{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 40000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.3426594734191895, "learning_rate": 2.25e-07, "loss": 1.5678, "step": 10 }, { "grad_norm": 1.0611402988433838, "learning_rate": 4.75e-07, "loss": 1.5584, "step": 20 }, { "grad_norm": 1.2228529453277588, "learning_rate": 7.25e-07, "loss": 1.5248, "step": 30 }, { "grad_norm": 1.1133203506469727, "learning_rate": 9.75e-07, "loss": 1.5147, "step": 40 }, { "grad_norm": 1.076000452041626, "learning_rate": 1.2250000000000001e-06, "loss": 1.5213, "step": 50 }, { "grad_norm": 1.6386210918426514, "learning_rate": 1.475e-06, "loss": 1.4499, "step": 60 }, { "grad_norm": 2.75483775138855, "learning_rate": 1.7250000000000002e-06, "loss": 1.3849, "step": 70 }, { "grad_norm": 3.2312285900115967, "learning_rate": 1.975e-06, "loss": 1.3288, "step": 80 }, { "grad_norm": 1.49522066116333, "learning_rate": 2.225e-06, "loss": 1.2965, "step": 90 }, { "grad_norm": 1.8922193050384521, "learning_rate": 2.4750000000000004e-06, "loss": 1.2571, "step": 100 }, { "grad_norm": 2.6854238510131836, "learning_rate": 2.725e-06, "loss": 1.2694, "step": 110 }, { "grad_norm": 1.9871408939361572, "learning_rate": 2.975e-06, "loss": 1.2836, "step": 120 }, { "grad_norm": 2.124488115310669, "learning_rate": 3.225e-06, "loss": 1.2554, "step": 130 }, { "grad_norm": 1.9910212755203247, "learning_rate": 3.4750000000000006e-06, "loss": 1.2024, "step": 140 }, { "grad_norm": 2.146791696548462, "learning_rate": 3.725e-06, "loss": 1.2324, "step": 150 }, { "grad_norm": 11.239241600036621, "learning_rate": 3.975e-06, "loss": 1.2675, "step": 160 }, { "grad_norm": 1.9998944997787476, "learning_rate": 4.225e-06, "loss": 1.1729, "step": 170 }, { "grad_norm": 2.497073173522949, "learning_rate": 4.475e-06, "loss": 1.2038, "step": 180 }, { "grad_norm": 2.8272454738616943, "learning_rate": 4.7250000000000005e-06, "loss": 1.1778, "step": 190 }, { "grad_norm": 2.5783464908599854, "learning_rate": 4.975000000000001e-06, "loss": 1.1909, "step": 200 }, { "grad_norm": 2.1639506816864014, "learning_rate": 5.225e-06, "loss": 1.1562, "step": 210 }, { "grad_norm": 2.215841293334961, "learning_rate": 5.475e-06, "loss": 1.1622, "step": 220 }, { "grad_norm": 3.9817605018615723, "learning_rate": 5.725e-06, "loss": 1.1645, "step": 230 }, { "grad_norm": 2.6633517742156982, "learning_rate": 5.975e-06, "loss": 1.1214, "step": 240 }, { "grad_norm": 2.082655906677246, "learning_rate": 6.2250000000000005e-06, "loss": 1.1642, "step": 250 }, { "grad_norm": 2.3182528018951416, "learning_rate": 6.475000000000001e-06, "loss": 1.144, "step": 260 }, { "grad_norm": 2.0284101963043213, "learning_rate": 6.725000000000001e-06, "loss": 1.1639, "step": 270 }, { "grad_norm": 2.5761773586273193, "learning_rate": 6.975000000000001e-06, "loss": 1.1453, "step": 280 }, { "grad_norm": 2.8645434379577637, "learning_rate": 7.2249999999999994e-06, "loss": 1.11, "step": 290 }, { "grad_norm": 2.805609941482544, "learning_rate": 7.4750000000000004e-06, "loss": 1.1373, "step": 300 }, { "grad_norm": 2.275893449783325, "learning_rate": 7.725e-06, "loss": 1.1426, "step": 310 }, { "grad_norm": 1.8804980516433716, "learning_rate": 7.975e-06, "loss": 1.12, "step": 320 }, { "grad_norm": 2.3131070137023926, "learning_rate": 8.225e-06, "loss": 1.106, "step": 330 }, { "grad_norm": 2.11425518989563, "learning_rate": 8.475000000000001e-06, "loss": 1.1727, "step": 340 }, { "grad_norm": 1.5693159103393555, "learning_rate": 8.725e-06, "loss": 1.1102, "step": 350 }, { "grad_norm": 2.1845128536224365, "learning_rate": 8.975e-06, "loss": 1.1149, "step": 360 }, { "grad_norm": 2.3118624687194824, "learning_rate": 9.225e-06, "loss": 1.1215, "step": 370 }, { "grad_norm": 1.6720075607299805, "learning_rate": 9.475e-06, "loss": 1.0913, "step": 380 }, { "grad_norm": 5.085855007171631, "learning_rate": 9.725000000000001e-06, "loss": 1.1309, "step": 390 }, { "grad_norm": 3.7987804412841797, "learning_rate": 9.975e-06, "loss": 1.122, "step": 400 }, { "grad_norm": 1.4088678359985352, "learning_rate": 1.0225e-05, "loss": 1.091, "step": 410 }, { "grad_norm": 1.817695140838623, "learning_rate": 1.0475e-05, "loss": 1.0917, "step": 420 }, { "grad_norm": 1.4823074340820312, "learning_rate": 1.0725e-05, "loss": 1.1185, "step": 430 }, { "grad_norm": 1.697500467300415, "learning_rate": 1.0975e-05, "loss": 1.1115, "step": 440 }, { "grad_norm": 2.183771848678589, "learning_rate": 1.1225e-05, "loss": 1.1041, "step": 450 }, { "grad_norm": 1.8258695602416992, "learning_rate": 1.1475000000000001e-05, "loss": 1.0869, "step": 460 }, { "grad_norm": 1.6938490867614746, "learning_rate": 1.1725e-05, "loss": 1.108, "step": 470 }, { "grad_norm": 1.6924278736114502, "learning_rate": 1.1975e-05, "loss": 1.1044, "step": 480 }, { "grad_norm": 1.6005676984786987, "learning_rate": 1.2225e-05, "loss": 1.0852, "step": 490 }, { "grad_norm": 1.8619517087936401, "learning_rate": 1.2475e-05, "loss": 1.0821, "step": 500 }, { "grad_norm": 2.342233896255493, "learning_rate": 1.2725000000000001e-05, "loss": 1.1169, "step": 510 }, { "grad_norm": 1.6643568277359009, "learning_rate": 1.2975e-05, "loss": 1.076, "step": 520 }, { "grad_norm": 1.961272120475769, "learning_rate": 1.3225000000000001e-05, "loss": 1.0527, "step": 530 }, { "grad_norm": 1.7056227922439575, "learning_rate": 1.3475000000000002e-05, "loss": 1.0663, "step": 540 }, { "grad_norm": 3.0993480682373047, "learning_rate": 1.3725000000000002e-05, "loss": 1.0804, "step": 550 }, { "grad_norm": 4.411482810974121, "learning_rate": 1.3975000000000003e-05, "loss": 1.0629, "step": 560 }, { "grad_norm": 2.6677801609039307, "learning_rate": 1.4225e-05, "loss": 1.0747, "step": 570 }, { "grad_norm": 2.0741782188415527, "learning_rate": 1.4475e-05, "loss": 1.0621, "step": 580 }, { "grad_norm": 2.46370005607605, "learning_rate": 1.4725e-05, "loss": 1.0609, "step": 590 }, { "grad_norm": 2.396771192550659, "learning_rate": 1.4975e-05, "loss": 1.021, "step": 600 }, { "grad_norm": 2.0176587104797363, "learning_rate": 1.5225e-05, "loss": 1.0082, "step": 610 }, { "grad_norm": 1.9911383390426636, "learning_rate": 1.5475e-05, "loss": 0.957, "step": 620 }, { "grad_norm": 2.496661424636841, "learning_rate": 1.5725e-05, "loss": 0.96, "step": 630 }, { "grad_norm": 3.7425341606140137, "learning_rate": 1.5975000000000002e-05, "loss": 0.8953, "step": 640 }, { "grad_norm": 2.4029791355133057, "learning_rate": 1.6225e-05, "loss": 0.8839, "step": 650 }, { "grad_norm": 4.102582931518555, "learning_rate": 1.6475e-05, "loss": 0.8774, "step": 660 }, { "grad_norm": 2.929990768432617, "learning_rate": 1.6725000000000003e-05, "loss": 0.8355, "step": 670 }, { "grad_norm": 3.1685848236083984, "learning_rate": 1.6975000000000003e-05, "loss": 0.8107, "step": 680 }, { "grad_norm": 2.144110679626465, "learning_rate": 1.7225e-05, "loss": 0.7757, "step": 690 }, { "grad_norm": 3.7501189708709717, "learning_rate": 1.7475e-05, "loss": 0.7782, "step": 700 }, { "grad_norm": 2.9469313621520996, "learning_rate": 1.7725e-05, "loss": 0.7133, "step": 710 }, { "grad_norm": 3.1264121532440186, "learning_rate": 1.7975e-05, "loss": 0.7254, "step": 720 }, { "grad_norm": 2.9018280506134033, "learning_rate": 1.8225e-05, "loss": 0.731, "step": 730 }, { "grad_norm": 2.8065426349639893, "learning_rate": 1.8475000000000002e-05, "loss": 0.7016, "step": 740 }, { "grad_norm": 3.2121236324310303, "learning_rate": 1.8725e-05, "loss": 0.6772, "step": 750 }, { "grad_norm": 2.3925118446350098, "learning_rate": 1.8975e-05, "loss": 0.657, "step": 760 }, { "grad_norm": 2.849717855453491, "learning_rate": 1.9225e-05, "loss": 0.6559, "step": 770 }, { "grad_norm": 2.8578758239746094, "learning_rate": 1.9475000000000002e-05, "loss": 0.6673, "step": 780 }, { "grad_norm": 2.8371853828430176, "learning_rate": 1.9725000000000002e-05, "loss": 0.6107, "step": 790 }, { "grad_norm": 3.3454151153564453, "learning_rate": 1.9975e-05, "loss": 0.6328, "step": 800 }, { "grad_norm": 3.9335694313049316, "learning_rate": 2.0225000000000004e-05, "loss": 0.6124, "step": 810 }, { "grad_norm": 3.1748478412628174, "learning_rate": 2.0475e-05, "loss": 0.5947, "step": 820 }, { "grad_norm": 2.738293170928955, "learning_rate": 2.0725e-05, "loss": 0.5292, "step": 830 }, { "grad_norm": 2.839280605316162, "learning_rate": 2.0975e-05, "loss": 0.5457, "step": 840 }, { "grad_norm": 3.3027169704437256, "learning_rate": 2.1225e-05, "loss": 0.5529, "step": 850 }, { "grad_norm": 3.0866003036499023, "learning_rate": 2.1475e-05, "loss": 0.5022, "step": 860 }, { "grad_norm": 3.217707633972168, "learning_rate": 2.1725e-05, "loss": 0.5137, "step": 870 }, { "grad_norm": 3.898993968963623, "learning_rate": 2.1975000000000002e-05, "loss": 0.488, "step": 880 }, { "grad_norm": 2.7182672023773193, "learning_rate": 2.2225e-05, "loss": 0.4621, "step": 890 }, { "grad_norm": 2.823652982711792, "learning_rate": 2.2475e-05, "loss": 0.4439, "step": 900 }, { "grad_norm": 1.9200247526168823, "learning_rate": 2.2725000000000003e-05, "loss": 0.4335, "step": 910 }, { "grad_norm": 2.961217164993286, "learning_rate": 2.2975000000000003e-05, "loss": 0.4577, "step": 920 }, { "grad_norm": 2.8610479831695557, "learning_rate": 2.3225000000000002e-05, "loss": 0.4505, "step": 930 }, { "grad_norm": 4.352377891540527, "learning_rate": 2.3475e-05, "loss": 0.4168, "step": 940 }, { "grad_norm": 2.5761353969573975, "learning_rate": 2.3725e-05, "loss": 0.428, "step": 950 }, { "grad_norm": 2.7179625034332275, "learning_rate": 2.3975e-05, "loss": 0.4046, "step": 960 }, { "grad_norm": 2.1742095947265625, "learning_rate": 2.4225e-05, "loss": 0.4292, "step": 970 }, { "grad_norm": 2.7539162635803223, "learning_rate": 2.4475000000000002e-05, "loss": 0.3763, "step": 980 }, { "grad_norm": 2.7107298374176025, "learning_rate": 2.4725e-05, "loss": 0.3656, "step": 990 }, { "grad_norm": 4.225486755371094, "learning_rate": 2.4975e-05, "loss": 0.3385, "step": 1000 }, { "grad_norm": 2.8445794582366943, "learning_rate": 2.5225e-05, "loss": 0.3668, "step": 1010 }, { "grad_norm": 2.800901174545288, "learning_rate": 2.5475e-05, "loss": 0.3689, "step": 1020 }, { "grad_norm": 3.1081080436706543, "learning_rate": 2.5725e-05, "loss": 0.3579, "step": 1030 }, { "grad_norm": 2.345898151397705, "learning_rate": 2.5974999999999998e-05, "loss": 0.342, "step": 1040 }, { "grad_norm": 2.6995162963867188, "learning_rate": 2.6225e-05, "loss": 0.3668, "step": 1050 }, { "grad_norm": 3.0032145977020264, "learning_rate": 2.6475e-05, "loss": 0.3235, "step": 1060 }, { "grad_norm": 3.2949724197387695, "learning_rate": 2.6725e-05, "loss": 0.344, "step": 1070 }, { "grad_norm": 2.7689476013183594, "learning_rate": 2.6975000000000002e-05, "loss": 0.2733, "step": 1080 }, { "grad_norm": 4.011098861694336, "learning_rate": 2.7225e-05, "loss": 0.3785, "step": 1090 }, { "grad_norm": 2.6909823417663574, "learning_rate": 2.7475e-05, "loss": 0.3319, "step": 1100 }, { "grad_norm": 3.2926783561706543, "learning_rate": 2.7725e-05, "loss": 0.3429, "step": 1110 }, { "grad_norm": 3.8819541931152344, "learning_rate": 2.7975000000000002e-05, "loss": 0.3402, "step": 1120 }, { "grad_norm": 2.0033819675445557, "learning_rate": 2.8225e-05, "loss": 0.3, "step": 1130 }, { "grad_norm": 2.404205083847046, "learning_rate": 2.8475e-05, "loss": 0.2708, "step": 1140 }, { "grad_norm": 2.1802289485931396, "learning_rate": 2.8725e-05, "loss": 0.2656, "step": 1150 }, { "grad_norm": 2.478653907775879, "learning_rate": 2.8975000000000003e-05, "loss": 0.2827, "step": 1160 }, { "grad_norm": 2.5176358222961426, "learning_rate": 2.9225000000000002e-05, "loss": 0.2706, "step": 1170 }, { "grad_norm": 2.647658348083496, "learning_rate": 2.9475e-05, "loss": 0.2676, "step": 1180 }, { "grad_norm": 2.7475979328155518, "learning_rate": 2.9725000000000004e-05, "loss": 0.263, "step": 1190 }, { "grad_norm": 3.6959683895111084, "learning_rate": 2.9975000000000004e-05, "loss": 0.2589, "step": 1200 }, { "grad_norm": 2.6263461112976074, "learning_rate": 3.0225000000000003e-05, "loss": 0.2395, "step": 1210 }, { "grad_norm": 2.893554925918579, "learning_rate": 3.0475000000000002e-05, "loss": 0.2269, "step": 1220 }, { "grad_norm": 2.5222160816192627, "learning_rate": 3.0725e-05, "loss": 0.2395, "step": 1230 }, { "grad_norm": 2.916205883026123, "learning_rate": 3.0975e-05, "loss": 0.2387, "step": 1240 }, { "grad_norm": 2.6902801990509033, "learning_rate": 3.122500000000001e-05, "loss": 0.2461, "step": 1250 }, { "grad_norm": 2.900451183319092, "learning_rate": 3.1475e-05, "loss": 0.2344, "step": 1260 }, { "grad_norm": 3.3921358585357666, "learning_rate": 3.1725e-05, "loss": 0.2703, "step": 1270 }, { "grad_norm": 2.8743741512298584, "learning_rate": 3.1975e-05, "loss": 0.2433, "step": 1280 }, { "grad_norm": 2.616668939590454, "learning_rate": 3.2225e-05, "loss": 0.2279, "step": 1290 }, { "grad_norm": 2.7857649326324463, "learning_rate": 3.2474999999999997e-05, "loss": 0.2225, "step": 1300 }, { "grad_norm": 2.690023183822632, "learning_rate": 3.2725e-05, "loss": 0.247, "step": 1310 }, { "grad_norm": 2.5297253131866455, "learning_rate": 3.2975e-05, "loss": 0.2441, "step": 1320 }, { "grad_norm": 2.2485928535461426, "learning_rate": 3.3225e-05, "loss": 0.2219, "step": 1330 }, { "grad_norm": 2.300481081008911, "learning_rate": 3.3475e-05, "loss": 0.2203, "step": 1340 }, { "grad_norm": 2.978123903274536, "learning_rate": 3.3725e-05, "loss": 0.1908, "step": 1350 }, { "grad_norm": 3.152855396270752, "learning_rate": 3.3975e-05, "loss": 0.2594, "step": 1360 }, { "grad_norm": 2.30363392829895, "learning_rate": 3.4225e-05, "loss": 0.2144, "step": 1370 }, { "grad_norm": 2.8560259342193604, "learning_rate": 3.4475000000000005e-05, "loss": 0.2322, "step": 1380 }, { "grad_norm": 2.6375668048858643, "learning_rate": 3.4725000000000004e-05, "loss": 0.2179, "step": 1390 }, { "grad_norm": 2.3711161613464355, "learning_rate": 3.4975e-05, "loss": 0.233, "step": 1400 }, { "grad_norm": 2.2280426025390625, "learning_rate": 3.5225e-05, "loss": 0.1838, "step": 1410 }, { "grad_norm": 2.1720826625823975, "learning_rate": 3.5475e-05, "loss": 0.2306, "step": 1420 }, { "grad_norm": 2.6238884925842285, "learning_rate": 3.5725e-05, "loss": 0.2397, "step": 1430 }, { "grad_norm": 2.2724032402038574, "learning_rate": 3.5975e-05, "loss": 0.2372, "step": 1440 }, { "grad_norm": 2.287733793258667, "learning_rate": 3.6225000000000006e-05, "loss": 0.202, "step": 1450 }, { "grad_norm": 2.5431671142578125, "learning_rate": 3.6475000000000006e-05, "loss": 0.2073, "step": 1460 }, { "grad_norm": 1.9215937852859497, "learning_rate": 3.6725000000000005e-05, "loss": 0.1796, "step": 1470 }, { "grad_norm": 2.223116636276245, "learning_rate": 3.6975000000000004e-05, "loss": 0.1815, "step": 1480 }, { "grad_norm": 2.7497174739837646, "learning_rate": 3.7225000000000004e-05, "loss": 0.215, "step": 1490 }, { "grad_norm": 2.190704584121704, "learning_rate": 3.7475e-05, "loss": 0.1889, "step": 1500 }, { "grad_norm": 2.674666166305542, "learning_rate": 3.7725e-05, "loss": 0.1915, "step": 1510 }, { "grad_norm": 2.2396607398986816, "learning_rate": 3.7975e-05, "loss": 0.1862, "step": 1520 }, { "grad_norm": 2.5526199340820312, "learning_rate": 3.8225e-05, "loss": 0.1732, "step": 1530 }, { "grad_norm": 1.764739990234375, "learning_rate": 3.8475e-05, "loss": 0.2518, "step": 1540 }, { "grad_norm": 2.1152284145355225, "learning_rate": 3.8725e-05, "loss": 0.2521, "step": 1550 }, { "grad_norm": 1.9804240465164185, "learning_rate": 3.8975e-05, "loss": 0.2362, "step": 1560 }, { "grad_norm": 2.125981569290161, "learning_rate": 3.9225e-05, "loss": 0.1959, "step": 1570 }, { "grad_norm": 2.0723536014556885, "learning_rate": 3.9475000000000004e-05, "loss": 0.2045, "step": 1580 }, { "grad_norm": 2.2145748138427734, "learning_rate": 3.9725e-05, "loss": 0.1849, "step": 1590 }, { "grad_norm": 1.8184689283370972, "learning_rate": 3.9975e-05, "loss": 0.1887, "step": 1600 }, { "grad_norm": 1.9053422212600708, "learning_rate": 4.0225e-05, "loss": 0.1945, "step": 1610 }, { "grad_norm": 1.8294774293899536, "learning_rate": 4.0475e-05, "loss": 0.1907, "step": 1620 }, { "grad_norm": 1.89945387840271, "learning_rate": 4.0725e-05, "loss": 0.2007, "step": 1630 }, { "grad_norm": 2.972592830657959, "learning_rate": 4.0975e-05, "loss": 0.2165, "step": 1640 }, { "grad_norm": 1.9451191425323486, "learning_rate": 4.1225e-05, "loss": 0.229, "step": 1650 }, { "grad_norm": 2.1198599338531494, "learning_rate": 4.1475000000000005e-05, "loss": 0.196, "step": 1660 }, { "grad_norm": 2.0144219398498535, "learning_rate": 4.1725000000000005e-05, "loss": 0.1704, "step": 1670 }, { "grad_norm": 1.5619484186172485, "learning_rate": 4.1975000000000004e-05, "loss": 0.2071, "step": 1680 }, { "grad_norm": 1.9323033094406128, "learning_rate": 4.2225e-05, "loss": 0.2041, "step": 1690 }, { "grad_norm": 1.9521551132202148, "learning_rate": 4.2475e-05, "loss": 0.2155, "step": 1700 }, { "grad_norm": 1.6346988677978516, "learning_rate": 4.2725e-05, "loss": 0.1655, "step": 1710 }, { "grad_norm": 2.333902597427368, "learning_rate": 4.2975e-05, "loss": 0.1947, "step": 1720 }, { "grad_norm": 2.301504373550415, "learning_rate": 4.322500000000001e-05, "loss": 0.2516, "step": 1730 }, { "grad_norm": 2.146275758743286, "learning_rate": 4.3475000000000006e-05, "loss": 0.2362, "step": 1740 }, { "grad_norm": 1.6660637855529785, "learning_rate": 4.3725000000000006e-05, "loss": 0.2102, "step": 1750 }, { "grad_norm": 1.4695221185684204, "learning_rate": 4.3975e-05, "loss": 0.195, "step": 1760 }, { "grad_norm": 2.54642915725708, "learning_rate": 4.4225e-05, "loss": 0.2115, "step": 1770 }, { "grad_norm": 1.9110864400863647, "learning_rate": 4.4475e-05, "loss": 0.1891, "step": 1780 }, { "grad_norm": 1.5354009866714478, "learning_rate": 4.4725e-05, "loss": 0.2032, "step": 1790 }, { "grad_norm": 1.9472720623016357, "learning_rate": 4.4975e-05, "loss": 0.2138, "step": 1800 }, { "grad_norm": 1.8614827394485474, "learning_rate": 4.5225e-05, "loss": 0.1944, "step": 1810 }, { "grad_norm": 2.519599199295044, "learning_rate": 4.5475e-05, "loss": 0.1885, "step": 1820 }, { "grad_norm": 2.547074794769287, "learning_rate": 4.5725e-05, "loss": 0.1827, "step": 1830 }, { "grad_norm": 1.5711215734481812, "learning_rate": 4.5975e-05, "loss": 0.183, "step": 1840 }, { "grad_norm": 1.9724048376083374, "learning_rate": 4.6225e-05, "loss": 0.185, "step": 1850 }, { "grad_norm": 1.3678338527679443, "learning_rate": 4.6475000000000005e-05, "loss": 0.2095, "step": 1860 }, { "grad_norm": 1.5370368957519531, "learning_rate": 4.6725000000000004e-05, "loss": 0.2014, "step": 1870 }, { "grad_norm": 1.9217031002044678, "learning_rate": 4.6975000000000003e-05, "loss": 0.187, "step": 1880 }, { "grad_norm": 1.3873637914657593, "learning_rate": 4.7225e-05, "loss": 0.1788, "step": 1890 }, { "grad_norm": 1.4921139478683472, "learning_rate": 4.7475e-05, "loss": 0.1867, "step": 1900 }, { "grad_norm": 1.4734078645706177, "learning_rate": 4.7725e-05, "loss": 0.1913, "step": 1910 }, { "grad_norm": 1.4935719966888428, "learning_rate": 4.7975e-05, "loss": 0.1693, "step": 1920 }, { "grad_norm": 1.4366475343704224, "learning_rate": 4.822500000000001e-05, "loss": 0.2372, "step": 1930 }, { "grad_norm": 1.8390486240386963, "learning_rate": 4.8475000000000006e-05, "loss": 0.1846, "step": 1940 }, { "grad_norm": 1.3911668062210083, "learning_rate": 4.8725000000000005e-05, "loss": 0.1593, "step": 1950 }, { "grad_norm": 1.450315237045288, "learning_rate": 4.8975000000000005e-05, "loss": 0.1937, "step": 1960 }, { "grad_norm": 1.8350298404693604, "learning_rate": 4.9225000000000004e-05, "loss": 0.1856, "step": 1970 }, { "grad_norm": 1.7923293113708496, "learning_rate": 4.9475e-05, "loss": 0.1732, "step": 1980 }, { "grad_norm": 1.8788368701934814, "learning_rate": 4.9725e-05, "loss": 0.186, "step": 1990 }, { "grad_norm": 1.7406036853790283, "learning_rate": 4.9975e-05, "loss": 0.173, "step": 2000 }, { "grad_norm": 1.6812522411346436, "learning_rate": 5.0225e-05, "loss": 0.1801, "step": 2010 }, { "grad_norm": 1.6396642923355103, "learning_rate": 5.047500000000001e-05, "loss": 0.1914, "step": 2020 }, { "grad_norm": 1.7596477270126343, "learning_rate": 5.0725e-05, "loss": 0.2102, "step": 2030 }, { "grad_norm": 2.0351991653442383, "learning_rate": 5.0975000000000006e-05, "loss": 0.1894, "step": 2040 }, { "grad_norm": 1.615128755569458, "learning_rate": 5.1225e-05, "loss": 0.1615, "step": 2050 }, { "grad_norm": 1.4212408065795898, "learning_rate": 5.1475000000000004e-05, "loss": 0.1839, "step": 2060 }, { "grad_norm": 1.6914554834365845, "learning_rate": 5.1725000000000004e-05, "loss": 0.1931, "step": 2070 }, { "grad_norm": 1.3586798906326294, "learning_rate": 5.197500000000001e-05, "loss": 0.1934, "step": 2080 }, { "grad_norm": 1.7830848693847656, "learning_rate": 5.2225e-05, "loss": 0.1897, "step": 2090 }, { "grad_norm": 1.544135332107544, "learning_rate": 5.247500000000001e-05, "loss": 0.1935, "step": 2100 }, { "grad_norm": 1.5708987712860107, "learning_rate": 5.2725e-05, "loss": 0.2091, "step": 2110 }, { "grad_norm": 1.694012999534607, "learning_rate": 5.297500000000001e-05, "loss": 0.2313, "step": 2120 }, { "grad_norm": 2.0928499698638916, "learning_rate": 5.3225e-05, "loss": 0.2263, "step": 2130 }, { "grad_norm": 1.7098162174224854, "learning_rate": 5.3475e-05, "loss": 0.207, "step": 2140 }, { "grad_norm": 1.8285406827926636, "learning_rate": 5.3725000000000005e-05, "loss": 0.2134, "step": 2150 }, { "grad_norm": 1.4276952743530273, "learning_rate": 5.3975e-05, "loss": 0.2203, "step": 2160 }, { "grad_norm": 1.877434492111206, "learning_rate": 5.4225000000000003e-05, "loss": 0.1755, "step": 2170 }, { "grad_norm": 1.8231861591339111, "learning_rate": 5.4474999999999996e-05, "loss": 0.1955, "step": 2180 }, { "grad_norm": 1.7123863697052002, "learning_rate": 5.4725e-05, "loss": 0.153, "step": 2190 }, { "grad_norm": 1.3425322771072388, "learning_rate": 5.4975e-05, "loss": 0.1654, "step": 2200 }, { "grad_norm": 2.1319870948791504, "learning_rate": 5.522500000000001e-05, "loss": 0.1922, "step": 2210 }, { "grad_norm": 1.8202077150344849, "learning_rate": 5.5475e-05, "loss": 0.2418, "step": 2220 }, { "grad_norm": 1.4796597957611084, "learning_rate": 5.5725000000000006e-05, "loss": 0.1702, "step": 2230 }, { "grad_norm": 1.4488469362258911, "learning_rate": 5.5975e-05, "loss": 0.1546, "step": 2240 }, { "grad_norm": 1.4228147268295288, "learning_rate": 5.6225000000000005e-05, "loss": 0.1776, "step": 2250 }, { "grad_norm": 1.4210416078567505, "learning_rate": 5.6475e-05, "loss": 0.1968, "step": 2260 }, { "grad_norm": 1.318648099899292, "learning_rate": 5.6725e-05, "loss": 0.2035, "step": 2270 }, { "grad_norm": 1.6862881183624268, "learning_rate": 5.6975e-05, "loss": 0.2053, "step": 2280 }, { "grad_norm": 1.6675561666488647, "learning_rate": 5.722500000000001e-05, "loss": 0.2092, "step": 2290 }, { "grad_norm": 1.602689266204834, "learning_rate": 5.7475e-05, "loss": 0.2266, "step": 2300 }, { "grad_norm": 1.790428876876831, "learning_rate": 5.772500000000001e-05, "loss": 0.1772, "step": 2310 }, { "grad_norm": 1.268574595451355, "learning_rate": 5.7975e-05, "loss": 0.1917, "step": 2320 }, { "grad_norm": 1.728223204612732, "learning_rate": 5.8225000000000006e-05, "loss": 0.1945, "step": 2330 }, { "grad_norm": 1.664332389831543, "learning_rate": 5.8475000000000005e-05, "loss": 0.2064, "step": 2340 }, { "grad_norm": 1.6534582376480103, "learning_rate": 5.8725000000000004e-05, "loss": 0.1722, "step": 2350 }, { "grad_norm": 2.1087543964385986, "learning_rate": 5.8975000000000004e-05, "loss": 0.1731, "step": 2360 }, { "grad_norm": 2.089299201965332, "learning_rate": 5.922500000000001e-05, "loss": 0.2037, "step": 2370 }, { "grad_norm": 1.302740216255188, "learning_rate": 5.9475e-05, "loss": 0.1701, "step": 2380 }, { "grad_norm": 1.089152455329895, "learning_rate": 5.9724999999999995e-05, "loss": 0.1642, "step": 2390 }, { "grad_norm": 1.690234661102295, "learning_rate": 5.9975e-05, "loss": 0.2108, "step": 2400 }, { "grad_norm": 1.3293975591659546, "learning_rate": 6.0225e-05, "loss": 0.2497, "step": 2410 }, { "grad_norm": 1.7480342388153076, "learning_rate": 6.0475000000000006e-05, "loss": 0.1988, "step": 2420 }, { "grad_norm": 1.7016792297363281, "learning_rate": 6.0725e-05, "loss": 0.1841, "step": 2430 }, { "grad_norm": 1.629548192024231, "learning_rate": 6.0975000000000005e-05, "loss": 0.218, "step": 2440 }, { "grad_norm": 1.3511013984680176, "learning_rate": 6.1225e-05, "loss": 0.2341, "step": 2450 }, { "grad_norm": 1.2721856832504272, "learning_rate": 6.1475e-05, "loss": 0.1627, "step": 2460 }, { "grad_norm": 1.6601649522781372, "learning_rate": 6.1725e-05, "loss": 0.2285, "step": 2470 }, { "grad_norm": 1.4752312898635864, "learning_rate": 6.1975e-05, "loss": 0.1591, "step": 2480 }, { "grad_norm": 1.4101492166519165, "learning_rate": 6.2225e-05, "loss": 0.1597, "step": 2490 }, { "grad_norm": 1.4799888134002686, "learning_rate": 6.2475e-05, "loss": 0.1797, "step": 2500 }, { "grad_norm": 1.1533913612365723, "learning_rate": 6.2725e-05, "loss": 0.169, "step": 2510 }, { "grad_norm": 1.169195294380188, "learning_rate": 6.297500000000001e-05, "loss": 0.1789, "step": 2520 }, { "grad_norm": 1.450256109237671, "learning_rate": 6.3225e-05, "loss": 0.1622, "step": 2530 }, { "grad_norm": 1.5524494647979736, "learning_rate": 6.347500000000001e-05, "loss": 0.1561, "step": 2540 }, { "grad_norm": 1.3230273723602295, "learning_rate": 6.3725e-05, "loss": 0.1855, "step": 2550 }, { "grad_norm": 1.2932367324829102, "learning_rate": 6.397500000000001e-05, "loss": 0.1756, "step": 2560 }, { "grad_norm": 1.2019833326339722, "learning_rate": 6.4225e-05, "loss": 0.2079, "step": 2570 }, { "grad_norm": 1.409037470817566, "learning_rate": 6.447500000000001e-05, "loss": 0.158, "step": 2580 }, { "grad_norm": 1.5124940872192383, "learning_rate": 6.4725e-05, "loss": 0.1959, "step": 2590 }, { "grad_norm": 1.128197193145752, "learning_rate": 6.497500000000001e-05, "loss": 0.1535, "step": 2600 }, { "grad_norm": 1.2888292074203491, "learning_rate": 6.5225e-05, "loss": 0.1719, "step": 2610 }, { "grad_norm": 1.4784741401672363, "learning_rate": 6.5475e-05, "loss": 0.1598, "step": 2620 }, { "grad_norm": 1.1589480638504028, "learning_rate": 6.5725e-05, "loss": 0.1707, "step": 2630 }, { "grad_norm": 1.175367832183838, "learning_rate": 6.5975e-05, "loss": 0.1944, "step": 2640 }, { "grad_norm": 1.2981537580490112, "learning_rate": 6.6225e-05, "loss": 0.1578, "step": 2650 }, { "grad_norm": 1.5885035991668701, "learning_rate": 6.6475e-05, "loss": 0.1708, "step": 2660 }, { "grad_norm": 1.5190941095352173, "learning_rate": 6.672500000000001e-05, "loss": 0.1538, "step": 2670 }, { "grad_norm": 0.9960178732872009, "learning_rate": 6.6975e-05, "loss": 0.1856, "step": 2680 }, { "grad_norm": 1.1171075105667114, "learning_rate": 6.722500000000001e-05, "loss": 0.1798, "step": 2690 }, { "grad_norm": 1.186639666557312, "learning_rate": 6.7475e-05, "loss": 0.1868, "step": 2700 }, { "grad_norm": 1.541678547859192, "learning_rate": 6.7725e-05, "loss": 0.1838, "step": 2710 }, { "grad_norm": 1.4054371118545532, "learning_rate": 6.7975e-05, "loss": 0.1743, "step": 2720 }, { "grad_norm": 1.5821988582611084, "learning_rate": 6.8225e-05, "loss": 0.1698, "step": 2730 }, { "grad_norm": 1.5916699171066284, "learning_rate": 6.8475e-05, "loss": 0.1836, "step": 2740 }, { "grad_norm": 1.1810917854309082, "learning_rate": 6.8725e-05, "loss": 0.1286, "step": 2750 }, { "grad_norm": 1.2796006202697754, "learning_rate": 6.8975e-05, "loss": 0.1668, "step": 2760 }, { "grad_norm": 1.2389726638793945, "learning_rate": 6.9225e-05, "loss": 0.1747, "step": 2770 }, { "grad_norm": 1.246410846710205, "learning_rate": 6.9475e-05, "loss": 0.1729, "step": 2780 }, { "grad_norm": 1.1990686655044556, "learning_rate": 6.9725e-05, "loss": 0.154, "step": 2790 }, { "grad_norm": 0.9104430675506592, "learning_rate": 6.997500000000001e-05, "loss": 0.17, "step": 2800 }, { "grad_norm": 1.0445033311843872, "learning_rate": 7.022500000000001e-05, "loss": 0.1665, "step": 2810 }, { "grad_norm": 1.153649926185608, "learning_rate": 7.0475e-05, "loss": 0.1731, "step": 2820 }, { "grad_norm": 1.172935128211975, "learning_rate": 7.072500000000001e-05, "loss": 0.1642, "step": 2830 }, { "grad_norm": 1.7028405666351318, "learning_rate": 7.0975e-05, "loss": 0.2107, "step": 2840 }, { "grad_norm": 1.4143717288970947, "learning_rate": 7.122500000000001e-05, "loss": 0.1324, "step": 2850 }, { "grad_norm": 1.0276389122009277, "learning_rate": 7.1475e-05, "loss": 0.1611, "step": 2860 }, { "grad_norm": 1.6902616024017334, "learning_rate": 7.172500000000001e-05, "loss": 0.1723, "step": 2870 }, { "grad_norm": 1.3310682773590088, "learning_rate": 7.1975e-05, "loss": 0.1658, "step": 2880 }, { "grad_norm": 1.3733688592910767, "learning_rate": 7.2225e-05, "loss": 0.2069, "step": 2890 }, { "grad_norm": 1.3055189847946167, "learning_rate": 7.2475e-05, "loss": 0.1864, "step": 2900 }, { "grad_norm": 1.1385434865951538, "learning_rate": 7.272499999999999e-05, "loss": 0.1968, "step": 2910 }, { "grad_norm": 1.164703130722046, "learning_rate": 7.2975e-05, "loss": 0.1338, "step": 2920 }, { "grad_norm": 1.2862591743469238, "learning_rate": 7.3225e-05, "loss": 0.18, "step": 2930 }, { "grad_norm": 0.8512696027755737, "learning_rate": 7.347500000000001e-05, "loss": 0.1661, "step": 2940 }, { "grad_norm": 1.2303428649902344, "learning_rate": 7.3725e-05, "loss": 0.1614, "step": 2950 }, { "grad_norm": 1.5450336933135986, "learning_rate": 7.397500000000001e-05, "loss": 0.1569, "step": 2960 }, { "grad_norm": 1.179665446281433, "learning_rate": 7.4225e-05, "loss": 0.1471, "step": 2970 }, { "grad_norm": 1.0603480339050293, "learning_rate": 7.447500000000001e-05, "loss": 0.1748, "step": 2980 }, { "grad_norm": 0.8578670024871826, "learning_rate": 7.4725e-05, "loss": 0.1724, "step": 2990 }, { "grad_norm": 1.1904118061065674, "learning_rate": 7.4975e-05, "loss": 0.1598, "step": 3000 }, { "grad_norm": 0.9909509420394897, "learning_rate": 7.5225e-05, "loss": 0.1609, "step": 3010 }, { "grad_norm": 1.1326698064804077, "learning_rate": 7.5475e-05, "loss": 0.1452, "step": 3020 }, { "grad_norm": 1.0651330947875977, "learning_rate": 7.5725e-05, "loss": 0.1512, "step": 3030 }, { "grad_norm": 1.3733209371566772, "learning_rate": 7.5975e-05, "loss": 0.1779, "step": 3040 }, { "grad_norm": 1.169141411781311, "learning_rate": 7.6225e-05, "loss": 0.1311, "step": 3050 }, { "grad_norm": 1.1044014692306519, "learning_rate": 7.6475e-05, "loss": 0.1956, "step": 3060 }, { "grad_norm": 0.9914211630821228, "learning_rate": 7.672500000000001e-05, "loss": 0.1869, "step": 3070 }, { "grad_norm": 1.1900849342346191, "learning_rate": 7.697500000000001e-05, "loss": 0.1491, "step": 3080 }, { "grad_norm": 1.1906745433807373, "learning_rate": 7.722500000000001e-05, "loss": 0.1471, "step": 3090 }, { "grad_norm": 1.300580382347107, "learning_rate": 7.747500000000001e-05, "loss": 0.1614, "step": 3100 }, { "grad_norm": 1.3671272993087769, "learning_rate": 7.7725e-05, "loss": 0.1998, "step": 3110 }, { "grad_norm": 1.1765789985656738, "learning_rate": 7.797500000000001e-05, "loss": 0.1976, "step": 3120 }, { "grad_norm": 1.7475860118865967, "learning_rate": 7.8225e-05, "loss": 0.1831, "step": 3130 }, { "grad_norm": 1.2040830850601196, "learning_rate": 7.8475e-05, "loss": 0.1738, "step": 3140 }, { "grad_norm": 1.082480549812317, "learning_rate": 7.8725e-05, "loss": 0.1799, "step": 3150 }, { "grad_norm": 1.4855351448059082, "learning_rate": 7.8975e-05, "loss": 0.1591, "step": 3160 }, { "grad_norm": 1.2044944763183594, "learning_rate": 7.9225e-05, "loss": 0.1713, "step": 3170 }, { "grad_norm": 1.2625677585601807, "learning_rate": 7.9475e-05, "loss": 0.1879, "step": 3180 }, { "grad_norm": 1.1849489212036133, "learning_rate": 7.9725e-05, "loss": 0.1711, "step": 3190 }, { "grad_norm": 1.109319806098938, "learning_rate": 7.9975e-05, "loss": 0.169, "step": 3200 }, { "grad_norm": 1.1486090421676636, "learning_rate": 8.022500000000001e-05, "loss": 0.1353, "step": 3210 }, { "grad_norm": 0.8770895600318909, "learning_rate": 8.0475e-05, "loss": 0.1461, "step": 3220 }, { "grad_norm": 1.536001443862915, "learning_rate": 8.072500000000001e-05, "loss": 0.1533, "step": 3230 }, { "grad_norm": 1.2847559452056885, "learning_rate": 8.0975e-05, "loss": 0.1532, "step": 3240 }, { "grad_norm": 1.0113424062728882, "learning_rate": 8.122500000000001e-05, "loss": 0.1878, "step": 3250 }, { "grad_norm": 1.1361004114151, "learning_rate": 8.1475e-05, "loss": 0.22, "step": 3260 }, { "grad_norm": 1.3338221311569214, "learning_rate": 8.172500000000001e-05, "loss": 0.1814, "step": 3270 }, { "grad_norm": 1.145363450050354, "learning_rate": 8.1975e-05, "loss": 0.1719, "step": 3280 }, { "grad_norm": 1.0885893106460571, "learning_rate": 8.2225e-05, "loss": 0.1796, "step": 3290 }, { "grad_norm": 1.267020344734192, "learning_rate": 8.2475e-05, "loss": 0.1719, "step": 3300 }, { "grad_norm": 1.206691861152649, "learning_rate": 8.2725e-05, "loss": 0.1681, "step": 3310 }, { "grad_norm": 1.4693042039871216, "learning_rate": 8.2975e-05, "loss": 0.1775, "step": 3320 }, { "grad_norm": 1.2660140991210938, "learning_rate": 8.3225e-05, "loss": 0.1512, "step": 3330 }, { "grad_norm": 1.0289838314056396, "learning_rate": 8.347500000000001e-05, "loss": 0.1442, "step": 3340 }, { "grad_norm": 0.9609429240226746, "learning_rate": 8.3725e-05, "loss": 0.1448, "step": 3350 }, { "grad_norm": 1.2478498220443726, "learning_rate": 8.397500000000001e-05, "loss": 0.1613, "step": 3360 }, { "grad_norm": 1.183092474937439, "learning_rate": 8.422500000000001e-05, "loss": 0.1565, "step": 3370 }, { "grad_norm": 1.1614521741867065, "learning_rate": 8.447500000000001e-05, "loss": 0.159, "step": 3380 }, { "grad_norm": 0.9941691756248474, "learning_rate": 8.4725e-05, "loss": 0.1674, "step": 3390 }, { "grad_norm": 1.0018359422683716, "learning_rate": 8.4975e-05, "loss": 0.1414, "step": 3400 }, { "grad_norm": 1.6718679666519165, "learning_rate": 8.5225e-05, "loss": 0.1619, "step": 3410 }, { "grad_norm": 1.1393837928771973, "learning_rate": 8.5475e-05, "loss": 0.1933, "step": 3420 }, { "grad_norm": 0.8575757145881653, "learning_rate": 8.5725e-05, "loss": 0.1504, "step": 3430 }, { "grad_norm": 1.175531268119812, "learning_rate": 8.5975e-05, "loss": 0.1797, "step": 3440 }, { "grad_norm": 0.9798291325569153, "learning_rate": 8.6225e-05, "loss": 0.1633, "step": 3450 }, { "grad_norm": 1.1075758934020996, "learning_rate": 8.6475e-05, "loss": 0.1517, "step": 3460 }, { "grad_norm": 1.4281535148620605, "learning_rate": 8.672500000000001e-05, "loss": 0.1963, "step": 3470 }, { "grad_norm": 0.947013795375824, "learning_rate": 8.6975e-05, "loss": 0.1399, "step": 3480 }, { "grad_norm": 1.5156246423721313, "learning_rate": 8.7225e-05, "loss": 0.1584, "step": 3490 }, { "grad_norm": 1.349194884300232, "learning_rate": 8.747500000000001e-05, "loss": 0.1726, "step": 3500 }, { "grad_norm": 1.557656168937683, "learning_rate": 8.7725e-05, "loss": 0.1669, "step": 3510 }, { "grad_norm": 1.353997826576233, "learning_rate": 8.797500000000001e-05, "loss": 0.167, "step": 3520 }, { "grad_norm": 1.1703020334243774, "learning_rate": 8.8225e-05, "loss": 0.1408, "step": 3530 }, { "grad_norm": 1.257754921913147, "learning_rate": 8.847500000000001e-05, "loss": 0.1717, "step": 3540 }, { "grad_norm": 1.4615027904510498, "learning_rate": 8.8725e-05, "loss": 0.1658, "step": 3550 }, { "grad_norm": 1.168090581893921, "learning_rate": 8.897500000000001e-05, "loss": 0.1792, "step": 3560 }, { "grad_norm": 1.0491012334823608, "learning_rate": 8.9225e-05, "loss": 0.1332, "step": 3570 }, { "grad_norm": 1.0409868955612183, "learning_rate": 8.9475e-05, "loss": 0.1668, "step": 3580 }, { "grad_norm": 1.000691533088684, "learning_rate": 8.9725e-05, "loss": 0.1655, "step": 3590 }, { "grad_norm": 0.802480936050415, "learning_rate": 8.9975e-05, "loss": 0.1491, "step": 3600 }, { "grad_norm": 1.5940778255462646, "learning_rate": 9.0225e-05, "loss": 0.1638, "step": 3610 }, { "grad_norm": 1.001261830329895, "learning_rate": 9.0475e-05, "loss": 0.1267, "step": 3620 }, { "grad_norm": 1.169386625289917, "learning_rate": 9.072500000000001e-05, "loss": 0.1353, "step": 3630 }, { "grad_norm": 1.7801941633224487, "learning_rate": 9.0975e-05, "loss": 0.1842, "step": 3640 }, { "grad_norm": 1.1861428022384644, "learning_rate": 9.122500000000001e-05, "loss": 0.1426, "step": 3650 }, { "grad_norm": 1.2493151426315308, "learning_rate": 9.1475e-05, "loss": 0.1774, "step": 3660 }, { "grad_norm": 1.0931211709976196, "learning_rate": 9.172500000000001e-05, "loss": 0.1899, "step": 3670 }, { "grad_norm": 0.9624337553977966, "learning_rate": 9.1975e-05, "loss": 0.1481, "step": 3680 }, { "grad_norm": 1.1499532461166382, "learning_rate": 9.2225e-05, "loss": 0.2161, "step": 3690 }, { "grad_norm": 0.8584557771682739, "learning_rate": 9.2475e-05, "loss": 0.1835, "step": 3700 }, { "grad_norm": 1.413686752319336, "learning_rate": 9.2725e-05, "loss": 0.1625, "step": 3710 }, { "grad_norm": 1.176853895187378, "learning_rate": 9.2975e-05, "loss": 0.2325, "step": 3720 }, { "grad_norm": 0.9389675259590149, "learning_rate": 9.3225e-05, "loss": 0.1418, "step": 3730 }, { "grad_norm": 1.1526861190795898, "learning_rate": 9.3475e-05, "loss": 0.1646, "step": 3740 }, { "grad_norm": 1.1761467456817627, "learning_rate": 9.3725e-05, "loss": 0.149, "step": 3750 }, { "grad_norm": 1.5374647378921509, "learning_rate": 9.397500000000001e-05, "loss": 0.1495, "step": 3760 }, { "grad_norm": 0.9686386585235596, "learning_rate": 9.422500000000001e-05, "loss": 0.1821, "step": 3770 }, { "grad_norm": 1.5518065690994263, "learning_rate": 9.4475e-05, "loss": 0.1398, "step": 3780 }, { "grad_norm": 1.1013659238815308, "learning_rate": 9.472500000000001e-05, "loss": 0.1648, "step": 3790 }, { "grad_norm": 0.8275545239448547, "learning_rate": 9.4975e-05, "loss": 0.1518, "step": 3800 }, { "grad_norm": 0.8118910789489746, "learning_rate": 9.522500000000001e-05, "loss": 0.1637, "step": 3810 }, { "grad_norm": 1.011211633682251, "learning_rate": 9.5475e-05, "loss": 0.1893, "step": 3820 }, { "grad_norm": 1.2738045454025269, "learning_rate": 9.572500000000001e-05, "loss": 0.1637, "step": 3830 }, { "grad_norm": 1.148013949394226, "learning_rate": 9.5975e-05, "loss": 0.1554, "step": 3840 }, { "grad_norm": 0.8054123520851135, "learning_rate": 9.622500000000001e-05, "loss": 0.1559, "step": 3850 }, { "grad_norm": 1.0602627992630005, "learning_rate": 9.6475e-05, "loss": 0.1451, "step": 3860 }, { "grad_norm": 1.080871343612671, "learning_rate": 9.6725e-05, "loss": 0.1529, "step": 3870 }, { "grad_norm": 1.11483633518219, "learning_rate": 9.6975e-05, "loss": 0.1487, "step": 3880 }, { "grad_norm": 2.460501194000244, "learning_rate": 9.7225e-05, "loss": 0.1765, "step": 3890 }, { "grad_norm": 1.0430244207382202, "learning_rate": 9.747500000000001e-05, "loss": 0.1419, "step": 3900 }, { "grad_norm": 0.8425542116165161, "learning_rate": 9.7725e-05, "loss": 0.173, "step": 3910 }, { "grad_norm": 0.9341223239898682, "learning_rate": 9.797500000000001e-05, "loss": 0.1737, "step": 3920 }, { "grad_norm": 1.1482388973236084, "learning_rate": 9.8225e-05, "loss": 0.1332, "step": 3930 }, { "grad_norm": 1.3832213878631592, "learning_rate": 9.847500000000001e-05, "loss": 0.1394, "step": 3940 }, { "grad_norm": 1.0939297676086426, "learning_rate": 9.8725e-05, "loss": 0.154, "step": 3950 }, { "grad_norm": 0.912337064743042, "learning_rate": 9.897500000000001e-05, "loss": 0.2031, "step": 3960 }, { "grad_norm": 1.1850663423538208, "learning_rate": 9.9225e-05, "loss": 0.1511, "step": 3970 }, { "grad_norm": 0.9213148355484009, "learning_rate": 9.9475e-05, "loss": 0.1269, "step": 3980 }, { "grad_norm": 1.0390311479568481, "learning_rate": 9.9725e-05, "loss": 0.1799, "step": 3990 }, { "grad_norm": 1.0389889478683472, "learning_rate": 9.9975e-05, "loss": 0.132, "step": 4000 }, { "grad_norm": 1.0111136436462402, "learning_rate": 9.999999653982884e-05, "loss": 0.1903, "step": 4010 }, { "grad_norm": 0.9220157265663147, "learning_rate": 9.999998457874392e-05, "loss": 0.178, "step": 4020 }, { "grad_norm": 1.10685133934021, "learning_rate": 9.999996407402913e-05, "loss": 0.1384, "step": 4030 }, { "grad_norm": 0.8706514835357666, "learning_rate": 9.999993502568801e-05, "loss": 0.1698, "step": 4040 }, { "grad_norm": 0.9460839629173279, "learning_rate": 9.999989743372548e-05, "loss": 0.1411, "step": 4050 }, { "grad_norm": 1.0447988510131836, "learning_rate": 9.999985129814798e-05, "loss": 0.1464, "step": 4060 }, { "grad_norm": 0.8997766971588135, "learning_rate": 9.99997966189634e-05, "loss": 0.1404, "step": 4070 }, { "grad_norm": 1.3907016515731812, "learning_rate": 9.999973339618107e-05, "loss": 0.1775, "step": 4080 }, { "grad_norm": 1.2763676643371582, "learning_rate": 9.999966162981179e-05, "loss": 0.1519, "step": 4090 }, { "grad_norm": 1.052891731262207, "learning_rate": 9.999958131986784e-05, "loss": 0.1666, "step": 4100 }, { "grad_norm": 1.0897619724273682, "learning_rate": 9.999949246636293e-05, "loss": 0.1337, "step": 4110 }, { "grad_norm": 0.8948440551757812, "learning_rate": 9.999939506931224e-05, "loss": 0.1424, "step": 4120 }, { "grad_norm": 1.1128064393997192, "learning_rate": 9.999928912873243e-05, "loss": 0.1419, "step": 4130 }, { "grad_norm": 0.7920106053352356, "learning_rate": 9.999917464464159e-05, "loss": 0.1285, "step": 4140 }, { "grad_norm": 0.8904651403427124, "learning_rate": 9.999905161705929e-05, "loss": 0.1399, "step": 4150 }, { "grad_norm": 1.2188864946365356, "learning_rate": 9.999892004600653e-05, "loss": 0.1568, "step": 4160 }, { "grad_norm": 0.9740485548973083, "learning_rate": 9.999877993150581e-05, "loss": 0.1332, "step": 4170 }, { "grad_norm": 0.8258699774742126, "learning_rate": 9.999863127358108e-05, "loss": 0.1676, "step": 4180 }, { "grad_norm": 0.9945643544197083, "learning_rate": 9.999847407225773e-05, "loss": 0.1485, "step": 4190 }, { "grad_norm": 0.9376386404037476, "learning_rate": 9.999830832756262e-05, "loss": 0.1683, "step": 4200 }, { "grad_norm": 1.0062787532806396, "learning_rate": 9.999813403952407e-05, "loss": 0.1677, "step": 4210 }, { "grad_norm": 0.8857281804084778, "learning_rate": 9.999795120817187e-05, "loss": 0.1692, "step": 4220 }, { "grad_norm": 1.3981996774673462, "learning_rate": 9.999775983353725e-05, "loss": 0.1464, "step": 4230 }, { "grad_norm": 1.0998586416244507, "learning_rate": 9.999755991565292e-05, "loss": 0.1611, "step": 4240 }, { "grad_norm": 0.9474404454231262, "learning_rate": 9.999735145455303e-05, "loss": 0.1626, "step": 4250 }, { "grad_norm": 0.8695165514945984, "learning_rate": 9.99971344502732e-05, "loss": 0.1335, "step": 4260 }, { "grad_norm": 0.9527442455291748, "learning_rate": 9.999690890285053e-05, "loss": 0.1433, "step": 4270 }, { "grad_norm": 1.0389082431793213, "learning_rate": 9.999667481232356e-05, "loss": 0.1286, "step": 4280 }, { "grad_norm": 1.1667630672454834, "learning_rate": 9.999643217873225e-05, "loss": 0.1921, "step": 4290 }, { "grad_norm": 1.1275607347488403, "learning_rate": 9.999618100211809e-05, "loss": 0.1543, "step": 4300 }, { "grad_norm": 0.947342574596405, "learning_rate": 9.999592128252402e-05, "loss": 0.1459, "step": 4310 }, { "grad_norm": 1.653929591178894, "learning_rate": 9.999565301999437e-05, "loss": 0.1492, "step": 4320 }, { "grad_norm": 1.0696467161178589, "learning_rate": 9.999537621457502e-05, "loss": 0.1609, "step": 4330 }, { "grad_norm": 1.1307257413864136, "learning_rate": 9.999509086631323e-05, "loss": 0.1295, "step": 4340 }, { "grad_norm": 1.0503162145614624, "learning_rate": 9.99947969752578e-05, "loss": 0.138, "step": 4350 }, { "grad_norm": 0.8730031847953796, "learning_rate": 9.999449454145891e-05, "loss": 0.1569, "step": 4360 }, { "grad_norm": 0.956855833530426, "learning_rate": 9.999418356496827e-05, "loss": 0.1386, "step": 4370 }, { "grad_norm": 0.8628953695297241, "learning_rate": 9.999386404583899e-05, "loss": 0.1337, "step": 4380 }, { "grad_norm": 1.1208082437515259, "learning_rate": 9.999353598412568e-05, "loss": 0.1511, "step": 4390 }, { "grad_norm": 1.1776121854782104, "learning_rate": 9.999319937988442e-05, "loss": 0.126, "step": 4400 }, { "grad_norm": 0.7995575070381165, "learning_rate": 9.999285423317268e-05, "loss": 0.167, "step": 4410 }, { "grad_norm": 0.7223640084266663, "learning_rate": 9.999250054404947e-05, "loss": 0.1254, "step": 4420 }, { "grad_norm": 0.9793170094490051, "learning_rate": 9.99921383125752e-05, "loss": 0.1465, "step": 4430 }, { "grad_norm": 0.714893102645874, "learning_rate": 9.99917675388118e-05, "loss": 0.1734, "step": 4440 }, { "grad_norm": 0.9832234382629395, "learning_rate": 9.99913882228226e-05, "loss": 0.1363, "step": 4450 }, { "grad_norm": 1.178597331047058, "learning_rate": 9.999100036467242e-05, "loss": 0.1604, "step": 4460 }, { "grad_norm": 1.1509236097335815, "learning_rate": 9.999060396442753e-05, "loss": 0.1354, "step": 4470 }, { "grad_norm": 1.089914321899414, "learning_rate": 9.999019902215566e-05, "loss": 0.1403, "step": 4480 }, { "grad_norm": 1.1061283349990845, "learning_rate": 9.998978553792602e-05, "loss": 0.1373, "step": 4490 }, { "grad_norm": 1.0683648586273193, "learning_rate": 9.998936351180926e-05, "loss": 0.1605, "step": 4500 }, { "grad_norm": 0.9987396597862244, "learning_rate": 9.998893294387747e-05, "loss": 0.1464, "step": 4510 }, { "grad_norm": 0.9771728515625, "learning_rate": 9.998849383420426e-05, "loss": 0.1258, "step": 4520 }, { "grad_norm": 1.1338224411010742, "learning_rate": 9.998804618286465e-05, "loss": 0.1142, "step": 4530 }, { "grad_norm": 1.1945017576217651, "learning_rate": 9.99875899899351e-05, "loss": 0.1378, "step": 4540 }, { "grad_norm": 0.9256120324134827, "learning_rate": 9.99871252554936e-05, "loss": 0.1318, "step": 4550 }, { "grad_norm": 1.5076935291290283, "learning_rate": 9.998665197961955e-05, "loss": 0.145, "step": 4560 }, { "grad_norm": 1.2068806886672974, "learning_rate": 9.998617016239379e-05, "loss": 0.1848, "step": 4570 }, { "grad_norm": 1.0651627779006958, "learning_rate": 9.998567980389869e-05, "loss": 0.1643, "step": 4580 }, { "grad_norm": 0.7126860022544861, "learning_rate": 9.998518090421802e-05, "loss": 0.1101, "step": 4590 }, { "grad_norm": 1.0340099334716797, "learning_rate": 9.998467346343703e-05, "loss": 0.1282, "step": 4600 }, { "grad_norm": 0.9176605939865112, "learning_rate": 9.998415748164243e-05, "loss": 0.131, "step": 4610 }, { "grad_norm": 1.1985607147216797, "learning_rate": 9.998363295892238e-05, "loss": 0.1333, "step": 4620 }, { "grad_norm": 0.9129343032836914, "learning_rate": 9.998309989536652e-05, "loss": 0.1185, "step": 4630 }, { "grad_norm": 0.7268670201301575, "learning_rate": 9.998255829106593e-05, "loss": 0.1074, "step": 4640 }, { "grad_norm": 1.2353579998016357, "learning_rate": 9.998200814611316e-05, "loss": 0.1504, "step": 4650 }, { "grad_norm": 0.9157158732414246, "learning_rate": 9.998144946060219e-05, "loss": 0.1552, "step": 4660 }, { "grad_norm": 0.91201251745224, "learning_rate": 9.998088223462852e-05, "loss": 0.1403, "step": 4670 }, { "grad_norm": 0.8304502367973328, "learning_rate": 9.998030646828905e-05, "loss": 0.1382, "step": 4680 }, { "grad_norm": 0.8799469470977783, "learning_rate": 9.997972216168217e-05, "loss": 0.1321, "step": 4690 }, { "grad_norm": 1.0399174690246582, "learning_rate": 9.997912931490771e-05, "loss": 0.1474, "step": 4700 }, { "grad_norm": 1.0595693588256836, "learning_rate": 9.9978527928067e-05, "loss": 0.1372, "step": 4710 }, { "grad_norm": 0.981183648109436, "learning_rate": 9.997791800126277e-05, "loss": 0.1634, "step": 4720 }, { "grad_norm": 0.9124171137809753, "learning_rate": 9.997729953459927e-05, "loss": 0.1467, "step": 4730 }, { "grad_norm": 0.9726876020431519, "learning_rate": 9.997667252818214e-05, "loss": 0.1213, "step": 4740 }, { "grad_norm": 1.188069224357605, "learning_rate": 9.997603698211855e-05, "loss": 0.1418, "step": 4750 }, { "grad_norm": 1.575325846672058, "learning_rate": 9.99753928965171e-05, "loss": 0.136, "step": 4760 }, { "grad_norm": 0.6705055236816406, "learning_rate": 9.997474027148781e-05, "loss": 0.148, "step": 4770 }, { "grad_norm": 0.890496015548706, "learning_rate": 9.997407910714223e-05, "loss": 0.1431, "step": 4780 }, { "grad_norm": 1.162657618522644, "learning_rate": 9.997340940359332e-05, "loss": 0.1358, "step": 4790 }, { "grad_norm": 0.8369739055633545, "learning_rate": 9.997273116095552e-05, "loss": 0.1299, "step": 4800 }, { "grad_norm": 0.7074972987174988, "learning_rate": 9.997204437934473e-05, "loss": 0.1416, "step": 4810 }, { "grad_norm": 0.8618525862693787, "learning_rate": 9.997134905887829e-05, "loss": 0.1401, "step": 4820 }, { "grad_norm": 0.9913590550422668, "learning_rate": 9.997064519967501e-05, "loss": 0.189, "step": 4830 }, { "grad_norm": 0.8403137922286987, "learning_rate": 9.996993280185517e-05, "loss": 0.1498, "step": 4840 }, { "grad_norm": 1.4645662307739258, "learning_rate": 9.99692118655405e-05, "loss": 0.1523, "step": 4850 }, { "grad_norm": 0.9440006613731384, "learning_rate": 9.996848239085417e-05, "loss": 0.1266, "step": 4860 }, { "grad_norm": 0.7700343728065491, "learning_rate": 9.996774437792085e-05, "loss": 0.1496, "step": 4870 }, { "grad_norm": 0.9653757214546204, "learning_rate": 9.996699782686664e-05, "loss": 0.1362, "step": 4880 }, { "grad_norm": 0.8223907947540283, "learning_rate": 9.996624273781909e-05, "loss": 0.1405, "step": 4890 }, { "grad_norm": 1.2616294622421265, "learning_rate": 9.996547911090725e-05, "loss": 0.14, "step": 4900 }, { "grad_norm": 1.0720597505569458, "learning_rate": 9.996470694626157e-05, "loss": 0.1727, "step": 4910 }, { "grad_norm": 1.2121844291687012, "learning_rate": 9.996392624401403e-05, "loss": 0.1225, "step": 4920 }, { "grad_norm": 0.9629889726638794, "learning_rate": 9.996313700429801e-05, "loss": 0.1376, "step": 4930 }, { "grad_norm": 0.9429501295089722, "learning_rate": 9.996233922724836e-05, "loss": 0.1194, "step": 4940 }, { "grad_norm": 0.7798393368721008, "learning_rate": 9.996153291300141e-05, "loss": 0.137, "step": 4950 }, { "grad_norm": 1.066105604171753, "learning_rate": 9.996071806169494e-05, "loss": 0.1552, "step": 4960 }, { "grad_norm": 0.9465988278388977, "learning_rate": 9.995989467346817e-05, "loss": 0.1346, "step": 4970 }, { "grad_norm": 0.9171234965324402, "learning_rate": 9.995906274846183e-05, "loss": 0.1329, "step": 4980 }, { "grad_norm": 0.976791501045227, "learning_rate": 9.995822228681803e-05, "loss": 0.1271, "step": 4990 }, { "grad_norm": 0.8455252051353455, "learning_rate": 9.99573732886804e-05, "loss": 0.1214, "step": 5000 }, { "grad_norm": 1.0003242492675781, "learning_rate": 9.995651575419402e-05, "loss": 0.1322, "step": 5010 }, { "grad_norm": 1.0489698648452759, "learning_rate": 9.995564968350541e-05, "loss": 0.1177, "step": 5020 }, { "grad_norm": 1.1392399072647095, "learning_rate": 9.995477507676256e-05, "loss": 0.1389, "step": 5030 }, { "grad_norm": 0.8040144443511963, "learning_rate": 9.995389193411493e-05, "loss": 0.1143, "step": 5040 }, { "grad_norm": 0.8556502461433411, "learning_rate": 9.995300025571339e-05, "loss": 0.1211, "step": 5050 }, { "grad_norm": 0.8533415198326111, "learning_rate": 9.995210004171034e-05, "loss": 0.103, "step": 5060 }, { "grad_norm": 1.060381531715393, "learning_rate": 9.995119129225956e-05, "loss": 0.1405, "step": 5070 }, { "grad_norm": 0.8363825678825378, "learning_rate": 9.995027400751637e-05, "loss": 0.0941, "step": 5080 }, { "grad_norm": 0.675698459148407, "learning_rate": 9.994934818763751e-05, "loss": 0.142, "step": 5090 }, { "grad_norm": 0.7856559157371521, "learning_rate": 9.994841383278115e-05, "loss": 0.1051, "step": 5100 }, { "grad_norm": 0.7441729307174683, "learning_rate": 9.994747094310695e-05, "loss": 0.1353, "step": 5110 }, { "grad_norm": 0.8222783207893372, "learning_rate": 9.994651951877604e-05, "loss": 0.114, "step": 5120 }, { "grad_norm": 1.1236411333084106, "learning_rate": 9.994555955995099e-05, "loss": 0.1115, "step": 5130 }, { "grad_norm": 0.5882471799850464, "learning_rate": 9.994459106679581e-05, "loss": 0.128, "step": 5140 }, { "grad_norm": 0.788363516330719, "learning_rate": 9.994361403947603e-05, "loss": 0.1466, "step": 5150 }, { "grad_norm": 1.0162791013717651, "learning_rate": 9.994262847815854e-05, "loss": 0.1417, "step": 5160 }, { "grad_norm": 1.1087762117385864, "learning_rate": 9.99416343830118e-05, "loss": 0.1388, "step": 5170 }, { "grad_norm": 0.73259037733078, "learning_rate": 9.994063175420565e-05, "loss": 0.11, "step": 5180 }, { "grad_norm": 0.6945235729217529, "learning_rate": 9.99396205919114e-05, "loss": 0.1502, "step": 5190 }, { "grad_norm": 1.0007797479629517, "learning_rate": 9.993860089630185e-05, "loss": 0.1183, "step": 5200 }, { "grad_norm": 0.9012766480445862, "learning_rate": 9.993757266755123e-05, "loss": 0.1198, "step": 5210 }, { "grad_norm": 1.20260751247406, "learning_rate": 9.993653590583522e-05, "loss": 0.156, "step": 5220 }, { "grad_norm": 0.6153502464294434, "learning_rate": 9.993549061133102e-05, "loss": 0.1051, "step": 5230 }, { "grad_norm": 0.840140700340271, "learning_rate": 9.993443678421719e-05, "loss": 0.1102, "step": 5240 }, { "grad_norm": 0.677542507648468, "learning_rate": 9.993337442467384e-05, "loss": 0.1167, "step": 5250 }, { "grad_norm": 0.8430750370025635, "learning_rate": 9.993230353288248e-05, "loss": 0.1462, "step": 5260 }, { "grad_norm": 0.7482090592384338, "learning_rate": 9.993122410902608e-05, "loss": 0.1205, "step": 5270 }, { "grad_norm": 0.8475335836410522, "learning_rate": 9.993013615328912e-05, "loss": 0.1495, "step": 5280 }, { "grad_norm": 1.3008928298950195, "learning_rate": 