{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6666666666666666, "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": 1.3409943580627441, "learning_rate": 3.0000000000000004e-07, "loss": 1.2932, "step": 10 }, { "grad_norm": 1.4018269777297974, "learning_rate": 6.333333333333333e-07, "loss": 1.2976, "step": 20 }, { "grad_norm": 1.290589451789856, "learning_rate": 9.666666666666668e-07, "loss": 1.2846, "step": 30 }, { "grad_norm": 1.0042083263397217, "learning_rate": 1.3e-06, "loss": 1.2801, "step": 40 }, { "grad_norm": 0.5740633606910706, "learning_rate": 1.6333333333333333e-06, "loss": 1.2625, "step": 50 }, { "grad_norm": 0.530264139175415, "learning_rate": 1.9666666666666668e-06, "loss": 1.2436, "step": 60 }, { "grad_norm": 0.33734890818595886, "learning_rate": 2.3e-06, "loss": 1.2123, "step": 70 }, { "grad_norm": 0.33657753467559814, "learning_rate": 2.6333333333333337e-06, "loss": 1.2147, "step": 80 }, { "grad_norm": 0.3979937434196472, "learning_rate": 2.966666666666667e-06, "loss": 1.2082, "step": 90 }, { "grad_norm": 0.4045668840408325, "learning_rate": 3.3e-06, "loss": 1.2188, "step": 100 }, { "grad_norm": 0.5521044731140137, "learning_rate": 3.633333333333334e-06, "loss": 1.1916, "step": 110 }, { "grad_norm": 0.6676170229911804, "learning_rate": 3.966666666666667e-06, "loss": 1.141, "step": 120 }, { "grad_norm": 0.5465882420539856, "learning_rate": 4.2999999999999995e-06, "loss": 1.1017, "step": 130 }, { "grad_norm": 0.5489553213119507, "learning_rate": 4.633333333333334e-06, "loss": 1.1023, "step": 140 }, { "grad_norm": 0.42917150259017944, "learning_rate": 4.966666666666667e-06, "loss": 1.091, "step": 150 }, { "grad_norm": 0.4442351758480072, "learning_rate": 5.3e-06, "loss": 1.076, "step": 160 }, { "grad_norm": 0.4976922869682312, "learning_rate": 5.633333333333333e-06, "loss": 1.0708, "step": 170 }, { "grad_norm": 0.509730875492096, "learning_rate": 5.9666666666666666e-06, "loss": 1.0648, "step": 180 }, { "grad_norm": 0.6341689229011536, "learning_rate": 6.300000000000001e-06, "loss": 1.0562, "step": 190 }, { "grad_norm": 0.6548420786857605, "learning_rate": 6.633333333333333e-06, "loss": 1.054, "step": 200 }, { "grad_norm": 0.5202745795249939, "learning_rate": 6.966666666666667e-06, "loss": 1.0531, "step": 210 }, { "grad_norm": 0.6620917320251465, "learning_rate": 7.2999999999999996e-06, "loss": 1.0517, "step": 220 }, { "grad_norm": 0.5149183869361877, "learning_rate": 7.633333333333334e-06, "loss": 1.0405, "step": 230 }, { "grad_norm": 0.5284705758094788, "learning_rate": 7.966666666666666e-06, "loss": 1.0414, "step": 240 }, { "grad_norm": 0.5993713140487671, "learning_rate": 8.3e-06, "loss": 1.0367, "step": 250 }, { "grad_norm": 0.5183808207511902, "learning_rate": 8.633333333333334e-06, "loss": 1.0409, "step": 260 }, { "grad_norm": 0.4466630220413208, "learning_rate": 8.966666666666668e-06, "loss": 1.0375, "step": 270 }, { "grad_norm": 0.4764702320098877, "learning_rate": 9.3e-06, "loss": 1.0359, "step": 280 }, { "grad_norm": 0.5116109848022461, "learning_rate": 9.633333333333335e-06, "loss": 1.0242, "step": 290 }, { "grad_norm": 0.5600728988647461, "learning_rate": 9.966666666666667e-06, "loss": 1.0277, "step": 300 }, { "grad_norm": 0.833690881729126, "learning_rate": 1.03e-05, "loss": 1.0204, "step": 310 }, { "grad_norm": 0.8469772338867188, "learning_rate": 1.0633333333333334e-05, "loss": 1.0047, "step": 320 }, { "grad_norm": 1.2423175573349, "learning_rate": 1.0966666666666666e-05, "loss": 0.9869, "step": 330 }, { "grad_norm": 0.8972920775413513, "learning_rate": 1.13e-05, "loss": 0.9614, "step": 340 }, { "grad_norm": 0.7836401462554932, "learning_rate": 1.1633333333333334e-05, "loss": 0.953, "step": 350 }, { "grad_norm": 1.4875415563583374, "learning_rate": 1.1966666666666668e-05, "loss": 0.9235, "step": 360 }, { "grad_norm": 1.079920768737793, "learning_rate": 1.23e-05, "loss": 0.9034, "step": 370 }, { "grad_norm": 1.5293974876403809, "learning_rate": 1.2633333333333333e-05, "loss": 0.8747, "step": 380 }, { "grad_norm": 1.3199079036712646, "learning_rate": 1.2966666666666669e-05, "loss": 0.8462, "step": 390 }, { "grad_norm": 1.616784691810608, "learning_rate": 1.3300000000000001e-05, "loss": 0.7956, "step": 400 }, { "grad_norm": 1.6455917358398438, "learning_rate": 1.3633333333333334e-05, "loss": 0.7655, "step": 410 }, { "grad_norm": 1.7950419187545776, "learning_rate": 1.3966666666666666e-05, "loss": 0.7124, "step": 420 }, { "grad_norm": 1.5076907873153687, "learning_rate": 1.43e-05, "loss": 0.6987, "step": 430 }, { "grad_norm": 1.964113473892212, "learning_rate": 1.4633333333333334e-05, "loss": 0.6736, "step": 440 }, { "grad_norm": 1.4842705726623535, "learning_rate": 1.4966666666666668e-05, "loss": 0.6651, "step": 450 }, { "grad_norm": 1.4045788049697876, "learning_rate": 1.53e-05, "loss": 0.6388, "step": 460 }, { "grad_norm": 1.4312362670898438, "learning_rate": 1.563333333333333e-05, "loss": 0.6282, "step": 470 }, { "grad_norm": 1.6672073602676392, "learning_rate": 1.5966666666666667e-05, "loss": 0.6091, "step": 480 }, { "grad_norm": 1.7698025703430176, "learning_rate": 1.63e-05, "loss": 0.5763, "step": 490 }, { "grad_norm": 1.8444621562957764, "learning_rate": 1.6633333333333336e-05, "loss": 0.5503, "step": 500 }, { "grad_norm": 1.9023514986038208, "learning_rate": 1.6966666666666668e-05, "loss": 0.5182, "step": 510 }, { "grad_norm": 1.448399543762207, "learning_rate": 1.73e-05, "loss": 0.4861, "step": 520 }, { "grad_norm": 1.6080480813980103, "learning_rate": 1.7633333333333336e-05, "loss": 0.4585, "step": 530 }, { "grad_norm": 1.7612723112106323, "learning_rate": 1.796666666666667e-05, "loss": 0.4439, "step": 540 }, { "grad_norm": 1.5014480352401733, "learning_rate": 1.83e-05, "loss": 0.4071, "step": 550 }, { "grad_norm": 2.058182954788208, "learning_rate": 1.8633333333333333e-05, "loss": 0.3941, "step": 560 }, { "grad_norm": 1.9121285676956177, "learning_rate": 1.896666666666667e-05, "loss": 0.3718, "step": 570 }, { "grad_norm": 1.9442795515060425, "learning_rate": 1.93e-05, "loss": 0.353, "step": 580 }, { "grad_norm": 1.7529265880584717, "learning_rate": 1.9633333333333334e-05, "loss": 0.3147, "step": 590 }, { "grad_norm": 1.845483660697937, "learning_rate": 1.9966666666666666e-05, "loss": 0.297, "step": 600 }, { "grad_norm": 2.2004470825195312, "learning_rate": 2.0300000000000002e-05, "loss": 0.2732, "step": 610 }, { "grad_norm": 1.791242003440857, "learning_rate": 2.0633333333333335e-05, "loss": 0.2611, "step": 620 }, { "grad_norm": 2.599447011947632, "learning_rate": 2.0966666666666667e-05, "loss": 0.2333, "step": 630 }, { "grad_norm": 2.062220573425293, "learning_rate": 2.13e-05, "loss": 0.2054, "step": 640 }, { "grad_norm": 1.9040703773498535, "learning_rate": 2.1633333333333332e-05, "loss": 0.1923, "step": 650 }, { "grad_norm": 2.4225902557373047, "learning_rate": 2.1966666666666668e-05, "loss": 0.1825, "step": 660 }, { "grad_norm": 1.7376368045806885, "learning_rate": 2.23e-05, "loss": 0.1576, "step": 670 }, { "grad_norm": 2.0857088565826416, "learning_rate": 2.2633333333333336e-05, "loss": 0.1454, "step": 680 }, { "grad_norm": 2.4043643474578857, "learning_rate": 2.2966666666666668e-05, "loss": 0.1453, "step": 690 }, { "grad_norm": 1.385374665260315, "learning_rate": 2.3300000000000004e-05, "loss": 0.1353, "step": 700 }, { "grad_norm": 2.503251552581787, "learning_rate": 2.3633333333333336e-05, "loss": 0.1506, "step": 710 }, { "grad_norm": 1.4551513195037842, "learning_rate": 2.396666666666667e-05, "loss": 0.135, "step": 720 }, { "grad_norm": 2.0421340465545654, "learning_rate": 2.43e-05, "loss": 0.1205, "step": 730 }, { "grad_norm": 1.8257296085357666, "learning_rate": 2.4633333333333334e-05, "loss": 0.1298, "step": 740 }, { "grad_norm": 1.8314361572265625, "learning_rate": 2.496666666666667e-05, "loss": 0.1302, "step": 750 }, { "grad_norm": 1.5614336729049683, "learning_rate": 2.5300000000000002e-05, "loss": 0.1309, "step": 760 }, { "grad_norm": 1.693200707435608, "learning_rate": 2.5633333333333338e-05, "loss": 0.1234, "step": 770 }, { "grad_norm": 1.62184476852417, "learning_rate": 2.5966666666666667e-05, "loss": 0.1273, "step": 780 }, { "grad_norm": 1.3120272159576416, "learning_rate": 2.6300000000000002e-05, "loss": 0.1226, "step": 790 }, { "grad_norm": 1.653055191040039, "learning_rate": 2.663333333333333e-05, "loss": 0.1232, "step": 800 }, { "grad_norm": 1.201752781867981, "learning_rate": 2.6966666666666667e-05, "loss": 0.1151, "step": 810 }, { "grad_norm": 1.6992968320846558, "learning_rate": 2.7300000000000003e-05, "loss": 0.1125, "step": 820 }, { "grad_norm": 1.6268329620361328, "learning_rate": 2.7633333333333332e-05, "loss": 0.1153, "step": 830 }, { "grad_norm": 1.4719772338867188, "learning_rate": 2.7966666666666668e-05, "loss": 0.1115, "step": 840 }, { "grad_norm": 1.4680430889129639, "learning_rate": 2.83e-05, "loss": 0.1124, "step": 850 }, { "grad_norm": 1.820801854133606, "learning_rate": 2.8633333333333336e-05, "loss": 0.1201, "step": 860 }, { "grad_norm": 1.3088511228561401, "learning_rate": 2.8966666666666668e-05, "loss": 0.1061, "step": 870 }, { "grad_norm": 2.017138957977295, "learning_rate": 2.93e-05, "loss": 0.1081, "step": 880 }, { "grad_norm": 1.6008727550506592, "learning_rate": 2.9633333333333336e-05, "loss": 0.1047, "step": 890 }, { "grad_norm": 1.4866894483566284, "learning_rate": 2.9966666666666672e-05, "loss": 0.0966, "step": 900 }, { "grad_norm": 1.825900673866272, "learning_rate": 3.03e-05, "loss": 0.0952, "step": 910 }, { "grad_norm": 1.5463889837265015, "learning_rate": 3.063333333333334e-05, "loss": 0.0982, "step": 920 }, { "grad_norm": 1.6832419633865356, "learning_rate": 3.096666666666666e-05, "loss": 0.0976, "step": 930 }, { "grad_norm": 1.2994757890701294, "learning_rate": 3.13e-05, "loss": 0.0935, "step": 940 }, { "grad_norm": 1.305489420890808, "learning_rate": 3.1633333333333334e-05, "loss": 0.0999, "step": 950 }, { "grad_norm": 1.4890505075454712, "learning_rate": 3.196666666666667e-05, "loss": 0.0998, "step": 960 }, { "grad_norm": 1.4655362367630005, "learning_rate": 3.2300000000000006e-05, "loss": 0.0978, "step": 970 }, { "grad_norm": 1.536659836769104, "learning_rate": 3.263333333333333e-05, "loss": 0.0984, "step": 980 }, { "grad_norm": 1.8284697532653809, "learning_rate": 3.296666666666667e-05, "loss": 0.1, "step": 990 }, { "grad_norm": 1.1327028274536133, "learning_rate": 3.33e-05, "loss": 0.1051, "step": 1000 }, { "grad_norm": 1.631208062171936, "learning_rate": 3.3633333333333335e-05, "loss": 0.0948, "step": 1010 }, { "grad_norm": 1.1060216426849365, "learning_rate": 3.396666666666667e-05, "loss": 0.0988, "step": 1020 }, { "grad_norm": 1.5067026615142822, "learning_rate": 3.430000000000001e-05, "loss": 0.102, "step": 1030 }, { "grad_norm": 1.2366249561309814, "learning_rate": 3.463333333333333e-05, "loss": 0.0939, "step": 1040 }, { "grad_norm": 1.249070644378662, "learning_rate": 3.496666666666667e-05, "loss": 0.0901, "step": 1050 }, { "grad_norm": 1.4778685569763184, "learning_rate": 3.53e-05, "loss": 0.0908, "step": 1060 }, { "grad_norm": 1.2828925848007202, "learning_rate": 3.563333333333334e-05, "loss": 0.0924, "step": 1070 }, { "grad_norm": 1.4040788412094116, "learning_rate": 3.596666666666667e-05, "loss": 0.0838, "step": 1080 }, { "grad_norm": 1.2230231761932373, "learning_rate": 3.63e-05, "loss": 0.0881, "step": 1090 }, { "grad_norm": 1.3488352298736572, "learning_rate": 3.6633333333333334e-05, "loss": 0.0955, "step": 1100 }, { "grad_norm": 1.0227181911468506, "learning_rate": 3.6966666666666666e-05, "loss": 0.0831, "step": 1110 }, { "grad_norm": 0.8721468448638916, "learning_rate": 3.73e-05, "loss": 0.0915, "step": 1120 }, { "grad_norm": 1.2423741817474365, "learning_rate": 3.763333333333334e-05, "loss": 0.0887, "step": 1130 }, { "grad_norm": 1.0989211797714233, "learning_rate": 3.796666666666667e-05, "loss": 0.0835, "step": 1140 }, { "grad_norm": 1.0175212621688843, "learning_rate": 3.83e-05, "loss": 0.0917, "step": 1150 }, { "grad_norm": 1.3825503587722778, "learning_rate": 3.8633333333333335e-05, "loss": 0.0858, "step": 1160 }, { "grad_norm": 1.1449437141418457, "learning_rate": 3.896666666666667e-05, "loss": 0.0902, "step": 1170 }, { "grad_norm": 1.2131868600845337, "learning_rate": 3.9300000000000007e-05, "loss": 0.0895, "step": 1180 }, { "grad_norm": 0.933954119682312, "learning_rate": 3.963333333333333e-05, "loss": 0.0837, "step": 1190 }, { "grad_norm": 1.2540720701217651, "learning_rate": 3.996666666666667e-05, "loss": 0.0909, "step": 1200 }, { "grad_norm": 1.4824661016464233, "learning_rate": 4.0300000000000004e-05, "loss": 0.0845, "step": 1210 }, { "grad_norm": 0.8873796463012695, "learning_rate": 4.0633333333333336e-05, "loss": 0.0775, "step": 1220 }, { "grad_norm": 1.157019853591919, "learning_rate": 4.096666666666667e-05, "loss": 0.081, "step": 1230 }, { "grad_norm": 1.2332080602645874, "learning_rate": 4.13e-05, "loss": 0.0869, "step": 1240 }, { "grad_norm": 0.9840149879455566, "learning_rate": 4.1633333333333333e-05, "loss": 0.0862, "step": 1250 }, { "grad_norm": 1.1323007345199585, "learning_rate": 4.196666666666667e-05, "loss": 0.0704, "step": 1260 }, { "grad_norm": 1.1442773342132568, "learning_rate": 4.23e-05, "loss": 0.0921, "step": 1270 }, { "grad_norm": 1.3689532279968262, "learning_rate": 4.263333333333334e-05, "loss": 0.0813, "step": 1280 }, { "grad_norm": 1.1174355745315552, "learning_rate": 4.296666666666666e-05, "loss": 0.0828, "step": 1290 }, { "grad_norm": 1.4927648305892944, "learning_rate": 4.33e-05, "loss": 0.0806, "step": 1300 }, { "grad_norm": 0.9332654476165771, "learning_rate": 4.3633333333333335e-05, "loss": 0.0765, "step": 1310 }, { "grad_norm": 0.9787437319755554, "learning_rate": 4.396666666666667e-05, "loss": 0.094, "step": 1320 }, { "grad_norm": 0.7872252464294434, "learning_rate": 4.43e-05, "loss": 0.0769, "step": 1330 }, { "grad_norm": 0.9678040742874146, "learning_rate": 4.463333333333334e-05, "loss": 0.07, "step": 1340 }, { "grad_norm": 0.8637449741363525, "learning_rate": 4.496666666666667e-05, "loss": 0.0801, "step": 1350 }, { "grad_norm": 0.9531744122505188, "learning_rate": 4.53e-05, "loss": 0.0808, "step": 1360 }, { "grad_norm": 1.139052152633667, "learning_rate": 4.5633333333333336e-05, "loss": 0.0744, "step": 1370 }, { "grad_norm": 0.9093393087387085, "learning_rate": 4.596666666666667e-05, "loss": 0.0721, "step": 1380 }, { "grad_norm": 0.9640454053878784, "learning_rate": 4.630000000000001e-05, "loss": 0.0683, "step": 1390 }, { "grad_norm": 1.0027737617492676, "learning_rate": 4.663333333333333e-05, "loss": 0.0683, "step": 1400 }, { "grad_norm": 1.1405223608016968, "learning_rate": 4.696666666666667e-05, "loss": 0.0712, "step": 1410 }, { "grad_norm": 0.8201662302017212, "learning_rate": 4.73e-05, "loss": 0.0736, "step": 1420 }, { "grad_norm": 1.0546902418136597, "learning_rate": 4.763333333333334e-05, "loss": 0.0825, "step": 1430 }, { "grad_norm": 0.9859091639518738, "learning_rate": 4.796666666666667e-05, "loss": 0.0725, "step": 1440 }, { "grad_norm": 0.941483199596405, "learning_rate": 4.83e-05, "loss": 0.0661, "step": 1450 }, { "grad_norm": 0.8626189827919006, "learning_rate": 4.8633333333333334e-05, "loss": 0.074, "step": 1460 }, { "grad_norm": 0.9490450620651245, "learning_rate": 4.8966666666666667e-05, "loss": 0.0657, "step": 1470 }, { "grad_norm": 0.9417278170585632, "learning_rate": 4.93e-05, "loss": 0.0687, "step": 1480 }, { "grad_norm": 0.9270307421684265, "learning_rate": 4.963333333333334e-05, "loss": 0.0758, "step": 1490 }, { "grad_norm": 1.002264380455017, "learning_rate": 4.996666666666667e-05, "loss": 0.0726, "step": 1500 }, { "grad_norm": 0.8564841747283936, "learning_rate": 5.03e-05, "loss": 0.0733, "step": 1510 }, { "grad_norm": 0.9816124439239502, "learning_rate": 5.0633333333333335e-05, "loss": 0.0719, "step": 1520 }, { "grad_norm": 0.9341065287590027, "learning_rate": 5.0966666666666674e-05, "loss": 0.0714, "step": 1530 }, { "grad_norm": 0.8709513545036316, "learning_rate": 5.130000000000001e-05, "loss": 0.072, "step": 1540 }, { "grad_norm": 1.033493995666504, "learning_rate": 5.163333333333333e-05, "loss": 0.0674, "step": 1550 }, { "grad_norm": 0.9912322759628296, "learning_rate": 5.196666666666667e-05, "loss": 0.0672, "step": 1560 }, { "grad_norm": 1.0924186706542969, "learning_rate": 5.2300000000000004e-05, "loss": 0.071, "step": 1570 }, { "grad_norm": 0.6413648724555969, "learning_rate": 5.2633333333333336e-05, "loss": 0.0623, "step": 1580 }, { "grad_norm": 0.997759222984314, "learning_rate": 5.296666666666666e-05, "loss": 0.0663, "step": 1590 }, { "grad_norm": 0.8930423259735107, "learning_rate": 5.330000000000001e-05, "loss": 0.0713, "step": 1600 }, { "grad_norm": 0.8437132239341736, "learning_rate": 5.3633333333333334e-05, "loss": 0.0636, "step": 1610 }, { "grad_norm": 0.9109311103820801, "learning_rate": 5.3966666666666666e-05, "loss": 0.0637, "step": 1620 }, { "grad_norm": 0.873157799243927, "learning_rate": 5.4300000000000005e-05, "loss": 0.0711, "step": 1630 }, { "grad_norm": 0.9040642976760864, "learning_rate": 5.463333333333334e-05, "loss": 0.0704, "step": 1640 }, { "grad_norm": 0.973240315914154, "learning_rate": 5.496666666666666e-05, "loss": 0.0646, "step": 1650 }, { "grad_norm": 0.8889073729515076, "learning_rate": 5.530000000000001e-05, "loss": 0.0636, "step": 1660 }, { "grad_norm": 0.8204637169837952, "learning_rate": 5.5633333333333335e-05, "loss": 0.0684, "step": 1670 }, { "grad_norm": 0.724500298500061, "learning_rate": 5.596666666666667e-05, "loss": 0.062, "step": 1680 }, { "grad_norm": 0.7349976897239685, "learning_rate": 5.63e-05, "loss": 0.0608, "step": 1690 }, { "grad_norm": 0.789164662361145, "learning_rate": 5.663333333333334e-05, "loss": 0.0721, "step": 1700 }, { "grad_norm": 0.8389636278152466, "learning_rate": 5.696666666666667e-05, "loss": 0.0656, "step": 1710 }, { "grad_norm": 0.8705915212631226, "learning_rate": 5.73e-05, "loss": 0.0679, "step": 1720 }, { "grad_norm": 0.6977961659431458, "learning_rate": 5.7633333333333336e-05, "loss": 0.0684, "step": 1730 }, { "grad_norm": 1.001558780670166, "learning_rate": 5.796666666666667e-05, "loss": 0.0618, "step": 1740 }, { "grad_norm": 0.8001946806907654, "learning_rate": 5.83e-05, "loss": 0.0597, "step": 1750 }, { "grad_norm": 0.769285261631012, "learning_rate": 5.863333333333334e-05, "loss": 0.0602, "step": 1760 }, { "grad_norm": 0.8238266706466675, "learning_rate": 5.896666666666667e-05, "loss": 0.0722, "step": 1770 }, { "grad_norm": 0.8632376194000244, "learning_rate": 5.93e-05, "loss": 0.0656, "step": 1780 }, { "grad_norm": 0.9744498133659363, "learning_rate": 5.9633333333333344e-05, "loss": 0.0646, "step": 1790 }, { "grad_norm": 0.7233326435089111, "learning_rate": 5.996666666666667e-05, "loss": 0.0716, "step": 1800 }, { "grad_norm": 0.9626665115356445, "learning_rate": 6.03e-05, "loss": 0.0625, "step": 1810 }, { "grad_norm": 0.8173403143882751, "learning_rate": 6.063333333333333e-05, "loss": 0.0674, "step": 1820 }, { "grad_norm": 0.7112985253334045, "learning_rate": 6.0966666666666674e-05, "loss": 0.0627, "step": 1830 }, { "grad_norm": 0.6004302501678467, "learning_rate": 6.13e-05, "loss": 0.0563, "step": 1840 }, { "grad_norm": 0.612101674079895, "learning_rate": 6.163333333333333e-05, "loss": 0.0602, "step": 1850 }, { "grad_norm": 0.6649537086486816, "learning_rate": 6.196666666666668e-05, "loss": 0.0577, "step": 1860 }, { "grad_norm": 1.043892502784729, "learning_rate": 6.23e-05, "loss": 0.0633, "step": 1870 }, { "grad_norm": 0.5617967247962952, "learning_rate": 6.263333333333333e-05, "loss": 0.0602, "step": 1880 }, { "grad_norm": 0.7084012031555176, "learning_rate": 6.296666666666667e-05, "loss": 0.0563, "step": 1890 }, { "grad_norm": 0.7328187227249146, "learning_rate": 6.330000000000001e-05, "loss": 0.0627, "step": 1900 }, { "grad_norm": 0.7772453427314758, "learning_rate": 6.363333333333334e-05, "loss": 0.0617, "step": 1910 }, { "grad_norm": 0.7743660807609558, "learning_rate": 6.396666666666667e-05, "loss": 0.0543, "step": 1920 }, { "grad_norm": 0.7355103492736816, "learning_rate": 6.43e-05, "loss": 0.0588, "step": 1930 }, { "grad_norm": 0.913023054599762, "learning_rate": 6.463333333333334e-05, "loss": 0.0725, "step": 1940 }, { "grad_norm": 0.6234018206596375, "learning_rate": 6.496666666666667e-05, "loss": 0.0587, "step": 1950 }, { "grad_norm": 0.5912435054779053, "learning_rate": 6.53e-05, "loss": 0.0607, "step": 1960 }, { "grad_norm": 0.9264614582061768, "learning_rate": 6.563333333333333e-05, "loss": 0.0629, "step": 1970 }, { "grad_norm": 0.7084673047065735, "learning_rate": 6.596666666666667e-05, "loss": 0.0665, "step": 1980 }, { "grad_norm": 0.7109888792037964, "learning_rate": 6.630000000000001e-05, "loss": 0.0548, "step": 1990 }, { "grad_norm": 1.070934534072876, "learning_rate": 6.663333333333333e-05, "loss": 0.0605, "step": 2000 }, { "grad_norm": 0.6817998886108398, "learning_rate": 6.696666666666666e-05, "loss": 0.0617, "step": 2010 }, { "grad_norm": 0.8135288953781128, "learning_rate": 6.730000000000001e-05, "loss": 0.061, "step": 2020 }, { "grad_norm": 0.6769553422927856, "learning_rate": 6.763333333333334e-05, "loss": 0.0565, "step": 2030 }, { "grad_norm": 0.6141065955162048, "learning_rate": 6.796666666666666e-05, "loss": 0.064, "step": 2040 }, { "grad_norm": 0.6450388431549072, "learning_rate": 6.83e-05, "loss": 0.0644, "step": 2050 }, { "grad_norm": 0.6130831837654114, "learning_rate": 6.863333333333334e-05, "loss": 0.0607, "step": 2060 }, { "grad_norm": 0.5850309729576111, "learning_rate": 6.896666666666667e-05, "loss": 0.0595, "step": 2070 }, { "grad_norm": 0.8597639203071594, "learning_rate": 6.93e-05, "loss": 0.0612, "step": 2080 }, { "grad_norm": 0.6991277933120728, "learning_rate": 6.963333333333334e-05, "loss": 0.0639, "step": 2090 }, { "grad_norm": 0.7476745843887329, "learning_rate": 6.996666666666667e-05, "loss": 0.0581, "step": 2100 }, { "grad_norm": 0.9618244767189026, "learning_rate": 7.03e-05, "loss": 0.0586, "step": 2110 }, { "grad_norm": 0.6543978452682495, "learning_rate": 7.063333333333333e-05, "loss": 0.0553, "step": 2120 }, { "grad_norm": 0.8125239610671997, "learning_rate": 7.096666666666667e-05, "loss": 0.0623, "step": 2130 }, { "grad_norm": 0.6939046382904053, "learning_rate": 7.13e-05, "loss": 0.0583, "step": 2140 }, { "grad_norm": 0.7810367345809937, "learning_rate": 7.163333333333334e-05, "loss": 0.0608, "step": 2150 }, { "grad_norm": 0.672482430934906, "learning_rate": 7.196666666666668e-05, "loss": 0.0493, "step": 2160 }, { "grad_norm": 0.6892238855361938, "learning_rate": 7.23e-05, "loss": 0.057, "step": 2170 }, { "grad_norm": 0.6374918222427368, "learning_rate": 7.263333333333334e-05, "loss": 0.0559, "step": 2180 }, { "grad_norm": 0.47278136014938354, "learning_rate": 7.296666666666667e-05, "loss": 0.0555, "step": 2190 }, { "grad_norm": 0.692745566368103, "learning_rate": 7.33e-05, "loss": 0.0515, "step": 2200 }, { "grad_norm": 0.5157440900802612, "learning_rate": 7.363333333333334e-05, "loss": 0.0519, "step": 2210 }, { "grad_norm": 0.648242712020874, "learning_rate": 7.396666666666667e-05, "loss": 0.0539, "step": 2220 }, { "grad_norm": 0.5529076457023621, "learning_rate": 7.43e-05, "loss": 0.0546, "step": 2230 }, { "grad_norm": 0.7032095193862915, "learning_rate": 7.463333333333334e-05, "loss": 0.0555, "step": 2240 }, { "grad_norm": 0.6388554573059082, "learning_rate": 7.496666666666667e-05, "loss": 0.0634, "step": 2250 }, { "grad_norm": 0.6875872611999512, "learning_rate": 7.53e-05, "loss": 0.0634, "step": 2260 }, { "grad_norm": 0.6595812439918518, "learning_rate": 7.563333333333333e-05, "loss": 0.0621, "step": 2270 }, { "grad_norm": 0.673121988773346, "learning_rate": 7.596666666666668e-05, "loss": 0.0537, "step": 2280 }, { "grad_norm": 0.5117025971412659, "learning_rate": 7.630000000000001e-05, "loss": 0.0501, "step": 2290 }, { "grad_norm": 0.6690002679824829, "learning_rate": 7.663333333333333e-05, "loss": 0.0542, "step": 2300 }, { "grad_norm": 0.5160834789276123, "learning_rate": 7.696666666666668e-05, "loss": 0.0564, "step": 2310 }, { "grad_norm": 0.554044783115387, "learning_rate": 7.730000000000001e-05, "loss": 0.0548, "step": 2320 }, { "grad_norm": 0.6758227944374084, "learning_rate": 7.763333333333334e-05, "loss": 0.0586, "step": 2330 }, { "grad_norm": 0.5237917900085449, "learning_rate": 7.796666666666666e-05, "loss": 0.0559, "step": 2340 }, { "grad_norm": 0.8799295425415039, "learning_rate": 7.83e-05, "loss": 0.0538, "step": 2350 }, { "grad_norm": 0.6447179913520813, "learning_rate": 7.863333333333334e-05, "loss": 0.0595, "step": 2360 }, { "grad_norm": 0.5708173513412476, "learning_rate": 7.896666666666667e-05, "loss": 0.0571, "step": 2370 }, { "grad_norm": 0.8418899774551392, "learning_rate": 7.93e-05, "loss": 0.0594, "step": 2380 }, { "grad_norm": 0.6246424913406372, "learning_rate": 7.963333333333334e-05, "loss": 0.0539, "step": 2390 }, { "grad_norm": 0.6540073752403259, "learning_rate": 7.996666666666667e-05, "loss": 0.0566, "step": 2400 }, { "grad_norm": 0.5650169849395752, "learning_rate": 8.030000000000001e-05, "loss": 0.0515, "step": 2410 }, { "grad_norm": 0.742441713809967, "learning_rate": 8.063333333333333e-05, "loss": 0.0543, "step": 2420 }, { "grad_norm": 0.6716034412384033, "learning_rate": 8.096666666666667e-05, "loss": 0.0579, "step": 2430 }, { "grad_norm": 0.622184693813324, "learning_rate": 8.13e-05, "loss": 0.0549, "step": 2440 }, { "grad_norm": 0.8220912218093872, "learning_rate": 8.163333333333334e-05, "loss": 0.0563, "step": 2450 }, { "grad_norm": 0.5408906936645508, "learning_rate": 8.196666666666668e-05, "loss": 0.0591, "step": 2460 }, { "grad_norm": 0.6848285794258118, "learning_rate": 8.23e-05, "loss": 0.053, "step": 2470 }, { "grad_norm": 0.898846447467804, "learning_rate": 8.263333333333334e-05, "loss": 0.0612, "step": 2480 }, { "grad_norm": 0.6678881049156189, "learning_rate": 8.296666666666667e-05, "loss": 0.0539, "step": 2490 }, { "grad_norm": 0.560133159160614, "learning_rate": 8.33e-05, "loss": 0.0547, "step": 2500 }, { "grad_norm": 0.6717010736465454, "learning_rate": 8.363333333333334e-05, "loss": 0.0571, "step": 2510 }, { "grad_norm": 0.6481582522392273, "learning_rate": 8.396666666666667e-05, "loss": 0.0536, "step": 2520 }, { "grad_norm": 0.6753191947937012, "learning_rate": 8.43e-05, "loss": 0.0519, "step": 2530 }, { "grad_norm": 0.5581200122833252, "learning_rate": 8.463333333333335e-05, "loss": 0.0513, "step": 2540 }, { "grad_norm": 0.7106221318244934, "learning_rate": 8.496666666666667e-05, "loss": 0.0577, "step": 2550 }, { "grad_norm": 0.7249637246131897, "learning_rate": 8.53e-05, "loss": 0.0491, "step": 2560 }, { "grad_norm": 0.6001949310302734, "learning_rate": 8.563333333333333e-05, "loss": 0.051, "step": 2570 }, { "grad_norm": 0.4366547763347626, "learning_rate": 8.596666666666668e-05, "loss": 0.0568, "step": 2580 }, { "grad_norm": 0.4842328131198883, "learning_rate": 8.63e-05, "loss": 0.049, "step": 2590 }, { "grad_norm": 0.6565221548080444, "learning_rate": 8.663333333333333e-05, "loss": 0.0542, "step": 2600 }, { "grad_norm": 0.7032201290130615, "learning_rate": 8.696666666666668e-05, "loss": 0.0553, "step": 2610 }, { "grad_norm": 0.5990325212478638, "learning_rate": 8.730000000000001e-05, "loss": 0.0515, "step": 2620 }, { "grad_norm": 0.6935186982154846, "learning_rate": 8.763333333333334e-05, "loss": 0.0559, "step": 2630 }, { "grad_norm": 0.6349155306816101, "learning_rate": 8.796666666666667e-05, "loss": 0.0545, "step": 2640 }, { "grad_norm": 0.5749071836471558, "learning_rate": 8.83e-05, "loss": 0.0553, "step": 2650 }, { "grad_norm": 0.5197382569313049, "learning_rate": 8.863333333333334e-05, "loss": 0.05, "step": 2660 }, { "grad_norm": 0.5119301080703735, "learning_rate": 8.896666666666667e-05, "loss": 0.0539, "step": 2670 }, { "grad_norm": 0.48509639501571655, "learning_rate": 8.93e-05, "loss": 0.0494, "step": 2680 }, { "grad_norm": 0.5243433117866516, "learning_rate": 8.963333333333333e-05, "loss": 0.0562, "step": 2690 }, { "grad_norm": 0.6515861749649048, "learning_rate": 8.996666666666667e-05, "loss": 0.0518, "step": 2700 }, { "grad_norm": 0.6514577865600586, "learning_rate": 9.030000000000001e-05, "loss": 0.0583, "step": 2710 }, { "grad_norm": 0.5067355632781982, "learning_rate": 9.063333333333333e-05, "loss": 0.0486, "step": 2720 }, { "grad_norm": 0.6687017679214478, "learning_rate": 9.096666666666666e-05, "loss": 0.0526, "step": 2730 }, { "grad_norm": 0.48689061403274536, "learning_rate": 9.130000000000001e-05, "loss": 0.0534, "step": 2740 }, { "grad_norm": 0.5722624063491821, "learning_rate": 9.163333333333334e-05, "loss": 0.0517, "step": 2750 }, { "grad_norm": 0.6039711833000183, "learning_rate": 9.196666666666666e-05, "loss": 0.0542, "step": 2760 }, { "grad_norm": 0.5194990634918213, "learning_rate": 9.230000000000001e-05, "loss": 0.0453, "step": 2770 }, { "grad_norm": 0.49638766050338745, "learning_rate": 9.263333333333334e-05, "loss": 0.0493, "step": 2780 }, { "grad_norm": 0.4542931616306305, "learning_rate": 9.296666666666667e-05, "loss": 0.052, "step": 2790 }, { "grad_norm": 0.5047942996025085, "learning_rate": 9.33e-05, "loss": 0.0492, "step": 2800 }, { "grad_norm": 0.5123155117034912, "learning_rate": 9.363333333333334e-05, "loss": 0.0475, "step": 2810 }, { "grad_norm": 0.565690279006958, "learning_rate": 9.396666666666667e-05, "loss": 0.0451, "step": 2820 }, { "grad_norm": 0.44891998171806335, "learning_rate": 9.43e-05, "loss": 0.0539, "step": 2830 }, { "grad_norm": 0.49138736724853516, "learning_rate": 9.463333333333333e-05, "loss": 0.0549, "step": 2840 }, { "grad_norm": 0.6099728941917419, "learning_rate": 9.496666666666667e-05, "loss": 0.0485, "step": 2850 }, { "grad_norm": 0.50760418176651, "learning_rate": 9.53e-05, "loss": 0.0475, "step": 2860 }, { "grad_norm": 0.437981516122818, "learning_rate": 9.563333333333334e-05, "loss": 0.0493, "step": 2870 }, { "grad_norm": 0.645258903503418, "learning_rate": 9.596666666666668e-05, "loss": 0.0464, "step": 2880 }, { "grad_norm": 0.4596323072910309, "learning_rate": 9.63e-05, "loss": 0.0461, "step": 2890 }, { "grad_norm": 0.5726641416549683, "learning_rate": 9.663333333333334e-05, "loss": 0.0488, "step": 2900 }, { "grad_norm": 0.5011798739433289, "learning_rate": 9.696666666666667e-05, "loss": 0.0465, "step": 2910 }, { "grad_norm": 0.5291911363601685, "learning_rate": 9.730000000000001e-05, "loss": 0.047, "step": 2920 }, { "grad_norm": 0.5504176020622253, "learning_rate": 9.763333333333334e-05, "loss": 0.0541, "step": 2930 }, { "grad_norm": 0.42448610067367554, "learning_rate": 9.796666666666667e-05, "loss": 0.0571, "step": 2940 }, { "grad_norm": 0.4829455614089966, "learning_rate": 9.83e-05, "loss": 0.0535, "step": 2950 }, { "grad_norm": 0.4329001009464264, "learning_rate": 9.863333333333334e-05, "loss": 0.0568, "step": 2960 }, { "grad_norm": 0.4769044816493988, "learning_rate": 9.896666666666667e-05, "loss": 0.0486, "step": 2970 }, { "grad_norm": 0.5948467254638672, "learning_rate": 9.93e-05, "loss": 0.0481, "step": 2980 }, { "grad_norm": 0.5535639524459839, "learning_rate": 9.963333333333333e-05, "loss": 0.0482, "step": 2990 }, { "grad_norm": 0.6949828863143921, "learning_rate": 9.996666666666668e-05, "loss": 0.0502, "step": 3000 }, { "grad_norm": 0.49713191390037537, "learning_rate": 9.999999384858465e-05, "loss": 0.0562, "step": 3010 }, { "grad_norm": 0.41346806287765503, "learning_rate": 9.999997258443473e-05, "loss": 0.0509, "step": 3020 }, { "grad_norm": 0.43312913179397583, "learning_rate": 9.999993613161331e-05, "loss": 0.0462, "step": 3030 }, { "grad_norm": 0.6210405230522156, "learning_rate": 9.999988449013146e-05, "loss": 0.0433, "step": 3040 }, { "grad_norm": 0.5242516994476318, "learning_rate": 9.99998176600049e-05, "loss": 0.047, "step": 3050 }, { "grad_norm": 0.4236348569393158, "learning_rate": 9.999973564125389e-05, "loss": 0.0448, "step": 3060 }, { "grad_norm": 0.5816516280174255, "learning_rate": 9.999963843390335e-05, "loss": 0.0511, "step": 3070 }, { "grad_norm": 0.47085490822792053, "learning_rate": 9.999952603798282e-05, "loss": 0.0482, "step": 3080 }, { "grad_norm": 0.5314513444900513, "learning_rate": 9.999939845352646e-05, "loss": 0.0489, "step": 3090 }, { "grad_norm": 0.5719561576843262, "learning_rate": 9.999925568057298e-05, "loss": 0.0553, "step": 3100 }, { "grad_norm": 0.5746577978134155, "learning_rate": 9.999909771916578e-05, "loss": 0.0547, "step": 3110 }, { "grad_norm": 0.5194207429885864, "learning_rate": 9.999892456935285e-05, "loss": 0.0436, "step": 3120 }, { "grad_norm": 0.4348747134208679, "learning_rate": 9.999873623118679e-05, "loss": 0.0557, "step": 3130 }, { "grad_norm": 0.42849013209342957, "learning_rate": 9.999853270472479e-05, "loss": 0.0456, "step": 3140 }, { "grad_norm": 0.7162014245986938, "learning_rate": 9.999831399002871e-05, "loss": 0.0518, "step": 3150 }, { "grad_norm": 0.6185081005096436, "learning_rate": 9.999808008716494e-05, "loss": 0.0459, "step": 3160 }, { "grad_norm": 0.5380639433860779, "learning_rate": 9.999783099620459e-05, "loss": 0.0507, "step": 3170 }, { "grad_norm": 0.4033256471157074, "learning_rate": 9.999756671722328e-05, "loss": 0.0508, "step": 3180 }, { "grad_norm": 0.5656639337539673, "learning_rate": 9.99972872503013e-05, "loss": 0.0513, "step": 3190 }, { "grad_norm": 0.5318040251731873, "learning_rate": 9.999699259552359e-05, "loss": 0.046, "step": 3200 }, { "grad_norm": 0.48459768295288086, "learning_rate": 9.99966827529796e-05, "loss": 0.0442, "step": 3210 }, { "grad_norm": 0.50738126039505, "learning_rate": 9.999635772276348e-05, "loss": 0.0488, "step": 3220 }, { "grad_norm": 0.44412481784820557, "learning_rate": 9.999601750497396e-05, "loss": 0.0538, "step": 3230 }, { "grad_norm": 0.5263833999633789, "learning_rate": 9.99956620997144e-05, "loss": 0.047, "step": 3240 }, { "grad_norm": 0.46966832876205444, "learning_rate": 9.999529150709275e-05, "loss": 0.0485, "step": 3250 }, { "grad_norm": 0.5718423128128052, "learning_rate": 9.999490572722158e-05, "loss": 0.051, "step": 3260 }, { "grad_norm": 0.5385769605636597, "learning_rate": 9.99945047602181e-05, "loss": 0.0493, "step": 3270 }, { "grad_norm": 0.5309423208236694, "learning_rate": 9.99940886062041e-05, "loss": 0.0432, "step": 3280 }, { "grad_norm": 0.3632584810256958, "learning_rate": 9.999365726530599e-05, "loss": 0.0465, "step": 3290 }, { "grad_norm": 0.4449458718299866, "learning_rate": 9.999321073765481e-05, "loss": 0.0481, "step": 3300 }, { "grad_norm": 0.41699162125587463, "learning_rate": 9.99927490233862e-05, "loss": 0.0476, "step": 3310 }, { "grad_norm": 0.4513562023639679, "learning_rate": 9.999227212264043e-05, "loss": 0.0503, "step": 3320 }, { "grad_norm": 0.4565102756023407, "learning_rate": 9.999178003556236e-05, "loss": 0.0483, "step": 3330 }, { "grad_norm": 0.4082600474357605, "learning_rate": 9.999127276230146e-05, "loss": 0.0483, "step": 3340 }, { "grad_norm": 0.41847455501556396, "learning_rate": 9.999075030301184e-05, "loss": 0.0475, "step": 3350 }, { "grad_norm": 0.4276575744152069, "learning_rate": 9.999021265785221e-05, "loss": 0.0429, "step": 3360 }, { "grad_norm": 0.4172125458717346, "learning_rate": 9.998965982698589e-05, "loss": 0.0449, "step": 3370 }, { "grad_norm": 0.4089154303073883, "learning_rate": 9.998909181058082e-05, "loss": 0.0426, "step": 3380 }, { "grad_norm": 0.35062670707702637, "learning_rate": 9.998850860880953e-05, "loss": 0.044, "step": 3390 }, { "grad_norm": 0.4332718551158905, "learning_rate": 9.998791022184922e-05, "loss": 0.045, "step": 3400 }, { "grad_norm": 0.5806475281715393, "learning_rate": 9.99872966498816e-05, "loss": 0.0509, "step": 3410 }, { "grad_norm": 0.44285985827445984, "learning_rate": 9.998666789309313e-05, "loss": 0.0432, "step": 3420 }, { "grad_norm": 0.4274218678474426, "learning_rate": 9.998602395167475e-05, "loss": 0.0414, "step": 3430 }, { "grad_norm": 0.6059900522232056, "learning_rate": 9.998536482582213e-05, "loss": 0.0474, "step": 3440 }, { "grad_norm": 0.5275575518608093, "learning_rate": 9.998469051573544e-05, "loss": 0.04, "step": 3450 }, { "grad_norm": 0.40877389907836914, "learning_rate": 9.998400102161954e-05, "loss": 0.0436, "step": 3460 }, { "grad_norm": 0.5050567984580994, "learning_rate": 9.998329634368388e-05, "loss": 0.0516, "step": 3470 }, { "grad_norm": 0.3676494061946869, "learning_rate": 9.998257648214253e-05, "loss": 0.0459, "step": 3480 }, { "grad_norm": 0.3629034459590912, "learning_rate": 9.998184143721417e-05, "loss": 0.0439, "step": 3490 }, { "grad_norm": 0.32410699129104614, "learning_rate": 9.998109120912206e-05, "loss": 0.0421, "step": 3500 }, { "grad_norm": 0.5103791356086731, "learning_rate": 9.998032579809411e-05, "loss": 0.0502, "step": 3510 }, { "grad_norm": 0.4009677767753601, "learning_rate": 9.997954520436286e-05, "loss": 0.0466, "step": 3520 }, { "grad_norm": 0.4283072352409363, "learning_rate": 9.997874942816538e-05, "loss": 0.0485, "step": 3530 }, { "grad_norm": 0.4755754768848419, "learning_rate": 9.997793846974345e-05, "loss": 0.0494, "step": 3540 }, { "grad_norm": 0.5817909240722656, "learning_rate": 9.997711232934341e-05, "loss": 0.0471, "step": 3550 }, { "grad_norm": 0.5930608510971069, "learning_rate": 9.99762710072162e-05, "loss": 0.0471, "step": 3560 }, { "grad_norm": 0.40242838859558105, "learning_rate": 9.997541450361743e-05, "loss": 0.0471, "step": 3570 }, { "grad_norm": 0.3893769085407257, "learning_rate": 9.997454281880723e-05, "loss": 0.0468, "step": 3580 }, { "grad_norm": 0.3425997495651245, "learning_rate": 9.997365595305044e-05, "loss": 0.0434, "step": 3590 }, { "grad_norm": 0.526712954044342, "learning_rate": 9.997275390661644e-05, "loss": 0.0456, "step": 3600 }, { "grad_norm": 0.45006272196769714, "learning_rate": 9.997183667977926e-05, "loss": 0.0445, "step": 3610 }, { "grad_norm": 0.44465991854667664, "learning_rate": 9.997090427281752e-05, "loss": 0.0414, "step": 3620 }, { "grad_norm": 0.3475283682346344, "learning_rate": 9.996995668601448e-05, "loss": 0.0458, "step": 3630 }, { "grad_norm": 0.5391725301742554, "learning_rate": 9.996899391965798e-05, "loss": 0.0407, "step": 3640 }, { "grad_norm": 0.4582381844520569, "learning_rate": 9.996801597404048e-05, "loss": 0.0411, "step": 3650 }, { "grad_norm": 0.51310795545578, "learning_rate": 9.996702284945905e-05, "loss": 0.042, "step": 3660 }, { "grad_norm": 0.39491528272628784, "learning_rate": 9.996601454621539e-05, "loss": 0.0402, "step": 3670 }, { "grad_norm": 0.3358951807022095, "learning_rate": 9.996499106461577e-05, "loss": 0.0398, "step": 3680 }, { "grad_norm": 0.34933358430862427, "learning_rate": 9.996395240497112e-05, "loss": 0.0459, "step": 3690 }, { "grad_norm": 0.4989163279533386, "learning_rate": 9.996289856759696e-05, "loss": 0.0372, "step": 3700 }, { "grad_norm": 0.3805118501186371, "learning_rate": 9.996182955281342e-05, "loss": 0.0387, "step": 3710 }, { "grad_norm": 0.43087902665138245, "learning_rate": 9.996074536094519e-05, "loss": 0.0413, "step": 3720 }, { "grad_norm": 0.4190708100795746, "learning_rate": 9.995964599232168e-05, "loss": 0.0438, "step": 3730 }, { "grad_norm": 0.45957276225090027, "learning_rate": 9.995853144727683e-05, "loss": 0.0444, "step": 3740 }, { "grad_norm": 0.3882060945034027, "learning_rate": 9.99574017261492e-05, "loss": 0.0435, "step": 3750 }, { "grad_norm": 0.5157021284103394, "learning_rate": 9.995625682928198e-05, "loss": 0.0451, "step": 3760 }, { "grad_norm": 0.3569749593734741, "learning_rate": 9.995509675702295e-05, "loss": 0.0369, "step": 3770 }, { "grad_norm": 0.3924664258956909, "learning_rate": 9.995392150972451e-05, "loss": 0.0415, "step": 3780 }, { "grad_norm": 0.3872543275356293, "learning_rate": 9.995273108774366e-05, "loss": 0.0398, "step": 3790 }, { "grad_norm": 0.38596799969673157, "learning_rate": 9.995152549144205e-05, "loss": 0.0423, "step": 3800 }, { "grad_norm": 0.370155930519104, "learning_rate": 9.995030472118587e-05, "loss": 0.0476, "step": 3810 }, { "grad_norm": 0.40164947509765625, "learning_rate": 9.9949068777346e-05, "loss": 0.0489, "step": 3820 }, { "grad_norm": 0.38970237970352173, "learning_rate": 9.994781766029786e-05, "loss": 0.0399, "step": 3830 }, { "grad_norm": 0.3389398157596588, "learning_rate": 9.994655137042151e-05, "loss": 0.0414, "step": 3840 }, { "grad_norm": 0.37786102294921875, "learning_rate": 9.99452699081016e-05, "loss": 0.044, "step": 3850 }, { "grad_norm": 0.34116846323013306, "learning_rate": 9.994397327372743e-05, "loss": 0.0427, "step": 3860 }, { "grad_norm": 0.5402582883834839, "learning_rate": 9.994266146769286e-05, "loss": 0.0376, "step": 3870 }, { "grad_norm": 0.4346312880516052, "learning_rate": 9.994133449039642e-05, "loss": 0.0398, "step": 3880 }, { "grad_norm": 0.39513513445854187, "learning_rate": 9.993999234224118e-05, "loss": 0.0441, "step": 3890 }, { "grad_norm": 0.3528674840927124, "learning_rate": 9.993863502363485e-05, "loss": 0.0486, "step": 3900 }, { "grad_norm": 0.4843061864376068, "learning_rate": 9.993726253498976e-05, "loss": 0.0452, "step": 3910 }, { "grad_norm": 0.38619351387023926, "learning_rate": 9.993587487672282e-05, "loss": 0.0491, "step": 3920 }, { "grad_norm": 0.4367417097091675, "learning_rate": 9.993447204925558e-05, "loss": 0.0392, "step": 3930 }, { "grad_norm": 0.5143662095069885, "learning_rate": 9.993305405301416e-05, "loss": 0.0432, "step": 3940 }, { "grad_norm": 0.37914103269577026, "learning_rate": 9.993162088842935e-05, "loss": 0.0414, "step": 3950 }, { "grad_norm": 0.4026878774166107, "learning_rate": 9.993017255593646e-05, "loss": 0.0404, "step": 3960 }, { "grad_norm": 0.353892982006073, "learning_rate": 9.992870905597548e-05, "loss": 0.0397, "step": 3970 }, { "grad_norm": 0.5037928223609924, "learning_rate": 9.9927230388991e-05, "loss": 0.0392, "step": 3980 }, { "grad_norm": 0.4537335932254791, "learning_rate": 9.992573655543215e-05, "loss": 0.0408, "step": 3990 }, { "grad_norm": 0.38810259103775024, "learning_rate": 9.992422755575277e-05, "loss": 0.045, "step": 4000 }, { "grad_norm": 0.4087066948413849, "learning_rate": 9.992270339041123e-05, "loss": 0.0393, "step": 4010 }, { "grad_norm": 0.3882775902748108, "learning_rate": 9.992116405987053e-05, "loss": 0.0377, "step": 4020 }, { "grad_norm": 0.4472566545009613, "learning_rate": 9.991960956459828e-05, "loss": 0.0432, "step": 4030 }, { "grad_norm": 0.49837973713874817, "learning_rate": 9.991803990506669e-05, "loss": 0.0384, "step": 4040 }, { "grad_norm": 0.42216408252716064, "learning_rate": 9.991645508175258e-05, "loss": 0.0421, "step": 4050 }, { "grad_norm": 0.4121020436286926, "learning_rate": 9.99148550951374e-05, "loss": 0.0419, "step": 4060 }, { "grad_norm": 0.45708101987838745, "learning_rate": 9.991323994570716e-05, "loss": 0.0375, "step": 4070 }, { "grad_norm": 0.48330315947532654, "learning_rate": 9.99116096339525e-05, "loss": 0.0413, "step": 4080 }, { "grad_norm": 0.45478254556655884, "learning_rate": 9.990996416036869e-05, "loss": 0.0402, "step": 4090 }, { "grad_norm": 0.37324628233909607, "learning_rate": 9.990830352545555e-05, "loss": 0.0395, "step": 4100 }, { "grad_norm": 0.38791781663894653, "learning_rate": 9.990662772971756e-05, "loss": 0.0445, "step": 4110 }, { "grad_norm": 0.35102948546409607, "learning_rate": 9.990493677366376e-05, "loss": 0.0442, "step": 4120 }, { "grad_norm": 0.42073872685432434, "learning_rate": 9.990323065780786e-05, "loss": 0.0461, "step": 4130 }, { "grad_norm": 0.39710474014282227, "learning_rate": 9.990150938266808e-05, "loss": 0.0399, "step": 4140 }, { "grad_norm": 0.3957497775554657, "learning_rate": 9.989977294876733e-05, "loss": 0.0426, "step": 4150 }, { "grad_norm": 0.3434363007545471, "learning_rate": 9.989802135663308e-05, "loss": 0.0399, "step": 4160 }, { "grad_norm": 0.43542248010635376, "learning_rate": 9.989625460679743e-05, "loss": 0.0442, "step": 4170 }, { "grad_norm": 0.4362136423587799, "learning_rate": 9.989447269979706e-05, "loss": 0.0429, "step": 4180 }, { "grad_norm": 0.4383696913719177, "learning_rate": 9.989267563617328e-05, "loss": 0.0441, "step": 4190 }, { "grad_norm": 0.4154229462146759, "learning_rate": 9.989086341647198e-05, "loss": 0.0373, "step": 4200 }, { "grad_norm": 0.4542798697948456, "learning_rate": 9.988903604124366e-05, "loss": 0.0371, "step": 4210 }, { "grad_norm": 0.4139091670513153, "learning_rate": 9.988719351104343e-05, "loss": 0.0368, "step": 4220 }, { "grad_norm": 0.36172670125961304, "learning_rate": 9.9885335826431e-05, "loss": 0.0386, "step": 4230 }, { "grad_norm": 0.4487634003162384, "learning_rate": 9.988346298797071e-05, "loss": 0.0368, "step": 4240 }, { "grad_norm": 0.3658663034439087, "learning_rate": 9.988157499623146e-05, "loss": 0.0398, "step": 4250 }, { "grad_norm": 0.29947733879089355, "learning_rate": 9.987967185178677e-05, "loss": 0.0392, "step": 4260 }, { "grad_norm": 0.4224274754524231, "learning_rate": 9.987775355521476e-05, "loss": 0.0416, "step": 4270 }, { "grad_norm": 0.47632452845573425, "learning_rate": 9.987582010709817e-05, "loss": 0.0423, "step": 4280 }, { "grad_norm": 0.3599952757358551, "learning_rate": 9.987387150802431e-05, "loss": 0.038, "step": 4290 }, { "grad_norm": 0.3750114142894745, "learning_rate": 9.987190775858517e-05, "loss": 0.0342, "step": 4300 }, { "grad_norm": 0.4251334071159363, "learning_rate": 9.98699288593772e-05, "loss": 0.0375, "step": 4310 }, { "grad_norm": 0.5118726491928101, "learning_rate": 9.986793481100161e-05, "loss": 0.0421, "step": 4320 }, { "grad_norm": 0.4013063609600067, "learning_rate": 9.986592561406412e-05, "loss": 0.0387, "step": 4330 }, { "grad_norm": 0.39799079298973083, "learning_rate": 9.986390126917503e-05, "loss": 0.0418, "step": 4340 }, { "grad_norm": 0.48451220989227295, "learning_rate": 9.986186177694933e-05, "loss": 0.0472, "step": 4350 }, { "grad_norm": 0.39511391520500183, "learning_rate": 9.985980713800656e-05, "loss": 0.0371, "step": 4360 }, { "grad_norm": 0.4370124936103821, "learning_rate": 9.985773735297084e-05, "loss": 0.0409, "step": 4370 }, { "grad_norm": 0.35660800337791443, "learning_rate": 9.985565242247092e-05, "loss": 0.042, "step": 4380 }, { "grad_norm": 0.4307158589363098, "learning_rate": 9.985355234714016e-05, "loss": 0.0463, "step": 4390 }, { "grad_norm": 0.38520991802215576, "learning_rate": 9.985143712761652e-05, "loss": 0.0385, "step": 4400 }, { "grad_norm": 0.43505391478538513, "learning_rate": 9.984930676454252e-05, "loss": 0.0427, "step": 4410 }, { "grad_norm": 0.37568730115890503, "learning_rate": 9.984716125856532e-05, "loss": 0.0441, "step": 4420 }, { "grad_norm": 0.32298043370246887, "learning_rate": 9.984500061033667e-05, "loss": 0.0388, "step": 4430 }, { "grad_norm": 0.37544673681259155, "learning_rate": 9.984282482051293e-05, "loss": 0.0398, "step": 4440 }, { "grad_norm": 0.39842307567596436, "learning_rate": 9.9840633889755e-05, "loss": 0.0415, "step": 4450 }, { "grad_norm": 0.40862372517585754, "learning_rate": 9.983842781872848e-05, "loss": 0.039, "step": 4460 }, { "grad_norm": 0.32705268263816833, "learning_rate": 9.98362066081035e-05, "loss": 0.0359, "step": 4470 }, { "grad_norm": 0.38580596446990967, "learning_rate": 9.983397025855479e-05, "loss": 0.0381, "step": 4480 }, { "grad_norm": 0.39721474051475525, "learning_rate": 9.983171877076171e-05, "loss": 0.0434, "step": 4490 }, { "grad_norm": 0.4050196707248688, "learning_rate": 9.98294521454082e-05, "loss": 0.0387, "step": 4500 }, { "grad_norm": 0.44031673669815063, "learning_rate": 9.98271703831828e-05, "loss": 0.038, "step": 4510 }, { "grad_norm": 0.43357351422309875, "learning_rate": 9.982487348477865e-05, "loss": 0.04, "step": 4520 }, { "grad_norm": 0.36262449622154236, "learning_rate": 9.982256145089347e-05, "loss": 0.0439, "step": 4530 }, { "grad_norm": 0.3927157521247864, "learning_rate": 9.982023428222962e-05, "loss": 0.04, "step": 4540 }, { "grad_norm": 0.3483410179615021, "learning_rate": 9.981789197949403e-05, "loss": 0.0341, "step": 4550 }, { "grad_norm": 0.32339203357696533, "learning_rate": 9.98155345433982e-05, "loss": 0.0379, "step": 4560 }, { "grad_norm": 0.37636318802833557, "learning_rate": 9.981316197465831e-05, "loss": 0.0355, "step": 4570 }, { "grad_norm": 0.32238516211509705, "learning_rate": 9.981077427399504e-05, "loss": 0.0428, "step": 4580 }, { "grad_norm": 0.3869650363922119, "learning_rate": 9.980837144213371e-05, "loss": 0.044, "step": 4590 }, { "grad_norm": 0.5506515502929688, "learning_rate": 9.980595347980426e-05, "loss": 0.0392, "step": 4600 }, { "grad_norm": 0.3600251376628876, "learning_rate": 9.980352038774119e-05, "loss": 0.0422, "step": 4610 }, { "grad_norm": 0.3580052852630615, "learning_rate": 9.98010721666836e-05, "loss": 0.0394, "step": 4620 }, { "grad_norm": 0.3095916509628296, "learning_rate": 9.979860881737523e-05, "loss": 0.0352, "step": 4630 }, { "grad_norm": 0.38373684883117676, "learning_rate": 9.979613034056434e-05, "loss": 0.0385, "step": 4640 }, { "grad_norm": 0.2672097980976105, "learning_rate": 9.979363673700386e-05, "loss": 0.0351, "step": 4650 }, { "grad_norm": 0.3109363317489624, "learning_rate": 9.979112800745124e-05, "loss": 0.0372, "step": 4660 }, { "grad_norm": 0.37667223811149597, "learning_rate": 9.978860415266861e-05, "loss": 0.0365, "step": 4670 }, { "grad_norm": 0.43733450770378113, "learning_rate": 9.978606517342262e-05, "loss": 0.034, "step": 4680 }, { "grad_norm": 0.36001718044281006, "learning_rate": 9.978351107048456e-05, "loss": 0.0375, "step": 4690 }, { "grad_norm": 0.49061185121536255, "learning_rate": 9.978094184463029e-05, "loss": 0.046, "step": 4700 }, { "grad_norm": 0.4112873673439026, "learning_rate": 9.977835749664029e-05, "loss": 0.0398, "step": 4710 }, { "grad_norm": 0.48449233174324036, "learning_rate": 9.97757580272996e-05, "loss": 0.0352, "step": 4720 }, { "grad_norm": 0.2888171672821045, "learning_rate": 9.977314343739786e-05, "loss": 0.0359, "step": 4730 }, { "grad_norm": 0.30549877882003784, "learning_rate": 9.977051372772934e-05, "loss": 0.038, "step": 4740 }, { "grad_norm": 0.3090793788433075, "learning_rate": 9.976786889909286e-05, "loss": 0.0361, "step": 4750 }, { "grad_norm": 0.327150821685791, "learning_rate": 9.976520895229185e-05, "loss": 0.0392, "step": 4760 }, { "grad_norm": 0.34092170000076294, "learning_rate": 9.976253388813433e-05, "loss": 0.0436, "step": 4770 }, { "grad_norm": 0.3138181269168854, "learning_rate": 9.975984370743293e-05, "loss": 0.0372, "step": 4780 }, { "grad_norm": 0.3465980589389801, "learning_rate": 9.975713841100485e-05, "loss": 0.0393, "step": 4790 }, { "grad_norm": 0.31443020701408386, "learning_rate": 9.975441799967187e-05, "loss": 0.0371, "step": 4800 }, { "grad_norm": 0.3832765221595764, "learning_rate": 9.975168247426039e-05, "loss": 0.0337, "step": 4810 }, { "grad_norm": 0.49052488803863525, "learning_rate": 9.974893183560139e-05, "loss": 0.0368, "step": 4820 }, { "grad_norm": 0.42763978242874146, "learning_rate": 9.974616608453045e-05, "loss": 0.0344, "step": 4830 }, { "grad_norm": 0.27213430404663086, "learning_rate": 9.974338522188772e-05, "loss": 0.0359, "step": 4840 }, { "grad_norm": 0.34958770871162415, "learning_rate": 9.974058924851797e-05, "loss": 0.0362, "step": 4850 }, { "grad_norm": 0.3363771140575409, "learning_rate": 9.973777816527051e-05, "loss": 0.0349, "step": 4860 }, { "grad_norm": 0.3931511640548706, "learning_rate": 9.973495197299931e-05, "loss": 0.0404, "step": 4870 }, { "grad_norm": 0.3751601576805115, "learning_rate": 9.973211067256287e-05, "loss": 0.0374, "step": 4880 }, { "grad_norm": 0.4205666184425354, "learning_rate": 9.97292542648243e-05, "loss": 0.0384, "step": 4890 }, { "grad_norm": 0.2562606632709503, "learning_rate": 9.972638275065131e-05, "loss": 0.0361, "step": 4900 }, { "grad_norm": 0.31570568680763245, "learning_rate": 9.972349613091621e-05, "loss": 0.033, "step": 4910 }, { "grad_norm": 0.2972767651081085, "learning_rate": 9.972059440649584e-05, "loss": 0.0372, "step": 4920 }, { "grad_norm": 0.4406643211841583, "learning_rate": 9.971767757827168e-05, "loss": 0.0392, "step": 4930 }, { "grad_norm": 0.4187912046909332, "learning_rate": 9.971474564712982e-05, "loss": 0.0341, "step": 4940 }, { "grad_norm": 0.39057663083076477, "learning_rate": 9.971179861396084e-05, "loss": 0.0376, "step": 4950 }, { "grad_norm": 0.3816169202327728, "learning_rate": 9.970883647966003e-05, "loss": 0.0396, "step": 4960 }, { "grad_norm": 0.3552718460559845, "learning_rate": 9.970585924512717e-05, "loss": 0.0318, "step": 4970 }, { "grad_norm": 0.25796258449554443, "learning_rate": 9.970286691126669e-05, "loss": 0.0396, "step": 4980 }, { "grad_norm": 0.38078033924102783, "learning_rate": 9.969985947898756e-05, "loss": 0.0377, "step": 4990 }, { "grad_norm": 0.34983304142951965, "learning_rate": 9.969683694920337e-05, "loss": 0.0361, "step": 5000 }, { "grad_norm": 0.41366100311279297, "learning_rate": 9.969379932283228e-05, "loss": 0.0387, "step": 5010 }, { "grad_norm": 0.41857433319091797, "learning_rate": 9.969074660079704e-05, "loss": 0.0397, "step": 5020 }, { "grad_norm": 0.3234754800796509, "learning_rate": 9.968767878402501e-05, "loss": 0.0336, "step": 5030 }, { "grad_norm": 0.4001924693584442, "learning_rate": 9.968459587344808e-05, "loss": 0.0326, "step": 5040 }, { "grad_norm": 0.3915397524833679, "learning_rate": 9.968149787000278e-05, "loss": 0.0367, "step": 5050 }, { "grad_norm": 0.3013749420642853, "learning_rate": 9.967838477463018e-05, "loss": 0.039, "step": 5060 }, { "grad_norm": 0.29826802015304565, "learning_rate": 9.967525658827597e-05, "loss": 0.0349, "step": 5070 }, { "grad_norm": 0.3880145251750946, "learning_rate": 9.967211331189042e-05, "loss": 0.0359, "step": 5080 }, { "grad_norm": 0.30261820554733276, "learning_rate": 9.966895494642834e-05, "loss": 0.0377, "step": 5090 }, { "grad_norm": 0.3755296766757965, "learning_rate": 9.96657814928492e-05, "loss": 0.0382, "step": 5100 }, { "grad_norm": 0.40052416920661926, "learning_rate": 9.966259295211697e-05, "loss": 0.0386, "step": 5110 }, { "grad_norm": 0.3210751712322235, "learning_rate": 9.965938932520028e-05, "loss": 0.0337, "step": 5120 }, { "grad_norm": 0.4238247871398926, "learning_rate": 9.965617061307229e-05, "loss": 0.0349, "step": 5130 }, { "grad_norm": 