020 }, { "grad_norm": 0.155208021402359, "learning_rate": 6.385478848334066e-05, "loss": 0.0127, "step": 132030 }, { "grad_norm": 0.16516371071338654, "learning_rate": 6.384949265334263e-05, "loss": 0.013, "step": 132040 }, { "grad_norm": 0.18457616865634918, "learning_rate": 6.384419665506019e-05, "loss": 0.016, "step": 132050 }, { "grad_norm": 0.06413548439741135, "learning_rate": 6.38389004885577e-05, "loss": 0.0171, "step": 132060 }, { "grad_norm": 0.1123218759894371, "learning_rate": 6.383360415389949e-05, "loss": 0.0114, "step": 132070 }, { "grad_norm": 0.12021854519844055, "learning_rate": 6.382830765114992e-05, "loss": 0.0131, "step": 132080 }, { "grad_norm": 0.10943297296762466, "learning_rate": 6.382301098037338e-05, "loss": 0.0126, "step": 132090 }, { "grad_norm": 0.3271220624446869, "learning_rate": 6.381771414163418e-05, "loss": 0.0151, "step": 132100 }, { "grad_norm": 0.12109902501106262, "learning_rate": 6.381241713499673e-05, "loss": 0.015, "step": 132110 }, { "grad_norm": 0.131669819355011, "learning_rate": 6.380711996052535e-05, "loss": 0.013, "step": 132120 }, { "grad_norm": 0.12797760963439941, "learning_rate": 6.380182261828445e-05, "loss": 0.0136, "step": 132130 }, { "grad_norm": 0.25638166069984436, "learning_rate": 6.379652510833836e-05, "loss": 0.0218, "step": 132140 }, { "grad_norm": 0.10185583680868149, "learning_rate": 6.379122743075145e-05, "loss": 0.0114, "step": 132150 }, { "grad_norm": 0.15586712956428528, "learning_rate": 6.378592958558814e-05, "loss": 0.0121, "step": 132160 }, { "grad_norm": 0.20635078847408295, "learning_rate": 6.378063157291275e-05, "loss": 0.015, "step": 132170 }, { "grad_norm": 0.11892857402563095, "learning_rate": 6.377533339278967e-05, "loss": 0.0153, "step": 132180 }, { "grad_norm": 0.1207534596323967, "learning_rate": 6.377003504528328e-05, "loss": 0.0153, "step": 132190 }, { "grad_norm": 0.14614947140216827, "learning_rate": 6.376473653045798e-05, "loss": 0.0185, "step": 132200 }, { "grad_norm": 0.7875719666481018, "learning_rate": 6.375943784837813e-05, "loss": 0.0203, "step": 132210 }, { "grad_norm": 0.13230864703655243, "learning_rate": 6.375413899910812e-05, "loss": 0.0141, "step": 132220 }, { "grad_norm": 0.10133618861436844, "learning_rate": 6.374883998271233e-05, "loss": 0.0124, "step": 132230 }, { "grad_norm": 0.10792959481477737, "learning_rate": 6.374354079925515e-05, "loss": 0.0146, "step": 132240 }, { "grad_norm": 0.5579195022583008, "learning_rate": 6.373824144880098e-05, "loss": 0.0196, "step": 132250 }, { "grad_norm": 0.06958629935979843, "learning_rate": 6.373294193141419e-05, "loss": 0.015, "step": 132260 }, { "grad_norm": 0.17052707076072693, "learning_rate": 6.372764224715922e-05, "loss": 0.0197, "step": 132270 }, { "grad_norm": 0.31531205773353577, "learning_rate": 6.37223423961004e-05, "loss": 0.0156, "step": 132280 }, { "grad_norm": 0.0959954634308815, "learning_rate": 6.371704237830219e-05, "loss": 0.0137, "step": 132290 }, { "grad_norm": 0.12951119244098663, "learning_rate": 6.371174219382897e-05, "loss": 0.0167, "step": 132300 }, { "grad_norm": 0.4519612789154053, "learning_rate": 6.37064418427451e-05, "loss": 0.0162, "step": 132310 }, { "grad_norm": 0.10163760930299759, "learning_rate": 6.370114132511505e-05, "loss": 0.0139, "step": 132320 }, { "grad_norm": 0.15507404506206512, "learning_rate": 6.369584064100318e-05, "loss": 0.0129, "step": 132330 }, { "grad_norm": 0.13697868585586548, "learning_rate": 6.369053979047395e-05, "loss": 0.015, "step": 132340 }, { "grad_norm": 0.15708027780056, "learning_rate": 6.368523877359169e-05, "loss": 0.0142, "step": 132350 }, { "grad_norm": 0.14049817621707916, "learning_rate": 6.36799375904209e-05, "loss": 0.0109, "step": 132360 }, { "grad_norm": 0.22900478541851044, "learning_rate": 6.367463624102592e-05, "loss": 0.0174, "step": 132370 }, { "grad_norm": 0.19266663491725922, "learning_rate": 6.366933472547121e-05, "loss": 0.0208, "step": 132380 }, { "grad_norm": 0.13598069548606873, "learning_rate": 6.366403304382118e-05, "loss": 0.0202, "step": 132390 }, { "grad_norm": 0.07628843933343887, "learning_rate": 6.365873119614024e-05, "loss": 0.0156, "step": 132400 }, { "grad_norm": 0.19182242453098297, "learning_rate": 6.365342918249283e-05, "loss": 0.0157, "step": 132410 }, { "grad_norm": 0.1735111027956009, "learning_rate": 6.364812700294334e-05, "loss": 0.0151, "step": 132420 }, { "grad_norm": 0.16181352734565735, "learning_rate": 6.364282465755626e-05, "loss": 0.0168, "step": 132430 }, { "grad_norm": 0.19716663658618927, "learning_rate": 6.363752214639595e-05, "loss": 0.0138, "step": 132440 }, { "grad_norm": 0.12498848885297775, "learning_rate": 6.363221946952688e-05, "loss": 0.0149, "step": 132450 }, { "grad_norm": 0.16595225036144257, "learning_rate": 6.362691662701347e-05, "loss": 0.0176, "step": 132460 }, { "grad_norm": 0.14634138345718384, "learning_rate": 6.362161361892017e-05, "loss": 0.0184, "step": 132470 }, { "grad_norm": 0.08876347541809082, "learning_rate": 6.361631044531138e-05, "loss": 0.015, "step": 132480 }, { "grad_norm": 0.16483156383037567, "learning_rate": 6.361100710625158e-05, "loss": 0.0223, "step": 132490 }, { "grad_norm": 0.1317223161458969, "learning_rate": 6.360570360180518e-05, "loss": 0.0171, "step": 132500 }, { "grad_norm": 0.10120637714862823, "learning_rate": 6.360039993203664e-05, "loss": 0.0137, "step": 132510 }, { "grad_norm": 0.09445735067129135, "learning_rate": 6.35950960970104e-05, "loss": 0.0119, "step": 132520 }, { "grad_norm": 0.0785437747836113, "learning_rate": 6.358979209679091e-05, "loss": 0.0144, "step": 132530 }, { "grad_norm": 0.1113806962966919, "learning_rate": 6.35844879314426e-05, "loss": 0.0137, "step": 132540 }, { "grad_norm": 0.11573170870542526, "learning_rate": 6.357918360102996e-05, "loss": 0.0154, "step": 132550 }, { "grad_norm": 0.09710977971553802, "learning_rate": 6.357387910561741e-05, "loss": 0.0129, "step": 132560 }, { "grad_norm": 0.3051455616950989, "learning_rate": 6.356857444526941e-05, "loss": 0.0203, "step": 132570 }, { "grad_norm": 0.14989367127418518, "learning_rate": 6.356326962005041e-05, "loss": 0.0302, "step": 132580 }, { "grad_norm": 0.08966211974620819, "learning_rate": 6.355796463002491e-05, "loss": 0.017, "step": 132590 }, { "grad_norm": 0.11701865494251251, "learning_rate": 6.35526594752573e-05, "loss": 0.0148, "step": 132600 }, { "grad_norm": 0.4339417815208435, "learning_rate": 6.35473541558121e-05, "loss": 0.017, "step": 132610 }, { "grad_norm": 0.053806938230991364, "learning_rate": 6.354204867175376e-05, "loss": 0.0137, "step": 132620 }, { "grad_norm": 0.0924120768904686, "learning_rate": 6.353674302314674e-05, "loss": 0.0123, "step": 132630 }, { "grad_norm": 0.07102454453706741, "learning_rate": 6.353143721005551e-05, "loss": 0.0161, "step": 132640 }, { "grad_norm": 0.5554786324501038, "learning_rate": 6.352613123254454e-05, "loss": 0.0226, "step": 132650 }, { "grad_norm": 0.08935004472732544, "learning_rate": 6.35208250906783e-05, "loss": 0.0124, "step": 132660 }, { "grad_norm": 0.13112996518611908, "learning_rate": 6.351551878452127e-05, "loss": 0.0149, "step": 132670 }, { "grad_norm": 0.2314896136522293, "learning_rate": 6.351021231413794e-05, "loss": 0.0206, "step": 132680 }, { "grad_norm": 0.12594665586948395, "learning_rate": 6.350490567959277e-05, "loss": 0.0167, "step": 132690 }, { "grad_norm": 0.11322697252035141, "learning_rate": 6.349959888095024e-05, "loss": 0.0176, "step": 132700 }, { "grad_norm": 0.241632878780365, "learning_rate": 6.349429191827486e-05, "loss": 0.0162, "step": 132710 }, { "grad_norm": 0.18710772693157196, "learning_rate": 6.348898479163105e-05, "loss": 0.0163, "step": 132720 }, { "grad_norm": 0.1426340937614441, "learning_rate": 6.348367750108337e-05, "loss": 0.0152, "step": 132730 }, { "grad_norm": 0.09738316386938095, "learning_rate": 6.347837004669625e-05, "loss": 0.0135, "step": 132740 }, { "grad_norm": 0.15061432123184204, "learning_rate": 6.347306242853425e-05, "loss": 0.015, "step": 132750 }, { "grad_norm": 0.21144218742847443, "learning_rate": 6.346775464666179e-05, "loss": 0.0253, "step": 132760 }, { "grad_norm": 0.19363148510456085, "learning_rate": 6.346244670114341e-05, "loss": 0.0182, "step": 132770 }, { "grad_norm": 0.23495101928710938, "learning_rate": 6.345713859204359e-05, "loss": 0.02, "step": 132780 }, { "grad_norm": 0.10767317563295364, "learning_rate": 6.345183031942681e-05, "loss": 0.0195, "step": 132790 }, { "grad_norm": 0.10213270038366318, "learning_rate": 6.344652188335762e-05, "loss": 0.0122, "step": 132800 }, { "grad_norm": 0.0897679552435875, "learning_rate": 6.344121328390047e-05, "loss": 0.012, "step": 132810 }, { "grad_norm": 0.5246968269348145, "learning_rate": 6.34359045211199e-05, "loss": 0.0191, "step": 132820 }, { "grad_norm": 0.1767932027578354, "learning_rate": 6.34305955950804e-05, "loss": 0.0112, "step": 132830 }, { "grad_norm": 0.11113735288381577, "learning_rate": 6.34252865058465e-05, "loss": 0.0179, "step": 132840 }, { "grad_norm": 0.11556914448738098, "learning_rate": 6.341997725348268e-05, "loss": 0.0179, "step": 132850 }, { "grad_norm": 0.1752251237630844, "learning_rate": 6.341466783805345e-05, "loss": 0.0193, "step": 132860 }, { "grad_norm": 0.19393625855445862, "learning_rate": 6.340935825962336e-05, "loss": 0.0229, "step": 132870 }, { "grad_norm": 0.08926175534725189, "learning_rate": 6.34040485182569e-05, "loss": 0.0139, "step": 132880 }, { "grad_norm": 0.1353200376033783, "learning_rate": 6.33987386140186e-05, "loss": 0.0164, "step": 132890 }, { "grad_norm": 0.1277129203081131, "learning_rate": 6.339342854697297e-05, "loss": 0.0171, "step": 132900 }, { "grad_norm": 0.20229656994342804, "learning_rate": 6.338811831718454e-05, "loss": 0.0149, "step": 132910 }, { "grad_norm": 0.10476810485124588, "learning_rate": 6.338280792471782e-05, "loss": 0.0171, "step": 132920 }, { "grad_norm": 0.10943479835987091, "learning_rate": 6.337749736963738e-05, "loss": 0.0167, "step": 132930 }, { "grad_norm": 0.11045479774475098, "learning_rate": 6.33721866520077e-05, "loss": 0.0132, "step": 132940 }, { "grad_norm": 0.1954442262649536, "learning_rate": 6.336687577189331e-05, "loss": 0.016, "step": 132950 }, { "grad_norm": 0.22721081972122192, "learning_rate": 6.336156472935877e-05, "loss": 0.0135, "step": 132960 }, { "grad_norm": 0.12773603200912476, "learning_rate": 6.33562535244686e-05, "loss": 0.0176, "step": 132970 }, { "grad_norm": 0.17153535783290863, "learning_rate": 6.335094215728735e-05, "loss": 0.0212, "step": 132980 }, { "grad_norm": 0.0768813043832779, "learning_rate": 6.334563062787953e-05, "loss": 0.0135, "step": 132990 }, { "grad_norm": 0.12335756421089172, "learning_rate": 6.334031893630972e-05, "loss": 0.0168, "step": 133000 }, { "grad_norm": 0.09671197086572647, "learning_rate": 6.333500708264243e-05, "loss": 0.0216, "step": 133010 }, { "grad_norm": 0.18505747616291046, "learning_rate": 6.332969506694222e-05, "loss": 0.0168, "step": 133020 }, { "grad_norm": 0.08684190362691879, "learning_rate": 6.332438288927362e-05, "loss": 0.0107, "step": 133030 }, { "grad_norm": 0.134268119931221, "learning_rate": 6.331907054970118e-05, "loss": 0.0201, "step": 133040 }, { "grad_norm": 0.3023369312286377, "learning_rate": 6.331375804828947e-05, "loss": 0.0213, "step": 133050 }, { "grad_norm": 0.08843742311000824, "learning_rate": 6.330844538510304e-05, "loss": 0.0125, "step": 133060 }, { "grad_norm": 0.19457952678203583, "learning_rate": 6.330313256020641e-05, "loss": 0.0146, "step": 133070 }, { "grad_norm": 0.1071019321680069, "learning_rate": 6.329781957366416e-05, "loss": 0.0334, "step": 133080 }, { "grad_norm": 0.10671307891607285, "learning_rate": 6.329250642554087e-05, "loss": 0.0152, "step": 133090 }, { "grad_norm": 0.23995709419250488, "learning_rate": 6.328719311590107e-05, "loss": 0.0178, "step": 133100 }, { "grad_norm": 0.11043597757816315, "learning_rate": 6.328187964480931e-05, "loss": 0.0126, "step": 133110 }, { "grad_norm": 0.113658607006073, "learning_rate": 6.327656601233018e-05, "loss": 0.0154, "step": 133120 }, { "grad_norm": 0.08549928665161133, "learning_rate": 6.327125221852825e-05, "loss": 0.0103, "step": 133130 }, { "grad_norm": 0.10219614207744598, "learning_rate": 6.326593826346806e-05, "loss": 0.0147, "step": 133140 }, { "grad_norm": 0.15848246216773987, "learning_rate": 6.326062414721419e-05, "loss": 0.0203, "step": 133150 }, { "grad_norm": 0.061702147126197815, "learning_rate": 6.325530986983124e-05, "loss": 0.014, "step": 133160 }, { "grad_norm": 0.14844605326652527, "learning_rate": 6.324999543138374e-05, "loss": 0.0167, "step": 133170 }, { "grad_norm": 0.13426271080970764, "learning_rate": 6.324468083193631e-05, "loss": 0.0165, "step": 133180 }, { "grad_norm": 0.11830157786607742, "learning_rate": 6.323936607155348e-05, "loss": 0.0165, "step": 133190 }, { "grad_norm": 0.17280694842338562, "learning_rate": 6.323405115029985e-05, "loss": 0.0129, "step": 133200 }, { "grad_norm": 0.2128733992576599, "learning_rate": 6.322873606824001e-05, "loss": 0.0216, "step": 133210 }, { "grad_norm": 0.7031431794166565, "learning_rate": 6.322342082543856e-05, "loss": 0.0189, "step": 133220 }, { "grad_norm": 0.2374652624130249, "learning_rate": 6.321810542196002e-05, "loss": 0.0209, "step": 133230 }, { "grad_norm": 0.07527456432580948, "learning_rate": 6.321278985786904e-05, "loss": 0.0145, "step": 133240 }, { "grad_norm": 0.10255256295204163, "learning_rate": 6.320747413323019e-05, "loss": 0.0154, "step": 133250 }, { "grad_norm": 0.13806720077991486, "learning_rate": 6.320215824810807e-05, "loss": 0.018, "step": 133260 }, { "grad_norm": 0.1411897987127304, "learning_rate": 6.319684220256723e-05, "loss": 0.0171, "step": 133270 }, { "grad_norm": 0.5360451340675354, "learning_rate": 6.319152599667232e-05, "loss": 0.0173, "step": 133280 }, { "grad_norm": 0.11102136224508286, "learning_rate": 6.318620963048791e-05, "loss": 0.0218, "step": 133290 }, { "grad_norm": 0.15158212184906006, "learning_rate": 6.318089310407862e-05, "loss": 0.0138, "step": 133300 }, { "grad_norm": 0.1661640852689743, "learning_rate": 6.317557641750901e-05, "loss": 0.0145, "step": 133310 }, { "grad_norm": 0.0982559323310852, "learning_rate": 6.317025957084372e-05, "loss": 0.0162, "step": 133320 }, { "grad_norm": 0.1817276030778885, "learning_rate": 6.316494256414733e-05, "loss": 0.0141, "step": 133330 }, { "grad_norm": 0.2092607021331787, "learning_rate": 6.315962539748447e-05, "loss": 0.0144, "step": 133340 }, { "grad_norm": 0.12437564134597778, "learning_rate": 6.315430807091973e-05, "loss": 0.0204, "step": 133350 }, { "grad_norm": 0.1393527388572693, "learning_rate": 6.314899058451773e-05, "loss": 0.0192, "step": 133360 }, { "grad_norm": 0.12535014748573303, "learning_rate": 6.314367293834309e-05, "loss": 0.0137, "step": 133370 }, { "grad_norm": 0.09099167585372925, "learning_rate": 6.313835513246038e-05, "loss": 0.0198, "step": 133380 }, { "grad_norm": 0.11866318434476852, "learning_rate": 6.313303716693428e-05, "loss": 0.0175, "step": 133390 }, { "grad_norm": 0.12892434000968933, "learning_rate": 6.312771904182937e-05, "loss": 0.017, "step": 133400 }, { "grad_norm": 0.07160770893096924, "learning_rate": 6.312240075721029e-05, "loss": 0.0236, "step": 133410 }, { "grad_norm": 0.106781005859375, "learning_rate": 6.311708231314164e-05, "loss": 0.0142, "step": 133420 }, { "grad_norm": 0.1288125365972519, "learning_rate": 6.311176370968806e-05, "loss": 0.023, "step": 133430 }, { "grad_norm": 0.11736598610877991, "learning_rate": 6.310644494691417e-05, "loss": 0.014, "step": 133440 }, { "grad_norm": 0.08207438886165619, "learning_rate": 6.31011260248846e-05, "loss": 0.0144, "step": 133450 }, { "grad_norm": 0.6274781227111816, "learning_rate": 6.309580694366397e-05, "loss": 0.0162, "step": 133460 }, { "grad_norm": 0.11114944517612457, "learning_rate": 6.309048770331692e-05, "loss": 0.0159, "step": 133470 }, { "grad_norm": 0.1882956475019455, "learning_rate": 6.30851683039081e-05, "loss": 0.0263, "step": 133480 }, { "grad_norm": 0.1832815557718277, "learning_rate": 6.307984874550213e-05, "loss": 0.0171, "step": 133490 }, { "grad_norm": 0.15985246002674103, "learning_rate": 6.307452902816364e-05, "loss": 0.0194, "step": 133500 }, { "grad_norm": 0.0727640762925148, "learning_rate": 6.306920915195728e-05, "loss": 0.021, "step": 133510 }, { "grad_norm": 0.21340596675872803, "learning_rate": 6.30638891169477e-05, "loss": 0.021, "step": 133520 }, { "grad_norm": 0.13605447113513947, "learning_rate": 6.305856892319952e-05, "loss": 0.0227, "step": 133530 }, { "grad_norm": 0.08041022717952728, "learning_rate": 6.305324857077739e-05, "loss": 0.0219, "step": 133540 }, { "grad_norm": 0.24331405758857727, "learning_rate": 6.304792805974599e-05, "loss": 0.0155, "step": 133550 }, { "grad_norm": 0.10611667484045029, "learning_rate": 6.304260739016992e-05, "loss": 0.0214, "step": 133560 }, { "grad_norm": 0.10113555938005447, "learning_rate": 6.303728656211387e-05, "loss": 0.0159, "step": 133570 }, { "grad_norm": 0.11264112591743469, "learning_rate": 6.303196557564249e-05, "loss": 0.0147, "step": 133580 }, { "grad_norm": 0.7945647835731506, "learning_rate": 6.302664443082041e-05, "loss": 0.0124, "step": 133590 }, { "grad_norm": 0.09605918824672699, "learning_rate": 6.30213231277123e-05, "loss": 0.0194, "step": 133600 }, { "grad_norm": 0.09060247242450714, "learning_rate": 6.30160016663828e-05, "loss": 0.0194, "step": 133610 }, { "grad_norm": 0.10608108341693878, "learning_rate": 6.301068004689662e-05, "loss": 0.0177, "step": 133620 }, { "grad_norm": 0.11646654456853867, "learning_rate": 6.300535826931838e-05, "loss": 0.012, "step": 133630 }, { "grad_norm": 0.4772145748138428, "learning_rate": 6.300003633371277e-05, "loss": 0.0178, "step": 133640 }, { "grad_norm": 0.09627481549978256, "learning_rate": 6.299471424014443e-05, "loss": 0.0158, "step": 133650 }, { "grad_norm": 0.14992700517177582, "learning_rate": 6.298939198867806e-05, "loss": 0.0157, "step": 133660 }, { "grad_norm": 0.13912080228328705, "learning_rate": 6.298406957937832e-05, "loss": 0.0119, "step": 133670 }, { "grad_norm": 0.3205471932888031, "learning_rate": 6.297874701230984e-05, "loss": 0.0172, "step": 133680 }, { "grad_norm": 0.11650905013084412, "learning_rate": 6.297342428753736e-05, "loss": 0.0189, "step": 133690 }, { "grad_norm": 0.13337060809135437, "learning_rate": 6.29681014051255e-05, "loss": 0.0192, "step": 133700 }, { "grad_norm": 0.14035919308662415, "learning_rate": 6.2962778365139e-05, "loss": 0.0185, "step": 133710 }, { "grad_norm": 0.3294000029563904, "learning_rate": 6.295745516764247e-05, "loss": 0.0142, "step": 133720 }, { "grad_norm": 0.15223604440689087, "learning_rate": 6.295213181270064e-05, "loss": 0.0153, "step": 133730 }, { "grad_norm": 0.13443653285503387, "learning_rate": 6.294680830037819e-05, "loss": 0.0113, "step": 133740 }, { "grad_norm": 0.20252077281475067, "learning_rate": 6.294148463073976e-05, "loss": 0.0142, "step": 133750 }, { "grad_norm": 0.1188170313835144, "learning_rate": 6.293616080385012e-05, "loss": 0.02, "step": 133760 }, { "grad_norm": 0.14343468844890594, "learning_rate": 6.293083681977387e-05, "loss": 0.0121, "step": 133770 }, { "grad_norm": 0.1127660945057869, "learning_rate": 6.292551267857578e-05, "loss": 0.0152, "step": 133780 }, { "grad_norm": 0.23874783515930176, "learning_rate": 6.292018838032047e-05, "loss": 0.017, "step": 133790 }, { "grad_norm": 0.12965546548366547, "learning_rate": 6.29148639250727e-05, "loss": 0.0185, "step": 133800 }, { "grad_norm": 0.08951053023338318, "learning_rate": 6.290953931289713e-05, "loss": 0.0117, "step": 133810 }, { "grad_norm": 0.10616511106491089, "learning_rate": 6.290421454385849e-05, "loss": 0.0165, "step": 133820 }, { "grad_norm": 0.09232180565595627, "learning_rate": 6.289888961802143e-05, "loss": 0.012, "step": 133830 }, { "grad_norm": 0.16942071914672852, "learning_rate": 6.28935645354507e-05, "loss": 0.0169, "step": 133840 }, { "grad_norm": 0.07595942169427872, "learning_rate": 6.288823929621098e-05, "loss": 0.014, "step": 133850 }, { "grad_norm": 0.13214443624019623, "learning_rate": 6.288291390036698e-05, "loss": 0.0126, "step": 133860 }, { "grad_norm": 0.16726773977279663, "learning_rate": 6.287758834798341e-05, "loss": 0.0125, "step": 133870 }, { "grad_norm": 0.09774893522262573, "learning_rate": 6.2872262639125e-05, "loss": 0.0108, "step": 133880 }, { "grad_norm": 0.13551396131515503, "learning_rate": 6.286693677385643e-05, "loss": 0.0184, "step": 133890 }, { "grad_norm": 0.13921114802360535, "learning_rate": 6.286161075224243e-05, "loss": 0.0173, "step": 133900 }, { "grad_norm": 0.15100044012069702, "learning_rate": 6.285628457434772e-05, "loss": 0.0148, "step": 133910 }, { "grad_norm": 0.16582971811294556, "learning_rate": 6.285095824023703e-05, "loss": 0.0174, "step": 133920 }, { "grad_norm": 0.11319773644208908, "learning_rate": 6.284563174997505e-05, "loss": 0.0158, "step": 133930 }, { "grad_norm": 0.09012899547815323, "learning_rate": 6.28403051036265e-05, "loss": 0.0204, "step": 133940 }, { "grad_norm": 0.13213516771793365, "learning_rate": 6.283497830125614e-05, "loss": 0.0153, "step": 133950 }, { "grad_norm": 0.2689349055290222, "learning_rate": 6.282965134292868e-05, "loss": 0.0166, "step": 133960 }, { "grad_norm": 0.12033719569444656, "learning_rate": 6.282432422870883e-05, "loss": 0.0174, "step": 133970 }, { "grad_norm": 0.16051742434501648, "learning_rate": 6.281899695866133e-05, "loss": 0.0249, "step": 133980 }, { "grad_norm": 0.14877590537071228, "learning_rate": 6.281366953285094e-05, "loss": 0.0205, "step": 133990 }, { "grad_norm": 0.08268215507268906, "learning_rate": 6.280834195134234e-05, "loss": 0.0141, "step": 134000 }, { "grad_norm": 0.21101729571819305, "learning_rate": 6.28030142142003e-05, "loss": 0.0169, "step": 134010 }, { "grad_norm": 0.31476813554763794, "learning_rate": 6.279768632148955e-05, "loss": 0.0191, "step": 134020 }, { "grad_norm": 0.13964802026748657, "learning_rate": 6.279235827327484e-05, "loss": 0.0135, "step": 134030 }, { "grad_norm": 0.10633116215467453, "learning_rate": 6.278703006962087e-05, "loss": 0.0177, "step": 134040 }, { "grad_norm": 0.14469760656356812, "learning_rate": 6.278170171059243e-05, "loss": 0.0162, "step": 134050 }, { "grad_norm": 0.193703293800354, "learning_rate": 6.277637319625425e-05, "loss": 0.0187, "step": 134060 }, { "grad_norm": 0.10666440427303314, "learning_rate": 6.277104452667106e-05, "loss": 0.0171, "step": 134070 }, { "grad_norm": 0.2177913933992386, "learning_rate": 6.276571570190764e-05, "loss": 0.0134, "step": 134080 }, { "grad_norm": 0.1109199970960617, "learning_rate": 6.27603867220287e-05, "loss": 0.0164, "step": 134090 }, { "grad_norm": 0.07464450597763062, "learning_rate": 6.275505758709903e-05, "loss": 0.0117, "step": 134100 }, { "grad_norm": 0.09602196514606476, "learning_rate": 6.274972829718335e-05, "loss": 0.0111, "step": 134110 }, { "grad_norm": 0.18481294810771942, "learning_rate": 6.274439885234645e-05, "loss": 0.0137, "step": 134120 }, { "grad_norm": 0.1279582381248474, "learning_rate": 6.273906925265306e-05, "loss": 0.0188, "step": 134130 }, { "grad_norm": 0.084302619099617, "learning_rate": 6.273373949816797e-05, "loss": 0.0171, "step": 134140 }, { "grad_norm": 0.28127411007881165, "learning_rate": 6.27284095889559e-05, "loss": 0.0199, "step": 134150 }, { "grad_norm": 0.1403573453426361, "learning_rate": 6.272307952508164e-05, "loss": 0.0165, "step": 134160 }, { "grad_norm": 0.1101175844669342, "learning_rate": 6.271774930660997e-05, "loss": 0.0149, "step": 134170 }, { "grad_norm": 0.09542228281497955, "learning_rate": 6.271241893360561e-05, "loss": 0.0174, "step": 134180 }, { "grad_norm": 0.3686491847038269, "learning_rate": 6.270708840613337e-05, "loss": 0.0169, "step": 134190 }, { "grad_norm": 0.08000776916742325, "learning_rate": 6.270175772425802e-05, "loss": 0.0192, "step": 134200 }, { "grad_norm": 0.1316685974597931, "learning_rate": 6.26964268880443e-05, "loss": 0.0154, "step": 134210 }, { "grad_norm": 0.12840068340301514, "learning_rate": 6.269109589755701e-05, "loss": 0.0137, "step": 134220 }, { "grad_norm": 0.06990808248519897, "learning_rate": 6.268576475286095e-05, "loss": 0.014, "step": 134230 }, { "grad_norm": 0.11476914584636688, "learning_rate": 6.268043345402085e-05, "loss": 0.0181, "step": 134240 }, { "grad_norm": 0.12152645736932755, "learning_rate": 6.26751020011015e-05, "loss": 0.0127, "step": 134250 }, { "grad_norm": 0.16025863587856293, "learning_rate": 6.266977039416771e-05, "loss": 0.0145, "step": 134260 }, { "grad_norm": 0.07846800237894058, "learning_rate": 6.266443863328424e-05, "loss": 0.0178, "step": 134270 }, { "grad_norm": 0.066872738301754, "learning_rate": 6.26591067185159e-05, "loss": 0.0144, "step": 134280 }, { "grad_norm": 0.14355701208114624, "learning_rate": 6.265377464992743e-05, "loss": 0.0176, "step": 134290 }, { "grad_norm": 0.10603616386651993, "learning_rate": 6.264844242758368e-05, "loss": 0.0155, "step": 134300 }, { "grad_norm": 0.12761791050434113, "learning_rate": 6.26431100515494e-05, "loss": 0.0207, "step": 134310 }, { "grad_norm": 0.16623681783676147, "learning_rate": 6.263777752188939e-05, "loss": 0.018, "step": 134320 }, { "grad_norm": 0.1455216258764267, "learning_rate": 6.263244483866846e-05, "loss": 0.0132, "step": 134330 }, { "grad_norm": 0.29419437050819397, "learning_rate": 6.26271120019514e-05, "loss": 0.016, "step": 134340 }, { "grad_norm": 0.11282046139240265, "learning_rate": 6.2621779011803e-05, "loss": 0.0153, "step": 134350 }, { "grad_norm": 0.14074677228927612, "learning_rate": 6.261644586828806e-05, "loss": 0.0156, "step": 134360 }, { "grad_norm": 0.22099721431732178, "learning_rate": 6.26111125714714e-05, "loss": 0.0159, "step": 134370 }, { "grad_norm": 0.16951441764831543, "learning_rate": 6.260577912141783e-05, "loss": 0.0163, "step": 134380 }, { "grad_norm": 0.10361094027757645, "learning_rate": 6.260044551819213e-05, "loss": 0.012, "step": 134390 }, { "grad_norm": 0.09931062161922455, "learning_rate": 6.259511176185913e-05, "loss": 0.0157, "step": 134400 }, { "grad_norm": 0.09596168994903564, "learning_rate": 6.258977785248362e-05, "loss": 0.0166, "step": 134410 }, { "grad_norm": 0.1279994249343872, "learning_rate": 6.258444379013041e-05, "loss": 0.0155, "step": 134420 }, { "grad_norm": 0.1507859081029892, "learning_rate": 6.257910957486433e-05, "loss": 0.014, "step": 134430 }, { "grad_norm": 0.18645814061164856, "learning_rate": 6.25737752067502e-05, "loss": 0.009, "step": 134440 }, { "grad_norm": 0.1817886084318161, "learning_rate": 6.256844068585284e-05, "loss": 0.012, "step": 134450 }, { "grad_norm": 0.1442972868680954, "learning_rate": 6.256310601223705e-05, "loss": 0.0195, "step": 134460 }, { "grad_norm": 0.20095331966876984, "learning_rate": 6.255777118596766e-05, "loss": 0.0161, "step": 134470 }, { "grad_norm": 0.07558725029230118, "learning_rate": 6.255243620710949e-05, "loss": 0.0262, "step": 134480 }, { "grad_norm": 0.0687684491276741, "learning_rate": 6.254710107572737e-05, "loss": 0.0143, "step": 134490 }, { "grad_norm": 0.1156325489282608, "learning_rate": 6.254176579188612e-05, "loss": 0.0151, "step": 134500 }, { "grad_norm": 0.09420692175626755, "learning_rate": 6.253643035565058e-05, "loss": 0.0116, "step": 134510 }, { "grad_norm": 0.08276817947626114, "learning_rate": 6.253109476708558e-05, "loss": 0.0193, "step": 134520 }, { "grad_norm": 0.20597806572914124, "learning_rate": 6.252575902625595e-05, "loss": 0.0166, "step": 134530 }, { "grad_norm": 0.09658998996019363, "learning_rate": 6.25204231332265e-05, "loss": 0.0117, "step": 134540 }, { "grad_norm": 0.07251031696796417, "learning_rate": 6.251508708806211e-05, "loss": 0.0207, "step": 134550 }, { "grad_norm": 0.14959146082401276, "learning_rate": 6.250975089082759e-05, "loss": 0.0342, "step": 134560 }, { "grad_norm": 0.1461031138896942, "learning_rate": 6.250441454158778e-05, "loss": 0.0141, "step": 134570 }, { "grad_norm": 0.1316603273153305, "learning_rate": 6.249907804040754e-05, "loss": 0.0176, "step": 134580 }, { "grad_norm": 0.24100519716739655, "learning_rate": 6.249374138735168e-05, "loss": 0.0158, "step": 134590 }, { "grad_norm": 0.1901606172323227, "learning_rate": 6.248840458248508e-05, "loss": 0.0207, "step": 134600 }, { "grad_norm": 0.11112773418426514, "learning_rate": 6.248306762587257e-05, "loss": 0.0168, "step": 134610 }, { "grad_norm": 0.07286206632852554, "learning_rate": 6.2477730517579e-05, "loss": 0.0137, "step": 134620 }, { "grad_norm": 0.08411084860563278, "learning_rate": 6.247239325766922e-05, "loss": 0.0158, "step": 134630 }, { "grad_norm": 0.16751104593276978, "learning_rate": 6.24670558462081e-05, "loss": 0.0183, "step": 134640 }, { "grad_norm": 0.12377684563398361, "learning_rate": 6.246171828326048e-05, "loss": 0.0192, "step": 134650 }, { "grad_norm": 0.13393081724643707, "learning_rate": 6.24563805688912e-05, "loss": 0.0159, "step": 134660 }, { "grad_norm": 0.06585218012332916, "learning_rate": 6.245104270316516e-05, "loss": 0.0121, "step": 134670 }, { "grad_norm": 0.1887381225824356, "learning_rate": 6.244570468614717e-05, "loss": 0.0172, "step": 134680 }, { "grad_norm": 0.09082817286252975, "learning_rate": 6.244036651790212e-05, "loss": 0.0162, "step": 134690 }, { "grad_norm": 0.1860501766204834, "learning_rate": 6.243502819849488e-05, "loss": 0.0175, "step": 134700 }, { "grad_norm": 0.10197297483682632, "learning_rate": 6.242968972799032e-05, "loss": 0.0115, "step": 134710 }, { "grad_norm": 0.16903464496135712, "learning_rate": 6.242435110645329e-05, "loss": 0.0196, "step": 134720 }, { "grad_norm": 0.19715866446495056, "learning_rate": 6.241901233394864e-05, "loss": 0.0147, "step": 134730 }, { "grad_norm": 0.10444547235965729, "learning_rate": 6.241367341054128e-05, "loss": 0.014, "step": 134740 }, { "grad_norm": 0.07763014733791351, "learning_rate": 6.240833433629606e-05, "loss": 0.0154, "step": 134750 }, { "grad_norm": 0.09606292098760605, "learning_rate": 6.240299511127787e-05, "loss": 0.0165, "step": 134760 }, { "grad_norm": 0.17572945356369019, "learning_rate": 6.239765573555157e-05, "loss": 0.0181, "step": 134770 }, { "grad_norm": 0.3911466598510742, "learning_rate": 6.239231620918207e-05, "loss": 0.016, "step": 134780 }, { "grad_norm": 0.0768217146396637, "learning_rate": 6.238697653223422e-05, "loss": 0.0198, "step": 134790 }, { "grad_norm": 0.21558824181556702, "learning_rate": 6.23816367047729e-05, "loss": 0.0197, "step": 134800 }, { "grad_norm": 0.2397594451904297, "learning_rate": 6.2376296726863e-05, "loss": 0.0156, "step": 134810 }, { "grad_norm": 0.11050695925951004, "learning_rate": 6.237095659856941e-05, "loss": 0.0162, "step": 134820 }, { "grad_norm": 0.255289226770401, "learning_rate": 6.236561631995703e-05, "loss": 0.0239, "step": 134830 }, { "grad_norm": 0.19148662686347961, "learning_rate": 6.236027589109073e-05, "loss": 0.0203, "step": 134840 }, { "grad_norm": 0.17711786925792694, "learning_rate": 6.235493531203542e-05, "loss": 0.0169, "step": 134850 }, { "grad_norm": 0.10191774368286133, "learning_rate": 6.234959458285595e-05, "loss": 0.0148, "step": 134860 }, { "grad_norm": 0.15982148051261902, "learning_rate": 6.234425370361728e-05, "loss": 0.0186, "step": 134870 }, { "grad_norm": 0.19207192957401276, "learning_rate": 6.233891267438425e-05, "loss": 0.0131, "step": 134880 }, { "grad_norm": 0.1592753827571869, "learning_rate": 6.23335714952218e-05, "loss": 0.015, "step": 134890 }, { "grad_norm": 0.09563827514648438, "learning_rate": 6.23282301661948e-05, "loss": 0.0166, "step": 134900 }, { "grad_norm": 0.08920090645551682, "learning_rate": 6.232288868736815e-05, "loss": 0.0153, "step": 134910 }, { "grad_norm": 0.32651275396347046, "learning_rate": 6.231754705880679e-05, "loss": 0.0141, "step": 134920 }, { "grad_norm": 0.07160872220993042, "learning_rate": 6.231220528057559e-05, "loss": 0.0136, "step": 134930 }, { "grad_norm": 0.6889591217041016, "learning_rate": 6.230686335273947e-05, "loss": 0.0187, "step": 134940 }, { "grad_norm": 0.12418459355831146, "learning_rate": 6.230152127536337e-05, "loss": 0.0212, "step": 134950 }, { "grad_norm": 0.171513170003891, "learning_rate": 6.229617904851214e-05, "loss": 0.0306, "step": 134960 }, { "grad_norm": 0.10311539471149445, "learning_rate": 6.229083667225073e-05, "loss": 0.0186, "step": 134970 }, { "grad_norm": 0.10763105005025864, "learning_rate": 6.228549414664406e-05, "loss": 0.0128, "step": 134980 }, { "grad_norm": 0.06877878308296204, "learning_rate": 6.228015147175703e-05, "loss": 0.0148, "step": 134990 }, { "grad_norm": 0.13693954050540924, "learning_rate": 6.227480864765456e-05, "loss": 0.0146, "step": 135000 }, { "grad_norm": 0.2566302716732025, "learning_rate": 6.226946567440158e-05, "loss": 0.0182, "step": 135010 }, { "grad_norm": 0.10518886148929596, "learning_rate": 6.2264122552063e-05, "loss": 0.0121, "step": 135020 }, { "grad_norm": 0.06816591322422028, "learning_rate": 6.225877928070377e-05, "loss": 0.013, "step": 135030 }, { "grad_norm": 0.30368396639823914, "learning_rate": 6.22534358603888e-05, "loss": 0.0175, "step": 135040 }, { "grad_norm": 0.1005859524011612, "learning_rate": 6.224809229118299e-05, "loss": 0.0148, "step": 135050 }, { "grad_norm": 0.11450503021478653, "learning_rate": 6.22427485731513e-05, "loss": 0.0194, "step": 135060 }, { "grad_norm": 0.4464927017688751, "learning_rate": 6.223740470635867e-05, "loss": 0.0173, "step": 135070 }, { "grad_norm": 0.06810624152421951, "learning_rate": 6.223206069087002e-05, "loss": 0.0169, "step": 135080 }, { "grad_norm": 0.797420859336853, "learning_rate": 6.222671652675027e-05, "loss": 0.0178, "step": 135090 }, { "grad_norm": 0.1120043396949768, "learning_rate": 6.222137221406439e-05, "loss": 0.0168, "step": 135100 }, { "grad_norm": 0.11150870472192764, "learning_rate": 6.221602775287728e-05, "loss": 0.014, "step": 135110 }, { "grad_norm": 0.13233166933059692, "learning_rate": 6.22106831432539e-05, "loss": 0.0184, "step": 135120 }, { "grad_norm": 0.18561770021915436, "learning_rate": 6.22053383852592e-05, "loss": 0.0176, "step": 135130 }, { "grad_norm": 0.0920807495713234, "learning_rate": 6.219999347895811e-05, "loss": 0.0166, "step": 135140 }, { "grad_norm": 0.07664527744054794, "learning_rate": 6.219464842441558e-05, "loss": 0.0153, "step": 135150 }, { "grad_norm": 0.2088567167520523, "learning_rate": 6.218930322169655e-05, "loss": 0.0153, "step": 135160 }, { "grad_norm": 0.17587047815322876, "learning_rate": 6.2183957870866e-05, "loss": 0.017, "step": 135170 }, { "grad_norm": 0.11883872002363205, "learning_rate": 6.217861237198884e-05, "loss": 0.0224, "step": 135180 }, { "grad_norm": 0.1472950428724289, "learning_rate": 6.217326672513004e-05, "loss": 0.0163, "step": 135190 }, { "grad_norm": 0.15635767579078674, "learning_rate": 6.216792093035457e-05, "loss": 0.0151, "step": 135200 }, { "grad_norm": 0.10108786076307297, "learning_rate": 6.216257498772734e-05, "loss": 0.0155, "step": 135210 }, { "grad_norm": 0.06924504786729813, "learning_rate": 6.21572288973134e-05, "loss": 0.012, "step": 135220 }, { "grad_norm": 0.09950423985719681, "learning_rate": 6.21518826591776e-05, "loss": 0.0145, "step": 135230 }, { "grad_norm": 0.06512041389942169, "learning_rate": 6.214653627338498e-05, "loss": 0.0154, "step": 135240 }, { "grad_norm": 0.13530467450618744, "learning_rate": 6.214118974000044e-05, "loss": 0.0173, "step": 135250 }, { "grad_norm": 0.06283390522003174, "learning_rate": 6.213584305908901e-05, "loss": 0.0127, "step": 135260 }, { "grad_norm": 0.1770239770412445, "learning_rate": 6.213049623071562e-05, "loss": 0.0143, "step": 135270 }, { "grad_norm": 0.06337015330791473, "learning_rate": 6.212514925494526e-05, "loss": 0.0122, "step": 135280 }, { "grad_norm": 0.2745843827724457, "learning_rate": 6.211980213184287e-05, "loss": 0.016, "step": 135290 }, { "grad_norm": 0.18126635253429413, "learning_rate": 6.211445486147346e-05, "loss": 0.0157, "step": 135300 }, { "grad_norm": 0.0889487937092781, "learning_rate": 6.210910744390197e-05, "loss": 0.0128, "step": 135310 }, { "grad_norm": 0.08733335137367249, "learning_rate": 6.210375987919341e-05, "loss": 0.0152, "step": 135320 }, { "grad_norm": 0.19346198439598083, "learning_rate": 6.209841216741271e-05, "loss": 0.0204, "step": 135330 }, { "grad_norm": 0.18029195070266724, "learning_rate": 6.20930643086249e-05, "loss": 0.0198, "step": 135340 }, { "grad_norm": 0.15252968668937683, "learning_rate": 6.208771630289493e-05, "loss": 0.0142, "step": 135350 }, { "grad_norm": 0.5861090421676636, "learning_rate": 6.208236815028781e-05, "loss": 0.0192, "step": 135360 }, { "grad_norm": 0.1299472600221634, "learning_rate": 6.207701985086849e-05, "loss": 0.0146, "step": 135370 }, { "grad_norm": 0.11512463539838791, "learning_rate": 6.2071671404702e-05, "loss": 0.0177, "step": 135380 }, { "grad_norm": 0.14877094328403473, "learning_rate": 6.206632281185328e-05, "loss": 0.021, "step": 135390 }, { "grad_norm": 0.10911516100168228, "learning_rate": 6.206097407238736e-05, "loss": 0.0108, "step": 135400 }, { "grad_norm": 0.19078165292739868, "learning_rate": 6.205562518636921e-05, "loss": 0.0188, "step": 135410 }, { "grad_norm": 0.17991256713867188, "learning_rate": 6.205027615386385e-05, "loss": 0.0163, "step": 135420 }, { "grad_norm": 0.08704536408185959, "learning_rate": 6.204492697493624e-05, "loss": 0.0213, "step": 135430 }, { "grad_norm": 0.2459384799003601, "learning_rate": 6.20395776496514e-05, "loss": 0.0231, "step": 135440 }, { "grad_norm": 0.15715640783309937, "learning_rate": 6.203422817807433e-05, "loss": 0.0153, "step": 135450 }, { "grad_norm": 0.11655066162347794, "learning_rate": 6.202887856027001e-05, "loss": 0.0167, "step": 135460 }, { "grad_norm": 0.15014885365962982, "learning_rate": 6.202352879630347e-05, "loss": 0.0133, "step": 135470 }, { "grad_norm": 0.08720294386148453, "learning_rate": 6.20181788862397e-05, "loss": 0.017, "step": 135480 }, { "grad_norm": 0.08066781610250473, "learning_rate": 6.201282883014371e-05, "loss": 0.0112, "step": 135490 }, { "grad_norm": 0.09893647581338882, "learning_rate": 6.20074786280805e-05, "loss": 0.0151, "step": 135500 }, { "grad_norm": 0.15172827243804932, "learning_rate": 6.200212828011511e-05, "loss": 0.0222, "step": 135510 }, { "grad_norm": 0.2651935815811157, "learning_rate": 6.199677778631251e-05, "loss": 0.0132, "step": 135520 }, { "grad_norm": 0.16905494034290314, "learning_rate": 6.199142714673774e-05, "loss": 0.0143, "step": 135530 }, { "grad_norm": 0.10615009814500809, "learning_rate": 6.19860763614558e-05, "loss": 0.0173, "step": 135540 }, { "grad_norm": 0.10320562869310379, "learning_rate": 6.198072543053173e-05, "loss": 0.0108, "step": 135550 }, { "grad_norm": 0.0888521820306778, "learning_rate": 6.197537435403052e-05, "loss": 0.0151, "step": 135560 }, { "grad_norm": 0.12330430746078491, "learning_rate": 6.197002313201721e-05, "loss": 0.0133, "step": 135570 }, { "grad_norm": 0.1889582723379135, "learning_rate": 6.196467176455683e-05, "loss": 0.0188, "step": 135580 }, { "grad_norm": 0.14333517849445343, "learning_rate": 6.195932025171436e-05, "loss": 0.0146, "step": 135590 }, { "grad_norm": 0.1064983606338501, "learning_rate": 6.195396859355488e-05, "loss": 0.0161, "step": 135600 }, { "grad_norm": 0.11222849786281586, "learning_rate": 6.194861679014341e-05, "loss": 0.0173, "step": 135610 }, { "grad_norm": 1.1623624563217163, "learning_rate": 6.194326484154493e-05, "loss": 0.0188, "step": 135620 }, { "grad_norm": 0.1082172691822052, "learning_rate": 6.193791274782453e-05, "loss": 0.0162, "step": 135630 }, { "grad_norm": 0.11888151615858078, "learning_rate": 6.19325605090472e-05, "loss": 0.02, "step": 135640 }, { "grad_norm": 0.7728049755096436, "learning_rate": 6.192720812527802e-05, "loss": 0.0143, "step": 135650 }, { "grad_norm": 0.14653539657592773, "learning_rate": 6.192185559658198e-05, "loss": 0.0145, "step": 135660 }, { "grad_norm": 0.1923896223306656, "learning_rate": 6.191650292302414e-05, "loss": 0.0132, "step": 135670 }, { "grad_norm": 0.12993238866329193, "learning_rate": 6.191115010466956e-05, "loss": 0.0122, "step": 135680 }, { "grad_norm": 0.11638893932104111, "learning_rate": 6.190579714158324e-05, "loss": 0.0168, "step": 135690 }, { "grad_norm": 0.10754574835300446, "learning_rate": 6.190044403383024e-05, "loss": 0.0157, "step": 135700 }, { "grad_norm": 0.17433591187000275, "learning_rate": 6.189509078147561e-05, "loss": 0.0188, "step": 135710 }, { "grad_norm": 0.11078416556119919, "learning_rate": 6.188973738458441e-05, "loss": 0.0121, "step": 135720 }, { "grad_norm": 0.04330397769808769, "learning_rate": 6.188438384322166e-05, "loss": 0.0106, "step": 135730 }, { "grad_norm": 0.11233196407556534, "learning_rate": 6.187903015745243e-05, "loss": 0.0132, "step": 135740 }, { "grad_norm": 0.15916982293128967, "learning_rate": 6.187367632734177e-05, "loss": 0.016, "step": 135750 }, { "grad_norm": 0.15209133923053741, "learning_rate": 6.186832235295474e-05, "loss": 0.015, "step": 135760 }, { "grad_norm": 0.4135335683822632, "learning_rate": 6.186296823435639e-05, "loss": 0.0194, "step": 135770 }, { "grad_norm": 0.10390457510948181, "learning_rate": 6.185761397161175e-05, "loss": 0.0169, "step": 135780 }, { "grad_norm": 0.11143146455287933, "learning_rate": 6.185225956478593e-05, "loss": 0.0184, "step": 135790 }, { "grad_norm": 0.09140907973051071, "learning_rate": 6.184690501394394e-05, "loss": 0.0148, "step": 135800 }, { "grad_norm": 0.14880235493183136, "learning_rate": 6.184155031915089e-05, "loss": 0.0181, "step": 135810 }, { "grad_norm": 0.18698784708976746, "learning_rate": 6.183619548047183e-05, "loss": 0.0137, "step": 135820 }, { "grad_norm": 0.08258425444364548, "learning_rate": 6.183084049797181e-05, "loss": 0.0223, "step": 135830 }, { "grad_norm": 0.09920784831047058, "learning_rate": 6.182548537171591e-05, "loss": 0.0166, "step": 135840 }, { "grad_norm": 0.23992004990577698, "learning_rate": 6.182013010176918e-05, "loss": 0.0158, "step": 135850 }, { "grad_norm": 0.1302836388349533, "learning_rate": 6.181477468819673e-05, "loss": 0.0297, "step": 135860 }, { "grad_norm": 0.16155889630317688, "learning_rate": 6.18094191310636e-05, "loss": 0.0133, "step": 135870 }, { "grad_norm": 0.15534985065460205, "learning_rate": 6.180406343043489e-05, "loss": 0.0138, "step": 135880 }, { "grad_norm": 0.08666267991065979, "learning_rate": 6.179870758637565e-05, "loss": 0.0177, "step": 135890 }, { "grad_norm": 0.27564355731010437, "learning_rate": 6.179335159895099e-05, "loss": 0.0138, "step": 135900 }, { "grad_norm": 0.24224239587783813, "learning_rate": 6.178799546822595e-05, "loss": 0.0166, "step": 135910 }, { "grad_norm": 0.15313969552516937, "learning_rate": 6.178263919426566e-05, "loss": 0.0179, "step": 135920 }, { "grad_norm": 0.1847677379846573, "learning_rate": 6.177728277713516e-05, "loss": 0.0191, "step": 135930 }, { "grad_norm": 0.10006438195705414, "learning_rate": 6.177192621689957e-05, "loss": 0.0151, "step": 135940 }, { "grad_norm": 0.22951513528823853, "learning_rate": 6.176656951362395e-05, "loss": 0.0152, "step": 135950 }, { "grad_norm": 0.07534025609493256, "learning_rate": 6.17612126673734e-05, "loss": 0.0201, "step": 135960 }, { "grad_norm": 0.08459293842315674, "learning_rate": 6.175585567821303e-05, "loss": 0.0218, "step": 135970 }, { "grad_norm": 0.2322390079498291, "learning_rate": 6.17504985462079e-05, "loss": 0.019, "step": 135980 }, { "grad_norm": 0.23127038776874542, "learning_rate": 6.174514127142312e-05, "loss": 0.0168, "step": 135990 }, { "grad_norm": 0.2035655677318573, "learning_rate": 6.173978385392377e-05, "loss": 0.0166, "step": 136000 }, { "grad_norm": 0.1354781687259674, "learning_rate": 6.173442629377498e-05, "loss": 0.0169, "step": 136010 }, { "grad_norm": 0.1418922394514084, "learning_rate": 6.172906859104183e-05, "loss": 0.0161, "step": 136020 }, { "grad_norm": 0.20665796101093292, "learning_rate": 6.172371074578942e-05, "loss": 0.019, "step": 136030 }, { "grad_norm": 0.1258135437965393, "learning_rate": 6.171835275808286e-05, "loss": 0.0166, "step": 136040 }, { "grad_norm": 0.09594345837831497, "learning_rate": 6.171299462798724e-05, "loss": 0.0165, "step": 136050 }, { "grad_norm": 0.09831715375185013, "learning_rate": 6.170763635556767e-05, "loss": 0.0123, "step": 136060 }, { "grad_norm": 0.8623060584068298, "learning_rate": 6.170227794088928e-05, "loss": 0.017, "step": 136070 }, { "grad_norm": 0.09866157174110413, "learning_rate": 6.169691938401716e-05, "loss": 0.0163, "step": 136080 }, { "grad_norm": 0.2825264632701874, "learning_rate": 6.169156068501643e-05, "loss": 0.0138, "step": 136090 }, { "grad_norm": 0.20137755572795868, "learning_rate": 6.16862018439522e-05, "loss": 0.0218, "step": 136100 }, { "grad_norm": 0.09514795243740082, "learning_rate": 6.168084286088958e-05, "loss": 0.0124, "step": 136110 }, { "grad_norm": 0.11184285581111908, "learning_rate": 6.167548373589368e-05, "loss": 0.0177, "step": 136120 }, { "grad_norm": 0.26376572251319885, "learning_rate": 6.167012446902964e-05, "loss": 0.0118, "step": 136130 }, { "grad_norm": 0.16042467951774597, "learning_rate": 6.166476506036257e-05, "loss": 0.024, "step": 136140 }, { "grad_norm": 0.10486756265163422, "learning_rate": 6.165940550995759e-05, "loss": 0.0162, "step": 136150 }, { "grad_norm": 0.10504622757434845, "learning_rate": 6.165404581787983e-05, "loss": 0.0232, "step": 136160 }, { "grad_norm": 0.10350406914949417, "learning_rate": 6.164868598419439e-05, "loss": 0.0134, "step": 136170 }, { "grad_norm": 0.1344565749168396, "learning_rate": 6.164332600896645e-05, "loss": 0.0143, "step": 136180 }, { "grad_norm": 0.1158519834280014, "learning_rate": 6.163796589226108e-05, "loss": 0.0163, "step": 136190 }, { "grad_norm": 0.1079728826880455, "learning_rate": 6.163260563414344e-05, "loss": 0.0139, "step": 136200 }, { "grad_norm": 0.09564144164323807, "learning_rate": 6.162724523467867e-05, "loss": 0.019, "step": 136210 }, { "grad_norm": 0.0970509797334671, "learning_rate": 6.162188469393188e-05, "loss": 0.0131, "step": 136220 }, { "grad_norm": 0.23364371061325073, "learning_rate": 6.161652401196822e-05, "loss": 0.0186, "step": 136230 }, { "grad_norm": 0.13483065366744995, "learning_rate": 6.161116318885283e-05, "loss": 0.0113, "step": 136240 }, { "grad_norm": 0.21290908753871918, "learning_rate": 6.160580222465085e-05, "loss": 0.0254, "step": 136250 }, { "grad_norm": 0.13461074233055115, "learning_rate": 6.16004411194274e-05, "loss": 0.0167, "step": 136260 }, { "grad_norm": 0.23323793709278107, "learning_rate": 6.159507987324765e-05, "loss": 0.0108, "step": 136270 }, { "grad_norm": 0.25707244873046875, "learning_rate": 6.158971848617673e-05, "loss": 0.0234, "step": 136280 }, { "grad_norm": 0.15711188316345215, "learning_rate": 6.158435695827979e-05, "loss": 0.0155, "step": 136290 }, { "grad_norm": 0.12456017732620239, "learning_rate": 6.157899528962197e-05, "loss": 0.0159, "step": 136300 }, { "grad_norm": 0.3117096722126007, "learning_rate": 6.157363348026842e-05, "loss": 0.0193, "step": 136310 }, { "grad_norm": 0.06624425947666168, "learning_rate": 6.15682715302843e-05, "loss": 0.0142, "step": 136320 }, { "grad_norm": 0.09189773350954056, "learning_rate": 6.156290943973476e-05, "loss": 0.0158, "step": 136330 }, { "grad_norm": 0.15536785125732422, "learning_rate": 6.155754720868496e-05, "loss": 0.0176, "step": 136340 }, { "grad_norm": 0.08590318262577057, "learning_rate": 6.155218483720004e-05, "loss": 0.013, "step": 136350 }, { "grad_norm": 0.1755184531211853, "learning_rate": 6.154682232534517e-05, "loss": 0.0136, "step": 136360 }, { "grad_norm": 0.07198463380336761, "learning_rate": 6.154145967318551e-05, "loss": 0.0176, "step": 136370 }, { "grad_norm": 0.07928343862295151, "learning_rate": 6.15360968807862e-05, "loss": 0.0097, "step": 136380 }, { "grad_norm": 0.12692220509052277, "learning_rate": 6.153073394821244e-05, "loss": 0.0125, "step": 136390 }, { "grad_norm": 0.11958735436201096, "learning_rate": 6.152537087552938e-05, "loss": 0.0149, "step": 136400 }, { "grad_norm": 0.11994636058807373, "learning_rate": 6.152000766280218e-05, "loss": 0.0154, "step": 136410 }, { "grad_norm": 0.10685025155544281, "learning_rate": 6.151464431009599e-05, "loss": 0.012, "step": 136420 }, { "grad_norm": 0.14344125986099243, "learning_rate": 6.150928081747603e-05, "loss": 0.0196, "step": 136430 }, { "grad_norm": 0.3190793991088867, "learning_rate": 6.150391718500741e-05, "loss": 0.0185, "step": 136440 }, { "grad_norm": 0.1496075540781021, "learning_rate": 6.149855341275536e-05, "loss": 0.0222, "step": 136450 }, { "grad_norm": 0.131259486079216, "learning_rate": 6.149318950078501e-05, "loss": 0.0216, "step": 136460 }, { "grad_norm": 0.172604501247406, "learning_rate": 6.148782544916157e-05, "loss": 0.0194, "step": 136470 }, { "grad_norm": 0.08576307445764542, "learning_rate": 6.14824612579502e-05, "loss": 0.0157, "step": 136480 }, { "grad_norm": 0.09788291156291962, "learning_rate": 6.147709692721608e-05, "loss": 0.019, "step": 136490 }, { "grad_norm": 0.11308565735816956, "learning_rate": 6.147173245702442e-05, "loss": 0.012, "step": 136500 }, { "grad_norm": 0.08152402937412262, "learning_rate": 6.146636784744036e-05, "loss": 0.0147, "step": 136510 }, { "grad_norm": 0.13828057050704956, "learning_rate": 6.14610030985291e-05, "loss": 0.0142, "step": 136520 }, { "grad_norm": 0.23999065160751343, "learning_rate": 6.145563821035582e-05, "loss": 0.0177, "step": 136530 }, { "grad_norm": 0.10897248238325119, "learning_rate": 6.145027318298576e-05, "loss": 0.0158, "step": 136540 }, { "grad_norm": 0.11127787083387375, "learning_rate": 6.144490801648404e-05, "loss": 0.0168, "step": 136550 }, { "grad_norm": 0.1576581597328186, "learning_rate": 6.143954271091589e-05, "loss": 0.0162, "step": 136560 }, { "grad_norm": 0.1452309489250183, "learning_rate": 6.14341772663465e-05, "loss": 0.0163, "step": 136570 }, { "grad_norm": 0.13052797317504883, "learning_rate": 6.142881168284104e-05, "loss": 0.0121, "step": 136580 }, { "grad_norm": 0.13326972723007202, "learning_rate": 6.142344596046475e-05, "loss": 0.0178, "step": 136590 }, { "grad_norm": 0.08802858740091324, "learning_rate": 6.14180800992828e-05, "loss": 0.0145, "step": 136600 }, { "grad_norm": 0.14554378390312195, "learning_rate": 6.141271409936039e-05, "loss": 0.0131, "step": 136610 }, { "grad_norm": 0.11928460001945496, "learning_rate": 6.140734796076273e-05, "loss": 0.0146, "step": 136620 }, { "grad_norm": 0.14267168939113617, "learning_rate": 6.140198168355502e-05, "loss": 0.0155, "step": 136630 }, { "grad_norm": 0.3661668002605438, "learning_rate": 6.139661526780247e-05, "loss": 0.0135, "step": 136640 }, { "grad_norm": 0.14020508527755737, "learning_rate": 6.13912487135703e-05, "loss": 0.016, "step": 136650 }, { "grad_norm": 0.13381154835224152, "learning_rate": 6.138588202092367e-05, "loss": 0.0161, "step": 136660 }, { "grad_norm": 0.09741595387458801, "learning_rate": 6.138051518992784e-05, "loss": 0.017, "step": 136670 }, { "grad_norm": 0.068262979388237, "learning_rate": 6.137514822064802e-05, "loss": 0.0144, "step": 136680 }, { "grad_norm": 0.2431197315454483, "learning_rate": 6.13697811131494e-05, "loss": 0.0195, "step": 136690 }, { "grad_norm": 0.11709296703338623, "learning_rate": 6.136441386749719e-05, "loss": 0.0154, "step": 136700 }, { "grad_norm": 0.29743072390556335, "learning_rate": 6.135904648375663e-05, "loss": 0.0125, "step": 136710 }, { "grad_norm": 0.12046976387500763, "learning_rate": 6.135367896199294e-05, "loss": 0.0145, "step": 136720 }, { "grad_norm": 0.11700005829334259, "learning_rate": 6.134831130227132e-05, "loss": 0.0191, "step": 136730 }, { "grad_norm": 0.16607077419757843, "learning_rate": 6.134294350465701e-05, "loss": 0.0184, "step": 136740 }, { "grad_norm": 0.11314594000577927, "learning_rate": 6.133757556921523e-05, "loss": 0.0141, "step": 136750 }, { "grad_norm": 0.10123779624700546, "learning_rate": 6.133220749601119e-05, "loss": 0.0189, "step": 136760 }, { "grad_norm": 0.09740787744522095, "learning_rate": 6.132683928511014e-05, "loss": 0.0159, "step": 136770 }, { "grad_norm": 0.09427497535943985, "learning_rate": 6.13214709365773e-05, "loss": 0.0132, "step": 136780 }, { "grad_norm": 0.23707681894302368, "learning_rate": 6.131610245047789e-05, "loss": 0.0186, "step": 136790 }, { "grad_norm": 0.10258286446332932, "learning_rate": 6.131073382687715e-05, "loss": 0.0173, "step": 136800 }, { "grad_norm": 0.15857383608818054, "learning_rate": 6.130536506584032e-05, "loss": 0.0145, "step": 136810 }, { "grad_norm": 0.1158180981874466, "learning_rate": 6.129999616743264e-05, "loss": 0.0122, "step": 136820 }, { "grad_norm": 0.13348537683486938, "learning_rate": 6.129462713171932e-05, "loss": 0.015, "step": 136830 }, { "grad_norm": 0.08514329791069031, "learning_rate": 6.128925795876562e-05, "loss": 0.0149, "step": 136840 }, { "grad_norm": 0.23104755580425262, "learning_rate": 6.128388864863678e-05, "loss": 0.0202, "step": 136850 }, { "grad_norm": 0.13599535822868347, "learning_rate": 6.127851920139804e-05, "loss": 0.019, "step": 136860 }, { "grad_norm": 0.09151464700698853, "learning_rate": 6.127314961711464e-05, "loss": 0.0115, "step": 136870 }, { "grad_norm": 0.09331901371479034, "learning_rate": 6.126777989585185e-05, "loss": 0.0195, "step": 136880 }, { "grad_norm": 0.1751076877117157, "learning_rate": 6.126241003767487e-05, "loss": 0.0184, "step": 136890 }, { "grad_norm": 0.07521181553602219, "learning_rate": 6.125704004264898e-05, "loss": 0.011, "step": 136900 }, { "grad_norm": 0.23723115026950836, "learning_rate": 6.125166991083942e-05, "loss": 0.0136, "step": 136910 }, { "grad_norm": 0.1065826341509819, "learning_rate": 6.124629964231145e-05, "loss": 0.0142, "step": 136920 }, { "grad_norm": 0.1382725089788437, "learning_rate": 6.124092923713033e-05, "loss": 0.0117, "step": 136930 }, { "grad_norm": 0.1353778839111328, "learning_rate": 6.12355586953613e-05, "loss": 0.0164, "step": 136940 }, { "grad_norm": 0.08550485223531723, "learning_rate": 6.123018801706963e-05, "loss": 0.0153, "step": 136950 }, { "grad_norm": 0.1742282658815384, "learning_rate": 6.122481720232058e-05, "loss": 0.0191, "step": 136960 }, { "grad_norm": 0.21124856173992157, "learning_rate": 6.121944625117939e-05, "loss": 0.0171, "step": 136970 }, { "grad_norm": 0.07647731900215149, "learning_rate": 6.121407516371134e-05, "loss": 0.0116, "step": 136980 }, { "grad_norm": 0.09597501903772354, "learning_rate": 6.120870393998168e-05, "loss": 0.0159, "step": 136990 }, { "grad_norm": 0.1135556548833847, "learning_rate": 6.120333258005569e-05, "loss": 0.0181, "step": 137000 }, { "grad_norm": 0.07201593369245529, "learning_rate": 6.119796108399862e-05, "loss": 0.0108, "step": 137010 }, { "grad_norm": 0.13312312960624695, "learning_rate": 6.119258945187577e-05, "loss": 0.0128, "step": 137020 }, { "grad_norm": 0.10832861065864563, "learning_rate": 6.118721768375238e-05, "loss": 0.0147, "step": 137030 }, { "grad_norm": 0.08256258815526962, "learning_rate": 6.118184577969374e-05, "loss": 0.0131, "step": 137040 }, { "grad_norm": 0.08042576164007187, "learning_rate": 6.117647373976512e-05, "loss": 0.0184, "step": 137050 }, { "grad_norm": 0.2745808959007263, "learning_rate": 6.117110156403178e-05, "loss": 0.0134, "step": 137060 }, { "grad_norm": 0.4340333938598633, "learning_rate": 6.116572925255902e-05, "loss": 0.0232, "step": 137070 }, { "grad_norm": 0.1653594672679901, "learning_rate": 6.116035680541208e-05, "loss": 0.0195, "step": 137080 }, { "grad_norm": 0.13291312754154205, "learning_rate": 6.11549842226563e-05, "loss": 0.0168, "step": 137090 }, { "grad_norm": 0.13168588280677795, "learning_rate": 6.114961150435691e-05, "loss": 0.0145, "step": 137100 }, { "grad_norm": 0.09184081852436066, "learning_rate": 6.114423865057922e-05, "loss": 0.0104, "step": 137110 }, { "grad_norm": 0.11188916116952896, "learning_rate": 6.11388656613885e-05, "loss": 0.0199, "step": 137120 }, { "grad_norm": 0.09079097211360931, "learning_rate": 6.113349253685006e-05, "loss": 0.0121, "step": 137130 }, { "grad_norm": 0.19698627293109894, "learning_rate": 6.112811927702917e-05, "loss": 0.0223, "step": 137140 }, { "grad_norm": 0.12527407705783844, "learning_rate": 6.112274588199111e-05, "loss": 0.0147, "step": 137150 }, { "grad_norm": 0.10776857286691666, "learning_rate": 6.111737235180121e-05, "loss": 0.0196, "step": 137160 }, { "grad_norm": 0.09718552976846695, "learning_rate": 6.111199868652472e-05, "loss": 0.017, "step": 137170 }, { "grad_norm": 0.1472916305065155, "learning_rate": 6.110662488622696e-05, "loss": 0.0216, "step": 137180 }, { "grad_norm": 0.12184568494558334, "learning_rate": 6.110125095097321e-05, "loss": 0.0136, "step": 137190 }, { "grad_norm": 0.31321659684181213, "learning_rate": 6.109587688082881e-05, "loss": 0.0199, "step": 137200 }, { "grad_norm": 0.12494364380836487, "learning_rate": 6.109050267585901e-05, "loss": 0.013, "step": 137210 }, { "grad_norm": 0.3458711802959442, "learning_rate": 6.108512833612913e-05, "loss": 0.0157, "step": 137220 }, { "grad_norm": 0.10664978623390198, "learning_rate": 6.107975386170448e-05, "loss": 0.0204, "step": 137230 }, { "grad_norm": 0.1984396129846573, "learning_rate": 6.107437925265035e-05, "loss": 0.0212, "step": 137240 }, { "grad_norm": 0.07980802655220032, "learning_rate": 6.106900450903209e-05, "loss": 0.0188, "step": 137250 }, { "grad_norm": 0.20651067793369293, "learning_rate": 6.106362963091493e-05, "loss": 0.0192, "step": 137260 }, { "grad_norm": 0.11881149560213089, "learning_rate": 6.105825461836425e-05, "loss": 0.0185, "step": 137270 }, { "grad_norm": 0.10034576803445816, "learning_rate": 6.10528794714453e-05, "loss": 0.0149, "step": 137280 }, { "grad_norm": 0.1737033873796463, "learning_rate": 6.104750419022348e-05, "loss": 0.0149, "step": 137290 }, { "grad_norm": 0.35883602499961853, "learning_rate": 6.104212877476403e-05, "loss": 0.0173, "step": 137300 }, { "grad_norm": 0.08003001660108566, "learning_rate": 6.103675322513228e-05, "loss": 0.0134, "step": 137310 }, { "grad_norm": 0.10879816114902496, "learning_rate": 6.1031377541393573e-05, "loss": 0.0199, "step": 137320 }, { "grad_norm": 0.09921803325414658, "learning_rate": 6.102600172361321e-05, "loss": 0.0178, "step": 137330 }, { "grad_norm": 0.15966498851776123, "learning_rate": 6.102062577185652e-05, "loss": 0.0132, "step": 137340 }, { "grad_norm": 0.10416419059038162, "learning_rate": 6.10152496861888e-05, "loss": 0.0095, "step": 137350 }, { "grad_norm": 0.27878516912460327, "learning_rate": 6.1009873466675424e-05, "loss": 0.0149, "step": 137360 }, { "grad_norm": 0.11998027563095093, "learning_rate": 6.100449711338169e-05, "loss": 0.0141, "step": 137370 }, { "grad_norm": 0.13469655811786652, "learning_rate": 6.0999120626372894e-05, "loss": 0.0156, "step": 137380 }, { "grad_norm": 0.0869053304195404, "learning_rate": 6.099374400571443e-05, "loss": 0.0097, "step": 137390 }, { "grad_norm": 0.22859425842761993, "learning_rate": 6.0988367251471577e-05, "loss": 0.0172, "step": 137400 }, { "grad_norm": 0.0960213914513588, "learning_rate": 6.09829903637097e-05, "loss": 0.0156, "step": 137410 }, { "grad_norm": 0.1207754909992218, "learning_rate": 6.097761334249411e-05, "loss": 0.015, "step": 137420 }, { "grad_norm": 0.13037732243537903, "learning_rate": 6.097223618789017e-05, "loss": 0.0143, "step": 137430 }, { "grad_norm": 0.18626533448696136, "learning_rate": 6.0966858899963175e-05, "loss": 0.0154, "step": 137440 }, { "grad_norm": 0.06410547345876694, "learning_rate": 6.096148147877851e-05, "loss": 0.0121, "step": 137450 }, { "grad_norm": 0.1265980303287506, "learning_rate": 6.0956103924401484e-05, "loss": 0.0155, "step": 137460 }, { "grad_norm": 0.10005945712327957, "learning_rate": 6.095072623689746e-05, "loss": 0.0129, "step": 137470 }, { "grad_norm": 0.11443312466144562, "learning_rate": 6.094534841633176e-05, "loss": 0.0148, "step": 137480 }, { "grad_norm": 0.34458884596824646, "learning_rate": 6.0939970462769756e-05, "loss": 0.0168, "step": 137490 }, { "grad_norm": 0.10278574377298355, "learning_rate": 6.093459237627678e-05, "loss": 0.0115, "step": 137500 }, { "grad_norm": 0.16803519427776337, "learning_rate": 6.092921415691818e-05, "loss": 0.0179, "step": 137510 }, { "grad_norm": 0.1890985369682312, "learning_rate": 6.092383580475931e-05, "loss": 0.0146, "step": 137520 }, { "grad_norm": 0.07386443763971329, "learning_rate": 6.091845731986552e-05, "loss": 0.012, "step": 137530 }, { "grad_norm": 0.07952989637851715, "learning_rate": 6.091307870230216e-05, "loss": 0.0138, "step": 137540 }, { "grad_norm": 0.08687824755907059, "learning_rate": 6.09076999521346e-05, "loss": 0.0156, "step": 137550 }, { "grad_norm": 0.11155518144369125, "learning_rate": 6.090232106942818e-05, "loss": 0.0136, "step": 137560 }, { "grad_norm": 0.3027096688747406, "learning_rate": 6.089694205424828e-05, "loss": 0.0114, "step": 137570 }, { "grad_norm": 0.16137030720710754, "learning_rate": 6.0891562906660226e-05, "loss": 0.016, "step": 137580 }, { "grad_norm": 0.15527528524398804, "learning_rate": 6.088618362672941e-05, "loss": 0.0159, "step": 137590 }, { "grad_norm": 0.2756875157356262, "learning_rate": 6.088080421452118e-05, "loss": 0.0192, "step": 137600 }, { "grad_norm": 0.6426250338554382, "learning_rate": 6.087542467010091e-05, "loss": 0.0153, "step": 137610 }, { "grad_norm": 0.10451429337263107, "learning_rate": 6.087004499353396e-05, "loss": 0.0196, "step": 137620 }, { "grad_norm": 0.4646431803703308, "learning_rate": 6.0864665184885694e-05, "loss": 0.0148, "step": 137630 }, { "grad_norm": 0.1364416480064392, "learning_rate": 6.085928524422151e-05, "loss": 0.0152, "step": 137640 }, { "grad_norm": 0.260180801153183, "learning_rate": 6.085390517160673e-05, "loss": 0.0162, "step": 137650 }, { "grad_norm": 0.10942576080560684, "learning_rate": 6.084852496710677e-05, "loss": 0.0224, "step": 137660 }, { "grad_norm": 0.08834236115217209, "learning_rate": 6.0843144630786975e-05, "loss": 0.0174, "step": 137670 }, { "grad_norm": 0.1041204184293747, "learning_rate": 6.083776416271275e-05, "loss": 0.0175, "step": 137680 }, { "grad_norm": 0.14608006179332733, "learning_rate": 6.0832383562949444e-05, "loss": 0.0192, "step": 137690 }, { "grad_norm": 0.11984565109014511, "learning_rate": 6.082700283156247e-05, "loss": 0.0169, "step": 137700 }, { "grad_norm": 0.18259058892726898, "learning_rate": 6.082162196861718e-05, "loss": 0.0181, "step": 137710 }, { "grad_norm": 0.13942871987819672, "learning_rate": 6.081624097417896e-05, "loss": 0.0202, "step": 137720 }, { "grad_norm": 0.10151062160730362, "learning_rate": 6.0810859848313206e-05, "loss": 0.014, "step": 137730 }, { "grad_norm": 0.06804186850786209, "learning_rate": 6.08054785910853e-05, "loss": 0.0115, "step": 137740 }, { "grad_norm": 0.4302012026309967, "learning_rate": 6.0800097202560624e-05, "loss": 0.0172, "step": 137750 }, { "grad_norm": 0.1449354588985443, "learning_rate": 6.079471568280456e-05, "loss": 0.0147, "step": 137760 }, { "grad_norm": 0.06674744933843613, "learning_rate": 6.078933403188253e-05, "loss": 0.0167, "step": 137770 }, { "grad_norm": 0.13062329590320587, "learning_rate": 6.078395224985989e-05, "loss": 0.015, "step": 137780 }, { "grad_norm": 0.10479342937469482, "learning_rate": 6.0778570336802044e-05, "loss": 0.0119, "step": 137790 }, { "grad_norm": 0.1282823532819748, "learning_rate": 6.077318829277441e-05, "loss": 0.021, "step": 137800 }, { "grad_norm": 0.12466434389352798, "learning_rate": 6.076780611784234e-05, "loss": 0.0179, "step": 137810 }, { "grad_norm": 0.11432933062314987, "learning_rate": 6.076242381207129e-05, "loss": 0.0146, "step": 137820 }, { "grad_norm": 0.06955333054065704, "learning_rate": 6.07570413755266e-05, "loss": 0.0128, "step": 137830 }, { "grad_norm": 0.14475588500499725, "learning_rate": 6.0751658808273724e-05, "loss": 0.0124, "step": 137840 }, { "grad_norm": 0.09282048791646957, "learning_rate": 6.074627611037802e-05, "loss": 0.0164, "step": 137850 }, { "grad_norm": 0.09803377836942673, "learning_rate": 6.074089328190493e-05, "loss": 0.0166, "step": 137860 }, { "grad_norm": 0.12004958838224411, "learning_rate": 6.0735510322919855e-05, "loss": 0.0112, "step": 137870 }, { "grad_norm": 0.1889181137084961, "learning_rate": 6.073012723348818e-05, "loss": 0.0115, "step": 137880 }, { "grad_norm": 0.2381286472082138, "learning_rate": 6.0724744013675325e-05, "loss": 0.0168, "step": 137890 }, { "grad_norm": 0.11639814078807831, "learning_rate": 6.071936066354671e-05, "loss": 0.0117, "step": 137900 }, { "grad_norm": 0.13822004199028015, "learning_rate": 6.071397718316775e-05, "loss": 0.018, "step": 137910 }, { "grad_norm": 0.1291055828332901, "learning_rate": 6.070859357260383e-05, "loss": 0.0144, "step": 137920 }, { "grad_norm": 0.14886626601219177, "learning_rate": 6.070320983192042e-05, "loss": 0.0142, "step": 137930 }, { "grad_norm": 0.0749405175447464, "learning_rate": 6.069782596118288e-05, "loss": 0.0178, "step": 137940 }, { "grad_norm": 0.12821874022483826, "learning_rate": 6.0692441960456657e-05, "loss": 0.0215, "step": 137950 }, { "grad_norm": 0.14054934680461884, "learning_rate": 6.068705782980717e-05, "loss": 0.0136, "step": 137960 }, { "grad_norm": 0.6394185423851013, "learning_rate": 6.0681673569299834e-05, "loss": 0.0176, "step": 137970 }, { "grad_norm": 0.13176898658275604, "learning_rate": 6.06762891790001e-05, "loss": 0.0167, "step": 137980 }, { "grad_norm": 0.11053510755300522, "learning_rate": 6.067090465897335e-05, "loss": 0.0211, "step": 137990 }, { "grad_norm": 0.23455984890460968, "learning_rate": 6.066552000928505e-05, "loss": 0.0173, "step": 138000 }, { "grad_norm": 0.3292752504348755, "learning_rate": 6.0660135230000604e-05, "loss": 0.019, "step": 138010 }, { "grad_norm": 0.16627062857151031, "learning_rate": 6.065475032118545e-05, "loss": 0.012, "step": 138020 }, { "grad_norm": 0.15235869586467743, "learning_rate": 6.064936528290503e-05, "loss": 0.0134, "step": 138030 }, { "grad_norm": 0.2092062383890152, "learning_rate": 6.064398011522474e-05, "loss": 0.0157, "step": 138040 }, { "grad_norm": 0.1461312621831894, "learning_rate": 6.063859481821007e-05, "loss": 0.0165, "step": 138050 }, { "grad_norm": 0.07650327682495117, "learning_rate": 6.063320939192641e-05, "loss": 0.0164, "step": 138060 }, { "grad_norm": 0.09592972695827484, "learning_rate": 6.0627823836439234e-05, "loss": 0.0179, "step": 138070 }, { "grad_norm": 0.10736212879419327, "learning_rate": 6.062243815181394e-05, "loss": 0.0173, "step": 138080 }, { "grad_norm": 0.18822981417179108, "learning_rate": 6.061705233811602e-05, "loss": 0.0193, "step": 138090 }, { "grad_norm": 0.09550769627094269, "learning_rate": 6.0611666395410885e-05, "loss": 0.015, "step": 138100 }, { "grad_norm": 0.07738831639289856, "learning_rate": 6.0606280323763966e-05, "loss": 0.016, "step": 138110 }, { "grad_norm": 0.12887854874134064, "learning_rate": 6.060089412324075e-05, "loss": 0.0152, "step": 138120 }, { "grad_norm": 0.08690322935581207, "learning_rate": 6.059550779390664e-05, "loss": 0.0129, "step": 138130 }, { "grad_norm": 0.1548640877008438, "learning_rate": 6.059012133582712e-05, "loss": 0.0134, "step": 138140 }, { "grad_norm": 0.059133123606443405, "learning_rate": 6.058473474906762e-05, "loss": 0.0104, "step": 138150 }, { "grad_norm": 0.12261830270290375, "learning_rate": 6.0579348033693604e-05, "loss": 0.0173, "step": 138160 }, { "grad_norm": 0.1732933670282364, "learning_rate": 6.0573961189770515e-05, "loss": 0.0165, "step": 138170 }, { "grad_norm": 0.19725307822227478, "learning_rate": 6.056857421736383e-05, "loss": 0.0245, "step": 138180 }, { "grad_norm": 0.08828285336494446, "learning_rate": 6.056318711653897e-05, "loss": 0.0128, "step": 138190 }, { "grad_norm": 0.1115797609090805, "learning_rate": 6.055779988736142e-05, "loss": 0.0157, "step": 138200 }, { "grad_norm": 0.09055470675230026, "learning_rate": 6.055241252989663e-05, "loss": 0.0171, "step": 138210 }, { "grad_norm": 0.1828366219997406, "learning_rate": 6.054702504421007e-05, "loss": 0.0152, "step": 138220 }, { "grad_norm": 0.23718154430389404, "learning_rate": 6.054163743036719e-05, "loss": 0.0192, "step": 138230 }, { "grad_norm": 0.21350951492786407, "learning_rate": 6.053624968843348e-05, "loss": 0.0181, "step": 138240 }, { "grad_norm": 0.12501275539398193, "learning_rate": 6.053086181847437e-05, "loss": 0.0122, "step": 138250 }, { "grad_norm": 0.14569421112537384, "learning_rate": 6.0525473820555356e-05, "loss": 0.0145, "step": 138260 }, { "grad_norm": 0.26767638325691223, "learning_rate": 6.052008569474189e-05, "loss": 0.0155, "step": 138270 }, { "grad_norm": 0.14156194031238556, "learning_rate": 6.0514697441099465e-05, "loss": 0.0152, "step": 138280 }, { "grad_norm": 0.147054523229599, "learning_rate": 6.050930905969352e-05, "loss": 0.0147, "step": 138290 }, { "grad_norm": 0.14696043729782104, "learning_rate": 6.050392055058955e-05, "loss": 0.0215, "step": 138300 }, { "grad_norm": 0.07688665390014648, "learning_rate": 6.049853191385304e-05, "loss": 0.0155, "step": 138310 }, { "grad_norm": 0.2421511709690094, "learning_rate": 6.049314314954946e-05, "loss": 0.0162, "step": 138320 }, { "grad_norm": 0.0972016453742981, "learning_rate": 6.048775425774425e-05, "loss": 0.0128, "step": 138330 }, { "grad_norm": 0.06418229639530182, "learning_rate": 6.0482365238502956e-05, "loss": 0.0104, "step": 138340 }, { "grad_norm": 0.11640223115682602, "learning_rate": 6.0476976091891025e-05, "loss": 0.0173, "step": 138350 }, { "grad_norm": 0.2740243077278137, "learning_rate": 6.047158681797394e-05, "loss": 0.0241, "step": 138360 }, { "grad_norm": 0.10133030265569687, "learning_rate": 6.046619741681718e-05, "loss": 0.0135, "step": 138370 }, { "grad_norm": 0.48694199323654175, "learning_rate": 6.046080788848624e-05, "loss": 0.0208, "step": 138380 }, { "grad_norm": 0.13453996181488037, "learning_rate": 6.0455418233046614e-05, "loss": 0.0125, "step": 138390 }, { "grad_norm": 0.07559369504451752, "learning_rate": 6.0450028450563776e-05, "loss": 0.0154, "step": 138400 }, { "grad_norm": 0.12430458515882492, "learning_rate": 6.044463854110324e-05, "loss": 0.0133, "step": 138410 }, { "grad_norm": 0.09085933119058609, "learning_rate": 6.043924850473048e-05, "loss": 0.0176, "step": 138420 }, { "grad_norm": 0.19507277011871338, "learning_rate": 6.043385834151099e-05, "loss": 0.0176, "step": 138430 }, { "grad_norm": 0.09273119270801544, "learning_rate": 6.042846805151027e-05, "loss": 0.0204, "step": 138440 }, { "grad_norm": 0.08926201611757278, "learning_rate": 6.042307763479381e-05, "loss": 0.0166, "step": 138450 }, { "grad_norm": 0.19023266434669495, "learning_rate": 6.0417687091427124e-05, "loss": 0.0274, "step": 138460 }, { "grad_norm": 0.1356634795665741, "learning_rate": 6.04122964214757e-05, "loss": 0.0115, "step": 138470 }, { "grad_norm": 0.17251308262348175, "learning_rate": 6.040690562500505e-05, "loss": 0.0172, "step": 138480 }, { "grad_norm": 0.09126905351877213, "learning_rate": 6.0401514702080666e-05, "loss": 0.0164, "step": 138490 }, { "grad_norm": 0.1385766714811325, "learning_rate": 6.039612365276807e-05, "loss": 0.0146, "step": 138500 }, { "grad_norm": 0.1502036452293396, "learning_rate": 6.0390732477132736e-05, "loss": 0.0107, "step": 138510 }, { "grad_norm": 0.1529809981584549, "learning_rate": 6.0385341175240205e-05, "loss": 0.0082, "step": 138520 }, { "grad_norm": 0.14904703199863434, "learning_rate": 6.0379949747155974e-05, "loss": 0.0158, "step": 138530 }, { "grad_norm": 0.10309142619371414, "learning_rate": 6.037455819294554e-05, "loss": 0.0135, "step": 138540 }, { "grad_norm": 0.09584265202283859, "learning_rate": 6.036916651267445e-05, "loss": 0.0122, "step": 138550 }, { "grad_norm": 0.079738549888134, "learning_rate": 6.036377470640817e-05, "loss": 0.0131, "step": 138560 }, { "grad_norm": 0.11469767987728119, "learning_rate": 6.035838277421226e-05, "loss": 0.0183, "step": 138570 }, { "grad_norm": 0.2841876745223999, "learning_rate": 6.0352990716152224e-05, "loss": 0.021, "step": 138580 }, { "grad_norm": 0.10695274919271469, "learning_rate": 6.034759853229356e-05, "loss": 0.0113, "step": 138590 }, { "grad_norm": 0.06610599160194397, "learning_rate": 6.034220622270181e-05, "loss": 0.0175, "step": 138600 }, { "grad_norm": 0.1498304307460785, "learning_rate": 6.033681378744249e-05, "loss": 0.0162, "step": 138610 }, { "grad_norm": 0.10936364531517029, "learning_rate": 6.033142122658113e-05, "loss": 0.0138, "step": 138620 }, { "grad_norm": 0.3883359730243683, "learning_rate": 6.032602854018324e-05, "loss": 0.0257, "step": 138630 }, { "grad_norm": 0.11606889963150024, "learning_rate": 6.032063572831437e-05, "loss": 0.0175, "step": 138640 }, { "grad_norm": 0.2908848822116852, "learning_rate": 6.031524279104001e-05, "loss": 0.0177, "step": 138650 }, { "grad_norm": 0.1503383368253708, "learning_rate": 6.030984972842571e-05, "loss": 0.0207, "step": 138660 }, { "grad_norm": 0.295225590467453, "learning_rate": 6.030445654053702e-05, "loss": 0.0171, "step": 138670 }, { "grad_norm": 0.13155166804790497, "learning_rate": 6.0299063227439445e-05, "loss": 0.0188, "step": 138680 }, { "grad_norm": 0.1076279729604721, "learning_rate": 6.0293669789198524e-05, "loss": 0.0188, "step": 138690 }, { "grad_norm": 0.11235720664262772, "learning_rate": 6.0288276225879805e-05, "loss": 0.0193, "step": 138700 }, { "grad_norm": 0.1742643564939499, "learning_rate": 6.0282882537548815e-05, "loss": 0.0172, "step": 138710 }, { "grad_norm": 0.12535783648490906, "learning_rate": 6.027748872427108e-05, "loss": 0.0191, "step": 138720 }, { "grad_norm": 0.19874469935894012, "learning_rate": 6.027209478611217e-05, "loss": 0.02, "step": 138730 }, { "grad_norm": 0.1290811449289322, "learning_rate": 6.026670072313759e-05, "loss": 0.0101, "step": 138740 }, { "grad_norm": 0.08011610805988312, "learning_rate": 6.026130653541292e-05, "loss": 0.0106, "step": 138750 }, { "grad_norm": 0.17220014333724976, "learning_rate": 6.025591222300368e-05, "loss": 0.0212, "step": 138760 }, { "grad_norm": 0.09963548183441162, "learning_rate": 6.025051778597543e-05, "loss": 0.0113, "step": 138770 }, { "grad_norm": 0.16198237240314484, "learning_rate": 6.0245123224393705e-05, "loss": 0.0231, "step": 138780 }, { "grad_norm": 0.07930637151002884, "learning_rate": 6.023972853832406e-05, "loss": 0.0162, "step": 138790 }, { "grad_norm": 0.1946713775396347, "learning_rate": 6.023433372783205e-05, "loss": 0.0157, "step": 138800 }, { "grad_norm": 0.18711300194263458, "learning_rate": 6.022893879298321e-05, "loss": 0.0138, "step": 138810 }, { "grad_norm": 0.09922084212303162, "learning_rate": 6.022354373384312e-05, "loss": 0.0114, "step": 138820 }, { "grad_norm": 0.12155605107545853, "learning_rate": 6.0218148550477324e-05, "loss": 0.0129, "step": 138830 }, { "grad_norm": 0.12961088120937347, "learning_rate": 6.021275324295137e-05, "loss": 0.0137, "step": 138840 }, { "grad_norm": 0.13487543165683746, "learning_rate": 6.0207357811330825e-05, "loss": 0.0124, "step": 138850 }, { "grad_norm": 0.1022874042391777, "learning_rate": 6.020196225568123e-05, "loss": 0.0152, "step": 138860 }, { "grad_norm": 0.15742094814777374, "learning_rate": 6.0196566576068184e-05, "loss": 0.0149, "step": 138870 }, { "grad_norm": 0.1322903335094452, "learning_rate": 6.01911707725572e-05, "loss": 0.0211, "step": 138880 }, { "grad_norm": 0.1608923375606537, "learning_rate": 6.0185774845213884e-05, "loss": 0.0144, "step": 138890 }, { "grad_norm": 0.16416682302951813, "learning_rate": 6.018037879410379e-05, "loss": 0.017, "step": 138900 }, { "grad_norm": 0.1294044703245163, "learning_rate": 6.0174982619292476e-05, "loss": 0.0172, "step": 138910 }, { "grad_norm": 0.08575914800167084, "learning_rate": 6.016958632084552e-05, "loss": 0.0127, "step": 138920 }, { "grad_norm": 0.11164458841085434, "learning_rate": 6.016418989882848e-05, "loss": 0.0152, "step": 138930 }, { "grad_norm": 0.12001658976078033, "learning_rate": 6.015879335330694e-05, "loss": 0.0137, "step": 138940 }, { "grad_norm": 0.10605805367231369, "learning_rate": 6.0153396684346475e-05, "loss": 0.0134, "step": 138950 }, { "grad_norm": 0.11070495843887329, "learning_rate": 6.014799989201265e-05, "loss": 0.0176, "step": 138960 }, { "grad_norm": 0.13523393869400024, "learning_rate": 6.0142602976371044e-05, "loss": 0.0196, "step": 138970 }, { "grad_norm": 0.11615967750549316, "learning_rate": 6.013720593748724e-05, "loss": 0.0148, "step": 138980 }, { "grad_norm": 0.10088559985160828, "learning_rate": 6.013180877542681e-05, "loss": 0.0164, "step": 138990 }, { "grad_norm": 0.12296384572982788, "learning_rate": 6.012641149025533e-05, "loss": 0.0167, "step": 139000 }, { "grad_norm": 0.09436769038438797, "learning_rate": 6.012101408203841e-05, "loss": 0.0179, "step": 139010 }, { "grad_norm": 0.09780348837375641, "learning_rate": 6.0115616550841593e-05, "loss": 0.0125, "step": 139020 }, { "grad_norm": 0.1048646792769432, "learning_rate": 6.01102188967305e-05, "loss": 0.0186, "step": 139030 }, { "grad_norm": 0.10372133553028107, "learning_rate": 6.010482111977068e-05, "loss": 0.0126, "step": 139040 }, { "grad_norm": 0.06316618621349335, "learning_rate": 6.0099423220027764e-05, "loss": 0.0134, "step": 139050 }, { "grad_norm": 0.11985548585653305, "learning_rate": 6.0094025197567304e-05, "loss": 0.0126, "step": 139060 }, { "grad_norm": 0.1765945553779602, "learning_rate": 6.0088627052454914e-05, "loss": 0.0153, "step": 139070 }, { "grad_norm": 0.29715803265571594, "learning_rate": 6.0083228784756196e-05, "loss": 0.0148, "step": 139080 }, { "grad_norm": 0.20852547883987427, "learning_rate": 6.00778303945367e-05, "loss": 0.032, "step": 139090 }, { "grad_norm": 0.09652888774871826, "learning_rate": 6.007243188186207e-05, "loss": 0.0163, "step": 139100 }, { "grad_norm": 0.17404213547706604, "learning_rate": 6.006703324679788e-05, "loss": 0.0195, "step": 139110 }, { "grad_norm": 0.1096988394856453, "learning_rate": 6.006163448940972e-05, "loss": 0.012, "step": 139120 }, { "grad_norm": 0.1270533800125122, "learning_rate": 6.0056235609763215e-05, "loss": 0.019, "step": 139130 }, { "grad_norm": 0.4801273047924042, "learning_rate": 6.0050836607923945e-05, "loss": 0.0231, "step": 139140 }, { "grad_norm": 0.2439935952425003, "learning_rate": 6.004543748395752e-05, "loss": 0.0216, "step": 139150 }, { "grad_norm": 0.05034250020980835, "learning_rate": 6.0040038237929554e-05, "loss": 0.0153, "step": 139160 }, { "grad_norm": 0.20833289623260498, "learning_rate": 6.0034638869905636e-05, "loss": 0.0201, "step": 139170 }, { "grad_norm": 0.18872222304344177, "learning_rate": 6.002923937995137e-05, "loss": 0.0293, "step": 139180 }, { "grad_norm": 0.11857201159000397, "learning_rate": 6.00238397681324e-05, "loss": 0.0176, "step": 139190 }, { "grad_norm": 0.15895603597164154, "learning_rate": 6.00184400345143e-05, "loss": 0.0204, "step": 139200 }, { "grad_norm": 0.16952921450138092, "learning_rate": 6.0013040179162694e-05, "loss": 0.0178, "step": 139210 }, { "grad_norm": 0.1010458916425705, "learning_rate": 6.00076402021432e-05, "loss": 0.0102, "step": 139220 }, { "grad_norm": 0.4679874777793884, "learning_rate": 6.000224010352142e-05, "loss": 0.0199, "step": 139230 }, { "grad_norm": 0.14939402043819427, "learning_rate": 5.9996839883363e-05, "loss": 0.0145, "step": 139240 }, { "grad_norm": 0.1587388664484024, "learning_rate": 5.999143954173352e-05, "loss": 0.0128, "step": 139250 }, { "grad_norm": 0.11543264985084534, "learning_rate": 5.998603907869862e-05, "loss": 0.0173, "step": 139260 }, { "grad_norm": 0.09599380940198898, "learning_rate": 5.998063849432391e-05, "loss": 0.0109, "step": 139270 }, { "grad_norm": 0.12310590595006943, "learning_rate": 5.997523778867503e-05, "loss": 0.0174, "step": 139280 }, { "grad_norm": 0.1578083336353302, "learning_rate": 5.996983696181758e-05, "loss": 0.0141, "step": 139290 }, { "grad_norm": 0.08277206867933273, "learning_rate": 5.996443601381722e-05, "loss": 0.0127, "step": 139300 }, { "grad_norm": 0.2262851595878601, "learning_rate": 5.995903494473953e-05, "loss": 0.0138, "step": 139310 }, { "grad_norm": 0.0634511336684227, "learning_rate": 5.995363375465017e-05, "loss": 0.0122, "step": 139320 }, { "grad_norm": 0.18023909628391266, "learning_rate": 5.9948232443614774e-05, "loss": 0.0139, "step": 139330 }, { "grad_norm": 0.09230446070432663, "learning_rate": 5.9942831011698944e-05, "loss": 0.0132, "step": 139340 }, { "grad_norm": 0.13861197233200073, "learning_rate": 5.993742945896834e-05, "loss": 0.0166, "step": 139350 }, { "grad_norm": 0.18467974662780762, "learning_rate": 5.993202778548858e-05, "loss": 0.0198, "step": 139360 }, { "grad_norm": 0.09881716966629028, "learning_rate": 5.992662599132531e-05, "loss": 0.0175, "step": 139370 }, { "grad_norm": 0.36569905281066895, "learning_rate": 5.9921224076544155e-05, "loss": 0.0146, "step": 139380 }, { "grad_norm": 0.11348399519920349, "learning_rate": 5.9915822041210775e-05, "loss": 0.0161, "step": 139390 }, { "grad_norm": 0.16561459004878998, "learning_rate": 5.991041988539079e-05, "loss": 0.0252, "step": 139400 }, { "grad_norm": 0.13573649525642395, "learning_rate": 5.990501760914984e-05, "loss": 0.0185, "step": 139410 }, { "grad_norm": 0.08872012048959732, "learning_rate": 5.989961521255357e-05, "loss": 0.0128, "step": 139420 }, { "grad_norm": 0.08232317864894867, "learning_rate": 5.989421269566763e-05, "loss": 0.0112, "step": 139430 }, { "grad_norm": 0.1468646228313446, "learning_rate": 5.9888810058557674e-05, "loss": 0.0184, "step": 139440 }, { "grad_norm": 0.1901644915342331, "learning_rate": 5.988340730128933e-05, "loss": 0.0183, "step": 139450 }, { "grad_norm": 0.10517986863851547, "learning_rate": 5.9878004423928256e-05, "loss": 0.0124, "step": 139460 }, { "grad_norm": 0.1035671979188919, "learning_rate": 5.987260142654012e-05, "loss": 0.0119, "step": 139470 }, { "grad_norm": 0.11039090156555176, "learning_rate": 5.986719830919053e-05, "loss": 0.0157, "step": 139480 }, { "grad_norm": 0.11025992035865784, "learning_rate": 5.986179507194518e-05, "loss": 0.013, "step": 139490 }, { "grad_norm": 0.11441954970359802, "learning_rate": 5.98563917148697e-05, "loss": 0.0194, "step": 139500 }, { "grad_norm": 0.10941976308822632, "learning_rate": 5.985098823802976e-05, "loss": 0.0149, "step": 139510 }, { "grad_norm": 0.08742102235555649, "learning_rate": 5.9845584641491e-05, "loss": 0.0155, "step": 139520 }, { "grad_norm": 0.140565425157547, "learning_rate": 5.984018092531911e-05, "loss": 0.0181, "step": 139530 }, { "grad_norm": 0.14164069294929504, "learning_rate": 5.983477708957972e-05, "loss": 0.015, "step": 139540 }, { "grad_norm": 0.0938040018081665, "learning_rate": 5.9829373134338503e-05, "loss": 0.015, "step": 139550 }, { "grad_norm": 0.10787329077720642, "learning_rate": 5.9823969059661135e-05, "loss": 0.0149, "step": 139560 }, { "grad_norm": 0.13510973751544952, "learning_rate": 5.981856486561325e-05, "loss": 0.0136, "step": 139570 }, { "grad_norm": 0.1402433216571808, "learning_rate": 5.981316055226054e-05, "loss": 0.0211, "step": 139580 }, { "grad_norm": 0.19186420738697052, "learning_rate": 5.980775611966866e-05, "loss": 0.0183, "step": 139590 }, { "grad_norm": 0.09138347953557968, "learning_rate": 5.980235156790329e-05, "loss": 0.0145, "step": 139600 }, { "grad_norm": 0.13914088904857635, "learning_rate": 5.979694689703008e-05, "loss": 0.0216, "step": 139610 }, { "grad_norm": 0.1766551434993744, "learning_rate": 5.979154210711473e-05, "loss": 0.0177, "step": 139620 }, { "grad_norm": 0.15763913094997406, "learning_rate": 5.978613719822289e-05, "loss": 0.0145, "step": 139630 }, { "grad_norm": 0.13152219355106354, "learning_rate": 5.978073217042025e-05, "loss": 0.0157, "step": 139640 }, { "grad_norm": 0.1864786595106125, "learning_rate": 5.9775327023772465e-05, "loss": 0.0179, "step": 139650 }, { "grad_norm": 0.09085522592067719, "learning_rate": 5.9769921758345246e-05, "loss": 0.0192, "step": 139660 }, { "grad_norm": 0.09181506186723709, "learning_rate": 5.976451637420424e-05, "loss": 0.0143, "step": 139670 }, { "grad_norm": 0.19964268803596497, "learning_rate": 5.975911087141516e-05, "loss": 0.0175, "step": 139680 }, { "grad_norm": 0.15872612595558167, "learning_rate": 5.975370525004365e-05, "loss": 0.0151, "step": 139690 }, { "grad_norm": 0.15179669857025146, "learning_rate": 5.974829951015542e-05, "loss": 0.0162, "step": 139700 }, { "grad_norm": 0.21394039690494537, "learning_rate": 5.974289365181613e-05, "loss": 0.0151, "step": 139710 }, { "grad_norm": 0.15461601316928864, "learning_rate": 5.9737487675091506e-05, "loss": 0.015, "step": 139720 }, { "grad_norm": 0.13733480870723724, "learning_rate": 5.973208158004719e-05, "loss": 0.014, "step": 139730 }, { "grad_norm": 0.09962024539709091, "learning_rate": 5.972667536674892e-05, "loss": 0.0141, "step": 139740 }, { "grad_norm": 0.12044335156679153, "learning_rate": 5.972126903526235e-05, "loss": 0.017, "step": 139750 }, { "grad_norm": 0.158617302775383, "learning_rate": 5.971586258565318e-05, "loss": 0.0099, "step": 139760 }, { "grad_norm": 0.1285577118396759, "learning_rate": 5.971045601798711e-05, "loss": 0.0124, "step": 139770 }, { "grad_norm": 0.1838189959526062, "learning_rate": 5.970504933232983e-05, "loss": 0.0145, "step": 139780 }, { "grad_norm": 0.1098409965634346, "learning_rate": 5.9699642528747044e-05, "loss": 0.016, "step": 139790 }, { "grad_norm": 0.11515399068593979, "learning_rate": 5.9694235607304426e-05, "loss": 0.0146, "step": 139800 }, { "grad_norm": 0.15380214154720306, "learning_rate": 5.9688828568067715e-05, "loss": 0.0163, "step": 139810 }, { "grad_norm": 0.11486736685037613, "learning_rate": 5.968342141110257e-05, "loss": 0.02, "step": 139820 }, { "grad_norm": 0.17220477759838104, "learning_rate": 5.9678014136474726e-05, "loss": 0.0174, "step": 139830 }, { "grad_norm": 0.07322155684232712, "learning_rate": 5.967260674424986e-05, "loss": 0.0111, "step": 139840 }, { "grad_norm": 0.1599053591489792, "learning_rate": 5.96671992344937e-05, "loss": 0.0112, "step": 139850 }, { "grad_norm": 0.10714141279459, "learning_rate": 5.966179160727193e-05, "loss": 0.0159, "step": 139860 }, { "grad_norm": 0.10776706784963608, "learning_rate": 5.965638386265028e-05, "loss": 0.0148, "step": 139870 }, { "grad_norm": 0.1851876676082611, "learning_rate": 5.965097600069446e-05, "loss": 0.0176, "step": 139880 }, { "grad_norm": 0.10479741543531418, "learning_rate": 5.964556802147015e-05, "loss": 0.0195, "step": 139890 }, { "grad_norm": 0.13402323424816132, "learning_rate": 5.964015992504308e-05, "loss": 0.0395, "step": 139900 }, { "grad_norm": 0.2017955332994461, "learning_rate": 5.963475171147898e-05, "loss": 0.0168, "step": 139910 }, { "grad_norm": 0.07687176764011383, "learning_rate": 5.9629343380843536e-05, "loss": 0.0169, "step": 139920 }, { "grad_norm": 0.06599919497966766, "learning_rate": 5.9623934933202486e-05, "loss": 0.0166, "step": 139930 }, { "grad_norm": 0.10116682201623917, "learning_rate": 5.961852636862154e-05, "loss": 0.0126, "step": 139940 }, { "grad_norm": 0.08767761290073395, "learning_rate": 5.961311768716641e-05, "loss": 0.016, "step": 139950 }, { "grad_norm": 0.08617699146270752, "learning_rate": 5.960770888890283e-05, "loss": 0.0143, "step": 139960 }, { "grad_norm": 0.08501426875591278, "learning_rate": 5.960229997389652e-05, "loss": 0.0152, "step": 139970 }, { "grad_norm": 0.1322690099477768, "learning_rate": 5.9596890942213204e-05, "loss": 0.0202, "step": 139980 }, { "grad_norm": 0.1219075620174408, "learning_rate": 5.959148179391859e-05, "loss": 0.0114, "step": 139990 }, { "grad_norm": 0.142394557595253, "learning_rate": 5.958607252907842e-05, "loss": 0.0144, "step": 140000 }, { "grad_norm": 0.06660550087690353, "learning_rate": 5.958066314775842e-05, "loss": 0.0127, "step": 140010 }, { "grad_norm": 0.05309951305389404, "learning_rate": 5.957525365002431e-05, "loss": 0.0161, "step": 140020 }, { "grad_norm": 0.1362919807434082, "learning_rate": 5.956984403594184e-05, "loss": 0.0157, "step": 140030 }, { "grad_norm": 0.08715998381376266, "learning_rate": 5.956443430557672e-05, "loss": 0.0144, "step": 140040 }, { "grad_norm": 0.12214304506778717, "learning_rate": 5.9559024458994695e-05, "loss": 0.016, "step": 140050 }, { "grad_norm": 0.08918526768684387, "learning_rate": 5.95536144962615e-05, "loss": 0.0126, "step": 140060 }, { "grad_norm": 0.5556427240371704, "learning_rate": 5.954820441744285e-05, "loss": 0.015, "step": 140070 }, { "grad_norm": 0.2989130914211273, "learning_rate": 5.9542794222604526e-05, "loss": 0.0139, "step": 140080 }, { "grad_norm": 0.15785464644432068, "learning_rate": 5.953738391181223e-05, "loss": 0.0147, "step": 140090 }, { "grad_norm": 0.24404875934123993, "learning_rate": 5.953197348513172e-05, "loss": 0.0127, "step": 140100 }, { "grad_norm": 0.15327145159244537, "learning_rate": 5.952656294262873e-05, "loss": 0.016, "step": 140110 }, { "grad_norm": 0.15236622095108032, "learning_rate": 5.952115228436901e-05, "loss": 0.0216, "step": 140120 }, { "grad_norm": 0.16195501387119293, "learning_rate": 5.95157415104183e-05, "loss": 0.0132, "step": 140130 }, { "grad_norm": 0.1284332424402237, "learning_rate": 5.9510330620842333e-05, "loss": 0.0145, "step": 140140 }, { "grad_norm": 0.07495042681694031, "learning_rate": 5.9504919615706886e-05, "loss": 0.0167, "step": 140150 }, { "grad_norm": 0.24549464881420135, "learning_rate": 5.949950849507768e-05, "loss": 0.0151, "step": 140160 }, { "grad_norm": 0.2368658483028412, "learning_rate": 5.949409725902049e-05, "loss": 0.0166, "step": 140170 }, { "grad_norm": 0.12773869931697845, "learning_rate": 5.948868590760104e-05, "loss": 0.0173, "step": 140180 }, { "grad_norm": 0.2023477405309677, "learning_rate": 5.9483274440885103e-05, "loss": 0.0128, "step": 140190 }, { "grad_norm": 0.10271958261728287, "learning_rate": 5.947786285893844e-05, "loss": 0.0131, "step": 140200 }, { "grad_norm": 0.1850070059299469, "learning_rate": 5.947245116182678e-05, "loss": 0.0143, "step": 140210 }, { "grad_norm": 0.14099571108818054, "learning_rate": 5.9467039349615903e-05, "loss": 0.0168, "step": 140220 }, { "grad_norm": 0.11802025884389877, "learning_rate": 5.946162742237154e-05, "loss": 0.0226, "step": 140230 }, { "grad_norm": 0.08986556529998779, "learning_rate": 5.9456215380159496e-05, "loss": 0.017, "step": 140240 }, { "grad_norm": 0.1381634920835495, "learning_rate": 5.945080322304549e-05, "loss": 0.0207, "step": 140250 }, { "grad_norm": 0.1111593097448349, "learning_rate": 5.944539095109531e-05, "loss": 0.0116, "step": 140260 }, { "grad_norm": 0.12480764836072922, "learning_rate": 5.943997856437471e-05, "loss": 0.0166, "step": 140270 }, { "grad_norm": 0.2321794331073761, "learning_rate": 5.943456606294947e-05, "loss": 0.0149, "step": 140280 }, { "grad_norm": 0.08592381328344345, "learning_rate": 5.9429153446885335e-05, "loss": 0.0146, "step": 140290 }, { "grad_norm": 0.1278807669878006, "learning_rate": 5.942374071624808e-05, "loss": 0.0161, "step": 140300 }, { "grad_norm": 0.10024937987327576, "learning_rate": 5.9418327871103476e-05, "loss": 0.0121, "step": 140310 }, { "grad_norm": 1.1432241201400757, "learning_rate": 5.9412914911517294e-05, "loss": 0.0192, "step": 140320 }, { "grad_norm": 0.2621101140975952, "learning_rate": 5.940750183755531e-05, "loss": 0.0148, "step": 140330 }, { "grad_norm": 0.16845254600048065, "learning_rate": 5.9402088649283306e-05, "loss": 0.0131, "step": 140340 }, { "grad_norm": 0.13744758069515228, "learning_rate": 5.939667534676704e-05, "loss": 0.0158, "step": 140350 }, { "grad_norm": 0.13493256270885468, "learning_rate": 5.93912619300723e-05, "loss": 0.0228, "step": 140360 }, { "grad_norm": 0.114899180829525, "learning_rate": 5.938584839926485e-05, "loss": 0.0112, "step": 140370 }, { "grad_norm": 0.141776904463768, "learning_rate": 5.938043475441049e-05, "loss": 0.016, "step": 140380 }, { "grad_norm": 0.1517983078956604, "learning_rate": 5.9375020995574996e-05, "loss": 0.0155, "step": 140390 }, { "grad_norm": 0.2226739078760147, "learning_rate": 5.9369607122824135e-05, "loss": 0.0176, "step": 140400 }, { "grad_norm": 0.06794174015522003, "learning_rate": 5.936419313622371e-05, "loss": 0.0111, "step": 140410 }, { "grad_norm": 0.1647164523601532, "learning_rate": 5.935877903583948e-05, "loss": 0.017, "step": 140420 }, { "grad_norm": 0.09775218367576599, "learning_rate": 5.9353364821737265e-05, "loss": 0.0144, "step": 140430 }, { "grad_norm": 0.12869137525558472, "learning_rate": 5.934795049398283e-05, "loss": 0.0169, "step": 140440 }, { "grad_norm": 0.4355826675891876, "learning_rate": 5.934253605264198e-05, "loss": 0.0152, "step": 140450 }, { "grad_norm": 0.5439556241035461, "learning_rate": 5.9337121497780476e-05, "loss": 0.0136, "step": 140460 }, { "grad_norm": 0.2015688568353653, "learning_rate": 5.933170682946416e-05, "loss": 0.0127, "step": 140470 }, { "grad_norm": 0.10401516407728195, "learning_rate": 5.9326292047758766e-05, "loss": 0.0171, "step": 140480 }, { "grad_norm": 0.13224217295646667, "learning_rate": 5.932087715273014e-05, "loss": 0.0184, "step": 140490 }, { "grad_norm": 0.12638238072395325, "learning_rate": 5.931546214444405e-05, "loss": 0.0214, "step": 140500 }, { "grad_norm": 0.08125496655702591, "learning_rate": 5.93100470229663e-05, "loss": 0.0213, "step": 140510 }, { "grad_norm": 0.1054220199584961, "learning_rate": 5.930463178836269e-05, "loss": 0.0178, "step": 140520 }, { "grad_norm": 0.10175991803407669, "learning_rate": 5.9299216440699005e-05, "loss": 0.0117, "step": 140530 }, { "grad_norm": 0.22622334957122803, "learning_rate": 5.929380098004107e-05, "loss": 0.0212, "step": 140540 }, { "grad_norm": 0.1464157998561859, "learning_rate": 5.9288385406454674e-05, "loss": 