9.992903966585747e-05, "loss": 0.1402, "step": 5290 }, { "grad_norm": 0.7475801110267639, "learning_rate": 9.992793464691852e-05, "loss": 0.1341, "step": 5300 }, { "grad_norm": 0.9961963295936584, "learning_rate": 9.992682109666105e-05, "loss": 0.1468, "step": 5310 }, { "grad_norm": 0.9029371738433838, "learning_rate": 9.992569901527538e-05, "loss": 0.1283, "step": 5320 }, { "grad_norm": 0.9746913313865662, "learning_rate": 9.99245684029532e-05, "loss": 0.1366, "step": 5330 }, { "grad_norm": 1.0848642587661743, "learning_rate": 9.992342925988774e-05, "loss": 0.1426, "step": 5340 }, { "grad_norm": 0.6137527823448181, "learning_rate": 9.992228158627361e-05, "loss": 0.16, "step": 5350 }, { "grad_norm": 1.056349277496338, "learning_rate": 9.992112538230693e-05, "loss": 0.1686, "step": 5360 }, { "grad_norm": 0.8254979848861694, "learning_rate": 9.991996064818527e-05, "loss": 0.141, "step": 5370 }, { "grad_norm": 1.1938321590423584, "learning_rate": 9.991878738410768e-05, "loss": 0.1373, "step": 5380 }, { "grad_norm": 0.7702672481536865, "learning_rate": 9.991760559027457e-05, "loss": 0.1418, "step": 5390 }, { "grad_norm": 0.8477912545204163, "learning_rate": 9.991641526688793e-05, "loss": 0.1001, "step": 5400 }, { "grad_norm": 0.8927893042564392, "learning_rate": 9.991521641415113e-05, "loss": 0.1108, "step": 5410 }, { "grad_norm": 0.8231714367866516, "learning_rate": 9.991400903226904e-05, "loss": 0.1324, "step": 5420 }, { "grad_norm": 0.9267168641090393, "learning_rate": 9.991279312144794e-05, "loss": 0.1233, "step": 5430 }, { "grad_norm": 1.6747153997421265, "learning_rate": 9.991156868189564e-05, "loss": 0.1749, "step": 5440 }, { "grad_norm": 0.8645277619361877, "learning_rate": 9.991033571382131e-05, "loss": 0.1272, "step": 5450 }, { "grad_norm": 0.8296321034431458, "learning_rate": 9.990909421743569e-05, "loss": 0.1052, "step": 5460 }, { "grad_norm": 0.9910750985145569, "learning_rate": 9.990784419295085e-05, "loss": 0.1048, "step": 5470 }, { "grad_norm": 0.6049179434776306, "learning_rate": 9.990658564058044e-05, "loss": 0.1253, "step": 5480 }, { "grad_norm": 0.9434384703636169, "learning_rate": 9.990531856053948e-05, "loss": 0.1238, "step": 5490 }, { "grad_norm": 0.8082303404808044, "learning_rate": 9.99040429530445e-05, "loss": 0.129, "step": 5500 }, { "grad_norm": 0.6439273953437805, "learning_rate": 9.990275881831346e-05, "loss": 0.1008, "step": 5510 }, { "grad_norm": 1.0379414558410645, "learning_rate": 9.990146615656577e-05, "loss": 0.1265, "step": 5520 }, { "grad_norm": 0.5780282020568848, "learning_rate": 9.990016496802233e-05, "loss": 0.1064, "step": 5530 }, { "grad_norm": 0.852206289768219, "learning_rate": 9.989885525290548e-05, "loss": 0.1342, "step": 5540 }, { "grad_norm": 1.322954773902893, "learning_rate": 9.989753701143897e-05, "loss": 0.1462, "step": 5550 }, { "grad_norm": 0.6917036175727844, "learning_rate": 9.989621024384812e-05, "loss": 0.134, "step": 5560 }, { "grad_norm": 0.692475438117981, "learning_rate": 9.989487495035959e-05, "loss": 0.1397, "step": 5570 }, { "grad_norm": 0.9798738956451416, "learning_rate": 9.989353113120156e-05, "loss": 0.1208, "step": 5580 }, { "grad_norm": 0.8597458004951477, "learning_rate": 9.989217878660366e-05, "loss": 0.1154, "step": 5590 }, { "grad_norm": 1.2124522924423218, "learning_rate": 9.989081791679695e-05, "loss": 0.1245, "step": 5600 }, { "grad_norm": 1.3898407220840454, "learning_rate": 9.988944852201397e-05, "loss": 0.1347, "step": 5610 }, { "grad_norm": 0.8765997290611267, "learning_rate": 9.988807060248873e-05, "loss": 0.1697, "step": 5620 }, { "grad_norm": 0.820453941822052, "learning_rate": 9.988668415845665e-05, "loss": 0.1405, "step": 5630 }, { "grad_norm": 0.8888039588928223, "learning_rate": 9.988528919015466e-05, "loss": 0.1268, "step": 5640 }, { "grad_norm": 0.9242334961891174, "learning_rate": 9.988388569782112e-05, "loss": 0.1068, "step": 5650 }, { "grad_norm": 0.8557186722755432, "learning_rate": 9.988247368169583e-05, "loss": 0.0992, "step": 5660 }, { "grad_norm": 0.8699058294296265, "learning_rate": 9.988105314202007e-05, "loss": 0.121, "step": 5670 }, { "grad_norm": 0.9741378426551819, "learning_rate": 9.987962407903659e-05, "loss": 0.119, "step": 5680 }, { "grad_norm": 0.6879732012748718, "learning_rate": 9.987818649298957e-05, "loss": 0.1388, "step": 5690 }, { "grad_norm": 1.0079014301300049, "learning_rate": 9.987674038412465e-05, "loss": 0.1114, "step": 5700 }, { "grad_norm": 0.8391895890235901, "learning_rate": 9.987528575268891e-05, "loss": 0.1102, "step": 5710 }, { "grad_norm": 1.0270448923110962, "learning_rate": 9.987382259893095e-05, "loss": 0.1395, "step": 5720 }, { "grad_norm": 0.7285240292549133, "learning_rate": 9.987235092310074e-05, "loss": 0.1299, "step": 5730 }, { "grad_norm": 0.8271536231040955, "learning_rate": 9.987087072544978e-05, "loss": 0.1107, "step": 5740 }, { "grad_norm": 0.9623945951461792, "learning_rate": 9.9869382006231e-05, "loss": 0.1292, "step": 5750 }, { "grad_norm": 0.9488499164581299, "learning_rate": 9.986788476569875e-05, "loss": 0.127, "step": 5760 }, { "grad_norm": 0.992723822593689, "learning_rate": 9.986637900410887e-05, "loss": 0.1252, "step": 5770 }, { "grad_norm": 1.4536932706832886, "learning_rate": 9.986486472171869e-05, "loss": 0.1343, "step": 5780 }, { "grad_norm": 0.9625011682510376, "learning_rate": 9.986334191878692e-05, "loss": 0.1385, "step": 5790 }, { "grad_norm": 0.8373525738716125, "learning_rate": 9.986181059557378e-05, "loss": 0.1458, "step": 5800 }, { "grad_norm": 0.8409951329231262, "learning_rate": 9.986027075234094e-05, "loss": 0.1195, "step": 5810 }, { "grad_norm": 0.9859687089920044, "learning_rate": 9.985872238935152e-05, "loss": 0.1261, "step": 5820 }, { "grad_norm": 0.7663140892982483, "learning_rate": 9.985716550687008e-05, "loss": 0.1299, "step": 5830 }, { "grad_norm": 1.3581867218017578, "learning_rate": 9.985560010516264e-05, "loss": 0.1254, "step": 5840 }, { "grad_norm": 0.7906031012535095, "learning_rate": 9.985402618449668e-05, "loss": 0.099, "step": 5850 }, { "grad_norm": 0.7540070414543152, "learning_rate": 9.985244374514118e-05, "loss": 0.1002, "step": 5860 }, { "grad_norm": 0.6904850602149963, "learning_rate": 9.985085278736651e-05, "loss": 0.1047, "step": 5870 }, { "grad_norm": 1.1079195737838745, "learning_rate": 9.984925331144452e-05, "loss": 0.112, "step": 5880 }, { "grad_norm": 0.7521240711212158, "learning_rate": 9.984764531764851e-05, "loss": 0.1003, "step": 5890 }, { "grad_norm": 0.6444527506828308, "learning_rate": 9.984602880625326e-05, "loss": 0.0946, "step": 5900 }, { "grad_norm": 0.6847091913223267, "learning_rate": 9.9844403777535e-05, "loss": 0.1384, "step": 5910 }, { "grad_norm": 0.6749064326286316, "learning_rate": 9.984277023177135e-05, "loss": 0.0924, "step": 5920 }, { "grad_norm": 0.7189170718193054, "learning_rate": 9.984112816924148e-05, "loss": 0.0956, "step": 5930 }, { "grad_norm": 0.8109831809997559, "learning_rate": 9.983947759022596e-05, "loss": 0.1127, "step": 5940 }, { "grad_norm": 0.5801900625228882, "learning_rate": 9.983781849500682e-05, "loss": 0.1101, "step": 5950 }, { "grad_norm": 0.8723872303962708, "learning_rate": 9.98361508838676e-05, "loss": 0.1093, "step": 5960 }, { "grad_norm": 0.8312374949455261, "learning_rate": 9.98344747570932e-05, "loss": 0.1069, "step": 5970 }, { "grad_norm": 1.023483157157898, "learning_rate": 9.983279011497004e-05, "loss": 0.1503, "step": 5980 }, { "grad_norm": 0.8804636001586914, "learning_rate": 9.983109695778596e-05, "loss": 0.1379, "step": 5990 }, { "grad_norm": 0.8683323860168457, "learning_rate": 9.982939528583032e-05, "loss": 0.1108, "step": 6000 }, { "grad_norm": 0.65106201171875, "learning_rate": 9.982768509939385e-05, "loss": 0.1008, "step": 6010 }, { "grad_norm": 1.2239795923233032, "learning_rate": 9.982596639876879e-05, "loss": 0.0931, "step": 6020 }, { "grad_norm": 0.7593650221824646, "learning_rate": 9.982423918424881e-05, "loss": 0.1091, "step": 6030 }, { "grad_norm": 0.9073904156684875, "learning_rate": 9.982250345612908e-05, "loss": 0.1113, "step": 6040 }, { "grad_norm": 0.7225806713104248, "learning_rate": 9.982075921470611e-05, "loss": 0.1194, "step": 6050 }, { "grad_norm": 0.653238832950592, "learning_rate": 9.981900646027802e-05, "loss": 0.1051, "step": 6060 }, { "grad_norm": 0.7486824989318848, "learning_rate": 9.981724519314425e-05, "loss": 0.1084, "step": 6070 }, { "grad_norm": 0.8365575075149536, "learning_rate": 9.981547541360581e-05, "loss": 0.1293, "step": 6080 }, { "grad_norm": 0.9656510949134827, "learning_rate": 9.981369712196508e-05, "loss": 0.1287, "step": 6090 }, { "grad_norm": 0.7241118550300598, "learning_rate": 9.981191031852592e-05, "loss": 0.1241, "step": 6100 }, { "grad_norm": 0.6503105759620667, "learning_rate": 9.981011500359362e-05, "loss": 0.1236, "step": 6110 }, { "grad_norm": 0.8336143493652344, "learning_rate": 9.9808311177475e-05, "loss": 0.1138, "step": 6120 }, { "grad_norm": 1.084769606590271, "learning_rate": 9.980649884047826e-05, "loss": 0.091, "step": 6130 }, { "grad_norm": 1.1764901876449585, "learning_rate": 9.980467799291307e-05, "loss": 0.0982, "step": 6140 }, { "grad_norm": 0.7258387207984924, "learning_rate": 9.980284863509058e-05, "loss": 0.112, "step": 6150 }, { "grad_norm": 0.7645953297615051, "learning_rate": 9.980101076732334e-05, "loss": 0.1295, "step": 6160 }, { "grad_norm": 1.088098406791687, "learning_rate": 9.979916438992544e-05, "loss": 0.1233, "step": 6170 }, { "grad_norm": 0.7395375967025757, "learning_rate": 9.979730950321237e-05, "loss": 0.1343, "step": 6180 }, { "grad_norm": 0.6850581765174866, "learning_rate": 9.979544610750104e-05, "loss": 0.0942, "step": 6190 }, { "grad_norm": 0.7372778654098511, "learning_rate": 9.97935742031099e-05, "loss": 0.1152, "step": 6200 }, { "grad_norm": 0.705208957195282, "learning_rate": 9.979169379035878e-05, "loss": 0.1282, "step": 6210 }, { "grad_norm": 0.8541697859764099, "learning_rate": 9.978980486956899e-05, "loss": 0.1009, "step": 6220 }, { "grad_norm": 0.9607712030410767, "learning_rate": 9.978790744106332e-05, "loss": 0.094, "step": 6230 }, { "grad_norm": 0.8388890624046326, "learning_rate": 9.978600150516594e-05, "loss": 0.0903, "step": 6240 }, { "grad_norm": 0.9612566232681274, "learning_rate": 9.978408706220259e-05, "loss": 0.0847, "step": 6250 }, { "grad_norm": 0.7367846369743347, "learning_rate": 9.978216411250032e-05, "loss": 0.1116, "step": 6260 }, { "grad_norm": 0.9141402244567871, "learning_rate": 9.978023265638778e-05, "loss": 0.1149, "step": 6270 }, { "grad_norm": 0.8614202737808228, "learning_rate": 9.977829269419495e-05, "loss": 0.1152, "step": 6280 }, { "grad_norm": 0.7045432329177856, "learning_rate": 9.977634422625335e-05, "loss": 0.0905, "step": 6290 }, { "grad_norm": 1.0019145011901855, "learning_rate": 9.97743872528959e-05, "loss": 0.1052, "step": 6300 }, { "grad_norm": 1.1030060052871704, "learning_rate": 9.9772421774457e-05, "loss": 0.1053, "step": 6310 }, { "grad_norm": 1.0137783288955688, "learning_rate": 9.977044779127252e-05, "loss": 0.125, "step": 6320 }, { "grad_norm": 0.9236677885055542, "learning_rate": 9.976846530367971e-05, "loss": 0.1098, "step": 6330 }, { "grad_norm": 0.9029295444488525, "learning_rate": 9.976647431201735e-05, "loss": 0.0976, "step": 6340 }, { "grad_norm": 0.9405393004417419, "learning_rate": 9.976447481662568e-05, "loss": 0.1337, "step": 6350 }, { "grad_norm": 0.9345237612724304, "learning_rate": 9.976246681784629e-05, "loss": 0.1366, "step": 6360 }, { "grad_norm": 0.9520278573036194, "learning_rate": 9.976045031602234e-05, "loss": 0.1116, "step": 6370 }, { "grad_norm": 0.881540834903717, "learning_rate": 9.975842531149837e-05, "loss": 0.1239, "step": 6380 }, { "grad_norm": 0.6556492447853088, "learning_rate": 9.975639180462043e-05, "loss": 0.1091, "step": 6390 }, { "grad_norm": 0.5849432349205017, "learning_rate": 9.975434979573596e-05, "loss": 0.1273, "step": 6400 }, { "grad_norm": 0.925284743309021, "learning_rate": 9.97522992851939e-05, "loss": 0.108, "step": 6410 }, { "grad_norm": 0.8121129870414734, "learning_rate": 9.975024027334461e-05, "loss": 0.098, "step": 6420 }, { "grad_norm": 1.046255111694336, "learning_rate": 9.974817276053993e-05, "loss": 0.1294, "step": 6430 }, { "grad_norm": 0.6514682769775391, "learning_rate": 9.974609674713315e-05, "loss": 0.1463, "step": 6440 }, { "grad_norm": 1.1543004512786865, "learning_rate": 9.9744012233479e-05, "loss": 0.14, "step": 6450 }, { "grad_norm": 0.7017199397087097, "learning_rate": 9.974191921993366e-05, "loss": 0.0972, "step": 6460 }, { "grad_norm": 0.7658441066741943, "learning_rate": 9.973981770685474e-05, "loss": 0.1223, "step": 6470 }, { "grad_norm": 0.843795657157898, "learning_rate": 9.97377076946014e-05, "loss": 0.0887, "step": 6480 }, { "grad_norm": 0.9538660049438477, "learning_rate": 9.973558918353412e-05, "loss": 0.0966, "step": 6490 }, { "grad_norm": 0.6984360218048096, "learning_rate": 9.973346217401494e-05, "loss": 0.122, "step": 6500 }, { "grad_norm": 0.6779630184173584, "learning_rate": 9.973132666640726e-05, "loss": 0.108, "step": 6510 }, { "grad_norm": 0.7405104041099548, "learning_rate": 9.972918266107602e-05, "loss": 0.1077, "step": 6520 }, { "grad_norm": 0.693120539188385, "learning_rate": 9.972703015838756e-05, "loss": 0.1095, "step": 6530 }, { "grad_norm": 0.8203225135803223, "learning_rate": 9.97248691587097e-05, "loss": 0.1214, "step": 6540 }, { "grad_norm": 1.0775237083435059, "learning_rate": 9.972269966241166e-05, "loss": 0.1011, "step": 6550 }, { "grad_norm": 1.0021036863327026, "learning_rate": 9.972052166986417e-05, "loss": 0.0926, "step": 6560 }, { "grad_norm": 1.371736764907837, "learning_rate": 9.971833518143938e-05, "loss": 0.1052, "step": 6570 }, { "grad_norm": 0.8303551077842712, "learning_rate": 9.971614019751093e-05, "loss": 0.1074, "step": 6580 }, { "grad_norm": 0.853589653968811, "learning_rate": 9.971393671845383e-05, "loss": 0.0956, "step": 6590 }, { "grad_norm": 0.7498068809509277, "learning_rate": 9.971172474464464e-05, "loss": 0.1056, "step": 6600 }, { "grad_norm": 0.5822083353996277, "learning_rate": 9.97095042764613e-05, "loss": 0.1125, "step": 6610 }, { "grad_norm": 0.7331120371818542, "learning_rate": 9.970727531428324e-05, "loss": 0.0943, "step": 6620 }, { "grad_norm": 0.6966016292572021, "learning_rate": 9.970503785849132e-05, "loss": 0.1175, "step": 6630 }, { "grad_norm": 0.6120558381080627, "learning_rate": 9.970279190946788e-05, "loss": 0.1103, "step": 6640 }, { "grad_norm": 0.9493107199668884, "learning_rate": 9.970053746759667e-05, "loss": 0.1237, "step": 6650 }, { "grad_norm": 0.7862327098846436, "learning_rate": 9.969827453326292e-05, "loss": 0.114, "step": 6660 }, { "grad_norm": 0.8472309708595276, "learning_rate": 9.969600310685332e-05, "loss": 0.1191, "step": 6670 }, { "grad_norm": 0.7310095429420471, "learning_rate": 9.969372318875596e-05, "loss": 0.1122, "step": 6680 }, { "grad_norm": 0.8303012251853943, "learning_rate": 9.969143477936043e-05, "loss": 0.1179, "step": 6690 }, { "grad_norm": 0.6633405089378357, "learning_rate": 9.968913787905775e-05, "loss": 0.0978, "step": 6700 }, { "grad_norm": 0.7294585108757019, "learning_rate": 9.968683248824045e-05, "loss": 0.0934, "step": 6710 }, { "grad_norm": 0.7431014776229858, "learning_rate": 9.968451860730238e-05, "loss": 0.11, "step": 6720 }, { "grad_norm": 0.5568671822547913, "learning_rate": 9.968219623663896e-05, "loss": 0.0985, "step": 6730 }, { "grad_norm": 0.5924029350280762, "learning_rate": 9.967986537664702e-05, "loss": 0.0904, "step": 6740 }, { "grad_norm": 0.8484299182891846, "learning_rate": 9.967752602772483e-05, "loss": 0.1275, "step": 6750 }, { "grad_norm": 0.5771042108535767, "learning_rate": 9.967517819027212e-05, "loss": 0.1542, "step": 6760 }, { "grad_norm": 0.7003229856491089, "learning_rate": 9.967282186469009e-05, "loss": 0.0863, "step": 6770 }, { "grad_norm": 0.7804136276245117, "learning_rate": 9.967045705138135e-05, "loss": 0.0986, "step": 6780 }, { "grad_norm": 0.9038034677505493, "learning_rate": 9.966808375074998e-05, "loss": 0.1231, "step": 6790 }, { "grad_norm": 0.9057369232177734, "learning_rate": 9.966570196320154e-05, "loss": 0.1061, "step": 6800 }, { "grad_norm": 0.8583527207374573, "learning_rate": 9.966331168914299e-05, "loss": 0.1198, "step": 6810 }, { "grad_norm": 0.8581283092498779, "learning_rate": 9.966091292898277e-05, "loss": 0.1026, "step": 6820 }, { "grad_norm": 0.7016920447349548, "learning_rate": 9.965850568313076e-05, "loss": 0.0931, "step": 6830 }, { "grad_norm": 1.0150614976882935, "learning_rate": 9.965608995199827e-05, "loss": 0.0928, "step": 6840 }, { "grad_norm": 0.9012296199798584, "learning_rate": 9.965366573599812e-05, "loss": 0.0922, "step": 6850 }, { "grad_norm": 0.6045441031455994, "learning_rate": 9.965123303554453e-05, "loss": 0.1116, "step": 6860 }, { "grad_norm": 0.9061306118965149, "learning_rate": 9.964879185105317e-05, "loss": 0.1013, "step": 6870 }, { "grad_norm": 0.7315981388092041, "learning_rate": 9.964634218294119e-05, "loss": 0.128, "step": 6880 }, { "grad_norm": 0.7219467759132385, "learning_rate": 9.964388403162714e-05, "loss": 0.0912, "step": 6890 }, { "grad_norm": 0.8771642446517944, "learning_rate": 9.96414173975311e-05, "loss": 0.1019, "step": 6900 }, { "grad_norm": 0.7830932140350342, "learning_rate": 9.963894228107451e-05, "loss": 0.0978, "step": 6910 }, { "grad_norm": 0.5676888227462769, "learning_rate": 9.963645868268032e-05, "loss": 0.09, "step": 6920 }, { "grad_norm": 1.047170639038086, "learning_rate": 9.963396660277289e-05, "loss": 0.1375, "step": 6930 }, { "grad_norm": 0.7574903964996338, "learning_rate": 9.963146604177807e-05, "loss": 0.0979, "step": 6940 }, { "grad_norm": 0.9704874157905579, "learning_rate": 9.962895700012311e-05, "loss": 0.1184, "step": 6950 }, { "grad_norm": 0.971198320388794, "learning_rate": 9.962643947823677e-05, "loss": 0.1006, "step": 6960 }, { "grad_norm": 0.7667521238327026, "learning_rate": 9.962391347654921e-05, "loss": 0.0914, "step": 6970 }, { "grad_norm": 0.7332438230514526, "learning_rate": 9.962137899549204e-05, "loss": 0.1057, "step": 6980 }, { "grad_norm": 0.6622326374053955, "learning_rate": 9.961883603549835e-05, "loss": 0.086, "step": 6990 }, { "grad_norm": 0.6018088459968567, "learning_rate": 9.961628459700267e-05, "loss": 0.0787, "step": 7000 }, { "grad_norm": 0.6755437850952148, "learning_rate": 9.961372468044095e-05, "loss": 0.0948, "step": 7010 }, { "grad_norm": 0.6310155391693115, "learning_rate": 9.961115628625062e-05, "loss": 0.0945, "step": 7020 }, { "grad_norm": 0.9381420612335205, "learning_rate": 9.960857941487056e-05, "loss": 0.1019, "step": 7030 }, { "grad_norm": 0.8643722534179688, "learning_rate": 9.960599406674106e-05, "loss": 0.1095, "step": 7040 }, { "grad_norm": 0.8444974422454834, "learning_rate": 9.960340024230393e-05, "loss": 0.104, "step": 7050 }, { "grad_norm": 0.5583925843238831, "learning_rate": 9.960079794200232e-05, "loss": 0.0949, "step": 7060 }, { "grad_norm": 0.7356012463569641, "learning_rate": 9.959818716628096e-05, "loss": 0.0926, "step": 7070 }, { "grad_norm": 1.0281096696853638, "learning_rate": 9.95955679155859e-05, "loss": 0.0999, "step": 7080 }, { "grad_norm": 1.251335620880127, "learning_rate": 9.959294019036472e-05, "loss": 0.095, "step": 7090 }, { "grad_norm": 0.737169086933136, "learning_rate": 9.959030399106646e-05, "loss": 0.1161, "step": 7100 }, { "grad_norm": 0.9912868142127991, "learning_rate": 9.958765931814153e-05, "loss": 0.09, "step": 7110 }, { "grad_norm": 0.5983666777610779, "learning_rate": 9.958500617204184e-05, "loss": 0.0862, "step": 7120 }, { "grad_norm": 0.7416418194770813, "learning_rate": 9.958234455322075e-05, "loss": 0.0958, "step": 7130 }, { "grad_norm": 0.7426813244819641, "learning_rate": 9.957967446213308e-05, "loss": 0.1095, "step": 7140 }, { "grad_norm": 0.9087061882019043, "learning_rate": 9.957699589923501e-05, "loss": 0.1325, "step": 7150 }, { "grad_norm": 0.7648048400878906, "learning_rate": 9.957430886498431e-05, "loss": 0.1111, "step": 7160 }, { "grad_norm": 1.0880796909332275, "learning_rate": 9.957161335984008e-05, "loss": 0.1352, "step": 7170 }, { "grad_norm": 0.840420663356781, "learning_rate": 9.956890938426291e-05, "loss": 0.1165, "step": 7180 }, { "grad_norm": 0.8662547469139099, "learning_rate": 9.956619693871482e-05, "loss": 0.1111, "step": 7190 }, { "grad_norm": 0.7112898826599121, "learning_rate": 9.956347602365934e-05, "loss": 0.0973, "step": 7200 }, { "grad_norm": 0.7277844548225403, "learning_rate": 9.956074663956135e-05, "loss": 0.1046, "step": 7210 }, { "grad_norm": 0.6772925853729248, "learning_rate": 9.955800878688726e-05, "loss": 0.1141, "step": 7220 }, { "grad_norm": 1.0284156799316406, "learning_rate": 9.955526246610489e-05, "loss": 0.1153, "step": 7230 }, { "grad_norm": 0.6874121427536011, "learning_rate": 9.955250767768349e-05, "loss": 0.1112, "step": 7240 }, { "grad_norm": 0.5856171250343323, "learning_rate": 9.95497444220938e-05, "loss": 0.0856, "step": 7250 }, { "grad_norm": 0.8924394249916077, "learning_rate": 9.954697269980797e-05, "loss": 0.0941, "step": 7260 }, { "grad_norm": 0.9136117696762085, "learning_rate": 9.954419251129962e-05, "loss": 0.1074, "step": 7270 }, { "grad_norm": 0.7872698307037354, "learning_rate": 9.95414038570438e-05, "loss": 0.1005, "step": 7280 }, { "grad_norm": 0.5870347619056702, "learning_rate": 9.953860673751703e-05, "loss": 0.0934, "step": 7290 }, { "grad_norm": 0.700964093208313, "learning_rate": 9.953580115319725e-05, "loss": 0.1119, "step": 7300 }, { "grad_norm": 1.0994155406951904, "learning_rate": 9.953298710456387e-05, "loss": 0.1183, "step": 7310 }, { "grad_norm": 0.9244146943092346, "learning_rate": 9.953016459209771e-05, "loss": 0.1131, "step": 7320 }, { "grad_norm": 0.8072935342788696, "learning_rate": 9.952733361628108e-05, "loss": 0.1183, "step": 7330 }, { "grad_norm": 0.983232855796814, "learning_rate": 9.952449417759772e-05, "loss": 0.1159, "step": 7340 }, { "grad_norm": 0.863714337348938, "learning_rate": 9.952164627653279e-05, "loss": 0.1167, "step": 7350 }, { "grad_norm": 1.1027182340621948, "learning_rate": 9.951878991357292e-05, "loss": 0.1194, "step": 7360 }, { "grad_norm": 0.6439850926399231, "learning_rate": 9.951592508920622e-05, "loss": 0.1127, "step": 7370 }, { "grad_norm": 0.93580162525177, "learning_rate": 9.951305180392219e-05, "loss": 0.0976, "step": 7380 }, { "grad_norm": 0.6626566648483276, "learning_rate": 9.951017005821178e-05, "loss": 0.1225, "step": 7390 }, { "grad_norm": 0.8584709763526917, "learning_rate": 9.95072798525674e-05, "loss": 0.0831, "step": 7400 }, { "grad_norm": 0.8314740061759949, "learning_rate": 9.950438118748293e-05, "loss": 0.1009, "step": 7410 }, { "grad_norm": 0.8883707523345947, "learning_rate": 9.950147406345366e-05, "loss": 0.121, "step": 7420 }, { "grad_norm": 0.5410902500152588, "learning_rate": 9.949855848097635e-05, "loss": 0.1061, "step": 7430 }, { "grad_norm": 0.9441075921058655, "learning_rate": 9.949563444054916e-05, "loss": 0.1044, "step": 7440 }, { "grad_norm": 0.6254064440727234, "learning_rate": 9.949270194267178e-05, "loss": 0.0975, "step": 7450 }, { "grad_norm": 0.7415446043014526, "learning_rate": 9.948976098784526e-05, "loss": 0.0896, "step": 7460 }, { "grad_norm": 1.015184998512268, "learning_rate": 9.948681157657213e-05, "loss": 0.0972, "step": 7470 }, { "grad_norm": 0.5615154504776001, "learning_rate": 9.948385370935638e-05, "loss": 0.1033, "step": 7480 }, { "grad_norm": 0.8285805583000183, "learning_rate": 9.94808873867034e-05, "loss": 0.1004, "step": 7490 }, { "grad_norm": 0.7084760069847107, "learning_rate": 9.947791260912009e-05, "loss": 0.1139, "step": 7500 }, { "grad_norm": 0.760371744632721, "learning_rate": 9.947492937711474e-05, "loss": 0.0958, "step": 7510 }, { "grad_norm": 0.8390057682991028, "learning_rate": 9.947193769119707e-05, "loss": 0.0916, "step": 7520 }, { "grad_norm": 0.5408549308776855, "learning_rate": 9.946893755187834e-05, "loss": 0.093, "step": 7530 }, { "grad_norm": 1.0908838510513306, "learning_rate": 9.946592895967115e-05, "loss": 0.1191, "step": 7540 }, { "grad_norm": 0.7613018155097961, "learning_rate": 9.94629119150896e-05, "loss": 0.114, "step": 7550 }, { "grad_norm": 0.9294500350952148, "learning_rate": 9.94598864186492e-05, "loss": 0.1292, "step": 7560 }, { "grad_norm": 0.8424614667892456, "learning_rate": 9.945685247086696e-05, "loss": 0.1052, "step": 7570 }, { "grad_norm": 0.685601532459259, "learning_rate": 9.945381007226129e-05, "loss": 0.0925, "step": 7580 }, { "grad_norm": 0.9752089977264404, "learning_rate": 9.945075922335203e-05, "loss": 0.1163, "step": 7590 }, { "grad_norm": 0.9874246716499329, "learning_rate": 9.944769992466049e-05, "loss": 0.1263, "step": 7600 }, { "grad_norm": 0.9374271035194397, "learning_rate": 9.944463217670945e-05, "loss": 0.1062, "step": 7610 }, { "grad_norm": 0.6305373311042786, "learning_rate": 9.944155598002307e-05, "loss": 0.0844, "step": 7620 }, { "grad_norm": 0.7922366857528687, "learning_rate": 9.943847133512701e-05, "loss": 0.1038, "step": 7630 }, { "grad_norm": 0.939365804195404, "learning_rate": 9.943537824254834e-05, "loss": 0.115, "step": 7640 }, { "grad_norm": 1.1197210550308228, "learning_rate": 9.943227670281559e-05, "loss": 0.1111, "step": 7650 }, { "grad_norm": 0.7283608317375183, "learning_rate": 9.942916671645873e-05, "loss": 0.0843, "step": 7660 }, { "grad_norm": 0.6528921127319336, "learning_rate": 9.942604828400916e-05, "loss": 0.081, "step": 7670 }, { "grad_norm": 0.7516102194786072, "learning_rate": 9.942292140599975e-05, "loss": 0.1003, "step": 7680 }, { "grad_norm": 0.7484530806541443, "learning_rate": 9.94197860829648e-05, "loss": 0.1039, "step": 7690 }, { "grad_norm": 0.876538097858429, "learning_rate": 9.941664231544004e-05, "loss": 0.0958, "step": 7700 }, { "grad_norm": 0.6864367127418518, "learning_rate": 9.941349010396264e-05, "loss": 0.0913, "step": 7710 }, { "grad_norm": 0.8464078903198242, "learning_rate": 9.941032944907125e-05, "loss": 0.0962, "step": 7720 }, { "grad_norm": 0.5962051749229431, "learning_rate": 9.940716035130596e-05, "loss": 0.0918, "step": 7730 }, { "grad_norm": 0.7081824541091919, "learning_rate": 9.940398281120821e-05, "loss": 0.091, "step": 7740 }, { "grad_norm": 0.7382912039756775, "learning_rate": 9.940079682932102e-05, "loss": 0.0809, "step": 7750 }, { "grad_norm": 0.7033389806747437, "learning_rate": 9.939760240618877e-05, "loss": 0.0835, "step": 7760 }, { "grad_norm": 0.7876982092857361, "learning_rate": 9.939439954235729e-05, "loss": 0.0994, "step": 7770 }, { "grad_norm": 0.7481704354286194, "learning_rate": 9.939118823837387e-05, "loss": 0.1056, "step": 7780 }, { "grad_norm": 0.9962781071662903, "learning_rate": 9.938796849478725e-05, "loss": 0.1043, "step": 7790 }, { "grad_norm": 0.723316490650177, "learning_rate": 9.938474031214755e-05, "loss": 0.1223, "step": 7800 }, { "grad_norm": 0.5671963095664978, "learning_rate": 9.938150369100643e-05, "loss": 0.0959, "step": 7810 }, { "grad_norm": 0.8839859962463379, "learning_rate": 9.93782586319169e-05, "loss": 0.0887, "step": 7820 }, { "grad_norm": 0.9103600382804871, "learning_rate": 9.937500513543348e-05, "loss": 0.0998, "step": 7830 }, { "grad_norm": 0.9751379489898682, "learning_rate": 9.937174320211207e-05, "loss": 0.124, "step": 7840 }, { "grad_norm": 0.8963457345962524, "learning_rate": 9.936847283251009e-05, "loss": 0.0998, "step": 7850 }, { "grad_norm": 0.6747804880142212, "learning_rate": 9.936519402718632e-05, "loss": 0.0966, "step": 7860 }, { "grad_norm": 0.9338706731796265, "learning_rate": 9.936190678670102e-05, "loss": 0.114, "step": 7870 }, { "grad_norm": 0.8322160243988037, "learning_rate": 9.935861111161593e-05, "loss": 0.1133, "step": 7880 }, { "grad_norm": 0.7005863189697266, "learning_rate": 9.935530700249416e-05, "loss": 0.1195, "step": 7890 }, { "grad_norm": 0.5847458243370056, "learning_rate": 9.935199445990028e-05, "loss": 0.0869, "step": 7900 }, { "grad_norm": 0.6225590705871582, "learning_rate": 9.934867348440033e-05, "loss": 0.1211, "step": 7910 }, { "grad_norm": 0.9123126864433289, "learning_rate": 9.934534407656176e-05, "loss": 0.0971, "step": 7920 }, { "grad_norm": 0.5939695835113525, "learning_rate": 9.93420062369535e-05, "loss": 0.0972, "step": 7930 }, { "grad_norm": 0.7382634282112122, "learning_rate": 9.933865996614589e-05, "loss": 0.1059, "step": 7940 }, { "grad_norm": 0.7648507356643677, "learning_rate": 9.933530526471068e-05, "loss": 0.1068, "step": 7950 }, { "grad_norm": 0.6637170314788818, "learning_rate": 9.933194213322114e-05, "loss": 0.094, "step": 7960 }, { "grad_norm": 0.5737982988357544, "learning_rate": 9.932857057225192e-05, "loss": 0.1068, "step": 7970 }, { "grad_norm": 0.9929165840148926, "learning_rate": 9.932519058237912e-05, "loss": 0.102, "step": 7980 }, { "grad_norm": 0.7136533260345459, "learning_rate": 9.932180216418032e-05, "loss": 0.0985, "step": 7990 }, { "grad_norm": 0.955701470375061, "learning_rate": 9.931840531823446e-05, "loss": 0.1105, "step": 8000 }, { "grad_norm": 0.654516339302063, "learning_rate": 9.9315000045122e-05, "loss": 0.0922, "step": 8010 }, { "grad_norm": 0.6714746356010437, "learning_rate": 9.931158634542481e-05, "loss": 0.1068, "step": 8020 }, { "grad_norm": 1.2192192077636719, "learning_rate": 9.930816421972617e-05, "loss": 0.103, "step": 8030 }, { "grad_norm": 0.5825273394584656, "learning_rate": 9.930473366861086e-05, "loss": 0.0766, "step": 8040 }, { "grad_norm": 0.9880362749099731, "learning_rate": 9.930129469266505e-05, "loss": 0.085, "step": 8050 }, { "grad_norm": 1.0032705068588257, "learning_rate": 9.929784729247638e-05, "loss": 0.1285, "step": 8060 }, { "grad_norm": 0.7178476452827454, "learning_rate": 9.929439146863389e-05, "loss": 0.0805, "step": 8070 }, { "grad_norm": 0.6700259447097778, "learning_rate": 9.92909272217281e-05, "loss": 0.1052, "step": 8080 }, { "grad_norm": 0.7232646346092224, "learning_rate": 9.928745455235097e-05, "loss": 0.1055, "step": 8090 }, { "grad_norm": 0.7158315777778625, "learning_rate": 9.928397346109588e-05, "loss": 0.1017, "step": 8100 }, { "grad_norm": 0.9042391777038574, "learning_rate": 9.928048394855762e-05, "loss": 0.1026, "step": 8110 }, { "grad_norm": 0.8404856324195862, "learning_rate": 9.92769860153325e-05, "loss": 0.0998, "step": 8120 }, { "grad_norm": 0.9068252444267273, "learning_rate": 9.927347966201819e-05, "loss": 0.0945, "step": 8130 }, { "grad_norm": 0.7767592668533325, "learning_rate": 9.926996488921383e-05, "loss": 0.0772, "step": 8140 }, { "grad_norm": 0.6338434815406799, "learning_rate": 9.926644169752001e-05, "loss": 0.0979, "step": 8150 }, { "grad_norm": 0.5157591104507446, "learning_rate": 9.926291008753875e-05, "loss": 0.1011, "step": 8160 }, { "grad_norm": 0.7151987552642822, "learning_rate": 9.92593700598735e-05, "loss": 0.0908, "step": 8170 }, { "grad_norm": 0.5046412944793701, "learning_rate": 9.925582161512915e-05, "loss": 0.0894, "step": 8180 }, { "grad_norm": 0.7494578957557678, "learning_rate": 9.925226475391205e-05, "loss": 0.0854, "step": 8190 }, { "grad_norm": 0.9613529443740845, "learning_rate": 9.924869947682993e-05, "loss": 0.0991, "step": 8200 }, { "grad_norm": 0.7585273385047913, "learning_rate": 9.924512578449204e-05, "loss": 0.0795, "step": 8210 }, { "grad_norm": 0.6874308586120605, "learning_rate": 9.924154367750901e-05, "loss": 0.1218, "step": 8220 }, { "grad_norm": 0.6890987753868103, "learning_rate": 9.923795315649293e-05, "loss": 0.1118, "step": 8230 }, { "grad_norm": 0.9691824316978455, "learning_rate": 9.92343542220573e-05, "loss": 0.1123, "step": 8240 }, { "grad_norm": 0.714965283870697, "learning_rate": 9.92307468748171e-05, "loss": 0.0966, "step": 8250 }, { "grad_norm": 1.096637487411499, "learning_rate": 9.922713111538873e-05, "loss": 0.1092, "step": 8260 }, { "grad_norm": 0.897088885307312, "learning_rate": 9.922350694439003e-05, "loss": 0.1169, "step": 8270 }, { "grad_norm": 0.6662712097167969, "learning_rate": 9.921987436244024e-05, "loss": 0.0902, "step": 8280 }, { "grad_norm": 0.651836633682251, "learning_rate": 9.921623337016008e-05, "loss": 0.1007, "step": 8290 }, { "grad_norm": 1.3325998783111572, "learning_rate": 9.921258396817172e-05, "loss": 0.1211, "step": 8300 }, { "grad_norm": 1.0808651447296143, "learning_rate": 9.920892615709874e-05, "loss": 0.1157, "step": 8310 }, { "grad_norm": 0.8673291206359863, "learning_rate": 9.920525993756612e-05, "loss": 0.1083, "step": 8320 }, { "grad_norm": 0.8227693438529968, "learning_rate": 9.920158531020036e-05, "loss": 0.1029, "step": 8330 }, { "grad_norm": 0.8529508709907532, "learning_rate": 9.919790227562933e-05, "loss": 0.0909, "step": 8340 }, { "grad_norm": 0.8856534361839294, "learning_rate": 9.919421083448237e-05, "loss": 0.1001, "step": 8350 }, { "grad_norm": 0.6411865949630737, "learning_rate": 9.919051098739022e-05, "loss": 0.0906, "step": 8360 }, { "grad_norm": 0.7953161597251892, "learning_rate": 9.918680273498514e-05, "loss": 0.1116, "step": 8370 }, { "grad_norm": 0.6292158961296082, "learning_rate": 9.918308607790072e-05, "loss": 0.0839, "step": 8380 }, { "grad_norm": 0.9518425464630127, "learning_rate": 9.917936101677205e-05, "loss": 0.089, "step": 8390 }, { "grad_norm": 0.6089425683021545, "learning_rate": 9.917562755223564e-05, "loss": 0.1074, "step": 8400 }, { "grad_norm": 0.5328935384750366, "learning_rate": 9.917188568492944e-05, "loss": 0.1122, "step": 8410 }, { "grad_norm": 0.7316062450408936, "learning_rate": 9.916813541549283e-05, "loss": 0.1121, "step": 8420 }, { "grad_norm": 0.9595744013786316, "learning_rate": 9.916437674456663e-05, "loss": 0.1195, "step": 8430 }, { "grad_norm": 0.801213264465332, "learning_rate": 9.916060967279308e-05, "loss": 0.1175, "step": 8440 }, { "grad_norm": 0.7613077759742737, "learning_rate": 9.91568342008159e-05, "loss": 0.1184, "step": 8450 }, { "grad_norm": 0.6945583820343018, "learning_rate": 9.915305032928019e-05, "loss": 0.13, "step": 8460 }, { "grad_norm": 0.788874089717865, "learning_rate": 9.914925805883253e-05, "loss": 0.1163, "step": 8470 }, { "grad_norm": 0.8268218636512756, "learning_rate": 9.914545739012088e-05, "loss": 0.1283, "step": 8480 }, { "grad_norm": 1.0315229892730713, "learning_rate": 9.91416483237947e-05, "loss": 0.1161, "step": 8490 }, { "grad_norm": 0.8155338168144226, "learning_rate": 9.913783086050485e-05, "loss": 0.1032, "step": 8500 }, { "grad_norm": 0.6779794692993164, "learning_rate": 9.913400500090364e-05, "loss": 0.108, "step": 8510 }, { "grad_norm": 0.9715902209281921, "learning_rate": 9.913017074564479e-05, "loss": 0.1131, "step": 8520 }, { "grad_norm": 0.7555238604545593, "learning_rate": 9.912632809538348e-05, "loss": 0.104, "step": 8530 }, { "grad_norm": 0.8795299530029297, "learning_rate": 9.912247705077629e-05, "loss": 0.0945, "step": 8540 }, { "grad_norm": 0.8100586533546448, "learning_rate": 9.911861761248127e-05, "loss": 0.0894, "step": 8550 }, { "grad_norm": 0.8075710535049438, "learning_rate": 9.91147497811579e-05, "loss": 0.0873, "step": 8560 }, { "grad_norm": 0.5781806111335754, "learning_rate": 9.911087355746709e-05, "loss": 0.1034, "step": 8570 }, { "grad_norm": 0.8852113485336304, "learning_rate": 9.910698894207117e-05, "loss": 0.0969, "step": 8580 }, { "grad_norm": 0.9935039281845093, "learning_rate": 9.910309593563392e-05, "loss": 0.1079, "step": 8590 }, { "grad_norm": 0.7119124531745911, "learning_rate": 9.909919453882057e-05, "loss": 0.0787, "step": 8600 }, { "grad_norm": 0.8139440417289734, "learning_rate": 9.90952847522977e-05, "loss": 0.0915, "step": 8610 }, { "grad_norm": 0.7322950959205627, "learning_rate": 9.909136657673346e-05, "loss": 0.1107, "step": 8620 }, { "grad_norm": 1.2230387926101685, "learning_rate": 9.908744001279731e-05, "loss": 0.1584, "step": 8630 }, { "grad_norm": 0.8736093044281006, "learning_rate": 9.90835050611602e-05, "loss": 0.0958, "step": 8640 }, { "grad_norm": 1.0110701322555542, "learning_rate": 9.90795617224945e-05, "loss": 0.097, "step": 8650 }, { "grad_norm": 0.8113943934440613, "learning_rate": 9.907560999747405e-05, "loss": 0.109, "step": 8660 }, { "grad_norm": 0.9234402179718018, "learning_rate": 9.907164988677408e-05, "loss": 0.0968, "step": 8670 }, { "grad_norm": 0.6547988057136536, "learning_rate": 9.906768139107124e-05, "loss": 0.0935, "step": 8680 }, { "grad_norm": 0.93763267993927, "learning_rate": 9.906370451104367e-05, "loss": 0.0906, "step": 8690 }, { "grad_norm": 0.5360924601554871, "learning_rate": 9.905971924737088e-05, "loss": 0.0748, "step": 8700 }, { "grad_norm": 0.6310835480690002, "learning_rate": 9.905572560073387e-05, "loss": 0.0971, "step": 8710 }, { "grad_norm": 0.7238646149635315, "learning_rate": 9.905172357181501e-05, "loss": 0.0902, "step": 8720 }, { "grad_norm": 0.9341732859611511, "learning_rate": 9.904771316129817e-05, "loss": 0.1041, "step": 8730 }, { "grad_norm": 0.823759138584137, "learning_rate": 9.904369436986862e-05, "loss": 0.0962, "step": 8740 }, { "grad_norm": 0.6461957097053528, "learning_rate": 9.903966719821303e-05, "loss": 0.0998, "step": 8750 }, { "grad_norm": 0.6497364640235901, "learning_rate": 9.903563164701956e-05, "loss": 0.0982, "step": 8760 }, { "grad_norm": 0.7201682925224304, "learning_rate": 9.903158771697778e-05, "loss": 0.1092, "step": 8770 }, { "grad_norm": 0.618614673614502, "learning_rate": 9.902753540877867e-05, "loss": 0.0982, "step": 8780 }, { "grad_norm": 0.7185437083244324, "learning_rate": 9.902347472311466e-05, "loss": 0.1219, "step": 8790 }, { "grad_norm": 0.6354465484619141, "learning_rate": 9.901940566067962e-05, "loss": 0.0905, "step": 8800 }, { "grad_norm": 0.6761928200721741, "learning_rate": 9.901532822216883e-05, "loss": 0.1126, "step": 8810 }, { "grad_norm": 0.6869699954986572, "learning_rate": 9.901124240827904e-05, "loss": 0.0887, "step": 8820 }, { "grad_norm": 0.9761304259300232, "learning_rate": 9.900714821970835e-05, "loss": 0.1208, "step": 8830 }, { "grad_norm": 0.7682822942733765, "learning_rate": 9.900304565715641e-05, "loss": 0.1069, "step": 8840 }, { "grad_norm": 0.5795685648918152, "learning_rate": 9.899893472132419e-05, "loss": 0.134, "step": 8850 }, { "grad_norm": 0.7858710289001465, "learning_rate": 9.899481541291415e-05, "loss": 0.0826, "step": 8860 }, { "grad_norm": 0.8336268067359924, "learning_rate": 9.899068773263016e-05, "loss": 0.0989, "step": 8870 }, { "grad_norm": 0.7345479130744934, "learning_rate": 9.898655168117754e-05, "loss": 0.1111, "step": 8880 }, { "grad_norm": 0.8231030702590942, "learning_rate": 9.898240725926302e-05, "loss": 0.1062, "step": 8890 }, { "grad_norm": 0.700225293636322, "learning_rate": 9.897825446759478e-05, "loss": 0.1186, "step": 8900 }, { "grad_norm": 1.4014745950698853, "learning_rate": 9.897409330688241e-05, "loss": 0.1066, "step": 8910 }, { "grad_norm": 0.7924684286117554, "learning_rate": 9.896992377783692e-05, "loss": 0.086, "step": 8920 }, { "grad_norm": 1.0422216653823853, "learning_rate": 9.89657458811708e-05, "loss": 0.1084, "step": 8930 }, { "grad_norm": 0.5873475074768066, "learning_rate": 9.896155961759792e-05, "loss": 0.0848, "step": 8940 }, { "grad_norm": 0.8307721614837646, "learning_rate": 9.895736498783361e-05, "loss": 0.0888, "step": 8950 }, { "grad_norm": 0.657787024974823, "learning_rate": 9.895316199259462e-05, "loss": 0.1038, "step": 8960 }, { "grad_norm": 0.6527426838874817, "learning_rate": 9.894895063259909e-05, "loss": 0.106, "step": 8970 }, { "grad_norm": 0.766675591468811, "learning_rate": 9.894473090856667e-05, "loss": 0.0881, "step": 8980 }, { "grad_norm": 0.8267131447792053, "learning_rate": 9.894050282121839e-05, "loss": 0.0983, "step": 8990 }, { "grad_norm": 0.8718323707580566, "learning_rate": 9.893626637127668e-05, "loss": 0.0996, "step": 9000 }, { "grad_norm": 0.8951252698898315, "learning_rate": 9.893202155946546e-05, "loss": 0.119, "step": 9010 }, { "grad_norm": 0.9010816812515259, "learning_rate": 9.892776838651006e-05, "loss": 0.1331, "step": 9020 }, { "grad_norm": 0.5317345261573792, "learning_rate": 9.892350685313722e-05, "loss": 0.0998, "step": 9030 }, { "grad_norm": 0.7378793358802795, "learning_rate": 9.891923696007513e-05, "loss": 0.1063, "step": 9040 }, { "grad_norm": 0.5833942890167236, "learning_rate": 9.891495870805336e-05, "loss": 0.0845, "step": 9050 }, { "grad_norm": 0.6494569182395935, "learning_rate": 9.891067209780298e-05, "loss": 0.1071, "step": 9060 }, { "grad_norm": 0.8134316205978394, "learning_rate": 9.890637713005646e-05, "loss": 0.1152, "step": 9070 }, { "grad_norm": 1.2516403198242188, "learning_rate": 9.890207380554767e-05, "loss": 0.1223, "step": 9080 }, { "grad_norm": 0.7558821439743042, "learning_rate": 9.889776212501196e-05, "loss": 0.1335, "step": 9090 }, { "grad_norm": 0.8313446044921875, "learning_rate": 9.889344208918605e-05, "loss": 0.0861, "step": 9100 }, { "grad_norm": 0.6323613524436951, "learning_rate": 9.888911369880812e-05, "loss": 0.0894, "step": 9110 }, { "grad_norm": 0.66416335105896, "learning_rate": 9.888477695461777e-05, "loss": 0.0993, "step": 9120 }, { "grad_norm": 0.6668763160705566, "learning_rate": 9.888043185735607e-05, "loss": 0.091, "step": 9130 }, { "grad_norm": 0.6472005844116211, "learning_rate": 9.887607840776542e-05, "loss": 0.1171, "step": 9140 }, { "grad_norm": 0.6034380793571472, "learning_rate": 9.887171660658975e-05, "loss": 0.0936, "step": 9150 }, { "grad_norm": 0.5815481543540955, "learning_rate": 9.886734645457435e-05, "loss": 0.1109, "step": 9160 }, { "grad_norm": 0.7708085179328918, "learning_rate": 9.886296795246597e-05, "loss": 0.1003, "step": 9170 }, { "grad_norm": 0.720365047454834, "learning_rate": 9.885858110101276e-05, "loss": 0.1053, "step": 9180 }, { "grad_norm": 0.8800206780433655, "learning_rate": 9.885418590096434e-05, "loss": 0.1068, "step": 9190 }, { "grad_norm": 0.5535581707954407, "learning_rate": 9.88497823530717e-05, "loss": 0.08, "step": 9200 }, { "grad_norm": 0.8560645580291748, "learning_rate": 9.884537045808732e-05, "loss": 0.0877, "step": 9210 }, { "grad_norm": 0.7484921813011169, "learning_rate": 9.884095021676502e-05, "loss": 0.1017, "step": 9220 }, { "grad_norm": 0.9301832318305969, "learning_rate": 9.883652162986017e-05, "loss": 0.0756, "step": 9230 }, { "grad_norm": 0.8669734597206116, "learning_rate": 9.883208469812943e-05, "loss": 0.0994, "step": 9240 }, { "grad_norm": 0.7217033505439758, "learning_rate": 9.882763942233098e-05, "loss": 0.0997, "step": 9250 }, { "grad_norm": 0.842746376991272, "learning_rate": 9.882318580322441e-05, "loss": 0.1228, "step": 9260 }, { "grad_norm": 0.5748803019523621, "learning_rate": 9.881872384157067e-05, "loss": 0.1082, "step": 9270 }, { "grad_norm": 0.9436133503913879, "learning_rate": 9.881425353813225e-05, "loss": 0.1206, "step": 9280 }, { "grad_norm": 0.8794638514518738, "learning_rate": 9.880977489367296e-05, "loss": 0.0946, "step": 9290 }, { "grad_norm": 0.8157941102981567, "learning_rate": 9.88052879089581e-05, "loss": 0.0873, "step": 9300 }, { "grad_norm": 0.7465366721153259, "learning_rate": 9.880079258475434e-05, "loss": 0.1074, "step": 9310 }, { "grad_norm": 0.7770203948020935, "learning_rate": 9.879628892182985e-05, "loss": 0.0981, "step": 9320 }, { "grad_norm": 0.7571409940719604, "learning_rate": 9.879177692095416e-05, "loss": 0.0832, "step": 9330 }, { "grad_norm": 1.0715452432632446, "learning_rate": 9.878725658289825e-05, "loss": 0.1037, "step": 9340 }, { "grad_norm": 0.8433256149291992, "learning_rate": 9.878272790843454e-05, "loss": 0.1027, "step": 9350 }, { "grad_norm": 0.6688000559806824, "learning_rate": 9.877819089833682e-05, "loss": 0.096, "step": 9360 }, { "grad_norm": 0.7323598861694336, "learning_rate": 9.877364555338038e-05, "loss": 0.1016, "step": 9370 }, { "grad_norm": 0.7176769375801086, "learning_rate": 9.876909187434186e-05, "loss": 0.1083, "step": 9380 }, { "grad_norm": 0.5774273872375488, "learning_rate": 9.876452986199939e-05, "loss": 0.1038, "step": 9390 }, { "grad_norm": 0.7579851746559143, "learning_rate": 9.875995951713248e-05, "loss": 0.0981, "step": 9400 }, { "grad_norm": 0.6723992824554443, "learning_rate": 9.875538084052207e-05, "loss": 0.0821, "step": 9410 }, { "grad_norm": 0.8092232942581177, "learning_rate": 9.875079383295053e-05, "loss": 0.104, "step": 9420 }, { "grad_norm": 0.6855385899543762, "learning_rate": 9.874619849520167e-05, "loss": 0.0854, "step": 9430 }, { "grad_norm": 0.8365980982780457, "learning_rate": 9.874159482806069e-05, "loss": 0.0977, "step": 9440 }, { "grad_norm": 0.7880380749702454, "learning_rate": 9.873698283231426e-05, "loss": 0.099, "step": 9450 }, { "grad_norm": 0.750499963760376, "learning_rate": 9.87323625087504e-05, "loss": 0.1278, "step": 9460 }, { "grad_norm": 0.8302763104438782, "learning_rate": 9.872773385815863e-05, "loss": 0.1237, "step": 9470 }, { "grad_norm": 0.5601010322570801, "learning_rate": 9.872309688132986e-05, "loss": 0.0834, "step": 9480 }, { "grad_norm": 0.940015435218811, "learning_rate": 9.871845157905639e-05, "loss": 0.0869, "step": 9490 }, { "grad_norm": 0.5881229043006897, "learning_rate": 9.871379795213201e-05, "loss": 0.0881, "step": 9500 }, { "grad_norm": 0.6432750821113586, "learning_rate": 9.87091360013519e-05, "loss": 0.0954, "step": 9510 }, { "grad_norm": 0.7410518527030945, "learning_rate": 9.870446572751262e-05, "loss": 0.1055, "step": 9520 }, { "grad_norm": 0.8095437288284302, "learning_rate": 9.869978713141224e-05, "loss": 0.1059, "step": 9530 }, { "grad_norm": 0.9179821610450745, "learning_rate": 9.869510021385016e-05, "loss": 0.1063, "step": 9540 }, { "grad_norm": 0.8533592224121094, "learning_rate": 9.869040497562727e-05, "loss": 0.1238, "step": 9550 }, { "grad_norm": 1.4444098472595215, "learning_rate": 9.868570141754587e-05, "loss": 0.1104, "step": 9560 }, { "grad_norm": 0.942704975605011, "learning_rate": 9.868098954040965e-05, "loss": 0.0934, "step": 9570 }, { "grad_norm": 0.5814160108566284, "learning_rate": 9.867626934502374e-05, "loss": 0.0953, "step": 9580 }, { "grad_norm": 0.9292078018188477, "learning_rate": 9.86715408321947e-05, "loss": 0.1022, "step": 9590 }, { "grad_norm": 0.738471269607544, "learning_rate": 9.86668040027305e-05, "loss": 0.0979, "step": 9600 }, { "grad_norm": 1.2479026317596436, "learning_rate": 9.866205885744053e-05, "loss": 0.0869, "step": 9610 }, { "grad_norm": 0.6442228555679321, "learning_rate": 9.865730539713563e-05, "loss": 0.1083, "step": 9620 }, { "grad_norm": 0.807628333568573, "learning_rate": 9.8652543622628e-05, "loss": 0.0918, "step": 9630 }, { "grad_norm": 0.9105769991874695, "learning_rate": 9.864777353473132e-05, "loss": 0.1267, "step": 9640 }, { "grad_norm": 0.68597811460495, "learning_rate": 9.864299513426068e-05, "loss": 0.1092, "step": 9650 }, { "grad_norm": 0.6838594079017639, "learning_rate": 9.863820842203254e-05, "loss": 0.0947, "step": 9660 }, { "grad_norm": 0.6495104432106018, "learning_rate": 9.863341339886483e-05, "loss": 0.1059, "step": 9670 }, { "grad_norm": 0.7303248047828674, "learning_rate": 9.86286100655769e-05, "loss": 0.099, "step": 9680 }, { "grad_norm": 0.592370867729187, "learning_rate": 9.862379842298953e-05, "loss": 0.1002, "step": 9690 }, { "grad_norm": 0.7901086807250977, "learning_rate": 9.861897847192485e-05, "loss": 0.0923, "step": 9700 }, { "grad_norm": 0.7598690390586853, "learning_rate": 9.86141502132065e-05, "loss": 0.0926, "step": 9710 }, { "grad_norm": 0.7846294641494751, "learning_rate": 9.860931364765946e-05, "loss": 0.0768, "step": 9720 }, { "grad_norm": 0.743218183517456, "learning_rate": 9.860446877611021e-05, "loss": 0.099, "step": 9730 }, { "grad_norm": 0.6770838499069214, "learning_rate": 9.859961559938655e-05, "loss": 0.0833, "step": 9740 }, { "grad_norm": 0.5930910706520081, "learning_rate": 9.85947541183178e-05, "loss": 0.0793, "step": 9750 }, { "grad_norm": 0.6023252010345459, "learning_rate": 9.858988433373463e-05, "loss": 0.0857, "step": 9760 }, { "grad_norm": 0.7443393468856812, "learning_rate": 9.858500624646918e-05, "loss": 0.0987, "step": 9770 }, { "grad_norm": 0.7207812666893005, "learning_rate": 9.858011985735497e-05, "loss": 0.124, "step": 9780 }, { "grad_norm": 0.5970950126647949, "learning_rate": 9.857522516722693e-05, "loss": 0.0925, "step": 9790 }, { "grad_norm": 0.8331255316734314, "learning_rate": 9.857032217692145e-05, "loss": 0.0773, "step": 9800 }, { "grad_norm": 0.6868201494216919, "learning_rate": 9.856541088727631e-05, "loss": 0.0761, "step": 9810 }, { "grad_norm": 0.7437664866447449, "learning_rate": 9.856049129913072e-05, "loss": 0.1076, "step": 9820 }, { "grad_norm": 1.0969713926315308, "learning_rate": 9.85555634133253e-05, "loss": 0.1276, "step": 9830 }, { "grad_norm": 0.7216771245002747, "learning_rate": 9.855062723070208e-05, "loss": 0.0919, "step": 9840 }, { "grad_norm": 0.8692975640296936, "learning_rate": 9.854568275210454e-05, "loss": 0.0988, "step": 9850 }, { "grad_norm": 0.5136711001396179, "learning_rate": 9.854072997837754e-05, "loss": 0.0826, "step": 9860 }, { "grad_norm": 0.9132865071296692, "learning_rate": 9.853576891036737e-05, "loss": 0.0757, "step": 9870 }, { "grad_norm": 0.9006978869438171, "learning_rate": 9.853079954892177e-05, "loss": 0.0795, "step": 9880 }, { "grad_norm": 0.7223334908485413, "learning_rate": 9.852582189488983e-05, "loss": 0.1018, "step": 9890 }, { "grad_norm": 0.5643168091773987, "learning_rate": 9.852083594912212e-05, "loss": 0.0826, "step": 9900 }, { "grad_norm": 0.8983790278434753, "learning_rate": 9.851584171247058e-05, "loss": 0.0984, "step": 9910 }, { "grad_norm": 0.5854384303092957, "learning_rate": 9.851083918578863e-05, "loss": 0.0902, "step": 9920 }, { "grad_norm": 0.6590430736541748, "learning_rate": 9.850582836993103e-05, "loss": 0.1147, "step": 9930 }, { "grad_norm": 0.7133288383483887, "learning_rate": 9.850080926575397e-05, "loss": 0.0884, "step": 9940 }, { "grad_norm": 0.6469534635543823, "learning_rate": 9.849578187411515e-05, "loss": 0.0998, "step": 9950 }, { "grad_norm": 0.7724837064743042, "learning_rate": 9.849074619587354e-05, "loss": 0.0974, "step": 9960 }, { "grad_norm": 0.7158581018447876, "learning_rate": 9.848570223188964e-05, "loss": 0.0994, "step": 9970 }, { "grad_norm": 0.7244992852210999, "learning_rate": 9.848064998302531e-05, "loss": 0.0916, "step": 9980 }, { "grad_norm": 0.8042113184928894, "learning_rate": 9.847558945014386e-05, "loss": 0.0873, "step": 9990 }, { "grad_norm": 0.5735450387001038, "learning_rate": 9.847052063410996e-05, "loss": 0.1166, "step": 10000 }, { "grad_norm": 1.0010111331939697, "learning_rate": 9.846544353578977e-05, "loss": 0.1058, "step": 10010 }, { "grad_norm": 0.8871595859527588, "learning_rate": 9.846035815605081e-05, "loss": 0.086, "step": 10020 }, { "grad_norm": 0.7393261194229126, "learning_rate": 9.845526449576204e-05, "loss": 0.0918, "step": 10030 }, { "grad_norm": 0.700742244720459, "learning_rate": 9.845016255579383e-05, "loss": 0.1025, "step": 10040 }, { "grad_norm": 0.6965211629867554, "learning_rate": 9.844505233701794e-05, "loss": 0.0945, "step": 10050 }, { "grad_norm": 0.6955482959747314, "learning_rate": 9.843993384030757e-05, "loss": 0.0811, "step": 10060 }, { "grad_norm": 0.7969945669174194, "learning_rate": 9.843480706653737e-05, "loss": 0.0797, "step": 10070 }, { "grad_norm": 0.6954423785209656, "learning_rate": 9.84296720165833e-05, "loss": 0.0926, "step": 10080 }, { "grad_norm": 0.5779935121536255, "learning_rate": 9.842452869132286e-05, "loss": 0.101, "step": 10090 }, { "grad_norm": 0.7609109878540039, "learning_rate": 9.841937709163489e-05, "loss": 0.0869, "step": 10100 }, { "grad_norm": 0.6712133288383484, "learning_rate": 9.841421721839962e-05, "loss": 0.0805, "step": 10110 }, { "grad_norm": 0.7561760544776917, "learning_rate": 9.840904907249879e-05, "loss": 0.0774, "step": 10120 }, { "grad_norm": 0.6563032269477844, "learning_rate": 9.840387265481545e-05, "loss": 0.1185, "step": 10130 }, { "grad_norm": 0.6898346543312073, "learning_rate": 9.839868796623411e-05, "loss": 0.1036, "step": 10140 }, { "grad_norm": 0.7847067713737488, "learning_rate": 9.839349500764072e-05, "loss": 0.1233, "step": 10150 }, { "grad_norm": 0.6048557758331299, "learning_rate": 9.83882937799226e-05, "loss": 0.0998, "step": 10160 }, { "grad_norm": 0.9034870862960815, "learning_rate": 9.838308428396849e-05, "loss": 0.0807, "step": 10170 }, { "grad_norm": 0.5826011896133423, "learning_rate": 9.837786652066854e-05, "loss": 0.0981, "step": 10180 }, { "grad_norm": 0.690026044845581, "learning_rate": 9.837264049091437e-05, "loss": 0.0809, "step": 10190 }, { "grad_norm": 0.7200329303741455, "learning_rate": 9.836740619559893e-05, "loss": 0.091, "step": 10200 }, { "grad_norm": 0.8169224262237549, "learning_rate": 9.836216363561659e-05, "loss": 0.0821, "step": 10210 }, { "grad_norm": 0.7252044081687927, "learning_rate": 9.835691281186322e-05, "loss": 0.1095, "step": 10220 }, { "grad_norm": 0.7920038104057312, "learning_rate": 9.8351653725236e-05, "loss": 0.1153, "step": 10230 }, { "grad_norm": 0.7298797369003296, "learning_rate": 9.83463863766336e-05, "loss": 0.1237, "step": 10240 }, { "grad_norm": 0.7437223196029663, "learning_rate": 9.834111076695602e-05, "loss": 0.0958, "step": 10250 }, { "grad_norm": 0.7686145901679993, "learning_rate": 9.833582689710477e-05, "loss": 0.114, "step": 10260 }, { "grad_norm": 0.8338329195976257, "learning_rate": 9.833053476798268e-05, "loss": 0.0966, "step": 10270 }, { "grad_norm": 0.8630142211914062, "learning_rate": 9.832523438049404e-05, "loss": 0.0914, "step": 10280 }, { "grad_norm": 0.7774743437767029, "learning_rate": 9.831992573554454e-05, "loss": 0.0857, "step": 