0.3406037986278534, "learning_rate": 9.965293681671077e-05, "loss": 0.0366, "step": 5140 }, { "grad_norm": 0.3410295248031616, "learning_rate": 9.964968793709804e-05, "loss": 0.035, "step": 5150 }, { "grad_norm": 0.342833012342453, "learning_rate": 9.964642397522106e-05, "loss": 0.0351, "step": 5160 }, { "grad_norm": 0.3669990003108978, "learning_rate": 9.96431449320713e-05, "loss": 0.0348, "step": 5170 }, { "grad_norm": 0.3081105947494507, "learning_rate": 9.963985080864486e-05, "loss": 0.0326, "step": 5180 }, { "grad_norm": 0.328521728515625, "learning_rate": 9.96365416059424e-05, "loss": 0.0385, "step": 5190 }, { "grad_norm": 0.2727031111717224, "learning_rate": 9.963321732496919e-05, "loss": 0.0341, "step": 5200 }, { "grad_norm": 0.35559502243995667, "learning_rate": 9.962987796673506e-05, "loss": 0.0406, "step": 5210 }, { "grad_norm": 0.3656843304634094, "learning_rate": 9.962652353225438e-05, "loss": 0.0426, "step": 5220 }, { "grad_norm": 0.3952522277832031, "learning_rate": 9.962315402254619e-05, "loss": 0.0393, "step": 5230 }, { "grad_norm": 0.32062584161758423, "learning_rate": 9.9619769438634e-05, "loss": 0.033, "step": 5240 }, { "grad_norm": 0.3699914216995239, "learning_rate": 9.9616369781546e-05, "loss": 0.0368, "step": 5250 }, { "grad_norm": 0.4781888723373413, "learning_rate": 9.961295505231491e-05, "loss": 0.0374, "step": 5260 }, { "grad_norm": 0.3728789687156677, "learning_rate": 9.960952525197804e-05, "loss": 0.0381, "step": 5270 }, { "grad_norm": 0.3024848401546478, "learning_rate": 9.960608038157724e-05, "loss": 0.0369, "step": 5280 }, { "grad_norm": 0.32407811284065247, "learning_rate": 9.960262044215901e-05, "loss": 0.037, "step": 5290 }, { "grad_norm": 0.2653365135192871, "learning_rate": 9.959914543477435e-05, "loss": 0.0402, "step": 5300 }, { "grad_norm": 0.29656344652175903, "learning_rate": 9.959565536047892e-05, "loss": 0.0418, "step": 5310 }, { "grad_norm": 0.4239679276943207, "learning_rate": 9.959215022033288e-05, "loss": 0.0426, "step": 5320 }, { "grad_norm": 0.35119569301605225, "learning_rate": 9.9588630015401e-05, "loss": 0.0367, "step": 5330 }, { "grad_norm": 0.38578638434410095, "learning_rate": 9.958509474675264e-05, "loss": 0.0414, "step": 5340 }, { "grad_norm": 0.32657355070114136, "learning_rate": 9.958154441546171e-05, "loss": 0.0407, "step": 5350 }, { "grad_norm": 0.4159374535083771, "learning_rate": 9.957797902260673e-05, "loss": 0.0339, "step": 5360 }, { "grad_norm": 0.3693059980869293, "learning_rate": 9.957439856927073e-05, "loss": 0.0354, "step": 5370 }, { "grad_norm": 0.2724316418170929, "learning_rate": 9.957080305654139e-05, "loss": 0.0358, "step": 5380 }, { "grad_norm": 0.35930395126342773, "learning_rate": 9.956719248551092e-05, "loss": 0.0358, "step": 5390 }, { "grad_norm": 0.34327206015586853, "learning_rate": 9.956356685727612e-05, "loss": 0.0317, "step": 5400 }, { "grad_norm": 0.3376636505126953, "learning_rate": 9.955992617293836e-05, "loss": 0.0351, "step": 5410 }, { "grad_norm": 0.34274038672447205, "learning_rate": 9.955627043360358e-05, "loss": 0.0329, "step": 5420 }, { "grad_norm": 0.2994603216648102, "learning_rate": 9.955259964038231e-05, "loss": 0.0365, "step": 5430 }, { "grad_norm": 0.2787672281265259, "learning_rate": 9.954891379438962e-05, "loss": 0.0358, "step": 5440 }, { "grad_norm": 0.37913763523101807, "learning_rate": 9.954521289674519e-05, "loss": 0.0353, "step": 5450 }, { "grad_norm": 0.3577515780925751, "learning_rate": 9.954149694857325e-05, "loss": 0.0338, "step": 5460 }, { "grad_norm": 0.2928103506565094, "learning_rate": 9.953776595100258e-05, "loss": 0.0326, "step": 5470 }, { "grad_norm": 0.4684181213378906, "learning_rate": 9.95340199051666e-05, "loss": 0.036, "step": 5480 }, { "grad_norm": 0.35733669996261597, "learning_rate": 9.953025881220325e-05, "loss": 0.0373, "step": 5490 }, { "grad_norm": 0.31079167127609253, "learning_rate": 9.952648267325504e-05, "loss": 0.0319, "step": 5500 }, { "grad_norm": 0.3933160901069641, "learning_rate": 9.952269148946905e-05, "loss": 0.0386, "step": 5510 }, { "grad_norm": 0.38780301809310913, "learning_rate": 9.951888526199697e-05, "loss": 0.0375, "step": 5520 }, { "grad_norm": 0.43761903047561646, "learning_rate": 9.951506399199501e-05, "loss": 0.0356, "step": 5530 }, { "grad_norm": 0.36253753304481506, "learning_rate": 9.951122768062399e-05, "loss": 0.0368, "step": 5540 }, { "grad_norm": 0.40080010890960693, "learning_rate": 9.950737632904927e-05, "loss": 0.0343, "step": 5550 }, { "grad_norm": 0.3452008068561554, "learning_rate": 9.950350993844077e-05, "loss": 0.0355, "step": 5560 }, { "grad_norm": 0.40946412086486816, "learning_rate": 9.949962850997303e-05, "loss": 0.0354, "step": 5570 }, { "grad_norm": 0.3044070899486542, "learning_rate": 9.949573204482512e-05, "loss": 0.0327, "step": 5580 }, { "grad_norm": 0.28367117047309875, "learning_rate": 9.949182054418064e-05, "loss": 0.0357, "step": 5590 }, { "grad_norm": 0.32383596897125244, "learning_rate": 9.948789400922787e-05, "loss": 0.0325, "step": 5600 }, { "grad_norm": 0.28029194474220276, "learning_rate": 9.948395244115953e-05, "loss": 0.0342, "step": 5610 }, { "grad_norm": 0.3512972891330719, "learning_rate": 9.9479995841173e-05, "loss": 0.0343, "step": 5620 }, { "grad_norm": 0.3724318742752075, "learning_rate": 9.947602421047017e-05, "loss": 0.0316, "step": 5630 }, { "grad_norm": 0.3316921293735504, "learning_rate": 9.947203755025753e-05, "loss": 0.0322, "step": 5640 }, { "grad_norm": 0.27900469303131104, "learning_rate": 9.946803586174611e-05, "loss": 0.0306, "step": 5650 }, { "grad_norm": 0.37775424122810364, "learning_rate": 9.946401914615151e-05, "loss": 0.0396, "step": 5660 }, { "grad_norm": 0.4283809959888458, "learning_rate": 9.945998740469394e-05, "loss": 0.032, "step": 5670 }, { "grad_norm": 0.34532442688941956, "learning_rate": 9.945594063859809e-05, "loss": 0.0362, "step": 5680 }, { "grad_norm": 0.33678585290908813, "learning_rate": 9.94518788490933e-05, "loss": 0.0311, "step": 5690 }, { "grad_norm": 0.4289318025112152, "learning_rate": 9.944780203741341e-05, "loss": 0.034, "step": 5700 }, { "grad_norm": 0.43957948684692383, "learning_rate": 9.944371020479686e-05, "loss": 0.0336, "step": 5710 }, { "grad_norm": 0.32831934094429016, "learning_rate": 9.943960335248662e-05, "loss": 0.0369, "step": 5720 }, { "grad_norm": 0.2606014609336853, "learning_rate": 9.943548148173027e-05, "loss": 0.0341, "step": 5730 }, { "grad_norm": 0.2588396668434143, "learning_rate": 9.943134459377992e-05, "loss": 0.0366, "step": 5740 }, { "grad_norm": 0.32817065715789795, "learning_rate": 9.942719268989222e-05, "loss": 0.0343, "step": 5750 }, { "grad_norm": 0.3155205249786377, "learning_rate": 9.942302577132844e-05, "loss": 0.0307, "step": 5760 }, { "grad_norm": 0.3319181203842163, "learning_rate": 9.941884383935438e-05, "loss": 0.0352, "step": 5770 }, { "grad_norm": 0.2803318202495575, "learning_rate": 9.941464689524039e-05, "loss": 0.0342, "step": 5780 }, { "grad_norm": 0.37721824645996094, "learning_rate": 9.941043494026139e-05, "loss": 0.0353, "step": 5790 }, { "grad_norm": 0.298725962638855, "learning_rate": 9.940620797569685e-05, "loss": 0.033, "step": 5800 }, { "grad_norm": 0.3156641721725464, "learning_rate": 9.940196600283082e-05, "loss": 0.0307, "step": 5810 }, { "grad_norm": 0.2644942104816437, "learning_rate": 9.939770902295192e-05, "loss": 0.0349, "step": 5820 }, { "grad_norm": 0.3563120365142822, "learning_rate": 9.939343703735329e-05, "loss": 0.0345, "step": 5830 }, { "grad_norm": 0.3468076288700104, "learning_rate": 9.938915004733264e-05, "loss": 0.0356, "step": 5840 }, { "grad_norm": 0.3416767418384552, "learning_rate": 9.938484805419224e-05, "loss": 0.0357, "step": 5850 }, { "grad_norm": 0.2277171015739441, "learning_rate": 9.938053105923894e-05, "loss": 0.0377, "step": 5860 }, { "grad_norm": 0.38225865364074707, "learning_rate": 9.937619906378413e-05, "loss": 0.0376, "step": 5870 }, { "grad_norm": 0.35922983288764954, "learning_rate": 9.937185206914374e-05, "loss": 0.0357, "step": 5880 }, { "grad_norm": 0.2836850583553314, "learning_rate": 9.936749007663829e-05, "loss": 0.0319, "step": 5890 }, { "grad_norm": 0.42863908410072327, "learning_rate": 9.93631130875928e-05, "loss": 0.0359, "step": 5900 }, { "grad_norm": 0.4289265275001526, "learning_rate": 9.935872110333692e-05, "loss": 0.0365, "step": 5910 }, { "grad_norm": 0.3604564964771271, "learning_rate": 9.935431412520484e-05, "loss": 0.0357, "step": 5920 }, { "grad_norm": 0.34228798747062683, "learning_rate": 9.934989215453523e-05, "loss": 0.0307, "step": 5930 }, { "grad_norm": 0.2853524684906006, "learning_rate": 9.934545519267139e-05, "loss": 0.0356, "step": 5940 }, { "grad_norm": 0.28032422065734863, "learning_rate": 9.934100324096117e-05, "loss": 0.0307, "step": 5950 }, { "grad_norm": 0.3661225140094757, "learning_rate": 9.933653630075692e-05, "loss": 0.0318, "step": 5960 }, { "grad_norm": 0.3960592746734619, "learning_rate": 9.93320543734156e-05, "loss": 0.035, "step": 5970 }, { "grad_norm": 0.23115798830986023, "learning_rate": 9.932755746029871e-05, "loss": 0.0315, "step": 5980 }, { "grad_norm": 0.34787777066230774, "learning_rate": 9.932304556277228e-05, "loss": 0.0338, "step": 5990 }, { "grad_norm": 0.4230327010154724, "learning_rate": 9.93185186822069e-05, "loss": 0.0373, "step": 6000 }, { "grad_norm": 0.30375298857688904, "learning_rate": 9.931397681997773e-05, "loss": 0.031, "step": 6010 }, { "grad_norm": 0.3451072573661804, "learning_rate": 9.930941997746446e-05, "loss": 0.032, "step": 6020 }, { "grad_norm": 0.31047409772872925, "learning_rate": 9.930484815605134e-05, "loss": 0.0312, "step": 6030 }, { "grad_norm": 0.2572951018810272, "learning_rate": 9.930026135712717e-05, "loss": 0.0317, "step": 6040 }, { "grad_norm": 0.33459600806236267, "learning_rate": 9.92956595820853e-05, "loss": 0.0325, "step": 6050 }, { "grad_norm": 0.3506987392902374, "learning_rate": 9.929104283232362e-05, "loss": 0.0307, "step": 6060 }, { "grad_norm": 0.33825263381004333, "learning_rate": 9.92864111092446e-05, "loss": 0.0322, "step": 6070 }, { "grad_norm": 0.37285566329956055, "learning_rate": 9.92817644142552e-05, "loss": 0.0351, "step": 6080 }, { "grad_norm": 0.31118044257164, "learning_rate": 9.927710274876698e-05, "loss": 0.0294, "step": 6090 }, { "grad_norm": 0.4206605851650238, "learning_rate": 9.927242611419603e-05, "loss": 0.0382, "step": 6100 }, { "grad_norm": 0.35129591822624207, "learning_rate": 9.926773451196301e-05, "loss": 0.0332, "step": 6110 }, { "grad_norm": 0.2999260127544403, "learning_rate": 9.926302794349306e-05, "loss": 0.0391, "step": 6120 }, { "grad_norm": 0.3350665867328644, "learning_rate": 9.925830641021594e-05, "loss": 0.033, "step": 6130 }, { "grad_norm": 0.3440190553665161, "learning_rate": 9.925356991356593e-05, "loss": 0.0328, "step": 6140 }, { "grad_norm": 0.30526918172836304, "learning_rate": 9.924881845498184e-05, "loss": 0.036, "step": 6150 }, { "grad_norm": 0.28284093737602234, "learning_rate": 9.924405203590705e-05, "loss": 0.0317, "step": 6160 }, { "grad_norm": 0.31751590967178345, "learning_rate": 9.923927065778946e-05, "loss": 0.036, "step": 6170 }, { "grad_norm": 0.4080774486064911, "learning_rate": 9.923447432208154e-05, "loss": 0.0331, "step": 6180 }, { "grad_norm": 0.3953138589859009, "learning_rate": 9.922966303024027e-05, "loss": 0.0306, "step": 6190 }, { "grad_norm": 0.30875417590141296, "learning_rate": 9.922483678372721e-05, "loss": 0.0336, "step": 6200 }, { "grad_norm": 0.38706648349761963, "learning_rate": 9.921999558400845e-05, "loss": 0.034, "step": 6210 }, { "grad_norm": 0.2922177016735077, "learning_rate": 9.92151394325546e-05, "loss": 0.0361, "step": 6220 }, { "grad_norm": 0.29725679755210876, "learning_rate": 9.921026833084084e-05, "loss": 0.0343, "step": 6230 }, { "grad_norm": 0.3925439715385437, "learning_rate": 9.920538228034689e-05, "loss": 0.0327, "step": 6240 }, { "grad_norm": 0.31094443798065186, "learning_rate": 9.920048128255699e-05, "loss": 0.0357, "step": 6250 }, { "grad_norm": 0.40422916412353516, "learning_rate": 9.919556533895995e-05, "loss": 0.0394, "step": 6260 }, { "grad_norm": 0.2863508462905884, "learning_rate": 9.919063445104907e-05, "loss": 0.0309, "step": 6270 }, { "grad_norm": 0.3202967643737793, "learning_rate": 9.918568862032227e-05, "loss": 0.0336, "step": 6280 }, { "grad_norm": 0.30895939469337463, "learning_rate": 9.918072784828194e-05, "loss": 0.0311, "step": 6290 }, { "grad_norm": 0.31117358803749084, "learning_rate": 9.917575213643501e-05, "loss": 0.0372, "step": 6300 }, { "grad_norm": 0.3306240737438202, "learning_rate": 9.917076148629302e-05, "loss": 0.0345, "step": 6310 }, { "grad_norm": 0.33253446221351624, "learning_rate": 9.916575589937196e-05, "loss": 0.0326, "step": 6320 }, { "grad_norm": 0.3375470042228699, "learning_rate": 9.916073537719239e-05, "loss": 0.0339, "step": 6330 }, { "grad_norm": 0.2498762458562851, "learning_rate": 9.915569992127944e-05, "loss": 0.0339, "step": 6340 }, { "grad_norm": 0.34006550908088684, "learning_rate": 9.915064953316273e-05, "loss": 0.0351, "step": 6350 }, { "grad_norm": 0.37306517362594604, "learning_rate": 9.914558421437645e-05, "loss": 0.0376, "step": 6360 }, { "grad_norm": 0.30782467126846313, "learning_rate": 9.914050396645929e-05, "loss": 0.0354, "step": 6370 }, { "grad_norm": 0.3211022615432739, "learning_rate": 9.913540879095452e-05, "loss": 0.0376, "step": 6380 }, { "grad_norm": 0.3714943528175354, "learning_rate": 9.913029868940987e-05, "loss": 0.0372, "step": 6390 }, { "grad_norm": 0.3647646903991699, "learning_rate": 9.912517366337772e-05, "loss": 0.0317, "step": 6400 }, { "grad_norm": 0.3446623682975769, "learning_rate": 9.912003371441487e-05, "loss": 0.0375, "step": 6410 }, { "grad_norm": 0.3416461646556854, "learning_rate": 9.911487884408271e-05, "loss": 0.0355, "step": 6420 }, { "grad_norm": 0.3306805491447449, "learning_rate": 9.910970905394719e-05, "loss": 0.0319, "step": 6430 }, { "grad_norm": 0.25647205114364624, "learning_rate": 9.91045243455787e-05, "loss": 0.0327, "step": 6440 }, { "grad_norm": 0.27095815539360046, "learning_rate": 9.909932472055225e-05, "loss": 0.031, "step": 6450 }, { "grad_norm": 0.338381290435791, "learning_rate": 9.909411018044734e-05, "loss": 0.0313, "step": 6460 }, { "grad_norm": 0.3162238895893097, "learning_rate": 9.908888072684802e-05, "loss": 0.0306, "step": 6470 }, { "grad_norm": 0.315406858921051, "learning_rate": 9.908363636134285e-05, "loss": 0.0353, "step": 6480 }, { "grad_norm": 0.342680424451828, "learning_rate": 9.907837708552493e-05, "loss": 0.0345, "step": 6490 }, { "grad_norm": 0.26771122217178345, "learning_rate": 9.90731029009919e-05, "loss": 0.0283, "step": 6500 }, { "grad_norm": 0.3400920629501343, "learning_rate": 9.906781380934589e-05, "loss": 0.0312, "step": 6510 }, { "grad_norm": 0.24385222792625427, "learning_rate": 9.906250981219362e-05, "loss": 0.0287, "step": 6520 }, { "grad_norm": 0.27690672874450684, "learning_rate": 9.905719091114628e-05, "loss": 0.0319, "step": 6530 }, { "grad_norm": 0.40467369556427, "learning_rate": 9.905185710781964e-05, "loss": 0.0336, "step": 6540 }, { "grad_norm": 0.2998357117176056, "learning_rate": 9.904650840383392e-05, "loss": 0.0327, "step": 6550 }, { "grad_norm": 0.32732391357421875, "learning_rate": 9.904114480081397e-05, "loss": 0.0358, "step": 6560 }, { "grad_norm": 0.2689487040042877, "learning_rate": 9.903576630038906e-05, "loss": 0.0347, "step": 6570 }, { "grad_norm": 0.3103993237018585, "learning_rate": 9.903037290419309e-05, "loss": 0.0301, "step": 6580 }, { "grad_norm": 0.2722451686859131, "learning_rate": 9.902496461386439e-05, "loss": 0.0327, "step": 6590 }, { "grad_norm": 0.2945043742656708, "learning_rate": 9.901954143104588e-05, "loss": 0.0368, "step": 6600 }, { "grad_norm": 0.3091038763523102, "learning_rate": 9.901410335738496e-05, "loss": 0.036, "step": 6610 }, { "grad_norm": 0.3637339770793915, "learning_rate": 9.900865039453358e-05, "loss": 0.0278, "step": 6620 }, { "grad_norm": 0.28492629528045654, "learning_rate": 9.900318254414821e-05, "loss": 0.0377, "step": 6630 }, { "grad_norm": 0.32217246294021606, "learning_rate": 9.899769980788985e-05, "loss": 0.0347, "step": 6640 }, { "grad_norm": 0.31840816140174866, "learning_rate": 9.899220218742398e-05, "loss": 0.0321, "step": 6650 }, { "grad_norm": 0.3371598422527313, "learning_rate": 9.898668968442066e-05, "loss": 0.0358, "step": 6660 }, { "grad_norm": 0.21782495081424713, "learning_rate": 9.898116230055443e-05, "loss": 0.0357, "step": 6670 }, { "grad_norm": 0.2916165888309479, "learning_rate": 9.897562003750437e-05, "loss": 0.0333, "step": 6680 }, { "grad_norm": 0.36927542090415955, "learning_rate": 9.897006289695407e-05, "loss": 0.0328, "step": 6690 }, { "grad_norm": 0.26930490136146545, "learning_rate": 9.896449088059164e-05, "loss": 0.0341, "step": 6700 }, { "grad_norm": 0.29737988114356995, "learning_rate": 9.89589039901097e-05, "loss": 0.0343, "step": 6710 }, { "grad_norm": 0.3037213683128357, "learning_rate": 9.895330222720542e-05, "loss": 0.0315, "step": 6720 }, { "grad_norm": 0.3183363080024719, "learning_rate": 9.894768559358047e-05, "loss": 0.0297, "step": 6730 }, { "grad_norm": 0.37189438939094543, "learning_rate": 9.894205409094101e-05, "loss": 0.0322, "step": 6740 }, { "grad_norm": 0.3187391459941864, "learning_rate": 9.893640772099777e-05, "loss": 0.0285, "step": 6750 }, { "grad_norm": 0.351421594619751, "learning_rate": 9.893074648546595e-05, "loss": 0.029, "step": 6760 }, { "grad_norm": 0.3324945569038391, "learning_rate": 9.892507038606528e-05, "loss": 0.0287, "step": 6770 }, { "grad_norm": 0.31905561685562134, "learning_rate": 9.891937942452003e-05, "loss": 0.0309, "step": 6780 }, { "grad_norm": 0.2676524817943573, "learning_rate": 9.891367360255895e-05, "loss": 0.0312, "step": 6790 }, { "grad_norm": 0.27930909395217896, "learning_rate": 9.890795292191532e-05, "loss": 0.0309, "step": 6800 }, { "grad_norm": 0.33420509099960327, "learning_rate": 9.890221738432694e-05, "loss": 0.0308, "step": 6810 }, { "grad_norm": 0.26467713713645935, "learning_rate": 9.88964669915361e-05, "loss": 0.0321, "step": 6820 }, { "grad_norm": 0.21701383590698242, "learning_rate": 9.889070174528963e-05, "loss": 0.0288, "step": 6830 }, { "grad_norm": 0.2978528141975403, "learning_rate": 9.888492164733883e-05, "loss": 0.0306, "step": 6840 }, { "grad_norm": 0.360443115234375, "learning_rate": 9.88791266994396e-05, "loss": 0.0351, "step": 6850 }, { "grad_norm": 0.3463997542858124, "learning_rate": 9.887331690335223e-05, "loss": 0.036, "step": 6860 }, { "grad_norm": 0.2631095051765442, "learning_rate": 9.886749226084163e-05, "loss": 0.0283, "step": 6870 }, { "grad_norm": 0.3081863224506378, "learning_rate": 9.886165277367714e-05, "loss": 0.0342, "step": 6880 }, { "grad_norm": 0.35896411538124084, "learning_rate": 9.885579844363265e-05, "loss": 0.0279, "step": 6890 }, { "grad_norm": 0.286037415266037, "learning_rate": 9.884992927248656e-05, "loss": 0.0284, "step": 6900 }, { "grad_norm": 0.3147907555103302, "learning_rate": 9.884404526202178e-05, "loss": 0.0282, "step": 6910 }, { "grad_norm": 0.34657639265060425, "learning_rate": 9.883814641402568e-05, "loss": 0.0295, "step": 6920 }, { "grad_norm": 0.3820229768753052, "learning_rate": 9.88322327302902e-05, "loss": 0.0313, "step": 6930 }, { "grad_norm": 0.2871992588043213, "learning_rate": 9.882630421261176e-05, "loss": 0.0279, "step": 6940 }, { "grad_norm": 0.2948105037212372, "learning_rate": 9.88203608627913e-05, "loss": 0.0301, "step": 6950 }, { "grad_norm": 0.30400049686431885, "learning_rate": 9.881440268263422e-05, "loss": 0.0323, "step": 6960 }, { "grad_norm": 0.22829778492450714, "learning_rate": 9.880842967395048e-05, "loss": 0.028, "step": 6970 }, { "grad_norm": 0.21446596086025238, "learning_rate": 9.880244183855452e-05, "loss": 0.0315, "step": 6980 }, { "grad_norm": 0.2835782766342163, "learning_rate": 9.879643917826527e-05, "loss": 0.0297, "step": 6990 }, { "grad_norm": 0.2574186623096466, "learning_rate": 9.87904216949062e-05, "loss": 0.0327, "step": 7000 }, { "grad_norm": 0.32211560010910034, "learning_rate": 9.878438939030526e-05, "loss": 0.0342, "step": 7010 }, { "grad_norm": 0.37616288661956787, "learning_rate": 9.877834226629489e-05, "loss": 0.033, "step": 7020 }, { "grad_norm": 0.3251686096191406, "learning_rate": 9.877228032471206e-05, "loss": 0.0274, "step": 7030 }, { "grad_norm": 0.35917752981185913, "learning_rate": 9.876620356739823e-05, "loss": 0.0289, "step": 7040 }, { "grad_norm": 0.33551594614982605, "learning_rate": 9.876011199619935e-05, "loss": 0.0297, "step": 7050 }, { "grad_norm": 0.32100170850753784, "learning_rate": 9.875400561296589e-05, "loss": 0.0314, "step": 7060 }, { "grad_norm": 0.39247632026672363, "learning_rate": 9.874788441955278e-05, "loss": 0.0292, "step": 7070 }, { "grad_norm": 0.25927114486694336, "learning_rate": 9.874174841781951e-05, "loss": 0.0301, "step": 7080 }, { "grad_norm": 0.26594218611717224, "learning_rate": 9.873559760963003e-05, "loss": 0.0298, "step": 7090 }, { "grad_norm": 0.2310691475868225, "learning_rate": 9.872943199685278e-05, "loss": 0.0258, "step": 7100 }, { "grad_norm": 0.3240968883037567, "learning_rate": 9.872325158136071e-05, "loss": 0.0302, "step": 7110 }, { "grad_norm": 0.24258121848106384, "learning_rate": 9.871705636503128e-05, "loss": 0.03, "step": 7120 }, { "grad_norm": 0.31522077322006226, "learning_rate": 9.871084634974641e-05, "loss": 0.0266, "step": 7130 }, { "grad_norm": 0.3872150778770447, "learning_rate": 9.870462153739257e-05, "loss": 0.0315, "step": 7140 }, { "grad_norm": 0.32647955417633057, "learning_rate": 9.869838192986067e-05, "loss": 0.0277, "step": 7150 }, { "grad_norm": 0.298687219619751, "learning_rate": 9.869212752904616e-05, "loss": 0.0287, "step": 7160 }, { "grad_norm": 0.4084342122077942, "learning_rate": 9.868585833684894e-05, "loss": 0.0267, "step": 7170 }, { "grad_norm": 0.38848501443862915, "learning_rate": 9.867957435517342e-05, "loss": 0.032, "step": 7180 }, { "grad_norm": 0.342729389667511, "learning_rate": 9.867327558592854e-05, "loss": 0.0294, "step": 7190 }, { "grad_norm": 0.2810225188732147, "learning_rate": 9.866696203102766e-05, "loss": 0.03, "step": 7200 }, { "grad_norm": 0.21295927464962006, "learning_rate": 9.86606336923887e-05, "loss": 0.0314, "step": 7210 }, { "grad_norm": 0.32007771730422974, "learning_rate": 9.865429057193403e-05, "loss": 0.0319, "step": 7220 }, { "grad_norm": 0.2807425856590271, "learning_rate": 9.864793267159053e-05, "loss": 0.032, "step": 7230 }, { "grad_norm": 0.30267590284347534, "learning_rate": 9.864155999328957e-05, "loss": 0.0274, "step": 7240 }, { "grad_norm": 0.2749980390071869, "learning_rate": 9.8635172538967e-05, "loss": 0.0276, "step": 7250 }, { "grad_norm": 0.288605660200119, "learning_rate": 9.862877031056312e-05, "loss": 0.0285, "step": 7260 }, { "grad_norm": 0.2730567455291748, "learning_rate": 9.862235331002279e-05, "loss": 0.0284, "step": 7270 }, { "grad_norm": 0.24366256594657898, "learning_rate": 9.861592153929533e-05, "loss": 0.0271, "step": 7280 }, { "grad_norm": 0.3140794634819031, "learning_rate": 9.860947500033455e-05, "loss": 0.0297, "step": 7290 }, { "grad_norm": 0.30340802669525146, "learning_rate": 9.86030136950987e-05, "loss": 0.0277, "step": 7300 }, { "grad_norm": 0.32342633605003357, "learning_rate": 9.85965376255506e-05, "loss": 0.0281, "step": 7310 }, { "grad_norm": 0.3216359317302704, "learning_rate": 9.859004679365747e-05, "loss": 0.0304, "step": 7320 }, { "grad_norm": 0.34430015087127686, "learning_rate": 9.858354120139108e-05, "loss": 0.0335, "step": 7330 }, { "grad_norm": 0.26155346632003784, "learning_rate": 9.857702085072764e-05, "loss": 0.0301, "step": 7340 }, { "grad_norm": 0.41682958602905273, "learning_rate": 9.857048574364787e-05, "loss": 0.0324, "step": 7350 }, { "grad_norm": 0.275655061006546, "learning_rate": 9.856393588213698e-05, "loss": 0.0281, "step": 7360 }, { "grad_norm": 0.2942330241203308, "learning_rate": 9.855737126818458e-05, "loss": 0.0305, "step": 7370 }, { "grad_norm": 0.2910158634185791, "learning_rate": 9.855079190378491e-05, "loss": 0.0302, "step": 7380 }, { "grad_norm": 0.3914359211921692, "learning_rate": 9.854419779093655e-05, "loss": 0.031, "step": 7390 }, { "grad_norm": 0.3008287847042084, "learning_rate": 9.853758893164264e-05, "loss": 0.0304, "step": 7400 }, { "grad_norm": 0.3547568619251251, "learning_rate": 9.853096532791078e-05, "loss": 0.0331, "step": 7410 }, { "grad_norm": 0.3412168025970459, "learning_rate": 9.852432698175304e-05, "loss": 0.0294, "step": 7420 }, { "grad_norm": 0.3696470558643341, "learning_rate": 9.851767389518597e-05, "loss": 0.0367, "step": 7430 }, { "grad_norm": 0.31847110390663147, "learning_rate": 9.85110060702306e-05, "loss": 0.0319, "step": 7440 }, { "grad_norm": 0.3443480134010315, "learning_rate": 9.850432350891245e-05, "loss": 0.0305, "step": 7450 }, { "grad_norm": 0.32419416308403015, "learning_rate": 9.84976262132615e-05, "loss": 0.0311, "step": 7460 }, { "grad_norm": 0.30363476276397705, "learning_rate": 9.849091418531222e-05, "loss": 0.0355, "step": 7470 }, { "grad_norm": 0.362386554479599, "learning_rate": 9.848418742710353e-05, "loss": 0.0306, "step": 7480 }, { "grad_norm": 0.4304051697254181, "learning_rate": 9.847744594067885e-05, "loss": 0.0344, "step": 7490 }, { "grad_norm": 0.3357637822628021, "learning_rate": 9.847068972808607e-05, "loss": 0.0333, "step": 7500 }, { "grad_norm": 0.3295206129550934, "learning_rate": 9.846391879137756e-05, "loss": 0.0281, "step": 7510 }, { "grad_norm": 0.29146185517311096, "learning_rate": 9.845713313261012e-05, "loss": 0.0272, "step": 7520 }, { "grad_norm": 0.2516167163848877, "learning_rate": 9.845033275384505e-05, "loss": 0.0296, "step": 7530 }, { "grad_norm": 0.26563677191734314, "learning_rate": 9.844351765714818e-05, "loss": 0.029, "step": 7540 }, { "grad_norm": 0.34071382880210876, "learning_rate": 9.843668784458971e-05, "loss": 0.0288, "step": 7550 }, { "grad_norm": 0.3154810667037964, "learning_rate": 9.842984331824437e-05, "loss": 0.0333, "step": 7560 }, { "grad_norm": 0.23538725078105927, "learning_rate": 9.842298408019133e-05, "loss": 0.0268, "step": 7570 }, { "grad_norm": 0.27176088094711304, "learning_rate": 9.841611013251429e-05, "loss": 0.0327, "step": 7580 }, { "grad_norm": 0.2794550955295563, "learning_rate": 9.840922147730133e-05, "loss": 0.0234, "step": 7590 }, { "grad_norm": 0.2870292365550995, "learning_rate": 9.840231811664506e-05, "loss": 0.0277, "step": 7600 }, { "grad_norm": 0.27878493070602417, "learning_rate": 9.839540005264252e-05, "loss": 0.0294, "step": 7610 }, { "grad_norm": 0.3701087236404419, "learning_rate": 9.838846728739527e-05, "loss": 0.0289, "step": 7620 }, { "grad_norm": 0.27764028310775757, "learning_rate": 9.838151982300927e-05, "loss": 0.0274, "step": 7630 }, { "grad_norm": 0.35061559081077576, "learning_rate": 9.8374557661595e-05, "loss": 0.0353, "step": 7640 }, { "grad_norm": 0.289299339056015, "learning_rate": 9.836758080526735e-05, "loss": 0.0332, "step": 7650 }, { "grad_norm": 0.33606818318367004, "learning_rate": 9.836058925614575e-05, "loss": 0.0254, "step": 7660 }, { "grad_norm": 0.2604127824306488, "learning_rate": 9.8353583016354e-05, "loss": 0.0295, "step": 7670 }, { "grad_norm": 0.2732633054256439, "learning_rate": 9.834656208802044e-05, "loss": 0.0311, "step": 7680 }, { "grad_norm": 0.26588037610054016, "learning_rate": 9.833952647327784e-05, "loss": 0.0302, "step": 7690 }, { "grad_norm": 0.3234092593193054, "learning_rate": 9.833247617426342e-05, "loss": 0.0301, "step": 7700 }, { "grad_norm": 0.2875858247280121, "learning_rate": 9.832541119311889e-05, "loss": 0.0298, "step": 7710 }, { "grad_norm": 0.2651088237762451, "learning_rate": 9.83183315319904e-05, "loss": 0.0245, "step": 7720 }, { "grad_norm": 0.2617790400981903, "learning_rate": 9.831123719302855e-05, "loss": 0.0287, "step": 7730 }, { "grad_norm": 0.26759836077690125, "learning_rate": 9.830412817838842e-05, "loss": 0.0277, "step": 7740 }, { "grad_norm": 0.3043751120567322, "learning_rate": 9.829700449022956e-05, "loss": 0.0288, "step": 7750 }, { "grad_norm": 0.32403793931007385, "learning_rate": 9.828986613071593e-05, "loss": 0.0304, "step": 7760 }, { "grad_norm": 0.26662757992744446, "learning_rate": 9.828271310201601e-05, "loss": 0.0272, "step": 7770 }, { "grad_norm": 0.21660564839839935, "learning_rate": 9.827554540630268e-05, "loss": 0.0281, "step": 7780 }, { "grad_norm": 0.2069849669933319, "learning_rate": 9.826836304575329e-05, "loss": 0.0256, "step": 7790 }, { "grad_norm": 0.32801732420921326, "learning_rate": 9.826116602254966e-05, "loss": 0.0311, "step": 7800 }, { "grad_norm": 0.3178703784942627, "learning_rate": 9.825395433887805e-05, "loss": 0.0313, "step": 7810 }, { "grad_norm": 0.2806498110294342, "learning_rate": 9.824672799692917e-05, "loss": 0.0274, "step": 7820 }, { "grad_norm": 0.2851501703262329, "learning_rate": 9.823948699889823e-05, "loss": 0.0285, "step": 7830 }, { "grad_norm": 0.3869074285030365, "learning_rate": 9.823223134698483e-05, "loss": 0.0284, "step": 7840 }, { "grad_norm": 0.4031084477901459, "learning_rate": 9.822496104339303e-05, "loss": 0.0259, "step": 7850 }, { "grad_norm": 0.3065429925918579, "learning_rate": 9.821767609033138e-05, "loss": 0.0283, "step": 7860 }, { "grad_norm": 0.2156037986278534, "learning_rate": 9.821037649001284e-05, "loss": 0.0282, "step": 7870 }, { "grad_norm": 0.3216991424560547, "learning_rate": 9.820306224465486e-05, "loss": 0.0314, "step": 7880 }, { "grad_norm": 0.24999511241912842, "learning_rate": 9.819573335647928e-05, "loss": 0.0259, "step": 7890 }, { "grad_norm": 0.22707845270633698, "learning_rate": 9.818838982771246e-05, "loss": 0.0268, "step": 7900 }, { "grad_norm": 0.292959064245224, "learning_rate": 9.818103166058514e-05, "loss": 0.0324, "step": 7910 }, { "grad_norm": 0.26074525713920593, "learning_rate": 9.817365885733254e-05, "loss": 0.0272, "step": 7920 }, { "grad_norm": 0.21737216413021088, "learning_rate": 9.816627142019434e-05, "loss": 0.0288, "step": 7930 }, { "grad_norm": 0.18364916741847992, "learning_rate": 9.815886935141463e-05, "loss": 0.0299, "step": 7940 }, { "grad_norm": 0.22927384078502655, "learning_rate": 9.8151452653242e-05, "loss": 0.0242, "step": 7950 }, { "grad_norm": 0.28923436999320984, "learning_rate": 9.814402132792939e-05, "loss": 0.0263, "step": 7960 }, { "grad_norm": 0.3171807527542114, "learning_rate": 9.813657537773428e-05, "loss": 0.0266, "step": 7970 }, { "grad_norm": 0.2683235704898834, "learning_rate": 9.812911480491854e-05, "loss": 0.0288, "step": 7980 }, { "grad_norm": 0.25860491394996643, "learning_rate": 9.81216396117485e-05, "loss": 0.0241, "step": 7990 }, { "grad_norm": 0.2866578996181488, "learning_rate": 9.811414980049491e-05, "loss": 0.0253, "step": 8000 }, { "grad_norm": 0.3496307134628296, "learning_rate": 9.810664537343301e-05, "loss": 0.027, "step": 8010 }, { "grad_norm": 0.3414653241634369, "learning_rate": 9.809912633284243e-05, "loss": 0.0287, "step": 8020 }, { "grad_norm": 0.2559238076210022, "learning_rate": 9.809159268100725e-05, "loss": 0.0254, "step": 8030 }, { "grad_norm": 0.2178637683391571, "learning_rate": 9.808404442021599e-05, "loss": 0.0279, "step": 8040 }, { "grad_norm": 0.26251450181007385, "learning_rate": 9.807648155276163e-05, "loss": 0.0253, "step": 8050 }, { "grad_norm": 0.38312774896621704, "learning_rate": 9.806890408094156e-05, "loss": 0.0256, "step": 8060 }, { "grad_norm": 0.23094558715820312, "learning_rate": 9.806131200705761e-05, "loss": 0.0262, "step": 8070 }, { "grad_norm": 0.28549057245254517, "learning_rate": 9.805370533341605e-05, "loss": 0.0266, "step": 8080 }, { "grad_norm": 0.264074444770813, "learning_rate": 9.804608406232762e-05, "loss": 0.0249, "step": 8090 }, { "grad_norm": 0.2705989480018616, "learning_rate": 9.803844819610741e-05, "loss": 0.0258, "step": 8100 }, { "grad_norm": 0.2862582802772522, "learning_rate": 9.803079773707504e-05, "loss": 0.0297, "step": 8110 }, { "grad_norm": 0.24136805534362793, "learning_rate": 9.802313268755447e-05, "loss": 0.028, "step": 8120 }, { "grad_norm": 0.2213464379310608, "learning_rate": 9.801545304987419e-05, "loss": 0.026, "step": 8130 }, { "grad_norm": 0.27490442991256714, "learning_rate": 9.800775882636704e-05, "loss": 0.0282, "step": 8140 }, { "grad_norm": 0.29307880997657776, "learning_rate": 9.800005001937034e-05, "loss": 0.0282, "step": 8150 }, { "grad_norm": 0.28472593426704407, "learning_rate": 9.79923266312258e-05, "loss": 0.0289, "step": 8160 }, { "grad_norm": 0.34690243005752563, "learning_rate": 9.79845886642796e-05, "loss": 0.0297, "step": 8170 }, { "grad_norm": 0.30912479758262634, "learning_rate": 9.797683612088233e-05, "loss": 0.0289, "step": 8180 }, { "grad_norm": 0.2814044952392578, "learning_rate": 9.796906900338898e-05, "loss": 0.0275, "step": 8190 }, { "grad_norm": 0.32253405451774597, "learning_rate": 9.796128731415903e-05, "loss": 0.0244, "step": 8200 }, { "grad_norm": 0.28466200828552246, "learning_rate": 9.795349105555634e-05, "loss": 0.0272, "step": 8210 }, { "grad_norm": 0.2599692642688751, "learning_rate": 9.794568022994922e-05, "loss": 0.0236, "step": 8220 }, { "grad_norm": 0.29395467042922974, "learning_rate": 9.793785483971034e-05, "loss": 0.0247, "step": 8230 }, { "grad_norm": 0.35641950368881226, "learning_rate": 9.793001488721691e-05, "loss": 0.0244, "step": 8240 }, { "grad_norm": 0.22591222822666168, "learning_rate": 9.792216037485047e-05, "loss": 0.0251, "step": 8250 }, { "grad_norm": 0.23377232253551483, "learning_rate": 9.791429130499704e-05, "loss": 0.0273, "step": 8260 }, { "grad_norm": 0.2855989634990692, "learning_rate": 9.790640768004698e-05, "loss": 0.0306, "step": 8270 }, { "grad_norm": 0.2909029424190521, "learning_rate": 9.789850950239518e-05, "loss": 0.029, "step": 8280 }, { "grad_norm": 0.3339901268482208, "learning_rate": 9.789059677444089e-05, "loss": 0.0289, "step": 8290 }, { "grad_norm": 0.42587971687316895, "learning_rate": 9.788266949858776e-05, "loss": 0.0262, "step": 8300 }, { "grad_norm": 0.3176492750644684, "learning_rate": 9.787472767724392e-05, "loss": 0.0293, "step": 8310 }, { "grad_norm": 0.2683604657649994, "learning_rate": 9.786677131282185e-05, "loss": 0.0276, "step": 8320 }, { "grad_norm": 0.29330340027809143, "learning_rate": 9.785880040773853e-05, "loss": 0.025, "step": 8330 }, { "grad_norm": 0.315416544675827, "learning_rate": 9.785081496441527e-05, "loss": 0.0256, "step": 8340 }, { "grad_norm": 0.21860909461975098, "learning_rate": 9.784281498527785e-05, "loss": 0.0247, "step": 8350 }, { "grad_norm": 0.2714749574661255, "learning_rate": 9.783480047275646e-05, "loss": 0.0235, "step": 8360 }, { "grad_norm": 0.36661073565483093, "learning_rate": 9.78267714292857e-05, "loss": 0.0253, "step": 8370 }, { "grad_norm": 0.2739230692386627, "learning_rate": 9.781872785730454e-05, "loss": 0.027, "step": 8380 }, { "grad_norm": 0.2830897271633148, "learning_rate": 9.781066975925646e-05, "loss": 0.026, "step": 8390 }, { "grad_norm": 0.22100189328193665, "learning_rate": 9.780259713758928e-05, "loss": 0.0245, "step": 8400 }, { "grad_norm": 0.32985079288482666, "learning_rate": 9.779450999475524e-05, "loss": 0.0247, "step": 8410 }, { "grad_norm": 0.33609288930892944, "learning_rate": 9.7786408333211e-05, "loss": 0.0267, "step": 8420 }, { "grad_norm": 0.28615397214889526, "learning_rate": 9.777829215541764e-05, "loss": 0.0241, "step": 8430 }, { "grad_norm": 0.30025020241737366, "learning_rate": 9.777016146384064e-05, "loss": 0.0263, "step": 8440 }, { "grad_norm": 0.23481041193008423, "learning_rate": 9.776201626094988e-05, "loss": 0.0263, "step": 8450 }, { "grad_norm": 0.2752722203731537, "learning_rate": 9.775385654921965e-05, "loss": 0.0274, "step": 8460 }, { "grad_norm": 0.25948959589004517, "learning_rate": 9.774568233112868e-05, "loss": 0.0284, "step": 8470 }, { "grad_norm": 0.22837062180042267, "learning_rate": 9.773749360916007e-05, "loss": 0.025, "step": 8480 }, { "grad_norm": 0.27387309074401855, "learning_rate": 9.772929038580134e-05, "loss": 0.0228, "step": 8490 }, { "grad_norm": 0.3017122745513916, "learning_rate": 9.772107266354439e-05, "loss": 0.0245, "step": 8500 }, { "grad_norm": 0.2967309355735779, "learning_rate": 9.77128404448856e-05, "loss": 0.027, "step": 8510 }, { "grad_norm": 0.34241509437561035, "learning_rate": 9.770459373232565e-05, "loss": 0.028, "step": 8520 }, { "grad_norm": 0.2630164921283722, "learning_rate": 9.769633252836969e-05, "loss": 0.0298, "step": 8530 }, { "grad_norm": 0.28461790084838867, "learning_rate": 9.768805683552724e-05, "loss": 0.0305, "step": 8540 }, { "grad_norm": 0.24441032111644745, "learning_rate": 9.767976665631228e-05, "loss": 0.027, "step": 8550 }, { "grad_norm": 0.29866012930870056, "learning_rate": 9.767146199324311e-05, "loss": 0.029, "step": 8560 }, { "grad_norm": 0.30913788080215454, "learning_rate": 9.766314284884249e-05, "loss": 0.0251, "step": 8570 }, { "grad_norm": 0.272126704454422, "learning_rate": 9.765480922563752e-05, "loss": 0.0277, "step": 8580 }, { "grad_norm": 0.22301889955997467, "learning_rate": 9.764646112615978e-05, "loss": 0.0269, "step": 8590 }, { "grad_norm": 0.2691085934638977, "learning_rate": 9.763809855294517e-05, "loss": 0.0246, "step": 8600 }, { "grad_norm": 0.269733726978302, "learning_rate": 9.762972150853404e-05, "loss": 0.0263, "step": 8610 }, { "grad_norm": 0.28248611092567444, "learning_rate": 9.762132999547111e-05, "loss": 0.026, "step": 8620 }, { "grad_norm": 0.25540366768836975, "learning_rate": 9.761292401630549e-05, "loss": 0.0223, "step": 8630 }, { "grad_norm": 0.21240611374378204, "learning_rate": 9.76045035735907e-05, "loss": 0.0241, "step": 8640 }, { "grad_norm": 0.23431488871574402, "learning_rate": 9.759606866988464e-05, "loss": 0.0257, "step": 8650 }, { "grad_norm": 0.261930376291275, "learning_rate": 9.758761930774963e-05, "loss": 0.0263, "step": 8660 }, { "grad_norm": 0.3126588463783264, "learning_rate": 9.757915548975235e-05, "loss": 0.0277, "step": 8670 }, { "grad_norm": 0.2715398073196411, "learning_rate": 9.757067721846389e-05, "loss": 0.0275, "step": 8680 }, { "grad_norm": 0.2792823910713196, "learning_rate": 9.756218449645971e-05, "loss": 0.0268, "step": 8690 }, { "grad_norm": 0.2766096591949463, "learning_rate": 9.75536773263197e-05, "loss": 0.0235, "step": 8700 }, { "grad_norm": 0.26456964015960693, "learning_rate": 9.75451557106281e-05, "loss": 0.0243, "step": 8710 }, { "grad_norm": 0.24190494418144226, "learning_rate": 9.753661965197354e-05, "loss": 0.023, "step": 8720 }, { "grad_norm": 0.28041931986808777, "learning_rate": 9.752806915294908e-05, "loss": 0.0264, "step": 8730 }, { "grad_norm": 0.21986833214759827, "learning_rate": 9.75195042161521e-05, "loss": 0.0253, "step": 8740 }, { "grad_norm": 0.33563533425331116, "learning_rate": 9.751092484418442e-05, "loss": 0.0304, "step": 8750 }, { "grad_norm": 0.20856404304504395, "learning_rate": 9.750233103965224e-05, "loss": 0.0262, "step": 8760 }, { "grad_norm": 0.21039685606956482, "learning_rate": 9.749372280516611e-05, "loss": 0.0231, "step": 8770 }, { "grad_norm": 0.28348100185394287, "learning_rate": 9.748510014334097e-05, "loss": 0.0233, "step": 8780 }, { "grad_norm": 0.2329883575439453, "learning_rate": 9.747646305679621e-05, "loss": 0.0229, "step": 8790 }, { "grad_norm": 0.2867863178253174, "learning_rate": 9.74678115481555e-05, "loss": 0.028, "step": 8800 }, { "grad_norm": 0.373161256313324, "learning_rate": 9.745914562004696e-05, "loss": 0.027, "step": 8810 }, { "grad_norm": 0.3572908639907837, "learning_rate": 9.745046527510307e-05, "loss": 0.0289, "step": 8820 }, { "grad_norm": 0.31740885972976685, "learning_rate": 9.744177051596068e-05, "loss": 0.0325, "step": 8830 }, { "grad_norm": 0.24094000458717346, "learning_rate": 9.743306134526105e-05, "loss": 0.0291, "step": 8840 }, { "grad_norm": 0.2866745889186859, "learning_rate": 9.742433776564977e-05, "loss": 0.0321, "step": 8850 }, { "grad_norm": 0.29295703768730164, "learning_rate": 9.741559977977683e-05, "loss": 0.0273, "step": 8860 }, { "grad_norm": 0.20020654797554016, "learning_rate": 9.740684739029661e-05, "loss": 0.027, "step": 8870 }, { "grad_norm": 0.23062194883823395, "learning_rate": 9.739808059986789e-05, "loss": 0.0238, "step": 8880 }, { "grad_norm": 0.2685215473175049, "learning_rate": 9.738929941115373e-05, "loss": 0.0286, "step": 8890 }, { "grad_norm": 0.23373660445213318, "learning_rate": 9.738050382682167e-05, "loss": 0.0231, "step": 8900 }, { "grad_norm": 0.19873349368572235, "learning_rate": 9.737169384954355e-05, "loss": 0.0247, "step": 8910 }, { "grad_norm": 0.19656549394130707, "learning_rate": 9.736286948199562e-05, "loss": 0.0229, "step": 8920 }, { "grad_norm": 0.26883891224861145, "learning_rate": 9.735403072685848e-05, "loss": 0.0239, "step": 8930 }, { "grad_norm": 0.2619089186191559, "learning_rate": 9.734517758681712e-05, "loss": 0.0229, "step": 8940 }, { "grad_norm": 0.37093478441238403, "learning_rate": 9.733631006456088e-05, "loss": 0.0265, "step": 8950 }, { "grad_norm": 0.295609712600708, "learning_rate": 9.732742816278348e-05, "loss": 0.0296, "step": 8960 }, { "grad_norm": 0.2909187972545624, "learning_rate": 9.731853188418302e-05, "loss": 0.0264, "step": 8970 }, { "grad_norm": 0.2272867113351822, "learning_rate": 9.730962123146194e-05, "loss": 0.026, "step": 8980 }, { "grad_norm": 0.23886997997760773, "learning_rate": 9.730069620732709e-05, "loss": 0.0274, "step": 8990 }, { "grad_norm": 0.23860760033130646, "learning_rate": 9.72917568144896e-05, "loss": 0.0301, "step": 9000 }, { "grad_norm": 0.2938750982284546, "learning_rate": 9.728280305566509e-05, "loss": 0.0279, "step": 9010 }, { "grad_norm": 0.25944027304649353, "learning_rate": 9.727383493357343e-05, "loss": 0.0224, "step": 9020 }, { "grad_norm": 0.3160631060600281, "learning_rate": 9.726485245093891e-05, "loss": 0.0283, "step": 9030 }, { "grad_norm": 0.29162925481796265, "learning_rate": 9.725585561049018e-05, "loss": 0.0272, "step": 9040 }, { "grad_norm": 0.2509813606739044, "learning_rate": 9.724684441496022e-05, "loss": 0.0232, "step": 9050 }, { "grad_norm": 0.194972425699234, "learning_rate": 9.72378188670864e-05, "loss": 0.0239, "step": 9060 }, { "grad_norm": 0.2572469711303711, "learning_rate": 9.722877896961047e-05, "loss": 0.0256, "step": 9070 }, { "grad_norm": 0.2768094539642334, "learning_rate": 9.721972472527848e-05, "loss": 0.0257, "step": 9080 }, { "grad_norm": 0.25002557039260864, "learning_rate": 9.721065613684089e-05, "loss": 0.0263, "step": 9090 }, { "grad_norm": 0.24552474915981293, "learning_rate": 9.72015732070525e-05, "loss": 0.025, "step": 9100 }, { "grad_norm": 0.3009200096130371, "learning_rate": 9.719247593867244e-05, "loss": 0.0255, "step": 9110 }, { "grad_norm": 0.2728825807571411, "learning_rate": 9.718336433446423e-05, "loss": 0.025, "step": 9120 }, { "grad_norm": 0.30432939529418945, "learning_rate": 9.717423839719574e-05, "loss": 0.0241, "step": 9130 }, { "grad_norm": 0.31839701533317566, "learning_rate": 9.71650981296392e-05, "loss": 0.0273, "step": 9140 }, { "grad_norm": 0.25372326374053955, "learning_rate": 9.715594353457118e-05, "loss": 0.0222, "step": 9150 }, { "grad_norm": 0.26152610778808594, "learning_rate": 9.714677461477257e-05, "loss": 0.0249, "step": 9160 }, { "grad_norm": 0.25475695729255676, "learning_rate": 9.713759137302869e-05, "loss": 0.0251, "step": 9170 }, { "grad_norm": 0.31092360615730286, "learning_rate": 9.712839381212914e-05, "loss": 0.0231, "step": 9180 }, { "grad_norm": 0.2575063705444336, "learning_rate": 9.71191819348679e-05, "loss": 0.0248, "step": 9190 }, { "grad_norm": 0.22370685636997223, "learning_rate": 9.710995574404331e-05, "loss": 0.0248, "step": 9200 }, { "grad_norm": 0.27329498529434204, "learning_rate": 9.710071524245802e-05, "loss": 0.0265, "step": 9210 }, { "grad_norm": 0.3020857274532318, "learning_rate": 9.709146043291906e-05, "loss": 0.0273, "step": 9220 }, { "grad_norm": 0.21980366110801697, "learning_rate": 9.70821913182378e-05, "loss": 0.0235, "step": 9230 }, { "grad_norm": 0.27141159772872925, "learning_rate": 9.707290790122995e-05, "loss": 0.026, "step": 9240 }, { "grad_norm": 0.24188312888145447, "learning_rate": 9.706361018471557e-05, "loss": 0.023, "step": 9250 }, { "grad_norm": 0.3038684129714966, "learning_rate": 9.705429817151906e-05, "loss": 0.022, "step": 9260 }, { "grad_norm": 0.17182745039463043, "learning_rate": 9.704497186446917e-05, "loss": 0.023, "step": 9270 }, { "grad_norm": 0.2454695701599121, "learning_rate": 9.703563126639896e-05, "loss": 0.025, "step": 9280 }, { "grad_norm": 0.3079718053340912, "learning_rate": 9.70262763801459e-05, "loss": 0.0242, "step": 9290 }, { "grad_norm": 0.3126790225505829, "learning_rate": 9.701690720855171e-05, "loss": 0.0258, "step": 9300 }, { "grad_norm": 0.22068674862384796, "learning_rate": 9.700752375446253e-05, "loss": 0.0222, "step": 9310 }, { "grad_norm": 0.20902153849601746, "learning_rate": 9.69981260207288e-05, "loss": 0.0225, "step": 9320 }, { "grad_norm": 0.22475716471672058, "learning_rate": 9.698871401020529e-05, "loss": 0.0217, "step": 9330 }, { "grad_norm": 0.2700927257537842, "learning_rate": 9.697928772575112e-05, "loss": 0.0208, "step": 9340 }, { "grad_norm": 0.23334349691867828, "learning_rate": 9.696984717022976e-05, "loss": 0.0249, "step": 9350 }, { "grad_norm": 0.23974038660526276, "learning_rate": 9.6960392346509e-05, "loss": 0.0225, "step": 9360 }, { "grad_norm": 0.2309287041425705, "learning_rate": 9.695092325746097e-05, "loss": 0.0194, "step": 9370 }, { "grad_norm": 0.25658899545669556, "learning_rate": 9.694143990596211e-05, "loss": 0.0228, "step": 9380 }, { "grad_norm": 0.22856222093105316, "learning_rate": 9.693194229489325e-05, "loss": 0.0203, "step": 9390 }, { "grad_norm": 0.2808919847011566, "learning_rate": 9.692243042713944e-05, "loss": 0.0234, "step": 9400 }, { "grad_norm": 0.2534981667995453, "learning_rate": 9.691290430559022e-05, "loss": 0.02, "step": 9410 }, { "grad_norm": 0.27763307094573975, "learning_rate": 9.690336393313932e-05, "loss": 0.0244, "step": 9420 }, { "grad_norm": 0.2839394211769104, "learning_rate": 9.689380931268487e-05, "loss": 0.0263, "step": 9430 }, { "grad_norm": 0.31455564498901367, "learning_rate": 9.688424044712932e-05, "loss": 0.0243, "step": 9440 }, { "grad_norm": 0.28494295477867126, "learning_rate": 9.687465733937942e-05, "loss": 0.0258, "step": 9450 }, { "grad_norm": 0.2735263705253601, "learning_rate": 9.686505999234627e-05, "loss": 0.0219, "step": 9460 }, { "grad_norm": 0.31209951639175415, "learning_rate": 9.685544840894529e-05, "loss": 0.0233, "step": 9470 }, { "grad_norm": 0.3096020221710205, "learning_rate": 9.684582259209624e-05, "loss": 0.0258, "step": 9480 }, { "grad_norm": 0.26423412561416626, "learning_rate": 9.683618254472317e-05, "loss": 0.0243, "step": 9490 }, { "grad_norm": 0.27202874422073364, "learning_rate": 9.682652826975449e-05, "loss": 0.0233, "step": 9500 }, { "grad_norm": 0.25743362307548523, "learning_rate": 9.681685977012291e-05, "loss": 0.0243, "step": 9510 }, { "grad_norm": 0.22061367332935333, "learning_rate": 9.680717704876546e-05, "loss": 0.0202, "step": 9520 }, { "grad_norm": 0.22621747851371765, "learning_rate": 9.679748010862349e-05, "loss": 0.0194, "step": 9530 }, { "grad_norm": 0.2238488346338272, "learning_rate": 9.678776895264267e-05, "loss": 0.0214, "step": 9540 }, { "grad_norm": 0.27744805812835693, "learning_rate": 9.6778043583773e-05, "loss": 0.0285, "step": 9550 }, { "grad_norm": 0.27572101354599, "learning_rate": 9.67683040049688e-05, "loss": 0.0254, "step": 9560 }, { "grad_norm": 0.3111267685890198, "learning_rate": 9.675855021918869e-05, "loss": 0.0235, "step": 9570 }, { "grad_norm": 0.2709175944328308, "learning_rate": 9.674878222939561e-05, "loss": 0.0272, "step": 9580 }, { "grad_norm": 0.2143869400024414, "learning_rate": 9.673900003855681e-05, "loss": 0.0261, "step": 9590 }, { "grad_norm": 0.20765914022922516, "learning_rate": 9.672920364964389e-05, "loss": 0.0214, "step": 9600 }, { "grad_norm": 0.24630039930343628, "learning_rate": 9.671939306563269e-05, "loss": 0.0258, "step": 9610 }, { "grad_norm": 0.2970157265663147, "learning_rate": 9.670956828950345e-05, "loss": 0.0199, "step": 9620 }, { "grad_norm": 0.20422594249248505, "learning_rate": 9.669972932424065e-05, "loss": 0.022, "step": 9630 }, { "grad_norm": 0.2494874894618988, "learning_rate": 9.668987617283312e-05, "loss": 0.0262, "step": 9640 }, { "grad_norm": 0.22991034388542175, "learning_rate": 9.668000883827397e-05, "loss": 0.0262, "step": 9650 }, { "grad_norm": 0.2760732173919678, "learning_rate": 9.667012732356067e-05, "loss": 0.0256, "step": 9660 }, { "grad_norm": 0.27585285902023315, "learning_rate": 9.666023163169493e-05, "loss": 0.0211, "step": 9670 }, { "grad_norm": 0.23761990666389465, "learning_rate": 9.665032176568281e-05, "loss": 0.0202, "step": 9680 }, { "grad_norm": 0.228504478931427, "learning_rate": 9.664039772853469e-05, "loss": 0.0251, "step": 9690 }, { "grad_norm": 0.27810612320899963, "learning_rate": 9.663045952326518e-05, "loss": 0.0236, "step": 9700 }, { "grad_norm": 0.25788038969039917, "learning_rate": 9.662050715289328e-05, "loss": 0.024, "step": 9710 }, { "grad_norm": 0.3318619132041931, "learning_rate": 9.661054062044226e-05, "loss": 0.0235, "step": 9720 }, { "grad_norm": 0.2068675458431244, "learning_rate": 9.660055992893968e-05, "loss": 0.0215, "step": 9730 }, { "grad_norm": 0.2709611654281616, "learning_rate": 9.659056508141739e-05, "loss": 0.0236, "step": 9740 }, { "grad_norm": 0.29869577288627625, "learning_rate": 9.658055608091161e-05, "loss": 0.0264, "step": 9750 }, { "grad_norm": 0.3235894739627838, "learning_rate": 9.657053293046276e-05, "loss": 0.0242, "step": 9760 }, { "grad_norm": 0.33867502212524414, "learning_rate": 9.656049563311564e-05, "loss": 0.0222, "step": 9770 }, { "grad_norm": 0.2540975511074066, "learning_rate": 9.655044419191929e-05, "loss": 0.0196, "step": 9780 }, { "grad_norm": 0.24688223004341125, "learning_rate": 9.654037860992711e-05, "loss": 0.0214, "step": 9790 }, { "grad_norm": 0.279769629240036, "learning_rate": 9.653029889019672e-05, "loss": 0.0227, "step": 9800 }, { "grad_norm": 0.2591441869735718, "learning_rate": 9.65202050357901e-05, "loss": 0.0303, "step": 9810 }, { "grad_norm": 0.2676081955432892, "learning_rate": 9.651009704977347e-05, "loss": 0.0237, "step": 9820 }, { "grad_norm": 0.23794572055339813, "learning_rate": 9.649997493521738e-05, "loss": 0.0257, "step": 9830 }, { "grad_norm": 0.2424449622631073, "learning_rate": 9.64898386951967e-05, "loss": 0.0256, "step": 9840 }, { "grad_norm": 0.25994762778282166, "learning_rate": 9.647968833279049e-05, "loss": 0.0266, "step": 9850 }, { "grad_norm": 0.2617906630039215, "learning_rate": 9.646952385108218e-05, "loss": 0.0223, "step": 9860 }, { "grad_norm": 0.27693215012550354, "learning_rate": 9.645934525315951e-05, "loss": 0.019, "step": 9870 }, { "grad_norm": 0.27863648533821106, "learning_rate": 9.644915254211442e-05, "loss": 0.0211, "step": 9880 }, { "grad_norm": 0.2454381287097931, "learning_rate": 9.643894572104321e-05, "loss": 0.0228, "step": 9890 }, { "grad_norm": 0.261979341506958, "learning_rate": 9.642872479304644e-05, "loss": 0.0229, "step": 9900 }, { "grad_norm": 0.23108142614364624, "learning_rate": 9.641848976122895e-05, "loss": 0.0234, "step": 9910 }, { "grad_norm": 0.2886497378349304, "learning_rate": 9.64082406286999e-05, "loss": 0.0222, "step": 9920 }, { "grad_norm": 0.22083401679992676, "learning_rate": 9.639797739857269e-05, "loss": 0.0229, "step": 9930 }, { "grad_norm": 0.26336610317230225, "learning_rate": 9.638770007396498e-05, "loss": 0.0194, "step": 9940 }, { "grad_norm": 0.22399553656578064, "learning_rate": 9.63774086579988e-05, "loss": 0.0281, "step": 9950 }, { "grad_norm": 0.2615184485912323, "learning_rate": 9.63671031538004e-05, "loss": 0.0244, "step": 9960 }, { "grad_norm": 0.28737738728523254, "learning_rate": 9.635678356450031e-05, "loss": 0.0218, "step": 9970 }, { "grad_norm": 0.2952464520931244, "learning_rate": 9.634644989323336e-05, "loss": 0.0201, "step": 9980 }, { "grad_norm": 0.26230376958847046, "learning_rate": 9.633610214313861e-05, "loss": 0.0228, "step": 9990 }, { "grad_norm": 0.2842937409877777, "learning_rate": 9.632574031735951e-05, "loss": 0.023, "step": 10000 }, { "grad_norm": 0.21320229768753052, "learning_rate": 9.631536441904364e-05, "loss": 0.0204, "step": 10010 }, { "grad_norm": 0.18646517395973206, "learning_rate": 9.630497445134293e-05, "loss": 0.0227, "step": 10020 }, { "grad_norm": 0.24699938297271729, "learning_rate": 9.62945704174136e-05, "loss": 0.0249, "step": 10030 }, { "grad_norm": 0.22841426730155945, "learning_rate": 9.628415232041612e-05, "loss": 0.0192, "step": 10040 }, { "grad_norm": 0.2767488956451416, "learning_rate": 9.627372016351524e-05, "loss": 0.0208, "step": 10050 }, { "grad_norm": 0.2440277487039566, "learning_rate": 9.626327394987995e-05, "loss": 0.0198, "step": 10060 }, { "grad_norm": 0.1703268587589264, "learning_rate": 9.625281368268355e-05, "loss": 0.0214, "step": 10070 }, { "grad_norm": 0.2161899209022522, "learning_rate": 9.624233936510357e-05, "loss": 0.0272, "step": 10080 }, { "grad_norm": 0.2666926085948944, "learning_rate": 9.623185100032187e-05, "loss": 0.0221, "step": 10090 }, { "grad_norm": 0.24558012187480927, "learning_rate": 9.62213485915245e-05, "loss": 0.022, "step": 10100 }, { "grad_norm": 0.22514750063419342, "learning_rate": 9.621083214190186e-05, "loss": 0.019, "step": 10110 }, { "grad_norm": 0.2468109428882599, "learning_rate": 9.62003016546485e-05, "loss": 0.0217, "step": 10120 }, { "grad_norm": 0.3119235336780548, "learning_rate": 9.618975713296339e-05, "loss": 0.0241, "step": 10130 }, { "grad_norm": 0.2540808320045471, "learning_rate": 9.61791985800496e-05, "loss": 0.0214, "step": 10140 }, { "grad_norm": 0.2623448371887207, "learning_rate": 9.616862599911458e-05, "loss": 0.0213, "step": 10150 }, { "grad_norm": 0.2816168963909149, "learning_rate": 9.615803939337e-05, "loss": 0.0236, "step": 10160 }, { "grad_norm": 0.23890972137451172, "learning_rate": 9.614743876603178e-05, "loss": 0.0238, "step": 10170 }, { "grad_norm": 0.30983272194862366, "learning_rate": 9.613682412032013e-05, "loss": 0.0238, "step": 10180 }, { "grad_norm": 0.25246214866638184, "learning_rate": 9.612619545945947e-05, "loss": 0.024, "step": 10190 }, { "grad_norm": 0.30085745453834534, "learning_rate": 9.611555278667852e-05, "loss": 0.0224, "step": 10200 }, { "grad_norm": 0.30035078525543213, "learning_rate": 9.610489610521024e-05, "loss": 0.0264, "step": 10210 }, { "grad_norm": 0.3021634519100189, "learning_rate": 9.609422541829187e-05, "loss": 0.0257, "step": 10220 }, { "grad_norm": 0.25775134563446045, "learning_rate": 9.608354072916486e-05, "loss": 0.0217, "step": 10230 }, { "grad_norm": 0.29226675629615784, "learning_rate": 9.607284204107493e-05, "loss": 0.02, "step": 10240 }, { "grad_norm": 0.36087870597839355, "learning_rate": 9.606212935727208e-05, "loss": 0.0201, "step": 10250 }, { "grad_norm": 0.23776382207870483, "learning_rate": 9.605140268101052e-05, "loss": 0.0261, "step": 10260 }, { "grad_norm": 0.23404501378536224, "learning_rate": 9.604066201554875e-05, "loss": 0.0256, "step": 10270 }, { "grad_norm": 0.32163286209106445, "learning_rate": 9.60299073641495e-05, "loss": 0.0262, "step": 10280 }, { "grad_norm": 0.2591401934623718, "learning_rate": 9.601913873007974e-05, "loss": 0.0206, "step": 10290 }, { "grad_norm": 0.2293136715888977, "learning_rate": 9.60083561166107e-05, "loss": 0.0221, "step": 10300 }, { "grad_norm": 0.30413544178009033, "learning_rate": 9.599755952701783e-05, "loss": 0.019, "step": 10310 }, { "grad_norm": 0.25975412130355835, "learning_rate": 9.598674896458089e-05, "loss": 0.0207, "step": 10320 }, { "grad_norm": 0.21435637772083282, "learning_rate": 9.597592443258383e-05, "loss": 0.0226, "step": 10330 }, { "grad_norm": 0.21193958818912506, "learning_rate": 9.596508593431483e-05, "loss": 0.0206, "step": 10340 }, { "grad_norm": 0.24184051156044006, "learning_rate": 9.59542334730664e-05, "loss": 0.0261, "step": 10350 }, { "grad_norm": 0.2696307599544525, "learning_rate": 9.594336705213516e-05, "loss": 0.0233, "step": 10360 }, { "grad_norm": 0.20043347775936127, "learning_rate": 9.593248667482208e-05, "loss": 0.0213, "step": 10370 }, { "grad_norm": 0.3055836856365204, "learning_rate": 9.592159234443233e-05, "loss": 0.019, "step": 10380 }, { "grad_norm": 0.31220743060112, "learning_rate": 9.59106840642753e-05, "loss": 0.0213, "step": 10390 }, { "grad_norm": 0.22095954418182373, "learning_rate": 9.589976183766467e-05, "loss": 0.0218, "step": 10400 }, { "grad_norm": 0.22150757908821106, "learning_rate": 9.58888256679183e-05, "loss": 0.021, "step": 10410 }, { "grad_norm": 0.24769799411296844, "learning_rate": 9.587787555835832e-05, "loss": 0.0186, "step": 10420 }, { "grad_norm": 0.33027222752571106, "learning_rate": 9.586691151231107e-05, "loss": 0.0232, "step": 10430 }, { "grad_norm": 0.29998984932899475, "learning_rate": 9.585593353310715e-05, "loss": 0.0214, "step": 10440 }, { "grad_norm": 0.29605287313461304, "learning_rate": 9.58449416240814e-05, "loss": 0.0207, "step": 10450 }, { "grad_norm": 0.2093871533870697, "learning_rate": 9.583393578857283e-05, "loss": 0.0216, "step": 10460 }, { "grad_norm": 0.22331809997558594, "learning_rate": 9.582291602992474e-05, "loss": 0.0231, "step": 10470 }, { "grad_norm": 0.17657454311847687, "learning_rate": 9.581188235148466e-05, "loss": 0.0217, "step": 10480 }, { "grad_norm": 0.17561835050582886, "learning_rate": 9.58008347566043e-05, "loss": 0.0195, "step": 10490 }, { "grad_norm": 0.2396654188632965, "learning_rate": 9.578977324863965e-05, "loss": 0.0196, "step": 10500 }, { "grad_norm": 0.2918781042098999, "learning_rate": 9.577869783095089e-05, "loss": 0.0214, "step": 10510 }, { "grad_norm": 0.31195297837257385, "learning_rate": 9.576760850690245e-05, "loss": 0.023, "step": 10520 }, { "grad_norm": 0.2430851310491562, "learning_rate": 9.575650527986298e-05, "loss": 0.0228, "step": 10530 }, { "grad_norm": 0.2528993487358093, "learning_rate": 9.574538815320531e-05, "loss": 0.0217, "step": 10540 }, { "grad_norm": 0.3403021991252899, "learning_rate": 9.573425713030656e-05, "loss": 0.0239, "step": 10550 }, { "grad_norm": 0.23839783668518066, "learning_rate": 9.572311221454806e-05, "loss": 0.0207, "step": 10560 }, { "grad_norm": 0.288766086101532, "learning_rate": 9.57119534093153e-05, "loss": 0.0224, "step": 10570 }, { "grad_norm": 0.21156710386276245, "learning_rate": 9.570078071799806e-05, "loss": 0.0192, "step": 10580 }, { "grad_norm": 0.30415865778923035, "learning_rate": 9.568959414399028e-05, "loss": 0.0248, "step": 10590 }, { "grad_norm": 0.236445352435112, "learning_rate": 9.567839369069018e-05, "loss": 0.0237, "step": 10600 }, { "grad_norm": 0.2035071700811386, "learning_rate": 9.566717936150013e-05, "loss": 0.0245, "step": 10610 }, { "grad_norm": 0.2861986458301544, "learning_rate": 9.565595115982678e-05, "loss": 0.0273, "step": 10620 }, { "grad_norm": 0.20240004360675812, "learning_rate": 9.564470908908094e-05, "loss": 0.0215, "step": 10630 }, { "grad_norm": 0.25257349014282227, "learning_rate": 9.563345315267764e-05, "loss": 0.0246, "step": 10640 }, { "grad_norm": 0.18968558311462402, "learning_rate": 9.562218335403616e-05, "loss": 0.0201, "step": 10650 }, { "grad_norm": 0.2485685795545578, "learning_rate": 9.561089969657999e-05, "loss": 0.0214, "step": 10660 }, { "grad_norm": 0.22852814197540283, "learning_rate": 9.559960218373673e-05, "loss": 0.0193, "step": 10670 }, { "grad_norm": 0.25968044996261597, "learning_rate": 9.558829081893836e-05, "loss": 0.0174, "step": 10680 }, { "grad_norm": 0.22292767465114594, "learning_rate": 9.55769656056209e-05, "loss": 0.0197, "step": 10690 }, { "grad_norm": 0.20375956594944, "learning_rate": 9.556562654722469e-05, "loss": 0.0219, "step": 10700 }, { "grad_norm": 0.2570340633392334, "learning_rate": 9.555427364719422e-05, "loss": 0.0186, "step": 10710 }, { "grad_norm": 0.19825661182403564, "learning_rate": 9.55429069089782e-05, "loss": 0.0224, "step": 10720 }, { "grad_norm": 0.21507947146892548, "learning_rate": 9.553152633602956e-05, "loss": 0.0223, "step": 10730 }, { "grad_norm": 0.2734997570514679, "learning_rate": 9.552013193180543e-05, "loss": 0.0201, "step": 10740 }, { "grad_norm": 0.19347400963306427, "learning_rate": 9.550872369976707e-05, "loss": 0.0205, "step": 10750 }, { "grad_norm": 0.24851839244365692, "learning_rate": 9.549730164338007e-05, "loss": 0.0205, "step": 10760 }, { "grad_norm": 0.1897653192281723, "learning_rate": 9.548586576611408e-05, "loss": 0.0221, "step": 10770 }, { "grad_norm": 0.26419514417648315, "learning_rate": 9.54744160714431e-05, "loss": 0.0197, "step": 10780 }, { "grad_norm": 0.28799721598625183, "learning_rate": 9.546295256284516e-05, "loss": 0.0251, "step": 10790 }, { "grad_norm": 0.21643613278865814, "learning_rate": 9.545147524380265e-05, "loss": 0.0208, "step": 10800 }, { "grad_norm": 0.21761153638362885, "learning_rate": 9.543998411780201e-05, "loss": 0.0184, "step": 10810 }, { "grad_norm": 0.2500767707824707, "learning_rate": 9.542847918833397e-05, "loss": 0.0232, "step": 10820 }, { "grad_norm": 0.23129954934120178, "learning_rate": 9.541696045889343e-05, "loss": 0.0225, "step": 10830 }, { "grad_norm": 0.2713587284088135, "learning_rate": 9.540542793297947e-05, "loss": 0.0227, "step": 10840 }, { "grad_norm": 0.26837271451950073, "learning_rate": 9.539388161409537e-05, "loss": 0.0183, "step": 10850 }, { "grad_norm": 0.23725038766860962, "learning_rate": 9.538232150574857e-05, "loss": 0.0232, "step": 10860 }, { "grad_norm": 0.20496739447116852, "learning_rate": 9.537074761145076e-05, "loss": 0.0211, "step": 10870 }, { "grad_norm": 0.20489861071109772, "learning_rate": 9.535915993471778e-05, "loss": 0.0205, "step": 10880 }, { "grad_norm": 0.24158085882663727, "learning_rate": 9.534755847906964e-05, "loss": 0.023, "step": 10890 }, { "grad_norm": 0.2469126582145691, "learning_rate": 9.533594324803057e-05, "loss": 0.0187, "step": 10900 }, { "grad_norm": 0.27128809690475464, "learning_rate": 9.532431424512895e-05, "loss": 0.0208, "step": 10910 }, { "grad_norm": 0.22587986290454865, "learning_rate": 9.531267147389741e-05, "loss": 0.0219, "step": 10920 }, { "grad_norm": 0.2897195518016815, "learning_rate": 9.530101493787266e-05, "loss": 0.0208, "step": 10930 }, { "grad_norm": 0.23130372166633606, "learning_rate": 9.528934464059571e-05, "loss": 0.0203, "step": 10940 }, { "grad_norm": 0.23403559625148773, "learning_rate": 9.527766058561163e-05, "loss": 0.0183, "step": 10950 }, { "grad_norm": 0.31039804220199585, "learning_rate": 9.526596277646976e-05, "loss": 0.0195, "step": 10960 }, { "grad_norm": 0.20981623232364655, "learning_rate": 9.525425121672358e-05, "loss": 0.0185, "step": 10970 }, { "grad_norm": 0.20594173669815063, "learning_rate": 9.524252590993074e-05, "loss": 0.022, "step": 10980 }, { "grad_norm": 0.20931924879550934, "learning_rate": 9.523078685965309e-05, "loss": 0.0237, "step": 10990 }, { "grad_norm": 0.2313784956932068, "learning_rate": 9.521903406945664e-05, "loss": 0.0218, "step": 11000 }, { "grad_norm": 0.26550155878067017, "learning_rate": 9.520726754291158e-05, "loss": 0.0192, "step": 11010 }, { "grad_norm": 0.22477690875530243, "learning_rate": 9.519548728359227e-05, "loss": 0.0192, "step": 11020 }, { "grad_norm": 0.206288143992424, "learning_rate": 9.518369329507726e-05, "loss": 0.0213, "step": 11030 }, { "grad_norm": 0.19712817668914795, "learning_rate": 9.51718855809492e-05, "loss": 0.0185, "step": 11040 }, { "grad_norm": 0.22105719149112701, "learning_rate": 9.516006414479502e-05, "loss": 0.0216, "step": 11050 }, { "grad_norm": 0.24299603700637817, "learning_rate": 9.514822899020572e-05, "loss": 0.0197, "step": 11060 }, { "grad_norm": 0.27770915627479553, "learning_rate": 9.513638012077654e-05, "loss": 0.0206, "step": 11070 }, { "grad_norm": 0.24340519309043884, "learning_rate": 9.512451754010683e-05, "loss": 0.0222, "step": 11080 }, { "grad_norm": 0.2264869660139084, "learning_rate": 9.511264125180013e-05, "loss": 0.0194, "step": 11090 }, { "grad_norm": 0.26486366987228394, "learning_rate": 9.510075125946414e-05, "loss": 0.0208, "step": 11100 }, { "grad_norm": 0.2877127230167389, "learning_rate": 9.508884756671075e-05, "loss": 0.0191, "step": 11110 }, { "grad_norm": 0.25378409028053284, "learning_rate": 9.507693017715596e-05, "loss": 0.0184, "step": 11120 }, { "grad_norm": 0.21884149312973022, "learning_rate": 9.506499909441997e-05, "loss": 0.0195, "step": 11130 }, { "grad_norm": 0.2156885415315628, "learning_rate": 9.505305432212713e-05, "loss": 0.0208, "step": 11140 }, { "grad_norm": 0.27469390630722046, "learning_rate": 9.504109586390595e-05, "loss": 0.0234, "step": 11150 }, { "grad_norm": 0.17537429928779602, "learning_rate": 9.502912372338908e-05, "loss": 0.0177, "step": 11160 }, { "grad_norm": 0.2303655445575714, "learning_rate": 9.501713790421335e-05, "loss": 0.0199, "step": 11170 }, { "grad_norm": 0.3044017255306244, "learning_rate": 9.500513841001974e-05, "loss": 0.023, "step": 11180 }, { "grad_norm": 0.3115254342556, "learning_rate": 9.499312524445336e-05, "loss": 0.0199, "step": 11190 }, { "grad_norm": 0.17428725957870483, "learning_rate": 9.498109841116351e-05, "loss": 0.0199, "step": 11200 }, { "grad_norm": 0.22277629375457764, "learning_rate": 9.496905791380363e-05, "loss": 0.024, "step": 11210 }, { "grad_norm": 0.3416803777217865, "learning_rate": 9.495700375603129e-05, "loss": 0.0199, "step": 11220 }, { "grad_norm": 0.2620771527290344, "learning_rate": 9.494493594150822e-05, "loss": 0.0187, "step": 11230 }, { "grad_norm": 0.18279294669628143, "learning_rate": 9.493285447390032e-05, "loss": 0.0179, "step": 11240 }, { "grad_norm": 0.25196772813796997, "learning_rate": 9.492075935687761e-05, "loss": 0.0204, "step": 11250 }, { "grad_norm": 0.2046557366847992, "learning_rate": 9.490865059411427e-05, "loss": 0.0231, "step": 11260 }, { "grad_norm": 0.3009183704853058, "learning_rate": 9.489652818928863e-05, "loss": 0.024, "step": 11270 }, { "grad_norm": 0.2327512800693512, "learning_rate": 9.488439214608315e-05, "loss": 0.0205, "step": 11280 }, { "grad_norm": 0.2553662359714508, "learning_rate": 9.487224246818444e-05, "loss": 0.0182, "step": 11290 }, { "grad_norm": 0.22339417040348053, "learning_rate": 9.486007915928325e-05, "loss": 0.019, "step": 11300 }, { "grad_norm": 0.21725863218307495, "learning_rate": 9.484790222307448e-05, "loss": 0.0241, "step": 11310 }, { "grad_norm": 0.20214000344276428, "learning_rate": 9.483571166325716e-05, "loss": 0.0178, "step": 11320 }, { "grad_norm": 0.20181487500667572, "learning_rate": 9.482350748353444e-05, "loss": 0.0203, "step": 11330 }, { "grad_norm": 0.2824574410915375, "learning_rate": 9.481128968761363e-05, "loss": 0.0225, "step": 11340 }, { "grad_norm": 0.24230293929576874, "learning_rate": 9.479905827920621e-05, "loss": 0.0238, "step": 11350 }, { "grad_norm": 0.19346241652965546, "learning_rate": 9.478681326202773e-05, "loss": 0.0179, "step": 11360 }, { "grad_norm": 0.2382003366947174, "learning_rate": 9.477455463979791e-05, "loss": 0.0175, "step": 11370 }, { "grad_norm": 0.2579376995563507, "learning_rate": 9.476228241624059e-05, "loss": 0.0208, "step": 11380 }, { "grad_norm": 0.23818336427211761, "learning_rate": 9.474999659508374e-05, "loss": 0.0196, "step": 11390 }, { "grad_norm": 0.2272234708070755, "learning_rate": 9.47376971800595e-05, "loss": 0.0187, "step": 11400 }, { "grad_norm": 0.22194844484329224, "learning_rate": 9.472538417490409e-05, "loss": 0.0214, "step": 11410 }, { "grad_norm": 0.301463782787323, "learning_rate": 9.471305758335784e-05, "loss": 0.0255, "step": 11420 }, { "grad_norm": 0.2925691604614258, "learning_rate": 9.47007174091653e-05, "loss": 0.0241, "step": 11430 }, { "grad_norm": 0.26111477613449097, "learning_rate": 9.468836365607507e-05, "loss": 0.0225, "step": 11440 }, { "grad_norm": 0.22888490557670593, "learning_rate": 9.467599632783988e-05, "loss": 0.02, "step": 11450 }, { "grad_norm": 0.28699415922164917, "learning_rate": 9.466361542821662e-05, "loss": 0.0215, "step": 11460 }, { "grad_norm": 0.1714620292186737, "learning_rate": 9.465122096096625e-05, "loss": 0.0207, "step": 11470 }, { "grad_norm": 0.20376504957675934, "learning_rate": 9.463881292985391e-05, "loss": 0.0202, "step": 11480 }, { "grad_norm": 0.20778138935565948, "learning_rate": 9.462639133864881e-05, "loss": 0.0172, "step": 11490 }, { "grad_norm": 0.2572818398475647, "learning_rate": 9.461395619112432e-05, "loss": 0.0229, "step": 11500 }, { "grad_norm": 0.28616493940353394, "learning_rate": 9.460150749105791e-05, "loss": 0.0207, "step": 11510 }, { "grad_norm": 0.2412489354610443, "learning_rate": 9.458904524223116e-05, "loss": 0.0218, "step": 11520 }, { "grad_norm": 0.3251907527446747, "learning_rate": 9.457656944842976e-05, "loss": 0.0214, "step": 11530 }, { "grad_norm": 0.22133620083332062, "learning_rate": 9.456408011344353e-05, "loss": 0.0227, "step": 11540 }, { "grad_norm": 0.21413926780223846, "learning_rate": 9.455157724106643e-05, "loss": 0.0156, "step": 11550 }, { "grad_norm": 0.18873924016952515, "learning_rate": 9.453906083509647e-05, "loss": 0.0223, "step": 11560 }, { "grad_norm": 0.25849947333335876, "learning_rate": 9.45265308993358e-05, "loss": 0.0208, "step": 11570 }, { "grad_norm": 0.27195805311203003, "learning_rate": 9.451398743759071e-05, "loss": 0.019, "step": 11580 }, { "grad_norm": 0.27973195910453796, "learning_rate": 9.450143045367156e-05, "loss": 0.0196, "step": 11590 }, { "grad_norm": 0.28090330958366394, "learning_rate": 9.448885995139283e-05, "loss": 0.0167, "step": 11600 }, { "grad_norm": 0.27176347374916077, "learning_rate": 9.44762759345731e-05, "loss": 0.02, "step": 11610 }, { "grad_norm": 0.2434348315000534, "learning_rate": 9.446367840703509e-05, "loss": 0.0229, "step": 11620 }, { "grad_norm": 0.2186182290315628, "learning_rate": 9.445106737260556e-05, "loss": 0.0178, "step": 11630 }, { "grad_norm": 0.23839423060417175, "learning_rate": 9.443844283511543e-05, "loss": 0.019, "step": 11640 }, { "grad_norm": 0.26961758732795715, "learning_rate": 9.442580479839968e-05, "loss": 0.0192, "step": 11650 }, { "grad_norm": 0.21457980573177338, "learning_rate": 9.441315326629745e-05, "loss": 0.0195, "step": 11660 }, { "grad_norm": 0.18954695761203766, "learning_rate": 9.44004882426519e-05, "loss": 0.0168, "step": 11670 }, { "grad_norm": 0.19075077772140503, "learning_rate": 9.438780973131037e-05, "loss": 0.0173, "step": 11680 }, { "grad_norm": 0.19790397584438324, "learning_rate": 9.437511773612423e-05, "loss": 0.023, "step": 11690 }, { "grad_norm": 0.22151848673820496, "learning_rate": 9.436241226094896e-05, "loss": 0.0214, "step": 11700 }, { "grad_norm": 0.38620254397392273, "learning_rate": 9.434969330964418e-05, "loss": 0.0203, "step": 11710 }, { "grad_norm": 0.28609856963157654, "learning_rate": 9.433696088607356e-05, "loss": 0.0225, "step": 11720 }, { "grad_norm": 0.2226434201002121, "learning_rate": 9.432421499410486e-05, "loss": 0.0167, "step": 11730 }, { "grad_norm": 0.14988462626934052, "learning_rate": 9.431145563760998e-05, "loss": 0.0178, "step": 11740 }, { "grad_norm": 0.2301434576511383, "learning_rate": 9.429868282046484e-05, "loss": 0.0188, "step": 11750 }, { "grad_norm": 0.2158951461315155, "learning_rate": 9.428589654654951e-05, "loss": 0.0198, "step": 11760 }, { "grad_norm": 0.255167692899704, "learning_rate": 9.42730968197481e-05, "loss": 0.0188, "step": 11770 }, { "grad_norm": 0.39529314637184143, "learning_rate": 9.426028364394883e-05, "loss": 0.0227, "step": 11780 }, { "grad_norm": 0.1977742612361908, "learning_rate": 9.424745702304402e-05, "loss": 0.0176, "step": 11790 }, { "grad_norm": 0.2568926215171814, "learning_rate": 9.423461696093006e-05, "loss": 0.0239, "step": 11800 }, { "grad_norm": 0.1836090385913849, "learning_rate": 9.422176346150741e-05, "loss": 0.0176, "step": 11810 }, { "grad_norm": 0.23888666927814484, "learning_rate": 9.420889652868063e-05, "loss": 0.0227, "step": 11820 }, { "grad_norm": 0.2707851827144623, "learning_rate": 9.419601616635836e-05, "loss": 0.0201, "step": 11830 }, { "grad_norm": 0.1999122053384781, "learning_rate": 9.418312237845331e-05, "loss": 0.018, "step": 11840 }, { "grad_norm": 0.1898421347141266, "learning_rate": 9.417021516888225e-05, "loss": 0.0208, "step": 11850 }, { "grad_norm": 0.23976866900920868, "learning_rate": 9.415729454156608e-05, "loss": 0.0219, "step": 11860 }, { "grad_norm": 0.21036432683467865, "learning_rate": 9.414436050042973e-05, "loss": 0.0213, "step": 11870 }, { "grad_norm": 0.23508863151073456, "learning_rate": 9.413141304940223e-05, "loss": 0.0194, "step": 11880 }, { "grad_norm": 0.20469745993614197, "learning_rate": 9.411845219241666e-05, "loss": 0.0205, "step": 11890 }, { "grad_norm": 0.2507587671279907, "learning_rate": 9.410547793341021e-05, "loss": 0.0286, "step": 11900 }, { "grad_norm": 0.26002630591392517, "learning_rate": 9.409249027632408e-05, "loss": 0.0193, "step": 11910 }, { "grad_norm": 0.34307941794395447, "learning_rate": 9.407948922510362e-05, "loss": 0.0208, "step": 11920 }, { "grad_norm": 0.34946009516716003, "learning_rate": 9.406647478369817e-05, "loss": 0.0224, "step": 11930 }, { "grad_norm": 0.19818659126758575, "learning_rate": 9.405344695606118e-05, "loss": 0.019, "step": 11940 }, { "grad_norm": 0.24285535514354706, "learning_rate": 9.404040574615018e-05, "loss": 0.0231, "step": 11950 }, { "grad_norm": 0.23419585824012756, "learning_rate": 9.402735115792674e-05, "loss": 0.0211, "step": 11960 }, { "grad_norm": 0.28434282541275024, "learning_rate": 9.401428319535649e-05, "loss": 0.0191, "step": 11970 }, { "grad_norm": 0.3188750147819519, "learning_rate": 9.400120186240912e-05, "loss": 0.0236, "step": 11980 }, { "grad_norm": 0.30453330278396606, "learning_rate": 9.398810716305844e-05, "loss": 0.02, "step": 11990 }, { "grad_norm": 0.178699791431427, "learning_rate": 9.397499910128222e-05, "loss": 0.0202, "step": 12000 }, { "grad_norm": 0.25042882561683655, "learning_rate": 9.396187768106237e-05, "loss": 0.023, "step": 12010 }, { "grad_norm": 0.19229824841022491, "learning_rate": 9.394874290638482e-05, "loss": 0.0201, "step": 12020 }, { "grad_norm": 0.21389175951480865, "learning_rate": 9.393559478123959e-05, "loss": 0.0203, "step": 12030 }, { "grad_norm": 0.2106197327375412, "learning_rate": 9.39224333096207e-05, "loss": 0.0211, "step": 12040 }, { "grad_norm": 0.31676894426345825, "learning_rate": 9.390925849552629e-05, "loss": 0.0185, "step": 12050 }, { "grad_norm": 0.2288636863231659, "learning_rate": 9.389607034295849e-05, "loss": 0.018, "step": 12060 }, { "grad_norm": 0.13705113530158997, "learning_rate": 9.388286885592355e-05, "loss": 0.0198, "step": 12070 }, { "grad_norm": 0.27348792552948, "learning_rate": 9.386965403843168e-05, "loss": 0.0184, "step": 12080 }, { "grad_norm": 0.2108476459980011, "learning_rate": 9.385642589449726e-05, "loss": 0.0173, "step": 12090 }, { "grad_norm": 0.1725032478570938, "learning_rate": 9.38431844281386e-05, "loss": 0.0185, "step": 12100 }, { "grad_norm": 0.1822620928287506, "learning_rate": 9.38299296433781e-05, "loss": 0.0177, "step": 12110 }, { "grad_norm": 0.18285177648067474, "learning_rate": 9.381666154424226e-05, "loss": 0.0209, "step": 12120 }, { "grad_norm": 0.17892472445964813, "learning_rate": 9.380338013476157e-05, "loss": 0.0176, "step": 12130 }, { "grad_norm": 0.20343731343746185, "learning_rate": 9.379008541897054e-05, "loss": 0.0156, "step": 12140 }, { "grad_norm": 0.21706080436706543, "learning_rate": 9.377677740090777e-05, "loss": 0.0173, "step": 12150 }, { "grad_norm": 0.19666706025600433, "learning_rate": 9.376345608461588e-05, "loss": 0.0205, "step": 12160 }, { "grad_norm": 0.26149308681488037, "learning_rate": 9.375012147414155e-05, "loss": 0.0204, "step": 12170 }, { "grad_norm": 0.3626407980918884, "learning_rate": 9.373677357353545e-05, "loss": 0.0187, "step": 12180 }, { "grad_norm": 0.18187591433525085, "learning_rate": 9.372341238685237e-05, "loss": 0.0174, "step": 12190 }, { "grad_norm": 0.2629121243953705, "learning_rate": 9.371003791815102e-05, "loss": 0.022, "step": 12200 }, { "grad_norm": 0.22706715762615204, "learning_rate": 9.369665017149429e-05, "loss": 0.0177, "step": 12210 }, { "grad_norm": 0.2553153932094574, "learning_rate": 9.368324915094895e-05, "loss": 0.0189, "step": 12220 }, { "grad_norm": 0.27068233489990234, "learning_rate": 9.366983486058591e-05, "loss": 0.0177, "step": 12230 }, { "grad_norm": 0.20746204257011414, "learning_rate": 9.365640730448009e-05, "loss": 0.0189, "step": 12240 }, { "grad_norm": 0.25995638966560364, "learning_rate": 9.36429664867104e-05, "loss": 0.023, "step": 12250 }, { "grad_norm": 0.206175297498703, "learning_rate": 9.362951241135982e-05, "loss": 0.0191, "step": 12260 }, { "grad_norm": 0.20049114525318146, "learning_rate": 9.361604508251534e-05, "loss": 0.022, "step": 12270 }, { "grad_norm": 0.23497067391872406, "learning_rate": 9.360256450426799e-05, "loss": 0.0208, "step": 12280 }, { "grad_norm": 0.2543128430843353, "learning_rate": 9.358907068071279e-05, "loss": 0.0201, "step": 12290 }, { "grad_norm": 0.21818190813064575, "learning_rate": 9.357556361594882e-05, "loss": 0.0195, "step": 12300 }, { "grad_norm": 0.22730080783367157, "learning_rate": 9.356204331407917e-05, "loss": 0.0208, "step": 12310 }, { "grad_norm": 0.2326461523771286, "learning_rate": 9.354850977921094e-05, "loss": 0.0157, "step": 12320 }, { "grad_norm": 0.1910976767539978, "learning_rate": 9.353496301545529e-05, "loss": 0.0195, "step": 12330 }, { "grad_norm": 0.23990578949451447, "learning_rate": 9.352140302692733e-05, "loss": 0.0232, "step": 12340 }, { "grad_norm": 0.17841969430446625, "learning_rate": 9.350782981774627e-05, "loss": 0.0181, "step": 12350 }, { "grad_norm": 0.28358274698257446, "learning_rate": 9.349424339203526e-05, "loss": 0.0237, "step": 12360 }, { "grad_norm": 0.3098521828651428, "learning_rate": 9.34806437539215e-05, "loss": 0.0187, "step": 12370 }, { "grad_norm": 0.21671871840953827, "learning_rate": 9.346703090753622e-05, "loss": 0.0193, "step": 12380 }, { "grad_norm": 0.1971525400876999, "learning_rate": 9.345340485701461e-05, "loss": 0.0185, "step": 12390 }, { "grad_norm": 0.16668559610843658, "learning_rate": 9.343976560649595e-05, "loss": 0.0217, "step": 12400 }, { "grad_norm": 0.20619092881679535, "learning_rate": 9.342611316012344e-05, "loss": 0.0186, "step": 12410 }, { "grad_norm": 0.19261892139911652, "learning_rate": 9.341244752204437e-05, "loss": 0.0176, "step": 12420 }, { "grad_norm": 0.2366665154695511, "learning_rate": 9.339876869640995e-05, "loss": 0.0186, "step": 12430 }, { "grad_norm": 0.1898844987154007, "learning_rate": 9.33850766873755e-05, "loss": 0.0189, "step": 12440 }, { "grad_norm": 0.2583731412887573, "learning_rate": 9.337137149910028e-05, "loss": 0.0263, "step": 12450 }, { "grad_norm": 0.26790666580200195, "learning_rate": 9.335765313574753e-05, "loss": 0.0203, "step": 12460 }, { "grad_norm": 0.30164650082588196, "learning_rate": 9.334392160148457e-05, "loss": 0.0236, "step": 12470 }, { "grad_norm": 0.2460765838623047, "learning_rate": 9.333017690048264e-05, "loss": 0.018, "step": 12480 }, { "grad_norm": 0.1946788877248764, "learning_rate": 9.331641903691706e-05, "loss": 0.0176, "step": 12490 }, { "grad_norm": 0.20961545407772064, "learning_rate": 9.330264801496707e-05, "loss": 0.0176, "step": 12500 }, { "grad_norm": 0.22169815003871918, "learning_rate": 9.328886383881594e-05, "loss": 0.0181, "step": 12510 }, { "grad_norm": 0.23392219841480255, "learning_rate": 9.327506651265095e-05, "loss": 0.0178, "step": 12520 }, { "grad_norm": 0.21396932005882263, "learning_rate": 9.326125604066338e-05, "loss": 0.022, "step": 12530 }, { "grad_norm": 0.20200613141059875, "learning_rate": 9.324743242704847e-05, "loss": 0.019, "step": 12540 }, { "grad_norm": 0.17662924528121948, "learning_rate": 9.323359567600546e-05, "loss": 0.0178, "step": 12550 }, { "grad_norm": 0.2284477949142456, "learning_rate": 9.321974579173761e-05, "loss": 0.015, "step": 12560 }, { "grad_norm": 0.2107972651720047, "learning_rate": 9.320588277845213e-05, "loss": 0.0178, "step": 12570 }, { "grad_norm": 0.21704347431659698, "learning_rate": 9.319200664036026e-05, "loss": 0.0168, "step": 12580 }, { "grad_norm": 0.2310636043548584, "learning_rate": 9.31781173816772e-05, "loss": 0.0167, "step": 12590 }, { "grad_norm": 0.17053933441638947, "learning_rate": 9.316421500662212e-05, "loss": 0.0183, "step": 12600 }, { "grad_norm": 0.27594414353370667, "learning_rate": 9.31502995194182e-05, "loss": 0.0181, "step": 12610 }, { "grad_norm": 0.2537136971950531, "learning_rate": 9.31363709242926e-05, "loss": 0.0192, "step": 12620 }, { "grad_norm": 0.27299875020980835, "learning_rate": 9.312242922547647e-05, "loss": 0.0182, "step": 12630 }, { "grad_norm": 0.26084569096565247, "learning_rate": 9.310847442720492e-05, "loss": 0.0179, "step": 12640 }, { "grad_norm": 0.24863143265247345, "learning_rate": 9.309450653371706e-05, "loss": 0.0178, "step": 12650 }, { "grad_norm": 0.2105208933353424, "learning_rate": 9.308052554925595e-05, "loss": 0.0159, "step": 12660 }, { "grad_norm": 0.23017339408397675, "learning_rate": 9.306653147806867e-05, "loss": 0.0163, "step": 12670 }, { "grad_norm": 0.2565554082393646, "learning_rate": 9.305252432440622e-05, "loss": 0.0203, "step": 12680 }, { "grad_norm": 0.22982409596443176, "learning_rate": 9.303850409252361e-05, "loss": 0.021, "step": 12690 }, { "grad_norm": 0.32730284333229065, "learning_rate": 9.302447078667985e-05, "loss": 0.0176, "step": 12700 }, { "grad_norm": 0.1748044788837433, "learning_rate": 9.301042441113783e-05, "loss": 0.0175, "step": 12710 }, { "grad_norm": 0.18458148837089539, "learning_rate": 9.299636497016451e-05, "loss": 0.0179, "step": 12720 }, { "grad_norm": 0.1981426179409027, "learning_rate": 9.298229246803076e-05, "loss": 0.0157, "step": 12730 }, { "grad_norm": 0.23914232850074768, "learning_rate": 9.296820690901144e-05, "loss": 0.0175, "step": 12740 }, { "grad_norm": 0.15699847042560577, "learning_rate": 9.295410829738539e-05, "loss": 0.0182, "step": 12750 }, { "grad_norm": 0.22517462074756622, "learning_rate": 9.293999663743535e-05, "loss": 0.0177, "step": 12760 }, { "grad_norm": 0.26979151368141174, "learning_rate": 9.292587193344813e-05, "loss": 0.0189, "step": 12770 }, { "grad_norm": 0.23430663347244263, "learning_rate": 9.291173418971437e-05, "loss": 0.0211, "step": 12780 }, { "grad_norm": 0.26493844389915466, "learning_rate": 9.28975834105288e-05, "loss": 0.0174, "step": 12790 }, { "grad_norm": 0.24645568430423737, "learning_rate": 9.288341960019004e-05, "loss": 0.02, "step": 12800 }, { "grad_norm": 0.1895293891429901, "learning_rate": 9.286924276300067e-05, "loss": 0.0166, "step": 12810 }, { "grad_norm": 0.2685120701789856, "learning_rate": 9.285505290326726e-05, "loss": 0.0203, "step": 12820 }, { "grad_norm": 0.2001100480556488, "learning_rate": 9.284085002530027e-05, "loss": 0.0171, "step": 12830 }, { "grad_norm": 0.21655787527561188, "learning_rate": 9.282663413341422e-05, "loss": 0.0173, "step": 12840 }, { "grad_norm": 0.2340943068265915, "learning_rate": 9.281240523192747e-05, "loss": 0.0172, "step": 12850 }, { "grad_norm": 0.21618707478046417, "learning_rate": 9.279816332516242e-05, "loss": 0.0188, "step": 12860 }, { "grad_norm": 0.26520639657974243, "learning_rate": 9.278390841744536e-05, "loss": 0.0175, "step": 12870 }, { "grad_norm": 0.19241797924041748, "learning_rate": 9.276964051310658e-05, "loss": 0.0166, "step": 12880 }, { "grad_norm": 0.22555850446224213, "learning_rate": 9.275535961648027e-05, "loss": 0.0198, "step": 12890 }, { "grad_norm": 0.21344318985939026, "learning_rate": 9.274106573190459e-05, "loss": 0.0179, "step": 12900 }, { "grad_norm": 0.24040445685386658, "learning_rate": 9.272675886372168e-05, "loss": 0.0176, "step": 12910 }, { "grad_norm": 0.2762855887413025, "learning_rate": 9.271243901627754e-05, "loss": 0.0182, "step": 12920 }, { "grad_norm": 0.2043735533952713, "learning_rate": 9.269810619392219e-05, "loss": 0.0179, "step": 12930 }, { "grad_norm": 0.22528837621212006, "learning_rate": 9.268376040100955e-05, "loss": 0.0202, "step": 12940 }, { "grad_norm": 0.23544466495513916, "learning_rate": 9.266940164189752e-05, "loss": 0.0171, "step": 12950 }, { "grad_norm": 0.2416199892759323, "learning_rate": 9.265502992094787e-05, "loss": 0.0187, "step": 12960 }, { "grad_norm": 0.26482725143432617, "learning_rate": 9.264064524252638e-05, "loss": 0.0169, "step": 12970 }, { "grad_norm": 0.2127596139907837, "learning_rate": 9.262624761100271e-05, "loss": 0.0193, "step": 12980 }, { "grad_norm": 0.24193450808525085, "learning_rate": 9.261183703075051e-05, "loss": 0.0242, "step": 12990 }, { "grad_norm": 0.21412479877471924, "learning_rate": 9.259741350614733e-05, "loss": 0.0168, "step": 13000 }, { "grad_norm": 0.18633344769477844, "learning_rate": 9.258297704157464e-05, "loss": 0.0183, "step": 13010 }, { "grad_norm": 0.2088470309972763, "learning_rate": 9.256852764141786e-05, "loss": 0.0189, "step": 13020 }, { "grad_norm": 0.2091560959815979, "learning_rate": 9.255406531006634e-05, "loss": 0.0177, "step": 13030 }, { "grad_norm": 0.19239622354507446, "learning_rate": 9.253959005191335e-05, "loss": 0.0219, "step": 13040 }, { "grad_norm": 0.21062783896923065, "learning_rate": 9.25251018713561e-05, "loss": 0.0205, "step": 13050 }, { "grad_norm": 0.18175527453422546, "learning_rate": 9.251060077279571e-05, "loss": 0.016, "step": 13060 }, { "grad_norm": 0.18987125158309937, "learning_rate": 9.249608676063724e-05, "loss": 0.0176, "step": 13070 }, { "grad_norm": 0.27277863025665283, "learning_rate": 9.248155983928964e-05, "loss": 0.0168, "step": 13080 }, { "grad_norm": 0.2922001779079437, "learning_rate": 9.246702001316583e-05, "loss": 0.0179, "step": 13090 }, { "grad_norm": 0.25792238116264343, "learning_rate": 9.245246728668262e-05, "loss": 0.0155, "step": 13100 }, { "grad_norm": 0.22047075629234314, "learning_rate": 9.243790166426073e-05, "loss": 0.0209, "step": 13110 }, { "grad_norm": 0.1479089856147766, "learning_rate": 9.242332315032484e-05, "loss": 0.0179, "step": 13120 }, { "grad_norm": 0.17811845242977142, "learning_rate": 9.240873174930349e-05, "loss": 0.0178, "step": 13130 }, { "grad_norm": 0.1687275469303131, "learning_rate": 9.239412746562917e-05, "loss": 0.0195, "step": 13140 }, { "grad_norm": 0.22178813815116882, "learning_rate": 9.237951030373828e-05, "loss": 0.0199, "step": 13150 }, { "grad_norm": 0.29291853308677673, "learning_rate": 9.236488026807113e-05, "loss": 0.0188, "step": 13160 }, { "grad_norm": 0.2229311466217041, "learning_rate": 9.235023736307193e-05, "loss": 0.0213, "step": 13170 }, { "grad_norm": 0.2918712794780731, "learning_rate": 9.233558159318881e-05, "loss": 0.0179, "step": 13180 }, { "grad_norm": 0.24563221633434296, "learning_rate": 9.232091296287382e-05, "loss": 0.0196, "step": 13190 }, { "grad_norm": 0.19480276107788086, "learning_rate": 9.230623147658288e-05, "loss": 0.0186, "step": 13200 }, { "grad_norm": 0.23967286944389343, "learning_rate": 9.229153713877586e-05, "loss": 0.0178, "step": 13210 }, { "grad_norm": 0.20350033044815063, "learning_rate": 9.227682995391649e-05, "loss": 0.02, "step": 13220 }, { "grad_norm": 0.28798019886016846, "learning_rate": 9.226210992647243e-05, "loss": 0.0213, "step": 13230 }, { "grad_norm": 0.2410394549369812, "learning_rate": 9.224737706091525e-05, "loss": 0.0224, "step": 13240 }, { "grad_norm": 0.15446338057518005, "learning_rate": 9.223263136172039e-05, "loss": 0.018, "step": 13250 }, { "grad_norm": 0.23401226103305817, "learning_rate": 9.22178728333672e-05, "loss": 0.0175, "step": 13260 }, { "grad_norm": 0.16846776008605957, "learning_rate": 9.220310148033897e-05, "loss": 0.0174, "step": 13270 }, { "grad_norm": 0.16731376945972443, "learning_rate": 9.21883173071228e-05, "loss": 0.0178, "step": 13280 }, { "grad_norm": 0.21082374453544617, "learning_rate": 9.217352031820976e-05, "loss": 0.0142, "step": 13290 }, { "grad_norm": 0.16853810846805573, "learning_rate": 9.215871051809477e-05, "loss": 0.0183, "step": 13300 }, { "grad_norm": 0.28896060585975647, "learning_rate": 9.214388791127666e-05, "loss": 0.0182, "step": 13310 }, { "grad_norm": 0.2429724782705307, "learning_rate": 9.212905250225814e-05, "loss": 0.0186, "step": 13320 }, { "grad_norm": 0.22670549154281616, "learning_rate": 9.211420429554583e-05, "loss": 0.0187, "step": 13330 }, { "grad_norm": 0.24197036027908325, "learning_rate": 9.209934329565022e-05, "loss": 0.0173, "step": 13340 }, { "grad_norm": 0.25024232268333435, "learning_rate": 9.208446950708568e-05, "loss": 0.0171, "step": 13350 }, { "grad_norm": 0.21584255993366241, "learning_rate": 9.20695829343705e-05, "loss": 0.0184, "step": 13360 }, { "grad_norm": 0.23744958639144897, "learning_rate": 9.205468358202678e-05, "loss": 0.0181, "step": 13370 }, { "grad_norm": 0.2093486189842224, "learning_rate": 9.203977145458059e-05, "loss": 0.017, "step": 13380 }, { "grad_norm": 0.23350675404071808, "learning_rate": 9.202484655656182e-05, "loss": 0.0152, "step": 13390 }, { "grad_norm": 0.195745050907135, "learning_rate": 9.200990889250427e-05, "loss": 0.0192, "step": 13400 }, { "grad_norm": 0.2241508662700653, "learning_rate": 9.19949584669456e-05, "loss": 0.0166, "step": 13410 }, { "grad_norm": 0.1851331889629364, "learning_rate": 9.197999528442738e-05, "loss": 0.0163, "step": 13420 }, { "grad_norm": 0.25323882699012756, "learning_rate": 9.196501934949499e-05, "loss": 0.0187, "step": 13430 }, { "grad_norm": 0.21738767623901367, "learning_rate": 9.195003066669776e-05, "loss": 0.0209, "step": 13440 }, { "grad_norm": 0.2131127566099167, "learning_rate": 9.193502924058884e-05, "loss": 0.0175, "step": 13450 }, { "grad_norm": 0.26615285873413086, "learning_rate": 9.192001507572526e-05, "loss": 0.0191, "step": 13460 }, { "grad_norm": 0.2423596829175949, "learning_rate": 9.190498817666793e-05, "loss": 0.0227, "step": 13470 }, { "grad_norm": 0.37432172894477844, "learning_rate": 9.188994854798163e-05, "loss": 0.019, "step": 13480 }, { "grad_norm": 0.20273643732070923, "learning_rate": 9.187489619423499e-05, "loss": 0.0164, "step": 13490 }, { "grad_norm": 0.17925502359867096, "learning_rate": 9.185983112000056e-05, "loss": 0.0162, "step": 13500 }, { "grad_norm": 0.21169701218605042, "learning_rate": 9.184475332985464e-05, "loss": 0.0177, "step": 13510 }, { "grad_norm": 0.2196681797504425, "learning_rate": 9.182966282837754e-05, "loss": 0.0152, "step": 13520 }, { "grad_norm": 0.22814275324344635, "learning_rate": 9.18145596201533e-05, "loss": 0.0152, "step": 13530 }, { "grad_norm": 0.21951165795326233, "learning_rate": 9.179944370976991e-05, "loss": 0.0147, "step": 13540 }, { "grad_norm": 0.198689267039299, "learning_rate": 9.178431510181918e-05, "loss": 0.0162, "step": 13550 }, { "grad_norm": 0.22830738127231598, "learning_rate": 9.176917380089675e-05, "loss": 0.022, "step": 13560 }, { "grad_norm": 0.24331767857074738, "learning_rate": 9.175401981160219e-05, "loss": 0.0179, "step": 13570 }, { "grad_norm": 0.3099052309989929, "learning_rate": 9.173885313853885e-05, "loss": 0.0171, "step": 13580 }, { "grad_norm": 0.2785036265850067, "learning_rate": 9.172367378631398e-05, "loss": 0.0184, "step": 13590 }, { "grad_norm": 0.27014437317848206, "learning_rate": 9.170848175953866e-05, "loss": 0.0166, "step": 13600 }, { "grad_norm": 0.2292494922876358, "learning_rate": 9.169327706282784e-05, "loss": 0.0181, "step": 13610 }, { "grad_norm": 0.35050785541534424, "learning_rate": 9.167805970080029e-05, "loss": 0.0172, "step": 13620 }, { "grad_norm": 0.2158464640378952, "learning_rate": 9.166282967807864e-05, "loss": 0.0191, "step": 13630 }, { "grad_norm": 0.21893194317817688, "learning_rate": 9.16475869992894e-05, "loss": 0.0176, "step": 13640 }, { "grad_norm": 0.2623606324195862, "learning_rate": 9.163233166906284e-05, "loss": 0.0171, "step": 13650 }, { "grad_norm": 0.19747573137283325, "learning_rate": 9.161706369203317e-05, "loss": 0.016, "step": 13660 }, { "grad_norm": 0.21809999644756317, "learning_rate": 9.16017830728384e-05, "loss": 0.0173, "step": 13670 }, { "grad_norm": 0.2610160708427429, "learning_rate": 9.158648981612035e-05, "loss": 0.0156, "step": 13680 }, { "grad_norm": 0.17865203320980072, "learning_rate": 9.157118392652472e-05, "loss": 0.0172, "step": 13690 }, { "grad_norm": 0.18325503170490265, "learning_rate": 9.155586540870104e-05, "loss": 0.016, "step": 13700 }, { "grad_norm": 0.21935522556304932, "learning_rate": 9.154053426730267e-05, "loss": 0.0179, "step": 13710 }, { "grad_norm": 0.272200345993042, "learning_rate": 9.15251905069868e-05, "loss": 0.0194, "step": 13720 }, { "grad_norm": 0.2666842043399811, "learning_rate": 9.150983413241446e-05, "loss": 0.0172, "step": 13730 }, { "grad_norm": 0.2319265902042389, "learning_rate": 9.149446514825051e-05, "loss": 0.02, "step": 13740 }, { "grad_norm": 0.23430608212947845, "learning_rate": 9.147908355916365e-05, "loss": 0.0161, "step": 13750 }, { "grad_norm": 0.21302464604377747, "learning_rate": 9.146368936982642e-05, "loss": 0.0166, "step": 13760 }, { "grad_norm": 0.17321933805942535, "learning_rate": 9.144828258491511e-05, "loss": 0.0201, "step": 13770 }, { "grad_norm": 0.2247217446565628, "learning_rate": 9.143286320910996e-05, "loss": 0.0209, "step": 13780 }, { "grad_norm": 0.29129037261009216, "learning_rate": 9.141743124709491e-05, "loss": 0.0196, "step": 13790 }, { "grad_norm": 0.25223758816719055, "learning_rate": 9.140198670355784e-05, "loss": 0.019, "step": 13800 }, { "grad_norm": 0.26697373390197754, "learning_rate": 9.138652958319034e-05, "loss": 0.0163, "step": 13810 }, { "grad_norm": 0.25239089131355286, "learning_rate": 9.137105989068791e-05, "loss": 0.0177, "step": 13820 }, { "grad_norm": 0.1914084553718567, "learning_rate": 9.135557763074983e-05, "loss": 0.0155, "step": 13830 }, { "grad_norm": 0.17219898104667664, "learning_rate": 9.13400828080792e-05, "loss": 0.0155, "step": 13840 }, { "grad_norm": 0.20189906656742096, "learning_rate": 9.132457542738292e-05, "loss": 0.0175, "step": 13850 }, { "grad_norm": 0.18364781141281128, "learning_rate": 9.130905549337174e-05, "loss": 0.019, "step": 13860 }, { "grad_norm": 0.21569961309432983, "learning_rate": 9.129352301076021e-05, "loss": 0.0176, "step": 13870 }, { "grad_norm": 0.21820808947086334, "learning_rate": 9.127797798426668e-05, "loss": 0.0182, "step": 13880 }, { "grad_norm": 0.24859902262687683, "learning_rate": 9.126242041861333e-05, "loss": 0.0149, "step": 13890 }, { "grad_norm": 0.26561594009399414, "learning_rate": 9.124685031852611e-05, "loss": 0.0179, "step": 13900 }, { "grad_norm": 0.2115054428577423, "learning_rate": 9.123126768873482e-05, "loss": 0.0149, "step": 13910 }, { "grad_norm": 0.21076783537864685, "learning_rate": 9.121567253397308e-05, "loss": 0.0154, "step": 13920 }, { "grad_norm": 0.2232057750225067, "learning_rate": 9.120006485897824e-05, "loss": 0.0152, "step": 13930 }, { "grad_norm": 0.2492823749780655, "learning_rate": 9.118444466849152e-05, "loss": 0.0163, "step": 13940 }, { "grad_norm": 0.18412922322750092, "learning_rate": 9.116881196725793e-05, "loss": 0.0182, "step": 13950 }, { "grad_norm": 0.2989897131919861, "learning_rate": 9.115316676002627e-05, "loss": 0.016, "step": 13960 }, { "grad_norm": 0.22460976243019104, "learning_rate": 9.113750905154911e-05, "loss": 0.017, "step": 13970 }, { "grad_norm": 0.2465679794549942, "learning_rate": 9.112183884658289e-05, "loss": 0.0185, "step": 13980 }, { "grad_norm": 0.23415116965770721, "learning_rate": 9.11061561498878e-05, "loss": 0.0195, "step": 13990 }, { "grad_norm": 0.2203013151884079, "learning_rate": 9.109046096622779e-05, "loss": 0.0211, "step": 14000 }, { "grad_norm": 0.23157034814357758, "learning_rate": 9.107475330037069e-05, "loss": 0.014, "step": 14010 }, { "grad_norm": 0.2106088399887085, "learning_rate": 9.105903315708806e-05, "loss": 0.0171, "step": 14020 }, { "grad_norm": 0.2497757375240326, "learning_rate": 9.104330054115524e-05, "loss": 0.0163, "step": 14030 }, { "grad_norm": 0.17736709117889404, "learning_rate": 9.102755545735141e-05, "loss": 0.0197, "step": 14040 }, { "grad_norm": 0.24250797927379608, "learning_rate": 9.10117979104595e-05, "loss": 0.0192, "step": 14050 }, { "grad_norm": 0.22558243572711945, "learning_rate": 9.099602790526624e-05, "loss": 0.0176, "step": 14060 }, { "grad_norm": 0.18382121622562408, "learning_rate": 9.098024544656212e-05, "loss": 0.0183, "step": 14070 }, { "grad_norm": 0.22038744390010834, "learning_rate": 9.096445053914148e-05, "loss": 0.0168, "step": 14080 }, { "grad_norm": 0.2185136377811432, "learning_rate": 9.094864318780236e-05, "loss": 0.0188, "step": 14090 }, { "grad_norm": 0.21107357740402222, "learning_rate": 9.093282339734663e-05, "loss": 0.0175, "step": 14100 }, { "grad_norm": 0.22706148028373718, "learning_rate": 9.091699117257992e-05, "loss": 0.0162, "step": 14110 }, { "grad_norm": 0.21090058982372284, "learning_rate": 9.090114651831163e-05, "loss": 0.0181, "step": 14120 }, { "grad_norm": 0.26566582918167114, "learning_rate": 9.088528943935497e-05, "loss": 0.0175, "step": 14130 }, { "grad_norm": 0.1879085749387741, "learning_rate": 9.086941994052689e-05, "loss": 0.0171, "step": 14140 }, { "grad_norm": 0.24638588726520538, "learning_rate": 9.085353802664813e-05, "loss": 0.0179, "step": 14150 }, { "grad_norm": 0.22964157164096832, "learning_rate": 9.08376437025432e-05, "loss": 0.0197, "step": 14160 }, { "grad_norm": 0.23122712969779968, "learning_rate": 9.082173697304035e-05, "loss": 0.017, "step": 14170 }, { "grad_norm": 0.23373879492282867, "learning_rate": 9.080581784297166e-05, "loss": 0.0216, "step": 14180 }, { "grad_norm": 0.24180062115192413, "learning_rate": 9.078988631717291e-05, "loss": 0.0175, "step": 14190 }, { "grad_norm": 0.213382288813591, "learning_rate": 9.077394240048369e-05, "loss": 0.0196, "step": 14200 }, { "grad_norm": 0.24612833559513092, "learning_rate": 9.075798609774736e-05, "loss": 0.0142, "step": 14210 }, { "grad_norm": 0.19768348336219788, "learning_rate": 9.0742017413811e-05, "loss": 0.0161, "step": 14220 }, { "grad_norm": 0.21749545633792877, "learning_rate": 9.072603635352548e-05, "loss": 0.016, "step": 14230 }, { "grad_norm": 0.20184111595153809, "learning_rate": 9.071004292174541e-05, "loss": 0.0183, "step": 14240 }, { "grad_norm": 0.19649778306484222, "learning_rate": 9.06940371233292e-05, "loss": 0.0188, "step": 14250 }, { "grad_norm": 0.2352328896522522, "learning_rate": 9.067801896313898e-05, "loss": 0.017, "step": 14260 }, { "grad_norm": 0.16892951726913452, "learning_rate": 9.066198844604064e-05, "loss": 0.0187, "step": 14270 }, { "grad_norm": 0.22742608189582825, "learning_rate": 9.06459455769038e-05, "loss": 0.0176, "step": 14280 }, { "grad_norm": 0.23805542290210724, "learning_rate": 9.062989036060193e-05, "loss": 0.0178, "step": 14290 }, { "grad_norm": 0.19140012562274933, "learning_rate": 9.061382280201212e-05, "loss": 0.0204, "step": 14300 }, { "grad_norm": 0.17034006118774414, "learning_rate": 9.059774290601528e-05, "loss": 0.0156, "step": 14310 }, { "grad_norm": 0.2727762758731842, "learning_rate": 9.058165067749606e-05, "loss": 0.0179, "step": 14320 }, { "grad_norm": 0.22248575091362, "learning_rate": 9.056554612134288e-05, "loss": 0.0174, "step": 14330 }, { "grad_norm": 0.17916861176490784, "learning_rate": 9.054942924244785e-05, "loss": 0.0158, "step": 14340 }, { "grad_norm": 0.22458846867084503, "learning_rate": 9.053330004570686e-05, "loss": 0.0176, "step": 14350 }, { "grad_norm": 0.22475259006023407, "learning_rate": 9.051715853601955e-05, "loss": 0.0181, "step": 14360 }, { "grad_norm": 0.18387079238891602, "learning_rate": 9.050100471828926e-05, "loss": 0.014, "step": 14370 }, { "grad_norm": 0.21298523247241974, "learning_rate": 9.048483859742311e-05, "loss": 0.0148, "step": 14380 }, { "grad_norm": 0.3328697383403778, "learning_rate": 9.046866017833193e-05, "loss": 0.0186, "step": 14390 }, { "grad_norm": 0.1982872635126114, "learning_rate": 9.045246946593029e-05, "loss": 0.0168, "step": 14400 }, { "grad_norm": 0.23502874374389648, "learning_rate": 9.043626646513652e-05, "loss": 0.0183, "step": 14410 }, { "grad_norm": 0.2302819788455963, "learning_rate": 9.042005118087267e-05, "loss": 0.0169, "step": 14420 }, { "grad_norm": 0.20800118148326874, "learning_rate": 9.040382361806448e-05, "loss": 0.0148, "step": 14430 }, { "grad_norm": 0.17737683653831482, "learning_rate": 9.038758378164148e-05, "loss": 0.015, "step": 14440 }, { "grad_norm": 0.1971062868833542, "learning_rate": 9.037133167653691e-05, "loss": 0.0161, "step": 14450 }, { "grad_norm": 0.20580241084098816, "learning_rate": 9.035506730768771e-05, "loss": 0.0168, "step": 14460 }, { "grad_norm": 0.19729848206043243, "learning_rate": 9.033879068003458e-05, "loss": 0.0174, "step": 14470 }, { "grad_norm": 0.22762610018253326, "learning_rate": 9.032250179852193e-05, "loss": 0.0161, "step": 14480 }, { "grad_norm": 0.16001790761947632, "learning_rate": 9.030620066809787e-05, "loss": 0.0171, "step": 14490 }, { "grad_norm": 0.19905957579612732, "learning_rate": 9.028988729371428e-05, "loss": 0.0143, "step": 14500 }, { "grad_norm": 0.24017615616321564, "learning_rate": 9.027356168032673e-05, "loss": 0.017, "step": 14510 }, { "grad_norm": 0.2014440894126892, "learning_rate": 9.02572238328945e-05, "loss": 0.0141, "step": 14520 }, { "grad_norm": 0.18788889050483704, "learning_rate": 9.02408737563806e-05, "loss": 0.0147, "step": 14530 }, { "grad_norm": 0.2750775218009949, "learning_rate": 9.022451145575174e-05, "loss": 0.0182, "step": 14540 }, { "grad_norm": 0.22969920933246613, "learning_rate": 9.02081369359784e-05, "loss": 0.0169, "step": 14550 }, { "grad_norm": 0.21618176996707916, "learning_rate": 9.019175020203465e-05, "loss": 0.0158, "step": 14560 }, { "grad_norm": 0.18880584836006165, "learning_rate": 9.017535125889842e-05, "loss": 0.0171, "step": 14570 }, { "grad_norm": 0.23030924797058105, "learning_rate": 9.015894011155124e-05, "loss": 0.0158, "step": 14580 }, { "grad_norm": 0.2626030147075653, "learning_rate": 9.014251676497838e-05, "loss": 0.0166, "step": 14590 }, { "grad_norm": 0.2218409776687622, "learning_rate": 9.012608122416884e-05, "loss": 0.0149, "step": 14600 }, { "grad_norm": 0.15640629827976227, "learning_rate": 9.010963349411529e-05, "loss": 0.0157, "step": 14610 }, { "grad_norm": 0.2926141619682312, "learning_rate": 9.00931735798141e-05, "loss": 0.0179, "step": 14620 }, { "grad_norm": 0.22980338335037231, "learning_rate": 9.00767014862654e-05, "loss": 0.0173, "step": 14630 }, { "grad_norm": 0.18593169748783112, "learning_rate": 9.006021721847295e-05, "loss": 0.0165, "step": 14640 }, { "grad_norm": 0.23980310559272766, "learning_rate": 9.004372078144423e-05, "loss": 0.0148, "step": 14650 }, { "grad_norm": 0.20433032512664795, "learning_rate": 9.002721218019043e-05, "loss": 0.0202, "step": 14660 }, { "grad_norm": 0.210842564702034, "learning_rate": 9.001069141972642e-05, "loss": 0.0155, "step": 14670 }, { "grad_norm": 0.1825425773859024, "learning_rate": 8.99941585050708e-05, "loss": 0.017, "step": 14680 }, { "grad_norm": 0.2634999752044678, "learning_rate": 8.997761344124578e-05, "loss": 0.0169, "step": 14690 }, { "grad_norm": 0.18641966581344604, "learning_rate": 8.996105623327737e-05, "loss": 0.0168, "step": 14700 }, { "grad_norm": 0.15901626646518707, "learning_rate": 8.994448688619517e-05, "loss": 0.0138, "step": 14710 }, { "grad_norm": 0.1493789553642273, "learning_rate": 8.992790540503253e-05, "loss": 0.0163, "step": 14720 }, { "grad_norm": 0.17090336978435516, "learning_rate": 8.991131179482648e-05, "loss": 0.0167, "step": 14730 }, { "grad_norm": 0.2294936627149582, "learning_rate": 8.989470606061768e-05, "loss": 0.019, "step": 14740 }, { "grad_norm": 0.20543819665908813, "learning_rate": 8.987808820745056e-05, "loss": 0.0162, "step": 14750 }, { "grad_norm": 0.1917973756790161, "learning_rate": 8.986145824037315e-05, "loss": 0.0138, "step": 14760 }, { "grad_norm": 0.18578708171844482, "learning_rate": 8.984481616443721e-05, "loss": 0.0166, "step": 14770 }, { "grad_norm": 0.18895868957042694, "learning_rate": 8.982816198469815e-05, "loss": 0.0154, "step": 14780 }, { "grad_norm": 0.26122504472732544, "learning_rate": 8.98114957062151e-05, "loss": 0.0188, "step": 14790 }, { "grad_norm": 0.18684130907058716, "learning_rate": 8.97948173340508e-05, "loss": 0.0174, "step": 14800 }, { "grad_norm": 0.19470994174480438, "learning_rate": 8.977812687327172e-05, "loss": 0.0154, "step": 14810 }, { "grad_norm": 0.21654631197452545, "learning_rate": 8.976142432894798e-05, "loss": 0.0183, "step": 14820 }, { "grad_norm": 0.13286633789539337, "learning_rate": 8.974470970615336e-05, "loss": 0.0171, "step": 14830 }, { "grad_norm": 0.18548937141895294, "learning_rate": 8.972798300996534e-05, "loss": 0.0173, "step": 14840 }, { "grad_norm": 0.23307941854000092, "learning_rate": 8.971124424546504e-05, "loss": 0.0162, "step": 14850 }, { "grad_norm": 0.1599283516407013, "learning_rate": 8.969449341773724e-05, "loss": 0.0135, "step": 14860 }, { "grad_norm": 0.19248218834400177, "learning_rate": 8.967773053187042e-05, "loss": 0.0152, "step": 14870 }, { "grad_norm": 0.20084917545318604, "learning_rate": 8.966095559295668e-05, "loss": 0.0172, "step": 14880 }, { "grad_norm": 0.21476879715919495, "learning_rate": 8.964416860609184e-05, "loss": 0.0163, "step": 14890 }, { "grad_norm": 0.18413971364498138, "learning_rate": 8.962736957637532e-05, "loss": 0.0147, "step": 14900 }, { "grad_norm": 0.219417542219162, "learning_rate": 8.96105585089102e-05, "loss": 0.0135, "step": 14910 }, { "grad_norm": 0.24737659096717834, "learning_rate": 8.959373540880329e-05, "loss": 0.0195, "step": 14920 }, { "grad_norm": 0.19265951216220856, "learning_rate": 8.957690028116495e-05, "loss": 0.018, "step": 14930 }, { "grad_norm": 0.20546934008598328, "learning_rate": 8.956005313110928e-05, "loss": 0.0192, "step": 14940 }, { "grad_norm": 0.21917493641376495, "learning_rate": 8.9543193963754e-05, "loss": 0.0148, "step": 14950 }, { "grad_norm": 0.2122030407190323, "learning_rate": 8.952632278422048e-05, "loss": 0.0141, "step": 14960 }, { "grad_norm": 0.20635296404361725, "learning_rate": 8.95094395976337e-05, "loss": 0.0207, "step": 14970 }, { "grad_norm": 0.18281425535678864, "learning_rate": 8.949254440912239e-05, "loss": 0.0163, "step": 14980 }, { "grad_norm": 0.1869470328092575, "learning_rate": 8.94756372238188e-05, "loss": 0.0149, "step": 14990 }, { "grad_norm": 0.19570378959178925, "learning_rate": 8.945871804685892e-05, "loss": 0.018, "step": 15000 }, { "grad_norm": 0.22860829532146454, "learning_rate": 8.944178688338236e-05, "loss": 0.0158, "step": 15010 }, { "grad_norm": 0.1931675374507904, "learning_rate": 8.942484373853233e-05, "loss": 0.0159, "step": 15020 }, { "grad_norm": 0.19285239279270172, "learning_rate": 8.940788861745572e-05, "loss": 0.0141, "step": 15030 }, { "grad_norm": 0.22303298115730286, "learning_rate": 8.939092152530308e-05, "loss": 0.0148, "step": 15040 }, { "grad_norm": 0.2180754542350769, "learning_rate": 8.937394246722853e-05, "loss": 0.0148, "step": 15050 }, { "grad_norm": 0.19640479981899261, "learning_rate": 8.935695144838984e-05, "loss": 0.0137, "step": 15060 }, { "grad_norm": 0.22813983261585236, "learning_rate": 8.933994847394849e-05, "loss": 0.0161, "step": 15070 }, { "grad_norm": 0.19255611300468445, "learning_rate": 8.932293354906949e-05, "loss": 0.0171, "step": 15080 }, { "grad_norm": 0.1827450692653656, "learning_rate": 8.930590667892153e-05, "loss": 0.0164, "step": 15090 }, { "grad_norm": 0.1481238603591919, "learning_rate": 8.928886786867696e-05, "loss": 0.0156, "step": 15100 }, { "grad_norm": 0.24410656094551086, "learning_rate": 8.927181712351168e-05, "loss": 0.0176, "step": 15110 }, { "grad_norm": 0.17143884301185608, "learning_rate": 8.925475444860527e-05, "loss": 0.0149, "step": 15120 }, { "grad_norm": 0.1986059844493866, "learning_rate": 8.923767984914092e-05, "loss": 0.0158, "step": 15130 }, { "grad_norm": 0.18132229149341583, "learning_rate": 8.922059333030545e-05, "loss": 0.02, "step": 15140 }, { "grad_norm": 0.2147589921951294, "learning_rate": 8.920349489728928e-05, "loss": 0.0202, "step": 15150 }, { "grad_norm": 0.2588476538658142, "learning_rate": 