0.0212, "step": 140550 }, { "grad_norm": 0.0672488585114479, "learning_rate": 5.9282969720005635e-05, "loss": 0.0137, "step": 140560 }, { "grad_norm": 0.10752002894878387, "learning_rate": 5.927755392075973e-05, "loss": 0.0146, "step": 140570 }, { "grad_norm": 0.10616806894540787, "learning_rate": 5.9272138008782805e-05, "loss": 0.0105, "step": 140580 }, { "grad_norm": 0.18513299524784088, "learning_rate": 5.9266721984140636e-05, "loss": 0.0173, "step": 140590 }, { "grad_norm": 0.08681151270866394, "learning_rate": 5.9261305846899063e-05, "loss": 0.0148, "step": 140600 }, { "grad_norm": 0.16712865233421326, "learning_rate": 5.925588959712387e-05, "loss": 0.0134, "step": 140610 }, { "grad_norm": 0.060059674084186554, "learning_rate": 5.925047323488088e-05, "loss": 0.0139, "step": 140620 }, { "grad_norm": 0.08615484833717346, "learning_rate": 5.924505676023591e-05, "loss": 0.0103, "step": 140630 }, { "grad_norm": 0.09138799458742142, "learning_rate": 5.9239640173254766e-05, "loss": 0.0084, "step": 140640 }, { "grad_norm": 0.08653005212545395, "learning_rate": 5.923422347400327e-05, "loss": 0.0155, "step": 140650 }, { "grad_norm": 0.14976397156715393, "learning_rate": 5.9228806662547234e-05, "loss": 0.0192, "step": 140660 }, { "grad_norm": 0.1354501098394394, "learning_rate": 5.9223389738952495e-05, "loss": 0.0156, "step": 140670 }, { "grad_norm": 0.14768186211585999, "learning_rate": 5.921797270328486e-05, "loss": 0.014, "step": 140680 }, { "grad_norm": 0.10309375077486038, "learning_rate": 5.921255555561015e-05, "loss": 0.016, "step": 140690 }, { "grad_norm": 0.0865187719464302, "learning_rate": 5.92071382959942e-05, "loss": 0.0171, "step": 140700 }, { "grad_norm": 0.11516379565000534, "learning_rate": 5.920172092450281e-05, "loss": 0.0163, "step": 140710 }, { "grad_norm": 0.20749610662460327, "learning_rate": 5.919630344120184e-05, "loss": 0.0164, "step": 140720 }, { "grad_norm": 0.11054585874080658, "learning_rate": 5.919088584615708e-05, "loss": 0.0164, "step": 140730 }, { "grad_norm": 0.19974057376384735, "learning_rate": 5.918546813943441e-05, "loss": 0.0174, "step": 140740 }, { "grad_norm": 0.13385836780071259, "learning_rate": 5.91800503210996e-05, "loss": 0.0154, "step": 140750 }, { "grad_norm": 0.15262456238269806, "learning_rate": 5.917463239121853e-05, "loss": 0.0185, "step": 140760 }, { "grad_norm": 0.227866068482399, "learning_rate": 5.916921434985702e-05, "loss": 0.0192, "step": 140770 }, { "grad_norm": 0.1503787338733673, "learning_rate": 5.916379619708088e-05, "loss": 0.0197, "step": 140780 }, { "grad_norm": 0.12811771035194397, "learning_rate": 5.915837793295598e-05, "loss": 0.0182, "step": 140790 }, { "grad_norm": 0.12924878299236298, "learning_rate": 5.9152959557548117e-05, "loss": 0.0122, "step": 140800 }, { "grad_norm": 0.6089279055595398, "learning_rate": 5.9147541070923174e-05, "loss": 0.0121, "step": 140810 }, { "grad_norm": 0.6139577627182007, "learning_rate": 5.914212247314696e-05, "loss": 0.0206, "step": 140820 }, { "grad_norm": 0.09109700471162796, "learning_rate": 5.913670376428533e-05, "loss": 0.014, "step": 140830 }, { "grad_norm": 0.09325173497200012, "learning_rate": 5.913128494440413e-05, "loss": 0.0173, "step": 140840 }, { "grad_norm": 0.10345077514648438, "learning_rate": 5.9125866013569186e-05, "loss": 0.0145, "step": 140850 }, { "grad_norm": 0.09930415451526642, "learning_rate": 5.9120446971846366e-05, "loss": 0.0141, "step": 140860 }, { "grad_norm": 0.07459211349487305, "learning_rate": 5.911502781930148e-05, "loss": 0.0147, "step": 140870 }, { "grad_norm": 0.12498396635055542, "learning_rate": 5.910960855600043e-05, "loss": 0.0153, "step": 140880 }, { "grad_norm": 0.07250145077705383, "learning_rate": 5.9104189182009016e-05, "loss": 0.0114, "step": 140890 }, { "grad_norm": 0.11859191954135895, "learning_rate": 5.909876969739312e-05, "loss": 0.0188, "step": 140900 }, { "grad_norm": 0.07155944406986237, "learning_rate": 5.909335010221857e-05, "loss": 0.0126, "step": 140910 }, { "grad_norm": 0.17935480177402496, "learning_rate": 5.9087930396551237e-05, "loss": 0.0156, "step": 140920 }, { "grad_norm": 0.07168471813201904, "learning_rate": 5.908251058045697e-05, "loss": 0.015, "step": 140930 }, { "grad_norm": 0.1211484968662262, "learning_rate": 5.907709065400162e-05, "loss": 0.0153, "step": 140940 }, { "grad_norm": 0.08273202180862427, "learning_rate": 5.907167061725105e-05, "loss": 0.0111, "step": 140950 }, { "grad_norm": 0.08274141699075699, "learning_rate": 5.906625047027111e-05, "loss": 0.0151, "step": 140960 }, { "grad_norm": 0.26837584376335144, "learning_rate": 5.906083021312768e-05, "loss": 0.0134, "step": 140970 }, { "grad_norm": 0.13641045987606049, "learning_rate": 5.90554098458866e-05, "loss": 0.016, "step": 140980 }, { "grad_norm": 0.15909060835838318, "learning_rate": 5.9049989368613745e-05, "loss": 0.0149, "step": 140990 }, { "grad_norm": 0.19708333909511566, "learning_rate": 5.904456878137498e-05, "loss": 0.0142, "step": 141000 }, { "grad_norm": 0.18692566454410553, "learning_rate": 5.903914808423614e-05, "loss": 0.0141, "step": 141010 }, { "grad_norm": 0.0997457206249237, "learning_rate": 5.903372727726314e-05, "loss": 0.012, "step": 141020 }, { "grad_norm": 0.107237309217453, "learning_rate": 5.9028306360521815e-05, "loss": 0.0176, "step": 141030 }, { "grad_norm": 0.10777591168880463, "learning_rate": 5.902288533407805e-05, "loss": 0.0131, "step": 141040 }, { "grad_norm": 0.43944093585014343, "learning_rate": 5.9017464197997695e-05, "loss": 0.0147, "step": 141050 }, { "grad_norm": 0.09870436042547226, "learning_rate": 5.901204295234666e-05, "loss": 0.0149, "step": 141060 }, { "grad_norm": 0.37947556376457214, "learning_rate": 5.900662159719077e-05, "loss": 0.0152, "step": 141070 }, { "grad_norm": 0.078074611723423, "learning_rate": 5.9001200132595934e-05, "loss": 0.0134, "step": 141080 }, { "grad_norm": 0.17321616411209106, "learning_rate": 5.899577855862802e-05, "loss": 0.0163, "step": 141090 }, { "grad_norm": 0.10905230045318604, "learning_rate": 5.89903568753529e-05, "loss": 0.0104, "step": 141100 }, { "grad_norm": 0.08825118839740753, "learning_rate": 5.8984935082836455e-05, "loss": 0.0167, "step": 141110 }, { "grad_norm": 0.10333997011184692, "learning_rate": 5.897951318114458e-05, "loss": 0.0129, "step": 141120 }, { "grad_norm": 0.1877516657114029, "learning_rate": 5.8974091170343114e-05, "loss": 0.016, "step": 141130 }, { "grad_norm": 0.14593657851219177, "learning_rate": 5.8968669050497995e-05, "loss": 0.0132, "step": 141140 }, { "grad_norm": 0.18510764837265015, "learning_rate": 5.8963246821675055e-05, "loss": 0.0152, "step": 141150 }, { "grad_norm": 0.15264779329299927, "learning_rate": 5.8957824483940225e-05, "loss": 0.0132, "step": 141160 }, { "grad_norm": 0.10262653976678848, "learning_rate": 5.895240203735936e-05, "loss": 0.0166, "step": 141170 }, { "grad_norm": 0.14265456795692444, "learning_rate": 5.894697948199835e-05, "loss": 0.0223, "step": 141180 }, { "grad_norm": 0.14800189435482025, "learning_rate": 5.8941556817923106e-05, "loss": 0.0159, "step": 141190 }, { "grad_norm": 0.15814225375652313, "learning_rate": 5.89361340451995e-05, "loss": 0.0208, "step": 141200 }, { "grad_norm": 0.25170615315437317, "learning_rate": 5.893071116389341e-05, "loss": 0.0158, "step": 141210 }, { "grad_norm": 0.1695844829082489, "learning_rate": 5.8925288174070783e-05, "loss": 0.0122, "step": 141220 }, { "grad_norm": 0.2748524844646454, "learning_rate": 5.891986507579745e-05, "loss": 0.0127, "step": 141230 }, { "grad_norm": 0.131617933511734, "learning_rate": 5.891444186913935e-05, "loss": 0.013, "step": 141240 }, { "grad_norm": 0.13181297481060028, "learning_rate": 5.890901855416235e-05, "loss": 0.015, "step": 141250 }, { "grad_norm": 1.032122015953064, "learning_rate": 5.890359513093238e-05, "loss": 0.0148, "step": 141260 }, { "grad_norm": 0.3108557164669037, "learning_rate": 5.8898171599515316e-05, "loss": 0.0167, "step": 141270 }, { "grad_norm": 0.13352419435977936, "learning_rate": 5.889274795997706e-05, "loss": 0.0143, "step": 141280 }, { "grad_norm": 0.16553500294685364, "learning_rate": 5.8887324212383523e-05, "loss": 0.0166, "step": 141290 }, { "grad_norm": 0.22775642573833466, "learning_rate": 5.8881900356800604e-05, "loss": 0.0197, "step": 141300 }, { "grad_norm": 0.10814119875431061, "learning_rate": 5.887647639329421e-05, "loss": 0.0168, "step": 141310 }, { "grad_norm": 0.08513914793729782, "learning_rate": 5.8871052321930255e-05, "loss": 0.0148, "step": 141320 }, { "grad_norm": 0.08450048416852951, "learning_rate": 5.886562814277463e-05, "loss": 0.0153, "step": 141330 }, { "grad_norm": 0.08152349293231964, "learning_rate": 5.886020385589326e-05, "loss": 0.0127, "step": 141340 }, { "grad_norm": 0.10155685245990753, "learning_rate": 5.885477946135204e-05, "loss": 0.0143, "step": 141350 }, { "grad_norm": 0.08839963376522064, "learning_rate": 5.88493549592169e-05, "loss": 0.0104, "step": 141360 }, { "grad_norm": 0.15991763770580292, "learning_rate": 5.8843930349553736e-05, "loss": 0.0146, "step": 141370 }, { "grad_norm": 0.36925816535949707, "learning_rate": 5.8838505632428466e-05, "loss": 0.0128, "step": 141380 }, { "grad_norm": 0.11680290102958679, "learning_rate": 5.8833080807907e-05, "loss": 0.0119, "step": 141390 }, { "grad_norm": 0.09733286499977112, "learning_rate": 5.8827655876055274e-05, "loss": 0.0093, "step": 141400 }, { "grad_norm": 0.08938135206699371, "learning_rate": 5.88222308369392e-05, "loss": 0.0217, "step": 141410 }, { "grad_norm": 0.08772674947977066, "learning_rate": 5.881680569062468e-05, "loss": 0.0151, "step": 141420 }, { "grad_norm": 0.1085706502199173, "learning_rate": 5.8811380437177654e-05, "loss": 0.0166, "step": 141430 }, { "grad_norm": 0.10503549128770828, "learning_rate": 5.880595507666402e-05, "loss": 0.0168, "step": 141440 }, { "grad_norm": 0.08137332648038864, "learning_rate": 5.8800529609149734e-05, "loss": 0.0139, "step": 141450 }, { "grad_norm": 0.1018388420343399, "learning_rate": 5.87951040347007e-05, "loss": 0.0126, "step": 141460 }, { "grad_norm": 0.20275312662124634, "learning_rate": 5.878967835338285e-05, "loss": 0.02, "step": 141470 }, { "grad_norm": 0.1591614931821823, "learning_rate": 5.87842525652621e-05, "loss": 0.0161, "step": 141480 }, { "grad_norm": 1.0813177824020386, "learning_rate": 5.87788266704044e-05, "loss": 0.0142, "step": 141490 }, { "grad_norm": 0.16188500821590424, "learning_rate": 5.877340066887567e-05, "loss": 0.0172, "step": 141500 }, { "grad_norm": 0.07556582987308502, "learning_rate": 5.876797456074182e-05, "loss": 0.0189, "step": 141510 }, { "grad_norm": 0.13865475356578827, "learning_rate": 5.8762548346068826e-05, "loss": 0.0188, "step": 141520 }, { "grad_norm": 0.10459568351507187, "learning_rate": 5.8757122024922576e-05, "loss": 0.0216, "step": 141530 }, { "grad_norm": 0.11721218377351761, "learning_rate": 5.8751695597369025e-05, "loss": 0.0176, "step": 141540 }, { "grad_norm": 0.09232106804847717, "learning_rate": 5.8746269063474114e-05, "loss": 0.0151, "step": 141550 }, { "grad_norm": 0.09397472441196442, "learning_rate": 5.8740842423303773e-05, "loss": 0.0181, "step": 141560 }, { "grad_norm": 0.09205315262079239, "learning_rate": 5.8735415676923965e-05, "loss": 0.0109, "step": 141570 }, { "grad_norm": 0.31761762499809265, "learning_rate": 5.872998882440058e-05, "loss": 0.0124, "step": 141580 }, { "grad_norm": 0.1891409456729889, "learning_rate": 5.872456186579961e-05, "loss": 0.0139, "step": 141590 }, { "grad_norm": 0.13527201116085052, "learning_rate": 5.871913480118696e-05, "loss": 0.0173, "step": 141600 }, { "grad_norm": 0.07432372123003006, "learning_rate": 5.87137076306286e-05, "loss": 0.0165, "step": 141610 }, { "grad_norm": 0.10579230636358261, "learning_rate": 5.870828035419046e-05, "loss": 0.0177, "step": 141620 }, { "grad_norm": 0.10641959309577942, "learning_rate": 5.870285297193849e-05, "loss": 0.0184, "step": 141630 }, { "grad_norm": 0.7217429280281067, "learning_rate": 5.869742548393864e-05, "loss": 0.0141, "step": 141640 }, { "grad_norm": 0.271190345287323, "learning_rate": 5.869199789025687e-05, "loss": 0.0164, "step": 141650 }, { "grad_norm": 0.16406705975532532, "learning_rate": 5.868657019095911e-05, "loss": 0.0126, "step": 141660 }, { "grad_norm": 0.08628792315721512, "learning_rate": 5.868114238611132e-05, "loss": 0.0164, "step": 141670 }, { "grad_norm": 0.08116069436073303, "learning_rate": 5.867571447577945e-05, "loss": 0.02, "step": 141680 }, { "grad_norm": 0.21566392481327057, "learning_rate": 5.867028646002946e-05, "loss": 0.015, "step": 141690 }, { "grad_norm": 0.17428691685199738, "learning_rate": 5.866485833892732e-05, "loss": 0.0178, "step": 141700 }, { "grad_norm": 0.11815300583839417, "learning_rate": 5.865943011253895e-05, "loss": 0.0127, "step": 141710 }, { "grad_norm": 0.08072513341903687, "learning_rate": 5.8654001780930334e-05, "loss": 0.0187, "step": 141720 }, { "grad_norm": 0.08706173300743103, "learning_rate": 5.8648573344167436e-05, "loss": 0.012, "step": 141730 }, { "grad_norm": 0.11251585185527802, "learning_rate": 5.864314480231619e-05, "loss": 0.0136, "step": 141740 }, { "grad_norm": 0.12384852021932602, "learning_rate": 5.863771615544258e-05, "loss": 0.0146, "step": 141750 }, { "grad_norm": 0.16249950230121613, "learning_rate": 5.8632287403612564e-05, "loss": 0.0127, "step": 141760 }, { "grad_norm": 0.09244554489850998, "learning_rate": 5.8626858546892106e-05, "loss": 0.0135, "step": 141770 }, { "grad_norm": 0.21643872559070587, "learning_rate": 5.862142958534717e-05, "loss": 0.012, "step": 141780 }, { "grad_norm": 0.12152918428182602, "learning_rate": 5.861600051904372e-05, "loss": 0.0129, "step": 141790 }, { "grad_norm": 0.13554561138153076, "learning_rate": 5.861057134804774e-05, "loss": 0.0141, "step": 141800 }, { "grad_norm": 0.16273772716522217, "learning_rate": 5.860514207242518e-05, "loss": 0.0141, "step": 141810 }, { "grad_norm": 0.2393989861011505, "learning_rate": 5.859971269224203e-05, "loss": 0.0164, "step": 141820 }, { "grad_norm": 0.24515482783317566, "learning_rate": 5.859428320756424e-05, "loss": 0.0176, "step": 141830 }, { "grad_norm": 0.2069588452577591, "learning_rate": 5.8588853618457795e-05, "loss": 0.0147, "step": 141840 }, { "grad_norm": 0.08870920538902283, "learning_rate": 5.858342392498868e-05, "loss": 0.0134, "step": 141850 }, { "grad_norm": 0.14021800458431244, "learning_rate": 5.857799412722285e-05, "loss": 0.0123, "step": 141860 }, { "grad_norm": 0.23434530198574066, "learning_rate": 5.857256422522629e-05, "loss": 0.0165, "step": 141870 }, { "grad_norm": 0.10044943541288376, "learning_rate": 5.856713421906498e-05, "loss": 0.0188, "step": 141880 }, { "grad_norm": 0.16403859853744507, "learning_rate": 5.8561704108804916e-05, "loss": 0.0179, "step": 141890 }, { "grad_norm": 0.08294078707695007, "learning_rate": 5.855627389451205e-05, "loss": 0.017, "step": 141900 }, { "grad_norm": 0.1891210377216339, "learning_rate": 5.855084357625238e-05, "loss": 0.0231, "step": 141910 }, { "grad_norm": 0.12211233377456665, "learning_rate": 5.854541315409189e-05, "loss": 0.017, "step": 141920 }, { "grad_norm": 0.10573994368314743, "learning_rate": 5.8539982628096556e-05, "loss": 0.0143, "step": 141930 }, { "grad_norm": 0.10743618756532669, "learning_rate": 5.853455199833238e-05, "loss": 0.0165, "step": 141940 }, { "grad_norm": 0.28116583824157715, "learning_rate": 5.852912126486534e-05, "loss": 0.0169, "step": 141950 }, { "grad_norm": 0.13051222264766693, "learning_rate": 5.85236904277614e-05, "loss": 0.016, "step": 141960 }, { "grad_norm": 0.1510370969772339, "learning_rate": 5.85182594870866e-05, "loss": 0.0135, "step": 141970 }, { "grad_norm": 0.12142506241798401, "learning_rate": 5.851282844290691e-05, "loss": 0.0125, "step": 141980 }, { "grad_norm": 0.21023313701152802, "learning_rate": 5.850739729528829e-05, "loss": 0.015, "step": 141990 }, { "grad_norm": 0.10798586159944534, "learning_rate": 5.8501966044296784e-05, "loss": 0.013, "step": 142000 }, { "grad_norm": 0.06489396840333939, "learning_rate": 5.849653468999834e-05, "loss": 0.0155, "step": 142010 }, { "grad_norm": 0.12087957561016083, "learning_rate": 5.8491103232459e-05, "loss": 0.02, "step": 142020 }, { "grad_norm": 0.15932999551296234, "learning_rate": 5.8485671671744726e-05, "loss": 0.018, "step": 142030 }, { "grad_norm": 0.11244548857212067, "learning_rate": 5.848024000792154e-05, "loss": 0.013, "step": 142040 }, { "grad_norm": 0.15422867238521576, "learning_rate": 5.8474808241055426e-05, "loss": 0.0176, "step": 142050 }, { "grad_norm": 0.10645399242639542, "learning_rate": 5.846937637121238e-05, "loss": 0.0185, "step": 142060 }, { "grad_norm": 0.09045510739088058, "learning_rate": 5.846394439845843e-05, "loss": 0.0149, "step": 142070 }, { "grad_norm": 0.09491837024688721, "learning_rate": 5.8458512322859546e-05, "loss": 0.0188, "step": 142080 }, { "grad_norm": 0.10988420993089676, "learning_rate": 5.8453080144481774e-05, "loss": 0.0163, "step": 142090 }, { "grad_norm": 0.10275460034608841, "learning_rate": 5.844764786339107e-05, "loss": 0.0141, "step": 142100 }, { "grad_norm": 0.11662594974040985, "learning_rate": 5.8442215479653496e-05, "loss": 0.0191, "step": 142110 }, { "grad_norm": 0.07571802288293839, "learning_rate": 5.843678299333501e-05, "loss": 0.0165, "step": 142120 }, { "grad_norm": 0.5248257517814636, "learning_rate": 5.843135040450167e-05, "loss": 0.0157, "step": 142130 }, { "grad_norm": 0.08192315697669983, "learning_rate": 5.842591771321945e-05, "loss": 0.0203, "step": 142140 }, { "grad_norm": 0.11095934361219406, "learning_rate": 5.842048491955436e-05, "loss": 0.012, "step": 142150 }, { "grad_norm": 0.10676997900009155, "learning_rate": 5.841505202357245e-05, "loss": 0.012, "step": 142160 }, { "grad_norm": 0.07075752317905426, "learning_rate": 5.84096190253397e-05, "loss": 0.0157, "step": 142170 }, { "grad_norm": 0.09371727705001831, "learning_rate": 5.840418592492215e-05, "loss": 0.0152, "step": 142180 }, { "grad_norm": 0.1114969253540039, "learning_rate": 5.83987527223858e-05, "loss": 0.0153, "step": 142190 }, { "grad_norm": 0.27090734243392944, "learning_rate": 5.8393319417796664e-05, "loss": 0.0234, "step": 142200 }, { "grad_norm": 0.18219445645809174, "learning_rate": 5.838788601122078e-05, "loss": 0.0146, "step": 142210 }, { "grad_norm": 0.22857190668582916, "learning_rate": 5.838245250272417e-05, "loss": 0.0149, "step": 142220 }, { "grad_norm": 0.0963795930147171, "learning_rate": 5.837701889237285e-05, "loss": 0.0159, "step": 142230 }, { "grad_norm": 0.16182467341423035, "learning_rate": 5.8371585180232823e-05, "loss": 0.015, "step": 142240 }, { "grad_norm": 0.08062035590410233, "learning_rate": 5.8366151366370145e-05, "loss": 0.0111, "step": 142250 }, { "grad_norm": 0.09114442765712738, "learning_rate": 5.836071745085082e-05, "loss": 0.0115, "step": 142260 }, { "grad_norm": 0.06891073286533356, "learning_rate": 5.8355283433740893e-05, "loss": 0.0154, "step": 142270 }, { "grad_norm": 0.11122462153434753, "learning_rate": 5.8349849315106376e-05, "loss": 0.0191, "step": 142280 }, { "grad_norm": 0.06385531276464462, "learning_rate": 5.8344415095013316e-05, "loss": 0.0137, "step": 142290 }, { "grad_norm": 0.08412857353687286, "learning_rate": 5.833898077352773e-05, "loss": 0.0115, "step": 142300 }, { "grad_norm": 0.2161054015159607, "learning_rate": 5.833354635071565e-05, "loss": 0.0177, "step": 142310 }, { "grad_norm": 0.13143140077590942, "learning_rate": 5.8328111826643114e-05, "loss": 0.0136, "step": 142320 }, { "grad_norm": 0.09484117478132248, "learning_rate": 5.832267720137616e-05, "loss": 0.0134, "step": 142330 }, { "grad_norm": 0.22573573887348175, "learning_rate": 5.8317242474980826e-05, "loss": 0.0151, "step": 142340 }, { "grad_norm": 0.13736848533153534, "learning_rate": 5.831180764752313e-05, "loss": 0.0201, "step": 142350 }, { "grad_norm": 0.14067049324512482, "learning_rate": 5.830637271906914e-05, "loss": 0.0099, "step": 142360 }, { "grad_norm": 0.19538775086402893, "learning_rate": 5.830093768968485e-05, "loss": 0.0155, "step": 142370 }, { "grad_norm": 0.06780251115560532, "learning_rate": 5.829550255943636e-05, "loss": 0.0129, "step": 142380 }, { "grad_norm": 0.11611487716436386, "learning_rate": 5.829006732838966e-05, "loss": 0.0154, "step": 142390 }, { "grad_norm": 0.09480096399784088, "learning_rate": 5.8284631996610827e-05, "loss": 0.013, "step": 142400 }, { "grad_norm": 0.17590084671974182, "learning_rate": 5.827919656416589e-05, "loss": 0.0184, "step": 142410 }, { "grad_norm": 0.09560075402259827, "learning_rate": 5.8273761031120896e-05, "loss": 0.0161, "step": 142420 }, { "grad_norm": 0.07866907864809036, "learning_rate": 5.82683253975419e-05, "loss": 0.0145, "step": 142430 }, { "grad_norm": 0.132740318775177, "learning_rate": 5.826288966349493e-05, "loss": 0.0248, "step": 142440 }, { "grad_norm": 0.08785248547792435, "learning_rate": 5.8257453829046074e-05, "loss": 0.0145, "step": 142450 }, { "grad_norm": 0.17976044118404388, "learning_rate": 5.825201789426135e-05, "loss": 0.0171, "step": 142460 }, { "grad_norm": 0.14768238365650177, "learning_rate": 5.8246581859206796e-05, "loss": 0.0136, "step": 142470 }, { "grad_norm": 0.45359736680984497, "learning_rate": 5.8241145723948506e-05, "loss": 0.0181, "step": 142480 }, { "grad_norm": 0.11573915183544159, "learning_rate": 5.8235709488552504e-05, "loss": 0.0167, "step": 142490 }, { "grad_norm": 0.16131705045700073, "learning_rate": 5.823027315308486e-05, "loss": 0.0125, "step": 142500 }, { "grad_norm": 0.06801220029592514, "learning_rate": 5.822483671761163e-05, "loss": 0.0137, "step": 142510 }, { "grad_norm": 0.11329922825098038, "learning_rate": 5.821940018219886e-05, "loss": 0.0186, "step": 142520 }, { "grad_norm": 0.22755533456802368, "learning_rate": 5.8213963546912634e-05, "loss": 0.0181, "step": 142530 }, { "grad_norm": 0.12587593495845795, "learning_rate": 5.820852681181898e-05, "loss": 0.0202, "step": 142540 }, { "grad_norm": 0.15874914824962616, "learning_rate": 5.8203089976983985e-05, "loss": 0.023, "step": 142550 }, { "grad_norm": 0.11818601936101913, "learning_rate": 5.8197653042473686e-05, "loss": 0.018, "step": 142560 }, { "grad_norm": 0.15148261189460754, "learning_rate": 5.819221600835418e-05, "loss": 0.0124, "step": 142570 }, { "grad_norm": 0.15013307332992554, "learning_rate": 5.8186778874691504e-05, "loss": 0.0164, "step": 142580 }, { "grad_norm": 0.12183510512113571, "learning_rate": 5.8181341641551736e-05, "loss": 0.0145, "step": 142590 }, { "grad_norm": 0.13525323569774628, "learning_rate": 5.817590430900094e-05, "loss": 0.0143, "step": 142600 }, { "grad_norm": 0.11359020322561264, "learning_rate": 5.81704668771052e-05, "loss": 0.0116, "step": 142610 }, { "grad_norm": 0.10344572365283966, "learning_rate": 5.8165029345930575e-05, "loss": 0.0178, "step": 142620 }, { "grad_norm": 0.1497708261013031, "learning_rate": 5.815959171554312e-05, "loss": 0.0244, "step": 142630 }, { "grad_norm": 0.1359218955039978, "learning_rate": 5.815415398600893e-05, "loss": 0.0148, "step": 142640 }, { "grad_norm": 0.11484482884407043, "learning_rate": 5.814871615739407e-05, "loss": 0.0172, "step": 142650 }, { "grad_norm": 0.10379769653081894, "learning_rate": 5.814327822976461e-05, "loss": 0.0166, "step": 142660 }, { "grad_norm": 0.11493199318647385, "learning_rate": 5.8137840203186636e-05, "loss": 0.0167, "step": 142670 }, { "grad_norm": 0.08397289365530014, "learning_rate": 5.8132402077726215e-05, "loss": 0.0135, "step": 142680 }, { "grad_norm": 0.13912931084632874, "learning_rate": 5.8126963853449425e-05, "loss": 0.0113, "step": 142690 }, { "grad_norm": 0.14874553680419922, "learning_rate": 5.812152553042237e-05, "loss": 0.0128, "step": 142700 }, { "grad_norm": 0.21143421530723572, "learning_rate": 5.8116087108711104e-05, "loss": 0.0181, "step": 142710 }, { "grad_norm": 0.08550455421209335, "learning_rate": 5.811064858838171e-05, "loss": 0.0166, "step": 142720 }, { "grad_norm": 0.08100500702857971, "learning_rate": 5.810520996950028e-05, "loss": 0.018, "step": 142730 }, { "grad_norm": 0.12840960919857025, "learning_rate": 5.809977125213289e-05, "loss": 0.0151, "step": 142740 }, { "grad_norm": 0.05996745079755783, "learning_rate": 5.809433243634565e-05, "loss": 0.0081, "step": 142750 }, { "grad_norm": 0.0940227210521698, "learning_rate": 5.8088893522204614e-05, "loss": 0.0097, "step": 142760 }, { "grad_norm": 0.17423343658447266, "learning_rate": 5.808345450977589e-05, "loss": 0.0156, "step": 142770 }, { "grad_norm": 0.2022884488105774, "learning_rate": 5.8078015399125573e-05, "loss": 0.0221, "step": 142780 }, { "grad_norm": 0.14525623619556427, "learning_rate": 5.8072576190319716e-05, "loss": 0.0136, "step": 142790 }, { "grad_norm": 0.10984179377555847, "learning_rate": 5.806713688342447e-05, "loss": 0.0218, "step": 142800 }, { "grad_norm": 0.1417614221572876, "learning_rate": 5.806169747850586e-05, "loss": 0.0164, "step": 142810 }, { "grad_norm": 0.0945950374007225, "learning_rate": 5.8056257975630036e-05, "loss": 0.0187, "step": 142820 }, { "grad_norm": 0.18781717121601105, "learning_rate": 5.8050818374863056e-05, "loss": 0.0154, "step": 142830 }, { "grad_norm": 0.09591129422187805, "learning_rate": 5.804537867627104e-05, "loss": 0.0122, "step": 142840 }, { "grad_norm": 0.1080113872885704, "learning_rate": 5.8039938879920076e-05, "loss": 0.0186, "step": 142850 }, { "grad_norm": 0.18385855853557587, "learning_rate": 5.803449898587627e-05, "loss": 0.015, "step": 142860 }, { "grad_norm": 0.1216120570898056, "learning_rate": 5.802905899420571e-05, "loss": 0.0122, "step": 142870 }, { "grad_norm": 0.08074580132961273, "learning_rate": 5.8023618904974494e-05, "loss": 0.0119, "step": 142880 }, { "grad_norm": 0.09299837052822113, "learning_rate": 5.801817871824875e-05, "loss": 0.0136, "step": 142890 }, { "grad_norm": 0.2777068316936493, "learning_rate": 5.801273843409455e-05, "loss": 0.017, "step": 142900 }, { "grad_norm": 0.11946485191583633, "learning_rate": 5.800729805257803e-05, "loss": 0.0158, "step": 142910 }, { "grad_norm": 0.15357813239097595, "learning_rate": 5.800185757376525e-05, "loss": 0.0145, "step": 142920 }, { "grad_norm": 0.13925114274024963, "learning_rate": 5.799641699772237e-05, "loss": 0.0149, "step": 142930 }, { "grad_norm": 0.09990258514881134, "learning_rate": 5.799097632451547e-05, "loss": 0.0154, "step": 142940 }, { "grad_norm": 0.08911434561014175, "learning_rate": 5.798553555421066e-05, "loss": 0.0128, "step": 142950 }, { "grad_norm": 0.08219058811664581, "learning_rate": 5.7980094686874054e-05, "loss": 0.0146, "step": 142960 }, { "grad_norm": 0.5459228754043579, "learning_rate": 5.7974653722571746e-05, "loss": 0.0185, "step": 142970 }, { "grad_norm": 0.10712769627571106, "learning_rate": 5.7969212661369885e-05, "loss": 0.017, "step": 142980 }, { "grad_norm": 0.25676265358924866, "learning_rate": 5.796377150333456e-05, "loss": 0.0158, "step": 142990 }, { "grad_norm": 0.45843446254730225, "learning_rate": 5.79583302485319e-05, "loss": 0.0183, "step": 143000 }, { "grad_norm": 0.21024519205093384, "learning_rate": 5.7952888897028e-05, "loss": 0.0144, "step": 143010 }, { "grad_norm": 0.1790638267993927, "learning_rate": 5.7947447448889e-05, "loss": 0.0157, "step": 143020 }, { "grad_norm": 0.1188083291053772, "learning_rate": 5.7942005904181016e-05, "loss": 0.0189, "step": 143030 }, { "grad_norm": 0.15516731142997742, "learning_rate": 5.793656426297015e-05, "loss": 0.0203, "step": 143040 }, { "grad_norm": 0.11783348768949509, "learning_rate": 5.793112252532255e-05, "loss": 0.0156, "step": 143050 }, { "grad_norm": 0.1042075827717781, "learning_rate": 5.7925680691304305e-05, "loss": 0.0146, "step": 143060 }, { "grad_norm": 0.08850222080945969, "learning_rate": 5.792023876098157e-05, "loss": 0.0187, "step": 143070 }, { "grad_norm": 0.14516371488571167, "learning_rate": 5.791479673442044e-05, "loss": 0.0229, "step": 143080 }, { "grad_norm": 0.09384093433618546, "learning_rate": 5.790935461168707e-05, "loss": 0.0125, "step": 143090 }, { "grad_norm": 0.22930847108364105, "learning_rate": 5.790391239284758e-05, "loss": 0.0218, "step": 143100 }, { "grad_norm": 0.09072723239660263, "learning_rate": 5.789847007796807e-05, "loss": 0.0104, "step": 143110 }, { "grad_norm": 0.20169806480407715, "learning_rate": 5.789302766711471e-05, "loss": 0.0225, "step": 143120 }, { "grad_norm": 0.15729452669620514, "learning_rate": 5.78875851603536e-05, "loss": 0.012, "step": 143130 }, { "grad_norm": 0.1395995020866394, "learning_rate": 5.7882142557750895e-05, "loss": 0.013, "step": 143140 }, { "grad_norm": 0.18475118279457092, "learning_rate": 5.7876699859372704e-05, "loss": 0.0174, "step": 143150 }, { "grad_norm": 0.08976823836565018, "learning_rate": 5.787125706528518e-05, "loss": 0.0117, "step": 143160 }, { "grad_norm": 0.16231201589107513, "learning_rate": 5.7865814175554445e-05, "loss": 0.0209, "step": 143170 }, { "grad_norm": 0.24208934605121613, "learning_rate": 5.7860371190246656e-05, "loss": 0.022, "step": 143180 }, { "grad_norm": 0.1481524109840393, "learning_rate": 5.785492810942792e-05, "loss": 0.0185, "step": 143190 }, { "grad_norm": 0.10234472900629044, "learning_rate": 5.78494849331644e-05, "loss": 0.0153, "step": 143200 }, { "grad_norm": 0.1784210205078125, "learning_rate": 5.784404166152222e-05, "loss": 0.0142, "step": 143210 }, { "grad_norm": 0.09770254790782928, "learning_rate": 5.7838598294567525e-05, "loss": 0.0181, "step": 143220 }, { "grad_norm": 0.09227512776851654, "learning_rate": 5.7833154832366465e-05, "loss": 0.0114, "step": 143230 }, { "grad_norm": 0.23456966876983643, "learning_rate": 5.7827711274985186e-05, "loss": 0.0123, "step": 143240 }, { "grad_norm": 0.09056717157363892, "learning_rate": 5.782226762248981e-05, "loss": 0.0142, "step": 143250 }, { "grad_norm": 0.09980004280805588, "learning_rate": 5.781682387494651e-05, "loss": 0.0132, "step": 143260 }, { "grad_norm": 0.10277003049850464, "learning_rate": 5.7811380032421405e-05, "loss": 0.0121, "step": 143270 }, { "grad_norm": 0.15322157740592957, "learning_rate": 5.780593609498067e-05, "loss": 0.017, "step": 143280 }, { "grad_norm": 0.09948983788490295, "learning_rate": 5.7800492062690435e-05, "loss": 0.013, "step": 143290 }, { "grad_norm": 0.20588675141334534, "learning_rate": 5.7795047935616866e-05, "loss": 0.0119, "step": 143300 }, { "grad_norm": 0.10957220196723938, "learning_rate": 5.778960371382609e-05, "loss": 0.0158, "step": 143310 }, { "grad_norm": 0.31167393922805786, "learning_rate": 5.7784159397384283e-05, "loss": 0.016, "step": 143320 }, { "grad_norm": 0.2614123821258545, "learning_rate": 5.777871498635758e-05, "loss": 0.0172, "step": 143330 }, { "grad_norm": 0.07703091204166412, "learning_rate": 5.777327048081217e-05, "loss": 0.0131, "step": 143340 }, { "grad_norm": 0.36687934398651123, "learning_rate": 5.7767825880814174e-05, "loss": 0.0191, "step": 143350 }, { "grad_norm": 0.5410085916519165, "learning_rate": 5.7762381186429745e-05, "loss": 0.0189, "step": 143360 }, { "grad_norm": 0.11932752281427383, "learning_rate": 5.775693639772507e-05, "loss": 0.0236, "step": 143370 }, { "grad_norm": 0.21568121016025543, "learning_rate": 5.775149151476629e-05, "loss": 0.0163, "step": 143380 }, { "grad_norm": 0.14342330396175385, "learning_rate": 5.774604653761957e-05, "loss": 0.0151, "step": 143390 }, { "grad_norm": 0.17019784450531006, "learning_rate": 5.774060146635106e-05, "loss": 0.0163, "step": 143400 }, { "grad_norm": 0.17907576262950897, "learning_rate": 5.7735156301026946e-05, "loss": 0.0226, "step": 143410 }, { "grad_norm": 0.11593668162822723, "learning_rate": 5.772971104171337e-05, "loss": 0.0174, "step": 143420 }, { "grad_norm": 0.106186643242836, "learning_rate": 5.772426568847653e-05, "loss": 0.015, "step": 143430 }, { "grad_norm": 0.2877826690673828, "learning_rate": 5.771882024138255e-05, "loss": 0.0152, "step": 143440 }, { "grad_norm": 0.09471463412046432, "learning_rate": 5.771337470049763e-05, "loss": 0.0122, "step": 143450 }, { "grad_norm": 0.14408254623413086, "learning_rate": 5.770792906588791e-05, "loss": 0.0157, "step": 143460 }, { "grad_norm": 0.4187453091144562, "learning_rate": 5.770248333761957e-05, "loss": 0.0091, "step": 143470 }, { "grad_norm": 0.1021246463060379, "learning_rate": 5.76970375157588e-05, "loss": 0.0191, "step": 143480 }, { "grad_norm": 0.13877682387828827, "learning_rate": 5.769159160037174e-05, "loss": 0.0172, "step": 143490 }, { "grad_norm": 0.16393336653709412, "learning_rate": 5.76861455915246e-05, "loss": 0.0135, "step": 143500 }, { "grad_norm": 0.19385404884815216, "learning_rate": 5.768069948928353e-05, "loss": 0.0176, "step": 143510 }, { "grad_norm": 0.22086642682552338, "learning_rate": 5.76752532937147e-05, "loss": 0.0177, "step": 143520 }, { "grad_norm": 0.32510486245155334, "learning_rate": 5.7669807004884304e-05, "loss": 0.0141, "step": 143530 }, { "grad_norm": 0.1799783706665039, "learning_rate": 5.76643606228585e-05, "loss": 0.0167, "step": 143540 }, { "grad_norm": 0.7554408311843872, "learning_rate": 5.7658914147703494e-05, "loss": 0.018, "step": 143550 }, { "grad_norm": 0.1133299171924591, "learning_rate": 5.765346757948543e-05, "loss": 0.0152, "step": 143560 }, { "grad_norm": 0.06755711883306503, "learning_rate": 5.764802091827053e-05, "loss": 0.0122, "step": 143570 }, { "grad_norm": 0.08878815919160843, "learning_rate": 5.764257416412494e-05, "loss": 0.013, "step": 143580 }, { "grad_norm": 0.13910506665706635, "learning_rate": 5.763712731711487e-05, "loss": 0.0152, "step": 143590 }, { "grad_norm": 0.5473304390907288, "learning_rate": 5.763168037730649e-05, "loss": 0.0145, "step": 143600 }, { "grad_norm": 0.12468405812978745, "learning_rate": 5.762623334476598e-05, "loss": 0.0171, "step": 143610 }, { "grad_norm": 0.06514015793800354, "learning_rate": 5.762078621955954e-05, "loss": 0.0121, "step": 143620 }, { "grad_norm": 0.10667099803686142, "learning_rate": 5.761533900175334e-05, "loss": 0.0148, "step": 143630 }, { "grad_norm": 0.11248353123664856, "learning_rate": 5.7609891691413595e-05, "loss": 0.01, "step": 143640 }, { "grad_norm": 0.09022518992424011, "learning_rate": 5.7604444288606474e-05, "loss": 0.0168, "step": 143650 }, { "grad_norm": 0.10126543045043945, "learning_rate": 5.759899679339819e-05, "loss": 0.0153, "step": 143660 }, { "grad_norm": 0.12083440274000168, "learning_rate": 5.7593549205854916e-05, "loss": 0.013, "step": 143670 }, { "grad_norm": 0.12449592351913452, "learning_rate": 5.758810152604284e-05, "loss": 0.0113, "step": 143680 }, { "grad_norm": 0.2245994210243225, "learning_rate": 5.758265375402817e-05, "loss": 0.0156, "step": 143690 }, { "grad_norm": 0.15560336410999298, "learning_rate": 5.757720588987708e-05, "loss": 0.0129, "step": 143700 }, { "grad_norm": 0.07842840254306793, "learning_rate": 5.7571757933655814e-05, "loss": 0.0149, "step": 143710 }, { "grad_norm": 0.09897900372743607, "learning_rate": 5.7566309885430526e-05, "loss": 0.0179, "step": 143720 }, { "grad_norm": 0.10660909861326218, "learning_rate": 5.756086174526743e-05, "loss": 0.0185, "step": 143730 }, { "grad_norm": 0.2397005558013916, "learning_rate": 5.755541351323272e-05, "loss": 0.0153, "step": 143740 }, { "grad_norm": 0.13589414954185486, "learning_rate": 5.754996518939262e-05, "loss": 0.0225, "step": 143750 }, { "grad_norm": 0.8005096912384033, "learning_rate": 5.754451677381329e-05, "loss": 0.0239, "step": 143760 }, { "grad_norm": 0.12544776499271393, "learning_rate": 5.753906826656097e-05, "loss": 0.0121, "step": 143770 }, { "grad_norm": 0.1017826497554779, "learning_rate": 5.753361966770187e-05, "loss": 0.0167, "step": 143780 }, { "grad_norm": 0.10920637845993042, "learning_rate": 5.752817097730214e-05, "loss": 0.0124, "step": 143790 }, { "grad_norm": 0.18158355355262756, "learning_rate": 5.7522722195428056e-05, "loss": 0.0147, "step": 143800 }, { "grad_norm": 0.13085927069187164, "learning_rate": 5.7517273322145785e-05, "loss": 0.0215, "step": 143810 }, { "grad_norm": 0.10953626036643982, "learning_rate": 5.751182435752155e-05, "loss": 0.0175, "step": 143820 }, { "grad_norm": 0.09129065275192261, "learning_rate": 5.750637530162155e-05, "loss": 0.0156, "step": 143830 }, { "grad_norm": 0.16519521176815033, "learning_rate": 5.7500926154512004e-05, "loss": 0.0189, "step": 143840 }, { "grad_norm": 0.5262969136238098, "learning_rate": 5.749547691625913e-05, "loss": 0.0145, "step": 143850 }, { "grad_norm": 0.10419507324695587, "learning_rate": 5.749002758692912e-05, "loss": 0.0162, "step": 143860 }, { "grad_norm": 0.07255261391401291, "learning_rate": 5.748457816658821e-05, "loss": 0.0124, "step": 143870 }, { "grad_norm": 0.0884169414639473, "learning_rate": 5.747912865530261e-05, "loss": 0.0141, "step": 143880 }, { "grad_norm": 0.1512785404920578, "learning_rate": 5.747367905313854e-05, "loss": 0.0115, "step": 143890 }, { "grad_norm": 0.20215897262096405, "learning_rate": 5.746822936016221e-05, "loss": 0.0185, "step": 143900 }, { "grad_norm": 0.27272462844848633, "learning_rate": 5.746277957643984e-05, "loss": 0.0158, "step": 143910 }, { "grad_norm": 0.10733620822429657, "learning_rate": 5.7457329702037646e-05, "loss": 0.0134, "step": 143920 }, { "grad_norm": 0.13270820677280426, "learning_rate": 5.745187973702186e-05, "loss": 0.0167, "step": 143930 }, { "grad_norm": 0.13014201819896698, "learning_rate": 5.74464296814587e-05, "loss": 0.0142, "step": 143940 }, { "grad_norm": 0.0813365951180458, "learning_rate": 5.7440979535414395e-05, "loss": 0.0177, "step": 143950 }, { "grad_norm": 0.10765165090560913, "learning_rate": 5.743552929895515e-05, "loss": 0.0189, "step": 143960 }, { "grad_norm": 0.3274138271808624, "learning_rate": 5.7430078972147226e-05, "loss": 0.0185, "step": 143970 }, { "grad_norm": 0.13588452339172363, "learning_rate": 5.74246285550568e-05, "loss": 0.018, "step": 143980 }, { "grad_norm": 0.10177785158157349, "learning_rate": 5.741917804775014e-05, "loss": 0.0112, "step": 143990 }, { "grad_norm": 0.174253910779953, "learning_rate": 5.7413727450293465e-05, "loss": 0.0192, "step": 144000 }, { "grad_norm": 0.17910444736480713, "learning_rate": 5.7408276762753e-05, "loss": 0.0169, "step": 144010 }, { "grad_norm": 0.09117602556943893, "learning_rate": 5.740282598519497e-05, "loss": 0.0136, "step": 144020 }, { "grad_norm": 0.13589200377464294, "learning_rate": 5.739737511768564e-05, "loss": 0.0171, "step": 144030 }, { "grad_norm": 0.4131356477737427, "learning_rate": 5.739192416029119e-05, "loss": 0.0147, "step": 144040 }, { "grad_norm": 0.10758762061595917, "learning_rate": 5.73864731130779e-05, "loss": 0.014, "step": 144050 }, { "grad_norm": 0.11211299151182175, "learning_rate": 5.738102197611197e-05, "loss": 0.017, "step": 144060 }, { "grad_norm": 0.09021065384149551, "learning_rate": 5.7375570749459676e-05, "loss": 0.0215, "step": 144070 }, { "grad_norm": 0.17578235268592834, "learning_rate": 5.737011943318722e-05, "loss": 0.0218, "step": 144080 }, { "grad_norm": 0.1027735248208046, "learning_rate": 5.736466802736085e-05, "loss": 0.0162, "step": 144090 }, { "grad_norm": 0.16378037631511688, "learning_rate": 5.735921653204682e-05, "loss": 0.0189, "step": 144100 }, { "grad_norm": 0.1575966626405716, "learning_rate": 5.735376494731135e-05, "loss": 0.0112, "step": 144110 }, { "grad_norm": 0.13571713864803314, "learning_rate": 5.73483132732207e-05, "loss": 0.0141, "step": 144120 }, { "grad_norm": 0.11106165498495102, "learning_rate": 5.734286150984111e-05, "loss": 0.0124, "step": 144130 }, { "grad_norm": 0.24251693487167358, "learning_rate": 5.733740965723882e-05, "loss": 0.0226, "step": 144140 }, { "grad_norm": 0.12832289934158325, "learning_rate": 5.7331957715480065e-05, "loss": 0.0159, "step": 144150 }, { "grad_norm": 0.1083158329129219, "learning_rate": 5.7326505684631105e-05, "loss": 0.0139, "step": 144160 }, { "grad_norm": 0.06971824914216995, "learning_rate": 5.732105356475819e-05, "loss": 0.0162, "step": 144170 }, { "grad_norm": 0.16142801940441132, "learning_rate": 5.731560135592755e-05, "loss": 0.0202, "step": 144180 }, { "grad_norm": 0.10154884308576584, "learning_rate": 5.731014905820547e-05, "loss": 0.0124, "step": 144190 }, { "grad_norm": 0.15406866371631622, "learning_rate": 5.730469667165814e-05, "loss": 0.0156, "step": 144200 }, { "grad_norm": 0.06293362379074097, "learning_rate": 5.729924419635189e-05, "loss": 0.0153, "step": 144210 }, { "grad_norm": 0.08784329146146774, "learning_rate": 5.72937916323529e-05, "loss": 0.0176, "step": 144220 }, { "grad_norm": 0.170359805226326, "learning_rate": 5.728833897972747e-05, "loss": 0.016, "step": 144230 }, { "grad_norm": 0.1748175173997879, "learning_rate": 5.7282886238541834e-05, "loss": 0.0175, "step": 144240 }, { "grad_norm": 0.1550837606191635, "learning_rate": 5.727743340886226e-05, "loss": 0.0175, "step": 144250 }, { "grad_norm": 0.14152014255523682, "learning_rate": 5.7271980490754994e-05, "loss": 0.013, "step": 144260 }, { "grad_norm": 0.2019459307193756, "learning_rate": 5.7266527484286295e-05, "loss": 0.0169, "step": 144270 }, { "grad_norm": 0.09585203230381012, "learning_rate": 5.726107438952244e-05, "loss": 0.0148, "step": 144280 }, { "grad_norm": 0.2090161144733429, "learning_rate": 5.725562120652965e-05, "loss": 0.0106, "step": 144290 }, { "grad_norm": 0.07341761887073517, "learning_rate": 5.725016793537424e-05, "loss": 0.0162, "step": 144300 }, { "grad_norm": 0.11779031902551651, "learning_rate": 5.724471457612244e-05, "loss": 0.022, "step": 144310 }, { "grad_norm": 0.16798283159732819, "learning_rate": 5.723926112884051e-05, "loss": 0.0132, "step": 144320 }, { "grad_norm": 0.15194523334503174, "learning_rate": 5.7233807593594726e-05, "loss": 0.0156, "step": 144330 }, { "grad_norm": 0.10672808438539505, "learning_rate": 5.722835397045134e-05, "loss": 0.0155, "step": 144340 }, { "grad_norm": 0.0968901515007019, "learning_rate": 5.722290025947664e-05, "loss": 0.0206, "step": 144350 }, { "grad_norm": 0.06950227916240692, "learning_rate": 5.721744646073688e-05, "loss": 0.0147, "step": 144360 }, { "grad_norm": 0.12483581155538559, "learning_rate": 5.7211992574298325e-05, "loss": 0.012, "step": 144370 }, { "grad_norm": 0.2332284301519394, "learning_rate": 5.720653860022726e-05, "loss": 0.013, "step": 144380 }, { "grad_norm": 0.13294830918312073, "learning_rate": 5.7201084538589945e-05, "loss": 0.012, "step": 144390 }, { "grad_norm": 0.09643932431936264, "learning_rate": 5.719563038945265e-05, "loss": 0.0153, "step": 144400 }, { "grad_norm": 0.10660646110773087, "learning_rate": 5.719017615288165e-05, "loss": 0.0139, "step": 144410 }, { "grad_norm": 0.10841784626245499, "learning_rate": 5.718472182894322e-05, "loss": 0.0107, "step": 144420 }, { "grad_norm": 0.16375787556171417, "learning_rate": 5.717926741770364e-05, "loss": 0.0114, "step": 144430 }, { "grad_norm": 0.11556445807218552, "learning_rate": 5.717381291922919e-05, "loss": 0.0166, "step": 144440 }, { "grad_norm": 0.07554787397384644, "learning_rate": 5.716835833358612e-05, "loss": 0.0104, "step": 144450 }, { "grad_norm": 0.0901021882891655, "learning_rate": 5.716290366084075e-05, "loss": 0.0131, "step": 144460 }, { "grad_norm": 0.07952728867530823, "learning_rate": 5.715744890105933e-05, "loss": 0.0113, "step": 144470 }, { "grad_norm": 0.08275801688432693, "learning_rate": 5.715199405430814e-05, "loss": 0.0167, "step": 144480 }, { "grad_norm": 0.08133421838283539, "learning_rate": 5.7146539120653486e-05, "loss": 0.0111, "step": 144490 }, { "grad_norm": 0.2044472098350525, "learning_rate": 5.7141084100161616e-05, "loss": 0.0141, "step": 144500 }, { "grad_norm": 0.11944638192653656, "learning_rate": 5.713562899289885e-05, "loss": 0.0167, "step": 144510 }, { "grad_norm": 0.2814723551273346, "learning_rate": 5.713017379893144e-05, "loss": 0.016, "step": 144520 }, { "grad_norm": 0.14525404572486877, "learning_rate": 5.712471851832569e-05, "loss": 0.0141, "step": 144530 }, { "grad_norm": 0.14589329063892365, "learning_rate": 5.7119263151147885e-05, "loss": 0.0233, "step": 144540 }, { "grad_norm": 0.10219953209161758, "learning_rate": 5.711380769746432e-05, "loss": 0.0111, "step": 144550 }, { "grad_norm": 0.11428643018007278, "learning_rate": 5.710835215734126e-05, "loss": 0.0118, "step": 144560 }, { "grad_norm": 0.0772143080830574, "learning_rate": 5.7102896530845015e-05, "loss": 0.0157, "step": 144570 }, { "grad_norm": 0.09006648510694504, "learning_rate": 5.709744081804187e-05, "loss": 0.0146, "step": 144580 }, { "grad_norm": 0.11318720877170563, "learning_rate": 5.7091985018998116e-05, "loss": 0.0189, "step": 144590 }, { "grad_norm": 0.13594724237918854, "learning_rate": 5.7086529133780054e-05, "loss": 0.0153, "step": 144600 }, { "grad_norm": 0.10021165758371353, "learning_rate": 5.708107316245397e-05, "loss": 0.0121, "step": 144610 }, { "grad_norm": 0.1425345242023468, "learning_rate": 5.7075617105086166e-05, "loss": 0.0138, "step": 144620 }, { "grad_norm": 0.1570414900779724, "learning_rate": 5.707016096174291e-05, "loss": 0.0093, "step": 144630 }, { "grad_norm": 0.15451663732528687, "learning_rate": 5.706470473249056e-05, "loss": 0.0178, "step": 144640 }, { "grad_norm": 0.1622103452682495, "learning_rate": 5.705924841739535e-05, "loss": 0.0176, "step": 144650 }, { "grad_norm": 0.17282742261886597, "learning_rate": 5.7053792016523625e-05, "loss": 0.016, "step": 144660 }, { "grad_norm": 0.16211426258087158, "learning_rate": 5.704833552994164e-05, "loss": 0.0157, "step": 144670 }, { "grad_norm": 0.07223177701234818, "learning_rate": 5.704287895771575e-05, "loss": 0.0156, "step": 144680 }, { "grad_norm": 0.13063305616378784, "learning_rate": 5.7037422299912226e-05, "loss": 0.0156, "step": 144690 }, { "grad_norm": 0.1956297606229782, "learning_rate": 5.703196555659737e-05, "loss": 0.0184, "step": 144700 }, { "grad_norm": 0.15449315309524536, "learning_rate": 5.702650872783749e-05, "loss": 0.0149, "step": 144710 }, { "grad_norm": 0.09047644585371017, "learning_rate": 5.702105181369891e-05, "loss": 0.0185, "step": 144720 }, { "grad_norm": 0.30743905901908875, "learning_rate": 5.701559481424791e-05, "loss": 0.0145, "step": 144730 }, { "grad_norm": 0.1255512535572052, "learning_rate": 5.701013772955082e-05, "loss": 0.0178, "step": 144740 }, { "grad_norm": 0.15553846955299377, "learning_rate": 5.7004680559673915e-05, "loss": 0.0174, "step": 144750 }, { "grad_norm": 0.16382354497909546, "learning_rate": 5.699922330468355e-05, "loss": 0.0148, "step": 144760 }, { "grad_norm": 0.10881976038217545, "learning_rate": 5.6993765964645993e-05, "loss": 0.0136, "step": 144770 }, { "grad_norm": 0.08227723836898804, "learning_rate": 5.69883085396276e-05, "loss": 0.0138, "step": 144780 }, { "grad_norm": 0.11401733756065369, "learning_rate": 5.6982851029694636e-05, "loss": 0.0158, "step": 144790 }, { "grad_norm": 0.10867775976657867, "learning_rate": 5.697739343491346e-05, "loss": 0.0152, "step": 144800 }, { "grad_norm": 0.14493024349212646, "learning_rate": 5.697193575535037e-05, "loss": 0.0111, "step": 144810 }, { "grad_norm": 0.1322643756866455, "learning_rate": 5.696647799107165e-05, "loss": 0.0171, "step": 144820 }, { "grad_norm": 0.1074151024222374, "learning_rate": 5.6961020142143676e-05, "loss": 0.0193, "step": 144830 }, { "grad_norm": 0.07817241549491882, "learning_rate": 5.695556220863271e-05, "loss": 0.0175, "step": 144840 }, { "grad_norm": 0.10715031623840332, "learning_rate": 5.695010419060511e-05, "loss": 0.0157, "step": 144850 }, { "grad_norm": 0.1248365268111229, "learning_rate": 5.6944646088127174e-05, "loss": 0.0116, "step": 144860 }, { "grad_norm": 0.11581148207187653, "learning_rate": 5.6939187901265244e-05, "loss": 0.0148, "step": 144870 }, { "grad_norm": 0.10697582364082336, "learning_rate": 5.6933729630085617e-05, "loss": 0.0123, "step": 144880 }, { "grad_norm": 0.11028406023979187, "learning_rate": 5.692827127465463e-05, "loss": 0.0165, "step": 144890 }, { "grad_norm": 0.09910901635885239, "learning_rate": 5.692281283503862e-05, "loss": 0.0144, "step": 144900 }, { "grad_norm": 0.18715444207191467, "learning_rate": 5.691735431130387e-05, "loss": 0.0152, "step": 144910 }, { "grad_norm": 0.0877513587474823, "learning_rate": 5.691189570351676e-05, "loss": 0.0177, "step": 144920 }, { "grad_norm": 0.10558311641216278, "learning_rate": 5.690643701174357e-05, "loss": 0.0178, "step": 144930 }, { "grad_norm": 0.12945054471492767, "learning_rate": 5.690097823605067e-05, "loss": 0.0163, "step": 144940 }, { "grad_norm": 0.29367953538894653, "learning_rate": 5.689551937650436e-05, "loss": 0.0177, "step": 144950 }, { "grad_norm": 0.0771709531545639, "learning_rate": 5.689006043317098e-05, "loss": 0.0154, "step": 144960 }, { "grad_norm": 0.09951929748058319, "learning_rate": 5.688460140611687e-05, "loss": 0.0147, "step": 144970 }, { "grad_norm": 0.11746915429830551, "learning_rate": 5.687914229540833e-05, "loss": 0.0124, "step": 144980 }, { "grad_norm": 0.16623525321483612, "learning_rate": 5.687368310111173e-05, "loss": 0.0157, "step": 144990 }, { "grad_norm": 0.10777555406093597, "learning_rate": 5.686822382329339e-05, "loss": 0.0184, "step": 145000 }, { "grad_norm": 0.07890982925891876, "learning_rate": 5.6862764462019646e-05, "loss": 0.0123, "step": 145010 }, { "grad_norm": 0.20153990387916565, "learning_rate": 5.685730501735682e-05, "loss": 0.0181, "step": 145020 }, { "grad_norm": 0.14894874393939972, "learning_rate": 5.685184548937129e-05, "loss": 0.0171, "step": 145030 }, { "grad_norm": 0.1870415359735489, "learning_rate": 5.684638587812936e-05, "loss": 0.0167, "step": 145040 }, { "grad_norm": 0.22221596539020538, "learning_rate": 5.684092618369736e-05, "loss": 0.0186, "step": 145050 }, { "grad_norm": 0.10006318986415863, "learning_rate": 5.683546640614167e-05, "loss": 0.0123, "step": 145060 }, { "grad_norm": 0.08696682751178741, "learning_rate": 5.6830006545528594e-05, "loss": 0.0124, "step": 145070 }, { "grad_norm": 0.11816420406103134, "learning_rate": 5.68245466019245e-05, "loss": 0.0152, "step": 145080 }, { "grad_norm": 0.12153097242116928, "learning_rate": 5.681908657539571e-05, "loss": 0.0131, "step": 145090 }, { "grad_norm": 0.12381460517644882, "learning_rate": 5.6813626466008585e-05, "loss": 0.0165, "step": 145100 }, { "grad_norm": 0.18723617494106293, "learning_rate": 5.6808166273829465e-05, "loss": 0.0224, "step": 145110 }, { "grad_norm": 0.1488027125597, "learning_rate": 5.68027059989247e-05, "loss": 0.0106, "step": 145120 }, { "grad_norm": 0.09459484368562698, "learning_rate": 5.6797245641360644e-05, "loss": 0.019, "step": 145130 }, { "grad_norm": 0.31879398226737976, "learning_rate": 5.6791785201203616e-05, "loss": 0.0143, "step": 145140 }, { "grad_norm": 0.09055708348751068, "learning_rate": 5.678632467852e-05, "loss": 0.0113, "step": 145150 }, { "grad_norm": 0.05663405358791351, "learning_rate": 5.678086407337612e-05, "loss": 0.0111, "step": 145160 }, { "grad_norm": 0.1705675721168518, "learning_rate": 5.6775403385838354e-05, "loss": 0.0161, "step": 145170 }, { "grad_norm": 0.0665634348988533, "learning_rate": 5.676994261597301e-05, "loss": 0.0132, "step": 145180 }, { "grad_norm": 0.07607413083314896, "learning_rate": 5.676448176384651e-05, "loss": 0.0109, "step": 145190 }, { "grad_norm": 0.1335114985704422, "learning_rate": 5.6759020829525154e-05, "loss": 0.0131, "step": 145200 }, { "grad_norm": 0.14106370508670807, "learning_rate": 5.675355981307531e-05, "loss": 0.0144, "step": 145210 }, { "grad_norm": 0.15022879838943481, "learning_rate": 5.6748098714563345e-05, "loss": 0.0156, "step": 145220 }, { "grad_norm": 0.15910549461841583, "learning_rate": 5.67426375340556e-05, "loss": 0.0129, "step": 145230 }, { "grad_norm": 0.14802095293998718, "learning_rate": 5.673717627161844e-05, "loss": 0.0137, "step": 145240 }, { "grad_norm": 0.08570988476276398, "learning_rate": 5.6731714927318236e-05, "loss": 0.0111, "step": 145250 }, { "grad_norm": 0.18156132102012634, "learning_rate": 5.6726253501221336e-05, "loss": 0.0104, "step": 145260 }, { "grad_norm": 0.08024875074625015, "learning_rate": 5.6720791993394104e-05, "loss": 0.0154, "step": 145270 }, { "grad_norm": 0.07788140326738358, "learning_rate": 5.671533040390291e-05, "loss": 0.0144, "step": 145280 }, { "grad_norm": 0.10240763425827026, "learning_rate": 5.670986873281412e-05, "loss": 0.0139, "step": 145290 }, { "grad_norm": 0.07487069070339203, "learning_rate": 5.670440698019407e-05, "loss": 0.019, "step": 145300 }, { "grad_norm": 0.26746511459350586, "learning_rate": 5.669894514610916e-05, "loss": 0.0199, "step": 145310 }, { "grad_norm": 0.10651995986700058, "learning_rate": 5.669348323062572e-05, "loss": 0.0194, "step": 145320 }, { "grad_norm": 0.24568410217761993, "learning_rate": 5.6688021233810164e-05, "loss": 0.0177, "step": 145330 }, { "grad_norm": 0.09093958884477615, "learning_rate": 5.668255915572882e-05, "loss": 0.0195, "step": 145340 }, { "grad_norm": 0.13394837081432343, "learning_rate": 5.667709699644808e-05, "loss": 0.0212, "step": 145350 }, { "grad_norm": 0.10069207847118378, "learning_rate": 5.6671634756034295e-05, "loss": 0.015, "step": 145360 }, { "grad_norm": 0.12036138027906418, "learning_rate": 5.666617243455387e-05, "loss": 0.0107, "step": 145370 }, { "grad_norm": 0.07985486090183258, "learning_rate": 5.666071003207314e-05, "loss": 0.0174, "step": 145380 }, { "grad_norm": 0.07063955813646317, "learning_rate": 5.66552475486585e-05, "loss": 0.0137, "step": 145390 }, { "grad_norm": 0.13063712418079376, "learning_rate": 5.664978498437632e-05, "loss": 0.0125, "step": 145400 }, { "grad_norm": 0.20720192790031433, "learning_rate": 5.6644322339292975e-05, "loss": 0.0131, "step": 145410 }, { "grad_norm": 0.08571026474237442, "learning_rate": 5.6638859613474825e-05, "loss": 0.0175, "step": 145420 }, { "grad_norm": 0.1399400234222412, "learning_rate": 5.663339680698828e-05, "loss": 0.0119, "step": 145430 }, { "grad_norm": 0.11725308001041412, "learning_rate": 5.662793391989969e-05, "loss": 0.0193, "step": 145440 }, { "grad_norm": 0.19926977157592773, "learning_rate": 5.662247095227546e-05, "loss": 0.014, "step": 145450 }, { "grad_norm": 0.15282417833805084, "learning_rate": 5.661700790418194e-05, "loss": 0.0152, "step": 145460 }, { "grad_norm": 0.11782414466142654, "learning_rate": 5.661154477568553e-05, "loss": 0.0139, "step": 145470 }, { "grad_norm": 0.0840103030204773, "learning_rate": 5.660608156685261e-05, "loss": 0.0132, "step": 145480 }, { "grad_norm": 0.20197974145412445, "learning_rate": 5.6600618277749575e-05, "loss": 0.0138, "step": 145490 }, { "grad_norm": 0.0840734913945198, "learning_rate": 5.659515490844277e-05, "loss": 0.0102, "step": 145500 }, { "grad_norm": 0.18349799513816833, "learning_rate": 5.658969145899863e-05, "loss": 0.0209, "step": 145510 }, { "grad_norm": 0.11136617511510849, "learning_rate": 5.6584227929483514e-05, "loss": 0.0126, "step": 145520 }, { "grad_norm": 0.08696899563074112, "learning_rate": 5.65787643199638e-05, "loss": 0.0218, "step": 145530 }, { "grad_norm": 0.08728407323360443, "learning_rate": 5.657330063050589e-05, "loss": 0.0108, "step": 145540 }, { "grad_norm": 0.1219850704073906, "learning_rate": 5.656783686117617e-05, "loss": 0.0162, "step": 145550 }, { "grad_norm": 0.0757894366979599, "learning_rate": 5.656237301204104e-05, "loss": 0.0132, "step": 145560 }, { "grad_norm": 0.06438016146421432, "learning_rate": 5.655690908316688e-05, "loss": 0.0142, "step": 145570 }, { "grad_norm": 0.09373068064451218, "learning_rate": 5.655144507462008e-05, "loss": 0.0116, "step": 145580 }, { "grad_norm": 0.197641059756279, "learning_rate": 5.654598098646703e-05, "loss": 0.0178, "step": 145590 }, { "grad_norm": 0.11535849422216415, "learning_rate": 5.654051681877415e-05, "loss": 0.0151, "step": 145600 }, { "grad_norm": 0.07817617803812027, "learning_rate": 5.6535052571607806e-05, "loss": 0.0121, "step": 145610 }, { "grad_norm": 0.10363923758268356, "learning_rate": 5.6529588245034396e-05, "loss": 0.0156, "step": 145620 }, { "grad_norm": 0.09299921989440918, "learning_rate": 5.652412383912034e-05, "loss": 0.0186, "step": 145630 }, { "grad_norm": 0.10158301144838333, "learning_rate": 5.651865935393199e-05, "loss": 0.0153, "step": 145640 }, { "grad_norm": 0.13562600314617157, "learning_rate": 5.6513194789535805e-05, "loss": 0.0184, "step": 145650 }, { "grad_norm": 0.10231997072696686, "learning_rate": 5.6507730145998126e-05, "loss": 0.0165, "step": 145660 }, { "grad_norm": 0.10233672708272934, "learning_rate": 5.65022654233854e-05, "loss": 0.0183, "step": 145670 }, { "grad_norm": 0.07665242999792099, "learning_rate": 5.649680062176401e-05, "loss": 0.0154, "step": 145680 }, { "grad_norm": 0.1494198590517044, "learning_rate": 5.6491335741200356e-05, "loss": 0.0182, "step": 145690 }, { "grad_norm": 0.08387499302625656, "learning_rate": 5.648587078176085e-05, "loss": 0.0126, "step": 145700 }, { "grad_norm": 0.08606582880020142, "learning_rate": 5.6480405743511876e-05, "loss": 0.0155, "step": 145710 }, { "grad_norm": 0.13237953186035156, "learning_rate": 5.6474940626519866e-05, "loss": 0.0153, "step": 145720 }, { "grad_norm": 0.1641160547733307, "learning_rate": 5.646947543085121e-05, "loss": 0.0149, "step": 145730 }, { "grad_norm": 0.06954699754714966, "learning_rate": 5.646401015657232e-05, "loss": 0.0122, "step": 145740 }, { "grad_norm": 0.128126323223114, "learning_rate": 5.645854480374959e-05, "loss": 0.0135, "step": 145750 }, { "grad_norm": 0.1463698446750641, "learning_rate": 5.6453079372449466e-05, "loss": 0.0147, "step": 145760 }, { "grad_norm": 0.16442342102527618, "learning_rate": 5.644761386273833e-05, "loss": 0.023, "step": 145770 }, { "grad_norm": 0.198249951004982, "learning_rate": 5.6442148274682584e-05, "loss": 0.0181, "step": 145780 }, { "grad_norm": 0.07855353504419327, "learning_rate": 5.643668260834867e-05, "loss": 0.0119, "step": 145790 }, { "grad_norm": 0.08781567215919495, "learning_rate": 5.643121686380297e-05, "loss": 0.0134, "step": 145800 }, { "grad_norm": 0.10150237381458282, "learning_rate": 5.642575104111192e-05, "loss": 0.0142, "step": 145810 }, { "grad_norm": 0.12651026248931885, "learning_rate": 5.642028514034192e-05, "loss": 0.0137, "step": 145820 }, { "grad_norm": 0.13518203794956207, "learning_rate": 5.641481916155942e-05, "loss": 0.0121, "step": 145830 }, { "grad_norm": 0.11676941066980362, "learning_rate": 5.640935310483078e-05, "loss": 0.0149, "step": 145840 }, { "grad_norm": 0.3501816987991333, "learning_rate": 5.6403886970222463e-05, "loss": 0.0138, "step": 145850 }, { "grad_norm": 0.06734307110309601, "learning_rate": 5.639842075780087e-05, "loss": 0.0157, "step": 145860 }, { "grad_norm": 0.09126701205968857, "learning_rate": 5.639295446763242e-05, "loss": 0.0175, "step": 145870 }, { "grad_norm": 0.1539909690618515, "learning_rate": 5.6387488099783546e-05, "loss": 0.0161, "step": 145880 }, { "grad_norm": 0.1399683952331543, "learning_rate": 5.638202165432066e-05, "loss": 0.0115, "step": 145890 }, { "grad_norm": 0.2968146800994873, "learning_rate": 5.637655513131018e-05, "loss": 0.0178, "step": 145900 }, { "grad_norm": 0.09891919046640396, "learning_rate": 5.637108853081854e-05, "loss": 0.0116, "step": 145910 }, { "grad_norm": 0.18892624974250793, "learning_rate": 5.6365621852912154e-05, "loss": 0.0184, "step": 145920 }, { "grad_norm": 0.14498236775398254, "learning_rate": 5.6360155097657464e-05, "loss": 0.0183, "step": 145930 }, { "grad_norm": 0.0875418558716774, "learning_rate": 5.635468826512087e-05, "loss": 0.0223, "step": 145940 }, { "grad_norm": 0.07885926216840744, "learning_rate": 5.634922135536882e-05, "loss": 0.0129, "step": 145950 }, { "grad_norm": 0.17740586400032043, "learning_rate": 5.634375436846773e-05, "loss": 0.015, "step": 145960 }, { "grad_norm": 0.22041574120521545, "learning_rate": 5.633828730448404e-05, "loss": 0.0159, "step": 145970 }, { "grad_norm": 0.10399430990219116, "learning_rate": 5.633282016348417e-05, "loss": 0.0156, "step": 145980 }, { "grad_norm": 0.07672068476676941, "learning_rate": 5.632735294553456e-05, "loss": 0.0132, "step": 145990 }, { "grad_norm": 0.10674820840358734, "learning_rate": 5.632188565070163e-05, "loss": 0.011, "step": 146000 }, { "grad_norm": 0.1063840389251709, "learning_rate": 5.6316418279051844e-05, "loss": 0.0144, "step": 146010 }, { "grad_norm": 0.09250543266534805, "learning_rate": 5.63109508306516e-05, "loss": 0.0115, "step": 146020 }, { "grad_norm": 0.16450393199920654, "learning_rate": 5.630548330556733e-05, "loss": 0.0232, "step": 146030 }, { "grad_norm": 0.5619001984596252, "learning_rate": 5.6300015703865494e-05, "loss": 0.0124, "step": 146040 }, { "grad_norm": 0.11138387024402618, "learning_rate": 5.6294548025612514e-05, "loss": 0.0199, "step": 146050 }, { "grad_norm": 0.15736648440361023, "learning_rate": 5.628908027087485e-05, "loss": 0.0135, "step": 146060 }, { "grad_norm": 0.1673724204301834, "learning_rate": 5.62836124397189e-05, "loss": 0.0144, "step": 146070 }, { "grad_norm": 0.3733821511268616, "learning_rate": 5.627814453221115e-05, "loss": 0.0129, "step": 146080 }, { "grad_norm": 0.07639773935079575, "learning_rate": 5.6272676548418e-05, "loss": 0.0148, "step": 146090 }, { "grad_norm": 0.06951843947172165, "learning_rate": 5.6267208488405924e-05, "loss": 0.017, "step": 146100 }, { "grad_norm": 0.09237265586853027, "learning_rate": 5.626174035224132e-05, "loss": 0.0243, "step": 146110 }, { "grad_norm": 0.17253315448760986, "learning_rate": 5.6256272139990676e-05, "loss": 0.0153, "step": 146120 }, { "grad_norm": 0.17453642189502716, "learning_rate": 5.6250803851720416e-05, "loss": 0.0197, "step": 146130 }, { "grad_norm": 0.10388466715812683, "learning_rate": 5.624533548749699e-05, "loss": 0.0207, "step": 146140 }, { "grad_norm": 0.2017974704504013, "learning_rate": 5.623986704738684e-05, "loss": 0.012, "step": 146150 }, { "grad_norm": 0.1539406031370163, "learning_rate": 5.623439853145641e-05, "loss": 0.02, "step": 146160 }, { "grad_norm": 0.12974408268928528, "learning_rate": 5.622892993977216e-05, "loss": 0.0133, "step": 146170 }, { "grad_norm": 0.16327166557312012, "learning_rate": 5.6223461272400525e-05, "loss": 0.0187, "step": 146180 }, { "grad_norm": 0.08843359351158142, "learning_rate": 5.6217992529407946e-05, "loss": 0.0176, "step": 146190 }, { "grad_norm": 0.31936633586883545, "learning_rate": 5.621252371086091e-05, "loss": 0.0235, "step": 146200 }, { "grad_norm": 0.12690983712673187, "learning_rate": 5.6207054816825835e-05, "loss": 0.0152, "step": 146210 }, { "grad_norm": 0.28783103823661804, "learning_rate": 5.6201585847369185e-05, "loss": 0.0185, "step": 146220 }, { "grad_norm": 0.11338311433792114, "learning_rate": 5.619611680255741e-05, "loss": 0.0189, "step": 146230 }, { "grad_norm": 0.15205812454223633, "learning_rate": 5.6190647682456965e-05, "loss": 0.0167, "step": 146240 }, { "grad_norm": 0.11245696246623993, "learning_rate": 5.618517848713432e-05, "loss": 0.0169, "step": 146250 }, { "grad_norm": 0.16438733041286469, "learning_rate": 5.6179709216655895e-05, "loss": 0.0105, "step": 146260 }, { "grad_norm": 0.13569901883602142, "learning_rate": 5.617423987108818e-05, "loss": 0.0129, "step": 146270 }, { "grad_norm": 0.09835081547498703, "learning_rate": 5.616877045049762e-05, "loss": 0.0138, "step": 146280 }, { "grad_norm": 0.10456738620996475, "learning_rate": 5.616330095495067e-05, "loss": 0.0173, "step": 146290 }, { "grad_norm": 0.13510702550411224, "learning_rate": 5.6157831384513796e-05, "loss": 0.0141, "step": 146300 }, { "grad_norm": 0.2143249213695526, "learning_rate": 5.615236173925347e-05, "loss": 0.0174, "step": 146310 }, { "grad_norm": 0.09962882846593857, "learning_rate": 5.614689201923612e-05, "loss": 0.014, "step": 146320 }, { "grad_norm": 0.3669474720954895, "learning_rate": 5.614142222452824e-05, "loss": 0.0191, "step": 146330 }, { "grad_norm": 0.1145888939499855, "learning_rate": 5.613595235519629e-05, "loss": 0.0133, "step": 146340 }, { "grad_norm": 0.10042132437229156, "learning_rate": 5.613048241130671e-05, "loss": 0.0191, "step": 146350 }, { "grad_norm": 0.09577550739049911, "learning_rate": 5.612501239292598e-05, "loss": 0.0194, "step": 146360 }, { "grad_norm": 0.0784691870212555, "learning_rate": 5.611954230012056e-05, "loss": 0.0154, "step": 146370 }, { "grad_norm": 0.07020308077335358, "learning_rate": 5.611407213295694e-05, "loss": 0.0152, "step": 146380 }, { "grad_norm": 0.09529642015695572, "learning_rate": 5.610860189150156e-05, "loss": 0.014, "step": 146390 }, { "grad_norm": 0.18413496017456055, "learning_rate": 5.610313157582091e-05, "loss": 0.0206, "step": 146400 }, { "grad_norm": 0.13352236151695251, "learning_rate": 5.6097661185981455e-05, "loss": 0.012, "step": 146410 }, { "grad_norm": 0.08302400261163712, "learning_rate": 5.609219072204964e-05, "loss": 0.0153, "step": 146420 }, { "grad_norm": 0.10069038718938828, "learning_rate": 5.6086720184091967e-05, "loss": 0.0138, "step": 146430 }, { "grad_norm": 0.11259807646274567, "learning_rate": 5.6081249572174884e-05, "loss": 0.0116, "step": 146440 }, { "grad_norm": 0.12690606713294983, "learning_rate": 5.6075778886364904e-05, "loss": 0.0118, "step": 146450 }, { "grad_norm": 0.1630120724439621, "learning_rate": 5.607030812672844e-05, "loss": 0.0169, "step": 146460 }, { "grad_norm": 0.1779734194278717, "learning_rate": 5.606483729333203e-05, "loss": 0.014, "step": 146470 }, { "grad_norm": 0.1579919010400772, "learning_rate": 5.6059366386242105e-05, "loss": 0.0209, "step": 146480 }, { "grad_norm": 0.0987214520573616, "learning_rate": 5.605389540552517e-05, "loss": 0.0177, "step": 146490 }, { "grad_norm": 0.1930692344903946, "learning_rate": 5.6048424351247686e-05, "loss": 0.0152, "step": 146500 }, { "grad_norm": 0.07616490125656128, "learning_rate": 5.604295322347613e-05, "loss": 0.0134, "step": 146510 }, { "grad_norm": 0.10911223292350769, "learning_rate": 5.603748202227699e-05, "loss": 0.0121, "step": 146520 }, { "grad_norm": 0.13169077038764954, "learning_rate": 5.603201074771674e-05, "loss": 0.0177, "step": 146530 }, { "grad_norm": 0.14105159044265747, "learning_rate": 5.6026539399861855e-05, "loss": 0.0135, "step": 146540 }, { "grad_norm": 0.13236606121063232, "learning_rate": 5.602106797877883e-05, "loss": 0.0156, "step": 146550 }, { "grad_norm": 0.10664016753435135, "learning_rate": 5.601559648453416e-05, "loss": 0.0138, "step": 146560 }, { "grad_norm": 0.17253488302230835, "learning_rate": 5.6010124917194306e-05, "loss": 0.0141, "step": 146570 }, { "grad_norm": 0.11683094501495361, "learning_rate": 5.600465327682576e-05, "loss": 0.0132, "step": 146580 }, { "grad_norm": 0.20266090333461761, "learning_rate": 5.599918156349501e-05, "loss": 0.0131, "step": 146590 }, { "grad_norm": 1.1732374429702759, "learning_rate": 5.599370977726852e-05, "loss": 0.0158, "step": 146600 }, { "grad_norm": 0.7731287479400635, "learning_rate": 5.5988237918212814e-05, "loss": 0.0149, "step": 146610 }, { "grad_norm": 0.06341296434402466, "learning_rate": 5.5982765986394356e-05, "loss": 0.0171, "step": 146620 }, { "grad_norm": 0.10021276026964188, "learning_rate": 5.5977293981879655e-05, "loss": 0.0151, "step": 146630 }, { "grad_norm": 0.12387906014919281, "learning_rate": 5.597182190473518e-05, "loss": 0.0253, "step": 146640 }, { "grad_norm": 0.09142766147851944, "learning_rate": 5.5966349755027436e-05, "loss": 0.01, "step": 146650 }, { "grad_norm": 0.08194611221551895, "learning_rate": 5.596087753282291e-05, "loss": 0.0119, "step": 146660 }, { "grad_norm": 0.08524205535650253, "learning_rate": 5.595540523818809e-05, "loss": 0.0155, "step": 146670 }, { "grad_norm": 0.39801865816116333, "learning_rate": 5.594993287118947e-05, "loss": 0.0208, "step": 146680 }, { "grad_norm": 0.20829012989997864, "learning_rate": 5.5944460431893545e-05, "loss": 0.013, "step": 146690 }, { "grad_norm": 0.1237136572599411, "learning_rate": 5.593898792036682e-05, "loss": 0.0147, "step": 146700 }, { "grad_norm": 0.11244098842144012, "learning_rate": 5.593351533667578e-05, "loss": 0.021, "step": 146710 }, { "grad_norm": 0.07770027965307236, "learning_rate": 5.592804268088694e-05, "loss": 0.0166, "step": 146720 }, { "grad_norm": 0.08116176724433899, "learning_rate": 5.5922569953066775e-05, "loss": 0.0132, "step": 146730 }, { "grad_norm": 0.06932453066110611, "learning_rate": 5.5917097153281785e-05, "loss": 0.0148, "step": 146740 }, { "grad_norm": 0.12713722884655, "learning_rate": 5.5911624281598496e-05, "loss": 0.0171, "step": 146750 }, { "grad_norm": 0.08048132807016373, "learning_rate": 5.5906151338083365e-05, "loss": 0.0141, "step": 146760 }, { "grad_norm": 0.14433170855045319, "learning_rate": 5.5900678322802946e-05, "loss": 0.0178, "step": 146770 }, { "grad_norm": 0.0996704027056694, "learning_rate": 5.589520523582369e-05, "loss": 0.0132, "step": 146780 }, { "grad_norm": 0.08274976909160614, "learning_rate": 5.5889732077212144e-05, "loss": 0.015, "step": 146790 }, { "grad_norm": 0.16672836244106293, "learning_rate": 5.588425884703478e-05, "loss": 0.0129, "step": 146800 }, { "grad_norm": 0.12390552461147308, "learning_rate": 5.587878554535813e-05, "loss": 0.0154, "step": 146810 }, { "grad_norm": 0.08875112235546112, "learning_rate": 5.587331217224867e-05, "loss": 0.0131, "step": 146820 }, { "grad_norm": 0.16206210851669312, "learning_rate": 5.586783872777292e-05, "loss": 0.0162, "step": 146830 }, { "grad_norm": 0.09435661882162094, "learning_rate": 5.586236521199739e-05, "loss": 0.0166, "step": 146840 }, { "grad_norm": 0.10297632217407227, "learning_rate": 5.5856891624988606e-05, "loss": 0.0096, "step": 146850 }, { "grad_norm": 0.09573410451412201, "learning_rate": 5.585141796681305e-05, "loss": 0.0184, "step": 146860 }, { "grad_norm": 0.11112247407436371, "learning_rate": 5.584594423753724e-05, "loss": 0.0144, "step": 146870 }, { "grad_norm": 0.17049618065357208, "learning_rate": 5.5840470437227673e-05, "loss": 0.0201, "step": 146880 }, { "grad_norm": 0.13944078981876373, "learning_rate": 5.583499656595089e-05, "loss": 0.0131, "step": 146890 }, { "grad_norm": 0.11291984468698502, "learning_rate": 5.582952262377339e-05, "loss": 0.0228, "step": 146900 }, { "grad_norm": 0.193595290184021, "learning_rate": 5.582404861076169e-05, "loss": 0.0166, "step": 146910 }, { "grad_norm": 0.19314348697662354, "learning_rate": 5.581857452698229e-05, "loss": 0.0178, "step": 146920 }, { "grad_norm": 0.21730676293373108, "learning_rate": 5.581310037250173e-05, "loss": 0.0138, "step": 146930 }, { "grad_norm": 0.32855555415153503, "learning_rate": 5.580762614738649e-05, "loss": 0.0163, "step": 146940 }, { "grad_norm": 0.1811763197183609, "learning_rate": 5.580215185170313e-05, "loss": 0.0121, "step": 146950 }, { "grad_norm": 0.08815772086381912, "learning_rate": 5.579667748551813e-05, "loss": 0.0105, "step": 146960 }, { "grad_norm": 0.16579940915107727, "learning_rate": 5.5791203048898043e-05, "loss": 0.0121, "step": 146970 }, { "grad_norm": 0.10778428614139557, "learning_rate": 5.5785728541909366e-05, "loss": 0.013, "step": 146980 }, { "grad_norm": 0.1837342083454132, "learning_rate": 5.578025396461861e-05, "loss": 0.0221, "step": 146990 }, { "grad_norm": 0.06842860579490662, "learning_rate": 5.577477931709233e-05, "loss": 0.0126, "step": 147000 }, { "grad_norm": 0.06247059628367424, "learning_rate": 5.576930459939702e-05, "loss": 0.0088, "step": 147010 }, { "grad_norm": 0.08876954764127731, "learning_rate": 5.576382981159921e-05, "loss": 0.0175, "step": 147020 }, { "grad_norm": 0.14505890011787415, "learning_rate": 5.5758354953765426e-05, "loss": 0.0126, "step": 147030 }, { "grad_norm": 0.10210204124450684, "learning_rate": 5.575288002596221e-05, "loss": 0.0104, "step": 147040 }, { "grad_norm": 0.1868004947900772, "learning_rate": 5.574740502825604e-05, "loss": 0.02, "step": 147050 }, { "grad_norm": 0.12502959370613098, "learning_rate": 5.57419299607135e-05, "loss": 0.0145, "step": 147060 }, { "grad_norm": 0.10789922624826431, "learning_rate": 5.573645482340107e-05, "loss": 0.0154, "step": 147070 }, { "grad_norm": 0.2137935608625412, "learning_rate": 5.57309796163853e-05, "loss": 0.0124, "step": 147080 }, { "grad_norm": 0.14493083953857422, "learning_rate": 5.5725504339732736e-05, "loss": 0.0151, "step": 147090 }, { "grad_norm": 0.17947734892368317, "learning_rate": 5.5720028993509865e-05, "loss": 0.0168, "step": 147100 }, { "grad_norm": 0.1103583574295044, "learning_rate": 5.571455357778326e-05, "loss": 0.0148, "step": 147110 }, { "grad_norm": 0.08735664933919907, "learning_rate": 5.570907809261942e-05, "loss": 0.0199, "step": 147120 }, { "grad_norm": 0.15481770038604736, "learning_rate": 5.5703602538084895e-05, "loss": 0.0205, "step": 147130 }, { "grad_norm": 0.13239528238773346, "learning_rate": 5.569812691424622e-05, "loss": 0.0144, "step": 147140 }, { "grad_norm": 0.08679452538490295, "learning_rate": 5.5692651221169914e-05, "loss": 0.0165, "step": 147150 }, { "grad_norm": 0.067934051156044, "learning_rate": 5.568717545892253e-05, "loss": 0.0118, "step": 147160 }, { "grad_norm": 0.10633350163698196, "learning_rate": 5.568169962757058e-05, "loss": 0.0118, "step": 147170 }, { "grad_norm": 0.11655721813440323, "learning_rate": 5.5676223727180635e-05, "loss": 0.0142, "step": 147180 }, { "grad_norm": 0.06923501193523407, "learning_rate": 5.56707477578192e-05, "loss": 0.0165, "step": 147190 }, { "grad_norm": 0.11512397229671478, "learning_rate": 5.5665271719552826e-05, "loss": 0.0117, "step": 147200 }, { "grad_norm": 0.21213681995868683, "learning_rate": 5.565979561244805e-05, "loss": 0.0164, "step": 147210 }, { "grad_norm": 0.08387356996536255, "learning_rate": 5.5654319436571425e-05, "loss": 0.0126, "step": 147220 }, { "grad_norm": 0.08656766265630722, "learning_rate": 5.5648843191989476e-05, "loss": 0.0156, "step": 147230 }, { "grad_norm": 0.48236656188964844, "learning_rate": 5.5643366878768734e-05, "loss": 0.0246, "step": 147240 }, { "grad_norm": 0.09428729116916656, "learning_rate": 5.5637890496975776e-05, "loss": 0.0182, "step": 147250 }, { "grad_norm": 0.18231189250946045, "learning_rate": 5.5632414046677104e-05, "loss": 0.0198, "step": 147260 }, { "grad_norm": 0.08963759243488312, "learning_rate": 5.562693752793931e-05, "loss": 0.0106, "step": 147270 }, { "grad_norm": 0.11370609700679779, "learning_rate": 5.562146094082888e-05, "loss": 0.0112, "step": 147280 }, { "grad_norm": 0.11564167588949203, "learning_rate": 5.5615984285412414e-05, "loss": 0.0173, "step": 147290 }, { "grad_norm": 0.12224537879228592, "learning_rate": 5.561050756175644e-05, "loss": 0.0134, "step": 147300 }, { "grad_norm": 0.061217524111270905, "learning_rate": 5.560503076992749e-05, "loss": 0.0139, "step": 147310 }, { "grad_norm": 0.10349559783935547, "learning_rate": 5.559955390999213e-05, "loss": 0.0147, "step": 147320 }, { "grad_norm": 0.11603857576847076, "learning_rate": 5.5594076982016895e-05, "loss": 0.0216, "step": 147330 }, { "grad_norm": 0.10467485338449478, "learning_rate": 5.558859998606836e-05, "loss": 0.0109, "step": 147340 }, { "grad_norm": 0.22196458280086517, "learning_rate": 5.558312292221304e-05, "loss": 0.0128, "step": 147350 }, { "grad_norm": 0.13206598162651062, "learning_rate": 5.557764579051752e-05, "loss": 0.0123, "step": 147360 }, { "grad_norm": 0.24730907380580902, "learning_rate": 5.5572168591048324e-05, "loss": 0.0199, "step": 147370 }, { "grad_norm": 0.31198257207870483, "learning_rate": 5.556669132387203e-05, "loss": 0.0197, "step": 147380 }, { "grad_norm": 0.10958272218704224, "learning_rate": 5.556121398905518e-05, "loss": 0.02, "step": 147390 }, { "grad_norm": 0.4538084864616394, "learning_rate": 5.555573658666432e-05, "loss": 0.0154, "step": 147400 }, { "grad_norm": 0.2696925401687622, "learning_rate": 5.5550259116766037e-05, "loss": 0.0141, "step": 147410 }, { "grad_norm": 0.11125235259532928, "learning_rate": 5.554478157942684e-05, "loss": 0.0109, "step": 147420 }, { "grad_norm": 0.13563567399978638, "learning_rate": 5.553930397471332e-05, "loss": 0.0145, "step": 147430 }, { "grad_norm": 0.10195794701576233, "learning_rate": 5.5533826302692026e-05, "loss": 0.0209, "step": 147440 }, { "grad_norm": 0.1374419927597046, "learning_rate": 5.5528348563429524e-05, "loss": 0.0157, "step": 147450 }, { "grad_norm": 0.08441617339849472, "learning_rate": 5.5522870756992364e-05, "loss": 0.0135, "step": 147460 }, { "grad_norm": 0.05140339955687523, "learning_rate": 5.551739288344711e-05, "loss": 0.014, "step": 147470 }, { "grad_norm": 0.22200952470302582, "learning_rate": 5.5511914942860323e-05, "loss": 0.0117, "step": 147480 }, { "grad_norm": 0.1983269602060318, "learning_rate": 5.550643693529856e-05, "loss": 0.015, "step": 147490 }, { "grad_norm": 0.19822421669960022, "learning_rate": 5.550095886082839e-05, "loss": 0.0133, "step": 147500 }, { "grad_norm": 0.1186329573392868, "learning_rate": 5.549548071951638e-05, "loss": 0.0132, "step": 147510 }, { "grad_norm": 0.1920613944530487, "learning_rate": 5.549000251142909e-05, "loss": 0.0129, "step": 147520 }, { "grad_norm": 0.08964870125055313, "learning_rate": 5.548452423663309e-05, "loss": 0.0135, "step": 147530 }, { "grad_norm": 0.1874375343322754, "learning_rate": 5.5479045895194934e-05, "loss": 0.0189, "step": 147540 }, { "grad_norm": 0.13413691520690918, "learning_rate": 5.54735674871812e-05, "loss": 0.0216, "step": 147550 }, { "grad_norm": 0.15558339655399323, "learning_rate": 5.5468089012658455e-05, "loss": 0.0162, "step": 147560 }, { "grad_norm": 0.12524022161960602, "learning_rate": 5.5462610471693255e-05, "loss": 0.0161, "step": 147570 }, { "grad_norm": 0.08587933331727982, "learning_rate": 5.5457131864352196e-05, "loss": 0.0178, "step": 147580 }, { "grad_norm": 0.10835359990596771, "learning_rate": 5.5451653190701826e-05, "loss": 0.0128, "step": 147590 }, { "grad_norm": 0.16882015764713287, "learning_rate": 5.544617445080873e-05, "loss": 0.014, "step": 147600 }, { "grad_norm": 0.3020373582839966, "learning_rate": 5.544069564473946e-05, "loss": 0.0144, "step": 147610 }, { "grad_norm": 0.12376975268125534, "learning_rate": 5.54352167725606e-05, "loss": 0.0206, "step": 147620 }, { "grad_norm": 0.14345954358577728, "learning_rate": 5.542973783433873e-05, "loss": 0.0142, "step": 147630 }, { "grad_norm": 0.07435710728168488, "learning_rate": 5.542425883014043e-05, "loss": 0.0112, "step": 147640 }, { "grad_norm": 0.07150337845087051, "learning_rate": 5.541877976003225e-05, "loss": 0.0166, "step": 147650 }, { "grad_norm": 0.15663976967334747, "learning_rate": 5.541330062408079e-05, "loss": 0.0122, "step": 147660 }, { "grad_norm": 0.17246532440185547, "learning_rate": 5.54078214223526e-05, "loss": 0.0164, "step": 147670 }, { "grad_norm": 0.11032747477293015, "learning_rate": 5.540234215491429e-05, "loss": 0.0151, "step": 147680 }, { "grad_norm": 0.14871065318584442, "learning_rate": 5.539686282183242e-05, "loss": 0.0202, "step": 147690 }, { "grad_norm": 0.13873887062072754, "learning_rate": 5.539138342317357e-05, "loss": 0.0179, "step": 147700 }, { "grad_norm": 0.09188326448202133, "learning_rate": 5.5385903959004316e-05, "loss": 0.0191, "step": 147710 }, { "grad_norm": 0.129940465092659, "learning_rate": 5.5380424429391245e-05, "loss": 0.0199, "step": 147720 }, { "grad_norm": 0.13288629055023193, "learning_rate": 5.5374944834400946e-05, "loss": 0.0118, "step": 147730 }, { "grad_norm": 0.12863677740097046, "learning_rate": 5.5369465174099975e-05, "loss": 0.0182, "step": 147740 }, { "grad_norm": 0.08206342160701752, "learning_rate": 5.536398544855494e-05, "loss": 0.0162, "step": 147750 }, { "grad_norm": 0.21324117481708527, "learning_rate": 5.535850565783243e-05, "loss": 0.017, "step": 147760 }, { "grad_norm": 0.10297494381666183, "learning_rate": 5.5353025801999e-05, "loss": 0.0176, "step": 147770 }, { "grad_norm": 0.13383173942565918, "learning_rate": 5.534754588112127e-05, "loss": 0.0148, "step": 147780 }, { "grad_norm": 0.10848536342382431, "learning_rate": 5.534206589526579e-05, "loss": 0.0116, "step": 147790 }, { "grad_norm": 0.10968947410583496, "learning_rate": 5.533658584449918e-05, "loss": 0.0111, "step": 147800 }, { "grad_norm": 0.05656638368964195, "learning_rate": 5.5331105728888e-05, "loss": 0.0124, "step": 147810 }, { "grad_norm": 0.14562015235424042, "learning_rate": 5.532562554849886e-05, "loss": 0.0177, "step": 147820 }, { "grad_norm": 0.10043754428625107, "learning_rate": 5.532014530339834e-05, "loss": 0.0106, "step": 147830 }, { "grad_norm": 0.40410396456718445, "learning_rate": 5.531466499365303e-05, "loss": 0.0129, "step": 147840 }, { "grad_norm": 0.15635348856449127, "learning_rate": 5.5309184619329524e-05, "loss": 0.0164, "step": 147850 }, { "grad_norm": 0.10150160640478134, "learning_rate": 5.530370418049442e-05, "loss": 0.0106, "step": 147860 }, { "grad_norm": 0.11577987670898438, "learning_rate": 5.52982236772143e-05, "loss": 0.0162, "step": 147870 }, { "grad_norm": 0.12756285071372986, "learning_rate": 5.529274310955575e-05, "loss": 0.0187, "step": 147880 }, { "grad_norm": 0.1015978679060936, "learning_rate": 5.5287262477585376e-05, "loss": 0.0119, "step": 147890 }, { "grad_norm": 0.12339767068624496, "learning_rate": 5.528178178136978e-05, "loss": 0.0169, "step": 147900 }, { "grad_norm": 0.22682848572731018, "learning_rate": 5.527630102097554e-05, "loss": 0.018, "step": 147910 }, { "grad_norm": 0.10938192903995514, "learning_rate": 5.5270820196469266e-05, "loss": 0.0125, "step": 147920 }, { "grad_norm": 0.09177755564451218, "learning_rate": 5.526533930791755e-05, "loss": 0.0099, "step": 147930 }, { "grad_norm": 0.08345299959182739, "learning_rate": 5.525985835538699e-05, "loss": 0.013, "step": 147940 }, { "grad_norm": 0.08124302327632904, "learning_rate": 5.525437733894417e-05, "loss": 0.0181, "step": 147950 }, { "grad_norm": 0.12756535410881042, "learning_rate": 5.524889625865572e-05, "loss": 0.0142, "step": 147960 }, { "grad_norm": 0.12838059663772583, "learning_rate": 5.5243415114588224e-05, "loss": 0.0163, "step": 147970 }, { "grad_norm": 0.14727571606636047, "learning_rate": 5.523793390680829e-05, "loss": 0.0164, "step": 147980 }, { "grad_norm": 0.09272918850183487, "learning_rate": 5.523245263538249e-05, "loss": 0.0146, "step": 147990 }, { "grad_norm": 0.10779318958520889, "learning_rate": 5.5226971300377486e-05, "loss": 0.0173, "step": 148000 }, { "grad_norm": 0.24851977825164795, "learning_rate": 5.5221489901859814e-05, "loss": 0.0141, "step": 148010 }, { "grad_norm": 0.8912979960441589, "learning_rate": 5.521600843989613e-05, "loss": 0.0152, "step": 148020 }, { "grad_norm": 0.19972604513168335, "learning_rate": 5.521052691455301e-05, "loss": 0.0166, "step": 148030 }, { "grad_norm": 0.1356048434972763, "learning_rate": 5.5205045325897065e-05, "loss": 0.0198, "step": 148040 }, { "grad_norm": 0.13850219547748566, "learning_rate": 5.51995636739949e-05, "loss": 0.0177, "step": 148050 }, { "grad_norm": 0.10918908566236496, "learning_rate": 5.5194081958913134e-05, "loss": 0.016, "step": 148060 }, { "grad_norm": 0.4849095344543457, "learning_rate": 5.518860018071838e-05, "loss": 0.0166, "step": 148070 }, { "grad_norm": 0.09443776309490204, "learning_rate": 5.518311833947721e-05, "loss": 0.0177, "step": 148080 }, { "grad_norm": 0.11661071330308914, "learning_rate": 5.5177636435256286e-05, "loss": 0.0138, "step": 148090 }, { "grad_norm": 0.12504123151302338, "learning_rate": 5.517215446812217e-05, "loss": 0.014, "step": 148100 }, { "grad_norm": 0.12894757091999054, "learning_rate": 5.516667243814151e-05, "loss": 0.0163, "step": 148110 }, { "grad_norm": 0.2647729516029358, "learning_rate": 5.51611903453809e-05, "loss": 0.0219, "step": 148120 }, { "grad_norm": 0.13083726167678833, "learning_rate": 5.515570818990693e-05, "loss": 0.0149, "step": 148130 }, { "grad_norm": 0.4067399799823761, "learning_rate": 5.5150225971786265e-05, "loss": 0.018, "step": 148140 }, { "grad_norm": 0.1576635241508484, "learning_rate": 5.514474369108548e-05, "loss": 0.0233, "step": 148150 }, { "grad_norm": 0.06658420711755753, "learning_rate": 5.513926134787121e-05, "loss": 0.0164, "step": 148160 }, { "grad_norm": 0.1813899725675583, "learning_rate": 5.5133778942210045e-05, "loss": 0.018, "step": 148170 }, { "grad_norm": 0.10821432620286942, "learning_rate": 5.512829647416864e-05, "loss": 0.0142, "step": 148180 }, { "grad_norm": 0.10850681364536285, "learning_rate": 5.512281394381358e-05, "loss": 0.0182, "step": 148190 }, { "grad_norm": 0.16163600981235504, "learning_rate": 5.5117331351211485e-05, "loss": 0.0128, "step": 148200 }, { "grad_norm": 0.805572509765625, "learning_rate": 5.5111848696429e-05, "loss": 0.0181, "step": 148210 }, { "grad_norm": 0.061823535710573196, "learning_rate": 5.5106365979532714e-05, "loss": 0.0141, "step": 148220 }, { "grad_norm": 0.1255832314491272, "learning_rate": 5.5100883200589274e-05, "loss": 0.0163, "step": 148230 }, { "grad_norm": 0.4467500150203705, "learning_rate": 5.5095400359665274e-05, "loss": 0.0151, "step": 148240 }, { "grad_norm": 0.15728828310966492, "learning_rate": 5.5089917456827355e-05, "loss": 0.0121, "step": 148250 }, { "grad_norm": 0.09601247310638428, "learning_rate": 5.5084434492142125e-05, "loss": 0.0143, "step": 148260 }, { "grad_norm": 0.10834052413702011, "learning_rate": 5.507895146567622e-05, "loss": 0.0105, "step": 148270 }, { "grad_norm": 0.20559293031692505, "learning_rate": 5.5073468377496265e-05, "loss": 0.016, "step": 148280 }, { "grad_norm": 0.12980033457279205, "learning_rate": 5.506798522766887e-05, "loss": 0.0152, "step": 148290 }, { "grad_norm": 0.0895535796880722, "learning_rate": 5.506250201626068e-05, "loss": 0.0165, "step": 148300 }, { "grad_norm": 0.09460600465536118, "learning_rate": 5.505701874333831e-05, "loss": 0.0124, "step": 148310 }, { "grad_norm": 0.08717084676027298, "learning_rate": 5.5051535408968366e-05, "loss": 0.0212, "step": 148320 }, { "grad_norm": 0.08925279229879379, "learning_rate": 5.504605201321752e-05, "loss": 0.0165, "step": 148330 }, { "grad_norm": 0.08012788742780685, "learning_rate": 5.504056855615237e-05, "loss": 0.0169, "step": 148340 }, { "grad_norm": 0.20332621037960052, "learning_rate": 5.503508503783956e-05, "loss": 0.0182, "step": 148350 }, { "grad_norm": 0.2921890914440155, "learning_rate": 5.5029601458345705e-05, "loss": 0.0162, "step": 148360 }, { "grad_norm": 0.1780405193567276, "learning_rate": 5.502411781773745e-05, "loss": 0.0145, "step": 148370 }, { "grad_norm": 0.25586071610450745, "learning_rate": 5.5018634116081415e-05, "loss": 0.0181, "step": 148380 }, { "grad_norm": 0.11794837564229965, "learning_rate": 5.501315035344423e-05, "loss": 0.0167, "step": 148390 }, { "grad_norm": 0.09425646811723709, "learning_rate": 5.500766652989254e-05, "loss": 0.0161, "step": 148400 }, { "grad_norm": 0.16140435636043549, "learning_rate": 5.500218264549298e-05, "loss": 0.0111, "step": 148410 }, { "grad_norm": 0.1058223620057106, "learning_rate": 5.499669870031218e-05, "loss": 0.0138, "step": 148420 }, { "grad_norm": 0.0868995413184166, "learning_rate": 5.499121469441677e-05, "loss": 0.0186, "step": 148430 }, { "grad_norm": 0.09212297201156616, "learning_rate": 5.498573062787339e-05, "loss": 0.0188, "step": 148440 }, { "grad_norm": 0.13705770671367645, "learning_rate": 5.4980246500748665e-05, "loss": 0.0112, "step": 148450 }, { "grad_norm": 0.13252094388008118, "learning_rate": 5.497476231310926e-05, "loss": 0.0155, "step": 148460 }, { "grad_norm": 0.16113296151161194, "learning_rate": 5.496927806502178e-05, "loss": 0.0163, "step": 148470 }, { "grad_norm": 0.9926427006721497, "learning_rate": 5.496379375655289e-05, "loss": 0.0187, "step": 148480 }, { "grad_norm": 0.13527950644493103, "learning_rate": 5.495830938776921e-05, "loss": 0.0173, "step": 148490 }, { "grad_norm": 0.18777316808700562, "learning_rate": 5.49528249587374e-05, "loss": 0.0175, "step": 148500 }, { "grad_norm": 0.12216820567846298, "learning_rate": 5.494734046952409e-05, "loss": 0.0176, "step": 148510 }, { "grad_norm": 0.15103687345981598, "learning_rate": 5.494185592019592e-05, "loss": 0.0204, "step": 148520 }, { "grad_norm": 0.08471620082855225, "learning_rate": 5.493637131081953e-05, "loss": 0.0143, "step": 148530 }, { "grad_norm": 0.18057596683502197, "learning_rate": 5.493088664146156e-05, "loss": 0.0204, "step": 148540 }, { "grad_norm": 0.13345755636692047, "learning_rate": 5.492540191218869e-05, "loss": 0.0186, "step": 148550 }, { "grad_norm": 0.2623766362667084, "learning_rate": 5.491991712306751e-05, "loss": 0.0187, "step": 148560 }, { "grad_norm": 0.10165415704250336, "learning_rate": 5.491443227416471e-05, "loss": 0.0126, "step": 148570 }, { "grad_norm": 0.45596903562545776, "learning_rate": 5.4908947365546895e-05, "loss": 0.0178, "step": 148580 }, { "grad_norm": 0.09642837941646576, "learning_rate": 5.4903462397280756e-05, "loss": 0.0134, "step": 148590 }, { "grad_norm": 0.5547844767570496, "learning_rate": 5.489797736943292e-05, "loss": 0.0196, "step": 148600 }, { "grad_norm": 0.20882803201675415, "learning_rate": 5.4892492282070016e-05, "loss": 0.0182, "step": 148610 }, { "grad_norm": 0.19503098726272583, "learning_rate": 5.4887007135258724e-05, "loss": 0.016, "step": 148620 }, { "grad_norm": 0.20776452124118805, "learning_rate": 5.488152192906567e-05, "loss": 0.0142, "step": 148630 }, { "grad_norm": 0.057189423590898514, "learning_rate": 5.4876036663557526e-05, "loss": 0.0195, "step": 148640 }, { "grad_norm": 0.09191692620515823, "learning_rate": 5.4870551338800936e-05, "loss": 0.0159, "step": 148650 }, { "grad_norm": 0.08299677819013596, "learning_rate": 5.486506595486254e-05, "loss": 0.0152, "step": 148660 }, { "grad_norm": 0.1075812503695488, "learning_rate": 5.485958051180902e-05, "loss": 0.0229, "step": 148670 }, { "grad_norm": 0.05859474092721939, "learning_rate": 5.485409500970697e-05, "loss": 0.0168, "step": 148680 }, { "grad_norm": 0.1603391170501709, "learning_rate": 5.484860944862311e-05, "loss": 0.0151, "step": 148690 }, { "grad_norm": 0.1844904124736786, "learning_rate": 5.484312382862407e-05, "loss": 0.0138, "step": 148700 }, { "grad_norm": 0.08221364766359329, "learning_rate": 5.4837638149776495e-05, "loss": 0.0126, "step": 148710 }, { "grad_norm": 0.18620702624320984, "learning_rate": 5.4832152412147044e-05, "loss": 0.0155, "step": 148720 }, { "grad_norm": 0.11510029435157776, "learning_rate": 5.482666661580238e-05, "loss": 0.0155, "step": 148730 }, { "grad_norm": 0.14309246838092804, "learning_rate": 5.4821180760809155e-05, "loss": 0.0138, "step": 148740 }, { "grad_norm": 0.14332643151283264, "learning_rate": 5.481569484723405e-05, "loss": 0.012, "step": 148750 }, { "grad_norm": 0.2085631787776947, "learning_rate": 5.4810208875143695e-05, "loss": 0.0172, "step": 148760 }, { "grad_norm": 0.14032268524169922, "learning_rate": 5.480472284460476e-05, "loss": 0.0188, "step": 148770 }, { "grad_norm": 0.12319779396057129, "learning_rate": 5.4799236755683914e-05, "loss": 0.0145, "step": 148780 }, { "grad_norm": 0.07505183666944504, "learning_rate": 5.479375060844779e-05, "loss": 0.0146, "step": 148790 }, { "grad_norm": 0.15017060935497284, "learning_rate": 5.47882644029631e-05, "loss": 0.0156, "step": 148800 }, { "grad_norm": 0.11369308829307556, "learning_rate": 5.4782778139296445e-05, "loss": 0.0286, "step": 148810 }, { "grad_norm": 0.35936984419822693, "learning_rate": 5.4777291817514544e-05, "loss": 0.0202, "step": 148820 }, { "grad_norm": 0.11350642144680023, "learning_rate": 5.477180543768403e-05, "loss": 0.0167, "step": 148830 }, { "grad_norm": 0.11074518412351608, "learning_rate": 5.4766318999871566e-05, "loss": 0.0126, "step": 148840 }, { "grad_norm": 0.14154115319252014, "learning_rate": 5.476083250414383e-05, "loss": 0.017, "step": 148850 }, { "grad_norm": 0.14044725894927979, "learning_rate": 5.475534595056748e-05, "loss": 0.0179, "step": 148860 }, { "grad_norm": 0.08654561638832092, "learning_rate": 5.474985933920921e-05, "loss": 0.0115, "step": 148870 }, { "grad_norm": 0.11913446336984634, "learning_rate": 5.474437267013563e-05, "loss": 0.0155, "step": 148880 }, { "grad_norm": 0.09862467646598816, "learning_rate": 5.4738885943413476e-05, "loss": 0.0258, "step": 148890 }, { "grad_norm": 0.2439592480659485, "learning_rate": 5.4733399159109356e-05, "loss": 0.0162, "step": 148900 }, { "grad_norm": 0.14522016048431396, "learning_rate": 5.4727912317289985e-05, "loss": 0.0151, "step": 148910 }, { "grad_norm": 0.17791195213794708, "learning_rate": 5.472242541802202e-05, "loss": 0.0136, "step": 148920 }, { "grad_norm": 0.25295719504356384, "learning_rate": 5.4716938461372106e-05, "loss": 0.0153, "step": 148930 }, { "grad_norm": 0.17948848009109497, "learning_rate": 5.471145144740695e-05, "loss": 0.0179, "step": 148940 }, { "grad_norm": 0.0949617326259613, "learning_rate": 5.47059643761932e-05, "loss": 0.014, "step": 148950 }, { "grad_norm": 0.11255095154047012, "learning_rate": 5.470047724779756e-05, "loss": 0.0187, "step": 148960 }, { "grad_norm": 0.05753468722105026, "learning_rate": 5.469499006228665e-05, "loss": 0.0158, "step": 148970 }, { "grad_norm": 0.17716512084007263, "learning_rate": 5.468950281972721e-05, "loss": 0.0143, "step": 148980 }, { "grad_norm": 0.1092396154999733, "learning_rate": 5.468401552018587e-05, "loss": 0.0146, "step": 148990 }, { "grad_norm": 0.19517238438129425, "learning_rate": 5.4678528163729326e-05, "loss": 0.0148, "step": 149000 }, { "grad_norm": 0.10128960758447647, "learning_rate": 5.467304075042423e-05, "loss": 0.0163, "step": 149010 }, { "grad_norm": 0.10905030369758606, "learning_rate": 5.46675532803373e-05, "loss": 0.0152, "step": 149020 }, { "grad_norm": 0.11025139689445496, "learning_rate": 5.466206575353518e-05, "loss": 0.0165, "step": 149030 }, { "grad_norm": 0.2735258936882019, "learning_rate": 5.465657817008456e-05, "loss": 0.0127, "step": 149040 }, { "grad_norm": 0.14310011267662048, "learning_rate": 