10290 }, { "grad_norm": 0.6753043532371521, "learning_rate": 9.831460883404128e-05, "loss": 0.0772, "step": 10300 }, { "grad_norm": 0.553542971611023, "learning_rate": 9.830928367689278e-05, "loss": 0.108, "step": 10310 }, { "grad_norm": 1.0435065031051636, "learning_rate": 9.830395026500896e-05, "loss": 0.1066, "step": 10320 }, { "grad_norm": 1.5101979970932007, "learning_rate": 9.829860859930115e-05, "loss": 0.1009, "step": 10330 }, { "grad_norm": 0.68012535572052, "learning_rate": 9.829325868068212e-05, "loss": 0.078, "step": 10340 }, { "grad_norm": 0.8037172555923462, "learning_rate": 9.8287900510066e-05, "loss": 0.083, "step": 10350 }, { "grad_norm": 0.697374701499939, "learning_rate": 9.828253408836834e-05, "loss": 0.0866, "step": 10360 }, { "grad_norm": 0.7213460803031921, "learning_rate": 9.827715941650615e-05, "loss": 0.1138, "step": 10370 }, { "grad_norm": 0.6065685153007507, "learning_rate": 9.82717764953978e-05, "loss": 0.1153, "step": 10380 }, { "grad_norm": 0.4628986418247223, "learning_rate": 9.826638532596308e-05, "loss": 0.0742, "step": 10390 }, { "grad_norm": 0.8130937218666077, "learning_rate": 9.82609859091232e-05, "loss": 0.0816, "step": 10400 }, { "grad_norm": 0.8737242221832275, "learning_rate": 9.825557824580076e-05, "loss": 0.0996, "step": 10410 }, { "grad_norm": 1.2011138200759888, "learning_rate": 9.82501623369198e-05, "loss": 0.1101, "step": 10420 }, { "grad_norm": 0.746999979019165, "learning_rate": 9.824473818340574e-05, "loss": 0.0896, "step": 10430 }, { "grad_norm": 0.7675411701202393, "learning_rate": 9.823930578618541e-05, "loss": 0.0879, "step": 10440 }, { "grad_norm": 0.8068604469299316, "learning_rate": 9.823386514618709e-05, "loss": 0.0924, "step": 10450 }, { "grad_norm": 0.6206934452056885, "learning_rate": 9.82284162643404e-05, "loss": 0.0824, "step": 10460 }, { "grad_norm": 1.0101569890975952, "learning_rate": 9.822295914157642e-05, "loss": 0.107, "step": 10470 }, { "grad_norm": 0.5467096567153931, "learning_rate": 9.821749377882763e-05, "loss": 0.0785, "step": 10480 }, { "grad_norm": 0.7053198218345642, "learning_rate": 9.821202017702791e-05, "loss": 0.1114, "step": 10490 }, { "grad_norm": 0.6924009323120117, "learning_rate": 9.820653833711253e-05, "loss": 0.0987, "step": 10500 }, { "grad_norm": 0.6686822772026062, "learning_rate": 9.820104826001822e-05, "loss": 0.0713, "step": 10510 }, { "grad_norm": 0.8701416850090027, "learning_rate": 9.819554994668305e-05, "loss": 0.102, "step": 10520 }, { "grad_norm": 0.7442052364349365, "learning_rate": 9.819004339804654e-05, "loss": 0.1054, "step": 10530 }, { "grad_norm": 1.1236679553985596, "learning_rate": 9.818452861504961e-05, "loss": 0.102, "step": 10540 }, { "grad_norm": 0.8057431578636169, "learning_rate": 9.81790055986346e-05, "loss": 0.1043, "step": 10550 }, { "grad_norm": 0.7209040522575378, "learning_rate": 9.817347434974523e-05, "loss": 0.0998, "step": 10560 }, { "grad_norm": 0.6833778023719788, "learning_rate": 9.816793486932664e-05, "loss": 0.1151, "step": 10570 }, { "grad_norm": 0.6225932240486145, "learning_rate": 9.816238715832538e-05, "loss": 0.0753, "step": 10580 }, { "grad_norm": 0.7222287654876709, "learning_rate": 9.815683121768939e-05, "loss": 0.102, "step": 10590 }, { "grad_norm": 0.8431231379508972, "learning_rate": 9.815126704836804e-05, "loss": 0.0654, "step": 10600 }, { "grad_norm": 0.41451558470726013, "learning_rate": 9.81456946513121e-05, "loss": 0.0795, "step": 10610 }, { "grad_norm": 0.5140631794929504, "learning_rate": 9.814011402747373e-05, "loss": 0.0806, "step": 10620 }, { "grad_norm": 0.5701113939285278, "learning_rate": 9.813452517780651e-05, "loss": 0.0895, "step": 10630 }, { "grad_norm": 0.7437645792961121, "learning_rate": 9.81289281032654e-05, "loss": 0.0922, "step": 10640 }, { "grad_norm": 0.6371653079986572, "learning_rate": 9.812332280480683e-05, "loss": 0.1133, "step": 10650 }, { "grad_norm": 0.9589364528656006, "learning_rate": 9.811770928338854e-05, "loss": 0.0943, "step": 10660 }, { "grad_norm": 0.6022529602050781, "learning_rate": 9.811208753996979e-05, "loss": 0.081, "step": 10670 }, { "grad_norm": 0.7232438325881958, "learning_rate": 9.810645757551113e-05, "loss": 0.0913, "step": 10680 }, { "grad_norm": 0.9034275412559509, "learning_rate": 9.810081939097459e-05, "loss": 0.0753, "step": 10690 }, { "grad_norm": 0.6311159133911133, "learning_rate": 9.809517298732356e-05, "loss": 0.0777, "step": 10700 }, { "grad_norm": 0.57738196849823, "learning_rate": 9.80895183655229e-05, "loss": 0.086, "step": 10710 }, { "grad_norm": 0.6764604449272156, "learning_rate": 9.808385552653877e-05, "loss": 0.0952, "step": 10720 }, { "grad_norm": 0.49527427554130554, "learning_rate": 9.807818447133886e-05, "loss": 0.0672, "step": 10730 }, { "grad_norm": 0.6943163275718689, "learning_rate": 9.807250520089215e-05, "loss": 0.084, "step": 10740 }, { "grad_norm": 0.5265636444091797, "learning_rate": 9.806681771616908e-05, "loss": 0.0677, "step": 10750 }, { "grad_norm": 0.6452435255050659, "learning_rate": 9.80611220181415e-05, "loss": 0.0698, "step": 10760 }, { "grad_norm": 0.572521984577179, "learning_rate": 9.805541810778264e-05, "loss": 0.0758, "step": 10770 }, { "grad_norm": 0.6271967887878418, "learning_rate": 9.804970598606716e-05, "loss": 0.0723, "step": 10780 }, { "grad_norm": 0.5941989421844482, "learning_rate": 9.804398565397106e-05, "loss": 0.1021, "step": 10790 }, { "grad_norm": 0.6579424738883972, "learning_rate": 9.803825711247183e-05, "loss": 0.0812, "step": 10800 }, { "grad_norm": 0.9447391033172607, "learning_rate": 9.803252036254831e-05, "loss": 0.0752, "step": 10810 }, { "grad_norm": 0.8198410868644714, "learning_rate": 9.802677540518076e-05, "loss": 0.0878, "step": 10820 }, { "grad_norm": 0.7470554709434509, "learning_rate": 9.802102224135081e-05, "loss": 0.0911, "step": 10830 }, { "grad_norm": 0.8763012290000916, "learning_rate": 9.801526087204155e-05, "loss": 0.0881, "step": 10840 }, { "grad_norm": 0.8642567992210388, "learning_rate": 9.800949129823743e-05, "loss": 0.0885, "step": 10850 }, { "grad_norm": 0.7573578953742981, "learning_rate": 9.80037135209243e-05, "loss": 0.0818, "step": 10860 }, { "grad_norm": 0.6147834658622742, "learning_rate": 9.799792754108946e-05, "loss": 0.1228, "step": 10870 }, { "grad_norm": 0.8187229633331299, "learning_rate": 9.799213335972152e-05, "loss": 0.0891, "step": 10880 }, { "grad_norm": 0.6115527749061584, "learning_rate": 9.798633097781058e-05, "loss": 0.0678, "step": 10890 }, { "grad_norm": 0.7065221667289734, "learning_rate": 9.79805203963481e-05, "loss": 0.0766, "step": 10900 }, { "grad_norm": 0.707631528377533, "learning_rate": 9.797470161632697e-05, "loss": 0.0735, "step": 10910 }, { "grad_norm": 1.0936063528060913, "learning_rate": 9.796887463874145e-05, "loss": 0.0843, "step": 10920 }, { "grad_norm": 0.583078920841217, "learning_rate": 9.796303946458718e-05, "loss": 0.0822, "step": 10930 }, { "grad_norm": 0.5872857570648193, "learning_rate": 9.795719609486127e-05, "loss": 0.0692, "step": 10940 }, { "grad_norm": 0.5686891078948975, "learning_rate": 9.795134453056219e-05, "loss": 0.0816, "step": 10950 }, { "grad_norm": 0.8526414036750793, "learning_rate": 9.794548477268979e-05, "loss": 0.0824, "step": 10960 }, { "grad_norm": 0.8245622515678406, "learning_rate": 9.793961682224537e-05, "loss": 0.0933, "step": 10970 }, { "grad_norm": 0.8951776623725891, "learning_rate": 9.793374068023156e-05, "loss": 0.0785, "step": 10980 }, { "grad_norm": 0.859546422958374, "learning_rate": 9.792785634765247e-05, "loss": 0.0966, "step": 10990 }, { "grad_norm": 0.8614386320114136, "learning_rate": 9.792196382551357e-05, "loss": 0.0906, "step": 11000 }, { "grad_norm": 0.6358842849731445, "learning_rate": 9.791606311482171e-05, "loss": 0.0955, "step": 11010 }, { "grad_norm": 0.7770469784736633, "learning_rate": 9.791015421658518e-05, "loss": 0.0847, "step": 11020 }, { "grad_norm": 0.6195667386054993, "learning_rate": 9.790423713181362e-05, "loss": 0.0832, "step": 11030 }, { "grad_norm": 0.6106101274490356, "learning_rate": 9.789831186151814e-05, "loss": 0.0783, "step": 11040 }, { "grad_norm": 0.6650472283363342, "learning_rate": 9.789237840671118e-05, "loss": 0.0943, "step": 11050 }, { "grad_norm": 0.5968204140663147, "learning_rate": 9.78864367684066e-05, "loss": 0.088, "step": 11060 }, { "grad_norm": 0.5963338017463684, "learning_rate": 9.788048694761968e-05, "loss": 0.0857, "step": 11070 }, { "grad_norm": 0.668365478515625, "learning_rate": 9.787452894536709e-05, "loss": 0.0707, "step": 11080 }, { "grad_norm": 1.4163156747817993, "learning_rate": 9.786856276266685e-05, "loss": 0.0872, "step": 11090 }, { "grad_norm": 1.1504135131835938, "learning_rate": 9.786258840053845e-05, "loss": 0.0831, "step": 11100 }, { "grad_norm": 0.8593558073043823, "learning_rate": 9.785660586000273e-05, "loss": 0.0966, "step": 11110 }, { "grad_norm": 0.7720941305160522, "learning_rate": 9.785061514208196e-05, "loss": 0.0851, "step": 11120 }, { "grad_norm": 0.6845059990882874, "learning_rate": 9.784461624779977e-05, "loss": 0.102, "step": 11130 }, { "grad_norm": 0.8046678900718689, "learning_rate": 9.783860917818123e-05, "loss": 0.099, "step": 11140 }, { "grad_norm": 0.6349353194236755, "learning_rate": 9.783259393425277e-05, "loss": 0.0987, "step": 11150 }, { "grad_norm": 0.8058755993843079, "learning_rate": 9.782657051704221e-05, "loss": 0.1073, "step": 11160 }, { "grad_norm": 0.5778304934501648, "learning_rate": 9.782053892757883e-05, "loss": 0.0921, "step": 11170 }, { "grad_norm": 0.906697154045105, "learning_rate": 9.781449916689324e-05, "loss": 0.0877, "step": 11180 }, { "grad_norm": 0.9028002619743347, "learning_rate": 9.780845123601746e-05, "loss": 0.0876, "step": 11190 }, { "grad_norm": 0.7063085436820984, "learning_rate": 9.780239513598492e-05, "loss": 0.0996, "step": 11200 }, { "grad_norm": 1.2979815006256104, "learning_rate": 9.779633086783047e-05, "loss": 0.1032, "step": 11210 }, { "grad_norm": 0.5877777338027954, "learning_rate": 9.779025843259031e-05, "loss": 0.0814, "step": 11220 }, { "grad_norm": 0.6613340377807617, "learning_rate": 9.778417783130204e-05, "loss": 0.0786, "step": 11230 }, { "grad_norm": 1.4965674877166748, "learning_rate": 9.777808906500468e-05, "loss": 0.0969, "step": 11240 }, { "grad_norm": 1.005271553993225, "learning_rate": 9.777199213473862e-05, "loss": 0.0939, "step": 11250 }, { "grad_norm": 0.9298380017280579, "learning_rate": 9.77658870415457e-05, "loss": 0.0943, "step": 11260 }, { "grad_norm": 0.871463418006897, "learning_rate": 9.775977378646906e-05, "loss": 0.0998, "step": 11270 }, { "grad_norm": 0.7017075419425964, "learning_rate": 9.775365237055331e-05, "loss": 0.0901, "step": 11280 }, { "grad_norm": 0.9866881966590881, "learning_rate": 9.774752279484445e-05, "loss": 0.0841, "step": 11290 }, { "grad_norm": 0.751624584197998, "learning_rate": 9.774138506038984e-05, "loss": 0.1185, "step": 11300 }, { "grad_norm": 0.7434491515159607, "learning_rate": 9.773523916823826e-05, "loss": 0.0943, "step": 11310 }, { "grad_norm": 0.7424829602241516, "learning_rate": 9.772908511943986e-05, "loss": 0.089, "step": 11320 }, { "grad_norm": 1.0654139518737793, "learning_rate": 9.77229229150462e-05, "loss": 0.0834, "step": 11330 }, { "grad_norm": 1.1444793939590454, "learning_rate": 9.771675255611024e-05, "loss": 0.0787, "step": 11340 }, { "grad_norm": 0.7065312266349792, "learning_rate": 9.771057404368632e-05, "loss": 0.0811, "step": 11350 }, { "grad_norm": 0.7241939902305603, "learning_rate": 9.770438737883018e-05, "loss": 0.1133, "step": 11360 }, { "grad_norm": 1.1541101932525635, "learning_rate": 9.769819256259898e-05, "loss": 0.0865, "step": 11370 }, { "grad_norm": 0.7072658538818359, "learning_rate": 9.769198959605119e-05, "loss": 0.0971, "step": 11380 }, { "grad_norm": 0.5443269610404968, "learning_rate": 9.768577848024678e-05, "loss": 0.0771, "step": 11390 }, { "grad_norm": 0.8593826293945312, "learning_rate": 9.767955921624702e-05, "loss": 0.0921, "step": 11400 }, { "grad_norm": 0.7345855236053467, "learning_rate": 9.767333180511465e-05, "loss": 0.081, "step": 11410 }, { "grad_norm": 0.6867089867591858, "learning_rate": 9.766709624791373e-05, "loss": 0.109, "step": 11420 }, { "grad_norm": 0.597951352596283, "learning_rate": 9.766085254570975e-05, "loss": 0.0787, "step": 11430 }, { "grad_norm": 0.7478621602058411, "learning_rate": 9.76546006995696e-05, "loss": 0.0788, "step": 11440 }, { "grad_norm": 0.7434982061386108, "learning_rate": 9.764834071056155e-05, "loss": 0.0906, "step": 11450 }, { "grad_norm": 0.7487280964851379, "learning_rate": 9.764207257975526e-05, "loss": 0.083, "step": 11460 }, { "grad_norm": 0.7838597297668457, "learning_rate": 9.763579630822179e-05, "loss": 0.1029, "step": 11470 }, { "grad_norm": 0.5239253044128418, "learning_rate": 9.762951189703356e-05, "loss": 0.0921, "step": 11480 }, { "grad_norm": 0.6020888090133667, "learning_rate": 9.762321934726442e-05, "loss": 0.0848, "step": 11490 }, { "grad_norm": 0.5461079478263855, "learning_rate": 9.761691865998959e-05, "loss": 0.0677, "step": 11500 }, { "grad_norm": 0.6652058959007263, "learning_rate": 9.76106098362857e-05, "loss": 0.076, "step": 11510 }, { "grad_norm": 0.5087058544158936, "learning_rate": 9.760429287723072e-05, "loss": 0.0882, "step": 11520 }, { "grad_norm": 0.4774642884731293, "learning_rate": 9.759796778390406e-05, "loss": 0.0662, "step": 11530 }, { "grad_norm": 0.8220255970954895, "learning_rate": 9.759163455738653e-05, "loss": 0.0934, "step": 11540 }, { "grad_norm": 0.7458997964859009, "learning_rate": 9.75852931987603e-05, "loss": 0.094, "step": 11550 }, { "grad_norm": 0.7510386109352112, "learning_rate": 9.757894370910891e-05, "loss": 0.096, "step": 11560 }, { "grad_norm": 0.7538068890571594, "learning_rate": 9.757258608951733e-05, "loss": 0.0757, "step": 11570 }, { "grad_norm": 0.6953954696655273, "learning_rate": 9.75662203410719e-05, "loss": 0.0862, "step": 11580 }, { "grad_norm": 0.8762327432632446, "learning_rate": 9.755984646486034e-05, "loss": 0.1126, "step": 11590 }, { "grad_norm": 0.8275863528251648, "learning_rate": 9.75534644619718e-05, "loss": 0.0853, "step": 11600 }, { "grad_norm": 0.6261857151985168, "learning_rate": 9.754707433349676e-05, "loss": 0.0873, "step": 11610 }, { "grad_norm": 0.8440598845481873, "learning_rate": 9.754067608052715e-05, "loss": 0.0773, "step": 11620 }, { "grad_norm": 0.8013345003128052, "learning_rate": 9.753426970415622e-05, "loss": 0.0773, "step": 11630 }, { "grad_norm": 0.4819238781929016, "learning_rate": 9.752785520547868e-05, "loss": 0.081, "step": 11640 }, { "grad_norm": 0.5628504753112793, "learning_rate": 9.752143258559056e-05, "loss": 0.0882, "step": 11650 }, { "grad_norm": 0.6923178434371948, "learning_rate": 9.751500184558933e-05, "loss": 0.0789, "step": 11660 }, { "grad_norm": 0.9634439945220947, "learning_rate": 9.750856298657383e-05, "loss": 0.0859, "step": 11670 }, { "grad_norm": 0.5025120973587036, "learning_rate": 9.750211600964428e-05, "loss": 0.0745, "step": 11680 }, { "grad_norm": 0.9251523613929749, "learning_rate": 9.749566091590226e-05, "loss": 0.078, "step": 11690 }, { "grad_norm": 0.7151030898094177, "learning_rate": 9.748919770645083e-05, "loss": 0.0621, "step": 11700 }, { "grad_norm": 0.5650019645690918, "learning_rate": 9.748272638239432e-05, "loss": 0.0858, "step": 11710 }, { "grad_norm": 0.5083168745040894, "learning_rate": 9.747624694483855e-05, "loss": 0.0722, "step": 11720 }, { "grad_norm": 0.6733271479606628, "learning_rate": 9.746975939489065e-05, "loss": 0.077, "step": 11730 }, { "grad_norm": 0.678138792514801, "learning_rate": 9.746326373365918e-05, "loss": 0.0678, "step": 11740 }, { "grad_norm": 0.8434913158416748, "learning_rate": 9.745675996225403e-05, "loss": 0.1019, "step": 11750 }, { "grad_norm": 0.5495457053184509, "learning_rate": 9.745024808178657e-05, "loss": 0.0774, "step": 11760 }, { "grad_norm": 0.8567472100257874, "learning_rate": 9.744372809336947e-05, "loss": 0.1012, "step": 11770 }, { "grad_norm": 0.5660176873207092, "learning_rate": 9.743719999811682e-05, "loss": 0.0617, "step": 11780 }, { "grad_norm": 0.6120133399963379, "learning_rate": 9.743066379714412e-05, "loss": 0.0842, "step": 11790 }, { "grad_norm": 0.6259734630584717, "learning_rate": 9.74241194915682e-05, "loss": 0.1202, "step": 11800 }, { "grad_norm": 0.522071361541748, "learning_rate": 9.741756708250731e-05, "loss": 0.1089, "step": 11810 }, { "grad_norm": 0.39956316351890564, "learning_rate": 9.741100657108109e-05, "loss": 0.0823, "step": 11820 }, { "grad_norm": 0.6691367030143738, "learning_rate": 9.740443795841054e-05, "loss": 0.0731, "step": 11830 }, { "grad_norm": 0.9245563745498657, "learning_rate": 9.739786124561805e-05, "loss": 0.0952, "step": 11840 }, { "grad_norm": 0.4499029815196991, "learning_rate": 9.73912764338274e-05, "loss": 0.0791, "step": 11850 }, { "grad_norm": 0.6878378391265869, "learning_rate": 9.738468352416377e-05, "loss": 0.0791, "step": 11860 }, { "grad_norm": 0.7206786870956421, "learning_rate": 9.737808251775369e-05, "loss": 0.0894, "step": 11870 }, { "grad_norm": 0.6785558462142944, "learning_rate": 9.737147341572512e-05, "loss": 0.1035, "step": 11880 }, { "grad_norm": 0.5591559410095215, "learning_rate": 9.736485621920735e-05, "loss": 0.0874, "step": 11890 }, { "grad_norm": 0.769014835357666, "learning_rate": 9.735823092933108e-05, "loss": 0.0635, "step": 11900 }, { "grad_norm": 0.6340753436088562, "learning_rate": 9.735159754722838e-05, "loss": 0.076, "step": 11910 }, { "grad_norm": 0.7333327531814575, "learning_rate": 9.734495607403275e-05, "loss": 0.0817, "step": 11920 }, { "grad_norm": 0.5347902774810791, "learning_rate": 9.733830651087901e-05, "loss": 0.0745, "step": 11930 }, { "grad_norm": 0.8642117977142334, "learning_rate": 9.733164885890338e-05, "loss": 0.0979, "step": 11940 }, { "grad_norm": 0.5237802267074585, "learning_rate": 9.732498311924349e-05, "loss": 0.0816, "step": 11950 }, { "grad_norm": 0.7240167260169983, "learning_rate": 9.731830929303833e-05, "loss": 0.0746, "step": 11960 }, { "grad_norm": 0.7403400540351868, "learning_rate": 9.731162738142827e-05, "loss": 0.1013, "step": 11970 }, { "grad_norm": 0.6832247376441956, "learning_rate": 9.730493738555506e-05, "loss": 0.111, "step": 11980 }, { "grad_norm": 0.6852920651435852, "learning_rate": 9.729823930656186e-05, "loss": 0.0873, "step": 11990 }, { "grad_norm": 0.6723257899284363, "learning_rate": 9.729153314559316e-05, "loss": 0.1237, "step": 12000 }, { "grad_norm": 0.8805858492851257, "learning_rate": 9.728481890379486e-05, "loss": 0.091, "step": 12010 }, { "grad_norm": 0.6520620584487915, "learning_rate": 9.727809658231428e-05, "loss": 0.0757, "step": 12020 }, { "grad_norm": 0.4958205819129944, "learning_rate": 9.727136618230003e-05, "loss": 0.0925, "step": 12030 }, { "grad_norm": 0.597855806350708, "learning_rate": 9.726462770490219e-05, "loss": 0.081, "step": 12040 }, { "grad_norm": 0.4602605104446411, "learning_rate": 9.725788115127214e-05, "loss": 0.0686, "step": 12050 }, { "grad_norm": 0.4734260141849518, "learning_rate": 9.725112652256274e-05, "loss": 0.0866, "step": 12060 }, { "grad_norm": 0.6187310814857483, "learning_rate": 9.724436381992812e-05, "loss": 0.0733, "step": 12070 }, { "grad_norm": 0.7311898469924927, "learning_rate": 9.723759304452387e-05, "loss": 0.0849, "step": 12080 }, { "grad_norm": 0.5180095434188843, "learning_rate": 9.72308141975069e-05, "loss": 0.1046, "step": 12090 }, { "grad_norm": 0.7204815149307251, "learning_rate": 9.722402728003557e-05, "loss": 0.0899, "step": 12100 }, { "grad_norm": 0.5707830190658569, "learning_rate": 9.721723229326953e-05, "loss": 0.0712, "step": 12110 }, { "grad_norm": 0.5861068964004517, "learning_rate": 9.721042923836992e-05, "loss": 0.0667, "step": 12120 }, { "grad_norm": 1.130186915397644, "learning_rate": 9.720361811649914e-05, "loss": 0.0815, "step": 12130 }, { "grad_norm": 0.7992662191390991, "learning_rate": 9.719679892882106e-05, "loss": 0.1001, "step": 12140 }, { "grad_norm": 0.8626496195793152, "learning_rate": 9.718997167650085e-05, "loss": 0.1061, "step": 12150 }, { "grad_norm": 0.6877887845039368, "learning_rate": 9.718313636070515e-05, "loss": 0.1045, "step": 12160 }, { "grad_norm": 0.7262735366821289, "learning_rate": 9.717629298260192e-05, "loss": 0.098, "step": 12170 }, { "grad_norm": 0.8845481276512146, "learning_rate": 9.716944154336047e-05, "loss": 0.0906, "step": 12180 }, { "grad_norm": 0.625058114528656, "learning_rate": 9.716258204415157e-05, "loss": 0.1023, "step": 12190 }, { "grad_norm": 0.6562637090682983, "learning_rate": 9.715571448614728e-05, "loss": 0.107, "step": 12200 }, { "grad_norm": 0.6236597299575806, "learning_rate": 9.71488388705211e-05, "loss": 0.0816, "step": 12210 }, { "grad_norm": 0.653003454208374, "learning_rate": 9.714195519844788e-05, "loss": 0.078, "step": 12220 }, { "grad_norm": 0.7485367655754089, "learning_rate": 9.713506347110386e-05, "loss": 0.0662, "step": 12230 }, { "grad_norm": 0.7469483017921448, "learning_rate": 9.712816368966662e-05, "loss": 0.095, "step": 12240 }, { "grad_norm": 0.46583694219589233, "learning_rate": 9.712125585531517e-05, "loss": 0.0776, "step": 12250 }, { "grad_norm": 0.5206833481788635, "learning_rate": 9.711433996922988e-05, "loss": 0.0616, "step": 12260 }, { "grad_norm": 0.5713030099868774, "learning_rate": 9.710741603259245e-05, "loss": 0.0723, "step": 12270 }, { "grad_norm": 1.0328154563903809, "learning_rate": 9.710048404658603e-05, "loss": 0.0929, "step": 12280 }, { "grad_norm": 0.61082524061203, "learning_rate": 9.709354401239508e-05, "loss": 0.1072, "step": 12290 }, { "grad_norm": 0.44916006922721863, "learning_rate": 9.708659593120546e-05, "loss": 0.0937, "step": 12300 }, { "grad_norm": 0.7566354274749756, "learning_rate": 9.707963980420443e-05, "loss": 0.0985, "step": 12310 }, { "grad_norm": 0.7629588842391968, "learning_rate": 9.707267563258058e-05, "loss": 0.1205, "step": 12320 }, { "grad_norm": 0.699417233467102, "learning_rate": 9.70657034175239e-05, "loss": 0.1163, "step": 12330 }, { "grad_norm": 0.8450615406036377, "learning_rate": 9.705872316022577e-05, "loss": 0.0931, "step": 12340 }, { "grad_norm": 0.7191195487976074, "learning_rate": 9.705173486187891e-05, "loss": 0.079, "step": 12350 }, { "grad_norm": 0.6944852471351624, "learning_rate": 9.704473852367741e-05, "loss": 0.0803, "step": 12360 }, { "grad_norm": 0.7114837765693665, "learning_rate": 9.70377341468168e-05, "loss": 0.1093, "step": 12370 }, { "grad_norm": 0.5071892738342285, "learning_rate": 9.703072173249389e-05, "loss": 0.0847, "step": 12380 }, { "grad_norm": 0.5999922156333923, "learning_rate": 9.702370128190693e-05, "loss": 0.0967, "step": 12390 }, { "grad_norm": 0.5221030116081238, "learning_rate": 9.701667279625552e-05, "loss": 0.0704, "step": 12400 }, { "grad_norm": 0.6143369674682617, "learning_rate": 9.700963627674065e-05, "loss": 0.092, "step": 12410 }, { "grad_norm": 0.8628034591674805, "learning_rate": 9.700259172456466e-05, "loss": 0.0909, "step": 12420 }, { "grad_norm": 0.8421007394790649, "learning_rate": 9.699553914093124e-05, "loss": 0.0861, "step": 12430 }, { "grad_norm": 0.8661388754844666, "learning_rate": 9.698847852704553e-05, "loss": 0.0771, "step": 12440 }, { "grad_norm": 0.8120477795600891, "learning_rate": 9.6981409884114e-05, "loss": 0.1011, "step": 12450 }, { "grad_norm": 0.6839486360549927, "learning_rate": 9.697433321334443e-05, "loss": 0.0824, "step": 12460 }, { "grad_norm": 0.7947761416435242, "learning_rate": 9.696724851594607e-05, "loss": 0.1007, "step": 12470 }, { "grad_norm": 0.8618309497833252, "learning_rate": 9.696015579312952e-05, "loss": 0.087, "step": 12480 }, { "grad_norm": 0.7995891571044922, "learning_rate": 9.695305504610668e-05, "loss": 0.092, "step": 12490 }, { "grad_norm": 0.8497946262359619, "learning_rate": 9.694594627609092e-05, "loss": 0.1025, "step": 12500 }, { "grad_norm": 0.801908016204834, "learning_rate": 9.693882948429691e-05, "loss": 0.0801, "step": 12510 }, { "grad_norm": 0.7751647233963013, "learning_rate": 9.693170467194071e-05, "loss": 0.0817, "step": 12520 }, { "grad_norm": 0.7698662877082825, "learning_rate": 9.692457184023977e-05, "loss": 0.0752, "step": 12530 }, { "grad_norm": 0.7246552109718323, "learning_rate": 9.691743099041291e-05, "loss": 0.0927, "step": 12540 }, { "grad_norm": 0.8171444535255432, "learning_rate": 9.691028212368027e-05, "loss": 0.0804, "step": 12550 }, { "grad_norm": 0.6071585416793823, "learning_rate": 9.690312524126342e-05, "loss": 0.084, "step": 12560 }, { "grad_norm": 0.708432674407959, "learning_rate": 9.689596034438527e-05, "loss": 0.0826, "step": 12570 }, { "grad_norm": 0.6841686367988586, "learning_rate": 9.688878743427012e-05, "loss": 0.0934, "step": 12580 }, { "grad_norm": 0.6742311716079712, "learning_rate": 9.688160651214359e-05, "loss": 0.0922, "step": 12590 }, { "grad_norm": 0.6568033695220947, "learning_rate": 9.687441757923273e-05, "loss": 0.0605, "step": 12600 }, { "grad_norm": 0.7475361227989197, "learning_rate": 9.68672206367659e-05, "loss": 0.0793, "step": 12610 }, { "grad_norm": 0.72279292345047, "learning_rate": 9.686001568597291e-05, "loss": 0.0719, "step": 12620 }, { "grad_norm": 0.5262613892555237, "learning_rate": 9.685280272808486e-05, "loss": 0.0841, "step": 12630 }, { "grad_norm": 0.7730474472045898, "learning_rate": 9.684558176433424e-05, "loss": 0.0836, "step": 12640 }, { "grad_norm": 0.5946832895278931, "learning_rate": 9.683835279595495e-05, "loss": 0.1011, "step": 12650 }, { "grad_norm": 0.8728222250938416, "learning_rate": 9.683111582418216e-05, "loss": 0.0779, "step": 12660 }, { "grad_norm": 0.8108968734741211, "learning_rate": 9.682387085025254e-05, "loss": 0.1083, "step": 12670 }, { "grad_norm": 0.8507341146469116, "learning_rate": 9.681661787540401e-05, "loss": 0.0805, "step": 12680 }, { "grad_norm": 0.7980782389640808, "learning_rate": 9.680935690087593e-05, "loss": 0.096, "step": 12690 }, { "grad_norm": 0.5163954496383667, "learning_rate": 9.680208792790901e-05, "loss": 0.0689, "step": 12700 }, { "grad_norm": 0.6993695497512817, "learning_rate": 9.679481095774529e-05, "loss": 0.1014, "step": 12710 }, { "grad_norm": 0.6646828055381775, "learning_rate": 9.678752599162822e-05, "loss": 0.0725, "step": 12720 }, { "grad_norm": 0.6717754602432251, "learning_rate": 9.678023303080259e-05, "loss": 0.0811, "step": 12730 }, { "grad_norm": 0.7289107441902161, "learning_rate": 9.677293207651459e-05, "loss": 0.0909, "step": 12740 }, { "grad_norm": 0.8553359508514404, "learning_rate": 9.676562313001173e-05, "loss": 0.0823, "step": 12750 }, { "grad_norm": 0.6299761533737183, "learning_rate": 9.675830619254293e-05, "loss": 0.0784, "step": 12760 }, { "grad_norm": 0.6923697590827942, "learning_rate": 9.675098126535843e-05, "loss": 0.0721, "step": 12770 }, { "grad_norm": 1.1491882801055908, "learning_rate": 9.674364834970988e-05, "loss": 0.0837, "step": 12780 }, { "grad_norm": 0.7996108531951904, "learning_rate": 9.673630744685028e-05, "loss": 0.072, "step": 12790 }, { "grad_norm": 0.6056500673294067, "learning_rate": 9.672895855803397e-05, "loss": 0.1059, "step": 12800 }, { "grad_norm": 0.6558142304420471, "learning_rate": 9.672160168451667e-05, "loss": 0.056, "step": 12810 }, { "grad_norm": 0.47917112708091736, "learning_rate": 9.671423682755549e-05, "loss": 0.0569, "step": 12820 }, { "grad_norm": 0.5677293539047241, "learning_rate": 9.670686398840888e-05, "loss": 0.0984, "step": 12830 }, { "grad_norm": 0.624017596244812, "learning_rate": 9.669948316833664e-05, "loss": 0.0675, "step": 12840 }, { "grad_norm": 0.6423555612564087, "learning_rate": 9.669209436859997e-05, "loss": 0.0878, "step": 12850 }, { "grad_norm": 0.7092372179031372, "learning_rate": 9.66846975904614e-05, "loss": 0.0895, "step": 12860 }, { "grad_norm": 0.5257102251052856, "learning_rate": 9.667729283518483e-05, "loss": 0.0754, "step": 12870 }, { "grad_norm": 0.5639511346817017, "learning_rate": 9.666988010403557e-05, "loss": 0.0882, "step": 12880 }, { "grad_norm": 0.8407220244407654, "learning_rate": 9.66624593982802e-05, "loss": 0.0714, "step": 12890 }, { "grad_norm": 0.7748295664787292, "learning_rate": 9.665503071918675e-05, "loss": 0.0704, "step": 12900 }, { "grad_norm": 0.8411682844161987, "learning_rate": 9.664759406802456e-05, "loss": 0.0872, "step": 12910 }, { "grad_norm": 0.6332447528839111, "learning_rate": 9.664014944606437e-05, "loss": 0.0727, "step": 12920 }, { "grad_norm": 0.5959235429763794, "learning_rate": 9.663269685457822e-05, "loss": 0.0736, "step": 12930 }, { "grad_norm": 0.5143741965293884, "learning_rate": 9.662523629483962e-05, "loss": 0.0762, "step": 12940 }, { "grad_norm": 0.6160115599632263, "learning_rate": 9.661776776812333e-05, "loss": 0.0633, "step": 12950 }, { "grad_norm": 0.7264766693115234, "learning_rate": 9.661029127570553e-05, "loss": 0.0627, "step": 12960 }, { "grad_norm": 0.6639412641525269, "learning_rate": 9.660280681886373e-05, "loss": 0.0775, "step": 12970 }, { "grad_norm": 0.6092992424964905, "learning_rate": 9.659531439887685e-05, "loss": 0.1076, "step": 12980 }, { "grad_norm": 1.0777708292007446, "learning_rate": 9.658781401702511e-05, "loss": 0.0929, "step": 12990 }, { "grad_norm": 0.5255779027938843, "learning_rate": 9.658030567459015e-05, "loss": 0.0786, "step": 13000 }, { "grad_norm": 0.6239049434661865, "learning_rate": 9.65727893728549e-05, "loss": 0.0696, "step": 13010 }, { "grad_norm": 0.7038301229476929, "learning_rate": 9.656526511310375e-05, "loss": 0.0689, "step": 13020 }, { "grad_norm": 0.8262446522712708, "learning_rate": 9.655773289662233e-05, "loss": 0.0735, "step": 13030 }, { "grad_norm": 0.8555121421813965, "learning_rate": 9.655019272469772e-05, "loss": 0.0774, "step": 13040 }, { "grad_norm": 0.7393469214439392, "learning_rate": 9.654264459861832e-05, "loss": 0.0877, "step": 13050 }, { "grad_norm": 0.6367297172546387, "learning_rate": 9.653508851967391e-05, "loss": 0.0797, "step": 13060 }, { "grad_norm": 0.899969756603241, "learning_rate": 9.65275244891556e-05, "loss": 0.099, "step": 13070 }, { "grad_norm": 0.5947976112365723, "learning_rate": 9.651995250835591e-05, "loss": 0.0714, "step": 13080 }, { "grad_norm": 0.6573998928070068, "learning_rate": 9.651237257856862e-05, "loss": 0.0596, "step": 13090 }, { "grad_norm": 0.63821941614151, "learning_rate": 9.6504784701089e-05, "loss": 0.0711, "step": 13100 }, { "grad_norm": 0.6497894525527954, "learning_rate": 9.649718887721357e-05, "loss": 0.0877, "step": 13110 }, { "grad_norm": 0.6810867190361023, "learning_rate": 9.648958510824028e-05, "loss": 0.0888, "step": 13120 }, { "grad_norm": 0.8263066411018372, "learning_rate": 9.648197339546837e-05, "loss": 0.0848, "step": 13130 }, { "grad_norm": 0.6600959300994873, "learning_rate": 9.647435374019851e-05, "loss": 0.1016, "step": 13140 }, { "grad_norm": 0.7852066159248352, "learning_rate": 9.646672614373266e-05, "loss": 0.0643, "step": 13150 }, { "grad_norm": 0.43821966648101807, "learning_rate": 9.645909060737418e-05, "loss": 0.0727, "step": 13160 }, { "grad_norm": 0.7708205580711365, "learning_rate": 9.645144713242778e-05, "loss": 0.0975, "step": 13170 }, { "grad_norm": 0.6374751329421997, "learning_rate": 9.64437957201995e-05, "loss": 0.0946, "step": 13180 }, { "grad_norm": 0.8301674127578735, "learning_rate": 9.643613637199678e-05, "loss": 0.0889, "step": 13190 }, { "grad_norm": 0.5554658770561218, "learning_rate": 9.642846908912839e-05, "loss": 0.0913, "step": 13200 }, { "grad_norm": 0.8718764185905457, "learning_rate": 9.642079387290444e-05, "loss": 0.0747, "step": 13210 }, { "grad_norm": 0.8317239880561829, "learning_rate": 9.641311072463644e-05, "loss": 0.0787, "step": 13220 }, { "grad_norm": 0.7097475528717041, "learning_rate": 9.640541964563722e-05, "loss": 0.0743, "step": 13230 }, { "grad_norm": 0.5897259712219238, "learning_rate": 9.639772063722096e-05, "loss": 0.0612, "step": 13240 }, { "grad_norm": 0.5616939663887024, "learning_rate": 9.639001370070324e-05, "loss": 0.0719, "step": 13250 }, { "grad_norm": 0.8753315210342407, "learning_rate": 9.638229883740095e-05, "loss": 0.0735, "step": 13260 }, { "grad_norm": 0.5513785481452942, "learning_rate": 9.637457604863233e-05, "loss": 0.0789, "step": 13270 }, { "grad_norm": 0.8379477262496948, "learning_rate": 9.636684533571703e-05, "loss": 0.0674, "step": 13280 }, { "grad_norm": 0.5731685757637024, "learning_rate": 9.635910669997599e-05, "loss": 0.0661, "step": 13290 }, { "grad_norm": 0.6795346736907959, "learning_rate": 9.635136014273154e-05, "loss": 0.1111, "step": 13300 }, { "grad_norm": 0.6826565265655518, "learning_rate": 9.634360566530735e-05, "loss": 0.074, "step": 13310 }, { "grad_norm": 0.5198861360549927, "learning_rate": 9.633584326902845e-05, "loss": 0.0654, "step": 13320 }, { "grad_norm": 0.5439481735229492, "learning_rate": 9.632807295522124e-05, "loss": 0.0672, "step": 13330 }, { "grad_norm": 0.4998805820941925, "learning_rate": 9.632029472521342e-05, "loss": 0.0596, "step": 13340 }, { "grad_norm": 0.6418207287788391, "learning_rate": 9.631250858033409e-05, "loss": 0.0862, "step": 13350 }, { "grad_norm": 0.6258972883224487, "learning_rate": 9.630471452191371e-05, "loss": 0.0861, "step": 13360 }, { "grad_norm": 0.7925549149513245, "learning_rate": 9.629691255128405e-05, "loss": 0.0761, "step": 13370 }, { "grad_norm": 0.6639201045036316, "learning_rate": 9.628910266977825e-05, "loss": 0.0857, "step": 13380 }, { "grad_norm": 0.6163544058799744, "learning_rate": 9.628128487873083e-05, "loss": 0.0688, "step": 13390 }, { "grad_norm": 0.5951204299926758, "learning_rate": 9.627345917947761e-05, "loss": 0.0746, "step": 13400 }, { "grad_norm": 0.9064527750015259, "learning_rate": 9.626562557335579e-05, "loss": 0.0914, "step": 13410 }, { "grad_norm": 0.5691927671432495, "learning_rate": 9.625778406170393e-05, "loss": 0.0879, "step": 13420 }, { "grad_norm": 0.5560926198959351, "learning_rate": 9.624993464586193e-05, "loss": 0.0997, "step": 13430 }, { "grad_norm": 0.7517156004905701, "learning_rate": 9.624207732717105e-05, "loss": 0.0893, "step": 13440 }, { "grad_norm": 0.7323119640350342, "learning_rate": 9.623421210697386e-05, "loss": 0.0705, "step": 13450 }, { "grad_norm": 0.758059024810791, "learning_rate": 9.622633898661434e-05, "loss": 0.0861, "step": 13460 }, { "grad_norm": 1.2613530158996582, "learning_rate": 9.621845796743778e-05, "loss": 0.1324, "step": 13470 }, { "grad_norm": 0.7133752107620239, "learning_rate": 9.621056905079082e-05, "loss": 0.0872, "step": 13480 }, { "grad_norm": 0.8120458126068115, "learning_rate": 9.620267223802149e-05, "loss": 0.0941, "step": 13490 }, { "grad_norm": 0.6869061589241028, "learning_rate": 9.619476753047911e-05, "loss": 0.0777, "step": 13500 }, { "grad_norm": 0.5442853569984436, "learning_rate": 9.618685492951438e-05, "loss": 0.0795, "step": 13510 }, { "grad_norm": 0.6511843204498291, "learning_rate": 9.617893443647938e-05, "loss": 0.0785, "step": 13520 }, { "grad_norm": 0.5586027503013611, "learning_rate": 9.617100605272746e-05, "loss": 0.0734, "step": 13530 }, { "grad_norm": 0.8875332474708557, "learning_rate": 9.616306977961338e-05, "loss": 0.0709, "step": 13540 }, { "grad_norm": 0.7981228828430176, "learning_rate": 9.615512561849326e-05, "loss": 0.0882, "step": 13550 }, { "grad_norm": 0.6465672850608826, "learning_rate": 9.61471735707245e-05, "loss": 0.0978, "step": 13560 }, { "grad_norm": 0.6422000527381897, "learning_rate": 9.613921363766592e-05, "loss": 0.0774, "step": 13570 }, { "grad_norm": 0.6363774538040161, "learning_rate": 9.613124582067763e-05, "loss": 0.093, "step": 13580 }, { "grad_norm": 0.6173588633537292, "learning_rate": 9.612327012112112e-05, "loss": 0.0806, "step": 13590 }, { "grad_norm": 0.6650611162185669, "learning_rate": 9.611528654035921e-05, "loss": 0.1016, "step": 13600 }, { "grad_norm": 0.750309944152832, "learning_rate": 9.610729507975611e-05, "loss": 0.0644, "step": 13610 }, { "grad_norm": 0.8461269736289978, "learning_rate": 9.609929574067731e-05, "loss": 0.0663, "step": 13620 }, { "grad_norm": 0.758895993232727, "learning_rate": 9.609128852448967e-05, "loss": 0.0731, "step": 13630 }, { "grad_norm": 0.7005458474159241, "learning_rate": 9.608327343256143e-05, "loss": 0.0976, "step": 13640 }, { "grad_norm": 0.8244765996932983, "learning_rate": 9.607525046626216e-05, "loss": 0.0692, "step": 13650 }, { "grad_norm": 0.8866346478462219, "learning_rate": 9.606721962696272e-05, "loss": 0.0872, "step": 13660 }, { "grad_norm": 0.814004123210907, "learning_rate": 9.60591809160354e-05, "loss": 0.098, "step": 13670 }, { "grad_norm": 1.066471815109253, "learning_rate": 9.605113433485378e-05, "loss": 0.0876, "step": 13680 }, { "grad_norm": 0.9139153957366943, "learning_rate": 9.604307988479279e-05, "loss": 0.0839, "step": 13690 }, { "grad_norm": 0.6648253798484802, "learning_rate": 9.603501756722876e-05, "loss": 0.0867, "step": 13700 }, { "grad_norm": 0.5476635098457336, "learning_rate": 9.602694738353927e-05, "loss": 0.0769, "step": 13710 }, { "grad_norm": 0.8447538018226624, "learning_rate": 9.601886933510331e-05, "loss": 0.0726, "step": 13720 }, { "grad_norm": 0.5359560251235962, "learning_rate": 9.60107834233012e-05, "loss": 0.1107, "step": 13730 }, { "grad_norm": 0.7587805390357971, "learning_rate": 9.60026896495146e-05, "loss": 0.0811, "step": 13740 }, { "grad_norm": 0.6511942148208618, "learning_rate": 9.599458801512652e-05, "loss": 0.0718, "step": 13750 }, { "grad_norm": 0.41139668226242065, "learning_rate": 9.598647852152129e-05, "loss": 0.0892, "step": 13760 }, { "grad_norm": 1.3006458282470703, "learning_rate": 9.597836117008462e-05, "loss": 0.0828, "step": 13770 }, { "grad_norm": 0.8124643564224243, "learning_rate": 9.597023596220356e-05, "loss": 0.0877, "step": 13780 }, { "grad_norm": 0.6113566756248474, "learning_rate": 9.596210289926643e-05, "loss": 0.0693, "step": 13790 }, { "grad_norm": 0.48924070596694946, "learning_rate": 9.5953961982663e-05, "loss": 0.0951, "step": 13800 }, { "grad_norm": 0.43748316168785095, "learning_rate": 9.594581321378431e-05, "loss": 0.0704, "step": 13810 }, { "grad_norm": 0.6833425164222717, "learning_rate": 9.593765659402276e-05, "loss": 0.084, "step": 13820 }, { "grad_norm": 0.6407827734947205, "learning_rate": 9.59294921247721e-05, "loss": 0.0766, "step": 13830 }, { "grad_norm": 0.45142248272895813, "learning_rate": 9.59213198074274e-05, "loss": 0.0952, "step": 13840 }, { "grad_norm": 0.7502511739730835, "learning_rate": 9.59131396433851e-05, "loss": 0.0759, "step": 13850 }, { "grad_norm": 0.5681416988372803, "learning_rate": 9.590495163404297e-05, "loss": 0.0774, "step": 13860 }, { "grad_norm": 0.6647101640701294, "learning_rate": 9.589675578080009e-05, "loss": 0.0765, "step": 13870 }, { "grad_norm": 0.7066798806190491, "learning_rate": 9.588855208505694e-05, "loss": 0.072, "step": 13880 }, { "grad_norm": 0.741698145866394, "learning_rate": 9.588034054821529e-05, "loss": 0.0701, "step": 13890 }, { "grad_norm": 0.6849616169929504, "learning_rate": 9.587212117167826e-05, "loss": 0.0687, "step": 13900 }, { "grad_norm": 0.6926918625831604, "learning_rate": 9.586389395685033e-05, "loss": 0.0856, "step": 13910 }, { "grad_norm": 0.7512252926826477, "learning_rate": 9.585565890513733e-05, "loss": 0.0716, "step": 13920 }, { "grad_norm": 0.5380191802978516, "learning_rate": 9.584741601794636e-05, "loss": 0.0875, "step": 13930 }, { "grad_norm": 0.47446730732917786, "learning_rate": 9.58391652966859e-05, "loss": 0.0848, "step": 13940 }, { "grad_norm": 0.8053243160247803, "learning_rate": 9.583090674276583e-05, "loss": 0.0722, "step": 13950 }, { "grad_norm": 0.45753931999206543, "learning_rate": 9.582264035759726e-05, "loss": 0.0972, "step": 13960 }, { "grad_norm": 0.6277456879615784, "learning_rate": 9.58143661425927e-05, "loss": 0.0592, "step": 13970 }, { "grad_norm": 0.6196413040161133, "learning_rate": 9.580608409916601e-05, "loss": 0.0633, "step": 13980 }, { "grad_norm": 0.7146477103233337, "learning_rate": 9.579779422873233e-05, "loss": 0.0836, "step": 13990 }, { "grad_norm": 0.7086893916130066, "learning_rate": 9.578949653270819e-05, "loss": 0.0757, "step": 14000 }, { "grad_norm": 0.5928334593772888, "learning_rate": 9.578119101251144e-05, "loss": 0.0889, "step": 14010 }, { "grad_norm": 0.6521613597869873, "learning_rate": 9.577287766956127e-05, "loss": 0.0827, "step": 14020 }, { "grad_norm": 0.7098095417022705, "learning_rate": 9.57645565052782e-05, "loss": 0.0651, "step": 14030 }, { "grad_norm": 0.632637083530426, "learning_rate": 9.575622752108407e-05, "loss": 0.0755, "step": 14040 }, { "grad_norm": 0.7037423253059387, "learning_rate": 9.57478907184021e-05, "loss": 0.1086, "step": 14050 }, { "grad_norm": 0.7031689882278442, "learning_rate": 9.573954609865681e-05, "loss": 0.0859, "step": 14060 }, { "grad_norm": 0.6831482648849487, "learning_rate": 9.573119366327408e-05, "loss": 0.0686, "step": 14070 }, { "grad_norm": 0.7334182262420654, "learning_rate": 9.57228334136811e-05, "loss": 0.0809, "step": 14080 }, { "grad_norm": 0.5740038156509399, "learning_rate": 9.571446535130641e-05, "loss": 0.0697, "step": 14090 }, { "grad_norm": 0.5966970324516296, "learning_rate": 9.570608947757988e-05, "loss": 0.0765, "step": 14100 }, { "grad_norm": 0.6257467865943909, "learning_rate": 9.569770579393274e-05, "loss": 0.0649, "step": 14110 }, { "grad_norm": 0.7471415996551514, "learning_rate": 9.56893143017975e-05, "loss": 0.0628, "step": 14120 }, { "grad_norm": 0.6007633209228516, "learning_rate": 9.568091500260806e-05, "loss": 0.0816, "step": 14130 }, { "grad_norm": 0.7184690237045288, "learning_rate": 9.567250789779961e-05, "loss": 0.0806, "step": 14140 }, { "grad_norm": 0.5987624526023865, "learning_rate": 9.566409298880872e-05, "loss": 0.0874, "step": 14150 }, { "grad_norm": 0.7656945586204529, "learning_rate": 9.565567027707326e-05, "loss": 0.1045, "step": 14160 }, { "grad_norm": 0.8480462431907654, "learning_rate": 9.56472397640324e-05, "loss": 0.0755, "step": 14170 }, { "grad_norm": 0.5916301608085632, "learning_rate": 9.563880145112675e-05, "loss": 0.0891, "step": 14180 }, { "grad_norm": 0.672315239906311, "learning_rate": 9.563035533979814e-05, "loss": 0.0571, "step": 14190 }, { "grad_norm": 0.7575505375862122, "learning_rate": 9.562190143148981e-05, "loss": 0.0817, "step": 14200 }, { "grad_norm": 0.6143346428871155, "learning_rate": 9.561343972764627e-05, "loss": 0.1114, "step": 14210 }, { "grad_norm": 0.6760562658309937, "learning_rate": 9.560497022971343e-05, "loss": 0.0719, "step": 14220 }, { "grad_norm": 0.7568971514701843, "learning_rate": 9.559649293913847e-05, "loss": 0.0787, "step": 14230 }, { "grad_norm": 0.6758793592453003, "learning_rate": 9.558800785736993e-05, "loss": 0.0709, "step": 14240 }, { "grad_norm": 0.7528777122497559, "learning_rate": 9.557951498585767e-05, "loss": 0.0765, "step": 14250 }, { "grad_norm": 1.162812352180481, "learning_rate": 9.557101432605293e-05, "loss": 0.076, "step": 14260 }, { "grad_norm": 0.9336472749710083, "learning_rate": 9.556250587940818e-05, "loss": 0.0913, "step": 14270 }, { "grad_norm": 0.6149274706840515, "learning_rate": 9.555398964737734e-05, "loss": 0.0957, "step": 14280 }, { "grad_norm": 0.7750711441040039, "learning_rate": 9.554546563141555e-05, "loss": 0.0963, "step": 14290 }, { "grad_norm": 0.5385413765907288, "learning_rate": 9.553693383297937e-05, "loss": 0.0758, "step": 14300 }, { "grad_norm": 0.594761848449707, "learning_rate": 9.552839425352663e-05, "loss": 0.0675, "step": 14310 }, { "grad_norm": 1.1306012868881226, "learning_rate": 9.551984689451652e-05, "loss": 0.0912, "step": 14320 }, { "grad_norm": 0.8249631524085999, "learning_rate": 9.551129175740953e-05, "loss": 0.0836, "step": 14330 }, { "grad_norm": 0.8112331032752991, "learning_rate": 9.550272884366754e-05, "loss": 0.0742, "step": 14340 }, { "grad_norm": 0.6747589707374573, "learning_rate": 9.549415815475369e-05, "loss": 0.08, "step": 14350 }, { "grad_norm": 0.6197712421417236, "learning_rate": 9.548557969213247e-05, "loss": 0.0761, "step": 14360 }, { "grad_norm": 0.7154834270477295, "learning_rate": 9.547699345726972e-05, "loss": 0.0924, "step": 14370 }, { "grad_norm": 0.5276948809623718, "learning_rate": 9.546839945163257e-05, "loss": 0.0838, "step": 14380 }, { "grad_norm": 0.4635521471500397, "learning_rate": 9.545979767668953e-05, "loss": 0.0733, "step": 14390 }, { "grad_norm": 0.733319878578186, "learning_rate": 9.54511881339104e-05, "loss": 0.0723, "step": 14400 }, { "grad_norm": 0.7873251438140869, "learning_rate": 9.54425708247663e-05, "loss": 0.0678, "step": 14410 }, { "grad_norm": 0.8751129508018494, "learning_rate": 9.543394575072972e-05, "loss": 0.0913, "step": 14420 }, { "grad_norm": 0.9974656105041504, "learning_rate": 9.542531291327441e-05, "loss": 0.0892, "step": 14430 }, { "grad_norm": 0.7318511009216309, "learning_rate": 9.541667231387552e-05, "loss": 0.0858, "step": 14440 }, { "grad_norm": 0.8378045558929443, "learning_rate": 9.540802395400949e-05, "loss": 0.0753, "step": 14450 }, { "grad_norm": 0.604774534702301, "learning_rate": 9.539936783515406e-05, "loss": 0.0759, "step": 14460 }, { "grad_norm": 0.6471384763717651, "learning_rate": 9.539070395878835e-05, "loss": 0.0748, "step": 14470 }, { "grad_norm": 0.7034491300582886, "learning_rate": 9.538203232639277e-05, "loss": 0.0869, "step": 14480 }, { "grad_norm": 0.5977133512496948, "learning_rate": 9.537335293944907e-05, "loss": 0.077, "step": 14490 }, { "grad_norm": 0.7281062006950378, "learning_rate": 9.536466579944032e-05, "loss": 0.0757, "step": 14500 }, { "grad_norm": 0.6482281684875488, "learning_rate": 9.535597090785091e-05, "loss": 0.0816, "step": 14510 }, { "grad_norm": 0.8723535537719727, "learning_rate": 9.534726826616656e-05, "loss": 0.0815, "step": 14520 }, { "grad_norm": 0.6080360412597656, "learning_rate": 9.53385578758743e-05, "loss": 0.0891, "step": 14530 }, { "grad_norm": 0.9196807742118835, "learning_rate": 9.532983973846252e-05, "loss": 0.0706, "step": 14540 }, { "grad_norm": 0.6867939233779907, "learning_rate": 9.53211138554209e-05, "loss": 0.0716, "step": 14550 }, { "grad_norm": 0.6872129440307617, "learning_rate": 9.531238022824047e-05, "loss": 0.0977, "step": 14560 }, { "grad_norm": 0.7142597436904907, "learning_rate": 9.530363885841355e-05, "loss": 0.0745, "step": 14570 }, { "grad_norm": 0.5690364837646484, "learning_rate": 9.52948897474338e-05, "loss": 0.0791, "step": 14580 }, { "grad_norm": 0.6274771094322205, "learning_rate": 9.528613289679622e-05, "loss": 0.0658, "step": 14590 }, { "grad_norm": 0.7076817750930786, "learning_rate": 9.52773683079971e-05, "loss": 0.0767, "step": 14600 }, { "grad_norm": 0.6363162994384766, "learning_rate": 9.526859598253407e-05, "loss": 0.0802, "step": 14610 }, { "grad_norm": 0.6687392592430115, "learning_rate": 9.525981592190609e-05, "loss": 0.0773, "step": 14620 }, { "grad_norm": 0.6686921715736389, "learning_rate": 9.525102812761342e-05, "loss": 0.067, "step": 14630 }, { "grad_norm": 0.6420778632164001, "learning_rate": 9.524223260115768e-05, "loss": 0.0737, "step": 14640 }, { "grad_norm": 0.5423732995986938, "learning_rate": 9.523342934404175e-05, "loss": 0.074, "step": 14650 }, { "grad_norm": 1.2105352878570557, "learning_rate": 9.522461835776989e-05, "loss": 0.0764, "step": 14660 }, { "grad_norm": 0.6520020365715027, "learning_rate": 9.521579964384764e-05, "loss": 0.0801, "step": 14670 }, { "grad_norm": 0.6237777471542358, "learning_rate": 9.52069732037819e-05, "loss": 0.0956, "step": 14680 }, { "grad_norm": 0.5739642381668091, "learning_rate": 9.519813903908083e-05, "loss": 0.072, "step": 14690 }, { "grad_norm": 0.7528777718544006, "learning_rate": 9.5189297151254e-05, "loss": 0.0831, "step": 14700 }, { "grad_norm": 0.5434434413909912, "learning_rate": 9.518044754181218e-05, "loss": 0.0664, "step": 14710 }, { "grad_norm": 0.7115609645843506, "learning_rate": 9.51715902122676e-05, "loss": 0.0763, "step": 14720 }, { "grad_norm": 0.7195811867713928, "learning_rate": 9.516272516413368e-05, "loss": 0.0836, "step": 14730 }, { "grad_norm": 0.7856733798980713, "learning_rate": 9.515385239892525e-05, "loss": 0.0794, "step": 14740 }, { "grad_norm": 0.9232592582702637, "learning_rate": 9.514497191815839e-05, "loss": 0.0788, "step": 14750 }, { "grad_norm": 0.5507126450538635, "learning_rate": 9.513608372335055e-05, "loss": 0.0738, "step": 14760 }, { "grad_norm": 0.5195257067680359, "learning_rate": 9.512718781602045e-05, "loss": 0.0781, "step": 14770 }, { "grad_norm": 0.6266555190086365, "learning_rate": 9.511828419768823e-05, "loss": 0.0779, "step": 14780 }, { "grad_norm": 0.6848190426826477, "learning_rate": 9.510937286987521e-05, "loss": 0.0773, "step": 14790 }, { "grad_norm": 0.6954227685928345, "learning_rate": 9.510045383410408e-05, "loss": 0.0728, "step": 14800 }, { "grad_norm": 0.5358026027679443, "learning_rate": 9.509152709189892e-05, "loss": 0.0771, "step": 14810 }, { "grad_norm": 0.6281309127807617, "learning_rate": 9.508259264478504e-05, "loss": 0.0702, "step": 14820 }, { "grad_norm": 0.7970719337463379, "learning_rate": 9.507365049428909e-05, "loss": 0.0864, "step": 14830 }, { "grad_norm": 0.7150878310203552, "learning_rate": 9.506470064193902e-05, "loss": 0.0952, "step": 14840 }, { "grad_norm": 1.114524006843567, "learning_rate": 9.505574308926414e-05, "loss": 0.1054, "step": 14850 }, { "grad_norm": 0.951076328754425, "learning_rate": 9.504677783779505e-05, "loss": 0.0779, "step": 14860 }, { "grad_norm": 0.563945472240448, "learning_rate": 9.503780488906365e-05, "loss": 0.0765, "step": 14870 }, { "grad_norm": 0.6033415794372559, "learning_rate": 9.502882424460319e-05, "loss": 0.0632, "step": 14880 }, { "grad_norm": 0.5822191834449768, "learning_rate": 9.501983590594821e-05, "loss": 0.0762, "step": 14890 }, { "grad_norm": 0.7310634255409241, "learning_rate": 9.501083987463455e-05, "loss": 0.0608, "step": 14900 }, { "grad_norm": 0.7937092185020447, "learning_rate": 9.500183615219942e-05, "loss": 0.0919, "step": 14910 }, { "grad_norm": 0.5976527333259583, "learning_rate": 9.49928247401813e-05, "loss": 0.0632, "step": 14920 }, { "grad_norm": 0.483794629573822, "learning_rate": 9.498380564011997e-05, "loss": 0.0812, "step": 14930 }, { "grad_norm": 0.5632884502410889, "learning_rate": 9.497477885355656e-05, "loss": 0.0712, "step": 14940 }, { "grad_norm": 0.5365797877311707, "learning_rate": 9.496574438203353e-05, "loss": 0.0848, "step": 14950 }, { "grad_norm": 0.610731840133667, "learning_rate": 9.495670222709459e-05, "loss": 0.0874, "step": 14960 }, { "grad_norm": 0.6000179648399353, "learning_rate": 9.494765239028483e-05, "loss": 0.0725, "step": 14970 }, { "grad_norm": 0.7736236453056335, "learning_rate": 9.493859487315057e-05, "loss": 0.0814, "step": 14980 }, { "grad_norm": 0.674365758895874, "learning_rate": 9.492952967723953e-05, "loss": 0.0626, "step": 14990 }, { "grad_norm": 0.7520804405212402, "learning_rate": 9.492045680410068e-05, "loss": 0.0838, "step": 15000 }, { "grad_norm": 0.4807124435901642, "learning_rate": 9.491137625528436e-05, "loss": 0.0657, "step": 15010 }, { "grad_norm": 0.6397553086280823, "learning_rate": 9.490228803234215e-05, "loss": 0.0593, "step": 15020 }, { "grad_norm": 0.7020514011383057, "learning_rate": 9.489319213682701e-05, "loss": 0.068, "step": 15030 }, { "grad_norm": 0.5716295838356018, "learning_rate": 9.488408857029316e-05, "loss": 0.0605, "step": 15040 }, { "grad_norm": 0.576718270778656, "learning_rate": 9.487497733429616e-05, "loss": 0.0626, "step": 15050 }, { "grad_norm": 0.6089158058166504, "learning_rate": 9.486585843039286e-05, "loss": 0.0717, "step": 15060 }, { "grad_norm": 0.843454122543335, "learning_rate": 9.485673186014143e-05, "loss": 0.0722, "step": 15070 }, { "grad_norm": 0.5496047735214233, "learning_rate": 9.484759762510137e-05, "loss": 0.0824, "step": 15080 }, { "grad_norm": 0.7371719479560852, "learning_rate": 9.483845572683346e-05, "loss": 0.0651, "step": 15090 }, { "grad_norm": 0.6968490481376648, "learning_rate": 9.48293061668998e-05, "loss": 0.0744, "step": 15100 }, { "grad_norm": 0.6263359189033508, "learning_rate": 9.48201489468638e-05, "loss": 0.0663, "step": 15110 }, { "grad_norm": 0.5693038105964661, "learning_rate": 9.481098406829016e-05, "loss": 0.0832, "step": 15120 }, { "grad_norm": 0.5542994737625122, "learning_rate": 9.480181153274495e-05, "loss": 0.086, "step": 15130 }, { "grad_norm": 0.5418812036514282, "learning_rate": 9.479263134179548e-05, "loss": 0.0705, "step": 15140 }, { "grad_norm": 0.7585984468460083, "learning_rate": 9.478344349701039e-05, "loss": 0.0739, "step": 15150 }, { "grad_norm": 0.641042947769165, "learning_rate": 9.477424799995964e-05, "loss": 0.094, "step": 15160 }, { "grad_norm": 0.8720000982284546, "learning_rate": 9.476504485221448e-05, "loss": 0.0961, "step": 15170 }, { "grad_norm": 1.2799240350723267, "learning_rate": 9.475583405534748e-05, "loss": 0.0861, "step": 15180 }, { "grad_norm": 0.6437321901321411, "learning_rate": 9.474661561093251e-05, "loss": 0.1066, "step": 15190 }, { "grad_norm": 0.5967101454734802, "learning_rate": 9.473738952054478e-05, "loss": 0.0947, "step": 15200 }, { "grad_norm": 1.0946258306503296, "learning_rate": 9.472815578576073e-05, "loss": 0.0996, "step": 15210 }, { "grad_norm": 0.7989858388900757, "learning_rate": 9.471891440815817e-05, "loss": 0.073, "step": 15220 }, { "grad_norm": 0.5913164615631104, "learning_rate": 9.470966538931621e-05, "loss": 0.104, "step": 15230 }, { "grad_norm": 0.7729251384735107, "learning_rate": 9.470040873081525e-05, "loss": 0.0711, "step": 15240 }, { "grad_norm": 0.7098260521888733, "learning_rate": 9.469114443423698e-05, "loss": 0.0895, "step": 