8.918638455528646e-05, "loss": 0.0176, "step": 15160 }, { "grad_norm": 0.23283778131008148, "learning_rate": 8.916926230949468e-05, "loss": 0.0181, "step": 15170 }, { "grad_norm": 0.2648862898349762, "learning_rate": 8.915212816511522e-05, "loss": 0.0164, "step": 15180 }, { "grad_norm": 0.2110104262828827, "learning_rate": 8.913498212735296e-05, "loss": 0.0198, "step": 15190 }, { "grad_norm": 0.24691520631313324, "learning_rate": 8.911782420141643e-05, "loss": 0.016, "step": 15200 }, { "grad_norm": 0.15366147458553314, "learning_rate": 8.910065439251775e-05, "loss": 0.0182, "step": 15210 }, { "grad_norm": 0.1988476812839508, "learning_rate": 8.908347270587268e-05, "loss": 0.0155, "step": 15220 }, { "grad_norm": 0.24126051366329193, "learning_rate": 8.906627914670054e-05, "loss": 0.0157, "step": 15230 }, { "grad_norm": 0.1933906376361847, "learning_rate": 8.904907372022427e-05, "loss": 0.0166, "step": 15240 }, { "grad_norm": 0.23052538931369781, "learning_rate": 8.903185643167042e-05, "loss": 0.0148, "step": 15250 }, { "grad_norm": 0.21419814229011536, "learning_rate": 8.901462728626919e-05, "loss": 0.0172, "step": 15260 }, { "grad_norm": 0.19982944428920746, "learning_rate": 8.899738628925429e-05, "loss": 0.0141, "step": 15270 }, { "grad_norm": 0.21320581436157227, "learning_rate": 8.898013344586312e-05, "loss": 0.0171, "step": 15280 }, { "grad_norm": 0.2594001889228821, "learning_rate": 8.896286876133661e-05, "loss": 0.0144, "step": 15290 }, { "grad_norm": 0.22878071665763855, "learning_rate": 8.894559224091933e-05, "loss": 0.0145, "step": 15300 }, { "grad_norm": 0.173085555434227, "learning_rate": 8.892830388985942e-05, "loss": 0.0141, "step": 15310 }, { "grad_norm": 0.24604995548725128, "learning_rate": 8.891100371340864e-05, "loss": 0.0161, "step": 15320 }, { "grad_norm": 0.21016032993793488, "learning_rate": 8.889369171682231e-05, "loss": 0.0146, "step": 15330 }, { "grad_norm": 0.24345511198043823, "learning_rate": 8.887636790535936e-05, "loss": 0.0161, "step": 15340 }, { "grad_norm": 0.26949355006217957, "learning_rate": 8.885903228428231e-05, "loss": 0.017, "step": 15350 }, { "grad_norm": 0.22191259264945984, "learning_rate": 8.884168485885727e-05, "loss": 0.0166, "step": 15360 }, { "grad_norm": 0.2008524388074875, "learning_rate": 8.882432563435393e-05, "loss": 0.0183, "step": 15370 }, { "grad_norm": 0.24217045307159424, "learning_rate": 8.880695461604556e-05, "loss": 0.0191, "step": 15380 }, { "grad_norm": 0.17606694996356964, "learning_rate": 8.878957180920901e-05, "loss": 0.0167, "step": 15390 }, { "grad_norm": 0.1924494206905365, "learning_rate": 8.877217721912473e-05, "loss": 0.0188, "step": 15400 }, { "grad_norm": 0.22953690588474274, "learning_rate": 8.875477085107673e-05, "loss": 0.0178, "step": 15410 }, { "grad_norm": 0.22758758068084717, "learning_rate": 8.87373527103526e-05, "loss": 0.0166, "step": 15420 }, { "grad_norm": 0.17671015858650208, "learning_rate": 8.871992280224353e-05, "loss": 0.0167, "step": 15430 }, { "grad_norm": 0.19471372663974762, "learning_rate": 8.870248113204422e-05, "loss": 0.018, "step": 15440 }, { "grad_norm": 0.6065483689308167, "learning_rate": 8.868502770505306e-05, "loss": 0.0154, "step": 15450 }, { "grad_norm": 0.19966688752174377, "learning_rate": 8.86675625265719e-05, "loss": 0.0179, "step": 15460 }, { "grad_norm": 0.201427161693573, "learning_rate": 8.865008560190618e-05, "loss": 0.0198, "step": 15470 }, { "grad_norm": 0.25346630811691284, "learning_rate": 8.863259693636496e-05, "loss": 0.0191, "step": 15480 }, { "grad_norm": 0.13502062857151031, "learning_rate": 8.861509653526083e-05, "loss": 0.0218, "step": 15490 }, { "grad_norm": 0.23682238161563873, "learning_rate": 8.859758440390993e-05, "loss": 0.02, "step": 15500 }, { "grad_norm": 0.1986692249774933, "learning_rate": 8.858006054763202e-05, "loss": 0.0189, "step": 15510 }, { "grad_norm": 0.20935127139091492, "learning_rate": 8.856252497175035e-05, "loss": 0.0177, "step": 15520 }, { "grad_norm": 0.14411357045173645, "learning_rate": 8.854497768159178e-05, "loss": 0.0156, "step": 15530 }, { "grad_norm": 0.18942229449748993, "learning_rate": 8.852741868248671e-05, "loss": 0.0155, "step": 15540 }, { "grad_norm": 0.270003080368042, "learning_rate": 8.85098479797691e-05, "loss": 0.0178, "step": 15550 }, { "grad_norm": 0.20625045895576477, "learning_rate": 8.849226557877646e-05, "loss": 0.0167, "step": 15560 }, { "grad_norm": 0.19137108325958252, "learning_rate": 8.84746714848499e-05, "loss": 0.0145, "step": 15570 }, { "grad_norm": 0.17116907238960266, "learning_rate": 8.845706570333397e-05, "loss": 0.0144, "step": 15580 }, { "grad_norm": 0.17588812112808228, "learning_rate": 8.84394482395769e-05, "loss": 0.016, "step": 15590 }, { "grad_norm": 0.1561909019947052, "learning_rate": 8.842181909893038e-05, "loss": 0.016, "step": 15600 }, { "grad_norm": 0.20123985409736633, "learning_rate": 8.840417828674969e-05, "loss": 0.0159, "step": 15610 }, { "grad_norm": 0.1989007145166397, "learning_rate": 8.838652580839364e-05, "loss": 0.013, "step": 15620 }, { "grad_norm": 0.15784704685211182, "learning_rate": 8.836886166922458e-05, "loss": 0.0168, "step": 15630 }, { "grad_norm": 0.16841182112693787, "learning_rate": 8.835118587460844e-05, "loss": 0.0163, "step": 15640 }, { "grad_norm": 0.20564918220043182, "learning_rate": 8.83334984299146e-05, "loss": 0.0167, "step": 15650 }, { "grad_norm": 0.23779746890068054, "learning_rate": 8.83157993405161e-05, "loss": 0.0175, "step": 15660 }, { "grad_norm": 0.17120246589183807, "learning_rate": 8.829808861178943e-05, "loss": 0.0168, "step": 15670 }, { "grad_norm": 0.19882643222808838, "learning_rate": 8.828036624911464e-05, "loss": 0.016, "step": 15680 }, { "grad_norm": 0.27655699849128723, "learning_rate": 8.826263225787532e-05, "loss": 0.0152, "step": 15690 }, { "grad_norm": 0.24084356427192688, "learning_rate": 8.824488664345858e-05, "loss": 0.0143, "step": 15700 }, { "grad_norm": 0.22857138514518738, "learning_rate": 8.822712941125508e-05, "loss": 0.0157, "step": 15710 }, { "grad_norm": 0.14316870272159576, "learning_rate": 8.820936056665898e-05, "loss": 0.0151, "step": 15720 }, { "grad_norm": 0.24181853234767914, "learning_rate": 8.819158011506801e-05, "loss": 0.0171, "step": 15730 }, { "grad_norm": 0.24201105535030365, "learning_rate": 8.81737880618834e-05, "loss": 0.0163, "step": 15740 }, { "grad_norm": 0.15717950463294983, "learning_rate": 8.815598441250987e-05, "loss": 0.0161, "step": 15750 }, { "grad_norm": 0.26285773515701294, "learning_rate": 8.813816917235576e-05, "loss": 0.0183, "step": 15760 }, { "grad_norm": 0.18913525342941284, "learning_rate": 8.812034234683282e-05, "loss": 0.017, "step": 15770 }, { "grad_norm": 0.1877979040145874, "learning_rate": 8.810250394135637e-05, "loss": 0.0165, "step": 15780 }, { "grad_norm": 0.23154231905937195, "learning_rate": 8.808465396134529e-05, "loss": 0.0159, "step": 15790 }, { "grad_norm": 0.19914856553077698, "learning_rate": 8.806679241222189e-05, "loss": 0.0168, "step": 15800 }, { "grad_norm": 0.24097314476966858, "learning_rate": 8.804891929941203e-05, "loss": 0.0154, "step": 15810 }, { "grad_norm": 0.1810251921415329, "learning_rate": 8.803103462834514e-05, "loss": 0.0185, "step": 15820 }, { "grad_norm": 0.16897618770599365, "learning_rate": 8.801313840445408e-05, "loss": 0.0135, "step": 15830 }, { "grad_norm": 0.1874612718820572, "learning_rate": 8.799523063317524e-05, "loss": 0.0188, "step": 15840 }, { "grad_norm": 0.1721617579460144, "learning_rate": 8.797731131994854e-05, "loss": 0.0156, "step": 15850 }, { "grad_norm": 0.24075952172279358, "learning_rate": 8.795938047021739e-05, "loss": 0.017, "step": 15860 }, { "grad_norm": 0.17625336349010468, "learning_rate": 8.794143808942872e-05, "loss": 0.0153, "step": 15870 }, { "grad_norm": 0.17968788743019104, "learning_rate": 8.792348418303296e-05, "loss": 0.0164, "step": 15880 }, { "grad_norm": 0.19727401435375214, "learning_rate": 8.790551875648398e-05, "loss": 0.0157, "step": 15890 }, { "grad_norm": 0.183676078915596, "learning_rate": 8.788754181523926e-05, "loss": 0.0182, "step": 15900 }, { "grad_norm": 0.18711458146572113, "learning_rate": 8.78695533647597e-05, "loss": 0.0169, "step": 15910 }, { "grad_norm": 0.1827874481678009, "learning_rate": 8.785155341050972e-05, "loss": 0.0203, "step": 15920 }, { "grad_norm": 0.29817062616348267, "learning_rate": 8.783354195795721e-05, "loss": 0.0159, "step": 15930 }, { "grad_norm": 0.22955390810966492, "learning_rate": 8.78155190125736e-05, "loss": 0.0154, "step": 15940 }, { "grad_norm": 0.23405328392982483, "learning_rate": 8.779748457983378e-05, "loss": 0.0191, "step": 15950 }, { "grad_norm": 0.23362132906913757, "learning_rate": 8.777943866521612e-05, "loss": 0.0151, "step": 15960 }, { "grad_norm": 0.1822373867034912, "learning_rate": 8.77613812742025e-05, "loss": 0.0153, "step": 15970 }, { "grad_norm": 0.1565285176038742, "learning_rate": 8.774331241227829e-05, "loss": 0.015, "step": 15980 }, { "grad_norm": 0.2089439183473587, "learning_rate": 8.772523208493232e-05, "loss": 0.0153, "step": 15990 }, { "grad_norm": 0.18043085932731628, "learning_rate": 8.770714029765692e-05, "loss": 0.0193, "step": 16000 }, { "grad_norm": 0.16770553588867188, "learning_rate": 8.768903705594789e-05, "loss": 0.0168, "step": 16010 }, { "grad_norm": 0.27399465441703796, "learning_rate": 8.767092236530453e-05, "loss": 0.0189, "step": 16020 }, { "grad_norm": 0.17062222957611084, "learning_rate": 8.76527962312296e-05, "loss": 0.0167, "step": 16030 }, { "grad_norm": 0.16855956614017487, "learning_rate": 8.763465865922934e-05, "loss": 0.015, "step": 16040 }, { "grad_norm": 0.20092900097370148, "learning_rate": 8.761650965481347e-05, "loss": 0.0155, "step": 16050 }, { "grad_norm": 0.19846290349960327, "learning_rate": 8.759834922349516e-05, "loss": 0.0165, "step": 16060 }, { "grad_norm": 0.20824801921844482, "learning_rate": 8.758017737079108e-05, "loss": 0.0138, "step": 16070 }, { "grad_norm": 0.1992178112268448, "learning_rate": 8.756199410222137e-05, "loss": 0.017, "step": 16080 }, { "grad_norm": 0.2405758947134018, "learning_rate": 8.754379942330963e-05, "loss": 0.0143, "step": 16090 }, { "grad_norm": 0.24901534616947174, "learning_rate": 8.75255933395829e-05, "loss": 0.0145, "step": 16100 }, { "grad_norm": 0.16219821572303772, "learning_rate": 8.750737585657171e-05, "loss": 0.0135, "step": 16110 }, { "grad_norm": 0.18079328536987305, "learning_rate": 8.748914697981008e-05, "loss": 0.0148, "step": 16120 }, { "grad_norm": 0.17693714797496796, "learning_rate": 8.747090671483542e-05, "loss": 0.016, "step": 16130 }, { "grad_norm": 0.21565929055213928, "learning_rate": 8.745265506718869e-05, "loss": 0.0154, "step": 16140 }, { "grad_norm": 0.233937606215477, "learning_rate": 8.74343920424142e-05, "loss": 0.0155, "step": 16150 }, { "grad_norm": 0.2384694218635559, "learning_rate": 8.741611764605982e-05, "loss": 0.0158, "step": 16160 }, { "grad_norm": 0.16872212290763855, "learning_rate": 8.739783188367682e-05, "loss": 0.0153, "step": 16170 }, { "grad_norm": 0.18897618353366852, "learning_rate": 8.737953476081991e-05, "loss": 0.0172, "step": 16180 }, { "grad_norm": 0.1768166869878769, "learning_rate": 8.73612262830473e-05, "loss": 0.0158, "step": 16190 }, { "grad_norm": 0.1895921528339386, "learning_rate": 8.734290645592061e-05, "loss": 0.0152, "step": 16200 }, { "grad_norm": 0.21544484794139862, "learning_rate": 8.732457528500493e-05, "loss": 0.0149, "step": 16210 }, { "grad_norm": 0.15938787162303925, "learning_rate": 8.730623277586875e-05, "loss": 0.016, "step": 16220 }, { "grad_norm": 0.14480537176132202, "learning_rate": 8.72878789340841e-05, "loss": 0.0149, "step": 16230 }, { "grad_norm": 0.25370094180107117, "learning_rate": 8.726951376522635e-05, "loss": 0.0167, "step": 16240 }, { "grad_norm": 0.21313831210136414, "learning_rate": 8.725113727487435e-05, "loss": 0.0122, "step": 16250 }, { "grad_norm": 0.1730213612318039, "learning_rate": 8.723274946861042e-05, "loss": 0.0166, "step": 16260 }, { "grad_norm": 0.20616716146469116, "learning_rate": 8.721435035202026e-05, "loss": 0.0148, "step": 16270 }, { "grad_norm": 0.22210009396076202, "learning_rate": 8.719593993069306e-05, "loss": 0.0185, "step": 16280 }, { "grad_norm": 0.1827392727136612, "learning_rate": 8.717751821022139e-05, "loss": 0.0141, "step": 16290 }, { "grad_norm": 0.23069880902767181, "learning_rate": 8.715908519620134e-05, "loss": 0.0161, "step": 16300 }, { "grad_norm": 0.20266062021255493, "learning_rate": 8.71406408942323e-05, "loss": 0.0166, "step": 16310 }, { "grad_norm": 0.23980361223220825, "learning_rate": 8.712218530991723e-05, "loss": 0.0158, "step": 16320 }, { "grad_norm": 0.1888474076986313, "learning_rate": 8.710371844886241e-05, "loss": 0.0143, "step": 16330 }, { "grad_norm": 0.16847030818462372, "learning_rate": 8.708524031667758e-05, "loss": 0.0138, "step": 16340 }, { "grad_norm": 0.2047513872385025, "learning_rate": 8.706675091897592e-05, "loss": 0.0162, "step": 16350 }, { "grad_norm": 0.1654508113861084, "learning_rate": 8.704825026137404e-05, "loss": 0.017, "step": 16360 }, { "grad_norm": 0.21469715237617493, "learning_rate": 8.702973834949192e-05, "loss": 0.0135, "step": 16370 }, { "grad_norm": 0.21048323810100555, "learning_rate": 8.701121518895301e-05, "loss": 0.0129, "step": 16380 }, { "grad_norm": 0.239071324467659, "learning_rate": 8.699268078538414e-05, "loss": 0.0136, "step": 16390 }, { "grad_norm": 0.13595813512802124, "learning_rate": 8.69741351444156e-05, "loss": 0.0179, "step": 16400 }, { "grad_norm": 0.18900099396705627, "learning_rate": 8.695557827168101e-05, "loss": 0.016, "step": 16410 }, { "grad_norm": 0.2541065216064453, "learning_rate": 8.693701017281753e-05, "loss": 0.0164, "step": 16420 }, { "grad_norm": 0.3018171489238739, "learning_rate": 8.691843085346563e-05, "loss": 0.0133, "step": 16430 }, { "grad_norm": 0.2026258260011673, "learning_rate": 8.689984031926919e-05, "loss": 0.0135, "step": 16440 }, { "grad_norm": 0.2711614966392517, "learning_rate": 8.688123857587555e-05, "loss": 0.0173, "step": 16450 }, { "grad_norm": 0.26419416069984436, "learning_rate": 8.686262562893544e-05, "loss": 0.0185, "step": 16460 }, { "grad_norm": 0.15856164693832397, "learning_rate": 8.684400148410294e-05, "loss": 0.014, "step": 16470 }, { "grad_norm": 0.17374126613140106, "learning_rate": 8.682536614703562e-05, "loss": 0.0165, "step": 16480 }, { "grad_norm": 0.14036087691783905, "learning_rate": 8.680671962339437e-05, "loss": 0.0145, "step": 16490 }, { "grad_norm": 0.3152315318584442, "learning_rate": 8.678806191884352e-05, "loss": 0.0156, "step": 16500 }, { "grad_norm": 0.1801459640264511, "learning_rate": 8.67693930390508e-05, "loss": 0.0122, "step": 16510 }, { "grad_norm": 0.1562817096710205, "learning_rate": 8.67507129896873e-05, "loss": 0.016, "step": 16520 }, { "grad_norm": 0.24163508415222168, "learning_rate": 8.673202177642757e-05, "loss": 0.0167, "step": 16530 }, { "grad_norm": 0.19927525520324707, "learning_rate": 8.671331940494945e-05, "loss": 0.0136, "step": 16540 }, { "grad_norm": 0.20604665577411652, "learning_rate": 8.669460588093427e-05, "loss": 0.0128, "step": 16550 }, { "grad_norm": 0.16649146378040314, "learning_rate": 8.667588121006667e-05, "loss": 0.0165, "step": 16560 }, { "grad_norm": 0.19990748167037964, "learning_rate": 8.665714539803475e-05, "loss": 0.0161, "step": 16570 }, { "grad_norm": 0.17887336015701294, "learning_rate": 8.663839845052993e-05, "loss": 0.0136, "step": 16580 }, { "grad_norm": 0.20955242216587067, "learning_rate": 8.661964037324703e-05, "loss": 0.0153, "step": 16590 }, { "grad_norm": 0.16096679866313934, "learning_rate": 8.660087117188427e-05, "loss": 0.0157, "step": 16600 }, { "grad_norm": 0.1325441300868988, "learning_rate": 8.658209085214325e-05, "loss": 0.0128, "step": 16610 }, { "grad_norm": 0.21701374650001526, "learning_rate": 8.656329941972891e-05, "loss": 0.0143, "step": 16620 }, { "grad_norm": 0.19403286278247833, "learning_rate": 8.654449688034963e-05, "loss": 0.0179, "step": 16630 }, { "grad_norm": 0.19598205387592316, "learning_rate": 8.652568323971706e-05, "loss": 0.0152, "step": 16640 }, { "grad_norm": 0.2251700460910797, "learning_rate": 8.650685850354636e-05, "loss": 0.0201, "step": 16650 }, { "grad_norm": 0.22157494723796844, "learning_rate": 8.648802267755593e-05, "loss": 0.016, "step": 16660 }, { "grad_norm": 0.17070917785167694, "learning_rate": 8.646917576746764e-05, "loss": 0.0196, "step": 16670 }, { "grad_norm": 0.2189841866493225, "learning_rate": 8.645031777900666e-05, "loss": 0.0148, "step": 16680 }, { "grad_norm": 0.18965278565883636, "learning_rate": 8.643144871790154e-05, "loss": 0.0142, "step": 16690 }, { "grad_norm": 0.16044025123119354, "learning_rate": 8.641256858988424e-05, "loss": 0.0144, "step": 16700 }, { "grad_norm": 0.14943727850914001, "learning_rate": 8.639367740069e-05, "loss": 0.0151, "step": 16710 }, { "grad_norm": 0.17754395306110382, "learning_rate": 8.63747751560575e-05, "loss": 0.0135, "step": 16720 }, { "grad_norm": 0.21049398183822632, "learning_rate": 8.635586186172871e-05, "loss": 0.0149, "step": 16730 }, { "grad_norm": 0.17809589207172394, "learning_rate": 8.633693752344902e-05, "loss": 0.0172, "step": 16740 }, { "grad_norm": 0.2097519189119339, "learning_rate": 8.631800214696713e-05, "loss": 0.0149, "step": 16750 }, { "grad_norm": 0.2415781170129776, "learning_rate": 8.629905573803511e-05, "loss": 0.0155, "step": 16760 }, { "grad_norm": 0.274127334356308, "learning_rate": 8.628009830240839e-05, "loss": 0.0164, "step": 16770 }, { "grad_norm": 0.23608514666557312, "learning_rate": 8.626112984584571e-05, "loss": 0.0165, "step": 16780 }, { "grad_norm": 0.17068035900592804, "learning_rate": 8.62421503741092e-05, "loss": 0.0143, "step": 16790 }, { "grad_norm": 0.17299747467041016, "learning_rate": 8.622315989296432e-05, "loss": 0.0147, "step": 16800 }, { "grad_norm": 0.17907561361789703, "learning_rate": 8.62041584081799e-05, "loss": 0.0116, "step": 16810 }, { "grad_norm": 0.24970760941505432, "learning_rate": 8.618514592552807e-05, "loss": 0.0146, "step": 16820 }, { "grad_norm": 0.19467633962631226, "learning_rate": 8.616612245078431e-05, "loss": 0.015, "step": 16830 }, { "grad_norm": 0.21927247941493988, "learning_rate": 8.614708798972746e-05, "loss": 0.0141, "step": 16840 }, { "grad_norm": 0.20855718851089478, "learning_rate": 8.61280425481397e-05, "loss": 0.0163, "step": 16850 }, { "grad_norm": 0.16303063929080963, "learning_rate": 8.61089861318065e-05, "loss": 0.0129, "step": 16860 }, { "grad_norm": 0.23376691341400146, "learning_rate": 8.608991874651673e-05, "loss": 0.0146, "step": 16870 }, { "grad_norm": 0.18266955018043518, "learning_rate": 8.607084039806255e-05, "loss": 0.0139, "step": 16880 }, { "grad_norm": 0.21190030872821808, "learning_rate": 8.605175109223944e-05, "loss": 0.0144, "step": 16890 }, { "grad_norm": 0.18038742244243622, "learning_rate": 8.603265083484624e-05, "loss": 0.0144, "step": 16900 }, { "grad_norm": 0.18079710006713867, "learning_rate": 8.60135396316851e-05, "loss": 0.0128, "step": 16910 }, { "grad_norm": 0.2323245406150818, "learning_rate": 8.599441748856152e-05, "loss": 0.0182, "step": 16920 }, { "grad_norm": 0.17718388140201569, "learning_rate": 8.597528441128427e-05, "loss": 0.0126, "step": 16930 }, { "grad_norm": 0.21379902958869934, "learning_rate": 8.595614040566549e-05, "loss": 0.0142, "step": 16940 }, { "grad_norm": 0.16303518414497375, "learning_rate": 8.593698547752063e-05, "loss": 0.0159, "step": 16950 }, { "grad_norm": 0.19656163454055786, "learning_rate": 8.591781963266843e-05, "loss": 0.0174, "step": 16960 }, { "grad_norm": 0.22497794032096863, "learning_rate": 8.5898642876931e-05, "loss": 0.017, "step": 16970 }, { "grad_norm": 0.21739748120307922, "learning_rate": 8.587945521613369e-05, "loss": 0.0141, "step": 16980 }, { "grad_norm": 0.230473592877388, "learning_rate": 8.586025665610524e-05, "loss": 0.0131, "step": 16990 }, { "grad_norm": 0.2047458291053772, "learning_rate": 8.584104720267765e-05, "loss": 0.0163, "step": 17000 }, { "grad_norm": 0.2574958801269531, "learning_rate": 8.582182686168625e-05, "loss": 0.0195, "step": 17010 }, { "grad_norm": 0.2092573344707489, "learning_rate": 8.580259563896967e-05, "loss": 0.0164, "step": 17020 }, { "grad_norm": 0.18015334010124207, "learning_rate": 8.578335354036983e-05, "loss": 0.0144, "step": 17030 }, { "grad_norm": 0.19187568128108978, "learning_rate": 8.576410057173201e-05, "loss": 0.0154, "step": 17040 }, { "grad_norm": 0.20787006616592407, "learning_rate": 8.574483673890474e-05, "loss": 0.015, "step": 17050 }, { "grad_norm": 0.21486352384090424, "learning_rate": 8.572556204773983e-05, "loss": 0.0153, "step": 17060 }, { "grad_norm": 0.17208783328533173, "learning_rate": 8.570627650409246e-05, "loss": 0.0158, "step": 17070 }, { "grad_norm": 0.19660937786102295, "learning_rate": 8.568698011382107e-05, "loss": 0.0132, "step": 17080 }, { "grad_norm": 0.17066003382205963, "learning_rate": 8.566767288278738e-05, "loss": 0.0171, "step": 17090 }, { "grad_norm": 0.1476115882396698, "learning_rate": 8.56483548168564e-05, "loss": 0.013, "step": 17100 }, { "grad_norm": 0.21268866956233978, "learning_rate": 8.562902592189648e-05, "loss": 0.0136, "step": 17110 }, { "grad_norm": 0.16706977784633636, "learning_rate": 8.560968620377921e-05, "loss": 0.015, "step": 17120 }, { "grad_norm": 0.20766694843769073, "learning_rate": 8.559033566837951e-05, "loss": 0.0143, "step": 17130 }, { "grad_norm": 0.18332284688949585, "learning_rate": 8.557097432157551e-05, "loss": 0.0158, "step": 17140 }, { "grad_norm": 0.25220832228660583, "learning_rate": 8.555160216924872e-05, "loss": 0.0144, "step": 17150 }, { "grad_norm": 0.1920144408941269, "learning_rate": 8.55322192172839e-05, "loss": 0.0165, "step": 17160 }, { "grad_norm": 0.2129826694726944, "learning_rate": 8.551282547156902e-05, "loss": 0.0129, "step": 17170 }, { "grad_norm": 0.1988976001739502, "learning_rate": 8.549342093799544e-05, "loss": 0.0133, "step": 17180 }, { "grad_norm": 0.18441089987754822, "learning_rate": 8.547400562245773e-05, "loss": 0.017, "step": 17190 }, { "grad_norm": 0.22001414000988007, "learning_rate": 8.545457953085374e-05, "loss": 0.0133, "step": 17200 }, { "grad_norm": 0.20350047945976257, "learning_rate": 8.543514266908463e-05, "loss": 0.0126, "step": 17210 }, { "grad_norm": 0.18411315977573395, "learning_rate": 8.541569504305478e-05, "loss": 0.016, "step": 17220 }, { "grad_norm": 0.2236071676015854, "learning_rate": 8.539623665867187e-05, "loss": 0.0181, "step": 17230 }, { "grad_norm": 0.16912561655044556, "learning_rate": 8.537676752184685e-05, "loss": 0.0158, "step": 17240 }, { "grad_norm": 0.22901912033557892, "learning_rate": 8.53572876384939e-05, "loss": 0.0144, "step": 17250 }, { "grad_norm": 0.20381100475788116, "learning_rate": 8.533779701453056e-05, "loss": 0.0173, "step": 17260 }, { "grad_norm": 0.23318757116794586, "learning_rate": 8.53182956558775e-05, "loss": 0.025, "step": 17270 }, { "grad_norm": 0.2114330530166626, "learning_rate": 8.529878356845877e-05, "loss": 0.0216, "step": 17280 }, { "grad_norm": 0.20636996626853943, "learning_rate": 8.527926075820158e-05, "loss": 0.0183, "step": 17290 }, { "grad_norm": 0.3009740710258484, "learning_rate": 8.525972723103648e-05, "loss": 0.0193, "step": 17300 }, { "grad_norm": 0.16768674552440643, "learning_rate": 8.524018299289722e-05, "loss": 0.0178, "step": 17310 }, { "grad_norm": 0.15977013111114502, "learning_rate": 8.522062804972083e-05, "loss": 0.0176, "step": 17320 }, { "grad_norm": 0.22208331525325775, "learning_rate": 8.520106240744759e-05, "loss": 0.0188, "step": 17330 }, { "grad_norm": 0.2799769639968872, "learning_rate": 8.518148607202102e-05, "loss": 0.0175, "step": 17340 }, { "grad_norm": 0.1926897019147873, "learning_rate": 8.51618990493879e-05, "loss": 0.0191, "step": 17350 }, { "grad_norm": 0.18070824444293976, "learning_rate": 8.514230134549823e-05, "loss": 0.0154, "step": 17360 }, { "grad_norm": 0.12492633610963821, "learning_rate": 8.51226929663053e-05, "loss": 0.0136, "step": 17370 }, { "grad_norm": 0.24923555552959442, "learning_rate": 8.51030739177656e-05, "loss": 0.0143, "step": 17380 }, { "grad_norm": 0.1786850094795227, "learning_rate": 8.508344420583889e-05, "loss": 0.0151, "step": 17390 }, { "grad_norm": 0.1840001493692398, "learning_rate": 8.506380383648816e-05, "loss": 0.0172, "step": 17400 }, { "grad_norm": 0.24508507549762726, "learning_rate": 8.504415281567963e-05, "loss": 0.0148, "step": 17410 }, { "grad_norm": 0.16506506502628326, "learning_rate": 8.502449114938275e-05, "loss": 0.0145, "step": 17420 }, { "grad_norm": 0.17777517437934875, "learning_rate": 8.500481884357025e-05, "loss": 0.0143, "step": 17430 }, { "grad_norm": 0.21596962213516235, "learning_rate": 8.498513590421801e-05, "loss": 0.0157, "step": 17440 }, { "grad_norm": 0.2398998886346817, "learning_rate": 8.496544233730522e-05, "loss": 0.017, "step": 17450 }, { "grad_norm": 0.14380888640880585, "learning_rate": 8.494573814881426e-05, "loss": 0.0131, "step": 17460 }, { "grad_norm": 0.19039617478847504, "learning_rate": 8.492602334473074e-05, "loss": 0.0162, "step": 17470 }, { "grad_norm": 0.18328002095222473, "learning_rate": 8.49062979310435e-05, "loss": 0.0152, "step": 17480 }, { "grad_norm": 0.17440980672836304, "learning_rate": 8.488656191374458e-05, "loss": 0.0149, "step": 17490 }, { "grad_norm": 0.15224866569042206, "learning_rate": 8.48668152988293e-05, "loss": 0.0149, "step": 17500 }, { "grad_norm": 0.182474747300148, "learning_rate": 8.484705809229612e-05, "loss": 0.0148, "step": 17510 }, { "grad_norm": 0.2156008780002594, "learning_rate": 8.482729030014677e-05, "loss": 0.013, "step": 17520 }, { "grad_norm": 0.19168466329574585, "learning_rate": 8.48075119283862e-05, "loss": 0.0151, "step": 17530 }, { "grad_norm": 0.21452461183071136, "learning_rate": 8.478772298302254e-05, "loss": 0.0134, "step": 17540 }, { "grad_norm": 0.13943615555763245, "learning_rate": 8.476792347006716e-05, "loss": 0.0137, "step": 17550 }, { "grad_norm": 0.1778298020362854, "learning_rate": 8.474811339553462e-05, "loss": 0.0144, "step": 17560 }, { "grad_norm": 0.21441002190113068, "learning_rate": 8.47282927654427e-05, "loss": 0.016, "step": 17570 }, { "grad_norm": 0.21756087243556976, "learning_rate": 8.470846158581238e-05, "loss": 0.0152, "step": 17580 }, { "grad_norm": 0.20060868561267853, "learning_rate": 8.468861986266787e-05, "loss": 0.0161, "step": 17590 }, { "grad_norm": 0.22246797382831573, "learning_rate": 8.466876760203654e-05, "loss": 0.0137, "step": 17600 }, { "grad_norm": 0.2446524202823639, "learning_rate": 8.464890480994898e-05, "loss": 0.016, "step": 17610 }, { "grad_norm": 0.16192401945590973, "learning_rate": 8.462903149243899e-05, "loss": 0.0149, "step": 17620 }, { "grad_norm": 0.26447898149490356, "learning_rate": 8.460914765554357e-05, "loss": 0.0166, "step": 17630 }, { "grad_norm": 0.23324547708034515, "learning_rate": 8.458925330530288e-05, "loss": 0.0156, "step": 17640 }, { "grad_norm": 0.25521981716156006, "learning_rate": 8.456934844776032e-05, "loss": 0.0126, "step": 17650 }, { "grad_norm": 0.15651030838489532, "learning_rate": 8.454943308896246e-05, "loss": 0.0129, "step": 17660 }, { "grad_norm": 0.21618057787418365, "learning_rate": 8.452950723495905e-05, "loss": 0.0152, "step": 17670 }, { "grad_norm": 0.21572807431221008, "learning_rate": 8.450957089180303e-05, "loss": 0.0177, "step": 17680 }, { "grad_norm": 0.23275405168533325, "learning_rate": 8.448962406555055e-05, "loss": 0.0149, "step": 17690 }, { "grad_norm": 0.20587804913520813, "learning_rate": 8.446966676226093e-05, "loss": 0.0158, "step": 17700 }, { "grad_norm": 0.21927987039089203, "learning_rate": 8.444969898799667e-05, "loss": 0.0136, "step": 17710 }, { "grad_norm": 0.1613103747367859, "learning_rate": 8.442972074882343e-05, "loss": 0.0166, "step": 17720 }, { "grad_norm": 0.18372711539268494, "learning_rate": 8.44097320508101e-05, "loss": 0.014, "step": 17730 }, { "grad_norm": 0.17765338718891144, "learning_rate": 8.43897329000287e-05, "loss": 0.0183, "step": 17740 }, { "grad_norm": 0.1649331897497177, "learning_rate": 8.436972330255448e-05, "loss": 0.0134, "step": 17750 }, { "grad_norm": 0.21857552230358124, "learning_rate": 8.434970326446579e-05, "loss": 0.0211, "step": 17760 }, { "grad_norm": 0.19466882944107056, "learning_rate": 8.432967279184418e-05, "loss": 0.0144, "step": 17770 }, { "grad_norm": 0.1765463650226593, "learning_rate": 8.430963189077441e-05, "loss": 0.0143, "step": 17780 }, { "grad_norm": 0.20863625407218933, "learning_rate": 8.428958056734437e-05, "loss": 0.0143, "step": 17790 }, { "grad_norm": 0.20499958097934723, "learning_rate": 8.426951882764513e-05, "loss": 0.0141, "step": 17800 }, { "grad_norm": 0.19222374260425568, "learning_rate": 8.424944667777089e-05, "loss": 0.0179, "step": 17810 }, { "grad_norm": 0.2233310490846634, "learning_rate": 8.422936412381905e-05, "loss": 0.0162, "step": 17820 }, { "grad_norm": 0.22461383044719696, "learning_rate": 8.420927117189017e-05, "loss": 0.013, "step": 17830 }, { "grad_norm": 0.19208909571170807, "learning_rate": 8.418916782808795e-05, "loss": 0.015, "step": 17840 }, { "grad_norm": 0.16311047971248627, "learning_rate": 8.416905409851926e-05, "loss": 0.0137, "step": 17850 }, { "grad_norm": 0.19757120311260223, "learning_rate": 8.41489299892941e-05, "loss": 0.0147, "step": 17860 }, { "grad_norm": 0.219007208943367, "learning_rate": 8.412879550652566e-05, "loss": 0.0177, "step": 17870 }, { "grad_norm": 0.2103228121995926, "learning_rate": 8.410865065633029e-05, "loss": 0.0144, "step": 17880 }, { "grad_norm": 0.2382318675518036, "learning_rate": 8.408849544482742e-05, "loss": 0.017, "step": 17890 }, { "grad_norm": 0.23297849297523499, "learning_rate": 8.406832987813968e-05, "loss": 0.0157, "step": 17900 }, { "grad_norm": 0.20323225855827332, "learning_rate": 8.404815396239286e-05, "loss": 0.0173, "step": 17910 }, { "grad_norm": 0.24990025162696838, "learning_rate": 8.402796770371587e-05, "loss": 0.0173, "step": 17920 }, { "grad_norm": 0.21142995357513428, "learning_rate": 8.400777110824071e-05, "loss": 0.0136, "step": 17930 }, { "grad_norm": 0.20246335864067078, "learning_rate": 8.398756418210263e-05, "loss": 0.0156, "step": 17940 }, { "grad_norm": 0.18478555977344513, "learning_rate": 8.396734693143993e-05, "loss": 0.0113, "step": 17950 }, { "grad_norm": 0.17914941906929016, "learning_rate": 8.39471193623941e-05, "loss": 0.0162, "step": 17960 }, { "grad_norm": 0.3457590341567993, "learning_rate": 8.392688148110974e-05, "loss": 0.0144, "step": 17970 }, { "grad_norm": 0.20722530782222748, "learning_rate": 8.390663329373456e-05, "loss": 0.013, "step": 17980 }, { "grad_norm": 0.24545671045780182, "learning_rate": 8.388637480641944e-05, "loss": 0.0162, "step": 17990 }, { "grad_norm": 0.18773916363716125, "learning_rate": 8.386610602531837e-05, "loss": 0.0145, "step": 18000 }, { "grad_norm": 0.19873075187206268, "learning_rate": 8.384582695658847e-05, "loss": 0.0153, "step": 18010 }, { "grad_norm": 0.16244113445281982, "learning_rate": 8.382553760638999e-05, "loss": 0.0125, "step": 18020 }, { "grad_norm": 0.17314477264881134, "learning_rate": 8.380523798088631e-05, "loss": 0.0125, "step": 18030 }, { "grad_norm": 0.19328732788562775, "learning_rate": 8.378492808624389e-05, "loss": 0.0143, "step": 18040 }, { "grad_norm": 0.18497174978256226, "learning_rate": 8.376460792863237e-05, "loss": 0.0147, "step": 18050 }, { "grad_norm": 0.1886579543352127, "learning_rate": 8.374427751422444e-05, "loss": 0.0157, "step": 18060 }, { "grad_norm": 0.21090707182884216, "learning_rate": 8.3723936849196e-05, "loss": 0.0166, "step": 18070 }, { "grad_norm": 0.219324991106987, "learning_rate": 8.370358593972595e-05, "loss": 0.0135, "step": 18080 }, { "grad_norm": 0.23930495977401733, "learning_rate": 8.36832247919964e-05, "loss": 0.0159, "step": 18090 }, { "grad_norm": 0.1941853016614914, "learning_rate": 8.36628534121925e-05, "loss": 0.0158, "step": 18100 }, { "grad_norm": 0.22383908927440643, "learning_rate": 8.364247180650254e-05, "loss": 0.0137, "step": 18110 }, { "grad_norm": 0.16728714108467102, "learning_rate": 8.362207998111794e-05, "loss": 0.0185, "step": 18120 }, { "grad_norm": 0.16648906469345093, "learning_rate": 8.360167794223318e-05, "loss": 0.0148, "step": 18130 }, { "grad_norm": 0.1553514152765274, "learning_rate": 8.358126569604586e-05, "loss": 0.0117, "step": 18140 }, { "grad_norm": 0.17497676610946655, "learning_rate": 8.356084324875668e-05, "loss": 0.0114, "step": 18150 }, { "grad_norm": 0.20871601998806, "learning_rate": 8.354041060656945e-05, "loss": 0.0142, "step": 18160 }, { "grad_norm": 0.17288286983966827, "learning_rate": 8.351996777569106e-05, "loss": 0.0128, "step": 18170 }, { "grad_norm": 0.19639484584331512, "learning_rate": 8.349951476233148e-05, "loss": 0.0149, "step": 18180 }, { "grad_norm": 0.1921977549791336, "learning_rate": 8.347905157270386e-05, "loss": 0.0165, "step": 18190 }, { "grad_norm": 0.22556458413600922, "learning_rate": 8.345857821302432e-05, "loss": 0.0161, "step": 18200 }, { "grad_norm": 0.16030456125736237, "learning_rate": 8.343809468951213e-05, "loss": 0.0138, "step": 18210 }, { "grad_norm": 0.14975617825984955, "learning_rate": 8.341760100838965e-05, "loss": 0.0132, "step": 18220 }, { "grad_norm": 0.21274641156196594, "learning_rate": 8.339709717588233e-05, "loss": 0.016, "step": 18230 }, { "grad_norm": 0.15842656791210175, "learning_rate": 8.33765831982187e-05, "loss": 0.0164, "step": 18240 }, { "grad_norm": 0.21159657835960388, "learning_rate": 8.335605908163035e-05, "loss": 0.0154, "step": 18250 }, { "grad_norm": 0.20894265174865723, "learning_rate": 8.333552483235196e-05, "loss": 0.0152, "step": 18260 }, { "grad_norm": 0.16322103142738342, "learning_rate": 8.33149804566213e-05, "loss": 0.0165, "step": 18270 }, { "grad_norm": 0.19822096824645996, "learning_rate": 8.329442596067921e-05, "loss": 0.0164, "step": 18280 }, { "grad_norm": 0.1880270093679428, "learning_rate": 8.32738613507696e-05, "loss": 0.0127, "step": 18290 }, { "grad_norm": 0.1914987862110138, "learning_rate": 8.325328663313946e-05, "loss": 0.0124, "step": 18300 }, { "grad_norm": 0.15001098811626434, "learning_rate": 8.323270181403884e-05, "loss": 0.0181, "step": 18310 }, { "grad_norm": 0.14210453629493713, "learning_rate": 8.321210689972086e-05, "loss": 0.0123, "step": 18320 }, { "grad_norm": 0.1896744966506958, "learning_rate": 8.319150189644174e-05, "loss": 0.0166, "step": 18330 }, { "grad_norm": 0.22478793561458588, "learning_rate": 8.31708868104607e-05, "loss": 0.0145, "step": 18340 }, { "grad_norm": 0.2202942818403244, "learning_rate": 8.315026164804007e-05, "loss": 0.0141, "step": 18350 }, { "grad_norm": 0.2524569034576416, "learning_rate": 8.312962641544524e-05, "loss": 0.0127, "step": 18360 }, { "grad_norm": 0.18218302726745605, "learning_rate": 8.310898111894465e-05, "loss": 0.012, "step": 18370 }, { "grad_norm": 0.23801879584789276, "learning_rate": 8.308832576480977e-05, "loss": 0.0149, "step": 18380 }, { "grad_norm": 0.17061831057071686, "learning_rate": 8.306766035931519e-05, "loss": 0.0116, "step": 18390 }, { "grad_norm": 0.19228629767894745, "learning_rate": 8.304698490873847e-05, "loss": 0.013, "step": 18400 }, { "grad_norm": 0.22220507264137268, "learning_rate": 8.30262994193603e-05, "loss": 0.0148, "step": 18410 }, { "grad_norm": 0.16693249344825745, "learning_rate": 8.300560389746438e-05, "loss": 0.0137, "step": 18420 }, { "grad_norm": 0.16961215436458588, "learning_rate": 8.298489834933745e-05, "loss": 0.016, "step": 18430 }, { "grad_norm": 0.20602042973041534, "learning_rate": 8.296418278126934e-05, "loss": 0.0148, "step": 18440 }, { "grad_norm": 0.22087234258651733, "learning_rate": 8.294345719955284e-05, "loss": 0.0158, "step": 18450 }, { "grad_norm": 0.18103636801242828, "learning_rate": 8.29227216104839e-05, "loss": 0.0122, "step": 18460 }, { "grad_norm": 0.16677148640155792, "learning_rate": 8.290197602036137e-05, "loss": 0.0126, "step": 18470 }, { "grad_norm": 0.1584021896123886, "learning_rate": 8.288122043548725e-05, "loss": 0.0125, "step": 18480 }, { "grad_norm": 0.19145570695400238, "learning_rate": 8.286045486216657e-05, "loss": 0.0176, "step": 18490 }, { "grad_norm": 0.2214629203081131, "learning_rate": 8.283967930670733e-05, "loss": 0.0156, "step": 18500 }, { "grad_norm": 0.21545736491680145, "learning_rate": 8.281889377542058e-05, "loss": 0.0145, "step": 18510 }, { "grad_norm": 0.22418957948684692, "learning_rate": 8.279809827462045e-05, "loss": 0.0145, "step": 18520 }, { "grad_norm": 0.1562269628047943, "learning_rate": 8.277729281062402e-05, "loss": 0.013, "step": 18530 }, { "grad_norm": 0.20594115555286407, "learning_rate": 8.27564773897515e-05, "loss": 0.0144, "step": 18540 }, { "grad_norm": 0.19701890647411346, "learning_rate": 8.273565201832602e-05, "loss": 0.0148, "step": 18550 }, { "grad_norm": 0.1463318169116974, "learning_rate": 8.27148167026738e-05, "loss": 0.0132, "step": 18560 }, { "grad_norm": 0.14141854643821716, "learning_rate": 8.269397144912405e-05, "loss": 0.0117, "step": 18570 }, { "grad_norm": 0.13947677612304688, "learning_rate": 8.267311626400899e-05, "loss": 0.0115, "step": 18580 }, { "grad_norm": 0.1903570592403412, "learning_rate": 8.26522511536639e-05, "loss": 0.0142, "step": 18590 }, { "grad_norm": 0.21554341912269592, "learning_rate": 8.263137612442706e-05, "loss": 0.013, "step": 18600 }, { "grad_norm": 0.16596032679080963, "learning_rate": 8.261049118263971e-05, "loss": 0.0156, "step": 18610 }, { "grad_norm": 0.23076866567134857, "learning_rate": 8.258959633464619e-05, "loss": 0.0151, "step": 18620 }, { "grad_norm": 0.44972681999206543, "learning_rate": 8.256869158679377e-05, "loss": 0.0174, "step": 18630 }, { "grad_norm": 0.17368218302726746, "learning_rate": 8.254777694543278e-05, "loss": 0.0139, "step": 18640 }, { "grad_norm": 0.16046375036239624, "learning_rate": 8.252685241691651e-05, "loss": 0.0138, "step": 18650 }, { "grad_norm": 0.15572717785835266, "learning_rate": 8.250591800760133e-05, "loss": 0.0133, "step": 18660 }, { "grad_norm": 0.15550094842910767, "learning_rate": 8.248497372384649e-05, "loss": 0.0118, "step": 18670 }, { "grad_norm": 0.2454672008752823, "learning_rate": 8.246401957201437e-05, "loss": 0.0149, "step": 18680 }, { "grad_norm": 0.2089288830757141, "learning_rate": 8.244305555847027e-05, "loss": 0.0142, "step": 18690 }, { "grad_norm": 0.1877906322479248, "learning_rate": 8.24220816895825e-05, "loss": 0.0142, "step": 18700 }, { "grad_norm": 0.1533493548631668, "learning_rate": 8.240109797172237e-05, "loss": 0.0124, "step": 18710 }, { "grad_norm": 0.20773102343082428, "learning_rate": 8.238010441126416e-05, "loss": 0.0124, "step": 18720 }, { "grad_norm": 0.19877521693706512, "learning_rate": 8.23591010145852e-05, "loss": 0.015, "step": 18730 }, { "grad_norm": 0.21221190690994263, "learning_rate": 8.233808778806571e-05, "loss": 0.0123, "step": 18740 }, { "grad_norm": 0.174085795879364, "learning_rate": 8.231706473808903e-05, "loss": 0.0129, "step": 18750 }, { "grad_norm": 0.16788165271282196, "learning_rate": 8.229603187104133e-05, "loss": 0.0111, "step": 18760 }, { "grad_norm": 0.2033764272928238, "learning_rate": 8.22749891933119e-05, "loss": 0.0125, "step": 18770 }, { "grad_norm": 0.17486941814422607, "learning_rate": 8.225393671129291e-05, "loss": 0.0132, "step": 18780 }, { "grad_norm": 0.17977885901927948, "learning_rate": 8.223287443137957e-05, "loss": 0.0112, "step": 18790 }, { "grad_norm": 0.1682451367378235, "learning_rate": 8.221180235997004e-05, "loss": 0.0133, "step": 18800 }, { "grad_norm": 0.14031682908535004, "learning_rate": 8.219072050346544e-05, "loss": 0.0134, "step": 18810 }, { "grad_norm": 0.13556478917598724, "learning_rate": 8.216962886826992e-05, "loss": 0.0162, "step": 18820 }, { "grad_norm": 0.19656401872634888, "learning_rate": 8.214852746079054e-05, "loss": 0.013, "step": 18830 }, { "grad_norm": 0.17102140188217163, "learning_rate": 8.212741628743732e-05, "loss": 0.0132, "step": 18840 }, { "grad_norm": 0.23731283843517303, "learning_rate": 8.210629535462333e-05, "loss": 0.0134, "step": 18850 }, { "grad_norm": 0.25438106060028076, "learning_rate": 8.208516466876453e-05, "loss": 0.013, "step": 18860 }, { "grad_norm": 0.15656302869319916, "learning_rate": 8.206402423627986e-05, "loss": 0.0125, "step": 18870 }, { "grad_norm": 0.18102291226387024, "learning_rate": 8.204287406359124e-05, "loss": 0.0152, "step": 18880 }, { "grad_norm": 0.15905703604221344, "learning_rate": 8.20217141571235e-05, "loss": 0.0156, "step": 18890 }, { "grad_norm": 0.21226248145103455, "learning_rate": 8.200054452330449e-05, "loss": 0.0132, "step": 18900 }, { "grad_norm": 0.2119407057762146, "learning_rate": 8.197936516856499e-05, "loss": 0.0144, "step": 18910 }, { "grad_norm": 0.22772064805030823, "learning_rate": 8.195817609933871e-05, "loss": 0.0141, "step": 18920 }, { "grad_norm": 0.25617796182632446, "learning_rate": 8.193697732206233e-05, "loss": 0.0147, "step": 18930 }, { "grad_norm": 0.2044711410999298, "learning_rate": 8.19157688431755e-05, "loss": 0.014, "step": 18940 }, { "grad_norm": 0.1760404109954834, "learning_rate": 8.189455066912077e-05, "loss": 0.0147, "step": 18950 }, { "grad_norm": 0.1744578778743744, "learning_rate": 8.187332280634369e-05, "loss": 0.0166, "step": 18960 }, { "grad_norm": 0.17901083827018738, "learning_rate": 8.18520852612927e-05, "loss": 0.0145, "step": 18970 }, { "grad_norm": 0.22220948338508606, "learning_rate": 8.183083804041921e-05, "loss": 0.0148, "step": 18980 }, { "grad_norm": 0.1635725051164627, "learning_rate": 8.180958115017757e-05, "loss": 0.0156, "step": 18990 }, { "grad_norm": 0.17216435074806213, "learning_rate": 8.178831459702505e-05, "loss": 0.0157, "step": 19000 }, { "grad_norm": 0.17520396411418915, "learning_rate": 8.17670383874219e-05, "loss": 0.0137, "step": 19010 }, { "grad_norm": 0.18210364878177643, "learning_rate": 8.174575252783124e-05, "loss": 0.0136, "step": 19020 }, { "grad_norm": 0.1878647357225418, "learning_rate": 8.172445702471914e-05, "loss": 0.0151, "step": 19030 }, { "grad_norm": 0.14308001101016998, "learning_rate": 8.170315188455466e-05, "loss": 0.011, "step": 19040 }, { "grad_norm": 0.1818666011095047, "learning_rate": 8.168183711380969e-05, "loss": 0.012, "step": 19050 }, { "grad_norm": 0.15583519637584686, "learning_rate": 8.166051271895913e-05, "loss": 0.0123, "step": 19060 }, { "grad_norm": 0.197065070271492, "learning_rate": 8.163917870648075e-05, "loss": 0.0144, "step": 19070 }, { "grad_norm": 0.15653961896896362, "learning_rate": 8.161783508285526e-05, "loss": 0.0171, "step": 19080 }, { "grad_norm": 0.21288014948368073, "learning_rate": 8.159648185456628e-05, "loss": 0.0145, "step": 19090 }, { "grad_norm": 0.22768355906009674, "learning_rate": 8.157511902810038e-05, "loss": 0.0129, "step": 19100 }, { "grad_norm": 0.19724513590335846, "learning_rate": 8.155374660994701e-05, "loss": 0.0126, "step": 19110 }, { "grad_norm": 0.18196941912174225, "learning_rate": 8.153236460659857e-05, "loss": 0.0144, "step": 19120 }, { "grad_norm": 0.17524836957454681, "learning_rate": 8.151097302455031e-05, "loss": 0.0127, "step": 19130 }, { "grad_norm": 0.1375587433576584, "learning_rate": 8.148957187030044e-05, "loss": 0.0113, "step": 19140 }, { "grad_norm": 0.16967584192752838, "learning_rate": 8.146816115035006e-05, "loss": 0.0129, "step": 19150 }, { "grad_norm": 0.2193153202533722, "learning_rate": 8.14467408712032e-05, "loss": 0.0117, "step": 19160 }, { "grad_norm": 0.22393405437469482, "learning_rate": 8.142531103936678e-05, "loss": 0.0129, "step": 19170 }, { "grad_norm": 0.1893235743045807, "learning_rate": 8.14038716613506e-05, "loss": 0.0157, "step": 19180 }, { "grad_norm": 0.17604349553585052, "learning_rate": 8.138242274366736e-05, "loss": 0.0138, "step": 19190 }, { "grad_norm": 0.22080367803573608, "learning_rate": 8.136096429283271e-05, "loss": 0.0136, "step": 19200 }, { "grad_norm": 0.1504085510969162, "learning_rate": 8.133949631536515e-05, "loss": 0.0117, "step": 19210 }, { "grad_norm": 0.15623079240322113, "learning_rate": 8.131801881778607e-05, "loss": 0.0109, "step": 19220 }, { "grad_norm": 0.24393001198768616, "learning_rate": 8.129653180661978e-05, "loss": 0.0129, "step": 19230 }, { "grad_norm": 0.16464675962924957, "learning_rate": 8.127503528839346e-05, "loss": 0.0121, "step": 19240 }, { "grad_norm": 0.19842086732387543, "learning_rate": 8.125352926963721e-05, "loss": 0.0153, "step": 19250 }, { "grad_norm": 0.2179587036371231, "learning_rate": 8.123201375688395e-05, "loss": 0.0108, "step": 19260 }, { "grad_norm": 0.17824122309684753, "learning_rate": 8.121048875666954e-05, "loss": 0.0127, "step": 19270 }, { "grad_norm": 0.17235732078552246, "learning_rate": 8.118895427553274e-05, "loss": 0.0147, "step": 19280 }, { "grad_norm": 0.13528646528720856, "learning_rate": 8.116741032001511e-05, "loss": 0.0148, "step": 19290 }, { "grad_norm": 0.1497543752193451, "learning_rate": 8.114585689666114e-05, "loss": 0.0113, "step": 19300 }, { "grad_norm": 0.1756291389465332, "learning_rate": 8.112429401201821e-05, "loss": 0.0124, "step": 19310 }, { "grad_norm": 0.24597600102424622, "learning_rate": 8.110272167263656e-05, "loss": 0.0135, "step": 19320 }, { "grad_norm": 0.15959666669368744, "learning_rate": 8.108113988506929e-05, "loss": 0.0165, "step": 19330 }, { "grad_norm": 0.14757995307445526, "learning_rate": 8.105954865587235e-05, "loss": 0.0142, "step": 19340 }, { "grad_norm": 0.16286198794841766, "learning_rate": 8.103794799160463e-05, "loss": 0.0124, "step": 19350 }, { "grad_norm": 0.22325968742370605, "learning_rate": 8.101633789882781e-05, "loss": 0.0134, "step": 19360 }, { "grad_norm": 0.16614790260791779, "learning_rate": 8.099471838410648e-05, "loss": 0.0126, "step": 19370 }, { "grad_norm": 0.19368641078472137, "learning_rate": 8.097308945400806e-05, "loss": 0.013, "step": 19380 }, { "grad_norm": 0.16594336926937103, "learning_rate": 8.095145111510288e-05, "loss": 0.015, "step": 19390 }, { "grad_norm": 0.17583014070987701, "learning_rate": 8.092980337396406e-05, "loss": 0.0133, "step": 19400 }, { "grad_norm": 0.2312123328447342, "learning_rate": 8.090814623716763e-05, "loss": 0.016, "step": 19410 }, { "grad_norm": 0.18187592923641205, "learning_rate": 8.088647971129246e-05, "loss": 0.0118, "step": 19420 }, { "grad_norm": 0.2259168177843094, "learning_rate": 8.086480380292026e-05, "loss": 0.0143, "step": 19430 }, { "grad_norm": 0.19626875221729279, "learning_rate": 8.084311851863562e-05, "loss": 0.013, "step": 19440 }, { "grad_norm": 0.2234029620885849, "learning_rate": 8.082142386502591e-05, "loss": 0.0134, "step": 19450 }, { "grad_norm": 0.1703065037727356, "learning_rate": 8.079971984868145e-05, "loss": 0.0117, "step": 19460 }, { "grad_norm": 0.1581679880619049, "learning_rate": 8.077800647619532e-05, "loss": 0.0148, "step": 19470 }, { "grad_norm": 0.19748832285404205, "learning_rate": 8.075628375416345e-05, "loss": 0.0143, "step": 19480 }, { "grad_norm": 0.14765974879264832, "learning_rate": 8.073455168918464e-05, "loss": 0.0122, "step": 19490 }, { "grad_norm": 0.18950392305850983, "learning_rate": 8.071281028786055e-05, "loss": 0.0122, "step": 19500 }, { "grad_norm": 0.20717236399650574, "learning_rate": 8.069105955679562e-05, "loss": 0.0159, "step": 19510 }, { "grad_norm": 0.2062862366437912, "learning_rate": 8.066929950259713e-05, "loss": 0.0138, "step": 19520 }, { "grad_norm": 0.18554501235485077, "learning_rate": 8.064753013187522e-05, "loss": 0.0146, "step": 19530 }, { "grad_norm": 0.15752996504306793, "learning_rate": 8.062575145124289e-05, "loss": 0.013, "step": 19540 }, { "grad_norm": 0.225843608379364, "learning_rate": 8.060396346731587e-05, "loss": 0.0163, "step": 19550 }, { "grad_norm": 0.20342853665351868, "learning_rate": 8.058216618671281e-05, "loss": 0.0127, "step": 19560 }, { "grad_norm": 0.13461482524871826, "learning_rate": 8.056035961605514e-05, "loss": 0.0144, "step": 19570 }, { "grad_norm": 0.2063598781824112, "learning_rate": 8.05385437619671e-05, "loss": 0.0102, "step": 19580 }, { "grad_norm": 0.21118704974651337, "learning_rate": 8.05167186310758e-05, "loss": 0.0144, "step": 19590 }, { "grad_norm": 0.2168005406856537, "learning_rate": 8.049488423001113e-05, "loss": 0.0158, "step": 19600 }, { "grad_norm": 0.14010019600391388, "learning_rate": 8.047304056540581e-05, "loss": 0.0137, "step": 19610 }, { "grad_norm": 0.18166162073612213, "learning_rate": 8.045118764389534e-05, "loss": 0.0154, "step": 19620 }, { "grad_norm": 0.15430934727191925, "learning_rate": 8.042932547211809e-05, "loss": 0.0116, "step": 19630 }, { "grad_norm": 0.198469877243042, "learning_rate": 8.04074540567152e-05, "loss": 0.0156, "step": 19640 }, { "grad_norm": 0.16560675203800201, "learning_rate": 8.038557340433063e-05, "loss": 0.0136, "step": 19650 }, { "grad_norm": 0.1902097463607788, "learning_rate": 8.036368352161115e-05, "loss": 0.0135, "step": 19660 }, { "grad_norm": 0.17075994610786438, "learning_rate": 8.034178441520633e-05, "loss": 0.0128, "step": 19670 }, { "grad_norm": 0.17169587314128876, "learning_rate": 8.031987609176852e-05, "loss": 0.0145, "step": 19680 }, { "grad_norm": 0.2513592541217804, "learning_rate": 8.02979585579529e-05, "loss": 0.0128, "step": 19690 }, { "grad_norm": 0.16993945837020874, "learning_rate": 8.027603182041745e-05, "loss": 0.0121, "step": 19700 }, { "grad_norm": 0.16619831323623657, "learning_rate": 8.025409588582292e-05, "loss": 0.0139, "step": 19710 }, { "grad_norm": 0.1607590764760971, "learning_rate": 8.023215076083288e-05, "loss": 0.013, "step": 19720 }, { "grad_norm": 0.1757594347000122, "learning_rate": 8.021019645211367e-05, "loss": 0.016, "step": 19730 }, { "grad_norm": 0.18187198042869568, "learning_rate": 8.018823296633441e-05, "loss": 0.0129, "step": 19740 }, { "grad_norm": 0.21748952567577362, "learning_rate": 8.016626031016708e-05, "loss": 0.0129, "step": 19750 }, { "grad_norm": 0.19877493381500244, "learning_rate": 8.014427849028636e-05, "loss": 0.0118, "step": 19760 }, { "grad_norm": 0.2016858160495758, "learning_rate": 8.012228751336974e-05, "loss": 0.0149, "step": 19770 }, { "grad_norm": 0.21192456781864166, "learning_rate": 8.01002873860975e-05, "loss": 0.0137, "step": 19780 }, { "grad_norm": 0.14802668988704681, "learning_rate": 8.00782781151527e-05, "loss": 0.0126, "step": 19790 }, { "grad_norm": 0.17750070989131927, "learning_rate": 8.005625970722119e-05, "loss": 0.0108, "step": 19800 }, { "grad_norm": 0.19011865556240082, "learning_rate": 8.003423216899158e-05, "loss": 0.0126, "step": 19810 }, { "grad_norm": 0.1797196865081787, "learning_rate": 8.001219550715522e-05, "loss": 0.0134, "step": 19820 }, { "grad_norm": 0.22280801832675934, "learning_rate": 7.999014972840632e-05, "loss": 0.0151, "step": 19830 }, { "grad_norm": 0.1881856471300125, "learning_rate": 7.996809483944174e-05, "loss": 0.0133, "step": 19840 }, { "grad_norm": 0.14256508648395538, "learning_rate": 7.994603084696124e-05, "loss": 0.012, "step": 19850 }, { "grad_norm": 0.19043007493019104, "learning_rate": 7.992395775766724e-05, "loss": 0.0149, "step": 19860 }, { "grad_norm": 0.13378222286701202, "learning_rate": 7.990187557826497e-05, "loss": 0.0148, "step": 19870 }, { "grad_norm": 0.21131668984889984, "learning_rate": 7.987978431546242e-05, "loss": 0.0145, "step": 19880 }, { "grad_norm": 0.17103995382785797, "learning_rate": 7.985768397597031e-05, "loss": 0.0131, "step": 19890 }, { "grad_norm": 0.19821582734584808, "learning_rate": 7.983557456650216e-05, "loss": 0.0141, "step": 19900 }, { "grad_norm": 0.19444768130779266, "learning_rate": 7.981345609377422e-05, "loss": 0.016, "step": 19910 }, { "grad_norm": 0.16000565886497498, "learning_rate": 7.97913285645055e-05, "loss": 0.0132, "step": 19920 }, { "grad_norm": 0.1964181363582611, "learning_rate": 7.976919198541776e-05, "loss": 0.0127, "step": 19930 }, { "grad_norm": 0.21764695644378662, "learning_rate": 7.974704636323548e-05, "loss": 0.0141, "step": 19940 }, { "grad_norm": 0.16310787200927734, "learning_rate": 7.972489170468597e-05, "loss": 0.011, "step": 19950 }, { "grad_norm": 0.19930334389209747, "learning_rate": 7.970272801649918e-05, "loss": 0.0136, "step": 19960 }, { "grad_norm": 0.17397473752498627, "learning_rate": 7.96805553054079e-05, "loss": 0.014, "step": 19970 }, { "grad_norm": 0.23535700142383575, "learning_rate": 7.965837357814756e-05, "loss": 0.0125, "step": 19980 }, { "grad_norm": 0.14898785948753357, "learning_rate": 7.963618284145643e-05, "loss": 0.0115, "step": 19990 }, { "grad_norm": 0.18823252618312836, "learning_rate": 7.961398310207544e-05, "loss": 0.0133, "step": 20000 }, { "grad_norm": 0.16020837426185608, "learning_rate": 7.95917743667483e-05, "loss": 0.016, "step": 20010 }, { "grad_norm": 0.15933600068092346, "learning_rate": 7.956955664222144e-05, "loss": 0.0135, "step": 20020 }, { "grad_norm": 0.1714819222688675, "learning_rate": 7.954732993524399e-05, "loss": 0.0124, "step": 20030 }, { "grad_norm": 0.2397737205028534, "learning_rate": 7.952509425256786e-05, "loss": 0.0117, "step": 20040 }, { "grad_norm": 0.17884042859077454, "learning_rate": 7.950284960094767e-05, "loss": 0.0115, "step": 20050 }, { "grad_norm": 0.16222406923770905, "learning_rate": 7.948059598714076e-05, "loss": 0.0136, "step": 20060 }, { "grad_norm": 0.21561439335346222, "learning_rate": 7.945833341790717e-05, "loss": 0.0148, "step": 20070 }, { "grad_norm": 0.1714911311864853, "learning_rate": 7.94360619000097e-05, "loss": 0.0124, "step": 20080 }, { "grad_norm": 0.14350572228431702, "learning_rate": 7.941378144021381e-05, "loss": 0.015, "step": 20090 }, { "grad_norm": 0.20168690383434296, "learning_rate": 7.939149204528777e-05, "loss": 0.0169, "step": 20100 }, { "grad_norm": 0.20321574807167053, "learning_rate": 7.936919372200246e-05, "loss": 0.0164, "step": 20110 }, { "grad_norm": 0.1915530115365982, "learning_rate": 7.934688647713158e-05, "loss": 0.016, "step": 20120 }, { "grad_norm": 0.2439250499010086, "learning_rate": 7.932457031745143e-05, "loss": 0.0153, "step": 20130 }, { "grad_norm": 0.19387829303741455, "learning_rate": 7.930224524974108e-05, "loss": 