5.4651090530052105e-05, "loss": 0.0143, "step": 149050 }, { "grad_norm": 0.09346210956573486, "learning_rate": 5.4645602833504526e-05, "loss": 0.0148, "step": 149060 }, { "grad_norm": 0.1564437448978424, "learning_rate": 5.464011508050848e-05, "loss": 0.0122, "step": 149070 }, { "grad_norm": 0.0856025218963623, "learning_rate": 5.4634627271130667e-05, "loss": 0.0149, "step": 149080 }, { "grad_norm": 0.2219318449497223, "learning_rate": 5.4629139405437736e-05, "loss": 0.0243, "step": 149090 }, { "grad_norm": 0.08365365117788315, "learning_rate": 5.4623651483496404e-05, "loss": 0.0134, "step": 149100 }, { "grad_norm": 0.13365453481674194, "learning_rate": 5.4618163505373335e-05, "loss": 0.0116, "step": 149110 }, { "grad_norm": 0.07492508739233017, "learning_rate": 5.461267547113522e-05, "loss": 0.0144, "step": 149120 }, { "grad_norm": 0.1223776713013649, "learning_rate": 5.4607187380848746e-05, "loss": 0.0107, "step": 149130 }, { "grad_norm": 0.1608324944972992, "learning_rate": 5.4601699234580604e-05, "loss": 0.0193, "step": 149140 }, { "grad_norm": 0.18367226421833038, "learning_rate": 5.459621103239746e-05, "loss": 0.0137, "step": 149150 }, { "grad_norm": 0.2293500155210495, "learning_rate": 5.459072277436601e-05, "loss": 0.0184, "step": 149160 }, { "grad_norm": 0.10078026354312897, "learning_rate": 5.4585234460552955e-05, "loss": 0.0118, "step": 149170 }, { "grad_norm": 0.10552740097045898, "learning_rate": 5.457974609102497e-05, "loss": 0.0163, "step": 149180 }, { "grad_norm": 0.09450460970401764, "learning_rate": 5.457425766584874e-05, "loss": 0.0127, "step": 149190 }, { "grad_norm": 0.16546177864074707, "learning_rate": 5.456876918509095e-05, "loss": 0.0168, "step": 149200 }, { "grad_norm": 0.0812034010887146, "learning_rate": 5.4563280648818325e-05, "loss": 0.0122, "step": 149210 }, { "grad_norm": 0.11297299712896347, "learning_rate": 5.4557792057097515e-05, "loss": 0.0121, "step": 149220 }, { "grad_norm": 0.09804640710353851, "learning_rate": 5.455230340999523e-05, "loss": 0.0147, "step": 149230 }, { "grad_norm": 0.08987505733966827, "learning_rate": 5.454681470757818e-05, "loss": 0.0142, "step": 149240 }, { "grad_norm": 0.16468925774097443, "learning_rate": 5.4541325949913e-05, "loss": 0.0141, "step": 149250 }, { "grad_norm": 0.16346478462219238, "learning_rate": 5.453583713706646e-05, "loss": 0.0157, "step": 149260 }, { "grad_norm": 0.11545498669147491, "learning_rate": 5.453034826910519e-05, "loss": 0.0131, "step": 149270 }, { "grad_norm": 0.09110376983880997, "learning_rate": 5.4524859346095926e-05, "loss": 0.0163, "step": 149280 }, { "grad_norm": 0.2613053023815155, "learning_rate": 5.4519370368105336e-05, "loss": 0.015, "step": 149290 }, { "grad_norm": 0.14332900941371918, "learning_rate": 5.4513881335200144e-05, "loss": 0.0112, "step": 149300 }, { "grad_norm": 0.05103851854801178, "learning_rate": 5.4508392247447014e-05, "loss": 0.009, "step": 149310 }, { "grad_norm": 0.15255245566368103, "learning_rate": 5.4502903104912685e-05, "loss": 0.0136, "step": 149320 }, { "grad_norm": 0.38526615500450134, "learning_rate": 5.4497413907663816e-05, "loss": 0.0125, "step": 149330 }, { "grad_norm": 0.1996082216501236, "learning_rate": 5.4491924655767126e-05, "loss": 0.0168, "step": 149340 }, { "grad_norm": 0.057264555245637894, "learning_rate": 5.448643534928931e-05, "loss": 0.0176, "step": 149350 }, { "grad_norm": 0.08726280927658081, "learning_rate": 5.448094598829707e-05, "loss": 0.0095, "step": 149360 }, { "grad_norm": 0.3834722340106964, "learning_rate": 5.447545657285711e-05, "loss": 0.0142, "step": 149370 }, { "grad_norm": 0.15556657314300537, "learning_rate": 5.446996710303611e-05, "loss": 0.0205, "step": 149380 }, { "grad_norm": 0.09228982776403427, "learning_rate": 5.446447757890081e-05, "loss": 0.015, "step": 149390 }, { "grad_norm": 0.16098494827747345, "learning_rate": 5.4458988000517885e-05, "loss": 0.0157, "step": 149400 }, { "grad_norm": 0.17504815757274628, "learning_rate": 5.445349836795404e-05, "loss": 0.0158, "step": 149410 }, { "grad_norm": 0.5114176869392395, "learning_rate": 5.4448008681275985e-05, "loss": 0.0134, "step": 149420 }, { "grad_norm": 0.1487496942281723, "learning_rate": 5.444251894055042e-05, "loss": 0.0166, "step": 149430 }, { "grad_norm": 0.0980713739991188, "learning_rate": 5.443702914584408e-05, "loss": 0.0124, "step": 149440 }, { "grad_norm": 0.09810221195220947, "learning_rate": 5.443153929722362e-05, "loss": 0.013, "step": 149450 }, { "grad_norm": 0.07018920034170151, "learning_rate": 5.4426049394755786e-05, "loss": 0.0171, "step": 149460 }, { "grad_norm": 0.4655076563358307, "learning_rate": 5.442055943850727e-05, "loss": 0.0196, "step": 149470 }, { "grad_norm": 0.08855587244033813, "learning_rate": 5.441506942854478e-05, "loss": 0.0182, "step": 149480 }, { "grad_norm": 0.23663875460624695, "learning_rate": 5.4409579364935035e-05, "loss": 0.0159, "step": 149490 }, { "grad_norm": 0.11224019527435303, "learning_rate": 5.4404089247744736e-05, "loss": 0.0193, "step": 149500 }, { "grad_norm": 0.12942467629909515, "learning_rate": 5.439859907704059e-05, "loss": 0.0154, "step": 149510 }, { "grad_norm": 0.42225781083106995, "learning_rate": 5.439310885288932e-05, "loss": 0.0139, "step": 149520 }, { "grad_norm": 0.20888333022594452, "learning_rate": 5.438761857535762e-05, "loss": 0.0202, "step": 149530 }, { "grad_norm": 0.09693828225135803, "learning_rate": 5.438212824451221e-05, "loss": 0.0178, "step": 149540 }, { "grad_norm": 0.1398838758468628, "learning_rate": 5.437663786041982e-05, "loss": 0.0135, "step": 149550 }, { "grad_norm": 0.10754575580358505, "learning_rate": 5.437114742314714e-05, "loss": 0.0111, "step": 149560 }, { "grad_norm": 0.11515092849731445, "learning_rate": 5.436565693276089e-05, "loss": 0.0141, "step": 149570 }, { "grad_norm": 0.09388484060764313, "learning_rate": 5.436016638932779e-05, "loss": 0.0153, "step": 149580 }, { "grad_norm": 0.07024654000997543, "learning_rate": 5.435467579291454e-05, "loss": 0.0138, "step": 149590 }, { "grad_norm": 0.12350204586982727, "learning_rate": 5.434918514358788e-05, "loss": 0.0161, "step": 149600 }, { "grad_norm": 0.07803989946842194, "learning_rate": 5.4343694441414507e-05, "loss": 0.0157, "step": 149610 }, { "grad_norm": 0.2722635269165039, "learning_rate": 5.433820368646115e-05, "loss": 0.0154, "step": 149620 }, { "grad_norm": 0.13328436017036438, "learning_rate": 5.433271287879451e-05, "loss": 0.0138, "step": 149630 }, { "grad_norm": 0.09815578907728195, "learning_rate": 5.432722201848134e-05, "loss": 0.0196, "step": 149640 }, { "grad_norm": 0.06613314151763916, "learning_rate": 5.432173110558834e-05, "loss": 0.0123, "step": 149650 }, { "grad_norm": 0.09697040170431137, "learning_rate": 5.4316240140182194e-05, "loss": 0.0121, "step": 149660 }, { "grad_norm": 0.12718632817268372, "learning_rate": 5.4310749122329686e-05, "loss": 0.0139, "step": 149670 }, { "grad_norm": 0.0859469324350357, "learning_rate": 5.430525805209749e-05, "loss": 0.0167, "step": 149680 }, { "grad_norm": 0.1217222586274147, "learning_rate": 5.4299766929552345e-05, "loss": 0.0139, "step": 149690 }, { "grad_norm": 0.0782899484038353, "learning_rate": 5.429427575476097e-05, "loss": 0.0112, "step": 149700 }, { "grad_norm": 0.18112370371818542, "learning_rate": 5.4288784527790096e-05, "loss": 0.0172, "step": 149710 }, { "grad_norm": 0.09373202174901962, "learning_rate": 5.4283293248706434e-05, "loss": 0.0151, "step": 149720 }, { "grad_norm": 0.09500639885663986, "learning_rate": 5.427780191757672e-05, "loss": 0.0135, "step": 149730 }, { "grad_norm": 0.09768566489219666, "learning_rate": 5.427231053446767e-05, "loss": 0.0145, "step": 149740 }, { "grad_norm": 0.08170420676469803, "learning_rate": 5.426681909944602e-05, "loss": 0.0148, "step": 149750 }, { "grad_norm": 0.18137256801128387, "learning_rate": 5.4261327612578485e-05, "loss": 0.0188, "step": 149760 }, { "grad_norm": 0.10195234417915344, "learning_rate": 5.425583607393181e-05, "loss": 0.0162, "step": 149770 }, { "grad_norm": 0.07344602793455124, "learning_rate": 5.425034448357268e-05, "loss": 0.0129, "step": 149780 }, { "grad_norm": 0.27840927243232727, "learning_rate": 5.424485284156787e-05, "loss": 0.0173, "step": 149790 }, { "grad_norm": 0.10720492154359818, "learning_rate": 5.423936114798409e-05, "loss": 0.0138, "step": 149800 }, { "grad_norm": 0.09912041574716568, "learning_rate": 5.423386940288807e-05, "loss": 0.0142, "step": 149810 }, { "grad_norm": 0.07304100692272186, "learning_rate": 5.422837760634654e-05, "loss": 0.0164, "step": 149820 }, { "grad_norm": 0.09732268750667572, "learning_rate": 5.4222885758426224e-05, "loss": 0.0115, "step": 149830 }, { "grad_norm": 0.11530086398124695, "learning_rate": 5.4217393859193855e-05, "loss": 0.0145, "step": 149840 }, { "grad_norm": 0.06716804206371307, "learning_rate": 5.421190190871618e-05, "loss": 0.0184, "step": 149850 }, { "grad_norm": 0.10317172110080719, "learning_rate": 5.420640990705991e-05, "loss": 0.0125, "step": 149860 }, { "grad_norm": 0.1633617877960205, "learning_rate": 5.42009178542918e-05, "loss": 0.0194, "step": 149870 }, { "grad_norm": 0.10544517636299133, "learning_rate": 5.4195425750478565e-05, "loss": 0.0229, "step": 149880 }, { "grad_norm": 0.10581018030643463, "learning_rate": 5.4189933595686936e-05, "loss": 0.0206, "step": 149890 }, { "grad_norm": 0.0841684341430664, "learning_rate": 5.4184441389983674e-05, "loss": 0.0115, "step": 149900 }, { "grad_norm": 0.098320871591568, "learning_rate": 5.417894913343548e-05, "loss": 0.0178, "step": 149910 }, { "grad_norm": 0.0627087727189064, "learning_rate": 5.417345682610914e-05, "loss": 0.0118, "step": 149920 }, { "grad_norm": 0.26421311497688293, "learning_rate": 5.416796446807133e-05, "loss": 0.0172, "step": 149930 }, { "grad_norm": 0.1313425898551941, "learning_rate": 5.416247205938883e-05, "loss": 0.0127, "step": 149940 }, { "grad_norm": 0.07785316556692123, "learning_rate": 5.415697960012835e-05, "loss": 0.0185, "step": 149950 }, { "grad_norm": 0.0994454175233841, "learning_rate": 5.4151487090356666e-05, "loss": 0.0134, "step": 149960 }, { "grad_norm": 0.0735468938946724, "learning_rate": 5.414599453014049e-05, "loss": 0.0162, "step": 149970 }, { "grad_norm": 0.09186217188835144, "learning_rate": 5.414050191954655e-05, "loss": 0.0196, "step": 149980 }, { "grad_norm": 0.04916100576519966, "learning_rate": 5.413500925864161e-05, "loss": 0.0125, "step": 149990 }, { "grad_norm": 0.15230953693389893, "learning_rate": 5.41295165474924e-05, "loss": 0.0134, "step": 150000 }, { "grad_norm": 0.09610769152641296, "learning_rate": 5.4124023786165676e-05, "loss": 0.0205, "step": 150010 }, { "grad_norm": 0.1212526187300682, "learning_rate": 5.4118530974728154e-05, "loss": 0.0135, "step": 150020 }, { "grad_norm": 0.14902308583259583, "learning_rate": 5.41130381132466e-05, "loss": 0.0145, "step": 150030 }, { "grad_norm": 0.12943564355373383, "learning_rate": 5.410754520178777e-05, "loss": 0.0166, "step": 150040 }, { "grad_norm": 0.13196244835853577, "learning_rate": 5.4102052240418356e-05, "loss": 0.0153, "step": 150050 }, { "grad_norm": 0.06974800676107407, "learning_rate": 5.4096559229205145e-05, "loss": 0.0176, "step": 150060 }, { "grad_norm": 0.0727878212928772, "learning_rate": 5.409106616821486e-05, "loss": 0.0103, "step": 150070 }, { "grad_norm": 0.08300753682851791, "learning_rate": 5.4085573057514285e-05, "loss": 0.0146, "step": 150080 }, { "grad_norm": 0.07265543192625046, "learning_rate": 5.408007989717012e-05, "loss": 0.0142, "step": 150090 }, { "grad_norm": 0.14244063198566437, "learning_rate": 5.407458668724914e-05, "loss": 0.0206, "step": 150100 }, { "grad_norm": 0.2378367781639099, "learning_rate": 5.406909342781809e-05, "loss": 0.019, "step": 150110 }, { "grad_norm": 0.1057257205247879, "learning_rate": 5.40636001189437e-05, "loss": 0.0144, "step": 150120 }, { "grad_norm": 0.2696717083454132, "learning_rate": 5.405810676069274e-05, "loss": 0.0186, "step": 150130 }, { "grad_norm": 0.1392456591129303, "learning_rate": 5.405261335313194e-05, "loss": 0.0168, "step": 150140 }, { "grad_norm": 0.08623373508453369, "learning_rate": 5.404711989632808e-05, "loss": 0.0125, "step": 150150 }, { "grad_norm": 0.1561983823776245, "learning_rate": 5.4041626390347875e-05, "loss": 0.017, "step": 150160 }, { "grad_norm": 0.08978274464607239, "learning_rate": 5.4036132835258115e-05, "loss": 0.0163, "step": 150170 }, { "grad_norm": 0.11203055083751678, "learning_rate": 5.40306392311255e-05, "loss": 0.0136, "step": 150180 }, { "grad_norm": 0.14576217532157898, "learning_rate": 5.402514557801684e-05, "loss": 0.0179, "step": 150190 }, { "grad_norm": 0.5075604319572449, "learning_rate": 5.401965187599886e-05, "loss": 0.0228, "step": 150200 }, { "grad_norm": 0.16741910576820374, "learning_rate": 5.401415812513829e-05, "loss": 0.0175, "step": 150210 }, { "grad_norm": 0.07538316398859024, "learning_rate": 5.400866432550193e-05, "loss": 0.0119, "step": 150220 }, { "grad_norm": 0.09611642360687256, "learning_rate": 5.40031704771565e-05, "loss": 0.0172, "step": 150230 }, { "grad_norm": 0.10219088941812515, "learning_rate": 5.3997676580168774e-05, "loss": 0.0163, "step": 150240 }, { "grad_norm": 0.08708347380161285, "learning_rate": 5.3992182634605506e-05, "loss": 0.0153, "step": 150250 }, { "grad_norm": 0.1300370991230011, "learning_rate": 5.398668864053345e-05, "loss": 0.0171, "step": 150260 }, { "grad_norm": 0.08549797534942627, "learning_rate": 5.3981194598019356e-05, "loss": 0.0281, "step": 150270 }, { "grad_norm": 0.08613982051610947, "learning_rate": 5.397570050713e-05, "loss": 0.0189, "step": 150280 }, { "grad_norm": 0.08709869533777237, "learning_rate": 5.397020636793213e-05, "loss": 0.012, "step": 150290 }, { "grad_norm": 0.1704987734556198, "learning_rate": 5.39647121804925e-05, "loss": 0.0132, "step": 150300 }, { "grad_norm": 0.2914181649684906, "learning_rate": 5.395921794487787e-05, "loss": 0.0176, "step": 150310 }, { "grad_norm": 0.2443007081747055, "learning_rate": 5.3953723661155e-05, "loss": 0.0128, "step": 150320 }, { "grad_norm": 0.08834056556224823, "learning_rate": 5.394822932939067e-05, "loss": 0.0156, "step": 150330 }, { "grad_norm": 0.09057930111885071, "learning_rate": 5.3942734949651606e-05, "loss": 0.0202, "step": 150340 }, { "grad_norm": 0.3057422935962677, "learning_rate": 5.3937240522004605e-05, "loss": 0.0127, "step": 150350 }, { "grad_norm": 0.11300088465213776, "learning_rate": 5.393174604651641e-05, "loss": 0.014, "step": 150360 }, { "grad_norm": 0.07521892338991165, "learning_rate": 5.392625152325378e-05, "loss": 0.0169, "step": 150370 }, { "grad_norm": 0.10507959127426147, "learning_rate": 5.39207569522835e-05, "loss": 0.0222, "step": 150380 }, { "grad_norm": 0.08666915446519852, "learning_rate": 5.391526233367231e-05, "loss": 0.013, "step": 150390 }, { "grad_norm": 0.13154692947864532, "learning_rate": 5.3909767667487e-05, "loss": 0.0175, "step": 150400 }, { "grad_norm": 0.0932016596198082, "learning_rate": 5.3904272953794296e-05, "loss": 0.0156, "step": 150410 }, { "grad_norm": 0.07969705760478973, "learning_rate": 5.389877819266102e-05, "loss": 0.0183, "step": 150420 }, { "grad_norm": 0.09803079068660736, "learning_rate": 5.389328338415388e-05, "loss": 0.0168, "step": 150430 }, { "grad_norm": 0.07809735834598541, "learning_rate": 5.38877885283397e-05, "loss": 0.0135, "step": 150440 }, { "grad_norm": 0.27835747599601746, "learning_rate": 5.3882293625285195e-05, "loss": 0.0103, "step": 150450 }, { "grad_norm": 0.09181277453899384, "learning_rate": 5.387679867505717e-05, "loss": 0.0157, "step": 150460 }, { "grad_norm": 0.07977498322725296, "learning_rate": 5.387130367772237e-05, "loss": 0.0117, "step": 150470 }, { "grad_norm": 0.10263369232416153, "learning_rate": 5.386580863334759e-05, "loss": 0.0174, "step": 150480 }, { "grad_norm": 0.11643073707818985, "learning_rate": 5.386031354199956e-05, "loss": 0.0164, "step": 150490 }, { "grad_norm": 0.14757953584194183, "learning_rate": 5.385481840374511e-05, "loss": 0.0121, "step": 150500 }, { "grad_norm": 0.08464988321065903, "learning_rate": 5.3849323218650946e-05, "loss": 0.0123, "step": 150510 }, { "grad_norm": 0.1375347226858139, "learning_rate": 5.384382798678389e-05, "loss": 0.0168, "step": 150520 }, { "grad_norm": 0.09667763859033585, "learning_rate": 5.383833270821068e-05, "loss": 0.018, "step": 150530 }, { "grad_norm": 0.1091034859418869, "learning_rate": 5.383283738299812e-05, "loss": 0.013, "step": 150540 }, { "grad_norm": 0.06998929381370544, "learning_rate": 5.3827342011212955e-05, "loss": 0.0151, "step": 150550 }, { "grad_norm": 0.11250150203704834, "learning_rate": 5.382184659292198e-05, "loss": 0.017, "step": 150560 }, { "grad_norm": 0.14614750444889069, "learning_rate": 5.381635112819195e-05, "loss": 0.0169, "step": 150570 }, { "grad_norm": 0.14321501553058624, "learning_rate": 5.381085561708966e-05, "loss": 0.0167, "step": 150580 }, { "grad_norm": 0.08651648461818695, "learning_rate": 5.380536005968186e-05, "loss": 0.0102, "step": 150590 }, { "grad_norm": 0.08173774927854538, "learning_rate": 5.3799864456035364e-05, "loss": 0.0141, "step": 150600 }, { "grad_norm": 0.09965354204177856, "learning_rate": 5.379436880621693e-05, "loss": 0.0178, "step": 150610 }, { "grad_norm": 0.08752793073654175, "learning_rate": 5.3788873110293314e-05, "loss": 0.0186, "step": 150620 }, { "grad_norm": 0.060848359018564224, "learning_rate": 5.378337736833132e-05, "loss": 0.0159, "step": 150630 }, { "grad_norm": 0.0762697383761406, "learning_rate": 5.3777881580397716e-05, "loss": 0.0159, "step": 150640 }, { "grad_norm": 0.09945079684257507, "learning_rate": 5.3772385746559294e-05, "loss": 0.0116, "step": 150650 }, { "grad_norm": 0.17551666498184204, "learning_rate": 5.376688986688281e-05, "loss": 0.0121, "step": 150660 }, { "grad_norm": 0.11305833607912064, "learning_rate": 5.3761393941435065e-05, "loss": 0.0143, "step": 150670 }, { "grad_norm": 0.1918969303369522, "learning_rate": 5.375589797028282e-05, "loss": 0.0214, "step": 150680 }, { "grad_norm": 0.09741517901420593, "learning_rate": 5.375040195349289e-05, "loss": 0.0188, "step": 150690 }, { "grad_norm": 0.18709057569503784, "learning_rate": 5.3744905891132025e-05, "loss": 0.0145, "step": 150700 }, { "grad_norm": 0.10406074672937393, "learning_rate": 5.373940978326701e-05, "loss": 0.0154, "step": 150710 }, { "grad_norm": 0.1268071085214615, "learning_rate": 5.373391362996465e-05, "loss": 0.0178, "step": 150720 }, { "grad_norm": 0.11501162499189377, "learning_rate": 5.37284174312917e-05, "loss": 0.0205, "step": 150730 }, { "grad_norm": 0.13631120324134827, "learning_rate": 5.3722921187314964e-05, "loss": 0.016, "step": 150740 }, { "grad_norm": 0.06358236074447632, "learning_rate": 5.371742489810122e-05, "loss": 0.0116, "step": 150750 }, { "grad_norm": 0.057058077305555344, "learning_rate": 5.3711928563717265e-05, "loss": 0.0139, "step": 150760 }, { "grad_norm": 0.11728749424219131, "learning_rate": 5.370643218422986e-05, "loss": 0.0117, "step": 150770 }, { "grad_norm": 0.1107531264424324, "learning_rate": 5.3700935759705804e-05, "loss": 0.0147, "step": 150780 }, { "grad_norm": 0.0735827088356018, "learning_rate": 5.369543929021189e-05, "loss": 0.0138, "step": 150790 }, { "grad_norm": 0.14093975722789764, "learning_rate": 5.3689942775814895e-05, "loss": 0.02, "step": 150800 }, { "grad_norm": 0.10638023912906647, "learning_rate": 5.368444621658162e-05, "loss": 0.0118, "step": 150810 }, { "grad_norm": 0.23869137465953827, "learning_rate": 5.367894961257882e-05, "loss": 0.015, "step": 150820 }, { "grad_norm": 0.10176016390323639, "learning_rate": 5.367345296387334e-05, "loss": 0.0149, "step": 150830 }, { "grad_norm": 0.10231173783540726, "learning_rate": 5.3667956270531916e-05, "loss": 0.0131, "step": 150840 }, { "grad_norm": 0.11044229567050934, "learning_rate": 5.3662459532621376e-05, "loss": 0.0142, "step": 150850 }, { "grad_norm": 0.14676398038864136, "learning_rate": 5.36569627502085e-05, "loss": 0.0179, "step": 150860 }, { "grad_norm": 0.31704822182655334, "learning_rate": 5.3651465923360045e-05, "loss": 0.0109, "step": 150870 }, { "grad_norm": 0.07836078107357025, "learning_rate": 5.364596905214285e-05, "loss": 0.0114, "step": 150880 }, { "grad_norm": 0.09477763622999191, "learning_rate": 5.364047213662369e-05, "loss": 0.0198, "step": 150890 }, { "grad_norm": 0.11078593879938126, "learning_rate": 5.363497517686935e-05, "loss": 0.0144, "step": 150900 }, { "grad_norm": 0.38135451078414917, "learning_rate": 5.362947817294664e-05, "loss": 0.0167, "step": 150910 }, { "grad_norm": 0.236544668674469, "learning_rate": 5.362398112492233e-05, "loss": 0.0161, "step": 150920 }, { "grad_norm": 0.09174078702926636, "learning_rate": 5.361848403286325e-05, "loss": 0.0133, "step": 150930 }, { "grad_norm": 0.09989449381828308, "learning_rate": 5.3612986896836145e-05, "loss": 0.0133, "step": 150940 }, { "grad_norm": 0.08863251656293869, "learning_rate": 5.360748971690786e-05, "loss": 0.0133, "step": 150950 }, { "grad_norm": 0.0569840744137764, "learning_rate": 5.360199249314515e-05, "loss": 0.0121, "step": 150960 }, { "grad_norm": 0.14015652239322662, "learning_rate": 5.359649522561484e-05, "loss": 0.0139, "step": 150970 }, { "grad_norm": 0.08611161261796951, "learning_rate": 5.359099791438371e-05, "loss": 0.0117, "step": 150980 }, { "grad_norm": 0.15410256385803223, "learning_rate": 5.3585500559518574e-05, "loss": 0.0147, "step": 150990 }, { "grad_norm": 0.13803288340568542, "learning_rate": 5.358000316108621e-05, "loss": 0.0111, "step": 151000 }, { "grad_norm": 0.08518091589212418, "learning_rate": 5.357450571915344e-05, "loss": 0.016, "step": 151010 }, { "grad_norm": 0.12584534287452698, "learning_rate": 5.356900823378705e-05, "loss": 0.0143, "step": 151020 }, { "grad_norm": 0.19639329612255096, "learning_rate": 5.356351070505383e-05, "loss": 0.0152, "step": 151030 }, { "grad_norm": 0.1365232616662979, "learning_rate": 5.3558013133020616e-05, "loss": 0.0211, "step": 151040 }, { "grad_norm": 0.08742634207010269, "learning_rate": 5.3552515517754155e-05, "loss": 0.0152, "step": 151050 }, { "grad_norm": 0.19013889133930206, "learning_rate": 5.354701785932129e-05, "loss": 0.0146, "step": 151060 }, { "grad_norm": 0.2546428442001343, "learning_rate": 5.354152015778879e-05, "loss": 0.0158, "step": 151070 }, { "grad_norm": 0.07083118706941605, "learning_rate": 5.3536022413223496e-05, "loss": 0.0166, "step": 151080 }, { "grad_norm": 0.09549345076084137, "learning_rate": 5.353052462569219e-05, "loss": 0.0141, "step": 151090 }, { "grad_norm": 0.08211251348257065, "learning_rate": 5.352502679526167e-05, "loss": 0.0176, "step": 151100 }, { "grad_norm": 0.10457319766283035, "learning_rate": 5.351952892199875e-05, "loss": 0.0188, "step": 151110 }, { "grad_norm": 0.12513306736946106, "learning_rate": 5.3514031005970235e-05, "loss": 0.0181, "step": 151120 }, { "grad_norm": 0.09420440346002579, "learning_rate": 5.350853304724291e-05, "loss": 0.017, "step": 151130 }, { "grad_norm": 0.20392198860645294, "learning_rate": 5.350303504588361e-05, "loss": 0.0186, "step": 151140 }, { "grad_norm": 0.22084417939186096, "learning_rate": 5.349753700195912e-05, "loss": 0.0167, "step": 151150 }, { "grad_norm": 0.13170339167118073, "learning_rate": 5.3492038915536255e-05, "loss": 0.0158, "step": 151160 }, { "grad_norm": 0.12577983736991882, "learning_rate": 5.348654078668183e-05, "loss": 0.0197, "step": 151170 }, { "grad_norm": 0.14897920191287994, "learning_rate": 5.3481042615462617e-05, "loss": 0.0128, "step": 151180 }, { "grad_norm": 0.12749910354614258, "learning_rate": 5.3475544401945465e-05, "loss": 0.0168, "step": 151190 }, { "grad_norm": 0.13635605573654175, "learning_rate": 5.3470046146197176e-05, "loss": 0.0189, "step": 151200 }, { "grad_norm": 0.11793992668390274, "learning_rate": 5.346454784828453e-05, "loss": 0.0136, "step": 151210 }, { "grad_norm": 0.0992022454738617, "learning_rate": 5.345904950827437e-05, "loss": 0.0141, "step": 151220 }, { "grad_norm": 0.15732984244823456, "learning_rate": 5.3453551126233495e-05, "loss": 0.012, "step": 151230 }, { "grad_norm": 0.08020619302988052, "learning_rate": 5.344805270222871e-05, "loss": 0.0169, "step": 151240 }, { "grad_norm": 0.10960473120212555, "learning_rate": 5.344255423632684e-05, "loss": 0.0133, "step": 151250 }, { "grad_norm": 0.15129157900810242, "learning_rate": 5.343705572859466e-05, "loss": 0.0209, "step": 151260 }, { "grad_norm": 0.08604533225297928, "learning_rate": 5.343155717909903e-05, "loss": 0.0122, "step": 151270 }, { "grad_norm": 0.09029286354780197, "learning_rate": 5.342605858790672e-05, "loss": 0.0124, "step": 151280 }, { "grad_norm": 0.0903453528881073, "learning_rate": 5.342055995508458e-05, "loss": 0.021, "step": 151290 }, { "grad_norm": 0.19534847140312195, "learning_rate": 5.341506128069941e-05, "loss": 0.0152, "step": 151300 }, { "grad_norm": 0.11693231761455536, "learning_rate": 5.3409562564818016e-05, "loss": 0.0124, "step": 151310 }, { "grad_norm": 0.14114652574062347, "learning_rate": 5.340406380750721e-05, "loss": 0.014, "step": 151320 }, { "grad_norm": 0.18723158538341522, "learning_rate": 5.339856500883382e-05, "loss": 0.0139, "step": 151330 }, { "grad_norm": 0.07835934311151505, "learning_rate": 5.339306616886466e-05, "loss": 0.0179, "step": 151340 }, { "grad_norm": 0.2346334308385849, "learning_rate": 5.338756728766654e-05, "loss": 0.0212, "step": 151350 }, { "grad_norm": 0.1663263589143753, "learning_rate": 5.338206836530628e-05, "loss": 0.014, "step": 151360 }, { "grad_norm": 0.5092107653617859, "learning_rate": 5.337656940185069e-05, "loss": 0.0198, "step": 151370 }, { "grad_norm": 0.1397888958454132, "learning_rate": 5.3371070397366594e-05, "loss": 0.0152, "step": 151380 }, { "grad_norm": 0.14654043316841125, "learning_rate": 5.336557135192082e-05, "loss": 0.0128, "step": 151390 }, { "grad_norm": 0.1458146870136261, "learning_rate": 5.336007226558016e-05, "loss": 0.0147, "step": 151400 }, { "grad_norm": 0.08210175484418869, "learning_rate": 5.335457313841148e-05, "loss": 0.0149, "step": 151410 }, { "grad_norm": 0.06144769862294197, "learning_rate": 5.3349073970481535e-05, "loss": 0.023, "step": 151420 }, { "grad_norm": 0.09521718323230743, "learning_rate": 5.3343574761857195e-05, "loss": 0.015, "step": 151430 }, { "grad_norm": 0.4238465130329132, "learning_rate": 5.333807551260526e-05, "loss": 0.0167, "step": 151440 }, { "grad_norm": 0.18802735209465027, "learning_rate": 5.333257622279256e-05, "loss": 0.0143, "step": 151450 }, { "grad_norm": 0.09142498672008514, "learning_rate": 5.332707689248591e-05, "loss": 0.0189, "step": 151460 }, { "grad_norm": 0.07890629768371582, "learning_rate": 5.332157752175214e-05, "loss": 0.0157, "step": 151470 }, { "grad_norm": 0.1044994592666626, "learning_rate": 5.331607811065806e-05, "loss": 0.0161, "step": 151480 }, { "grad_norm": 0.098338782787323, "learning_rate": 5.331057865927051e-05, "loss": 0.0135, "step": 151490 }, { "grad_norm": 0.08953644335269928, "learning_rate": 5.33050791676563e-05, "loss": 0.0188, "step": 151500 }, { "grad_norm": 0.13621732592582703, "learning_rate": 5.3299579635882246e-05, "loss": 0.0163, "step": 151510 }, { "grad_norm": 0.14434292912483215, "learning_rate": 5.329408006401519e-05, "loss": 0.017, "step": 151520 }, { "grad_norm": 0.135265052318573, "learning_rate": 5.328858045212195e-05, "loss": 0.0162, "step": 151530 }, { "grad_norm": 0.12702006101608276, "learning_rate": 5.328308080026936e-05, "loss": 0.0176, "step": 151540 }, { "grad_norm": 0.1008491963148117, "learning_rate": 5.3277581108524234e-05, "loss": 0.0196, "step": 151550 }, { "grad_norm": 0.0918256863951683, "learning_rate": 5.327208137695341e-05, "loss": 0.0151, "step": 151560 }, { "grad_norm": 0.09818849712610245, "learning_rate": 5.326658160562369e-05, "loss": 0.0164, "step": 151570 }, { "grad_norm": 0.08796454966068268, "learning_rate": 5.326108179460194e-05, "loss": 0.0175, "step": 151580 }, { "grad_norm": 0.31091806292533875, "learning_rate": 5.325558194395496e-05, "loss": 0.0171, "step": 151590 }, { "grad_norm": 0.11923656612634659, "learning_rate": 5.325008205374957e-05, "loss": 0.0159, "step": 151600 }, { "grad_norm": 0.09291182458400726, "learning_rate": 5.324458212405264e-05, "loss": 0.0153, "step": 151610 }, { "grad_norm": 0.21893338859081268, "learning_rate": 5.323908215493095e-05, "loss": 0.0114, "step": 151620 }, { "grad_norm": 0.1137666255235672, "learning_rate": 5.3233582146451375e-05, "loss": 0.0127, "step": 151630 }, { "grad_norm": 0.15487168729305267, "learning_rate": 5.322808209868071e-05, "loss": 0.0172, "step": 151640 }, { "grad_norm": 0.12312621623277664, "learning_rate": 5.322258201168581e-05, "loss": 0.0169, "step": 151650 }, { "grad_norm": 0.1489683836698532, "learning_rate": 5.321708188553349e-05, "loss": 0.0158, "step": 151660 }, { "grad_norm": 0.06025518849492073, "learning_rate": 5.321158172029058e-05, "loss": 0.0108, "step": 151670 }, { "grad_norm": 0.39907562732696533, "learning_rate": 5.320608151602393e-05, "loss": 0.012, "step": 151680 }, { "grad_norm": 0.1265852004289627, "learning_rate": 5.3200581272800366e-05, "loss": 0.0172, "step": 151690 }, { "grad_norm": 0.12808138132095337, "learning_rate": 5.3195080990686706e-05, "loss": 0.0122, "step": 151700 }, { "grad_norm": 0.12378499656915665, "learning_rate": 5.3189580669749793e-05, "loss": 0.0192, "step": 151710 }, { "grad_norm": 0.10373418033123016, "learning_rate": 5.318408031005648e-05, "loss": 0.014, "step": 151720 }, { "grad_norm": 0.0781085193157196, "learning_rate": 5.317857991167358e-05, "loss": 0.0127, "step": 151730 }, { "grad_norm": 0.12247279286384583, "learning_rate": 5.317307947466793e-05, "loss": 0.0127, "step": 151740 }, { "grad_norm": 0.2528548538684845, "learning_rate": 5.316757899910637e-05, "loss": 0.0194, "step": 151750 }, { "grad_norm": 0.16230694949626923, "learning_rate": 5.3162078485055725e-05, "loss": 0.0156, "step": 151760 }, { "grad_norm": 0.13586845993995667, "learning_rate": 5.315657793258285e-05, "loss": 0.0121, "step": 151770 }, { "grad_norm": 0.15281660854816437, "learning_rate": 5.3151077341754576e-05, "loss": 0.0182, "step": 151780 }, { "grad_norm": 0.11298371851444244, "learning_rate": 5.3145576712637735e-05, "loss": 0.0125, "step": 151790 }, { "grad_norm": 0.11079557240009308, "learning_rate": 5.314007604529915e-05, "loss": 0.0131, "step": 151800 }, { "grad_norm": 0.39672785997390747, "learning_rate": 5.31345753398057e-05, "loss": 0.0205, "step": 151810 }, { "grad_norm": 0.1331707239151001, "learning_rate": 5.312907459622418e-05, "loss": 0.0167, "step": 151820 }, { "grad_norm": 0.15843185782432556, "learning_rate": 5.312357381462146e-05, "loss": 0.0238, "step": 151830 }, { "grad_norm": 0.1309862732887268, "learning_rate": 5.3118072995064375e-05, "loss": 0.0158, "step": 151840 }, { "grad_norm": 0.0662728026509285, "learning_rate": 5.3112572137619754e-05, "loss": 0.019, "step": 151850 }, { "grad_norm": 0.1201958879828453, "learning_rate": 5.3107071242354435e-05, "loss": 0.0126, "step": 151860 }, { "grad_norm": 0.1631220281124115, "learning_rate": 5.310157030933528e-05, "loss": 0.0204, "step": 151870 }, { "grad_norm": 0.09017764776945114, "learning_rate": 5.3096069338629105e-05, "loss": 0.0168, "step": 151880 }, { "grad_norm": 0.32765981554985046, "learning_rate": 5.309056833030277e-05, "loss": 0.0191, "step": 151890 }, { "grad_norm": 0.16251777112483978, "learning_rate": 5.308506728442311e-05, "loss": 0.015, "step": 151900 }, { "grad_norm": 0.22924946248531342, "learning_rate": 5.307956620105697e-05, "loss": 0.0163, "step": 151910 }, { "grad_norm": 0.08543560653924942, "learning_rate": 5.3074065080271196e-05, "loss": 0.0196, "step": 151920 }, { "grad_norm": 0.13248172402381897, "learning_rate": 5.306856392213263e-05, "loss": 0.0169, "step": 151930 }, { "grad_norm": 0.08011269569396973, "learning_rate": 5.3063062726708115e-05, "loss": 0.0139, "step": 151940 }, { "grad_norm": 0.051640912890434265, "learning_rate": 5.305756149406449e-05, "loss": 0.0118, "step": 151950 }, { "grad_norm": 0.10449332743883133, "learning_rate": 5.3052060224268606e-05, "loss": 0.0254, "step": 151960 }, { "grad_norm": 0.20581795275211334, "learning_rate": 5.304655891738732e-05, "loss": 0.0177, "step": 151970 }, { "grad_norm": 0.1639934629201889, "learning_rate": 5.3041057573487464e-05, "loss": 0.0173, "step": 151980 }, { "grad_norm": 0.14935818314552307, "learning_rate": 5.303555619263587e-05, "loss": 0.0165, "step": 151990 }, { "grad_norm": 0.13919739425182343, "learning_rate": 5.303005477489943e-05, "loss": 0.0172, "step": 152000 }, { "grad_norm": 0.14678578078746796, "learning_rate": 5.302455332034494e-05, "loss": 0.0221, "step": 152010 }, { "grad_norm": 0.17497527599334717, "learning_rate": 5.301905182903928e-05, "loss": 0.0148, "step": 152020 }, { "grad_norm": 0.08334311097860336, "learning_rate": 5.3013550301049286e-05, "loss": 0.0172, "step": 152030 }, { "grad_norm": 0.20017345249652863, "learning_rate": 5.3008048736441814e-05, "loss": 0.0133, "step": 152040 }, { "grad_norm": 0.19424372911453247, "learning_rate": 5.3002547135283706e-05, "loss": 0.0193, "step": 152050 }, { "grad_norm": 0.2052268236875534, "learning_rate": 5.299704549764183e-05, "loss": 0.0193, "step": 152060 }, { "grad_norm": 0.8860464096069336, "learning_rate": 5.299154382358301e-05, "loss": 0.0162, "step": 152070 }, { "grad_norm": 0.191798135638237, "learning_rate": 5.2986042113174104e-05, "loss": 0.0171, "step": 152080 }, { "grad_norm": 0.12318574637174606, "learning_rate": 5.298054036648198e-05, "loss": 0.0183, "step": 152090 }, { "grad_norm": 0.10787923634052277, "learning_rate": 5.297503858357346e-05, "loss": 0.0126, "step": 152100 }, { "grad_norm": 0.11687487363815308, "learning_rate": 5.2969536764515425e-05, "loss": 0.0225, "step": 152110 }, { "grad_norm": 0.17958104610443115, "learning_rate": 5.2964034909374704e-05, "loss": 0.0186, "step": 152120 }, { "grad_norm": 0.09857156127691269, "learning_rate": 5.295853301821817e-05, "loss": 0.0175, "step": 152130 }, { "grad_norm": 0.2509155869483948, "learning_rate": 5.295303109111267e-05, "loss": 0.0178, "step": 152140 }, { "grad_norm": 0.06165466085076332, "learning_rate": 5.2947529128125036e-05, "loss": 0.0114, "step": 152150 }, { "grad_norm": 0.13249367475509644, "learning_rate": 5.294202712932217e-05, "loss": 0.0144, "step": 152160 }, { "grad_norm": 0.2257525622844696, "learning_rate": 5.293652509477087e-05, "loss": 0.0145, "step": 152170 }, { "grad_norm": 0.07276974618434906, "learning_rate": 5.293102302453803e-05, "loss": 0.0174, "step": 152180 }, { "grad_norm": 0.27634933590888977, "learning_rate": 5.2925520918690486e-05, "loss": 0.0143, "step": 152190 }, { "grad_norm": 0.09816053509712219, "learning_rate": 5.29200187772951e-05, "loss": 0.0144, "step": 152200 }, { "grad_norm": 0.08571749180555344, "learning_rate": 5.291451660041873e-05, "loss": 0.0095, "step": 152210 }, { "grad_norm": 0.16679883003234863, "learning_rate": 5.290901438812824e-05, "loss": 0.0282, "step": 152220 }, { "grad_norm": 0.24212580919265747, "learning_rate": 5.2903512140490476e-05, "loss": 0.018, "step": 152230 }, { "grad_norm": 0.09944948554039001, "learning_rate": 5.2898009857572285e-05, "loss": 0.0162, "step": 152240 }, { "grad_norm": 0.1616537868976593, "learning_rate": 5.289250753944055e-05, "loss": 0.0166, "step": 152250 }, { "grad_norm": 0.1481146663427353, "learning_rate": 5.288700518616212e-05, "loss": 0.0179, "step": 152260 }, { "grad_norm": 0.14075779914855957, "learning_rate": 5.288150279780385e-05, "loss": 0.0125, "step": 152270 }, { "grad_norm": 0.09848514944314957, "learning_rate": 5.287600037443259e-05, "loss": 0.0142, "step": 152280 }, { "grad_norm": 0.09069666266441345, "learning_rate": 5.287049791611521e-05, "loss": 0.0133, "step": 152290 }, { "grad_norm": 0.10347218066453934, "learning_rate": 5.2864995422918586e-05, "loss": 0.0174, "step": 152300 }, { "grad_norm": 0.12124628573656082, "learning_rate": 5.285949289490955e-05, "loss": 0.0207, "step": 152310 }, { "grad_norm": 0.10278545320034027, "learning_rate": 5.285399033215499e-05, "loss": 0.0126, "step": 152320 }, { "grad_norm": 0.14539003372192383, "learning_rate": 5.284848773472173e-05, "loss": 0.0182, "step": 152330 }, { "grad_norm": 0.07273095846176147, "learning_rate": 5.284298510267667e-05, "loss": 0.0112, "step": 152340 }, { "grad_norm": 0.12195876985788345, "learning_rate": 5.283748243608665e-05, "loss": 0.021, "step": 152350 }, { "grad_norm": 0.09389834105968475, "learning_rate": 5.2831979735018547e-05, "loss": 0.0156, "step": 152360 }, { "grad_norm": 0.10959893465042114, "learning_rate": 5.282647699953921e-05, "loss": 0.0132, "step": 152370 }, { "grad_norm": 0.21562781929969788, "learning_rate": 5.2820974229715515e-05, "loss": 0.015, "step": 152380 }, { "grad_norm": 0.0875542163848877, "learning_rate": 5.2815471425614313e-05, "loss": 0.0092, "step": 152390 }, { "grad_norm": 0.11980888247489929, "learning_rate": 5.2809968587302474e-05, "loss": 0.0174, "step": 152400 }, { "grad_norm": 0.17437604069709778, "learning_rate": 5.280446571484687e-05, "loss": 0.0135, "step": 152410 }, { "grad_norm": 0.1281406581401825, "learning_rate": 5.2798962808314356e-05, "loss": 0.0132, "step": 152420 }, { "grad_norm": 0.10309693217277527, "learning_rate": 5.27934598677718e-05, "loss": 0.0319, "step": 152430 }, { "grad_norm": 0.13607992231845856, "learning_rate": 5.278795689328607e-05, "loss": 0.0112, "step": 152440 }, { "grad_norm": 0.11236272007226944, "learning_rate": 5.2782453884924034e-05, "loss": 0.0124, "step": 152450 }, { "grad_norm": 0.3340766429901123, "learning_rate": 5.277695084275256e-05, "loss": 0.0171, "step": 152460 }, { "grad_norm": 0.13815182447433472, "learning_rate": 5.27714477668385e-05, "loss": 0.0125, "step": 152470 }, { "grad_norm": 0.21310186386108398, "learning_rate": 5.276594465724875e-05, "loss": 0.0172, "step": 152480 }, { "grad_norm": 0.09545807540416718, "learning_rate": 5.276044151405014e-05, "loss": 0.0224, "step": 152490 }, { "grad_norm": 0.07655700296163559, "learning_rate": 5.2754938337309576e-05, "loss": 0.0118, "step": 152500 }, { "grad_norm": 0.24882245063781738, "learning_rate": 5.2749435127093895e-05, "loss": 0.0241, "step": 152510 }, { "grad_norm": 0.08553605526685715, "learning_rate": 5.274393188347e-05, "loss": 0.0177, "step": 152520 }, { "grad_norm": 0.07990024983882904, "learning_rate": 5.273842860650472e-05, "loss": 0.0154, "step": 152530 }, { "grad_norm": 0.11268875747919083, "learning_rate": 5.2732925296264965e-05, "loss": 0.0175, "step": 152540 }, { "grad_norm": 0.134487122297287, "learning_rate": 5.272742195281758e-05, "loss": 0.0143, "step": 152550 }, { "grad_norm": 0.1312326192855835, "learning_rate": 5.272191857622944e-05, "loss": 0.0169, "step": 152560 }, { "grad_norm": 0.07627318799495697, "learning_rate": 5.2716415166567426e-05, "loss": 0.0162, "step": 152570 }, { "grad_norm": 0.10261909663677216, "learning_rate": 5.2710911723898406e-05, "loss": 0.0124, "step": 152580 }, { "grad_norm": 0.09514786303043365, "learning_rate": 5.270540824828924e-05, "loss": 0.0162, "step": 152590 }, { "grad_norm": 0.14961741864681244, "learning_rate": 5.2699904739806814e-05, "loss": 0.0164, "step": 152600 }, { "grad_norm": 0.12186122685670853, "learning_rate": 5.2694401198517985e-05, "loss": 0.0135, "step": 152610 }, { "grad_norm": 0.06668910384178162, "learning_rate": 5.268889762448965e-05, "loss": 0.015, "step": 152620 }, { "grad_norm": 0.06433824449777603, "learning_rate": 5.268339401778867e-05, "loss": 0.0126, "step": 152630 }, { "grad_norm": 0.0719415619969368, "learning_rate": 5.267789037848192e-05, "loss": 0.0151, "step": 152640 }, { "grad_norm": 0.15307338535785675, "learning_rate": 5.2672386706636256e-05, "loss": 0.0104, "step": 152650 }, { "grad_norm": 0.15752562880516052, "learning_rate": 5.26668830023186e-05, "loss": 0.0166, "step": 152660 }, { "grad_norm": 0.11613895744085312, "learning_rate": 5.2661379265595766e-05, "loss": 0.0102, "step": 152670 }, { "grad_norm": 0.07971271872520447, "learning_rate": 5.265587549653469e-05, "loss": 0.015, "step": 152680 }, { "grad_norm": 0.13829024136066437, "learning_rate": 5.26503716952022e-05, "loss": 0.0108, "step": 152690 }, { "grad_norm": 0.09824438393115997, "learning_rate": 5.26448678616652e-05, "loss": 0.0154, "step": 152700 }, { "grad_norm": 0.09642525762319565, "learning_rate": 5.2639363995990564e-05, "loss": 0.0149, "step": 152710 }, { "grad_norm": 0.12600141763687134, "learning_rate": 5.2633860098245144e-05, "loss": 0.0251, "step": 152720 }, { "grad_norm": 0.21711353957653046, "learning_rate": 5.2628356168495865e-05, "loss": 0.0175, "step": 152730 }, { "grad_norm": 0.09085221588611603, "learning_rate": 5.262285220680955e-05, "loss": 0.013, "step": 152740 }, { "grad_norm": 0.08318129926919937, "learning_rate": 5.261734821325313e-05, "loss": 0.0145, "step": 152750 }, { "grad_norm": 0.13649626076221466, "learning_rate": 5.261184418789343e-05, "loss": 0.0163, "step": 152760 }, { "grad_norm": 0.17958687245845795, "learning_rate": 5.260634013079737e-05, "loss": 0.0189, "step": 152770 }, { "grad_norm": 0.08055032789707184, "learning_rate": 5.260083604203182e-05, "loss": 0.0137, "step": 152780 }, { "grad_norm": 0.12693960964679718, "learning_rate": 5.259533192166366e-05, "loss": 0.0109, "step": 152790 }, { "grad_norm": 0.14079035818576813, "learning_rate": 5.258982776975976e-05, "loss": 0.0152, "step": 152800 }, { "grad_norm": 0.16906499862670898, "learning_rate": 5.2584323586387e-05, "loss": 0.0203, "step": 152810 }, { "grad_norm": 0.1268940567970276, "learning_rate": 5.2578819371612284e-05, "loss": 0.0147, "step": 152820 }, { "grad_norm": 0.08534939587116241, "learning_rate": 5.257331512550247e-05, "loss": 0.0179, "step": 152830 }, { "grad_norm": 0.22035621106624603, "learning_rate": 5.256781084812444e-05, "loss": 0.0184, "step": 152840 }, { "grad_norm": 0.13759160041809082, "learning_rate": 5.256230653954509e-05, "loss": 0.017, "step": 152850 }, { "grad_norm": 0.11562369763851166, "learning_rate": 5.25568021998313e-05, "loss": 0.0158, "step": 152860 }, { "grad_norm": 0.11861096322536469, "learning_rate": 5.255129782904996e-05, "loss": 0.0117, "step": 152870 }, { "grad_norm": 0.05680383741855621, "learning_rate": 5.254579342726792e-05, "loss": 0.0101, "step": 152880 }, { "grad_norm": 0.07627981901168823, "learning_rate": 5.25402889945521e-05, "loss": 0.0161, "step": 152890 }, { "grad_norm": 0.135771706700325, "learning_rate": 5.253478453096936e-05, "loss": 0.0125, "step": 152900 }, { "grad_norm": 0.10293816775083542, "learning_rate": 5.25292800365866e-05, "loss": 0.0111, "step": 152910 }, { "grad_norm": 0.14364799857139587, "learning_rate": 5.25237755114707e-05, "loss": 0.0137, "step": 152920 }, { "grad_norm": 0.2338544875383377, "learning_rate": 5.251827095568855e-05, "loss": 0.0158, "step": 152930 }, { "grad_norm": 0.39611825346946716, "learning_rate": 5.251276636930702e-05, "loss": 0.0183, "step": 152940 }, { "grad_norm": 0.10840430110692978, "learning_rate": 5.250726175239302e-05, "loss": 0.016, "step": 152950 }, { "grad_norm": 0.07745418697595596, "learning_rate": 5.250175710501342e-05, "loss": 0.0131, "step": 152960 }, { "grad_norm": 0.20201057195663452, "learning_rate": 5.2496252427235093e-05, "loss": 0.0132, "step": 152970 }, { "grad_norm": 0.09673323482275009, "learning_rate": 5.249074771912495e-05, "loss": 0.0118, "step": 152980 }, { "grad_norm": 0.16606810688972473, "learning_rate": 5.248524298074987e-05, "loss": 0.0133, "step": 152990 }, { "grad_norm": 0.10886639356613159, "learning_rate": 5.2479738212176755e-05, "loss": 0.0154, "step": 153000 }, { "grad_norm": 0.12523384392261505, "learning_rate": 5.247423341347246e-05, "loss": 0.0143, "step": 153010 }, { "grad_norm": 0.08716604113578796, "learning_rate": 5.246872858470391e-05, "loss": 0.0155, "step": 153020 }, { "grad_norm": 0.14790047705173492, "learning_rate": 5.246322372593797e-05, "loss": 0.0167, "step": 153030 }, { "grad_norm": 0.19167223572731018, "learning_rate": 5.245771883724153e-05, "loss": 0.0233, "step": 153040 }, { "grad_norm": 0.13591548800468445, "learning_rate": 5.2452213918681494e-05, "loss": 0.011, "step": 153050 }, { "grad_norm": 0.5803980231285095, "learning_rate": 5.2446708970324734e-05, "loss": 0.016, "step": 153060 }, { "grad_norm": 0.12745261192321777, "learning_rate": 5.244120399223816e-05, "loss": 0.012, "step": 153070 }, { "grad_norm": 0.3570973575115204, "learning_rate": 5.2435698984488645e-05, "loss": 0.0191, "step": 153080 }, { "grad_norm": 0.06443952769041061, "learning_rate": 5.243019394714309e-05, "loss": 0.014, "step": 153090 }, { "grad_norm": 0.07019776850938797, "learning_rate": 5.242468888026838e-05, "loss": 0.0147, "step": 153100 }, { "grad_norm": 0.1181524470448494, "learning_rate": 5.241918378393143e-05, "loss": 0.0192, "step": 153110 }, { "grad_norm": 0.06729315966367722, "learning_rate": 5.24136786581991e-05, "loss": 0.0146, "step": 153120 }, { "grad_norm": 0.10482876747846603, "learning_rate": 5.240817350313828e-05, "loss": 0.0127, "step": 153130 }, { "grad_norm": 0.23332011699676514, "learning_rate": 5.240266831881591e-05, "loss": 0.016, "step": 153140 }, { "grad_norm": 0.09920115023851395, "learning_rate": 5.239716310529882e-05, "loss": 0.0152, "step": 153150 }, { "grad_norm": 0.14466434717178345, "learning_rate": 5.2391657862653953e-05, "loss": 0.0187, "step": 153160 }, { "grad_norm": 0.09362883865833282, "learning_rate": 5.238615259094818e-05, "loss": 0.0161, "step": 153170 }, { "grad_norm": 0.10298386216163635, "learning_rate": 5.2380647290248396e-05, "loss": 0.0095, "step": 153180 }, { "grad_norm": 0.07186304777860641, "learning_rate": 5.237514196062151e-05, "loss": 0.0151, "step": 153190 }, { "grad_norm": 0.14719437062740326, "learning_rate": 5.2369636602134397e-05, "loss": 0.0154, "step": 153200 }, { "grad_norm": 0.13837340474128723, "learning_rate": 5.236413121485397e-05, "loss": 0.0145, "step": 153210 }, { "grad_norm": 0.09962540119886398, "learning_rate": 5.2358625798847105e-05, "loss": 0.012, "step": 153220 }, { "grad_norm": 0.20022451877593994, "learning_rate": 5.235312035418073e-05, "loss": 0.0136, "step": 153230 }, { "grad_norm": 0.11666645109653473, "learning_rate": 5.2347614880921704e-05, "loss": 0.0144, "step": 153240 }, { "grad_norm": 0.08297380059957504, "learning_rate": 5.234210937913695e-05, "loss": 0.0107, "step": 153250 }, { "grad_norm": 0.06501177698373795, "learning_rate": 5.233660384889334e-05, "loss": 0.0125, "step": 153260 }, { "grad_norm": 0.08993778377771378, "learning_rate": 5.233109829025781e-05, "loss": 0.0144, "step": 153270 }, { "grad_norm": 0.07808135449886322, "learning_rate": 5.232559270329721e-05, "loss": 0.0208, "step": 153280 }, { "grad_norm": 0.09402422606945038, "learning_rate": 5.232008708807848e-05, "loss": 0.0126, "step": 153290 }, { "grad_norm": 0.07856597751379013, "learning_rate": 5.231458144466849e-05, "loss": 0.0133, "step": 153300 }, { "grad_norm": 0.3751954138278961, "learning_rate": 5.230907577313416e-05, "loss": 0.0169, "step": 153310 }, { "grad_norm": 0.10640888661146164, "learning_rate": 5.230357007354237e-05, "loss": 0.0131, "step": 153320 }, { "grad_norm": 0.10229373723268509, "learning_rate": 5.229806434596004e-05, "loss": 0.0157, "step": 153330 }, { "grad_norm": 0.06905501335859299, "learning_rate": 5.229255859045406e-05, "loss": 0.0109, "step": 153340 }, { "grad_norm": 0.12437703460454941, "learning_rate": 5.228705280709132e-05, "loss": 0.0194, "step": 153350 }, { "grad_norm": 0.13027937710285187, "learning_rate": 5.228154699593873e-05, "loss": 0.0126, "step": 153360 }, { "grad_norm": 0.13468322157859802, "learning_rate": 5.2276041157063196e-05, "loss": 0.0159, "step": 153370 }, { "grad_norm": 0.11076437681913376, "learning_rate": 5.22705352905316e-05, "loss": 0.0218, "step": 153380 }, { "grad_norm": 0.08688879758119583, "learning_rate": 5.226502939641087e-05, "loss": 0.0173, "step": 153390 }, { "grad_norm": 0.194171741604805, "learning_rate": 5.225952347476788e-05, "loss": 0.0185, "step": 153400 }, { "grad_norm": 0.10241594910621643, "learning_rate": 5.225401752566956e-05, "loss": 0.0137, "step": 153410 }, { "grad_norm": 0.12472289800643921, "learning_rate": 5.2248511549182786e-05, "loss": 0.0136, "step": 153420 }, { "grad_norm": 0.15952876210212708, "learning_rate": 5.224300554537449e-05, "loss": 0.0133, "step": 153430 }, { "grad_norm": 0.0627143383026123, "learning_rate": 5.223749951431155e-05, "loss": 0.0158, "step": 153440 }, { "grad_norm": 0.07344695180654526, "learning_rate": 5.223199345606088e-05, "loss": 0.0134, "step": 153450 }, { "grad_norm": 0.1986064314842224, "learning_rate": 5.222648737068938e-05, "loss": 0.0158, "step": 153460 }, { "grad_norm": 0.08666630834341049, "learning_rate": 5.222098125826396e-05, "loss": 0.011, "step": 153470 }, { "grad_norm": 0.08427216112613678, "learning_rate": 5.221547511885152e-05, "loss": 0.0148, "step": 153480 }, { "grad_norm": 0.2165200412273407, "learning_rate": 5.220996895251895e-05, "loss": 0.0187, "step": 153490 }, { "grad_norm": 0.09064082056283951, "learning_rate": 5.2204462759333196e-05, "loss": 0.0162, "step": 153500 }, { "grad_norm": 0.09101345390081406, "learning_rate": 5.219895653936113e-05, "loss": 0.0111, "step": 153510 }, { "grad_norm": 0.28686845302581787, "learning_rate": 5.2193450292669654e-05, "loss": 0.0207, "step": 153520 }, { "grad_norm": 0.22620879113674164, "learning_rate": 5.218794401932571e-05, "loss": 0.0209, "step": 153530 }, { "grad_norm": 0.1596132516860962, "learning_rate": 5.2182437719396146e-05, "loss": 0.0219, "step": 153540 }, { "grad_norm": 0.10665204375982285, "learning_rate": 5.217693139294794e-05, "loss": 0.0153, "step": 153550 }, { "grad_norm": 0.08941168338060379, "learning_rate": 5.217142504004793e-05, "loss": 0.0154, "step": 153560 }, { "grad_norm": 0.12452290952205658, "learning_rate": 5.216591866076308e-05, "loss": 0.0181, "step": 153570 }, { "grad_norm": 0.14819751679897308, "learning_rate": 5.216041225516025e-05, "loss": 0.0139, "step": 153580 }, { "grad_norm": 0.2031726986169815, "learning_rate": 5.215490582330639e-05, "loss": 0.0148, "step": 153590 }, { "grad_norm": 0.10478783398866653, "learning_rate": 5.214939936526839e-05, "loss": 0.0121, "step": 153600 }, { "grad_norm": 0.1881178468465805, "learning_rate": 5.214389288111314e-05, "loss": 0.0171, "step": 153610 }, { "grad_norm": 0.07342567294836044, "learning_rate": 5.21383863709076e-05, "loss": 0.0132, "step": 153620 }, { "grad_norm": 0.09236231446266174, "learning_rate": 5.213287983471862e-05, "loss": 0.0195, "step": 153630 }, { "grad_norm": 0.12423837929964066, "learning_rate": 5.212737327261315e-05, "loss": 0.0171, "step": 153640 }, { "grad_norm": 0.12076068669557571, "learning_rate": 5.212186668465806e-05, "loss": 0.0161, "step": 153650 }, { "grad_norm": 0.09133617579936981, "learning_rate": 5.211636007092031e-05, "loss": 0.0149, "step": 153660 }, { "grad_norm": 0.06004045158624649, "learning_rate": 5.211085343146679e-05, "loss": 0.0125, "step": 153670 }, { "grad_norm": 0.06569675356149673, "learning_rate": 5.210534676636439e-05, "loss": 0.0155, "step": 153680 }, { "grad_norm": 0.07859525084495544, "learning_rate": 5.2099840075680064e-05, "loss": 0.0118, "step": 153690 }, { "grad_norm": 0.09015180915594101, "learning_rate": 5.209433335948067e-05, "loss": 0.0232, "step": 153700 }, { "grad_norm": 0.1437399834394455, "learning_rate": 5.208882661783318e-05, "loss": 0.0159, "step": 153710 }, { "grad_norm": 0.10267327725887299, "learning_rate": 5.208331985080445e-05, "loss": 0.0164, "step": 153720 }, { "grad_norm": 0.08929942548274994, "learning_rate": 5.207781305846142e-05, "loss": 0.0167, "step": 153730 }, { "grad_norm": 0.08433347195386887, "learning_rate": 5.207230624087099e-05, "loss": 0.0108, "step": 153740 }, { "grad_norm": 0.17293542623519897, "learning_rate": 5.206679939810011e-05, "loss": 0.018, "step": 153750 }, { "grad_norm": 0.14774003624916077, "learning_rate": 5.2061292530215656e-05, "loss": 0.013, "step": 153760 }, { "grad_norm": 0.13975198566913605, "learning_rate": 5.205578563728454e-05, "loss": 0.0148, "step": 153770 }, { "grad_norm": 0.08298024535179138, "learning_rate": 5.20502787193737e-05, "loss": 0.0125, "step": 153780 }, { "grad_norm": 0.11479754745960236, "learning_rate": 5.204477177655003e-05, "loss": 0.017, "step": 153790 }, { "grad_norm": 0.11643625050783157, "learning_rate": 5.203926480888045e-05, "loss": 0.0133, "step": 153800 }, { "grad_norm": 0.14074446260929108, "learning_rate": 5.203375781643188e-05, "loss": 0.0168, "step": 153810 }, { "grad_norm": 0.1628917008638382, "learning_rate": 5.202825079927124e-05, "loss": 0.0133, "step": 153820 }, { "grad_norm": 0.10656394064426422, "learning_rate": 5.202274375746543e-05, "loss": 0.0098, "step": 153830 }, { "grad_norm": 0.10711158066987991, "learning_rate": 5.2017236691081374e-05, "loss": 0.0122, "step": 153840 }, { "grad_norm": 0.20194442570209503, "learning_rate": 5.2011729600185986e-05, "loss": 0.013, "step": 153850 }, { "grad_norm": 0.11266841739416122, "learning_rate": 5.200622248484619e-05, "loss": 0.0166, "step": 153860 }, { "grad_norm": 0.06615506857633591, "learning_rate": 5.200071534512889e-05, "loss": 0.0131, "step": 153870 }, { "grad_norm": 0.08468074351549149, "learning_rate": 5.199520818110102e-05, "loss": 0.0205, "step": 153880 }, { "grad_norm": 0.09886528551578522, "learning_rate": 5.1989700992829484e-05, "loss": 0.0127, "step": 153890 }, { "grad_norm": 0.05106283724308014, "learning_rate": 5.198419378038118e-05, "loss": 0.0137, "step": 153900 }, { "grad_norm": 0.08636683225631714, "learning_rate": 5.1978686543823076e-05, "loss": 0.0154, "step": 153910 }, { "grad_norm": 0.08928239345550537, "learning_rate": 5.197317928322206e-05, "loss": 0.0097, "step": 153920 }, { "grad_norm": 0.10259271413087845, "learning_rate": 5.196767199864504e-05, "loss": 0.0104, "step": 153930 }, { "grad_norm": 0.1182594746351242, "learning_rate": 5.196216469015895e-05, "loss": 0.0152, "step": 153940 }, { "grad_norm": 0.17478537559509277, "learning_rate": 5.195665735783071e-05, "loss": 0.0176, "step": 153950 }, { "grad_norm": 0.20141421258449554, "learning_rate": 5.1951150001727244e-05, "loss": 0.0127, "step": 153960 }, { "grad_norm": 0.10753463208675385, "learning_rate": 5.194564262191545e-05, "loss": 0.0169, "step": 153970 }, { "grad_norm": 0.10954604297876358, "learning_rate": 5.194013521846226e-05, "loss": 0.0162, "step": 153980 }, { "grad_norm": 0.15166498720645905, "learning_rate": 5.1934627791434595e-05, "loss": 0.029, "step": 153990 }, { "grad_norm": 0.1335342973470688, "learning_rate": 5.192912034089939e-05, "loss": 0.0195, "step": 154000 }, { "grad_norm": 0.10924966633319855, "learning_rate": 5.192361286692354e-05, "loss": 0.023, "step": 154010 }, { "grad_norm": 0.16631321609020233, "learning_rate": 5.191810536957399e-05, "loss": 0.0159, "step": 154020 }, { "grad_norm": 0.07835660874843597, "learning_rate": 5.191259784891763e-05, "loss": 0.0132, "step": 154030 }, { "grad_norm": 0.12099508196115494, "learning_rate": 5.190709030502141e-05, "loss": 0.0122, "step": 154040 }, { "grad_norm": 0.10250890254974365, "learning_rate": 5.1901582737952236e-05, "loss": 0.0123, "step": 154050 }, { "grad_norm": 0.07736462354660034, "learning_rate": 5.1896075147777045e-05, "loss": 0.0136, "step": 154060 }, { "grad_norm": 0.09335549920797348, "learning_rate": 5.189056753456275e-05, "loss": 0.0179, "step": 154070 }, { "grad_norm": 0.12714940309524536, "learning_rate": 5.188505989837627e-05, "loss": 0.0112, "step": 154080 }, { "grad_norm": 0.13971129059791565, "learning_rate": 5.187955223928454e-05, "loss": 0.0159, "step": 154090 }, { "grad_norm": 0.13497890532016754, "learning_rate": 5.1874044557354475e-05, "loss": 0.0105, "step": 154100 }, { "grad_norm": 0.20502229034900665, "learning_rate": 5.186853685265298e-05, "loss": 0.0105, "step": 154110 }, { "grad_norm": 0.061841998249292374, "learning_rate": 5.186302912524702e-05, "loss": 0.0187, "step": 154120 }, { "grad_norm": 0.10702966898679733, "learning_rate": 5.185752137520349e-05, "loss": 0.0158, "step": 154130 }, { "grad_norm": 0.05952597036957741, "learning_rate": 5.185201360258932e-05, "loss": 0.0173, "step": 154140 }, { "grad_norm": 0.15602217614650726, "learning_rate": 5.184650580747143e-05, "loss": 0.0152, "step": 154150 }, { "grad_norm": 0.09395094215869904, "learning_rate": 5.184099798991676e-05, "loss": 0.0165, "step": 154160 }, { "grad_norm": 0.13370849192142487, "learning_rate": 5.1835490149992236e-05, "loss": 0.0157, "step": 154170 }, { "grad_norm": 0.08349261432886124, "learning_rate": 5.1829982287764754e-05, "loss": 0.0126, "step": 154180 }, { "grad_norm": 0.1683802604675293, "learning_rate": 5.1824474403301274e-05, "loss": 0.011, "step": 154190 }, { "grad_norm": 0.13056515157222748, "learning_rate": 5.18189664966687e-05, "loss": 0.0125, "step": 154200 }, { "grad_norm": 0.15535220503807068, "learning_rate": 5.181345856793397e-05, "loss": 0.0133, "step": 154210 }, { "grad_norm": 0.06567078828811646, "learning_rate": 5.180795061716399e-05, "loss": 0.0151, "step": 154220 }, { "grad_norm": 0.15443553030490875, "learning_rate": 5.180244264442573e-05, "loss": 0.0109, "step": 154230 }, { "grad_norm": 0.1650485247373581, "learning_rate": 5.179693464978608e-05, "loss": 0.0136, "step": 154240 }, { "grad_norm": 0.12195950001478195, "learning_rate": 5.179142663331197e-05, "loss": 0.0156, "step": 154250 }, { "grad_norm": 0.10278505831956863, "learning_rate": 5.1785918595070337e-05, "loss": 0.0179, "step": 154260 }, { "grad_norm": 0.12352548539638519, "learning_rate": 5.17804105351281e-05, "loss": 0.0131, "step": 154270 }, { "grad_norm": 0.21342390775680542, "learning_rate": 5.177490245355221e-05, "loss": 0.019, "step": 154280 }, { "grad_norm": 0.06144294515252113, "learning_rate": 5.176939435040957e-05, "loss": 0.0148, "step": 154290 }, { "grad_norm": 0.05997951328754425, "learning_rate": 5.176388622576712e-05, "loss": 0.0118, "step": 154300 }, { "grad_norm": 0.07051306962966919, "learning_rate": 5.175837807969178e-05, "loss": 0.0152, "step": 154310 }, { "grad_norm": 0.06448874622583389, "learning_rate": 5.17528699122505e-05, "loss": 0.0133, "step": 154320 }, { "grad_norm": 0.14522776007652283, "learning_rate": 5.17473617235102e-05, "loss": 0.0118, "step": 154330 }, { "grad_norm": 0.10134236514568329, "learning_rate": 5.1741853513537785e-05, "loss": 0.0125, "step": 154340 }, { "grad_norm": 0.15704426169395447, "learning_rate": 5.173634528240022e-05, "loss": 0.0213, "step": 154350 }, { "grad_norm": 0.07000929862260818, "learning_rate": 5.1730837030164406e-05, "loss": 0.0131, "step": 154360 }, { "grad_norm": 0.1328677535057068, "learning_rate": 5.1725328756897306e-05, "loss": 0.0158, "step": 154370 }, { "grad_norm": 0.07657411694526672, "learning_rate": 5.1719820462665815e-05, "loss": 0.0122, "step": 154380 }, { "grad_norm": 0.10121425986289978, "learning_rate": 5.171431214753689e-05, "loss": 0.0193, "step": 154390 }, { "grad_norm": 0.13392984867095947, "learning_rate": 5.1708803811577466e-05, "loss": 0.0178, "step": 154400 }, { "grad_norm": 0.1490030437707901, "learning_rate": 5.170329545485444e-05, "loss": 0.0161, "step": 154410 }, { "grad_norm": 0.06530784070491791, "learning_rate": 5.169778707743478e-05, "loss": 0.0161, "step": 154420 }, { "grad_norm": 0.18256893754005432, "learning_rate": 5.16922786793854e-05, "loss": 0.0168, "step": 154430 }, { "grad_norm": 0.2372869849205017, "learning_rate": 5.168677026077324e-05, "loss": 0.0111, "step": 154440 }, { "grad_norm": 0.21594040095806122, "learning_rate": 5.1681261821665214e-05, "loss": 0.0255, "step": 154450 }, { "grad_norm": 0.1582459658384323, "learning_rate": 5.1675753362128285e-05, "loss": 0.0215, "step": 154460 }, { "grad_norm": 0.10272442549467087, "learning_rate": 5.167024488222937e-05, "loss": 0.0151, "step": 154470 }, { "grad_norm": 0.15594661235809326, "learning_rate": 5.166473638203539e-05, "loss": 0.0143, "step": 154480 }, { "grad_norm": 0.0940205454826355, "learning_rate": 5.1659227861613314e-05, "loss": 0.0108, "step": 154490 }, { "grad_norm": 0.10889297723770142, "learning_rate": 5.1653719321030036e-05, "loss": 0.0208, "step": 154500 }, { "grad_norm": 0.10911719501018524, "learning_rate": 5.1648210760352514e-05, "loss": 0.0133, "step": 154510 }, { "grad_norm": 0.1499861627817154, "learning_rate": 5.164270217964766e-05, "loss": 0.0151, "step": 154520 }, { "grad_norm": 0.13116878271102905, "learning_rate": 5.163719357898245e-05, "loss": 0.0153, "step": 154530 }, { "grad_norm": 0.14989492297172546, "learning_rate": 5.163168495842376e-05, "loss": 0.0187, "step": 154540 }, { "grad_norm": 0.12729090452194214, "learning_rate": 5.1626176318038576e-05, "loss": 0.0139, "step": 154550 }, { "grad_norm": 0.07165464013814926, "learning_rate": 5.162066765789382e-05, "loss": 0.0128, "step": 154560 }, { "grad_norm": 0.12235093116760254, "learning_rate": 5.161515897805641e-05, "loss": 0.0153, "step": 154570 }, { "grad_norm": 0.13593602180480957, "learning_rate": 5.16096502785933e-05, "loss": 0.0087, "step": 154580 }, { "grad_norm": 0.18026235699653625, "learning_rate": 5.160414155957142e-05, "loss": 0.0202, "step": 154590 }, { "grad_norm": 0.06569741666316986, "learning_rate": 5.1598632821057705e-05, "loss": 0.0163, "step": 154600 }, { "grad_norm": 0.1325611174106598, "learning_rate": 5.159312406311908e-05, "loss": 0.0136, "step": 154610 }, { "grad_norm": 0.14499329030513763, "learning_rate": 5.158761528582251e-05, "loss": 0.0152, "step": 154620 }, { "grad_norm": 0.1556313931941986, "learning_rate": 5.158210648923489e-05, "loss": 0.0142, "step": 154630 }, { "grad_norm": 0.06484642624855042, "learning_rate": 5.157659767342321e-05, "loss": 0.0124, "step": 154640 }, { "grad_norm": 0.12428215891122818, "learning_rate": 5.157108883845437e-05, "loss": 0.0158, "step": 154650 }, { "grad_norm": 0.08651730418205261, "learning_rate": 5.156557998439532e-05, "loss": 0.0165, "step": 154660 }, { "grad_norm": 0.07022877037525177, "learning_rate": 5.1560071111313e-05, "loss": 0.0144, "step": 154670 }, { "grad_norm": 0.09752435237169266, "learning_rate": 5.1554562219274325e-05, "loss": 0.0121, "step": 154680 }, { "grad_norm": 0.113207146525383, "learning_rate": 5.1549053308346273e-05, "loss": 0.0106, "step": 154690 }, { "grad_norm": 0.2590562105178833, "learning_rate": 5.154354437859574e-05, "loss": 0.017, "step": 154700 }, { "grad_norm": 0.15251073241233826, "learning_rate": 5.15380354300897e-05, "loss": 0.0154, "step": 154710 }, { "grad_norm": 0.1384100466966629, "learning_rate": 5.153252646289507e-05, "loss": 0.0173, "step": 154720 }, { "grad_norm": 0.16256795823574066, "learning_rate": 5.15270174770788e-05, "loss": 0.0121, "step": 154730 }, { "grad_norm": 0.07316306978464127, "learning_rate": 5.152150847270781e-05, "loss": 0.0153, "step": 154740 }, { "grad_norm": 0.08555444329977036, "learning_rate": 5.1515999449849085e-05, "loss": 0.0128, "step": 154750 }, { "grad_norm": 0.09786584973335266, "learning_rate": 5.1510490408569514e-05, "loss": 0.0163, "step": 154760 }, { "grad_norm": 0.18090829253196716, "learning_rate": 5.150498134893607e-05, "loss": 0.0159, "step": 154770 }, { "grad_norm": 0.0983714684844017, "learning_rate": 5.1499472271015666e-05, "loss": 0.0144, "step": 154780 }, { "grad_norm": 0.22969728708267212, "learning_rate": 5.149396317487528e-05, "loss": 0.0183, "step": 154790 }, { "grad_norm": 0.22673848271369934, "learning_rate": 5.148845406058182e-05, "loss": 0.0195, "step": 154800 }, { "grad_norm": 0.08820810168981552, "learning_rate": 5.148294492820224e-05, "loss": 0.0141, "step": 154810 }, { "grad_norm": 0.13854651153087616, "learning_rate": 5.147743577780346e-05, "loss": 0.0135, "step": 154820 }, { "grad_norm": 0.10710911452770233, "learning_rate": 5.147192660945247e-05, "loss": 0.0193, "step": 154830 }, { "grad_norm": 0.2067396491765976, "learning_rate": 5.146641742321616e-05, "loss": 0.0148, "step": 154840 }, { "grad_norm": 0.08179356902837753, "learning_rate": 5.146090821916151e-05, "loss": 0.0164, "step": 154850 }, { "grad_norm": 0.1537964642047882, "learning_rate": 5.1455398997355434e-05, "loss": 0.0194, "step": 154860 }, { "grad_norm": 0.11662694811820984, "learning_rate": 5.144988975786489e-05, "loss": 0.0188, "step": 154870 }, { "grad_norm": 0.11395362764596939, "learning_rate": 5.144438050075683e-05, "loss": 0.0147, "step": 154880 }, { "grad_norm": 0.16084560751914978, "learning_rate": 5.143887122609816e-05, "loss": 0.0185, "step": 154890 }, { "grad_norm": 0.20042182505130768, "learning_rate": 5.143336193395586e-05, "loss": 0.0124, "step": 154900 }, { "grad_norm": 0.09746246784925461, "learning_rate": 5.142785262439685e-05, "loss": 0.0117, "step": 154910 }, { "grad_norm": 0.10974277555942535, "learning_rate": 5.1422343297488096e-05, "loss": 0.0181, "step": 154920 }, { "grad_norm": 0.11335013061761856, "learning_rate": 5.1416833953296506e-05, "loss": 0.0205, "step": 154930 }, { "grad_norm": 0.18090204894542694, "learning_rate": 5.1411324591889065e-05, "loss": 0.0197, "step": 154940 }, { "grad_norm": 0.07955144345760345, "learning_rate": 5.1405815213332684e-05, "loss": 0.0124, "step": 154950 }, { "grad_norm": 0.10614059120416641, "learning_rate": 5.140030581769434e-05, "loss": 0.0188, "step": 154960 }, { "grad_norm": 0.11075881868600845, "learning_rate": 5.139479640504096e-05, "loss": 0.0172, "step": 154970 }, { "grad_norm": 0.0618385411798954, "learning_rate": 5.1389286975439464e-05, "loss": 0.0162, "step": 154980 }, { "grad_norm": 0.06893245875835419, "learning_rate": 5.1383777528956834e-05, "loss": 0.0146, "step": 154990 }, { "grad_norm": 0.1278872936964035, "learning_rate": 5.137826806565998e-05, "loss": 0.015, "step": 155000 }, { "grad_norm": 0.13888534903526306, "learning_rate": 5.13727585856159e-05, "loss": 0.0174, "step": 155010 }, { "grad_norm": 0.13389559090137482, "learning_rate": 5.136724908889148e-05, "loss": 0.0168, "step": 155020 }, { "grad_norm": 0.2712150812149048, "learning_rate": 5.136173957555371e-05, "loss": 0.0204, "step": 155030 }, { "grad_norm": 0.27248212695121765, "learning_rate": 5.135623004566952e-05, "loss": 0.0171, "step": 155040 }, { "grad_norm": 0.537645161151886, "learning_rate": 5.1350720499305835e-05, "loss": 0.0172, "step": 155050 }, { "grad_norm": 0.13056257367134094, "learning_rate": 5.134521093652964e-05, "loss": 0.0189, "step": 155060 }, { "grad_norm": 0.26933804154396057, "learning_rate": 5.133970135740785e-05, "loss": 0.0157, "step": 155070 }, { "grad_norm": 0.4158160090446472, "learning_rate": 5.133419176200743e-05, "loss": 0.0167, "step": 155080 }, { "grad_norm": 0.13808901607990265, "learning_rate": 5.1328682150395314e-05, "loss": 0.0142, "step": 155090 }, { "grad_norm": 0.0752161368727684, "learning_rate": 5.1323172522638465e-05, "loss": 0.0147, "step": 155100 }, { "grad_norm": 0.08136949688196182, "learning_rate": 5.1317662878803806e-05, "loss": 0.0119, "step": 155110 }, { "grad_norm": 0.0832638069987297, "learning_rate": 5.131215321895831e-05, "loss": 0.0287, "step": 155120 }, { "grad_norm": 0.10551346838474274, "learning_rate": 5.130664354316893e-05, "loss": 0.0177, "step": 155130 }, { "grad_norm": 0.08103303611278534, "learning_rate": 5.130113385150257e-05, "loss": 0.0154, "step": 155140 }, { "grad_norm": 0.1420152336359024, "learning_rate": 5.129562414402622e-05, "loss": 0.0184, "step": 155150 }, { "grad_norm": 0.21739336848258972, "learning_rate": 5.12901144208068e-05, "loss": 0.0162, "step": 155160 }, { "grad_norm": 0.1236189529299736, "learning_rate": 5.1284604681911285e-05, "loss": 0.0122, "step": 155170 }, { "grad_norm": 0.0801469087600708, "learning_rate": 5.12790949274066e-05, "loss": 0.0179, "step": 155180 }, { "grad_norm": 0.19912558794021606, "learning_rate": 5.1273585157359716e-05, "loss": 0.0136, "step": 155190 }, { "grad_norm": 0.15770581364631653, "learning_rate": 5.126807537183755e-05, "loss": 0.0145, "step": 155200 }, { "grad_norm": 0.09828092902898788, "learning_rate": 5.1262565570907095e-05, "loss": 0.0138, "step": 155210 }, { "grad_norm": 0.07133636623620987, "learning_rate": 5.1257055754635264e-05, "loss": 0.0153, "step": 155220 }, { "grad_norm": 0.19027996063232422, "learning_rate": 5.125154592308902e-05, "loss": 0.0302, "step": 155230 }, { "grad_norm": 0.11974913626909256, "learning_rate": 5.1246036076335316e-05, "loss": 0.0145, "step": 155240 }, { "grad_norm": 0.12014953047037125, "learning_rate": 5.124052621444108e-05, "loss": 0.0104, "step": 155250 }, { "grad_norm": 0.0733722522854805, "learning_rate": 5.12350163374733e-05, "loss": 0.0138, "step": 155260 }, { "grad_norm": 0.17164911329746246, "learning_rate": 5.122950644549889e-05, "loss": 0.0159, "step": 155270 }, { "grad_norm": 0.2823238968849182, "learning_rate": 5.122399653858483e-05, "loss": 0.0196, "step": 155280 }, { "grad_norm": 0.13334940373897552, "learning_rate": 5.121848661679805e-05, "loss": 0.0178, "step": 155290 }, { "grad_norm": 0.08259464800357819, "learning_rate": 5.1212976680205505e-05, "loss": 0.0142, "step": 155300 }, { "grad_norm": 0.15752747654914856, "learning_rate": 5.1207466728874156e-05, "loss": 0.0147, "step": 155310 }, { "grad_norm": 0.07533631473779678, "learning_rate": 5.120195676287093e-05, "loss": 0.0113, "step": 155320 }, { "grad_norm": 0.11049909889698029, "learning_rate": 5.119644678226281e-05, "loss": 0.0158, "step": 155330 }, { "grad_norm": 0.1067538633942604, "learning_rate": 5.119093678711672e-05, "loss": 0.0125, "step": 155340 }, { "grad_norm": 0.1104951873421669, "learning_rate": 5.118542677749964e-05, "loss": 0.0138, "step": 155350 }, { "grad_norm": 0.1384247988462448, "learning_rate": 5.117991675347849e-05, "loss": 0.0138, "step": 155360 }, { "grad_norm": 0.09264574199914932, "learning_rate": 5.1174406715120236e-05, "loss": 0.0138, "step": 155370 }, { "grad_norm": 0.1469135582447052, "learning_rate": 5.116889666249185e-05, "loss": 0.0126, "step": 155380 }, { "grad_norm": 0.09624073654413223, "learning_rate": 5.116338659566025e-05, "loss": 0.0161, "step": 155390 }, { "grad_norm": 0.08536243438720703, "learning_rate": 5.11578765146924e-05, "loss": 0.0133, "step": 155400 }, { "grad_norm": 0.16956092417240143, "learning_rate": 5.115236641965526e-05, "loss": 0.0155, "step": 155410 }, { "grad_norm": 0.1445063203573227, "learning_rate": 5.11468563106158e-05, "loss": 0.0122, "step": 155420 }, { "grad_norm": 0.170977383852005, "learning_rate": 5.114134618764093e-05, "loss": 0.0173, "step": 155430 }, { "grad_norm": 0.12669691443443298, "learning_rate": 5.1135836050797634e-05, "loss": 0.0148, "step": 155440 }, { "grad_norm": 0.09142539650201797, "learning_rate": 5.113032590015285e-05, "loss": 0.0112, "step": 155450 }, { "grad_norm": 0.22767102718353271, "learning_rate": 5.112481573577356e-05, "loss": 0.0176, "step": 155460 }, { "grad_norm": 0.14042282104492188, "learning_rate": 5.111930555772668e-05, "loss": 0.0165, "step": 155470 }, { "grad_norm": 0.09090536832809448, "learning_rate": 5.111379536607919e-05, "loss": 0.0122, "step": 155480 }, { "grad_norm": 0.08745075017213821, "learning_rate": 5.110828516089802e-05, "loss": 0.0185, "step": 155490 }, { "grad_norm": 0.1347077637910843, "learning_rate": 5.1102774942250154e-05, "loss": 0.0172, "step": 155500 }, { "grad_norm": 0.1634717583656311, "learning_rate": 5.109726471020252e-05, "loss": 0.0154, "step": 155510 }, { "grad_norm": 0.1286703646183014, "learning_rate": 5.1091754464822095e-05, "loss": 0.0112, "step": 155520 }, { "grad_norm": 0.1923484057188034, "learning_rate": 5.1086244206175815e-05, "loss": 0.0168, "step": 155530 }, { "grad_norm": 0.08209358155727386, "learning_rate": 5.108073393433065e-05, "loss": 0.0163, "step": 155540 }, { "grad_norm": 0.0738782063126564, "learning_rate": 5.1075223649353534e-05, "loss": 0.012, "step": 155550 }, { "grad_norm": 0.13125574588775635, "learning_rate": 5.106971335131144e-05, "loss": 0.014, "step": 155560 }, { "grad_norm": 0.15905164182186127, "learning_rate": 5.106420304027132e-05, "loss": 0.0195, "step": 155570 }, { "grad_norm": 0.069176584482193, "learning_rate": 5.105869271630014e-05, "loss": 0.0142, "step": 155580 }, { "grad_norm": 0.1098523885011673, "learning_rate": 5.105318237946482e-05, "loss": 0.0188, "step": 155590 }, { "grad_norm": 0.1229640394449234, "learning_rate": 5.1047672029832364e-05, "loss": 0.0142, "step": 155600 }, { "grad_norm": 0.12760311365127563, "learning_rate": 5.10421616674697e-05, "loss": 0.0117, "step": 155610 }, { "grad_norm": 0.267068475484848, "learning_rate": 5.1036651292443774e-05, "loss": 0.0116, "step": 155620 }, { "grad_norm": 0.13188138604164124, "learning_rate": 5.103114090482156e-05, "loss": 0.013, "step": 155630 }, { "grad_norm": 0.08076154440641403, "learning_rate": 5.102563050467002e-05, "loss": 0.0133, "step": 155640 }, { "grad_norm": 0.11174396425485611, "learning_rate": 5.102012009205609e-05, "loss": 0.0162, "step": 155650 }, { "grad_norm": 0.10962902009487152, "learning_rate": 5.101460966704674e-05, "loss": 0.0149, "step": 155660 }, { "grad_norm": 0.07763482630252838, "learning_rate": 5.1009099229708935e-05, "loss": 0.0137, "step": 155670 }, { "grad_norm": 0.08482719957828522, "learning_rate": 5.10035887801096e-05, "loss": 0.0125, "step": 155680 }, { "grad_norm": 0.06654250621795654, "learning_rate": 5.0998078318315736e-05, "loss": 0.0126, "step": 155690 }, { "grad_norm": 0.07708597183227539, "learning_rate": 5.099256784439428e-05, "loss": 0.0122, "step": 155700 }, { "grad_norm": 0.17218360304832458, "learning_rate": 5.0987057358412174e-05, "loss": 0.0153, "step": 155710 }, { "grad_norm": 0.15328264236450195, "learning_rate": 5.098154686043639e-05, "loss": 0.0223, "step": 155720 }, { "grad_norm": 0.1585853099822998, "learning_rate": 5.0976036350533885e-05, "loss": 0.0169, "step": 155730 }, { "grad_norm": 0.07536037266254425, "learning_rate": 5.097052582877162e-05, "loss": 0.0152, "step": 155740 }, { "grad_norm": 0.1214260682463646, "learning_rate": 5.096501529521655e-05, "loss": 0.0128, "step": 155750 }, { "grad_norm": 0.12939713895320892, "learning_rate": 5.0959504749935626e-05, "loss": 0.0151, "step": 155760 }, { "grad_norm": 0.1621655970811844, "learning_rate": 5.095399419299584e-05, "loss": 0.0121, "step": 155770 }, { "grad_norm": 0.10691821575164795, "learning_rate": 5.0948483624464085e-05, "loss": 0.013, "step": 155780 }, { "grad_norm": 0.23857365548610687, "learning_rate": 5.0942973044407386e-05, "loss": 0.0194, "step": 155790 }, { "grad_norm": 0.08082393556833267, "learning_rate": 5.093746245289266e-05, "loss": 0.0164, "step": 155800 }, { "grad_norm": 0.08831904828548431, "learning_rate": 5.093195184998689e-05, "loss": 0.0127, "step": 155810 }, { "grad_norm": 0.16040660440921783, "learning_rate": 5.0926441235757004e-05, "loss": 0.0119, "step": 155820 }, { "grad_norm": 0.11952187120914459, "learning_rate": 5.092093061027001e-05, "loss": 0.0163, "step": 155830 }, { "grad_norm": 0.11214961856603622, "learning_rate": 5.0915419973592825e-05, "loss": 0.0146, "step": 155840 }, { "grad_norm": 0.08070733398199081, "learning_rate": 5.090990932579243e-05, "loss": 0.0152, "step": 155850 }, { "grad_norm": 0.19493548572063446, "learning_rate": 5.090439866693577e-05, "loss": 0.015, "step": 155860 }, { "grad_norm": 0.04752594232559204, "learning_rate": 5.0898887997089814e-05, "loss": 0.0132, "step": 155870 }, { "grad_norm": 0.27071982622146606, "learning_rate": 5.0893377316321534e-05, "loss": 0.0153, "step": 155880 }, { "grad_norm": 0.07974924892187119, "learning_rate": 5.0887866624697856e-05, "loss": 0.0119, "step": 155890 }, { "grad_norm": 0.057219285517930984, "learning_rate": 5.0882355922285776e-05, "loss": 0.0161, "step": 155900 }, { "grad_norm": 0.1768416315317154, "learning_rate": 5.087684520915221e-05, "loss": 0.0175, "step": 155910 }, { "grad_norm": 0.07641292363405228, "learning_rate": 5.0871334485364185e-05, "loss": 0.0148, "step": 155920 }, { "grad_norm": 0.06663710623979568, "learning_rate": 5.0865823750988606e-05, "loss": 0.0114, "step": 155930 }, { "grad_norm": 0.06135397031903267, "learning_rate": 5.0860313006092444e-05, "loss": 0.0152, "step": 155940 }, { "grad_norm": 0.14584758877754211, "learning_rate": 5.085480225074268e-05, "loss": 0.0178, "step": 155950 }, { "grad_norm": 0.09718872606754303, "learning_rate": 5.0849291485006244e-05, "loss": 0.0148, "step": 155960 }, { "grad_norm": 0.10804114490747452, "learning_rate": 5.084378070895014e-05, "loss": 0.0119, "step": 155970 }, { "grad_norm": 0.4692486822605133, "learning_rate": 5.083826992264128e-05, "loss": 0.0133, "step": 155980 }, { "grad_norm": 0.07858887314796448, "learning_rate": 5.0832759126146665e-05, "loss": 0.0114, "step": 155990 }, { "grad_norm": 0.08481350541114807, "learning_rate": 5.082724831953323e-05, "loss": 0.0118, "step": 156000 }, { "grad_norm": 0.16341416537761688, "learning_rate": 5.082173750286795e-05, "loss": 0.0211, "step": 156010 }, { "grad_norm": 0.0937533900141716, "learning_rate": 5.081622667621779e-05, "loss": 0.0103, "step": 156020 }, { "grad_norm": 0.13907769322395325, "learning_rate": 5.081071583964969e-05, "loss": 0.0173, "step": 156030 }, { "grad_norm": 0.08436891436576843, "learning_rate": 5.080520499323065e-05, "loss": 0.0159, "step": 156040 }, { "grad_norm": 0.16923430562019348, "learning_rate": 5.079969413702759e-05, "loss": 0.0146, "step": 156050 }, { "grad_norm": 0.196259543299675, "learning_rate": 5.07941832711075e-05, "loss": 0.0148, "step": 156060 }, { "grad_norm": 0.2474023401737213, "learning_rate": 5.0788672395537316e-05, "loss": 0.0095, "step": 156070 }, { "grad_norm": 0.08413099497556686, "learning_rate": 5.078316151038404e-05, "loss": 0.0136, "step": 156080 }, { "grad_norm": 0.08069055527448654, "learning_rate": 5.077765061571461e-05, "loss": 0.0161, "step": 156090 }, { "grad_norm": 0.12343244254589081, "learning_rate": 5.077213971159598e-05, "loss": 0.0148, "step": 156100 }, { "grad_norm": 0.18218877911567688, "learning_rate": 5.076662879809513e-05, "loss": 0.0178, "step": 156110 }, { "grad_norm": 0.054948221892118454, "learning_rate": 5.0761117875279007e-05, "loss": 0.0124, "step": 156120 }, { "grad_norm": 0.1166505217552185, "learning_rate": 5.0755606943214604e-05, "loss": 0.0154, "step": 156130 }, { "grad_norm": 0.15387320518493652, "learning_rate": 5.075009600196884e-05, "loss": 0.0107, "step": 156140 }, { "grad_norm": 0.10793402045965195, "learning_rate": 5.0744585051608716e-05, "loss": 0.0115, "step": 156150 }, { "grad_norm": 0.13799220323562622, "learning_rate": 5.073907409220117e-05, "loss": 0.0153, "step": 156160 }, { "grad_norm": 0.15957336127758026, "learning_rate": 5.073356312381319e-05, "loss": 0.0203, "step": 156170 }, { "grad_norm": 0.15226635336875916, "learning_rate": 5.0728052146511704e-05, "loss": 0.0177, "step": 156180 }, { "grad_norm": 0.07642250508069992, "learning_rate": 5.0722541160363726e-05, "loss": 0.0129, "step": 156190 }, { "grad_norm": 0.13650387525558472, "learning_rate": 5.0717030165436164e-05, "loss": 0.0167, "step": 156200 }, { "grad_norm": 0.16581310331821442, "learning_rate": 5.071151916179603e-05, "loss": 0.015, "step": 156210 }, { "grad_norm": 0.1524878889322281, "learning_rate": 5.070600814951024e-05, "loss": 0.0139, "step": 156220 }, { "grad_norm": 0.08264154195785522, "learning_rate": 5.0700497128645804e-05, "loss": 0.01, "step": 156230 }, { "grad_norm": 0.1260799914598465, "learning_rate": 5.069498609926967e-05, "loss": 0.0172, "step": 156240 }, { "grad_norm": 0.21516326069831848, "learning_rate": 5.0689475061448786e-05, "loss": 0.0176, "step": 156250 }, { "grad_norm": 0.20249205827713013, "learning_rate": 5.068396401525014e-05, "loss": 0.0113, "step": 156260 }, { "grad_norm": 0.06772587448358536, "learning_rate": 5.0678452960740684e-05, "loss": 0.0102, "step": 156270 }, { "grad_norm": 0.07502169907093048, "learning_rate": 5.067294189798737e-05, "loss": 0.0103, "step": 156280 }, { "grad_norm": 0.09016861766576767, "learning_rate": 5.06674308270572e-05, "loss": 0.0143, "step": 156290 }, { "grad_norm": 0.14484372735023499, "learning_rate": 5.066191974801708e-05, "loss": 0.013, "step": 156300 }, { "grad_norm": 0.4776536524295807, "learning_rate": 5.065640866093404e-05, "loss": 0.0187, "step": 156310 }, { "grad_norm": 0.12240827083587646, "learning_rate": 5.065089756587501e-05, "loss": 0.016, "step": 156320 }, { "grad_norm": 0.13647116720676422, "learning_rate": 5.0645386462906965e-05, "loss": 0.0145, "step": 156330 }, { "grad_norm": 0.16517309844493866, "learning_rate": 5.063987535209685e-05, "loss": 0.0121, "step": 156340 }, { "grad_norm": 0.09201209247112274, "learning_rate": 5.0634364233511655e-05, "loss": 0.0185, "step": 156350 }, { "grad_norm": 0.2856767773628235, "learning_rate": 5.0628853107218336e-05, "loss": 0.0183, "step": 156360 }, { "grad_norm": 0.2128065526485443, "learning_rate": 5.062334197328386e-05, "loss": 0.0146, "step": 156370 }, { "grad_norm": 0.08842888474464417, "learning_rate": 5.0617830831775184e-05, "loss": 0.0114, "step": 156380 }, { "grad_norm": 0.1623016744852066, "learning_rate": 5.061231968275927e-05, "loss": 0.0183, "step": 156390 }, { "grad_norm": 0.15222081542015076, "learning_rate": 5.0606808526303115e-05, "loss": 0.0104, "step": 156400 }, { "grad_norm": 0.14784225821495056, "learning_rate": 5.0601297362473646e-05, "loss": 0.0198, "step": 156410 }, { "grad_norm": 0.18562136590480804, "learning_rate": 5.0595786191337854e-05, "loss": 0.0173, "step": 156420 }, { "grad_norm": 0.14812953770160675, "learning_rate": 5.059027501296271e-05, "loss": 0.0133, "step": 156430 }, { "grad_norm": 0.11553843319416046, "learning_rate": 5.058476382741515e-05, "loss": 0.0203, "step": 156440 }, { "grad_norm": 0.13866014778614044, "learning_rate": 5.057925263476216e-05, "loss": 0.0151, "step": 156450 }, { "grad_norm": 0.29285648465156555, "learning_rate": 5.0573741435070707e-05, "loss": 0.0177, "step": 156460 }, { "grad_norm": 0.07615531980991364, "learning_rate": 5.056823022840775e-05, "loss": 0.0178, "step": 156470 }, { "grad_norm": 0.06734439730644226, "learning_rate": 5.056271901484026e-05, "loss": 0.012, "step": 156480 }, { "grad_norm": 0.13995422422885895, "learning_rate": 5.0557207794435213e-05, "loss": 0.0173, "step": 156490 }, { "grad_norm": 0.12410502135753632, "learning_rate": 5.0551696567259566e-05, "loss": 0.0168, "step": 156500 }, { "grad_norm": 0.14313729107379913, "learning_rate": 5.054618533338027e-05, "loss": 0.0126, "step": 156510 }, { "grad_norm": 0.08898133784532547, "learning_rate": 5.054067409286431e-05, "loss": 0.0189, "step": 156520 }, { "grad_norm": 0.14905893802642822, "learning_rate": 5.053516284577865e-05, "loss": 0.0124, "step": 156530 }, { "grad_norm": 0.16462966799736023, "learning_rate": 5.052965159219026e-05, "loss": 0.0238, "step": 156540 }, { "grad_norm": 0.12188166379928589, "learning_rate": 5.0524140332166094e-05, "loss": 0.0152, "step": 156550 }, { "grad_norm": 0.24404415488243103, "learning_rate": 5.051862906577314e-05, "loss": 0.0112, "step": 156560 }, { "grad_norm": 0.16275587677955627, "learning_rate": 5.051311779307835e-05, "loss": 0.0124, "step": 156570 }, { "grad_norm": 0.10484062135219574, "learning_rate": 5.050760651414869e-05, "loss": 0.0143, "step": 156580 }, { "grad_norm": 0.09564241766929626, "learning_rate": 5.050209522905114e-05, "loss": 0.0141, "step": 156590 }, { "grad_norm": 0.061835721135139465, "learning_rate": 5.049658393785265e-05, "loss": 0.0157, "step": 156600 }, { "grad_norm": 0.08451510220766068, "learning_rate": 5.04910726406202e-05, "loss": 0.0166, "step": 156610 }, { "grad_norm": 0.08094117045402527, "learning_rate": 5.048556133742075e-05, "loss": 0.0144, "step": 156620 }, { "grad_norm": 0.15077827870845795, "learning_rate": 5.048005002832128e-05, "loss": 0.0187, "step": 156630 }, { "grad_norm": 0.3928839862346649, "learning_rate": 5.047453871338873e-05, "loss": 0.0138, "step": 156640 }, { "grad_norm": 0.11309415847063065, "learning_rate": 5.046902739269011e-05, "loss": 0.0212, "step": 156650 }, { "grad_norm": 0.12008526921272278, "learning_rate": 5.046351606629236e-05, "loss": 0.0181, "step": 156660 }, { "grad_norm": 0.1403588354587555, "learning_rate": 5.0458004734262445e-05, "loss": 0.0165, "step": 156670 }, { "grad_norm": 0.4539472162723541, "learning_rate": 5.0452493396667345e-05, "loss": 0.0174, "step": 156680 }, { "grad_norm": 0.1295393854379654, "learning_rate": 5.044698205357401e-05, "loss": 0.0117, "step": 156690 }, { "grad_norm": 0.08875777572393417, "learning_rate": 5.0441470705049435e-05, "loss": 0.012, "step": 156700 }, { "grad_norm": 0.10720917582511902, "learning_rate": 5.0435959351160565e-05, "loss": 0.015, "step": 156710 }, { "grad_norm": 0.10173407942056656, "learning_rate": 5.0430447991974385e-05, "loss": 0.0151, "step": 156720 }, { "grad_norm": 0.3875148892402649, "learning_rate": 5.042493662755785e-05, "loss": 0.0097, "step": 156730 }, { "grad_norm": 0.2759803235530853, "learning_rate": 5.041942525797794e-05, "loss": 0.0178, "step": 156740 }, { "grad_norm": 0.1475081592798233, "learning_rate": 5.041391388330161e-05, "loss": 0.017, "step": 156750 }, { "grad_norm": 0.18356956541538239, "learning_rate": 5.0408402503595845e-05, "loss": 0.0161, "step": 156760 }, { "grad_norm": 0.099421925842762, "learning_rate": 5.04028911189276e-05, "loss": 0.0141, "step": 156770 }, { "grad_norm": 0.1560952365398407, "learning_rate": 5.039737972936384e-05, "loss": 0.0135, "step": 156780 }, { "grad_norm": 0.1308957040309906, "learning_rate": 5.0391868334971546e-05, "loss": 0.0137, "step": 156790 }, { "grad_norm": 0.09339924901723862, "learning_rate": 5.038635693581768e-05, "loss": 0.0179, "step": 156800 }, { "grad_norm": 0.07375016063451767, "learning_rate": 5.038084553196921e-05, "loss": 0.0126, "step": 156810 }, { "grad_norm": 0.14597663283348083, "learning_rate": 5.0375334123493126e-05, "loss": 0.0134, "step": 156820 }, { "grad_norm": 0.085727758705616, "learning_rate": 5.036982271045635e-05, "loss": 0.0115, "step": 156830 }, { "grad_norm": 0.30742886662483215, "learning_rate": 5.0364311292925906e-05, "loss": 0.0161, "step": 156840 }, { "grad_norm": 0.13441872596740723, "learning_rate": 5.035879987096871e-05, "loss": 0.0192, "step": 156850 }, { "grad_norm": 0.07804377377033234, "learning_rate": 5.035328844465178e-05, "loss": 0.0207, "step": 156860 }, { "grad_norm": 0.27774763107299805, "learning_rate": 5.034777701404204e-05, "loss": 0.0203, "step": 156870 }, { "grad_norm": 0.09927504509687424, "learning_rate": 5.03422655792065e-05, "loss": 0.0119, "step": 156880 }, { "grad_norm": 0.2972140908241272, "learning_rate": 5.03367541402121e-05, "loss": 0.0165, "step": 156890 }, { "grad_norm": 0.08638945966959, "learning_rate": 5.033124269712583e-05, "loss": 0.0167, "step": 156900 }, { "grad_norm": 0.122479148209095, "learning_rate": 5.032573125001463e-05, "loss": 0.0154, "step": 156910 }, { "grad_norm": 0.1486314833164215, "learning_rate": 5.03202197989455e-05, "loss": 0.0158, "step": 156920 }, { "grad_norm": 0.18325701355934143, "learning_rate": 5.031470834398538e-05, "loss": 0.0168, "step": 156930 }, { "grad_norm": 0.13667163252830505, "learning_rate": 5.030919688520128e-05, "loss": 0.0189, "step": 156940 }, { "grad_norm": 0.11693636327981949, "learning_rate": 5.0303685422660127e-05, "loss": 0.0131, "step": 156950 }, { "grad_norm": 0.09316379576921463, "learning_rate": 5.029817395642893e-05, "loss": 0.0203, "step": 156960 }, { "grad_norm": 0.07224850356578827, "learning_rate": 5.0292662486574624e-05, "loss": 0.0236, "step": 156970 }, { "grad_norm": 0.18344101309776306, "learning_rate": 5.02871510131642e-05, "loss": 0.0156, "step": 156980 }, { "grad_norm": 0.11846131831407547, "learning_rate": 5.028163953626461e-05, "loss": 0.0135, "step": 156990 }, { "grad_norm": 0.1322011500597, "learning_rate": 5.027612805594284e-05, "loss": 0.019, "step": 157000 }, { "grad_norm": 0.12189602106809616, "learning_rate": 5.027061657226585e-05, "loss": 0.0132, "step": 157010 }, { "grad_norm": 0.1227070689201355, "learning_rate": 5.0265105085300624e-05, "loss": 0.0144, "step": 157020 }, { "grad_norm": 0.25075551867485046, "learning_rate": 5.025959359511411e-05, "loss": 0.0154, "step": 157030 }, { "grad_norm": 0.1428881734609604, "learning_rate": 5.0254082101773294e-05, "loss": 0.0111, "step": 157040 }, { "grad_norm": 0.4432365596294403, "learning_rate": 5.0248570605345137e-05, "loss": 0.0149, "step": 157050 }, { "grad_norm": 0.07009067386388779, "learning_rate": 5.0243059105896616e-05, "loss": 0.013, "step": 157060 }, { "grad_norm": 0.16135339438915253, "learning_rate": 5.0237547603494695e-05, "loss": 0.0141, "step": 157070 }, { "grad_norm": 0.12608349323272705, "learning_rate": 5.023203609820635e-05, "loss": 0.0139, "step": 157080 }, { "grad_norm": 0.11860748380422592, "learning_rate": 5.0226524590098544e-05, "loss": 0.0127, "step": 157090 }, { "grad_norm": 0.10317521542310715, "learning_rate": 5.0221013079238246e-05, "loss": 0.015, "step": 157100 }, { "grad_norm": 0.18769246339797974, "learning_rate": 5.0215501565692434e-05, "loss": 0.0129, "step": 157110 }, { "grad_norm": 0.1141408160328865, "learning_rate": 5.0209990049528064e-05, "loss": 0.011, "step": 157120 }, { "grad_norm": 0.2722535729408264, "learning_rate": 5.020447853081214e-05, "loss": 0.0211, "step": 157130 }, { "grad_norm": 0.16143620014190674, "learning_rate": 5.0198967009611596e-05, "loss": 0.0166, "step": 157140 }, { "grad_norm": 0.1596352458000183, "learning_rate": 5.019345548599341e-05, "loss": 0.0133, "step": 157150 }, { "grad_norm": 0.14736635982990265, "learning_rate": 5.018794396002457e-05, "loss": 0.0213, "step": 157160 }, { "grad_norm": 0.1882055103778839, "learning_rate": 5.018243243177202e-05, "loss": 0.0157, "step": 157170 }, { "grad_norm": 0.11034747958183289, "learning_rate": 5.017692090130276e-05, "loss": 0.0148, "step": 157180 }, { "grad_norm": 0.07293616980314255, "learning_rate": 5.017140936868373e-05, "loss": 0.0154, "step": 157190 }, { "grad_norm": 0.12016903609037399, "learning_rate": 5.016589783398192e-05, "loss": 0.0154, "step": 157200 }, { "grad_norm": 0.0722426101565361, "learning_rate": 5.0160386297264285e-05, "loss": 0.0133, "step": 157210 }, { "grad_norm": 0.1157994419336319, "learning_rate": 5.0154874758597814e-05, "loss": 0.0155, "step": 157220 }, { "grad_norm": 0.08901607245206833, "learning_rate": 5.014936321804947e-05, "loss": 0.0129, "step": 157230 }, { "grad_norm": 0.07370409369468689, "learning_rate": 5.014385167568622e-05, "loss": 0.0149, "step": 157240 }, { "grad_norm": 0.0945078507065773, "learning_rate": 5.0138340131575045e-05, "loss": 0.0124, "step": 157250 }, { "grad_norm": 0.30922409892082214, "learning_rate": 5.0132828585782886e-05, "loss": 0.0209, "step": 157260 }, { "grad_norm": 0.10024122148752213, "learning_rate": 5.0127317038376756e-05, "loss": 0.0178, "step": 157270 }, { "grad_norm": 0.11766966432332993, "learning_rate": 5.012180548942358e-05, "loss": 0.0145, "step": 157280 }, { "grad_norm": 0.10470899939537048, "learning_rate": 5.0116293938990375e-05, "loss": 0.0138, "step": 157290 }, { "grad_norm": 0.10603491961956024, "learning_rate": 5.0110782387144085e-05, "loss": 0.0155, "step": 157300 }, { "grad_norm": 0.12521253526210785, "learning_rate": 5.010527083395167e-05, "loss": 0.0144, "step": 157310 }, { "grad_norm": 0.06885499507188797, "learning_rate": 5.009975927948013e-05, "loss": 0.0139, "step": 157320 }, { "grad_norm": 0.1484794169664383, "learning_rate": 5.0094247723796403e-05, "loss": 0.0218, "step": 157330 }, { "grad_norm": 0.14001941680908203, "learning_rate": 5.00887361669675e-05, "loss": 0.0191, "step": 157340 }, { "grad_norm": 0.10734976828098297, "learning_rate": 5.0083224609060356e-05, "loss": 0.0201, "step": 157350 }, { "grad_norm": 0.10934934765100479, "learning_rate": 5.0077713050141964e-05, "loss": 0.0149, "step": 157360 }, { "grad_norm": 0.10552393645048141, "learning_rate": 5.0072201490279266e-05, "loss": 0.0097, "step": 157370 }, { "grad_norm": 0.0805220678448677, "learning_rate": 5.0066689929539265e-05, "loss": 0.0145, "step": 157380 }, { "grad_norm": 0.12160088121891022, "learning_rate": 5.0061178367988925e-05, "loss": 0.017, "step": 157390 }, { "grad_norm": 0.09655290842056274, "learning_rate": 5.005566680569519e-05, "loss": 0.0124, "step": 157400 }, { "grad_norm": 0.1629292219877243, "learning_rate": 5.005015524272507e-05, "loss": 0.0148, "step": 157410 }, { "grad_norm": 0.09117494523525238, "learning_rate": 5.00446436791455e-05, "loss": 0.0149, "step": 157420 }, { "grad_norm": 0.07757781445980072, "learning_rate": 5.003913211502348e-05, "loss": 0.0123, "step": 157430 }, { "grad_norm": 0.11327559500932693, "learning_rate": 5.0033620550425966e-05, "loss": 0.0115, "step": 157440 }, { "grad_norm": 0.18714340031147003, "learning_rate": 5.002810898541993e-05, "loss": 0.0157, "step": 157450 }, { "grad_norm": 0.049916304647922516, "learning_rate": 5.002259742007234e-05, "loss": 0.0117, "step": 157460 }, { "grad_norm": 0.1338634490966797, "learning_rate": 5.001708585445016e-05, "loss": 0.0178, "step": 157470 }, { "grad_norm": 0.05188620090484619, "learning_rate": 5.00115742886204e-05, "loss": 0.0131, "step": 157480 }, { "grad_norm": 0.06271452456712723, "learning_rate": 5.000606272264997e-05, "loss": 0.0121, "step": 157490 }, { "grad_norm": 0.07275572419166565, "learning_rate": 5.000055115660589e-05, "loss": 0.0124, "step": 157500 }, { "grad_norm": 0.06847888231277466, "learning_rate": 4.9995039590555116e-05, "loss": 0.0135, "step": 157510 }, { "grad_norm": 0.10977216064929962, "learning_rate": 4.99895280245646e-05, "loss": 0.0157, "step": 157520 }, { "grad_norm": 0.10835565626621246, "learning_rate": 4.998401645870133e-05, "loss": 0.0111, "step": 157530 }, { "grad_norm": 0.11809404194355011, "learning_rate": 4.997850489303228e-05, "loss": 0.0098, "step": 157540 }, { "grad_norm": 0.07604659348726273, "learning_rate": 