15250 }, { "grad_norm": 0.8419852256774902, "learning_rate": 9.468187250116445e-05, "loss": 0.0718, "step": 15260 }, { "grad_norm": 0.5897564888000488, "learning_rate": 9.467259293318197e-05, "loss": 0.0691, "step": 15270 }, { "grad_norm": 0.8218228220939636, "learning_rate": 9.466330573187514e-05, "loss": 0.0883, "step": 15280 }, { "grad_norm": 0.9415950179100037, "learning_rate": 9.46540108988309e-05, "loss": 0.0721, "step": 15290 }, { "grad_norm": 0.8096191883087158, "learning_rate": 9.46447084356375e-05, "loss": 0.0764, "step": 15300 }, { "grad_norm": 0.6247458457946777, "learning_rate": 9.463539834388447e-05, "loss": 0.0774, "step": 15310 }, { "grad_norm": 0.5412017703056335, "learning_rate": 9.462608062516263e-05, "loss": 0.0652, "step": 15320 }, { "grad_norm": 0.4560605585575104, "learning_rate": 9.461675528106413e-05, "loss": 0.0817, "step": 15330 }, { "grad_norm": 0.7938639521598816, "learning_rate": 9.460742231318244e-05, "loss": 0.0708, "step": 15340 }, { "grad_norm": 0.515921413898468, "learning_rate": 9.459808172311229e-05, "loss": 0.0861, "step": 15350 }, { "grad_norm": 0.6725958585739136, "learning_rate": 9.458873351244972e-05, "loss": 0.0893, "step": 15360 }, { "grad_norm": 0.898387610912323, "learning_rate": 9.457937768279211e-05, "loss": 0.0974, "step": 15370 }, { "grad_norm": 0.9113718271255493, "learning_rate": 9.45700142357381e-05, "loss": 0.0822, "step": 15380 }, { "grad_norm": 0.5770963430404663, "learning_rate": 9.456064317288765e-05, "loss": 0.068, "step": 15390 }, { "grad_norm": 0.5114596486091614, "learning_rate": 9.455126449584201e-05, "loss": 0.0522, "step": 15400 }, { "grad_norm": 0.5929690003395081, "learning_rate": 9.454187820620375e-05, "loss": 0.0771, "step": 15410 }, { "grad_norm": 0.5943490862846375, "learning_rate": 9.453248430557673e-05, "loss": 0.1101, "step": 15420 }, { "grad_norm": 0.6514558792114258, "learning_rate": 9.452308279556611e-05, "loss": 0.0888, "step": 15430 }, { "grad_norm": 0.7042309045791626, "learning_rate": 9.451367367777835e-05, "loss": 0.0712, "step": 15440 }, { "grad_norm": 0.6436219811439514, "learning_rate": 9.450425695382122e-05, "loss": 0.0735, "step": 15450 }, { "grad_norm": 0.6407597064971924, "learning_rate": 9.449483262530375e-05, "loss": 0.0767, "step": 15460 }, { "grad_norm": 0.9511303901672363, "learning_rate": 9.448540069383633e-05, "loss": 0.0781, "step": 15470 }, { "grad_norm": 0.9550337791442871, "learning_rate": 9.447596116103061e-05, "loss": 0.0834, "step": 15480 }, { "grad_norm": 0.7014877796173096, "learning_rate": 9.446651402849955e-05, "loss": 0.0707, "step": 15490 }, { "grad_norm": 5.889959335327148, "learning_rate": 9.44570592978574e-05, "loss": 0.1003, "step": 15500 }, { "grad_norm": 0.6604233384132385, "learning_rate": 9.444759697071972e-05, "loss": 0.0855, "step": 15510 }, { "grad_norm": 0.6364761590957642, "learning_rate": 9.443812704870336e-05, "loss": 0.0871, "step": 15520 }, { "grad_norm": 0.5870062708854675, "learning_rate": 9.442864953342649e-05, "loss": 0.0843, "step": 15530 }, { "grad_norm": 0.907164990901947, "learning_rate": 9.441916442650852e-05, "loss": 0.0813, "step": 15540 }, { "grad_norm": 0.6608607769012451, "learning_rate": 9.440967172957023e-05, "loss": 0.0714, "step": 15550 }, { "grad_norm": 0.6217159032821655, "learning_rate": 9.440017144423364e-05, "loss": 0.0851, "step": 15560 }, { "grad_norm": 0.6908700466156006, "learning_rate": 9.439066357212209e-05, "loss": 0.0683, "step": 15570 }, { "grad_norm": 0.7665643692016602, "learning_rate": 9.438114811486022e-05, "loss": 0.0733, "step": 15580 }, { "grad_norm": 0.8209888339042664, "learning_rate": 9.4371625074074e-05, "loss": 0.091, "step": 15590 }, { "grad_norm": 0.7247982025146484, "learning_rate": 9.436209445139059e-05, "loss": 0.094, "step": 15600 }, { "grad_norm": 0.6202503442764282, "learning_rate": 9.435255624843855e-05, "loss": 0.0818, "step": 15610 }, { "grad_norm": 0.7439670562744141, "learning_rate": 9.43430104668477e-05, "loss": 0.0602, "step": 15620 }, { "grad_norm": 0.7113103866577148, "learning_rate": 9.433345710824914e-05, "loss": 0.0583, "step": 15630 }, { "grad_norm": 0.7470187544822693, "learning_rate": 9.432389617427529e-05, "loss": 0.0572, "step": 15640 }, { "grad_norm": 0.6606983542442322, "learning_rate": 9.431432766655984e-05, "loss": 0.0657, "step": 15650 }, { "grad_norm": 0.6439732909202576, "learning_rate": 9.430475158673778e-05, "loss": 0.066, "step": 15660 }, { "grad_norm": 0.6940078139305115, "learning_rate": 9.429516793644542e-05, "loss": 0.0699, "step": 15670 }, { "grad_norm": 0.7818444967269897, "learning_rate": 9.428557671732034e-05, "loss": 0.0746, "step": 15680 }, { "grad_norm": 0.6428760886192322, "learning_rate": 9.42759779310014e-05, "loss": 0.0822, "step": 15690 }, { "grad_norm": 0.7076892852783203, "learning_rate": 9.426637157912879e-05, "loss": 0.0629, "step": 15700 }, { "grad_norm": 0.6533300280570984, "learning_rate": 9.425675766334397e-05, "loss": 0.0748, "step": 15710 }, { "grad_norm": 0.809769868850708, "learning_rate": 9.424713618528968e-05, "loss": 0.0942, "step": 15720 }, { "grad_norm": 0.6544210314750671, "learning_rate": 9.423750714661e-05, "loss": 0.0629, "step": 15730 }, { "grad_norm": 0.49316921830177307, "learning_rate": 9.422787054895022e-05, "loss": 0.0781, "step": 15740 }, { "grad_norm": 0.56609046459198, "learning_rate": 9.4218226393957e-05, "loss": 0.077, "step": 15750 }, { "grad_norm": 0.603759765625, "learning_rate": 9.420857468327828e-05, "loss": 0.0693, "step": 15760 }, { "grad_norm": 0.520902693271637, "learning_rate": 9.419891541856323e-05, "loss": 0.06, "step": 15770 }, { "grad_norm": 0.721537172794342, "learning_rate": 9.41892486014624e-05, "loss": 0.0786, "step": 15780 }, { "grad_norm": 0.9180787801742554, "learning_rate": 9.417957423362756e-05, "loss": 0.0929, "step": 15790 }, { "grad_norm": 0.6712152361869812, "learning_rate": 9.416989231671178e-05, "loss": 0.0959, "step": 15800 }, { "grad_norm": 0.5709459185600281, "learning_rate": 9.416020285236946e-05, "loss": 0.0995, "step": 15810 }, { "grad_norm": 0.5895217061042786, "learning_rate": 9.415050584225626e-05, "loss": 0.0677, "step": 15820 }, { "grad_norm": 0.7203216552734375, "learning_rate": 9.414080128802914e-05, "loss": 0.0951, "step": 15830 }, { "grad_norm": 0.4307329058647156, "learning_rate": 9.413108919134632e-05, "loss": 0.0976, "step": 15840 }, { "grad_norm": 0.6682922840118408, "learning_rate": 9.412136955386734e-05, "loss": 0.0669, "step": 15850 }, { "grad_norm": 0.591634213924408, "learning_rate": 9.411164237725303e-05, "loss": 0.0692, "step": 15860 }, { "grad_norm": 0.533290684223175, "learning_rate": 9.41019076631655e-05, "loss": 0.0679, "step": 15870 }, { "grad_norm": 0.5191143751144409, "learning_rate": 9.409216541326815e-05, "loss": 0.0799, "step": 15880 }, { "grad_norm": 0.49654752016067505, "learning_rate": 9.408241562922564e-05, "loss": 0.0757, "step": 15890 }, { "grad_norm": 0.787734866142273, "learning_rate": 9.407265831270395e-05, "loss": 0.0628, "step": 15900 }, { "grad_norm": 0.746163547039032, "learning_rate": 9.406289346537035e-05, "loss": 0.0673, "step": 15910 }, { "grad_norm": 0.6230735182762146, "learning_rate": 9.405312108889339e-05, "loss": 0.0676, "step": 15920 }, { "grad_norm": 0.6117263436317444, "learning_rate": 9.404334118494288e-05, "loss": 0.06, "step": 15930 }, { "grad_norm": 0.6633436679840088, "learning_rate": 9.403355375518995e-05, "loss": 0.069, "step": 15940 }, { "grad_norm": 0.5508829355239868, "learning_rate": 9.4023758801307e-05, "loss": 0.0805, "step": 15950 }, { "grad_norm": 0.5768089294433594, "learning_rate": 9.401395632496774e-05, "loss": 0.0585, "step": 15960 }, { "grad_norm": 0.5603706240653992, "learning_rate": 9.400414632784711e-05, "loss": 0.0538, "step": 15970 }, { "grad_norm": 0.5833896398544312, "learning_rate": 9.39943288116214e-05, "loss": 0.0615, "step": 15980 }, { "grad_norm": 0.6535240411758423, "learning_rate": 9.398450377796815e-05, "loss": 0.0607, "step": 15990 }, { "grad_norm": 0.6697086095809937, "learning_rate": 9.397467122856616e-05, "loss": 0.0594, "step": 16000 }, { "grad_norm": 0.7242302894592285, "learning_rate": 9.396483116509558e-05, "loss": 0.0621, "step": 16010 }, { "grad_norm": 0.6036511063575745, "learning_rate": 9.39549835892378e-05, "loss": 0.0639, "step": 16020 }, { "grad_norm": 0.6160491704940796, "learning_rate": 9.39451285026755e-05, "loss": 0.0783, "step": 16030 }, { "grad_norm": 0.5291743278503418, "learning_rate": 9.393526590709262e-05, "loss": 0.0638, "step": 16040 }, { "grad_norm": 0.5942814350128174, "learning_rate": 9.392539580417444e-05, "loss": 0.0578, "step": 16050 }, { "grad_norm": 0.8881676197052002, "learning_rate": 9.391551819560747e-05, "loss": 0.0668, "step": 16060 }, { "grad_norm": 0.5578470230102539, "learning_rate": 9.390563308307955e-05, "loss": 0.0696, "step": 16070 }, { "grad_norm": 0.5564722418785095, "learning_rate": 9.389574046827974e-05, "loss": 0.0822, "step": 16080 }, { "grad_norm": 0.45694565773010254, "learning_rate": 9.388584035289845e-05, "loss": 0.0765, "step": 16090 }, { "grad_norm": 0.6778779029846191, "learning_rate": 9.387593273862732e-05, "loss": 0.0688, "step": 16100 }, { "grad_norm": 0.6430210471153259, "learning_rate": 9.386601762715929e-05, "loss": 0.0701, "step": 16110 }, { "grad_norm": 0.6544694900512695, "learning_rate": 9.38560950201886e-05, "loss": 0.0643, "step": 16120 }, { "grad_norm": 0.7904616594314575, "learning_rate": 9.384616491941071e-05, "loss": 0.0798, "step": 16130 }, { "grad_norm": 0.6747731566429138, "learning_rate": 9.383622732652245e-05, "loss": 0.0699, "step": 16140 }, { "grad_norm": 0.9077541828155518, "learning_rate": 9.382628224322187e-05, "loss": 0.0906, "step": 16150 }, { "grad_norm": 0.7514163255691528, "learning_rate": 9.381632967120829e-05, "loss": 0.0856, "step": 16160 }, { "grad_norm": 0.5805565714836121, "learning_rate": 9.380636961218235e-05, "loss": 0.0798, "step": 16170 }, { "grad_norm": 0.8805339336395264, "learning_rate": 9.379640206784597e-05, "loss": 0.0627, "step": 16180 }, { "grad_norm": 0.7633301615715027, "learning_rate": 9.378642703990229e-05, "loss": 0.0914, "step": 16190 }, { "grad_norm": 0.536226212978363, "learning_rate": 9.37764445300558e-05, "loss": 0.0633, "step": 16200 }, { "grad_norm": 0.4995524287223816, "learning_rate": 9.376645454001222e-05, "loss": 0.1034, "step": 16210 }, { "grad_norm": 0.7895928025245667, "learning_rate": 9.375645707147858e-05, "loss": 0.073, "step": 16220 }, { "grad_norm": 0.8142479062080383, "learning_rate": 9.374645212616316e-05, "loss": 0.085, "step": 16230 }, { "grad_norm": 0.6997902393341064, "learning_rate": 9.373643970577555e-05, "loss": 0.0637, "step": 16240 }, { "grad_norm": 0.7187997102737427, "learning_rate": 9.372641981202659e-05, "loss": 0.0785, "step": 16250 }, { "grad_norm": 0.7481949925422668, "learning_rate": 9.37163924466284e-05, "loss": 0.0915, "step": 16260 }, { "grad_norm": 0.720382034778595, "learning_rate": 9.370635761129438e-05, "loss": 0.1132, "step": 16270 }, { "grad_norm": 0.9123202562332153, "learning_rate": 9.36963153077392e-05, "loss": 0.0736, "step": 16280 }, { "grad_norm": 0.584080159664154, "learning_rate": 9.368626553767888e-05, "loss": 0.0582, "step": 16290 }, { "grad_norm": 0.8777958154678345, "learning_rate": 9.367620830283057e-05, "loss": 0.0791, "step": 16300 }, { "grad_norm": 0.6876343488693237, "learning_rate": 9.366614360491281e-05, "loss": 0.0999, "step": 16310 }, { "grad_norm": 0.5131940245628357, "learning_rate": 9.365607144564539e-05, "loss": 0.0669, "step": 16320 }, { "grad_norm": 0.5486665368080139, "learning_rate": 9.364599182674934e-05, "loss": 0.0802, "step": 16330 }, { "grad_norm": 0.6957598924636841, "learning_rate": 9.3635904749947e-05, "loss": 0.0683, "step": 16340 }, { "grad_norm": 0.7205453515052795, "learning_rate": 9.362581021696202e-05, "loss": 0.0621, "step": 16350 }, { "grad_norm": 0.5596363544464111, "learning_rate": 9.361570822951921e-05, "loss": 0.0657, "step": 16360 }, { "grad_norm": 0.682066798210144, "learning_rate": 9.360559878934476e-05, "loss": 0.0554, "step": 16370 }, { "grad_norm": 0.5809022188186646, "learning_rate": 9.359548189816611e-05, "loss": 0.0782, "step": 16380 }, { "grad_norm": 0.4947965145111084, "learning_rate": 9.358535755771193e-05, "loss": 0.0572, "step": 16390 }, { "grad_norm": 0.6175491213798523, "learning_rate": 9.357522576971221e-05, "loss": 0.0664, "step": 16400 }, { "grad_norm": 0.624290943145752, "learning_rate": 9.356508653589819e-05, "loss": 0.0608, "step": 16410 }, { "grad_norm": 0.8851645588874817, "learning_rate": 9.355493985800237e-05, "loss": 0.0634, "step": 16420 }, { "grad_norm": 0.5072843432426453, "learning_rate": 9.354478573775857e-05, "loss": 0.0785, "step": 16430 }, { "grad_norm": 0.5967523455619812, "learning_rate": 9.353462417690186e-05, "loss": 0.0683, "step": 16440 }, { "grad_norm": 0.5610153675079346, "learning_rate": 9.352445517716853e-05, "loss": 0.0688, "step": 16450 }, { "grad_norm": 0.6705263257026672, "learning_rate": 9.351427874029621e-05, "loss": 0.069, "step": 16460 }, { "grad_norm": 0.6496067047119141, "learning_rate": 9.350409486802379e-05, "loss": 0.0897, "step": 16470 }, { "grad_norm": 0.6841572523117065, "learning_rate": 9.349390356209138e-05, "loss": 0.0698, "step": 16480 }, { "grad_norm": 0.6002643704414368, "learning_rate": 9.348370482424042e-05, "loss": 0.0626, "step": 16490 }, { "grad_norm": 0.8101145625114441, "learning_rate": 9.347349865621357e-05, "loss": 0.0783, "step": 16500 }, { "grad_norm": 0.5773817896842957, "learning_rate": 9.346328505975481e-05, "loss": 0.0688, "step": 16510 }, { "grad_norm": 0.5705987811088562, "learning_rate": 9.345306403660936e-05, "loss": 0.0625, "step": 16520 }, { "grad_norm": 0.8034189939498901, "learning_rate": 9.344283558852371e-05, "loss": 0.0682, "step": 16530 }, { "grad_norm": 0.7335330843925476, "learning_rate": 9.343259971724563e-05, "loss": 0.0687, "step": 16540 }, { "grad_norm": 0.42320260405540466, "learning_rate": 9.342235642452413e-05, "loss": 0.0646, "step": 16550 }, { "grad_norm": 0.7819838523864746, "learning_rate": 9.341210571210954e-05, "loss": 0.1002, "step": 16560 }, { "grad_norm": 0.5251914262771606, "learning_rate": 9.340184758175338e-05, "loss": 0.0649, "step": 16570 }, { "grad_norm": 0.5240599513053894, "learning_rate": 9.339158203520854e-05, "loss": 0.0625, "step": 16580 }, { "grad_norm": 0.5401110649108887, "learning_rate": 9.338130907422908e-05, "loss": 0.0789, "step": 16590 }, { "grad_norm": 0.7780466675758362, "learning_rate": 9.337102870057037e-05, "loss": 0.0785, "step": 16600 }, { "grad_norm": 0.5013154745101929, "learning_rate": 9.336074091598907e-05, "loss": 0.0591, "step": 16610 }, { "grad_norm": 0.7003335356712341, "learning_rate": 9.335044572224306e-05, "loss": 0.0728, "step": 16620 }, { "grad_norm": 0.702305793762207, "learning_rate": 9.334014312109151e-05, "loss": 0.067, "step": 16630 }, { "grad_norm": 0.6025710701942444, "learning_rate": 9.332983311429486e-05, "loss": 0.057, "step": 16640 }, { "grad_norm": 0.48436611890792847, "learning_rate": 9.33195157036148e-05, "loss": 0.0681, "step": 16650 }, { "grad_norm": 0.6232336163520813, "learning_rate": 9.330919089081432e-05, "loss": 0.0897, "step": 16660 }, { "grad_norm": 0.8306937217712402, "learning_rate": 9.32988586776576e-05, "loss": 0.0676, "step": 16670 }, { "grad_norm": 0.692804217338562, "learning_rate": 9.328851906591016e-05, "loss": 0.0681, "step": 16680 }, { "grad_norm": 0.68687504529953, "learning_rate": 9.327817205733875e-05, "loss": 0.0902, "step": 16690 }, { "grad_norm": 0.6984761357307434, "learning_rate": 9.326781765371142e-05, "loss": 0.0613, "step": 16700 }, { "grad_norm": 0.498328298330307, "learning_rate": 9.325745585679741e-05, "loss": 0.0691, "step": 16710 }, { "grad_norm": 0.5404256582260132, "learning_rate": 9.32470866683673e-05, "loss": 0.0669, "step": 16720 }, { "grad_norm": 0.5658946633338928, "learning_rate": 9.323671009019288e-05, "loss": 0.048, "step": 16730 }, { "grad_norm": 0.6267743706703186, "learning_rate": 9.322632612404725e-05, "loss": 0.0648, "step": 16740 }, { "grad_norm": 0.6957098245620728, "learning_rate": 9.321593477170471e-05, "loss": 0.0825, "step": 16750 }, { "grad_norm": 0.6981226205825806, "learning_rate": 9.320553603494088e-05, "loss": 0.0652, "step": 16760 }, { "grad_norm": 0.874060869216919, "learning_rate": 9.319512991553261e-05, "loss": 0.0712, "step": 16770 }, { "grad_norm": 0.6672017574310303, "learning_rate": 9.318471641525803e-05, "loss": 0.0733, "step": 16780 }, { "grad_norm": 0.6838167905807495, "learning_rate": 9.317429553589652e-05, "loss": 0.063, "step": 16790 }, { "grad_norm": 0.41493549942970276, "learning_rate": 9.316386727922873e-05, "loss": 0.0597, "step": 16800 }, { "grad_norm": 0.7923299670219421, "learning_rate": 9.315343164703656e-05, "loss": 0.0697, "step": 16810 }, { "grad_norm": 0.7194953560829163, "learning_rate": 9.314298864110316e-05, "loss": 0.0707, "step": 16820 }, { "grad_norm": 0.6271785497665405, "learning_rate": 9.313253826321295e-05, "loss": 0.068, "step": 16830 }, { "grad_norm": 0.6728053689002991, "learning_rate": 9.312208051515165e-05, "loss": 0.0658, "step": 16840 }, { "grad_norm": 0.7498440146446228, "learning_rate": 9.311161539870618e-05, "loss": 0.0649, "step": 16850 }, { "grad_norm": 0.738482654094696, "learning_rate": 9.310114291566474e-05, "loss": 0.0962, "step": 16860 }, { "grad_norm": 0.7637571692466736, "learning_rate": 9.309066306781679e-05, "loss": 0.0822, "step": 16870 }, { "grad_norm": 0.6074163317680359, "learning_rate": 9.308017585695306e-05, "loss": 0.0769, "step": 16880 }, { "grad_norm": 0.906410276889801, "learning_rate": 9.306968128486552e-05, "loss": 0.081, "step": 16890 }, { "grad_norm": 0.5658301711082458, "learning_rate": 9.30591793533474e-05, "loss": 0.0609, "step": 16900 }, { "grad_norm": 0.602830171585083, "learning_rate": 9.304867006419321e-05, "loss": 0.0647, "step": 16910 }, { "grad_norm": 0.6115087866783142, "learning_rate": 9.303815341919868e-05, "loss": 0.067, "step": 16920 }, { "grad_norm": 0.5763307809829712, "learning_rate": 9.302762942016084e-05, "loss": 0.0709, "step": 16930 }, { "grad_norm": 0.6263272166252136, "learning_rate": 9.301709806887792e-05, "loss": 0.0605, "step": 16940 }, { "grad_norm": 1.5651391744613647, "learning_rate": 9.300655936714948e-05, "loss": 0.0697, "step": 16950 }, { "grad_norm": 0.7158195972442627, "learning_rate": 9.299601331677627e-05, "loss": 0.06, "step": 16960 }, { "grad_norm": 0.8227112293243408, "learning_rate": 9.298545991956033e-05, "loss": 0.0731, "step": 16970 }, { "grad_norm": 0.9584242105484009, "learning_rate": 9.297489917730493e-05, "loss": 0.0802, "step": 16980 }, { "grad_norm": 0.6843316555023193, "learning_rate": 9.296433109181464e-05, "loss": 0.08, "step": 16990 }, { "grad_norm": 0.6386792659759521, "learning_rate": 9.295375566489523e-05, "loss": 0.0864, "step": 17000 }, { "grad_norm": 0.5148645639419556, "learning_rate": 9.294317289835379e-05, "loss": 0.0756, "step": 17010 }, { "grad_norm": 0.6730923056602478, "learning_rate": 9.293258279399859e-05, "loss": 0.0751, "step": 17020 }, { "grad_norm": 0.6581635475158691, "learning_rate": 9.292198535363919e-05, "loss": 0.0678, "step": 17030 }, { "grad_norm": 0.7924745082855225, "learning_rate": 9.291138057908641e-05, "loss": 0.0704, "step": 17040 }, { "grad_norm": 0.5201916098594666, "learning_rate": 9.290076847215234e-05, "loss": 0.0709, "step": 17050 }, { "grad_norm": 0.557521402835846, "learning_rate": 9.289014903465025e-05, "loss": 0.0772, "step": 17060 }, { "grad_norm": 0.6296219825744629, "learning_rate": 9.287952226839475e-05, "loss": 0.0794, "step": 17070 }, { "grad_norm": 0.5344809889793396, "learning_rate": 9.286888817520164e-05, "loss": 0.0773, "step": 17080 }, { "grad_norm": 0.8383042812347412, "learning_rate": 9.285824675688803e-05, "loss": 0.0841, "step": 17090 }, { "grad_norm": 0.6311007142066956, "learning_rate": 9.28475980152722e-05, "loss": 0.0694, "step": 17100 }, { "grad_norm": 0.6467316150665283, "learning_rate": 9.283694195217379e-05, "loss": 0.0865, "step": 17110 }, { "grad_norm": 0.7070531845092773, "learning_rate": 9.282627856941356e-05, "loss": 0.0924, "step": 17120 }, { "grad_norm": 0.7749756574630737, "learning_rate": 9.281560786881363e-05, "loss": 0.1206, "step": 17130 }, { "grad_norm": 0.6749516129493713, "learning_rate": 9.280492985219733e-05, "loss": 0.084, "step": 17140 }, { "grad_norm": 0.5702797174453735, "learning_rate": 9.279424452138924e-05, "loss": 0.0821, "step": 17150 }, { "grad_norm": 0.6244213581085205, "learning_rate": 9.278355187821517e-05, "loss": 0.0885, "step": 17160 }, { "grad_norm": 0.7253963351249695, "learning_rate": 9.277285192450224e-05, "loss": 0.0842, "step": 17170 }, { "grad_norm": 0.608838677406311, "learning_rate": 9.276214466207875e-05, "loss": 0.0632, "step": 17180 }, { "grad_norm": 0.4207419753074646, "learning_rate": 9.275143009277427e-05, "loss": 0.0543, "step": 17190 }, { "grad_norm": 0.7396284937858582, "learning_rate": 9.274070821841964e-05, "loss": 0.0634, "step": 17200 }, { "grad_norm": 0.5695433020591736, "learning_rate": 9.272997904084696e-05, "loss": 0.0705, "step": 17210 }, { "grad_norm": 0.7633906006813049, "learning_rate": 9.271924256188951e-05, "loss": 0.0636, "step": 17220 }, { "grad_norm": 0.5102940797805786, "learning_rate": 9.270849878338189e-05, "loss": 0.0645, "step": 17230 }, { "grad_norm": 0.6745217442512512, "learning_rate": 9.269774770715991e-05, "loss": 0.085, "step": 17240 }, { "grad_norm": 1.1384609937667847, "learning_rate": 9.268698933506061e-05, "loss": 0.0778, "step": 17250 }, { "grad_norm": 0.6832176446914673, "learning_rate": 9.267622366892235e-05, "loss": 0.0639, "step": 17260 }, { "grad_norm": 0.39319807291030884, "learning_rate": 9.266545071058465e-05, "loss": 0.0712, "step": 17270 }, { "grad_norm": 0.6669896841049194, "learning_rate": 9.265467046188833e-05, "loss": 0.0783, "step": 17280 }, { "grad_norm": 0.5303493142127991, "learning_rate": 9.264388292467543e-05, "loss": 0.0734, "step": 17290 }, { "grad_norm": 0.5372130274772644, "learning_rate": 9.263308810078926e-05, "loss": 0.08, "step": 17300 }, { "grad_norm": 0.7770669460296631, "learning_rate": 9.262228599207434e-05, "loss": 0.0643, "step": 17310 }, { "grad_norm": 0.5164933204650879, "learning_rate": 9.261147660037647e-05, "loss": 0.0692, "step": 17320 }, { "grad_norm": 0.6354234218597412, "learning_rate": 9.26006599275427e-05, "loss": 0.0628, "step": 17330 }, { "grad_norm": 0.658703088760376, "learning_rate": 9.258983597542124e-05, "loss": 0.0655, "step": 17340 }, { "grad_norm": 0.8940509557723999, "learning_rate": 9.257900474586167e-05, "loss": 0.068, "step": 17350 }, { "grad_norm": 0.6216449737548828, "learning_rate": 9.256816624071471e-05, "loss": 0.0608, "step": 17360 }, { "grad_norm": 0.5889434218406677, "learning_rate": 9.25573204618324e-05, "loss": 0.0601, "step": 17370 }, { "grad_norm": 0.41465872526168823, "learning_rate": 9.254646741106796e-05, "loss": 0.0643, "step": 17380 }, { "grad_norm": 0.7007591724395752, "learning_rate": 9.253560709027589e-05, "loss": 0.0602, "step": 17390 }, { "grad_norm": 0.6279572248458862, "learning_rate": 9.252473950131192e-05, "loss": 0.0673, "step": 17400 }, { "grad_norm": 1.330439805984497, "learning_rate": 9.251386464603302e-05, "loss": 0.0844, "step": 17410 }, { "grad_norm": 0.723328173160553, "learning_rate": 9.250298252629741e-05, "loss": 0.0746, "step": 17420 }, { "grad_norm": 0.5837368369102478, "learning_rate": 9.249209314396454e-05, "loss": 0.0666, "step": 17430 }, { "grad_norm": 0.8119161128997803, "learning_rate": 9.248119650089513e-05, "loss": 0.108, "step": 17440 }, { "grad_norm": 0.6762596964836121, "learning_rate": 9.247029259895108e-05, "loss": 0.0617, "step": 17450 }, { "grad_norm": 0.6120361685752869, "learning_rate": 9.24593814399956e-05, "loss": 0.0647, "step": 17460 }, { "grad_norm": 0.7686606049537659, "learning_rate": 9.244846302589309e-05, "loss": 0.0614, "step": 17470 }, { "grad_norm": 0.5257925987243652, "learning_rate": 9.243753735850923e-05, "loss": 0.0551, "step": 17480 }, { "grad_norm": 0.8428086042404175, "learning_rate": 9.24266044397109e-05, "loss": 0.0624, "step": 17490 }, { "grad_norm": 0.5789021253585815, "learning_rate": 9.241566427136624e-05, "loss": 0.087, "step": 17500 }, { "grad_norm": 0.7111076712608337, "learning_rate": 9.240471685534463e-05, "loss": 0.0674, "step": 17510 }, { "grad_norm": 0.5372388362884521, "learning_rate": 9.239376219351667e-05, "loss": 0.0688, "step": 17520 }, { "grad_norm": 0.8844457864761353, "learning_rate": 9.238280028775425e-05, "loss": 0.0753, "step": 17530 }, { "grad_norm": 0.6628541350364685, "learning_rate": 9.237183113993041e-05, "loss": 0.0757, "step": 17540 }, { "grad_norm": 0.7726335525512695, "learning_rate": 9.236085475191952e-05, "loss": 0.0568, "step": 17550 }, { "grad_norm": 0.8071607947349548, "learning_rate": 9.234987112559709e-05, "loss": 0.0865, "step": 17560 }, { "grad_norm": 0.5704105496406555, "learning_rate": 9.233888026283999e-05, "loss": 0.0755, "step": 17570 }, { "grad_norm": 0.6932694315910339, "learning_rate": 9.232788216552619e-05, "loss": 0.0682, "step": 17580 }, { "grad_norm": 0.8769665360450745, "learning_rate": 9.231687683553502e-05, "loss": 0.0643, "step": 17590 }, { "grad_norm": 0.5220070481300354, "learning_rate": 9.230586427474698e-05, "loss": 0.0507, "step": 17600 }, { "grad_norm": 0.4498666524887085, "learning_rate": 9.229484448504379e-05, "loss": 0.0701, "step": 17610 }, { "grad_norm": 0.9221429228782654, "learning_rate": 9.228381746830843e-05, "loss": 0.0694, "step": 17620 }, { "grad_norm": 2.3972034454345703, "learning_rate": 9.227278322642514e-05, "loss": 0.0667, "step": 17630 }, { "grad_norm": 0.7535490393638611, "learning_rate": 9.226174176127937e-05, "loss": 0.0721, "step": 17640 }, { "grad_norm": 0.6855373382568359, "learning_rate": 9.22506930747578e-05, "loss": 0.0693, "step": 17650 }, { "grad_norm": 0.634380042552948, "learning_rate": 9.223963716874831e-05, "loss": 0.0659, "step": 17660 }, { "grad_norm": 0.6997556090354919, "learning_rate": 9.222857404514012e-05, "loss": 0.0646, "step": 17670 }, { "grad_norm": 0.8580924272537231, "learning_rate": 9.221750370582355e-05, "loss": 0.0714, "step": 17680 }, { "grad_norm": 1.153288722038269, "learning_rate": 9.220642615269028e-05, "loss": 0.0529, "step": 17690 }, { "grad_norm": 0.5514541268348694, "learning_rate": 9.219534138763311e-05, "loss": 0.059, "step": 17700 }, { "grad_norm": 0.6099024415016174, "learning_rate": 9.218424941254613e-05, "loss": 0.0756, "step": 17710 }, { "grad_norm": 0.9077996611595154, "learning_rate": 9.217315022932468e-05, "loss": 0.0716, "step": 17720 }, { "grad_norm": 0.6251293420791626, "learning_rate": 9.216204383986528e-05, "loss": 0.0614, "step": 17730 }, { "grad_norm": 0.7469528317451477, "learning_rate": 9.215093024606574e-05, "loss": 0.0847, "step": 17740 }, { "grad_norm": 0.7456029653549194, "learning_rate": 9.213980944982506e-05, "loss": 0.0956, "step": 17750 }, { "grad_norm": 0.8042892813682556, "learning_rate": 9.212868145304346e-05, "loss": 0.0631, "step": 17760 }, { "grad_norm": 0.754037618637085, "learning_rate": 9.211754625762241e-05, "loss": 0.0976, "step": 17770 }, { "grad_norm": 0.6019042730331421, "learning_rate": 9.210640386546463e-05, "loss": 0.0583, "step": 17780 }, { "grad_norm": 0.6794756650924683, "learning_rate": 9.209525427847405e-05, "loss": 0.0623, "step": 17790 }, { "grad_norm": 0.6040306687355042, "learning_rate": 9.208409749855583e-05, "loss": 0.0723, "step": 17800 }, { "grad_norm": 0.8136261105537415, "learning_rate": 9.207293352761633e-05, "loss": 0.0617, "step": 17810 }, { "grad_norm": 0.5900750756263733, "learning_rate": 9.206176236756319e-05, "loss": 0.0652, "step": 17820 }, { "grad_norm": 0.9503437876701355, "learning_rate": 9.205058402030525e-05, "loss": 0.0787, "step": 17830 }, { "grad_norm": 0.7836813926696777, "learning_rate": 9.203939848775259e-05, "loss": 0.0935, "step": 17840 }, { "grad_norm": 0.8397489190101624, "learning_rate": 9.202820577181652e-05, "loss": 0.0609, "step": 17850 }, { "grad_norm": 0.6767406463623047, "learning_rate": 9.201700587440953e-05, "loss": 0.0702, "step": 17860 }, { "grad_norm": 0.730832040309906, "learning_rate": 9.200579879744544e-05, "loss": 0.072, "step": 17870 }, { "grad_norm": 0.623552680015564, "learning_rate": 9.199458454283918e-05, "loss": 0.0706, "step": 17880 }, { "grad_norm": 0.6712446212768555, "learning_rate": 9.198336311250697e-05, "loss": 0.076, "step": 17890 }, { "grad_norm": 0.7218803763389587, "learning_rate": 9.197213450836626e-05, "loss": 0.0624, "step": 17900 }, { "grad_norm": 1.0561237335205078, "learning_rate": 9.19608987323357e-05, "loss": 0.0633, "step": 17910 }, { "grad_norm": 0.949665904045105, "learning_rate": 9.194965578633517e-05, "loss": 0.0644, "step": 17920 }, { "grad_norm": 0.5194064974784851, "learning_rate": 9.193840567228582e-05, "loss": 0.0697, "step": 17930 }, { "grad_norm": 0.7479503750801086, "learning_rate": 9.192714839210994e-05, "loss": 0.0598, "step": 17940 }, { "grad_norm": 0.751947283744812, "learning_rate": 9.19158839477311e-05, "loss": 0.0792, "step": 17950 }, { "grad_norm": 0.6769455671310425, "learning_rate": 9.190461234107411e-05, "loss": 0.0719, "step": 17960 }, { "grad_norm": 0.6388373374938965, "learning_rate": 9.189333357406496e-05, "loss": 0.1028, "step": 17970 }, { "grad_norm": 0.517700731754303, "learning_rate": 9.188204764863089e-05, "loss": 0.0882, "step": 17980 }, { "grad_norm": 0.5410895347595215, "learning_rate": 9.187075456670033e-05, "loss": 0.084, "step": 17990 }, { "grad_norm": 0.6114312410354614, "learning_rate": 9.1859454330203e-05, "loss": 0.053, "step": 18000 }, { "grad_norm": 0.6185637712478638, "learning_rate": 9.18481469410698e-05, "loss": 0.0526, "step": 18010 }, { "grad_norm": 0.6108859777450562, "learning_rate": 9.183683240123281e-05, "loss": 0.0737, "step": 18020 }, { "grad_norm": 0.8671708703041077, "learning_rate": 9.182551071262541e-05, "loss": 0.0676, "step": 18030 }, { "grad_norm": 0.7721058130264282, "learning_rate": 9.181418187718218e-05, "loss": 0.0781, "step": 18040 }, { "grad_norm": 0.4796210825443268, "learning_rate": 9.180284589683888e-05, "loss": 0.0862, "step": 18050 }, { "grad_norm": 0.45511487126350403, "learning_rate": 9.17915027735325e-05, "loss": 0.0726, "step": 18060 }, { "grad_norm": 0.5704712271690369, "learning_rate": 9.178015250920133e-05, "loss": 0.0728, "step": 18070 }, { "grad_norm": 0.3987959623336792, "learning_rate": 9.176879510578477e-05, "loss": 0.0655, "step": 18080 }, { "grad_norm": 0.5551433563232422, "learning_rate": 9.17574305652235e-05, "loss": 0.0598, "step": 18090 }, { "grad_norm": 0.48342302441596985, "learning_rate": 9.174605888945942e-05, "loss": 0.0611, "step": 18100 }, { "grad_norm": 0.6893698573112488, "learning_rate": 9.173468008043564e-05, "loss": 0.0892, "step": 18110 }, { "grad_norm": 0.5339382886886597, "learning_rate": 9.172329414009648e-05, "loss": 0.0653, "step": 18120 }, { "grad_norm": 0.45560041069984436, "learning_rate": 9.171190107038747e-05, "loss": 0.0629, "step": 18130 }, { "grad_norm": 0.7123005390167236, "learning_rate": 9.170050087325541e-05, "loss": 0.0641, "step": 18140 }, { "grad_norm": 0.6869412064552307, "learning_rate": 9.168909355064824e-05, "loss": 0.0728, "step": 18150 }, { "grad_norm": 0.5181896686553955, "learning_rate": 9.167767910451519e-05, "loss": 0.0741, "step": 18160 }, { "grad_norm": 0.7981396317481995, "learning_rate": 9.166625753680669e-05, "loss": 0.0927, "step": 18170 }, { "grad_norm": 0.5731774568557739, "learning_rate": 9.165482884947431e-05, "loss": 0.0591, "step": 18180 }, { "grad_norm": 0.7093178629875183, "learning_rate": 9.164339304447098e-05, "loss": 0.0831, "step": 18190 }, { "grad_norm": 0.6205394268035889, "learning_rate": 9.163195012375072e-05, "loss": 0.0783, "step": 18200 }, { "grad_norm": 0.8228815197944641, "learning_rate": 9.16205000892688e-05, "loss": 0.0711, "step": 18210 }, { "grad_norm": 0.6996854543685913, "learning_rate": 9.160904294298175e-05, "loss": 0.0667, "step": 18220 }, { "grad_norm": 0.5928246974945068, "learning_rate": 9.159757868684727e-05, "loss": 0.0701, "step": 18230 }, { "grad_norm": 0.7256582379341125, "learning_rate": 9.15861073228243e-05, "loss": 0.0728, "step": 18240 }, { "grad_norm": 0.8110372424125671, "learning_rate": 9.157462885287296e-05, "loss": 0.0902, "step": 18250 }, { "grad_norm": 0.9004707336425781, "learning_rate": 9.156314327895461e-05, "loss": 0.0577, "step": 18260 }, { "grad_norm": 0.6206316947937012, "learning_rate": 9.155165060303185e-05, "loss": 0.0627, "step": 18270 }, { "grad_norm": 0.6227447986602783, "learning_rate": 9.154015082706841e-05, "loss": 0.0638, "step": 18280 }, { "grad_norm": 0.6815391182899475, "learning_rate": 9.152864395302936e-05, "loss": 0.0884, "step": 18290 }, { "grad_norm": 0.46585139632225037, "learning_rate": 9.151712998288085e-05, "loss": 0.0627, "step": 18300 }, { "grad_norm": 0.7261255383491516, "learning_rate": 9.150560891859031e-05, "loss": 0.0618, "step": 18310 }, { "grad_norm": 0.6609123349189758, "learning_rate": 9.14940807621264e-05, "loss": 0.0663, "step": 18320 }, { "grad_norm": 0.6110312342643738, "learning_rate": 9.148254551545894e-05, "loss": 0.0671, "step": 18330 }, { "grad_norm": 0.520423412322998, "learning_rate": 9.147100318055901e-05, "loss": 0.0515, "step": 18340 }, { "grad_norm": 0.5421935319900513, "learning_rate": 9.145945375939888e-05, "loss": 0.0572, "step": 18350 }, { "grad_norm": 0.5092663764953613, "learning_rate": 9.144789725395203e-05, "loss": 0.0531, "step": 18360 }, { "grad_norm": 0.5769091248512268, "learning_rate": 9.14363336661931e-05, "loss": 0.0556, "step": 18370 }, { "grad_norm": 0.5437376499176025, "learning_rate": 9.142476299809806e-05, "loss": 0.0461, "step": 18380 }, { "grad_norm": 0.8120520710945129, "learning_rate": 9.1413185251644e-05, "loss": 0.0676, "step": 18390 }, { "grad_norm": 0.55147784948349, "learning_rate": 9.140160042880923e-05, "loss": 0.0711, "step": 18400 }, { "grad_norm": 1.0182024240493774, "learning_rate": 9.139000853157327e-05, "loss": 0.0757, "step": 18410 }, { "grad_norm": 0.5398492217063904, "learning_rate": 9.137840956191688e-05, "loss": 0.0812, "step": 18420 }, { "grad_norm": 0.41514596343040466, "learning_rate": 9.136680352182199e-05, "loss": 0.0804, "step": 18430 }, { "grad_norm": 0.5609413385391235, "learning_rate": 9.135519041327177e-05, "loss": 0.067, "step": 18440 }, { "grad_norm": 0.5717300176620483, "learning_rate": 9.134357023825058e-05, "loss": 0.0669, "step": 18450 }, { "grad_norm": 0.4658317565917969, "learning_rate": 9.133194299874398e-05, "loss": 0.062, "step": 18460 }, { "grad_norm": 0.5694889426231384, "learning_rate": 9.132030869673876e-05, "loss": 0.0635, "step": 18470 }, { "grad_norm": 0.5818838477134705, "learning_rate": 9.130866733422288e-05, "loss": 0.0951, "step": 18480 }, { "grad_norm": 0.5314889550209045, "learning_rate": 9.129701891318556e-05, "loss": 0.0598, "step": 18490 }, { "grad_norm": 0.44801655411720276, "learning_rate": 9.128536343561718e-05, "loss": 0.0523, "step": 18500 }, { "grad_norm": 0.4418388903141022, "learning_rate": 9.127370090350934e-05, "loss": 0.0712, "step": 18510 }, { "grad_norm": 0.617035448551178, "learning_rate": 9.126203131885487e-05, "loss": 0.0519, "step": 18520 }, { "grad_norm": 0.6926045417785645, "learning_rate": 9.125035468364775e-05, "loss": 0.057, "step": 18530 }, { "grad_norm": 0.4543774724006653, "learning_rate": 9.123867099988322e-05, "loss": 0.0777, "step": 18540 }, { "grad_norm": 0.4338540732860565, "learning_rate": 9.122698026955769e-05, "loss": 0.0639, "step": 18550 }, { "grad_norm": 0.6128022074699402, "learning_rate": 9.12152824946688e-05, "loss": 0.0794, "step": 18560 }, { "grad_norm": 0.563491702079773, "learning_rate": 9.120357767721538e-05, "loss": 0.0745, "step": 18570 }, { "grad_norm": 0.7469806671142578, "learning_rate": 9.119186581919745e-05, "loss": 0.0784, "step": 18580 }, { "grad_norm": 0.723395049571991, "learning_rate": 9.118014692261624e-05, "loss": 0.0609, "step": 18590 }, { "grad_norm": 0.5241453647613525, "learning_rate": 9.116842098947422e-05, "loss": 0.0918, "step": 18600 }, { "grad_norm": 0.7737153768539429, "learning_rate": 9.115668802177499e-05, "loss": 0.073, "step": 18610 }, { "grad_norm": 0.6155425310134888, "learning_rate": 9.114494802152342e-05, "loss": 0.0714, "step": 18620 }, { "grad_norm": 0.4265763759613037, "learning_rate": 9.113320099072555e-05, "loss": 0.0724, "step": 18630 }, { "grad_norm": 0.6456373929977417, "learning_rate": 9.112144693138864e-05, "loss": 0.0773, "step": 18640 }, { "grad_norm": 0.6607341766357422, "learning_rate": 9.110968584552111e-05, "loss": 0.0661, "step": 18650 }, { "grad_norm": 0.44007623195648193, "learning_rate": 9.109791773513264e-05, "loss": 0.0567, "step": 18660 }, { "grad_norm": 0.8596996665000916, "learning_rate": 9.108614260223403e-05, "loss": 0.0789, "step": 18670 }, { "grad_norm": 0.6906821727752686, "learning_rate": 9.107436044883738e-05, "loss": 0.0606, "step": 18680 }, { "grad_norm": 0.516899585723877, "learning_rate": 9.10625712769559e-05, "loss": 0.0565, "step": 18690 }, { "grad_norm": 0.452340692281723, "learning_rate": 9.105077508860406e-05, "loss": 0.0461, "step": 18700 }, { "grad_norm": 0.3900512754917145, "learning_rate": 9.103897188579751e-05, "loss": 0.0527, "step": 18710 }, { "grad_norm": 0.5770500302314758, "learning_rate": 9.102716167055308e-05, "loss": 0.0566, "step": 18720 }, { "grad_norm": 1.302507996559143, "learning_rate": 9.10153444448888e-05, "loss": 0.0794, "step": 18730 }, { "grad_norm": 0.5251820683479309, "learning_rate": 9.100352021082393e-05, "loss": 0.0869, "step": 18740 }, { "grad_norm": 0.871849775314331, "learning_rate": 9.099168897037891e-05, "loss": 0.08, "step": 18750 }, { "grad_norm": 0.8755308389663696, "learning_rate": 9.097985072557538e-05, "loss": 0.0649, "step": 18760 }, { "grad_norm": 0.5238458514213562, "learning_rate": 9.096800547843615e-05, "loss": 0.0672, "step": 18770 }, { "grad_norm": 0.7512543201446533, "learning_rate": 9.095615323098526e-05, "loss": 0.074, "step": 18780 }, { "grad_norm": 0.5159833431243896, "learning_rate": 9.094429398524795e-05, "loss": 0.0577, "step": 18790 }, { "grad_norm": 0.4983659088611603, "learning_rate": 9.093242774325061e-05, "loss": 0.077, "step": 18800 }, { "grad_norm": 0.7443787455558777, "learning_rate": 9.092055450702088e-05, "loss": 0.0751, "step": 18810 }, { "grad_norm": 0.828294038772583, "learning_rate": 9.090867427858756e-05, "loss": 0.0641, "step": 18820 }, { "grad_norm": 0.568107545375824, "learning_rate": 9.089678705998066e-05, "loss": 0.0583, "step": 18830 }, { "grad_norm": 0.8351099491119385, "learning_rate": 9.088489285323139e-05, "loss": 0.0718, "step": 18840 }, { "grad_norm": 0.9588242173194885, "learning_rate": 9.087299166037212e-05, "loss": 0.0624, "step": 18850 }, { "grad_norm": 0.5841375589370728, "learning_rate": 9.086108348343647e-05, "loss": 0.0475, "step": 18860 }, { "grad_norm": 0.6113753914833069, "learning_rate": 9.08491683244592e-05, "loss": 0.0576, "step": 18870 }, { "grad_norm": 0.9316718578338623, "learning_rate": 9.08372461854763e-05, "loss": 0.0698, "step": 18880 }, { "grad_norm": 0.6338059902191162, "learning_rate": 9.082531706852492e-05, "loss": 0.0679, "step": 18890 }, { "grad_norm": 0.555725634098053, "learning_rate": 9.081338097564342e-05, "loss": 0.0515, "step": 18900 }, { "grad_norm": 0.5537071824073792, "learning_rate": 9.080143790887137e-05, "loss": 0.0579, "step": 18910 }, { "grad_norm": 0.6454886198043823, "learning_rate": 9.07894878702495e-05, "loss": 0.0547, "step": 18920 }, { "grad_norm": 0.674269437789917, "learning_rate": 9.077753086181974e-05, "loss": 0.0719, "step": 18930 }, { "grad_norm": 0.4943781793117523, "learning_rate": 9.076556688562524e-05, "loss": 0.0871, "step": 18940 }, { "grad_norm": 0.5560541152954102, "learning_rate": 9.075359594371029e-05, "loss": 0.0643, "step": 18950 }, { "grad_norm": 0.6247986555099487, "learning_rate": 9.07416180381204e-05, "loss": 0.0497, "step": 18960 }, { "grad_norm": 0.5199994444847107, "learning_rate": 9.072963317090228e-05, "loss": 0.0498, "step": 18970 }, { "grad_norm": 0.48714718222618103, "learning_rate": 9.071764134410382e-05, "loss": 0.0533, "step": 18980 }, { "grad_norm": 0.4456281065940857, "learning_rate": 9.070564255977407e-05, "loss": 0.0618, "step": 18990 }, { "grad_norm": 0.5539214015007019, "learning_rate": 9.06936368199633e-05, "loss": 0.0517, "step": 19000 }, { "grad_norm": 0.6855767369270325, "learning_rate": 9.0681624126723e-05, "loss": 0.0675, "step": 19010 }, { "grad_norm": 0.562073826789856, "learning_rate": 9.066960448210576e-05, "loss": 0.0703, "step": 19020 }, { "grad_norm": 0.5332403182983398, "learning_rate": 9.065757788816543e-05, "loss": 0.0632, "step": 19030 }, { "grad_norm": 0.6673846244812012, "learning_rate": 9.064554434695705e-05, "loss": 0.0748, "step": 19040 }, { "grad_norm": 0.5894826650619507, "learning_rate": 9.063350386053677e-05, "loss": 0.0576, "step": 19050 }, { "grad_norm": 0.4642564654350281, "learning_rate": 9.062145643096202e-05, "loss": 0.0713, "step": 19060 }, { "grad_norm": 0.594599187374115, "learning_rate": 9.060940206029136e-05, "loss": 0.0567, "step": 19070 }, { "grad_norm": 0.7774783968925476, "learning_rate": 9.059734075058457e-05, "loss": 0.0771, "step": 19080 }, { "grad_norm": 0.6718998551368713, "learning_rate": 9.058527250390257e-05, "loss": 0.0754, "step": 19090 }, { "grad_norm": 0.6238294839859009, "learning_rate": 9.057319732230752e-05, "loss": 0.0619, "step": 19100 }, { "grad_norm": 0.7435464262962341, "learning_rate": 9.056111520786273e-05, "loss": 0.0767, "step": 19110 }, { "grad_norm": 0.5009673833847046, "learning_rate": 9.054902616263268e-05, "loss": 0.0585, "step": 19120 }, { "grad_norm": 0.8663815259933472, "learning_rate": 9.05369301886831e-05, "loss": 0.0767, "step": 19130 }, { "grad_norm": 0.7037065029144287, "learning_rate": 9.052482728808083e-05, "loss": 0.0863, "step": 19140 }, { "grad_norm": 0.43375346064567566, "learning_rate": 9.051271746289391e-05, "loss": 0.068, "step": 19150 }, { "grad_norm": 0.7473809123039246, "learning_rate": 9.050060071519162e-05, "loss": 0.0656, "step": 19160 }, { "grad_norm": 0.6555063724517822, "learning_rate": 9.048847704704437e-05, "loss": 0.0628, "step": 19170 }, { "grad_norm": 0.5760678648948669, "learning_rate": 9.047634646052376e-05, "loss": 0.0735, "step": 19180 }, { "grad_norm": 0.570476770401001, "learning_rate": 9.046420895770256e-05, "loss": 0.0668, "step": 19190 }, { "grad_norm": 0.7380223870277405, "learning_rate": 9.045206454065473e-05, "loss": 0.0707, "step": 19200 }, { "grad_norm": 0.5209107398986816, "learning_rate": 9.043991321145546e-05, "loss": 0.0575, "step": 19210 }, { "grad_norm": 0.8641029596328735, "learning_rate": 9.042775497218105e-05, "loss": 0.0692, "step": 19220 }, { "grad_norm": 0.49759745597839355, "learning_rate": 9.041558982490901e-05, "loss": 0.0666, "step": 19230 }, { "grad_norm": 0.4948166310787201, "learning_rate": 9.040341777171805e-05, "loss": 0.0701, "step": 19240 }, { "grad_norm": 0.6774736046791077, "learning_rate": 9.039123881468802e-05, "loss": 0.0656, "step": 19250 }, { "grad_norm": 0.5143533945083618, "learning_rate": 9.037905295589998e-05, "loss": 0.0506, "step": 19260 }, { "grad_norm": 0.5676798224449158, "learning_rate": 9.036686019743617e-05, "loss": 0.0465, "step": 19270 }, { "grad_norm": 0.6215192675590515, "learning_rate": 9.035466054137997e-05, "loss": 0.0739, "step": 19280 }, { "grad_norm": 0.6356722116470337, "learning_rate": 9.0342453989816e-05, "loss": 0.0731, "step": 19290 }, { "grad_norm": 0.5352442264556885, "learning_rate": 9.033024054483e-05, "loss": 0.0571, "step": 19300 }, { "grad_norm": 0.5391924977302551, "learning_rate": 9.031802020850894e-05, "loss": 0.0612, "step": 19310 }, { "grad_norm": 0.6871293783187866, "learning_rate": 9.030579298294092e-05, "loss": 0.0522, "step": 19320 }, { "grad_norm": 0.47703102231025696, "learning_rate": 9.029355887021524e-05, "loss": 0.0577, "step": 19330 }, { "grad_norm": 0.4091772735118866, "learning_rate": 9.028131787242238e-05, "loss": 0.0661, "step": 19340 }, { "grad_norm": 0.712080717086792, "learning_rate": 9.026906999165399e-05, "loss": 0.0575, "step": 19350 }, { "grad_norm": 0.5490865111351013, "learning_rate": 9.025681523000291e-05, "loss": 0.0624, "step": 19360 }, { "grad_norm": 0.6160407662391663, "learning_rate": 9.024455358956315e-05, "loss": 0.0654, "step": 19370 }, { "grad_norm": 0.7033779621124268, "learning_rate": 9.023228507242984e-05, "loss": 0.0568, "step": 19380 }, { "grad_norm": 0.5340075492858887, "learning_rate": 9.022000968069937e-05, "loss": 0.0524, "step": 19390 }, { "grad_norm": 0.6969432830810547, "learning_rate": 9.020772741646928e-05, "loss": 0.0589, "step": 19400 }, { "grad_norm": 0.6764618754386902, "learning_rate": 9.019543828183826e-05, "loss": 0.0592, "step": 19410 }, { "grad_norm": 0.6378481388092041, "learning_rate": 9.018314227890616e-05, "loss": 0.066, "step": 19420 }, { "grad_norm": 0.6196190714836121, "learning_rate": 9.017083940977408e-05, "loss": 0.053, "step": 19430 }, { "grad_norm": 0.6823040246963501, "learning_rate": 9.015852967654422e-05, "loss": 0.073, "step": 19440 }, { "grad_norm": 0.6060167551040649, "learning_rate": 9.014621308131996e-05, "loss": 0.0713, "step": 19450 }, { "grad_norm": 0.4655073285102844, "learning_rate": 9.01338896262059e-05, "loss": 0.0598, "step": 19460 }, { "grad_norm": 0.44978755712509155, "learning_rate": 9.012155931330777e-05, "loss": 0.062, "step": 19470 }, { "grad_norm": 0.5360910892486572, "learning_rate": 9.010922214473246e-05, "loss": 0.0727, "step": 19480 }, { "grad_norm": 0.6492568850517273, "learning_rate": 9.009687812258808e-05, "loss": 0.0692, "step": 19490 }, { "grad_norm": 0.9392902255058289, "learning_rate": 9.00845272489839e-05, "loss": 0.0664, "step": 19500 }, { "grad_norm": 0.5027825236320496, "learning_rate": 9.007216952603031e-05, "loss": 0.0596, "step": 19510 }, { "grad_norm": 0.4999372065067291, "learning_rate": 9.005980495583894e-05, "loss": 0.0602, "step": 19520 }, { "grad_norm": 0.5183752179145813, "learning_rate": 9.004743354052252e-05, "loss": 0.069, "step": 19530 }, { "grad_norm": 0.8409597873687744, "learning_rate": 9.003505528219503e-05, "loss": 0.0608, "step": 19540 }, { "grad_norm": 0.7863728404045105, "learning_rate": 9.002267018297154e-05, "loss": 0.0788, "step": 19550 }, { "grad_norm": 0.5498166680335999, "learning_rate": 9.001027824496834e-05, "loss": 0.0664, "step": 19560 }, { "grad_norm": 0.546129047870636, "learning_rate": 8.999787947030287e-05, "loss": 0.0612, "step": 19570 }, { "grad_norm": 0.7612246870994568, "learning_rate": 8.998547386109376e-05, "loss": 0.0666, "step": 19580 }, { "grad_norm": 0.5650148987770081, "learning_rate": 8.997306141946073e-05, "loss": 0.0582, "step": 19590 }, { "grad_norm": 0.4756816029548645, "learning_rate": 8.996064214752481e-05, "loss": 0.0516, "step": 19600 }, { "grad_norm": 0.510148823261261, "learning_rate": 8.994821604740806e-05, "loss": 0.0672, "step": 19610 }, { "grad_norm": 0.6311298608779907, "learning_rate": 8.993578312123377e-05, "loss": 0.0638, "step": 19620 }, { "grad_norm": 0.9102712869644165, "learning_rate": 8.992334337112639e-05, "loss": 0.0903, "step": 19630 }, { "grad_norm": 1.0025094747543335, "learning_rate": 8.991089679921154e-05, "loss": 0.0636, "step": 19640 }, { "grad_norm": 0.5947380661964417, "learning_rate": 8.989844340761599e-05, "loss": 0.0821, "step": 19650 }, { "grad_norm": 0.6008804440498352, "learning_rate": 8.988598319846768e-05, "loss": 0.0636, "step": 19660 }, { "grad_norm": 0.9125691056251526, "learning_rate": 8.987351617389574e-05, "loss": 0.0771, "step": 19670 }, { "grad_norm": 0.49695223569869995, "learning_rate": 8.98610423360304e-05, "loss": 0.0753, "step": 19680 }, { "grad_norm": 0.527684211730957, "learning_rate": 8.984856168700317e-05, "loss": 0.0771, "step": 19690 }, { "grad_norm": 0.501270592212677, "learning_rate": 8.983607422894658e-05, "loss": 0.0621, "step": 19700 }, { "grad_norm": 0.507274329662323, "learning_rate": 8.982357996399442e-05, "loss": 0.0662, "step": 19710 }, { "grad_norm": 0.6475759744644165, "learning_rate": 8.981107889428164e-05, "loss": 0.0499, "step": 19720 }, { "grad_norm": 0.6387499570846558, "learning_rate": 8.979857102194428e-05, "loss": 0.0605, "step": 19730 }, { "grad_norm": 0.5681970119476318, "learning_rate": 8.978605634911968e-05, "loss": 0.0567, "step": 19740 }, { "grad_norm": 0.8651335835456848, "learning_rate": 8.977353487794616e-05, "loss": 0.0699, "step": 19750 }, { "grad_norm": 0.8734700083732605, "learning_rate": 8.976100661056334e-05, "loss": 0.0723, "step": 19760 }, { "grad_norm": 0.4655676484107971, "learning_rate": 8.974847154911197e-05, "loss": 0.0518, "step": 19770 }, { "grad_norm": 0.4197981655597687, "learning_rate": 8.973592969573393e-05, "loss": 0.0483, "step": 19780 }, { "grad_norm": 0.5185088515281677, "learning_rate": 8.972338105257228e-05, "loss": 0.0621, "step": 19790 }, { "grad_norm": 0.6441845297813416, "learning_rate": 8.971082562177125e-05, "loss": 0.0557, "step": 19800 }, { "grad_norm": 0.6651083827018738, "learning_rate": 8.96982634054762e-05, "loss": 0.0563, "step": 19810 }, { "grad_norm": 1.1688061952590942, "learning_rate": 8.96856944058337e-05, "loss": 0.063, "step": 19820 }, { "grad_norm": 0.6663194298744202, "learning_rate": 8.967311862499144e-05, "loss": 0.062, "step": 19830 }, { "grad_norm": 0.4434698820114136, "learning_rate": 8.966053606509825e-05, "loss": 0.074, "step": 19840 }, { "grad_norm": 0.5813297033309937, "learning_rate": 8.964794672830417e-05, "loss": 0.0591, "step": 19850 }, { "grad_norm": 0.5553680062294006, "learning_rate": 8.963535061676038e-05, "loss": 0.0417, "step": 19860 }, { "grad_norm": 0.5900445580482483, "learning_rate": 8.962274773261918e-05, "loss": 0.0533, "step": 19870 }, { "grad_norm": 0.3554530143737793, "learning_rate": 8.961013807803409e-05, "loss": 0.0629, "step": 19880 }, { "grad_norm": 0.6450086832046509, "learning_rate": 8.959752165515973e-05, "loss": 0.0475, "step": 19890 }, { "grad_norm": 0.5100235939025879, "learning_rate": 8.958489846615193e-05, "loss": 0.0664, "step": 19900 }, { "grad_norm": 0.46056219935417175, "learning_rate": 8.957226851316762e-05, "loss": 0.0541, "step": 19910 }, { "grad_norm": 0.584564745426178, "learning_rate": 8.955963179836493e-05, "loss": 0.0473, "step": 19920 }, { "grad_norm": 0.4954194724559784, "learning_rate": 8.954698832390312e-05, "loss": 0.0547, "step": 19930 }, { "grad_norm": 0.5514769554138184, "learning_rate": 8.953433809194263e-05, "loss": 0.0705, "step": 19940 }, { "grad_norm": 0.9402923583984375, "learning_rate": 8.9521681104645e-05, "loss": 0.0623, "step": 19950 }, { "grad_norm": 0.4009716808795929, "learning_rate": 8.9509017364173e-05, "loss": 0.0516, "step": 19960 }, { "grad_norm": 0.7013669013977051, "learning_rate": 8.949634687269052e-05, "loss": 0.0786, "step": 19970 }, { "grad_norm": 0.5605018138885498, "learning_rate": 8.948366963236259e-05, "loss": 0.0551, "step": 19980 }, { "grad_norm": 0.7137072086334229, "learning_rate": 8.947098564535538e-05, "loss": 0.0685, "step": 19990 }, { "grad_norm": 0.8225098848342896, "learning_rate": 8.945829491383627e-05, "loss": 0.0834, "step": 20000 }, { "grad_norm": 0.6243898868560791, "learning_rate": 8.944559743997374e-05, "loss": 0.0821, "step": 20010 }, { "grad_norm": 0.5370494723320007, "learning_rate": 8.943289322593746e-05, "loss": 0.059, "step": 20020 }, { "grad_norm": 0.5499917268753052, "learning_rate": 8.942018227389821e-05, "loss": 0.0561, "step": 20030 }, { "grad_norm": 0.49408793449401855, "learning_rate": 8.940746458602795e-05, "loss": 0.0763, "step": 20040 }, { "grad_norm": 0.5492050647735596, "learning_rate": 8.939474016449979e-05, "loss": 0.0572, "step": 20050 }, { "grad_norm": 1.1043319702148438, "learning_rate": 8.938200901148799e-05, "loss": 0.0579, "step": 20060 }, { "grad_norm": 0.6246229410171509, "learning_rate": 8.936927112916795e-05, "loss": 0.0753, "step": 20070 }, { "grad_norm": 0.7544718384742737, "learning_rate": 8.935652651971622e-05, "loss": 0.0637, "step": 20080 }, { "grad_norm": 0.7651641964912415, "learning_rate": 8.934377518531052e-05, "loss": 0.0709, "step": 20090 }, { "grad_norm": 0.6361972093582153, "learning_rate": 8.933101712812967e-05, "loss": 0.0606, "step": 20100 }, { "grad_norm": 0.4157719612121582, "learning_rate": 8.931825235035374e-05, "loss": 0.068, "step": 20110 }, { "grad_norm": 0.8426913022994995, "learning_rate": 8.930548085416382e-05, "loss": 0.058, "step": 20120 }, { "grad_norm": 0.4876193106174469, "learning_rate": 8.92927026417422e-05, "loss": 0.0507, "step": 20130 }, { "grad_norm": 0.4518603980541229, "learning_rate": 8.92799177152724e-05, "loss": 0.0505, "step": 20140 }, { "grad_norm": 0.5875030159950256, "learning_rate": 8.926712607693895e-05, "loss": 0.0645, "step": 20150 }, { "grad_norm": 0.5337037444114685, "learning_rate": 8.925432772892762e-05, "loss": 0.057, "step": 20160 }, { "grad_norm": 0.45306965708732605, "learning_rate": 8.924152267342529e-05, "loss": 0.0602, "step": 20170 }, { "grad_norm": 0.7140997648239136, "learning_rate": 8.922871091261998e-05, "loss": 0.0575, "step": 20180 }, { "grad_norm": 0.7310110926628113, "learning_rate": 8.92158924487009e-05, "loss": 0.0554, "step": 20190 }, { "grad_norm": 0.4800836443901062, "learning_rate": 8.920306728385834e-05, "loss": 0.0565, "step": 20200 }, { "grad_norm": 0.6410326957702637, "learning_rate": 8.919023542028379e-05, "loss": 0.0475, "step": 20210 }, { "grad_norm": 0.7726832628250122, "learning_rate": 8.917739686016988e-05, "loss": 0.059, "step": 20220 }, { "grad_norm": 0.5145350098609924, "learning_rate": 8.916455160571033e-05, "loss": 0.0466, "step": 20230 }, { "grad_norm": 0.5194502472877502, "learning_rate": 8.915169965910008e-05, "loss": 0.0597, "step": 20240 }, { "grad_norm": 0.477666437625885, "learning_rate": 8.913884102253514e-05, "loss": 0.0594, "step": 20250 }, { "grad_norm": 0.8985059261322021, "learning_rate": 8.912597569821273e-05, "loss": 0.0886, "step": 20260 }, { "grad_norm": 0.5221349000930786, "learning_rate": 8.911310368833118e-05, "loss": 0.0677, "step": 20270 }, { "grad_norm": 0.6239999532699585, "learning_rate": 8.910022499508994e-05, "loss": 0.058, "step": 20280 }, { "grad_norm": 0.5335757732391357, "learning_rate": 8.908733962068965e-05, "loss": 0.0759, "step": 20290 }, { "grad_norm": 0.606099545955658, "learning_rate": 8.907444756733207e-05, "loss": 0.0758, "step": 20300 }, { "grad_norm": 0.6715884208679199, "learning_rate": 8.906154883722006e-05, "loss": 0.0659, "step": 20310 }, { "grad_norm": 0.939964234828949, "learning_rate": 8.904864343255773e-05, "loss": 0.0741, "step": 20320 }, { "grad_norm": 0.45836010575294495, "learning_rate": 8.90357313555502e-05, "loss": 0.0669, "step": 20330 }, { "grad_norm": 0.6889458894729614, "learning_rate": 8.90228126084038e-05, "loss": 0.0802, "step": 20340 }, { "grad_norm": 0.5801084041595459, "learning_rate": 8.900988719332601e-05, "loss": 0.0715, "step": 20350 }, { "grad_norm": 0.7492280602455139, "learning_rate": 8.899695511252542e-05, "loss": 0.0882, "step": 20360 }, { "grad_norm": 0.6240035891532898, "learning_rate": 8.898401636821176e-05, "loss": 0.0689, "step": 20370 }, { "grad_norm": 0.6272207498550415, "learning_rate": 8.897107096259593e-05, "loss": 0.1003, "step": 20380 }, { "grad_norm": 0.6752526760101318, "learning_rate": 8.895811889788994e-05, "loss": 0.0712, "step": 20390 }, { "grad_norm": 0.7192416787147522, "learning_rate": 8.894516017630692e-05, "loss": 0.0634, "step": 20400 }, { "grad_norm": 0.5782979726791382, "learning_rate": 8.893219480006118e-05, "loss": 0.0711, "step": 20410 }, { "grad_norm": 0.660591185092926, "learning_rate": 8.891922277136817e-05, "loss": 0.0608, "step": 20420 }, { "grad_norm": 1.0018795728683472, "learning_rate": 8.890624409244441e-05, "loss": 0.0618, "step": 20430 }, { "grad_norm": 0.5189362168312073, "learning_rate": 8.889325876550763e-05, "loss": 0.046, "step": 20440 }, { "grad_norm": 0.4457656443119049, "learning_rate": 8.888026679277666e-05, "loss": 0.0544, "step": 20450 }, { "grad_norm": 0.42417237162590027, "learning_rate": 8.886726817647147e-05, "loss": 0.0651, "step": 20460 }, { "grad_norm": 0.5183926224708557, "learning_rate": 8.885426291881319e-05, "loss": 0.0685, "step": 20470 }, { "grad_norm": 0.6332104206085205, "learning_rate": 8.884125102202401e-05, "loss": 0.0427, "step": 20480 }, { "grad_norm": 0.5539050698280334, "learning_rate": 8.882823248832736e-05, "loss": 0.0614, "step": 20490 }, { "grad_norm": 0.643974781036377, "learning_rate": 8.881520731994772e-05, "loss": 0.0609, "step": 20500 }, { "grad_norm": 0.5086473822593689, "learning_rate": 8.880217551911077e-05, "loss": 0.0513, "step": 20510 }, { "grad_norm": 1.2259094715118408, "learning_rate": 8.878913708804323e-05, "loss": 0.0673, "step": 20520 }, { "grad_norm": 0.5133638381958008, "learning_rate": 8.877609202897308e-05, "loss": 0.0587, "step": 20530 }, { "grad_norm": 0.5859725475311279, "learning_rate": 8.876304034412933e-05, "loss": 0.0763, "step": 20540 }, { "grad_norm": 0.7326239943504333, "learning_rate": 8.874998203574214e-05, "loss": 0.0801, "step": 20550 }, { "grad_norm": 0.7457078099250793, "learning_rate": 8.873691710604284e-05, "loss": 0.0711, "step": 20560 }, { "grad_norm": 0.3855946660041809, "learning_rate": 8.872384555726387e-05, "loss": 0.0912, "step": 20570 }, { "grad_norm": 0.5846999287605286, "learning_rate": 8.871076739163878e-05, "loss": 0.053, "step": 20580 }, { "grad_norm": 0.5228148102760315, "learning_rate": 8.86976826114023e-05, "loss": 0.0636, "step": 20590 }, { "grad_norm": 0.7016137838363647, "learning_rate": 8.868459121879023e-05, "loss": 0.0726, "step": 20600 }, { "grad_norm": 0.413457453250885, "learning_rate": 8.867149321603956e-05, "loss": 0.0485, "step": 20610 }, { "grad_norm": 0.7759842872619629, "learning_rate": 8.865838860538835e-05, "loss": 0.059, "step": 20620 }, { "grad_norm": 0.5944617390632629, "learning_rate": 8.864527738907585e-05, "loss": 0.068, "step": 20630 }, { "grad_norm": 0.6485181450843811, "learning_rate": 8.863215956934239e-05, "loss": 0.0619, "step": 20640 }, { "grad_norm": 0.4966452121734619, "learning_rate": 8.861903514842947e-05, "loss": 0.06, "step": 20650 }, { "grad_norm": 0.8632218241691589, "learning_rate": 8.860590412857966e-05, "loss": 0.064, "step": 20660 }, { "grad_norm": 0.4014282822608948, "learning_rate": 8.85927665120367e-05, "loss": 0.0519, "step": 20670 }, { "grad_norm": 0.6076270937919617, "learning_rate": 8.857962230104546e-05, "loss": 0.0727, "step": 20680 }, { "grad_norm": 0.4481952488422394, "learning_rate": 8.856647149785193e-05, "loss": 0.0764, "step": 20690 }, { "grad_norm": 0.49438756704330444, "learning_rate": 8.855331410470322e-05, "loss": 0.0814, "step": 20700 }, { "grad_norm": 1.3530442714691162, "learning_rate": 8.854015012384756e-05, "loss": 0.0715, "step": 20710 }, { "grad_norm": 0.5372995138168335, "learning_rate": 8.852697955753433e-05, "loss": 0.0552, "step": 20720 }, { "grad_norm": 0.552340030670166, "learning_rate": 8.851380240801399e-05, "loss": 0.0658, "step": 20730 }, { "grad_norm": 0.49919429421424866, "learning_rate": 8.850061867753818e-05, "loss": 0.0545, "step": 20740 }, { "grad_norm": 0.6616458296775818, "learning_rate": 8.848742836835963e-05, "loss": 0.0571, "step": 20750 }, { "grad_norm": 0.5667692422866821, "learning_rate": 8.847423148273221e-05, "loss": 0.0574, "step": 20760 }, { "grad_norm": 0.37669655680656433, "learning_rate": 8.846102802291092e-05, "loss": 0.0616, "step": 20770 }, { "grad_norm": 0.40785902738571167, "learning_rate": 8.844781799115183e-05, "loss": 0.0575, "step": 20780 }, { "grad_norm": 0.623694658279419, "learning_rate": 8.84346013897122e-05, "loss": 0.0688, "step": 20790 }, { "grad_norm": 0.5003803372383118, "learning_rate": 8.842137822085038e-05, "loss": 0.0636, "step": 20800 }, { "grad_norm": 0.5962779521942139, "learning_rate": 8.840814848682585e-05, "loss": 0.059, "step": 20810 }, { "grad_norm": 0.7796601057052612, "learning_rate": 8.83949121898992e-05, "loss": 0.0626, "step": 20820 }, { "grad_norm": 0.7351930737495422, "learning_rate": 8.838166933233217e-05, "loss": 0.0623, "step": 20830 }, { "grad_norm": 0.8955584764480591, "learning_rate": 8.83684199163876e-05, "loss": 0.0737, "step": 20840 }, { "grad_norm": 0.510408878326416, "learning_rate": 8.835516394432943e-05, "loss": 0.0548, "step": 20850 }, { "grad_norm": 0.7084617018699646, "learning_rate": 8.834190141842276e-05, "loss": 0.0729, "step": 20860 }, { "grad_norm": 0.5627443790435791, "learning_rate": 8.83286323409338e-05, "loss": 0.0597, "step": 20870 }, { "grad_norm": 0.7703986763954163, "learning_rate": 8.831535671412986e-05, "loss": 0.0601, "step": 20880 }, { "grad_norm": 0.6708890795707703, "learning_rate": 8.830207454027938e-05, "loss": 0.0641, "step": 20890 }, { "grad_norm": 0.5512528419494629, "learning_rate": 8.828878582165192e-05, "loss": 0.0805, "step": 20900 }, { "grad_norm": 0.7466639876365662, "learning_rate": 8.827549056051818e-05, "loss": 0.0636, "step": 20910 }, { "grad_norm": 0.555346667766571, "learning_rate": 8.826218875914993e-05, "loss": 0.0509, "step": 20920 }, { "grad_norm": 0.686839759349823, "learning_rate": 8.82488804198201e-05, "loss": 0.0643, "step": 20930 }, { "grad_norm": 0.5347676277160645, "learning_rate": 8.82355655448027e-05, "loss": 0.0509, "step": 20940 }, { "grad_norm": 0.5710241794586182, "learning_rate": 8.822224413637293e-05, "loss": 0.0548, "step": 20950 }, { "grad_norm": 0.5803826451301575, "learning_rate": 8.820891619680697e-05, "loss": 0.0593, "step": 20960 }, { "grad_norm": 0.8530887365341187, "learning_rate": 8.819558172838227e-05, "loss": 0.0653, "step": 20970 }, { "grad_norm": 0.5591057538986206, "learning_rate": 8.818224073337731e-05, "loss": 0.0508, "step": 20980 }, { "grad_norm": 0.5776038765907288, "learning_rate": 8.816889321407169e-05, "loss": 0.0474, "step": 20990 }, { "grad_norm": 0.6539303660392761, "learning_rate": 8.815553917274615e-05, "loss": 0.0471, "step": 21000 }, { "grad_norm": 0.4173564612865448, "learning_rate": 8.81421786116825e-05, "loss": 0.0537, "step": 21010 }, { "grad_norm": 0.5757914781570435, "learning_rate": 8.812881153316373e-05, "loss": 0.0871, "step": 21020 }, { "grad_norm": 0.7287484407424927, "learning_rate": 8.81154379394739e-05, "loss": 0.0552, "step": 21030 }, { "grad_norm": 0.7726491689682007, "learning_rate": 8.810205783289818e-05, "loss": 0.0497, "step": 21040 }, { "grad_norm": 0.42335620522499084, "learning_rate": 8.808867121572286e-05, "loss": 0.065, "step": 21050 }, { "grad_norm": 0.5191537141799927, "learning_rate": 8.807527809023537e-05, "loss": 0.0572, "step": 21060 }, { "grad_norm": 0.8115302920341492, "learning_rate": 8.80618784587242e-05, "loss": 0.0764, "step": 21070 }, { "grad_norm": 0.4268556535243988, "learning_rate": 8.804847232347902e-05, "loss": 0.0466, "step": 21080 }, { "grad_norm": 0.5171899795532227, "learning_rate": 8.803505968679054e-05, "loss": 0.0617, "step": 21090 }, { "grad_norm": 0.6130486130714417, "learning_rate": 8.802164055095061e-05, "loss": 0.0669, "step": 21100 }, { "grad_norm": 0.6660966277122498, "learning_rate": 8.80082149182522e-05, "loss": 0.0476, "step": 21110 }, { "grad_norm": 0.4735492765903473, "learning_rate": 8.79947827909894e-05, "loss": 0.0485, "step": 21120 }, { "grad_norm": 0.5671480298042297, "learning_rate": 8.798134417145738e-05, "loss": 0.0544, "step": 21130 }, { "grad_norm": 0.5576417446136475, "learning_rate": 8.796789906195243e-05, "loss": 0.0521, "step": 21140 }, { "grad_norm": 0.6121723651885986, "learning_rate": 8.795444746477195e-05, "loss": 0.0524, "step": 21150 }, { "grad_norm": 0.4526224434375763, "learning_rate": 8.794098938221446e-05, "loss": 0.0579, "step": 21160 }, { "grad_norm": 0.5770092606544495, "learning_rate": 8.792752481657957e-05, "loss": 0.0596, "step": 21170 }, { "grad_norm": 0.6330416202545166, "learning_rate": 8.791405377016802e-05, "loss": 0.055, "step": 21180 }, { "grad_norm": 0.7199046611785889, "learning_rate": 8.790057624528163e-05, "loss": 0.0601, "step": 21190 }, { "grad_norm": 0.5796276330947876, "learning_rate": 8.788709224422333e-05, "loss": 0.0777, "step": 21200 }, { "grad_norm": 0.6625869274139404, "learning_rate": 8.787360176929717e-05, "loss": 0.1032, "step": 21210 }, { "grad_norm": 0.5338436961174011, "learning_rate": 8.786010482280834e-05, "loss": 0.0663, "step": 21220 }, { "grad_norm": 0.49985256791114807, "learning_rate": 8.784660140706306e-05, "loss": 0.0691, "step": 21230 }, { "grad_norm": 0.7633554339408875, "learning_rate": 8.783309152436872e-05, "loss": 0.0705, "step": 21240 }, { "grad_norm": 0.676211416721344, "learning_rate": 8.781957517703375e-05, "loss": 0.0596, "step": 21250 }, { "grad_norm": 0.5982766151428223, "learning_rate": 8.780605236736776e-05, "loss": 0.0526, "step": 21260 }, { "grad_norm": 0.6315072178840637, "learning_rate": 8.779252309768142e-05, "loss": 0.0729, "step": 21270 }, { "grad_norm": 0.5278592109680176, "learning_rate": 8.777898737028652e-05, "loss": 0.066, "step": 21280 }, { "grad_norm": 0.5372454524040222, "learning_rate": 8.776544518749591e-05, "loss": 0.0797, "step": 21290 }, { "grad_norm": 0.5788964629173279, "learning_rate": 8.775189655162364e-05, "loss": 0.0577, "step": 21300 }, { "grad_norm": 0.5674257278442383, "learning_rate": 8.773834146498474e-05, "loss": 0.0808, "step": 21310 }, { "grad_norm": 0.521350085735321, "learning_rate": 8.772477992989545e-05, "loss": 0.0648, "step": 21320 }, { "grad_norm": 0.8971951603889465, "learning_rate": 8.771121194867304e-05, "loss": 0.0707, "step": 21330 }, { "grad_norm": 0.6461492776870728, "learning_rate": 8.769763752363589e-05, "loss": 0.0569, "step": 21340 }, { "grad_norm": 0.421587198972702, "learning_rate": 8.768405665710352e-05, "loss": 0.0443, "step": 21350 }, { "grad_norm": 0.35140058398246765, "learning_rate": 8.767046935139655e-05, "loss": 0.0515, "step": 21360 }, { "grad_norm": 0.394243061542511, "learning_rate": 8.765687560883666e-05, "loss": 0.0587, "step": 21370 }, { "grad_norm": 0.7140743136405945, "learning_rate": 8.764327543174664e-05, "loss": 0.0504, "step": 21380 }, { "grad_norm": 0.5861976146697998, "learning_rate": 8.762966882245038e-05, "loss": 0.0474, "step": 21390 }, { "grad_norm": 0.5930838584899902, "learning_rate": 8.761605578327291e-05, "loss": 0.1098, "step": 21400 }, { "grad_norm": 0.6017702221870422, "learning_rate": 8.76024363165403e-05, "loss": 0.0692, "step": 21410 }, { "grad_norm": 0.5239245891571045, "learning_rate": 8.758881042457976e-05, "loss": 0.0608, "step": 21420 }, { "grad_norm": 0.4352739751338959, "learning_rate": 8.757517810971957e-05, "loss": 0.0538, "step": 21430 }, { "grad_norm": 0.5670722126960754, "learning_rate": 8.756153937428913e-05, "loss": 0.0707, "step": 21440 }, { "grad_norm": 0.5181528925895691, "learning_rate": 8.754789422061889e-05, "loss": 0.0572, "step": 21450 }, { "grad_norm": 0.47972342371940613, "learning_rate": 8.753424265104052e-05, "loss": 0.0459, "step": 21460 }, { "grad_norm": 0.4141260087490082, "learning_rate": 8.752058466788659e-05, "loss": 0.0547, "step": 21470 }, { "grad_norm": 0.7614762783050537, "learning_rate": 8.750692027349097e-05, "loss": 0.0409, "step": 21480 }, { "grad_norm": 0.8288816213607788, "learning_rate": 8.749324947018847e-05, "loss": 0.0584, "step": 21490 }, { "grad_norm": 0.42283815145492554, "learning_rate": 8.747957226031507e-05, "loss": 0.0608, "step": 21500 }, { "grad_norm": 0.5053452849388123, "learning_rate": 8.746588864620787e-05, "loss": 0.0717, "step": 21510 }, { "grad_norm": 0.6002009510993958, "learning_rate": 8.745219863020498e-05, "loss": 0.0587, "step": 21520 }, { "grad_norm": 0.54496830701828, "learning_rate": 8.743850221464564e-05, "loss": 0.0695, "step": 21530 }, { "grad_norm": 0.8493882417678833, "learning_rate": 8.742479940187026e-05, "loss": 0.0808, "step": 21540 }, { "grad_norm": 0.49896591901779175, "learning_rate": 8.74110901942202e-05, "loss": 0.0792, "step": 21550 }, { "grad_norm": 0.5150942206382751, "learning_rate": 8.739737459403803e-05, "loss": 0.0697, "step": 21560 }, { "grad_norm": 0.5396903157234192, "learning_rate": 8.738365260366737e-05, "loss": 0.064, "step": 21570 }, { "grad_norm": 0.6287738084793091, "learning_rate": 8.736992422545292e-05, "loss": 0.0586, "step": 21580 }, { "grad_norm": 0.48064497113227844, "learning_rate": 8.73561894617405e-05, "loss": 0.0741, "step": 21590 }, { "grad_norm": 0.7048019170761108, "learning_rate": 8.734244831487697e-05, "loss": 0.0745, "step": 21600 }, { "grad_norm": 0.7130262851715088, "learning_rate": 8.732870078721035e-05, "loss": 0.062, "step": 21610 }, { "grad_norm": 0.5515240430831909, "learning_rate": 8.731494688108972e-05, "loss": 0.0697, "step": 21620 }, { "grad_norm": 0.7344058156013489, "learning_rate": 8.730118659886523e-05, "loss": 0.0877, "step": 21630 }, { "grad_norm": 0.49009838700294495, "learning_rate": 8.728741994288814e-05, "loss": 0.0632, "step": 21640 }, { "grad_norm": 0.7136932611465454, "learning_rate": 8.727364691551079e-05, "loss": 0.0742, "step": 21650 }, { "grad_norm": 0.6005926132202148, "learning_rate": 8.725986751908661e-05, "loss": 0.0736, "step": 21660 }, { "grad_norm": 0.7626722455024719, "learning_rate": 8.724608175597016e-05, "loss": 0.0898, "step": 21670 }, { "grad_norm": 0.5580800175666809, "learning_rate": 8.723228962851699e-05, "loss": 0.0638, "step": 21680 }, { "grad_norm": 0.5164566040039062, "learning_rate": 8.721849113908385e-05, "loss": 0.0626, "step": 21690 }, { "grad_norm": 0.6217962503433228, "learning_rate": 8.720468629002848e-05, "loss": 0.0688, "step": 21700 }, { "grad_norm": 1.1892591714859009, "learning_rate": 8.719087508370978e-05, "loss": 0.0578, "step": 21710 }, { "grad_norm": 0.5711930990219116, "learning_rate": 8.717705752248772e-05, "loss": 0.0621, "step": 21720 }, { "grad_norm": 0.6688865423202515, "learning_rate": 8.71632336087233e-05, "loss": 0.0679, "step": 21730 }, { "grad_norm": 0.5257223844528198, "learning_rate": 8.71494033447787e-05, "loss": 0.068, "step": 21740 }, { "grad_norm": 0.7705744504928589, "learning_rate": 8.713556673301708e-05, "loss": 0.0759, "step": 21750 }, { "grad_norm": 0.5754894018173218, "learning_rate": 8.712172377580278e-05, "loss": 0.0685, "step": 21760 }, { "grad_norm": 0.9511322975158691, "learning_rate": 8.710787447550114e-05, "loss": 0.07, "step": 21770 }, { "grad_norm": 0.5976829528808594, "learning_rate": 8.70940188344787e-05, "loss": 0.0666, "step": 21780 }, { "grad_norm": 0.9404848217964172, "learning_rate": 8.708015685510293e-05, "loss": 0.0915, "step": 21790 }, { "grad_norm": 0.7081054449081421, "learning_rate": 8.706628853974252e-05, "loss": 0.0686, "step": 21800 }, { "grad_norm": 0.8244065046310425, "learning_rate": 8.705241389076715e-05, "loss": 0.0678, "step": 21810 }, { "grad_norm": 0.46262824535369873, "learning_rate": 8.703853291054764e-05, "loss": 0.0623, "step": 21820 }, { "grad_norm": 0.6114557385444641, "learning_rate": 8.702464560145587e-05, "loss": 0.0603, "step": 21830 }, { "grad_norm": 0.6565998196601868, "learning_rate": 8.701075196586476e-05, "loss": 0.0576, "step": 21840 }, { "grad_norm": 0.8840348720550537, "learning_rate": 8.699685200614842e-05, "loss": 0.0954, "step": 21850 }, { "grad_norm": 0.7361170649528503, "learning_rate": 8.698294572468193e-05, "loss": 0.0649, "step": 21860 }, { "grad_norm": 0.5458929538726807, "learning_rate": 8.696903312384148e-05, "loss": 0.0566, "step": 21870 }, { "grad_norm": 0.6949689984321594, "learning_rate": 8.695511420600439e-05, "loss": 0.0775, "step": 21880 }, { "grad_norm": 0.394422322511673, "learning_rate": 8.694118897354901e-05, "loss": 0.0545, "step": 21890 }, { "grad_norm": 1.1813582181930542, "learning_rate": 8.692725742885478e-05, "loss": 0.0805, "step": 21900 }, { "grad_norm": 0.6634201407432556, "learning_rate": 8.691331957430221e-05, "loss": 0.0476, "step": 21910 }, { "grad_norm": 0.5312386155128479, "learning_rate": 8.68993754122729e-05, "loss": 0.061, "step": 21920 }, { "grad_norm": 0.5794894695281982, "learning_rate": 8.688542494514955e-05, "loss": 0.0697, "step": 21930 }, { "grad_norm": 0.7271232008934021, "learning_rate": 8.68714681753159e-05, "loss": 0.0504, "step": 21940 }, { "grad_norm": 0.5416128635406494, "learning_rate": 8.685750510515676e-05, "loss": 0.0577, "step": 21950 }, { "grad_norm": 0.6829200983047485, "learning_rate": 8.684353573705805e-05, "loss": 0.0564, "step": 21960 }, { "grad_norm": 0.6064825654029846, "learning_rate": 8.682956007340677e-05, "loss": 0.0471, "step": 21970 }, { "grad_norm": 0.7039556503295898, "learning_rate": 8.681557811659095e-05, "loss": 0.0547, "step": 21980 }, { "grad_norm": 0.5873735547065735, "learning_rate": 8.680158986899974e-05, "loss": 0.0567, "step": 21990 }, { "grad_norm": 0.7589130997657776, "learning_rate": 8.678759533302335e-05, "loss": 0.0715, "step": 22000 }, { "grad_norm": 0.6061754822731018, "learning_rate": 8.677359451105308e-05, "loss": 0.079, "step": 22010 }, { "grad_norm": 0.5571337938308716, "learning_rate": 8.675958740548123e-05, "loss": 0.0638, "step": 22020 }, { "grad_norm": 0.6447767019271851, "learning_rate": 8.674557401870129e-05, "loss": 0.0717, "step": 22030 }, { "grad_norm": 0.5989105701446533, "learning_rate": 8.673155435310775e-05, "loss": 0.0591, "step": 22040 }, { "grad_norm": 0.8005304336547852, "learning_rate": 8.671752841109617e-05, "loss": 0.0602, "step": 22050 }, { "grad_norm": 0.8927066922187805, "learning_rate": 8.670349619506321e-05, "loss": 0.0746, "step": 22060 }, { "grad_norm": 0.7577258348464966, "learning_rate": 8.66894577074066e-05, "loss": 0.055, "step": 22070 }, { "grad_norm": 0.6947263479232788, "learning_rate": 8.667541295052513e-05, "loss": 0.0801, "step": 22080 }, { "grad_norm": 0.45651164650917053, "learning_rate": 8.666136192681865e-05, "loss": 0.0591, "step": 22090 }, { "grad_norm": 0.6230955123901367, "learning_rate": 8.664730463868811e-05, "loss": 0.05, "step": 22100 }, { "grad_norm": 0.7753889560699463, "learning_rate": 8.663324108853552e-05, "loss": 0.0639, "step": 22110 }, { "grad_norm": 0.72808438539505, "learning_rate": 8.661917127876395e-05, "loss": 0.072, "step": 22120 }, { "grad_norm": 0.6767024397850037, "learning_rate": 8.660509521177754e-05, "loss": 0.059, "step": 22130 }, { "grad_norm": 0.5572755932807922, "learning_rate": 8.65910128899815e-05, "loss": 0.0659, "step": 22140 }, { "grad_norm": 0.4727124273777008, "learning_rate": 8.657692431578214e-05, "loss": 0.0628, "step": 22150 }, { "grad_norm": 0.5439937114715576, "learning_rate": 8.656282949158679e-05, "loss": 0.0787, "step": 22160 }, { "grad_norm": 11.219111442565918, "learning_rate": 8.654872841980388e-05, "loss": 0.1076, "step": 22170 }, { "grad_norm": 0.5078825354576111, "learning_rate": 8.653462110284289e-05, "loss": 0.3797, "step": 22180 }, { "grad_norm": 0.6767317652702332, "learning_rate": 8.652050754311437e-05, "loss": 0.0651, "step": 22190 }, { "grad_norm": 0.7002315521240234, "learning_rate": 8.650638774302995e-05, "loss": 0.0905, "step": 22200 }, { "grad_norm": 0.8259180188179016, "learning_rate": 8.649226170500233e-05, "loss": 0.0491, "step": 22210 }, { "grad_norm": 1.1728787422180176, "learning_rate": 8.647812943144524e-05, "loss": 0.0806, "step": 22220 }, { "grad_norm": 0.7804592847824097, "learning_rate": 8.646399092477351e-05, "loss": 0.0553, "step": 22230 }, { "grad_norm": 0.6792881488800049, "learning_rate": 8.644984618740301e-05, "loss": 0.069, "step": 22240 }, { "grad_norm": 0.4699406325817108, "learning_rate": 8.643569522175073e-05, "loss": 0.0827, "step": 22250 }, { "grad_norm": 0.7308951616287231, "learning_rate": 8.642153803023463e-05, "loss": 0.0697, "step": 22260 }, { "grad_norm": 0.5936901569366455, "learning_rate": 8.640737461527383e-05, "loss": 0.0684, "step": 22270 }, { "grad_norm": 0.4958588778972626, "learning_rate": 8.639320497928845e-05, "loss": 0.0516, "step": 22280 }, { "grad_norm": 0.577332079410553, "learning_rate": 8.637902912469969e-05, "loss": 0.0559, "step": 22290 }, { "grad_norm": 0.4179581105709076, "learning_rate": 8.636484705392982e-05, "loss": 0.0526, "step": 22300 }, { "grad_norm": 1.2594698667526245, "learning_rate": 8.635065876940216e-05, "loss": 0.0697, "step": 22310 }, { "grad_norm": 0.6352766156196594, "learning_rate": 8.633646427354112e-05, "loss": 0.0558, "step": 22320 }, { "grad_norm": 0.7250261902809143, "learning_rate": 8.632226356877213e-05, "loss": 0.0538, "step": 22330 }, { "grad_norm": 0.5320909023284912, "learning_rate": 8.630805665752173e-05, "loss": 0.0615, "step": 22340 }, { "grad_norm": 0.600367546081543, "learning_rate": 8.629384354221748e-05, "loss": 0.0686, "step": 22350 }, { "grad_norm": 0.8238775730133057, "learning_rate": 8.627962422528797e-05, "loss": 0.0621, "step": 22360 }, { "grad_norm": 0.42056846618652344, "learning_rate": 8.626539870916296e-05, "loss": 0.0774, "step": 22370 }, { "grad_norm": 0.3680470883846283, "learning_rate": 8.625116699627317e-05, "loss": 0.0502, "step": 22380 }, { "grad_norm": 0.7238022089004517, "learning_rate": 8.623692908905041e-05, "loss": 0.0611, "step": 22390 }, { "grad_norm": 0.6522291898727417, "learning_rate": 8.622268498992755e-05, "loss": 0.0664, "step": 22400 }, { "grad_norm": 0.4765101373195648, "learning_rate": 8.620843470133851e-05, "loss": 0.0633, "step": 22410 }, { "grad_norm": 0.7143649458885193, "learning_rate": 8.619417822571829e-05, "loss": 0.0726, "step": 22420 }, { "grad_norm": 0.686448335647583, "learning_rate": 8.617991556550292e-05, "loss": 0.0476, "step": 22430 }, { "grad_norm": 0.5008071064949036, "learning_rate": 8.616564672312952e-05, "loss": 0.0576, "step": 22440 }, { "grad_norm": 0.6546361446380615, "learning_rate": 8.61513717010362e-05, "loss": 0.0753, "step": 22450 }, { "grad_norm": 0.5493152141571045, "learning_rate": 8.613709050166221e-05, "loss": 0.0571, "step": 22460 }, { "grad_norm": 0.8865336775779724, "learning_rate": 8.61228031274478e-05, "loss": 0.06, "step": 22470 }, { "grad_norm": 0.7256611585617065, "learning_rate": 8.610850958083431e-05, "loss": 0.0557, "step": 22480 }, { "grad_norm": 0.7036254405975342, "learning_rate": 8.609420986426409e-05, "loss": 0.0648, "step": 22490 }, { "grad_norm": 0.6148037910461426, "learning_rate": 8.60799039801806e-05, "loss": 0.0666, "step": 22500 }, { "grad_norm": 0.47885391116142273, "learning_rate": 8.606559193102828e-05, "loss": 0.051, "step": 22510 }, { "grad_norm": 0.894219160079956, "learning_rate": 8.605127371925273e-05, "loss": 0.0551, "step": 22520 }, { "grad_norm": 0.6715536713600159, "learning_rate": 8.603694934730047e-05, "loss": 0.0565, "step": 22530 }, { "grad_norm": 0.5588095784187317, "learning_rate": 8.602261881761919e-05, "loss": 0.0574, "step": 22540 }, { "grad_norm": 0.5659414529800415, "learning_rate": 8.600828213265759e-05, "loss": 0.0516, "step": 22550 }, { "grad_norm": 0.7329782843589783, "learning_rate": 8.599393929486539e-05, "loss": 0.0815, "step": 22560 }, { "grad_norm": 0.6880636215209961, "learning_rate": 8.59795903066934e-05, "loss": 0.0604, "step": 22570 }, { "grad_norm": 0.9357970952987671, "learning_rate": 8.596523517059347e-05, "loss": 0.0623, "step": 22580 }, { "grad_norm": 1.1542925834655762, "learning_rate": 8.59508738890185e-05, "loss": 0.0582, "step": 22590 }, { "grad_norm": 0.5667861104011536, "learning_rate": 8.593650646442246e-05, "loss": 0.0548, "step": 22600 }, { "grad_norm": 0.6539766192436218, "learning_rate": 8.59221328992603e-05, "loss": 0.0698, "step": 22610 }, { "grad_norm": 0.6250693798065186, "learning_rate": 8.590775319598813e-05, "loss": 0.069, "step": 22620 }, { "grad_norm": 0.7263991236686707, "learning_rate": 8.589336735706301e-05, "loss": 0.0633, "step": 22630 }, { "grad_norm": 0.6218357682228088, "learning_rate": 8.587897538494307e-05, "loss": 0.062, "step": 22640 }, { "grad_norm": 0.6537426114082336, "learning_rate": 8.586457728208756e-05, "loss": 0.0669, "step": 22650 }, { "grad_norm": 0.8093648552894592, "learning_rate": 8.585017305095667e-05, "loss": 0.0587, "step": 22660 }, { "grad_norm": 0.6275246143341064, "learning_rate": 8.583576269401173e-05, "loss": 0.0527, "step": 22670 }, { "grad_norm": 0.8097725510597229, "learning_rate": 8.582134621371504e-05, "loss": 0.0557, "step": 22680 }, { "grad_norm": 0.9636636972427368, "learning_rate": 8.580692361253e-05, "loss": 0.0772, "step": 22690 }, { "grad_norm": 0.5616360306739807, "learning_rate": 8.579249489292104e-05, "loss": 0.0709, "step": 22700 }, { "grad_norm": 0.5450718998908997, "learning_rate": 8.577806005735363e-05, "loss": 0.0521, "step": 22710 }, { "grad_norm": 0.5302026867866516, "learning_rate": 8.576361910829429e-05, "loss": 0.0547, "step": 22720 }, { "grad_norm": 0.4670838713645935, "learning_rate": 8.574917204821057e-05, "loss": 0.0443, "step": 22730 }, { "grad_norm": 0.6017994284629822, "learning_rate": 8.57347188795711e-05, "loss": 0.0499, "step": 22740 }, { "grad_norm": 0.6564391851425171, "learning_rate": 8.572025960484551e-05, "loss": 0.0578, "step": 22750 }, { "grad_norm": 0.5716729760169983, "learning_rate": 8.57057942265045e-05, "loss": 0.0575, "step": 22760 }, { "grad_norm": 0.5495119690895081, "learning_rate": 8.569132274701984e-05, "loss": 0.0733, "step": 22770 }, { "grad_norm": 0.4163214862346649, "learning_rate": 8.567684516886427e-05, "loss": 0.0495, "step": 22780 }, { "grad_norm": 0.7890225648880005, "learning_rate": 8.56623614945116e-05, "loss": 0.0615, "step": 22790 }, { "grad_norm": 0.5007728934288025, "learning_rate": 8.564787172643675e-05, "loss": 0.0779, "step": 22800 }, { "grad_norm": 0.6040037274360657, "learning_rate": 8.563337586711559e-05, "loss": 0.0448, "step": 22810 }, { "grad_norm": 0.606113612651825, "learning_rate": 8.561887391902506e-05, "loss": 0.0619, "step": 22820 }, { "grad_norm": 0.7039556503295898, "learning_rate": 8.560436588464316e-05, "loss": 0.0865, "step": 22830 }, { "grad_norm": 0.6916776299476624, "learning_rate": 8.55898517664489e-05, "loss": 0.0501, "step": 22840 }, { "grad_norm": 0.45975780487060547, "learning_rate": 8.557533156692236e-05, "loss": 0.0558, "step": 22850 }, { "grad_norm": 0.6211679577827454, "learning_rate": 8.556080528854467e-05, "loss": 0.0586, "step": 22860 }, { "grad_norm": 0.5889313817024231, "learning_rate": 8.554627293379791e-05, "loss": 0.0548, "step": 22870 }, { "grad_norm": 0.4917896091938019, "learning_rate": 8.553173450516531e-05, "loss": 0.0561, "step": 22880 }, { "grad_norm": 0.4133954644203186, "learning_rate": 8.551719000513108e-05, "loss": 0.0411, "step": 22890 }, { "grad_norm": 0.6826973557472229, "learning_rate": 8.550263943618049e-05, "loss": 0.0769, "step": 22900 }, { "grad_norm": 0.5914101004600525, "learning_rate": 8.54880828007998e-05, "loss": 0.0874, "step": 22910 }, { "grad_norm": 0.6139667630195618, "learning_rate": 8.547352010147637e-05, "loss": 0.0519, "step": 22920 }, { "grad_norm": 0.49609842896461487, "learning_rate": 8.545895134069855e-05, "loss": 0.0547, "step": 22930 }, { "grad_norm": 0.527489423751831, "learning_rate": 8.544437652095576e-05, "loss": 0.0563, "step": 22940 }, { "grad_norm": 0.7019833922386169, "learning_rate": 8.542979564473843e-05, "loss": 0.0594, "step": 22950 }, { "grad_norm": 0.6349751949310303, "learning_rate": 8.541520871453802e-05, "loss": 0.0609, "step": 22960 }, { "grad_norm": 0.5181769728660583, "learning_rate": 8.540061573284705e-05, "loss": 0.0619, "step": 22970 }, { "grad_norm": 0.7255668044090271, "learning_rate": 8.538601670215906e-05, "loss": 0.0694, "step": 22980 }, { "grad_norm": 0.41523730754852295, "learning_rate": 8.537141162496864e-05, "loss": 0.0438, "step": 22990 }, { "grad_norm": 0.743994414806366, "learning_rate": 8.535680050377137e-05, "loss": 0.0749, "step": 23000 }, { "grad_norm": 1.0159450769424438, "learning_rate": 8.534218334106391e-05, "loss": 0.058, "step": 23010 }, { "grad_norm": 0.5104672312736511, "learning_rate": 8.532756013934393e-05, "loss": 0.0639, "step": 23020 }, { "grad_norm": 0.4735672175884247, "learning_rate": 8.531293090111012e-05, "loss": 0.0675, "step": 23030 }, { "grad_norm": 0.5620094537734985, "learning_rate": 8.529829562886225e-05, "loss": 0.0426, "step": 23040 }, { "grad_norm": 0.6196703910827637, "learning_rate": 8.528365432510105e-05, "loss": 0.0507, "step": 23050 }, { "grad_norm": 0.39135614037513733, "learning_rate": 8.526900699232833e-05, "loss": 0.0541, "step": 23060 }, { "grad_norm": 0.38068705797195435, "learning_rate": 8.525435363304695e-05, "loss": 0.0618, "step": 23070 }, { "grad_norm": 0.45958077907562256, "learning_rate": 8.523969424976072e-05, "loss": 0.0529, "step": 23080 }, { "grad_norm": 0.5724307298660278, "learning_rate": 8.522502884497457e-05, "loss": 0.0633, "step": 23090 }, { "grad_norm": 0.8153622150421143, "learning_rate": 8.521035742119437e-05, "loss": 0.0589, "step": 23100 }, { "grad_norm": 0.4216693937778473, "learning_rate": 8.519567998092712e-05, "loss": 0.0534, "step": 23110 }, { "grad_norm": 0.45535820722579956, "learning_rate": 8.518099652668075e-05, "loss": 0.0504, "step": 23120 }, { "grad_norm": 0.5701044797897339, "learning_rate": 8.516630706096429e-05, "loss": 0.0528, "step": 23130 }, { "grad_norm": 0.5750473737716675, "learning_rate": 8.515161158628773e-05, "loss": 0.0482, "step": 23140 }, { "grad_norm": 0.7400012016296387, "learning_rate": 8.513691010516216e-05, "loss": 0.0578, "step": 23150 }, { "grad_norm": 0.7493351697921753, "learning_rate": 8.512220262009966e-05, "loss": 0.0635, "step": 23160 }, { "grad_norm": 0.6741169095039368, "learning_rate": 8.510748913361332e-05, "loss": 0.0654, "step": 23170 }, { "grad_norm": 0.7321332097053528, "learning_rate": 8.509276964821726e-05, "loss": 0.0622, "step": 23180 }, { "grad_norm": 0.4660942852497101, "learning_rate": 8.507804416642669e-05, "loss": 0.0568, "step": 23190 }, { "grad_norm": 0.6724201440811157, "learning_rate": 8.506331269075774e-05, "loss": 0.0529, "step": 23200 }, { "grad_norm": 0.44469454884529114, "learning_rate": 8.504857522372765e-05, "loss": 0.0747, "step": 23210 }, { "grad_norm": 0.507215678691864, "learning_rate": 8.503383176785461e-05, "loss": 0.0606, "step": 23220 }, { "grad_norm": 0.8304296731948853, "learning_rate": 8.501908232565792e-05, "loss": 0.0633, "step": 23230 }, { "grad_norm": 0.9604566097259521, "learning_rate": 8.50043268996578e-05, "loss": 0.0802, "step": 23240 }, { "grad_norm": 0.8057485222816467, "learning_rate": 8.498956549237562e-05, "loss": 0.0746, "step": 23250 }, { "grad_norm": 0.3850315809249878, "learning_rate": 8.497479810633366e-05, "loss": 0.0807, "step": 23260 }, { "grad_norm": 0.5496267676353455, "learning_rate": 8.496002474405525e-05, "loss": 0.0593, "step": 23270 }, { "grad_norm": 0.5741512179374695, "learning_rate": 8.494524540806478e-05, "loss": 0.0566, "step": 23280 }, { "grad_norm": 0.6105430126190186, "learning_rate": 8.493046010088761e-05, "loss": 0.0591, "step": 23290 }, { "grad_norm": 0.8345263600349426, "learning_rate": 8.491566882505018e-05, "loss": 0.0633, "step": 23300 }, { "grad_norm": 0.6198033094406128, "learning_rate": 8.490087158307988e-05, "loss": 0.0559, "step": 23310 }, { "grad_norm": 0.4845809042453766, "learning_rate": 8.488606837750518e-05, "loss": 0.06, "step": 23320 }, { "grad_norm": 0.7687404155731201, "learning_rate": 8.487125921085552e-05, "loss": 0.0564, "step": 23330 }, { "grad_norm": 0.5295619368553162, "learning_rate": 8.485644408566141e-05, "loss": 0.0534, "step": 23340 }, { "grad_norm": 0.44232869148254395, "learning_rate": 8.484162300445431e-05, "loss": 0.0695, "step": 23350 }, { "grad_norm": 0.8720763325691223, "learning_rate": 8.482679596976676e-05, "loss": 0.0662, "step": 23360 }, { "grad_norm": 0.7056252956390381, "learning_rate": 8.48119629841323e-05, "loss": 0.0563, "step": 23370 }, { "grad_norm": 0.6441484093666077, "learning_rate": 8.479712405008547e-05, "loss": 0.0591, "step": 23380 }, { "grad_norm": 0.3949601650238037, "learning_rate": 8.478227917016184e-05, "loss": 0.056, "step": 23390 }, { "grad_norm": 0.4689229726791382, "learning_rate": 8.476742834689801e-05, "loss": 0.0517, "step": 23400 }, { "grad_norm": 0.512578547000885, "learning_rate": 8.475257158283157e-05, "loss": 0.0574, "step": 23410 }, { "grad_norm": 0.576245129108429, "learning_rate": 8.473770888050112e-05, "loss": 0.0601, "step": 23420 }, { "grad_norm": 0.6677529215812683, "learning_rate": 8.47228402424463e-05, "loss": 0.0556, "step": 23430 }, { "grad_norm": 0.5844681262969971, "learning_rate": 8.470796567120775e-05, "loss": 0.0558, "step": 23440 }, { "grad_norm": 0.45732370018959045, "learning_rate": 8.469308516932714e-05, "loss": 0.0484, "step": 23450 }, { "grad_norm": 0.6338613033294678, "learning_rate": 8.467819873934714e-05, "loss": 0.0465, "step": 23460 }, { "grad_norm": 0.4297735095024109, "learning_rate": 8.466330638381143e-05, "loss": 0.0453, "step": 23470 }, { "grad_norm": 0.6144276261329651, "learning_rate": 8.464840810526469e-05, "loss": 0.0644, "step": 23480 }, { "grad_norm": 0.5161672830581665, "learning_rate": 8.463350390625264e-05, "loss": 0.052, "step": 23490 }, { "grad_norm": 0.5042910575866699, "learning_rate": 8.4618593789322e-05, "loss": 0.0772, "step": 23500 }, { "grad_norm": 0.46863624453544617, "learning_rate": 8.46036777570205e-05, "loss": 0.0668, "step": 23510 }, { "grad_norm": 0.5748704075813293, "learning_rate": 8.458875581189688e-05, "loss": 0.0494, "step": 23520 }, { "grad_norm": 0.5321336388587952, "learning_rate": 8.457382795650092e-05, "loss": 0.063, "step": 23530 }, { "grad_norm": 0.5218977928161621, "learning_rate": 8.455889419338335e-05, "loss": 0.0444, "step": 23540 }, { "grad_norm": 0.47055453062057495, "learning_rate": 8.454395452509593e-05, "loss": 0.0489, "step": 23550 }, { "grad_norm": 0.6565519571304321, "learning_rate": 8.452900895419146e-05, "loss": 0.0571, "step": 23560 }, { "grad_norm": 0.3631455898284912, "learning_rate": 8.451405748322376e-05, "loss": 0.0731, "step": 23570 }, { "grad_norm": 0.37103137373924255, "learning_rate": 8.449910011474759e-05, "loss": 0.0546, "step": 23580 }, { "grad_norm": 0.5445656776428223, "learning_rate": 8.448413685131876e-05, "loss": 0.0598, "step": 23590 }, { "grad_norm": 0.6625922322273254, "learning_rate": 8.446916769549407e-05, "loss": 0.0589, "step": 23600 }, { "grad_norm": 0.663645327091217, "learning_rate": 8.445419264983136e-05, "loss": 0.0742, "step": 23610 }, { "grad_norm": 0.6729543805122375, "learning_rate": 8.443921171688947e-05, "loss": 0.0689, "step": 23620 }, { "grad_norm": 0.7291763424873352, "learning_rate": 8.442422489922819e-05, "loss": 0.0773, "step": 23630 }, { "grad_norm": 0.6150374412536621, "learning_rate": 8.440923219940838e-05, "loss": 0.0568, "step": 23640 }, { "grad_norm": 0.6113916635513306, "learning_rate": 8.439423361999189e-05, "loss": 0.0508, "step": 23650 }, { "grad_norm": 0.686606228351593, "learning_rate": 8.437922916354155e-05, "loss": 0.094, "step": 23660 }, { "grad_norm": 0.5708193182945251, "learning_rate": 8.436421883262123e-05, "loss": 0.048, "step": 23670 }, { "grad_norm": 0.46887969970703125, "learning_rate": 8.434920262979577e-05, "loss": 0.0546, "step": 23680 }, { "grad_norm": 0.5539233088493347, "learning_rate": 8.433418055763104e-05, "loss": 0.0496, "step": 23690 }, { "grad_norm": 0.6689127087593079, "learning_rate": 8.431915261869389e-05, "loss": 0.0534, "step": 23700 }, { "grad_norm": 0.6790589094161987, "learning_rate": 8.43041188155522e-05, "loss": 0.0738, "step": 23710 }, { "grad_norm": 0.5265712738037109, "learning_rate": 8.428907915077481e-05, "loss": 0.0539, "step": 23720 }, { "grad_norm": 0.7497503161430359, "learning_rate": 8.42740336269316e-05, "loss": 0.0569, "step": 23730 }, { "grad_norm": 0.7796579599380493, "learning_rate": 8.425898224659345e-05, "loss": 0.0586, "step": 23740 }, { "grad_norm": 0.5725399851799011, "learning_rate": 8.42439250123322e-05, "loss": 0.0911, "step": 23750 }, { "grad_norm": 0.678447425365448, "learning_rate": 8.422886192672076e-05, "loss": 0.0508, "step": 23760 }, { "grad_norm": 0.37652331590652466, "learning_rate": 8.421379299233297e-05, "loss": 0.0464, "step": 23770 }, { "grad_norm": 0.6679572463035583, "learning_rate": 8.419871821174371e-05, "loss": 0.0544, "step": 23780 }, { "grad_norm": 0.5653269290924072, "learning_rate": 8.418363758752884e-05, "loss": 0.0547, "step": 23790 }, { "grad_norm": 0.6390396952629089, "learning_rate": 8.416855112226523e-05, "loss": 0.0591, "step": 23800 }, { "grad_norm": 0.6203336119651794, "learning_rate": 8.415345881853075e-05, "loss": 0.0611, "step": 23810 }, { "grad_norm": 0.7409602999687195, "learning_rate": 8.413836067890426e-05, "loss": 0.0551, "step": 23820 }, { "grad_norm": 0.5891879200935364, "learning_rate": 8.41232567059656e-05, "loss": 0.052, "step": 23830 }, { "grad_norm": 0.49388259649276733, "learning_rate": 8.410814690229565e-05, "loss": 0.0512, "step": 23840 }, { "grad_norm": 0.5603510737419128, "learning_rate": 8.409303127047626e-05, "loss": 0.0493, "step": 23850 }, { "grad_norm": 0.8212195634841919, "learning_rate": 8.407790981309028e-05, "loss": 0.0667, "step": 23860 }, { "grad_norm": 0.4634336233139038, "learning_rate": 8.406278253272153e-05, "loss": 0.0537, "step": 23870 }, { "grad_norm": 0.6150299906730652, "learning_rate": 8.404764943195487e-05, "loss": 0.0632, "step": 23880 }, { "grad_norm": 0.5312031507492065, "learning_rate": 8.403251051337613e-05, "loss": 0.0484, "step": 23890 }, { "grad_norm": 0.958672046661377, "learning_rate": 8.401736577957214e-05, "loss": 0.0565, "step": 23900 }, { "grad_norm": 0.6477164626121521, "learning_rate": 8.40022152331307e-05, "loss": 0.0597, "step": 23910 }, { "grad_norm": 0.5619948506355286, "learning_rate": 8.398705887664064e-05, "loss": 0.053, "step": 23920 }, { "grad_norm": 0.6298164129257202, "learning_rate": 8.397189671269177e-05, "loss": 0.0576, "step": 23930 }, { "grad_norm": 0.4107424318790436, "learning_rate": 8.395672874387488e-05, "loss": 0.0521, "step": 23940 }, { "grad_norm": 0.4368799030780792, "learning_rate": 8.394155497278177e-05, "loss": 0.0606, "step": 23950 }, { "grad_norm": 0.5185745358467102, "learning_rate": 8.392637540200523e-05, "loss": 0.0635, "step": 23960 }, { "grad_norm": 0.5549556612968445, "learning_rate": 8.391119003413902e-05, "loss": 0.0607, "step": 23970 }, { "grad_norm": 0.6164408922195435, "learning_rate": 8.38959988717779e-05, "loss": 0.0578, "step": 23980 }, { "grad_norm": 0.4902772307395935, "learning_rate": 8.388080191751764e-05, "loss": 0.0641, "step": 23990 }, { "grad_norm": 0.7203797698020935, "learning_rate": 8.386559917395496e-05, "loss": 0.0576, "step": 24000 }, { "grad_norm": 0.5941953659057617, "learning_rate": 8.385039064368761e-05, "loss": 0.0641, "step": 24010 }, { "grad_norm": 0.8926677107810974, "learning_rate": 8.383517632931431e-05, "loss": 0.0627, "step": 24020 }, { "grad_norm": 0.6805779933929443, "learning_rate": 8.381995623343477e-05, "loss": 0.0563, "step": 24030 }, { "grad_norm": 0.7268481850624084, "learning_rate": 8.380473035864968e-05, "loss": 0.057, "step": 24040 }, { "grad_norm": 0.6876078844070435, "learning_rate": 8.378949870756076e-05, "loss": 0.059, "step": 24050 }, { "grad_norm": 0.5758152008056641, "learning_rate": 8.377426128277063e-05, "loss": 0.044, "step": 24060 }, { "grad_norm": 0.9297917485237122, "learning_rate": 8.375901808688298e-05, "loss": 0.0677, "step": 24070 }, { "grad_norm": 0.6371437907218933, "learning_rate": 8.374376912250246e-05, "loss": 0.0708, "step": 24080 }, { "grad_norm": 0.6548125743865967, "learning_rate": 8.372851439223468e-05, "loss": 0.0503, "step": 24090 }, { "grad_norm": 0.511476457118988, "learning_rate": 8.371325389868627e-05, "loss": 0.0622, "step": 24100 }, { "grad_norm": 0.5076619386672974, "learning_rate": 8.369798764446482e-05, "loss": 0.0703, "step": 24110 }, { "grad_norm": 0.5850763916969299, "learning_rate": 8.368271563217893e-05, "loss": 0.0567, "step": 24120 }, { "grad_norm": 0.7657737731933594, "learning_rate": 8.366743786443817e-05, "loss": 0.0703, "step": 24130 }, { "grad_norm": 0.5319310426712036, "learning_rate": 8.365215434385309e-05, "loss": 0.0784, "step": 24140 }, { "grad_norm": 0.9230803847312927, "learning_rate": 8.36368650730352e-05, "loss": 0.0479, "step": 24150 }, { "grad_norm": 0.4910469651222229, "learning_rate": 8.362157005459705e-05, "loss": 0.0543, "step": 24160 }, { "grad_norm": 0.5511177182197571, "learning_rate": 8.360626929115213e-05, "loss": 0.0513, "step": 24170 }, { "grad_norm": 0.5857527852058411, "learning_rate": 8.359096278531492e-05, "loss": 0.0505, "step": 24180 }, { "grad_norm": 0.9097776412963867, "learning_rate": 8.357565053970088e-05, "loss": 0.0739, "step": 24190 }, { "grad_norm": 0.5545198917388916, "learning_rate": 8.356033255692647e-05, "loss": 0.06, "step": 24200 }, { "grad_norm": 0.7675286531448364, "learning_rate": 8.354500883960911e-05, "loss": 0.0543, "step": 24210 }, { "grad_norm": 0.531624436378479, "learning_rate": 8.352967939036717e-05, "loss": 0.047, "step": 24220 }, { "grad_norm": 0.4205546975135803, "learning_rate": 8.35143442118201e-05, "loss": 0.0411, "step": 24230 }, { "grad_norm": 0.6582947373390198, "learning_rate": 8.349900330658819e-05, "loss": 0.054, "step": 24240 }, { "grad_norm": 0.6030747294425964, "learning_rate": 8.348365667729284e-05, "loss": 0.0525, "step": 24250 }, { "grad_norm": 0.42922115325927734, "learning_rate": 8.346830432655633e-05, "loss": 0.0554, "step": 24260 }, { "grad_norm": 0.5717368721961975, "learning_rate": 8.345294625700195e-05, "loss": 0.0486, "step": 24270 }, { "grad_norm": 0.4574267566204071, "learning_rate": 8.343758247125402e-05, "loss": 0.0425, "step": 24280 }, { "grad_norm": 0.525412380695343, "learning_rate": 8.342221297193776e-05, "loss": 0.0507, "step": 24290 }, { "grad_norm": 0.9974703788757324, "learning_rate": 8.34068377616794e-05, "loss": 0.0491, "step": 24300 }, { "grad_norm": 0.500464916229248, "learning_rate": 8.339145684310615e-05, "loss": 0.0572, "step": 24310 }, { "grad_norm": 0.42500972747802734, "learning_rate": 8.337607021884618e-05, "loss": 0.0415, "step": 24320 }, { "grad_norm": 0.694705069065094, "learning_rate": 8.336067789152867e-05, "loss": 0.0493, "step": 24330 }, { "grad_norm": 0.6705901026725769, "learning_rate": 8.334527986378369e-05, "loss": 0.0566, "step": 24340 }, { "grad_norm": 0.6024774312973022, "learning_rate": 8.332987613824239e-05, "loss": 0.0585, "step": 24350 }, { "grad_norm": 0.8544214367866516, "learning_rate": 8.331446671753685e-05, "loss": 0.0776, "step": 24360 }, { "grad_norm": 0.5409985184669495, "learning_rate": 8.329905160430007e-05, "loss": 0.0507, "step": 24370 }, { "grad_norm": 0.9164049625396729, "learning_rate": 8.328363080116611e-05, "loss": 0.0619, "step": 24380 }, { "grad_norm": 0.511623203754425, "learning_rate": 8.326820431076997e-05, "loss": 0.0744, "step": 24390 }, { "grad_norm": 0.6539996862411499, "learning_rate": 8.325277213574759e-05, "loss": 0.059, "step": 24400 }, { "grad_norm": 0.6015225648880005, "learning_rate": 8.32373342787359e-05, "loss": 0.0954, "step": 24410 }, { "grad_norm": 0.4937165379524231, "learning_rate": 8.322189074237285e-05, "loss": 0.0517, "step": 24420 }, { "grad_norm": 0.518849790096283, "learning_rate": 8.32064415292973e-05, "loss": 0.0529, "step": 24430 }, { "grad_norm": 0.6164225339889526, "learning_rate": 8.319098664214907e-05, "loss": 0.0517, "step": 24440 }, { "grad_norm": 0.767159640789032, "learning_rate": 8.3175526083569e-05, "loss": 0.0524, "step": 24450 }, { "grad_norm": 0.6859629154205322, "learning_rate": 8.316005985619889e-05, "loss": 0.0544, "step": 24460 }, { "grad_norm": 0.5149483680725098, "learning_rate": 8.314458796268147e-05, "loss": 0.0505, "step": 24470 }, { "grad_norm": 0.5094712972640991, "learning_rate": 8.312911040566047e-05, "loss": 0.0519, "step": 24480 }, { "grad_norm": 0.5372845530509949, "learning_rate": 8.31136271877806e-05, "loss": 0.0752, "step": 24490 }, { "grad_norm": 0.48130497336387634, "learning_rate": 8.309813831168748e-05, "loss": 0.0565, "step": 24500 }, { "grad_norm": 0.38843590021133423, "learning_rate": 8.308264378002777e-05, "loss": 0.063, "step": 24510 }, { "grad_norm": 0.5769661664962769, "learning_rate": 8.306714359544906e-05, "loss": 0.0636, "step": 24520 }, { "grad_norm": 0.558430016040802, "learning_rate": 8.30516377605999e-05, "loss": 0.0563, "step": 24530 }, { "grad_norm": 0.5016301870346069, "learning_rate": 8.30361262781298e-05, "loss": 0.0612, "step": 24540 }, { "grad_norm": 0.5075740814208984, "learning_rate": 8.302060915068924e-05, "loss": 0.0482, "step": 24550 }, { "grad_norm": 0.47783833742141724, "learning_rate": 8.300508638092972e-05, "loss": 0.0504, "step": 24560 }, { "grad_norm": 0.5712860822677612, "learning_rate": 8.298955797150361e-05, "loss": 0.0651, "step": 24570 }, { "grad_norm": 0.5805451273918152, "learning_rate": 8.297402392506433e-05, "loss": 0.056, "step": 24580 }, { "grad_norm": 0.5868285894393921, "learning_rate": 8.295848424426617e-05, "loss": 0.0606, "step": 24590 }, { "grad_norm": 0.587644100189209, "learning_rate": 8.29429389317645e-05, "loss": 0.0605, "step": 24600 }, { "grad_norm": 0.6799212694168091, "learning_rate": 8.292738799021556e-05, "loss": 0.0549, "step": 24610 }, { "grad_norm": 0.5977173447608948, "learning_rate": 8.291183142227656e-05, "loss": 0.0452, "step": 24620 }, { "grad_norm": 0.4226408004760742, "learning_rate": 8.289626923060572e-05, "loss": 0.0611, "step": 24630 }, { "grad_norm": 0.585571825504303, "learning_rate": 8.288070141786218e-05, "loss": 0.0519, "step": 24640 }, { "grad_norm": 0.5037949085235596, "learning_rate": 8.286512798670605e-05, "loss": 0.0509, "step": 24650 }, { "grad_norm": 0.5013058185577393, "learning_rate": 8.284954893979842e-05, "loss": 0.0568, "step": 24660 }, { "grad_norm": 0.5839572548866272, "learning_rate": 8.283396427980131e-05, "loss": 0.056, "step": 24670 }, { "grad_norm": 0.6115303039550781, "learning_rate": 8.281837400937771e-05, "loss": 0.0653, "step": 24680 }, { "grad_norm": 0.5859594345092773, "learning_rate": 8.28027781311916e-05, "loss": 0.0598, "step": 24690 }, { "grad_norm": 0.6718143820762634, "learning_rate": 8.278717664790785e-05, "loss": 0.0682, "step": 24700 }, { "grad_norm": 0.4941374957561493, "learning_rate": 8.277156956219234e-05, "loss": 0.0838, "step": 24710 }, { "grad_norm": 0.7008110880851746, "learning_rate": 8.275595687671189e-05, "loss": 0.0468, "step": 24720 }, { "grad_norm": 0.502301812171936, "learning_rate": 8.27403385941343e-05, "loss": 0.0461, "step": 24730 }, { "grad_norm": 0.6134041547775269, "learning_rate": 8.272471471712828e-05, "loss": 0.0524, "step": 24740 }, { "grad_norm": 0.7758094072341919, "learning_rate": 8.270908524836355e-05, "loss": 0.0576, "step": 24750 }, { "grad_norm": 0.6961936354637146, "learning_rate": 8.269345019051074e-05, "loss": 0.0492, "step": 24760 }, { "grad_norm": 0.47630763053894043, "learning_rate": 8.267780954624147e-05, "loss": 0.056, "step": 24770 }, { "grad_norm": 0.6672959327697754, "learning_rate": 8.266216331822827e-05, "loss": 0.0682, "step": 24780 }, { "grad_norm": 0.8264604806900024, "learning_rate": 8.264651150914469e-05, "loss": 0.0709, "step": 24790 }, { "grad_norm": 0.7731257677078247, "learning_rate": 8.263085412166517e-05, "loss": 0.0648, "step": 24800 }, { "grad_norm": 0.543178379535675, "learning_rate": 8.261519115846514e-05, "loss": 0.0583, "step": 24810 }, { "grad_norm": 0.6337549686431885, "learning_rate": 8.259952262222096e-05, "loss": 0.0655, "step": 24820 }, { "grad_norm": 1.939210295677185, "learning_rate": 8.258384851560997e-05, "loss": 0.0521, "step": 24830 }, { "grad_norm": 1.084863543510437, "learning_rate": 8.256816884131044e-05, "loss": 0.0505, "step": 24840 }, { "grad_norm": 1.0270273685455322, "learning_rate": 8.255248360200159e-05, "loss": 0.0619, "step": 24850 }, { "grad_norm": 0.8156452775001526, "learning_rate": 8.253679280036359e-05, "loss": 0.0744, "step": 24860 }, { "grad_norm": 0.5198655724525452, "learning_rate": 8.252109643907762e-05, "loss": 0.0656, "step": 24870 }, { "grad_norm": 0.5813058018684387, "learning_rate": 8.250539452082569e-05, "loss": 0.0635, "step": 24880 }, { "grad_norm": 0.7952348589897156, "learning_rate": 8.248968704829087e-05, "loss": 0.061, "step": 24890 }, { "grad_norm": 0.5005083084106445, "learning_rate": 8.247397402415714e-05, "loss": 0.0444, "step": 24900 }, { "grad_norm": 0.49925240874290466, "learning_rate": 8.24582554511094e-05, "loss": 0.0486, "step": 24910 }, { "grad_norm": 0.5074062347412109, "learning_rate": 8.244253133183355e-05, "loss": 0.0732, "step": 24920 }, { "grad_norm": 0.4765985906124115, "learning_rate": 8.24268016690164e-05, "loss": 0.0521, "step": 24930 }, { "grad_norm": 0.4537082314491272, "learning_rate": 8.241106646534571e-05, "loss": 0.0463, "step": 24940 }, { "grad_norm": 0.5202295780181885, "learning_rate": 8.23953257235102e-05, "loss": 0.0443, "step": 24950 }, { "grad_norm": 0.6029784679412842, "learning_rate": 8.237957944619956e-05, "loss": 0.0429, "step": 24960 }, { "grad_norm": 0.7449153661727905, "learning_rate": 8.236382763610437e-05, "loss": 0.0536, "step": 24970 }, { "grad_norm": 0.5232363939285278, "learning_rate": 8.234807029591619e-05, "loss": 0.0574, "step": 24980 }, { "grad_norm": 0.5022523999214172, "learning_rate": 8.233230742832752e-05, "loss": 0.0648, "step": 24990 }, { "grad_norm": 0.7033634781837463, "learning_rate": 8.231653903603178e-05, "loss": 0.0563, "step": 25000 }, { "grad_norm": 0.7782810926437378, "learning_rate": 8.23007651217234e-05, "loss": 0.06, "step": 25010 }, { "grad_norm": 0.6049478054046631, "learning_rate": 8.228498568809769e-05, "loss": 0.077, "step": 25020 }, { "grad_norm": 0.577958881855011, "learning_rate": 8.22692007378509e-05, "loss": 0.0598, "step": 25030 }, { "grad_norm": 0.7532017827033997, "learning_rate": 8.225341027368028e-05, "loss": 0.0558, "step": 25040 }, { "grad_norm": 0.5371041893959045, "learning_rate": 8.223761429828399e-05, "loss": 0.0524, "step": 25050 }, { "grad_norm": 0.47950026392936707, "learning_rate": 8.22218128143611e-05, "loss": 0.0522, "step": 25060 }, { "grad_norm": 0.3876975476741791, "learning_rate": 8.220600582461166e-05, "loss": 0.053, "step": 25070 }, { "grad_norm": 0.766292929649353, "learning_rate": 8.219019333173668e-05, "loss": 0.0747, "step": 25080 }, { "grad_norm": 0.6987185478210449, "learning_rate": 8.217437533843805e-05, "loss": 0.0642, "step": 25090 }, { "grad_norm": 0.7290134429931641, "learning_rate": 8.215855184741867e-05, "loss": 0.0544, "step": 25100 }, { "grad_norm": 0.8204655647277832, "learning_rate": 8.21427228613823e-05, "loss": 0.0586, "step": 25110 }, { "grad_norm": 0.844447910785675, "learning_rate": 8.21268883830337e-05, "loss": 0.0556, "step": 25120 }, { "grad_norm": 0.4904274642467499, "learning_rate": 8.211104841507855e-05, "loss": 0.0614, "step": 25130 }, { "grad_norm": 0.47750622034072876, "learning_rate": 8.209520296022346e-05, "loss": 0.0491, "step": 25140 }, { "grad_norm": 0.5271225571632385, "learning_rate": 8.207935202117599e-05, "loss": 0.0503, "step": 25150 }, { "grad_norm": 0.42755326628685, "learning_rate": 8.206349560064463e-05, "loss": 0.0489, "step": 25160 }, { "grad_norm": 0.5187305808067322, "learning_rate": 8.204763370133881e-05, "loss": 