0.0146, "step": 20140 }, { "grad_norm": 0.19536292552947998, "learning_rate": 7.927991128078232e-05, "loss": 0.0163, "step": 20150 }, { "grad_norm": 0.2315545231103897, "learning_rate": 7.925756841735958e-05, "loss": 0.0121, "step": 20160 }, { "grad_norm": 0.2087385654449463, "learning_rate": 7.923521666626008e-05, "loss": 0.0118, "step": 20170 }, { "grad_norm": 0.143747478723526, "learning_rate": 7.921285603427366e-05, "loss": 0.0125, "step": 20180 }, { "grad_norm": 0.15774601697921753, "learning_rate": 7.91904865281929e-05, "loss": 0.0125, "step": 20190 }, { "grad_norm": 0.17218251526355743, "learning_rate": 7.916810815481307e-05, "loss": 0.0118, "step": 20200 }, { "grad_norm": 0.22163359820842743, "learning_rate": 7.914572092093211e-05, "loss": 0.0124, "step": 20210 }, { "grad_norm": 0.1918443739414215, "learning_rate": 7.912332483335068e-05, "loss": 0.0146, "step": 20220 }, { "grad_norm": 0.1440751701593399, "learning_rate": 7.910091989887213e-05, "loss": 0.0106, "step": 20230 }, { "grad_norm": 0.21371963620185852, "learning_rate": 7.907850612430248e-05, "loss": 0.0109, "step": 20240 }, { "grad_norm": 0.1898081749677658, "learning_rate": 7.905608351645044e-05, "loss": 0.0142, "step": 20250 }, { "grad_norm": 0.22363169491291046, "learning_rate": 7.90336520821274e-05, "loss": 0.0152, "step": 20260 }, { "grad_norm": 0.23510876297950745, "learning_rate": 7.901121182814746e-05, "loss": 0.0117, "step": 20270 }, { "grad_norm": 0.21623186767101288, "learning_rate": 7.898876276132736e-05, "loss": 0.0127, "step": 20280 }, { "grad_norm": 0.21464785933494568, "learning_rate": 7.896630488848654e-05, "loss": 0.013, "step": 20290 }, { "grad_norm": 0.16909578442573547, "learning_rate": 7.89438382164471e-05, "loss": 0.0121, "step": 20300 }, { "grad_norm": 0.14415471255779266, "learning_rate": 7.892136275203383e-05, "loss": 0.0166, "step": 20310 }, { "grad_norm": 0.21509721875190735, "learning_rate": 7.889887850207418e-05, "loss": 0.0124, "step": 20320 }, { "grad_norm": 0.19273434579372406, "learning_rate": 7.887638547339827e-05, "loss": 0.0132, "step": 20330 }, { "grad_norm": 0.15882517397403717, "learning_rate": 7.885388367283891e-05, "loss": 0.0145, "step": 20340 }, { "grad_norm": 0.18550314009189606, "learning_rate": 7.88313731072315e-05, "loss": 0.0148, "step": 20350 }, { "grad_norm": 0.1482803225517273, "learning_rate": 7.88088537834142e-05, "loss": 0.0124, "step": 20360 }, { "grad_norm": 0.20852744579315186, "learning_rate": 7.878632570822778e-05, "loss": 0.0139, "step": 20370 }, { "grad_norm": 0.15360260009765625, "learning_rate": 7.876378888851567e-05, "loss": 0.0141, "step": 20380 }, { "grad_norm": 0.1548977941274643, "learning_rate": 7.874124333112396e-05, "loss": 0.0152, "step": 20390 }, { "grad_norm": 0.1717674881219864, "learning_rate": 7.871868904290138e-05, "loss": 0.0137, "step": 20400 }, { "grad_norm": 0.1951676905155182, "learning_rate": 7.869612603069935e-05, "loss": 0.0132, "step": 20410 }, { "grad_norm": 0.1987815648317337, "learning_rate": 7.867355430137192e-05, "loss": 0.0099, "step": 20420 }, { "grad_norm": 0.16982588171958923, "learning_rate": 7.865097386177577e-05, "loss": 0.0096, "step": 20430 }, { "grad_norm": 0.16288337111473083, "learning_rate": 7.862838471877023e-05, "loss": 0.013, "step": 20440 }, { "grad_norm": 0.19783851504325867, "learning_rate": 7.860578687921731e-05, "loss": 0.0117, "step": 20450 }, { "grad_norm": 0.18966840207576752, "learning_rate": 7.858318034998164e-05, "loss": 0.0135, "step": 20460 }, { "grad_norm": 0.2013414055109024, "learning_rate": 7.856056513793046e-05, "loss": 0.0134, "step": 20470 }, { "grad_norm": 0.16279907524585724, "learning_rate": 7.85379412499337e-05, "loss": 0.0137, "step": 20480 }, { "grad_norm": 0.18188521265983582, "learning_rate": 7.851530869286389e-05, "loss": 0.0163, "step": 20490 }, { "grad_norm": 0.1432977020740509, "learning_rate": 7.849266747359619e-05, "loss": 0.013, "step": 20500 }, { "grad_norm": 0.20708243548870087, "learning_rate": 7.847001759900843e-05, "loss": 0.0113, "step": 20510 }, { "grad_norm": 0.16951817274093628, "learning_rate": 7.844735907598102e-05, "loss": 0.0168, "step": 20520 }, { "grad_norm": 0.16113369166851044, "learning_rate": 7.842469191139703e-05, "loss": 0.0151, "step": 20530 }, { "grad_norm": 0.16427677869796753, "learning_rate": 7.840201611214215e-05, "loss": 0.0095, "step": 20540 }, { "grad_norm": 0.16463884711265564, "learning_rate": 7.837933168510469e-05, "loss": 0.0135, "step": 20550 }, { "grad_norm": 0.18257789313793182, "learning_rate": 7.835663863717559e-05, "loss": 0.0146, "step": 20560 }, { "grad_norm": 0.2227262705564499, "learning_rate": 7.833393697524838e-05, "loss": 0.0132, "step": 20570 }, { "grad_norm": 0.2148093730211258, "learning_rate": 7.831122670621922e-05, "loss": 0.0147, "step": 20580 }, { "grad_norm": 0.17331402003765106, "learning_rate": 7.82885078369869e-05, "loss": 0.0122, "step": 20590 }, { "grad_norm": 0.14037664234638214, "learning_rate": 7.826578037445283e-05, "loss": 0.0119, "step": 20600 }, { "grad_norm": 0.19912059605121613, "learning_rate": 7.824304432552097e-05, "loss": 0.0127, "step": 20610 }, { "grad_norm": 0.16311021149158478, "learning_rate": 7.822029969709798e-05, "loss": 0.0094, "step": 20620 }, { "grad_norm": 0.19246454536914825, "learning_rate": 7.819754649609306e-05, "loss": 0.0135, "step": 20630 }, { "grad_norm": 0.2234046757221222, "learning_rate": 7.817478472941802e-05, "loss": 0.0131, "step": 20640 }, { "grad_norm": 0.19695481657981873, "learning_rate": 7.815201440398727e-05, "loss": 0.0156, "step": 20650 }, { "grad_norm": 0.17279158532619476, "learning_rate": 7.812923552671789e-05, "loss": 0.0157, "step": 20660 }, { "grad_norm": 0.17194996774196625, "learning_rate": 7.810644810452945e-05, "loss": 0.0141, "step": 20670 }, { "grad_norm": 0.15895040333271027, "learning_rate": 7.808365214434417e-05, "loss": 0.0135, "step": 20680 }, { "grad_norm": 0.13684864342212677, "learning_rate": 7.80608476530869e-05, "loss": 0.0125, "step": 20690 }, { "grad_norm": 0.17378540337085724, "learning_rate": 7.8038034637685e-05, "loss": 0.0142, "step": 20700 }, { "grad_norm": 0.1502588987350464, "learning_rate": 7.801521310506848e-05, "loss": 0.0132, "step": 20710 }, { "grad_norm": 0.15435048937797546, "learning_rate": 7.799238306216994e-05, "loss": 0.0124, "step": 20720 }, { "grad_norm": 0.2150384485721588, "learning_rate": 7.796954451592448e-05, "loss": 0.0152, "step": 20730 }, { "grad_norm": 0.17259204387664795, "learning_rate": 7.794669747326992e-05, "loss": 0.0151, "step": 20740 }, { "grad_norm": 0.21191270649433136, "learning_rate": 7.792384194114654e-05, "loss": 0.0122, "step": 20750 }, { "grad_norm": 0.18239204585552216, "learning_rate": 7.790097792649729e-05, "loss": 0.0137, "step": 20760 }, { "grad_norm": 0.24034015834331512, "learning_rate": 7.787810543626762e-05, "loss": 0.0165, "step": 20770 }, { "grad_norm": 0.14808164536952972, "learning_rate": 7.785522447740558e-05, "loss": 0.011, "step": 20780 }, { "grad_norm": 0.17794863879680634, "learning_rate": 7.783233505686182e-05, "loss": 0.0121, "step": 20790 }, { "grad_norm": 0.1435529589653015, "learning_rate": 7.780943718158955e-05, "loss": 0.0122, "step": 20800 }, { "grad_norm": 0.2432141900062561, "learning_rate": 7.778653085854453e-05, "loss": 0.0115, "step": 20810 }, { "grad_norm": 0.17531511187553406, "learning_rate": 7.77636160946851e-05, "loss": 0.0121, "step": 20820 }, { "grad_norm": 0.1754736602306366, "learning_rate": 7.774069289697215e-05, "loss": 0.0146, "step": 20830 }, { "grad_norm": 0.1633952558040619, "learning_rate": 7.771776127236913e-05, "loss": 0.0117, "step": 20840 }, { "grad_norm": 0.1429428607225418, "learning_rate": 7.769482122784212e-05, "loss": 0.0125, "step": 20850 }, { "grad_norm": 0.18260623514652252, "learning_rate": 7.767187277035963e-05, "loss": 0.0131, "step": 20860 }, { "grad_norm": 0.19557423889636993, "learning_rate": 7.764891590689285e-05, "loss": 0.0123, "step": 20870 }, { "grad_norm": 0.20297329127788544, "learning_rate": 7.762595064441542e-05, "loss": 0.0131, "step": 20880 }, { "grad_norm": 0.172795370221138, "learning_rate": 7.760297698990362e-05, "loss": 0.0127, "step": 20890 }, { "grad_norm": 0.12658677995204926, "learning_rate": 7.757999495033623e-05, "loss": 0.01, "step": 20900 }, { "grad_norm": 0.16464506089687347, "learning_rate": 7.755700453269456e-05, "loss": 0.0136, "step": 20910 }, { "grad_norm": 0.15891386568546295, "learning_rate": 7.753400574396254e-05, "loss": 0.0124, "step": 20920 }, { "grad_norm": 0.14975062012672424, "learning_rate": 7.751099859112655e-05, "loss": 0.0113, "step": 20930 }, { "grad_norm": 0.1977827250957489, "learning_rate": 7.748798308117557e-05, "loss": 0.0154, "step": 20940 }, { "grad_norm": 0.17700375616550446, "learning_rate": 7.746495922110112e-05, "loss": 0.0109, "step": 20950 }, { "grad_norm": 0.2471134513616562, "learning_rate": 7.744192701789723e-05, "loss": 0.012, "step": 20960 }, { "grad_norm": 0.20402798056602478, "learning_rate": 7.741888647856046e-05, "loss": 0.0132, "step": 20970 }, { "grad_norm": 0.11566800624132156, "learning_rate": 7.739583761008994e-05, "loss": 0.0126, "step": 20980 }, { "grad_norm": 0.23263414204120636, "learning_rate": 7.73727804194873e-05, "loss": 0.0124, "step": 20990 }, { "grad_norm": 0.1849532127380371, "learning_rate": 7.734971491375671e-05, "loss": 0.0153, "step": 21000 }, { "grad_norm": 0.1970638632774353, "learning_rate": 7.732664109990485e-05, "loss": 0.0122, "step": 21010 }, { "grad_norm": 0.17937348783016205, "learning_rate": 7.730355898494095e-05, "loss": 0.0137, "step": 21020 }, { "grad_norm": 0.18044660985469818, "learning_rate": 7.728046857587673e-05, "loss": 0.0121, "step": 21030 }, { "grad_norm": 0.16604310274124146, "learning_rate": 7.725736987972647e-05, "loss": 0.0132, "step": 21040 }, { "grad_norm": 0.21968591213226318, "learning_rate": 7.723426290350691e-05, "loss": 0.0157, "step": 21050 }, { "grad_norm": 0.1733166128396988, "learning_rate": 7.721114765423736e-05, "loss": 0.0115, "step": 21060 }, { "grad_norm": 0.1717037558555603, "learning_rate": 7.718802413893963e-05, "loss": 0.016, "step": 21070 }, { "grad_norm": 0.14537109434604645, "learning_rate": 7.716489236463802e-05, "loss": 0.0128, "step": 21080 }, { "grad_norm": 0.14465565979480743, "learning_rate": 7.714175233835936e-05, "loss": 0.012, "step": 21090 }, { "grad_norm": 0.15834325551986694, "learning_rate": 7.711860406713299e-05, "loss": 0.0112, "step": 21100 }, { "grad_norm": 0.16257193684577942, "learning_rate": 7.70954475579907e-05, "loss": 0.0146, "step": 21110 }, { "grad_norm": 0.16397523880004883, "learning_rate": 7.707228281796688e-05, "loss": 0.011, "step": 21120 }, { "grad_norm": 0.18260625004768372, "learning_rate": 7.704910985409833e-05, "loss": 0.014, "step": 21130 }, { "grad_norm": 0.19432596862316132, "learning_rate": 7.702592867342439e-05, "loss": 0.0141, "step": 21140 }, { "grad_norm": 0.14887279272079468, "learning_rate": 7.700273928298691e-05, "loss": 0.0131, "step": 21150 }, { "grad_norm": 0.13985984027385712, "learning_rate": 7.697954168983021e-05, "loss": 0.0126, "step": 21160 }, { "grad_norm": 0.15809126198291779, "learning_rate": 7.695633590100109e-05, "loss": 0.0127, "step": 21170 }, { "grad_norm": 0.17151756584644318, "learning_rate": 7.693312192354886e-05, "loss": 0.0134, "step": 21180 }, { "grad_norm": 0.1590651124715805, "learning_rate": 7.690989976452532e-05, "loss": 0.0141, "step": 21190 }, { "grad_norm": 0.19783005118370056, "learning_rate": 7.688666943098475e-05, "loss": 0.0132, "step": 21200 }, { "grad_norm": 0.21689698100090027, "learning_rate": 7.686343092998389e-05, "loss": 0.0108, "step": 21210 }, { "grad_norm": 0.21114593744277954, "learning_rate": 7.684018426858202e-05, "loss": 0.013, "step": 21220 }, { "grad_norm": 0.1366564929485321, "learning_rate": 7.681692945384084e-05, "loss": 0.0128, "step": 21230 }, { "grad_norm": 0.16639478504657745, "learning_rate": 7.679366649282456e-05, "loss": 0.0115, "step": 21240 }, { "grad_norm": 0.15636441111564636, "learning_rate": 7.677039539259983e-05, "loss": 0.0113, "step": 21250 }, { "grad_norm": 0.1252758651971817, "learning_rate": 7.674711616023581e-05, "loss": 0.0116, "step": 21260 }, { "grad_norm": 0.1305607110261917, "learning_rate": 7.672382880280413e-05, "loss": 0.0132, "step": 21270 }, { "grad_norm": 0.21144995093345642, "learning_rate": 7.670053332737885e-05, "loss": 0.0113, "step": 21280 }, { "grad_norm": 0.14970237016677856, "learning_rate": 7.667722974103654e-05, "loss": 0.0128, "step": 21290 }, { "grad_norm": 0.18742920458316803, "learning_rate": 7.66539180508562e-05, "loss": 0.0114, "step": 21300 }, { "grad_norm": 0.20268219709396362, "learning_rate": 7.663059826391932e-05, "loss": 0.0146, "step": 21310 }, { "grad_norm": 0.18376143276691437, "learning_rate": 7.660727038730981e-05, "loss": 0.0125, "step": 21320 }, { "grad_norm": 0.11423268914222717, "learning_rate": 7.65839344281141e-05, "loss": 0.0089, "step": 21330 }, { "grad_norm": 0.1986813098192215, "learning_rate": 7.656059039342101e-05, "loss": 0.013, "step": 21340 }, { "grad_norm": 0.17664888501167297, "learning_rate": 7.653723829032187e-05, "loss": 0.0141, "step": 21350 }, { "grad_norm": 0.1851118505001068, "learning_rate": 7.65138781259104e-05, "loss": 0.011, "step": 21360 }, { "grad_norm": 0.129022479057312, "learning_rate": 7.649050990728279e-05, "loss": 0.0111, "step": 21370 }, { "grad_norm": 0.173222616314888, "learning_rate": 7.646713364153774e-05, "loss": 0.0105, "step": 21380 }, { "grad_norm": 0.1933252364397049, "learning_rate": 7.64437493357763e-05, "loss": 0.013, "step": 21390 }, { "grad_norm": 0.15459072589874268, "learning_rate": 7.642035699710202e-05, "loss": 0.0117, "step": 21400 }, { "grad_norm": 0.13028892874717712, "learning_rate": 7.639695663262089e-05, "loss": 0.0115, "step": 21410 }, { "grad_norm": 0.14310713112354279, "learning_rate": 7.637354824944128e-05, "loss": 0.0122, "step": 21420 }, { "grad_norm": 0.1602892428636551, "learning_rate": 7.635013185467408e-05, "loss": 0.0117, "step": 21430 }, { "grad_norm": 0.16178129613399506, "learning_rate": 7.632670745543256e-05, "loss": 0.012, "step": 21440 }, { "grad_norm": 0.15610232949256897, "learning_rate": 7.630327505883242e-05, "loss": 0.0158, "step": 21450 }, { "grad_norm": 0.178521990776062, "learning_rate": 7.627983467199182e-05, "loss": 0.0111, "step": 21460 }, { "grad_norm": 0.1685800403356552, "learning_rate": 7.625638630203132e-05, "loss": 0.0103, "step": 21470 }, { "grad_norm": 0.14680565893650055, "learning_rate": 7.623292995607394e-05, "loss": 0.0111, "step": 21480 }, { "grad_norm": 0.11726298928260803, "learning_rate": 7.620946564124507e-05, "loss": 0.0116, "step": 21490 }, { "grad_norm": 0.12938754260540009, "learning_rate": 7.618599336467256e-05, "loss": 0.0128, "step": 21500 }, { "grad_norm": 0.14968307316303253, "learning_rate": 7.616251313348666e-05, "loss": 0.01, "step": 21510 }, { "grad_norm": 0.14377884566783905, "learning_rate": 7.613902495482005e-05, "loss": 0.0118, "step": 21520 }, { "grad_norm": 0.13845188915729523, "learning_rate": 7.611552883580784e-05, "loss": 0.0126, "step": 21530 }, { "grad_norm": 0.1591366082429886, "learning_rate": 7.609202478358748e-05, "loss": 0.0129, "step": 21540 }, { "grad_norm": 0.218962162733078, "learning_rate": 7.606851280529895e-05, "loss": 0.0121, "step": 21550 }, { "grad_norm": 0.18619735538959503, "learning_rate": 7.604499290808449e-05, "loss": 0.0124, "step": 21560 }, { "grad_norm": 0.189217671751976, "learning_rate": 7.602146509908888e-05, "loss": 0.0106, "step": 21570 }, { "grad_norm": 0.2005598545074463, "learning_rate": 7.599792938545921e-05, "loss": 0.0149, "step": 21580 }, { "grad_norm": 0.19676022231578827, "learning_rate": 7.597438577434506e-05, "loss": 0.0134, "step": 21590 }, { "grad_norm": 0.18533079326152802, "learning_rate": 7.595083427289831e-05, "loss": 0.0141, "step": 21600 }, { "grad_norm": 0.19535128772258759, "learning_rate": 7.59272748882733e-05, "loss": 0.0121, "step": 21610 }, { "grad_norm": 0.19317680597305298, "learning_rate": 7.590370762762675e-05, "loss": 0.0129, "step": 21620 }, { "grad_norm": 0.22840264439582825, "learning_rate": 7.588013249811777e-05, "loss": 0.0126, "step": 21630 }, { "grad_norm": 0.1635420024394989, "learning_rate": 7.585654950690786e-05, "loss": 0.0129, "step": 21640 }, { "grad_norm": 0.17025958001613617, "learning_rate": 7.583295866116091e-05, "loss": 0.0106, "step": 21650 }, { "grad_norm": 0.17257638275623322, "learning_rate": 7.580935996804321e-05, "loss": 0.0124, "step": 21660 }, { "grad_norm": 0.18423010408878326, "learning_rate": 7.57857534347234e-05, "loss": 0.0135, "step": 21670 }, { "grad_norm": 0.2009989321231842, "learning_rate": 7.576213906837254e-05, "loss": 0.0131, "step": 21680 }, { "grad_norm": 0.23913054168224335, "learning_rate": 7.573851687616403e-05, "loss": 0.0137, "step": 21690 }, { "grad_norm": 0.18988600373268127, "learning_rate": 7.571488686527368e-05, "loss": 0.0145, "step": 21700 }, { "grad_norm": 0.17714907228946686, "learning_rate": 7.569124904287968e-05, "loss": 0.0128, "step": 21710 }, { "grad_norm": 0.18879935145378113, "learning_rate": 7.566760341616254e-05, "loss": 0.0187, "step": 21720 }, { "grad_norm": 0.14503033459186554, "learning_rate": 7.564394999230519e-05, "loss": 0.0154, "step": 21730 }, { "grad_norm": 0.1514173448085785, "learning_rate": 7.562028877849294e-05, "loss": 0.0116, "step": 21740 }, { "grad_norm": 0.19175000488758087, "learning_rate": 7.559661978191341e-05, "loss": 0.0122, "step": 21750 }, { "grad_norm": 0.13511724770069122, "learning_rate": 7.557294300975664e-05, "loss": 0.0137, "step": 21760 }, { "grad_norm": 0.18006043136119843, "learning_rate": 7.554925846921499e-05, "loss": 0.015, "step": 21770 }, { "grad_norm": 0.18870270252227783, "learning_rate": 7.552556616748321e-05, "loss": 0.0126, "step": 21780 }, { "grad_norm": 0.17639651894569397, "learning_rate": 7.550186611175838e-05, "loss": 0.0114, "step": 21790 }, { "grad_norm": 0.1747010350227356, "learning_rate": 7.547815830923998e-05, "loss": 0.0118, "step": 21800 }, { "grad_norm": 0.15640074014663696, "learning_rate": 7.54544427671298e-05, "loss": 0.0135, "step": 21810 }, { "grad_norm": 0.24674680829048157, "learning_rate": 7.543071949263198e-05, "loss": 0.0113, "step": 21820 }, { "grad_norm": 0.19583919644355774, "learning_rate": 7.540698849295305e-05, "loss": 0.0127, "step": 21830 }, { "grad_norm": 0.17406322062015533, "learning_rate": 7.538324977530183e-05, "loss": 0.0121, "step": 21840 }, { "grad_norm": 0.19419142603874207, "learning_rate": 7.535950334688955e-05, "loss": 0.0139, "step": 21850 }, { "grad_norm": 0.20584779977798462, "learning_rate": 7.533574921492972e-05, "loss": 0.0126, "step": 21860 }, { "grad_norm": 0.15359751880168915, "learning_rate": 7.531198738663824e-05, "loss": 0.014, "step": 21870 }, { "grad_norm": 0.19086605310440063, "learning_rate": 7.528821786923333e-05, "loss": 0.0118, "step": 21880 }, { "grad_norm": 0.19253429770469666, "learning_rate": 7.52644406699355e-05, "loss": 0.0123, "step": 21890 }, { "grad_norm": 0.3760901987552643, "learning_rate": 7.524065579596766e-05, "loss": 0.0134, "step": 21900 }, { "grad_norm": 0.1737489402294159, "learning_rate": 7.521686325455506e-05, "loss": 0.0116, "step": 21910 }, { "grad_norm": 0.18661975860595703, "learning_rate": 7.51930630529252e-05, "loss": 0.0118, "step": 21920 }, { "grad_norm": 0.17255744338035583, "learning_rate": 7.516925519830797e-05, "loss": 0.0132, "step": 21930 }, { "grad_norm": 0.14751549065113068, "learning_rate": 7.514543969793557e-05, "loss": 0.0132, "step": 21940 }, { "grad_norm": 0.1426813304424286, "learning_rate": 7.512161655904251e-05, "loss": 0.0108, "step": 21950 }, { "grad_norm": 0.1409102976322174, "learning_rate": 7.509778578886563e-05, "loss": 0.0116, "step": 21960 }, { "grad_norm": 0.15614323318004608, "learning_rate": 7.507394739464412e-05, "loss": 0.014, "step": 21970 }, { "grad_norm": 0.17394378781318665, "learning_rate": 7.50501013836194e-05, "loss": 0.0151, "step": 21980 }, { "grad_norm": 0.1560397893190384, "learning_rate": 7.50262477630353e-05, "loss": 0.012, "step": 21990 }, { "grad_norm": 0.1588214933872223, "learning_rate": 7.500238654013794e-05, "loss": 0.0101, "step": 22000 }, { "grad_norm": 0.14757207036018372, "learning_rate": 7.497851772217566e-05, "loss": 0.0116, "step": 22010 }, { "grad_norm": 0.17326520383358002, "learning_rate": 7.495464131639924e-05, "loss": 0.0123, "step": 22020 }, { "grad_norm": 0.14887601137161255, "learning_rate": 7.493075733006166e-05, "loss": 0.0117, "step": 22030 }, { "grad_norm": 0.1813371628522873, "learning_rate": 7.490686577041828e-05, "loss": 0.013, "step": 22040 }, { "grad_norm": 0.1344877928495407, "learning_rate": 7.488296664472668e-05, "loss": 0.011, "step": 22050 }, { "grad_norm": 0.12156510353088379, "learning_rate": 7.485905996024682e-05, "loss": 0.0102, "step": 22060 }, { "grad_norm": 0.18202263116836548, "learning_rate": 7.483514572424093e-05, "loss": 0.0122, "step": 22070 }, { "grad_norm": 0.18970726430416107, "learning_rate": 7.481122394397349e-05, "loss": 0.0105, "step": 22080 }, { "grad_norm": 0.22966159880161285, "learning_rate": 7.478729462671131e-05, "loss": 0.0109, "step": 22090 }, { "grad_norm": 0.14310118556022644, "learning_rate": 7.47633577797235e-05, "loss": 0.0148, "step": 22100 }, { "grad_norm": 0.22116245329380035, "learning_rate": 7.473941341028144e-05, "loss": 0.0113, "step": 22110 }, { "grad_norm": 0.15427906811237335, "learning_rate": 7.471546152565879e-05, "loss": 0.0117, "step": 22120 }, { "grad_norm": 0.24027226865291595, "learning_rate": 7.46915021331315e-05, "loss": 0.0156, "step": 22130 }, { "grad_norm": 0.14472393691539764, "learning_rate": 7.466753523997778e-05, "loss": 0.0123, "step": 22140 }, { "grad_norm": 0.2938302159309387, "learning_rate": 7.464356085347819e-05, "loss": 0.0122, "step": 22150 }, { "grad_norm": 0.20824630558490753, "learning_rate": 7.461957898091548e-05, "loss": 0.0107, "step": 22160 }, { "grad_norm": 0.1405724436044693, "learning_rate": 7.459558962957473e-05, "loss": 0.0121, "step": 22170 }, { "grad_norm": 0.1785513013601303, "learning_rate": 7.457159280674326e-05, "loss": 0.0129, "step": 22180 }, { "grad_norm": 0.17288343608379364, "learning_rate": 7.454758851971066e-05, "loss": 0.011, "step": 22190 }, { "grad_norm": 0.15892523527145386, "learning_rate": 7.45235767757688e-05, "loss": 0.0127, "step": 22200 }, { "grad_norm": 0.1713247448205948, "learning_rate": 7.449955758221183e-05, "loss": 0.0123, "step": 22210 }, { "grad_norm": 0.1660010814666748, "learning_rate": 7.447553094633615e-05, "loss": 0.0139, "step": 22220 }, { "grad_norm": 0.22075694799423218, "learning_rate": 7.445149687544039e-05, "loss": 0.0119, "step": 22230 }, { "grad_norm": 0.1619454175233841, "learning_rate": 7.44274553768255e-05, "loss": 0.0132, "step": 22240 }, { "grad_norm": 0.18113099038600922, "learning_rate": 7.440340645779464e-05, "loss": 0.0175, "step": 22250 }, { "grad_norm": 0.1577557623386383, "learning_rate": 7.437935012565322e-05, "loss": 0.0136, "step": 22260 }, { "grad_norm": 0.13912491500377655, "learning_rate": 7.435528638770893e-05, "loss": 0.011, "step": 22270 }, { "grad_norm": 0.18002864718437195, "learning_rate": 7.433121525127171e-05, "loss": 0.0116, "step": 22280 }, { "grad_norm": 0.1419007033109665, "learning_rate": 7.430713672365371e-05, "loss": 0.0123, "step": 22290 }, { "grad_norm": 0.12496695667505264, "learning_rate": 7.428305081216938e-05, "loss": 0.0101, "step": 22300 }, { "grad_norm": 0.17858773469924927, "learning_rate": 7.425895752413536e-05, "loss": 0.0116, "step": 22310 }, { "grad_norm": 0.1858450472354889, "learning_rate": 7.423485686687057e-05, "loss": 0.0113, "step": 22320 }, { "grad_norm": 0.18587982654571533, "learning_rate": 7.421074884769616e-05, "loss": 0.0144, "step": 22330 }, { "grad_norm": 0.1669781506061554, "learning_rate": 7.418663347393548e-05, "loss": 0.013, "step": 22340 }, { "grad_norm": 0.15790146589279175, "learning_rate": 7.416251075291418e-05, "loss": 0.0107, "step": 22350 }, { "grad_norm": 0.1801101267337799, "learning_rate": 7.413838069196007e-05, "loss": 0.012, "step": 22360 }, { "grad_norm": 0.19033674895763397, "learning_rate": 7.411424329840324e-05, "loss": 0.0113, "step": 22370 }, { "grad_norm": 0.1760331094264984, "learning_rate": 7.409009857957601e-05, "loss": 0.0104, "step": 22380 }, { "grad_norm": 0.24575220048427582, "learning_rate": 7.40659465428129e-05, "loss": 0.0146, "step": 22390 }, { "grad_norm": 0.1551329344511032, "learning_rate": 7.404178719545063e-05, "loss": 0.0135, "step": 22400 }, { "grad_norm": 0.1806325763463974, "learning_rate": 7.401762054482822e-05, "loss": 0.0125, "step": 22410 }, { "grad_norm": 0.12041919678449631, "learning_rate": 7.39934465982868e-05, "loss": 0.013, "step": 22420 }, { "grad_norm": 0.17324374616146088, "learning_rate": 7.396926536316984e-05, "loss": 0.0132, "step": 22430 }, { "grad_norm": 0.1749378889799118, "learning_rate": 7.394507684682293e-05, "loss": 0.0145, "step": 22440 }, { "grad_norm": 0.1940423846244812, "learning_rate": 7.392088105659393e-05, "loss": 0.0116, "step": 22450 }, { "grad_norm": 0.2075999528169632, "learning_rate": 7.389667799983284e-05, "loss": 0.0108, "step": 22460 }, { "grad_norm": 0.13052906095981598, "learning_rate": 7.387246768389193e-05, "loss": 0.0107, "step": 22470 }, { "grad_norm": 0.2081504911184311, "learning_rate": 7.384825011612563e-05, "loss": 0.0106, "step": 22480 }, { "grad_norm": 0.15046757459640503, "learning_rate": 7.382402530389066e-05, "loss": 0.0104, "step": 22490 }, { "grad_norm": 0.17905877530574799, "learning_rate": 7.379979325454582e-05, "loss": 0.0123, "step": 22500 }, { "grad_norm": 0.1715562790632248, "learning_rate": 7.37755539754522e-05, "loss": 0.0115, "step": 22510 }, { "grad_norm": 0.12567193806171417, "learning_rate": 7.375130747397302e-05, "loss": 0.0104, "step": 22520 }, { "grad_norm": 0.12104231864213943, "learning_rate": 7.372705375747377e-05, "loss": 0.0118, "step": 22530 }, { "grad_norm": 0.1889556348323822, "learning_rate": 7.370279283332205e-05, "loss": 0.0123, "step": 22540 }, { "grad_norm": 0.13925382494926453, "learning_rate": 7.36785247088877e-05, "loss": 0.0102, "step": 22550 }, { "grad_norm": 0.188195139169693, "learning_rate": 7.365424939154275e-05, "loss": 0.0113, "step": 22560 }, { "grad_norm": 0.17504571378231049, "learning_rate": 7.362996688866138e-05, "loss": 0.0119, "step": 22570 }, { "grad_norm": 0.10947179794311523, "learning_rate": 7.360567720761999e-05, "loss": 0.0094, "step": 22580 }, { "grad_norm": 0.14903965592384338, "learning_rate": 7.358138035579711e-05, "loss": 0.01, "step": 22590 }, { "grad_norm": 0.21834591031074524, "learning_rate": 7.355707634057354e-05, "loss": 0.0119, "step": 22600 }, { "grad_norm": 0.16476909816265106, "learning_rate": 7.353276516933215e-05, "loss": 0.0112, "step": 22610 }, { "grad_norm": 0.20400598645210266, "learning_rate": 7.350844684945806e-05, "loss": 0.0102, "step": 22620 }, { "grad_norm": 0.16072043776512146, "learning_rate": 7.348412138833851e-05, "loss": 0.0111, "step": 22630 }, { "grad_norm": 0.17533530294895172, "learning_rate": 7.345978879336295e-05, "loss": 0.0118, "step": 22640 }, { "grad_norm": 0.1374957263469696, "learning_rate": 7.343544907192296e-05, "loss": 0.0137, "step": 22650 }, { "grad_norm": 0.13098597526550293, "learning_rate": 7.341110223141235e-05, "loss": 0.0121, "step": 22660 }, { "grad_norm": 0.201637864112854, "learning_rate": 7.3386748279227e-05, "loss": 0.016, "step": 22670 }, { "grad_norm": 0.15712393820285797, "learning_rate": 7.336238722276501e-05, "loss": 0.0125, "step": 22680 }, { "grad_norm": 0.23508912324905396, "learning_rate": 7.333801906942663e-05, "loss": 0.0125, "step": 22690 }, { "grad_norm": 0.13728126883506775, "learning_rate": 7.331364382661428e-05, "loss": 0.0131, "step": 22700 }, { "grad_norm": 0.1740349382162094, "learning_rate": 7.328926150173248e-05, "loss": 0.0138, "step": 22710 }, { "grad_norm": 0.1470784842967987, "learning_rate": 7.326487210218795e-05, "loss": 0.011, "step": 22720 }, { "grad_norm": 0.16224507987499237, "learning_rate": 7.324047563538955e-05, "loss": 0.0113, "step": 22730 }, { "grad_norm": 0.1723995804786682, "learning_rate": 7.321607210874828e-05, "loss": 0.0119, "step": 22740 }, { "grad_norm": 0.15825122594833374, "learning_rate": 7.31916615296773e-05, "loss": 0.011, "step": 22750 }, { "grad_norm": 0.21725349128246307, "learning_rate": 7.316724390559188e-05, "loss": 0.0102, "step": 22760 }, { "grad_norm": 0.16769151389598846, "learning_rate": 7.314281924390946e-05, "loss": 0.0135, "step": 22770 }, { "grad_norm": 0.20859095454216003, "learning_rate": 7.311838755204959e-05, "loss": 0.0117, "step": 22780 }, { "grad_norm": 0.18368493020534515, "learning_rate": 7.3093948837434e-05, "loss": 0.0113, "step": 22790 }, { "grad_norm": 0.1560908854007721, "learning_rate": 7.306950310748651e-05, "loss": 0.0119, "step": 22800 }, { "grad_norm": 0.2338186353445053, "learning_rate": 7.304505036963311e-05, "loss": 0.014, "step": 22810 }, { "grad_norm": 0.16892550885677338, "learning_rate": 7.302059063130186e-05, "loss": 0.0091, "step": 22820 }, { "grad_norm": 0.20232340693473816, "learning_rate": 7.2996123899923e-05, "loss": 0.0116, "step": 22830 }, { "grad_norm": 0.1686621606349945, "learning_rate": 7.297165018292886e-05, "loss": 0.0103, "step": 22840 }, { "grad_norm": 0.22443228960037231, "learning_rate": 7.294716948775396e-05, "loss": 0.0142, "step": 22850 }, { "grad_norm": 0.17994746565818787, "learning_rate": 7.292268182183484e-05, "loss": 0.0129, "step": 22860 }, { "grad_norm": 0.20750127732753754, "learning_rate": 7.28981871926102e-05, "loss": 0.0129, "step": 22870 }, { "grad_norm": 0.15031135082244873, "learning_rate": 7.28736856075209e-05, "loss": 0.012, "step": 22880 }, { "grad_norm": 0.17473061382770538, "learning_rate": 7.284917707400985e-05, "loss": 0.011, "step": 22890 }, { "grad_norm": 0.18782617151737213, "learning_rate": 7.282466159952212e-05, "loss": 0.0142, "step": 22900 }, { "grad_norm": 0.13879907131195068, "learning_rate": 7.280013919150483e-05, "loss": 0.0123, "step": 22910 }, { "grad_norm": 0.14000850915908813, "learning_rate": 7.277560985740728e-05, "loss": 0.013, "step": 22920 }, { "grad_norm": 0.18398405611515045, "learning_rate": 7.275107360468079e-05, "loss": 0.0115, "step": 22930 }, { "grad_norm": 0.16797271370887756, "learning_rate": 7.272653044077885e-05, "loss": 0.0125, "step": 22940 }, { "grad_norm": 0.19461485743522644, "learning_rate": 7.270198037315703e-05, "loss": 0.0143, "step": 22950 }, { "grad_norm": 0.2843029499053955, "learning_rate": 7.267742340927297e-05, "loss": 0.0134, "step": 22960 }, { "grad_norm": 0.16121266782283783, "learning_rate": 7.265285955658645e-05, "loss": 0.0134, "step": 22970 }, { "grad_norm": 0.12052371352910995, "learning_rate": 7.26282888225593e-05, "loss": 0.0114, "step": 22980 }, { "grad_norm": 0.15333124995231628, "learning_rate": 7.260371121465548e-05, "loss": 0.0096, "step": 22990 }, { "grad_norm": 0.18748906254768372, "learning_rate": 7.2579126740341e-05, "loss": 0.0122, "step": 23000 }, { "grad_norm": 0.1267554759979248, "learning_rate": 7.2554535407084e-05, "loss": 0.011, "step": 23010 }, { "grad_norm": 0.211029052734375, "learning_rate": 7.252993722235464e-05, "loss": 0.0112, "step": 23020 }, { "grad_norm": 0.13090677559375763, "learning_rate": 7.250533219362523e-05, "loss": 0.0097, "step": 23030 }, { "grad_norm": 0.153087317943573, "learning_rate": 7.248072032837012e-05, "loss": 0.0117, "step": 23040 }, { "grad_norm": 0.17736949026584625, "learning_rate": 7.245610163406575e-05, "loss": 0.01, "step": 23050 }, { "grad_norm": 0.20081038773059845, "learning_rate": 7.243147611819061e-05, "loss": 0.0127, "step": 23060 }, { "grad_norm": 0.19183449447155, "learning_rate": 7.240684378822531e-05, "loss": 0.0126, "step": 23070 }, { "grad_norm": 0.17746706306934357, "learning_rate": 7.238220465165248e-05, "loss": 0.0132, "step": 23080 }, { "grad_norm": 0.21774281561374664, "learning_rate": 7.235755871595684e-05, "loss": 0.0123, "step": 23090 }, { "grad_norm": 0.15074512362480164, "learning_rate": 7.233290598862517e-05, "loss": 0.0114, "step": 23100 }, { "grad_norm": 0.16752979159355164, "learning_rate": 7.230824647714635e-05, "loss": 0.0109, "step": 23110 }, { "grad_norm": 0.16502957046031952, "learning_rate": 7.228358018901124e-05, "loss": 0.0098, "step": 23120 }, { "grad_norm": 0.17359527945518494, "learning_rate": 7.225890713171286e-05, "loss": 0.0116, "step": 23130 }, { "grad_norm": 0.15578356385231018, "learning_rate": 7.223422731274618e-05, "loss": 0.0103, "step": 23140 }, { "grad_norm": 0.13405634462833405, "learning_rate": 7.220954073960832e-05, "loss": 0.011, "step": 23150 }, { "grad_norm": 0.16401828825473785, "learning_rate": 7.218484741979838e-05, "loss": 0.0106, "step": 23160 }, { "grad_norm": 0.1413487046957016, "learning_rate": 7.216014736081756e-05, "loss": 0.0102, "step": 23170 }, { "grad_norm": 0.16507241129875183, "learning_rate": 7.213544057016906e-05, "loss": 0.0103, "step": 23180 }, { "grad_norm": 0.20376083254814148, "learning_rate": 7.211072705535819e-05, "loss": 0.0136, "step": 23190 }, { "grad_norm": 0.17869792878627777, "learning_rate": 7.208600682389224e-05, "loss": 0.0156, "step": 23200 }, { "grad_norm": 0.1763104796409607, "learning_rate": 7.206127988328055e-05, "loss": 0.0103, "step": 23210 }, { "grad_norm": 0.17413882911205292, "learning_rate": 7.203654624103453e-05, "loss": 0.0096, "step": 23220 }, { "grad_norm": 0.2339128851890564, "learning_rate": 7.201180590466761e-05, "loss": 0.0116, "step": 23230 }, { "grad_norm": 0.15292124450206757, "learning_rate": 7.198705888169523e-05, "loss": 0.0118, "step": 23240 }, { "grad_norm": 0.13795657455921173, "learning_rate": 7.196230517963491e-05, "loss": 0.0091, "step": 23250 }, { "grad_norm": 0.17857733368873596, "learning_rate": 7.193754480600615e-05, "loss": 0.0099, "step": 23260 }, { "grad_norm": 0.12599767744541168, "learning_rate": 7.19127777683305e-05, "loss": 0.0098, "step": 23270 }, { "grad_norm": 0.16473855078220367, "learning_rate": 7.188800407413156e-05, "loss": 0.0133, "step": 23280 }, { "grad_norm": 0.18616081774234772, "learning_rate": 7.186322373093489e-05, "loss": 0.0115, "step": 23290 }, { "grad_norm": 0.14844922721385956, "learning_rate": 7.18384367462681e-05, "loss": 0.0113, "step": 23300 }, { "grad_norm": 0.14664813876152039, "learning_rate": 7.181364312766085e-05, "loss": 0.0121, "step": 23310 }, { "grad_norm": 0.22185179591178894, "learning_rate": 7.178884288264477e-05, "loss": 0.0134, "step": 23320 }, { "grad_norm": 0.1766119748353958, "learning_rate": 7.176403601875353e-05, "loss": 0.012, "step": 23330 }, { "grad_norm": 0.16931365430355072, "learning_rate": 7.173922254352279e-05, "loss": 0.0108, "step": 23340 }, { "grad_norm": 0.1827923208475113, "learning_rate": 7.171440246449024e-05, "loss": 0.0123, "step": 23350 }, { "grad_norm": 0.16851960122585297, "learning_rate": 7.168957578919555e-05, "loss": 0.0141, "step": 23360 }, { "grad_norm": 0.12390138953924179, "learning_rate": 7.16647425251804e-05, "loss": 0.014, "step": 23370 }, { "grad_norm": 0.12317881733179092, "learning_rate": 7.163990267998852e-05, "loss": 0.0158, "step": 23380 }, { "grad_norm": 0.15085311233997345, "learning_rate": 7.161505626116556e-05, "loss": 0.0128, "step": 23390 }, { "grad_norm": 0.188528910279274, "learning_rate": 7.159020327625923e-05, "loss": 0.0117, "step": 23400 }, { "grad_norm": 0.15639448165893555, "learning_rate": 7.15653437328192e-05, "loss": 0.0127, "step": 23410 }, { "grad_norm": 0.16873562335968018, "learning_rate": 7.154047763839713e-05, "loss": 0.0151, "step": 23420 }, { "grad_norm": 0.24084924161434174, "learning_rate": 7.15156050005467e-05, "loss": 0.0153, "step": 23430 }, { "grad_norm": 0.1649945229291916, "learning_rate": 7.149072582682357e-05, "loss": 0.0114, "step": 23440 }, { "grad_norm": 0.13813960552215576, "learning_rate": 7.146584012478535e-05, "loss": 0.0139, "step": 23450 }, { "grad_norm": 0.21215388178825378, "learning_rate": 7.144094790199169e-05, "loss": 0.0132, "step": 23460 }, { "grad_norm": 0.17439880967140198, "learning_rate": 7.141604916600415e-05, "loss": 0.0123, "step": 23470 }, { "grad_norm": 0.15887263417243958, "learning_rate": 7.139114392438635e-05, "loss": 0.0107, "step": 23480 }, { "grad_norm": 0.18467043340206146, "learning_rate": 7.136623218470382e-05, "loss": 0.0091, "step": 23490 }, { "grad_norm": 0.23317134380340576, "learning_rate": 7.13413139545241e-05, "loss": 0.0112, "step": 23500 }, { "grad_norm": 0.12570923566818237, "learning_rate": 7.131638924141668e-05, "loss": 0.0124, "step": 23510 }, { "grad_norm": 0.2159280627965927, "learning_rate": 7.129145805295304e-05, "loss": 0.0126, "step": 23520 }, { "grad_norm": 0.1632796823978424, "learning_rate": 7.126652039670661e-05, "loss": 0.0136, "step": 23530 }, { "grad_norm": 0.1533907949924469, "learning_rate": 7.124157628025278e-05, "loss": 0.0129, "step": 23540 }, { "grad_norm": 0.1636648178100586, "learning_rate": 7.121662571116894e-05, "loss": 0.0109, "step": 23550 }, { "grad_norm": 0.18439266085624695, "learning_rate": 7.119166869703441e-05, "loss": 0.009, "step": 23560 }, { "grad_norm": 0.17363785207271576, "learning_rate": 7.116670524543044e-05, "loss": 0.0104, "step": 23570 }, { "grad_norm": 0.5763436555862427, "learning_rate": 7.114173536394032e-05, "loss": 0.0172, "step": 23580 }, { "grad_norm": 0.19986064732074738, "learning_rate": 7.111675906014917e-05, "loss": 0.0126, "step": 23590 }, { "grad_norm": 0.2046046108007431, "learning_rate": 7.109177634164421e-05, "loss": 0.011, "step": 23600 }, { "grad_norm": 0.20825861394405365, "learning_rate": 7.106678721601449e-05, "loss": 0.0104, "step": 23610 }, { "grad_norm": 0.18584531545639038, "learning_rate": 7.104179169085103e-05, "loss": 0.0109, "step": 23620 }, { "grad_norm": 0.14219674468040466, "learning_rate": 7.101678977374683e-05, "loss": 0.0114, "step": 23630 }, { "grad_norm": 0.16503114998340607, "learning_rate": 7.099178147229685e-05, "loss": 0.012, "step": 23640 }, { "grad_norm": 0.14622929692268372, "learning_rate": 7.096676679409789e-05, "loss": 0.0103, "step": 23650 }, { "grad_norm": 0.14497458934783936, "learning_rate": 7.094174574674877e-05, "loss": 0.0099, "step": 23660 }, { "grad_norm": 0.20721295475959778, "learning_rate": 7.091671833785025e-05, "loss": 0.011, "step": 23670 }, { "grad_norm": 0.17211541533470154, "learning_rate": 7.089168457500493e-05, "loss": 0.0096, "step": 23680 }, { "grad_norm": 0.20502124726772308, "learning_rate": 7.086664446581747e-05, "loss": 0.0096, "step": 23690 }, { "grad_norm": 0.12819524109363556, "learning_rate": 7.084159801789438e-05, "loss": 0.0106, "step": 23700 }, { "grad_norm": 0.16310255229473114, "learning_rate": 7.081654523884411e-05, "loss": 0.0103, "step": 23710 }, { "grad_norm": 0.18584245443344116, "learning_rate": 7.0791486136277e-05, "loss": 0.0122, "step": 23720 }, { "grad_norm": 0.1984422355890274, "learning_rate": 7.07664207178054e-05, "loss": 0.0112, "step": 23730 }, { "grad_norm": 0.16079163551330566, "learning_rate": 7.074134899104345e-05, "loss": 0.0125, "step": 23740 }, { "grad_norm": 0.15527112782001495, "learning_rate": 7.071627096360735e-05, "loss": 0.011, "step": 23750 }, { "grad_norm": 0.19263045489788055, "learning_rate": 7.069118664311511e-05, "loss": 0.0134, "step": 23760 }, { "grad_norm": 0.314259797334671, "learning_rate": 7.06660960371867e-05, "loss": 0.0125, "step": 23770 }, { "grad_norm": 0.2271421253681183, "learning_rate": 7.064099915344396e-05, "loss": 0.0119, "step": 23780 }, { "grad_norm": 0.22278203070163727, "learning_rate": 7.061589599951066e-05, "loss": 0.0112, "step": 23790 }, { "grad_norm": 0.14477333426475525, "learning_rate": 7.05907865830125e-05, "loss": 0.0117, "step": 23800 }, { "grad_norm": 0.22522729635238647, "learning_rate": 7.056567091157703e-05, "loss": 0.0108, "step": 23810 }, { "grad_norm": 0.18953977525234222, "learning_rate": 7.054054899283375e-05, "loss": 0.0132, "step": 23820 }, { "grad_norm": 0.17472168803215027, "learning_rate": 7.051542083441403e-05, "loss": 0.0094, "step": 23830 }, { "grad_norm": 0.14251646399497986, "learning_rate": 7.049028644395113e-05, "loss": 0.0104, "step": 23840 }, { "grad_norm": 0.12777183949947357, "learning_rate": 7.046514582908024e-05, "loss": 0.0108, "step": 23850 }, { "grad_norm": 0.1861669272184372, "learning_rate": 7.043999899743838e-05, "loss": 0.0103, "step": 23860 }, { "grad_norm": 0.1429280936717987, "learning_rate": 7.041484595666451e-05, "loss": 0.0116, "step": 23870 }, { "grad_norm": 0.12824885547161102, "learning_rate": 7.038968671439948e-05, "loss": 0.0102, "step": 23880 }, { "grad_norm": 0.18769249320030212, "learning_rate": 7.036452127828596e-05, "loss": 0.0127, "step": 23890 }, { "grad_norm": 0.17568187415599823, "learning_rate": 7.033934965596859e-05, "loss": 0.0117, "step": 23900 }, { "grad_norm": 0.15580984950065613, "learning_rate": 7.031417185509381e-05, "loss": 0.011, "step": 23910 }, { "grad_norm": 0.16340461373329163, "learning_rate": 7.028898788331e-05, "loss": 0.01, "step": 23920 }, { "grad_norm": 0.18306197226047516, "learning_rate": 7.026379774826736e-05, "loss": 0.0099, "step": 23930 }, { "grad_norm": 0.160444438457489, "learning_rate": 7.0238601457618e-05, "loss": 0.0112, "step": 23940 }, { "grad_norm": 0.17667743563652039, "learning_rate": 7.02133990190159e-05, "loss": 0.0117, "step": 23950 }, { "grad_norm": 0.17828941345214844, "learning_rate": 7.018819044011687e-05, "loss": 0.0119, "step": 23960 }, { "grad_norm": 0.12197089940309525, "learning_rate": 7.016297572857863e-05, "loss": 0.0085, "step": 23970 }, { "grad_norm": 0.1715010106563568, "learning_rate": 7.013775489206072e-05, "loss": 0.0085, "step": 23980 }, { "grad_norm": 0.19831928610801697, "learning_rate": 7.01125279382246e-05, "loss": 0.0107, "step": 23990 }, { "grad_norm": 0.17121922969818115, "learning_rate": 7.008729487473351e-05, "loss": 0.011, "step": 24000 }, { "grad_norm": 0.16703198850154877, "learning_rate": 7.006205570925263e-05, "loss": 0.0102, "step": 24010 }, { "grad_norm": 0.13932228088378906, "learning_rate": 7.003681044944892e-05, "loss": 0.0124, "step": 24020 }, { "grad_norm": 0.11682816594839096, "learning_rate": 7.001155910299126e-05, "loss": 0.0095, "step": 24030 }, { "grad_norm": 0.1747809499502182, "learning_rate": 6.99863016775503e-05, "loss": 0.0121, "step": 24040 }, { "grad_norm": 0.17871850728988647, "learning_rate": 6.996103818079859e-05, "loss": 0.0116, "step": 24050 }, { "grad_norm": 0.15700621902942657, "learning_rate": 6.993576862041054e-05, "loss": 0.0142, "step": 24060 }, { "grad_norm": 0.20004938542842865, "learning_rate": 6.991049300406235e-05, "loss": 0.0111, "step": 24070 }, { "grad_norm": 0.13137249648571014, "learning_rate": 6.988521133943209e-05, "loss": 0.0091, "step": 24080 }, { "grad_norm": 0.2499111294746399, "learning_rate": 6.985992363419966e-05, "loss": 0.0136, "step": 24090 }, { "grad_norm": 0.15881101787090302, "learning_rate": 6.983462989604682e-05, "loss": 0.0124, "step": 24100 }, { "grad_norm": 0.12548868358135223, "learning_rate": 6.980933013265709e-05, "loss": 0.0105, "step": 24110 }, { "grad_norm": 0.15080714225769043, "learning_rate": 6.978402435171592e-05, "loss": 0.0119, "step": 24120 }, { "grad_norm": 0.20335225760936737, "learning_rate": 6.975871256091052e-05, "loss": 0.0102, "step": 24130 }, { "grad_norm": 0.16042613983154297, "learning_rate": 6.973339476792995e-05, "loss": 0.01, "step": 24140 }, { "grad_norm": 0.13215391337871552, "learning_rate": 6.970807098046505e-05, "loss": 0.0099, "step": 24150 }, { "grad_norm": 0.15222732722759247, "learning_rate": 6.968274120620858e-05, "loss": 0.0122, "step": 24160 }, { "grad_norm": 0.16139432787895203, "learning_rate": 6.965740545285499e-05, "loss": 0.0094, "step": 24170 }, { "grad_norm": 0.22044889628887177, "learning_rate": 6.963206372810068e-05, "loss": 0.0139, "step": 24180 }, { "grad_norm": 0.17919054627418518, "learning_rate": 6.960671603964375e-05, "loss": 0.0134, "step": 24190 }, { "grad_norm": 0.14107970893383026, "learning_rate": 6.958136239518418e-05, "loss": 0.0103, "step": 24200 }, { "grad_norm": 0.15469670295715332, "learning_rate": 6.955600280242371e-05, "loss": 0.0106, "step": 24210 }, { "grad_norm": 0.14845925569534302, "learning_rate": 6.953063726906596e-05, "loss": 0.0137, "step": 24220 }, { "grad_norm": 0.158721461892128, "learning_rate": 6.950526580281626e-05, "loss": 0.0127, "step": 24230 }, { "grad_norm": 0.17607218027114868, "learning_rate": 6.947988841138184e-05, "loss": 0.0104, "step": 24240 }, { "grad_norm": 0.15800365805625916, "learning_rate": 6.945450510247165e-05, "loss": 0.0113, "step": 24250 }, { "grad_norm": 0.1355629861354828, "learning_rate": 6.942911588379647e-05, "loss": 0.0106, "step": 24260 }, { "grad_norm": 0.18515510857105255, "learning_rate": 6.940372076306888e-05, "loss": 0.0098, "step": 24270 }, { "grad_norm": 0.1384684145450592, "learning_rate": 6.937831974800326e-05, "loss": 0.01, "step": 24280 }, { "grad_norm": 0.16107390820980072, "learning_rate": 6.935291284631574e-05, "loss": 0.013, "step": 24290 }, { "grad_norm": 0.10127615928649902, "learning_rate": 6.932750006572428e-05, "loss": 0.0109, "step": 24300 }, { "grad_norm": 0.5654267072677612, "learning_rate": 6.930208141394863e-05, "loss": 0.0111, "step": 24310 }, { "grad_norm": 0.15375381708145142, "learning_rate": 6.927665689871026e-05, "loss": 0.0109, "step": 24320 }, { "grad_norm": 0.12444242835044861, "learning_rate": 6.925122652773253e-05, "loss": 0.0111, "step": 24330 }, { "grad_norm": 0.15553639829158783, "learning_rate": 6.922579030874046e-05, "loss": 0.0113, "step": 24340 }, { "grad_norm": 0.14643502235412598, "learning_rate": 6.920034824946093e-05, "loss": 0.0104, "step": 24350 }, { "grad_norm": 0.15679141879081726, "learning_rate": 6.917490035762255e-05, "loss": 0.0122, "step": 24360 }, { "grad_norm": 0.168912872672081, "learning_rate": 6.914944664095573e-05, "loss": 0.0117, "step": 24370 }, { "grad_norm": 0.13377171754837036, "learning_rate": 6.912398710719264e-05, "loss": 0.0104, "step": 24380 }, { "grad_norm": 0.14731770753860474, "learning_rate": 6.90985217640672e-05, "loss": 0.011, "step": 24390 }, { "grad_norm": 0.15897846221923828, "learning_rate": 6.90730506193151e-05, "loss": 0.0104, "step": 24400 }, { "grad_norm": 0.16850464046001434, "learning_rate": 6.904757368067384e-05, "loss": 0.0115, "step": 24410 }, { "grad_norm": 0.11259033530950546, "learning_rate": 6.90220909558826e-05, "loss": 0.0102, "step": 24420 }, { "grad_norm": 0.1859765499830246, "learning_rate": 6.899660245268237e-05, "loss": 0.0128, "step": 24430 }, { "grad_norm": 0.162866473197937, "learning_rate": 6.897110817881592e-05, "loss": 0.0136, "step": 24440 }, { "grad_norm": 0.1620102971792221, "learning_rate": 6.894560814202769e-05, "loss": 0.0123, "step": 24450 }, { "grad_norm": 0.17919126152992249, "learning_rate": 6.892010235006394e-05, "loss": 0.0105, "step": 24460 }, { "grad_norm": 0.2487703114748001, "learning_rate": 6.889459081067264e-05, "loss": 0.0105, "step": 24470 }, { "grad_norm": 0.1594577580690384, "learning_rate": 6.886907353160356e-05, "loss": 0.0122, "step": 24480 }, { "grad_norm": 0.17152267694473267, "learning_rate": 6.884355052060814e-05, "loss": 0.014, "step": 24490 }, { "grad_norm": 0.17827998101711273, "learning_rate": 6.88180217854396e-05, "loss": 0.0157, "step": 24500 }, { "grad_norm": 0.16063331067562103, "learning_rate": 6.87924873338529e-05, "loss": 0.0119, "step": 24510 }, { "grad_norm": 0.13556474447250366, "learning_rate": 6.876694717360475e-05, "loss": 0.0124, "step": 24520 }, { "grad_norm": 0.15202440321445465, "learning_rate": 6.874140131245355e-05, "loss": 0.0111, "step": 24530 }, { "grad_norm": 0.14644457399845123, "learning_rate": 6.871584975815948e-05, "loss": 0.0148, "step": 24540 }, { "grad_norm": 0.1423528641462326, "learning_rate": 6.86902925184844e-05, "loss": 0.0114, "step": 24550 }, { "grad_norm": 0.1934996247291565, "learning_rate": 6.866472960119195e-05, "loss": 0.0118, "step": 24560 }, { "grad_norm": 0.18993361294269562, "learning_rate": 6.863916101404748e-05, "loss": 0.015, "step": 24570 }, { "grad_norm": 0.16360034048557281, "learning_rate": 6.8613586764818e-05, "loss": 0.0112, "step": 24580 }, { "grad_norm": 0.16797387599945068, "learning_rate": 6.858800686127233e-05, "loss": 0.0098, "step": 24590 }, { "grad_norm": 0.1930767446756363, "learning_rate": 6.856242131118097e-05, "loss": 0.0138, "step": 24600 }, { "grad_norm": 0.19186945259571075, "learning_rate": 6.853683012231614e-05, "loss": 0.0117, "step": 24610 }, { "grad_norm": 0.20375369489192963, "learning_rate": 6.851123330245173e-05, "loss": 0.013, "step": 24620 }, { "grad_norm": 0.1632332056760788, "learning_rate": 6.848563085936343e-05, "loss": 0.0098, "step": 24630 }, { "grad_norm": 0.18363292515277863, "learning_rate": 6.846002280082853e-05, "loss": 0.0097, "step": 24640 }, { "grad_norm": 0.17361091077327728, "learning_rate": 6.843440913462614e-05, "loss": 0.0111, "step": 24650 }, { "grad_norm": 0.17607784271240234, "learning_rate": 6.840878986853698e-05, "loss": 0.0113, "step": 24660 }, { "grad_norm": 0.17829598486423492, "learning_rate": 6.838316501034352e-05, "loss": 0.0114, "step": 24670 }, { "grad_norm": 0.16381937265396118, "learning_rate": 6.83575345678299e-05, "loss": 0.0118, "step": 24680 }, { "grad_norm": 0.1665642112493515, "learning_rate": 6.833189854878196e-05, "loss": 0.0119, "step": 24690 }, { "grad_norm": 0.1395958960056305, "learning_rate": 6.83062569609873e-05, "loss": 0.0113, "step": 24700 }, { "grad_norm": 0.17569592595100403, "learning_rate": 6.828060981223512e-05, "loss": 0.0096, "step": 24710 }, { "grad_norm": 0.1770053654909134, "learning_rate": 6.825495711031634e-05, "loss": 0.013, "step": 24720 }, { "grad_norm": 0.1943073272705078, "learning_rate": 6.822929886302359e-05, "loss": 0.0108, "step": 24730 }, { "grad_norm": 0.23522807657718658, "learning_rate": 6.820363507815116e-05, "loss": 0.0104, "step": 24740 }, { "grad_norm": 0.1738400161266327, "learning_rate": 6.817796576349501e-05, "loss": 0.0112, "step": 24750 }, { "grad_norm": 0.17125411331653595, "learning_rate": 6.815229092685285e-05, "loss": 0.0138, "step": 24760 }, { "grad_norm": 0.131092831492424, "learning_rate": 6.812661057602399e-05, "loss": 0.0095, "step": 24770 }, { "grad_norm": 0.13276070356369019, "learning_rate": 6.810092471880943e-05, "loss": 0.0101, "step": 24780 }, { "grad_norm": 0.1530490517616272, "learning_rate": 6.807523336301187e-05, "loss": 0.0117, "step": 24790 }, { "grad_norm": 0.20622505247592926, "learning_rate": 6.804953651643566e-05, "loss": 0.0109, "step": 24800 }, { "grad_norm": 0.24835462868213654, "learning_rate": 6.802383418688685e-05, "loss": 0.0133, "step": 24810 }, { "grad_norm": 0.6826775074005127, "learning_rate": 6.799812638217309e-05, "loss": 0.01, "step": 24820 }, { "grad_norm": 0.20080111920833588, "learning_rate": 6.797241311010373e-05, "loss": 0.0111, "step": 24830 }, { "grad_norm": 0.15835897624492645, "learning_rate": 6.794669437848982e-05, "loss": 0.0104, "step": 24840 }, { "grad_norm": 0.16997842490673065, "learning_rate": 6.792097019514402e-05, "loss": 0.0119, "step": 24850 }, { "grad_norm": 0.10773400962352753, "learning_rate": 6.789524056788064e-05, "loss": 0.0084, "step": 24860 }, { "grad_norm": 0.1656419038772583, "learning_rate": 6.786950550451567e-05, "loss": 0.0085, "step": 24870 }, { "grad_norm": 0.12358113378286362, "learning_rate": 6.784376501286676e-05, "loss": 0.0105, "step": 24880 }, { "grad_norm": 0.12907825410366058, "learning_rate": 6.781801910075316e-05, "loss": 0.0081, "step": 24890 }, { "grad_norm": 0.17780262231826782, "learning_rate": 6.779226777599581e-05, "loss": 0.0107, "step": 24900 }, { "grad_norm": 0.15165843069553375, "learning_rate": 6.776651104641729e-05, "loss": 0.0097, "step": 24910 }, { "grad_norm": 0.17261967062950134, "learning_rate": 6.774074891984183e-05, "loss": 0.0124, "step": 24920 }, { "grad_norm": 0.16701839864253998, "learning_rate": 6.771498140409526e-05, "loss": 0.011, "step": 24930 }, { "grad_norm": 0.1680184155702591, "learning_rate": 6.768920850700506e-05, "loss": 0.0112, "step": 24940 }, { "grad_norm": 0.15243291854858398, "learning_rate": 6.766343023640039e-05, "loss": 0.0108, "step": 24950 }, { "grad_norm": 0.1543406844139099, "learning_rate": 6.763764660011198e-05, "loss": 0.0107, "step": 24960 }, { "grad_norm": 0.164527028799057, "learning_rate": 6.761185760597223e-05, "loss": 0.0092, "step": 24970 }, { "grad_norm": 0.1712827980518341, "learning_rate": 6.758606326181515e-05, "loss": 0.0139, "step": 24980 }, { "grad_norm": 0.09431389719247818, "learning_rate": 6.75602635754764e-05, "loss": 0.0123, "step": 24990 }, { "grad_norm": 0.22923798859119415, "learning_rate": 6.75344585547932e-05, "loss": 0.0123, "step": 25000 }, { "grad_norm": 0.12871837615966797, "learning_rate": 6.750864820760449e-05, "loss": 0.0111, "step": 25010 }, { "grad_norm": 0.15529975295066833, "learning_rate": 6.748283254175072e-05, "loss": 0.0099, "step": 25020 }, { "grad_norm": 0.16034051775932312, "learning_rate": 6.745701156507404e-05, "loss": 0.0113, "step": 25030 }, { "grad_norm": 0.15979060530662537, "learning_rate": 6.743118528541818e-05, "loss": 0.0104, "step": 25040 }, { "grad_norm": 0.15842434763908386, "learning_rate": 6.740535371062846e-05, "loss": 0.0104, "step": 25050 }, { "grad_norm": 0.16099430620670319, "learning_rate": 6.737951684855185e-05, "loss": 0.0101, "step": 25060 }, { "grad_norm": 0.19790741801261902, "learning_rate": 6.735367470703691e-05, "loss": 0.0125, "step": 25070 }, { "grad_norm": 0.20027127861976624, "learning_rate": 6.732782729393379e-05, "loss": 0.0107, "step": 25080 }, { "grad_norm": 0.16228605806827545, "learning_rate": 6.730197461709425e-05, "loss": 0.0104, "step": 25090 }, { "grad_norm": 0.1622474193572998, "learning_rate": 6.727611668437164e-05, "loss": 0.0121, "step": 25100 }, { "grad_norm": 0.1837078034877777, "learning_rate": 6.725025350362094e-05, "loss": 0.0106, "step": 25110 }, { "grad_norm": 0.17330123484134674, "learning_rate": 6.72243850826987e-05, "loss": 0.0121, "step": 25120 }, { "grad_norm": 