4.9972993327624434e-05, "loss": 0.0076, "step": 157550 }, { "grad_norm": 0.1700049489736557, "learning_rate": 4.996748176254472e-05, "loss": 0.0137, "step": 157560 }, { "grad_norm": 0.186066135764122, "learning_rate": 4.996197019786014e-05, "loss": 0.012, "step": 157570 }, { "grad_norm": 0.08889748156070709, "learning_rate": 4.995645863363766e-05, "loss": 0.0119, "step": 157580 }, { "grad_norm": 0.08969295769929886, "learning_rate": 4.995094706994425e-05, "loss": 0.0139, "step": 157590 }, { "grad_norm": 0.0857299342751503, "learning_rate": 4.994543550684688e-05, "loss": 0.0181, "step": 157600 }, { "grad_norm": 0.13815198838710785, "learning_rate": 4.9939923944412506e-05, "loss": 0.0108, "step": 157610 }, { "grad_norm": 0.08493931591510773, "learning_rate": 4.993441238270813e-05, "loss": 0.0185, "step": 157620 }, { "grad_norm": 0.08685217052698135, "learning_rate": 4.9928900821800705e-05, "loss": 0.018, "step": 157630 }, { "grad_norm": 0.06757359951734543, "learning_rate": 4.992338926175719e-05, "loss": 0.0133, "step": 157640 }, { "grad_norm": 0.06662184745073318, "learning_rate": 4.991787770264458e-05, "loss": 0.0144, "step": 157650 }, { "grad_norm": 0.20165656507015228, "learning_rate": 4.991236614452983e-05, "loss": 0.0171, "step": 157660 }, { "grad_norm": 0.11042362451553345, "learning_rate": 4.990685458747991e-05, "loss": 0.0155, "step": 157670 }, { "grad_norm": 0.13983361423015594, "learning_rate": 4.990134303156179e-05, "loss": 0.0175, "step": 157680 }, { "grad_norm": 0.28565528988838196, "learning_rate": 4.989583147684246e-05, "loss": 0.0117, "step": 157690 }, { "grad_norm": 0.11156169325113297, "learning_rate": 4.989031992338886e-05, "loss": 0.0161, "step": 157700 }, { "grad_norm": 0.087398961186409, "learning_rate": 4.9884808371267994e-05, "loss": 0.0107, "step": 157710 }, { "grad_norm": 0.12587425112724304, "learning_rate": 4.9879296820546814e-05, "loss": 0.0148, "step": 157720 }, { "grad_norm": 0.19884949922561646, "learning_rate": 4.9873785271292275e-05, "loss": 0.0145, "step": 157730 }, { "grad_norm": 0.08060722798109055, "learning_rate": 4.986827372357136e-05, "loss": 0.0118, "step": 157740 }, { "grad_norm": 0.2104557901620865, "learning_rate": 4.986276217745107e-05, "loss": 0.0159, "step": 157750 }, { "grad_norm": 0.13659648597240448, "learning_rate": 4.985725063299834e-05, "loss": 0.0149, "step": 157760 }, { "grad_norm": 0.1814238429069519, "learning_rate": 4.9851739090280135e-05, "loss": 0.0128, "step": 157770 }, { "grad_norm": 0.1127181127667427, "learning_rate": 4.984622754936344e-05, "loss": 0.0147, "step": 157780 }, { "grad_norm": 0.10179127752780914, "learning_rate": 4.984071601031526e-05, "loss": 0.0164, "step": 157790 }, { "grad_norm": 0.12015680223703384, "learning_rate": 4.9835204473202496e-05, "loss": 0.0127, "step": 157800 }, { "grad_norm": 0.13908608257770538, "learning_rate": 4.982969293809216e-05, "loss": 0.0136, "step": 157810 }, { "grad_norm": 0.13602879643440247, "learning_rate": 4.9824181405051214e-05, "loss": 0.0156, "step": 157820 }, { "grad_norm": 0.12420987337827682, "learning_rate": 4.981866987414665e-05, "loss": 0.0174, "step": 157830 }, { "grad_norm": 0.13009019196033478, "learning_rate": 4.98131583454454e-05, "loss": 0.0138, "step": 157840 }, { "grad_norm": 0.0742361918091774, "learning_rate": 4.9807646819014446e-05, "loss": 0.0138, "step": 157850 }, { "grad_norm": 0.24202944338321686, "learning_rate": 4.980213529492077e-05, "loss": 0.0137, "step": 157860 }, { "grad_norm": 0.0736936405301094, "learning_rate": 4.979662377323135e-05, "loss": 0.0162, "step": 157870 }, { "grad_norm": 0.07826320827007294, "learning_rate": 4.979111225401313e-05, "loss": 0.016, "step": 157880 }, { "grad_norm": 0.19201883673667908, "learning_rate": 4.97856007373331e-05, "loss": 0.0178, "step": 157890 }, { "grad_norm": 0.11861463636159897, "learning_rate": 4.978008922325824e-05, "loss": 0.0135, "step": 157900 }, { "grad_norm": 0.13326308131217957, "learning_rate": 4.9774577711855474e-05, "loss": 0.0117, "step": 157910 }, { "grad_norm": 0.1441580057144165, "learning_rate": 4.9769066203191804e-05, "loss": 0.0173, "step": 157920 }, { "grad_norm": 0.10528367757797241, "learning_rate": 4.9763554697334206e-05, "loss": 0.0109, "step": 157930 }, { "grad_norm": 0.2547963261604309, "learning_rate": 4.9758043194349654e-05, "loss": 0.0185, "step": 157940 }, { "grad_norm": 0.28350579738616943, "learning_rate": 4.975253169430509e-05, "loss": 0.014, "step": 157950 }, { "grad_norm": 0.21702840924263, "learning_rate": 4.974702019726749e-05, "loss": 0.015, "step": 157960 }, { "grad_norm": 0.3584892153739929, "learning_rate": 4.974150870330387e-05, "loss": 0.0175, "step": 157970 }, { "grad_norm": 0.1781434863805771, "learning_rate": 4.973599721248113e-05, "loss": 0.0167, "step": 157980 }, { "grad_norm": 0.0971161425113678, "learning_rate": 4.973048572486628e-05, "loss": 0.0146, "step": 157990 }, { "grad_norm": 1.1138736009597778, "learning_rate": 4.972497424052628e-05, "loss": 0.0196, "step": 158000 }, { "grad_norm": 0.06909734755754471, "learning_rate": 4.971946275952812e-05, "loss": 0.0194, "step": 158010 }, { "grad_norm": 0.2488040030002594, "learning_rate": 4.971395128193874e-05, "loss": 0.0155, "step": 158020 }, { "grad_norm": 0.11710835248231888, "learning_rate": 4.970843980782512e-05, "loss": 0.0191, "step": 158030 }, { "grad_norm": 0.17038005590438843, "learning_rate": 4.970292833725424e-05, "loss": 0.0161, "step": 158040 }, { "grad_norm": 0.09645406156778336, "learning_rate": 4.9697416870293055e-05, "loss": 0.0177, "step": 158050 }, { "grad_norm": 0.1597353219985962, "learning_rate": 4.969190540700854e-05, "loss": 0.0252, "step": 158060 }, { "grad_norm": 0.20306560397148132, "learning_rate": 4.9686393947467656e-05, "loss": 0.0168, "step": 158070 }, { "grad_norm": 0.11633739620447159, "learning_rate": 4.968088249173741e-05, "loss": 0.0146, "step": 158080 }, { "grad_norm": 0.13116341829299927, "learning_rate": 4.9675371039884724e-05, "loss": 0.0172, "step": 158090 }, { "grad_norm": 0.09425127506256104, "learning_rate": 4.966985959197658e-05, "loss": 0.0215, "step": 158100 }, { "grad_norm": 0.09823963791131973, "learning_rate": 4.9664348148079976e-05, "loss": 0.0144, "step": 158110 }, { "grad_norm": 0.05980531498789787, "learning_rate": 4.9658836708261844e-05, "loss": 0.0135, "step": 158120 }, { "grad_norm": 0.41013479232788086, "learning_rate": 4.965332527258916e-05, "loss": 0.0171, "step": 158130 }, { "grad_norm": 0.14146694540977478, "learning_rate": 4.9647813841128895e-05, "loss": 0.0202, "step": 158140 }, { "grad_norm": 0.10780177265405655, "learning_rate": 4.964230241394806e-05, "loss": 0.0141, "step": 158150 }, { "grad_norm": 0.1033477708697319, "learning_rate": 4.963679099111356e-05, "loss": 0.0115, "step": 158160 }, { "grad_norm": 0.11023928225040436, "learning_rate": 4.9631279572692396e-05, "loss": 0.0126, "step": 158170 }, { "grad_norm": 0.19937288761138916, "learning_rate": 4.9625768158751526e-05, "loss": 0.0165, "step": 158180 }, { "grad_norm": 0.11051797866821289, "learning_rate": 4.9620256749357956e-05, "loss": 0.0175, "step": 158190 }, { "grad_norm": 0.13882693648338318, "learning_rate": 4.961474534457861e-05, "loss": 0.014, "step": 158200 }, { "grad_norm": 0.06506811082363129, "learning_rate": 4.960923394448045e-05, "loss": 0.0112, "step": 158210 }, { "grad_norm": 0.49743953347206116, "learning_rate": 4.9603722549130505e-05, "loss": 0.0149, "step": 158220 }, { "grad_norm": 0.3128504455089569, "learning_rate": 4.959821115859568e-05, "loss": 0.0175, "step": 158230 }, { "grad_norm": 0.05492601916193962, "learning_rate": 4.959269977294297e-05, "loss": 0.0165, "step": 158240 }, { "grad_norm": 0.24900947511196136, "learning_rate": 4.9587188392239334e-05, "loss": 0.0117, "step": 158250 }, { "grad_norm": 0.173191636800766, "learning_rate": 4.958167701655178e-05, "loss": 0.0193, "step": 158260 }, { "grad_norm": 0.19460900127887726, "learning_rate": 4.957616564594722e-05, "loss": 0.0144, "step": 158270 }, { "grad_norm": 0.23797833919525146, "learning_rate": 4.9570654280492654e-05, "loss": 0.0119, "step": 158280 }, { "grad_norm": 0.10852567106485367, "learning_rate": 4.956514292025504e-05, "loss": 0.0154, "step": 158290 }, { "grad_norm": 0.12343413382768631, "learning_rate": 4.955963156530136e-05, "loss": 0.0141, "step": 158300 }, { "grad_norm": 0.14885307848453522, "learning_rate": 4.955412021569856e-05, "loss": 0.0148, "step": 158310 }, { "grad_norm": 0.32802626490592957, "learning_rate": 4.954860887151362e-05, "loss": 0.0165, "step": 158320 }, { "grad_norm": 0.1289692521095276, "learning_rate": 4.954309753281352e-05, "loss": 0.0197, "step": 158330 }, { "grad_norm": 0.1535702347755432, "learning_rate": 4.9537586199665215e-05, "loss": 0.0169, "step": 158340 }, { "grad_norm": 0.13256274163722992, "learning_rate": 4.9532074872135665e-05, "loss": 0.0154, "step": 158350 }, { "grad_norm": 0.06171330809593201, "learning_rate": 4.952656355029186e-05, "loss": 0.0117, "step": 158360 }, { "grad_norm": 0.35343071818351746, "learning_rate": 4.952105223420074e-05, "loss": 0.0224, "step": 158370 }, { "grad_norm": 0.3066432774066925, "learning_rate": 4.95155409239293e-05, "loss": 0.0139, "step": 158380 }, { "grad_norm": 0.15105748176574707, "learning_rate": 4.951002961954449e-05, "loss": 0.0104, "step": 158390 }, { "grad_norm": 0.08972354978322983, "learning_rate": 4.950451832111328e-05, "loss": 0.0137, "step": 158400 }, { "grad_norm": 0.12202204018831253, "learning_rate": 4.949900702870264e-05, "loss": 0.0125, "step": 158410 }, { "grad_norm": 0.12146482616662979, "learning_rate": 4.949349574237954e-05, "loss": 0.0153, "step": 158420 }, { "grad_norm": 0.1566338837146759, "learning_rate": 4.948798446221095e-05, "loss": 0.0155, "step": 158430 }, { "grad_norm": 0.0746845230460167, "learning_rate": 4.9482473188263825e-05, "loss": 0.0153, "step": 158440 }, { "grad_norm": 0.2837418019771576, "learning_rate": 4.947696192060514e-05, "loss": 0.0102, "step": 158450 }, { "grad_norm": 0.11183515191078186, "learning_rate": 4.947145065930186e-05, "loss": 0.0134, "step": 158460 }, { "grad_norm": 0.18450932204723358, "learning_rate": 4.9465939404420966e-05, "loss": 0.0193, "step": 158470 }, { "grad_norm": 0.11721710860729218, "learning_rate": 4.94604281560294e-05, "loss": 0.0156, "step": 158480 }, { "grad_norm": 0.17861394584178925, "learning_rate": 4.945491691419414e-05, "loss": 0.0128, "step": 158490 }, { "grad_norm": 0.05011218786239624, "learning_rate": 4.944940567898216e-05, "loss": 0.0121, "step": 158500 }, { "grad_norm": 0.06012852117419243, "learning_rate": 4.944389445046043e-05, "loss": 0.0111, "step": 158510 }, { "grad_norm": 0.25699540972709656, "learning_rate": 4.943838322869589e-05, "loss": 0.0165, "step": 158520 }, { "grad_norm": 0.07186874747276306, "learning_rate": 4.943287201375552e-05, "loss": 0.0107, "step": 158530 }, { "grad_norm": 0.3332613706588745, "learning_rate": 4.9427360805706316e-05, "loss": 0.0208, "step": 158540 }, { "grad_norm": 0.15866807103157043, "learning_rate": 4.9421849604615203e-05, "loss": 0.0154, "step": 158550 }, { "grad_norm": 0.14733298122882843, "learning_rate": 4.941633841054916e-05, "loss": 0.0249, "step": 158560 }, { "grad_norm": 0.15027742087841034, "learning_rate": 4.941082722357515e-05, "loss": 0.0181, "step": 158570 }, { "grad_norm": 0.13476283848285675, "learning_rate": 4.940531604376017e-05, "loss": 0.0129, "step": 158580 }, { "grad_norm": 0.08728595823049545, "learning_rate": 4.939980487117115e-05, "loss": 0.0147, "step": 158590 }, { "grad_norm": 0.1705053150653839, "learning_rate": 4.939429370587506e-05, "loss": 0.0163, "step": 158600 }, { "grad_norm": 0.12009286880493164, "learning_rate": 4.938878254793889e-05, "loss": 0.0195, "step": 158610 }, { "grad_norm": 0.15948215126991272, "learning_rate": 4.938327139742957e-05, "loss": 0.0159, "step": 158620 }, { "grad_norm": 0.12662675976753235, "learning_rate": 4.937776025441409e-05, "loss": 0.0149, "step": 158630 }, { "grad_norm": 0.08183420449495316, "learning_rate": 4.937224911895941e-05, "loss": 0.011, "step": 158640 }, { "grad_norm": 0.10340335220098495, "learning_rate": 4.9366737991132514e-05, "loss": 0.018, "step": 158650 }, { "grad_norm": 0.11634411662817001, "learning_rate": 4.936122687100033e-05, "loss": 0.0137, "step": 158660 }, { "grad_norm": 0.16461895406246185, "learning_rate": 4.935571575862984e-05, "loss": 0.0135, "step": 158670 }, { "grad_norm": 0.09585689753293991, "learning_rate": 4.935020465408804e-05, "loss": 0.014, "step": 158680 }, { "grad_norm": 0.13127501308918, "learning_rate": 4.934469355744183e-05, "loss": 0.0156, "step": 158690 }, { "grad_norm": 0.0968119278550148, "learning_rate": 4.9339182468758225e-05, "loss": 0.0168, "step": 158700 }, { "grad_norm": 0.11573731899261475, "learning_rate": 4.933367138810418e-05, "loss": 0.0163, "step": 158710 }, { "grad_norm": 0.10494935512542725, "learning_rate": 4.932816031554667e-05, "loss": 0.0175, "step": 158720 }, { "grad_norm": 0.19640889763832092, "learning_rate": 4.932264925115263e-05, "loss": 0.0178, "step": 158730 }, { "grad_norm": 0.10131736844778061, "learning_rate": 4.931713819498904e-05, "loss": 0.0151, "step": 158740 }, { "grad_norm": 0.1483370065689087, "learning_rate": 4.931162714712286e-05, "loss": 0.0127, "step": 158750 }, { "grad_norm": 0.280985563993454, "learning_rate": 4.930611610762109e-05, "loss": 0.013, "step": 158760 }, { "grad_norm": 0.07077572494745255, "learning_rate": 4.930060507655063e-05, "loss": 0.0137, "step": 158770 }, { "grad_norm": 0.10838886350393295, "learning_rate": 4.929509405397849e-05, "loss": 0.0151, "step": 158780 }, { "grad_norm": 0.08230222761631012, "learning_rate": 4.9289583039971644e-05, "loss": 0.0109, "step": 158790 }, { "grad_norm": 0.46071937680244446, "learning_rate": 4.9284072034597015e-05, "loss": 0.0151, "step": 158800 }, { "grad_norm": 0.11090035736560822, "learning_rate": 4.9278561037921576e-05, "loss": 0.0134, "step": 158810 }, { "grad_norm": 0.1736079305410385, "learning_rate": 4.927305005001231e-05, "loss": 0.0153, "step": 158820 }, { "grad_norm": 0.070504330098629, "learning_rate": 4.926753907093619e-05, "loss": 0.0155, "step": 158830 }, { "grad_norm": 0.12846507132053375, "learning_rate": 4.9262028100760146e-05, "loss": 0.0132, "step": 158840 }, { "grad_norm": 0.33329296112060547, "learning_rate": 4.925651713955115e-05, "loss": 0.0104, "step": 158850 }, { "grad_norm": 0.1534387618303299, "learning_rate": 4.92510061873762e-05, "loss": 0.0115, "step": 158860 }, { "grad_norm": 0.09252570569515228, "learning_rate": 4.9245495244302206e-05, "loss": 0.0141, "step": 158870 }, { "grad_norm": 0.09361197799444199, "learning_rate": 4.9239984310396164e-05, "loss": 0.0155, "step": 158880 }, { "grad_norm": 0.06526494026184082, "learning_rate": 4.9234473385725025e-05, "loss": 0.0141, "step": 158890 }, { "grad_norm": 0.13699480891227722, "learning_rate": 4.922896247035578e-05, "loss": 0.0127, "step": 158900 }, { "grad_norm": 0.1262766420841217, "learning_rate": 4.9223451564355345e-05, "loss": 0.0124, "step": 158910 }, { "grad_norm": 0.13343919813632965, "learning_rate": 4.9217940667790704e-05, "loss": 0.0141, "step": 158920 }, { "grad_norm": 0.16508854925632477, "learning_rate": 4.921242978072885e-05, "loss": 0.0164, "step": 158930 }, { "grad_norm": 0.11285191029310226, "learning_rate": 4.92069189032367e-05, "loss": 0.0171, "step": 158940 }, { "grad_norm": 0.13801121711730957, "learning_rate": 4.9201408035381225e-05, "loss": 0.0163, "step": 158950 }, { "grad_norm": 0.09040949493646622, "learning_rate": 4.91958971772294e-05, "loss": 0.0157, "step": 158960 }, { "grad_norm": 0.445629745721817, "learning_rate": 4.9190386328848206e-05, "loss": 0.0165, "step": 158970 }, { "grad_norm": 0.21025756001472473, "learning_rate": 4.918487549030456e-05, "loss": 0.0196, "step": 158980 }, { "grad_norm": 0.09478508681058884, "learning_rate": 4.9179364661665445e-05, "loss": 0.0139, "step": 158990 }, { "grad_norm": 0.13639630377292633, "learning_rate": 4.9173853842997836e-05, "loss": 0.0133, "step": 159000 }, { "grad_norm": 0.07754699140787125, "learning_rate": 4.9168343034368686e-05, "loss": 0.0122, "step": 159010 }, { "grad_norm": 0.0960865318775177, "learning_rate": 4.916283223584494e-05, "loss": 0.0134, "step": 159020 }, { "grad_norm": 0.11966516822576523, "learning_rate": 4.915732144749357e-05, "loss": 0.0102, "step": 159030 }, { "grad_norm": 0.46396395564079285, "learning_rate": 4.9151810669381556e-05, "loss": 0.0142, "step": 159040 }, { "grad_norm": 0.1585092693567276, "learning_rate": 4.9146299901575836e-05, "loss": 0.01, "step": 159050 }, { "grad_norm": 0.07418674975633621, "learning_rate": 4.9140789144143377e-05, "loss": 0.0179, "step": 159060 }, { "grad_norm": 0.27270984649658203, "learning_rate": 4.913527839715113e-05, "loss": 0.023, "step": 159070 }, { "grad_norm": 0.1412113904953003, "learning_rate": 4.912976766066609e-05, "loss": 0.0145, "step": 159080 }, { "grad_norm": 0.17011940479278564, "learning_rate": 4.912425693475519e-05, "loss": 0.0137, "step": 159090 }, { "grad_norm": 0.16094830632209778, "learning_rate": 4.911874621948538e-05, "loss": 0.0152, "step": 159100 }, { "grad_norm": 0.39069706201553345, "learning_rate": 4.9113235514923646e-05, "loss": 0.0166, "step": 159110 }, { "grad_norm": 0.08432938903570175, "learning_rate": 4.910772482113693e-05, "loss": 0.0157, "step": 159120 }, { "grad_norm": 0.08962514251470566, "learning_rate": 4.910221413819221e-05, "loss": 0.0135, "step": 159130 }, { "grad_norm": 0.1010599136352539, "learning_rate": 4.9096703466156426e-05, "loss": 0.0143, "step": 159140 }, { "grad_norm": 0.1905852109193802, "learning_rate": 4.9091192805096556e-05, "loss": 0.015, "step": 159150 }, { "grad_norm": 0.09552178531885147, "learning_rate": 4.908568215507954e-05, "loss": 0.0147, "step": 159160 }, { "grad_norm": 0.14606517553329468, "learning_rate": 4.908017151617237e-05, "loss": 0.0196, "step": 159170 }, { "grad_norm": 0.21717029809951782, "learning_rate": 4.907466088844197e-05, "loss": 0.0151, "step": 159180 }, { "grad_norm": 0.07826266437768936, "learning_rate": 4.9069150271955314e-05, "loss": 0.0162, "step": 159190 }, { "grad_norm": 0.07476163655519485, "learning_rate": 4.906363966677936e-05, "loss": 0.0114, "step": 159200 }, { "grad_norm": 0.1359165906906128, "learning_rate": 4.9058129072981073e-05, "loss": 0.0179, "step": 159210 }, { "grad_norm": 0.06244704872369766, "learning_rate": 4.905261849062741e-05, "loss": 0.0172, "step": 159220 }, { "grad_norm": 0.20233745872974396, "learning_rate": 4.904710791978532e-05, "loss": 0.0159, "step": 159230 }, { "grad_norm": 0.13509611785411835, "learning_rate": 4.9041597360521766e-05, "loss": 0.0131, "step": 159240 }, { "grad_norm": 0.13718901574611664, "learning_rate": 4.9036086812903725e-05, "loss": 0.0149, "step": 159250 }, { "grad_norm": 0.17136403918266296, "learning_rate": 4.903057627699812e-05, "loss": 0.0139, "step": 159260 }, { "grad_norm": 0.2877620458602905, "learning_rate": 4.902506575287194e-05, "loss": 0.0244, "step": 159270 }, { "grad_norm": 0.07819168269634247, "learning_rate": 4.901955524059211e-05, "loss": 0.012, "step": 159280 }, { "grad_norm": 0.07619377970695496, "learning_rate": 4.901404474022565e-05, "loss": 0.0125, "step": 159290 }, { "grad_norm": 0.048052623867988586, "learning_rate": 4.900853425183945e-05, "loss": 0.0159, "step": 159300 }, { "grad_norm": 0.11868283897638321, "learning_rate": 4.900302377550049e-05, "loss": 0.0201, "step": 159310 }, { "grad_norm": 0.12589649856090546, "learning_rate": 4.899751331127576e-05, "loss": 0.0135, "step": 159320 }, { "grad_norm": 0.10783841460943222, "learning_rate": 4.8992002859232166e-05, "loss": 0.0206, "step": 159330 }, { "grad_norm": 0.0706724002957344, "learning_rate": 4.898649241943668e-05, "loss": 0.0122, "step": 159340 }, { "grad_norm": 0.08097659796476364, "learning_rate": 4.8980981991956285e-05, "loss": 0.017, "step": 159350 }, { "grad_norm": 0.12082535773515701, "learning_rate": 4.897547157685793e-05, "loss": 0.0126, "step": 159360 }, { "grad_norm": 0.11192402243614197, "learning_rate": 4.896996117420855e-05, "loss": 0.016, "step": 159370 }, { "grad_norm": 0.17763960361480713, "learning_rate": 4.896445078407511e-05, "loss": 0.0231, "step": 159380 }, { "grad_norm": 0.11539031565189362, "learning_rate": 4.895894040652457e-05, "loss": 0.0143, "step": 159390 }, { "grad_norm": 0.19652904570102692, "learning_rate": 4.895343004162391e-05, "loss": 0.016, "step": 159400 }, { "grad_norm": 0.23651087284088135, "learning_rate": 4.894791968944004e-05, "loss": 0.0143, "step": 159410 }, { "grad_norm": 0.08803147077560425, "learning_rate": 4.894240935003994e-05, "loss": 0.0125, "step": 159420 }, { "grad_norm": 0.19279301166534424, "learning_rate": 4.893689902349058e-05, "loss": 0.0184, "step": 159430 }, { "grad_norm": 0.2432875782251358, "learning_rate": 4.893138870985889e-05, "loss": 0.019, "step": 159440 }, { "grad_norm": 0.1542809009552002, "learning_rate": 4.8925878409211834e-05, "loss": 0.0173, "step": 159450 }, { "grad_norm": 0.13275973498821259, "learning_rate": 4.8920368121616366e-05, "loss": 0.0137, "step": 159460 }, { "grad_norm": 0.12134815752506256, "learning_rate": 4.891485784713947e-05, "loss": 0.0131, "step": 159470 }, { "grad_norm": 0.1034131795167923, "learning_rate": 4.8909347585848046e-05, "loss": 0.0147, "step": 159480 }, { "grad_norm": 0.07504834234714508, "learning_rate": 4.890383733780909e-05, "loss": 0.0141, "step": 159490 }, { "grad_norm": 0.10046958923339844, "learning_rate": 4.889832710308956e-05, "loss": 0.0202, "step": 159500 }, { "grad_norm": 0.06915870308876038, "learning_rate": 4.889281688175637e-05, "loss": 0.0105, "step": 159510 }, { "grad_norm": 0.12945859134197235, "learning_rate": 4.88873066738765e-05, "loss": 0.0109, "step": 159520 }, { "grad_norm": 0.12681575119495392, "learning_rate": 4.888179647951692e-05, "loss": 0.014, "step": 159530 }, { "grad_norm": 0.08668245375156403, "learning_rate": 4.887628629874458e-05, "loss": 0.0133, "step": 159540 }, { "grad_norm": 0.08782228082418442, "learning_rate": 4.887077613162639e-05, "loss": 0.0101, "step": 159550 }, { "grad_norm": 0.16056986153125763, "learning_rate": 4.886526597822935e-05, "loss": 0.0158, "step": 159560 }, { "grad_norm": 0.10922635346651077, "learning_rate": 4.8859755838620426e-05, "loss": 0.011, "step": 159570 }, { "grad_norm": 0.16070640087127686, "learning_rate": 4.885424571286651e-05, "loss": 0.012, "step": 159580 }, { "grad_norm": 0.10563171654939651, "learning_rate": 4.88487356010346e-05, "loss": 0.0164, "step": 159590 }, { "grad_norm": 0.1227644830942154, "learning_rate": 4.8843225503191634e-05, "loss": 0.0142, "step": 159600 }, { "grad_norm": 0.09134188294410706, "learning_rate": 4.883771541940459e-05, "loss": 0.0143, "step": 159610 }, { "grad_norm": 0.11931659281253815, "learning_rate": 4.883220534974038e-05, "loss": 0.03, "step": 159620 }, { "grad_norm": 0.09232911467552185, "learning_rate": 4.8826695294265986e-05, "loss": 0.0132, "step": 159630 }, { "grad_norm": 0.051021184772253036, "learning_rate": 4.882118525304837e-05, "loss": 0.0096, "step": 159640 }, { "grad_norm": 0.17889921367168427, "learning_rate": 4.881567522615444e-05, "loss": 0.0158, "step": 159650 }, { "grad_norm": 0.07270260155200958, "learning_rate": 4.881016521365118e-05, "loss": 0.013, "step": 159660 }, { "grad_norm": 0.12509669363498688, "learning_rate": 4.880465521560553e-05, "loss": 0.0227, "step": 159670 }, { "grad_norm": 0.18600678443908691, "learning_rate": 4.879914523208447e-05, "loss": 0.0193, "step": 159680 }, { "grad_norm": 0.12755030393600464, "learning_rate": 4.879363526315491e-05, "loss": 0.018, "step": 159690 }, { "grad_norm": 0.1376175731420517, "learning_rate": 4.878812530888382e-05, "loss": 0.0133, "step": 159700 }, { "grad_norm": 0.08383357524871826, "learning_rate": 4.878261536933815e-05, "loss": 0.0152, "step": 159710 }, { "grad_norm": 0.0752391442656517, "learning_rate": 4.877710544458488e-05, "loss": 0.0129, "step": 159720 }, { "grad_norm": 0.09061828255653381, "learning_rate": 4.877159553469091e-05, "loss": 0.0145, "step": 159730 }, { "grad_norm": 0.09716653823852539, "learning_rate": 4.8766085639723205e-05, "loss": 0.0151, "step": 159740 }, { "grad_norm": 0.11222995817661285, "learning_rate": 4.8760575759748743e-05, "loss": 0.0148, "step": 159750 }, { "grad_norm": 0.08446008712053299, "learning_rate": 4.875506589483445e-05, "loss": 0.0135, "step": 159760 }, { "grad_norm": 0.11890272051095963, "learning_rate": 4.8749556045047284e-05, "loss": 0.0154, "step": 159770 }, { "grad_norm": 0.1156875267624855, "learning_rate": 4.8744046210454184e-05, "loss": 0.0112, "step": 159780 }, { "grad_norm": 0.0993054211139679, "learning_rate": 4.873853639112212e-05, "loss": 0.0163, "step": 159790 }, { "grad_norm": 0.1041811928153038, "learning_rate": 4.873302658711803e-05, "loss": 0.0185, "step": 159800 }, { "grad_norm": 0.12525543570518494, "learning_rate": 4.872751679850886e-05, "loss": 0.0189, "step": 159810 }, { "grad_norm": 0.15832851827144623, "learning_rate": 4.872200702536156e-05, "loss": 0.0115, "step": 159820 }, { "grad_norm": 0.16662713885307312, "learning_rate": 4.8716497267743075e-05, "loss": 0.0214, "step": 159830 }, { "grad_norm": 0.12297353148460388, "learning_rate": 4.871098752572037e-05, "loss": 0.0157, "step": 159840 }, { "grad_norm": 0.12983757257461548, "learning_rate": 4.8705477799360375e-05, "loss": 0.0129, "step": 159850 }, { "grad_norm": 0.07585883885622025, "learning_rate": 4.869996808873006e-05, "loss": 0.0175, "step": 159860 }, { "grad_norm": 0.19647981226444244, "learning_rate": 4.869445839389634e-05, "loss": 0.0189, "step": 159870 }, { "grad_norm": 0.0709855780005455, "learning_rate": 4.868894871492619e-05, "loss": 0.0135, "step": 159880 }, { "grad_norm": 0.16542337834835052, "learning_rate": 4.868343905188658e-05, "loss": 0.0205, "step": 159890 }, { "grad_norm": 0.1329207420349121, "learning_rate": 4.8677929404844394e-05, "loss": 0.0157, "step": 159900 }, { "grad_norm": 0.09097471088171005, "learning_rate": 4.867241977386662e-05, "loss": 0.0194, "step": 159910 }, { "grad_norm": 0.15336447954177856, "learning_rate": 4.86669101590202e-05, "loss": 0.0181, "step": 159920 }, { "grad_norm": 0.07959548383951187, "learning_rate": 4.86614005603721e-05, "loss": 0.0141, "step": 159930 }, { "grad_norm": 0.177122563123703, "learning_rate": 4.8655890977989224e-05, "loss": 0.016, "step": 159940 }, { "grad_norm": 0.13242432475090027, "learning_rate": 4.865038141193854e-05, "loss": 0.0189, "step": 159950 }, { "grad_norm": 0.11203442513942719, "learning_rate": 4.8644871862287e-05, "loss": 0.0155, "step": 159960 }, { "grad_norm": 0.11343926936388016, "learning_rate": 4.863936232910157e-05, "loss": 0.0161, "step": 159970 }, { "grad_norm": 0.08320446312427521, "learning_rate": 4.863385281244915e-05, "loss": 0.0135, "step": 159980 }, { "grad_norm": 0.19585780799388885, "learning_rate": 4.86283433123967e-05, "loss": 0.0145, "step": 159990 }, { "grad_norm": 0.10139840841293335, "learning_rate": 4.8622833829011205e-05, "loss": 0.0192, "step": 160000 }, { "grad_norm": 0.09149350970983505, "learning_rate": 4.8617324362359555e-05, "loss": 0.0137, "step": 160010 }, { "grad_norm": 0.10291777551174164, "learning_rate": 4.8611814912508715e-05, "loss": 0.0126, "step": 160020 }, { "grad_norm": 0.15700621902942657, "learning_rate": 4.8606305479525636e-05, "loss": 0.0123, "step": 160030 }, { "grad_norm": 0.0906241238117218, "learning_rate": 4.8600796063477286e-05, "loss": 0.0113, "step": 160040 }, { "grad_norm": 0.07577816396951675, "learning_rate": 4.859528666443057e-05, "loss": 0.0143, "step": 160050 }, { "grad_norm": 0.12470731139183044, "learning_rate": 4.858977728245244e-05, "loss": 0.0186, "step": 160060 }, { "grad_norm": 0.10590014606714249, "learning_rate": 4.8584267917609865e-05, "loss": 0.0126, "step": 160070 }, { "grad_norm": 0.09733769297599792, "learning_rate": 4.857875856996975e-05, "loss": 0.0117, "step": 160080 }, { "grad_norm": 0.12814128398895264, "learning_rate": 4.857324923959907e-05, "loss": 0.0104, "step": 160090 }, { "grad_norm": 0.09444978833198547, "learning_rate": 4.8567739926564746e-05, "loss": 0.0138, "step": 160100 }, { "grad_norm": 0.08935682475566864, "learning_rate": 4.8562230630933764e-05, "loss": 0.0145, "step": 160110 }, { "grad_norm": 0.10492294281721115, "learning_rate": 4.855672135277301e-05, "loss": 0.015, "step": 160120 }, { "grad_norm": 0.15679733455181122, "learning_rate": 4.855121209214947e-05, "loss": 0.0144, "step": 160130 }, { "grad_norm": 0.16875779628753662, "learning_rate": 4.8545702849130076e-05, "loss": 0.0151, "step": 160140 }, { "grad_norm": 0.09644544869661331, "learning_rate": 4.854019362378175e-05, "loss": 0.0122, "step": 160150 }, { "grad_norm": 0.10120651125907898, "learning_rate": 4.853468441617145e-05, "loss": 0.0163, "step": 160160 }, { "grad_norm": 0.10703402012586594, "learning_rate": 4.8529175226366114e-05, "loss": 0.0092, "step": 160170 }, { "grad_norm": 0.10330267250537872, "learning_rate": 4.852366605443271e-05, "loss": 0.013, "step": 160180 }, { "grad_norm": 0.08653794974088669, "learning_rate": 4.8518156900438136e-05, "loss": 0.0134, "step": 160190 }, { "grad_norm": 0.08800660073757172, "learning_rate": 4.851264776444936e-05, "loss": 0.0147, "step": 160200 }, { "grad_norm": 0.16844625771045685, "learning_rate": 4.8507138646533336e-05, "loss": 0.0142, "step": 160210 }, { "grad_norm": 0.0944291427731514, "learning_rate": 4.8501629546756965e-05, "loss": 0.0192, "step": 160220 }, { "grad_norm": 0.16663935780525208, "learning_rate": 4.849612046518721e-05, "loss": 0.0147, "step": 160230 }, { "grad_norm": 0.09282603114843369, "learning_rate": 4.8490611401891014e-05, "loss": 0.0131, "step": 160240 }, { "grad_norm": 0.22883827984333038, "learning_rate": 4.8485102356935333e-05, "loss": 0.0182, "step": 160250 }, { "grad_norm": 0.15471284091472626, "learning_rate": 4.847959333038706e-05, "loss": 0.0129, "step": 160260 }, { "grad_norm": 0.07815590500831604, "learning_rate": 4.8474084322313174e-05, "loss": 0.014, "step": 160270 }, { "grad_norm": 0.12307275086641312, "learning_rate": 4.84685753327806e-05, "loss": 0.0153, "step": 160280 }, { "grad_norm": 0.15871202945709229, "learning_rate": 4.8463066361856304e-05, "loss": 0.0156, "step": 160290 }, { "grad_norm": 0.08212015777826309, "learning_rate": 4.845755740960719e-05, "loss": 0.0117, "step": 160300 }, { "grad_norm": 0.07742168009281158, "learning_rate": 4.845204847610019e-05, "loss": 0.0155, "step": 160310 }, { "grad_norm": 0.08096970617771149, "learning_rate": 4.84465395614023e-05, "loss": 0.016, "step": 160320 }, { "grad_norm": 0.15741047263145447, "learning_rate": 4.844103066558039e-05, "loss": 0.0132, "step": 160330 }, { "grad_norm": 0.10940541326999664, "learning_rate": 4.843552178870144e-05, "loss": 0.0136, "step": 160340 }, { "grad_norm": 0.1086326390504837, "learning_rate": 4.843001293083237e-05, "loss": 0.0111, "step": 160350 }, { "grad_norm": 0.07508556544780731, "learning_rate": 4.8424504092040146e-05, "loss": 0.0125, "step": 160360 }, { "grad_norm": 0.07625310122966766, "learning_rate": 4.8418995272391665e-05, "loss": 0.0107, "step": 160370 }, { "grad_norm": 0.08846919238567352, "learning_rate": 4.8413486471953885e-05, "loss": 0.0131, "step": 160380 }, { "grad_norm": 0.09340768307447433, "learning_rate": 4.840797769079377e-05, "loss": 0.0114, "step": 160390 }, { "grad_norm": 0.05302005261182785, "learning_rate": 4.8402468928978196e-05, "loss": 0.0124, "step": 160400 }, { "grad_norm": 0.11295569688081741, "learning_rate": 4.839696018657414e-05, "loss": 0.0168, "step": 160410 }, { "grad_norm": 0.08327606320381165, "learning_rate": 4.839145146364853e-05, "loss": 0.0081, "step": 160420 }, { "grad_norm": 0.09736546128988266, "learning_rate": 4.838594276026832e-05, "loss": 0.0098, "step": 160430 }, { "grad_norm": 0.06092170625925064, "learning_rate": 4.838043407650041e-05, "loss": 0.0164, "step": 160440 }, { "grad_norm": 0.08901051431894302, "learning_rate": 4.837492541241176e-05, "loss": 0.0175, "step": 160450 }, { "grad_norm": 0.12026440352201462, "learning_rate": 4.8369416768069306e-05, "loss": 0.0162, "step": 160460 }, { "grad_norm": 0.09899874776601791, "learning_rate": 4.836390814353997e-05, "loss": 0.0151, "step": 160470 }, { "grad_norm": 0.10666003823280334, "learning_rate": 4.835839953889071e-05, "loss": 0.0161, "step": 160480 }, { "grad_norm": 0.12411024421453476, "learning_rate": 4.835289095418843e-05, "loss": 0.0122, "step": 160490 }, { "grad_norm": 0.09530618786811829, "learning_rate": 4.834738238950009e-05, "loss": 0.018, "step": 160500 }, { "grad_norm": 0.09216927736997604, "learning_rate": 4.83418738448926e-05, "loss": 0.0161, "step": 160510 }, { "grad_norm": 0.09855306148529053, "learning_rate": 4.833636532043293e-05, "loss": 0.0092, "step": 160520 }, { "grad_norm": 0.5133212804794312, "learning_rate": 4.833085681618798e-05, "loss": 0.0175, "step": 160530 }, { "grad_norm": 0.17281199991703033, "learning_rate": 4.832534833222469e-05, "loss": 0.0125, "step": 160540 }, { "grad_norm": 0.1286899596452713, "learning_rate": 4.831983986861e-05, "loss": 0.0149, "step": 160550 }, { "grad_norm": 0.1609204262495041, "learning_rate": 4.831433142541085e-05, "loss": 0.0137, "step": 160560 }, { "grad_norm": 0.25602006912231445, "learning_rate": 4.830882300269417e-05, "loss": 0.0135, "step": 160570 }, { "grad_norm": 0.12911906838417053, "learning_rate": 4.830331460052687e-05, "loss": 0.0162, "step": 160580 }, { "grad_norm": 0.13546517491340637, "learning_rate": 4.8297806218975906e-05, "loss": 0.0128, "step": 160590 }, { "grad_norm": 0.0663185715675354, "learning_rate": 4.829229785810821e-05, "loss": 0.0097, "step": 160600 }, { "grad_norm": 0.18225426971912384, "learning_rate": 4.82867895179907e-05, "loss": 0.0157, "step": 160610 }, { "grad_norm": 0.22204144299030304, "learning_rate": 4.828128119869032e-05, "loss": 0.011, "step": 160620 }, { "grad_norm": 0.06807631254196167, "learning_rate": 4.827577290027399e-05, "loss": 0.0221, "step": 160630 }, { "grad_norm": 0.1188775971531868, "learning_rate": 4.827026462280866e-05, "loss": 0.014, "step": 160640 }, { "grad_norm": 0.21770286560058594, "learning_rate": 4.826475636636124e-05, "loss": 0.0152, "step": 160650 }, { "grad_norm": 0.07161993533372879, "learning_rate": 4.825924813099866e-05, "loss": 0.0133, "step": 160660 }, { "grad_norm": 0.1875649243593216, "learning_rate": 4.825373991678786e-05, "loss": 0.0178, "step": 160670 }, { "grad_norm": 0.07502929866313934, "learning_rate": 4.824823172379579e-05, "loss": 0.0141, "step": 160680 }, { "grad_norm": 0.20484724640846252, "learning_rate": 4.824272355208934e-05, "loss": 0.0119, "step": 160690 }, { "grad_norm": 0.053134430199861526, "learning_rate": 4.8237215401735456e-05, "loss": 0.015, "step": 160700 }, { "grad_norm": 0.07146137207746506, "learning_rate": 4.823170727280109e-05, "loss": 0.0152, "step": 160710 }, { "grad_norm": 0.15832212567329407, "learning_rate": 4.822619916535313e-05, "loss": 0.014, "step": 160720 }, { "grad_norm": 0.1053500771522522, "learning_rate": 4.822069107945853e-05, "loss": 0.0123, "step": 160730 }, { "grad_norm": 0.1364041119813919, "learning_rate": 4.82151830151842e-05, "loss": 0.014, "step": 160740 }, { "grad_norm": 0.143165722489357, "learning_rate": 4.8209674972597114e-05, "loss": 0.0171, "step": 160750 }, { "grad_norm": 0.08117113262414932, "learning_rate": 4.820416695176414e-05, "loss": 0.0121, "step": 160760 }, { "grad_norm": 0.08515364676713943, "learning_rate": 4.8198658952752236e-05, "loss": 0.0169, "step": 160770 }, { "grad_norm": 0.0812104120850563, "learning_rate": 4.8193150975628345e-05, "loss": 0.0194, "step": 160780 }, { "grad_norm": 0.17455387115478516, "learning_rate": 4.818764302045936e-05, "loss": 0.013, "step": 160790 }, { "grad_norm": 0.11524619162082672, "learning_rate": 4.818213508731222e-05, "loss": 0.0128, "step": 160800 }, { "grad_norm": 0.11881104856729507, "learning_rate": 4.817662717625386e-05, "loss": 0.0207, "step": 160810 }, { "grad_norm": 0.12002680450677872, "learning_rate": 4.8171119287351213e-05, "loss": 0.0123, "step": 160820 }, { "grad_norm": 0.16092130541801453, "learning_rate": 4.8165611420671183e-05, "loss": 0.0184, "step": 160830 }, { "grad_norm": 0.13101954758167267, "learning_rate": 4.81601035762807e-05, "loss": 0.0155, "step": 160840 }, { "grad_norm": 0.16438348591327667, "learning_rate": 4.815459575424671e-05, "loss": 0.0215, "step": 160850 }, { "grad_norm": 0.15940554440021515, "learning_rate": 4.814908795463611e-05, "loss": 0.0142, "step": 160860 }, { "grad_norm": 0.07818388193845749, "learning_rate": 4.814358017751584e-05, "loss": 0.0172, "step": 160870 }, { "grad_norm": 0.14014169573783875, "learning_rate": 4.8138072422952827e-05, "loss": 0.019, "step": 160880 }, { "grad_norm": 0.10749505460262299, "learning_rate": 4.813256469101401e-05, "loss": 0.013, "step": 160890 }, { "grad_norm": 0.1113574430346489, "learning_rate": 4.812705698176627e-05, "loss": 0.0129, "step": 160900 }, { "grad_norm": 0.07368806004524231, "learning_rate": 4.812154929527656e-05, "loss": 0.0152, "step": 160910 }, { "grad_norm": 0.19243185222148895, "learning_rate": 4.81160416316118e-05, "loss": 0.0121, "step": 160920 }, { "grad_norm": 0.16360554099082947, "learning_rate": 4.811053399083893e-05, "loss": 0.0188, "step": 160930 }, { "grad_norm": 0.19201427698135376, "learning_rate": 4.810502637302484e-05, "loss": 0.012, "step": 160940 }, { "grad_norm": 0.29911530017852783, "learning_rate": 4.809951877823646e-05, "loss": 0.0168, "step": 160950 }, { "grad_norm": 0.08895602077245712, "learning_rate": 4.809401120654075e-05, "loss": 0.0147, "step": 160960 }, { "grad_norm": 0.09088456630706787, "learning_rate": 4.808850365800458e-05, "loss": 0.0141, "step": 160970 }, { "grad_norm": 0.08069208264350891, "learning_rate": 4.80829961326949e-05, "loss": 0.0155, "step": 160980 }, { "grad_norm": 0.16279852390289307, "learning_rate": 4.807748863067862e-05, "loss": 0.015, "step": 160990 }, { "grad_norm": 0.1534314751625061, "learning_rate": 4.807198115202269e-05, "loss": 0.0214, "step": 161000 }, { "grad_norm": 0.1778637319803238, "learning_rate": 4.806647369679399e-05, "loss": 0.0128, "step": 161010 }, { "grad_norm": 0.09660652279853821, "learning_rate": 4.806096626505946e-05, "loss": 0.0099, "step": 161020 }, { "grad_norm": 0.062470462173223495, "learning_rate": 4.805545885688604e-05, "loss": 0.0104, "step": 161030 }, { "grad_norm": 0.1436895728111267, "learning_rate": 4.804995147234061e-05, "loss": 0.0107, "step": 161040 }, { "grad_norm": 0.18280309438705444, "learning_rate": 4.804444411149012e-05, "loss": 0.0115, "step": 161050 }, { "grad_norm": 0.06427203863859177, "learning_rate": 4.8038936774401475e-05, "loss": 0.0114, "step": 161060 }, { "grad_norm": 0.08245015144348145, "learning_rate": 4.8033429461141615e-05, "loss": 0.0128, "step": 161070 }, { "grad_norm": 0.13652442395687103, "learning_rate": 4.802792217177743e-05, "loss": 0.0121, "step": 161080 }, { "grad_norm": 0.08941390365362167, "learning_rate": 4.8022414906375846e-05, "loss": 0.0114, "step": 161090 }, { "grad_norm": 0.16229289770126343, "learning_rate": 4.80169076650038e-05, "loss": 0.0126, "step": 161100 }, { "grad_norm": 0.10325726866722107, "learning_rate": 4.8011400447728205e-05, "loss": 0.0118, "step": 161110 }, { "grad_norm": 0.055094871670007706, "learning_rate": 4.800589325461595e-05, "loss": 0.0106, "step": 161120 }, { "grad_norm": 0.07836663722991943, "learning_rate": 4.800038608573398e-05, "loss": 0.0166, "step": 161130 }, { "grad_norm": 0.14009587466716766, "learning_rate": 4.799487894114922e-05, "loss": 0.0153, "step": 161140 }, { "grad_norm": 0.1244383230805397, "learning_rate": 4.7989371820928583e-05, "loss": 0.0113, "step": 161150 }, { "grad_norm": 0.0931384265422821, "learning_rate": 4.798386472513895e-05, "loss": 0.0091, "step": 161160 }, { "grad_norm": 0.11524409055709839, "learning_rate": 4.7978357653847274e-05, "loss": 0.0155, "step": 161170 }, { "grad_norm": 0.1438729465007782, "learning_rate": 4.7972850607120466e-05, "loss": 0.0129, "step": 161180 }, { "grad_norm": 0.07512291520833969, "learning_rate": 4.796734358502544e-05, "loss": 0.0131, "step": 161190 }, { "grad_norm": 0.06484723091125488, "learning_rate": 4.79618365876291e-05, "loss": 0.0149, "step": 161200 }, { "grad_norm": 0.07727029919624329, "learning_rate": 4.7956329614998374e-05, "loss": 0.0112, "step": 161210 }, { "grad_norm": 0.22928042709827423, "learning_rate": 4.7950822667200163e-05, "loss": 0.0117, "step": 161220 }, { "grad_norm": 0.07818150520324707, "learning_rate": 4.794531574430141e-05, "loss": 0.0135, "step": 161230 }, { "grad_norm": 0.0909726694226265, "learning_rate": 4.793980884636899e-05, "loss": 0.0141, "step": 161240 }, { "grad_norm": 0.1331484168767929, "learning_rate": 4.7934301973469856e-05, "loss": 0.0146, "step": 161250 }, { "grad_norm": 0.7979176640510559, "learning_rate": 4.792879512567089e-05, "loss": 0.0166, "step": 161260 }, { "grad_norm": 0.27085456252098083, "learning_rate": 4.792328830303903e-05, "loss": 0.0157, "step": 161270 }, { "grad_norm": 0.23898372054100037, "learning_rate": 4.7917781505641175e-05, "loss": 0.0153, "step": 161280 }, { "grad_norm": 0.1097908616065979, "learning_rate": 4.7912274733544235e-05, "loss": 0.0145, "step": 161290 }, { "grad_norm": 0.09248705208301544, "learning_rate": 4.790676798681512e-05, "loss": 0.0115, "step": 161300 }, { "grad_norm": 0.14016002416610718, "learning_rate": 4.790126126552076e-05, "loss": 0.0139, "step": 161310 }, { "grad_norm": 0.16500495374202728, "learning_rate": 4.7895754569728066e-05, "loss": 0.0157, "step": 161320 }, { "grad_norm": 0.19337384402751923, "learning_rate": 4.7890247899503934e-05, "loss": 0.0215, "step": 161330 }, { "grad_norm": 0.2245512455701828, "learning_rate": 4.7884741254915275e-05, "loss": 0.0171, "step": 161340 }, { "grad_norm": 0.14886827766895294, "learning_rate": 4.7879234636029024e-05, "loss": 0.0117, "step": 161350 }, { "grad_norm": 0.06723202019929886, "learning_rate": 4.7873728042912057e-05, "loss": 0.0145, "step": 161360 }, { "grad_norm": 0.12962254881858826, "learning_rate": 4.78682214756313e-05, "loss": 0.0154, "step": 161370 }, { "grad_norm": 0.06603516638278961, "learning_rate": 4.7862714934253675e-05, "loss": 0.0174, "step": 161380 }, { "grad_norm": 0.10754817724227905, "learning_rate": 4.785720841884609e-05, "loss": 0.0134, "step": 161390 }, { "grad_norm": 0.06330955773591995, "learning_rate": 4.785170192947543e-05, "loss": 0.0146, "step": 161400 }, { "grad_norm": 0.12695682048797607, "learning_rate": 4.7846195466208624e-05, "loss": 0.0085, "step": 161410 }, { "grad_norm": 0.08044762164354324, "learning_rate": 4.784068902911259e-05, "loss": 0.0123, "step": 161420 }, { "grad_norm": 0.09855548292398453, "learning_rate": 4.78351826182542e-05, "loss": 0.0156, "step": 161430 }, { "grad_norm": 0.1502295285463333, "learning_rate": 4.7829676233700396e-05, "loss": 0.0146, "step": 161440 }, { "grad_norm": 0.15686559677124023, "learning_rate": 4.7824169875518066e-05, "loss": 0.0136, "step": 161450 }, { "grad_norm": 0.197770357131958, "learning_rate": 4.781866354377415e-05, "loss": 0.0114, "step": 161460 }, { "grad_norm": 0.21981842815876007, "learning_rate": 4.781315723853551e-05, "loss": 0.0164, "step": 161470 }, { "grad_norm": 0.11623015254735947, "learning_rate": 4.7807650959869075e-05, "loss": 0.0127, "step": 161480 }, { "grad_norm": 0.11909837275743484, "learning_rate": 4.780214470784175e-05, "loss": 0.0203, "step": 161490 }, { "grad_norm": 0.07395562529563904, "learning_rate": 4.779663848252047e-05, "loss": 0.0159, "step": 161500 }, { "grad_norm": 0.10061877965927124, "learning_rate": 4.779113228397208e-05, "loss": 0.0202, "step": 161510 }, { "grad_norm": 0.10220920294523239, "learning_rate": 4.778562611226353e-05, "loss": 0.0112, "step": 161520 }, { "grad_norm": 0.09354131668806076, "learning_rate": 4.7780119967461726e-05, "loss": 0.0142, "step": 161530 }, { "grad_norm": 0.12866157293319702, "learning_rate": 4.7774613849633546e-05, "loss": 0.0165, "step": 161540 }, { "grad_norm": 0.08987855166196823, "learning_rate": 4.7769107758845896e-05, "loss": 0.0142, "step": 161550 }, { "grad_norm": 0.07830806076526642, "learning_rate": 4.776360169516571e-05, "loss": 0.0171, "step": 161560 }, { "grad_norm": 0.07818891853094101, "learning_rate": 4.7758095658659886e-05, "loss": 0.0118, "step": 161570 }, { "grad_norm": 0.12343993782997131, "learning_rate": 4.775258964939529e-05, "loss": 0.0093, "step": 161580 }, { "grad_norm": 0.12355256080627441, "learning_rate": 4.7747083667438865e-05, "loss": 0.0113, "step": 161590 }, { "grad_norm": 0.07673162966966629, "learning_rate": 4.774157771285751e-05, "loss": 0.0116, "step": 161600 }, { "grad_norm": 0.0926830917596817, "learning_rate": 4.773607178571811e-05, "loss": 0.016, "step": 161610 }, { "grad_norm": 0.17917613685131073, "learning_rate": 4.773056588608756e-05, "loss": 0.0214, "step": 161620 }, { "grad_norm": 0.15355844795703888, "learning_rate": 4.772506001403279e-05, "loss": 0.014, "step": 161630 }, { "grad_norm": 0.06642365455627441, "learning_rate": 4.77195541696207e-05, "loss": 0.0151, "step": 161640 }, { "grad_norm": 0.2758347690105438, "learning_rate": 4.7714048352918164e-05, "loss": 0.0153, "step": 161650 }, { "grad_norm": 0.1748768538236618, "learning_rate": 4.77085425639921e-05, "loss": 0.0144, "step": 161660 }, { "grad_norm": 0.12024268507957458, "learning_rate": 4.770303680290942e-05, "loss": 0.0121, "step": 161670 }, { "grad_norm": 0.09882012009620667, "learning_rate": 4.7697531069737e-05, "loss": 0.012, "step": 161680 }, { "grad_norm": 0.08172624558210373, "learning_rate": 4.769202536454175e-05, "loss": 0.0156, "step": 161690 }, { "grad_norm": 0.15575596690177917, "learning_rate": 4.768651968739057e-05, "loss": 0.0187, "step": 161700 }, { "grad_norm": 0.18211135268211365, "learning_rate": 4.768101403835038e-05, "loss": 0.0149, "step": 161710 }, { "grad_norm": 0.09140272438526154, "learning_rate": 4.7675508417488035e-05, "loss": 0.0131, "step": 161720 }, { "grad_norm": 0.13005079329013824, "learning_rate": 4.767000282487046e-05, "loss": 0.0135, "step": 161730 }, { "grad_norm": 0.2732502818107605, "learning_rate": 4.7664497260564576e-05, "loss": 0.0154, "step": 161740 }, { "grad_norm": 0.10827823728322983, "learning_rate": 4.765899172463723e-05, "loss": 0.0144, "step": 161750 }, { "grad_norm": 0.08799377083778381, "learning_rate": 4.765348621715534e-05, "loss": 0.0171, "step": 161760 }, { "grad_norm": 0.10398037731647491, "learning_rate": 4.764798073818582e-05, "loss": 0.0174, "step": 161770 }, { "grad_norm": 0.10308049619197845, "learning_rate": 4.7642475287795566e-05, "loss": 0.0118, "step": 161780 }, { "grad_norm": 0.15382051467895508, "learning_rate": 4.7636969866051446e-05, "loss": 0.015, "step": 161790 }, { "grad_norm": 0.06649608165025711, "learning_rate": 4.7631464473020373e-05, "loss": 0.015, "step": 161800 }, { "grad_norm": 0.1373852789402008, "learning_rate": 4.7625959108769245e-05, "loss": 0.0154, "step": 161810 }, { "grad_norm": 0.10261056572198868, "learning_rate": 4.762045377336497e-05, "loss": 0.0103, "step": 161820 }, { "grad_norm": 0.0792509987950325, "learning_rate": 4.761494846687442e-05, "loss": 0.0127, "step": 161830 }, { "grad_norm": 0.08442133665084839, "learning_rate": 4.7609443189364484e-05, "loss": 0.0105, "step": 161840 }, { "grad_norm": 0.23930981755256653, "learning_rate": 4.760393794090208e-05, "loss": 0.0135, "step": 161850 }, { "grad_norm": 0.08412202447652817, "learning_rate": 4.7598432721554106e-05, "loss": 0.012, "step": 161860 }, { "grad_norm": 0.09907028824090958, "learning_rate": 4.7592927531387424e-05, "loss": 0.0156, "step": 161870 }, { "grad_norm": 0.4449041187763214, "learning_rate": 4.758742237046894e-05, "loss": 0.0118, "step": 161880 }, { "grad_norm": 0.29326802492141724, "learning_rate": 4.758191723886557e-05, "loss": 0.0188, "step": 161890 }, { "grad_norm": 0.16086186468601227, "learning_rate": 4.757641213664418e-05, "loss": 0.0197, "step": 161900 }, { "grad_norm": 0.2293139100074768, "learning_rate": 4.7570907063871666e-05, "loss": 0.0155, "step": 161910 }, { "grad_norm": 0.16445408761501312, "learning_rate": 4.7565402020614935e-05, "loss": 0.0199, "step": 161920 }, { "grad_norm": 0.07470729202032089, "learning_rate": 4.755989700694086e-05, "loss": 0.0124, "step": 161930 }, { "grad_norm": 0.1537066102027893, "learning_rate": 4.755439202291635e-05, "loss": 0.014, "step": 161940 }, { "grad_norm": 0.26624834537506104, "learning_rate": 4.754888706860827e-05, "loss": 0.0133, "step": 161950 }, { "grad_norm": 0.05823301151394844, "learning_rate": 4.7543382144083544e-05, "loss": 0.0208, "step": 161960 }, { "grad_norm": 0.20865172147750854, "learning_rate": 4.753787724940903e-05, "loss": 0.0102, "step": 161970 }, { "grad_norm": 0.2669254541397095, "learning_rate": 4.753237238465164e-05, "loss": 0.0146, "step": 161980 }, { "grad_norm": 0.09857779741287231, "learning_rate": 4.752686754987826e-05, "loss": 0.0107, "step": 161990 }, { "grad_norm": 0.15923848748207092, "learning_rate": 4.752136274515576e-05, "loss": 0.0162, "step": 162000 }, { "grad_norm": 0.12353336811065674, "learning_rate": 4.7515857970551045e-05, "loss": 0.012, "step": 162010 }, { "grad_norm": 0.13711650669574738, "learning_rate": 4.7510353226131014e-05, "loss": 0.011, "step": 162020 }, { "grad_norm": 0.11009115725755692, "learning_rate": 4.750484851196254e-05, "loss": 0.012, "step": 162030 }, { "grad_norm": 0.12444918602705002, "learning_rate": 4.74993438281125e-05, "loss": 0.0143, "step": 162040 }, { "grad_norm": 0.1477126181125641, "learning_rate": 4.7493839174647795e-05, "loss": 0.0168, "step": 162050 }, { "grad_norm": 0.1341046839952469, "learning_rate": 4.748833455163532e-05, "loss": 0.0187, "step": 162060 }, { "grad_norm": 0.1848071664571762, "learning_rate": 4.748282995914196e-05, "loss": 0.0134, "step": 162070 }, { "grad_norm": 0.13765615224838257, "learning_rate": 4.7477325397234575e-05, "loss": 0.0132, "step": 162080 }, { "grad_norm": 0.10655226558446884, "learning_rate": 4.747182086598007e-05, "loss": 0.014, "step": 162090 }, { "grad_norm": 0.11092812567949295, "learning_rate": 4.746631636544534e-05, "loss": 0.0213, "step": 162100 }, { "grad_norm": 0.13768194615840912, "learning_rate": 4.746081189569724e-05, "loss": 0.019, "step": 162110 }, { "grad_norm": 0.07573354989290237, "learning_rate": 4.7455307456802685e-05, "loss": 0.0126, "step": 162120 }, { "grad_norm": 0.17047882080078125, "learning_rate": 4.744980304882854e-05, "loss": 0.0179, "step": 162130 }, { "grad_norm": 0.1836443692445755, "learning_rate": 4.7444298671841716e-05, "loss": 0.0241, "step": 162140 }, { "grad_norm": 0.0822155624628067, "learning_rate": 4.743879432590905e-05, "loss": 0.0149, "step": 162150 }, { "grad_norm": 0.13210037350654602, "learning_rate": 4.743329001109746e-05, "loss": 0.0182, "step": 162160 }, { "grad_norm": 0.14604395627975464, "learning_rate": 4.742778572747384e-05, "loss": 0.0151, "step": 162170 }, { "grad_norm": 0.14140257239341736, "learning_rate": 4.7422281475105025e-05, "loss": 0.0149, "step": 162180 }, { "grad_norm": 0.06584099680185318, "learning_rate": 4.7416777254057924e-05, "loss": 0.0118, "step": 162190 }, { "grad_norm": 0.08500893414020538, "learning_rate": 4.7411273064399424e-05, "loss": 0.0096, "step": 162200 }, { "grad_norm": 0.16252215206623077, "learning_rate": 4.7405768906196415e-05, "loss": 0.0106, "step": 162210 }, { "grad_norm": 0.06262930482625961, "learning_rate": 4.740026477951575e-05, "loss": 0.016, "step": 162220 }, { "grad_norm": 0.1634882241487503, "learning_rate": 4.739476068442432e-05, "loss": 0.0127, "step": 162230 }, { "grad_norm": 0.1202324628829956, "learning_rate": 4.738925662098902e-05, "loss": 0.014, "step": 162240 }, { "grad_norm": 0.11958383023738861, "learning_rate": 4.738375258927671e-05, "loss": 0.0169, "step": 162250 }, { "grad_norm": 0.15047334134578705, "learning_rate": 4.737824858935427e-05, "loss": 0.0121, "step": 162260 }, { "grad_norm": 0.1290251612663269, "learning_rate": 4.737274462128858e-05, "loss": 0.0142, "step": 162270 }, { "grad_norm": 0.08964484184980392, "learning_rate": 4.7367240685146546e-05, "loss": 0.0129, "step": 162280 }, { "grad_norm": 0.12785190343856812, "learning_rate": 4.736173678099501e-05, "loss": 0.0092, "step": 162290 }, { "grad_norm": 0.14814884960651398, "learning_rate": 4.735623290890086e-05, "loss": 0.0115, "step": 162300 }, { "grad_norm": 0.0755804181098938, "learning_rate": 4.735072906893099e-05, "loss": 0.0141, "step": 162310 }, { "grad_norm": 0.12633204460144043, "learning_rate": 4.734522526115225e-05, "loss": 0.0155, "step": 162320 }, { "grad_norm": 0.098869189620018, "learning_rate": 4.733972148563153e-05, "loss": 0.016, "step": 162330 }, { "grad_norm": 0.12085097283124924, "learning_rate": 4.73342177424357e-05, "loss": 0.0141, "step": 162340 }, { "grad_norm": 0.10813922435045242, "learning_rate": 4.732871403163166e-05, "loss": 0.0156, "step": 162350 }, { "grad_norm": 0.10978168249130249, "learning_rate": 4.732321035328625e-05, "loss": 0.0142, "step": 162360 }, { "grad_norm": 0.1977674663066864, "learning_rate": 4.7317706707466365e-05, "loss": 0.0209, "step": 162370 }, { "grad_norm": 0.11147972196340561, "learning_rate": 4.731220309423887e-05, "loss": 0.012, "step": 162380 }, { "grad_norm": 0.13453538715839386, "learning_rate": 4.730669951367067e-05, "loss": 0.0153, "step": 162390 }, { "grad_norm": 0.07685104012489319, "learning_rate": 4.730119596582859e-05, "loss": 0.0121, "step": 162400 }, { "grad_norm": 0.17961041629314423, "learning_rate": 4.729569245077953e-05, "loss": 0.0134, "step": 162410 }, { "grad_norm": 0.3177962303161621, "learning_rate": 4.7290188968590385e-05, "loss": 0.0122, "step": 162420 }, { "grad_norm": 0.19698019325733185, "learning_rate": 4.728468551932798e-05, "loss": 0.0163, "step": 162430 }, { "grad_norm": 0.10612570494413376, "learning_rate": 4.72791821030592e-05, "loss": 0.0138, "step": 162440 }, { "grad_norm": 0.12848767638206482, "learning_rate": 4.727367871985093e-05, "loss": 0.0134, "step": 162450 }, { "grad_norm": 0.11682768166065216, "learning_rate": 4.7268175369770066e-05, "loss": 0.0117, "step": 162460 }, { "grad_norm": 0.09624873846769333, "learning_rate": 4.7262672052883425e-05, "loss": 0.0177, "step": 162470 }, { "grad_norm": 0.11805964261293411, "learning_rate": 4.72571687692579e-05, "loss": 0.0121, "step": 162480 }, { "grad_norm": 0.15658922493457794, "learning_rate": 4.725166551896039e-05, "loss": 0.019, "step": 162490 }, { "grad_norm": 0.08300770074129105, "learning_rate": 4.724616230205772e-05, "loss": 0.017, "step": 162500 }, { "grad_norm": 0.09977899491786957, "learning_rate": 4.724065911861677e-05, "loss": 0.0137, "step": 162510 }, { "grad_norm": 0.10273914039134979, "learning_rate": 4.723515596870442e-05, "loss": 0.0151, "step": 162520 }, { "grad_norm": 0.09983590245246887, "learning_rate": 4.722965285238756e-05, "loss": 0.0141, "step": 162530 }, { "grad_norm": 0.07041700184345245, "learning_rate": 4.7224149769733007e-05, "loss": 0.0111, "step": 162540 }, { "grad_norm": 0.10614906251430511, "learning_rate": 4.721864672080766e-05, "loss": 0.0146, "step": 162550 }, { "grad_norm": 0.14283034205436707, "learning_rate": 4.721314370567839e-05, "loss": 0.018, "step": 162560 }, { "grad_norm": 0.08545304089784622, "learning_rate": 4.720764072441205e-05, "loss": 0.0121, "step": 162570 }, { "grad_norm": 0.06752676516771317, "learning_rate": 4.720213777707551e-05, "loss": 0.0166, "step": 162580 }, { "grad_norm": 0.08391569554805756, "learning_rate": 4.7196634863735625e-05, "loss": 0.0128, "step": 162590 }, { "grad_norm": 0.1641133725643158, "learning_rate": 4.7191131984459294e-05, "loss": 0.0216, "step": 162600 }, { "grad_norm": 0.10790902376174927, "learning_rate": 4.718562913931336e-05, "loss": 0.0109, "step": 162610 }, { "grad_norm": 0.14644429087638855, "learning_rate": 4.718012632836468e-05, "loss": 0.0138, "step": 162620 }, { "grad_norm": 0.11253193020820618, "learning_rate": 4.717462355168014e-05, "loss": 0.0187, "step": 162630 }, { "grad_norm": 0.0933515802025795, "learning_rate": 4.7169120809326574e-05, "loss": 0.0105, "step": 162640 }, { "grad_norm": 0.09287738054990768, "learning_rate": 4.7163618101370876e-05, "loss": 0.0113, "step": 162650 }, { "grad_norm": 0.13399174809455872, "learning_rate": 4.715811542787989e-05, "loss": 0.0162, "step": 162660 }, { "grad_norm": 0.11417143046855927, "learning_rate": 4.715261278892049e-05, "loss": 0.018, "step": 162670 }, { "grad_norm": 0.0641699731349945, "learning_rate": 4.714711018455952e-05, "loss": 0.0176, "step": 162680 }, { "grad_norm": 0.19840966165065765, "learning_rate": 4.714160761486387e-05, "loss": 0.0147, "step": 162690 }, { "grad_norm": 0.15798471868038177, "learning_rate": 4.713610507990038e-05, "loss": 0.0187, "step": 162700 }, { "grad_norm": 0.3638732433319092, "learning_rate": 4.7130602579735923e-05, "loss": 0.0187, "step": 162710 }, { "grad_norm": 0.12358827143907547, "learning_rate": 4.7125100114437337e-05, "loss": 0.0145, "step": 162720 }, { "grad_norm": 0.05819198861718178, "learning_rate": 4.711959768407152e-05, "loss": 0.0137, "step": 162730 }, { "grad_norm": 0.1181386411190033, "learning_rate": 4.711409528870531e-05, "loss": 0.0128, "step": 162740 }, { "grad_norm": 0.1034308671951294, "learning_rate": 4.710859292840555e-05, "loss": 0.0121, "step": 162750 }, { "grad_norm": 0.11811546981334686, "learning_rate": 4.710309060323912e-05, "loss": 0.0114, "step": 162760 }, { "grad_norm": 0.11285314708948135, "learning_rate": 4.709758831327288e-05, "loss": 0.0159, "step": 162770 }, { "grad_norm": 0.1510782092809677, "learning_rate": 4.709208605857369e-05, "loss": 0.0155, "step": 162780 }, { "grad_norm": 0.14269593358039856, "learning_rate": 4.708658383920838e-05, "loss": 0.0184, "step": 162790 }, { "grad_norm": 0.08549647033214569, "learning_rate": 4.708108165524383e-05, "loss": 0.0135, "step": 162800 }, { "grad_norm": 0.35333889722824097, "learning_rate": 4.7075579506746917e-05, "loss": 0.0156, "step": 162810 }, { "grad_norm": 0.10724321752786636, "learning_rate": 4.707007739378446e-05, "loss": 0.0166, "step": 162820 }, { "grad_norm": 0.09152097254991531, "learning_rate": 4.706457531642332e-05, "loss": 0.0136, "step": 162830 }, { "grad_norm": 0.06305861473083496, "learning_rate": 4.705907327473036e-05, "loss": 0.0139, "step": 162840 }, { "grad_norm": 0.27580899000167847, "learning_rate": 4.705357126877246e-05, "loss": 0.0189, "step": 162850 }, { "grad_norm": 0.13018129765987396, "learning_rate": 4.7048069298616434e-05, "loss": 0.0168, "step": 162860 }, { "grad_norm": 0.0836523100733757, "learning_rate": 4.704256736432915e-05, "loss": 0.0152, "step": 162870 }, { "grad_norm": 0.10978614538908005, "learning_rate": 4.7037065465977485e-05, "loss": 0.0166, "step": 162880 }, { "grad_norm": 0.08827633410692215, "learning_rate": 4.7031563603628256e-05, "loss": 0.0167, "step": 162890 }, { "grad_norm": 0.0979287326335907, "learning_rate": 4.7026061777348326e-05, "loss": 0.0106, "step": 162900 }, { "grad_norm": 0.10055143386125565, "learning_rate": 4.702055998720456e-05, "loss": 0.0201, "step": 162910 }, { "grad_norm": 0.20137110352516174, "learning_rate": 4.7015058233263815e-05, "loss": 0.0149, "step": 162920 }, { "grad_norm": 0.11237872391939163, "learning_rate": 4.700955651559291e-05, "loss": 0.018, "step": 162930 }, { "grad_norm": 0.19525665044784546, "learning_rate": 4.700405483425872e-05, "loss": 0.0168, "step": 162940 }, { "grad_norm": 0.08224686235189438, "learning_rate": 4.699855318932811e-05, "loss": 0.0149, "step": 162950 }, { "grad_norm": 0.08831794559955597, "learning_rate": 4.69930515808679e-05, "loss": 0.0132, "step": 162960 }, { "grad_norm": 0.19035083055496216, "learning_rate": 4.698755000894495e-05, "loss": 0.0135, "step": 162970 }, { "grad_norm": 0.07898271083831787, "learning_rate": 4.698204847362611e-05, "loss": 0.013, "step": 162980 }, { "grad_norm": 0.1304837167263031, "learning_rate": 4.6976546974978256e-05, "loss": 0.0212, "step": 162990 }, { "grad_norm": 0.1824943572282791, "learning_rate": 4.6971045513068176e-05, "loss": 0.0147, "step": 163000 }, { "grad_norm": 0.23355722427368164, "learning_rate": 4.696554408796276e-05, "loss": 0.0125, "step": 163010 }, { "grad_norm": 0.1491895169019699, "learning_rate": 4.6960042699728855e-05, "loss": 0.0136, "step": 163020 }, { "grad_norm": 0.3203468322753906, "learning_rate": 4.6954541348433316e-05, "loss": 0.024, "step": 163030 }, { "grad_norm": 0.08118993788957596, "learning_rate": 4.694904003414296e-05, "loss": 0.0203, "step": 163040 }, { "grad_norm": 0.1066608801484108, "learning_rate": 4.6943538756924647e-05, "loss": 0.0102, "step": 163050 }, { "grad_norm": 0.05374867096543312, "learning_rate": 4.693803751684524e-05, "loss": 0.0118, "step": 163060 }, { "grad_norm": 0.12638135254383087, "learning_rate": 4.693253631397155e-05, "loss": 0.0146, "step": 163070 }, { "grad_norm": 0.08854157477617264, "learning_rate": 4.6927035148370434e-05, "loss": 0.0161, "step": 163080 }, { "grad_norm": 0.18655118346214294, "learning_rate": 4.6921534020108746e-05, "loss": 0.0204, "step": 163090 }, { "grad_norm": 0.12204229831695557, "learning_rate": 4.691603292925335e-05, "loss": 0.0162, "step": 163100 }, { "grad_norm": 0.0720599889755249, "learning_rate": 4.691053187587105e-05, "loss": 0.0127, "step": 163110 }, { "grad_norm": 0.1147075966000557, "learning_rate": 4.690503086002869e-05, "loss": 0.0123, "step": 163120 }, { "grad_norm": 0.08679351210594177, "learning_rate": 4.689952988179316e-05, "loss": 0.0197, "step": 163130 }, { "grad_norm": 0.1735379695892334, "learning_rate": 4.6894028941231236e-05, "loss": 0.0173, "step": 163140 }, { "grad_norm": 0.09418459981679916, "learning_rate": 4.688852803840981e-05, "loss": 0.015, "step": 163150 }, { "grad_norm": 0.0766821876168251, "learning_rate": 4.688302717339569e-05, "loss": 0.0107, "step": 163160 }, { "grad_norm": 0.05879239737987518, "learning_rate": 4.6877526346255763e-05, "loss": 0.0119, "step": 163170 }, { "grad_norm": 0.05025111138820648, "learning_rate": 4.687202555705682e-05, "loss": 0.011, "step": 163180 }, { "grad_norm": 0.05787859112024307, "learning_rate": 4.686652480586571e-05, "loss": 0.0159, "step": 163190 }, { "grad_norm": 0.056571148335933685, "learning_rate": 4.686102409274931e-05, "loss": 0.0115, "step": 163200 }, { "grad_norm": 0.20500501990318298, "learning_rate": 4.685552341777441e-05, "loss": 0.0159, "step": 163210 }, { "grad_norm": 0.0718003660440445, "learning_rate": 4.685002278100787e-05, "loss": 0.0135, "step": 163220 }, { "grad_norm": 0.142069011926651, "learning_rate": 4.684452218251652e-05, "loss": 0.0139, "step": 163230 }, { "grad_norm": 0.15986795723438263, "learning_rate": 4.683902162236723e-05, "loss": 0.0128, "step": 163240 }, { "grad_norm": 0.08199362456798553, "learning_rate": 4.683352110062679e-05, "loss": 0.0156, "step": 163250 }, { "grad_norm": 0.0748877301812172, "learning_rate": 4.682802061736206e-05, "loss": 0.0156, "step": 163260 }, { "grad_norm": 0.08267389237880707, "learning_rate": 4.682252017263988e-05, "loss": 0.011, "step": 163270 }, { "grad_norm": 0.14657655358314514, "learning_rate": 4.6817019766527095e-05, "loss": 0.0099, "step": 163280 }, { "grad_norm": 0.06349551677703857, "learning_rate": 4.6811519399090506e-05, "loss": 0.0129, "step": 163290 }, { "grad_norm": 0.1547335535287857, "learning_rate": 4.680601907039696e-05, "loss": 0.0142, "step": 163300 }, { "grad_norm": 0.1770123392343521, "learning_rate": 4.680051878051331e-05, "loss": 0.0132, "step": 163310 }, { "grad_norm": 0.06812764704227448, "learning_rate": 4.679501852950638e-05, "loss": 0.014, "step": 163320 }, { "grad_norm": 0.11143157631158829, "learning_rate": 4.6789518317443e-05, "loss": 0.0177, "step": 163330 }, { "grad_norm": 0.08979782462120056, "learning_rate": 4.6784018144389994e-05, "loss": 0.0157, "step": 163340 }, { "grad_norm": 0.1523895561695099, "learning_rate": 4.677851801041421e-05, "loss": 0.0102, "step": 163350 }, { "grad_norm": 0.11892152577638626, "learning_rate": 4.6773017915582484e-05, "loss": 0.0155, "step": 163360 }, { "grad_norm": 0.05711865425109863, "learning_rate": 4.676751785996162e-05, "loss": 0.0117, "step": 163370 }, { "grad_norm": 0.09543472528457642, "learning_rate": 4.676201784361848e-05, "loss": 0.0175, "step": 163380 }, { "grad_norm": 0.07749902456998825, "learning_rate": 4.6756517866619864e-05, "loss": 0.0194, "step": 163390 }, { "grad_norm": 0.10578692704439163, "learning_rate": 4.6751017929032635e-05, "loss": 0.0159, "step": 163400 }, { "grad_norm": 0.10760363191366196, "learning_rate": 4.67455180309236e-05, "loss": 0.013, "step": 163410 }, { "grad_norm": 0.26156187057495117, "learning_rate": 4.674001817235959e-05, "loss": 0.018, "step": 163420 }, { "grad_norm": 0.26246488094329834, "learning_rate": 4.6734518353407435e-05, "loss": 0.016, "step": 163430 }, { "grad_norm": 0.209719717502594, "learning_rate": 4.672901857413398e-05, "loss": 0.0158, "step": 163440 }, { "grad_norm": 0.10677753388881683, "learning_rate": 4.672351883460603e-05, "loss": 0.0112, "step": 163450 }, { "grad_norm": 0.09995748847723007, "learning_rate": 4.671801913489041e-05, "loss": 0.0172, "step": 163460 }, { "grad_norm": 0.21285380423069, "learning_rate": 4.671251947505396e-05, "loss": 0.0164, "step": 163470 }, { "grad_norm": 0.17989489436149597, "learning_rate": 4.6707019855163505e-05, "loss": 0.0163, "step": 163480 }, { "grad_norm": 0.10478521138429642, "learning_rate": 4.670152027528588e-05, "loss": 0.0148, "step": 163490 }, { "grad_norm": 0.24133439362049103, "learning_rate": 4.669602073548788e-05, "loss": 0.0132, "step": 163500 }, { "grad_norm": 0.07861169427633286, "learning_rate": 4.669052123583634e-05, "loss": 0.0169, "step": 163510 }, { "grad_norm": 0.2875151038169861, "learning_rate": 4.6685021776398126e-05, "loss": 0.0159, "step": 163520 }, { "grad_norm": 0.13892953097820282, "learning_rate": 4.667952235724e-05, "loss": 0.0161, "step": 163530 }, { "grad_norm": 0.24113784730434418, "learning_rate": 4.667402297842882e-05, "loss": 0.0144, "step": 163540 }, { "grad_norm": 0.0689358189702034, "learning_rate": 4.666852364003139e-05, "loss": 0.0111, "step": 163550 }, { "grad_norm": 0.1892528086900711, "learning_rate": 4.666302434211457e-05, "loss": 0.0166, "step": 163560 }, { "grad_norm": 0.3582090735435486, "learning_rate": 4.665752508474513e-05, "loss": 0.0124, "step": 163570 }, { "grad_norm": 0.09237365424633026, "learning_rate": 4.665202586798992e-05, "loss": 0.0111, "step": 163580 }, { "grad_norm": 0.06434981524944305, "learning_rate": 4.6646526691915755e-05, "loss": 0.0088, "step": 163590 }, { "grad_norm": 0.0932386964559555, "learning_rate": 4.664102755658948e-05, "loss": 0.0123, "step": 163600 }, { "grad_norm": 0.12160881608724594, "learning_rate": 4.6635528462077866e-05, "loss": 0.0114, "step": 163610 }, { "grad_norm": 0.061907872557640076, "learning_rate": 4.663002940844776e-05, "loss": 0.0098, "step": 163620 }, { "grad_norm": 0.11788845807313919, "learning_rate": 4.6624530395766e-05, "loss": 0.0146, "step": 163630 }, { "grad_norm": 0.10310281068086624, "learning_rate": 4.661903142409936e-05, "loss": 0.0125, "step": 163640 }, { "grad_norm": 0.08323375135660172, "learning_rate": 4.6613532493514685e-05, "loss": 0.0201, "step": 163650 }, { "grad_norm": 0.10949553549289703, "learning_rate": 4.660803360407878e-05, "loss": 0.0115, "step": 163660 }, { "grad_norm": 0.71239173412323, "learning_rate": 4.66025347558585e-05, "loss": 0.0157, "step": 163670 }, { "grad_norm": 0.09782083332538605, "learning_rate": 4.6597035948920606e-05, "loss": 0.0158, "step": 163680 }, { "grad_norm": 0.13310885429382324, "learning_rate": 4.659153718333194e-05, "loss": 0.0127, "step": 163690 }, { "grad_norm": 0.1101999506354332, "learning_rate": 4.658603845915933e-05, "loss": 0.0128, "step": 163700 }, { "grad_norm": 0.1388559192419052, "learning_rate": 4.658053977646956e-05, "loss": 0.0135, "step": 163710 }, { "grad_norm": 0.22091610729694366, "learning_rate": 4.657504113532946e-05, "loss": 0.0228, "step": 163720 }, { "grad_norm": 0.20005378127098083, "learning_rate": 4.656954253580584e-05, "loss": 0.0133, "step": 163730 }, { "grad_norm": 0.14829213917255402, "learning_rate": 4.656404397796554e-05, "loss": 0.0147, "step": 163740 }, { "grad_norm": 0.15694840252399445, "learning_rate": 4.6558545461875324e-05, "loss": 0.0117, "step": 163750 }, { "grad_norm": 0.12534648180007935, "learning_rate": 4.6553046987602035e-05, "loss": 0.0134, "step": 163760 }, { "grad_norm": 0.1312282681465149, "learning_rate": 4.6547548555212505e-05, "loss": 0.0149, "step": 163770 }, { "grad_norm": 0.2770776152610779, "learning_rate": 4.6542050164773495e-05, "loss": 0.0166, "step": 163780 }, { "grad_norm": 0.11372493207454681, "learning_rate": 4.653655181635184e-05, "loss": 0.0172, "step": 163790 }, { "grad_norm": 0.0947912260890007, "learning_rate": 4.6531053510014344e-05, "loss": 0.0103, "step": 163800 }, { "grad_norm": 0.7143014073371887, "learning_rate": 4.652555524582785e-05, "loss": 0.0171, "step": 163810 }, { "grad_norm": 0.19498898088932037, "learning_rate": 4.652005702385912e-05, "loss": 0.015, "step": 163820 }, { "grad_norm": 0.1505073606967926, "learning_rate": 4.6514558844174986e-05, "loss": 0.0183, "step": 163830 }, { "grad_norm": 0.10294323414564133, "learning_rate": 4.650906070684227e-05, "loss": 0.0116, "step": 163840 }, { "grad_norm": 0.17077700793743134, "learning_rate": 4.6503562611927734e-05, "loss": 0.0156, "step": 163850 }, { "grad_norm": 0.09568599611520767, "learning_rate": 4.6498064559498224e-05, "loss": 0.0148, "step": 163860 }, { "grad_norm": 0.23411259055137634, "learning_rate": 4.649256654962053e-05, "loss": 0.0134, "step": 163870 }, { "grad_norm": 0.0892353430390358, "learning_rate": 4.6487068582361494e-05, "loss": 0.0144, "step": 163880 }, { "grad_norm": 0.3696456849575043, "learning_rate": 4.648157065778786e-05, "loss": 0.0203, "step": 163890 }, { "grad_norm": 0.566776692867279, "learning_rate": 4.6476072775966474e-05, "loss": 0.0139, "step": 163900 }, { "grad_norm": 0.20662173628807068, "learning_rate": 4.647057493696412e-05, "loss": 0.0136, "step": 163910 }, { "grad_norm": 0.08684610575437546, "learning_rate": 4.646507714084764e-05, "loss": 0.0202, "step": 163920 }, { "grad_norm": 0.1616184264421463, "learning_rate": 4.645957938768379e-05, "loss": 0.0172, "step": 163930 }, { "grad_norm": 0.197796031832695, "learning_rate": 4.645408167753939e-05, "loss": 0.0137, "step": 163940 }, { "grad_norm": 0.08574465662240982, "learning_rate": 4.644858401048127e-05, "loss": 0.0125, "step": 163950 }, { "grad_norm": 0.11806670576334, "learning_rate": 4.6443086386576185e-05, "loss": 0.0137, "step": 163960 }, { "grad_norm": 0.06288644671440125, "learning_rate": 4.6437588805890956e-05, "loss": 0.0134, "step": 163970 }, { "grad_norm": 0.14396771788597107, "learning_rate": 4.643209126849239e-05, "loss": 0.0136, "step": 163980 }, { "grad_norm": 0.12396621704101562, "learning_rate": 4.64265937744473e-05, "loss": 0.0101, "step": 163990 }, { "grad_norm": 0.662406861782074, "learning_rate": 4.6421096323822455e-05, "loss": 0.0143, "step": 164000 }, { "grad_norm": 0.09242844581604004, "learning_rate": 4.641559891668467e-05, "loss": 0.0118, "step": 164010 }, { "grad_norm": 0.14796507358551025, "learning_rate": 4.6410101553100745e-05, "loss": 0.0158, "step": 164020 }, { "grad_norm": 0.5643093585968018, "learning_rate": 4.6404604233137474e-05, "loss": 0.0172, "step": 164030 }, { "grad_norm": 0.054406728595495224, "learning_rate": 4.6399106956861656e-05, "loss": 0.0103, "step": 164040 }, { "grad_norm": 0.04847550019621849, "learning_rate": 4.639360972434008e-05, "loss": 0.0135, "step": 164050 }, { "grad_norm": 0.07070325314998627, "learning_rate": 4.638811253563956e-05, "loss": 0.0111, "step": 164060 }, { "grad_norm": 0.18617881834506989, "learning_rate": 4.63826153908269e-05, "loss": 0.0165, "step": 164070 }, { "grad_norm": 0.13769283890724182, "learning_rate": 4.637711828996885e-05, "loss": 0.0093, "step": 164080 }, { "grad_norm": 0.12041405588388443, "learning_rate": 4.637162123313225e-05, "loss": 0.012, "step": 164090 }, { "grad_norm": 0.1078411117196083, "learning_rate": 4.636612422038387e-05, "loss": 0.0159, "step": 164100 }, { "grad_norm": 0.12416789680719376, "learning_rate": 4.636062725179052e-05, "loss": 0.0119, "step": 164110 }, { "grad_norm": 0.07938583195209503, "learning_rate": 4.635513032741897e-05, "loss": 0.0192, "step": 164120 }, { "grad_norm": 0.1130300909280777, "learning_rate": 4.634963344733604e-05, "loss": 0.0191, "step": 164130 }, { "grad_norm": 0.21319402754306793, "learning_rate": 4.63441366116085e-05, "loss": 0.0125, "step": 164140 }, { "grad_norm": 0.1088237464427948, "learning_rate": 4.6338639820303166e-05, "loss": 0.0155, "step": 164150 }, { "grad_norm": 0.1857554167509079, "learning_rate": 4.633314307348681e-05, "loss": 0.0132, "step": 164160 }, { "grad_norm": 0.1037997379899025, "learning_rate": 4.632764637122622e-05, "loss": 0.0158, "step": 164170 }, { "grad_norm": 0.07005921751260757, "learning_rate": 4.6322149713588194e-05, "loss": 0.0131, "step": 164180 }, { "grad_norm": 0.08077196776866913, "learning_rate": 4.631665310063953e-05, "loss": 0.0156, "step": 164190 }, { "grad_norm": 0.08124613761901855, "learning_rate": 4.6311156532447016e-05, "loss": 0.0153, "step": 164200 }, { "grad_norm": 0.13843241333961487, "learning_rate": 4.630566000907741e-05, "loss": 0.0182, "step": 164210 }, { "grad_norm": 0.12829574942588806, "learning_rate": 4.630016353059753e-05, "loss": 0.0095, "step": 164220 }, { "grad_norm": 0.13498364388942719, "learning_rate": 4.629466709707417e-05, "loss": 0.0114, "step": 164230 }, { "grad_norm": 0.12126173824071884, "learning_rate": 4.62891707085741e-05, "loss": 0.0132, "step": 164240 }, { "grad_norm": 0.13368405401706696, "learning_rate": 4.62836743651641e-05, "loss": 0.0169, "step": 164250 }, { "grad_norm": 0.15368565917015076, "learning_rate": 4.627817806691096e-05, "loss": 0.0148, "step": 164260 }, { "grad_norm": 0.1263275146484375, "learning_rate": 4.62726818138815e-05, "loss": 0.0205, "step": 164270 }, { "grad_norm": 0.11338971555233002, "learning_rate": 4.6267185606142444e-05, "loss": 0.0158, "step": 164280 }, { "grad_norm": 0.2000257521867752, "learning_rate": 4.6261689443760616e-05, "loss": 0.0156, "step": 164290 }, { "grad_norm": 0.06773374229669571, "learning_rate": 4.625619332680278e-05, "loss": 0.0136, "step": 164300 }, { "grad_norm": 0.1295095980167389, "learning_rate": 4.6250697255335764e-05, "loss": 0.0125, "step": 164310 }, { "grad_norm": 0.3886997699737549, "learning_rate": 4.6245201229426285e-05, "loss": 0.0159, "step": 164320 }, { "grad_norm": 0.12860539555549622, "learning_rate": 4.6239705249141166e-05, "loss": 0.0169, "step": 164330 }, { "grad_norm": 0.06894701719284058, "learning_rate": 4.623420931454719e-05, "loss": 0.013, "step": 164340 }, { "grad_norm": 0.07246217876672745, "learning_rate": 4.622871342571111e-05, "loss": 0.0171, "step": 164350 }, { "grad_norm": 0.1335720270872116, "learning_rate": 4.622321758269972e-05, "loss": 0.0185, "step": 164360 }, { "grad_norm": 0.28044626116752625, "learning_rate": 4.6217721785579806e-05, "loss": 0.0168, "step": 164370 }, { "grad_norm": 0.11572499573230743, "learning_rate": 4.621222603441816e-05, "loss": 0.0132, "step": 164380 }, { "grad_norm": 0.09780042618513107, "learning_rate": 4.6206730329281525e-05, "loss": 0.0094, "step": 164390 }, { "grad_norm": 0.07091225683689117, "learning_rate": 4.62012346702367e-05, "loss": 0.0098, "step": 164400 }, { "grad_norm": 0.12862259149551392, "learning_rate": 4.619573905735048e-05, "loss": 0.0153, "step": 164410 }, { "grad_norm": 0.08930016309022903, "learning_rate": 4.61902434906896e-05, "loss": 0.0185, "step": 164420 }, { "grad_norm": 0.125261589884758, "learning_rate": 4.618474797032086e-05, "loss": 0.015, "step": 164430 }, { "grad_norm": 0.12742023169994354, "learning_rate": 4.6179252496311036e-05, "loss": 0.0163, "step": 164440 }, { "grad_norm": 0.10354825854301453, "learning_rate": 4.617375706872692e-05, "loss": 0.0165, "step": 164450 }, { "grad_norm": 0.10222303122282028, "learning_rate": 4.616826168763525e-05, "loss": 0.0115, "step": 164460 }, { "grad_norm": 0.1704077571630478, "learning_rate": 4.616276635310281e-05, "loss": 0.016, "step": 164470 }, { "grad_norm": 0.22978746891021729, "learning_rate": 4.61572710651964e-05, "loss": 0.0129, "step": 164480 }, { "grad_norm": 0.12422434240579605, "learning_rate": 4.615177582398278e-05, "loss": 0.0181, "step": 164490 }, { "grad_norm": 0.08227492868900299, "learning_rate": 4.61462806295287e-05, "loss": 0.0168, "step": 164500 }, { "grad_norm": 0.05125294253230095, "learning_rate": 4.614078548190095e-05, "loss": 0.0171, "step": 164510 }, { "grad_norm": 0.1278081238269806, "learning_rate": 4.613529038116632e-05, "loss": 0.0107, "step": 164520 }, { "grad_norm": 0.1499093770980835, "learning_rate": 4.6129795327391536e-05, "loss": 0.0153, "step": 164530 }, { "grad_norm": 0.10235781222581863, "learning_rate": 4.61243003206434e-05, "loss": 0.0147, "step": 164540 }, { "grad_norm": 0.12692314386367798, "learning_rate": 4.611880536098867e-05, "loss": 0.0126, "step": 164550 }, { "grad_norm": 0.1091608852148056, "learning_rate": 4.611331044849414e-05, "loss": 0.0178, "step": 164560 }, { "grad_norm": 0.4001905620098114, "learning_rate": 4.6107815583226535e-05, "loss": 0.0191, "step": 164570 }, { "grad_norm": 0.11054087430238724, "learning_rate": 4.610232076525265e-05, "loss": 0.009, "step": 164580 }, { "grad_norm": 0.14927417039871216, "learning_rate": 4.6096825994639264e-05, "loss": 0.0111, "step": 164590 }, { "grad_norm": 0.09631859511137009, "learning_rate": 4.609133127145311e-05, "loss": 0.0151, "step": 164600 }, { "grad_norm": 0.10103966295719147, "learning_rate": 4.608583659576097e-05, "loss": 0.0145, "step": 164610 }, { "grad_norm": 0.07376226782798767, "learning_rate": 4.6080341967629614e-05, "loss": 0.0183, "step": 164620 }, { "grad_norm": 0.13657087087631226, "learning_rate": 4.607484738712583e-05, "loss": 0.0169, "step": 164630 }, { "grad_norm": 0.11756652593612671, "learning_rate": 4.6069352854316324e-05, "loss": 0.0108, "step": 164640 }, { "grad_norm": 0.1049397811293602, "learning_rate": 4.606385836926789e-05, "loss": 0.0113, "step": 164650 }, { "grad_norm": 0.14485995471477509, "learning_rate": 4.6058363932047325e-05, "loss": 0.0141, "step": 164660 }, { "grad_norm": 0.10921449959278107, "learning_rate": 4.6052869542721334e-05, "loss": 0.0136, "step": 164670 }, { "grad_norm": 0.18889279663562775, "learning_rate": 4.60473752013567e-05, "loss": 0.0113, "step": 164680 }, { "grad_norm": 0.07061358541250229, "learning_rate": 4.6041880908020196e-05, "loss": 0.0158, "step": 164690 }, { "grad_norm": 0.2674305737018585, "learning_rate": 4.603638666277859e-05, "loss": 0.0183, "step": 164700 }, { "grad_norm": 0.12229006737470627, "learning_rate": 4.603089246569862e-05, "loss": 0.0149, "step": 164710 }, { "grad_norm": 0.11848139762878418, "learning_rate": 4.602539831684704e-05, "loss": 0.0122, "step": 164720 }, { "grad_norm": 0.1515754759311676, "learning_rate": 4.6019904216290636e-05, "loss": 0.0123, "step": 164730 }, { "grad_norm": 0.10489605367183685, "learning_rate": 4.601441016409616e-05, "loss": 0.0118, "step": 164740 }, { "grad_norm": 0.0859958603978157, "learning_rate": 4.600891616033035e-05, "loss": 0.0154, "step": 164750 }, { "grad_norm": 0.12126696854829788, "learning_rate": 4.6003422205059976e-05, "loss": 0.0162, "step": 164760 }, { "grad_norm": 0.07425307482481003, "learning_rate": 4.599792829835181e-05, "loss": 0.0109, "step": 164770 }, { "grad_norm": 0.07449278235435486, "learning_rate": 4.599243444027259e-05, "loss": 0.0116, "step": 164780 }, { "grad_norm": 0.09077659249305725, "learning_rate": 4.5986940630889066e-05, "loss": 0.0093, "step": 164790 }, { "grad_norm": 0.08697737008333206, "learning_rate": 4.5981446870268e-05, "loss": 0.0145, "step": 164800 }, { "grad_norm": 0.07169897854328156, "learning_rate": 4.597595315847616e-05, "loss": 0.017, "step": 164810 }, { "grad_norm": 0.05671899765729904, "learning_rate": 4.597045949558029e-05, "loss": 0.0132, "step": 164820 }, { "grad_norm": 0.14762334525585175, "learning_rate": 4.5964965881647134e-05, "loss": 0.0136, "step": 164830 }, { "grad_norm": 0.09750711172819138, "learning_rate": 4.595947231674346e-05, "loss": 0.0158, "step": 164840 }, { "grad_norm": 0.0768904760479927, "learning_rate": 4.5953978800936e-05, "loss": 0.0145, "step": 164850 }, { "grad_norm": 0.12065400928258896, "learning_rate": 4.594848533429153e-05, "loss": 0.0151, "step": 164860 }, { "grad_norm": 0.1309071183204651, "learning_rate": 4.5942991916876776e-05, "loss": 0.0201, "step": 164870 }, { "grad_norm": 0.15662814676761627, "learning_rate": 4.5937498548758514e-05, "loss": 0.0223, "step": 164880 }, { "grad_norm": 0.1171841248869896, "learning_rate": 4.593200523000347e-05, "loss": 0.0149, "step": 164890 }, { "grad_norm": 0.17898032069206238, "learning_rate": 4.592651196067841e-05, "loss": 0.0118, "step": 164900 }, { "grad_norm": 0.10367310792207718, "learning_rate": 4.5921018740850075e-05, "loss": 0.0113, "step": 164910 }, { "grad_norm": 0.08251751959323883, "learning_rate": 4.5915525570585205e-05, "loss": 0.0134, "step": 164920 }, { "grad_norm": 0.08313797414302826, "learning_rate": 4.5910032449950556e-05, "loss": 0.0122, "step": 164930 }, { "grad_norm": 0.12806183099746704, "learning_rate": 4.590453937901289e-05, "loss": 0.0107, "step": 164940 }, { "grad_norm": 0.1449667066335678, "learning_rate": 4.589904635783893e-05, "loss": 0.0164, "step": 164950 }, { "grad_norm": 0.15179717540740967, "learning_rate": 4.589355338649542e-05, "loss": 0.0127, "step": 164960 }, { "grad_norm": 0.077681764960289, "learning_rate": 4.588806046504912e-05, "loss": 0.0107, "step": 164970 }, { "grad_norm": 0.09161663800477982, "learning_rate": 4.588256759356678e-05, "loss": 0.0154, "step": 164980 }, { "grad_norm": 0.0826011672616005, "learning_rate": 4.5877074772115116e-05, "loss": 0.0155, "step": 164990 }, { "grad_norm": 0.16274182498455048, "learning_rate": 4.587158200076087e-05, "loss": 0.0156, "step": 165000 }, { "grad_norm": 0.12659665942192078, "learning_rate": 4.586608927957082e-05, "loss": 0.0178, "step": 165010 }, { "grad_norm": 0.21577399969100952, "learning_rate": 4.586059660861171e-05, "loss": 0.0137, "step": 165020 }, { "grad_norm": 0.15532249212265015, "learning_rate": 4.585510398795023e-05, "loss": 0.0143, "step": 165030 }, { "grad_norm": 0.1047993078827858, "learning_rate": 4.584961141765315e-05, "loss": 0.0107, "step": 165040 }, { "grad_norm": 0.08253828436136246, "learning_rate": 4.584411889778723e-05, "loss": 0.018, "step": 165050 }, { "grad_norm": 0.11881187558174133, "learning_rate": 4.583862642841917e-05, "loss": 0.0164, "step": 165060 }, { "grad_norm": 0.15170779824256897, "learning_rate": 4.583313400961573e-05, "loss": 0.0139, "step": 165070 }, { "grad_norm": 0.12420006841421127, "learning_rate": 4.5827641641443645e-05, "loss": 0.0109, "step": 165080 }, { "grad_norm": 0.1151583343744278, "learning_rate": 4.5822149323969676e-05, "loss": 0.0149, "step": 165090 }, { "grad_norm": 0.21030743420124054, "learning_rate": 4.581665705726051e-05, "loss": 0.0157, "step": 165100 }, { "grad_norm": 0.07404553890228271, "learning_rate": 4.581116484138291e-05, "loss": 0.0129, "step": 165110 }, { "grad_norm": 0.06947270035743713, "learning_rate": 4.580567267640362e-05, "loss": 0.0119, "step": 165120 }, { "grad_norm": 0.07982737571001053, "learning_rate": 4.580018056238939e-05, "loss": 0.0143, "step": 165130 }, { "grad_norm": 0.10265390574932098, "learning_rate": 4.57946884994069e-05, "loss": 0.0199, "step": 165140 }, { "grad_norm": 0.0732414498925209, "learning_rate": 4.5789196487522926e-05, "loss": 0.0148, "step": 165150 }, { "grad_norm": 0.08430738002061844, "learning_rate": 4.5783704526804204e-05, "loss": 0.016, "step": 165160 }, { "grad_norm": 0.09442855417728424, "learning_rate": 4.577821261731744e-05, "loss": 0.0132, "step": 165170 }, { "grad_norm": 0.06065520644187927, "learning_rate": 4.577272075912937e-05, "loss": 0.011, "step": 165180 }, { "grad_norm": 0.07762127369642258, "learning_rate": 4.576722895230673e-05, "loss": 0.0174, "step": 165190 }, { "grad_norm": 0.20071610808372498, "learning_rate": 4.576173719691628e-05, "loss": 0.0152, "step": 165200 }, { "grad_norm": 0.0962061807513237, "learning_rate": 4.5756245493024706e-05, "loss": 0.0134, "step": 165210 }, { "grad_norm": 0.12515844404697418, "learning_rate": 4.5750753840698754e-05, "loss": 0.0146, "step": 165220 }, { "grad_norm": 0.0640467181801796, "learning_rate": 4.574526224000518e-05, "loss": 0.014, "step": 165230 }, { "grad_norm": 0.11797534674406052, "learning_rate": 4.573977069101066e-05, "loss": 0.0143, "step": 165240 }, { "grad_norm": 0.0857706069946289, "learning_rate": 4.573427919378196e-05, "loss": 0.0128, "step": 165250 }, { "grad_norm": 0.15034744143486023, "learning_rate": 4.572878774838578e-05, "loss": 0.0143, "step": 165260 }, { "grad_norm": 0.09785298258066177, "learning_rate": 4.5723296354888894e-05, "loss": 0.0126, "step": 165270 }, { "grad_norm": 0.10317999124526978, "learning_rate": 4.571780501335797e-05, "loss": 0.0164, "step": 165280 }, { "grad_norm": 0.26345306634902954, "learning_rate": 4.5712313723859756e-05, "loss": 0.0183, "step": 165290 }, { "grad_norm": 0.1437520980834961, "learning_rate": 4.5706822486461004e-05, "loss": 0.0204, "step": 165300 }, { "grad_norm": 0.11404864490032196, "learning_rate": 4.570133130122839e-05, "loss": 0.0168, "step": 165310 }, { "grad_norm": 0.09406369924545288, "learning_rate": 4.569584016822865e-05, "loss": 0.0161, "step": 165320 }, { "grad_norm": 0.205752432346344, "learning_rate": 4.569034908752853e-05, "loss": 0.0147, "step": 165330 }, { "grad_norm": 0.08154415339231491, "learning_rate": 4.5684858059194754e-05, "loss": 0.0076, "step": 165340 }, { "grad_norm": 0.10684038698673248, "learning_rate": 4.5679367083294006e-05, "loss": 0.0127, "step": 165350 }, { "grad_norm": 0.06642898917198181, "learning_rate": 4.5673876159893025e-05, "loss": 0.015, "step": 165360 }, { "grad_norm": 0.29267850518226624, "learning_rate": 4.566838528905855e-05, "loss": 0.0194, "step": 165370 }, { "grad_norm": 0.12124817073345184, "learning_rate": 4.566289447085727e-05, "loss": 0.0127, "step": 165380 }, { "grad_norm": 0.08569636195898056, "learning_rate": 4.565740370535591e-05, "loss": 0.0146, "step": 165390 }, { "grad_norm": 0.12954865396022797, "learning_rate": 4.56519129926212e-05, "loss": 0.018, "step": 165400 }, { "grad_norm": 0.09812096506357193, "learning_rate": 4.564642233271987e-05, "loss": 0.0178, "step": 165410 }, { "grad_norm": 0.164106547832489, "learning_rate": 4.564093172571861e-05, "loss": 0.0196, "step": 165420 }, { "grad_norm": 0.12170698493719101, "learning_rate": 4.563544117168413e-05, "loss": 0.019, "step": 165430 }, { "grad_norm": 0.09142036736011505, "learning_rate": 4.5629950670683166e-05, "loss": 0.0172, "step": 165440 }, { "grad_norm": 0.2533232271671295, "learning_rate": 4.562446022278245e-05, "loss": 0.0132, "step": 165450 }, { "grad_norm": 0.09147845208644867, "learning_rate": 4.5618969828048655e-05, "loss": 0.0166, "step": 165460 }, { "grad_norm": 0.1703251153230667, "learning_rate": 4.561347948654851e-05, "loss": 0.0167, "step": 165470 }, { "grad_norm": 0.07528239488601685, "learning_rate": 4.560798919834874e-05, "loss": 0.02, "step": 165480 }, { "grad_norm": 0.19332870841026306, "learning_rate": 4.560249896351605e-05, "loss": 0.0171, "step": 165490 }, { "grad_norm": 0.1487593948841095, "learning_rate": 4.5597008782117134e-05, "loss": 0.0186, "step": 165500 }, { "grad_norm": 0.07238736003637314, "learning_rate": 4.5591518654218726e-05, "loss": 0.0121, "step": 165510 }, { "grad_norm": 0.13535280525684357, "learning_rate": 4.558602857988753e-05, "loss": 0.0124, "step": 165520 }, { "grad_norm": 0.1901242434978485, "learning_rate": 4.558053855919026e-05, "loss": 0.0161, "step": 165530 }, { "grad_norm": 0.11020450294017792, "learning_rate": 4.5575048592193606e-05, "loss": 0.0146, "step": 165540 }, { "grad_norm": 0.06719060987234116, "learning_rate": 4.55695586789643e-05, "loss": 0.0174, "step": 165550 }, { "grad_norm": 0.6208780407905579, "learning_rate": 4.5564068819569034e-05, "loss": 0.0118, "step": 165560 }, { "grad_norm": 0.1080193817615509, "learning_rate": 4.555857901407453e-05, "loss": 0.0173, "step": 165570 }, { "grad_norm": 0.05808354914188385, "learning_rate": 4.555308926254747e-05, "loss": 0.0149, "step": 165580 }, { "grad_norm": 0.07890791445970535, "learning_rate": 4.554759956505458e-05, "loss": 0.0158, "step": 165590 }, { "grad_norm": 0.08310743421316147, "learning_rate": 4.554210992166256e-05, "loss": 0.0142, "step": 165600 }, { "grad_norm": 0.07862541824579239, "learning_rate": 4.553662033243812e-05, "loss": 0.021, "step": 165610 }, { "grad_norm": 0.13946294784545898, "learning_rate": 4.5531130797447954e-05, "loss": 0.0181, "step": 165620 }, { "grad_norm": 0.2767152190208435, "learning_rate": 4.5525641316758754e-05, "loss": 0.0208, "step": 165630 }, { "grad_norm": 0.07348153740167618, "learning_rate": 4.5520151890437244e-05, "loss": 0.0181, "step": 165640 }, { "grad_norm": 0.13795731961727142, "learning_rate": 4.551466251855013e-05, "loss": 0.0197, "step": 165650 }, { "grad_norm": 0.07330244779586792, "learning_rate": 4.55091732011641e-05, "loss": 0.0166, "step": 165660 }, { "grad_norm": 0.0717916414141655, "learning_rate": 4.550368393834584e-05, "loss": 0.0241, "step": 165670 }, { "grad_norm": 0.07290470600128174, "learning_rate": 4.5498194730162066e-05, "loss": 0.0115, "step": 165680 }, { "grad_norm": 0.17606206238269806, "learning_rate": 4.5492705576679484e-05, "loss": 0.0208, "step": 165690 }, { "grad_norm": 0.09081262350082397, "learning_rate": 4.5487216477964786e-05, "loss": 0.014, "step": 165700 }, { "grad_norm": 0.19773292541503906, "learning_rate": 4.548172743408466e-05, "loss": 0.0151, "step": 165710 }, { "grad_norm": 0.09651742875576019, "learning_rate": 4.547623844510582e-05, "loss": 0.019, "step": 165720 }, { "grad_norm": 0.10407716780900955, "learning_rate": 4.547074951109497e-05, "loss": 0.014, "step": 165730 }, { "grad_norm": 0.20942994952201843, "learning_rate": 4.5465260632118764e-05, "loss": 0.0194, "step": 165740 }, { "grad_norm": 0.2014431655406952, "learning_rate": 4.545977180824393e-05, "loss": 0.0122, "step": 165750 }, { "grad_norm": 0.1471719592809677, "learning_rate": 4.545428303953715e-05, "loss": 0.0136, "step": 165760 }, { "grad_norm": 0.13665221631526947, "learning_rate": 4.5448794326065145e-05, "loss": 0.0143, "step": 165770 }, { "grad_norm": 0.09401119500398636, "learning_rate": 4.5443305667894565e-05, "loss": 0.0158, "step": 165780 }, { "grad_norm": 0.10127527266740799, "learning_rate": 4.543781706509212e-05, "loss": 0.0107, "step": 165790 }, { "grad_norm": 0.14040417969226837, "learning_rate": 4.5432328517724534e-05, "loss": 0.0161, "step": 165800 }, { "grad_norm": 0.07662434130907059, "learning_rate": 4.542684002585844e-05, "loss": 0.0106, "step": 165810 }, { "grad_norm": 0.21797063946723938, "learning_rate": 4.5421351589560564e-05, "loss": 0.0156, "step": 165820 }, { "grad_norm": 0.0858120322227478, "learning_rate": 4.541586320889759e-05, "loss": 0.0202, "step": 165830 }, { "grad_norm": 0.08490242809057236, "learning_rate": 4.541037488393622e-05, "loss": 0.0151, "step": 165840 }, { "grad_norm": 0.062419455498456955, "learning_rate": 4.540488661474311e-05, "loss": 0.0129, "step": 165850 }, { "grad_norm": 0.06945613771677017, "learning_rate": 4.5399398401384965e-05, "loss": 0.0143, "step": 165860 }, { "grad_norm": 0.10276332497596741, "learning_rate": 4.5393910243928494e-05, "loss": 0.0151, "step": 165870 }, { "grad_norm": 0.17588163912296295, "learning_rate": 4.538842214244035e-05, "loss": 0.0181, "step": 165880 }, { "grad_norm": 0.08265800774097443, "learning_rate": 4.538293409698722e-05, "loss": 0.0126, "step": 165890 }, { "grad_norm": 0.08234496414661407, "learning_rate": 4.53774461076358e-05, "loss": 0.0119, "step": 165900 }, { "grad_norm": 0.8551492691040039, "learning_rate": 4.53719581744528e-05, "loss": 0.0156, "step": 165910 }, { "grad_norm": 0.089483842253685, "learning_rate": 4.536647029750486e-05, "loss": 0.0115, "step": 165920 }, { "grad_norm": 0.08976241201162338, "learning_rate": 4.536098247685868e-05, "loss": 0.0165, "step": 165930 }, { "grad_norm": 0.14184170961380005, "learning_rate": 4.535549471258096e-05, "loss": 0.0114, "step": 165940 }, { "grad_norm": 0.14160622656345367, "learning_rate": 4.535000700473834e-05, "loss": 0.0171, "step": 165950 }, { "grad_norm": 0.30638185143470764, "learning_rate": 4.534451935339753e-05, "loss": 0.011, "step": 165960 }, { "grad_norm": 0.13650532066822052, "learning_rate": 4.5339031758625204e-05, "loss": 0.0153, "step": 165970 }, { "grad_norm": 0.05744950473308563, "learning_rate": 4.5333544220488066e-05, "loss": 0.0149, "step": 165980 }, { "grad_norm": 0.07565081119537354, "learning_rate": 4.532805673905275e-05, "loss": 0.0139, "step": 165990 }, { "grad_norm": 0.18620942533016205, "learning_rate": 4.5322569314385955e-05, "loss": 0.0112, "step": 166000 }, { "grad_norm": 0.10720306634902954, "learning_rate": 4.531708194655436e-05, "loss": 0.0156, "step": 166010 }, { "grad_norm": 0.10885979235172272, "learning_rate": 4.531159463562466e-05, "loss": 0.013, "step": 166020 }, { "grad_norm": 0.08564496040344238, "learning_rate": 4.5306107381663494e-05, "loss": 0.0102, "step": 166030 }, { "grad_norm": 0.31798040866851807, "learning_rate": 4.530062018473755e-05, "loss": 0.0135, "step": 166040 }, { "grad_norm": 0.13681411743164062, "learning_rate": 