0.082, "step": 25170 }, { "grad_norm": 0.47208428382873535, "learning_rate": 8.203176632596892e-05, "loss": 0.0476, "step": 25180 }, { "grad_norm": 0.6045830249786377, "learning_rate": 8.20158934772462e-05, "loss": 0.0455, "step": 25190 }, { "grad_norm": 0.4282638430595398, "learning_rate": 8.200001515788294e-05, "loss": 0.0535, "step": 25200 }, { "grad_norm": 0.5689007043838501, "learning_rate": 8.198413137059228e-05, "loss": 0.0443, "step": 25210 }, { "grad_norm": 0.5227928757667542, "learning_rate": 8.196824211808835e-05, "loss": 0.0573, "step": 25220 }, { "grad_norm": 0.46236100792884827, "learning_rate": 8.195234740308617e-05, "loss": 0.0434, "step": 25230 }, { "grad_norm": 0.8655121326446533, "learning_rate": 8.193644722830171e-05, "loss": 0.052, "step": 25240 }, { "grad_norm": 0.44673383235931396, "learning_rate": 8.19205415964519e-05, "loss": 0.0389, "step": 25250 }, { "grad_norm": 0.542629063129425, "learning_rate": 8.190463051025456e-05, "loss": 0.0416, "step": 25260 }, { "grad_norm": 0.7115727066993713, "learning_rate": 8.188871397242843e-05, "loss": 0.0584, "step": 25270 }, { "grad_norm": 0.47936609387397766, "learning_rate": 8.187279198569326e-05, "loss": 0.0727, "step": 25280 }, { "grad_norm": 0.5776700973510742, "learning_rate": 8.185686455276966e-05, "loss": 0.0639, "step": 25290 }, { "grad_norm": 0.524560272693634, "learning_rate": 8.184093167637921e-05, "loss": 0.0548, "step": 25300 }, { "grad_norm": 0.43159741163253784, "learning_rate": 8.182499335924437e-05, "loss": 0.0441, "step": 25310 }, { "grad_norm": 0.589410126209259, "learning_rate": 8.18090496040886e-05, "loss": 0.0859, "step": 25320 }, { "grad_norm": 0.5733356475830078, "learning_rate": 8.179310041363621e-05, "loss": 0.046, "step": 25330 }, { "grad_norm": 0.5726913213729858, "learning_rate": 8.17771457906125e-05, "loss": 0.0568, "step": 25340 }, { "grad_norm": 0.7073231935501099, "learning_rate": 8.176118573774371e-05, "loss": 0.05, "step": 25350 }, { "grad_norm": 0.6054450869560242, "learning_rate": 8.174522025775692e-05, "loss": 0.0686, "step": 25360 }, { "grad_norm": 0.5846814513206482, "learning_rate": 8.172924935338022e-05, "loss": 0.0592, "step": 25370 }, { "grad_norm": 0.7362290024757385, "learning_rate": 8.171327302734262e-05, "loss": 0.0739, "step": 25380 }, { "grad_norm": 0.49544286727905273, "learning_rate": 8.169729128237401e-05, "loss": 0.053, "step": 25390 }, { "grad_norm": 0.6581268310546875, "learning_rate": 8.168130412120525e-05, "loss": 0.0517, "step": 25400 }, { "grad_norm": 1.4225058555603027, "learning_rate": 8.16653115465681e-05, "loss": 0.0551, "step": 25410 }, { "grad_norm": 0.5898613333702087, "learning_rate": 8.164931356119526e-05, "loss": 0.0586, "step": 25420 }, { "grad_norm": 0.7016025185585022, "learning_rate": 8.163331016782032e-05, "loss": 0.0535, "step": 25430 }, { "grad_norm": 0.716249942779541, "learning_rate": 8.161730136917785e-05, "loss": 0.057, "step": 25440 }, { "grad_norm": 0.6657708287239075, "learning_rate": 8.160128716800333e-05, "loss": 0.0424, "step": 25450 }, { "grad_norm": 0.5291245579719543, "learning_rate": 8.158526756703313e-05, "loss": 0.0526, "step": 25460 }, { "grad_norm": 0.6583539843559265, "learning_rate": 8.156924256900455e-05, "loss": 0.0657, "step": 25470 }, { "grad_norm": 0.5680755376815796, "learning_rate": 8.155321217665584e-05, "loss": 0.0527, "step": 25480 }, { "grad_norm": 0.5707153677940369, "learning_rate": 8.153717639272614e-05, "loss": 0.0438, "step": 25490 }, { "grad_norm": 0.6601756811141968, "learning_rate": 8.152113521995555e-05, "loss": 0.0529, "step": 25500 }, { "grad_norm": 0.6251322627067566, "learning_rate": 8.150508866108505e-05, "loss": 0.0788, "step": 25510 }, { "grad_norm": 0.688053548336029, "learning_rate": 8.148903671885657e-05, "loss": 0.0651, "step": 25520 }, { "grad_norm": 0.6377131938934326, "learning_rate": 8.147297939601292e-05, "loss": 0.0828, "step": 25530 }, { "grad_norm": 0.6827023029327393, "learning_rate": 8.145691669529792e-05, "loss": 0.0545, "step": 25540 }, { "grad_norm": 0.6658210754394531, "learning_rate": 8.144084861945618e-05, "loss": 0.0663, "step": 25550 }, { "grad_norm": 0.5795366764068604, "learning_rate": 8.142477517123333e-05, "loss": 0.0617, "step": 25560 }, { "grad_norm": 0.8271415829658508, "learning_rate": 8.140869635337586e-05, "loss": 0.073, "step": 25570 }, { "grad_norm": 0.5655683875083923, "learning_rate": 8.139261216863123e-05, "loss": 0.0492, "step": 25580 }, { "grad_norm": 0.5952126979827881, "learning_rate": 8.137652261974776e-05, "loss": 0.0422, "step": 25590 }, { "grad_norm": 0.4883631765842438, "learning_rate": 8.136042770947472e-05, "loss": 0.0606, "step": 25600 }, { "grad_norm": 0.47580286860466003, "learning_rate": 8.134432744056228e-05, "loss": 0.0521, "step": 25610 }, { "grad_norm": 0.5164011716842651, "learning_rate": 8.132822181576158e-05, "loss": 0.0527, "step": 25620 }, { "grad_norm": 0.634967029094696, "learning_rate": 8.131211083782459e-05, "loss": 0.0554, "step": 25630 }, { "grad_norm": 0.5667328238487244, "learning_rate": 8.129599450950424e-05, "loss": 0.0511, "step": 25640 }, { "grad_norm": 0.8054759502410889, "learning_rate": 8.127987283355438e-05, "loss": 0.0524, "step": 25650 }, { "grad_norm": 0.6459496021270752, "learning_rate": 8.126374581272976e-05, "loss": 0.0586, "step": 25660 }, { "grad_norm": 0.42176708579063416, "learning_rate": 8.124761344978605e-05, "loss": 0.0426, "step": 25670 }, { "grad_norm": 0.37616652250289917, "learning_rate": 8.12314757474798e-05, "loss": 0.0591, "step": 25680 }, { "grad_norm": 0.6766027808189392, "learning_rate": 8.121533270856856e-05, "loss": 0.0506, "step": 25690 }, { "grad_norm": 0.4624263048171997, "learning_rate": 8.119918433581069e-05, "loss": 0.042, "step": 25700 }, { "grad_norm": 0.7827884554862976, "learning_rate": 8.118303063196551e-05, "loss": 0.0606, "step": 25710 }, { "grad_norm": 0.8496668338775635, "learning_rate": 8.116687159979326e-05, "loss": 0.0493, "step": 25720 }, { "grad_norm": 0.4315830171108246, "learning_rate": 8.115070724205508e-05, "loss": 0.0491, "step": 25730 }, { "grad_norm": 0.6851984858512878, "learning_rate": 8.113453756151296e-05, "loss": 0.0482, "step": 25740 }, { "grad_norm": 0.7144718766212463, "learning_rate": 8.111836256092995e-05, "loss": 0.0443, "step": 25750 }, { "grad_norm": 0.8847464919090271, "learning_rate": 8.110218224306985e-05, "loss": 0.054, "step": 25760 }, { "grad_norm": 0.7685368657112122, "learning_rate": 8.108599661069745e-05, "loss": 0.0697, "step": 25770 }, { "grad_norm": 0.5545656085014343, "learning_rate": 8.106980566657845e-05, "loss": 0.0657, "step": 25780 }, { "grad_norm": 0.7478880286216736, "learning_rate": 8.10536094134794e-05, "loss": 0.0658, "step": 25790 }, { "grad_norm": 0.6158279776573181, "learning_rate": 8.103740785416783e-05, "loss": 0.0627, "step": 25800 }, { "grad_norm": 0.8006550669670105, "learning_rate": 8.102120099141212e-05, "loss": 0.0647, "step": 25810 }, { "grad_norm": 0.49405089020729065, "learning_rate": 8.100498882798163e-05, "loss": 0.0417, "step": 25820 }, { "grad_norm": 0.5244649648666382, "learning_rate": 8.09887713666465e-05, "loss": 0.0484, "step": 25830 }, { "grad_norm": 0.7058099508285522, "learning_rate": 8.09725486101779e-05, "loss": 0.0532, "step": 25840 }, { "grad_norm": 0.5900492072105408, "learning_rate": 8.095632056134784e-05, "loss": 0.0521, "step": 25850 }, { "grad_norm": 0.39316603541374207, "learning_rate": 8.094008722292925e-05, "loss": 0.0436, "step": 25860 }, { "grad_norm": 0.45976927876472473, "learning_rate": 8.092384859769598e-05, "loss": 0.05, "step": 25870 }, { "grad_norm": 0.5837084054946899, "learning_rate": 8.090760468842275e-05, "loss": 0.0499, "step": 25880 }, { "grad_norm": 0.7529478669166565, "learning_rate": 8.089135549788521e-05, "loss": 0.0485, "step": 25890 }, { "grad_norm": 0.4333232641220093, "learning_rate": 8.087510102885987e-05, "loss": 0.039, "step": 25900 }, { "grad_norm": 0.8457750678062439, "learning_rate": 8.085884128412422e-05, "loss": 0.0531, "step": 25910 }, { "grad_norm": 0.8396565318107605, "learning_rate": 8.084257626645659e-05, "loss": 0.0527, "step": 25920 }, { "grad_norm": 0.5372565388679504, "learning_rate": 8.08263059786362e-05, "loss": 0.0495, "step": 25930 }, { "grad_norm": 0.6347891092300415, "learning_rate": 8.081003042344325e-05, "loss": 0.0558, "step": 25940 }, { "grad_norm": 0.521631121635437, "learning_rate": 8.079374960365872e-05, "loss": 0.0509, "step": 25950 }, { "grad_norm": 0.515491247177124, "learning_rate": 8.077746352206463e-05, "loss": 0.0505, "step": 25960 }, { "grad_norm": 0.9754478335380554, "learning_rate": 8.076117218144377e-05, "loss": 0.0616, "step": 25970 }, { "grad_norm": 0.7582111954689026, "learning_rate": 8.074487558457991e-05, "loss": 0.0552, "step": 25980 }, { "grad_norm": 0.47954121232032776, "learning_rate": 8.072857373425768e-05, "loss": 0.0537, "step": 25990 }, { "grad_norm": 0.4831122159957886, "learning_rate": 8.071226663326264e-05, "loss": 0.0472, "step": 26000 }, { "grad_norm": 0.8785590529441833, "learning_rate": 8.069595428438121e-05, "loss": 0.0557, "step": 26010 }, { "grad_norm": 0.6901411414146423, "learning_rate": 8.067963669040072e-05, "loss": 0.0673, "step": 26020 }, { "grad_norm": 1.5136687755584717, "learning_rate": 8.066331385410942e-05, "loss": 0.0522, "step": 26030 }, { "grad_norm": 0.6424103379249573, "learning_rate": 8.064698577829641e-05, "loss": 0.061, "step": 26040 }, { "grad_norm": 0.6062220931053162, "learning_rate": 8.063065246575175e-05, "loss": 0.0688, "step": 26050 }, { "grad_norm": 0.5052341818809509, "learning_rate": 8.061431391926631e-05, "loss": 0.0571, "step": 26060 }, { "grad_norm": 0.5174107551574707, "learning_rate": 8.059797014163195e-05, "loss": 0.0428, "step": 26070 }, { "grad_norm": 0.681979238986969, "learning_rate": 8.058162113564133e-05, "loss": 0.0548, "step": 26080 }, { "grad_norm": 0.6446349024772644, "learning_rate": 8.056526690408806e-05, "loss": 0.044, "step": 26090 }, { "grad_norm": 0.4459543526172638, "learning_rate": 8.054890744976666e-05, "loss": 0.044, "step": 26100 }, { "grad_norm": 0.525368332862854, "learning_rate": 8.053254277547248e-05, "loss": 0.0494, "step": 26110 }, { "grad_norm": 0.4953129291534424, "learning_rate": 8.051617288400182e-05, "loss": 0.0503, "step": 26120 }, { "grad_norm": 0.891243040561676, "learning_rate": 8.049979777815182e-05, "loss": 0.0531, "step": 26130 }, { "grad_norm": 0.5126921534538269, "learning_rate": 8.048341746072054e-05, "loss": 0.0642, "step": 26140 }, { "grad_norm": 0.5870969295501709, "learning_rate": 8.046703193450696e-05, "loss": 0.0449, "step": 26150 }, { "grad_norm": 0.5793577432632446, "learning_rate": 8.04506412023109e-05, "loss": 0.0601, "step": 26160 }, { "grad_norm": 0.5778521299362183, "learning_rate": 8.043424526693306e-05, "loss": 0.0464, "step": 26170 }, { "grad_norm": 0.6208720207214355, "learning_rate": 8.04178441311751e-05, "loss": 0.0711, "step": 26180 }, { "grad_norm": 0.4601674973964691, "learning_rate": 8.04014377978395e-05, "loss": 0.0586, "step": 26190 }, { "grad_norm": 0.6746564507484436, "learning_rate": 8.038502626972967e-05, "loss": 0.0566, "step": 26200 }, { "grad_norm": 0.45926135778427124, "learning_rate": 8.036860954964989e-05, "loss": 0.0441, "step": 26210 }, { "grad_norm": 0.5555881857872009, "learning_rate": 8.035218764040531e-05, "loss": 0.0455, "step": 26220 }, { "grad_norm": 0.7983447909355164, "learning_rate": 8.033576054480199e-05, "loss": 0.0439, "step": 26230 }, { "grad_norm": 0.8004482388496399, "learning_rate": 8.031932826564688e-05, "loss": 0.0481, "step": 26240 }, { "grad_norm": 0.7474438548088074, "learning_rate": 8.030289080574782e-05, "loss": 0.0821, "step": 26250 }, { "grad_norm": 0.6048551797866821, "learning_rate": 8.028644816791349e-05, "loss": 0.0585, "step": 26260 }, { "grad_norm": 0.5713956356048584, "learning_rate": 8.027000035495351e-05, "loss": 0.0522, "step": 26270 }, { "grad_norm": 0.6029292345046997, "learning_rate": 8.025354736967836e-05, "loss": 0.053, "step": 26280 }, { "grad_norm": 0.7069141864776611, "learning_rate": 8.023708921489941e-05, "loss": 0.055, "step": 26290 }, { "grad_norm": 0.47170647978782654, "learning_rate": 8.022062589342887e-05, "loss": 0.0552, "step": 26300 }, { "grad_norm": 0.6971204876899719, "learning_rate": 8.020415740807993e-05, "loss": 0.0399, "step": 26310 }, { "grad_norm": 0.44881296157836914, "learning_rate": 8.018768376166656e-05, "loss": 0.043, "step": 26320 }, { "grad_norm": 0.5898566842079163, "learning_rate": 8.017120495700368e-05, "loss": 0.069, "step": 26330 }, { "grad_norm": 0.6261188983917236, "learning_rate": 8.015472099690704e-05, "loss": 0.0397, "step": 26340 }, { "grad_norm": 0.8016960024833679, "learning_rate": 8.013823188419332e-05, "loss": 0.0698, "step": 26350 }, { "grad_norm": 0.4620378613471985, "learning_rate": 8.012173762168006e-05, "loss": 0.0403, "step": 26360 }, { "grad_norm": 1.6243385076522827, "learning_rate": 8.010523821218567e-05, "loss": 0.044, "step": 26370 }, { "grad_norm": 0.41259077191352844, "learning_rate": 8.008873365852945e-05, "loss": 0.0394, "step": 26380 }, { "grad_norm": 0.5544365048408508, "learning_rate": 8.007222396353157e-05, "loss": 0.0462, "step": 26390 }, { "grad_norm": 0.61443692445755, "learning_rate": 8.00557091300131e-05, "loss": 0.0758, "step": 26400 }, { "grad_norm": 0.5648439526557922, "learning_rate": 8.003918916079597e-05, "loss": 0.0594, "step": 26410 }, { "grad_norm": 0.5670267343521118, "learning_rate": 8.002266405870298e-05, "loss": 0.0451, "step": 26420 }, { "grad_norm": 0.7045540809631348, "learning_rate": 8.000613382655782e-05, "loss": 0.0623, "step": 26430 }, { "grad_norm": 0.7228793501853943, "learning_rate": 7.998959846718505e-05, "loss": 0.0483, "step": 26440 }, { "grad_norm": 0.6329235434532166, "learning_rate": 7.997305798341012e-05, "loss": 0.0466, "step": 26450 }, { "grad_norm": 0.46921953558921814, "learning_rate": 7.995651237805937e-05, "loss": 0.0457, "step": 26460 }, { "grad_norm": 0.666300892829895, "learning_rate": 7.993996165395996e-05, "loss": 0.059, "step": 26470 }, { "grad_norm": 0.6902782320976257, "learning_rate": 7.992340581393996e-05, "loss": 0.0566, "step": 26480 }, { "grad_norm": 0.3583124577999115, "learning_rate": 7.990684486082831e-05, "loss": 0.0533, "step": 26490 }, { "grad_norm": 0.6715186238288879, "learning_rate": 7.989027879745482e-05, "loss": 0.0485, "step": 26500 }, { "grad_norm": 0.7952657341957092, "learning_rate": 7.98737076266502e-05, "loss": 0.0462, "step": 26510 }, { "grad_norm": 0.47797930240631104, "learning_rate": 7.985713135124598e-05, "loss": 0.054, "step": 26520 }, { "grad_norm": 0.7740814685821533, "learning_rate": 7.98405499740746e-05, "loss": 0.0676, "step": 26530 }, { "grad_norm": 0.5657539963722229, "learning_rate": 7.98239634979694e-05, "loss": 0.0452, "step": 26540 }, { "grad_norm": 0.7787570357322693, "learning_rate": 7.98073719257645e-05, "loss": 0.0509, "step": 26550 }, { "grad_norm": 0.5954111218452454, "learning_rate": 7.979077526029499e-05, "loss": 0.0561, "step": 26560 }, { "grad_norm": 0.5211132764816284, "learning_rate": 7.977417350439675e-05, "loss": 0.0447, "step": 26570 }, { "grad_norm": 0.4983132481575012, "learning_rate": 7.97575666609066e-05, "loss": 0.0502, "step": 26580 }, { "grad_norm": 0.813994288444519, "learning_rate": 7.974095473266216e-05, "loss": 0.0447, "step": 26590 }, { "grad_norm": 0.5312252044677734, "learning_rate": 7.972433772250198e-05, "loss": 0.0508, "step": 26600 }, { "grad_norm": 0.47847414016723633, "learning_rate": 7.970771563326544e-05, "loss": 0.052, "step": 26610 }, { "grad_norm": 0.5464851260185242, "learning_rate": 7.96910884677928e-05, "loss": 0.0629, "step": 26620 }, { "grad_norm": 0.5391753315925598, "learning_rate": 7.967445622892523e-05, "loss": 0.0584, "step": 26630 }, { "grad_norm": 0.6593781113624573, "learning_rate": 7.965781891950465e-05, "loss": 0.0692, "step": 26640 }, { "grad_norm": 0.7802566885948181, "learning_rate": 7.964117654237397e-05, "loss": 0.0574, "step": 26650 }, { "grad_norm": 0.7598226070404053, "learning_rate": 7.962452910037692e-05, "loss": 0.0558, "step": 26660 }, { "grad_norm": 0.5680957436561584, "learning_rate": 7.96078765963581e-05, "loss": 0.0603, "step": 26670 }, { "grad_norm": 0.6176034808158875, "learning_rate": 7.95912190331629e-05, "loss": 0.0508, "step": 26680 }, { "grad_norm": 0.5872812867164612, "learning_rate": 7.957455641363772e-05, "loss": 0.0463, "step": 26690 }, { "grad_norm": 0.46297258138656616, "learning_rate": 7.955788874062968e-05, "loss": 0.0593, "step": 26700 }, { "grad_norm": 0.49740561842918396, "learning_rate": 7.95412160169869e-05, "loss": 0.0532, "step": 26710 }, { "grad_norm": 0.5247524976730347, "learning_rate": 7.952453824555824e-05, "loss": 0.0625, "step": 26720 }, { "grad_norm": 0.711382269859314, "learning_rate": 7.95078554291935e-05, "loss": 0.0639, "step": 26730 }, { "grad_norm": 0.5292142629623413, "learning_rate": 7.94911675707433e-05, "loss": 0.058, "step": 26740 }, { "grad_norm": 0.5646471977233887, "learning_rate": 7.947447467305915e-05, "loss": 0.0462, "step": 26750 }, { "grad_norm": 0.4308475852012634, "learning_rate": 7.94577767389934e-05, "loss": 0.0625, "step": 26760 }, { "grad_norm": 0.5552541017532349, "learning_rate": 7.944107377139928e-05, "loss": 0.0677, "step": 26770 }, { "grad_norm": 0.4238233268260956, "learning_rate": 7.942436577313088e-05, "loss": 0.0656, "step": 26780 }, { "grad_norm": 0.5616858601570129, "learning_rate": 7.940765274704312e-05, "loss": 0.062, "step": 26790 }, { "grad_norm": 1.109894871711731, "learning_rate": 7.939093469599181e-05, "loss": 0.0612, "step": 26800 }, { "grad_norm": 0.5722556114196777, "learning_rate": 7.93742116228336e-05, "loss": 0.0547, "step": 26810 }, { "grad_norm": 0.4427660405635834, "learning_rate": 7.935748353042602e-05, "loss": 0.0538, "step": 26820 }, { "grad_norm": 0.48043200373649597, "learning_rate": 7.934075042162744e-05, "loss": 0.0561, "step": 26830 }, { "grad_norm": 0.5112820863723755, "learning_rate": 7.932401229929705e-05, "loss": 0.0494, "step": 26840 }, { "grad_norm": 0.6049957871437073, "learning_rate": 7.9307269166295e-05, "loss": 0.0429, "step": 26850 }, { "grad_norm": 0.5652236342430115, "learning_rate": 7.92905210254822e-05, "loss": 0.0534, "step": 26860 }, { "grad_norm": 0.3457721769809723, "learning_rate": 7.927376787972045e-05, "loss": 0.0498, "step": 26870 }, { "grad_norm": 0.54903644323349, "learning_rate": 7.92570097318724e-05, "loss": 0.0503, "step": 26880 }, { "grad_norm": 0.8588812351226807, "learning_rate": 7.924024658480158e-05, "loss": 0.0467, "step": 26890 }, { "grad_norm": 0.6685551404953003, "learning_rate": 7.922347844137233e-05, "loss": 0.064, "step": 26900 }, { "grad_norm": 0.958220362663269, "learning_rate": 7.92067053044499e-05, "loss": 0.0785, "step": 26910 }, { "grad_norm": 0.546288788318634, "learning_rate": 7.918992717690031e-05, "loss": 0.0531, "step": 26920 }, { "grad_norm": 1.857681155204773, "learning_rate": 7.917314406159053e-05, "loss": 0.1096, "step": 26930 }, { "grad_norm": 0.5870792269706726, "learning_rate": 7.915635596138832e-05, "loss": 0.0569, "step": 26940 }, { "grad_norm": 0.6670536398887634, "learning_rate": 7.913956287916228e-05, "loss": 0.063, "step": 26950 }, { "grad_norm": 0.5605659484863281, "learning_rate": 7.912276481778193e-05, "loss": 0.0651, "step": 26960 }, { "grad_norm": 0.5154249668121338, "learning_rate": 7.910596178011759e-05, "loss": 0.0624, "step": 26970 }, { "grad_norm": 0.5214511752128601, "learning_rate": 7.908915376904043e-05, "loss": 0.049, "step": 26980 }, { "grad_norm": 0.5164859294891357, "learning_rate": 7.907234078742247e-05, "loss": 0.0401, "step": 26990 }, { "grad_norm": 0.5708712935447693, "learning_rate": 7.90555228381366e-05, "loss": 0.0455, "step": 27000 }, { "grad_norm": 0.9746940732002258, "learning_rate": 7.903869992405656e-05, "loss": 0.0664, "step": 27010 }, { "grad_norm": 0.48380374908447266, "learning_rate": 7.902187204805691e-05, "loss": 0.0489, "step": 27020 }, { "grad_norm": 0.5672171115875244, "learning_rate": 7.900503921301308e-05, "loss": 0.0676, "step": 27030 }, { "grad_norm": 0.4751311242580414, "learning_rate": 7.898820142180133e-05, "loss": 0.0606, "step": 27040 }, { "grad_norm": 0.6418725252151489, "learning_rate": 7.897135867729879e-05, "loss": 0.0706, "step": 27050 }, { "grad_norm": 0.6546995639801025, "learning_rate": 7.89545109823834e-05, "loss": 0.0507, "step": 27060 }, { "grad_norm": 0.4401167333126068, "learning_rate": 7.8937658339934e-05, "loss": 0.0706, "step": 27070 }, { "grad_norm": 0.4125717878341675, "learning_rate": 7.892080075283026e-05, "loss": 0.0594, "step": 27080 }, { "grad_norm": 0.5766555070877075, "learning_rate": 7.890393822395263e-05, "loss": 0.0502, "step": 27090 }, { "grad_norm": 0.4124499559402466, "learning_rate": 7.88870707561825e-05, "loss": 0.0452, "step": 27100 }, { "grad_norm": 0.4881872236728668, "learning_rate": 7.887019835240203e-05, "loss": 0.0705, "step": 27110 }, { "grad_norm": 0.7108585834503174, "learning_rate": 7.885332101549427e-05, "loss": 0.0596, "step": 27120 }, { "grad_norm": 0.6403623819351196, "learning_rate": 7.883643874834308e-05, "loss": 0.0571, "step": 27130 }, { "grad_norm": 0.42752018570899963, "learning_rate": 7.881955155383321e-05, "loss": 0.0478, "step": 27140 }, { "grad_norm": 0.5165519118309021, "learning_rate": 7.880265943485017e-05, "loss": 0.0473, "step": 27150 }, { "grad_norm": 0.5588263869285583, "learning_rate": 7.878576239428038e-05, "loss": 0.0483, "step": 27160 }, { "grad_norm": 0.5842504501342773, "learning_rate": 7.87688604350111e-05, "loss": 0.0732, "step": 27170 }, { "grad_norm": 0.5761439204216003, "learning_rate": 7.875195355993042e-05, "loss": 0.0479, "step": 27180 }, { "grad_norm": 0.5273188948631287, "learning_rate": 7.873504177192724e-05, "loss": 0.0794, "step": 27190 }, { "grad_norm": 0.524787187576294, "learning_rate": 7.87181250738913e-05, "loss": 0.0565, "step": 27200 }, { "grad_norm": 0.7384615540504456, "learning_rate": 7.870120346871324e-05, "loss": 0.0526, "step": 27210 }, { "grad_norm": 0.5823633074760437, "learning_rate": 7.86842769592845e-05, "loss": 0.0407, "step": 27220 }, { "grad_norm": 0.5197446942329407, "learning_rate": 7.866734554849732e-05, "loss": 0.0547, "step": 27230 }, { "grad_norm": 0.5304838418960571, "learning_rate": 7.865040923924486e-05, "loss": 0.0802, "step": 27240 }, { "grad_norm": 0.4367210268974304, "learning_rate": 7.863346803442104e-05, "loss": 0.0486, "step": 27250 }, { "grad_norm": 0.5447648167610168, "learning_rate": 7.861652193692067e-05, "loss": 0.0485, "step": 27260 }, { "grad_norm": 0.41351377964019775, "learning_rate": 7.859957094963937e-05, "loss": 0.0431, "step": 27270 }, { "grad_norm": 0.5202351212501526, "learning_rate": 7.858261507547357e-05, "loss": 0.0592, "step": 27280 }, { "grad_norm": 0.47332969307899475, "learning_rate": 7.856565431732061e-05, "loss": 0.0483, "step": 27290 }, { "grad_norm": 0.5733317732810974, "learning_rate": 7.854868867807859e-05, "loss": 0.0693, "step": 27300 }, { "grad_norm": 0.4352481961250305, "learning_rate": 7.85317181606465e-05, "loss": 0.059, "step": 27310 }, { "grad_norm": 0.4983659088611603, "learning_rate": 7.85147427679241e-05, "loss": 0.0622, "step": 27320 }, { "grad_norm": 0.4521951675415039, "learning_rate": 7.849776250281205e-05, "loss": 0.0514, "step": 27330 }, { "grad_norm": 0.6328049898147583, "learning_rate": 7.84807773682118e-05, "loss": 0.0616, "step": 27340 }, { "grad_norm": 0.7539600729942322, "learning_rate": 7.846378736702565e-05, "loss": 0.0641, "step": 27350 }, { "grad_norm": 0.40834343433380127, "learning_rate": 7.844679250215671e-05, "loss": 0.0445, "step": 27360 }, { "grad_norm": 0.6073178052902222, "learning_rate": 7.842979277650898e-05, "loss": 0.0496, "step": 27370 }, { "grad_norm": 0.7784979939460754, "learning_rate": 7.84127881929872e-05, "loss": 0.0638, "step": 27380 }, { "grad_norm": 0.4332987070083618, "learning_rate": 7.839577875449704e-05, "loss": 0.0527, "step": 27390 }, { "grad_norm": 0.5874393582344055, "learning_rate": 7.837876446394489e-05, "loss": 0.0551, "step": 27400 }, { "grad_norm": 0.49755650758743286, "learning_rate": 7.836174532423805e-05, "loss": 0.0452, "step": 27410 }, { "grad_norm": 0.5590549111366272, "learning_rate": 7.834472133828466e-05, "loss": 0.0596, "step": 27420 }, { "grad_norm": 0.5934829711914062, "learning_rate": 7.832769250899359e-05, "loss": 0.0622, "step": 27430 }, { "grad_norm": 0.503244936466217, "learning_rate": 7.831065883927464e-05, "loss": 0.0445, "step": 27440 }, { "grad_norm": 0.4162977933883667, "learning_rate": 7.829362033203841e-05, "loss": 0.0428, "step": 27450 }, { "grad_norm": 0.6851633191108704, "learning_rate": 7.827657699019628e-05, "loss": 0.0448, "step": 27460 }, { "grad_norm": 0.7481182813644409, "learning_rate": 7.825952881666052e-05, "loss": 0.0522, "step": 27470 }, { "grad_norm": 0.38910794258117676, "learning_rate": 7.824247581434418e-05, "loss": 0.0518, "step": 27480 }, { "grad_norm": 0.7245668172836304, "learning_rate": 7.822541798616116e-05, "loss": 0.0579, "step": 27490 }, { "grad_norm": 0.523284912109375, "learning_rate": 7.820835533502617e-05, "loss": 0.0469, "step": 27500 }, { "grad_norm": 0.47811001539230347, "learning_rate": 7.819128786385475e-05, "loss": 0.0566, "step": 27510 }, { "grad_norm": 0.7239187955856323, "learning_rate": 7.817421557556329e-05, "loss": 0.0936, "step": 27520 }, { "grad_norm": 0.3836422562599182, "learning_rate": 7.815713847306893e-05, "loss": 0.0423, "step": 27530 }, { "grad_norm": 0.7077396512031555, "learning_rate": 7.81400565592897e-05, "loss": 0.059, "step": 27540 }, { "grad_norm": 0.40957921743392944, "learning_rate": 7.812296983714444e-05, "loss": 0.0514, "step": 27550 }, { "grad_norm": 0.7060549855232239, "learning_rate": 7.810587830955281e-05, "loss": 0.0559, "step": 27560 }, { "grad_norm": 0.6667201519012451, "learning_rate": 7.808878197943528e-05, "loss": 0.0525, "step": 27570 }, { "grad_norm": 0.5481546521186829, "learning_rate": 7.807168084971312e-05, "loss": 0.0548, "step": 27580 }, { "grad_norm": 0.5251937508583069, "learning_rate": 7.805457492330849e-05, "loss": 0.0482, "step": 27590 }, { "grad_norm": 0.4940166473388672, "learning_rate": 7.803746420314428e-05, "loss": 0.0525, "step": 27600 }, { "grad_norm": 0.41673535108566284, "learning_rate": 7.802034869214428e-05, "loss": 0.0625, "step": 27610 }, { "grad_norm": 0.726103663444519, "learning_rate": 7.800322839323303e-05, "loss": 0.0643, "step": 27620 }, { "grad_norm": 0.5714308023452759, "learning_rate": 7.798610330933593e-05, "loss": 0.0552, "step": 27630 }, { "grad_norm": 0.5683056712150574, "learning_rate": 7.796897344337922e-05, "loss": 0.0792, "step": 27640 }, { "grad_norm": 0.43696367740631104, "learning_rate": 7.795183879828989e-05, "loss": 0.0543, "step": 27650 }, { "grad_norm": 0.5230207443237305, "learning_rate": 7.793469937699579e-05, "loss": 0.0509, "step": 27660 }, { "grad_norm": 0.5043750405311584, "learning_rate": 7.791755518242558e-05, "loss": 0.0684, "step": 27670 }, { "grad_norm": 0.6022247672080994, "learning_rate": 7.790040621750876e-05, "loss": 0.0537, "step": 27680 }, { "grad_norm": 0.45675474405288696, "learning_rate": 7.788325248517558e-05, "loss": 0.042, "step": 27690 }, { "grad_norm": 0.5960133075714111, "learning_rate": 7.786609398835715e-05, "loss": 0.0572, "step": 27700 }, { "grad_norm": 0.4533807039260864, "learning_rate": 7.784893072998541e-05, "loss": 0.0569, "step": 27710 }, { "grad_norm": 0.5842493772506714, "learning_rate": 7.783176271299306e-05, "loss": 0.0651, "step": 27720 }, { "grad_norm": 0.5611975193023682, "learning_rate": 7.781458994031368e-05, "loss": 0.0632, "step": 27730 }, { "grad_norm": 0.5355972051620483, "learning_rate": 7.779741241488161e-05, "loss": 0.0532, "step": 27740 }, { "grad_norm": 0.6637497544288635, "learning_rate": 7.7780230139632e-05, "loss": 0.0629, "step": 27750 }, { "grad_norm": 0.45654526352882385, "learning_rate": 7.776304311750087e-05, "loss": 0.0411, "step": 27760 }, { "grad_norm": 0.7136979103088379, "learning_rate": 7.7745851351425e-05, "loss": 0.0528, "step": 27770 }, { "grad_norm": 0.7655240297317505, "learning_rate": 7.772865484434197e-05, "loss": 0.0647, "step": 27780 }, { "grad_norm": 0.5169385671615601, "learning_rate": 7.77114535991902e-05, "loss": 0.0429, "step": 27790 }, { "grad_norm": 0.5160502195358276, "learning_rate": 7.769424761890893e-05, "loss": 0.0544, "step": 27800 }, { "grad_norm": 0.6278141140937805, "learning_rate": 7.767703690643817e-05, "loss": 0.06, "step": 27810 }, { "grad_norm": 0.4988095164299011, "learning_rate": 7.76598214647188e-05, "loss": 0.0477, "step": 27820 }, { "grad_norm": 0.43379512429237366, "learning_rate": 7.764260129669241e-05, "loss": 0.0612, "step": 27830 }, { "grad_norm": 0.5213671922683716, "learning_rate": 7.76253764053015e-05, "loss": 0.0545, "step": 27840 }, { "grad_norm": 0.7467885613441467, "learning_rate": 7.760814679348932e-05, "loss": 0.0534, "step": 27850 }, { "grad_norm": 0.6857855916023254, "learning_rate": 7.759091246419992e-05, "loss": 0.0584, "step": 27860 }, { "grad_norm": 0.44964244961738586, "learning_rate": 7.75736734203782e-05, "loss": 0.0602, "step": 27870 }, { "grad_norm": 0.4742697477340698, "learning_rate": 7.755642966496985e-05, "loss": 0.0476, "step": 27880 }, { "grad_norm": 0.4354233741760254, "learning_rate": 7.753918120092132e-05, "loss": 0.0474, "step": 27890 }, { "grad_norm": 0.6445034742355347, "learning_rate": 7.752192803117993e-05, "loss": 0.062, "step": 27900 }, { "grad_norm": 0.556145191192627, "learning_rate": 7.750467015869377e-05, "loss": 0.055, "step": 27910 }, { "grad_norm": 0.5817632675170898, "learning_rate": 7.748740758641174e-05, "loss": 0.061, "step": 27920 }, { "grad_norm": 0.44095635414123535, "learning_rate": 7.74701403172835e-05, "loss": 0.06, "step": 27930 }, { "grad_norm": 0.8861122727394104, "learning_rate": 7.745286835425962e-05, "loss": 0.0856, "step": 27940 }, { "grad_norm": 0.822792649269104, "learning_rate": 7.743559170029138e-05, "loss": 0.0554, "step": 27950 }, { "grad_norm": 0.5078284740447998, "learning_rate": 7.741831035833087e-05, "loss": 0.0716, "step": 27960 }, { "grad_norm": 0.6617388725280762, "learning_rate": 7.740102433133102e-05, "loss": 0.0772, "step": 27970 }, { "grad_norm": 0.43832752108573914, "learning_rate": 7.738373362224553e-05, "loss": 0.0544, "step": 27980 }, { "grad_norm": 0.6619330644607544, "learning_rate": 7.73664382340289e-05, "loss": 0.0556, "step": 27990 }, { "grad_norm": 0.5857419371604919, "learning_rate": 7.734913816963647e-05, "loss": 0.053, "step": 28000 }, { "grad_norm": 0.4810680150985718, "learning_rate": 7.73318334320243e-05, "loss": 0.0566, "step": 28010 }, { "grad_norm": 0.5566508769989014, "learning_rate": 7.731452402414934e-05, "loss": 0.0491, "step": 28020 }, { "grad_norm": 0.8637070655822754, "learning_rate": 7.729720994896928e-05, "loss": 0.06, "step": 28030 }, { "grad_norm": 0.6585515141487122, "learning_rate": 7.727989120944262e-05, "loss": 0.0649, "step": 28040 }, { "grad_norm": 0.8398230075836182, "learning_rate": 7.726256780852865e-05, "loss": 0.061, "step": 28050 }, { "grad_norm": 0.786975085735321, "learning_rate": 7.724523974918749e-05, "loss": 0.0577, "step": 28060 }, { "grad_norm": 1.1509703397750854, "learning_rate": 7.722790703438002e-05, "loss": 0.0461, "step": 28070 }, { "grad_norm": 0.8004499673843384, "learning_rate": 7.72105696670679e-05, "loss": 0.0496, "step": 28080 }, { "grad_norm": 0.5986136794090271, "learning_rate": 7.719322765021364e-05, "loss": 0.0503, "step": 28090 }, { "grad_norm": 0.30809444189071655, "learning_rate": 7.717588098678051e-05, "loss": 0.0418, "step": 28100 }, { "grad_norm": 0.5883771181106567, "learning_rate": 7.715852967973258e-05, "loss": 0.0527, "step": 28110 }, { "grad_norm": 0.43505576252937317, "learning_rate": 7.714117373203474e-05, "loss": 0.0517, "step": 28120 }, { "grad_norm": 0.6085708737373352, "learning_rate": 7.712381314665259e-05, "loss": 0.0589, "step": 28130 }, { "grad_norm": 0.6336249113082886, "learning_rate": 7.710644792655261e-05, "loss": 0.0418, "step": 28140 }, { "grad_norm": 0.5679900050163269, "learning_rate": 7.708907807470207e-05, "loss": 0.0598, "step": 28150 }, { "grad_norm": 0.5910090804100037, "learning_rate": 7.707170359406896e-05, "loss": 0.0425, "step": 28160 }, { "grad_norm": 0.4542761743068695, "learning_rate": 7.705432448762213e-05, "loss": 0.0524, "step": 28170 }, { "grad_norm": 0.5032370686531067, "learning_rate": 7.703694075833117e-05, "loss": 0.0445, "step": 28180 }, { "grad_norm": 0.833810567855835, "learning_rate": 7.70195524091665e-05, "loss": 0.0564, "step": 28190 }, { "grad_norm": 0.594531774520874, "learning_rate": 7.70021594430993e-05, "loss": 0.0573, "step": 28200 }, { "grad_norm": 0.5144063830375671, "learning_rate": 7.698476186310157e-05, "loss": 0.0559, "step": 28210 }, { "grad_norm": 0.8057174682617188, "learning_rate": 7.696735967214608e-05, "loss": 0.0717, "step": 28220 }, { "grad_norm": 0.4129098951816559, "learning_rate": 7.694995287320636e-05, "loss": 0.0394, "step": 28230 }, { "grad_norm": 0.5348847508430481, "learning_rate": 7.693254146925679e-05, "loss": 0.0583, "step": 28240 }, { "grad_norm": 0.41581425070762634, "learning_rate": 7.691512546327251e-05, "loss": 0.0382, "step": 28250 }, { "grad_norm": 0.6092197299003601, "learning_rate": 7.689770485822939e-05, "loss": 0.0712, "step": 28260 }, { "grad_norm": 0.7998915314674377, "learning_rate": 7.688027965710416e-05, "loss": 0.065, "step": 28270 }, { "grad_norm": 0.43888458609580994, "learning_rate": 7.686284986287433e-05, "loss": 0.0547, "step": 28280 }, { "grad_norm": 0.3971706032752991, "learning_rate": 7.684541547851817e-05, "loss": 0.0392, "step": 28290 }, { "grad_norm": 0.5457953214645386, "learning_rate": 7.68279765070147e-05, "loss": 0.058, "step": 28300 }, { "grad_norm": 0.9437524676322937, "learning_rate": 7.68105329513438e-05, "loss": 0.0807, "step": 28310 }, { "grad_norm": 0.89663165807724, "learning_rate": 7.67930848144861e-05, "loss": 0.0654, "step": 28320 }, { "grad_norm": 0.42750486731529236, "learning_rate": 7.6775632099423e-05, "loss": 0.0495, "step": 28330 }, { "grad_norm": 0.7481943964958191, "learning_rate": 7.675817480913667e-05, "loss": 0.0622, "step": 28340 }, { "grad_norm": 0.7533428072929382, "learning_rate": 7.674071294661011e-05, "loss": 0.0508, "step": 28350 }, { "grad_norm": 0.5885940790176392, "learning_rate": 7.672324651482707e-05, "loss": 0.0572, "step": 28360 }, { "grad_norm": 0.6891503930091858, "learning_rate": 7.670577551677209e-05, "loss": 0.0571, "step": 28370 }, { "grad_norm": 0.6869279742240906, "learning_rate": 7.668829995543047e-05, "loss": 0.0547, "step": 28380 }, { "grad_norm": 0.5401954650878906, "learning_rate": 7.667081983378832e-05, "loss": 0.0542, "step": 28390 }, { "grad_norm": 0.6517198085784912, "learning_rate": 7.66533351548325e-05, "loss": 0.0572, "step": 28400 }, { "grad_norm": 0.6283939480781555, "learning_rate": 7.663584592155069e-05, "loss": 0.0495, "step": 28410 }, { "grad_norm": 0.5620317459106445, "learning_rate": 7.661835213693129e-05, "loss": 0.0521, "step": 28420 }, { "grad_norm": 0.579947829246521, "learning_rate": 7.660085380396353e-05, "loss": 0.0714, "step": 28430 }, { "grad_norm": 0.5576977729797363, "learning_rate": 7.658335092563738e-05, "loss": 0.0536, "step": 28440 }, { "grad_norm": 0.4828713834285736, "learning_rate": 7.656584350494362e-05, "loss": 0.0519, "step": 28450 }, { "grad_norm": 0.6197406053543091, "learning_rate": 7.654833154487378e-05, "loss": 0.0513, "step": 28460 }, { "grad_norm": 0.7389736771583557, "learning_rate": 7.653081504842017e-05, "loss": 0.0508, "step": 28470 }, { "grad_norm": 0.6564584970474243, "learning_rate": 7.65132940185759e-05, "loss": 0.0501, "step": 28480 }, { "grad_norm": 0.6260964274406433, "learning_rate": 7.649576845833481e-05, "loss": 0.0636, "step": 28490 }, { "grad_norm": 0.39380067586898804, "learning_rate": 7.647823837069156e-05, "loss": 0.0479, "step": 28500 }, { "grad_norm": 0.4255894422531128, "learning_rate": 7.646070375864156e-05, "loss": 0.0562, "step": 28510 }, { "grad_norm": 0.5473397970199585, "learning_rate": 7.644316462518099e-05, "loss": 0.0483, "step": 28520 }, { "grad_norm": 0.6523059010505676, "learning_rate": 7.642562097330679e-05, "loss": 0.0693, "step": 28530 }, { "grad_norm": 0.739006757736206, "learning_rate": 7.640807280601671e-05, "loss": 0.0581, "step": 28540 }, { "grad_norm": 0.4267621636390686, "learning_rate": 7.639052012630927e-05, "loss": 0.0509, "step": 28550 }, { "grad_norm": 0.6087300777435303, "learning_rate": 7.63729629371837e-05, "loss": 0.0567, "step": 28560 }, { "grad_norm": 0.6088545322418213, "learning_rate": 7.635540124164009e-05, "loss": 0.0448, "step": 28570 }, { "grad_norm": 0.6497215032577515, "learning_rate": 7.633783504267922e-05, "loss": 0.0522, "step": 28580 }, { "grad_norm": 0.7605544924736023, "learning_rate": 7.632026434330269e-05, "loss": 0.0498, "step": 28590 }, { "grad_norm": 0.559249222278595, "learning_rate": 7.630268914651282e-05, "loss": 0.0611, "step": 28600 }, { "grad_norm": 1.069284439086914, "learning_rate": 7.628510945531278e-05, "loss": 0.0462, "step": 28610 }, { "grad_norm": 0.5837466716766357, "learning_rate": 7.626752527270641e-05, "loss": 0.0547, "step": 28620 }, { "grad_norm": 0.6088914275169373, "learning_rate": 7.62499366016984e-05, "loss": 0.0833, "step": 28630 }, { "grad_norm": 0.5121813416481018, "learning_rate": 7.623234344529416e-05, "loss": 0.052, "step": 28640 }, { "grad_norm": 0.768269419670105, "learning_rate": 7.62147458064999e-05, "loss": 0.0567, "step": 28650 }, { "grad_norm": 0.6191715598106384, "learning_rate": 7.619714368832254e-05, "loss": 0.0557, "step": 28660 }, { "grad_norm": 0.656878650188446, "learning_rate": 7.61795370937698e-05, "loss": 0.0537, "step": 28670 }, { "grad_norm": 0.5064024329185486, "learning_rate": 7.61619260258502e-05, "loss": 0.056, "step": 28680 }, { "grad_norm": 0.5368398427963257, "learning_rate": 7.614431048757298e-05, "loss": 0.059, "step": 28690 }, { "grad_norm": 0.8257727026939392, "learning_rate": 7.612669048194814e-05, "loss": 0.057, "step": 28700 }, { "grad_norm": 0.672538697719574, "learning_rate": 7.610906601198646e-05, "loss": 0.0502, "step": 28710 }, { "grad_norm": 0.8526982069015503, "learning_rate": 7.60914370806995e-05, "loss": 0.0433, "step": 28720 }, { "grad_norm": 0.7014760375022888, "learning_rate": 7.607380369109953e-05, "loss": 0.0519, "step": 28730 }, { "grad_norm": 0.44971805810928345, "learning_rate": 7.605616584619961e-05, "loss": 0.0475, "step": 28740 }, { "grad_norm": 0.5289646983146667, "learning_rate": 7.603852354901362e-05, "loss": 0.0502, "step": 28750 }, { "grad_norm": 0.6250002384185791, "learning_rate": 7.602087680255609e-05, "loss": 0.0533, "step": 28760 }, { "grad_norm": 0.6198046207427979, "learning_rate": 7.600322560984238e-05, "loss": 0.0556, "step": 28770 }, { "grad_norm": 0.5866119265556335, "learning_rate": 7.598556997388863e-05, "loss": 0.0708, "step": 28780 }, { "grad_norm": 0.871617317199707, "learning_rate": 7.596790989771166e-05, "loss": 0.0652, "step": 28790 }, { "grad_norm": 0.49055078625679016, "learning_rate": 7.595024538432914e-05, "loss": 0.0452, "step": 28800 }, { "grad_norm": 0.4850771129131317, "learning_rate": 7.59325764367594e-05, "loss": 0.0406, "step": 28810 }, { "grad_norm": 0.5789682865142822, "learning_rate": 7.59149030580216e-05, "loss": 0.0527, "step": 28820 }, { "grad_norm": 0.5368995070457458, "learning_rate": 7.589722525113562e-05, "loss": 0.0561, "step": 28830 }, { "grad_norm": 0.5236459374427795, "learning_rate": 7.587954301912216e-05, "loss": 0.0502, "step": 28840 }, { "grad_norm": 0.6859441995620728, "learning_rate": 7.586185636500263e-05, "loss": 0.0396, "step": 28850 }, { "grad_norm": 0.5380666851997375, "learning_rate": 7.584416529179914e-05, "loss": 0.0545, "step": 28860 }, { "grad_norm": 0.4937480390071869, "learning_rate": 7.582646980253465e-05, "loss": 0.0546, "step": 28870 }, { "grad_norm": 0.5032433867454529, "learning_rate": 7.580876990023282e-05, "loss": 0.047, "step": 28880 }, { "grad_norm": 0.6546592712402344, "learning_rate": 7.579106558791809e-05, "loss": 0.0554, "step": 28890 }, { "grad_norm": 0.6302303075790405, "learning_rate": 7.577335686861565e-05, "loss": 0.0547, "step": 28900 }, { "grad_norm": 0.52598637342453, "learning_rate": 7.575564374535141e-05, "loss": 0.0722, "step": 28910 }, { "grad_norm": 0.5370291471481323, "learning_rate": 7.573792622115207e-05, "loss": 0.0509, "step": 28920 }, { "grad_norm": 0.42519885301589966, "learning_rate": 7.572020429904507e-05, "loss": 0.0423, "step": 28930 }, { "grad_norm": 0.6695528626441956, "learning_rate": 7.570247798205861e-05, "loss": 0.0783, "step": 28940 }, { "grad_norm": 0.5607414841651917, "learning_rate": 7.568474727322164e-05, "loss": 0.043, "step": 28950 }, { "grad_norm": 0.4677278697490692, "learning_rate": 7.566701217556384e-05, "loss": 0.0436, "step": 28960 }, { "grad_norm": 0.4205225110054016, "learning_rate": 7.564927269211564e-05, "loss": 0.0437, "step": 28970 }, { "grad_norm": 0.4900592565536499, "learning_rate": 7.563152882590824e-05, "loss": 0.0469, "step": 28980 }, { "grad_norm": 0.6282466650009155, "learning_rate": 7.56137805799736e-05, "loss": 0.0716, "step": 28990 }, { "grad_norm": 0.48834481835365295, "learning_rate": 7.559602795734439e-05, "loss": 0.0466, "step": 29000 }, { "grad_norm": 0.6092322468757629, "learning_rate": 7.557827096105408e-05, "loss": 0.0579, "step": 29010 }, { "grad_norm": 0.638506293296814, "learning_rate": 7.55605095941368e-05, "loss": 0.0441, "step": 29020 }, { "grad_norm": 0.44136884808540344, "learning_rate": 7.554274385962752e-05, "loss": 0.0389, "step": 29030 }, { "grad_norm": 0.4606383144855499, "learning_rate": 7.552497376056191e-05, "loss": 0.0428, "step": 29040 }, { "grad_norm": 0.6888620853424072, "learning_rate": 7.550719929997639e-05, "loss": 0.0963, "step": 29050 }, { "grad_norm": 0.5685014128684998, "learning_rate": 7.548942048090813e-05, "loss": 0.0577, "step": 29060 }, { "grad_norm": 0.41657283902168274, "learning_rate": 7.547163730639506e-05, "loss": 0.0469, "step": 29070 }, { "grad_norm": 0.7812302112579346, "learning_rate": 7.545384977947583e-05, "loss": 0.0557, "step": 29080 }, { "grad_norm": 0.5564094185829163, "learning_rate": 7.543605790318981e-05, "loss": 0.0465, "step": 29090 }, { "grad_norm": 0.5079571008682251, "learning_rate": 7.54182616805772e-05, "loss": 0.0553, "step": 29100 }, { "grad_norm": 0.7187867760658264, "learning_rate": 7.540046111467885e-05, "loss": 0.0539, "step": 29110 }, { "grad_norm": 0.6129151582717896, "learning_rate": 7.53826562085364e-05, "loss": 0.0502, "step": 29120 }, { "grad_norm": 0.6026313304901123, "learning_rate": 7.536484696519221e-05, "loss": 0.0523, "step": 29130 }, { "grad_norm": 0.40680667757987976, "learning_rate": 7.534703338768942e-05, "loss": 0.0454, "step": 29140 }, { "grad_norm": 0.5769581198692322, "learning_rate": 7.532921547907185e-05, "loss": 0.0535, "step": 29150 }, { "grad_norm": 0.5846493244171143, "learning_rate": 7.531139324238412e-05, "loss": 0.0521, "step": 29160 }, { "grad_norm": 0.30987030267715454, "learning_rate": 7.529356668067157e-05, "loss": 0.0736, "step": 29170 }, { "grad_norm": 0.4516614079475403, "learning_rate": 7.527573579698023e-05, "loss": 0.0575, "step": 29180 }, { "grad_norm": 0.5477419495582581, "learning_rate": 7.525790059435693e-05, "loss": 0.0422, "step": 29190 }, { "grad_norm": 0.29437053203582764, "learning_rate": 7.524006107584926e-05, "loss": 0.0602, "step": 29200 }, { "grad_norm": 0.6715206503868103, "learning_rate": 7.522221724450544e-05, "loss": 0.043, "step": 29210 }, { "grad_norm": 0.5783339142799377, "learning_rate": 7.520436910337451e-05, "loss": 0.0589, "step": 29220 }, { "grad_norm": 0.5513633489608765, "learning_rate": 7.518651665550627e-05, "loss": 0.0485, "step": 29230 }, { "grad_norm": 0.7453587055206299, "learning_rate": 7.516865990395117e-05, "loss": 0.0459, "step": 29240 }, { "grad_norm": 0.49919435381889343, "learning_rate": 7.515079885176047e-05, "loss": 0.0426, "step": 29250 }, { "grad_norm": 0.7576665282249451, "learning_rate": 7.513293350198612e-05, "loss": 0.0593, "step": 29260 }, { "grad_norm": 0.4270918369293213, "learning_rate": 7.511506385768081e-05, "loss": 0.0428, "step": 29270 }, { "grad_norm": 0.615449070930481, "learning_rate": 7.509718992189801e-05, "loss": 0.0448, "step": 29280 }, { "grad_norm": 0.4977574348449707, "learning_rate": 7.507931169769182e-05, "loss": 0.0473, "step": 29290 }, { "grad_norm": 0.7234824299812317, "learning_rate": 7.506142918811722e-05, "loss": 0.0618, "step": 29300 }, { "grad_norm": 0.5520404577255249, "learning_rate": 7.504354239622978e-05, "loss": 0.0521, "step": 29310 }, { "grad_norm": 0.5742981433868408, "learning_rate": 7.50256513250859e-05, "loss": 0.0496, "step": 29320 }, { "grad_norm": 0.6193253397941589, "learning_rate": 7.500775597774265e-05, "loss": 0.0559, "step": 29330 }, { "grad_norm": 0.5186882615089417, "learning_rate": 7.498985635725788e-05, "loss": 0.048, "step": 29340 }, { "grad_norm": 0.3820967972278595, "learning_rate": 7.497195246669012e-05, "loss": 0.0472, "step": 29350 }, { "grad_norm": 0.6631262898445129, "learning_rate": 7.495404430909868e-05, "loss": 0.0527, "step": 29360 }, { "grad_norm": 0.48175591230392456, "learning_rate": 7.493613188754356e-05, "loss": 0.0418, "step": 29370 }, { "grad_norm": 0.6309952735900879, "learning_rate": 7.49182152050855e-05, "loss": 0.0479, "step": 29380 }, { "grad_norm": 0.4868457615375519, "learning_rate": 7.490029426478598e-05, "loss": 0.0471, "step": 29390 }, { "grad_norm": 0.36560487747192383, "learning_rate": 7.488236906970719e-05, "loss": 0.064, "step": 29400 }, { "grad_norm": 0.48480796813964844, "learning_rate": 7.486443962291207e-05, "loss": 0.0414, "step": 29410 }, { "grad_norm": 0.68849116563797, "learning_rate": 7.484650592746424e-05, "loss": 0.0696, "step": 29420 }, { "grad_norm": 0.6195644736289978, "learning_rate": 7.482856798642811e-05, "loss": 0.0586, "step": 29430 }, { "grad_norm": 0.6328918933868408, "learning_rate": 7.481062580286878e-05, "loss": 0.0505, "step": 29440 }, { "grad_norm": 0.514101505279541, "learning_rate": 7.479267937985208e-05, "loss": 0.0503, "step": 29450 }, { "grad_norm": 0.3248741030693054, "learning_rate": 7.477472872044456e-05, "loss": 0.0731, "step": 29460 }, { "grad_norm": 0.6835294365882874, "learning_rate": 7.475677382771347e-05, "loss": 0.0692, "step": 29470 }, { "grad_norm": 0.6518325805664062, "learning_rate": 7.473881470472683e-05, "loss": 0.0526, "step": 29480 }, { "grad_norm": 0.6682626605033875, "learning_rate": 7.472085135455337e-05, "loss": 0.039, "step": 29490 }, { "grad_norm": 0.5851399302482605, "learning_rate": 7.470288378026256e-05, "loss": 0.0581, "step": 29500 }, { "grad_norm": 0.5814422965049744, "learning_rate": 7.468491198492451e-05, "loss": 0.0476, "step": 29510 }, { "grad_norm": 0.6076379418373108, "learning_rate": 7.466693597161013e-05, "loss": 0.0548, "step": 29520 }, { "grad_norm": 0.4369398355484009, "learning_rate": 7.464895574339104e-05, "loss": 0.0405, "step": 29530 }, { "grad_norm": 0.42513468861579895, "learning_rate": 7.463097130333958e-05, "loss": 0.046, "step": 29540 }, { "grad_norm": 0.49078479409217834, "learning_rate": 7.461298265452876e-05, "loss": 0.0465, "step": 29550 }, { "grad_norm": 0.5786937475204468, "learning_rate": 7.459498980003239e-05, "loss": 0.0464, "step": 29560 }, { "grad_norm": 0.43314650654792786, "learning_rate": 7.457699274292493e-05, "loss": 0.0395, "step": 29570 }, { "grad_norm": 0.4894212782382965, "learning_rate": 7.455899148628159e-05, "loss": 0.0495, "step": 29580 }, { "grad_norm": 0.668125331401825, "learning_rate": 7.45409860331783e-05, "loss": 0.0505, "step": 29590 }, { "grad_norm": 0.770045280456543, "learning_rate": 7.452297638669169e-05, "loss": 0.0677, "step": 29600 }, { "grad_norm": 1.0999289751052856, "learning_rate": 7.450496254989911e-05, "loss": 0.0602, "step": 29610 }, { "grad_norm": 0.5166067481040955, "learning_rate": 7.448694452587866e-05, "loss": 0.0463, "step": 29620 }, { "grad_norm": 0.5630406141281128, "learning_rate": 7.44689223177091e-05, "loss": 0.0565, "step": 29630 }, { "grad_norm": 0.5306178331375122, "learning_rate": 7.445089592846994e-05, "loss": 0.041, "step": 29640 }, { "grad_norm": 0.6613732576370239, "learning_rate": 7.443286536124141e-05, "loss": 0.076, "step": 29650 }, { "grad_norm": 0.47750869393348694, "learning_rate": 7.441483061910443e-05, "loss": 0.0379, "step": 29660 }, { "grad_norm": 0.48479458689689636, "learning_rate": 7.439679170514064e-05, "loss": 0.0835, "step": 29670 }, { "grad_norm": 0.40018102526664734, "learning_rate": 7.43787486224324e-05, "loss": 0.0438, "step": 29680 }, { "grad_norm": 0.47166717052459717, "learning_rate": 7.436070137406276e-05, "loss": 0.0421, "step": 29690 }, { "grad_norm": 0.5306294560432434, "learning_rate": 7.434264996311556e-05, "loss": 0.0954, "step": 29700 }, { "grad_norm": 0.4447936415672302, "learning_rate": 7.432459439267525e-05, "loss": 0.0428, "step": 29710 }, { "grad_norm": 0.4869143068790436, "learning_rate": 7.430653466582701e-05, "loss": 0.0473, "step": 29720 }, { "grad_norm": 1.1219894886016846, "learning_rate": 7.42884707856568e-05, "loss": 0.0522, "step": 29730 }, { "grad_norm": 0.7495091557502747, "learning_rate": 7.427040275525122e-05, "loss": 0.0455, "step": 29740 }, { "grad_norm": 0.2937484681606293, "learning_rate": 7.42523305776976e-05, "loss": 0.0412, "step": 29750 }, { "grad_norm": 0.6602367162704468, "learning_rate": 7.4234254256084e-05, "loss": 0.0628, "step": 29760 }, { "grad_norm": 0.5187426805496216, "learning_rate": 7.421617379349915e-05, "loss": 0.0427, "step": 29770 }, { "grad_norm": 0.4692125618457794, "learning_rate": 7.41980891930325e-05, "loss": 0.0451, "step": 29780 }, { "grad_norm": 0.6319959163665771, "learning_rate": 7.418000045777425e-05, "loss": 0.0532, "step": 29790 }, { "grad_norm": 0.627991795539856, "learning_rate": 7.416190759081523e-05, "loss": 0.0434, "step": 29800 }, { "grad_norm": 0.46493133902549744, "learning_rate": 7.414381059524704e-05, "loss": 0.0498, "step": 29810 }, { "grad_norm": 0.9039298892021179, "learning_rate": 7.412570947416195e-05, "loss": 0.0508, "step": 29820 }, { "grad_norm": 0.5829097628593445, "learning_rate": 7.410760423065295e-05, "loss": 0.0462, "step": 29830 }, { "grad_norm": 0.6488021612167358, "learning_rate": 7.408949486781372e-05, "loss": 0.048, "step": 29840 }, { "grad_norm": 0.6022322773933411, "learning_rate": 7.407138138873868e-05, "loss": 0.0454, "step": 29850 }, { "grad_norm": 0.549929141998291, "learning_rate": 7.405326379652292e-05, "loss": 0.0366, "step": 29860 }, { "grad_norm": 0.5992811322212219, "learning_rate": 7.403514209426222e-05, "loss": 0.0635, "step": 29870 }, { "grad_norm": 0.5110329985618591, "learning_rate": 7.40170162850531e-05, "loss": 0.0494, "step": 29880 }, { "grad_norm": 0.763399600982666, "learning_rate": 7.399888637199278e-05, "loss": 0.0389, "step": 29890 }, { "grad_norm": 0.5641098022460938, "learning_rate": 7.398075235817914e-05, "loss": 0.0796, "step": 29900 }, { "grad_norm": 0.43278732895851135, "learning_rate": 7.39626142467108e-05, "loss": 0.0405, "step": 29910 }, { "grad_norm": 0.4768240451812744, "learning_rate": 7.394447204068706e-05, "loss": 0.0592, "step": 29920 }, { "grad_norm": 0.6034666895866394, "learning_rate": 7.392632574320793e-05, "loss": 0.0411, "step": 29930 }, { "grad_norm": 0.4266379475593567, "learning_rate": 7.390817535737411e-05, "loss": 0.0539, "step": 29940 }, { "grad_norm": 0.6156547665596008, "learning_rate": 7.389002088628703e-05, "loss": 0.0493, "step": 29950 }, { "grad_norm": 0.5695153474807739, "learning_rate": 7.387186233304877e-05, "loss": 0.0455, "step": 29960 }, { "grad_norm": 0.7372891306877136, "learning_rate": 7.385369970076212e-05, "loss": 0.0382, "step": 29970 }, { "grad_norm": 0.4604862928390503, "learning_rate": 7.38355329925306e-05, "loss": 0.05, "step": 29980 }, { "grad_norm": 0.5944568514823914, "learning_rate": 7.381736221145838e-05, "loss": 0.0472, "step": 29990 }, { "grad_norm": 0.6910271048545837, "learning_rate": 7.37991873606504e-05, "loss": 0.0551, "step": 30000 }, { "grad_norm": 0.6453610062599182, "learning_rate": 7.378100844321218e-05, "loss": 0.0438, "step": 30010 }, { "grad_norm": 0.7859434485435486, "learning_rate": 7.376282546225004e-05, "loss": 0.0531, "step": 30020 }, { "grad_norm": 0.5469463467597961, "learning_rate": 7.374463842087094e-05, "loss": 0.0475, "step": 30030 }, { "grad_norm": 0.36168381571769714, "learning_rate": 7.372644732218254e-05, "loss": 0.0579, "step": 30040 }, { "grad_norm": 0.5312907099723816, "learning_rate": 7.370825216929322e-05, "loss": 0.0471, "step": 30050 }, { "grad_norm": 0.9723027944564819, "learning_rate": 7.369005296531205e-05, "loss": 0.0508, "step": 30060 }, { "grad_norm": 0.6765477061271667, "learning_rate": 7.367184971334873e-05, "loss": 0.0769, "step": 30070 }, { "grad_norm": 0.3533386290073395, "learning_rate": 7.365364241651371e-05, "loss": 0.0365, "step": 30080 }, { "grad_norm": 0.47876131534576416, "learning_rate": 7.363543107791815e-05, "loss": 0.0496, "step": 30090 }, { "grad_norm": 0.35026198625564575, "learning_rate": 7.361721570067384e-05, "loss": 0.064, "step": 30100 }, { "grad_norm": 0.6286191344261169, "learning_rate": 7.359899628789331e-05, "loss": 0.0607, "step": 30110 }, { "grad_norm": 0.6472609639167786, "learning_rate": 7.358077284268974e-05, "loss": 0.0487, "step": 30120 }, { "grad_norm": 0.4688190519809723, "learning_rate": 