0.13495180010795593, "learning_rate": 6.719851142946305e-05, "loss": 0.0113, "step": 25130 }, { "grad_norm": 0.1893640160560608, "learning_rate": 6.717263255177372e-05, "loss": 0.0141, "step": 25140 }, { "grad_norm": 0.17785687744617462, "learning_rate": 6.714674845749205e-05, "loss": 0.0105, "step": 25150 }, { "grad_norm": 0.20819811522960663, "learning_rate": 6.712085915448092e-05, "loss": 0.0094, "step": 25160 }, { "grad_norm": 0.1558191180229187, "learning_rate": 6.709496465060486e-05, "loss": 0.01, "step": 25170 }, { "grad_norm": 0.1419392079114914, "learning_rate": 6.706906495372987e-05, "loss": 0.0104, "step": 25180 }, { "grad_norm": 0.14692221581935883, "learning_rate": 6.704316007172365e-05, "loss": 0.0114, "step": 25190 }, { "grad_norm": 0.1367548704147339, "learning_rate": 6.701725001245539e-05, "loss": 0.0084, "step": 25200 }, { "grad_norm": 0.16768132150173187, "learning_rate": 6.699133478379588e-05, "loss": 0.0091, "step": 25210 }, { "grad_norm": 0.16635535657405853, "learning_rate": 6.69654143936175e-05, "loss": 0.014, "step": 25220 }, { "grad_norm": 0.20107769966125488, "learning_rate": 6.693948884979419e-05, "loss": 0.0114, "step": 25230 }, { "grad_norm": 0.1590781956911087, "learning_rate": 6.691355816020142e-05, "loss": 0.01, "step": 25240 }, { "grad_norm": 0.182684987783432, "learning_rate": 6.688762233271624e-05, "loss": 0.0095, "step": 25250 }, { "grad_norm": 0.12868814170360565, "learning_rate": 6.68616813752173e-05, "loss": 0.0099, "step": 25260 }, { "grad_norm": 0.20184777677059174, "learning_rate": 6.683573529558477e-05, "loss": 0.0115, "step": 25270 }, { "grad_norm": 0.15900979936122894, "learning_rate": 6.680978410170037e-05, "loss": 0.0101, "step": 25280 }, { "grad_norm": 0.26713383197784424, "learning_rate": 6.678382780144741e-05, "loss": 0.013, "step": 25290 }, { "grad_norm": 0.1508447378873825, "learning_rate": 6.675786640271071e-05, "loss": 0.0114, "step": 25300 }, { "grad_norm": 0.15635646879673004, "learning_rate": 6.673189991337665e-05, "loss": 0.0127, "step": 25310 }, { "grad_norm": 0.1450241208076477, "learning_rate": 6.670592834133317e-05, "loss": 0.0129, "step": 25320 }, { "grad_norm": 0.17202886939048767, "learning_rate": 6.667995169446979e-05, "loss": 0.0097, "step": 25330 }, { "grad_norm": 0.2107439935207367, "learning_rate": 6.665396998067747e-05, "loss": 0.0111, "step": 25340 }, { "grad_norm": 0.17894704639911652, "learning_rate": 6.66279832078488e-05, "loss": 0.0094, "step": 25350 }, { "grad_norm": 0.20352692902088165, "learning_rate": 6.660199138387786e-05, "loss": 0.0106, "step": 25360 }, { "grad_norm": 0.17342983186244965, "learning_rate": 6.65759945166603e-05, "loss": 0.0124, "step": 25370 }, { "grad_norm": 0.12950168550014496, "learning_rate": 6.654999261409326e-05, "loss": 0.012, "step": 25380 }, { "grad_norm": 0.20561183989048004, "learning_rate": 6.652398568407544e-05, "loss": 0.0117, "step": 25390 }, { "grad_norm": 0.15396729111671448, "learning_rate": 6.649797373450707e-05, "loss": 0.0111, "step": 25400 }, { "grad_norm": 0.1239575743675232, "learning_rate": 6.647195677328988e-05, "loss": 0.0144, "step": 25410 }, { "grad_norm": 0.11635764688253403, "learning_rate": 6.644593480832712e-05, "loss": 0.0126, "step": 25420 }, { "grad_norm": 0.16757458448410034, "learning_rate": 6.641990784752363e-05, "loss": 0.0097, "step": 25430 }, { "grad_norm": 0.14023087918758392, "learning_rate": 6.639387589878566e-05, "loss": 0.013, "step": 25440 }, { "grad_norm": 0.11095253378152847, "learning_rate": 6.636783897002103e-05, "loss": 0.0088, "step": 25450 }, { "grad_norm": 0.15164944529533386, "learning_rate": 6.63417970691391e-05, "loss": 0.012, "step": 25460 }, { "grad_norm": 0.14355899393558502, "learning_rate": 6.63157502040507e-05, "loss": 0.0141, "step": 25470 }, { "grad_norm": 0.20373253524303436, "learning_rate": 6.628969838266819e-05, "loss": 0.0146, "step": 25480 }, { "grad_norm": 0.14014476537704468, "learning_rate": 6.626364161290541e-05, "loss": 0.0132, "step": 25490 }, { "grad_norm": 0.1861034482717514, "learning_rate": 6.623757990267774e-05, "loss": 0.0106, "step": 25500 }, { "grad_norm": 0.1144380122423172, "learning_rate": 6.621151325990201e-05, "loss": 0.011, "step": 25510 }, { "grad_norm": 0.17930154502391815, "learning_rate": 6.618544169249657e-05, "loss": 0.0141, "step": 25520 }, { "grad_norm": 0.15454809367656708, "learning_rate": 6.615936520838133e-05, "loss": 0.012, "step": 25530 }, { "grad_norm": 0.17539145052433014, "learning_rate": 6.613328381547759e-05, "loss": 0.0126, "step": 25540 }, { "grad_norm": 0.1514122188091278, "learning_rate": 6.610719752170821e-05, "loss": 0.0089, "step": 25550 }, { "grad_norm": 0.15907134115695953, "learning_rate": 6.60811063349975e-05, "loss": 0.0104, "step": 25560 }, { "grad_norm": 0.12535740435123444, "learning_rate": 6.605501026327127e-05, "loss": 0.0117, "step": 25570 }, { "grad_norm": 0.18020330369472504, "learning_rate": 6.602890931445685e-05, "loss": 0.0119, "step": 25580 }, { "grad_norm": 0.198246031999588, "learning_rate": 6.6002803496483e-05, "loss": 0.0135, "step": 25590 }, { "grad_norm": 0.21660524606704712, "learning_rate": 6.597669281727997e-05, "loss": 0.013, "step": 25600 }, { "grad_norm": 0.16434717178344727, "learning_rate": 6.595057728477949e-05, "loss": 0.0107, "step": 25610 }, { "grad_norm": 0.14502571523189545, "learning_rate": 6.59244569069148e-05, "loss": 0.0111, "step": 25620 }, { "grad_norm": 0.1783359795808792, "learning_rate": 6.589833169162054e-05, "loss": 0.012, "step": 25630 }, { "grad_norm": 0.1640496701002121, "learning_rate": 6.587220164683291e-05, "loss": 0.0105, "step": 25640 }, { "grad_norm": 0.13130079209804535, "learning_rate": 6.58460667804895e-05, "loss": 0.0111, "step": 25650 }, { "grad_norm": 0.14838863909244537, "learning_rate": 6.581992710052938e-05, "loss": 0.0117, "step": 25660 }, { "grad_norm": 0.18984675407409668, "learning_rate": 6.579378261489311e-05, "loss": 0.0159, "step": 25670 }, { "grad_norm": 0.1566551774740219, "learning_rate": 6.576763333152268e-05, "loss": 0.0118, "step": 25680 }, { "grad_norm": 0.09921462833881378, "learning_rate": 6.574147925836159e-05, "loss": 0.0085, "step": 25690 }, { "grad_norm": 0.17940911650657654, "learning_rate": 6.571532040335472e-05, "loss": 0.0122, "step": 25700 }, { "grad_norm": 0.17799319326877594, "learning_rate": 6.568915677444845e-05, "loss": 0.013, "step": 25710 }, { "grad_norm": 0.1432897448539734, "learning_rate": 6.56629883795906e-05, "loss": 0.0098, "step": 25720 }, { "grad_norm": 0.17128075659275055, "learning_rate": 6.563681522673043e-05, "loss": 0.0119, "step": 25730 }, { "grad_norm": 0.16848117113113403, "learning_rate": 6.561063732381867e-05, "loss": 0.0128, "step": 25740 }, { "grad_norm": 0.2049059122800827, "learning_rate": 6.558445467880745e-05, "loss": 0.0121, "step": 25750 }, { "grad_norm": 0.25369036197662354, "learning_rate": 6.55582672996504e-05, "loss": 0.0126, "step": 25760 }, { "grad_norm": 0.12546488642692566, "learning_rate": 6.553207519430253e-05, "loss": 0.0104, "step": 25770 }, { "grad_norm": 0.18090243637561798, "learning_rate": 6.550587837072032e-05, "loss": 0.0103, "step": 25780 }, { "grad_norm": 0.280173659324646, "learning_rate": 6.547967683686166e-05, "loss": 0.0126, "step": 25790 }, { "grad_norm": 0.1447996199131012, "learning_rate": 6.545347060068591e-05, "loss": 0.0102, "step": 25800 }, { "grad_norm": 0.1582382619380951, "learning_rate": 6.542725967015382e-05, "loss": 0.0114, "step": 25810 }, { "grad_norm": 0.15358088910579681, "learning_rate": 6.540104405322757e-05, "loss": 0.0107, "step": 25820 }, { "grad_norm": 0.1112448051571846, "learning_rate": 6.537482375787077e-05, "loss": 0.0123, "step": 25830 }, { "grad_norm": 0.15006554126739502, "learning_rate": 6.534859879204845e-05, "loss": 0.0117, "step": 25840 }, { "grad_norm": 0.10311344265937805, "learning_rate": 6.532236916372709e-05, "loss": 0.011, "step": 25850 }, { "grad_norm": 0.16268883645534515, "learning_rate": 6.529613488087454e-05, "loss": 0.0144, "step": 25860 }, { "grad_norm": 0.13913439214229584, "learning_rate": 6.526989595146009e-05, "loss": 0.0093, "step": 25870 }, { "grad_norm": 0.1968732327222824, "learning_rate": 6.524365238345441e-05, "loss": 0.0132, "step": 25880 }, { "grad_norm": 0.17848339676856995, "learning_rate": 6.521740418482964e-05, "loss": 0.0109, "step": 25890 }, { "grad_norm": 0.1551397144794464, "learning_rate": 6.519115136355925e-05, "loss": 0.0105, "step": 25900 }, { "grad_norm": 0.2006198614835739, "learning_rate": 6.51648939276182e-05, "loss": 0.0123, "step": 25910 }, { "grad_norm": 0.17419004440307617, "learning_rate": 6.513863188498277e-05, "loss": 0.0119, "step": 25920 }, { "grad_norm": 0.12141752988100052, "learning_rate": 6.511236524363068e-05, "loss": 0.0084, "step": 25930 }, { "grad_norm": 0.20103000104427338, "learning_rate": 6.508609401154104e-05, "loss": 0.01, "step": 25940 }, { "grad_norm": 0.12503036856651306, "learning_rate": 6.505981819669439e-05, "loss": 0.0109, "step": 25950 }, { "grad_norm": 0.13108153641223907, "learning_rate": 6.503353780707258e-05, "loss": 0.0096, "step": 25960 }, { "grad_norm": 0.14540526270866394, "learning_rate": 6.500725285065895e-05, "loss": 0.0078, "step": 25970 }, { "grad_norm": 0.13446126878261566, "learning_rate": 6.498096333543813e-05, "loss": 0.0098, "step": 25980 }, { "grad_norm": 0.13886648416519165, "learning_rate": 6.49546692693962e-05, "loss": 0.0093, "step": 25990 }, { "grad_norm": 0.12078068405389786, "learning_rate": 6.492837066052059e-05, "loss": 0.0087, "step": 26000 }, { "grad_norm": 0.1635040044784546, "learning_rate": 6.490206751680014e-05, "loss": 0.0151, "step": 26010 }, { "grad_norm": 0.17023468017578125, "learning_rate": 6.487575984622505e-05, "loss": 0.0088, "step": 26020 }, { "grad_norm": 0.17209765315055847, "learning_rate": 6.484944765678689e-05, "loss": 0.0122, "step": 26030 }, { "grad_norm": 0.1673210710287094, "learning_rate": 6.482313095647861e-05, "loss": 0.0097, "step": 26040 }, { "grad_norm": 0.15107110142707825, "learning_rate": 6.479680975329451e-05, "loss": 0.009, "step": 26050 }, { "grad_norm": 0.15358613431453705, "learning_rate": 6.477048405523031e-05, "loss": 0.0102, "step": 26060 }, { "grad_norm": 0.20796670019626617, "learning_rate": 6.474415387028304e-05, "loss": 0.0109, "step": 26070 }, { "grad_norm": 0.11308877170085907, "learning_rate": 6.471781920645114e-05, "loss": 0.0106, "step": 26080 }, { "grad_norm": 0.21033675968647003, "learning_rate": 6.469148007173434e-05, "loss": 0.0121, "step": 26090 }, { "grad_norm": 0.15872140228748322, "learning_rate": 6.466513647413381e-05, "loss": 0.0111, "step": 26100 }, { "grad_norm": 0.14171026647090912, "learning_rate": 6.463878842165203e-05, "loss": 0.0093, "step": 26110 }, { "grad_norm": 0.11856114119291306, "learning_rate": 6.461243592229286e-05, "loss": 0.0084, "step": 26120 }, { "grad_norm": 0.20780117809772491, "learning_rate": 6.458607898406146e-05, "loss": 0.0104, "step": 26130 }, { "grad_norm": 0.1308564841747284, "learning_rate": 6.455971761496439e-05, "loss": 0.0095, "step": 26140 }, { "grad_norm": 0.1497523933649063, "learning_rate": 6.453335182300953e-05, "loss": 0.0108, "step": 26150 }, { "grad_norm": 0.1378156542778015, "learning_rate": 6.450698161620612e-05, "loss": 0.0095, "step": 26160 }, { "grad_norm": 0.1562722772359848, "learning_rate": 6.448060700256473e-05, "loss": 0.0119, "step": 26170 }, { "grad_norm": 0.18039649724960327, "learning_rate": 6.445422799009726e-05, "loss": 0.0082, "step": 26180 }, { "grad_norm": 0.14946924149990082, "learning_rate": 6.442784458681699e-05, "loss": 0.0087, "step": 26190 }, { "grad_norm": 0.11544221639633179, "learning_rate": 6.440145680073847e-05, "loss": 0.0115, "step": 26200 }, { "grad_norm": 0.13610732555389404, "learning_rate": 6.437506463987762e-05, "loss": 0.0126, "step": 26210 }, { "grad_norm": 0.11169607937335968, "learning_rate": 6.434866811225168e-05, "loss": 0.0109, "step": 26220 }, { "grad_norm": 0.19146093726158142, "learning_rate": 6.432226722587923e-05, "loss": 0.0139, "step": 26230 }, { "grad_norm": 0.18908607959747314, "learning_rate": 6.429586198878015e-05, "loss": 0.0101, "step": 26240 }, { "grad_norm": 0.14198841154575348, "learning_rate": 6.426945240897566e-05, "loss": 0.0087, "step": 26250 }, { "grad_norm": 0.13669657707214355, "learning_rate": 6.424303849448829e-05, "loss": 0.0115, "step": 26260 }, { "grad_norm": 0.14527133107185364, "learning_rate": 6.42166202533419e-05, "loss": 0.0099, "step": 26270 }, { "grad_norm": 0.12976787984371185, "learning_rate": 6.419019769356164e-05, "loss": 0.0082, "step": 26280 }, { "grad_norm": 0.15863244235515594, "learning_rate": 6.416377082317398e-05, "loss": 0.009, "step": 26290 }, { "grad_norm": 0.1434449851512909, "learning_rate": 6.413733965020674e-05, "loss": 0.0119, "step": 26300 }, { "grad_norm": 0.13021379709243774, "learning_rate": 6.411090418268896e-05, "loss": 0.0096, "step": 26310 }, { "grad_norm": 0.15689249336719513, "learning_rate": 6.408446442865109e-05, "loss": 0.0099, "step": 26320 }, { "grad_norm": 0.1530851125717163, "learning_rate": 6.405802039612479e-05, "loss": 0.0113, "step": 26330 }, { "grad_norm": 0.2088930904865265, "learning_rate": 6.403157209314308e-05, "loss": 0.0118, "step": 26340 }, { "grad_norm": 0.17105770111083984, "learning_rate": 6.400511952774024e-05, "loss": 0.014, "step": 26350 }, { "grad_norm": 0.155032679438591, "learning_rate": 6.397866270795187e-05, "loss": 0.0097, "step": 26360 }, { "grad_norm": 0.169908344745636, "learning_rate": 6.395220164181489e-05, "loss": 0.0089, "step": 26370 }, { "grad_norm": 0.17218080163002014, "learning_rate": 6.39257363373674e-05, "loss": 0.0111, "step": 26380 }, { "grad_norm": 0.16341139376163483, "learning_rate": 6.389926680264892e-05, "loss": 0.0117, "step": 26390 }, { "grad_norm": 0.11131316423416138, "learning_rate": 6.387279304570017e-05, "loss": 0.0093, "step": 26400 }, { "grad_norm": 0.15495672821998596, "learning_rate": 6.384631507456319e-05, "loss": 0.0106, "step": 26410 }, { "grad_norm": 0.12723760306835175, "learning_rate": 6.381983289728126e-05, "loss": 0.01, "step": 26420 }, { "grad_norm": 0.10397426784038544, "learning_rate": 6.3793346521899e-05, "loss": 0.0099, "step": 26430 }, { "grad_norm": 0.14344927668571472, "learning_rate": 6.376685595646226e-05, "loss": 0.0116, "step": 26440 }, { "grad_norm": 0.12059877067804337, "learning_rate": 6.374036120901816e-05, "loss": 0.01, "step": 26450 }, { "grad_norm": 0.16448760032653809, "learning_rate": 6.371386228761514e-05, "loss": 0.0103, "step": 26460 }, { "grad_norm": 0.17599819600582123, "learning_rate": 6.368735920030283e-05, "loss": 0.0094, "step": 26470 }, { "grad_norm": 0.14219145476818085, "learning_rate": 6.366085195513218e-05, "loss": 0.0108, "step": 26480 }, { "grad_norm": 0.14177338778972626, "learning_rate": 6.363434056015543e-05, "loss": 0.0118, "step": 26490 }, { "grad_norm": 0.12093307822942734, "learning_rate": 6.360782502342599e-05, "loss": 0.009, "step": 26500 }, { "grad_norm": 0.133374884724617, "learning_rate": 6.358130535299862e-05, "loss": 0.0109, "step": 26510 }, { "grad_norm": 0.16065704822540283, "learning_rate": 6.355478155692926e-05, "loss": 0.0099, "step": 26520 }, { "grad_norm": 0.14233280718326569, "learning_rate": 6.352825364327517e-05, "loss": 0.0099, "step": 26530 }, { "grad_norm": 0.15172870457172394, "learning_rate": 6.350172162009482e-05, "loss": 0.0135, "step": 26540 }, { "grad_norm": 0.1519414782524109, "learning_rate": 6.347518549544793e-05, "loss": 0.0113, "step": 26550 }, { "grad_norm": 0.12334699928760529, "learning_rate": 6.344864527739547e-05, "loss": 0.0108, "step": 26560 }, { "grad_norm": 0.16323243081569672, "learning_rate": 6.342210097399966e-05, "loss": 0.0139, "step": 26570 }, { "grad_norm": 0.23056140542030334, "learning_rate": 6.339555259332398e-05, "loss": 0.0147, "step": 26580 }, { "grad_norm": 0.2130890041589737, "learning_rate": 6.33690001434331e-05, "loss": 0.0103, "step": 26590 }, { "grad_norm": 0.21326710283756256, "learning_rate": 6.334244363239296e-05, "loss": 0.0112, "step": 26600 }, { "grad_norm": 0.1875673532485962, "learning_rate": 6.331588306827073e-05, "loss": 0.013, "step": 26610 }, { "grad_norm": 0.14274565875530243, "learning_rate": 6.328931845913483e-05, "loss": 0.0129, "step": 26620 }, { "grad_norm": 0.25591957569122314, "learning_rate": 6.326274981305484e-05, "loss": 0.0104, "step": 26630 }, { "grad_norm": 0.1166575625538826, "learning_rate": 6.323617713810166e-05, "loss": 0.0103, "step": 26640 }, { "grad_norm": 0.1487191766500473, "learning_rate": 6.320960044234734e-05, "loss": 0.0096, "step": 26650 }, { "grad_norm": 0.14105810225009918, "learning_rate": 6.318301973386518e-05, "loss": 0.0103, "step": 26660 }, { "grad_norm": 0.1590202897787094, "learning_rate": 6.315643502072971e-05, "loss": 0.0105, "step": 26670 }, { "grad_norm": 0.14710581302642822, "learning_rate": 6.312984631101667e-05, "loss": 0.009, "step": 26680 }, { "grad_norm": 0.12992490828037262, "learning_rate": 6.310325361280297e-05, "loss": 0.0096, "step": 26690 }, { "grad_norm": 0.1754453033208847, "learning_rate": 6.30766569341668e-05, "loss": 0.0104, "step": 26700 }, { "grad_norm": 0.14969195425510406, "learning_rate": 6.305005628318753e-05, "loss": 0.0118, "step": 26710 }, { "grad_norm": 0.18842464685440063, "learning_rate": 6.302345166794572e-05, "loss": 0.0101, "step": 26720 }, { "grad_norm": 0.18546561896800995, "learning_rate": 6.299684309652316e-05, "loss": 0.0096, "step": 26730 }, { "grad_norm": 0.15282337367534637, "learning_rate": 6.297023057700283e-05, "loss": 0.0089, "step": 26740 }, { "grad_norm": 0.1518387794494629, "learning_rate": 6.294361411746891e-05, "loss": 0.0107, "step": 26750 }, { "grad_norm": 0.13774053752422333, "learning_rate": 6.291699372600677e-05, "loss": 0.0096, "step": 26760 }, { "grad_norm": 0.14947807788848877, "learning_rate": 6.2890369410703e-05, "loss": 0.0091, "step": 26770 }, { "grad_norm": 0.18124040961265564, "learning_rate": 6.286374117964534e-05, "loss": 0.0123, "step": 26780 }, { "grad_norm": 0.1256067305803299, "learning_rate": 6.283710904092277e-05, "loss": 0.0076, "step": 26790 }, { "grad_norm": 0.14810851216316223, "learning_rate": 6.281047300262542e-05, "loss": 0.0112, "step": 26800 }, { "grad_norm": 0.13465215265750885, "learning_rate": 6.278383307284461e-05, "loss": 0.0101, "step": 26810 }, { "grad_norm": 0.20757386088371277, "learning_rate": 6.275718925967284e-05, "loss": 0.0106, "step": 26820 }, { "grad_norm": 0.18485184013843536, "learning_rate": 6.273054157120382e-05, "loss": 0.0114, "step": 26830 }, { "grad_norm": 0.17105865478515625, "learning_rate": 6.270389001553238e-05, "loss": 0.0101, "step": 26840 }, { "grad_norm": 0.1757960170507431, "learning_rate": 6.26772346007546e-05, "loss": 0.0095, "step": 26850 }, { "grad_norm": 0.10743670165538788, "learning_rate": 6.265057533496767e-05, "loss": 0.009, "step": 26860 }, { "grad_norm": 0.22108349204063416, "learning_rate": 6.262391222626997e-05, "loss": 0.013, "step": 26870 }, { "grad_norm": 0.17412255704402924, "learning_rate": 6.259724528276106e-05, "loss": 0.0117, "step": 26880 }, { "grad_norm": 0.1578100174665451, "learning_rate": 6.257057451254162e-05, "loss": 0.0127, "step": 26890 }, { "grad_norm": 0.11584234982728958, "learning_rate": 6.254389992371357e-05, "loss": 0.0106, "step": 26900 }, { "grad_norm": 0.18269562721252441, "learning_rate": 6.25172215243799e-05, "loss": 0.0097, "step": 26910 }, { "grad_norm": 0.13444401323795319, "learning_rate": 6.249053932264486e-05, "loss": 0.0093, "step": 26920 }, { "grad_norm": 0.20541752874851227, "learning_rate": 6.246385332661376e-05, "loss": 0.0115, "step": 26930 }, { "grad_norm": 0.12951697409152985, "learning_rate": 6.24371635443931e-05, "loss": 0.0107, "step": 26940 }, { "grad_norm": 0.15568575263023376, "learning_rate": 6.241046998409054e-05, "loss": 0.0127, "step": 26950 }, { "grad_norm": 0.15839561820030212, "learning_rate": 6.238377265381489e-05, "loss": 0.0105, "step": 26960 }, { "grad_norm": 0.18894773721694946, "learning_rate": 6.235707156167607e-05, "loss": 0.0102, "step": 26970 }, { "grad_norm": 0.14707505702972412, "learning_rate": 6.233036671578519e-05, "loss": 0.0104, "step": 26980 }, { "grad_norm": 0.1278727501630783, "learning_rate": 6.230365812425445e-05, "loss": 0.0107, "step": 26990 }, { "grad_norm": 0.1583043932914734, "learning_rate": 6.227694579519724e-05, "loss": 0.0095, "step": 27000 }, { "grad_norm": 0.10000037401914597, "learning_rate": 6.225022973672805e-05, "loss": 0.0099, "step": 27010 }, { "grad_norm": 0.1532788723707199, "learning_rate": 6.222350995696253e-05, "loss": 0.0097, "step": 27020 }, { "grad_norm": 0.18468177318572998, "learning_rate": 6.21967864640174e-05, "loss": 0.0105, "step": 27030 }, { "grad_norm": 0.12312000244855881, "learning_rate": 6.217005926601059e-05, "loss": 0.0087, "step": 27040 }, { "grad_norm": 0.11225666850805283, "learning_rate": 6.214332837106111e-05, "loss": 0.0088, "step": 27050 }, { "grad_norm": 0.18372578918933868, "learning_rate": 6.21165937872891e-05, "loss": 0.0115, "step": 27060 }, { "grad_norm": 0.16427044570446014, "learning_rate": 6.208985552281582e-05, "loss": 0.0097, "step": 27070 }, { "grad_norm": 0.16742584109306335, "learning_rate": 6.206311358576364e-05, "loss": 0.0123, "step": 27080 }, { "grad_norm": 0.1345333606004715, "learning_rate": 6.203636798425608e-05, "loss": 0.0087, "step": 27090 }, { "grad_norm": 0.1504017412662506, "learning_rate": 6.20096187264177e-05, "loss": 0.0091, "step": 27100 }, { "grad_norm": 0.12451709061861038, "learning_rate": 6.198286582037425e-05, "loss": 0.0084, "step": 27110 }, { "grad_norm": 0.1351533830165863, "learning_rate": 6.195610927425256e-05, "loss": 0.0095, "step": 27120 }, { "grad_norm": 0.13844281435012817, "learning_rate": 6.192934909618056e-05, "loss": 0.0095, "step": 27130 }, { "grad_norm": 0.14024847745895386, "learning_rate": 6.190258529428728e-05, "loss": 0.0086, "step": 27140 }, { "grad_norm": 0.18496021628379822, "learning_rate": 6.187581787670285e-05, "loss": 0.0101, "step": 27150 }, { "grad_norm": 0.1483844667673111, "learning_rate": 6.184904685155852e-05, "loss": 0.0116, "step": 27160 }, { "grad_norm": 0.16765375435352325, "learning_rate": 6.18222722269866e-05, "loss": 0.0122, "step": 27170 }, { "grad_norm": 0.12199889868497849, "learning_rate": 6.179549401112053e-05, "loss": 0.0099, "step": 27180 }, { "grad_norm": 0.1610552966594696, "learning_rate": 6.176871221209482e-05, "loss": 0.0128, "step": 27190 }, { "grad_norm": 0.11035660654306412, "learning_rate": 6.174192683804508e-05, "loss": 0.0085, "step": 27200 }, { "grad_norm": 0.14147846400737762, "learning_rate": 6.1715137897108e-05, "loss": 0.0102, "step": 27210 }, { "grad_norm": 0.13188697397708893, "learning_rate": 6.168834539742134e-05, "loss": 0.0085, "step": 27220 }, { "grad_norm": 0.14263899624347687, "learning_rate": 6.166154934712397e-05, "loss": 0.0089, "step": 27230 }, { "grad_norm": 0.11893629282712936, "learning_rate": 6.163474975435581e-05, "loss": 0.0087, "step": 27240 }, { "grad_norm": 0.14067909121513367, "learning_rate": 6.160794662725787e-05, "loss": 0.0104, "step": 27250 }, { "grad_norm": 0.14711898565292358, "learning_rate": 6.158113997397222e-05, "loss": 0.0099, "step": 27260 }, { "grad_norm": 0.13909848034381866, "learning_rate": 6.155432980264205e-05, "loss": 0.0091, "step": 27270 }, { "grad_norm": 0.0882066935300827, "learning_rate": 6.152751612141156e-05, "loss": 0.0101, "step": 27280 }, { "grad_norm": 0.10727140307426453, "learning_rate": 6.150069893842602e-05, "loss": 0.0102, "step": 27290 }, { "grad_norm": 0.17133969068527222, "learning_rate": 6.147387826183182e-05, "loss": 0.0127, "step": 27300 }, { "grad_norm": 0.16673888266086578, "learning_rate": 6.144705409977635e-05, "loss": 0.0106, "step": 27310 }, { "grad_norm": 0.15818116068840027, "learning_rate": 6.142022646040808e-05, "loss": 0.0105, "step": 27320 }, { "grad_norm": 0.15253213047981262, "learning_rate": 6.139339535187653e-05, "loss": 0.0081, "step": 27330 }, { "grad_norm": 0.13986855745315552, "learning_rate": 6.136656078233232e-05, "loss": 0.0094, "step": 27340 }, { "grad_norm": 0.1791655570268631, "learning_rate": 6.133972275992707e-05, "loss": 0.0081, "step": 27350 }, { "grad_norm": 0.15878309309482574, "learning_rate": 6.131288129281342e-05, "loss": 0.0092, "step": 27360 }, { "grad_norm": 0.19032178819179535, "learning_rate": 6.128603638914516e-05, "loss": 0.011, "step": 27370 }, { "grad_norm": 0.16488806903362274, "learning_rate": 6.125918805707704e-05, "loss": 0.0108, "step": 27380 }, { "grad_norm": 0.1168365478515625, "learning_rate": 6.123233630476485e-05, "loss": 0.0101, "step": 27390 }, { "grad_norm": 0.18678684532642365, "learning_rate": 6.120548114036547e-05, "loss": 0.0098, "step": 27400 }, { "grad_norm": 0.16289658844470978, "learning_rate": 6.117862257203679e-05, "loss": 0.0101, "step": 27410 }, { "grad_norm": 0.10929526388645172, "learning_rate": 6.115176060793771e-05, "loss": 0.0097, "step": 27420 }, { "grad_norm": 0.14360985159873962, "learning_rate": 6.112489525622822e-05, "loss": 0.0105, "step": 27430 }, { "grad_norm": 0.15959127247333527, "learning_rate": 6.109802652506928e-05, "loss": 0.0105, "step": 27440 }, { "grad_norm": 0.17778116464614868, "learning_rate": 6.107115442262291e-05, "loss": 0.0117, "step": 27450 }, { "grad_norm": 0.11981134116649628, "learning_rate": 6.104427895705214e-05, "loss": 0.0086, "step": 27460 }, { "grad_norm": 0.11101911216974258, "learning_rate": 6.101740013652103e-05, "loss": 0.0112, "step": 27470 }, { "grad_norm": 0.14534416794776917, "learning_rate": 6.099051796919465e-05, "loss": 0.0103, "step": 27480 }, { "grad_norm": 0.1346181035041809, "learning_rate": 6.096363246323911e-05, "loss": 0.0093, "step": 27490 }, { "grad_norm": 0.1757805496454239, "learning_rate": 6.0936743626821504e-05, "loss": 0.0101, "step": 27500 }, { "grad_norm": 0.1485069990158081, "learning_rate": 6.090985146810996e-05, "loss": 0.0089, "step": 27510 }, { "grad_norm": 0.13551391661167145, "learning_rate": 6.088295599527357e-05, "loss": 0.0089, "step": 27520 }, { "grad_norm": 0.1573822796344757, "learning_rate": 6.085605721648252e-05, "loss": 0.0082, "step": 27530 }, { "grad_norm": 0.13945643603801727, "learning_rate": 6.082915513990792e-05, "loss": 0.0105, "step": 27540 }, { "grad_norm": 0.19370941817760468, "learning_rate": 6.080224977372192e-05, "loss": 0.01, "step": 27550 }, { "grad_norm": 0.12952940165996552, "learning_rate": 6.0775341126097666e-05, "loss": 0.0087, "step": 27560 }, { "grad_norm": 0.15455538034439087, "learning_rate": 6.074842920520926e-05, "loss": 0.0102, "step": 27570 }, { "grad_norm": 0.18173076212406158, "learning_rate": 6.072151401923186e-05, "loss": 0.0097, "step": 27580 }, { "grad_norm": 0.0927848070859909, "learning_rate": 6.069459557634159e-05, "loss": 0.0109, "step": 27590 }, { "grad_norm": 0.1418600231409073, "learning_rate": 6.066767388471557e-05, "loss": 0.0094, "step": 27600 }, { "grad_norm": 0.12877675890922546, "learning_rate": 6.064074895253188e-05, "loss": 0.0121, "step": 27610 }, { "grad_norm": 0.1961016207933426, "learning_rate": 6.061382078796961e-05, "loss": 0.0137, "step": 27620 }, { "grad_norm": 0.19288033246994019, "learning_rate": 6.0586889399208814e-05, "loss": 0.0109, "step": 27630 }, { "grad_norm": 0.20355454087257385, "learning_rate": 6.0559954794430565e-05, "loss": 0.0097, "step": 27640 }, { "grad_norm": 0.11328686773777008, "learning_rate": 6.053301698181687e-05, "loss": 0.009, "step": 27650 }, { "grad_norm": 0.13693800568580627, "learning_rate": 6.0506075969550725e-05, "loss": 0.0094, "step": 27660 }, { "grad_norm": 0.1314649134874344, "learning_rate": 6.047913176581609e-05, "loss": 0.0081, "step": 27670 }, { "grad_norm": 0.1810685396194458, "learning_rate": 6.0452184378797904e-05, "loss": 0.0076, "step": 27680 }, { "grad_norm": 0.17078685760498047, "learning_rate": 6.042523381668209e-05, "loss": 0.0095, "step": 27690 }, { "grad_norm": 0.1516137272119522, "learning_rate": 6.03982800876555e-05, "loss": 0.01, "step": 27700 }, { "grad_norm": 0.19596350193023682, "learning_rate": 6.0371323199905975e-05, "loss": 0.0086, "step": 27710 }, { "grad_norm": 0.17418746650218964, "learning_rate": 6.03443631616223e-05, "loss": 0.008, "step": 27720 }, { "grad_norm": 0.19962242245674133, "learning_rate": 6.031739998099421e-05, "loss": 0.0082, "step": 27730 }, { "grad_norm": 0.14206264913082123, "learning_rate": 6.029043366621243e-05, "loss": 0.0094, "step": 27740 }, { "grad_norm": 0.18790645897388458, "learning_rate": 6.0263464225468615e-05, "loss": 0.0101, "step": 27750 }, { "grad_norm": 0.1612071841955185, "learning_rate": 6.023649166695534e-05, "loss": 0.009, "step": 27760 }, { "grad_norm": 0.1758749932050705, "learning_rate": 6.0209515998866186e-05, "loss": 0.0112, "step": 27770 }, { "grad_norm": 0.12636691331863403, "learning_rate": 6.018253722939563e-05, "loss": 0.008, "step": 27780 }, { "grad_norm": 0.13885140419006348, "learning_rate": 6.015555536673914e-05, "loss": 0.0097, "step": 27790 }, { "grad_norm": 0.1310536116361618, "learning_rate": 6.0128570419093054e-05, "loss": 0.0099, "step": 27800 }, { "grad_norm": 0.19410450756549835, "learning_rate": 6.010158239465471e-05, "loss": 0.0097, "step": 27810 }, { "grad_norm": 0.16835527122020721, "learning_rate": 6.007459130162235e-05, "loss": 0.0111, "step": 27820 }, { "grad_norm": 0.13291800022125244, "learning_rate": 6.004759714819516e-05, "loss": 0.0088, "step": 27830 }, { "grad_norm": 0.1302768588066101, "learning_rate": 6.002059994257323e-05, "loss": 0.0091, "step": 27840 }, { "grad_norm": 0.19653068482875824, "learning_rate": 5.999359969295764e-05, "loss": 0.0103, "step": 27850 }, { "grad_norm": 0.145920529961586, "learning_rate": 5.9966596407550314e-05, "loss": 0.0102, "step": 27860 }, { "grad_norm": 0.17836420238018036, "learning_rate": 5.993959009455416e-05, "loss": 0.0127, "step": 27870 }, { "grad_norm": 0.18088433146476746, "learning_rate": 5.991258076217298e-05, "loss": 0.0108, "step": 27880 }, { "grad_norm": 0.16877490282058716, "learning_rate": 5.988556841861147e-05, "loss": 0.0129, "step": 27890 }, { "grad_norm": 0.13466811180114746, "learning_rate": 5.985855307207531e-05, "loss": 0.0087, "step": 27900 }, { "grad_norm": 0.2149573266506195, "learning_rate": 5.9831534730771e-05, "loss": 0.0127, "step": 27910 }, { "grad_norm": 0.14185313880443573, "learning_rate": 5.980451340290605e-05, "loss": 0.0089, "step": 27920 }, { "grad_norm": 0.14752046763896942, "learning_rate": 5.97774890966888e-05, "loss": 0.0115, "step": 27930 }, { "grad_norm": 0.15626288950443268, "learning_rate": 5.975046182032851e-05, "loss": 0.0088, "step": 27940 }, { "grad_norm": 0.15622073411941528, "learning_rate": 5.972343158203537e-05, "loss": 0.0125, "step": 27950 }, { "grad_norm": 0.2309648096561432, "learning_rate": 5.969639839002045e-05, "loss": 0.0097, "step": 27960 }, { "grad_norm": 0.13034358620643616, "learning_rate": 5.966936225249572e-05, "loss": 0.0102, "step": 27970 }, { "grad_norm": 0.20221267640590668, "learning_rate": 5.9642323177674044e-05, "loss": 0.0097, "step": 27980 }, { "grad_norm": 0.1703231930732727, "learning_rate": 5.9615281173769154e-05, "loss": 0.0102, "step": 27990 }, { "grad_norm": 0.11277788877487183, "learning_rate": 5.958823624899574e-05, "loss": 0.0095, "step": 28000 }, { "grad_norm": 0.1379043459892273, "learning_rate": 5.956118841156933e-05, "loss": 0.0111, "step": 28010 }, { "grad_norm": 0.17198573052883148, "learning_rate": 5.953413766970631e-05, "loss": 0.0097, "step": 28020 }, { "grad_norm": 0.13428471982479095, "learning_rate": 5.9507084031624e-05, "loss": 0.0101, "step": 28030 }, { "grad_norm": 0.1305200755596161, "learning_rate": 5.948002750554058e-05, "loss": 0.009, "step": 28040 }, { "grad_norm": 0.17269131541252136, "learning_rate": 5.9452968099675124e-05, "loss": 0.012, "step": 28050 }, { "grad_norm": 0.15746262669563293, "learning_rate": 5.9425905822247527e-05, "loss": 0.0085, "step": 28060 }, { "grad_norm": 0.1342504620552063, "learning_rate": 5.939884068147864e-05, "loss": 0.0088, "step": 28070 }, { "grad_norm": 0.13903330266475677, "learning_rate": 5.937177268559011e-05, "loss": 0.0087, "step": 28080 }, { "grad_norm": 0.13885918259620667, "learning_rate": 5.934470184280448e-05, "loss": 0.0072, "step": 28090 }, { "grad_norm": 0.18353670835494995, "learning_rate": 5.931762816134516e-05, "loss": 0.0123, "step": 28100 }, { "grad_norm": 0.1931857019662857, "learning_rate": 5.9290551649436434e-05, "loss": 0.0082, "step": 28110 }, { "grad_norm": 0.15542519092559814, "learning_rate": 5.9263472315303416e-05, "loss": 0.0085, "step": 28120 }, { "grad_norm": 0.14996159076690674, "learning_rate": 5.9236390167172096e-05, "loss": 0.0087, "step": 28130 }, { "grad_norm": 0.13856706023216248, "learning_rate": 5.920930521326932e-05, "loss": 0.0098, "step": 28140 }, { "grad_norm": 0.12953415513038635, "learning_rate": 5.918221746182276e-05, "loss": 0.0075, "step": 28150 }, { "grad_norm": 0.12357208132743835, "learning_rate": 5.9155126921061e-05, "loss": 0.0093, "step": 28160 }, { "grad_norm": 0.17798827588558197, "learning_rate": 5.91280335992134e-05, "loss": 0.0112, "step": 28170 }, { "grad_norm": 0.11415532976388931, "learning_rate": 5.91009375045102e-05, "loss": 0.009, "step": 28180 }, { "grad_norm": 0.16471323370933533, "learning_rate": 5.9073838645182476e-05, "loss": 0.0116, "step": 28190 }, { "grad_norm": 0.18751539289951324, "learning_rate": 5.904673702946217e-05, "loss": 0.0093, "step": 28200 }, { "grad_norm": 0.1372060477733612, "learning_rate": 5.9019632665582004e-05, "loss": 0.009, "step": 28210 }, { "grad_norm": 0.21642136573791504, "learning_rate": 5.899252556177559e-05, "loss": 0.0101, "step": 28220 }, { "grad_norm": 0.15814834833145142, "learning_rate": 5.896541572627735e-05, "loss": 0.009, "step": 28230 }, { "grad_norm": 0.1612922102212906, "learning_rate": 5.893830316732253e-05, "loss": 0.0112, "step": 28240 }, { "grad_norm": 0.22164000570774078, "learning_rate": 5.8911187893147214e-05, "loss": 0.0097, "step": 28250 }, { "grad_norm": 0.1568300575017929, "learning_rate": 5.888406991198828e-05, "loss": 0.0106, "step": 28260 }, { "grad_norm": 0.1349438726902008, "learning_rate": 5.885694923208349e-05, "loss": 0.0079, "step": 28270 }, { "grad_norm": 0.10907037556171417, "learning_rate": 5.882982586167138e-05, "loss": 0.0101, "step": 28280 }, { "grad_norm": 0.1747833490371704, "learning_rate": 5.880269980899131e-05, "loss": 0.0084, "step": 28290 }, { "grad_norm": 0.1506744921207428, "learning_rate": 5.8775571082283465e-05, "loss": 0.0102, "step": 28300 }, { "grad_norm": 0.09556593745946884, "learning_rate": 5.8748439689788824e-05, "loss": 0.0111, "step": 28310 }, { "grad_norm": 0.14382223784923553, "learning_rate": 5.87213056397492e-05, "loss": 0.0106, "step": 28320 }, { "grad_norm": 0.12460514903068542, "learning_rate": 5.869416894040719e-05, "loss": 0.008, "step": 28330 }, { "grad_norm": 0.13985835015773773, "learning_rate": 5.866702960000621e-05, "loss": 0.0083, "step": 28340 }, { "grad_norm": 0.1667737066745758, "learning_rate": 5.863988762679048e-05, "loss": 0.0081, "step": 28350 }, { "grad_norm": 0.165554940700531, "learning_rate": 5.8612743029005e-05, "loss": 0.0079, "step": 28360 }, { "grad_norm": 0.24155211448669434, "learning_rate": 5.858559581489561e-05, "loss": 0.0079, "step": 28370 }, { "grad_norm": 0.1526772677898407, "learning_rate": 5.85584459927089e-05, "loss": 0.0095, "step": 28380 }, { "grad_norm": 0.1966041475534439, "learning_rate": 5.853129357069227e-05, "loss": 0.0104, "step": 28390 }, { "grad_norm": 0.13618722558021545, "learning_rate": 5.8504138557093913e-05, "loss": 0.0081, "step": 28400 }, { "grad_norm": 0.117268405854702, "learning_rate": 5.8476980960162784e-05, "loss": 0.0121, "step": 28410 }, { "grad_norm": 0.18981099128723145, "learning_rate": 5.844982078814868e-05, "loss": 0.0086, "step": 28420 }, { "grad_norm": 0.19096235930919647, "learning_rate": 5.842265804930211e-05, "loss": 0.0098, "step": 28430 }, { "grad_norm": 0.21802517771720886, "learning_rate": 5.839549275187444e-05, "loss": 0.0094, "step": 28440 }, { "grad_norm": 0.2338905781507492, "learning_rate": 5.836832490411771e-05, "loss": 0.0107, "step": 28450 }, { "grad_norm": 0.21222993731498718, "learning_rate": 5.834115451428485e-05, "loss": 0.0091, "step": 28460 }, { "grad_norm": 0.16241928935050964, "learning_rate": 5.831398159062946e-05, "loss": 0.0116, "step": 28470 }, { "grad_norm": 0.16079668700695038, "learning_rate": 5.828680614140599e-05, "loss": 0.0101, "step": 28480 }, { "grad_norm": 0.14661946892738342, "learning_rate": 5.825962817486962e-05, "loss": 0.0093, "step": 28490 }, { "grad_norm": 0.18763498961925507, "learning_rate": 5.823244769927629e-05, "loss": 0.0099, "step": 28500 }, { "grad_norm": 0.12279319018125534, "learning_rate": 5.8205264722882716e-05, "loss": 0.0079, "step": 28510 }, { "grad_norm": 0.12747424840927124, "learning_rate": 5.817807925394636e-05, "loss": 0.0089, "step": 28520 }, { "grad_norm": 0.15497778356075287, "learning_rate": 5.815089130072546e-05, "loss": 0.0106, "step": 28530 }, { "grad_norm": 0.17195966839790344, "learning_rate": 5.8123700871479e-05, "loss": 0.0095, "step": 28540 }, { "grad_norm": 0.16314750909805298, "learning_rate": 5.809650797446671e-05, "loss": 0.0097, "step": 28550 }, { "grad_norm": 0.1193024069070816, "learning_rate": 5.806931261794907e-05, "loss": 0.0102, "step": 28560 }, { "grad_norm": 0.12990956008434296, "learning_rate": 5.804211481018731e-05, "loss": 0.0099, "step": 28570 }, { "grad_norm": 0.09661617875099182, "learning_rate": 5.801491455944341e-05, "loss": 0.009, "step": 28580 }, { "grad_norm": 0.1669945865869522, "learning_rate": 5.79877118739801e-05, "loss": 0.009, "step": 28590 }, { "grad_norm": 0.1159871369600296, "learning_rate": 5.7960506762060816e-05, "loss": 0.0088, "step": 28600 }, { "grad_norm": 0.2761957347393036, "learning_rate": 5.793329923194977e-05, "loss": 0.0089, "step": 28610 }, { "grad_norm": 0.12401419132947922, "learning_rate": 5.790608929191187e-05, "loss": 0.0096, "step": 28620 }, { "grad_norm": 0.13750149309635162, "learning_rate": 5.78788769502128e-05, "loss": 0.0124, "step": 28630 }, { "grad_norm": 0.14558929204940796, "learning_rate": 5.785166221511894e-05, "loss": 0.0082, "step": 28640 }, { "grad_norm": 0.1472577601671219, "learning_rate": 5.7824445094897415e-05, "loss": 0.0099, "step": 28650 }, { "grad_norm": 0.10099607706069946, "learning_rate": 5.7797225597816065e-05, "loss": 0.0075, "step": 28660 }, { "grad_norm": 0.08834551274776459, "learning_rate": 5.777000373214345e-05, "loss": 0.0077, "step": 28670 }, { "grad_norm": 0.15895093977451324, "learning_rate": 5.774277950614885e-05, "loss": 0.0101, "step": 28680 }, { "grad_norm": 0.1706102043390274, "learning_rate": 5.771555292810227e-05, "loss": 0.0102, "step": 28690 }, { "grad_norm": 0.14959591627120972, "learning_rate": 5.768832400627444e-05, "loss": 0.01, "step": 28700 }, { "grad_norm": 0.15032066404819489, "learning_rate": 5.7661092748936775e-05, "loss": 0.0102, "step": 28710 }, { "grad_norm": 0.1807149052619934, "learning_rate": 5.76338591643614e-05, "loss": 0.0079, "step": 28720 }, { "grad_norm": 0.177439346909523, "learning_rate": 5.760662326082118e-05, "loss": 0.0102, "step": 28730 }, { "grad_norm": 0.13104327023029327, "learning_rate": 5.757938504658965e-05, "loss": 0.0086, "step": 28740 }, { "grad_norm": 0.1667434722185135, "learning_rate": 5.755214452994107e-05, "loss": 0.01, "step": 28750 }, { "grad_norm": 0.11513946950435638, "learning_rate": 5.752490171915039e-05, "loss": 0.0085, "step": 28760 }, { "grad_norm": 0.1728157252073288, "learning_rate": 5.749765662249324e-05, "loss": 0.0092, "step": 28770 }, { "grad_norm": 0.12589134275913239, "learning_rate": 5.747040924824596e-05, "loss": 0.0111, "step": 28780 }, { "grad_norm": 0.1356787234544754, "learning_rate": 5.7443159604685613e-05, "loss": 0.0083, "step": 28790 }, { "grad_norm": 0.1250733882188797, "learning_rate": 5.74159077000899e-05, "loss": 0.0086, "step": 28800 }, { "grad_norm": 0.1432030349969864, "learning_rate": 5.7388653542737235e-05, "loss": 0.0091, "step": 28810 }, { "grad_norm": 0.1138804703950882, "learning_rate": 5.736139714090672e-05, "loss": 0.0112, "step": 28820 }, { "grad_norm": 0.08995441347360611, "learning_rate": 5.73341385028781e-05, "loss": 0.0078, "step": 28830 }, { "grad_norm": 0.14702442288398743, "learning_rate": 5.7306877636931855e-05, "loss": 0.0105, "step": 28840 }, { "grad_norm": 0.1182124987244606, "learning_rate": 5.7279614551349125e-05, "loss": 0.009, "step": 28850 }, { "grad_norm": 0.1394113153219223, "learning_rate": 5.725234925441169e-05, "loss": 0.0088, "step": 28860 }, { "grad_norm": 0.1420663297176361, "learning_rate": 5.7225081754402044e-05, "loss": 0.0089, "step": 28870 }, { "grad_norm": 0.13879817724227905, "learning_rate": 5.7197812059603326e-05, "loss": 0.0093, "step": 28880 }, { "grad_norm": 0.09610181301832199, "learning_rate": 5.717054017829934e-05, "loss": 0.0083, "step": 28890 }, { "grad_norm": 0.1528598517179489, "learning_rate": 5.7143266118774584e-05, "loss": 0.0097, "step": 28900 }, { "grad_norm": 0.19391806423664093, "learning_rate": 5.711598988931418e-05, "loss": 0.0086, "step": 28910 }, { "grad_norm": 0.1360250860452652, "learning_rate": 5.7088711498203954e-05, "loss": 0.0067, "step": 28920 }, { "grad_norm": 0.13648079335689545, "learning_rate": 5.706143095373033e-05, "loss": 0.0104, "step": 28930 }, { "grad_norm": 0.08695212006568909, "learning_rate": 5.703414826418042e-05, "loss": 0.0073, "step": 28940 }, { "grad_norm": 0.13381962478160858, "learning_rate": 5.7006863437842007e-05, "loss": 0.0086, "step": 28950 }, { "grad_norm": 0.14613103866577148, "learning_rate": 5.697957648300348e-05, "loss": 0.0114, "step": 28960 }, { "grad_norm": 0.15583699941635132, "learning_rate": 5.695228740795391e-05, "loss": 0.0091, "step": 28970 }, { "grad_norm": 0.14314420521259308, "learning_rate": 5.6924996220982985e-05, "loss": 0.0099, "step": 28980 }, { "grad_norm": 0.1852773278951645, "learning_rate": 5.6897702930381045e-05, "loss": 0.0105, "step": 28990 }, { "grad_norm": 0.16114766895771027, "learning_rate": 5.687040754443908e-05, "loss": 0.0092, "step": 29000 }, { "grad_norm": 0.1211831197142601, "learning_rate": 5.6843110071448725e-05, "loss": 0.0091, "step": 29010 }, { "grad_norm": 0.23434337973594666, "learning_rate": 5.6815810519702194e-05, "loss": 0.0094, "step": 29020 }, { "grad_norm": 0.1530357152223587, "learning_rate": 5.6788508897492396e-05, "loss": 0.0083, "step": 29030 }, { "grad_norm": 0.14513076841831207, "learning_rate": 5.676120521311282e-05, "loss": 0.0131, "step": 29040 }, { "grad_norm": 0.11756022274494171, "learning_rate": 5.6733899474857634e-05, "loss": 0.0064, "step": 29050 }, { "grad_norm": 0.09825494885444641, "learning_rate": 5.670659169102157e-05, "loss": 0.0082, "step": 29060 }, { "grad_norm": 0.17105188965797424, "learning_rate": 5.6679281869900044e-05, "loss": 0.01, "step": 29070 }, { "grad_norm": 0.1558821201324463, "learning_rate": 5.6651970019789045e-05, "loss": 0.0087, "step": 29080 }, { "grad_norm": 0.11952544003725052, "learning_rate": 5.662465614898519e-05, "loss": 0.0094, "step": 29090 }, { "grad_norm": 0.21019327640533447, "learning_rate": 5.6597340265785695e-05, "loss": 0.0098, "step": 29100 }, { "grad_norm": 0.17267361283302307, "learning_rate": 5.657002237848843e-05, "loss": 0.0081, "step": 29110 }, { "grad_norm": 0.17585085332393646, "learning_rate": 5.654270249539183e-05, "loss": 0.0094, "step": 29120 }, { "grad_norm": 0.11256231367588043, "learning_rate": 5.651538062479498e-05, "loss": 0.0101, "step": 29130 }, { "grad_norm": 0.15657632052898407, "learning_rate": 5.648805677499751e-05, "loss": 0.0114, "step": 29140 }, { "grad_norm": 0.18555797636508942, "learning_rate": 5.646073095429969e-05, "loss": 0.0101, "step": 29150 }, { "grad_norm": 0.18918566405773163, "learning_rate": 5.643340317100241e-05, "loss": 0.0105, "step": 29160 }, { "grad_norm": 0.17497022449970245, "learning_rate": 5.64060734334071e-05, "loss": 0.0096, "step": 29170 }, { "grad_norm": 0.1441657692193985, "learning_rate": 5.637874174981583e-05, "loss": 0.0091, "step": 29180 }, { "grad_norm": 0.14733131229877472, "learning_rate": 5.635140812853124e-05, "loss": 0.0094, "step": 29190 }, { "grad_norm": 0.1494186818599701, "learning_rate": 5.6324072577856544e-05, "loss": 0.008, "step": 29200 }, { "grad_norm": 0.14981749653816223, "learning_rate": 5.629673510609559e-05, "loss": 0.0081, "step": 29210 }, { "grad_norm": 0.17620575428009033, "learning_rate": 5.626939572155276e-05, "loss": 0.0099, "step": 29220 }, { "grad_norm": 0.2219356745481491, "learning_rate": 5.6242054432533054e-05, "loss": 0.0101, "step": 29230 }, { "grad_norm": 0.15584969520568848, "learning_rate": 5.621471124734201e-05, "loss": 0.0118, "step": 29240 }, { "grad_norm": 0.16988924145698547, "learning_rate": 5.6187366174285794e-05, "loss": 0.0102, "step": 29250 }, { "grad_norm": 0.12163927406072617, "learning_rate": 5.616001922167109e-05, "loss": 0.0077, "step": 29260 }, { "grad_norm": 0.08522210270166397, "learning_rate": 5.61326703978052e-05, "loss": 0.0078, "step": 29270 }, { "grad_norm": 0.15230847895145416, "learning_rate": 5.6105319710995964e-05, "loss": 0.008, "step": 29280 }, { "grad_norm": 0.18996210396289825, "learning_rate": 5.60779671695518e-05, "loss": 0.0102, "step": 29290 }, { "grad_norm": 0.1587129533290863, "learning_rate": 5.6050612781781684e-05, "loss": 0.0107, "step": 29300 }, { "grad_norm": 0.1640380620956421, "learning_rate": 5.602325655599516e-05, "loss": 0.0084, "step": 29310 }, { "grad_norm": 0.16209305822849274, "learning_rate": 5.599589850050234e-05, "loss": 0.0094, "step": 29320 }, { "grad_norm": 0.14265407621860504, "learning_rate": 5.5968538623613874e-05, "loss": 0.0104, "step": 29330 }, { "grad_norm": 0.10305136442184448, "learning_rate": 5.594117693364095e-05, "loss": 0.0077, "step": 29340 }, { "grad_norm": 0.2949110269546509, "learning_rate": 5.591381343889535e-05, "loss": 0.0092, "step": 29350 }, { "grad_norm": 0.16062259674072266, "learning_rate": 5.5886448147689355e-05, "loss": 0.0101, "step": 29360 }, { "grad_norm": 0.13613101840019226, "learning_rate": 5.585908106833585e-05, "loss": 0.0092, "step": 29370 }, { "grad_norm": 0.1726338118314743, "learning_rate": 5.5831712209148226e-05, "loss": 0.0087, "step": 29380 }, { "grad_norm": 0.15846596658229828, "learning_rate": 5.58043415784404e-05, "loss": 0.0087, "step": 29390 }, { "grad_norm": 0.15655212104320526, "learning_rate": 5.577696918452686e-05, "loss": 0.0092, "step": 29400 }, { "grad_norm": 0.1499212086200714, "learning_rate": 5.5749595035722604e-05, "loss": 0.008, "step": 29410 }, { "grad_norm": 0.14675629138946533, "learning_rate": 5.5722219140343193e-05, "loss": 0.0077, "step": 29420 }, { "grad_norm": 0.12646639347076416, "learning_rate": 5.56948415067047e-05, "loss": 0.0091, "step": 29430 }, { "grad_norm": 0.11616289615631104, "learning_rate": 5.5667462143123704e-05, "loss": 0.0091, "step": 29440 }, { "grad_norm": 0.14012645184993744, "learning_rate": 5.564008105791737e-05, "loss": 0.0085, "step": 29450 }, { "grad_norm": 0.1307409107685089, "learning_rate": 5.5612698259403316e-05, "loss": 0.0096, "step": 29460 }, { "grad_norm": 0.1627029925584793, "learning_rate": 5.5585313755899724e-05, "loss": 0.0077, "step": 29470 }, { "grad_norm": 0.1538277268409729, "learning_rate": 5.5557927555725285e-05, "loss": 0.0096, "step": 29480 }, { "grad_norm": 0.10988488793373108, "learning_rate": 5.55305396671992e-05, "loss": 0.0077, "step": 29490 }, { "grad_norm": 0.16877970099449158, "learning_rate": 5.55031500986412e-05, "loss": 0.0094, "step": 29500 }, { "grad_norm": 0.0885067880153656, "learning_rate": 5.547575885837149e-05, "loss": 0.0098, "step": 29510 }, { "grad_norm": 0.16271597146987915, "learning_rate": 5.5448365954710825e-05, "loss": 0.0092, "step": 29520 }, { "grad_norm": 0.18359683454036713, "learning_rate": 5.5420971395980446e-05, "loss": 0.0084, "step": 29530 }, { "grad_norm": 0.16879217326641083, "learning_rate": 5.539357519050209e-05, "loss": 0.0118, "step": 29540 }, { "grad_norm": 0.13348248600959778, "learning_rate": 5.536617734659799e-05, "loss": 0.0084, "step": 29550 }, { "grad_norm": 0.22321072220802307, "learning_rate": 5.533877787259091e-05, "loss": 0.0103, "step": 29560 }, { "grad_norm": 0.13665379583835602, "learning_rate": 5.5311376776804044e-05, "loss": 0.0086, "step": 29570 }, { "grad_norm": 0.12562698125839233, "learning_rate": 5.528397406756118e-05, "loss": 0.0086, "step": 29580 }, { "grad_norm": 0.1583940088748932, "learning_rate": 5.525656975318652e-05, "loss": 0.0083, "step": 29590 }, { "grad_norm": 0.19829806685447693, "learning_rate": 5.522916384200474e-05, "loss": 0.0104, "step": 29600 }, { "grad_norm": 0.17261961102485657, "learning_rate": 5.520175634234106e-05, "loss": 0.0112, "step": 29610 }, { "grad_norm": 0.14479856193065643, "learning_rate": 5.517434726252113e-05, "loss": 0.01, "step": 29620 }, { "grad_norm": 0.13560862839221954, "learning_rate": 5.514693661087113e-05, "loss": 0.0079, "step": 29630 }, { "grad_norm": 0.16518263518810272, "learning_rate": 5.511952439571769e-05, "loss": 0.0087, "step": 29640 }, { "grad_norm": 0.09760603308677673, "learning_rate": 5.509211062538791e-05, "loss": 0.0086, "step": 29650 }, { "grad_norm": 0.15030759572982788, "learning_rate": 5.506469530820939e-05, "loss": 0.0083, "step": 29660 }, { "grad_norm": 0.1291397660970688, "learning_rate": 5.503727845251014e-05, "loss": 0.0077, "step": 29670 }, { "grad_norm": 0.17114421725273132, "learning_rate": 5.50098600666187e-05, "loss": 0.0072, "step": 29680 }, { "grad_norm": 0.1383432149887085, "learning_rate": 5.498244015886406e-05, "loss": 0.0089, "step": 29690 }, { "grad_norm": 0.15665224194526672, "learning_rate": 5.495501873757565e-05, "loss": 0.0098, "step": 29700 }, { "grad_norm": 0.16983947157859802, "learning_rate": 5.492759581108336e-05, "loss": 0.009, "step": 29710 }, { "grad_norm": 0.12586601078510284, "learning_rate": 5.490017138771759e-05, "loss": 0.0088, "step": 29720 }, { "grad_norm": 0.13476479053497314, "learning_rate": 5.487274547580912e-05, "loss": 0.0119, "step": 29730 }, { "grad_norm": 0.1290171593427658, "learning_rate": 5.484531808368923e-05, "loss": 0.0075, "step": 29740 }, { "grad_norm": 0.17045141756534576, "learning_rate": 5.4817889219689656e-05, "loss": 0.0099, "step": 29750 }, { "grad_norm": 0.11157448589801788, "learning_rate": 5.4790458892142536e-05, "loss": 0.0089, "step": 29760 }, { "grad_norm": 0.14087089896202087, "learning_rate": 5.476302710938048e-05, "loss": 0.0111, "step": 29770 }, { "grad_norm": 0.1558552086353302, "learning_rate": 5.473559387973657e-05, "loss": 0.0096, "step": 29780 }, { "grad_norm": 0.1464914083480835, "learning_rate": 5.470815921154425e-05, "loss": 0.0098, "step": 29790 }, { "grad_norm": 0.1705850511789322, "learning_rate": 5.468072311313749e-05, "loss": 0.0125, "step": 29800 }, { "grad_norm": 0.16910332441329956, "learning_rate": 5.465328559285063e-05, "loss": 0.0071, "step": 29810 }, { "grad_norm": 0.13436482846736908, "learning_rate": 5.462584665901849e-05, "loss": 0.0088, "step": 29820 }, { "grad_norm": 0.17028506100177765, "learning_rate": 5.4598406319976235e-05, "loss": 0.0104, "step": 29830 }, { "grad_norm": 0.13407348096370697, "learning_rate": 5.457096458405958e-05, "loss": 0.0091, "step": 29840 }, { "grad_norm": 0.14533597230911255, "learning_rate": 5.454352145960457e-05, "loss": 0.0101, "step": 29850 }, { "grad_norm": 0.14092817902565002, "learning_rate": 5.4516076954947715e-05, "loss": 0.0113, "step": 29860 }, { "grad_norm": 0.15093886852264404, "learning_rate": 5.448863107842591e-05, "loss": 0.0097, "step": 29870 }, { "grad_norm": 0.15621499717235565, "learning_rate": 5.446118383837651e-05, "loss": 0.0117, "step": 29880 }, { "grad_norm": 0.15075983107089996, "learning_rate": 5.443373524313722e-05, "loss": 0.0112, "step": 29890 }, { "grad_norm": 0.2137628197669983, "learning_rate": 5.440628530104626e-05, "loss": 0.0089, "step": 29900 }, { "grad_norm": 0.1340322196483612, "learning_rate": 5.4378834020442146e-05, "loss": 0.0109, "step": 29910 }, { "grad_norm": 0.0994606763124466, "learning_rate": 5.4351381409663884e-05, "loss": 0.0086, "step": 29920 }, { "grad_norm": 0.151324063539505, "learning_rate": 5.432392747705084e-05, "loss": 0.0112, "step": 29930 }, { "grad_norm": 0.09338458627462387, "learning_rate": 5.429647223094278e-05, "loss": 0.0099, "step": 29940 }, { "grad_norm": 0.11794301867485046, "learning_rate": 5.4269015679679924e-05, "loss": 0.0089, "step": 29950 }, { "grad_norm": 0.11790595203638077, "learning_rate": 5.424155783160281e-05, "loss": 0.0082, "step": 29960 }, { "grad_norm": 0.13061057031154633, "learning_rate": 5.4214098695052415e-05, "loss": 0.0078, "step": 29970 }, { "grad_norm": 0.10929935425519943, "learning_rate": 5.418663827837012e-05, "loss": 0.0071, "step": 29980 }, { "grad_norm": 0.16549895703792572, "learning_rate": 5.415917658989763e-05, "loss": 0.0076, "step": 29990 }, { "grad_norm": 0.10336269438266754, "learning_rate": 5.413171363797713e-05, "loss": 0.0092, "step": 30000 }, { "grad_norm": 0.1355941742658615, "learning_rate": 5.4104249430951116e-05, "loss": 0.0079, "step": 30010 }, { "grad_norm": 0.13503240048885345, "learning_rate": 5.4076783977162494e-05, "loss": 0.0125, "step": 30020 }, { "grad_norm": 0.13805261254310608, "learning_rate": 5.4049317284954525e-05, "loss": 0.0091, "step": 30030 }, { "grad_norm": 0.13187868893146515, "learning_rate": 5.4021849362670884e-05, "loss": 0.0084, "step": 30040 }, { "grad_norm": 0.09412156790494919, "learning_rate": 5.3994380218655604e-05, "loss": 0.0104, "step": 30050 }, { "grad_norm": 0.14690960943698883, "learning_rate": 5.396690986125309e-05, "loss": 0.0103, "step": 30060 }, { "grad_norm": 0.17469796538352966, "learning_rate": 5.3939438298808075e-05, "loss": 0.0084, "step": 30070 }, { "grad_norm": 0.12309994548559189, "learning_rate": 5.3911965539665744e-05, "loss": 0.0078, "step": 30080 }, { "grad_norm": 0.14869773387908936, "learning_rate": 5.388449159217156e-05, "loss": 0.0109, "step": 30090 }, { "grad_norm": 0.1776612251996994, "learning_rate": 