4.529513304491353e-05, "loss": 0.0152, "step": 166050 }, { "grad_norm": 0.16015625, "learning_rate": 4.528964596225806e-05, "loss": 0.014, "step": 166060 }, { "grad_norm": 0.05149540305137634, "learning_rate": 4.528415893683785e-05, "loss": 0.0102, "step": 166070 }, { "grad_norm": 0.3240056335926056, "learning_rate": 4.5278671968719555e-05, "loss": 0.0205, "step": 166080 }, { "grad_norm": 0.12151096016168594, "learning_rate": 4.527318505796988e-05, "loss": 0.0151, "step": 166090 }, { "grad_norm": 0.2506868541240692, "learning_rate": 4.526769820465544e-05, "loss": 0.0201, "step": 166100 }, { "grad_norm": 0.20444121956825256, "learning_rate": 4.526221140884294e-05, "loss": 0.0219, "step": 166110 }, { "grad_norm": 0.14897049963474274, "learning_rate": 4.525672467059906e-05, "loss": 0.015, "step": 166120 }, { "grad_norm": 0.11730443686246872, "learning_rate": 4.5251237989990444e-05, "loss": 0.0113, "step": 166130 }, { "grad_norm": 0.06917387247085571, "learning_rate": 4.524575136708376e-05, "loss": 0.0135, "step": 166140 }, { "grad_norm": 0.07171345502138138, "learning_rate": 4.524026480194569e-05, "loss": 0.0172, "step": 166150 }, { "grad_norm": 0.17383183538913727, "learning_rate": 4.5234778294642914e-05, "loss": 0.0146, "step": 166160 }, { "grad_norm": 0.072920061647892, "learning_rate": 4.522929184524207e-05, "loss": 0.0157, "step": 166170 }, { "grad_norm": 0.0888282060623169, "learning_rate": 4.522380545380982e-05, "loss": 0.0181, "step": 166180 }, { "grad_norm": 0.0926932692527771, "learning_rate": 4.521831912041286e-05, "loss": 0.0138, "step": 166190 }, { "grad_norm": 1.9187225103378296, "learning_rate": 4.5212832845117836e-05, "loss": 0.0148, "step": 166200 }, { "grad_norm": 0.15050280094146729, "learning_rate": 4.52073466279914e-05, "loss": 0.0154, "step": 166210 }, { "grad_norm": 0.11059987545013428, "learning_rate": 4.520186046910024e-05, "loss": 0.0175, "step": 166220 }, { "grad_norm": 0.10839008539915085, "learning_rate": 4.5196374368511e-05, "loss": 0.012, "step": 166230 }, { "grad_norm": 0.23137173056602478, "learning_rate": 4.519088832629036e-05, "loss": 0.0137, "step": 166240 }, { "grad_norm": 0.06202003359794617, "learning_rate": 4.518540234250495e-05, "loss": 0.0117, "step": 166250 }, { "grad_norm": 0.1887863129377365, "learning_rate": 4.5179916417221455e-05, "loss": 0.0103, "step": 166260 }, { "grad_norm": 0.16481894254684448, "learning_rate": 4.517443055050652e-05, "loss": 0.012, "step": 166270 }, { "grad_norm": 0.14971935749053955, "learning_rate": 4.516894474242682e-05, "loss": 0.0116, "step": 166280 }, { "grad_norm": 0.08871060609817505, "learning_rate": 4.5163458993048994e-05, "loss": 0.0118, "step": 166290 }, { "grad_norm": 0.2681106626987457, "learning_rate": 4.515797330243972e-05, "loss": 0.0137, "step": 166300 }, { "grad_norm": 0.050747405737638474, "learning_rate": 4.515248767066562e-05, "loss": 0.0159, "step": 166310 }, { "grad_norm": 0.11487697809934616, "learning_rate": 4.5147002097793395e-05, "loss": 0.0118, "step": 166320 }, { "grad_norm": 0.08705766499042511, "learning_rate": 4.514151658388966e-05, "loss": 0.016, "step": 166330 }, { "grad_norm": 0.041774772107601166, "learning_rate": 4.5136031129021096e-05, "loss": 0.0166, "step": 166340 }, { "grad_norm": 0.1021328717470169, "learning_rate": 4.513054573325434e-05, "loss": 0.0111, "step": 166350 }, { "grad_norm": 0.18423548340797424, "learning_rate": 4.512506039665605e-05, "loss": 0.0129, "step": 166360 }, { "grad_norm": 0.10842817276716232, "learning_rate": 4.5119575119292894e-05, "loss": 0.0152, "step": 166370 }, { "grad_norm": 0.2038458287715912, "learning_rate": 4.511408990123149e-05, "loss": 0.0146, "step": 166380 }, { "grad_norm": 0.07549681514501572, "learning_rate": 4.5108604742538505e-05, "loss": 0.0113, "step": 166390 }, { "grad_norm": 0.06195775792002678, "learning_rate": 4.5103119643280605e-05, "loss": 0.0137, "step": 166400 }, { "grad_norm": 0.150613933801651, "learning_rate": 4.5097634603524425e-05, "loss": 0.0163, "step": 166410 }, { "grad_norm": 0.14974947273731232, "learning_rate": 4.5092149623336596e-05, "loss": 0.0135, "step": 166420 }, { "grad_norm": 0.06520790606737137, "learning_rate": 4.508666470278379e-05, "loss": 0.016, "step": 166430 }, { "grad_norm": 0.17360720038414001, "learning_rate": 4.508117984193267e-05, "loss": 0.02, "step": 166440 }, { "grad_norm": 0.189132422208786, "learning_rate": 4.507569504084983e-05, "loss": 0.0181, "step": 166450 }, { "grad_norm": 0.09730388224124908, "learning_rate": 4.507021029960195e-05, "loss": 0.0135, "step": 166460 }, { "grad_norm": 0.10386423766613007, "learning_rate": 4.506472561825567e-05, "loss": 0.0145, "step": 166470 }, { "grad_norm": 0.10239727050065994, "learning_rate": 4.505924099687765e-05, "loss": 0.0144, "step": 166480 }, { "grad_norm": 0.14927427470684052, "learning_rate": 4.50537564355345e-05, "loss": 0.0149, "step": 166490 }, { "grad_norm": 0.09369169175624847, "learning_rate": 4.504827193429287e-05, "loss": 0.0095, "step": 166500 }, { "grad_norm": 0.06118876114487648, "learning_rate": 4.5042787493219445e-05, "loss": 0.0153, "step": 166510 }, { "grad_norm": 0.08570536226034164, "learning_rate": 4.50373031123808e-05, "loss": 0.0165, "step": 166520 }, { "grad_norm": 0.12026933580636978, "learning_rate": 4.5031818791843615e-05, "loss": 0.0113, "step": 166530 }, { "grad_norm": 0.15227045118808746, "learning_rate": 4.502633453167452e-05, "loss": 0.0141, "step": 166540 }, { "grad_norm": 0.1851135492324829, "learning_rate": 4.502085033194018e-05, "loss": 0.0134, "step": 166550 }, { "grad_norm": 0.08882587403059006, "learning_rate": 4.501536619270719e-05, "loss": 0.0131, "step": 166560 }, { "grad_norm": 0.4708741009235382, "learning_rate": 4.50098821140422e-05, "loss": 0.0155, "step": 166570 }, { "grad_norm": 0.2744427025318146, "learning_rate": 4.500439809601188e-05, "loss": 0.0137, "step": 166580 }, { "grad_norm": 0.069420225918293, "learning_rate": 4.499891413868282e-05, "loss": 0.0127, "step": 166590 }, { "grad_norm": 0.20212654769420624, "learning_rate": 4.499343024212167e-05, "loss": 0.0148, "step": 166600 }, { "grad_norm": 0.17218434810638428, "learning_rate": 4.4987946406395074e-05, "loss": 0.0116, "step": 166610 }, { "grad_norm": 0.13417193293571472, "learning_rate": 4.4982462631569674e-05, "loss": 0.0173, "step": 166620 }, { "grad_norm": 0.10329639911651611, "learning_rate": 4.497697891771208e-05, "loss": 0.0177, "step": 166630 }, { "grad_norm": 0.0781199261546135, "learning_rate": 4.497149526488893e-05, "loss": 0.0103, "step": 166640 }, { "grad_norm": 0.09519768506288528, "learning_rate": 4.4966011673166864e-05, "loss": 0.0097, "step": 166650 }, { "grad_norm": 0.05727454647421837, "learning_rate": 4.4960528142612526e-05, "loss": 0.0128, "step": 166660 }, { "grad_norm": 0.17915919423103333, "learning_rate": 4.495504467329251e-05, "loss": 0.018, "step": 166670 }, { "grad_norm": 0.09149942547082901, "learning_rate": 4.494956126527346e-05, "loss": 0.0137, "step": 166680 }, { "grad_norm": 0.11940629035234451, "learning_rate": 4.494407791862204e-05, "loss": 0.0158, "step": 166690 }, { "grad_norm": 0.09042169153690338, "learning_rate": 4.4938594633404815e-05, "loss": 0.0157, "step": 166700 }, { "grad_norm": 0.18388791382312775, "learning_rate": 4.4933111409688446e-05, "loss": 0.0166, "step": 166710 }, { "grad_norm": 0.07642225176095963, "learning_rate": 4.4927628247539554e-05, "loss": 0.0122, "step": 166720 }, { "grad_norm": 0.4904954135417938, "learning_rate": 4.49221451470248e-05, "loss": 0.0166, "step": 166730 }, { "grad_norm": 0.08216685056686401, "learning_rate": 4.491666210821075e-05, "loss": 0.0127, "step": 166740 }, { "grad_norm": 0.07482829689979553, "learning_rate": 4.491117913116405e-05, "loss": 0.0143, "step": 166750 }, { "grad_norm": 0.23680520057678223, "learning_rate": 4.490569621595136e-05, "loss": 0.018, "step": 166760 }, { "grad_norm": 0.08607172966003418, "learning_rate": 4.490021336263924e-05, "loss": 0.0146, "step": 166770 }, { "grad_norm": 0.05558279901742935, "learning_rate": 4.4894730571294345e-05, "loss": 0.0131, "step": 166780 }, { "grad_norm": 0.08880944550037384, "learning_rate": 4.4889247841983295e-05, "loss": 0.0117, "step": 166790 }, { "grad_norm": 0.22475561499595642, "learning_rate": 4.488376517477273e-05, "loss": 0.0116, "step": 166800 }, { "grad_norm": 0.22038152813911438, "learning_rate": 4.487828256972923e-05, "loss": 0.0169, "step": 166810 }, { "grad_norm": 0.09777584671974182, "learning_rate": 4.4872800026919434e-05, "loss": 0.0164, "step": 166820 }, { "grad_norm": 0.11996261775493622, "learning_rate": 4.486731754640997e-05, "loss": 0.0118, "step": 166830 }, { "grad_norm": 0.08986026048660278, "learning_rate": 4.486183512826743e-05, "loss": 0.0144, "step": 166840 }, { "grad_norm": 0.06019558385014534, "learning_rate": 4.4856352772558444e-05, "loss": 0.0136, "step": 166850 }, { "grad_norm": 0.0542772114276886, "learning_rate": 4.4850870479349625e-05, "loss": 0.0141, "step": 166860 }, { "grad_norm": 0.10780642181634903, "learning_rate": 4.484538824870762e-05, "loss": 0.0108, "step": 166870 }, { "grad_norm": 0.19894084334373474, "learning_rate": 4.4839906080698984e-05, "loss": 0.0168, "step": 166880 }, { "grad_norm": 0.19093741476535797, "learning_rate": 4.4834423975390364e-05, "loss": 0.0186, "step": 166890 }, { "grad_norm": 0.08075211197137833, "learning_rate": 4.4828941932848375e-05, "loss": 0.0166, "step": 166900 }, { "grad_norm": 0.11123894900083542, "learning_rate": 4.482345995313964e-05, "loss": 0.0178, "step": 166910 }, { "grad_norm": 0.07450329512357712, "learning_rate": 4.481797803633073e-05, "loss": 0.0146, "step": 166920 }, { "grad_norm": 0.1057223305106163, "learning_rate": 4.481249618248828e-05, "loss": 0.0148, "step": 166930 }, { "grad_norm": 0.07908646017313004, "learning_rate": 4.480701439167892e-05, "loss": 0.0165, "step": 166940 }, { "grad_norm": 0.11635939031839371, "learning_rate": 4.4801532663969234e-05, "loss": 0.0193, "step": 166950 }, { "grad_norm": 0.07955275475978851, "learning_rate": 4.479605099942582e-05, "loss": 0.0101, "step": 166960 }, { "grad_norm": 0.09679765999317169, "learning_rate": 4.4790569398115304e-05, "loss": 0.0163, "step": 166970 }, { "grad_norm": 0.11317604035139084, "learning_rate": 4.478508786010429e-05, "loss": 0.0168, "step": 166980 }, { "grad_norm": 0.6006643772125244, "learning_rate": 4.477960638545939e-05, "loss": 0.0188, "step": 166990 }, { "grad_norm": 0.1784907728433609, "learning_rate": 4.477412497424719e-05, "loss": 0.0172, "step": 167000 }, { "grad_norm": 0.08895444869995117, "learning_rate": 4.4768643626534314e-05, "loss": 0.0186, "step": 167010 }, { "grad_norm": 0.06368862092494965, "learning_rate": 4.476316234238734e-05, "loss": 0.017, "step": 167020 }, { "grad_norm": 0.375124990940094, "learning_rate": 4.475768112187291e-05, "loss": 0.0155, "step": 167030 }, { "grad_norm": 0.1331186145544052, "learning_rate": 4.475219996505759e-05, "loss": 0.0125, "step": 167040 }, { "grad_norm": 0.07230454683303833, "learning_rate": 4.4746718872008004e-05, "loss": 0.0149, "step": 167050 }, { "grad_norm": 0.1735202968120575, "learning_rate": 4.4741237842790734e-05, "loss": 0.0125, "step": 167060 }, { "grad_norm": 0.09069781005382538, "learning_rate": 4.4735756877472403e-05, "loss": 0.0143, "step": 167070 }, { "grad_norm": 0.11523409187793732, "learning_rate": 4.473027597611959e-05, "loss": 0.0109, "step": 167080 }, { "grad_norm": 0.06931524723768234, "learning_rate": 4.4724795138798893e-05, "loss": 0.0156, "step": 167090 }, { "grad_norm": 0.09940711408853531, "learning_rate": 4.4719314365576915e-05, "loss": 0.0152, "step": 167100 }, { "grad_norm": 0.1424129754304886, "learning_rate": 4.471383365652026e-05, "loss": 0.0112, "step": 167110 }, { "grad_norm": 0.11012671887874603, "learning_rate": 4.4708353011695524e-05, "loss": 0.0136, "step": 167120 }, { "grad_norm": 0.09987810254096985, "learning_rate": 4.470287243116928e-05, "loss": 0.0095, "step": 167130 }, { "grad_norm": 0.10253249108791351, "learning_rate": 4.469739191500813e-05, "loss": 0.0125, "step": 167140 }, { "grad_norm": 0.1364651769399643, "learning_rate": 4.4691911463278716e-05, "loss": 0.0156, "step": 167150 }, { "grad_norm": 0.078359454870224, "learning_rate": 4.468643107604755e-05, "loss": 0.0117, "step": 167160 }, { "grad_norm": 0.14593657851219177, "learning_rate": 4.468095075338127e-05, "loss": 0.0137, "step": 167170 }, { "grad_norm": 0.10202480852603912, "learning_rate": 4.467547049534646e-05, "loss": 0.0118, "step": 167180 }, { "grad_norm": 0.09726685285568237, "learning_rate": 4.4669990302009736e-05, "loss": 0.0082, "step": 167190 }, { "grad_norm": 0.1471458524465561, "learning_rate": 4.466451017343763e-05, "loss": 0.0156, "step": 167200 }, { "grad_norm": 0.0939645916223526, "learning_rate": 4.465903010969677e-05, "loss": 0.0107, "step": 167210 }, { "grad_norm": 0.12917938828468323, "learning_rate": 4.4653550110853735e-05, "loss": 0.0141, "step": 167220 }, { "grad_norm": 0.14079976081848145, "learning_rate": 4.464807017697514e-05, "loss": 0.0168, "step": 167230 }, { "grad_norm": 0.17896468937397003, "learning_rate": 4.464259030812752e-05, "loss": 0.0114, "step": 167240 }, { "grad_norm": 0.16173361241817474, "learning_rate": 4.4637110504377485e-05, "loss": 0.0146, "step": 167250 }, { "grad_norm": 0.10717982053756714, "learning_rate": 4.463163076579165e-05, "loss": 0.0118, "step": 167260 }, { "grad_norm": 0.17591288685798645, "learning_rate": 4.462615109243654e-05, "loss": 0.0136, "step": 167270 }, { "grad_norm": 0.1604386419057846, "learning_rate": 4.462067148437877e-05, "loss": 0.012, "step": 167280 }, { "grad_norm": 0.08534497767686844, "learning_rate": 4.4615191941684916e-05, "loss": 0.0136, "step": 167290 }, { "grad_norm": 0.08484260737895966, "learning_rate": 4.460971246442159e-05, "loss": 0.0176, "step": 167300 }, { "grad_norm": 0.09680985659360886, "learning_rate": 4.460423305265533e-05, "loss": 0.0082, "step": 167310 }, { "grad_norm": 0.06824658811092377, "learning_rate": 4.459875370645273e-05, "loss": 0.0137, "step": 167320 }, { "grad_norm": 0.09123905748128891, "learning_rate": 4.4593274425880395e-05, "loss": 0.0148, "step": 167330 }, { "grad_norm": 0.2496875822544098, "learning_rate": 4.458779521100485e-05, "loss": 0.0203, "step": 167340 }, { "grad_norm": 0.10284475237131119, "learning_rate": 4.458231606189271e-05, "loss": 0.012, "step": 167350 }, { "grad_norm": 0.10347341001033783, "learning_rate": 4.4576836978610546e-05, "loss": 0.0103, "step": 167360 }, { "grad_norm": 0.06042517349123955, "learning_rate": 4.457135796122495e-05, "loss": 0.0111, "step": 167370 }, { "grad_norm": 0.07769464701414108, "learning_rate": 4.4565879009802464e-05, "loss": 0.0125, "step": 167380 }, { "grad_norm": 0.09861599653959274, "learning_rate": 4.456040012440967e-05, "loss": 0.0128, "step": 167390 }, { "grad_norm": 0.1357201635837555, "learning_rate": 4.455492130511319e-05, "loss": 0.0121, "step": 167400 }, { "grad_norm": 0.1182207316160202, "learning_rate": 4.454944255197952e-05, "loss": 0.0167, "step": 167410 }, { "grad_norm": 0.10731968283653259, "learning_rate": 4.4543963865075275e-05, "loss": 0.0132, "step": 167420 }, { "grad_norm": 0.06962931156158447, "learning_rate": 4.453848524446702e-05, "loss": 0.0134, "step": 167430 }, { "grad_norm": 0.05445640906691551, "learning_rate": 4.453300669022134e-05, "loss": 0.015, "step": 167440 }, { "grad_norm": 0.07344464957714081, "learning_rate": 4.452752820240477e-05, "loss": 0.0171, "step": 167450 }, { "grad_norm": 0.19439519941806793, "learning_rate": 4.452204978108391e-05, "loss": 0.0256, "step": 167460 }, { "grad_norm": 0.10547887533903122, "learning_rate": 4.451657142632533e-05, "loss": 0.0161, "step": 167470 }, { "grad_norm": 0.14203625917434692, "learning_rate": 4.451109313819556e-05, "loss": 0.0151, "step": 167480 }, { "grad_norm": 0.1886889934539795, "learning_rate": 4.45056149167612e-05, "loss": 0.0204, "step": 167490 }, { "grad_norm": 0.06854136288166046, "learning_rate": 4.45001367620888e-05, "loss": 0.013, "step": 167500 }, { "grad_norm": 0.07017741352319717, "learning_rate": 4.449465867424496e-05, "loss": 0.019, "step": 167510 }, { "grad_norm": 0.07465435564517975, "learning_rate": 4.448918065329619e-05, "loss": 0.0168, "step": 167520 }, { "grad_norm": 0.12248612195253372, "learning_rate": 4.4483702699309074e-05, "loss": 0.0171, "step": 167530 }, { "grad_norm": 0.0966111272573471, "learning_rate": 4.4478224812350176e-05, "loss": 0.0127, "step": 167540 }, { "grad_norm": 0.06597088277339935, "learning_rate": 4.447274699248609e-05, "loss": 0.0127, "step": 167550 }, { "grad_norm": 0.06232792139053345, "learning_rate": 4.446726923978332e-05, "loss": 0.0105, "step": 167560 }, { "grad_norm": 0.10200733691453934, "learning_rate": 4.4461791554308454e-05, "loss": 0.026, "step": 167570 }, { "grad_norm": 0.10061812400817871, "learning_rate": 4.445631393612807e-05, "loss": 0.0155, "step": 167580 }, { "grad_norm": 0.07342424243688583, "learning_rate": 4.445083638530869e-05, "loss": 0.011, "step": 167590 }, { "grad_norm": 0.0734291598200798, "learning_rate": 4.4445358901916886e-05, "loss": 0.0123, "step": 167600 }, { "grad_norm": 0.10473036020994186, "learning_rate": 4.443988148601921e-05, "loss": 0.0118, "step": 167610 }, { "grad_norm": 0.10772328823804855, "learning_rate": 4.4434404137682245e-05, "loss": 0.0107, "step": 167620 }, { "grad_norm": 0.11176414042711258, "learning_rate": 4.442892685697251e-05, "loss": 0.0087, "step": 167630 }, { "grad_norm": 0.05573917180299759, "learning_rate": 4.442344964395657e-05, "loss": 0.01, "step": 167640 }, { "grad_norm": 0.05678108334541321, "learning_rate": 4.4417972498700986e-05, "loss": 0.0134, "step": 167650 }, { "grad_norm": 0.08096213638782501, "learning_rate": 4.441249542127231e-05, "loss": 0.0143, "step": 167660 }, { "grad_norm": 0.06785622984170914, "learning_rate": 4.4407018411737084e-05, "loss": 0.017, "step": 167670 }, { "grad_norm": 0.08166012167930603, "learning_rate": 4.4401541470161864e-05, "loss": 0.015, "step": 167680 }, { "grad_norm": 0.11564169824123383, "learning_rate": 4.439606459661321e-05, "loss": 0.0183, "step": 167690 }, { "grad_norm": 0.2413763552904129, "learning_rate": 4.439058779115766e-05, "loss": 0.0183, "step": 167700 }, { "grad_norm": 0.12249595671892166, "learning_rate": 4.438511105386176e-05, "loss": 0.0114, "step": 167710 }, { "grad_norm": 0.21319018304347992, "learning_rate": 4.437963438479207e-05, "loss": 0.0199, "step": 167720 }, { "grad_norm": 0.10577221214771271, "learning_rate": 4.4374157784015114e-05, "loss": 0.0143, "step": 167730 }, { "grad_norm": 0.1828373670578003, "learning_rate": 4.4368681251597465e-05, "loss": 0.0171, "step": 167740 }, { "grad_norm": 0.08078467845916748, "learning_rate": 4.436320478760565e-05, "loss": 0.017, "step": 167750 }, { "grad_norm": 0.17733997106552124, "learning_rate": 4.435772839210622e-05, "loss": 0.0122, "step": 167760 }, { "grad_norm": 0.1430450677871704, "learning_rate": 4.4352252065165715e-05, "loss": 0.016, "step": 167770 }, { "grad_norm": 0.2024618536233902, "learning_rate": 4.4346775806850685e-05, "loss": 0.0126, "step": 167780 }, { "grad_norm": 0.15703818202018738, "learning_rate": 4.434129961722767e-05, "loss": 0.0167, "step": 167790 }, { "grad_norm": 0.17979997396469116, "learning_rate": 4.43358234963632e-05, "loss": 0.0136, "step": 167800 }, { "grad_norm": 0.11764761060476303, "learning_rate": 4.433034744432382e-05, "loss": 0.0121, "step": 167810 }, { "grad_norm": 0.11252244561910629, "learning_rate": 4.4324871461176074e-05, "loss": 0.017, "step": 167820 }, { "grad_norm": 0.09983213245868683, "learning_rate": 4.431939554698651e-05, "loss": 0.0178, "step": 167830 }, { "grad_norm": 0.3969132602214813, "learning_rate": 4.431391970182164e-05, "loss": 0.0154, "step": 167840 }, { "grad_norm": 0.2180873453617096, "learning_rate": 4.4308443925748005e-05, "loss": 0.0214, "step": 167850 }, { "grad_norm": 0.14458976686000824, "learning_rate": 4.4302968218832166e-05, "loss": 0.0167, "step": 167860 }, { "grad_norm": 0.22645303606987, "learning_rate": 4.429749258114065e-05, "loss": 0.0129, "step": 167870 }, { "grad_norm": 0.11099954694509506, "learning_rate": 4.429201701273996e-05, "loss": 0.0155, "step": 167880 }, { "grad_norm": 0.16094262897968292, "learning_rate": 4.428654151369666e-05, "loss": 0.0142, "step": 167890 }, { "grad_norm": 0.14328517019748688, "learning_rate": 4.4281066084077296e-05, "loss": 0.0208, "step": 167900 }, { "grad_norm": 0.13195036351680756, "learning_rate": 4.427559072394836e-05, "loss": 0.0139, "step": 167910 }, { "grad_norm": 0.1591905802488327, "learning_rate": 4.4270115433376394e-05, "loss": 0.0169, "step": 167920 }, { "grad_norm": 0.1661054641008377, "learning_rate": 4.426464021242794e-05, "loss": 0.0164, "step": 167930 }, { "grad_norm": 0.08789175748825073, "learning_rate": 4.425916506116954e-05, "loss": 0.0129, "step": 167940 }, { "grad_norm": 0.07763378322124481, "learning_rate": 4.4253689979667684e-05, "loss": 0.0197, "step": 167950 }, { "grad_norm": 0.2054547667503357, "learning_rate": 4.424821496798892e-05, "loss": 0.0166, "step": 167960 }, { "grad_norm": 0.08247584849596024, "learning_rate": 4.4242740026199805e-05, "loss": 0.0153, "step": 167970 }, { "grad_norm": 0.1024889275431633, "learning_rate": 4.42372651543668e-05, "loss": 0.0131, "step": 167980 }, { "grad_norm": 0.06857410073280334, "learning_rate": 4.4231790352556476e-05, "loss": 0.0117, "step": 167990 }, { "grad_norm": 0.0650959238409996, "learning_rate": 4.4226315620835334e-05, "loss": 0.0145, "step": 168000 }, { "grad_norm": 0.18946093320846558, "learning_rate": 4.422084095926994e-05, "loss": 0.0165, "step": 168010 }, { "grad_norm": 0.12097349017858505, "learning_rate": 4.4215366367926756e-05, "loss": 0.0142, "step": 168020 }, { "grad_norm": 0.08071383088827133, "learning_rate": 4.420989184687234e-05, "loss": 0.0123, "step": 168030 }, { "grad_norm": 0.07695125043392181, "learning_rate": 4.420441739617322e-05, "loss": 0.0261, "step": 168040 }, { "grad_norm": 0.22212080657482147, "learning_rate": 4.419894301589588e-05, "loss": 0.0136, "step": 168050 }, { "grad_norm": 0.23317034542560577, "learning_rate": 4.419346870610686e-05, "loss": 0.0092, "step": 168060 }, { "grad_norm": 0.07145293056964874, "learning_rate": 4.4187994466872676e-05, "loss": 0.0126, "step": 168070 }, { "grad_norm": 0.10817300528287888, "learning_rate": 4.418252029825987e-05, "loss": 0.0171, "step": 168080 }, { "grad_norm": 0.15658116340637207, "learning_rate": 4.417704620033491e-05, "loss": 0.0174, "step": 168090 }, { "grad_norm": 0.08079994469881058, "learning_rate": 4.4171572173164345e-05, "loss": 0.012, "step": 168100 }, { "grad_norm": 0.16539888083934784, "learning_rate": 4.4166098216814676e-05, "loss": 0.0136, "step": 168110 }, { "grad_norm": 0.07615014165639877, "learning_rate": 4.416062433135244e-05, "loss": 0.0114, "step": 168120 }, { "grad_norm": 0.09773293137550354, "learning_rate": 4.415515051684412e-05, "loss": 0.0151, "step": 168130 }, { "grad_norm": 0.1579495072364807, "learning_rate": 4.414967677335623e-05, "loss": 0.0138, "step": 168140 }, { "grad_norm": 0.16185331344604492, "learning_rate": 4.4144203100955314e-05, "loss": 0.0162, "step": 168150 }, { "grad_norm": 0.11924362182617188, "learning_rate": 4.4138729499707844e-05, "loss": 0.0119, "step": 168160 }, { "grad_norm": 0.099802166223526, "learning_rate": 4.413325596968034e-05, "loss": 0.0136, "step": 168170 }, { "grad_norm": 0.17362232506275177, "learning_rate": 4.4127782510939305e-05, "loss": 0.0141, "step": 168180 }, { "grad_norm": 0.2198292762041092, "learning_rate": 4.412230912355129e-05, "loss": 0.0149, "step": 168190 }, { "grad_norm": 0.08488120883703232, "learning_rate": 4.411683580758275e-05, "loss": 0.0114, "step": 168200 }, { "grad_norm": 0.1480478048324585, "learning_rate": 4.41113625631002e-05, "loss": 0.016, "step": 168210 }, { "grad_norm": 0.10674389451742172, "learning_rate": 4.4105889390170186e-05, "loss": 0.0137, "step": 168220 }, { "grad_norm": 0.18006494641304016, "learning_rate": 4.410041628885915e-05, "loss": 0.0137, "step": 168230 }, { "grad_norm": 0.10485739260911942, "learning_rate": 4.4094943259233635e-05, "loss": 0.0152, "step": 168240 }, { "grad_norm": 0.07612331211566925, "learning_rate": 4.408947030136012e-05, "loss": 0.0168, "step": 168250 }, { "grad_norm": 0.10322307795286179, "learning_rate": 4.408399741530516e-05, "loss": 0.0137, "step": 168260 }, { "grad_norm": 0.093967005610466, "learning_rate": 4.4078524601135184e-05, "loss": 0.0115, "step": 168270 }, { "grad_norm": 0.07755576074123383, "learning_rate": 4.4073051858916734e-05, "loss": 0.0116, "step": 168280 }, { "grad_norm": 0.10915940999984741, "learning_rate": 4.406757918871631e-05, "loss": 0.0208, "step": 168290 }, { "grad_norm": 0.1133117750287056, "learning_rate": 4.4062106590600386e-05, "loss": 0.0132, "step": 168300 }, { "grad_norm": 0.0642893835902214, "learning_rate": 4.405663406463548e-05, "loss": 0.0106, "step": 168310 }, { "grad_norm": 0.14611464738845825, "learning_rate": 4.405116161088807e-05, "loss": 0.0146, "step": 168320 }, { "grad_norm": 0.09539072960615158, "learning_rate": 4.404568922942469e-05, "loss": 0.0147, "step": 168330 }, { "grad_norm": 0.07068277895450592, "learning_rate": 4.404021692031178e-05, "loss": 0.0118, "step": 168340 }, { "grad_norm": 0.21546894311904907, "learning_rate": 4.403474468361587e-05, "loss": 0.0103, "step": 168350 }, { "grad_norm": 0.10082825273275375, "learning_rate": 4.4029272519403455e-05, "loss": 0.0084, "step": 168360 }, { "grad_norm": 0.10512357205152512, "learning_rate": 4.402380042774101e-05, "loss": 0.0145, "step": 168370 }, { "grad_norm": 0.05923695117235184, "learning_rate": 4.401832840869502e-05, "loss": 0.0151, "step": 168380 }, { "grad_norm": 0.07044187188148499, "learning_rate": 4.401285646233199e-05, "loss": 0.0185, "step": 168390 }, { "grad_norm": 0.07820320874452591, "learning_rate": 4.400738458871842e-05, "loss": 0.0114, "step": 168400 }, { "grad_norm": 0.08527620881795883, "learning_rate": 4.4001912787920784e-05, "loss": 0.0116, "step": 168410 }, { "grad_norm": 0.0736866295337677, "learning_rate": 4.399644106000556e-05, "loss": 0.0147, "step": 168420 }, { "grad_norm": 0.09307968616485596, "learning_rate": 4.3990969405039244e-05, "loss": 0.0135, "step": 168430 }, { "grad_norm": 0.09642309695482254, "learning_rate": 4.398549782308834e-05, "loss": 0.0132, "step": 168440 }, { "grad_norm": 0.06259016692638397, "learning_rate": 4.3980026314219306e-05, "loss": 0.0114, "step": 168450 }, { "grad_norm": 0.18152642250061035, "learning_rate": 4.3974554878498634e-05, "loss": 0.0177, "step": 168460 }, { "grad_norm": 0.12053132802248001, "learning_rate": 4.396908351599281e-05, "loss": 0.0191, "step": 168470 }, { "grad_norm": 0.09079915285110474, "learning_rate": 4.396361222676832e-05, "loss": 0.0183, "step": 168480 }, { "grad_norm": 0.1958445906639099, "learning_rate": 4.395814101089164e-05, "loss": 0.0163, "step": 168490 }, { "grad_norm": 0.07307229936122894, "learning_rate": 4.395266986842924e-05, "loss": 0.0131, "step": 168500 }, { "grad_norm": 0.061893582344055176, "learning_rate": 4.3947198799447625e-05, "loss": 0.019, "step": 168510 }, { "grad_norm": 0.0914997011423111, "learning_rate": 4.394172780401325e-05, "loss": 0.008, "step": 168520 }, { "grad_norm": 0.12820377945899963, "learning_rate": 4.393625688219261e-05, "loss": 0.0182, "step": 168530 }, { "grad_norm": 0.06709734350442886, "learning_rate": 4.393078603405218e-05, "loss": 0.0124, "step": 168540 }, { "grad_norm": 0.07370111346244812, "learning_rate": 4.392531525965841e-05, "loss": 0.0169, "step": 168550 }, { "grad_norm": 0.09367454797029495, "learning_rate": 4.39198445590778e-05, "loss": 0.0171, "step": 168560 }, { "grad_norm": 0.09454098343849182, "learning_rate": 4.3914373932376816e-05, "loss": 0.015, "step": 168570 }, { "grad_norm": 0.12368748337030411, "learning_rate": 4.390890337962196e-05, "loss": 0.0161, "step": 168580 }, { "grad_norm": 0.16106900572776794, "learning_rate": 4.3903432900879656e-05, "loss": 0.0149, "step": 168590 }, { "grad_norm": 0.0937390848994255, "learning_rate": 4.38979624962164e-05, "loss": 0.0129, "step": 168600 }, { "grad_norm": 0.13451921939849854, "learning_rate": 4.3892492165698684e-05, "loss": 0.0218, "step": 168610 }, { "grad_norm": 0.08526875823736191, "learning_rate": 4.388702190939293e-05, "loss": 0.0134, "step": 168620 }, { "grad_norm": 0.1044604629278183, "learning_rate": 4.3881551727365626e-05, "loss": 0.0136, "step": 168630 }, { "grad_norm": 0.06549612432718277, "learning_rate": 4.387608161968325e-05, "loss": 0.014, "step": 168640 }, { "grad_norm": 0.07257392257452011, "learning_rate": 4.387061158641229e-05, "loss": 0.0161, "step": 168650 }, { "grad_norm": 0.07884617894887924, "learning_rate": 4.386514162761916e-05, "loss": 0.0151, "step": 168660 }, { "grad_norm": 0.07107222825288773, "learning_rate": 4.3859671743370355e-05, "loss": 0.0195, "step": 168670 }, { "grad_norm": 0.09066975116729736, "learning_rate": 4.3854201933732353e-05, "loss": 0.0132, "step": 168680 }, { "grad_norm": 0.3440973460674286, "learning_rate": 4.384873219877158e-05, "loss": 0.0169, "step": 168690 }, { "grad_norm": 0.1264248788356781, "learning_rate": 4.384326253855452e-05, "loss": 0.0143, "step": 168700 }, { "grad_norm": 0.06004004180431366, "learning_rate": 4.3837792953147636e-05, "loss": 0.0152, "step": 168710 }, { "grad_norm": 0.14315451681613922, "learning_rate": 4.38323234426174e-05, "loss": 0.0169, "step": 168720 }, { "grad_norm": 0.10611332952976227, "learning_rate": 4.382685400703024e-05, "loss": 0.0143, "step": 168730 }, { "grad_norm": 0.11833896487951279, "learning_rate": 4.3821384646452635e-05, "loss": 0.0138, "step": 168740 }, { "grad_norm": 0.07405419647693634, "learning_rate": 4.381591536095103e-05, "loss": 0.0102, "step": 168750 }, { "grad_norm": 0.09354928880929947, "learning_rate": 4.381044615059192e-05, "loss": 0.0122, "step": 168760 }, { "grad_norm": 0.107783243060112, "learning_rate": 4.380497701544172e-05, "loss": 0.0152, "step": 168770 }, { "grad_norm": 0.08523130416870117, "learning_rate": 4.379950795556689e-05, "loss": 0.0151, "step": 168780 }, { "grad_norm": 0.09679669886827469, "learning_rate": 4.3794038971033914e-05, "loss": 0.0194, "step": 168790 }, { "grad_norm": 0.13405461609363556, "learning_rate": 4.37885700619092e-05, "loss": 0.0132, "step": 168800 }, { "grad_norm": 0.08457488566637039, "learning_rate": 4.3783101228259225e-05, "loss": 0.0106, "step": 168810 }, { "grad_norm": 0.07196106761693954, "learning_rate": 4.377763247015044e-05, "loss": 0.0154, "step": 168820 }, { "grad_norm": 0.13781476020812988, "learning_rate": 4.3772163787649324e-05, "loss": 0.014, "step": 168830 }, { "grad_norm": 0.13248804211616516, "learning_rate": 4.376669518082227e-05, "loss": 0.016, "step": 168840 }, { "grad_norm": 0.1285495162010193, "learning_rate": 4.376122664973575e-05, "loss": 0.0168, "step": 168850 }, { "grad_norm": 0.09079433232545853, "learning_rate": 4.375575819445625e-05, "loss": 0.0121, "step": 168860 }, { "grad_norm": 0.06947021186351776, "learning_rate": 4.375028981505015e-05, "loss": 0.0144, "step": 168870 }, { "grad_norm": 0.07444901764392853, "learning_rate": 4.3744821511583945e-05, "loss": 0.0185, "step": 168880 }, { "grad_norm": 0.04648392274975777, "learning_rate": 4.373935328412405e-05, "loss": 0.0123, "step": 168890 }, { "grad_norm": 0.05440989509224892, "learning_rate": 4.373388513273695e-05, "loss": 0.014, "step": 168900 }, { "grad_norm": 0.10220775753259659, "learning_rate": 4.372841705748905e-05, "loss": 0.0112, "step": 168910 }, { "grad_norm": 0.09350630640983582, "learning_rate": 4.3722949058446786e-05, "loss": 0.0168, "step": 168920 }, { "grad_norm": 0.2009490579366684, "learning_rate": 4.371748113567665e-05, "loss": 0.0151, "step": 168930 }, { "grad_norm": 0.10581041872501373, "learning_rate": 4.371201328924502e-05, "loss": 0.0164, "step": 168940 }, { "grad_norm": 0.10040197521448135, "learning_rate": 4.370654551921838e-05, "loss": 0.0142, "step": 168950 }, { "grad_norm": 0.10291216522455215, "learning_rate": 4.370107782566314e-05, "loss": 0.0135, "step": 168960 }, { "grad_norm": 0.18686427175998688, "learning_rate": 4.369561020864578e-05, "loss": 0.0138, "step": 168970 }, { "grad_norm": 0.08193603903055191, "learning_rate": 4.3690142668232686e-05, "loss": 0.0181, "step": 168980 }, { "grad_norm": 0.15741924941539764, "learning_rate": 4.368467520449032e-05, "loss": 0.0157, "step": 168990 }, { "grad_norm": 0.14808617532253265, "learning_rate": 4.3679207817485124e-05, "loss": 0.0158, "step": 169000 }, { "grad_norm": 0.4202447831630707, "learning_rate": 4.36737405072835e-05, "loss": 0.0138, "step": 169010 }, { "grad_norm": 0.1919669508934021, "learning_rate": 4.366827327395191e-05, "loss": 0.0178, "step": 169020 }, { "grad_norm": 0.2702859044075012, "learning_rate": 4.366280611755678e-05, "loss": 0.0119, "step": 169030 }, { "grad_norm": 0.09593556076288223, "learning_rate": 4.365733903816455e-05, "loss": 0.0163, "step": 169040 }, { "grad_norm": 0.09566152840852737, "learning_rate": 4.365187203584162e-05, "loss": 0.0153, "step": 169050 }, { "grad_norm": 0.07299435138702393, "learning_rate": 4.364640511065445e-05, "loss": 0.0137, "step": 169060 }, { "grad_norm": 0.0977703183889389, "learning_rate": 4.3640938262669445e-05, "loss": 0.0114, "step": 169070 }, { "grad_norm": 0.07090242952108383, "learning_rate": 4.363547149195307e-05, "loss": 0.0129, "step": 169080 }, { "grad_norm": 0.15392906963825226, "learning_rate": 4.36300047985717e-05, "loss": 0.0127, "step": 169090 }, { "grad_norm": 0.20920920372009277, "learning_rate": 4.362453818259179e-05, "loss": 0.0118, "step": 169100 }, { "grad_norm": 0.07215454429388046, "learning_rate": 4.3619071644079765e-05, "loss": 0.0113, "step": 169110 }, { "grad_norm": 0.08487886190414429, "learning_rate": 4.3613605183102054e-05, "loss": 0.0142, "step": 169120 }, { "grad_norm": 0.09584177285432816, "learning_rate": 4.360813879972505e-05, "loss": 0.0162, "step": 169130 }, { "grad_norm": 0.127626433968544, "learning_rate": 4.36026724940152e-05, "loss": 0.0154, "step": 169140 }, { "grad_norm": 0.14793814718723297, "learning_rate": 4.3597206266038933e-05, "loss": 0.0111, "step": 169150 }, { "grad_norm": 0.07716608792543411, "learning_rate": 4.3591740115862636e-05, "loss": 0.0142, "step": 169160 }, { "grad_norm": 0.13468576967716217, "learning_rate": 4.358627404355276e-05, "loss": 0.0147, "step": 169170 }, { "grad_norm": 0.09519333392381668, "learning_rate": 4.3580808049175714e-05, "loss": 0.0172, "step": 169180 }, { "grad_norm": 0.11155279725790024, "learning_rate": 4.35753421327979e-05, "loss": 0.0115, "step": 169190 }, { "grad_norm": 0.03885781019926071, "learning_rate": 4.356987629448574e-05, "loss": 0.0116, "step": 169200 }, { "grad_norm": 0.5634893774986267, "learning_rate": 4.3564410534305666e-05, "loss": 0.0147, "step": 169210 }, { "grad_norm": 0.23170028626918793, "learning_rate": 4.355894485232408e-05, "loss": 0.0145, "step": 169220 }, { "grad_norm": 0.10954861342906952, "learning_rate": 4.355347924860739e-05, "loss": 0.0136, "step": 169230 }, { "grad_norm": 0.17426444590091705, "learning_rate": 4.354801372322201e-05, "loss": 0.0201, "step": 169240 }, { "grad_norm": 0.2148808240890503, "learning_rate": 4.3542548276234374e-05, "loss": 0.0145, "step": 169250 }, { "grad_norm": 0.04992468282580376, "learning_rate": 4.3537082907710855e-05, "loss": 0.0165, "step": 169260 }, { "grad_norm": 0.0866013690829277, "learning_rate": 4.353161761771788e-05, "loss": 0.0159, "step": 169270 }, { "grad_norm": 0.1456785500049591, "learning_rate": 4.352615240632186e-05, "loss": 0.0186, "step": 169280 }, { "grad_norm": 0.06521961838006973, "learning_rate": 4.352068727358922e-05, "loss": 0.0159, "step": 169290 }, { "grad_norm": 0.1510002315044403, "learning_rate": 4.3515222219586325e-05, "loss": 0.0151, "step": 169300 }, { "grad_norm": 0.2085297554731369, "learning_rate": 4.35097572443796e-05, "loss": 0.0159, "step": 169310 }, { "grad_norm": 0.13948184251785278, "learning_rate": 4.350429234803546e-05, "loss": 0.0132, "step": 169320 }, { "grad_norm": 0.08359073847532272, "learning_rate": 4.349882753062032e-05, "loss": 0.0121, "step": 169330 }, { "grad_norm": 0.17305108904838562, "learning_rate": 4.3493362792200536e-05, "loss": 0.0142, "step": 169340 }, { "grad_norm": 0.10540008544921875, "learning_rate": 4.3487898132842544e-05, "loss": 0.0148, "step": 169350 }, { "grad_norm": 0.08458419889211655, "learning_rate": 4.348243355261276e-05, "loss": 0.0118, "step": 169360 }, { "grad_norm": 0.06986629217863083, "learning_rate": 4.347696905157753e-05, "loss": 0.0242, "step": 169370 }, { "grad_norm": 0.13760985434055328, "learning_rate": 4.3471504629803294e-05, "loss": 0.0105, "step": 169380 }, { "grad_norm": 0.09824580699205399, "learning_rate": 4.346604028735644e-05, "loss": 0.0169, "step": 169390 }, { "grad_norm": 0.07778452336788177, "learning_rate": 4.3460576024303384e-05, "loss": 0.0117, "step": 169400 }, { "grad_norm": 0.23314249515533447, "learning_rate": 4.3455111840710483e-05, "loss": 0.0088, "step": 169410 }, { "grad_norm": 0.11543798446655273, "learning_rate": 4.344964773664416e-05, "loss": 0.0134, "step": 169420 }, { "grad_norm": 0.09801921993494034, "learning_rate": 4.3444183712170814e-05, "loss": 0.0138, "step": 169430 }, { "grad_norm": 0.17446669936180115, "learning_rate": 4.3438719767356807e-05, "loss": 0.0193, "step": 169440 }, { "grad_norm": 0.12126146256923676, "learning_rate": 4.343325590226855e-05, "loss": 0.0156, "step": 169450 }, { "grad_norm": 0.09502262622117996, "learning_rate": 4.3427792116972434e-05, "loss": 0.012, "step": 169460 }, { "grad_norm": 0.12693190574645996, "learning_rate": 4.342232841153487e-05, "loss": 0.0144, "step": 169470 }, { "grad_norm": 0.12715142965316772, "learning_rate": 4.34168647860222e-05, "loss": 0.0143, "step": 169480 }, { "grad_norm": 0.13674326241016388, "learning_rate": 4.341140124050085e-05, "loss": 0.0156, "step": 169490 }, { "grad_norm": 0.09500175714492798, "learning_rate": 4.340593777503721e-05, "loss": 0.0138, "step": 169500 }, { "grad_norm": 0.10420508682727814, "learning_rate": 4.340047438969763e-05, "loss": 0.0112, "step": 169510 }, { "grad_norm": 0.15471462905406952, "learning_rate": 4.3395011084548515e-05, "loss": 0.014, "step": 169520 }, { "grad_norm": 0.10537571460008621, "learning_rate": 4.3389547859656256e-05, "loss": 0.017, "step": 169530 }, { "grad_norm": 0.14712820947170258, "learning_rate": 4.338408471508725e-05, "loss": 0.0132, "step": 169540 }, { "grad_norm": 0.0714384987950325, "learning_rate": 4.337862165090784e-05, "loss": 0.0135, "step": 169550 }, { "grad_norm": 0.16526581346988678, "learning_rate": 4.3373158667184425e-05, "loss": 0.0131, "step": 169560 }, { "grad_norm": 0.10542213916778564, "learning_rate": 4.336769576398342e-05, "loss": 0.0125, "step": 169570 }, { "grad_norm": 0.058553650975227356, "learning_rate": 4.3362232941371144e-05, "loss": 0.0148, "step": 169580 }, { "grad_norm": 0.09754208475351334, "learning_rate": 4.335677019941401e-05, "loss": 0.0142, "step": 169590 }, { "grad_norm": 0.08201830089092255, "learning_rate": 4.335130753817839e-05, "loss": 0.0142, "step": 169600 }, { "grad_norm": 0.10873664915561676, "learning_rate": 4.3345844957730675e-05, "loss": 0.0152, "step": 169610 }, { "grad_norm": 0.1977691948413849, "learning_rate": 4.334038245813721e-05, "loss": 0.0177, "step": 169620 }, { "grad_norm": 0.07927840948104858, "learning_rate": 4.333492003946439e-05, "loss": 0.0124, "step": 169630 }, { "grad_norm": 0.08439034968614578, "learning_rate": 4.332945770177858e-05, "loss": 0.0167, "step": 169640 }, { "grad_norm": 0.10284597426652908, "learning_rate": 4.332399544514618e-05, "loss": 0.0109, "step": 169650 }, { "grad_norm": 0.08856828510761261, "learning_rate": 4.331853326963352e-05, "loss": 0.0155, "step": 169660 }, { "grad_norm": 0.12537087500095367, "learning_rate": 4.331307117530698e-05, "loss": 0.0163, "step": 169670 }, { "grad_norm": 0.06404202431440353, "learning_rate": 4.330760916223296e-05, "loss": 0.0116, "step": 169680 }, { "grad_norm": 0.10410824418067932, "learning_rate": 4.3302147230477796e-05, "loss": 0.0173, "step": 169690 }, { "grad_norm": 0.5610575675964355, "learning_rate": 4.329668538010786e-05, "loss": 0.0098, "step": 169700 }, { "grad_norm": 0.09699076414108276, "learning_rate": 4.329122361118953e-05, "loss": 0.0138, "step": 169710 }, { "grad_norm": 0.07775720208883286, "learning_rate": 4.328576192378919e-05, "loss": 0.0146, "step": 169720 }, { "grad_norm": 0.13993258774280548, "learning_rate": 4.328030031797315e-05, "loss": 0.0178, "step": 169730 }, { "grad_norm": 0.08680348098278046, "learning_rate": 4.3274838793807814e-05, "loss": 0.0177, "step": 169740 }, { "grad_norm": 0.06502468138933182, "learning_rate": 4.3269377351359546e-05, "loss": 0.0126, "step": 169750 }, { "grad_norm": 0.18275102972984314, "learning_rate": 4.32639159906947e-05, "loss": 0.0175, "step": 169760 }, { "grad_norm": 0.13669629395008087, "learning_rate": 4.325845471187962e-05, "loss": 0.0152, "step": 169770 }, { "grad_norm": 0.23880226910114288, "learning_rate": 4.325299351498068e-05, "loss": 0.0133, "step": 169780 }, { "grad_norm": 0.20458856225013733, "learning_rate": 4.324753240006425e-05, "loss": 0.0141, "step": 169790 }, { "grad_norm": 0.24363726377487183, "learning_rate": 4.324207136719668e-05, "loss": 0.0155, "step": 169800 }, { "grad_norm": 0.08516134321689606, "learning_rate": 4.323661041644431e-05, "loss": 0.0122, "step": 169810 }, { "grad_norm": 0.10952642560005188, "learning_rate": 4.323114954787352e-05, "loss": 0.0123, "step": 169820 }, { "grad_norm": 0.06601034104824066, "learning_rate": 4.322568876155065e-05, "loss": 0.013, "step": 169830 }, { "grad_norm": 0.17166657745838165, "learning_rate": 4.322022805754206e-05, "loss": 0.0185, "step": 169840 }, { "grad_norm": 0.21288210153579712, "learning_rate": 4.321476743591409e-05, "loss": 0.0138, "step": 169850 }, { "grad_norm": 0.08527464419603348, "learning_rate": 4.3209306896733117e-05, "loss": 0.0116, "step": 169860 }, { "grad_norm": 0.08338645845651627, "learning_rate": 4.320384644006546e-05, "loss": 0.0091, "step": 169870 }, { "grad_norm": 0.15953071415424347, "learning_rate": 4.31983860659775e-05, "loss": 0.0112, "step": 169880 }, { "grad_norm": 0.1496579796075821, "learning_rate": 4.319292577453557e-05, "loss": 0.0139, "step": 169890 }, { "grad_norm": 0.11781926453113556, "learning_rate": 4.3187465565806005e-05, "loss": 0.0098, "step": 169900 }, { "grad_norm": 0.3143309950828552, "learning_rate": 4.318200543985517e-05, "loss": 0.0199, "step": 169910 }, { "grad_norm": 0.1326499730348587, "learning_rate": 4.317654539674942e-05, "loss": 0.0186, "step": 169920 }, { "grad_norm": 0.08626215904951096, "learning_rate": 4.317108543655508e-05, "loss": 0.0164, "step": 169930 }, { "grad_norm": 0.28117886185646057, "learning_rate": 4.316562555933849e-05, "loss": 0.0177, "step": 169940 }, { "grad_norm": 0.24409928917884827, "learning_rate": 4.3160165765166005e-05, "loss": 0.0162, "step": 169950 }, { "grad_norm": 0.1747608780860901, "learning_rate": 4.315470605410397e-05, "loss": 0.0222, "step": 169960 }, { "grad_norm": 0.22400982677936554, "learning_rate": 4.314924642621873e-05, "loss": 0.0167, "step": 169970 }, { "grad_norm": 0.18403422832489014, "learning_rate": 4.3143786881576584e-05, "loss": 0.0171, "step": 169980 }, { "grad_norm": 0.269787073135376, "learning_rate": 4.313832742024392e-05, "loss": 0.0197, "step": 169990 }, { "grad_norm": 0.2562250792980194, "learning_rate": 4.3132868042287064e-05, "loss": 0.0135, "step": 170000 }, { "grad_norm": 0.07463277131319046, "learning_rate": 4.312740874777233e-05, "loss": 0.0108, "step": 170010 }, { "grad_norm": 0.26977765560150146, "learning_rate": 4.312194953676607e-05, "loss": 0.0228, "step": 170020 }, { "grad_norm": 0.10424908995628357, "learning_rate": 4.311649040933461e-05, "loss": 0.0124, "step": 170030 }, { "grad_norm": 0.20493170619010925, "learning_rate": 4.311103136554432e-05, "loss": 0.0161, "step": 170040 }, { "grad_norm": 0.21304935216903687, "learning_rate": 4.310557240546148e-05, "loss": 0.0126, "step": 170050 }, { "grad_norm": 0.1259348839521408, "learning_rate": 4.310011352915244e-05, "loss": 0.0157, "step": 170060 }, { "grad_norm": 0.3365173041820526, "learning_rate": 4.309465473668356e-05, "loss": 0.0182, "step": 170070 }, { "grad_norm": 0.4038749039173126, "learning_rate": 4.3089196028121124e-05, "loss": 0.011, "step": 170080 }, { "grad_norm": 0.49717575311660767, "learning_rate": 4.308373740353147e-05, "loss": 0.0164, "step": 170090 }, { "grad_norm": 0.08633528649806976, "learning_rate": 4.307827886298096e-05, "loss": 0.0162, "step": 170100 }, { "grad_norm": 0.2569139003753662, "learning_rate": 4.30728204065359e-05, "loss": 0.0118, "step": 170110 }, { "grad_norm": 0.3032132089138031, "learning_rate": 4.30673620342626e-05, "loss": 0.0126, "step": 170120 }, { "grad_norm": 0.19600635766983032, "learning_rate": 4.30619037462274e-05, "loss": 0.0157, "step": 170130 }, { "grad_norm": 0.09668814390897751, "learning_rate": 4.305644554249662e-05, "loss": 0.0183, "step": 170140 }, { "grad_norm": 0.08388002216815948, "learning_rate": 4.305098742313658e-05, "loss": 0.0121, "step": 170150 }, { "grad_norm": 0.08958350121974945, "learning_rate": 4.30455293882136e-05, "loss": 0.0189, "step": 170160 }, { "grad_norm": 0.202832892537117, "learning_rate": 4.3040071437794e-05, "loss": 0.0141, "step": 170170 }, { "grad_norm": 0.1363568902015686, "learning_rate": 4.303461357194413e-05, "loss": 0.0134, "step": 170180 }, { "grad_norm": 0.11639479547739029, "learning_rate": 4.3029155790730255e-05, "loss": 0.0161, "step": 170190 }, { "grad_norm": 0.0969608873128891, "learning_rate": 4.302369809421873e-05, "loss": 0.0085, "step": 170200 }, { "grad_norm": 0.11449194699525833, "learning_rate": 4.3018240482475866e-05, "loss": 0.0113, "step": 170210 }, { "grad_norm": 0.04220336675643921, "learning_rate": 4.301278295556797e-05, "loss": 0.013, "step": 170220 }, { "grad_norm": 0.11085984855890274, "learning_rate": 4.300732551356133e-05, "loss": 0.0208, "step": 170230 }, { "grad_norm": 0.08562668412923813, "learning_rate": 4.3001868156522315e-05, "loss": 0.0147, "step": 170240 }, { "grad_norm": 0.09453747421503067, "learning_rate": 4.299641088451721e-05, "loss": 0.0145, "step": 170250 }, { "grad_norm": 0.051202382892370224, "learning_rate": 4.299095369761232e-05, "loss": 0.0196, "step": 170260 }, { "grad_norm": 0.1710200160741806, "learning_rate": 4.2985496595873956e-05, "loss": 0.0189, "step": 170270 }, { "grad_norm": 0.07727526873350143, "learning_rate": 4.298003957936843e-05, "loss": 0.012, "step": 170280 }, { "grad_norm": 0.07674337178468704, "learning_rate": 4.2974582648162064e-05, "loss": 0.0094, "step": 170290 }, { "grad_norm": 0.06918320059776306, "learning_rate": 4.296912580232114e-05, "loss": 0.0134, "step": 170300 }, { "grad_norm": 0.07738805562257767, "learning_rate": 4.296366904191197e-05, "loss": 0.0192, "step": 170310 }, { "grad_norm": 0.052749112248420715, "learning_rate": 4.295821236700089e-05, "loss": 0.0142, "step": 170320 }, { "grad_norm": 0.06852728128433228, "learning_rate": 4.295275577765415e-05, "loss": 0.0151, "step": 170330 }, { "grad_norm": 0.1051894873380661, "learning_rate": 4.294729927393809e-05, "loss": 0.015, "step": 170340 }, { "grad_norm": 0.11850859969854355, "learning_rate": 4.2941842855919e-05, "loss": 0.0182, "step": 170350 }, { "grad_norm": 0.27426469326019287, "learning_rate": 4.2936386523663205e-05, "loss": 0.0144, "step": 170360 }, { "grad_norm": 0.057666756212711334, "learning_rate": 4.293093027723696e-05, "loss": 0.0117, "step": 170370 }, { "grad_norm": 0.0665675550699234, "learning_rate": 4.292547411670658e-05, "loss": 0.0128, "step": 170380 }, { "grad_norm": 0.15798233449459076, "learning_rate": 4.2920018042138397e-05, "loss": 0.0158, "step": 170390 }, { "grad_norm": 0.11481022834777832, "learning_rate": 4.291456205359866e-05, "loss": 0.0158, "step": 170400 }, { "grad_norm": 0.14101648330688477, "learning_rate": 4.290910615115368e-05, "loss": 0.0187, "step": 170410 }, { "grad_norm": 0.13361765444278717, "learning_rate": 4.290365033486976e-05, "loss": 0.0242, "step": 170420 }, { "grad_norm": 0.19162321090698242, "learning_rate": 4.289819460481321e-05, "loss": 0.012, "step": 170430 }, { "grad_norm": 0.07742757350206375, "learning_rate": 4.289273896105027e-05, "loss": 0.0123, "step": 170440 }, { "grad_norm": 0.07985255122184753, "learning_rate": 4.2887283403647275e-05, "loss": 0.0125, "step": 170450 }, { "grad_norm": 0.12362871319055557, "learning_rate": 4.2881827932670506e-05, "loss": 0.0136, "step": 170460 }, { "grad_norm": 0.22366930544376373, "learning_rate": 4.287637254818625e-05, "loss": 0.0136, "step": 170470 }, { "grad_norm": 0.38468852639198303, "learning_rate": 4.287091725026078e-05, "loss": 0.0127, "step": 170480 }, { "grad_norm": 0.1502823382616043, "learning_rate": 4.28654620389604e-05, "loss": 0.0128, "step": 170490 }, { "grad_norm": 0.1329609751701355, "learning_rate": 4.2860006914351395e-05, "loss": 0.0156, "step": 170500 }, { "grad_norm": 0.10790030658245087, "learning_rate": 4.285455187650005e-05, "loss": 0.018, "step": 170510 }, { "grad_norm": 0.10897823423147202, "learning_rate": 4.284909692547264e-05, "loss": 0.0156, "step": 170520 }, { "grad_norm": 0.09236577898263931, "learning_rate": 4.284364206133545e-05, "loss": 0.0124, "step": 170530 }, { "grad_norm": 0.14075379073619843, "learning_rate": 4.2838187284154784e-05, "loss": 0.0097, "step": 170540 }, { "grad_norm": 0.06512415409088135, "learning_rate": 4.28327325939969e-05, "loss": 0.0165, "step": 170550 }, { "grad_norm": 0.05230589210987091, "learning_rate": 4.282727799092807e-05, "loss": 0.0143, "step": 170560 }, { "grad_norm": 0.05670732259750366, "learning_rate": 4.282182347501459e-05, "loss": 0.0091, "step": 170570 }, { "grad_norm": 0.46367427706718445, "learning_rate": 4.281636904632273e-05, "loss": 0.0129, "step": 170580 }, { "grad_norm": 0.17706675827503204, "learning_rate": 4.2810914704918774e-05, "loss": 0.0221, "step": 170590 }, { "grad_norm": 0.09122566133737564, "learning_rate": 4.280546045086898e-05, "loss": 0.0165, "step": 170600 }, { "grad_norm": 0.0723317414522171, "learning_rate": 4.280000628423966e-05, "loss": 0.0178, "step": 170610 }, { "grad_norm": 0.06541626155376434, "learning_rate": 4.2794552205097035e-05, "loss": 0.0122, "step": 170620 }, { "grad_norm": 0.11328795552253723, "learning_rate": 4.2789098213507415e-05, "loss": 0.0163, "step": 170630 }, { "grad_norm": 0.08837244659662247, "learning_rate": 4.2783644309537066e-05, "loss": 0.0134, "step": 170640 }, { "grad_norm": 0.17439958453178406, "learning_rate": 4.277819049325224e-05, "loss": 0.0199, "step": 170650 }, { "grad_norm": 0.09378871321678162, "learning_rate": 4.277273676471922e-05, "loss": 0.0185, "step": 170660 }, { "grad_norm": 0.09785623848438263, "learning_rate": 4.2767283124004275e-05, "loss": 0.0197, "step": 170670 }, { "grad_norm": 0.16709138453006744, "learning_rate": 4.276182957117367e-05, "loss": 0.0131, "step": 170680 }, { "grad_norm": 0.08225053548812866, "learning_rate": 4.275637610629367e-05, "loss": 0.0095, "step": 170690 }, { "grad_norm": 0.10385428369045258, "learning_rate": 4.275092272943053e-05, "loss": 0.0157, "step": 170700 }, { "grad_norm": 0.07167080789804459, "learning_rate": 4.274546944065055e-05, "loss": 0.0117, "step": 170710 }, { "grad_norm": 0.16913871467113495, "learning_rate": 4.2740016240019936e-05, "loss": 0.0106, "step": 170720 }, { "grad_norm": 0.06688228249549866, "learning_rate": 4.2734563127604986e-05, "loss": 0.009, "step": 170730 }, { "grad_norm": 0.07591714709997177, "learning_rate": 4.272911010347195e-05, "loss": 0.0114, "step": 170740 }, { "grad_norm": 0.1962841898202896, "learning_rate": 4.2723657167687114e-05, "loss": 0.0138, "step": 170750 }, { "grad_norm": 0.16505759954452515, "learning_rate": 4.271820432031669e-05, "loss": 0.0137, "step": 170760 }, { "grad_norm": 0.058761369436979294, "learning_rate": 4.2712751561426964e-05, "loss": 0.0142, "step": 170770 }, { "grad_norm": 0.09429913014173508, "learning_rate": 4.270729889108421e-05, "loss": 0.0134, "step": 170780 }, { "grad_norm": 0.1254139542579651, "learning_rate": 4.270184630935463e-05, "loss": 0.0167, "step": 170790 }, { "grad_norm": 0.09412485361099243, "learning_rate": 4.269639381630452e-05, "loss": 0.0114, "step": 170800 }, { "grad_norm": 0.09313551336526871, "learning_rate": 4.26909414120001e-05, "loss": 0.0134, "step": 170810 }, { "grad_norm": 0.057105448096990585, "learning_rate": 4.268548909650768e-05, "loss": 0.0142, "step": 170820 }, { "grad_norm": 0.0651569738984108, "learning_rate": 4.268003686989345e-05, "loss": 0.0106, "step": 170830 }, { "grad_norm": 0.0691990777850151, "learning_rate": 4.2674584732223685e-05, "loss": 0.0091, "step": 170840 }, { "grad_norm": 0.06828074902296066, "learning_rate": 4.266913268356463e-05, "loss": 0.0136, "step": 170850 }, { "grad_norm": 0.07897531986236572, "learning_rate": 4.266368072398255e-05, "loss": 0.0119, "step": 170860 }, { "grad_norm": 0.18100081384181976, "learning_rate": 4.265822885354366e-05, "loss": 0.0161, "step": 170870 }, { "grad_norm": 0.17581142485141754, "learning_rate": 4.265277707231422e-05, "loss": 0.014, "step": 170880 }, { "grad_norm": 0.07005386799573898, "learning_rate": 4.26473253803605e-05, "loss": 0.0164, "step": 170890 }, { "grad_norm": 0.08428460359573364, "learning_rate": 4.264187377774869e-05, "loss": 0.014, "step": 170900 }, { "grad_norm": 0.11887308955192566, "learning_rate": 4.263642226454507e-05, "loss": 0.0069, "step": 170910 }, { "grad_norm": 0.11933223158121109, "learning_rate": 4.2630970840815865e-05, "loss": 0.0116, "step": 170920 }, { "grad_norm": 0.4756585955619812, "learning_rate": 4.2625519506627345e-05, "loss": 0.0167, "step": 170930 }, { "grad_norm": 0.18530353903770447, "learning_rate": 4.26200682620457e-05, "loss": 0.0165, "step": 170940 }, { "grad_norm": 0.11448249965906143, "learning_rate": 4.26146171071372e-05, "loss": 0.0173, "step": 170950 }, { "grad_norm": 0.09067118912935257, "learning_rate": 4.2609166041968096e-05, "loss": 0.0093, "step": 170960 }, { "grad_norm": 0.10590369254350662, "learning_rate": 4.260371506660458e-05, "loss": 0.0152, "step": 170970 }, { "grad_norm": 0.17369671165943146, "learning_rate": 4.2598264181112907e-05, "loss": 0.0142, "step": 170980 }, { "grad_norm": 0.06387853622436523, "learning_rate": 4.259281338555931e-05, "loss": 0.0108, "step": 170990 }, { "grad_norm": 0.07608062028884888, "learning_rate": 4.258736268001006e-05, "loss": 0.0161, "step": 171000 }, { "grad_norm": 0.06294885277748108, "learning_rate": 4.258191206453131e-05, "loss": 0.01, "step": 171010 }, { "grad_norm": 0.09145735204219818, "learning_rate": 4.257646153918935e-05, "loss": 0.013, "step": 171020 }, { "grad_norm": 0.14266055822372437, "learning_rate": 4.257101110405041e-05, "loss": 0.0128, "step": 171030 }, { "grad_norm": 0.10189449042081833, "learning_rate": 4.256556075918067e-05, "loss": 0.0167, "step": 171040 }, { "grad_norm": 0.08327537029981613, "learning_rate": 4.256011050464639e-05, "loss": 0.0094, "step": 171050 }, { "grad_norm": 0.1020013689994812, "learning_rate": 4.255466034051379e-05, "loss": 0.0137, "step": 171060 }, { "grad_norm": 0.31051352620124817, "learning_rate": 4.2549210266849114e-05, "loss": 0.0121, "step": 171070 }, { "grad_norm": 0.11923477798700333, "learning_rate": 4.254376028371856e-05, "loss": 0.0135, "step": 171080 }, { "grad_norm": 0.10371814668178558, "learning_rate": 4.253831039118835e-05, "loss": 0.0127, "step": 171090 }, { "grad_norm": 0.08221545070409775, "learning_rate": 4.253286058932473e-05, "loss": 0.013, "step": 171100 }, { "grad_norm": 0.13305321335792542, "learning_rate": 4.252741087819389e-05, "loss": 0.0117, "step": 171110 }, { "grad_norm": 0.14858943223953247, "learning_rate": 4.252196125786206e-05, "loss": 0.0121, "step": 171120 }, { "grad_norm": 0.09068646281957626, "learning_rate": 4.251651172839546e-05, "loss": 0.011, "step": 171130 }, { "grad_norm": 0.09893818944692612, "learning_rate": 4.251106228986033e-05, "loss": 0.0112, "step": 171140 }, { "grad_norm": 0.1970691978931427, "learning_rate": 4.250561294232285e-05, "loss": 0.0126, "step": 171150 }, { "grad_norm": 0.08102095127105713, "learning_rate": 4.2500163685849234e-05, "loss": 0.0179, "step": 171160 }, { "grad_norm": 0.08446076512336731, "learning_rate": 4.249471452050573e-05, "loss": 0.0093, "step": 171170 }, { "grad_norm": 0.07753191888332367, "learning_rate": 4.248926544635854e-05, "loss": 0.0124, "step": 171180 }, { "grad_norm": 0.1992044299840927, "learning_rate": 4.248381646347384e-05, "loss": 0.0163, "step": 171190 }, { "grad_norm": 0.1433616429567337, "learning_rate": 4.247836757191788e-05, "loss": 0.0121, "step": 171200 }, { "grad_norm": 0.047855574637651443, "learning_rate": 4.247291877175685e-05, "loss": 0.015, "step": 171210 }, { "grad_norm": 0.08654972910881042, "learning_rate": 4.246747006305697e-05, "loss": 0.014, "step": 171220 }, { "grad_norm": 0.1120559424161911, "learning_rate": 4.246202144588444e-05, "loss": 0.0159, "step": 171230 }, { "grad_norm": 0.19933223724365234, "learning_rate": 4.245657292030546e-05, "loss": 0.0155, "step": 171240 }, { "grad_norm": 0.1856081634759903, "learning_rate": 4.2451124486386244e-05, "loss": 0.0144, "step": 171250 }, { "grad_norm": 0.1515270471572876, "learning_rate": 4.2445676144193e-05, "loss": 0.0123, "step": 171260 }, { "grad_norm": 0.40158435702323914, "learning_rate": 4.244022789379191e-05, "loss": 0.0173, "step": 171270 }, { "grad_norm": 0.08646170794963837, "learning_rate": 4.24347797352492e-05, "loss": 0.0142, "step": 171280 }, { "grad_norm": 0.0655069500207901, "learning_rate": 4.242933166863105e-05, "loss": 0.011, "step": 171290 }, { "grad_norm": 0.10248139500617981, "learning_rate": 4.242388369400368e-05, "loss": 0.0111, "step": 171300 }, { "grad_norm": 0.17917099595069885, "learning_rate": 4.2418435811433267e-05, "loss": 0.0197, "step": 171310 }, { "grad_norm": 0.22902445495128632, "learning_rate": 4.241298802098602e-05, "loss": 0.0171, "step": 171320 }, { "grad_norm": 0.1684633195400238, "learning_rate": 4.240754032272813e-05, "loss": 0.0152, "step": 171330 }, { "grad_norm": 0.11874749511480331, "learning_rate": 4.24020927167258e-05, "loss": 0.0121, "step": 171340 }, { "grad_norm": 0.08145739883184433, "learning_rate": 4.239664520304522e-05, "loss": 0.0105, "step": 171350 }, { "grad_norm": 0.07090058922767639, "learning_rate": 4.239119778175256e-05, "loss": 0.0104, "step": 171360 }, { "grad_norm": 0.059709444642066956, "learning_rate": 4.238575045291404e-05, "loss": 0.0122, "step": 171370 }, { "grad_norm": 0.27112722396850586, "learning_rate": 4.2380303216595845e-05, "loss": 0.0122, "step": 171380 }, { "grad_norm": 0.11483819037675858, "learning_rate": 4.2374856072864166e-05, "loss": 0.0159, "step": 171390 }, { "grad_norm": 0.1173122301697731, "learning_rate": 4.236940902178517e-05, "loss": 0.0139, "step": 171400 }, { "grad_norm": 0.09830272942781448, "learning_rate": 4.236396206342506e-05, "loss": 0.0142, "step": 171410 }, { "grad_norm": 0.17115512490272522, "learning_rate": 4.2358515197850036e-05, "loss": 0.0101, "step": 171420 }, { "grad_norm": 0.09873181581497192, "learning_rate": 4.235306842512626e-05, "loss": 0.0145, "step": 171430 }, { "grad_norm": 0.13710208237171173, "learning_rate": 4.2347621745319915e-05, "loss": 0.0151, "step": 171440 }, { "grad_norm": 0.09241931140422821, "learning_rate": 4.234217515849719e-05, "loss": 0.014, "step": 171450 }, { "grad_norm": 0.14483100175857544, "learning_rate": 4.233672866472429e-05, "loss": 0.0155, "step": 171460 }, { "grad_norm": 0.12349475175142288, "learning_rate": 4.233128226406735e-05, "loss": 0.0116, "step": 171470 }, { "grad_norm": 0.3885982930660248, "learning_rate": 4.2325835956592576e-05, "loss": 0.016, "step": 171480 }, { "grad_norm": 0.09840809553861618, "learning_rate": 4.2320389742366135e-05, "loss": 0.0135, "step": 171490 }, { "grad_norm": 0.19034349918365479, "learning_rate": 4.231494362145423e-05, "loss": 0.0179, "step": 171500 }, { "grad_norm": 0.09507421404123306, "learning_rate": 4.2309497593923e-05, "loss": 0.0137, "step": 171510 }, { "grad_norm": 0.09226329624652863, "learning_rate": 4.2304051659838625e-05, "loss": 0.0178, "step": 171520 }, { "grad_norm": 1.073192834854126, "learning_rate": 4.229860581926732e-05, "loss": 0.0181, "step": 171530 }, { "grad_norm": 0.28001296520233154, "learning_rate": 4.2293160072275194e-05, "loss": 0.0168, "step": 171540 }, { "grad_norm": 0.08318115770816803, "learning_rate": 4.228771441892846e-05, "loss": 0.0143, "step": 171550 }, { "grad_norm": 0.06115107610821724, "learning_rate": 4.2282268859293274e-05, "loss": 0.0113, "step": 171560 }, { "grad_norm": 0.06655410677194595, "learning_rate": 4.2276823393435824e-05, "loss": 0.0149, "step": 171570 }, { "grad_norm": 0.3416507840156555, "learning_rate": 4.2271378021422246e-05, "loss": 0.0169, "step": 171580 }, { "grad_norm": 0.0804290696978569, "learning_rate": 4.226593274331873e-05, "loss": 0.0128, "step": 171590 }, { "grad_norm": 0.24044324457645416, "learning_rate": 4.2260487559191444e-05, "loss": 0.0178, "step": 171600 }, { "grad_norm": 0.1064104288816452, "learning_rate": 4.225504246910652e-05, "loss": 0.0132, "step": 171610 }, { "grad_norm": 0.052678100764751434, "learning_rate": 4.224959747313014e-05, "loss": 0.012, "step": 171620 }, { "grad_norm": 0.1147439256310463, "learning_rate": 4.224415257132848e-05, "loss": 0.0145, "step": 171630 }, { "grad_norm": 0.1297350376844406, "learning_rate": 4.2238707763767696e-05, "loss": 0.0156, "step": 171640 }, { "grad_norm": 0.2711683213710785, "learning_rate": 4.223326305051393e-05, "loss": 0.0145, "step": 171650 }, { "grad_norm": 0.19071492552757263, "learning_rate": 4.2227818431633345e-05, "loss": 0.0155, "step": 171660 }, { "grad_norm": 0.17356793582439423, "learning_rate": 4.2222373907192124e-05, "loss": 0.0173, "step": 171670 }, { "grad_norm": 0.10230676829814911, "learning_rate": 4.221692947725639e-05, "loss": 0.0119, "step": 171680 }, { "grad_norm": 0.10603737831115723, "learning_rate": 4.22114851418923e-05, "loss": 0.0161, "step": 171690 }, { "grad_norm": 0.1419316679239273, "learning_rate": 4.2206040901166035e-05, "loss": 0.0163, "step": 171700 }, { "grad_norm": 0.2890738248825073, "learning_rate": 4.220059675514374e-05, "loss": 0.0183, "step": 171710 }, { "grad_norm": 0.06648923456668854, "learning_rate": 4.219515270389154e-05, "loss": 0.0143, "step": 171720 }, { "grad_norm": 0.17846453189849854, "learning_rate": 4.218970874747561e-05, "loss": 0.0154, "step": 171730 }, { "grad_norm": 0.11052192002534866, "learning_rate": 4.218426488596209e-05, "loss": 0.0157, "step": 171740 }, { "grad_norm": 0.06608334928750992, "learning_rate": 4.2178821119417145e-05, "loss": 0.0104, "step": 171750 }, { "grad_norm": 0.195216104388237, "learning_rate": 4.217337744790689e-05, "loss": 0.0181, "step": 171760 }, { "grad_norm": 0.07611877471208572, "learning_rate": 4.216793387149749e-05, "loss": 0.0132, "step": 171770 }, { "grad_norm": 0.10700737684965134, "learning_rate": 4.216249039025511e-05, "loss": 0.0116, "step": 171780 }, { "grad_norm": 0.2553168833255768, "learning_rate": 4.2157047004245856e-05, "loss": 0.0151, "step": 171790 }, { "grad_norm": 0.2916371524333954, "learning_rate": 4.215160371353588e-05, "loss": 0.0139, "step": 171800 }, { "grad_norm": 0.061486002057790756, "learning_rate": 4.214616051819132e-05, "loss": 0.0153, "step": 171810 }, { "grad_norm": 0.125982865691185, "learning_rate": 4.2140717418278355e-05, "loss": 0.0192, "step": 171820 }, { "grad_norm": 0.06766712665557861, "learning_rate": 4.2135274413863075e-05, "loss": 0.0121, "step": 171830 }, { "grad_norm": 0.14970283210277557, "learning_rate": 4.212983150501164e-05, "loss": 0.0129, "step": 171840 }, { "grad_norm": 0.15562336146831512, "learning_rate": 4.212438869179019e-05, "loss": 0.0165, "step": 171850 }, { "grad_norm": 0.16561272740364075, "learning_rate": 4.211894597426484e-05, "loss": 0.0141, "step": 171860 }, { "grad_norm": 0.11488991975784302, "learning_rate": 4.211350335250174e-05, "loss": 0.0143, "step": 171870 }, { "grad_norm": 1.062574863433838, "learning_rate": 4.210806082656702e-05, "loss": 0.0125, "step": 171880 }, { "grad_norm": 0.9485182166099548, "learning_rate": 4.210261839652683e-05, "loss": 0.0171, "step": 171890 }, { "grad_norm": 0.15667085349559784, "learning_rate": 4.209717606244726e-05, "loss": 0.0146, "step": 171900 }, { "grad_norm": 0.3074174225330353, "learning_rate": 4.2091733824394466e-05, "loss": 0.0144, "step": 171910 }, { "grad_norm": 0.213270366191864, "learning_rate": 4.2086291682434584e-05, "loss": 0.017, "step": 171920 }, { "grad_norm": 0.19040384888648987, "learning_rate": 4.208084963663372e-05, "loss": 0.0131, "step": 171930 }, { "grad_norm": 0.16971172392368317, "learning_rate": 4.207540768705801e-05, "loss": 0.0126, "step": 171940 }, { "grad_norm": 0.25892242789268494, "learning_rate": 4.206996583377358e-05, "loss": 0.0186, "step": 171950 }, { "grad_norm": 0.04587559774518013, "learning_rate": 4.206452407684656e-05, "loss": 0.0098, "step": 171960 }, { "grad_norm": 0.11574586480855942, "learning_rate": 4.205908241634306e-05, "loss": 0.0127, "step": 171970 }, { "grad_norm": 0.17898261547088623, "learning_rate": 4.20536408523292e-05, "loss": 0.0129, "step": 171980 }, { "grad_norm": 0.09971249848604202, "learning_rate": 4.2048199384871116e-05, "loss": 0.014, "step": 171990 }, { "grad_norm": 0.08768578618764877, "learning_rate": 4.2042758014034904e-05, "loss": 0.0135, "step": 172000 }, { "grad_norm": 0.11472950130701065, "learning_rate": 4.203731673988671e-05, "loss": 0.0144, "step": 172010 }, { "grad_norm": 0.10632673650979996, "learning_rate": 4.203187556249262e-05, "loss": 0.0159, "step": 172020 }, { "grad_norm": 0.17798742651939392, "learning_rate": 4.2026434481918786e-05, "loss": 0.011, "step": 172030 }, { "grad_norm": 0.08662689477205276, "learning_rate": 4.202099349823128e-05, "loss": 0.0102, "step": 172040 }, { "grad_norm": 0.11355497688055038, "learning_rate": 4.201555261149625e-05, "loss": 0.0135, "step": 172050 }, { "grad_norm": 0.10154319554567337, "learning_rate": 4.201011182177979e-05, "loss": 0.0101, "step": 172060 }, { "grad_norm": 0.33823826909065247, "learning_rate": 4.2004671129148017e-05, "loss": 0.01, "step": 172070 }, { "grad_norm": 0.147205650806427, "learning_rate": 4.1999230533667035e-05, "loss": 0.0172, "step": 172080 }, { "grad_norm": 0.10944528877735138, "learning_rate": 4.199379003540296e-05, "loss": 0.0106, "step": 172090 }, { "grad_norm": 0.08059067279100418, "learning_rate": 4.19883496344219e-05, "loss": 0.0132, "step": 172100 }, { "grad_norm": 0.206875279545784, "learning_rate": 4.198290933078994e-05, "loss": 0.0205, "step": 172110 }, { "grad_norm": 0.10879316180944443, "learning_rate": 4.19774691245732e-05, "loss": 0.013, "step": 172120 }, { "grad_norm": 0.13010066747665405, "learning_rate": 4.1972029015837805e-05, "loss": 0.0108, "step": 172130 }, { "grad_norm": 0.14836961030960083, "learning_rate": 4.1966589004649834e-05, "loss": 0.0107, "step": 172140 }, { "grad_norm": 0.08973715454339981, "learning_rate": 4.1961149091075376e-05, "loss": 0.0107, "step": 172150 }, { "grad_norm": 0.14651992917060852, "learning_rate": 4.195570927518055e-05, "loss": 0.019, "step": 172160 }, { "grad_norm": 0.23789876699447632, "learning_rate": 4.195026955703147e-05, "loss": 0.0102, "step": 172170 }, { "grad_norm": 0.2582230865955353, "learning_rate": 4.1944829936694195e-05, "loss": 0.011, "step": 172180 }, { "grad_norm": 0.14594906568527222, "learning_rate": 4.193939041423484e-05, "loss": 0.0145, "step": 172190 }, { "grad_norm": 0.10989674180746078, "learning_rate": 4.193395098971951e-05, "loss": 0.0177, "step": 172200 }, { "grad_norm": 0.11707297712564468, "learning_rate": 4.19285116632143e-05, "loss": 0.0124, "step": 172210 }, { "grad_norm": 0.15881389379501343, "learning_rate": 4.192307243478528e-05, "loss": 0.0122, "step": 172220 }, { "grad_norm": 0.08558361977338791, "learning_rate": 4.191763330449856e-05, "loss": 0.019, "step": 172230 }, { "grad_norm": 0.13857685029506683, "learning_rate": 4.191219427242024e-05, "loss": 0.014, "step": 172240 }, { "grad_norm": 0.09190009534358978, "learning_rate": 4.1906755338616376e-05, "loss": 0.0101, "step": 172250 }, { "grad_norm": 0.07511521875858307, "learning_rate": 4.190131650315308e-05, "loss": 0.0128, "step": 172260 }, { "grad_norm": 0.10255512595176697, "learning_rate": 4.189587776609643e-05, "loss": 0.0127, "step": 172270 }, { "grad_norm": 0.07406622171401978, "learning_rate": 4.1890439127512544e-05, "loss": 0.0105, "step": 172280 }, { "grad_norm": 0.06063447892665863, "learning_rate": 4.188500058746745e-05, "loss": 0.0128, "step": 172290 }, { "grad_norm": 0.12194913625717163, "learning_rate": 4.187956214602726e-05, "loss": 0.0112, "step": 172300 }, { "grad_norm": 0.24958577752113342, "learning_rate": 4.187412380325809e-05, "loss": 0.0182, "step": 172310 }, { "grad_norm": 0.29180610179901123, "learning_rate": 4.186868555922596e-05, "loss": 0.011, "step": 172320 }, { "grad_norm": 1.0297473669052124, "learning_rate": 4.186324741399697e-05, "loss": 0.0159, "step": 172330 }, { "grad_norm": 0.12822693586349487, "learning_rate": 4.185780936763722e-05, "loss": 0.0142, "step": 172340 }, { "grad_norm": 0.2691901624202728, "learning_rate": 4.185237142021279e-05, "loss": 0.0111, "step": 172350 }, { "grad_norm": 0.0683746486902237, "learning_rate": 4.184693357178972e-05, "loss": 0.0174, "step": 172360 }, { "grad_norm": 0.0814594253897667, "learning_rate": 4.1841495822434096e-05, "loss": 0.0149, "step": 172370 }, { "grad_norm": 0.33958497643470764, "learning_rate": 4.1836058172212006e-05, "loss": 0.0137, "step": 172380 }, { "grad_norm": 0.19620002806186676, "learning_rate": 4.183062062118954e-05, "loss": 0.0135, "step": 172390 }, { "grad_norm": 0.1609063744544983, "learning_rate": 4.1825183169432726e-05, "loss": 0.0172, "step": 172400 }, { "grad_norm": 0.1081048846244812, "learning_rate": 4.181974581700765e-05, "loss": 0.0175, "step": 172410 }, { "grad_norm": 0.16493700444698334, "learning_rate": 4.181430856398041e-05, "loss": 0.0163, "step": 172420 }, { "grad_norm": 0.09677969664335251, "learning_rate": 4.180887141041703e-05, "loss": 0.0181, "step": 172430 }, { "grad_norm": 0.10455245524644852, "learning_rate": 4.18034343563836e-05, "loss": 0.0158, "step": 172440 }, { "grad_norm": 0.1763758659362793, "learning_rate": 4.179799740194617e-05, "loss": 0.0118, "step": 172450 }, { "grad_norm": 0.13429993391036987, "learning_rate": 4.1792560547170844e-05, "loss": 0.0162, "step": 172460 }, { "grad_norm": 0.07118792831897736, "learning_rate": 4.178712379212363e-05, "loss": 0.008, "step": 172470 }, { "grad_norm": 0.12896953523159027, "learning_rate": 4.178168713687062e-05, "loss": 0.0139, "step": 172480 }, { "grad_norm": 0.06532953679561615, "learning_rate": 4.177625058147789e-05, "loss": 0.0132, "step": 172490 }, { "grad_norm": 0.21294701099395752, "learning_rate": 4.1770814126011474e-05, "loss": 0.0165, "step": 172500 }, { "grad_norm": 0.09966130554676056, "learning_rate": 4.176537777053742e-05, "loss": 0.0102, "step": 172510 }, { "grad_norm": 0.10281381756067276, "learning_rate": 4.17599415151218e-05, "loss": 0.0138, "step": 172520 }, { "grad_norm": 0.16330642998218536, "learning_rate": 4.17545053598307e-05, "loss": 0.0151, "step": 172530 }, { "grad_norm": 0.1988316923379898, "learning_rate": 4.1749069304730116e-05, "loss": 0.0144, "step": 172540 }, { "grad_norm": 0.14821481704711914, "learning_rate": 4.174363334988613e-05, "loss": 0.0103, "step": 172550 }, { "grad_norm": 0.12718036770820618, "learning_rate": 4.1738197495364815e-05, "loss": 0.0169, "step": 172560 }, { "grad_norm": 0.3251430094242096, "learning_rate": 4.1732761741232184e-05, "loss": 0.0121, "step": 172570 }, { "grad_norm": 0.2870536148548126, "learning_rate": 4.1727326087554307e-05, "loss": 0.0161, "step": 172580 }, { "grad_norm": 0.18769913911819458, "learning_rate": 4.1721890534397224e-05, "loss": 0.0126, "step": 172590 }, { "grad_norm": 0.08785738795995712, "learning_rate": 4.171645508182701e-05, "loss": 0.015, "step": 172600 }, { "grad_norm": 0.07799611240625381, "learning_rate": 4.171101972990966e-05, "loss": 0.0129, "step": 172610 }, { "grad_norm": 0.11448036134243011, "learning_rate": 4.170558447871126e-05, "loss": 0.0189, "step": 172620 }, { "grad_norm": 0.1389615833759308, "learning_rate": 4.170014932829782e-05, "loss": 0.0156, "step": 172630 }, { "grad_norm": 0.17371349036693573, "learning_rate": 4.169471427873543e-05, "loss": 0.0155, "step": 172640 }, { "grad_norm": 0.12672872841358185, "learning_rate": 4.16892793300901e-05, "loss": 0.0164, "step": 172650 }, { "grad_norm": 0.07398949563503265, "learning_rate": 4.1683844482427855e-05, "loss": 0.0144, "step": 172660 }, { "grad_norm": 0.07073935866355896, "learning_rate": 4.1678409735814765e-05, "loss": 0.0112, "step": 172670 }, { "grad_norm": 0.12812979519367218, "learning_rate": 4.167297509031686e-05, "loss": 0.0133, "step": 172680 }, { "grad_norm": 0.20659813284873962, "learning_rate": 4.166754054600015e-05, "loss": 0.0135, "step": 172690 }, { "grad_norm": 0.17377683520317078, "learning_rate": 4.166210610293069e-05, "loss": 0.0132, "step": 172700 }, { "grad_norm": 0.08454551547765732, "learning_rate": 4.165667176117453e-05, "loss": 0.0114, "step": 172710 }, { "grad_norm": 0.0778360441327095, "learning_rate": 4.165123752079768e-05, "loss": 0.0145, "step": 172720 }, { "grad_norm": 0.14570403099060059, "learning_rate": 4.1645803381866165e-05, "loss": 0.0142, "step": 172730 }, { "grad_norm": 0.25839075446128845, "learning_rate": 4.164036934444604e-05, "loss": 0.0156, "step": 172740 }, { "grad_norm": 0.09079430997371674, "learning_rate": 4.163493540860331e-05, "loss": 0.0143, "step": 172750 }, { "grad_norm": 0.1252485066652298, "learning_rate": 4.162950157440403e-05, "loss": 0.0098, "step": 172760 }, { "grad_norm": 0.1323385089635849, "learning_rate": 4.162406784191419e-05, "loss": 0.0139, "step": 172770 }, { "grad_norm": 0.5273572206497192, "learning_rate": 4.161863421119985e-05, "loss": 0.0128, "step": 172780 }, { "grad_norm": 0.11356174945831299, "learning_rate": 4.1613200682326994e-05, "loss": 0.01, "step": 172790 }, { "grad_norm": 0.269519567489624, "learning_rate": 4.160776725536169e-05, "loss": 0.0143, "step": 172800 }, { "grad_norm": 0.13993678987026215, "learning_rate": 4.1602333930369934e-05, "loss": 0.0083, "step": 172810 }, { "grad_norm": 0.15170495212078094, "learning_rate": 4.159690070741774e-05, "loss": 0.0142, "step": 172820 }, { "grad_norm": 0.1417302042245865, "learning_rate": 4.159146758657114e-05, "loss": 0.0106, "step": 172830 }, { "grad_norm": 0.11957783997058868, "learning_rate": 4.158603456789615e-05, "loss": 0.0122, "step": 172840 }, { "grad_norm": 0.15572649240493774, "learning_rate": 4.1580601651458795e-05, "loss": 0.0145, "step": 172850 }, { "grad_norm": 0.05649802088737488, "learning_rate": 4.157516883732506e-05, "loss": 0.0116, "step": 172860 }, { "grad_norm": 0.12040199339389801, "learning_rate": 4.156973612556098e-05, "loss": 0.0145, "step": 172870 }, { "grad_norm": 0.14121069014072418, "learning_rate": 4.156430351623259e-05, "loss": 0.0229, "step": 172880 }, { "grad_norm": 0.11900431662797928, "learning_rate": 4.1558871009405854e-05, "loss": 0.0129, "step": 172890 }, { "grad_norm": 0.08503440767526627, "learning_rate": 4.15534386051468e-05, "loss": 0.0189, "step": 172900 }, { "grad_norm": 0.10312678664922714, "learning_rate": 4.154800630352145e-05, "loss": 0.0177, "step": 172910 }, { "grad_norm": 0.14249663054943085, "learning_rate": 4.154257410459582e-05, "loss": 0.0159, "step": 172920 }, { "grad_norm": 0.10920369625091553, "learning_rate": 4.153714200843587e-05, "loss": 0.0111, "step": 172930 }, { "grad_norm": 0.09663582593202591, "learning_rate": 4.153171001510765e-05, "loss": 0.0157, "step": 172940 }, { "grad_norm": 0.13181118667125702, "learning_rate": 4.152627812467714e-05, "loss": 0.0128, "step": 172950 }, { "grad_norm": 0.15216989815235138, "learning_rate": 4.152084633721038e-05, "loss": 0.0161, "step": 172960 }, { "grad_norm": 0.1345244199037552, "learning_rate": 4.151541465277332e-05, "loss": 0.016, "step": 172970 }, { "grad_norm": 0.09155690670013428, "learning_rate": 4.1509983071431977e-05, "loss": 0.0179, "step": 172980 }, { "grad_norm": 0.10068155825138092, "learning_rate": 4.150455159325238e-05, "loss": 0.0122, "step": 172990 }, { "grad_norm": 0.13043221831321716, "learning_rate": 4.149912021830048e-05, "loss": 0.0129, "step": 173000 }, { "grad_norm": 0.12525580823421478, "learning_rate": 4.149368894664229e-05, "loss": 0.02, "step": 173010 }, { "grad_norm": 0.13839150965213776, "learning_rate": 4.148825777834382e-05, "loss": 0.0127, "step": 173020 }, { "grad_norm": 0.1547180712223053, "learning_rate": 4.148282671347107e-05, "loss": 0.021, "step": 173030 }, { "grad_norm": 0.12374240159988403, "learning_rate": 4.1477395752089997e-05, "loss": 0.0158, "step": 173040 }, { "grad_norm": 0.07901754975318909, "learning_rate": 4.1471964894266615e-05, "loss": 0.0108, "step": 173050 }, { "grad_norm": 0.07325224578380585, "learning_rate": 4.1466534140066927e-05, "loss": 0.0141, "step": 173060 }, { "grad_norm": 0.09374186396598816, "learning_rate": 4.146110348955689e-05, "loss": 0.0148, "step": 173070 }, { "grad_norm": 0.1401207596063614, "learning_rate": 4.1455672942802504e-05, "loss": 0.0088, "step": 173080 }, { "grad_norm": 0.1257949024438858, "learning_rate": 4.1450242499869765e-05, "loss": 0.0147, "step": 173090 }, { "grad_norm": 0.07118294388055801, "learning_rate": 4.144481216082466e-05, "loss": 0.0142, "step": 173100 }, { "grad_norm": 0.10389859974384308, "learning_rate": 4.143938192573315e-05, "loss": 0.013, "step": 173110 }, { "grad_norm": 0.09390342980623245, "learning_rate": 4.143395179466124e-05, "loss": 0.0162, "step": 173120 }, { "grad_norm": 0.14175866544246674, "learning_rate": 4.142852176767492e-05, "loss": 0.0191, "step": 173130 }, { "grad_norm": 0.10913334041833878, "learning_rate": 4.142309184484014e-05, "loss": 0.0149, "step": 173140 }, { "grad_norm": 0.0844852402806282, "learning_rate": 4.1417662026222884e-05, "loss": 0.0124, "step": 173150 }, { "grad_norm": 0.1061672493815422, "learning_rate": 4.141223231188913e-05, "loss": 0.0201, "step": 173160 }, { "grad_norm": 0.24230827391147614, "learning_rate": 4.140680270190489e-05, "loss": 0.0178, "step": 173170 }, { "grad_norm": 0.13140437006950378, "learning_rate": 4.140137319633609e-05, "loss": 0.0155, "step": 173180 }, { "grad_norm": 0.07981971651315689, "learning_rate": 4.139594379524872e-05, "loss": 0.0149, "step": 173190 }, { "grad_norm": 0.1276475191116333, "learning_rate": 4.139051449870878e-05, "loss": 0.0117, "step": 173200 }, { "grad_norm": 0.12747855484485626, "learning_rate": 4.13850853067822e-05, "loss": 0.0132, "step": 173210 }, { "grad_norm": 0.06839092820882797, "learning_rate": 4.137965621953497e-05, "loss": 0.0121, "step": 173220 }, { "grad_norm": 0.08438421040773392, "learning_rate": 4.1374227237033044e-05, "loss": 0.0161, "step": 173230 }, { "grad_norm": 0.1796739399433136, "learning_rate": 4.136879835934243e-05, "loss": 0.0145, "step": 173240 }, { "grad_norm": 0.06281217932701111, "learning_rate": 4.136336958652904e-05, "loss": 0.0125, "step": 173250 }, { "grad_norm": 0.05920477956533432, "learning_rate": 4.135794091865887e-05, "loss": 0.0192, "step": 173260 }, { "grad_norm": 0.15344491600990295, "learning_rate": 4.135251235579787e-05, "loss": 0.0148, "step": 173270 }, { "grad_norm": 0.1090971976518631, "learning_rate": 4.134708389801204e-05, "loss": 0.0115, "step": 173280 }, { "grad_norm": 0.07934261113405228, "learning_rate": 4.134165554536728e-05, "loss": 0.0165, "step": 173290 }, { "grad_norm": 0.12644527852535248, "learning_rate": 4.133622729792959e-05, "loss": 0.0172, "step": 173300 }, { "grad_norm": 0.060992270708084106, "learning_rate": 4.1330799155764925e-05, "loss": 0.0125, "step": 173310 }, { "grad_norm": 0.09810487926006317, "learning_rate": 4.1325371118939224e-05, "loss": 0.0132, "step": 173320 }, { "grad_norm": 0.2739485204219818, "learning_rate": 4.131994318751845e-05, "loss": 0.0191, "step": 173330 }, { "grad_norm": 0.09804918617010117, "learning_rate": 4.131451536156857e-05, "loss": 0.0136, "step": 173340 }, { "grad_norm": 0.24344965815544128, "learning_rate": 4.130908764115554e-05, "loss": 0.0163, "step": 173350 }, { "grad_norm": 0.09556464105844498, "learning_rate": 4.130366002634528e-05, "loss": 0.0139, "step": 173360 }, { "grad_norm": 0.2875368595123291, "learning_rate": 4.1298232517203764e-05, "loss": 0.013, "step": 173370 }, { "grad_norm": 0.10960564762353897, "learning_rate": 4.129280511379695e-05, "loss": 0.014, "step": 173380 }, { "grad_norm": 0.13295680284500122, "learning_rate": 4.128737781619078e-05, "loss": 0.0166, "step": 173390 }, { "grad_norm": 0.11291792243719101, "learning_rate": 4.128195062445118e-05, "loss": 0.0172, "step": 173400 }, { "grad_norm": 0.23585429787635803, "learning_rate": 4.127652353864411e-05, "loss": 0.0194, "step": 173410 }, { "grad_norm": 0.06401023268699646, "learning_rate": 4.1271096558835524e-05, "loss": 0.0086, "step": 173420 }, { "grad_norm": 0.18786969780921936, "learning_rate": 4.126566968509136e-05, "loss": 0.0128, "step": 173430 }, { "grad_norm": 0.09592420607805252, "learning_rate": 4.126024291747755e-05, "loss": 0.0108, "step": 173440 }, { "grad_norm": 0.09795533120632172, "learning_rate": 4.125481625606004e-05, "loss": 0.0166, "step": 173450 }, { "grad_norm": 0.06904099136590958, "learning_rate": 4.124938970090477e-05, "loss": 0.0119, "step": 173460 }, { "grad_norm": 0.2448839247226715, "learning_rate": 4.1243963252077675e-05, "loss": 0.0143, "step": 173470 }, { "grad_norm": 0.08008865267038345, "learning_rate": 4.1238536909644694e-05, "loss": 0.0137, "step": 173480 }, { "grad_norm": 0.04128451272845268, "learning_rate": 4.123311067367177e-05, "loss": 0.0142, "step": 173490 }, { "grad_norm": 0.08777709305286407, "learning_rate": 4.1227684544224816e-05, "loss": 0.0156, "step": 173500 }, { "grad_norm": 0.10655059665441513, "learning_rate": 4.1222258521369784e-05, "loss": 0.023, "step": 173510 }, { "grad_norm": 0.09572996944189072, "learning_rate": 4.1216832605172605e-05, "loss": 0.0131, "step": 173520 }, { "grad_norm": 0.25457414984703064, "learning_rate": 4.1211406795699186e-05, "loss": 0.0147, "step": 173530 }, { "grad_norm": 0.11856149137020111, "learning_rate": 4.120598109301548e-05, "loss": 0.016, "step": 173540 }, { "grad_norm": 0.06429441273212433, "learning_rate": 4.120055549718741e-05, "loss": 0.0108, "step": 173550 }, { "grad_norm": 0.08785238116979599, "learning_rate": 4.11951300082809e-05, "loss": 0.0131, "step": 173560 }, { "grad_norm": 0.10936155170202255, "learning_rate": 4.118970462636187e-05, "loss": 0.0154, "step": 173570 }, { "grad_norm": 0.09977562725543976, "learning_rate": 4.118427935149624e-05, "loss": 0.0094, "step": 173580 }, { "grad_norm": 0.08012434840202332, "learning_rate": 4.117885418374995e-05, "loss": 0.0107, "step": 173590 }, { "grad_norm": 0.12357717752456665, "learning_rate": 4.117342912318891e-05, "loss": 0.0088, "step": 173600 }, { "grad_norm": 0.14270631968975067, "learning_rate": 4.116800416987903e-05, "loss": 0.0133, "step": 173610 }, { "grad_norm": 0.09117395430803299, "learning_rate": 4.116257932388624e-05, "loss": 0.0149, "step": 173620 }, { "grad_norm": 0.06734819710254669, "learning_rate": 4.1157154585276484e-05, "loss": 0.0087, "step": 173630 }, { "grad_norm": 0.10394202172756195, "learning_rate": 4.115172995411561e-05, "loss": 0.0106, "step": 173640 }, { "grad_norm": 0.06264941394329071, "learning_rate": 4.114630543046958e-05, "loss": 0.0129, "step": 173650 }, { "grad_norm": 0.2186499536037445, "learning_rate": 4.11408810144043e-05, "loss": 0.0139, "step": 173660 }, { "grad_norm": 1.1384270191192627, "learning_rate": 4.1135456705985704e-05, "loss": 0.0168, "step": 173670 }, { "grad_norm": 0.11636862903833389, "learning_rate": 4.1130032505279644e-05, "loss": 0.015, "step": 173680 }, { "grad_norm": 0.07617221027612686, "learning_rate": 4.1124608412352076e-05, "loss": 0.0084, "step": 173690 }, { "grad_norm": 0.31123384833335876, "learning_rate": 4.111918442726892e-05, "loss": 0.0145, "step": 173700 }, { "grad_norm": 0.05516538769006729, "learning_rate": 4.1113760550096023e-05, "loss": 0.0112, "step": 173710 }, { "grad_norm": 0.057738859206438065, "learning_rate": 4.110833678089934e-05, "loss": 0.0156, "step": 173720 }, { "grad_norm": 0.22566600143909454, "learning_rate": 4.1102913119744754e-05, "loss": 0.0177, "step": 173730 }, { "grad_norm": 0.509624719619751, "learning_rate": 4.1097489566698185e-05, "loss": 0.013, "step": 173740 }, { "grad_norm": 0.11143341660499573, "learning_rate": 4.1092066121825516e-05, "loss": 0.0151, "step": 173750 }, { "grad_norm": 0.06630333513021469, "learning_rate": 4.108664278519265e-05, "loss": 0.0153, "step": 173760 }, { "grad_norm": 0.09295668452978134, "learning_rate": 4.1081219556865515e-05, "loss": 0.0101, "step": 173770 }, { "grad_norm": 0.07491957396268845, "learning_rate": 4.1075796436909955e-05, "loss": 0.0113, "step": 173780 }, { "grad_norm": 0.1546718180179596, "learning_rate": 4.10703734253919e-05, "loss": 0.013, "step": 173790 }, { "grad_norm": 0.10069502145051956, "learning_rate": 4.106495052237723e-05, "loss": 0.0073, "step": 173800 }, { "grad_norm": 0.3173098862171173, "learning_rate": 4.105952772793187e-05, "loss": 0.0149, "step": 173810 }, { "grad_norm": 0.05113162472844124, "learning_rate": 4.105410504212167e-05, "loss": 0.0105, "step": 173820 }, { "grad_norm": 0.05771787092089653, "learning_rate": 4.1048682465012546e-05, "loss": 0.0123, "step": 173830 }, { "grad_norm": 0.08512593060731888, "learning_rate": 4.104325999667037e-05, "loss": 0.017, "step": 173840 }, { "grad_norm": 0.09069246798753738, "learning_rate": 4.1037837637161076e-05, "loss": 0.0103, "step": 173850 }, { "grad_norm": 0.15614749491214752, "learning_rate": 4.103241538655049e-05, "loss": 0.0163, "step": 173860 }, { "grad_norm": 0.2089117467403412, "learning_rate": 4.102699324490452e-05, "loss": 0.0184, "step": 173870 }, { "grad_norm": 0.07391361147165298, "learning_rate": 4.102157121228908e-05, "loss": 0.0108, "step": 173880 }, { "grad_norm": 0.1693754941225052, "learning_rate": 4.101614928877001e-05, "loss": 0.0117, "step": 173890 }, { "grad_norm": 0.09032542258501053, "learning_rate": 4.10107274744132e-05, "loss": 0.0083, "step": 173900 }, { "grad_norm": 0.10936950892210007, "learning_rate": 4.100530576928454e-05, "loss": 0.0106, "step": 173910 }, { "grad_norm": 0.1209188848733902, "learning_rate": 4.099988417344993e-05, "loss": 0.0163, "step": 173920 }, { "grad_norm": 0.06834255158901215, "learning_rate": 4.09944626869752e-05, "loss": 0.0154, "step": 173930 }, { "grad_norm": 0.06301773339509964, "learning_rate": 4.0989041309926254e-05, "loss": 0.016, "step": 173940 }, { "grad_norm": 0.1163356676697731, "learning_rate": 4.098362004236899e-05, "loss": 0.0134, "step": 173950 }, { "grad_norm": 0.4281946122646332, "learning_rate": 4.0978198884369226e-05, "loss": 0.0158, "step": 173960 }, { "grad_norm": 0.09364087879657745, "learning_rate": 4.097277783599287e-05, "loss": 0.0134, "step": 173970 }, { "grad_norm": 0.16991454362869263, "learning_rate": 4.096735689730577e-05, "loss": 0.0111, "step": 173980 }, { "grad_norm": 0.07229039818048477, "learning_rate": 4.096193606837385e-05, "loss": 0.0124, "step": 173990 }, { "grad_norm": 0.07877857983112335, "learning_rate": 4.095651534926292e-05, "loss": 0.0156, "step": 174000 }, { "grad_norm": 0.1964758038520813, "learning_rate": 4.095109474003886e-05, "loss": 0.018, "step": 174010 }, { "grad_norm": 0.15902179479599, "learning_rate": 4.094567424076756e-05, "loss": 0.0163, "step": 174020 }, { "grad_norm": 0.12118478864431381, "learning_rate": 4.094025385151485e-05, "loss": 0.0135, "step": 174030 }, { "grad_norm": 0.08547522872686386, "learning_rate": 4.093483357234662e-05, "loss": 0.0168, "step": 174040 }, { "grad_norm": 0.13134905695915222, "learning_rate": 4.092941340332871e-05, "loss": 0.0129, "step": 174050 }, { "grad_norm": 0.2702239453792572, "learning_rate": 4.092399334452701e-05, "loss": 0.0223, "step": 174060 }, { "grad_norm": 0.11861354112625122, "learning_rate": 4.091857339600734e-05, "loss": 0.0123, "step": 174070 }, { "grad_norm": 0.08062449842691422, "learning_rate": 4.091315355783558e-05, "loss": 0.0127, "step": 174080 }, { "grad_norm": 0.06606034189462662, "learning_rate": 4.09077338300776e-05, "loss": 0.018, "step": 174090 }, { "grad_norm": 0.12087550759315491, "learning_rate": 4.090231421279923e-05, "loss": 0.013, "step": 174100 }, { "grad_norm": 0.07277371734380722, "learning_rate": 4.0896894706066324e-05, "loss": 0.0128, "step": 174110 }, { "grad_norm": 0.1065029427409172, "learning_rate": 4.089147530994474e-05, "loss": 0.014, "step": 174120 }, { "grad_norm": 0.20462460815906525, "learning_rate": 4.0886056024500355e-05, "loss": 0.0168, "step": 174130 }, { "grad_norm": 0.12851133942604065, "learning_rate": 4.088063684979898e-05, "loss": 0.014, "step": 174140 }, { "grad_norm": 0.19479690492153168, "learning_rate": 4.087521778590647e-05, "loss": 0.0151, "step": 174150 }, { "grad_norm": 0.0935564860701561, "learning_rate": 4.086979883288867e-05, "loss": 0.0126, "step": 174160 }, { "grad_norm": 0.2753884494304657, "learning_rate": 4.086437999081146e-05, "loss": 0.015, "step": 174170 }, { "grad_norm": 0.12895886600017548, "learning_rate": 4.085896125974065e-05, "loss": 0.011, "step": 174180 }, { "grad_norm": 0.12637114524841309, "learning_rate": 4.085354263974209e-05, "loss": 0.0292, "step": 174190 }, { "grad_norm": 0.6403467655181885, "learning_rate": 4.0848124130881625e-05, "loss": 0.0146, "step": 174200 }, { "grad_norm": 0.15931761264801025, "learning_rate": 4.084270573322509e-05, "loss": 0.0186, "step": 174210 }, { "grad_norm": 0.07124873250722885, "learning_rate": 4.083728744683833e-05, "loss": 0.0169, "step": 174220 }, { "grad_norm": 0.2108766883611679, "learning_rate": 4.0831869271787157e-05, "loss": 0.0174, "step": 174230 }, { "grad_norm": 0.17356941103935242, "learning_rate": 4.0826451208137454e-05, "loss": 0.0189, "step": 174240 }, { "grad_norm": 0.0840068832039833, "learning_rate": 4.082103325595502e-05, "loss": 0.0117, "step": 174250 }, { "grad_norm": 0.09739131480455399, "learning_rate": 4.08156154153057e-05, "loss": 0.0177, "step": 174260 }, { "grad_norm": 0.07111109793186188, "learning_rate": 4.081019768625533e-05, "loss": 0.014, "step": 174270 }, { "grad_norm": 0.07948340475559235, "learning_rate": 4.0804780068869726e-05, "loss": 0.0144, "step": 174280 }, { "grad_norm": 0.23996895551681519, "learning_rate": 4.079936256321472e-05, "loss": 0.0153, "step": 174290 }, { "grad_norm": 0.15024976432323456, "learning_rate": 4.079394516935616e-05, "loss": 0.0153, "step": 174300 }, { "grad_norm": 0.0892832949757576, "learning_rate": 4.078852788735986e-05, "loss": 0.0169, "step": 174310 }, { "grad_norm": 0.08940370380878448, "learning_rate": 4.078311071729163e-05, "loss": 0.0142, "step": 174320 }, { "grad_norm": 0.19110353291034698, "learning_rate": 4.077769365921731e-05, "loss": 0.0119, "step": 174330 }, { "grad_norm": 0.06812626868486404, "learning_rate": 4.0772276713202735e-05, "loss": 0.0114, "step": 174340 }, { "grad_norm": 0.0712224543094635, "learning_rate": 4.076685987931369e-05, "loss": 0.0071, "step": 174350 }, { "grad_norm": 0.10722296684980392, "learning_rate": 4.0761443157616016e-05, "loss": 0.0156, "step": 174360 }, { "grad_norm": 0.08642451465129852, "learning_rate": 4.075602654817554e-05, "loss": 0.0141, "step": 174370 }, { "grad_norm": 0.14297077059745789, "learning_rate": 4.075061005105807e-05, "loss": 0.0147, "step": 174380 }, { "grad_norm": 0.17531666159629822, "learning_rate": 4.074519366632942e-05, "loss": 0.0128, "step": 174390 }, { "grad_norm": 0.08098473399877548, "learning_rate": 4.07397773940554e-05, "loss": 0.0176, "step": 174400 }, { "grad_norm": 0.06212299317121506, "learning_rate": 4.0734361234301837e-05, "loss": 0.0078, "step": 174410 }, { "grad_norm": 0.11501622945070267, "learning_rate": 4.072894518713452e-05, "loss": 0.0168, "step": 174420 }, { "grad_norm": 0.23700065910816193, "learning_rate": 4.072352925261927e-05, "loss": 0.0158, "step": 174430 }, { "grad_norm": 0.09162556380033493, "learning_rate": 4.0718113430821904e-05, "loss": 0.0119, "step": 174440 }, { "grad_norm": 0.12718792259693146, "learning_rate": 4.071269772180824e-05, "loss": 0.0202, "step": 174450 }, { "grad_norm": 0.04970484972000122, "learning_rate": 4.070728212564405e-05, "loss": 0.011, "step": 174460 }, { "grad_norm": 0.05469265952706337, "learning_rate": 4.070186664239515e-05, "loss": 0.0105, "step": 174470 }, { "grad_norm": 0.12018833309412003, "learning_rate": 4.069645127212736e-05, "loss": 0.0131, "step": 174480 }, { "grad_norm": 0.07823044806718826, "learning_rate": 4.069103601490648e-05, "loss": 0.0126, "step": 174490 }, { "grad_norm": 0.08917832374572754, "learning_rate": 4.068562087079828e-05, "loss": 0.0151, "step": 174500 }, { "grad_norm": 0.06415777653455734, "learning_rate": 4.068020583986859e-05, "loss": 0.0183, "step": 174510 }, { "grad_norm": 0.05578971654176712, "learning_rate": 4.067479092218321e-05, "loss": 0.0145, "step": 174520 }, { "grad_norm": 0.1884126216173172, "learning_rate": 4.066937611780791e-05, "loss": 0.0148, "step": 174530 }, { "grad_norm": 0.24516798555850983, "learning_rate": 4.06639614268085e-05, "loss": 0.0116, "step": 174540 }, { "grad_norm": 0.2263590693473816, "learning_rate": 4.0658546849250774e-05, "loss": 