{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.16666666666666666, "eval_steps": 500, "global_step": 50000, "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 } ], "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 }