5.3857016464671385e-05, "loss": 0.0095, "step": 30100 }, { "grad_norm": 0.16429154574871063, "learning_rate": 5.382954016551146e-05, "loss": 0.0106, "step": 30110 }, { "grad_norm": 0.14833566546440125, "learning_rate": 5.380206270303835e-05, "loss": 0.0093, "step": 30120 }, { "grad_norm": 0.15002252161502838, "learning_rate": 5.377458408559897e-05, "loss": 0.0089, "step": 30130 }, { "grad_norm": 0.16130809485912323, "learning_rate": 5.374710432154061e-05, "loss": 0.0125, "step": 30140 }, { "grad_norm": 0.17991903424263, "learning_rate": 5.3719623419210886e-05, "loss": 0.0097, "step": 30150 }, { "grad_norm": 0.1781352013349533, "learning_rate": 5.3692141386957786e-05, "loss": 0.0096, "step": 30160 }, { "grad_norm": 0.16062434017658234, "learning_rate": 5.3664658233129616e-05, "loss": 0.0102, "step": 30170 }, { "grad_norm": 0.13923218846321106, "learning_rate": 5.363717396607504e-05, "loss": 0.0087, "step": 30180 }, { "grad_norm": 0.1527981162071228, "learning_rate": 5.360968859414305e-05, "loss": 0.0085, "step": 30190 }, { "grad_norm": 0.14214833080768585, "learning_rate": 5.358220212568295e-05, "loss": 0.0108, "step": 30200 }, { "grad_norm": 0.14322465658187866, "learning_rate": 5.355471456904444e-05, "loss": 0.0099, "step": 30210 }, { "grad_norm": 0.15302196145057678, "learning_rate": 5.3527225932577495e-05, "loss": 0.0107, "step": 30220 }, { "grad_norm": 0.13275688886642456, "learning_rate": 5.349973622463246e-05, "loss": 0.0083, "step": 30230 }, { "grad_norm": 0.119534432888031, "learning_rate": 5.3472245453559956e-05, "loss": 0.0091, "step": 30240 }, { "grad_norm": 0.18476727604866028, "learning_rate": 5.3444753627710955e-05, "loss": 0.0122, "step": 30250 }, { "grad_norm": 0.1977226734161377, "learning_rate": 5.341726075543676e-05, "loss": 0.009, "step": 30260 }, { "grad_norm": 0.20465780794620514, "learning_rate": 5.338976684508898e-05, "loss": 0.0102, "step": 30270 }, { "grad_norm": 0.19133885204792023, "learning_rate": 5.336227190501953e-05, "loss": 0.0113, "step": 30280 }, { "grad_norm": 0.1287716031074524, "learning_rate": 5.3334775943580664e-05, "loss": 0.0099, "step": 30290 }, { "grad_norm": 0.17929761111736298, "learning_rate": 5.330727896912491e-05, "loss": 0.0083, "step": 30300 }, { "grad_norm": 0.12163344770669937, "learning_rate": 5.327978099000511e-05, "loss": 0.0078, "step": 30310 }, { "grad_norm": 0.14043210446834564, "learning_rate": 5.3252282014574465e-05, "loss": 0.0107, "step": 30320 }, { "grad_norm": 0.09496947377920151, "learning_rate": 5.322478205118641e-05, "loss": 0.0083, "step": 30330 }, { "grad_norm": 0.15527188777923584, "learning_rate": 5.3197281108194704e-05, "loss": 0.0101, "step": 30340 }, { "grad_norm": 0.16352921724319458, "learning_rate": 5.316977919395342e-05, "loss": 0.0086, "step": 30350 }, { "grad_norm": 0.18152935802936554, "learning_rate": 5.314227631681691e-05, "loss": 0.0102, "step": 30360 }, { "grad_norm": 0.15311649441719055, "learning_rate": 5.311477248513982e-05, "loss": 0.01, "step": 30370 }, { "grad_norm": 0.16320796310901642, "learning_rate": 5.30872677072771e-05, "loss": 0.0081, "step": 30380 }, { "grad_norm": 0.17895640432834625, "learning_rate": 5.3059761991583954e-05, "loss": 0.0092, "step": 30390 }, { "grad_norm": 0.12542809545993805, "learning_rate": 5.303225534641592e-05, "loss": 0.0071, "step": 30400 }, { "grad_norm": 0.12036345154047012, "learning_rate": 5.300474778012875e-05, "loss": 0.0091, "step": 30410 }, { "grad_norm": 0.11302000284194946, "learning_rate": 5.297723930107855e-05, "loss": 0.0091, "step": 30420 }, { "grad_norm": 0.11045386642217636, "learning_rate": 5.294972991762167e-05, "loss": 0.0056, "step": 30430 }, { "grad_norm": 0.11105021089315414, "learning_rate": 5.292221963811472e-05, "loss": 0.0064, "step": 30440 }, { "grad_norm": 0.15129269659519196, "learning_rate": 5.28947084709146e-05, "loss": 0.0071, "step": 30450 }, { "grad_norm": 0.10995443910360336, "learning_rate": 5.2867196424378465e-05, "loss": 0.0066, "step": 30460 }, { "grad_norm": 0.17592564225196838, "learning_rate": 5.2839683506863765e-05, "loss": 0.0081, "step": 30470 }, { "grad_norm": 0.09984711557626724, "learning_rate": 5.281216972672821e-05, "loss": 0.0071, "step": 30480 }, { "grad_norm": 0.16995196044445038, "learning_rate": 5.278465509232973e-05, "loss": 0.0085, "step": 30490 }, { "grad_norm": 0.1815129518508911, "learning_rate": 5.275713961202655e-05, "loss": 0.0089, "step": 30500 }, { "grad_norm": 0.14397267997264862, "learning_rate": 5.2729623294177165e-05, "loss": 0.0083, "step": 30510 }, { "grad_norm": 0.12290389090776443, "learning_rate": 5.270210614714028e-05, "loss": 0.0089, "step": 30520 }, { "grad_norm": 0.11927606165409088, "learning_rate": 5.267458817927491e-05, "loss": 0.011, "step": 30530 }, { "grad_norm": 0.1250564455986023, "learning_rate": 5.264706939894026e-05, "loss": 0.009, "step": 30540 }, { "grad_norm": 0.1596885621547699, "learning_rate": 5.261954981449584e-05, "loss": 0.0092, "step": 30550 }, { "grad_norm": 0.1380947083234787, "learning_rate": 5.2592029434301324e-05, "loss": 0.0116, "step": 30560 }, { "grad_norm": 0.12301519513130188, "learning_rate": 5.256450826671672e-05, "loss": 0.0124, "step": 30570 }, { "grad_norm": 0.14696165919303894, "learning_rate": 5.253698632010221e-05, "loss": 0.0101, "step": 30580 }, { "grad_norm": 0.15963053703308105, "learning_rate": 5.2509463602818246e-05, "loss": 0.0082, "step": 30590 }, { "grad_norm": 0.1393347531557083, "learning_rate": 5.248194012322549e-05, "loss": 0.0083, "step": 30600 }, { "grad_norm": 0.12066706269979477, "learning_rate": 5.245441588968486e-05, "loss": 0.0097, "step": 30610 }, { "grad_norm": 0.11982277780771255, "learning_rate": 5.242689091055748e-05, "loss": 0.0069, "step": 30620 }, { "grad_norm": 0.12444431334733963, "learning_rate": 5.239936519420473e-05, "loss": 0.0096, "step": 30630 }, { "grad_norm": 0.1089920923113823, "learning_rate": 5.2371838748988175e-05, "loss": 0.0086, "step": 30640 }, { "grad_norm": 0.16336967051029205, "learning_rate": 5.234431158326965e-05, "loss": 0.0077, "step": 30650 }, { "grad_norm": 0.1636052280664444, "learning_rate": 5.231678370541115e-05, "loss": 0.0098, "step": 30660 }, { "grad_norm": 0.11604250967502594, "learning_rate": 5.228925512377495e-05, "loss": 0.0079, "step": 30670 }, { "grad_norm": 0.14752201735973358, "learning_rate": 5.2261725846723465e-05, "loss": 0.0083, "step": 30680 }, { "grad_norm": 0.1391010880470276, "learning_rate": 5.22341958826194e-05, "loss": 0.0095, "step": 30690 }, { "grad_norm": 0.150456503033638, "learning_rate": 5.22066652398256e-05, "loss": 0.0088, "step": 30700 }, { "grad_norm": 0.13101176917552948, "learning_rate": 5.2179133926705185e-05, "loss": 0.009, "step": 30710 }, { "grad_norm": 0.12121345102787018, "learning_rate": 5.215160195162141e-05, "loss": 0.01, "step": 30720 }, { "grad_norm": 0.14741013944149017, "learning_rate": 5.212406932293776e-05, "loss": 0.0096, "step": 30730 }, { "grad_norm": 0.1682111620903015, "learning_rate": 5.209653604901795e-05, "loss": 0.0097, "step": 30740 }, { "grad_norm": 0.09808894246816635, "learning_rate": 5.206900213822584e-05, "loss": 0.0068, "step": 30750 }, { "grad_norm": 0.12602174282073975, "learning_rate": 5.204146759892551e-05, "loss": 0.0081, "step": 30760 }, { "grad_norm": 0.15925189852714539, "learning_rate": 5.2013932439481216e-05, "loss": 0.0095, "step": 30770 }, { "grad_norm": 0.156436488032341, "learning_rate": 5.198639666825743e-05, "loss": 0.0109, "step": 30780 }, { "grad_norm": 0.2125389575958252, "learning_rate": 5.195886029361877e-05, "loss": 0.0073, "step": 30790 }, { "grad_norm": 0.13378499448299408, "learning_rate": 5.193132332393009e-05, "loss": 0.0102, "step": 30800 }, { "grad_norm": 0.15587392449378967, "learning_rate": 5.1903785767556376e-05, "loss": 0.0091, "step": 30810 }, { "grad_norm": 0.17695119976997375, "learning_rate": 5.187624763286282e-05, "loss": 0.0092, "step": 30820 }, { "grad_norm": 0.1928357034921646, "learning_rate": 5.184870892821475e-05, "loss": 0.0096, "step": 30830 }, { "grad_norm": 0.1298125982284546, "learning_rate": 5.182116966197773e-05, "loss": 0.0075, "step": 30840 }, { "grad_norm": 0.19459466636180878, "learning_rate": 5.1793629842517466e-05, "loss": 0.0082, "step": 30850 }, { "grad_norm": 0.12080803513526917, "learning_rate": 5.17660894781998e-05, "loss": 0.0102, "step": 30860 }, { "grad_norm": 0.17154094576835632, "learning_rate": 5.173854857739079e-05, "loss": 0.0115, "step": 30870 }, { "grad_norm": 0.12156889587640762, "learning_rate": 5.171100714845661e-05, "loss": 0.0099, "step": 30880 }, { "grad_norm": 0.1327996402978897, "learning_rate": 5.1683465199763646e-05, "loss": 0.0081, "step": 30890 }, { "grad_norm": 0.1315133273601532, "learning_rate": 5.16559227396784e-05, "loss": 0.0092, "step": 30900 }, { "grad_norm": 0.15760394930839539, "learning_rate": 5.1628379776567556e-05, "loss": 0.0097, "step": 30910 }, { "grad_norm": 0.12377595156431198, "learning_rate": 5.160083631879792e-05, "loss": 0.0092, "step": 30920 }, { "grad_norm": 0.09465891867876053, "learning_rate": 5.1573292374736484e-05, "loss": 0.0085, "step": 30930 }, { "grad_norm": 0.10498922318220139, "learning_rate": 5.1545747952750356e-05, "loss": 0.0077, "step": 30940 }, { "grad_norm": 0.12158244848251343, "learning_rate": 5.151820306120682e-05, "loss": 0.008, "step": 30950 }, { "grad_norm": 0.12887521088123322, "learning_rate": 5.149065770847328e-05, "loss": 0.0087, "step": 30960 }, { "grad_norm": 0.13322246074676514, "learning_rate": 5.1463111902917297e-05, "loss": 0.0083, "step": 30970 }, { "grad_norm": 0.12286029756069183, "learning_rate": 5.143556565290654e-05, "loss": 0.0078, "step": 30980 }, { "grad_norm": 0.11585300415754318, "learning_rate": 5.140801896680882e-05, "loss": 0.0105, "step": 30990 }, { "grad_norm": 0.11426711082458496, "learning_rate": 5.1380471852992144e-05, "loss": 0.008, "step": 31000 }, { "grad_norm": 0.1441556066274643, "learning_rate": 5.135292431982457e-05, "loss": 0.0091, "step": 31010 }, { "grad_norm": 0.15070781111717224, "learning_rate": 5.1325376375674294e-05, "loss": 0.0093, "step": 31020 }, { "grad_norm": 0.11201663315296173, "learning_rate": 5.129782802890968e-05, "loss": 0.0079, "step": 31030 }, { "grad_norm": 0.15370844304561615, "learning_rate": 5.127027928789916e-05, "loss": 0.011, "step": 31040 }, { "grad_norm": 0.11308003962039948, "learning_rate": 5.124273016101135e-05, "loss": 0.0082, "step": 31050 }, { "grad_norm": 0.13749033212661743, "learning_rate": 5.121518065661492e-05, "loss": 0.0077, "step": 31060 }, { "grad_norm": 0.10332407057285309, "learning_rate": 5.11876307830787e-05, "loss": 0.009, "step": 31070 }, { "grad_norm": 0.10987216979265213, "learning_rate": 5.1160080548771596e-05, "loss": 0.0106, "step": 31080 }, { "grad_norm": 0.12232471257448196, "learning_rate": 5.1132529962062656e-05, "loss": 0.0077, "step": 31090 }, { "grad_norm": 0.16002243757247925, "learning_rate": 5.110497903132101e-05, "loss": 0.0093, "step": 31100 }, { "grad_norm": 0.16475671529769897, "learning_rate": 5.107742776491592e-05, "loss": 0.0088, "step": 31110 }, { "grad_norm": 0.16094276309013367, "learning_rate": 5.104987617121673e-05, "loss": 0.0095, "step": 31120 }, { "grad_norm": 0.14461717009544373, "learning_rate": 5.102232425859287e-05, "loss": 0.0085, "step": 31130 }, { "grad_norm": 0.11230067908763885, "learning_rate": 5.09947720354139e-05, "loss": 0.0083, "step": 31140 }, { "grad_norm": 0.15558016300201416, "learning_rate": 5.096721951004942e-05, "loss": 0.0084, "step": 31150 }, { "grad_norm": 0.1929575502872467, "learning_rate": 5.0939666690869227e-05, "loss": 0.0094, "step": 31160 }, { "grad_norm": 0.12198149412870407, "learning_rate": 5.0912113586243096e-05, "loss": 0.0086, "step": 31170 }, { "grad_norm": 0.13168774545192719, "learning_rate": 5.0884560204540935e-05, "loss": 0.0091, "step": 31180 }, { "grad_norm": 0.1456492692232132, "learning_rate": 5.0857006554132736e-05, "loss": 0.0079, "step": 31190 }, { "grad_norm": 0.13445141911506653, "learning_rate": 5.0829452643388575e-05, "loss": 0.0079, "step": 31200 }, { "grad_norm": 0.12142378836870193, "learning_rate": 5.08018984806786e-05, "loss": 0.0093, "step": 31210 }, { "grad_norm": 0.14097630977630615, "learning_rate": 5.0774344074373036e-05, "loss": 0.0095, "step": 31220 }, { "grad_norm": 0.14456595480442047, "learning_rate": 5.07467894328422e-05, "loss": 0.009, "step": 31230 }, { "grad_norm": 0.12963595986366272, "learning_rate": 5.0719234564456454e-05, "loss": 0.0107, "step": 31240 }, { "grad_norm": 0.1360125094652176, "learning_rate": 5.0691679477586216e-05, "loss": 0.008, "step": 31250 }, { "grad_norm": 0.1229148805141449, "learning_rate": 5.0664124180602035e-05, "loss": 0.0075, "step": 31260 }, { "grad_norm": 0.11107556521892548, "learning_rate": 5.063656868187447e-05, "loss": 0.0086, "step": 31270 }, { "grad_norm": 0.11655775457620621, "learning_rate": 5.060901298977413e-05, "loss": 0.0111, "step": 31280 }, { "grad_norm": 0.16421949863433838, "learning_rate": 5.0581457112671725e-05, "loss": 0.0095, "step": 31290 }, { "grad_norm": 0.19433259963989258, "learning_rate": 5.0553901058938016e-05, "loss": 0.0115, "step": 31300 }, { "grad_norm": 0.12029293179512024, "learning_rate": 5.052634483694377e-05, "loss": 0.007, "step": 31310 }, { "grad_norm": 0.15750651061534882, "learning_rate": 5.049878845505988e-05, "loss": 0.0083, "step": 31320 }, { "grad_norm": 0.15562117099761963, "learning_rate": 5.047123192165721e-05, "loss": 0.009, "step": 31330 }, { "grad_norm": 0.21003100275993347, "learning_rate": 5.0443675245106735e-05, "loss": 0.009, "step": 31340 }, { "grad_norm": 0.15292595326900482, "learning_rate": 5.0416118433779426e-05, "loss": 0.0096, "step": 31350 }, { "grad_norm": 0.12458266317844391, "learning_rate": 5.038856149604633e-05, "loss": 0.01, "step": 31360 }, { "grad_norm": 0.1384914517402649, "learning_rate": 5.03610044402785e-05, "loss": 0.0098, "step": 31370 }, { "grad_norm": 0.1216164380311966, "learning_rate": 5.033344727484707e-05, "loss": 0.0095, "step": 31380 }, { "grad_norm": 0.14786569774150848, "learning_rate": 5.030589000812315e-05, "loss": 0.0126, "step": 31390 }, { "grad_norm": 0.1457679718732834, "learning_rate": 5.027833264847793e-05, "loss": 0.0089, "step": 31400 }, { "grad_norm": 0.13066412508487701, "learning_rate": 5.025077520428258e-05, "loss": 0.0098, "step": 31410 }, { "grad_norm": 0.16426227986812592, "learning_rate": 5.022321768390837e-05, "loss": 0.0111, "step": 31420 }, { "grad_norm": 0.11766323447227478, "learning_rate": 5.0195660095726516e-05, "loss": 0.0079, "step": 31430 }, { "grad_norm": 0.11077671498060226, "learning_rate": 5.016810244810829e-05, "loss": 0.0088, "step": 31440 }, { "grad_norm": 0.4936646819114685, "learning_rate": 5.0140544749424976e-05, "loss": 0.0083, "step": 31450 }, { "grad_norm": 0.16693612933158875, "learning_rate": 5.0112987008047874e-05, "loss": 0.0101, "step": 31460 }, { "grad_norm": 0.11414411664009094, "learning_rate": 5.008542923234831e-05, "loss": 0.0084, "step": 31470 }, { "grad_norm": 0.11563949286937714, "learning_rate": 5.00578714306976e-05, "loss": 0.0084, "step": 31480 }, { "grad_norm": 0.1739262044429779, "learning_rate": 5.0030313611467084e-05, "loss": 0.0093, "step": 31490 }, { "grad_norm": 0.16373059153556824, "learning_rate": 5.0002755783028074e-05, "loss": 0.0095, "step": 31500 }, { "grad_norm": 0.10895588994026184, "learning_rate": 4.997519795375194e-05, "loss": 0.0072, "step": 31510 }, { "grad_norm": 0.11419501155614853, "learning_rate": 4.9947640132010016e-05, "loss": 0.0072, "step": 31520 }, { "grad_norm": 0.15012823045253754, "learning_rate": 4.9920082326173625e-05, "loss": 0.0077, "step": 31530 }, { "grad_norm": 0.1237712949514389, "learning_rate": 4.9892524544614114e-05, "loss": 0.0102, "step": 31540 }, { "grad_norm": 0.11448884010314941, "learning_rate": 4.986496679570283e-05, "loss": 0.0124, "step": 31550 }, { "grad_norm": 0.13758747279644012, "learning_rate": 4.983740908781105e-05, "loss": 0.0078, "step": 31560 }, { "grad_norm": 0.15551336109638214, "learning_rate": 4.9809851429310116e-05, "loss": 0.0071, "step": 31570 }, { "grad_norm": 0.17234541475772858, "learning_rate": 4.9782293828571275e-05, "loss": 0.0072, "step": 31580 }, { "grad_norm": 0.16729746758937836, "learning_rate": 4.9754736293965846e-05, "loss": 0.0073, "step": 31590 }, { "grad_norm": 0.1466628611087799, "learning_rate": 4.972717883386502e-05, "loss": 0.0088, "step": 31600 }, { "grad_norm": 0.13040341436862946, "learning_rate": 4.9699621456640075e-05, "loss": 0.0074, "step": 31610 }, { "grad_norm": 0.17846138775348663, "learning_rate": 4.9672064170662214e-05, "loss": 0.0116, "step": 31620 }, { "grad_norm": 0.15124651789665222, "learning_rate": 4.9644506984302583e-05, "loss": 0.0069, "step": 31630 }, { "grad_norm": 0.1697724312543869, "learning_rate": 4.9616949905932356e-05, "loss": 0.0108, "step": 31640 }, { "grad_norm": 0.16581417620182037, "learning_rate": 4.9589392943922615e-05, "loss": 0.0093, "step": 31650 }, { "grad_norm": 0.12122616171836853, "learning_rate": 4.956183610664447e-05, "loss": 0.0095, "step": 31660 }, { "grad_norm": 0.18258793652057648, "learning_rate": 4.9534279402468945e-05, "loss": 0.0103, "step": 31670 }, { "grad_norm": 0.11156520247459412, "learning_rate": 4.9506722839767036e-05, "loss": 0.009, "step": 31680 }, { "grad_norm": 0.1451719105243683, "learning_rate": 4.947916642690972e-05, "loss": 0.0073, "step": 31690 }, { "grad_norm": 0.15209558606147766, "learning_rate": 4.9451610172267874e-05, "loss": 0.0096, "step": 31700 }, { "grad_norm": 0.165350079536438, "learning_rate": 4.9424054084212376e-05, "loss": 0.0075, "step": 31710 }, { "grad_norm": 0.14584851264953613, "learning_rate": 4.939649817111407e-05, "loss": 0.0072, "step": 31720 }, { "grad_norm": 0.1802072674036026, "learning_rate": 4.936894244134365e-05, "loss": 0.0092, "step": 31730 }, { "grad_norm": 0.12484602630138397, "learning_rate": 4.9341386903271886e-05, "loss": 0.0097, "step": 31740 }, { "grad_norm": 0.23751679062843323, "learning_rate": 4.931383156526936e-05, "loss": 0.0103, "step": 31750 }, { "grad_norm": 0.11639759689569473, "learning_rate": 4.92862764357067e-05, "loss": 0.0091, "step": 31760 }, { "grad_norm": 0.1960509717464447, "learning_rate": 4.925872152295443e-05, "loss": 0.0103, "step": 31770 }, { "grad_norm": 0.10358352959156036, "learning_rate": 4.923116683538296e-05, "loss": 0.0085, "step": 31780 }, { "grad_norm": 0.18661317229270935, "learning_rate": 4.920361238136273e-05, "loss": 0.0093, "step": 31790 }, { "grad_norm": 0.13916148245334625, "learning_rate": 4.9176058169264014e-05, "loss": 0.009, "step": 31800 }, { "grad_norm": 0.10826461762189865, "learning_rate": 4.9148504207457074e-05, "loss": 0.0092, "step": 31810 }, { "grad_norm": 0.12830103933811188, "learning_rate": 4.912095050431208e-05, "loss": 0.008, "step": 31820 }, { "grad_norm": 0.1203383356332779, "learning_rate": 4.909339706819911e-05, "loss": 0.0091, "step": 31830 }, { "grad_norm": 0.1504511535167694, "learning_rate": 4.906584390748819e-05, "loss": 0.008, "step": 31840 }, { "grad_norm": 0.13724319636821747, "learning_rate": 4.9038291030549195e-05, "loss": 0.0085, "step": 31850 }, { "grad_norm": 0.13427422940731049, "learning_rate": 4.9010738445751995e-05, "loss": 0.0083, "step": 31860 }, { "grad_norm": 0.18421593308448792, "learning_rate": 4.8983186161466364e-05, "loss": 0.0084, "step": 31870 }, { "grad_norm": 0.12770594656467438, "learning_rate": 4.89556341860619e-05, "loss": 0.0082, "step": 31880 }, { "grad_norm": 0.09195824712514877, "learning_rate": 4.892808252790822e-05, "loss": 0.0069, "step": 31890 }, { "grad_norm": 0.1445271521806717, "learning_rate": 4.890053119537475e-05, "loss": 0.0078, "step": 31900 }, { "grad_norm": 0.11959756910800934, "learning_rate": 4.887298019683087e-05, "loss": 0.0091, "step": 31910 }, { "grad_norm": 0.142107754945755, "learning_rate": 4.884542954064587e-05, "loss": 0.0078, "step": 31920 }, { "grad_norm": 0.15202714502811432, "learning_rate": 4.881787923518887e-05, "loss": 0.0074, "step": 31930 }, { "grad_norm": 0.13579359650611877, "learning_rate": 4.879032928882896e-05, "loss": 0.01, "step": 31940 }, { "grad_norm": 0.19527456164360046, "learning_rate": 4.876277970993505e-05, "loss": 0.0078, "step": 31950 }, { "grad_norm": 0.18692053854465485, "learning_rate": 4.873523050687602e-05, "loss": 0.0113, "step": 31960 }, { "grad_norm": 0.15965811908245087, "learning_rate": 4.870768168802056e-05, "loss": 0.008, "step": 31970 }, { "grad_norm": 0.14408844709396362, "learning_rate": 4.868013326173728e-05, "loss": 0.0098, "step": 31980 }, { "grad_norm": 0.12757398188114166, "learning_rate": 4.865258523639468e-05, "loss": 0.0068, "step": 31990 }, { "grad_norm": 0.16718854010105133, "learning_rate": 4.862503762036109e-05, "loss": 0.0078, "step": 32000 }, { "grad_norm": 0.20566560328006744, "learning_rate": 4.859749042200478e-05, "loss": 0.0109, "step": 32010 }, { "grad_norm": 0.13022272288799286, "learning_rate": 4.856994364969384e-05, "loss": 0.0085, "step": 32020 }, { "grad_norm": 0.1364019215106964, "learning_rate": 4.854239731179625e-05, "loss": 0.0082, "step": 32030 }, { "grad_norm": 0.14658920466899872, "learning_rate": 4.85148514166799e-05, "loss": 0.0071, "step": 32040 }, { "grad_norm": 0.11317313462495804, "learning_rate": 4.8487305972712456e-05, "loss": 0.0069, "step": 32050 }, { "grad_norm": 0.12312735617160797, "learning_rate": 4.8459760988261526e-05, "loss": 0.0085, "step": 32060 }, { "grad_norm": 0.11399148404598236, "learning_rate": 4.843221647169453e-05, "loss": 0.0067, "step": 32070 }, { "grad_norm": 0.14482147991657257, "learning_rate": 4.840467243137878e-05, "loss": 0.0093, "step": 32080 }, { "grad_norm": 0.14841710031032562, "learning_rate": 4.837712887568143e-05, "loss": 0.0094, "step": 32090 }, { "grad_norm": 0.11084672063589096, "learning_rate": 4.8349585812969464e-05, "loss": 0.0069, "step": 32100 }, { "grad_norm": 0.10244743525981903, "learning_rate": 4.8322043251609775e-05, "loss": 0.0071, "step": 32110 }, { "grad_norm": 0.12586361169815063, "learning_rate": 4.8294501199969015e-05, "loss": 0.0079, "step": 32120 }, { "grad_norm": 0.16646693646907806, "learning_rate": 4.826695966641376e-05, "loss": 0.0081, "step": 32130 }, { "grad_norm": 0.10850372165441513, "learning_rate": 4.823941865931043e-05, "loss": 0.0082, "step": 32140 }, { "grad_norm": 0.09271184355020523, "learning_rate": 4.82118781870252e-05, "loss": 0.0089, "step": 32150 }, { "grad_norm": 0.14088532328605652, "learning_rate": 4.8184338257924185e-05, "loss": 0.0071, "step": 32160 }, { "grad_norm": 0.20075276494026184, "learning_rate": 4.815679888037324e-05, "loss": 0.0076, "step": 32170 }, { "grad_norm": 0.14180243015289307, "learning_rate": 4.8129260062738135e-05, "loss": 0.0085, "step": 32180 }, { "grad_norm": 0.13171835243701935, "learning_rate": 4.810172181338445e-05, "loss": 0.0062, "step": 32190 }, { "grad_norm": 0.12141643464565277, "learning_rate": 4.807418414067753e-05, "loss": 0.007, "step": 32200 }, { "grad_norm": 0.14662732183933258, "learning_rate": 4.804664705298264e-05, "loss": 0.0117, "step": 32210 }, { "grad_norm": 0.12022975832223892, "learning_rate": 4.80191105586648e-05, "loss": 0.0082, "step": 32220 }, { "grad_norm": 0.18553917109966278, "learning_rate": 4.799157466608886e-05, "loss": 0.0088, "step": 32230 }, { "grad_norm": 0.2063138633966446, "learning_rate": 4.796403938361951e-05, "loss": 0.009, "step": 32240 }, { "grad_norm": 0.1101706326007843, "learning_rate": 4.793650471962123e-05, "loss": 0.0064, "step": 32250 }, { "grad_norm": 0.14278559386730194, "learning_rate": 4.790897068245835e-05, "loss": 0.0095, "step": 32260 }, { "grad_norm": 0.12546329200267792, "learning_rate": 4.7881437280494954e-05, "loss": 0.0082, "step": 32270 }, { "grad_norm": 0.1900077909231186, "learning_rate": 4.7853904522094965e-05, "loss": 0.0098, "step": 32280 }, { "grad_norm": 0.1407705843448639, "learning_rate": 4.782637241562215e-05, "loss": 0.0091, "step": 32290 }, { "grad_norm": 0.11747023463249207, "learning_rate": 4.779884096943997e-05, "loss": 0.0087, "step": 32300 }, { "grad_norm": 0.19713810086250305, "learning_rate": 4.777131019191182e-05, "loss": 0.0073, "step": 32310 }, { "grad_norm": 0.13056251406669617, "learning_rate": 4.774378009140076e-05, "loss": 0.006, "step": 32320 }, { "grad_norm": 0.10717102885246277, "learning_rate": 4.7716250676269735e-05, "loss": 0.008, "step": 32330 }, { "grad_norm": 0.10618135333061218, "learning_rate": 4.7688721954881485e-05, "loss": 0.0075, "step": 32340 }, { "grad_norm": 0.09055940061807632, "learning_rate": 4.7661193935598446e-05, "loss": 0.0079, "step": 32350 }, { "grad_norm": 0.1002425029873848, "learning_rate": 4.763366662678296e-05, "loss": 0.0082, "step": 32360 }, { "grad_norm": 0.15010704100131989, "learning_rate": 4.7606140036797064e-05, "loss": 0.008, "step": 32370 }, { "grad_norm": 0.13866105675697327, "learning_rate": 4.7578614174002614e-05, "loss": 0.0072, "step": 32380 }, { "grad_norm": 0.1689925640821457, "learning_rate": 4.755108904676125e-05, "loss": 0.0106, "step": 32390 }, { "grad_norm": 0.1225447952747345, "learning_rate": 4.752356466343436e-05, "loss": 0.0092, "step": 32400 }, { "grad_norm": 0.14154060184955597, "learning_rate": 4.7496041032383174e-05, "loss": 0.0109, "step": 32410 }, { "grad_norm": 0.15406401455402374, "learning_rate": 4.746851816196858e-05, "loss": 0.0087, "step": 32420 }, { "grad_norm": 0.16673970222473145, "learning_rate": 4.744099606055135e-05, "loss": 0.0075, "step": 32430 }, { "grad_norm": 0.09149692207574844, "learning_rate": 4.741347473649193e-05, "loss": 0.0071, "step": 32440 }, { "grad_norm": 0.15565694868564606, "learning_rate": 4.738595419815058e-05, "loss": 0.0084, "step": 32450 }, { "grad_norm": 0.10588236898183823, "learning_rate": 4.7358434453887365e-05, "loss": 0.0102, "step": 32460 }, { "grad_norm": 0.12180153280496597, "learning_rate": 4.7330915512061976e-05, "loss": 0.0097, "step": 32470 }, { "grad_norm": 0.15840336680412292, "learning_rate": 4.730339738103402e-05, "loss": 0.0089, "step": 32480 }, { "grad_norm": 0.12282919138669968, "learning_rate": 4.727588006916271e-05, "loss": 0.0083, "step": 32490 }, { "grad_norm": 0.13146135210990906, "learning_rate": 4.724836358480711e-05, "loss": 0.0071, "step": 32500 }, { "grad_norm": 0.130620539188385, "learning_rate": 4.722084793632601e-05, "loss": 0.0081, "step": 32510 }, { "grad_norm": 0.15068359673023224, "learning_rate": 4.719333313207792e-05, "loss": 0.0079, "step": 32520 }, { "grad_norm": 0.13779020309448242, "learning_rate": 4.716581918042114e-05, "loss": 0.0074, "step": 32530 }, { "grad_norm": 0.16398966312408447, "learning_rate": 4.7138306089713636e-05, "loss": 0.012, "step": 32540 }, { "grad_norm": 0.16652879118919373, "learning_rate": 4.7110793868313183e-05, "loss": 0.0085, "step": 32550 }, { "grad_norm": 0.10353954136371613, "learning_rate": 4.708328252457729e-05, "loss": 0.0073, "step": 32560 }, { "grad_norm": 0.13440635800361633, "learning_rate": 4.7055772066863135e-05, "loss": 0.0088, "step": 32570 }, { "grad_norm": 0.16336217522621155, "learning_rate": 4.702826250352771e-05, "loss": 0.0099, "step": 32580 }, { "grad_norm": 0.10445163398981094, "learning_rate": 4.7000753842927653e-05, "loss": 0.0099, "step": 32590 }, { "grad_norm": 0.1053512692451477, "learning_rate": 4.6973246093419384e-05, "loss": 0.0112, "step": 32600 }, { "grad_norm": 0.12129240483045578, "learning_rate": 4.694573926335906e-05, "loss": 0.0065, "step": 32610 }, { "grad_norm": 0.14631444215774536, "learning_rate": 4.6918233361102476e-05, "loss": 0.0062, "step": 32620 }, { "grad_norm": 0.1251402497291565, "learning_rate": 4.689072839500525e-05, "loss": 0.0095, "step": 32630 }, { "grad_norm": 0.14422094821929932, "learning_rate": 4.6863224373422635e-05, "loss": 0.0072, "step": 32640 }, { "grad_norm": 0.22137703001499176, "learning_rate": 4.683572130470962e-05, "loss": 0.0065, "step": 32650 }, { "grad_norm": 0.14240820705890656, "learning_rate": 4.680821919722094e-05, "loss": 0.008, "step": 32660 }, { "grad_norm": 0.11389458924531937, "learning_rate": 4.6780718059310975e-05, "loss": 0.0079, "step": 32670 }, { "grad_norm": 0.1291392743587494, "learning_rate": 4.675321789933389e-05, "loss": 0.0102, "step": 32680 }, { "grad_norm": 0.13346102833747864, "learning_rate": 4.6725718725643464e-05, "loss": 0.0082, "step": 32690 }, { "grad_norm": 0.13118436932563782, "learning_rate": 4.669822054659323e-05, "loss": 0.0076, "step": 32700 }, { "grad_norm": 0.14133423566818237, "learning_rate": 4.667072337053644e-05, "loss": 0.0066, "step": 32710 }, { "grad_norm": 0.12170569598674774, "learning_rate": 4.6643227205825965e-05, "loss": 0.0079, "step": 32720 }, { "grad_norm": 0.12059202790260315, "learning_rate": 4.6615732060814454e-05, "loss": 0.0098, "step": 32730 }, { "grad_norm": 0.1366928219795227, "learning_rate": 4.658823794385417e-05, "loss": 0.0059, "step": 32740 }, { "grad_norm": 0.16090616583824158, "learning_rate": 4.6560744863297115e-05, "loss": 0.009, "step": 32750 }, { "grad_norm": 0.15259893238544464, "learning_rate": 4.653325282749498e-05, "loss": 0.0079, "step": 32760 }, { "grad_norm": 0.12005897611379623, "learning_rate": 4.6505761844799075e-05, "loss": 0.0094, "step": 32770 }, { "grad_norm": 0.17838597297668457, "learning_rate": 4.647827192356048e-05, "loss": 0.0085, "step": 32780 }, { "grad_norm": 0.1943388134241104, "learning_rate": 4.645078307212989e-05, "loss": 0.0076, "step": 32790 }, { "grad_norm": 0.09344248473644257, "learning_rate": 4.642329529885768e-05, "loss": 0.0076, "step": 32800 }, { "grad_norm": 0.11692114174365997, "learning_rate": 4.639580861209393e-05, "loss": 0.007, "step": 32810 }, { "grad_norm": 0.1433800756931305, "learning_rate": 4.636832302018835e-05, "loss": 0.0069, "step": 32820 }, { "grad_norm": 0.16124001145362854, "learning_rate": 4.6340838531490365e-05, "loss": 0.0109, "step": 32830 }, { "grad_norm": 0.17380660772323608, "learning_rate": 4.6313355154349e-05, "loss": 0.0072, "step": 32840 }, { "grad_norm": 0.11058055609464645, "learning_rate": 4.6285872897113025e-05, "loss": 0.0075, "step": 32850 }, { "grad_norm": 0.15884427726268768, "learning_rate": 4.625839176813077e-05, "loss": 0.0068, "step": 32860 }, { "grad_norm": 0.08822338283061981, "learning_rate": 4.623091177575031e-05, "loss": 0.007, "step": 32870 }, { "grad_norm": 0.11447353661060333, "learning_rate": 4.620343292831936e-05, "loss": 0.0099, "step": 32880 }, { "grad_norm": 0.07198995351791382, "learning_rate": 4.6175955234185206e-05, "loss": 0.0063, "step": 32890 }, { "grad_norm": 0.12690114974975586, "learning_rate": 4.614847870169492e-05, "loss": 0.0088, "step": 32900 }, { "grad_norm": 0.1282849907875061, "learning_rate": 4.612100333919509e-05, "loss": 0.0061, "step": 32910 }, { "grad_norm": 0.14572015404701233, "learning_rate": 4.609352915503202e-05, "loss": 0.0081, "step": 32920 }, { "grad_norm": 0.1389801800251007, "learning_rate": 4.606605615755166e-05, "loss": 0.0072, "step": 32930 }, { "grad_norm": 0.16309131681919098, "learning_rate": 4.6038584355099576e-05, "loss": 0.0103, "step": 32940 }, { "grad_norm": 0.14228329062461853, "learning_rate": 4.6011113756020964e-05, "loss": 0.0085, "step": 32950 }, { "grad_norm": 0.1393682211637497, "learning_rate": 4.598364436866066e-05, "loss": 0.007, "step": 32960 }, { "grad_norm": 0.1629355251789093, "learning_rate": 4.595617620136316e-05, "loss": 0.0067, "step": 32970 }, { "grad_norm": 0.11775395274162292, "learning_rate": 4.592870926247257e-05, "loss": 0.0078, "step": 32980 }, { "grad_norm": 0.12219362705945969, "learning_rate": 4.5901243560332594e-05, "loss": 0.009, "step": 32990 }, { "grad_norm": 0.08063459396362305, "learning_rate": 4.587377910328662e-05, "loss": 0.0084, "step": 33000 }, { "grad_norm": 0.13642741739749908, "learning_rate": 4.5846315899677586e-05, "loss": 0.0069, "step": 33010 }, { "grad_norm": 0.13871058821678162, "learning_rate": 4.5818853957848114e-05, "loss": 0.0082, "step": 33020 }, { "grad_norm": 0.1168137639760971, "learning_rate": 4.579139328614043e-05, "loss": 0.0104, "step": 33030 }, { "grad_norm": 0.08562269061803818, "learning_rate": 4.576393389289633e-05, "loss": 0.0092, "step": 33040 }, { "grad_norm": 0.14482447504997253, "learning_rate": 4.573647578645728e-05, "loss": 0.0096, "step": 33050 }, { "grad_norm": 0.11319229006767273, "learning_rate": 4.57090189751643e-05, "loss": 0.0084, "step": 33060 }, { "grad_norm": 0.13597126305103302, "learning_rate": 4.568156346735806e-05, "loss": 0.0081, "step": 33070 }, { "grad_norm": 0.10996027290821075, "learning_rate": 4.565410927137882e-05, "loss": 0.0105, "step": 33080 }, { "grad_norm": 0.13970650732517242, "learning_rate": 4.562665639556644e-05, "loss": 0.0106, "step": 33090 }, { "grad_norm": 0.13901926577091217, "learning_rate": 4.559920484826037e-05, "loss": 0.0089, "step": 33100 }, { "grad_norm": 0.14276081323623657, "learning_rate": 4.5571754637799665e-05, "loss": 0.0079, "step": 33110 }, { "grad_norm": 0.16927126049995422, "learning_rate": 4.554430577252298e-05, "loss": 0.0079, "step": 33120 }, { "grad_norm": 0.14765608310699463, "learning_rate": 4.551685826076858e-05, "loss": 0.008, "step": 33130 }, { "grad_norm": 0.28408634662628174, "learning_rate": 4.5489412110874246e-05, "loss": 0.0125, "step": 33140 }, { "grad_norm": 0.1761709600687027, "learning_rate": 4.5461967331177444e-05, "loss": 0.0099, "step": 33150 }, { "grad_norm": 0.11380459368228912, "learning_rate": 4.5434523930015115e-05, "loss": 0.0082, "step": 33160 }, { "grad_norm": 0.1570453643798828, "learning_rate": 4.540708191572388e-05, "loss": 0.009, "step": 33170 }, { "grad_norm": 0.16311872005462646, "learning_rate": 4.537964129663991e-05, "loss": 0.0098, "step": 33180 }, { "grad_norm": 0.1775088757276535, "learning_rate": 4.535220208109889e-05, "loss": 0.0076, "step": 33190 }, { "grad_norm": 0.16654586791992188, "learning_rate": 4.5324764277436194e-05, "loss": 0.008, "step": 33200 }, { "grad_norm": 0.09914485365152359, "learning_rate": 4.529732789398664e-05, "loss": 0.0112, "step": 33210 }, { "grad_norm": 0.09306420385837555, "learning_rate": 4.526989293908472e-05, "loss": 0.0083, "step": 33220 }, { "grad_norm": 0.1365428864955902, "learning_rate": 4.524245942106442e-05, "loss": 0.008, "step": 33230 }, { "grad_norm": 0.12701191008090973, "learning_rate": 4.5215027348259345e-05, "loss": 0.0077, "step": 33240 }, { "grad_norm": 0.0873037576675415, "learning_rate": 4.5187596729002616e-05, "loss": 0.0077, "step": 33250 }, { "grad_norm": 0.1407930999994278, "learning_rate": 4.516016757162693e-05, "loss": 0.0108, "step": 33260 }, { "grad_norm": 0.12454428523778915, "learning_rate": 4.513273988446457e-05, "loss": 0.0085, "step": 33270 }, { "grad_norm": 0.1282428354024887, "learning_rate": 4.5105313675847296e-05, "loss": 0.0085, "step": 33280 }, { "grad_norm": 0.12907466292381287, "learning_rate": 4.5077888954106495e-05, "loss": 0.0093, "step": 33290 }, { "grad_norm": 0.13438816368579865, "learning_rate": 4.505046572757309e-05, "loss": 0.0072, "step": 33300 }, { "grad_norm": 0.1705635040998459, "learning_rate": 4.502304400457749e-05, "loss": 0.008, "step": 33310 }, { "grad_norm": 0.11090387403964996, "learning_rate": 4.499562379344973e-05, "loss": 0.0066, "step": 33320 }, { "grad_norm": 0.18341000378131866, "learning_rate": 4.4968205102519306e-05, "loss": 0.0111, "step": 33330 }, { "grad_norm": 0.1238308772444725, "learning_rate": 4.494078794011532e-05, "loss": 0.0076, "step": 33340 }, { "grad_norm": 0.15207140147686005, "learning_rate": 4.491337231456639e-05, "loss": 0.0078, "step": 33350 }, { "grad_norm": 0.13741303980350494, "learning_rate": 4.4885958234200634e-05, "loss": 0.0073, "step": 33360 }, { "grad_norm": 0.12023897469043732, "learning_rate": 4.485854570734575e-05, "loss": 0.0086, "step": 33370 }, { "grad_norm": 0.13464632630348206, "learning_rate": 4.483113474232891e-05, "loss": 0.008, "step": 33380 }, { "grad_norm": 0.1205194815993309, "learning_rate": 4.480372534747688e-05, "loss": 0.0098, "step": 33390 }, { "grad_norm": 0.1335202157497406, "learning_rate": 4.477631753111588e-05, "loss": 0.0085, "step": 33400 }, { "grad_norm": 0.15186652541160583, "learning_rate": 4.4748911301571686e-05, "loss": 0.009, "step": 33410 }, { "grad_norm": 0.15773442387580872, "learning_rate": 4.472150666716961e-05, "loss": 0.0071, "step": 33420 }, { "grad_norm": 0.10634848475456238, "learning_rate": 4.469410363623442e-05, "loss": 0.0075, "step": 33430 }, { "grad_norm": 0.14765650033950806, "learning_rate": 4.466670221709044e-05, "loss": 0.0119, "step": 33440 }, { "grad_norm": 0.13364309072494507, "learning_rate": 4.463930241806154e-05, "loss": 0.0074, "step": 33450 }, { "grad_norm": 0.11791519075632095, "learning_rate": 4.4611904247471006e-05, "loss": 0.0066, "step": 33460 }, { "grad_norm": 0.11600565910339355, "learning_rate": 4.458450771364171e-05, "loss": 0.0096, "step": 33470 }, { "grad_norm": 0.18439564108848572, "learning_rate": 4.4557112824895965e-05, "loss": 0.0105, "step": 33480 }, { "grad_norm": 0.12334108352661133, "learning_rate": 4.452971958955563e-05, "loss": 0.007, "step": 33490 }, { "grad_norm": 0.10221710801124573, "learning_rate": 4.450232801594208e-05, "loss": 0.0072, "step": 33500 }, { "grad_norm": 0.11935111880302429, "learning_rate": 4.447493811237609e-05, "loss": 0.0108, "step": 33510 }, { "grad_norm": 0.13445092737674713, "learning_rate": 4.444754988717804e-05, "loss": 0.0069, "step": 33520 }, { "grad_norm": 0.14660455286502838, "learning_rate": 4.442016334866771e-05, "loss": 0.0097, "step": 33530 }, { "grad_norm": 0.11704234778881073, "learning_rate": 4.4392778505164445e-05, "loss": 0.0061, "step": 33540 }, { "grad_norm": 0.14008381962776184, "learning_rate": 4.436539536498702e-05, "loss": 0.0082, "step": 33550 }, { "grad_norm": 0.09721247106790543, "learning_rate": 4.433801393645369e-05, "loss": 0.0074, "step": 33560 }, { "grad_norm": 0.16536292433738708, "learning_rate": 4.431063422788226e-05, "loss": 0.0066, "step": 33570 }, { "grad_norm": 0.1312534511089325, "learning_rate": 4.428325624758991e-05, "loss": 0.0082, "step": 33580 }, { "grad_norm": 0.1337660849094391, "learning_rate": 4.4255880003893366e-05, "loss": 0.0078, "step": 33590 }, { "grad_norm": 0.11012689024209976, "learning_rate": 4.422850550510884e-05, "loss": 0.0065, "step": 33600 }, { "grad_norm": 0.14705367386341095, "learning_rate": 4.4201132759551934e-05, "loss": 0.0095, "step": 33610 }, { "grad_norm": 0.09315816313028336, "learning_rate": 4.4173761775537804e-05, "loss": 0.0072, "step": 33620 }, { "grad_norm": 0.13306888937950134, "learning_rate": 4.414639256138099e-05, "loss": 0.0083, "step": 33630 }, { "grad_norm": 0.10817690938711166, "learning_rate": 4.411902512539557e-05, "loss": 0.0075, "step": 33640 }, { "grad_norm": 0.1276145875453949, "learning_rate": 4.4091659475895044e-05, "loss": 0.0092, "step": 33650 }, { "grad_norm": 0.12676368653774261, "learning_rate": 4.406429562119235e-05, "loss": 0.008, "step": 33660 }, { "grad_norm": 0.10508456826210022, "learning_rate": 4.4036933569599945e-05, "loss": 0.008, "step": 33670 }, { "grad_norm": 0.12538066506385803, "learning_rate": 4.400957332942965e-05, "loss": 0.0069, "step": 33680 }, { "grad_norm": 0.18319331109523773, "learning_rate": 4.3982214908992844e-05, "loss": 0.0106, "step": 33690 }, { "grad_norm": 0.10475094616413116, "learning_rate": 4.3954858316600235e-05, "loss": 0.0064, "step": 33700 }, { "grad_norm": 0.12228339910507202, "learning_rate": 4.392750356056205e-05, "loss": 0.0083, "step": 33710 }, { "grad_norm": 0.11369737237691879, "learning_rate": 4.390015064918798e-05, "loss": 0.0085, "step": 33720 }, { "grad_norm": 0.11382140964269638, "learning_rate": 4.387279959078705e-05, "loss": 0.0069, "step": 33730 }, { "grad_norm": 0.11927065253257751, "learning_rate": 4.384545039366786e-05, "loss": 0.0071, "step": 33740 }, { "grad_norm": 0.07724978029727936, "learning_rate": 4.381810306613831e-05, "loss": 0.008, "step": 33750 }, { "grad_norm": 0.12947262823581696, "learning_rate": 4.3790757616505826e-05, "loss": 0.007, "step": 33760 }, { "grad_norm": 0.12312886863946915, "learning_rate": 4.376341405307725e-05, "loss": 0.0075, "step": 33770 }, { "grad_norm": 0.13520151376724243, "learning_rate": 4.37360723841588e-05, "loss": 0.0079, "step": 33780 }, { "grad_norm": 0.15814030170440674, "learning_rate": 4.370873261805619e-05, "loss": 0.0087, "step": 33790 }, { "grad_norm": 0.158869206905365, "learning_rate": 4.368139476307449e-05, "loss": 0.0093, "step": 33800 }, { "grad_norm": 0.1363781839609146, "learning_rate": 4.365405882751822e-05, "loss": 0.008, "step": 33810 }, { "grad_norm": 0.16634191572666168, "learning_rate": 4.3626724819691326e-05, "loss": 0.0063, "step": 33820 }, { "grad_norm": 0.14438165724277496, "learning_rate": 4.359939274789715e-05, "loss": 0.0097, "step": 33830 }, { "grad_norm": 0.11817781627178192, "learning_rate": 4.357206262043848e-05, "loss": 0.0072, "step": 33840 }, { "grad_norm": 0.1524733603000641, "learning_rate": 4.354473444561745e-05, "loss": 0.0088, "step": 33850 }, { "grad_norm": 0.12007039040327072, "learning_rate": 4.3517408231735644e-05, "loss": 0.0056, "step": 33860 }, { "grad_norm": 0.10612867772579193, "learning_rate": 4.3490083987094086e-05, "loss": 0.0065, "step": 33870 }, { "grad_norm": 0.20579460263252258, "learning_rate": 4.34627617199931e-05, "loss": 0.0077, "step": 33880 }, { "grad_norm": 0.19559720158576965, "learning_rate": 4.3435441438732526e-05, "loss": 0.0083, "step": 33890 }, { "grad_norm": 0.1662520170211792, "learning_rate": 4.340812315161149e-05, "loss": 0.0094, "step": 33900 }, { "grad_norm": 0.13170196115970612, "learning_rate": 4.338080686692859e-05, "loss": 0.0058, "step": 33910 }, { "grad_norm": 0.10686212033033371, "learning_rate": 4.3353492592981816e-05, "loss": 0.008, "step": 33920 }, { "grad_norm": 0.12958794832229614, "learning_rate": 4.3326180338068485e-05, "loss": 0.0099, "step": 33930 }, { "grad_norm": 0.09682626277208328, "learning_rate": 4.3298870110485356e-05, "loss": 0.0067, "step": 33940 }, { "grad_norm": 0.1385529488325119, "learning_rate": 4.3271561918528567e-05, "loss": 0.0066, "step": 33950 }, { "grad_norm": 0.12641383707523346, "learning_rate": 4.324425577049359e-05, "loss": 0.0081, "step": 33960 }, { "grad_norm": 0.154867023229599, "learning_rate": 4.321695167467535e-05, "loss": 0.0094, "step": 33970 }, { "grad_norm": 0.19231891632080078, "learning_rate": 4.3189649639368093e-05, "loss": 0.0107, "step": 33980 }, { "grad_norm": 0.18294808268547058, "learning_rate": 4.316234967286547e-05, "loss": 0.0077, "step": 33990 }, { "grad_norm": 0.12869961559772491, "learning_rate": 4.313505178346046e-05, "loss": 0.0094, "step": 34000 }, { "grad_norm": 0.13040682673454285, "learning_rate": 4.3107755979445465e-05, "loss": 0.0067, "step": 34010 }, { "grad_norm": 0.12458106130361557, "learning_rate": 4.308046226911224e-05, "loss": 0.0079, "step": 34020 }, { "grad_norm": 0.1065613329410553, "learning_rate": 4.305317066075185e-05, "loss": 0.0098, "step": 34030 }, { "grad_norm": 0.1305847316980362, "learning_rate": 4.302588116265482e-05, "loss": 0.0081, "step": 34040 }, { "grad_norm": 0.09874769300222397, "learning_rate": 4.299859378311094e-05, "loss": 0.007, "step": 34050 }, { "grad_norm": 0.13599000871181488, "learning_rate": 4.2971308530409424e-05, "loss": 0.0112, "step": 34060 }, { "grad_norm": 0.11861316859722137, "learning_rate": 4.2944025412838765e-05, "loss": 0.0074, "step": 34070 }, { "grad_norm": 0.1254570335149765, "learning_rate": 4.291674443868689e-05, "loss": 0.0074, "step": 34080 }, { "grad_norm": 0.11759398877620697, "learning_rate": 4.288946561624104e-05, "loss": 0.0086, "step": 34090 }, { "grad_norm": 0.10181578248739243, "learning_rate": 4.2862188953787794e-05, "loss": 0.0065, "step": 34100 }, { "grad_norm": 0.144558846950531, "learning_rate": 4.283491445961308e-05, "loss": 0.0101, "step": 34110 }, { "grad_norm": 0.1098392903804779, "learning_rate": 4.2807642142002155e-05, "loss": 0.0092, "step": 34120 }, { "grad_norm": 0.20912599563598633, "learning_rate": 4.278037200923966e-05, "loss": 0.01, "step": 34130 }, { "grad_norm": 0.12511901557445526, "learning_rate": 4.275310406960953e-05, "loss": 0.0069, "step": 34140 }, { "grad_norm": 0.11960165202617645, "learning_rate": 4.272583833139502e-05, "loss": 0.0087, "step": 34150 }, { "grad_norm": 0.16597497463226318, "learning_rate": 4.2698574802878794e-05, "loss": 0.0068, "step": 34160 }, { "grad_norm": 0.13457193970680237, "learning_rate": 4.2671313492342734e-05, "loss": 0.0078, "step": 34170 }, { "grad_norm": 0.13419216871261597, "learning_rate": 4.264405440806813e-05, "loss": 0.0083, "step": 34180 }, { "grad_norm": 0.1332281231880188, "learning_rate": 4.26167975583356e-05, "loss": 0.007, "step": 34190 }, { "grad_norm": 0.16882677376270294, "learning_rate": 4.2589542951425e-05, "loss": 0.0068, "step": 34200 }, { "grad_norm": 0.10466647148132324, "learning_rate": 4.2562290595615615e-05, "loss": 0.0062, "step": 34210 }, { "grad_norm": 0.14152312278747559, "learning_rate": 4.2535040499185946e-05, "loss": 0.0073, "step": 34220 }, { "grad_norm": 0.1539088934659958, "learning_rate": 4.250779267041387e-05, "loss": 0.0078, "step": 34230 }, { "grad_norm": 0.16063842177391052, "learning_rate": 4.248054711757657e-05, "loss": 0.0099, "step": 34240 }, { "grad_norm": 0.10601302981376648, "learning_rate": 4.245330384895052e-05, "loss": 0.0079, "step": 34250 }, { "grad_norm": 0.1582925170660019, "learning_rate": 4.242606287281151e-05, "loss": 0.0094, "step": 34260 }, { "grad_norm": 0.14229658246040344, "learning_rate": 4.2398824197434595e-05, "loss": 0.0087, "step": 34270 }, { "grad_norm": 0.11216456443071365, "learning_rate": 4.23715878310942e-05, "loss": 0.0087, "step": 34280 }, { "grad_norm": 0.20030075311660767, "learning_rate": 4.234435378206402e-05, "loss": 0.0081, "step": 34290 }, { "grad_norm": 0.13434365391731262, "learning_rate": 4.2317122058617006e-05, "loss": 0.0066, "step": 34300 }, { "grad_norm": 0.14401564002037048, "learning_rate": 4.2289892669025485e-05, "loss": 0.0071, "step": 34310 }, { "grad_norm": 0.13933831453323364, "learning_rate": 4.226266562156097e-05, "loss": 0.0099, "step": 34320 }, { "grad_norm": 0.09844581037759781, "learning_rate": 4.223544092449435e-05, "loss": 0.0084, "step": 34330 }, { "grad_norm": 0.11509919911623001, "learning_rate": 4.2208218586095784e-05, "loss": 0.0074, "step": 34340 }, { "grad_norm": 0.1087879091501236, "learning_rate": 4.218099861463466e-05, "loss": 0.0082, "step": 34350 }, { "grad_norm": 0.15493716299533844, "learning_rate": 4.215378101837972e-05, "loss": 0.0081, "step": 34360 }, { "grad_norm": 0.12501280009746552, "learning_rate": 4.2126565805598937e-05, "loss": 0.0071, "step": 34370 }, { "grad_norm": 0.145077183842659, "learning_rate": 4.209935298455957e-05, "loss": 0.0079, "step": 34380 }, { "grad_norm": 0.20251423120498657, "learning_rate": 4.207214256352817e-05, "loss": 0.0119, "step": 34390 }, { "grad_norm": 0.12973961234092712, "learning_rate": 4.2044934550770524e-05, "loss": 0.0064, "step": 34400 }, { "grad_norm": 0.10071495920419693, "learning_rate": 4.201772895455174e-05, "loss": 0.0059, "step": 34410 }, { "grad_norm": 0.11046213656663895, "learning_rate": 4.199052578313613e-05, "loss": 0.0063, "step": 34420 }, { "grad_norm": 0.1259777694940567, "learning_rate": 4.1963325044787294e-05, "loss": 0.0077, "step": 34430 }, { "grad_norm": 0.1010337769985199, "learning_rate": 4.193612674776814e-05, "loss": 0.006, "step": 34440 }, { "grad_norm": 0.17622892558574677, "learning_rate": 4.1908930900340745e-05, "loss": 0.0083, "step": 34450 }, { "grad_norm": 0.148279070854187, "learning_rate": 4.1881737510766536e-05, "loss": 0.0075, "step": 34460 }, { "grad_norm": 0.11851528286933899, "learning_rate": 4.185454658730609e-05, "loss": 0.0072, "step": 34470 }, { "grad_norm": 0.11994662135839462, "learning_rate": 4.1827358138219355e-05, "loss": 0.0084, "step": 34480 }, { "grad_norm": 0.16091382503509521, "learning_rate": 4.1800172171765404e-05, "loss": 0.0072, "step": 34490 }, { "grad_norm": 0.09345506876707077, "learning_rate": 4.177298869620264e-05, "loss": 0.0078, "step": 34500 }, { "grad_norm": 0.1247178241610527, "learning_rate": 4.1745807719788705e-05, "loss": 0.0083, "step": 34510 }, { "grad_norm": 0.14229801297187805, "learning_rate": 4.1718629250780445e-05, "loss": 0.0083, "step": 34520 }, { "grad_norm": 0.10941078513860703, "learning_rate": 4.1691453297433956e-05, "loss": 0.008, "step": 34530 }, { "grad_norm": 0.10526470839977264, "learning_rate": 4.166427986800457e-05, "loss": 0.0069, "step": 34540 }, { "grad_norm": 0.09941883385181427, "learning_rate": 4.163710897074688e-05, "loss": 0.0074, "step": 34550 }, { "grad_norm": 0.11425173282623291, "learning_rate": 4.1609940613914686e-05, "loss": 0.0079, "step": 34560 }, { "grad_norm": 0.157799631357193, "learning_rate": 4.1582774805760996e-05, "loss": 0.0084, "step": 34570 }, { "grad_norm": 0.11055468022823334, "learning_rate": 4.155561155453809e-05, "loss": 0.0095, "step": 34580 }, { "grad_norm": 0.12736450135707855, "learning_rate": 4.15284508684974e-05, "loss": 0.0074, "step": 34590 }, { "grad_norm": 0.1606854349374771, "learning_rate": 4.1501292755889675e-05, "loss": 0.0059, "step": 34600 }, { "grad_norm": 0.09983600676059723, "learning_rate": 4.1474137224964833e-05, "loss": 0.0077, "step": 34610 }, { "grad_norm": 0.13942135870456696, "learning_rate": 4.144698428397197e-05, "loss": 0.0076, "step": 34620 }, { "grad_norm": 0.1316649317741394, "learning_rate": 4.1419833941159466e-05, "loss": 0.009, "step": 34630 }, { "grad_norm": 0.12614111602306366, "learning_rate": 4.1392686204774846e-05, "loss": 0.0092, "step": 34640 }, { "grad_norm": 0.16033563017845154, "learning_rate": 4.13655410830649e-05, "loss": 0.0106, "step": 34650 }, { "grad_norm": 0.10426435619592667, "learning_rate": 4.1338398584275594e-05, "loss": 0.007, "step": 34660 }, { "grad_norm": 0.2229520082473755, "learning_rate": 4.1311258716652104e-05, "loss": 0.0066, "step": 34670 }, { "grad_norm": 0.17303821444511414, "learning_rate": 4.128412148843881e-05, "loss": 0.0089, "step": 34680 }, { "grad_norm": 0.13827525079250336, "learning_rate": 4.125698690787926e-05, "loss": 0.0072, "step": 34690 }, { "grad_norm": 0.11920956522226334, "learning_rate": 4.1229854983216245e-05, "loss": 0.0084, "step": 34700 }, { "grad_norm": 0.10788995772600174, "learning_rate": 4.120272572269175e-05, "loss": 0.0078, "step": 34710 }, { "grad_norm": 0.19850941002368927, "learning_rate": 4.117559913454687e-05, "loss": 0.0074, "step": 34720 }, { "grad_norm": 0.1665463149547577, "learning_rate": 4.114847522702201e-05, "loss": 0.0072, "step": 34730 }, { "grad_norm": 0.1253158003091812, "learning_rate": 4.112135400835664e-05, "loss": 0.0087, "step": 34740 }, { "grad_norm": 0.09035281836986542, "learning_rate": 4.109423548678949e-05, "loss": 0.0081, "step": 34750 }, { "grad_norm": 0.14195893704891205, "learning_rate": 4.106711967055848e-05, "loss": 0.0068, "step": 34760 }, { "grad_norm": 0.2172786295413971, "learning_rate": 4.1040006567900636e-05, "loss": 0.0071, "step": 34770 }, { "grad_norm": 0.09826207906007767, "learning_rate": 4.101289618705224e-05, "loss": 0.0061, "step": 34780 }, { "grad_norm": 0.11914803087711334, "learning_rate": 4.0985788536248675e-05, "loss": 0.0068, "step": 34790 }, { "grad_norm": 0.09826754778623581, "learning_rate": 4.095868362372454e-05, "loss": 0.0066, "step": 34800 }, { "grad_norm": 0.11957152187824249, "learning_rate": 4.0931581457713614e-05, "loss": 0.0091, "step": 34810 }, { "grad_norm": 0.13045302033424377, "learning_rate": 4.09044820464488e-05, "loss": 0.0066, "step": 34820 }, { "grad_norm": 0.09643863141536713, "learning_rate": 4.087738539816219e-05, "loss": 0.0087, "step": 34830 }, { "grad_norm": 0.1631782352924347, "learning_rate": 4.085029152108501e-05, "loss": 0.0068, "step": 34840 }, { "grad_norm": 0.1367868185043335, "learning_rate": 4.0823200423447714e-05, "loss": 0.0063, "step": 34850 }, { "grad_norm": 0.09270889312028885, "learning_rate": 4.079611211347981e-05, "loss": 0.0056, "step": 34860 }, { "grad_norm": 0.12723076343536377, "learning_rate": 4.076902659941002e-05, "loss": 0.0094, "step": 34870 }, { "grad_norm": 0.10432921350002289, "learning_rate": 4.074194388946624e-05, "loss": 0.0068, "step": 34880 }, { "grad_norm": 0.11487031728029251, "learning_rate": 4.071486399187545e-05, "loss": 0.0077, "step": 34890 }, { "grad_norm": 0.13163360953330994, "learning_rate": 4.0687786914863836e-05, "loss": 0.0062, "step": 34900 }, { "grad_norm": 0.1356772631406784, "learning_rate": 4.0660712666656666e-05, "loss": 0.0068, "step": 34910 }, { "grad_norm": 0.10866617411375046, "learning_rate": 4.0633641255478394e-05, "loss": 0.006, "step": 34920 }, { "grad_norm": 0.13603520393371582, "learning_rate": 4.0606572689552624e-05, "loss": 0.0071, "step": 34930 }, { "grad_norm": 0.11029678583145142, "learning_rate": 4.0579506977102036e-05, "loss": 0.0079, "step": 34940 }, { "grad_norm": 0.160651296377182, "learning_rate": 4.055244412634849e-05, "loss": 0.0071, "step": 34950 }, { "grad_norm": 0.16457994282245636, "learning_rate": 4.052538414551298e-05, "loss": 0.0075, "step": 34960 }, { "grad_norm": 0.10955900698900223, "learning_rate": 4.0498327042815596e-05, "loss": 0.007, "step": 34970 }, { "grad_norm": 0.12350951880216599, "learning_rate": 4.047127282647559e-05, "loss": 0.0095, "step": 34980 }, { "grad_norm": 0.2803334891796112, "learning_rate": 4.04442215047113e-05, "loss": 0.0075, "step": 34990 }, { "grad_norm": 0.156516894698143, "learning_rate": 4.041717308574023e-05, "loss": 0.0077, "step": 35000 }, { "grad_norm": 0.14094401895999908, "learning_rate": 4.039012757777893e-05, "loss": 0.0123, "step": 35010 }, { "grad_norm": 0.16907261312007904, "learning_rate": 4.036308498904314e-05, "loss": 0.0077, "step": 35020 }, { "grad_norm": 0.14641693234443665, "learning_rate": 4.033604532774771e-05, "loss": 0.0089, "step": 35030 }, { "grad_norm": 0.12041180580854416, "learning_rate": 4.030900860210652e-05, "loss": 0.0077, "step": 35040 }, { "grad_norm": 0.1278047114610672, "learning_rate": 4.028197482033266e-05, "loss": 0.0064, "step": 35050 }, { "grad_norm": 0.12763164937496185, "learning_rate": 