0.0176, "step": 174550 }, { "grad_norm": 0.07752037048339844, "learning_rate": 4.0653132385200545e-05, "loss": 0.0128, "step": 174560 }, { "grad_norm": 0.07167519629001617, "learning_rate": 4.0647718034723556e-05, "loss": 0.0159, "step": 174570 }, { "grad_norm": 0.11694102734327316, "learning_rate": 4.0642303797885626e-05, "loss": 0.0095, "step": 174580 }, { "grad_norm": 0.19408908486366272, "learning_rate": 4.063688967475256e-05, "loss": 0.0175, "step": 174590 }, { "grad_norm": 0.05137784406542778, "learning_rate": 4.06314756653901e-05, "loss": 0.0157, "step": 174600 }, { "grad_norm": 1.2315995693206787, "learning_rate": 4.0626061769864065e-05, "loss": 0.0183, "step": 174610 }, { "grad_norm": 0.1252683848142624, "learning_rate": 4.062064798824022e-05, "loss": 0.0118, "step": 174620 }, { "grad_norm": 0.14560864865779877, "learning_rate": 4.061523432058439e-05, "loss": 0.0177, "step": 174630 }, { "grad_norm": 0.2337680160999298, "learning_rate": 4.06098207669623e-05, "loss": 0.0125, "step": 174640 }, { "grad_norm": 0.08854162693023682, "learning_rate": 4.060440732743975e-05, "loss": 0.0097, "step": 174650 }, { "grad_norm": 0.07895202934741974, "learning_rate": 4.059899400208254e-05, "loss": 0.0129, "step": 174660 }, { "grad_norm": 0.3178735077381134, "learning_rate": 4.059358079095641e-05, "loss": 0.0143, "step": 174670 }, { "grad_norm": 0.11561967432498932, "learning_rate": 4.058816769412716e-05, "loss": 0.0121, "step": 174680 }, { "grad_norm": 0.11059124022722244, "learning_rate": 4.058275471166055e-05, "loss": 0.0154, "step": 174690 }, { "grad_norm": 0.1596577763557434, "learning_rate": 4.057734184362239e-05, "loss": 0.0116, "step": 174700 }, { "grad_norm": 0.07723759859800339, "learning_rate": 4.05719290900784e-05, "loss": 0.0168, "step": 174710 }, { "grad_norm": 0.10633006691932678, "learning_rate": 4.0566516451094364e-05, "loss": 0.0147, "step": 174720 }, { "grad_norm": 0.09307857602834702, "learning_rate": 4.056110392673609e-05, "loss": 0.0122, "step": 174730 }, { "grad_norm": 0.10689998418092728, "learning_rate": 4.0555691517069295e-05, "loss": 0.0161, "step": 174740 }, { "grad_norm": 0.09352244436740875, "learning_rate": 4.055027922215977e-05, "loss": 0.013, "step": 174750 }, { "grad_norm": 0.11146131157875061, "learning_rate": 4.054486704207326e-05, "loss": 0.0209, "step": 174760 }, { "grad_norm": 0.09150845557451248, "learning_rate": 4.053945497687557e-05, "loss": 0.0139, "step": 174770 }, { "grad_norm": 0.11387994885444641, "learning_rate": 4.0534043026632416e-05, "loss": 0.0123, "step": 174780 }, { "grad_norm": 0.17726458609104156, "learning_rate": 4.052863119140957e-05, "loss": 0.0151, "step": 174790 }, { "grad_norm": 0.07238343358039856, "learning_rate": 4.0523219471272805e-05, "loss": 0.0151, "step": 174800 }, { "grad_norm": 0.0840802863240242, "learning_rate": 4.0517807866287896e-05, "loss": 0.0095, "step": 174810 }, { "grad_norm": 0.2628291845321655, "learning_rate": 4.051239637652054e-05, "loss": 0.0145, "step": 174820 }, { "grad_norm": 0.10478318482637405, "learning_rate": 4.0506985002036534e-05, "loss": 0.0108, "step": 174830 }, { "grad_norm": 0.16001994907855988, "learning_rate": 4.0501573742901635e-05, "loss": 0.0134, "step": 174840 }, { "grad_norm": 0.05928583815693855, "learning_rate": 4.0496162599181576e-05, "loss": 0.0181, "step": 174850 }, { "grad_norm": 0.4496844410896301, "learning_rate": 4.049075157094211e-05, "loss": 0.014, "step": 174860 }, { "grad_norm": 0.12098950892686844, "learning_rate": 4.048534065824898e-05, "loss": 0.0177, "step": 174870 }, { "grad_norm": 0.24632155895233154, "learning_rate": 4.047992986116797e-05, "loss": 0.0152, "step": 174880 }, { "grad_norm": 0.13423104584217072, "learning_rate": 4.047451917976479e-05, "loss": 0.0155, "step": 174890 }, { "grad_norm": 0.18547871708869934, "learning_rate": 4.0469108614105186e-05, "loss": 0.0216, "step": 174900 }, { "grad_norm": 0.08963152766227722, "learning_rate": 4.0463698164254926e-05, "loss": 0.0191, "step": 174910 }, { "grad_norm": 0.06770087778568268, "learning_rate": 4.045828783027973e-05, "loss": 0.0143, "step": 174920 }, { "grad_norm": 0.08753690868616104, "learning_rate": 4.045287761224535e-05, "loss": 0.012, "step": 174930 }, { "grad_norm": 0.17526666820049286, "learning_rate": 4.0447467510217504e-05, "loss": 0.014, "step": 174940 }, { "grad_norm": 0.10424112528562546, "learning_rate": 4.0442057524261975e-05, "loss": 0.0138, "step": 174950 }, { "grad_norm": 0.0674312561750412, "learning_rate": 4.043664765444445e-05, "loss": 0.0098, "step": 174960 }, { "grad_norm": 0.09113551676273346, "learning_rate": 4.0431237900830694e-05, "loss": 0.0118, "step": 174970 }, { "grad_norm": 0.09690990298986435, "learning_rate": 4.042582826348643e-05, "loss": 0.0169, "step": 174980 }, { "grad_norm": 0.05298605188727379, "learning_rate": 4.042041874247739e-05, "loss": 0.0123, "step": 174990 }, { "grad_norm": 0.06353944540023804, "learning_rate": 4.04150093378693e-05, "loss": 0.011, "step": 175000 }, { "grad_norm": 0.06847719848155975, "learning_rate": 4.0409600049727915e-05, "loss": 0.0123, "step": 175010 }, { "grad_norm": 0.0707084983587265, "learning_rate": 4.0404190878118936e-05, "loss": 0.0112, "step": 175020 }, { "grad_norm": 0.07689573615789413, "learning_rate": 4.0398781823108086e-05, "loss": 0.0147, "step": 175030 }, { "grad_norm": 0.0503820963203907, "learning_rate": 4.039337288476111e-05, "loss": 0.0149, "step": 175040 }, { "grad_norm": 0.07081872224807739, "learning_rate": 4.0387964063143726e-05, "loss": 0.015, "step": 175050 }, { "grad_norm": 0.12401974201202393, "learning_rate": 4.038255535832166e-05, "loss": 0.0165, "step": 175060 }, { "grad_norm": 0.07479910552501678, "learning_rate": 4.037714677036062e-05, "loss": 0.0172, "step": 175070 }, { "grad_norm": 0.23226507008075714, "learning_rate": 4.037173829932632e-05, "loss": 0.0124, "step": 175080 }, { "grad_norm": 0.07610074430704117, "learning_rate": 4.036632994528452e-05, "loss": 0.0125, "step": 175090 }, { "grad_norm": 0.06990476697683334, "learning_rate": 4.036092170830089e-05, "loss": 0.0154, "step": 175100 }, { "grad_norm": 0.12389940768480301, "learning_rate": 4.035551358844115e-05, "loss": 0.0113, "step": 175110 }, { "grad_norm": 0.05194637179374695, "learning_rate": 4.0350105585771034e-05, "loss": 0.0126, "step": 175120 }, { "grad_norm": 0.07748328149318695, "learning_rate": 4.034469770035627e-05, "loss": 0.0153, "step": 175130 }, { "grad_norm": 0.1668209284543991, "learning_rate": 4.033928993226251e-05, "loss": 0.0164, "step": 175140 }, { "grad_norm": 0.08871997892856598, "learning_rate": 4.0333882281555514e-05, "loss": 0.0125, "step": 175150 }, { "grad_norm": 0.07359711825847626, "learning_rate": 4.0328474748300985e-05, "loss": 0.0118, "step": 175160 }, { "grad_norm": 0.11765874922275543, "learning_rate": 4.03230673325646e-05, "loss": 0.0161, "step": 175170 }, { "grad_norm": 0.08049051463603973, "learning_rate": 4.031766003441209e-05, "loss": 0.015, "step": 175180 }, { "grad_norm": 0.1627187877893448, "learning_rate": 4.0312252853909146e-05, "loss": 0.0117, "step": 175190 }, { "grad_norm": 0.08049292862415314, "learning_rate": 4.03068457911215e-05, "loss": 0.0193, "step": 175200 }, { "grad_norm": 0.09510047733783722, "learning_rate": 4.030143884611481e-05, "loss": 0.0211, "step": 175210 }, { "grad_norm": 0.15902231633663177, "learning_rate": 4.029603201895479e-05, "loss": 0.014, "step": 175220 }, { "grad_norm": 0.09020872414112091, "learning_rate": 4.029062530970717e-05, "loss": 0.0132, "step": 175230 }, { "grad_norm": 0.11983469128608704, "learning_rate": 4.028521871843759e-05, "loss": 0.0174, "step": 175240 }, { "grad_norm": 0.08723198622465134, "learning_rate": 4.027981224521178e-05, "loss": 0.0176, "step": 175250 }, { "grad_norm": 0.16329249739646912, "learning_rate": 4.0274405890095434e-05, "loss": 0.0108, "step": 175260 }, { "grad_norm": 0.23034806549549103, "learning_rate": 4.0268999653154255e-05, "loss": 0.0143, "step": 175270 }, { "grad_norm": 0.15261930227279663, "learning_rate": 4.026359353445389e-05, "loss": 0.0178, "step": 175280 }, { "grad_norm": 0.07763433456420898, "learning_rate": 4.0258187534060065e-05, "loss": 0.0157, "step": 175290 }, { "grad_norm": 0.09788001328706741, "learning_rate": 4.0252781652038485e-05, "loss": 0.0119, "step": 175300 }, { "grad_norm": 0.11151216924190521, "learning_rate": 4.024737588845478e-05, "loss": 0.0175, "step": 175310 }, { "grad_norm": 0.0494876392185688, "learning_rate": 4.024197024337467e-05, "loss": 0.0125, "step": 175320 }, { "grad_norm": 0.07859236001968384, "learning_rate": 4.023656471686383e-05, "loss": 0.0192, "step": 175330 }, { "grad_norm": 0.1517627090215683, "learning_rate": 4.023115930898797e-05, "loss": 0.0149, "step": 175340 }, { "grad_norm": 0.1344856321811676, "learning_rate": 4.0225754019812735e-05, "loss": 0.0174, "step": 175350 }, { "grad_norm": 0.0981186181306839, "learning_rate": 4.02203488494038e-05, "loss": 0.0134, "step": 175360 }, { "grad_norm": 0.18521074950695038, "learning_rate": 4.021494379782688e-05, "loss": 0.0133, "step": 175370 }, { "grad_norm": 0.39454859495162964, "learning_rate": 4.020953886514764e-05, "loss": 0.0138, "step": 175380 }, { "grad_norm": 0.17825612425804138, "learning_rate": 4.020413405143173e-05, "loss": 0.0112, "step": 175390 }, { "grad_norm": 0.3832428455352783, "learning_rate": 4.019872935674484e-05, "loss": 0.0158, "step": 175400 }, { "grad_norm": 0.049632441252470016, "learning_rate": 4.0193324781152655e-05, "loss": 0.0139, "step": 175410 }, { "grad_norm": 0.1084444597363472, "learning_rate": 4.0187920324720824e-05, "loss": 0.0118, "step": 175420 }, { "grad_norm": 0.1324031800031662, "learning_rate": 4.018251598751502e-05, "loss": 0.0144, "step": 175430 }, { "grad_norm": 0.07236433029174805, "learning_rate": 4.017711176960092e-05, "loss": 0.0148, "step": 175440 }, { "grad_norm": 0.09233225882053375, "learning_rate": 4.01717076710442e-05, "loss": 0.0159, "step": 175450 }, { "grad_norm": 0.40378889441490173, "learning_rate": 4.016630369191049e-05, "loss": 0.0158, "step": 175460 }, { "grad_norm": 0.1789543181657791, "learning_rate": 4.016089983226548e-05, "loss": 0.0156, "step": 175470 }, { "grad_norm": 0.11669356375932693, "learning_rate": 4.0155496092174845e-05, "loss": 0.0135, "step": 175480 }, { "grad_norm": 0.08786390721797943, "learning_rate": 4.0150092471704204e-05, "loss": 0.0151, "step": 175490 }, { "grad_norm": 0.1465854048728943, "learning_rate": 4.014468897091924e-05, "loss": 0.013, "step": 175500 }, { "grad_norm": 0.06621842086315155, "learning_rate": 4.013928558988562e-05, "loss": 0.0166, "step": 175510 }, { "grad_norm": 0.06510975956916809, "learning_rate": 4.013388232866901e-05, "loss": 0.0144, "step": 175520 }, { "grad_norm": 0.10255969315767288, "learning_rate": 4.0128479187335014e-05, "loss": 0.0168, "step": 175530 }, { "grad_norm": 0.09936657547950745, "learning_rate": 4.0123076165949316e-05, "loss": 0.0176, "step": 175540 }, { "grad_norm": 0.15798591077327728, "learning_rate": 4.011767326457758e-05, "loss": 0.0156, "step": 175550 }, { "grad_norm": 0.10162319242954254, "learning_rate": 4.0112270483285446e-05, "loss": 0.0137, "step": 175560 }, { "grad_norm": 0.12053146958351135, "learning_rate": 4.010686782213855e-05, "loss": 0.0137, "step": 175570 }, { "grad_norm": 0.1686740517616272, "learning_rate": 4.010146528120256e-05, "loss": 0.0152, "step": 175580 }, { "grad_norm": 0.10859960317611694, "learning_rate": 4.00960628605431e-05, "loss": 0.0162, "step": 175590 }, { "grad_norm": 0.08647730946540833, "learning_rate": 4.0090660560225846e-05, "loss": 0.0167, "step": 175600 }, { "grad_norm": 0.19102084636688232, "learning_rate": 4.008525838031641e-05, "loss": 0.0118, "step": 175610 }, { "grad_norm": 0.06386260688304901, "learning_rate": 4.0079856320880457e-05, "loss": 0.0133, "step": 175620 }, { "grad_norm": 0.24535523355007172, "learning_rate": 4.00744543819836e-05, "loss": 0.0173, "step": 175630 }, { "grad_norm": 0.09363703429698944, "learning_rate": 4.00690525636915e-05, "loss": 0.0162, "step": 175640 }, { "grad_norm": 0.17435939610004425, "learning_rate": 4.006365086606979e-05, "loss": 0.0173, "step": 175650 }, { "grad_norm": 0.17089837789535522, "learning_rate": 4.0058249289184105e-05, "loss": 0.016, "step": 175660 }, { "grad_norm": 0.4561161398887634, "learning_rate": 4.0052847833100074e-05, "loss": 0.0124, "step": 175670 }, { "grad_norm": 0.12095379084348679, "learning_rate": 4.0047446497883335e-05, "loss": 0.0122, "step": 175680 }, { "grad_norm": 0.17750808596611023, "learning_rate": 4.004204528359951e-05, "loss": 0.0174, "step": 175690 }, { "grad_norm": 0.32904931902885437, "learning_rate": 4.003664419031425e-05, "loss": 0.0181, "step": 175700 }, { "grad_norm": 0.16028450429439545, "learning_rate": 4.0031243218093164e-05, "loss": 0.0169, "step": 175710 }, { "grad_norm": 0.0976792573928833, "learning_rate": 4.002584236700189e-05, "loss": 0.0117, "step": 175720 }, { "grad_norm": 0.17738127708435059, "learning_rate": 4.002044163710604e-05, "loss": 0.0124, "step": 175730 }, { "grad_norm": 0.1275952309370041, "learning_rate": 4.0015041028471245e-05, "loss": 0.022, "step": 175740 }, { "grad_norm": 0.13121215999126434, "learning_rate": 4.000964054116313e-05, "loss": 0.0142, "step": 175750 }, { "grad_norm": 0.11413027346134186, "learning_rate": 4.0004240175247324e-05, "loss": 0.0148, "step": 175760 }, { "grad_norm": 0.12580358982086182, "learning_rate": 3.9998839930789436e-05, "loss": 0.0131, "step": 175770 }, { "grad_norm": 0.12349849194288254, "learning_rate": 3.9993439807855074e-05, "loss": 0.0187, "step": 175780 }, { "grad_norm": 0.1950920671224594, "learning_rate": 3.998803980650987e-05, "loss": 0.016, "step": 175790 }, { "grad_norm": 0.12175533175468445, "learning_rate": 3.998263992681945e-05, "loss": 0.0146, "step": 175800 }, { "grad_norm": 0.09860040992498398, "learning_rate": 3.99772401688494e-05, "loss": 0.0132, "step": 175810 }, { "grad_norm": 0.06690989434719086, "learning_rate": 3.997184053266534e-05, "loss": 0.0117, "step": 175820 }, { "grad_norm": 0.09024856239557266, "learning_rate": 3.996644101833289e-05, "loss": 0.0102, "step": 175830 }, { "grad_norm": 0.1441049724817276, "learning_rate": 3.996104162591767e-05, "loss": 0.0146, "step": 175840 }, { "grad_norm": 0.1646593064069748, "learning_rate": 3.9955642355485256e-05, "loss": 0.0192, "step": 175850 }, { "grad_norm": 0.07735231518745422, "learning_rate": 3.995024320710127e-05, "loss": 0.0185, "step": 175860 }, { "grad_norm": 0.13588592410087585, "learning_rate": 3.9944844180831334e-05, "loss": 0.0109, "step": 175870 }, { "grad_norm": 0.06351003795862198, "learning_rate": 3.9939445276741014e-05, "loss": 0.0107, "step": 175880 }, { "grad_norm": 0.8159831762313843, "learning_rate": 3.993404649489594e-05, "loss": 0.0137, "step": 175890 }, { "grad_norm": 0.06359362602233887, "learning_rate": 3.99286478353617e-05, "loss": 0.0129, "step": 175900 }, { "grad_norm": 0.12021102011203766, "learning_rate": 3.9923249298203924e-05, "loss": 0.0162, "step": 175910 }, { "grad_norm": 0.1128435954451561, "learning_rate": 3.991785088348816e-05, "loss": 0.0151, "step": 175920 }, { "grad_norm": 0.12026005983352661, "learning_rate": 3.991245259128002e-05, "loss": 0.0168, "step": 175930 }, { "grad_norm": 0.08816404640674591, "learning_rate": 3.990705442164513e-05, "loss": 0.0119, "step": 175940 }, { "grad_norm": 0.16949614882469177, "learning_rate": 3.990165637464904e-05, "loss": 0.0146, "step": 175950 }, { "grad_norm": 0.05715477839112282, "learning_rate": 3.989625845035736e-05, "loss": 0.0117, "step": 175960 }, { "grad_norm": 0.07927803695201874, "learning_rate": 3.989086064883567e-05, "loss": 0.0174, "step": 175970 }, { "grad_norm": 0.05422087758779526, "learning_rate": 3.98854629701496e-05, "loss": 0.0123, "step": 175980 }, { "grad_norm": 0.0778641477227211, "learning_rate": 3.988006541436468e-05, "loss": 0.0129, "step": 175990 }, { "grad_norm": 0.32954171299934387, "learning_rate": 3.987466798154652e-05, "loss": 0.0097, "step": 176000 }, { "grad_norm": 0.5972574949264526, "learning_rate": 3.98692706717607e-05, "loss": 0.0125, "step": 176010 }, { "grad_norm": 0.2475002408027649, "learning_rate": 3.9863873485072834e-05, "loss": 0.0121, "step": 176020 }, { "grad_norm": 0.0835241973400116, "learning_rate": 3.985847642154845e-05, "loss": 0.0114, "step": 176030 }, { "grad_norm": 0.08219512552022934, "learning_rate": 3.9853079481253155e-05, "loss": 0.0115, "step": 176040 }, { "grad_norm": 0.12212089449167252, "learning_rate": 3.984768266425254e-05, "loss": 0.0177, "step": 176050 }, { "grad_norm": 0.1875649392604828, "learning_rate": 3.984228597061215e-05, "loss": 0.0134, "step": 176060 }, { "grad_norm": 0.07282457500696182, "learning_rate": 3.983688940039758e-05, "loss": 0.0157, "step": 176070 }, { "grad_norm": 0.09091036766767502, "learning_rate": 3.9831492953674386e-05, "loss": 0.0107, "step": 176080 }, { "grad_norm": 0.08016981929540634, "learning_rate": 3.982609663050819e-05, "loss": 0.0135, "step": 176090 }, { "grad_norm": 0.08818657696247101, "learning_rate": 3.982070043096449e-05, "loss": 0.014, "step": 176100 }, { "grad_norm": 0.10436265170574188, "learning_rate": 3.98153043551089e-05, "loss": 0.011, "step": 176110 }, { "grad_norm": 0.14106768369674683, "learning_rate": 3.9809908403006996e-05, "loss": 0.0127, "step": 176120 }, { "grad_norm": 0.4758909046649933, "learning_rate": 3.9804512574724304e-05, "loss": 0.0133, "step": 176130 }, { "grad_norm": 0.06356363743543625, "learning_rate": 3.9799116870326415e-05, "loss": 0.0134, "step": 176140 }, { "grad_norm": 0.1037377268075943, "learning_rate": 3.979372128987888e-05, "loss": 0.0114, "step": 176150 }, { "grad_norm": 0.25124308466911316, "learning_rate": 3.97883258334473e-05, "loss": 0.0191, "step": 176160 }, { "grad_norm": 0.08223887532949448, "learning_rate": 3.9782930501097174e-05, "loss": 0.0164, "step": 176170 }, { "grad_norm": 0.09734512865543365, "learning_rate": 3.9777535292894085e-05, "loss": 0.014, "step": 176180 }, { "grad_norm": 0.1871231645345688, "learning_rate": 3.977214020890362e-05, "loss": 0.0153, "step": 176190 }, { "grad_norm": 0.22297845780849457, "learning_rate": 3.976674524919129e-05, "loss": 0.0188, "step": 176200 }, { "grad_norm": 0.12158312648534775, "learning_rate": 3.976135041382266e-05, "loss": 0.0103, "step": 176210 }, { "grad_norm": 0.13054700195789337, "learning_rate": 3.975595570286329e-05, "loss": 0.0144, "step": 176220 }, { "grad_norm": 0.09858293831348419, "learning_rate": 3.975056111637876e-05, "loss": 0.0118, "step": 176230 }, { "grad_norm": 0.30275535583496094, "learning_rate": 3.974516665443456e-05, "loss": 0.0137, "step": 176240 }, { "grad_norm": 0.15386833250522614, "learning_rate": 3.973977231709627e-05, "loss": 0.0106, "step": 176250 }, { "grad_norm": 0.07859275490045547, "learning_rate": 3.973437810442943e-05, "loss": 0.014, "step": 176260 }, { "grad_norm": 0.1056126058101654, "learning_rate": 3.9728984016499614e-05, "loss": 0.0154, "step": 176270 }, { "grad_norm": 0.13238084316253662, "learning_rate": 3.972359005337232e-05, "loss": 0.0095, "step": 176280 }, { "grad_norm": 0.12538088858127594, "learning_rate": 3.971819621511311e-05, "loss": 0.0116, "step": 176290 }, { "grad_norm": 0.06602029502391815, "learning_rate": 3.9712802501787525e-05, "loss": 0.0115, "step": 176300 }, { "grad_norm": 0.2646222710609436, "learning_rate": 3.970740891346111e-05, "loss": 0.0154, "step": 176310 }, { "grad_norm": 0.15288516879081726, "learning_rate": 3.970201545019938e-05, "loss": 0.0172, "step": 176320 }, { "grad_norm": 0.1059907004237175, "learning_rate": 3.9696622112067896e-05, "loss": 0.0137, "step": 176330 }, { "grad_norm": 0.09251923859119415, "learning_rate": 3.9691228899132176e-05, "loss": 0.0129, "step": 176340 }, { "grad_norm": 0.06762640923261642, "learning_rate": 3.968583581145777e-05, "loss": 0.0137, "step": 176350 }, { "grad_norm": 0.12107201665639877, "learning_rate": 3.968044284911018e-05, "loss": 0.0116, "step": 176360 }, { "grad_norm": 0.0699080303311348, "learning_rate": 3.967505001215496e-05, "loss": 0.0126, "step": 176370 }, { "grad_norm": 0.0574205107986927, "learning_rate": 3.966965730065763e-05, "loss": 0.0119, "step": 176380 }, { "grad_norm": 0.0702875480055809, "learning_rate": 3.9664264714683725e-05, "loss": 0.0101, "step": 176390 }, { "grad_norm": 0.0568726547062397, "learning_rate": 3.965887225429874e-05, "loss": 0.0127, "step": 176400 }, { "grad_norm": 0.10041477531194687, "learning_rate": 3.9653479919568246e-05, "loss": 0.0181, "step": 176410 }, { "grad_norm": 0.10699886828660965, "learning_rate": 3.964808771055773e-05, "loss": 0.0137, "step": 176420 }, { "grad_norm": 0.1893591731786728, "learning_rate": 3.9642695627332725e-05, "loss": 0.0151, "step": 176430 }, { "grad_norm": 0.1148991659283638, "learning_rate": 3.963730366995875e-05, "loss": 0.0115, "step": 176440 }, { "grad_norm": 0.08641421049833298, "learning_rate": 3.9631911838501305e-05, "loss": 0.0092, "step": 176450 }, { "grad_norm": 0.07371451705694199, "learning_rate": 3.962652013302593e-05, "loss": 0.0105, "step": 176460 }, { "grad_norm": 0.11456634849309921, "learning_rate": 3.9621128553598124e-05, "loss": 0.0117, "step": 176470 }, { "grad_norm": 0.1099887415766716, "learning_rate": 3.961573710028342e-05, "loss": 0.0133, "step": 176480 }, { "grad_norm": 0.07271060347557068, "learning_rate": 3.961034577314729e-05, "loss": 0.0108, "step": 176490 }, { "grad_norm": 0.12061239778995514, "learning_rate": 3.960495457225528e-05, "loss": 0.0148, "step": 176500 }, { "grad_norm": 0.12498492002487183, "learning_rate": 3.959956349767291e-05, "loss": 0.0159, "step": 176510 }, { "grad_norm": 0.10977207869291306, "learning_rate": 3.9594172549465636e-05, "loss": 0.0159, "step": 176520 }, { "grad_norm": 0.09568868577480316, "learning_rate": 3.9588781727698986e-05, "loss": 0.0141, "step": 176530 }, { "grad_norm": 0.06490054726600647, "learning_rate": 3.9583391032438475e-05, "loss": 0.0169, "step": 176540 }, { "grad_norm": 0.14059101045131683, "learning_rate": 3.957800046374962e-05, "loss": 0.0118, "step": 176550 }, { "grad_norm": 0.30052539706230164, "learning_rate": 3.957261002169787e-05, "loss": 0.0156, "step": 176560 }, { "grad_norm": 0.11155841499567032, "learning_rate": 3.956721970634876e-05, "loss": 0.0175, "step": 176570 }, { "grad_norm": 0.10242150723934174, "learning_rate": 3.956182951776778e-05, "loss": 0.0132, "step": 176580 }, { "grad_norm": 0.22734960913658142, "learning_rate": 3.955643945602044e-05, "loss": 0.0126, "step": 176590 }, { "grad_norm": 0.09271740913391113, "learning_rate": 3.955104952117221e-05, "loss": 0.0117, "step": 176600 }, { "grad_norm": 0.12013103812932968, "learning_rate": 3.954565971328859e-05, "loss": 0.0127, "step": 176610 }, { "grad_norm": 0.08266016095876694, "learning_rate": 3.9540270032435103e-05, "loss": 0.0118, "step": 176620 }, { "grad_norm": 0.3617279827594757, "learning_rate": 3.9534880478677175e-05, "loss": 0.0167, "step": 176630 }, { "grad_norm": 0.05430837720632553, "learning_rate": 3.952949105208034e-05, "loss": 0.015, "step": 176640 }, { "grad_norm": 0.22238844633102417, "learning_rate": 3.9524101752710075e-05, "loss": 0.0193, "step": 176650 }, { "grad_norm": 0.16315320134162903, "learning_rate": 3.951871258063188e-05, "loss": 0.0215, "step": 176660 }, { "grad_norm": 0.101485975086689, "learning_rate": 3.951332353591121e-05, "loss": 0.0124, "step": 176670 }, { "grad_norm": 0.10981747508049011, "learning_rate": 3.950793461861355e-05, "loss": 0.016, "step": 176680 }, { "grad_norm": 0.1002514511346817, "learning_rate": 3.950254582880442e-05, "loss": 0.0106, "step": 176690 }, { "grad_norm": 0.10532841086387634, "learning_rate": 3.949715716654925e-05, "loss": 0.0156, "step": 176700 }, { "grad_norm": 0.13681060075759888, "learning_rate": 3.949176863191353e-05, "loss": 0.0118, "step": 176710 }, { "grad_norm": 0.25828003883361816, "learning_rate": 3.948638022496274e-05, "loss": 0.0177, "step": 176720 }, { "grad_norm": 0.10166580975055695, "learning_rate": 3.948099194576239e-05, "loss": 0.0153, "step": 176730 }, { "grad_norm": 0.06468251347541809, "learning_rate": 3.947560379437788e-05, "loss": 0.0139, "step": 176740 }, { "grad_norm": 0.22561520338058472, "learning_rate": 3.947021577087473e-05, "loss": 0.0179, "step": 176750 }, { "grad_norm": 0.0868208184838295, "learning_rate": 3.946482787531843e-05, "loss": 0.0116, "step": 176760 }, { "grad_norm": 0.06368863582611084, "learning_rate": 3.9459440107774385e-05, "loss": 0.0141, "step": 176770 }, { "grad_norm": 0.11835659295320511, "learning_rate": 3.94540524683081e-05, "loss": 0.015, "step": 176780 }, { "grad_norm": 0.11664307862520218, "learning_rate": 3.9448664956985036e-05, "loss": 0.0126, "step": 176790 }, { "grad_norm": 0.05940873175859451, "learning_rate": 3.9443277573870676e-05, "loss": 0.0131, "step": 176800 }, { "grad_norm": 0.09436013549566269, "learning_rate": 3.9437890319030436e-05, "loss": 0.0144, "step": 176810 }, { "grad_norm": 0.09727030992507935, "learning_rate": 3.9432503192529803e-05, "loss": 0.0121, "step": 176820 }, { "grad_norm": 0.09583064913749695, "learning_rate": 3.9427116194434256e-05, "loss": 0.0115, "step": 176830 }, { "grad_norm": 0.13427041471004486, "learning_rate": 3.9421729324809205e-05, "loss": 0.01, "step": 176840 }, { "grad_norm": 0.1436285674571991, "learning_rate": 3.941634258372013e-05, "loss": 0.0161, "step": 176850 }, { "grad_norm": 0.07874026894569397, "learning_rate": 3.941095597123249e-05, "loss": 0.0127, "step": 176860 }, { "grad_norm": 0.10501739382743835, "learning_rate": 3.940556948741175e-05, "loss": 0.013, "step": 176870 }, { "grad_norm": 0.08934084326028824, "learning_rate": 3.940018313232332e-05, "loss": 0.0126, "step": 176880 }, { "grad_norm": 0.05120185762643814, "learning_rate": 3.9394796906032675e-05, "loss": 0.0138, "step": 176890 }, { "grad_norm": 0.10451419651508331, "learning_rate": 3.9389410808605265e-05, "loss": 0.0112, "step": 176900 }, { "grad_norm": 0.1696261763572693, "learning_rate": 3.938402484010655e-05, "loss": 0.0098, "step": 176910 }, { "grad_norm": 0.07609675824642181, "learning_rate": 3.937863900060194e-05, "loss": 0.0157, "step": 176920 }, { "grad_norm": 0.053078606724739075, "learning_rate": 3.937325329015688e-05, "loss": 0.008, "step": 176930 }, { "grad_norm": 0.0670950785279274, "learning_rate": 3.936786770883686e-05, "loss": 0.0173, "step": 176940 }, { "grad_norm": 0.0755949541926384, "learning_rate": 3.936248225670726e-05, "loss": 0.0154, "step": 176950 }, { "grad_norm": 0.10280685126781464, "learning_rate": 3.935709693383354e-05, "loss": 0.0145, "step": 176960 }, { "grad_norm": 0.1322178989648819, "learning_rate": 3.935171174028115e-05, "loss": 0.0152, "step": 176970 }, { "grad_norm": 0.10458528995513916, "learning_rate": 3.934632667611553e-05, "loss": 0.013, "step": 176980 }, { "grad_norm": 0.11322692036628723, "learning_rate": 3.9340941741402084e-05, "loss": 0.0185, "step": 176990 }, { "grad_norm": 0.12230568379163742, "learning_rate": 3.9335556936206255e-05, "loss": 0.0137, "step": 177000 }, { "grad_norm": 0.15195903182029724, "learning_rate": 3.933017226059349e-05, "loss": 0.0156, "step": 177010 }, { "grad_norm": 0.1102721244096756, "learning_rate": 3.932478771462921e-05, "loss": 0.0153, "step": 177020 }, { "grad_norm": 0.12313207238912582, "learning_rate": 3.931940329837882e-05, "loss": 0.0102, "step": 177030 }, { "grad_norm": 0.12445493787527084, "learning_rate": 3.9314019011907774e-05, "loss": 0.0126, "step": 177040 }, { "grad_norm": 0.0984802395105362, "learning_rate": 3.930863485528149e-05, "loss": 0.0122, "step": 177050 }, { "grad_norm": 0.0640687644481659, "learning_rate": 3.930325082856538e-05, "loss": 0.0164, "step": 177060 }, { "grad_norm": 0.09023647755384445, "learning_rate": 3.929786693182487e-05, "loss": 0.0114, "step": 177070 }, { "grad_norm": 0.09190133213996887, "learning_rate": 3.929248316512539e-05, "loss": 0.0112, "step": 177080 }, { "grad_norm": 0.1406460702419281, "learning_rate": 3.9287099528532346e-05, "loss": 0.0132, "step": 177090 }, { "grad_norm": 0.09326710551977158, "learning_rate": 3.928171602211116e-05, "loss": 0.0128, "step": 177100 }, { "grad_norm": 0.08880529552698135, "learning_rate": 3.927633264592724e-05, "loss": 0.0216, "step": 177110 }, { "grad_norm": 0.13435165584087372, "learning_rate": 3.9270949400046e-05, "loss": 0.0177, "step": 177120 }, { "grad_norm": 0.15703411400318146, "learning_rate": 3.9265566284532854e-05, "loss": 0.0106, "step": 177130 }, { "grad_norm": 0.13202185928821564, "learning_rate": 3.9260183299453226e-05, "loss": 0.0139, "step": 177140 }, { "grad_norm": 0.09549591690301895, "learning_rate": 3.925480044487251e-05, "loss": 0.0166, "step": 177150 }, { "grad_norm": 0.28084343671798706, "learning_rate": 3.92494177208561e-05, "loss": 0.0139, "step": 177160 }, { "grad_norm": 0.12504692375659943, "learning_rate": 3.9244035127469414e-05, "loss": 0.0101, "step": 177170 }, { "grad_norm": 0.09047079086303711, "learning_rate": 3.923865266477787e-05, "loss": 0.0114, "step": 177180 }, { "grad_norm": 0.09181391447782516, "learning_rate": 3.923327033284685e-05, "loss": 0.0153, "step": 177190 }, { "grad_norm": 0.14516986906528473, "learning_rate": 3.9227888131741756e-05, "loss": 0.0142, "step": 177200 }, { "grad_norm": 0.14774630963802338, "learning_rate": 3.9222506061527995e-05, "loss": 0.0117, "step": 177210 }, { "grad_norm": 0.10921600461006165, "learning_rate": 3.921712412227097e-05, "loss": 0.0145, "step": 177220 }, { "grad_norm": 0.13342128694057465, "learning_rate": 3.921174231403606e-05, "loss": 0.0155, "step": 177230 }, { "grad_norm": 0.11073622107505798, "learning_rate": 3.920636063688866e-05, "loss": 0.0166, "step": 177240 }, { "grad_norm": 0.08728156983852386, "learning_rate": 3.9200979090894166e-05, "loss": 0.0126, "step": 177250 }, { "grad_norm": 0.17865507304668427, "learning_rate": 3.9195597676118e-05, "loss": 0.0122, "step": 177260 }, { "grad_norm": 0.10811537504196167, "learning_rate": 3.91902163926255e-05, "loss": 0.0115, "step": 177270 }, { "grad_norm": 0.44587647914886475, "learning_rate": 3.918483524048207e-05, "loss": 0.0108, "step": 177280 }, { "grad_norm": 0.0664987713098526, "learning_rate": 3.917945421975312e-05, "loss": 0.0123, "step": 177290 }, { "grad_norm": 0.06900837272405624, "learning_rate": 3.9174073330504024e-05, "loss": 0.0153, "step": 177300 }, { "grad_norm": 0.08692364394664764, "learning_rate": 3.916869257280015e-05, "loss": 0.0125, "step": 177310 }, { "grad_norm": 0.11056861281394958, "learning_rate": 3.9163311946706873e-05, "loss": 0.0116, "step": 177320 }, { "grad_norm": 0.11800209432840347, "learning_rate": 3.915793145228961e-05, "loss": 0.0141, "step": 177330 }, { "grad_norm": 0.11236980557441711, "learning_rate": 3.91525510896137e-05, "loss": 0.016, "step": 177340 }, { "grad_norm": 0.12017536908388138, "learning_rate": 3.914717085874454e-05, "loss": 0.0118, "step": 177350 }, { "grad_norm": 0.10032474994659424, "learning_rate": 3.91417907597475e-05, "loss": 0.0211, "step": 177360 }, { "grad_norm": 0.09518329054117203, "learning_rate": 3.913641079268797e-05, "loss": 0.0103, "step": 177370 }, { "grad_norm": 0.09427344799041748, "learning_rate": 3.913103095763129e-05, "loss": 0.0123, "step": 177380 }, { "grad_norm": 0.4129995107650757, "learning_rate": 3.912565125464284e-05, "loss": 0.0167, "step": 177390 }, { "grad_norm": 0.06768077611923218, "learning_rate": 3.912027168378802e-05, "loss": 0.0138, "step": 177400 }, { "grad_norm": 0.07306017726659775, "learning_rate": 3.911489224513215e-05, "loss": 0.0201, "step": 177410 }, { "grad_norm": 0.09595100581645966, "learning_rate": 3.910951293874061e-05, "loss": 0.0106, "step": 177420 }, { "grad_norm": 0.12250877916812897, "learning_rate": 3.9104133764678774e-05, "loss": 0.0143, "step": 177430 }, { "grad_norm": 0.12390798330307007, "learning_rate": 3.909875472301203e-05, "loss": 0.0163, "step": 177440 }, { "grad_norm": 0.07448119670152664, "learning_rate": 3.909337581380568e-05, "loss": 0.0129, "step": 177450 }, { "grad_norm": 0.13726487755775452, "learning_rate": 3.908799703712511e-05, "loss": 0.0203, "step": 177460 }, { "grad_norm": 0.08618812263011932, "learning_rate": 3.9082618393035675e-05, "loss": 0.0153, "step": 177470 }, { "grad_norm": 0.0889221653342247, "learning_rate": 3.907723988160276e-05, "loss": 0.0146, "step": 177480 }, { "grad_norm": 0.14876137673854828, "learning_rate": 3.907186150289167e-05, "loss": 0.0154, "step": 177490 }, { "grad_norm": 0.10258185863494873, "learning_rate": 3.906648325696778e-05, "loss": 0.014, "step": 177500 }, { "grad_norm": 0.17721378803253174, "learning_rate": 3.906110514389647e-05, "loss": 0.0168, "step": 177510 }, { "grad_norm": 0.20040424168109894, "learning_rate": 3.9055727163743036e-05, "loss": 0.0161, "step": 177520 }, { "grad_norm": 0.3110126256942749, "learning_rate": 3.905034931657285e-05, "loss": 0.0164, "step": 177530 }, { "grad_norm": 0.04937228187918663, "learning_rate": 3.904497160245125e-05, "loss": 0.0164, "step": 177540 }, { "grad_norm": 0.10903752595186234, "learning_rate": 3.903959402144362e-05, "loss": 0.018, "step": 177550 }, { "grad_norm": 0.16891272366046906, "learning_rate": 3.903421657361525e-05, "loss": 0.0219, "step": 177560 }, { "grad_norm": 0.13930782675743103, "learning_rate": 3.902883925903149e-05, "loss": 0.0137, "step": 177570 }, { "grad_norm": 0.10442612320184708, "learning_rate": 3.902346207775772e-05, "loss": 0.0146, "step": 177580 }, { "grad_norm": 0.14376771450042725, "learning_rate": 3.901808502985924e-05, "loss": 0.016, "step": 177590 }, { "grad_norm": 0.145144984126091, "learning_rate": 3.901270811540138e-05, "loss": 0.0146, "step": 177600 }, { "grad_norm": 0.08722217381000519, "learning_rate": 3.9007331334449495e-05, "loss": 0.0168, "step": 177610 }, { "grad_norm": 0.07461220771074295, "learning_rate": 3.900195468706893e-05, "loss": 0.0152, "step": 177620 }, { "grad_norm": 0.1599961519241333, "learning_rate": 3.899657817332498e-05, "loss": 0.0176, "step": 177630 }, { "grad_norm": 0.31970497965812683, "learning_rate": 3.8991201793283e-05, "loss": 0.0169, "step": 177640 }, { "grad_norm": 0.08390647172927856, "learning_rate": 3.898582554700833e-05, "loss": 0.0148, "step": 177650 }, { "grad_norm": 0.2417486011981964, "learning_rate": 3.898044943456626e-05, "loss": 0.0137, "step": 177660 }, { "grad_norm": 0.11946486681699753, "learning_rate": 3.897507345602213e-05, "loss": 0.0152, "step": 177670 }, { "grad_norm": 0.09762975573539734, "learning_rate": 3.896969761144126e-05, "loss": 0.0143, "step": 177680 }, { "grad_norm": 0.17251098155975342, "learning_rate": 3.896432190088901e-05, "loss": 0.0161, "step": 177690 }, { "grad_norm": 0.35424721240997314, "learning_rate": 3.8958946324430644e-05, "loss": 0.0108, "step": 177700 }, { "grad_norm": 0.21016989648342133, "learning_rate": 3.895357088213151e-05, "loss": 0.0196, "step": 177710 }, { "grad_norm": 0.1291424185037613, "learning_rate": 3.894819557405692e-05, "loss": 0.0148, "step": 177720 }, { "grad_norm": 0.12629517912864685, "learning_rate": 3.894282040027219e-05, "loss": 0.0134, "step": 177730 }, { "grad_norm": 0.0762491524219513, "learning_rate": 3.893744536084263e-05, "loss": 0.0129, "step": 177740 }, { "grad_norm": 0.11746494472026825, "learning_rate": 3.893207045583355e-05, "loss": 0.0144, "step": 177750 }, { "grad_norm": 0.23832716047763824, "learning_rate": 3.892669568531027e-05, "loss": 0.0152, "step": 177760 }, { "grad_norm": 0.22261203825473785, "learning_rate": 3.89213210493381e-05, "loss": 0.009, "step": 177770 }, { "grad_norm": 0.06897926330566406, "learning_rate": 3.891594654798232e-05, "loss": 0.0108, "step": 177780 }, { "grad_norm": 0.08293585479259491, "learning_rate": 3.891057218130826e-05, "loss": 0.0172, "step": 177790 }, { "grad_norm": 0.067166268825531, "learning_rate": 3.8905197949381226e-05, "loss": 0.0168, "step": 177800 }, { "grad_norm": 0.13925404846668243, "learning_rate": 3.889982385226651e-05, "loss": 0.0164, "step": 177810 }, { "grad_norm": 0.15776698291301727, "learning_rate": 3.88944498900294e-05, "loss": 0.0138, "step": 177820 }, { "grad_norm": 0.08502274006605148, "learning_rate": 3.888907606273523e-05, "loss": 0.014, "step": 177830 }, { "grad_norm": 0.24024413526058197, "learning_rate": 3.8883702370449264e-05, "loss": 0.0131, "step": 177840 }, { "grad_norm": 0.1171172633767128, "learning_rate": 3.887832881323681e-05, "loss": 0.0121, "step": 177850 }, { "grad_norm": 0.08617110550403595, "learning_rate": 3.887295539116317e-05, "loss": 0.0189, "step": 177860 }, { "grad_norm": 0.10413450747728348, "learning_rate": 3.8867582104293626e-05, "loss": 0.0147, "step": 177870 }, { "grad_norm": 0.07070209830999374, "learning_rate": 3.886220895269346e-05, "loss": 0.0184, "step": 177880 }, { "grad_norm": 0.04452076926827431, "learning_rate": 3.885683593642797e-05, "loss": 0.0082, "step": 177890 }, { "grad_norm": 0.04198768734931946, "learning_rate": 3.885146305556247e-05, "loss": 0.0102, "step": 177900 }, { "grad_norm": 0.07762367278337479, "learning_rate": 3.8846090310162195e-05, "loss": 0.0124, "step": 177910 }, { "grad_norm": 0.05211000517010689, "learning_rate": 3.8840717700292453e-05, "loss": 0.0128, "step": 177920 }, { "grad_norm": 0.44720444083213806, "learning_rate": 3.883534522601853e-05, "loss": 0.0152, "step": 177930 }, { "grad_norm": 0.08739012479782104, "learning_rate": 3.882997288740573e-05, "loss": 0.0118, "step": 177940 }, { "grad_norm": 0.20234887301921844, "learning_rate": 3.8824600684519276e-05, "loss": 0.0162, "step": 177950 }, { "grad_norm": 0.14225326478481293, "learning_rate": 3.881922861742447e-05, "loss": 0.0135, "step": 177960 }, { "grad_norm": 0.07868689298629761, "learning_rate": 3.8813856686186626e-05, "loss": 0.0121, "step": 177970 }, { "grad_norm": 0.07427767664194107, "learning_rate": 3.8808484890870964e-05, "loss": 0.0119, "step": 177980 }, { "grad_norm": 0.05594661086797714, "learning_rate": 3.880311323154277e-05, "loss": 0.0082, "step": 177990 }, { "grad_norm": 0.25964561104774475, "learning_rate": 3.879774170826732e-05, "loss": 0.0135, "step": 178000 }, { "grad_norm": 0.0861363336443901, "learning_rate": 3.8792370321109915e-05, "loss": 0.0161, "step": 178010 }, { "grad_norm": 0.07098454236984253, "learning_rate": 3.878699907013577e-05, "loss": 0.0132, "step": 178020 }, { "grad_norm": 0.08101367205381393, "learning_rate": 3.8781627955410164e-05, "loss": 0.0123, "step": 178030 }, { "grad_norm": 0.15622399747371674, "learning_rate": 3.87762569769984e-05, "loss": 0.0119, "step": 178040 }, { "grad_norm": 0.1911545991897583, "learning_rate": 3.877088613496568e-05, "loss": 0.0123, "step": 178050 }, { "grad_norm": 0.0724591612815857, "learning_rate": 3.8765515429377306e-05, "loss": 0.0129, "step": 178060 }, { "grad_norm": 0.07839687168598175, "learning_rate": 3.876014486029852e-05, "loss": 0.0141, "step": 178070 }, { "grad_norm": 0.12836147844791412, "learning_rate": 3.875477442779461e-05, "loss": 0.0107, "step": 178080 }, { "grad_norm": 0.2655600309371948, "learning_rate": 3.874940413193079e-05, "loss": 0.0209, "step": 178090 }, { "grad_norm": 0.12399900704622269, "learning_rate": 3.874403397277233e-05, "loss": 0.0175, "step": 178100 }, { "grad_norm": 0.08219838887453079, "learning_rate": 3.873866395038448e-05, "loss": 0.0151, "step": 178110 }, { "grad_norm": 0.13210529088974, "learning_rate": 3.873329406483252e-05, "loss": 0.013, "step": 178120 }, { "grad_norm": 0.10569901019334793, "learning_rate": 3.872792431618165e-05, "loss": 0.0088, "step": 178130 }, { "grad_norm": 0.10553359985351562, "learning_rate": 3.872255470449715e-05, "loss": 0.0151, "step": 178140 }, { "grad_norm": 0.09039313346147537, "learning_rate": 3.8717185229844276e-05, "loss": 0.0145, "step": 178150 }, { "grad_norm": 0.05997099354863167, "learning_rate": 3.8711815892288236e-05, "loss": 0.0088, "step": 178160 }, { "grad_norm": 0.15296316146850586, "learning_rate": 3.870644669189428e-05, "loss": 0.0179, "step": 178170 }, { "grad_norm": 0.10711739957332611, "learning_rate": 3.8701077628727666e-05, "loss": 0.0119, "step": 178180 }, { "grad_norm": 0.274679958820343, "learning_rate": 3.869570870285365e-05, "loss": 0.0139, "step": 178190 }, { "grad_norm": 0.26516392827033997, "learning_rate": 3.8690339914337424e-05, "loss": 0.0181, "step": 178200 }, { "grad_norm": 0.12497007846832275, "learning_rate": 3.868497126324425e-05, "loss": 0.0158, "step": 178210 }, { "grad_norm": 0.10526420921087265, "learning_rate": 3.867960274963937e-05, "loss": 0.0133, "step": 178220 }, { "grad_norm": 0.06470349431037903, "learning_rate": 3.8674234373587984e-05, "loss": 0.0121, "step": 178230 }, { "grad_norm": 0.122550368309021, "learning_rate": 3.8668866135155354e-05, "loss": 0.0134, "step": 178240 }, { "grad_norm": 0.16378535330295563, "learning_rate": 3.866349803440669e-05, "loss": 0.0203, "step": 178250 }, { "grad_norm": 0.09828949719667435, "learning_rate": 3.865813007140725e-05, "loss": 0.0125, "step": 178260 }, { "grad_norm": 0.10013251006603241, "learning_rate": 3.8652762246222225e-05, "loss": 0.011, "step": 178270 }, { "grad_norm": 0.06783870607614517, "learning_rate": 3.864739455891685e-05, "loss": 0.0125, "step": 178280 }, { "grad_norm": 0.170048788189888, "learning_rate": 3.8642027009556355e-05, "loss": 0.0169, "step": 178290 }, { "grad_norm": 0.14590515196323395, "learning_rate": 3.863665959820596e-05, "loss": 0.0141, "step": 178300 }, { "grad_norm": 0.14906685054302216, "learning_rate": 3.863129232493086e-05, "loss": 0.017, "step": 178310 }, { "grad_norm": 0.08020821958780289, "learning_rate": 3.86259251897963e-05, "loss": 0.0096, "step": 178320 }, { "grad_norm": 0.07314173877239227, "learning_rate": 3.8620558192867504e-05, "loss": 0.0172, "step": 178330 }, { "grad_norm": 0.13064298033714294, "learning_rate": 3.8615191334209654e-05, "loss": 0.0123, "step": 178340 }, { "grad_norm": 0.09112795442342758, "learning_rate": 3.860982461388798e-05, "loss": 0.0306, "step": 178350 }, { "grad_norm": 0.16482573747634888, "learning_rate": 3.8604458031967706e-05, "loss": 0.0095, "step": 178360 }, { "grad_norm": 0.1445361226797104, "learning_rate": 3.859909158851401e-05, "loss": 0.0143, "step": 178370 }, { "grad_norm": 0.11848839372396469, "learning_rate": 3.859372528359211e-05, "loss": 0.0138, "step": 178380 }, { "grad_norm": 0.1374588906764984, "learning_rate": 3.8588359117267225e-05, "loss": 0.0154, "step": 178390 }, { "grad_norm": 0.12152532488107681, "learning_rate": 3.858299308960456e-05, "loss": 0.0187, "step": 178400 }, { "grad_norm": 0.18555879592895508, "learning_rate": 3.85776272006693e-05, "loss": 0.0188, "step": 178410 }, { "grad_norm": 0.09731388837099075, "learning_rate": 3.857226145052665e-05, "loss": 0.0142, "step": 178420 }, { "grad_norm": 0.10490664094686508, "learning_rate": 3.8566895839241814e-05, "loss": 0.0142, "step": 178430 }, { "grad_norm": 0.14236238598823547, "learning_rate": 3.856153036688e-05, "loss": 0.0133, "step": 178440 }, { "grad_norm": 0.1398921012878418, "learning_rate": 3.855616503350638e-05, "loss": 0.0176, "step": 178450 }, { "grad_norm": 0.19937527179718018, "learning_rate": 3.855079983918616e-05, "loss": 0.0231, "step": 178460 }, { "grad_norm": 0.12491186708211899, "learning_rate": 3.854543478398454e-05, "loss": 0.0097, "step": 178470 }, { "grad_norm": 2.0656590461730957, "learning_rate": 3.854006986796669e-05, "loss": 0.0184, "step": 178480 }, { "grad_norm": 0.2330005168914795, "learning_rate": 3.8534705091197824e-05, "loss": 0.015, "step": 178490 }, { "grad_norm": 0.10720840096473694, "learning_rate": 3.852934045374311e-05, "loss": 0.0129, "step": 178500 }, { "grad_norm": 0.1630605012178421, "learning_rate": 3.852397595566774e-05, "loss": 0.0112, "step": 178510 }, { "grad_norm": 0.12403947860002518, "learning_rate": 3.851861159703689e-05, "loss": 0.0093, "step": 178520 }, { "grad_norm": 0.13579480350017548, "learning_rate": 3.851324737791576e-05, "loss": 0.0122, "step": 178530 }, { "grad_norm": 0.09979815781116486, "learning_rate": 3.850788329836952e-05, "loss": 0.0172, "step": 178540 }, { "grad_norm": 0.21245253086090088, "learning_rate": 3.850251935846334e-05, "loss": 0.0179, "step": 178550 }, { "grad_norm": 0.26102954149246216, "learning_rate": 3.849715555826241e-05, "loss": 0.0143, "step": 178560 }, { "grad_norm": 0.09343242645263672, "learning_rate": 3.8491791897831903e-05, "loss": 0.0155, "step": 178570 }, { "grad_norm": 0.17937028408050537, "learning_rate": 3.8486428377236993e-05, "loss": 0.0129, "step": 178580 }, { "grad_norm": 0.08052541315555573, "learning_rate": 3.8481064996542836e-05, "loss": 0.014, "step": 178590 }, { "grad_norm": 0.0912446603178978, "learning_rate": 3.847570175581462e-05, "loss": 0.0127, "step": 178600 }, { "grad_norm": 0.2650527358055115, "learning_rate": 3.847033865511752e-05, "loss": 0.0145, "step": 178610 }, { "grad_norm": 0.1879325807094574, "learning_rate": 3.846497569451668e-05, "loss": 0.016, "step": 178620 }, { "grad_norm": 0.15487287938594818, "learning_rate": 3.845961287407727e-05, "loss": 0.0141, "step": 178630 }, { "grad_norm": 0.14529114961624146, "learning_rate": 3.845425019386446e-05, "loss": 0.0157, "step": 178640 }, { "grad_norm": 0.1307639479637146, "learning_rate": 3.844888765394343e-05, "loss": 0.0133, "step": 178650 }, { "grad_norm": 0.09996423125267029, "learning_rate": 3.844352525437931e-05, "loss": 0.0167, "step": 178660 }, { "grad_norm": 0.10225720703601837, "learning_rate": 3.843816299523726e-05, "loss": 0.0156, "step": 178670 }, { "grad_norm": 0.1132054328918457, "learning_rate": 3.843280087658245e-05, "loss": 0.0143, "step": 178680 }, { "grad_norm": 0.14752711355686188, "learning_rate": 3.842743889848005e-05, "loss": 0.0229, "step": 178690 }, { "grad_norm": 0.0614563524723053, "learning_rate": 3.842207706099518e-05, "loss": 0.0115, "step": 178700 }, { "grad_norm": 0.15913423895835876, "learning_rate": 3.8416715364193e-05, "loss": 0.0222, "step": 178710 }, { "grad_norm": 0.11347267031669617, "learning_rate": 3.841135380813868e-05, "loss": 0.012, "step": 178720 }, { "grad_norm": 0.11874494701623917, "learning_rate": 3.840599239289733e-05, "loss": 0.0082, "step": 178730 }, { "grad_norm": 0.12522844970226288, "learning_rate": 3.8400631118534134e-05, "loss": 0.0154, "step": 178740 }, { "grad_norm": 0.11426962167024612, "learning_rate": 3.839526998511421e-05, "loss": 0.0112, "step": 178750 }, { "grad_norm": 0.1372041553258896, "learning_rate": 3.838990899270273e-05, "loss": 0.0208, "step": 178760 }, { "grad_norm": 0.08609359711408615, "learning_rate": 3.838454814136481e-05, "loss": 0.0162, "step": 178770 }, { "grad_norm": 0.09936109930276871, "learning_rate": 3.8379187431165586e-05, "loss": 0.0132, "step": 178780 }, { "grad_norm": 0.07390548288822174, "learning_rate": 3.8373826862170225e-05, "loss": 0.0107, "step": 178790 }, { "grad_norm": 0.23211754858493805, "learning_rate": 3.836846643444383e-05, "loss": 0.0131, "step": 178800 }, { "grad_norm": 0.12082918733358383, "learning_rate": 3.8363106148051556e-05, "loss": 0.0166, "step": 178810 }, { "grad_norm": 0.19809959828853607, "learning_rate": 3.8357746003058514e-05, "loss": 0.0195, "step": 178820 }, { "grad_norm": 0.08203679323196411, "learning_rate": 3.835238599952988e-05, "loss": 0.0108, "step": 178830 }, { "grad_norm": 0.06439997255802155, "learning_rate": 3.8347026137530736e-05, "loss": 0.0128, "step": 178840 }, { "grad_norm": 0.07540611177682877, "learning_rate": 3.834166641712622e-05, "loss": 0.0145, "step": 178850 }, { "grad_norm": 0.16448158025741577, "learning_rate": 3.833630683838149e-05, "loss": 0.0137, "step": 178860 }, { "grad_norm": 0.15644019842147827, "learning_rate": 3.833094740136162e-05, "loss": 0.0135, "step": 178870 }, { "grad_norm": 0.11339075118303299, "learning_rate": 3.832558810613176e-05, "loss": 0.0139, "step": 178880 }, { "grad_norm": 0.0799371749162674, "learning_rate": 3.832022895275702e-05, "loss": 0.0126, "step": 178890 }, { "grad_norm": 0.0783083587884903, "learning_rate": 3.8314869941302554e-05, "loss": 0.0151, "step": 178900 }, { "grad_norm": 0.05860588327050209, "learning_rate": 3.830951107183343e-05, "loss": 0.0146, "step": 178910 }, { "grad_norm": 0.09603232890367508, "learning_rate": 3.830415234441478e-05, "loss": 0.0148, "step": 178920 }, { "grad_norm": 0.07168231159448624, "learning_rate": 3.829879375911174e-05, "loss": 0.0131, "step": 178930 }, { "grad_norm": 0.12475262582302094, "learning_rate": 3.829343531598939e-05, "loss": 0.0168, "step": 178940 }, { "grad_norm": 0.15999998152256012, "learning_rate": 3.828807701511285e-05, "loss": 0.0186, "step": 178950 }, { "grad_norm": 0.08787482976913452, "learning_rate": 3.828271885654723e-05, "loss": 0.0112, "step": 178960 }, { "grad_norm": 0.08398980647325516, "learning_rate": 3.827736084035766e-05, "loss": 0.0128, "step": 178970 }, { "grad_norm": 0.1266966462135315, "learning_rate": 3.827200296660921e-05, "loss": 0.0174, "step": 178980 }, { "grad_norm": 0.1133415475487709, "learning_rate": 3.8266645235367e-05, "loss": 0.0131, "step": 178990 }, { "grad_norm": 0.12848803400993347, "learning_rate": 3.8261287646696106e-05, "loss": 0.017, "step": 179000 }, { "grad_norm": 0.06479045748710632, "learning_rate": 3.8255930200661695e-05, "loss": 0.0122, "step": 179010 }, { "grad_norm": 0.13389155268669128, "learning_rate": 3.825057289732879e-05, "loss": 0.0144, "step": 179020 }, { "grad_norm": 0.08403977006673813, "learning_rate": 3.824521573676251e-05, "loss": 0.0125, "step": 179030 }, { "grad_norm": 0.11793902516365051, "learning_rate": 3.8239858719027985e-05, "loss": 0.0173, "step": 179040 }, { "grad_norm": 0.13258300721645355, "learning_rate": 3.823450184419026e-05, "loss": 0.0129, "step": 179050 }, { "grad_norm": 0.06952225416898727, "learning_rate": 3.8229145112314454e-05, "loss": 0.0098, "step": 179060 }, { "grad_norm": 0.05327184498310089, "learning_rate": 3.8223788523465635e-05, "loss": 0.0102, "step": 179070 }, { "grad_norm": 0.13315026462078094, "learning_rate": 3.821843207770893e-05, "loss": 0.0189, "step": 179080 }, { "grad_norm": 0.188589408993721, "learning_rate": 3.821307577510938e-05, "loss": 0.0148, "step": 179090 }, { "grad_norm": 0.10302866250276566, "learning_rate": 3.8207719615732084e-05, "loss": 0.0102, "step": 179100 }, { "grad_norm": 0.09059485793113708, "learning_rate": 3.8202363599642146e-05, "loss": 0.017, "step": 179110 }, { "grad_norm": 0.11199209094047546, "learning_rate": 3.819700772690462e-05, "loss": 0.0172, "step": 179120 }, { "grad_norm": 0.11175256222486496, "learning_rate": 3.8191651997584585e-05, "loss": 0.0147, "step": 179130 }, { "grad_norm": 0.18973949551582336, "learning_rate": 3.8186296411747133e-05, "loss": 0.0169, "step": 179140 }, { "grad_norm": 0.11287695169448853, "learning_rate": 3.8180940969457345e-05, "loss": 0.0233, "step": 179150 }, { "grad_norm": 0.1837257593870163, "learning_rate": 3.817558567078028e-05, "loss": 0.0158, "step": 179160 }, { "grad_norm": 0.33593428134918213, "learning_rate": 3.8170230515781006e-05, "loss": 0.0176, "step": 179170 }, { "grad_norm": 0.06957712769508362, "learning_rate": 3.81648755045246e-05, "loss": 0.0202, "step": 179180 }, { "grad_norm": 0.1379140466451645, "learning_rate": 3.815952063707613e-05, "loss": 0.0167, "step": 179190 }, { "grad_norm": 0.050216637551784515, "learning_rate": 3.8154165913500665e-05, "loss": 0.0111, "step": 179200 }, { "grad_norm": 0.16953293979167938, "learning_rate": 3.814881133386327e-05, "loss": 0.0142, "step": 179210 }, { "grad_norm": 0.11363812536001205, "learning_rate": 3.8143456898229006e-05, "loss": 0.0134, "step": 179220 }, { "grad_norm": 0.09470388293266296, "learning_rate": 3.813810260666293e-05, "loss": 0.0127, "step": 179230 }, { "grad_norm": 0.07227140665054321, "learning_rate": 3.813274845923012e-05, "loss": 0.01, "step": 179240 }, { "grad_norm": 0.118209607899189, "learning_rate": 3.812739445599562e-05, "loss": 0.017, "step": 179250 }, { "grad_norm": 0.1514015942811966, "learning_rate": 3.812204059702447e-05, "loss": 0.0147, "step": 179260 }, { "grad_norm": 0.07678141444921494, "learning_rate": 3.811668688238175e-05, "loss": 0.0208, "step": 179270 }, { "grad_norm": 0.10503491759300232, "learning_rate": 3.811133331213251e-05, "loss": 0.0131, "step": 179280 }, { "grad_norm": 0.08582286536693573, "learning_rate": 3.8105979886341795e-05, "loss": 0.0111, "step": 179290 }, { "grad_norm": 0.1395944058895111, "learning_rate": 3.810062660507465e-05, "loss": 0.0204, "step": 179300 }, { "grad_norm": 0.05285784229636192, "learning_rate": 3.8095273468396115e-05, "loss": 0.0094, "step": 179310 }, { "grad_norm": 0.07072566449642181, "learning_rate": 3.8089920476371266e-05, "loss": 0.0116, "step": 179320 }, { "grad_norm": 0.12149769067764282, "learning_rate": 3.8084567629065126e-05, "loss": 0.0088, "step": 179330 }, { "grad_norm": 0.16816458106040955, "learning_rate": 3.807921492654273e-05, "loss": 0.0116, "step": 179340 }, { "grad_norm": 0.10993128269910812, "learning_rate": 3.8073862368869124e-05, "loss": 0.0152, "step": 179350 }, { "grad_norm": 0.12898226082324982, "learning_rate": 3.806850995610937e-05, "loss": 0.0126, "step": 179360 }, { "grad_norm": 0.2169269174337387, "learning_rate": 3.806315768832847e-05, "loss": 0.0202, "step": 179370 }, { "grad_norm": 0.1109815388917923, "learning_rate": 3.805780556559148e-05, "loss": 0.0171, "step": 179380 }, { "grad_norm": 0.07289204746484756, "learning_rate": 3.805245358796341e-05, "loss": 0.0119, "step": 179390 }, { "grad_norm": 0.07954330742359161, "learning_rate": 3.804710175550934e-05, "loss": 0.0174, "step": 179400 }, { "grad_norm": 0.08094729483127594, "learning_rate": 3.804175006829425e-05, "loss": 0.0111, "step": 179410 }, { "grad_norm": 0.09538698196411133, "learning_rate": 3.8036398526383186e-05, "loss": 0.0156, "step": 179420 }, { "grad_norm": 0.0830344632267952, "learning_rate": 3.80310471298412e-05, "loss": 0.0139, "step": 179430 }, { "grad_norm": 0.1975034773349762, "learning_rate": 3.802569587873326e-05, "loss": 0.0144, "step": 179440 }, { "grad_norm": 0.09937272220849991, "learning_rate": 3.802034477312443e-05, "loss": 0.0149, "step": 179450 }, { "grad_norm": 0.14261573553085327, "learning_rate": 3.801499381307972e-05, "loss": 0.0191, "step": 179460 }, { "grad_norm": 0.15785950422286987, "learning_rate": 3.8009642998664175e-05, "loss": 0.0137, "step": 179470 }, { "grad_norm": 0.1075570210814476, "learning_rate": 3.800429232994276e-05, "loss": 0.014, "step": 179480 }, { "grad_norm": 0.06846819818019867, "learning_rate": 3.799894180698052e-05, "loss": 0.0091, "step": 179490 }, { "grad_norm": 0.18600958585739136, "learning_rate": 3.79935914298425e-05, "loss": 0.0169, "step": 179500 }, { "grad_norm": 0.2056894600391388, "learning_rate": 3.7988241198593646e-05, "loss": 0.0147, "step": 179510 }, { "grad_norm": 0.20047755539417267, "learning_rate": 3.7982891113299e-05, "loss": 0.0173, "step": 179520 }, { "grad_norm": 0.12283997982740402, "learning_rate": 3.797754117402358e-05, "loss": 0.0187, "step": 179530 }, { "grad_norm": 0.12052567303180695, "learning_rate": 3.7972191380832414e-05, "loss": 0.0192, "step": 179540 }, { "grad_norm": 0.05775127187371254, "learning_rate": 3.7966841733790445e-05, "loss": 0.0135, "step": 179550 }, { "grad_norm": 0.06632022559642792, "learning_rate": 3.796149223296272e-05, "loss": 0.0193, "step": 179560 }, { "grad_norm": 0.12470193207263947, "learning_rate": 3.7956142878414246e-05, "loss": 0.0137, "step": 179570 }, { "grad_norm": 0.06373490393161774, "learning_rate": 3.795079367020999e-05, "loss": 0.0118, "step": 179580 }, { "grad_norm": 0.2906757891178131, "learning_rate": 3.794544460841496e-05, "loss": 0.01, "step": 179590 }, { "grad_norm": 0.11843602359294891, "learning_rate": 3.794009569309417e-05, "loss": 0.0133, "step": 179600 }, { "grad_norm": 0.0342605859041214, "learning_rate": 3.7934746924312615e-05, "loss": 0.0128, "step": 179610 }, { "grad_norm": 0.08676106482744217, "learning_rate": 3.792939830213526e-05, "loss": 0.0099, "step": 179620 }, { "grad_norm": 0.10187127441167831, "learning_rate": 3.792404982662711e-05, "loss": 0.0189, "step": 179630 }, { "grad_norm": 0.06879492104053497, "learning_rate": 3.791870149785316e-05, "loss": 0.0155, "step": 179640 }, { "grad_norm": 0.08920270204544067, "learning_rate": 3.7913353315878407e-05, "loss": 0.0133, "step": 179650 }, { "grad_norm": 0.15537117421627045, "learning_rate": 3.790800528076781e-05, "loss": 0.0171, "step": 179660 }, { "grad_norm": 0.1381116658449173, "learning_rate": 3.790265739258636e-05, "loss": 0.0131, "step": 179670 }, { "grad_norm": 0.197050079703331, "learning_rate": 3.789730965139907e-05, "loss": 0.012, "step": 179680 }, { "grad_norm": 0.15804998576641083, "learning_rate": 3.789196205727088e-05, "loss": 0.0141, "step": 179690 }, { "grad_norm": 0.08793164044618607, "learning_rate": 3.788661461026677e-05, "loss": 0.0075, "step": 179700 }, { "grad_norm": 0.052793800830841064, "learning_rate": 3.7881267310451726e-05, "loss": 0.0114, "step": 179710 }, { "grad_norm": 0.08336115628480911, "learning_rate": 3.787592015789076e-05, "loss": 0.0154, "step": 179720 }, { "grad_norm": 0.11517488211393356, "learning_rate": 3.7870573152648784e-05, "loss": 0.014, "step": 179730 }, { "grad_norm": 0.21427907049655914, "learning_rate": 3.78652262947908e-05, "loss": 0.0137, "step": 179740 }, { "grad_norm": 0.07971873134374619, "learning_rate": 3.785987958438179e-05, "loss": 0.0155, "step": 179750 }, { "grad_norm": 0.06461046636104584, "learning_rate": 3.785453302148669e-05, "loss": 0.0079, "step": 179760 }, { "grad_norm": 0.16153034567832947, "learning_rate": 3.7849186606170474e-05, "loss": 0.0113, "step": 179770 }, { "grad_norm": 0.10546264797449112, "learning_rate": 3.784384033849812e-05, "loss": 0.0169, "step": 179780 }, { "grad_norm": 0.12926612794399261, "learning_rate": 3.78384942185346e-05, "loss": 0.0152, "step": 179790 }, { "grad_norm": 0.07105815410614014, "learning_rate": 3.7833148246344835e-05, "loss": 0.0109, "step": 179800 }, { "grad_norm": 0.08608443289995193, "learning_rate": 3.782780242199381e-05, "loss": 0.01, "step": 179810 }, { "grad_norm": 0.21825113892555237, "learning_rate": 3.7822456745546486e-05, "loss": 0.0133, "step": 179820 }, { "grad_norm": 0.06685511022806168, "learning_rate": 3.7817111217067816e-05, "loss": 0.0131, "step": 179830 }, { "grad_norm": 0.06105932593345642, "learning_rate": 3.781176583662272e-05, "loss": 0.0138, "step": 179840 }, { "grad_norm": 0.16588512063026428, "learning_rate": 3.780642060427619e-05, "loss": 0.0094, "step": 179850 }, { "grad_norm": 0.0816139355301857, "learning_rate": 3.7801075520093176e-05, "loss": 0.0121, "step": 179860 }, { "grad_norm": 0.18732425570487976, "learning_rate": 3.77957305841386e-05, "loss": 0.0123, "step": 179870 }, { "grad_norm": 0.06376605480909348, "learning_rate": 3.779038579647742e-05, "loss": 0.011, "step": 179880 }, { "grad_norm": 0.05531824380159378, "learning_rate": 3.7785041157174584e-05, "loss": 0.009, "step": 179890 }, { "grad_norm": 0.10401636362075806, "learning_rate": 3.7779696666295025e-05, "loss": 0.0151, "step": 179900 }, { "grad_norm": 0.058157823979854584, "learning_rate": 3.77743523239037e-05, "loss": 0.0112, "step": 179910 }, { "grad_norm": 0.06082044169306755, "learning_rate": 3.776900813006553e-05, "loss": 0.0122, "step": 179920 }, { "grad_norm": 0.06268928200006485, "learning_rate": 3.776366408484547e-05, "loss": 0.0088, "step": 179930 }, { "grad_norm": 0.06296366453170776, "learning_rate": 3.775832018830843e-05, "loss": 0.0129, "step": 179940 }, { "grad_norm": 0.279958575963974, "learning_rate": 3.775297644051936e-05, "loss": 0.012, "step": 179950 }, { "grad_norm": 0.13411742448806763, "learning_rate": 3.774763284154319e-05, "loss": 0.0108, "step": 179960 }, { "grad_norm": 0.19524580240249634, "learning_rate": 3.7742289391444855e-05, "loss": 0.0128, "step": 179970 }, { "grad_norm": 0.07418278604745865, "learning_rate": 3.773694609028927e-05, "loss": 0.0137, "step": 179980 }, { "grad_norm": 0.0739382952451706, "learning_rate": 3.773160293814138e-05, "loss": 0.016, "step": 179990 }, { "grad_norm": 0.24307212233543396, "learning_rate": 3.772625993506609e-05, "loss": 0.0139, "step": 180000 }, { "grad_norm": 0.08570325374603271, "learning_rate": 3.772091708112832e-05, "loss": 0.0164, "step": 180010 }, { "grad_norm": 0.14248719811439514, "learning_rate": 3.7715574376393006e-05, "loss": 0.0144, "step": 180020 }, { "grad_norm": 0.060111239552497864, "learning_rate": 3.771023182092507e-05, "loss": 0.0125, "step": 180030 }, { "grad_norm": 0.08670012652873993, "learning_rate": 3.770488941478942e-05, "loss": 0.0141, "step": 180040 }, { "grad_norm": 0.12635356187820435, "learning_rate": 3.769954715805096e-05, "loss": 0.0136, "step": 180050 }, { "grad_norm": 0.23120711743831635, "learning_rate": 3.769420505077462e-05, "loss": 0.013, "step": 180060 }, { "grad_norm": 0.1390899121761322, "learning_rate": 3.768886309302533e-05, "loss": 0.0102, "step": 180070 }, { "grad_norm": 0.10634827613830566, "learning_rate": 3.768352128486795e-05, "loss": 0.0155, "step": 180080 }, { "grad_norm": 0.12690120935440063, "learning_rate": 3.7678179626367424e-05, "loss": 0.015, "step": 180090 }, { "grad_norm": 0.09744180738925934, "learning_rate": 3.767283811758864e-05, "loss": 0.0185, "step": 180100 }, { "grad_norm": 0.1225784495472908, "learning_rate": 3.7667496758596535e-05, "loss": 0.013, "step": 180110 }, { "grad_norm": 0.10020848363637924, "learning_rate": 3.766215554945597e-05, "loss": 0.0173, "step": 180120 }, { "grad_norm": 0.09387937933206558, "learning_rate": 3.7656814490231865e-05, "loss": 0.016, "step": 180130 }, { "grad_norm": 0.06873328238725662, "learning_rate": 3.765147358098913e-05, "loss": 0.0107, "step": 180140 }, { "grad_norm": 0.11120818555355072, "learning_rate": 3.764613282179263e-05, "loss": 0.015, "step": 180150 }, { "grad_norm": 0.12123921513557434, "learning_rate": 3.764079221270729e-05, "loss": 0.0111, "step": 180160 }, { "grad_norm": 0.10229731351137161, "learning_rate": 3.763545175379799e-05, "loss": 0.0176, "step": 180170 }, { "grad_norm": 0.1355731338262558, "learning_rate": 3.7630111445129645e-05, "loss": 0.0131, "step": 180180 }, { "grad_norm": 0.17386415600776672, "learning_rate": 3.7624771286767104e-05, "loss": 0.0116, "step": 180190 }, { "grad_norm": 0.08338634669780731, "learning_rate": 3.761943127877527e-05, "loss": 0.0158, "step": 180200 }, { "grad_norm": 0.19151721894741058, "learning_rate": 3.761409142121905e-05, "loss": 0.0138, "step": 180210 }, { "grad_norm": 0.22579866647720337, "learning_rate": 3.7608751714163325e-05, "loss": 0.0102, "step": 180220 }, { "grad_norm": 0.18465088307857513, "learning_rate": 3.7603412157672946e-05, "loss": 0.0106, "step": 180230 }, { "grad_norm": 0.07525420933961868, "learning_rate": 3.7598072751812815e-05, "loss": 0.0133, "step": 180240 }, { "grad_norm": 0.08718311786651611, "learning_rate": 3.759273349664783e-05, "loss": 0.011, "step": 180250 }, { "grad_norm": 0.07235898077487946, "learning_rate": 3.758739439224282e-05, "loss": 0.0108, "step": 180260 }, { "grad_norm": 0.06194305047392845, "learning_rate": 3.7582055438662694e-05, "loss": 0.0138, "step": 180270 }, { "grad_norm": 0.148709237575531, "learning_rate": 3.757671663597231e-05, "loss": 0.0164, "step": 180280 }, { "grad_norm": 0.16948913037776947, "learning_rate": 3.7571377984236585e-05, "loss": 0.0151, "step": 180290 }, { "grad_norm": 0.10498162358999252, "learning_rate": 3.756603948352032e-05, "loss": 0.0152, "step": 180300 }, { "grad_norm": 0.04938796907663345, "learning_rate": 3.756070113388842e-05, "loss": 0.0096, "step": 180310 }, { "grad_norm": 0.1632992923259735, "learning_rate": 3.7555362935405764e-05, "loss": 0.0155, "step": 180320 }, { "grad_norm": 0.1156364157795906, "learning_rate": 3.755002488813718e-05, "loss": 0.012, "step": 180330 }, { "grad_norm": 0.06553276628255844, "learning_rate": 3.754468699214754e-05, "loss": 0.0149, "step": 180340 }, { "grad_norm": 0.12324810773134232, "learning_rate": 3.753934924750172e-05, "loss": 0.0131, "step": 180350 }, { "grad_norm": 0.14569449424743652, "learning_rate": 3.753401165426459e-05, "loss": 0.0159, "step": 180360 }, { "grad_norm": 0.14131642878055573, "learning_rate": 3.752867421250097e-05, "loss": 0.0132, "step": 180370 }, { "grad_norm": 0.07629072666168213, "learning_rate": 3.752333692227573e-05, "loss": 0.014, "step": 180380 }, { "grad_norm": 0.12160839140415192, "learning_rate": 3.7517999783653754e-05, "loss": 0.012, "step": 180390 }, { "grad_norm": 0.3380425274372101, "learning_rate": 3.751266279669984e-05, "loss": 0.0235, "step": 180400 }, { "grad_norm": 0.08893588930368423, "learning_rate": 3.750732596147886e-05, "loss": 0.0156, "step": 180410 }, { "grad_norm": 0.11553272604942322, "learning_rate": 3.750198927805567e-05, "loss": 0.0194, "step": 180420 }, { "grad_norm": 0.1038627102971077, "learning_rate": 3.749665274649513e-05, "loss": 0.0121, "step": 180430 }, { "grad_norm": 0.10275866836309433, "learning_rate": 3.749131636686204e-05, "loss": 0.0134, "step": 180440 }, { "grad_norm": 0.05937758833169937, "learning_rate": 3.748598013922127e-05, "loss": 0.0173, "step": 180450 }, { "grad_norm": 0.25741979479789734, "learning_rate": 3.748064406363767e-05, "loss": 0.0184, "step": 180460 }, { "grad_norm": 0.07145527005195618, "learning_rate": 3.747530814017605e-05, "loss": 0.0116, "step": 180470 }, { "grad_norm": 0.06087591126561165, "learning_rate": 3.746997236890126e-05, "loss": 0.0135, "step": 180480 }, { "grad_norm": 0.0915568545460701, "learning_rate": 3.746463674987814e-05, "loss": 0.0134, "step": 180490 }, { "grad_norm": 0.24188880622386932, "learning_rate": 3.7459301283171536e-05, "loss": 0.0124, "step": 180500 }, { "grad_norm": 0.08623575419187546, "learning_rate": 3.745396596884624e-05, "loss": 0.0108, "step": 180510 }, { "grad_norm": 0.10371819138526917, "learning_rate": 3.744863080696711e-05, "loss": 0.0202, "step": 180520 }, { "grad_norm": 0.10201545804738998, "learning_rate": 3.744329579759895e-05, "loss": 0.0188, "step": 180530 }, { "grad_norm": 0.09290697425603867, "learning_rate": 3.743796094080664e-05, "loss": 0.0068, "step": 180540 }, { "grad_norm": 0.19180817902088165, "learning_rate": 3.7432626236654936e-05, "loss": 0.0146, "step": 180550 }, { "grad_norm": 0.11569438129663467, "learning_rate": 3.7427291685208694e-05, "loss": 0.0133, "step": 180560 }, { "grad_norm": 0.08710773289203644, "learning_rate": 3.742195728653273e-05, "loss": 0.0144, "step": 180570 }, { "grad_norm": 0.09378976374864578, "learning_rate": 3.741662304069187e-05, "loss": 0.0099, "step": 180580 }, { "grad_norm": 0.09772850573062897, "learning_rate": 3.7411288947750903e-05, "loss": 0.0151, "step": 180590 }, { "grad_norm": 0.08511371910572052, "learning_rate": 3.740595500777466e-05, "loss": 0.0172, "step": 180600 }, { "grad_norm": 0.0600549690425396, "learning_rate": 3.7400621220827974e-05, "loss": 0.0178, "step": 180610 }, { "grad_norm": 0.14012515544891357, "learning_rate": 3.7395287586975633e-05, "loss": 0.0152, "step": 180620 }, { "grad_norm": 0.24728479981422424, "learning_rate": 3.7389954106282435e-05, "loss": 0.0228, "step": 180630 }, { "grad_norm": 0.15762126445770264, "learning_rate": 3.7384620778813214e-05, "loss": 0.0101, "step": 180640 }, { "grad_norm": 0.09346745908260345, "learning_rate": 3.737928760463275e-05, "loss": 0.0135, "step": 180650 }, { "grad_norm": 0.1214175820350647, "learning_rate": 3.7373954583805864e-05, "loss": 0.0129, "step": 180660 }, { "grad_norm": 0.13643397390842438, "learning_rate": 3.7368621716397344e-05, "loss": 0.0147, "step": 180670 }, { "grad_norm": 0.07209794223308563, "learning_rate": 3.7363289002472e-05, "loss": 0.0169, "step": 180680 }, { "grad_norm": 0.06833119690418243, "learning_rate": 3.735795644209463e-05, "loss": 0.0134, "step": 180690 }, { "grad_norm": 0.13101205229759216, "learning_rate": 3.735262403533002e-05, "loss": 0.0136, "step": 180700 }, { "grad_norm": 0.14850923418998718, "learning_rate": 3.734729178224298e-05, "loss": 0.0127, "step": 180710 }, { "grad_norm": 0.09299503266811371, "learning_rate": 3.7341959682898274e-05, "loss": 0.0103, "step": 180720 }, { "grad_norm": 0.14510776102542877, "learning_rate": 3.733662773736071e-05, "loss": 0.0116, "step": 180730 }, { "grad_norm": 0.08948090672492981, "learning_rate": 3.7331295945695087e-05, "loss": 0.019, "step": 180740 }, { "grad_norm": 0.07181455940008163, "learning_rate": 3.732596430796618e-05, "loss": 0.0129, "step": 180750 }, { "grad_norm": 0.08436305820941925, "learning_rate": 3.7320632824238766e-05, "loss": 0.0102, "step": 180760 }, { "grad_norm": 0.14491644501686096, "learning_rate": 3.7315301494577626e-05, "loss": 0.0146, "step": 180770 }, { "grad_norm": 0.1075538918375969, "learning_rate": 3.730997031904757e-05, "loss": 0.0151, "step": 180780 }, { "grad_norm": 0.15397875010967255, "learning_rate": 3.730463929771335e-05, "loss": 0.0128, "step": 180790 }, { "grad_norm": 0.10775582492351532, "learning_rate": 3.7299308430639735e-05, "loss": 0.0169, "step": 180800 }, { "grad_norm": 0.07514413446187973, "learning_rate": 3.7293977717891515e-05, "loss": 0.0118, "step": 180810 }, { "grad_norm": 0.1898985356092453, "learning_rate": 3.728864715953349e-05, "loss": 0.0235, "step": 180820 }, { "grad_norm": 0.08814548701047897, "learning_rate": 3.7283316755630374e-05, "loss": 0.0172, "step": 180830 }, { "grad_norm": 0.11165012419223785, "learning_rate": 3.7277986506246974e-05, "loss": 0.015, "step": 180840 }, { "grad_norm": 0.0557774156332016, "learning_rate": 3.727265641144805e-05, "loss": 0.0111, "step": 180850 }, { "grad_norm": 0.10888976603746414, "learning_rate": 3.726732647129838e-05, "loss": 0.0134, "step": 180860 }, { "grad_norm": 0.6815385222434998, "learning_rate": 3.72619966858627e-05, "loss": 0.0142, "step": 180870 }, { "grad_norm": 0.1573089361190796, "learning_rate": 3.725666705520579e-05, "loss": 0.0133, "step": 180880 }, { "grad_norm": 0.07842500507831573, "learning_rate": 3.725133757939242e-05, "loss": 0.0182, "step": 180890 }, { "grad_norm": 0.11616259813308716, "learning_rate": 3.7246008258487314e-05, "loss": 0.0125, "step": 180900 }, { "grad_norm": 0.07621477544307709, "learning_rate": 3.7240679092555256e-05, "loss": 0.0114, "step": 180910 }, { "grad_norm": 0.2176208794116974, "learning_rate": 3.7235350081660985e-05, "loss": 0.0171, "step": 180920 }, { "grad_norm": 0.054980162531137466, "learning_rate": 3.723002122586929e-05, "loss": 0.0077, "step": 180930 }, { "grad_norm": 0.23547276854515076, "learning_rate": 3.722469252524487e-05, "loss": 0.0164, "step": 180940 }, { "grad_norm": 0.07104655355215073, "learning_rate": 3.72193639798525e-05, "loss": 0.0129, "step": 180950 }, { "grad_norm": 0.04838063195347786, "learning_rate": 3.7214035589756944e-05, "loss": 0.0105, "step": 180960 }, { "grad_norm": 0.07669416815042496, "learning_rate": 3.720870735502291e-05, "loss": 0.0129, "step": 180970 }, { "grad_norm": 0.2654251754283905, "learning_rate": 3.720337927571515e-05, "loss": 0.0146, "step": 180980 }, { "grad_norm": 0.11080904304981232, "learning_rate": 3.719805135189842e-05, "loss": 0.016, "step": 180990 }, { "grad_norm": 0.13869324326515198, "learning_rate": 3.719272358363748e-05, "loss": 0.0107, "step": 181000 }, { "grad_norm": 0.07687480747699738, "learning_rate": 3.7187395970997014e-05, "loss": 0.0102, "step": 181010 }, { "grad_norm": 0.16911019384860992, "learning_rate": 3.718206851404178e-05, "loss": 0.016, "step": 181020 }, { "grad_norm": 0.11066202074289322, "learning_rate": 3.717674121283654e-05, "loss": 0.0106, "step": 181030 }, { "grad_norm": 0.12573225796222687, "learning_rate": 3.717141406744599e-05, "loss": 0.009, "step": 181040 }, { "grad_norm": 0.04195665940642357, "learning_rate": 3.716608707793485e-05, "loss": 0.0135, "step": 181050 }, { "grad_norm": 0.21398340165615082, "learning_rate": 3.716076024436789e-05, "loss": 0.0144, "step": 181060 }, { "grad_norm": 0.12516045570373535, "learning_rate": 3.715543356680982e-05, "loss": 0.0163, "step": 181070 }, { "grad_norm": 0.2334388941526413, "learning_rate": 3.715010704532535e-05, "loss": 0.0158, "step": 181080 }, { "grad_norm": 0.0932031199336052, "learning_rate": 3.714478067997921e-05, "loss": 0.0088, "step": 181090 }, { "grad_norm": 0.18102498352527618, "learning_rate": 3.713945447083612e-05, "loss": 0.0184, "step": 181100 }, { "grad_norm": 0.09931889176368713, "learning_rate": 3.7134128417960816e-05, "loss": 0.0137, "step": 181110 }, { "grad_norm": 0.25038859248161316, "learning_rate": 3.712880252141798e-05, "loss": 0.0163, "step": 181120 }, { "grad_norm": 0.08404505997896194, "learning_rate": 3.712347678127235e-05, "loss": 0.015, "step": 181130 }, { "grad_norm": 0.05134015530347824, "learning_rate": 3.7118151197588655e-05, "loss": 0.0115, "step": 181140 }, { "grad_norm": 0.10438355803489685, "learning_rate": 3.7112825770431567e-05, "loss": 0.0155, "step": 181150 }, { "grad_norm": 0.08988446742296219, "learning_rate": 3.710750049986581e-05, "loss": 0.0116, "step": 181160 }, { "grad_norm": 0.13240335881710052, "learning_rate": 3.7102175385956095e-05, "loss": 0.0165, "step": 181170 }, { "grad_norm": 0.07157173752784729, "learning_rate": 3.7096850428767147e-05, "loss": 0.0155, "step": 181180 }, { "grad_norm": 0.2838505804538727, "learning_rate": 3.709152562836363e-05, "loss": 0.0172, "step": 181190 }, { "grad_norm": 0.12559275329113007, "learning_rate": 3.708620098481026e-05, "loss": 0.0097, "step": 181200 }, { "grad_norm": 0.19487226009368896, "learning_rate": 3.708087649817177e-05, "loss": 0.0142, "step": 181210 }, { "grad_norm": 0.08208457380533218, "learning_rate": 3.7075552168512804e-05, "loss": 0.0133, "step": 181220 }, { "grad_norm": 0.16461916267871857, "learning_rate": 3.7070227995898076e-05, "loss": 0.0129, "step": 181230 }, { "grad_norm": 0.0910579040646553, "learning_rate": 3.706490398039229e-05, "loss": 0.012, "step": 181240 }, { "grad_norm": 0.7377015352249146, "learning_rate": 3.705958012206016e-05, "loss": 0.0129, "step": 181250 }, { "grad_norm": 0.08604061603546143, "learning_rate": 3.7054256420966317e-05, "loss": 0.0167, "step": 181260 }, { "grad_norm": 0.11789306998252869, "learning_rate": 3.704893287717548e-05, "loss": 0.0146, "step": 181270 }, { "grad_norm": 0.11880455166101456, "learning_rate": 3.704360949075235e-05, "loss": 0.009, "step": 181280 }, { "grad_norm": 0.039252862334251404, "learning_rate": 3.703828626176161e-05, "loss": 0.0142, "step": 181290 }, { "grad_norm": 0.1330060362815857, "learning_rate": 3.703296319026791e-05, "loss": 0.0204, "step": 181300 }, { "grad_norm": 0.06987357139587402, "learning_rate": 3.702764027633595e-05, "loss": 0.0138, "step": 181310 }, { "grad_norm": 0.2046755850315094, "learning_rate": 3.7022317520030416e-05, "loss": 0.017, "step": 181320 }, { "grad_norm": 0.1325555294752121, "learning_rate": 3.7016994921415984e-05, "loss": 0.0157, "step": 181330 }, { "grad_norm": 0.07056573778390884, "learning_rate": 3.7011672480557306e-05, "loss": 0.0123, "step": 181340 }, { "grad_norm": 0.36331871151924133, "learning_rate": 3.700635019751908e-05, "loss": 0.0133, "step": 181350 }, { "grad_norm": 0.3180936276912689, "learning_rate": 3.700102807236596e-05, "loss": 0.0188, "step": 181360 }, { "grad_norm": 0.049173351377248764, "learning_rate": 3.699570610516263e-05, "loss": 0.0126, "step": 181370 }, { "grad_norm": 0.05552572384476662, "learning_rate": 3.6990384295973736e-05, "loss": 0.0112, "step": 181380 }, { "grad_norm": 0.3569159507751465, "learning_rate": 3.698506264486397e-05, "loss": 0.0162, "step": 181390 }, { "grad_norm": 0.07621746510267258, "learning_rate": 3.697974115189797e-05, "loss": 0.0116, "step": 181400 }, { "grad_norm": 0.07995524257421494, "learning_rate": 3.6974419817140424e-05, "loss": 0.0111, "step": 181410 }, { "grad_norm": 0.052979886531829834, "learning_rate": 3.696909864065596e-05, "loss": 0.0114, "step": 181420 }, { "grad_norm": 0.07412345707416534, "learning_rate": 3.696377762250926e-05, "loss": 0.0125, "step": 181430 }, { "grad_norm": 0.06589456647634506, "learning_rate": 3.695845676276496e-05, "loss": 0.0127, "step": 181440 }, { "grad_norm": 0.06042889878153801, "learning_rate": 3.6953136061487737e-05, "loss": 0.0107, "step": 181450 }, { "grad_norm": 0.08627468347549438, "learning_rate": 3.694781551874222e-05, "loss": 0.0087, "step": 181460 }, { "grad_norm": 0.338621586561203, "learning_rate": 3.694249513459307e-05, "loss": 0.0141, "step": 181470 }, { "grad_norm": 0.08721254765987396, "learning_rate": 3.693717490910493e-05, "loss": 0.0128, "step": 181480 }, { "grad_norm": 0.07236401736736298, "learning_rate": 3.693185484234245e-05, "loss": 0.0154, "step": 181490 }, { "grad_norm": 0.06422148644924164, "learning_rate": 3.692653493437028e-05, "loss": 0.0176, "step": 181500 }, { "grad_norm": 0.07140909880399704, "learning_rate": 3.692121518525305e-05, "loss": 0.0124, "step": 181510 }, { "grad_norm": 0.11546757817268372, "learning_rate": 3.691589559505538e-05, "loss": 0.0156, "step": 181520 }, { "grad_norm": 0.08224333077669144, "learning_rate": 3.6910576163841976e-05, "loss": 0.0105, "step": 181530 }, { "grad_norm": 0.0995955616235733, "learning_rate": 3.69052568916774e-05, "loss": 0.0115, "step": 181540 }, { "grad_norm": 0.1565602421760559, "learning_rate": 3.689993777862631e-05, "loss": 0.018, "step": 181550 }, { "grad_norm": 0.16230256855487823, "learning_rate": 3.689461882475336e-05, "loss": 0.0145, "step": 181560 }, { "grad_norm": 0.08201687783002853, "learning_rate": 3.6889300030123174e-05, "loss": 0.017, "step": 181570 }, { "grad_norm": 0.07942213863134384, "learning_rate": 3.688398139480036e-05, "loss": 0.0142, "step": 181580 }, { "grad_norm": 0.10536503791809082, "learning_rate": 3.687866291884955e-05, "loss": 0.0124, "step": 181590 }, { "grad_norm": 0.0791398286819458, "learning_rate": 3.687334460233539e-05, "loss": 0.0121, "step": 181600 }, { "grad_norm": 0.09115490317344666, "learning_rate": 3.686802644532247e-05, "loss": 0.0113, "step": 181610 }, { "grad_norm": 0.23304873704910278, "learning_rate": 3.6862708447875424e-05, "loss": 0.013, "step": 181620 }, { "grad_norm": 0.09394761174917221, "learning_rate": 3.685739061005888e-05, "loss": 0.0149, "step": 181630 }, { "grad_norm": 0.6852003931999207, "learning_rate": 3.685207293193747e-05, "loss": 0.0172, "step": 181640 }, { "grad_norm": 0.09901674091815948, "learning_rate": 3.684675541357576e-05, "loss": 0.0153, "step": 181650 }, { "grad_norm": 0.23568108677864075, "learning_rate": 3.6841438055038405e-05, "loss": 0.0218, "step": 181660 }, { "grad_norm": 0.1127261370420456, "learning_rate": 3.6836120856390004e-05, "loss": 0.0112, "step": 181670 }, { "grad_norm": 0.21107149124145508, "learning_rate": 3.6830803817695156e-05, "loss": 0.018, "step": 181680 }, { "grad_norm": 0.10520943999290466, "learning_rate": 3.6825486939018466e-05, "loss": 0.0165, "step": 181690 }, { "grad_norm": 0.0940946564078331, "learning_rate": 3.682017022042456e-05, "loss": 0.0146, "step": 181700 }, { "grad_norm": 0.09674453735351562, "learning_rate": 3.681485366197803e-05, "loss": 0.0106, "step": 181710 }, { "grad_norm": 0.1339041143655777, "learning_rate": 3.6809537263743476e-05, "loss": 0.0169, "step": 181720 }, { "grad_norm": 0.09306381642818451, "learning_rate": 3.680422102578549e-05, "loss": 0.0097, "step": 181730 }, { "grad_norm": 0.15426775813102722, "learning_rate": 3.6798904948168676e-05, "loss": 0.0112, "step": 181740 }, { "grad_norm": 0.08361419290304184, "learning_rate": 3.6793589030957654e-05, "loss": 0.0147, "step": 181750 }, { "grad_norm": 0.1308351457118988, "learning_rate": 3.678827327421697e-05, "loss": 0.0131, "step": 181760 }, { "grad_norm": 0.08045092970132828, "learning_rate": 3.678295767801124e-05, "loss": 0.0127, "step": 181770 }, { "grad_norm": 0.06004759296774864, "learning_rate": 3.677764224240509e-05, "loss": 0.0154, "step": 181780 }, { "grad_norm": 0.15195846557617188, "learning_rate": 3.6772326967463026e-05, "loss": 0.0163, "step": 181790 }, { "grad_norm": 0.09813559055328369, "learning_rate": 3.676701185324969e-05, "loss": 0.0137, "step": 181800 }, { "grad_norm": 0.21465715765953064, "learning_rate": 3.6761696899829656e-05, "loss": 0.0112, "step": 181810 }, { "grad_norm": 0.14026421308517456, "learning_rate": 3.675638210726752e-05, "loss": 0.0127, "step": 181820 }, { "grad_norm": 0.10501345247030258, "learning_rate": 3.675106747562783e-05, "loss": 0.0101, "step": 181830 }, { "grad_norm": 0.08571842312812805, "learning_rate": 3.674575300497517e-05, "loss": 0.0121, "step": 181840 }, { "grad_norm": 0.08211398124694824, "learning_rate": 3.674043869537416e-05, "loss": 0.0115, "step": 181850 }, { "grad_norm": 0.0939708799123764, "learning_rate": 3.67351245468893e-05, "loss": 0.0161, "step": 181860 }, { "grad_norm": 0.13138221204280853, "learning_rate": 3.672981055958521e-05, "loss": 0.0152, "step": 181870 }, { "grad_norm": 0.23288396000862122, "learning_rate": 3.672449673352645e-05, "loss": 0.016, "step": 181880 }, { "grad_norm": 0.11111312359571457, "learning_rate": 3.671918306877762e-05, "loss": 0.0146, "step": 181890 }, { "grad_norm": 0.2202318012714386, "learning_rate": 3.6713869565403216e-05, "loss": 0.0145, "step": 181900 }, { "grad_norm": 0.10418011993169785, "learning_rate": 3.6708556223467846e-05, "loss": 0.016, "step": 181910 }, { "grad_norm": 0.13136012852191925, "learning_rate": 3.670324304303609e-05, "loss": 0.0144, "step": 181920 }, { "grad_norm": 0.07880818098783493, "learning_rate": 3.669793002417248e-05, "loss": 0.0075, "step": 181930 }, { "grad_norm": 0.1412672996520996, "learning_rate": 3.669261716694156e-05, "loss": 0.0088, "step": 181940 }, { "grad_norm": 0.2655748426914215, "learning_rate": 3.6687304471407914e-05, "loss": 0.015, "step": 181950 }, { "grad_norm": 0.17183895409107208, "learning_rate": 3.6681991937636105e-05, "loss": 0.0162, "step": 181960 }, { "grad_norm": 0.1353483945131302, "learning_rate": 3.6676679565690645e-05, "loss": 0.014, "step": 181970 }, { "grad_norm": 0.1382344365119934, "learning_rate": 3.667136735563611e-05, "loss": 0.0185, "step": 181980 }, { "grad_norm": 0.12287898361682892, "learning_rate": 3.666605530753705e-05, "loss": 0.0119, "step": 181990 }, { "grad_norm": 0.2282361537218094, "learning_rate": 3.6660743421458025e-05, "loss": 0.0147, "step": 182000 }, { "grad_norm": 0.09654845297336578, "learning_rate": 3.665543169746354e-05, "loss": 0.016, "step": 182010 }, { "grad_norm": 0.10386194288730621, "learning_rate": 3.665012013561816e-05, "loss": 0.0137, "step": 182020 }, { "grad_norm": 0.13704095780849457, "learning_rate": 3.6644808735986434e-05, "loss": 0.0161, "step": 182030 }, { "grad_norm": 0.12988872826099396, "learning_rate": 3.66394974986329e-05, "loss": 0.0096, "step": 182040 }, { "grad_norm": 0.16809509694576263, "learning_rate": 3.6634186423622065e-05, "loss": 0.0139, "step": 182050 }, { "grad_norm": 0.09176413714885712, "learning_rate": 3.662887551101849e-05, "loss": 0.0124, "step": 182060 }, { "grad_norm": 0.07622791826725006, "learning_rate": 3.662356476088671e-05, "loss": 0.0128, "step": 182070 }, { "grad_norm": 0.07987140864133835, "learning_rate": 3.661825417329125e-05, "loss": 0.0118, "step": 182080 }, { "grad_norm": 0.19873134791851044, "learning_rate": 3.6612943748296614e-05, "loss": 0.011, "step": 182090 }, { "grad_norm": 0.13680006563663483, "learning_rate": 3.6607633485967375e-05, "loss": 0.0157, "step": 182100 }, { "grad_norm": 0.07649790495634079, "learning_rate": 3.660232338636802e-05, "loss": 0.0138, "step": 182110 }, { "grad_norm": 0.11648839712142944, "learning_rate": 3.659701344956309e-05, "loss": 0.0152, "step": 182120 }, { "grad_norm": 0.1313691884279251, "learning_rate": 3.659170367561709e-05, "loss": 0.0162, "step": 182130 }, { "grad_norm": 0.06900165230035782, "learning_rate": 3.658639406459457e-05, "loss": 0.0151, "step": 182140 }, { "grad_norm": 0.1603567749261856, "learning_rate": 3.6581084616560016e-05, "loss": 0.0145, "step": 182150 }, { "grad_norm": 0.23264490067958832, "learning_rate": 3.6575775331577954e-05, "loss": 0.0143, "step": 182160 }, { "grad_norm": 0.11276934295892715, "learning_rate": 3.657046620971291e-05, "loss": 0.0314, "step": 182170 }, { "grad_norm": 0.21927042305469513, "learning_rate": 3.656515725102936e-05, "loss": 0.0127, "step": 182180 }, { "grad_norm": 0.10429613292217255, "learning_rate": 3.6559848455591836e-05, "loss": 0.015, "step": 182190 }, { "grad_norm": 0.43639427423477173, "learning_rate": 3.655453982346485e-05, "loss": 0.0158, "step": 182200 }, { "grad_norm": 0.06437492370605469, "learning_rate": 3.654923135471291e-05, "loss": 0.015, "step": 182210 }, { "grad_norm": 0.3968084752559662, "learning_rate": 3.654392304940048e-05, "loss": 0.0156, "step": 182220 }, { "grad_norm": 0.08072572946548462, "learning_rate": 3.653861490759211e-05, "loss": 0.0131, "step": 182230 }, { "grad_norm": 0.5710911154747009, "learning_rate": 3.653330692935229e-05, "loss": 0.0113, "step": 182240 }, { "grad_norm": 0.22530142962932587, "learning_rate": 3.6527999114745484e-05, "loss": 0.0142, "step": 182250 }, { "grad_norm": 0.1272730678319931, "learning_rate": 3.6522691463836215e-05, "loss": 0.0121, "step": 182260 }, { "grad_norm": 0.12414753437042236, "learning_rate": 3.651738397668896e-05, "loss": 0.0126, "step": 182270 }, { "grad_norm": 0.08071642369031906, "learning_rate": 3.651207665336825e-05, "loss": 0.0163, "step": 182280 }, { "grad_norm": 0.10374417901039124, "learning_rate": 3.650676949393851e-05, "loss": 0.0128, "step": 182290 }, { "grad_norm": 0.12336569279432297, "learning_rate": 3.650146249846427e-05, "loss": 0.0151, "step": 182300 }, { "grad_norm": 0.08420661091804504, "learning_rate": 3.649615566701001e-05, "loss": 0.016, "step": 182310 }, { "grad_norm": 0.08022833615541458, "learning_rate": 3.6490848999640215e-05, "loss": 0.0139, "step": 182320 }, { "grad_norm": 0.14464335143566132, "learning_rate": 3.648554249641934e-05, "loss": 0.0168, "step": 182330 }, { "grad_norm": 0.16591139137744904, "learning_rate": 3.6480236157411893e-05, "loss": 0.0125, "step": 182340 }, { "grad_norm": 0.09066733717918396, "learning_rate": 3.647492998268236e-05, "loss": 0.0135, "step": 182350 }, { "grad_norm": 0.0627393051981926, "learning_rate": 3.646962397229518e-05, "loss": 0.0098, "step": 182360 }, { "grad_norm": 0.08236940205097198, "learning_rate": 3.646431812631484e-05, "loss": 0.019, "step": 182370 }, { "grad_norm": 0.10685442388057709, "learning_rate": 3.6459012444805804e-05, "loss": 0.0131, "step": 182380 }, { "grad_norm": 0.09757271409034729, "learning_rate": 3.645370692783259e-05, "loss": 0.0142, "step": 182390 }, { "grad_norm": 0.10276689380407333, "learning_rate": 3.644840157545959e-05, "loss": 0.018, "step": 182400 }, { "grad_norm": 0.08280409872531891, "learning_rate": 3.644309638775132e-05, "loss": 0.0121, "step": 182410 }, { "grad_norm": 0.09163950383663177, "learning_rate": 3.6437791364772246e-05, "loss": 0.0189, "step": 182420 }, { "grad_norm": 0.11285525560379028, "learning_rate": 3.6432486506586786e-05, "loss": 0.0119, "step": 182430 }, { "grad_norm": 0.12798409163951874, "learning_rate": 3.642718181325942e-05, "loss": 0.0135, "step": 182440 }, { "grad_norm": 0.06709589064121246, "learning_rate": 3.642187728485462e-05, "loss": 0.0124, "step": 182450 }, { "grad_norm": 0.18128491938114166, "learning_rate": 3.6416572921436846e-05, "loss": 0.014, "step": 182460 }, { "grad_norm": 0.0970025360584259, "learning_rate": 3.6411268723070516e-05, "loss": 0.0143, "step": 182470 }, { "grad_norm": 0.09496604651212692, "learning_rate": 3.640596468982011e-05, "loss": 0.0116, "step": 182480 }, { "grad_norm": 0.17903399467468262, "learning_rate": 3.6400660821750073e-05, "loss": 0.0158, "step": 182490 }, { "grad_norm": 0.07504930347204208, "learning_rate": 3.639535711892483e-05, "loss": 0.0175, "step": 182500 }, { "grad_norm": 0.07031582295894623, "learning_rate": 3.639005358140886e-05, "loss": 0.0171, "step": 182510 }, { "grad_norm": 0.1139092817902565, "learning_rate": 3.638475020926657e-05, "loss": 0.0153, "step": 182520 }, { "grad_norm": 0.09844208508729935, "learning_rate": 3.6379447002562445e-05, "loss": 0.0121, "step": 182530 }, { "grad_norm": 0.0612870454788208, "learning_rate": 3.637414396136088e-05, "loss": 0.0152, "step": 182540 }, { "grad_norm": 0.07538135349750519, "learning_rate": 3.636884108572633e-05, "loss": 0.0111, "step": 182550 }, { "grad_norm": 0.03661186620593071, "learning_rate": 3.636353837572325e-05, "loss": 0.007, "step": 182560 }, { "grad_norm": 0.04757808893918991, "learning_rate": 3.6358235831416034e-05, "loss": 0.0127, "step": 182570 }, { "grad_norm": 0.08578237891197205, "learning_rate": 3.635293345286913e-05, "loss": 0.0173, "step": 182580 }, { "grad_norm": 0.07673756033182144, "learning_rate": 3.634763124014698e-05, "loss": 0.0144, "step": 182590 }, { "grad_norm": 0.15473517775535583, "learning_rate": 3.634232919331401e-05, "loss": 0.0101, "step": 182600 }, { "grad_norm": 0.17374926805496216, "learning_rate": 3.633702731243462e-05, "loss": 0.0145, "step": 182610 }, { "grad_norm": 0.09507699310779572, "learning_rate": 3.6331725597573245e-05, "loss": 0.0104, "step": 182620 }, { "grad_norm": 0.1603206843137741, "learning_rate": 3.632642404879431e-05, "loss": 0.0152, "step": 182630 }, { "grad_norm": 0.08811911940574646, "learning_rate": 3.6321122666162255e-05, "loss": 0.0132, "step": 182640 }, { "grad_norm": 0.10186687856912613, "learning_rate": 3.6315821449741454e-05, "loss": 0.015, "step": 182650 }, { "grad_norm": 0.18263249099254608, "learning_rate": 3.631052039959634e-05, "loss": 0.0118, "step": 182660 }, { "grad_norm": 0.06912794709205627, "learning_rate": 3.630521951579135e-05, "loss": 0.0111, "step": 182670 }, { "grad_norm": 0.09769349545240402, "learning_rate": 3.629991879839085e-05, "loss": 0.0099, "step": 182680 }, { "grad_norm": 0.10633940994739532, "learning_rate": 3.629461824745928e-05, "loss": 0.0138, "step": 182690 }, { "grad_norm": 0.10621915757656097, "learning_rate": 3.628931786306103e-05, "loss": 0.0164, "step": 182700 }, { "grad_norm": 0.10696237534284592, "learning_rate": 3.628401764526054e-05, "loss": 0.0109, "step": 182710 }, { "grad_norm": 0.1173257976770401, "learning_rate": 3.627871759412216e-05, "loss": 0.0125, "step": 182720 }, { "grad_norm": 0.0631280243396759, "learning_rate": 3.627341770971031e-05, "loss": 0.0137, "step": 182730 }, { "grad_norm": 0.07715447247028351, "learning_rate": 3.626811799208942e-05, "loss": 0.0148, "step": 182740 }, { "grad_norm": 0.13398440182209015, "learning_rate": 3.6262818441323854e-05, "loss": 0.014, "step": 182750 }, { "grad_norm": 0.05344126373529434, "learning_rate": 3.6257519057478004e-05, "loss": 0.0144, "step": 182760 }, { "grad_norm": 0.125515878200531, "learning_rate": 3.625221984061627e-05, "loss": 0.0114, "step": 182770 }, { "grad_norm": 0.04965164139866829, "learning_rate": 3.624692079080306e-05, "loss": 0.0113, "step": 182780 }, { "grad_norm": 0.104847252368927, "learning_rate": 3.624162190810274e-05, "loss": 0.0135, "step": 182790 }, { "grad_norm": 0.07032009214162827, "learning_rate": 3.6236323192579704e-05, "loss": 0.012, "step": 182800 }, { "grad_norm": 0.06505874544382095, "learning_rate": 3.623102464429834e-05, "loss": 0.0116, "step": 182810 }, { "grad_norm": 0.051421791315078735, "learning_rate": 3.622572626332301e-05, "loss": 0.0162, "step": 182820 }, { "grad_norm": 0.0782088041305542, "learning_rate": 3.622042804971814e-05, "loss": 0.0112, "step": 182830 }, { "grad_norm": 0.10591492801904678, "learning_rate": 3.6215130003548045e-05, "loss": 0.0114, "step": 182840 }, { "grad_norm": 0.11709161102771759, "learning_rate": 3.620983212487716e-05, "loss": 0.0141, "step": 182850 }, { "grad_norm": 0.062352459877729416, "learning_rate": 3.620453441376982e-05, "loss": 0.0091, "step": 182860 }, { "grad_norm": 0.11511804908514023, "learning_rate": 3.6199236870290435e-05, "loss": 0.013, "step": 182870 }, { "grad_norm": 0.3281092047691345, "learning_rate": 3.6193939494503336e-05, "loss": 0.0145, "step": 182880 }, { "grad_norm": 0.1269715279340744, "learning_rate": 3.618864228647291e-05, "loss": 0.0165, "step": 182890 }, { "grad_norm": 0.1935773491859436, "learning_rate": 3.618334524626351e-05, "loss": 0.0159, "step": 182900 }, { "grad_norm": 0.06662129610776901, "learning_rate": 3.617804837393952e-05, "loss": 0.0114, "step": 182910 }, { "grad_norm": 0.08555945754051208, "learning_rate": 3.6172751669565294e-05, "loss": 0.0164, "step": 182920 }, { "grad_norm": 0.08741482347249985, "learning_rate": 3.6167455133205186e-05, "loss": 0.0138, "step": 182930 }, { "grad_norm": 0.11059696972370148, "learning_rate": 3.616215876492356e-05, "loss": 0.0107, "step": 182940 }, { "grad_norm": 0.04834273084998131, "learning_rate": 3.615686256478477e-05, "loss": 0.0086, "step": 182950 }, { "grad_norm": 0.0694572925567627, "learning_rate": 3.6151566532853175e-05, "loss": 0.0126, "step": 182960 }, { "grad_norm": 0.07481694966554642, "learning_rate": 3.6146270669193116e-05, "loss": 0.0162, "step": 182970 }, { "grad_norm": 0.14682495594024658, "learning_rate": 3.614097497386894e-05, "loss": 0.0175, "step": 182980 }, { "grad_norm": 0.24691441655158997, "learning_rate": 3.613567944694503e-05, "loss": 0.0123, "step": 182990 }, { "grad_norm": 0.08783343434333801, "learning_rate": 3.613038408848568e-05, "loss": 0.0157, "step": 183000 }, { "grad_norm": 0.06583744287490845, "learning_rate": 3.612508889855526e-05, "loss": 0.0096, "step": 183010 }, { "grad_norm": 0.1039351150393486, "learning_rate": 3.611979387721811e-05, "loss": 0.0186, "step": 183020 }, { "grad_norm": 0.11201706528663635, "learning_rate": 3.611449902453859e-05, "loss": 0.0156, "step": 183030 }, { "grad_norm": 0.07965945452451706, "learning_rate": 3.610920434058099e-05, "loss": 0.0154, "step": 183040 }, { "grad_norm": 0.14730225503444672, "learning_rate": 