7.356254536817702e-05, "loss": 0.0418, "step": 30130 }, { "grad_norm": 0.4129399359226227, "learning_rate": 7.354431386746973e-05, "loss": 0.0572, "step": 30140 }, { "grad_norm": 0.521628201007843, "learning_rate": 7.352607834368316e-05, "loss": 0.043, "step": 30150 }, { "grad_norm": 0.49493706226348877, "learning_rate": 7.350783879993324e-05, "loss": 0.0512, "step": 30160 }, { "grad_norm": 0.4154100716114044, "learning_rate": 7.348959523933658e-05, "loss": 0.0613, "step": 30170 }, { "grad_norm": 0.4990423619747162, "learning_rate": 7.347134766501057e-05, "loss": 0.0498, "step": 30180 }, { "grad_norm": 0.5884373784065247, "learning_rate": 7.345309608007315e-05, "loss": 0.0509, "step": 30190 }, { "grad_norm": 0.48470011353492737, "learning_rate": 7.343484048764308e-05, "loss": 0.0423, "step": 30200 }, { "grad_norm": 0.4686812162399292, "learning_rate": 7.341658089083972e-05, "loss": 0.0404, "step": 30210 }, { "grad_norm": 0.5146093964576721, "learning_rate": 7.339831729278313e-05, "loss": 0.0582, "step": 30220 }, { "grad_norm": 0.5694398283958435, "learning_rate": 7.338004969659404e-05, "loss": 0.0434, "step": 30230 }, { "grad_norm": 0.5065194964408875, "learning_rate": 7.336177810539391e-05, "loss": 0.0402, "step": 30240 }, { "grad_norm": 0.41985851526260376, "learning_rate": 7.334350252230485e-05, "loss": 0.0462, "step": 30250 }, { "grad_norm": 0.6483628153800964, "learning_rate": 7.332522295044965e-05, "loss": 0.0348, "step": 30260 }, { "grad_norm": 0.4972793459892273, "learning_rate": 7.33069393929518e-05, "loss": 0.0411, "step": 30270 }, { "grad_norm": 0.5178481340408325, "learning_rate": 7.328865185293545e-05, "loss": 0.0304, "step": 30280 }, { "grad_norm": 0.622421383857727, "learning_rate": 7.327036033352546e-05, "loss": 0.0579, "step": 30290 }, { "grad_norm": 0.5213474035263062, "learning_rate": 7.325206483784733e-05, "loss": 0.0522, "step": 30300 }, { "grad_norm": 0.670924186706543, "learning_rate": 7.323376536902724e-05, "loss": 0.052, "step": 30310 }, { "grad_norm": 0.4607028663158417, "learning_rate": 7.321546193019213e-05, "loss": 0.0556, "step": 30320 }, { "grad_norm": 0.49157196283340454, "learning_rate": 7.31971545244695e-05, "loss": 0.0366, "step": 30330 }, { "grad_norm": 0.6191896200180054, "learning_rate": 7.31788431549876e-05, "loss": 0.07, "step": 30340 }, { "grad_norm": 0.6117685437202454, "learning_rate": 7.316052782487534e-05, "loss": 0.0452, "step": 30350 }, { "grad_norm": 0.49307122826576233, "learning_rate": 7.314220853726234e-05, "loss": 0.0492, "step": 30360 }, { "grad_norm": 0.7322137951850891, "learning_rate": 7.312388529527884e-05, "loss": 0.056, "step": 30370 }, { "grad_norm": 0.5207671523094177, "learning_rate": 7.310555810205577e-05, "loss": 0.0416, "step": 30380 }, { "grad_norm": 0.5089696645736694, "learning_rate": 7.308722696072476e-05, "loss": 0.0591, "step": 30390 }, { "grad_norm": 0.4823630750179291, "learning_rate": 7.306889187441811e-05, "loss": 0.052, "step": 30400 }, { "grad_norm": 0.7319954633712769, "learning_rate": 7.305055284626876e-05, "loss": 0.0527, "step": 30410 }, { "grad_norm": 0.6642930507659912, "learning_rate": 7.303220987941037e-05, "loss": 0.0398, "step": 30420 }, { "grad_norm": 0.6302623152732849, "learning_rate": 7.301386297697726e-05, "loss": 0.0424, "step": 30430 }, { "grad_norm": 1.0326836109161377, "learning_rate": 7.299551214210438e-05, "loss": 0.0572, "step": 30440 }, { "grad_norm": 0.6334959864616394, "learning_rate": 7.297715737792738e-05, "loss": 0.0394, "step": 30450 }, { "grad_norm": 0.9572920203208923, "learning_rate": 7.295879868758265e-05, "loss": 0.0418, "step": 30460 }, { "grad_norm": 0.4760129749774933, "learning_rate": 7.294043607420713e-05, "loss": 0.048, "step": 30470 }, { "grad_norm": 0.6474470496177673, "learning_rate": 7.292206954093852e-05, "loss": 0.0408, "step": 30480 }, { "grad_norm": 0.6184828281402588, "learning_rate": 7.290369909091515e-05, "loss": 0.0521, "step": 30490 }, { "grad_norm": 0.523918867111206, "learning_rate": 7.2885324727276e-05, "loss": 0.0836, "step": 30500 }, { "grad_norm": 0.751838743686676, "learning_rate": 7.286694645316076e-05, "loss": 0.0519, "step": 30510 }, { "grad_norm": 0.3915726840496063, "learning_rate": 7.284856427170982e-05, "loss": 0.0476, "step": 30520 }, { "grad_norm": 0.5601533055305481, "learning_rate": 7.283017818606414e-05, "loss": 0.0466, "step": 30530 }, { "grad_norm": 0.4097653031349182, "learning_rate": 7.28117881993654e-05, "loss": 0.0413, "step": 30540 }, { "grad_norm": 1.3710845708847046, "learning_rate": 7.279339431475598e-05, "loss": 0.0508, "step": 30550 }, { "grad_norm": 0.3615245223045349, "learning_rate": 7.277499653537887e-05, "loss": 0.0343, "step": 30560 }, { "grad_norm": 0.5561213493347168, "learning_rate": 7.275659486437776e-05, "loss": 0.0578, "step": 30570 }, { "grad_norm": 0.7341307401657104, "learning_rate": 7.273818930489695e-05, "loss": 0.0476, "step": 30580 }, { "grad_norm": 0.7139369249343872, "learning_rate": 7.271977986008151e-05, "loss": 0.0516, "step": 30590 }, { "grad_norm": 0.5780345797538757, "learning_rate": 7.270136653307705e-05, "loss": 0.056, "step": 30600 }, { "grad_norm": 0.4522034227848053, "learning_rate": 7.268294932702994e-05, "loss": 0.0655, "step": 30610 }, { "grad_norm": 0.5028349757194519, "learning_rate": 7.266452824508719e-05, "loss": 0.0416, "step": 30620 }, { "grad_norm": 0.2565953731536865, "learning_rate": 7.264610329039643e-05, "loss": 0.0496, "step": 30630 }, { "grad_norm": 0.4125237464904785, "learning_rate": 7.262767446610599e-05, "loss": 0.0344, "step": 30640 }, { "grad_norm": 0.40523162484169006, "learning_rate": 7.260924177536485e-05, "loss": 0.0363, "step": 30650 }, { "grad_norm": 0.6576168537139893, "learning_rate": 7.259080522132265e-05, "loss": 0.0551, "step": 30660 }, { "grad_norm": 0.5914615392684937, "learning_rate": 7.257236480712972e-05, "loss": 0.041, "step": 30670 }, { "grad_norm": 0.3128407895565033, "learning_rate": 7.255392053593697e-05, "loss": 0.0648, "step": 30680 }, { "grad_norm": 0.607868492603302, "learning_rate": 7.253547241089607e-05, "loss": 0.042, "step": 30690 }, { "grad_norm": 0.41011109948158264, "learning_rate": 7.251702043515927e-05, "loss": 0.0604, "step": 30700 }, { "grad_norm": 0.3491290509700775, "learning_rate": 7.249856461187952e-05, "loss": 0.0419, "step": 30710 }, { "grad_norm": 0.42259255051612854, "learning_rate": 7.248010494421042e-05, "loss": 0.0451, "step": 30720 }, { "grad_norm": 0.49831539392471313, "learning_rate": 7.246164143530622e-05, "loss": 0.0585, "step": 30730 }, { "grad_norm": 0.6026462316513062, "learning_rate": 7.244317408832181e-05, "loss": 0.0468, "step": 30740 }, { "grad_norm": 0.6549220681190491, "learning_rate": 7.242470290641279e-05, "loss": 0.054, "step": 30750 }, { "grad_norm": 0.7466177344322205, "learning_rate": 7.240622789273536e-05, "loss": 0.0445, "step": 30760 }, { "grad_norm": 0.5937085747718811, "learning_rate": 7.238774905044638e-05, "loss": 0.0362, "step": 30770 }, { "grad_norm": 0.6639177203178406, "learning_rate": 7.236926638270341e-05, "loss": 0.0504, "step": 30780 }, { "grad_norm": 0.4017050862312317, "learning_rate": 7.23507798926646e-05, "loss": 0.0542, "step": 30790 }, { "grad_norm": 0.512866199016571, "learning_rate": 7.23322895834888e-05, "loss": 0.0414, "step": 30800 }, { "grad_norm": 0.49303725361824036, "learning_rate": 7.231379545833552e-05, "loss": 0.0521, "step": 30810 }, { "grad_norm": 0.5636654496192932, "learning_rate": 7.229529752036487e-05, "loss": 0.0467, "step": 30820 }, { "grad_norm": 0.5962358713150024, "learning_rate": 7.227679577273765e-05, "loss": 0.0581, "step": 30830 }, { "grad_norm": 0.5329880118370056, "learning_rate": 7.225829021861529e-05, "loss": 0.0412, "step": 30840 }, { "grad_norm": 0.5441898107528687, "learning_rate": 7.223978086115992e-05, "loss": 0.0563, "step": 30850 }, { "grad_norm": 0.533840000629425, "learning_rate": 7.222126770353425e-05, "loss": 0.0727, "step": 30860 }, { "grad_norm": 0.4577081799507141, "learning_rate": 7.22027507489017e-05, "loss": 0.0456, "step": 30870 }, { "grad_norm": 0.9207549095153809, "learning_rate": 7.218423000042627e-05, "loss": 0.0506, "step": 30880 }, { "grad_norm": 0.5321143269538879, "learning_rate": 7.216570546127268e-05, "loss": 0.0705, "step": 30890 }, { "grad_norm": 0.6290050745010376, "learning_rate": 7.214717713460626e-05, "loss": 0.0512, "step": 30900 }, { "grad_norm": 0.49346840381622314, "learning_rate": 7.2128645023593e-05, "loss": 0.0453, "step": 30910 }, { "grad_norm": 0.428723007440567, "learning_rate": 7.211010913139951e-05, "loss": 0.0597, "step": 30920 }, { "grad_norm": 0.4573637843132019, "learning_rate": 7.209156946119308e-05, "loss": 0.0394, "step": 30930 }, { "grad_norm": 0.32704266905784607, "learning_rate": 7.207302601614166e-05, "loss": 0.0412, "step": 30940 }, { "grad_norm": 0.5035837888717651, "learning_rate": 7.205447879941378e-05, "loss": 0.0448, "step": 30950 }, { "grad_norm": 0.6770862340927124, "learning_rate": 7.203592781417866e-05, "loss": 0.0699, "step": 30960 }, { "grad_norm": 0.5303424596786499, "learning_rate": 7.201737306360617e-05, "loss": 0.064, "step": 30970 }, { "grad_norm": 0.7414591908454895, "learning_rate": 7.19988145508668e-05, "loss": 0.0514, "step": 30980 }, { "grad_norm": 0.5781143307685852, "learning_rate": 7.198025227913168e-05, "loss": 0.0522, "step": 30990 }, { "grad_norm": 0.43529942631721497, "learning_rate": 7.196168625157261e-05, "loss": 0.0535, "step": 31000 }, { "grad_norm": 0.5086681842803955, "learning_rate": 7.194311647136201e-05, "loss": 0.0452, "step": 31010 }, { "grad_norm": 0.38984519243240356, "learning_rate": 7.192454294167297e-05, "loss": 0.0419, "step": 31020 }, { "grad_norm": 0.42761337757110596, "learning_rate": 7.190596566567917e-05, "loss": 0.0442, "step": 31030 }, { "grad_norm": 0.44461482763290405, "learning_rate": 7.188738464655496e-05, "loss": 0.0432, "step": 31040 }, { "grad_norm": 0.5233903527259827, "learning_rate": 7.186879988747533e-05, "loss": 0.0378, "step": 31050 }, { "grad_norm": 0.7642790675163269, "learning_rate": 7.185021139161592e-05, "loss": 0.042, "step": 31060 }, { "grad_norm": 0.45114797353744507, "learning_rate": 7.1831619162153e-05, "loss": 0.0566, "step": 31070 }, { "grad_norm": 0.4261466860771179, "learning_rate": 7.181302320226345e-05, "loss": 0.043, "step": 31080 }, { "grad_norm": 0.7670255303382874, "learning_rate": 7.179442351512482e-05, "loss": 0.0424, "step": 31090 }, { "grad_norm": 0.49760720133781433, "learning_rate": 7.177582010391528e-05, "loss": 0.0525, "step": 31100 }, { "grad_norm": 0.6087490320205688, "learning_rate": 7.175721297181366e-05, "loss": 0.0522, "step": 31110 }, { "grad_norm": 0.4523579478263855, "learning_rate": 7.173860212199942e-05, "loss": 0.0365, "step": 31120 }, { "grad_norm": 0.49971723556518555, "learning_rate": 7.171998755765263e-05, "loss": 0.0558, "step": 31130 }, { "grad_norm": 0.6303145885467529, "learning_rate": 7.170136928195398e-05, "loss": 0.0544, "step": 31140 }, { "grad_norm": 0.7170969247817993, "learning_rate": 7.168274729808489e-05, "loss": 0.0652, "step": 31150 }, { "grad_norm": 0.4102283716201782, "learning_rate": 7.166412160922728e-05, "loss": 0.0658, "step": 31160 }, { "grad_norm": 0.5796474814414978, "learning_rate": 7.164549221856382e-05, "loss": 0.0469, "step": 31170 }, { "grad_norm": 0.5401419997215271, "learning_rate": 7.162685912927775e-05, "loss": 0.0467, "step": 31180 }, { "grad_norm": 1.2519314289093018, "learning_rate": 7.160822234455294e-05, "loss": 0.061, "step": 31190 }, { "grad_norm": 1.1510075330734253, "learning_rate": 7.158958186757391e-05, "loss": 0.0431, "step": 31200 }, { "grad_norm": 0.8488655686378479, "learning_rate": 7.157093770152582e-05, "loss": 0.0507, "step": 31210 }, { "grad_norm": 0.5408819317817688, "learning_rate": 7.155228984959446e-05, "loss": 0.0605, "step": 31220 }, { "grad_norm": 0.47108492255210876, "learning_rate": 7.153363831496621e-05, "loss": 0.0391, "step": 31230 }, { "grad_norm": 0.3993987739086151, "learning_rate": 7.151498310082811e-05, "loss": 0.0511, "step": 31240 }, { "grad_norm": 0.4164508879184723, "learning_rate": 7.149632421036784e-05, "loss": 0.0498, "step": 31250 }, { "grad_norm": 0.7529287934303284, "learning_rate": 7.147766164677369e-05, "loss": 0.0521, "step": 31260 }, { "grad_norm": 0.6139137744903564, "learning_rate": 7.145899541323459e-05, "loss": 0.0484, "step": 31270 }, { "grad_norm": 0.534092128276825, "learning_rate": 7.144032551294007e-05, "loss": 0.0434, "step": 31280 }, { "grad_norm": 0.5775997042655945, "learning_rate": 7.14216519490803e-05, "loss": 0.0636, "step": 31290 }, { "grad_norm": 0.7801694869995117, "learning_rate": 7.140297472484609e-05, "loss": 0.0569, "step": 31300 }, { "grad_norm": 0.5731073021888733, "learning_rate": 7.138429384342891e-05, "loss": 0.0513, "step": 31310 }, { "grad_norm": 0.7975532412528992, "learning_rate": 7.136560930802074e-05, "loss": 0.0603, "step": 31320 }, { "grad_norm": 0.4969448149204254, "learning_rate": 7.134692112181431e-05, "loss": 0.047, "step": 31330 }, { "grad_norm": 0.8305109739303589, "learning_rate": 7.13282292880029e-05, "loss": 0.0603, "step": 31340 }, { "grad_norm": 0.5191242098808289, "learning_rate": 7.130953380978043e-05, "loss": 0.0508, "step": 31350 }, { "grad_norm": 0.3428443968296051, "learning_rate": 7.129083469034144e-05, "loss": 0.0434, "step": 31360 }, { "grad_norm": 0.39006027579307556, "learning_rate": 7.127213193288112e-05, "loss": 0.0471, "step": 31370 }, { "grad_norm": 0.479794979095459, "learning_rate": 7.125342554059522e-05, "loss": 0.0435, "step": 31380 }, { "grad_norm": 0.5881751179695129, "learning_rate": 7.12347155166802e-05, "loss": 0.0445, "step": 31390 }, { "grad_norm": 0.6929100155830383, "learning_rate": 7.121600186433306e-05, "loss": 0.0437, "step": 31400 }, { "grad_norm": 0.817488968372345, "learning_rate": 7.119728458675148e-05, "loss": 0.0444, "step": 31410 }, { "grad_norm": 0.5169036388397217, "learning_rate": 7.117856368713369e-05, "loss": 0.0505, "step": 31420 }, { "grad_norm": 0.6735900640487671, "learning_rate": 7.115983916867861e-05, "loss": 0.0506, "step": 31430 }, { "grad_norm": 0.5931878685951233, "learning_rate": 7.114111103458574e-05, "loss": 0.0501, "step": 31440 }, { "grad_norm": 0.6359471678733826, "learning_rate": 7.11223792880552e-05, "loss": 0.0613, "step": 31450 }, { "grad_norm": 0.63620525598526, "learning_rate": 7.110364393228773e-05, "loss": 0.0618, "step": 31460 }, { "grad_norm": 0.5444961190223694, "learning_rate": 7.108490497048471e-05, "loss": 0.0528, "step": 31470 }, { "grad_norm": 0.5698848962783813, "learning_rate": 7.10661624058481e-05, "loss": 0.0412, "step": 31480 }, { "grad_norm": 0.5369821190834045, "learning_rate": 7.10474162415805e-05, "loss": 0.0593, "step": 31490 }, { "grad_norm": 0.6283092498779297, "learning_rate": 7.102866648088511e-05, "loss": 0.0491, "step": 31500 }, { "grad_norm": 0.6350674033164978, "learning_rate": 7.100991312696576e-05, "loss": 0.0414, "step": 31510 }, { "grad_norm": 0.661550760269165, "learning_rate": 7.099115618302686e-05, "loss": 0.0559, "step": 31520 }, { "grad_norm": 0.5650786757469177, "learning_rate": 7.097239565227349e-05, "loss": 0.0337, "step": 31530 }, { "grad_norm": 0.44127562642097473, "learning_rate": 7.09536315379113e-05, "loss": 0.0577, "step": 31540 }, { "grad_norm": 0.5243874192237854, "learning_rate": 7.093486384314656e-05, "loss": 0.0577, "step": 31550 }, { "grad_norm": 0.3709259331226349, "learning_rate": 7.091609257118616e-05, "loss": 0.0408, "step": 31560 }, { "grad_norm": 0.4720003306865692, "learning_rate": 7.08973177252376e-05, "loss": 0.0659, "step": 31570 }, { "grad_norm": 0.5060707926750183, "learning_rate": 7.087853930850898e-05, "loss": 0.0496, "step": 31580 }, { "grad_norm": 0.647515594959259, "learning_rate": 7.085975732420903e-05, "loss": 0.052, "step": 31590 }, { "grad_norm": 0.5824308395385742, "learning_rate": 7.084097177554706e-05, "loss": 0.0444, "step": 31600 }, { "grad_norm": 0.9154977798461914, "learning_rate": 7.082218266573301e-05, "loss": 0.0565, "step": 31610 }, { "grad_norm": 0.4368111193180084, "learning_rate": 7.080338999797743e-05, "loss": 0.0516, "step": 31620 }, { "grad_norm": 0.5685699582099915, "learning_rate": 7.07845937754915e-05, "loss": 0.0535, "step": 31630 }, { "grad_norm": 0.43540796637535095, "learning_rate": 7.076579400148693e-05, "loss": 0.0426, "step": 31640 }, { "grad_norm": 0.5934962034225464, "learning_rate": 7.074699067917611e-05, "loss": 0.05, "step": 31650 }, { "grad_norm": 0.655725359916687, "learning_rate": 7.072818381177201e-05, "loss": 0.0398, "step": 31660 }, { "grad_norm": 0.8258194327354431, "learning_rate": 7.070937340248823e-05, "loss": 0.0374, "step": 31670 }, { "grad_norm": 0.6001187562942505, "learning_rate": 7.069055945453893e-05, "loss": 0.0488, "step": 31680 }, { "grad_norm": 0.44187119603157043, "learning_rate": 7.067174197113892e-05, "loss": 0.0413, "step": 31690 }, { "grad_norm": 0.6473723649978638, "learning_rate": 7.065292095550355e-05, "loss": 0.0465, "step": 31700 }, { "grad_norm": 0.48658865690231323, "learning_rate": 7.063409641084887e-05, "loss": 0.0468, "step": 31710 }, { "grad_norm": 0.4742107391357422, "learning_rate": 7.061526834039145e-05, "loss": 0.0502, "step": 31720 }, { "grad_norm": 0.576642632484436, "learning_rate": 7.05964367473485e-05, "loss": 0.0556, "step": 31730 }, { "grad_norm": 0.7181727290153503, "learning_rate": 7.057760163493783e-05, "loss": 0.0536, "step": 31740 }, { "grad_norm": 0.6833740472793579, "learning_rate": 7.055876300637783e-05, "loss": 0.0746, "step": 31750 }, { "grad_norm": 0.7703845500946045, "learning_rate": 7.053992086488753e-05, "loss": 0.0409, "step": 31760 }, { "grad_norm": 0.5997470021247864, "learning_rate": 7.052107521368651e-05, "loss": 0.0542, "step": 31770 }, { "grad_norm": 0.42781001329421997, "learning_rate": 7.0502226055995e-05, "loss": 0.0419, "step": 31780 }, { "grad_norm": 0.4499622881412506, "learning_rate": 7.048337339503379e-05, "loss": 0.0409, "step": 31790 }, { "grad_norm": 0.4884495139122009, "learning_rate": 7.046451723402427e-05, "loss": 0.047, "step": 31800 }, { "grad_norm": 0.5164876580238342, "learning_rate": 7.044565757618848e-05, "loss": 0.0646, "step": 31810 }, { "grad_norm": 0.6167012453079224, "learning_rate": 7.042679442474899e-05, "loss": 0.0458, "step": 31820 }, { "grad_norm": 0.5796486139297485, "learning_rate": 7.040792778292902e-05, "loss": 0.0414, "step": 31830 }, { "grad_norm": 0.5955703854560852, "learning_rate": 7.038905765395234e-05, "loss": 0.0615, "step": 31840 }, { "grad_norm": 0.3900468647480011, "learning_rate": 7.037018404104334e-05, "loss": 0.0534, "step": 31850 }, { "grad_norm": 0.589620053768158, "learning_rate": 7.035130694742702e-05, "loss": 0.0489, "step": 31860 }, { "grad_norm": 0.5823613405227661, "learning_rate": 7.033242637632897e-05, "loss": 0.0419, "step": 31870 }, { "grad_norm": 0.5973843336105347, "learning_rate": 7.031354233097534e-05, "loss": 0.0442, "step": 31880 }, { "grad_norm": 0.5304083824157715, "learning_rate": 7.029465481459289e-05, "loss": 0.0439, "step": 31890 }, { "grad_norm": 0.41852739453315735, "learning_rate": 7.027576383040898e-05, "loss": 0.045, "step": 31900 }, { "grad_norm": 0.3681059777736664, "learning_rate": 7.025686938165159e-05, "loss": 0.0553, "step": 31910 }, { "grad_norm": 0.40802502632141113, "learning_rate": 7.023797147154924e-05, "loss": 0.0382, "step": 31920 }, { "grad_norm": 0.9870336651802063, "learning_rate": 7.021907010333111e-05, "loss": 0.0591, "step": 31930 }, { "grad_norm": 0.5759120583534241, "learning_rate": 7.020016528022685e-05, "loss": 0.0616, "step": 31940 }, { "grad_norm": 0.5667083263397217, "learning_rate": 7.018125700546683e-05, "loss": 0.0477, "step": 31950 }, { "grad_norm": 0.7329322099685669, "learning_rate": 7.016234528228196e-05, "loss": 0.0444, "step": 31960 }, { "grad_norm": 0.4317202568054199, "learning_rate": 7.014343011390372e-05, "loss": 0.061, "step": 31970 }, { "grad_norm": 0.5859937071800232, "learning_rate": 7.01245115035642e-05, "loss": 0.0574, "step": 31980 }, { "grad_norm": 0.48863083124160767, "learning_rate": 7.010558945449606e-05, "loss": 0.0387, "step": 31990 }, { "grad_norm": 0.6202539205551147, "learning_rate": 7.008666396993258e-05, "loss": 0.0524, "step": 32000 }, { "grad_norm": 0.4552992284297943, "learning_rate": 7.006773505310759e-05, "loss": 0.0441, "step": 32010 }, { "grad_norm": 0.4733272194862366, "learning_rate": 7.004880270725553e-05, "loss": 0.036, "step": 32020 }, { "grad_norm": 0.6784123182296753, "learning_rate": 7.002986693561144e-05, "loss": 0.051, "step": 32030 }, { "grad_norm": 0.5833758115768433, "learning_rate": 7.001092774141089e-05, "loss": 0.0331, "step": 32040 }, { "grad_norm": 0.872124433517456, "learning_rate": 6.999198512789009e-05, "loss": 0.0603, "step": 32050 }, { "grad_norm": 0.5675405263900757, "learning_rate": 6.997303909828584e-05, "loss": 0.0388, "step": 32060 }, { "grad_norm": 0.647888720035553, "learning_rate": 6.995408965583544e-05, "loss": 0.0459, "step": 32070 }, { "grad_norm": 0.4640538990497589, "learning_rate": 6.993513680377688e-05, "loss": 0.0488, "step": 32080 }, { "grad_norm": 0.5346035957336426, "learning_rate": 6.991618054534868e-05, "loss": 0.0436, "step": 32090 }, { "grad_norm": 0.5936759114265442, "learning_rate": 6.989722088378991e-05, "loss": 0.0448, "step": 32100 }, { "grad_norm": 0.5574720501899719, "learning_rate": 6.987825782234027e-05, "loss": 0.0452, "step": 32110 }, { "grad_norm": 0.6253948211669922, "learning_rate": 6.985929136424006e-05, "loss": 0.0463, "step": 32120 }, { "grad_norm": 0.5357709527015686, "learning_rate": 6.984032151273012e-05, "loss": 0.0469, "step": 32130 }, { "grad_norm": 0.568518877029419, "learning_rate": 6.982134827105186e-05, "loss": 0.0395, "step": 32140 }, { "grad_norm": 0.6198900938034058, "learning_rate": 6.980237164244729e-05, "loss": 0.0443, "step": 32150 }, { "grad_norm": 0.4856678247451782, "learning_rate": 6.9783391630159e-05, "loss": 0.0428, "step": 32160 }, { "grad_norm": 0.465214341878891, "learning_rate": 6.976440823743015e-05, "loss": 0.0427, "step": 32170 }, { "grad_norm": 0.6379620432853699, "learning_rate": 6.974542146750451e-05, "loss": 0.0497, "step": 32180 }, { "grad_norm": 0.7236629128456116, "learning_rate": 6.972643132362637e-05, "loss": 0.0512, "step": 32190 }, { "grad_norm": 0.5078441500663757, "learning_rate": 6.970743780904064e-05, "loss": 0.0496, "step": 32200 }, { "grad_norm": 0.5628373026847839, "learning_rate": 6.968844092699277e-05, "loss": 0.0635, "step": 32210 }, { "grad_norm": 0.48998013138771057, "learning_rate": 6.966944068072883e-05, "loss": 0.0489, "step": 32220 }, { "grad_norm": 0.5155541300773621, "learning_rate": 6.965043707349545e-05, "loss": 0.0562, "step": 32230 }, { "grad_norm": 0.6041528582572937, "learning_rate": 6.963143010853982e-05, "loss": 0.0581, "step": 32240 }, { "grad_norm": 0.5102199912071228, "learning_rate": 6.961241978910971e-05, "loss": 0.0505, "step": 32250 }, { "grad_norm": 0.43858712911605835, "learning_rate": 6.959340611845344e-05, "loss": 0.0568, "step": 32260 }, { "grad_norm": 0.3786903917789459, "learning_rate": 6.957438909981995e-05, "loss": 0.0367, "step": 32270 }, { "grad_norm": 0.7722460627555847, "learning_rate": 6.955536873645872e-05, "loss": 0.0492, "step": 32280 }, { "grad_norm": 0.9877246618270874, "learning_rate": 6.953634503161982e-05, "loss": 0.0579, "step": 32290 }, { "grad_norm": 0.4876300096511841, "learning_rate": 6.951731798855387e-05, "loss": 0.0416, "step": 32300 }, { "grad_norm": 0.7277383804321289, "learning_rate": 6.949828761051208e-05, "loss": 0.062, "step": 32310 }, { "grad_norm": 0.340151846408844, "learning_rate": 6.947925390074622e-05, "loss": 0.0332, "step": 32320 }, { "grad_norm": 0.5340989232063293, "learning_rate": 6.946021686250863e-05, "loss": 0.0405, "step": 32330 }, { "grad_norm": 0.5743739008903503, "learning_rate": 6.94411764990522e-05, "loss": 0.0485, "step": 32340 }, { "grad_norm": 0.4492305815219879, "learning_rate": 6.942213281363044e-05, "loss": 0.0435, "step": 32350 }, { "grad_norm": 0.784281849861145, "learning_rate": 6.940308580949737e-05, "loss": 0.0466, "step": 32360 }, { "grad_norm": 0.43943727016448975, "learning_rate": 6.93840354899076e-05, "loss": 0.0403, "step": 32370 }, { "grad_norm": 0.8607339262962341, "learning_rate": 6.936498185811633e-05, "loss": 0.0491, "step": 32380 }, { "grad_norm": 0.7819705605506897, "learning_rate": 6.93459249173793e-05, "loss": 0.052, "step": 32390 }, { "grad_norm": 0.5633111000061035, "learning_rate": 6.932686467095279e-05, "loss": 0.0411, "step": 32400 }, { "grad_norm": 0.4593973755836487, "learning_rate": 6.930780112209373e-05, "loss": 0.0689, "step": 32410 }, { "grad_norm": 0.6373540163040161, "learning_rate": 6.92887342740595e-05, "loss": 0.0504, "step": 32420 }, { "grad_norm": 0.3290124237537384, "learning_rate": 6.926966413010816e-05, "loss": 0.0411, "step": 32430 }, { "grad_norm": 0.43515896797180176, "learning_rate": 6.925059069349824e-05, "loss": 0.0371, "step": 32440 }, { "grad_norm": 0.46044406294822693, "learning_rate": 6.923151396748886e-05, "loss": 0.0543, "step": 32450 }, { "grad_norm": 0.46577170491218567, "learning_rate": 6.921243395533974e-05, "loss": 0.0544, "step": 32460 }, { "grad_norm": 0.45022642612457275, "learning_rate": 6.919335066031109e-05, "loss": 0.0505, "step": 32470 }, { "grad_norm": 0.7660912275314331, "learning_rate": 6.917426408566379e-05, "loss": 0.0538, "step": 32480 }, { "grad_norm": 0.3314548432826996, "learning_rate": 6.915517423465916e-05, "loss": 0.0435, "step": 32490 }, { "grad_norm": 0.5379778742790222, "learning_rate": 6.913608111055914e-05, "loss": 0.0411, "step": 32500 }, { "grad_norm": 0.4670667350292206, "learning_rate": 6.911698471662623e-05, "loss": 0.0374, "step": 32510 }, { "grad_norm": 0.8989704847335815, "learning_rate": 6.90978850561235e-05, "loss": 0.0498, "step": 32520 }, { "grad_norm": 0.3790876269340515, "learning_rate": 6.907878213231454e-05, "loss": 0.0377, "step": 32530 }, { "grad_norm": 0.4998003840446472, "learning_rate": 6.90596759484635e-05, "loss": 0.0483, "step": 32540 }, { "grad_norm": 0.48955580592155457, "learning_rate": 6.904056650783514e-05, "loss": 0.0549, "step": 32550 }, { "grad_norm": 0.6187520027160645, "learning_rate": 6.902145381369471e-05, "loss": 0.0401, "step": 32560 }, { "grad_norm": 0.5744567513465881, "learning_rate": 6.900233786930808e-05, "loss": 0.0516, "step": 32570 }, { "grad_norm": 0.7463847398757935, "learning_rate": 6.898321867794161e-05, "loss": 0.0522, "step": 32580 }, { "grad_norm": 0.5959059596061707, "learning_rate": 6.896409624286226e-05, "loss": 0.0392, "step": 32590 }, { "grad_norm": 0.35803645849227905, "learning_rate": 6.894497056733754e-05, "loss": 0.0418, "step": 32600 }, { "grad_norm": 0.5125581622123718, "learning_rate": 6.89258416546355e-05, "loss": 0.0471, "step": 32610 }, { "grad_norm": 0.7851459980010986, "learning_rate": 6.890670950802474e-05, "loss": 0.0716, "step": 32620 }, { "grad_norm": 0.615053117275238, "learning_rate": 6.88875741307744e-05, "loss": 0.0518, "step": 32630 }, { "grad_norm": 0.49181970953941345, "learning_rate": 6.886843552615425e-05, "loss": 0.0499, "step": 32640 }, { "grad_norm": 0.36682116985321045, "learning_rate": 6.884929369743451e-05, "loss": 0.0653, "step": 32650 }, { "grad_norm": 0.6505691409111023, "learning_rate": 6.8830148647886e-05, "loss": 0.0487, "step": 32660 }, { "grad_norm": 0.4478042423725128, "learning_rate": 6.88110003807801e-05, "loss": 0.0421, "step": 32670 }, { "grad_norm": 0.4282994866371155, "learning_rate": 6.87918488993887e-05, "loss": 0.0431, "step": 32680 }, { "grad_norm": 0.5161975622177124, "learning_rate": 6.877269420698431e-05, "loss": 0.0373, "step": 32690 }, { "grad_norm": 0.7587684988975525, "learning_rate": 6.875353630683989e-05, "loss": 0.0401, "step": 32700 }, { "grad_norm": 0.48530328273773193, "learning_rate": 6.873437520222905e-05, "loss": 0.059, "step": 32710 }, { "grad_norm": 0.5563420653343201, "learning_rate": 6.871521089642585e-05, "loss": 0.0373, "step": 32720 }, { "grad_norm": 0.4722117483615875, "learning_rate": 6.869604339270498e-05, "loss": 0.0428, "step": 32730 }, { "grad_norm": 0.6833545565605164, "learning_rate": 6.867687269434164e-05, "loss": 0.0624, "step": 32740 }, { "grad_norm": 0.7153725028038025, "learning_rate": 6.865769880461156e-05, "loss": 0.0616, "step": 32750 }, { "grad_norm": 0.49789300560951233, "learning_rate": 6.863852172679104e-05, "loss": 0.0336, "step": 32760 }, { "grad_norm": 0.39830195903778076, "learning_rate": 6.861934146415693e-05, "loss": 0.0563, "step": 32770 }, { "grad_norm": 0.28970247507095337, "learning_rate": 6.86001580199866e-05, "loss": 0.0609, "step": 32780 }, { "grad_norm": 0.6062127947807312, "learning_rate": 6.858097139755798e-05, "loss": 0.0502, "step": 32790 }, { "grad_norm": 0.5783467292785645, "learning_rate": 6.856178160014955e-05, "loss": 0.0587, "step": 32800 }, { "grad_norm": 0.5841614007949829, "learning_rate": 6.85425886310403e-05, "loss": 0.057, "step": 32810 }, { "grad_norm": 0.6609606146812439, "learning_rate": 6.852339249350979e-05, "loss": 0.053, "step": 32820 }, { "grad_norm": 0.4669267237186432, "learning_rate": 6.850419319083812e-05, "loss": 0.044, "step": 32830 }, { "grad_norm": 0.5995356440544128, "learning_rate": 6.848499072630592e-05, "loss": 0.05, "step": 32840 }, { "grad_norm": 0.4980786442756653, "learning_rate": 6.846578510319439e-05, "loss": 0.0451, "step": 32850 }, { "grad_norm": 0.5614827275276184, "learning_rate": 6.844657632478519e-05, "loss": 0.049, "step": 32860 }, { "grad_norm": 0.5963683128356934, "learning_rate": 6.842736439436063e-05, "loss": 0.04, "step": 32870 }, { "grad_norm": 0.40620744228363037, "learning_rate": 6.84081493152035e-05, "loss": 0.0505, "step": 32880 }, { "grad_norm": 0.5160734057426453, "learning_rate": 6.83889310905971e-05, "loss": 0.0434, "step": 32890 }, { "grad_norm": 0.5105470418930054, "learning_rate": 6.836970972382533e-05, "loss": 0.0376, "step": 32900 }, { "grad_norm": 0.46117615699768066, "learning_rate": 6.835048521817257e-05, "loss": 0.0507, "step": 32910 }, { "grad_norm": 0.40369608998298645, "learning_rate": 6.833125757692379e-05, "loss": 0.0683, "step": 32920 }, { "grad_norm": 0.4140407145023346, "learning_rate": 6.831202680336441e-05, "loss": 0.0449, "step": 32930 }, { "grad_norm": 0.6959244012832642, "learning_rate": 6.829279290078052e-05, "loss": 0.0484, "step": 32940 }, { "grad_norm": 0.35989779233932495, "learning_rate": 6.827355587245863e-05, "loss": 0.039, "step": 32950 }, { "grad_norm": 0.4821512699127197, "learning_rate": 6.82543157216858e-05, "loss": 0.0623, "step": 32960 }, { "grad_norm": 0.4773100018501282, "learning_rate": 6.823507245174969e-05, "loss": 0.0375, "step": 32970 }, { "grad_norm": 0.8411384224891663, "learning_rate": 6.821582606593841e-05, "loss": 0.0674, "step": 32980 }, { "grad_norm": 0.4565982520580292, "learning_rate": 6.81965765675407e-05, "loss": 0.0416, "step": 32990 }, { "grad_norm": 0.5511131882667542, "learning_rate": 6.81773239598457e-05, "loss": 0.0485, "step": 33000 }, { "grad_norm": 0.7124353051185608, "learning_rate": 6.815806824614319e-05, "loss": 0.0449, "step": 33010 }, { "grad_norm": 0.5395668148994446, "learning_rate": 6.813880942972343e-05, "loss": 0.0404, "step": 33020 }, { "grad_norm": 0.5307108759880066, "learning_rate": 6.811954751387726e-05, "loss": 0.0519, "step": 33030 }, { "grad_norm": 0.4718398451805115, "learning_rate": 6.810028250189598e-05, "loss": 0.047, "step": 33040 }, { "grad_norm": 0.4012863337993622, "learning_rate": 6.808101439707147e-05, "loss": 0.0426, "step": 33050 }, { "grad_norm": 0.5795778036117554, "learning_rate": 6.806174320269609e-05, "loss": 0.0495, "step": 33060 }, { "grad_norm": 0.3803556561470032, "learning_rate": 6.804246892206281e-05, "loss": 0.0427, "step": 33070 }, { "grad_norm": 0.8154512643814087, "learning_rate": 6.802319155846506e-05, "loss": 0.049, "step": 33080 }, { "grad_norm": 0.5108463764190674, "learning_rate": 6.800391111519679e-05, "loss": 0.0496, "step": 33090 }, { "grad_norm": 0.6640890836715698, "learning_rate": 6.798462759555253e-05, "loss": 0.0467, "step": 33100 }, { "grad_norm": 0.43295422196388245, "learning_rate": 6.79653410028273e-05, "loss": 0.0579, "step": 33110 }, { "grad_norm": 0.5215921401977539, "learning_rate": 6.794605134031663e-05, "loss": 0.0561, "step": 33120 }, { "grad_norm": 0.38847097754478455, "learning_rate": 6.792675861131661e-05, "loss": 0.0491, "step": 33130 }, { "grad_norm": 0.42019134759902954, "learning_rate": 6.790746281912386e-05, "loss": 0.0389, "step": 33140 }, { "grad_norm": 0.5898679494857788, "learning_rate": 6.788816396703546e-05, "loss": 0.0589, "step": 33150 }, { "grad_norm": 0.6782401204109192, "learning_rate": 6.78688620583491e-05, "loss": 0.0555, "step": 33160 }, { "grad_norm": 0.44546326994895935, "learning_rate": 6.784955709636292e-05, "loss": 0.0355, "step": 33170 }, { "grad_norm": 0.4429086744785309, "learning_rate": 6.783024908437564e-05, "loss": 0.0419, "step": 33180 }, { "grad_norm": 1.3811832666397095, "learning_rate": 6.781093802568641e-05, "loss": 0.0491, "step": 33190 }, { "grad_norm": 0.5926406383514404, "learning_rate": 6.779162392359504e-05, "loss": 0.0394, "step": 33200 }, { "grad_norm": 0.4308987855911255, "learning_rate": 6.777230678140172e-05, "loss": 0.0374, "step": 33210 }, { "grad_norm": 0.44524574279785156, "learning_rate": 6.775298660240726e-05, "loss": 0.0396, "step": 33220 }, { "grad_norm": 0.48147958517074585, "learning_rate": 6.773366338991292e-05, "loss": 0.044, "step": 33230 }, { "grad_norm": 0.380649209022522, "learning_rate": 6.771433714722052e-05, "loss": 0.0407, "step": 33240 }, { "grad_norm": 0.5373631119728088, "learning_rate": 6.769500787763239e-05, "loss": 0.0408, "step": 33250 }, { "grad_norm": 0.3213772773742676, "learning_rate": 6.76756755844514e-05, "loss": 0.0412, "step": 33260 }, { "grad_norm": 0.3212015926837921, "learning_rate": 6.765634027098087e-05, "loss": 0.0423, "step": 33270 }, { "grad_norm": 0.33873215317726135, "learning_rate": 6.763700194052468e-05, "loss": 0.0552, "step": 33280 }, { "grad_norm": 0.4073248505592346, "learning_rate": 6.761766059638723e-05, "loss": 0.0586, "step": 33290 }, { "grad_norm": 0.5903339982032776, "learning_rate": 6.759831624187345e-05, "loss": 0.054, "step": 33300 }, { "grad_norm": 0.6039458513259888, "learning_rate": 6.757896888028871e-05, "loss": 0.0426, "step": 33310 }, { "grad_norm": 0.47373658418655396, "learning_rate": 6.7559618514939e-05, "loss": 0.0504, "step": 33320 }, { "grad_norm": 0.6603068709373474, "learning_rate": 6.754026514913073e-05, "loss": 0.0491, "step": 33330 }, { "grad_norm": 0.5789900422096252, "learning_rate": 6.752090878617087e-05, "loss": 0.0753, "step": 33340 }, { "grad_norm": 0.4708380699157715, "learning_rate": 6.75015494293669e-05, "loss": 0.0345, "step": 33350 }, { "grad_norm": 0.5962815284729004, "learning_rate": 6.74821870820268e-05, "loss": 0.0593, "step": 33360 }, { "grad_norm": 0.546952486038208, "learning_rate": 6.746282174745907e-05, "loss": 0.0435, "step": 33370 }, { "grad_norm": 0.5163534879684448, "learning_rate": 6.744345342897271e-05, "loss": 0.0521, "step": 33380 }, { "grad_norm": 0.3742934763431549, "learning_rate": 6.742408212987724e-05, "loss": 0.0431, "step": 33390 }, { "grad_norm": 0.5053467154502869, "learning_rate": 6.740470785348269e-05, "loss": 0.0496, "step": 33400 }, { "grad_norm": 0.41561374068260193, "learning_rate": 6.738533060309958e-05, "loss": 0.0527, "step": 33410 }, { "grad_norm": 0.5922508835792542, "learning_rate": 6.736595038203894e-05, "loss": 0.0453, "step": 33420 }, { "grad_norm": 0.7498428821563721, "learning_rate": 6.734656719361236e-05, "loss": 0.0635, "step": 33430 }, { "grad_norm": 0.5393611192703247, "learning_rate": 6.732718104113189e-05, "loss": 0.0629, "step": 33440 }, { "grad_norm": 0.4727462828159332, "learning_rate": 6.730779192791006e-05, "loss": 0.0455, "step": 33450 }, { "grad_norm": 0.4953755736351013, "learning_rate": 6.728839985725997e-05, "loss": 0.0703, "step": 33460 }, { "grad_norm": 0.4663527011871338, "learning_rate": 6.726900483249517e-05, "loss": 0.0566, "step": 33470 }, { "grad_norm": 0.4822884500026703, "learning_rate": 6.724960685692976e-05, "loss": 0.0459, "step": 33480 }, { "grad_norm": 0.5119364857673645, "learning_rate": 6.723020593387833e-05, "loss": 0.0498, "step": 33490 }, { "grad_norm": 0.5936525464057922, "learning_rate": 6.721080206665593e-05, "loss": 0.0573, "step": 33500 }, { "grad_norm": 0.3783983886241913, "learning_rate": 6.719139525857819e-05, "loss": 0.0361, "step": 33510 }, { "grad_norm": 0.6682852506637573, "learning_rate": 6.717198551296117e-05, "loss": 0.045, "step": 33520 }, { "grad_norm": 0.43982213735580444, "learning_rate": 6.715257283312148e-05, "loss": 0.0443, "step": 33530 }, { "grad_norm": 0.3821674883365631, "learning_rate": 6.713315722237623e-05, "loss": 0.0366, "step": 33540 }, { "grad_norm": 0.4315384328365326, "learning_rate": 6.7113738684043e-05, "loss": 0.0394, "step": 33550 }, { "grad_norm": 0.4685547351837158, "learning_rate": 6.709431722143989e-05, "loss": 0.0451, "step": 33560 }, { "grad_norm": 0.7438979148864746, "learning_rate": 6.70748928378855e-05, "loss": 0.0491, "step": 33570 }, { "grad_norm": 0.4618809223175049, "learning_rate": 6.705546553669891e-05, "loss": 0.0363, "step": 33580 }, { "grad_norm": 0.431601881980896, "learning_rate": 6.703603532119974e-05, "loss": 0.0495, "step": 33590 }, { "grad_norm": 0.43028581142425537, "learning_rate": 6.701660219470808e-05, "loss": 0.049, "step": 33600 }, { "grad_norm": 0.542759895324707, "learning_rate": 6.69971661605445e-05, "loss": 0.0414, "step": 33610 }, { "grad_norm": 0.7784504294395447, "learning_rate": 6.697772722203008e-05, "loss": 0.0445, "step": 33620 }, { "grad_norm": 0.5541027188301086, "learning_rate": 6.695828538248643e-05, "loss": 0.043, "step": 33630 }, { "grad_norm": 0.4609900712966919, "learning_rate": 6.693884064523563e-05, "loss": 0.047, "step": 33640 }, { "grad_norm": 0.6201260685920715, "learning_rate": 6.691939301360023e-05, "loss": 0.0492, "step": 33650 }, { "grad_norm": 0.5164437294006348, "learning_rate": 6.689994249090333e-05, "loss": 0.041, "step": 33660 }, { "grad_norm": 0.40346184372901917, "learning_rate": 6.688048908046845e-05, "loss": 0.0385, "step": 33670 }, { "grad_norm": 0.6172037124633789, "learning_rate": 6.686103278561969e-05, "loss": 0.0395, "step": 33680 }, { "grad_norm": 0.4575740694999695, "learning_rate": 6.684157360968156e-05, "loss": 0.0462, "step": 33690 }, { "grad_norm": 0.7521384954452515, "learning_rate": 6.682211155597911e-05, "loss": 0.0484, "step": 33700 }, { "grad_norm": 0.5380110740661621, "learning_rate": 6.680264662783789e-05, "loss": 0.0401, "step": 33710 }, { "grad_norm": 0.5540778636932373, "learning_rate": 6.678317882858391e-05, "loss": 0.0732, "step": 33720 }, { "grad_norm": 0.6521415710449219, "learning_rate": 6.67637081615437e-05, "loss": 0.0476, "step": 33730 }, { "grad_norm": 0.56541508436203, "learning_rate": 6.674423463004427e-05, "loss": 0.0426, "step": 33740 }, { "grad_norm": 0.6375212073326111, "learning_rate": 6.672475823741308e-05, "loss": 0.0486, "step": 33750 }, { "grad_norm": 0.590222179889679, "learning_rate": 6.670527898697811e-05, "loss": 0.0537, "step": 33760 }, { "grad_norm": 0.5655605792999268, "learning_rate": 6.668579688206788e-05, "loss": 0.0445, "step": 33770 }, { "grad_norm": 0.5341977477073669, "learning_rate": 6.666631192601131e-05, "loss": 0.0495, "step": 33780 }, { "grad_norm": 0.46761491894721985, "learning_rate": 6.664682412213785e-05, "loss": 0.0458, "step": 33790 }, { "grad_norm": 0.4576612114906311, "learning_rate": 6.662733347377745e-05, "loss": 0.0419, "step": 33800 }, { "grad_norm": 0.39236027002334595, "learning_rate": 6.660783998426051e-05, "loss": 0.0406, "step": 33810 }, { "grad_norm": 0.6558806300163269, "learning_rate": 6.658834365691794e-05, "loss": 0.0459, "step": 33820 }, { "grad_norm": 0.5346777439117432, "learning_rate": 6.656884449508115e-05, "loss": 0.0668, "step": 33830 }, { "grad_norm": 0.5192512273788452, "learning_rate": 6.654934250208198e-05, "loss": 0.0417, "step": 33840 }, { "grad_norm": 0.47741371393203735, "learning_rate": 6.65298376812528e-05, "loss": 0.0475, "step": 33850 }, { "grad_norm": 0.6401151418685913, "learning_rate": 6.651033003592646e-05, "loss": 0.0478, "step": 33860 }, { "grad_norm": 0.4774327576160431, "learning_rate": 6.649081956943626e-05, "loss": 0.0394, "step": 33870 }, { "grad_norm": 0.5798813104629517, "learning_rate": 6.647130628511604e-05, "loss": 0.0401, "step": 33880 }, { "grad_norm": 0.3284071683883667, "learning_rate": 6.645179018630005e-05, "loss": 0.0533, "step": 33890 }, { "grad_norm": 0.7225894927978516, "learning_rate": 6.643227127632309e-05, "loss": 0.0657, "step": 33900 }, { "grad_norm": 0.3934440314769745, "learning_rate": 6.641274955852038e-05, "loss": 0.0445, "step": 33910 }, { "grad_norm": 0.5049543976783752, "learning_rate": 6.639322503622768e-05, "loss": 0.0478, "step": 33920 }, { "grad_norm": 0.5020114183425903, "learning_rate": 6.637369771278116e-05, "loss": 0.0562, "step": 33930 }, { "grad_norm": 0.48464691638946533, "learning_rate": 6.635416759151751e-05, "loss": 0.0495, "step": 33940 }, { "grad_norm": 0.4526793956756592, "learning_rate": 6.633463467577394e-05, "loss": 0.0546, "step": 33950 }, { "grad_norm": 0.5818130970001221, "learning_rate": 6.631509896888803e-05, "loss": 0.0489, "step": 33960 }, { "grad_norm": 0.6869866847991943, "learning_rate": 6.629556047419794e-05, "loss": 0.0638, "step": 33970 }, { "grad_norm": 0.5834659934043884, "learning_rate": 6.627601919504223e-05, "loss": 0.0569, "step": 33980 }, { "grad_norm": 0.550663948059082, "learning_rate": 6.625647513476001e-05, "loss": 0.0439, "step": 33990 }, { "grad_norm": 0.757373034954071, "learning_rate": 6.62369282966908e-05, "loss": 0.0548, "step": 34000 }, { "grad_norm": 0.576371967792511, "learning_rate": 6.621737868417464e-05, "loss": 0.043, "step": 34010 }, { "grad_norm": 0.6513667106628418, "learning_rate": 6.619782630055198e-05, "loss": 0.0676, "step": 34020 }, { "grad_norm": 0.6952188611030579, "learning_rate": 6.617827114916382e-05, "loss": 0.0563, "step": 34030 }, { "grad_norm": 0.4689405858516693, "learning_rate": 6.615871323335161e-05, "loss": 0.06, "step": 34040 }, { "grad_norm": 0.5391433835029602, "learning_rate": 6.613915255645725e-05, "loss": 0.053, "step": 34050 }, { "grad_norm": 0.45975932478904724, "learning_rate": 6.611958912182312e-05, "loss": 0.0393, "step": 34060 }, { "grad_norm": 0.5516810417175293, "learning_rate": 6.610002293279207e-05, "loss": 0.0499, "step": 34070 }, { "grad_norm": 0.43242159485816956, "learning_rate": 6.608045399270746e-05, "loss": 0.04, "step": 34080 }, { "grad_norm": 0.4535503089427948, "learning_rate": 6.606088230491304e-05, "loss": 0.0406, "step": 34090 }, { "grad_norm": 0.9324324727058411, "learning_rate": 6.604130787275312e-05, "loss": 0.0512, "step": 34100 }, { "grad_norm": 0.46248409152030945, "learning_rate": 6.602173069957242e-05, "loss": 0.0456, "step": 34110 }, { "grad_norm": 0.6931774616241455, "learning_rate": 6.600215078871612e-05, "loss": 0.0752, "step": 34120 }, { "grad_norm": 0.5166140198707581, "learning_rate": 6.598256814352992e-05, "loss": 0.0445, "step": 34130 }, { "grad_norm": 0.4688939154148102, "learning_rate": 6.596298276735995e-05, "loss": 0.0422, "step": 34140 }, { "grad_norm": 0.47248324751853943, "learning_rate": 6.594339466355282e-05, "loss": 0.0504, "step": 34150 }, { "grad_norm": 0.6409335136413574, "learning_rate": 6.592380383545558e-05, "loss": 0.0388, "step": 34160 }, { "grad_norm": 0.3455584943294525, "learning_rate": 6.590421028641577e-05, "loss": 0.0428, "step": 34170 }, { "grad_norm": 0.42421770095825195, "learning_rate": 6.588461401978143e-05, "loss": 0.0501, "step": 34180 }, { "grad_norm": 0.39902475476264954, "learning_rate": 6.586501503890099e-05, "loss": 0.0438, "step": 34190 }, { "grad_norm": 0.501176118850708, "learning_rate": 6.584541334712338e-05, "loss": 0.0416, "step": 34200 }, { "grad_norm": 0.670672595500946, "learning_rate": 6.582580894779802e-05, "loss": 0.0321, "step": 34210 }, { "grad_norm": 0.6583379507064819, "learning_rate": 6.580620184427473e-05, "loss": 0.0501, "step": 34220 }, { "grad_norm": 0.5883367657661438, "learning_rate": 6.578659203990385e-05, "loss": 0.046, "step": 34230 }, { "grad_norm": 0.5498661994934082, "learning_rate": 6.576697953803615e-05, "loss": 0.043, "step": 34240 }, { "grad_norm": 0.43037453293800354, "learning_rate": 6.57473643420229e-05, "loss": 0.0379, "step": 34250 }, { "grad_norm": 0.47237008810043335, "learning_rate": 6.572774645521574e-05, "loss": 0.0464, "step": 34260 }, { "grad_norm": 0.44977959990501404, "learning_rate": 6.570812588096688e-05, "loss": 0.0447, "step": 34270 }, { "grad_norm": 0.47644513845443726, "learning_rate": 6.568850262262893e-05, "loss": 0.0355, "step": 34280 }, { "grad_norm": 0.4979630708694458, "learning_rate": 6.566887668355497e-05, "loss": 0.0475, "step": 34290 }, { "grad_norm": 0.47303566336631775, "learning_rate": 6.564924806709851e-05, "loss": 0.0478, "step": 34300 }, { "grad_norm": 0.49072888493537903, "learning_rate": 6.562961677661359e-05, "loss": 0.0504, "step": 34310 }, { "grad_norm": 0.49418869614601135, "learning_rate": 6.56099828154546e-05, "loss": 0.0501, "step": 34320 }, { "grad_norm": 0.668373703956604, "learning_rate": 6.55903461869765e-05, "loss": 0.0558, "step": 34330 }, { "grad_norm": 0.5776866674423218, "learning_rate": 6.557070689453465e-05, "loss": 0.0397, "step": 34340 }, { "grad_norm": 0.7112862467765808, "learning_rate": 6.555106494148482e-05, "loss": 0.0385, "step": 34350 }, { "grad_norm": 0.7261610627174377, "learning_rate": 6.553142033118333e-05, "loss": 0.0513, "step": 34360 }, { "grad_norm": 0.4950915575027466, "learning_rate": 6.551177306698688e-05, "loss": 0.0438, "step": 34370 }, { "grad_norm": 0.5595852732658386, "learning_rate": 6.549212315225267e-05, "loss": 0.0401, "step": 34380 }, { "grad_norm": 0.41426223516464233, "learning_rate": 6.547247059033833e-05, "loss": 0.0416, "step": 34390 }, { "grad_norm": 0.5142017006874084, "learning_rate": 6.545281538460193e-05, "loss": 0.0506, "step": 34400 }, { "grad_norm": 0.47200044989585876, "learning_rate": 6.543315753840202e-05, "loss": 0.0381, "step": 34410 }, { "grad_norm": 0.5717721581459045, "learning_rate": 6.541349705509758e-05, "loss": 0.046, "step": 34420 }, { "grad_norm": 0.4991680085659027, "learning_rate": 6.539383393804805e-05, "loss": 0.0432, "step": 34430 }, { "grad_norm": 0.5112307667732239, "learning_rate": 6.537416819061333e-05, "loss": 0.0421, "step": 34440 }, { "grad_norm": 0.5387911200523376, "learning_rate": 6.535449981615375e-05, "loss": 0.0495, "step": 34450 }, { "grad_norm": 0.4672732651233673, "learning_rate": 6.53348288180301e-05, "loss": 0.0403, "step": 34460 }, { "grad_norm": 0.46517133712768555, "learning_rate": 6.531515519960361e-05, "loss": 0.0409, "step": 34470 }, { "grad_norm": 0.5800597071647644, "learning_rate": 6.529547896423597e-05, "loss": 0.0549, "step": 34480 }, { "grad_norm": 0.33106526732444763, "learning_rate": 6.52758001152893e-05, "loss": 0.0782, "step": 34490 }, { "grad_norm": 0.6415481567382812, "learning_rate": 6.525611865612618e-05, "loss": 0.0513, "step": 34500 }, { "grad_norm": 0.3450855314731598, "learning_rate": 6.523643459010966e-05, "loss": 0.043, "step": 34510 }, { "grad_norm": 0.45390790700912476, "learning_rate": 6.521674792060317e-05, "loss": 0.0406, "step": 34520 }, { "grad_norm": 0.6186999678611755, "learning_rate": 6.519705865097063e-05, "loss": 0.0514, "step": 34530 }, { "grad_norm": 0.3381117284297943, "learning_rate": 6.517736678457641e-05, "loss": 0.0405, "step": 34540 }, { "grad_norm": 0.558711588382721, "learning_rate": 6.515767232478534e-05, "loss": 0.0444, "step": 34550 }, { "grad_norm": 0.4731491208076477, "learning_rate": 6.51379752749626e-05, "loss": 0.0388, "step": 34560 }, { "grad_norm": 0.6652040481567383, "learning_rate": 6.511827563847393e-05, "loss": 0.0405, "step": 34570 }, { "grad_norm": 0.550664484500885, "learning_rate": 6.509857341868542e-05, "loss": 0.0559, "step": 34580 }, { "grad_norm": 0.5432375073432922, "learning_rate": 6.507886861896367e-05, "loss": 0.0395, "step": 34590 }, { "grad_norm": 0.4649987816810608, "learning_rate": 6.505916124267567e-05, "loss": 0.0372, "step": 34600 }, { "grad_norm": 0.4566579759120941, "learning_rate": 6.503945129318891e-05, "loss": 0.0571, "step": 34610 }, { "grad_norm": 0.3959982097148895, "learning_rate": 6.501973877387122e-05, "loss": 0.0378, "step": 34620 }, { "grad_norm": 0.6243671774864197, "learning_rate": 6.500002368809098e-05, "loss": 0.0349, "step": 34630 }, { "grad_norm": 0.4566273093223572, "learning_rate": 6.498030603921694e-05, "loss": 0.0392, "step": 34640 }, { "grad_norm": 0.5698886513710022, "learning_rate": 6.496058583061832e-05, "loss": 0.0413, "step": 34650 }, { "grad_norm": 0.5200355648994446, "learning_rate": 6.494086306566475e-05, "loss": 0.0411, "step": 34660 }, { "grad_norm": 0.7958352565765381, "learning_rate": 6.492113774772632e-05, "loss": 0.0605, "step": 34670 }, { "grad_norm": 0.44362759590148926, "learning_rate": 6.490140988017354e-05, "loss": 0.0425, "step": 34680 }, { "grad_norm": 0.3785235285758972, "learning_rate": 6.488167946637736e-05, "loss": 0.0493, "step": 34690 }, { "grad_norm": 0.9079923033714294, "learning_rate": 6.486194650970915e-05, "loss": 0.0388, "step": 34700 }, { "grad_norm": 0.6298394203186035, "learning_rate": 6.48422110135408e-05, "loss": 0.047, "step": 34710 }, { "grad_norm": 0.6137121915817261, "learning_rate": 6.482247298124451e-05, "loss": 0.041, "step": 34720 }, { "grad_norm": 0.3644055724143982, "learning_rate": 6.480273241619297e-05, "loss": 0.0469, "step": 34730 }, { "grad_norm": 0.687587320804596, "learning_rate": 6.478298932175933e-05, "loss": 0.0484, "step": 34740 }, { "grad_norm": 0.47886011004447937, "learning_rate": 6.476324370131712e-05, "loss": 0.0494, "step": 34750 }, { "grad_norm": 0.5577468276023865, "learning_rate": 6.474349555824036e-05, "loss": 0.0365, "step": 34760 }, { "grad_norm": 0.5495826601982117, "learning_rate": 6.472374489590342e-05, "loss": 0.0563, "step": 34770 }, { "grad_norm": 0.2946663796901703, "learning_rate": 6.470399171768118e-05, "loss": 0.0352, "step": 34780 }, { "grad_norm": 0.5120938420295715, "learning_rate": 6.468423602694891e-05, "loss": 0.0393, "step": 34790 }, { "grad_norm": 0.45027363300323486, "learning_rate": 6.466447782708232e-05, "loss": 0.0523, "step": 34800 }, { "grad_norm": 0.5467864274978638, "learning_rate": 6.464471712145754e-05, "loss": 0.0402, "step": 34810 }, { "grad_norm": 0.5104268789291382, "learning_rate": 6.462495391345114e-05, "loss": 0.0303, "step": 34820 }, { "grad_norm": 0.2837316393852234, "learning_rate": 6.46051882064401e-05, "loss": 0.0394, "step": 34830 }, { "grad_norm": 0.5485310554504395, "learning_rate": 6.458542000380186e-05, "loss": 0.0551, "step": 34840 }, { "grad_norm": 0.4940124750137329, "learning_rate": 6.456564930891424e-05, "loss": 0.0478, "step": 34850 }, { "grad_norm": 0.8468431830406189, "learning_rate": 6.454587612515555e-05, "loss": 0.0432, "step": 34860 }, { "grad_norm": 0.430270254611969, "learning_rate": 6.452610045590444e-05, "loss": 0.0387, "step": 34870 }, { "grad_norm": 0.4548019468784332, "learning_rate": 6.450632230454005e-05, "loss": 0.0627, "step": 34880 }, { "grad_norm": 0.4251865744590759, "learning_rate": 6.448654167444195e-05, "loss": 0.0425, "step": 34890 }, { "grad_norm": 0.5355004072189331, "learning_rate": 6.446675856899005e-05, "loss": 0.044, "step": 34900 }, { "grad_norm": 0.5154012441635132, "learning_rate": 6.444697299156481e-05, "loss": 0.0456, "step": 34910 }, { "grad_norm": 0.5027465224266052, "learning_rate": 6.442718494554701e-05, "loss": 0.0458, "step": 34920 }, { "grad_norm": 0.5302962064743042, "learning_rate": 6.440739443431787e-05, "loss": 0.043, "step": 34930 }, { "grad_norm": 0.6156870126724243, "learning_rate": 6.438760146125906e-05, "loss": 0.0337, "step": 34940 }, { "grad_norm": 0.6559486389160156, "learning_rate": 6.436780602975267e-05, "loss": 0.0419, "step": 34950 }, { "grad_norm": 0.4644981622695923, "learning_rate": 6.43480081431812e-05, "loss": 0.0444, "step": 34960 }, { "grad_norm": 0.39894595742225647, "learning_rate": 6.432820780492756e-05, "loss": 0.0445, "step": 34970 }, { "grad_norm": 0.6742246747016907, "learning_rate": 6.430840501837506e-05, "loss": 0.0419, "step": 34980 }, { "grad_norm": 0.6699027419090271, "learning_rate": 6.428859978690748e-05, "loss": 0.0501, "step": 34990 }, { "grad_norm": 0.5285466909408569, "learning_rate": 6.426879211390901e-05, "loss": 0.0454, "step": 35000 }, { "grad_norm": 0.6627843379974365, "learning_rate": 6.424898200276422e-05, "loss": 0.0553, "step": 35010 }, { "grad_norm": 0.47524121403694153, "learning_rate": 6.42291694568581e-05, "loss": 0.0477, "step": 35020 }, { "grad_norm": 0.43455591797828674, "learning_rate": 6.42093544795761e-05, "loss": 0.0588, "step": 35030 }, { "grad_norm": 0.3392829895019531, "learning_rate": 6.418953707430403e-05, "loss": 0.0459, "step": 35040 }, { "grad_norm": 0.6668872833251953, "learning_rate": 6.416971724442819e-05, "loss": 0.049, "step": 35050 }, { "grad_norm": 0.39268526434898376, "learning_rate": 6.414989499333519e-05, "loss": 0.0467, "step": 35060 }, { "grad_norm": 0.6481648683547974, "learning_rate": 6.413007032441214e-05, "loss": 0.0646, "step": 35070 }, { "grad_norm": 0.31916776299476624, "learning_rate": 