4.0254943990638246e-05, "loss": 0.0065, "step": 35060 }, { "grad_norm": 0.15323811769485474, "learning_rate": 4.022791612123454e-05, "loss": 0.0082, "step": 35070 }, { "grad_norm": 0.10118595510721207, "learning_rate": 4.020089122033192e-05, "loss": 0.0119, "step": 35080 }, { "grad_norm": 0.1902007907629013, "learning_rate": 4.01738692961398e-05, "loss": 0.0089, "step": 35090 }, { "grad_norm": 0.17794427275657654, "learning_rate": 4.014685035686675e-05, "loss": 0.0078, "step": 35100 }, { "grad_norm": 0.15489310026168823, "learning_rate": 4.011983441072039e-05, "loss": 0.0107, "step": 35110 }, { "grad_norm": 0.1613932102918625, "learning_rate": 4.0092821465907485e-05, "loss": 0.0081, "step": 35120 }, { "grad_norm": 0.1051371693611145, "learning_rate": 4.006581153063383e-05, "loss": 0.007, "step": 35130 }, { "grad_norm": 0.1605580747127533, "learning_rate": 4.003880461310432e-05, "loss": 0.0069, "step": 35140 }, { "grad_norm": 0.12068233639001846, "learning_rate": 4.001180072152298e-05, "loss": 0.007, "step": 35150 }, { "grad_norm": 0.1203436404466629, "learning_rate": 3.998479986409285e-05, "loss": 0.0073, "step": 35160 }, { "grad_norm": 0.10654669255018234, "learning_rate": 3.995780204901607e-05, "loss": 0.0066, "step": 35170 }, { "grad_norm": 0.14147965610027313, "learning_rate": 3.993080728449391e-05, "loss": 0.0097, "step": 35180 }, { "grad_norm": 0.11061286926269531, "learning_rate": 3.990381557872661e-05, "loss": 0.0103, "step": 35190 }, { "grad_norm": 0.11108201742172241, "learning_rate": 3.987682693991359e-05, "loss": 0.0103, "step": 35200 }, { "grad_norm": 0.08096151798963547, "learning_rate": 3.9849841376253226e-05, "loss": 0.0072, "step": 35210 }, { "grad_norm": 0.1438518613576889, "learning_rate": 3.982285889594306e-05, "loss": 0.0075, "step": 35220 }, { "grad_norm": 0.13857311010360718, "learning_rate": 3.9795879507179665e-05, "loss": 0.007, "step": 35230 }, { "grad_norm": 0.16567234694957733, "learning_rate": 3.9768903218158634e-05, "loss": 0.008, "step": 35240 }, { "grad_norm": 0.1341148316860199, "learning_rate": 3.974193003707468e-05, "loss": 0.0071, "step": 35250 }, { "grad_norm": 0.11318539083003998, "learning_rate": 3.971495997212152e-05, "loss": 0.0078, "step": 35260 }, { "grad_norm": 0.14059056341648102, "learning_rate": 3.9687993031491985e-05, "loss": 0.0072, "step": 35270 }, { "grad_norm": 0.11348053067922592, "learning_rate": 3.966102922337787e-05, "loss": 0.0055, "step": 35280 }, { "grad_norm": 0.11457528173923492, "learning_rate": 3.963406855597009e-05, "loss": 0.0079, "step": 35290 }, { "grad_norm": 0.1231054738163948, "learning_rate": 3.960711103745861e-05, "loss": 0.0088, "step": 35300 }, { "grad_norm": 0.19759848713874817, "learning_rate": 3.958015667603237e-05, "loss": 0.0082, "step": 35310 }, { "grad_norm": 0.1663767248392105, "learning_rate": 3.955320547987943e-05, "loss": 0.0083, "step": 35320 }, { "grad_norm": 0.12246493995189667, "learning_rate": 3.952625745718681e-05, "loss": 0.0085, "step": 35330 }, { "grad_norm": 0.11206039041280746, "learning_rate": 3.949931261614064e-05, "loss": 0.0088, "step": 35340 }, { "grad_norm": 0.11012540012598038, "learning_rate": 3.947237096492605e-05, "loss": 0.008, "step": 35350 }, { "grad_norm": 0.13101744651794434, "learning_rate": 3.944543251172719e-05, "loss": 0.0066, "step": 35360 }, { "grad_norm": 0.1387193500995636, "learning_rate": 3.941849726472725e-05, "loss": 0.0098, "step": 35370 }, { "grad_norm": 0.1127999871969223, "learning_rate": 3.939156523210846e-05, "loss": 0.0061, "step": 35380 }, { "grad_norm": 0.12621094286441803, "learning_rate": 3.9364636422052046e-05, "loss": 0.0078, "step": 35390 }, { "grad_norm": 0.23415936529636383, "learning_rate": 3.933771084273828e-05, "loss": 0.0084, "step": 35400 }, { "grad_norm": 0.11109693348407745, "learning_rate": 3.931078850234643e-05, "loss": 0.0064, "step": 35410 }, { "grad_norm": 0.1434287279844284, "learning_rate": 3.928386940905483e-05, "loss": 0.0107, "step": 35420 }, { "grad_norm": 0.11402348428964615, "learning_rate": 3.925695357104073e-05, "loss": 0.0084, "step": 35430 }, { "grad_norm": 0.14428526163101196, "learning_rate": 3.923004099648049e-05, "loss": 0.0114, "step": 35440 }, { "grad_norm": 0.17113228142261505, "learning_rate": 3.920313169354944e-05, "loss": 0.0082, "step": 35450 }, { "grad_norm": 0.097653329372406, "learning_rate": 3.9176225670421897e-05, "loss": 0.0089, "step": 35460 }, { "grad_norm": 0.15898698568344116, "learning_rate": 3.9149322935271224e-05, "loss": 0.0081, "step": 35470 }, { "grad_norm": 0.1297447830438614, "learning_rate": 3.9122423496269725e-05, "loss": 0.0078, "step": 35480 }, { "grad_norm": 0.11241266131401062, "learning_rate": 3.909552736158877e-05, "loss": 0.0052, "step": 35490 }, { "grad_norm": 0.11944648623466492, "learning_rate": 3.90686345393987e-05, "loss": 0.0089, "step": 35500 }, { "grad_norm": 0.13407287001609802, "learning_rate": 3.9041745037868816e-05, "loss": 0.0099, "step": 35510 }, { "grad_norm": 0.1265448033809662, "learning_rate": 3.9014858865167465e-05, "loss": 0.0079, "step": 35520 }, { "grad_norm": 0.14959008991718292, "learning_rate": 3.8987976029461935e-05, "loss": 0.0075, "step": 35530 }, { "grad_norm": 0.1370944082736969, "learning_rate": 3.896109653891853e-05, "loss": 0.0082, "step": 35540 }, { "grad_norm": 0.14825837314128876, "learning_rate": 3.893422040170254e-05, "loss": 0.0069, "step": 35550 }, { "grad_norm": 0.08543507009744644, "learning_rate": 3.8907347625978207e-05, "loss": 0.0077, "step": 35560 }, { "grad_norm": 0.1257164031267166, "learning_rate": 3.88804782199088e-05, "loss": 0.0078, "step": 35570 }, { "grad_norm": 0.14669454097747803, "learning_rate": 3.8853612191656495e-05, "loss": 0.0067, "step": 35580 }, { "grad_norm": 0.25058043003082275, "learning_rate": 3.88267495493825e-05, "loss": 0.0076, "step": 35590 }, { "grad_norm": 0.19463807344436646, "learning_rate": 3.8799890301247004e-05, "loss": 0.0074, "step": 35600 }, { "grad_norm": 0.12228431552648544, "learning_rate": 3.8773034455409096e-05, "loss": 0.008, "step": 35610 }, { "grad_norm": 0.15154018998146057, "learning_rate": 3.8746182020026904e-05, "loss": 0.0091, "step": 35620 }, { "grad_norm": 0.1360769420862198, "learning_rate": 3.871933300325745e-05, "loss": 0.0099, "step": 35630 }, { "grad_norm": 0.1791931390762329, "learning_rate": 3.869248741325679e-05, "loss": 0.0104, "step": 35640 }, { "grad_norm": 0.15201504528522491, "learning_rate": 3.866564525817992e-05, "loss": 0.007, "step": 35650 }, { "grad_norm": 0.07836383581161499, "learning_rate": 3.8638806546180725e-05, "loss": 0.0077, "step": 35660 }, { "grad_norm": 0.1151650920510292, "learning_rate": 3.861197128541213e-05, "loss": 0.0069, "step": 35670 }, { "grad_norm": 0.12348178029060364, "learning_rate": 3.858513948402599e-05, "loss": 0.0079, "step": 35680 }, { "grad_norm": 0.16399936378002167, "learning_rate": 3.8558311150173077e-05, "loss": 0.0092, "step": 35690 }, { "grad_norm": 0.12401127815246582, "learning_rate": 3.853148629200312e-05, "loss": 0.008, "step": 35700 }, { "grad_norm": 0.11274367570877075, "learning_rate": 3.850466491766482e-05, "loss": 0.0071, "step": 35710 }, { "grad_norm": 0.11869209259748459, "learning_rate": 3.847784703530583e-05, "loss": 0.0101, "step": 35720 }, { "grad_norm": 0.12218038737773895, "learning_rate": 3.845103265307266e-05, "loss": 0.0091, "step": 35730 }, { "grad_norm": 0.12479884922504425, "learning_rate": 3.842422177911086e-05, "loss": 0.0072, "step": 35740 }, { "grad_norm": 0.11848518252372742, "learning_rate": 3.8397414421564826e-05, "loss": 0.0073, "step": 35750 }, { "grad_norm": 0.13410590589046478, "learning_rate": 3.8370610588577935e-05, "loss": 0.0071, "step": 35760 }, { "grad_norm": 0.1470874696969986, "learning_rate": 3.834381028829251e-05, "loss": 0.0083, "step": 35770 }, { "grad_norm": 0.1347421407699585, "learning_rate": 3.8317013528849745e-05, "loss": 0.0057, "step": 35780 }, { "grad_norm": 0.13560247421264648, "learning_rate": 3.8290220318389815e-05, "loss": 0.0061, "step": 35790 }, { "grad_norm": 0.14469033479690552, "learning_rate": 3.8263430665051746e-05, "loss": 0.0078, "step": 35800 }, { "grad_norm": 0.2041354477405548, "learning_rate": 3.8236644576973554e-05, "loss": 0.0088, "step": 35810 }, { "grad_norm": 0.1791732758283615, "learning_rate": 3.820986206229217e-05, "loss": 0.0069, "step": 35820 }, { "grad_norm": 0.12100465595722198, "learning_rate": 3.8183083129143384e-05, "loss": 0.0068, "step": 35830 }, { "grad_norm": 0.11279579252004623, "learning_rate": 3.815630778566193e-05, "loss": 0.0076, "step": 35840 }, { "grad_norm": 0.1012563407421112, "learning_rate": 3.812953603998145e-05, "loss": 0.0073, "step": 35850 }, { "grad_norm": 0.09293825924396515, "learning_rate": 3.8102767900234504e-05, "loss": 0.0057, "step": 35860 }, { "grad_norm": 0.11769795417785645, "learning_rate": 3.807600337455256e-05, "loss": 0.0099, "step": 35870 }, { "grad_norm": 0.1731603443622589, "learning_rate": 3.804924247106593e-05, "loss": 0.0059, "step": 35880 }, { "grad_norm": 0.1345290094614029, "learning_rate": 3.8022485197903925e-05, "loss": 0.0062, "step": 35890 }, { "grad_norm": 0.10879077017307281, "learning_rate": 3.799573156319464e-05, "loss": 0.006, "step": 35900 }, { "grad_norm": 0.1283910721540451, "learning_rate": 3.796898157506515e-05, "loss": 0.0094, "step": 35910 }, { "grad_norm": 0.1819138526916504, "learning_rate": 3.794223524164143e-05, "loss": 0.0061, "step": 35920 }, { "grad_norm": 0.10958180576562881, "learning_rate": 3.7915492571048245e-05, "loss": 0.0068, "step": 35930 }, { "grad_norm": 0.09319625794887543, "learning_rate": 3.788875357140937e-05, "loss": 0.0062, "step": 35940 }, { "grad_norm": 0.16355441510677338, "learning_rate": 3.786201825084736e-05, "loss": 0.0075, "step": 35950 }, { "grad_norm": 0.10152479261159897, "learning_rate": 3.783528661748372e-05, "loss": 0.0064, "step": 35960 }, { "grad_norm": 0.1663065403699875, "learning_rate": 3.780855867943882e-05, "loss": 0.008, "step": 35970 }, { "grad_norm": 0.12261281907558441, "learning_rate": 3.778183444483189e-05, "loss": 0.0082, "step": 35980 }, { "grad_norm": 0.22608354687690735, "learning_rate": 3.775511392178108e-05, "loss": 0.0086, "step": 35990 }, { "grad_norm": 0.11343361437320709, "learning_rate": 3.772839711840332e-05, "loss": 0.0085, "step": 36000 }, { "grad_norm": 0.12466990947723389, "learning_rate": 3.7701684042814515e-05, "loss": 0.0071, "step": 36010 }, { "grad_norm": 0.1639922708272934, "learning_rate": 3.76749747031294e-05, "loss": 0.0066, "step": 36020 }, { "grad_norm": 0.10039724409580231, "learning_rate": 3.764826910746152e-05, "loss": 0.0097, "step": 36030 }, { "grad_norm": 0.11328635364770889, "learning_rate": 3.762156726392338e-05, "loss": 0.0076, "step": 36040 }, { "grad_norm": 0.14457908272743225, "learning_rate": 3.759486918062625e-05, "loss": 0.0069, "step": 36050 }, { "grad_norm": 0.09576483815908432, "learning_rate": 3.756817486568033e-05, "loss": 0.008, "step": 36060 }, { "grad_norm": 0.13656000792980194, "learning_rate": 3.7541484327194654e-05, "loss": 0.0069, "step": 36070 }, { "grad_norm": 0.13985979557037354, "learning_rate": 3.751479757327707e-05, "loss": 0.0078, "step": 36080 }, { "grad_norm": 0.18548311293125153, "learning_rate": 3.7488114612034345e-05, "loss": 0.0083, "step": 36090 }, { "grad_norm": 0.1345152109861374, "learning_rate": 3.7461435451572044e-05, "loss": 0.0104, "step": 36100 }, { "grad_norm": 0.10714475065469742, "learning_rate": 3.743476009999459e-05, "loss": 0.008, "step": 36110 }, { "grad_norm": 0.10640059411525726, "learning_rate": 3.7408088565405245e-05, "loss": 0.0071, "step": 36120 }, { "grad_norm": 0.14819355309009552, "learning_rate": 3.738142085590612e-05, "loss": 0.0071, "step": 36130 }, { "grad_norm": 0.10004594177007675, "learning_rate": 3.7354756979598194e-05, "loss": 0.006, "step": 36140 }, { "grad_norm": 0.16215449571609497, "learning_rate": 3.7328096944581187e-05, "loss": 0.0076, "step": 36150 }, { "grad_norm": 0.09871073812246323, "learning_rate": 3.730144075895377e-05, "loss": 0.0062, "step": 36160 }, { "grad_norm": 0.08158784359693527, "learning_rate": 3.727478843081335e-05, "loss": 0.0091, "step": 36170 }, { "grad_norm": 0.10629922151565552, "learning_rate": 3.72481399682562e-05, "loss": 0.0068, "step": 36180 }, { "grad_norm": 0.13596868515014648, "learning_rate": 3.722149537937747e-05, "loss": 0.0074, "step": 36190 }, { "grad_norm": 0.1231282502412796, "learning_rate": 3.7194854672271015e-05, "loss": 0.0052, "step": 36200 }, { "grad_norm": 0.09762272238731384, "learning_rate": 3.7168217855029644e-05, "loss": 0.0059, "step": 36210 }, { "grad_norm": 0.10628896206617355, "learning_rate": 3.7141584935744856e-05, "loss": 0.0054, "step": 36220 }, { "grad_norm": 0.11206149309873581, "learning_rate": 3.7114955922507055e-05, "loss": 0.0097, "step": 36230 }, { "grad_norm": 0.1867494285106659, "learning_rate": 3.708833082340545e-05, "loss": 0.0077, "step": 36240 }, { "grad_norm": 0.08440199494361877, "learning_rate": 3.7061709646528034e-05, "loss": 0.0072, "step": 36250 }, { "grad_norm": 0.1612088680267334, "learning_rate": 3.7035092399961604e-05, "loss": 0.0076, "step": 36260 }, { "grad_norm": 0.08390357345342636, "learning_rate": 3.700847909179177e-05, "loss": 0.0069, "step": 36270 }, { "grad_norm": 0.1260763555765152, "learning_rate": 3.698186973010297e-05, "loss": 0.0083, "step": 36280 }, { "grad_norm": 0.14135706424713135, "learning_rate": 3.695526432297844e-05, "loss": 0.0072, "step": 36290 }, { "grad_norm": 0.13472460210323334, "learning_rate": 3.692866287850017e-05, "loss": 0.0074, "step": 36300 }, { "grad_norm": 0.1516205370426178, "learning_rate": 3.6902065404749006e-05, "loss": 0.0065, "step": 36310 }, { "grad_norm": 0.1463129222393036, "learning_rate": 3.6875471909804516e-05, "loss": 0.0062, "step": 36320 }, { "grad_norm": 0.13805963099002838, "learning_rate": 3.6848882401745135e-05, "loss": 0.0072, "step": 36330 }, { "grad_norm": 0.1506083905696869, "learning_rate": 3.682229688864806e-05, "loss": 0.0083, "step": 36340 }, { "grad_norm": 0.15848015248775482, "learning_rate": 3.6795715378589235e-05, "loss": 0.007, "step": 36350 }, { "grad_norm": 0.08936440199613571, "learning_rate": 3.676913787964345e-05, "loss": 0.0052, "step": 36360 }, { "grad_norm": 0.10827995091676712, "learning_rate": 3.674256439988423e-05, "loss": 0.0068, "step": 36370 }, { "grad_norm": 0.22791720926761627, "learning_rate": 3.6715994947383904e-05, "loss": 0.0079, "step": 36380 }, { "grad_norm": 0.13279999792575836, "learning_rate": 3.668942953021357e-05, "loss": 0.0103, "step": 36390 }, { "grad_norm": 0.10433155298233032, "learning_rate": 3.66628681564431e-05, "loss": 0.0065, "step": 36400 }, { "grad_norm": 0.15845690667629242, "learning_rate": 3.663631083414114e-05, "loss": 0.0063, "step": 36410 }, { "grad_norm": 0.12968966364860535, "learning_rate": 3.660975757137509e-05, "loss": 0.0069, "step": 36420 }, { "grad_norm": 0.1253787875175476, "learning_rate": 3.658320837621114e-05, "loss": 0.0079, "step": 36430 }, { "grad_norm": 0.1350501924753189, "learning_rate": 3.655666325671426e-05, "loss": 0.0068, "step": 36440 }, { "grad_norm": 0.14791668951511383, "learning_rate": 3.65301222209481e-05, "loss": 0.0066, "step": 36450 }, { "grad_norm": 0.16780593991279602, "learning_rate": 3.650358527697519e-05, "loss": 0.0065, "step": 36460 }, { "grad_norm": 0.1569937765598297, "learning_rate": 3.64770524328567e-05, "loss": 0.0072, "step": 36470 }, { "grad_norm": 0.08969135582447052, "learning_rate": 3.645052369665265e-05, "loss": 0.0072, "step": 36480 }, { "grad_norm": 0.16686853766441345, "learning_rate": 3.6423999076421724e-05, "loss": 0.0066, "step": 36490 }, { "grad_norm": 0.117096908390522, "learning_rate": 3.639747858022142e-05, "loss": 0.0064, "step": 36500 }, { "grad_norm": 0.13365019857883453, "learning_rate": 3.637096221610799e-05, "loss": 0.0052, "step": 36510 }, { "grad_norm": 0.07966294139623642, "learning_rate": 3.634444999213638e-05, "loss": 0.0049, "step": 36520 }, { "grad_norm": 0.09822756797075272, "learning_rate": 3.6317941916360296e-05, "loss": 0.0052, "step": 36530 }, { "grad_norm": 0.1518258899450302, "learning_rate": 3.629143799683221e-05, "loss": 0.0076, "step": 36540 }, { "grad_norm": 0.14962010085582733, "learning_rate": 3.626493824160331e-05, "loss": 0.0064, "step": 36550 }, { "grad_norm": 0.1433212161064148, "learning_rate": 3.623844265872352e-05, "loss": 0.0054, "step": 36560 }, { "grad_norm": 0.12320403009653091, "learning_rate": 3.621195125624149e-05, "loss": 0.0058, "step": 36570 }, { "grad_norm": 0.0950976312160492, "learning_rate": 3.618546404220463e-05, "loss": 0.0098, "step": 36580 }, { "grad_norm": 0.11947789043188095, "learning_rate": 3.615898102465903e-05, "loss": 0.0088, "step": 36590 }, { "grad_norm": 0.13936592638492584, "learning_rate": 3.6132502211649544e-05, "loss": 0.0078, "step": 36600 }, { "grad_norm": 0.13247184455394745, "learning_rate": 3.610602761121975e-05, "loss": 0.0079, "step": 36610 }, { "grad_norm": 0.06558853387832642, "learning_rate": 3.6079557231411897e-05, "loss": 0.007, "step": 36620 }, { "grad_norm": 0.1397782415151596, "learning_rate": 3.6053091080267035e-05, "loss": 0.0056, "step": 36630 }, { "grad_norm": 0.1354225128889084, "learning_rate": 3.602662916582483e-05, "loss": 0.0061, "step": 36640 }, { "grad_norm": 0.1587260514497757, "learning_rate": 3.600017149612375e-05, "loss": 0.0095, "step": 36650 }, { "grad_norm": 0.11156457662582397, "learning_rate": 3.5973718079200935e-05, "loss": 0.0053, "step": 36660 }, { "grad_norm": 0.13765788078308105, "learning_rate": 3.5947268923092216e-05, "loss": 0.0069, "step": 36670 }, { "grad_norm": 0.0948808342218399, "learning_rate": 3.592082403583216e-05, "loss": 0.0066, "step": 36680 }, { "grad_norm": 0.11222776770591736, "learning_rate": 3.5894383425454004e-05, "loss": 0.008, "step": 36690 }, { "grad_norm": 0.10375398397445679, "learning_rate": 3.586794709998975e-05, "loss": 0.0066, "step": 36700 }, { "grad_norm": 0.1185312569141388, "learning_rate": 3.584151506747002e-05, "loss": 0.0063, "step": 36710 }, { "grad_norm": 0.11273569613695145, "learning_rate": 3.581508733592418e-05, "loss": 0.0072, "step": 36720 }, { "grad_norm": 0.08732655644416809, "learning_rate": 3.5788663913380297e-05, "loss": 0.006, "step": 36730 }, { "grad_norm": 0.14437878131866455, "learning_rate": 3.576224480786506e-05, "loss": 0.0068, "step": 36740 }, { "grad_norm": 0.09144413471221924, "learning_rate": 3.573583002740393e-05, "loss": 0.0091, "step": 36750 }, { "grad_norm": 0.13064774870872498, "learning_rate": 3.570941958002103e-05, "loss": 0.0096, "step": 36760 }, { "grad_norm": 0.1307470053434372, "learning_rate": 3.568301347373912e-05, "loss": 0.0051, "step": 36770 }, { "grad_norm": 0.12341591715812683, "learning_rate": 3.5656611716579726e-05, "loss": 0.0072, "step": 36780 }, { "grad_norm": 0.1615581065416336, "learning_rate": 3.5630214316562946e-05, "loss": 0.0076, "step": 36790 }, { "grad_norm": 0.14746738970279694, "learning_rate": 3.560382128170766e-05, "loss": 0.009, "step": 36800 }, { "grad_norm": 0.10075698047876358, "learning_rate": 3.5577432620031374e-05, "loss": 0.0063, "step": 36810 }, { "grad_norm": 0.17118196189403534, "learning_rate": 3.5551048339550216e-05, "loss": 0.0082, "step": 36820 }, { "grad_norm": 0.1677771359682083, "learning_rate": 3.55246684482791e-05, "loss": 0.0099, "step": 36830 }, { "grad_norm": 0.12772054970264435, "learning_rate": 3.5498292954231496e-05, "loss": 0.0075, "step": 36840 }, { "grad_norm": 0.126637801527977, "learning_rate": 3.54719218654196e-05, "loss": 0.0077, "step": 36850 }, { "grad_norm": 0.1491449922323227, "learning_rate": 3.544555518985425e-05, "loss": 0.0074, "step": 36860 }, { "grad_norm": 0.14344386756420135, "learning_rate": 3.541919293554494e-05, "loss": 0.0067, "step": 36870 }, { "grad_norm": 0.10394579917192459, "learning_rate": 3.539283511049985e-05, "loss": 0.0072, "step": 36880 }, { "grad_norm": 0.18113799393177032, "learning_rate": 3.5366481722725755e-05, "loss": 0.0061, "step": 36890 }, { "grad_norm": 0.16529962420463562, "learning_rate": 3.534013278022816e-05, "loss": 0.0071, "step": 36900 }, { "grad_norm": 0.1391511857509613, "learning_rate": 3.531378829101113e-05, "loss": 0.0105, "step": 36910 }, { "grad_norm": 0.09078815579414368, "learning_rate": 3.528744826307746e-05, "loss": 0.0061, "step": 36920 }, { "grad_norm": 0.08631620556116104, "learning_rate": 3.5261112704428554e-05, "loss": 0.0061, "step": 36930 }, { "grad_norm": 0.14050248265266418, "learning_rate": 3.523478162306443e-05, "loss": 0.0073, "step": 36940 }, { "grad_norm": 0.14735718071460724, "learning_rate": 3.520845502698381e-05, "loss": 0.0103, "step": 36950 }, { "grad_norm": 0.10980305075645447, "learning_rate": 3.5182132924184005e-05, "loss": 0.0114, "step": 36960 }, { "grad_norm": 0.14128433167934418, "learning_rate": 3.5155815322660966e-05, "loss": 0.0079, "step": 36970 }, { "grad_norm": 0.1586560308933258, "learning_rate": 3.512950223040931e-05, "loss": 0.006, "step": 36980 }, { "grad_norm": 0.13566209375858307, "learning_rate": 3.5103193655422216e-05, "loss": 0.0066, "step": 36990 }, { "grad_norm": 0.1387576162815094, "learning_rate": 3.5076889605691596e-05, "loss": 0.0088, "step": 37000 }, { "grad_norm": 0.1675083041191101, "learning_rate": 3.505059008920787e-05, "loss": 0.0071, "step": 37010 }, { "grad_norm": 0.1081133782863617, "learning_rate": 3.502429511396016e-05, "loss": 0.0066, "step": 37020 }, { "grad_norm": 0.10868696123361588, "learning_rate": 3.4998004687936196e-05, "loss": 0.0068, "step": 37030 }, { "grad_norm": 0.10556480288505554, "learning_rate": 3.497171881912229e-05, "loss": 0.0089, "step": 37040 }, { "grad_norm": 0.10793177038431168, "learning_rate": 3.494543751550342e-05, "loss": 0.0076, "step": 37050 }, { "grad_norm": 0.1302759349346161, "learning_rate": 3.491916078506313e-05, "loss": 0.0098, "step": 37060 }, { "grad_norm": 0.12941662967205048, "learning_rate": 3.489288863578361e-05, "loss": 0.0062, "step": 37070 }, { "grad_norm": 0.15737994015216827, "learning_rate": 3.4866621075645646e-05, "loss": 0.0066, "step": 37080 }, { "grad_norm": 0.16145141422748566, "learning_rate": 3.4840358112628614e-05, "loss": 0.0085, "step": 37090 }, { "grad_norm": 0.14678581058979034, "learning_rate": 3.481409975471053e-05, "loss": 0.0068, "step": 37100 }, { "grad_norm": 0.12147446721792221, "learning_rate": 3.4787846009867986e-05, "loss": 0.006, "step": 37110 }, { "grad_norm": 0.16713844239711761, "learning_rate": 3.476159688607615e-05, "loss": 0.0052, "step": 37120 }, { "grad_norm": 0.10929717868566513, "learning_rate": 3.4735352391308854e-05, "loss": 0.0065, "step": 37130 }, { "grad_norm": 0.12070167064666748, "learning_rate": 3.4709112533538446e-05, "loss": 0.0057, "step": 37140 }, { "grad_norm": 0.12387269735336304, "learning_rate": 3.4682877320735934e-05, "loss": 0.007, "step": 37150 }, { "grad_norm": 0.14145198464393616, "learning_rate": 3.465664676087085e-05, "loss": 0.0071, "step": 37160 }, { "grad_norm": 0.09708660840988159, "learning_rate": 3.463042086191136e-05, "loss": 0.008, "step": 37170 }, { "grad_norm": 0.0760851502418518, "learning_rate": 3.460419963182423e-05, "loss": 0.0072, "step": 37180 }, { "grad_norm": 0.1700795739889145, "learning_rate": 3.457798307857473e-05, "loss": 0.0105, "step": 37190 }, { "grad_norm": 0.10200081020593643, "learning_rate": 3.455177121012678e-05, "loss": 0.0081, "step": 37200 }, { "grad_norm": 0.1152409017086029, "learning_rate": 3.452556403444285e-05, "loss": 0.0098, "step": 37210 }, { "grad_norm": 0.10869649052619934, "learning_rate": 3.4499361559483975e-05, "loss": 0.0081, "step": 37220 }, { "grad_norm": 0.10979956388473511, "learning_rate": 3.44731637932098e-05, "loss": 0.0092, "step": 37230 }, { "grad_norm": 0.15196092426776886, "learning_rate": 3.44469707435785e-05, "loss": 0.0057, "step": 37240 }, { "grad_norm": 0.08077824115753174, "learning_rate": 3.4420782418546835e-05, "loss": 0.0069, "step": 37250 }, { "grad_norm": 0.1103692576289177, "learning_rate": 3.439459882607012e-05, "loss": 0.0065, "step": 37260 }, { "grad_norm": 0.19839991629123688, "learning_rate": 3.436841997410225e-05, "loss": 0.0064, "step": 37270 }, { "grad_norm": 0.16059467196464539, "learning_rate": 3.434224587059567e-05, "loss": 0.009, "step": 37280 }, { "grad_norm": 0.14140890538692474, "learning_rate": 3.431607652350136e-05, "loss": 0.0056, "step": 37290 }, { "grad_norm": 0.1360873430967331, "learning_rate": 3.428991194076891e-05, "loss": 0.0074, "step": 37300 }, { "grad_norm": 0.11616801470518112, "learning_rate": 3.4263752130346394e-05, "loss": 0.0057, "step": 37310 }, { "grad_norm": 0.10649307072162628, "learning_rate": 3.4237597100180515e-05, "loss": 0.0095, "step": 37320 }, { "grad_norm": 0.125951886177063, "learning_rate": 3.4211446858216427e-05, "loss": 0.0085, "step": 37330 }, { "grad_norm": 0.13866932690143585, "learning_rate": 3.4185301412397915e-05, "loss": 0.0078, "step": 37340 }, { "grad_norm": 0.07170236855745316, "learning_rate": 3.415916077066729e-05, "loss": 0.0078, "step": 37350 }, { "grad_norm": 0.10552391409873962, "learning_rate": 3.413302494096535e-05, "loss": 0.0077, "step": 37360 }, { "grad_norm": 0.09272938966751099, "learning_rate": 3.410689393123151e-05, "loss": 0.0058, "step": 37370 }, { "grad_norm": 0.16376835107803345, "learning_rate": 3.408076774940364e-05, "loss": 0.0071, "step": 37380 }, { "grad_norm": 0.07835295051336288, "learning_rate": 3.40546464034182e-05, "loss": 0.0081, "step": 37390 }, { "grad_norm": 0.12876825034618378, "learning_rate": 3.4028529901210185e-05, "loss": 0.0078, "step": 37400 }, { "grad_norm": 0.09757491201162338, "learning_rate": 3.4002418250713086e-05, "loss": 0.0091, "step": 37410 }, { "grad_norm": 0.15743950009346008, "learning_rate": 3.3976311459858936e-05, "loss": 0.0062, "step": 37420 }, { "grad_norm": 0.14022889733314514, "learning_rate": 3.395020953657826e-05, "loss": 0.0057, "step": 37430 }, { "grad_norm": 0.10844560712575912, "learning_rate": 3.3924112488800165e-05, "loss": 0.0094, "step": 37440 }, { "grad_norm": 0.11271502077579498, "learning_rate": 3.389802032445225e-05, "loss": 0.0075, "step": 37450 }, { "grad_norm": 0.09826766699552536, "learning_rate": 3.38719330514606e-05, "loss": 0.0086, "step": 37460 }, { "grad_norm": 0.13393853604793549, "learning_rate": 3.3845850677749866e-05, "loss": 0.0086, "step": 37470 }, { "grad_norm": 0.13969865441322327, "learning_rate": 3.3819773211243157e-05, "loss": 0.0067, "step": 37480 }, { "grad_norm": 0.1403890699148178, "learning_rate": 3.379370065986213e-05, "loss": 0.0076, "step": 37490 }, { "grad_norm": 0.11010415852069855, "learning_rate": 3.3767633031526955e-05, "loss": 0.0077, "step": 37500 }, { "grad_norm": 0.11215680837631226, "learning_rate": 3.374157033415626e-05, "loss": 0.0069, "step": 37510 }, { "grad_norm": 0.09760154038667679, "learning_rate": 3.371551257566723e-05, "loss": 0.0095, "step": 37520 }, { "grad_norm": 0.17087791860103607, "learning_rate": 3.36894597639755e-05, "loss": 0.0068, "step": 37530 }, { "grad_norm": 0.11585631221532822, "learning_rate": 3.366341190699523e-05, "loss": 0.0068, "step": 37540 }, { "grad_norm": 0.0947478637099266, "learning_rate": 3.36373690126391e-05, "loss": 0.0066, "step": 37550 }, { "grad_norm": 0.09682121127843857, "learning_rate": 3.3611331088818234e-05, "loss": 0.0068, "step": 37560 }, { "grad_norm": 0.13202980160713196, "learning_rate": 3.3585298143442265e-05, "loss": 0.0095, "step": 37570 }, { "grad_norm": 0.11170762777328491, "learning_rate": 3.35592701844193e-05, "loss": 0.007, "step": 37580 }, { "grad_norm": 0.16133680939674377, "learning_rate": 3.353324721965596e-05, "loss": 0.0068, "step": 37590 }, { "grad_norm": 0.09161177277565002, "learning_rate": 3.350722925705736e-05, "loss": 0.0049, "step": 37600 }, { "grad_norm": 0.09545434266328812, "learning_rate": 3.348121630452703e-05, "loss": 0.007, "step": 37610 }, { "grad_norm": 0.1413215696811676, "learning_rate": 3.3455208369967044e-05, "loss": 0.0058, "step": 37620 }, { "grad_norm": 0.12626560032367706, "learning_rate": 3.34292054612779e-05, "loss": 0.0058, "step": 37630 }, { "grad_norm": 0.15247602760791779, "learning_rate": 3.340320758635861e-05, "loss": 0.0056, "step": 37640 }, { "grad_norm": 0.12281513959169388, "learning_rate": 3.337721475310666e-05, "loss": 0.0083, "step": 37650 }, { "grad_norm": 0.11012974381446838, "learning_rate": 3.335122696941795e-05, "loss": 0.0072, "step": 37660 }, { "grad_norm": 0.13896717131137848, "learning_rate": 3.332524424318692e-05, "loss": 0.008, "step": 37670 }, { "grad_norm": 0.15927226841449738, "learning_rate": 3.32992665823064e-05, "loss": 0.0104, "step": 37680 }, { "grad_norm": 0.13294734060764313, "learning_rate": 3.327329399466774e-05, "loss": 0.0091, "step": 37690 }, { "grad_norm": 0.1688428670167923, "learning_rate": 3.324732648816072e-05, "loss": 0.0055, "step": 37700 }, { "grad_norm": 0.10882576555013657, "learning_rate": 3.322136407067358e-05, "loss": 0.0062, "step": 37710 }, { "grad_norm": 0.11416974663734436, "learning_rate": 3.3195406750093036e-05, "loss": 0.0064, "step": 37720 }, { "grad_norm": 0.10251650214195251, "learning_rate": 3.3169454534304205e-05, "loss": 0.0056, "step": 37730 }, { "grad_norm": 0.15380913019180298, "learning_rate": 3.3143507431190725e-05, "loss": 0.007, "step": 37740 }, { "grad_norm": 0.10356797277927399, "learning_rate": 3.311756544863459e-05, "loss": 0.007, "step": 37750 }, { "grad_norm": 0.08939073234796524, "learning_rate": 3.309162859451633e-05, "loss": 0.0068, "step": 37760 }, { "grad_norm": 0.14250344038009644, "learning_rate": 3.306569687671487e-05, "loss": 0.0056, "step": 37770 }, { "grad_norm": 0.15017390251159668, "learning_rate": 3.303977030310756e-05, "loss": 0.0078, "step": 37780 }, { "grad_norm": 0.1305500864982605, "learning_rate": 3.3013848881570245e-05, "loss": 0.0055, "step": 37790 }, { "grad_norm": 0.12914340198040009, "learning_rate": 3.298793261997712e-05, "loss": 0.006, "step": 37800 }, { "grad_norm": 0.07763019949197769, "learning_rate": 3.2962021526200893e-05, "loss": 0.0066, "step": 37810 }, { "grad_norm": 0.1341439038515091, "learning_rate": 3.293611560811268e-05, "loss": 0.0069, "step": 37820 }, { "grad_norm": 0.18862155079841614, "learning_rate": 3.291021487358199e-05, "loss": 0.0057, "step": 37830 }, { "grad_norm": 0.1494808793067932, "learning_rate": 3.28843193304768e-05, "loss": 0.006, "step": 37840 }, { "grad_norm": 0.1869671791791916, "learning_rate": 3.2858428986663456e-05, "loss": 0.0085, "step": 37850 }, { "grad_norm": 0.07325652986764908, "learning_rate": 3.283254385000681e-05, "loss": 0.0058, "step": 37860 }, { "grad_norm": 0.12663641571998596, "learning_rate": 3.2806663928370076e-05, "loss": 0.0058, "step": 37870 }, { "grad_norm": 0.09363584965467453, "learning_rate": 3.278078922961485e-05, "loss": 0.0057, "step": 37880 }, { "grad_norm": 0.06581281870603561, "learning_rate": 3.275491976160123e-05, "loss": 0.0062, "step": 37890 }, { "grad_norm": 0.1318918615579605, "learning_rate": 3.2729055532187645e-05, "loss": 0.0057, "step": 37900 }, { "grad_norm": 0.13737726211547852, "learning_rate": 3.270319654923097e-05, "loss": 0.0064, "step": 37910 }, { "grad_norm": 0.12293432652950287, "learning_rate": 3.2677342820586506e-05, "loss": 0.0058, "step": 37920 }, { "grad_norm": 0.1427280157804489, "learning_rate": 3.2651494354107905e-05, "loss": 0.0066, "step": 37930 }, { "grad_norm": 0.10916595160961151, "learning_rate": 3.2625651157647266e-05, "loss": 0.0058, "step": 37940 }, { "grad_norm": 0.0973946824669838, "learning_rate": 3.259981323905505e-05, "loss": 0.0056, "step": 37950 }, { "grad_norm": 0.1018475890159607, "learning_rate": 3.257398060618014e-05, "loss": 0.007, "step": 37960 }, { "grad_norm": 0.10460755974054337, "learning_rate": 3.254815326686983e-05, "loss": 0.0069, "step": 37970 }, { "grad_norm": 0.17231127619743347, "learning_rate": 3.2522331228969774e-05, "loss": 0.0071, "step": 37980 }, { "grad_norm": 0.07293780893087387, "learning_rate": 3.2496514500324006e-05, "loss": 0.0056, "step": 37990 }, { "grad_norm": 0.16890381276607513, "learning_rate": 3.247070308877498e-05, "loss": 0.0091, "step": 38000 }, { "grad_norm": 0.16217070817947388, "learning_rate": 3.2444897002163515e-05, "loss": 0.0065, "step": 38010 }, { "grad_norm": 0.1316230148077011, "learning_rate": 3.241909624832885e-05, "loss": 0.0063, "step": 38020 }, { "grad_norm": 0.11816340684890747, "learning_rate": 3.239330083510852e-05, "loss": 0.0055, "step": 38030 }, { "grad_norm": 0.1174665093421936, "learning_rate": 3.236751077033855e-05, "loss": 0.0056, "step": 38040 }, { "grad_norm": 0.16081377863883972, "learning_rate": 3.234172606185322e-05, "loss": 0.0076, "step": 38050 }, { "grad_norm": 0.10412835329771042, "learning_rate": 3.231594671748528e-05, "loss": 0.006, "step": 38060 }, { "grad_norm": 0.08428890258073807, "learning_rate": 3.2290172745065815e-05, "loss": 0.0049, "step": 38070 }, { "grad_norm": 0.18319891393184662, "learning_rate": 3.226440415242426e-05, "loss": 0.0071, "step": 38080 }, { "grad_norm": 0.12845052778720856, "learning_rate": 3.223864094738846e-05, "loss": 0.0077, "step": 38090 }, { "grad_norm": 0.1588849127292633, "learning_rate": 3.221288313778456e-05, "loss": 0.0092, "step": 38100 }, { "grad_norm": 0.0926247388124466, "learning_rate": 3.2187130731437125e-05, "loss": 0.0068, "step": 38110 }, { "grad_norm": 0.11667146533727646, "learning_rate": 3.216138373616905e-05, "loss": 0.0065, "step": 38120 }, { "grad_norm": 0.13544712960720062, "learning_rate": 3.21356421598016e-05, "loss": 0.0083, "step": 38130 }, { "grad_norm": 0.10994580388069153, "learning_rate": 3.210990601015438e-05, "loss": 0.0048, "step": 38140 }, { "grad_norm": 0.11213801056146622, "learning_rate": 3.208417529504535e-05, "loss": 0.0055, "step": 38150 }, { "grad_norm": 0.10129038989543915, "learning_rate": 3.205845002229084e-05, "loss": 0.0077, "step": 38160 }, { "grad_norm": 0.1191839873790741, "learning_rate": 3.203273019970547e-05, "loss": 0.0088, "step": 38170 }, { "grad_norm": 0.1042160838842392, "learning_rate": 3.200701583510227e-05, "loss": 0.0055, "step": 38180 }, { "grad_norm": 0.10988601297140121, "learning_rate": 3.198130693629261e-05, "loss": 0.0071, "step": 38190 }, { "grad_norm": 0.12964323163032532, "learning_rate": 3.195560351108612e-05, "loss": 0.0065, "step": 38200 }, { "grad_norm": 0.1668119877576828, "learning_rate": 3.1929905567290865e-05, "loss": 0.007, "step": 38210 }, { "grad_norm": 0.1143944188952446, "learning_rate": 3.1904213112713164e-05, "loss": 0.007, "step": 38220 }, { "grad_norm": 0.09431826323270798, "learning_rate": 3.187852615515774e-05, "loss": 0.0081, "step": 38230 }, { "grad_norm": 0.10894148051738739, "learning_rate": 3.1852844702427606e-05, "loss": 0.0059, "step": 38240 }, { "grad_norm": 0.1554187834262848, "learning_rate": 3.18271687623241e-05, "loss": 0.0075, "step": 38250 }, { "grad_norm": 0.09049777686595917, "learning_rate": 3.1801498342646896e-05, "loss": 0.007, "step": 38260 }, { "grad_norm": 0.1320122331380844, "learning_rate": 3.177583345119398e-05, "loss": 0.0065, "step": 38270 }, { "grad_norm": 0.1240546852350235, "learning_rate": 3.17501740957617e-05, "loss": 0.0064, "step": 38280 }, { "grad_norm": 0.09544296562671661, "learning_rate": 3.172452028414467e-05, "loss": 0.0066, "step": 38290 }, { "grad_norm": 0.14805816113948822, "learning_rate": 3.169887202413583e-05, "loss": 0.0058, "step": 38300 }, { "grad_norm": 0.137800395488739, "learning_rate": 3.167322932352646e-05, "loss": 0.0069, "step": 38310 }, { "grad_norm": 0.10838611423969269, "learning_rate": 3.164759219010613e-05, "loss": 0.0061, "step": 38320 }, { "grad_norm": 0.09396657347679138, "learning_rate": 3.1621960631662725e-05, "loss": 0.0077, "step": 38330 }, { "grad_norm": 0.09364219754934311, "learning_rate": 3.159633465598245e-05, "loss": 0.0092, "step": 38340 }, { "grad_norm": 0.1546711027622223, "learning_rate": 3.1570714270849767e-05, "loss": 0.0071, "step": 38350 }, { "grad_norm": 0.12945564091205597, "learning_rate": 3.1545099484047516e-05, "loss": 0.0061, "step": 38360 }, { "grad_norm": 0.24111242592334747, "learning_rate": 3.151949030335674e-05, "loss": 0.0076, "step": 38370 }, { "grad_norm": 0.19180868566036224, "learning_rate": 3.149388673655687e-05, "loss": 0.0068, "step": 38380 }, { "grad_norm": 0.12605281174182892, "learning_rate": 3.146828879142559e-05, "loss": 0.0058, "step": 38390 }, { "grad_norm": 0.110804982483387, "learning_rate": 3.1442696475738866e-05, "loss": 0.0059, "step": 38400 }, { "grad_norm": 0.16742242872714996, "learning_rate": 3.141710979727098e-05, "loss": 0.0089, "step": 38410 }, { "grad_norm": 0.11603760719299316, "learning_rate": 3.139152876379447e-05, "loss": 0.0065, "step": 38420 }, { "grad_norm": 0.1477184295654297, "learning_rate": 3.1365953383080214e-05, "loss": 0.0071, "step": 38430 }, { "grad_norm": 0.1640712171792984, "learning_rate": 3.134038366289731e-05, "loss": 0.0058, "step": 38440 }, { "grad_norm": 0.15138214826583862, "learning_rate": 3.131481961101317e-05, "loss": 0.0063, "step": 38450 }, { "grad_norm": 0.11197535693645477, "learning_rate": 3.128926123519349e-05, "loss": 0.0079, "step": 38460 }, { "grad_norm": 0.10483185946941376, "learning_rate": 3.1263708543202194e-05, "loss": 0.0077, "step": 38470 }, { "grad_norm": 0.14087358117103577, "learning_rate": 3.123816154280155e-05, "loss": 0.007, "step": 38480 }, { "grad_norm": 0.13359613716602325, "learning_rate": 3.121262024175207e-05, "loss": 0.0074, "step": 38490 }, { "grad_norm": 0.07941890507936478, "learning_rate": 3.118708464781248e-05, "loss": 0.0087, "step": 38500 }, { "grad_norm": 0.09331407397985458, "learning_rate": 3.116155476873987e-05, "loss": 0.008, "step": 38510 }, { "grad_norm": 0.13695932924747467, "learning_rate": 3.11360306122895e-05, "loss": 0.006, "step": 38520 }, { "grad_norm": 0.14312633872032166, "learning_rate": 3.1110512186214975e-05, "loss": 0.007, "step": 38530 }, { "grad_norm": 0.0958724394440651, "learning_rate": 3.1084999498268095e-05, "loss": 0.008, "step": 38540 }, { "grad_norm": 0.13362810015678406, "learning_rate": 3.1059492556198934e-05, "loss": 0.0091, "step": 38550 }, { "grad_norm": 0.11039809882640839, "learning_rate": 3.103399136775586e-05, "loss": 0.0055, "step": 38560 }, { "grad_norm": 0.10257239639759064, "learning_rate": 3.100849594068541e-05, "loss": 0.007, "step": 38570 }, { "grad_norm": 0.14257267117500305, "learning_rate": 3.0983006282732484e-05, "loss": 0.0065, "step": 38580 }, { "grad_norm": 0.1439438760280609, "learning_rate": 3.0957522401640116e-05, "loss": 0.0071, "step": 38590 }, { "grad_norm": 0.08090196549892426, "learning_rate": 3.0932044305149645e-05, "loss": 0.0067, "step": 38600 }, { "grad_norm": 0.07952888309955597, "learning_rate": 3.090657200100068e-05, "loss": 0.0068, "step": 38610 }, { "grad_norm": 0.12599816918373108, "learning_rate": 3.088110549693099e-05, "loss": 0.0098, "step": 38620 }, { "grad_norm": 0.084544837474823, "learning_rate": 3.085564480067667e-05, "loss": 0.0063, "step": 38630 }, { "grad_norm": 0.11306051164865494, "learning_rate": 3.0830189919971955e-05, "loss": 0.0065, "step": 38640 }, { "grad_norm": 0.11927969008684158, "learning_rate": 3.080474086254939e-05, "loss": 0.007, "step": 38650 }, { "grad_norm": 0.144496887922287, "learning_rate": 3.077929763613975e-05, "loss": 0.0068, "step": 38660 }, { "grad_norm": 0.1444762945175171, "learning_rate": 3.075386024847198e-05, "loss": 0.0065, "step": 38670 }, { "grad_norm": 0.11979537457227707, "learning_rate": 3.072842870727331e-05, "loss": 0.0082, "step": 38680 }, { "grad_norm": 0.12391337752342224, "learning_rate": 3.070300302026916e-05, "loss": 0.007, "step": 38690 }, { "grad_norm": 0.08279018849134445, "learning_rate": 3.067758319518318e-05, "loss": 0.006, "step": 38700 }, { "grad_norm": 0.10797414928674698, "learning_rate": 3.065216923973725e-05, "loss": 0.0072, "step": 38710 }, { "grad_norm": 0.09692847728729248, "learning_rate": 3.062676116165145e-05, "loss": 0.0072, "step": 38720 }, { "grad_norm": 0.1610303819179535, "learning_rate": 3.06013589686441e-05, "loss": 0.0075, "step": 38730 }, { "grad_norm": 0.11898607760667801, "learning_rate": 3.05759626684317e-05, "loss": 0.008, "step": 38740 }, { "grad_norm": 0.16863633692264557, "learning_rate": 3.055057226872896e-05, "loss": 0.0085, "step": 38750 }, { "grad_norm": 0.10865630209445953, "learning_rate": 3.052518777724887e-05, "loss": 0.0059, "step": 38760 }, { "grad_norm": 0.1596652865409851, "learning_rate": 3.04998092017025e-05, "loss": 0.0079, "step": 38770 }, { "grad_norm": 0.101386658847332, "learning_rate": 3.0474436549799246e-05, "loss": 0.0068, "step": 38780 }, { "grad_norm": 0.10304393619298935, "learning_rate": 3.044906982924661e-05, "loss": 0.0064, "step": 38790 }, { "grad_norm": 0.10013653337955475, "learning_rate": 3.0423709047750337e-05, "loss": 0.0065, "step": 38800 }, { "grad_norm": 0.12681624293327332, "learning_rate": 3.03983542130144e-05, "loss": 0.0065, "step": 38810 }, { "grad_norm": 0.09869085252285004, "learning_rate": 3.0373005332740877e-05, "loss": 0.0066, "step": 38820 }, { "grad_norm": 0.1473652720451355, "learning_rate": 3.034766241463013e-05, "loss": 0.0071, "step": 38830 }, { "grad_norm": 0.10324922204017639, "learning_rate": 3.032232546638064e-05, "loss": 0.0066, "step": 38840 }, { "grad_norm": 0.09889817982912064, "learning_rate": 3.0296994495689114e-05, "loss": 0.0077, "step": 38850 }, { "grad_norm": 0.11263620108366013, "learning_rate": 3.0271669510250444e-05, "loss": 0.0094, "step": 38860 }, { "grad_norm": 0.14369843900203705, "learning_rate": 3.024635051775766e-05, "loss": 0.0065, "step": 38870 }, { "grad_norm": 0.17481087148189545, "learning_rate": 3.022103752590205e-05, "loss": 0.0074, "step": 38880 }, { "grad_norm": 0.11772217601537704, "learning_rate": 3.0195730542372992e-05, "loss": 0.0068, "step": 38890 }, { "grad_norm": 0.1190761849284172, "learning_rate": 3.0170429574858084e-05, "loss": 0.0079, "step": 38900 }, { "grad_norm": 0.10098757594823837, "learning_rate": 3.0145134631043127e-05, "loss": 0.0072, "step": 38910 }, { "grad_norm": 0.13090744614601135, "learning_rate": 3.0119845718612018e-05, "loss": 0.0064, "step": 38920 }, { "grad_norm": 0.10893768072128296, "learning_rate": 3.009456284524688e-05, "loss": 0.0055, "step": 38930 }, { "grad_norm": 0.08774928748607635, "learning_rate": 3.0069286018627967e-05, "loss": 0.0061, "step": 38940 }, { "grad_norm": 0.13919804990291595, "learning_rate": 3.0044015246433743e-05, "loss": 0.0083, "step": 38950 }, { "grad_norm": 0.09468366950750351, "learning_rate": 3.0018750536340755e-05, "loss": 0.0067, "step": 38960 }, { "grad_norm": 0.11890067160129547, "learning_rate": 2.999349189602378e-05, "loss": 0.0055, "step": 38970 }, { "grad_norm": 0.09815024584531784, "learning_rate": 2.9968239333155733e-05, "loss": 0.0065, "step": 38980 }, { "grad_norm": 0.09865466505289078, "learning_rate": 2.994299285540767e-05, "loss": 0.0076, "step": 38990 }, { "grad_norm": 0.13073867559432983, "learning_rate": 2.9917752470448813e-05, "loss": 0.0061, "step": 39000 }, { "grad_norm": 0.1498776227235794, "learning_rate": 2.9892518185946495e-05, "loss": 0.0075, "step": 39010 }, { "grad_norm": 0.10415022820234299, "learning_rate": 2.986729000956624e-05, "loss": 0.0055, "step": 39020 }, { "grad_norm": 0.09161114692687988, "learning_rate": 2.9842067948971736e-05, "loss": 0.0067, "step": 39030 }, { "grad_norm": 0.1613493114709854, "learning_rate": 2.9816852011824727e-05, "loss": 0.0086, "step": 39040 }, { "grad_norm": 0.13444006443023682, "learning_rate": 2.979164220578519e-05, "loss": 0.0056, "step": 39050 }, { "grad_norm": 0.14974646270275116, "learning_rate": 2.9766438538511165e-05, "loss": 0.0076, "step": 39060 }, { "grad_norm": 0.137882262468338, "learning_rate": 2.9741241017658873e-05, "loss": 0.0076, "step": 39070 }, { "grad_norm": 0.11090516299009323, "learning_rate": 2.971604965088267e-05, "loss": 0.0053, "step": 39080 }, { "grad_norm": 0.10387637466192245, "learning_rate": 2.9690864445835008e-05, "loss": 0.0068, "step": 39090 }, { "grad_norm": 0.09021613746881485, "learning_rate": 2.966568541016651e-05, "loss": 0.0068, "step": 39100 }, { "grad_norm": 0.1425737887620926, "learning_rate": 2.9640512551525867e-05, "loss": 0.0052, "step": 39110 }, { "grad_norm": 0.14365984499454498, "learning_rate": 2.961534587755995e-05, "loss": 0.0088, "step": 39120 }, { "grad_norm": 0.09719514846801758, "learning_rate": 2.959018539591375e-05, "loss": 0.0086, "step": 39130 }, { "grad_norm": 0.15312020480632782, "learning_rate": 2.9565031114230325e-05, "loss": 0.0068, "step": 39140 }, { "grad_norm": 0.09198985248804092, "learning_rate": 2.9539883040150895e-05, "loss": 0.0049, "step": 39150 }, { "grad_norm": 0.17112354934215546, "learning_rate": 2.9514741181314774e-05, "loss": 0.0067, "step": 39160 }, { "grad_norm": 0.08647598326206207, "learning_rate": 2.94896055453594e-05, "loss": 0.0055, "step": 39170 }, { "grad_norm": 0.09235130250453949, "learning_rate": 2.9464476139920332e-05, "loss": 0.0057, "step": 39180 }, { "grad_norm": 0.09608003497123718, "learning_rate": 2.9439352972631186e-05, "loss": 0.0078, "step": 39190 }, { "grad_norm": 0.16062594950199127, "learning_rate": 2.9414236051123757e-05, "loss": 0.0064, "step": 39200 }, { "grad_norm": 0.16799132525920868, "learning_rate": 2.938912538302785e-05, "loss": 0.0063, "step": 39210 }, { "grad_norm": 0.10433726012706757, "learning_rate": 2.9364020975971464e-05, "loss": 0.0072, "step": 39220 }, { "grad_norm": 0.13397331535816193, "learning_rate": 2.9338922837580657e-05, "loss": 0.0055, "step": 39230 }, { "grad_norm": 0.22167237102985382, "learning_rate": 2.931383097547955e-05, "loss": 0.0106, "step": 39240 }, { "grad_norm": 0.10755990445613861, "learning_rate": 2.928874539729043e-05, "loss": 0.0043, "step": 39250 }, { "grad_norm": 0.12663798034191132, "learning_rate": 2.926366611063358e-05, "loss": 0.0048, "step": 39260 }, { "grad_norm": 0.11928591877222061, "learning_rate": 2.9238593123127463e-05, "loss": 0.0047, "step": 39270 }, { "grad_norm": 0.10731702297925949, "learning_rate": 2.9213526442388583e-05, "loss": 0.0083, "step": 39280 }, { "grad_norm": 0.14762090146541595, "learning_rate": 2.9188466076031545e-05, "loss": 0.0084, "step": 39290 }, { "grad_norm": 0.11717171967029572, "learning_rate": 2.9163412031669012e-05, "loss": 0.0064, "step": 39300 }, { "grad_norm": 0.1172061413526535, "learning_rate": 2.913836431691175e-05, "loss": 0.0076, "step": 39310 }, { "grad_norm": 0.131857767701149, "learning_rate": 2.9113322939368583e-05, "loss": 0.0069, "step": 39320 }, { "grad_norm": 0.09248086810112, "learning_rate": 2.9088287906646427e-05, "loss": 0.008, "step": 39330 }, { "grad_norm": 0.11714860796928406, "learning_rate": 2.906325922635024e-05, "loss": 0.0053, "step": 39340 }, { "grad_norm": 0.257415771484375, "learning_rate": 2.903823690608313e-05, "loss": 0.008, "step": 39350 }, { "grad_norm": 0.12913592159748077, "learning_rate": 2.9013220953446174e-05, "loss": 0.0075, "step": 39360 }, { "grad_norm": 0.1071137934923172, "learning_rate": 2.8988211376038564e-05, "loss": 0.0052, "step": 39370 }, { "grad_norm": 0.3282824456691742, "learning_rate": 2.8963208181457564e-05, "loss": 0.006, "step": 39380 }, { "grad_norm": 0.10713055729866028, "learning_rate": 2.8938211377298453e-05, "loss": 0.0065, "step": 39390 }, { "grad_norm": 0.1393366903066635, "learning_rate": 2.8913220971154652e-05, "loss": 0.0055, "step": 39400 }, { "grad_norm": 0.16344556212425232, "learning_rate": 2.888823697061753e-05, "loss": 0.0065, "step": 39410 }, { "grad_norm": 0.11965668946504593, "learning_rate": 2.8863259383276618e-05, "loss": 0.0059, "step": 39420 }, { "grad_norm": 0.13126720488071442, "learning_rate": 2.8838288216719395e-05, "loss": 0.008, "step": 39430 }, { "grad_norm": 0.10391475260257721, "learning_rate": 2.8813323478531484e-05, "loss": 0.0084, "step": 39440 }, { "grad_norm": 0.08540089428424835, "learning_rate": 2.8788365176296496e-05, "loss": 0.0064, "step": 39450 }, { "grad_norm": 0.12683479487895966, "learning_rate": 2.876341331759611e-05, "loss": 0.0056, "step": 39460 }, { "grad_norm": 0.13696612417697906, "learning_rate": 2.8738467910010036e-05, "loss": 0.0056, "step": 39470 }, { "grad_norm": 0.1289096474647522, "learning_rate": 2.8713528961116032e-05, "loss": 0.0057, "step": 39480 }, { "grad_norm": 0.11299335211515427, "learning_rate": 2.8688596478489875e-05, "loss": 0.0069, "step": 39490 }, { "grad_norm": 0.13159984350204468, "learning_rate": 2.8663670469705434e-05, "loss": 0.0057, "step": 39500 }, { "grad_norm": 0.18043431639671326, "learning_rate": 2.8638750942334546e-05, "loss": 0.0105, "step": 39510 }, { "grad_norm": 0.15355703234672546, "learning_rate": 2.8613837903947115e-05, "loss": 0.0081, "step": 39520 }, { "grad_norm": 0.132408007979393, "learning_rate": 2.858893136211106e-05, "loss": 0.0054, "step": 39530 }, { "grad_norm": 0.10902559012174606, "learning_rate": 2.8564031324392315e-05, "loss": 0.0065, "step": 39540 }, { "grad_norm": 0.09491493552923203, "learning_rate": 2.85391377983549e-05, "loss": 0.0056, "step": 39550 }, { "grad_norm": 0.12559708952903748, "learning_rate": 2.851425079156075e-05, "loss": 0.0054, "step": 39560 }, { "grad_norm": 0.11759500950574875, "learning_rate": 2.848937031156994e-05, "loss": 0.0059, "step": 39570 }, { "grad_norm": 0.14830181002616882, "learning_rate": 2.846449636594044e-05, "loss": 0.0077, "step": 39580 }, { "grad_norm": 0.14699554443359375, "learning_rate": 2.843962896222836e-05, "loss": 0.0077, "step": 39590 }, { "grad_norm": 0.13067728281021118, "learning_rate": 2.8414768107987722e-05, "loss": 0.0057, "step": 39600 }, { "grad_norm": 0.11412951350212097, "learning_rate": 2.838991381077061e-05, "loss": 0.0066, "step": 39610 }, { "grad_norm": 0.17204080522060394, "learning_rate": 2.83650660781271e-05, "loss": 0.0049, "step": 39620 }, { "grad_norm": 0.1361851990222931, "learning_rate": 2.8340224917605285e-05, "loss": 0.0058, "step": 39630 }, { "grad_norm": 0.12692902982234955, "learning_rate": 2.831539033675122e-05, "loss": 0.006, "step": 39640 }, { "grad_norm": 0.10396310687065125, "learning_rate": 2.8290562343109038e-05, "loss": 0.0046, "step": 39650 }, { "grad_norm": 0.11130504310131073, "learning_rate": 2.826574094422082e-05, "loss": 0.0064, "step": 39660 }, { "grad_norm": 0.12165484577417374, "learning_rate": 2.8240926147626645e-05, "loss": 0.0069, "step": 39670 }, { "grad_norm": 0.1143483817577362, "learning_rate": 2.8216117960864586e-05, "loss": 0.0072, "step": 39680 }, { "grad_norm": 0.24515694379806519, "learning_rate": 2.8191316391470703e-05, "loss": 0.0065, "step": 39690 }, { "grad_norm": 0.10586488991975784, "learning_rate": 2.816652144697911e-05, "loss": 0.0067, "step": 39700 }, { "grad_norm": 0.09483800828456879, "learning_rate": 2.8141733134921783e-05, "loss": 0.007, "step": 39710 }, { "grad_norm": 0.08686671406030655, "learning_rate": 2.811695146282884e-05, "loss": 0.0052, "step": 39720 }, { "grad_norm": 0.15082742273807526, "learning_rate": 2.8092176438228212e-05, "loss": 0.0071, "step": 39730 }, { "grad_norm": 0.12295253574848175, "learning_rate": 2.806740806864598e-05, "loss": 0.0089, "step": 39740 }, { "grad_norm": 0.10172870010137558, "learning_rate": 2.804264636160604e-05, "loss": 0.0049, "step": 39750 }, { "grad_norm": 0.11858488619327545, "learning_rate": 2.8017891324630402e-05, "loss": 0.0063, "step": 39760 }, { "grad_norm": 0.08705195784568787, "learning_rate": 2.7993142965238976e-05, "loss": 0.0062, "step": 39770 }, { "grad_norm": 0.10987970232963562, "learning_rate": 2.7968401290949665e-05, "loss": 0.006, "step": 39780 }, { "grad_norm": 0.12476957589387894, "learning_rate": 2.7943666309278328e-05, "loss": 0.0081, "step": 39790 }, { "grad_norm": 0.24269619584083557, "learning_rate": 2.7918938027738783e-05, "loss": 0.0067, "step": 39800 }, { "grad_norm": 0.08510522544384003, "learning_rate": 2.789421645384287e-05, "loss": 0.0069, "step": 39810 }, { "grad_norm": 0.10297483205795288, "learning_rate": 2.786950159510032e-05, "loss": 0.0078, "step": 39820 }, { "grad_norm": 0.10112371295690536, "learning_rate": 2.7844793459018876e-05, "loss": 0.0061, "step": 39830 }, { "grad_norm": 0.13950401544570923, "learning_rate": 2.7820092053104195e-05, "loss": 0.0064, "step": 39840 }, { "grad_norm": 0.1379530429840088, "learning_rate": 2.7795397384859933e-05, "loss": 0.0084, "step": 39850 }, { "grad_norm": 0.10481324046850204, "learning_rate": 2.7770709461787638e-05, "loss": 0.0062, "step": 39860 }, { "grad_norm": 0.12570089101791382, "learning_rate": 2.7746028291386915e-05, "loss": 0.0057, "step": 39870 }, { "grad_norm": 0.11382874101400375, "learning_rate": 2.772135388115519e-05, "loss": 0.0063, "step": 39880 }, { "grad_norm": 0.13827571272850037, "learning_rate": 2.7696686238587945e-05, "loss": 0.0097, "step": 39890 }, { "grad_norm": 0.12058261036872864, "learning_rate": 2.7672025371178505e-05, "loss": 0.0078, "step": 39900 }, { "grad_norm": 0.1501762717962265, "learning_rate": 2.7647371286418238e-05, "loss": 0.0067, "step": 39910 }, { "grad_norm": 0.09086267650127411, "learning_rate": 2.762272399179639e-05, "loss": 0.0076, "step": 39920 }, { "grad_norm": 0.09816907346248627, "learning_rate": 2.7598083494800154e-05, "loss": 0.0085, "step": 39930 }, { "grad_norm": 0.1151476725935936, "learning_rate": 2.7573449802914664e-05, "loss": 0.0052, "step": 39940 }, { "grad_norm": 0.1178794726729393, "learning_rate": 2.7548822923622964e-05, "loss": 0.0051, "step": 39950 }, { "grad_norm": 0.10875263065099716, "learning_rate": 2.752420286440609e-05, "loss": 0.0069, "step": 39960 }, { "grad_norm": 0.1414417177438736, "learning_rate": 2.749958963274295e-05, "loss": 0.0052, "step": 39970 }, { "grad_norm": 0.1147313341498375, "learning_rate": 2.747498323611039e-05, "loss": 0.006, "step": 39980 }, { "grad_norm": 0.1301727145910263, "learning_rate": 2.7450383681983184e-05, "loss": 0.0054, "step": 39990 }, { "grad_norm": 0.09649036824703217, "learning_rate": 2.742579097783403e-05, "loss": 0.005, "step": 40000 } ], "logging_steps": 10, "max_steps": 60000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 48, "trial_name": null, "trial_params": null }