3.610390982540969e-05, "loss": 0.0111, "step": 183050 }, { "grad_norm": 0.08623678237199783, "learning_rate": 3.609861547908901e-05, "loss": 0.0146, "step": 183060 }, { "grad_norm": 0.08425083756446838, "learning_rate": 3.6093321301683266e-05, "loss": 0.0123, "step": 183070 }, { "grad_norm": 0.10139855742454529, "learning_rate": 3.608802729325679e-05, "loss": 0.0121, "step": 183080 }, { "grad_norm": 0.09455994516611099, "learning_rate": 3.608273345387392e-05, "loss": 0.0154, "step": 183090 }, { "grad_norm": 0.07241128385066986, "learning_rate": 3.6077439783599003e-05, "loss": 0.015, "step": 183100 }, { "grad_norm": 0.08788103610277176, "learning_rate": 3.607214628249631e-05, "loss": 0.0134, "step": 183110 }, { "grad_norm": 0.15972043573856354, "learning_rate": 3.60668529506302e-05, "loss": 0.0121, "step": 183120 }, { "grad_norm": 0.07271479070186615, "learning_rate": 3.6061559788064995e-05, "loss": 0.0161, "step": 183130 }, { "grad_norm": 0.2766696810722351, "learning_rate": 3.605626679486498e-05, "loss": 0.0131, "step": 183140 }, { "grad_norm": 0.19433344900608063, "learning_rate": 3.605097397109449e-05, "loss": 0.0139, "step": 183150 }, { "grad_norm": 0.08348120003938675, "learning_rate": 3.6045681316817835e-05, "loss": 0.0109, "step": 183160 }, { "grad_norm": 0.1333535760641098, "learning_rate": 3.6040388832099345e-05, "loss": 0.0131, "step": 183170 }, { "grad_norm": 0.07984792441129684, "learning_rate": 3.6035096517003296e-05, "loss": 0.0165, "step": 183180 }, { "grad_norm": 0.11097489297389984, "learning_rate": 3.6029804371594e-05, "loss": 0.0153, "step": 183190 }, { "grad_norm": 0.11294487118721008, "learning_rate": 3.6024512395935787e-05, "loss": 0.0125, "step": 183200 }, { "grad_norm": 0.06043127551674843, "learning_rate": 3.6019220590092954e-05, "loss": 0.0144, "step": 183210 }, { "grad_norm": 0.11871388554573059, "learning_rate": 3.601392895412976e-05, "loss": 0.018, "step": 183220 }, { "grad_norm": 0.09252213686704636, "learning_rate": 3.600863748811055e-05, "loss": 0.0176, "step": 183230 }, { "grad_norm": 0.07467619329690933, "learning_rate": 3.600334619209962e-05, "loss": 0.0143, "step": 183240 }, { "grad_norm": 0.4519101083278656, "learning_rate": 3.599805506616122e-05, "loss": 0.0117, "step": 183250 }, { "grad_norm": 0.1610850840806961, "learning_rate": 3.5992764110359686e-05, "loss": 0.0169, "step": 183260 }, { "grad_norm": 0.07967840135097504, "learning_rate": 3.598747332475928e-05, "loss": 0.0142, "step": 183270 }, { "grad_norm": 0.14235316216945648, "learning_rate": 3.598218270942433e-05, "loss": 0.0124, "step": 183280 }, { "grad_norm": 0.0797697901725769, "learning_rate": 3.597689226441908e-05, "loss": 0.0149, "step": 183290 }, { "grad_norm": 0.13428068161010742, "learning_rate": 3.597160198980782e-05, "loss": 0.0099, "step": 183300 }, { "grad_norm": 0.08134565502405167, "learning_rate": 3.596631188565486e-05, "loss": 0.0113, "step": 183310 }, { "grad_norm": 0.07956650108098984, "learning_rate": 3.596102195202445e-05, "loss": 0.0125, "step": 183320 }, { "grad_norm": 0.06238391622900963, "learning_rate": 3.595573218898087e-05, "loss": 0.0109, "step": 183330 }, { "grad_norm": 0.09957478195428848, "learning_rate": 3.595044259658841e-05, "loss": 0.0149, "step": 183340 }, { "grad_norm": 0.09371329843997955, "learning_rate": 3.594515317491135e-05, "loss": 0.0111, "step": 183350 }, { "grad_norm": 0.06386794149875641, "learning_rate": 3.5939863924013934e-05, "loss": 0.0111, "step": 183360 }, { "grad_norm": 0.13892054557800293, "learning_rate": 3.593457484396044e-05, "loss": 0.0102, "step": 183370 }, { "grad_norm": 0.08583791553974152, "learning_rate": 3.5929285934815174e-05, "loss": 0.0111, "step": 183380 }, { "grad_norm": 0.2382541447877884, "learning_rate": 3.592399719664234e-05, "loss": 0.0189, "step": 183390 }, { "grad_norm": 0.09031680971384048, "learning_rate": 3.5918708629506234e-05, "loss": 0.0162, "step": 183400 }, { "grad_norm": 0.1961221843957901, "learning_rate": 3.591342023347111e-05, "loss": 0.0297, "step": 183410 }, { "grad_norm": 0.0770321860909462, "learning_rate": 3.590813200860125e-05, "loss": 0.0156, "step": 183420 }, { "grad_norm": 0.07266402244567871, "learning_rate": 3.590284395496087e-05, "loss": 0.0136, "step": 183430 }, { "grad_norm": 0.1189754456281662, "learning_rate": 3.5897556072614244e-05, "loss": 0.0132, "step": 183440 }, { "grad_norm": 0.1959686428308487, "learning_rate": 3.589226836162565e-05, "loss": 0.0158, "step": 183450 }, { "grad_norm": 0.08855043351650238, "learning_rate": 3.58869808220593e-05, "loss": 0.0155, "step": 183460 }, { "grad_norm": 0.22212016582489014, "learning_rate": 3.588169345397945e-05, "loss": 0.013, "step": 183470 }, { "grad_norm": 0.06308304518461227, "learning_rate": 3.587640625745036e-05, "loss": 0.0104, "step": 183480 }, { "grad_norm": 0.11672741174697876, "learning_rate": 3.5871119232536275e-05, "loss": 0.0156, "step": 183490 }, { "grad_norm": 0.09940445423126221, "learning_rate": 3.586583237930144e-05, "loss": 0.0149, "step": 183500 }, { "grad_norm": 0.08984662592411041, "learning_rate": 3.5860545697810065e-05, "loss": 0.0118, "step": 183510 }, { "grad_norm": 0.07834333181381226, "learning_rate": 3.5855259188126414e-05, "loss": 0.0095, "step": 183520 }, { "grad_norm": 0.15551182627677917, "learning_rate": 3.5849972850314726e-05, "loss": 0.0173, "step": 183530 }, { "grad_norm": 0.08182331919670105, "learning_rate": 3.5844686684439234e-05, "loss": 0.0091, "step": 183540 }, { "grad_norm": 0.11203526705503464, "learning_rate": 3.583940069056415e-05, "loss": 0.0144, "step": 183550 }, { "grad_norm": 0.08155277371406555, "learning_rate": 3.583411486875373e-05, "loss": 0.011, "step": 183560 }, { "grad_norm": 0.07147955894470215, "learning_rate": 3.582882921907218e-05, "loss": 0.0141, "step": 183570 }, { "grad_norm": 0.2207721620798111, "learning_rate": 3.582354374158374e-05, "loss": 0.01, "step": 183580 }, { "grad_norm": 0.05991141498088837, "learning_rate": 3.581825843635262e-05, "loss": 0.0129, "step": 183590 }, { "grad_norm": 0.0681571289896965, "learning_rate": 3.581297330344306e-05, "loss": 0.0182, "step": 183600 }, { "grad_norm": 0.09033667296171188, "learning_rate": 3.5807688342919256e-05, "loss": 0.0152, "step": 183610 }, { "grad_norm": 0.1248108297586441, "learning_rate": 3.580240355484545e-05, "loss": 0.0122, "step": 183620 }, { "grad_norm": 0.06222086399793625, "learning_rate": 3.579711893928585e-05, "loss": 0.0131, "step": 183630 }, { "grad_norm": 0.4366881549358368, "learning_rate": 3.5791834496304656e-05, "loss": 0.0149, "step": 183640 }, { "grad_norm": 0.10514994710683823, "learning_rate": 3.578655022596607e-05, "loss": 0.0124, "step": 183650 }, { "grad_norm": 0.0712532103061676, "learning_rate": 3.578126612833435e-05, "loss": 0.0158, "step": 183660 }, { "grad_norm": 0.0805274173617363, "learning_rate": 3.577598220347366e-05, "loss": 0.0119, "step": 183670 }, { "grad_norm": 0.0727730393409729, "learning_rate": 3.577069845144821e-05, "loss": 0.0097, "step": 183680 }, { "grad_norm": 0.1259654313325882, "learning_rate": 3.5765414872322213e-05, "loss": 0.0149, "step": 183690 }, { "grad_norm": 0.06506020575761795, "learning_rate": 3.5760131466159875e-05, "loss": 0.0147, "step": 183700 }, { "grad_norm": 0.07055844366550446, "learning_rate": 3.575484823302537e-05, "loss": 0.0136, "step": 183710 }, { "grad_norm": 0.18799802660942078, "learning_rate": 3.57495651729829e-05, "loss": 0.0087, "step": 183720 }, { "grad_norm": 0.11379387229681015, "learning_rate": 3.574428228609668e-05, "loss": 0.014, "step": 183730 }, { "grad_norm": 0.13343502581119537, "learning_rate": 3.573899957243091e-05, "loss": 0.0189, "step": 183740 }, { "grad_norm": 0.09561067074537277, "learning_rate": 3.573371703204973e-05, "loss": 0.0109, "step": 183750 }, { "grad_norm": 0.11153502017259598, "learning_rate": 3.5728434665017366e-05, "loss": 0.0165, "step": 183760 }, { "grad_norm": 0.06715720146894455, "learning_rate": 3.5723152471398005e-05, "loss": 0.0113, "step": 183770 }, { "grad_norm": 0.2894130349159241, "learning_rate": 3.571787045125581e-05, "loss": 0.0154, "step": 183780 }, { "grad_norm": 0.08510484546422958, "learning_rate": 3.571258860465497e-05, "loss": 0.0112, "step": 183790 }, { "grad_norm": 0.2243991643190384, "learning_rate": 3.570730693165967e-05, "loss": 0.0125, "step": 183800 }, { "grad_norm": 0.21204988658428192, "learning_rate": 3.570202543233411e-05, "loss": 0.0163, "step": 183810 }, { "grad_norm": 0.07846271246671677, "learning_rate": 3.569674410674241e-05, "loss": 0.0175, "step": 183820 }, { "grad_norm": 0.1003948524594307, "learning_rate": 3.569146295494878e-05, "loss": 0.0132, "step": 183830 }, { "grad_norm": 0.2130974382162094, "learning_rate": 3.568618197701739e-05, "loss": 0.0175, "step": 183840 }, { "grad_norm": 0.24523155391216278, "learning_rate": 3.568090117301241e-05, "loss": 0.0128, "step": 183850 }, { "grad_norm": 0.06702958792448044, "learning_rate": 3.567562054299799e-05, "loss": 0.0133, "step": 183860 }, { "grad_norm": 0.06334616243839264, "learning_rate": 3.56703400870383e-05, "loss": 0.0111, "step": 183870 }, { "grad_norm": 0.07481563836336136, "learning_rate": 3.566505980519753e-05, "loss": 0.0107, "step": 183880 }, { "grad_norm": 0.14747408032417297, "learning_rate": 3.5659779697539794e-05, "loss": 0.0128, "step": 183890 }, { "grad_norm": 0.08073624968528748, "learning_rate": 3.565449976412928e-05, "loss": 0.0102, "step": 183900 }, { "grad_norm": 0.10353532433509827, "learning_rate": 3.5649220005030134e-05, "loss": 0.0172, "step": 183910 }, { "grad_norm": 0.12058411538600922, "learning_rate": 3.564394042030654e-05, "loss": 0.0135, "step": 183920 }, { "grad_norm": 0.2179989516735077, "learning_rate": 3.5638661010022604e-05, "loss": 0.0176, "step": 183930 }, { "grad_norm": 0.10025092959403992, "learning_rate": 3.5633381774242485e-05, "loss": 0.0154, "step": 183940 }, { "grad_norm": 0.06146501749753952, "learning_rate": 3.562810271303037e-05, "loss": 0.0134, "step": 183950 }, { "grad_norm": 0.05829741433262825, "learning_rate": 3.5622823826450355e-05, "loss": 0.0119, "step": 183960 }, { "grad_norm": 0.05916905775666237, "learning_rate": 3.5617545114566606e-05, "loss": 0.0104, "step": 183970 }, { "grad_norm": 0.06480561196804047, "learning_rate": 3.5612266577443256e-05, "loss": 0.0139, "step": 183980 }, { "grad_norm": 0.1661282479763031, "learning_rate": 3.5606988215144485e-05, "loss": 0.0113, "step": 183990 }, { "grad_norm": 0.07029557228088379, "learning_rate": 3.560171002773437e-05, "loss": 0.0129, "step": 184000 }, { "grad_norm": 0.12837430834770203, "learning_rate": 3.559643201527707e-05, "loss": 0.0147, "step": 184010 }, { "grad_norm": 0.06554895639419556, "learning_rate": 3.5591154177836736e-05, "loss": 0.0125, "step": 184020 }, { "grad_norm": 0.1023612767457962, "learning_rate": 3.5585876515477474e-05, "loss": 0.0115, "step": 184030 }, { "grad_norm": 0.11598971486091614, "learning_rate": 3.5580599028263416e-05, "loss": 0.0159, "step": 184040 }, { "grad_norm": 0.07734265178442001, "learning_rate": 3.557532171625869e-05, "loss": 0.0114, "step": 184050 }, { "grad_norm": 0.055797476321458817, "learning_rate": 3.557004457952745e-05, "loss": 0.012, "step": 184060 }, { "grad_norm": 0.15469013154506683, "learning_rate": 3.556476761813377e-05, "loss": 0.0204, "step": 184070 }, { "grad_norm": 0.1972174048423767, "learning_rate": 3.555949083214179e-05, "loss": 0.0178, "step": 184080 }, { "grad_norm": 0.09715595096349716, "learning_rate": 3.555421422161566e-05, "loss": 0.0134, "step": 184090 }, { "grad_norm": 0.06689880043268204, "learning_rate": 3.554893778661944e-05, "loss": 0.0201, "step": 184100 }, { "grad_norm": 0.1885470747947693, "learning_rate": 3.554366152721727e-05, "loss": 0.0118, "step": 184110 }, { "grad_norm": 0.13135789334774017, "learning_rate": 3.553838544347326e-05, "loss": 0.0154, "step": 184120 }, { "grad_norm": 0.29099905490875244, "learning_rate": 3.553310953545155e-05, "loss": 0.0146, "step": 184130 }, { "grad_norm": 0.07100645452737808, "learning_rate": 3.552783380321618e-05, "loss": 0.0136, "step": 184140 }, { "grad_norm": 0.18801085650920868, "learning_rate": 3.5522558246831306e-05, "loss": 0.0159, "step": 184150 }, { "grad_norm": 0.09369699656963348, "learning_rate": 3.5517282866361016e-05, "loss": 0.0169, "step": 184160 }, { "grad_norm": 0.06840737909078598, "learning_rate": 3.551200766186944e-05, "loss": 0.0155, "step": 184170 }, { "grad_norm": 0.10296505689620972, "learning_rate": 3.550673263342062e-05, "loss": 0.0114, "step": 184180 }, { "grad_norm": 0.14692875742912292, "learning_rate": 3.550145778107869e-05, "loss": 0.0143, "step": 184190 }, { "grad_norm": 0.06103593483567238, "learning_rate": 3.5496183104907735e-05, "loss": 0.0146, "step": 184200 }, { "grad_norm": 0.1983984410762787, "learning_rate": 3.549090860497186e-05, "loss": 0.0188, "step": 184210 }, { "grad_norm": 0.05506482720375061, "learning_rate": 3.548563428133513e-05, "loss": 0.0133, "step": 184220 }, { "grad_norm": 0.08788188546895981, "learning_rate": 3.548036013406165e-05, "loss": 0.0176, "step": 184230 }, { "grad_norm": 0.1370910406112671, "learning_rate": 3.547508616321551e-05, "loss": 0.0144, "step": 184240 }, { "grad_norm": 0.06372177600860596, "learning_rate": 3.546981236886079e-05, "loss": 0.0129, "step": 184250 }, { "grad_norm": 0.11910109221935272, "learning_rate": 3.5464538751061556e-05, "loss": 0.0141, "step": 184260 }, { "grad_norm": 0.10823671519756317, "learning_rate": 3.545926530988191e-05, "loss": 0.0127, "step": 184270 }, { "grad_norm": 0.135049507021904, "learning_rate": 3.545399204538591e-05, "loss": 0.0178, "step": 184280 }, { "grad_norm": 0.43659937381744385, "learning_rate": 3.544871895763765e-05, "loss": 0.0156, "step": 184290 }, { "grad_norm": 0.1093502938747406, "learning_rate": 3.544344604670118e-05, "loss": 0.0146, "step": 184300 }, { "grad_norm": 0.08433324098587036, "learning_rate": 3.54381733126406e-05, "loss": 0.0127, "step": 184310 }, { "grad_norm": 0.06459072977304459, "learning_rate": 3.5432900755519946e-05, "loss": 0.0125, "step": 184320 }, { "grad_norm": 0.13805820047855377, "learning_rate": 3.5427628375403324e-05, "loss": 0.0098, "step": 184330 }, { "grad_norm": 0.07704328000545502, "learning_rate": 3.542235617235477e-05, "loss": 0.0145, "step": 184340 }, { "grad_norm": 0.24708126485347748, "learning_rate": 3.541708414643834e-05, "loss": 0.0127, "step": 184350 }, { "grad_norm": 0.0807284563779831, "learning_rate": 3.5411812297718106e-05, "loss": 0.0135, "step": 184360 }, { "grad_norm": 0.08776576817035675, "learning_rate": 3.5406540626258136e-05, "loss": 0.0124, "step": 184370 }, { "grad_norm": 0.08300846070051193, "learning_rate": 3.5401269132122474e-05, "loss": 0.0158, "step": 184380 }, { "grad_norm": 0.0959397405385971, "learning_rate": 3.539599781537517e-05, "loss": 0.0152, "step": 184390 }, { "grad_norm": 0.11495845019817352, "learning_rate": 3.5390726676080276e-05, "loss": 0.015, "step": 184400 }, { "grad_norm": 0.13347730040550232, "learning_rate": 3.538545571430186e-05, "loss": 0.0162, "step": 184410 }, { "grad_norm": 0.12744610011577606, "learning_rate": 3.5380184930103955e-05, "loss": 0.0164, "step": 184420 }, { "grad_norm": 0.14077846705913544, "learning_rate": 3.53749143235506e-05, "loss": 0.013, "step": 184430 }, { "grad_norm": 0.06667326390743256, "learning_rate": 3.536964389470584e-05, "loss": 0.0141, "step": 184440 }, { "grad_norm": 0.10419896990060806, "learning_rate": 3.536437364363374e-05, "loss": 0.0125, "step": 184450 }, { "grad_norm": 0.055984016507864, "learning_rate": 3.53591035703983e-05, "loss": 0.0116, "step": 184460 }, { "grad_norm": 0.11605052649974823, "learning_rate": 3.535383367506357e-05, "loss": 0.0141, "step": 184470 }, { "grad_norm": 0.08627112954854965, "learning_rate": 3.5348563957693595e-05, "loss": 0.013, "step": 184480 }, { "grad_norm": 0.10474426299333572, "learning_rate": 3.534329441835242e-05, "loss": 0.0169, "step": 184490 }, { "grad_norm": 0.06989360600709915, "learning_rate": 3.533802505710403e-05, "loss": 0.0179, "step": 184500 }, { "grad_norm": 0.06425723433494568, "learning_rate": 3.533275587401249e-05, "loss": 0.0193, "step": 184510 }, { "grad_norm": 0.06415529549121857, "learning_rate": 3.532748686914183e-05, "loss": 0.0117, "step": 184520 }, { "grad_norm": 0.21772824227809906, "learning_rate": 3.5322218042556034e-05, "loss": 0.0153, "step": 184530 }, { "grad_norm": 0.10544084012508392, "learning_rate": 3.531694939431914e-05, "loss": 0.0163, "step": 184540 }, { "grad_norm": 0.09120295196771622, "learning_rate": 3.531168092449519e-05, "loss": 0.0117, "step": 184550 }, { "grad_norm": 0.11781467497348785, "learning_rate": 3.53064126331482e-05, "loss": 0.012, "step": 184560 }, { "grad_norm": 0.04190269857645035, "learning_rate": 3.5301144520342145e-05, "loss": 0.0109, "step": 184570 }, { "grad_norm": 0.31318989396095276, "learning_rate": 3.529587658614106e-05, "loss": 0.0107, "step": 184580 }, { "grad_norm": 0.7920392751693726, "learning_rate": 3.5290608830608984e-05, "loss": 0.0238, "step": 184590 }, { "grad_norm": 0.30970367789268494, "learning_rate": 3.5285341253809877e-05, "loss": 0.0121, "step": 184600 }, { "grad_norm": 0.22825728356838226, "learning_rate": 3.528007385580777e-05, "loss": 0.0102, "step": 184610 }, { "grad_norm": 0.07193978130817413, "learning_rate": 3.5274806636666656e-05, "loss": 0.0163, "step": 184620 }, { "grad_norm": 0.19773299992084503, "learning_rate": 3.5269539596450574e-05, "loss": 0.0144, "step": 184630 }, { "grad_norm": 0.1229340136051178, "learning_rate": 3.5264272735223464e-05, "loss": 0.0136, "step": 184640 }, { "grad_norm": 0.194325253367424, "learning_rate": 3.525900605304936e-05, "loss": 0.0174, "step": 184650 }, { "grad_norm": 0.1884865015745163, "learning_rate": 3.5253739549992266e-05, "loss": 0.0117, "step": 184660 }, { "grad_norm": 0.07580292224884033, "learning_rate": 3.524847322611614e-05, "loss": 0.0121, "step": 184670 }, { "grad_norm": 0.11098426580429077, "learning_rate": 3.524320708148499e-05, "loss": 0.0142, "step": 184680 }, { "grad_norm": 0.09828941524028778, "learning_rate": 3.523794111616281e-05, "loss": 0.0103, "step": 184690 }, { "grad_norm": 0.07340487092733383, "learning_rate": 3.52326753302136e-05, "loss": 0.0105, "step": 184700 }, { "grad_norm": 0.17381322383880615, "learning_rate": 3.522740972370131e-05, "loss": 0.0122, "step": 184710 }, { "grad_norm": 0.07249685376882553, "learning_rate": 3.522214429668994e-05, "loss": 0.0149, "step": 184720 }, { "grad_norm": 0.0689517930150032, "learning_rate": 3.521687904924347e-05, "loss": 0.0098, "step": 184730 }, { "grad_norm": 0.071865014731884, "learning_rate": 3.5211613981425894e-05, "loss": 0.0121, "step": 184740 }, { "grad_norm": 0.10065040737390518, "learning_rate": 3.5206349093301156e-05, "loss": 0.0126, "step": 184750 }, { "grad_norm": 0.05344584211707115, "learning_rate": 3.5201084384933236e-05, "loss": 0.0114, "step": 184760 }, { "grad_norm": 0.06083204597234726, "learning_rate": 3.519581985638614e-05, "loss": 0.0115, "step": 184770 }, { "grad_norm": 0.09579893201589584, "learning_rate": 3.5190555507723785e-05, "loss": 0.0131, "step": 184780 }, { "grad_norm": 0.1200236976146698, "learning_rate": 3.518529133901017e-05, "loss": 0.0153, "step": 184790 }, { "grad_norm": 0.21335020661354065, "learning_rate": 3.5180027350309245e-05, "loss": 0.017, "step": 184800 }, { "grad_norm": 0.16031020879745483, "learning_rate": 3.517476354168501e-05, "loss": 0.0133, "step": 184810 }, { "grad_norm": 0.14859411120414734, "learning_rate": 3.5169499913201365e-05, "loss": 0.0127, "step": 184820 }, { "grad_norm": 0.0879666730761528, "learning_rate": 3.516423646492231e-05, "loss": 0.0218, "step": 184830 }, { "grad_norm": 0.06195935979485512, "learning_rate": 3.5158973196911795e-05, "loss": 0.0154, "step": 184840 }, { "grad_norm": 0.07455185800790787, "learning_rate": 3.515371010923375e-05, "loss": 0.0118, "step": 184850 }, { "grad_norm": 0.09911854565143585, "learning_rate": 3.514844720195215e-05, "loss": 0.0099, "step": 184860 }, { "grad_norm": 0.19517865777015686, "learning_rate": 3.5143184475130943e-05, "loss": 0.0164, "step": 184870 }, { "grad_norm": 0.060625821352005005, "learning_rate": 3.513792192883408e-05, "loss": 0.011, "step": 184880 }, { "grad_norm": 0.15089404582977295, "learning_rate": 3.513265956312549e-05, "loss": 0.0094, "step": 184890 }, { "grad_norm": 0.28840431571006775, "learning_rate": 3.5127397378069114e-05, "loss": 0.0122, "step": 184900 }, { "grad_norm": 0.060373835265636444, "learning_rate": 3.5122135373728917e-05, "loss": 0.0117, "step": 184910 }, { "grad_norm": 0.0497889369726181, "learning_rate": 3.511687355016882e-05, "loss": 0.0148, "step": 184920 }, { "grad_norm": 0.083066426217556, "learning_rate": 3.511161190745275e-05, "loss": 0.014, "step": 184930 }, { "grad_norm": 0.1073855459690094, "learning_rate": 3.510635044564466e-05, "loss": 0.0123, "step": 184940 }, { "grad_norm": 0.08341794461011887, "learning_rate": 3.510108916480847e-05, "loss": 0.0171, "step": 184950 }, { "grad_norm": 0.14115722477436066, "learning_rate": 3.509582806500813e-05, "loss": 0.0198, "step": 184960 }, { "grad_norm": 0.06506645679473877, "learning_rate": 3.509056714630753e-05, "loss": 0.0273, "step": 184970 }, { "grad_norm": 0.3202148377895355, "learning_rate": 3.508530640877063e-05, "loss": 0.0121, "step": 184980 }, { "grad_norm": 0.151128351688385, "learning_rate": 3.508004585246133e-05, "loss": 0.0103, "step": 184990 }, { "grad_norm": 0.14985987544059753, "learning_rate": 3.507478547744356e-05, "loss": 0.0129, "step": 185000 }, { "grad_norm": 0.09855193644762039, "learning_rate": 3.5069525283781233e-05, "loss": 0.0109, "step": 185010 }, { "grad_norm": 0.2053036242723465, "learning_rate": 3.5064265271538285e-05, "loss": 0.0086, "step": 185020 }, { "grad_norm": 0.06521136313676834, "learning_rate": 3.505900544077861e-05, "loss": 0.0098, "step": 185030 }, { "grad_norm": 0.0810810849070549, "learning_rate": 3.505374579156613e-05, "loss": 0.0136, "step": 185040 }, { "grad_norm": 0.14455881714820862, "learning_rate": 3.5048486323964735e-05, "loss": 0.013, "step": 185050 }, { "grad_norm": 0.07812394946813583, "learning_rate": 3.504322703803836e-05, "loss": 0.0133, "step": 185060 }, { "grad_norm": 0.052216652780771255, "learning_rate": 3.503796793385089e-05, "loss": 0.0129, "step": 185070 }, { "grad_norm": 0.10572987794876099, "learning_rate": 3.503270901146625e-05, "loss": 0.0152, "step": 185080 }, { "grad_norm": 0.09957516193389893, "learning_rate": 3.502745027094832e-05, "loss": 0.0205, "step": 185090 }, { "grad_norm": 0.09142632782459259, "learning_rate": 3.5022191712361004e-05, "loss": 0.0125, "step": 185100 }, { "grad_norm": 0.06998409330844879, "learning_rate": 3.5016933335768194e-05, "loss": 0.0107, "step": 185110 }, { "grad_norm": 0.22650521993637085, "learning_rate": 3.50116751412338e-05, "loss": 0.0147, "step": 185120 }, { "grad_norm": 0.06721282750368118, "learning_rate": 3.500641712882171e-05, "loss": 0.0168, "step": 185130 }, { "grad_norm": 0.1701705902814865, "learning_rate": 3.5001159298595784e-05, "loss": 0.0131, "step": 185140 }, { "grad_norm": 0.1370822638273239, "learning_rate": 3.499590165061994e-05, "loss": 0.0161, "step": 185150 }, { "grad_norm": 0.14883409440517426, "learning_rate": 3.499064418495809e-05, "loss": 0.0151, "step": 185160 }, { "grad_norm": 0.1380893588066101, "learning_rate": 3.4985386901674054e-05, "loss": 0.015, "step": 185170 }, { "grad_norm": 0.10019829124212265, "learning_rate": 3.4980129800831745e-05, "loss": 0.0134, "step": 185180 }, { "grad_norm": 0.07479246705770493, "learning_rate": 3.497487288249504e-05, "loss": 0.0153, "step": 185190 }, { "grad_norm": 0.20786291360855103, "learning_rate": 3.496961614672783e-05, "loss": 0.0205, "step": 185200 }, { "grad_norm": 0.25680288672447205, "learning_rate": 3.496435959359396e-05, "loss": 0.0145, "step": 185210 }, { "grad_norm": 0.10978354513645172, "learning_rate": 3.495910322315732e-05, "loss": 0.01, "step": 185220 }, { "grad_norm": 0.11963348090648651, "learning_rate": 3.495384703548179e-05, "loss": 0.011, "step": 185230 }, { "grad_norm": 0.09331522136926651, "learning_rate": 3.494859103063121e-05, "loss": 0.0109, "step": 185240 }, { "grad_norm": 0.05995829775929451, "learning_rate": 3.494333520866946e-05, "loss": 0.0108, "step": 185250 }, { "grad_norm": 0.10085059702396393, "learning_rate": 3.49380795696604e-05, "loss": 0.0133, "step": 185260 }, { "grad_norm": 0.12074071913957596, "learning_rate": 3.4932824113667916e-05, "loss": 0.0204, "step": 185270 }, { "grad_norm": 0.08104394376277924, "learning_rate": 3.492756884075582e-05, "loss": 0.0108, "step": 185280 }, { "grad_norm": 0.06191881746053696, "learning_rate": 3.4922313750988e-05, "loss": 0.0119, "step": 185290 }, { "grad_norm": 0.06386497616767883, "learning_rate": 3.491705884442832e-05, "loss": 0.0121, "step": 185300 }, { "grad_norm": 0.10993970185518265, "learning_rate": 3.49118041211406e-05, "loss": 0.0138, "step": 185310 }, { "grad_norm": 0.14490872621536255, "learning_rate": 3.49065495811887e-05, "loss": 0.0116, "step": 185320 }, { "grad_norm": 0.0930967628955841, "learning_rate": 3.4901295224636475e-05, "loss": 0.0174, "step": 185330 }, { "grad_norm": 0.07231307774782181, "learning_rate": 3.489604105154779e-05, "loss": 0.0125, "step": 185340 }, { "grad_norm": 0.11544221639633179, "learning_rate": 3.489078706198644e-05, "loss": 0.0163, "step": 185350 }, { "grad_norm": 0.07055962085723877, "learning_rate": 3.48855332560163e-05, "loss": 0.0143, "step": 185360 }, { "grad_norm": 0.10093438625335693, "learning_rate": 3.48802796337012e-05, "loss": 0.0108, "step": 185370 }, { "grad_norm": 0.1846667379140854, "learning_rate": 3.4875026195105e-05, "loss": 0.0114, "step": 185380 }, { "grad_norm": 0.13368327915668488, "learning_rate": 3.486977294029149e-05, "loss": 0.0141, "step": 185390 }, { "grad_norm": 0.09017928689718246, "learning_rate": 3.486451986932452e-05, "loss": 0.012, "step": 185400 }, { "grad_norm": 0.13578858971595764, "learning_rate": 3.4859266982267944e-05, "loss": 0.0121, "step": 185410 }, { "grad_norm": 0.08096735924482346, "learning_rate": 3.4854014279185556e-05, "loss": 0.0152, "step": 185420 }, { "grad_norm": 0.12612830102443695, "learning_rate": 3.484876176014119e-05, "loss": 0.0149, "step": 185430 }, { "grad_norm": 0.06822744011878967, "learning_rate": 3.484350942519867e-05, "loss": 0.0149, "step": 185440 }, { "grad_norm": 0.07042191177606583, "learning_rate": 3.483825727442185e-05, "loss": 0.0174, "step": 185450 }, { "grad_norm": 0.14631788432598114, "learning_rate": 3.48330053078745e-05, "loss": 0.0108, "step": 185460 }, { "grad_norm": 0.07881069928407669, "learning_rate": 3.4827753525620444e-05, "loss": 0.0134, "step": 185470 }, { "grad_norm": 0.10608047991991043, "learning_rate": 3.482250192772354e-05, "loss": 0.0098, "step": 185480 }, { "grad_norm": 0.1187247633934021, "learning_rate": 3.481725051424754e-05, "loss": 0.0176, "step": 185490 }, { "grad_norm": 0.16047799587249756, "learning_rate": 3.481199928525628e-05, "loss": 0.0121, "step": 185500 }, { "grad_norm": 0.15065257251262665, "learning_rate": 3.480674824081357e-05, "loss": 0.0165, "step": 185510 }, { "grad_norm": 0.07252974808216095, "learning_rate": 3.4801497380983237e-05, "loss": 0.016, "step": 185520 }, { "grad_norm": 0.09717118740081787, "learning_rate": 3.479624670582904e-05, "loss": 0.0173, "step": 185530 }, { "grad_norm": 0.1154583990573883, "learning_rate": 3.479099621541479e-05, "loss": 0.0153, "step": 185540 }, { "grad_norm": 0.135457843542099, "learning_rate": 3.4785745909804334e-05, "loss": 0.0143, "step": 185550 }, { "grad_norm": 0.10006260126829147, "learning_rate": 3.47804957890614e-05, "loss": 0.0131, "step": 185560 }, { "grad_norm": 0.06989967077970505, "learning_rate": 3.477524585324981e-05, "loss": 0.0116, "step": 185570 }, { "grad_norm": 0.151707261800766, "learning_rate": 3.476999610243336e-05, "loss": 0.0152, "step": 185580 }, { "grad_norm": 0.11580369621515274, "learning_rate": 3.4764746536675856e-05, "loss": 0.0131, "step": 185590 }, { "grad_norm": 0.11234871298074722, "learning_rate": 3.475949715604104e-05, "loss": 0.0133, "step": 185600 }, { "grad_norm": 0.20799174904823303, "learning_rate": 3.475424796059273e-05, "loss": 0.013, "step": 185610 }, { "grad_norm": 0.059803739190101624, "learning_rate": 3.474899895039471e-05, "loss": 0.0138, "step": 185620 }, { "grad_norm": 0.11080843210220337, "learning_rate": 3.474375012551076e-05, "loss": 0.0116, "step": 185630 }, { "grad_norm": 0.09668572247028351, "learning_rate": 3.473850148600463e-05, "loss": 0.0116, "step": 185640 }, { "grad_norm": 0.06282156705856323, "learning_rate": 3.473325303194012e-05, "loss": 0.0151, "step": 185650 }, { "grad_norm": 0.05380634963512421, "learning_rate": 3.472800476338101e-05, "loss": 0.0128, "step": 185660 }, { "grad_norm": 0.29183414578437805, "learning_rate": 3.472275668039106e-05, "loss": 0.0144, "step": 185670 }, { "grad_norm": 0.07968681305646896, "learning_rate": 3.4717508783034034e-05, "loss": 0.0131, "step": 185680 }, { "grad_norm": 0.06400745362043381, "learning_rate": 3.4712261071373696e-05, "loss": 0.0167, "step": 185690 }, { "grad_norm": 0.1818530112504959, "learning_rate": 3.470701354547384e-05, "loss": 0.0153, "step": 185700 }, { "grad_norm": 0.04877745732665062, "learning_rate": 3.4701766205398206e-05, "loss": 0.0134, "step": 185710 }, { "grad_norm": 0.07713718712329865, "learning_rate": 3.469651905121054e-05, "loss": 0.0137, "step": 185720 }, { "grad_norm": 0.08828478306531906, "learning_rate": 3.469127208297464e-05, "loss": 0.0113, "step": 185730 }, { "grad_norm": 0.11301087588071823, "learning_rate": 3.4686025300754215e-05, "loss": 0.012, "step": 185740 }, { "grad_norm": 0.0776020884513855, "learning_rate": 3.468077870461306e-05, "loss": 0.0129, "step": 185750 }, { "grad_norm": 0.09970976412296295, "learning_rate": 3.467553229461489e-05, "loss": 0.0126, "step": 185760 }, { "grad_norm": 0.09375789761543274, "learning_rate": 3.467028607082349e-05, "loss": 0.0096, "step": 185770 }, { "grad_norm": 0.06312237679958344, "learning_rate": 3.466504003330257e-05, "loss": 0.0155, "step": 185780 }, { "grad_norm": 0.08832195401191711, "learning_rate": 3.46597941821159e-05, "loss": 0.0124, "step": 185790 }, { "grad_norm": 0.0916948914527893, "learning_rate": 3.4654548517327225e-05, "loss": 0.0169, "step": 185800 }, { "grad_norm": 0.08268138766288757, "learning_rate": 3.464930303900026e-05, "loss": 0.0119, "step": 185810 }, { "grad_norm": 0.1446968913078308, "learning_rate": 3.464405774719875e-05, "loss": 0.0103, "step": 185820 }, { "grad_norm": 0.1046992763876915, "learning_rate": 3.4638812641986454e-05, "loss": 0.0136, "step": 185830 }, { "grad_norm": 0.4726480543613434, "learning_rate": 3.4633567723427086e-05, "loss": 0.0179, "step": 185840 }, { "grad_norm": 0.12520943582057953, "learning_rate": 3.462832299158436e-05, "loss": 0.0149, "step": 185850 }, { "grad_norm": 0.11065094918012619, "learning_rate": 3.462307844652204e-05, "loss": 0.0131, "step": 185860 }, { "grad_norm": 0.23822087049484253, "learning_rate": 3.461783408830384e-05, "loss": 0.0124, "step": 185870 }, { "grad_norm": 0.14944598078727722, "learning_rate": 3.461258991699346e-05, "loss": 0.0144, "step": 185880 }, { "grad_norm": 0.15664750337600708, "learning_rate": 3.4607345932654656e-05, "loss": 0.0087, "step": 185890 }, { "grad_norm": 0.07181987166404724, "learning_rate": 3.460210213535112e-05, "loss": 0.0113, "step": 185900 }, { "grad_norm": 0.12586179375648499, "learning_rate": 3.45968585251466e-05, "loss": 0.0137, "step": 185910 }, { "grad_norm": 0.14493079483509064, "learning_rate": 3.459161510210478e-05, "loss": 0.015, "step": 185920 }, { "grad_norm": 0.08300477266311646, "learning_rate": 3.4586371866289366e-05, "loss": 0.012, "step": 185930 }, { "grad_norm": 0.07815441489219666, "learning_rate": 3.4581128817764107e-05, "loss": 0.012, "step": 185940 }, { "grad_norm": 0.08425053209066391, "learning_rate": 3.45758859565927e-05, "loss": 0.0126, "step": 185950 }, { "grad_norm": 0.08439558744430542, "learning_rate": 3.4570643282838824e-05, "loss": 0.0125, "step": 185960 }, { "grad_norm": 0.08706759661436081, "learning_rate": 3.45654007965662e-05, "loss": 0.0111, "step": 185970 }, { "grad_norm": 0.08323623985052109, "learning_rate": 3.456015849783856e-05, "loss": 0.0111, "step": 185980 }, { "grad_norm": 0.0870073139667511, "learning_rate": 3.455491638671954e-05, "loss": 0.0105, "step": 185990 }, { "grad_norm": 0.062095604836940765, "learning_rate": 3.454967446327286e-05, "loss": 0.0099, "step": 186000 }, { "grad_norm": 0.08660731464624405, "learning_rate": 3.454443272756224e-05, "loss": 0.0148, "step": 186010 }, { "grad_norm": 0.09079543501138687, "learning_rate": 3.453919117965136e-05, "loss": 0.0128, "step": 186020 }, { "grad_norm": 0.2476610690355301, "learning_rate": 3.453394981960389e-05, "loss": 0.0168, "step": 186030 }, { "grad_norm": 0.24363288283348083, "learning_rate": 3.452870864748354e-05, "loss": 0.0177, "step": 186040 }, { "grad_norm": 0.21960105001926422, "learning_rate": 3.4523467663354e-05, "loss": 0.018, "step": 186050 }, { "grad_norm": 0.07522421330213547, "learning_rate": 3.451822686727893e-05, "loss": 0.0126, "step": 186060 }, { "grad_norm": 0.14744426310062408, "learning_rate": 3.4512986259322005e-05, "loss": 0.0197, "step": 186070 }, { "grad_norm": 0.07400009036064148, "learning_rate": 3.450774583954692e-05, "loss": 0.0166, "step": 186080 }, { "grad_norm": 0.1246132031083107, "learning_rate": 3.450250560801739e-05, "loss": 0.0123, "step": 186090 }, { "grad_norm": 0.11532454192638397, "learning_rate": 3.449726556479701e-05, "loss": 0.014, "step": 186100 }, { "grad_norm": 0.19981133937835693, "learning_rate": 3.449202570994949e-05, "loss": 0.0114, "step": 186110 }, { "grad_norm": 0.10130488127470016, "learning_rate": 3.4486786043538535e-05, "loss": 0.0171, "step": 186120 }, { "grad_norm": 0.0905490592122078, "learning_rate": 3.448154656562774e-05, "loss": 0.018, "step": 186130 }, { "grad_norm": 0.059361159801483154, "learning_rate": 3.4476307276280815e-05, "loss": 0.0111, "step": 186140 }, { "grad_norm": 0.054299063980579376, "learning_rate": 3.447106817556141e-05, "loss": 0.0105, "step": 186150 }, { "grad_norm": 0.19163094460964203, "learning_rate": 3.446582926353321e-05, "loss": 0.0157, "step": 186160 }, { "grad_norm": 0.10022298246622086, "learning_rate": 3.446059054025983e-05, "loss": 0.0131, "step": 186170 }, { "grad_norm": 0.13960711658000946, "learning_rate": 3.445535200580495e-05, "loss": 0.0157, "step": 186180 }, { "grad_norm": 0.05483591929078102, "learning_rate": 3.4450113660232244e-05, "loss": 0.0125, "step": 186190 }, { "grad_norm": 0.3385617434978485, "learning_rate": 3.444487550360531e-05, "loss": 0.0167, "step": 186200 }, { "grad_norm": 0.10221098363399506, "learning_rate": 3.443963753598782e-05, "loss": 0.0148, "step": 186210 }, { "grad_norm": 0.07327147573232651, "learning_rate": 3.443439975744344e-05, "loss": 0.0121, "step": 186220 }, { "grad_norm": 0.06196596100926399, "learning_rate": 3.442916216803581e-05, "loss": 0.0133, "step": 186230 }, { "grad_norm": 0.06038226932287216, "learning_rate": 3.442392476782854e-05, "loss": 0.0111, "step": 186240 }, { "grad_norm": 0.14360415935516357, "learning_rate": 3.441868755688529e-05, "loss": 0.0212, "step": 186250 }, { "grad_norm": 0.12383418530225754, "learning_rate": 3.44134505352697e-05, "loss": 0.0121, "step": 186260 }, { "grad_norm": 0.3610582947731018, "learning_rate": 3.440821370304542e-05, "loss": 0.0129, "step": 186270 }, { "grad_norm": 0.08445216715335846, "learning_rate": 3.440297706027606e-05, "loss": 0.0132, "step": 186280 }, { "grad_norm": 0.11205735802650452, "learning_rate": 3.4397740607025234e-05, "loss": 0.0166, "step": 186290 }, { "grad_norm": 0.10200851410627365, "learning_rate": 3.439250434335662e-05, "loss": 0.0178, "step": 186300 }, { "grad_norm": 0.07871738821268082, "learning_rate": 3.438726826933379e-05, "loss": 0.0084, "step": 186310 }, { "grad_norm": 0.14580896496772766, "learning_rate": 3.438203238502039e-05, "loss": 0.0104, "step": 186320 }, { "grad_norm": 0.21030621230602264, "learning_rate": 3.437679669048005e-05, "loss": 0.0145, "step": 186330 }, { "grad_norm": 0.06955191493034363, "learning_rate": 3.43715611857764e-05, "loss": 0.0111, "step": 186340 }, { "grad_norm": 0.07627276331186295, "learning_rate": 3.436632587097302e-05, "loss": 0.0113, "step": 186350 }, { "grad_norm": 0.09646020084619522, "learning_rate": 3.436109074613353e-05, "loss": 0.0127, "step": 186360 }, { "grad_norm": 0.10076672583818436, "learning_rate": 3.435585581132157e-05, "loss": 0.0215, "step": 186370 }, { "grad_norm": 0.3293757438659668, "learning_rate": 3.435062106660073e-05, "loss": 0.0135, "step": 186380 }, { "grad_norm": 0.12223944067955017, "learning_rate": 3.434538651203461e-05, "loss": 0.0149, "step": 186390 }, { "grad_norm": 0.07621601223945618, "learning_rate": 3.4340152147686824e-05, "loss": 0.0106, "step": 186400 }, { "grad_norm": 0.17874296009540558, "learning_rate": 3.4334917973620987e-05, "loss": 0.0144, "step": 186410 }, { "grad_norm": 0.09268946945667267, "learning_rate": 3.4329683989900684e-05, "loss": 0.015, "step": 186420 }, { "grad_norm": 0.06851826608181, "learning_rate": 3.4324450196589506e-05, "loss": 0.0132, "step": 186430 }, { "grad_norm": 0.06255878508090973, "learning_rate": 3.431921659375107e-05, "loss": 0.011, "step": 186440 }, { "grad_norm": 0.3415505290031433, "learning_rate": 3.431398318144895e-05, "loss": 0.0153, "step": 186450 }, { "grad_norm": 0.0987846851348877, "learning_rate": 3.430874995974675e-05, "loss": 0.0148, "step": 186460 }, { "grad_norm": 0.06281139701604843, "learning_rate": 3.430351692870804e-05, "loss": 0.0112, "step": 186470 }, { "grad_norm": 0.0934961661696434, "learning_rate": 3.429828408839644e-05, "loss": 0.0085, "step": 186480 }, { "grad_norm": 0.19650696218013763, "learning_rate": 3.429305143887549e-05, "loss": 0.0126, "step": 186490 }, { "grad_norm": 0.06856313347816467, "learning_rate": 3.428781898020882e-05, "loss": 0.0133, "step": 186500 }, { "grad_norm": 0.20462948083877563, "learning_rate": 3.4282586712459976e-05, "loss": 0.0095, "step": 186510 }, { "grad_norm": 0.10167406499385834, "learning_rate": 3.427735463569253e-05, "loss": 0.0118, "step": 186520 }, { "grad_norm": 0.39655014872550964, "learning_rate": 3.427212274997007e-05, "loss": 0.0121, "step": 186530 }, { "grad_norm": 0.07884323596954346, "learning_rate": 3.426689105535617e-05, "loss": 0.0155, "step": 186540 }, { "grad_norm": 0.1978330910205841, "learning_rate": 3.426165955191442e-05, "loss": 0.0117, "step": 186550 }, { "grad_norm": 0.06312301009893417, "learning_rate": 3.425642823970835e-05, "loss": 0.0115, "step": 186560 }, { "grad_norm": 0.18822544813156128, "learning_rate": 3.4251197118801534e-05, "loss": 0.012, "step": 186570 }, { "grad_norm": 0.17881280183792114, "learning_rate": 3.4245966189257536e-05, "loss": 0.012, "step": 186580 }, { "grad_norm": 0.12925243377685547, "learning_rate": 3.424073545113996e-05, "loss": 0.014, "step": 186590 }, { "grad_norm": 0.10963655263185501, "learning_rate": 3.42355049045123e-05, "loss": 0.0085, "step": 186600 }, { "grad_norm": 0.05123920366168022, "learning_rate": 3.423027454943813e-05, "loss": 0.014, "step": 186610 }, { "grad_norm": 0.07742400467395782, "learning_rate": 3.4225044385981054e-05, "loss": 0.0109, "step": 186620 }, { "grad_norm": 0.05510835349559784, "learning_rate": 3.4219814414204545e-05, "loss": 0.0087, "step": 186630 }, { "grad_norm": 0.14778941869735718, "learning_rate": 3.42145846341722e-05, "loss": 0.0124, "step": 186640 }, { "grad_norm": 0.09034418314695358, "learning_rate": 3.420935504594754e-05, "loss": 0.0112, "step": 186650 }, { "grad_norm": 0.28412020206451416, "learning_rate": 3.420412564959416e-05, "loss": 0.0146, "step": 186660 }, { "grad_norm": 0.08655502647161484, "learning_rate": 3.419889644517555e-05, "loss": 0.0206, "step": 186670 }, { "grad_norm": 0.06619873642921448, "learning_rate": 3.419366743275526e-05, "loss": 0.0159, "step": 186680 }, { "grad_norm": 0.21721158921718597, "learning_rate": 3.4188438612396855e-05, "loss": 0.0126, "step": 186690 }, { "grad_norm": 0.08652205020189285, "learning_rate": 3.418320998416383e-05, "loss": 0.0148, "step": 186700 }, { "grad_norm": 0.07871148735284805, "learning_rate": 3.417798154811973e-05, "loss": 0.0105, "step": 186710 }, { "grad_norm": 0.09775559604167938, "learning_rate": 3.417275330432811e-05, "loss": 0.0189, "step": 186720 }, { "grad_norm": 0.2045925259590149, "learning_rate": 3.4167525252852484e-05, "loss": 0.0151, "step": 186730 }, { "grad_norm": 0.08923546224832535, "learning_rate": 3.416229739375636e-05, "loss": 0.0112, "step": 186740 }, { "grad_norm": 0.059467777609825134, "learning_rate": 3.4157069727103274e-05, "loss": 0.0179, "step": 186750 }, { "grad_norm": 0.06047675386071205, "learning_rate": 3.415184225295677e-05, "loss": 0.0106, "step": 186760 }, { "grad_norm": 0.10384296625852585, "learning_rate": 3.414661497138033e-05, "loss": 0.0155, "step": 186770 }, { "grad_norm": 0.23613479733467102, "learning_rate": 3.4141387882437484e-05, "loss": 0.0154, "step": 186780 }, { "grad_norm": 0.05295710638165474, "learning_rate": 3.4136160986191744e-05, "loss": 0.0159, "step": 186790 }, { "grad_norm": 0.0601591132581234, "learning_rate": 3.4130934282706653e-05, "loss": 0.0108, "step": 186800 }, { "grad_norm": 0.23369081318378448, "learning_rate": 3.4125707772045666e-05, "loss": 0.0103, "step": 186810 }, { "grad_norm": 0.07348848879337311, "learning_rate": 3.412048145427232e-05, "loss": 0.0112, "step": 186820 }, { "grad_norm": 0.13158755004405975, "learning_rate": 3.411525532945012e-05, "loss": 0.0159, "step": 186830 }, { "grad_norm": 0.1652757078409195, "learning_rate": 3.411002939764258e-05, "loss": 0.0115, "step": 186840 }, { "grad_norm": 0.06711474806070328, "learning_rate": 3.410480365891318e-05, "loss": 0.0167, "step": 186850 }, { "grad_norm": 0.12411963939666748, "learning_rate": 3.4099578113325406e-05, "loss": 0.021, "step": 186860 }, { "grad_norm": 0.09521503746509552, "learning_rate": 3.40943527609428e-05, "loss": 0.019, "step": 186870 }, { "grad_norm": 0.08381135016679764, "learning_rate": 3.40891276018288e-05, "loss": 0.0118, "step": 186880 }, { "grad_norm": 0.08340678364038467, "learning_rate": 3.4083902636046926e-05, "loss": 0.019, "step": 186890 }, { "grad_norm": 0.12307596951723099, "learning_rate": 3.407867786366066e-05, "loss": 0.009, "step": 186900 }, { "grad_norm": 0.08047757297754288, "learning_rate": 3.407345328473351e-05, "loss": 0.0094, "step": 186910 }, { "grad_norm": 0.10840930789709091, "learning_rate": 3.406822889932892e-05, "loss": 0.0115, "step": 186920 }, { "grad_norm": 0.13967560231685638, "learning_rate": 3.406300470751039e-05, "loss": 0.0183, "step": 186930 }, { "grad_norm": 0.11254921555519104, "learning_rate": 3.405778070934142e-05, "loss": 0.0177, "step": 186940 }, { "grad_norm": 0.08782779425382614, "learning_rate": 3.4052556904885446e-05, "loss": 0.0151, "step": 186950 }, { "grad_norm": 0.10234230756759644, "learning_rate": 3.404733329420596e-05, "loss": 0.013, "step": 186960 }, { "grad_norm": 0.05779870226979256, "learning_rate": 3.4042109877366446e-05, "loss": 0.0093, "step": 186970 }, { "grad_norm": 0.17413614690303802, "learning_rate": 3.4036886654430375e-05, "loss": 0.0166, "step": 186980 }, { "grad_norm": 0.1666446179151535, "learning_rate": 3.403166362546119e-05, "loss": 0.0179, "step": 186990 }, { "grad_norm": 0.11491641402244568, "learning_rate": 3.4026440790522364e-05, "loss": 0.0148, "step": 187000 }, { "grad_norm": 0.05102037638425827, "learning_rate": 3.402121814967738e-05, "loss": 0.0109, "step": 187010 }, { "grad_norm": 0.12449248880147934, "learning_rate": 3.401599570298967e-05, "loss": 0.0081, "step": 187020 }, { "grad_norm": 0.058105841279029846, "learning_rate": 3.40107734505227e-05, "loss": 0.0106, "step": 187030 }, { "grad_norm": 0.08403424918651581, "learning_rate": 3.4005551392339926e-05, "loss": 0.0112, "step": 187040 }, { "grad_norm": 0.09189268201589584, "learning_rate": 3.4000329528504825e-05, "loss": 0.0144, "step": 187050 }, { "grad_norm": 0.11132572591304779, "learning_rate": 3.3995107859080814e-05, "loss": 0.0121, "step": 187060 }, { "grad_norm": 0.10302679985761642, "learning_rate": 3.398988638413134e-05, "loss": 0.0194, "step": 187070 }, { "grad_norm": 0.08797436207532883, "learning_rate": 3.398466510371988e-05, "loss": 0.0135, "step": 187080 }, { "grad_norm": 0.13741514086723328, "learning_rate": 3.397944401790987e-05, "loss": 0.0119, "step": 187090 }, { "grad_norm": 0.44266924262046814, "learning_rate": 3.3974223126764715e-05, "loss": 0.0167, "step": 187100 }, { "grad_norm": 0.11954700946807861, "learning_rate": 3.396900243034789e-05, "loss": 0.0164, "step": 187110 }, { "grad_norm": 0.17495174705982208, "learning_rate": 3.396378192872283e-05, "loss": 0.0152, "step": 187120 }, { "grad_norm": 0.14257673919200897, "learning_rate": 3.3958561621952946e-05, "loss": 0.0132, "step": 187130 }, { "grad_norm": 0.09498561918735504, "learning_rate": 3.395334151010169e-05, "loss": 0.0147, "step": 187140 }, { "grad_norm": 0.06535622477531433, "learning_rate": 3.3948121593232486e-05, "loss": 0.0129, "step": 187150 }, { "grad_norm": 0.0660572350025177, "learning_rate": 3.394290187140876e-05, "loss": 0.0086, "step": 187160 }, { "grad_norm": 0.10859210789203644, "learning_rate": 3.3937682344693935e-05, "loss": 0.013, "step": 187170 }, { "grad_norm": 0.18799899518489838, "learning_rate": 3.393246301315144e-05, "loss": 0.019, "step": 187180 }, { "grad_norm": 0.21281588077545166, "learning_rate": 3.3927243876844686e-05, "loss": 0.0147, "step": 187190 }, { "grad_norm": 0.17064698040485382, "learning_rate": 3.392202493583709e-05, "loss": 0.0173, "step": 187200 }, { "grad_norm": 0.1304396390914917, "learning_rate": 3.391680619019208e-05, "loss": 0.0142, "step": 187210 }, { "grad_norm": 0.06799020618200302, "learning_rate": 3.391158763997305e-05, "loss": 0.0135, "step": 187220 }, { "grad_norm": 0.033796459436416626, "learning_rate": 3.3906369285243434e-05, "loss": 0.0122, "step": 187230 }, { "grad_norm": 0.18087954819202423, "learning_rate": 3.390115112606661e-05, "loss": 0.0142, "step": 187240 }, { "grad_norm": 0.16466733813285828, "learning_rate": 3.3895933162506e-05, "loss": 0.0106, "step": 187250 }, { "grad_norm": 0.10330938547849655, "learning_rate": 3.3890715394625025e-05, "loss": 0.0148, "step": 187260 }, { "grad_norm": 0.72233647108078, "learning_rate": 3.3885497822487055e-05, "loss": 0.012, "step": 187270 }, { "grad_norm": 0.05974867567420006, "learning_rate": 3.3880280446155496e-05, "loss": 0.012, "step": 187280 }, { "grad_norm": 0.11810482293367386, "learning_rate": 3.387506326569375e-05, "loss": 0.015, "step": 187290 }, { "grad_norm": 0.10649214684963226, "learning_rate": 3.386984628116523e-05, "loss": 0.0134, "step": 187300 }, { "grad_norm": 0.13907165825366974, "learning_rate": 3.386462949263329e-05, "loss": 0.022, "step": 187310 }, { "grad_norm": 0.09202444553375244, "learning_rate": 3.3859412900161336e-05, "loss": 0.0138, "step": 187320 }, { "grad_norm": 0.32132473587989807, "learning_rate": 3.385419650381277e-05, "loss": 0.012, "step": 187330 }, { "grad_norm": 0.08266313374042511, "learning_rate": 3.384898030365094e-05, "loss": 0.0172, "step": 187340 }, { "grad_norm": 0.08297775685787201, "learning_rate": 3.3843764299739255e-05, "loss": 0.0132, "step": 187350 }, { "grad_norm": 0.1319277286529541, "learning_rate": 3.383854849214108e-05, "loss": 0.0165, "step": 187360 }, { "grad_norm": 0.13591435551643372, "learning_rate": 3.383333288091983e-05, "loss": 0.0174, "step": 187370 }, { "grad_norm": 0.08341985195875168, "learning_rate": 3.382811746613882e-05, "loss": 0.0103, "step": 187380 }, { "grad_norm": 0.1344980001449585, "learning_rate": 3.3822902247861454e-05, "loss": 0.0243, "step": 187390 }, { "grad_norm": 0.07965084910392761, "learning_rate": 3.381768722615112e-05, "loss": 0.0171, "step": 187400 }, { "grad_norm": 0.08288899809122086, "learning_rate": 3.381247240107114e-05, "loss": 0.0128, "step": 187410 }, { "grad_norm": 0.13070018589496613, "learning_rate": 3.380725777268491e-05, "loss": 0.0163, "step": 187420 }, { "grad_norm": 0.11268968880176544, "learning_rate": 3.380204334105579e-05, "loss": 0.0139, "step": 187430 }, { "grad_norm": 0.13369761407375336, "learning_rate": 3.379682910624715e-05, "loss": 0.0115, "step": 187440 }, { "grad_norm": 0.1277664303779602, "learning_rate": 3.379161506832231e-05, "loss": 0.0166, "step": 187450 }, { "grad_norm": 0.054559867829084396, "learning_rate": 3.378640122734466e-05, "loss": 0.0123, "step": 187460 }, { "grad_norm": 0.08218082785606384, "learning_rate": 3.3781187583377536e-05, "loss": 0.0107, "step": 187470 }, { "grad_norm": 0.1282362937927246, "learning_rate": 3.377597413648432e-05, "loss": 0.0163, "step": 187480 }, { "grad_norm": 0.0759279727935791, "learning_rate": 3.377076088672832e-05, "loss": 0.0162, "step": 187490 }, { "grad_norm": 0.09982235729694366, "learning_rate": 3.3765547834172896e-05, "loss": 0.0187, "step": 187500 }, { "grad_norm": 0.07898125797510147, "learning_rate": 3.3760334978881414e-05, "loss": 0.0141, "step": 187510 }, { "grad_norm": 0.08442001789808273, "learning_rate": 3.3755122320917173e-05, "loss": 0.0102, "step": 187520 }, { "grad_norm": 0.1321909874677658, "learning_rate": 3.374990986034353e-05, "loss": 0.0132, "step": 187530 }, { "grad_norm": 0.10501312464475632, "learning_rate": 3.374469759722383e-05, "loss": 0.0096, "step": 187540 }, { "grad_norm": 0.07764802128076553, "learning_rate": 3.3739485531621415e-05, "loss": 0.0139, "step": 187550 }, { "grad_norm": 0.04834025725722313, "learning_rate": 3.373427366359959e-05, "loss": 0.015, "step": 187560 }, { "grad_norm": 0.06913590431213379, "learning_rate": 3.37290619932217e-05, "loss": 0.0148, "step": 187570 }, { "grad_norm": 0.0584455206990242, "learning_rate": 3.372385052055108e-05, "loss": 0.0115, "step": 187580 }, { "grad_norm": 0.12102609872817993, "learning_rate": 3.3718639245651026e-05, "loss": 0.0145, "step": 187590 }, { "grad_norm": 0.05839953199028969, "learning_rate": 3.3713428168584884e-05, "loss": 0.026, "step": 187600 }, { "grad_norm": 0.17057791352272034, "learning_rate": 3.370821728941596e-05, "loss": 0.0198, "step": 187610 }, { "grad_norm": 0.06976745277643204, "learning_rate": 3.37030066082076e-05, "loss": 0.0157, "step": 187620 }, { "grad_norm": 0.12026460468769073, "learning_rate": 3.3697796125023076e-05, "loss": 0.011, "step": 187630 }, { "grad_norm": 0.10151660442352295, "learning_rate": 3.3692585839925726e-05, "loss": 0.0142, "step": 187640 }, { "grad_norm": 0.1490112990140915, "learning_rate": 3.3687375752978867e-05, "loss": 0.015, "step": 187650 }, { "grad_norm": 0.15216167271137238, "learning_rate": 3.368216586424578e-05, "loss": 0.0174, "step": 187660 }, { "grad_norm": 0.1751721054315567, "learning_rate": 3.3676956173789784e-05, "loss": 0.0247, "step": 187670 }, { "grad_norm": 0.10264594852924347, "learning_rate": 3.3671746681674176e-05, "loss": 0.0147, "step": 187680 }, { "grad_norm": 0.10993095487356186, "learning_rate": 3.366653738796229e-05, "loss": 0.0147, "step": 187690 }, { "grad_norm": 0.1707526445388794, "learning_rate": 3.366132829271737e-05, "loss": 0.0183, "step": 187700 }, { "grad_norm": 0.12463212758302689, "learning_rate": 3.365611939600274e-05, "loss": 0.0146, "step": 187710 }, { "grad_norm": 0.1771264374256134, "learning_rate": 3.36509106978817e-05, "loss": 0.0171, "step": 187720 }, { "grad_norm": 0.19007229804992676, "learning_rate": 3.364570219841753e-05, "loss": 0.0195, "step": 187730 }, { "grad_norm": 0.19174478948116302, "learning_rate": 3.364049389767351e-05, "loss": 0.0141, "step": 187740 }, { "grad_norm": 0.1344849020242691, "learning_rate": 3.3635285795712934e-05, "loss": 0.0142, "step": 187750 }, { "grad_norm": 0.0984177440404892, "learning_rate": 3.36300778925991e-05, "loss": 0.0095, "step": 187760 }, { "grad_norm": 0.09417089819908142, "learning_rate": 3.3624870188395274e-05, "loss": 0.0124, "step": 187770 }, { "grad_norm": 0.21760743856430054, "learning_rate": 3.361966268316473e-05, "loss": 0.018, "step": 187780 }, { "grad_norm": 0.13473336398601532, "learning_rate": 3.3614455376970744e-05, "loss": 0.0233, "step": 187790 }, { "grad_norm": 0.09891398251056671, "learning_rate": 3.3609248269876604e-05, "loss": 0.0104, "step": 187800 }, { "grad_norm": 0.16146567463874817, "learning_rate": 3.360404136194558e-05, "loss": 0.0195, "step": 187810 }, { "grad_norm": 0.1313544660806656, "learning_rate": 3.359883465324092e-05, "loss": 0.0142, "step": 187820 }, { "grad_norm": 0.2257879227399826, "learning_rate": 3.3593628143825916e-05, "loss": 0.0146, "step": 187830 }, { "grad_norm": 0.11059003323316574, "learning_rate": 3.3588421833763806e-05, "loss": 0.0115, "step": 187840 }, { "grad_norm": 0.14289788901805878, "learning_rate": 3.358321572311788e-05, "loss": 0.0138, "step": 187850 }, { "grad_norm": 0.04583834856748581, "learning_rate": 3.357800981195137e-05, "loss": 0.0185, "step": 187860 }, { "grad_norm": 0.102397620677948, "learning_rate": 3.357280410032756e-05, "loss": 0.0135, "step": 187870 }, { "grad_norm": 0.23128393292427063, "learning_rate": 3.3567598588309676e-05, "loss": 0.016, "step": 187880 }, { "grad_norm": 0.16595835983753204, "learning_rate": 3.356239327596099e-05, "loss": 0.0153, "step": 187890 }, { "grad_norm": 0.08743423968553543, "learning_rate": 3.355718816334475e-05, "loss": 0.0114, "step": 187900 }, { "grad_norm": 0.18795596063137054, "learning_rate": 3.355198325052419e-05, "loss": 0.0182, "step": 187910 }, { "grad_norm": 0.051731668412685394, "learning_rate": 3.3546778537562564e-05, "loss": 0.0133, "step": 187920 }, { "grad_norm": 0.05119454115629196, "learning_rate": 3.354157402452312e-05, "loss": 0.0099, "step": 187930 }, { "grad_norm": 0.15643233060836792, "learning_rate": 3.353636971146909e-05, "loss": 0.0137, "step": 187940 }, { "grad_norm": 0.10564538836479187, "learning_rate": 3.35311655984637e-05, "loss": 0.0161, "step": 187950 }, { "grad_norm": 0.21971721947193146, "learning_rate": 3.352596168557021e-05, "loss": 0.0194, "step": 187960 }, { "grad_norm": 0.09645799547433853, "learning_rate": 3.3520757972851846e-05, "loss": 0.0128, "step": 187970 }, { "grad_norm": 0.18780812621116638, "learning_rate": 3.3515554460371815e-05, "loss": 0.0156, "step": 187980 }, { "grad_norm": 0.15862245857715607, "learning_rate": 3.3510351148193364e-05, "loss": 0.0196, "step": 187990 }, { "grad_norm": 0.10269345343112946, "learning_rate": 3.3505148036379716e-05, "loss": 0.0137, "step": 188000 }, { "grad_norm": 0.060628972947597504, "learning_rate": 3.3499945124994115e-05, "loss": 0.0136, "step": 188010 }, { "grad_norm": 0.15176984667778015, "learning_rate": 3.3494742414099735e-05, "loss": 0.0126, "step": 188020 }, { "grad_norm": 0.07250723242759705, "learning_rate": 3.348953990375981e-05, "loss": 0.015, "step": 188030 }, { "grad_norm": 0.12360326945781708, "learning_rate": 3.348433759403757e-05, "loss": 0.0163, "step": 188040 }, { "grad_norm": 0.09695669263601303, "learning_rate": 3.347913548499625e-05, "loss": 0.0198, "step": 188050 }, { "grad_norm": 0.0907076820731163, "learning_rate": 3.347393357669901e-05, "loss": 0.0212, "step": 188060 }, { "grad_norm": 0.06418294459581375, "learning_rate": 3.346873186920907e-05, "loss": 0.011, "step": 188070 }, { "grad_norm": 0.08202304691076279, "learning_rate": 3.3463530362589676e-05, "loss": 0.0155, "step": 188080 }, { "grad_norm": 0.12075649201869965, "learning_rate": 3.3458329056903975e-05, "loss": 0.0161, "step": 188090 }, { "grad_norm": 0.27037855982780457, "learning_rate": 3.3453127952215195e-05, "loss": 0.0132, "step": 188100 }, { "grad_norm": 0.12653489410877228, "learning_rate": 3.344792704858654e-05, "loss": 0.0166, "step": 188110 }, { "grad_norm": 0.10023168474435806, "learning_rate": 3.344272634608121e-05, "loss": 0.0119, "step": 188120 }, { "grad_norm": 0.2305683046579361, "learning_rate": 3.343752584476237e-05, "loss": 0.0113, "step": 188130 }, { "grad_norm": 0.13161346316337585, "learning_rate": 3.3432325544693234e-05, "loss": 0.0111, "step": 188140 }, { "grad_norm": 0.06776443868875504, "learning_rate": 3.3427125445936994e-05, "loss": 0.0133, "step": 188150 }, { "grad_norm": 0.05778350681066513, "learning_rate": 3.342192554855682e-05, "loss": 0.0116, "step": 188160 }, { "grad_norm": 0.3566524088382721, "learning_rate": 3.3416725852615894e-05, "loss": 0.0131, "step": 188170 }, { "grad_norm": 0.053367048501968384, "learning_rate": 3.341152635817741e-05, "loss": 0.013, "step": 188180 }, { "grad_norm": 0.1138165220618248, "learning_rate": 3.340632706530456e-05, "loss": 0.0131, "step": 188190 }, { "grad_norm": 0.18562325835227966, "learning_rate": 3.340112797406049e-05, "loss": 0.0159, "step": 188200 }, { "grad_norm": 0.12905851006507874, "learning_rate": 3.339592908450838e-05, "loss": 0.0141, "step": 188210 }, { "grad_norm": 0.21247899532318115, "learning_rate": 3.3390730396711426e-05, "loss": 0.0167, "step": 188220 }, { "grad_norm": 0.1477205753326416, "learning_rate": 3.3385531910732765e-05, "loss": 0.0119, "step": 188230 }, { "grad_norm": 0.0854615569114685, "learning_rate": 3.338033362663557e-05, "loss": 0.0072, "step": 188240 }, { "grad_norm": 0.07412886619567871, "learning_rate": 3.3375135544483025e-05, "loss": 0.0117, "step": 188250 }, { "grad_norm": 0.07826566696166992, "learning_rate": 3.3369937664338296e-05, "loss": 0.0124, "step": 188260 }, { "grad_norm": 0.10601780563592911, "learning_rate": 3.3364739986264506e-05, "loss": 0.0162, "step": 188270 }, { "grad_norm": 0.09240500628948212, "learning_rate": 3.335954251032483e-05, "loss": 0.0148, "step": 188280 }, { "grad_norm": 0.11453377455472946, "learning_rate": 3.335434523658244e-05, "loss": 0.0123, "step": 188290 }, { "grad_norm": 0.05422466620802879, "learning_rate": 3.334914816510046e-05, "loss": 0.0142, "step": 188300 }, { "grad_norm": 0.1403363049030304, "learning_rate": 3.334395129594205e-05, "loss": 0.0224, "step": 188310 }, { "grad_norm": 0.11251305788755417, "learning_rate": 3.333875462917034e-05, "loss": 0.0144, "step": 188320 }, { "grad_norm": 0.10812894999980927, "learning_rate": 3.333355816484853e-05, "loss": 0.0209, "step": 188330 }, { "grad_norm": 0.08647070080041885, "learning_rate": 3.33283619030397e-05, "loss": 0.0126, "step": 188340 }, { "grad_norm": 0.1559220850467682, "learning_rate": 3.3323165843807016e-05, "loss": 0.0097, "step": 188350 }, { "grad_norm": 0.1297753006219864, "learning_rate": 3.3317969987213604e-05, "loss": 0.0171, "step": 188360 }, { "grad_norm": 0.14042618870735168, "learning_rate": 3.3312774333322636e-05, "loss": 0.01, "step": 188370 }, { "grad_norm": 0.07478182762861252, "learning_rate": 3.330757888219719e-05, "loss": 0.0112, "step": 188380 }, { "grad_norm": 0.08131960779428482, "learning_rate": 3.3302383633900425e-05, "loss": 0.0085, "step": 188390 }, { "grad_norm": 0.07435497641563416, "learning_rate": 3.329718858849548e-05, "loss": 0.0155, "step": 188400 }, { "grad_norm": 0.026377681642770767, "learning_rate": 3.3291993746045444e-05, "loss": 0.012, "step": 188410 }, { "grad_norm": 0.05118143558502197, "learning_rate": 3.328679910661347e-05, "loss": 0.0143, "step": 188420 }, { "grad_norm": 0.0700179934501648, "learning_rate": 3.328160467026265e-05, "loss": 0.0144, "step": 188430 }, { "grad_norm": 0.15563462674617767, "learning_rate": 3.3276410437056146e-05, "loss": 0.0151, "step": 188440 }, { "grad_norm": 0.07988227158784866, "learning_rate": 3.327121640705702e-05, "loss": 0.0121, "step": 188450 }, { "grad_norm": 0.07306627929210663, "learning_rate": 3.326602258032842e-05, "loss": 0.0151, "step": 188460 }, { "grad_norm": 0.1611955314874649, "learning_rate": 3.326082895693344e-05, "loss": 0.0177, "step": 188470 }, { "grad_norm": 0.0963425263762474, "learning_rate": 3.3255635536935205e-05, "loss": 0.0111, "step": 188480 }, { "grad_norm": 0.06962727755308151, "learning_rate": 3.325044232039679e-05, "loss": 0.0149, "step": 188490 }, { "grad_norm": 0.10625340044498444, "learning_rate": 3.324524930738132e-05, "loss": 0.0254, "step": 188500 }, { "grad_norm": 0.1486767679452896, "learning_rate": 3.32400564979519e-05, "loss": 0.0152, "step": 188510 }, { "grad_norm": 0.07903289794921875, "learning_rate": 3.323486389217161e-05, "loss": 0.013, "step": 188520 }, { "grad_norm": 0.17670543491840363, "learning_rate": 3.3229671490103555e-05, "loss": 0.0132, "step": 188530 }, { "grad_norm": 0.060246244072914124, "learning_rate": 3.322447929181083e-05, "loss": 0.0118, "step": 188540 }, { "grad_norm": 0.088985875248909, "learning_rate": 3.3219287297356507e-05, "loss": 0.015, "step": 188550 }, { "grad_norm": 0.17890626192092896, "learning_rate": 3.32140955068037e-05, "loss": 0.0133, "step": 188560 }, { "grad_norm": 0.09386872500181198, "learning_rate": 3.320890392021547e-05, "loss": 0.0192, "step": 188570 }, { "grad_norm": 0.5898045897483826, "learning_rate": 3.320371253765492e-05, "loss": 0.0119, "step": 188580 }, { "grad_norm": 0.2626536190509796, "learning_rate": 3.319852135918511e-05, "loss": 0.0233, "step": 188590 }, { "grad_norm": 0.06073864921927452, "learning_rate": 3.319333038486914e-05, "loss": 0.0148, "step": 188600 }, { "grad_norm": 0.08948324620723724, "learning_rate": 3.318813961477008e-05, "loss": 0.0105, "step": 188610 }, { "grad_norm": 0.10161270946264267, "learning_rate": 3.3182949048950974e-05, "loss": 0.0127, "step": 188620 }, { "grad_norm": 0.1376010775566101, "learning_rate": 3.317775868747492e-05, "loss": 0.0149, "step": 188630 }, { "grad_norm": 0.24184538424015045, "learning_rate": 3.3172568530405e-05, "loss": 0.019, "step": 188640 }, { "grad_norm": 0.13776619732379913, "learning_rate": 3.3167378577804254e-05, "loss": 0.0108, "step": 188650 }, { "grad_norm": 0.16788378357887268, "learning_rate": 3.316218882973574e-05, "loss": 0.0157, "step": 188660 }, { "grad_norm": 0.11020203679800034, "learning_rate": 3.315699928626252e-05, "loss": 0.0171, "step": 188670 }, { "grad_norm": 0.16682705283164978, "learning_rate": 3.315180994744769e-05, "loss": 0.015, "step": 188680 }, { "grad_norm": 0.3505641520023346, "learning_rate": 3.314662081335427e-05, "loss": 0.0135, "step": 188690 }, { "grad_norm": 0.07988253980875015, "learning_rate": 3.3141431884045304e-05, "loss": 0.0214, "step": 188700 }, { "grad_norm": 0.06635773181915283, "learning_rate": 3.313624315958386e-05, "loss": 0.0138, "step": 188710 }, { "grad_norm": 0.14021362364292145, "learning_rate": 3.3131054640033e-05, "loss": 0.0144, "step": 188720 }, { "grad_norm": 0.054054033011198044, "learning_rate": 3.3125866325455735e-05, "loss": 0.0171, "step": 188730 }, { "grad_norm": 0.170057013630867, "learning_rate": 3.312067821591513e-05, "loss": 0.0179, "step": 188740 }, { "grad_norm": 0.09620226174592972, "learning_rate": 3.311549031147422e-05, "loss": 0.0195, "step": 188750 }, { "grad_norm": 0.08381272852420807, "learning_rate": 3.3110302612196064e-05, "loss": 0.0127, "step": 188760 }, { "grad_norm": 0.08645889163017273, "learning_rate": 3.310511511814365e-05, "loss": 0.015, "step": 188770 }, { "grad_norm": 0.09437496960163116, "learning_rate": 3.3099927829380054e-05, "loss": 0.0152, "step": 188780 }, { "grad_norm": 0.08670583367347717, "learning_rate": 3.30947407459683e-05, "loss": 0.0167, "step": 188790 }, { "grad_norm": 0.061215467751026154, "learning_rate": 3.308955386797139e-05, "loss": 0.0116, "step": 188800 }, { "grad_norm": 0.2869122624397278, "learning_rate": 3.308436719545237e-05, "loss": 0.0135, "step": 188810 }, { "grad_norm": 0.263451486825943, "learning_rate": 3.307918072847425e-05, "loss": 0.0164, "step": 188820 }, { "grad_norm": 0.07016804814338684, "learning_rate": 3.307399446710009e-05, "loss": 0.0171, "step": 188830 }, { "grad_norm": 0.11015737056732178, "learning_rate": 3.306880841139286e-05, "loss": 0.0125, "step": 188840 }, { "grad_norm": 0.1105053499341011, "learning_rate": 3.3063622561415585e-05, "loss": 0.0149, "step": 188850 }, { "grad_norm": 0.040567368268966675, "learning_rate": 3.3058436917231304e-05, "loss": 0.0116, "step": 188860 }, { "grad_norm": 0.156538188457489, "learning_rate": 3.3053251478903e-05, "loss": 0.0135, "step": 188870 }, { "grad_norm": 0.09400134533643723, "learning_rate": 3.304806624649368e-05, "loss": 0.0125, "step": 188880 }, { "grad_norm": 0.13672228157520294, "learning_rate": 3.3042881220066366e-05, "loss": 0.0168, "step": 188890 }, { "grad_norm": 0.1257491260766983, "learning_rate": 3.303769639968408e-05, "loss": 0.0183, "step": 188900 }, { "grad_norm": 0.05513109266757965, "learning_rate": 3.303251178540978e-05, "loss": 0.0102, "step": 188910 }, { "grad_norm": 0.11341026425361633, "learning_rate": 3.302732737730647e-05, "loss": 0.0098, "step": 188920 }, { "grad_norm": 0.09137076139450073, "learning_rate": 3.302214317543716e-05, "loss": 0.0134, "step": 188930 }, { "grad_norm": 0.08214081823825836, "learning_rate": 3.301695917986487e-05, "loss": 0.0093, "step": 188940 }, { "grad_norm": 0.06665825098752975, "learning_rate": 3.301177539065253e-05, "loss": 0.0098, "step": 188950 }, { "grad_norm": 0.06160819157958031, "learning_rate": 3.300659180786316e-05, "loss": 0.0161, "step": 188960 }, { "grad_norm": 0.05625885725021362, "learning_rate": 3.3001408431559764e-05, "loss": 0.0167, "step": 188970 }, { "grad_norm": 0.07541772723197937, "learning_rate": 3.299622526180529e-05, "loss": 0.0159, "step": 188980 }, { "grad_norm": 0.09199530631303787, "learning_rate": 3.299104229866273e-05, "loss": 0.0143, "step": 188990 }, { "grad_norm": 0.10170910507440567, "learning_rate": 3.2985859542195066e-05, "loss": 0.0112, "step": 189000 }, { "grad_norm": 0.0767320990562439, "learning_rate": 3.2980676992465294e-05, "loss": 0.013, "step": 189010 }, { "grad_norm": 0.0588814951479435, "learning_rate": 3.297549464953634e-05, "loss": 0.0106, "step": 189020 }, { "grad_norm": 0.09770244359970093, "learning_rate": 3.29703125134712e-05, "loss": 0.0119, "step": 189030 }, { "grad_norm": 0.061267051845788956, "learning_rate": 3.296513058433287e-05, "loss": 0.0133, "step": 189040 }, { "grad_norm": 0.1000460833311081, "learning_rate": 3.295994886218427e-05, "loss": 0.008, "step": 189050 }, { "grad_norm": 0.05395670235157013, "learning_rate": 3.295476734708837e-05, "loss": 0.0134, "step": 189060 }, { "grad_norm": 0.10980275273323059, "learning_rate": 3.294958603910814e-05, "loss": 0.0142, "step": 189070 }, { "grad_norm": 0.12224497646093369, "learning_rate": 3.2944404938306555e-05, "loss": 0.0103, "step": 189080 }, { "grad_norm": 0.09280701726675034, "learning_rate": 3.293922404474654e-05, "loss": 0.0142, "step": 189090 }, { "grad_norm": 0.18412083387374878, "learning_rate": 3.293404335849106e-05, "loss": 0.016, "step": 189100 }, { "grad_norm": 0.09328268468379974, "learning_rate": 3.2928862879603084e-05, "loss": 0.016, "step": 189110 }, { "grad_norm": 0.06330840289592743, "learning_rate": 3.292368260814552e-05, "loss": 0.0167, "step": 189120 }, { "grad_norm": 0.07257387787103653, "learning_rate": 3.291850254418134e-05, "loss": 0.0129, "step": 189130 }, { "grad_norm": 0.07694470137357712, "learning_rate": 3.2913322687773476e-05, "loss": 0.0135, "step": 189140 }, { "grad_norm": 0.057695068418979645, "learning_rate": 3.29081430389849e-05, "loss": 0.0107, "step": 189150 }, { "grad_norm": 0.09091401845216751, "learning_rate": 3.29029635978785e-05, "loss": 0.0147, "step": 189160 }, { "grad_norm": 0.07087284326553345, "learning_rate": 3.289778436451724e-05, "loss": 0.0117, "step": 189170 }, { "grad_norm": 0.07014832645654678, "learning_rate": 3.2892605338964043e-05, "loss": 0.0143, "step": 189180 }, { "grad_norm": 0.08794008195400238, "learning_rate": 3.288742652128186e-05, "loss": 0.0155, "step": 189190 }, { "grad_norm": 0.13773450255393982, "learning_rate": 3.288224791153358e-05, "loss": 0.014, "step": 189200 }, { "grad_norm": 0.08999897539615631, "learning_rate": 3.287706950978216e-05, "loss": 0.0101, "step": 189210 }, { "grad_norm": 0.1294410675764084, "learning_rate": 3.287189131609051e-05, "loss": 0.0126, "step": 189220 }, { "grad_norm": 0.08806441724300385, "learning_rate": 3.286671333052156e-05, "loss": 0.0107, "step": 189230 }, { "grad_norm": 0.07904268801212311, "learning_rate": 3.286153555313821e-05, "loss": 0.0121, "step": 189240 }, { "grad_norm": 0.2004845142364502, "learning_rate": 3.285635798400338e-05, "loss": 0.0154, "step": 189250 }, { "grad_norm": 0.11464407294988632, "learning_rate": 3.285118062318e-05, "loss": 0.0125, "step": 189260 }, { "grad_norm": 0.12523239850997925, "learning_rate": 3.284600347073097e-05, "loss": 0.0192, "step": 189270 }, { "grad_norm": 0.0890258252620697, "learning_rate": 3.2840826526719175e-05, "loss": 0.013, "step": 189280 }, { "grad_norm": 0.0688433051109314, "learning_rate": 3.2835649791207555e-05, "loss": 0.0134, "step": 189290 }, { "grad_norm": 0.3971644937992096, "learning_rate": 3.283047326425898e-05, "loss": 0.0129, "step": 189300 }, { "grad_norm": 0.05764397233724594, "learning_rate": 3.282529694593638e-05, "loss": 0.0093, "step": 189310 }, { "grad_norm": 0.08405497670173645, "learning_rate": 3.2820120836302635e-05, "loss": 0.0133, "step": 189320 }, { "grad_norm": 0.07329435646533966, "learning_rate": 3.281494493542064e-05, "loss": 0.0141, "step": 189330 }, { "grad_norm": 0.09990441054105759, "learning_rate": 3.2809769243353285e-05, "loss": 0.0125, "step": 189340 }, { "grad_norm": 0.27857205271720886, "learning_rate": 3.280459376016347e-05, "loss": 0.0125, "step": 189350 }, { "grad_norm": 0.09689328074455261, "learning_rate": 3.279941848591408e-05, "loss": 0.0128, "step": 189360 }, { "grad_norm": 0.1355282962322235, "learning_rate": 3.2794243420667986e-05, "loss": 0.0156, "step": 189370 }, { "grad_norm": 0.07945053279399872, "learning_rate": 3.278906856448808e-05, "loss": 0.0141, "step": 189380 }, { "grad_norm": 0.09334901720285416, "learning_rate": 3.2783893917437246e-05, "loss": 0.0149, "step": 189390 }, { "grad_norm": 0.22638705372810364, "learning_rate": 3.2778719479578356e-05, "loss": 0.0148, "step": 189400 }, { "grad_norm": 0.0699748694896698, "learning_rate": 3.277354525097428e-05, "loss": 0.014, "step": 189410 }, { "grad_norm": 0.041244398802518845, "learning_rate": 3.276837123168789e-05, "loss": 0.0139, "step": 189420 }, { "grad_norm": 0.10224953293800354, "learning_rate": 3.276319742178208e-05, "loss": 0.0116, "step": 189430 }, { "grad_norm": 0.06786640733480453, "learning_rate": 3.275802382131967e-05, "loss": 0.0118, "step": 189440 }, { "grad_norm": 0.17918603122234344, "learning_rate": 3.2752850430363555e-05, "loss": 0.0155, "step": 189450 }, { "grad_norm": 0.07637445628643036, "learning_rate": 3.2747677248976595e-05, "loss": 0.0106, "step": 189460 }, { "grad_norm": 0.057429008185863495, "learning_rate": 3.274250427722165e-05, "loss": 0.016, "step": 189470 }, { "grad_norm": 0.08128145337104797, "learning_rate": 3.273733151516157e-05, "loss": 0.0135, "step": 189480 }, { "grad_norm": 0.11192499101161957, "learning_rate": 3.2732158962859204e-05, "loss": 0.0125, "step": 189490 }, { "grad_norm": 0.087766632437706, "learning_rate": 3.272698662037743e-05, "loss": 0.0129, "step": 189500 }, { "grad_norm": 0.06603274494409561, "learning_rate": 3.2721814487779055e-05, "loss": 0.0159, "step": 189510 }, { "grad_norm": 0.14982755482196808, "learning_rate": 3.2716642565126944e-05, "loss": 0.0105, "step": 189520 }, { "grad_norm": 0.0901491641998291, "learning_rate": 3.271147085248395e-05, "loss": 0.0154, "step": 189530 }, { "grad_norm": 0.08935833722352982, "learning_rate": 3.270629934991293e-05, "loss": 0.0099, "step": 189540 }, { "grad_norm": 0.08299906551837921, "learning_rate": 3.2701128057476675e-05, "loss": 0.0178, "step": 189550 }, { "grad_norm": 0.07047022879123688, "learning_rate": 3.269595697523805e-05, "loss": 0.0109, "step": 189560 }, { "grad_norm": 0.2385275661945343, "learning_rate": 3.2690786103259886e-05, "loss": 0.017, "step": 189570 }, { "grad_norm": 0.10061722248792648, "learning_rate": 3.268561544160503e-05, "loss": 0.0194, "step": 189580 }, { "grad_norm": 0.08458875119686127, "learning_rate": 3.268044499033628e-05, "loss": 0.0109, "step": 189590 }, { "grad_norm": 0.08769010007381439, "learning_rate": 3.2675274749516474e-05, "loss": 0.0076, "step": 189600 }, { "grad_norm": 0.07586000859737396, "learning_rate": 3.267010471920845e-05, "loss": 0.0115, "step": 189610 }, { "grad_norm": 0.055203430354595184, "learning_rate": 3.2664934899475006e-05, "loss": 0.011, "step": 189620 }, { "grad_norm": 0.10080733150243759, "learning_rate": 3.265976529037896e-05, "loss": 0.0144, "step": 189630 }, { "grad_norm": 0.07066699117422104, "learning_rate": 3.265459589198314e-05, "loss": 0.0119, "step": 189640 }, { "grad_norm": 0.05359729006886482, "learning_rate": 3.264942670435038e-05, "loss": 0.0118, "step": 189650 }, { "grad_norm": 0.10785553604364395, "learning_rate": 3.264425772754345e-05, "loss": 0.0102, "step": 189660 }, { "grad_norm": 0.11438971012830734, "learning_rate": 3.2639088961625166e-05, "loss": 0.0168, "step": 189670 }, { "grad_norm": 0.05237286165356636, "learning_rate": 3.263392040665838e-05, "loss": 0.0201, "step": 189680 }, { "grad_norm": 0.115660160779953, "learning_rate": 3.262875206270582e-05, "loss": 0.0145, "step": 189690 }, { "grad_norm": 0.09681818634271622, "learning_rate": 3.262358392983033e-05, "loss": 0.0269, "step": 189700 }, { "grad_norm": 0.4368288516998291, "learning_rate": 3.26184160080947e-05, "loss": 0.0172, "step": 189710 }, { "grad_norm": 0.11358924955129623, "learning_rate": 3.261324829756175e-05, "loss": 0.0169, "step": 189720 }, { "grad_norm": 0.04093506187200546, "learning_rate": 3.260808079829422e-05, "loss": 0.0105, "step": 189730 }, { "grad_norm": 0.5188801288604736, "learning_rate": 3.2602913510354937e-05, "loss": 0.019, "step": 189740 }, { "grad_norm": 0.15588752925395966, "learning_rate": 3.25977464338067e-05, "loss": 0.0124, "step": 189750 }, { "grad_norm": 0.07800319045782089, "learning_rate": 3.259257956871225e-05, "loss": 0.012, "step": 189760 }, { "grad_norm": 0.04888299107551575, "learning_rate": 3.2587412915134395e-05, "loss": 0.0149, "step": 189770 }, { "grad_norm": 0.18218660354614258, "learning_rate": 3.2582246473135915e-05, "loss": 0.0099, "step": 189780 }, { "grad_norm": 0.0870354026556015, "learning_rate": 3.257708024277961e-05, "loss": 0.0093, "step": 189790 }, { "grad_norm": 0.14353205263614655, "learning_rate": 3.2571914224128195e-05, "loss": 0.0202, "step": 189800 }, { "grad_norm": 0.07353948056697845, "learning_rate": 3.256674841724449e-05, "loss": 0.0123, "step": 189810 }, { "grad_norm": 0.28427308797836304, "learning_rate": 3.2561582822191274e-05, "loss": 0.0195, "step": 189820 }, { "grad_norm": 0.10264795273542404, "learning_rate": 3.255641743903126e-05, "loss": 0.0138, "step": 189830 }, { "grad_norm": 0.1293007880449295, "learning_rate": 3.255125226782725e-05, "loss": 0.02, "step": 189840 }, { "grad_norm": 0.13896605372428894, "learning_rate": 3.2546087308641995e-05, "loss": 0.019, "step": 189850 }, { "grad_norm": 0.12329131364822388, "learning_rate": 3.254092256153829e-05, "loss": 0.0136, "step": 189860 }, { "grad_norm": 0.08950083702802658, "learning_rate": 3.253575802657883e-05, "loss": 0.0119, "step": 189870 }, { "grad_norm": 0.07192225754261017, "learning_rate": 3.253059370382639e-05, "loss": 0.0093, "step": 189880 }, { "grad_norm": 0.09648063778877258, "learning_rate": 3.2525429593343745e-05, "loss": 0.0121, "step": 189890 }, { "grad_norm": 0.08960802108049393, "learning_rate": 3.252026569519365e-05, "loss": 0.0166, "step": 189900 }, { "grad_norm": 0.18577630817890167, "learning_rate": 3.25151020094388e-05, "loss": 0.0102, "step": 189910 }, { "grad_norm": 0.13880357146263123, "learning_rate": 3.250993853614198e-05, "loss": 0.013, "step": 189920 }, { "grad_norm": 0.09379233419895172, "learning_rate": 3.250477527536593e-05, "loss": 0.0128, "step": 189930 }, { "grad_norm": 0.13656450808048248, "learning_rate": 3.249961222717337e-05, "loss": 0.0164, "step": 189940 }, { "grad_norm": 0.16223999857902527, "learning_rate": 3.249444939162704e-05, "loss": 0.0136, "step": 189950 }, { "grad_norm": 0.08155100047588348, "learning_rate": 3.248928676878967e-05, "loss": 0.0159, "step": 189960 }, { "grad_norm": 0.07901652902364731, "learning_rate": 3.2484124358724025e-05, "loss": 0.0116, "step": 189970 }, { "grad_norm": 0.3721226751804352, "learning_rate": 3.247896216149279e-05, "loss": 0.0191, "step": 189980 }, { "grad_norm": 0.10102632641792297, "learning_rate": 3.2473800177158703e-05, "loss": 0.0106, "step": 189990 }, { "grad_norm": 0.18056859076023102, "learning_rate": 3.24686384057845e-05, "loss": 0.0139, "step": 190000 }, { "grad_norm": 0.08081318438053131, "learning_rate": 3.246347684743288e-05, "loss": 0.0267, "step": 190010 }, { "grad_norm": 0.08696623146533966, "learning_rate": 3.2458315502166595e-05, "loss": 0.0135, "step": 190020 }, { "grad_norm": 0.21039097011089325, "learning_rate": 3.245315437004832e-05, "loss": 0.0173, "step": 190030 }, { "grad_norm": 0.25608712434768677, "learning_rate": 3.24479934511408e-05, "loss": 0.0145, "step": 190040 }, { "grad_norm": 0.07685185223817825, "learning_rate": 3.2442832745506726e-05, "loss": 0.0152, "step": 190050 }, { "grad_norm": 0.11848994344472885, "learning_rate": 3.24376722532088e-05, "loss": 0.0145, "step": 190060 }, { "grad_norm": 0.10117973387241364, "learning_rate": 3.2432511974309766e-05, "loss": 0.0098, "step": 190070 }, { "grad_norm": 0.20684674382209778, "learning_rate": 3.2427351908872275e-05, "loss": 0.0115, "step": 190080 }, { "grad_norm": 0.1598733365535736, "learning_rate": 3.242219205695904e-05, "loss": 0.0159, "step": 190090 }, { "grad_norm": 0.08562607318162918, "learning_rate": 3.2417032418632786e-05, "loss": 0.0097, "step": 190100 }, { "grad_norm": 0.08671686053276062, "learning_rate": 3.2411872993956196e-05, "loss": 0.0119, "step": 190110 }, { "grad_norm": 0.12096557021141052, "learning_rate": 3.2406713782991935e-05, "loss": 0.0179, "step": 190120 }, { "grad_norm": 0.1410549432039261, "learning_rate": 3.240155478580272e-05, "loss": 0.0133, "step": 190130 }, { "grad_norm": 0.19545839726924896, "learning_rate": 3.2396396002451235e-05, "loss": 0.0179, "step": 190140 }, { "grad_norm": 0.10172299295663834, "learning_rate": 3.239123743300016e-05, "loss": 0.0118, "step": 190150 }, { "grad_norm": 0.12407368421554565, "learning_rate": 3.238607907751217e-05, "loss": 0.0116, "step": 190160 }, { "grad_norm": 0.1223587691783905, "learning_rate": 3.238092093604994e-05, "loss": 0.0123, "step": 190170 }, { "grad_norm": 0.06499360501766205, "learning_rate": 3.237576300867619e-05, "loss": 0.0137, "step": 190180 }, { "grad_norm": 0.2221072018146515, "learning_rate": 3.237060529545353e-05, "loss": 0.0162, "step": 190190 }, { "grad_norm": 0.08915414661169052, "learning_rate": 3.236544779644466e-05, "loss": 0.011, "step": 190200 }, { "grad_norm": 0.09881598502397537, "learning_rate": 3.236029051171224e-05, "loss": 0.0106, "step": 190210 }, { "grad_norm": 0.11377615481615067, "learning_rate": 3.235513344131899e-05, "loss": 0.0112, "step": 190220 }, { "grad_norm": 0.11242452263832092, "learning_rate": 3.234997658532748e-05, "loss": 0.0163, "step": 190230 }, { "grad_norm": 0.13876844942569733, "learning_rate": 3.234481994380044e-05, "loss": 0.0137, "step": 190240 }, { "grad_norm": 0.16598673164844513, "learning_rate": 3.2339663516800514e-05, "loss": 0.0106, "step": 190250 }, { "grad_norm": 0.09448165446519852, "learning_rate": 3.2334507304390335e-05, "loss": 0.01, "step": 190260 }, { "grad_norm": 0.1541796177625656, "learning_rate": 3.2329351306632576e-05, "loss": 0.0138, "step": 190270 }, { "grad_norm": 0.0664118304848671, "learning_rate": 3.232419552358988e-05, "loss": 0.0118, "step": 190280 }, { "grad_norm": 0.062223877757787704, "learning_rate": 3.231903995532492e-05, "loss": 0.0095, "step": 190290 }, { "grad_norm": 0.08384998142719269, "learning_rate": 3.231388460190029e-05, "loss": 0.0115, "step": 190300 }, { "grad_norm": 0.054020803421735764, "learning_rate": 3.230872946337866e-05, "loss": 0.0121, "step": 190310 }, { "grad_norm": 0.07202015072107315, "learning_rate": 3.23035745398227e-05, "loss": 0.013, "step": 190320 }, { "grad_norm": 0.06306794285774231, "learning_rate": 3.229841983129499e-05, "loss": 0.0095, "step": 190330 }, { "grad_norm": 0.09968280792236328, "learning_rate": 3.229326533785819e-05, "loss": 0.0171, "step": 190340 }, { "grad_norm": 0.06275980919599533, "learning_rate": 3.228811105957494e-05, "loss": 0.0111, "step": 190350 }, { "grad_norm": 0.07380688935518265, "learning_rate": 3.2282956996507885e-05, "loss": 0.012, "step": 190360 }, { "grad_norm": 0.11325013637542725, "learning_rate": 3.227780314871961e-05, "loss": 0.0146, "step": 190370 }, { "grad_norm": 0.28690215945243835, "learning_rate": 3.227264951627275e-05, "loss": 0.0125, "step": 190380 }, { "grad_norm": 0.1098400354385376, "learning_rate": 3.226749609922997e-05, "loss": 0.0103, "step": 190390 }, { "grad_norm": 0.09215099364519119, "learning_rate": 3.226234289765382e-05, "loss": 0.0124, "step": 190400 }, { "grad_norm": 0.14499931037425995, "learning_rate": 3.2257189911606955e-05, "loss": 0.0121, "step": 190410 }, { "grad_norm": 0.10270806401968002, "learning_rate": 3.225203714115199e-05, "loss": 0.0166, "step": 190420 }, { "grad_norm": 0.0672360211610794, "learning_rate": 3.224688458635155e-05, "loss": 0.012, "step": 190430 }, { "grad_norm": 0.14336225390434265, "learning_rate": 3.22417322472682e-05, "loss": 0.0134, "step": 190440 }, { "grad_norm": 0.06293465197086334, "learning_rate": 3.2236580123964574e-05, "loss": 0.0131, "step": 190450 }, { "grad_norm": 0.08263921737670898, "learning_rate": 3.223142821650327e-05, "loss": 0.0132, "step": 190460 }, { "grad_norm": 0.15304957330226898, "learning_rate": 3.222627652494691e-05, "loss": 0.0134, "step": 190470 }, { "grad_norm": 0.14689645171165466, "learning_rate": 3.222112504935805e-05, "loss": 0.0166, "step": 190480 }, { "grad_norm": 0.1087551936507225, "learning_rate": 3.22159737897993e-05, "loss": 0.0142, "step": 190490 }, { "grad_norm": 0.32580018043518066, "learning_rate": 3.2210822746333286e-05, "loss": 0.0147, "step": 190500 }, { "grad_norm": 0.18463356792926788, "learning_rate": 3.220567191902255e-05, "loss": 0.011, "step": 190510 }, { "grad_norm": 0.1048242449760437, "learning_rate": 3.2200521307929695e-05, "loss": 0.0178, "step": 190520 }, { "grad_norm": 0.1071142628788948, "learning_rate": 3.219537091311733e-05, "loss": 0.0127, "step": 190530 }, { "grad_norm": 0.06504739820957184, "learning_rate": 3.219022073464802e-05, "loss": 0.0143, "step": 190540 }, { "grad_norm": 0.11850254982709885, "learning_rate": 3.218507077258434e-05, "loss": 0.0102, "step": 190550 }, { "grad_norm": 0.29234325885772705, "learning_rate": 3.217992102698885e-05, "loss": 0.0124, "step": 190560 }, { "grad_norm": 0.10860617458820343, "learning_rate": 3.217477149792418e-05, "loss": 0.0182, "step": 190570 }, { "grad_norm": 0.10864121466875076, "learning_rate": 3.216962218545284e-05, "loss": 0.0183, "step": 190580 }, { "grad_norm": 0.13939473032951355, "learning_rate": 3.2164473089637424e-05, "loss": 0.0169, "step": 190590 }, { "grad_norm": 0.09130599349737167, "learning_rate": 3.21593242105405e-05, "loss": 0.0123, "step": 190600 }, { "grad_norm": 0.07029204815626144, "learning_rate": 3.215417554822465e-05, "loss": 0.0092, "step": 190610 }, { "grad_norm": 0.1007460355758667, "learning_rate": 3.21490271027524e-05, "loss": 0.0121, "step": 190620 }, { "grad_norm": 0.07872121036052704, "learning_rate": 3.2143878874186325e-05, "loss": 0.0121, "step": 190630 }, { "grad_norm": 0.07964808493852615, "learning_rate": 3.213873086258899e-05, "loss": 0.0076, "step": 190640 }, { "grad_norm": 0.07302787899971008, "learning_rate": 3.2133583068022946e-05, "loss": 0.0135, "step": 190650 }, { "grad_norm": 0.09454663842916489, "learning_rate": 3.212843549055071e-05, "loss": 0.0118, "step": 190660 }, { "grad_norm": 0.1822887510061264, "learning_rate": 3.212328813023487e-05, "loss": 0.0171, "step": 190670 }, { "grad_norm": 0.07322009652853012, "learning_rate": 3.211814098713796e-05, "loss": 0.0122, "step": 190680 }, { "grad_norm": 0.06399980932474136, "learning_rate": 3.211299406132253e-05, "loss": 0.0132, "step": 190690 }, { "grad_norm": 0.10862711817026138, "learning_rate": 3.210784735285109e-05, "loss": 0.0124, "step": 190700 }, { "grad_norm": 0.06530331075191498, "learning_rate": 3.210270086178621e-05, "loss": 0.0157, "step": 190710 }, { "grad_norm": 0.06289958953857422, "learning_rate": 3.20975545881904e-05, "loss": 0.0126, "step": 190720 }, { "grad_norm": 0.06779484450817108, "learning_rate": 3.209240853212622e-05, "loss": 0.0111, "step": 190730 }, { "grad_norm": 0.08037640154361725, "learning_rate": 3.2087262693656175e-05, "loss": 0.012, "step": 190740 }, { "grad_norm": 0.09647397696971893, "learning_rate": 3.208211707284281e-05, "loss": 0.0096, "step": 190750 }, { "grad_norm": 0.1654861718416214, "learning_rate": 3.2076971669748625e-05, "loss": 0.0152, "step": 190760 }, { "grad_norm": 0.14355459809303284, "learning_rate": 3.2071826484436165e-05, "loss": 0.0145, "step": 190770 }, { "grad_norm": 0.10387977212667465, "learning_rate": 3.2066681516967935e-05, "loss": 0.0119, "step": 190780 }, { "grad_norm": 0.05060248449444771, "learning_rate": 3.2061536767406466e-05, "loss": 0.0143, "step": 190790 }, { "grad_norm": 0.08906219154596329, "learning_rate": 3.2056392235814255e-05, "loss": 0.0151, "step": 190800 }, { "grad_norm": 0.08604512363672256, "learning_rate": 3.205124792225382e-05, "loss": 0.0093, "step": 190810 }, { "grad_norm": 0.15392877161502838, "learning_rate": 3.2046103826787674e-05, "loss": 0.0145, "step": 190820 }, { "grad_norm": 0.17454612255096436, "learning_rate": 3.204095994947831e-05, "loss": 0.0102, "step": 190830 }, { "grad_norm": 0.0751076266169548, "learning_rate": 3.2035816290388244e-05, "loss": 0.0111, "step": 190840 }, { "grad_norm": 0.14176112413406372, "learning_rate": 3.203067284957997e-05, "loss": 0.0155, "step": 190850 }, { "grad_norm": 0.08172067999839783, "learning_rate": 3.2025529627115994e-05, "loss": 0.0138, "step": 190860 }, { "grad_norm": 0.21736736595630646, "learning_rate": 3.20203866230588e-05, "loss": 0.0189, "step": 190870 }, { "grad_norm": 0.2730112671852112, "learning_rate": 3.201524383747086e-05, "loss": 0.0127, "step": 190880 }, { "grad_norm": 0.22734922170639038, "learning_rate": 3.2010101270414736e-05, "loss": 0.0164, "step": 190890 }, { "grad_norm": 0.14627934992313385, "learning_rate": 3.200495892195283e-05, "loss": 0.0096, "step": 190900 }, { "grad_norm": 0.07333184778690338, "learning_rate": 3.199981679214767e-05, "loss": 0.0143, "step": 190910 }, { "grad_norm": 0.06568080186843872, "learning_rate": 3.1994674881061726e-05, "loss": 0.0116, "step": 190920 }, { "grad_norm": 0.06908056139945984, "learning_rate": 3.1989533188757504e-05, "loss": 0.0155, "step": 190930 }, { "grad_norm": 0.0692249983549118, "learning_rate": 3.1984391715297444e-05, "loss": 0.0163, "step": 190940 }, { "grad_norm": 0.049253374338150024, "learning_rate": 3.197925046074402e-05, "loss": 0.0111, "step": 190950 }, { "grad_norm": 0.08782292157411575, "learning_rate": 3.197410942515975e-05, "loss": 0.0105, "step": 190960 }, { "grad_norm": 0.10512223094701767, "learning_rate": 3.1968968608607055e-05, "loss": 0.012, "step": 190970 }, { "grad_norm": 0.07819406688213348, "learning_rate": 3.19638280111484e-05, "loss": 0.01, "step": 190980 }, { "grad_norm": 0.13547492027282715, "learning_rate": 3.1958687632846274e-05, "loss": 0.0139, "step": 190990 }, { "grad_norm": 0.13735227286815643, "learning_rate": 3.195354747376314e-05, "loss": 0.0149, "step": 191000 }, { "grad_norm": 0.09853493422269821, "learning_rate": 3.1948407533961425e-05, "loss": 0.0136, "step": 191010 }, { "grad_norm": 0.163326233625412, "learning_rate": 3.194326781350361e-05, "loss": 0.0108, "step": 191020 }, { "grad_norm": 0.07800911366939545, "learning_rate": 3.1938128312452146e-05, "loss": 0.0124, "step": 191030 }, { "grad_norm": 0.08858387917280197, "learning_rate": 3.193298903086946e-05, "loss": 0.0173, "step": 191040 }, { "grad_norm": 0.11880091577768326, "learning_rate": 3.192784996881803e-05, "loss": 0.0135, "step": 191050 }, { "grad_norm": 0.06785809993743896, "learning_rate": 3.192271112636027e-05, "loss": 0.0099, "step": 191060 }, { "grad_norm": 0.11125101149082184, "learning_rate": 3.191757250355866e-05, "loss": 0.0169, "step": 191070 }, { "grad_norm": 0.07081549614667892, "learning_rate": 3.19124341004756e-05, "loss": 0.0119, "step": 191080 }, { "grad_norm": 0.09253812581300735, "learning_rate": 3.190729591717354e-05, "loss": 0.0142, "step": 191090 }, { "grad_norm": 0.08207842707633972, "learning_rate": 3.1902157953714914e-05, "loss": 0.0094, "step": 191100 }, { "grad_norm": 0.13916034996509552, "learning_rate": 3.189702021016219e-05, "loss": 0.0113, "step": 191110 }, { "grad_norm": 0.05520579218864441, "learning_rate": 3.189188268657773e-05, "loss": 0.0152, "step": 191120 }, { "grad_norm": 0.04959767311811447, "learning_rate": 3.1886745383024e-05, "loss": 0.0165, "step": 191130 }, { "grad_norm": 0.0651058629155159, "learning_rate": 3.1881608299563426e-05, "loss": 0.0111, "step": 191140 }, { "grad_norm": 0.07486053556203842, "learning_rate": 3.18764714362584e-05, "loss": 0.015, "step": 191150 }, { "grad_norm": 0.16459476947784424, "learning_rate": 3.1871334793171356e-05, "loss": 0.0123, "step": 191160 }, { "grad_norm": 0.17881333827972412, "learning_rate": 3.186619837036472e-05, "loss": 0.0113, "step": 191170 }, { "grad_norm": 0.06190909817814827, "learning_rate": 3.186106216790091e-05, "loss": 0.0124, "step": 191180 }, { "grad_norm": 0.28220629692077637, "learning_rate": 3.18559261858423e-05, "loss": 0.0103, "step": 191190 }, { "grad_norm": 0.06476398557424545, "learning_rate": 3.185079042425132e-05, "loss": 0.0096, "step": 191200 }, { "grad_norm": 0.19519785046577454, "learning_rate": 3.184565488319039e-05, "loss": 0.0108, "step": 191210 }, { "grad_norm": 0.08381586521863937, "learning_rate": 3.184051956272187e-05, "loss": 0.0155, "step": 191220 }, { "grad_norm": 0.11666611582040787, "learning_rate": 3.183538446290819e-05, "loss": 0.0133, "step": 191230 }, { "grad_norm": 0.05385403707623482, "learning_rate": 3.183024958381173e-05, "loss": 0.013, "step": 191240 }, { "grad_norm": 0.09528185427188873, "learning_rate": 3.182511492549492e-05, "loss": 0.0115, "step": 191250 }, { "grad_norm": 0.1953740268945694, "learning_rate": 3.181998048802011e-05, "loss": 0.0156, "step": 191260 }, { "grad_norm": 0.06857625395059586, "learning_rate": 3.1814846271449684e-05, "loss": 0.0123, "step": 191270 }, { "grad_norm": 0.4091016948223114, "learning_rate": 3.180971227584608e-05, "loss": 0.0115, "step": 191280 }, { "grad_norm": 0.10669361054897308, "learning_rate": 3.180457850127162e-05, "loss": 0.014, "step": 191290 }, { "grad_norm": 0.053218621760606766, "learning_rate": 3.1799444947788715e-05, "loss": 0.0132, "step": 191300 }, { "grad_norm": 0.10289230942726135, "learning_rate": 3.1794311615459735e-05, "loss": 0.0177, "step": 191310 }, { "grad_norm": 0.07454270869493484, "learning_rate": 3.178917850434707e-05, "loss": 0.0133, "step": 191320 }, { "grad_norm": 0.14484120905399323, "learning_rate": 3.178404561451307e-05, "loss": 0.0163, "step": 191330 }, { "grad_norm": 0.05854416266083717, "learning_rate": 3.177891294602011e-05, "loss": 0.0144, "step": 191340 }, { "grad_norm": 0.11864319443702698, "learning_rate": 3.1773780498930564e-05, "loss": 0.0136, "step": 191350 }, { "grad_norm": 0.12133611738681793, "learning_rate": 3.176864827330682e-05, "loss": 0.0147, "step": 191360 }, { "grad_norm": 0.0850478857755661, "learning_rate": 3.176351626921118e-05, "loss": 0.0085, "step": 191370 }, { "grad_norm": 0.050997357815504074, "learning_rate": 3.1758384486706047e-05, "loss": 0.0154, "step": 191380 }, { "grad_norm": 0.12905605137348175, "learning_rate": 3.175325292585376e-05, "loss": 0.0141, "step": 191390 }, { "grad_norm": 0.09023205935955048, "learning_rate": 3.1748121586716696e-05, "loss": 0.0133, "step": 191400 }, { "grad_norm": 0.12812459468841553, "learning_rate": 3.174299046935716e-05, "loss": 0.0107, "step": 191410 }, { "grad_norm": 0.05186474695801735, "learning_rate": 3.173785957383754e-05, "loss": 0.01, "step": 191420 }, { "grad_norm": 0.14241313934326172, "learning_rate": 3.173272890022018e-05, "loss": 0.0128, "step": 191430 }, { "grad_norm": 0.07007235288619995, "learning_rate": 3.17275984485674e-05, "loss": 0.0132, "step": 191440 }, { "grad_norm": 0.11749093979597092, "learning_rate": 3.172246821894155e-05, "loss": 0.0128, "step": 191450 }, { "grad_norm": 0.10517653077840805, "learning_rate": 3.171733821140497e-05, "loss": 0.0136, "step": 191460 }, { "grad_norm": 0.0566866472363472, "learning_rate": 3.1712208426019984e-05, "loss": 0.0129, "step": 191470 }, { "grad_norm": 0.09424833953380585, "learning_rate": 3.170707886284895e-05, "loss": 0.0165, "step": 191480 }, { "grad_norm": 0.14124296605587006, "learning_rate": 3.1701949521954167e-05, "loss": 0.0121, "step": 191490 }, { "grad_norm": 0.1344970464706421, "learning_rate": 3.169682040339798e-05, "loss": 0.0124, "step": 191500 }, { "grad_norm": 0.0822962149977684, "learning_rate": 3.1691691507242694e-05, "loss": 0.0117, "step": 191510 }, { "grad_norm": 0.13366375863552094, "learning_rate": 3.1686562833550655e-05, "loss": 0.0146, "step": 191520 }, { "grad_norm": 0.11661072075366974, "learning_rate": 3.1681434382384165e-05, "loss": 0.0204, "step": 191530 }, { "grad_norm": 0.06504345685243607, "learning_rate": 3.1676306153805533e-05, "loss": 0.0175, "step": 191540 }, { "grad_norm": 