6.411024324104653e-05, "loss": 0.0412, "step": 35080 }, { "grad_norm": 0.5329338908195496, "learning_rate": 6.409041374662628e-05, "loss": 0.0435, "step": 35090 }, { "grad_norm": 0.370484322309494, "learning_rate": 6.407058184453967e-05, "loss": 0.0429, "step": 35100 }, { "grad_norm": 0.5366207361221313, "learning_rate": 6.405074753817548e-05, "loss": 0.0608, "step": 35110 }, { "grad_norm": 0.43772780895233154, "learning_rate": 6.40309108309228e-05, "loss": 0.0538, "step": 35120 }, { "grad_norm": 0.4070051610469818, "learning_rate": 6.401107172617122e-05, "loss": 0.0494, "step": 35130 }, { "grad_norm": 0.8192091584205627, "learning_rate": 6.399123022731068e-05, "loss": 0.0625, "step": 35140 }, { "grad_norm": 0.6562430262565613, "learning_rate": 6.397138633773157e-05, "loss": 0.0454, "step": 35150 }, { "grad_norm": 0.6039839386940002, "learning_rate": 6.395154006082463e-05, "loss": 0.0558, "step": 35160 }, { "grad_norm": 0.5808266401290894, "learning_rate": 6.393169139998109e-05, "loss": 0.0544, "step": 35170 }, { "grad_norm": 0.922042965888977, "learning_rate": 6.39118403585925e-05, "loss": 0.0571, "step": 35180 }, { "grad_norm": 0.45491814613342285, "learning_rate": 6.38919869400509e-05, "loss": 0.0405, "step": 35190 }, { "grad_norm": 0.5762380957603455, "learning_rate": 6.387213114774865e-05, "loss": 0.0665, "step": 35200 }, { "grad_norm": 0.6476483941078186, "learning_rate": 6.385227298507863e-05, "loss": 0.0692, "step": 35210 }, { "grad_norm": 0.44239628314971924, "learning_rate": 6.3832412455434e-05, "loss": 0.0501, "step": 35220 }, { "grad_norm": 0.6344821453094482, "learning_rate": 6.381254956220841e-05, "loss": 0.0492, "step": 35230 }, { "grad_norm": 0.5833265781402588, "learning_rate": 6.379268430879586e-05, "loss": 0.0527, "step": 35240 }, { "grad_norm": 0.5520291328430176, "learning_rate": 6.37728166985908e-05, "loss": 0.0409, "step": 35250 }, { "grad_norm": 0.4207690358161926, "learning_rate": 6.375294673498804e-05, "loss": 0.0482, "step": 35260 }, { "grad_norm": 0.7429574728012085, "learning_rate": 6.373307442138284e-05, "loss": 0.0502, "step": 35270 }, { "grad_norm": 0.3326079249382019, "learning_rate": 6.371319976117081e-05, "loss": 0.0418, "step": 35280 }, { "grad_norm": 0.469893217086792, "learning_rate": 6.3693322757748e-05, "loss": 0.0498, "step": 35290 }, { "grad_norm": 0.4097830653190613, "learning_rate": 6.367344341451086e-05, "loss": 0.0307, "step": 35300 }, { "grad_norm": 0.5255021452903748, "learning_rate": 6.36535617348562e-05, "loss": 0.037, "step": 35310 }, { "grad_norm": 0.6276051998138428, "learning_rate": 6.363367772218128e-05, "loss": 0.0769, "step": 35320 }, { "grad_norm": 0.46518853306770325, "learning_rate": 6.36137913798837e-05, "loss": 0.0485, "step": 35330 }, { "grad_norm": 0.5296504497528076, "learning_rate": 6.359390271136151e-05, "loss": 0.0565, "step": 35340 }, { "grad_norm": 0.4595714211463928, "learning_rate": 6.357401172001314e-05, "loss": 0.0543, "step": 35350 }, { "grad_norm": 0.6709229946136475, "learning_rate": 6.355411840923742e-05, "loss": 0.047, "step": 35360 }, { "grad_norm": 0.6279711723327637, "learning_rate": 6.353422278243358e-05, "loss": 0.0541, "step": 35370 }, { "grad_norm": 0.4104306399822235, "learning_rate": 6.351432484300121e-05, "loss": 0.0386, "step": 35380 }, { "grad_norm": 0.5845974087715149, "learning_rate": 6.349442459434036e-05, "loss": 0.0605, "step": 35390 }, { "grad_norm": 0.526170015335083, "learning_rate": 6.34745220398514e-05, "loss": 0.0426, "step": 35400 }, { "grad_norm": 0.6811028122901917, "learning_rate": 6.345461718293518e-05, "loss": 0.0482, "step": 35410 }, { "grad_norm": 0.6667479276657104, "learning_rate": 6.343471002699286e-05, "loss": 0.0471, "step": 35420 }, { "grad_norm": 0.4534580111503601, "learning_rate": 6.341480057542602e-05, "loss": 0.0297, "step": 35430 }, { "grad_norm": 0.6037793755531311, "learning_rate": 6.339488883163667e-05, "loss": 0.0414, "step": 35440 }, { "grad_norm": 0.4877546727657318, "learning_rate": 6.337497479902716e-05, "loss": 0.0379, "step": 35450 }, { "grad_norm": 0.3807693421840668, "learning_rate": 6.335505848100027e-05, "loss": 0.044, "step": 35460 }, { "grad_norm": 0.45512232184410095, "learning_rate": 6.333513988095915e-05, "loss": 0.0562, "step": 35470 }, { "grad_norm": 0.37151843309402466, "learning_rate": 6.331521900230735e-05, "loss": 0.0646, "step": 35480 }, { "grad_norm": 0.5905119776725769, "learning_rate": 6.329529584844878e-05, "loss": 0.0486, "step": 35490 }, { "grad_norm": 0.5093281269073486, "learning_rate": 6.327537042278777e-05, "loss": 0.0712, "step": 35500 }, { "grad_norm": 0.749320387840271, "learning_rate": 6.325544272872905e-05, "loss": 0.047, "step": 35510 }, { "grad_norm": 0.5664107203483582, "learning_rate": 6.323551276967771e-05, "loss": 0.0596, "step": 35520 }, { "grad_norm": 0.4640924334526062, "learning_rate": 6.321558054903922e-05, "loss": 0.037, "step": 35530 }, { "grad_norm": 0.4734708070755005, "learning_rate": 6.319564607021947e-05, "loss": 0.0407, "step": 35540 }, { "grad_norm": 0.4715246260166168, "learning_rate": 6.31757093366247e-05, "loss": 0.0452, "step": 35550 }, { "grad_norm": 0.5079176425933838, "learning_rate": 6.315577035166154e-05, "loss": 0.0781, "step": 35560 }, { "grad_norm": 0.5711513757705688, "learning_rate": 6.313582911873708e-05, "loss": 0.0312, "step": 35570 }, { "grad_norm": 0.34044644236564636, "learning_rate": 6.311588564125865e-05, "loss": 0.035, "step": 35580 }, { "grad_norm": 0.6389060616493225, "learning_rate": 6.30959399226341e-05, "loss": 0.044, "step": 35590 }, { "grad_norm": 0.5656521320343018, "learning_rate": 6.30759919662716e-05, "loss": 0.0323, "step": 35600 }, { "grad_norm": 0.286536306142807, "learning_rate": 6.30560417755797e-05, "loss": 0.0377, "step": 35610 }, { "grad_norm": 0.5654929876327515, "learning_rate": 6.303608935396735e-05, "loss": 0.0504, "step": 35620 }, { "grad_norm": 0.5064629316329956, "learning_rate": 6.301613470484386e-05, "loss": 0.038, "step": 35630 }, { "grad_norm": 0.5407440066337585, "learning_rate": 6.299617783161893e-05, "loss": 0.0529, "step": 35640 }, { "grad_norm": 0.5527203679084778, "learning_rate": 6.297621873770266e-05, "loss": 0.0441, "step": 35650 }, { "grad_norm": 0.5226134061813354, "learning_rate": 6.29562574265055e-05, "loss": 0.05, "step": 35660 }, { "grad_norm": 0.6040967106819153, "learning_rate": 6.293629390143834e-05, "loss": 0.0416, "step": 35670 }, { "grad_norm": 0.40755993127822876, "learning_rate": 6.291632816591232e-05, "loss": 0.0383, "step": 35680 }, { "grad_norm": 0.5420337915420532, "learning_rate": 6.28963602233391e-05, "loss": 0.0408, "step": 35690 }, { "grad_norm": 0.5515817403793335, "learning_rate": 6.287639007713062e-05, "loss": 0.042, "step": 35700 }, { "grad_norm": 0.5372301936149597, "learning_rate": 6.285641773069926e-05, "loss": 0.0528, "step": 35710 }, { "grad_norm": 0.6104236841201782, "learning_rate": 6.283644318745773e-05, "loss": 0.0387, "step": 35720 }, { "grad_norm": 0.7184553146362305, "learning_rate": 6.281646645081912e-05, "loss": 0.0433, "step": 35730 }, { "grad_norm": 0.511245608329773, "learning_rate": 6.279648752419693e-05, "loss": 0.0507, "step": 35740 }, { "grad_norm": 0.44312477111816406, "learning_rate": 6.2776506411005e-05, "loss": 0.0407, "step": 35750 }, { "grad_norm": 0.6216403841972351, "learning_rate": 6.275652311465758e-05, "loss": 0.0476, "step": 35760 }, { "grad_norm": 0.6241663694381714, "learning_rate": 6.273653763856926e-05, "loss": 0.0373, "step": 35770 }, { "grad_norm": 0.4242338538169861, "learning_rate": 6.271654998615501e-05, "loss": 0.0505, "step": 35780 }, { "grad_norm": 0.5400751233100891, "learning_rate": 6.269656016083013e-05, "loss": 0.0385, "step": 35790 }, { "grad_norm": 0.4997502267360687, "learning_rate": 6.267656816601038e-05, "loss": 0.0392, "step": 35800 }, { "grad_norm": 0.6972205638885498, "learning_rate": 6.265657400511185e-05, "loss": 0.0525, "step": 35810 }, { "grad_norm": 0.562938392162323, "learning_rate": 6.263657768155098e-05, "loss": 0.0573, "step": 35820 }, { "grad_norm": 0.7920035123825073, "learning_rate": 6.261657919874457e-05, "loss": 0.0418, "step": 35830 }, { "grad_norm": 0.3667515814304352, "learning_rate": 6.259657856010986e-05, "loss": 0.0436, "step": 35840 }, { "grad_norm": 0.5729233026504517, "learning_rate": 6.257657576906439e-05, "loss": 0.0475, "step": 35850 }, { "grad_norm": 0.43907496333122253, "learning_rate": 6.255657082902609e-05, "loss": 0.0388, "step": 35860 }, { "grad_norm": 0.39755263924598694, "learning_rate": 6.253656374341325e-05, "loss": 0.0479, "step": 35870 }, { "grad_norm": 0.3939215838909149, "learning_rate": 6.251655451564457e-05, "loss": 0.0435, "step": 35880 }, { "grad_norm": 0.5181946754455566, "learning_rate": 6.249654314913902e-05, "loss": 0.0647, "step": 35890 }, { "grad_norm": 0.5771373510360718, "learning_rate": 6.247652964731604e-05, "loss": 0.036, "step": 35900 }, { "grad_norm": 0.5719823837280273, "learning_rate": 6.245651401359537e-05, "loss": 0.0452, "step": 35910 }, { "grad_norm": 0.6695706844329834, "learning_rate": 6.243649625139715e-05, "loss": 0.0476, "step": 35920 }, { "grad_norm": 0.7255530953407288, "learning_rate": 6.241647636414185e-05, "loss": 0.0404, "step": 35930 }, { "grad_norm": 0.4588411748409271, "learning_rate": 6.239645435525034e-05, "loss": 0.0344, "step": 35940 }, { "grad_norm": 0.5573229789733887, "learning_rate": 6.237643022814381e-05, "loss": 0.0629, "step": 35950 }, { "grad_norm": 0.5467061400413513, "learning_rate": 6.235640398624386e-05, "loss": 0.0524, "step": 35960 }, { "grad_norm": 0.6708900928497314, "learning_rate": 6.233637563297243e-05, "loss": 0.0445, "step": 35970 }, { "grad_norm": 0.7267605662345886, "learning_rate": 6.23163451717518e-05, "loss": 0.0381, "step": 35980 }, { "grad_norm": 0.45442521572113037, "learning_rate": 6.229631260600463e-05, "loss": 0.0602, "step": 35990 }, { "grad_norm": 0.40113815665245056, "learning_rate": 6.227627793915392e-05, "loss": 0.0362, "step": 36000 }, { "grad_norm": 0.3561828136444092, "learning_rate": 6.225624117462309e-05, "loss": 0.0437, "step": 36010 }, { "grad_norm": 0.593137800693512, "learning_rate": 6.223620231583586e-05, "loss": 0.0716, "step": 36020 }, { "grad_norm": 0.6641594171524048, "learning_rate": 6.221616136621629e-05, "loss": 0.0531, "step": 36030 }, { "grad_norm": 0.2919638752937317, "learning_rate": 6.219611832918887e-05, "loss": 0.0386, "step": 36040 }, { "grad_norm": 0.3941786587238312, "learning_rate": 6.217607320817838e-05, "loss": 0.0592, "step": 36050 }, { "grad_norm": 0.5102508068084717, "learning_rate": 6.215602600661001e-05, "loss": 0.0422, "step": 36060 }, { "grad_norm": 0.44162869453430176, "learning_rate": 6.213597672790925e-05, "loss": 0.0549, "step": 36070 }, { "grad_norm": 0.7122750878334045, "learning_rate": 6.2115925375502e-05, "loss": 0.0523, "step": 36080 }, { "grad_norm": 0.3854220509529114, "learning_rate": 6.209587195281447e-05, "loss": 0.0471, "step": 36090 }, { "grad_norm": 0.873443067073822, "learning_rate": 6.207581646327324e-05, "loss": 0.0468, "step": 36100 }, { "grad_norm": 0.3446284234523773, "learning_rate": 6.205575891030526e-05, "loss": 0.0365, "step": 36110 }, { "grad_norm": 0.9116955995559692, "learning_rate": 6.203569929733781e-05, "loss": 0.0488, "step": 36120 }, { "grad_norm": 0.45934009552001953, "learning_rate": 6.201563762779852e-05, "loss": 0.0433, "step": 36130 }, { "grad_norm": 0.585118293762207, "learning_rate": 6.199557390511538e-05, "loss": 0.035, "step": 36140 }, { "grad_norm": 0.5796310305595398, "learning_rate": 6.197550813271675e-05, "loss": 0.0457, "step": 36150 }, { "grad_norm": 0.49207988381385803, "learning_rate": 6.195544031403131e-05, "loss": 0.0359, "step": 36160 }, { "grad_norm": 0.41359105706214905, "learning_rate": 6.19353704524881e-05, "loss": 0.0393, "step": 36170 }, { "grad_norm": 0.5466599464416504, "learning_rate": 6.191529855151652e-05, "loss": 0.0394, "step": 36180 }, { "grad_norm": 0.6901848912239075, "learning_rate": 6.189522461454629e-05, "loss": 0.0511, "step": 36190 }, { "grad_norm": 0.5164117217063904, "learning_rate": 6.187514864500752e-05, "loss": 0.049, "step": 36200 }, { "grad_norm": 0.40674278140068054, "learning_rate": 6.185507064633062e-05, "loss": 0.0397, "step": 36210 }, { "grad_norm": 0.7100276947021484, "learning_rate": 6.18349906219464e-05, "loss": 0.0401, "step": 36220 }, { "grad_norm": 0.5165338516235352, "learning_rate": 6.181490857528596e-05, "loss": 0.051, "step": 36230 }, { "grad_norm": 0.646584689617157, "learning_rate": 6.179482450978077e-05, "loss": 0.0391, "step": 36240 }, { "grad_norm": 0.6013262271881104, "learning_rate": 6.177473842886269e-05, "loss": 0.0447, "step": 36250 }, { "grad_norm": 0.6463924646377563, "learning_rate": 6.175465033596382e-05, "loss": 0.039, "step": 36260 }, { "grad_norm": 0.46445026993751526, "learning_rate": 6.173456023451671e-05, "loss": 0.0352, "step": 36270 }, { "grad_norm": 0.4439867436885834, "learning_rate": 6.171446812795422e-05, "loss": 0.0386, "step": 36280 }, { "grad_norm": 0.3987475037574768, "learning_rate": 6.169437401970949e-05, "loss": 0.0435, "step": 36290 }, { "grad_norm": 0.47871407866477966, "learning_rate": 6.16742779132161e-05, "loss": 0.0427, "step": 36300 }, { "grad_norm": 0.6753706932067871, "learning_rate": 6.165417981190789e-05, "loss": 0.0472, "step": 36310 }, { "grad_norm": 0.3900323808193207, "learning_rate": 6.16340797192191e-05, "loss": 0.0403, "step": 36320 }, { "grad_norm": 0.5657809376716614, "learning_rate": 6.161397763858427e-05, "loss": 0.0516, "step": 36330 }, { "grad_norm": 0.5723017454147339, "learning_rate": 6.159387357343834e-05, "loss": 0.036, "step": 36340 }, { "grad_norm": 0.3799423277378082, "learning_rate": 6.157376752721648e-05, "loss": 0.0399, "step": 36350 }, { "grad_norm": 0.6083239316940308, "learning_rate": 6.155365950335428e-05, "loss": 0.0438, "step": 36360 }, { "grad_norm": 0.5774778723716736, "learning_rate": 6.153354950528768e-05, "loss": 0.0357, "step": 36370 }, { "grad_norm": 0.4509536623954773, "learning_rate": 6.151343753645293e-05, "loss": 0.0432, "step": 36380 }, { "grad_norm": 0.2999398112297058, "learning_rate": 6.149332360028657e-05, "loss": 0.0342, "step": 36390 }, { "grad_norm": 0.30307114124298096, "learning_rate": 6.147320770022555e-05, "loss": 0.0362, "step": 36400 }, { "grad_norm": 0.52382493019104, "learning_rate": 6.145308983970715e-05, "loss": 0.0468, "step": 36410 }, { "grad_norm": 0.49172505736351013, "learning_rate": 6.143297002216892e-05, "loss": 0.036, "step": 36420 }, { "grad_norm": 0.5626898407936096, "learning_rate": 6.141284825104882e-05, "loss": 0.0309, "step": 36430 }, { "grad_norm": 0.4868556559085846, "learning_rate": 6.13927245297851e-05, "loss": 0.039, "step": 36440 }, { "grad_norm": 0.3674430251121521, "learning_rate": 6.137259886181633e-05, "loss": 0.0344, "step": 36450 }, { "grad_norm": 0.6540480852127075, "learning_rate": 6.135247125058145e-05, "loss": 0.0369, "step": 36460 }, { "grad_norm": 0.6084543466567993, "learning_rate": 6.133234169951974e-05, "loss": 0.0474, "step": 36470 }, { "grad_norm": 0.6426849365234375, "learning_rate": 6.131221021207078e-05, "loss": 0.0374, "step": 36480 }, { "grad_norm": 0.590886652469635, "learning_rate": 6.129207679167448e-05, "loss": 0.0487, "step": 36490 }, { "grad_norm": 0.7510046362876892, "learning_rate": 6.127194144177109e-05, "loss": 0.0467, "step": 36500 }, { "grad_norm": 0.5428792238235474, "learning_rate": 6.125180416580118e-05, "loss": 0.0438, "step": 36510 }, { "grad_norm": 0.4579862952232361, "learning_rate": 6.123166496720571e-05, "loss": 0.037, "step": 36520 }, { "grad_norm": 0.5465682148933411, "learning_rate": 6.121152384942588e-05, "loss": 0.048, "step": 36530 }, { "grad_norm": 0.578715443611145, "learning_rate": 6.119138081590324e-05, "loss": 0.0413, "step": 36540 }, { "grad_norm": 0.47955331206321716, "learning_rate": 6.117123587007971e-05, "loss": 0.0448, "step": 36550 }, { "grad_norm": 0.6062302589416504, "learning_rate": 6.11510890153975e-05, "loss": 0.034, "step": 36560 }, { "grad_norm": 0.43397876620292664, "learning_rate": 6.113094025529916e-05, "loss": 0.0479, "step": 36570 }, { "grad_norm": 0.47402116656303406, "learning_rate": 6.111078959322757e-05, "loss": 0.0515, "step": 36580 }, { "grad_norm": 0.8267166018486023, "learning_rate": 6.109063703262592e-05, "loss": 0.0386, "step": 36590 }, { "grad_norm": 0.5363077521324158, "learning_rate": 6.107048257693772e-05, "loss": 0.0382, "step": 36600 }, { "grad_norm": 0.448027640581131, "learning_rate": 6.105032622960683e-05, "loss": 0.0319, "step": 36610 }, { "grad_norm": 0.5829205513000488, "learning_rate": 6.103016799407743e-05, "loss": 0.056, "step": 36620 }, { "grad_norm": 0.5872648358345032, "learning_rate": 6.1010007873793984e-05, "loss": 0.0416, "step": 36630 }, { "grad_norm": 0.5362212061882019, "learning_rate": 6.098984587220131e-05, "loss": 0.055, "step": 36640 }, { "grad_norm": 0.35571253299713135, "learning_rate": 6.096968199274456e-05, "loss": 0.0442, "step": 36650 }, { "grad_norm": 0.5455938577651978, "learning_rate": 6.0949516238869166e-05, "loss": 0.0355, "step": 36660 }, { "grad_norm": 0.37813660502433777, "learning_rate": 6.092934861402092e-05, "loss": 0.0368, "step": 36670 }, { "grad_norm": 0.38766762614250183, "learning_rate": 6.0909179121645924e-05, "loss": 0.0351, "step": 36680 }, { "grad_norm": 0.5206385254859924, "learning_rate": 6.0889007765190576e-05, "loss": 0.0444, "step": 36690 }, { "grad_norm": 0.5806698203086853, "learning_rate": 6.086883454810162e-05, "loss": 0.0351, "step": 36700 }, { "grad_norm": 0.2799379229545593, "learning_rate": 6.0848659473826084e-05, "loss": 0.0491, "step": 36710 }, { "grad_norm": 0.5267837643623352, "learning_rate": 6.082848254581138e-05, "loss": 0.0359, "step": 36720 }, { "grad_norm": 0.5850389003753662, "learning_rate": 6.080830376750517e-05, "loss": 0.0402, "step": 36730 }, { "grad_norm": 0.49902573227882385, "learning_rate": 6.0788123142355445e-05, "loss": 0.0401, "step": 36740 }, { "grad_norm": 0.42249494791030884, "learning_rate": 6.076794067381052e-05, "loss": 0.0474, "step": 36750 }, { "grad_norm": 0.5641534924507141, "learning_rate": 6.074775636531905e-05, "loss": 0.0417, "step": 36760 }, { "grad_norm": 0.7904433012008667, "learning_rate": 6.072757022032997e-05, "loss": 0.0466, "step": 36770 }, { "grad_norm": 0.40064626932144165, "learning_rate": 6.070738224229253e-05, "loss": 0.0366, "step": 36780 }, { "grad_norm": 0.6123993992805481, "learning_rate": 6.0687192434656314e-05, "loss": 0.0395, "step": 36790 }, { "grad_norm": 0.5588334202766418, "learning_rate": 6.066700080087121e-05, "loss": 0.0632, "step": 36800 }, { "grad_norm": 0.5115469694137573, "learning_rate": 6.0646807344387424e-05, "loss": 0.0383, "step": 36810 }, { "grad_norm": 0.544879138469696, "learning_rate": 6.062661206865543e-05, "loss": 0.043, "step": 36820 }, { "grad_norm": 0.44891175627708435, "learning_rate": 6.06064149771261e-05, "loss": 0.0431, "step": 36830 }, { "grad_norm": 0.4341658055782318, "learning_rate": 6.058621607325051e-05, "loss": 0.0458, "step": 36840 }, { "grad_norm": 0.3929447829723358, "learning_rate": 6.056601536048014e-05, "loss": 0.0371, "step": 36850 }, { "grad_norm": 0.5701108574867249, "learning_rate": 6.0545812842266725e-05, "loss": 0.0325, "step": 36860 }, { "grad_norm": 0.6095906496047974, "learning_rate": 6.052560852206232e-05, "loss": 0.0443, "step": 36870 }, { "grad_norm": 0.6087625622749329, "learning_rate": 6.05054024033193e-05, "loss": 0.0266, "step": 36880 }, { "grad_norm": 0.66153883934021, "learning_rate": 6.048519448949032e-05, "loss": 0.0409, "step": 36890 }, { "grad_norm": 0.505803644657135, "learning_rate": 6.046498478402839e-05, "loss": 0.0606, "step": 36900 }, { "grad_norm": 0.3789033591747284, "learning_rate": 6.044477329038677e-05, "loss": 0.0332, "step": 36910 }, { "grad_norm": 0.5113815069198608, "learning_rate": 6.042456001201906e-05, "loss": 0.0447, "step": 36920 }, { "grad_norm": 0.38188984990119934, "learning_rate": 6.040434495237917e-05, "loss": 0.0459, "step": 36930 }, { "grad_norm": 0.4176854193210602, "learning_rate": 6.0384128114921256e-05, "loss": 0.0568, "step": 36940 }, { "grad_norm": 0.7063339352607727, "learning_rate": 6.036390950309987e-05, "loss": 0.0451, "step": 36950 }, { "grad_norm": 0.5220387578010559, "learning_rate": 6.0343689120369805e-05, "loss": 0.0359, "step": 36960 }, { "grad_norm": 0.696071445941925, "learning_rate": 6.032346697018616e-05, "loss": 0.0342, "step": 36970 }, { "grad_norm": 0.6365885734558105, "learning_rate": 6.0303243056004375e-05, "loss": 0.0354, "step": 36980 }, { "grad_norm": 0.30126187205314636, "learning_rate": 6.0283017381280136e-05, "loss": 0.0558, "step": 36990 }, { "grad_norm": 0.47410818934440613, "learning_rate": 6.026278994946948e-05, "loss": 0.0497, "step": 37000 }, { "grad_norm": 0.5718415379524231, "learning_rate": 6.02425607640287e-05, "loss": 0.0495, "step": 37010 }, { "grad_norm": 0.5726228952407837, "learning_rate": 6.022232982841441e-05, "loss": 0.0436, "step": 37020 }, { "grad_norm": 0.4686565399169922, "learning_rate": 6.020209714608355e-05, "loss": 0.0372, "step": 37030 }, { "grad_norm": 0.5923793911933899, "learning_rate": 6.018186272049331e-05, "loss": 0.0351, "step": 37040 }, { "grad_norm": 0.32620948553085327, "learning_rate": 6.01616265551012e-05, "loss": 0.041, "step": 37050 }, { "grad_norm": 0.6249624490737915, "learning_rate": 6.014138865336503e-05, "loss": 0.0439, "step": 37060 }, { "grad_norm": 0.47158393263816833, "learning_rate": 6.0121149018742905e-05, "loss": 0.0351, "step": 37070 }, { "grad_norm": 0.3716740608215332, "learning_rate": 6.010090765469325e-05, "loss": 0.0344, "step": 37080 }, { "grad_norm": 0.24213756620883942, "learning_rate": 6.008066456467473e-05, "loss": 0.0301, "step": 37090 }, { "grad_norm": 0.4958357512950897, "learning_rate": 6.0060419752146335e-05, "loss": 0.0358, "step": 37100 }, { "grad_norm": 0.5911210775375366, "learning_rate": 6.0040173220567353e-05, "loss": 0.0495, "step": 37110 }, { "grad_norm": 0.42600932717323303, "learning_rate": 6.001992497339737e-05, "loss": 0.0467, "step": 37120 }, { "grad_norm": 0.4671316146850586, "learning_rate": 5.999967501409626e-05, "loss": 0.0334, "step": 37130 }, { "grad_norm": 0.5569872260093689, "learning_rate": 5.997942334612418e-05, "loss": 0.042, "step": 37140 }, { "grad_norm": 0.43538570404052734, "learning_rate": 5.995916997294158e-05, "loss": 0.0296, "step": 37150 }, { "grad_norm": 0.5059002041816711, "learning_rate": 5.9938914898009214e-05, "loss": 0.036, "step": 37160 }, { "grad_norm": 0.3190186321735382, "learning_rate": 5.991865812478813e-05, "loss": 0.0382, "step": 37170 }, { "grad_norm": 0.4427020251750946, "learning_rate": 5.989839965673964e-05, "loss": 0.0442, "step": 37180 }, { "grad_norm": 0.7765408158302307, "learning_rate": 5.987813949732539e-05, "loss": 0.0505, "step": 37190 }, { "grad_norm": 0.48798951506614685, "learning_rate": 5.9857877650007255e-05, "loss": 0.0443, "step": 37200 }, { "grad_norm": 0.49539852142333984, "learning_rate": 5.983761411824744e-05, "loss": 0.0334, "step": 37210 }, { "grad_norm": 0.5440638065338135, "learning_rate": 5.981734890550844e-05, "loss": 0.0318, "step": 37220 }, { "grad_norm": 0.44508180022239685, "learning_rate": 5.979708201525301e-05, "loss": 0.0393, "step": 37230 }, { "grad_norm": 0.31015831232070923, "learning_rate": 5.977681345094422e-05, "loss": 0.0692, "step": 37240 }, { "grad_norm": 0.6232216358184814, "learning_rate": 5.97565432160454e-05, "loss": 0.0741, "step": 37250 }, { "grad_norm": 0.5219430923461914, "learning_rate": 5.9736271314020186e-05, "loss": 0.0637, "step": 37260 }, { "grad_norm": 0.45538222789764404, "learning_rate": 5.971599774833251e-05, "loss": 0.0434, "step": 37270 }, { "grad_norm": 0.4065968692302704, "learning_rate": 5.9695722522446525e-05, "loss": 0.0437, "step": 37280 }, { "grad_norm": 0.586835503578186, "learning_rate": 5.9675445639826765e-05, "loss": 0.0422, "step": 37290 }, { "grad_norm": 0.349581241607666, "learning_rate": 5.965516710393796e-05, "loss": 0.0319, "step": 37300 }, { "grad_norm": 0.4292939603328705, "learning_rate": 5.963488691824516e-05, "loss": 0.0363, "step": 37310 }, { "grad_norm": 0.36810240149497986, "learning_rate": 5.96146050862137e-05, "loss": 0.0463, "step": 37320 }, { "grad_norm": 0.46279770135879517, "learning_rate": 5.959432161130919e-05, "loss": 0.0363, "step": 37330 }, { "grad_norm": 0.4365394413471222, "learning_rate": 5.9574036496997545e-05, "loss": 0.043, "step": 37340 }, { "grad_norm": 0.3822559118270874, "learning_rate": 5.955374974674488e-05, "loss": 0.0371, "step": 37350 }, { "grad_norm": 0.3990321457386017, "learning_rate": 5.9533461364017696e-05, "loss": 0.0467, "step": 37360 }, { "grad_norm": 0.4642932415008545, "learning_rate": 5.9513171352282716e-05, "loss": 0.0414, "step": 37370 }, { "grad_norm": 0.4652380049228668, "learning_rate": 5.949287971500692e-05, "loss": 0.0304, "step": 37380 }, { "grad_norm": 0.45132964849472046, "learning_rate": 5.947258645565762e-05, "loss": 0.0311, "step": 37390 }, { "grad_norm": 0.5928606986999512, "learning_rate": 5.945229157770237e-05, "loss": 0.0355, "step": 37400 }, { "grad_norm": 0.5834494829177856, "learning_rate": 5.9431995084609006e-05, "loss": 0.0395, "step": 37410 }, { "grad_norm": 0.5616679787635803, "learning_rate": 5.941169697984564e-05, "loss": 0.032, "step": 37420 }, { "grad_norm": 0.5652944445610046, "learning_rate": 5.9391397266880675e-05, "loss": 0.0423, "step": 37430 }, { "grad_norm": 0.31877395510673523, "learning_rate": 5.937109594918279e-05, "loss": 0.0312, "step": 37440 }, { "grad_norm": 0.5382317304611206, "learning_rate": 5.9350793030220884e-05, "loss": 0.0682, "step": 37450 }, { "grad_norm": 0.3745928406715393, "learning_rate": 5.933048851346421e-05, "loss": 0.0566, "step": 37460 }, { "grad_norm": 0.4849753975868225, "learning_rate": 5.931018240238224e-05, "loss": 0.0427, "step": 37470 }, { "grad_norm": 0.6333637237548828, "learning_rate": 5.928987470044471e-05, "loss": 0.0403, "step": 37480 }, { "grad_norm": 0.4733121395111084, "learning_rate": 5.9269565411121695e-05, "loss": 0.0357, "step": 37490 }, { "grad_norm": 0.5967952609062195, "learning_rate": 5.924925453788347e-05, "loss": 0.0442, "step": 37500 }, { "grad_norm": 0.4948769211769104, "learning_rate": 5.92289420842006e-05, "loss": 0.0391, "step": 37510 }, { "grad_norm": 0.8927011489868164, "learning_rate": 5.9208628053543945e-05, "loss": 0.0489, "step": 37520 }, { "grad_norm": 0.6510568857192993, "learning_rate": 5.918831244938462e-05, "loss": 0.0587, "step": 37530 }, { "grad_norm": 0.37866702675819397, "learning_rate": 5.9167995275194e-05, "loss": 0.0606, "step": 37540 }, { "grad_norm": 0.5833227634429932, "learning_rate": 5.914767653444373e-05, "loss": 0.041, "step": 37550 }, { "grad_norm": 0.41837945580482483, "learning_rate": 5.912735623060572e-05, "loss": 0.0599, "step": 37560 }, { "grad_norm": 0.4015866816043854, "learning_rate": 5.910703436715217e-05, "loss": 0.0362, "step": 37570 }, { "grad_norm": 0.257644385099411, "learning_rate": 5.908671094755552e-05, "loss": 0.0495, "step": 37580 }, { "grad_norm": 0.5257370471954346, "learning_rate": 5.906638597528851e-05, "loss": 0.0448, "step": 37590 }, { "grad_norm": 0.48335710167884827, "learning_rate": 5.9046059453824076e-05, "loss": 0.0347, "step": 37600 }, { "grad_norm": 0.4192202389240265, "learning_rate": 5.9025731386635505e-05, "loss": 0.0467, "step": 37610 }, { "grad_norm": 0.6148442625999451, "learning_rate": 5.900540177719629e-05, "loss": 0.0406, "step": 37620 }, { "grad_norm": 0.6365336775779724, "learning_rate": 5.898507062898021e-05, "loss": 0.0394, "step": 37630 }, { "grad_norm": 0.4815109372138977, "learning_rate": 5.8964737945461316e-05, "loss": 0.0432, "step": 37640 }, { "grad_norm": 0.6889711618423462, "learning_rate": 5.8944403730113885e-05, "loss": 0.0497, "step": 37650 }, { "grad_norm": 0.414187490940094, "learning_rate": 5.892406798641248e-05, "loss": 0.0315, "step": 37660 }, { "grad_norm": 0.4323122799396515, "learning_rate": 5.890373071783193e-05, "loss": 0.0396, "step": 37670 }, { "grad_norm": 0.5963699817657471, "learning_rate": 5.888339192784732e-05, "loss": 0.0401, "step": 37680 }, { "grad_norm": 0.3983745872974396, "learning_rate": 5.8863051619934003e-05, "loss": 0.0362, "step": 37690 }, { "grad_norm": 0.4784690737724304, "learning_rate": 5.8842709797567554e-05, "loss": 0.0322, "step": 37700 }, { "grad_norm": 0.3920377492904663, "learning_rate": 5.8822366464223855e-05, "loss": 0.0504, "step": 37710 }, { "grad_norm": 0.5909147262573242, "learning_rate": 5.880202162337901e-05, "loss": 0.0458, "step": 37720 }, { "grad_norm": 0.3980712294578552, "learning_rate": 5.8781675278509405e-05, "loss": 0.0381, "step": 37730 }, { "grad_norm": 0.43754127621650696, "learning_rate": 5.8761327433091696e-05, "loss": 0.0492, "step": 37740 }, { "grad_norm": 0.499524861574173, "learning_rate": 5.874097809060275e-05, "loss": 0.0413, "step": 37750 }, { "grad_norm": 0.37147191166877747, "learning_rate": 5.87206272545197e-05, "loss": 0.0382, "step": 37760 }, { "grad_norm": 0.34958234429359436, "learning_rate": 5.8700274928319955e-05, "loss": 0.0348, "step": 37770 }, { "grad_norm": 0.4181925058364868, "learning_rate": 5.867992111548118e-05, "loss": 0.0407, "step": 37780 }, { "grad_norm": 0.3088415563106537, "learning_rate": 5.865956581948131e-05, "loss": 0.0416, "step": 37790 }, { "grad_norm": 0.6805109977722168, "learning_rate": 5.863920904379845e-05, "loss": 0.0583, "step": 37800 }, { "grad_norm": 0.4722338318824768, "learning_rate": 5.8618850791911064e-05, "loss": 0.0286, "step": 37810 }, { "grad_norm": 0.4109574854373932, "learning_rate": 5.859849106729779e-05, "loss": 0.0379, "step": 37820 }, { "grad_norm": 0.4618489146232605, "learning_rate": 5.857812987343758e-05, "loss": 0.029, "step": 37830 }, { "grad_norm": 0.4088805317878723, "learning_rate": 5.855776721380957e-05, "loss": 0.0421, "step": 37840 }, { "grad_norm": 0.3996375799179077, "learning_rate": 5.8537403091893217e-05, "loss": 0.0374, "step": 37850 }, { "grad_norm": 0.5554091334342957, "learning_rate": 5.851703751116816e-05, "loss": 0.0442, "step": 37860 }, { "grad_norm": 0.5145927667617798, "learning_rate": 5.8496670475114336e-05, "loss": 0.0298, "step": 37870 }, { "grad_norm": 0.4359380304813385, "learning_rate": 5.84763019872119e-05, "loss": 0.036, "step": 37880 }, { "grad_norm": 0.5442459583282471, "learning_rate": 5.845593205094131e-05, "loss": 0.0388, "step": 37890 }, { "grad_norm": 0.46593594551086426, "learning_rate": 5.843556066978318e-05, "loss": 0.0403, "step": 37900 }, { "grad_norm": 0.41958028078079224, "learning_rate": 5.8415187847218455e-05, "loss": 0.0439, "step": 37910 }, { "grad_norm": 0.5459396839141846, "learning_rate": 5.839481358672827e-05, "loss": 0.0404, "step": 37920 }, { "grad_norm": 0.6222456693649292, "learning_rate": 5.837443789179407e-05, "loss": 0.0444, "step": 37930 }, { "grad_norm": 0.4108741581439972, "learning_rate": 5.8354060765897445e-05, "loss": 0.0402, "step": 37940 }, { "grad_norm": 0.4173455238342285, "learning_rate": 5.8333682212520334e-05, "loss": 0.0432, "step": 37950 }, { "grad_norm": 0.4157247245311737, "learning_rate": 5.831330223514486e-05, "loss": 0.0373, "step": 37960 }, { "grad_norm": 0.5234596133232117, "learning_rate": 5.8292920837253396e-05, "loss": 0.0371, "step": 37970 }, { "grad_norm": 0.6065647006034851, "learning_rate": 5.827253802232857e-05, "loss": 0.0421, "step": 37980 }, { "grad_norm": 0.5376636385917664, "learning_rate": 5.825215379385327e-05, "loss": 0.0309, "step": 37990 }, { "grad_norm": 0.5602231621742249, "learning_rate": 5.823176815531057e-05, "loss": 0.0543, "step": 38000 }, { "grad_norm": 0.8680509924888611, "learning_rate": 5.8211381110183826e-05, "loss": 0.0458, "step": 38010 }, { "grad_norm": 0.5081421732902527, "learning_rate": 5.8190992661956645e-05, "loss": 0.049, "step": 38020 }, { "grad_norm": 0.29946717619895935, "learning_rate": 5.817060281411284e-05, "loss": 0.0416, "step": 38030 }, { "grad_norm": 0.37616363167762756, "learning_rate": 5.815021157013647e-05, "loss": 0.0336, "step": 38040 }, { "grad_norm": 0.39559879899024963, "learning_rate": 5.8129818933511856e-05, "loss": 0.0353, "step": 38050 }, { "grad_norm": 0.6068143844604492, "learning_rate": 5.8109424907723544e-05, "loss": 0.0354, "step": 38060 }, { "grad_norm": 0.4777062237262726, "learning_rate": 5.80890294962563e-05, "loss": 0.0497, "step": 38070 }, { "grad_norm": 0.41504016518592834, "learning_rate": 5.806863270259515e-05, "loss": 0.0371, "step": 38080 }, { "grad_norm": 0.46713659167289734, "learning_rate": 5.804823453022536e-05, "loss": 0.0437, "step": 38090 }, { "grad_norm": 0.5475659966468811, "learning_rate": 5.80278349826324e-05, "loss": 0.0472, "step": 38100 }, { "grad_norm": 0.35638168454170227, "learning_rate": 5.8007434063302014e-05, "loss": 0.0384, "step": 38110 }, { "grad_norm": 0.4601535499095917, "learning_rate": 5.7987031775720136e-05, "loss": 0.0441, "step": 38120 }, { "grad_norm": 0.5225940942764282, "learning_rate": 5.7966628123372976e-05, "loss": 0.0436, "step": 38130 }, { "grad_norm": 0.4760628938674927, "learning_rate": 5.7946223109746956e-05, "loss": 0.0433, "step": 38140 }, { "grad_norm": 0.29088085889816284, "learning_rate": 5.7925816738328754e-05, "loss": 0.0347, "step": 38150 }, { "grad_norm": 0.4336841106414795, "learning_rate": 5.790540901260521e-05, "loss": 0.0476, "step": 38160 }, { "grad_norm": 0.3652149438858032, "learning_rate": 5.788499993606351e-05, "loss": 0.0491, "step": 38170 }, { "grad_norm": 0.4711284935474396, "learning_rate": 5.786458951219096e-05, "loss": 0.0409, "step": 38180 }, { "grad_norm": 0.48297762870788574, "learning_rate": 5.784417774447517e-05, "loss": 0.0677, "step": 38190 }, { "grad_norm": 0.6058173775672913, "learning_rate": 5.782376463640393e-05, "loss": 0.042, "step": 38200 }, { "grad_norm": 0.5471176505088806, "learning_rate": 5.780335019146531e-05, "loss": 0.0378, "step": 38210 }, { "grad_norm": 0.42554524540901184, "learning_rate": 5.778293441314755e-05, "loss": 0.044, "step": 38220 }, { "grad_norm": 0.4157654941082001, "learning_rate": 5.776251730493917e-05, "loss": 0.0355, "step": 38230 }, { "grad_norm": 0.4665908217430115, "learning_rate": 5.774209887032887e-05, "loss": 0.0341, "step": 38240 }, { "grad_norm": 0.4195568263530731, "learning_rate": 5.772167911280565e-05, "loss": 0.0412, "step": 38250 }, { "grad_norm": 0.4352262616157532, "learning_rate": 5.770125803585864e-05, "loss": 0.056, "step": 38260 }, { "grad_norm": 0.45459720492362976, "learning_rate": 5.768083564297726e-05, "loss": 0.038, "step": 38270 }, { "grad_norm": 0.4869898855686188, "learning_rate": 5.766041193765114e-05, "loss": 0.0306, "step": 38280 }, { "grad_norm": 0.47529706358909607, "learning_rate": 5.763998692337015e-05, "loss": 0.0293, "step": 38290 }, { "grad_norm": 0.5126249194145203, "learning_rate": 5.761956060362433e-05, "loss": 0.0329, "step": 38300 }, { "grad_norm": 0.590864360332489, "learning_rate": 5.7599132981904e-05, "loss": 0.0428, "step": 38310 }, { "grad_norm": 0.7564600110054016, "learning_rate": 5.75787040616997e-05, "loss": 0.0406, "step": 38320 }, { "grad_norm": 0.4288097023963928, "learning_rate": 5.755827384650212e-05, "loss": 0.0339, "step": 38330 }, { "grad_norm": 0.29964935779571533, "learning_rate": 5.753784233980228e-05, "loss": 0.0389, "step": 38340 }, { "grad_norm": 0.706881582736969, "learning_rate": 5.751740954509135e-05, "loss": 0.0408, "step": 38350 }, { "grad_norm": 0.37663862109184265, "learning_rate": 5.7496975465860715e-05, "loss": 0.0339, "step": 38360 }, { "grad_norm": 0.7576355338096619, "learning_rate": 5.747654010560202e-05, "loss": 0.0713, "step": 38370 }, { "grad_norm": 0.5571715235710144, "learning_rate": 5.7456103467807097e-05, "loss": 0.0455, "step": 38380 }, { "grad_norm": 0.5564838647842407, "learning_rate": 5.7435665555968046e-05, "loss": 0.0336, "step": 38390 }, { "grad_norm": 0.7033531069755554, "learning_rate": 5.74152263735771e-05, "loss": 0.0502, "step": 38400 }, { "grad_norm": 0.5233127474784851, "learning_rate": 5.739478592412677e-05, "loss": 0.0457, "step": 38410 }, { "grad_norm": 0.6447079181671143, "learning_rate": 5.7374344211109766e-05, "loss": 0.039, "step": 38420 }, { "grad_norm": 0.7429726123809814, "learning_rate": 5.735390123801904e-05, "loss": 0.0283, "step": 38430 }, { "grad_norm": 0.40380731225013733, "learning_rate": 5.7333457008347704e-05, "loss": 0.0371, "step": 38440 }, { "grad_norm": 0.550425112247467, "learning_rate": 5.7313011525589156e-05, "loss": 0.0323, "step": 38450 }, { "grad_norm": 0.36377382278442383, "learning_rate": 5.729256479323694e-05, "loss": 0.0468, "step": 38460 }, { "grad_norm": 0.5031350255012512, "learning_rate": 5.727211681478485e-05, "loss": 0.042, "step": 38470 }, { "grad_norm": 0.3098162114620209, "learning_rate": 5.7251667593726886e-05, "loss": 0.0416, "step": 38480 }, { "grad_norm": 0.699055016040802, "learning_rate": 5.723121713355728e-05, "loss": 0.0423, "step": 38490 }, { "grad_norm": 0.4805642068386078, "learning_rate": 5.721076543777044e-05, "loss": 0.0322, "step": 38500 }, { "grad_norm": 0.4972936511039734, "learning_rate": 5.7190312509860986e-05, "loss": 0.0332, "step": 38510 }, { "grad_norm": 0.4751169681549072, "learning_rate": 5.716985835332379e-05, "loss": 0.0329, "step": 38520 }, { "grad_norm": 0.39567917585372925, "learning_rate": 5.714940297165389e-05, "loss": 0.0287, "step": 38530 }, { "grad_norm": 0.4807588756084442, "learning_rate": 5.712894636834656e-05, "loss": 0.0331, "step": 38540 }, { "grad_norm": 0.6106812357902527, "learning_rate": 5.7108488546897276e-05, "loss": 0.0405, "step": 38550 }, { "grad_norm": 0.3945685923099518, "learning_rate": 5.708802951080172e-05, "loss": 0.0422, "step": 38560 }, { "grad_norm": 0.4382278025150299, "learning_rate": 5.706756926355576e-05, "loss": 0.0528, "step": 38570 }, { "grad_norm": 0.5319151282310486, "learning_rate": 5.704710780865554e-05, "loss": 0.0503, "step": 38580 }, { "grad_norm": 0.6974313855171204, "learning_rate": 5.7026645149597325e-05, "loss": 0.0418, "step": 38590 }, { "grad_norm": 0.6088300943374634, "learning_rate": 5.700618128987764e-05, "loss": 0.052, "step": 38600 }, { "grad_norm": 0.5383033752441406, "learning_rate": 5.698571623299317e-05, "loss": 0.0411, "step": 38610 }, { "grad_norm": 0.6111306548118591, "learning_rate": 5.696524998244086e-05, "loss": 0.0451, "step": 38620 }, { "grad_norm": 0.5794963240623474, "learning_rate": 5.6944782541717836e-05, "loss": 0.0464, "step": 38630 }, { "grad_norm": 0.47020068764686584, "learning_rate": 5.69243139143214e-05, "loss": 0.0525, "step": 38640 }, { "grad_norm": 0.3766089379787445, "learning_rate": 5.6903844103749125e-05, "loss": 0.045, "step": 38650 }, { "grad_norm": 0.45176035165786743, "learning_rate": 5.688337311349869e-05, "loss": 0.0617, "step": 38660 }, { "grad_norm": 0.6235688924789429, "learning_rate": 5.6862900947068074e-05, "loss": 0.0426, "step": 38670 }, { "grad_norm": 0.7905659079551697, "learning_rate": 5.6842427607955374e-05, "loss": 0.0749, "step": 38680 }, { "grad_norm": 0.49107012152671814, "learning_rate": 5.682195309965893e-05, "loss": 0.0479, "step": 38690 }, { "grad_norm": 0.38324445486068726, "learning_rate": 5.6801477425677294e-05, "loss": 0.0467, "step": 38700 }, { "grad_norm": 0.4988499879837036, "learning_rate": 5.678100058950917e-05, "loss": 0.0281, "step": 38710 }, { "grad_norm": 0.5839582681655884, "learning_rate": 5.676052259465352e-05, "loss": 0.0382, "step": 38720 }, { "grad_norm": 0.47472691535949707, "learning_rate": 5.674004344460945e-05, "loss": 0.0417, "step": 38730 }, { "grad_norm": 0.3569970428943634, "learning_rate": 5.6719563142876295e-05, "loss": 0.0409, "step": 38740 }, { "grad_norm": 0.45832139253616333, "learning_rate": 5.669908169295359e-05, "loss": 0.029, "step": 38750 }, { "grad_norm": 0.6222115159034729, "learning_rate": 5.667859909834105e-05, "loss": 0.0615, "step": 38760 }, { "grad_norm": 0.550894558429718, "learning_rate": 5.6658115362538565e-05, "loss": 0.0357, "step": 38770 }, { "grad_norm": 0.2960491478443146, "learning_rate": 5.663763048904628e-05, "loss": 0.0392, "step": 38780 }, { "grad_norm": 0.5009835362434387, "learning_rate": 5.661714448136447e-05, "loss": 0.0322, "step": 38790 }, { "grad_norm": 0.3809974193572998, "learning_rate": 5.659665734299366e-05, "loss": 0.0344, "step": 38800 }, { "grad_norm": 0.49357885122299194, "learning_rate": 5.6576169077434516e-05, "loss": 0.0324, "step": 38810 }, { "grad_norm": 0.4194801449775696, "learning_rate": 5.6555679688187944e-05, "loss": 0.0374, "step": 38820 }, { "grad_norm": 0.37308818101882935, "learning_rate": 5.6535189178755e-05, "loss": 0.0353, "step": 38830 }, { "grad_norm": 0.3469720184803009, "learning_rate": 5.651469755263698e-05, "loss": 0.0323, "step": 38840 }, { "grad_norm": 1.0770370960235596, "learning_rate": 5.6494204813335316e-05, "loss": 0.0416, "step": 38850 }, { "grad_norm": 0.6731308698654175, "learning_rate": 5.647371096435168e-05, "loss": 0.0762, "step": 38860 }, { "grad_norm": 0.4500412344932556, "learning_rate": 5.645321600918788e-05, "loss": 0.0287, "step": 38870 }, { "grad_norm": 0.40989184379577637, "learning_rate": 5.643271995134597e-05, "loss": 0.0437, "step": 38880 }, { "grad_norm": 0.5586013793945312, "learning_rate": 5.641222279432814e-05, "loss": 0.038, "step": 38890 }, { "grad_norm": 0.4706622362136841, "learning_rate": 5.6391724541636834e-05, "loss": 0.0425, "step": 38900 }, { "grad_norm": 0.7213220596313477, "learning_rate": 5.6371225196774605e-05, "loss": 0.0422, "step": 38910 }, { "grad_norm": 0.4053627848625183, "learning_rate": 5.635072476324423e-05, "loss": 0.036, "step": 38920 }, { "grad_norm": 0.496115505695343, "learning_rate": 5.63302232445487e-05, "loss": 0.0368, "step": 38930 }, { "grad_norm": 0.5104112029075623, "learning_rate": 5.6309720644191144e-05, "loss": 0.0489, "step": 38940 }, { "grad_norm": 0.40040823817253113, "learning_rate": 5.628921696567491e-05, "loss": 0.035, "step": 38950 }, { "grad_norm": 0.5308087468147278, "learning_rate": 5.62687122125035e-05, "loss": 0.0332, "step": 38960 }, { "grad_norm": 0.4214068055152893, "learning_rate": 5.624820638818062e-05, "loss": 0.0469, "step": 38970 }, { "grad_norm": 0.5841588973999023, "learning_rate": 5.6227699496210164e-05, "loss": 0.055, "step": 38980 }, { "grad_norm": 0.5009936690330505, "learning_rate": 5.6207191540096195e-05, "loss": 0.0397, "step": 38990 }, { "grad_norm": 0.5224179029464722, "learning_rate": 5.618668252334296e-05, "loss": 0.0358, "step": 39000 }, { "grad_norm": 0.6118584275245667, "learning_rate": 5.616617244945488e-05, "loss": 0.0466, "step": 39010 }, { "grad_norm": 0.5988094210624695, "learning_rate": 5.614566132193656e-05, "loss": 0.0481, "step": 39020 }, { "grad_norm": 0.2983669340610504, "learning_rate": 5.612514914429282e-05, "loss": 0.0286, "step": 39030 }, { "grad_norm": 0.32392579317092896, "learning_rate": 5.610463592002863e-05, "loss": 0.0495, "step": 39040 }, { "grad_norm": 0.5127257108688354, "learning_rate": 5.608412165264909e-05, "loss": 0.036, "step": 39050 }, { "grad_norm": 0.485983282327652, "learning_rate": 5.606360634565959e-05, "loss": 0.0369, "step": 39060 }, { "grad_norm": 0.378805547952652, "learning_rate": 5.604309000256559e-05, "loss": 0.03, "step": 39070 }, { "grad_norm": 0.41595926880836487, "learning_rate": 5.6022572626872785e-05, "loss": 0.0337, "step": 39080 }, { "grad_norm": 0.47726795077323914, "learning_rate": 5.600205422208704e-05, "loss": 0.0445, "step": 39090 }, { "grad_norm": 0.4930351674556732, "learning_rate": 5.59815347917144e-05, "loss": 0.0406, "step": 39100 }, { "grad_norm": 0.47680315375328064, "learning_rate": 5.596101433926103e-05, "loss": 0.0497, "step": 39110 }, { "grad_norm": 0.3027793765068054, "learning_rate": 5.5940492868233364e-05, "loss": 0.0456, "step": 39120 }, { "grad_norm": 0.7933788299560547, "learning_rate": 5.591997038213793e-05, "loss": 0.0488, "step": 39130 }, { "grad_norm": 0.419924259185791, "learning_rate": 5.5899446884481475e-05, "loss": 0.0349, "step": 39140 }, { "grad_norm": 0.5259824395179749, "learning_rate": 5.5878922378770906e-05, "loss": 0.0448, "step": 39150 }, { "grad_norm": 0.5797459483146667, "learning_rate": 5.5858396868513285e-05, "loss": 0.0353, "step": 39160 }, { "grad_norm": 0.6999917030334473, "learning_rate": 5.583787035721586e-05, "loss": 0.0467, "step": 39170 }, { "grad_norm": 0.38591063022613525, "learning_rate": 5.581734284838606e-05, "loss": 0.0386, "step": 39180 }, { "grad_norm": 0.3507244288921356, "learning_rate": 5.579681434553147e-05, "loss": 0.0388, "step": 39190 }, { "grad_norm": 0.456969678401947, "learning_rate": 5.5776284852159854e-05, "loss": 0.0444, "step": 39200 }, { "grad_norm": 0.4261348843574524, "learning_rate": 5.575575437177913e-05, "loss": 0.0414, "step": 39210 }, { "grad_norm": 0.6286771893501282, "learning_rate": 5.573522290789742e-05, "loss": 0.0384, "step": 39220 }, { "grad_norm": 0.7089576125144958, "learning_rate": 5.571469046402298e-05, "loss": 0.046, "step": 39230 }, { "grad_norm": 0.4783311188220978, "learning_rate": 5.5694157043664205e-05, "loss": 0.04, "step": 39240 }, { "grad_norm": 0.4580082595348358, "learning_rate": 5.567362265032975e-05, "loss": 0.0497, "step": 39250 }, { "grad_norm": 0.5695842504501343, "learning_rate": 5.565308728752836e-05, "loss": 0.0333, "step": 39260 }, { "grad_norm": 0.3948034346103668, "learning_rate": 5.5632550958768956e-05, "loss": 0.0311, "step": 39270 }, { "grad_norm": 0.4128513038158417, "learning_rate": 5.5612013667560636e-05, "loss": 0.0368, "step": 39280 }, { "grad_norm": 0.4666301906108856, "learning_rate": 5.5591475417412676e-05, "loss": 0.0377, "step": 39290 }, { "grad_norm": 0.6184660196304321, "learning_rate": 5.557093621183451e-05, "loss": 0.0374, "step": 39300 }, { "grad_norm": 0.4974169433116913, "learning_rate": 5.55503960543357e-05, "loss": 0.0361, "step": 39310 }, { "grad_norm": 0.4993094503879547, "learning_rate": 5.552985494842601e-05, "loss": 0.0316, "step": 39320 }, { "grad_norm": 0.4084608554840088, "learning_rate": 5.550931289761534e-05, "loss": 0.0354, "step": 39330 }, { "grad_norm": 0.5272881388664246, "learning_rate": 5.548876990541378e-05, "loss": 0.0382, "step": 39340 }, { "grad_norm": 0.43738189339637756, "learning_rate": 5.5468225975331564e-05, "loss": 0.0381, "step": 39350 }, { "grad_norm": 0.45507365465164185, "learning_rate": 5.544768111087909e-05, "loss": 0.0369, "step": 39360 }, { "grad_norm": 0.442108690738678, "learning_rate": 5.54271353155669e-05, "loss": 0.0499, "step": 39370 }, { "grad_norm": 0.46773675084114075, "learning_rate": 5.5406588592905715e-05, "loss": 0.0371, "step": 39380 }, { "grad_norm": 0.47127559781074524, "learning_rate": 5.5386040946406416e-05, "loss": 0.0396, "step": 39390 }, { "grad_norm": 0.4181945323944092, "learning_rate": 5.536549237958004e-05, "loss": 0.0361, "step": 39400 }, { "grad_norm": 0.5175321102142334, "learning_rate": 5.5344942895937744e-05, "loss": 0.0457, "step": 39410 }, { "grad_norm": 0.40114811062812805, "learning_rate": 5.5324392498990904e-05, "loss": 0.0458, "step": 39420 }, { "grad_norm": 0.4428647458553314, "learning_rate": 5.5303841192251e-05, "loss": 0.04, "step": 39430 }, { "grad_norm": 0.6708800792694092, "learning_rate": 5.52832889792297e-05, "loss": 0.0454, "step": 39440 }, { "grad_norm": 0.6082895398139954, "learning_rate": 5.526273586343881e-05, "loss": 0.0321, "step": 39450 }, { "grad_norm": 0.4448365867137909, "learning_rate": 5.5242181848390306e-05, "loss": 0.0432, "step": 39460 }, { "grad_norm": 0.5716751217842102, "learning_rate": 5.5221626937596285e-05, "loss": 0.0399, "step": 39470 }, { "grad_norm": 0.46400848031044006, "learning_rate": 5.520107113456903e-05, "loss": 0.0329, "step": 39480 }, { "grad_norm": 0.46922752261161804, "learning_rate": 5.5180514442820974e-05, "loss": 0.0379, "step": 39490 }, { "grad_norm": 0.3022518455982208, "learning_rate": 5.515995686586469e-05, "loss": 0.0364, "step": 39500 }, { "grad_norm": 0.38760247826576233, "learning_rate": 5.5139398407212916e-05, "loss": 0.0427, "step": 39510 }, { "grad_norm": 0.5235208868980408, "learning_rate": 5.511883907037849e-05, "loss": 0.0547, "step": 39520 }, { "grad_norm": 1.0074224472045898, "learning_rate": 5.509827885887449e-05, "loss": 0.0523, "step": 39530 }, { "grad_norm": 0.47696352005004883, "learning_rate": 5.507771777621406e-05, "loss": 0.0273, "step": 39540 }, { "grad_norm": 0.5825628042221069, "learning_rate": 5.505715582591052e-05, "loss": 0.0538, "step": 39550 }, { "grad_norm": 0.5552605390548706, "learning_rate": 5.50365930114774e-05, "loss": 0.0343, "step": 39560 }, { "grad_norm": 0.447710245847702, "learning_rate": 5.5016029336428255e-05, "loss": 0.0489, "step": 39570 }, { "grad_norm": 0.4242534637451172, "learning_rate": 5.49954648042769e-05, "loss": 0.0545, "step": 39580 }, { "grad_norm": 0.3921734094619751, "learning_rate": 5.4974899418537226e-05, "loss": 0.0282, "step": 39590 }, { "grad_norm": 0.46449968218803406, "learning_rate": 5.4954333182723316e-05, "loss": 0.0444, "step": 39600 }, { "grad_norm": 0.7591481804847717, "learning_rate": 5.493376610034937e-05, "loss": 0.0416, "step": 39610 }, { "grad_norm": 0.3711221516132355, "learning_rate": 5.4913198174929735e-05, "loss": 0.0349, "step": 39620 }, { "grad_norm": 0.48273369669914246, "learning_rate": 5.489262940997889e-05, "loss": 0.0477, "step": 39630 }, { "grad_norm": 0.669812023639679, "learning_rate": 5.487205980901151e-05, "loss": 0.0369, "step": 39640 }, { "grad_norm": 0.6205584406852722, "learning_rate": 5.485148937554234e-05, "loss": 0.0398, "step": 39650 }, { "grad_norm": 0.38346853852272034, "learning_rate": 5.483091811308635e-05, "loss": 0.0301, "step": 39660 }, { "grad_norm": 0.34580349922180176, "learning_rate": 5.4810346025158564e-05, "loss": 0.0391, "step": 39670 }, { "grad_norm": 0.31956374645233154, "learning_rate": 5.478977311527421e-05, "loss": 0.0281, "step": 39680 }, { "grad_norm": 0.5228403210639954, "learning_rate": 5.476919938694863e-05, "loss": 0.0399, "step": 39690 }, { "grad_norm": 0.7169438600540161, "learning_rate": 5.474862484369733e-05, "loss": 0.0617, "step": 39700 }, { "grad_norm": 0.44685694575309753, "learning_rate": 5.472804948903589e-05, "loss": 0.0354, "step": 39710 }, { "grad_norm": 0.8371110558509827, "learning_rate": 5.470747332648013e-05, "loss": 0.0732, "step": 39720 }, { "grad_norm": 0.47564560174942017, "learning_rate": 5.468689635954591e-05, "loss": 0.033, "step": 39730 }, { "grad_norm": 0.6792428493499756, "learning_rate": 5.46663185917493e-05, "loss": 0.0436, "step": 39740 }, { "grad_norm": 0.6197739839553833, "learning_rate": 5.464574002660645e-05, "loss": 0.0506, "step": 39750 }, { "grad_norm": 0.4962458312511444, "learning_rate": 5.4625160667633724e-05, "loss": 0.0432, "step": 39760 }, { "grad_norm": 0.7617383003234863, "learning_rate": 5.4604580518347505e-05, "loss": 0.0466, "step": 39770 }, { "grad_norm": 0.5233536958694458, "learning_rate": 5.458399958226442e-05, "loss": 0.042, "step": 39780 }, { "grad_norm": 0.7076195478439331, "learning_rate": 5.456341786290119e-05, "loss": 0.0453, "step": 39790 }, { "grad_norm": 0.6534068584442139, "learning_rate": 5.454283536377465e-05, "loss": 0.0765, "step": 39800 }, { "grad_norm": 0.5727410316467285, "learning_rate": 5.452225208840179e-05, "loss": 0.0424, "step": 39810 }, { "grad_norm": 0.5812073945999146, "learning_rate": 5.450166804029975e-05, "loss": 0.0476, "step": 39820 }, { "grad_norm": 0.37716299295425415, "learning_rate": 5.448108322298574e-05, "loss": 0.0408, "step": 39830 }, { "grad_norm": 0.5824719071388245, "learning_rate": 5.446049763997717e-05, "loss": 0.035, "step": 39840 }, { "grad_norm": 0.6880974769592285, "learning_rate": 5.4439911294791546e-05, "loss": 0.0439, "step": 39850 }, { "grad_norm": 0.4839688837528229, "learning_rate": 5.441932419094652e-05, "loss": 0.0402, "step": 39860 }, { "grad_norm": 0.580384373664856, "learning_rate": 5.439873633195985e-05, "loss": 0.0349, "step": 39870 }, { "grad_norm": 0.7261250615119934, "learning_rate": 5.437814772134947e-05, "loss": 0.0458, "step": 39880 }, { "grad_norm": 0.4319758713245392, "learning_rate": 5.4357558362633366e-05, "loss": 0.0269, "step": 39890 }, { "grad_norm": 0.5072149038314819, "learning_rate": 5.4336968259329726e-05, "loss": 0.0495, "step": 39900 }, { "grad_norm": 0.4981139898300171, "learning_rate": 5.431637741495681e-05, "loss": 0.0372, "step": 39910 }, { "grad_norm": 0.5512427091598511, "learning_rate": 5.429578583303307e-05, "loss": 0.0376, "step": 39920 }, { "grad_norm": 0.4006814956665039, "learning_rate": 5.427519351707701e-05, "loss": 0.0439, "step": 39930 }, { "grad_norm": 0.49178245663642883, "learning_rate": 5.42546004706073e-05, "loss": 0.0428, "step": 39940 }, { "grad_norm": 0.5232398509979248, "learning_rate": 5.4234006697142735e-05, "loss": 0.0357, "step": 39950 }, { "grad_norm": 0.49318185448646545, "learning_rate": 5.421341220020224e-05, "loss": 0.0443, "step": 39960 }, { "grad_norm": 0.3116013705730438, "learning_rate": 5.419281698330482e-05, "loss": 0.0334, "step": 39970 }, { "grad_norm": 0.37963756918907166, "learning_rate": 5.4172221049969665e-05, "loss": 0.0322, "step": 39980 }, { "grad_norm": 0.45508772134780884, "learning_rate": 5.415162440371604e-05, "loss": 0.0356, "step": 39990 }, { "grad_norm": 0.5294517874717712, "learning_rate": 5.413102704806334e-05, "loss": 0.0487, "step": 40000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 40000, "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": 4, "trial_name": null, "trial_params": null }