0.08379276096820831, "learning_rate": 3.167117814787709e-05, "loss": 0.0152, "step": 191550 }, { "grad_norm": 0.13052089512348175, "learning_rate": 3.1666050364661136e-05, "loss": 0.0159, "step": 191560 }, { "grad_norm": 0.1091337725520134, "learning_rate": 3.166092280421999e-05, "loss": 0.0113, "step": 191570 }, { "grad_norm": 0.17773044109344482, "learning_rate": 3.165579546661593e-05, "loss": 0.0143, "step": 191580 }, { "grad_norm": 0.27807143330574036, "learning_rate": 3.165066835191128e-05, "loss": 0.0154, "step": 191590 }, { "grad_norm": 0.08175297826528549, "learning_rate": 3.164554146016836e-05, "loss": 0.0105, "step": 191600 }, { "grad_norm": 0.13222751021385193, "learning_rate": 3.164041479144941e-05, "loss": 0.0151, "step": 191610 }, { "grad_norm": 0.2301613986492157, "learning_rate": 3.163528834581676e-05, "loss": 0.0193, "step": 191620 }, { "grad_norm": 0.0871552973985672, "learning_rate": 3.163016212333269e-05, "loss": 0.0124, "step": 191630 }, { "grad_norm": 0.3045822083950043, "learning_rate": 3.1625036124059517e-05, "loss": 0.0154, "step": 191640 }, { "grad_norm": 0.05846729129552841, "learning_rate": 3.161991034805949e-05, "loss": 0.0127, "step": 191650 }, { "grad_norm": 0.18053528666496277, "learning_rate": 3.1614784795394895e-05, "loss": 0.016, "step": 191660 }, { "grad_norm": 0.06657266616821289, "learning_rate": 3.160965946612803e-05, "loss": 0.0104, "step": 191670 }, { "grad_norm": 0.10592621564865112, "learning_rate": 3.160453436032119e-05, "loss": 0.0141, "step": 191680 }, { "grad_norm": 0.10490065068006516, "learning_rate": 3.1599409478036595e-05, "loss": 0.0079, "step": 191690 }, { "grad_norm": 0.06712656468153, "learning_rate": 3.159428481933655e-05, "loss": 0.0094, "step": 191700 }, { "grad_norm": 0.06223522499203682, "learning_rate": 3.1589160384283354e-05, "loss": 0.0278, "step": 191710 }, { "grad_norm": 0.061082012951374054, "learning_rate": 3.158403617293921e-05, "loss": 0.0155, "step": 191720 }, { "grad_norm": 0.058526888489723206, "learning_rate": 3.1578912185366414e-05, "loss": 0.0095, "step": 191730 }, { "grad_norm": 0.1314283013343811, "learning_rate": 3.157378842162724e-05, "loss": 0.0166, "step": 191740 }, { "grad_norm": 0.10339855402708054, "learning_rate": 3.156866488178394e-05, "loss": 0.0125, "step": 191750 }, { "grad_norm": 0.07105425000190735, "learning_rate": 3.1563541565898744e-05, "loss": 0.0126, "step": 191760 }, { "grad_norm": 0.05835302174091339, "learning_rate": 3.1558418474033934e-05, "loss": 0.0113, "step": 191770 }, { "grad_norm": 0.10267894715070724, "learning_rate": 3.1553295606251765e-05, "loss": 0.0169, "step": 191780 }, { "grad_norm": 0.08193469792604446, "learning_rate": 3.1548172962614456e-05, "loss": 0.0122, "step": 191790 }, { "grad_norm": 0.07866290956735611, "learning_rate": 3.154305054318426e-05, "loss": 0.0163, "step": 191800 }, { "grad_norm": 0.062005478888750076, "learning_rate": 3.1537928348023434e-05, "loss": 0.0125, "step": 191810 }, { "grad_norm": 0.08541126549243927, "learning_rate": 3.153280637719422e-05, "loss": 0.0128, "step": 191820 }, { "grad_norm": 0.1304345279932022, "learning_rate": 3.1527684630758835e-05, "loss": 0.0084, "step": 191830 }, { "grad_norm": 0.21101881563663483, "learning_rate": 3.1522563108779516e-05, "loss": 0.0158, "step": 191840 }, { "grad_norm": 0.062137335538864136, "learning_rate": 3.1517441811318524e-05, "loss": 0.0122, "step": 191850 }, { "grad_norm": 0.07646746933460236, "learning_rate": 3.151232073843804e-05, "loss": 0.0151, "step": 191860 }, { "grad_norm": 0.14708806574344635, "learning_rate": 3.150719989020032e-05, "loss": 0.0163, "step": 191870 }, { "grad_norm": 0.14980098605155945, "learning_rate": 3.1502079266667575e-05, "loss": 0.015, "step": 191880 }, { "grad_norm": 0.05713095888495445, "learning_rate": 3.149695886790206e-05, "loss": 0.0142, "step": 191890 }, { "grad_norm": 0.043363407254219055, "learning_rate": 3.1491838693965935e-05, "loss": 0.0179, "step": 191900 }, { "grad_norm": 0.12169207632541656, "learning_rate": 3.148671874492145e-05, "loss": 0.0212, "step": 191910 }, { "grad_norm": 0.0792168453335762, "learning_rate": 3.1481599020830835e-05, "loss": 0.0115, "step": 191920 }, { "grad_norm": 0.15157979726791382, "learning_rate": 3.1476479521756255e-05, "loss": 0.0105, "step": 191930 }, { "grad_norm": 0.09262619912624359, "learning_rate": 3.147136024775993e-05, "loss": 0.0218, "step": 191940 }, { "grad_norm": 0.1062120571732521, "learning_rate": 3.146624119890408e-05, "loss": 0.0133, "step": 191950 }, { "grad_norm": 0.11653832346200943, "learning_rate": 3.146112237525092e-05, "loss": 0.0149, "step": 191960 }, { "grad_norm": 0.08132798224687576, "learning_rate": 3.1456003776862597e-05, "loss": 0.0096, "step": 191970 }, { "grad_norm": 0.08865479379892349, "learning_rate": 3.145088540380134e-05, "loss": 0.012, "step": 191980 }, { "grad_norm": 0.0963347777724266, "learning_rate": 3.1445767256129346e-05, "loss": 0.0108, "step": 191990 }, { "grad_norm": 0.12963750958442688, "learning_rate": 3.14406493339088e-05, "loss": 0.0156, "step": 192000 }, { "grad_norm": 0.04867808520793915, "learning_rate": 3.1435531637201886e-05, "loss": 0.0141, "step": 192010 }, { "grad_norm": 0.0964055135846138, "learning_rate": 3.1430414166070785e-05, "loss": 0.0127, "step": 192020 }, { "grad_norm": 0.16936339437961578, "learning_rate": 3.14252969205777e-05, "loss": 0.0113, "step": 192030 }, { "grad_norm": 0.07080455869436264, "learning_rate": 3.142017990078478e-05, "loss": 0.0147, "step": 192040 }, { "grad_norm": 0.05874938517808914, "learning_rate": 3.1415063106754214e-05, "loss": 0.0145, "step": 192050 }, { "grad_norm": 0.12826654314994812, "learning_rate": 3.140994653854818e-05, "loss": 0.0153, "step": 192060 }, { "grad_norm": 0.1369192898273468, "learning_rate": 3.140483019622887e-05, "loss": 0.0099, "step": 192070 }, { "grad_norm": 0.09524809569120407, "learning_rate": 3.13997140798584e-05, "loss": 0.018, "step": 192080 }, { "grad_norm": 0.18731293082237244, "learning_rate": 3.139459818949898e-05, "loss": 0.0073, "step": 192090 }, { "grad_norm": 0.06021459400653839, "learning_rate": 3.138948252521275e-05, "loss": 0.0101, "step": 192100 }, { "grad_norm": 0.05387602001428604, "learning_rate": 3.1384367087061896e-05, "loss": 0.0127, "step": 192110 }, { "grad_norm": 0.08343812078237534, "learning_rate": 3.137925187510854e-05, "loss": 0.0106, "step": 192120 }, { "grad_norm": 0.05811578035354614, "learning_rate": 3.137413688941485e-05, "loss": 0.0115, "step": 192130 }, { "grad_norm": 0.12459717690944672, "learning_rate": 3.1369022130043e-05, "loss": 0.0138, "step": 192140 }, { "grad_norm": 0.07447143644094467, "learning_rate": 3.136390759705512e-05, "loss": 0.0165, "step": 192150 }, { "grad_norm": 0.1336621642112732, "learning_rate": 3.135879329051335e-05, "loss": 0.0111, "step": 192160 }, { "grad_norm": 0.09857337176799774, "learning_rate": 3.135367921047985e-05, "loss": 0.0108, "step": 192170 }, { "grad_norm": 0.09001373499631882, "learning_rate": 3.134856535701675e-05, "loss": 0.0138, "step": 192180 }, { "grad_norm": 0.04770286753773689, "learning_rate": 3.1343451730186184e-05, "loss": 0.0116, "step": 192190 }, { "grad_norm": 0.09630925208330154, "learning_rate": 3.133833833005029e-05, "loss": 0.0134, "step": 192200 }, { "grad_norm": 0.12806954979896545, "learning_rate": 3.133322515667122e-05, "loss": 0.008, "step": 192210 }, { "grad_norm": 0.0938120111823082, "learning_rate": 3.1328112210111085e-05, "loss": 0.0118, "step": 192220 }, { "grad_norm": 0.05943809822201729, "learning_rate": 3.132299949043202e-05, "loss": 0.0118, "step": 192230 }, { "grad_norm": 0.07694079726934433, "learning_rate": 3.1317886997696136e-05, "loss": 0.0127, "step": 192240 }, { "grad_norm": 0.5614924430847168, "learning_rate": 3.131277473196555e-05, "loss": 0.0193, "step": 192250 }, { "grad_norm": 0.15491046011447906, "learning_rate": 3.130766269330241e-05, "loss": 0.0102, "step": 192260 }, { "grad_norm": 0.08123526722192764, "learning_rate": 3.1302550881768824e-05, "loss": 0.0157, "step": 192270 }, { "grad_norm": 0.07923956215381622, "learning_rate": 3.12974392974269e-05, "loss": 0.0148, "step": 192280 }, { "grad_norm": 0.26607027649879456, "learning_rate": 3.129232794033873e-05, "loss": 0.0128, "step": 192290 }, { "grad_norm": 0.08599235862493515, "learning_rate": 3.128721681056644e-05, "loss": 0.0116, "step": 192300 }, { "grad_norm": 0.06749361753463745, "learning_rate": 3.128210590817214e-05, "loss": 0.0097, "step": 192310 }, { "grad_norm": 0.08719280362129211, "learning_rate": 3.127699523321794e-05, "loss": 0.0134, "step": 192320 }, { "grad_norm": 0.11416628211736679, "learning_rate": 3.1271884785765906e-05, "loss": 0.0133, "step": 192330 }, { "grad_norm": 0.5521237850189209, "learning_rate": 3.1266774565878154e-05, "loss": 0.0143, "step": 192340 }, { "grad_norm": 0.11204732954502106, "learning_rate": 3.12616645736168e-05, "loss": 0.0136, "step": 192350 }, { "grad_norm": 0.11469721049070358, "learning_rate": 3.125655480904389e-05, "loss": 0.0106, "step": 192360 }, { "grad_norm": 0.07724836468696594, "learning_rate": 3.1251445272221536e-05, "loss": 0.0121, "step": 192370 }, { "grad_norm": 0.12118643522262573, "learning_rate": 3.1246335963211825e-05, "loss": 0.014, "step": 192380 }, { "grad_norm": 0.17232224345207214, "learning_rate": 3.1241226882076855e-05, "loss": 0.0128, "step": 192390 }, { "grad_norm": 0.0758100375533104, "learning_rate": 3.1236118028878673e-05, "loss": 0.0113, "step": 192400 }, { "grad_norm": 0.05102860927581787, "learning_rate": 3.123100940367937e-05, "loss": 0.0096, "step": 192410 }, { "grad_norm": 0.07426083087921143, "learning_rate": 3.122590100654105e-05, "loss": 0.0155, "step": 192420 }, { "grad_norm": 0.08123311400413513, "learning_rate": 3.122079283752574e-05, "loss": 0.0115, "step": 192430 }, { "grad_norm": 0.18252629041671753, "learning_rate": 3.1215684896695515e-05, "loss": 0.0158, "step": 192440 }, { "grad_norm": 0.12926550209522247, "learning_rate": 3.121057718411246e-05, "loss": 0.0132, "step": 192450 }, { "grad_norm": 0.07485201954841614, "learning_rate": 3.120546969983866e-05, "loss": 0.0137, "step": 192460 }, { "grad_norm": 0.16183099150657654, "learning_rate": 3.120036244393612e-05, "loss": 0.0157, "step": 192470 }, { "grad_norm": 0.16047431528568268, "learning_rate": 3.119525541646693e-05, "loss": 0.0167, "step": 192480 }, { "grad_norm": 0.05545085296034813, "learning_rate": 3.1190148617493154e-05, "loss": 0.013, "step": 192490 }, { "grad_norm": 0.09201636165380478, "learning_rate": 3.118504204707682e-05, "loss": 0.0106, "step": 192500 }, { "grad_norm": 0.07000643014907837, "learning_rate": 3.117993570527998e-05, "loss": 0.0105, "step": 192510 }, { "grad_norm": 0.058631040155887604, "learning_rate": 3.11748295921647e-05, "loss": 0.012, "step": 192520 }, { "grad_norm": 0.10558649152517319, "learning_rate": 3.116972370779303e-05, "loss": 0.0157, "step": 192530 }, { "grad_norm": 0.2786913514137268, "learning_rate": 3.116461805222698e-05, "loss": 0.014, "step": 192540 }, { "grad_norm": 0.04997656121850014, "learning_rate": 3.11595126255286e-05, "loss": 0.0135, "step": 192550 }, { "grad_norm": 0.07056476920843124, "learning_rate": 3.115440742775993e-05, "loss": 0.0111, "step": 192560 }, { "grad_norm": 0.24132850766181946, "learning_rate": 3.1149302458983026e-05, "loss": 0.0134, "step": 192570 }, { "grad_norm": 0.08266200125217438, "learning_rate": 3.114419771925987e-05, "loss": 0.0148, "step": 192580 }, { "grad_norm": 0.07449892163276672, "learning_rate": 3.113909320865252e-05, "loss": 0.0119, "step": 192590 }, { "grad_norm": 0.08380363136529922, "learning_rate": 3.113398892722301e-05, "loss": 0.0151, "step": 192600 }, { "grad_norm": 0.08893382549285889, "learning_rate": 3.112888487503333e-05, "loss": 0.0118, "step": 192610 }, { "grad_norm": 0.1378265917301178, "learning_rate": 3.112378105214552e-05, "loss": 0.0142, "step": 192620 }, { "grad_norm": 0.19338077306747437, "learning_rate": 3.1118677458621586e-05, "loss": 0.0167, "step": 192630 }, { "grad_norm": 0.04134218767285347, "learning_rate": 3.1113574094523566e-05, "loss": 0.0127, "step": 192640 }, { "grad_norm": 0.06683547794818878, "learning_rate": 3.110847095991344e-05, "loss": 0.0081, "step": 192650 }, { "grad_norm": 0.14558584988117218, "learning_rate": 3.110336805485322e-05, "loss": 0.0126, "step": 192660 }, { "grad_norm": 0.31678569316864014, "learning_rate": 3.1098265379404954e-05, "loss": 0.0118, "step": 192670 }, { "grad_norm": 0.10087304562330246, "learning_rate": 3.109316293363058e-05, "loss": 0.0162, "step": 192680 }, { "grad_norm": 0.07081694155931473, "learning_rate": 3.108806071759212e-05, "loss": 0.0132, "step": 192690 }, { "grad_norm": 0.08872600644826889, "learning_rate": 3.108295873135158e-05, "loss": 0.0132, "step": 192700 }, { "grad_norm": 0.07990634441375732, "learning_rate": 3.107785697497099e-05, "loss": 0.0115, "step": 192710 }, { "grad_norm": 0.15246430039405823, "learning_rate": 3.107275544851227e-05, "loss": 0.0154, "step": 192720 }, { "grad_norm": 0.10287946462631226, "learning_rate": 3.106765415203744e-05, "loss": 0.0119, "step": 192730 }, { "grad_norm": 0.12321262061595917, "learning_rate": 3.106255308560851e-05, "loss": 0.0119, "step": 192740 }, { "grad_norm": 0.10842669755220413, "learning_rate": 3.105745224928742e-05, "loss": 0.0117, "step": 192750 }, { "grad_norm": 0.091094009578228, "learning_rate": 3.1052351643136166e-05, "loss": 0.0129, "step": 192760 }, { "grad_norm": 0.12609457969665527, "learning_rate": 3.104725126721674e-05, "loss": 0.0147, "step": 192770 }, { "grad_norm": 0.10866589844226837, "learning_rate": 3.104215112159111e-05, "loss": 0.0166, "step": 192780 }, { "grad_norm": 0.06324508041143417, "learning_rate": 3.1037051206321235e-05, "loss": 0.0128, "step": 192790 }, { "grad_norm": 0.0912817195057869, "learning_rate": 3.103195152146908e-05, "loss": 0.0264, "step": 192800 }, { "grad_norm": 0.16085542738437653, "learning_rate": 3.1026852067096646e-05, "loss": 0.0145, "step": 192810 }, { "grad_norm": 0.23876386880874634, "learning_rate": 3.1021752843265863e-05, "loss": 0.0163, "step": 192820 }, { "grad_norm": 0.09069845080375671, "learning_rate": 3.1016653850038703e-05, "loss": 0.0191, "step": 192830 }, { "grad_norm": 0.11477050930261612, "learning_rate": 3.1011555087477114e-05, "loss": 0.0137, "step": 192840 }, { "grad_norm": 0.08200155943632126, "learning_rate": 3.100645655564307e-05, "loss": 0.0152, "step": 192850 }, { "grad_norm": 0.13106001913547516, "learning_rate": 3.100135825459851e-05, "loss": 0.0231, "step": 192860 }, { "grad_norm": 0.06735099852085114, "learning_rate": 3.099626018440538e-05, "loss": 0.0162, "step": 192870 }, { "grad_norm": 0.0692671611905098, "learning_rate": 3.0991162345125626e-05, "loss": 0.0133, "step": 192880 }, { "grad_norm": 0.05177628993988037, "learning_rate": 3.098606473682121e-05, "loss": 0.0135, "step": 192890 }, { "grad_norm": 0.062178291380405426, "learning_rate": 3.098096735955406e-05, "loss": 0.0139, "step": 192900 }, { "grad_norm": 0.19745682179927826, "learning_rate": 3.097587021338609e-05, "loss": 0.0122, "step": 192910 }, { "grad_norm": 0.04739203304052353, "learning_rate": 3.0970773298379286e-05, "loss": 0.0119, "step": 192920 }, { "grad_norm": 0.12017642706632614, "learning_rate": 3.096567661459554e-05, "loss": 0.0128, "step": 192930 }, { "grad_norm": 0.09029453992843628, "learning_rate": 3.09605801620968e-05, "loss": 0.0176, "step": 192940 }, { "grad_norm": 0.098316490650177, "learning_rate": 3.095548394094497e-05, "loss": 0.015, "step": 192950 }, { "grad_norm": 0.05195852741599083, "learning_rate": 3.095038795120202e-05, "loss": 0.0094, "step": 192960 }, { "grad_norm": 0.055558301508426666, "learning_rate": 3.094529219292982e-05, "loss": 0.0098, "step": 192970 }, { "grad_norm": 0.11070480942726135, "learning_rate": 3.0940196666190323e-05, "loss": 0.0155, "step": 192980 }, { "grad_norm": 0.0958605408668518, "learning_rate": 3.0935101371045436e-05, "loss": 0.0178, "step": 192990 }, { "grad_norm": 0.1297767460346222, "learning_rate": 3.0930006307557047e-05, "loss": 0.0121, "step": 193000 }, { "grad_norm": 0.1978663057088852, "learning_rate": 3.0924911475787096e-05, "loss": 0.0118, "step": 193010 }, { "grad_norm": 0.0768435150384903, "learning_rate": 3.091981687579749e-05, "loss": 0.0113, "step": 193020 }, { "grad_norm": 0.09089866280555725, "learning_rate": 3.091472250765012e-05, "loss": 0.0115, "step": 193030 }, { "grad_norm": 0.07536239176988602, "learning_rate": 3.090962837140688e-05, "loss": 0.015, "step": 193040 }, { "grad_norm": 0.11075359582901001, "learning_rate": 3.090453446712967e-05, "loss": 0.0109, "step": 193050 }, { "grad_norm": 0.10891420394182205, "learning_rate": 3.089944079488044e-05, "loss": 0.0121, "step": 193060 }, { "grad_norm": 0.3308716118335724, "learning_rate": 3.0894347354721e-05, "loss": 0.0203, "step": 193070 }, { "grad_norm": 0.1728089600801468, "learning_rate": 3.0889254146713286e-05, "loss": 0.0141, "step": 193080 }, { "grad_norm": 0.41880956292152405, "learning_rate": 3.088416117091917e-05, "loss": 0.015, "step": 193090 }, { "grad_norm": 0.4147465229034424, "learning_rate": 3.087906842740057e-05, "loss": 0.0142, "step": 193100 }, { "grad_norm": 0.14272207021713257, "learning_rate": 3.0873975916219325e-05, "loss": 0.0155, "step": 193110 }, { "grad_norm": 0.10520780086517334, "learning_rate": 3.0868883637437327e-05, "loss": 0.0165, "step": 193120 }, { "grad_norm": 0.12090113013982773, "learning_rate": 3.0863791591116473e-05, "loss": 0.014, "step": 193130 }, { "grad_norm": 0.10307391732931137, "learning_rate": 3.08586997773186e-05, "loss": 0.0132, "step": 193140 }, { "grad_norm": 0.05040854960680008, "learning_rate": 3.08536081961056e-05, "loss": 0.0142, "step": 193150 }, { "grad_norm": 0.07311967760324478, "learning_rate": 3.0848516847539336e-05, "loss": 0.0121, "step": 193160 }, { "grad_norm": 0.11614755541086197, "learning_rate": 3.084342573168169e-05, "loss": 0.0115, "step": 193170 }, { "grad_norm": 0.07130278646945953, "learning_rate": 3.0838334848594495e-05, "loss": 0.0136, "step": 193180 }, { "grad_norm": 0.19827118515968323, "learning_rate": 3.083324419833962e-05, "loss": 0.0129, "step": 193190 }, { "grad_norm": 0.10420311987400055, "learning_rate": 3.082815378097892e-05, "loss": 0.0138, "step": 193200 }, { "grad_norm": 0.08817262947559357, "learning_rate": 3.082306359657428e-05, "loss": 0.016, "step": 193210 }, { "grad_norm": 0.0721122995018959, "learning_rate": 3.0817973645187505e-05, "loss": 0.0142, "step": 193220 }, { "grad_norm": 0.06908596307039261, "learning_rate": 3.081288392688045e-05, "loss": 0.0118, "step": 193230 }, { "grad_norm": 0.14705698192119598, "learning_rate": 3.0807794441715e-05, "loss": 0.024, "step": 193240 }, { "grad_norm": 0.131950244307518, "learning_rate": 3.080270518975295e-05, "loss": 0.0113, "step": 193250 }, { "grad_norm": 0.09959548711776733, "learning_rate": 3.079761617105615e-05, "loss": 0.0164, "step": 193260 }, { "grad_norm": 0.0813436210155487, "learning_rate": 3.0792527385686445e-05, "loss": 0.0109, "step": 193270 }, { "grad_norm": 0.04863734543323517, "learning_rate": 3.078743883370569e-05, "loss": 0.0185, "step": 193280 }, { "grad_norm": 0.07601074874401093, "learning_rate": 3.078235051517567e-05, "loss": 0.0154, "step": 193290 }, { "grad_norm": 0.0603690929710865, "learning_rate": 3.077726243015824e-05, "loss": 0.0127, "step": 193300 }, { "grad_norm": 0.07280798256397247, "learning_rate": 3.0772174578715236e-05, "loss": 0.0118, "step": 193310 }, { "grad_norm": 0.2499954253435135, "learning_rate": 3.0767086960908444e-05, "loss": 0.0174, "step": 193320 }, { "grad_norm": 0.22543027997016907, "learning_rate": 3.076199957679971e-05, "loss": 0.0152, "step": 193330 }, { "grad_norm": 0.0539887510240078, "learning_rate": 3.075691242645084e-05, "loss": 0.0144, "step": 193340 }, { "grad_norm": 0.11043057590723038, "learning_rate": 3.075182550992367e-05, "loss": 0.0134, "step": 193350 }, { "grad_norm": 0.21914826333522797, "learning_rate": 3.074673882727999e-05, "loss": 0.013, "step": 193360 }, { "grad_norm": 0.05885922536253929, "learning_rate": 3.0741652378581585e-05, "loss": 0.0151, "step": 193370 }, { "grad_norm": 0.08203389495611191, "learning_rate": 3.0736566163890325e-05, "loss": 0.0115, "step": 193380 }, { "grad_norm": 0.08143788576126099, "learning_rate": 3.073148018326795e-05, "loss": 0.0083, "step": 193390 }, { "grad_norm": 0.053285565227270126, "learning_rate": 3.072639443677629e-05, "loss": 0.0171, "step": 193400 }, { "grad_norm": 0.041181355714797974, "learning_rate": 3.0721308924477124e-05, "loss": 0.0109, "step": 193410 }, { "grad_norm": 0.06301455944776535, "learning_rate": 3.0716223646432274e-05, "loss": 0.01, "step": 193420 }, { "grad_norm": 0.09204502403736115, "learning_rate": 3.0711138602703504e-05, "loss": 0.0084, "step": 193430 }, { "grad_norm": 0.1602962166070938, "learning_rate": 3.0706053793352606e-05, "loss": 0.0157, "step": 193440 }, { "grad_norm": 0.15259036421775818, "learning_rate": 3.0700969218441386e-05, "loss": 0.0152, "step": 193450 }, { "grad_norm": 0.08423150330781937, "learning_rate": 3.0695884878031604e-05, "loss": 0.0153, "step": 193460 }, { "grad_norm": 0.0770726203918457, "learning_rate": 3.069080077218504e-05, "loss": 0.0113, "step": 193470 }, { "grad_norm": 0.06774123013019562, "learning_rate": 3.068571690096348e-05, "loss": 0.0188, "step": 193480 }, { "grad_norm": 0.0654936358332634, "learning_rate": 3.068063326442871e-05, "loss": 0.0106, "step": 193490 }, { "grad_norm": 0.06429482996463776, "learning_rate": 3.067554986264247e-05, "loss": 0.0131, "step": 193500 }, { "grad_norm": 0.10118528455495834, "learning_rate": 3.0670466695666544e-05, "loss": 0.0079, "step": 193510 }, { "grad_norm": 0.10475943237543106, "learning_rate": 3.06653837635627e-05, "loss": 0.0128, "step": 193520 }, { "grad_norm": 0.07230609655380249, "learning_rate": 3.066030106639272e-05, "loss": 0.0211, "step": 193530 }, { "grad_norm": 0.07138821482658386, "learning_rate": 3.0655218604218315e-05, "loss": 0.0182, "step": 193540 }, { "grad_norm": 0.09500598162412643, "learning_rate": 3.065013637710127e-05, "loss": 0.0099, "step": 193550 }, { "grad_norm": 0.21297790110111237, "learning_rate": 3.064505438510335e-05, "loss": 0.0149, "step": 193560 }, { "grad_norm": 0.17255426943302155, "learning_rate": 3.063997262828629e-05, "loss": 0.0251, "step": 193570 }, { "grad_norm": 0.17861145734786987, "learning_rate": 3.0634891106711836e-05, "loss": 0.0151, "step": 193580 }, { "grad_norm": 0.057441651821136475, "learning_rate": 3.062980982044174e-05, "loss": 0.0111, "step": 193590 }, { "grad_norm": 0.07819370925426483, "learning_rate": 3.062472876953775e-05, "loss": 0.013, "step": 193600 }, { "grad_norm": 0.15561208128929138, "learning_rate": 3.0619647954061594e-05, "loss": 0.0127, "step": 193610 }, { "grad_norm": 0.07642487436532974, "learning_rate": 3.0614567374075004e-05, "loss": 0.0135, "step": 193620 }, { "grad_norm": 0.06325715780258179, "learning_rate": 3.060948702963974e-05, "loss": 0.018, "step": 193630 }, { "grad_norm": 0.12992751598358154, "learning_rate": 3.0604406920817504e-05, "loss": 0.0168, "step": 193640 }, { "grad_norm": 0.07969452440738678, "learning_rate": 3.059932704767004e-05, "loss": 0.0144, "step": 193650 }, { "grad_norm": 0.14277897775173187, "learning_rate": 3.059424741025907e-05, "loss": 0.0182, "step": 193660 }, { "grad_norm": 0.10058169066905975, "learning_rate": 3.058916800864632e-05, "loss": 0.0183, "step": 193670 }, { "grad_norm": 0.10720477998256683, "learning_rate": 3.05840888428935e-05, "loss": 0.0206, "step": 193680 }, { "grad_norm": 0.08932270109653473, "learning_rate": 3.0579009913062334e-05, "loss": 0.0098, "step": 193690 }, { "grad_norm": 0.0842830091714859, "learning_rate": 3.057393121921453e-05, "loss": 0.0118, "step": 193700 }, { "grad_norm": 0.08324195444583893, "learning_rate": 3.0568852761411814e-05, "loss": 0.0117, "step": 193710 }, { "grad_norm": 0.17314966022968292, "learning_rate": 3.0563774539715864e-05, "loss": 0.014, "step": 193720 }, { "grad_norm": 0.2502899765968323, "learning_rate": 3.055869655418843e-05, "loss": 0.0202, "step": 193730 }, { "grad_norm": 0.06095978245139122, "learning_rate": 3.055361880489117e-05, "loss": 0.0159, "step": 193740 }, { "grad_norm": 0.11456616222858429, "learning_rate": 3.0548541291885805e-05, "loss": 0.0169, "step": 193750 }, { "grad_norm": 0.2100641280412674, "learning_rate": 3.054346401523403e-05, "loss": 0.0137, "step": 193760 }, { "grad_norm": 0.1762220710515976, "learning_rate": 3.0538386974997536e-05, "loss": 0.0163, "step": 193770 }, { "grad_norm": 0.1610131412744522, "learning_rate": 3.053331017123802e-05, "loss": 0.0151, "step": 193780 }, { "grad_norm": 0.08871090412139893, "learning_rate": 3.052823360401716e-05, "loss": 0.0137, "step": 193790 }, { "grad_norm": 0.06114361807703972, "learning_rate": 3.052315727339664e-05, "loss": 0.0106, "step": 193800 }, { "grad_norm": 0.04813535884022713, "learning_rate": 3.051808117943817e-05, "loss": 0.0153, "step": 193810 }, { "grad_norm": 0.12880702316761017, "learning_rate": 3.0513005322203392e-05, "loss": 0.0161, "step": 193820 }, { "grad_norm": 0.057031456381082535, "learning_rate": 3.050792970175399e-05, "loss": 0.0114, "step": 193830 }, { "grad_norm": 0.06641924381256104, "learning_rate": 3.0502854318151653e-05, "loss": 0.0127, "step": 193840 }, { "grad_norm": 0.07888400554656982, "learning_rate": 3.0497779171458063e-05, "loss": 0.0154, "step": 193850 }, { "grad_norm": 0.10207097232341766, "learning_rate": 3.049270426173485e-05, "loss": 0.0151, "step": 193860 }, { "grad_norm": 0.09485193341970444, "learning_rate": 3.048762958904369e-05, "loss": 0.0118, "step": 193870 }, { "grad_norm": 0.14642342925071716, "learning_rate": 3.048255515344628e-05, "loss": 0.015, "step": 193880 }, { "grad_norm": 0.04065338522195816, "learning_rate": 3.0477480955004233e-05, "loss": 0.0112, "step": 193890 }, { "grad_norm": 0.047365330159664154, "learning_rate": 3.0472406993779223e-05, "loss": 0.0149, "step": 193900 }, { "grad_norm": 0.08391782641410828, "learning_rate": 3.046733326983291e-05, "loss": 0.0099, "step": 193910 }, { "grad_norm": 0.07878198474645615, "learning_rate": 3.0462259783226954e-05, "loss": 0.0127, "step": 193920 }, { "grad_norm": 0.13648733496665955, "learning_rate": 3.0457186534022962e-05, "loss": 0.0175, "step": 193930 }, { "grad_norm": 0.1371980607509613, "learning_rate": 3.0452113522282618e-05, "loss": 0.0151, "step": 193940 }, { "grad_norm": 0.15705612301826477, "learning_rate": 3.0447040748067567e-05, "loss": 0.0087, "step": 193950 }, { "grad_norm": 0.1239374503493309, "learning_rate": 3.044196821143941e-05, "loss": 0.0122, "step": 193960 }, { "grad_norm": 0.07074077427387238, "learning_rate": 3.0436895912459808e-05, "loss": 0.0121, "step": 193970 }, { "grad_norm": 0.08237256854772568, "learning_rate": 3.0431823851190383e-05, "loss": 0.0123, "step": 193980 }, { "grad_norm": 0.055204037576913834, "learning_rate": 3.04267520276928e-05, "loss": 0.0096, "step": 193990 }, { "grad_norm": 0.05187305808067322, "learning_rate": 3.0421680442028645e-05, "loss": 0.0086, "step": 194000 }, { "grad_norm": 0.08802944421768188, "learning_rate": 3.0416609094259545e-05, "loss": 0.0136, "step": 194010 }, { "grad_norm": 0.09436172246932983, "learning_rate": 3.0411537984447163e-05, "loss": 0.0165, "step": 194020 }, { "grad_norm": 0.19203293323516846, "learning_rate": 3.040646711265307e-05, "loss": 0.0129, "step": 194030 }, { "grad_norm": 0.13316288590431213, "learning_rate": 3.0401396478938892e-05, "loss": 0.0104, "step": 194040 }, { "grad_norm": 0.1464243084192276, "learning_rate": 3.039632608336626e-05, "loss": 0.0123, "step": 194050 }, { "grad_norm": 0.3143026828765869, "learning_rate": 3.03912559259968e-05, "loss": 0.0118, "step": 194060 }, { "grad_norm": 0.09575064480304718, "learning_rate": 3.038618600689206e-05, "loss": 0.0178, "step": 194070 }, { "grad_norm": 0.13407182693481445, "learning_rate": 3.038111632611369e-05, "loss": 0.0131, "step": 194080 }, { "grad_norm": 0.10823870450258255, "learning_rate": 3.0376046883723275e-05, "loss": 0.0151, "step": 194090 }, { "grad_norm": 0.09924686700105667, "learning_rate": 3.037097767978243e-05, "loss": 0.0129, "step": 194100 }, { "grad_norm": 0.05309636518359184, "learning_rate": 3.0365908714352733e-05, "loss": 0.0146, "step": 194110 }, { "grad_norm": 0.13794636726379395, "learning_rate": 3.0360839987495776e-05, "loss": 0.0138, "step": 194120 }, { "grad_norm": 0.6233665347099304, "learning_rate": 3.035577149927318e-05, "loss": 0.0146, "step": 194130 }, { "grad_norm": 0.2189297080039978, "learning_rate": 3.035070324974648e-05, "loss": 0.0129, "step": 194140 }, { "grad_norm": 0.05506867542862892, "learning_rate": 3.0345635238977294e-05, "loss": 0.0125, "step": 194150 }, { "grad_norm": 0.047136079519987106, "learning_rate": 3.0340567467027193e-05, "loss": 0.015, "step": 194160 }, { "grad_norm": 0.07850368320941925, "learning_rate": 3.0335499933957777e-05, "loss": 0.0118, "step": 194170 }, { "grad_norm": 0.16085733473300934, "learning_rate": 3.033043263983059e-05, "loss": 0.0147, "step": 194180 }, { "grad_norm": 0.1410415768623352, "learning_rate": 3.0325365584707216e-05, "loss": 0.0121, "step": 194190 }, { "grad_norm": 0.11543109267950058, "learning_rate": 3.0320298768649246e-05, "loss": 0.0116, "step": 194200 }, { "grad_norm": 0.1383790820837021, "learning_rate": 3.03152321917182e-05, "loss": 0.0144, "step": 194210 }, { "grad_norm": 0.12439191341400146, "learning_rate": 3.0310165853975674e-05, "loss": 0.0095, "step": 194220 }, { "grad_norm": 0.5029171705245972, "learning_rate": 3.0305099755483224e-05, "loss": 0.0121, "step": 194230 }, { "grad_norm": 0.0634826272726059, "learning_rate": 3.0300033896302426e-05, "loss": 0.014, "step": 194240 }, { "grad_norm": 0.1307435780763626, "learning_rate": 3.0294968276494796e-05, "loss": 0.0121, "step": 194250 }, { "grad_norm": 0.0693371593952179, "learning_rate": 3.0289902896121903e-05, "loss": 0.0107, "step": 194260 }, { "grad_norm": 0.06509946286678314, "learning_rate": 3.028483775524531e-05, "loss": 0.0108, "step": 194270 }, { "grad_norm": 0.09992468357086182, "learning_rate": 3.0279772853926556e-05, "loss": 0.012, "step": 194280 }, { "grad_norm": 0.06964760273694992, "learning_rate": 3.0274708192227173e-05, "loss": 0.0144, "step": 194290 }, { "grad_norm": 0.09078145027160645, "learning_rate": 3.02696437702087e-05, "loss": 0.0126, "step": 194300 }, { "grad_norm": 0.09077732264995575, "learning_rate": 3.02645795879327e-05, "loss": 0.0114, "step": 194310 }, { "grad_norm": 0.082001693546772, "learning_rate": 3.025951564546069e-05, "loss": 0.0136, "step": 194320 }, { "grad_norm": 0.0511941984295845, "learning_rate": 3.0254451942854196e-05, "loss": 0.0122, "step": 194330 }, { "grad_norm": 0.07443481683731079, "learning_rate": 3.024938848017476e-05, "loss": 0.0294, "step": 194340 }, { "grad_norm": 0.13049490749835968, "learning_rate": 3.0244325257483897e-05, "loss": 0.0154, "step": 194350 }, { "grad_norm": 0.10371310263872147, "learning_rate": 3.023926227484314e-05, "loss": 0.0111, "step": 194360 }, { "grad_norm": 0.6212798357009888, "learning_rate": 3.023419953231399e-05, "loss": 0.0122, "step": 194370 }, { "grad_norm": 0.11410152167081833, "learning_rate": 3.0229137029957993e-05, "loss": 0.0108, "step": 194380 }, { "grad_norm": 0.15588894486427307, "learning_rate": 3.0224074767836637e-05, "loss": 0.0144, "step": 194390 }, { "grad_norm": 0.10975398123264313, "learning_rate": 3.0219012746011454e-05, "loss": 0.0163, "step": 194400 }, { "grad_norm": 0.11312080919742584, "learning_rate": 3.021395096454393e-05, "loss": 0.0109, "step": 194410 }, { "grad_norm": 0.15543852746486664, "learning_rate": 3.0208889423495594e-05, "loss": 0.0174, "step": 194420 }, { "grad_norm": 0.19082070887088776, "learning_rate": 3.020382812292793e-05, "loss": 0.0173, "step": 194430 }, { "grad_norm": 0.06513717025518417, "learning_rate": 3.0198767062902455e-05, "loss": 0.0136, "step": 194440 }, { "grad_norm": 0.1154998168349266, "learning_rate": 3.0193706243480656e-05, "loss": 0.0105, "step": 194450 }, { "grad_norm": 0.09556050598621368, "learning_rate": 3.0188645664724014e-05, "loss": 0.013, "step": 194460 }, { "grad_norm": 0.11161264032125473, "learning_rate": 3.018358532669403e-05, "loss": 0.0107, "step": 194470 }, { "grad_norm": 0.15974824130535126, "learning_rate": 3.0178525229452204e-05, "loss": 0.014, "step": 194480 }, { "grad_norm": 0.079057477414608, "learning_rate": 3.0173465373060018e-05, "loss": 0.0158, "step": 194490 }, { "grad_norm": 0.07241007685661316, "learning_rate": 3.0168405757578934e-05, "loss": 0.0153, "step": 194500 }, { "grad_norm": 0.31189000606536865, "learning_rate": 3.016334638307044e-05, "loss": 0.0156, "step": 194510 }, { "grad_norm": 0.1692381352186203, "learning_rate": 3.015828724959604e-05, "loss": 0.0181, "step": 194520 }, { "grad_norm": 0.09132332354784012, "learning_rate": 3.0153228357217166e-05, "loss": 0.0165, "step": 194530 }, { "grad_norm": 0.07822779566049576, "learning_rate": 3.01481697059953e-05, "loss": 0.0175, "step": 194540 }, { "grad_norm": 0.09604644775390625, "learning_rate": 3.0143111295991922e-05, "loss": 0.0137, "step": 194550 }, { "grad_norm": 0.1246941015124321, "learning_rate": 3.0138053127268506e-05, "loss": 0.0139, "step": 194560 }, { "grad_norm": 0.10748615860939026, "learning_rate": 3.013299519988647e-05, "loss": 0.0175, "step": 194570 }, { "grad_norm": 0.1720854490995407, "learning_rate": 3.012793751390731e-05, "loss": 0.0167, "step": 194580 }, { "grad_norm": 0.08503196388483047, "learning_rate": 3.012288006939249e-05, "loss": 0.015, "step": 194590 }, { "grad_norm": 0.16258695721626282, "learning_rate": 3.0117822866403424e-05, "loss": 0.0107, "step": 194600 }, { "grad_norm": 0.06893172115087509, "learning_rate": 3.0112765905001583e-05, "loss": 0.0132, "step": 194610 }, { "grad_norm": 0.07500805705785751, "learning_rate": 3.0107709185248407e-05, "loss": 0.0119, "step": 194620 }, { "grad_norm": 0.13982349634170532, "learning_rate": 3.010265270720537e-05, "loss": 0.0145, "step": 194630 }, { "grad_norm": 0.09686604142189026, "learning_rate": 3.009759647093387e-05, "loss": 0.0178, "step": 194640 }, { "grad_norm": 0.083826944231987, "learning_rate": 3.0092540476495362e-05, "loss": 0.0135, "step": 194650 }, { "grad_norm": 0.11002078652381897, "learning_rate": 3.0087484723951304e-05, "loss": 0.014, "step": 194660 }, { "grad_norm": 0.10071070492267609, "learning_rate": 3.0082429213363084e-05, "loss": 0.0117, "step": 194670 }, { "grad_norm": 0.06742372363805771, "learning_rate": 3.0077373944792158e-05, "loss": 0.0146, "step": 194680 }, { "grad_norm": 0.13818670809268951, "learning_rate": 3.0072318918299947e-05, "loss": 0.0153, "step": 194690 }, { "grad_norm": 0.09866032004356384, "learning_rate": 3.0067264133947897e-05, "loss": 0.0132, "step": 194700 }, { "grad_norm": 0.06385965645313263, "learning_rate": 3.0062209591797386e-05, "loss": 0.0102, "step": 194710 }, { "grad_norm": 0.16237585246562958, "learning_rate": 3.0057155291909855e-05, "loss": 0.0138, "step": 194720 }, { "grad_norm": 0.1343536525964737, "learning_rate": 3.0052101234346718e-05, "loss": 0.0169, "step": 194730 }, { "grad_norm": 0.2004028856754303, "learning_rate": 3.00470474191694e-05, "loss": 0.0188, "step": 194740 }, { "grad_norm": 0.08669586479663849, "learning_rate": 3.0041993846439274e-05, "loss": 0.0076, "step": 194750 }, { "grad_norm": 0.1473059207201004, "learning_rate": 3.003694051621777e-05, "loss": 0.0115, "step": 194760 }, { "grad_norm": 0.11330032348632812, "learning_rate": 3.003188742856631e-05, "loss": 0.0113, "step": 194770 }, { "grad_norm": 0.06455081701278687, "learning_rate": 3.0026834583546244e-05, "loss": 0.013, "step": 194780 }, { "grad_norm": 0.12941515445709229, "learning_rate": 3.0021781981218998e-05, "loss": 0.0102, "step": 194790 }, { "grad_norm": 0.1008070558309555, "learning_rate": 3.0016729621645967e-05, "loss": 0.0116, "step": 194800 }, { "grad_norm": 0.09471125900745392, "learning_rate": 3.0011677504888554e-05, "loss": 0.0152, "step": 194810 }, { "grad_norm": 0.14329229295253754, "learning_rate": 3.000662563100811e-05, "loss": 0.01, "step": 194820 }, { "grad_norm": 0.10892070084810257, "learning_rate": 3.0001574000066045e-05, "loss": 0.0139, "step": 194830 }, { "grad_norm": 0.17808759212493896, "learning_rate": 2.9996522612123755e-05, "loss": 0.0105, "step": 194840 }, { "grad_norm": 0.15327590703964233, "learning_rate": 2.999147146724258e-05, "loss": 0.0093, "step": 194850 }, { "grad_norm": 0.10172733664512634, "learning_rate": 2.9986420565483925e-05, "loss": 0.0101, "step": 194860 }, { "grad_norm": 0.1052720844745636, "learning_rate": 2.998136990690915e-05, "loss": 0.0117, "step": 194870 }, { "grad_norm": 0.09687182307243347, "learning_rate": 2.997631949157966e-05, "loss": 0.0125, "step": 194880 }, { "grad_norm": 0.1243574395775795, "learning_rate": 2.997126931955676e-05, "loss": 0.01, "step": 194890 }, { "grad_norm": 0.0871216431260109, "learning_rate": 2.9966219390901863e-05, "loss": 0.009, "step": 194900 }, { "grad_norm": 0.13685142993927002, "learning_rate": 2.9961169705676324e-05, "loss": 0.0131, "step": 194910 }, { "grad_norm": 0.10116638988256454, "learning_rate": 2.9956120263941478e-05, "loss": 0.011, "step": 194920 }, { "grad_norm": 0.24830284714698792, "learning_rate": 2.9951071065758697e-05, "loss": 0.015, "step": 194930 }, { "grad_norm": 0.2402850091457367, "learning_rate": 2.994602211118932e-05, "loss": 0.011, "step": 194940 }, { "grad_norm": 0.10391630232334137, "learning_rate": 2.9940973400294743e-05, "loss": 0.0114, "step": 194950 }, { "grad_norm": 0.092475026845932, "learning_rate": 2.993592493313625e-05, "loss": 0.0088, "step": 194960 }, { "grad_norm": 0.10540097951889038, "learning_rate": 2.993087670977521e-05, "loss": 0.0104, "step": 194970 }, { "grad_norm": 0.07464906573295593, "learning_rate": 2.9925828730272976e-05, "loss": 0.0108, "step": 194980 }, { "grad_norm": 0.05253443494439125, "learning_rate": 2.992078099469089e-05, "loss": 0.0111, "step": 194990 }, { "grad_norm": 0.19241958856582642, "learning_rate": 2.991573350309026e-05, "loss": 0.0153, "step": 195000 }, { "grad_norm": 0.08695973455905914, "learning_rate": 2.991068625553242e-05, "loss": 0.0119, "step": 195010 }, { "grad_norm": 0.03598088026046753, "learning_rate": 2.9905639252078727e-05, "loss": 0.0109, "step": 195020 }, { "grad_norm": 0.12836118042469025, "learning_rate": 2.9900592492790485e-05, "loss": 0.0147, "step": 195030 }, { "grad_norm": 0.07458854466676712, "learning_rate": 2.989554597772901e-05, "loss": 0.0114, "step": 195040 }, { "grad_norm": 0.09128211438655853, "learning_rate": 2.9890499706955638e-05, "loss": 0.0093, "step": 195050 }, { "grad_norm": 0.11350052058696747, "learning_rate": 2.988545368053169e-05, "loss": 0.015, "step": 195060 }, { "grad_norm": 0.17394594848155975, "learning_rate": 2.988040789851847e-05, "loss": 0.0105, "step": 195070 }, { "grad_norm": 0.08569498360157013, "learning_rate": 2.9875362360977278e-05, "loss": 0.0104, "step": 195080 }, { "grad_norm": 0.09398705512285233, "learning_rate": 2.9870317067969445e-05, "loss": 0.0163, "step": 195090 }, { "grad_norm": 0.8632879257202148, "learning_rate": 2.9865272019556256e-05, "loss": 0.0165, "step": 195100 }, { "grad_norm": 0.08000004291534424, "learning_rate": 2.9860227215799034e-05, "loss": 0.0094, "step": 195110 }, { "grad_norm": 0.17871920764446259, "learning_rate": 2.9855182656759057e-05, "loss": 0.0204, "step": 195120 }, { "grad_norm": 0.09162009507417679, "learning_rate": 2.9850138342497635e-05, "loss": 0.0105, "step": 195130 }, { "grad_norm": 0.10488429665565491, "learning_rate": 2.9845094273076057e-05, "loss": 0.0113, "step": 195140 }, { "grad_norm": 0.10569893568754196, "learning_rate": 2.9840050448555614e-05, "loss": 0.0118, "step": 195150 }, { "grad_norm": 0.13763892650604248, "learning_rate": 2.9835006868997594e-05, "loss": 0.0133, "step": 195160 }, { "grad_norm": 0.06570009142160416, "learning_rate": 2.9829963534463268e-05, "loss": 0.0127, "step": 195170 }, { "grad_norm": 0.0877063125371933, "learning_rate": 2.9824920445013937e-05, "loss": 0.0107, "step": 195180 }, { "grad_norm": 0.1726713478565216, "learning_rate": 2.9819877600710873e-05, "loss": 0.0135, "step": 195190 }, { "grad_norm": 0.5784580707550049, "learning_rate": 2.981483500161535e-05, "loss": 0.0311, "step": 195200 }, { "grad_norm": 0.10064489394426346, "learning_rate": 2.980979264778863e-05, "loss": 0.0156, "step": 195210 }, { "grad_norm": 0.13795794546604156, "learning_rate": 2.9804750539291992e-05, "loss": 0.0166, "step": 195220 }, { "grad_norm": 0.10581028461456299, "learning_rate": 2.979970867618672e-05, "loss": 0.0171, "step": 195230 }, { "grad_norm": 0.16208449006080627, "learning_rate": 2.9794667058534047e-05, "loss": 0.0115, "step": 195240 }, { "grad_norm": 0.14005111157894135, "learning_rate": 2.9789625686395245e-05, "loss": 0.0153, "step": 195250 }, { "grad_norm": 0.07569295912981033, "learning_rate": 2.978458455983157e-05, "loss": 0.0121, "step": 195260 }, { "grad_norm": 0.18033823370933533, "learning_rate": 2.97795436789043e-05, "loss": 0.0141, "step": 195270 }, { "grad_norm": 0.08474742621183395, "learning_rate": 2.977450304367464e-05, "loss": 0.012, "step": 195280 }, { "grad_norm": 0.09059303253889084, "learning_rate": 2.9769462654203874e-05, "loss": 0.0119, "step": 195290 }, { "grad_norm": 0.09172669798135757, "learning_rate": 2.976442251055323e-05, "loss": 0.0126, "step": 195300 }, { "grad_norm": 0.08850747346878052, "learning_rate": 2.9759382612783982e-05, "loss": 0.0128, "step": 195310 }, { "grad_norm": 0.1369434893131256, "learning_rate": 2.9754342960957326e-05, "loss": 0.0143, "step": 195320 }, { "grad_norm": 0.1383151412010193, "learning_rate": 2.9749303555134512e-05, "loss": 0.0187, "step": 195330 }, { "grad_norm": 0.11016619205474854, "learning_rate": 2.974426439537681e-05, "loss": 0.0147, "step": 195340 }, { "grad_norm": 0.10430923104286194, "learning_rate": 2.973922548174539e-05, "loss": 0.0112, "step": 195350 }, { "grad_norm": 0.08053325116634369, "learning_rate": 2.9734186814301513e-05, "loss": 0.0115, "step": 195360 }, { "grad_norm": 0.16252173483371735, "learning_rate": 2.97291483931064e-05, "loss": 0.0128, "step": 195370 }, { "grad_norm": 0.09749425947666168, "learning_rate": 2.9724110218221298e-05, "loss": 0.0177, "step": 195380 }, { "grad_norm": 0.09055919945240021, "learning_rate": 2.9719072289707374e-05, "loss": 0.0115, "step": 195390 }, { "grad_norm": 0.08083388954401016, "learning_rate": 2.9714034607625875e-05, "loss": 0.0146, "step": 195400 }, { "grad_norm": 0.13500910997390747, "learning_rate": 2.970899717203802e-05, "loss": 0.0122, "step": 195410 }, { "grad_norm": 0.06447175145149231, "learning_rate": 2.970395998300499e-05, "loss": 0.0152, "step": 195420 }, { "grad_norm": 0.1430426836013794, "learning_rate": 2.9698923040588017e-05, "loss": 0.0154, "step": 195430 }, { "grad_norm": 0.12191756069660187, "learning_rate": 2.9693886344848294e-05, "loss": 0.0108, "step": 195440 }, { "grad_norm": 0.07720009237527847, "learning_rate": 2.9688849895847036e-05, "loss": 0.0095, "step": 195450 }, { "grad_norm": 0.08367638289928436, "learning_rate": 2.968381369364542e-05, "loss": 0.0126, "step": 195460 }, { "grad_norm": 0.08509274572134018, "learning_rate": 2.9678777738304642e-05, "loss": 0.0173, "step": 195470 }, { "grad_norm": 0.10916808992624283, "learning_rate": 2.967374202988592e-05, "loss": 0.0107, "step": 195480 }, { "grad_norm": 0.052532147616147995, "learning_rate": 2.966870656845041e-05, "loss": 0.0155, "step": 195490 }, { "grad_norm": 0.07812700420618057, "learning_rate": 2.9663671354059307e-05, "loss": 0.0127, "step": 195500 }, { "grad_norm": 0.11082214117050171, "learning_rate": 2.9658636386773797e-05, "loss": 0.0104, "step": 195510 }, { "grad_norm": 0.14243000745773315, "learning_rate": 2.965360166665508e-05, "loss": 0.0174, "step": 195520 }, { "grad_norm": 0.18411868810653687, "learning_rate": 2.964856719376429e-05, "loss": 0.0134, "step": 195530 }, { "grad_norm": 0.13268817961215973, "learning_rate": 2.9643532968162636e-05, "loss": 0.01, "step": 195540 }, { "grad_norm": 0.12130294740200043, "learning_rate": 2.9638498989911283e-05, "loss": 0.0116, "step": 195550 }, { "grad_norm": 0.10890472680330276, "learning_rate": 2.963346525907138e-05, "loss": 0.0118, "step": 195560 }, { "grad_norm": 0.051068950444459915, "learning_rate": 2.9628431775704097e-05, "loss": 0.0131, "step": 195570 }, { "grad_norm": 0.20536313951015472, "learning_rate": 2.9623398539870607e-05, "loss": 0.0104, "step": 195580 }, { "grad_norm": 0.09862605482339859, "learning_rate": 2.9618365551632076e-05, "loss": 0.0165, "step": 195590 }, { "grad_norm": 0.09477578848600388, "learning_rate": 2.9613332811049633e-05, "loss": 0.0131, "step": 195600 }, { "grad_norm": 0.09212811291217804, "learning_rate": 2.9608300318184446e-05, "loss": 0.0084, "step": 195610 }, { "grad_norm": 0.06274933367967606, "learning_rate": 2.9603268073097657e-05, "loss": 0.0082, "step": 195620 }, { "grad_norm": 0.806212306022644, "learning_rate": 2.9598236075850444e-05, "loss": 0.0107, "step": 195630 }, { "grad_norm": 0.08157247304916382, "learning_rate": 2.9593204326503904e-05, "loss": 0.0117, "step": 195640 }, { "grad_norm": 0.05152556300163269, "learning_rate": 2.9588172825119198e-05, "loss": 0.0133, "step": 195650 }, { "grad_norm": 0.13125015795230865, "learning_rate": 2.9583141571757482e-05, "loss": 0.014, "step": 195660 }, { "grad_norm": 0.07777941972017288, "learning_rate": 2.9578110566479855e-05, "loss": 0.0096, "step": 195670 }, { "grad_norm": 0.15195751190185547, "learning_rate": 2.957307980934747e-05, "loss": 0.0146, "step": 195680 }, { "grad_norm": 0.07298668473958969, "learning_rate": 2.9568049300421447e-05, "loss": 0.011, "step": 195690 }, { "grad_norm": 0.0918571725487709, "learning_rate": 2.9563019039762935e-05, "loss": 0.0119, "step": 195700 }, { "grad_norm": 0.13115064799785614, "learning_rate": 2.955798902743302e-05, "loss": 0.0153, "step": 195710 }, { "grad_norm": 0.048692021518945694, "learning_rate": 2.9552959263492835e-05, "loss": 0.0103, "step": 195720 }, { "grad_norm": 0.05385694280266762, "learning_rate": 2.954792974800351e-05, "loss": 0.0118, "step": 195730 }, { "grad_norm": 0.20327824354171753, "learning_rate": 2.954290048102616e-05, "loss": 0.0133, "step": 195740 }, { "grad_norm": 0.09088591486215591, "learning_rate": 2.9537871462621858e-05, "loss": 0.011, "step": 195750 }, { "grad_norm": 0.05738937109708786, "learning_rate": 2.953284269285175e-05, "loss": 0.0117, "step": 195760 }, { "grad_norm": 0.05299486964941025, "learning_rate": 2.9527814171776923e-05, "loss": 0.0102, "step": 195770 }, { "grad_norm": 0.1475902497768402, "learning_rate": 2.9522785899458493e-05, "loss": 0.0167, "step": 195780 }, { "grad_norm": 0.06698776036500931, "learning_rate": 2.951775787595754e-05, "loss": 0.0084, "step": 195790 }, { "grad_norm": 0.2637841999530792, "learning_rate": 2.9512730101335175e-05, "loss": 0.0125, "step": 195800 }, { "grad_norm": 0.038362495601177216, "learning_rate": 2.9507702575652476e-05, "loss": 0.0104, "step": 195810 }, { "grad_norm": 0.08121839910745621, "learning_rate": 2.9502675298970527e-05, "loss": 0.0131, "step": 195820 }, { "grad_norm": 0.0858301892876625, "learning_rate": 2.9497648271350442e-05, "loss": 0.0168, "step": 195830 }, { "grad_norm": 0.1315009742975235, "learning_rate": 2.949262149285329e-05, "loss": 0.0119, "step": 195840 }, { "grad_norm": 0.19621002674102783, "learning_rate": 2.948759496354014e-05, "loss": 0.0184, "step": 195850 }, { "grad_norm": 0.06563730537891388, "learning_rate": 2.948256868347207e-05, "loss": 0.0086, "step": 195860 }, { "grad_norm": 0.2795974612236023, "learning_rate": 2.9477542652710188e-05, "loss": 0.0175, "step": 195870 }, { "grad_norm": 0.07600849866867065, "learning_rate": 2.9472516871315525e-05, "loss": 0.0145, "step": 195880 }, { "grad_norm": 0.06351998448371887, "learning_rate": 2.946749133934915e-05, "loss": 0.0135, "step": 195890 }, { "grad_norm": 0.10376477986574173, "learning_rate": 2.9462466056872152e-05, "loss": 0.0122, "step": 195900 }, { "grad_norm": 0.12173100560903549, "learning_rate": 2.9457441023945597e-05, "loss": 0.0117, "step": 195910 }, { "grad_norm": 0.04590436816215515, "learning_rate": 2.945241624063051e-05, "loss": 0.0103, "step": 195920 }, { "grad_norm": 0.08457951247692108, "learning_rate": 2.944739170698797e-05, "loss": 0.0092, "step": 195930 }, { "grad_norm": 0.08478598296642303, "learning_rate": 2.944236742307902e-05, "loss": 0.0127, "step": 195940 }, { "grad_norm": 0.08172619342803955, "learning_rate": 2.9437343388964737e-05, "loss": 0.0129, "step": 195950 }, { "grad_norm": 0.1664576381444931, "learning_rate": 2.9432319604706126e-05, "loss": 0.0141, "step": 195960 }, { "grad_norm": 0.11610013246536255, "learning_rate": 2.942729607036425e-05, "loss": 0.0153, "step": 195970 }, { "grad_norm": 0.06233232840895653, "learning_rate": 2.9422272786000172e-05, "loss": 0.0142, "step": 195980 }, { "grad_norm": 0.1780783236026764, "learning_rate": 2.941724975167489e-05, "loss": 0.018, "step": 195990 }, { "grad_norm": 0.037147700786590576, "learning_rate": 2.9412226967449452e-05, "loss": 0.0075, "step": 196000 }, { "grad_norm": 0.03530716150999069, "learning_rate": 2.9407204433384905e-05, "loss": 0.0093, "step": 196010 }, { "grad_norm": 0.0822959914803505, "learning_rate": 2.940218214954228e-05, "loss": 0.0116, "step": 196020 }, { "grad_norm": 0.20844238996505737, "learning_rate": 2.9397160115982562e-05, "loss": 0.0119, "step": 196030 }, { "grad_norm": 0.0583011768758297, "learning_rate": 2.939213833276681e-05, "loss": 0.0133, "step": 196040 }, { "grad_norm": 0.07505908608436584, "learning_rate": 2.9387116799956048e-05, "loss": 0.0117, "step": 196050 }, { "grad_norm": 0.07792467623949051, "learning_rate": 2.9382095517611264e-05, "loss": 0.0151, "step": 196060 }, { "grad_norm": 0.15561337769031525, "learning_rate": 2.9377074485793486e-05, "loss": 0.012, "step": 196070 }, { "grad_norm": 0.13612446188926697, "learning_rate": 2.9372053704563728e-05, "loss": 0.0131, "step": 196080 }, { "grad_norm": 0.1661333292722702, "learning_rate": 2.9367033173983006e-05, "loss": 0.0094, "step": 196090 }, { "grad_norm": 0.14406327903270721, "learning_rate": 2.936201289411229e-05, "loss": 0.0147, "step": 196100 }, { "grad_norm": 0.13531774282455444, "learning_rate": 2.9356992865012612e-05, "loss": 0.0204, "step": 196110 }, { "grad_norm": 0.06563237309455872, "learning_rate": 2.935197308674498e-05, "loss": 0.0135, "step": 196120 }, { "grad_norm": 0.08999569714069366, "learning_rate": 2.9346953559370337e-05, "loss": 0.0118, "step": 196130 }, { "grad_norm": 0.09937280416488647, "learning_rate": 2.9341934282949724e-05, "loss": 0.0104, "step": 196140 }, { "grad_norm": 0.1642063558101654, "learning_rate": 2.9336915257544106e-05, "loss": 0.0141, "step": 196150 }, { "grad_norm": 0.047477032989263535, "learning_rate": 2.9331896483214494e-05, "loss": 0.0106, "step": 196160 }, { "grad_norm": 0.09987935423851013, "learning_rate": 2.9326877960021836e-05, "loss": 0.0109, "step": 196170 }, { "grad_norm": 0.13157156109809875, "learning_rate": 2.9321859688027132e-05, "loss": 0.0166, "step": 196180 }, { "grad_norm": 0.07493949681520462, "learning_rate": 2.9316841667291356e-05, "loss": 0.0136, "step": 196190 }, { "grad_norm": 0.07354862242937088, "learning_rate": 2.9311823897875503e-05, "loss": 0.0113, "step": 196200 }, { "grad_norm": 0.1063285619020462, "learning_rate": 2.9306806379840502e-05, "loss": 0.0118, "step": 196210 }, { "grad_norm": 0.1293686479330063, "learning_rate": 2.9301789113247334e-05, "loss": 0.012, "step": 196220 }, { "grad_norm": 0.07328727096319199, "learning_rate": 2.9296772098157e-05, "loss": 0.0165, "step": 196230 }, { "grad_norm": 0.10911484062671661, "learning_rate": 2.9291755334630405e-05, "loss": 0.0162, "step": 196240 }, { "grad_norm": 0.057859960943460464, "learning_rate": 2.928673882272854e-05, "loss": 0.0102, "step": 196250 }, { "grad_norm": 0.05796321853995323, "learning_rate": 2.928172256251235e-05, "loss": 0.0097, "step": 196260 }, { "grad_norm": 0.06676974892616272, "learning_rate": 2.9276706554042814e-05, "loss": 0.0135, "step": 196270 }, { "grad_norm": 0.07969058305025101, "learning_rate": 2.927169079738084e-05, "loss": 0.0133, "step": 196280 }, { "grad_norm": 0.055192604660987854, "learning_rate": 2.9266675292587396e-05, "loss": 0.0151, "step": 196290 }, { "grad_norm": 0.06531354784965515, "learning_rate": 2.9261660039723438e-05, "loss": 0.0099, "step": 196300 }, { "grad_norm": 0.12701188027858734, "learning_rate": 2.925664503884987e-05, "loss": 0.0119, "step": 196310 }, { "grad_norm": 0.09655971825122833, "learning_rate": 2.9251630290027654e-05, "loss": 0.0154, "step": 196320 }, { "grad_norm": 0.054292093962430954, "learning_rate": 2.9246615793317723e-05, "loss": 0.0125, "step": 196330 }, { "grad_norm": 0.05534975975751877, "learning_rate": 2.9241601548781016e-05, "loss": 0.0111, "step": 196340 }, { "grad_norm": 0.10144544392824173, "learning_rate": 2.9236587556478435e-05, "loss": 0.0189, "step": 196350 }, { "grad_norm": 0.20355947315692902, "learning_rate": 2.9231573816470914e-05, "loss": 0.0158, "step": 196360 }, { "grad_norm": 0.07367628812789917, "learning_rate": 2.9226560328819396e-05, "loss": 0.0138, "step": 196370 }, { "grad_norm": 0.10731668025255203, "learning_rate": 2.9221547093584777e-05, "loss": 0.0173, "step": 196380 }, { "grad_norm": 0.05631445348262787, "learning_rate": 2.9216534110827976e-05, "loss": 0.0169, "step": 196390 }, { "grad_norm": 0.06793589144945145, "learning_rate": 2.921152138060991e-05, "loss": 0.0106, "step": 196400 }, { "grad_norm": 0.061266206204891205, "learning_rate": 2.9206508902991485e-05, "loss": 0.0141, "step": 196410 }, { "grad_norm": 0.06045805662870407, "learning_rate": 2.920149667803362e-05, "loss": 0.0094, "step": 196420 }, { "grad_norm": 0.08109552413225174, "learning_rate": 2.919648470579719e-05, "loss": 0.0102, "step": 196430 }, { "grad_norm": 0.10987601429224014, "learning_rate": 2.9191472986343138e-05, "loss": 0.0141, "step": 196440 }, { "grad_norm": 0.10641881078481674, "learning_rate": 2.91864615197323e-05, "loss": 0.0106, "step": 196450 }, { "grad_norm": 0.06679897755384445, "learning_rate": 2.9181450306025633e-05, "loss": 0.0133, "step": 196460 }, { "grad_norm": 0.08258429169654846, "learning_rate": 2.917643934528398e-05, "loss": 0.0176, "step": 196470 }, { "grad_norm": 0.05923759192228317, "learning_rate": 2.9171428637568282e-05, "loss": 0.0106, "step": 196480 }, { "grad_norm": 0.05527986213564873, "learning_rate": 2.916641818293936e-05, "loss": 0.0109, "step": 196490 }, { "grad_norm": 0.14380130171775818, "learning_rate": 2.9161407981458144e-05, "loss": 0.0122, "step": 196500 }, { "grad_norm": 0.07849674671888351, "learning_rate": 2.9156398033185472e-05, "loss": 0.0108, "step": 196510 }, { "grad_norm": 0.09349183738231659, "learning_rate": 2.9151388338182282e-05, "loss": 0.0115, "step": 196520 }, { "grad_norm": 0.1216014176607132, "learning_rate": 2.9146378896509374e-05, "loss": 0.0112, "step": 196530 }, { "grad_norm": 0.16526740789413452, "learning_rate": 2.9141369708227673e-05, "loss": 0.0105, "step": 196540 }, { "grad_norm": 0.09485948085784912, "learning_rate": 2.913636077339802e-05, "loss": 0.0157, "step": 196550 }, { "grad_norm": 0.07224380970001221, "learning_rate": 2.9131352092081286e-05, "loss": 0.0152, "step": 196560 }, { "grad_norm": 0.12075314670801163, "learning_rate": 2.912634366433831e-05, "loss": 0.0162, "step": 196570 }, { "grad_norm": 0.11193390190601349, "learning_rate": 2.9121335490229972e-05, "loss": 0.0134, "step": 196580 }, { "grad_norm": 0.07969692349433899, "learning_rate": 2.9116327569817138e-05, "loss": 0.0152, "step": 196590 }, { "grad_norm": 0.061948005110025406, "learning_rate": 2.911131990316063e-05, "loss": 0.0122, "step": 196600 }, { "grad_norm": 0.06234394386410713, "learning_rate": 2.910631249032129e-05, "loss": 0.0115, "step": 196610 }, { "grad_norm": 0.05715988576412201, "learning_rate": 2.9101305331360016e-05, "loss": 0.0146, "step": 196620 }, { "grad_norm": 0.08938008546829224, "learning_rate": 2.9096298426337587e-05, "loss": 0.0144, "step": 196630 }, { "grad_norm": 0.09011559188365936, "learning_rate": 2.9091291775314877e-05, "loss": 0.0112, "step": 196640 }, { "grad_norm": 0.11194532364606857, "learning_rate": 2.9086285378352696e-05, "loss": 0.0115, "step": 196650 }, { "grad_norm": 0.07847779244184494, "learning_rate": 2.908127923551194e-05, "loss": 0.0128, "step": 196660 }, { "grad_norm": 0.06988367438316345, "learning_rate": 2.907627334685334e-05, "loss": 0.0122, "step": 196670 }, { "grad_norm": 0.18558438122272491, "learning_rate": 2.90712677124378e-05, "loss": 0.0133, "step": 196680 }, { "grad_norm": 0.1027863621711731, "learning_rate": 2.9066262332326115e-05, "loss": 0.0112, "step": 196690 }, { "grad_norm": 0.07129154354333878, "learning_rate": 2.9061257206579106e-05, "loss": 0.0111, "step": 196700 }, { "grad_norm": 0.10095677524805069, "learning_rate": 2.905625233525758e-05, "loss": 0.0093, "step": 196710 }, { "grad_norm": 0.07487958669662476, "learning_rate": 2.9051247718422374e-05, "loss": 0.0109, "step": 196720 }, { "grad_norm": 0.09585308283567429, "learning_rate": 2.9046243356134295e-05, "loss": 0.0171, "step": 196730 }, { "grad_norm": 0.07457458227872849, "learning_rate": 2.9041239248454132e-05, "loss": 0.0159, "step": 196740 }, { "grad_norm": 0.12066895514726639, "learning_rate": 2.903623539544269e-05, "loss": 0.0122, "step": 196750 }, { "grad_norm": 0.09948164969682693, "learning_rate": 2.9031231797160795e-05, "loss": 0.0138, "step": 196760 }, { "grad_norm": 0.09628460556268692, "learning_rate": 2.9026228453669234e-05, "loss": 0.0148, "step": 196770 }, { "grad_norm": 0.09586472064256668, "learning_rate": 2.9021225365028802e-05, "loss": 0.0105, "step": 196780 }, { "grad_norm": 0.056593939661979675, "learning_rate": 2.9016222531300263e-05, "loss": 0.0098, "step": 196790 }, { "grad_norm": 0.080986388027668, "learning_rate": 2.901121995254446e-05, "loss": 0.0176, "step": 196800 }, { "grad_norm": 0.07093062251806259, "learning_rate": 2.9006217628822146e-05, "loss": 0.0121, "step": 196810 }, { "grad_norm": 0.05186658725142479, "learning_rate": 2.900121556019411e-05, "loss": 0.0109, "step": 196820 }, { "grad_norm": 0.06155075877904892, "learning_rate": 2.8996213746721113e-05, "loss": 0.0128, "step": 196830 }, { "grad_norm": 0.16408973932266235, "learning_rate": 2.8991212188463967e-05, "loss": 0.0136, "step": 196840 }, { "grad_norm": 0.1058548092842102, "learning_rate": 2.8986210885483433e-05, "loss": 0.0123, "step": 196850 }, { "grad_norm": 0.04871874302625656, "learning_rate": 2.8981209837840272e-05, "loss": 0.0134, "step": 196860 }, { "grad_norm": 0.14546631276607513, "learning_rate": 2.897620904559526e-05, "loss": 0.011, "step": 196870 }, { "grad_norm": 0.07462581247091293, "learning_rate": 2.8971208508809134e-05, "loss": 0.012, "step": 196880 }, { "grad_norm": 0.1825595498085022, "learning_rate": 2.89662082275427e-05, "loss": 0.0141, "step": 196890 }, { "grad_norm": 0.12118816375732422, "learning_rate": 2.8961208201856698e-05, "loss": 0.0117, "step": 196900 }, { "grad_norm": 0.05291849002242088, "learning_rate": 2.895620843181188e-05, "loss": 0.0123, "step": 196910 }, { "grad_norm": 0.10606470704078674, "learning_rate": 2.8951208917468975e-05, "loss": 0.0141, "step": 196920 }, { "grad_norm": 0.15467417240142822, "learning_rate": 2.8946209658888778e-05, "loss": 0.0198, "step": 196930 }, { "grad_norm": 0.12064540386199951, "learning_rate": 2.8941210656132013e-05, "loss": 0.0139, "step": 196940 }, { "grad_norm": 0.09307267516851425, "learning_rate": 2.8936211909259416e-05, "loss": 0.0095, "step": 196950 }, { "grad_norm": 0.08684470504522324, "learning_rate": 2.8931213418331714e-05, "loss": 0.0142, "step": 196960 }, { "grad_norm": 0.0816081240773201, "learning_rate": 2.8926215183409683e-05, "loss": 0.0134, "step": 196970 }, { "grad_norm": 0.09014184027910233, "learning_rate": 2.8921217204554035e-05, "loss": 0.0169, "step": 196980 }, { "grad_norm": 0.08446343243122101, "learning_rate": 2.89162194818255e-05, "loss": 0.0087, "step": 196990 }, { "grad_norm": 0.09398844838142395, "learning_rate": 2.891122201528478e-05, "loss": 0.0113, "step": 197000 }, { "grad_norm": 0.05727965384721756, "learning_rate": 2.8906224804992666e-05, "loss": 0.0117, "step": 197010 }, { "grad_norm": 0.05878710001707077, "learning_rate": 2.89012278510098e-05, "loss": 0.0112, "step": 197020 }, { "grad_norm": 0.12821024656295776, "learning_rate": 2.8896231153396953e-05, "loss": 0.0171, "step": 197030 }, { "grad_norm": 0.12244030088186264, "learning_rate": 2.8891234712214798e-05, "loss": 0.0116, "step": 197040 }, { "grad_norm": 0.06968706101179123, "learning_rate": 2.8886238527524122e-05, "loss": 0.011, "step": 197050 }, { "grad_norm": 0.49611034989356995, "learning_rate": 2.8881242599385534e-05, "loss": 0.0153, "step": 197060 }, { "grad_norm": 0.1662241667509079, "learning_rate": 2.8876246927859808e-05, "loss": 0.012, "step": 197070 }, { "grad_norm": 0.21164382994174957, "learning_rate": 2.887125151300763e-05, "loss": 0.0202, "step": 197080 }, { "grad_norm": 0.09697063267230988, "learning_rate": 2.88662563548897e-05, "loss": 0.0098, "step": 197090 }, { "grad_norm": 0.1397366225719452, "learning_rate": 2.886126145356668e-05, "loss": 0.0125, "step": 197100 }, { "grad_norm": 0.12081081420183182, "learning_rate": 2.885626680909931e-05, "loss": 0.0144, "step": 197110 }, { "grad_norm": 0.16208110749721527, "learning_rate": 2.8851272421548263e-05, "loss": 0.0166, "step": 197120 }, { "grad_norm": 0.06687392294406891, "learning_rate": 2.8846278290974227e-05, "loss": 0.016, "step": 197130 }, { "grad_norm": 0.10574277490377426, "learning_rate": 2.8841284417437864e-05, "loss": 0.0131, "step": 197140 }, { "grad_norm": 0.1023416593670845, "learning_rate": 2.8836290800999888e-05, "loss": 0.0105, "step": 197150 }, { "grad_norm": 0.07853226363658905, "learning_rate": 2.883129744172096e-05, "loss": 0.0162, "step": 197160 }, { "grad_norm": 0.07446834444999695, "learning_rate": 2.8826304339661757e-05, "loss": 0.0074, "step": 197170 }, { "grad_norm": 0.10332591086626053, "learning_rate": 2.882131149488293e-05, "loss": 0.0132, "step": 197180 }, { "grad_norm": 0.16235919296741486, "learning_rate": 2.88163189074452e-05, "loss": 0.0149, "step": 197190 }, { "grad_norm": 0.053974997252225876, "learning_rate": 2.8811326577409157e-05, "loss": 0.0119, "step": 197200 }, { "grad_norm": 0.09485036134719849, "learning_rate": 2.8806334504835525e-05, "loss": 0.015, "step": 197210 }, { "grad_norm": 0.06786736100912094, "learning_rate": 2.8801342689784916e-05, "loss": 0.0161, "step": 197220 }, { "grad_norm": 0.07049494236707687, "learning_rate": 2.8796351132318046e-05, "loss": 0.0157, "step": 197230 }, { "grad_norm": 0.05725189670920372, "learning_rate": 2.879135983249549e-05, "loss": 0.0163, "step": 197240 }, { "grad_norm": 0.11751303821802139, "learning_rate": 2.878636879037796e-05, "loss": 0.0103, "step": 197250 }, { "grad_norm": 0.18038605153560638, "learning_rate": 2.878137800602608e-05, "loss": 0.0101, "step": 197260 }, { "grad_norm": 0.07607915997505188, "learning_rate": 2.877638747950049e-05, "loss": 0.0111, "step": 197270 }, { "grad_norm": 0.10274843126535416, "learning_rate": 2.8771397210861807e-05, "loss": 0.0098, "step": 197280 }, { "grad_norm": 0.26854920387268066, "learning_rate": 2.876640720017071e-05, "loss": 0.0154, "step": 197290 }, { "grad_norm": 0.2603137493133545, "learning_rate": 2.8761417447487816e-05, "loss": 0.0115, "step": 197300 }, { "grad_norm": 0.05692819133400917, "learning_rate": 2.8756427952873744e-05, "loss": 0.0115, "step": 197310 }, { "grad_norm": 0.22050577402114868, "learning_rate": 2.8751438716389124e-05, "loss": 0.0147, "step": 197320 }, { "grad_norm": 0.08343278616666794, "learning_rate": 2.8746449738094617e-05, "loss": 0.0173, "step": 197330 }, { "grad_norm": 0.08853666484355927, "learning_rate": 2.8741461018050776e-05, "loss": 0.0191, "step": 197340 }, { "grad_norm": 0.08269936591386795, "learning_rate": 2.8736472556318272e-05, "loss": 0.0155, "step": 197350 }, { "grad_norm": 0.06690890341997147, "learning_rate": 2.8731484352957684e-05, "loss": 0.0143, "step": 197360 }, { "grad_norm": 0.05760273337364197, "learning_rate": 2.8726496408029678e-05, "loss": 0.0141, "step": 197370 }, { "grad_norm": 0.13820545375347137, "learning_rate": 2.8721508721594792e-05, "loss": 0.0111, "step": 197380 }, { "grad_norm": 0.08887473493814468, "learning_rate": 2.871652129371368e-05, "loss": 0.0098, "step": 197390 }, { "grad_norm": 0.12141834199428558, "learning_rate": 2.8711534124446915e-05, "loss": 0.0124, "step": 197400 }, { "grad_norm": 0.12339719384908676, "learning_rate": 2.8706547213855144e-05, "loss": 0.0099, "step": 197410 }, { "grad_norm": 0.0956655889749527, "learning_rate": 2.870156056199889e-05, "loss": 0.0129, "step": 197420 }, { "grad_norm": 0.12078306823968887, "learning_rate": 2.8696574168938805e-05, "loss": 0.0134, "step": 197430 }, { "grad_norm": 0.1384868323802948, "learning_rate": 2.869158803473546e-05, "loss": 0.0178, "step": 197440 }, { "grad_norm": 0.05511803179979324, "learning_rate": 2.868660215944944e-05, "loss": 0.0197, "step": 197450 }, { "grad_norm": 0.061922844499349594, "learning_rate": 2.8681616543141304e-05, "loss": 0.0134, "step": 197460 }, { "grad_norm": 0.07743890583515167, "learning_rate": 2.867663118587167e-05, "loss": 0.0137, "step": 197470 }, { "grad_norm": 0.12027987837791443, "learning_rate": 2.8671646087701108e-05, "loss": 0.0117, "step": 197480 }, { "grad_norm": 0.11274606734514236, "learning_rate": 2.8666661248690186e-05, "loss": 0.0123, "step": 197490 }, { "grad_norm": 0.07941561937332153, "learning_rate": 2.866167666889944e-05, "loss": 0.0094, "step": 197500 }, { "grad_norm": 0.20392821729183197, "learning_rate": 2.86566923483895e-05, "loss": 0.0142, "step": 197510 }, { "grad_norm": 0.1433165818452835, "learning_rate": 2.865170828722089e-05, "loss": 0.0128, "step": 197520 }, { "grad_norm": 0.046346262097358704, "learning_rate": 2.864672448545419e-05, "loss": 0.0155, "step": 197530 }, { "grad_norm": 0.1494833081960678, "learning_rate": 2.864174094314992e-05, "loss": 0.014, "step": 197540 }, { "grad_norm": 0.0821707472205162, "learning_rate": 2.863675766036869e-05, "loss": 0.0159, "step": 197550 }, { "grad_norm": 0.09468613564968109, "learning_rate": 2.8631774637171022e-05, "loss": 0.0124, "step": 197560 }, { "grad_norm": 0.0507422611117363, "learning_rate": 2.8626791873617463e-05, "loss": 0.0119, "step": 197570 }, { "grad_norm": 0.13963837921619415, "learning_rate": 2.862180936976857e-05, "loss": 0.0235, "step": 197580 }, { "grad_norm": 0.07337971031665802, "learning_rate": 2.8616827125684853e-05, "loss": 0.0143, "step": 197590 }, { "grad_norm": 0.18910035490989685, "learning_rate": 2.8611845141426895e-05, "loss": 0.0109, "step": 197600 }, { "grad_norm": 0.13446994125843048, "learning_rate": 2.8606863417055212e-05, "loss": 0.014, "step": 197610 }, { "grad_norm": 0.09381727129220963, "learning_rate": 2.860188195263034e-05, "loss": 0.0111, "step": 197620 }, { "grad_norm": 0.29467055201530457, "learning_rate": 2.8596900748212785e-05, "loss": 0.015, "step": 197630 }, { "grad_norm": 0.0740220844745636, "learning_rate": 2.859191980386311e-05, "loss": 0.0214, "step": 197640 }, { "grad_norm": 0.08979740738868713, "learning_rate": 2.858693911964182e-05, "loss": 0.0183, "step": 197650 }, { "grad_norm": 0.07375887781381607, "learning_rate": 2.8581958695609447e-05, "loss": 0.0207, "step": 197660 }, { "grad_norm": 0.1271897256374359, "learning_rate": 2.857697853182647e-05, "loss": 0.0123, "step": 197670 }, { "grad_norm": 0.14510871469974518, "learning_rate": 2.8571998628353453e-05, "loss": 0.0115, "step": 197680 }, { "grad_norm": 0.08664453774690628, "learning_rate": 2.856701898525088e-05, "loss": 0.0105, "step": 197690 }, { "grad_norm": 0.12518896162509918, "learning_rate": 2.856203960257927e-05, "loss": 0.0164, "step": 197700 }, { "grad_norm": 0.0838671326637268, "learning_rate": 2.8557060480399095e-05, "loss": 0.0141, "step": 197710 }, { "grad_norm": 0.09289126843214035, "learning_rate": 2.85520816187709e-05, "loss": 0.0168, "step": 197720 }, { "grad_norm": 0.060838483273983, "learning_rate": 2.8547103017755167e-05, "loss": 0.016, "step": 197730 }, { "grad_norm": 0.19762317836284637, "learning_rate": 2.8542124677412385e-05, "loss": 0.0105, "step": 197740 }, { "grad_norm": 0.06662491708993912, "learning_rate": 2.8537146597803027e-05, "loss": 0.0085, "step": 197750 }, { "grad_norm": 0.052644699811935425, "learning_rate": 2.8532168778987643e-05, "loss": 0.0124, "step": 197760 }, { "grad_norm": 0.08097334951162338, "learning_rate": 2.852719122102664e-05, "loss": 0.0115, "step": 197770 }, { "grad_norm": 0.1321706771850586, "learning_rate": 2.8522213923980556e-05, "loss": 0.0103, "step": 197780 }, { "grad_norm": 0.14280973374843597, "learning_rate": 2.8517236887909826e-05, "loss": 0.0119, "step": 197790 }, { "grad_norm": 0.06202089414000511, "learning_rate": 2.8512260112874996e-05, "loss": 0.0169, "step": 197800 }, { "grad_norm": 0.4394383728504181, "learning_rate": 2.8507283598936452e-05, "loss": 0.0141, "step": 197810 }, { "grad_norm": 0.07752896845340729, "learning_rate": 2.850230734615472e-05, "loss": 0.0127, "step": 197820 }, { "grad_norm": 0.06991366297006607, "learning_rate": 2.8497331354590262e-05, "loss": 0.0123, "step": 197830 }, { "grad_norm": 0.12742002308368683, "learning_rate": 2.8492355624303525e-05, "loss": 0.016, "step": 197840 }, { "grad_norm": 0.16388118267059326, "learning_rate": 2.848738015535495e-05, "loss": 0.0155, "step": 197850 }, { "grad_norm": 0.08288360387086868, "learning_rate": 2.848240494780504e-05, "loss": 0.0117, "step": 197860 }, { "grad_norm": 0.07754620909690857, "learning_rate": 2.8477430001714223e-05, "loss": 0.0139, "step": 197870 }, { "grad_norm": 0.2514651119709015, "learning_rate": 2.8472455317142955e-05, "loss": 0.014, "step": 197880 }, { "grad_norm": 0.14849454164505005, "learning_rate": 2.8467480894151666e-05, "loss": 0.0123, "step": 197890 }, { "grad_norm": 0.05825161561369896, "learning_rate": 2.846250673280084e-05, "loss": 0.0165, "step": 197900 }, { "grad_norm": 0.06142051890492439, "learning_rate": 2.8457532833150867e-05, "loss": 0.0159, "step": 197910 }, { "grad_norm": 0.07239852845668793, "learning_rate": 2.845255919526222e-05, "loss": 0.0101, "step": 197920 }, { "grad_norm": 0.07645361125469208, "learning_rate": 2.8447585819195312e-05, "loss": 0.0136, "step": 197930 }, { "grad_norm": 0.05894743651151657, "learning_rate": 2.844261270501062e-05, "loss": 0.01, "step": 197940 }, { "grad_norm": 0.2830185294151306, "learning_rate": 2.8437639852768496e-05, "loss": 0.0182, "step": 197950 }, { "grad_norm": 0.0863175243139267, "learning_rate": 2.843266726252942e-05, "loss": 0.0182, "step": 197960 }, { "grad_norm": 0.1945209801197052, "learning_rate": 2.842769493435381e-05, "loss": 0.0147, "step": 197970 }, { "grad_norm": 0.04515886679291725, "learning_rate": 2.8422722868302065e-05, "loss": 0.0132, "step": 197980 }, { "grad_norm": 0.12133122235536575, "learning_rate": 2.8417751064434596e-05, "loss": 0.0144, "step": 197990 }, { "grad_norm": 0.11540612578392029, "learning_rate": 2.8412779522811848e-05, "loss": 0.0097, "step": 198000 }, { "grad_norm": 0.22337503731250763, "learning_rate": 2.840780824349421e-05, "loss": 0.0083, "step": 198010 }, { "grad_norm": 0.05321527644991875, "learning_rate": 2.840283722654209e-05, "loss": 0.0083, "step": 198020 }, { "grad_norm": 0.04878682270646095, "learning_rate": 2.8397866472015867e-05, "loss": 0.0143, "step": 198030 }, { "grad_norm": 0.06322353333234787, "learning_rate": 2.8392895979975975e-05, "loss": 0.0145, "step": 198040 }, { "grad_norm": 0.09000475704669952, "learning_rate": 2.8387925750482807e-05, "loss": 0.0166, "step": 198050 }, { "grad_norm": 0.07893607765436172, "learning_rate": 2.838295578359674e-05, "loss": 0.0137, "step": 198060 }, { "grad_norm": 0.18930424749851227, "learning_rate": 2.837798607937816e-05, "loss": 0.0137, "step": 198070 }, { "grad_norm": 0.07423602044582367, "learning_rate": 2.8373016637887497e-05, "loss": 0.0098, "step": 198080 }, { "grad_norm": 0.0883326530456543, "learning_rate": 2.836804745918506e-05, "loss": 0.0103, "step": 198090 }, { "grad_norm": 0.10725565999746323, "learning_rate": 2.836307854333129e-05, "loss": 0.0128, "step": 198100 }, { "grad_norm": 0.07294309884309769, "learning_rate": 2.835810989038653e-05, "loss": 0.0132, "step": 198110 }, { "grad_norm": 0.07152645289897919, "learning_rate": 2.8353141500411202e-05, "loss": 0.0132, "step": 198120 }, { "grad_norm": 0.09688978642225266, "learning_rate": 2.8348173373465602e-05, "loss": 0.0107, "step": 198130 }, { "grad_norm": 0.234793558716774, "learning_rate": 2.8343205509610166e-05, "loss": 0.0162, "step": 198140 }, { "grad_norm": 0.09284241497516632, "learning_rate": 2.8338237908905225e-05, "loss": 0.0162, "step": 198150 }, { "grad_norm": 0.18438315391540527, "learning_rate": 2.8333270571411147e-05, "loss": 0.0147, "step": 198160 }, { "grad_norm": 0.10766971111297607, "learning_rate": 2.8328303497188268e-05, "loss": 0.0107, "step": 198170 }, { "grad_norm": 0.2219102531671524, "learning_rate": 2.8323336686296982e-05, "loss": 0.012, "step": 198180 }, { "grad_norm": 0.11406802386045456, "learning_rate": 2.8318370138797622e-05, "loss": 0.0125, "step": 198190 }, { "grad_norm": 0.08171200752258301, "learning_rate": 2.831340385475053e-05, "loss": 0.0116, "step": 198200 }, { "grad_norm": 0.08329711109399796, "learning_rate": 2.830843783421605e-05, "loss": 0.0117, "step": 198210 }, { "grad_norm": 0.08249317109584808, "learning_rate": 2.830347207725454e-05, "loss": 0.0162, "step": 198220 }, { "grad_norm": 0.10292914509773254, "learning_rate": 2.8298506583926333e-05, "loss": 0.0162, "step": 198230 }, { "grad_norm": 0.14515379071235657, "learning_rate": 2.829354135429176e-05, "loss": 0.014, "step": 198240 }, { "grad_norm": 0.10262659192085266, "learning_rate": 2.8288576388411132e-05, "loss": 0.0171, "step": 198250 }, { "grad_norm": 0.12631994485855103, "learning_rate": 2.8283611686344825e-05, "loss": 0.0159, "step": 198260 }, { "grad_norm": 0.10347571969032288, "learning_rate": 2.8278647248153134e-05, "loss": 0.0218, "step": 198270 }, { "grad_norm": 0.07088945060968399, "learning_rate": 2.8273683073896384e-05, "loss": 0.0102, "step": 198280 }, { "grad_norm": 0.08246149867773056, "learning_rate": 2.8268719163634888e-05, "loss": 0.0167, "step": 198290 }, { "grad_norm": 0.13408401608467102, "learning_rate": 2.8263755517428987e-05, "loss": 0.0154, "step": 198300 }, { "grad_norm": 0.12043865025043488, "learning_rate": 2.8258792135338974e-05, "loss": 0.0122, "step": 198310 }, { "grad_norm": 0.2688068151473999, "learning_rate": 2.8253829017425172e-05, "loss": 0.0127, "step": 198320 }, { "grad_norm": 0.17705829441547394, "learning_rate": 2.8248866163747878e-05, "loss": 0.015, "step": 198330 }, { "grad_norm": 0.08623620867729187, "learning_rate": 2.824390357436737e-05, "loss": 0.0107, "step": 198340 }, { "grad_norm": 0.07632762938737869, "learning_rate": 2.8238941249343996e-05, "loss": 0.0139, "step": 198350 }, { "grad_norm": 0.062102071940898895, "learning_rate": 2.823397918873804e-05, "loss": 0.0144, "step": 198360 }, { "grad_norm": 0.06179801747202873, "learning_rate": 2.822901739260978e-05, "loss": 0.0112, "step": 198370 }, { "grad_norm": 0.1505834460258484, "learning_rate": 2.82240558610195e-05, "loss": 0.0106, "step": 198380 }, { "grad_norm": 0.06471798568964005, "learning_rate": 2.821909459402752e-05, "loss": 0.0185, "step": 198390 }, { "grad_norm": 0.08497530221939087, "learning_rate": 2.8214133591694098e-05, "loss": 0.0149, "step": 198400 }, { "grad_norm": 0.08438640087842941, "learning_rate": 2.8209172854079524e-05, "loss": 0.0131, "step": 198410 }, { "grad_norm": 0.08358771353960037, "learning_rate": 2.8204212381244054e-05, "loss": 0.0113, "step": 198420 }, { "grad_norm": 0.10579011589288712, "learning_rate": 2.8199252173248007e-05, "loss": 0.0124, "step": 198430 }, { "grad_norm": 0.07492579519748688, "learning_rate": 2.8194292230151632e-05, "loss": 0.0099, "step": 198440 }, { "grad_norm": 0.08628899604082108, "learning_rate": 2.8189332552015192e-05, "loss": 0.0113, "step": 198450 }, { "grad_norm": 0.1043139100074768, "learning_rate": 2.8184373138898923e-05, "loss": 0.0115, "step": 198460 }, { "grad_norm": 0.11232662945985794, "learning_rate": 2.817941399086317e-05, "loss": 0.015, "step": 198470 }, { "grad_norm": 0.06731437891721725, "learning_rate": 2.8174455107968094e-05, "loss": 0.0128, "step": 198480 }, { "grad_norm": 0.19342678785324097, "learning_rate": 2.8169496490274018e-05, "loss": 0.0144, "step": 198490 }, { "grad_norm": 0.07195664197206497, "learning_rate": 2.8164538137841146e-05, "loss": 0.0125, "step": 198500 }, { "grad_norm": 0.06177019327878952, "learning_rate": 2.8159580050729795e-05, "loss": 0.0096, "step": 198510 }, { "grad_norm": 0.16720253229141235, "learning_rate": 2.8154622229000128e-05, "loss": 0.0176, "step": 198520 }, { "grad_norm": 0.08001359552145004, "learning_rate": 2.814966467271245e-05, "loss": 0.0125, "step": 198530 }, { "grad_norm": 0.07177598774433136, "learning_rate": 2.814470738192697e-05, "loss": 0.0148, "step": 198540 }, { "grad_norm": 0.18563251197338104, "learning_rate": 2.8139750356703932e-05, "loss": 0.0137, "step": 198550 }, { "grad_norm": 0.1509101837873459, "learning_rate": 2.8134793597103548e-05, "loss": 0.0122, "step": 198560 }, { "grad_norm": 0.10112090408802032, "learning_rate": 2.8129837103186086e-05, "loss": 0.0101, "step": 198570 }, { "grad_norm": 0.1108926460146904, "learning_rate": 2.812488087501175e-05, "loss": 0.0147, "step": 198580 }, { "grad_norm": 0.19484129548072815, "learning_rate": 2.8119924912640766e-05, "loss": 0.0129, "step": 198590 }, { "grad_norm": 0.12368015944957733, "learning_rate": 2.8114969216133334e-05, "loss": 0.0156, "step": 198600 }, { "grad_norm": 0.05035683512687683, "learning_rate": 2.8110013785549705e-05, "loss": 0.0093, "step": 198610 }, { "grad_norm": 0.054728057235479355, "learning_rate": 2.8105058620950077e-05, "loss": 0.0088, "step": 198620 }, { "grad_norm": 0.06494411826133728, "learning_rate": 2.8100103722394662e-05, "loss": 0.0123, "step": 198630 }, { "grad_norm": 0.07025399059057236, "learning_rate": 2.809514908994364e-05, "loss": 0.0094, "step": 198640 }, { "grad_norm": 0.09794454276561737, "learning_rate": 2.809019472365728e-05, "loss": 0.0126, "step": 198650 }, { "grad_norm": 0.10962457209825516, "learning_rate": 2.8085240623595698e-05, "loss": 0.0099, "step": 198660 }, { "grad_norm": 0.09797685593366623, "learning_rate": 2.808028678981916e-05, "loss": 0.0137, "step": 198670 }, { "grad_norm": 0.05079331994056702, "learning_rate": 2.807533322238781e-05, "loss": 0.0092, "step": 198680 }, { "grad_norm": 0.0839480385184288, "learning_rate": 2.8070379921361902e-05, "loss": 0.0139, "step": 198690 }, { "grad_norm": 0.11146638542413712, "learning_rate": 2.806542688680154e-05, "loss": 0.0118, "step": 198700 }, { "grad_norm": 0.07211994379758835, "learning_rate": 2.8060474118766978e-05, "loss": 0.0126, "step": 198710 }, { "grad_norm": 0.07436840236186981, "learning_rate": 2.8055521617318357e-05, "loss": 0.0131, "step": 198720 }, { "grad_norm": 0.1034180223941803, "learning_rate": 2.805056938251588e-05, "loss": 0.0106, "step": 198730 }, { "grad_norm": 0.19936862587928772, "learning_rate": 2.804561741441968e-05, "loss": 0.0114, "step": 198740 }, { "grad_norm": 0.16086509823799133, "learning_rate": 2.8040665713089982e-05, "loss": 0.0122, "step": 198750 }, { "grad_norm": 0.10470753163099289, "learning_rate": 2.803571427858693e-05, "loss": 0.009, "step": 198760 }, { "grad_norm": 0.09001333266496658, "learning_rate": 2.8030763110970682e-05, "loss": 0.0099, "step": 198770 }, { "grad_norm": 0.06226348504424095, "learning_rate": 2.8025812210301382e-05, "loss": 0.0089, "step": 198780 }, { "grad_norm": 0.07784715294837952, "learning_rate": 2.8020861576639256e-05, "loss": 0.0125, "step": 198790 }, { "grad_norm": 0.10679129511117935, "learning_rate": 2.8015911210044378e-05, "loss": 0.0132, "step": 198800 }, { "grad_norm": 0.10143927484750748, "learning_rate": 2.801096111057695e-05, "loss": 0.0177, "step": 198810 }, { "grad_norm": 0.08041594177484512, "learning_rate": 2.8006011278297085e-05, "loss": 0.0137, "step": 198820 }, { "grad_norm": 0.20818117260932922, "learning_rate": 2.8001061713264986e-05, "loss": 0.0174, "step": 198830 }, { "grad_norm": 0.10041085630655289, "learning_rate": 2.7996112415540727e-05, "loss": 0.0166, "step": 198840 }, { "grad_norm": 0.1020314022898674, "learning_rate": 2.7991163385184495e-05, "loss": 0.01, "step": 198850 }, { "grad_norm": 0.11953981965780258, "learning_rate": 2.79862146222564e-05, "loss": 0.0144, "step": 198860 }, { "grad_norm": 0.0942995697259903, "learning_rate": 2.7981266126816587e-05, "loss": 0.0117, "step": 198870 }, { "grad_norm": 0.2253362536430359, "learning_rate": 2.797631789892516e-05, "loss": 0.016, "step": 198880 }, { "grad_norm": 0.04835544154047966, "learning_rate": 2.7971369938642278e-05, "loss": 0.0124, "step": 198890 }, { "grad_norm": 0.08975034952163696, "learning_rate": 2.7966422246028055e-05, "loss": 0.0152, "step": 198900 }, { "grad_norm": 0.054419152438640594, "learning_rate": 2.7961474821142608e-05, "loss": 0.0134, "step": 198910 }, { "grad_norm": 0.09043745696544647, "learning_rate": 2.7956527664046027e-05, "loss": 0.0129, "step": 198920 }, { "grad_norm": 0.10672514140605927, "learning_rate": 2.795158077479847e-05, "loss": 0.013, "step": 198930 }, { "grad_norm": 0.05063808709383011, "learning_rate": 2.7946634153460016e-05, "loss": 0.0101, "step": 198940 }, { "grad_norm": 0.09380925446748734, "learning_rate": 2.7941687800090783e-05, "loss": 0.0123, "step": 198950 }, { "grad_norm": 0.11888062208890915, "learning_rate": 2.7936741714750848e-05, "loss": 0.0142, "step": 198960 }, { "grad_norm": 0.09825782477855682, "learning_rate": 2.7931795897500358e-05, "loss": 0.0159, "step": 198970 }, { "grad_norm": 0.05621163919568062, "learning_rate": 2.792685034839938e-05, "loss": 0.014, "step": 198980 }, { "grad_norm": 0.14386478066444397, "learning_rate": 2.7921905067508012e-05, "loss": 0.0088, "step": 198990 }, { "grad_norm": 0.056766316294670105, "learning_rate": 2.791696005488632e-05, "loss": 0.0093, "step": 199000 }, { "grad_norm": 0.1021701842546463, "learning_rate": 2.7912015310594436e-05, "loss": 0.0164, "step": 199010 }, { "grad_norm": 0.07812493294477463, "learning_rate": 2.7907070834692416e-05, "loss": 0.0085, "step": 199020 }, { "grad_norm": 0.11085664480924606, "learning_rate": 2.790212662724035e-05, "loss": 0.0124, "step": 199030 }, { "grad_norm": 0.05824935436248779, "learning_rate": 2.7897182688298307e-05, "loss": 0.0143, "step": 199040 }, { "grad_norm": 0.06809072941541672, "learning_rate": 2.789223901792634e-05, "loss": 0.0121, "step": 199050 }, { "grad_norm": 0.0616694800555706, "learning_rate": 2.7887295616184566e-05, "loss": 0.0159, "step": 199060 }, { "grad_norm": 0.5749078392982483, "learning_rate": 2.788235248313303e-05, "loss": 0.0136, "step": 199070 }, { "grad_norm": 0.05511048436164856, "learning_rate": 2.787740961883179e-05, "loss": 0.0115, "step": 199080 }, { "grad_norm": 0.12152931094169617, "learning_rate": 2.787246702334089e-05, "loss": 0.0181, "step": 199090 }, { "grad_norm": 0.10713773965835571, "learning_rate": 2.7867524696720425e-05, "loss": 0.0126, "step": 199100 }, { "grad_norm": 0.1644231081008911, "learning_rate": 2.7862582639030445e-05, "loss": 0.0108, "step": 199110 }, { "grad_norm": 0.06170474365353584, "learning_rate": 2.7857640850330974e-05, "loss": 0.01, "step": 199120 }, { "grad_norm": 0.08261367678642273, "learning_rate": 2.785269933068205e-05, "loss": 0.0124, "step": 199130 }, { "grad_norm": 0.07368739694356918, "learning_rate": 2.784775808014377e-05, "loss": 0.0113, "step": 199140 }, { "grad_norm": 0.06162931025028229, "learning_rate": 2.784281709877614e-05, "loss": 0.0127, "step": 199150 }, { "grad_norm": 0.08124223351478577, "learning_rate": 2.7837876386639205e-05, "loss": 0.0174, "step": 199160 }, { "grad_norm": 0.37043729424476624, "learning_rate": 2.783293594379297e-05, "loss": 0.0087, "step": 199170 }, { "grad_norm": 0.05454012751579285, "learning_rate": 2.7827995770297537e-05, "loss": 0.0148, "step": 199180 }, { "grad_norm": 0.08410772681236267, "learning_rate": 2.7823055866212848e-05, "loss": 0.013, "step": 199190 }, { "grad_norm": 0.18869160115718842, "learning_rate": 2.781811623159899e-05, "loss": 0.0114, "step": 199200 }, { "grad_norm": 0.09329104423522949, "learning_rate": 2.7813176866515943e-05, "loss": 0.0139, "step": 199210 }, { "grad_norm": 0.0805583968758583, "learning_rate": 2.7808237771023782e-05, "loss": 0.0114, "step": 199220 }, { "grad_norm": 0.09872692078351974, "learning_rate": 2.7803298945182443e-05, "loss": 0.0162, "step": 199230 }, { "grad_norm": 0.18512649834156036, "learning_rate": 2.7798360389052003e-05, "loss": 0.0163, "step": 199240 }, { "grad_norm": 0.22890453040599823, "learning_rate": 2.779342210269243e-05, "loss": 0.0135, "step": 199250 }, { "grad_norm": 0.07785895466804504, "learning_rate": 2.778848408616378e-05, "loss": 0.0078, "step": 199260 }, { "grad_norm": 0.04936693236231804, "learning_rate": 2.778354633952598e-05, "loss": 0.0113, "step": 199270 }, { "grad_norm": 0.09360210597515106, "learning_rate": 2.7778608862839084e-05, "loss": 0.0115, "step": 199280 }, { "grad_norm": 0.15696588158607483, "learning_rate": 2.777367165616308e-05, "loss": 0.0127, "step": 199290 }, { "grad_norm": 0.07925073057413101, "learning_rate": 2.776873471955794e-05, "loss": 0.0148, "step": 199300 }, { "grad_norm": 0.17287202179431915, "learning_rate": 2.776379805308365e-05, "loss": 0.0137, "step": 199310 }, { "grad_norm": 0.08098938316106796, "learning_rate": 2.775886165680023e-05, "loss": 0.011, "step": 199320 }, { "grad_norm": 0.08105290681123734, "learning_rate": 2.7753925530767633e-05, "loss": 0.01, "step": 199330 }, { "grad_norm": 0.0781637579202652, "learning_rate": 2.7748989675045843e-05, "loss": 0.011, "step": 199340 }, { "grad_norm": 0.10682343691587448, "learning_rate": 2.7744054089694828e-05, "loss": 0.0111, "step": 199350 }, { "grad_norm": 0.31772124767303467, "learning_rate": 2.77391187747746e-05, "loss": 0.0149, "step": 199360 }, { "grad_norm": 0.07747769355773926, "learning_rate": 2.7734183730345058e-05, "loss": 0.0114, "step": 199370 }, { "grad_norm": 0.053203657269477844, "learning_rate": 2.772924895646623e-05, "loss": 0.0087, "step": 199380 }, { "grad_norm": 0.0748036801815033, "learning_rate": 2.772431445319803e-05, "loss": 0.0125, "step": 199390 }, { "grad_norm": 0.06860906630754471, "learning_rate": 2.771938022060048e-05, "loss": 0.0113, "step": 199400 }, { "grad_norm": 0.21072465181350708, "learning_rate": 2.771444625873346e-05, "loss": 0.0195, "step": 199410 }, { "grad_norm": 0.1602131873369217, "learning_rate": 2.770951256765698e-05, "loss": 0.0106, "step": 199420 }, { "grad_norm": 0.0679631233215332, "learning_rate": 2.7704579147430965e-05, "loss": 0.0098, "step": 199430 }, { "grad_norm": 0.10269494354724884, "learning_rate": 2.7699645998115366e-05, "loss": 0.0131, "step": 199440 }, { "grad_norm": 0.09324628114700317, "learning_rate": 2.7694713119770106e-05, "loss": 0.0152, "step": 199450 }, { "grad_norm": 0.10113801062107086, "learning_rate": 2.7689780512455156e-05, "loss": 0.0143, "step": 199460 }, { "grad_norm": 0.10070336610078812, "learning_rate": 2.7684848176230437e-05, "loss": 0.0142, "step": 199470 }, { "grad_norm": 0.08810260146856308, "learning_rate": 2.7679916111155886e-05, "loss": 0.0131, "step": 199480 }, { "grad_norm": 0.2969464361667633, "learning_rate": 2.76749843172914e-05, "loss": 0.0083, "step": 199490 }, { "grad_norm": 0.09994868189096451, "learning_rate": 2.767005279469696e-05, "loss": 0.0137, "step": 199500 }, { "grad_norm": 0.08384473621845245, "learning_rate": 2.766512154343246e-05, "loss": 0.0145, "step": 199510 }, { "grad_norm": 0.16278301179409027, "learning_rate": 2.766019056355782e-05, "loss": 0.0297, "step": 199520 }, { "grad_norm": 0.07579148560762405, "learning_rate": 2.765525985513294e-05, "loss": 0.0145, "step": 199530 }, { "grad_norm": 0.17910172045230865, "learning_rate": 2.7650329418217786e-05, "loss": 0.0123, "step": 199540 }, { "grad_norm": 0.05241759121417999, "learning_rate": 2.7645399252872195e-05, "loss": 0.0107, "step": 199550 }, { "grad_norm": 0.05815223604440689, "learning_rate": 2.7640469359156124e-05, "loss": 0.0113, "step": 199560 }, { "grad_norm": 0.1058991551399231, "learning_rate": 2.7635539737129447e-05, "loss": 0.0163, "step": 199570 }, { "grad_norm": 0.07843251526355743, "learning_rate": 2.7630610386852107e-05, "loss": 0.0175, "step": 199580 }, { "grad_norm": 0.06958871334791183, "learning_rate": 2.7625681308383944e-05, "loss": 0.013, "step": 199590 }, { "grad_norm": 0.055795442312955856, "learning_rate": 2.762075250178489e-05, "loss": 0.0138, "step": 199600 }, { "grad_norm": 0.22874337434768677, "learning_rate": 2.7615823967114823e-05, "loss": 0.0191, "step": 199610 }, { "grad_norm": 0.09153883159160614, "learning_rate": 2.7610895704433632e-05, "loss": 0.0125, "step": 199620 }, { "grad_norm": 0.12981484830379486, "learning_rate": 2.7605967713801173e-05, "loss": 0.0112, "step": 199630 }, { "grad_norm": 0.07311660051345825, "learning_rate": 2.760103999527738e-05, "loss": 0.0115, "step": 199640 }, { "grad_norm": 0.08627527952194214, "learning_rate": 2.7596112548922088e-05, "loss": 0.0141, "step": 199650 }, { "grad_norm": 0.07954037189483643, "learning_rate": 2.7591185374795186e-05, "loss": 0.0156, "step": 199660 }, { "grad_norm": 0.09963161498308182, "learning_rate": 2.7586258472956515e-05, "loss": 0.0144, "step": 199670 }, { "grad_norm": 0.07715888321399689, "learning_rate": 2.758133184346599e-05, "loss": 0.0112, "step": 199680 }, { "grad_norm": 0.2091515064239502, "learning_rate": 2.7576405486383448e-05, "loss": 0.0137, "step": 199690 }, { "grad_norm": 0.21062906086444855, "learning_rate": 2.757147940176875e-05, "loss": 0.009, "step": 199700 }, { "grad_norm": 0.054769065231084824, "learning_rate": 2.756655358968174e-05, "loss": 0.0188, "step": 199710 }, { "grad_norm": 0.04182562977075577, "learning_rate": 2.7561628050182298e-05, "loss": 0.0098, "step": 199720 }, { "grad_norm": 0.1250893920660019, "learning_rate": 2.755670278333026e-05, "loss": 0.0115, "step": 199730 }, { "grad_norm": 0.12458380311727524, "learning_rate": 2.7551777789185472e-05, "loss": 0.0126, "step": 199740 }, { "grad_norm": 0.16842769086360931, "learning_rate": 2.7546853067807786e-05, "loss": 0.0114, "step": 199750 }, { "grad_norm": 0.07462997734546661, "learning_rate": 2.7541928619257007e-05, "loss": 0.0099, "step": 199760 }, { "grad_norm": 0.10438333451747894, "learning_rate": 2.753700444359303e-05, "loss": 0.0139, "step": 199770 }, { "grad_norm": 0.19773486256599426, "learning_rate": 2.753208054087565e-05, "loss": 0.0133, "step": 199780 }, { "grad_norm": 0.0552460141479969, "learning_rate": 2.7527156911164698e-05, "loss": 0.0119, "step": 199790 }, { "grad_norm": 0.18378406763076782, "learning_rate": 2.7522233554519994e-05, "loss": 0.0169, "step": 199800 }, { "grad_norm": 0.07639019191265106, "learning_rate": 2.75173104710014e-05, "loss": 0.0119, "step": 199810 }, { "grad_norm": 0.09262725710868835, "learning_rate": 2.7512387660668703e-05, "loss": 0.0125, "step": 199820 }, { "grad_norm": 0.07553408294916153, "learning_rate": 2.7507465123581732e-05, "loss": 0.0165, "step": 199830 }, { "grad_norm": 0.19822324812412262, "learning_rate": 2.7502542859800273e-05, "loss": 0.012, "step": 199840 }, { "grad_norm": 0.13755272328853607, "learning_rate": 2.749762086938418e-05, "loss": 0.0104, "step": 199850 }, { "grad_norm": 0.07655218243598938, "learning_rate": 2.749269915239324e-05, "loss": 0.0115, "step": 199860 }, { "grad_norm": 0.0825933963060379, "learning_rate": 2.748777770888726e-05, "loss": 0.011, "step": 199870 }, { "grad_norm": 0.09477688372135162, "learning_rate": 2.7482856538926004e-05, "loss": 0.0133, "step": 199880 }, { "grad_norm": 0.09030502289533615, "learning_rate": 2.747793564256933e-05, "loss": 0.0072, "step": 199890 }, { "grad_norm": 0.47895973920822144, "learning_rate": 2.7473015019877e-05, "loss": 0.015, "step": 199900 }, { "grad_norm": 0.0823904275894165, "learning_rate": 2.7468094670908813e-05, "loss": 0.0183, "step": 199910 }, { "grad_norm": 0.44433340430259705, "learning_rate": 2.7463174595724523e-05, "loss": 0.0134, "step": 199920 }, { "grad_norm": 0.05668520927429199, "learning_rate": 2.745825479438398e-05, "loss": 0.0182, "step": 199930 }, { "grad_norm": 0.31136584281921387, "learning_rate": 2.7453335266946888e-05, "loss": 0.0199, "step": 199940 }, { "grad_norm": 0.09806109219789505, "learning_rate": 2.7448416013473076e-05, "loss": 0.024, "step": 199950 }, { "grad_norm": 0.0979510024189949, "learning_rate": 2.7443497034022282e-05, "loss": 0.0139, "step": 199960 }, { "grad_norm": 0.11474896222352982, "learning_rate": 2.7438578328654335e-05, "loss": 0.0145, "step": 199970 }, { "grad_norm": 0.08093073964118958, "learning_rate": 2.7433659897428927e-05, "loss": 0.0103, "step": 199980 }, { "grad_norm": 0.07370444387197495, "learning_rate": 2.742874174040587e-05, "loss": 0.0099, "step": 199990 }, { "grad_norm": 0.13039012253284454, "learning_rate": 2.742382385764492e-05, "loss": 0.0151, "step": 200000 } ], "logging_steps": 10, "max_steps": 300000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 50000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }