{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 100000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.5012474060058594, "learning_rate": 1.8e-07, "loss": 1.5525, "step": 10 }, { "grad_norm": 1.2146625518798828, "learning_rate": 3.8e-07, "loss": 1.5673, "step": 20 }, { "grad_norm": 1.0245391130447388, "learning_rate": 5.8e-07, "loss": 1.532, "step": 30 }, { "grad_norm": 1.2960489988327026, "learning_rate": 7.8e-07, "loss": 1.5514, "step": 40 }, { "grad_norm": 1.2018564939498901, "learning_rate": 9.8e-07, "loss": 1.5242, "step": 50 }, { "grad_norm": 1.1285942792892456, "learning_rate": 1.18e-06, "loss": 1.5026, "step": 60 }, { "grad_norm": 1.4625284671783447, "learning_rate": 1.3800000000000001e-06, "loss": 1.4507, "step": 70 }, { "grad_norm": 2.491309642791748, "learning_rate": 1.5800000000000003e-06, "loss": 1.369, "step": 80 }, { "grad_norm": 1.7699050903320312, "learning_rate": 1.7800000000000001e-06, "loss": 1.282, "step": 90 }, { "grad_norm": 2.565260648727417, "learning_rate": 1.98e-06, "loss": 1.2342, "step": 100 }, { "grad_norm": 1.815170407295227, "learning_rate": 2.1800000000000003e-06, "loss": 1.1986, "step": 110 }, { "grad_norm": 2.2113091945648193, "learning_rate": 2.38e-06, "loss": 1.236, "step": 120 }, { "grad_norm": 2.0573925971984863, "learning_rate": 2.5800000000000003e-06, "loss": 1.2135, "step": 130 }, { "grad_norm": 3.3078083992004395, "learning_rate": 2.78e-06, "loss": 1.2257, "step": 140 }, { "grad_norm": 2.314305067062378, "learning_rate": 2.9800000000000003e-06, "loss": 1.2289, "step": 150 }, { "grad_norm": 1.9644147157669067, "learning_rate": 3.1800000000000005e-06, "loss": 1.2049, "step": 160 }, { "grad_norm": 2.068976879119873, "learning_rate": 3.38e-06, "loss": 1.1663, "step": 170 }, { "grad_norm": 2.035574436187744, "learning_rate": 3.58e-06, "loss": 1.1568, "step": 180 }, { "grad_norm": 2.5194318294525146, "learning_rate": 3.7800000000000002e-06, "loss": 1.1833, "step": 190 }, { "grad_norm": 1.6756360530853271, "learning_rate": 3.98e-06, "loss": 1.1714, "step": 200 }, { "grad_norm": 2.106815814971924, "learning_rate": 4.18e-06, "loss": 1.1695, "step": 210 }, { "grad_norm": 1.9852526187896729, "learning_rate": 4.38e-06, "loss": 1.1379, "step": 220 }, { "grad_norm": 1.5099138021469116, "learning_rate": 4.58e-06, "loss": 1.136, "step": 230 }, { "grad_norm": 1.867788314819336, "learning_rate": 4.780000000000001e-06, "loss": 1.1248, "step": 240 }, { "grad_norm": 1.9127310514450073, "learning_rate": 4.98e-06, "loss": 1.1251, "step": 250 }, { "grad_norm": 2.1361348628997803, "learning_rate": 5.18e-06, "loss": 1.1177, "step": 260 }, { "grad_norm": 2.23775577545166, "learning_rate": 5.38e-06, "loss": 1.1193, "step": 270 }, { "grad_norm": 1.9133384227752686, "learning_rate": 5.580000000000001e-06, "loss": 1.106, "step": 280 }, { "grad_norm": 2.159597873687744, "learning_rate": 5.78e-06, "loss": 1.101, "step": 290 }, { "grad_norm": 2.0629966259002686, "learning_rate": 5.98e-06, "loss": 1.0899, "step": 300 }, { "grad_norm": 2.2393853664398193, "learning_rate": 6.18e-06, "loss": 1.1086, "step": 310 }, { "grad_norm": 1.6421185731887817, "learning_rate": 6.38e-06, "loss": 1.0902, "step": 320 }, { "grad_norm": 1.9355559349060059, "learning_rate": 6.58e-06, "loss": 1.0683, "step": 330 }, { "grad_norm": 1.5663796663284302, "learning_rate": 6.78e-06, "loss": 1.0987, "step": 340 }, { "grad_norm": 2.4907419681549072, "learning_rate": 6.98e-06, "loss": 1.0911, "step": 350 }, { "grad_norm": 1.7491058111190796, "learning_rate": 7.180000000000001e-06, "loss": 1.0778, "step": 360 }, { "grad_norm": 1.748526692390442, "learning_rate": 7.3800000000000005e-06, "loss": 1.0571, "step": 370 }, { "grad_norm": 1.6035287380218506, "learning_rate": 7.580000000000001e-06, "loss": 1.1052, "step": 380 }, { "grad_norm": 2.3182404041290283, "learning_rate": 7.78e-06, "loss": 1.0707, "step": 390 }, { "grad_norm": 1.8089001178741455, "learning_rate": 7.98e-06, "loss": 1.0695, "step": 400 }, { "grad_norm": 1.584087610244751, "learning_rate": 8.18e-06, "loss": 1.0992, "step": 410 }, { "grad_norm": 1.9932029247283936, "learning_rate": 8.380000000000001e-06, "loss": 1.0764, "step": 420 }, { "grad_norm": 4.949777126312256, "learning_rate": 8.580000000000001e-06, "loss": 1.0448, "step": 430 }, { "grad_norm": 2.0596437454223633, "learning_rate": 8.78e-06, "loss": 1.0755, "step": 440 }, { "grad_norm": 1.8806339502334595, "learning_rate": 8.98e-06, "loss": 1.0911, "step": 450 }, { "grad_norm": 1.625296950340271, "learning_rate": 9.180000000000002e-06, "loss": 1.0797, "step": 460 }, { "grad_norm": 2.0157721042633057, "learning_rate": 9.38e-06, "loss": 1.0609, "step": 470 }, { "grad_norm": 1.7422446012496948, "learning_rate": 9.58e-06, "loss": 1.0609, "step": 480 }, { "grad_norm": 1.859240174293518, "learning_rate": 9.78e-06, "loss": 1.0865, "step": 490 }, { "grad_norm": 2.216200828552246, "learning_rate": 9.980000000000001e-06, "loss": 1.0272, "step": 500 }, { "grad_norm": 1.9646296501159668, "learning_rate": 1.018e-05, "loss": 1.0761, "step": 510 }, { "grad_norm": 1.8490190505981445, "learning_rate": 1.038e-05, "loss": 1.0793, "step": 520 }, { "grad_norm": 1.879969596862793, "learning_rate": 1.058e-05, "loss": 1.0605, "step": 530 }, { "grad_norm": 2.3852312564849854, "learning_rate": 1.0780000000000002e-05, "loss": 1.0438, "step": 540 }, { "grad_norm": 1.9657657146453857, "learning_rate": 1.098e-05, "loss": 1.0231, "step": 550 }, { "grad_norm": 2.594092607498169, "learning_rate": 1.118e-05, "loss": 1.0158, "step": 560 }, { "grad_norm": 4.822360992431641, "learning_rate": 1.1380000000000001e-05, "loss": 0.9903, "step": 570 }, { "grad_norm": 2.538457155227661, "learning_rate": 1.1580000000000001e-05, "loss": 0.9999, "step": 580 }, { "grad_norm": 3.5164666175842285, "learning_rate": 1.178e-05, "loss": 0.9958, "step": 590 }, { "grad_norm": 2.6633036136627197, "learning_rate": 1.198e-05, "loss": 0.9674, "step": 600 }, { "grad_norm": 2.179823637008667, "learning_rate": 1.2180000000000002e-05, "loss": 0.9665, "step": 610 }, { "grad_norm": 2.2518794536590576, "learning_rate": 1.238e-05, "loss": 0.9183, "step": 620 }, { "grad_norm": 2.3668229579925537, "learning_rate": 1.258e-05, "loss": 0.9052, "step": 630 }, { "grad_norm": 2.5538995265960693, "learning_rate": 1.278e-05, "loss": 0.874, "step": 640 }, { "grad_norm": 2.36499285697937, "learning_rate": 1.2980000000000001e-05, "loss": 0.8575, "step": 650 }, { "grad_norm": 3.4141745567321777, "learning_rate": 1.3180000000000001e-05, "loss": 0.861, "step": 660 }, { "grad_norm": 3.2661240100860596, "learning_rate": 1.338e-05, "loss": 0.8228, "step": 670 }, { "grad_norm": 2.9554224014282227, "learning_rate": 1.358e-05, "loss": 0.8104, "step": 680 }, { "grad_norm": 2.683201313018799, "learning_rate": 1.3780000000000002e-05, "loss": 0.7841, "step": 690 }, { "grad_norm": 3.1050238609313965, "learning_rate": 1.3980000000000002e-05, "loss": 0.7959, "step": 700 }, { "grad_norm": 2.910602569580078, "learning_rate": 1.4180000000000001e-05, "loss": 0.7791, "step": 710 }, { "grad_norm": 2.955834150314331, "learning_rate": 1.4380000000000001e-05, "loss": 0.7284, "step": 720 }, { "grad_norm": 2.559694290161133, "learning_rate": 1.4580000000000003e-05, "loss": 0.7408, "step": 730 }, { "grad_norm": 2.489654779434204, "learning_rate": 1.4779999999999999e-05, "loss": 0.7242, "step": 740 }, { "grad_norm": 2.2240114212036133, "learning_rate": 1.4979999999999999e-05, "loss": 0.6954, "step": 750 }, { "grad_norm": 3.076728105545044, "learning_rate": 1.518e-05, "loss": 0.6786, "step": 760 }, { "grad_norm": 3.0400826930999756, "learning_rate": 1.538e-05, "loss": 0.6937, "step": 770 }, { "grad_norm": 3.052520275115967, "learning_rate": 1.558e-05, "loss": 0.6674, "step": 780 }, { "grad_norm": 3.1089158058166504, "learning_rate": 1.578e-05, "loss": 0.6164, "step": 790 }, { "grad_norm": 3.6023967266082764, "learning_rate": 1.598e-05, "loss": 0.6909, "step": 800 }, { "grad_norm": 2.909393072128296, "learning_rate": 1.618e-05, "loss": 0.6051, "step": 810 }, { "grad_norm": 3.5538527965545654, "learning_rate": 1.6380000000000002e-05, "loss": 0.5828, "step": 820 }, { "grad_norm": 4.017760753631592, "learning_rate": 1.658e-05, "loss": 0.5581, "step": 830 }, { "grad_norm": 3.1244137287139893, "learning_rate": 1.6780000000000002e-05, "loss": 0.5734, "step": 840 }, { "grad_norm": 3.541074514389038, "learning_rate": 1.698e-05, "loss": 0.5996, "step": 850 }, { "grad_norm": 4.022465705871582, "learning_rate": 1.718e-05, "loss": 0.5678, "step": 860 }, { "grad_norm": 3.0104103088378906, "learning_rate": 1.7380000000000003e-05, "loss": 0.5548, "step": 870 }, { "grad_norm": 3.0066187381744385, "learning_rate": 1.758e-05, "loss": 0.5235, "step": 880 }, { "grad_norm": 2.7671992778778076, "learning_rate": 1.7780000000000003e-05, "loss": 0.4983, "step": 890 }, { "grad_norm": 3.1867687702178955, "learning_rate": 1.798e-05, "loss": 0.4737, "step": 900 }, { "grad_norm": 3.0164501667022705, "learning_rate": 1.818e-05, "loss": 0.5089, "step": 910 }, { "grad_norm": 3.0226709842681885, "learning_rate": 1.838e-05, "loss": 0.5159, "step": 920 }, { "grad_norm": 2.300595760345459, "learning_rate": 1.858e-05, "loss": 0.4902, "step": 930 }, { "grad_norm": 4.1822028160095215, "learning_rate": 1.878e-05, "loss": 0.4972, "step": 940 }, { "grad_norm": 4.099733352661133, "learning_rate": 1.898e-05, "loss": 0.4952, "step": 950 }, { "grad_norm": 2.693007230758667, "learning_rate": 1.918e-05, "loss": 0.4559, "step": 960 }, { "grad_norm": 2.976850748062134, "learning_rate": 1.938e-05, "loss": 0.4897, "step": 970 }, { "grad_norm": 2.8378570079803467, "learning_rate": 1.9580000000000002e-05, "loss": 0.4388, "step": 980 }, { "grad_norm": 2.9693796634674072, "learning_rate": 1.978e-05, "loss": 0.4219, "step": 990 }, { "grad_norm": 4.528493881225586, "learning_rate": 1.9980000000000002e-05, "loss": 0.4259, "step": 1000 }, { "grad_norm": 3.414433717727661, "learning_rate": 2.0180000000000003e-05, "loss": 0.4443, "step": 1010 }, { "grad_norm": 3.057544708251953, "learning_rate": 2.038e-05, "loss": 0.4521, "step": 1020 }, { "grad_norm": 3.560657024383545, "learning_rate": 2.0580000000000003e-05, "loss": 0.4131, "step": 1030 }, { "grad_norm": 2.192399024963379, "learning_rate": 2.078e-05, "loss": 0.4234, "step": 1040 }, { "grad_norm": 3.586073875427246, "learning_rate": 2.098e-05, "loss": 0.4513, "step": 1050 }, { "grad_norm": 3.4540517330169678, "learning_rate": 2.118e-05, "loss": 0.4026, "step": 1060 }, { "grad_norm": 3.1791276931762695, "learning_rate": 2.138e-05, "loss": 0.4305, "step": 1070 }, { "grad_norm": 2.223522901535034, "learning_rate": 2.158e-05, "loss": 0.3517, "step": 1080 }, { "grad_norm": 3.3637428283691406, "learning_rate": 2.178e-05, "loss": 0.4503, "step": 1090 }, { "grad_norm": 2.332475423812866, "learning_rate": 2.198e-05, "loss": 0.4013, "step": 1100 }, { "grad_norm": 2.4115912914276123, "learning_rate": 2.218e-05, "loss": 0.4089, "step": 1110 }, { "grad_norm": 3.197923183441162, "learning_rate": 2.2380000000000003e-05, "loss": 0.3901, "step": 1120 }, { "grad_norm": 2.956510305404663, "learning_rate": 2.258e-05, "loss": 0.3813, "step": 1130 }, { "grad_norm": 3.0005507469177246, "learning_rate": 2.2780000000000002e-05, "loss": 0.3486, "step": 1140 }, { "grad_norm": 2.569953203201294, "learning_rate": 2.298e-05, "loss": 0.3446, "step": 1150 }, { "grad_norm": 3.000540256500244, "learning_rate": 2.318e-05, "loss": 0.3677, "step": 1160 }, { "grad_norm": 2.9901797771453857, "learning_rate": 2.3380000000000003e-05, "loss": 0.3264, "step": 1170 }, { "grad_norm": 3.1437904834747314, "learning_rate": 2.358e-05, "loss": 0.3593, "step": 1180 }, { "grad_norm": 2.3716585636138916, "learning_rate": 2.3780000000000003e-05, "loss": 0.3407, "step": 1190 }, { "grad_norm": 1.979577660560608, "learning_rate": 2.398e-05, "loss": 0.3512, "step": 1200 }, { "grad_norm": 2.4803481101989746, "learning_rate": 2.418e-05, "loss": 0.3573, "step": 1210 }, { "grad_norm": 2.7531981468200684, "learning_rate": 2.438e-05, "loss": 0.3124, "step": 1220 }, { "grad_norm": 3.115539789199829, "learning_rate": 2.4580000000000002e-05, "loss": 0.3298, "step": 1230 }, { "grad_norm": 3.1382100582122803, "learning_rate": 2.478e-05, "loss": 0.3156, "step": 1240 }, { "grad_norm": 2.328143835067749, "learning_rate": 2.498e-05, "loss": 0.3137, "step": 1250 }, { "grad_norm": 2.782494306564331, "learning_rate": 2.5180000000000003e-05, "loss": 0.3119, "step": 1260 }, { "grad_norm": 3.0141594409942627, "learning_rate": 2.5380000000000004e-05, "loss": 0.3024, "step": 1270 }, { "grad_norm": 3.6324706077575684, "learning_rate": 2.5580000000000002e-05, "loss": 0.3131, "step": 1280 }, { "grad_norm": 3.070713520050049, "learning_rate": 2.5779999999999997e-05, "loss": 0.287, "step": 1290 }, { "grad_norm": 3.363847255706787, "learning_rate": 2.598e-05, "loss": 0.2972, "step": 1300 }, { "grad_norm": 2.8572022914886475, "learning_rate": 2.618e-05, "loss": 0.2679, "step": 1310 }, { "grad_norm": 2.777320623397827, "learning_rate": 2.6379999999999998e-05, "loss": 0.3181, "step": 1320 }, { "grad_norm": 3.105717420578003, "learning_rate": 2.658e-05, "loss": 0.2373, "step": 1330 }, { "grad_norm": 2.0183348655700684, "learning_rate": 2.678e-05, "loss": 0.2494, "step": 1340 }, { "grad_norm": 2.5448684692382812, "learning_rate": 2.698e-05, "loss": 0.2357, "step": 1350 }, { "grad_norm": 3.3505444526672363, "learning_rate": 2.718e-05, "loss": 0.2912, "step": 1360 }, { "grad_norm": 2.263275384902954, "learning_rate": 2.738e-05, "loss": 0.2438, "step": 1370 }, { "grad_norm": 3.2260284423828125, "learning_rate": 2.758e-05, "loss": 0.253, "step": 1380 }, { "grad_norm": 3.4103755950927734, "learning_rate": 2.778e-05, "loss": 0.2465, "step": 1390 }, { "grad_norm": 3.0476861000061035, "learning_rate": 2.798e-05, "loss": 0.2819, "step": 1400 }, { "grad_norm": 2.611722230911255, "learning_rate": 2.818e-05, "loss": 0.2132, "step": 1410 }, { "grad_norm": 3.0690524578094482, "learning_rate": 2.8380000000000003e-05, "loss": 0.2489, "step": 1420 }, { "grad_norm": 2.908590316772461, "learning_rate": 2.858e-05, "loss": 0.2381, "step": 1430 }, { "grad_norm": 2.4160943031311035, "learning_rate": 2.8780000000000002e-05, "loss": 0.229, "step": 1440 }, { "grad_norm": 1.7800403833389282, "learning_rate": 2.898e-05, "loss": 0.1948, "step": 1450 }, { "grad_norm": 3.3379688262939453, "learning_rate": 2.9180000000000002e-05, "loss": 0.235, "step": 1460 }, { "grad_norm": 2.2461602687835693, "learning_rate": 2.9380000000000003e-05, "loss": 0.1991, "step": 1470 }, { "grad_norm": 2.0999789237976074, "learning_rate": 2.958e-05, "loss": 0.1676, "step": 1480 }, { "grad_norm": 2.347161293029785, "learning_rate": 2.9780000000000003e-05, "loss": 0.2206, "step": 1490 }, { "grad_norm": 2.8292829990386963, "learning_rate": 2.998e-05, "loss": 0.1901, "step": 1500 }, { "grad_norm": 2.489220380783081, "learning_rate": 3.0180000000000002e-05, "loss": 0.2261, "step": 1510 }, { "grad_norm": 2.447613000869751, "learning_rate": 3.0380000000000004e-05, "loss": 0.1999, "step": 1520 }, { "grad_norm": 2.476598024368286, "learning_rate": 3.058e-05, "loss": 0.2029, "step": 1530 }, { "grad_norm": 2.0163443088531494, "learning_rate": 3.078e-05, "loss": 0.2321, "step": 1540 }, { "grad_norm": 2.3641507625579834, "learning_rate": 3.0980000000000005e-05, "loss": 0.2793, "step": 1550 }, { "grad_norm": 2.9125075340270996, "learning_rate": 3.118e-05, "loss": 0.2468, "step": 1560 }, { "grad_norm": 2.3041059970855713, "learning_rate": 3.138e-05, "loss": 0.1786, "step": 1570 }, { "grad_norm": 2.304486036300659, "learning_rate": 3.1580000000000006e-05, "loss": 0.2144, "step": 1580 }, { "grad_norm": 2.4906227588653564, "learning_rate": 3.1780000000000004e-05, "loss": 0.2019, "step": 1590 }, { "grad_norm": 2.0444188117980957, "learning_rate": 3.198e-05, "loss": 0.192, "step": 1600 }, { "grad_norm": 2.422886610031128, "learning_rate": 3.218e-05, "loss": 0.1945, "step": 1610 }, { "grad_norm": 2.27366042137146, "learning_rate": 3.238e-05, "loss": 0.1989, "step": 1620 }, { "grad_norm": 2.4706530570983887, "learning_rate": 3.2579999999999996e-05, "loss": 0.195, "step": 1630 }, { "grad_norm": 2.5368151664733887, "learning_rate": 3.278e-05, "loss": 0.226, "step": 1640 }, { "grad_norm": 3.5510096549987793, "learning_rate": 3.298e-05, "loss": 0.2474, "step": 1650 }, { "grad_norm": 2.3665149211883545, "learning_rate": 3.318e-05, "loss": 0.1968, "step": 1660 }, { "grad_norm": 2.116311550140381, "learning_rate": 3.338e-05, "loss": 0.218, "step": 1670 }, { "grad_norm": 2.028407335281372, "learning_rate": 3.358e-05, "loss": 0.2265, "step": 1680 }, { "grad_norm": 2.1220433712005615, "learning_rate": 3.378e-05, "loss": 0.2066, "step": 1690 }, { "grad_norm": 2.1847188472747803, "learning_rate": 3.398e-05, "loss": 0.226, "step": 1700 }, { "grad_norm": 2.2854065895080566, "learning_rate": 3.418e-05, "loss": 0.1794, "step": 1710 }, { "grad_norm": 2.3652238845825195, "learning_rate": 3.438e-05, "loss": 0.1976, "step": 1720 }, { "grad_norm": 2.1055116653442383, "learning_rate": 3.4580000000000004e-05, "loss": 0.2288, "step": 1730 }, { "grad_norm": 2.4365599155426025, "learning_rate": 3.478e-05, "loss": 0.2127, "step": 1740 }, { "grad_norm": 2.304931163787842, "learning_rate": 3.498e-05, "loss": 0.1886, "step": 1750 }, { "grad_norm": 2.3138813972473145, "learning_rate": 3.518e-05, "loss": 0.1883, "step": 1760 }, { "grad_norm": 3.1194801330566406, "learning_rate": 3.5380000000000003e-05, "loss": 0.2064, "step": 1770 }, { "grad_norm": 2.2941689491271973, "learning_rate": 3.558e-05, "loss": 0.1951, "step": 1780 }, { "grad_norm": 1.951239824295044, "learning_rate": 3.578e-05, "loss": 0.1928, "step": 1790 }, { "grad_norm": 2.508244037628174, "learning_rate": 3.5980000000000004e-05, "loss": 0.2038, "step": 1800 }, { "grad_norm": 1.5855786800384521, "learning_rate": 3.618e-05, "loss": 0.185, "step": 1810 }, { "grad_norm": 1.9359112977981567, "learning_rate": 3.638e-05, "loss": 0.2126, "step": 1820 }, { "grad_norm": 2.7339859008789062, "learning_rate": 3.6580000000000006e-05, "loss": 0.173, "step": 1830 }, { "grad_norm": 2.4343347549438477, "learning_rate": 3.6780000000000004e-05, "loss": 0.1873, "step": 1840 }, { "grad_norm": 2.19831919670105, "learning_rate": 3.698e-05, "loss": 0.2005, "step": 1850 }, { "grad_norm": 1.4627195596694946, "learning_rate": 3.7180000000000007e-05, "loss": 0.233, "step": 1860 }, { "grad_norm": 2.2731926441192627, "learning_rate": 3.7380000000000005e-05, "loss": 0.2055, "step": 1870 }, { "grad_norm": 2.305877447128296, "learning_rate": 3.758e-05, "loss": 0.2018, "step": 1880 }, { "grad_norm": 2.298448085784912, "learning_rate": 3.778000000000001e-05, "loss": 0.2044, "step": 1890 }, { "grad_norm": 1.8457739353179932, "learning_rate": 3.7980000000000006e-05, "loss": 0.1974, "step": 1900 }, { "grad_norm": 1.5577192306518555, "learning_rate": 3.818e-05, "loss": 0.2089, "step": 1910 }, { "grad_norm": 2.170569896697998, "learning_rate": 3.838e-05, "loss": 0.1819, "step": 1920 }, { "grad_norm": 1.7001880407333374, "learning_rate": 3.858e-05, "loss": 0.2387, "step": 1930 }, { "grad_norm": 1.7247507572174072, "learning_rate": 3.878e-05, "loss": 0.195, "step": 1940 }, { "grad_norm": 1.400775671005249, "learning_rate": 3.898e-05, "loss": 0.1677, "step": 1950 }, { "grad_norm": 2.1194846630096436, "learning_rate": 3.918e-05, "loss": 0.1727, "step": 1960 }, { "grad_norm": 2.724498748779297, "learning_rate": 3.938e-05, "loss": 0.1891, "step": 1970 }, { "grad_norm": 1.7648111581802368, "learning_rate": 3.958e-05, "loss": 0.1858, "step": 1980 }, { "grad_norm": 1.6467453241348267, "learning_rate": 3.978e-05, "loss": 0.1899, "step": 1990 }, { "grad_norm": 2.688725471496582, "learning_rate": 3.998e-05, "loss": 0.1797, "step": 2000 }, { "grad_norm": 1.3593108654022217, "learning_rate": 4.018e-05, "loss": 0.1782, "step": 2010 }, { "grad_norm": 2.1957039833068848, "learning_rate": 4.038e-05, "loss": 0.1771, "step": 2020 }, { "grad_norm": 1.7742526531219482, "learning_rate": 4.058e-05, "loss": 0.2093, "step": 2030 }, { "grad_norm": 1.9274981021881104, "learning_rate": 4.078e-05, "loss": 0.1963, "step": 2040 }, { "grad_norm": 1.7410703897476196, "learning_rate": 4.0980000000000004e-05, "loss": 0.1683, "step": 2050 }, { "grad_norm": 1.8081207275390625, "learning_rate": 4.118e-05, "loss": 0.1846, "step": 2060 }, { "grad_norm": 1.5967745780944824, "learning_rate": 4.138e-05, "loss": 0.1697, "step": 2070 }, { "grad_norm": 1.761831521987915, "learning_rate": 4.1580000000000005e-05, "loss": 0.1691, "step": 2080 }, { "grad_norm": 1.4057347774505615, "learning_rate": 4.178e-05, "loss": 0.1983, "step": 2090 }, { "grad_norm": 1.5048596858978271, "learning_rate": 4.198e-05, "loss": 0.189, "step": 2100 }, { "grad_norm": 1.9168219566345215, "learning_rate": 4.2180000000000006e-05, "loss": 0.1802, "step": 2110 }, { "grad_norm": 1.7692962884902954, "learning_rate": 4.2380000000000004e-05, "loss": 0.2283, "step": 2120 }, { "grad_norm": 2.1260554790496826, "learning_rate": 4.258e-05, "loss": 0.2142, "step": 2130 }, { "grad_norm": 1.823845386505127, "learning_rate": 4.278e-05, "loss": 0.2074, "step": 2140 }, { "grad_norm": 1.9063102006912231, "learning_rate": 4.2980000000000005e-05, "loss": 0.2245, "step": 2150 }, { "grad_norm": 1.7605113983154297, "learning_rate": 4.318e-05, "loss": 0.1991, "step": 2160 }, { "grad_norm": 1.9768701791763306, "learning_rate": 4.338e-05, "loss": 0.188, "step": 2170 }, { "grad_norm": 1.4491922855377197, "learning_rate": 4.3580000000000006e-05, "loss": 0.1867, "step": 2180 }, { "grad_norm": 1.8857464790344238, "learning_rate": 4.3780000000000004e-05, "loss": 0.1538, "step": 2190 }, { "grad_norm": 1.5342018604278564, "learning_rate": 4.398e-05, "loss": 0.1594, "step": 2200 }, { "grad_norm": 1.563970923423767, "learning_rate": 4.418000000000001e-05, "loss": 0.2024, "step": 2210 }, { "grad_norm": 2.1911370754241943, "learning_rate": 4.438e-05, "loss": 0.2336, "step": 2220 }, { "grad_norm": 1.925463080406189, "learning_rate": 4.458e-05, "loss": 0.175, "step": 2230 }, { "grad_norm": 1.578927993774414, "learning_rate": 4.478e-05, "loss": 0.1417, "step": 2240 }, { "grad_norm": 1.8371862173080444, "learning_rate": 4.498e-05, "loss": 0.174, "step": 2250 }, { "grad_norm": 1.8293713331222534, "learning_rate": 4.518e-05, "loss": 0.2004, "step": 2260 }, { "grad_norm": 1.5329355001449585, "learning_rate": 4.538e-05, "loss": 0.2083, "step": 2270 }, { "grad_norm": 1.825599193572998, "learning_rate": 4.558e-05, "loss": 0.2013, "step": 2280 }, { "grad_norm": 1.7724578380584717, "learning_rate": 4.578e-05, "loss": 0.1852, "step": 2290 }, { "grad_norm": 1.4620435237884521, "learning_rate": 4.5980000000000004e-05, "loss": 0.2241, "step": 2300 }, { "grad_norm": 1.7978534698486328, "learning_rate": 4.618e-05, "loss": 0.1639, "step": 2310 }, { "grad_norm": 1.1567622423171997, "learning_rate": 4.638e-05, "loss": 0.1971, "step": 2320 }, { "grad_norm": 1.7834441661834717, "learning_rate": 4.6580000000000005e-05, "loss": 0.1566, "step": 2330 }, { "grad_norm": 1.6117660999298096, "learning_rate": 4.678e-05, "loss": 0.199, "step": 2340 }, { "grad_norm": 1.748262643814087, "learning_rate": 4.698e-05, "loss": 0.1658, "step": 2350 }, { "grad_norm": 2.4168825149536133, "learning_rate": 4.718e-05, "loss": 0.1639, "step": 2360 }, { "grad_norm": 1.9433375597000122, "learning_rate": 4.7380000000000004e-05, "loss": 0.2122, "step": 2370 }, { "grad_norm": 1.7732341289520264, "learning_rate": 4.758e-05, "loss": 0.1733, "step": 2380 }, { "grad_norm": 1.6695702075958252, "learning_rate": 4.778e-05, "loss": 0.1698, "step": 2390 }, { "grad_norm": 2.1211626529693604, "learning_rate": 4.7980000000000005e-05, "loss": 0.212, "step": 2400 }, { "grad_norm": 1.7896133661270142, "learning_rate": 4.818e-05, "loss": 0.232, "step": 2410 }, { "grad_norm": 2.844330310821533, "learning_rate": 4.838e-05, "loss": 0.1907, "step": 2420 }, { "grad_norm": 1.813476800918579, "learning_rate": 4.8580000000000006e-05, "loss": 0.1707, "step": 2430 }, { "grad_norm": 1.6307790279388428, "learning_rate": 4.8780000000000004e-05, "loss": 0.1837, "step": 2440 }, { "grad_norm": 1.6099331378936768, "learning_rate": 4.898e-05, "loss": 0.196, "step": 2450 }, { "grad_norm": 1.4627485275268555, "learning_rate": 4.918000000000001e-05, "loss": 0.1475, "step": 2460 }, { "grad_norm": 1.6647454500198364, "learning_rate": 4.9380000000000005e-05, "loss": 0.2161, "step": 2470 }, { "grad_norm": 1.7234411239624023, "learning_rate": 4.958e-05, "loss": 0.1603, "step": 2480 }, { "grad_norm": 1.3777894973754883, "learning_rate": 4.978e-05, "loss": 0.1746, "step": 2490 }, { "grad_norm": 1.5055820941925049, "learning_rate": 4.9980000000000006e-05, "loss": 0.1715, "step": 2500 }, { "grad_norm": 1.2411346435546875, "learning_rate": 5.0180000000000004e-05, "loss": 0.1948, "step": 2510 }, { "grad_norm": 1.327649712562561, "learning_rate": 5.038e-05, "loss": 0.1758, "step": 2520 }, { "grad_norm": 1.6146681308746338, "learning_rate": 5.058000000000001e-05, "loss": 0.1538, "step": 2530 }, { "grad_norm": 1.637036681175232, "learning_rate": 5.0780000000000005e-05, "loss": 0.1581, "step": 2540 }, { "grad_norm": 1.4311394691467285, "learning_rate": 5.098e-05, "loss": 0.186, "step": 2550 }, { "grad_norm": 1.7461345195770264, "learning_rate": 5.118000000000001e-05, "loss": 0.1881, "step": 2560 }, { "grad_norm": 1.4374252557754517, "learning_rate": 5.1380000000000006e-05, "loss": 0.2123, "step": 2570 }, { "grad_norm": 1.330580711364746, "learning_rate": 5.1580000000000004e-05, "loss": 0.1683, "step": 2580 }, { "grad_norm": 1.5097509622573853, "learning_rate": 5.178000000000001e-05, "loss": 0.1844, "step": 2590 }, { "grad_norm": 1.4273509979248047, "learning_rate": 5.198000000000001e-05, "loss": 0.1808, "step": 2600 }, { "grad_norm": 1.423295021057129, "learning_rate": 5.2180000000000005e-05, "loss": 0.1637, "step": 2610 }, { "grad_norm": 1.7052562236785889, "learning_rate": 5.238000000000001e-05, "loss": 0.1721, "step": 2620 }, { "grad_norm": 1.262094497680664, "learning_rate": 5.258000000000001e-05, "loss": 0.131, "step": 2630 }, { "grad_norm": 1.3316586017608643, "learning_rate": 5.2780000000000006e-05, "loss": 0.2054, "step": 2640 }, { "grad_norm": 1.4320149421691895, "learning_rate": 5.2980000000000004e-05, "loss": 0.1426, "step": 2650 }, { "grad_norm": 1.311105489730835, "learning_rate": 5.318000000000001e-05, "loss": 0.1507, "step": 2660 }, { "grad_norm": 1.7971792221069336, "learning_rate": 5.338000000000001e-05, "loss": 0.1464, "step": 2670 }, { "grad_norm": 1.1024998426437378, "learning_rate": 5.3580000000000005e-05, "loss": 0.1573, "step": 2680 }, { "grad_norm": 1.2363228797912598, "learning_rate": 5.378e-05, "loss": 0.1843, "step": 2690 }, { "grad_norm": 1.0134878158569336, "learning_rate": 5.3979999999999995e-05, "loss": 0.1963, "step": 2700 }, { "grad_norm": 1.6746656894683838, "learning_rate": 5.418e-05, "loss": 0.1799, "step": 2710 }, { "grad_norm": 1.761023998260498, "learning_rate": 5.438e-05, "loss": 0.1707, "step": 2720 }, { "grad_norm": 1.579875111579895, "learning_rate": 5.4579999999999996e-05, "loss": 0.15, "step": 2730 }, { "grad_norm": 1.371948480606079, "learning_rate": 5.478e-05, "loss": 0.1948, "step": 2740 }, { "grad_norm": 1.0666640996932983, "learning_rate": 5.498e-05, "loss": 0.1433, "step": 2750 }, { "grad_norm": 1.0300315618515015, "learning_rate": 5.518e-05, "loss": 0.1448, "step": 2760 }, { "grad_norm": 1.5343259572982788, "learning_rate": 5.538e-05, "loss": 0.1736, "step": 2770 }, { "grad_norm": 1.5649316310882568, "learning_rate": 5.558e-05, "loss": 0.2027, "step": 2780 }, { "grad_norm": 1.134095549583435, "learning_rate": 5.578e-05, "loss": 0.1447, "step": 2790 }, { "grad_norm": 1.4656163454055786, "learning_rate": 5.5979999999999996e-05, "loss": 0.1718, "step": 2800 }, { "grad_norm": 1.3107445240020752, "learning_rate": 5.618e-05, "loss": 0.1854, "step": 2810 }, { "grad_norm": 1.9199272394180298, "learning_rate": 5.638e-05, "loss": 0.1726, "step": 2820 }, { "grad_norm": 1.0706117153167725, "learning_rate": 5.658e-05, "loss": 0.1539, "step": 2830 }, { "grad_norm": 1.7643694877624512, "learning_rate": 5.678e-05, "loss": 0.1821, "step": 2840 }, { "grad_norm": 1.8026481866836548, "learning_rate": 5.698e-05, "loss": 0.1485, "step": 2850 }, { "grad_norm": 1.5425713062286377, "learning_rate": 5.718e-05, "loss": 0.1513, "step": 2860 }, { "grad_norm": 2.2407069206237793, "learning_rate": 5.738e-05, "loss": 0.1793, "step": 2870 }, { "grad_norm": 1.2791683673858643, "learning_rate": 5.758e-05, "loss": 0.1509, "step": 2880 }, { "grad_norm": 1.3382351398468018, "learning_rate": 5.778e-05, "loss": 0.2154, "step": 2890 }, { "grad_norm": 1.8914626836776733, "learning_rate": 5.7980000000000004e-05, "loss": 0.1652, "step": 2900 }, { "grad_norm": 0.9768365621566772, "learning_rate": 5.818e-05, "loss": 0.2084, "step": 2910 }, { "grad_norm": 1.118532657623291, "learning_rate": 5.838e-05, "loss": 0.146, "step": 2920 }, { "grad_norm": 1.2673780918121338, "learning_rate": 5.858e-05, "loss": 0.1739, "step": 2930 }, { "grad_norm": 1.4097108840942383, "learning_rate": 5.878e-05, "loss": 0.1853, "step": 2940 }, { "grad_norm": 1.465994954109192, "learning_rate": 5.898e-05, "loss": 0.1655, "step": 2950 }, { "grad_norm": 1.6712486743927002, "learning_rate": 5.918e-05, "loss": 0.1606, "step": 2960 }, { "grad_norm": 1.202526330947876, "learning_rate": 5.9380000000000004e-05, "loss": 0.1519, "step": 2970 }, { "grad_norm": 1.1506075859069824, "learning_rate": 5.958e-05, "loss": 0.1871, "step": 2980 }, { "grad_norm": 1.1076087951660156, "learning_rate": 5.978e-05, "loss": 0.1791, "step": 2990 }, { "grad_norm": 1.230447769165039, "learning_rate": 5.9980000000000005e-05, "loss": 0.1608, "step": 3000 }, { "grad_norm": 1.0246939659118652, "learning_rate": 6.018e-05, "loss": 0.1621, "step": 3010 }, { "grad_norm": 1.0347527265548706, "learning_rate": 6.038e-05, "loss": 0.1478, "step": 3020 }, { "grad_norm": 1.1781933307647705, "learning_rate": 6.0580000000000006e-05, "loss": 0.172, "step": 3030 }, { "grad_norm": 1.5311596393585205, "learning_rate": 6.0780000000000004e-05, "loss": 0.1822, "step": 3040 }, { "grad_norm": 1.6025725603103638, "learning_rate": 6.098e-05, "loss": 0.158, "step": 3050 }, { "grad_norm": 1.0310349464416504, "learning_rate": 6.118000000000001e-05, "loss": 0.2106, "step": 3060 }, { "grad_norm": 1.2412116527557373, "learning_rate": 6.138e-05, "loss": 0.1828, "step": 3070 }, { "grad_norm": 1.2308166027069092, "learning_rate": 6.158e-05, "loss": 0.1682, "step": 3080 }, { "grad_norm": 1.1965445280075073, "learning_rate": 6.178000000000001e-05, "loss": 0.1444, "step": 3090 }, { "grad_norm": 1.2821794748306274, "learning_rate": 6.198e-05, "loss": 0.1484, "step": 3100 }, { "grad_norm": 1.7308801412582397, "learning_rate": 6.218e-05, "loss": 0.1911, "step": 3110 }, { "grad_norm": 1.3271489143371582, "learning_rate": 6.238000000000001e-05, "loss": 0.1691, "step": 3120 }, { "grad_norm": 1.2411469221115112, "learning_rate": 6.258e-05, "loss": 0.144, "step": 3130 }, { "grad_norm": 1.0459628105163574, "learning_rate": 6.278e-05, "loss": 0.155, "step": 3140 }, { "grad_norm": 1.3057786226272583, "learning_rate": 6.298000000000001e-05, "loss": 0.1463, "step": 3150 }, { "grad_norm": 1.3008325099945068, "learning_rate": 6.318e-05, "loss": 0.1608, "step": 3160 }, { "grad_norm": 1.768121600151062, "learning_rate": 6.338e-05, "loss": 0.1614, "step": 3170 }, { "grad_norm": 1.4740065336227417, "learning_rate": 6.358000000000001e-05, "loss": 0.1784, "step": 3180 }, { "grad_norm": 1.1220133304595947, "learning_rate": 6.378e-05, "loss": 0.1455, "step": 3190 }, { "grad_norm": 1.2463538646697998, "learning_rate": 6.398000000000001e-05, "loss": 0.1431, "step": 3200 }, { "grad_norm": 1.3256611824035645, "learning_rate": 6.418000000000001e-05, "loss": 0.1344, "step": 3210 }, { "grad_norm": 1.1783268451690674, "learning_rate": 6.438e-05, "loss": 0.1459, "step": 3220 }, { "grad_norm": 1.413588285446167, "learning_rate": 6.458000000000001e-05, "loss": 0.1669, "step": 3230 }, { "grad_norm": 0.9773035645484924, "learning_rate": 6.478000000000001e-05, "loss": 0.1843, "step": 3240 }, { "grad_norm": 1.334753394126892, "learning_rate": 6.498e-05, "loss": 0.1677, "step": 3250 }, { "grad_norm": 1.2168118953704834, "learning_rate": 6.518000000000001e-05, "loss": 0.2005, "step": 3260 }, { "grad_norm": 1.4621636867523193, "learning_rate": 6.538000000000001e-05, "loss": 0.152, "step": 3270 }, { "grad_norm": 1.4570032358169556, "learning_rate": 6.558e-05, "loss": 0.1822, "step": 3280 }, { "grad_norm": 1.7833729982376099, "learning_rate": 6.578000000000001e-05, "loss": 0.1804, "step": 3290 }, { "grad_norm": 1.4840397834777832, "learning_rate": 6.598e-05, "loss": 0.1688, "step": 3300 }, { "grad_norm": 1.0332310199737549, "learning_rate": 6.618e-05, "loss": 0.1495, "step": 3310 }, { "grad_norm": 1.3835996389389038, "learning_rate": 6.638e-05, "loss": 0.1766, "step": 3320 }, { "grad_norm": 1.2760679721832275, "learning_rate": 6.658e-05, "loss": 0.1527, "step": 3330 }, { "grad_norm": 1.3417202234268188, "learning_rate": 6.678e-05, "loss": 0.1445, "step": 3340 }, { "grad_norm": 1.0973843336105347, "learning_rate": 6.698e-05, "loss": 0.1526, "step": 3350 }, { "grad_norm": 1.039154052734375, "learning_rate": 6.718e-05, "loss": 0.1668, "step": 3360 }, { "grad_norm": 0.9965974688529968, "learning_rate": 6.738e-05, "loss": 0.1583, "step": 3370 }, { "grad_norm": 1.2868777513504028, "learning_rate": 6.758e-05, "loss": 0.1482, "step": 3380 }, { "grad_norm": 1.0655351877212524, "learning_rate": 6.778e-05, "loss": 0.1546, "step": 3390 }, { "grad_norm": 1.0684820413589478, "learning_rate": 6.798e-05, "loss": 0.1711, "step": 3400 }, { "grad_norm": 1.437104344367981, "learning_rate": 6.818e-05, "loss": 0.1573, "step": 3410 }, { "grad_norm": 1.9106245040893555, "learning_rate": 6.838e-05, "loss": 0.204, "step": 3420 }, { "grad_norm": 1.0013025999069214, "learning_rate": 6.858e-05, "loss": 0.1746, "step": 3430 }, { "grad_norm": 1.4474050998687744, "learning_rate": 6.878e-05, "loss": 0.1901, "step": 3440 }, { "grad_norm": 1.234022617340088, "learning_rate": 6.898e-05, "loss": 0.1717, "step": 3450 }, { "grad_norm": 1.3721383810043335, "learning_rate": 6.918e-05, "loss": 0.1419, "step": 3460 }, { "grad_norm": 1.5615460872650146, "learning_rate": 6.938e-05, "loss": 0.1624, "step": 3470 }, { "grad_norm": 1.2918823957443237, "learning_rate": 6.958e-05, "loss": 0.1331, "step": 3480 }, { "grad_norm": 1.6937296390533447, "learning_rate": 6.978e-05, "loss": 0.1498, "step": 3490 }, { "grad_norm": 1.4963815212249756, "learning_rate": 6.998e-05, "loss": 0.1785, "step": 3500 }, { "grad_norm": 1.5047812461853027, "learning_rate": 7.018e-05, "loss": 0.1638, "step": 3510 }, { "grad_norm": 1.960795521736145, "learning_rate": 7.038e-05, "loss": 0.1717, "step": 3520 }, { "grad_norm": 1.193882942199707, "learning_rate": 7.058e-05, "loss": 0.1615, "step": 3530 }, { "grad_norm": 1.5679564476013184, "learning_rate": 7.078e-05, "loss": 0.1623, "step": 3540 }, { "grad_norm": 1.7234268188476562, "learning_rate": 7.098e-05, "loss": 0.1879, "step": 3550 }, { "grad_norm": 1.388874888420105, "learning_rate": 7.118e-05, "loss": 0.1844, "step": 3560 }, { "grad_norm": 1.1911578178405762, "learning_rate": 7.138e-05, "loss": 0.1491, "step": 3570 }, { "grad_norm": 1.4365423917770386, "learning_rate": 7.158e-05, "loss": 0.1726, "step": 3580 }, { "grad_norm": 1.4166555404663086, "learning_rate": 7.178000000000001e-05, "loss": 0.1702, "step": 3590 }, { "grad_norm": 1.0059136152267456, "learning_rate": 7.198e-05, "loss": 0.1552, "step": 3600 }, { "grad_norm": 1.4463592767715454, "learning_rate": 7.218e-05, "loss": 0.1717, "step": 3610 }, { "grad_norm": 0.7899258732795715, "learning_rate": 7.238000000000001e-05, "loss": 0.1608, "step": 3620 }, { "grad_norm": 1.220356822013855, "learning_rate": 7.258e-05, "loss": 0.1571, "step": 3630 }, { "grad_norm": 1.779317855834961, "learning_rate": 7.278e-05, "loss": 0.1827, "step": 3640 }, { "grad_norm": 0.8999063372612, "learning_rate": 7.298000000000001e-05, "loss": 0.1391, "step": 3650 }, { "grad_norm": 2.003122568130493, "learning_rate": 7.318e-05, "loss": 0.1964, "step": 3660 }, { "grad_norm": 1.410516619682312, "learning_rate": 7.338e-05, "loss": 0.1917, "step": 3670 }, { "grad_norm": 1.3287616968154907, "learning_rate": 7.358000000000001e-05, "loss": 0.1517, "step": 3680 }, { "grad_norm": 1.3219640254974365, "learning_rate": 7.378e-05, "loss": 0.2093, "step": 3690 }, { "grad_norm": 0.9244419932365417, "learning_rate": 7.398e-05, "loss": 0.1603, "step": 3700 }, { "grad_norm": 1.5125168561935425, "learning_rate": 7.418000000000001e-05, "loss": 0.1719, "step": 3710 }, { "grad_norm": 1.334072470664978, "learning_rate": 7.438e-05, "loss": 0.2206, "step": 3720 }, { "grad_norm": 1.3584039211273193, "learning_rate": 7.458000000000001e-05, "loss": 0.1519, "step": 3730 }, { "grad_norm": 0.8140402436256409, "learning_rate": 7.478e-05, "loss": 0.1604, "step": 3740 }, { "grad_norm": 1.1773486137390137, "learning_rate": 7.498e-05, "loss": 0.1447, "step": 3750 }, { "grad_norm": 1.1757171154022217, "learning_rate": 7.518000000000001e-05, "loss": 0.1446, "step": 3760 }, { "grad_norm": 1.3454327583312988, "learning_rate": 7.538e-05, "loss": 0.1664, "step": 3770 }, { "grad_norm": 1.4402409791946411, "learning_rate": 7.558e-05, "loss": 0.1374, "step": 3780 }, { "grad_norm": 1.2982863187789917, "learning_rate": 7.578000000000001e-05, "loss": 0.1571, "step": 3790 }, { "grad_norm": 0.9112211465835571, "learning_rate": 7.598e-05, "loss": 0.1471, "step": 3800 }, { "grad_norm": 1.0370618104934692, "learning_rate": 7.618e-05, "loss": 0.1572, "step": 3810 }, { "grad_norm": 1.4923752546310425, "learning_rate": 7.638000000000001e-05, "loss": 0.2015, "step": 3820 }, { "grad_norm": 1.5001946687698364, "learning_rate": 7.658e-05, "loss": 0.1663, "step": 3830 }, { "grad_norm": 0.9578271508216858, "learning_rate": 7.678000000000001e-05, "loss": 0.1592, "step": 3840 }, { "grad_norm": 1.2109283208847046, "learning_rate": 7.698000000000001e-05, "loss": 0.1472, "step": 3850 }, { "grad_norm": 1.2254716157913208, "learning_rate": 7.718e-05, "loss": 0.1429, "step": 3860 }, { "grad_norm": 1.0213395357131958, "learning_rate": 7.738000000000001e-05, "loss": 0.1554, "step": 3870 }, { "grad_norm": 1.4949971437454224, "learning_rate": 7.758000000000001e-05, "loss": 0.1726, "step": 3880 }, { "grad_norm": 0.9370759129524231, "learning_rate": 7.778e-05, "loss": 0.1568, "step": 3890 }, { "grad_norm": 0.9354826807975769, "learning_rate": 7.798000000000001e-05, "loss": 0.1232, "step": 3900 }, { "grad_norm": 0.8594309091567993, "learning_rate": 7.818000000000001e-05, "loss": 0.1668, "step": 3910 }, { "grad_norm": 1.071624994277954, "learning_rate": 7.838e-05, "loss": 0.184, "step": 3920 }, { "grad_norm": 1.0268324613571167, "learning_rate": 7.858000000000001e-05, "loss": 0.1219, "step": 3930 }, { "grad_norm": 2.005343198776245, "learning_rate": 7.878e-05, "loss": 0.154, "step": 3940 }, { "grad_norm": 1.0231436491012573, "learning_rate": 7.897999999999999e-05, "loss": 0.1529, "step": 3950 }, { "grad_norm": 1.0937352180480957, "learning_rate": 7.918e-05, "loss": 0.1906, "step": 3960 }, { "grad_norm": 1.1466468572616577, "learning_rate": 7.938e-05, "loss": 0.1482, "step": 3970 }, { "grad_norm": 1.3131920099258423, "learning_rate": 7.958e-05, "loss": 0.153, "step": 3980 }, { "grad_norm": 1.3712188005447388, "learning_rate": 7.978e-05, "loss": 0.1989, "step": 3990 }, { "grad_norm": 1.2049126625061035, "learning_rate": 7.998e-05, "loss": 0.1344, "step": 4000 }, { "grad_norm": 1.7069916725158691, "learning_rate": 8.018e-05, "loss": 0.2009, "step": 4010 }, { "grad_norm": 0.9849252700805664, "learning_rate": 8.038e-05, "loss": 0.1931, "step": 4020 }, { "grad_norm": 1.183049201965332, "learning_rate": 8.058e-05, "loss": 0.1359, "step": 4030 }, { "grad_norm": 0.9986336827278137, "learning_rate": 8.078e-05, "loss": 0.1658, "step": 4040 }, { "grad_norm": 1.1819989681243896, "learning_rate": 8.098e-05, "loss": 0.115, "step": 4050 }, { "grad_norm": 1.1661499738693237, "learning_rate": 8.118e-05, "loss": 0.1615, "step": 4060 }, { "grad_norm": 1.5022082328796387, "learning_rate": 8.138e-05, "loss": 0.1236, "step": 4070 }, { "grad_norm": 1.2467286586761475, "learning_rate": 8.158e-05, "loss": 0.2027, "step": 4080 }, { "grad_norm": 1.5378745794296265, "learning_rate": 8.178e-05, "loss": 0.1473, "step": 4090 }, { "grad_norm": 1.4121419191360474, "learning_rate": 8.198e-05, "loss": 0.1659, "step": 4100 }, { "grad_norm": 1.1808445453643799, "learning_rate": 8.218e-05, "loss": 0.134, "step": 4110 }, { "grad_norm": 0.8156430125236511, "learning_rate": 8.238000000000001e-05, "loss": 0.1553, "step": 4120 }, { "grad_norm": 1.2520188093185425, "learning_rate": 8.258e-05, "loss": 0.1576, "step": 4130 }, { "grad_norm": 1.1240702867507935, "learning_rate": 8.278e-05, "loss": 0.1576, "step": 4140 }, { "grad_norm": 1.2347127199172974, "learning_rate": 8.298000000000001e-05, "loss": 0.1526, "step": 4150 }, { "grad_norm": 1.1798745393753052, "learning_rate": 8.318e-05, "loss": 0.165, "step": 4160 }, { "grad_norm": 0.8989346623420715, "learning_rate": 8.338e-05, "loss": 0.1283, "step": 4170 }, { "grad_norm": 1.3072926998138428, "learning_rate": 8.358e-05, "loss": 0.1782, "step": 4180 }, { "grad_norm": 1.1346895694732666, "learning_rate": 8.378e-05, "loss": 0.1463, "step": 4190 }, { "grad_norm": 1.1478803157806396, "learning_rate": 8.398e-05, "loss": 0.1632, "step": 4200 }, { "grad_norm": 0.9917253851890564, "learning_rate": 8.418e-05, "loss": 0.1706, "step": 4210 }, { "grad_norm": 1.2503101825714111, "learning_rate": 8.438e-05, "loss": 0.2013, "step": 4220 }, { "grad_norm": 1.496206283569336, "learning_rate": 8.458e-05, "loss": 0.1611, "step": 4230 }, { "grad_norm": 1.3017332553863525, "learning_rate": 8.478e-05, "loss": 0.1578, "step": 4240 }, { "grad_norm": 0.769720733165741, "learning_rate": 8.498e-05, "loss": 0.1633, "step": 4250 }, { "grad_norm": 0.9450863599777222, "learning_rate": 8.518000000000001e-05, "loss": 0.1516, "step": 4260 }, { "grad_norm": 0.893190860748291, "learning_rate": 8.538e-05, "loss": 0.1256, "step": 4270 }, { "grad_norm": 0.9646791815757751, "learning_rate": 8.558e-05, "loss": 0.1395, "step": 4280 }, { "grad_norm": 0.9270627498626709, "learning_rate": 8.578000000000001e-05, "loss": 0.1802, "step": 4290 }, { "grad_norm": 1.2329161167144775, "learning_rate": 8.598e-05, "loss": 0.1628, "step": 4300 }, { "grad_norm": 1.3186681270599365, "learning_rate": 8.618e-05, "loss": 0.1872, "step": 4310 }, { "grad_norm": 1.135085940361023, "learning_rate": 8.638000000000001e-05, "loss": 0.1592, "step": 4320 }, { "grad_norm": 1.1887229681015015, "learning_rate": 8.658e-05, "loss": 0.1421, "step": 4330 }, { "grad_norm": 1.1730358600616455, "learning_rate": 8.678e-05, "loss": 0.1424, "step": 4340 }, { "grad_norm": 1.3606189489364624, "learning_rate": 8.698000000000001e-05, "loss": 0.1788, "step": 4350 }, { "grad_norm": 1.1684802770614624, "learning_rate": 8.718e-05, "loss": 0.1886, "step": 4360 }, { "grad_norm": 1.215255856513977, "learning_rate": 8.738000000000001e-05, "loss": 0.1675, "step": 4370 }, { "grad_norm": 1.1084082126617432, "learning_rate": 8.758000000000001e-05, "loss": 0.152, "step": 4380 }, { "grad_norm": 1.0562762022018433, "learning_rate": 8.778e-05, "loss": 0.1501, "step": 4390 }, { "grad_norm": 0.9628294110298157, "learning_rate": 8.798000000000001e-05, "loss": 0.1426, "step": 4400 }, { "grad_norm": 0.9839140772819519, "learning_rate": 8.818000000000001e-05, "loss": 0.1608, "step": 4410 }, { "grad_norm": 0.9281678795814514, "learning_rate": 8.838e-05, "loss": 0.1443, "step": 4420 }, { "grad_norm": 0.8492934107780457, "learning_rate": 8.858000000000001e-05, "loss": 0.1482, "step": 4430 }, { "grad_norm": 1.3261876106262207, "learning_rate": 8.878000000000001e-05, "loss": 0.1489, "step": 4440 }, { "grad_norm": 0.900682270526886, "learning_rate": 8.898e-05, "loss": 0.13, "step": 4450 }, { "grad_norm": 1.3449084758758545, "learning_rate": 8.918000000000001e-05, "loss": 0.1818, "step": 4460 }, { "grad_norm": 1.3589637279510498, "learning_rate": 8.938e-05, "loss": 0.1501, "step": 4470 }, { "grad_norm": 1.1185253858566284, "learning_rate": 8.958e-05, "loss": 0.1361, "step": 4480 }, { "grad_norm": 1.1264721155166626, "learning_rate": 8.978000000000001e-05, "loss": 0.1492, "step": 4490 }, { "grad_norm": 1.1675959825515747, "learning_rate": 8.998e-05, "loss": 0.179, "step": 4500 }, { "grad_norm": 1.1882388591766357, "learning_rate": 9.018000000000001e-05, "loss": 0.1696, "step": 4510 }, { "grad_norm": 1.2302823066711426, "learning_rate": 9.038000000000001e-05, "loss": 0.1504, "step": 4520 }, { "grad_norm": 0.9393045902252197, "learning_rate": 9.058e-05, "loss": 0.1529, "step": 4530 }, { "grad_norm": 0.8356147408485413, "learning_rate": 9.078000000000001e-05, "loss": 0.1546, "step": 4540 }, { "grad_norm": 0.8099928498268127, "learning_rate": 9.098000000000001e-05, "loss": 0.1483, "step": 4550 }, { "grad_norm": 1.1101744174957275, "learning_rate": 9.118e-05, "loss": 0.1593, "step": 4560 }, { "grad_norm": 0.7569586038589478, "learning_rate": 9.138e-05, "loss": 0.1683, "step": 4570 }, { "grad_norm": 1.5757242441177368, "learning_rate": 9.158e-05, "loss": 0.1894, "step": 4580 }, { "grad_norm": 0.820000171661377, "learning_rate": 9.178e-05, "loss": 0.1464, "step": 4590 }, { "grad_norm": 0.9819680452346802, "learning_rate": 9.198e-05, "loss": 0.1487, "step": 4600 }, { "grad_norm": 1.27237069606781, "learning_rate": 9.218e-05, "loss": 0.1614, "step": 4610 }, { "grad_norm": 1.0018446445465088, "learning_rate": 9.238e-05, "loss": 0.1399, "step": 4620 }, { "grad_norm": 1.4350857734680176, "learning_rate": 9.258e-05, "loss": 0.1308, "step": 4630 }, { "grad_norm": 1.5398279428482056, "learning_rate": 9.278e-05, "loss": 0.1402, "step": 4640 }, { "grad_norm": 1.4775776863098145, "learning_rate": 9.298e-05, "loss": 0.1564, "step": 4650 }, { "grad_norm": 1.1562808752059937, "learning_rate": 9.318e-05, "loss": 0.1624, "step": 4660 }, { "grad_norm": 1.0243970155715942, "learning_rate": 9.338e-05, "loss": 0.1503, "step": 4670 }, { "grad_norm": 1.0430004596710205, "learning_rate": 9.358e-05, "loss": 0.1459, "step": 4680 }, { "grad_norm": 0.6623867154121399, "learning_rate": 9.378e-05, "loss": 0.1879, "step": 4690 }, { "grad_norm": 0.9096137285232544, "learning_rate": 9.398e-05, "loss": 0.1438, "step": 4700 }, { "grad_norm": 1.1110996007919312, "learning_rate": 9.418e-05, "loss": 0.1507, "step": 4710 }, { "grad_norm": 1.0748286247253418, "learning_rate": 9.438e-05, "loss": 0.1761, "step": 4720 }, { "grad_norm": 0.7281737327575684, "learning_rate": 9.458e-05, "loss": 0.1606, "step": 4730 }, { "grad_norm": 0.8584475517272949, "learning_rate": 9.478e-05, "loss": 0.1392, "step": 4740 }, { "grad_norm": 1.1682658195495605, "learning_rate": 9.498e-05, "loss": 0.1735, "step": 4750 }, { "grad_norm": 1.2095717191696167, "learning_rate": 9.518000000000001e-05, "loss": 0.1729, "step": 4760 }, { "grad_norm": 1.0367568731307983, "learning_rate": 9.538e-05, "loss": 0.164, "step": 4770 }, { "grad_norm": 1.1720809936523438, "learning_rate": 9.558e-05, "loss": 0.1747, "step": 4780 }, { "grad_norm": 1.2915488481521606, "learning_rate": 9.578000000000001e-05, "loss": 0.1705, "step": 4790 }, { "grad_norm": 1.1261248588562012, "learning_rate": 9.598e-05, "loss": 0.1413, "step": 4800 }, { "grad_norm": 0.7206113338470459, "learning_rate": 9.618e-05, "loss": 0.147, "step": 4810 }, { "grad_norm": 1.0400115251541138, "learning_rate": 9.638000000000001e-05, "loss": 0.1537, "step": 4820 }, { "grad_norm": 1.0907080173492432, "learning_rate": 9.658e-05, "loss": 0.2282, "step": 4830 }, { "grad_norm": 1.032419204711914, "learning_rate": 9.678e-05, "loss": 0.1671, "step": 4840 }, { "grad_norm": 1.6597310304641724, "learning_rate": 9.698000000000001e-05, "loss": 0.1514, "step": 4850 }, { "grad_norm": 0.9939576983451843, "learning_rate": 9.718e-05, "loss": 0.1534, "step": 4860 }, { "grad_norm": 1.075304388999939, "learning_rate": 9.738e-05, "loss": 0.1781, "step": 4870 }, { "grad_norm": 0.8459386825561523, "learning_rate": 9.758000000000001e-05, "loss": 0.1581, "step": 4880 }, { "grad_norm": 1.1606593132019043, "learning_rate": 9.778e-05, "loss": 0.1458, "step": 4890 }, { "grad_norm": 1.1434354782104492, "learning_rate": 9.798000000000001e-05, "loss": 0.1529, "step": 4900 }, { "grad_norm": 0.9275431036949158, "learning_rate": 9.818000000000001e-05, "loss": 0.1561, "step": 4910 }, { "grad_norm": 1.0689730644226074, "learning_rate": 9.838e-05, "loss": 0.1399, "step": 4920 }, { "grad_norm": 1.1134532690048218, "learning_rate": 9.858000000000001e-05, "loss": 0.149, "step": 4930 }, { "grad_norm": 0.8319103717803955, "learning_rate": 9.878e-05, "loss": 0.1358, "step": 4940 }, { "grad_norm": 0.8945499062538147, "learning_rate": 9.898e-05, "loss": 0.1542, "step": 4950 }, { "grad_norm": 1.1195900440216064, "learning_rate": 9.918000000000001e-05, "loss": 0.1844, "step": 4960 }, { "grad_norm": 1.254982829093933, "learning_rate": 9.938e-05, "loss": 0.1463, "step": 4970 }, { "grad_norm": 1.2328035831451416, "learning_rate": 9.958e-05, "loss": 0.1472, "step": 4980 }, { "grad_norm": 0.870928168296814, "learning_rate": 9.978000000000001e-05, "loss": 0.1327, "step": 4990 }, { "grad_norm": 0.9403820633888245, "learning_rate": 9.998e-05, "loss": 0.122, "step": 5000 }, { "grad_norm": 1.0019904375076294, "learning_rate": 9.999999778549045e-05, "loss": 0.1297, "step": 5010 }, { "grad_norm": 1.2080109119415283, "learning_rate": 9.999999013039593e-05, "loss": 0.1418, "step": 5020 }, { "grad_norm": 1.0325815677642822, "learning_rate": 9.999997700737766e-05, "loss": 0.1609, "step": 5030 }, { "grad_norm": 1.690227746963501, "learning_rate": 9.999995841643709e-05, "loss": 0.155, "step": 5040 }, { "grad_norm": 1.0926035642623901, "learning_rate": 9.999993435757623e-05, "loss": 0.1521, "step": 5050 }, { "grad_norm": 0.9352501630783081, "learning_rate": 9.999990483079773e-05, "loss": 0.1314, "step": 5060 }, { "grad_norm": 1.3137941360473633, "learning_rate": 9.999986983610481e-05, "loss": 0.1591, "step": 5070 }, { "grad_norm": 1.0324677228927612, "learning_rate": 9.99998293735013e-05, "loss": 0.1399, "step": 5080 }, { "grad_norm": 0.936154842376709, "learning_rate": 9.999978344299161e-05, "loss": 0.2008, "step": 5090 }, { "grad_norm": 0.9394804239273071, "learning_rate": 9.99997320445808e-05, "loss": 0.1295, "step": 5100 }, { "grad_norm": 0.9066133499145508, "learning_rate": 9.999967517827444e-05, "loss": 0.171, "step": 5110 }, { "grad_norm": 1.2529761791229248, "learning_rate": 9.999961284407879e-05, "loss": 0.1567, "step": 5120 }, { "grad_norm": 1.220461368560791, "learning_rate": 9.999954504200067e-05, "loss": 0.1367, "step": 5130 }, { "grad_norm": 0.9663711786270142, "learning_rate": 9.999947177204744e-05, "loss": 0.1658, "step": 5140 }, { "grad_norm": 1.0384024381637573, "learning_rate": 9.999939303422718e-05, "loss": 0.1597, "step": 5150 }, { "grad_norm": 1.099417805671692, "learning_rate": 9.999930882854847e-05, "loss": 0.1542, "step": 5160 }, { "grad_norm": 1.010582447052002, "learning_rate": 9.999921915502051e-05, "loss": 0.1577, "step": 5170 }, { "grad_norm": 0.9573379158973694, "learning_rate": 9.99991240136531e-05, "loss": 0.1362, "step": 5180 }, { "grad_norm": 1.0766419172286987, "learning_rate": 9.999902340445668e-05, "loss": 0.1754, "step": 5190 }, { "grad_norm": 1.0932250022888184, "learning_rate": 9.999891732744224e-05, "loss": 0.1314, "step": 5200 }, { "grad_norm": 1.0141373872756958, "learning_rate": 9.999880578262135e-05, "loss": 0.1343, "step": 5210 }, { "grad_norm": 0.8893674612045288, "learning_rate": 9.999868877000624e-05, "loss": 0.1552, "step": 5220 }, { "grad_norm": 0.9639114737510681, "learning_rate": 9.99985662896097e-05, "loss": 0.1459, "step": 5230 }, { "grad_norm": 0.8453996181488037, "learning_rate": 9.999843834144513e-05, "loss": 0.135, "step": 5240 }, { "grad_norm": 0.9926575422286987, "learning_rate": 9.99983049255265e-05, "loss": 0.1241, "step": 5250 }, { "grad_norm": 0.9374807476997375, "learning_rate": 9.999816604186843e-05, "loss": 0.1471, "step": 5260 }, { "grad_norm": 1.2511835098266602, "learning_rate": 9.999802169048609e-05, "loss": 0.1584, "step": 5270 }, { "grad_norm": 1.027848243713379, "learning_rate": 9.999787187139527e-05, "loss": 0.1608, "step": 5280 }, { "grad_norm": 1.791734218597412, "learning_rate": 9.999771658461234e-05, "loss": 0.1692, "step": 5290 }, { "grad_norm": 1.1778756380081177, "learning_rate": 9.999755583015431e-05, "loss": 0.146, "step": 5300 }, { "grad_norm": 0.7394344210624695, "learning_rate": 9.999738960803874e-05, "loss": 0.1577, "step": 5310 }, { "grad_norm": 1.1809402704238892, "learning_rate": 9.99972179182838e-05, "loss": 0.1654, "step": 5320 }, { "grad_norm": 1.3563941717147827, "learning_rate": 9.99970407609083e-05, "loss": 0.1807, "step": 5330 }, { "grad_norm": 1.4944872856140137, "learning_rate": 9.999685813593159e-05, "loss": 0.1872, "step": 5340 }, { "grad_norm": 0.7708336710929871, "learning_rate": 9.999667004337362e-05, "loss": 0.1752, "step": 5350 }, { "grad_norm": 1.1358566284179688, "learning_rate": 9.9996476483255e-05, "loss": 0.1867, "step": 5360 }, { "grad_norm": 0.9001571536064148, "learning_rate": 9.999627745559688e-05, "loss": 0.1584, "step": 5370 }, { "grad_norm": 0.7894822955131531, "learning_rate": 9.999607296042101e-05, "loss": 0.1464, "step": 5380 }, { "grad_norm": 1.0933775901794434, "learning_rate": 9.99958629977498e-05, "loss": 0.1646, "step": 5390 }, { "grad_norm": 1.221946358680725, "learning_rate": 9.999564756760615e-05, "loss": 0.1524, "step": 5400 }, { "grad_norm": 1.0180553197860718, "learning_rate": 9.999542667001366e-05, "loss": 0.1435, "step": 5410 }, { "grad_norm": 0.9368793964385986, "learning_rate": 9.999520030499647e-05, "loss": 0.1524, "step": 5420 }, { "grad_norm": 1.1239811182022095, "learning_rate": 9.999496847257936e-05, "loss": 0.1406, "step": 5430 }, { "grad_norm": 0.8192836046218872, "learning_rate": 9.999473117278764e-05, "loss": 0.1787, "step": 5440 }, { "grad_norm": 0.9946401119232178, "learning_rate": 9.999448840564731e-05, "loss": 0.1675, "step": 5450 }, { "grad_norm": 1.225852370262146, "learning_rate": 9.999424017118488e-05, "loss": 0.1543, "step": 5460 }, { "grad_norm": 1.1775858402252197, "learning_rate": 9.999398646942751e-05, "loss": 0.1318, "step": 5470 }, { "grad_norm": 0.8133357763290405, "learning_rate": 9.999372730040296e-05, "loss": 0.1297, "step": 5480 }, { "grad_norm": 1.0658870935440063, "learning_rate": 9.999346266413953e-05, "loss": 0.1456, "step": 5490 }, { "grad_norm": 1.2661042213439941, "learning_rate": 9.99931925606662e-05, "loss": 0.1485, "step": 5500 }, { "grad_norm": 1.0346708297729492, "learning_rate": 9.99929169900125e-05, "loss": 0.1401, "step": 5510 }, { "grad_norm": 1.2552354335784912, "learning_rate": 9.999263595220855e-05, "loss": 0.1438, "step": 5520 }, { "grad_norm": 0.9892277717590332, "learning_rate": 9.99923494472851e-05, "loss": 0.1353, "step": 5530 }, { "grad_norm": 1.316953420639038, "learning_rate": 9.999205747527348e-05, "loss": 0.1609, "step": 5540 }, { "grad_norm": 1.3832358121871948, "learning_rate": 9.999176003620561e-05, "loss": 0.1741, "step": 5550 }, { "grad_norm": 0.8655797243118286, "learning_rate": 9.999145713011405e-05, "loss": 0.1564, "step": 5560 }, { "grad_norm": 0.9292653799057007, "learning_rate": 9.999114875703186e-05, "loss": 0.1488, "step": 5570 }, { "grad_norm": 1.093984603881836, "learning_rate": 9.999083491699281e-05, "loss": 0.1546, "step": 5580 }, { "grad_norm": 1.021715521812439, "learning_rate": 9.999051561003123e-05, "loss": 0.1353, "step": 5590 }, { "grad_norm": 1.001710057258606, "learning_rate": 9.999019083618202e-05, "loss": 0.1517, "step": 5600 }, { "grad_norm": 1.1143486499786377, "learning_rate": 9.99898605954807e-05, "loss": 0.1527, "step": 5610 }, { "grad_norm": 1.0707193613052368, "learning_rate": 9.998952488796338e-05, "loss": 0.1942, "step": 5620 }, { "grad_norm": 0.9530632495880127, "learning_rate": 9.998918371366676e-05, "loss": 0.1594, "step": 5630 }, { "grad_norm": 1.241216778755188, "learning_rate": 9.99888370726282e-05, "loss": 0.1539, "step": 5640 }, { "grad_norm": 1.1965723037719727, "learning_rate": 9.998848496488556e-05, "loss": 0.1401, "step": 5650 }, { "grad_norm": 1.208155632019043, "learning_rate": 9.998812739047736e-05, "loss": 0.1802, "step": 5660 }, { "grad_norm": 1.0846831798553467, "learning_rate": 9.99877643494427e-05, "loss": 0.1673, "step": 5670 }, { "grad_norm": 1.0572212934494019, "learning_rate": 9.998739584182128e-05, "loss": 0.1725, "step": 5680 }, { "grad_norm": 1.205645203590393, "learning_rate": 9.998702186765342e-05, "loss": 0.1874, "step": 5690 }, { "grad_norm": 1.1166167259216309, "learning_rate": 9.998664242698e-05, "loss": 0.1552, "step": 5700 }, { "grad_norm": 1.1210567951202393, "learning_rate": 9.998625751984251e-05, "loss": 0.1295, "step": 5710 }, { "grad_norm": 1.43648099899292, "learning_rate": 9.998586714628307e-05, "loss": 0.1495, "step": 5720 }, { "grad_norm": 0.9181079864501953, "learning_rate": 9.998547130634432e-05, "loss": 0.1544, "step": 5730 }, { "grad_norm": 0.9851597547531128, "learning_rate": 9.99850700000696e-05, "loss": 0.133, "step": 5740 }, { "grad_norm": 1.039941668510437, "learning_rate": 9.998466322750278e-05, "loss": 0.1388, "step": 5750 }, { "grad_norm": 1.1226661205291748, "learning_rate": 9.998425098868834e-05, "loss": 0.1181, "step": 5760 }, { "grad_norm": 0.9138882756233215, "learning_rate": 9.998383328367136e-05, "loss": 0.1554, "step": 5770 }, { "grad_norm": 0.8415870666503906, "learning_rate": 9.99834101124975e-05, "loss": 0.1423, "step": 5780 }, { "grad_norm": 0.9686135649681091, "learning_rate": 9.998298147521309e-05, "loss": 0.1241, "step": 5790 }, { "grad_norm": 0.961361825466156, "learning_rate": 9.998254737186496e-05, "loss": 0.1584, "step": 5800 }, { "grad_norm": 1.2501837015151978, "learning_rate": 9.99821078025006e-05, "loss": 0.1323, "step": 5810 }, { "grad_norm": 1.0210368633270264, "learning_rate": 9.998166276716807e-05, "loss": 0.1335, "step": 5820 }, { "grad_norm": 0.9968952536582947, "learning_rate": 9.998121226591606e-05, "loss": 0.1551, "step": 5830 }, { "grad_norm": 1.1637003421783447, "learning_rate": 9.998075629879382e-05, "loss": 0.1368, "step": 5840 }, { "grad_norm": 0.9901126027107239, "learning_rate": 9.99802948658512e-05, "loss": 0.1206, "step": 5850 }, { "grad_norm": 0.9167219400405884, "learning_rate": 9.99798279671387e-05, "loss": 0.118, "step": 5860 }, { "grad_norm": 1.003381609916687, "learning_rate": 9.997935560270734e-05, "loss": 0.126, "step": 5870 }, { "grad_norm": 1.025883674621582, "learning_rate": 9.997887777260879e-05, "loss": 0.118, "step": 5880 }, { "grad_norm": 1.1779203414916992, "learning_rate": 9.997839447689532e-05, "loss": 0.1223, "step": 5890 }, { "grad_norm": 0.8557911515235901, "learning_rate": 9.997790571561978e-05, "loss": 0.1262, "step": 5900 }, { "grad_norm": 0.842242419719696, "learning_rate": 9.99774114888356e-05, "loss": 0.1529, "step": 5910 }, { "grad_norm": 0.7897462844848633, "learning_rate": 9.997691179659684e-05, "loss": 0.1139, "step": 5920 }, { "grad_norm": 0.789458155632019, "learning_rate": 9.997640663895815e-05, "loss": 0.0991, "step": 5930 }, { "grad_norm": 0.8673627972602844, "learning_rate": 9.997589601597477e-05, "loss": 0.1325, "step": 5940 }, { "grad_norm": 0.6666049361228943, "learning_rate": 9.997537992770252e-05, "loss": 0.1178, "step": 5950 }, { "grad_norm": 1.0037775039672852, "learning_rate": 9.997485837419788e-05, "loss": 0.1316, "step": 5960 }, { "grad_norm": 1.8260738849639893, "learning_rate": 9.997433135551786e-05, "loss": 0.1209, "step": 5970 }, { "grad_norm": 0.9506875872612, "learning_rate": 9.997379887172009e-05, "loss": 0.1734, "step": 5980 }, { "grad_norm": 0.9832735061645508, "learning_rate": 9.997326092286281e-05, "loss": 0.1805, "step": 5990 }, { "grad_norm": 0.9315097332000732, "learning_rate": 9.997271750900486e-05, "loss": 0.1238, "step": 6000 }, { "grad_norm": 0.751462996006012, "learning_rate": 9.997216863020565e-05, "loss": 0.1207, "step": 6010 }, { "grad_norm": 0.9272622466087341, "learning_rate": 9.99716142865252e-05, "loss": 0.1301, "step": 6020 }, { "grad_norm": 0.7624633312225342, "learning_rate": 9.997105447802415e-05, "loss": 0.1431, "step": 6030 }, { "grad_norm": 0.721828281879425, "learning_rate": 9.997048920476373e-05, "loss": 0.1132, "step": 6040 }, { "grad_norm": 0.9278378486633301, "learning_rate": 9.996991846680572e-05, "loss": 0.1491, "step": 6050 }, { "grad_norm": 0.7210906147956848, "learning_rate": 9.996934226421257e-05, "loss": 0.1401, "step": 6060 }, { "grad_norm": 0.6985068917274475, "learning_rate": 9.996876059704726e-05, "loss": 0.1303, "step": 6070 }, { "grad_norm": 1.4394667148590088, "learning_rate": 9.996817346537343e-05, "loss": 0.1526, "step": 6080 }, { "grad_norm": 0.9238962531089783, "learning_rate": 9.996758086925526e-05, "loss": 0.122, "step": 6090 }, { "grad_norm": 0.7858859896659851, "learning_rate": 9.996698280875759e-05, "loss": 0.1445, "step": 6100 }, { "grad_norm": 1.0383386611938477, "learning_rate": 9.99663792839458e-05, "loss": 0.1349, "step": 6110 }, { "grad_norm": 0.8486121892929077, "learning_rate": 9.99657702948859e-05, "loss": 0.1256, "step": 6120 }, { "grad_norm": 0.7430689334869385, "learning_rate": 9.996515584164448e-05, "loss": 0.113, "step": 6130 }, { "grad_norm": 0.9288655519485474, "learning_rate": 9.996453592428873e-05, "loss": 0.1293, "step": 6140 }, { "grad_norm": 0.868072509765625, "learning_rate": 9.996391054288646e-05, "loss": 0.1179, "step": 6150 }, { "grad_norm": 0.8938488364219666, "learning_rate": 9.996327969750605e-05, "loss": 0.1424, "step": 6160 }, { "grad_norm": 0.9442999958992004, "learning_rate": 9.996264338821649e-05, "loss": 0.1478, "step": 6170 }, { "grad_norm": 0.8771108388900757, "learning_rate": 9.996200161508735e-05, "loss": 0.1632, "step": 6180 }, { "grad_norm": 0.7808802127838135, "learning_rate": 9.996135437818885e-05, "loss": 0.1275, "step": 6190 }, { "grad_norm": 0.8146441578865051, "learning_rate": 9.996070167759175e-05, "loss": 0.1262, "step": 6200 }, { "grad_norm": 1.1473326683044434, "learning_rate": 9.996004351336743e-05, "loss": 0.1176, "step": 6210 }, { "grad_norm": 0.9273423552513123, "learning_rate": 9.995937988558785e-05, "loss": 0.1147, "step": 6220 }, { "grad_norm": 0.9426174163818359, "learning_rate": 9.995871079432561e-05, "loss": 0.1185, "step": 6230 }, { "grad_norm": 0.9078529477119446, "learning_rate": 9.995803623965389e-05, "loss": 0.1018, "step": 6240 }, { "grad_norm": 0.8676959872245789, "learning_rate": 9.995735622164641e-05, "loss": 0.1258, "step": 6250 }, { "grad_norm": 1.3232240676879883, "learning_rate": 9.995667074037758e-05, "loss": 0.1539, "step": 6260 }, { "grad_norm": 1.0557873249053955, "learning_rate": 9.995597979592232e-05, "loss": 0.1331, "step": 6270 }, { "grad_norm": 1.0973550081253052, "learning_rate": 9.995528338835625e-05, "loss": 0.1477, "step": 6280 }, { "grad_norm": 1.602622151374817, "learning_rate": 9.995458151775547e-05, "loss": 0.1429, "step": 6290 }, { "grad_norm": 1.7283166646957397, "learning_rate": 9.995387418419677e-05, "loss": 0.1473, "step": 6300 }, { "grad_norm": 0.8297176361083984, "learning_rate": 9.99531613877575e-05, "loss": 0.1275, "step": 6310 }, { "grad_norm": 1.3952515125274658, "learning_rate": 9.995244312851559e-05, "loss": 0.1454, "step": 6320 }, { "grad_norm": 0.7226715087890625, "learning_rate": 9.995171940654961e-05, "loss": 0.1267, "step": 6330 }, { "grad_norm": 0.8705762028694153, "learning_rate": 9.995099022193871e-05, "loss": 0.1222, "step": 6340 }, { "grad_norm": 0.9036403894424438, "learning_rate": 9.995025557476261e-05, "loss": 0.1425, "step": 6350 }, { "grad_norm": 0.8775654435157776, "learning_rate": 9.994951546510165e-05, "loss": 0.1555, "step": 6360 }, { "grad_norm": 0.8087652325630188, "learning_rate": 9.994876989303679e-05, "loss": 0.1506, "step": 6370 }, { "grad_norm": 1.1848372220993042, "learning_rate": 9.994801885864955e-05, "loss": 0.1427, "step": 6380 }, { "grad_norm": 0.9206727147102356, "learning_rate": 9.994726236202205e-05, "loss": 0.1297, "step": 6390 }, { "grad_norm": 0.8685625791549683, "learning_rate": 9.994650040323704e-05, "loss": 0.1427, "step": 6400 }, { "grad_norm": 0.8497742414474487, "learning_rate": 9.994573298237784e-05, "loss": 0.1325, "step": 6410 }, { "grad_norm": 0.7926694750785828, "learning_rate": 9.994496009952837e-05, "loss": 0.1223, "step": 6420 }, { "grad_norm": 1.3108410835266113, "learning_rate": 9.994418175477316e-05, "loss": 0.1523, "step": 6430 }, { "grad_norm": 0.7679789662361145, "learning_rate": 9.994339794819733e-05, "loss": 0.1646, "step": 6440 }, { "grad_norm": 0.6809723377227783, "learning_rate": 9.994260867988658e-05, "loss": 0.1433, "step": 6450 }, { "grad_norm": 0.8946380615234375, "learning_rate": 9.994181394992723e-05, "loss": 0.1257, "step": 6460 }, { "grad_norm": 1.0296484231948853, "learning_rate": 9.994101375840618e-05, "loss": 0.1469, "step": 6470 }, { "grad_norm": 0.7981522679328918, "learning_rate": 9.994020810541098e-05, "loss": 0.1208, "step": 6480 }, { "grad_norm": 0.8506094217300415, "learning_rate": 9.99393969910297e-05, "loss": 0.1092, "step": 6490 }, { "grad_norm": 0.9478693008422852, "learning_rate": 9.993858041535104e-05, "loss": 0.1546, "step": 6500 }, { "grad_norm": 0.7474284768104553, "learning_rate": 9.99377583784643e-05, "loss": 0.1295, "step": 6510 }, { "grad_norm": 0.876118004322052, "learning_rate": 9.993693088045939e-05, "loss": 0.1312, "step": 6520 }, { "grad_norm": 0.9267895221710205, "learning_rate": 9.99360979214268e-05, "loss": 0.1395, "step": 6530 }, { "grad_norm": 1.0178271532058716, "learning_rate": 9.99352595014576e-05, "loss": 0.1359, "step": 6540 }, { "grad_norm": 1.167556881904602, "learning_rate": 9.993441562064354e-05, "loss": 0.1405, "step": 6550 }, { "grad_norm": 0.8275800347328186, "learning_rate": 9.993356627907685e-05, "loss": 0.1242, "step": 6560 }, { "grad_norm": 0.7834093570709229, "learning_rate": 9.99327114768504e-05, "loss": 0.1183, "step": 6570 }, { "grad_norm": 0.9138591289520264, "learning_rate": 9.99318512140577e-05, "loss": 0.1254, "step": 6580 }, { "grad_norm": 0.8003849983215332, "learning_rate": 9.993098549079284e-05, "loss": 0.1223, "step": 6590 }, { "grad_norm": 0.7128949761390686, "learning_rate": 9.993011430715047e-05, "loss": 0.1101, "step": 6600 }, { "grad_norm": 0.8169552683830261, "learning_rate": 9.992923766322586e-05, "loss": 0.1493, "step": 6610 }, { "grad_norm": 0.6897605657577515, "learning_rate": 9.99283555591149e-05, "loss": 0.1268, "step": 6620 }, { "grad_norm": 0.885418713092804, "learning_rate": 9.992746799491404e-05, "loss": 0.1381, "step": 6630 }, { "grad_norm": 0.9048051238059998, "learning_rate": 9.992657497072033e-05, "loss": 0.1484, "step": 6640 }, { "grad_norm": 1.067857027053833, "learning_rate": 9.992567648663147e-05, "loss": 0.1321, "step": 6650 }, { "grad_norm": 0.9048311114311218, "learning_rate": 9.992477254274568e-05, "loss": 0.1291, "step": 6660 }, { "grad_norm": 1.0096852779388428, "learning_rate": 9.992386313916183e-05, "loss": 0.1254, "step": 6670 }, { "grad_norm": 1.0372956991195679, "learning_rate": 9.992294827597934e-05, "loss": 0.1288, "step": 6680 }, { "grad_norm": 0.8786035776138306, "learning_rate": 9.992202795329831e-05, "loss": 0.1317, "step": 6690 }, { "grad_norm": 1.0538312196731567, "learning_rate": 9.992110217121936e-05, "loss": 0.1172, "step": 6700 }, { "grad_norm": 0.9533098936080933, "learning_rate": 9.992017092984372e-05, "loss": 0.1165, "step": 6710 }, { "grad_norm": 0.7246508598327637, "learning_rate": 9.991923422927326e-05, "loss": 0.1249, "step": 6720 }, { "grad_norm": 0.820712149143219, "learning_rate": 9.991829206961037e-05, "loss": 0.1205, "step": 6730 }, { "grad_norm": 0.7651575803756714, "learning_rate": 9.991734445095813e-05, "loss": 0.1311, "step": 6740 }, { "grad_norm": 1.1023621559143066, "learning_rate": 9.991639137342015e-05, "loss": 0.1667, "step": 6750 }, { "grad_norm": 1.0471198558807373, "learning_rate": 9.991543283710064e-05, "loss": 0.1783, "step": 6760 }, { "grad_norm": 1.047382116317749, "learning_rate": 9.991446884210445e-05, "loss": 0.1306, "step": 6770 }, { "grad_norm": 0.6033477783203125, "learning_rate": 9.9913499388537e-05, "loss": 0.1088, "step": 6780 }, { "grad_norm": 0.7281407117843628, "learning_rate": 9.99125244765043e-05, "loss": 0.1422, "step": 6790 }, { "grad_norm": 0.9299976229667664, "learning_rate": 9.991154410611296e-05, "loss": 0.1227, "step": 6800 }, { "grad_norm": 0.9512494206428528, "learning_rate": 9.99105582774702e-05, "loss": 0.1291, "step": 6810 }, { "grad_norm": 0.7498908042907715, "learning_rate": 9.990956699068384e-05, "loss": 0.1157, "step": 6820 }, { "grad_norm": 0.5994188785552979, "learning_rate": 9.990857024586224e-05, "loss": 0.1146, "step": 6830 }, { "grad_norm": 0.8667819499969482, "learning_rate": 9.990756804311446e-05, "loss": 0.1341, "step": 6840 }, { "grad_norm": 0.8136551976203918, "learning_rate": 9.990656038255006e-05, "loss": 0.1102, "step": 6850 }, { "grad_norm": 1.1777822971343994, "learning_rate": 9.990554726427926e-05, "loss": 0.1477, "step": 6860 }, { "grad_norm": 0.779658854007721, "learning_rate": 9.990452868841284e-05, "loss": 0.1062, "step": 6870 }, { "grad_norm": 0.8954547643661499, "learning_rate": 9.99035046550622e-05, "loss": 0.1719, "step": 6880 }, { "grad_norm": 0.8771800994873047, "learning_rate": 9.99024751643393e-05, "loss": 0.124, "step": 6890 }, { "grad_norm": 0.841768741607666, "learning_rate": 9.990144021635677e-05, "loss": 0.1174, "step": 6900 }, { "grad_norm": 0.8459254503250122, "learning_rate": 9.990039981122775e-05, "loss": 0.1212, "step": 6910 }, { "grad_norm": 0.9611098170280457, "learning_rate": 9.989935394906602e-05, "loss": 0.1308, "step": 6920 }, { "grad_norm": 1.2849098443984985, "learning_rate": 9.989830262998598e-05, "loss": 0.1569, "step": 6930 }, { "grad_norm": 0.7680016756057739, "learning_rate": 9.989724585410259e-05, "loss": 0.1336, "step": 6940 }, { "grad_norm": 1.1031455993652344, "learning_rate": 9.989618362153139e-05, "loss": 0.1403, "step": 6950 }, { "grad_norm": 0.9487519264221191, "learning_rate": 9.989511593238859e-05, "loss": 0.1248, "step": 6960 }, { "grad_norm": 1.0604908466339111, "learning_rate": 9.98940427867909e-05, "loss": 0.1467, "step": 6970 }, { "grad_norm": 0.792131245136261, "learning_rate": 9.989296418485573e-05, "loss": 0.1333, "step": 6980 }, { "grad_norm": 0.6842859983444214, "learning_rate": 9.989188012670101e-05, "loss": 0.1106, "step": 6990 }, { "grad_norm": 0.6299747824668884, "learning_rate": 9.989079061244528e-05, "loss": 0.0982, "step": 7000 }, { "grad_norm": 1.0192755460739136, "learning_rate": 9.988969564220769e-05, "loss": 0.1117, "step": 7010 }, { "grad_norm": 1.130700707435608, "learning_rate": 9.988859521610801e-05, "loss": 0.1307, "step": 7020 }, { "grad_norm": 0.9129413366317749, "learning_rate": 9.988748933426656e-05, "loss": 0.106, "step": 7030 }, { "grad_norm": 0.9109381437301636, "learning_rate": 9.988637799680428e-05, "loss": 0.1412, "step": 7040 }, { "grad_norm": 0.9632829427719116, "learning_rate": 9.98852612038427e-05, "loss": 0.1342, "step": 7050 }, { "grad_norm": 0.7449760437011719, "learning_rate": 9.988413895550397e-05, "loss": 0.1155, "step": 7060 }, { "grad_norm": 0.9442205429077148, "learning_rate": 9.98830112519108e-05, "loss": 0.1182, "step": 7070 }, { "grad_norm": 0.6277509927749634, "learning_rate": 9.98818780931865e-05, "loss": 0.124, "step": 7080 }, { "grad_norm": 0.9411152601242065, "learning_rate": 9.988073947945502e-05, "loss": 0.1263, "step": 7090 }, { "grad_norm": 0.907970130443573, "learning_rate": 9.987959541084087e-05, "loss": 0.15, "step": 7100 }, { "grad_norm": 0.9305227398872375, "learning_rate": 9.987844588746915e-05, "loss": 0.1116, "step": 7110 }, { "grad_norm": 0.6791557669639587, "learning_rate": 9.987729090946558e-05, "loss": 0.1173, "step": 7120 }, { "grad_norm": 0.8109787702560425, "learning_rate": 9.987613047695647e-05, "loss": 0.1203, "step": 7130 }, { "grad_norm": 0.7769937515258789, "learning_rate": 9.987496459006871e-05, "loss": 0.1304, "step": 7140 }, { "grad_norm": 0.8207823634147644, "learning_rate": 9.987379324892982e-05, "loss": 0.1564, "step": 7150 }, { "grad_norm": 0.8597224354743958, "learning_rate": 9.987261645366788e-05, "loss": 0.1205, "step": 7160 }, { "grad_norm": 0.7620813250541687, "learning_rate": 9.987143420441158e-05, "loss": 0.1474, "step": 7170 }, { "grad_norm": 1.0995652675628662, "learning_rate": 9.987024650129022e-05, "loss": 0.131, "step": 7180 }, { "grad_norm": 0.7862362861633301, "learning_rate": 9.986905334443368e-05, "loss": 0.1242, "step": 7190 }, { "grad_norm": 0.9790805578231812, "learning_rate": 9.986785473397245e-05, "loss": 0.1327, "step": 7200 }, { "grad_norm": 0.7905594110488892, "learning_rate": 9.98666506700376e-05, "loss": 0.1183, "step": 7210 }, { "grad_norm": 0.7613375186920166, "learning_rate": 9.986544115276081e-05, "loss": 0.1288, "step": 7220 }, { "grad_norm": 0.9750333428382874, "learning_rate": 9.986422618227433e-05, "loss": 0.1127, "step": 7230 }, { "grad_norm": 0.8286240100860596, "learning_rate": 9.986300575871106e-05, "loss": 0.1203, "step": 7240 }, { "grad_norm": 0.7224409580230713, "learning_rate": 9.986177988220444e-05, "loss": 0.0976, "step": 7250 }, { "grad_norm": 0.7010834813117981, "learning_rate": 9.986054855288856e-05, "loss": 0.0982, "step": 7260 }, { "grad_norm": 0.8476609587669373, "learning_rate": 9.985931177089802e-05, "loss": 0.116, "step": 7270 }, { "grad_norm": 0.7460498213768005, "learning_rate": 9.985806953636814e-05, "loss": 0.1091, "step": 7280 }, { "grad_norm": 0.7786434292793274, "learning_rate": 9.985682184943471e-05, "loss": 0.1292, "step": 7290 }, { "grad_norm": 0.6912874579429626, "learning_rate": 9.98555687102342e-05, "loss": 0.1201, "step": 7300 }, { "grad_norm": 0.8059238791465759, "learning_rate": 9.985431011890367e-05, "loss": 0.1369, "step": 7310 }, { "grad_norm": 0.8677680492401123, "learning_rate": 9.985304607558075e-05, "loss": 0.1108, "step": 7320 }, { "grad_norm": 0.8089348077774048, "learning_rate": 9.985177658040364e-05, "loss": 0.1251, "step": 7330 }, { "grad_norm": 0.8723015189170837, "learning_rate": 9.985050163351119e-05, "loss": 0.1174, "step": 7340 }, { "grad_norm": 0.6247819662094116, "learning_rate": 9.984922123504286e-05, "loss": 0.1307, "step": 7350 }, { "grad_norm": 0.8720806241035461, "learning_rate": 9.984793538513862e-05, "loss": 0.1229, "step": 7360 }, { "grad_norm": 0.9379742741584778, "learning_rate": 9.984664408393912e-05, "loss": 0.1267, "step": 7370 }, { "grad_norm": 0.6716248393058777, "learning_rate": 9.984534733158556e-05, "loss": 0.1138, "step": 7380 }, { "grad_norm": 0.8679938912391663, "learning_rate": 9.984404512821977e-05, "loss": 0.1269, "step": 7390 }, { "grad_norm": 0.5659937262535095, "learning_rate": 9.984273747398411e-05, "loss": 0.107, "step": 7400 }, { "grad_norm": 0.8841966986656189, "learning_rate": 9.984142436902165e-05, "loss": 0.118, "step": 7410 }, { "grad_norm": 1.0121030807495117, "learning_rate": 9.984010581347596e-05, "loss": 0.1076, "step": 7420 }, { "grad_norm": 0.774074375629425, "learning_rate": 9.983878180749121e-05, "loss": 0.116, "step": 7430 }, { "grad_norm": 0.8944800496101379, "learning_rate": 9.983745235121222e-05, "loss": 0.1122, "step": 7440 }, { "grad_norm": 0.9779520630836487, "learning_rate": 9.983611744478438e-05, "loss": 0.1192, "step": 7450 }, { "grad_norm": 0.7238088250160217, "learning_rate": 9.983477708835365e-05, "loss": 0.1011, "step": 7460 }, { "grad_norm": 0.7158451080322266, "learning_rate": 9.983343128206664e-05, "loss": 0.1215, "step": 7470 }, { "grad_norm": 0.7888866066932678, "learning_rate": 9.983208002607049e-05, "loss": 0.1031, "step": 7480 }, { "grad_norm": 0.7091410756111145, "learning_rate": 9.9830723320513e-05, "loss": 0.116, "step": 7490 }, { "grad_norm": 0.9252912402153015, "learning_rate": 9.982936116554254e-05, "loss": 0.1507, "step": 7500 }, { "grad_norm": 0.7599365711212158, "learning_rate": 9.982799356130803e-05, "loss": 0.1024, "step": 7510 }, { "grad_norm": 0.5730023384094238, "learning_rate": 9.982662050795908e-05, "loss": 0.0959, "step": 7520 }, { "grad_norm": 0.8175016045570374, "learning_rate": 9.982524200564583e-05, "loss": 0.1143, "step": 7530 }, { "grad_norm": 1.096090316772461, "learning_rate": 9.982385805451901e-05, "loss": 0.1337, "step": 7540 }, { "grad_norm": 0.8793091177940369, "learning_rate": 9.982246865472998e-05, "loss": 0.1198, "step": 7550 }, { "grad_norm": 1.0680394172668457, "learning_rate": 9.982107380643069e-05, "loss": 0.1344, "step": 7560 }, { "grad_norm": 0.8386809229850769, "learning_rate": 9.981967350977368e-05, "loss": 0.122, "step": 7570 }, { "grad_norm": 0.7698608040809631, "learning_rate": 9.981826776491208e-05, "loss": 0.1094, "step": 7580 }, { "grad_norm": 1.1371766328811646, "learning_rate": 9.98168565719996e-05, "loss": 0.1463, "step": 7590 }, { "grad_norm": 0.7297532558441162, "learning_rate": 9.98154399311906e-05, "loss": 0.1376, "step": 7600 }, { "grad_norm": 0.8856207728385925, "learning_rate": 9.981401784263997e-05, "loss": 0.1192, "step": 7610 }, { "grad_norm": 1.0700023174285889, "learning_rate": 9.981259030650326e-05, "loss": 0.1076, "step": 7620 }, { "grad_norm": 0.7908790707588196, "learning_rate": 9.981115732293655e-05, "loss": 0.1088, "step": 7630 }, { "grad_norm": 0.9601367712020874, "learning_rate": 9.980971889209659e-05, "loss": 0.134, "step": 7640 }, { "grad_norm": 0.9675263166427612, "learning_rate": 9.980827501414064e-05, "loss": 0.1185, "step": 7650 }, { "grad_norm": 0.7067102193832397, "learning_rate": 9.980682568922663e-05, "loss": 0.1136, "step": 7660 }, { "grad_norm": 1.0756850242614746, "learning_rate": 9.980537091751304e-05, "loss": 0.1014, "step": 7670 }, { "grad_norm": 0.6811105608940125, "learning_rate": 9.980391069915897e-05, "loss": 0.1115, "step": 7680 }, { "grad_norm": 0.5906137824058533, "learning_rate": 9.98024450343241e-05, "loss": 0.126, "step": 7690 }, { "grad_norm": 1.2311028242111206, "learning_rate": 9.980097392316872e-05, "loss": 0.1434, "step": 7700 }, { "grad_norm": 0.9351291060447693, "learning_rate": 9.97994973658537e-05, "loss": 0.1099, "step": 7710 }, { "grad_norm": 0.8663374185562134, "learning_rate": 9.979801536254054e-05, "loss": 0.1239, "step": 7720 }, { "grad_norm": 0.5719385743141174, "learning_rate": 9.979652791339127e-05, "loss": 0.1198, "step": 7730 }, { "grad_norm": 0.8115229606628418, "learning_rate": 9.97950350185686e-05, "loss": 0.1033, "step": 7740 }, { "grad_norm": 0.7557669878005981, "learning_rate": 9.979353667823574e-05, "loss": 0.1074, "step": 7750 }, { "grad_norm": 0.8390995860099792, "learning_rate": 9.979203289255658e-05, "loss": 0.097, "step": 7760 }, { "grad_norm": 1.0162521600723267, "learning_rate": 9.979052366169557e-05, "loss": 0.1455, "step": 7770 }, { "grad_norm": 0.6871316432952881, "learning_rate": 9.978900898581775e-05, "loss": 0.112, "step": 7780 }, { "grad_norm": 1.001885175704956, "learning_rate": 9.978748886508875e-05, "loss": 0.128, "step": 7790 }, { "grad_norm": 0.851946234703064, "learning_rate": 9.978596329967484e-05, "loss": 0.1284, "step": 7800 }, { "grad_norm": 0.9966526627540588, "learning_rate": 9.978443228974284e-05, "loss": 0.1207, "step": 7810 }, { "grad_norm": 0.8110262751579285, "learning_rate": 9.978289583546015e-05, "loss": 0.1079, "step": 7820 }, { "grad_norm": 0.8469088673591614, "learning_rate": 9.978135393699484e-05, "loss": 0.1287, "step": 7830 }, { "grad_norm": 1.0399794578552246, "learning_rate": 9.977980659451548e-05, "loss": 0.1359, "step": 7840 }, { "grad_norm": 0.8678767681121826, "learning_rate": 9.977825380819135e-05, "loss": 0.1217, "step": 7850 }, { "grad_norm": 0.7992406487464905, "learning_rate": 9.97766955781922e-05, "loss": 0.1177, "step": 7860 }, { "grad_norm": 1.220746636390686, "learning_rate": 9.977513190468848e-05, "loss": 0.1468, "step": 7870 }, { "grad_norm": 0.7085089087486267, "learning_rate": 9.977356278785116e-05, "loss": 0.1466, "step": 7880 }, { "grad_norm": 0.8579751253128052, "learning_rate": 9.977198822785184e-05, "loss": 0.1426, "step": 7890 }, { "grad_norm": 0.8789287805557251, "learning_rate": 9.977040822486273e-05, "loss": 0.1204, "step": 7900 }, { "grad_norm": 0.5928671360015869, "learning_rate": 9.97688227790566e-05, "loss": 0.1459, "step": 7910 }, { "grad_norm": 0.8472351431846619, "learning_rate": 9.976723189060684e-05, "loss": 0.114, "step": 7920 }, { "grad_norm": 0.6632356643676758, "learning_rate": 9.976563555968742e-05, "loss": 0.1177, "step": 7930 }, { "grad_norm": 0.8595002293586731, "learning_rate": 9.976403378647292e-05, "loss": 0.1175, "step": 7940 }, { "grad_norm": 0.7229486703872681, "learning_rate": 9.97624265711385e-05, "loss": 0.1273, "step": 7950 }, { "grad_norm": 0.9305367469787598, "learning_rate": 9.976081391385993e-05, "loss": 0.1036, "step": 7960 }, { "grad_norm": 0.7361112236976624, "learning_rate": 9.975919581481356e-05, "loss": 0.1375, "step": 7970 }, { "grad_norm": 1.2267382144927979, "learning_rate": 9.975757227417634e-05, "loss": 0.1289, "step": 7980 }, { "grad_norm": 0.9147441387176514, "learning_rate": 9.975594329212586e-05, "loss": 0.1333, "step": 7990 }, { "grad_norm": 1.1056863069534302, "learning_rate": 9.97543088688402e-05, "loss": 0.13, "step": 8000 }, { "grad_norm": 0.777895450592041, "learning_rate": 9.975266900449814e-05, "loss": 0.1234, "step": 8010 }, { "grad_norm": 0.7477298974990845, "learning_rate": 9.975102369927898e-05, "loss": 0.1443, "step": 8020 }, { "grad_norm": 1.5404783487319946, "learning_rate": 9.974937295336269e-05, "loss": 0.1368, "step": 8030 }, { "grad_norm": 0.9218682646751404, "learning_rate": 9.974771676692975e-05, "loss": 0.1115, "step": 8040 }, { "grad_norm": 0.7724891901016235, "learning_rate": 9.974605514016131e-05, "loss": 0.1068, "step": 8050 }, { "grad_norm": 1.00066077709198, "learning_rate": 9.974438807323907e-05, "loss": 0.146, "step": 8060 }, { "grad_norm": 0.8093965649604797, "learning_rate": 9.974271556634535e-05, "loss": 0.099, "step": 8070 }, { "grad_norm": 0.5770211815834045, "learning_rate": 9.974103761966302e-05, "loss": 0.1095, "step": 8080 }, { "grad_norm": 0.9652886986732483, "learning_rate": 9.973935423337563e-05, "loss": 0.129, "step": 8090 }, { "grad_norm": 0.662965714931488, "learning_rate": 9.973766540766722e-05, "loss": 0.1141, "step": 8100 }, { "grad_norm": 0.785734236240387, "learning_rate": 9.97359711427225e-05, "loss": 0.1127, "step": 8110 }, { "grad_norm": 0.8307560682296753, "learning_rate": 9.973427143872677e-05, "loss": 0.1123, "step": 8120 }, { "grad_norm": 0.8438348174095154, "learning_rate": 9.973256629586589e-05, "loss": 0.1183, "step": 8130 }, { "grad_norm": 0.7222525477409363, "learning_rate": 9.973085571432632e-05, "loss": 0.1041, "step": 8140 }, { "grad_norm": 0.7395462989807129, "learning_rate": 9.972913969429513e-05, "loss": 0.1283, "step": 8150 }, { "grad_norm": 0.8497434854507446, "learning_rate": 9.972741823596e-05, "loss": 0.127, "step": 8160 }, { "grad_norm": 0.7928845882415771, "learning_rate": 9.972569133950917e-05, "loss": 0.1146, "step": 8170 }, { "grad_norm": 0.8306027054786682, "learning_rate": 9.972395900513151e-05, "loss": 0.1097, "step": 8180 }, { "grad_norm": 0.9338010549545288, "learning_rate": 9.972222123301645e-05, "loss": 0.1141, "step": 8190 }, { "grad_norm": 0.836601197719574, "learning_rate": 9.972047802335403e-05, "loss": 0.1075, "step": 8200 }, { "grad_norm": 0.8903421759605408, "learning_rate": 9.971872937633488e-05, "loss": 0.0893, "step": 8210 }, { "grad_norm": 0.8305997848510742, "learning_rate": 9.971697529215024e-05, "loss": 0.1271, "step": 8220 }, { "grad_norm": 1.2355844974517822, "learning_rate": 9.971521577099192e-05, "loss": 0.129, "step": 8230 }, { "grad_norm": 1.4038933515548706, "learning_rate": 9.971345081305236e-05, "loss": 0.1158, "step": 8240 }, { "grad_norm": 0.5942561030387878, "learning_rate": 9.971168041852456e-05, "loss": 0.1075, "step": 8250 }, { "grad_norm": 1.0584590435028076, "learning_rate": 9.970990458760215e-05, "loss": 0.1155, "step": 8260 }, { "grad_norm": 0.8966363668441772, "learning_rate": 9.970812332047929e-05, "loss": 0.1245, "step": 8270 }, { "grad_norm": 0.8896788954734802, "learning_rate": 9.97063366173508e-05, "loss": 0.1066, "step": 8280 }, { "grad_norm": 0.6199048757553101, "learning_rate": 9.970454447841207e-05, "loss": 0.1333, "step": 8290 }, { "grad_norm": 0.9888023138046265, "learning_rate": 9.970274690385909e-05, "loss": 0.132, "step": 8300 }, { "grad_norm": 1.1640733480453491, "learning_rate": 9.970094389388844e-05, "loss": 0.1255, "step": 8310 }, { "grad_norm": 0.8685014247894287, "learning_rate": 9.969913544869728e-05, "loss": 0.1221, "step": 8320 }, { "grad_norm": 0.714165210723877, "learning_rate": 9.96973215684834e-05, "loss": 0.0975, "step": 8330 }, { "grad_norm": 1.0964475870132446, "learning_rate": 9.969550225344513e-05, "loss": 0.1075, "step": 8340 }, { "grad_norm": 0.8092615604400635, "learning_rate": 9.969367750378147e-05, "loss": 0.1227, "step": 8350 }, { "grad_norm": 0.7243706583976746, "learning_rate": 9.969184731969194e-05, "loss": 0.1205, "step": 8360 }, { "grad_norm": 1.0241942405700684, "learning_rate": 9.96900117013767e-05, "loss": 0.1358, "step": 8370 }, { "grad_norm": 0.9080133438110352, "learning_rate": 9.96881706490365e-05, "loss": 0.103, "step": 8380 }, { "grad_norm": 0.9740006923675537, "learning_rate": 9.968632416287265e-05, "loss": 0.1343, "step": 8390 }, { "grad_norm": 0.6454100608825684, "learning_rate": 9.96844722430871e-05, "loss": 0.1137, "step": 8400 }, { "grad_norm": 0.7086679935455322, "learning_rate": 9.968261488988235e-05, "loss": 0.1247, "step": 8410 }, { "grad_norm": 0.6946403980255127, "learning_rate": 9.968075210346155e-05, "loss": 0.1193, "step": 8420 }, { "grad_norm": 0.6763221025466919, "learning_rate": 9.967888388402839e-05, "loss": 0.1395, "step": 8430 }, { "grad_norm": 0.9765684604644775, "learning_rate": 9.967701023178717e-05, "loss": 0.1359, "step": 8440 }, { "grad_norm": 1.0695669651031494, "learning_rate": 9.967513114694282e-05, "loss": 0.1322, "step": 8450 }, { "grad_norm": 0.9928604960441589, "learning_rate": 9.967324662970079e-05, "loss": 0.1476, "step": 8460 }, { "grad_norm": 1.145950436592102, "learning_rate": 9.96713566802672e-05, "loss": 0.1149, "step": 8470 }, { "grad_norm": 1.011491060256958, "learning_rate": 9.966946129884873e-05, "loss": 0.1379, "step": 8480 }, { "grad_norm": 1.1056150197982788, "learning_rate": 9.966756048565265e-05, "loss": 0.1223, "step": 8490 }, { "grad_norm": 0.8844866752624512, "learning_rate": 9.966565424088681e-05, "loss": 0.1176, "step": 8500 }, { "grad_norm": 1.0323196649551392, "learning_rate": 9.96637425647597e-05, "loss": 0.1319, "step": 8510 }, { "grad_norm": 0.959927499294281, "learning_rate": 9.966182545748038e-05, "loss": 0.136, "step": 8520 }, { "grad_norm": 0.9265884160995483, "learning_rate": 9.96599029192585e-05, "loss": 0.1126, "step": 8530 }, { "grad_norm": 0.7836909294128418, "learning_rate": 9.965797495030428e-05, "loss": 0.1212, "step": 8540 }, { "grad_norm": 0.7496338486671448, "learning_rate": 9.96560415508286e-05, "loss": 0.1067, "step": 8550 }, { "grad_norm": 0.9275850653648376, "learning_rate": 9.965410272104286e-05, "loss": 0.1179, "step": 8560 }, { "grad_norm": 1.179119348526001, "learning_rate": 9.96521584611591e-05, "loss": 0.1435, "step": 8570 }, { "grad_norm": 0.9702613353729248, "learning_rate": 9.965020877138994e-05, "loss": 0.1285, "step": 8580 }, { "grad_norm": 1.0295568704605103, "learning_rate": 9.964825365194861e-05, "loss": 0.1302, "step": 8590 }, { "grad_norm": 1.106991171836853, "learning_rate": 9.96462931030489e-05, "loss": 0.1107, "step": 8600 }, { "grad_norm": 2.2576959133148193, "learning_rate": 9.96443271249052e-05, "loss": 0.1335, "step": 8610 }, { "grad_norm": 0.7826164960861206, "learning_rate": 9.964235571773255e-05, "loss": 0.1308, "step": 8620 }, { "grad_norm": 1.1666662693023682, "learning_rate": 9.96403788817465e-05, "loss": 0.165, "step": 8630 }, { "grad_norm": 0.9140579104423523, "learning_rate": 9.963839661716325e-05, "loss": 0.1136, "step": 8640 }, { "grad_norm": 0.7420567274093628, "learning_rate": 9.963640892419958e-05, "loss": 0.0962, "step": 8650 }, { "grad_norm": 0.7164888978004456, "learning_rate": 9.963441580307286e-05, "loss": 0.105, "step": 8660 }, { "grad_norm": 1.1173795461654663, "learning_rate": 9.963241725400104e-05, "loss": 0.106, "step": 8670 }, { "grad_norm": 0.6314429640769958, "learning_rate": 9.963041327720271e-05, "loss": 0.1015, "step": 8680 }, { "grad_norm": 0.631750762462616, "learning_rate": 9.962840387289697e-05, "loss": 0.1039, "step": 8690 }, { "grad_norm": 0.6227891445159912, "learning_rate": 9.962638904130363e-05, "loss": 0.0847, "step": 8700 }, { "grad_norm": 0.9309449791908264, "learning_rate": 9.962436878264298e-05, "loss": 0.1017, "step": 8710 }, { "grad_norm": 0.67230623960495, "learning_rate": 9.962234309713598e-05, "loss": 0.0852, "step": 8720 }, { "grad_norm": 0.6945277452468872, "learning_rate": 9.962031198500414e-05, "loss": 0.1072, "step": 8730 }, { "grad_norm": 0.7134501934051514, "learning_rate": 9.961827544646958e-05, "loss": 0.1164, "step": 8740 }, { "grad_norm": 0.9285884499549866, "learning_rate": 9.961623348175501e-05, "loss": 0.1038, "step": 8750 }, { "grad_norm": 0.7845903038978577, "learning_rate": 9.961418609108377e-05, "loss": 0.1225, "step": 8760 }, { "grad_norm": 0.836473286151886, "learning_rate": 9.961213327467971e-05, "loss": 0.1113, "step": 8770 }, { "grad_norm": 0.9852554798126221, "learning_rate": 9.961007503276736e-05, "loss": 0.1166, "step": 8780 }, { "grad_norm": 0.7179824113845825, "learning_rate": 9.960801136557179e-05, "loss": 0.1362, "step": 8790 }, { "grad_norm": 0.6981139779090881, "learning_rate": 9.960594227331866e-05, "loss": 0.1051, "step": 8800 }, { "grad_norm": 0.8729488849639893, "learning_rate": 9.960386775623429e-05, "loss": 0.123, "step": 8810 }, { "grad_norm": 0.824259340763092, "learning_rate": 9.96017878145455e-05, "loss": 0.1232, "step": 8820 }, { "grad_norm": 0.9057918190956116, "learning_rate": 9.959970244847977e-05, "loss": 0.1377, "step": 8830 }, { "grad_norm": 0.7064768671989441, "learning_rate": 9.959761165826518e-05, "loss": 0.1241, "step": 8840 }, { "grad_norm": 0.6860164999961853, "learning_rate": 9.959551544413033e-05, "loss": 0.1141, "step": 8850 }, { "grad_norm": 0.8583148717880249, "learning_rate": 9.959341380630448e-05, "loss": 0.0969, "step": 8860 }, { "grad_norm": 0.6554011106491089, "learning_rate": 9.959130674501746e-05, "loss": 0.119, "step": 8870 }, { "grad_norm": 0.6850699782371521, "learning_rate": 9.958919426049968e-05, "loss": 0.1041, "step": 8880 }, { "grad_norm": 0.6985675096511841, "learning_rate": 9.958707635298219e-05, "loss": 0.1296, "step": 8890 }, { "grad_norm": 1.121345043182373, "learning_rate": 9.958495302269657e-05, "loss": 0.131, "step": 8900 }, { "grad_norm": 0.8658716678619385, "learning_rate": 9.958282426987503e-05, "loss": 0.1401, "step": 8910 }, { "grad_norm": 0.748931348323822, "learning_rate": 9.95806900947504e-05, "loss": 0.0971, "step": 8920 }, { "grad_norm": 0.6728017926216125, "learning_rate": 9.957855049755604e-05, "loss": 0.1269, "step": 8930 }, { "grad_norm": 0.8509067893028259, "learning_rate": 9.957640547852593e-05, "loss": 0.1021, "step": 8940 }, { "grad_norm": 1.0803554058074951, "learning_rate": 9.957425503789466e-05, "loss": 0.0999, "step": 8950 }, { "grad_norm": 0.9941617846488953, "learning_rate": 9.957209917589738e-05, "loss": 0.1332, "step": 8960 }, { "grad_norm": 1.0748333930969238, "learning_rate": 9.956993789276987e-05, "loss": 0.1231, "step": 8970 }, { "grad_norm": 0.8804616332054138, "learning_rate": 9.956777118874847e-05, "loss": 0.1293, "step": 8980 }, { "grad_norm": 0.9812070727348328, "learning_rate": 9.956559906407016e-05, "loss": 0.1309, "step": 8990 }, { "grad_norm": 0.8302607536315918, "learning_rate": 9.956342151897245e-05, "loss": 0.1062, "step": 9000 }, { "grad_norm": 1.6869239807128906, "learning_rate": 9.956123855369346e-05, "loss": 0.1562, "step": 9010 }, { "grad_norm": 0.9391546249389648, "learning_rate": 9.955905016847196e-05, "loss": 0.1409, "step": 9020 }, { "grad_norm": 0.8369888067245483, "learning_rate": 9.955685636354723e-05, "loss": 0.1094, "step": 9030 }, { "grad_norm": 0.6963891983032227, "learning_rate": 9.95546571391592e-05, "loss": 0.1327, "step": 9040 }, { "grad_norm": 0.9462390542030334, "learning_rate": 9.955245249554837e-05, "loss": 0.1193, "step": 9050 }, { "grad_norm": 1.1133322715759277, "learning_rate": 9.955024243295582e-05, "loss": 0.1298, "step": 9060 }, { "grad_norm": 0.9167202711105347, "learning_rate": 9.954802695162328e-05, "loss": 0.1278, "step": 9070 }, { "grad_norm": 1.4912103414535522, "learning_rate": 9.954580605179302e-05, "loss": 0.1193, "step": 9080 }, { "grad_norm": 0.646638035774231, "learning_rate": 9.954357973370788e-05, "loss": 0.1457, "step": 9090 }, { "grad_norm": 0.7574346661567688, "learning_rate": 9.954134799761135e-05, "loss": 0.1038, "step": 9100 }, { "grad_norm": 0.6291149854660034, "learning_rate": 9.953911084374748e-05, "loss": 0.1005, "step": 9110 }, { "grad_norm": 0.7383108139038086, "learning_rate": 9.953686827236093e-05, "loss": 0.13, "step": 9120 }, { "grad_norm": 0.7200573086738586, "learning_rate": 9.953462028369695e-05, "loss": 0.1261, "step": 9130 }, { "grad_norm": 0.9287683963775635, "learning_rate": 9.953236687800136e-05, "loss": 0.1303, "step": 9140 }, { "grad_norm": 0.6133270263671875, "learning_rate": 9.95301080555206e-05, "loss": 0.1114, "step": 9150 }, { "grad_norm": 0.9350899457931519, "learning_rate": 9.952784381650171e-05, "loss": 0.1258, "step": 9160 }, { "grad_norm": 0.9420683979988098, "learning_rate": 9.952557416119226e-05, "loss": 0.1125, "step": 9170 }, { "grad_norm": 0.6672725677490234, "learning_rate": 9.95232990898405e-05, "loss": 0.1029, "step": 9180 }, { "grad_norm": 0.7794475555419922, "learning_rate": 9.95210186026952e-05, "loss": 0.1259, "step": 9190 }, { "grad_norm": 0.7985951900482178, "learning_rate": 9.951873270000576e-05, "loss": 0.0938, "step": 9200 }, { "grad_norm": 0.851832389831543, "learning_rate": 9.951644138202216e-05, "loss": 0.1038, "step": 9210 }, { "grad_norm": 0.9110016822814941, "learning_rate": 9.951414464899498e-05, "loss": 0.1205, "step": 9220 }, { "grad_norm": 0.9439889192581177, "learning_rate": 9.951184250117538e-05, "loss": 0.0895, "step": 9230 }, { "grad_norm": 1.155942440032959, "learning_rate": 9.950953493881513e-05, "loss": 0.1298, "step": 9240 }, { "grad_norm": 0.9210226535797119, "learning_rate": 9.950722196216658e-05, "loss": 0.1089, "step": 9250 }, { "grad_norm": 0.7920499444007874, "learning_rate": 9.950490357148265e-05, "loss": 0.1415, "step": 9260 }, { "grad_norm": 0.5549044013023376, "learning_rate": 9.950257976701692e-05, "loss": 0.1084, "step": 9270 }, { "grad_norm": 0.6812148690223694, "learning_rate": 9.950025054902348e-05, "loss": 0.1261, "step": 9280 }, { "grad_norm": 0.6389753818511963, "learning_rate": 9.949791591775706e-05, "loss": 0.0982, "step": 9290 }, { "grad_norm": 0.8688968420028687, "learning_rate": 9.949557587347298e-05, "loss": 0.1127, "step": 9300 }, { "grad_norm": 0.9613819718360901, "learning_rate": 9.949323041642713e-05, "loss": 0.1456, "step": 9310 }, { "grad_norm": 0.7879213094711304, "learning_rate": 9.949087954687602e-05, "loss": 0.1079, "step": 9320 }, { "grad_norm": 1.0570451021194458, "learning_rate": 9.948852326507672e-05, "loss": 0.1209, "step": 9330 }, { "grad_norm": 1.2241922616958618, "learning_rate": 9.948616157128694e-05, "loss": 0.1066, "step": 9340 }, { "grad_norm": 0.6760350465774536, "learning_rate": 9.948379446576493e-05, "loss": 0.0985, "step": 9350 }, { "grad_norm": 0.6921988129615784, "learning_rate": 9.948142194876952e-05, "loss": 0.1034, "step": 9360 }, { "grad_norm": 0.9795412421226501, "learning_rate": 9.947904402056024e-05, "loss": 0.1174, "step": 9370 }, { "grad_norm": 0.8625786304473877, "learning_rate": 9.947666068139708e-05, "loss": 0.1339, "step": 9380 }, { "grad_norm": 0.6297959089279175, "learning_rate": 9.947427193154071e-05, "loss": 0.1045, "step": 9390 }, { "grad_norm": 0.8743233680725098, "learning_rate": 9.947187777125233e-05, "loss": 0.1166, "step": 9400 }, { "grad_norm": 0.6835300326347351, "learning_rate": 9.946947820079377e-05, "loss": 0.1053, "step": 9410 }, { "grad_norm": 0.8944941759109497, "learning_rate": 9.946707322042747e-05, "loss": 0.1299, "step": 9420 }, { "grad_norm": 0.6777350306510925, "learning_rate": 9.94646628304164e-05, "loss": 0.1027, "step": 9430 }, { "grad_norm": 1.1236889362335205, "learning_rate": 9.946224703102418e-05, "loss": 0.1068, "step": 9440 }, { "grad_norm": 0.7125857472419739, "learning_rate": 9.945982582251498e-05, "loss": 0.0988, "step": 9450 }, { "grad_norm": 0.6218726634979248, "learning_rate": 9.94573992051536e-05, "loss": 0.1479, "step": 9460 }, { "grad_norm": 0.7648109793663025, "learning_rate": 9.94549671792054e-05, "loss": 0.123, "step": 9470 }, { "grad_norm": 0.6903023719787598, "learning_rate": 9.945252974493635e-05, "loss": 0.0929, "step": 9480 }, { "grad_norm": 0.883831799030304, "learning_rate": 9.9450086902613e-05, "loss": 0.1007, "step": 9490 }, { "grad_norm": 0.5251662731170654, "learning_rate": 9.944763865250248e-05, "loss": 0.0943, "step": 9500 }, { "grad_norm": 0.7490226626396179, "learning_rate": 9.944518499487254e-05, "loss": 0.1204, "step": 9510 }, { "grad_norm": 0.8511526584625244, "learning_rate": 9.944272592999151e-05, "loss": 0.1151, "step": 9520 }, { "grad_norm": 0.9133392572402954, "learning_rate": 9.94402614581283e-05, "loss": 0.1169, "step": 9530 }, { "grad_norm": 1.1175918579101562, "learning_rate": 9.943779157955244e-05, "loss": 0.1189, "step": 9540 }, { "grad_norm": 0.7206315994262695, "learning_rate": 9.943531629453403e-05, "loss": 0.1392, "step": 9550 }, { "grad_norm": 0.9925403594970703, "learning_rate": 9.943283560334375e-05, "loss": 0.1247, "step": 9560 }, { "grad_norm": 0.6978933215141296, "learning_rate": 9.943034950625288e-05, "loss": 0.0945, "step": 9570 }, { "grad_norm": 0.6782399415969849, "learning_rate": 9.942785800353332e-05, "loss": 0.1106, "step": 9580 }, { "grad_norm": 0.7005835771560669, "learning_rate": 9.942536109545751e-05, "loss": 0.1031, "step": 9590 }, { "grad_norm": 0.8286450505256653, "learning_rate": 9.942285878229853e-05, "loss": 0.1165, "step": 9600 }, { "grad_norm": 0.6734827756881714, "learning_rate": 9.942035106433001e-05, "loss": 0.0845, "step": 9610 }, { "grad_norm": 0.6243906617164612, "learning_rate": 9.94178379418262e-05, "loss": 0.1016, "step": 9620 }, { "grad_norm": 0.9771730899810791, "learning_rate": 9.941531941506194e-05, "loss": 0.1012, "step": 9630 }, { "grad_norm": 0.721775472164154, "learning_rate": 9.941279548431263e-05, "loss": 0.1379, "step": 9640 }, { "grad_norm": 0.9556573033332825, "learning_rate": 9.941026614985431e-05, "loss": 0.1034, "step": 9650 }, { "grad_norm": 0.8283907175064087, "learning_rate": 9.940773141196357e-05, "loss": 0.0924, "step": 9660 }, { "grad_norm": 0.8084217309951782, "learning_rate": 9.94051912709176e-05, "loss": 0.1204, "step": 9670 }, { "grad_norm": 0.8145257830619812, "learning_rate": 9.940264572699421e-05, "loss": 0.1093, "step": 9680 }, { "grad_norm": 0.8224850296974182, "learning_rate": 9.940009478047174e-05, "loss": 0.1134, "step": 9690 }, { "grad_norm": 0.8174322247505188, "learning_rate": 9.939753843162918e-05, "loss": 0.111, "step": 9700 }, { "grad_norm": 0.6566669344902039, "learning_rate": 9.939497668074609e-05, "loss": 0.0917, "step": 9710 }, { "grad_norm": 0.7340297102928162, "learning_rate": 9.93924095281026e-05, "loss": 0.0886, "step": 9720 }, { "grad_norm": 0.9729318022727966, "learning_rate": 9.938983697397948e-05, "loss": 0.1159, "step": 9730 }, { "grad_norm": 0.7033445835113525, "learning_rate": 9.938725901865805e-05, "loss": 0.1028, "step": 9740 }, { "grad_norm": 1.2624092102050781, "learning_rate": 9.93846756624202e-05, "loss": 0.1089, "step": 9750 }, { "grad_norm": 0.8768384456634521, "learning_rate": 9.938208690554849e-05, "loss": 0.1078, "step": 9760 }, { "grad_norm": 0.6570026874542236, "learning_rate": 9.9379492748326e-05, "loss": 0.1242, "step": 9770 }, { "grad_norm": 0.6011269092559814, "learning_rate": 9.937689319103641e-05, "loss": 0.1159, "step": 9780 }, { "grad_norm": 0.9253286123275757, "learning_rate": 9.937428823396404e-05, "loss": 0.1087, "step": 9790 }, { "grad_norm": 0.7789440155029297, "learning_rate": 9.937167787739372e-05, "loss": 0.0926, "step": 9800 }, { "grad_norm": 0.9072915315628052, "learning_rate": 9.936906212161095e-05, "loss": 0.0915, "step": 9810 }, { "grad_norm": 0.6810943484306335, "learning_rate": 9.936644096690176e-05, "loss": 0.1117, "step": 9820 }, { "grad_norm": 1.4472684860229492, "learning_rate": 9.936381441355282e-05, "loss": 0.1398, "step": 9830 }, { "grad_norm": 0.8197895288467407, "learning_rate": 9.936118246185136e-05, "loss": 0.1086, "step": 9840 }, { "grad_norm": 0.6963949799537659, "learning_rate": 9.935854511208518e-05, "loss": 0.1126, "step": 9850 }, { "grad_norm": 0.6545697450637817, "learning_rate": 9.935590236454272e-05, "loss": 0.0966, "step": 9860 }, { "grad_norm": 1.1380804777145386, "learning_rate": 9.935325421951298e-05, "loss": 0.1085, "step": 9870 }, { "grad_norm": 0.8374897837638855, "learning_rate": 9.935060067728557e-05, "loss": 0.0939, "step": 9880 }, { "grad_norm": 0.7912716269493103, "learning_rate": 9.934794173815067e-05, "loss": 0.1152, "step": 9890 }, { "grad_norm": 0.6038432717323303, "learning_rate": 9.934527740239906e-05, "loss": 0.0921, "step": 9900 }, { "grad_norm": 0.7763460278511047, "learning_rate": 9.934260767032209e-05, "loss": 0.1165, "step": 9910 }, { "grad_norm": 0.7776538729667664, "learning_rate": 9.933993254221172e-05, "loss": 0.0995, "step": 9920 }, { "grad_norm": 1.039384365081787, "learning_rate": 9.933725201836053e-05, "loss": 0.126, "step": 9930 }, { "grad_norm": 0.8051326274871826, "learning_rate": 9.933456609906162e-05, "loss": 0.1158, "step": 9940 }, { "grad_norm": 1.0329238176345825, "learning_rate": 9.933187478460875e-05, "loss": 0.1209, "step": 9950 }, { "grad_norm": 0.9953287839889526, "learning_rate": 9.93291780752962e-05, "loss": 0.1229, "step": 9960 }, { "grad_norm": 0.8715243935585022, "learning_rate": 9.932647597141893e-05, "loss": 0.1062, "step": 9970 }, { "grad_norm": 0.895285427570343, "learning_rate": 9.932376847327239e-05, "loss": 0.116, "step": 9980 }, { "grad_norm": 0.7422624230384827, "learning_rate": 9.932105558115268e-05, "loss": 0.1112, "step": 9990 }, { "grad_norm": 0.8075588941574097, "learning_rate": 9.931833729535651e-05, "loss": 0.1439, "step": 10000 }, { "grad_norm": 0.6635860204696655, "learning_rate": 9.931561361618111e-05, "loss": 0.1148, "step": 10010 }, { "grad_norm": 0.9174380302429199, "learning_rate": 9.931288454392435e-05, "loss": 0.1061, "step": 10020 }, { "grad_norm": 0.534555196762085, "learning_rate": 9.931015007888467e-05, "loss": 0.1012, "step": 10030 }, { "grad_norm": 1.011292815208435, "learning_rate": 9.930741022136112e-05, "loss": 0.1115, "step": 10040 }, { "grad_norm": 1.447161316871643, "learning_rate": 9.930466497165333e-05, "loss": 0.1198, "step": 10050 }, { "grad_norm": 0.8361740708351135, "learning_rate": 9.93019143300615e-05, "loss": 0.082, "step": 10060 }, { "grad_norm": 0.8564516305923462, "learning_rate": 9.929915829688644e-05, "loss": 0.1065, "step": 10070 }, { "grad_norm": 1.0384271144866943, "learning_rate": 9.929639687242955e-05, "loss": 0.1238, "step": 10080 }, { "grad_norm": 1.0140039920806885, "learning_rate": 9.929363005699281e-05, "loss": 0.1341, "step": 10090 }, { "grad_norm": 0.8809318542480469, "learning_rate": 9.92908578508788e-05, "loss": 0.1133, "step": 10100 }, { "grad_norm": 0.6887377500534058, "learning_rate": 9.928808025439069e-05, "loss": 0.1055, "step": 10110 }, { "grad_norm": 0.6361767053604126, "learning_rate": 9.928529726783223e-05, "loss": 0.0951, "step": 10120 }, { "grad_norm": 0.5162762999534607, "learning_rate": 9.928250889150774e-05, "loss": 0.1359, "step": 10130 }, { "grad_norm": 0.7143911123275757, "learning_rate": 9.92797151257222e-05, "loss": 0.1182, "step": 10140 }, { "grad_norm": 0.8303850293159485, "learning_rate": 9.927691597078108e-05, "loss": 0.1563, "step": 10150 }, { "grad_norm": 0.976688802242279, "learning_rate": 9.927411142699053e-05, "loss": 0.1106, "step": 10160 }, { "grad_norm": 0.5963495969772339, "learning_rate": 9.927130149465725e-05, "loss": 0.0983, "step": 10170 }, { "grad_norm": 1.1290552616119385, "learning_rate": 9.92684861740885e-05, "loss": 0.1093, "step": 10180 }, { "grad_norm": 0.8171343207359314, "learning_rate": 9.926566546559217e-05, "loss": 0.0956, "step": 10190 }, { "grad_norm": 0.9520787000656128, "learning_rate": 9.926283936947673e-05, "loss": 0.1137, "step": 10200 }, { "grad_norm": 0.7823962569236755, "learning_rate": 9.926000788605126e-05, "loss": 0.1042, "step": 10210 }, { "grad_norm": 0.7724796533584595, "learning_rate": 9.92571710156254e-05, "loss": 0.1367, "step": 10220 }, { "grad_norm": 0.6247060298919678, "learning_rate": 9.925432875850936e-05, "loss": 0.1188, "step": 10230 }, { "grad_norm": 0.847324788570404, "learning_rate": 9.925148111501396e-05, "loss": 0.1341, "step": 10240 }, { "grad_norm": 0.8726259469985962, "learning_rate": 9.924862808545066e-05, "loss": 0.1192, "step": 10250 }, { "grad_norm": 0.8343139886856079, "learning_rate": 9.924576967013141e-05, "loss": 0.1203, "step": 10260 }, { "grad_norm": 0.8313607573509216, "learning_rate": 9.924290586936887e-05, "loss": 0.1181, "step": 10270 }, { "grad_norm": 0.7463793754577637, "learning_rate": 9.924003668347614e-05, "loss": 0.1007, "step": 10280 }, { "grad_norm": 0.888272225856781, "learning_rate": 9.923716211276704e-05, "loss": 0.1032, "step": 10290 }, { "grad_norm": 1.0299919843673706, "learning_rate": 9.923428215755594e-05, "loss": 0.1016, "step": 10300 }, { "grad_norm": 0.6457698345184326, "learning_rate": 9.923139681815775e-05, "loss": 0.1268, "step": 10310 }, { "grad_norm": 0.6544174551963806, "learning_rate": 9.922850609488801e-05, "loss": 0.1102, "step": 10320 }, { "grad_norm": 0.7687867879867554, "learning_rate": 9.922560998806287e-05, "loss": 0.1157, "step": 10330 }, { "grad_norm": 0.8977317810058594, "learning_rate": 9.922270849799905e-05, "loss": 0.1038, "step": 10340 }, { "grad_norm": 0.6284701228141785, "learning_rate": 9.92198016250138e-05, "loss": 0.1029, "step": 10350 }, { "grad_norm": 0.7273046970367432, "learning_rate": 9.921688936942506e-05, "loss": 0.1023, "step": 10360 }, { "grad_norm": 1.1062486171722412, "learning_rate": 9.921397173155129e-05, "loss": 0.132, "step": 10370 }, { "grad_norm": 0.7048616409301758, "learning_rate": 9.921104871171157e-05, "loss": 0.1424, "step": 10380 }, { "grad_norm": 0.592090368270874, "learning_rate": 9.920812031022554e-05, "loss": 0.1056, "step": 10390 }, { "grad_norm": 1.1818903684616089, "learning_rate": 9.920518652741348e-05, "loss": 0.1067, "step": 10400 }, { "grad_norm": 0.7368386387825012, "learning_rate": 9.920224736359618e-05, "loss": 0.1158, "step": 10410 }, { "grad_norm": 0.7428275346755981, "learning_rate": 9.91993028190951e-05, "loss": 0.1211, "step": 10420 }, { "grad_norm": 0.7377288341522217, "learning_rate": 9.919635289423222e-05, "loss": 0.0944, "step": 10430 }, { "grad_norm": 0.7383182048797607, "learning_rate": 9.919339758933015e-05, "loss": 0.1072, "step": 10440 }, { "grad_norm": 0.7498071789741516, "learning_rate": 9.919043690471209e-05, "loss": 0.0905, "step": 10450 }, { "grad_norm": 0.76882004737854, "learning_rate": 9.91874708407018e-05, "loss": 0.1135, "step": 10460 }, { "grad_norm": 0.8410624265670776, "learning_rate": 9.918449939762367e-05, "loss": 0.1038, "step": 10470 }, { "grad_norm": 1.004152774810791, "learning_rate": 9.91815225758026e-05, "loss": 0.1016, "step": 10480 }, { "grad_norm": 0.6359772086143494, "learning_rate": 9.917854037556419e-05, "loss": 0.1392, "step": 10490 }, { "grad_norm": 0.7547083497047424, "learning_rate": 9.917555279723454e-05, "loss": 0.1049, "step": 10500 }, { "grad_norm": 0.8769236207008362, "learning_rate": 9.917255984114036e-05, "loss": 0.0897, "step": 10510 }, { "grad_norm": 1.120335340499878, "learning_rate": 9.916956150760896e-05, "loss": 0.1162, "step": 10520 }, { "grad_norm": 0.750665545463562, "learning_rate": 9.916655779696826e-05, "loss": 0.1151, "step": 10530 }, { "grad_norm": 0.6536281108856201, "learning_rate": 9.916354870954671e-05, "loss": 0.1043, "step": 10540 }, { "grad_norm": 0.7093473672866821, "learning_rate": 9.91605342456734e-05, "loss": 0.109, "step": 10550 }, { "grad_norm": 0.7781521081924438, "learning_rate": 9.915751440567795e-05, "loss": 0.1085, "step": 10560 }, { "grad_norm": 0.9541229009628296, "learning_rate": 9.915448918989066e-05, "loss": 0.1466, "step": 10570 }, { "grad_norm": 0.8136950731277466, "learning_rate": 9.915145859864232e-05, "loss": 0.098, "step": 10580 }, { "grad_norm": 0.7315027117729187, "learning_rate": 9.914842263226437e-05, "loss": 0.1103, "step": 10590 }, { "grad_norm": 0.8679319620132446, "learning_rate": 9.914538129108882e-05, "loss": 0.0892, "step": 10600 }, { "grad_norm": 0.6893255114555359, "learning_rate": 9.914233457544825e-05, "loss": 0.0898, "step": 10610 }, { "grad_norm": 0.5771347880363464, "learning_rate": 9.913928248567586e-05, "loss": 0.091, "step": 10620 }, { "grad_norm": 0.8795440793037415, "learning_rate": 9.913622502210542e-05, "loss": 0.0995, "step": 10630 }, { "grad_norm": 0.6640200018882751, "learning_rate": 9.913316218507128e-05, "loss": 0.0901, "step": 10640 }, { "grad_norm": 0.9421930909156799, "learning_rate": 9.91300939749084e-05, "loss": 0.1335, "step": 10650 }, { "grad_norm": 0.7874806523323059, "learning_rate": 9.91270203919523e-05, "loss": 0.1086, "step": 10660 }, { "grad_norm": 0.6908004879951477, "learning_rate": 9.912394143653912e-05, "loss": 0.1077, "step": 10670 }, { "grad_norm": 0.7358015179634094, "learning_rate": 9.912085710900555e-05, "loss": 0.1022, "step": 10680 }, { "grad_norm": 0.832926332950592, "learning_rate": 9.911776740968892e-05, "loss": 0.1001, "step": 10690 }, { "grad_norm": 0.7693156003952026, "learning_rate": 9.911467233892709e-05, "loss": 0.095, "step": 10700 }, { "grad_norm": 0.8833759427070618, "learning_rate": 9.911157189705853e-05, "loss": 0.0953, "step": 10710 }, { "grad_norm": 0.7522829174995422, "learning_rate": 9.910846608442229e-05, "loss": 0.1076, "step": 10720 }, { "grad_norm": 0.542607843875885, "learning_rate": 9.910535490135805e-05, "loss": 0.0975, "step": 10730 }, { "grad_norm": 0.8606292009353638, "learning_rate": 9.910223834820603e-05, "loss": 0.1153, "step": 10740 }, { "grad_norm": 0.958458423614502, "learning_rate": 9.909911642530703e-05, "loss": 0.1021, "step": 10750 }, { "grad_norm": 0.7489601969718933, "learning_rate": 9.909598913300249e-05, "loss": 0.0989, "step": 10760 }, { "grad_norm": 0.6961759924888611, "learning_rate": 9.909285647163438e-05, "loss": 0.0961, "step": 10770 }, { "grad_norm": 1.0517387390136719, "learning_rate": 9.908971844154531e-05, "loss": 0.0986, "step": 10780 }, { "grad_norm": 0.7875208258628845, "learning_rate": 9.908657504307843e-05, "loss": 0.1027, "step": 10790 }, { "grad_norm": 0.7027023434638977, "learning_rate": 9.908342627657751e-05, "loss": 0.0942, "step": 10800 }, { "grad_norm": 0.9286534786224365, "learning_rate": 9.908027214238689e-05, "loss": 0.087, "step": 10810 }, { "grad_norm": 0.744388997554779, "learning_rate": 9.90771126408515e-05, "loss": 0.085, "step": 10820 }, { "grad_norm": 0.7828751802444458, "learning_rate": 9.907394777231685e-05, "loss": 0.1168, "step": 10830 }, { "grad_norm": 0.5766028761863708, "learning_rate": 9.907077753712905e-05, "loss": 0.0971, "step": 10840 }, { "grad_norm": 0.8571682572364807, "learning_rate": 9.906760193563482e-05, "loss": 0.1065, "step": 10850 }, { "grad_norm": 0.7804542779922485, "learning_rate": 9.906442096818139e-05, "loss": 0.1075, "step": 10860 }, { "grad_norm": 0.7929694056510925, "learning_rate": 9.906123463511665e-05, "loss": 0.1376, "step": 10870 }, { "grad_norm": 0.7652131915092468, "learning_rate": 9.905804293678907e-05, "loss": 0.1143, "step": 10880 }, { "grad_norm": 0.6483983397483826, "learning_rate": 9.905484587354766e-05, "loss": 0.0768, "step": 10890 }, { "grad_norm": 0.8571088314056396, "learning_rate": 9.905164344574205e-05, "loss": 0.0943, "step": 10900 }, { "grad_norm": 0.5763403177261353, "learning_rate": 9.904843565372248e-05, "loss": 0.0904, "step": 10910 }, { "grad_norm": 0.6459569931030273, "learning_rate": 9.904522249783972e-05, "loss": 0.0816, "step": 10920 }, { "grad_norm": 0.5484317541122437, "learning_rate": 9.904200397844517e-05, "loss": 0.109, "step": 10930 }, { "grad_norm": 0.7472326159477234, "learning_rate": 9.903878009589078e-05, "loss": 0.097, "step": 10940 }, { "grad_norm": 1.1499087810516357, "learning_rate": 9.903555085052915e-05, "loss": 0.0898, "step": 10950 }, { "grad_norm": 0.927630603313446, "learning_rate": 9.903231624271338e-05, "loss": 0.1101, "step": 10960 }, { "grad_norm": 0.8524746894836426, "learning_rate": 9.902907627279724e-05, "loss": 0.0984, "step": 10970 }, { "grad_norm": 0.6759746074676514, "learning_rate": 9.902583094113504e-05, "loss": 0.0946, "step": 10980 }, { "grad_norm": 0.8950694799423218, "learning_rate": 9.902258024808168e-05, "loss": 0.1007, "step": 10990 }, { "grad_norm": 0.7657372355461121, "learning_rate": 9.901932419399264e-05, "loss": 0.0999, "step": 11000 }, { "grad_norm": 0.6284401416778564, "learning_rate": 9.9016062779224e-05, "loss": 0.1076, "step": 11010 }, { "grad_norm": 0.9271935820579529, "learning_rate": 9.901279600413242e-05, "loss": 0.1105, "step": 11020 }, { "grad_norm": 1.081339716911316, "learning_rate": 9.900952386907518e-05, "loss": 0.1099, "step": 11030 }, { "grad_norm": 0.6274111270904541, "learning_rate": 9.90062463744101e-05, "loss": 0.0897, "step": 11040 }, { "grad_norm": 0.9863210916519165, "learning_rate": 9.900296352049558e-05, "loss": 0.1079, "step": 11050 }, { "grad_norm": 1.2015612125396729, "learning_rate": 9.899967530769065e-05, "loss": 0.0973, "step": 11060 }, { "grad_norm": 0.9442397952079773, "learning_rate": 9.899638173635489e-05, "loss": 0.0895, "step": 11070 }, { "grad_norm": 0.6987829208374023, "learning_rate": 9.899308280684849e-05, "loss": 0.0937, "step": 11080 }, { "grad_norm": 1.0580841302871704, "learning_rate": 9.898977851953222e-05, "loss": 0.109, "step": 11090 }, { "grad_norm": 0.8869590163230896, "learning_rate": 9.898646887476741e-05, "loss": 0.1038, "step": 11100 }, { "grad_norm": 1.080464243888855, "learning_rate": 9.898315387291603e-05, "loss": 0.1044, "step": 11110 }, { "grad_norm": 0.7338188886642456, "learning_rate": 9.89798335143406e-05, "loss": 0.0965, "step": 11120 }, { "grad_norm": 0.7095860242843628, "learning_rate": 9.897650779940419e-05, "loss": 0.126, "step": 11130 }, { "grad_norm": 0.6210036873817444, "learning_rate": 9.897317672847054e-05, "loss": 0.1054, "step": 11140 }, { "grad_norm": 1.320188045501709, "learning_rate": 9.89698403019039e-05, "loss": 0.1241, "step": 11150 }, { "grad_norm": 1.3613945245742798, "learning_rate": 9.896649852006917e-05, "loss": 0.1161, "step": 11160 }, { "grad_norm": 0.7565872073173523, "learning_rate": 9.896315138333177e-05, "loss": 0.0991, "step": 11170 }, { "grad_norm": 0.6074925065040588, "learning_rate": 9.895979889205774e-05, "loss": 0.0956, "step": 11180 }, { "grad_norm": 0.610516369342804, "learning_rate": 9.895644104661372e-05, "loss": 0.0991, "step": 11190 }, { "grad_norm": 0.617689311504364, "learning_rate": 9.895307784736691e-05, "loss": 0.1261, "step": 11200 }, { "grad_norm": 0.6525757312774658, "learning_rate": 9.894970929468512e-05, "loss": 0.1021, "step": 11210 }, { "grad_norm": 0.8138195276260376, "learning_rate": 9.89463353889367e-05, "loss": 0.0935, "step": 11220 }, { "grad_norm": 0.7867052555084229, "learning_rate": 9.894295613049065e-05, "loss": 0.0961, "step": 11230 }, { "grad_norm": 0.8161771893501282, "learning_rate": 9.893957151971649e-05, "loss": 0.1065, "step": 11240 }, { "grad_norm": 0.9354140758514404, "learning_rate": 9.893618155698436e-05, "loss": 0.1022, "step": 11250 }, { "grad_norm": 0.6808926463127136, "learning_rate": 9.8932786242665e-05, "loss": 0.0796, "step": 11260 }, { "grad_norm": 0.9054293036460876, "learning_rate": 9.89293855771297e-05, "loss": 0.0942, "step": 11270 }, { "grad_norm": 0.5981770157814026, "learning_rate": 9.892597956075036e-05, "loss": 0.0927, "step": 11280 }, { "grad_norm": 0.7748412489891052, "learning_rate": 9.892256819389947e-05, "loss": 0.0961, "step": 11290 }, { "grad_norm": 0.6127803921699524, "learning_rate": 9.891915147695006e-05, "loss": 0.1193, "step": 11300 }, { "grad_norm": 0.5147935152053833, "learning_rate": 9.891572941027577e-05, "loss": 0.1069, "step": 11310 }, { "grad_norm": 1.0172711610794067, "learning_rate": 9.89123019942509e-05, "loss": 0.109, "step": 11320 }, { "grad_norm": 0.666699230670929, "learning_rate": 9.89088692292502e-05, "loss": 0.1006, "step": 11330 }, { "grad_norm": 0.9771139621734619, "learning_rate": 9.89054311156491e-05, "loss": 0.1124, "step": 11340 }, { "grad_norm": 0.624506413936615, "learning_rate": 9.890198765382357e-05, "loss": 0.0858, "step": 11350 }, { "grad_norm": 0.7509952783584595, "learning_rate": 9.889853884415021e-05, "loss": 0.1152, "step": 11360 }, { "grad_norm": 0.7631583213806152, "learning_rate": 9.889508468700614e-05, "loss": 0.0849, "step": 11370 }, { "grad_norm": 0.7522403597831726, "learning_rate": 9.889162518276915e-05, "loss": 0.1185, "step": 11380 }, { "grad_norm": 0.6930897235870361, "learning_rate": 9.888816033181752e-05, "loss": 0.0832, "step": 11390 }, { "grad_norm": 1.0743483304977417, "learning_rate": 9.888469013453018e-05, "loss": 0.1119, "step": 11400 }, { "grad_norm": 0.8286439180374146, "learning_rate": 9.888121459128663e-05, "loss": 0.1063, "step": 11410 }, { "grad_norm": 1.1226645708084106, "learning_rate": 9.887773370246693e-05, "loss": 0.1211, "step": 11420 }, { "grad_norm": 0.8443123698234558, "learning_rate": 9.887424746845177e-05, "loss": 0.1055, "step": 11430 }, { "grad_norm": 0.8280768394470215, "learning_rate": 9.887075588962239e-05, "loss": 0.1082, "step": 11440 }, { "grad_norm": 0.9749234914779663, "learning_rate": 9.88672589663606e-05, "loss": 0.1005, "step": 11450 }, { "grad_norm": 0.8266438841819763, "learning_rate": 9.886375669904886e-05, "loss": 0.1017, "step": 11460 }, { "grad_norm": 0.9254885315895081, "learning_rate": 9.886024908807014e-05, "loss": 0.1194, "step": 11470 }, { "grad_norm": 0.6811612248420715, "learning_rate": 9.885673613380806e-05, "loss": 0.1223, "step": 11480 }, { "grad_norm": 0.742536187171936, "learning_rate": 9.885321783664676e-05, "loss": 0.1073, "step": 11490 }, { "grad_norm": 0.711036741733551, "learning_rate": 9.884969419697101e-05, "loss": 0.0867, "step": 11500 }, { "grad_norm": 0.8175861835479736, "learning_rate": 9.884616521516614e-05, "loss": 0.0993, "step": 11510 }, { "grad_norm": 0.7830586433410645, "learning_rate": 9.88426308916181e-05, "loss": 0.1114, "step": 11520 }, { "grad_norm": 0.7568894624710083, "learning_rate": 9.883909122671335e-05, "loss": 0.0937, "step": 11530 }, { "grad_norm": 0.8784993290901184, "learning_rate": 9.883554622083904e-05, "loss": 0.1248, "step": 11540 }, { "grad_norm": 0.8266193270683289, "learning_rate": 9.88319958743828e-05, "loss": 0.1247, "step": 11550 }, { "grad_norm": 0.7025633454322815, "learning_rate": 9.882844018773291e-05, "loss": 0.1057, "step": 11560 }, { "grad_norm": 0.7963440418243408, "learning_rate": 9.882487916127823e-05, "loss": 0.0888, "step": 11570 }, { "grad_norm": 0.8738244771957397, "learning_rate": 9.882131279540815e-05, "loss": 0.0989, "step": 11580 }, { "grad_norm": 0.7718095779418945, "learning_rate": 9.881774109051271e-05, "loss": 0.1164, "step": 11590 }, { "grad_norm": 0.8488677740097046, "learning_rate": 9.881416404698252e-05, "loss": 0.0974, "step": 11600 }, { "grad_norm": 0.6365064382553101, "learning_rate": 9.881058166520873e-05, "loss": 0.1093, "step": 11610 }, { "grad_norm": 0.9237587451934814, "learning_rate": 9.880699394558311e-05, "loss": 0.093, "step": 11620 }, { "grad_norm": 0.9913944602012634, "learning_rate": 9.880340088849801e-05, "loss": 0.1075, "step": 11630 }, { "grad_norm": 0.7769346237182617, "learning_rate": 9.879980249434637e-05, "loss": 0.1032, "step": 11640 }, { "grad_norm": 0.8126174211502075, "learning_rate": 9.879619876352168e-05, "loss": 0.1107, "step": 11650 }, { "grad_norm": 0.7458279728889465, "learning_rate": 9.879258969641809e-05, "loss": 0.1091, "step": 11660 }, { "grad_norm": 0.9638131260871887, "learning_rate": 9.878897529343023e-05, "loss": 0.1075, "step": 11670 }, { "grad_norm": 0.773190975189209, "learning_rate": 9.878535555495338e-05, "loss": 0.1031, "step": 11680 }, { "grad_norm": 0.968411386013031, "learning_rate": 9.87817304813834e-05, "loss": 0.1092, "step": 11690 }, { "grad_norm": 0.7275693416595459, "learning_rate": 9.877810007311671e-05, "loss": 0.11, "step": 11700 }, { "grad_norm": 0.6900421977043152, "learning_rate": 9.877446433055035e-05, "loss": 0.1362, "step": 11710 }, { "grad_norm": 0.9718241691589355, "learning_rate": 9.877082325408191e-05, "loss": 0.1037, "step": 11720 }, { "grad_norm": 1.1703181266784668, "learning_rate": 9.876717684410954e-05, "loss": 0.1131, "step": 11730 }, { "grad_norm": 0.8391615748405457, "learning_rate": 9.876352510103204e-05, "loss": 0.098, "step": 11740 }, { "grad_norm": 1.0247907638549805, "learning_rate": 9.875986802524875e-05, "loss": 0.1319, "step": 11750 }, { "grad_norm": 0.8152320384979248, "learning_rate": 9.87562056171596e-05, "loss": 0.111, "step": 11760 }, { "grad_norm": 0.8764631748199463, "learning_rate": 9.875253787716511e-05, "loss": 0.1169, "step": 11770 }, { "grad_norm": 0.6780527234077454, "learning_rate": 9.874886480566637e-05, "loss": 0.0924, "step": 11780 }, { "grad_norm": 0.616722047328949, "learning_rate": 9.874518640306507e-05, "loss": 0.097, "step": 11790 }, { "grad_norm": 0.6075611710548401, "learning_rate": 9.874150266976347e-05, "loss": 0.1338, "step": 11800 }, { "grad_norm": 0.9226823449134827, "learning_rate": 9.873781360616443e-05, "loss": 0.1105, "step": 11810 }, { "grad_norm": 0.7523388266563416, "learning_rate": 9.873411921267137e-05, "loss": 0.1178, "step": 11820 }, { "grad_norm": 0.7134493589401245, "learning_rate": 9.873041948968829e-05, "loss": 0.0932, "step": 11830 }, { "grad_norm": 0.7263906598091125, "learning_rate": 9.872671443761981e-05, "loss": 0.097, "step": 11840 }, { "grad_norm": 0.6408348679542542, "learning_rate": 9.872300405687109e-05, "loss": 0.0852, "step": 11850 }, { "grad_norm": 0.8190735578536987, "learning_rate": 9.871928834784792e-05, "loss": 0.0895, "step": 11860 }, { "grad_norm": 0.6823885440826416, "learning_rate": 9.871556731095661e-05, "loss": 0.0968, "step": 11870 }, { "grad_norm": 0.5951379537582397, "learning_rate": 9.871184094660411e-05, "loss": 0.1136, "step": 11880 }, { "grad_norm": 0.9723508358001709, "learning_rate": 9.870810925519791e-05, "loss": 0.109, "step": 11890 }, { "grad_norm": 0.7926439046859741, "learning_rate": 9.870437223714612e-05, "loss": 0.1056, "step": 11900 }, { "grad_norm": 0.8091016411781311, "learning_rate": 9.87006298928574e-05, "loss": 0.0905, "step": 11910 }, { "grad_norm": 0.5466582775115967, "learning_rate": 9.869688222274103e-05, "loss": 0.0829, "step": 11920 }, { "grad_norm": 0.7805463671684265, "learning_rate": 9.869312922720681e-05, "loss": 0.0863, "step": 11930 }, { "grad_norm": 1.238135576248169, "learning_rate": 9.868937090666521e-05, "loss": 0.1102, "step": 11940 }, { "grad_norm": 0.7577864527702332, "learning_rate": 9.86856072615272e-05, "loss": 0.099, "step": 11950 }, { "grad_norm": 0.6667923331260681, "learning_rate": 9.868183829220438e-05, "loss": 0.0997, "step": 11960 }, { "grad_norm": 0.9932202696800232, "learning_rate": 9.867806399910893e-05, "loss": 0.1051, "step": 11970 }, { "grad_norm": 0.7182154655456543, "learning_rate": 9.867428438265356e-05, "loss": 0.1283, "step": 11980 }, { "grad_norm": 0.8174052834510803, "learning_rate": 9.867049944325165e-05, "loss": 0.1112, "step": 11990 }, { "grad_norm": 0.6426248550415039, "learning_rate": 9.86667091813171e-05, "loss": 0.153, "step": 12000 }, { "grad_norm": 0.7268158793449402, "learning_rate": 9.866291359726438e-05, "loss": 0.1075, "step": 12010 }, { "grad_norm": 0.7338287830352783, "learning_rate": 9.865911269150861e-05, "loss": 0.0927, "step": 12020 }, { "grad_norm": 0.5791228413581848, "learning_rate": 9.865530646446544e-05, "loss": 0.093, "step": 12030 }, { "grad_norm": 0.9894895553588867, "learning_rate": 9.86514949165511e-05, "loss": 0.1029, "step": 12040 }, { "grad_norm": 0.7926651239395142, "learning_rate": 9.864767804818243e-05, "loss": 0.1052, "step": 12050 }, { "grad_norm": 0.5623246431350708, "learning_rate": 9.86438558597768e-05, "loss": 0.0924, "step": 12060 }, { "grad_norm": 0.8127744793891907, "learning_rate": 9.864002835175225e-05, "loss": 0.1144, "step": 12070 }, { "grad_norm": 0.6601552367210388, "learning_rate": 9.863619552452734e-05, "loss": 0.1102, "step": 12080 }, { "grad_norm": 1.0438287258148193, "learning_rate": 9.863235737852119e-05, "loss": 0.1128, "step": 12090 }, { "grad_norm": 0.9125041365623474, "learning_rate": 9.862851391415356e-05, "loss": 0.0974, "step": 12100 }, { "grad_norm": 0.8950212001800537, "learning_rate": 9.862466513184477e-05, "loss": 0.1033, "step": 12110 }, { "grad_norm": 0.7335769534111023, "learning_rate": 9.86208110320157e-05, "loss": 0.0788, "step": 12120 }, { "grad_norm": 1.1577171087265015, "learning_rate": 9.861695161508784e-05, "loss": 0.1116, "step": 12130 }, { "grad_norm": 0.9521828889846802, "learning_rate": 9.861308688148324e-05, "loss": 0.1086, "step": 12140 }, { "grad_norm": 0.8539612889289856, "learning_rate": 9.860921683162455e-05, "loss": 0.118, "step": 12150 }, { "grad_norm": 0.9020950794219971, "learning_rate": 9.860534146593499e-05, "loss": 0.1094, "step": 12160 }, { "grad_norm": 0.7546865940093994, "learning_rate": 9.860146078483836e-05, "loss": 0.109, "step": 12170 }, { "grad_norm": 0.8086910247802734, "learning_rate": 9.859757478875905e-05, "loss": 0.1137, "step": 12180 }, { "grad_norm": 0.7223188877105713, "learning_rate": 9.859368347812204e-05, "loss": 0.1126, "step": 12190 }, { "grad_norm": 0.6892596483230591, "learning_rate": 9.858978685335285e-05, "loss": 0.1129, "step": 12200 }, { "grad_norm": 0.8918875455856323, "learning_rate": 9.858588491487763e-05, "loss": 0.0771, "step": 12210 }, { "grad_norm": 0.9071831703186035, "learning_rate": 9.858197766312308e-05, "loss": 0.0939, "step": 12220 }, { "grad_norm": 0.6949076652526855, "learning_rate": 9.857806509851649e-05, "loss": 0.0827, "step": 12230 }, { "grad_norm": 0.6469933986663818, "learning_rate": 9.857414722148574e-05, "loss": 0.1064, "step": 12240 }, { "grad_norm": 1.0746058225631714, "learning_rate": 9.857022403245928e-05, "loss": 0.093, "step": 12250 }, { "grad_norm": 0.6070605516433716, "learning_rate": 9.856629553186615e-05, "loss": 0.0864, "step": 12260 }, { "grad_norm": 0.8219984769821167, "learning_rate": 9.856236172013595e-05, "loss": 0.0917, "step": 12270 }, { "grad_norm": 0.6769071817398071, "learning_rate": 9.85584225976989e-05, "loss": 0.1018, "step": 12280 }, { "grad_norm": 0.7362833023071289, "learning_rate": 9.855447816498575e-05, "loss": 0.103, "step": 12290 }, { "grad_norm": 1.012410283088684, "learning_rate": 9.855052842242787e-05, "loss": 0.1219, "step": 12300 }, { "grad_norm": 0.9796984791755676, "learning_rate": 9.85465733704572e-05, "loss": 0.119, "step": 12310 }, { "grad_norm": 0.9646763801574707, "learning_rate": 9.854261300950624e-05, "loss": 0.1346, "step": 12320 }, { "grad_norm": 0.8365620374679565, "learning_rate": 9.853864734000813e-05, "loss": 0.139, "step": 12330 }, { "grad_norm": 0.757664680480957, "learning_rate": 9.85346763623965e-05, "loss": 0.1225, "step": 12340 }, { "grad_norm": 0.888119637966156, "learning_rate": 9.853070007710564e-05, "loss": 0.1047, "step": 12350 }, { "grad_norm": 0.6333223581314087, "learning_rate": 9.85267184845704e-05, "loss": 0.0913, "step": 12360 }, { "grad_norm": 1.1212773323059082, "learning_rate": 9.852273158522616e-05, "loss": 0.1198, "step": 12370 }, { "grad_norm": 0.8257589340209961, "learning_rate": 9.851873937950896e-05, "loss": 0.1061, "step": 12380 }, { "grad_norm": 0.825764000415802, "learning_rate": 9.851474186785537e-05, "loss": 0.1112, "step": 12390 }, { "grad_norm": 0.7188701629638672, "learning_rate": 9.851073905070254e-05, "loss": 0.0767, "step": 12400 }, { "grad_norm": 0.6524165272712708, "learning_rate": 9.850673092848824e-05, "loss": 0.0995, "step": 12410 }, { "grad_norm": 0.9336461424827576, "learning_rate": 9.850271750165077e-05, "loss": 0.1029, "step": 12420 }, { "grad_norm": 0.7432175874710083, "learning_rate": 9.849869877062902e-05, "loss": 0.0921, "step": 12430 }, { "grad_norm": 0.8444298505783081, "learning_rate": 9.849467473586252e-05, "loss": 0.0854, "step": 12440 }, { "grad_norm": 0.7287114262580872, "learning_rate": 9.849064539779127e-05, "loss": 0.1105, "step": 12450 }, { "grad_norm": 0.9708828330039978, "learning_rate": 9.848661075685594e-05, "loss": 0.098, "step": 12460 }, { "grad_norm": 0.6952724456787109, "learning_rate": 9.848257081349778e-05, "loss": 0.1223, "step": 12470 }, { "grad_norm": 0.8116122484207153, "learning_rate": 9.847852556815856e-05, "loss": 0.1014, "step": 12480 }, { "grad_norm": 0.8632642030715942, "learning_rate": 9.847447502128067e-05, "loss": 0.1039, "step": 12490 }, { "grad_norm": 0.600843071937561, "learning_rate": 9.847041917330708e-05, "loss": 0.112, "step": 12500 }, { "grad_norm": 0.49701884388923645, "learning_rate": 9.846635802468132e-05, "loss": 0.0711, "step": 12510 }, { "grad_norm": 0.7632274031639099, "learning_rate": 9.84622915758475e-05, "loss": 0.0805, "step": 12520 }, { "grad_norm": 0.7884507179260254, "learning_rate": 9.845821982725034e-05, "loss": 0.0838, "step": 12530 }, { "grad_norm": 0.6349719166755676, "learning_rate": 9.845414277933514e-05, "loss": 0.0988, "step": 12540 }, { "grad_norm": 0.49430301785469055, "learning_rate": 9.845006043254771e-05, "loss": 0.0826, "step": 12550 }, { "grad_norm": 0.7683109641075134, "learning_rate": 9.844597278733451e-05, "loss": 0.089, "step": 12560 }, { "grad_norm": 0.8270992636680603, "learning_rate": 9.844187984414259e-05, "loss": 0.1009, "step": 12570 }, { "grad_norm": 0.6921969056129456, "learning_rate": 9.84377816034195e-05, "loss": 0.0997, "step": 12580 }, { "grad_norm": 0.9006268978118896, "learning_rate": 9.843367806561345e-05, "loss": 0.1164, "step": 12590 }, { "grad_norm": 0.729813814163208, "learning_rate": 9.842956923117317e-05, "loss": 0.096, "step": 12600 }, { "grad_norm": 0.6496661901473999, "learning_rate": 9.842545510054802e-05, "loss": 0.0976, "step": 12610 }, { "grad_norm": 0.8188395500183105, "learning_rate": 9.842133567418792e-05, "loss": 0.0987, "step": 12620 }, { "grad_norm": 0.6164426207542419, "learning_rate": 9.841721095254333e-05, "loss": 0.0963, "step": 12630 }, { "grad_norm": 1.3616766929626465, "learning_rate": 9.841308093606537e-05, "loss": 0.104, "step": 12640 }, { "grad_norm": 0.8063573241233826, "learning_rate": 9.840894562520565e-05, "loss": 0.1085, "step": 12650 }, { "grad_norm": 0.7272081971168518, "learning_rate": 9.840480502041642e-05, "loss": 0.0804, "step": 12660 }, { "grad_norm": 1.12986421585083, "learning_rate": 9.840065912215049e-05, "loss": 0.1215, "step": 12670 }, { "grad_norm": 1.035656213760376, "learning_rate": 9.839650793086124e-05, "loss": 0.0915, "step": 12680 }, { "grad_norm": 0.6597587466239929, "learning_rate": 9.839235144700265e-05, "loss": 0.1028, "step": 12690 }, { "grad_norm": 1.335968017578125, "learning_rate": 9.838818967102926e-05, "loss": 0.0829, "step": 12700 }, { "grad_norm": 0.7940053939819336, "learning_rate": 9.83840226033962e-05, "loss": 0.106, "step": 12710 }, { "grad_norm": 0.7575740814208984, "learning_rate": 9.837985024455918e-05, "loss": 0.0841, "step": 12720 }, { "grad_norm": 0.7014760375022888, "learning_rate": 9.837567259497447e-05, "loss": 0.0808, "step": 12730 }, { "grad_norm": 0.8305689096450806, "learning_rate": 9.837148965509894e-05, "loss": 0.0945, "step": 12740 }, { "grad_norm": 0.7173106074333191, "learning_rate": 9.836730142539001e-05, "loss": 0.081, "step": 12750 }, { "grad_norm": 0.6250150203704834, "learning_rate": 9.836310790630574e-05, "loss": 0.0974, "step": 12760 }, { "grad_norm": 0.8723942637443542, "learning_rate": 9.83589090983047e-05, "loss": 0.1103, "step": 12770 }, { "grad_norm": 1.0921318531036377, "learning_rate": 9.835470500184605e-05, "loss": 0.1254, "step": 12780 }, { "grad_norm": 0.7604947686195374, "learning_rate": 9.835049561738957e-05, "loss": 0.0876, "step": 12790 }, { "grad_norm": 0.7019551992416382, "learning_rate": 9.834628094539558e-05, "loss": 0.1229, "step": 12800 }, { "grad_norm": 0.7513987421989441, "learning_rate": 9.834206098632499e-05, "loss": 0.0882, "step": 12810 }, { "grad_norm": 0.834708034992218, "learning_rate": 9.833783574063931e-05, "loss": 0.0929, "step": 12820 }, { "grad_norm": 0.7582963705062866, "learning_rate": 9.833360520880058e-05, "loss": 0.115, "step": 12830 }, { "grad_norm": 0.6747496128082275, "learning_rate": 9.832936939127144e-05, "loss": 0.0817, "step": 12840 }, { "grad_norm": 0.7840268015861511, "learning_rate": 9.832512828851515e-05, "loss": 0.0786, "step": 12850 }, { "grad_norm": 0.6674308180809021, "learning_rate": 9.832088190099546e-05, "loss": 0.0882, "step": 12860 }, { "grad_norm": 0.8370606899261475, "learning_rate": 9.831663022917679e-05, "loss": 0.0925, "step": 12870 }, { "grad_norm": 0.5077400803565979, "learning_rate": 9.831237327352407e-05, "loss": 0.0853, "step": 12880 }, { "grad_norm": 0.6509312391281128, "learning_rate": 9.830811103450286e-05, "loss": 0.09, "step": 12890 }, { "grad_norm": 0.829326868057251, "learning_rate": 9.830384351257924e-05, "loss": 0.0951, "step": 12900 }, { "grad_norm": 0.7870033383369446, "learning_rate": 9.829957070821993e-05, "loss": 0.0949, "step": 12910 }, { "grad_norm": 0.6149786114692688, "learning_rate": 9.829529262189218e-05, "loss": 0.0917, "step": 12920 }, { "grad_norm": 0.8465380668640137, "learning_rate": 9.829100925406385e-05, "loss": 0.0982, "step": 12930 }, { "grad_norm": 0.8918927907943726, "learning_rate": 9.828672060520333e-05, "loss": 0.0902, "step": 12940 }, { "grad_norm": 0.5109763741493225, "learning_rate": 9.828242667577966e-05, "loss": 0.0778, "step": 12950 }, { "grad_norm": 0.8425917029380798, "learning_rate": 9.82781274662624e-05, "loss": 0.1003, "step": 12960 }, { "grad_norm": 0.9031098484992981, "learning_rate": 9.82738229771217e-05, "loss": 0.1016, "step": 12970 }, { "grad_norm": 0.8883136510848999, "learning_rate": 9.826951320882829e-05, "loss": 0.1159, "step": 12980 }, { "grad_norm": 0.5969290137290955, "learning_rate": 9.826519816185351e-05, "loss": 0.0815, "step": 12990 }, { "grad_norm": 0.6718582510948181, "learning_rate": 9.826087783666921e-05, "loss": 0.0732, "step": 13000 }, { "grad_norm": 0.668758749961853, "learning_rate": 9.825655223374787e-05, "loss": 0.0817, "step": 13010 }, { "grad_norm": 0.7269002199172974, "learning_rate": 9.825222135356253e-05, "loss": 0.0962, "step": 13020 }, { "grad_norm": 0.8675873875617981, "learning_rate": 9.82478851965868e-05, "loss": 0.0942, "step": 13030 }, { "grad_norm": 0.7726803421974182, "learning_rate": 9.82435437632949e-05, "loss": 0.101, "step": 13040 }, { "grad_norm": 0.8919403553009033, "learning_rate": 9.823919705416158e-05, "loss": 0.0935, "step": 13050 }, { "grad_norm": 0.7390616536140442, "learning_rate": 9.82348450696622e-05, "loss": 0.0846, "step": 13060 }, { "grad_norm": 0.8063105344772339, "learning_rate": 9.823048781027268e-05, "loss": 0.0982, "step": 13070 }, { "grad_norm": 0.9884769320487976, "learning_rate": 9.822612527646953e-05, "loss": 0.1001, "step": 13080 }, { "grad_norm": 1.0010490417480469, "learning_rate": 9.822175746872984e-05, "loss": 0.0976, "step": 13090 }, { "grad_norm": 0.7120330333709717, "learning_rate": 9.821738438753123e-05, "loss": 0.0829, "step": 13100 }, { "grad_norm": 0.7963780164718628, "learning_rate": 9.821300603335196e-05, "loss": 0.112, "step": 13110 }, { "grad_norm": 0.7323761582374573, "learning_rate": 9.820862240667085e-05, "loss": 0.1016, "step": 13120 }, { "grad_norm": 0.8338249325752258, "learning_rate": 9.820423350796726e-05, "loss": 0.0969, "step": 13130 }, { "grad_norm": 0.6795777082443237, "learning_rate": 9.819983933772118e-05, "loss": 0.1187, "step": 13140 }, { "grad_norm": 0.9416393041610718, "learning_rate": 9.819543989641314e-05, "loss": 0.0877, "step": 13150 }, { "grad_norm": 0.6675598621368408, "learning_rate": 9.819103518452423e-05, "loss": 0.0969, "step": 13160 }, { "grad_norm": 0.8978418707847595, "learning_rate": 9.818662520253618e-05, "loss": 0.1225, "step": 13170 }, { "grad_norm": 0.6404996514320374, "learning_rate": 9.818220995093126e-05, "loss": 0.1092, "step": 13180 }, { "grad_norm": 1.2347018718719482, "learning_rate": 9.817778943019228e-05, "loss": 0.1074, "step": 13190 }, { "grad_norm": 0.5849775671958923, "learning_rate": 9.81733636408027e-05, "loss": 0.1014, "step": 13200 }, { "grad_norm": 0.6151407361030579, "learning_rate": 9.816893258324649e-05, "loss": 0.0968, "step": 13210 }, { "grad_norm": 0.6757089495658875, "learning_rate": 9.816449625800823e-05, "loss": 0.0978, "step": 13220 }, { "grad_norm": 1.2746092081069946, "learning_rate": 9.816005466557308e-05, "loss": 0.0901, "step": 13230 }, { "grad_norm": 0.9705587029457092, "learning_rate": 9.815560780642674e-05, "loss": 0.0741, "step": 13240 }, { "grad_norm": 0.7503018379211426, "learning_rate": 9.815115568105555e-05, "loss": 0.0998, "step": 13250 }, { "grad_norm": 0.8353606462478638, "learning_rate": 9.814669828994638e-05, "loss": 0.0821, "step": 13260 }, { "grad_norm": 0.6047099232673645, "learning_rate": 9.814223563358665e-05, "loss": 0.0731, "step": 13270 }, { "grad_norm": 0.4663007855415344, "learning_rate": 9.813776771246443e-05, "loss": 0.0846, "step": 13280 }, { "grad_norm": 0.6818801164627075, "learning_rate": 9.813329452706829e-05, "loss": 0.0825, "step": 13290 }, { "grad_norm": 0.7244171500205994, "learning_rate": 9.812881607788744e-05, "loss": 0.1235, "step": 13300 }, { "grad_norm": 0.7992056012153625, "learning_rate": 9.812433236541163e-05, "loss": 0.0904, "step": 13310 }, { "grad_norm": 0.6421704888343811, "learning_rate": 9.811984339013116e-05, "loss": 0.0849, "step": 13320 }, { "grad_norm": 0.7281742691993713, "learning_rate": 9.811534915253698e-05, "loss": 0.0953, "step": 13330 }, { "grad_norm": 0.5401804447174072, "learning_rate": 9.811084965312056e-05, "loss": 0.0776, "step": 13340 }, { "grad_norm": 0.7067269086837769, "learning_rate": 9.810634489237396e-05, "loss": 0.1046, "step": 13350 }, { "grad_norm": 0.7424843311309814, "learning_rate": 9.81018348707898e-05, "loss": 0.0981, "step": 13360 }, { "grad_norm": 0.9124297499656677, "learning_rate": 9.809731958886131e-05, "loss": 0.1114, "step": 13370 }, { "grad_norm": 0.5591872334480286, "learning_rate": 9.809279904708224e-05, "loss": 0.0907, "step": 13380 }, { "grad_norm": 0.5454598665237427, "learning_rate": 9.808827324594699e-05, "loss": 0.0878, "step": 13390 }, { "grad_norm": 0.7287442088127136, "learning_rate": 9.808374218595046e-05, "loss": 0.1011, "step": 13400 }, { "grad_norm": 0.542995810508728, "learning_rate": 9.80792058675882e-05, "loss": 0.0973, "step": 13410 }, { "grad_norm": 0.9890803098678589, "learning_rate": 9.807466429135627e-05, "loss": 0.0953, "step": 13420 }, { "grad_norm": 0.6417820453643799, "learning_rate": 9.807011745775132e-05, "loss": 0.1137, "step": 13430 }, { "grad_norm": 0.634846031665802, "learning_rate": 9.806556536727061e-05, "loss": 0.0988, "step": 13440 }, { "grad_norm": 0.6398778557777405, "learning_rate": 9.806100802041193e-05, "loss": 0.0918, "step": 13450 }, { "grad_norm": 0.8958592414855957, "learning_rate": 9.805644541767368e-05, "loss": 0.1099, "step": 13460 }, { "grad_norm": 1.1624759435653687, "learning_rate": 9.805187755955478e-05, "loss": 0.1449, "step": 13470 }, { "grad_norm": 0.5513485670089722, "learning_rate": 9.804730444655483e-05, "loss": 0.0916, "step": 13480 }, { "grad_norm": 0.8739791512489319, "learning_rate": 9.804272607917388e-05, "loss": 0.0865, "step": 13490 }, { "grad_norm": 0.7167856693267822, "learning_rate": 9.803814245791265e-05, "loss": 0.085, "step": 13500 }, { "grad_norm": 0.7442106008529663, "learning_rate": 9.803355358327239e-05, "loss": 0.0977, "step": 13510 }, { "grad_norm": 0.709531307220459, "learning_rate": 9.802895945575492e-05, "loss": 0.1013, "step": 13520 }, { "grad_norm": 0.8674237132072449, "learning_rate": 9.802436007586266e-05, "loss": 0.1034, "step": 13530 }, { "grad_norm": 0.7355494499206543, "learning_rate": 9.801975544409858e-05, "loss": 0.0742, "step": 13540 }, { "grad_norm": 1.1350960731506348, "learning_rate": 9.801514556096625e-05, "loss": 0.1042, "step": 13550 }, { "grad_norm": 0.7587882280349731, "learning_rate": 9.801053042696977e-05, "loss": 0.1082, "step": 13560 }, { "grad_norm": 0.7523324489593506, "learning_rate": 9.800591004261388e-05, "loss": 0.1029, "step": 13570 }, { "grad_norm": 0.6743282079696655, "learning_rate": 9.800128440840385e-05, "loss": 0.0817, "step": 13580 }, { "grad_norm": 0.6156467795372009, "learning_rate": 9.799665352484552e-05, "loss": 0.0832, "step": 13590 }, { "grad_norm": 0.7317350506782532, "learning_rate": 9.799201739244532e-05, "loss": 0.119, "step": 13600 }, { "grad_norm": 0.7033385038375854, "learning_rate": 9.798737601171025e-05, "loss": 0.0793, "step": 13610 }, { "grad_norm": 0.7141819000244141, "learning_rate": 9.79827293831479e-05, "loss": 0.0859, "step": 13620 }, { "grad_norm": 0.869807779788971, "learning_rate": 9.797807750726638e-05, "loss": 0.0837, "step": 13630 }, { "grad_norm": 0.8348325490951538, "learning_rate": 9.797342038457446e-05, "loss": 0.1058, "step": 13640 }, { "grad_norm": 0.6120017170906067, "learning_rate": 9.796875801558141e-05, "loss": 0.0846, "step": 13650 }, { "grad_norm": 0.6865276098251343, "learning_rate": 9.79640904007971e-05, "loss": 0.0862, "step": 13660 }, { "grad_norm": 0.6246527433395386, "learning_rate": 9.795941754073199e-05, "loss": 0.111, "step": 13670 }, { "grad_norm": 0.9060617685317993, "learning_rate": 9.795473943589705e-05, "loss": 0.0921, "step": 13680 }, { "grad_norm": 0.8822085857391357, "learning_rate": 9.795005608680394e-05, "loss": 0.1035, "step": 13690 }, { "grad_norm": 0.725045919418335, "learning_rate": 9.794536749396477e-05, "loss": 0.0935, "step": 13700 }, { "grad_norm": 1.0422370433807373, "learning_rate": 9.79406736578923e-05, "loss": 0.1075, "step": 13710 }, { "grad_norm": 0.6858001947402954, "learning_rate": 9.793597457909984e-05, "loss": 0.0716, "step": 13720 }, { "grad_norm": 0.6307905912399292, "learning_rate": 9.793127025810127e-05, "loss": 0.1186, "step": 13730 }, { "grad_norm": 0.8295782804489136, "learning_rate": 9.792656069541104e-05, "loss": 0.1046, "step": 13740 }, { "grad_norm": 0.7912870645523071, "learning_rate": 9.79218458915442e-05, "loss": 0.0853, "step": 13750 }, { "grad_norm": 0.9891495108604431, "learning_rate": 9.791712584701634e-05, "loss": 0.1212, "step": 13760 }, { "grad_norm": 0.9253432750701904, "learning_rate": 9.791240056234364e-05, "loss": 0.1159, "step": 13770 }, { "grad_norm": 0.7418815493583679, "learning_rate": 9.790767003804283e-05, "loss": 0.0932, "step": 13780 }, { "grad_norm": 0.5998997688293457, "learning_rate": 9.790293427463126e-05, "loss": 0.084, "step": 13790 }, { "grad_norm": 0.728608250617981, "learning_rate": 9.789819327262684e-05, "loss": 0.1147, "step": 13800 }, { "grad_norm": 0.7772880792617798, "learning_rate": 9.7893447032548e-05, "loss": 0.1084, "step": 13810 }, { "grad_norm": 0.6307204961776733, "learning_rate": 9.78886955549138e-05, "loss": 0.1027, "step": 13820 }, { "grad_norm": 0.6902729272842407, "learning_rate": 9.788393884024387e-05, "loss": 0.1022, "step": 13830 }, { "grad_norm": 0.5415530800819397, "learning_rate": 9.787917688905836e-05, "loss": 0.1124, "step": 13840 }, { "grad_norm": 0.9573285579681396, "learning_rate": 9.787440970187807e-05, "loss": 0.098, "step": 13850 }, { "grad_norm": 0.8002467155456543, "learning_rate": 9.786963727922429e-05, "loss": 0.0927, "step": 13860 }, { "grad_norm": 0.9968522787094116, "learning_rate": 9.786485962161897e-05, "loss": 0.0874, "step": 13870 }, { "grad_norm": 0.7990223169326782, "learning_rate": 9.786007672958455e-05, "loss": 0.0781, "step": 13880 }, { "grad_norm": 0.7025278210639954, "learning_rate": 9.78552886036441e-05, "loss": 0.0739, "step": 13890 }, { "grad_norm": 0.785295307636261, "learning_rate": 9.785049524432124e-05, "loss": 0.078, "step": 13900 }, { "grad_norm": 0.8606801629066467, "learning_rate": 9.784569665214016e-05, "loss": 0.1074, "step": 13910 }, { "grad_norm": 0.8379659056663513, "learning_rate": 9.784089282762563e-05, "loss": 0.096, "step": 13920 }, { "grad_norm": 0.4727950096130371, "learning_rate": 9.7836083771303e-05, "loss": 0.093, "step": 13930 }, { "grad_norm": 0.5432344079017639, "learning_rate": 9.783126948369817e-05, "loss": 0.0925, "step": 13940 }, { "grad_norm": 0.7275602221488953, "learning_rate": 9.78264499653376e-05, "loss": 0.0832, "step": 13950 }, { "grad_norm": 1.2274068593978882, "learning_rate": 9.782162521674838e-05, "loss": 0.123, "step": 13960 }, { "grad_norm": 0.6397491097450256, "learning_rate": 9.781679523845812e-05, "loss": 0.0735, "step": 13970 }, { "grad_norm": 0.8157382607460022, "learning_rate": 9.781196003099502e-05, "loss": 0.0832, "step": 13980 }, { "grad_norm": 1.0276471376419067, "learning_rate": 9.780711959488786e-05, "loss": 0.1094, "step": 13990 }, { "grad_norm": 0.7714353799819946, "learning_rate": 9.780227393066599e-05, "loss": 0.0819, "step": 14000 }, { "grad_norm": 0.6892088651657104, "learning_rate": 9.77974230388593e-05, "loss": 0.0944, "step": 14010 }, { "grad_norm": 0.9490200281143188, "learning_rate": 9.779256691999829e-05, "loss": 0.1276, "step": 14020 }, { "grad_norm": 0.6877714395523071, "learning_rate": 9.778770557461403e-05, "loss": 0.0891, "step": 14030 }, { "grad_norm": 0.7763106226921082, "learning_rate": 9.778283900323812e-05, "loss": 0.0919, "step": 14040 }, { "grad_norm": 0.6545308232307434, "learning_rate": 9.777796720640277e-05, "loss": 0.1396, "step": 14050 }, { "grad_norm": 0.8268513679504395, "learning_rate": 9.777309018464078e-05, "loss": 0.0936, "step": 14060 }, { "grad_norm": 0.8413519859313965, "learning_rate": 9.776820793848547e-05, "loss": 0.0742, "step": 14070 }, { "grad_norm": 0.8192014098167419, "learning_rate": 9.776332046847075e-05, "loss": 0.0935, "step": 14080 }, { "grad_norm": 0.6019310355186462, "learning_rate": 9.775842777513111e-05, "loss": 0.074, "step": 14090 }, { "grad_norm": 0.7724180817604065, "learning_rate": 9.775352985900163e-05, "loss": 0.0812, "step": 14100 }, { "grad_norm": 0.7046501636505127, "learning_rate": 9.774862672061791e-05, "loss": 0.0755, "step": 14110 }, { "grad_norm": 0.7000584602355957, "learning_rate": 9.774371836051616e-05, "loss": 0.0938, "step": 14120 }, { "grad_norm": 0.6570112109184265, "learning_rate": 9.773880477923315e-05, "loss": 0.0797, "step": 14130 }, { "grad_norm": 0.6103637218475342, "learning_rate": 9.773388597730623e-05, "loss": 0.0906, "step": 14140 }, { "grad_norm": 0.7732371687889099, "learning_rate": 9.77289619552733e-05, "loss": 0.1027, "step": 14150 }, { "grad_norm": 0.5172204375267029, "learning_rate": 9.772403271367285e-05, "loss": 0.1028, "step": 14160 }, { "grad_norm": 0.8006776571273804, "learning_rate": 9.771909825304396e-05, "loss": 0.1044, "step": 14170 }, { "grad_norm": 1.0537868738174438, "learning_rate": 9.771415857392619e-05, "loss": 0.1139, "step": 14180 }, { "grad_norm": 1.14788019657135, "learning_rate": 9.770921367685978e-05, "loss": 0.0758, "step": 14190 }, { "grad_norm": 1.0938217639923096, "learning_rate": 9.770426356238551e-05, "loss": 0.0994, "step": 14200 }, { "grad_norm": 0.7311357259750366, "learning_rate": 9.769930823104469e-05, "loss": 0.1299, "step": 14210 }, { "grad_norm": 0.7175320982933044, "learning_rate": 9.769434768337926e-05, "loss": 0.0873, "step": 14220 }, { "grad_norm": 0.7130507230758667, "learning_rate": 9.768938191993164e-05, "loss": 0.0739, "step": 14230 }, { "grad_norm": 0.8408972024917603, "learning_rate": 9.768441094124494e-05, "loss": 0.0803, "step": 14240 }, { "grad_norm": 0.5726144313812256, "learning_rate": 9.767943474786275e-05, "loss": 0.0833, "step": 14250 }, { "grad_norm": 1.0743283033370972, "learning_rate": 9.767445334032923e-05, "loss": 0.0837, "step": 14260 }, { "grad_norm": 0.7622365355491638, "learning_rate": 9.766946671918919e-05, "loss": 0.0891, "step": 14270 }, { "grad_norm": 0.871010422706604, "learning_rate": 9.766447488498796e-05, "loss": 0.1034, "step": 14280 }, { "grad_norm": 0.7474870085716248, "learning_rate": 9.765947783827139e-05, "loss": 0.1079, "step": 14290 }, { "grad_norm": 0.8612173199653625, "learning_rate": 9.765447557958599e-05, "loss": 0.1096, "step": 14300 }, { "grad_norm": 0.7246222496032715, "learning_rate": 9.764946810947879e-05, "loss": 0.0886, "step": 14310 }, { "grad_norm": 1.3313300609588623, "learning_rate": 9.764445542849738e-05, "loss": 0.1221, "step": 14320 }, { "grad_norm": 1.2705864906311035, "learning_rate": 9.763943753718998e-05, "loss": 0.104, "step": 14330 }, { "grad_norm": 0.8348579406738281, "learning_rate": 9.76344144361053e-05, "loss": 0.0983, "step": 14340 }, { "grad_norm": 0.6073992848396301, "learning_rate": 9.762938612579269e-05, "loss": 0.0753, "step": 14350 }, { "grad_norm": 0.8272507786750793, "learning_rate": 9.762435260680202e-05, "loss": 0.0969, "step": 14360 }, { "grad_norm": 0.8708973526954651, "learning_rate": 9.761931387968373e-05, "loss": 0.1146, "step": 14370 }, { "grad_norm": 0.7699762582778931, "learning_rate": 9.76142699449889e-05, "loss": 0.0983, "step": 14380 }, { "grad_norm": 0.7698308825492859, "learning_rate": 9.760922080326908e-05, "loss": 0.0843, "step": 14390 }, { "grad_norm": 0.901005208492279, "learning_rate": 9.760416645507644e-05, "loss": 0.0704, "step": 14400 }, { "grad_norm": 0.9459502696990967, "learning_rate": 9.759910690096375e-05, "loss": 0.076, "step": 14410 }, { "grad_norm": 0.7755318880081177, "learning_rate": 9.759404214148429e-05, "loss": 0.1045, "step": 14420 }, { "grad_norm": 0.8898059129714966, "learning_rate": 9.758897217719191e-05, "loss": 0.0994, "step": 14430 }, { "grad_norm": 0.7176892161369324, "learning_rate": 9.758389700864113e-05, "loss": 0.1078, "step": 14440 }, { "grad_norm": 0.7349691390991211, "learning_rate": 9.757881663638688e-05, "loss": 0.0896, "step": 14450 }, { "grad_norm": 0.47904860973358154, "learning_rate": 9.757373106098478e-05, "loss": 0.0827, "step": 14460 }, { "grad_norm": 0.8592793941497803, "learning_rate": 9.756864028299097e-05, "loss": 0.0956, "step": 14470 }, { "grad_norm": 0.7002658247947693, "learning_rate": 9.75635443029622e-05, "loss": 0.0952, "step": 14480 }, { "grad_norm": 0.7097303867340088, "learning_rate": 9.755844312145572e-05, "loss": 0.1022, "step": 14490 }, { "grad_norm": 0.630409836769104, "learning_rate": 9.755333673902941e-05, "loss": 0.0899, "step": 14500 }, { "grad_norm": 0.660174548625946, "learning_rate": 9.75482251562417e-05, "loss": 0.0868, "step": 14510 }, { "grad_norm": 0.769241213798523, "learning_rate": 9.754310837365155e-05, "loss": 0.0729, "step": 14520 }, { "grad_norm": 0.7489430904388428, "learning_rate": 9.753798639181856e-05, "loss": 0.0906, "step": 14530 }, { "grad_norm": 0.6417179703712463, "learning_rate": 9.753285921130286e-05, "loss": 0.0691, "step": 14540 }, { "grad_norm": 0.7082772850990295, "learning_rate": 9.752772683266512e-05, "loss": 0.0889, "step": 14550 }, { "grad_norm": 0.6629830598831177, "learning_rate": 9.752258925646665e-05, "loss": 0.0789, "step": 14560 }, { "grad_norm": 0.5148946046829224, "learning_rate": 9.751744648326926e-05, "loss": 0.0757, "step": 14570 }, { "grad_norm": 0.7255206108093262, "learning_rate": 9.751229851363536e-05, "loss": 0.0957, "step": 14580 }, { "grad_norm": 0.6071600914001465, "learning_rate": 9.750714534812793e-05, "loss": 0.0714, "step": 14590 }, { "grad_norm": 0.8189387917518616, "learning_rate": 9.750198698731053e-05, "loss": 0.0864, "step": 14600 }, { "grad_norm": 0.7735449075698853, "learning_rate": 9.749682343174722e-05, "loss": 0.0984, "step": 14610 }, { "grad_norm": 0.6503742337226868, "learning_rate": 9.749165468200272e-05, "loss": 0.0869, "step": 14620 }, { "grad_norm": 0.926214337348938, "learning_rate": 9.748648073864229e-05, "loss": 0.0929, "step": 14630 }, { "grad_norm": 0.8338154554367065, "learning_rate": 9.748130160223168e-05, "loss": 0.0856, "step": 14640 }, { "grad_norm": 0.9925368428230286, "learning_rate": 9.747611727333734e-05, "loss": 0.0912, "step": 14650 }, { "grad_norm": 0.7837947607040405, "learning_rate": 9.74709277525262e-05, "loss": 0.0829, "step": 14660 }, { "grad_norm": 0.8493066430091858, "learning_rate": 9.746573304036576e-05, "loss": 0.0869, "step": 14670 }, { "grad_norm": 1.0534039735794067, "learning_rate": 9.746053313742412e-05, "loss": 0.0893, "step": 14680 }, { "grad_norm": 0.6441314220428467, "learning_rate": 9.745532804426994e-05, "loss": 0.0729, "step": 14690 }, { "grad_norm": 0.7998584508895874, "learning_rate": 9.745011776147242e-05, "loss": 0.0879, "step": 14700 }, { "grad_norm": 0.6733270883560181, "learning_rate": 9.744490228960138e-05, "loss": 0.0879, "step": 14710 }, { "grad_norm": 0.917946457862854, "learning_rate": 9.743968162922713e-05, "loss": 0.088, "step": 14720 }, { "grad_norm": 0.6256473064422607, "learning_rate": 9.743445578092064e-05, "loss": 0.0872, "step": 14730 }, { "grad_norm": 0.534142255783081, "learning_rate": 9.742922474525338e-05, "loss": 0.1034, "step": 14740 }, { "grad_norm": 1.1627488136291504, "learning_rate": 9.742398852279741e-05, "loss": 0.0923, "step": 14750 }, { "grad_norm": 1.1280261278152466, "learning_rate": 9.741874711412535e-05, "loss": 0.1033, "step": 14760 }, { "grad_norm": 0.6368110179901123, "learning_rate": 9.741350051981042e-05, "loss": 0.1079, "step": 14770 }, { "grad_norm": 0.8327208757400513, "learning_rate": 9.740824874042633e-05, "loss": 0.1016, "step": 14780 }, { "grad_norm": 0.7431718111038208, "learning_rate": 9.740299177654746e-05, "loss": 0.0948, "step": 14790 }, { "grad_norm": 0.817742109298706, "learning_rate": 9.739772962874867e-05, "loss": 0.0945, "step": 14800 }, { "grad_norm": 0.684403657913208, "learning_rate": 9.739246229760541e-05, "loss": 0.0935, "step": 14810 }, { "grad_norm": 0.5217265486717224, "learning_rate": 9.738718978369376e-05, "loss": 0.091, "step": 14820 }, { "grad_norm": 0.9694015383720398, "learning_rate": 9.738191208759025e-05, "loss": 0.1024, "step": 14830 }, { "grad_norm": 0.8043459057807922, "learning_rate": 9.73766292098721e-05, "loss": 0.1142, "step": 14840 }, { "grad_norm": 0.7912941575050354, "learning_rate": 9.737134115111699e-05, "loss": 0.1168, "step": 14850 }, { "grad_norm": 0.5942757725715637, "learning_rate": 9.736604791190323e-05, "loss": 0.0974, "step": 14860 }, { "grad_norm": 0.6049342155456543, "learning_rate": 9.73607494928097e-05, "loss": 0.0943, "step": 14870 }, { "grad_norm": 0.5584937930107117, "learning_rate": 9.735544589441581e-05, "loss": 0.081, "step": 14880 }, { "grad_norm": 0.7871374487876892, "learning_rate": 9.735013711730154e-05, "loss": 0.0852, "step": 14890 }, { "grad_norm": 0.7459621429443359, "learning_rate": 9.734482316204747e-05, "loss": 0.0835, "step": 14900 }, { "grad_norm": 1.0145949125289917, "learning_rate": 9.733950402923473e-05, "loss": 0.1065, "step": 14910 }, { "grad_norm": 0.8392410278320312, "learning_rate": 9.7334179719445e-05, "loss": 0.0991, "step": 14920 }, { "grad_norm": 0.7530577778816223, "learning_rate": 9.732885023326053e-05, "loss": 0.1018, "step": 14930 }, { "grad_norm": 0.5594677925109863, "learning_rate": 9.732351557126418e-05, "loss": 0.0958, "step": 14940 }, { "grad_norm": 0.8167734742164612, "learning_rate": 9.731817573403929e-05, "loss": 0.1093, "step": 14950 }, { "grad_norm": 0.7729814648628235, "learning_rate": 9.731283072216985e-05, "loss": 0.092, "step": 14960 }, { "grad_norm": 0.9881871938705444, "learning_rate": 9.730748053624039e-05, "loss": 0.1009, "step": 14970 }, { "grad_norm": 0.9819440841674805, "learning_rate": 9.730212517683598e-05, "loss": 0.0919, "step": 14980 }, { "grad_norm": 1.2195541858673096, "learning_rate": 9.729676464454228e-05, "loss": 0.0815, "step": 14990 }, { "grad_norm": 0.7405465841293335, "learning_rate": 9.72913989399455e-05, "loss": 0.0929, "step": 15000 }, { "grad_norm": 0.9798779487609863, "learning_rate": 9.728602806363242e-05, "loss": 0.078, "step": 15010 }, { "grad_norm": 0.5950163006782532, "learning_rate": 9.728065201619043e-05, "loss": 0.0888, "step": 15020 }, { "grad_norm": 0.6275384426116943, "learning_rate": 9.727527079820742e-05, "loss": 0.0803, "step": 15030 }, { "grad_norm": 0.6893172264099121, "learning_rate": 9.726988441027186e-05, "loss": 0.083, "step": 15040 }, { "grad_norm": 0.6586472392082214, "learning_rate": 9.726449285297281e-05, "loss": 0.0742, "step": 15050 }, { "grad_norm": 0.6616076231002808, "learning_rate": 9.72590961268999e-05, "loss": 0.0902, "step": 15060 }, { "grad_norm": 0.6501769423484802, "learning_rate": 9.725369423264328e-05, "loss": 0.0785, "step": 15070 }, { "grad_norm": 0.7492568492889404, "learning_rate": 9.72482871707937e-05, "loss": 0.1005, "step": 15080 }, { "grad_norm": 0.7050536870956421, "learning_rate": 9.724287494194247e-05, "loss": 0.0794, "step": 15090 }, { "grad_norm": 0.9688053131103516, "learning_rate": 9.723745754668147e-05, "loss": 0.0982, "step": 15100 }, { "grad_norm": 0.840621292591095, "learning_rate": 9.723203498560313e-05, "loss": 0.0696, "step": 15110 }, { "grad_norm": 0.5570762753486633, "learning_rate": 9.722660725930046e-05, "loss": 0.0977, "step": 15120 }, { "grad_norm": 1.1241858005523682, "learning_rate": 9.722117436836702e-05, "loss": 0.1124, "step": 15130 }, { "grad_norm": 0.6732636094093323, "learning_rate": 9.721573631339696e-05, "loss": 0.082, "step": 15140 }, { "grad_norm": 0.6229876279830933, "learning_rate": 9.721029309498494e-05, "loss": 0.0742, "step": 15150 }, { "grad_norm": 0.942916214466095, "learning_rate": 9.720484471372627e-05, "loss": 0.127, "step": 15160 }, { "grad_norm": 0.9137147068977356, "learning_rate": 9.719939117021673e-05, "loss": 0.1142, "step": 15170 }, { "grad_norm": 0.8187100887298584, "learning_rate": 9.719393246505275e-05, "loss": 0.0902, "step": 15180 }, { "grad_norm": 0.8825961351394653, "learning_rate": 9.718846859883128e-05, "loss": 0.1241, "step": 15190 }, { "grad_norm": 0.6611846685409546, "learning_rate": 9.718299957214982e-05, "loss": 0.0873, "step": 15200 }, { "grad_norm": 0.9640697836875916, "learning_rate": 9.717752538560646e-05, "loss": 0.1068, "step": 15210 }, { "grad_norm": 0.648556649684906, "learning_rate": 9.717204603979986e-05, "loss": 0.0715, "step": 15220 }, { "grad_norm": 0.6724408864974976, "learning_rate": 9.716656153532922e-05, "loss": 0.1003, "step": 15230 }, { "grad_norm": 0.6555762887001038, "learning_rate": 9.716107187279434e-05, "loss": 0.074, "step": 15240 }, { "grad_norm": 0.580376923084259, "learning_rate": 9.715557705279555e-05, "loss": 0.0854, "step": 15250 }, { "grad_norm": 0.6910600662231445, "learning_rate": 9.715007707593372e-05, "loss": 0.0868, "step": 15260 }, { "grad_norm": 0.5151798725128174, "learning_rate": 9.714457194281036e-05, "loss": 0.0693, "step": 15270 }, { "grad_norm": 0.47859352827072144, "learning_rate": 9.713906165402751e-05, "loss": 0.0917, "step": 15280 }, { "grad_norm": 0.8657810688018799, "learning_rate": 9.713354621018774e-05, "loss": 0.0914, "step": 15290 }, { "grad_norm": 0.7584747076034546, "learning_rate": 9.712802561189422e-05, "loss": 0.0819, "step": 15300 }, { "grad_norm": 0.7612912654876709, "learning_rate": 9.712249985975069e-05, "loss": 0.0764, "step": 15310 }, { "grad_norm": 1.1356087923049927, "learning_rate": 9.71169689543614e-05, "loss": 0.0792, "step": 15320 }, { "grad_norm": 0.7186538577079773, "learning_rate": 9.711143289633123e-05, "loss": 0.0925, "step": 15330 }, { "grad_norm": 0.8753007054328918, "learning_rate": 9.710589168626561e-05, "loss": 0.097, "step": 15340 }, { "grad_norm": 0.5665281414985657, "learning_rate": 9.710034532477048e-05, "loss": 0.0836, "step": 15350 }, { "grad_norm": 0.8163571357727051, "learning_rate": 9.709479381245239e-05, "loss": 0.0926, "step": 15360 }, { "grad_norm": 0.9029859304428101, "learning_rate": 9.708923714991847e-05, "loss": 0.1017, "step": 15370 }, { "grad_norm": 0.8515869379043579, "learning_rate": 9.708367533777638e-05, "loss": 0.0932, "step": 15380 }, { "grad_norm": 0.6870287656784058, "learning_rate": 9.707810837663431e-05, "loss": 0.076, "step": 15390 }, { "grad_norm": 0.6358727216720581, "learning_rate": 9.707253626710113e-05, "loss": 0.0787, "step": 15400 }, { "grad_norm": 0.8179126977920532, "learning_rate": 9.706695900978613e-05, "loss": 0.0755, "step": 15410 }, { "grad_norm": 0.6751254200935364, "learning_rate": 9.706137660529926e-05, "loss": 0.1096, "step": 15420 }, { "grad_norm": 0.5425547957420349, "learning_rate": 9.705578905425101e-05, "loss": 0.0953, "step": 15430 }, { "grad_norm": 0.9011279940605164, "learning_rate": 9.705019635725241e-05, "loss": 0.083, "step": 15440 }, { "grad_norm": 0.5448620319366455, "learning_rate": 9.704459851491508e-05, "loss": 0.0898, "step": 15450 }, { "grad_norm": 1.0122700929641724, "learning_rate": 9.703899552785118e-05, "loss": 0.0883, "step": 15460 }, { "grad_norm": 0.6829877495765686, "learning_rate": 9.703338739667346e-05, "loss": 0.0877, "step": 15470 }, { "grad_norm": 1.131113052368164, "learning_rate": 9.70277741219952e-05, "loss": 0.0926, "step": 15480 }, { "grad_norm": 1.1395291090011597, "learning_rate": 9.702215570443027e-05, "loss": 0.0716, "step": 15490 }, { "grad_norm": 0.709406316280365, "learning_rate": 9.701653214459309e-05, "loss": 0.0717, "step": 15500 }, { "grad_norm": 0.8779828548431396, "learning_rate": 9.701090344309865e-05, "loss": 0.0899, "step": 15510 }, { "grad_norm": 0.6689621806144714, "learning_rate": 9.700526960056247e-05, "loss": 0.0905, "step": 15520 }, { "grad_norm": 0.808188796043396, "learning_rate": 9.699963061760068e-05, "loss": 0.0884, "step": 15530 }, { "grad_norm": 0.6928560733795166, "learning_rate": 9.699398649482997e-05, "loss": 0.0864, "step": 15540 }, { "grad_norm": 0.8405932188034058, "learning_rate": 9.698833723286753e-05, "loss": 0.0909, "step": 15550 }, { "grad_norm": 0.9030756950378418, "learning_rate": 9.698268283233118e-05, "loss": 0.0977, "step": 15560 }, { "grad_norm": 0.8249684572219849, "learning_rate": 9.697702329383929e-05, "loss": 0.0778, "step": 15570 }, { "grad_norm": 0.7565657496452332, "learning_rate": 9.697135861801074e-05, "loss": 0.0881, "step": 15580 }, { "grad_norm": 0.7601709961891174, "learning_rate": 9.696568880546505e-05, "loss": 0.0685, "step": 15590 }, { "grad_norm": 0.8825244307518005, "learning_rate": 9.696001385682223e-05, "loss": 0.0834, "step": 15600 }, { "grad_norm": 0.8816746473312378, "learning_rate": 9.695433377270291e-05, "loss": 0.0846, "step": 15610 }, { "grad_norm": 0.7138751745223999, "learning_rate": 9.694864855372824e-05, "loss": 0.079, "step": 15620 }, { "grad_norm": 0.7068935036659241, "learning_rate": 9.694295820051995e-05, "loss": 0.0758, "step": 15630 }, { "grad_norm": 0.7399963736534119, "learning_rate": 9.693726271370032e-05, "loss": 0.084, "step": 15640 }, { "grad_norm": 0.7366583347320557, "learning_rate": 9.693156209389221e-05, "loss": 0.0811, "step": 15650 }, { "grad_norm": 0.558510959148407, "learning_rate": 9.692585634171905e-05, "loss": 0.084, "step": 15660 }, { "grad_norm": 0.6591649055480957, "learning_rate": 9.692014545780476e-05, "loss": 0.0818, "step": 15670 }, { "grad_norm": 0.9244884848594666, "learning_rate": 9.691442944277393e-05, "loss": 0.0813, "step": 15680 }, { "grad_norm": 0.6152047514915466, "learning_rate": 9.690870829725162e-05, "loss": 0.0837, "step": 15690 }, { "grad_norm": 1.604455828666687, "learning_rate": 9.69029820218635e-05, "loss": 0.0776, "step": 15700 }, { "grad_norm": 0.8105626702308655, "learning_rate": 9.689725061723579e-05, "loss": 0.1002, "step": 15710 }, { "grad_norm": 0.809847891330719, "learning_rate": 9.689151408399527e-05, "loss": 0.1104, "step": 15720 }, { "grad_norm": 0.9968097805976868, "learning_rate": 9.688577242276924e-05, "loss": 0.0969, "step": 15730 }, { "grad_norm": 0.620740532875061, "learning_rate": 9.688002563418566e-05, "loss": 0.1187, "step": 15740 }, { "grad_norm": 0.5757347345352173, "learning_rate": 9.687427371887293e-05, "loss": 0.099, "step": 15750 }, { "grad_norm": 0.8649154305458069, "learning_rate": 9.686851667746012e-05, "loss": 0.0933, "step": 15760 }, { "grad_norm": 0.501143753528595, "learning_rate": 9.686275451057677e-05, "loss": 0.0681, "step": 15770 }, { "grad_norm": 0.7148818373680115, "learning_rate": 9.685698721885308e-05, "loss": 0.085, "step": 15780 }, { "grad_norm": 0.5109237432479858, "learning_rate": 9.68512148029197e-05, "loss": 0.0905, "step": 15790 }, { "grad_norm": 0.8388933539390564, "learning_rate": 9.684543726340791e-05, "loss": 0.0941, "step": 15800 }, { "grad_norm": 0.8046023845672607, "learning_rate": 9.683965460094952e-05, "loss": 0.1033, "step": 15810 }, { "grad_norm": 0.6212950348854065, "learning_rate": 9.683386681617694e-05, "loss": 0.092, "step": 15820 }, { "grad_norm": 0.7786453366279602, "learning_rate": 9.68280739097231e-05, "loss": 0.1067, "step": 15830 }, { "grad_norm": 0.9136140942573547, "learning_rate": 9.682227588222148e-05, "loss": 0.1292, "step": 15840 }, { "grad_norm": 0.5622973442077637, "learning_rate": 9.681647273430618e-05, "loss": 0.0824, "step": 15850 }, { "grad_norm": 0.9240359663963318, "learning_rate": 9.681066446661182e-05, "loss": 0.0864, "step": 15860 }, { "grad_norm": 0.7080281376838684, "learning_rate": 9.680485107977357e-05, "loss": 0.0748, "step": 15870 }, { "grad_norm": 0.6401684284210205, "learning_rate": 9.679903257442716e-05, "loss": 0.0956, "step": 15880 }, { "grad_norm": 0.7280483245849609, "learning_rate": 9.679320895120891e-05, "loss": 0.0746, "step": 15890 }, { "grad_norm": 0.8138735890388489, "learning_rate": 9.67873802107557e-05, "loss": 0.0867, "step": 15900 }, { "grad_norm": 0.7550488114356995, "learning_rate": 9.67815463537049e-05, "loss": 0.0757, "step": 15910 }, { "grad_norm": 0.48932209610939026, "learning_rate": 9.677570738069457e-05, "loss": 0.0831, "step": 15920 }, { "grad_norm": 0.5978405475616455, "learning_rate": 9.676986329236318e-05, "loss": 0.0789, "step": 15930 }, { "grad_norm": 0.8603209853172302, "learning_rate": 9.676401408934987e-05, "loss": 0.0837, "step": 15940 }, { "grad_norm": 0.7073023319244385, "learning_rate": 9.675815977229428e-05, "loss": 0.0789, "step": 15950 }, { "grad_norm": 0.7929479479789734, "learning_rate": 9.675230034183664e-05, "loss": 0.0918, "step": 15960 }, { "grad_norm": 1.2562741041183472, "learning_rate": 9.674643579861773e-05, "loss": 0.0655, "step": 15970 }, { "grad_norm": 0.9763041138648987, "learning_rate": 9.674056614327886e-05, "loss": 0.0867, "step": 15980 }, { "grad_norm": 0.6723517179489136, "learning_rate": 9.673469137646198e-05, "loss": 0.0856, "step": 15990 }, { "grad_norm": 0.6273314356803894, "learning_rate": 9.67288114988095e-05, "loss": 0.0778, "step": 16000 }, { "grad_norm": 0.6609240770339966, "learning_rate": 9.672292651096447e-05, "loss": 0.0649, "step": 16010 }, { "grad_norm": 0.592934250831604, "learning_rate": 9.671703641357042e-05, "loss": 0.0669, "step": 16020 }, { "grad_norm": 0.6313994526863098, "learning_rate": 9.67111412072715e-05, "loss": 0.0858, "step": 16030 }, { "grad_norm": 0.7693423628807068, "learning_rate": 9.670524089271242e-05, "loss": 0.0734, "step": 16040 }, { "grad_norm": 0.629192054271698, "learning_rate": 9.669933547053842e-05, "loss": 0.0711, "step": 16050 }, { "grad_norm": 1.0769572257995605, "learning_rate": 9.669342494139531e-05, "loss": 0.0832, "step": 16060 }, { "grad_norm": 0.8484664559364319, "learning_rate": 9.668750930592943e-05, "loss": 0.0732, "step": 16070 }, { "grad_norm": 0.6069327592849731, "learning_rate": 9.668158856478775e-05, "loss": 0.0701, "step": 16080 }, { "grad_norm": 0.6333803534507751, "learning_rate": 9.66756627186177e-05, "loss": 0.0802, "step": 16090 }, { "grad_norm": 0.5408791303634644, "learning_rate": 9.666973176806737e-05, "loss": 0.0779, "step": 16100 }, { "grad_norm": 0.576647162437439, "learning_rate": 9.666379571378534e-05, "loss": 0.0769, "step": 16110 }, { "grad_norm": 0.9368566870689392, "learning_rate": 9.665785455642076e-05, "loss": 0.0759, "step": 16120 }, { "grad_norm": 0.9756987690925598, "learning_rate": 9.665190829662337e-05, "loss": 0.0863, "step": 16130 }, { "grad_norm": 0.9379574060440063, "learning_rate": 9.664595693504342e-05, "loss": 0.076, "step": 16140 }, { "grad_norm": 0.6803876757621765, "learning_rate": 9.664000047233175e-05, "loss": 0.0975, "step": 16150 }, { "grad_norm": 0.7618082761764526, "learning_rate": 9.663403890913976e-05, "loss": 0.0892, "step": 16160 }, { "grad_norm": 0.7510375380516052, "learning_rate": 9.662807224611938e-05, "loss": 0.0991, "step": 16170 }, { "grad_norm": 0.8179499506950378, "learning_rate": 9.662210048392311e-05, "loss": 0.0785, "step": 16180 }, { "grad_norm": 0.8513506650924683, "learning_rate": 9.661612362320405e-05, "loss": 0.1192, "step": 16190 }, { "grad_norm": 0.5275939702987671, "learning_rate": 9.661014166461579e-05, "loss": 0.08, "step": 16200 }, { "grad_norm": 0.4063563644886017, "learning_rate": 9.66041546088125e-05, "loss": 0.1107, "step": 16210 }, { "grad_norm": 0.8429636359214783, "learning_rate": 9.659816245644895e-05, "loss": 0.0793, "step": 16220 }, { "grad_norm": 0.6130996942520142, "learning_rate": 9.65921652081804e-05, "loss": 0.0842, "step": 16230 }, { "grad_norm": 0.6435204148292542, "learning_rate": 9.658616286466271e-05, "loss": 0.0807, "step": 16240 }, { "grad_norm": 0.7048453092575073, "learning_rate": 9.65801554265523e-05, "loss": 0.0766, "step": 16250 }, { "grad_norm": 0.705902099609375, "learning_rate": 9.657414289450612e-05, "loss": 0.1137, "step": 16260 }, { "grad_norm": 0.6426945328712463, "learning_rate": 9.656812526918171e-05, "loss": 0.1035, "step": 16270 }, { "grad_norm": 0.8331275582313538, "learning_rate": 9.656210255123712e-05, "loss": 0.078, "step": 16280 }, { "grad_norm": 0.671990692615509, "learning_rate": 9.6556074741331e-05, "loss": 0.0669, "step": 16290 }, { "grad_norm": 0.727232813835144, "learning_rate": 9.655004184012256e-05, "loss": 0.085, "step": 16300 }, { "grad_norm": 0.7509024143218994, "learning_rate": 9.654400384827152e-05, "loss": 0.0995, "step": 16310 }, { "grad_norm": 0.7627646923065186, "learning_rate": 9.653796076643818e-05, "loss": 0.0774, "step": 16320 }, { "grad_norm": 0.5176739692687988, "learning_rate": 9.653191259528344e-05, "loss": 0.0874, "step": 16330 }, { "grad_norm": 0.7895996570587158, "learning_rate": 9.65258593354687e-05, "loss": 0.0869, "step": 16340 }, { "grad_norm": 0.6481452584266663, "learning_rate": 9.651980098765591e-05, "loss": 0.0803, "step": 16350 }, { "grad_norm": 0.5474938750267029, "learning_rate": 9.651373755250765e-05, "loss": 0.0718, "step": 16360 }, { "grad_norm": 1.0570935010910034, "learning_rate": 9.650766903068697e-05, "loss": 0.07, "step": 16370 }, { "grad_norm": 0.8439607620239258, "learning_rate": 9.650159542285753e-05, "loss": 0.099, "step": 16380 }, { "grad_norm": 1.035445213317871, "learning_rate": 9.649551672968353e-05, "loss": 0.0859, "step": 16390 }, { "grad_norm": 0.6831916570663452, "learning_rate": 9.648943295182973e-05, "loss": 0.0943, "step": 16400 }, { "grad_norm": 0.7349274754524231, "learning_rate": 9.648334408996144e-05, "loss": 0.077, "step": 16410 }, { "grad_norm": 0.5822612047195435, "learning_rate": 9.647725014474452e-05, "loss": 0.0643, "step": 16420 }, { "grad_norm": 1.0596188306808472, "learning_rate": 9.64711511168454e-05, "loss": 0.0965, "step": 16430 }, { "grad_norm": 0.7179012298583984, "learning_rate": 9.646504700693108e-05, "loss": 0.0714, "step": 16440 }, { "grad_norm": 0.7787632346153259, "learning_rate": 9.645893781566907e-05, "loss": 0.0827, "step": 16450 }, { "grad_norm": 0.7065779566764832, "learning_rate": 9.645282354372744e-05, "loss": 0.0839, "step": 16460 }, { "grad_norm": 1.403751015663147, "learning_rate": 9.644670419177491e-05, "loss": 0.0992, "step": 16470 }, { "grad_norm": 0.6424018740653992, "learning_rate": 9.644057976048062e-05, "loss": 0.0832, "step": 16480 }, { "grad_norm": 0.9100737571716309, "learning_rate": 9.643445025051435e-05, "loss": 0.0823, "step": 16490 }, { "grad_norm": 0.8741660714149475, "learning_rate": 9.642831566254641e-05, "loss": 0.0812, "step": 16500 }, { "grad_norm": 0.47820135951042175, "learning_rate": 9.642217599724769e-05, "loss": 0.0681, "step": 16510 }, { "grad_norm": 0.5606174468994141, "learning_rate": 9.64160312552896e-05, "loss": 0.0846, "step": 16520 }, { "grad_norm": 1.0119917392730713, "learning_rate": 9.64098814373441e-05, "loss": 0.0872, "step": 16530 }, { "grad_norm": 0.7340121865272522, "learning_rate": 9.640372654408374e-05, "loss": 0.0795, "step": 16540 }, { "grad_norm": 0.6426247358322144, "learning_rate": 9.639756657618162e-05, "loss": 0.0819, "step": 16550 }, { "grad_norm": 0.8447701930999756, "learning_rate": 9.639140153431138e-05, "loss": 0.1008, "step": 16560 }, { "grad_norm": 1.0207210779190063, "learning_rate": 9.638523141914721e-05, "loss": 0.0683, "step": 16570 }, { "grad_norm": 0.6900017261505127, "learning_rate": 9.637905623136388e-05, "loss": 0.0746, "step": 16580 }, { "grad_norm": 0.8714759349822998, "learning_rate": 9.637287597163669e-05, "loss": 0.0917, "step": 16590 }, { "grad_norm": 0.6679915189743042, "learning_rate": 9.63666906406415e-05, "loss": 0.097, "step": 16600 }, { "grad_norm": 0.8000161051750183, "learning_rate": 9.636050023905473e-05, "loss": 0.0759, "step": 16610 }, { "grad_norm": 0.7255744934082031, "learning_rate": 9.635430476755336e-05, "loss": 0.0858, "step": 16620 }, { "grad_norm": 0.6932341456413269, "learning_rate": 9.63481042268149e-05, "loss": 0.0719, "step": 16630 }, { "grad_norm": 0.7093508839607239, "learning_rate": 9.634189861751745e-05, "loss": 0.0627, "step": 16640 }, { "grad_norm": 0.6146634817123413, "learning_rate": 9.633568794033967e-05, "loss": 0.0727, "step": 16650 }, { "grad_norm": 0.6332418322563171, "learning_rate": 9.63294721959607e-05, "loss": 0.0761, "step": 16660 }, { "grad_norm": 0.8716710209846497, "learning_rate": 9.63232513850603e-05, "loss": 0.0729, "step": 16670 }, { "grad_norm": 0.7573118805885315, "learning_rate": 9.631702550831878e-05, "loss": 0.0727, "step": 16680 }, { "grad_norm": 0.6542959213256836, "learning_rate": 9.631079456641698e-05, "loss": 0.0881, "step": 16690 }, { "grad_norm": 0.639053463935852, "learning_rate": 9.630455856003632e-05, "loss": 0.0676, "step": 16700 }, { "grad_norm": 0.546222984790802, "learning_rate": 9.629831748985876e-05, "loss": 0.0786, "step": 16710 }, { "grad_norm": 0.9462377429008484, "learning_rate": 9.629207135656679e-05, "loss": 0.0845, "step": 16720 }, { "grad_norm": 0.6608410477638245, "learning_rate": 9.628582016084353e-05, "loss": 0.0539, "step": 16730 }, { "grad_norm": 0.6564825773239136, "learning_rate": 9.627956390337254e-05, "loss": 0.0736, "step": 16740 }, { "grad_norm": 0.6086409687995911, "learning_rate": 9.627330258483802e-05, "loss": 0.0801, "step": 16750 }, { "grad_norm": 0.6574465036392212, "learning_rate": 9.62670362059247e-05, "loss": 0.0856, "step": 16760 }, { "grad_norm": 0.6946197748184204, "learning_rate": 9.626076476731786e-05, "loss": 0.0723, "step": 16770 }, { "grad_norm": 0.6184017062187195, "learning_rate": 9.625448826970336e-05, "loss": 0.0787, "step": 16780 }, { "grad_norm": 0.6053348183631897, "learning_rate": 9.624820671376755e-05, "loss": 0.0667, "step": 16790 }, { "grad_norm": 0.5338664650917053, "learning_rate": 9.62419201001974e-05, "loss": 0.0816, "step": 16800 }, { "grad_norm": 0.6462382078170776, "learning_rate": 9.623562842968037e-05, "loss": 0.0901, "step": 16810 }, { "grad_norm": 0.7145140171051025, "learning_rate": 9.622933170290454e-05, "loss": 0.0816, "step": 16820 }, { "grad_norm": 0.6025509834289551, "learning_rate": 9.622302992055849e-05, "loss": 0.0668, "step": 16830 }, { "grad_norm": 0.5775575041770935, "learning_rate": 9.62167230833314e-05, "loss": 0.0727, "step": 16840 }, { "grad_norm": 0.6770910620689392, "learning_rate": 9.621041119191295e-05, "loss": 0.0621, "step": 16850 }, { "grad_norm": 0.7785433530807495, "learning_rate": 9.620409424699342e-05, "loss": 0.0865, "step": 16860 }, { "grad_norm": 0.6994750499725342, "learning_rate": 9.619777224926359e-05, "loss": 0.0803, "step": 16870 }, { "grad_norm": 0.617652177810669, "learning_rate": 9.619144519941485e-05, "loss": 0.0767, "step": 16880 }, { "grad_norm": 0.8206660151481628, "learning_rate": 9.618511309813912e-05, "loss": 0.0835, "step": 16890 }, { "grad_norm": 0.5437554121017456, "learning_rate": 9.617877594612886e-05, "loss": 0.066, "step": 16900 }, { "grad_norm": 0.7041560411453247, "learning_rate": 9.617243374407707e-05, "loss": 0.0798, "step": 16910 }, { "grad_norm": 0.44662004709243774, "learning_rate": 9.616608649267736e-05, "loss": 0.0834, "step": 16920 }, { "grad_norm": 0.42985594272613525, "learning_rate": 9.615973419262385e-05, "loss": 0.0724, "step": 16930 }, { "grad_norm": 0.5794316530227661, "learning_rate": 9.615337684461119e-05, "loss": 0.0894, "step": 16940 }, { "grad_norm": 0.6571738719940186, "learning_rate": 9.614701444933465e-05, "loss": 0.0749, "step": 16950 }, { "grad_norm": 0.6188997626304626, "learning_rate": 9.614064700748997e-05, "loss": 0.0783, "step": 16960 }, { "grad_norm": 0.8199189305305481, "learning_rate": 9.613427451977352e-05, "loss": 0.093, "step": 16970 }, { "grad_norm": 0.6873540282249451, "learning_rate": 9.612789698688216e-05, "loss": 0.0806, "step": 16980 }, { "grad_norm": 0.6778824329376221, "learning_rate": 9.612151440951334e-05, "loss": 0.0905, "step": 16990 }, { "grad_norm": 0.6669266223907471, "learning_rate": 9.611512678836506e-05, "loss": 0.0813, "step": 17000 }, { "grad_norm": 0.5836851000785828, "learning_rate": 9.610873412413584e-05, "loss": 0.0858, "step": 17010 }, { "grad_norm": 1.0064671039581299, "learning_rate": 9.610233641752476e-05, "loss": 0.0873, "step": 17020 }, { "grad_norm": 0.7493147253990173, "learning_rate": 9.609593366923151e-05, "loss": 0.0722, "step": 17030 }, { "grad_norm": 0.717962920665741, "learning_rate": 9.608952587995625e-05, "loss": 0.0745, "step": 17040 }, { "grad_norm": 0.9446314573287964, "learning_rate": 9.608311305039972e-05, "loss": 0.0734, "step": 17050 }, { "grad_norm": 0.6662717461585999, "learning_rate": 9.607669518126326e-05, "loss": 0.0813, "step": 17060 }, { "grad_norm": 0.5594919919967651, "learning_rate": 9.607027227324866e-05, "loss": 0.0792, "step": 17070 }, { "grad_norm": 0.7587931752204895, "learning_rate": 9.606384432705837e-05, "loss": 0.0802, "step": 17080 }, { "grad_norm": 0.7191274166107178, "learning_rate": 9.60574113433953e-05, "loss": 0.0735, "step": 17090 }, { "grad_norm": 0.8002609610557556, "learning_rate": 9.6050973322963e-05, "loss": 0.084, "step": 17100 }, { "grad_norm": 0.7377089262008667, "learning_rate": 9.604453026646547e-05, "loss": 0.0759, "step": 17110 }, { "grad_norm": 0.8409256339073181, "learning_rate": 9.603808217460735e-05, "loss": 0.0828, "step": 17120 }, { "grad_norm": 0.7313737869262695, "learning_rate": 9.603162904809377e-05, "loss": 0.1176, "step": 17130 }, { "grad_norm": 0.5294074416160583, "learning_rate": 9.602517088763045e-05, "loss": 0.0757, "step": 17140 }, { "grad_norm": 0.45190393924713135, "learning_rate": 9.601870769392365e-05, "loss": 0.0621, "step": 17150 }, { "grad_norm": 0.5801727175712585, "learning_rate": 9.601223946768017e-05, "loss": 0.086, "step": 17160 }, { "grad_norm": 1.0465035438537598, "learning_rate": 9.600576620960734e-05, "loss": 0.0991, "step": 17170 }, { "grad_norm": 0.5102530121803284, "learning_rate": 9.599928792041308e-05, "loss": 0.0731, "step": 17180 }, { "grad_norm": 0.490426629781723, "learning_rate": 9.599280460080587e-05, "loss": 0.062, "step": 17190 }, { "grad_norm": 0.6534616351127625, "learning_rate": 9.59863162514947e-05, "loss": 0.0721, "step": 17200 }, { "grad_norm": 0.48189884424209595, "learning_rate": 9.597982287318911e-05, "loss": 0.071, "step": 17210 }, { "grad_norm": 0.7744833827018738, "learning_rate": 9.597332446659923e-05, "loss": 0.0707, "step": 17220 }, { "grad_norm": 0.810249388217926, "learning_rate": 9.59668210324357e-05, "loss": 0.0828, "step": 17230 }, { "grad_norm": 0.6602053642272949, "learning_rate": 9.596031257140974e-05, "loss": 0.0867, "step": 17240 }, { "grad_norm": 0.7691343426704407, "learning_rate": 9.59537990842331e-05, "loss": 0.0783, "step": 17250 }, { "grad_norm": 0.6551119685173035, "learning_rate": 9.594728057161806e-05, "loss": 0.0831, "step": 17260 }, { "grad_norm": 0.7958602905273438, "learning_rate": 9.594075703427752e-05, "loss": 0.0814, "step": 17270 }, { "grad_norm": 0.7898664474487305, "learning_rate": 9.593422847292486e-05, "loss": 0.0939, "step": 17280 }, { "grad_norm": 0.5764762163162231, "learning_rate": 9.592769488827402e-05, "loss": 0.0654, "step": 17290 }, { "grad_norm": 0.9383295178413391, "learning_rate": 9.592115628103952e-05, "loss": 0.0933, "step": 17300 }, { "grad_norm": 0.7167173027992249, "learning_rate": 9.591461265193643e-05, "loss": 0.0725, "step": 17310 }, { "grad_norm": 0.480713814496994, "learning_rate": 9.590806400168032e-05, "loss": 0.0689, "step": 17320 }, { "grad_norm": 0.5782924890518188, "learning_rate": 9.590151033098735e-05, "loss": 0.066, "step": 17330 }, { "grad_norm": 0.6383687853813171, "learning_rate": 9.589495164057423e-05, "loss": 0.0703, "step": 17340 }, { "grad_norm": 0.8692727088928223, "learning_rate": 9.58883879311582e-05, "loss": 0.0718, "step": 17350 }, { "grad_norm": 0.7201029658317566, "learning_rate": 9.588181920345705e-05, "loss": 0.0775, "step": 17360 }, { "grad_norm": 0.6738185882568359, "learning_rate": 9.587524545818913e-05, "loss": 0.0583, "step": 17370 }, { "grad_norm": 0.5370697975158691, "learning_rate": 9.586866669607335e-05, "loss": 0.0661, "step": 17380 }, { "grad_norm": 0.7720847725868225, "learning_rate": 9.586208291782915e-05, "loss": 0.0768, "step": 17390 }, { "grad_norm": 0.6878800988197327, "learning_rate": 9.58554941241765e-05, "loss": 0.0649, "step": 17400 }, { "grad_norm": 0.5273878574371338, "learning_rate": 9.584890031583596e-05, "loss": 0.0762, "step": 17410 }, { "grad_norm": 1.0272510051727295, "learning_rate": 9.584230149352861e-05, "loss": 0.1015, "step": 17420 }, { "grad_norm": 0.4656578004360199, "learning_rate": 9.58356976579761e-05, "loss": 0.0653, "step": 17430 }, { "grad_norm": 1.3012231588363647, "learning_rate": 9.58290888099006e-05, "loss": 0.1272, "step": 17440 }, { "grad_norm": 0.839185357093811, "learning_rate": 9.582247495002486e-05, "loss": 0.0804, "step": 17450 }, { "grad_norm": 0.49543672800064087, "learning_rate": 9.581585607907214e-05, "loss": 0.075, "step": 17460 }, { "grad_norm": 0.7057957649230957, "learning_rate": 9.580923219776628e-05, "loss": 0.0837, "step": 17470 }, { "grad_norm": 0.9272152781486511, "learning_rate": 9.580260330683167e-05, "loss": 0.0755, "step": 17480 }, { "grad_norm": 0.5017886161804199, "learning_rate": 9.579596940699322e-05, "loss": 0.0786, "step": 17490 }, { "grad_norm": 0.5510516166687012, "learning_rate": 9.578933049897643e-05, "loss": 0.0901, "step": 17500 }, { "grad_norm": 0.6196696162223816, "learning_rate": 9.578268658350728e-05, "loss": 0.065, "step": 17510 }, { "grad_norm": 0.8089232444763184, "learning_rate": 9.577603766131235e-05, "loss": 0.0875, "step": 17520 }, { "grad_norm": 0.958017885684967, "learning_rate": 9.576938373311878e-05, "loss": 0.0871, "step": 17530 }, { "grad_norm": 0.8081335425376892, "learning_rate": 9.576272479965421e-05, "loss": 0.0942, "step": 17540 }, { "grad_norm": 0.9375270009040833, "learning_rate": 9.575606086164687e-05, "loss": 0.0685, "step": 17550 }, { "grad_norm": 0.6041695475578308, "learning_rate": 9.57493919198255e-05, "loss": 0.1073, "step": 17560 }, { "grad_norm": 0.5682409405708313, "learning_rate": 9.57427179749194e-05, "loss": 0.0887, "step": 17570 }, { "grad_norm": 0.80363929271698, "learning_rate": 9.573603902765846e-05, "loss": 0.0685, "step": 17580 }, { "grad_norm": 0.8411824703216553, "learning_rate": 9.572935507877304e-05, "loss": 0.0777, "step": 17590 }, { "grad_norm": 0.5136776566505432, "learning_rate": 9.57226661289941e-05, "loss": 0.0709, "step": 17600 }, { "grad_norm": 0.7133339047431946, "learning_rate": 9.571597217905315e-05, "loss": 0.0737, "step": 17610 }, { "grad_norm": 0.7121514678001404, "learning_rate": 9.57092732296822e-05, "loss": 0.0746, "step": 17620 }, { "grad_norm": 0.6712501645088196, "learning_rate": 9.570256928161385e-05, "loss": 0.0628, "step": 17630 }, { "grad_norm": 0.898370623588562, "learning_rate": 9.569586033558126e-05, "loss": 0.0794, "step": 17640 }, { "grad_norm": 0.5448424220085144, "learning_rate": 9.568914639231807e-05, "loss": 0.0817, "step": 17650 }, { "grad_norm": 0.6784566640853882, "learning_rate": 9.568242745255852e-05, "loss": 0.0739, "step": 17660 }, { "grad_norm": 0.7435540556907654, "learning_rate": 9.567570351703739e-05, "loss": 0.0812, "step": 17670 }, { "grad_norm": 0.7656102180480957, "learning_rate": 9.566897458649001e-05, "loss": 0.0788, "step": 17680 }, { "grad_norm": 0.616260290145874, "learning_rate": 9.566224066165221e-05, "loss": 0.0671, "step": 17690 }, { "grad_norm": 0.6843709349632263, "learning_rate": 9.565550174326043e-05, "loss": 0.0675, "step": 17700 }, { "grad_norm": 0.5411096215248108, "learning_rate": 9.564875783205162e-05, "loss": 0.0759, "step": 17710 }, { "grad_norm": 0.4935464859008789, "learning_rate": 9.564200892876328e-05, "loss": 0.0721, "step": 17720 }, { "grad_norm": 0.561926543712616, "learning_rate": 9.563525503413348e-05, "loss": 0.0742, "step": 17730 }, { "grad_norm": 0.8436631560325623, "learning_rate": 9.562849614890079e-05, "loss": 0.0743, "step": 17740 }, { "grad_norm": 0.5735068917274475, "learning_rate": 9.562173227380436e-05, "loss": 0.0944, "step": 17750 }, { "grad_norm": 0.5131365060806274, "learning_rate": 9.561496340958389e-05, "loss": 0.0664, "step": 17760 }, { "grad_norm": 0.5761950612068176, "learning_rate": 9.560818955697959e-05, "loss": 0.0886, "step": 17770 }, { "grad_norm": 0.4174179434776306, "learning_rate": 9.560141071673228e-05, "loss": 0.0638, "step": 17780 }, { "grad_norm": 0.6001603007316589, "learning_rate": 9.559462688958323e-05, "loss": 0.0753, "step": 17790 }, { "grad_norm": 0.7144585251808167, "learning_rate": 9.558783807627434e-05, "loss": 0.0808, "step": 17800 }, { "grad_norm": 1.3229955434799194, "learning_rate": 9.558104427754801e-05, "loss": 0.0755, "step": 17810 }, { "grad_norm": 0.5630087852478027, "learning_rate": 9.557424549414722e-05, "loss": 0.0766, "step": 17820 }, { "grad_norm": 0.948325514793396, "learning_rate": 9.556744172681546e-05, "loss": 0.0972, "step": 17830 }, { "grad_norm": 0.6481444835662842, "learning_rate": 9.556063297629677e-05, "loss": 0.0898, "step": 17840 }, { "grad_norm": 0.6728043556213379, "learning_rate": 9.555381924333578e-05, "loss": 0.0729, "step": 17850 }, { "grad_norm": 0.8117936849594116, "learning_rate": 9.554700052867758e-05, "loss": 0.0859, "step": 17860 }, { "grad_norm": 0.5755231380462646, "learning_rate": 9.554017683306789e-05, "loss": 0.0676, "step": 17870 }, { "grad_norm": 0.9883171916007996, "learning_rate": 9.553334815725294e-05, "loss": 0.0973, "step": 17880 }, { "grad_norm": 0.6385284662246704, "learning_rate": 9.552651450197949e-05, "loss": 0.0788, "step": 17890 }, { "grad_norm": 0.7022650241851807, "learning_rate": 9.551967586799486e-05, "loss": 0.0748, "step": 17900 }, { "grad_norm": 0.49759432673454285, "learning_rate": 9.551283225604692e-05, "loss": 0.0611, "step": 17910 }, { "grad_norm": 0.6642858386039734, "learning_rate": 9.550598366688406e-05, "loss": 0.0744, "step": 17920 }, { "grad_norm": 0.4258267879486084, "learning_rate": 9.549913010125526e-05, "loss": 0.0623, "step": 17930 }, { "grad_norm": 0.44421613216400146, "learning_rate": 9.549227155990999e-05, "loss": 0.0545, "step": 17940 }, { "grad_norm": 0.7693401575088501, "learning_rate": 9.548540804359828e-05, "loss": 0.0725, "step": 17950 }, { "grad_norm": 0.8947684168815613, "learning_rate": 9.547853955307077e-05, "loss": 0.0757, "step": 17960 }, { "grad_norm": 0.8902791142463684, "learning_rate": 9.547166608907853e-05, "loss": 0.1018, "step": 17970 }, { "grad_norm": 0.5266231894493103, "learning_rate": 9.546478765237326e-05, "loss": 0.0986, "step": 17980 }, { "grad_norm": 0.7645797729492188, "learning_rate": 9.545790424370715e-05, "loss": 0.0957, "step": 17990 }, { "grad_norm": 0.8759085536003113, "learning_rate": 9.5451015863833e-05, "loss": 0.0786, "step": 18000 }, { "grad_norm": 0.8057957291603088, "learning_rate": 9.544412251350408e-05, "loss": 0.0691, "step": 18010 }, { "grad_norm": 0.6951627135276794, "learning_rate": 9.543722419347422e-05, "loss": 0.071, "step": 18020 }, { "grad_norm": 0.49850842356681824, "learning_rate": 9.543032090449788e-05, "loss": 0.0745, "step": 18030 }, { "grad_norm": 0.6174628734588623, "learning_rate": 9.542341264732992e-05, "loss": 0.0907, "step": 18040 }, { "grad_norm": 0.8613621592521667, "learning_rate": 9.541649942272585e-05, "loss": 0.0947, "step": 18050 }, { "grad_norm": 0.5228797793388367, "learning_rate": 9.54095812314417e-05, "loss": 0.0779, "step": 18060 }, { "grad_norm": 0.7884936332702637, "learning_rate": 9.540265807423401e-05, "loss": 0.0723, "step": 18070 }, { "grad_norm": 0.4714964032173157, "learning_rate": 9.53957299518599e-05, "loss": 0.0743, "step": 18080 }, { "grad_norm": 0.8195388913154602, "learning_rate": 9.5388796865077e-05, "loss": 0.077, "step": 18090 }, { "grad_norm": 0.6221393942832947, "learning_rate": 9.538185881464353e-05, "loss": 0.0745, "step": 18100 }, { "grad_norm": 0.7344996333122253, "learning_rate": 9.537491580131821e-05, "loss": 0.079, "step": 18110 }, { "grad_norm": 0.7895506024360657, "learning_rate": 9.53679678258603e-05, "loss": 0.084, "step": 18120 }, { "grad_norm": 0.6282979249954224, "learning_rate": 9.536101488902966e-05, "loss": 0.0563, "step": 18130 }, { "grad_norm": 0.7120851278305054, "learning_rate": 9.535405699158663e-05, "loss": 0.0642, "step": 18140 }, { "grad_norm": 0.4402574300765991, "learning_rate": 9.53470941342921e-05, "loss": 0.0728, "step": 18150 }, { "grad_norm": 0.6178858280181885, "learning_rate": 9.534012631790756e-05, "loss": 0.0687, "step": 18160 }, { "grad_norm": 0.8735440373420715, "learning_rate": 9.533315354319494e-05, "loss": 0.0988, "step": 18170 }, { "grad_norm": 1.356397032737732, "learning_rate": 9.532617581091682e-05, "loss": 0.0959, "step": 18180 }, { "grad_norm": 0.9219657778739929, "learning_rate": 9.531919312183629e-05, "loss": 0.102, "step": 18190 }, { "grad_norm": 0.8042499423027039, "learning_rate": 9.531220547671688e-05, "loss": 0.0812, "step": 18200 }, { "grad_norm": 0.8878005743026733, "learning_rate": 9.530521287632285e-05, "loss": 0.0955, "step": 18210 }, { "grad_norm": 0.9850731492042542, "learning_rate": 9.529821532141884e-05, "loss": 0.074, "step": 18220 }, { "grad_norm": 0.5830186605453491, "learning_rate": 9.52912128127701e-05, "loss": 0.0741, "step": 18230 }, { "grad_norm": 0.8139116168022156, "learning_rate": 9.528420535114244e-05, "loss": 0.0978, "step": 18240 }, { "grad_norm": 0.6267507076263428, "learning_rate": 9.527719293730215e-05, "loss": 0.0854, "step": 18250 }, { "grad_norm": 0.732174277305603, "learning_rate": 9.527017557201611e-05, "loss": 0.0696, "step": 18260 }, { "grad_norm": 0.6291735768318176, "learning_rate": 9.526315325605176e-05, "loss": 0.0758, "step": 18270 }, { "grad_norm": 0.636790931224823, "learning_rate": 9.525612599017699e-05, "loss": 0.0624, "step": 18280 }, { "grad_norm": 0.9214754104614258, "learning_rate": 9.524909377516033e-05, "loss": 0.0961, "step": 18290 }, { "grad_norm": 0.7375367283821106, "learning_rate": 9.524205661177081e-05, "loss": 0.0787, "step": 18300 }, { "grad_norm": 0.6875525116920471, "learning_rate": 9.523501450077801e-05, "loss": 0.072, "step": 18310 }, { "grad_norm": 0.7766441702842712, "learning_rate": 9.522796744295202e-05, "loss": 0.0723, "step": 18320 }, { "grad_norm": 0.5645381212234497, "learning_rate": 9.522091543906352e-05, "loss": 0.0801, "step": 18330 }, { "grad_norm": 1.1631635427474976, "learning_rate": 9.521385848988369e-05, "loss": 0.0884, "step": 18340 }, { "grad_norm": 0.8901931047439575, "learning_rate": 9.520679659618428e-05, "loss": 0.0787, "step": 18350 }, { "grad_norm": 0.6391176581382751, "learning_rate": 9.519972975873754e-05, "loss": 0.0613, "step": 18360 }, { "grad_norm": 0.5968806147575378, "learning_rate": 9.519265797831633e-05, "loss": 0.0591, "step": 18370 }, { "grad_norm": 0.8743019104003906, "learning_rate": 9.518558125569399e-05, "loss": 0.0562, "step": 18380 }, { "grad_norm": 1.0424753427505493, "learning_rate": 9.517849959164442e-05, "loss": 0.0845, "step": 18390 }, { "grad_norm": 0.5652055144309998, "learning_rate": 9.517141298694205e-05, "loss": 0.0762, "step": 18400 }, { "grad_norm": 0.9414704442024231, "learning_rate": 9.516432144236188e-05, "loss": 0.0784, "step": 18410 }, { "grad_norm": 0.8702095746994019, "learning_rate": 9.515722495867941e-05, "loss": 0.0963, "step": 18420 }, { "grad_norm": 0.5444852113723755, "learning_rate": 9.515012353667072e-05, "loss": 0.0992, "step": 18430 }, { "grad_norm": 0.4915858507156372, "learning_rate": 9.51430171771124e-05, "loss": 0.0771, "step": 18440 }, { "grad_norm": 0.6500811576843262, "learning_rate": 9.513590588078159e-05, "loss": 0.077, "step": 18450 }, { "grad_norm": 0.9423644542694092, "learning_rate": 9.512878964845597e-05, "loss": 0.0841, "step": 18460 }, { "grad_norm": 0.7610445022583008, "learning_rate": 9.512166848091377e-05, "loss": 0.0643, "step": 18470 }, { "grad_norm": 0.5797930359840393, "learning_rate": 9.511454237893376e-05, "loss": 0.09, "step": 18480 }, { "grad_norm": 0.46338650584220886, "learning_rate": 9.51074113432952e-05, "loss": 0.0734, "step": 18490 }, { "grad_norm": 0.5636799335479736, "learning_rate": 9.510027537477797e-05, "loss": 0.0817, "step": 18500 }, { "grad_norm": 0.543668806552887, "learning_rate": 9.509313447416242e-05, "loss": 0.0845, "step": 18510 }, { "grad_norm": 0.5912137031555176, "learning_rate": 9.508598864222949e-05, "loss": 0.0659, "step": 18520 }, { "grad_norm": 0.6859591007232666, "learning_rate": 9.507883787976062e-05, "loss": 0.0772, "step": 18530 }, { "grad_norm": 0.5468846559524536, "learning_rate": 9.507168218753781e-05, "loss": 0.0824, "step": 18540 }, { "grad_norm": 0.6124027371406555, "learning_rate": 9.506452156634362e-05, "loss": 0.0611, "step": 18550 }, { "grad_norm": 0.6019365191459656, "learning_rate": 9.505735601696109e-05, "loss": 0.0783, "step": 18560 }, { "grad_norm": 0.6673015356063843, "learning_rate": 9.505018554017385e-05, "loss": 0.0778, "step": 18570 }, { "grad_norm": 1.0481928586959839, "learning_rate": 9.504301013676604e-05, "loss": 0.088, "step": 18580 }, { "grad_norm": 0.6909290552139282, "learning_rate": 9.503582980752238e-05, "loss": 0.0653, "step": 18590 }, { "grad_norm": 0.5841122269630432, "learning_rate": 9.502864455322809e-05, "loss": 0.0938, "step": 18600 }, { "grad_norm": 0.7394177317619324, "learning_rate": 9.502145437466891e-05, "loss": 0.0736, "step": 18610 }, { "grad_norm": 0.7825600504875183, "learning_rate": 9.501425927263116e-05, "loss": 0.0779, "step": 18620 }, { "grad_norm": 0.6538883447647095, "learning_rate": 9.500705924790172e-05, "loss": 0.0884, "step": 18630 }, { "grad_norm": 0.7872423529624939, "learning_rate": 9.499985430126794e-05, "loss": 0.1053, "step": 18640 }, { "grad_norm": 0.7546918392181396, "learning_rate": 9.499264443351775e-05, "loss": 0.0835, "step": 18650 }, { "grad_norm": 0.9908331036567688, "learning_rate": 9.498542964543961e-05, "loss": 0.0785, "step": 18660 }, { "grad_norm": 0.722748875617981, "learning_rate": 9.497820993782252e-05, "loss": 0.1003, "step": 18670 }, { "grad_norm": 0.6174325346946716, "learning_rate": 9.497098531145601e-05, "loss": 0.0694, "step": 18680 }, { "grad_norm": 0.7685341238975525, "learning_rate": 9.496375576713017e-05, "loss": 0.0806, "step": 18690 }, { "grad_norm": 1.0880881547927856, "learning_rate": 9.49565213056356e-05, "loss": 0.0716, "step": 18700 }, { "grad_norm": 0.7709265351295471, "learning_rate": 9.494928192776342e-05, "loss": 0.0788, "step": 18710 }, { "grad_norm": 0.5977085828781128, "learning_rate": 9.494203763430538e-05, "loss": 0.0693, "step": 18720 }, { "grad_norm": 1.0691419839859009, "learning_rate": 9.493478842605366e-05, "loss": 0.0852, "step": 18730 }, { "grad_norm": 0.5672874450683594, "learning_rate": 9.492753430380105e-05, "loss": 0.1018, "step": 18740 }, { "grad_norm": 0.593226969242096, "learning_rate": 9.492027526834083e-05, "loss": 0.1053, "step": 18750 }, { "grad_norm": 0.6002535820007324, "learning_rate": 9.491301132046684e-05, "loss": 0.0702, "step": 18760 }, { "grad_norm": 0.7822251319885254, "learning_rate": 9.490574246097345e-05, "loss": 0.0899, "step": 18770 }, { "grad_norm": 0.7520727515220642, "learning_rate": 9.48984686906556e-05, "loss": 0.0679, "step": 18780 }, { "grad_norm": 0.5937962532043457, "learning_rate": 9.489119001030871e-05, "loss": 0.0773, "step": 18790 }, { "grad_norm": 0.8174864649772644, "learning_rate": 9.488390642072878e-05, "loss": 0.0877, "step": 18800 }, { "grad_norm": 0.8394585251808167, "learning_rate": 9.48766179227123e-05, "loss": 0.0935, "step": 18810 }, { "grad_norm": 0.7467166185379028, "learning_rate": 9.486932451705636e-05, "loss": 0.0675, "step": 18820 }, { "grad_norm": 0.9755411148071289, "learning_rate": 9.486202620455857e-05, "loss": 0.0648, "step": 18830 }, { "grad_norm": 0.7941814661026001, "learning_rate": 9.485472298601704e-05, "loss": 0.073, "step": 18840 }, { "grad_norm": 0.8604807257652283, "learning_rate": 9.484741486223043e-05, "loss": 0.0812, "step": 18850 }, { "grad_norm": 0.839301347732544, "learning_rate": 9.484010183399797e-05, "loss": 0.0679, "step": 18860 }, { "grad_norm": 0.7571041584014893, "learning_rate": 9.483278390211938e-05, "loss": 0.0815, "step": 18870 }, { "grad_norm": 0.570949137210846, "learning_rate": 9.482546106739496e-05, "loss": 0.0691, "step": 18880 }, { "grad_norm": 0.6847254633903503, "learning_rate": 9.48181333306255e-05, "loss": 0.072, "step": 18890 }, { "grad_norm": 0.8774123191833496, "learning_rate": 9.481080069261237e-05, "loss": 0.0726, "step": 18900 }, { "grad_norm": 0.9025694131851196, "learning_rate": 9.480346315415745e-05, "loss": 0.0689, "step": 18910 }, { "grad_norm": 0.7022468447685242, "learning_rate": 9.479612071606314e-05, "loss": 0.0738, "step": 18920 }, { "grad_norm": 0.5303649306297302, "learning_rate": 9.478877337913244e-05, "loss": 0.0931, "step": 18930 }, { "grad_norm": 0.5266226530075073, "learning_rate": 9.478142114416881e-05, "loss": 0.1037, "step": 18940 }, { "grad_norm": 1.0466172695159912, "learning_rate": 9.47740640119763e-05, "loss": 0.0829, "step": 18950 }, { "grad_norm": 0.7453996539115906, "learning_rate": 9.476670198335947e-05, "loss": 0.0678, "step": 18960 }, { "grad_norm": 0.6400222778320312, "learning_rate": 9.47593350591234e-05, "loss": 0.0604, "step": 18970 }, { "grad_norm": 0.5034556984901428, "learning_rate": 9.475196324007376e-05, "loss": 0.0704, "step": 18980 }, { "grad_norm": 0.6021028757095337, "learning_rate": 9.474458652701669e-05, "loss": 0.0739, "step": 18990 }, { "grad_norm": 0.5888953804969788, "learning_rate": 9.473720492075892e-05, "loss": 0.0689, "step": 19000 }, { "grad_norm": 1.0074914693832397, "learning_rate": 9.472981842210768e-05, "loss": 0.0603, "step": 19010 }, { "grad_norm": 0.7823606729507446, "learning_rate": 9.472242703187074e-05, "loss": 0.083, "step": 19020 }, { "grad_norm": 0.5470935702323914, "learning_rate": 9.471503075085643e-05, "loss": 0.0653, "step": 19030 }, { "grad_norm": 0.699946403503418, "learning_rate": 9.470762957987359e-05, "loss": 0.0825, "step": 19040 }, { "grad_norm": 0.5738539099693298, "learning_rate": 9.470022351973158e-05, "loss": 0.0646, "step": 19050 }, { "grad_norm": 0.6211488246917725, "learning_rate": 9.469281257124034e-05, "loss": 0.0683, "step": 19060 }, { "grad_norm": 0.9402207732200623, "learning_rate": 9.46853967352103e-05, "loss": 0.0836, "step": 19070 }, { "grad_norm": 0.862953782081604, "learning_rate": 9.467797601245246e-05, "loss": 0.0933, "step": 19080 }, { "grad_norm": 0.6313472390174866, "learning_rate": 9.467055040377834e-05, "loss": 0.0766, "step": 19090 }, { "grad_norm": 0.7822346687316895, "learning_rate": 9.466311990999999e-05, "loss": 0.0642, "step": 19100 }, { "grad_norm": 0.9510399699211121, "learning_rate": 9.465568453193e-05, "loss": 0.0838, "step": 19110 }, { "grad_norm": 0.5207825303077698, "learning_rate": 9.464824427038148e-05, "loss": 0.0731, "step": 19120 }, { "grad_norm": 0.8336910605430603, "learning_rate": 9.46407991261681e-05, "loss": 0.0806, "step": 19130 }, { "grad_norm": 0.9271414875984192, "learning_rate": 9.463334910010404e-05, "loss": 0.0953, "step": 19140 }, { "grad_norm": 0.7302889823913574, "learning_rate": 9.462589419300403e-05, "loss": 0.0734, "step": 19150 }, { "grad_norm": 0.5707947611808777, "learning_rate": 9.461843440568333e-05, "loss": 0.0677, "step": 19160 }, { "grad_norm": 0.6661115288734436, "learning_rate": 9.461096973895773e-05, "loss": 0.0633, "step": 19170 }, { "grad_norm": 0.5208153128623962, "learning_rate": 9.460350019364355e-05, "loss": 0.0769, "step": 19180 }, { "grad_norm": 0.5378293991088867, "learning_rate": 9.459602577055764e-05, "loss": 0.061, "step": 19190 }, { "grad_norm": 0.9593559503555298, "learning_rate": 9.45885464705174e-05, "loss": 0.0805, "step": 19200 }, { "grad_norm": 0.7575443983078003, "learning_rate": 9.458106229434076e-05, "loss": 0.0645, "step": 19210 }, { "grad_norm": 0.7181740403175354, "learning_rate": 9.457357324284617e-05, "loss": 0.0797, "step": 19220 }, { "grad_norm": 0.5927838683128357, "learning_rate": 9.456607931685262e-05, "loss": 0.0705, "step": 19230 }, { "grad_norm": 0.5201430916786194, "learning_rate": 9.455858051717965e-05, "loss": 0.0803, "step": 19240 }, { "grad_norm": 0.683699905872345, "learning_rate": 9.45510768446473e-05, "loss": 0.0796, "step": 19250 }, { "grad_norm": 0.5535622835159302, "learning_rate": 9.454356830007618e-05, "loss": 0.0633, "step": 19260 }, { "grad_norm": 0.5288880467414856, "learning_rate": 9.45360548842874e-05, "loss": 0.0542, "step": 19270 }, { "grad_norm": 0.7818023562431335, "learning_rate": 9.452853659810261e-05, "loss": 0.0881, "step": 19280 }, { "grad_norm": 0.6630284786224365, "learning_rate": 9.452101344234401e-05, "loss": 0.0874, "step": 19290 }, { "grad_norm": 0.5243117213249207, "learning_rate": 9.451348541783431e-05, "loss": 0.0623, "step": 19300 }, { "grad_norm": 0.7211890816688538, "learning_rate": 9.450595252539678e-05, "loss": 0.0725, "step": 19310 }, { "grad_norm": 0.6681658625602722, "learning_rate": 9.449841476585518e-05, "loss": 0.0619, "step": 19320 }, { "grad_norm": 0.659248948097229, "learning_rate": 9.449087214003384e-05, "loss": 0.0802, "step": 19330 }, { "grad_norm": 0.6797370910644531, "learning_rate": 9.448332464875765e-05, "loss": 0.0857, "step": 19340 }, { "grad_norm": 0.6026895642280579, "learning_rate": 9.447577229285192e-05, "loss": 0.0664, "step": 19350 }, { "grad_norm": 0.7367994785308838, "learning_rate": 9.446821507314261e-05, "loss": 0.0638, "step": 19360 }, { "grad_norm": 0.7803398370742798, "learning_rate": 9.446065299045617e-05, "loss": 0.06, "step": 19370 }, { "grad_norm": 0.6828271150588989, "learning_rate": 9.445308604561955e-05, "loss": 0.0751, "step": 19380 }, { "grad_norm": 0.5588394999504089, "learning_rate": 9.444551423946028e-05, "loss": 0.0646, "step": 19390 }, { "grad_norm": 0.850143551826477, "learning_rate": 9.443793757280638e-05, "loss": 0.0768, "step": 19400 }, { "grad_norm": 0.9804726839065552, "learning_rate": 9.443035604648646e-05, "loss": 0.0892, "step": 19410 }, { "grad_norm": 0.7057348489761353, "learning_rate": 9.44227696613296e-05, "loss": 0.0606, "step": 19420 }, { "grad_norm": 0.42114925384521484, "learning_rate": 9.441517841816542e-05, "loss": 0.0692, "step": 19430 }, { "grad_norm": 0.763391375541687, "learning_rate": 9.440758231782413e-05, "loss": 0.0806, "step": 19440 }, { "grad_norm": 0.49750804901123047, "learning_rate": 9.439998136113639e-05, "loss": 0.0986, "step": 19450 }, { "grad_norm": 0.5078548789024353, "learning_rate": 9.439237554893344e-05, "loss": 0.0764, "step": 19460 }, { "grad_norm": 1.5257207155227661, "learning_rate": 9.438476488204705e-05, "loss": 0.0712, "step": 19470 }, { "grad_norm": 0.8513360023498535, "learning_rate": 9.43771493613095e-05, "loss": 0.0789, "step": 19480 }, { "grad_norm": 0.6049618721008301, "learning_rate": 9.436952898755362e-05, "loss": 0.0815, "step": 19490 }, { "grad_norm": 0.7289863228797913, "learning_rate": 9.436190376161276e-05, "loss": 0.0641, "step": 19500 }, { "grad_norm": 0.7163481712341309, "learning_rate": 9.43542736843208e-05, "loss": 0.0806, "step": 19510 }, { "grad_norm": 0.5352818369865417, "learning_rate": 9.434663875651216e-05, "loss": 0.079, "step": 19520 }, { "grad_norm": 0.7007880806922913, "learning_rate": 9.433899897902177e-05, "loss": 0.0734, "step": 19530 }, { "grad_norm": 0.8785184621810913, "learning_rate": 9.433135435268511e-05, "loss": 0.0703, "step": 19540 }, { "grad_norm": 0.926040530204773, "learning_rate": 9.432370487833819e-05, "loss": 0.0927, "step": 19550 }, { "grad_norm": 0.8527601957321167, "learning_rate": 9.431605055681756e-05, "loss": 0.0642, "step": 19560 }, { "grad_norm": 0.5517381429672241, "learning_rate": 9.430839138896026e-05, "loss": 0.0824, "step": 19570 }, { "grad_norm": 0.6539831757545471, "learning_rate": 9.43007273756039e-05, "loss": 0.0637, "step": 19580 }, { "grad_norm": 0.6277626156806946, "learning_rate": 9.429305851758658e-05, "loss": 0.0628, "step": 19590 }, { "grad_norm": 1.1845752000808716, "learning_rate": 9.428538481574699e-05, "loss": 0.0691, "step": 19600 }, { "grad_norm": 0.6209822297096252, "learning_rate": 9.42777062709243e-05, "loss": 0.0671, "step": 19610 }, { "grad_norm": 1.2683488130569458, "learning_rate": 9.427002288395821e-05, "loss": 0.0916, "step": 19620 }, { "grad_norm": 0.6785003542900085, "learning_rate": 9.426233465568898e-05, "loss": 0.1125, "step": 19630 }, { "grad_norm": 0.5753723978996277, "learning_rate": 9.42546415869574e-05, "loss": 0.0728, "step": 19640 }, { "grad_norm": 0.6768969297409058, "learning_rate": 9.424694367860473e-05, "loss": 0.0816, "step": 19650 }, { "grad_norm": 0.6097058057785034, "learning_rate": 9.423924093147284e-05, "loss": 0.0613, "step": 19660 }, { "grad_norm": 0.7414233088493347, "learning_rate": 9.423153334640407e-05, "loss": 0.0859, "step": 19670 }, { "grad_norm": 0.7869282364845276, "learning_rate": 9.42238209242413e-05, "loss": 0.082, "step": 19680 }, { "grad_norm": 0.8458155393600464, "learning_rate": 9.421610366582798e-05, "loss": 0.094, "step": 19690 }, { "grad_norm": 0.7527723908424377, "learning_rate": 9.420838157200803e-05, "loss": 0.0786, "step": 19700 }, { "grad_norm": 0.5043975710868835, "learning_rate": 9.420065464362594e-05, "loss": 0.0703, "step": 19710 }, { "grad_norm": 0.4860634505748749, "learning_rate": 9.419292288152673e-05, "loss": 0.0685, "step": 19720 }, { "grad_norm": 0.7811315655708313, "learning_rate": 9.418518628655588e-05, "loss": 0.0752, "step": 19730 }, { "grad_norm": 0.76888507604599, "learning_rate": 9.417744485955951e-05, "loss": 0.0709, "step": 19740 }, { "grad_norm": 0.889941394329071, "learning_rate": 9.41696986013842e-05, "loss": 0.0938, "step": 19750 }, { "grad_norm": 0.782135546207428, "learning_rate": 9.416194751287705e-05, "loss": 0.1052, "step": 19760 }, { "grad_norm": 1.0203098058700562, "learning_rate": 9.415419159488572e-05, "loss": 0.0727, "step": 19770 }, { "grad_norm": 0.6186936497688293, "learning_rate": 9.414643084825837e-05, "loss": 0.0735, "step": 19780 }, { "grad_norm": 0.6150769591331482, "learning_rate": 9.413866527384372e-05, "loss": 0.0885, "step": 19790 }, { "grad_norm": 0.46080076694488525, "learning_rate": 9.4130894872491e-05, "loss": 0.0752, "step": 19800 }, { "grad_norm": 0.5535436868667603, "learning_rate": 9.412311964504998e-05, "loss": 0.0735, "step": 19810 }, { "grad_norm": 0.6921910643577576, "learning_rate": 9.411533959237091e-05, "loss": 0.0817, "step": 19820 }, { "grad_norm": 0.5932953953742981, "learning_rate": 9.410755471530464e-05, "loss": 0.0837, "step": 19830 }, { "grad_norm": 0.4947398006916046, "learning_rate": 9.40997650147025e-05, "loss": 0.0913, "step": 19840 }, { "grad_norm": 0.790018618106842, "learning_rate": 9.409197049141637e-05, "loss": 0.0634, "step": 19850 }, { "grad_norm": 0.627713680267334, "learning_rate": 9.408417114629863e-05, "loss": 0.0637, "step": 19860 }, { "grad_norm": 0.7035484313964844, "learning_rate": 9.40763669802022e-05, "loss": 0.0715, "step": 19870 }, { "grad_norm": 0.5631149411201477, "learning_rate": 9.406855799398056e-05, "loss": 0.0784, "step": 19880 }, { "grad_norm": 0.6199910640716553, "learning_rate": 9.406074418848767e-05, "loss": 0.0665, "step": 19890 }, { "grad_norm": 0.5596885085105896, "learning_rate": 9.405292556457805e-05, "loss": 0.0794, "step": 19900 }, { "grad_norm": 0.6267533302307129, "learning_rate": 9.404510212310671e-05, "loss": 0.0734, "step": 19910 }, { "grad_norm": 0.5696802735328674, "learning_rate": 9.403727386492924e-05, "loss": 0.0621, "step": 19920 }, { "grad_norm": 0.742728054523468, "learning_rate": 9.40294407909017e-05, "loss": 0.0712, "step": 19930 }, { "grad_norm": 0.7339962124824524, "learning_rate": 9.40216029018807e-05, "loss": 0.0804, "step": 19940 }, { "grad_norm": 0.8270772099494934, "learning_rate": 9.401376019872338e-05, "loss": 0.0769, "step": 19950 }, { "grad_norm": 0.7441871762275696, "learning_rate": 9.400591268228746e-05, "loss": 0.0668, "step": 19960 }, { "grad_norm": 0.5431694388389587, "learning_rate": 9.399806035343106e-05, "loss": 0.0763, "step": 19970 }, { "grad_norm": 0.5371382832527161, "learning_rate": 9.399020321301294e-05, "loss": 0.0574, "step": 19980 }, { "grad_norm": 0.5581750869750977, "learning_rate": 9.398234126189234e-05, "loss": 0.0728, "step": 19990 }, { "grad_norm": 0.9231567978858948, "learning_rate": 9.397447450092902e-05, "loss": 0.0841, "step": 20000 }, { "grad_norm": 0.5562914609909058, "learning_rate": 9.39666029309833e-05, "loss": 0.085, "step": 20010 }, { "grad_norm": 0.7332001328468323, "learning_rate": 9.395872655291596e-05, "loss": 0.0628, "step": 20020 }, { "grad_norm": 0.7716360688209534, "learning_rate": 9.395084536758838e-05, "loss": 0.0631, "step": 20030 }, { "grad_norm": 0.5220313668251038, "learning_rate": 9.394295937586243e-05, "loss": 0.0803, "step": 20040 }, { "grad_norm": 0.6387038826942444, "learning_rate": 9.393506857860052e-05, "loss": 0.0702, "step": 20050 }, { "grad_norm": 0.6401796340942383, "learning_rate": 9.392717297666555e-05, "loss": 0.0668, "step": 20060 }, { "grad_norm": 0.5996093153953552, "learning_rate": 9.391927257092101e-05, "loss": 0.0754, "step": 20070 }, { "grad_norm": 0.5215421915054321, "learning_rate": 9.391136736223085e-05, "loss": 0.0618, "step": 20080 }, { "grad_norm": 0.7747317552566528, "learning_rate": 9.390345735145956e-05, "loss": 0.0765, "step": 20090 }, { "grad_norm": 0.6294176578521729, "learning_rate": 9.389554253947219e-05, "loss": 0.0639, "step": 20100 }, { "grad_norm": 0.5341600775718689, "learning_rate": 9.388762292713428e-05, "loss": 0.0787, "step": 20110 }, { "grad_norm": 0.565431535243988, "learning_rate": 9.38796985153119e-05, "loss": 0.0635, "step": 20120 }, { "grad_norm": 0.7446627020835876, "learning_rate": 9.387176930487169e-05, "loss": 0.0593, "step": 20130 }, { "grad_norm": 0.5114668011665344, "learning_rate": 9.386383529668072e-05, "loss": 0.0551, "step": 20140 }, { "grad_norm": 0.8835233449935913, "learning_rate": 9.385589649160669e-05, "loss": 0.079, "step": 20150 }, { "grad_norm": 0.48272234201431274, "learning_rate": 9.384795289051775e-05, "loss": 0.0733, "step": 20160 }, { "grad_norm": 0.4557362198829651, "learning_rate": 9.384000449428261e-05, "loss": 0.0703, "step": 20170 }, { "grad_norm": 0.8447526097297668, "learning_rate": 9.383205130377048e-05, "loss": 0.082, "step": 20180 }, { "grad_norm": 0.8228936195373535, "learning_rate": 9.382409331985114e-05, "loss": 0.0678, "step": 20190 }, { "grad_norm": 0.7276959419250488, "learning_rate": 9.381613054339482e-05, "loss": 0.0895, "step": 20200 }, { "grad_norm": 0.5288962721824646, "learning_rate": 9.380816297527235e-05, "loss": 0.0593, "step": 20210 }, { "grad_norm": 0.5513678193092346, "learning_rate": 9.380019061635506e-05, "loss": 0.0661, "step": 20220 }, { "grad_norm": 0.8348037600517273, "learning_rate": 9.379221346751474e-05, "loss": 0.0621, "step": 20230 }, { "grad_norm": 0.6552679538726807, "learning_rate": 9.378423152962382e-05, "loss": 0.0748, "step": 20240 }, { "grad_norm": 0.7090774178504944, "learning_rate": 9.377624480355517e-05, "loss": 0.0557, "step": 20250 }, { "grad_norm": 0.7821017503738403, "learning_rate": 9.376825329018219e-05, "loss": 0.0826, "step": 20260 }, { "grad_norm": 0.44957002997398376, "learning_rate": 9.376025699037884e-05, "loss": 0.0618, "step": 20270 }, { "grad_norm": 0.8130098581314087, "learning_rate": 9.37522559050196e-05, "loss": 0.0618, "step": 20280 }, { "grad_norm": 0.46586495637893677, "learning_rate": 9.37442500349794e-05, "loss": 0.0843, "step": 20290 }, { "grad_norm": 0.5500568747520447, "learning_rate": 9.373623938113381e-05, "loss": 0.0819, "step": 20300 }, { "grad_norm": 0.6119130253791809, "learning_rate": 9.372822394435883e-05, "loss": 0.0741, "step": 20310 }, { "grad_norm": 0.5553300380706787, "learning_rate": 9.372020372553102e-05, "loss": 0.0661, "step": 20320 }, { "grad_norm": 0.5099859237670898, "learning_rate": 9.371217872552746e-05, "loss": 0.0625, "step": 20330 }, { "grad_norm": 0.5270673036575317, "learning_rate": 9.370414894522576e-05, "loss": 0.0771, "step": 20340 }, { "grad_norm": 0.8688459396362305, "learning_rate": 9.369611438550406e-05, "loss": 0.0701, "step": 20350 }, { "grad_norm": 0.8680895566940308, "learning_rate": 9.368807504724095e-05, "loss": 0.0828, "step": 20360 }, { "grad_norm": 1.0587208271026611, "learning_rate": 9.368003093131565e-05, "loss": 0.0894, "step": 20370 }, { "grad_norm": 0.619983971118927, "learning_rate": 9.367198203860785e-05, "loss": 0.1019, "step": 20380 }, { "grad_norm": 0.7932840585708618, "learning_rate": 9.366392836999774e-05, "loss": 0.0752, "step": 20390 }, { "grad_norm": 0.7435505986213684, "learning_rate": 9.365586992636607e-05, "loss": 0.0691, "step": 20400 }, { "grad_norm": 0.6426588296890259, "learning_rate": 9.364780670859412e-05, "loss": 0.0701, "step": 20410 }, { "grad_norm": 0.6737703084945679, "learning_rate": 9.363973871756364e-05, "loss": 0.0693, "step": 20420 }, { "grad_norm": 0.814439594745636, "learning_rate": 9.363166595415696e-05, "loss": 0.0578, "step": 20430 }, { "grad_norm": 0.5010234713554382, "learning_rate": 9.362358841925686e-05, "loss": 0.0649, "step": 20440 }, { "grad_norm": 0.8409552574157715, "learning_rate": 9.361550611374674e-05, "loss": 0.0641, "step": 20450 }, { "grad_norm": 1.0304675102233887, "learning_rate": 9.360741903851043e-05, "loss": 0.0813, "step": 20460 }, { "grad_norm": 0.7445641160011292, "learning_rate": 9.359932719443236e-05, "loss": 0.0697, "step": 20470 }, { "grad_norm": 0.9519016742706299, "learning_rate": 9.35912305823974e-05, "loss": 0.0641, "step": 20480 }, { "grad_norm": 0.5508914589881897, "learning_rate": 9.358312920329101e-05, "loss": 0.0695, "step": 20490 }, { "grad_norm": 0.5447830557823181, "learning_rate": 9.357502305799914e-05, "loss": 0.0585, "step": 20500 }, { "grad_norm": 0.6497121453285217, "learning_rate": 9.356691214740824e-05, "loss": 0.0609, "step": 20510 }, { "grad_norm": 0.7739286422729492, "learning_rate": 9.355879647240535e-05, "loss": 0.0668, "step": 20520 }, { "grad_norm": 0.5580950379371643, "learning_rate": 9.355067603387798e-05, "loss": 0.0586, "step": 20530 }, { "grad_norm": 0.47025159001350403, "learning_rate": 9.354255083271412e-05, "loss": 0.0661, "step": 20540 }, { "grad_norm": 0.5736057758331299, "learning_rate": 9.353442086980239e-05, "loss": 0.0787, "step": 20550 }, { "grad_norm": 0.9452510476112366, "learning_rate": 9.352628614603185e-05, "loss": 0.0812, "step": 20560 }, { "grad_norm": 0.8153768181800842, "learning_rate": 9.351814666229209e-05, "loss": 0.0984, "step": 20570 }, { "grad_norm": 0.543253481388092, "learning_rate": 9.351000241947324e-05, "loss": 0.0596, "step": 20580 }, { "grad_norm": 0.9370946884155273, "learning_rate": 9.350185341846594e-05, "loss": 0.0756, "step": 20590 }, { "grad_norm": 0.6989810466766357, "learning_rate": 9.349369966016134e-05, "loss": 0.0787, "step": 20600 }, { "grad_norm": 0.5603529810905457, "learning_rate": 9.348554114545117e-05, "loss": 0.0558, "step": 20610 }, { "grad_norm": 0.8501703143119812, "learning_rate": 9.347737787522758e-05, "loss": 0.0661, "step": 20620 }, { "grad_norm": 0.728526771068573, "learning_rate": 9.346920985038332e-05, "loss": 0.0783, "step": 20630 }, { "grad_norm": 0.9387527704238892, "learning_rate": 9.346103707181162e-05, "loss": 0.0664, "step": 20640 }, { "grad_norm": 0.5585402250289917, "learning_rate": 9.345285954040626e-05, "loss": 0.0624, "step": 20650 }, { "grad_norm": 0.7251256108283997, "learning_rate": 9.34446772570615e-05, "loss": 0.0662, "step": 20660 }, { "grad_norm": 0.4659869372844696, "learning_rate": 9.343649022267214e-05, "loss": 0.0475, "step": 20670 }, { "grad_norm": 0.48778682947158813, "learning_rate": 9.342829843813353e-05, "loss": 0.0581, "step": 20680 }, { "grad_norm": 0.6449577808380127, "learning_rate": 9.342010190434149e-05, "loss": 0.069, "step": 20690 }, { "grad_norm": 0.6661026477813721, "learning_rate": 9.34119006221924e-05, "loss": 0.0756, "step": 20700 }, { "grad_norm": 0.905950129032135, "learning_rate": 9.340369459258313e-05, "loss": 0.0681, "step": 20710 }, { "grad_norm": 0.6859092712402344, "learning_rate": 9.339548381641106e-05, "loss": 0.0722, "step": 20720 }, { "grad_norm": 0.6079756617546082, "learning_rate": 9.338726829457413e-05, "loss": 0.0604, "step": 20730 }, { "grad_norm": 0.4142446219921112, "learning_rate": 9.337904802797078e-05, "loss": 0.0539, "step": 20740 }, { "grad_norm": 0.8136045932769775, "learning_rate": 9.337082301749993e-05, "loss": 0.0621, "step": 20750 }, { "grad_norm": 0.7398823499679565, "learning_rate": 9.336259326406109e-05, "loss": 0.0805, "step": 20760 }, { "grad_norm": 0.7659249305725098, "learning_rate": 9.335435876855427e-05, "loss": 0.0571, "step": 20770 }, { "grad_norm": 0.8553189635276794, "learning_rate": 9.334611953187994e-05, "loss": 0.0704, "step": 20780 }, { "grad_norm": 0.8482860326766968, "learning_rate": 9.333787555493914e-05, "loss": 0.0762, "step": 20790 }, { "grad_norm": 0.6025634407997131, "learning_rate": 9.332962683863345e-05, "loss": 0.0579, "step": 20800 }, { "grad_norm": 0.7261701822280884, "learning_rate": 9.332137338386489e-05, "loss": 0.0754, "step": 20810 }, { "grad_norm": 0.9521201252937317, "learning_rate": 9.33131151915361e-05, "loss": 0.0665, "step": 20820 }, { "grad_norm": 0.6992783546447754, "learning_rate": 9.330485226255012e-05, "loss": 0.0782, "step": 20830 }, { "grad_norm": 0.6992504596710205, "learning_rate": 9.329658459781061e-05, "loss": 0.0753, "step": 20840 }, { "grad_norm": 0.6707754135131836, "learning_rate": 9.328831219822172e-05, "loss": 0.0669, "step": 20850 }, { "grad_norm": 0.5749484896659851, "learning_rate": 9.328003506468808e-05, "loss": 0.0722, "step": 20860 }, { "grad_norm": 1.0431764125823975, "learning_rate": 9.327175319811488e-05, "loss": 0.0595, "step": 20870 }, { "grad_norm": 0.5806134343147278, "learning_rate": 9.326346659940781e-05, "loss": 0.0703, "step": 20880 }, { "grad_norm": 0.8538147807121277, "learning_rate": 9.325517526947308e-05, "loss": 0.0754, "step": 20890 }, { "grad_norm": 0.9907940030097961, "learning_rate": 9.32468792092174e-05, "loss": 0.0804, "step": 20900 }, { "grad_norm": 0.7129705548286438, "learning_rate": 9.323857841954803e-05, "loss": 0.0791, "step": 20910 }, { "grad_norm": 0.8089710474014282, "learning_rate": 9.323027290137276e-05, "loss": 0.0622, "step": 20920 }, { "grad_norm": 0.5784865021705627, "learning_rate": 9.322196265559981e-05, "loss": 0.0773, "step": 20930 }, { "grad_norm": 0.8927255272865295, "learning_rate": 9.321364768313803e-05, "loss": 0.0688, "step": 20940 }, { "grad_norm": 0.4302178919315338, "learning_rate": 9.32053279848967e-05, "loss": 0.064, "step": 20950 }, { "grad_norm": 0.6125414967536926, "learning_rate": 9.319700356178567e-05, "loss": 0.067, "step": 20960 }, { "grad_norm": 0.8259851336479187, "learning_rate": 9.318867441471527e-05, "loss": 0.0627, "step": 20970 }, { "grad_norm": 0.7265821695327759, "learning_rate": 9.318034054459637e-05, "loss": 0.0752, "step": 20980 }, { "grad_norm": 0.737758457660675, "learning_rate": 9.317200195234034e-05, "loss": 0.0701, "step": 20990 }, { "grad_norm": 0.794211208820343, "learning_rate": 9.316365863885909e-05, "loss": 0.0625, "step": 21000 }, { "grad_norm": 0.9251758456230164, "learning_rate": 9.315531060506502e-05, "loss": 0.0736, "step": 21010 }, { "grad_norm": 0.6550325155258179, "learning_rate": 9.314695785187108e-05, "loss": 0.078, "step": 21020 }, { "grad_norm": 0.774599015712738, "learning_rate": 9.313860038019069e-05, "loss": 0.0677, "step": 21030 }, { "grad_norm": 0.7187937498092651, "learning_rate": 9.313023819093782e-05, "loss": 0.0648, "step": 21040 }, { "grad_norm": 0.806169331073761, "learning_rate": 9.312187128502695e-05, "loss": 0.0839, "step": 21050 }, { "grad_norm": 0.7177576422691345, "learning_rate": 9.311349966337307e-05, "loss": 0.0656, "step": 21060 }, { "grad_norm": 0.5693023204803467, "learning_rate": 9.310512332689169e-05, "loss": 0.055, "step": 21070 }, { "grad_norm": 0.6634539365768433, "learning_rate": 9.309674227649883e-05, "loss": 0.0655, "step": 21080 }, { "grad_norm": 0.6618765592575073, "learning_rate": 9.308835651311103e-05, "loss": 0.0702, "step": 21090 }, { "grad_norm": 1.0420764684677124, "learning_rate": 9.307996603764533e-05, "loss": 0.0664, "step": 21100 }, { "grad_norm": 0.5542659759521484, "learning_rate": 9.307157085101932e-05, "loss": 0.0615, "step": 21110 }, { "grad_norm": 0.5507504940032959, "learning_rate": 9.306317095415109e-05, "loss": 0.0612, "step": 21120 }, { "grad_norm": 0.7653120756149292, "learning_rate": 9.305476634795922e-05, "loss": 0.0635, "step": 21130 }, { "grad_norm": 0.6185083389282227, "learning_rate": 9.304635703336284e-05, "loss": 0.0535, "step": 21140 }, { "grad_norm": 0.8118137121200562, "learning_rate": 9.303794301128157e-05, "loss": 0.073, "step": 21150 }, { "grad_norm": 0.3515920042991638, "learning_rate": 9.302952428263555e-05, "loss": 0.0638, "step": 21160 }, { "grad_norm": 0.5334591865539551, "learning_rate": 9.302110084834545e-05, "loss": 0.0753, "step": 21170 }, { "grad_norm": 0.6279579401016235, "learning_rate": 9.301267270933245e-05, "loss": 0.0688, "step": 21180 }, { "grad_norm": 0.7414092421531677, "learning_rate": 9.300423986651823e-05, "loss": 0.0702, "step": 21190 }, { "grad_norm": 0.5753052830696106, "learning_rate": 9.299580232082501e-05, "loss": 0.0807, "step": 21200 }, { "grad_norm": 0.5923434495925903, "learning_rate": 9.298736007317547e-05, "loss": 0.1101, "step": 21210 }, { "grad_norm": 0.6634958386421204, "learning_rate": 9.297891312449288e-05, "loss": 0.0767, "step": 21220 }, { "grad_norm": 0.9870584607124329, "learning_rate": 9.297046147570094e-05, "loss": 0.0841, "step": 21230 }, { "grad_norm": 0.7738531827926636, "learning_rate": 9.296200512772396e-05, "loss": 0.0888, "step": 21240 }, { "grad_norm": 0.6282006502151489, "learning_rate": 9.295354408148668e-05, "loss": 0.0585, "step": 21250 }, { "grad_norm": 0.5353670716285706, "learning_rate": 9.294507833791441e-05, "loss": 0.0656, "step": 21260 }, { "grad_norm": 0.5040200352668762, "learning_rate": 9.293660789793295e-05, "loss": 0.0688, "step": 21270 }, { "grad_norm": 0.47628527879714966, "learning_rate": 9.292813276246858e-05, "loss": 0.0692, "step": 21280 }, { "grad_norm": 0.5924477577209473, "learning_rate": 9.291965293244816e-05, "loss": 0.085, "step": 21290 }, { "grad_norm": 0.4636029601097107, "learning_rate": 9.291116840879904e-05, "loss": 0.0575, "step": 21300 }, { "grad_norm": 0.9792137145996094, "learning_rate": 9.290267919244904e-05, "loss": 0.0931, "step": 21310 }, { "grad_norm": 0.6778801679611206, "learning_rate": 9.289418528432655e-05, "loss": 0.0672, "step": 21320 }, { "grad_norm": 0.766366183757782, "learning_rate": 9.288568668536045e-05, "loss": 0.0728, "step": 21330 }, { "grad_norm": 0.6964833736419678, "learning_rate": 9.287718339648013e-05, "loss": 0.0678, "step": 21340 }, { "grad_norm": 0.45620086789131165, "learning_rate": 9.28686754186155e-05, "loss": 0.0639, "step": 21350 }, { "grad_norm": 0.8325190544128418, "learning_rate": 9.286016275269698e-05, "loss": 0.0741, "step": 21360 }, { "grad_norm": 0.4979463219642639, "learning_rate": 9.285164539965551e-05, "loss": 0.0649, "step": 21370 }, { "grad_norm": 0.5680466294288635, "learning_rate": 9.284312336042251e-05, "loss": 0.0502, "step": 21380 }, { "grad_norm": 1.0209112167358398, "learning_rate": 9.283459663592996e-05, "loss": 0.0721, "step": 21390 }, { "grad_norm": 0.589161217212677, "learning_rate": 9.282606522711033e-05, "loss": 0.1313, "step": 21400 }, { "grad_norm": 0.6091513633728027, "learning_rate": 9.281752913489657e-05, "loss": 0.0882, "step": 21410 }, { "grad_norm": 1.117629885673523, "learning_rate": 9.280898836022222e-05, "loss": 0.0777, "step": 21420 }, { "grad_norm": 0.8877532482147217, "learning_rate": 9.280044290402126e-05, "loss": 0.0741, "step": 21430 }, { "grad_norm": 0.7744385600090027, "learning_rate": 9.279189276722821e-05, "loss": 0.0749, "step": 21440 }, { "grad_norm": 0.7479994893074036, "learning_rate": 9.278333795077812e-05, "loss": 0.0699, "step": 21450 }, { "grad_norm": 0.6615599393844604, "learning_rate": 9.27747784556065e-05, "loss": 0.0511, "step": 21460 }, { "grad_norm": 0.7211073637008667, "learning_rate": 9.276621428264942e-05, "loss": 0.0659, "step": 21470 }, { "grad_norm": 0.7603253126144409, "learning_rate": 9.275764543284345e-05, "loss": 0.0658, "step": 21480 }, { "grad_norm": 0.7489781379699707, "learning_rate": 9.274907190712566e-05, "loss": 0.0712, "step": 21490 }, { "grad_norm": 0.716621458530426, "learning_rate": 9.274049370643363e-05, "loss": 0.0721, "step": 21500 }, { "grad_norm": 0.5891606211662292, "learning_rate": 9.273191083170547e-05, "loss": 0.0821, "step": 21510 }, { "grad_norm": 0.7085055708885193, "learning_rate": 9.27233232838798e-05, "loss": 0.0716, "step": 21520 }, { "grad_norm": 0.5281147956848145, "learning_rate": 9.27147310638957e-05, "loss": 0.0577, "step": 21530 }, { "grad_norm": 0.7147297859191895, "learning_rate": 9.270613417269286e-05, "loss": 0.0818, "step": 21540 }, { "grad_norm": 0.8110277056694031, "learning_rate": 9.269753261121138e-05, "loss": 0.0668, "step": 21550 }, { "grad_norm": 0.6222120523452759, "learning_rate": 9.268892638039194e-05, "loss": 0.0761, "step": 21560 }, { "grad_norm": 0.7231652736663818, "learning_rate": 9.268031548117569e-05, "loss": 0.0628, "step": 21570 }, { "grad_norm": 0.613433837890625, "learning_rate": 9.26716999145043e-05, "loss": 0.065, "step": 21580 }, { "grad_norm": 0.6665284037590027, "learning_rate": 9.266307968131998e-05, "loss": 0.0673, "step": 21590 }, { "grad_norm": 0.629257082939148, "learning_rate": 9.26544547825654e-05, "loss": 0.077, "step": 21600 }, { "grad_norm": 0.45013678073883057, "learning_rate": 9.264582521918376e-05, "loss": 0.0675, "step": 21610 }, { "grad_norm": 0.8599061965942383, "learning_rate": 9.263719099211881e-05, "loss": 0.0695, "step": 21620 }, { "grad_norm": 0.6445178389549255, "learning_rate": 9.262855210231476e-05, "loss": 0.0749, "step": 21630 }, { "grad_norm": 0.8245641589164734, "learning_rate": 9.261990855071633e-05, "loss": 0.0637, "step": 21640 }, { "grad_norm": 0.7054548859596252, "learning_rate": 9.261126033826878e-05, "loss": 0.0743, "step": 21650 }, { "grad_norm": 0.5771211385726929, "learning_rate": 9.260260746591786e-05, "loss": 0.0714, "step": 21660 }, { "grad_norm": 0.8951902389526367, "learning_rate": 9.259394993460985e-05, "loss": 0.1033, "step": 21670 }, { "grad_norm": 0.5859652161598206, "learning_rate": 9.258528774529151e-05, "loss": 0.0638, "step": 21680 }, { "grad_norm": 0.7873304486274719, "learning_rate": 9.257662089891013e-05, "loss": 0.0647, "step": 21690 }, { "grad_norm": 0.7005412578582764, "learning_rate": 9.25679493964135e-05, "loss": 0.067, "step": 21700 }, { "grad_norm": 0.5909721851348877, "learning_rate": 9.255927323874994e-05, "loss": 0.0562, "step": 21710 }, { "grad_norm": 0.641142725944519, "learning_rate": 9.255059242686822e-05, "loss": 0.0676, "step": 21720 }, { "grad_norm": 0.8786341547966003, "learning_rate": 9.254190696171769e-05, "loss": 0.0675, "step": 21730 }, { "grad_norm": 0.57558673620224, "learning_rate": 9.25332168442482e-05, "loss": 0.0723, "step": 21740 }, { "grad_norm": 0.468736469745636, "learning_rate": 9.252452207541004e-05, "loss": 0.0875, "step": 21750 }, { "grad_norm": 0.5183502435684204, "learning_rate": 9.251582265615409e-05, "loss": 0.0597, "step": 21760 }, { "grad_norm": 0.59132981300354, "learning_rate": 9.250711858743169e-05, "loss": 0.0674, "step": 21770 }, { "grad_norm": 0.7316772937774658, "learning_rate": 9.24984098701947e-05, "loss": 0.0725, "step": 21780 }, { "grad_norm": 0.9357244372367859, "learning_rate": 9.248969650539552e-05, "loss": 0.0919, "step": 21790 }, { "grad_norm": 0.6155796647071838, "learning_rate": 9.2480978493987e-05, "loss": 0.0672, "step": 21800 }, { "grad_norm": 0.5458794236183167, "learning_rate": 9.247225583692256e-05, "loss": 0.0632, "step": 21810 }, { "grad_norm": 0.49635735154151917, "learning_rate": 9.246352853515607e-05, "loss": 0.0728, "step": 21820 }, { "grad_norm": 0.5802704095840454, "learning_rate": 9.245479658964194e-05, "loss": 0.065, "step": 21830 }, { "grad_norm": 0.6434354782104492, "learning_rate": 9.244606000133507e-05, "loss": 0.0683, "step": 21840 }, { "grad_norm": 0.6924486756324768, "learning_rate": 9.24373187711909e-05, "loss": 0.0789, "step": 21850 }, { "grad_norm": 0.6804031133651733, "learning_rate": 9.242857290016537e-05, "loss": 0.0806, "step": 21860 }, { "grad_norm": 0.7665066719055176, "learning_rate": 9.241982238921488e-05, "loss": 0.0728, "step": 21870 }, { "grad_norm": 0.6765313744544983, "learning_rate": 9.24110672392964e-05, "loss": 0.0836, "step": 21880 }, { "grad_norm": 0.5590128302574158, "learning_rate": 9.240230745136737e-05, "loss": 0.0616, "step": 21890 }, { "grad_norm": 0.9623154401779175, "learning_rate": 9.239354302638575e-05, "loss": 0.0772, "step": 21900 }, { "grad_norm": 1.0012823343276978, "learning_rate": 9.238477396531e-05, "loss": 0.0606, "step": 21910 }, { "grad_norm": 0.48953375220298767, "learning_rate": 9.23760002690991e-05, "loss": 0.0736, "step": 21920 }, { "grad_norm": 0.7689464092254639, "learning_rate": 9.236722193871252e-05, "loss": 0.0617, "step": 21930 }, { "grad_norm": 0.4845414459705353, "learning_rate": 9.235843897511023e-05, "loss": 0.0681, "step": 21940 }, { "grad_norm": 0.7603110074996948, "learning_rate": 9.234965137925276e-05, "loss": 0.0843, "step": 21950 }, { "grad_norm": 0.44880834221839905, "learning_rate": 9.234085915210108e-05, "loss": 0.0507, "step": 21960 }, { "grad_norm": 0.648504912853241, "learning_rate": 9.23320622946167e-05, "loss": 0.0684, "step": 21970 }, { "grad_norm": 0.6772283315658569, "learning_rate": 9.232326080776163e-05, "loss": 0.0663, "step": 21980 }, { "grad_norm": 0.6242740750312805, "learning_rate": 9.23144546924984e-05, "loss": 0.0639, "step": 21990 }, { "grad_norm": 0.6844095587730408, "learning_rate": 9.230564394979e-05, "loss": 0.0706, "step": 22000 }, { "grad_norm": 0.5957431197166443, "learning_rate": 9.22968285806e-05, "loss": 0.0766, "step": 22010 }, { "grad_norm": 0.6512158513069153, "learning_rate": 9.228800858589242e-05, "loss": 0.066, "step": 22020 }, { "grad_norm": 0.9082592725753784, "learning_rate": 9.227918396663179e-05, "loss": 0.074, "step": 22030 }, { "grad_norm": 0.9891961216926575, "learning_rate": 9.227035472378319e-05, "loss": 0.0737, "step": 22040 }, { "grad_norm": 0.6272591948509216, "learning_rate": 9.226152085831213e-05, "loss": 0.0724, "step": 22050 }, { "grad_norm": 0.8527829051017761, "learning_rate": 9.22526823711847e-05, "loss": 0.0731, "step": 22060 }, { "grad_norm": 0.5582748651504517, "learning_rate": 9.224383926336745e-05, "loss": 0.0646, "step": 22070 }, { "grad_norm": 0.5797166228294373, "learning_rate": 9.223499153582744e-05, "loss": 0.0789, "step": 22080 }, { "grad_norm": 0.9896573424339294, "learning_rate": 9.222613918953226e-05, "loss": 0.0741, "step": 22090 }, { "grad_norm": 0.7993379831314087, "learning_rate": 9.221728222544999e-05, "loss": 0.0655, "step": 22100 }, { "grad_norm": 0.8315528035163879, "learning_rate": 9.22084206445492e-05, "loss": 0.0672, "step": 22110 }, { "grad_norm": 0.6362498998641968, "learning_rate": 9.2199554447799e-05, "loss": 0.0907, "step": 22120 }, { "grad_norm": 0.496742844581604, "learning_rate": 9.219068363616897e-05, "loss": 0.0631, "step": 22130 }, { "grad_norm": 0.8057672381401062, "learning_rate": 9.218180821062919e-05, "loss": 0.079, "step": 22140 }, { "grad_norm": 0.7513682246208191, "learning_rate": 9.21729281721503e-05, "loss": 0.0624, "step": 22150 }, { "grad_norm": 0.7161015272140503, "learning_rate": 9.216404352170339e-05, "loss": 0.0723, "step": 22160 }, { "grad_norm": 0.5230936408042908, "learning_rate": 9.215515426026007e-05, "loss": 0.0521, "step": 22170 }, { "grad_norm": 0.6819182634353638, "learning_rate": 9.214626038879246e-05, "loss": 0.0742, "step": 22180 }, { "grad_norm": 0.6765508651733398, "learning_rate": 9.21373619082732e-05, "loss": 0.0641, "step": 22190 }, { "grad_norm": 0.47867146134376526, "learning_rate": 9.212845881967535e-05, "loss": 0.0858, "step": 22200 }, { "grad_norm": 0.42181167006492615, "learning_rate": 9.211955112397262e-05, "loss": 0.0616, "step": 22210 }, { "grad_norm": 0.6563255786895752, "learning_rate": 9.211063882213909e-05, "loss": 0.0841, "step": 22220 }, { "grad_norm": 0.6204162836074829, "learning_rate": 9.210172191514942e-05, "loss": 0.0608, "step": 22230 }, { "grad_norm": 0.7162397503852844, "learning_rate": 9.209280040397874e-05, "loss": 0.0746, "step": 22240 }, { "grad_norm": 0.46393755078315735, "learning_rate": 9.208387428960268e-05, "loss": 0.0807, "step": 22250 }, { "grad_norm": 0.5932866334915161, "learning_rate": 9.20749435729974e-05, "loss": 0.0669, "step": 22260 }, { "grad_norm": 0.7282512784004211, "learning_rate": 9.206600825513957e-05, "loss": 0.0591, "step": 22270 }, { "grad_norm": 0.6484500169754028, "learning_rate": 9.20570683370063e-05, "loss": 0.0601, "step": 22280 }, { "grad_norm": 0.624518096446991, "learning_rate": 9.204812381957528e-05, "loss": 0.0491, "step": 22290 }, { "grad_norm": 0.5125452280044556, "learning_rate": 9.203917470382465e-05, "loss": 0.0544, "step": 22300 }, { "grad_norm": 1.0382882356643677, "learning_rate": 9.203022099073309e-05, "loss": 0.0646, "step": 22310 }, { "grad_norm": 0.891837477684021, "learning_rate": 9.202126268127976e-05, "loss": 0.0696, "step": 22320 }, { "grad_norm": 0.6292453408241272, "learning_rate": 9.20122997764443e-05, "loss": 0.0507, "step": 22330 }, { "grad_norm": 0.7474984526634216, "learning_rate": 9.200333227720692e-05, "loss": 0.067, "step": 22340 }, { "grad_norm": 0.9825013279914856, "learning_rate": 9.199436018454826e-05, "loss": 0.0733, "step": 22350 }, { "grad_norm": 0.8241647481918335, "learning_rate": 9.198538349944952e-05, "loss": 0.0691, "step": 22360 }, { "grad_norm": 1.0989387035369873, "learning_rate": 9.197640222289234e-05, "loss": 0.0757, "step": 22370 }, { "grad_norm": 0.6577708721160889, "learning_rate": 9.196741635585895e-05, "loss": 0.0645, "step": 22380 }, { "grad_norm": 0.8185371160507202, "learning_rate": 9.195842589933199e-05, "loss": 0.0852, "step": 22390 }, { "grad_norm": 0.8832830786705017, "learning_rate": 9.194943085429466e-05, "loss": 0.0759, "step": 22400 }, { "grad_norm": 0.45010077953338623, "learning_rate": 9.194043122173065e-05, "loss": 0.0615, "step": 22410 }, { "grad_norm": 0.5614982843399048, "learning_rate": 9.193142700262413e-05, "loss": 0.081, "step": 22420 }, { "grad_norm": 0.7156761884689331, "learning_rate": 9.192241819795979e-05, "loss": 0.0628, "step": 22430 }, { "grad_norm": 0.5824816226959229, "learning_rate": 9.191340480872284e-05, "loss": 0.0746, "step": 22440 }, { "grad_norm": 0.5599823594093323, "learning_rate": 9.190438683589895e-05, "loss": 0.0839, "step": 22450 }, { "grad_norm": 0.9621607065200806, "learning_rate": 9.189536428047432e-05, "loss": 0.0652, "step": 22460 }, { "grad_norm": 0.9698213338851929, "learning_rate": 9.188633714343564e-05, "loss": 0.0631, "step": 22470 }, { "grad_norm": 0.6326019763946533, "learning_rate": 9.18773054257701e-05, "loss": 0.0643, "step": 22480 }, { "grad_norm": 0.5831296443939209, "learning_rate": 9.18682691284654e-05, "loss": 0.0616, "step": 22490 }, { "grad_norm": 0.6871567964553833, "learning_rate": 9.185922825250974e-05, "loss": 0.0631, "step": 22500 }, { "grad_norm": 0.4738117456436157, "learning_rate": 9.185018279889181e-05, "loss": 0.0511, "step": 22510 }, { "grad_norm": 0.6467530727386475, "learning_rate": 9.184113276860082e-05, "loss": 0.0736, "step": 22520 }, { "grad_norm": 0.7416959404945374, "learning_rate": 9.183207816262645e-05, "loss": 0.0675, "step": 22530 }, { "grad_norm": 0.7844415903091431, "learning_rate": 9.182301898195891e-05, "loss": 0.0616, "step": 22540 }, { "grad_norm": 0.6610749363899231, "learning_rate": 9.181395522758889e-05, "loss": 0.0682, "step": 22550 }, { "grad_norm": 0.8159846067428589, "learning_rate": 9.180488690050759e-05, "loss": 0.1031, "step": 22560 }, { "grad_norm": 0.9175958633422852, "learning_rate": 9.179581400170671e-05, "loss": 0.0755, "step": 22570 }, { "grad_norm": 0.7328087687492371, "learning_rate": 9.178673653217845e-05, "loss": 0.0665, "step": 22580 }, { "grad_norm": 0.7804659605026245, "learning_rate": 9.177765449291551e-05, "loss": 0.0644, "step": 22590 }, { "grad_norm": 0.4570940434932709, "learning_rate": 9.176856788491109e-05, "loss": 0.0654, "step": 22600 }, { "grad_norm": 0.5807808637619019, "learning_rate": 9.175947670915887e-05, "loss": 0.0722, "step": 22610 }, { "grad_norm": 0.7307921051979065, "learning_rate": 9.175038096665309e-05, "loss": 0.0736, "step": 22620 }, { "grad_norm": 0.7664946913719177, "learning_rate": 9.17412806583884e-05, "loss": 0.0717, "step": 22630 }, { "grad_norm": 0.5955625176429749, "learning_rate": 9.173217578536002e-05, "loss": 0.0653, "step": 22640 }, { "grad_norm": 0.4404740035533905, "learning_rate": 9.172306634856362e-05, "loss": 0.0572, "step": 22650 }, { "grad_norm": 0.5913869738578796, "learning_rate": 9.171395234899545e-05, "loss": 0.0607, "step": 22660 }, { "grad_norm": 0.6703005433082581, "learning_rate": 9.170483378765214e-05, "loss": 0.0503, "step": 22670 }, { "grad_norm": 0.6551108956336975, "learning_rate": 9.169571066553091e-05, "loss": 0.0616, "step": 22680 }, { "grad_norm": 0.6083523035049438, "learning_rate": 9.168658298362946e-05, "loss": 0.0846, "step": 22690 }, { "grad_norm": 0.793948233127594, "learning_rate": 9.167745074294598e-05, "loss": 0.0733, "step": 22700 }, { "grad_norm": 0.7101596593856812, "learning_rate": 9.166831394447913e-05, "loss": 0.0636, "step": 22710 }, { "grad_norm": 0.6498521566390991, "learning_rate": 9.165917258922812e-05, "loss": 0.0681, "step": 22720 }, { "grad_norm": 0.6833184957504272, "learning_rate": 9.165002667819262e-05, "loss": 0.0558, "step": 22730 }, { "grad_norm": 0.5166665315628052, "learning_rate": 9.164087621237282e-05, "loss": 0.0559, "step": 22740 }, { "grad_norm": 0.627591609954834, "learning_rate": 9.163172119276942e-05, "loss": 0.0638, "step": 22750 }, { "grad_norm": 0.45467185974121094, "learning_rate": 9.162256162038358e-05, "loss": 0.0565, "step": 22760 }, { "grad_norm": 0.7288874983787537, "learning_rate": 9.161339749621698e-05, "loss": 0.0847, "step": 22770 }, { "grad_norm": 0.6007886528968811, "learning_rate": 9.160422882127177e-05, "loss": 0.0662, "step": 22780 }, { "grad_norm": 0.7494064569473267, "learning_rate": 9.159505559655069e-05, "loss": 0.0673, "step": 22790 }, { "grad_norm": 0.8470434546470642, "learning_rate": 9.158587782305684e-05, "loss": 0.0727, "step": 22800 }, { "grad_norm": 0.8307350277900696, "learning_rate": 9.157669550179391e-05, "loss": 0.0588, "step": 22810 }, { "grad_norm": 0.7006117701530457, "learning_rate": 9.156750863376609e-05, "loss": 0.0683, "step": 22820 }, { "grad_norm": 0.6861932277679443, "learning_rate": 9.155831721997801e-05, "loss": 0.0952, "step": 22830 }, { "grad_norm": 0.45426705479621887, "learning_rate": 9.154912126143484e-05, "loss": 0.0597, "step": 22840 }, { "grad_norm": 0.6007707715034485, "learning_rate": 9.153992075914224e-05, "loss": 0.0592, "step": 22850 }, { "grad_norm": 0.6377313733100891, "learning_rate": 9.153071571410635e-05, "loss": 0.0545, "step": 22860 }, { "grad_norm": 0.622797966003418, "learning_rate": 9.152150612733384e-05, "loss": 0.0657, "step": 22870 }, { "grad_norm": 0.5922026038169861, "learning_rate": 9.151229199983184e-05, "loss": 0.0588, "step": 22880 }, { "grad_norm": 0.8427172899246216, "learning_rate": 9.150307333260802e-05, "loss": 0.0587, "step": 22890 }, { "grad_norm": 0.7557138204574585, "learning_rate": 9.149385012667048e-05, "loss": 0.0763, "step": 22900 }, { "grad_norm": 0.6152352690696716, "learning_rate": 9.148462238302788e-05, "loss": 0.0951, "step": 22910 }, { "grad_norm": 0.5348989367485046, "learning_rate": 9.147539010268936e-05, "loss": 0.0707, "step": 22920 }, { "grad_norm": 0.9524750709533691, "learning_rate": 9.14661532866645e-05, "loss": 0.0575, "step": 22930 }, { "grad_norm": 0.5737044811248779, "learning_rate": 9.145691193596348e-05, "loss": 0.0618, "step": 22940 }, { "grad_norm": 1.2555816173553467, "learning_rate": 9.144766605159691e-05, "loss": 0.0601, "step": 22950 }, { "grad_norm": 0.7377685308456421, "learning_rate": 9.14384156345759e-05, "loss": 0.0668, "step": 22960 }, { "grad_norm": 0.7362896203994751, "learning_rate": 9.142916068591204e-05, "loss": 0.0721, "step": 22970 }, { "grad_norm": 0.6724982857704163, "learning_rate": 9.141990120661746e-05, "loss": 0.0681, "step": 22980 }, { "grad_norm": 0.6032174229621887, "learning_rate": 9.141063719770475e-05, "loss": 0.0491, "step": 22990 }, { "grad_norm": 0.8642526268959045, "learning_rate": 9.140136866018704e-05, "loss": 0.0682, "step": 23000 }, { "grad_norm": 0.8197963237762451, "learning_rate": 9.139209559507788e-05, "loss": 0.0538, "step": 23010 }, { "grad_norm": 1.1529065370559692, "learning_rate": 9.13828180033914e-05, "loss": 0.073, "step": 23020 }, { "grad_norm": 0.7428745627403259, "learning_rate": 9.137353588614212e-05, "loss": 0.0555, "step": 23030 }, { "grad_norm": 0.5965142846107483, "learning_rate": 9.136424924434519e-05, "loss": 0.0669, "step": 23040 }, { "grad_norm": 0.44277358055114746, "learning_rate": 9.135495807901615e-05, "loss": 0.0659, "step": 23050 }, { "grad_norm": 0.4121267795562744, "learning_rate": 9.134566239117108e-05, "loss": 0.0721, "step": 23060 }, { "grad_norm": 0.6288734078407288, "learning_rate": 9.13363621818265e-05, "loss": 0.0707, "step": 23070 }, { "grad_norm": 0.4657118618488312, "learning_rate": 9.132705745199953e-05, "loss": 0.0647, "step": 23080 }, { "grad_norm": 0.5382657051086426, "learning_rate": 9.131774820270768e-05, "loss": 0.0742, "step": 23090 }, { "grad_norm": 0.7558544278144836, "learning_rate": 9.130843443496901e-05, "loss": 0.0851, "step": 23100 }, { "grad_norm": 1.0346988439559937, "learning_rate": 9.129911614980206e-05, "loss": 0.0886, "step": 23110 }, { "grad_norm": 0.6366576552391052, "learning_rate": 9.128979334822584e-05, "loss": 0.0717, "step": 23120 }, { "grad_norm": 0.8426340222358704, "learning_rate": 9.128046603125992e-05, "loss": 0.0668, "step": 23130 }, { "grad_norm": 1.1287175416946411, "learning_rate": 9.12711341999243e-05, "loss": 0.0568, "step": 23140 }, { "grad_norm": 0.689199686050415, "learning_rate": 9.12617978552395e-05, "loss": 0.07, "step": 23150 }, { "grad_norm": 0.7660415172576904, "learning_rate": 9.12524569982265e-05, "loss": 0.0826, "step": 23160 }, { "grad_norm": 0.7360414266586304, "learning_rate": 9.124311162990684e-05, "loss": 0.0582, "step": 23170 }, { "grad_norm": 0.6248835325241089, "learning_rate": 9.12337617513025e-05, "loss": 0.0662, "step": 23180 }, { "grad_norm": 0.6521707773208618, "learning_rate": 9.122440736343596e-05, "loss": 0.0687, "step": 23190 }, { "grad_norm": 0.9284383654594421, "learning_rate": 9.12150484673302e-05, "loss": 0.0652, "step": 23200 }, { "grad_norm": 0.644334077835083, "learning_rate": 9.120568506400873e-05, "loss": 0.0939, "step": 23210 }, { "grad_norm": 0.8932468295097351, "learning_rate": 9.119631715449548e-05, "loss": 0.0721, "step": 23220 }, { "grad_norm": 0.7849669456481934, "learning_rate": 9.118694473981493e-05, "loss": 0.0811, "step": 23230 }, { "grad_norm": 0.8003984093666077, "learning_rate": 9.117756782099203e-05, "loss": 0.0718, "step": 23240 }, { "grad_norm": 0.7222186326980591, "learning_rate": 9.11681863990522e-05, "loss": 0.0709, "step": 23250 }, { "grad_norm": 0.5605329275131226, "learning_rate": 9.115880047502142e-05, "loss": 0.0988, "step": 23260 }, { "grad_norm": 0.852624773979187, "learning_rate": 9.114941004992609e-05, "loss": 0.0594, "step": 23270 }, { "grad_norm": 0.7516064643859863, "learning_rate": 9.114001512479317e-05, "loss": 0.0726, "step": 23280 }, { "grad_norm": 0.5895059704780579, "learning_rate": 9.113061570065003e-05, "loss": 0.0747, "step": 23290 }, { "grad_norm": 0.8567159175872803, "learning_rate": 9.112121177852459e-05, "loss": 0.0712, "step": 23300 }, { "grad_norm": 0.7830417156219482, "learning_rate": 9.111180335944527e-05, "loss": 0.071, "step": 23310 }, { "grad_norm": 0.5750876069068909, "learning_rate": 9.110239044444093e-05, "loss": 0.0739, "step": 23320 }, { "grad_norm": 0.865447998046875, "learning_rate": 9.109297303454099e-05, "loss": 0.0758, "step": 23330 }, { "grad_norm": 0.519143283367157, "learning_rate": 9.108355113077526e-05, "loss": 0.049, "step": 23340 }, { "grad_norm": 0.6543761491775513, "learning_rate": 9.107412473417419e-05, "loss": 0.0756, "step": 23350 }, { "grad_norm": 0.49950703978538513, "learning_rate": 9.106469384576858e-05, "loss": 0.0566, "step": 23360 }, { "grad_norm": 0.6019634008407593, "learning_rate": 9.105525846658978e-05, "loss": 0.0611, "step": 23370 }, { "grad_norm": 0.6156604290008545, "learning_rate": 9.104581859766965e-05, "loss": 0.0603, "step": 23380 }, { "grad_norm": 0.6930670142173767, "learning_rate": 9.10363742400405e-05, "loss": 0.0711, "step": 23390 }, { "grad_norm": 0.5452879667282104, "learning_rate": 9.102692539473518e-05, "loss": 0.0655, "step": 23400 }, { "grad_norm": 1.0673350095748901, "learning_rate": 9.101747206278697e-05, "loss": 0.0754, "step": 23410 }, { "grad_norm": 0.783387303352356, "learning_rate": 9.100801424522968e-05, "loss": 0.0649, "step": 23420 }, { "grad_norm": 0.611560046672821, "learning_rate": 9.099855194309762e-05, "loss": 0.0676, "step": 23430 }, { "grad_norm": 0.716949462890625, "learning_rate": 9.098908515742554e-05, "loss": 0.0641, "step": 23440 }, { "grad_norm": 0.8762778639793396, "learning_rate": 9.097961388924873e-05, "loss": 0.0628, "step": 23450 }, { "grad_norm": 0.8526438474655151, "learning_rate": 9.097013813960298e-05, "loss": 0.0719, "step": 23460 }, { "grad_norm": 0.6165217757225037, "learning_rate": 9.09606579095245e-05, "loss": 0.0625, "step": 23470 }, { "grad_norm": 0.672297477722168, "learning_rate": 9.095117320005008e-05, "loss": 0.0709, "step": 23480 }, { "grad_norm": 0.8328626155853271, "learning_rate": 9.094168401221691e-05, "loss": 0.0581, "step": 23490 }, { "grad_norm": 0.6183393597602844, "learning_rate": 9.093219034706273e-05, "loss": 0.0951, "step": 23500 }, { "grad_norm": 0.48098909854888916, "learning_rate": 9.092269220562577e-05, "loss": 0.0729, "step": 23510 }, { "grad_norm": 1.130825161933899, "learning_rate": 9.09131895889447e-05, "loss": 0.0693, "step": 23520 }, { "grad_norm": 0.5192562341690063, "learning_rate": 9.090368249805873e-05, "loss": 0.0753, "step": 23530 }, { "grad_norm": 0.6723483800888062, "learning_rate": 9.089417093400754e-05, "loss": 0.066, "step": 23540 }, { "grad_norm": 0.996096670627594, "learning_rate": 9.088465489783131e-05, "loss": 0.0721, "step": 23550 }, { "grad_norm": 0.9744309186935425, "learning_rate": 9.087513439057068e-05, "loss": 0.0701, "step": 23560 }, { "grad_norm": 0.46440112590789795, "learning_rate": 9.08656094132668e-05, "loss": 0.0571, "step": 23570 }, { "grad_norm": 0.7763239145278931, "learning_rate": 9.085607996696134e-05, "loss": 0.0671, "step": 23580 }, { "grad_norm": 0.6523996591567993, "learning_rate": 9.084654605269639e-05, "loss": 0.0829, "step": 23590 }, { "grad_norm": 0.602449357509613, "learning_rate": 9.083700767151457e-05, "loss": 0.0739, "step": 23600 }, { "grad_norm": 0.7204505205154419, "learning_rate": 9.082746482445898e-05, "loss": 0.08, "step": 23610 }, { "grad_norm": 0.5613015294075012, "learning_rate": 9.081791751257325e-05, "loss": 0.0599, "step": 23620 }, { "grad_norm": 0.46973034739494324, "learning_rate": 9.080836573690142e-05, "loss": 0.0653, "step": 23630 }, { "grad_norm": 0.6484370231628418, "learning_rate": 9.079880949848805e-05, "loss": 0.0607, "step": 23640 }, { "grad_norm": 0.4949551522731781, "learning_rate": 9.078924879837822e-05, "loss": 0.0556, "step": 23650 }, { "grad_norm": 0.7255868315696716, "learning_rate": 9.077968363761747e-05, "loss": 0.1046, "step": 23660 }, { "grad_norm": 0.8033621311187744, "learning_rate": 9.077011401725182e-05, "loss": 0.0566, "step": 23670 }, { "grad_norm": 0.6095101237297058, "learning_rate": 9.07605399383278e-05, "loss": 0.0665, "step": 23680 }, { "grad_norm": 0.6483529806137085, "learning_rate": 9.075096140189243e-05, "loss": 0.0628, "step": 23690 }, { "grad_norm": 0.5664818286895752, "learning_rate": 9.074137840899318e-05, "loss": 0.0585, "step": 23700 }, { "grad_norm": 0.927013099193573, "learning_rate": 9.073179096067804e-05, "loss": 0.0871, "step": 23710 }, { "grad_norm": 0.7228860259056091, "learning_rate": 9.072219905799549e-05, "loss": 0.0616, "step": 23720 }, { "grad_norm": 0.6001428961753845, "learning_rate": 9.071260270199447e-05, "loss": 0.0617, "step": 23730 }, { "grad_norm": 0.7669336199760437, "learning_rate": 9.070300189372441e-05, "loss": 0.0565, "step": 23740 }, { "grad_norm": 0.46181270480155945, "learning_rate": 9.069339663423528e-05, "loss": 0.0967, "step": 23750 }, { "grad_norm": 0.8242732286453247, "learning_rate": 9.068378692457747e-05, "loss": 0.0652, "step": 23760 }, { "grad_norm": 0.47924166917800903, "learning_rate": 9.067417276580189e-05, "loss": 0.0453, "step": 23770 }, { "grad_norm": 0.7401602268218994, "learning_rate": 9.066455415895993e-05, "loss": 0.0656, "step": 23780 }, { "grad_norm": 0.8537379503250122, "learning_rate": 9.065493110510346e-05, "loss": 0.0644, "step": 23790 }, { "grad_norm": 0.6246068477630615, "learning_rate": 9.064530360528484e-05, "loss": 0.057, "step": 23800 }, { "grad_norm": 0.7245602011680603, "learning_rate": 9.063567166055695e-05, "loss": 0.063, "step": 23810 }, { "grad_norm": 0.7313910126686096, "learning_rate": 9.062603527197308e-05, "loss": 0.0587, "step": 23820 }, { "grad_norm": 0.6080215573310852, "learning_rate": 9.06163944405871e-05, "loss": 0.0565, "step": 23830 }, { "grad_norm": 0.5184410810470581, "learning_rate": 9.060674916745327e-05, "loss": 0.0685, "step": 23840 }, { "grad_norm": 0.6993257999420166, "learning_rate": 9.05970994536264e-05, "loss": 0.0637, "step": 23850 }, { "grad_norm": 0.5788268446922302, "learning_rate": 9.05874453001618e-05, "loss": 0.0628, "step": 23860 }, { "grad_norm": 0.7041254639625549, "learning_rate": 9.057778670811517e-05, "loss": 0.0684, "step": 23870 }, { "grad_norm": 0.5881869196891785, "learning_rate": 9.056812367854281e-05, "loss": 0.0575, "step": 23880 }, { "grad_norm": 0.7403483390808105, "learning_rate": 9.055845621250143e-05, "loss": 0.057, "step": 23890 }, { "grad_norm": 0.6323526501655579, "learning_rate": 9.054878431104825e-05, "loss": 0.0586, "step": 23900 }, { "grad_norm": 0.7199260592460632, "learning_rate": 9.0539107975241e-05, "loss": 0.0552, "step": 23910 }, { "grad_norm": 0.6441857218742371, "learning_rate": 9.052942720613784e-05, "loss": 0.0631, "step": 23920 }, { "grad_norm": 0.6448526382446289, "learning_rate": 9.051974200479745e-05, "loss": 0.0731, "step": 23930 }, { "grad_norm": 0.8204861283302307, "learning_rate": 9.051005237227901e-05, "loss": 0.064, "step": 23940 }, { "grad_norm": 0.8888016939163208, "learning_rate": 9.050035830964215e-05, "loss": 0.0771, "step": 23950 }, { "grad_norm": 0.8192788362503052, "learning_rate": 9.049065981794698e-05, "loss": 0.0674, "step": 23960 }, { "grad_norm": 0.8093031048774719, "learning_rate": 9.048095689825414e-05, "loss": 0.0683, "step": 23970 }, { "grad_norm": 0.9111888408660889, "learning_rate": 9.047124955162472e-05, "loss": 0.0731, "step": 23980 }, { "grad_norm": 0.5215030908584595, "learning_rate": 9.046153777912028e-05, "loss": 0.0812, "step": 23990 }, { "grad_norm": 0.8889768719673157, "learning_rate": 9.045182158180292e-05, "loss": 0.0676, "step": 24000 }, { "grad_norm": 0.8553900718688965, "learning_rate": 9.044210096073516e-05, "loss": 0.0679, "step": 24010 }, { "grad_norm": 0.7041555643081665, "learning_rate": 9.043237591698004e-05, "loss": 0.0597, "step": 24020 }, { "grad_norm": 0.6221044063568115, "learning_rate": 9.04226464516011e-05, "loss": 0.0617, "step": 24030 }, { "grad_norm": 0.764584481716156, "learning_rate": 9.041291256566229e-05, "loss": 0.0738, "step": 24040 }, { "grad_norm": 0.6904083490371704, "learning_rate": 9.040317426022814e-05, "loss": 0.0554, "step": 24050 }, { "grad_norm": 0.5492669343948364, "learning_rate": 9.03934315363636e-05, "loss": 0.0538, "step": 24060 }, { "grad_norm": 0.6830490827560425, "learning_rate": 9.038368439513409e-05, "loss": 0.0667, "step": 24070 }, { "grad_norm": 0.6452053785324097, "learning_rate": 9.03739328376056e-05, "loss": 0.071, "step": 24080 }, { "grad_norm": 0.9370554685592651, "learning_rate": 9.036417686484451e-05, "loss": 0.0614, "step": 24090 }, { "grad_norm": 0.6541611552238464, "learning_rate": 9.035441647791773e-05, "loss": 0.0773, "step": 24100 }, { "grad_norm": 0.5755256414413452, "learning_rate": 9.034465167789263e-05, "loss": 0.0549, "step": 24110 }, { "grad_norm": 0.7277663946151733, "learning_rate": 9.033488246583706e-05, "loss": 0.064, "step": 24120 }, { "grad_norm": 0.44442829489707947, "learning_rate": 9.032510884281941e-05, "loss": 0.0736, "step": 24130 }, { "grad_norm": 0.6249781847000122, "learning_rate": 9.031533080990848e-05, "loss": 0.0818, "step": 24140 }, { "grad_norm": 0.8012821078300476, "learning_rate": 9.030554836817358e-05, "loss": 0.0702, "step": 24150 }, { "grad_norm": 0.63719242811203, "learning_rate": 9.029576151868451e-05, "loss": 0.0729, "step": 24160 }, { "grad_norm": 0.49372872710227966, "learning_rate": 9.028597026251155e-05, "loss": 0.0572, "step": 24170 }, { "grad_norm": 0.7112248539924622, "learning_rate": 9.027617460072547e-05, "loss": 0.068, "step": 24180 }, { "grad_norm": 0.8014315366744995, "learning_rate": 9.026637453439745e-05, "loss": 0.0668, "step": 24190 }, { "grad_norm": 0.569343090057373, "learning_rate": 9.025657006459927e-05, "loss": 0.0739, "step": 24200 }, { "grad_norm": 0.6308473944664001, "learning_rate": 9.024676119240311e-05, "loss": 0.0654, "step": 24210 }, { "grad_norm": 1.1127392053604126, "learning_rate": 9.023694791888166e-05, "loss": 0.0654, "step": 24220 }, { "grad_norm": 0.9541243314743042, "learning_rate": 9.022713024510808e-05, "loss": 0.0575, "step": 24230 }, { "grad_norm": 0.640657901763916, "learning_rate": 9.021730817215601e-05, "loss": 0.063, "step": 24240 }, { "grad_norm": 1.1528334617614746, "learning_rate": 9.02074817010996e-05, "loss": 0.0684, "step": 24250 }, { "grad_norm": 0.6347309350967407, "learning_rate": 9.019765083301342e-05, "loss": 0.0692, "step": 24260 }, { "grad_norm": 0.587225615978241, "learning_rate": 9.01878155689726e-05, "loss": 0.06, "step": 24270 }, { "grad_norm": 0.5324854850769043, "learning_rate": 9.017797591005268e-05, "loss": 0.0527, "step": 24280 }, { "grad_norm": 0.6813539862632751, "learning_rate": 9.016813185732972e-05, "loss": 0.0623, "step": 24290 }, { "grad_norm": 0.5645701885223389, "learning_rate": 9.015828341188027e-05, "loss": 0.0664, "step": 24300 }, { "grad_norm": 0.5273630023002625, "learning_rate": 9.01484305747813e-05, "loss": 0.0491, "step": 24310 }, { "grad_norm": 0.8803101181983948, "learning_rate": 9.013857334711033e-05, "loss": 0.0507, "step": 24320 }, { "grad_norm": 0.833070695400238, "learning_rate": 9.012871172994534e-05, "loss": 0.0595, "step": 24330 }, { "grad_norm": 0.4483625888824463, "learning_rate": 9.011884572436476e-05, "loss": 0.0517, "step": 24340 }, { "grad_norm": 0.686194121837616, "learning_rate": 9.010897533144754e-05, "loss": 0.0692, "step": 24350 }, { "grad_norm": 0.7267826199531555, "learning_rate": 9.009910055227306e-05, "loss": 0.0875, "step": 24360 }, { "grad_norm": 0.5122624635696411, "learning_rate": 9.008922138792124e-05, "loss": 0.059, "step": 24370 }, { "grad_norm": 0.7669130563735962, "learning_rate": 9.007933783947244e-05, "loss": 0.0616, "step": 24380 }, { "grad_norm": 0.7172131538391113, "learning_rate": 9.006944990800752e-05, "loss": 0.0781, "step": 24390 }, { "grad_norm": 0.9519791007041931, "learning_rate": 9.005955759460779e-05, "loss": 0.0593, "step": 24400 }, { "grad_norm": 0.6672049760818481, "learning_rate": 9.004966090035508e-05, "loss": 0.1075, "step": 24410 }, { "grad_norm": 0.8726165294647217, "learning_rate": 9.003975982633166e-05, "loss": 0.0662, "step": 24420 }, { "grad_norm": 0.7662742137908936, "learning_rate": 9.00298543736203e-05, "loss": 0.0597, "step": 24430 }, { "grad_norm": 0.5761677026748657, "learning_rate": 9.001994454330427e-05, "loss": 0.0689, "step": 24440 }, { "grad_norm": 0.5550775527954102, "learning_rate": 9.001003033646727e-05, "loss": 0.0543, "step": 24450 }, { "grad_norm": 0.6770970821380615, "learning_rate": 9.00001117541935e-05, "loss": 0.0554, "step": 24460 }, { "grad_norm": 0.6116522550582886, "learning_rate": 8.999018879756764e-05, "loss": 0.0557, "step": 24470 }, { "grad_norm": 0.5062376856803894, "learning_rate": 8.998026146767487e-05, "loss": 0.056, "step": 24480 }, { "grad_norm": 0.6143461465835571, "learning_rate": 8.99703297656008e-05, "loss": 0.0917, "step": 24490 }, { "grad_norm": 0.7194730043411255, "learning_rate": 8.996039369243156e-05, "loss": 0.0537, "step": 24500 }, { "grad_norm": 0.7643340229988098, "learning_rate": 8.995045324925378e-05, "loss": 0.077, "step": 24510 }, { "grad_norm": 0.576680600643158, "learning_rate": 8.994050843715448e-05, "loss": 0.0666, "step": 24520 }, { "grad_norm": 0.8276187181472778, "learning_rate": 8.993055925722121e-05, "loss": 0.0749, "step": 24530 }, { "grad_norm": 0.6014428734779358, "learning_rate": 8.992060571054202e-05, "loss": 0.0631, "step": 24540 }, { "grad_norm": 1.0489619970321655, "learning_rate": 8.991064779820542e-05, "loss": 0.0795, "step": 24550 }, { "grad_norm": 0.5231780409812927, "learning_rate": 8.990068552130036e-05, "loss": 0.0601, "step": 24560 }, { "grad_norm": 0.5873991847038269, "learning_rate": 8.989071888091634e-05, "loss": 0.0802, "step": 24570 }, { "grad_norm": 0.8726377487182617, "learning_rate": 8.988074787814329e-05, "loss": 0.0687, "step": 24580 }, { "grad_norm": 0.5472860336303711, "learning_rate": 8.987077251407158e-05, "loss": 0.0582, "step": 24590 }, { "grad_norm": 0.5678046345710754, "learning_rate": 8.986079278979216e-05, "loss": 0.0618, "step": 24600 }, { "grad_norm": 0.5765443444252014, "learning_rate": 8.985080870639635e-05, "loss": 0.0693, "step": 24610 }, { "grad_norm": 0.739202082157135, "learning_rate": 8.984082026497603e-05, "loss": 0.0559, "step": 24620 }, { "grad_norm": 0.455244243144989, "learning_rate": 8.98308274666235e-05, "loss": 0.0694, "step": 24630 }, { "grad_norm": 0.8153599500656128, "learning_rate": 8.982083031243155e-05, "loss": 0.0719, "step": 24640 }, { "grad_norm": 0.7613582611083984, "learning_rate": 8.98108288034935e-05, "loss": 0.058, "step": 24650 }, { "grad_norm": 0.615393877029419, "learning_rate": 8.980082294090305e-05, "loss": 0.0601, "step": 24660 }, { "grad_norm": 0.5579047799110413, "learning_rate": 8.979081272575443e-05, "loss": 0.0613, "step": 24670 }, { "grad_norm": 0.9644839763641357, "learning_rate": 8.978079815914236e-05, "loss": 0.0755, "step": 24680 }, { "grad_norm": 0.7730404734611511, "learning_rate": 8.977077924216202e-05, "loss": 0.0594, "step": 24690 }, { "grad_norm": 0.5827149152755737, "learning_rate": 8.976075597590905e-05, "loss": 0.058, "step": 24700 }, { "grad_norm": 0.5808035135269165, "learning_rate": 8.975072836147958e-05, "loss": 0.0831, "step": 24710 }, { "grad_norm": 0.4108971655368805, "learning_rate": 8.974069639997025e-05, "loss": 0.0456, "step": 24720 }, { "grad_norm": 0.6490606069564819, "learning_rate": 8.973066009247808e-05, "loss": 0.0584, "step": 24730 }, { "grad_norm": 0.5087289810180664, "learning_rate": 8.972061944010066e-05, "loss": 0.0589, "step": 24740 }, { "grad_norm": 0.8840323686599731, "learning_rate": 8.971057444393603e-05, "loss": 0.0575, "step": 24750 }, { "grad_norm": 0.7459710836410522, "learning_rate": 8.970052510508268e-05, "loss": 0.0649, "step": 24760 }, { "grad_norm": 0.754507839679718, "learning_rate": 8.969047142463959e-05, "loss": 0.0709, "step": 24770 }, { "grad_norm": 0.6065576672554016, "learning_rate": 8.968041340370621e-05, "loss": 0.0606, "step": 24780 }, { "grad_norm": 0.6339989304542542, "learning_rate": 8.96703510433825e-05, "loss": 0.0668, "step": 24790 }, { "grad_norm": 1.0168102979660034, "learning_rate": 8.966028434476883e-05, "loss": 0.0673, "step": 24800 }, { "grad_norm": 0.5064639449119568, "learning_rate": 8.96502133089661e-05, "loss": 0.0654, "step": 24810 }, { "grad_norm": 0.5313208699226379, "learning_rate": 8.964013793707564e-05, "loss": 0.0706, "step": 24820 }, { "grad_norm": 0.4612565040588379, "learning_rate": 8.963005823019932e-05, "loss": 0.0554, "step": 24830 }, { "grad_norm": 0.6946414113044739, "learning_rate": 8.961997418943939e-05, "loss": 0.0582, "step": 24840 }, { "grad_norm": 0.7682949900627136, "learning_rate": 8.960988581589865e-05, "loss": 0.0667, "step": 24850 }, { "grad_norm": 0.7892441749572754, "learning_rate": 8.959979311068037e-05, "loss": 0.0821, "step": 24860 }, { "grad_norm": 0.8050259947776794, "learning_rate": 8.958969607488823e-05, "loss": 0.0784, "step": 24870 }, { "grad_norm": 0.6511209011077881, "learning_rate": 8.957959470962647e-05, "loss": 0.0696, "step": 24880 }, { "grad_norm": 0.6213780641555786, "learning_rate": 8.956948901599971e-05, "loss": 0.065, "step": 24890 }, { "grad_norm": 0.6692705750465393, "learning_rate": 8.955937899511315e-05, "loss": 0.0729, "step": 24900 }, { "grad_norm": 0.5111452341079712, "learning_rate": 8.954926464807238e-05, "loss": 0.0567, "step": 24910 }, { "grad_norm": 0.6484368443489075, "learning_rate": 8.953914597598347e-05, "loss": 0.0859, "step": 24920 }, { "grad_norm": 0.4008750319480896, "learning_rate": 8.952902297995303e-05, "loss": 0.0674, "step": 24930 }, { "grad_norm": 0.7355577945709229, "learning_rate": 8.951889566108804e-05, "loss": 0.0568, "step": 24940 }, { "grad_norm": 0.5502939820289612, "learning_rate": 8.950876402049606e-05, "loss": 0.0486, "step": 24950 }, { "grad_norm": 0.5397590398788452, "learning_rate": 8.949862805928504e-05, "loss": 0.0536, "step": 24960 }, { "grad_norm": 0.6881162524223328, "learning_rate": 8.948848777856343e-05, "loss": 0.0545, "step": 24970 }, { "grad_norm": 0.4612123370170593, "learning_rate": 8.947834317944017e-05, "loss": 0.0631, "step": 24980 }, { "grad_norm": 0.52192622423172, "learning_rate": 8.946819426302466e-05, "loss": 0.0766, "step": 24990 }, { "grad_norm": 0.5895112752914429, "learning_rate": 8.945804103042676e-05, "loss": 0.0567, "step": 25000 }, { "grad_norm": 0.861078679561615, "learning_rate": 8.944788348275681e-05, "loss": 0.0805, "step": 25010 }, { "grad_norm": 0.6125785708427429, "learning_rate": 8.943772162112565e-05, "loss": 0.089, "step": 25020 }, { "grad_norm": 0.47669270634651184, "learning_rate": 8.942755544664454e-05, "loss": 0.0752, "step": 25030 }, { "grad_norm": 0.5135210752487183, "learning_rate": 8.941738496042525e-05, "loss": 0.0744, "step": 25040 }, { "grad_norm": 0.7101136445999146, "learning_rate": 8.940721016357999e-05, "loss": 0.0826, "step": 25050 }, { "grad_norm": 1.0962014198303223, "learning_rate": 8.939703105722148e-05, "loss": 0.0757, "step": 25060 }, { "grad_norm": 0.8002169132232666, "learning_rate": 8.93868476424629e-05, "loss": 0.0664, "step": 25070 }, { "grad_norm": 0.6405255198478699, "learning_rate": 8.937665992041786e-05, "loss": 0.0778, "step": 25080 }, { "grad_norm": 0.8372513055801392, "learning_rate": 8.93664678922005e-05, "loss": 0.0813, "step": 25090 }, { "grad_norm": 0.9109705090522766, "learning_rate": 8.93562715589254e-05, "loss": 0.0724, "step": 25100 }, { "grad_norm": 0.5603980422019958, "learning_rate": 8.934607092170762e-05, "loss": 0.0719, "step": 25110 }, { "grad_norm": 0.8050946593284607, "learning_rate": 8.933586598166266e-05, "loss": 0.0649, "step": 25120 }, { "grad_norm": 0.5469125509262085, "learning_rate": 8.932565673990655e-05, "loss": 0.0647, "step": 25130 }, { "grad_norm": 0.7402544617652893, "learning_rate": 8.931544319755574e-05, "loss": 0.0731, "step": 25140 }, { "grad_norm": 0.6481015086174011, "learning_rate": 8.930522535572718e-05, "loss": 0.0626, "step": 25150 }, { "grad_norm": 0.9095219969749451, "learning_rate": 8.929500321553826e-05, "loss": 0.0644, "step": 25160 }, { "grad_norm": 0.7818571925163269, "learning_rate": 8.928477677810686e-05, "loss": 0.0785, "step": 25170 }, { "grad_norm": 0.6421324014663696, "learning_rate": 8.927454604455137e-05, "loss": 0.057, "step": 25180 }, { "grad_norm": 0.9791440367698669, "learning_rate": 8.926431101599053e-05, "loss": 0.0749, "step": 25190 }, { "grad_norm": 0.6397925019264221, "learning_rate": 8.925407169354369e-05, "loss": 0.0747, "step": 25200 }, { "grad_norm": 0.6145147681236267, "learning_rate": 8.92438280783306e-05, "loss": 0.06, "step": 25210 }, { "grad_norm": 0.6613629460334778, "learning_rate": 8.923358017147146e-05, "loss": 0.0632, "step": 25220 }, { "grad_norm": 0.6529679894447327, "learning_rate": 8.922332797408697e-05, "loss": 0.0542, "step": 25230 }, { "grad_norm": 0.43674468994140625, "learning_rate": 8.921307148729831e-05, "loss": 0.0642, "step": 25240 }, { "grad_norm": 0.46219146251678467, "learning_rate": 8.920281071222712e-05, "loss": 0.0518, "step": 25250 }, { "grad_norm": 0.7485427260398865, "learning_rate": 8.919254564999548e-05, "loss": 0.0567, "step": 25260 }, { "grad_norm": 0.5616505742073059, "learning_rate": 8.918227630172598e-05, "loss": 0.0677, "step": 25270 }, { "grad_norm": 0.751234769821167, "learning_rate": 8.917200266854165e-05, "loss": 0.0888, "step": 25280 }, { "grad_norm": 0.6984492540359497, "learning_rate": 8.9161724751566e-05, "loss": 0.0594, "step": 25290 }, { "grad_norm": 0.7298682928085327, "learning_rate": 8.915144255192302e-05, "loss": 0.0675, "step": 25300 }, { "grad_norm": 0.6528634428977966, "learning_rate": 8.914115607073714e-05, "loss": 0.0532, "step": 25310 }, { "grad_norm": 0.9782761335372925, "learning_rate": 8.913086530913327e-05, "loss": 0.082, "step": 25320 }, { "grad_norm": 0.6668620109558105, "learning_rate": 8.912057026823681e-05, "loss": 0.0682, "step": 25330 }, { "grad_norm": 0.8750947713851929, "learning_rate": 8.91102709491736e-05, "loss": 0.0679, "step": 25340 }, { "grad_norm": 0.6201645731925964, "learning_rate": 8.909996735306996e-05, "loss": 0.056, "step": 25350 }, { "grad_norm": 0.6765099167823792, "learning_rate": 8.908965948105268e-05, "loss": 0.0835, "step": 25360 }, { "grad_norm": 0.6270646452903748, "learning_rate": 8.907934733424901e-05, "loss": 0.0573, "step": 25370 }, { "grad_norm": 0.7121774554252625, "learning_rate": 8.906903091378666e-05, "loss": 0.0731, "step": 25380 }, { "grad_norm": 0.7866692543029785, "learning_rate": 8.905871022079384e-05, "loss": 0.074, "step": 25390 }, { "grad_norm": 1.054587721824646, "learning_rate": 8.90483852563992e-05, "loss": 0.0617, "step": 25400 }, { "grad_norm": 1.1369032859802246, "learning_rate": 8.903805602173185e-05, "loss": 0.0575, "step": 25410 }, { "grad_norm": 1.1450034379959106, "learning_rate": 8.902772251792137e-05, "loss": 0.0673, "step": 25420 }, { "grad_norm": 0.737481415271759, "learning_rate": 8.901738474609786e-05, "loss": 0.0671, "step": 25430 }, { "grad_norm": 0.6497815847396851, "learning_rate": 8.900704270739179e-05, "loss": 0.0635, "step": 25440 }, { "grad_norm": 0.4904618561267853, "learning_rate": 8.89966964029342e-05, "loss": 0.0685, "step": 25450 }, { "grad_norm": 0.798392653465271, "learning_rate": 8.898634583385652e-05, "loss": 0.0631, "step": 25460 }, { "grad_norm": 0.7455979585647583, "learning_rate": 8.897599100129065e-05, "loss": 0.0674, "step": 25470 }, { "grad_norm": 0.5591856837272644, "learning_rate": 8.896563190636903e-05, "loss": 0.0635, "step": 25480 }, { "grad_norm": 0.6680186986923218, "learning_rate": 8.895526855022448e-05, "loss": 0.0591, "step": 25490 }, { "grad_norm": 0.3705199062824249, "learning_rate": 8.894490093399033e-05, "loss": 0.0483, "step": 25500 }, { "grad_norm": 0.5195206999778748, "learning_rate": 8.893452905880035e-05, "loss": 0.0722, "step": 25510 }, { "grad_norm": 0.6889044642448425, "learning_rate": 8.892415292578883e-05, "loss": 0.0769, "step": 25520 }, { "grad_norm": 0.6468475461006165, "learning_rate": 8.891377253609046e-05, "loss": 0.0885, "step": 25530 }, { "grad_norm": 0.6240228414535522, "learning_rate": 8.890338789084043e-05, "loss": 0.065, "step": 25540 }, { "grad_norm": 0.735396146774292, "learning_rate": 8.88929989911744e-05, "loss": 0.0654, "step": 25550 }, { "grad_norm": 0.7277148962020874, "learning_rate": 8.888260583822847e-05, "loss": 0.069, "step": 25560 }, { "grad_norm": 0.920173168182373, "learning_rate": 8.887220843313921e-05, "loss": 0.0787, "step": 25570 }, { "grad_norm": 0.4949610233306885, "learning_rate": 8.88618067770437e-05, "loss": 0.0574, "step": 25580 }, { "grad_norm": 0.6549193263053894, "learning_rate": 8.885140087107942e-05, "loss": 0.0623, "step": 25590 }, { "grad_norm": 0.8905200958251953, "learning_rate": 8.884099071638436e-05, "loss": 0.0678, "step": 25600 }, { "grad_norm": 0.7285430431365967, "learning_rate": 8.883057631409695e-05, "loss": 0.0709, "step": 25610 }, { "grad_norm": 0.5034683346748352, "learning_rate": 8.882015766535608e-05, "loss": 0.0773, "step": 25620 }, { "grad_norm": 0.541135847568512, "learning_rate": 8.880973477130115e-05, "loss": 0.0684, "step": 25630 }, { "grad_norm": 0.37008050084114075, "learning_rate": 8.879930763307197e-05, "loss": 0.0572, "step": 25640 }, { "grad_norm": 1.164826512336731, "learning_rate": 8.878887625180884e-05, "loss": 0.0682, "step": 25650 }, { "grad_norm": 0.8509688973426819, "learning_rate": 8.877844062865253e-05, "loss": 0.0715, "step": 25660 }, { "grad_norm": 0.5466315150260925, "learning_rate": 8.876800076474424e-05, "loss": 0.0521, "step": 25670 }, { "grad_norm": 0.39452987909317017, "learning_rate": 8.875755666122568e-05, "loss": 0.0558, "step": 25680 }, { "grad_norm": 0.7429144382476807, "learning_rate": 8.8747108319239e-05, "loss": 0.0498, "step": 25690 }, { "grad_norm": 0.5501088500022888, "learning_rate": 8.87366557399268e-05, "loss": 0.054, "step": 25700 }, { "grad_norm": 0.6903983354568481, "learning_rate": 8.872619892443217e-05, "loss": 0.0712, "step": 25710 }, { "grad_norm": 0.37081024050712585, "learning_rate": 8.871573787389865e-05, "loss": 0.063, "step": 25720 }, { "grad_norm": 0.5137118101119995, "learning_rate": 8.870527258947024e-05, "loss": 0.0643, "step": 25730 }, { "grad_norm": 0.8612167239189148, "learning_rate": 8.869480307229143e-05, "loss": 0.0663, "step": 25740 }, { "grad_norm": 0.5314634442329407, "learning_rate": 8.868432932350712e-05, "loss": 0.0485, "step": 25750 }, { "grad_norm": 0.5200369358062744, "learning_rate": 8.867385134426272e-05, "loss": 0.0581, "step": 25760 }, { "grad_norm": 0.6266282796859741, "learning_rate": 8.866336913570407e-05, "loss": 0.0583, "step": 25770 }, { "grad_norm": 0.6581119298934937, "learning_rate": 8.865288269897751e-05, "loss": 0.0696, "step": 25780 }, { "grad_norm": 0.4405023455619812, "learning_rate": 8.864239203522981e-05, "loss": 0.0748, "step": 25790 }, { "grad_norm": 0.4665422737598419, "learning_rate": 8.863189714560822e-05, "loss": 0.0712, "step": 25800 }, { "grad_norm": 0.6986181735992432, "learning_rate": 8.862139803126043e-05, "loss": 0.075, "step": 25810 }, { "grad_norm": 0.7120136618614197, "learning_rate": 8.861089469333463e-05, "loss": 0.0588, "step": 25820 }, { "grad_norm": 0.4904029667377472, "learning_rate": 8.860038713297944e-05, "loss": 0.0528, "step": 25830 }, { "grad_norm": 0.73731929063797, "learning_rate": 8.858987535134394e-05, "loss": 0.0648, "step": 25840 }, { "grad_norm": 0.5558491349220276, "learning_rate": 8.857935934957769e-05, "loss": 0.0535, "step": 25850 }, { "grad_norm": 0.7454808354377747, "learning_rate": 8.856883912883071e-05, "loss": 0.0691, "step": 25860 }, { "grad_norm": 0.7641825675964355, "learning_rate": 8.855831469025346e-05, "loss": 0.0597, "step": 25870 }, { "grad_norm": 0.6620141267776489, "learning_rate": 8.854778603499689e-05, "loss": 0.0669, "step": 25880 }, { "grad_norm": 0.5190871953964233, "learning_rate": 8.85372531642124e-05, "loss": 0.056, "step": 25890 }, { "grad_norm": 0.5164166688919067, "learning_rate": 8.852671607905185e-05, "loss": 0.0472, "step": 25900 }, { "grad_norm": 0.40901950001716614, "learning_rate": 8.851617478066754e-05, "loss": 0.0561, "step": 25910 }, { "grad_norm": 0.627755343914032, "learning_rate": 8.850562927021227e-05, "loss": 0.0604, "step": 25920 }, { "grad_norm": 0.4865238070487976, "learning_rate": 8.849507954883928e-05, "loss": 0.0616, "step": 25930 }, { "grad_norm": 0.7112263441085815, "learning_rate": 8.848452561770226e-05, "loss": 0.0639, "step": 25940 }, { "grad_norm": 0.5538089871406555, "learning_rate": 8.847396747795538e-05, "loss": 0.0497, "step": 25950 }, { "grad_norm": 0.4553898572921753, "learning_rate": 8.846340513075327e-05, "loss": 0.0571, "step": 25960 }, { "grad_norm": 0.6150168776512146, "learning_rate": 8.845283857725099e-05, "loss": 0.0713, "step": 25970 }, { "grad_norm": 0.5473587512969971, "learning_rate": 8.844226781860409e-05, "loss": 0.0462, "step": 25980 }, { "grad_norm": 0.27799198031425476, "learning_rate": 8.84316928559686e-05, "loss": 0.059, "step": 25990 }, { "grad_norm": 0.6574429869651794, "learning_rate": 8.842111369050094e-05, "loss": 0.0542, "step": 26000 }, { "grad_norm": 0.8638331294059753, "learning_rate": 8.841053032335808e-05, "loss": 0.0707, "step": 26010 }, { "grad_norm": 0.5790804028511047, "learning_rate": 8.839994275569735e-05, "loss": 0.0849, "step": 26020 }, { "grad_norm": 0.9933264255523682, "learning_rate": 8.838935098867662e-05, "loss": 0.0632, "step": 26030 }, { "grad_norm": 0.7594457864761353, "learning_rate": 8.837875502345418e-05, "loss": 0.0759, "step": 26040 }, { "grad_norm": 0.5957802534103394, "learning_rate": 8.83681548611888e-05, "loss": 0.0915, "step": 26050 }, { "grad_norm": 0.6952584981918335, "learning_rate": 8.835755050303969e-05, "loss": 0.0693, "step": 26060 }, { "grad_norm": 0.8482658863067627, "learning_rate": 8.834694195016653e-05, "loss": 0.0482, "step": 26070 }, { "grad_norm": 0.6031813621520996, "learning_rate": 8.833632920372942e-05, "loss": 0.0565, "step": 26080 }, { "grad_norm": 0.8511720299720764, "learning_rate": 8.832571226488903e-05, "loss": 0.0526, "step": 26090 }, { "grad_norm": 0.6329925656318665, "learning_rate": 8.831509113480634e-05, "loss": 0.046, "step": 26100 }, { "grad_norm": 0.6198030710220337, "learning_rate": 8.83044658146429e-05, "loss": 0.0578, "step": 26110 }, { "grad_norm": 0.6833028197288513, "learning_rate": 8.829383630556067e-05, "loss": 0.0602, "step": 26120 }, { "grad_norm": 0.5630984306335449, "learning_rate": 8.828320260872207e-05, "loss": 0.062, "step": 26130 }, { "grad_norm": 0.7579520344734192, "learning_rate": 8.827256472529e-05, "loss": 0.0902, "step": 26140 }, { "grad_norm": 0.7149955034255981, "learning_rate": 8.826192265642778e-05, "loss": 0.0587, "step": 26150 }, { "grad_norm": 0.8593595027923584, "learning_rate": 8.825127640329923e-05, "loss": 0.0537, "step": 26160 }, { "grad_norm": 0.6260187029838562, "learning_rate": 8.824062596706861e-05, "loss": 0.0502, "step": 26170 }, { "grad_norm": 0.7226477861404419, "learning_rate": 8.822997134890062e-05, "loss": 0.0732, "step": 26180 }, { "grad_norm": 0.7470952272415161, "learning_rate": 8.821931254996044e-05, "loss": 0.0676, "step": 26190 }, { "grad_norm": 0.5582023859024048, "learning_rate": 8.82086495714137e-05, "loss": 0.0606, "step": 26200 }, { "grad_norm": 0.5510509014129639, "learning_rate": 8.81979824144265e-05, "loss": 0.0523, "step": 26210 }, { "grad_norm": 0.6812443733215332, "learning_rate": 8.818731108016536e-05, "loss": 0.0669, "step": 26220 }, { "grad_norm": 0.6425209641456604, "learning_rate": 8.81766355697973e-05, "loss": 0.0647, "step": 26230 }, { "grad_norm": 0.8429871797561646, "learning_rate": 8.816595588448977e-05, "loss": 0.0631, "step": 26240 }, { "grad_norm": 0.6308342218399048, "learning_rate": 8.81552720254107e-05, "loss": 0.0752, "step": 26250 }, { "grad_norm": 0.6642541885375977, "learning_rate": 8.814458399372842e-05, "loss": 0.0542, "step": 26260 }, { "grad_norm": 0.500108540058136, "learning_rate": 8.813389179061181e-05, "loss": 0.0531, "step": 26270 }, { "grad_norm": 0.4890458583831787, "learning_rate": 8.812319541723012e-05, "loss": 0.0624, "step": 26280 }, { "grad_norm": 0.561985969543457, "learning_rate": 8.811249487475309e-05, "loss": 0.055, "step": 26290 }, { "grad_norm": 0.5521506667137146, "learning_rate": 8.810179016435092e-05, "loss": 0.0649, "step": 26300 }, { "grad_norm": 0.6015092134475708, "learning_rate": 8.809108128719428e-05, "loss": 0.0464, "step": 26310 }, { "grad_norm": 0.3631734848022461, "learning_rate": 8.808036824445424e-05, "loss": 0.045, "step": 26320 }, { "grad_norm": 0.9572807550430298, "learning_rate": 8.806965103730238e-05, "loss": 0.0739, "step": 26330 }, { "grad_norm": 0.6344606280326843, "learning_rate": 8.805892966691074e-05, "loss": 0.0526, "step": 26340 }, { "grad_norm": 0.5441756844520569, "learning_rate": 8.804820413445175e-05, "loss": 0.0682, "step": 26350 }, { "grad_norm": 0.47050318121910095, "learning_rate": 8.803747444109837e-05, "loss": 0.0505, "step": 26360 }, { "grad_norm": 0.5455765128135681, "learning_rate": 8.802674058802399e-05, "loss": 0.0502, "step": 26370 }, { "grad_norm": 0.5837761759757996, "learning_rate": 8.801600257640241e-05, "loss": 0.0569, "step": 26380 }, { "grad_norm": 0.6566590070724487, "learning_rate": 8.800526040740795e-05, "loss": 0.0559, "step": 26390 }, { "grad_norm": 0.6936398148536682, "learning_rate": 8.799451408221535e-05, "loss": 0.0937, "step": 26400 }, { "grad_norm": 0.5151925683021545, "learning_rate": 8.798376360199982e-05, "loss": 0.0568, "step": 26410 }, { "grad_norm": 0.7423343658447266, "learning_rate": 8.797300896793701e-05, "loss": 0.0438, "step": 26420 }, { "grad_norm": 0.732240617275238, "learning_rate": 8.796225018120302e-05, "loss": 0.0654, "step": 26430 }, { "grad_norm": 0.4429579973220825, "learning_rate": 8.795148724297444e-05, "loss": 0.0492, "step": 26440 }, { "grad_norm": 0.47808754444122314, "learning_rate": 8.794072015442825e-05, "loss": 0.0509, "step": 26450 }, { "grad_norm": 0.7944058179855347, "learning_rate": 8.792994891674198e-05, "loss": 0.0521, "step": 26460 }, { "grad_norm": 0.8404315114021301, "learning_rate": 8.79191735310935e-05, "loss": 0.0866, "step": 26470 }, { "grad_norm": 0.6070707440376282, "learning_rate": 8.790839399866122e-05, "loss": 0.0609, "step": 26480 }, { "grad_norm": 0.5793141722679138, "learning_rate": 8.789761032062397e-05, "loss": 0.0549, "step": 26490 }, { "grad_norm": 0.503459095954895, "learning_rate": 8.788682249816103e-05, "loss": 0.043, "step": 26500 }, { "grad_norm": 0.6482664346694946, "learning_rate": 8.787603053245215e-05, "loss": 0.0564, "step": 26510 }, { "grad_norm": 0.6922561526298523, "learning_rate": 8.78652344246775e-05, "loss": 0.0902, "step": 26520 }, { "grad_norm": 0.7989283800125122, "learning_rate": 8.785443417601776e-05, "loss": 0.0693, "step": 26530 }, { "grad_norm": 0.5809928774833679, "learning_rate": 8.784362978765401e-05, "loss": 0.0521, "step": 26540 }, { "grad_norm": 0.8081580400466919, "learning_rate": 8.783282126076779e-05, "loss": 0.059, "step": 26550 }, { "grad_norm": 0.6364911198616028, "learning_rate": 8.782200859654112e-05, "loss": 0.0503, "step": 26560 }, { "grad_norm": 0.8810157179832458, "learning_rate": 8.781119179615646e-05, "loss": 0.047, "step": 26570 }, { "grad_norm": 0.5390669703483582, "learning_rate": 8.780037086079674e-05, "loss": 0.0586, "step": 26580 }, { "grad_norm": 0.4941972494125366, "learning_rate": 8.778954579164527e-05, "loss": 0.0456, "step": 26590 }, { "grad_norm": 0.5546192526817322, "learning_rate": 8.777871658988588e-05, "loss": 0.0549, "step": 26600 }, { "grad_norm": 1.0292696952819824, "learning_rate": 8.776788325670285e-05, "loss": 0.0619, "step": 26610 }, { "grad_norm": 0.5029594898223877, "learning_rate": 8.775704579328089e-05, "loss": 0.0766, "step": 26620 }, { "grad_norm": 0.623504638671875, "learning_rate": 8.774620420080517e-05, "loss": 0.07, "step": 26630 }, { "grad_norm": 0.6595993638038635, "learning_rate": 8.773535848046131e-05, "loss": 0.0728, "step": 26640 }, { "grad_norm": 0.5390982031822205, "learning_rate": 8.772450863343538e-05, "loss": 0.0566, "step": 26650 }, { "grad_norm": 0.694746732711792, "learning_rate": 8.77136546609139e-05, "loss": 0.0536, "step": 26660 }, { "grad_norm": 0.7670135498046875, "learning_rate": 8.770279656408385e-05, "loss": 0.0837, "step": 26670 }, { "grad_norm": 0.7144254446029663, "learning_rate": 8.769193434413265e-05, "loss": 0.0529, "step": 26680 }, { "grad_norm": 0.650834321975708, "learning_rate": 8.76810680022482e-05, "loss": 0.059, "step": 26690 }, { "grad_norm": 0.725711464881897, "learning_rate": 8.767019753961878e-05, "loss": 0.0466, "step": 26700 }, { "grad_norm": 0.63492351770401, "learning_rate": 8.765932295743321e-05, "loss": 0.0559, "step": 26710 }, { "grad_norm": 0.3691171407699585, "learning_rate": 8.764844425688068e-05, "loss": 0.0553, "step": 26720 }, { "grad_norm": 0.9217612147331238, "learning_rate": 8.763756143915092e-05, "loss": 0.0594, "step": 26730 }, { "grad_norm": 0.6528527736663818, "learning_rate": 8.7626674505434e-05, "loss": 0.0711, "step": 26740 }, { "grad_norm": 0.7568041086196899, "learning_rate": 8.761578345692053e-05, "loss": 0.0511, "step": 26750 }, { "grad_norm": 0.6712261438369751, "learning_rate": 8.760488829480156e-05, "loss": 0.0606, "step": 26760 }, { "grad_norm": 0.7332906723022461, "learning_rate": 8.759398902026854e-05, "loss": 0.0644, "step": 26770 }, { "grad_norm": 0.9650977849960327, "learning_rate": 8.758308563451339e-05, "loss": 0.0642, "step": 26780 }, { "grad_norm": 0.6508862972259521, "learning_rate": 8.75721781387285e-05, "loss": 0.0695, "step": 26790 }, { "grad_norm": 0.43295544385910034, "learning_rate": 8.75612665341067e-05, "loss": 0.0577, "step": 26800 }, { "grad_norm": 0.881504237651825, "learning_rate": 8.755035082184126e-05, "loss": 0.0613, "step": 26810 }, { "grad_norm": 0.5052363872528076, "learning_rate": 8.753943100312592e-05, "loss": 0.048, "step": 26820 }, { "grad_norm": 0.4286463260650635, "learning_rate": 8.752850707915484e-05, "loss": 0.0625, "step": 26830 }, { "grad_norm": 0.6384840607643127, "learning_rate": 8.751757905112264e-05, "loss": 0.0524, "step": 26840 }, { "grad_norm": 0.8419556021690369, "learning_rate": 8.75066469202244e-05, "loss": 0.0593, "step": 26850 }, { "grad_norm": 0.5673034191131592, "learning_rate": 8.749571068765567e-05, "loss": 0.0551, "step": 26860 }, { "grad_norm": 0.35602128505706787, "learning_rate": 8.748477035461238e-05, "loss": 0.0605, "step": 26870 }, { "grad_norm": 0.7989517450332642, "learning_rate": 8.747382592229095e-05, "loss": 0.0625, "step": 26880 }, { "grad_norm": 0.6227601170539856, "learning_rate": 8.746287739188828e-05, "loss": 0.06, "step": 26890 }, { "grad_norm": 0.7756356000900269, "learning_rate": 8.745192476460165e-05, "loss": 0.078, "step": 26900 }, { "grad_norm": 0.43828174471855164, "learning_rate": 8.744096804162882e-05, "loss": 0.0895, "step": 26910 }, { "grad_norm": 1.6464377641677856, "learning_rate": 8.743000722416804e-05, "loss": 0.0598, "step": 26920 }, { "grad_norm": 0.6500731110572815, "learning_rate": 8.741904231341793e-05, "loss": 0.0776, "step": 26930 }, { "grad_norm": 0.7296499609947205, "learning_rate": 8.740807331057762e-05, "loss": 0.0801, "step": 26940 }, { "grad_norm": 0.6515093445777893, "learning_rate": 8.739710021684667e-05, "loss": 0.0616, "step": 26950 }, { "grad_norm": 0.7695393562316895, "learning_rate": 8.738612303342503e-05, "loss": 0.0762, "step": 26960 }, { "grad_norm": 0.6000638008117676, "learning_rate": 8.73751417615132e-05, "loss": 0.0768, "step": 26970 }, { "grad_norm": 0.5154496431350708, "learning_rate": 8.736415640231208e-05, "loss": 0.0596, "step": 26980 }, { "grad_norm": 0.47019389271736145, "learning_rate": 8.735316695702297e-05, "loss": 0.0398, "step": 26990 }, { "grad_norm": 0.5808984637260437, "learning_rate": 8.734217342684769e-05, "loss": 0.0579, "step": 27000 }, { "grad_norm": 0.7981351613998413, "learning_rate": 8.733117581298847e-05, "loss": 0.0676, "step": 27010 }, { "grad_norm": 0.34913375973701477, "learning_rate": 8.732017411664796e-05, "loss": 0.0489, "step": 27020 }, { "grad_norm": 0.6613159775733948, "learning_rate": 8.730916833902936e-05, "loss": 0.0748, "step": 27030 }, { "grad_norm": 0.5231829285621643, "learning_rate": 8.729815848133618e-05, "loss": 0.0695, "step": 27040 }, { "grad_norm": 0.6132832169532776, "learning_rate": 8.728714454477247e-05, "loss": 0.082, "step": 27050 }, { "grad_norm": 0.6890149116516113, "learning_rate": 8.727612653054269e-05, "loss": 0.072, "step": 27060 }, { "grad_norm": 0.5698198676109314, "learning_rate": 8.726510443985176e-05, "loss": 0.0857, "step": 27070 }, { "grad_norm": 0.5404027700424194, "learning_rate": 8.725407827390503e-05, "loss": 0.0485, "step": 27080 }, { "grad_norm": 0.6207746267318726, "learning_rate": 8.724304803390833e-05, "loss": 0.0517, "step": 27090 }, { "grad_norm": 0.5725296139717102, "learning_rate": 8.723201372106788e-05, "loss": 0.0609, "step": 27100 }, { "grad_norm": 0.4212014973163605, "learning_rate": 8.722097533659038e-05, "loss": 0.07, "step": 27110 }, { "grad_norm": 0.5687318444252014, "learning_rate": 8.720993288168299e-05, "loss": 0.0598, "step": 27120 }, { "grad_norm": 0.4481736123561859, "learning_rate": 8.719888635755327e-05, "loss": 0.059, "step": 27130 }, { "grad_norm": 0.5860828757286072, "learning_rate": 8.718783576540928e-05, "loss": 0.0502, "step": 27140 }, { "grad_norm": 0.68625408411026, "learning_rate": 8.717678110645948e-05, "loss": 0.0501, "step": 27150 }, { "grad_norm": 0.6010351181030273, "learning_rate": 8.716572238191279e-05, "loss": 0.0579, "step": 27160 }, { "grad_norm": 0.7650757431983948, "learning_rate": 8.715465959297857e-05, "loss": 0.0813, "step": 27170 }, { "grad_norm": 0.6051108241081238, "learning_rate": 8.714359274086665e-05, "loss": 0.0574, "step": 27180 }, { "grad_norm": 0.6366086602210999, "learning_rate": 8.713252182678726e-05, "loss": 0.0996, "step": 27190 }, { "grad_norm": 0.8396928906440735, "learning_rate": 8.712144685195112e-05, "loss": 0.066, "step": 27200 }, { "grad_norm": 0.736702024936676, "learning_rate": 8.711036781756936e-05, "loss": 0.0516, "step": 27210 }, { "grad_norm": 0.5670284628868103, "learning_rate": 8.709928472485357e-05, "loss": 0.0527, "step": 27220 }, { "grad_norm": 0.5210492014884949, "learning_rate": 8.708819757501579e-05, "loss": 0.0461, "step": 27230 }, { "grad_norm": 0.700473964214325, "learning_rate": 8.707710636926846e-05, "loss": 0.0765, "step": 27240 }, { "grad_norm": 0.4925887882709503, "learning_rate": 8.706601110882455e-05, "loss": 0.047, "step": 27250 }, { "grad_norm": 0.7978522181510925, "learning_rate": 8.705491179489738e-05, "loss": 0.0539, "step": 27260 }, { "grad_norm": 0.45475512742996216, "learning_rate": 8.704380842870077e-05, "loss": 0.041, "step": 27270 }, { "grad_norm": 0.42053377628326416, "learning_rate": 8.703270101144895e-05, "loss": 0.0372, "step": 27280 }, { "grad_norm": 0.5601793527603149, "learning_rate": 8.702158954435664e-05, "loss": 0.0562, "step": 27290 }, { "grad_norm": 0.4217994809150696, "learning_rate": 8.701047402863896e-05, "loss": 0.0693, "step": 27300 }, { "grad_norm": 0.48205360770225525, "learning_rate": 8.699935446551148e-05, "loss": 0.0673, "step": 27310 }, { "grad_norm": 0.5723078846931458, "learning_rate": 8.698823085619022e-05, "loss": 0.0646, "step": 27320 }, { "grad_norm": 0.35590824484825134, "learning_rate": 8.697710320189166e-05, "loss": 0.0545, "step": 27330 }, { "grad_norm": 0.5102130770683289, "learning_rate": 8.696597150383268e-05, "loss": 0.0661, "step": 27340 }, { "grad_norm": 0.8836430311203003, "learning_rate": 8.695483576323063e-05, "loss": 0.0549, "step": 27350 }, { "grad_norm": 0.6869199872016907, "learning_rate": 8.69436959813033e-05, "loss": 0.0548, "step": 27360 }, { "grad_norm": 0.48243165016174316, "learning_rate": 8.693255215926892e-05, "loss": 0.0634, "step": 27370 }, { "grad_norm": 0.7895901799201965, "learning_rate": 8.692140429834617e-05, "loss": 0.0536, "step": 27380 }, { "grad_norm": 0.6868446469306946, "learning_rate": 8.691025239975415e-05, "loss": 0.0699, "step": 27390 }, { "grad_norm": 0.8648872375488281, "learning_rate": 8.689909646471243e-05, "loss": 0.081, "step": 27400 }, { "grad_norm": 0.843585193157196, "learning_rate": 8.688793649444099e-05, "loss": 0.0733, "step": 27410 }, { "grad_norm": 0.6830079555511475, "learning_rate": 8.687677249016029e-05, "loss": 0.0614, "step": 27420 }, { "grad_norm": 0.6057128310203552, "learning_rate": 8.686560445309118e-05, "loss": 0.058, "step": 27430 }, { "grad_norm": 0.4901363253593445, "learning_rate": 8.685443238445499e-05, "loss": 0.0528, "step": 27440 }, { "grad_norm": 0.7904564142227173, "learning_rate": 8.68432562854735e-05, "loss": 0.0628, "step": 27450 }, { "grad_norm": 0.7462652921676636, "learning_rate": 8.683207615736887e-05, "loss": 0.057, "step": 27460 }, { "grad_norm": 0.5680897235870361, "learning_rate": 8.682089200136379e-05, "loss": 0.052, "step": 27470 }, { "grad_norm": 0.5922772884368896, "learning_rate": 8.680970381868132e-05, "loss": 0.0583, "step": 27480 }, { "grad_norm": 0.6232773065567017, "learning_rate": 8.679851161054498e-05, "loss": 0.0651, "step": 27490 }, { "grad_norm": 0.6654723286628723, "learning_rate": 8.678731537817873e-05, "loss": 0.0503, "step": 27500 }, { "grad_norm": 0.5384201407432556, "learning_rate": 8.677611512280697e-05, "loss": 0.0705, "step": 27510 }, { "grad_norm": 0.624618649482727, "learning_rate": 8.676491084565457e-05, "loss": 0.0939, "step": 27520 }, { "grad_norm": 0.6487059593200684, "learning_rate": 8.675370254794678e-05, "loss": 0.0577, "step": 27530 }, { "grad_norm": 0.4338919222354889, "learning_rate": 8.674249023090935e-05, "loss": 0.0524, "step": 27540 }, { "grad_norm": 0.3846522569656372, "learning_rate": 8.673127389576843e-05, "loss": 0.0583, "step": 27550 }, { "grad_norm": 0.6526152491569519, "learning_rate": 8.67200535437506e-05, "loss": 0.0547, "step": 27560 }, { "grad_norm": 0.7216732501983643, "learning_rate": 8.670882917608296e-05, "loss": 0.0698, "step": 27570 }, { "grad_norm": 0.6951150894165039, "learning_rate": 8.669760079399292e-05, "loss": 0.0587, "step": 27580 }, { "grad_norm": 1.0557268857955933, "learning_rate": 8.668636839870845e-05, "loss": 0.0593, "step": 27590 }, { "grad_norm": 0.5494657158851624, "learning_rate": 8.667513199145789e-05, "loss": 0.0651, "step": 27600 }, { "grad_norm": 0.5060943961143494, "learning_rate": 8.666389157347002e-05, "loss": 0.0744, "step": 27610 }, { "grad_norm": 0.6319811344146729, "learning_rate": 8.66526471459741e-05, "loss": 0.0627, "step": 27620 }, { "grad_norm": 0.6404223442077637, "learning_rate": 8.66413987101998e-05, "loss": 0.0576, "step": 27630 }, { "grad_norm": 0.45069798827171326, "learning_rate": 8.663014626737723e-05, "loss": 0.0722, "step": 27640 }, { "grad_norm": 0.7354658246040344, "learning_rate": 8.661888981873691e-05, "loss": 0.0792, "step": 27650 }, { "grad_norm": 1.0894672870635986, "learning_rate": 8.660762936550988e-05, "loss": 0.0752, "step": 27660 }, { "grad_norm": 0.7755995988845825, "learning_rate": 8.659636490892753e-05, "loss": 0.0671, "step": 27670 }, { "grad_norm": 0.7615054845809937, "learning_rate": 8.658509645022174e-05, "loss": 0.0692, "step": 27680 }, { "grad_norm": 0.5214378833770752, "learning_rate": 8.657382399062481e-05, "loss": 0.046, "step": 27690 }, { "grad_norm": 0.6693738698959351, "learning_rate": 8.656254753136946e-05, "loss": 0.0592, "step": 27700 }, { "grad_norm": 0.7322148680686951, "learning_rate": 8.655126707368891e-05, "loss": 0.066, "step": 27710 }, { "grad_norm": 0.6545186042785645, "learning_rate": 8.653998261881672e-05, "loss": 0.0504, "step": 27720 }, { "grad_norm": 0.6075306534767151, "learning_rate": 8.652869416798699e-05, "loss": 0.0841, "step": 27730 }, { "grad_norm": 0.8194342255592346, "learning_rate": 8.651740172243417e-05, "loss": 0.0636, "step": 27740 }, { "grad_norm": 0.568179726600647, "learning_rate": 8.65061052833932e-05, "loss": 0.0631, "step": 27750 }, { "grad_norm": 0.6693857312202454, "learning_rate": 8.649480485209945e-05, "loss": 0.0482, "step": 27760 }, { "grad_norm": 0.6263279318809509, "learning_rate": 8.64835004297887e-05, "loss": 0.0608, "step": 27770 }, { "grad_norm": 0.8442308902740479, "learning_rate": 8.64721920176972e-05, "loss": 0.0928, "step": 27780 }, { "grad_norm": 0.8901479244232178, "learning_rate": 8.646087961706164e-05, "loss": 0.0593, "step": 27790 }, { "grad_norm": 0.5784755349159241, "learning_rate": 8.644956322911908e-05, "loss": 0.0631, "step": 27800 }, { "grad_norm": 1.197954535484314, "learning_rate": 8.643824285510709e-05, "loss": 0.0742, "step": 27810 }, { "grad_norm": 0.9347872138023376, "learning_rate": 8.642691849626364e-05, "loss": 0.0602, "step": 27820 }, { "grad_norm": 0.676547110080719, "learning_rate": 8.641559015382717e-05, "loss": 0.0844, "step": 27830 }, { "grad_norm": 0.5089895129203796, "learning_rate": 8.640425782903649e-05, "loss": 0.0547, "step": 27840 }, { "grad_norm": 0.7822219729423523, "learning_rate": 8.639292152313091e-05, "loss": 0.0513, "step": 27850 }, { "grad_norm": 0.7273155450820923, "learning_rate": 8.638158123735015e-05, "loss": 0.0666, "step": 27860 }, { "grad_norm": 0.8805534243583679, "learning_rate": 8.637023697293436e-05, "loss": 0.0645, "step": 27870 }, { "grad_norm": 0.6644729971885681, "learning_rate": 8.635888873112414e-05, "loss": 0.0535, "step": 27880 }, { "grad_norm": 0.5062054395675659, "learning_rate": 8.634753651316052e-05, "loss": 0.0653, "step": 27890 }, { "grad_norm": 0.6080183386802673, "learning_rate": 8.633618032028496e-05, "loss": 0.0692, "step": 27900 }, { "grad_norm": 0.4850911498069763, "learning_rate": 8.632482015373934e-05, "loss": 0.0602, "step": 27910 }, { "grad_norm": 0.6151760816574097, "learning_rate": 8.6313456014766e-05, "loss": 0.0685, "step": 27920 }, { "grad_norm": 0.6571665406227112, "learning_rate": 8.630208790460771e-05, "loss": 0.0873, "step": 27930 }, { "grad_norm": 0.9935288429260254, "learning_rate": 8.629071582450768e-05, "loss": 0.0882, "step": 27940 }, { "grad_norm": 0.6930984854698181, "learning_rate": 8.62793397757095e-05, "loss": 0.0628, "step": 27950 }, { "grad_norm": 0.8323072195053101, "learning_rate": 8.626795975945729e-05, "loss": 0.073, "step": 27960 }, { "grad_norm": 1.0872440338134766, "learning_rate": 8.625657577699551e-05, "loss": 0.0828, "step": 27970 }, { "grad_norm": 0.6163191199302673, "learning_rate": 8.624518782956914e-05, "loss": 0.0539, "step": 27980 }, { "grad_norm": 0.7864904999732971, "learning_rate": 8.62337959184235e-05, "loss": 0.0723, "step": 27990 }, { "grad_norm": 0.6624781489372253, "learning_rate": 8.622240004480441e-05, "loss": 0.0596, "step": 28000 }, { "grad_norm": 0.9578549265861511, "learning_rate": 8.621100020995814e-05, "loss": 0.0718, "step": 28010 }, { "grad_norm": 0.5350413918495178, "learning_rate": 8.619959641513132e-05, "loss": 0.0641, "step": 28020 }, { "grad_norm": 0.6487065553665161, "learning_rate": 8.618818866157105e-05, "loss": 0.0774, "step": 28030 }, { "grad_norm": 0.7522705793380737, "learning_rate": 8.617677695052487e-05, "loss": 0.0653, "step": 28040 }, { "grad_norm": 0.6045382618904114, "learning_rate": 8.616536128324078e-05, "loss": 0.0625, "step": 28050 }, { "grad_norm": 0.4711821973323822, "learning_rate": 8.615394166096712e-05, "loss": 0.0615, "step": 28060 }, { "grad_norm": 0.6630086898803711, "learning_rate": 8.614251808495279e-05, "loss": 0.0532, "step": 28070 }, { "grad_norm": 0.5064035058021545, "learning_rate": 8.6131090556447e-05, "loss": 0.0506, "step": 28080 }, { "grad_norm": 0.7094131112098694, "learning_rate": 8.611965907669947e-05, "loss": 0.054, "step": 28090 }, { "grad_norm": 0.4691004455089569, "learning_rate": 8.610822364696034e-05, "loss": 0.0496, "step": 28100 }, { "grad_norm": 0.8593310713768005, "learning_rate": 8.609678426848015e-05, "loss": 0.0578, "step": 28110 }, { "grad_norm": 0.6459159255027771, "learning_rate": 8.60853409425099e-05, "loss": 0.0559, "step": 28120 }, { "grad_norm": 0.4254169464111328, "learning_rate": 8.607389367030104e-05, "loss": 0.0658, "step": 28130 }, { "grad_norm": 0.7387769222259521, "learning_rate": 8.606244245310538e-05, "loss": 0.0662, "step": 28140 }, { "grad_norm": 0.7347027659416199, "learning_rate": 8.605098729217525e-05, "loss": 0.0716, "step": 28150 }, { "grad_norm": 0.8002103567123413, "learning_rate": 8.603952818876335e-05, "loss": 0.0549, "step": 28160 }, { "grad_norm": 0.5289435386657715, "learning_rate": 8.602806514412281e-05, "loss": 0.068, "step": 28170 }, { "grad_norm": 0.6121037602424622, "learning_rate": 8.601659815950726e-05, "loss": 0.0498, "step": 28180 }, { "grad_norm": 0.929263710975647, "learning_rate": 8.600512723617067e-05, "loss": 0.0587, "step": 28190 }, { "grad_norm": 0.4984990358352661, "learning_rate": 8.59936523753675e-05, "loss": 0.0563, "step": 28200 }, { "grad_norm": 0.5382267236709595, "learning_rate": 8.598217357835264e-05, "loss": 0.0496, "step": 28210 }, { "grad_norm": 0.775170624256134, "learning_rate": 8.597069084638135e-05, "loss": 0.0665, "step": 28220 }, { "grad_norm": 0.9698160886764526, "learning_rate": 8.595920418070939e-05, "loss": 0.0473, "step": 28230 }, { "grad_norm": 0.6555355191230774, "learning_rate": 8.594771358259295e-05, "loss": 0.0515, "step": 28240 }, { "grad_norm": 0.8159235715866089, "learning_rate": 8.593621905328858e-05, "loss": 0.0507, "step": 28250 }, { "grad_norm": 0.6012915372848511, "learning_rate": 8.592472059405333e-05, "loss": 0.0699, "step": 28260 }, { "grad_norm": 0.8554127812385559, "learning_rate": 8.591321820614464e-05, "loss": 0.0805, "step": 28270 }, { "grad_norm": 0.581131637096405, "learning_rate": 8.590171189082041e-05, "loss": 0.0583, "step": 28280 }, { "grad_norm": 0.5303578972816467, "learning_rate": 8.589020164933894e-05, "loss": 0.0488, "step": 28290 }, { "grad_norm": 0.8097518682479858, "learning_rate": 8.587868748295898e-05, "loss": 0.0726, "step": 28300 }, { "grad_norm": 0.9595269560813904, "learning_rate": 8.586716939293971e-05, "loss": 0.0823, "step": 28310 }, { "grad_norm": 0.9741398096084595, "learning_rate": 8.58556473805407e-05, "loss": 0.0693, "step": 28320 }, { "grad_norm": 0.8259432315826416, "learning_rate": 8.584412144702202e-05, "loss": 0.0566, "step": 28330 }, { "grad_norm": 0.7551407814025879, "learning_rate": 8.58325915936441e-05, "loss": 0.0578, "step": 28340 }, { "grad_norm": 0.44712957739830017, "learning_rate": 8.582105782166783e-05, "loss": 0.0575, "step": 28350 }, { "grad_norm": 0.5467517375946045, "learning_rate": 8.580952013235455e-05, "loss": 0.0476, "step": 28360 }, { "grad_norm": 0.8758025765419006, "learning_rate": 8.579797852696596e-05, "loss": 0.0654, "step": 28370 }, { "grad_norm": 0.7443386316299438, "learning_rate": 8.578643300676428e-05, "loss": 0.058, "step": 28380 }, { "grad_norm": 0.8012157082557678, "learning_rate": 8.577488357301209e-05, "loss": 0.059, "step": 28390 }, { "grad_norm": 0.5654558539390564, "learning_rate": 8.576333022697242e-05, "loss": 0.0506, "step": 28400 }, { "grad_norm": 0.7806534171104431, "learning_rate": 8.575177296990873e-05, "loss": 0.0613, "step": 28410 }, { "grad_norm": 0.8064211010932922, "learning_rate": 8.574021180308489e-05, "loss": 0.0611, "step": 28420 }, { "grad_norm": 0.7732758522033691, "learning_rate": 8.572864672776523e-05, "loss": 0.0959, "step": 28430 }, { "grad_norm": 0.49165233969688416, "learning_rate": 8.571707774521447e-05, "loss": 0.0404, "step": 28440 }, { "grad_norm": 0.5515072345733643, "learning_rate": 8.57055048566978e-05, "loss": 0.0623, "step": 28450 }, { "grad_norm": 0.6302871108055115, "learning_rate": 8.569392806348078e-05, "loss": 0.044, "step": 28460 }, { "grad_norm": 0.4671187400817871, "learning_rate": 8.568234736682947e-05, "loss": 0.0559, "step": 28470 }, { "grad_norm": 0.6887956261634827, "learning_rate": 8.567076276801029e-05, "loss": 0.0624, "step": 28480 }, { "grad_norm": 0.5986855030059814, "learning_rate": 8.565917426829013e-05, "loss": 0.068, "step": 28490 }, { "grad_norm": 0.610286295413971, "learning_rate": 8.564758186893628e-05, "loss": 0.0585, "step": 28500 }, { "grad_norm": 0.6495296955108643, "learning_rate": 8.563598557121649e-05, "loss": 0.0567, "step": 28510 }, { "grad_norm": 0.47180306911468506, "learning_rate": 8.562438537639888e-05, "loss": 0.0625, "step": 28520 }, { "grad_norm": 0.9214743971824646, "learning_rate": 8.561278128575206e-05, "loss": 0.0762, "step": 28530 }, { "grad_norm": 0.5694481134414673, "learning_rate": 8.5601173300545e-05, "loss": 0.0671, "step": 28540 }, { "grad_norm": 0.6264482140541077, "learning_rate": 8.558956142204717e-05, "loss": 0.0583, "step": 28550 }, { "grad_norm": 0.5227018594741821, "learning_rate": 8.55779456515284e-05, "loss": 0.0565, "step": 28560 }, { "grad_norm": 0.7932324409484863, "learning_rate": 8.556632599025898e-05, "loss": 0.0728, "step": 28570 }, { "grad_norm": 0.6208620071411133, "learning_rate": 8.555470243950964e-05, "loss": 0.0558, "step": 28580 }, { "grad_norm": 0.7124881744384766, "learning_rate": 8.554307500055148e-05, "loss": 0.0591, "step": 28590 }, { "grad_norm": 0.711031436920166, "learning_rate": 8.553144367465609e-05, "loss": 0.0806, "step": 28600 }, { "grad_norm": 0.5292753577232361, "learning_rate": 8.551980846309544e-05, "loss": 0.0528, "step": 28610 }, { "grad_norm": 0.6141519546508789, "learning_rate": 8.550816936714193e-05, "loss": 0.0587, "step": 28620 }, { "grad_norm": 0.4930410087108612, "learning_rate": 8.549652638806841e-05, "loss": 0.084, "step": 28630 }, { "grad_norm": 0.5351129174232483, "learning_rate": 8.548487952714812e-05, "loss": 0.0473, "step": 28640 }, { "grad_norm": 0.5046980381011963, "learning_rate": 8.547322878565478e-05, "loss": 0.0623, "step": 28650 }, { "grad_norm": 0.6483556628227234, "learning_rate": 8.546157416486245e-05, "loss": 0.0455, "step": 28660 }, { "grad_norm": 0.5957053899765015, "learning_rate": 8.54499156660457e-05, "loss": 0.0566, "step": 28670 }, { "grad_norm": 1.1049420833587646, "learning_rate": 8.543825329047947e-05, "loss": 0.0568, "step": 28680 }, { "grad_norm": 0.68559730052948, "learning_rate": 8.542658703943913e-05, "loss": 0.0564, "step": 28690 }, { "grad_norm": 0.9316090941429138, "learning_rate": 8.541491691420051e-05, "loss": 0.0804, "step": 28700 }, { "grad_norm": 0.6598924994468689, "learning_rate": 8.54032429160398e-05, "loss": 0.0509, "step": 28710 }, { "grad_norm": 0.6129103302955627, "learning_rate": 8.539156504623369e-05, "loss": 0.0485, "step": 28720 }, { "grad_norm": 1.201121211051941, "learning_rate": 8.537988330605923e-05, "loss": 0.0523, "step": 28730 }, { "grad_norm": 0.7868419289588928, "learning_rate": 8.536819769679393e-05, "loss": 0.0502, "step": 28740 }, { "grad_norm": 0.8266069293022156, "learning_rate": 8.53565082197157e-05, "loss": 0.0605, "step": 28750 }, { "grad_norm": 0.4653393626213074, "learning_rate": 8.534481487610289e-05, "loss": 0.0547, "step": 28760 }, { "grad_norm": 0.6968278884887695, "learning_rate": 8.533311766723428e-05, "loss": 0.0604, "step": 28770 }, { "grad_norm": 0.5546461343765259, "learning_rate": 8.532141659438901e-05, "loss": 0.078, "step": 28780 }, { "grad_norm": 0.498382031917572, "learning_rate": 8.530971165884675e-05, "loss": 0.0673, "step": 28790 }, { "grad_norm": 0.6751869320869446, "learning_rate": 8.529800286188752e-05, "loss": 0.048, "step": 28800 }, { "grad_norm": 0.4791843295097351, "learning_rate": 8.528629020479175e-05, "loss": 0.0552, "step": 28810 }, { "grad_norm": 0.4950236976146698, "learning_rate": 8.527457368884033e-05, "loss": 0.0544, "step": 28820 }, { "grad_norm": 0.6391816735267639, "learning_rate": 8.526285331531458e-05, "loss": 0.054, "step": 28830 }, { "grad_norm": 0.6223805546760559, "learning_rate": 8.525112908549621e-05, "loss": 0.0584, "step": 28840 }, { "grad_norm": 0.5351467132568359, "learning_rate": 8.523940100066735e-05, "loss": 0.055, "step": 28850 }, { "grad_norm": 0.8027659058570862, "learning_rate": 8.52276690621106e-05, "loss": 0.0633, "step": 28860 }, { "grad_norm": 0.9202778339385986, "learning_rate": 8.521593327110889e-05, "loss": 0.0591, "step": 28870 }, { "grad_norm": 0.4917513430118561, "learning_rate": 8.520419362894569e-05, "loss": 0.0523, "step": 28880 }, { "grad_norm": 0.5635359287261963, "learning_rate": 8.51924501369048e-05, "loss": 0.0558, "step": 28890 }, { "grad_norm": 0.5574584603309631, "learning_rate": 8.518070279627047e-05, "loss": 0.0524, "step": 28900 }, { "grad_norm": 0.5424038171768188, "learning_rate": 8.516895160832737e-05, "loss": 0.068, "step": 28910 }, { "grad_norm": 0.6094598174095154, "learning_rate": 8.515719657436061e-05, "loss": 0.0475, "step": 28920 }, { "grad_norm": 0.6853286027908325, "learning_rate": 8.514543769565568e-05, "loss": 0.0483, "step": 28930 }, { "grad_norm": 0.5709862112998962, "learning_rate": 8.513367497349853e-05, "loss": 0.0794, "step": 28940 }, { "grad_norm": 0.5415998101234436, "learning_rate": 8.51219084091755e-05, "loss": 0.0611, "step": 28950 }, { "grad_norm": 0.7507919073104858, "learning_rate": 8.511013800397338e-05, "loss": 0.0652, "step": 28960 }, { "grad_norm": 0.6597094535827637, "learning_rate": 8.509836375917937e-05, "loss": 0.0594, "step": 28970 }, { "grad_norm": 0.5931372046470642, "learning_rate": 8.508658567608104e-05, "loss": 0.0502, "step": 28980 }, { "grad_norm": 0.6178900599479675, "learning_rate": 8.507480375596647e-05, "loss": 0.0733, "step": 28990 }, { "grad_norm": 0.6219725608825684, "learning_rate": 8.506301800012408e-05, "loss": 0.0781, "step": 29000 }, { "grad_norm": 0.7355655431747437, "learning_rate": 8.505122840984278e-05, "loss": 0.0599, "step": 29010 }, { "grad_norm": 0.4945739507675171, "learning_rate": 8.503943498641182e-05, "loss": 0.0519, "step": 29020 }, { "grad_norm": 0.44953835010528564, "learning_rate": 8.502763773112095e-05, "loss": 0.0576, "step": 29030 }, { "grad_norm": 0.7315079569816589, "learning_rate": 8.501583664526026e-05, "loss": 0.0519, "step": 29040 }, { "grad_norm": 0.598014235496521, "learning_rate": 8.500403173012032e-05, "loss": 0.0919, "step": 29050 }, { "grad_norm": 0.6427421569824219, "learning_rate": 8.499222298699211e-05, "loss": 0.0466, "step": 29060 }, { "grad_norm": 0.4542202353477478, "learning_rate": 8.498041041716701e-05, "loss": 0.054, "step": 29070 }, { "grad_norm": 0.9276020526885986, "learning_rate": 8.496859402193681e-05, "loss": 0.056, "step": 29080 }, { "grad_norm": 0.5962585210800171, "learning_rate": 8.495677380259374e-05, "loss": 0.0529, "step": 29090 }, { "grad_norm": 0.8051475882530212, "learning_rate": 8.494494976043045e-05, "loss": 0.0726, "step": 29100 }, { "grad_norm": 0.5186856389045715, "learning_rate": 8.493312189673998e-05, "loss": 0.0593, "step": 29110 }, { "grad_norm": 0.5810421705245972, "learning_rate": 8.492129021281584e-05, "loss": 0.0615, "step": 29120 }, { "grad_norm": 0.5492552518844604, "learning_rate": 8.490945470995188e-05, "loss": 0.05, "step": 29130 }, { "grad_norm": 0.45091524720191956, "learning_rate": 8.489761538944247e-05, "loss": 0.049, "step": 29140 }, { "grad_norm": 0.6688235998153687, "learning_rate": 8.48857722525823e-05, "loss": 0.0752, "step": 29150 }, { "grad_norm": 0.6929740309715271, "learning_rate": 8.487392530066652e-05, "loss": 0.0547, "step": 29160 }, { "grad_norm": 0.7658695578575134, "learning_rate": 8.486207453499069e-05, "loss": 0.0684, "step": 29170 }, { "grad_norm": 0.457388311624527, "learning_rate": 8.485021995685082e-05, "loss": 0.0694, "step": 29180 }, { "grad_norm": 0.7554125189781189, "learning_rate": 8.483836156754328e-05, "loss": 0.0532, "step": 29190 }, { "grad_norm": 0.698671817779541, "learning_rate": 8.482649936836491e-05, "loss": 0.0821, "step": 29200 }, { "grad_norm": 0.39446982741355896, "learning_rate": 8.481463336061293e-05, "loss": 0.0405, "step": 29210 }, { "grad_norm": 0.6849027276039124, "learning_rate": 8.480276354558496e-05, "loss": 0.0476, "step": 29220 }, { "grad_norm": 0.6728796362876892, "learning_rate": 8.479088992457913e-05, "loss": 0.0534, "step": 29230 }, { "grad_norm": 0.5440759062767029, "learning_rate": 8.477901249889387e-05, "loss": 0.0584, "step": 29240 }, { "grad_norm": 0.5529820322990417, "learning_rate": 8.47671312698281e-05, "loss": 0.0461, "step": 29250 }, { "grad_norm": 0.6339840292930603, "learning_rate": 8.475524623868112e-05, "loss": 0.0616, "step": 29260 }, { "grad_norm": 0.4354795217514038, "learning_rate": 8.474335740675266e-05, "loss": 0.0499, "step": 29270 }, { "grad_norm": 0.5105798244476318, "learning_rate": 8.473146477534289e-05, "loss": 0.0381, "step": 29280 }, { "grad_norm": 0.583057165145874, "learning_rate": 8.471956834575232e-05, "loss": 0.0544, "step": 29290 }, { "grad_norm": 0.7092256546020508, "learning_rate": 8.470766811928197e-05, "loss": 0.058, "step": 29300 }, { "grad_norm": 0.5182752013206482, "learning_rate": 8.469576409723323e-05, "loss": 0.0569, "step": 29310 }, { "grad_norm": 0.9172281622886658, "learning_rate": 8.468385628090788e-05, "loss": 0.0539, "step": 29320 }, { "grad_norm": 0.6870521306991577, "learning_rate": 8.467194467160815e-05, "loss": 0.0696, "step": 29330 }, { "grad_norm": 0.6869657635688782, "learning_rate": 8.466002927063667e-05, "loss": 0.0606, "step": 29340 }, { "grad_norm": 0.6090614199638367, "learning_rate": 8.464811007929651e-05, "loss": 0.0546, "step": 29350 }, { "grad_norm": 0.6983537077903748, "learning_rate": 8.463618709889114e-05, "loss": 0.0523, "step": 29360 }, { "grad_norm": 0.6588764190673828, "learning_rate": 8.462426033072442e-05, "loss": 0.0568, "step": 29370 }, { "grad_norm": 0.8573976755142212, "learning_rate": 8.461232977610061e-05, "loss": 0.0487, "step": 29380 }, { "grad_norm": 0.38141554594039917, "learning_rate": 8.46003954363245e-05, "loss": 0.0546, "step": 29390 }, { "grad_norm": 0.5302128195762634, "learning_rate": 8.458845731270115e-05, "loss": 0.059, "step": 29400 }, { "grad_norm": 0.42444872856140137, "learning_rate": 8.45765154065361e-05, "loss": 0.0482, "step": 29410 }, { "grad_norm": 1.188137412071228, "learning_rate": 8.456456971913532e-05, "loss": 0.0836, "step": 29420 }, { "grad_norm": 0.5321111083030701, "learning_rate": 8.455262025180517e-05, "loss": 0.0502, "step": 29430 }, { "grad_norm": 0.3424254357814789, "learning_rate": 8.454066700585242e-05, "loss": 0.0557, "step": 29440 }, { "grad_norm": 0.4105274975299835, "learning_rate": 8.452870998258423e-05, "loss": 0.0635, "step": 29450 }, { "grad_norm": 0.5101051926612854, "learning_rate": 8.451674918330825e-05, "loss": 0.0685, "step": 29460 }, { "grad_norm": 0.5354620218276978, "learning_rate": 8.450478460933246e-05, "loss": 0.0751, "step": 29470 }, { "grad_norm": 0.6226847767829895, "learning_rate": 8.449281626196532e-05, "loss": 0.0564, "step": 29480 }, { "grad_norm": 0.6374140381813049, "learning_rate": 8.448084414251564e-05, "loss": 0.0491, "step": 29490 }, { "grad_norm": 0.6846103072166443, "learning_rate": 8.446886825229271e-05, "loss": 0.0692, "step": 29500 }, { "grad_norm": 0.5737400054931641, "learning_rate": 8.445688859260615e-05, "loss": 0.065, "step": 29510 }, { "grad_norm": 0.5657024383544922, "learning_rate": 8.444490516476606e-05, "loss": 0.0871, "step": 29520 }, { "grad_norm": 0.7003991603851318, "learning_rate": 8.443291797008293e-05, "loss": 0.0572, "step": 29530 }, { "grad_norm": 0.615995466709137, "learning_rate": 8.442092700986765e-05, "loss": 0.0459, "step": 29540 }, { "grad_norm": 0.7752625346183777, "learning_rate": 8.440893228543156e-05, "loss": 0.0491, "step": 29550 }, { "grad_norm": 0.4197361469268799, "learning_rate": 8.439693379808638e-05, "loss": 0.0581, "step": 29560 }, { "grad_norm": 0.7505691051483154, "learning_rate": 8.43849315491442e-05, "loss": 0.0555, "step": 29570 }, { "grad_norm": 0.45352667570114136, "learning_rate": 8.437292553991763e-05, "loss": 0.0508, "step": 29580 }, { "grad_norm": 0.46393707394599915, "learning_rate": 8.436091577171959e-05, "loss": 0.0614, "step": 29590 }, { "grad_norm": 0.7904545664787292, "learning_rate": 8.434890224586347e-05, "loss": 0.0664, "step": 29600 }, { "grad_norm": 0.5844821333885193, "learning_rate": 8.433688496366303e-05, "loss": 0.0642, "step": 29610 }, { "grad_norm": 0.5640271306037903, "learning_rate": 8.432486392643248e-05, "loss": 0.0524, "step": 29620 }, { "grad_norm": 0.8080571293830872, "learning_rate": 8.431283913548643e-05, "loss": 0.0705, "step": 29630 }, { "grad_norm": 0.6388671398162842, "learning_rate": 8.430081059213985e-05, "loss": 0.0522, "step": 29640 }, { "grad_norm": 0.7765347361564636, "learning_rate": 8.428877829770823e-05, "loss": 0.0939, "step": 29650 }, { "grad_norm": 0.7823896408081055, "learning_rate": 8.427674225350735e-05, "loss": 0.0534, "step": 29660 }, { "grad_norm": 0.4814890921115875, "learning_rate": 8.426470246085347e-05, "loss": 0.0978, "step": 29670 }, { "grad_norm": 0.5849226117134094, "learning_rate": 8.425265892106324e-05, "loss": 0.0656, "step": 29680 }, { "grad_norm": 0.883385419845581, "learning_rate": 8.424061163545374e-05, "loss": 0.0565, "step": 29690 }, { "grad_norm": 0.6300733089447021, "learning_rate": 8.422856060534243e-05, "loss": 0.1036, "step": 29700 }, { "grad_norm": 0.5651357173919678, "learning_rate": 8.421650583204718e-05, "loss": 0.0557, "step": 29710 }, { "grad_norm": 0.4657493829727173, "learning_rate": 8.420444731688633e-05, "loss": 0.061, "step": 29720 }, { "grad_norm": 0.9745444059371948, "learning_rate": 8.419238506117852e-05, "loss": 0.0506, "step": 29730 }, { "grad_norm": 0.4138009548187256, "learning_rate": 8.418031906624289e-05, "loss": 0.0699, "step": 29740 }, { "grad_norm": 0.8187844157218933, "learning_rate": 8.416824933339898e-05, "loss": 0.0547, "step": 29750 }, { "grad_norm": 0.7778370380401611, "learning_rate": 8.415617586396667e-05, "loss": 0.0736, "step": 29760 }, { "grad_norm": 0.6737480163574219, "learning_rate": 8.414409865926632e-05, "loss": 0.0549, "step": 29770 }, { "grad_norm": 0.5272465944290161, "learning_rate": 8.413201772061867e-05, "loss": 0.056, "step": 29780 }, { "grad_norm": 0.5503995418548584, "learning_rate": 8.411993304934488e-05, "loss": 0.0551, "step": 29790 }, { "grad_norm": 0.7057262659072876, "learning_rate": 8.410784464676654e-05, "loss": 0.0558, "step": 29800 }, { "grad_norm": 0.6808003783226013, "learning_rate": 8.409575251420556e-05, "loss": 0.058, "step": 29810 }, { "grad_norm": 0.8485104441642761, "learning_rate": 8.408365665298435e-05, "loss": 0.0581, "step": 29820 }, { "grad_norm": 0.9931054711341858, "learning_rate": 8.40715570644257e-05, "loss": 0.0623, "step": 29830 }, { "grad_norm": 0.6116315722465515, "learning_rate": 8.40594537498528e-05, "loss": 0.0603, "step": 29840 }, { "grad_norm": 0.8432446122169495, "learning_rate": 8.404734671058924e-05, "loss": 0.0514, "step": 29850 }, { "grad_norm": 0.62088942527771, "learning_rate": 8.403523594795902e-05, "loss": 0.0525, "step": 29860 }, { "grad_norm": 0.8037418723106384, "learning_rate": 8.402312146328659e-05, "loss": 0.0706, "step": 29870 }, { "grad_norm": 0.5918948650360107, "learning_rate": 8.401100325789675e-05, "loss": 0.0556, "step": 29880 }, { "grad_norm": 0.4791649878025055, "learning_rate": 8.399888133311472e-05, "loss": 0.0526, "step": 29890 }, { "grad_norm": 0.849177360534668, "learning_rate": 8.398675569026613e-05, "loss": 0.1008, "step": 29900 }, { "grad_norm": 0.719860851764679, "learning_rate": 8.397462633067705e-05, "loss": 0.0581, "step": 29910 }, { "grad_norm": 0.5832352638244629, "learning_rate": 8.396249325567392e-05, "loss": 0.0695, "step": 29920 }, { "grad_norm": 0.7683268189430237, "learning_rate": 8.395035646658357e-05, "loss": 0.0496, "step": 29930 }, { "grad_norm": 0.475237101316452, "learning_rate": 8.39382159647333e-05, "loss": 0.0682, "step": 29940 }, { "grad_norm": 0.8472764492034912, "learning_rate": 8.392607175145075e-05, "loss": 0.0521, "step": 29950 }, { "grad_norm": 0.505450963973999, "learning_rate": 8.3913923828064e-05, "loss": 0.0461, "step": 29960 }, { "grad_norm": 0.7435784935951233, "learning_rate": 8.390177219590152e-05, "loss": 0.0533, "step": 29970 }, { "grad_norm": 0.5060569047927856, "learning_rate": 8.388961685629222e-05, "loss": 0.0545, "step": 29980 }, { "grad_norm": 0.9787224531173706, "learning_rate": 8.387745781056536e-05, "loss": 0.066, "step": 29990 }, { "grad_norm": 0.9585809707641602, "learning_rate": 8.386529506005065e-05, "loss": 0.0682, "step": 30000 }, { "grad_norm": 0.5974560379981995, "learning_rate": 8.38531286060782e-05, "loss": 0.0426, "step": 30010 }, { "grad_norm": 0.3024219572544098, "learning_rate": 8.384095844997849e-05, "loss": 0.0392, "step": 30020 }, { "grad_norm": 0.4539470076560974, "learning_rate": 8.382878459308245e-05, "loss": 0.0407, "step": 30030 }, { "grad_norm": 0.4472258985042572, "learning_rate": 8.381660703672138e-05, "loss": 0.0621, "step": 30040 }, { "grad_norm": 0.6402504444122314, "learning_rate": 8.380442578222702e-05, "loss": 0.0503, "step": 30050 }, { "grad_norm": 0.715416431427002, "learning_rate": 8.379224083093146e-05, "loss": 0.0544, "step": 30060 }, { "grad_norm": 0.6658074259757996, "learning_rate": 8.378005218416727e-05, "loss": 0.0918, "step": 30070 }, { "grad_norm": 0.5378523468971252, "learning_rate": 8.376785984326735e-05, "loss": 0.044, "step": 30080 }, { "grad_norm": 0.6169776320457458, "learning_rate": 8.375566380956506e-05, "loss": 0.0484, "step": 30090 }, { "grad_norm": 0.5499141812324524, "learning_rate": 8.374346408439411e-05, "loss": 0.0809, "step": 30100 }, { "grad_norm": 1.474642038345337, "learning_rate": 8.373126066908868e-05, "loss": 0.0876, "step": 30110 }, { "grad_norm": 0.5469075441360474, "learning_rate": 8.371905356498326e-05, "loss": 0.0578, "step": 30120 }, { "grad_norm": 0.6462624669075012, "learning_rate": 8.370684277341288e-05, "loss": 0.0437, "step": 30130 }, { "grad_norm": 0.82696133852005, "learning_rate": 8.369462829571282e-05, "loss": 0.0696, "step": 30140 }, { "grad_norm": 0.8096324801445007, "learning_rate": 8.36824101332189e-05, "loss": 0.0455, "step": 30150 }, { "grad_norm": 0.6963064670562744, "learning_rate": 8.367018828726721e-05, "loss": 0.062, "step": 30160 }, { "grad_norm": 0.7136299014091492, "learning_rate": 8.365796275919438e-05, "loss": 0.0767, "step": 30170 }, { "grad_norm": 0.8095769882202148, "learning_rate": 8.364573355033734e-05, "loss": 0.0627, "step": 30180 }, { "grad_norm": 1.1449350118637085, "learning_rate": 8.363350066203346e-05, "loss": 0.0725, "step": 30190 }, { "grad_norm": 0.7244926691055298, "learning_rate": 8.362126409562053e-05, "loss": 0.0489, "step": 30200 }, { "grad_norm": 0.3648633360862732, "learning_rate": 8.360902385243667e-05, "loss": 0.0504, "step": 30210 }, { "grad_norm": 0.738550066947937, "learning_rate": 8.359677993382052e-05, "loss": 0.0579, "step": 30220 }, { "grad_norm": 0.5619854927062988, "learning_rate": 8.358453234111103e-05, "loss": 0.0638, "step": 30230 }, { "grad_norm": 0.6515202522277832, "learning_rate": 8.357228107564756e-05, "loss": 0.0588, "step": 30240 }, { "grad_norm": 0.35680389404296875, "learning_rate": 8.356002613876993e-05, "loss": 0.0567, "step": 30250 }, { "grad_norm": 0.4236852824687958, "learning_rate": 8.35477675318183e-05, "loss": 0.0471, "step": 30260 }, { "grad_norm": 0.6745830178260803, "learning_rate": 8.353550525613323e-05, "loss": 0.0404, "step": 30270 }, { "grad_norm": 0.8191913962364197, "learning_rate": 8.352323931305572e-05, "loss": 0.0465, "step": 30280 }, { "grad_norm": 0.7007126212120056, "learning_rate": 8.351096970392717e-05, "loss": 0.0507, "step": 30290 }, { "grad_norm": 0.4786253869533539, "learning_rate": 8.349869643008937e-05, "loss": 0.0462, "step": 30300 }, { "grad_norm": 0.5636030435562134, "learning_rate": 8.348641949288449e-05, "loss": 0.0549, "step": 30310 }, { "grad_norm": 0.8807996511459351, "learning_rate": 8.34741388936551e-05, "loss": 0.0688, "step": 30320 }, { "grad_norm": 0.36287248134613037, "learning_rate": 8.346185463374423e-05, "loss": 0.0536, "step": 30330 }, { "grad_norm": 0.5869978070259094, "learning_rate": 8.344956671449524e-05, "loss": 0.0735, "step": 30340 }, { "grad_norm": 0.5865240097045898, "learning_rate": 8.343727513725192e-05, "loss": 0.0458, "step": 30350 }, { "grad_norm": 0.6403157711029053, "learning_rate": 8.342497990335847e-05, "loss": 0.0531, "step": 30360 }, { "grad_norm": 0.9851850867271423, "learning_rate": 8.341268101415946e-05, "loss": 0.0546, "step": 30370 }, { "grad_norm": 0.6380877494812012, "learning_rate": 8.34003784709999e-05, "loss": 0.0418, "step": 30380 }, { "grad_norm": 0.8085491061210632, "learning_rate": 8.338807227522516e-05, "loss": 0.0608, "step": 30390 }, { "grad_norm": 0.8545091152191162, "learning_rate": 8.337576242818103e-05, "loss": 0.0543, "step": 30400 }, { "grad_norm": 0.5315381288528442, "learning_rate": 8.336344893121372e-05, "loss": 0.0606, "step": 30410 }, { "grad_norm": 0.5783113241195679, "learning_rate": 8.335113178566977e-05, "loss": 0.044, "step": 30420 }, { "grad_norm": 0.7452690601348877, "learning_rate": 8.33388109928962e-05, "loss": 0.0587, "step": 30430 }, { "grad_norm": 0.7334824204444885, "learning_rate": 8.33264865542404e-05, "loss": 0.0582, "step": 30440 }, { "grad_norm": 0.5761064887046814, "learning_rate": 8.331415847105013e-05, "loss": 0.058, "step": 30450 }, { "grad_norm": 0.7378759980201721, "learning_rate": 8.330182674467357e-05, "loss": 0.0486, "step": 30460 }, { "grad_norm": 0.7604551911354065, "learning_rate": 8.32894913764593e-05, "loss": 0.0499, "step": 30470 }, { "grad_norm": 0.4048168957233429, "learning_rate": 8.327715236775633e-05, "loss": 0.0476, "step": 30480 }, { "grad_norm": 0.7241169214248657, "learning_rate": 8.326480971991398e-05, "loss": 0.0478, "step": 30490 }, { "grad_norm": 0.6829982399940491, "learning_rate": 8.325246343428206e-05, "loss": 0.087, "step": 30500 }, { "grad_norm": 0.7217009663581848, "learning_rate": 8.324011351221072e-05, "loss": 0.0566, "step": 30510 }, { "grad_norm": 0.48233574628829956, "learning_rate": 8.322775995505057e-05, "loss": 0.0526, "step": 30520 }, { "grad_norm": 0.8394272923469543, "learning_rate": 8.321540276415254e-05, "loss": 0.051, "step": 30530 }, { "grad_norm": 0.6346523761749268, "learning_rate": 8.320304194086798e-05, "loss": 0.0688, "step": 30540 }, { "grad_norm": 0.7638611197471619, "learning_rate": 8.31906774865487e-05, "loss": 0.0551, "step": 30550 }, { "grad_norm": 0.5280537009239197, "learning_rate": 8.317830940254682e-05, "loss": 0.0475, "step": 30560 }, { "grad_norm": 0.7298116683959961, "learning_rate": 8.316593769021491e-05, "loss": 0.0704, "step": 30570 }, { "grad_norm": 0.6217509508132935, "learning_rate": 8.315356235090592e-05, "loss": 0.0585, "step": 30580 }, { "grad_norm": 0.748466968536377, "learning_rate": 8.314118338597319e-05, "loss": 0.0493, "step": 30590 }, { "grad_norm": 0.6889912486076355, "learning_rate": 8.312880079677048e-05, "loss": 0.063, "step": 30600 }, { "grad_norm": 0.6779010891914368, "learning_rate": 8.311641458465191e-05, "loss": 0.0774, "step": 30610 }, { "grad_norm": 0.6492477655410767, "learning_rate": 8.310402475097205e-05, "loss": 0.0509, "step": 30620 }, { "grad_norm": 0.8883407711982727, "learning_rate": 8.309163129708581e-05, "loss": 0.0549, "step": 30630 }, { "grad_norm": 1.000390648841858, "learning_rate": 8.307923422434852e-05, "loss": 0.0486, "step": 30640 }, { "grad_norm": 0.7365864515304565, "learning_rate": 8.30668335341159e-05, "loss": 0.0474, "step": 30650 }, { "grad_norm": 0.6632148027420044, "learning_rate": 8.305442922774408e-05, "loss": 0.0545, "step": 30660 }, { "grad_norm": 0.6067889928817749, "learning_rate": 8.304202130658959e-05, "loss": 0.0631, "step": 30670 }, { "grad_norm": 0.5586692094802856, "learning_rate": 8.302960977200931e-05, "loss": 0.0643, "step": 30680 }, { "grad_norm": 1.2473618984222412, "learning_rate": 8.301719462536058e-05, "loss": 0.0621, "step": 30690 }, { "grad_norm": 0.7428673505783081, "learning_rate": 8.300477586800108e-05, "loss": 0.0764, "step": 30700 }, { "grad_norm": 0.7509905099868774, "learning_rate": 8.299235350128892e-05, "loss": 0.0486, "step": 30710 }, { "grad_norm": 0.4735362231731415, "learning_rate": 8.297992752658258e-05, "loss": 0.0517, "step": 30720 }, { "grad_norm": 0.6178908348083496, "learning_rate": 8.296749794524094e-05, "loss": 0.0768, "step": 30730 }, { "grad_norm": 0.8235010504722595, "learning_rate": 8.295506475862332e-05, "loss": 0.052, "step": 30740 }, { "grad_norm": 0.4685142934322357, "learning_rate": 8.294262796808933e-05, "loss": 0.0575, "step": 30750 }, { "grad_norm": 0.9316502809524536, "learning_rate": 8.293018757499909e-05, "loss": 0.0394, "step": 30760 }, { "grad_norm": 0.47582072019577026, "learning_rate": 8.291774358071305e-05, "loss": 0.0508, "step": 30770 }, { "grad_norm": 1.0347458124160767, "learning_rate": 8.290529598659205e-05, "loss": 0.0584, "step": 30780 }, { "grad_norm": 0.5332809686660767, "learning_rate": 8.289284479399738e-05, "loss": 0.0661, "step": 30790 }, { "grad_norm": 0.5595408082008362, "learning_rate": 8.288039000429064e-05, "loss": 0.0492, "step": 30800 }, { "grad_norm": 0.6357834339141846, "learning_rate": 8.286793161883388e-05, "loss": 0.0508, "step": 30810 }, { "grad_norm": 0.6886292099952698, "learning_rate": 8.285546963898954e-05, "loss": 0.0504, "step": 30820 }, { "grad_norm": 0.7591543793678284, "learning_rate": 8.284300406612044e-05, "loss": 0.063, "step": 30830 }, { "grad_norm": 0.4399099051952362, "learning_rate": 8.283053490158978e-05, "loss": 0.0582, "step": 30840 }, { "grad_norm": 0.6063951849937439, "learning_rate": 8.28180621467612e-05, "loss": 0.0695, "step": 30850 }, { "grad_norm": 0.5931396484375, "learning_rate": 8.280558580299867e-05, "loss": 0.0823, "step": 30860 }, { "grad_norm": 1.1604316234588623, "learning_rate": 8.279310587166661e-05, "loss": 0.0572, "step": 30870 }, { "grad_norm": 1.0641039609909058, "learning_rate": 8.278062235412978e-05, "loss": 0.0522, "step": 30880 }, { "grad_norm": 0.6821928024291992, "learning_rate": 8.276813525175339e-05, "loss": 0.076, "step": 30890 }, { "grad_norm": 0.4824386537075043, "learning_rate": 8.2755644565903e-05, "loss": 0.0571, "step": 30900 }, { "grad_norm": 0.7191377878189087, "learning_rate": 8.274315029794454e-05, "loss": 0.0586, "step": 30910 }, { "grad_norm": 0.5479428768157959, "learning_rate": 8.273065244924443e-05, "loss": 0.0589, "step": 30920 }, { "grad_norm": 0.4656144082546234, "learning_rate": 8.271815102116936e-05, "loss": 0.0451, "step": 30930 }, { "grad_norm": 0.4688870310783386, "learning_rate": 8.270564601508648e-05, "loss": 0.0509, "step": 30940 }, { "grad_norm": 0.46094387769699097, "learning_rate": 8.269313743236333e-05, "loss": 0.0555, "step": 30950 }, { "grad_norm": 1.0318650007247925, "learning_rate": 8.268062527436783e-05, "loss": 0.0715, "step": 30960 }, { "grad_norm": 0.47274765372276306, "learning_rate": 8.266810954246828e-05, "loss": 0.0584, "step": 30970 }, { "grad_norm": 0.80938720703125, "learning_rate": 8.265559023803338e-05, "loss": 0.0548, "step": 30980 }, { "grad_norm": 0.5059708952903748, "learning_rate": 8.264306736243223e-05, "loss": 0.048, "step": 30990 }, { "grad_norm": 0.6973666548728943, "learning_rate": 8.263054091703432e-05, "loss": 0.0636, "step": 31000 }, { "grad_norm": 0.6295280456542969, "learning_rate": 8.26180109032095e-05, "loss": 0.0634, "step": 31010 }, { "grad_norm": 0.708530604839325, "learning_rate": 8.260547732232807e-05, "loss": 0.0546, "step": 31020 }, { "grad_norm": 0.702016294002533, "learning_rate": 8.259294017576063e-05, "loss": 0.0516, "step": 31030 }, { "grad_norm": 0.4899987280368805, "learning_rate": 8.258039946487828e-05, "loss": 0.0416, "step": 31040 }, { "grad_norm": 0.6024342179298401, "learning_rate": 8.256785519105241e-05, "loss": 0.0552, "step": 31050 }, { "grad_norm": 0.5638710260391235, "learning_rate": 8.255530735565488e-05, "loss": 0.0532, "step": 31060 }, { "grad_norm": 0.5777182579040527, "learning_rate": 8.254275596005787e-05, "loss": 0.0547, "step": 31070 }, { "grad_norm": 0.7313320636749268, "learning_rate": 8.2530201005634e-05, "loss": 0.0433, "step": 31080 }, { "grad_norm": 0.7790588736534119, "learning_rate": 8.251764249375626e-05, "loss": 0.0626, "step": 31090 }, { "grad_norm": 0.5365225076675415, "learning_rate": 8.250508042579803e-05, "loss": 0.0451, "step": 31100 }, { "grad_norm": 0.5205854177474976, "learning_rate": 8.249251480313307e-05, "loss": 0.0751, "step": 31110 }, { "grad_norm": 0.6952489018440247, "learning_rate": 8.247994562713555e-05, "loss": 0.0587, "step": 31120 }, { "grad_norm": 0.6452251672744751, "learning_rate": 8.246737289917998e-05, "loss": 0.0554, "step": 31130 }, { "grad_norm": 0.5530452132225037, "learning_rate": 8.245479662064135e-05, "loss": 0.068, "step": 31140 }, { "grad_norm": 0.40111085772514343, "learning_rate": 8.244221679289496e-05, "loss": 0.0516, "step": 31150 }, { "grad_norm": 0.7329855561256409, "learning_rate": 8.242963341731652e-05, "loss": 0.0678, "step": 31160 }, { "grad_norm": 0.9857061505317688, "learning_rate": 8.24170464952821e-05, "loss": 0.0502, "step": 31170 }, { "grad_norm": 0.762641429901123, "learning_rate": 8.240445602816824e-05, "loss": 0.0467, "step": 31180 }, { "grad_norm": 0.5891866683959961, "learning_rate": 8.239186201735179e-05, "loss": 0.0613, "step": 31190 }, { "grad_norm": 0.5943124890327454, "learning_rate": 8.237926446420998e-05, "loss": 0.0554, "step": 31200 }, { "grad_norm": 0.6294950842857361, "learning_rate": 8.236666337012052e-05, "loss": 0.0534, "step": 31210 }, { "grad_norm": 0.8288459181785583, "learning_rate": 8.23540587364614e-05, "loss": 0.0632, "step": 31220 }, { "grad_norm": 0.7174407839775085, "learning_rate": 8.234145056461107e-05, "loss": 0.0499, "step": 31230 }, { "grad_norm": 0.5672682523727417, "learning_rate": 8.232883885594832e-05, "loss": 0.0481, "step": 31240 }, { "grad_norm": 0.6310124397277832, "learning_rate": 8.231622361185236e-05, "loss": 0.0577, "step": 31250 }, { "grad_norm": 0.6221425533294678, "learning_rate": 8.230360483370278e-05, "loss": 0.059, "step": 31260 }, { "grad_norm": 0.514126718044281, "learning_rate": 8.229098252287953e-05, "loss": 0.056, "step": 31270 }, { "grad_norm": 0.6071417331695557, "learning_rate": 8.2278356680763e-05, "loss": 0.055, "step": 31280 }, { "grad_norm": 0.7466873526573181, "learning_rate": 8.22657273087339e-05, "loss": 0.0557, "step": 31290 }, { "grad_norm": 0.6219695210456848, "learning_rate": 8.225309440817336e-05, "loss": 0.0752, "step": 31300 }, { "grad_norm": 0.661135196685791, "learning_rate": 8.224045798046293e-05, "loss": 0.0556, "step": 31310 }, { "grad_norm": 0.603130578994751, "learning_rate": 8.22278180269845e-05, "loss": 0.0644, "step": 31320 }, { "grad_norm": 0.5540425181388855, "learning_rate": 8.221517454912031e-05, "loss": 0.0442, "step": 31330 }, { "grad_norm": 1.008595585823059, "learning_rate": 8.220252754825308e-05, "loss": 0.0648, "step": 31340 }, { "grad_norm": 0.7762852907180786, "learning_rate": 8.218987702576586e-05, "loss": 0.0526, "step": 31350 }, { "grad_norm": 0.8167858123779297, "learning_rate": 8.217722298304207e-05, "loss": 0.0513, "step": 31360 }, { "grad_norm": 0.6463524103164673, "learning_rate": 8.216456542146557e-05, "loss": 0.0709, "step": 31370 }, { "grad_norm": 0.6784887909889221, "learning_rate": 8.215190434242055e-05, "loss": 0.0633, "step": 31380 }, { "grad_norm": 0.45406755805015564, "learning_rate": 8.213923974729161e-05, "loss": 0.0477, "step": 31390 }, { "grad_norm": 0.6137409806251526, "learning_rate": 8.212657163746373e-05, "loss": 0.0503, "step": 31400 }, { "grad_norm": 0.439651221036911, "learning_rate": 8.211390001432227e-05, "loss": 0.0556, "step": 31410 }, { "grad_norm": 0.6000168919563293, "learning_rate": 8.210122487925297e-05, "loss": 0.0548, "step": 31420 }, { "grad_norm": 0.5458800196647644, "learning_rate": 8.208854623364202e-05, "loss": 0.0496, "step": 31430 }, { "grad_norm": 0.6844146847724915, "learning_rate": 8.207586407887589e-05, "loss": 0.0616, "step": 31440 }, { "grad_norm": 1.1617285013198853, "learning_rate": 8.206317841634148e-05, "loss": 0.0674, "step": 31450 }, { "grad_norm": 0.6875506639480591, "learning_rate": 8.205048924742609e-05, "loss": 0.0634, "step": 31460 }, { "grad_norm": 0.4106326997280121, "learning_rate": 8.203779657351738e-05, "loss": 0.0679, "step": 31470 }, { "grad_norm": 0.524097204208374, "learning_rate": 8.20251003960034e-05, "loss": 0.05, "step": 31480 }, { "grad_norm": 0.5439702272415161, "learning_rate": 8.201240071627258e-05, "loss": 0.065, "step": 31490 }, { "grad_norm": 0.6326367259025574, "learning_rate": 8.199969753571377e-05, "loss": 0.0615, "step": 31500 }, { "grad_norm": 0.864160418510437, "learning_rate": 8.198699085571615e-05, "loss": 0.0518, "step": 31510 }, { "grad_norm": 0.7301937937736511, "learning_rate": 8.197428067766928e-05, "loss": 0.0528, "step": 31520 }, { "grad_norm": 0.5787374973297119, "learning_rate": 8.196156700296316e-05, "loss": 0.0488, "step": 31530 }, { "grad_norm": 0.6527935862541199, "learning_rate": 8.194884983298814e-05, "loss": 0.064, "step": 31540 }, { "grad_norm": 0.7120985388755798, "learning_rate": 8.193612916913491e-05, "loss": 0.0718, "step": 31550 }, { "grad_norm": 0.4340418875217438, "learning_rate": 8.192340501279463e-05, "loss": 0.0468, "step": 31560 }, { "grad_norm": 0.6621752381324768, "learning_rate": 8.191067736535876e-05, "loss": 0.0735, "step": 31570 }, { "grad_norm": 0.5967679023742676, "learning_rate": 8.18979462282192e-05, "loss": 0.0551, "step": 31580 }, { "grad_norm": 0.5848417282104492, "learning_rate": 8.188521160276819e-05, "loss": 0.0521, "step": 31590 }, { "grad_norm": 0.4717271625995636, "learning_rate": 8.187247349039837e-05, "loss": 0.0507, "step": 31600 }, { "grad_norm": 0.499920517206192, "learning_rate": 8.185973189250278e-05, "loss": 0.0558, "step": 31610 }, { "grad_norm": 0.6361565589904785, "learning_rate": 8.184698681047482e-05, "loss": 0.0554, "step": 31620 }, { "grad_norm": 1.2358442544937134, "learning_rate": 8.183423824570827e-05, "loss": 0.0556, "step": 31630 }, { "grad_norm": 0.7115122079849243, "learning_rate": 8.182148619959725e-05, "loss": 0.0552, "step": 31640 }, { "grad_norm": 0.46247681975364685, "learning_rate": 8.180873067353636e-05, "loss": 0.0511, "step": 31650 }, { "grad_norm": 0.8850142955780029, "learning_rate": 8.179597166892052e-05, "loss": 0.0468, "step": 31660 }, { "grad_norm": 0.7183693647384644, "learning_rate": 8.178320918714501e-05, "loss": 0.041, "step": 31670 }, { "grad_norm": 0.6362981796264648, "learning_rate": 8.177044322960554e-05, "loss": 0.0566, "step": 31680 }, { "grad_norm": 0.4978984594345093, "learning_rate": 8.175767379769816e-05, "loss": 0.052, "step": 31690 }, { "grad_norm": 0.8456287384033203, "learning_rate": 8.174490089281932e-05, "loss": 0.0601, "step": 31700 }, { "grad_norm": 0.6100579500198364, "learning_rate": 8.173212451636584e-05, "loss": 0.0597, "step": 31710 }, { "grad_norm": 0.6865125298500061, "learning_rate": 8.171934466973493e-05, "loss": 0.0565, "step": 31720 }, { "grad_norm": 0.7462185025215149, "learning_rate": 8.170656135432418e-05, "loss": 0.054, "step": 31730 }, { "grad_norm": 0.5639727711677551, "learning_rate": 8.169377457153155e-05, "loss": 0.0619, "step": 31740 }, { "grad_norm": 0.9038207530975342, "learning_rate": 8.168098432275539e-05, "loss": 0.077, "step": 31750 }, { "grad_norm": 0.6791274547576904, "learning_rate": 8.166819060939442e-05, "loss": 0.051, "step": 31760 }, { "grad_norm": 0.5331553816795349, "learning_rate": 8.165539343284772e-05, "loss": 0.0714, "step": 31770 }, { "grad_norm": 0.6689344644546509, "learning_rate": 8.16425927945148e-05, "loss": 0.0589, "step": 31780 }, { "grad_norm": 0.5980545282363892, "learning_rate": 8.162978869579551e-05, "loss": 0.0465, "step": 31790 }, { "grad_norm": 0.7747064828872681, "learning_rate": 8.161698113809007e-05, "loss": 0.0502, "step": 31800 }, { "grad_norm": 0.588306188583374, "learning_rate": 8.160417012279911e-05, "loss": 0.0561, "step": 31810 }, { "grad_norm": 0.7279757857322693, "learning_rate": 8.159135565132363e-05, "loss": 0.0469, "step": 31820 }, { "grad_norm": 0.728237509727478, "learning_rate": 8.157853772506498e-05, "loss": 0.0541, "step": 31830 }, { "grad_norm": 0.5371814966201782, "learning_rate": 8.156571634542494e-05, "loss": 0.0742, "step": 31840 }, { "grad_norm": 0.5826811790466309, "learning_rate": 8.15528915138056e-05, "loss": 0.051, "step": 31850 }, { "grad_norm": 0.5819937586784363, "learning_rate": 8.154006323160949e-05, "loss": 0.0634, "step": 31860 }, { "grad_norm": 0.7375007271766663, "learning_rate": 8.152723150023949e-05, "loss": 0.0692, "step": 31870 }, { "grad_norm": 0.8676941990852356, "learning_rate": 8.151439632109886e-05, "loss": 0.055, "step": 31880 }, { "grad_norm": 0.6672599911689758, "learning_rate": 8.150155769559122e-05, "loss": 0.0554, "step": 31890 }, { "grad_norm": 0.8549590110778809, "learning_rate": 8.148871562512058e-05, "loss": 0.0468, "step": 31900 }, { "grad_norm": 0.5492894053459167, "learning_rate": 8.147587011109136e-05, "loss": 0.0606, "step": 31910 }, { "grad_norm": 0.7295194268226624, "learning_rate": 8.14630211549083e-05, "loss": 0.0462, "step": 31920 }, { "grad_norm": 0.7249582409858704, "learning_rate": 8.145016875797655e-05, "loss": 0.0565, "step": 31930 }, { "grad_norm": 0.6448867321014404, "learning_rate": 8.143731292170164e-05, "loss": 0.0625, "step": 31940 }, { "grad_norm": 0.626338541507721, "learning_rate": 8.142445364748944e-05, "loss": 0.055, "step": 31950 }, { "grad_norm": 0.7814924716949463, "learning_rate": 8.141159093674624e-05, "loss": 0.0543, "step": 31960 }, { "grad_norm": 0.7556452751159668, "learning_rate": 8.139872479087869e-05, "loss": 0.0613, "step": 31970 }, { "grad_norm": 0.8462814688682556, "learning_rate": 8.13858552112938e-05, "loss": 0.0683, "step": 31980 }, { "grad_norm": 0.6133204698562622, "learning_rate": 8.137298219939895e-05, "loss": 0.0542, "step": 31990 }, { "grad_norm": 0.7497602701187134, "learning_rate": 8.136010575660196e-05, "loss": 0.0576, "step": 32000 }, { "grad_norm": 0.7544711828231812, "learning_rate": 8.134722588431095e-05, "loss": 0.0679, "step": 32010 }, { "grad_norm": 0.6066392660140991, "learning_rate": 8.133434258393444e-05, "loss": 0.0556, "step": 32020 }, { "grad_norm": 0.5704143643379211, "learning_rate": 8.132145585688134e-05, "loss": 0.057, "step": 32030 }, { "grad_norm": 0.5556191205978394, "learning_rate": 8.130856570456093e-05, "loss": 0.0512, "step": 32040 }, { "grad_norm": 0.6091664433479309, "learning_rate": 8.129567212838283e-05, "loss": 0.0662, "step": 32050 }, { "grad_norm": 0.8834907412528992, "learning_rate": 8.128277512975708e-05, "loss": 0.0497, "step": 32060 }, { "grad_norm": 0.9607659578323364, "learning_rate": 8.126987471009408e-05, "loss": 0.0583, "step": 32070 }, { "grad_norm": 0.4185747802257538, "learning_rate": 8.125697087080459e-05, "loss": 0.0455, "step": 32080 }, { "grad_norm": 0.4549284875392914, "learning_rate": 8.124406361329976e-05, "loss": 0.0424, "step": 32090 }, { "grad_norm": 0.7753152251243591, "learning_rate": 8.123115293899111e-05, "loss": 0.0586, "step": 32100 }, { "grad_norm": 0.7106359601020813, "learning_rate": 8.121823884929055e-05, "loss": 0.0469, "step": 32110 }, { "grad_norm": 0.6681147217750549, "learning_rate": 8.12053213456103e-05, "loss": 0.0544, "step": 32120 }, { "grad_norm": 0.5379635095596313, "learning_rate": 8.119240042936303e-05, "loss": 0.0522, "step": 32130 }, { "grad_norm": 0.8804593682289124, "learning_rate": 8.117947610196175e-05, "loss": 0.0524, "step": 32140 }, { "grad_norm": 0.7985565066337585, "learning_rate": 8.116654836481982e-05, "loss": 0.0535, "step": 32150 }, { "grad_norm": 0.6452004313468933, "learning_rate": 8.115361721935106e-05, "loss": 0.0578, "step": 32160 }, { "grad_norm": 0.5065868496894836, "learning_rate": 8.114068266696953e-05, "loss": 0.0586, "step": 32170 }, { "grad_norm": 0.9944193959236145, "learning_rate": 8.112774470908978e-05, "loss": 0.0482, "step": 32180 }, { "grad_norm": 0.7445652484893799, "learning_rate": 8.111480334712665e-05, "loss": 0.0501, "step": 32190 }, { "grad_norm": 0.6059375405311584, "learning_rate": 8.110185858249542e-05, "loss": 0.0514, "step": 32200 }, { "grad_norm": 0.4958922564983368, "learning_rate": 8.108891041661168e-05, "loss": 0.0608, "step": 32210 }, { "grad_norm": 0.8201939463615417, "learning_rate": 8.107595885089146e-05, "loss": 0.0493, "step": 32220 }, { "grad_norm": 0.8680042028427124, "learning_rate": 8.106300388675108e-05, "loss": 0.061, "step": 32230 }, { "grad_norm": 0.6825832724571228, "learning_rate": 8.105004552560731e-05, "loss": 0.0547, "step": 32240 }, { "grad_norm": 0.5283468961715698, "learning_rate": 8.103708376887724e-05, "loss": 0.0395, "step": 32250 }, { "grad_norm": 0.516544759273529, "learning_rate": 8.102411861797836e-05, "loss": 0.0613, "step": 32260 }, { "grad_norm": 0.6920638084411621, "learning_rate": 8.101115007432851e-05, "loss": 0.0501, "step": 32270 }, { "grad_norm": 0.6356926560401917, "learning_rate": 8.09981781393459e-05, "loss": 0.0589, "step": 32280 }, { "grad_norm": 0.7384229302406311, "learning_rate": 8.098520281444915e-05, "loss": 0.0712, "step": 32290 }, { "grad_norm": 0.894065797328949, "learning_rate": 8.09722241010572e-05, "loss": 0.0551, "step": 32300 }, { "grad_norm": 0.8356561064720154, "learning_rate": 8.095924200058939e-05, "loss": 0.0706, "step": 32310 }, { "grad_norm": 0.4696984887123108, "learning_rate": 8.094625651446541e-05, "loss": 0.0581, "step": 32320 }, { "grad_norm": 0.5360325574874878, "learning_rate": 8.093326764410536e-05, "loss": 0.0519, "step": 32330 }, { "grad_norm": 0.32181069254875183, "learning_rate": 8.092027539092966e-05, "loss": 0.0471, "step": 32340 }, { "grad_norm": 0.5807729959487915, "learning_rate": 8.090727975635913e-05, "loss": 0.0497, "step": 32350 }, { "grad_norm": 0.440530389547348, "learning_rate": 8.089428074181497e-05, "loss": 0.0469, "step": 32360 }, { "grad_norm": 0.4543727934360504, "learning_rate": 8.088127834871871e-05, "loss": 0.0594, "step": 32370 }, { "grad_norm": 0.6184815168380737, "learning_rate": 8.086827257849226e-05, "loss": 0.0528, "step": 32380 }, { "grad_norm": 0.5954731702804565, "learning_rate": 8.085526343255795e-05, "loss": 0.0563, "step": 32390 }, { "grad_norm": 0.989349901676178, "learning_rate": 8.084225091233842e-05, "loss": 0.0624, "step": 32400 }, { "grad_norm": 0.4989190995693207, "learning_rate": 8.082923501925668e-05, "loss": 0.0699, "step": 32410 }, { "grad_norm": 0.8007518649101257, "learning_rate": 8.081621575473617e-05, "loss": 0.053, "step": 32420 }, { "grad_norm": 0.6742886900901794, "learning_rate": 8.080319312020064e-05, "loss": 0.0531, "step": 32430 }, { "grad_norm": 0.4607202112674713, "learning_rate": 8.079016711707421e-05, "loss": 0.0506, "step": 32440 }, { "grad_norm": 0.4681565761566162, "learning_rate": 8.077713774678139e-05, "loss": 0.0572, "step": 32450 }, { "grad_norm": 0.6922878623008728, "learning_rate": 8.076410501074707e-05, "loss": 0.0702, "step": 32460 }, { "grad_norm": 0.6521017551422119, "learning_rate": 8.075106891039647e-05, "loss": 0.0725, "step": 32470 }, { "grad_norm": 0.593413233757019, "learning_rate": 8.073802944715523e-05, "loss": 0.06, "step": 32480 }, { "grad_norm": 0.6761391162872314, "learning_rate": 8.072498662244929e-05, "loss": 0.0566, "step": 32490 }, { "grad_norm": 0.7162612676620483, "learning_rate": 8.0711940437705e-05, "loss": 0.0592, "step": 32500 }, { "grad_norm": 0.5103864669799805, "learning_rate": 8.06988908943491e-05, "loss": 0.0425, "step": 32510 }, { "grad_norm": 0.8528596758842468, "learning_rate": 8.068583799380863e-05, "loss": 0.0609, "step": 32520 }, { "grad_norm": 0.6067702770233154, "learning_rate": 8.067278173751104e-05, "loss": 0.0521, "step": 32530 }, { "grad_norm": 0.7361693382263184, "learning_rate": 8.065972212688417e-05, "loss": 0.0611, "step": 32540 }, { "grad_norm": 0.5436917543411255, "learning_rate": 8.064665916335618e-05, "loss": 0.0706, "step": 32550 }, { "grad_norm": 0.464987576007843, "learning_rate": 8.063359284835564e-05, "loss": 0.0507, "step": 32560 }, { "grad_norm": 0.5131824612617493, "learning_rate": 8.062052318331142e-05, "loss": 0.0616, "step": 32570 }, { "grad_norm": 0.4690153896808624, "learning_rate": 8.060745016965283e-05, "loss": 0.0572, "step": 32580 }, { "grad_norm": 0.4848041832447052, "learning_rate": 8.059437380880952e-05, "loss": 0.0447, "step": 32590 }, { "grad_norm": 0.8345783352851868, "learning_rate": 8.058129410221146e-05, "loss": 0.0448, "step": 32600 }, { "grad_norm": 0.621216356754303, "learning_rate": 8.056821105128908e-05, "loss": 0.0524, "step": 32610 }, { "grad_norm": 0.5261797904968262, "learning_rate": 8.05551246574731e-05, "loss": 0.0492, "step": 32620 }, { "grad_norm": 0.6494306921958923, "learning_rate": 8.05420349221946e-05, "loss": 0.0517, "step": 32630 }, { "grad_norm": 0.5216360092163086, "learning_rate": 8.05289418468851e-05, "loss": 0.047, "step": 32640 }, { "grad_norm": 0.4543169438838959, "learning_rate": 8.051584543297642e-05, "loss": 0.0645, "step": 32650 }, { "grad_norm": 0.47398996353149414, "learning_rate": 8.050274568190074e-05, "loss": 0.0471, "step": 32660 }, { "grad_norm": 0.41480743885040283, "learning_rate": 8.048964259509067e-05, "loss": 0.0427, "step": 32670 }, { "grad_norm": 0.37810787558555603, "learning_rate": 8.047653617397914e-05, "loss": 0.0583, "step": 32680 }, { "grad_norm": 0.6993769407272339, "learning_rate": 8.046342641999941e-05, "loss": 0.0516, "step": 32690 }, { "grad_norm": 0.673272430896759, "learning_rate": 8.045031333458517e-05, "loss": 0.0479, "step": 32700 }, { "grad_norm": 0.5135578513145447, "learning_rate": 8.043719691917047e-05, "loss": 0.0618, "step": 32710 }, { "grad_norm": 0.5532351136207581, "learning_rate": 8.042407717518966e-05, "loss": 0.0599, "step": 32720 }, { "grad_norm": 0.5076521039009094, "learning_rate": 8.041095410407751e-05, "loss": 0.0555, "step": 32730 }, { "grad_norm": 0.4647543132305145, "learning_rate": 8.039782770726913e-05, "loss": 0.0725, "step": 32740 }, { "grad_norm": 0.7093020081520081, "learning_rate": 8.038469798620004e-05, "loss": 0.0672, "step": 32750 }, { "grad_norm": 0.4467257857322693, "learning_rate": 8.037156494230604e-05, "loss": 0.0447, "step": 32760 }, { "grad_norm": 0.6374499201774597, "learning_rate": 8.035842857702338e-05, "loss": 0.0667, "step": 32770 }, { "grad_norm": 0.45962896943092346, "learning_rate": 8.03452888917886e-05, "loss": 0.07, "step": 32780 }, { "grad_norm": 0.7281491160392761, "learning_rate": 8.033214588803866e-05, "loss": 0.0545, "step": 32790 }, { "grad_norm": 0.8671857714653015, "learning_rate": 8.031899956721083e-05, "loss": 0.0748, "step": 32800 }, { "grad_norm": 0.7587715983390808, "learning_rate": 8.030584993074282e-05, "loss": 0.0681, "step": 32810 }, { "grad_norm": 0.7433728575706482, "learning_rate": 8.02926969800726e-05, "loss": 0.0632, "step": 32820 }, { "grad_norm": 0.713270902633667, "learning_rate": 8.027954071663859e-05, "loss": 0.0448, "step": 32830 }, { "grad_norm": 0.6475414633750916, "learning_rate": 8.026638114187954e-05, "loss": 0.0584, "step": 32840 }, { "grad_norm": 0.6174284815788269, "learning_rate": 8.025321825723456e-05, "loss": 0.0661, "step": 32850 }, { "grad_norm": 0.8640555739402771, "learning_rate": 8.02400520641431e-05, "loss": 0.0658, "step": 32860 }, { "grad_norm": 0.45168912410736084, "learning_rate": 8.022688256404501e-05, "loss": 0.0586, "step": 32870 }, { "grad_norm": 0.6611167192459106, "learning_rate": 8.02137097583805e-05, "loss": 0.0603, "step": 32880 }, { "grad_norm": 0.5874803066253662, "learning_rate": 8.02005336485901e-05, "loss": 0.0515, "step": 32890 }, { "grad_norm": 0.8136375546455383, "learning_rate": 8.018735423611476e-05, "loss": 0.0591, "step": 32900 }, { "grad_norm": 0.6971474289894104, "learning_rate": 8.017417152239574e-05, "loss": 0.0551, "step": 32910 }, { "grad_norm": 0.6804592609405518, "learning_rate": 8.01609855088747e-05, "loss": 0.0649, "step": 32920 }, { "grad_norm": 0.6901130080223083, "learning_rate": 8.014779619699362e-05, "loss": 0.0443, "step": 32930 }, { "grad_norm": 0.4871353507041931, "learning_rate": 8.013460358819489e-05, "loss": 0.0432, "step": 32940 }, { "grad_norm": 0.35949692130088806, "learning_rate": 8.01214076839212e-05, "loss": 0.0441, "step": 32950 }, { "grad_norm": 0.706446647644043, "learning_rate": 8.010820848561565e-05, "loss": 0.0721, "step": 32960 }, { "grad_norm": 0.6336079239845276, "learning_rate": 8.009500599472171e-05, "loss": 0.0605, "step": 32970 }, { "grad_norm": 0.5135526657104492, "learning_rate": 8.008180021268314e-05, "loss": 0.0526, "step": 32980 }, { "grad_norm": 0.8831173777580261, "learning_rate": 8.006859114094414e-05, "loss": 0.039, "step": 32990 }, { "grad_norm": 0.6091956496238708, "learning_rate": 8.005537878094921e-05, "loss": 0.0532, "step": 33000 }, { "grad_norm": 0.6190789937973022, "learning_rate": 8.004216313414323e-05, "loss": 0.0532, "step": 33010 }, { "grad_norm": 0.6955604553222656, "learning_rate": 8.002894420197149e-05, "loss": 0.0488, "step": 33020 }, { "grad_norm": 0.46490082144737244, "learning_rate": 8.001572198587954e-05, "loss": 0.0594, "step": 33030 }, { "grad_norm": 0.47949251532554626, "learning_rate": 8.000249648731338e-05, "loss": 0.0559, "step": 33040 }, { "grad_norm": 0.8235985636711121, "learning_rate": 7.998926770771928e-05, "loss": 0.0469, "step": 33050 }, { "grad_norm": 0.5105790495872498, "learning_rate": 7.997603564854397e-05, "loss": 0.0569, "step": 33060 }, { "grad_norm": 0.5291262269020081, "learning_rate": 7.996280031123448e-05, "loss": 0.053, "step": 33070 }, { "grad_norm": 0.7746485471725464, "learning_rate": 7.994956169723818e-05, "loss": 0.0607, "step": 33080 }, { "grad_norm": 0.47163236141204834, "learning_rate": 7.993631980800285e-05, "loss": 0.0436, "step": 33090 }, { "grad_norm": 0.3917408585548401, "learning_rate": 7.992307464497659e-05, "loss": 0.0428, "step": 33100 }, { "grad_norm": 0.5102298259735107, "learning_rate": 7.990982620960787e-05, "loss": 0.0666, "step": 33110 }, { "grad_norm": 0.6161292195320129, "learning_rate": 7.989657450334554e-05, "loss": 0.0603, "step": 33120 }, { "grad_norm": 0.4616718292236328, "learning_rate": 7.988331952763877e-05, "loss": 0.0559, "step": 33130 }, { "grad_norm": 0.5486434102058411, "learning_rate": 7.987006128393709e-05, "loss": 0.0514, "step": 33140 }, { "grad_norm": 0.44665542244911194, "learning_rate": 7.985679977369043e-05, "loss": 0.0653, "step": 33150 }, { "grad_norm": 0.4228389263153076, "learning_rate": 7.984353499834902e-05, "loss": 0.0512, "step": 33160 }, { "grad_norm": 0.5068491697311401, "learning_rate": 7.983026695936351e-05, "loss": 0.0563, "step": 33170 }, { "grad_norm": 0.40407463908195496, "learning_rate": 7.981699565818486e-05, "loss": 0.0519, "step": 33180 }, { "grad_norm": 0.5293968915939331, "learning_rate": 7.980372109626437e-05, "loss": 0.0506, "step": 33190 }, { "grad_norm": 0.701977550983429, "learning_rate": 7.979044327505375e-05, "loss": 0.0679, "step": 33200 }, { "grad_norm": 0.9923244714736938, "learning_rate": 7.977716219600506e-05, "loss": 0.0507, "step": 33210 }, { "grad_norm": 0.5452499389648438, "learning_rate": 7.97638778605707e-05, "loss": 0.0435, "step": 33220 }, { "grad_norm": 0.693054735660553, "learning_rate": 7.975059027020338e-05, "loss": 0.0548, "step": 33230 }, { "grad_norm": 0.724010169506073, "learning_rate": 7.973729942635623e-05, "loss": 0.0552, "step": 33240 }, { "grad_norm": 0.4567502737045288, "learning_rate": 7.972400533048273e-05, "loss": 0.0365, "step": 33250 }, { "grad_norm": 0.4407713711261749, "learning_rate": 7.97107079840367e-05, "loss": 0.0547, "step": 33260 }, { "grad_norm": 0.2613040506839752, "learning_rate": 7.969740738847231e-05, "loss": 0.0547, "step": 33270 }, { "grad_norm": 0.44928526878356934, "learning_rate": 7.968410354524411e-05, "loss": 0.0582, "step": 33280 }, { "grad_norm": 0.5735220909118652, "learning_rate": 7.967079645580697e-05, "loss": 0.0563, "step": 33290 }, { "grad_norm": 0.6645241379737854, "learning_rate": 7.965748612161612e-05, "loss": 0.0543, "step": 33300 }, { "grad_norm": 0.5347357988357544, "learning_rate": 7.96441725441272e-05, "loss": 0.048, "step": 33310 }, { "grad_norm": 0.47803962230682373, "learning_rate": 7.963085572479614e-05, "loss": 0.0474, "step": 33320 }, { "grad_norm": 0.6154311895370483, "learning_rate": 7.961753566507924e-05, "loss": 0.0469, "step": 33330 }, { "grad_norm": 0.6000067591667175, "learning_rate": 7.960421236643316e-05, "loss": 0.0762, "step": 33340 }, { "grad_norm": 0.45833373069763184, "learning_rate": 7.959088583031496e-05, "loss": 0.0534, "step": 33350 }, { "grad_norm": 0.717689573764801, "learning_rate": 7.957755605818194e-05, "loss": 0.0518, "step": 33360 }, { "grad_norm": 0.44779032468795776, "learning_rate": 7.956422305149185e-05, "loss": 0.047, "step": 33370 }, { "grad_norm": 0.6933895945549011, "learning_rate": 7.95508868117028e-05, "loss": 0.0577, "step": 33380 }, { "grad_norm": 0.5556763410568237, "learning_rate": 7.953754734027318e-05, "loss": 0.0518, "step": 33390 }, { "grad_norm": 0.5398297905921936, "learning_rate": 7.952420463866182e-05, "loss": 0.0609, "step": 33400 }, { "grad_norm": 0.632603645324707, "learning_rate": 7.951085870832782e-05, "loss": 0.0624, "step": 33410 }, { "grad_norm": 0.3958321809768677, "learning_rate": 7.949750955073067e-05, "loss": 0.0481, "step": 33420 }, { "grad_norm": 0.795534074306488, "learning_rate": 7.948415716733022e-05, "loss": 0.064, "step": 33430 }, { "grad_norm": 0.7428968548774719, "learning_rate": 7.947080155958669e-05, "loss": 0.0563, "step": 33440 }, { "grad_norm": 0.5694753527641296, "learning_rate": 7.94574427289606e-05, "loss": 0.0392, "step": 33450 }, { "grad_norm": 0.4779495596885681, "learning_rate": 7.944408067691284e-05, "loss": 0.0763, "step": 33460 }, { "grad_norm": 0.41722574830055237, "learning_rate": 7.943071540490473e-05, "loss": 0.0466, "step": 33470 }, { "grad_norm": 0.5444866418838501, "learning_rate": 7.94173469143978e-05, "loss": 0.0508, "step": 33480 }, { "grad_norm": 0.6286319494247437, "learning_rate": 7.940397520685406e-05, "loss": 0.0435, "step": 33490 }, { "grad_norm": 0.5448167324066162, "learning_rate": 7.939060028373577e-05, "loss": 0.0654, "step": 33500 }, { "grad_norm": 0.7798380851745605, "learning_rate": 7.937722214650565e-05, "loss": 0.052, "step": 33510 }, { "grad_norm": 0.8805353045463562, "learning_rate": 7.936384079662666e-05, "loss": 0.0478, "step": 33520 }, { "grad_norm": 0.4121832251548767, "learning_rate": 7.93504562355622e-05, "loss": 0.0425, "step": 33530 }, { "grad_norm": 0.4413534700870514, "learning_rate": 7.933706846477599e-05, "loss": 0.0406, "step": 33540 }, { "grad_norm": 0.5614041090011597, "learning_rate": 7.932367748573206e-05, "loss": 0.0459, "step": 33550 }, { "grad_norm": 1.0889077186584473, "learning_rate": 7.931028329989485e-05, "loss": 0.0667, "step": 33560 }, { "grad_norm": 0.7044821977615356, "learning_rate": 7.929688590872913e-05, "loss": 0.0485, "step": 33570 }, { "grad_norm": 0.3825686275959015, "learning_rate": 7.928348531370003e-05, "loss": 0.0476, "step": 33580 }, { "grad_norm": 0.5396252870559692, "learning_rate": 7.927008151627297e-05, "loss": 0.0558, "step": 33590 }, { "grad_norm": 0.6127467751502991, "learning_rate": 7.925667451791383e-05, "loss": 0.0638, "step": 33600 }, { "grad_norm": 0.4240770637989044, "learning_rate": 7.924326432008874e-05, "loss": 0.0345, "step": 33610 }, { "grad_norm": 0.8075342774391174, "learning_rate": 7.922985092426422e-05, "loss": 0.064, "step": 33620 }, { "grad_norm": 0.4244768023490906, "learning_rate": 7.921643433190717e-05, "loss": 0.0496, "step": 33630 }, { "grad_norm": 0.4581005573272705, "learning_rate": 7.920301454448478e-05, "loss": 0.0413, "step": 33640 }, { "grad_norm": 0.5016934275627136, "learning_rate": 7.918959156346461e-05, "loss": 0.0486, "step": 33650 }, { "grad_norm": 0.5504421591758728, "learning_rate": 7.91761653903146e-05, "loss": 0.0494, "step": 33660 }, { "grad_norm": 0.45795831084251404, "learning_rate": 7.916273602650302e-05, "loss": 0.0392, "step": 33670 }, { "grad_norm": 0.8047927021980286, "learning_rate": 7.914930347349847e-05, "loss": 0.0442, "step": 33680 }, { "grad_norm": 0.717911958694458, "learning_rate": 7.913586773276992e-05, "loss": 0.0417, "step": 33690 }, { "grad_norm": 0.8969458341598511, "learning_rate": 7.912242880578667e-05, "loss": 0.05, "step": 33700 }, { "grad_norm": 0.4240056574344635, "learning_rate": 7.910898669401839e-05, "loss": 0.038, "step": 33710 }, { "grad_norm": 0.4971449375152588, "learning_rate": 7.909554139893511e-05, "loss": 0.0639, "step": 33720 }, { "grad_norm": 0.7381962537765503, "learning_rate": 7.908209292200715e-05, "loss": 0.046, "step": 33730 }, { "grad_norm": 0.6223669648170471, "learning_rate": 7.906864126470523e-05, "loss": 0.054, "step": 33740 }, { "grad_norm": 0.46285778284072876, "learning_rate": 7.905518642850041e-05, "loss": 0.0438, "step": 33750 }, { "grad_norm": 0.6315364837646484, "learning_rate": 7.904172841486409e-05, "loss": 0.0525, "step": 33760 }, { "grad_norm": 0.6992430090904236, "learning_rate": 7.902826722526801e-05, "loss": 0.0568, "step": 33770 }, { "grad_norm": 0.8864247798919678, "learning_rate": 7.901480286118427e-05, "loss": 0.0563, "step": 33780 }, { "grad_norm": 0.6273237466812134, "learning_rate": 7.900133532408531e-05, "loss": 0.05, "step": 33790 }, { "grad_norm": 0.6563764810562134, "learning_rate": 7.898786461544395e-05, "loss": 0.0546, "step": 33800 }, { "grad_norm": 0.5622920989990234, "learning_rate": 7.897439073673325e-05, "loss": 0.0486, "step": 33810 }, { "grad_norm": 0.4896616041660309, "learning_rate": 7.896091368942677e-05, "loss": 0.0538, "step": 33820 }, { "grad_norm": 0.7181726098060608, "learning_rate": 7.894743347499832e-05, "loss": 0.0702, "step": 33830 }, { "grad_norm": 0.764378547668457, "learning_rate": 7.893395009492203e-05, "loss": 0.0576, "step": 33840 }, { "grad_norm": 0.7286031246185303, "learning_rate": 7.892046355067248e-05, "loss": 0.0591, "step": 33850 }, { "grad_norm": 1.3925459384918213, "learning_rate": 7.890697384372451e-05, "loss": 0.0414, "step": 33860 }, { "grad_norm": 0.7832208871841431, "learning_rate": 7.889348097555336e-05, "loss": 0.0569, "step": 33870 }, { "grad_norm": 0.5743293762207031, "learning_rate": 7.887998494763455e-05, "loss": 0.044, "step": 33880 }, { "grad_norm": 1.065563678741455, "learning_rate": 7.886648576144404e-05, "loss": 0.0632, "step": 33890 }, { "grad_norm": 0.7071296572685242, "learning_rate": 7.885298341845802e-05, "loss": 0.06, "step": 33900 }, { "grad_norm": 0.5576117634773254, "learning_rate": 7.883947792015311e-05, "loss": 0.047, "step": 33910 }, { "grad_norm": 0.7696178555488586, "learning_rate": 7.882596926800628e-05, "loss": 0.0623, "step": 33920 }, { "grad_norm": 0.6811059713363647, "learning_rate": 7.881245746349477e-05, "loss": 0.0516, "step": 33930 }, { "grad_norm": 0.5976905822753906, "learning_rate": 7.879894250809623e-05, "loss": 0.0516, "step": 33940 }, { "grad_norm": 0.930791437625885, "learning_rate": 7.878542440328865e-05, "loss": 0.0558, "step": 33950 }, { "grad_norm": 0.46210840344429016, "learning_rate": 7.877190315055031e-05, "loss": 0.0494, "step": 33960 }, { "grad_norm": 0.6667014956474304, "learning_rate": 7.875837875135991e-05, "loss": 0.0499, "step": 33970 }, { "grad_norm": 0.5917971730232239, "learning_rate": 7.874485120719646e-05, "loss": 0.0628, "step": 33980 }, { "grad_norm": 0.5009516477584839, "learning_rate": 7.873132051953928e-05, "loss": 0.042, "step": 33990 }, { "grad_norm": 0.974977433681488, "learning_rate": 7.87177866898681e-05, "loss": 0.0635, "step": 34000 }, { "grad_norm": 0.5052922964096069, "learning_rate": 7.870424971966294e-05, "loss": 0.054, "step": 34010 }, { "grad_norm": 0.6959660649299622, "learning_rate": 7.869070961040419e-05, "loss": 0.0771, "step": 34020 }, { "grad_norm": 0.5003424882888794, "learning_rate": 7.867716636357257e-05, "loss": 0.0586, "step": 34030 }, { "grad_norm": 1.1501665115356445, "learning_rate": 7.866361998064915e-05, "loss": 0.07, "step": 34040 }, { "grad_norm": 0.5885269641876221, "learning_rate": 7.865007046311534e-05, "loss": 0.0472, "step": 34050 }, { "grad_norm": 0.4961516261100769, "learning_rate": 7.86365178124529e-05, "loss": 0.0422, "step": 34060 }, { "grad_norm": 0.5439490079879761, "learning_rate": 7.862296203014394e-05, "loss": 0.0546, "step": 34070 }, { "grad_norm": 0.5273604989051819, "learning_rate": 7.860940311767088e-05, "loss": 0.0429, "step": 34080 }, { "grad_norm": 0.4855992794036865, "learning_rate": 7.85958410765165e-05, "loss": 0.0492, "step": 34090 }, { "grad_norm": 0.6094950437545776, "learning_rate": 7.858227590816394e-05, "loss": 0.0497, "step": 34100 }, { "grad_norm": 0.73515784740448, "learning_rate": 7.856870761409664e-05, "loss": 0.0433, "step": 34110 }, { "grad_norm": 1.320656418800354, "learning_rate": 7.855513619579846e-05, "loss": 0.0677, "step": 34120 }, { "grad_norm": 0.8745254874229431, "learning_rate": 7.85415616547535e-05, "loss": 0.0619, "step": 34130 }, { "grad_norm": 0.41252025961875916, "learning_rate": 7.852798399244627e-05, "loss": 0.0476, "step": 34140 }, { "grad_norm": 0.4986049234867096, "learning_rate": 7.851440321036161e-05, "loss": 0.0499, "step": 34150 }, { "grad_norm": 0.6066179871559143, "learning_rate": 7.850081930998468e-05, "loss": 0.0391, "step": 34160 }, { "grad_norm": 0.44202056527137756, "learning_rate": 7.8487232292801e-05, "loss": 0.0397, "step": 34170 }, { "grad_norm": 0.5468878149986267, "learning_rate": 7.847364216029642e-05, "loss": 0.0522, "step": 34180 }, { "grad_norm": 0.562357485294342, "learning_rate": 7.846004891395716e-05, "loss": 0.0435, "step": 34190 }, { "grad_norm": 0.5705817937850952, "learning_rate": 7.844645255526972e-05, "loss": 0.0452, "step": 34200 }, { "grad_norm": 0.5150830745697021, "learning_rate": 7.843285308572101e-05, "loss": 0.0499, "step": 34210 }, { "grad_norm": 0.3964110314846039, "learning_rate": 7.841925050679823e-05, "loss": 0.0576, "step": 34220 }, { "grad_norm": 0.7326869964599609, "learning_rate": 7.840564481998895e-05, "loss": 0.0514, "step": 34230 }, { "grad_norm": 0.6501555442810059, "learning_rate": 7.839203602678105e-05, "loss": 0.0473, "step": 34240 }, { "grad_norm": 0.5131869912147522, "learning_rate": 7.837842412866279e-05, "loss": 0.0413, "step": 34250 }, { "grad_norm": 0.6684926748275757, "learning_rate": 7.836480912712272e-05, "loss": 0.045, "step": 34260 }, { "grad_norm": 0.8456944227218628, "learning_rate": 7.835119102364976e-05, "loss": 0.0477, "step": 34270 }, { "grad_norm": 0.39425477385520935, "learning_rate": 7.83375698197332e-05, "loss": 0.0379, "step": 34280 }, { "grad_norm": 0.591113269329071, "learning_rate": 7.83239455168626e-05, "loss": 0.0621, "step": 34290 }, { "grad_norm": 0.6503347754478455, "learning_rate": 7.83103181165279e-05, "loss": 0.0531, "step": 34300 }, { "grad_norm": 0.3996175229549408, "learning_rate": 7.829668762021937e-05, "loss": 0.0428, "step": 34310 }, { "grad_norm": 0.4224504232406616, "learning_rate": 7.828305402942764e-05, "loss": 0.0619, "step": 34320 }, { "grad_norm": 0.6993439793586731, "learning_rate": 7.826941734564363e-05, "loss": 0.0807, "step": 34330 }, { "grad_norm": 0.5428266525268555, "learning_rate": 7.825577757035865e-05, "loss": 0.0454, "step": 34340 }, { "grad_norm": 0.5303205251693726, "learning_rate": 7.824213470506431e-05, "loss": 0.0464, "step": 34350 }, { "grad_norm": 0.6656293272972107, "learning_rate": 7.822848875125257e-05, "loss": 0.0466, "step": 34360 }, { "grad_norm": 0.7545396089553833, "learning_rate": 7.821483971041576e-05, "loss": 0.0488, "step": 34370 }, { "grad_norm": 0.67264324426651, "learning_rate": 7.820118758404649e-05, "loss": 0.0531, "step": 34380 }, { "grad_norm": 0.414943665266037, "learning_rate": 7.818753237363776e-05, "loss": 0.052, "step": 34390 }, { "grad_norm": 0.5979828238487244, "learning_rate": 7.817387408068286e-05, "loss": 0.0663, "step": 34400 }, { "grad_norm": 1.0235629081726074, "learning_rate": 7.816021270667544e-05, "loss": 0.0525, "step": 34410 }, { "grad_norm": 0.6089292168617249, "learning_rate": 7.81465482531095e-05, "loss": 0.0493, "step": 34420 }, { "grad_norm": 0.4845200479030609, "learning_rate": 7.813288072147938e-05, "loss": 0.0613, "step": 34430 }, { "grad_norm": 0.8538415431976318, "learning_rate": 7.811921011327972e-05, "loss": 0.0472, "step": 34440 }, { "grad_norm": 0.4928704500198364, "learning_rate": 7.810553643000549e-05, "loss": 0.0506, "step": 34450 }, { "grad_norm": 0.475894570350647, "learning_rate": 7.809185967315206e-05, "loss": 0.0413, "step": 34460 }, { "grad_norm": 0.5827860832214355, "learning_rate": 7.80781798442151e-05, "loss": 0.0456, "step": 34470 }, { "grad_norm": 0.5515769124031067, "learning_rate": 7.806449694469061e-05, "loss": 0.055, "step": 34480 }, { "grad_norm": 0.4079605042934418, "learning_rate": 7.805081097607492e-05, "loss": 0.0735, "step": 34490 }, { "grad_norm": 0.7780126929283142, "learning_rate": 7.803712193986474e-05, "loss": 0.0494, "step": 34500 }, { "grad_norm": 0.7549233436584473, "learning_rate": 7.802342983755702e-05, "loss": 0.0494, "step": 34510 }, { "grad_norm": 0.5722954869270325, "learning_rate": 7.800973467064918e-05, "loss": 0.0492, "step": 34520 }, { "grad_norm": 0.6036261320114136, "learning_rate": 7.799603644063884e-05, "loss": 0.0452, "step": 34530 }, { "grad_norm": 0.6044749617576599, "learning_rate": 7.798233514902405e-05, "loss": 0.0584, "step": 34540 }, { "grad_norm": 0.7752850651741028, "learning_rate": 7.796863079730318e-05, "loss": 0.0587, "step": 34550 }, { "grad_norm": 0.5736677646636963, "learning_rate": 7.795492338697488e-05, "loss": 0.057, "step": 34560 }, { "grad_norm": 0.5446323752403259, "learning_rate": 7.794121291953819e-05, "loss": 0.0448, "step": 34570 }, { "grad_norm": 0.5303813219070435, "learning_rate": 7.792749939649246e-05, "loss": 0.067, "step": 34580 }, { "grad_norm": 0.957830011844635, "learning_rate": 7.79137828193374e-05, "loss": 0.0554, "step": 34590 }, { "grad_norm": 0.6070699095726013, "learning_rate": 7.790006318957301e-05, "loss": 0.0683, "step": 34600 }, { "grad_norm": 0.5063735842704773, "learning_rate": 7.788634050869965e-05, "loss": 0.057, "step": 34610 }, { "grad_norm": 0.3908594846725464, "learning_rate": 7.787261477821803e-05, "loss": 0.0324, "step": 34620 }, { "grad_norm": 0.3943997621536255, "learning_rate": 7.785888599962916e-05, "loss": 0.0333, "step": 34630 }, { "grad_norm": 0.37123870849609375, "learning_rate": 7.784515417443439e-05, "loss": 0.0487, "step": 34640 }, { "grad_norm": 0.7837179899215698, "learning_rate": 7.783141930413545e-05, "loss": 0.0546, "step": 34650 }, { "grad_norm": 0.7984529733657837, "learning_rate": 7.78176813902343e-05, "loss": 0.0587, "step": 34660 }, { "grad_norm": 0.7395409345626831, "learning_rate": 7.780394043423336e-05, "loss": 0.0627, "step": 34670 }, { "grad_norm": 0.608328640460968, "learning_rate": 7.77901964376353e-05, "loss": 0.0483, "step": 34680 }, { "grad_norm": 0.8465337157249451, "learning_rate": 7.777644940194316e-05, "loss": 0.0481, "step": 34690 }, { "grad_norm": 0.5764308571815491, "learning_rate": 7.776269932866023e-05, "loss": 0.0461, "step": 34700 }, { "grad_norm": 0.6712399125099182, "learning_rate": 7.774894621929026e-05, "loss": 0.0585, "step": 34710 }, { "grad_norm": 0.7594621777534485, "learning_rate": 7.773519007533725e-05, "loss": 0.0619, "step": 34720 }, { "grad_norm": 0.5137913227081299, "learning_rate": 7.772143089830556e-05, "loss": 0.041, "step": 34730 }, { "grad_norm": 0.6290614604949951, "learning_rate": 7.770766868969985e-05, "loss": 0.0594, "step": 34740 }, { "grad_norm": 0.46263769268989563, "learning_rate": 7.769390345102518e-05, "loss": 0.0399, "step": 34750 }, { "grad_norm": 0.6381781697273254, "learning_rate": 7.768013518378683e-05, "loss": 0.0438, "step": 34760 }, { "grad_norm": 0.9351321458816528, "learning_rate": 7.766636388949053e-05, "loss": 0.0588, "step": 34770 }, { "grad_norm": 0.26952457427978516, "learning_rate": 7.765258956964229e-05, "loss": 0.0362, "step": 34780 }, { "grad_norm": 0.5264546275138855, "learning_rate": 7.76388122257484e-05, "loss": 0.0592, "step": 34790 }, { "grad_norm": 0.7053637504577637, "learning_rate": 7.762503185931558e-05, "loss": 0.0634, "step": 34800 }, { "grad_norm": 0.807819664478302, "learning_rate": 7.76112484718508e-05, "loss": 0.0442, "step": 34810 }, { "grad_norm": 0.3997638523578644, "learning_rate": 7.75974620648614e-05, "loss": 0.0457, "step": 34820 }, { "grad_norm": 0.4984797239303589, "learning_rate": 7.758367263985503e-05, "loss": 0.0483, "step": 34830 }, { "grad_norm": 0.5521815419197083, "learning_rate": 7.75698801983397e-05, "loss": 0.0569, "step": 34840 }, { "grad_norm": 0.7328259944915771, "learning_rate": 7.755608474182372e-05, "loss": 0.051, "step": 34850 }, { "grad_norm": 0.47788339853286743, "learning_rate": 7.754228627181574e-05, "loss": 0.0449, "step": 34860 }, { "grad_norm": 0.47704097628593445, "learning_rate": 7.752848478982476e-05, "loss": 0.0467, "step": 34870 }, { "grad_norm": 0.5588858127593994, "learning_rate": 7.751468029736006e-05, "loss": 0.0685, "step": 34880 }, { "grad_norm": 0.6893874406814575, "learning_rate": 7.750087279593129e-05, "loss": 0.0632, "step": 34890 }, { "grad_norm": 0.576888382434845, "learning_rate": 7.748706228704843e-05, "loss": 0.0581, "step": 34900 }, { "grad_norm": 0.8542589545249939, "learning_rate": 7.747324877222176e-05, "loss": 0.0415, "step": 34910 }, { "grad_norm": 0.49753060936927795, "learning_rate": 7.745943225296188e-05, "loss": 0.0544, "step": 34920 }, { "grad_norm": 0.5057033896446228, "learning_rate": 7.744561273077981e-05, "loss": 0.0513, "step": 34930 }, { "grad_norm": 0.5965380668640137, "learning_rate": 7.743179020718678e-05, "loss": 0.0465, "step": 34940 }, { "grad_norm": 0.6275066137313843, "learning_rate": 7.741796468369443e-05, "loss": 0.0476, "step": 34950 }, { "grad_norm": 0.44629719853401184, "learning_rate": 7.740413616181466e-05, "loss": 0.0489, "step": 34960 }, { "grad_norm": 0.4571866989135742, "learning_rate": 7.739030464305978e-05, "loss": 0.0494, "step": 34970 }, { "grad_norm": 0.6151777505874634, "learning_rate": 7.737647012894235e-05, "loss": 0.049, "step": 34980 }, { "grad_norm": 0.6477625966072083, "learning_rate": 7.736263262097532e-05, "loss": 0.0447, "step": 34990 }, { "grad_norm": 0.4227541387081146, "learning_rate": 7.734879212067192e-05, "loss": 0.0434, "step": 35000 }, { "grad_norm": 0.7556278109550476, "learning_rate": 7.733494862954573e-05, "loss": 0.0631, "step": 35010 }, { "grad_norm": 0.4002894461154938, "learning_rate": 7.732110214911066e-05, "loss": 0.0603, "step": 35020 }, { "grad_norm": 0.47550633549690247, "learning_rate": 7.730725268088094e-05, "loss": 0.0474, "step": 35030 }, { "grad_norm": 0.634392261505127, "learning_rate": 7.729340022637112e-05, "loss": 0.0601, "step": 35040 }, { "grad_norm": 0.7046347856521606, "learning_rate": 7.727954478709607e-05, "loss": 0.0719, "step": 35050 }, { "grad_norm": 0.524796187877655, "learning_rate": 7.726568636457103e-05, "loss": 0.0476, "step": 35060 }, { "grad_norm": 0.6124864220619202, "learning_rate": 7.725182496031153e-05, "loss": 0.0567, "step": 35070 }, { "grad_norm": 0.5092042684555054, "learning_rate": 7.723796057583341e-05, "loss": 0.0517, "step": 35080 }, { "grad_norm": 0.4403247535228729, "learning_rate": 7.722409321265287e-05, "loss": 0.0524, "step": 35090 }, { "grad_norm": 0.5950377583503723, "learning_rate": 7.721022287228645e-05, "loss": 0.0608, "step": 35100 }, { "grad_norm": 0.49725815653800964, "learning_rate": 7.719634955625094e-05, "loss": 0.0617, "step": 35110 }, { "grad_norm": 0.6465283036231995, "learning_rate": 7.718247326606355e-05, "loss": 0.0665, "step": 35120 }, { "grad_norm": 0.5200223922729492, "learning_rate": 7.716859400324175e-05, "loss": 0.0479, "step": 35130 }, { "grad_norm": 0.668935239315033, "learning_rate": 7.715471176930335e-05, "loss": 0.0676, "step": 35140 }, { "grad_norm": 0.558784544467926, "learning_rate": 7.714082656576651e-05, "loss": 0.0537, "step": 35150 }, { "grad_norm": 0.9365920424461365, "learning_rate": 7.712693839414968e-05, "loss": 0.0524, "step": 35160 }, { "grad_norm": 0.6632643938064575, "learning_rate": 7.711304725597164e-05, "loss": 0.0607, "step": 35170 }, { "grad_norm": 0.37818512320518494, "learning_rate": 7.709915315275151e-05, "loss": 0.07, "step": 35180 }, { "grad_norm": 0.51436847448349, "learning_rate": 7.708525608600876e-05, "loss": 0.0375, "step": 35190 }, { "grad_norm": 0.8565710186958313, "learning_rate": 7.707135605726311e-05, "loss": 0.0818, "step": 35200 }, { "grad_norm": 1.2652043104171753, "learning_rate": 7.705745306803466e-05, "loss": 0.0778, "step": 35210 }, { "grad_norm": 0.6782556772232056, "learning_rate": 7.704354711984383e-05, "loss": 0.0592, "step": 35220 }, { "grad_norm": 0.6320639252662659, "learning_rate": 7.702963821421133e-05, "loss": 0.0541, "step": 35230 }, { "grad_norm": 0.6994426846504211, "learning_rate": 7.701572635265826e-05, "loss": 0.0558, "step": 35240 }, { "grad_norm": 0.4312247931957245, "learning_rate": 7.700181153670596e-05, "loss": 0.053, "step": 35250 }, { "grad_norm": 0.5765292644500732, "learning_rate": 7.698789376787616e-05, "loss": 0.0519, "step": 35260 }, { "grad_norm": 0.5019671320915222, "learning_rate": 7.697397304769085e-05, "loss": 0.052, "step": 35270 }, { "grad_norm": 0.6865214109420776, "learning_rate": 7.696004937767241e-05, "loss": 0.0553, "step": 35280 }, { "grad_norm": 0.6692317724227905, "learning_rate": 7.694612275934352e-05, "loss": 0.0551, "step": 35290 }, { "grad_norm": 0.47118043899536133, "learning_rate": 7.693219319422714e-05, "loss": 0.0427, "step": 35300 }, { "grad_norm": 0.6874503493309021, "learning_rate": 7.69182606838466e-05, "loss": 0.0457, "step": 35310 }, { "grad_norm": 0.6593464612960815, "learning_rate": 7.690432522972558e-05, "loss": 0.0743, "step": 35320 }, { "grad_norm": 0.46261167526245117, "learning_rate": 7.689038683338796e-05, "loss": 0.0542, "step": 35330 }, { "grad_norm": 0.41695237159729004, "learning_rate": 7.687644549635808e-05, "loss": 0.0514, "step": 35340 }, { "grad_norm": 0.4448486864566803, "learning_rate": 7.686250122016053e-05, "loss": 0.0596, "step": 35350 }, { "grad_norm": 0.7632777690887451, "learning_rate": 7.684855400632022e-05, "loss": 0.0441, "step": 35360 }, { "grad_norm": 0.46865716576576233, "learning_rate": 7.683460385636243e-05, "loss": 0.0494, "step": 35370 }, { "grad_norm": 1.8537484407424927, "learning_rate": 7.68206507718127e-05, "loss": 0.0451, "step": 35380 }, { "grad_norm": 0.3733409643173218, "learning_rate": 7.680669475419693e-05, "loss": 0.0545, "step": 35390 }, { "grad_norm": 0.6357145309448242, "learning_rate": 7.679273580504132e-05, "loss": 0.0436, "step": 35400 }, { "grad_norm": 0.6528716087341309, "learning_rate": 7.67787739258724e-05, "loss": 0.0515, "step": 35410 }, { "grad_norm": 0.8632883429527283, "learning_rate": 7.676480911821705e-05, "loss": 0.0582, "step": 35420 }, { "grad_norm": 0.42117103934288025, "learning_rate": 7.675084138360238e-05, "loss": 0.0508, "step": 35430 }, { "grad_norm": 0.5239145755767822, "learning_rate": 7.673687072355592e-05, "loss": 0.0545, "step": 35440 }, { "grad_norm": 0.4802842438220978, "learning_rate": 7.67228971396055e-05, "loss": 0.0544, "step": 35450 }, { "grad_norm": 0.529340386390686, "learning_rate": 7.670892063327922e-05, "loss": 0.043, "step": 35460 }, { "grad_norm": 0.5753985643386841, "learning_rate": 7.669494120610552e-05, "loss": 0.0503, "step": 35470 }, { "grad_norm": 0.30972278118133545, "learning_rate": 7.66809588596132e-05, "loss": 0.0687, "step": 35480 }, { "grad_norm": 0.6562963128089905, "learning_rate": 7.666697359533132e-05, "loss": 0.0579, "step": 35490 }, { "grad_norm": 0.6920831799507141, "learning_rate": 7.665298541478932e-05, "loss": 0.0797, "step": 35500 }, { "grad_norm": 0.6789314150810242, "learning_rate": 7.66389943195169e-05, "loss": 0.0553, "step": 35510 }, { "grad_norm": 0.7382079362869263, "learning_rate": 7.66250003110441e-05, "loss": 0.0728, "step": 35520 }, { "grad_norm": 0.4493366777896881, "learning_rate": 7.66110033909013e-05, "loss": 0.0375, "step": 35530 }, { "grad_norm": 0.531278133392334, "learning_rate": 7.659700356061918e-05, "loss": 0.0416, "step": 35540 }, { "grad_norm": 0.5491594672203064, "learning_rate": 7.658300082172875e-05, "loss": 0.0492, "step": 35550 }, { "grad_norm": 0.44038012623786926, "learning_rate": 7.65689951757613e-05, "loss": 0.0682, "step": 35560 }, { "grad_norm": 0.5785505175590515, "learning_rate": 7.655498662424849e-05, "loss": 0.0322, "step": 35570 }, { "grad_norm": 0.5331071615219116, "learning_rate": 7.654097516872227e-05, "loss": 0.0392, "step": 35580 }, { "grad_norm": 0.48743635416030884, "learning_rate": 7.65269608107149e-05, "loss": 0.0568, "step": 35590 }, { "grad_norm": 0.4524826109409332, "learning_rate": 7.6512943551759e-05, "loss": 0.0437, "step": 35600 }, { "grad_norm": 0.65291827917099, "learning_rate": 7.649892339338743e-05, "loss": 0.0503, "step": 35610 }, { "grad_norm": 0.48731741309165955, "learning_rate": 7.648490033713344e-05, "loss": 0.0694, "step": 35620 }, { "grad_norm": 0.3516795039176941, "learning_rate": 7.647087438453058e-05, "loss": 0.0425, "step": 35630 }, { "grad_norm": 0.6208215951919556, "learning_rate": 7.64568455371127e-05, "loss": 0.0502, "step": 35640 }, { "grad_norm": 0.631807267665863, "learning_rate": 7.644281379641396e-05, "loss": 0.0541, "step": 35650 }, { "grad_norm": 0.5848708152770996, "learning_rate": 7.642877916396887e-05, "loss": 0.0465, "step": 35660 }, { "grad_norm": 0.705730676651001, "learning_rate": 7.641474164131221e-05, "loss": 0.0513, "step": 35670 }, { "grad_norm": 0.5207118988037109, "learning_rate": 7.640070122997913e-05, "loss": 0.0504, "step": 35680 }, { "grad_norm": 0.5205183029174805, "learning_rate": 7.638665793150505e-05, "loss": 0.0384, "step": 35690 }, { "grad_norm": 0.6241444945335388, "learning_rate": 7.637261174742574e-05, "loss": 0.0654, "step": 35700 }, { "grad_norm": 0.4300824701786041, "learning_rate": 7.635856267927725e-05, "loss": 0.046, "step": 35710 }, { "grad_norm": 0.5331668257713318, "learning_rate": 7.634451072859597e-05, "loss": 0.0471, "step": 35720 }, { "grad_norm": 0.5550295114517212, "learning_rate": 7.633045589691863e-05, "loss": 0.0464, "step": 35730 }, { "grad_norm": 0.575972855091095, "learning_rate": 7.63163981857822e-05, "loss": 0.0507, "step": 35740 }, { "grad_norm": 0.6481486558914185, "learning_rate": 7.630233759672403e-05, "loss": 0.0419, "step": 35750 }, { "grad_norm": 0.7481361627578735, "learning_rate": 7.628827413128177e-05, "loss": 0.047, "step": 35760 }, { "grad_norm": 0.5550068616867065, "learning_rate": 7.627420779099338e-05, "loss": 0.038, "step": 35770 }, { "grad_norm": 0.6617854237556458, "learning_rate": 7.626013857739711e-05, "loss": 0.0452, "step": 35780 }, { "grad_norm": 0.5909620523452759, "learning_rate": 7.624606649203158e-05, "loss": 0.045, "step": 35790 }, { "grad_norm": 0.7428339719772339, "learning_rate": 7.623199153643569e-05, "loss": 0.0415, "step": 35800 }, { "grad_norm": 0.7669971585273743, "learning_rate": 7.621791371214863e-05, "loss": 0.0434, "step": 35810 }, { "grad_norm": 0.6659590005874634, "learning_rate": 7.620383302070995e-05, "loss": 0.0532, "step": 35820 }, { "grad_norm": 0.6740964651107788, "learning_rate": 7.61897494636595e-05, "loss": 0.0506, "step": 35830 }, { "grad_norm": 0.8750871419906616, "learning_rate": 7.617566304253739e-05, "loss": 0.0452, "step": 35840 }, { "grad_norm": 0.6661349534988403, "learning_rate": 7.616157375888416e-05, "loss": 0.0473, "step": 35850 }, { "grad_norm": 0.4589477479457855, "learning_rate": 7.614748161424053e-05, "loss": 0.0441, "step": 35860 }, { "grad_norm": 0.8423811197280884, "learning_rate": 7.613338661014763e-05, "loss": 0.0533, "step": 35870 }, { "grad_norm": 0.6141271591186523, "learning_rate": 7.611928874814686e-05, "loss": 0.0428, "step": 35880 }, { "grad_norm": 0.5252729058265686, "learning_rate": 7.610518802977996e-05, "loss": 0.0785, "step": 35890 }, { "grad_norm": 0.783600389957428, "learning_rate": 7.609108445658893e-05, "loss": 0.043, "step": 35900 }, { "grad_norm": 0.39787864685058594, "learning_rate": 7.607697803011612e-05, "loss": 0.0487, "step": 35910 }, { "grad_norm": 0.5136784911155701, "learning_rate": 7.606286875190421e-05, "loss": 0.0532, "step": 35920 }, { "grad_norm": 0.7596614956855774, "learning_rate": 7.604875662349617e-05, "loss": 0.0441, "step": 35930 }, { "grad_norm": 0.46102190017700195, "learning_rate": 7.603464164643525e-05, "loss": 0.0311, "step": 35940 }, { "grad_norm": 0.5779945850372314, "learning_rate": 7.602052382226507e-05, "loss": 0.0693, "step": 35950 }, { "grad_norm": 0.6012935042381287, "learning_rate": 7.600640315252954e-05, "loss": 0.0643, "step": 35960 }, { "grad_norm": 1.5177571773529053, "learning_rate": 7.599227963877284e-05, "loss": 0.0525, "step": 35970 }, { "grad_norm": 0.6142071485519409, "learning_rate": 7.597815328253954e-05, "loss": 0.041, "step": 35980 }, { "grad_norm": 0.45384353399276733, "learning_rate": 7.596402408537443e-05, "loss": 0.0664, "step": 35990 }, { "grad_norm": 1.1276699304580688, "learning_rate": 7.59498920488227e-05, "loss": 0.0476, "step": 36000 }, { "grad_norm": 0.4281141459941864, "learning_rate": 7.593575717442979e-05, "loss": 0.0432, "step": 36010 }, { "grad_norm": 0.49885672330856323, "learning_rate": 7.592161946374147e-05, "loss": 0.0717, "step": 36020 }, { "grad_norm": 0.5182499885559082, "learning_rate": 7.590747891830381e-05, "loss": 0.0609, "step": 36030 }, { "grad_norm": 0.42672696709632874, "learning_rate": 7.58933355396632e-05, "loss": 0.0426, "step": 36040 }, { "grad_norm": 0.40177348256111145, "learning_rate": 7.587918932936636e-05, "loss": 0.0689, "step": 36050 }, { "grad_norm": 0.753188967704773, "learning_rate": 7.586504028896028e-05, "loss": 0.0546, "step": 36060 }, { "grad_norm": 0.4234880805015564, "learning_rate": 7.585088841999228e-05, "loss": 0.06, "step": 36070 }, { "grad_norm": 0.6545063257217407, "learning_rate": 7.583673372400999e-05, "loss": 0.0491, "step": 36080 }, { "grad_norm": 0.8833786249160767, "learning_rate": 7.582257620256134e-05, "loss": 0.0444, "step": 36090 }, { "grad_norm": 0.6520196795463562, "learning_rate": 7.580841585719458e-05, "loss": 0.0596, "step": 36100 }, { "grad_norm": 0.6122443675994873, "learning_rate": 7.579425268945825e-05, "loss": 0.0533, "step": 36110 }, { "grad_norm": 1.0380134582519531, "learning_rate": 7.578008670090127e-05, "loss": 0.0497, "step": 36120 }, { "grad_norm": 0.6421599984169006, "learning_rate": 7.576591789307272e-05, "loss": 0.0631, "step": 36130 }, { "grad_norm": 0.4678109288215637, "learning_rate": 7.575174626752216e-05, "loss": 0.05, "step": 36140 }, { "grad_norm": 0.7364804148674011, "learning_rate": 7.573757182579934e-05, "loss": 0.0551, "step": 36150 }, { "grad_norm": 0.5555671453475952, "learning_rate": 7.572339456945435e-05, "loss": 0.0415, "step": 36160 }, { "grad_norm": 0.37062206864356995, "learning_rate": 7.570921450003762e-05, "loss": 0.0462, "step": 36170 }, { "grad_norm": 0.4585171341896057, "learning_rate": 7.569503161909983e-05, "loss": 0.053, "step": 36180 }, { "grad_norm": 0.5026857852935791, "learning_rate": 7.568084592819202e-05, "loss": 0.0475, "step": 36190 }, { "grad_norm": 0.6343719959259033, "learning_rate": 7.566665742886551e-05, "loss": 0.0596, "step": 36200 }, { "grad_norm": 0.3653203547000885, "learning_rate": 7.565246612267194e-05, "loss": 0.0368, "step": 36210 }, { "grad_norm": 0.5107782483100891, "learning_rate": 7.563827201116325e-05, "loss": 0.0495, "step": 36220 }, { "grad_norm": 0.5868180990219116, "learning_rate": 7.562407509589164e-05, "loss": 0.0648, "step": 36230 }, { "grad_norm": 0.7330650091171265, "learning_rate": 7.560987537840973e-05, "loss": 0.0579, "step": 36240 }, { "grad_norm": 0.28582072257995605, "learning_rate": 7.559567286027036e-05, "loss": 0.0475, "step": 36250 }, { "grad_norm": 0.7589055299758911, "learning_rate": 7.558146754302668e-05, "loss": 0.0519, "step": 36260 }, { "grad_norm": 0.6627779603004456, "learning_rate": 7.556725942823217e-05, "loss": 0.0508, "step": 36270 }, { "grad_norm": 0.37073084712028503, "learning_rate": 7.555304851744061e-05, "loss": 0.0502, "step": 36280 }, { "grad_norm": 0.8883762359619141, "learning_rate": 7.553883481220608e-05, "loss": 0.0482, "step": 36290 }, { "grad_norm": 0.6018204092979431, "learning_rate": 7.552461831408298e-05, "loss": 0.0528, "step": 36300 }, { "grad_norm": 0.5850959420204163, "learning_rate": 7.551039902462599e-05, "loss": 0.0498, "step": 36310 }, { "grad_norm": 0.41754046082496643, "learning_rate": 7.549617694539011e-05, "loss": 0.054, "step": 36320 }, { "grad_norm": 0.5301085710525513, "learning_rate": 7.548195207793066e-05, "loss": 0.0586, "step": 36330 }, { "grad_norm": 0.5757991671562195, "learning_rate": 7.546772442380323e-05, "loss": 0.0433, "step": 36340 }, { "grad_norm": 0.5238051414489746, "learning_rate": 7.545349398456376e-05, "loss": 0.0485, "step": 36350 }, { "grad_norm": 0.6333210468292236, "learning_rate": 7.543926076176845e-05, "loss": 0.0419, "step": 36360 }, { "grad_norm": 0.7835444808006287, "learning_rate": 7.542502475697385e-05, "loss": 0.0464, "step": 36370 }, { "grad_norm": 0.5055364370346069, "learning_rate": 7.541078597173675e-05, "loss": 0.0493, "step": 36380 }, { "grad_norm": 0.597962498664856, "learning_rate": 7.539654440761431e-05, "loss": 0.045, "step": 36390 }, { "grad_norm": 0.553058385848999, "learning_rate": 7.538230006616395e-05, "loss": 0.0359, "step": 36400 }, { "grad_norm": 0.5145840644836426, "learning_rate": 7.536805294894342e-05, "loss": 0.0589, "step": 36410 }, { "grad_norm": 0.45378729701042175, "learning_rate": 7.535380305751076e-05, "loss": 0.0476, "step": 36420 }, { "grad_norm": 0.7068535685539246, "learning_rate": 7.533955039342431e-05, "loss": 0.0468, "step": 36430 }, { "grad_norm": 0.49627429246902466, "learning_rate": 7.532529495824274e-05, "loss": 0.0505, "step": 36440 }, { "grad_norm": 0.5344480276107788, "learning_rate": 7.531103675352497e-05, "loss": 0.0512, "step": 36450 }, { "grad_norm": 1.1808969974517822, "learning_rate": 7.529677578083029e-05, "loss": 0.0505, "step": 36460 }, { "grad_norm": 0.8106597661972046, "learning_rate": 7.528251204171823e-05, "loss": 0.0524, "step": 36470 }, { "grad_norm": 0.7342141270637512, "learning_rate": 7.526824553774867e-05, "loss": 0.0477, "step": 36480 }, { "grad_norm": 0.5473117828369141, "learning_rate": 7.525397627048178e-05, "loss": 0.0616, "step": 36490 }, { "grad_norm": 0.6320523023605347, "learning_rate": 7.523970424147802e-05, "loss": 0.0574, "step": 36500 }, { "grad_norm": 0.4908519983291626, "learning_rate": 7.522542945229813e-05, "loss": 0.0377, "step": 36510 }, { "grad_norm": 0.6630715727806091, "learning_rate": 7.521115190450321e-05, "loss": 0.0453, "step": 36520 }, { "grad_norm": 0.7546815872192383, "learning_rate": 7.519687159965462e-05, "loss": 0.0557, "step": 36530 }, { "grad_norm": 0.70189368724823, "learning_rate": 7.518258853931403e-05, "loss": 0.0503, "step": 36540 }, { "grad_norm": 0.3613411486148834, "learning_rate": 7.516830272504342e-05, "loss": 0.0361, "step": 36550 }, { "grad_norm": 0.4595264792442322, "learning_rate": 7.515401415840509e-05, "loss": 0.0474, "step": 36560 }, { "grad_norm": 0.6682825088500977, "learning_rate": 7.513972284096155e-05, "loss": 0.0436, "step": 36570 }, { "grad_norm": 0.508323609828949, "learning_rate": 7.512542877427574e-05, "loss": 0.0368, "step": 36580 }, { "grad_norm": 0.8242812752723694, "learning_rate": 7.51111319599108e-05, "loss": 0.0454, "step": 36590 }, { "grad_norm": 0.529893159866333, "learning_rate": 7.509683239943024e-05, "loss": 0.0505, "step": 36600 }, { "grad_norm": 0.48370203375816345, "learning_rate": 7.50825300943978e-05, "loss": 0.0449, "step": 36610 }, { "grad_norm": 0.7233507037162781, "learning_rate": 7.50682250463776e-05, "loss": 0.0618, "step": 36620 }, { "grad_norm": 0.44856640696525574, "learning_rate": 7.5053917256934e-05, "loss": 0.0432, "step": 36630 }, { "grad_norm": 0.6092673540115356, "learning_rate": 7.503960672763166e-05, "loss": 0.0515, "step": 36640 }, { "grad_norm": 0.5568799376487732, "learning_rate": 7.502529346003559e-05, "loss": 0.0514, "step": 36650 }, { "grad_norm": 0.35035672783851624, "learning_rate": 7.501097745571106e-05, "loss": 0.0418, "step": 36660 }, { "grad_norm": 0.4402039051055908, "learning_rate": 7.499665871622361e-05, "loss": 0.0593, "step": 36670 }, { "grad_norm": 0.4147014617919922, "learning_rate": 7.498233724313917e-05, "loss": 0.0473, "step": 36680 }, { "grad_norm": 0.5034570097923279, "learning_rate": 7.496801303802387e-05, "loss": 0.0453, "step": 36690 }, { "grad_norm": 0.4845857322216034, "learning_rate": 7.495368610244423e-05, "loss": 0.0454, "step": 36700 }, { "grad_norm": 0.7652280330657959, "learning_rate": 7.493935643796697e-05, "loss": 0.0525, "step": 36710 }, { "grad_norm": 0.5441318154335022, "learning_rate": 7.49250240461592e-05, "loss": 0.0386, "step": 36720 }, { "grad_norm": 0.4761449098587036, "learning_rate": 7.491068892858826e-05, "loss": 0.0418, "step": 36730 }, { "grad_norm": 0.5039743781089783, "learning_rate": 7.489635108682184e-05, "loss": 0.0447, "step": 36740 }, { "grad_norm": 0.5954833030700684, "learning_rate": 7.48820105224279e-05, "loss": 0.06, "step": 36750 }, { "grad_norm": 0.6520915031433105, "learning_rate": 7.48676672369747e-05, "loss": 0.0403, "step": 36760 }, { "grad_norm": 0.4916776120662689, "learning_rate": 7.485332123203079e-05, "loss": 0.0439, "step": 36770 }, { "grad_norm": 0.5206270217895508, "learning_rate": 7.483897250916504e-05, "loss": 0.0397, "step": 36780 }, { "grad_norm": 0.7628073692321777, "learning_rate": 7.48246210699466e-05, "loss": 0.0548, "step": 36790 }, { "grad_norm": 0.5550000667572021, "learning_rate": 7.481026691594492e-05, "loss": 0.0604, "step": 36800 }, { "grad_norm": 0.5484527349472046, "learning_rate": 7.479591004872975e-05, "loss": 0.0436, "step": 36810 }, { "grad_norm": 0.559083104133606, "learning_rate": 7.478155046987114e-05, "loss": 0.0571, "step": 36820 }, { "grad_norm": 0.645405650138855, "learning_rate": 7.476718818093944e-05, "loss": 0.0493, "step": 36830 }, { "grad_norm": 0.6210358738899231, "learning_rate": 7.475282318350527e-05, "loss": 0.0579, "step": 36840 }, { "grad_norm": 0.7990049123764038, "learning_rate": 7.473845547913957e-05, "loss": 0.0583, "step": 36850 }, { "grad_norm": 0.39916086196899414, "learning_rate": 7.472408506941357e-05, "loss": 0.0427, "step": 36860 }, { "grad_norm": 0.46214261651039124, "learning_rate": 7.470971195589881e-05, "loss": 0.0374, "step": 36870 }, { "grad_norm": 0.7638441920280457, "learning_rate": 7.46953361401671e-05, "loss": 0.0471, "step": 36880 }, { "grad_norm": 0.3973270654678345, "learning_rate": 7.468095762379055e-05, "loss": 0.0344, "step": 36890 }, { "grad_norm": 0.4368070363998413, "learning_rate": 7.466657640834158e-05, "loss": 0.0557, "step": 36900 }, { "grad_norm": 0.8295287489891052, "learning_rate": 7.46521924953929e-05, "loss": 0.0476, "step": 36910 }, { "grad_norm": 0.5169959664344788, "learning_rate": 7.463780588651752e-05, "loss": 0.0568, "step": 36920 }, { "grad_norm": 0.5373004674911499, "learning_rate": 7.462341658328873e-05, "loss": 0.046, "step": 36930 }, { "grad_norm": 0.5324411988258362, "learning_rate": 7.460902458728012e-05, "loss": 0.0612, "step": 36940 }, { "grad_norm": 1.318938970565796, "learning_rate": 7.459462990006558e-05, "loss": 0.0628, "step": 36950 }, { "grad_norm": 0.6329230070114136, "learning_rate": 7.458023252321929e-05, "loss": 0.0394, "step": 36960 }, { "grad_norm": 0.5703490376472473, "learning_rate": 7.456583245831573e-05, "loss": 0.04, "step": 36970 }, { "grad_norm": 0.5209431052207947, "learning_rate": 7.455142970692968e-05, "loss": 0.0479, "step": 36980 }, { "grad_norm": 0.44132566452026367, "learning_rate": 7.453702427063615e-05, "loss": 0.0637, "step": 36990 }, { "grad_norm": 0.6963248252868652, "learning_rate": 7.452261615101057e-05, "loss": 0.047, "step": 37000 }, { "grad_norm": 0.6616309285163879, "learning_rate": 7.450820534962856e-05, "loss": 0.0658, "step": 37010 }, { "grad_norm": 0.39600858092308044, "learning_rate": 7.449379186806604e-05, "loss": 0.0438, "step": 37020 }, { "grad_norm": 0.5208168625831604, "learning_rate": 7.44793757078993e-05, "loss": 0.0461, "step": 37030 }, { "grad_norm": 0.690659761428833, "learning_rate": 7.44649568707048e-05, "loss": 0.0465, "step": 37040 }, { "grad_norm": 0.4589040279388428, "learning_rate": 7.445053535805942e-05, "loss": 0.0415, "step": 37050 }, { "grad_norm": 0.5978549122810364, "learning_rate": 7.443611117154024e-05, "loss": 0.0429, "step": 37060 }, { "grad_norm": 0.4500546157360077, "learning_rate": 7.442168431272469e-05, "loss": 0.0363, "step": 37070 }, { "grad_norm": 0.5942654013633728, "learning_rate": 7.440725478319044e-05, "loss": 0.0409, "step": 37080 }, { "grad_norm": 0.7582845687866211, "learning_rate": 7.439282258451553e-05, "loss": 0.0524, "step": 37090 }, { "grad_norm": 0.5641847848892212, "learning_rate": 7.43783877182782e-05, "loss": 0.047, "step": 37100 }, { "grad_norm": 0.9449962973594666, "learning_rate": 7.436395018605707e-05, "loss": 0.0519, "step": 37110 }, { "grad_norm": 0.6717395782470703, "learning_rate": 7.434950998943094e-05, "loss": 0.0747, "step": 37120 }, { "grad_norm": 0.3831614553928375, "learning_rate": 7.433506712997902e-05, "loss": 0.0475, "step": 37130 }, { "grad_norm": 0.769290030002594, "learning_rate": 7.432062160928077e-05, "loss": 0.0628, "step": 37140 }, { "grad_norm": 0.5724731087684631, "learning_rate": 7.430617342891588e-05, "loss": 0.0348, "step": 37150 }, { "grad_norm": 0.48826169967651367, "learning_rate": 7.429172259046442e-05, "loss": 0.0492, "step": 37160 }, { "grad_norm": 0.38288572430610657, "learning_rate": 7.42772690955067e-05, "loss": 0.0519, "step": 37170 }, { "grad_norm": 0.4792249798774719, "learning_rate": 7.426281294562332e-05, "loss": 0.0537, "step": 37180 }, { "grad_norm": 0.8742226958274841, "learning_rate": 7.424835414239519e-05, "loss": 0.065, "step": 37190 }, { "grad_norm": 0.44370606541633606, "learning_rate": 7.423389268740352e-05, "loss": 0.0534, "step": 37200 }, { "grad_norm": 0.6433680057525635, "learning_rate": 7.42194285822298e-05, "loss": 0.0501, "step": 37210 }, { "grad_norm": 0.7853065729141235, "learning_rate": 7.420496182845576e-05, "loss": 0.0426, "step": 37220 }, { "grad_norm": 1.0017377138137817, "learning_rate": 7.419049242766348e-05, "loss": 0.051, "step": 37230 }, { "grad_norm": 0.6897883415222168, "learning_rate": 7.417602038143533e-05, "loss": 0.0725, "step": 37240 }, { "grad_norm": 0.7114195227622986, "learning_rate": 7.416154569135393e-05, "loss": 0.0733, "step": 37250 }, { "grad_norm": 0.4020247161388397, "learning_rate": 7.414706835900222e-05, "loss": 0.0748, "step": 37260 }, { "grad_norm": 0.8123992681503296, "learning_rate": 7.413258838596345e-05, "loss": 0.0397, "step": 37270 }, { "grad_norm": 0.4433222711086273, "learning_rate": 7.411810577382106e-05, "loss": 0.0476, "step": 37280 }, { "grad_norm": 0.7948765754699707, "learning_rate": 7.410362052415889e-05, "loss": 0.0584, "step": 37290 }, { "grad_norm": 1.0881836414337158, "learning_rate": 7.408913263856102e-05, "loss": 0.0436, "step": 37300 }, { "grad_norm": 0.5786188840866089, "learning_rate": 7.407464211861183e-05, "loss": 0.0492, "step": 37310 }, { "grad_norm": 0.6717128157615662, "learning_rate": 7.406014896589597e-05, "loss": 0.0558, "step": 37320 }, { "grad_norm": 0.4131741225719452, "learning_rate": 7.404565318199838e-05, "loss": 0.0496, "step": 37330 }, { "grad_norm": 0.5580568909645081, "learning_rate": 7.403115476850432e-05, "loss": 0.0541, "step": 37340 }, { "grad_norm": 0.3878609836101532, "learning_rate": 7.401665372699932e-05, "loss": 0.0435, "step": 37350 }, { "grad_norm": 0.3460920453071594, "learning_rate": 7.400215005906916e-05, "loss": 0.0404, "step": 37360 }, { "grad_norm": 0.5113314390182495, "learning_rate": 7.398764376629996e-05, "loss": 0.0485, "step": 37370 }, { "grad_norm": 0.5572539567947388, "learning_rate": 7.39731348502781e-05, "loss": 0.0375, "step": 37380 }, { "grad_norm": 0.5005744695663452, "learning_rate": 7.395862331259028e-05, "loss": 0.0354, "step": 37390 }, { "grad_norm": 0.6678351163864136, "learning_rate": 7.394410915482341e-05, "loss": 0.0502, "step": 37400 }, { "grad_norm": 0.7481545805931091, "learning_rate": 7.392959237856476e-05, "loss": 0.0593, "step": 37410 }, { "grad_norm": 0.5990611910820007, "learning_rate": 7.391507298540187e-05, "loss": 0.0421, "step": 37420 }, { "grad_norm": 0.6186842918395996, "learning_rate": 7.390055097692257e-05, "loss": 0.046, "step": 37430 }, { "grad_norm": 0.3975127935409546, "learning_rate": 7.388602635471494e-05, "loss": 0.0396, "step": 37440 }, { "grad_norm": 0.6177820563316345, "learning_rate": 7.387149912036738e-05, "loss": 0.0931, "step": 37450 }, { "grad_norm": 0.4064792990684509, "learning_rate": 7.385696927546854e-05, "loss": 0.063, "step": 37460 }, { "grad_norm": 0.5564593076705933, "learning_rate": 7.384243682160743e-05, "loss": 0.0436, "step": 37470 }, { "grad_norm": 0.707462728023529, "learning_rate": 7.382790176037329e-05, "loss": 0.041, "step": 37480 }, { "grad_norm": 0.4521186649799347, "learning_rate": 7.381336409335562e-05, "loss": 0.0392, "step": 37490 }, { "grad_norm": 0.46546047925949097, "learning_rate": 7.379882382214426e-05, "loss": 0.0462, "step": 37500 }, { "grad_norm": 0.4643953740596771, "learning_rate": 7.378428094832931e-05, "loss": 0.0432, "step": 37510 }, { "grad_norm": 0.8584717512130737, "learning_rate": 7.376973547350114e-05, "loss": 0.05, "step": 37520 }, { "grad_norm": 0.6938144564628601, "learning_rate": 7.375518739925046e-05, "loss": 0.0577, "step": 37530 }, { "grad_norm": 0.5278977751731873, "learning_rate": 7.374063672716818e-05, "loss": 0.0653, "step": 37540 }, { "grad_norm": 0.43358665704727173, "learning_rate": 7.372608345884558e-05, "loss": 0.0419, "step": 37550 }, { "grad_norm": 0.46836057305336, "learning_rate": 7.371152759587415e-05, "loss": 0.0753, "step": 37560 }, { "grad_norm": 0.6458834409713745, "learning_rate": 7.369696913984572e-05, "loss": 0.0527, "step": 37570 }, { "grad_norm": 0.7156946659088135, "learning_rate": 7.368240809235237e-05, "loss": 0.0472, "step": 37580 }, { "grad_norm": 0.6197140216827393, "learning_rate": 7.366784445498649e-05, "loss": 0.0491, "step": 37590 }, { "grad_norm": 0.6746407747268677, "learning_rate": 7.36532782293407e-05, "loss": 0.0523, "step": 37600 }, { "grad_norm": 0.5643788576126099, "learning_rate": 7.363870941700797e-05, "loss": 0.0666, "step": 37610 }, { "grad_norm": 0.5519173741340637, "learning_rate": 7.362413801958152e-05, "loss": 0.0506, "step": 37620 }, { "grad_norm": 0.5882431268692017, "learning_rate": 7.360956403865486e-05, "loss": 0.0512, "step": 37630 }, { "grad_norm": 0.39027076959609985, "learning_rate": 7.359498747582177e-05, "loss": 0.0473, "step": 37640 }, { "grad_norm": 0.7130259871482849, "learning_rate": 7.358040833267634e-05, "loss": 0.0564, "step": 37650 }, { "grad_norm": 0.5415861010551453, "learning_rate": 7.356582661081288e-05, "loss": 0.0353, "step": 37660 }, { "grad_norm": 0.4427546560764313, "learning_rate": 7.355124231182607e-05, "loss": 0.0478, "step": 37670 }, { "grad_norm": 0.9908180832862854, "learning_rate": 7.353665543731083e-05, "loss": 0.0431, "step": 37680 }, { "grad_norm": 0.4193418323993683, "learning_rate": 7.352206598886229e-05, "loss": 0.0515, "step": 37690 }, { "grad_norm": 0.6635802984237671, "learning_rate": 7.350747396807601e-05, "loss": 0.0473, "step": 37700 }, { "grad_norm": 0.5438148975372314, "learning_rate": 7.349287937654772e-05, "loss": 0.0449, "step": 37710 }, { "grad_norm": 0.7885289192199707, "learning_rate": 7.347828221587345e-05, "loss": 0.0581, "step": 37720 }, { "grad_norm": 0.6432181596755981, "learning_rate": 7.346368248764955e-05, "loss": 0.052, "step": 37730 }, { "grad_norm": 0.6054680347442627, "learning_rate": 7.34490801934726e-05, "loss": 0.0477, "step": 37740 }, { "grad_norm": 0.505750834941864, "learning_rate": 7.343447533493947e-05, "loss": 0.0469, "step": 37750 }, { "grad_norm": 0.9344393610954285, "learning_rate": 7.341986791364739e-05, "loss": 0.0462, "step": 37760 }, { "grad_norm": 0.4683530032634735, "learning_rate": 7.340525793119374e-05, "loss": 0.0508, "step": 37770 }, { "grad_norm": 0.5495766401290894, "learning_rate": 7.33906453891763e-05, "loss": 0.0528, "step": 37780 }, { "grad_norm": 0.48416829109191895, "learning_rate": 7.337603028919303e-05, "loss": 0.0536, "step": 37790 }, { "grad_norm": 0.5620005130767822, "learning_rate": 7.336141263284225e-05, "loss": 0.0665, "step": 37800 }, { "grad_norm": 0.5418936014175415, "learning_rate": 7.334679242172249e-05, "loss": 0.0427, "step": 37810 }, { "grad_norm": 0.5648353099822998, "learning_rate": 7.33321696574326e-05, "loss": 0.0528, "step": 37820 }, { "grad_norm": 0.46536168456077576, "learning_rate": 7.331754434157174e-05, "loss": 0.0437, "step": 37830 }, { "grad_norm": 0.5540425181388855, "learning_rate": 7.330291647573928e-05, "loss": 0.0458, "step": 37840 }, { "grad_norm": 0.4650634229183197, "learning_rate": 7.32882860615349e-05, "loss": 0.0394, "step": 37850 }, { "grad_norm": 0.5909016728401184, "learning_rate": 7.327365310055858e-05, "loss": 0.0559, "step": 37860 }, { "grad_norm": 0.578218400478363, "learning_rate": 7.325901759441055e-05, "loss": 0.0432, "step": 37870 }, { "grad_norm": 0.5953149199485779, "learning_rate": 7.324437954469132e-05, "loss": 0.042, "step": 37880 }, { "grad_norm": 0.5010259747505188, "learning_rate": 7.32297389530017e-05, "loss": 0.0393, "step": 37890 }, { "grad_norm": 0.6784442663192749, "learning_rate": 7.321509582094274e-05, "loss": 0.0449, "step": 37900 }, { "grad_norm": 0.8220787048339844, "learning_rate": 7.320045015011582e-05, "loss": 0.0516, "step": 37910 }, { "grad_norm": 0.48012182116508484, "learning_rate": 7.318580194212253e-05, "loss": 0.0573, "step": 37920 }, { "grad_norm": 0.6227657198905945, "learning_rate": 7.317115119856483e-05, "loss": 0.0436, "step": 37930 }, { "grad_norm": 0.7207933664321899, "learning_rate": 7.315649792104485e-05, "loss": 0.0561, "step": 37940 }, { "grad_norm": 0.8501777648925781, "learning_rate": 7.314184211116507e-05, "loss": 0.0517, "step": 37950 }, { "grad_norm": 0.8433561325073242, "learning_rate": 7.312718377052823e-05, "loss": 0.0457, "step": 37960 }, { "grad_norm": 0.5965112447738647, "learning_rate": 7.311252290073737e-05, "loss": 0.0427, "step": 37970 }, { "grad_norm": 0.5852581858634949, "learning_rate": 7.309785950339572e-05, "loss": 0.0482, "step": 37980 }, { "grad_norm": 0.6628265380859375, "learning_rate": 7.308319358010692e-05, "loss": 0.0387, "step": 37990 }, { "grad_norm": 0.4789709448814392, "learning_rate": 7.306852513247474e-05, "loss": 0.0618, "step": 38000 }, { "grad_norm": 0.5638805627822876, "learning_rate": 7.305385416210336e-05, "loss": 0.0446, "step": 38010 }, { "grad_norm": 0.7524433135986328, "learning_rate": 7.303918067059711e-05, "loss": 0.0529, "step": 38020 }, { "grad_norm": 0.5709922313690186, "learning_rate": 7.302450465956075e-05, "loss": 0.0482, "step": 38030 }, { "grad_norm": 0.8792669773101807, "learning_rate": 7.300982613059914e-05, "loss": 0.0428, "step": 38040 }, { "grad_norm": 0.2958456575870514, "learning_rate": 7.299514508531757e-05, "loss": 0.0298, "step": 38050 }, { "grad_norm": 0.8408222794532776, "learning_rate": 7.298046152532148e-05, "loss": 0.049, "step": 38060 }, { "grad_norm": 0.33565977215766907, "learning_rate": 7.296577545221668e-05, "loss": 0.0671, "step": 38070 }, { "grad_norm": 0.4745369255542755, "learning_rate": 7.29510868676092e-05, "loss": 0.0349, "step": 38080 }, { "grad_norm": 0.5616754293441772, "learning_rate": 7.293639577310538e-05, "loss": 0.0525, "step": 38090 }, { "grad_norm": 0.6489152312278748, "learning_rate": 7.292170217031179e-05, "loss": 0.044, "step": 38100 }, { "grad_norm": 0.33522793650627136, "learning_rate": 7.290700606083532e-05, "loss": 0.0439, "step": 38110 }, { "grad_norm": 0.5600237250328064, "learning_rate": 7.28923074462831e-05, "loss": 0.0548, "step": 38120 }, { "grad_norm": 0.7031562924385071, "learning_rate": 7.287760632826255e-05, "loss": 0.0595, "step": 38130 }, { "grad_norm": 0.4587860405445099, "learning_rate": 7.286290270838139e-05, "loss": 0.0522, "step": 38140 }, { "grad_norm": 0.4350912570953369, "learning_rate": 7.284819658824756e-05, "loss": 0.0409, "step": 38150 }, { "grad_norm": 0.43136149644851685, "learning_rate": 7.283348796946929e-05, "loss": 0.0518, "step": 38160 }, { "grad_norm": 0.2885681390762329, "learning_rate": 7.28187768536551e-05, "loss": 0.0599, "step": 38170 }, { "grad_norm": 0.43851128220558167, "learning_rate": 7.280406324241379e-05, "loss": 0.0456, "step": 38180 }, { "grad_norm": 0.6731852293014526, "learning_rate": 7.278934713735441e-05, "loss": 0.0661, "step": 38190 }, { "grad_norm": 0.424345999956131, "learning_rate": 7.277462854008629e-05, "loss": 0.0442, "step": 38200 }, { "grad_norm": 0.719321072101593, "learning_rate": 7.275990745221904e-05, "loss": 0.0437, "step": 38210 }, { "grad_norm": 0.4708038568496704, "learning_rate": 7.274518387536252e-05, "loss": 0.0496, "step": 38220 }, { "grad_norm": 0.5269182920455933, "learning_rate": 7.273045781112689e-05, "loss": 0.0433, "step": 38230 }, { "grad_norm": 0.33227673172950745, "learning_rate": 7.271572926112257e-05, "loss": 0.0542, "step": 38240 }, { "grad_norm": 0.32769548892974854, "learning_rate": 7.270099822696024e-05, "loss": 0.0528, "step": 38250 }, { "grad_norm": 0.3329704999923706, "learning_rate": 7.26862647102509e-05, "loss": 0.0715, "step": 38260 }, { "grad_norm": 0.49345919489860535, "learning_rate": 7.267152871260573e-05, "loss": 0.037, "step": 38270 }, { "grad_norm": 0.5326594114303589, "learning_rate": 7.265679023563629e-05, "loss": 0.0418, "step": 38280 }, { "grad_norm": 0.4523646831512451, "learning_rate": 7.264204928095433e-05, "loss": 0.0438, "step": 38290 }, { "grad_norm": 0.6031946539878845, "learning_rate": 7.262730585017188e-05, "loss": 0.0358, "step": 38300 }, { "grad_norm": 0.6264281868934631, "learning_rate": 7.261255994490131e-05, "loss": 0.0461, "step": 38310 }, { "grad_norm": 0.6331522464752197, "learning_rate": 7.259781156675517e-05, "loss": 0.0476, "step": 38320 }, { "grad_norm": 0.24607881903648376, "learning_rate": 7.258306071734633e-05, "loss": 0.0342, "step": 38330 }, { "grad_norm": 0.4551982581615448, "learning_rate": 7.256830739828792e-05, "loss": 0.05, "step": 38340 }, { "grad_norm": 0.6277351975440979, "learning_rate": 7.255355161119336e-05, "loss": 0.0524, "step": 38350 }, { "grad_norm": 0.7084808349609375, "learning_rate": 7.253879335767628e-05, "loss": 0.0518, "step": 38360 }, { "grad_norm": 1.081127405166626, "learning_rate": 7.252403263935067e-05, "loss": 0.0894, "step": 38370 }, { "grad_norm": 0.6711466908454895, "learning_rate": 7.25092694578307e-05, "loss": 0.054, "step": 38380 }, { "grad_norm": 0.47202467918395996, "learning_rate": 7.249450381473087e-05, "loss": 0.0426, "step": 38390 }, { "grad_norm": 1.0321215391159058, "learning_rate": 7.247973571166593e-05, "loss": 0.0711, "step": 38400 }, { "grad_norm": 0.4228329062461853, "learning_rate": 7.246496515025089e-05, "loss": 0.0444, "step": 38410 }, { "grad_norm": 0.6093732714653015, "learning_rate": 7.245019213210101e-05, "loss": 0.0383, "step": 38420 }, { "grad_norm": 0.5794056057929993, "learning_rate": 7.24354166588319e-05, "loss": 0.0435, "step": 38430 }, { "grad_norm": 0.9721533060073853, "learning_rate": 7.242063873205935e-05, "loss": 0.0464, "step": 38440 }, { "grad_norm": 0.8285031914710999, "learning_rate": 7.240585835339946e-05, "loss": 0.0482, "step": 38450 }, { "grad_norm": 0.7308717370033264, "learning_rate": 7.239107552446857e-05, "loss": 0.0568, "step": 38460 }, { "grad_norm": 0.5072964429855347, "learning_rate": 7.237629024688334e-05, "loss": 0.0414, "step": 38470 }, { "grad_norm": 0.5884511470794678, "learning_rate": 7.236150252226064e-05, "loss": 0.053, "step": 38480 }, { "grad_norm": 0.4504544734954834, "learning_rate": 7.234671235221765e-05, "loss": 0.0479, "step": 38490 }, { "grad_norm": 0.5316330194473267, "learning_rate": 7.233191973837179e-05, "loss": 0.0409, "step": 38500 }, { "grad_norm": 0.5179964303970337, "learning_rate": 7.231712468234078e-05, "loss": 0.0428, "step": 38510 }, { "grad_norm": 0.6296061873435974, "learning_rate": 7.230232718574254e-05, "loss": 0.0466, "step": 38520 }, { "grad_norm": 0.5811984539031982, "learning_rate": 7.228752725019535e-05, "loss": 0.0395, "step": 38530 }, { "grad_norm": 0.7512699961662292, "learning_rate": 7.227272487731769e-05, "loss": 0.0378, "step": 38540 }, { "grad_norm": 0.5626817345619202, "learning_rate": 7.225792006872831e-05, "loss": 0.034, "step": 38550 }, { "grad_norm": 0.48706337809562683, "learning_rate": 7.224311282604628e-05, "loss": 0.0314, "step": 38560 }, { "grad_norm": 0.34414657950401306, "learning_rate": 7.222830315089085e-05, "loss": 0.0486, "step": 38570 }, { "grad_norm": 0.6691610813140869, "learning_rate": 7.22134910448816e-05, "loss": 0.067, "step": 38580 }, { "grad_norm": 0.598763644695282, "learning_rate": 7.219867650963839e-05, "loss": 0.0392, "step": 38590 }, { "grad_norm": 0.3188248872756958, "learning_rate": 7.21838595467813e-05, "loss": 0.0524, "step": 38600 }, { "grad_norm": 0.7399426102638245, "learning_rate": 7.216904015793067e-05, "loss": 0.053, "step": 38610 }, { "grad_norm": 0.4787661135196686, "learning_rate": 7.215421834470713e-05, "loss": 0.0501, "step": 38620 }, { "grad_norm": 0.3645237684249878, "learning_rate": 7.213939410873158e-05, "loss": 0.0579, "step": 38630 }, { "grad_norm": 0.6209582090377808, "learning_rate": 7.212456745162519e-05, "loss": 0.0621, "step": 38640 }, { "grad_norm": 0.5781618356704712, "learning_rate": 7.210973837500937e-05, "loss": 0.0503, "step": 38650 }, { "grad_norm": 0.6891327500343323, "learning_rate": 7.209490688050578e-05, "loss": 0.0797, "step": 38660 }, { "grad_norm": 0.6189484596252441, "learning_rate": 7.208007296973641e-05, "loss": 0.0447, "step": 38670 }, { "grad_norm": 0.9244091510772705, "learning_rate": 7.206523664432345e-05, "loss": 0.0792, "step": 38680 }, { "grad_norm": 0.4277634620666504, "learning_rate": 7.205039790588939e-05, "loss": 0.0523, "step": 38690 }, { "grad_norm": 0.44156163930892944, "learning_rate": 7.203555675605697e-05, "loss": 0.0579, "step": 38700 }, { "grad_norm": 0.6194189190864563, "learning_rate": 7.202071319644917e-05, "loss": 0.0424, "step": 38710 }, { "grad_norm": 0.5031735897064209, "learning_rate": 7.200586722868932e-05, "loss": 0.0486, "step": 38720 }, { "grad_norm": 1.0435187816619873, "learning_rate": 7.199101885440088e-05, "loss": 0.0508, "step": 38730 }, { "grad_norm": 0.4259663224220276, "learning_rate": 7.197616807520771e-05, "loss": 0.038, "step": 38740 }, { "grad_norm": 0.8536447286605835, "learning_rate": 7.196131489273381e-05, "loss": 0.0462, "step": 38750 }, { "grad_norm": 0.5648546814918518, "learning_rate": 7.194645930860355e-05, "loss": 0.0744, "step": 38760 }, { "grad_norm": 0.7823297381401062, "learning_rate": 7.19316013244415e-05, "loss": 0.0505, "step": 38770 }, { "grad_norm": 0.5098614692687988, "learning_rate": 7.191674094187248e-05, "loss": 0.0468, "step": 38780 }, { "grad_norm": 0.6015757918357849, "learning_rate": 7.190187816252165e-05, "loss": 0.0482, "step": 38790 }, { "grad_norm": 0.40865767002105713, "learning_rate": 7.188701298801435e-05, "loss": 0.0516, "step": 38800 }, { "grad_norm": 0.783567488193512, "learning_rate": 7.18721454199762e-05, "loss": 0.0439, "step": 38810 }, { "grad_norm": 0.5598047971725464, "learning_rate": 7.185727546003312e-05, "loss": 0.0396, "step": 38820 }, { "grad_norm": 0.4233526587486267, "learning_rate": 7.184240310981126e-05, "loss": 0.0565, "step": 38830 }, { "grad_norm": 0.48612916469573975, "learning_rate": 7.182752837093704e-05, "loss": 0.0343, "step": 38840 }, { "grad_norm": 0.26375067234039307, "learning_rate": 7.181265124503711e-05, "loss": 0.0444, "step": 38850 }, { "grad_norm": 0.9165248274803162, "learning_rate": 7.179777173373847e-05, "loss": 0.0903, "step": 38860 }, { "grad_norm": 0.5210519433021545, "learning_rate": 7.178288983866826e-05, "loss": 0.0484, "step": 38870 }, { "grad_norm": 0.3343445360660553, "learning_rate": 7.176800556145397e-05, "loss": 0.0528, "step": 38880 }, { "grad_norm": 0.6303686499595642, "learning_rate": 7.175311890372334e-05, "loss": 0.041, "step": 38890 }, { "grad_norm": 0.5989748239517212, "learning_rate": 7.17382298671043e-05, "loss": 0.0528, "step": 38900 }, { "grad_norm": 0.6537275314331055, "learning_rate": 7.172333845322515e-05, "loss": 0.0463, "step": 38910 }, { "grad_norm": 0.6603087186813354, "learning_rate": 7.170844466371436e-05, "loss": 0.0419, "step": 38920 }, { "grad_norm": 0.7065642476081848, "learning_rate": 7.16935485002007e-05, "loss": 0.0625, "step": 38930 }, { "grad_norm": 0.5932770371437073, "learning_rate": 7.167864996431319e-05, "loss": 0.0638, "step": 38940 }, { "grad_norm": 0.6587347984313965, "learning_rate": 7.166374905768111e-05, "loss": 0.039, "step": 38950 }, { "grad_norm": 0.6103933453559875, "learning_rate": 7.164884578193401e-05, "loss": 0.0476, "step": 38960 }, { "grad_norm": 0.7686602473258972, "learning_rate": 7.163394013870168e-05, "loss": 0.06, "step": 38970 }, { "grad_norm": 0.5587769746780396, "learning_rate": 7.161903212961419e-05, "loss": 0.0618, "step": 38980 }, { "grad_norm": 0.45120370388031006, "learning_rate": 7.160412175630183e-05, "loss": 0.0532, "step": 38990 }, { "grad_norm": 0.6461181640625, "learning_rate": 7.158920902039521e-05, "loss": 0.0413, "step": 39000 }, { "grad_norm": 0.7146856188774109, "learning_rate": 7.157429392352514e-05, "loss": 0.0553, "step": 39010 }, { "grad_norm": 0.5794014930725098, "learning_rate": 7.155937646732275e-05, "loss": 0.0544, "step": 39020 }, { "grad_norm": 0.41435977816581726, "learning_rate": 7.154445665341933e-05, "loss": 0.0407, "step": 39030 }, { "grad_norm": 0.39479660987854004, "learning_rate": 7.152953448344654e-05, "loss": 0.0427, "step": 39040 }, { "grad_norm": 0.7639192342758179, "learning_rate": 7.151460995903624e-05, "loss": 0.042, "step": 39050 }, { "grad_norm": 0.5753194689750671, "learning_rate": 7.149968308182052e-05, "loss": 0.0461, "step": 39060 }, { "grad_norm": 0.5419573187828064, "learning_rate": 7.14847538534318e-05, "loss": 0.0432, "step": 39070 }, { "grad_norm": 0.5406008958816528, "learning_rate": 7.14698222755027e-05, "loss": 0.0412, "step": 39080 }, { "grad_norm": 0.6337255835533142, "learning_rate": 7.14548883496661e-05, "loss": 0.0555, "step": 39090 }, { "grad_norm": 0.400088906288147, "learning_rate": 7.143995207755517e-05, "loss": 0.0428, "step": 39100 }, { "grad_norm": 0.6724062561988831, "learning_rate": 7.142501346080333e-05, "loss": 0.0551, "step": 39110 }, { "grad_norm": 0.500503420829773, "learning_rate": 7.141007250104421e-05, "loss": 0.0593, "step": 39120 }, { "grad_norm": 0.6998113989830017, "learning_rate": 7.139512919991176e-05, "loss": 0.0543, "step": 39130 }, { "grad_norm": 0.3811032474040985, "learning_rate": 7.138018355904014e-05, "loss": 0.0428, "step": 39140 }, { "grad_norm": 0.5105065703392029, "learning_rate": 7.13652355800638e-05, "loss": 0.0567, "step": 39150 }, { "grad_norm": 0.5345547795295715, "learning_rate": 7.135028526461743e-05, "loss": 0.0393, "step": 39160 }, { "grad_norm": 0.6546329259872437, "learning_rate": 7.133533261433594e-05, "loss": 0.0397, "step": 39170 }, { "grad_norm": 0.614482581615448, "learning_rate": 7.132037763085457e-05, "loss": 0.0363, "step": 39180 }, { "grad_norm": 0.62531578540802, "learning_rate": 7.130542031580875e-05, "loss": 0.0521, "step": 39190 }, { "grad_norm": 0.6512128710746765, "learning_rate": 7.12904606708342e-05, "loss": 0.043, "step": 39200 }, { "grad_norm": 0.4687196910381317, "learning_rate": 7.127549869756687e-05, "loss": 0.0521, "step": 39210 }, { "grad_norm": 0.37457582354545593, "learning_rate": 7.126053439764299e-05, "loss": 0.0425, "step": 39220 }, { "grad_norm": 0.4602105915546417, "learning_rate": 7.124556777269904e-05, "loss": 0.0494, "step": 39230 }, { "grad_norm": 0.5400976538658142, "learning_rate": 7.123059882437174e-05, "loss": 0.0391, "step": 39240 }, { "grad_norm": 0.544532835483551, "learning_rate": 7.121562755429807e-05, "loss": 0.0549, "step": 39250 }, { "grad_norm": 0.47585803270339966, "learning_rate": 7.120065396411528e-05, "loss": 0.0367, "step": 39260 }, { "grad_norm": 0.36478662490844727, "learning_rate": 7.118567805546084e-05, "loss": 0.0372, "step": 39270 }, { "grad_norm": 0.42540818452835083, "learning_rate": 7.117069982997248e-05, "loss": 0.0374, "step": 39280 }, { "grad_norm": 0.42705702781677246, "learning_rate": 7.115571928928825e-05, "loss": 0.0396, "step": 39290 }, { "grad_norm": 0.4161660671234131, "learning_rate": 7.114073643504635e-05, "loss": 0.041, "step": 39300 }, { "grad_norm": 0.6657724976539612, "learning_rate": 7.11257512688853e-05, "loss": 0.0417, "step": 39310 }, { "grad_norm": 0.4504674971103668, "learning_rate": 7.111076379244384e-05, "loss": 0.0471, "step": 39320 }, { "grad_norm": 0.716223418712616, "learning_rate": 7.109577400736101e-05, "loss": 0.0526, "step": 39330 }, { "grad_norm": 0.4912082254886627, "learning_rate": 7.108078191527605e-05, "loss": 0.0464, "step": 39340 }, { "grad_norm": 0.6158248782157898, "learning_rate": 7.106578751782847e-05, "loss": 0.0448, "step": 39350 }, { "grad_norm": 0.49315157532691956, "learning_rate": 7.105079081665803e-05, "loss": 0.0409, "step": 39360 }, { "grad_norm": 0.522064208984375, "learning_rate": 7.103579181340476e-05, "loss": 0.0603, "step": 39370 }, { "grad_norm": 0.400621622800827, "learning_rate": 7.102079050970893e-05, "loss": 0.0402, "step": 39380 }, { "grad_norm": 0.5406233668327332, "learning_rate": 7.100578690721105e-05, "loss": 0.0465, "step": 39390 }, { "grad_norm": 0.9256009459495544, "learning_rate": 7.09907810075519e-05, "loss": 0.0603, "step": 39400 }, { "grad_norm": 0.7401074767112732, "learning_rate": 7.097577281237249e-05, "loss": 0.0565, "step": 39410 }, { "grad_norm": 0.4590173661708832, "learning_rate": 7.09607623233141e-05, "loss": 0.0528, "step": 39420 }, { "grad_norm": 0.6823176145553589, "learning_rate": 7.094574954201827e-05, "loss": 0.0413, "step": 39430 }, { "grad_norm": 1.3380917310714722, "learning_rate": 7.093073447012675e-05, "loss": 0.0575, "step": 39440 }, { "grad_norm": 0.5231727361679077, "learning_rate": 7.09157171092816e-05, "loss": 0.0596, "step": 39450 }, { "grad_norm": 0.7505602836608887, "learning_rate": 7.090069746112504e-05, "loss": 0.0504, "step": 39460 }, { "grad_norm": 0.7941405773162842, "learning_rate": 7.088567552729965e-05, "loss": 0.0553, "step": 39470 }, { "grad_norm": 0.8985861539840698, "learning_rate": 7.087065130944818e-05, "loss": 0.0611, "step": 39480 }, { "grad_norm": 0.5362274646759033, "learning_rate": 7.085562480921366e-05, "loss": 0.0353, "step": 39490 }, { "grad_norm": 0.21273046731948853, "learning_rate": 7.084059602823937e-05, "loss": 0.0363, "step": 39500 }, { "grad_norm": 0.47779685258865356, "learning_rate": 7.082556496816882e-05, "loss": 0.0542, "step": 39510 }, { "grad_norm": 0.40584996342658997, "learning_rate": 7.081053163064582e-05, "loss": 0.0635, "step": 39520 }, { "grad_norm": 0.6561947464942932, "learning_rate": 7.079549601731434e-05, "loss": 0.0564, "step": 39530 }, { "grad_norm": 0.45216628909111023, "learning_rate": 7.07804581298187e-05, "loss": 0.0405, "step": 39540 }, { "grad_norm": 0.869432806968689, "learning_rate": 7.07654179698034e-05, "loss": 0.0644, "step": 39550 }, { "grad_norm": 0.4880632162094116, "learning_rate": 7.075037553891321e-05, "loss": 0.043, "step": 39560 }, { "grad_norm": 0.3539785146713257, "learning_rate": 7.073533083879315e-05, "loss": 0.052, "step": 39570 }, { "grad_norm": 0.5001407265663147, "learning_rate": 7.072028387108849e-05, "loss": 0.0565, "step": 39580 }, { "grad_norm": 0.4373812973499298, "learning_rate": 7.070523463744472e-05, "loss": 0.0516, "step": 39590 }, { "grad_norm": 0.4821024537086487, "learning_rate": 7.069018313950763e-05, "loss": 0.0518, "step": 39600 }, { "grad_norm": 0.5945042967796326, "learning_rate": 7.067512937892322e-05, "loss": 0.0515, "step": 39610 }, { "grad_norm": 0.49526289105415344, "learning_rate": 7.066007335733775e-05, "loss": 0.0492, "step": 39620 }, { "grad_norm": 0.5396716594696045, "learning_rate": 7.064501507639772e-05, "loss": 0.0557, "step": 39630 }, { "grad_norm": 0.4144168198108673, "learning_rate": 7.062995453774987e-05, "loss": 0.0474, "step": 39640 }, { "grad_norm": 0.6459541320800781, "learning_rate": 7.061489174304121e-05, "loss": 0.0337, "step": 39650 }, { "grad_norm": 0.5221291184425354, "learning_rate": 7.0599826693919e-05, "loss": 0.0321, "step": 39660 }, { "grad_norm": 0.6908265352249146, "learning_rate": 7.05847593920307e-05, "loss": 0.0431, "step": 39670 }, { "grad_norm": 0.3886229395866394, "learning_rate": 7.056968983902406e-05, "loss": 0.0313, "step": 39680 }, { "grad_norm": 0.6028208136558533, "learning_rate": 7.055461803654706e-05, "loss": 0.0419, "step": 39690 }, { "grad_norm": 0.770945131778717, "learning_rate": 7.053954398624794e-05, "loss": 0.0759, "step": 39700 }, { "grad_norm": 0.4465915858745575, "learning_rate": 7.052446768977518e-05, "loss": 0.0382, "step": 39710 }, { "grad_norm": 0.9533628225326538, "learning_rate": 7.050938914877748e-05, "loss": 0.073, "step": 39720 }, { "grad_norm": 0.42660272121429443, "learning_rate": 7.049430836490382e-05, "loss": 0.035, "step": 39730 }, { "grad_norm": 0.47157832980155945, "learning_rate": 7.047922533980342e-05, "loss": 0.0425, "step": 39740 }, { "grad_norm": 0.51576828956604, "learning_rate": 7.046414007512571e-05, "loss": 0.0599, "step": 39750 }, { "grad_norm": 0.5910781621932983, "learning_rate": 7.044905257252042e-05, "loss": 0.0597, "step": 39760 }, { "grad_norm": 0.6582907438278198, "learning_rate": 7.04339628336375e-05, "loss": 0.0539, "step": 39770 }, { "grad_norm": 0.9239331483840942, "learning_rate": 7.041887086012711e-05, "loss": 0.0463, "step": 39780 }, { "grad_norm": 0.5978990197181702, "learning_rate": 7.040377665363969e-05, "loss": 0.0587, "step": 39790 }, { "grad_norm": 0.7425133585929871, "learning_rate": 7.038868021582594e-05, "loss": 0.0825, "step": 39800 }, { "grad_norm": 0.8151009678840637, "learning_rate": 7.037358154833679e-05, "loss": 0.0496, "step": 39810 }, { "grad_norm": 0.5139080882072449, "learning_rate": 7.035848065282339e-05, "loss": 0.0417, "step": 39820 }, { "grad_norm": 0.579353392124176, "learning_rate": 7.034337753093714e-05, "loss": 0.0482, "step": 39830 }, { "grad_norm": 0.6980199813842773, "learning_rate": 7.032827218432972e-05, "loss": 0.0551, "step": 39840 }, { "grad_norm": 0.6832898855209351, "learning_rate": 7.031316461465302e-05, "loss": 0.0489, "step": 39850 }, { "grad_norm": 0.552950382232666, "learning_rate": 7.029805482355915e-05, "loss": 0.0603, "step": 39860 }, { "grad_norm": 0.5962393879890442, "learning_rate": 7.028294281270055e-05, "loss": 0.0422, "step": 39870 }, { "grad_norm": 0.6370553374290466, "learning_rate": 7.026782858372982e-05, "loss": 0.0571, "step": 39880 }, { "grad_norm": 0.5773964524269104, "learning_rate": 7.025271213829982e-05, "loss": 0.0372, "step": 39890 }, { "grad_norm": 0.5813879370689392, "learning_rate": 7.023759347806366e-05, "loss": 0.0501, "step": 39900 }, { "grad_norm": 0.7149528861045837, "learning_rate": 7.022247260467472e-05, "loss": 0.0536, "step": 39910 }, { "grad_norm": 0.5787070393562317, "learning_rate": 7.020734951978658e-05, "loss": 0.0399, "step": 39920 }, { "grad_norm": 0.393403023481369, "learning_rate": 7.019222422505307e-05, "loss": 0.0531, "step": 39930 }, { "grad_norm": 0.7004271149635315, "learning_rate": 7.017709672212827e-05, "loss": 0.0398, "step": 39940 }, { "grad_norm": 0.4882737398147583, "learning_rate": 7.016196701266652e-05, "loss": 0.046, "step": 39950 }, { "grad_norm": 0.6509738564491272, "learning_rate": 7.014683509832238e-05, "loss": 0.0553, "step": 39960 }, { "grad_norm": 0.5881996154785156, "learning_rate": 7.013170098075063e-05, "loss": 0.0444, "step": 39970 }, { "grad_norm": 0.46960338950157166, "learning_rate": 7.011656466160632e-05, "loss": 0.0397, "step": 39980 }, { "grad_norm": 0.5722256898880005, "learning_rate": 7.010142614254475e-05, "loss": 0.0461, "step": 39990 }, { "grad_norm": 0.5382997393608093, "learning_rate": 7.008628542522147e-05, "loss": 0.0483, "step": 40000 }, { "grad_norm": 0.5089513659477234, "learning_rate": 7.007114251129218e-05, "loss": 0.0565, "step": 40010 }, { "grad_norm": 0.5009747743606567, "learning_rate": 7.005599740241293e-05, "loss": 0.0394, "step": 40020 }, { "grad_norm": 0.6115933656692505, "learning_rate": 7.004085010023996e-05, "loss": 0.0424, "step": 40030 }, { "grad_norm": 0.5134349465370178, "learning_rate": 7.002570060642976e-05, "loss": 0.0502, "step": 40040 }, { "grad_norm": 0.6082241535186768, "learning_rate": 7.001054892263903e-05, "loss": 0.0468, "step": 40050 }, { "grad_norm": 0.7432689666748047, "learning_rate": 6.999539505052477e-05, "loss": 0.0349, "step": 40060 }, { "grad_norm": 0.48680534958839417, "learning_rate": 6.998023899174418e-05, "loss": 0.0759, "step": 40070 }, { "grad_norm": 0.557243824005127, "learning_rate": 6.996508074795467e-05, "loss": 0.0419, "step": 40080 }, { "grad_norm": 0.4887785017490387, "learning_rate": 6.994992032081396e-05, "loss": 0.0387, "step": 40090 }, { "grad_norm": 0.3928995728492737, "learning_rate": 6.993475771197995e-05, "loss": 0.0493, "step": 40100 }, { "grad_norm": 0.5339394211769104, "learning_rate": 6.991959292311082e-05, "loss": 0.0355, "step": 40110 }, { "grad_norm": 0.6319310069084167, "learning_rate": 6.990442595586495e-05, "loss": 0.0382, "step": 40120 }, { "grad_norm": 0.3162306547164917, "learning_rate": 6.988925681190098e-05, "loss": 0.0456, "step": 40130 }, { "grad_norm": 0.646583080291748, "learning_rate": 6.987408549287778e-05, "loss": 0.0664, "step": 40140 }, { "grad_norm": 0.49571701884269714, "learning_rate": 6.985891200045449e-05, "loss": 0.0485, "step": 40150 }, { "grad_norm": 0.5205399990081787, "learning_rate": 6.984373633629045e-05, "loss": 0.0389, "step": 40160 }, { "grad_norm": 0.6367200613021851, "learning_rate": 6.98285585020452e-05, "loss": 0.0502, "step": 40170 }, { "grad_norm": 0.3189334571361542, "learning_rate": 6.981337849937864e-05, "loss": 0.0432, "step": 40180 }, { "grad_norm": 0.4581611752510071, "learning_rate": 6.979819632995078e-05, "loss": 0.0346, "step": 40190 }, { "grad_norm": 0.5112259984016418, "learning_rate": 6.978301199542193e-05, "loss": 0.0478, "step": 40200 }, { "grad_norm": 0.49225491285324097, "learning_rate": 6.976782549745263e-05, "loss": 0.0455, "step": 40210 }, { "grad_norm": 0.43159592151641846, "learning_rate": 6.975263683770367e-05, "loss": 0.0367, "step": 40220 }, { "grad_norm": 0.480012446641922, "learning_rate": 6.973744601783604e-05, "loss": 0.0497, "step": 40230 }, { "grad_norm": 0.7385151982307434, "learning_rate": 6.972225303951097e-05, "loss": 0.0454, "step": 40240 }, { "grad_norm": 0.5719454884529114, "learning_rate": 6.970705790438998e-05, "loss": 0.0557, "step": 40250 }, { "grad_norm": 0.5203103423118591, "learning_rate": 6.969186061413477e-05, "loss": 0.0405, "step": 40260 }, { "grad_norm": 0.7172430157661438, "learning_rate": 6.967666117040727e-05, "loss": 0.0518, "step": 40270 }, { "grad_norm": 0.72115159034729, "learning_rate": 6.966145957486972e-05, "loss": 0.0394, "step": 40280 }, { "grad_norm": 0.48412591218948364, "learning_rate": 6.964625582918449e-05, "loss": 0.0538, "step": 40290 }, { "grad_norm": 0.34197738766670227, "learning_rate": 6.963104993501425e-05, "loss": 0.044, "step": 40300 }, { "grad_norm": 0.5559509992599487, "learning_rate": 6.961584189402192e-05, "loss": 0.0525, "step": 40310 }, { "grad_norm": 0.44404590129852295, "learning_rate": 6.96006317078706e-05, "loss": 0.047, "step": 40320 }, { "grad_norm": 0.600018322467804, "learning_rate": 6.958541937822367e-05, "loss": 0.0637, "step": 40330 }, { "grad_norm": 0.6914344429969788, "learning_rate": 6.957020490674473e-05, "loss": 0.0508, "step": 40340 }, { "grad_norm": 0.2705029845237732, "learning_rate": 6.95549882950976e-05, "loss": 0.0506, "step": 40350 }, { "grad_norm": 0.2885328531265259, "learning_rate": 6.953976954494635e-05, "loss": 0.0478, "step": 40360 }, { "grad_norm": 0.7000731229782104, "learning_rate": 6.952454865795528e-05, "loss": 0.04, "step": 40370 }, { "grad_norm": 0.5469340682029724, "learning_rate": 6.950932563578892e-05, "loss": 0.0381, "step": 40380 }, { "grad_norm": 0.5898646712303162, "learning_rate": 6.949410048011206e-05, "loss": 0.048, "step": 40390 }, { "grad_norm": 0.8982322812080383, "learning_rate": 6.947887319258966e-05, "loss": 0.0518, "step": 40400 }, { "grad_norm": 0.8559104204177856, "learning_rate": 6.946364377488696e-05, "loss": 0.055, "step": 40410 }, { "grad_norm": 0.49183350801467896, "learning_rate": 6.944841222866947e-05, "loss": 0.0444, "step": 40420 }, { "grad_norm": 0.4646814465522766, "learning_rate": 6.943317855560284e-05, "loss": 0.0551, "step": 40430 }, { "grad_norm": 0.5081223845481873, "learning_rate": 6.941794275735302e-05, "loss": 0.0342, "step": 40440 }, { "grad_norm": 0.4592117667198181, "learning_rate": 6.94027048355862e-05, "loss": 0.037, "step": 40450 }, { "grad_norm": 0.8040829300880432, "learning_rate": 6.938746479196871e-05, "loss": 0.0753, "step": 40460 }, { "grad_norm": 0.5084847211837769, "learning_rate": 6.937222262816724e-05, "loss": 0.0505, "step": 40470 }, { "grad_norm": 0.6678411960601807, "learning_rate": 6.935697834584865e-05, "loss": 0.0476, "step": 40480 }, { "grad_norm": 0.4378424286842346, "learning_rate": 6.934173194667998e-05, "loss": 0.0534, "step": 40490 }, { "grad_norm": 0.6645277142524719, "learning_rate": 6.93264834323286e-05, "loss": 0.0408, "step": 40500 }, { "grad_norm": 0.4710036516189575, "learning_rate": 6.931123280446206e-05, "loss": 0.0498, "step": 40510 }, { "grad_norm": 0.6040021181106567, "learning_rate": 6.929598006474811e-05, "loss": 0.0357, "step": 40520 }, { "grad_norm": 0.5626167058944702, "learning_rate": 6.92807252148548e-05, "loss": 0.0298, "step": 40530 }, { "grad_norm": 0.4023161232471466, "learning_rate": 6.926546825645037e-05, "loss": 0.0338, "step": 40540 }, { "grad_norm": 0.44359976053237915, "learning_rate": 6.92502091912033e-05, "loss": 0.0451, "step": 40550 }, { "grad_norm": 0.6494433879852295, "learning_rate": 6.923494802078229e-05, "loss": 0.0494, "step": 40560 }, { "grad_norm": 0.6073871850967407, "learning_rate": 6.921968474685628e-05, "loss": 0.062, "step": 40570 }, { "grad_norm": 0.5236456394195557, "learning_rate": 6.920441937109445e-05, "loss": 0.0502, "step": 40580 }, { "grad_norm": 0.6938263177871704, "learning_rate": 6.91891518951662e-05, "loss": 0.0416, "step": 40590 }, { "grad_norm": 0.6592438817024231, "learning_rate": 6.917388232074114e-05, "loss": 0.0481, "step": 40600 }, { "grad_norm": 0.5049991011619568, "learning_rate": 6.915861064948914e-05, "loss": 0.057, "step": 40610 }, { "grad_norm": 0.8062890768051147, "learning_rate": 6.914333688308025e-05, "loss": 0.0455, "step": 40620 }, { "grad_norm": 0.6568120121955872, "learning_rate": 6.912806102318487e-05, "loss": 0.0473, "step": 40630 }, { "grad_norm": 0.5400828719139099, "learning_rate": 6.911278307147347e-05, "loss": 0.0453, "step": 40640 }, { "grad_norm": 1.2338945865631104, "learning_rate": 6.909750302961684e-05, "loss": 0.0511, "step": 40650 }, { "grad_norm": 0.5860038995742798, "learning_rate": 6.9082220899286e-05, "loss": 0.0369, "step": 40660 }, { "grad_norm": 0.5042128562927246, "learning_rate": 6.906693668215219e-05, "loss": 0.0486, "step": 40670 }, { "grad_norm": 0.35969167947769165, "learning_rate": 6.905165037988683e-05, "loss": 0.0458, "step": 40680 }, { "grad_norm": 0.7009468078613281, "learning_rate": 6.903636199416164e-05, "loss": 0.0529, "step": 40690 }, { "grad_norm": 0.5641099214553833, "learning_rate": 6.902107152664851e-05, "loss": 0.0571, "step": 40700 }, { "grad_norm": 0.4703878164291382, "learning_rate": 6.90057789790196e-05, "loss": 0.0544, "step": 40710 }, { "grad_norm": 0.43825840950012207, "learning_rate": 6.899048435294728e-05, "loss": 0.0495, "step": 40720 }, { "grad_norm": 0.44230785965919495, "learning_rate": 6.897518765010415e-05, "loss": 0.0418, "step": 40730 }, { "grad_norm": 0.6905311346054077, "learning_rate": 6.895988887216302e-05, "loss": 0.0535, "step": 40740 }, { "grad_norm": 0.8123583793640137, "learning_rate": 6.894458802079694e-05, "loss": 0.0432, "step": 40750 }, { "grad_norm": 0.5989980697631836, "learning_rate": 6.892928509767922e-05, "loss": 0.0669, "step": 40760 }, { "grad_norm": 0.47933995723724365, "learning_rate": 6.891398010448333e-05, "loss": 0.0412, "step": 40770 }, { "grad_norm": 0.4914722144603729, "learning_rate": 6.8898673042883e-05, "loss": 0.0431, "step": 40780 }, { "grad_norm": 0.5094951391220093, "learning_rate": 6.888336391455222e-05, "loss": 0.0427, "step": 40790 }, { "grad_norm": 0.5458118319511414, "learning_rate": 6.886805272116513e-05, "loss": 0.0381, "step": 40800 }, { "grad_norm": 0.4239861071109772, "learning_rate": 6.885273946439617e-05, "loss": 0.0388, "step": 40810 }, { "grad_norm": 0.38708367943763733, "learning_rate": 6.883742414591998e-05, "loss": 0.0459, "step": 40820 }, { "grad_norm": 0.6517436504364014, "learning_rate": 6.88221067674114e-05, "loss": 0.0437, "step": 40830 }, { "grad_norm": 0.7538542747497559, "learning_rate": 6.88067873305455e-05, "loss": 0.0451, "step": 40840 }, { "grad_norm": 0.40614280104637146, "learning_rate": 6.879146583699765e-05, "loss": 0.0383, "step": 40850 }, { "grad_norm": 0.533366858959198, "learning_rate": 6.877614228844334e-05, "loss": 0.0399, "step": 40860 }, { "grad_norm": 0.44205451011657715, "learning_rate": 6.876081668655832e-05, "loss": 0.0495, "step": 40870 }, { "grad_norm": 0.96088707447052, "learning_rate": 6.874548903301861e-05, "loss": 0.051, "step": 40880 }, { "grad_norm": 0.4563271403312683, "learning_rate": 6.873015932950039e-05, "loss": 0.0438, "step": 40890 }, { "grad_norm": 0.546510636806488, "learning_rate": 6.871482757768012e-05, "loss": 0.0407, "step": 40900 }, { "grad_norm": 0.8272100687026978, "learning_rate": 6.869949377923442e-05, "loss": 0.0835, "step": 40910 }, { "grad_norm": 0.4423938989639282, "learning_rate": 6.868415793584022e-05, "loss": 0.047, "step": 40920 }, { "grad_norm": 0.6800862550735474, "learning_rate": 6.86688200491746e-05, "loss": 0.0459, "step": 40930 }, { "grad_norm": 0.5589090585708618, "learning_rate": 6.865348012091488e-05, "loss": 0.061, "step": 40940 }, { "grad_norm": 0.6367959380149841, "learning_rate": 6.86381381527386e-05, "loss": 0.0482, "step": 40950 }, { "grad_norm": 0.6418368220329285, "learning_rate": 6.862279414632359e-05, "loss": 0.0462, "step": 40960 }, { "grad_norm": 0.974134087562561, "learning_rate": 6.860744810334778e-05, "loss": 0.0377, "step": 40970 }, { "grad_norm": 0.6933169960975647, "learning_rate": 6.859210002548943e-05, "loss": 0.0338, "step": 40980 }, { "grad_norm": 0.4350048303604126, "learning_rate": 6.8576749914427e-05, "loss": 0.0346, "step": 40990 }, { "grad_norm": 0.42193102836608887, "learning_rate": 6.85613977718391e-05, "loss": 0.0567, "step": 41000 }, { "grad_norm": 0.7817487120628357, "learning_rate": 6.854604359940464e-05, "loss": 0.0468, "step": 41010 }, { "grad_norm": 0.7049188017845154, "learning_rate": 6.853068739880276e-05, "loss": 0.0476, "step": 41020 }, { "grad_norm": 0.5803288221359253, "learning_rate": 6.851532917171276e-05, "loss": 0.0401, "step": 41030 }, { "grad_norm": 0.38709205389022827, "learning_rate": 6.849996891981418e-05, "loss": 0.0433, "step": 41040 }, { "grad_norm": 0.36507001519203186, "learning_rate": 6.848460664478684e-05, "loss": 0.0442, "step": 41050 }, { "grad_norm": 0.5185824632644653, "learning_rate": 6.846924234831069e-05, "loss": 0.0537, "step": 41060 }, { "grad_norm": 0.49078014492988586, "learning_rate": 6.845387603206597e-05, "loss": 0.039, "step": 41070 }, { "grad_norm": 0.5500463843345642, "learning_rate": 6.843850769773312e-05, "loss": 0.0436, "step": 41080 }, { "grad_norm": 0.7089939117431641, "learning_rate": 6.842313734699278e-05, "loss": 0.0431, "step": 41090 }, { "grad_norm": 0.6553078293800354, "learning_rate": 6.840776498152584e-05, "loss": 0.0406, "step": 41100 }, { "grad_norm": 0.532368004322052, "learning_rate": 6.83923906030134e-05, "loss": 0.042, "step": 41110 }, { "grad_norm": 0.5635348558425903, "learning_rate": 6.837701421313677e-05, "loss": 0.0447, "step": 41120 }, { "grad_norm": 0.5320530533790588, "learning_rate": 6.83616358135775e-05, "loss": 0.042, "step": 41130 }, { "grad_norm": 0.442190945148468, "learning_rate": 6.834625540601734e-05, "loss": 0.0428, "step": 41140 }, { "grad_norm": 0.9061349630355835, "learning_rate": 6.833087299213829e-05, "loss": 0.0468, "step": 41150 }, { "grad_norm": 0.8201615810394287, "learning_rate": 6.83154885736225e-05, "loss": 0.0362, "step": 41160 }, { "grad_norm": 0.7202713489532471, "learning_rate": 6.830010215215244e-05, "loss": 0.0492, "step": 41170 }, { "grad_norm": 0.6529033780097961, "learning_rate": 6.828471372941072e-05, "loss": 0.0451, "step": 41180 }, { "grad_norm": 0.7894836068153381, "learning_rate": 6.826932330708018e-05, "loss": 0.0481, "step": 41190 }, { "grad_norm": 1.605133056640625, "learning_rate": 6.825393088684393e-05, "loss": 0.0465, "step": 41200 }, { "grad_norm": 0.613349437713623, "learning_rate": 6.823853647038524e-05, "loss": 0.0441, "step": 41210 }, { "grad_norm": 0.5749487280845642, "learning_rate": 6.822314005938762e-05, "loss": 0.0635, "step": 41220 }, { "grad_norm": 0.30115386843681335, "learning_rate": 6.820774165553479e-05, "loss": 0.0333, "step": 41230 }, { "grad_norm": 0.4423600435256958, "learning_rate": 6.819234126051072e-05, "loss": 0.0418, "step": 41240 }, { "grad_norm": 0.4572508931159973, "learning_rate": 6.817693887599956e-05, "loss": 0.0559, "step": 41250 }, { "grad_norm": 0.6760064363479614, "learning_rate": 6.816153450368568e-05, "loss": 0.0489, "step": 41260 }, { "grad_norm": 0.4733411967754364, "learning_rate": 6.814612814525372e-05, "loss": 0.0438, "step": 41270 }, { "grad_norm": 0.5284804701805115, "learning_rate": 6.813071980238846e-05, "loss": 0.0405, "step": 41280 }, { "grad_norm": 0.32610639929771423, "learning_rate": 6.811530947677492e-05, "loss": 0.0356, "step": 41290 }, { "grad_norm": 0.45765334367752075, "learning_rate": 6.809989717009839e-05, "loss": 0.0388, "step": 41300 }, { "grad_norm": 0.6852368116378784, "learning_rate": 6.808448288404431e-05, "loss": 0.0354, "step": 41310 }, { "grad_norm": 1.0775601863861084, "learning_rate": 6.806906662029838e-05, "loss": 0.0608, "step": 41320 }, { "grad_norm": 0.37077391147613525, "learning_rate": 6.805364838054647e-05, "loss": 0.0348, "step": 41330 }, { "grad_norm": 0.34415385127067566, "learning_rate": 6.803822816647474e-05, "loss": 0.0402, "step": 41340 }, { "grad_norm": 0.43032538890838623, "learning_rate": 6.802280597976949e-05, "loss": 0.0428, "step": 41350 }, { "grad_norm": 0.3912650942802429, "learning_rate": 6.800738182211727e-05, "loss": 0.0403, "step": 41360 }, { "grad_norm": 0.5019071698188782, "learning_rate": 6.799195569520485e-05, "loss": 0.0307, "step": 41370 }, { "grad_norm": 0.690303385257721, "learning_rate": 6.79765276007192e-05, "loss": 0.0464, "step": 41380 }, { "grad_norm": 0.4501653015613556, "learning_rate": 6.796109754034753e-05, "loss": 0.0482, "step": 41390 }, { "grad_norm": 0.5052114725112915, "learning_rate": 6.794566551577724e-05, "loss": 0.0516, "step": 41400 }, { "grad_norm": 0.6039555668830872, "learning_rate": 6.793023152869594e-05, "loss": 0.0372, "step": 41410 }, { "grad_norm": 0.639258861541748, "learning_rate": 6.791479558079147e-05, "loss": 0.0405, "step": 41420 }, { "grad_norm": 0.20329667627811432, "learning_rate": 6.789935767375191e-05, "loss": 0.0511, "step": 41430 }, { "grad_norm": 0.7217090725898743, "learning_rate": 6.78839178092655e-05, "loss": 0.0411, "step": 41440 }, { "grad_norm": 0.5644355416297913, "learning_rate": 6.786847598902072e-05, "loss": 0.0429, "step": 41450 }, { "grad_norm": 0.7029252052307129, "learning_rate": 6.785303221470627e-05, "loss": 0.0587, "step": 41460 }, { "grad_norm": 0.38191747665405273, "learning_rate": 6.783758648801108e-05, "loss": 0.033, "step": 41470 }, { "grad_norm": 0.3859906792640686, "learning_rate": 6.782213881062422e-05, "loss": 0.0299, "step": 41480 }, { "grad_norm": 0.4301736652851105, "learning_rate": 6.780668918423508e-05, "loss": 0.0484, "step": 41490 }, { "grad_norm": 0.6436577439308167, "learning_rate": 6.779123761053317e-05, "loss": 0.0469, "step": 41500 }, { "grad_norm": 0.6222675442695618, "learning_rate": 6.777578409120826e-05, "loss": 0.0422, "step": 41510 }, { "grad_norm": 0.7594345808029175, "learning_rate": 6.776032862795033e-05, "loss": 0.0503, "step": 41520 }, { "grad_norm": 0.6657999157905579, "learning_rate": 6.774487122244956e-05, "loss": 0.0581, "step": 41530 }, { "grad_norm": 0.48006191849708557, "learning_rate": 6.772941187639637e-05, "loss": 0.0617, "step": 41540 }, { "grad_norm": 0.5576595067977905, "learning_rate": 6.771395059148134e-05, "loss": 0.0424, "step": 41550 }, { "grad_norm": 0.5729661583900452, "learning_rate": 6.76984873693953e-05, "loss": 0.046, "step": 41560 }, { "grad_norm": 0.561429500579834, "learning_rate": 6.768302221182931e-05, "loss": 0.0398, "step": 41570 }, { "grad_norm": 0.38967278599739075, "learning_rate": 6.766755512047457e-05, "loss": 0.0423, "step": 41580 }, { "grad_norm": 0.5124437212944031, "learning_rate": 6.765208609702259e-05, "loss": 0.063, "step": 41590 }, { "grad_norm": 0.37477272748947144, "learning_rate": 6.763661514316499e-05, "loss": 0.0437, "step": 41600 }, { "grad_norm": 0.5240551829338074, "learning_rate": 6.76211422605937e-05, "loss": 0.0372, "step": 41610 }, { "grad_norm": 0.9348548650741577, "learning_rate": 6.760566745100076e-05, "loss": 0.0451, "step": 41620 }, { "grad_norm": 0.3800657093524933, "learning_rate": 6.75901907160785e-05, "loss": 0.049, "step": 41630 }, { "grad_norm": 0.6660665273666382, "learning_rate": 6.757471205751943e-05, "loss": 0.0545, "step": 41640 }, { "grad_norm": 0.5239313244819641, "learning_rate": 6.75592314770163e-05, "loss": 0.0431, "step": 41650 }, { "grad_norm": 0.48511311411857605, "learning_rate": 6.7543748976262e-05, "loss": 0.0402, "step": 41660 }, { "grad_norm": 0.5757350921630859, "learning_rate": 6.752826455694968e-05, "loss": 0.059, "step": 41670 }, { "grad_norm": 0.9599480628967285, "learning_rate": 6.751277822077271e-05, "loss": 0.0477, "step": 41680 }, { "grad_norm": 0.6549193859100342, "learning_rate": 6.749728996942465e-05, "loss": 0.0529, "step": 41690 }, { "grad_norm": 0.42295345664024353, "learning_rate": 6.748179980459924e-05, "loss": 0.0429, "step": 41700 }, { "grad_norm": 0.3428593575954437, "learning_rate": 6.746630772799052e-05, "loss": 0.0648, "step": 41710 }, { "grad_norm": 0.43183034658432007, "learning_rate": 6.745081374129262e-05, "loss": 0.0449, "step": 41720 }, { "grad_norm": 0.48934105038642883, "learning_rate": 6.743531784619998e-05, "loss": 0.0473, "step": 41730 }, { "grad_norm": 0.660232424736023, "learning_rate": 6.741982004440719e-05, "loss": 0.0499, "step": 41740 }, { "grad_norm": 0.4600193202495575, "learning_rate": 6.740432033760907e-05, "loss": 0.0475, "step": 41750 }, { "grad_norm": 0.8607430458068848, "learning_rate": 6.738881872750066e-05, "loss": 0.0473, "step": 41760 }, { "grad_norm": 0.5438328981399536, "learning_rate": 6.737331521577715e-05, "loss": 0.0428, "step": 41770 }, { "grad_norm": 0.641002357006073, "learning_rate": 6.735780980413403e-05, "loss": 0.0475, "step": 41780 }, { "grad_norm": 0.6402473449707031, "learning_rate": 6.734230249426692e-05, "loss": 0.0449, "step": 41790 }, { "grad_norm": 0.5979411602020264, "learning_rate": 6.732679328787168e-05, "loss": 0.0563, "step": 41800 }, { "grad_norm": 0.6565465927124023, "learning_rate": 6.731128218664438e-05, "loss": 0.0523, "step": 41810 }, { "grad_norm": 0.5847198963165283, "learning_rate": 6.729576919228129e-05, "loss": 0.0489, "step": 41820 }, { "grad_norm": 0.481851726770401, "learning_rate": 6.728025430647888e-05, "loss": 0.0495, "step": 41830 }, { "grad_norm": 0.4096800684928894, "learning_rate": 6.726473753093383e-05, "loss": 0.0393, "step": 41840 }, { "grad_norm": 0.7617120146751404, "learning_rate": 6.724921886734305e-05, "loss": 0.0388, "step": 41850 }, { "grad_norm": 0.44406193494796753, "learning_rate": 6.723369831740363e-05, "loss": 0.0416, "step": 41860 }, { "grad_norm": 0.9267803430557251, "learning_rate": 6.721817588281288e-05, "loss": 0.0581, "step": 41870 }, { "grad_norm": 0.3614042401313782, "learning_rate": 6.720265156526828e-05, "loss": 0.0459, "step": 41880 }, { "grad_norm": 0.5165209770202637, "learning_rate": 6.718712536646758e-05, "loss": 0.0346, "step": 41890 }, { "grad_norm": 0.42173242568969727, "learning_rate": 6.71715972881087e-05, "loss": 0.0375, "step": 41900 }, { "grad_norm": 0.6154616475105286, "learning_rate": 6.715606733188974e-05, "loss": 0.045, "step": 41910 }, { "grad_norm": 0.4581485390663147, "learning_rate": 6.714053549950909e-05, "loss": 0.0366, "step": 41920 }, { "grad_norm": 0.765896201133728, "learning_rate": 6.71250017926652e-05, "loss": 0.0422, "step": 41930 }, { "grad_norm": 0.5215998291969299, "learning_rate": 6.71094662130569e-05, "loss": 0.0365, "step": 41940 }, { "grad_norm": 0.6354628801345825, "learning_rate": 6.709392876238307e-05, "loss": 0.0419, "step": 41950 }, { "grad_norm": 0.4005032777786255, "learning_rate": 6.707838944234289e-05, "loss": 0.0448, "step": 41960 }, { "grad_norm": 0.4541216492652893, "learning_rate": 6.706284825463574e-05, "loss": 0.0386, "step": 41970 }, { "grad_norm": 0.4989551603794098, "learning_rate": 6.704730520096114e-05, "loss": 0.0433, "step": 41980 }, { "grad_norm": 0.34744778275489807, "learning_rate": 6.703176028301888e-05, "loss": 0.0378, "step": 41990 }, { "grad_norm": 0.3455585837364197, "learning_rate": 6.701621350250892e-05, "loss": 0.0413, "step": 42000 }, { "grad_norm": 0.4255872070789337, "learning_rate": 6.700066486113144e-05, "loss": 0.0502, "step": 42010 }, { "grad_norm": 0.4613243341445923, "learning_rate": 6.69851143605868e-05, "loss": 0.038, "step": 42020 }, { "grad_norm": 0.6496278643608093, "learning_rate": 6.696956200257561e-05, "loss": 0.0503, "step": 42030 }, { "grad_norm": 0.34890681505203247, "learning_rate": 6.695400778879863e-05, "loss": 0.0316, "step": 42040 }, { "grad_norm": 0.5357744693756104, "learning_rate": 6.693845172095683e-05, "loss": 0.0398, "step": 42050 }, { "grad_norm": 0.7095522880554199, "learning_rate": 6.692289380075142e-05, "loss": 0.0548, "step": 42060 }, { "grad_norm": 0.7514999508857727, "learning_rate": 6.690733402988379e-05, "loss": 0.042, "step": 42070 }, { "grad_norm": 0.3130633234977722, "learning_rate": 6.689177241005553e-05, "loss": 0.0383, "step": 42080 }, { "grad_norm": 0.4946473240852356, "learning_rate": 6.687620894296844e-05, "loss": 0.0428, "step": 42090 }, { "grad_norm": 0.36793890595436096, "learning_rate": 6.686064363032451e-05, "loss": 0.042, "step": 42100 }, { "grad_norm": 0.33893221616744995, "learning_rate": 6.684507647382596e-05, "loss": 0.0511, "step": 42110 }, { "grad_norm": 0.47625795006752014, "learning_rate": 6.682950747517513e-05, "loss": 0.0577, "step": 42120 }, { "grad_norm": 0.645612359046936, "learning_rate": 6.68139366360747e-05, "loss": 0.0569, "step": 42130 }, { "grad_norm": 0.6559462547302246, "learning_rate": 6.679836395822744e-05, "loss": 0.0503, "step": 42140 }, { "grad_norm": 0.7306621670722961, "learning_rate": 6.678278944333633e-05, "loss": 0.0436, "step": 42150 }, { "grad_norm": 0.5519349575042725, "learning_rate": 6.676721309310462e-05, "loss": 0.0456, "step": 42160 }, { "grad_norm": 0.486824095249176, "learning_rate": 6.67516349092357e-05, "loss": 0.0555, "step": 42170 }, { "grad_norm": 0.45890986919403076, "learning_rate": 6.673605489343315e-05, "loss": 0.0457, "step": 42180 }, { "grad_norm": 0.6550080180168152, "learning_rate": 6.672047304740081e-05, "loss": 0.0491, "step": 42190 }, { "grad_norm": 0.42314091324806213, "learning_rate": 6.670488937284268e-05, "loss": 0.0475, "step": 42200 }, { "grad_norm": 0.5048230290412903, "learning_rate": 6.668930387146294e-05, "loss": 0.0514, "step": 42210 }, { "grad_norm": 0.5139121413230896, "learning_rate": 6.667371654496605e-05, "loss": 0.0394, "step": 42220 }, { "grad_norm": 0.6493424773216248, "learning_rate": 6.665812739505659e-05, "loss": 0.036, "step": 42230 }, { "grad_norm": 0.47495293617248535, "learning_rate": 6.664253642343935e-05, "loss": 0.0495, "step": 42240 }, { "grad_norm": 0.3860386610031128, "learning_rate": 6.662694363181935e-05, "loss": 0.053, "step": 42250 }, { "grad_norm": 0.5049498677253723, "learning_rate": 6.66113490219018e-05, "loss": 0.043, "step": 42260 }, { "grad_norm": 0.7174449563026428, "learning_rate": 6.659575259539207e-05, "loss": 0.0448, "step": 42270 }, { "grad_norm": 0.7194159626960754, "learning_rate": 6.65801543539958e-05, "loss": 0.0456, "step": 42280 }, { "grad_norm": 0.5476118922233582, "learning_rate": 6.65645542994188e-05, "loss": 0.0386, "step": 42290 }, { "grad_norm": 0.5814276337623596, "learning_rate": 6.654895243336702e-05, "loss": 0.0408, "step": 42300 }, { "grad_norm": 0.3316667377948761, "learning_rate": 6.65333487575467e-05, "loss": 0.0384, "step": 42310 }, { "grad_norm": 0.5290560126304626, "learning_rate": 6.651774327366422e-05, "loss": 0.0383, "step": 42320 }, { "grad_norm": 0.7002332210540771, "learning_rate": 6.650213598342613e-05, "loss": 0.0485, "step": 42330 }, { "grad_norm": 0.4581928849220276, "learning_rate": 6.648652688853931e-05, "loss": 0.0469, "step": 42340 }, { "grad_norm": 0.48807668685913086, "learning_rate": 6.647091599071066e-05, "loss": 0.0405, "step": 42350 }, { "grad_norm": 0.42490655183792114, "learning_rate": 6.645530329164743e-05, "loss": 0.0396, "step": 42360 }, { "grad_norm": 0.5255283117294312, "learning_rate": 6.643968879305694e-05, "loss": 0.0438, "step": 42370 }, { "grad_norm": 0.6174297332763672, "learning_rate": 6.642407249664683e-05, "loss": 0.0552, "step": 42380 }, { "grad_norm": 0.6854400634765625, "learning_rate": 6.640845440412483e-05, "loss": 0.0476, "step": 42390 }, { "grad_norm": 0.7515572309494019, "learning_rate": 6.639283451719893e-05, "loss": 0.0505, "step": 42400 }, { "grad_norm": 0.5224952101707458, "learning_rate": 6.637721283757729e-05, "loss": 0.0433, "step": 42410 }, { "grad_norm": 0.5033630132675171, "learning_rate": 6.636158936696828e-05, "loss": 0.0491, "step": 42420 }, { "grad_norm": 0.6237773299217224, "learning_rate": 6.634596410708047e-05, "loss": 0.0485, "step": 42430 }, { "grad_norm": 0.44330453872680664, "learning_rate": 6.633033705962257e-05, "loss": 0.0346, "step": 42440 }, { "grad_norm": 0.31616440415382385, "learning_rate": 6.631470822630359e-05, "loss": 0.0531, "step": 42450 }, { "grad_norm": 0.5860431790351868, "learning_rate": 6.629907760883263e-05, "loss": 0.0388, "step": 42460 }, { "grad_norm": 0.4649695158004761, "learning_rate": 6.628344520891905e-05, "loss": 0.0358, "step": 42470 }, { "grad_norm": 0.7211846709251404, "learning_rate": 6.62678110282724e-05, "loss": 0.0369, "step": 42480 }, { "grad_norm": 0.5432628393173218, "learning_rate": 6.625217506860237e-05, "loss": 0.0397, "step": 42490 }, { "grad_norm": 0.49168360233306885, "learning_rate": 6.623653733161892e-05, "loss": 0.0403, "step": 42500 }, { "grad_norm": 0.551996648311615, "learning_rate": 6.622089781903217e-05, "loss": 0.051, "step": 42510 }, { "grad_norm": 0.5032615661621094, "learning_rate": 6.620525653255241e-05, "loss": 0.0503, "step": 42520 }, { "grad_norm": 0.39995092153549194, "learning_rate": 6.618961347389015e-05, "loss": 0.044, "step": 42530 }, { "grad_norm": 0.3370911180973053, "learning_rate": 6.617396864475613e-05, "loss": 0.0426, "step": 42540 }, { "grad_norm": 0.6885951161384583, "learning_rate": 6.61583220468612e-05, "loss": 0.0336, "step": 42550 }, { "grad_norm": 0.393963098526001, "learning_rate": 6.614267368191645e-05, "loss": 0.0411, "step": 42560 }, { "grad_norm": 0.6010159254074097, "learning_rate": 6.61270235516332e-05, "loss": 0.036, "step": 42570 }, { "grad_norm": 0.6699177026748657, "learning_rate": 6.611137165772287e-05, "loss": 0.0369, "step": 42580 }, { "grad_norm": 0.40856683254241943, "learning_rate": 6.60957180018972e-05, "loss": 0.0326, "step": 42590 }, { "grad_norm": 0.6205872893333435, "learning_rate": 6.608006258586797e-05, "loss": 0.0292, "step": 42600 }, { "grad_norm": 0.3433400094509125, "learning_rate": 6.606440541134731e-05, "loss": 0.035, "step": 42610 }, { "grad_norm": 0.49841538071632385, "learning_rate": 6.60487464800474e-05, "loss": 0.041, "step": 42620 }, { "grad_norm": 0.36606961488723755, "learning_rate": 6.603308579368071e-05, "loss": 0.0389, "step": 42630 }, { "grad_norm": 0.44026970863342285, "learning_rate": 6.601742335395987e-05, "loss": 0.0426, "step": 42640 }, { "grad_norm": 0.5204870104789734, "learning_rate": 6.600175916259769e-05, "loss": 0.0478, "step": 42650 }, { "grad_norm": 0.7216585278511047, "learning_rate": 6.598609322130718e-05, "loss": 0.0749, "step": 42660 }, { "grad_norm": 0.5334904789924622, "learning_rate": 6.597042553180155e-05, "loss": 0.0428, "step": 42670 }, { "grad_norm": 0.5474067330360413, "learning_rate": 6.59547560957942e-05, "loss": 0.0348, "step": 42680 }, { "grad_norm": 0.3895580470561981, "learning_rate": 6.59390849149987e-05, "loss": 0.0538, "step": 42690 }, { "grad_norm": 0.3380317687988281, "learning_rate": 6.592341199112886e-05, "loss": 0.0415, "step": 42700 }, { "grad_norm": 0.706634521484375, "learning_rate": 6.590773732589861e-05, "loss": 0.0527, "step": 42710 }, { "grad_norm": 0.5586513876914978, "learning_rate": 6.589206092102213e-05, "loss": 0.0445, "step": 42720 }, { "grad_norm": 0.4018407464027405, "learning_rate": 6.587638277821375e-05, "loss": 0.0464, "step": 42730 }, { "grad_norm": 0.4252001643180847, "learning_rate": 6.586070289918803e-05, "loss": 0.0329, "step": 42740 }, { "grad_norm": 0.4396626651287079, "learning_rate": 6.584502128565968e-05, "loss": 0.0393, "step": 42750 }, { "grad_norm": 0.49311473965644836, "learning_rate": 6.582933793934363e-05, "loss": 0.0452, "step": 42760 }, { "grad_norm": 0.6602224707603455, "learning_rate": 6.581365286195498e-05, "loss": 0.0388, "step": 42770 }, { "grad_norm": 0.5007394552230835, "learning_rate": 6.579796605520903e-05, "loss": 0.0552, "step": 42780 }, { "grad_norm": 0.7997175455093384, "learning_rate": 6.578227752082127e-05, "loss": 0.0607, "step": 42790 }, { "grad_norm": 0.5479656457901001, "learning_rate": 6.576658726050735e-05, "loss": 0.0467, "step": 42800 }, { "grad_norm": 0.37798821926116943, "learning_rate": 6.575089527598316e-05, "loss": 0.0486, "step": 42810 }, { "grad_norm": 0.6089691519737244, "learning_rate": 6.573520156896476e-05, "loss": 0.0667, "step": 42820 }, { "grad_norm": 0.4655319154262543, "learning_rate": 6.571950614116835e-05, "loss": 0.0446, "step": 42830 }, { "grad_norm": 0.44995537400245667, "learning_rate": 6.570380899431039e-05, "loss": 0.0588, "step": 42840 }, { "grad_norm": 0.5153194069862366, "learning_rate": 6.568811013010749e-05, "loss": 0.0487, "step": 42850 }, { "grad_norm": 0.32563814520835876, "learning_rate": 6.567240955027645e-05, "loss": 0.0408, "step": 42860 }, { "grad_norm": 0.651756227016449, "learning_rate": 6.565670725653427e-05, "loss": 0.0427, "step": 42870 }, { "grad_norm": 0.41674211621284485, "learning_rate": 6.564100325059812e-05, "loss": 0.0476, "step": 42880 }, { "grad_norm": 0.6474977731704712, "learning_rate": 6.562529753418538e-05, "loss": 0.0482, "step": 42890 }, { "grad_norm": 1.1234726905822754, "learning_rate": 6.56095901090136e-05, "loss": 0.0565, "step": 42900 }, { "grad_norm": 0.4349644184112549, "learning_rate": 6.55938809768005e-05, "loss": 0.0521, "step": 42910 }, { "grad_norm": 0.5181159377098083, "learning_rate": 6.557817013926402e-05, "loss": 0.0462, "step": 42920 }, { "grad_norm": 0.8893530368804932, "learning_rate": 6.556245759812227e-05, "loss": 0.0601, "step": 42930 }, { "grad_norm": 0.6156738996505737, "learning_rate": 6.554674335509357e-05, "loss": 0.0559, "step": 42940 }, { "grad_norm": 0.4728456735610962, "learning_rate": 6.553102741189638e-05, "loss": 0.0333, "step": 42950 }, { "grad_norm": 0.6273009777069092, "learning_rate": 6.55153097702494e-05, "loss": 0.0444, "step": 42960 }, { "grad_norm": 0.5430031418800354, "learning_rate": 6.549959043187146e-05, "loss": 0.0373, "step": 42970 }, { "grad_norm": 0.7350622415542603, "learning_rate": 6.54838693984816e-05, "loss": 0.0424, "step": 42980 }, { "grad_norm": 0.5186254978179932, "learning_rate": 6.546814667179909e-05, "loss": 0.0449, "step": 42990 }, { "grad_norm": 0.371356338262558, "learning_rate": 6.545242225354328e-05, "loss": 0.044, "step": 43000 }, { "grad_norm": 0.8548389673233032, "learning_rate": 6.543669614543383e-05, "loss": 0.0598, "step": 43010 }, { "grad_norm": 0.7039634585380554, "learning_rate": 6.542096834919049e-05, "loss": 0.0381, "step": 43020 }, { "grad_norm": 0.414620965719223, "learning_rate": 6.540523886653324e-05, "loss": 0.0338, "step": 43030 }, { "grad_norm": 0.6783522367477417, "learning_rate": 6.538950769918222e-05, "loss": 0.0498, "step": 43040 }, { "grad_norm": 0.5787374973297119, "learning_rate": 6.537377484885779e-05, "loss": 0.0461, "step": 43050 }, { "grad_norm": 0.5527325868606567, "learning_rate": 6.535804031728045e-05, "loss": 0.0321, "step": 43060 }, { "grad_norm": 0.4724065065383911, "learning_rate": 6.534230410617092e-05, "loss": 0.0393, "step": 43070 }, { "grad_norm": 0.5644825100898743, "learning_rate": 6.532656621725007e-05, "loss": 0.0427, "step": 43080 }, { "grad_norm": 0.7711220383644104, "learning_rate": 6.531082665223898e-05, "loss": 0.0456, "step": 43090 }, { "grad_norm": 0.7048136591911316, "learning_rate": 6.529508541285889e-05, "loss": 0.0534, "step": 43100 }, { "grad_norm": 0.5389816761016846, "learning_rate": 6.527934250083128e-05, "loss": 0.0448, "step": 43110 }, { "grad_norm": 0.5059049725532532, "learning_rate": 6.526359791787775e-05, "loss": 0.0498, "step": 43120 }, { "grad_norm": 0.4494723975658417, "learning_rate": 6.524785166572009e-05, "loss": 0.0367, "step": 43130 }, { "grad_norm": 0.5238962769508362, "learning_rate": 6.523210374608029e-05, "loss": 0.0373, "step": 43140 }, { "grad_norm": 0.5218603610992432, "learning_rate": 6.521635416068054e-05, "loss": 0.035, "step": 43150 }, { "grad_norm": 0.44878312945365906, "learning_rate": 6.520060291124317e-05, "loss": 0.0353, "step": 43160 }, { "grad_norm": 0.6820785999298096, "learning_rate": 6.518484999949072e-05, "loss": 0.0446, "step": 43170 }, { "grad_norm": 0.518032431602478, "learning_rate": 6.516909542714591e-05, "loss": 0.0521, "step": 43180 }, { "grad_norm": 0.32958894968032837, "learning_rate": 6.515333919593161e-05, "loss": 0.0368, "step": 43190 }, { "grad_norm": 0.4929386079311371, "learning_rate": 6.513758130757094e-05, "loss": 0.0427, "step": 43200 }, { "grad_norm": 0.41983914375305176, "learning_rate": 6.512182176378713e-05, "loss": 0.041, "step": 43210 }, { "grad_norm": 0.6520940065383911, "learning_rate": 6.510606056630362e-05, "loss": 0.0622, "step": 43220 }, { "grad_norm": 0.3834204077720642, "learning_rate": 6.509029771684405e-05, "loss": 0.0365, "step": 43230 }, { "grad_norm": 0.4615090489387512, "learning_rate": 6.50745332171322e-05, "loss": 0.041, "step": 43240 }, { "grad_norm": 0.5553041100502014, "learning_rate": 6.505876706889207e-05, "loss": 0.0463, "step": 43250 }, { "grad_norm": 0.5095438361167908, "learning_rate": 6.504299927384781e-05, "loss": 0.0367, "step": 43260 }, { "grad_norm": 0.5549650192260742, "learning_rate": 6.502722983372378e-05, "loss": 0.0461, "step": 43270 }, { "grad_norm": 0.5141852498054504, "learning_rate": 6.501145875024446e-05, "loss": 0.0392, "step": 43280 }, { "grad_norm": 0.5335288643836975, "learning_rate": 6.499568602513459e-05, "loss": 0.0443, "step": 43290 }, { "grad_norm": 0.5120536684989929, "learning_rate": 6.497991166011903e-05, "loss": 0.0308, "step": 43300 }, { "grad_norm": 0.40067994594573975, "learning_rate": 6.49641356569229e-05, "loss": 0.0402, "step": 43310 }, { "grad_norm": 0.36693188548088074, "learning_rate": 6.494835801727135e-05, "loss": 0.0401, "step": 43320 }, { "grad_norm": 0.4912545084953308, "learning_rate": 6.493257874288986e-05, "loss": 0.0423, "step": 43330 }, { "grad_norm": 0.8194305300712585, "learning_rate": 6.491679783550399e-05, "loss": 0.0381, "step": 43340 }, { "grad_norm": 0.5227460265159607, "learning_rate": 6.490101529683957e-05, "loss": 0.0378, "step": 43350 }, { "grad_norm": 1.0071382522583008, "learning_rate": 6.488523112862248e-05, "loss": 0.0528, "step": 43360 }, { "grad_norm": 0.5499758720397949, "learning_rate": 6.486944533257891e-05, "loss": 0.0292, "step": 43370 }, { "grad_norm": 0.7647159099578857, "learning_rate": 6.485365791043517e-05, "loss": 0.0494, "step": 43380 }, { "grad_norm": 0.33416298031806946, "learning_rate": 6.483786886391772e-05, "loss": 0.044, "step": 43390 }, { "grad_norm": 0.5584033727645874, "learning_rate": 6.482207819475323e-05, "loss": 0.0425, "step": 43400 }, { "grad_norm": 0.9074093699455261, "learning_rate": 6.480628590466857e-05, "loss": 0.038, "step": 43410 }, { "grad_norm": 0.7602803707122803, "learning_rate": 6.479049199539074e-05, "loss": 0.0455, "step": 43420 }, { "grad_norm": 0.6405401229858398, "learning_rate": 6.477469646864693e-05, "loss": 0.0529, "step": 43430 }, { "grad_norm": 0.5692764520645142, "learning_rate": 6.475889932616454e-05, "loss": 0.0482, "step": 43440 }, { "grad_norm": 0.5099360346794128, "learning_rate": 6.474310056967111e-05, "loss": 0.036, "step": 43450 }, { "grad_norm": 1.9128302335739136, "learning_rate": 6.472730020089437e-05, "loss": 0.0473, "step": 43460 }, { "grad_norm": 0.6497701406478882, "learning_rate": 6.471149822156222e-05, "loss": 0.0386, "step": 43470 }, { "grad_norm": 0.6028927564620972, "learning_rate": 6.469569463340275e-05, "loss": 0.0428, "step": 43480 }, { "grad_norm": 0.44558480381965637, "learning_rate": 6.467988943814418e-05, "loss": 0.039, "step": 43490 }, { "grad_norm": 0.4775140881538391, "learning_rate": 6.4664082637515e-05, "loss": 0.0379, "step": 43500 }, { "grad_norm": 0.7557408809661865, "learning_rate": 6.464827423324377e-05, "loss": 0.0339, "step": 43510 }, { "grad_norm": 0.6953355669975281, "learning_rate": 6.463246422705931e-05, "loss": 0.0694, "step": 43520 }, { "grad_norm": 0.4059837758541107, "learning_rate": 6.461665262069055e-05, "loss": 0.0492, "step": 43530 }, { "grad_norm": 0.4424385726451874, "learning_rate": 6.460083941586665e-05, "loss": 0.0415, "step": 43540 }, { "grad_norm": 0.5041056871414185, "learning_rate": 6.45850246143169e-05, "loss": 0.0425, "step": 43550 }, { "grad_norm": 0.6774024963378906, "learning_rate": 6.456920821777077e-05, "loss": 0.0729, "step": 43560 }, { "grad_norm": 0.2997308671474457, "learning_rate": 6.455339022795796e-05, "loss": 0.0324, "step": 43570 }, { "grad_norm": 0.5179464221000671, "learning_rate": 6.453757064660825e-05, "loss": 0.0394, "step": 43580 }, { "grad_norm": 0.30417537689208984, "learning_rate": 6.452174947545169e-05, "loss": 0.0327, "step": 43590 }, { "grad_norm": 0.9001305103302002, "learning_rate": 6.450592671621842e-05, "loss": 0.0721, "step": 43600 }, { "grad_norm": 0.5535735487937927, "learning_rate": 6.449010237063882e-05, "loss": 0.0394, "step": 43610 }, { "grad_norm": 0.5412998795509338, "learning_rate": 6.447427644044343e-05, "loss": 0.0406, "step": 43620 }, { "grad_norm": 0.3862716257572174, "learning_rate": 6.44584489273629e-05, "loss": 0.0314, "step": 43630 }, { "grad_norm": 0.9255921244621277, "learning_rate": 6.444261983312817e-05, "loss": 0.0525, "step": 43640 }, { "grad_norm": 0.6295282244682312, "learning_rate": 6.442678915947023e-05, "loss": 0.0475, "step": 43650 }, { "grad_norm": 1.3995131254196167, "learning_rate": 6.441095690812032e-05, "loss": 0.0672, "step": 43660 }, { "grad_norm": 0.5805985331535339, "learning_rate": 6.439512308080982e-05, "loss": 0.0522, "step": 43670 }, { "grad_norm": 0.3936976194381714, "learning_rate": 6.43792876792703e-05, "loss": 0.051, "step": 43680 }, { "grad_norm": 0.45658713579177856, "learning_rate": 6.43634507052335e-05, "loss": 0.045, "step": 43690 }, { "grad_norm": 0.46290987730026245, "learning_rate": 6.434761216043135e-05, "loss": 0.047, "step": 43700 }, { "grad_norm": 0.5488128066062927, "learning_rate": 6.433177204659589e-05, "loss": 0.0462, "step": 43710 }, { "grad_norm": 0.43138599395751953, "learning_rate": 6.431593036545938e-05, "loss": 0.0399, "step": 43720 }, { "grad_norm": 0.29193928837776184, "learning_rate": 6.430008711875426e-05, "loss": 0.0381, "step": 43730 }, { "grad_norm": 0.5825775265693665, "learning_rate": 6.42842423082131e-05, "loss": 0.0406, "step": 43740 }, { "grad_norm": 0.4596838355064392, "learning_rate": 6.426839593556868e-05, "loss": 0.035, "step": 43750 }, { "grad_norm": 0.6722753047943115, "learning_rate": 6.425254800255392e-05, "loss": 0.0377, "step": 43760 }, { "grad_norm": 0.4142323136329651, "learning_rate": 6.423669851090196e-05, "loss": 0.041, "step": 43770 }, { "grad_norm": 0.43763238191604614, "learning_rate": 6.422084746234604e-05, "loss": 0.0396, "step": 43780 }, { "grad_norm": 1.0890623331069946, "learning_rate": 6.420499485861965e-05, "loss": 0.0372, "step": 43790 }, { "grad_norm": 0.6454383134841919, "learning_rate": 6.418914070145636e-05, "loss": 0.0536, "step": 43800 }, { "grad_norm": 0.6055180430412292, "learning_rate": 6.417328499258996e-05, "loss": 0.0413, "step": 43810 }, { "grad_norm": 0.4377211332321167, "learning_rate": 6.415742773375446e-05, "loss": 0.0494, "step": 43820 }, { "grad_norm": 0.6056092381477356, "learning_rate": 6.414156892668393e-05, "loss": 0.0414, "step": 43830 }, { "grad_norm": 0.625998318195343, "learning_rate": 6.412570857311267e-05, "loss": 0.0455, "step": 43840 }, { "grad_norm": 0.4699645936489105, "learning_rate": 6.410984667477518e-05, "loss": 0.0454, "step": 43850 }, { "grad_norm": 0.6239978075027466, "learning_rate": 6.409398323340607e-05, "loss": 0.0385, "step": 43860 }, { "grad_norm": 0.5603727698326111, "learning_rate": 6.407811825074012e-05, "loss": 0.0456, "step": 43870 }, { "grad_norm": 0.8368421792984009, "learning_rate": 6.406225172851234e-05, "loss": 0.0595, "step": 43880 }, { "grad_norm": 0.35894137620925903, "learning_rate": 6.404638366845786e-05, "loss": 0.0334, "step": 43890 }, { "grad_norm": 0.33045485615730286, "learning_rate": 6.403051407231196e-05, "loss": 0.0433, "step": 43900 }, { "grad_norm": 0.4733526110649109, "learning_rate": 6.401464294181016e-05, "loss": 0.0406, "step": 43910 }, { "grad_norm": 0.5761218070983887, "learning_rate": 6.399877027868808e-05, "loss": 0.0387, "step": 43920 }, { "grad_norm": 0.5701918005943298, "learning_rate": 6.39828960846815e-05, "loss": 0.0501, "step": 43930 }, { "grad_norm": 0.35749438405036926, "learning_rate": 6.396702036152644e-05, "loss": 0.0345, "step": 43940 }, { "grad_norm": 0.35402587056159973, "learning_rate": 6.395114311095905e-05, "loss": 0.0284, "step": 43950 }, { "grad_norm": 0.7060163021087646, "learning_rate": 6.393526433471559e-05, "loss": 0.0533, "step": 43960 }, { "grad_norm": 0.8273815512657166, "learning_rate": 6.391938403453262e-05, "loss": 0.0348, "step": 43970 }, { "grad_norm": 0.4951455891132355, "learning_rate": 6.390350221214671e-05, "loss": 0.0525, "step": 43980 }, { "grad_norm": 0.5445528626441956, "learning_rate": 6.38876188692947e-05, "loss": 0.0433, "step": 43990 }, { "grad_norm": 0.5995427966117859, "learning_rate": 6.387173400771358e-05, "loss": 0.0464, "step": 44000 }, { "grad_norm": 0.5877513289451599, "learning_rate": 6.385584762914047e-05, "loss": 0.0441, "step": 44010 }, { "grad_norm": 0.7529405355453491, "learning_rate": 6.38399597353127e-05, "loss": 0.0389, "step": 44020 }, { "grad_norm": 0.8031070828437805, "learning_rate": 6.382407032796775e-05, "loss": 0.0502, "step": 44030 }, { "grad_norm": 0.8800229430198669, "learning_rate": 6.380817940884325e-05, "loss": 0.0457, "step": 44040 }, { "grad_norm": 0.5843180418014526, "learning_rate": 6.379228697967702e-05, "loss": 0.0678, "step": 44050 }, { "grad_norm": 0.4556906521320343, "learning_rate": 6.3776393042207e-05, "loss": 0.0355, "step": 44060 }, { "grad_norm": 0.4601978063583374, "learning_rate": 6.376049759817137e-05, "loss": 0.0427, "step": 44070 }, { "grad_norm": 0.656682014465332, "learning_rate": 6.37446006493084e-05, "loss": 0.0527, "step": 44080 }, { "grad_norm": 0.4983794093132019, "learning_rate": 6.372870219735656e-05, "loss": 0.041, "step": 44090 }, { "grad_norm": 0.38116025924682617, "learning_rate": 6.371280224405452e-05, "loss": 0.047, "step": 44100 }, { "grad_norm": 0.48705190420150757, "learning_rate": 6.369690079114101e-05, "loss": 0.0368, "step": 44110 }, { "grad_norm": 0.6969823837280273, "learning_rate": 6.368099784035504e-05, "loss": 0.0414, "step": 44120 }, { "grad_norm": 0.3719163239002228, "learning_rate": 6.366509339343572e-05, "loss": 0.0428, "step": 44130 }, { "grad_norm": 0.5127790570259094, "learning_rate": 6.364918745212232e-05, "loss": 0.037, "step": 44140 }, { "grad_norm": 0.4395209848880768, "learning_rate": 6.363328001815431e-05, "loss": 0.0425, "step": 44150 }, { "grad_norm": 0.4718007445335388, "learning_rate": 6.361737109327128e-05, "loss": 0.0408, "step": 44160 }, { "grad_norm": 0.5705053210258484, "learning_rate": 6.360146067921301e-05, "loss": 0.0546, "step": 44170 }, { "grad_norm": 0.4083486795425415, "learning_rate": 6.35855487777195e-05, "loss": 0.0525, "step": 44180 }, { "grad_norm": 0.5647221207618713, "learning_rate": 6.356963539053076e-05, "loss": 0.0363, "step": 44190 }, { "grad_norm": 0.36919406056404114, "learning_rate": 6.355372051938711e-05, "loss": 0.0388, "step": 44200 }, { "grad_norm": 0.7648324966430664, "learning_rate": 6.353780416602894e-05, "loss": 0.051, "step": 44210 }, { "grad_norm": 0.648564875125885, "learning_rate": 6.352188633219689e-05, "loss": 0.0683, "step": 44220 }, { "grad_norm": 0.680253267288208, "learning_rate": 6.350596701963166e-05, "loss": 0.0591, "step": 44230 }, { "grad_norm": 0.6043264865875244, "learning_rate": 6.349004623007419e-05, "loss": 0.041, "step": 44240 }, { "grad_norm": 0.575336217880249, "learning_rate": 6.347412396526555e-05, "loss": 0.0415, "step": 44250 }, { "grad_norm": 0.4301414489746094, "learning_rate": 6.345820022694696e-05, "loss": 0.0427, "step": 44260 }, { "grad_norm": 0.6433024406433105, "learning_rate": 6.344227501685984e-05, "loss": 0.032, "step": 44270 }, { "grad_norm": 0.5147823095321655, "learning_rate": 6.342634833674572e-05, "loss": 0.0514, "step": 44280 }, { "grad_norm": 0.47004833817481995, "learning_rate": 6.341042018834635e-05, "loss": 0.0353, "step": 44290 }, { "grad_norm": 0.3952903151512146, "learning_rate": 6.339449057340359e-05, "loss": 0.0472, "step": 44300 }, { "grad_norm": 0.506009042263031, "learning_rate": 6.337855949365945e-05, "loss": 0.0371, "step": 44310 }, { "grad_norm": 0.8895264863967896, "learning_rate": 6.336262695085619e-05, "loss": 0.0386, "step": 44320 }, { "grad_norm": 0.4917137920856476, "learning_rate": 6.334669294673612e-05, "loss": 0.0349, "step": 44330 }, { "grad_norm": 0.2676810026168823, "learning_rate": 6.333075748304179e-05, "loss": 0.0273, "step": 44340 }, { "grad_norm": 0.2936895787715912, "learning_rate": 6.331482056151585e-05, "loss": 0.0318, "step": 44350 }, { "grad_norm": 0.3961887061595917, "learning_rate": 6.329888218390117e-05, "loss": 0.0317, "step": 44360 }, { "grad_norm": 0.46938902139663696, "learning_rate": 6.328294235194072e-05, "loss": 0.0343, "step": 44370 }, { "grad_norm": 0.3821921944618225, "learning_rate": 6.326700106737765e-05, "loss": 0.0412, "step": 44380 }, { "grad_norm": 0.4947367310523987, "learning_rate": 6.325105833195531e-05, "loss": 0.0264, "step": 44390 }, { "grad_norm": 0.5610930919647217, "learning_rate": 6.323511414741715e-05, "loss": 0.0459, "step": 44400 }, { "grad_norm": 0.558598518371582, "learning_rate": 6.321916851550678e-05, "loss": 0.0435, "step": 44410 }, { "grad_norm": 0.5227365493774414, "learning_rate": 6.320322143796806e-05, "loss": 0.0391, "step": 44420 }, { "grad_norm": 0.4253247082233429, "learning_rate": 6.318727291654488e-05, "loss": 0.0433, "step": 44430 }, { "grad_norm": 0.4083815813064575, "learning_rate": 6.317132295298134e-05, "loss": 0.0381, "step": 44440 }, { "grad_norm": 0.48256880044937134, "learning_rate": 6.315537154902173e-05, "loss": 0.0292, "step": 44450 }, { "grad_norm": 0.43131983280181885, "learning_rate": 6.313941870641048e-05, "loss": 0.0353, "step": 44460 }, { "grad_norm": 0.5510905385017395, "learning_rate": 6.312346442689214e-05, "loss": 0.0498, "step": 44470 }, { "grad_norm": 0.6570072770118713, "learning_rate": 6.310750871221147e-05, "loss": 0.0436, "step": 44480 }, { "grad_norm": 0.33451640605926514, "learning_rate": 6.309155156411335e-05, "loss": 0.0234, "step": 44490 }, { "grad_norm": 0.546701192855835, "learning_rate": 6.307559298434284e-05, "loss": 0.0395, "step": 44500 }, { "grad_norm": 0.598980724811554, "learning_rate": 6.305963297464512e-05, "loss": 0.0398, "step": 44510 }, { "grad_norm": 0.6056735515594482, "learning_rate": 6.304367153676561e-05, "loss": 0.0485, "step": 44520 }, { "grad_norm": 0.7089329361915588, "learning_rate": 6.302770867244978e-05, "loss": 0.0406, "step": 44530 }, { "grad_norm": 0.508097767829895, "learning_rate": 6.301174438344328e-05, "loss": 0.0355, "step": 44540 }, { "grad_norm": 0.4623335003852844, "learning_rate": 6.299577867149202e-05, "loss": 0.0357, "step": 44550 }, { "grad_norm": 0.7796452045440674, "learning_rate": 6.297981153834192e-05, "loss": 0.0581, "step": 44560 }, { "grad_norm": 0.579284131526947, "learning_rate": 6.296384298573916e-05, "loss": 0.0489, "step": 44570 }, { "grad_norm": 0.3395442068576813, "learning_rate": 6.294787301543001e-05, "loss": 0.0375, "step": 44580 }, { "grad_norm": 1.3526713848114014, "learning_rate": 6.293190162916095e-05, "loss": 0.0418, "step": 44590 }, { "grad_norm": 0.46552109718322754, "learning_rate": 6.291592882867855e-05, "loss": 0.0343, "step": 44600 }, { "grad_norm": 0.5975745916366577, "learning_rate": 6.28999546157296e-05, "loss": 0.0483, "step": 44610 }, { "grad_norm": 0.5740965008735657, "learning_rate": 6.288397899206102e-05, "loss": 0.0395, "step": 44620 }, { "grad_norm": 0.47641250491142273, "learning_rate": 6.286800195941984e-05, "loss": 0.0491, "step": 44630 }, { "grad_norm": 0.6035071611404419, "learning_rate": 6.285202351955334e-05, "loss": 0.0473, "step": 44640 }, { "grad_norm": 0.5857117772102356, "learning_rate": 6.283604367420887e-05, "loss": 0.0278, "step": 44650 }, { "grad_norm": 0.489309161901474, "learning_rate": 6.282006242513394e-05, "loss": 0.0384, "step": 44660 }, { "grad_norm": 0.3791789710521698, "learning_rate": 6.280407977407628e-05, "loss": 0.0451, "step": 44670 }, { "grad_norm": 0.23192337155342102, "learning_rate": 6.27880957227837e-05, "loss": 0.0414, "step": 44680 }, { "grad_norm": 1.0653886795043945, "learning_rate": 6.27721102730042e-05, "loss": 0.0448, "step": 44690 }, { "grad_norm": 0.5251147747039795, "learning_rate": 6.275612342648592e-05, "loss": 0.0463, "step": 44700 }, { "grad_norm": 0.7459490895271301, "learning_rate": 6.274013518497716e-05, "loss": 0.0335, "step": 44710 }, { "grad_norm": 0.579126238822937, "learning_rate": 6.272414555022636e-05, "loss": 0.0373, "step": 44720 }, { "grad_norm": 0.45373639464378357, "learning_rate": 6.270815452398215e-05, "loss": 0.0432, "step": 44730 }, { "grad_norm": 0.36837849020957947, "learning_rate": 6.269216210799326e-05, "loss": 0.0607, "step": 44740 }, { "grad_norm": 0.37148386240005493, "learning_rate": 6.26761683040086e-05, "loss": 0.0344, "step": 44750 }, { "grad_norm": 0.5533735156059265, "learning_rate": 6.266017311377723e-05, "loss": 0.0433, "step": 44760 }, { "grad_norm": 0.4175671339035034, "learning_rate": 6.264417653904839e-05, "loss": 0.0351, "step": 44770 }, { "grad_norm": 0.38530758023262024, "learning_rate": 6.262817858157139e-05, "loss": 0.0351, "step": 44780 }, { "grad_norm": 0.5466518998146057, "learning_rate": 6.261217924309576e-05, "loss": 0.0404, "step": 44790 }, { "grad_norm": 0.5291089415550232, "learning_rate": 6.259617852537118e-05, "loss": 0.0472, "step": 44800 }, { "grad_norm": 1.1042388677597046, "learning_rate": 6.258017643014747e-05, "loss": 0.068, "step": 44810 }, { "grad_norm": 0.3495115637779236, "learning_rate": 6.256417295917456e-05, "loss": 0.0526, "step": 44820 }, { "grad_norm": 0.658675491809845, "learning_rate": 6.254816811420258e-05, "loss": 0.0417, "step": 44830 }, { "grad_norm": 0.9400171041488647, "learning_rate": 6.253216189698183e-05, "loss": 0.0506, "step": 44840 }, { "grad_norm": 0.5834140181541443, "learning_rate": 6.251615430926267e-05, "loss": 0.0403, "step": 44850 }, { "grad_norm": 0.3788054585456848, "learning_rate": 6.25001453527957e-05, "loss": 0.038, "step": 44860 }, { "grad_norm": 0.34909898042678833, "learning_rate": 6.248413502933164e-05, "loss": 0.045, "step": 44870 }, { "grad_norm": 0.49355730414390564, "learning_rate": 6.246812334062133e-05, "loss": 0.0385, "step": 44880 }, { "grad_norm": 0.47881147265434265, "learning_rate": 6.245211028841579e-05, "loss": 0.0417, "step": 44890 }, { "grad_norm": 0.5476759076118469, "learning_rate": 6.24360958744662e-05, "loss": 0.0391, "step": 44900 }, { "grad_norm": 0.5274955034255981, "learning_rate": 6.242008010052387e-05, "loss": 0.0417, "step": 44910 }, { "grad_norm": 0.4665059447288513, "learning_rate": 6.240406296834024e-05, "loss": 0.0445, "step": 44920 }, { "grad_norm": 0.6661362051963806, "learning_rate": 6.238804447966694e-05, "loss": 0.0365, "step": 44930 }, { "grad_norm": 0.4515610933303833, "learning_rate": 6.237202463625573e-05, "loss": 0.0594, "step": 44940 }, { "grad_norm": 0.44551578164100647, "learning_rate": 6.235600343985848e-05, "loss": 0.04, "step": 44950 }, { "grad_norm": 0.5694113969802856, "learning_rate": 6.233998089222729e-05, "loss": 0.0523, "step": 44960 }, { "grad_norm": 0.4018631875514984, "learning_rate": 6.232395699511433e-05, "loss": 0.0488, "step": 44970 }, { "grad_norm": 0.6850451231002808, "learning_rate": 6.230793175027199e-05, "loss": 0.0385, "step": 44980 }, { "grad_norm": 0.5034547448158264, "learning_rate": 6.22919051594527e-05, "loss": 0.0443, "step": 44990 }, { "grad_norm": 0.6563150882720947, "learning_rate": 6.227587722440917e-05, "loss": 0.0417, "step": 45000 }, { "grad_norm": 0.3946092128753662, "learning_rate": 6.225984794689414e-05, "loss": 0.0359, "step": 45010 }, { "grad_norm": 0.5963165163993835, "learning_rate": 6.22438173286606e-05, "loss": 0.0384, "step": 45020 }, { "grad_norm": 0.4361182153224945, "learning_rate": 6.222778537146159e-05, "loss": 0.0286, "step": 45030 }, { "grad_norm": 0.47042933106422424, "learning_rate": 6.221175207705037e-05, "loss": 0.0541, "step": 45040 }, { "grad_norm": 0.623105525970459, "learning_rate": 6.219571744718027e-05, "loss": 0.0757, "step": 45050 }, { "grad_norm": 0.8204020857810974, "learning_rate": 6.217968148360487e-05, "loss": 0.0466, "step": 45060 }, { "grad_norm": 0.39847221970558167, "learning_rate": 6.216364418807782e-05, "loss": 0.0347, "step": 45070 }, { "grad_norm": 0.736095130443573, "learning_rate": 6.214760556235292e-05, "loss": 0.0295, "step": 45080 }, { "grad_norm": 0.4370041787624359, "learning_rate": 6.213156560818414e-05, "loss": 0.0453, "step": 45090 }, { "grad_norm": 0.7945849299430847, "learning_rate": 6.211552432732559e-05, "loss": 0.0473, "step": 45100 }, { "grad_norm": 0.39864373207092285, "learning_rate": 6.20994817215315e-05, "loss": 0.0324, "step": 45110 }, { "grad_norm": 0.6261030435562134, "learning_rate": 6.208343779255628e-05, "loss": 0.0562, "step": 45120 }, { "grad_norm": 0.4193212687969208, "learning_rate": 6.206739254215449e-05, "loss": 0.0348, "step": 45130 }, { "grad_norm": 0.8751237392425537, "learning_rate": 6.205134597208077e-05, "loss": 0.0441, "step": 45140 }, { "grad_norm": 0.47100481390953064, "learning_rate": 6.203529808409e-05, "loss": 0.0372, "step": 45150 }, { "grad_norm": 0.35079893469810486, "learning_rate": 6.201924887993712e-05, "loss": 0.0351, "step": 45160 }, { "grad_norm": 0.7256351709365845, "learning_rate": 6.200319836137725e-05, "loss": 0.0411, "step": 45170 }, { "grad_norm": 0.2943079471588135, "learning_rate": 6.198714653016565e-05, "loss": 0.0357, "step": 45180 }, { "grad_norm": 0.41639232635498047, "learning_rate": 6.197109338805774e-05, "loss": 0.0328, "step": 45190 }, { "grad_norm": 0.9100534915924072, "learning_rate": 6.195503893680903e-05, "loss": 0.0459, "step": 45200 }, { "grad_norm": 0.5846815705299377, "learning_rate": 6.193898317817524e-05, "loss": 0.0339, "step": 45210 }, { "grad_norm": 0.5916298627853394, "learning_rate": 6.192292611391221e-05, "loss": 0.0431, "step": 45220 }, { "grad_norm": 0.6508750319480896, "learning_rate": 6.190686774577591e-05, "loss": 0.0447, "step": 45230 }, { "grad_norm": 0.7413817644119263, "learning_rate": 6.189080807552245e-05, "loss": 0.0417, "step": 45240 }, { "grad_norm": 0.4556681513786316, "learning_rate": 6.187474710490809e-05, "loss": 0.036, "step": 45250 }, { "grad_norm": 0.45222896337509155, "learning_rate": 6.185868483568926e-05, "loss": 0.0495, "step": 45260 }, { "grad_norm": 0.4543890655040741, "learning_rate": 6.184262126962245e-05, "loss": 0.0526, "step": 45270 }, { "grad_norm": 0.5073235630989075, "learning_rate": 6.182655640846442e-05, "loss": 0.0508, "step": 45280 }, { "grad_norm": 0.3422025740146637, "learning_rate": 6.181049025397196e-05, "loss": 0.0372, "step": 45290 }, { "grad_norm": 0.5130838751792908, "learning_rate": 6.179442280790202e-05, "loss": 0.0453, "step": 45300 }, { "grad_norm": 0.9403830766677856, "learning_rate": 6.177835407201174e-05, "loss": 0.0369, "step": 45310 }, { "grad_norm": 0.8262252807617188, "learning_rate": 6.176228404805839e-05, "loss": 0.0495, "step": 45320 }, { "grad_norm": 0.5020814538002014, "learning_rate": 6.174621273779932e-05, "loss": 0.0436, "step": 45330 }, { "grad_norm": 0.5707901120185852, "learning_rate": 6.17301401429921e-05, "loss": 0.0417, "step": 45340 }, { "grad_norm": 0.4767991304397583, "learning_rate": 6.171406626539436e-05, "loss": 0.042, "step": 45350 }, { "grad_norm": 0.6571478247642517, "learning_rate": 6.169799110676398e-05, "loss": 0.0532, "step": 45360 }, { "grad_norm": 0.6020030379295349, "learning_rate": 6.168191466885885e-05, "loss": 0.0448, "step": 45370 }, { "grad_norm": 0.6253277063369751, "learning_rate": 6.166583695343711e-05, "loss": 0.0273, "step": 45380 }, { "grad_norm": 0.5232621431350708, "learning_rate": 6.164975796225698e-05, "loss": 0.0341, "step": 45390 }, { "grad_norm": 0.6781899333000183, "learning_rate": 6.163367769707683e-05, "loss": 0.0429, "step": 45400 }, { "grad_norm": 0.48014721274375916, "learning_rate": 6.161759615965519e-05, "loss": 0.0425, "step": 45410 }, { "grad_norm": 0.301883339881897, "learning_rate": 6.16015133517507e-05, "loss": 0.0606, "step": 45420 }, { "grad_norm": 0.40712204575538635, "learning_rate": 6.158542927512214e-05, "loss": 0.0299, "step": 45430 }, { "grad_norm": 0.5063052773475647, "learning_rate": 6.156934393152846e-05, "loss": 0.0369, "step": 45440 }, { "grad_norm": 0.5365954637527466, "learning_rate": 6.15532573227287e-05, "loss": 0.0365, "step": 45450 }, { "grad_norm": 0.365740031003952, "learning_rate": 6.153716945048212e-05, "loss": 0.0325, "step": 45460 }, { "grad_norm": 0.558070957660675, "learning_rate": 6.152108031654802e-05, "loss": 0.0462, "step": 45470 }, { "grad_norm": 0.5491302013397217, "learning_rate": 6.15049899226859e-05, "loss": 0.0429, "step": 45480 }, { "grad_norm": 0.4388715326786041, "learning_rate": 6.148889827065537e-05, "loss": 0.0434, "step": 45490 }, { "grad_norm": 0.7170516848564148, "learning_rate": 6.147280536221622e-05, "loss": 0.0516, "step": 45500 }, { "grad_norm": 0.6162199378013611, "learning_rate": 6.145671119912832e-05, "loss": 0.0539, "step": 45510 }, { "grad_norm": 0.7023391723632812, "learning_rate": 6.144061578315169e-05, "loss": 0.0429, "step": 45520 }, { "grad_norm": 0.5057665109634399, "learning_rate": 6.142451911604654e-05, "loss": 0.0396, "step": 45530 }, { "grad_norm": 0.49886995553970337, "learning_rate": 6.140842119957315e-05, "loss": 0.0376, "step": 45540 }, { "grad_norm": 0.6046223044395447, "learning_rate": 6.139232203549197e-05, "loss": 0.0327, "step": 45550 }, { "grad_norm": 0.6743777990341187, "learning_rate": 6.137622162556357e-05, "loss": 0.0467, "step": 45560 }, { "grad_norm": 0.35168278217315674, "learning_rate": 6.13601199715487e-05, "loss": 0.0294, "step": 45570 }, { "grad_norm": 0.6188446879386902, "learning_rate": 6.134401707520816e-05, "loss": 0.0349, "step": 45580 }, { "grad_norm": 0.8611143231391907, "learning_rate": 6.132791293830298e-05, "loss": 0.0415, "step": 45590 }, { "grad_norm": 0.4227142035961151, "learning_rate": 6.131180756259428e-05, "loss": 0.0316, "step": 45600 }, { "grad_norm": 0.49922674894332886, "learning_rate": 6.129570094984331e-05, "loss": 0.0413, "step": 45610 }, { "grad_norm": 1.1932177543640137, "learning_rate": 6.127959310181145e-05, "loss": 0.0627, "step": 45620 }, { "grad_norm": 0.7691528797149658, "learning_rate": 6.126348402026026e-05, "loss": 0.05, "step": 45630 }, { "grad_norm": 0.4453265368938446, "learning_rate": 6.12473737069514e-05, "loss": 0.0383, "step": 45640 }, { "grad_norm": 0.5726106762886047, "learning_rate": 6.123126216364665e-05, "loss": 0.032, "step": 45650 }, { "grad_norm": 0.37548309564590454, "learning_rate": 6.121514939210797e-05, "loss": 0.0363, "step": 45660 }, { "grad_norm": 0.6973494291305542, "learning_rate": 6.119903539409741e-05, "loss": 0.0429, "step": 45670 }, { "grad_norm": 0.6070818901062012, "learning_rate": 6.118292017137716e-05, "loss": 0.0421, "step": 45680 }, { "grad_norm": 0.514407217502594, "learning_rate": 6.116680372570959e-05, "loss": 0.0494, "step": 45690 }, { "grad_norm": 0.5659774541854858, "learning_rate": 6.115068605885713e-05, "loss": 0.0436, "step": 45700 }, { "grad_norm": 0.44843265414237976, "learning_rate": 6.113456717258243e-05, "loss": 0.0468, "step": 45710 }, { "grad_norm": 0.5351880788803101, "learning_rate": 6.11184470686482e-05, "loss": 0.0325, "step": 45720 }, { "grad_norm": 0.49016451835632324, "learning_rate": 6.110232574881731e-05, "loss": 0.0314, "step": 45730 }, { "grad_norm": 0.6098645925521851, "learning_rate": 6.108620321485277e-05, "loss": 0.0401, "step": 45740 }, { "grad_norm": 0.4952126443386078, "learning_rate": 6.107007946851773e-05, "loss": 0.0489, "step": 45750 }, { "grad_norm": 0.4228462278842926, "learning_rate": 6.105395451157542e-05, "loss": 0.0543, "step": 45760 }, { "grad_norm": 0.4802439212799072, "learning_rate": 6.103782834578928e-05, "loss": 0.0433, "step": 45770 }, { "grad_norm": 0.40686145424842834, "learning_rate": 6.102170097292281e-05, "loss": 0.0343, "step": 45780 }, { "grad_norm": 0.45889121294021606, "learning_rate": 6.1005572394739716e-05, "loss": 0.0577, "step": 45790 }, { "grad_norm": 0.7374529242515564, "learning_rate": 6.0989442613003765e-05, "loss": 0.0445, "step": 45800 }, { "grad_norm": 0.7373717427253723, "learning_rate": 6.097331162947888e-05, "loss": 0.0524, "step": 45810 }, { "grad_norm": 0.5185517072677612, "learning_rate": 6.095717944592914e-05, "loss": 0.0432, "step": 45820 }, { "grad_norm": 0.38885772228240967, "learning_rate": 6.094104606411873e-05, "loss": 0.0557, "step": 45830 }, { "grad_norm": 0.514132559299469, "learning_rate": 6.0924911485811966e-05, "loss": 0.0435, "step": 45840 }, { "grad_norm": 0.3646095097064972, "learning_rate": 6.090877571277331e-05, "loss": 0.0419, "step": 45850 }, { "grad_norm": 0.6713000535964966, "learning_rate": 6.089263874676736e-05, "loss": 0.0406, "step": 45860 }, { "grad_norm": 0.4804949164390564, "learning_rate": 6.0876500589558796e-05, "loss": 0.0442, "step": 45870 }, { "grad_norm": 0.6221182942390442, "learning_rate": 6.086036124291248e-05, "loss": 0.0325, "step": 45880 }, { "grad_norm": 0.3360651731491089, "learning_rate": 6.084422070859339e-05, "loss": 0.0434, "step": 45890 }, { "grad_norm": 0.48758193850517273, "learning_rate": 6.082807898836663e-05, "loss": 0.0656, "step": 45900 }, { "grad_norm": 0.3812839090824127, "learning_rate": 6.081193608399742e-05, "loss": 0.0419, "step": 45910 }, { "grad_norm": 0.43968167901039124, "learning_rate": 6.0795791997251164e-05, "loss": 0.0379, "step": 45920 }, { "grad_norm": 0.5050171613693237, "learning_rate": 6.0779646729893294e-05, "loss": 0.0405, "step": 45930 }, { "grad_norm": 0.48804163932800293, "learning_rate": 6.0763500283689476e-05, "loss": 0.0372, "step": 45940 }, { "grad_norm": 0.4265117049217224, "learning_rate": 6.0747352660405455e-05, "loss": 0.0289, "step": 45950 }, { "grad_norm": 0.49989479780197144, "learning_rate": 6.073120386180709e-05, "loss": 0.04, "step": 45960 }, { "grad_norm": 0.7234792709350586, "learning_rate": 6.0715053889660425e-05, "loss": 0.0373, "step": 45970 }, { "grad_norm": 0.5057922601699829, "learning_rate": 6.069890274573157e-05, "loss": 0.0426, "step": 45980 }, { "grad_norm": 0.4918573796749115, "learning_rate": 6.068275043178679e-05, "loss": 0.0319, "step": 45990 }, { "grad_norm": 0.7554152607917786, "learning_rate": 6.066659694959248e-05, "loss": 0.0374, "step": 46000 }, { "grad_norm": 0.4295148551464081, "learning_rate": 6.0650442300915176e-05, "loss": 0.0372, "step": 46010 }, { "grad_norm": 0.5879769325256348, "learning_rate": 6.063428648752152e-05, "loss": 0.0431, "step": 46020 }, { "grad_norm": 0.6177473068237305, "learning_rate": 6.0618129511178266e-05, "loss": 0.0458, "step": 46030 }, { "grad_norm": 0.4385383725166321, "learning_rate": 6.060197137365234e-05, "loss": 0.035, "step": 46040 }, { "grad_norm": 0.5663703680038452, "learning_rate": 6.058581207671077e-05, "loss": 0.0491, "step": 46050 }, { "grad_norm": 0.42419618368148804, "learning_rate": 6.056965162212072e-05, "loss": 0.0415, "step": 46060 }, { "grad_norm": 0.6041064858436584, "learning_rate": 6.055349001164943e-05, "loss": 0.0568, "step": 46070 }, { "grad_norm": 0.6191592216491699, "learning_rate": 6.0537327247064347e-05, "loss": 0.0475, "step": 46080 }, { "grad_norm": 0.3722659945487976, "learning_rate": 6.0521163330133014e-05, "loss": 0.0566, "step": 46090 }, { "grad_norm": 0.7138701677322388, "learning_rate": 6.050499826262306e-05, "loss": 0.0439, "step": 46100 }, { "grad_norm": 0.9136433005332947, "learning_rate": 6.0488832046302294e-05, "loss": 0.0464, "step": 46110 }, { "grad_norm": 0.711907148361206, "learning_rate": 6.0472664682938626e-05, "loss": 0.0374, "step": 46120 }, { "grad_norm": 0.6785475015640259, "learning_rate": 6.045649617430009e-05, "loss": 0.0352, "step": 46130 }, { "grad_norm": 0.48925578594207764, "learning_rate": 6.0440326522154866e-05, "loss": 0.0376, "step": 46140 }, { "grad_norm": 1.4654407501220703, "learning_rate": 6.0424155728271224e-05, "loss": 0.0503, "step": 46150 }, { "grad_norm": 0.40831711888313293, "learning_rate": 6.040798379441758e-05, "loss": 0.0334, "step": 46160 }, { "grad_norm": 0.5350956916809082, "learning_rate": 6.0391810722362485e-05, "loss": 0.0405, "step": 46170 }, { "grad_norm": 0.368590772151947, "learning_rate": 6.037563651387458e-05, "loss": 0.0351, "step": 46180 }, { "grad_norm": 0.502582311630249, "learning_rate": 6.0359461170722666e-05, "loss": 0.0475, "step": 46190 }, { "grad_norm": 0.6258621215820312, "learning_rate": 6.034328469467566e-05, "loss": 0.036, "step": 46200 }, { "grad_norm": 0.6396428346633911, "learning_rate": 6.0327107087502596e-05, "loss": 0.06, "step": 46210 }, { "grad_norm": 0.5144876837730408, "learning_rate": 6.031092835097262e-05, "loss": 0.0348, "step": 46220 }, { "grad_norm": 0.4119630455970764, "learning_rate": 6.0294748486855024e-05, "loss": 0.0415, "step": 46230 }, { "grad_norm": 0.4705018401145935, "learning_rate": 6.0278567496919216e-05, "loss": 0.0415, "step": 46240 }, { "grad_norm": 0.609961986541748, "learning_rate": 6.026238538293472e-05, "loss": 0.0406, "step": 46250 }, { "grad_norm": 0.4719609320163727, "learning_rate": 6.024620214667118e-05, "loss": 0.0344, "step": 46260 }, { "grad_norm": 0.6451053619384766, "learning_rate": 6.0230017789898384e-05, "loss": 0.0342, "step": 46270 }, { "grad_norm": 0.5123656392097473, "learning_rate": 6.021383231438622e-05, "loss": 0.0354, "step": 46280 }, { "grad_norm": 0.8736794590950012, "learning_rate": 6.0197645721904704e-05, "loss": 0.0702, "step": 46290 }, { "grad_norm": 0.46068626642227173, "learning_rate": 6.0181458014224e-05, "loss": 0.0322, "step": 46300 }, { "grad_norm": 0.4918520450592041, "learning_rate": 6.016526919311435e-05, "loss": 0.0375, "step": 46310 }, { "grad_norm": 0.4165835976600647, "learning_rate": 6.014907926034613e-05, "loss": 0.0372, "step": 46320 }, { "grad_norm": 0.3533317446708679, "learning_rate": 6.0132888217689875e-05, "loss": 0.0301, "step": 46330 }, { "grad_norm": 0.33789631724357605, "learning_rate": 6.01166960669162e-05, "loss": 0.0501, "step": 46340 }, { "grad_norm": 0.5769772529602051, "learning_rate": 6.0100502809795845e-05, "loss": 0.0276, "step": 46350 }, { "grad_norm": 0.7959001064300537, "learning_rate": 6.00843084480997e-05, "loss": 0.0393, "step": 46360 }, { "grad_norm": 0.3669402599334717, "learning_rate": 6.0068112983598736e-05, "loss": 0.0468, "step": 46370 }, { "grad_norm": 2.1153244972229004, "learning_rate": 6.005191641806407e-05, "loss": 0.0378, "step": 46380 }, { "grad_norm": 0.30682191252708435, "learning_rate": 6.003571875326694e-05, "loss": 0.0315, "step": 46390 }, { "grad_norm": 0.38332197070121765, "learning_rate": 6.00195199909787e-05, "loss": 0.0468, "step": 46400 }, { "grad_norm": 0.38579174876213074, "learning_rate": 6.000332013297082e-05, "loss": 0.0363, "step": 46410 }, { "grad_norm": 0.5022239685058594, "learning_rate": 5.998711918101487e-05, "loss": 0.0409, "step": 46420 }, { "grad_norm": 0.3695491850376129, "learning_rate": 5.997091713688261e-05, "loss": 0.0344, "step": 46430 }, { "grad_norm": 0.6741548180580139, "learning_rate": 5.9954714002345836e-05, "loss": 0.0355, "step": 46440 }, { "grad_norm": 0.7668101191520691, "learning_rate": 5.993850977917649e-05, "loss": 0.0369, "step": 46450 }, { "grad_norm": 0.41579338908195496, "learning_rate": 5.992230446914667e-05, "loss": 0.0383, "step": 46460 }, { "grad_norm": 0.651203989982605, "learning_rate": 5.990609807402855e-05, "loss": 0.0385, "step": 46470 }, { "grad_norm": 0.8960958123207092, "learning_rate": 5.988989059559443e-05, "loss": 0.0383, "step": 46480 }, { "grad_norm": 0.4399777054786682, "learning_rate": 5.987368203561675e-05, "loss": 0.0497, "step": 46490 }, { "grad_norm": 0.8031578063964844, "learning_rate": 5.9857472395868055e-05, "loss": 0.0313, "step": 46500 }, { "grad_norm": 0.5028109550476074, "learning_rate": 5.9841261678120983e-05, "loss": 0.0329, "step": 46510 }, { "grad_norm": 0.3525160551071167, "learning_rate": 5.982504988414834e-05, "loss": 0.0321, "step": 46520 }, { "grad_norm": 0.5086719989776611, "learning_rate": 5.9808837015723015e-05, "loss": 0.0316, "step": 46530 }, { "grad_norm": 0.894873857498169, "learning_rate": 5.9792623074618016e-05, "loss": 0.0466, "step": 46540 }, { "grad_norm": 0.5458965301513672, "learning_rate": 5.977640806260648e-05, "loss": 0.0366, "step": 46550 }, { "grad_norm": 1.0964030027389526, "learning_rate": 5.976019198146166e-05, "loss": 0.0391, "step": 46560 }, { "grad_norm": 0.7272026538848877, "learning_rate": 5.974397483295692e-05, "loss": 0.0576, "step": 46570 }, { "grad_norm": 0.740049421787262, "learning_rate": 5.972775661886575e-05, "loss": 0.0354, "step": 46580 }, { "grad_norm": 0.44595617055892944, "learning_rate": 5.971153734096173e-05, "loss": 0.0391, "step": 46590 }, { "grad_norm": 0.602143406867981, "learning_rate": 5.969531700101859e-05, "loss": 0.0458, "step": 46600 }, { "grad_norm": 0.5841566324234009, "learning_rate": 5.9679095600810155e-05, "loss": 0.0451, "step": 46610 }, { "grad_norm": 0.5059265494346619, "learning_rate": 5.9662873142110384e-05, "loss": 0.0469, "step": 46620 }, { "grad_norm": 0.5205808281898499, "learning_rate": 5.964664962669333e-05, "loss": 0.0376, "step": 46630 }, { "grad_norm": 0.6823062300682068, "learning_rate": 5.9630425056333186e-05, "loss": 0.0382, "step": 46640 }, { "grad_norm": 0.5724559426307678, "learning_rate": 5.961419943280422e-05, "loss": 0.0365, "step": 46650 }, { "grad_norm": 0.4045714735984802, "learning_rate": 5.959797275788087e-05, "loss": 0.0444, "step": 46660 }, { "grad_norm": 0.511551558971405, "learning_rate": 5.958174503333765e-05, "loss": 0.0453, "step": 46670 }, { "grad_norm": 0.31283310055732727, "learning_rate": 5.9565516260949195e-05, "loss": 0.0409, "step": 46680 }, { "grad_norm": 0.44228407740592957, "learning_rate": 5.9549286442490273e-05, "loss": 0.0306, "step": 46690 }, { "grad_norm": 0.5114235281944275, "learning_rate": 5.953305557973572e-05, "loss": 0.0489, "step": 46700 }, { "grad_norm": 0.3286876976490021, "learning_rate": 5.9516823674460564e-05, "loss": 0.0308, "step": 46710 }, { "grad_norm": 0.47636476159095764, "learning_rate": 5.9500590728439875e-05, "loss": 0.0434, "step": 46720 }, { "grad_norm": 0.7064223289489746, "learning_rate": 5.948435674344886e-05, "loss": 0.0443, "step": 46730 }, { "grad_norm": 0.44908809661865234, "learning_rate": 5.946812172126285e-05, "loss": 0.0337, "step": 46740 }, { "grad_norm": 0.3687613606452942, "learning_rate": 5.945188566365729e-05, "loss": 0.0466, "step": 46750 }, { "grad_norm": 0.5168828964233398, "learning_rate": 5.943564857240773e-05, "loss": 0.0555, "step": 46760 }, { "grad_norm": 0.6955902576446533, "learning_rate": 5.941941044928981e-05, "loss": 0.0337, "step": 46770 }, { "grad_norm": 0.6446154713630676, "learning_rate": 5.940317129607935e-05, "loss": 0.0658, "step": 46780 }, { "grad_norm": 0.8887546062469482, "learning_rate": 5.9386931114552204e-05, "loss": 0.0386, "step": 46790 }, { "grad_norm": 0.3874961733818054, "learning_rate": 5.937068990648438e-05, "loss": 0.0567, "step": 46800 }, { "grad_norm": 0.7304533123970032, "learning_rate": 5.935444767365199e-05, "loss": 0.0497, "step": 46810 }, { "grad_norm": 0.5493643283843994, "learning_rate": 5.933820441783129e-05, "loss": 0.0354, "step": 46820 }, { "grad_norm": 0.3671071231365204, "learning_rate": 5.932196014079857e-05, "loss": 0.0334, "step": 46830 }, { "grad_norm": 0.22937160730361938, "learning_rate": 5.930571484433032e-05, "loss": 0.0303, "step": 46840 }, { "grad_norm": 0.5682726502418518, "learning_rate": 5.928946853020309e-05, "loss": 0.0449, "step": 46850 }, { "grad_norm": 0.43812674283981323, "learning_rate": 5.927322120019352e-05, "loss": 0.0365, "step": 46860 }, { "grad_norm": 0.31317874789237976, "learning_rate": 5.925697285607844e-05, "loss": 0.0278, "step": 46870 }, { "grad_norm": 0.5650526285171509, "learning_rate": 5.924072349963472e-05, "loss": 0.0291, "step": 46880 }, { "grad_norm": 0.4450380802154541, "learning_rate": 5.922447313263938e-05, "loss": 0.0373, "step": 46890 }, { "grad_norm": 0.533662736415863, "learning_rate": 5.920822175686952e-05, "loss": 0.0539, "step": 46900 }, { "grad_norm": 0.495441198348999, "learning_rate": 5.919196937410239e-05, "loss": 0.0304, "step": 46910 }, { "grad_norm": 0.5498186945915222, "learning_rate": 5.917571598611529e-05, "loss": 0.04, "step": 46920 }, { "grad_norm": 0.9872904419898987, "learning_rate": 5.9159461594685686e-05, "loss": 0.0516, "step": 46930 }, { "grad_norm": 0.5338671207427979, "learning_rate": 5.914320620159114e-05, "loss": 0.04, "step": 46940 }, { "grad_norm": 0.35807251930236816, "learning_rate": 5.912694980860931e-05, "loss": 0.0486, "step": 46950 }, { "grad_norm": 0.29199180006980896, "learning_rate": 5.9110692417517964e-05, "loss": 0.0311, "step": 46960 }, { "grad_norm": 0.5379152894020081, "learning_rate": 5.9094434030095e-05, "loss": 0.039, "step": 46970 }, { "grad_norm": 0.5436850190162659, "learning_rate": 5.907817464811841e-05, "loss": 0.0451, "step": 46980 }, { "grad_norm": 0.5257582068443298, "learning_rate": 5.906191427336627e-05, "loss": 0.0316, "step": 46990 }, { "grad_norm": 0.41693419218063354, "learning_rate": 5.9045652907616835e-05, "loss": 0.0621, "step": 47000 }, { "grad_norm": 0.4069702625274658, "learning_rate": 5.902939055264838e-05, "loss": 0.0439, "step": 47010 }, { "grad_norm": 0.3174087703227997, "learning_rate": 5.901312721023935e-05, "loss": 0.0436, "step": 47020 }, { "grad_norm": 0.6057451367378235, "learning_rate": 5.8996862882168294e-05, "loss": 0.0696, "step": 47030 }, { "grad_norm": 0.6434809565544128, "learning_rate": 5.8980597570213826e-05, "loss": 0.073, "step": 47040 }, { "grad_norm": 0.8128032684326172, "learning_rate": 5.896433127615471e-05, "loss": 0.0431, "step": 47050 }, { "grad_norm": 1.2439943552017212, "learning_rate": 5.894806400176981e-05, "loss": 0.04, "step": 47060 }, { "grad_norm": 0.6268391013145447, "learning_rate": 5.893179574883808e-05, "loss": 0.0478, "step": 47070 }, { "grad_norm": 0.4696190357208252, "learning_rate": 5.8915526519138585e-05, "loss": 0.0405, "step": 47080 }, { "grad_norm": 0.4708162248134613, "learning_rate": 5.889925631445053e-05, "loss": 0.0377, "step": 47090 }, { "grad_norm": 0.4464063346385956, "learning_rate": 5.888298513655318e-05, "loss": 0.0373, "step": 47100 }, { "grad_norm": 0.40289121866226196, "learning_rate": 5.886671298722592e-05, "loss": 0.0382, "step": 47110 }, { "grad_norm": 0.4008645713329315, "learning_rate": 5.885043986824828e-05, "loss": 0.035, "step": 47120 }, { "grad_norm": 0.5800312757492065, "learning_rate": 5.883416578139982e-05, "loss": 0.0444, "step": 47130 }, { "grad_norm": 0.48354554176330566, "learning_rate": 5.881789072846029e-05, "loss": 0.0334, "step": 47140 }, { "grad_norm": 0.6348511576652527, "learning_rate": 5.8801614711209474e-05, "loss": 0.0338, "step": 47150 }, { "grad_norm": 0.7917037606239319, "learning_rate": 5.878533773142732e-05, "loss": 0.0392, "step": 47160 }, { "grad_norm": 0.32908597588539124, "learning_rate": 5.8769059790893834e-05, "loss": 0.0337, "step": 47170 }, { "grad_norm": 0.7900880575180054, "learning_rate": 5.8752780891389145e-05, "loss": 0.0602, "step": 47180 }, { "grad_norm": 0.31418222188949585, "learning_rate": 5.873650103469351e-05, "loss": 0.0396, "step": 47190 }, { "grad_norm": 0.42894434928894043, "learning_rate": 5.872022022258725e-05, "loss": 0.0365, "step": 47200 }, { "grad_norm": 0.6606318354606628, "learning_rate": 5.8703938456850814e-05, "loss": 0.0502, "step": 47210 }, { "grad_norm": 0.27205464243888855, "learning_rate": 5.868765573926477e-05, "loss": 0.0267, "step": 47220 }, { "grad_norm": 0.5112336277961731, "learning_rate": 5.867137207160974e-05, "loss": 0.0335, "step": 47230 }, { "grad_norm": 0.38165879249572754, "learning_rate": 5.86550874556665e-05, "loss": 0.0359, "step": 47240 }, { "grad_norm": 0.9148063659667969, "learning_rate": 5.863880189321592e-05, "loss": 0.0288, "step": 47250 }, { "grad_norm": 0.48080945014953613, "learning_rate": 5.862251538603894e-05, "loss": 0.0394, "step": 47260 }, { "grad_norm": 0.34346437454223633, "learning_rate": 5.8606227935916656e-05, "loss": 0.0369, "step": 47270 }, { "grad_norm": 0.49248045682907104, "learning_rate": 5.858993954463021e-05, "loss": 0.0577, "step": 47280 }, { "grad_norm": 0.6122331619262695, "learning_rate": 5.8573650213960896e-05, "loss": 0.0296, "step": 47290 }, { "grad_norm": 0.49907317757606506, "learning_rate": 5.8557359945690084e-05, "loss": 0.0423, "step": 47300 }, { "grad_norm": 0.8070854544639587, "learning_rate": 5.8541068741599246e-05, "loss": 0.0364, "step": 47310 }, { "grad_norm": 0.5649891495704651, "learning_rate": 5.8524776603469985e-05, "loss": 0.0417, "step": 47320 }, { "grad_norm": 0.49311479926109314, "learning_rate": 5.850848353308397e-05, "loss": 0.0283, "step": 47330 }, { "grad_norm": 0.33464300632476807, "learning_rate": 5.849218953222297e-05, "loss": 0.0439, "step": 47340 }, { "grad_norm": 0.6575952768325806, "learning_rate": 5.847589460266891e-05, "loss": 0.049, "step": 47350 }, { "grad_norm": 0.8154641389846802, "learning_rate": 5.845959874620375e-05, "loss": 0.044, "step": 47360 }, { "grad_norm": 0.4968276917934418, "learning_rate": 5.84433019646096e-05, "loss": 0.0355, "step": 47370 }, { "grad_norm": 0.40822991728782654, "learning_rate": 5.842700425966863e-05, "loss": 0.0295, "step": 47380 }, { "grad_norm": 0.34337785840034485, "learning_rate": 5.841070563316315e-05, "loss": 0.045, "step": 47390 }, { "grad_norm": 0.47355473041534424, "learning_rate": 5.839440608687554e-05, "loss": 0.0336, "step": 47400 }, { "grad_norm": 0.41458675265312195, "learning_rate": 5.837810562258831e-05, "loss": 0.0354, "step": 47410 }, { "grad_norm": 0.46843403577804565, "learning_rate": 5.836180424208405e-05, "loss": 0.0391, "step": 47420 }, { "grad_norm": 0.623774528503418, "learning_rate": 5.834550194714543e-05, "loss": 0.032, "step": 47430 }, { "grad_norm": 0.493404746055603, "learning_rate": 5.832919873955528e-05, "loss": 0.0306, "step": 47440 }, { "grad_norm": 0.4691556394100189, "learning_rate": 5.831289462109648e-05, "loss": 0.0408, "step": 47450 }, { "grad_norm": 0.7887482047080994, "learning_rate": 5.829658959355202e-05, "loss": 0.0573, "step": 47460 }, { "grad_norm": 0.5981613993644714, "learning_rate": 5.8280283658704995e-05, "loss": 0.0376, "step": 47470 }, { "grad_norm": 0.6420438885688782, "learning_rate": 5.826397681833859e-05, "loss": 0.034, "step": 47480 }, { "grad_norm": 0.503362238407135, "learning_rate": 5.824766907423612e-05, "loss": 0.0441, "step": 47490 }, { "grad_norm": 0.46293339133262634, "learning_rate": 5.8231360428180945e-05, "loss": 0.0381, "step": 47500 }, { "grad_norm": 0.5160019993782043, "learning_rate": 5.821505088195658e-05, "loss": 0.0363, "step": 47510 }, { "grad_norm": 0.5611217617988586, "learning_rate": 5.819874043734661e-05, "loss": 0.0447, "step": 47520 }, { "grad_norm": 1.1123214960098267, "learning_rate": 5.8182429096134695e-05, "loss": 0.037, "step": 47530 }, { "grad_norm": 0.5682719349861145, "learning_rate": 5.816611686010465e-05, "loss": 0.033, "step": 47540 }, { "grad_norm": 0.556789755821228, "learning_rate": 5.814980373104033e-05, "loss": 0.0374, "step": 47550 }, { "grad_norm": 0.4959907829761505, "learning_rate": 5.813348971072572e-05, "loss": 0.0432, "step": 47560 }, { "grad_norm": 0.5573338270187378, "learning_rate": 5.811717480094492e-05, "loss": 0.0359, "step": 47570 }, { "grad_norm": 0.4886205494403839, "learning_rate": 5.810085900348209e-05, "loss": 0.0347, "step": 47580 }, { "grad_norm": 0.4968734085559845, "learning_rate": 5.8084542320121483e-05, "loss": 0.0372, "step": 47590 }, { "grad_norm": 0.4159528613090515, "learning_rate": 5.8068224752647497e-05, "loss": 0.0345, "step": 47600 }, { "grad_norm": 0.5880458950996399, "learning_rate": 5.805190630284457e-05, "loss": 0.0305, "step": 47610 }, { "grad_norm": 0.4548424482345581, "learning_rate": 5.803558697249729e-05, "loss": 0.0359, "step": 47620 }, { "grad_norm": 0.3644905090332031, "learning_rate": 5.8019266763390276e-05, "loss": 0.0256, "step": 47630 }, { "grad_norm": 0.5028567314147949, "learning_rate": 5.800294567730833e-05, "loss": 0.0329, "step": 47640 }, { "grad_norm": 0.4902764558792114, "learning_rate": 5.798662371603628e-05, "loss": 0.0311, "step": 47650 }, { "grad_norm": 0.6249067783355713, "learning_rate": 5.797030088135904e-05, "loss": 0.0374, "step": 47660 }, { "grad_norm": 0.8114420771598816, "learning_rate": 5.7953977175061705e-05, "loss": 0.0355, "step": 47670 }, { "grad_norm": 0.5691797137260437, "learning_rate": 5.793765259892938e-05, "loss": 0.0405, "step": 47680 }, { "grad_norm": 0.5530675649642944, "learning_rate": 5.792132715474729e-05, "loss": 0.0296, "step": 47690 }, { "grad_norm": 0.5473280549049377, "learning_rate": 5.790500084430078e-05, "loss": 0.0232, "step": 47700 }, { "grad_norm": 0.643637478351593, "learning_rate": 5.7888673669375274e-05, "loss": 0.0338, "step": 47710 }, { "grad_norm": 1.7806148529052734, "learning_rate": 5.787234563175625e-05, "loss": 0.0372, "step": 47720 }, { "grad_norm": 0.8986713886260986, "learning_rate": 5.7856016733229355e-05, "loss": 0.0432, "step": 47730 }, { "grad_norm": 0.36645928025245667, "learning_rate": 5.7839686975580297e-05, "loss": 0.0316, "step": 47740 }, { "grad_norm": 0.45028558373451233, "learning_rate": 5.782335636059484e-05, "loss": 0.036, "step": 47750 }, { "grad_norm": 0.5780682563781738, "learning_rate": 5.780702489005889e-05, "loss": 0.0281, "step": 47760 }, { "grad_norm": 0.6230888962745667, "learning_rate": 5.779069256575845e-05, "loss": 0.0302, "step": 47770 }, { "grad_norm": 0.5425339937210083, "learning_rate": 5.7774359389479574e-05, "loss": 0.0596, "step": 47780 }, { "grad_norm": 0.6965950131416321, "learning_rate": 5.775802536300845e-05, "loss": 0.0345, "step": 47790 }, { "grad_norm": 0.6177119612693787, "learning_rate": 5.774169048813134e-05, "loss": 0.0337, "step": 47800 }, { "grad_norm": 0.4641689956188202, "learning_rate": 5.77253547666346e-05, "loss": 0.0377, "step": 47810 }, { "grad_norm": 0.483353853225708, "learning_rate": 5.770901820030465e-05, "loss": 0.0351, "step": 47820 }, { "grad_norm": 0.4589986801147461, "learning_rate": 5.769268079092809e-05, "loss": 0.0445, "step": 47830 }, { "grad_norm": 0.5055543780326843, "learning_rate": 5.767634254029151e-05, "loss": 0.0519, "step": 47840 }, { "grad_norm": 0.5410589575767517, "learning_rate": 5.7660003450181655e-05, "loss": 0.0344, "step": 47850 }, { "grad_norm": 0.49686309695243835, "learning_rate": 5.764366352238534e-05, "loss": 0.035, "step": 47860 }, { "grad_norm": 0.33427613973617554, "learning_rate": 5.7627322758689474e-05, "loss": 0.029, "step": 47870 }, { "grad_norm": 0.3627042770385742, "learning_rate": 5.761098116088105e-05, "loss": 0.0293, "step": 47880 }, { "grad_norm": 0.7069175839424133, "learning_rate": 5.759463873074717e-05, "loss": 0.0344, "step": 47890 }, { "grad_norm": 0.6708053350448608, "learning_rate": 5.757829547007504e-05, "loss": 0.0367, "step": 47900 }, { "grad_norm": 0.4930576980113983, "learning_rate": 5.756195138065189e-05, "loss": 0.0462, "step": 47910 }, { "grad_norm": 0.48064324259757996, "learning_rate": 5.754560646426511e-05, "loss": 0.0508, "step": 47920 }, { "grad_norm": 0.5163628458976746, "learning_rate": 5.752926072270216e-05, "loss": 0.0435, "step": 47930 }, { "grad_norm": 0.8762794733047485, "learning_rate": 5.7512914157750563e-05, "loss": 0.0494, "step": 47940 }, { "grad_norm": 0.6609182953834534, "learning_rate": 5.749656677119798e-05, "loss": 0.0419, "step": 47950 }, { "grad_norm": 0.6197870969772339, "learning_rate": 5.7480218564832125e-05, "loss": 0.0465, "step": 47960 }, { "grad_norm": 0.8353673815727234, "learning_rate": 5.746386954044082e-05, "loss": 0.0522, "step": 47970 }, { "grad_norm": 0.7277426719665527, "learning_rate": 5.744751969981195e-05, "loss": 0.0493, "step": 47980 }, { "grad_norm": 0.3353672921657562, "learning_rate": 5.7431169044733526e-05, "loss": 0.0331, "step": 47990 }, { "grad_norm": 0.49139246344566345, "learning_rate": 5.741481757699364e-05, "loss": 0.0394, "step": 48000 }, { "grad_norm": 0.5496563911437988, "learning_rate": 5.7398465298380434e-05, "loss": 0.0282, "step": 48010 }, { "grad_norm": 0.4020997881889343, "learning_rate": 5.7382112210682193e-05, "loss": 0.0309, "step": 48020 }, { "grad_norm": 0.3509879410266876, "learning_rate": 5.7365758315687266e-05, "loss": 0.0469, "step": 48030 }, { "grad_norm": 0.3841764032840729, "learning_rate": 5.734940361518407e-05, "loss": 0.0272, "step": 48040 }, { "grad_norm": 0.42376184463500977, "learning_rate": 5.733304811096116e-05, "loss": 0.0347, "step": 48050 }, { "grad_norm": 0.5273138880729675, "learning_rate": 5.731669180480713e-05, "loss": 0.0476, "step": 48060 }, { "grad_norm": 0.6142171025276184, "learning_rate": 5.730033469851067e-05, "loss": 0.0345, "step": 48070 }, { "grad_norm": 0.602709949016571, "learning_rate": 5.7283976793860607e-05, "loss": 0.0407, "step": 48080 }, { "grad_norm": 0.7131620645523071, "learning_rate": 5.726761809264577e-05, "loss": 0.0416, "step": 48090 }, { "grad_norm": 0.6107317209243774, "learning_rate": 5.7251258596655155e-05, "loss": 0.0346, "step": 48100 }, { "grad_norm": 0.5555548071861267, "learning_rate": 5.72348983076778e-05, "loss": 0.0403, "step": 48110 }, { "grad_norm": 0.3646657168865204, "learning_rate": 5.7218537227502854e-05, "loss": 0.0414, "step": 48120 }, { "grad_norm": 0.3722534477710724, "learning_rate": 5.720217535791951e-05, "loss": 0.0327, "step": 48130 }, { "grad_norm": 0.7997023463249207, "learning_rate": 5.718581270071711e-05, "loss": 0.0342, "step": 48140 }, { "grad_norm": 0.30892014503479004, "learning_rate": 5.716944925768505e-05, "loss": 0.0381, "step": 48150 }, { "grad_norm": 0.8929995894432068, "learning_rate": 5.7153085030612786e-05, "loss": 0.0327, "step": 48160 }, { "grad_norm": 0.3718774914741516, "learning_rate": 5.713672002128989e-05, "loss": 0.0315, "step": 48170 }, { "grad_norm": 0.33506256341934204, "learning_rate": 5.712035423150602e-05, "loss": 0.0327, "step": 48180 }, { "grad_norm": 0.5871257185935974, "learning_rate": 5.710398766305094e-05, "loss": 0.0444, "step": 48190 }, { "grad_norm": 0.4968559443950653, "learning_rate": 5.708762031771442e-05, "loss": 0.0413, "step": 48200 }, { "grad_norm": 0.756089448928833, "learning_rate": 5.707125219728642e-05, "loss": 0.035, "step": 48210 }, { "grad_norm": 0.42640504240989685, "learning_rate": 5.7054883303556905e-05, "loss": 0.0299, "step": 48220 }, { "grad_norm": 0.5022427439689636, "learning_rate": 5.703851363831595e-05, "loss": 0.0302, "step": 48230 }, { "grad_norm": 0.3512842655181885, "learning_rate": 5.702214320335374e-05, "loss": 0.0336, "step": 48240 }, { "grad_norm": 0.4586591124534607, "learning_rate": 5.7005772000460514e-05, "loss": 0.0537, "step": 48250 }, { "grad_norm": 0.46326538920402527, "learning_rate": 5.698940003142656e-05, "loss": 0.0458, "step": 48260 }, { "grad_norm": 0.5026319622993469, "learning_rate": 5.697302729804236e-05, "loss": 0.0342, "step": 48270 }, { "grad_norm": 0.41689831018447876, "learning_rate": 5.695665380209837e-05, "loss": 0.0468, "step": 48280 }, { "grad_norm": 0.7561361789703369, "learning_rate": 5.6940279545385165e-05, "loss": 0.0408, "step": 48290 }, { "grad_norm": 0.4784672260284424, "learning_rate": 5.692390452969344e-05, "loss": 0.0318, "step": 48300 }, { "grad_norm": 0.4608076214790344, "learning_rate": 5.690752875681392e-05, "loss": 0.034, "step": 48310 }, { "grad_norm": 0.3699593245983124, "learning_rate": 5.6891152228537435e-05, "loss": 0.0309, "step": 48320 }, { "grad_norm": 0.2474759966135025, "learning_rate": 5.687477494665492e-05, "loss": 0.0369, "step": 48330 }, { "grad_norm": 0.5737731456756592, "learning_rate": 5.685839691295733e-05, "loss": 0.0417, "step": 48340 }, { "grad_norm": 0.4271846115589142, "learning_rate": 5.6842018129235786e-05, "loss": 0.037, "step": 48350 }, { "grad_norm": 0.5160698890686035, "learning_rate": 5.6825638597281404e-05, "loss": 0.0293, "step": 48360 }, { "grad_norm": 0.42695313692092896, "learning_rate": 5.680925831888546e-05, "loss": 0.0302, "step": 48370 }, { "grad_norm": 0.3380265235900879, "learning_rate": 5.6792877295839274e-05, "loss": 0.0307, "step": 48380 }, { "grad_norm": 0.38567018508911133, "learning_rate": 5.6776495529934224e-05, "loss": 0.0273, "step": 48390 }, { "grad_norm": 0.3509846329689026, "learning_rate": 5.6760113022961824e-05, "loss": 0.0386, "step": 48400 }, { "grad_norm": 0.5245848298072815, "learning_rate": 5.6743729776713617e-05, "loss": 0.0274, "step": 48410 }, { "grad_norm": 0.6442736983299255, "learning_rate": 5.672734579298126e-05, "loss": 0.0497, "step": 48420 }, { "grad_norm": 0.483984112739563, "learning_rate": 5.671096107355649e-05, "loss": 0.0364, "step": 48430 }, { "grad_norm": 0.762944221496582, "learning_rate": 5.66945756202311e-05, "loss": 0.0482, "step": 48440 }, { "grad_norm": 0.4519008994102478, "learning_rate": 5.667818943479699e-05, "loss": 0.0373, "step": 48450 }, { "grad_norm": 0.4329519271850586, "learning_rate": 5.666180251904612e-05, "loss": 0.0401, "step": 48460 }, { "grad_norm": 0.5199269652366638, "learning_rate": 5.6645414874770555e-05, "loss": 0.0337, "step": 48470 }, { "grad_norm": 0.42963510751724243, "learning_rate": 5.66290265037624e-05, "loss": 0.0341, "step": 48480 }, { "grad_norm": 0.4355095326900482, "learning_rate": 5.661263740781386e-05, "loss": 0.0314, "step": 48490 }, { "grad_norm": 1.020163655281067, "learning_rate": 5.6596247588717254e-05, "loss": 0.0389, "step": 48500 }, { "grad_norm": 0.46826300024986267, "learning_rate": 5.6579857048264926e-05, "loss": 0.0414, "step": 48510 }, { "grad_norm": 0.5379703640937805, "learning_rate": 5.6563465788249314e-05, "loss": 0.0382, "step": 48520 }, { "grad_norm": 0.6119928359985352, "learning_rate": 5.6547073810462956e-05, "loss": 0.0406, "step": 48530 }, { "grad_norm": 0.5260797142982483, "learning_rate": 5.653068111669846e-05, "loss": 0.0346, "step": 48540 }, { "grad_norm": 0.5656941533088684, "learning_rate": 5.651428770874848e-05, "loss": 0.0325, "step": 48550 }, { "grad_norm": 0.8091406226158142, "learning_rate": 5.64978935884058e-05, "loss": 0.0344, "step": 48560 }, { "grad_norm": 0.563309371471405, "learning_rate": 5.6481498757463244e-05, "loss": 0.0328, "step": 48570 }, { "grad_norm": 0.5622743368148804, "learning_rate": 5.646510321771373e-05, "loss": 0.0462, "step": 48580 }, { "grad_norm": 0.5503963232040405, "learning_rate": 5.644870697095024e-05, "loss": 0.055, "step": 48590 }, { "grad_norm": 0.690356969833374, "learning_rate": 5.643231001896586e-05, "loss": 0.0334, "step": 48600 }, { "grad_norm": 0.6374477744102478, "learning_rate": 5.6415912363553726e-05, "loss": 0.0359, "step": 48610 }, { "grad_norm": 0.6009634137153625, "learning_rate": 5.639951400650706e-05, "loss": 0.0566, "step": 48620 }, { "grad_norm": 0.19683276116847992, "learning_rate": 5.6383114949619165e-05, "loss": 0.0321, "step": 48630 }, { "grad_norm": 0.4999937117099762, "learning_rate": 5.636671519468342e-05, "loss": 0.0433, "step": 48640 }, { "grad_norm": 0.46277153491973877, "learning_rate": 5.635031474349327e-05, "loss": 0.0324, "step": 48650 }, { "grad_norm": 0.6813939213752747, "learning_rate": 5.6333913597842246e-05, "loss": 0.0378, "step": 48660 }, { "grad_norm": 0.8857858777046204, "learning_rate": 5.6317511759523955e-05, "loss": 0.0402, "step": 48670 }, { "grad_norm": 0.506359338760376, "learning_rate": 5.630110923033207e-05, "loss": 0.0387, "step": 48680 }, { "grad_norm": 0.7327402234077454, "learning_rate": 5.628470601206036e-05, "loss": 0.0426, "step": 48690 }, { "grad_norm": 0.4475754499435425, "learning_rate": 5.626830210650266e-05, "loss": 0.0407, "step": 48700 }, { "grad_norm": 0.3259465992450714, "learning_rate": 5.625189751545285e-05, "loss": 0.0378, "step": 48710 }, { "grad_norm": 1.0861355066299438, "learning_rate": 5.6235492240704936e-05, "loss": 0.0239, "step": 48720 }, { "grad_norm": 0.536125123500824, "learning_rate": 5.621908628405296e-05, "loss": 0.0484, "step": 48730 }, { "grad_norm": 0.49721160531044006, "learning_rate": 5.620267964729106e-05, "loss": 0.0333, "step": 48740 }, { "grad_norm": 0.4191444516181946, "learning_rate": 5.618627233221344e-05, "loss": 0.0329, "step": 48750 }, { "grad_norm": 0.4538092613220215, "learning_rate": 5.616986434061438e-05, "loss": 0.0353, "step": 48760 }, { "grad_norm": 0.5553292632102966, "learning_rate": 5.615345567428822e-05, "loss": 0.0353, "step": 48770 }, { "grad_norm": 0.4328831136226654, "learning_rate": 5.613704633502941e-05, "loss": 0.0343, "step": 48780 }, { "grad_norm": 1.2714548110961914, "learning_rate": 5.612063632463245e-05, "loss": 0.0339, "step": 48790 }, { "grad_norm": 0.6377741694450378, "learning_rate": 5.610422564489188e-05, "loss": 0.0486, "step": 48800 }, { "grad_norm": 0.5339638590812683, "learning_rate": 5.608781429760239e-05, "loss": 0.027, "step": 48810 }, { "grad_norm": 0.3639584481716156, "learning_rate": 5.607140228455866e-05, "loss": 0.0354, "step": 48820 }, { "grad_norm": 0.4407941699028015, "learning_rate": 5.605498960755553e-05, "loss": 0.0297, "step": 48830 }, { "grad_norm": 0.4010593891143799, "learning_rate": 5.603857626838782e-05, "loss": 0.0389, "step": 48840 }, { "grad_norm": 0.6231701970100403, "learning_rate": 5.60221622688505e-05, "loss": 0.0478, "step": 48850 }, { "grad_norm": 0.5664529800415039, "learning_rate": 5.6005747610738565e-05, "loss": 0.0397, "step": 48860 }, { "grad_norm": 0.45600345730781555, "learning_rate": 5.59893322958471e-05, "loss": 0.0282, "step": 48870 }, { "grad_norm": 0.9791207313537598, "learning_rate": 5.5972916325971256e-05, "loss": 0.0314, "step": 48880 }, { "grad_norm": 0.407922625541687, "learning_rate": 5.595649970290628e-05, "loss": 0.0349, "step": 48890 }, { "grad_norm": 0.625444233417511, "learning_rate": 5.5940082428447426e-05, "loss": 0.0477, "step": 48900 }, { "grad_norm": 0.3968881070613861, "learning_rate": 5.592366450439012e-05, "loss": 0.0506, "step": 48910 }, { "grad_norm": 0.5080453753471375, "learning_rate": 5.590724593252975e-05, "loss": 0.044, "step": 48920 }, { "grad_norm": 0.45054367184638977, "learning_rate": 5.589082671466184e-05, "loss": 0.0404, "step": 48930 }, { "grad_norm": 0.4294191300868988, "learning_rate": 5.587440685258199e-05, "loss": 0.0477, "step": 48940 }, { "grad_norm": 0.5574612617492676, "learning_rate": 5.585798634808583e-05, "loss": 0.0317, "step": 48950 }, { "grad_norm": 0.4041530191898346, "learning_rate": 5.584156520296909e-05, "loss": 0.0459, "step": 48960 }, { "grad_norm": 0.4575566351413727, "learning_rate": 5.582514341902757e-05, "loss": 0.0331, "step": 48970 }, { "grad_norm": 0.5069553256034851, "learning_rate": 5.580872099805713e-05, "loss": 0.04, "step": 48980 }, { "grad_norm": 0.6326920390129089, "learning_rate": 5.5792297941853674e-05, "loss": 0.0387, "step": 48990 }, { "grad_norm": 0.39833611249923706, "learning_rate": 5.5775874252213247e-05, "loss": 0.0405, "step": 49000 }, { "grad_norm": 0.42402204871177673, "learning_rate": 5.575944993093189e-05, "loss": 0.0366, "step": 49010 }, { "grad_norm": 0.47556057572364807, "learning_rate": 5.574302497980574e-05, "loss": 0.0327, "step": 49020 }, { "grad_norm": 0.8837756514549255, "learning_rate": 5.5726599400631005e-05, "loss": 0.047, "step": 49030 }, { "grad_norm": 0.6472008228302002, "learning_rate": 5.571017319520399e-05, "loss": 0.0328, "step": 49040 }, { "grad_norm": 0.7033142447471619, "learning_rate": 5.5693746365320984e-05, "loss": 0.0351, "step": 49050 }, { "grad_norm": 0.716214656829834, "learning_rate": 5.567731891277848e-05, "loss": 0.0362, "step": 49060 }, { "grad_norm": 0.4281844198703766, "learning_rate": 5.566089083937288e-05, "loss": 0.0353, "step": 49070 }, { "grad_norm": 0.7364916205406189, "learning_rate": 5.564446214690079e-05, "loss": 0.0387, "step": 49080 }, { "grad_norm": 0.6866326928138733, "learning_rate": 5.56280328371588e-05, "loss": 0.0383, "step": 49090 }, { "grad_norm": 0.6066648960113525, "learning_rate": 5.5611602911943596e-05, "loss": 0.0313, "step": 49100 }, { "grad_norm": 0.6047119498252869, "learning_rate": 5.559517237305194e-05, "loss": 0.0453, "step": 49110 }, { "grad_norm": 0.343569278717041, "learning_rate": 5.557874122228064e-05, "loss": 0.032, "step": 49120 }, { "grad_norm": 0.3954274654388428, "learning_rate": 5.55623094614266e-05, "loss": 0.0339, "step": 49130 }, { "grad_norm": 0.5883297920227051, "learning_rate": 5.5545877092286743e-05, "loss": 0.0455, "step": 49140 }, { "grad_norm": 0.4780648350715637, "learning_rate": 5.552944411665811e-05, "loss": 0.029, "step": 49150 }, { "grad_norm": 0.3599778711795807, "learning_rate": 5.5513010536337795e-05, "loss": 0.0401, "step": 49160 }, { "grad_norm": 0.6141192317008972, "learning_rate": 5.5496576353122934e-05, "loss": 0.0348, "step": 49170 }, { "grad_norm": 0.7982378602027893, "learning_rate": 5.548014156881074e-05, "loss": 0.0332, "step": 49180 }, { "grad_norm": 0.42287078499794006, "learning_rate": 5.5463706185198494e-05, "loss": 0.0214, "step": 49190 }, { "grad_norm": 0.34313589334487915, "learning_rate": 5.544727020408358e-05, "loss": 0.0367, "step": 49200 }, { "grad_norm": 0.3345785439014435, "learning_rate": 5.5430833627263366e-05, "loss": 0.0439, "step": 49210 }, { "grad_norm": 1.0383179187774658, "learning_rate": 5.5414396456535354e-05, "loss": 0.0516, "step": 49220 }, { "grad_norm": 0.4685083031654358, "learning_rate": 5.539795869369709e-05, "loss": 0.0321, "step": 49230 }, { "grad_norm": 0.5074538588523865, "learning_rate": 5.538152034054618e-05, "loss": 0.0376, "step": 49240 }, { "grad_norm": 0.34581229090690613, "learning_rate": 5.536508139888028e-05, "loss": 0.043, "step": 49250 }, { "grad_norm": 0.40339434146881104, "learning_rate": 5.534864187049716e-05, "loss": 0.034, "step": 49260 }, { "grad_norm": 0.634080708026886, "learning_rate": 5.5332201757194615e-05, "loss": 0.0413, "step": 49270 }, { "grad_norm": 0.43401700258255005, "learning_rate": 5.531576106077049e-05, "loss": 0.0243, "step": 49280 }, { "grad_norm": 0.37169352173805237, "learning_rate": 5.529931978302272e-05, "loss": 0.0303, "step": 49290 }, { "grad_norm": 0.48898977041244507, "learning_rate": 5.528287792574932e-05, "loss": 0.0312, "step": 49300 }, { "grad_norm": 0.3549465835094452, "learning_rate": 5.5266435490748324e-05, "loss": 0.0343, "step": 49310 }, { "grad_norm": 0.4506746530532837, "learning_rate": 5.524999247981787e-05, "loss": 0.0448, "step": 49320 }, { "grad_norm": 0.44002822041511536, "learning_rate": 5.523354889475613e-05, "loss": 0.0382, "step": 49330 }, { "grad_norm": 0.5908084511756897, "learning_rate": 5.521710473736134e-05, "loss": 0.0465, "step": 49340 }, { "grad_norm": 0.9031105041503906, "learning_rate": 5.5200660009431826e-05, "loss": 0.0303, "step": 49350 }, { "grad_norm": 0.3861353099346161, "learning_rate": 5.518421471276596e-05, "loss": 0.037, "step": 49360 }, { "grad_norm": 0.3906310200691223, "learning_rate": 5.5167768849162174e-05, "loss": 0.0384, "step": 49370 }, { "grad_norm": 0.45671626925468445, "learning_rate": 5.515132242041893e-05, "loss": 0.037, "step": 49380 }, { "grad_norm": 0.7360225915908813, "learning_rate": 5.513487542833483e-05, "loss": 0.0415, "step": 49390 }, { "grad_norm": 0.5943183898925781, "learning_rate": 5.5118427874708466e-05, "loss": 0.0456, "step": 49400 }, { "grad_norm": 1.2734026908874512, "learning_rate": 5.510197976133853e-05, "loss": 0.0436, "step": 49410 }, { "grad_norm": 0.544469952583313, "learning_rate": 5.508553109002376e-05, "loss": 0.0345, "step": 49420 }, { "grad_norm": 0.6314529180526733, "learning_rate": 5.5069081862562957e-05, "loss": 0.0486, "step": 49430 }, { "grad_norm": 0.658308744430542, "learning_rate": 5.5052632080754965e-05, "loss": 0.0553, "step": 49440 }, { "grad_norm": 0.5053730607032776, "learning_rate": 5.503618174639874e-05, "loss": 0.0323, "step": 49450 }, { "grad_norm": 0.7022899985313416, "learning_rate": 5.501973086129325e-05, "loss": 0.0435, "step": 49460 }, { "grad_norm": 0.5908603668212891, "learning_rate": 5.500327942723753e-05, "loss": 0.0384, "step": 49470 }, { "grad_norm": 0.5978016257286072, "learning_rate": 5.498682744603071e-05, "loss": 0.0408, "step": 49480 }, { "grad_norm": 0.40422526001930237, "learning_rate": 5.497037491947194e-05, "loss": 0.0487, "step": 49490 }, { "grad_norm": 1.0561925172805786, "learning_rate": 5.4953921849360424e-05, "loss": 0.0415, "step": 49500 }, { "grad_norm": 0.6208935976028442, "learning_rate": 5.493746823749547e-05, "loss": 0.043, "step": 49510 }, { "grad_norm": 0.7511846423149109, "learning_rate": 5.49210140856764e-05, "loss": 0.0516, "step": 49520 }, { "grad_norm": 0.61715167760849, "learning_rate": 5.4904559395702647e-05, "loss": 0.0485, "step": 49530 }, { "grad_norm": 0.31451815366744995, "learning_rate": 5.488810416937364e-05, "loss": 0.0375, "step": 49540 }, { "grad_norm": 0.6374552845954895, "learning_rate": 5.4871648408488926e-05, "loss": 0.0383, "step": 49550 }, { "grad_norm": 0.42200636863708496, "learning_rate": 5.485519211484807e-05, "loss": 0.0404, "step": 49560 }, { "grad_norm": 1.272242784500122, "learning_rate": 5.4838735290250676e-05, "loss": 0.0481, "step": 49570 }, { "grad_norm": 0.6999809741973877, "learning_rate": 5.482227793649649e-05, "loss": 0.0483, "step": 49580 }, { "grad_norm": 0.49818310141563416, "learning_rate": 5.480582005538524e-05, "loss": 0.0359, "step": 49590 }, { "grad_norm": 0.4260682165622711, "learning_rate": 5.478936164871671e-05, "loss": 0.039, "step": 49600 }, { "grad_norm": 0.24754944443702698, "learning_rate": 5.477290271829081e-05, "loss": 0.0304, "step": 49610 }, { "grad_norm": 0.5089380741119385, "learning_rate": 5.475644326590744e-05, "loss": 0.0435, "step": 49620 }, { "grad_norm": 0.7974507212638855, "learning_rate": 5.473998329336658e-05, "loss": 0.0335, "step": 49630 }, { "grad_norm": 0.4953862130641937, "learning_rate": 5.4723522802468286e-05, "loss": 0.0376, "step": 49640 }, { "grad_norm": 0.7585790157318115, "learning_rate": 5.470706179501264e-05, "loss": 0.0546, "step": 49650 }, { "grad_norm": 0.28757503628730774, "learning_rate": 5.4690600272799776e-05, "loss": 0.0311, "step": 49660 }, { "grad_norm": 0.49198490381240845, "learning_rate": 5.467413823762993e-05, "loss": 0.0374, "step": 49670 }, { "grad_norm": 0.6183354258537292, "learning_rate": 5.465767569130335e-05, "loss": 0.0352, "step": 49680 }, { "grad_norm": 0.697134792804718, "learning_rate": 5.464121263562036e-05, "loss": 0.0334, "step": 49690 }, { "grad_norm": 0.3985814154148102, "learning_rate": 5.4624749072381343e-05, "loss": 0.0348, "step": 49700 }, { "grad_norm": 0.5676522850990295, "learning_rate": 5.460828500338672e-05, "loss": 0.0391, "step": 49710 }, { "grad_norm": 0.5880598425865173, "learning_rate": 5.459182043043698e-05, "loss": 0.0374, "step": 49720 }, { "grad_norm": 0.5249411463737488, "learning_rate": 5.457535535533265e-05, "loss": 0.0436, "step": 49730 }, { "grad_norm": 0.4142945408821106, "learning_rate": 5.4558889779874334e-05, "loss": 0.0519, "step": 49740 }, { "grad_norm": 0.4579707086086273, "learning_rate": 5.454242370586269e-05, "loss": 0.034, "step": 49750 }, { "grad_norm": 0.4002920091152191, "learning_rate": 5.452595713509843e-05, "loss": 0.044, "step": 49760 }, { "grad_norm": 0.4849781394004822, "learning_rate": 5.4509490069382274e-05, "loss": 0.0434, "step": 49770 }, { "grad_norm": 0.5791706442832947, "learning_rate": 5.449302251051509e-05, "loss": 0.0354, "step": 49780 }, { "grad_norm": 0.4698072373867035, "learning_rate": 5.44765544602977e-05, "loss": 0.0402, "step": 49790 }, { "grad_norm": 0.4044833183288574, "learning_rate": 5.446008592053107e-05, "loss": 0.036, "step": 49800 }, { "grad_norm": 0.278633177280426, "learning_rate": 5.4443616893016135e-05, "loss": 0.0257, "step": 49810 }, { "grad_norm": 0.5524589419364929, "learning_rate": 5.4427147379553934e-05, "loss": 0.0364, "step": 49820 }, { "grad_norm": 0.4085260033607483, "learning_rate": 5.441067738194556e-05, "loss": 0.0335, "step": 49830 }, { "grad_norm": 0.49241623282432556, "learning_rate": 5.439420690199214e-05, "loss": 0.0436, "step": 49840 }, { "grad_norm": 0.5601370930671692, "learning_rate": 5.437773594149484e-05, "loss": 0.0353, "step": 49850 }, { "grad_norm": 0.4153933525085449, "learning_rate": 5.436126450225495e-05, "loss": 0.0397, "step": 49860 }, { "grad_norm": 0.5688595771789551, "learning_rate": 5.434479258607371e-05, "loss": 0.0481, "step": 49870 }, { "grad_norm": 0.425173819065094, "learning_rate": 5.43283201947525e-05, "loss": 0.0369, "step": 49880 }, { "grad_norm": 0.20398737490177155, "learning_rate": 5.431184733009268e-05, "loss": 0.0279, "step": 49890 }, { "grad_norm": 0.43293043971061707, "learning_rate": 5.4295373993895736e-05, "loss": 0.0489, "step": 49900 }, { "grad_norm": 0.5167643427848816, "learning_rate": 5.4278900187963157e-05, "loss": 0.0328, "step": 49910 }, { "grad_norm": 0.5152075886726379, "learning_rate": 5.426242591409646e-05, "loss": 0.0441, "step": 49920 }, { "grad_norm": 0.5273408889770508, "learning_rate": 5.42459511740973e-05, "loss": 0.034, "step": 49930 }, { "grad_norm": 0.26172497868537903, "learning_rate": 5.422947596976729e-05, "loss": 0.0338, "step": 49940 }, { "grad_norm": 0.6180300712585449, "learning_rate": 5.4213000302908134e-05, "loss": 0.0475, "step": 49950 }, { "grad_norm": 0.5980921983718872, "learning_rate": 5.419652417532162e-05, "loss": 0.031, "step": 49960 }, { "grad_norm": 0.48191943764686584, "learning_rate": 5.4180047588809534e-05, "loss": 0.0374, "step": 49970 }, { "grad_norm": 0.5687133073806763, "learning_rate": 5.4163570545173704e-05, "loss": 0.0538, "step": 49980 }, { "grad_norm": 0.3373957872390747, "learning_rate": 5.414709304621608e-05, "loss": 0.0396, "step": 49990 }, { "grad_norm": 0.8523396849632263, "learning_rate": 5.413061509373858e-05, "loss": 0.0509, "step": 50000 }, { "grad_norm": 0.5883761048316956, "learning_rate": 5.411413668954324e-05, "loss": 0.0446, "step": 50010 }, { "grad_norm": 0.36262163519859314, "learning_rate": 5.4097657835432083e-05, "loss": 0.0332, "step": 50020 }, { "grad_norm": 0.6462728381156921, "learning_rate": 5.408117853320723e-05, "loss": 0.0361, "step": 50030 }, { "grad_norm": 0.39720964431762695, "learning_rate": 5.406469878467084e-05, "loss": 0.0493, "step": 50040 }, { "grad_norm": 0.5354596376419067, "learning_rate": 5.404821859162509e-05, "loss": 0.0311, "step": 50050 }, { "grad_norm": 0.5994185209274292, "learning_rate": 5.403173795587225e-05, "loss": 0.0564, "step": 50060 }, { "grad_norm": 1.2904597520828247, "learning_rate": 5.4015256879214606e-05, "loss": 0.034, "step": 50070 }, { "grad_norm": 0.5160858035087585, "learning_rate": 5.39987753634545e-05, "loss": 0.0306, "step": 50080 }, { "grad_norm": 0.47026392817497253, "learning_rate": 5.3982293410394335e-05, "loss": 0.0447, "step": 50090 }, { "grad_norm": 0.38916465640068054, "learning_rate": 5.396581102183655e-05, "loss": 0.0342, "step": 50100 }, { "grad_norm": 0.959515392780304, "learning_rate": 5.3949328199583634e-05, "loss": 0.045, "step": 50110 }, { "grad_norm": 0.6960296630859375, "learning_rate": 5.393284494543813e-05, "loss": 0.0348, "step": 50120 }, { "grad_norm": 0.5650599598884583, "learning_rate": 5.391636126120262e-05, "loss": 0.0386, "step": 50130 }, { "grad_norm": 0.4246739447116852, "learning_rate": 5.389987714867971e-05, "loss": 0.0285, "step": 50140 }, { "grad_norm": 0.8440026044845581, "learning_rate": 5.3883392609672114e-05, "loss": 0.0419, "step": 50150 }, { "grad_norm": 0.5990544557571411, "learning_rate": 5.3866907645982545e-05, "loss": 0.0557, "step": 50160 }, { "grad_norm": 0.741486668586731, "learning_rate": 5.385042225941375e-05, "loss": 0.0352, "step": 50170 }, { "grad_norm": 0.8403356671333313, "learning_rate": 5.3833936451768585e-05, "loss": 0.051, "step": 50180 }, { "grad_norm": 0.4251073896884918, "learning_rate": 5.3817450224849884e-05, "loss": 0.0373, "step": 50190 }, { "grad_norm": 0.24752992391586304, "learning_rate": 5.380096358046056e-05, "loss": 0.041, "step": 50200 }, { "grad_norm": 0.4795016050338745, "learning_rate": 5.378447652040359e-05, "loss": 0.0385, "step": 50210 }, { "grad_norm": 0.5545260906219482, "learning_rate": 5.376798904648195e-05, "loss": 0.0348, "step": 50220 }, { "grad_norm": 0.5114573240280151, "learning_rate": 5.375150116049869e-05, "loss": 0.033, "step": 50230 }, { "grad_norm": 0.7621128559112549, "learning_rate": 5.373501286425691e-05, "loss": 0.0389, "step": 50240 }, { "grad_norm": 0.368707537651062, "learning_rate": 5.3718524159559726e-05, "loss": 0.0597, "step": 50250 }, { "grad_norm": 0.20687571167945862, "learning_rate": 5.370203504821034e-05, "loss": 0.035, "step": 50260 }, { "grad_norm": 0.5643267035484314, "learning_rate": 5.368554553201196e-05, "loss": 0.0373, "step": 50270 }, { "grad_norm": 0.5563617944717407, "learning_rate": 5.3669055612767874e-05, "loss": 0.038, "step": 50280 }, { "grad_norm": 0.4798167645931244, "learning_rate": 5.3652565292281376e-05, "loss": 0.024, "step": 50290 }, { "grad_norm": 0.6857990622520447, "learning_rate": 5.3636074572355824e-05, "loss": 0.0324, "step": 50300 }, { "grad_norm": 0.41297441720962524, "learning_rate": 5.361958345479463e-05, "loss": 0.037, "step": 50310 }, { "grad_norm": 0.33958718180656433, "learning_rate": 5.360309194140123e-05, "loss": 0.0227, "step": 50320 }, { "grad_norm": 0.5194748044013977, "learning_rate": 5.3586600033979106e-05, "loss": 0.0388, "step": 50330 }, { "grad_norm": 0.574190616607666, "learning_rate": 5.357010773433181e-05, "loss": 0.0445, "step": 50340 }, { "grad_norm": 0.7276344299316406, "learning_rate": 5.3553615044262895e-05, "loss": 0.0352, "step": 50350 }, { "grad_norm": 0.23941200971603394, "learning_rate": 5.3537121965575964e-05, "loss": 0.03, "step": 50360 }, { "grad_norm": 0.39396941661834717, "learning_rate": 5.352062850007471e-05, "loss": 0.044, "step": 50370 }, { "grad_norm": 0.3443875312805176, "learning_rate": 5.350413464956282e-05, "loss": 0.0295, "step": 50380 }, { "grad_norm": 0.39099353551864624, "learning_rate": 5.348764041584403e-05, "loss": 0.0365, "step": 50390 }, { "grad_norm": 0.5306743383407593, "learning_rate": 5.347114580072212e-05, "loss": 0.0308, "step": 50400 }, { "grad_norm": 0.3303665816783905, "learning_rate": 5.345465080600093e-05, "loss": 0.0291, "step": 50410 }, { "grad_norm": 0.4857589900493622, "learning_rate": 5.3438155433484326e-05, "loss": 0.0435, "step": 50420 }, { "grad_norm": 0.2839222550392151, "learning_rate": 5.3421659684976197e-05, "loss": 0.0312, "step": 50430 }, { "grad_norm": 0.4740869700908661, "learning_rate": 5.340516356228052e-05, "loss": 0.0277, "step": 50440 }, { "grad_norm": 0.7144643664360046, "learning_rate": 5.338866706720128e-05, "loss": 0.0532, "step": 50450 }, { "grad_norm": 0.48283860087394714, "learning_rate": 5.337217020154249e-05, "loss": 0.0461, "step": 50460 }, { "grad_norm": 0.3444746732711792, "learning_rate": 5.335567296710825e-05, "loss": 0.0285, "step": 50470 }, { "grad_norm": 0.43764787912368774, "learning_rate": 5.333917536570265e-05, "loss": 0.0427, "step": 50480 }, { "grad_norm": 0.3801107108592987, "learning_rate": 5.332267739912986e-05, "loss": 0.039, "step": 50490 }, { "grad_norm": 0.4830816090106964, "learning_rate": 5.330617906919405e-05, "loss": 0.026, "step": 50500 }, { "grad_norm": 0.3492610454559326, "learning_rate": 5.328968037769949e-05, "loss": 0.0332, "step": 50510 }, { "grad_norm": 0.4192330539226532, "learning_rate": 5.327318132645042e-05, "loss": 0.0319, "step": 50520 }, { "grad_norm": 0.5636561512947083, "learning_rate": 5.325668191725116e-05, "loss": 0.0325, "step": 50530 }, { "grad_norm": 0.6887425780296326, "learning_rate": 5.324018215190606e-05, "loss": 0.0429, "step": 50540 }, { "grad_norm": 0.3543698787689209, "learning_rate": 5.3223682032219515e-05, "loss": 0.0369, "step": 50550 }, { "grad_norm": 0.3032907247543335, "learning_rate": 5.320718155999595e-05, "loss": 0.031, "step": 50560 }, { "grad_norm": 0.4944036900997162, "learning_rate": 5.3190680737039835e-05, "loss": 0.0278, "step": 50570 }, { "grad_norm": 0.3244442939758301, "learning_rate": 5.317417956515567e-05, "loss": 0.0352, "step": 50580 }, { "grad_norm": 0.48005548119544983, "learning_rate": 5.315767804614798e-05, "loss": 0.04, "step": 50590 }, { "grad_norm": 0.4162504971027374, "learning_rate": 5.3141176181821395e-05, "loss": 0.0295, "step": 50600 }, { "grad_norm": 0.397670716047287, "learning_rate": 5.31246739739805e-05, "loss": 0.0235, "step": 50610 }, { "grad_norm": 0.7200835943222046, "learning_rate": 5.310817142442994e-05, "loss": 0.0348, "step": 50620 }, { "grad_norm": 0.6245437264442444, "learning_rate": 5.309166853497445e-05, "loss": 0.0341, "step": 50630 }, { "grad_norm": 0.4473285675048828, "learning_rate": 5.307516530741873e-05, "loss": 0.0354, "step": 50640 }, { "grad_norm": 0.34942173957824707, "learning_rate": 5.305866174356754e-05, "loss": 0.0555, "step": 50650 }, { "grad_norm": 0.5437353253364563, "learning_rate": 5.304215784522571e-05, "loss": 0.04, "step": 50660 }, { "grad_norm": 0.5247125029563904, "learning_rate": 5.302565361419808e-05, "loss": 0.0374, "step": 50670 }, { "grad_norm": 0.3573973774909973, "learning_rate": 5.3009149052289507e-05, "loss": 0.0328, "step": 50680 }, { "grad_norm": 0.38040444254875183, "learning_rate": 5.299264416130493e-05, "loss": 0.0347, "step": 50690 }, { "grad_norm": 0.40872451663017273, "learning_rate": 5.297613894304928e-05, "loss": 0.0475, "step": 50700 }, { "grad_norm": 0.5830802917480469, "learning_rate": 5.2959633399327534e-05, "loss": 0.0347, "step": 50710 }, { "grad_norm": 0.4877784550189972, "learning_rate": 5.294312753194476e-05, "loss": 0.0351, "step": 50720 }, { "grad_norm": 0.46108904480934143, "learning_rate": 5.292662134270596e-05, "loss": 0.0447, "step": 50730 }, { "grad_norm": 0.5989847183227539, "learning_rate": 5.291011483341626e-05, "loss": 0.0332, "step": 50740 }, { "grad_norm": 0.536078929901123, "learning_rate": 5.2893608005880767e-05, "loss": 0.03, "step": 50750 }, { "grad_norm": 0.5773659348487854, "learning_rate": 5.287710086190467e-05, "loss": 0.0326, "step": 50760 }, { "grad_norm": 0.8738128542900085, "learning_rate": 5.286059340329314e-05, "loss": 0.0333, "step": 50770 }, { "grad_norm": 0.6311296820640564, "learning_rate": 5.284408563185141e-05, "loss": 0.0378, "step": 50780 }, { "grad_norm": 0.46876177191734314, "learning_rate": 5.282757754938476e-05, "loss": 0.0356, "step": 50790 }, { "grad_norm": 0.38858136534690857, "learning_rate": 5.281106915769849e-05, "loss": 0.0269, "step": 50800 }, { "grad_norm": 0.45930221676826477, "learning_rate": 5.2794560458597897e-05, "loss": 0.0405, "step": 50810 }, { "grad_norm": 0.6541167497634888, "learning_rate": 5.27780514538884e-05, "loss": 0.0336, "step": 50820 }, { "grad_norm": 0.3421066701412201, "learning_rate": 5.2761542145375365e-05, "loss": 0.0291, "step": 50830 }, { "grad_norm": 0.438493013381958, "learning_rate": 5.274503253486421e-05, "loss": 0.0381, "step": 50840 }, { "grad_norm": 0.5527797937393188, "learning_rate": 5.272852262416046e-05, "loss": 0.0398, "step": 50850 }, { "grad_norm": 0.4970887303352356, "learning_rate": 5.2712012415069555e-05, "loss": 0.0486, "step": 50860 }, { "grad_norm": 0.5622032284736633, "learning_rate": 5.269550190939705e-05, "loss": 0.0465, "step": 50870 }, { "grad_norm": 0.4747670590877533, "learning_rate": 5.267899110894852e-05, "loss": 0.0318, "step": 50880 }, { "grad_norm": 0.3503338098526001, "learning_rate": 5.266248001552955e-05, "loss": 0.055, "step": 50890 }, { "grad_norm": 0.4111098647117615, "learning_rate": 5.264596863094575e-05, "loss": 0.0355, "step": 50900 }, { "grad_norm": 0.3666824698448181, "learning_rate": 5.262945695700282e-05, "loss": 0.0327, "step": 50910 }, { "grad_norm": 0.3038397431373596, "learning_rate": 5.261294499550643e-05, "loss": 0.0314, "step": 50920 }, { "grad_norm": 0.4784640371799469, "learning_rate": 5.25964327482623e-05, "loss": 0.0421, "step": 50930 }, { "grad_norm": 0.6227097511291504, "learning_rate": 5.257992021707617e-05, "loss": 0.0381, "step": 50940 }, { "grad_norm": 0.5005767345428467, "learning_rate": 5.256340740375387e-05, "loss": 0.0303, "step": 50950 }, { "grad_norm": 0.4637351930141449, "learning_rate": 5.254689431010117e-05, "loss": 0.0434, "step": 50960 }, { "grad_norm": 0.4228936433792114, "learning_rate": 5.253038093792395e-05, "loss": 0.0305, "step": 50970 }, { "grad_norm": 0.33720970153808594, "learning_rate": 5.251386728902806e-05, "loss": 0.0265, "step": 50980 }, { "grad_norm": 0.44191455841064453, "learning_rate": 5.2497353365219446e-05, "loss": 0.0263, "step": 50990 }, { "grad_norm": 0.4846437871456146, "learning_rate": 5.2480839168304e-05, "loss": 0.038, "step": 51000 }, { "grad_norm": 0.5377728343009949, "learning_rate": 5.246432470008773e-05, "loss": 0.0366, "step": 51010 }, { "grad_norm": 0.5073580145835876, "learning_rate": 5.2447809962376616e-05, "loss": 0.0364, "step": 51020 }, { "grad_norm": 0.6189533472061157, "learning_rate": 5.243129495697666e-05, "loss": 0.0339, "step": 51030 }, { "grad_norm": 0.34049054980278015, "learning_rate": 5.241477968569397e-05, "loss": 0.0322, "step": 51040 }, { "grad_norm": 0.8140254616737366, "learning_rate": 5.23982641503346e-05, "loss": 0.0444, "step": 51050 }, { "grad_norm": 0.5427805185317993, "learning_rate": 5.238174835270465e-05, "loss": 0.0373, "step": 51060 }, { "grad_norm": 0.4352009892463684, "learning_rate": 5.23652322946103e-05, "loss": 0.0302, "step": 51070 }, { "grad_norm": 0.6791859865188599, "learning_rate": 5.234871597785771e-05, "loss": 0.0456, "step": 51080 }, { "grad_norm": 0.33739563822746277, "learning_rate": 5.233219940425306e-05, "loss": 0.0289, "step": 51090 }, { "grad_norm": 0.6071344614028931, "learning_rate": 5.2315682575602576e-05, "loss": 0.0353, "step": 51100 }, { "grad_norm": 0.44420522451400757, "learning_rate": 5.229916549371254e-05, "loss": 0.0345, "step": 51110 }, { "grad_norm": 0.4277038276195526, "learning_rate": 5.228264816038924e-05, "loss": 0.0493, "step": 51120 }, { "grad_norm": 0.4541788101196289, "learning_rate": 5.2266130577438935e-05, "loss": 0.0314, "step": 51130 }, { "grad_norm": 0.9759305119514465, "learning_rate": 5.224961274666802e-05, "loss": 0.0361, "step": 51140 }, { "grad_norm": 0.5096675753593445, "learning_rate": 5.223309466988283e-05, "loss": 0.033, "step": 51150 }, { "grad_norm": 0.2821183502674103, "learning_rate": 5.221657634888976e-05, "loss": 0.0378, "step": 51160 }, { "grad_norm": 0.4798392951488495, "learning_rate": 5.2200057785495235e-05, "loss": 0.0327, "step": 51170 }, { "grad_norm": 1.4802261590957642, "learning_rate": 5.2183538981505706e-05, "loss": 0.0362, "step": 51180 }, { "grad_norm": 0.4437905251979828, "learning_rate": 5.216701993872762e-05, "loss": 0.0377, "step": 51190 }, { "grad_norm": 0.6168032884597778, "learning_rate": 5.215050065896752e-05, "loss": 0.0332, "step": 51200 }, { "grad_norm": 0.5725601315498352, "learning_rate": 5.2133981144031875e-05, "loss": 0.0438, "step": 51210 }, { "grad_norm": 0.4898439049720764, "learning_rate": 5.211746139572727e-05, "loss": 0.0393, "step": 51220 }, { "grad_norm": 0.5703569650650024, "learning_rate": 5.210094141586026e-05, "loss": 0.0483, "step": 51230 }, { "grad_norm": 0.6290965676307678, "learning_rate": 5.2084421206237476e-05, "loss": 0.0318, "step": 51240 }, { "grad_norm": 0.5392075777053833, "learning_rate": 5.20679007686655e-05, "loss": 0.0449, "step": 51250 }, { "grad_norm": 0.5733930468559265, "learning_rate": 5.205138010495102e-05, "loss": 0.0552, "step": 51260 }, { "grad_norm": 0.4310639798641205, "learning_rate": 5.20348592169007e-05, "loss": 0.032, "step": 51270 }, { "grad_norm": 0.7705879211425781, "learning_rate": 5.201833810632123e-05, "loss": 0.0284, "step": 51280 }, { "grad_norm": 0.39881834387779236, "learning_rate": 5.200181677501933e-05, "loss": 0.0278, "step": 51290 }, { "grad_norm": 0.4822031855583191, "learning_rate": 5.1985295224801777e-05, "loss": 0.0473, "step": 51300 }, { "grad_norm": 0.4112773835659027, "learning_rate": 5.196877345747531e-05, "loss": 0.0267, "step": 51310 }, { "grad_norm": 0.9401116371154785, "learning_rate": 5.195225147484674e-05, "loss": 0.0348, "step": 51320 }, { "grad_norm": 0.6305873394012451, "learning_rate": 5.193572927872288e-05, "loss": 0.0448, "step": 51330 }, { "grad_norm": 0.38819238543510437, "learning_rate": 5.191920687091059e-05, "loss": 0.0391, "step": 51340 }, { "grad_norm": 0.5143709182739258, "learning_rate": 5.1902684253216695e-05, "loss": 0.0332, "step": 51350 }, { "grad_norm": 0.7980589866638184, "learning_rate": 5.1886161427448134e-05, "loss": 0.0387, "step": 51360 }, { "grad_norm": 0.3321519196033478, "learning_rate": 5.186963839541179e-05, "loss": 0.0269, "step": 51370 }, { "grad_norm": 0.465671569108963, "learning_rate": 5.185311515891459e-05, "loss": 0.0458, "step": 51380 }, { "grad_norm": 0.6703530550003052, "learning_rate": 5.1836591719763516e-05, "loss": 0.0358, "step": 51390 }, { "grad_norm": 0.4378802478313446, "learning_rate": 5.182006807976554e-05, "loss": 0.0297, "step": 51400 }, { "grad_norm": 0.6722027659416199, "learning_rate": 5.180354424072763e-05, "loss": 0.0461, "step": 51410 }, { "grad_norm": 0.7281163930892944, "learning_rate": 5.178702020445686e-05, "loss": 0.0383, "step": 51420 }, { "grad_norm": 0.37970879673957825, "learning_rate": 5.177049597276024e-05, "loss": 0.0386, "step": 51430 }, { "grad_norm": 0.518791913986206, "learning_rate": 5.175397154744482e-05, "loss": 0.0279, "step": 51440 }, { "grad_norm": 0.3862539529800415, "learning_rate": 5.1737446930317736e-05, "loss": 0.0521, "step": 51450 }, { "grad_norm": 1.230769157409668, "learning_rate": 5.172092212318604e-05, "loss": 0.0477, "step": 51460 }, { "grad_norm": 0.7228604555130005, "learning_rate": 5.170439712785692e-05, "loss": 0.037, "step": 51470 }, { "grad_norm": 0.5881946086883545, "learning_rate": 5.1687871946137467e-05, "loss": 0.032, "step": 51480 }, { "grad_norm": 0.3907802999019623, "learning_rate": 5.16713465798349e-05, "loss": 0.0368, "step": 51490 }, { "grad_norm": 0.5236990451812744, "learning_rate": 5.165482103075637e-05, "loss": 0.0351, "step": 51500 }, { "grad_norm": 0.5398353934288025, "learning_rate": 5.163829530070909e-05, "loss": 0.0399, "step": 51510 }, { "grad_norm": 0.6162451505661011, "learning_rate": 5.162176939150032e-05, "loss": 0.034, "step": 51520 }, { "grad_norm": 0.4700557291507721, "learning_rate": 5.160524330493728e-05, "loss": 0.0308, "step": 51530 }, { "grad_norm": 0.6160517930984497, "learning_rate": 5.158871704282724e-05, "loss": 0.0316, "step": 51540 }, { "grad_norm": 0.5862358212471008, "learning_rate": 5.15721906069775e-05, "loss": 0.0292, "step": 51550 }, { "grad_norm": 0.7198232412338257, "learning_rate": 5.155566399919536e-05, "loss": 0.0263, "step": 51560 }, { "grad_norm": 0.5939303636550903, "learning_rate": 5.153913722128813e-05, "loss": 0.0286, "step": 51570 }, { "grad_norm": 0.4634632170200348, "learning_rate": 5.1522610275063176e-05, "loss": 0.0359, "step": 51580 }, { "grad_norm": 0.5826991200447083, "learning_rate": 5.150608316232786e-05, "loss": 0.0267, "step": 51590 }, { "grad_norm": 0.3123319149017334, "learning_rate": 5.148955588488955e-05, "loss": 0.0306, "step": 51600 }, { "grad_norm": 0.27517250180244446, "learning_rate": 5.147302844455564e-05, "loss": 0.0329, "step": 51610 }, { "grad_norm": 0.472941130399704, "learning_rate": 5.145650084313356e-05, "loss": 0.0368, "step": 51620 }, { "grad_norm": 0.6594412922859192, "learning_rate": 5.1439973082430734e-05, "loss": 0.0375, "step": 51630 }, { "grad_norm": 0.41132938861846924, "learning_rate": 5.142344516425461e-05, "loss": 0.0253, "step": 51640 }, { "grad_norm": 0.5651567578315735, "learning_rate": 5.1406917090412673e-05, "loss": 0.0378, "step": 51650 }, { "grad_norm": 1.2997735738754272, "learning_rate": 5.1390388862712404e-05, "loss": 0.0281, "step": 51660 }, { "grad_norm": 0.6442931890487671, "learning_rate": 5.137386048296129e-05, "loss": 0.0495, "step": 51670 }, { "grad_norm": 0.5473443269729614, "learning_rate": 5.135733195296686e-05, "loss": 0.0271, "step": 51680 }, { "grad_norm": 0.5485981106758118, "learning_rate": 5.134080327453664e-05, "loss": 0.0381, "step": 51690 }, { "grad_norm": 0.33751180768013, "learning_rate": 5.1324274449478204e-05, "loss": 0.0391, "step": 51700 }, { "grad_norm": 0.5715444087982178, "learning_rate": 5.130774547959909e-05, "loss": 0.0346, "step": 51710 }, { "grad_norm": 0.5521735548973083, "learning_rate": 5.129121636670692e-05, "loss": 0.0352, "step": 51720 }, { "grad_norm": 0.5272700190544128, "learning_rate": 5.1274687112609246e-05, "loss": 0.0436, "step": 51730 }, { "grad_norm": 0.41997992992401123, "learning_rate": 5.125815771911372e-05, "loss": 0.049, "step": 51740 }, { "grad_norm": 0.27517345547676086, "learning_rate": 5.1241628188027956e-05, "loss": 0.0475, "step": 51750 }, { "grad_norm": 0.49335071444511414, "learning_rate": 5.1225098521159585e-05, "loss": 0.0364, "step": 51760 }, { "grad_norm": 0.7715234160423279, "learning_rate": 5.12085687203163e-05, "loss": 0.0379, "step": 51770 }, { "grad_norm": 0.35750967264175415, "learning_rate": 5.119203878730575e-05, "loss": 0.0249, "step": 51780 }, { "grad_norm": 0.5043384432792664, "learning_rate": 5.117550872393564e-05, "loss": 0.0295, "step": 51790 }, { "grad_norm": 0.3861471116542816, "learning_rate": 5.1158978532013645e-05, "loss": 0.0323, "step": 51800 }, { "grad_norm": 0.47999274730682373, "learning_rate": 5.114244821334751e-05, "loss": 0.0365, "step": 51810 }, { "grad_norm": 0.4810296893119812, "learning_rate": 5.112591776974497e-05, "loss": 0.039, "step": 51820 }, { "grad_norm": 0.46175751090049744, "learning_rate": 5.110938720301374e-05, "loss": 0.0264, "step": 51830 }, { "grad_norm": 0.5612702369689941, "learning_rate": 5.1092856514961595e-05, "loss": 0.0465, "step": 51840 }, { "grad_norm": 0.39344313740730286, "learning_rate": 5.107632570739631e-05, "loss": 0.0356, "step": 51850 }, { "grad_norm": 0.5159633755683899, "learning_rate": 5.1059794782125645e-05, "loss": 0.0381, "step": 51860 }, { "grad_norm": 0.5424600839614868, "learning_rate": 5.104326374095744e-05, "loss": 0.0364, "step": 51870 }, { "grad_norm": 0.5590403079986572, "learning_rate": 5.102673258569947e-05, "loss": 0.0339, "step": 51880 }, { "grad_norm": 0.5357734560966492, "learning_rate": 5.101020131815956e-05, "loss": 0.0348, "step": 51890 }, { "grad_norm": 0.31324276328086853, "learning_rate": 5.099366994014556e-05, "loss": 0.0313, "step": 51900 }, { "grad_norm": 0.4766724705696106, "learning_rate": 5.097713845346532e-05, "loss": 0.0365, "step": 51910 }, { "grad_norm": 0.30627018213272095, "learning_rate": 5.096060685992668e-05, "loss": 0.032, "step": 51920 }, { "grad_norm": 0.6238638162612915, "learning_rate": 5.0944075161337514e-05, "loss": 0.0433, "step": 51930 }, { "grad_norm": 0.57297682762146, "learning_rate": 5.09275433595057e-05, "loss": 0.0442, "step": 51940 }, { "grad_norm": 0.4911951720714569, "learning_rate": 5.0911011456239157e-05, "loss": 0.0271, "step": 51950 }, { "grad_norm": 0.40849706530570984, "learning_rate": 5.089447945334574e-05, "loss": 0.0375, "step": 51960 }, { "grad_norm": 0.5946373343467712, "learning_rate": 5.087794735263343e-05, "loss": 0.0462, "step": 51970 }, { "grad_norm": 0.4054684340953827, "learning_rate": 5.086141515591011e-05, "loss": 0.0363, "step": 51980 }, { "grad_norm": 0.4317164123058319, "learning_rate": 5.084488286498371e-05, "loss": 0.035, "step": 51990 }, { "grad_norm": 0.6671044826507568, "learning_rate": 5.08283504816622e-05, "loss": 0.039, "step": 52000 }, { "grad_norm": 0.6209673285484314, "learning_rate": 5.081181800775353e-05, "loss": 0.0319, "step": 52010 }, { "grad_norm": 0.6144553422927856, "learning_rate": 5.079528544506565e-05, "loss": 0.0478, "step": 52020 }, { "grad_norm": 0.6712954640388489, "learning_rate": 5.077875279540657e-05, "loss": 0.0423, "step": 52030 }, { "grad_norm": 0.6657041907310486, "learning_rate": 5.076222006058424e-05, "loss": 0.0369, "step": 52040 }, { "grad_norm": 0.38783952593803406, "learning_rate": 5.074568724240667e-05, "loss": 0.0274, "step": 52050 }, { "grad_norm": 0.35123005509376526, "learning_rate": 5.072915434268187e-05, "loss": 0.0361, "step": 52060 }, { "grad_norm": 0.6083913445472717, "learning_rate": 5.071262136321785e-05, "loss": 0.0367, "step": 52070 }, { "grad_norm": 0.45327913761138916, "learning_rate": 5.069608830582262e-05, "loss": 0.0386, "step": 52080 }, { "grad_norm": 0.446470707654953, "learning_rate": 5.067955517230423e-05, "loss": 0.0329, "step": 52090 }, { "grad_norm": 0.6950271129608154, "learning_rate": 5.06630219644707e-05, "loss": 0.0453, "step": 52100 }, { "grad_norm": 0.38778501749038696, "learning_rate": 5.064648868413008e-05, "loss": 0.0312, "step": 52110 }, { "grad_norm": 0.7739783525466919, "learning_rate": 5.062995533309044e-05, "loss": 0.031, "step": 52120 }, { "grad_norm": 0.6635921597480774, "learning_rate": 5.061342191315983e-05, "loss": 0.0517, "step": 52130 }, { "grad_norm": 0.6075290441513062, "learning_rate": 5.0596888426146325e-05, "loss": 0.0403, "step": 52140 }, { "grad_norm": 0.4279981553554535, "learning_rate": 5.058035487385798e-05, "loss": 0.0394, "step": 52150 }, { "grad_norm": 0.562356173992157, "learning_rate": 5.0563821258102915e-05, "loss": 0.0724, "step": 52160 }, { "grad_norm": 0.5275769233703613, "learning_rate": 5.05472875806892e-05, "loss": 0.0405, "step": 52170 }, { "grad_norm": 0.2821164131164551, "learning_rate": 5.053075384342494e-05, "loss": 0.0489, "step": 52180 }, { "grad_norm": 0.4975047707557678, "learning_rate": 5.0514220048118234e-05, "loss": 0.0254, "step": 52190 }, { "grad_norm": 0.33837804198265076, "learning_rate": 5.049768619657721e-05, "loss": 0.0441, "step": 52200 }, { "grad_norm": 0.3734913468360901, "learning_rate": 5.048115229060996e-05, "loss": 0.0266, "step": 52210 }, { "grad_norm": 0.41836291551589966, "learning_rate": 5.046461833202462e-05, "loss": 0.0293, "step": 52220 }, { "grad_norm": 0.500074028968811, "learning_rate": 5.044808432262932e-05, "loss": 0.0323, "step": 52230 }, { "grad_norm": 0.3890972137451172, "learning_rate": 5.043155026423219e-05, "loss": 0.042, "step": 52240 }, { "grad_norm": 0.33568641543388367, "learning_rate": 5.041501615864137e-05, "loss": 0.0362, "step": 52250 }, { "grad_norm": 0.5428494811058044, "learning_rate": 5.039848200766502e-05, "loss": 0.0297, "step": 52260 }, { "grad_norm": 0.5897196531295776, "learning_rate": 5.038194781311125e-05, "loss": 0.0428, "step": 52270 }, { "grad_norm": 0.5633234977722168, "learning_rate": 5.0365413576788264e-05, "loss": 0.0361, "step": 52280 }, { "grad_norm": 0.3417055606842041, "learning_rate": 5.0348879300504194e-05, "loss": 0.0516, "step": 52290 }, { "grad_norm": 0.5173901319503784, "learning_rate": 5.033234498606721e-05, "loss": 0.0286, "step": 52300 }, { "grad_norm": 0.4583854377269745, "learning_rate": 5.0315810635285476e-05, "loss": 0.0384, "step": 52310 }, { "grad_norm": 0.4541206359863281, "learning_rate": 5.0299276249967164e-05, "loss": 0.0307, "step": 52320 }, { "grad_norm": 0.46654146909713745, "learning_rate": 5.0282741831920454e-05, "loss": 0.0298, "step": 52330 }, { "grad_norm": 0.6909177303314209, "learning_rate": 5.026620738295351e-05, "loss": 0.0372, "step": 52340 }, { "grad_norm": 0.3003823757171631, "learning_rate": 5.024967290487455e-05, "loss": 0.0289, "step": 52350 }, { "grad_norm": 0.5751645565032959, "learning_rate": 5.023313839949172e-05, "loss": 0.0365, "step": 52360 }, { "grad_norm": 0.4205668270587921, "learning_rate": 5.0216603868613224e-05, "loss": 0.035, "step": 52370 }, { "grad_norm": 0.2518739402294159, "learning_rate": 5.020006931404726e-05, "loss": 0.0323, "step": 52380 }, { "grad_norm": 0.548818051815033, "learning_rate": 5.0183534737602014e-05, "loss": 0.0328, "step": 52390 }, { "grad_norm": 0.3456385135650635, "learning_rate": 5.0167000141085686e-05, "loss": 0.0369, "step": 52400 }, { "grad_norm": 1.0282325744628906, "learning_rate": 5.015046552630647e-05, "loss": 0.0282, "step": 52410 }, { "grad_norm": 0.6089572310447693, "learning_rate": 5.013393089507258e-05, "loss": 0.0382, "step": 52420 }, { "grad_norm": 0.5389156937599182, "learning_rate": 5.011739624919219e-05, "loss": 0.0373, "step": 52430 }, { "grad_norm": 0.36703792214393616, "learning_rate": 5.0100861590473545e-05, "loss": 0.0398, "step": 52440 }, { "grad_norm": 0.5634084343910217, "learning_rate": 5.008432692072482e-05, "loss": 0.0478, "step": 52450 }, { "grad_norm": 0.6116580367088318, "learning_rate": 5.006779224175423e-05, "loss": 0.0286, "step": 52460 }, { "grad_norm": 0.38093024492263794, "learning_rate": 5.005125755536998e-05, "loss": 0.0406, "step": 52470 }, { "grad_norm": 0.6948017477989197, "learning_rate": 5.003472286338029e-05, "loss": 0.0392, "step": 52480 }, { "grad_norm": 0.23440459370613098, "learning_rate": 5.001818816759335e-05, "loss": 0.0219, "step": 52490 }, { "grad_norm": 0.6789817214012146, "learning_rate": 5.000165346981738e-05, "loss": 0.0567, "step": 52500 }, { "grad_norm": 0.5701590776443481, "learning_rate": 4.9985118771860595e-05, "loss": 0.0435, "step": 52510 }, { "grad_norm": 0.5523441433906555, "learning_rate": 4.996858407553119e-05, "loss": 0.0609, "step": 52520 }, { "grad_norm": 0.750613808631897, "learning_rate": 4.99520493826374e-05, "loss": 0.0459, "step": 52530 }, { "grad_norm": 0.4102987051010132, "learning_rate": 4.9935514694987384e-05, "loss": 0.028, "step": 52540 }, { "grad_norm": 0.6378567814826965, "learning_rate": 4.991898001438941e-05, "loss": 0.0636, "step": 52550 }, { "grad_norm": 0.9610705375671387, "learning_rate": 4.990244534265166e-05, "loss": 0.0389, "step": 52560 }, { "grad_norm": 0.708709180355072, "learning_rate": 4.988591068158234e-05, "loss": 0.0541, "step": 52570 }, { "grad_norm": 0.4169381260871887, "learning_rate": 4.986937603298965e-05, "loss": 0.0575, "step": 52580 }, { "grad_norm": 0.6493057608604431, "learning_rate": 4.985284139868181e-05, "loss": 0.0491, "step": 52590 }, { "grad_norm": 0.6217213273048401, "learning_rate": 4.9836306780467e-05, "loss": 0.0454, "step": 52600 }, { "grad_norm": 0.47910764813423157, "learning_rate": 4.981977218015343e-05, "loss": 0.0373, "step": 52610 }, { "grad_norm": 0.4928431808948517, "learning_rate": 4.980323759954932e-05, "loss": 0.0523, "step": 52620 }, { "grad_norm": 0.5508522987365723, "learning_rate": 4.978670304046284e-05, "loss": 0.0375, "step": 52630 }, { "grad_norm": 0.6471503973007202, "learning_rate": 4.9770168504702195e-05, "loss": 0.0401, "step": 52640 }, { "grad_norm": 0.7647477388381958, "learning_rate": 4.975363399407559e-05, "loss": 0.045, "step": 52650 }, { "grad_norm": 0.7330384254455566, "learning_rate": 4.9737099510391194e-05, "loss": 0.0366, "step": 52660 }, { "grad_norm": 0.5263122916221619, "learning_rate": 4.972056505545719e-05, "loss": 0.0382, "step": 52670 }, { "grad_norm": 0.31246739625930786, "learning_rate": 4.9704030631081775e-05, "loss": 0.0347, "step": 52680 }, { "grad_norm": 0.5593954920768738, "learning_rate": 4.968749623907313e-05, "loss": 0.0336, "step": 52690 }, { "grad_norm": 0.3829071819782257, "learning_rate": 4.967096188123943e-05, "loss": 0.0288, "step": 52700 }, { "grad_norm": 0.31809183955192566, "learning_rate": 4.965442755938884e-05, "loss": 0.0317, "step": 52710 }, { "grad_norm": 0.4480666518211365, "learning_rate": 4.963789327532954e-05, "loss": 0.0392, "step": 52720 }, { "grad_norm": 0.3409181237220764, "learning_rate": 4.962135903086968e-05, "loss": 0.0252, "step": 52730 }, { "grad_norm": 0.4162476360797882, "learning_rate": 4.960482482781741e-05, "loss": 0.0331, "step": 52740 }, { "grad_norm": 0.6500039100646973, "learning_rate": 4.958829066798092e-05, "loss": 0.0396, "step": 52750 }, { "grad_norm": 0.4894169867038727, "learning_rate": 4.957175655316835e-05, "loss": 0.0311, "step": 52760 }, { "grad_norm": 0.7323266267776489, "learning_rate": 4.9555222485187834e-05, "loss": 0.0459, "step": 52770 }, { "grad_norm": 0.8457822203636169, "learning_rate": 4.953868846584752e-05, "loss": 0.0366, "step": 52780 }, { "grad_norm": 0.5370662212371826, "learning_rate": 4.9522154496955544e-05, "loss": 0.0355, "step": 52790 }, { "grad_norm": 0.42694658041000366, "learning_rate": 4.950562058032001e-05, "loss": 0.0277, "step": 52800 }, { "grad_norm": 0.578191339969635, "learning_rate": 4.948908671774909e-05, "loss": 0.0334, "step": 52810 }, { "grad_norm": 0.5330528616905212, "learning_rate": 4.947255291105087e-05, "loss": 0.0304, "step": 52820 }, { "grad_norm": 0.23596727848052979, "learning_rate": 4.945601916203348e-05, "loss": 0.0196, "step": 52830 }, { "grad_norm": 0.5344364047050476, "learning_rate": 4.9439485472505016e-05, "loss": 0.0344, "step": 52840 }, { "grad_norm": 0.36589881777763367, "learning_rate": 4.942295184427356e-05, "loss": 0.0335, "step": 52850 }, { "grad_norm": 0.5899412035942078, "learning_rate": 4.9406418279147234e-05, "loss": 0.0346, "step": 52860 }, { "grad_norm": 0.38988929986953735, "learning_rate": 4.9389884778934116e-05, "loss": 0.0352, "step": 52870 }, { "grad_norm": 0.6342248916625977, "learning_rate": 4.937335134544228e-05, "loss": 0.0287, "step": 52880 }, { "grad_norm": 0.5960299968719482, "learning_rate": 4.935681798047979e-05, "loss": 0.0493, "step": 52890 }, { "grad_norm": 0.48870953917503357, "learning_rate": 4.934028468585473e-05, "loss": 0.0237, "step": 52900 }, { "grad_norm": 0.36614301800727844, "learning_rate": 4.932375146337514e-05, "loss": 0.0298, "step": 52910 }, { "grad_norm": 0.5707033276557922, "learning_rate": 4.930721831484908e-05, "loss": 0.0487, "step": 52920 }, { "grad_norm": 0.5188475847244263, "learning_rate": 4.929068524208456e-05, "loss": 0.04, "step": 52930 }, { "grad_norm": 0.6064600348472595, "learning_rate": 4.927415224688965e-05, "loss": 0.0415, "step": 52940 }, { "grad_norm": 0.33627623319625854, "learning_rate": 4.9257619331072375e-05, "loss": 0.0265, "step": 52950 }, { "grad_norm": 0.4151187837123871, "learning_rate": 4.924108649644072e-05, "loss": 0.0349, "step": 52960 }, { "grad_norm": 0.5142866969108582, "learning_rate": 4.922455374480271e-05, "loss": 0.0386, "step": 52970 }, { "grad_norm": 0.36998888850212097, "learning_rate": 4.920802107796634e-05, "loss": 0.0334, "step": 52980 }, { "grad_norm": 0.4162953197956085, "learning_rate": 4.9191488497739564e-05, "loss": 0.0414, "step": 52990 }, { "grad_norm": 0.515900731086731, "learning_rate": 4.917495600593042e-05, "loss": 0.0387, "step": 53000 }, { "grad_norm": 0.4257829189300537, "learning_rate": 4.915842360434684e-05, "loss": 0.0613, "step": 53010 }, { "grad_norm": 0.48351800441741943, "learning_rate": 4.914189129479679e-05, "loss": 0.0313, "step": 53020 }, { "grad_norm": 0.2536469101905823, "learning_rate": 4.912535907908822e-05, "loss": 0.0288, "step": 53030 }, { "grad_norm": 0.5598856806755066, "learning_rate": 4.910882695902907e-05, "loss": 0.0471, "step": 53040 }, { "grad_norm": 0.28093788027763367, "learning_rate": 4.9092294936427234e-05, "loss": 0.0375, "step": 53050 }, { "grad_norm": 0.424979031085968, "learning_rate": 4.9075763013090685e-05, "loss": 0.0436, "step": 53060 }, { "grad_norm": 0.3794650137424469, "learning_rate": 4.9059231190827305e-05, "loss": 0.0401, "step": 53070 }, { "grad_norm": 0.5331365466117859, "learning_rate": 4.904269947144499e-05, "loss": 0.0377, "step": 53080 }, { "grad_norm": 0.5123594403266907, "learning_rate": 4.902616785675161e-05, "loss": 0.0362, "step": 53090 }, { "grad_norm": 0.5349284410476685, "learning_rate": 4.900963634855505e-05, "loss": 0.0361, "step": 53100 }, { "grad_norm": 0.7021744847297668, "learning_rate": 4.8993104948663173e-05, "loss": 0.0409, "step": 53110 }, { "grad_norm": 0.9947371482849121, "learning_rate": 4.8976573658883823e-05, "loss": 0.0433, "step": 53120 }, { "grad_norm": 0.4363841116428375, "learning_rate": 4.896004248102484e-05, "loss": 0.0348, "step": 53130 }, { "grad_norm": 0.5058912038803101, "learning_rate": 4.8943511416894045e-05, "loss": 0.0279, "step": 53140 }, { "grad_norm": 0.3614865243434906, "learning_rate": 4.892698046829926e-05, "loss": 0.0342, "step": 53150 }, { "grad_norm": 0.4731334447860718, "learning_rate": 4.8910449637048275e-05, "loss": 0.0321, "step": 53160 }, { "grad_norm": 0.6598422527313232, "learning_rate": 4.8893918924948884e-05, "loss": 0.0359, "step": 53170 }, { "grad_norm": 0.5346551537513733, "learning_rate": 4.8877388333808835e-05, "loss": 0.0306, "step": 53180 }, { "grad_norm": 0.38741907477378845, "learning_rate": 4.8860857865435925e-05, "loss": 0.0335, "step": 53190 }, { "grad_norm": 0.7090893983840942, "learning_rate": 4.884432752163789e-05, "loss": 0.0327, "step": 53200 }, { "grad_norm": 0.3594973683357239, "learning_rate": 4.8827797304222466e-05, "loss": 0.0282, "step": 53210 }, { "grad_norm": 0.3470994830131531, "learning_rate": 4.8811267214997366e-05, "loss": 0.0302, "step": 53220 }, { "grad_norm": 0.35534101724624634, "learning_rate": 4.879473725577029e-05, "loss": 0.031, "step": 53230 }, { "grad_norm": 0.2966793179512024, "learning_rate": 4.877820742834891e-05, "loss": 0.0453, "step": 53240 }, { "grad_norm": 0.34738367795944214, "learning_rate": 4.8761677734540956e-05, "loss": 0.0455, "step": 53250 }, { "grad_norm": 0.5189946293830872, "learning_rate": 4.874514817615407e-05, "loss": 0.0257, "step": 53260 }, { "grad_norm": 0.486799418926239, "learning_rate": 4.872861875499588e-05, "loss": 0.0269, "step": 53270 }, { "grad_norm": 0.37393149733543396, "learning_rate": 4.871208947287404e-05, "loss": 0.0257, "step": 53280 }, { "grad_norm": 0.5074824690818787, "learning_rate": 4.869556033159615e-05, "loss": 0.0313, "step": 53290 }, { "grad_norm": 0.3320681154727936, "learning_rate": 4.8679031332969835e-05, "loss": 0.022, "step": 53300 }, { "grad_norm": 0.31314200162887573, "learning_rate": 4.8662502478802635e-05, "loss": 0.031, "step": 53310 }, { "grad_norm": 0.42357736825942993, "learning_rate": 4.8645973770902176e-05, "loss": 0.0358, "step": 53320 }, { "grad_norm": 0.24929159879684448, "learning_rate": 4.862944521107599e-05, "loss": 0.0248, "step": 53330 }, { "grad_norm": 0.7550547122955322, "learning_rate": 4.861291680113162e-05, "loss": 0.0242, "step": 53340 }, { "grad_norm": 0.4394071698188782, "learning_rate": 4.8596388542876563e-05, "loss": 0.035, "step": 53350 }, { "grad_norm": 0.4989111125469208, "learning_rate": 4.857986043811836e-05, "loss": 0.0334, "step": 53360 }, { "grad_norm": 0.823056697845459, "learning_rate": 4.856333248866448e-05, "loss": 0.0365, "step": 53370 }, { "grad_norm": 0.6681874394416809, "learning_rate": 4.85468046963224e-05, "loss": 0.0408, "step": 53380 }, { "grad_norm": 0.40012624859809875, "learning_rate": 4.853027706289957e-05, "loss": 0.0445, "step": 53390 }, { "grad_norm": 0.8697823286056519, "learning_rate": 4.851374959020344e-05, "loss": 0.0458, "step": 53400 }, { "grad_norm": 0.38515999913215637, "learning_rate": 4.849722228004141e-05, "loss": 0.0369, "step": 53410 }, { "grad_norm": 0.40134039521217346, "learning_rate": 4.848069513422091e-05, "loss": 0.036, "step": 53420 }, { "grad_norm": 0.269569456577301, "learning_rate": 4.8464168154549294e-05, "loss": 0.0367, "step": 53430 }, { "grad_norm": 0.5720899701118469, "learning_rate": 4.844764134283392e-05, "loss": 0.0395, "step": 53440 }, { "grad_norm": 0.4950227439403534, "learning_rate": 4.843111470088218e-05, "loss": 0.0333, "step": 53450 }, { "grad_norm": 0.6069079637527466, "learning_rate": 4.841458823050137e-05, "loss": 0.0321, "step": 53460 }, { "grad_norm": 0.6074604988098145, "learning_rate": 4.8398061933498816e-05, "loss": 0.0381, "step": 53470 }, { "grad_norm": 0.3681807816028595, "learning_rate": 4.83815358116818e-05, "loss": 0.0337, "step": 53480 }, { "grad_norm": 0.6748741269111633, "learning_rate": 4.8365009866857584e-05, "loss": 0.0273, "step": 53490 }, { "grad_norm": 0.5314276218414307, "learning_rate": 4.834848410083342e-05, "loss": 0.0453, "step": 53500 }, { "grad_norm": 0.8993976712226868, "learning_rate": 4.833195851541657e-05, "loss": 0.0324, "step": 53510 }, { "grad_norm": 0.34447506070137024, "learning_rate": 4.831543311241423e-05, "loss": 0.032, "step": 53520 }, { "grad_norm": 0.42328324913978577, "learning_rate": 4.829890789363359e-05, "loss": 0.0337, "step": 53530 }, { "grad_norm": 0.5429836511611938, "learning_rate": 4.828238286088182e-05, "loss": 0.028, "step": 53540 }, { "grad_norm": 0.5081155300140381, "learning_rate": 4.826585801596609e-05, "loss": 0.0407, "step": 53550 }, { "grad_norm": 0.8398640751838684, "learning_rate": 4.8249333360693486e-05, "loss": 0.0343, "step": 53560 }, { "grad_norm": 0.7291114330291748, "learning_rate": 4.8232808896871185e-05, "loss": 0.0471, "step": 53570 }, { "grad_norm": 0.28935953974723816, "learning_rate": 4.821628462630624e-05, "loss": 0.0261, "step": 53580 }, { "grad_norm": 0.30539631843566895, "learning_rate": 4.8199760550805735e-05, "loss": 0.0349, "step": 53590 }, { "grad_norm": 1.1456245183944702, "learning_rate": 4.818323667217669e-05, "loss": 0.0339, "step": 53600 }, { "grad_norm": 0.4100203514099121, "learning_rate": 4.8166712992226165e-05, "loss": 0.0285, "step": 53610 }, { "grad_norm": 0.3726944327354431, "learning_rate": 4.8150189512761154e-05, "loss": 0.0311, "step": 53620 }, { "grad_norm": 0.44358810782432556, "learning_rate": 4.813366623558862e-05, "loss": 0.0328, "step": 53630 }, { "grad_norm": 0.48637402057647705, "learning_rate": 4.811714316251554e-05, "loss": 0.0341, "step": 53640 }, { "grad_norm": 0.6143826842308044, "learning_rate": 4.810062029534886e-05, "loss": 0.0251, "step": 53650 }, { "grad_norm": 0.7315083146095276, "learning_rate": 4.8084097635895505e-05, "loss": 0.04, "step": 53660 }, { "grad_norm": 0.6388356685638428, "learning_rate": 4.8067575185962335e-05, "loss": 0.0302, "step": 53670 }, { "grad_norm": 0.5559186935424805, "learning_rate": 4.805105294735625e-05, "loss": 0.044, "step": 53680 }, { "grad_norm": 0.2683783173561096, "learning_rate": 4.803453092188406e-05, "loss": 0.0323, "step": 53690 }, { "grad_norm": 0.6945298910140991, "learning_rate": 4.801800911135263e-05, "loss": 0.041, "step": 53700 }, { "grad_norm": 0.2447488158941269, "learning_rate": 4.8001487517568744e-05, "loss": 0.0388, "step": 53710 }, { "grad_norm": 0.28395578265190125, "learning_rate": 4.798496614233918e-05, "loss": 0.0341, "step": 53720 }, { "grad_norm": 0.2630408704280853, "learning_rate": 4.796844498747069e-05, "loss": 0.0319, "step": 53730 }, { "grad_norm": 0.4003467261791229, "learning_rate": 4.795192405477e-05, "loss": 0.0287, "step": 53740 }, { "grad_norm": 0.2595003843307495, "learning_rate": 4.79354033460438e-05, "loss": 0.0279, "step": 53750 }, { "grad_norm": 0.4924631714820862, "learning_rate": 4.79188828630988e-05, "loss": 0.0439, "step": 53760 }, { "grad_norm": 0.3268970549106598, "learning_rate": 4.790236260774165e-05, "loss": 0.0337, "step": 53770 }, { "grad_norm": 0.5157552361488342, "learning_rate": 4.788584258177896e-05, "loss": 0.0318, "step": 53780 }, { "grad_norm": 0.3939521312713623, "learning_rate": 4.7869322787017356e-05, "loss": 0.0297, "step": 53790 }, { "grad_norm": 0.8380522131919861, "learning_rate": 4.78528032252634e-05, "loss": 0.0367, "step": 53800 }, { "grad_norm": 0.5445830225944519, "learning_rate": 4.783628389832366e-05, "loss": 0.0454, "step": 53810 }, { "grad_norm": 0.3003499209880829, "learning_rate": 4.7819764808004644e-05, "loss": 0.0375, "step": 53820 }, { "grad_norm": 0.8295403122901917, "learning_rate": 4.780324595611289e-05, "loss": 0.0325, "step": 53830 }, { "grad_norm": 0.507655680179596, "learning_rate": 4.7786727344454854e-05, "loss": 0.0351, "step": 53840 }, { "grad_norm": 0.3858433961868286, "learning_rate": 4.777020897483697e-05, "loss": 0.0375, "step": 53850 }, { "grad_norm": 0.6400659680366516, "learning_rate": 4.775369084906569e-05, "loss": 0.0301, "step": 53860 }, { "grad_norm": 0.6988999843597412, "learning_rate": 4.773717296894742e-05, "loss": 0.0481, "step": 53870 }, { "grad_norm": 0.8680821657180786, "learning_rate": 4.772065533628848e-05, "loss": 0.0547, "step": 53880 }, { "grad_norm": 0.4931732714176178, "learning_rate": 4.7704137952895265e-05, "loss": 0.042, "step": 53890 }, { "grad_norm": 0.6259525418281555, "learning_rate": 4.768762082057407e-05, "loss": 0.0306, "step": 53900 }, { "grad_norm": 0.7200382351875305, "learning_rate": 4.7671103941131186e-05, "loss": 0.0345, "step": 53910 }, { "grad_norm": 0.4300471842288971, "learning_rate": 4.765458731637287e-05, "loss": 0.0345, "step": 53920 }, { "grad_norm": 0.4665313959121704, "learning_rate": 4.763807094810537e-05, "loss": 0.0481, "step": 53930 }, { "grad_norm": 0.3664546608924866, "learning_rate": 4.762155483813487e-05, "loss": 0.0396, "step": 53940 }, { "grad_norm": 0.8022248148918152, "learning_rate": 4.760503898826754e-05, "loss": 0.0336, "step": 53950 }, { "grad_norm": 0.5024925470352173, "learning_rate": 4.758852340030957e-05, "loss": 0.0232, "step": 53960 }, { "grad_norm": 0.7623918652534485, "learning_rate": 4.757200807606705e-05, "loss": 0.0411, "step": 53970 }, { "grad_norm": 0.6366125345230103, "learning_rate": 4.755549301734607e-05, "loss": 0.0415, "step": 53980 }, { "grad_norm": 0.7300599217414856, "learning_rate": 4.753897822595271e-05, "loss": 0.0287, "step": 53990 }, { "grad_norm": 0.5950002670288086, "learning_rate": 4.7522463703692994e-05, "loss": 0.0312, "step": 54000 }, { "grad_norm": 0.37420547008514404, "learning_rate": 4.75059494523729e-05, "loss": 0.0312, "step": 54010 }, { "grad_norm": 0.7052971720695496, "learning_rate": 4.748943547379843e-05, "loss": 0.024, "step": 54020 }, { "grad_norm": 1.0995805263519287, "learning_rate": 4.7472921769775535e-05, "loss": 0.0297, "step": 54030 }, { "grad_norm": 0.8189511299133301, "learning_rate": 4.745640834211011e-05, "loss": 0.0352, "step": 54040 }, { "grad_norm": 0.46757158637046814, "learning_rate": 4.7439895192608054e-05, "loss": 0.0238, "step": 54050 }, { "grad_norm": 0.5604574680328369, "learning_rate": 4.7423382323075186e-05, "loss": 0.0276, "step": 54060 }, { "grad_norm": 0.5444274544715881, "learning_rate": 4.740686973531737e-05, "loss": 0.0357, "step": 54070 }, { "grad_norm": 0.27037313580513, "learning_rate": 4.7390357431140375e-05, "loss": 0.0463, "step": 54080 }, { "grad_norm": 0.27910369634628296, "learning_rate": 4.737384541234998e-05, "loss": 0.0266, "step": 54090 }, { "grad_norm": 0.4671325385570526, "learning_rate": 4.735733368075189e-05, "loss": 0.0418, "step": 54100 }, { "grad_norm": 0.5469764471054077, "learning_rate": 4.734082223815182e-05, "loss": 0.0396, "step": 54110 }, { "grad_norm": 0.3428659737110138, "learning_rate": 4.7324311086355436e-05, "loss": 0.0283, "step": 54120 }, { "grad_norm": 0.4858717918395996, "learning_rate": 4.730780022716836e-05, "loss": 0.0326, "step": 54130 }, { "grad_norm": 0.6368493437767029, "learning_rate": 4.729128966239619e-05, "loss": 0.0319, "step": 54140 }, { "grad_norm": 0.4213298261165619, "learning_rate": 4.7274779393844536e-05, "loss": 0.0259, "step": 54150 }, { "grad_norm": 0.4003543555736542, "learning_rate": 4.725826942331891e-05, "loss": 0.0297, "step": 54160 }, { "grad_norm": 0.6263589859008789, "learning_rate": 4.724175975262481e-05, "loss": 0.0339, "step": 54170 }, { "grad_norm": 0.2612852156162262, "learning_rate": 4.722525038356772e-05, "loss": 0.0272, "step": 54180 }, { "grad_norm": 0.6340018510818481, "learning_rate": 4.7208741317953075e-05, "loss": 0.0349, "step": 54190 }, { "grad_norm": 0.2866368591785431, "learning_rate": 4.7192232557586264e-05, "loss": 0.0198, "step": 54200 }, { "grad_norm": 0.44172197580337524, "learning_rate": 4.717572410427269e-05, "loss": 0.027, "step": 54210 }, { "grad_norm": 0.5798938274383545, "learning_rate": 4.71592159598177e-05, "loss": 0.0367, "step": 54220 }, { "grad_norm": 0.5543927550315857, "learning_rate": 4.714270812602657e-05, "loss": 0.0369, "step": 54230 }, { "grad_norm": 0.4691001772880554, "learning_rate": 4.712620060470458e-05, "loss": 0.0376, "step": 54240 }, { "grad_norm": 0.3942660689353943, "learning_rate": 4.7109693397656986e-05, "loss": 0.0263, "step": 54250 }, { "grad_norm": 0.47672879695892334, "learning_rate": 4.709318650668894e-05, "loss": 0.0353, "step": 54260 }, { "grad_norm": 0.7800202369689941, "learning_rate": 4.7076679933605676e-05, "loss": 0.0454, "step": 54270 }, { "grad_norm": 0.45202112197875977, "learning_rate": 4.70601736802123e-05, "loss": 0.0325, "step": 54280 }, { "grad_norm": 0.479722261428833, "learning_rate": 4.704366774831391e-05, "loss": 0.0353, "step": 54290 }, { "grad_norm": 0.39515766501426697, "learning_rate": 4.702716213971557e-05, "loss": 0.0274, "step": 54300 }, { "grad_norm": 0.557340681552887, "learning_rate": 4.70106568562223e-05, "loss": 0.0288, "step": 54310 }, { "grad_norm": 1.3980717658996582, "learning_rate": 4.6994151899639105e-05, "loss": 0.0359, "step": 54320 }, { "grad_norm": 0.6148605346679688, "learning_rate": 4.697764727177093e-05, "loss": 0.0382, "step": 54330 }, { "grad_norm": 0.5215862393379211, "learning_rate": 4.696114297442272e-05, "loss": 0.0347, "step": 54340 }, { "grad_norm": 0.5047242045402527, "learning_rate": 4.694463900939933e-05, "loss": 0.0332, "step": 54350 }, { "grad_norm": 0.5593169331550598, "learning_rate": 4.692813537850564e-05, "loss": 0.0494, "step": 54360 }, { "grad_norm": 0.7557565569877625, "learning_rate": 4.6911632083546445e-05, "loss": 0.0359, "step": 54370 }, { "grad_norm": 1.0149463415145874, "learning_rate": 4.6895129126326525e-05, "loss": 0.0367, "step": 54380 }, { "grad_norm": 0.6731070876121521, "learning_rate": 4.687862650865059e-05, "loss": 0.048, "step": 54390 }, { "grad_norm": 0.3670400381088257, "learning_rate": 4.686212423232339e-05, "loss": 0.0462, "step": 54400 }, { "grad_norm": 0.508354902267456, "learning_rate": 4.6845622299149565e-05, "loss": 0.0337, "step": 54410 }, { "grad_norm": 0.4281359016895294, "learning_rate": 4.682912071093374e-05, "loss": 0.0279, "step": 54420 }, { "grad_norm": 0.3340253531932831, "learning_rate": 4.681261946948052e-05, "loss": 0.0242, "step": 54430 }, { "grad_norm": 0.4681253731250763, "learning_rate": 4.679611857659443e-05, "loss": 0.0329, "step": 54440 }, { "grad_norm": 0.6580877900123596, "learning_rate": 4.6779618034079975e-05, "loss": 0.0342, "step": 54450 }, { "grad_norm": 0.47614815831184387, "learning_rate": 4.6763117843741675e-05, "loss": 0.0256, "step": 54460 }, { "grad_norm": 0.5256355404853821, "learning_rate": 4.674661800738394e-05, "loss": 0.0316, "step": 54470 }, { "grad_norm": 0.3681196868419647, "learning_rate": 4.673011852681117e-05, "loss": 0.0308, "step": 54480 }, { "grad_norm": 0.7888350486755371, "learning_rate": 4.671361940382772e-05, "loss": 0.0664, "step": 54490 }, { "grad_norm": 0.7516223192214966, "learning_rate": 4.669712064023791e-05, "loss": 0.0347, "step": 54500 }, { "grad_norm": 0.5160568356513977, "learning_rate": 4.668062223784602e-05, "loss": 0.0266, "step": 54510 }, { "grad_norm": 0.6258777379989624, "learning_rate": 4.6664124198456274e-05, "loss": 0.0283, "step": 54520 }, { "grad_norm": 0.49234679341316223, "learning_rate": 4.6647626523872905e-05, "loss": 0.0283, "step": 54530 }, { "grad_norm": 0.2895647883415222, "learning_rate": 4.6631129215900066e-05, "loss": 0.0279, "step": 54540 }, { "grad_norm": 0.24980343878269196, "learning_rate": 4.661463227634187e-05, "loss": 0.0281, "step": 54550 }, { "grad_norm": 0.6283552646636963, "learning_rate": 4.659813570700238e-05, "loss": 0.0331, "step": 54560 }, { "grad_norm": 0.3104020655155182, "learning_rate": 4.658163950968568e-05, "loss": 0.0281, "step": 54570 }, { "grad_norm": 0.5061855316162109, "learning_rate": 4.6565143686195726e-05, "loss": 0.0249, "step": 54580 }, { "grad_norm": 0.332788348197937, "learning_rate": 4.654864823833649e-05, "loss": 0.0225, "step": 54590 }, { "grad_norm": 0.729218065738678, "learning_rate": 4.65321531679119e-05, "loss": 0.0388, "step": 54600 }, { "grad_norm": 0.39239993691444397, "learning_rate": 4.6515658476725834e-05, "loss": 0.0347, "step": 54610 }, { "grad_norm": 0.6680811643600464, "learning_rate": 4.649916416658212e-05, "loss": 0.0391, "step": 54620 }, { "grad_norm": 0.23801182210445404, "learning_rate": 4.648267023928455e-05, "loss": 0.0382, "step": 54630 }, { "grad_norm": 0.517554759979248, "learning_rate": 4.646617669663687e-05, "loss": 0.0399, "step": 54640 }, { "grad_norm": 0.41340798139572144, "learning_rate": 4.644968354044277e-05, "loss": 0.0273, "step": 54650 }, { "grad_norm": 0.4238346219062805, "learning_rate": 4.643319077250596e-05, "loss": 0.0265, "step": 54660 }, { "grad_norm": 0.40162426233291626, "learning_rate": 4.6416698394630055e-05, "loss": 0.0273, "step": 54670 }, { "grad_norm": 0.441607266664505, "learning_rate": 4.640020640861862e-05, "loss": 0.027, "step": 54680 }, { "grad_norm": 0.49997225403785706, "learning_rate": 4.6383714816275195e-05, "loss": 0.0403, "step": 54690 }, { "grad_norm": 0.3883686363697052, "learning_rate": 4.636722361940329e-05, "loss": 0.0297, "step": 54700 }, { "grad_norm": 0.41443929076194763, "learning_rate": 4.635073281980632e-05, "loss": 0.0361, "step": 54710 }, { "grad_norm": 0.5499587059020996, "learning_rate": 4.633424241928775e-05, "loss": 0.0286, "step": 54720 }, { "grad_norm": 0.5298710465431213, "learning_rate": 4.6317752419650914e-05, "loss": 0.0311, "step": 54730 }, { "grad_norm": 0.5219287872314453, "learning_rate": 4.630126282269914e-05, "loss": 0.0282, "step": 54740 }, { "grad_norm": 0.3879542350769043, "learning_rate": 4.628477363023569e-05, "loss": 0.0313, "step": 54750 }, { "grad_norm": 0.4562376141548157, "learning_rate": 4.626828484406383e-05, "loss": 0.0408, "step": 54760 }, { "grad_norm": 0.5005574226379395, "learning_rate": 4.625179646598669e-05, "loss": 0.0353, "step": 54770 }, { "grad_norm": 0.40344691276550293, "learning_rate": 4.6235308497807487e-05, "loss": 0.0247, "step": 54780 }, { "grad_norm": 0.23936347663402557, "learning_rate": 4.6218820941329276e-05, "loss": 0.0275, "step": 54790 }, { "grad_norm": 0.4263826310634613, "learning_rate": 4.620233379835513e-05, "loss": 0.0248, "step": 54800 }, { "grad_norm": 0.4938463270664215, "learning_rate": 4.618584707068804e-05, "loss": 0.0419, "step": 54810 }, { "grad_norm": 0.27925407886505127, "learning_rate": 4.616936076013099e-05, "loss": 0.0358, "step": 54820 }, { "grad_norm": 0.36849862337112427, "learning_rate": 4.615287486848688e-05, "loss": 0.0267, "step": 54830 }, { "grad_norm": 0.3647521138191223, "learning_rate": 4.6136389397558584e-05, "loss": 0.0275, "step": 54840 }, { "grad_norm": 0.578216016292572, "learning_rate": 4.6119904349148935e-05, "loss": 0.0329, "step": 54850 }, { "grad_norm": 0.44888490438461304, "learning_rate": 4.6103419725060726e-05, "loss": 0.0323, "step": 54860 }, { "grad_norm": 0.5755162239074707, "learning_rate": 4.608693552709667e-05, "loss": 0.0339, "step": 54870 }, { "grad_norm": 1.0132148265838623, "learning_rate": 4.607045175705947e-05, "loss": 0.0392, "step": 54880 }, { "grad_norm": 0.5762080550193787, "learning_rate": 4.605396841675176e-05, "loss": 0.0299, "step": 54890 }, { "grad_norm": 0.3012998700141907, "learning_rate": 4.6037485507976094e-05, "loss": 0.0312, "step": 54900 }, { "grad_norm": 0.33219143748283386, "learning_rate": 4.602100303253509e-05, "loss": 0.0285, "step": 54910 }, { "grad_norm": 0.4350661039352417, "learning_rate": 4.600452099223122e-05, "loss": 0.0324, "step": 54920 }, { "grad_norm": 0.6120577454566956, "learning_rate": 4.598803938886691e-05, "loss": 0.0344, "step": 54930 }, { "grad_norm": 0.5223218202590942, "learning_rate": 4.597155822424459e-05, "loss": 0.0533, "step": 54940 }, { "grad_norm": 0.9300928115844727, "learning_rate": 4.59550775001666e-05, "loss": 0.0382, "step": 54950 }, { "grad_norm": 0.49920934438705444, "learning_rate": 4.5938597218435234e-05, "loss": 0.0322, "step": 54960 }, { "grad_norm": 0.4957734942436218, "learning_rate": 4.592211738085279e-05, "loss": 0.0437, "step": 54970 }, { "grad_norm": 0.5467409491539001, "learning_rate": 4.590563798922146e-05, "loss": 0.0437, "step": 54980 }, { "grad_norm": 0.36589497327804565, "learning_rate": 4.5889159045343404e-05, "loss": 0.0294, "step": 54990 }, { "grad_norm": 0.4115303158760071, "learning_rate": 4.5872680551020734e-05, "loss": 0.0535, "step": 55000 }, { "grad_norm": 0.38561511039733887, "learning_rate": 4.585620250805552e-05, "loss": 0.0367, "step": 55010 }, { "grad_norm": 0.6741690635681152, "learning_rate": 4.583972491824975e-05, "loss": 0.0337, "step": 55020 }, { "grad_norm": 0.36495622992515564, "learning_rate": 4.582324778340541e-05, "loss": 0.032, "step": 55030 }, { "grad_norm": 0.43279916048049927, "learning_rate": 4.5806771105324424e-05, "loss": 0.0525, "step": 55040 }, { "grad_norm": 0.395621120929718, "learning_rate": 4.579029488580865e-05, "loss": 0.0403, "step": 55050 }, { "grad_norm": 0.5373966693878174, "learning_rate": 4.5773819126659876e-05, "loss": 0.0368, "step": 55060 }, { "grad_norm": 0.39746788144111633, "learning_rate": 4.575734382967991e-05, "loss": 0.0428, "step": 55070 }, { "grad_norm": 0.25816506147384644, "learning_rate": 4.574086899667044e-05, "loss": 0.0252, "step": 55080 }, { "grad_norm": 0.8030422329902649, "learning_rate": 4.5724394629433106e-05, "loss": 0.04, "step": 55090 }, { "grad_norm": 0.8241357803344727, "learning_rate": 4.570792072976958e-05, "loss": 0.0417, "step": 55100 }, { "grad_norm": 0.4064077138900757, "learning_rate": 4.569144729948138e-05, "loss": 0.032, "step": 55110 }, { "grad_norm": 0.45671939849853516, "learning_rate": 4.5674974340370033e-05, "loss": 0.0288, "step": 55120 }, { "grad_norm": 0.5151647925376892, "learning_rate": 4.565850185423699e-05, "loss": 0.0366, "step": 55130 }, { "grad_norm": 0.4900144934654236, "learning_rate": 4.564202984288365e-05, "loss": 0.0334, "step": 55140 }, { "grad_norm": 0.6050332188606262, "learning_rate": 4.5625558308111356e-05, "loss": 0.0254, "step": 55150 }, { "grad_norm": 0.43128523230552673, "learning_rate": 4.560908725172145e-05, "loss": 0.0261, "step": 55160 }, { "grad_norm": 0.760159969329834, "learning_rate": 4.559261667551516e-05, "loss": 0.0378, "step": 55170 }, { "grad_norm": 1.0605593919754028, "learning_rate": 4.5576146581293685e-05, "loss": 0.0279, "step": 55180 }, { "grad_norm": 0.6070608496665955, "learning_rate": 4.5559676970858164e-05, "loss": 0.0427, "step": 55190 }, { "grad_norm": 0.5054159164428711, "learning_rate": 4.5543207846009705e-05, "loss": 0.0388, "step": 55200 }, { "grad_norm": 0.6065922975540161, "learning_rate": 4.552673920854933e-05, "loss": 0.0421, "step": 55210 }, { "grad_norm": 0.3997404873371124, "learning_rate": 4.5510271060278006e-05, "loss": 0.042, "step": 55220 }, { "grad_norm": 0.5638282299041748, "learning_rate": 4.5493803402996704e-05, "loss": 0.0347, "step": 55230 }, { "grad_norm": 0.5462096929550171, "learning_rate": 4.5477336238506295e-05, "loss": 0.0309, "step": 55240 }, { "grad_norm": 0.49391889572143555, "learning_rate": 4.546086956860759e-05, "loss": 0.0288, "step": 55250 }, { "grad_norm": 0.3060339689254761, "learning_rate": 4.5444403395101374e-05, "loss": 0.0296, "step": 55260 }, { "grad_norm": 0.5219467282295227, "learning_rate": 4.542793771978834e-05, "loss": 0.0316, "step": 55270 }, { "grad_norm": 0.9106680154800415, "learning_rate": 4.541147254446916e-05, "loss": 0.0314, "step": 55280 }, { "grad_norm": 0.588253378868103, "learning_rate": 4.539500787094445e-05, "loss": 0.0361, "step": 55290 }, { "grad_norm": 0.5462825298309326, "learning_rate": 4.537854370101475e-05, "loss": 0.0462, "step": 55300 }, { "grad_norm": 0.28514131903648376, "learning_rate": 4.5362080036480566e-05, "loss": 0.0251, "step": 55310 }, { "grad_norm": 0.45154905319213867, "learning_rate": 4.534561687914235e-05, "loss": 0.0534, "step": 55320 }, { "grad_norm": 0.7943061590194702, "learning_rate": 4.5329154230800466e-05, "loss": 0.0501, "step": 55330 }, { "grad_norm": 0.39923012256622314, "learning_rate": 4.531269209325525e-05, "loss": 0.0355, "step": 55340 }, { "grad_norm": 0.34545889496803284, "learning_rate": 4.529623046830696e-05, "loss": 0.0328, "step": 55350 }, { "grad_norm": 0.38030487298965454, "learning_rate": 4.5279769357755866e-05, "loss": 0.035, "step": 55360 }, { "grad_norm": 0.5742738842964172, "learning_rate": 4.5263308763402084e-05, "loss": 0.0262, "step": 55370 }, { "grad_norm": 0.44628024101257324, "learning_rate": 4.524684868704574e-05, "loss": 0.0287, "step": 55380 }, { "grad_norm": 0.5828417539596558, "learning_rate": 4.523038913048688e-05, "loss": 0.036, "step": 55390 }, { "grad_norm": 0.2905600666999817, "learning_rate": 4.521393009552549e-05, "loss": 0.0326, "step": 55400 }, { "grad_norm": 0.48816919326782227, "learning_rate": 4.5197471583961484e-05, "loss": 0.0391, "step": 55410 }, { "grad_norm": 0.415549635887146, "learning_rate": 4.51810135975948e-05, "loss": 0.0291, "step": 55420 }, { "grad_norm": 0.21009409427642822, "learning_rate": 4.516455613822521e-05, "loss": 0.0259, "step": 55430 }, { "grad_norm": 0.1853850781917572, "learning_rate": 4.514809920765249e-05, "loss": 0.0307, "step": 55440 }, { "grad_norm": 0.5690603256225586, "learning_rate": 4.5131642807676356e-05, "loss": 0.0338, "step": 55450 }, { "grad_norm": 0.27812835574150085, "learning_rate": 4.511518694009644e-05, "loss": 0.0284, "step": 55460 }, { "grad_norm": 0.35191574692726135, "learning_rate": 4.509873160671231e-05, "loss": 0.0246, "step": 55470 }, { "grad_norm": 0.32207968831062317, "learning_rate": 4.508227680932356e-05, "loss": 0.0322, "step": 55480 }, { "grad_norm": 0.7770389914512634, "learning_rate": 4.5065822549729605e-05, "loss": 0.0343, "step": 55490 }, { "grad_norm": 0.6237019896507263, "learning_rate": 4.504936882972989e-05, "loss": 0.0278, "step": 55500 }, { "grad_norm": 0.3467557430267334, "learning_rate": 4.503291565112374e-05, "loss": 0.0269, "step": 55510 }, { "grad_norm": 0.4226987063884735, "learning_rate": 4.501646301571048e-05, "loss": 0.0308, "step": 55520 }, { "grad_norm": 0.6543102264404297, "learning_rate": 4.500001092528933e-05, "loss": 0.0437, "step": 55530 }, { "grad_norm": 0.5442014336585999, "learning_rate": 4.4983559381659455e-05, "loss": 0.0388, "step": 55540 }, { "grad_norm": 0.32987552881240845, "learning_rate": 4.496710838661999e-05, "loss": 0.0352, "step": 55550 }, { "grad_norm": 0.7011414170265198, "learning_rate": 4.495065794196999e-05, "loss": 0.0375, "step": 55560 }, { "grad_norm": 0.6809697151184082, "learning_rate": 4.4934208049508455e-05, "loss": 0.0404, "step": 55570 }, { "grad_norm": 0.5646066665649414, "learning_rate": 4.4917758711034303e-05, "loss": 0.0439, "step": 55580 }, { "grad_norm": 1.3677581548690796, "learning_rate": 4.490130992834642e-05, "loss": 0.0452, "step": 55590 }, { "grad_norm": 0.8268740773200989, "learning_rate": 4.488486170324359e-05, "loss": 0.0365, "step": 55600 }, { "grad_norm": 0.48411503434181213, "learning_rate": 4.486841403752463e-05, "loss": 0.0335, "step": 55610 }, { "grad_norm": 0.522384762763977, "learning_rate": 4.4851966932988174e-05, "loss": 0.0617, "step": 55620 }, { "grad_norm": 0.4656311571598053, "learning_rate": 4.4835520391432894e-05, "loss": 0.0292, "step": 55630 }, { "grad_norm": 0.3669193685054779, "learning_rate": 4.481907441465733e-05, "loss": 0.0319, "step": 55640 }, { "grad_norm": 0.5751237273216248, "learning_rate": 4.480262900446e-05, "loss": 0.0312, "step": 55650 }, { "grad_norm": 0.3862683176994324, "learning_rate": 4.478618416263933e-05, "loss": 0.028, "step": 55660 }, { "grad_norm": 0.48799192905426025, "learning_rate": 4.476973989099373e-05, "loss": 0.0323, "step": 55670 }, { "grad_norm": 0.4090180993080139, "learning_rate": 4.475329619132153e-05, "loss": 0.0394, "step": 55680 }, { "grad_norm": 0.5426142811775208, "learning_rate": 4.4736853065420956e-05, "loss": 0.0321, "step": 55690 }, { "grad_norm": 0.442511647939682, "learning_rate": 4.472041051509023e-05, "loss": 0.0331, "step": 55700 }, { "grad_norm": 0.27453798055648804, "learning_rate": 4.470396854212746e-05, "loss": 0.0311, "step": 55710 }, { "grad_norm": 0.37314295768737793, "learning_rate": 4.468752714833072e-05, "loss": 0.0296, "step": 55720 }, { "grad_norm": 0.944052517414093, "learning_rate": 4.467108633549803e-05, "loss": 0.0316, "step": 55730 }, { "grad_norm": 0.3831089735031128, "learning_rate": 4.4654646105427335e-05, "loss": 0.0344, "step": 55740 }, { "grad_norm": 0.39694008231163025, "learning_rate": 4.463820645991651e-05, "loss": 0.0387, "step": 55750 }, { "grad_norm": 0.904818058013916, "learning_rate": 4.462176740076334e-05, "loss": 0.0686, "step": 55760 }, { "grad_norm": 0.48572802543640137, "learning_rate": 4.4605328929765624e-05, "loss": 0.0391, "step": 55770 }, { "grad_norm": 0.3647749423980713, "learning_rate": 4.458889104872102e-05, "loss": 0.0299, "step": 55780 }, { "grad_norm": 0.3392406702041626, "learning_rate": 4.457245375942715e-05, "loss": 0.0271, "step": 55790 }, { "grad_norm": 0.5334383249282837, "learning_rate": 4.455601706368157e-05, "loss": 0.0343, "step": 55800 }, { "grad_norm": 0.46799957752227783, "learning_rate": 4.453958096328179e-05, "loss": 0.0304, "step": 55810 }, { "grad_norm": 0.5153801441192627, "learning_rate": 4.452314546002523e-05, "loss": 0.0268, "step": 55820 }, { "grad_norm": 0.3727766275405884, "learning_rate": 4.4506710555709246e-05, "loss": 0.0302, "step": 55830 }, { "grad_norm": 0.24389460682868958, "learning_rate": 4.449027625213114e-05, "loss": 0.0326, "step": 55840 }, { "grad_norm": 0.4576377868652344, "learning_rate": 4.4473842551088135e-05, "loss": 0.0286, "step": 55850 }, { "grad_norm": 1.321211576461792, "learning_rate": 4.445740945437738e-05, "loss": 0.0329, "step": 55860 }, { "grad_norm": 0.5754765868186951, "learning_rate": 4.444097696379602e-05, "loss": 0.0409, "step": 55870 }, { "grad_norm": 0.4693811535835266, "learning_rate": 4.442454508114106e-05, "loss": 0.0314, "step": 55880 }, { "grad_norm": 0.41083914041519165, "learning_rate": 4.4408113808209465e-05, "loss": 0.0299, "step": 55890 }, { "grad_norm": 0.3687366843223572, "learning_rate": 4.439168314679813e-05, "loss": 0.0376, "step": 55900 }, { "grad_norm": 0.38709932565689087, "learning_rate": 4.43752530987039e-05, "loss": 0.0261, "step": 55910 }, { "grad_norm": 0.31497061252593994, "learning_rate": 4.4358823665723515e-05, "loss": 0.0576, "step": 55920 }, { "grad_norm": 0.5539577007293701, "learning_rate": 4.434239484965371e-05, "loss": 0.0247, "step": 55930 }, { "grad_norm": 0.48635733127593994, "learning_rate": 4.4325966652291103e-05, "loss": 0.0349, "step": 55940 }, { "grad_norm": 0.575778603553772, "learning_rate": 4.430953907543225e-05, "loss": 0.0313, "step": 55950 }, { "grad_norm": 0.37851449847221375, "learning_rate": 4.4293112120873645e-05, "loss": 0.0283, "step": 55960 }, { "grad_norm": 0.5351945161819458, "learning_rate": 4.427668579041172e-05, "loss": 0.0361, "step": 55970 }, { "grad_norm": 0.47832798957824707, "learning_rate": 4.4260260085842816e-05, "loss": 0.0335, "step": 55980 }, { "grad_norm": 0.6169758439064026, "learning_rate": 4.4243835008963256e-05, "loss": 0.0456, "step": 55990 }, { "grad_norm": 0.3756570816040039, "learning_rate": 4.422741056156925e-05, "loss": 0.0303, "step": 56000 }, { "grad_norm": 0.43896305561065674, "learning_rate": 4.421098674545693e-05, "loss": 0.0292, "step": 56010 }, { "grad_norm": 0.71889328956604, "learning_rate": 4.419456356242241e-05, "loss": 0.0315, "step": 56020 }, { "grad_norm": 0.6734185814857483, "learning_rate": 4.4178141014261685e-05, "loss": 0.0454, "step": 56030 }, { "grad_norm": 0.5791629552841187, "learning_rate": 4.416171910277071e-05, "loss": 0.0416, "step": 56040 }, { "grad_norm": 0.538993239402771, "learning_rate": 4.414529782974533e-05, "loss": 0.0363, "step": 56050 }, { "grad_norm": 0.4996951222419739, "learning_rate": 4.41288771969814e-05, "loss": 0.0255, "step": 56060 }, { "grad_norm": 0.39014720916748047, "learning_rate": 4.411245720627464e-05, "loss": 0.0313, "step": 56070 }, { "grad_norm": 0.21115945279598236, "learning_rate": 4.40960378594207e-05, "loss": 0.0391, "step": 56080 }, { "grad_norm": 0.4584598243236542, "learning_rate": 4.407961915821519e-05, "loss": 0.0291, "step": 56090 }, { "grad_norm": 0.8940001726150513, "learning_rate": 4.406320110445362e-05, "loss": 0.0317, "step": 56100 }, { "grad_norm": 0.3843633234500885, "learning_rate": 4.404678369993144e-05, "loss": 0.0298, "step": 56110 }, { "grad_norm": 0.5212532877922058, "learning_rate": 4.403036694644406e-05, "loss": 0.0316, "step": 56120 }, { "grad_norm": 0.5444071888923645, "learning_rate": 4.4013950845786764e-05, "loss": 0.0241, "step": 56130 }, { "grad_norm": 0.48557186126708984, "learning_rate": 4.399753539975482e-05, "loss": 0.0424, "step": 56140 }, { "grad_norm": 0.4242027699947357, "learning_rate": 4.398112061014337e-05, "loss": 0.0237, "step": 56150 }, { "grad_norm": 0.3585607707500458, "learning_rate": 4.396470647874753e-05, "loss": 0.0279, "step": 56160 }, { "grad_norm": 0.7659003734588623, "learning_rate": 4.394829300736229e-05, "loss": 0.0287, "step": 56170 }, { "grad_norm": 0.4682641327381134, "learning_rate": 4.393188019778265e-05, "loss": 0.0372, "step": 56180 }, { "grad_norm": 0.2936476171016693, "learning_rate": 4.391546805180347e-05, "loss": 0.0321, "step": 56190 }, { "grad_norm": 0.8751198053359985, "learning_rate": 4.389905657121955e-05, "loss": 0.0312, "step": 56200 }, { "grad_norm": 0.4044307470321655, "learning_rate": 4.3882645757825635e-05, "loss": 0.0414, "step": 56210 }, { "grad_norm": 0.37897735834121704, "learning_rate": 4.386623561341637e-05, "loss": 0.0271, "step": 56220 }, { "grad_norm": 0.5498656630516052, "learning_rate": 4.384982613978637e-05, "loss": 0.0363, "step": 56230 }, { "grad_norm": 0.8314926624298096, "learning_rate": 4.383341733873012e-05, "loss": 0.0305, "step": 56240 }, { "grad_norm": 0.48516032099723816, "learning_rate": 4.381700921204209e-05, "loss": 0.0292, "step": 56250 }, { "grad_norm": 0.3078881800174713, "learning_rate": 4.380060176151663e-05, "loss": 0.0247, "step": 56260 }, { "grad_norm": 0.9413109421730042, "learning_rate": 4.378419498894805e-05, "loss": 0.0371, "step": 56270 }, { "grad_norm": 0.6395618915557861, "learning_rate": 4.376778889613056e-05, "loss": 0.0345, "step": 56280 }, { "grad_norm": 0.6535065770149231, "learning_rate": 4.375138348485831e-05, "loss": 0.0302, "step": 56290 }, { "grad_norm": 0.5071475505828857, "learning_rate": 4.3734978756925346e-05, "loss": 0.0453, "step": 56300 }, { "grad_norm": 0.5258886814117432, "learning_rate": 4.3718574714125706e-05, "loss": 0.0301, "step": 56310 }, { "grad_norm": 1.7460874319076538, "learning_rate": 4.370217135825329e-05, "loss": 0.0517, "step": 56320 }, { "grad_norm": 0.8150551915168762, "learning_rate": 4.368576869110194e-05, "loss": 0.0286, "step": 56330 }, { "grad_norm": 0.3917509913444519, "learning_rate": 4.366936671446544e-05, "loss": 0.0333, "step": 56340 }, { "grad_norm": 0.8710410594940186, "learning_rate": 4.3652965430137474e-05, "loss": 0.0477, "step": 56350 }, { "grad_norm": 0.40890398621559143, "learning_rate": 4.3636564839911646e-05, "loss": 0.0227, "step": 56360 }, { "grad_norm": 0.5804592967033386, "learning_rate": 4.3620164945581545e-05, "loss": 0.0425, "step": 56370 }, { "grad_norm": 0.38474515080451965, "learning_rate": 4.360376574894061e-05, "loss": 0.0434, "step": 56380 }, { "grad_norm": 0.37440118193626404, "learning_rate": 4.358736725178224e-05, "loss": 0.0307, "step": 56390 }, { "grad_norm": 0.6674050688743591, "learning_rate": 4.357096945589974e-05, "loss": 0.0388, "step": 56400 }, { "grad_norm": 0.4279480576515198, "learning_rate": 4.355457236308636e-05, "loss": 0.0323, "step": 56410 }, { "grad_norm": 0.311716765165329, "learning_rate": 4.353817597513526e-05, "loss": 0.0218, "step": 56420 }, { "grad_norm": 0.5729249119758606, "learning_rate": 4.352178029383948e-05, "loss": 0.0375, "step": 56430 }, { "grad_norm": 0.2628120183944702, "learning_rate": 4.3505385320992105e-05, "loss": 0.0315, "step": 56440 }, { "grad_norm": 0.696582555770874, "learning_rate": 4.348899105838602e-05, "loss": 0.0277, "step": 56450 }, { "grad_norm": 0.2251901626586914, "learning_rate": 4.3472597507814087e-05, "loss": 0.0339, "step": 56460 }, { "grad_norm": 0.35947200655937195, "learning_rate": 4.3456204671069066e-05, "loss": 0.0327, "step": 56470 }, { "grad_norm": 0.5341354608535767, "learning_rate": 4.343981254994367e-05, "loss": 0.0253, "step": 56480 }, { "grad_norm": 0.3929655849933624, "learning_rate": 4.34234211462305e-05, "loss": 0.0354, "step": 56490 }, { "grad_norm": 0.41204988956451416, "learning_rate": 4.340703046172213e-05, "loss": 0.0464, "step": 56500 }, { "grad_norm": 0.38763049244880676, "learning_rate": 4.339064049821097e-05, "loss": 0.0342, "step": 56510 }, { "grad_norm": 0.39841485023498535, "learning_rate": 4.3374251257489446e-05, "loss": 0.0432, "step": 56520 }, { "grad_norm": 0.2883642911911011, "learning_rate": 4.335786274134984e-05, "loss": 0.0288, "step": 56530 }, { "grad_norm": 0.3436058759689331, "learning_rate": 4.334147495158438e-05, "loss": 0.0298, "step": 56540 }, { "grad_norm": 0.29898837208747864, "learning_rate": 4.33250878899852e-05, "loss": 0.0306, "step": 56550 }, { "grad_norm": 0.48520293831825256, "learning_rate": 4.330870155834437e-05, "loss": 0.0237, "step": 56560 }, { "grad_norm": 0.43880563974380493, "learning_rate": 4.3292315958453884e-05, "loss": 0.0314, "step": 56570 }, { "grad_norm": 0.36787301301956177, "learning_rate": 4.3275931092105646e-05, "loss": 0.0205, "step": 56580 }, { "grad_norm": 0.48089420795440674, "learning_rate": 4.325954696109147e-05, "loss": 0.0413, "step": 56590 }, { "grad_norm": 0.40912550687789917, "learning_rate": 4.32431635672031e-05, "loss": 0.0263, "step": 56600 }, { "grad_norm": 0.41048505902290344, "learning_rate": 4.3226780912232214e-05, "loss": 0.0344, "step": 56610 }, { "grad_norm": 0.32105353474617004, "learning_rate": 4.321039899797036e-05, "loss": 0.0269, "step": 56620 }, { "grad_norm": 0.6504315733909607, "learning_rate": 4.319401782620908e-05, "loss": 0.031, "step": 56630 }, { "grad_norm": 0.39864203333854675, "learning_rate": 4.317763739873978e-05, "loss": 0.0264, "step": 56640 }, { "grad_norm": 0.3343406021595001, "learning_rate": 4.31612577173538e-05, "loss": 0.0272, "step": 56650 }, { "grad_norm": 0.5672971606254578, "learning_rate": 4.314487878384239e-05, "loss": 0.0337, "step": 56660 }, { "grad_norm": 0.5417550206184387, "learning_rate": 4.3128500599996736e-05, "loss": 0.0336, "step": 56670 }, { "grad_norm": 0.3546617031097412, "learning_rate": 4.31121231676079e-05, "loss": 0.033, "step": 56680 }, { "grad_norm": 0.45190876722335815, "learning_rate": 4.309574648846694e-05, "loss": 0.0368, "step": 56690 }, { "grad_norm": 0.4870181977748871, "learning_rate": 4.3079370564364755e-05, "loss": 0.0241, "step": 56700 }, { "grad_norm": 0.4147794246673584, "learning_rate": 4.306299539709221e-05, "loss": 0.0299, "step": 56710 }, { "grad_norm": 0.4220578968524933, "learning_rate": 4.3046620988440046e-05, "loss": 0.023, "step": 56720 }, { "grad_norm": 0.3881849944591522, "learning_rate": 4.303024734019897e-05, "loss": 0.0299, "step": 56730 }, { "grad_norm": 0.6203158497810364, "learning_rate": 4.3013874454159565e-05, "loss": 0.0317, "step": 56740 }, { "grad_norm": 0.4315352439880371, "learning_rate": 4.299750233211233e-05, "loss": 0.0276, "step": 56750 }, { "grad_norm": 0.7102642059326172, "learning_rate": 4.2981130975847715e-05, "loss": 0.0354, "step": 56760 }, { "grad_norm": 0.4805290699005127, "learning_rate": 4.2964760387156076e-05, "loss": 0.0256, "step": 56770 }, { "grad_norm": 0.3818122148513794, "learning_rate": 4.294839056782767e-05, "loss": 0.0413, "step": 56780 }, { "grad_norm": 0.3869178593158722, "learning_rate": 4.293202151965266e-05, "loss": 0.0232, "step": 56790 }, { "grad_norm": 0.4739108085632324, "learning_rate": 4.291565324442115e-05, "loss": 0.0306, "step": 56800 }, { "grad_norm": 0.3821718990802765, "learning_rate": 4.2899285743923135e-05, "loss": 0.0323, "step": 56810 }, { "grad_norm": 0.6318532228469849, "learning_rate": 4.288291901994857e-05, "loss": 0.0397, "step": 56820 }, { "grad_norm": 0.291171133518219, "learning_rate": 4.286655307428729e-05, "loss": 0.0426, "step": 56830 }, { "grad_norm": 0.4454280734062195, "learning_rate": 4.285018790872903e-05, "loss": 0.0314, "step": 56840 }, { "grad_norm": 0.37313348054885864, "learning_rate": 4.283382352506347e-05, "loss": 0.033, "step": 56850 }, { "grad_norm": 0.6493011116981506, "learning_rate": 4.2817459925080205e-05, "loss": 0.0351, "step": 56860 }, { "grad_norm": 0.5222212076187134, "learning_rate": 4.28010971105687e-05, "loss": 0.0319, "step": 56870 }, { "grad_norm": 0.4478955864906311, "learning_rate": 4.278473508331841e-05, "loss": 0.0398, "step": 56880 }, { "grad_norm": 0.2580925524234772, "learning_rate": 4.276837384511864e-05, "loss": 0.0253, "step": 56890 }, { "grad_norm": 0.3296954035758972, "learning_rate": 4.275201339775864e-05, "loss": 0.0315, "step": 56900 }, { "grad_norm": 0.43424108624458313, "learning_rate": 4.273565374302756e-05, "loss": 0.0352, "step": 56910 }, { "grad_norm": 0.6589232683181763, "learning_rate": 4.2719294882714454e-05, "loss": 0.0349, "step": 56920 }, { "grad_norm": 0.35268092155456543, "learning_rate": 4.270293681860831e-05, "loss": 0.0322, "step": 56930 }, { "grad_norm": 0.6159883737564087, "learning_rate": 4.268657955249802e-05, "loss": 0.0291, "step": 56940 }, { "grad_norm": 0.3927861452102661, "learning_rate": 4.267022308617241e-05, "loss": 0.0383, "step": 56950 }, { "grad_norm": 0.5492995977401733, "learning_rate": 4.265386742142018e-05, "loss": 0.0422, "step": 56960 }, { "grad_norm": 0.33356550335884094, "learning_rate": 4.263751256002996e-05, "loss": 0.0305, "step": 56970 }, { "grad_norm": 0.47760534286499023, "learning_rate": 4.26211585037903e-05, "loss": 0.0372, "step": 56980 }, { "grad_norm": 0.3528470993041992, "learning_rate": 4.2604805254489656e-05, "loss": 0.0369, "step": 56990 }, { "grad_norm": 0.4169338345527649, "learning_rate": 4.258845281391638e-05, "loss": 0.0285, "step": 57000 }, { "grad_norm": 0.4668065309524536, "learning_rate": 4.257210118385876e-05, "loss": 0.0261, "step": 57010 }, { "grad_norm": 0.4751929044723511, "learning_rate": 4.2555750366104994e-05, "loss": 0.0323, "step": 57020 }, { "grad_norm": 0.39722901582717896, "learning_rate": 4.2539400362443184e-05, "loss": 0.0404, "step": 57030 }, { "grad_norm": 0.36168166995048523, "learning_rate": 4.252305117466133e-05, "loss": 0.0333, "step": 57040 }, { "grad_norm": 0.6918168067932129, "learning_rate": 4.2506702804547356e-05, "loss": 0.0395, "step": 57050 }, { "grad_norm": 0.4464077055454254, "learning_rate": 4.249035525388907e-05, "loss": 0.0393, "step": 57060 }, { "grad_norm": 0.37099218368530273, "learning_rate": 4.247400852447427e-05, "loss": 0.0348, "step": 57070 }, { "grad_norm": 0.38528698682785034, "learning_rate": 4.245766261809059e-05, "loss": 0.0315, "step": 57080 }, { "grad_norm": 0.30325859785079956, "learning_rate": 4.244131753652558e-05, "loss": 0.039, "step": 57090 }, { "grad_norm": 0.7927608489990234, "learning_rate": 4.242497328156672e-05, "loss": 0.0367, "step": 57100 }, { "grad_norm": 0.41398751735687256, "learning_rate": 4.240862985500139e-05, "loss": 0.0376, "step": 57110 }, { "grad_norm": 0.456826776266098, "learning_rate": 4.2392287258616894e-05, "loss": 0.0291, "step": 57120 }, { "grad_norm": 0.5393294095993042, "learning_rate": 4.2375945494200396e-05, "loss": 0.0538, "step": 57130 }, { "grad_norm": 0.3855471611022949, "learning_rate": 4.2359604563539054e-05, "loss": 0.0327, "step": 57140 }, { "grad_norm": 0.6019389629364014, "learning_rate": 4.2343264468419885e-05, "loss": 0.0502, "step": 57150 }, { "grad_norm": 0.3230873644351959, "learning_rate": 4.232692521062979e-05, "loss": 0.0301, "step": 57160 }, { "grad_norm": 0.29595789313316345, "learning_rate": 4.231058679195562e-05, "loss": 0.0317, "step": 57170 }, { "grad_norm": 0.3602553904056549, "learning_rate": 4.229424921418411e-05, "loss": 0.0363, "step": 57180 }, { "grad_norm": 0.2704041302204132, "learning_rate": 4.2277912479101916e-05, "loss": 0.043, "step": 57190 }, { "grad_norm": 0.24199019372463226, "learning_rate": 4.226157658849562e-05, "loss": 0.0258, "step": 57200 }, { "grad_norm": 0.42840856313705444, "learning_rate": 4.2245241544151664e-05, "loss": 0.0465, "step": 57210 }, { "grad_norm": 0.3226681649684906, "learning_rate": 4.2228907347856425e-05, "loss": 0.0366, "step": 57220 }, { "grad_norm": 0.46971601247787476, "learning_rate": 4.2212574001396214e-05, "loss": 0.0303, "step": 57230 }, { "grad_norm": 0.510884702205658, "learning_rate": 4.21962415065572e-05, "loss": 0.0352, "step": 57240 }, { "grad_norm": 0.6120411157608032, "learning_rate": 4.217990986512548e-05, "loss": 0.0312, "step": 57250 }, { "grad_norm": 0.5331808924674988, "learning_rate": 4.216357907888704e-05, "loss": 0.0214, "step": 57260 }, { "grad_norm": 0.21099990606307983, "learning_rate": 4.2147249149627824e-05, "loss": 0.0282, "step": 57270 }, { "grad_norm": 1.31477689743042, "learning_rate": 4.213092007913364e-05, "loss": 0.0322, "step": 57280 }, { "grad_norm": 0.5774071216583252, "learning_rate": 4.2114591869190205e-05, "loss": 0.0319, "step": 57290 }, { "grad_norm": 0.5070744156837463, "learning_rate": 4.209826452158315e-05, "loss": 0.032, "step": 57300 }, { "grad_norm": 0.474216490983963, "learning_rate": 4.2081938038098e-05, "loss": 0.0295, "step": 57310 }, { "grad_norm": 0.6393231153488159, "learning_rate": 4.2065612420520185e-05, "loss": 0.0416, "step": 57320 }, { "grad_norm": 0.442217081785202, "learning_rate": 4.204928767063509e-05, "loss": 0.0218, "step": 57330 }, { "grad_norm": 0.4065779447555542, "learning_rate": 4.2032963790227944e-05, "loss": 0.0326, "step": 57340 }, { "grad_norm": 0.4821713864803314, "learning_rate": 4.2016640781083887e-05, "loss": 0.0423, "step": 57350 }, { "grad_norm": 0.48696765303611755, "learning_rate": 4.2000318644988004e-05, "loss": 0.0354, "step": 57360 }, { "grad_norm": 0.3466052711009979, "learning_rate": 4.198399738372524e-05, "loss": 0.0314, "step": 57370 }, { "grad_norm": 0.3082248568534851, "learning_rate": 4.196767699908045e-05, "loss": 0.0282, "step": 57380 }, { "grad_norm": 0.3977924883365631, "learning_rate": 4.195135749283845e-05, "loss": 0.0418, "step": 57390 }, { "grad_norm": 0.5531509518623352, "learning_rate": 4.1935038866783894e-05, "loss": 0.0222, "step": 57400 }, { "grad_norm": 0.5090299248695374, "learning_rate": 4.1918721122701355e-05, "loss": 0.0281, "step": 57410 }, { "grad_norm": 0.4625193476676941, "learning_rate": 4.190240426237533e-05, "loss": 0.039, "step": 57420 }, { "grad_norm": 0.5069612264633179, "learning_rate": 4.188608828759019e-05, "loss": 0.0253, "step": 57430 }, { "grad_norm": 0.33715900778770447, "learning_rate": 4.1869773200130246e-05, "loss": 0.0401, "step": 57440 }, { "grad_norm": 0.8660805225372314, "learning_rate": 4.185345900177967e-05, "loss": 0.0368, "step": 57450 }, { "grad_norm": 0.43515971302986145, "learning_rate": 4.183714569432258e-05, "loss": 0.0644, "step": 57460 }, { "grad_norm": 0.28471478819847107, "learning_rate": 4.182083327954295e-05, "loss": 0.0314, "step": 57470 }, { "grad_norm": 0.378571480512619, "learning_rate": 4.1804521759224714e-05, "loss": 0.0292, "step": 57480 }, { "grad_norm": 0.41691768169403076, "learning_rate": 4.178821113515165e-05, "loss": 0.0358, "step": 57490 }, { "grad_norm": 0.4555296003818512, "learning_rate": 4.1771901409107476e-05, "loss": 0.0326, "step": 57500 }, { "grad_norm": 0.23811522126197815, "learning_rate": 4.175559258287577e-05, "loss": 0.0441, "step": 57510 }, { "grad_norm": 0.49004217982292175, "learning_rate": 4.173928465824008e-05, "loss": 0.0332, "step": 57520 }, { "grad_norm": 0.49003148078918457, "learning_rate": 4.17229776369838e-05, "loss": 0.029, "step": 57530 }, { "grad_norm": 0.7825912833213806, "learning_rate": 4.170667152089025e-05, "loss": 0.0324, "step": 57540 }, { "grad_norm": 0.5083895921707153, "learning_rate": 4.1690366311742626e-05, "loss": 0.038, "step": 57550 }, { "grad_norm": 0.35410943627357483, "learning_rate": 4.1674062011324055e-05, "loss": 0.0337, "step": 57560 }, { "grad_norm": 0.4540114998817444, "learning_rate": 4.165775862141752e-05, "loss": 0.0357, "step": 57570 }, { "grad_norm": 0.3434520661830902, "learning_rate": 4.164145614380598e-05, "loss": 0.0362, "step": 57580 }, { "grad_norm": 0.355459988117218, "learning_rate": 4.162515458027223e-05, "loss": 0.0341, "step": 57590 }, { "grad_norm": 0.22319065034389496, "learning_rate": 4.160885393259899e-05, "loss": 0.0257, "step": 57600 }, { "grad_norm": 0.42942482233047485, "learning_rate": 4.159255420256886e-05, "loss": 0.0277, "step": 57610 }, { "grad_norm": 0.5626619458198547, "learning_rate": 4.157625539196436e-05, "loss": 0.0264, "step": 57620 }, { "grad_norm": 0.43706566095352173, "learning_rate": 4.155995750256791e-05, "loss": 0.0314, "step": 57630 }, { "grad_norm": 0.5415272116661072, "learning_rate": 4.15436605361618e-05, "loss": 0.0296, "step": 57640 }, { "grad_norm": 0.33054542541503906, "learning_rate": 4.152736449452827e-05, "loss": 0.0255, "step": 57650 }, { "grad_norm": 0.20617544651031494, "learning_rate": 4.151106937944943e-05, "loss": 0.0278, "step": 57660 }, { "grad_norm": 0.8129100799560547, "learning_rate": 4.149477519270727e-05, "loss": 0.0255, "step": 57670 }, { "grad_norm": 0.49829939007759094, "learning_rate": 4.147848193608371e-05, "loss": 0.0285, "step": 57680 }, { "grad_norm": 0.5462833642959595, "learning_rate": 4.146218961136056e-05, "loss": 0.0223, "step": 57690 }, { "grad_norm": 0.5737144947052002, "learning_rate": 4.14458982203195e-05, "loss": 0.0439, "step": 57700 }, { "grad_norm": 0.47191283106803894, "learning_rate": 4.1429607764742165e-05, "loss": 0.0371, "step": 57710 }, { "grad_norm": 0.18313759565353394, "learning_rate": 4.141331824641003e-05, "loss": 0.0261, "step": 57720 }, { "grad_norm": 0.6240636110305786, "learning_rate": 4.139702966710452e-05, "loss": 0.0345, "step": 57730 }, { "grad_norm": 0.368286669254303, "learning_rate": 4.13807420286069e-05, "loss": 0.033, "step": 57740 }, { "grad_norm": 0.2493850737810135, "learning_rate": 4.136445533269837e-05, "loss": 0.0319, "step": 57750 }, { "grad_norm": 0.650396466255188, "learning_rate": 4.134816958116003e-05, "loss": 0.0377, "step": 57760 }, { "grad_norm": 0.727554440498352, "learning_rate": 4.133188477577283e-05, "loss": 0.0304, "step": 57770 }, { "grad_norm": 0.46513277292251587, "learning_rate": 4.131560091831771e-05, "loss": 0.0259, "step": 57780 }, { "grad_norm": 0.5801138281822205, "learning_rate": 4.1299318010575405e-05, "loss": 0.0287, "step": 57790 }, { "grad_norm": 0.30680590867996216, "learning_rate": 4.128303605432661e-05, "loss": 0.0251, "step": 57800 }, { "grad_norm": 0.3901071846485138, "learning_rate": 4.126675505135188e-05, "loss": 0.0269, "step": 57810 }, { "grad_norm": 0.4947914183139801, "learning_rate": 4.125047500343169e-05, "loss": 0.0366, "step": 57820 }, { "grad_norm": 0.4627628028392792, "learning_rate": 4.1234195912346376e-05, "loss": 0.0273, "step": 57830 }, { "grad_norm": 0.328255295753479, "learning_rate": 4.1217917779876235e-05, "loss": 0.029, "step": 57840 }, { "grad_norm": 0.6679262518882751, "learning_rate": 4.120164060780141e-05, "loss": 0.0278, "step": 57850 }, { "grad_norm": 0.4114679992198944, "learning_rate": 4.1185364397901935e-05, "loss": 0.028, "step": 57860 }, { "grad_norm": 0.431401789188385, "learning_rate": 4.116908915195776e-05, "loss": 0.0219, "step": 57870 }, { "grad_norm": 0.5655409693717957, "learning_rate": 4.115281487174872e-05, "loss": 0.0322, "step": 57880 }, { "grad_norm": 0.36642950773239136, "learning_rate": 4.113654155905452e-05, "loss": 0.0359, "step": 57890 }, { "grad_norm": 0.15479235351085663, "learning_rate": 4.112026921565483e-05, "loss": 0.0328, "step": 57900 }, { "grad_norm": 0.4292246699333191, "learning_rate": 4.1103997843329147e-05, "loss": 0.0305, "step": 57910 }, { "grad_norm": 0.3690046966075897, "learning_rate": 4.108772744385689e-05, "loss": 0.0238, "step": 57920 }, { "grad_norm": 0.3870871663093567, "learning_rate": 4.1071458019017346e-05, "loss": 0.039, "step": 57930 }, { "grad_norm": 0.4458240866661072, "learning_rate": 4.1055189570589734e-05, "loss": 0.029, "step": 57940 }, { "grad_norm": 0.4427388906478882, "learning_rate": 4.103892210035315e-05, "loss": 0.024, "step": 57950 }, { "grad_norm": 0.34051257371902466, "learning_rate": 4.102265561008656e-05, "loss": 0.0229, "step": 57960 }, { "grad_norm": 0.8451854586601257, "learning_rate": 4.100639010156885e-05, "loss": 0.0209, "step": 57970 }, { "grad_norm": 0.6075581312179565, "learning_rate": 4.099012557657882e-05, "loss": 0.0314, "step": 57980 }, { "grad_norm": 0.5647826790809631, "learning_rate": 4.09738620368951e-05, "loss": 0.0343, "step": 57990 }, { "grad_norm": 0.46151024103164673, "learning_rate": 4.095759948429626e-05, "loss": 0.0235, "step": 58000 }, { "grad_norm": 0.40562722086906433, "learning_rate": 4.094133792056075e-05, "loss": 0.0313, "step": 58010 }, { "grad_norm": 0.45837169885635376, "learning_rate": 4.092507734746688e-05, "loss": 0.026, "step": 58020 }, { "grad_norm": 0.3884168565273285, "learning_rate": 4.090881776679293e-05, "loss": 0.0302, "step": 58030 }, { "grad_norm": 0.320519357919693, "learning_rate": 4.0892559180316994e-05, "loss": 0.0315, "step": 58040 }, { "grad_norm": 0.48065391182899475, "learning_rate": 4.08763015898171e-05, "loss": 0.0313, "step": 58050 }, { "grad_norm": 0.21028278768062592, "learning_rate": 4.0860044997071145e-05, "loss": 0.0281, "step": 58060 }, { "grad_norm": 0.5283288359642029, "learning_rate": 4.084378940385693e-05, "loss": 0.0302, "step": 58070 }, { "grad_norm": 0.4667746126651764, "learning_rate": 4.082753481195211e-05, "loss": 0.0387, "step": 58080 }, { "grad_norm": 0.6350712776184082, "learning_rate": 4.0811281223134324e-05, "loss": 0.0305, "step": 58090 }, { "grad_norm": 0.33050933480262756, "learning_rate": 4.0795028639181e-05, "loss": 0.032, "step": 58100 }, { "grad_norm": 0.4079333543777466, "learning_rate": 4.077877706186951e-05, "loss": 0.0287, "step": 58110 }, { "grad_norm": 0.6986764669418335, "learning_rate": 4.076252649297709e-05, "loss": 0.0407, "step": 58120 }, { "grad_norm": 0.6126006245613098, "learning_rate": 4.07462769342809e-05, "loss": 0.0321, "step": 58130 }, { "grad_norm": 0.8069866895675659, "learning_rate": 4.073002838755793e-05, "loss": 0.0361, "step": 58140 }, { "grad_norm": 0.6384527087211609, "learning_rate": 4.0713780854585115e-05, "loss": 0.0329, "step": 58150 }, { "grad_norm": 0.31703394651412964, "learning_rate": 4.069753433713927e-05, "loss": 0.0313, "step": 58160 }, { "grad_norm": 1.1678261756896973, "learning_rate": 4.068128883699709e-05, "loss": 0.0336, "step": 58170 }, { "grad_norm": 0.2654670476913452, "learning_rate": 4.0665044355935155e-05, "loss": 0.0256, "step": 58180 }, { "grad_norm": 0.49873554706573486, "learning_rate": 4.064880089572993e-05, "loss": 0.0261, "step": 58190 }, { "grad_norm": 0.49718156456947327, "learning_rate": 4.063255845815779e-05, "loss": 0.0286, "step": 58200 }, { "grad_norm": 0.43572190403938293, "learning_rate": 4.0616317044994945e-05, "loss": 0.0296, "step": 58210 }, { "grad_norm": 0.9427915811538696, "learning_rate": 4.0600076658017585e-05, "loss": 0.0268, "step": 58220 }, { "grad_norm": 0.5636535286903381, "learning_rate": 4.0583837299001706e-05, "loss": 0.0248, "step": 58230 }, { "grad_norm": 0.351535439491272, "learning_rate": 4.056759896972322e-05, "loss": 0.0372, "step": 58240 }, { "grad_norm": 0.39424172043800354, "learning_rate": 4.055136167195793e-05, "loss": 0.0329, "step": 58250 }, { "grad_norm": 0.5460142493247986, "learning_rate": 4.053512540748152e-05, "loss": 0.0264, "step": 58260 }, { "grad_norm": 0.34533822536468506, "learning_rate": 4.051889017806955e-05, "loss": 0.0342, "step": 58270 }, { "grad_norm": 0.44164443016052246, "learning_rate": 4.0502655985497516e-05, "loss": 0.0387, "step": 58280 }, { "grad_norm": 0.40792521834373474, "learning_rate": 4.048642283154074e-05, "loss": 0.0322, "step": 58290 }, { "grad_norm": 0.25721409916877747, "learning_rate": 4.0470190717974464e-05, "loss": 0.0321, "step": 58300 }, { "grad_norm": 0.5103589296340942, "learning_rate": 4.045395964657381e-05, "loss": 0.0448, "step": 58310 }, { "grad_norm": 1.283676266670227, "learning_rate": 4.043772961911376e-05, "loss": 0.034, "step": 58320 }, { "grad_norm": 0.522139847278595, "learning_rate": 4.042150063736924e-05, "loss": 0.0349, "step": 58330 }, { "grad_norm": 0.7295383214950562, "learning_rate": 4.0405272703114976e-05, "loss": 0.0486, "step": 58340 }, { "grad_norm": 0.4917377233505249, "learning_rate": 4.038904581812569e-05, "loss": 0.043, "step": 58350 }, { "grad_norm": 0.41746190190315247, "learning_rate": 4.03728199841759e-05, "loss": 0.0326, "step": 58360 }, { "grad_norm": 0.28972089290618896, "learning_rate": 4.0356595203040044e-05, "loss": 0.0178, "step": 58370 }, { "grad_norm": 0.49450352787971497, "learning_rate": 4.0340371476492426e-05, "loss": 0.0348, "step": 58380 }, { "grad_norm": 0.518925130367279, "learning_rate": 4.032414880630727e-05, "loss": 0.0292, "step": 58390 }, { "grad_norm": 0.8258159160614014, "learning_rate": 4.0307927194258646e-05, "loss": 0.0361, "step": 58400 }, { "grad_norm": 0.5234627723693848, "learning_rate": 4.029170664212054e-05, "loss": 0.0272, "step": 58410 }, { "grad_norm": 0.596439778804779, "learning_rate": 4.0275487151666794e-05, "loss": 0.0349, "step": 58420 }, { "grad_norm": 0.308449387550354, "learning_rate": 4.025926872467116e-05, "loss": 0.0365, "step": 58430 }, { "grad_norm": 0.3054339587688446, "learning_rate": 4.024305136290725e-05, "loss": 0.0336, "step": 58440 }, { "grad_norm": 0.24125856161117554, "learning_rate": 4.022683506814858e-05, "loss": 0.0251, "step": 58450 }, { "grad_norm": 0.27696695923805237, "learning_rate": 4.021061984216853e-05, "loss": 0.0361, "step": 58460 }, { "grad_norm": 0.7468804121017456, "learning_rate": 4.019440568674035e-05, "loss": 0.0504, "step": 58470 }, { "grad_norm": 0.44934988021850586, "learning_rate": 4.017819260363724e-05, "loss": 0.0256, "step": 58480 }, { "grad_norm": 0.6673471331596375, "learning_rate": 4.016198059463223e-05, "loss": 0.0256, "step": 58490 }, { "grad_norm": 0.7101934552192688, "learning_rate": 4.0145769661498233e-05, "loss": 0.0277, "step": 58500 }, { "grad_norm": 0.5776350498199463, "learning_rate": 4.012955980600804e-05, "loss": 0.0323, "step": 58510 }, { "grad_norm": 0.41675901412963867, "learning_rate": 4.0113351029934355e-05, "loss": 0.0355, "step": 58520 }, { "grad_norm": 0.385142058134079, "learning_rate": 4.0097143335049714e-05, "loss": 0.026, "step": 58530 }, { "grad_norm": 0.41656118631362915, "learning_rate": 4.0080936723126606e-05, "loss": 0.03, "step": 58540 }, { "grad_norm": 0.6642564535140991, "learning_rate": 4.006473119593735e-05, "loss": 0.0297, "step": 58550 }, { "grad_norm": 0.6011072397232056, "learning_rate": 4.004852675525415e-05, "loss": 0.0328, "step": 58560 }, { "grad_norm": 0.6350217461585999, "learning_rate": 4.00323234028491e-05, "loss": 0.0451, "step": 58570 }, { "grad_norm": 0.5734363794326782, "learning_rate": 4.0016121140494186e-05, "loss": 0.0272, "step": 58580 }, { "grad_norm": 0.40243011713027954, "learning_rate": 3.999991996996122e-05, "loss": 0.0261, "step": 58590 }, { "grad_norm": 0.3327464163303375, "learning_rate": 3.9983719893022e-05, "loss": 0.0285, "step": 58600 }, { "grad_norm": 0.23688510060310364, "learning_rate": 3.996752091144812e-05, "loss": 0.036, "step": 58610 }, { "grad_norm": 0.9981302618980408, "learning_rate": 3.9951323027011055e-05, "loss": 0.0376, "step": 58620 }, { "grad_norm": 0.6912528872489929, "learning_rate": 3.993512624148219e-05, "loss": 0.0406, "step": 58630 }, { "grad_norm": 0.39106449484825134, "learning_rate": 3.99189305566328e-05, "loss": 0.0298, "step": 58640 }, { "grad_norm": 0.2565540671348572, "learning_rate": 3.990273597423401e-05, "loss": 0.0248, "step": 58650 }, { "grad_norm": 0.6874910593032837, "learning_rate": 3.988654249605681e-05, "loss": 0.0337, "step": 58660 }, { "grad_norm": 0.4515158534049988, "learning_rate": 3.987035012387213e-05, "loss": 0.0278, "step": 58670 }, { "grad_norm": 1.1586641073226929, "learning_rate": 3.9854158859450735e-05, "loss": 0.0288, "step": 58680 }, { "grad_norm": 0.3922828137874603, "learning_rate": 3.983796870456328e-05, "loss": 0.0281, "step": 58690 }, { "grad_norm": 0.5975247025489807, "learning_rate": 3.982177966098028e-05, "loss": 0.0336, "step": 58700 }, { "grad_norm": 0.494573712348938, "learning_rate": 3.980559173047216e-05, "loss": 0.0257, "step": 58710 }, { "grad_norm": 0.6380792856216431, "learning_rate": 3.9789404914809176e-05, "loss": 0.0378, "step": 58720 }, { "grad_norm": 0.8324628472328186, "learning_rate": 3.9773219215761544e-05, "loss": 0.0282, "step": 58730 }, { "grad_norm": 0.3003259003162384, "learning_rate": 3.975703463509928e-05, "loss": 0.026, "step": 58740 }, { "grad_norm": 0.30439090728759766, "learning_rate": 3.974085117459231e-05, "loss": 0.0267, "step": 58750 }, { "grad_norm": 0.3284868896007538, "learning_rate": 3.972466883601043e-05, "loss": 0.0274, "step": 58760 }, { "grad_norm": 0.7030460238456726, "learning_rate": 3.9708487621123316e-05, "loss": 0.0274, "step": 58770 }, { "grad_norm": 0.4096240699291229, "learning_rate": 3.969230753170051e-05, "loss": 0.021, "step": 58780 }, { "grad_norm": 0.51985764503479, "learning_rate": 3.967612856951146e-05, "loss": 0.0333, "step": 58790 }, { "grad_norm": 0.3816528916358948, "learning_rate": 3.9659950736325476e-05, "loss": 0.0248, "step": 58800 }, { "grad_norm": 0.3063020706176758, "learning_rate": 3.964377403391174e-05, "loss": 0.0267, "step": 58810 }, { "grad_norm": 0.5989251136779785, "learning_rate": 3.9627598464039303e-05, "loss": 0.0348, "step": 58820 }, { "grad_norm": 0.25791582465171814, "learning_rate": 3.9611424028477096e-05, "loss": 0.0253, "step": 58830 }, { "grad_norm": 0.3404653072357178, "learning_rate": 3.959525072899394e-05, "loss": 0.0262, "step": 58840 }, { "grad_norm": 0.3816676437854767, "learning_rate": 3.9579078567358506e-05, "loss": 0.0221, "step": 58850 }, { "grad_norm": 0.47171705961227417, "learning_rate": 3.956290754533939e-05, "loss": 0.0346, "step": 58860 }, { "grad_norm": 0.2982821464538574, "learning_rate": 3.954673766470501e-05, "loss": 0.0237, "step": 58870 }, { "grad_norm": 0.3381100594997406, "learning_rate": 3.9530568927223674e-05, "loss": 0.0363, "step": 58880 }, { "grad_norm": 0.6925659775733948, "learning_rate": 3.95144013346636e-05, "loss": 0.0441, "step": 58890 }, { "grad_norm": 0.4193187952041626, "learning_rate": 3.949823488879282e-05, "loss": 0.027, "step": 58900 }, { "grad_norm": 0.3695870637893677, "learning_rate": 3.9482069591379276e-05, "loss": 0.0309, "step": 58910 }, { "grad_norm": 0.33614328503608704, "learning_rate": 3.9465905444190784e-05, "loss": 0.0219, "step": 58920 }, { "grad_norm": 0.9236588478088379, "learning_rate": 3.944974244899505e-05, "loss": 0.0282, "step": 58930 }, { "grad_norm": 0.3630269467830658, "learning_rate": 3.9433580607559616e-05, "loss": 0.0276, "step": 58940 }, { "grad_norm": 0.9177111387252808, "learning_rate": 3.941741992165193e-05, "loss": 0.0384, "step": 58950 }, { "grad_norm": 0.4677365720272064, "learning_rate": 3.940126039303928e-05, "loss": 0.0282, "step": 58960 }, { "grad_norm": 0.4992145895957947, "learning_rate": 3.938510202348886e-05, "loss": 0.0291, "step": 58970 }, { "grad_norm": 0.31465649604797363, "learning_rate": 3.93689448147677e-05, "loss": 0.0351, "step": 58980 }, { "grad_norm": 0.7153583765029907, "learning_rate": 3.935278876864278e-05, "loss": 0.0393, "step": 58990 }, { "grad_norm": 0.537705659866333, "learning_rate": 3.933663388688087e-05, "loss": 0.0289, "step": 59000 }, { "grad_norm": 0.8205650448799133, "learning_rate": 3.9320480171248645e-05, "loss": 0.0327, "step": 59010 }, { "grad_norm": 0.41703447699546814, "learning_rate": 3.930432762351265e-05, "loss": 0.0261, "step": 59020 }, { "grad_norm": 0.3819284737110138, "learning_rate": 3.928817624543931e-05, "loss": 0.0241, "step": 59030 }, { "grad_norm": 0.21588094532489777, "learning_rate": 3.9272026038794885e-05, "loss": 0.03, "step": 59040 }, { "grad_norm": 0.8863736391067505, "learning_rate": 3.9255877005345586e-05, "loss": 0.0306, "step": 59050 }, { "grad_norm": 0.47574087977409363, "learning_rate": 3.923972914685743e-05, "loss": 0.0291, "step": 59060 }, { "grad_norm": 0.492206335067749, "learning_rate": 3.922358246509631e-05, "loss": 0.0294, "step": 59070 }, { "grad_norm": 0.31822967529296875, "learning_rate": 3.920743696182801e-05, "loss": 0.05, "step": 59080 }, { "grad_norm": 0.4279882311820984, "learning_rate": 3.9191292638818164e-05, "loss": 0.0275, "step": 59090 }, { "grad_norm": 0.44392603635787964, "learning_rate": 3.9175149497832316e-05, "loss": 0.0308, "step": 59100 }, { "grad_norm": 0.3617464601993561, "learning_rate": 3.915900754063584e-05, "loss": 0.0266, "step": 59110 }, { "grad_norm": 0.4341704547405243, "learning_rate": 3.9142866768994014e-05, "loss": 0.0255, "step": 59120 }, { "grad_norm": 0.6087809801101685, "learning_rate": 3.912672718467193e-05, "loss": 0.0297, "step": 59130 }, { "grad_norm": 0.5582113265991211, "learning_rate": 3.911058878943463e-05, "loss": 0.0263, "step": 59140 }, { "grad_norm": 0.3995397984981537, "learning_rate": 3.9094451585046974e-05, "loss": 0.036, "step": 59150 }, { "grad_norm": 0.27457568049430847, "learning_rate": 3.90783155732737e-05, "loss": 0.0331, "step": 59160 }, { "grad_norm": 0.23705355823040009, "learning_rate": 3.9062180755879373e-05, "loss": 0.0298, "step": 59170 }, { "grad_norm": 0.3049056828022003, "learning_rate": 3.904604713462855e-05, "loss": 0.0334, "step": 59180 }, { "grad_norm": 0.3974856436252594, "learning_rate": 3.902991471128554e-05, "loss": 0.0218, "step": 59190 }, { "grad_norm": 0.4475879967212677, "learning_rate": 3.901378348761456e-05, "loss": 0.0227, "step": 59200 }, { "grad_norm": 1.1844905614852905, "learning_rate": 3.899765346537971e-05, "loss": 0.0327, "step": 59210 }, { "grad_norm": 0.5617716908454895, "learning_rate": 3.898152464634493e-05, "loss": 0.0216, "step": 59220 }, { "grad_norm": 0.20350994169712067, "learning_rate": 3.896539703227402e-05, "loss": 0.0318, "step": 59230 }, { "grad_norm": 1.1491634845733643, "learning_rate": 3.894927062493073e-05, "loss": 0.0286, "step": 59240 }, { "grad_norm": 0.38674572110176086, "learning_rate": 3.893314542607859e-05, "loss": 0.0327, "step": 59250 }, { "grad_norm": 0.6504107117652893, "learning_rate": 3.891702143748103e-05, "loss": 0.0281, "step": 59260 }, { "grad_norm": 0.6221770644187927, "learning_rate": 3.890089866090133e-05, "loss": 0.0312, "step": 59270 }, { "grad_norm": 0.5193603038787842, "learning_rate": 3.888477709810267e-05, "loss": 0.0296, "step": 59280 }, { "grad_norm": 0.6005814671516418, "learning_rate": 3.886865675084806e-05, "loss": 0.0347, "step": 59290 }, { "grad_norm": 0.40415865182876587, "learning_rate": 3.885253762090043e-05, "loss": 0.0307, "step": 59300 }, { "grad_norm": 0.3219727873802185, "learning_rate": 3.8836419710022534e-05, "loss": 0.0313, "step": 59310 }, { "grad_norm": 0.40708810091018677, "learning_rate": 3.882030301997698e-05, "loss": 0.0407, "step": 59320 }, { "grad_norm": 0.5056127905845642, "learning_rate": 3.880418755252629e-05, "loss": 0.0372, "step": 59330 }, { "grad_norm": 0.49524787068367004, "learning_rate": 3.87880733094328e-05, "loss": 0.0285, "step": 59340 }, { "grad_norm": 0.3590370714664459, "learning_rate": 3.877196029245877e-05, "loss": 0.0332, "step": 59350 }, { "grad_norm": 0.4815409183502197, "learning_rate": 3.875584850336627e-05, "loss": 0.028, "step": 59360 }, { "grad_norm": 0.43177467584609985, "learning_rate": 3.873973794391728e-05, "loss": 0.0294, "step": 59370 }, { "grad_norm": 0.6624674797058105, "learning_rate": 3.872362861587361e-05, "loss": 0.0375, "step": 59380 }, { "grad_norm": 0.42388468980789185, "learning_rate": 3.870752052099698e-05, "loss": 0.0255, "step": 59390 }, { "grad_norm": 0.6926313638687134, "learning_rate": 3.8691413661048916e-05, "loss": 0.0354, "step": 59400 }, { "grad_norm": 0.4832577705383301, "learning_rate": 3.867530803779085e-05, "loss": 0.0276, "step": 59410 }, { "grad_norm": 0.5529965162277222, "learning_rate": 3.8659203652984054e-05, "loss": 0.0499, "step": 59420 }, { "grad_norm": 0.5511907339096069, "learning_rate": 3.864310050838972e-05, "loss": 0.0356, "step": 59430 }, { "grad_norm": 0.6833990216255188, "learning_rate": 3.8626998605768835e-05, "loss": 0.0283, "step": 59440 }, { "grad_norm": 0.5411245822906494, "learning_rate": 3.8610897946882294e-05, "loss": 0.0353, "step": 59450 }, { "grad_norm": 0.40370434522628784, "learning_rate": 3.859479853349082e-05, "loss": 0.0521, "step": 59460 }, { "grad_norm": 0.2677776515483856, "learning_rate": 3.857870036735503e-05, "loss": 0.028, "step": 59470 }, { "grad_norm": 0.28392913937568665, "learning_rate": 3.856260345023538e-05, "loss": 0.0315, "step": 59480 }, { "grad_norm": 0.8024232983589172, "learning_rate": 3.854650778389224e-05, "loss": 0.0286, "step": 59490 }, { "grad_norm": 0.3109496235847473, "learning_rate": 3.8530413370085786e-05, "loss": 0.033, "step": 59500 }, { "grad_norm": 0.38555872440338135, "learning_rate": 3.851432021057608e-05, "loss": 0.0287, "step": 59510 }, { "grad_norm": 0.33028703927993774, "learning_rate": 3.849822830712306e-05, "loss": 0.0313, "step": 59520 }, { "grad_norm": 0.3254498243331909, "learning_rate": 3.848213766148649e-05, "loss": 0.0298, "step": 59530 }, { "grad_norm": 0.48735561966896057, "learning_rate": 3.846604827542602e-05, "loss": 0.0222, "step": 59540 }, { "grad_norm": 0.49257808923721313, "learning_rate": 3.844996015070117e-05, "loss": 0.0262, "step": 59550 }, { "grad_norm": 0.3190963864326477, "learning_rate": 3.843387328907132e-05, "loss": 0.0268, "step": 59560 }, { "grad_norm": 0.4920717477798462, "learning_rate": 3.8417787692295696e-05, "loss": 0.0289, "step": 59570 }, { "grad_norm": 0.25316551327705383, "learning_rate": 3.84017033621334e-05, "loss": 0.0196, "step": 59580 }, { "grad_norm": 0.8483221530914307, "learning_rate": 3.838562030034338e-05, "loss": 0.0387, "step": 59590 }, { "grad_norm": 0.4059063196182251, "learning_rate": 3.836953850868447e-05, "loss": 0.0235, "step": 59600 }, { "grad_norm": 0.7021512985229492, "learning_rate": 3.835345798891533e-05, "loss": 0.0324, "step": 59610 }, { "grad_norm": 0.6148982644081116, "learning_rate": 3.833737874279452e-05, "loss": 0.0289, "step": 59620 }, { "grad_norm": 1.1281284093856812, "learning_rate": 3.8321300772080433e-05, "loss": 0.0255, "step": 59630 }, { "grad_norm": 0.26895201206207275, "learning_rate": 3.8305224078531345e-05, "loss": 0.0258, "step": 59640 }, { "grad_norm": 0.7043973207473755, "learning_rate": 3.828914866390537e-05, "loss": 0.0432, "step": 59650 }, { "grad_norm": 0.7394375801086426, "learning_rate": 3.827307452996048e-05, "loss": 0.033, "step": 59660 }, { "grad_norm": 0.4378630518913269, "learning_rate": 3.825700167845454e-05, "loss": 0.0372, "step": 59670 }, { "grad_norm": 0.194795161485672, "learning_rate": 3.8240930111145214e-05, "loss": 0.0258, "step": 59680 }, { "grad_norm": 0.24530740082263947, "learning_rate": 3.8224859829790116e-05, "loss": 0.0399, "step": 59690 }, { "grad_norm": 0.3072943091392517, "learning_rate": 3.820879083614664e-05, "loss": 0.0269, "step": 59700 }, { "grad_norm": 0.6520752310752869, "learning_rate": 3.819272313197208e-05, "loss": 0.0347, "step": 59710 }, { "grad_norm": 0.35430029034614563, "learning_rate": 3.817665671902355e-05, "loss": 0.0303, "step": 59720 }, { "grad_norm": 0.5058355331420898, "learning_rate": 3.816059159905808e-05, "loss": 0.0313, "step": 59730 }, { "grad_norm": 0.3699784576892853, "learning_rate": 3.8144527773832476e-05, "loss": 0.0269, "step": 59740 }, { "grad_norm": 0.5207605957984924, "learning_rate": 3.812846524510352e-05, "loss": 0.035, "step": 59750 }, { "grad_norm": 0.4011480212211609, "learning_rate": 3.811240401462775e-05, "loss": 0.029, "step": 59760 }, { "grad_norm": 0.30578988790512085, "learning_rate": 3.809634408416162e-05, "loss": 0.0229, "step": 59770 }, { "grad_norm": 0.31734204292297363, "learning_rate": 3.8080285455461394e-05, "loss": 0.027, "step": 59780 }, { "grad_norm": 0.36297643184661865, "learning_rate": 3.806422813028323e-05, "loss": 0.0497, "step": 59790 }, { "grad_norm": 0.6800527572631836, "learning_rate": 3.8048172110383114e-05, "loss": 0.0355, "step": 59800 }, { "grad_norm": 0.4876583218574524, "learning_rate": 3.8032117397516944e-05, "loss": 0.0302, "step": 59810 }, { "grad_norm": 0.5442116856575012, "learning_rate": 3.801606399344042e-05, "loss": 0.033, "step": 59820 }, { "grad_norm": 0.5043653249740601, "learning_rate": 3.800001189990913e-05, "loss": 0.0342, "step": 59830 }, { "grad_norm": 0.5727059245109558, "learning_rate": 3.798396111867847e-05, "loss": 0.0271, "step": 59840 }, { "grad_norm": 0.5103516578674316, "learning_rate": 3.796791165150378e-05, "loss": 0.03, "step": 59850 }, { "grad_norm": 0.34005409479141235, "learning_rate": 3.795186350014018e-05, "loss": 0.043, "step": 59860 }, { "grad_norm": 0.5486010313034058, "learning_rate": 3.793581666634266e-05, "loss": 0.0365, "step": 59870 }, { "grad_norm": 0.38496264815330505, "learning_rate": 3.791977115186609e-05, "loss": 0.037, "step": 59880 }, { "grad_norm": 0.4322469234466553, "learning_rate": 3.79037269584652e-05, "loss": 0.0313, "step": 59890 }, { "grad_norm": 0.4201471209526062, "learning_rate": 3.7887684087894545e-05, "loss": 0.0242, "step": 59900 }, { "grad_norm": 0.6335984468460083, "learning_rate": 3.787164254190854e-05, "loss": 0.0376, "step": 59910 }, { "grad_norm": 0.5383214950561523, "learning_rate": 3.7855602322261473e-05, "loss": 0.0305, "step": 59920 }, { "grad_norm": 0.4138793349266052, "learning_rate": 3.783956343070746e-05, "loss": 0.0255, "step": 59930 }, { "grad_norm": 0.34561070799827576, "learning_rate": 3.7823525869000526e-05, "loss": 0.0256, "step": 59940 }, { "grad_norm": 0.4870399534702301, "learning_rate": 3.78074896388945e-05, "loss": 0.0309, "step": 59950 }, { "grad_norm": 0.5938989520072937, "learning_rate": 3.7791454742143075e-05, "loss": 0.0406, "step": 59960 }, { "grad_norm": 0.8906267285346985, "learning_rate": 3.7775421180499795e-05, "loss": 0.032, "step": 59970 }, { "grad_norm": 0.22280602157115936, "learning_rate": 3.7759388955718086e-05, "loss": 0.0293, "step": 59980 }, { "grad_norm": 0.5605113506317139, "learning_rate": 3.774335806955117e-05, "loss": 0.0301, "step": 59990 }, { "grad_norm": 0.34149307012557983, "learning_rate": 3.7727328523752215e-05, "loss": 0.0283, "step": 60000 }, { "grad_norm": 0.6512458324432373, "learning_rate": 3.771130032007415e-05, "loss": 0.0243, "step": 60010 }, { "grad_norm": 0.5144569873809814, "learning_rate": 3.769527346026982e-05, "loss": 0.0317, "step": 60020 }, { "grad_norm": 0.4992750585079193, "learning_rate": 3.767924794609188e-05, "loss": 0.0222, "step": 60030 }, { "grad_norm": 0.8706034421920776, "learning_rate": 3.7663223779292854e-05, "loss": 0.027, "step": 60040 }, { "grad_norm": 0.23365652561187744, "learning_rate": 3.764720096162512e-05, "loss": 0.0315, "step": 60050 }, { "grad_norm": 0.22592894732952118, "learning_rate": 3.763117949484092e-05, "loss": 0.0205, "step": 60060 }, { "grad_norm": 0.35975778102874756, "learning_rate": 3.7615159380692344e-05, "loss": 0.0278, "step": 60070 }, { "grad_norm": 0.4452230632305145, "learning_rate": 3.759914062093132e-05, "loss": 0.0282, "step": 60080 }, { "grad_norm": 0.8747525811195374, "learning_rate": 3.7583123217309615e-05, "loss": 0.0492, "step": 60090 }, { "grad_norm": 0.19513554871082306, "learning_rate": 3.7567107171578904e-05, "loss": 0.0308, "step": 60100 }, { "grad_norm": 0.38565728068351746, "learning_rate": 3.755109248549066e-05, "loss": 0.0331, "step": 60110 }, { "grad_norm": 0.33998093008995056, "learning_rate": 3.75350791607962e-05, "loss": 0.0234, "step": 60120 }, { "grad_norm": 0.38311365246772766, "learning_rate": 3.751906719924676e-05, "loss": 0.0238, "step": 60130 }, { "grad_norm": 0.42139431834220886, "learning_rate": 3.750305660259337e-05, "loss": 0.0249, "step": 60140 }, { "grad_norm": 0.7042794227600098, "learning_rate": 3.748704737258693e-05, "loss": 0.042, "step": 60150 }, { "grad_norm": 0.40003374218940735, "learning_rate": 3.747103951097816e-05, "loss": 0.0209, "step": 60160 }, { "grad_norm": 0.18564632534980774, "learning_rate": 3.745503301951768e-05, "loss": 0.0276, "step": 60170 }, { "grad_norm": 0.39538562297821045, "learning_rate": 3.743902789995592e-05, "loss": 0.0342, "step": 60180 }, { "grad_norm": 0.269853413105011, "learning_rate": 3.7423024154043183e-05, "loss": 0.0305, "step": 60190 }, { "grad_norm": 0.4094776511192322, "learning_rate": 3.740702178352963e-05, "loss": 0.0409, "step": 60200 }, { "grad_norm": 0.48454272747039795, "learning_rate": 3.739102079016523e-05, "loss": 0.0339, "step": 60210 }, { "grad_norm": 0.7681587338447571, "learning_rate": 3.7375021175699846e-05, "loss": 0.0354, "step": 60220 }, { "grad_norm": 0.4241175949573517, "learning_rate": 3.735902294188317e-05, "loss": 0.0233, "step": 60230 }, { "grad_norm": 0.5124614834785461, "learning_rate": 3.734302609046473e-05, "loss": 0.0326, "step": 60240 }, { "grad_norm": 0.5550968647003174, "learning_rate": 3.732703062319391e-05, "loss": 0.0298, "step": 60250 }, { "grad_norm": 0.3137231469154358, "learning_rate": 3.731103654181999e-05, "loss": 0.0244, "step": 60260 }, { "grad_norm": 0.5908140540122986, "learning_rate": 3.729504384809203e-05, "loss": 0.0335, "step": 60270 }, { "grad_norm": 0.4450008273124695, "learning_rate": 3.727905254375898e-05, "loss": 0.0286, "step": 60280 }, { "grad_norm": 0.360933393239975, "learning_rate": 3.7263062630569614e-05, "loss": 0.0322, "step": 60290 }, { "grad_norm": 0.3330884575843811, "learning_rate": 3.724707411027256e-05, "loss": 0.0292, "step": 60300 }, { "grad_norm": 0.37833884358406067, "learning_rate": 3.723108698461631e-05, "loss": 0.0273, "step": 60310 }, { "grad_norm": 0.4410476088523865, "learning_rate": 3.72151012553492e-05, "loss": 0.0209, "step": 60320 }, { "grad_norm": 0.3163236081600189, "learning_rate": 3.719911692421939e-05, "loss": 0.0229, "step": 60330 }, { "grad_norm": 0.35874828696250916, "learning_rate": 3.71831339929749e-05, "loss": 0.0406, "step": 60340 }, { "grad_norm": 0.48550882935523987, "learning_rate": 3.716715246336362e-05, "loss": 0.0278, "step": 60350 }, { "grad_norm": 0.3017290532588959, "learning_rate": 3.715117233713324e-05, "loss": 0.0258, "step": 60360 }, { "grad_norm": 0.693209171295166, "learning_rate": 3.713519361603135e-05, "loss": 0.0258, "step": 60370 }, { "grad_norm": 0.7654157876968384, "learning_rate": 3.711921630180532e-05, "loss": 0.0245, "step": 60380 }, { "grad_norm": 0.27608272433280945, "learning_rate": 3.710324039620245e-05, "loss": 0.0278, "step": 60390 }, { "grad_norm": 0.5641591548919678, "learning_rate": 3.708726590096982e-05, "loss": 0.0314, "step": 60400 }, { "grad_norm": 0.4762183129787445, "learning_rate": 3.707129281785437e-05, "loss": 0.0271, "step": 60410 }, { "grad_norm": 0.4166819453239441, "learning_rate": 3.705532114860291e-05, "loss": 0.0301, "step": 60420 }, { "grad_norm": 0.23250162601470947, "learning_rate": 3.703935089496207e-05, "loss": 0.0217, "step": 60430 }, { "grad_norm": 0.41572868824005127, "learning_rate": 3.702338205867829e-05, "loss": 0.0236, "step": 60440 }, { "grad_norm": 0.5706062316894531, "learning_rate": 3.7007414641497976e-05, "loss": 0.023, "step": 60450 }, { "grad_norm": 0.780346155166626, "learning_rate": 3.699144864516726e-05, "loss": 0.0311, "step": 60460 }, { "grad_norm": 0.5459393262863159, "learning_rate": 3.6975484071432153e-05, "loss": 0.0373, "step": 60470 }, { "grad_norm": 0.3332422971725464, "learning_rate": 3.695952092203854e-05, "loss": 0.0365, "step": 60480 }, { "grad_norm": 0.6884810328483582, "learning_rate": 3.69435591987321e-05, "loss": 0.0307, "step": 60490 }, { "grad_norm": 0.28101295232772827, "learning_rate": 3.6927598903258374e-05, "loss": 0.0238, "step": 60500 }, { "grad_norm": 0.33671167492866516, "learning_rate": 3.691164003736281e-05, "loss": 0.028, "step": 60510 }, { "grad_norm": 0.34963977336883545, "learning_rate": 3.689568260279061e-05, "loss": 0.0246, "step": 60520 }, { "grad_norm": 0.4010445475578308, "learning_rate": 3.687972660128686e-05, "loss": 0.032, "step": 60530 }, { "grad_norm": 0.6305937767028809, "learning_rate": 3.686377203459648e-05, "loss": 0.0287, "step": 60540 }, { "grad_norm": 0.5924372673034668, "learning_rate": 3.6847818904464226e-05, "loss": 0.0391, "step": 60550 }, { "grad_norm": 0.4581129252910614, "learning_rate": 3.683186721263474e-05, "loss": 0.049, "step": 60560 }, { "grad_norm": 0.6249569654464722, "learning_rate": 3.6815916960852435e-05, "loss": 0.026, "step": 60570 }, { "grad_norm": 0.49944764375686646, "learning_rate": 3.679996815086165e-05, "loss": 0.0292, "step": 60580 }, { "grad_norm": 0.5123986601829529, "learning_rate": 3.678402078440649e-05, "loss": 0.0236, "step": 60590 }, { "grad_norm": 0.3898443579673767, "learning_rate": 3.676807486323096e-05, "loss": 0.0284, "step": 60600 }, { "grad_norm": 0.41051343083381653, "learning_rate": 3.6752130389078864e-05, "loss": 0.0393, "step": 60610 }, { "grad_norm": 0.8386613726615906, "learning_rate": 3.673618736369388e-05, "loss": 0.028, "step": 60620 }, { "grad_norm": 0.4803575277328491, "learning_rate": 3.6720245788819474e-05, "loss": 0.0288, "step": 60630 }, { "grad_norm": 0.6708992719650269, "learning_rate": 3.6704305666199044e-05, "loss": 0.0362, "step": 60640 }, { "grad_norm": 0.41275888681411743, "learning_rate": 3.668836699757576e-05, "loss": 0.0269, "step": 60650 }, { "grad_norm": 0.22351881861686707, "learning_rate": 3.667242978469265e-05, "loss": 0.0448, "step": 60660 }, { "grad_norm": 0.40236571431159973, "learning_rate": 3.6656494029292575e-05, "loss": 0.0288, "step": 60670 }, { "grad_norm": 0.4959982931613922, "learning_rate": 3.664055973311825e-05, "loss": 0.0332, "step": 60680 }, { "grad_norm": 0.6918027400970459, "learning_rate": 3.662462689791221e-05, "loss": 0.0313, "step": 60690 }, { "grad_norm": 0.2525852620601654, "learning_rate": 3.660869552541689e-05, "loss": 0.0249, "step": 60700 }, { "grad_norm": 0.5613102912902832, "learning_rate": 3.659276561737448e-05, "loss": 0.0327, "step": 60710 }, { "grad_norm": 0.3281805217266083, "learning_rate": 3.657683717552708e-05, "loss": 0.0326, "step": 60720 }, { "grad_norm": 0.2884448766708374, "learning_rate": 3.6560910201616574e-05, "loss": 0.0243, "step": 60730 }, { "grad_norm": 0.3608170747756958, "learning_rate": 3.654498469738472e-05, "loss": 0.03, "step": 60740 }, { "grad_norm": 0.4440949559211731, "learning_rate": 3.6529060664573126e-05, "loss": 0.0278, "step": 60750 }, { "grad_norm": 0.4618605673313141, "learning_rate": 3.6513138104923176e-05, "loss": 0.038, "step": 60760 }, { "grad_norm": 0.5919512510299683, "learning_rate": 3.6497217020176176e-05, "loss": 0.0326, "step": 60770 }, { "grad_norm": 0.6462788581848145, "learning_rate": 3.648129741207323e-05, "loss": 0.0343, "step": 60780 }, { "grad_norm": 0.42171230912208557, "learning_rate": 3.646537928235527e-05, "loss": 0.0261, "step": 60790 }, { "grad_norm": 0.4223651587963104, "learning_rate": 3.6449462632763075e-05, "loss": 0.029, "step": 60800 }, { "grad_norm": 0.716827392578125, "learning_rate": 3.643354746503729e-05, "loss": 0.0214, "step": 60810 }, { "grad_norm": 0.3150990605354309, "learning_rate": 3.6417633780918345e-05, "loss": 0.022, "step": 60820 }, { "grad_norm": 0.39365920424461365, "learning_rate": 3.640172158214655e-05, "loss": 0.0178, "step": 60830 }, { "grad_norm": 0.2598009705543518, "learning_rate": 3.6385810870462035e-05, "loss": 0.0261, "step": 60840 }, { "grad_norm": 0.43970727920532227, "learning_rate": 3.636990164760479e-05, "loss": 0.0313, "step": 60850 }, { "grad_norm": 0.42925360798835754, "learning_rate": 3.6353993915314596e-05, "loss": 0.0351, "step": 60860 }, { "grad_norm": 0.5031397342681885, "learning_rate": 3.6338087675331124e-05, "loss": 0.016, "step": 60870 }, { "grad_norm": 0.5618613362312317, "learning_rate": 3.632218292939383e-05, "loss": 0.0258, "step": 60880 }, { "grad_norm": 0.39712288975715637, "learning_rate": 3.6306279679242033e-05, "loss": 0.0278, "step": 60890 }, { "grad_norm": 0.5129222869873047, "learning_rate": 3.629037792661492e-05, "loss": 0.0251, "step": 60900 }, { "grad_norm": 0.2881194055080414, "learning_rate": 3.627447767325147e-05, "loss": 0.0384, "step": 60910 }, { "grad_norm": 0.3869175314903259, "learning_rate": 3.62585789208905e-05, "loss": 0.0243, "step": 60920 }, { "grad_norm": 0.29760149121284485, "learning_rate": 3.624268167127068e-05, "loss": 0.0222, "step": 60930 }, { "grad_norm": 0.5793107151985168, "learning_rate": 3.622678592613051e-05, "loss": 0.0327, "step": 60940 }, { "grad_norm": 0.35861343145370483, "learning_rate": 3.6210891687208304e-05, "loss": 0.0261, "step": 60950 }, { "grad_norm": 0.20623648166656494, "learning_rate": 3.619499895624228e-05, "loss": 0.0277, "step": 60960 }, { "grad_norm": 0.3134140968322754, "learning_rate": 3.61791077349704e-05, "loss": 0.0223, "step": 60970 }, { "grad_norm": 0.5367245078086853, "learning_rate": 3.616321802513053e-05, "loss": 0.0279, "step": 60980 }, { "grad_norm": 0.31942662596702576, "learning_rate": 3.6147329828460325e-05, "loss": 0.037, "step": 60990 }, { "grad_norm": 0.3945963680744171, "learning_rate": 3.6131443146697306e-05, "loss": 0.0253, "step": 61000 }, { "grad_norm": 0.5355116128921509, "learning_rate": 3.611555798157879e-05, "loss": 0.0252, "step": 61010 }, { "grad_norm": 0.4033777117729187, "learning_rate": 3.6099674334842e-05, "loss": 0.0279, "step": 61020 }, { "grad_norm": 0.6674099564552307, "learning_rate": 3.6083792208223934e-05, "loss": 0.0209, "step": 61030 }, { "grad_norm": 0.3928966522216797, "learning_rate": 3.606791160346142e-05, "loss": 0.0282, "step": 61040 }, { "grad_norm": 0.5819295644760132, "learning_rate": 3.6052032522291134e-05, "loss": 0.0437, "step": 61050 }, { "grad_norm": 0.37951165437698364, "learning_rate": 3.603615496644962e-05, "loss": 0.0224, "step": 61060 }, { "grad_norm": 0.5463579297065735, "learning_rate": 3.60202789376732e-05, "loss": 0.0323, "step": 61070 }, { "grad_norm": 0.5102512240409851, "learning_rate": 3.6004404437698036e-05, "loss": 0.0258, "step": 61080 }, { "grad_norm": 0.3731391429901123, "learning_rate": 3.5988531468260184e-05, "loss": 0.0242, "step": 61090 }, { "grad_norm": 0.37014487385749817, "learning_rate": 3.597266003109546e-05, "loss": 0.0299, "step": 61100 }, { "grad_norm": 0.5218746662139893, "learning_rate": 3.5956790127939557e-05, "loss": 0.0259, "step": 61110 }, { "grad_norm": 0.5358002781867981, "learning_rate": 3.594092176052796e-05, "loss": 0.0314, "step": 61120 }, { "grad_norm": 0.4634091854095459, "learning_rate": 3.592505493059603e-05, "loss": 0.0336, "step": 61130 }, { "grad_norm": 0.5033720135688782, "learning_rate": 3.5909189639878905e-05, "loss": 0.0367, "step": 61140 }, { "grad_norm": 0.7945721745491028, "learning_rate": 3.5893325890111626e-05, "loss": 0.0384, "step": 61150 }, { "grad_norm": 0.3529479503631592, "learning_rate": 3.5877463683029034e-05, "loss": 0.0226, "step": 61160 }, { "grad_norm": 0.28012123703956604, "learning_rate": 3.586160302036578e-05, "loss": 0.0278, "step": 61170 }, { "grad_norm": 0.47591471672058105, "learning_rate": 3.584574390385636e-05, "loss": 0.0326, "step": 61180 }, { "grad_norm": 0.5226933360099792, "learning_rate": 3.582988633523511e-05, "loss": 0.0308, "step": 61190 }, { "grad_norm": 0.35530874133110046, "learning_rate": 3.581403031623616e-05, "loss": 0.0296, "step": 61200 }, { "grad_norm": 0.33213239908218384, "learning_rate": 3.579817584859354e-05, "loss": 0.0301, "step": 61210 }, { "grad_norm": 0.530200719833374, "learning_rate": 3.578232293404107e-05, "loss": 0.0238, "step": 61220 }, { "grad_norm": 0.7620400190353394, "learning_rate": 3.5766471574312385e-05, "loss": 0.0438, "step": 61230 }, { "grad_norm": 0.535585343837738, "learning_rate": 3.5750621771140964e-05, "loss": 0.0306, "step": 61240 }, { "grad_norm": 0.45411548018455505, "learning_rate": 3.5734773526260115e-05, "loss": 0.0351, "step": 61250 }, { "grad_norm": 0.7060451507568359, "learning_rate": 3.5718926841402996e-05, "loss": 0.0265, "step": 61260 }, { "grad_norm": 0.21715113520622253, "learning_rate": 3.5703081718302555e-05, "loss": 0.0257, "step": 61270 }, { "grad_norm": 0.7436778545379639, "learning_rate": 3.568723815869162e-05, "loss": 0.0344, "step": 61280 }, { "grad_norm": 0.24142581224441528, "learning_rate": 3.5671396164302777e-05, "loss": 0.0284, "step": 61290 }, { "grad_norm": 0.5746600031852722, "learning_rate": 3.565555573686853e-05, "loss": 0.0297, "step": 61300 }, { "grad_norm": 0.32318633794784546, "learning_rate": 3.563971687812113e-05, "loss": 0.0366, "step": 61310 }, { "grad_norm": 0.43160754442214966, "learning_rate": 3.5623879589792706e-05, "loss": 0.0228, "step": 61320 }, { "grad_norm": 0.41741490364074707, "learning_rate": 3.5608043873615185e-05, "loss": 0.0266, "step": 61330 }, { "grad_norm": 0.35126426815986633, "learning_rate": 3.559220973132035e-05, "loss": 0.0299, "step": 61340 }, { "grad_norm": 0.4712933599948883, "learning_rate": 3.557637716463981e-05, "loss": 0.0282, "step": 61350 }, { "grad_norm": 0.3730274736881256, "learning_rate": 3.556054617530497e-05, "loss": 0.0289, "step": 61360 }, { "grad_norm": 0.333132803440094, "learning_rate": 3.5544716765047106e-05, "loss": 0.0261, "step": 61370 }, { "grad_norm": 0.39040645956993103, "learning_rate": 3.5528888935597274e-05, "loss": 0.0304, "step": 61380 }, { "grad_norm": 0.45090487599372864, "learning_rate": 3.551306268868636e-05, "loss": 0.0344, "step": 61390 }, { "grad_norm": 0.35917145013809204, "learning_rate": 3.5497238026045166e-05, "loss": 0.0285, "step": 61400 }, { "grad_norm": 0.39447250962257385, "learning_rate": 3.548141494940421e-05, "loss": 0.0291, "step": 61410 }, { "grad_norm": 0.33086708188056946, "learning_rate": 3.546559346049388e-05, "loss": 0.0247, "step": 61420 }, { "grad_norm": 0.32887744903564453, "learning_rate": 3.544977356104441e-05, "loss": 0.0225, "step": 61430 }, { "grad_norm": 0.4823644757270813, "learning_rate": 3.543395525278582e-05, "loss": 0.0289, "step": 61440 }, { "grad_norm": 0.3793609142303467, "learning_rate": 3.5418138537447985e-05, "loss": 0.03, "step": 61450 }, { "grad_norm": 0.3780703544616699, "learning_rate": 3.5402323416760574e-05, "loss": 0.0455, "step": 61460 }, { "grad_norm": 0.4188382625579834, "learning_rate": 3.538650989245316e-05, "loss": 0.0237, "step": 61470 }, { "grad_norm": 0.45819780230522156, "learning_rate": 3.537069796625504e-05, "loss": 0.0294, "step": 61480 }, { "grad_norm": 0.3530000150203705, "learning_rate": 3.53548876398954e-05, "loss": 0.0209, "step": 61490 }, { "grad_norm": 0.3120771646499634, "learning_rate": 3.5339078915103214e-05, "loss": 0.0228, "step": 61500 }, { "grad_norm": 0.6216593384742737, "learning_rate": 3.532327179360733e-05, "loss": 0.028, "step": 61510 }, { "grad_norm": 0.38070565462112427, "learning_rate": 3.530746627713636e-05, "loss": 0.0249, "step": 61520 }, { "grad_norm": 0.34667298197746277, "learning_rate": 3.5291662367418805e-05, "loss": 0.0233, "step": 61530 }, { "grad_norm": 0.3337670564651489, "learning_rate": 3.527586006618293e-05, "loss": 0.0233, "step": 61540 }, { "grad_norm": 0.5377850532531738, "learning_rate": 3.526005937515687e-05, "loss": 0.0218, "step": 61550 }, { "grad_norm": 0.4620932340621948, "learning_rate": 3.524426029606856e-05, "loss": 0.0307, "step": 61560 }, { "grad_norm": 0.3222367763519287, "learning_rate": 3.5228462830645746e-05, "loss": 0.0231, "step": 61570 }, { "grad_norm": 0.43861207365989685, "learning_rate": 3.5212666980616036e-05, "loss": 0.0286, "step": 61580 }, { "grad_norm": 0.6133649349212646, "learning_rate": 3.519687274770681e-05, "loss": 0.0207, "step": 61590 }, { "grad_norm": 1.587517261505127, "learning_rate": 3.518108013364535e-05, "loss": 0.0381, "step": 61600 }, { "grad_norm": 0.5198895335197449, "learning_rate": 3.516528914015868e-05, "loss": 0.0314, "step": 61610 }, { "grad_norm": 0.6973796486854553, "learning_rate": 3.5149499768973694e-05, "loss": 0.0329, "step": 61620 }, { "grad_norm": 0.2777363061904907, "learning_rate": 3.513371202181708e-05, "loss": 0.0352, "step": 61630 }, { "grad_norm": 0.3508583903312683, "learning_rate": 3.511792590041537e-05, "loss": 0.0278, "step": 61640 }, { "grad_norm": 0.4057922065258026, "learning_rate": 3.510214140649488e-05, "loss": 0.0245, "step": 61650 }, { "grad_norm": 0.2683665156364441, "learning_rate": 3.508635854178184e-05, "loss": 0.034, "step": 61660 }, { "grad_norm": 0.4783669412136078, "learning_rate": 3.50705773080022e-05, "loss": 0.0352, "step": 61670 }, { "grad_norm": 0.6205143332481384, "learning_rate": 3.505479770688178e-05, "loss": 0.0263, "step": 61680 }, { "grad_norm": 0.46301090717315674, "learning_rate": 3.503901974014622e-05, "loss": 0.0218, "step": 61690 }, { "grad_norm": 0.5576381087303162, "learning_rate": 3.502324340952095e-05, "loss": 0.0245, "step": 61700 }, { "grad_norm": 0.4532322883605957, "learning_rate": 3.500746871673126e-05, "loss": 0.0337, "step": 61710 }, { "grad_norm": 0.5935972929000854, "learning_rate": 3.499169566350227e-05, "loss": 0.0336, "step": 61720 }, { "grad_norm": 0.3775467872619629, "learning_rate": 3.4975924251558876e-05, "loss": 0.0265, "step": 61730 }, { "grad_norm": 0.4104669988155365, "learning_rate": 3.496015448262582e-05, "loss": 0.0327, "step": 61740 }, { "grad_norm": 0.4120265245437622, "learning_rate": 3.494438635842764e-05, "loss": 0.0311, "step": 61750 }, { "grad_norm": 0.4276120662689209, "learning_rate": 3.492861988068874e-05, "loss": 0.0248, "step": 61760 }, { "grad_norm": 0.5384520292282104, "learning_rate": 3.491285505113332e-05, "loss": 0.0362, "step": 61770 }, { "grad_norm": 0.8128005862236023, "learning_rate": 3.4897091871485376e-05, "loss": 0.0341, "step": 61780 }, { "grad_norm": 0.314085453748703, "learning_rate": 3.488133034346875e-05, "loss": 0.0414, "step": 61790 }, { "grad_norm": 0.25380581617355347, "learning_rate": 3.486557046880713e-05, "loss": 0.0241, "step": 61800 }, { "grad_norm": 0.6216807961463928, "learning_rate": 3.484981224922395e-05, "loss": 0.0319, "step": 61810 }, { "grad_norm": 0.5758399367332458, "learning_rate": 3.4834055686442544e-05, "loss": 0.0277, "step": 61820 }, { "grad_norm": 0.5048778057098389, "learning_rate": 3.4818300782186e-05, "loss": 0.0418, "step": 61830 }, { "grad_norm": 0.23587608337402344, "learning_rate": 3.4802547538177236e-05, "loss": 0.0297, "step": 61840 }, { "grad_norm": 0.694584310054779, "learning_rate": 3.478679595613905e-05, "loss": 0.0269, "step": 61850 }, { "grad_norm": 0.4575558602809906, "learning_rate": 3.4771046037793975e-05, "loss": 0.0337, "step": 61860 }, { "grad_norm": 0.48034608364105225, "learning_rate": 3.4755297784864416e-05, "loss": 0.0306, "step": 61870 }, { "grad_norm": 0.8168765306472778, "learning_rate": 3.473955119907257e-05, "loss": 0.0415, "step": 61880 }, { "grad_norm": 0.5408153533935547, "learning_rate": 3.472380628214046e-05, "loss": 0.0337, "step": 61890 }, { "grad_norm": 0.3851965665817261, "learning_rate": 3.4708063035789906e-05, "loss": 0.0202, "step": 61900 }, { "grad_norm": 0.37429091334342957, "learning_rate": 3.4692321461742614e-05, "loss": 0.0288, "step": 61910 }, { "grad_norm": 0.5539854168891907, "learning_rate": 3.4676581561720023e-05, "loss": 0.0303, "step": 61920 }, { "grad_norm": 0.3133904039859772, "learning_rate": 3.466084333744344e-05, "loss": 0.0273, "step": 61930 }, { "grad_norm": 0.3989672362804413, "learning_rate": 3.4645106790633964e-05, "loss": 0.0294, "step": 61940 }, { "grad_norm": 0.4310055375099182, "learning_rate": 3.4629371923012526e-05, "loss": 0.023, "step": 61950 }, { "grad_norm": 0.3273913860321045, "learning_rate": 3.4613638736299856e-05, "loss": 0.0246, "step": 61960 }, { "grad_norm": 0.25434157252311707, "learning_rate": 3.459790723221652e-05, "loss": 0.0212, "step": 61970 }, { "grad_norm": 0.5060449242591858, "learning_rate": 3.45821774124829e-05, "loss": 0.0317, "step": 61980 }, { "grad_norm": 0.489896684885025, "learning_rate": 3.456644927881918e-05, "loss": 0.0262, "step": 61990 }, { "grad_norm": 0.33120760321617126, "learning_rate": 3.4550722832945354e-05, "loss": 0.0307, "step": 62000 }, { "grad_norm": 0.38206014037132263, "learning_rate": 3.453499807658126e-05, "loss": 0.0281, "step": 62010 }, { "grad_norm": 0.3838197588920593, "learning_rate": 3.451927501144653e-05, "loss": 0.0231, "step": 62020 }, { "grad_norm": 0.4784593880176544, "learning_rate": 3.450355363926059e-05, "loss": 0.0293, "step": 62030 }, { "grad_norm": 0.5281883478164673, "learning_rate": 3.448783396174273e-05, "loss": 0.0197, "step": 62040 }, { "grad_norm": 0.6918438673019409, "learning_rate": 3.4472115980612034e-05, "loss": 0.0248, "step": 62050 }, { "grad_norm": 0.4221719205379486, "learning_rate": 3.4456399697587396e-05, "loss": 0.0271, "step": 62060 }, { "grad_norm": 0.303338885307312, "learning_rate": 3.4440685114387506e-05, "loss": 0.0268, "step": 62070 }, { "grad_norm": 0.5184349417686462, "learning_rate": 3.442497223273091e-05, "loss": 0.0299, "step": 62080 }, { "grad_norm": 0.38642317056655884, "learning_rate": 3.4409261054335926e-05, "loss": 0.0277, "step": 62090 }, { "grad_norm": 0.3821783661842346, "learning_rate": 3.439355158092069e-05, "loss": 0.0231, "step": 62100 }, { "grad_norm": 0.34453561902046204, "learning_rate": 3.437784381420322e-05, "loss": 0.0242, "step": 62110 }, { "grad_norm": 0.3634488880634308, "learning_rate": 3.436213775590126e-05, "loss": 0.016, "step": 62120 }, { "grad_norm": 0.2770400941371918, "learning_rate": 3.4346433407732405e-05, "loss": 0.0279, "step": 62130 }, { "grad_norm": 0.4920543134212494, "learning_rate": 3.433073077141405e-05, "loss": 0.0245, "step": 62140 }, { "grad_norm": 1.760764241218567, "learning_rate": 3.431502984866343e-05, "loss": 0.0293, "step": 62150 }, { "grad_norm": 0.4316096901893616, "learning_rate": 3.429933064119753e-05, "loss": 0.0269, "step": 62160 }, { "grad_norm": 0.468458890914917, "learning_rate": 3.4283633150733255e-05, "loss": 0.0328, "step": 62170 }, { "grad_norm": 0.22730764746665955, "learning_rate": 3.4267937378987225e-05, "loss": 0.0205, "step": 62180 }, { "grad_norm": 0.510995090007782, "learning_rate": 3.425224332767592e-05, "loss": 0.0258, "step": 62190 }, { "grad_norm": 0.5592499375343323, "learning_rate": 3.4236550998515595e-05, "loss": 0.0374, "step": 62200 }, { "grad_norm": 0.2790827751159668, "learning_rate": 3.422086039322235e-05, "loss": 0.031, "step": 62210 }, { "grad_norm": 0.7268064618110657, "learning_rate": 3.420517151351209e-05, "loss": 0.0315, "step": 62220 }, { "grad_norm": 0.31494495272636414, "learning_rate": 3.418948436110053e-05, "loss": 0.0223, "step": 62230 }, { "grad_norm": 0.515012800693512, "learning_rate": 3.417379893770319e-05, "loss": 0.0318, "step": 62240 }, { "grad_norm": 0.513543963432312, "learning_rate": 3.41581152450354e-05, "loss": 0.0289, "step": 62250 }, { "grad_norm": 0.46295732259750366, "learning_rate": 3.414243328481231e-05, "loss": 0.0311, "step": 62260 }, { "grad_norm": 0.23638516664505005, "learning_rate": 3.412675305874888e-05, "loss": 0.0241, "step": 62270 }, { "grad_norm": 0.5075637698173523, "learning_rate": 3.4111074568559876e-05, "loss": 0.0241, "step": 62280 }, { "grad_norm": 0.47178328037261963, "learning_rate": 3.4095397815959846e-05, "loss": 0.0229, "step": 62290 }, { "grad_norm": 0.3267945349216461, "learning_rate": 3.407972280266322e-05, "loss": 0.0286, "step": 62300 }, { "grad_norm": 0.655913770198822, "learning_rate": 3.4064049530384166e-05, "loss": 0.0313, "step": 62310 }, { "grad_norm": 0.26161113381385803, "learning_rate": 3.40483780008367e-05, "loss": 0.0217, "step": 62320 }, { "grad_norm": 0.3654254972934723, "learning_rate": 3.4032708215734635e-05, "loss": 0.0282, "step": 62330 }, { "grad_norm": 0.4919222891330719, "learning_rate": 3.4017040176791584e-05, "loss": 0.0359, "step": 62340 }, { "grad_norm": 0.272250235080719, "learning_rate": 3.4001373885720976e-05, "loss": 0.0446, "step": 62350 }, { "grad_norm": 0.4383549094200134, "learning_rate": 3.398570934423608e-05, "loss": 0.0383, "step": 62360 }, { "grad_norm": 0.2513928711414337, "learning_rate": 3.397004655404994e-05, "loss": 0.0315, "step": 62370 }, { "grad_norm": 0.33900928497314453, "learning_rate": 3.395438551687541e-05, "loss": 0.0368, "step": 62380 }, { "grad_norm": 0.8334594964981079, "learning_rate": 3.393872623442515e-05, "loss": 0.0275, "step": 62390 }, { "grad_norm": 0.43671566247940063, "learning_rate": 3.392306870841164e-05, "loss": 0.0268, "step": 62400 }, { "grad_norm": 0.4632948338985443, "learning_rate": 3.390741294054715e-05, "loss": 0.0357, "step": 62410 }, { "grad_norm": 0.65592360496521, "learning_rate": 3.38917589325438e-05, "loss": 0.0361, "step": 62420 }, { "grad_norm": 0.5093851685523987, "learning_rate": 3.387610668611349e-05, "loss": 0.025, "step": 62430 }, { "grad_norm": 0.40850546956062317, "learning_rate": 3.38604562029679e-05, "loss": 0.0308, "step": 62440 }, { "grad_norm": 0.7069917321205139, "learning_rate": 3.384480748481855e-05, "loss": 0.0333, "step": 62450 }, { "grad_norm": 0.4582878053188324, "learning_rate": 3.382916053337676e-05, "loss": 0.0321, "step": 62460 }, { "grad_norm": 0.5700192451477051, "learning_rate": 3.381351535035366e-05, "loss": 0.0379, "step": 62470 }, { "grad_norm": 0.49074041843414307, "learning_rate": 3.379787193746019e-05, "loss": 0.0269, "step": 62480 }, { "grad_norm": 0.5137712359428406, "learning_rate": 3.378223029640709e-05, "loss": 0.0256, "step": 62490 }, { "grad_norm": 0.37735167145729065, "learning_rate": 3.376659042890489e-05, "loss": 0.0277, "step": 62500 }, { "grad_norm": 0.41193586587905884, "learning_rate": 3.375095233666397e-05, "loss": 0.033, "step": 62510 }, { "grad_norm": 0.6912750005722046, "learning_rate": 3.373531602139447e-05, "loss": 0.0672, "step": 62520 }, { "grad_norm": 0.40609416365623474, "learning_rate": 3.371968148480635e-05, "loss": 0.0194, "step": 62530 }, { "grad_norm": 0.42919665575027466, "learning_rate": 3.370404872860937e-05, "loss": 0.0256, "step": 62540 }, { "grad_norm": 0.3255288302898407, "learning_rate": 3.3688417754513145e-05, "loss": 0.0209, "step": 62550 }, { "grad_norm": 0.39714187383651733, "learning_rate": 3.367278856422703e-05, "loss": 0.0471, "step": 62560 }, { "grad_norm": 0.32408615946769714, "learning_rate": 3.365716115946021e-05, "loss": 0.0209, "step": 62570 }, { "grad_norm": 0.5224564075469971, "learning_rate": 3.3641535541921666e-05, "loss": 0.0336, "step": 62580 }, { "grad_norm": 0.4213831126689911, "learning_rate": 3.36259117133202e-05, "loss": 0.0293, "step": 62590 }, { "grad_norm": 0.4527692496776581, "learning_rate": 3.36102896753644e-05, "loss": 0.0419, "step": 62600 }, { "grad_norm": 0.4921444058418274, "learning_rate": 3.35946694297627e-05, "loss": 0.0431, "step": 62610 }, { "grad_norm": 0.40427568554878235, "learning_rate": 3.357905097822328e-05, "loss": 0.0343, "step": 62620 }, { "grad_norm": 0.2955072820186615, "learning_rate": 3.356343432245416e-05, "loss": 0.0246, "step": 62630 }, { "grad_norm": 0.3206224739551544, "learning_rate": 3.354781946416315e-05, "loss": 0.017, "step": 62640 }, { "grad_norm": 0.5237351059913635, "learning_rate": 3.3532206405057864e-05, "loss": 0.0232, "step": 62650 }, { "grad_norm": 0.531683087348938, "learning_rate": 3.351659514684573e-05, "loss": 0.0268, "step": 62660 }, { "grad_norm": 0.40890562534332275, "learning_rate": 3.350098569123395e-05, "loss": 0.0217, "step": 62670 }, { "grad_norm": 0.6425703167915344, "learning_rate": 3.348537803992958e-05, "loss": 0.028, "step": 62680 }, { "grad_norm": 0.3029231131076813, "learning_rate": 3.346977219463945e-05, "loss": 0.0186, "step": 62690 }, { "grad_norm": 0.3633541464805603, "learning_rate": 3.345416815707017e-05, "loss": 0.0352, "step": 62700 }, { "grad_norm": 0.2855181396007538, "learning_rate": 3.3438565928928185e-05, "loss": 0.0267, "step": 62710 }, { "grad_norm": 0.2591353952884674, "learning_rate": 3.3422965511919736e-05, "loss": 0.0363, "step": 62720 }, { "grad_norm": 0.6133602261543274, "learning_rate": 3.3407366907750846e-05, "loss": 0.0208, "step": 62730 }, { "grad_norm": 0.8151361346244812, "learning_rate": 3.339177011812737e-05, "loss": 0.0413, "step": 62740 }, { "grad_norm": 0.35287031531333923, "learning_rate": 3.3376175144754935e-05, "loss": 0.0291, "step": 62750 }, { "grad_norm": 0.7114384174346924, "learning_rate": 3.3360581989339004e-05, "loss": 0.0285, "step": 62760 }, { "grad_norm": 0.38809582591056824, "learning_rate": 3.3344990653584804e-05, "loss": 0.0246, "step": 62770 }, { "grad_norm": 0.3259795308113098, "learning_rate": 3.332940113919739e-05, "loss": 0.0278, "step": 62780 }, { "grad_norm": 0.4154856503009796, "learning_rate": 3.331381344788161e-05, "loss": 0.0208, "step": 62790 }, { "grad_norm": 0.40113911032676697, "learning_rate": 3.3298227581342076e-05, "loss": 0.0287, "step": 62800 }, { "grad_norm": 0.4446221888065338, "learning_rate": 3.328264354128329e-05, "loss": 0.0217, "step": 62810 }, { "grad_norm": 0.4229682981967926, "learning_rate": 3.3267061329409465e-05, "loss": 0.0231, "step": 62820 }, { "grad_norm": 0.2371945083141327, "learning_rate": 3.325148094742466e-05, "loss": 0.019, "step": 62830 }, { "grad_norm": 0.23481670022010803, "learning_rate": 3.323590239703271e-05, "loss": 0.0273, "step": 62840 }, { "grad_norm": 0.7058129906654358, "learning_rate": 3.322032567993728e-05, "loss": 0.0334, "step": 62850 }, { "grad_norm": 0.36344414949417114, "learning_rate": 3.3204750797841774e-05, "loss": 0.0253, "step": 62860 }, { "grad_norm": 0.34544849395751953, "learning_rate": 3.31891777524495e-05, "loss": 0.0295, "step": 62870 }, { "grad_norm": 0.366265207529068, "learning_rate": 3.317360654546347e-05, "loss": 0.024, "step": 62880 }, { "grad_norm": 0.444357305765152, "learning_rate": 3.3158037178586534e-05, "loss": 0.0294, "step": 62890 }, { "grad_norm": 0.37097737193107605, "learning_rate": 3.3142469653521316e-05, "loss": 0.0262, "step": 62900 }, { "grad_norm": 0.6812760829925537, "learning_rate": 3.3126903971970284e-05, "loss": 0.031, "step": 62910 }, { "grad_norm": 0.6468125581741333, "learning_rate": 3.311134013563564e-05, "loss": 0.0235, "step": 62920 }, { "grad_norm": 0.23984184861183167, "learning_rate": 3.309577814621947e-05, "loss": 0.0229, "step": 62930 }, { "grad_norm": 0.36344948410987854, "learning_rate": 3.308021800542359e-05, "loss": 0.0287, "step": 62940 }, { "grad_norm": 0.41782742738723755, "learning_rate": 3.306465971494963e-05, "loss": 0.0261, "step": 62950 }, { "grad_norm": 0.46517306566238403, "learning_rate": 3.304910327649901e-05, "loss": 0.0327, "step": 62960 }, { "grad_norm": 0.40130314230918884, "learning_rate": 3.303354869177297e-05, "loss": 0.033, "step": 62970 }, { "grad_norm": 0.2092520147562027, "learning_rate": 3.301799596247255e-05, "loss": 0.0272, "step": 62980 }, { "grad_norm": 0.5562899112701416, "learning_rate": 3.300244509029854e-05, "loss": 0.0374, "step": 62990 }, { "grad_norm": 0.4464012384414673, "learning_rate": 3.2986896076951585e-05, "loss": 0.0214, "step": 63000 }, { "grad_norm": 0.4212311804294586, "learning_rate": 3.297134892413211e-05, "loss": 0.024, "step": 63010 }, { "grad_norm": 0.3614446818828583, "learning_rate": 3.295580363354031e-05, "loss": 0.027, "step": 63020 }, { "grad_norm": 0.49947813153266907, "learning_rate": 3.2940260206876194e-05, "loss": 0.0327, "step": 63030 }, { "grad_norm": 0.4501672685146332, "learning_rate": 3.292471864583958e-05, "loss": 0.0302, "step": 63040 }, { "grad_norm": 0.46052873134613037, "learning_rate": 3.290917895213004e-05, "loss": 0.0307, "step": 63050 }, { "grad_norm": 0.46381479501724243, "learning_rate": 3.289364112744701e-05, "loss": 0.0317, "step": 63060 }, { "grad_norm": 0.41873109340667725, "learning_rate": 3.287810517348967e-05, "loss": 0.0242, "step": 63070 }, { "grad_norm": 0.21691082417964935, "learning_rate": 3.2862571091957014e-05, "loss": 0.0223, "step": 63080 }, { "grad_norm": 0.7774670124053955, "learning_rate": 3.284703888454781e-05, "loss": 0.0268, "step": 63090 }, { "grad_norm": 0.45352840423583984, "learning_rate": 3.2831508552960646e-05, "loss": 0.0271, "step": 63100 }, { "grad_norm": 0.46420982480049133, "learning_rate": 3.281598009889387e-05, "loss": 0.0374, "step": 63110 }, { "grad_norm": 0.32668617367744446, "learning_rate": 3.28004535240457e-05, "loss": 0.023, "step": 63120 }, { "grad_norm": 0.2907572388648987, "learning_rate": 3.278492883011407e-05, "loss": 0.0339, "step": 63130 }, { "grad_norm": 0.28866469860076904, "learning_rate": 3.276940601879675e-05, "loss": 0.0309, "step": 63140 }, { "grad_norm": 0.3369692862033844, "learning_rate": 3.275388509179127e-05, "loss": 0.0232, "step": 63150 }, { "grad_norm": 0.4060382544994354, "learning_rate": 3.2738366050794985e-05, "loss": 0.0277, "step": 63160 }, { "grad_norm": 0.29368895292282104, "learning_rate": 3.2722848897505034e-05, "loss": 0.0221, "step": 63170 }, { "grad_norm": 0.47953110933303833, "learning_rate": 3.2707333633618356e-05, "loss": 0.0247, "step": 63180 }, { "grad_norm": 0.25877252221107483, "learning_rate": 3.269182026083167e-05, "loss": 0.0156, "step": 63190 }, { "grad_norm": 0.5311033129692078, "learning_rate": 3.2676308780841516e-05, "loss": 0.0269, "step": 63200 }, { "grad_norm": 0.21234317123889923, "learning_rate": 3.266079919534416e-05, "loss": 0.0267, "step": 63210 }, { "grad_norm": 0.4565725326538086, "learning_rate": 3.264529150603575e-05, "loss": 0.0428, "step": 63220 }, { "grad_norm": 0.7285224795341492, "learning_rate": 3.262978571461217e-05, "loss": 0.0312, "step": 63230 }, { "grad_norm": 0.5198431611061096, "learning_rate": 3.26142818227691e-05, "loss": 0.0275, "step": 63240 }, { "grad_norm": 0.38271182775497437, "learning_rate": 3.259877983220203e-05, "loss": 0.0256, "step": 63250 }, { "grad_norm": 0.582359254360199, "learning_rate": 3.2583279744606246e-05, "loss": 0.0237, "step": 63260 }, { "grad_norm": 0.4368899166584015, "learning_rate": 3.25677815616768e-05, "loss": 0.0286, "step": 63270 }, { "grad_norm": 0.5078150629997253, "learning_rate": 3.255228528510856e-05, "loss": 0.0282, "step": 63280 }, { "grad_norm": 0.3620127737522125, "learning_rate": 3.253679091659617e-05, "loss": 0.024, "step": 63290 }, { "grad_norm": 0.27176961302757263, "learning_rate": 3.252129845783404e-05, "loss": 0.0259, "step": 63300 }, { "grad_norm": 0.44168728590011597, "learning_rate": 3.2505807910516456e-05, "loss": 0.0244, "step": 63310 }, { "grad_norm": 0.4253149628639221, "learning_rate": 3.249031927633742e-05, "loss": 0.0231, "step": 63320 }, { "grad_norm": 0.4736720323562622, "learning_rate": 3.2474832556990737e-05, "loss": 0.0284, "step": 63330 }, { "grad_norm": 0.4729926884174347, "learning_rate": 3.245934775417002e-05, "loss": 0.025, "step": 63340 }, { "grad_norm": 0.4712546765804291, "learning_rate": 3.244386486956866e-05, "loss": 0.0285, "step": 63350 }, { "grad_norm": 0.30275458097457886, "learning_rate": 3.2428383904879846e-05, "loss": 0.0259, "step": 63360 }, { "grad_norm": 0.3825109899044037, "learning_rate": 3.241290486179652e-05, "loss": 0.0207, "step": 63370 }, { "grad_norm": 0.4004799425601959, "learning_rate": 3.239742774201151e-05, "loss": 0.0249, "step": 63380 }, { "grad_norm": 0.3265618085861206, "learning_rate": 3.238195254721733e-05, "loss": 0.0293, "step": 63390 }, { "grad_norm": 0.28665876388549805, "learning_rate": 3.236647927910633e-05, "loss": 0.0211, "step": 63400 }, { "grad_norm": 0.5277240872383118, "learning_rate": 3.2351007939370646e-05, "loss": 0.0321, "step": 63410 }, { "grad_norm": 0.4772679805755615, "learning_rate": 3.2335538529702184e-05, "loss": 0.0212, "step": 63420 }, { "grad_norm": 0.5091335773468018, "learning_rate": 3.232007105179268e-05, "loss": 0.0315, "step": 63430 }, { "grad_norm": 1.024162769317627, "learning_rate": 3.230460550733363e-05, "loss": 0.0435, "step": 63440 }, { "grad_norm": 0.40111416578292847, "learning_rate": 3.2289141898016315e-05, "loss": 0.0225, "step": 63450 }, { "grad_norm": 0.8608583211898804, "learning_rate": 3.227368022553181e-05, "loss": 0.0259, "step": 63460 }, { "grad_norm": 0.4210636615753174, "learning_rate": 3.225822049157099e-05, "loss": 0.0351, "step": 63470 }, { "grad_norm": 0.34018123149871826, "learning_rate": 3.2242762697824494e-05, "loss": 0.024, "step": 63480 }, { "grad_norm": 0.4302554428577423, "learning_rate": 3.222730684598279e-05, "loss": 0.0232, "step": 63490 }, { "grad_norm": 0.4522785544395447, "learning_rate": 3.221185293773605e-05, "loss": 0.0301, "step": 63500 }, { "grad_norm": 0.19707313179969788, "learning_rate": 3.2196400974774354e-05, "loss": 0.0244, "step": 63510 }, { "grad_norm": 0.38334235548973083, "learning_rate": 3.2180950958787485e-05, "loss": 0.0256, "step": 63520 }, { "grad_norm": 0.6968550086021423, "learning_rate": 3.2165502891465024e-05, "loss": 0.0435, "step": 63530 }, { "grad_norm": 0.25206664204597473, "learning_rate": 3.215005677449636e-05, "loss": 0.0218, "step": 63540 }, { "grad_norm": 0.4436051547527313, "learning_rate": 3.2134612609570644e-05, "loss": 0.0362, "step": 63550 }, { "grad_norm": 0.45689043402671814, "learning_rate": 3.211917039837682e-05, "loss": 0.0236, "step": 63560 }, { "grad_norm": 0.2802789807319641, "learning_rate": 3.210373014260365e-05, "loss": 0.0213, "step": 63570 }, { "grad_norm": 0.3257076144218445, "learning_rate": 3.208829184393965e-05, "loss": 0.0492, "step": 63580 }, { "grad_norm": 0.3532268702983856, "learning_rate": 3.207285550407312e-05, "loss": 0.0292, "step": 63590 }, { "grad_norm": 0.4795745015144348, "learning_rate": 3.205742112469217e-05, "loss": 0.0273, "step": 63600 }, { "grad_norm": 0.30249521136283875, "learning_rate": 3.204198870748465e-05, "loss": 0.0486, "step": 63610 }, { "grad_norm": 0.182064488530159, "learning_rate": 3.202655825413824e-05, "loss": 0.0212, "step": 63620 }, { "grad_norm": 0.8190656900405884, "learning_rate": 3.2011129766340416e-05, "loss": 0.0255, "step": 63630 }, { "grad_norm": 0.2677903473377228, "learning_rate": 3.1995703245778384e-05, "loss": 0.0202, "step": 63640 }, { "grad_norm": 0.39390668272972107, "learning_rate": 3.1980278694139186e-05, "loss": 0.0282, "step": 63650 }, { "grad_norm": 0.5582869648933411, "learning_rate": 3.196485611310962e-05, "loss": 0.0404, "step": 63660 }, { "grad_norm": 0.5974416732788086, "learning_rate": 3.1949435504376257e-05, "loss": 0.0462, "step": 63670 }, { "grad_norm": 0.43028250336647034, "learning_rate": 3.1934016869625505e-05, "loss": 0.0269, "step": 63680 }, { "grad_norm": 0.4086562991142273, "learning_rate": 3.191860021054349e-05, "loss": 0.0246, "step": 63690 }, { "grad_norm": 0.222966268658638, "learning_rate": 3.190318552881618e-05, "loss": 0.0383, "step": 63700 }, { "grad_norm": 0.37867245078086853, "learning_rate": 3.1887772826129285e-05, "loss": 0.0182, "step": 63710 }, { "grad_norm": 0.3045664131641388, "learning_rate": 3.1872362104168335e-05, "loss": 0.0249, "step": 63720 }, { "grad_norm": 0.5572736859321594, "learning_rate": 3.185695336461861e-05, "loss": 0.0264, "step": 63730 }, { "grad_norm": 0.30200016498565674, "learning_rate": 3.184154660916519e-05, "loss": 0.0246, "step": 63740 }, { "grad_norm": 0.24024955928325653, "learning_rate": 3.1826141839492906e-05, "loss": 0.0248, "step": 63750 }, { "grad_norm": 0.5613534450531006, "learning_rate": 3.181073905728644e-05, "loss": 0.0252, "step": 63760 }, { "grad_norm": 0.261852890253067, "learning_rate": 3.1795338264230214e-05, "loss": 0.0277, "step": 63770 }, { "grad_norm": 0.6631220579147339, "learning_rate": 3.1779939462008426e-05, "loss": 0.0494, "step": 63780 }, { "grad_norm": 0.22018055617809296, "learning_rate": 3.1764542652305066e-05, "loss": 0.0231, "step": 63790 }, { "grad_norm": 0.4869028627872467, "learning_rate": 3.17491478368039e-05, "loss": 0.0327, "step": 63800 }, { "grad_norm": 0.49438250064849854, "learning_rate": 3.173375501718847e-05, "loss": 0.0325, "step": 63810 }, { "grad_norm": 0.4120804965496063, "learning_rate": 3.171836419514215e-05, "loss": 0.0448, "step": 63820 }, { "grad_norm": 0.3045046925544739, "learning_rate": 3.170297537234803e-05, "loss": 0.0265, "step": 63830 }, { "grad_norm": 0.43217548727989197, "learning_rate": 3.168758855048901e-05, "loss": 0.0448, "step": 63840 }, { "grad_norm": 4.4208598136901855, "learning_rate": 3.1672203731247785e-05, "loss": 0.03, "step": 63850 }, { "grad_norm": 0.33540186285972595, "learning_rate": 3.16568209163068e-05, "loss": 0.0238, "step": 63860 }, { "grad_norm": 0.9103405475616455, "learning_rate": 3.1641440107348295e-05, "loss": 0.0208, "step": 63870 }, { "grad_norm": 0.32062309980392456, "learning_rate": 3.162606130605429e-05, "loss": 0.0229, "step": 63880 }, { "grad_norm": 0.5148916840553284, "learning_rate": 3.16106845141066e-05, "loss": 0.0277, "step": 63890 }, { "grad_norm": 0.4473513066768646, "learning_rate": 3.159530973318682e-05, "loss": 0.0332, "step": 63900 }, { "grad_norm": 0.5099276900291443, "learning_rate": 3.1579936964976286e-05, "loss": 0.0248, "step": 63910 }, { "grad_norm": 0.2582869231700897, "learning_rate": 3.156456621115615e-05, "loss": 0.0381, "step": 63920 }, { "grad_norm": 0.3912651836872101, "learning_rate": 3.154919747340734e-05, "loss": 0.0232, "step": 63930 }, { "grad_norm": 0.3811185956001282, "learning_rate": 3.153383075341054e-05, "loss": 0.03, "step": 63940 }, { "grad_norm": 0.434771865606308, "learning_rate": 3.1518466052846264e-05, "loss": 0.0262, "step": 63950 }, { "grad_norm": 0.39434173703193665, "learning_rate": 3.150310337339474e-05, "loss": 0.0311, "step": 63960 }, { "grad_norm": 0.6612309813499451, "learning_rate": 3.148774271673603e-05, "loss": 0.0295, "step": 63970 }, { "grad_norm": 0.2880444824695587, "learning_rate": 3.147238408454995e-05, "loss": 0.0399, "step": 63980 }, { "grad_norm": 0.8458864688873291, "learning_rate": 3.1457027478516086e-05, "loss": 0.0269, "step": 63990 }, { "grad_norm": 0.6227560639381409, "learning_rate": 3.144167290031382e-05, "loss": 0.0308, "step": 64000 }, { "grad_norm": 0.4528515636920929, "learning_rate": 3.142632035162227e-05, "loss": 0.0331, "step": 64010 }, { "grad_norm": 0.5523889660835266, "learning_rate": 3.141096983412044e-05, "loss": 0.0265, "step": 64020 }, { "grad_norm": 0.4218192994594574, "learning_rate": 3.139562134948699e-05, "loss": 0.0325, "step": 64030 }, { "grad_norm": 0.6584253907203674, "learning_rate": 3.1380274899400424e-05, "loss": 0.028, "step": 64040 }, { "grad_norm": 0.2693951725959778, "learning_rate": 3.1364930485539e-05, "loss": 0.0181, "step": 64050 }, { "grad_norm": 0.7468874454498291, "learning_rate": 3.134958810958076e-05, "loss": 0.0319, "step": 64060 }, { "grad_norm": 0.5118517279624939, "learning_rate": 3.13342477732035e-05, "loss": 0.0206, "step": 64070 }, { "grad_norm": 0.42051905393600464, "learning_rate": 3.131890947808487e-05, "loss": 0.0345, "step": 64080 }, { "grad_norm": 0.30910810828208923, "learning_rate": 3.13035732259022e-05, "loss": 0.024, "step": 64090 }, { "grad_norm": 0.8883373141288757, "learning_rate": 3.1288239018332656e-05, "loss": 0.026, "step": 64100 }, { "grad_norm": 0.42503777146339417, "learning_rate": 3.1272906857053164e-05, "loss": 0.031, "step": 64110 }, { "grad_norm": 0.49838384985923767, "learning_rate": 3.125757674374042e-05, "loss": 0.0288, "step": 64120 }, { "grad_norm": 0.9479433298110962, "learning_rate": 3.124224868007088e-05, "loss": 0.0341, "step": 64130 }, { "grad_norm": 0.3187352120876312, "learning_rate": 3.1226922667720854e-05, "loss": 0.025, "step": 64140 }, { "grad_norm": 0.39416947960853577, "learning_rate": 3.121159870836633e-05, "loss": 0.0192, "step": 64150 }, { "grad_norm": 0.4899378716945648, "learning_rate": 3.119627680368312e-05, "loss": 0.0211, "step": 64160 }, { "grad_norm": 0.5587776303291321, "learning_rate": 3.1180956955346816e-05, "loss": 0.0309, "step": 64170 }, { "grad_norm": 0.32144594192504883, "learning_rate": 3.116563916503276e-05, "loss": 0.0268, "step": 64180 }, { "grad_norm": 0.2758913040161133, "learning_rate": 3.1150323434416095e-05, "loss": 0.0309, "step": 64190 }, { "grad_norm": 0.541075587272644, "learning_rate": 3.1135009765171696e-05, "loss": 0.0242, "step": 64200 }, { "grad_norm": 0.3278258144855499, "learning_rate": 3.111969815897429e-05, "loss": 0.0349, "step": 64210 }, { "grad_norm": 0.8398790955543518, "learning_rate": 3.1104388617498295e-05, "loss": 0.0219, "step": 64220 }, { "grad_norm": 0.33123424649238586, "learning_rate": 3.1089081142417966e-05, "loss": 0.0206, "step": 64230 }, { "grad_norm": 0.3637530505657196, "learning_rate": 3.1073775735407285e-05, "loss": 0.0203, "step": 64240 }, { "grad_norm": 0.33759546279907227, "learning_rate": 3.1058472398140035e-05, "loss": 0.0281, "step": 64250 }, { "grad_norm": 0.5384575724601746, "learning_rate": 3.1043171132289736e-05, "loss": 0.035, "step": 64260 }, { "grad_norm": 0.49599605798721313, "learning_rate": 3.1027871939529774e-05, "loss": 0.0282, "step": 64270 }, { "grad_norm": 0.34102410078048706, "learning_rate": 3.1012574821533214e-05, "loss": 0.0172, "step": 64280 }, { "grad_norm": 0.20304274559020996, "learning_rate": 3.099727977997291e-05, "loss": 0.0174, "step": 64290 }, { "grad_norm": 0.5512796640396118, "learning_rate": 3.098198681652154e-05, "loss": 0.0294, "step": 64300 }, { "grad_norm": 0.4577024579048157, "learning_rate": 3.0966695932851475e-05, "loss": 0.0244, "step": 64310 }, { "grad_norm": 0.4912836253643036, "learning_rate": 3.0951407130634926e-05, "loss": 0.03, "step": 64320 }, { "grad_norm": 2.1543288230895996, "learning_rate": 3.0936120411543864e-05, "loss": 0.029, "step": 64330 }, { "grad_norm": 0.40542855858802795, "learning_rate": 3.0920835777250015e-05, "loss": 0.0259, "step": 64340 }, { "grad_norm": 0.4739944636821747, "learning_rate": 3.090555322942488e-05, "loss": 0.021, "step": 64350 }, { "grad_norm": 1.038339614868164, "learning_rate": 3.089027276973973e-05, "loss": 0.03, "step": 64360 }, { "grad_norm": 0.4696024954319, "learning_rate": 3.0874994399865615e-05, "loss": 0.0237, "step": 64370 }, { "grad_norm": 0.5105630159378052, "learning_rate": 3.085971812147337e-05, "loss": 0.0213, "step": 64380 }, { "grad_norm": 0.40249496698379517, "learning_rate": 3.0844443936233557e-05, "loss": 0.0208, "step": 64390 }, { "grad_norm": 0.3152383267879486, "learning_rate": 3.082917184581657e-05, "loss": 0.0258, "step": 64400 }, { "grad_norm": 0.2039627581834793, "learning_rate": 3.0813901851892513e-05, "loss": 0.0274, "step": 64410 }, { "grad_norm": 0.45995259284973145, "learning_rate": 3.0798633956131316e-05, "loss": 0.0327, "step": 64420 }, { "grad_norm": 0.5738227963447571, "learning_rate": 3.078336816020264e-05, "loss": 0.0219, "step": 64430 }, { "grad_norm": 0.543293833732605, "learning_rate": 3.0768104465775935e-05, "loss": 0.0343, "step": 64440 }, { "grad_norm": 0.33903518319129944, "learning_rate": 3.0752842874520396e-05, "loss": 0.0175, "step": 64450 }, { "grad_norm": 0.44372159242630005, "learning_rate": 3.073758338810503e-05, "loss": 0.0211, "step": 64460 }, { "grad_norm": 0.5389677882194519, "learning_rate": 3.072232600819859e-05, "loss": 0.0322, "step": 64470 }, { "grad_norm": 0.7503824234008789, "learning_rate": 3.070707073646959e-05, "loss": 0.0279, "step": 64480 }, { "grad_norm": 0.25018373131752014, "learning_rate": 3.069181757458633e-05, "loss": 0.0322, "step": 64490 }, { "grad_norm": 0.5344921946525574, "learning_rate": 3.067656652421687e-05, "loss": 0.0356, "step": 64500 }, { "grad_norm": 0.20884104073047638, "learning_rate": 3.0661317587029025e-05, "loss": 0.0266, "step": 64510 }, { "grad_norm": 0.5027360916137695, "learning_rate": 3.064607076469043e-05, "loss": 0.0305, "step": 64520 }, { "grad_norm": 0.47523441910743713, "learning_rate": 3.063082605886845e-05, "loss": 0.0311, "step": 64530 }, { "grad_norm": 0.2953278422355652, "learning_rate": 3.061558347123019e-05, "loss": 0.021, "step": 64540 }, { "grad_norm": 0.22958320379257202, "learning_rate": 3.060034300344259e-05, "loss": 0.0214, "step": 64550 }, { "grad_norm": 0.4650949537754059, "learning_rate": 3.058510465717231e-05, "loss": 0.0344, "step": 64560 }, { "grad_norm": 0.27299386262893677, "learning_rate": 3.0569868434085804e-05, "loss": 0.0218, "step": 64570 }, { "grad_norm": 0.36687272787094116, "learning_rate": 3.055463433584924e-05, "loss": 0.0251, "step": 64580 }, { "grad_norm": 0.3571387529373169, "learning_rate": 3.0539402364128656e-05, "loss": 0.0219, "step": 64590 }, { "grad_norm": 0.5792713165283203, "learning_rate": 3.052417252058977e-05, "loss": 0.027, "step": 64600 }, { "grad_norm": 1.9093399047851562, "learning_rate": 3.0508944806898092e-05, "loss": 0.035, "step": 64610 }, { "grad_norm": 0.6033287644386292, "learning_rate": 3.04937192247189e-05, "loss": 0.0299, "step": 64620 }, { "grad_norm": 0.28557470440864563, "learning_rate": 3.047849577571725e-05, "loss": 0.0232, "step": 64630 }, { "grad_norm": 0.4911213219165802, "learning_rate": 3.046327446155795e-05, "loss": 0.0259, "step": 64640 }, { "grad_norm": 0.3570463955402374, "learning_rate": 3.0448055283905587e-05, "loss": 0.0323, "step": 64650 }, { "grad_norm": 0.7886012196540833, "learning_rate": 3.0432838244424494e-05, "loss": 0.0247, "step": 64660 }, { "grad_norm": 0.4913657605648041, "learning_rate": 3.04176233447788e-05, "loss": 0.029, "step": 64670 }, { "grad_norm": 0.3324507772922516, "learning_rate": 3.040241058663238e-05, "loss": 0.0219, "step": 64680 }, { "grad_norm": 0.5362446904182434, "learning_rate": 3.038719997164887e-05, "loss": 0.0215, "step": 64690 }, { "grad_norm": 0.4259251058101654, "learning_rate": 3.0371991501491683e-05, "loss": 0.0213, "step": 64700 }, { "grad_norm": 0.45156461000442505, "learning_rate": 3.0356785177823975e-05, "loss": 0.0243, "step": 64710 }, { "grad_norm": 0.5250648856163025, "learning_rate": 3.0341581002308726e-05, "loss": 0.0236, "step": 64720 }, { "grad_norm": 0.3502635061740875, "learning_rate": 3.0326378976608627e-05, "loss": 0.0215, "step": 64730 }, { "grad_norm": 0.33703479170799255, "learning_rate": 3.0311179102386146e-05, "loss": 0.0274, "step": 64740 }, { "grad_norm": 0.31311631202697754, "learning_rate": 3.0295981381303513e-05, "loss": 0.0286, "step": 64750 }, { "grad_norm": 0.3260096311569214, "learning_rate": 3.0280785815022727e-05, "loss": 0.0221, "step": 64760 }, { "grad_norm": 0.41314274072647095, "learning_rate": 3.0265592405205544e-05, "loss": 0.019, "step": 64770 }, { "grad_norm": 0.6007036566734314, "learning_rate": 3.0250401153513517e-05, "loss": 0.0179, "step": 64780 }, { "grad_norm": 0.372677206993103, "learning_rate": 3.0235212061607936e-05, "loss": 0.0279, "step": 64790 }, { "grad_norm": 0.8029396533966064, "learning_rate": 3.0220025131149842e-05, "loss": 0.035, "step": 64800 }, { "grad_norm": 0.7374178767204285, "learning_rate": 3.020484036380007e-05, "loss": 0.0231, "step": 64810 }, { "grad_norm": 0.40961962938308716, "learning_rate": 3.0189657761219188e-05, "loss": 0.0294, "step": 64820 }, { "grad_norm": 0.5515984296798706, "learning_rate": 3.017447732506753e-05, "loss": 0.0176, "step": 64830 }, { "grad_norm": 0.4592876434326172, "learning_rate": 3.015929905700524e-05, "loss": 0.0339, "step": 64840 }, { "grad_norm": 0.4276443421840668, "learning_rate": 3.014412295869218e-05, "loss": 0.0296, "step": 64850 }, { "grad_norm": 0.7562962174415588, "learning_rate": 3.0128949031787977e-05, "loss": 0.0261, "step": 64860 }, { "grad_norm": 0.6691019535064697, "learning_rate": 3.0113777277952022e-05, "loss": 0.0325, "step": 64870 }, { "grad_norm": 0.30634805560112, "learning_rate": 3.0098607698843496e-05, "loss": 0.0225, "step": 64880 }, { "grad_norm": 0.43581944704055786, "learning_rate": 3.0083440296121312e-05, "loss": 0.0271, "step": 64890 }, { "grad_norm": 0.478030264377594, "learning_rate": 3.0068275071444142e-05, "loss": 0.0269, "step": 64900 }, { "grad_norm": 0.3726937174797058, "learning_rate": 3.0053112026470442e-05, "loss": 0.0264, "step": 64910 }, { "grad_norm": 0.35499152541160583, "learning_rate": 3.0037951162858434e-05, "loss": 0.025, "step": 64920 }, { "grad_norm": 0.49177926778793335, "learning_rate": 3.002279248226607e-05, "loss": 0.0301, "step": 64930 }, { "grad_norm": 0.28083932399749756, "learning_rate": 3.000763598635109e-05, "loss": 0.0329, "step": 64940 }, { "grad_norm": 0.4877203404903412, "learning_rate": 2.999248167677098e-05, "loss": 0.0287, "step": 64950 }, { "grad_norm": 0.42264240980148315, "learning_rate": 2.9977329555182976e-05, "loss": 0.0243, "step": 64960 }, { "grad_norm": 0.4453458786010742, "learning_rate": 2.9962179623244124e-05, "loss": 0.024, "step": 64970 }, { "grad_norm": 0.5996779799461365, "learning_rate": 2.9947031882611175e-05, "loss": 0.0273, "step": 64980 }, { "grad_norm": 0.378772497177124, "learning_rate": 2.9931886334940685e-05, "loss": 0.0197, "step": 64990 }, { "grad_norm": 0.5091874003410339, "learning_rate": 2.9916742981888923e-05, "loss": 0.0261, "step": 65000 }, { "grad_norm": 0.3308754861354828, "learning_rate": 2.990160182511196e-05, "loss": 0.0188, "step": 65010 }, { "grad_norm": 0.4569283425807953, "learning_rate": 2.9886462866265574e-05, "loss": 0.0259, "step": 65020 }, { "grad_norm": 0.2256559133529663, "learning_rate": 2.9871326107005392e-05, "loss": 0.0251, "step": 65030 }, { "grad_norm": 0.3718191683292389, "learning_rate": 2.9856191548986735e-05, "loss": 0.0214, "step": 65040 }, { "grad_norm": 0.4144100844860077, "learning_rate": 2.9841059193864672e-05, "loss": 0.0212, "step": 65050 }, { "grad_norm": 0.24264109134674072, "learning_rate": 2.982592904329407e-05, "loss": 0.0204, "step": 65060 }, { "grad_norm": 0.37558409571647644, "learning_rate": 2.981080109892954e-05, "loss": 0.0248, "step": 65070 }, { "grad_norm": 1.1032638549804688, "learning_rate": 2.9795675362425434e-05, "loss": 0.0273, "step": 65080 }, { "grad_norm": 0.5323411226272583, "learning_rate": 2.97805518354359e-05, "loss": 0.0398, "step": 65090 }, { "grad_norm": 0.5650411248207092, "learning_rate": 2.9765430519614825e-05, "loss": 0.0303, "step": 65100 }, { "grad_norm": 0.7783190011978149, "learning_rate": 2.975031141661584e-05, "loss": 0.0276, "step": 65110 }, { "grad_norm": 0.2810123860836029, "learning_rate": 2.9735194528092347e-05, "loss": 0.0255, "step": 65120 }, { "grad_norm": 0.21876420080661774, "learning_rate": 2.9720079855697525e-05, "loss": 0.0251, "step": 65130 }, { "grad_norm": 0.34676212072372437, "learning_rate": 2.970496740108427e-05, "loss": 0.0324, "step": 65140 }, { "grad_norm": 0.6257414221763611, "learning_rate": 2.968985716590527e-05, "loss": 0.0202, "step": 65150 }, { "grad_norm": 0.4631596505641937, "learning_rate": 2.9674749151812953e-05, "loss": 0.0276, "step": 65160 }, { "grad_norm": 0.5425687432289124, "learning_rate": 2.965964336045951e-05, "loss": 0.0239, "step": 65170 }, { "grad_norm": 0.8018279075622559, "learning_rate": 2.9644539793496894e-05, "loss": 0.0281, "step": 65180 }, { "grad_norm": 0.3142433166503906, "learning_rate": 2.9629438452576807e-05, "loss": 0.0198, "step": 65190 }, { "grad_norm": 0.5458557605743408, "learning_rate": 2.9614339339350698e-05, "loss": 0.0259, "step": 65200 }, { "grad_norm": 0.22540299594402313, "learning_rate": 2.959924245546979e-05, "loss": 0.023, "step": 65210 }, { "grad_norm": 0.564355731010437, "learning_rate": 2.9584147802585038e-05, "loss": 0.0263, "step": 65220 }, { "grad_norm": 0.7809672951698303, "learning_rate": 2.956905538234721e-05, "loss": 0.0384, "step": 65230 }, { "grad_norm": 0.3505743145942688, "learning_rate": 2.9553965196406764e-05, "loss": 0.0255, "step": 65240 }, { "grad_norm": 0.6003190875053406, "learning_rate": 2.9538877246413943e-05, "loss": 0.0313, "step": 65250 }, { "grad_norm": 0.5011919736862183, "learning_rate": 2.952379153401874e-05, "loss": 0.0184, "step": 65260 }, { "grad_norm": 0.3887672424316406, "learning_rate": 2.9508708060870914e-05, "loss": 0.0275, "step": 65270 }, { "grad_norm": 0.29262712597846985, "learning_rate": 2.9493626828619948e-05, "loss": 0.024, "step": 65280 }, { "grad_norm": 0.7817725539207458, "learning_rate": 2.9478547838915137e-05, "loss": 0.0262, "step": 65290 }, { "grad_norm": 0.29081496596336365, "learning_rate": 2.946347109340548e-05, "loss": 0.0239, "step": 65300 }, { "grad_norm": 0.4118783473968506, "learning_rate": 2.944839659373976e-05, "loss": 0.0275, "step": 65310 }, { "grad_norm": 0.4706691801548004, "learning_rate": 2.943332434156647e-05, "loss": 0.0243, "step": 65320 }, { "grad_norm": 0.223039910197258, "learning_rate": 2.9418254338533912e-05, "loss": 0.0269, "step": 65330 }, { "grad_norm": 0.5479937195777893, "learning_rate": 2.940318658629011e-05, "loss": 0.0271, "step": 65340 }, { "grad_norm": 0.514360249042511, "learning_rate": 2.938812108648287e-05, "loss": 0.0228, "step": 65350 }, { "grad_norm": 0.4288012385368347, "learning_rate": 2.937305784075972e-05, "loss": 0.023, "step": 65360 }, { "grad_norm": 0.5438146591186523, "learning_rate": 2.9357996850767937e-05, "loss": 0.0302, "step": 65370 }, { "grad_norm": 0.6406678557395935, "learning_rate": 2.9342938118154595e-05, "loss": 0.023, "step": 65380 }, { "grad_norm": 0.5134913921356201, "learning_rate": 2.9327881644566485e-05, "loss": 0.0218, "step": 65390 }, { "grad_norm": 0.2979176342487335, "learning_rate": 2.9312827431650157e-05, "loss": 0.0302, "step": 65400 }, { "grad_norm": 0.6523600220680237, "learning_rate": 2.92977754810519e-05, "loss": 0.0228, "step": 65410 }, { "grad_norm": 0.7210808992385864, "learning_rate": 2.928272579441781e-05, "loss": 0.0381, "step": 65420 }, { "grad_norm": 0.6811108589172363, "learning_rate": 2.9267678373393676e-05, "loss": 0.0387, "step": 65430 }, { "grad_norm": 0.4047669768333435, "learning_rate": 2.925263321962507e-05, "loss": 0.0218, "step": 65440 }, { "grad_norm": 0.25142204761505127, "learning_rate": 2.9237590334757292e-05, "loss": 0.0238, "step": 65450 }, { "grad_norm": 0.695768415927887, "learning_rate": 2.922254972043543e-05, "loss": 0.0238, "step": 65460 }, { "grad_norm": 0.42219677567481995, "learning_rate": 2.9207511378304265e-05, "loss": 0.0235, "step": 65470 }, { "grad_norm": 0.6363299489021301, "learning_rate": 2.9192475310008428e-05, "loss": 0.0284, "step": 65480 }, { "grad_norm": 0.27646782994270325, "learning_rate": 2.917744151719219e-05, "loss": 0.0224, "step": 65490 }, { "grad_norm": 1.0081851482391357, "learning_rate": 2.916241000149964e-05, "loss": 0.0277, "step": 65500 }, { "grad_norm": 0.6882208585739136, "learning_rate": 2.9147380764574622e-05, "loss": 0.0242, "step": 65510 }, { "grad_norm": 0.19700515270233154, "learning_rate": 2.9132353808060674e-05, "loss": 0.0219, "step": 65520 }, { "grad_norm": 0.398360937833786, "learning_rate": 2.9117329133601134e-05, "loss": 0.0194, "step": 65530 }, { "grad_norm": 0.283997118473053, "learning_rate": 2.9102306742839093e-05, "loss": 0.0274, "step": 65540 }, { "grad_norm": 0.43172648549079895, "learning_rate": 2.9087286637417375e-05, "loss": 0.0354, "step": 65550 }, { "grad_norm": 0.6912000775337219, "learning_rate": 2.9072268818978537e-05, "loss": 0.0254, "step": 65560 }, { "grad_norm": 0.3485567569732666, "learning_rate": 2.905725328916493e-05, "loss": 0.0196, "step": 65570 }, { "grad_norm": 0.2930862009525299, "learning_rate": 2.9042240049618608e-05, "loss": 0.0201, "step": 65580 }, { "grad_norm": 0.38565945625305176, "learning_rate": 2.9027229101981413e-05, "loss": 0.0347, "step": 65590 }, { "grad_norm": 0.7768843770027161, "learning_rate": 2.901222044789488e-05, "loss": 0.0274, "step": 65600 }, { "grad_norm": 0.3963560163974762, "learning_rate": 2.89972140890004e-05, "loss": 0.0224, "step": 65610 }, { "grad_norm": 0.4370405077934265, "learning_rate": 2.8982210026938988e-05, "loss": 0.0219, "step": 65620 }, { "grad_norm": 0.5092225670814514, "learning_rate": 2.896720826335151e-05, "loss": 0.03, "step": 65630 }, { "grad_norm": 0.41976192593574524, "learning_rate": 2.8952208799878493e-05, "loss": 0.0199, "step": 65640 }, { "grad_norm": 0.4633350074291229, "learning_rate": 2.89372116381603e-05, "loss": 0.0211, "step": 65650 }, { "grad_norm": 0.6445507407188416, "learning_rate": 2.8922216779836932e-05, "loss": 0.0448, "step": 65660 }, { "grad_norm": 0.38240885734558105, "learning_rate": 2.8907224226548285e-05, "loss": 0.0214, "step": 65670 }, { "grad_norm": 0.3163401782512665, "learning_rate": 2.8892233979933868e-05, "loss": 0.0304, "step": 65680 }, { "grad_norm": 0.5379208922386169, "learning_rate": 2.8877246041633026e-05, "loss": 0.0368, "step": 65690 }, { "grad_norm": 0.38936397433280945, "learning_rate": 2.886226041328478e-05, "loss": 0.0296, "step": 65700 }, { "grad_norm": 0.32456839084625244, "learning_rate": 2.8847277096527957e-05, "loss": 0.0239, "step": 65710 }, { "grad_norm": 0.34488844871520996, "learning_rate": 2.8832296093001105e-05, "loss": 0.0401, "step": 65720 }, { "grad_norm": 0.44225645065307617, "learning_rate": 2.8817317404342547e-05, "loss": 0.025, "step": 65730 }, { "grad_norm": 0.3569585978984833, "learning_rate": 2.880234103219029e-05, "loss": 0.0248, "step": 65740 }, { "grad_norm": 0.6251403093338013, "learning_rate": 2.878736697818215e-05, "loss": 0.0336, "step": 65750 }, { "grad_norm": 0.7539202570915222, "learning_rate": 2.877239524395568e-05, "loss": 0.028, "step": 65760 }, { "grad_norm": 0.3541563153266907, "learning_rate": 2.8757425831148134e-05, "loss": 0.0225, "step": 65770 }, { "grad_norm": 0.3251822888851166, "learning_rate": 2.874245874139658e-05, "loss": 0.0185, "step": 65780 }, { "grad_norm": 0.5472338199615479, "learning_rate": 2.8727493976337737e-05, "loss": 0.0383, "step": 65790 }, { "grad_norm": 0.215046226978302, "learning_rate": 2.8712531537608218e-05, "loss": 0.026, "step": 65800 }, { "grad_norm": 0.36899513006210327, "learning_rate": 2.8697571426844217e-05, "loss": 0.0297, "step": 65810 }, { "grad_norm": 0.34914109110832214, "learning_rate": 2.86826136456818e-05, "loss": 0.029, "step": 65820 }, { "grad_norm": 0.47741439938545227, "learning_rate": 2.8667658195756685e-05, "loss": 0.0265, "step": 65830 }, { "grad_norm": 0.45842912793159485, "learning_rate": 2.8652705078704412e-05, "loss": 0.0312, "step": 65840 }, { "grad_norm": 0.4126265347003937, "learning_rate": 2.863775429616018e-05, "loss": 0.0289, "step": 65850 }, { "grad_norm": 0.323968768119812, "learning_rate": 2.862280584975906e-05, "loss": 0.0188, "step": 65860 }, { "grad_norm": 0.5426278710365295, "learning_rate": 2.8607859741135735e-05, "loss": 0.0276, "step": 65870 }, { "grad_norm": 0.3134753406047821, "learning_rate": 2.8592915971924723e-05, "loss": 0.0225, "step": 65880 }, { "grad_norm": 0.3313499987125397, "learning_rate": 2.8577974543760212e-05, "loss": 0.0256, "step": 65890 }, { "grad_norm": 0.6477792263031006, "learning_rate": 2.8563035458276222e-05, "loss": 0.0267, "step": 65900 }, { "grad_norm": 0.4416424036026001, "learning_rate": 2.8548098717106424e-05, "loss": 0.0234, "step": 65910 }, { "grad_norm": 0.7527157664299011, "learning_rate": 2.85331643218843e-05, "loss": 0.0265, "step": 65920 }, { "grad_norm": 0.593618631362915, "learning_rate": 2.8518232274243046e-05, "loss": 0.0303, "step": 65930 }, { "grad_norm": 0.27685102820396423, "learning_rate": 2.8503302575815638e-05, "loss": 0.023, "step": 65940 }, { "grad_norm": 0.439047634601593, "learning_rate": 2.8488375228234716e-05, "loss": 0.018, "step": 65950 }, { "grad_norm": 0.305162638425827, "learning_rate": 2.8473450233132738e-05, "loss": 0.0269, "step": 65960 }, { "grad_norm": 0.3552190363407135, "learning_rate": 2.8458527592141898e-05, "loss": 0.0233, "step": 65970 }, { "grad_norm": 0.3354339897632599, "learning_rate": 2.8443607306894073e-05, "loss": 0.0229, "step": 65980 }, { "grad_norm": 0.43528759479522705, "learning_rate": 2.8428689379020946e-05, "loss": 0.0288, "step": 65990 }, { "grad_norm": 0.5347820520401001, "learning_rate": 2.841377381015391e-05, "loss": 0.0295, "step": 66000 }, { "grad_norm": 0.36731529235839844, "learning_rate": 2.8398860601924144e-05, "loss": 0.0224, "step": 66010 }, { "grad_norm": 0.3770519196987152, "learning_rate": 2.838394975596248e-05, "loss": 0.0312, "step": 66020 }, { "grad_norm": 0.31349170207977295, "learning_rate": 2.83690412738996e-05, "loss": 0.0183, "step": 66030 }, { "grad_norm": 0.45641136169433594, "learning_rate": 2.835413515736579e-05, "loss": 0.0215, "step": 66040 }, { "grad_norm": 0.3277398347854614, "learning_rate": 2.8339231407991267e-05, "loss": 0.0295, "step": 66050 }, { "grad_norm": 0.6593809127807617, "learning_rate": 2.832433002740581e-05, "loss": 0.0193, "step": 66060 }, { "grad_norm": 0.52291339635849, "learning_rate": 2.8309431017239047e-05, "loss": 0.0199, "step": 66070 }, { "grad_norm": 0.3990444540977478, "learning_rate": 2.829453437912028e-05, "loss": 0.0197, "step": 66080 }, { "grad_norm": 0.39198827743530273, "learning_rate": 2.8279640114678617e-05, "loss": 0.0326, "step": 66090 }, { "grad_norm": 0.5510401129722595, "learning_rate": 2.8264748225542836e-05, "loss": 0.0282, "step": 66100 }, { "grad_norm": 0.8128154277801514, "learning_rate": 2.8249858713341516e-05, "loss": 0.0341, "step": 66110 }, { "grad_norm": 0.37130019068717957, "learning_rate": 2.823497157970294e-05, "loss": 0.0288, "step": 66120 }, { "grad_norm": 0.38248559832572937, "learning_rate": 2.8220086826255164e-05, "loss": 0.0258, "step": 66130 }, { "grad_norm": 0.22382043302059174, "learning_rate": 2.8205204454625934e-05, "loss": 0.0245, "step": 66140 }, { "grad_norm": 0.5158146619796753, "learning_rate": 2.8190324466442798e-05, "loss": 0.0273, "step": 66150 }, { "grad_norm": 0.5226659774780273, "learning_rate": 2.8175446863332965e-05, "loss": 0.0448, "step": 66160 }, { "grad_norm": 0.6075853109359741, "learning_rate": 2.816057164692345e-05, "loss": 0.0223, "step": 66170 }, { "grad_norm": 0.36359158158302307, "learning_rate": 2.8145698818840992e-05, "loss": 0.0267, "step": 66180 }, { "grad_norm": 0.4358406960964203, "learning_rate": 2.8130828380712065e-05, "loss": 0.0392, "step": 66190 }, { "grad_norm": 0.3566080331802368, "learning_rate": 2.811596033416285e-05, "loss": 0.0204, "step": 66200 }, { "grad_norm": 0.4035719931125641, "learning_rate": 2.8101094680819307e-05, "loss": 0.0236, "step": 66210 }, { "grad_norm": 0.5062714219093323, "learning_rate": 2.808623142230715e-05, "loss": 0.0168, "step": 66220 }, { "grad_norm": 0.35740602016448975, "learning_rate": 2.8071370560251757e-05, "loss": 0.0238, "step": 66230 }, { "grad_norm": 0.4305228590965271, "learning_rate": 2.805651209627831e-05, "loss": 0.0257, "step": 66240 }, { "grad_norm": 0.512042760848999, "learning_rate": 2.804165603201171e-05, "loss": 0.0244, "step": 66250 }, { "grad_norm": 0.5814520120620728, "learning_rate": 2.802680236907661e-05, "loss": 0.0245, "step": 66260 }, { "grad_norm": 0.40274927020072937, "learning_rate": 2.8011951109097344e-05, "loss": 0.033, "step": 66270 }, { "grad_norm": 0.3870392143726349, "learning_rate": 2.7997102253698072e-05, "loss": 0.0271, "step": 66280 }, { "grad_norm": 0.5150160193443298, "learning_rate": 2.7982255804502588e-05, "loss": 0.0299, "step": 66290 }, { "grad_norm": 0.541028618812561, "learning_rate": 2.7967411763134516e-05, "loss": 0.0203, "step": 66300 }, { "grad_norm": 0.33911681175231934, "learning_rate": 2.7952570131217166e-05, "loss": 0.026, "step": 66310 }, { "grad_norm": 0.35200005769729614, "learning_rate": 2.7937730910373618e-05, "loss": 0.0233, "step": 66320 }, { "grad_norm": 0.538017988204956, "learning_rate": 2.7922894102226628e-05, "loss": 0.0243, "step": 66330 }, { "grad_norm": 0.48929518461227417, "learning_rate": 2.790805970839876e-05, "loss": 0.0296, "step": 66340 }, { "grad_norm": 0.37811174988746643, "learning_rate": 2.789322773051225e-05, "loss": 0.0213, "step": 66350 }, { "grad_norm": 0.6695576310157776, "learning_rate": 2.7878398170189124e-05, "loss": 0.0325, "step": 66360 }, { "grad_norm": 0.47570398449897766, "learning_rate": 2.7863571029051112e-05, "loss": 0.0329, "step": 66370 }, { "grad_norm": 0.354780912399292, "learning_rate": 2.784874630871971e-05, "loss": 0.0257, "step": 66380 }, { "grad_norm": 0.4088517725467682, "learning_rate": 2.7833924010816086e-05, "loss": 0.0269, "step": 66390 }, { "grad_norm": 0.48507434129714966, "learning_rate": 2.7819104136961226e-05, "loss": 0.0247, "step": 66400 }, { "grad_norm": 0.5395270586013794, "learning_rate": 2.780428668877577e-05, "loss": 0.0267, "step": 66410 }, { "grad_norm": 0.5112636089324951, "learning_rate": 2.7789471667880152e-05, "loss": 0.0263, "step": 66420 }, { "grad_norm": 0.7421238422393799, "learning_rate": 2.7774659075894523e-05, "loss": 0.0355, "step": 66430 }, { "grad_norm": 0.658112108707428, "learning_rate": 2.7759848914438778e-05, "loss": 0.0195, "step": 66440 }, { "grad_norm": 0.47415977716445923, "learning_rate": 2.7745041185132496e-05, "loss": 0.0523, "step": 66450 }, { "grad_norm": 0.35149314999580383, "learning_rate": 2.773023588959506e-05, "loss": 0.0214, "step": 66460 }, { "grad_norm": 0.621269941329956, "learning_rate": 2.7715433029445555e-05, "loss": 0.0232, "step": 66470 }, { "grad_norm": 0.16126041114330292, "learning_rate": 2.7700632606302766e-05, "loss": 0.0251, "step": 66480 }, { "grad_norm": 0.6361221671104431, "learning_rate": 2.768583462178528e-05, "loss": 0.018, "step": 66490 }, { "grad_norm": 0.3833577036857605, "learning_rate": 2.767103907751136e-05, "loss": 0.0387, "step": 66500 }, { "grad_norm": 0.31863096356391907, "learning_rate": 2.7656245975099065e-05, "loss": 0.0257, "step": 66510 }, { "grad_norm": 0.38670313358306885, "learning_rate": 2.7641455316166093e-05, "loss": 0.023, "step": 66520 }, { "grad_norm": 0.5585145950317383, "learning_rate": 2.7626667102329973e-05, "loss": 0.0255, "step": 66530 }, { "grad_norm": 0.39041292667388916, "learning_rate": 2.761188133520788e-05, "loss": 0.033, "step": 66540 }, { "grad_norm": 0.24630945920944214, "learning_rate": 2.7597098016416784e-05, "loss": 0.0227, "step": 66550 }, { "grad_norm": 0.6008287072181702, "learning_rate": 2.7582317147573368e-05, "loss": 0.0291, "step": 66560 }, { "grad_norm": 0.38287580013275146, "learning_rate": 2.7567538730294064e-05, "loss": 0.0183, "step": 66570 }, { "grad_norm": 0.47288957238197327, "learning_rate": 2.7552762766194973e-05, "loss": 0.0298, "step": 66580 }, { "grad_norm": 0.6698023080825806, "learning_rate": 2.753798925689202e-05, "loss": 0.0295, "step": 66590 }, { "grad_norm": 0.30668729543685913, "learning_rate": 2.7523218204000774e-05, "loss": 0.0224, "step": 66600 }, { "grad_norm": 0.38796621561050415, "learning_rate": 2.7508449609136612e-05, "loss": 0.035, "step": 66610 }, { "grad_norm": 0.3984353542327881, "learning_rate": 2.749368347391454e-05, "loss": 0.0331, "step": 66620 }, { "grad_norm": 0.39480018615722656, "learning_rate": 2.7478919799949443e-05, "loss": 0.0248, "step": 66630 }, { "grad_norm": 0.27118808031082153, "learning_rate": 2.7464158588855803e-05, "loss": 0.0276, "step": 66640 }, { "grad_norm": 0.5705381035804749, "learning_rate": 2.7449399842247918e-05, "loss": 0.0204, "step": 66650 }, { "grad_norm": 0.500881016254425, "learning_rate": 2.7434643561739736e-05, "loss": 0.0231, "step": 66660 }, { "grad_norm": 0.5247453451156616, "learning_rate": 2.7419889748945016e-05, "loss": 0.0242, "step": 66670 }, { "grad_norm": 0.24039198458194733, "learning_rate": 2.7405138405477203e-05, "loss": 0.031, "step": 66680 }, { "grad_norm": 0.3948383033275604, "learning_rate": 2.73903895329495e-05, "loss": 0.0281, "step": 66690 }, { "grad_norm": 0.3198521137237549, "learning_rate": 2.737564313297478e-05, "loss": 0.0186, "step": 66700 }, { "grad_norm": 0.5287935733795166, "learning_rate": 2.736089920716572e-05, "loss": 0.0236, "step": 66710 }, { "grad_norm": 0.3494676351547241, "learning_rate": 2.7346157757134706e-05, "loss": 0.0195, "step": 66720 }, { "grad_norm": 0.6085984706878662, "learning_rate": 2.733141878449379e-05, "loss": 0.0373, "step": 66730 }, { "grad_norm": 0.27208641171455383, "learning_rate": 2.7316682290854844e-05, "loss": 0.0269, "step": 66740 }, { "grad_norm": 0.6900826692581177, "learning_rate": 2.7301948277829414e-05, "loss": 0.0302, "step": 66750 }, { "grad_norm": 0.6295361518859863, "learning_rate": 2.7287216747028814e-05, "loss": 0.0244, "step": 66760 }, { "grad_norm": 0.32454460859298706, "learning_rate": 2.7272487700064025e-05, "loss": 0.0239, "step": 66770 }, { "grad_norm": 0.3796382248401642, "learning_rate": 2.725776113854583e-05, "loss": 0.0235, "step": 66780 }, { "grad_norm": 0.2927260994911194, "learning_rate": 2.7243037064084664e-05, "loss": 0.0224, "step": 66790 }, { "grad_norm": 0.5563279986381531, "learning_rate": 2.7228315478290766e-05, "loss": 0.0239, "step": 66800 }, { "grad_norm": 0.3378130793571472, "learning_rate": 2.721359638277402e-05, "loss": 0.0191, "step": 66810 }, { "grad_norm": 0.5695446729660034, "learning_rate": 2.7198879779144148e-05, "loss": 0.0286, "step": 66820 }, { "grad_norm": 0.22628875076770782, "learning_rate": 2.7184165669010485e-05, "loss": 0.0224, "step": 66830 }, { "grad_norm": 0.3671002984046936, "learning_rate": 2.716945405398218e-05, "loss": 0.0206, "step": 66840 }, { "grad_norm": 0.44173985719680786, "learning_rate": 2.715474493566803e-05, "loss": 0.0278, "step": 66850 }, { "grad_norm": 0.4306660294532776, "learning_rate": 2.7140038315676652e-05, "loss": 0.0242, "step": 66860 }, { "grad_norm": 0.5228588581085205, "learning_rate": 2.7125334195616275e-05, "loss": 0.0276, "step": 66870 }, { "grad_norm": 0.41024133563041687, "learning_rate": 2.7110632577094997e-05, "loss": 0.0227, "step": 66880 }, { "grad_norm": 0.3294629752635956, "learning_rate": 2.709593346172051e-05, "loss": 0.0158, "step": 66890 }, { "grad_norm": 0.42819371819496155, "learning_rate": 2.708123685110032e-05, "loss": 0.0251, "step": 66900 }, { "grad_norm": 0.45121070742607117, "learning_rate": 2.7066542746841588e-05, "loss": 0.0205, "step": 66910 }, { "grad_norm": 0.22495482861995697, "learning_rate": 2.7051851150551267e-05, "loss": 0.026, "step": 66920 }, { "grad_norm": 0.35543957352638245, "learning_rate": 2.7037162063835996e-05, "loss": 0.0209, "step": 66930 }, { "grad_norm": 0.5427163243293762, "learning_rate": 2.702247548830218e-05, "loss": 0.0186, "step": 66940 }, { "grad_norm": 0.5824455618858337, "learning_rate": 2.7007791425555883e-05, "loss": 0.03, "step": 66950 }, { "grad_norm": 0.4792860746383667, "learning_rate": 2.6993109877202945e-05, "loss": 0.0281, "step": 66960 }, { "grad_norm": 0.5645502209663391, "learning_rate": 2.6978430844848935e-05, "loss": 0.0195, "step": 66970 }, { "grad_norm": 0.4354967772960663, "learning_rate": 2.69637543300991e-05, "loss": 0.0244, "step": 66980 }, { "grad_norm": 0.5516055822372437, "learning_rate": 2.694908033455848e-05, "loss": 0.025, "step": 66990 }, { "grad_norm": 0.4083826243877411, "learning_rate": 2.6934408859831733e-05, "loss": 0.0228, "step": 67000 }, { "grad_norm": 0.5515359044075012, "learning_rate": 2.69197399075234e-05, "loss": 0.0255, "step": 67010 }, { "grad_norm": 0.36372461915016174, "learning_rate": 2.6905073479237584e-05, "loss": 0.0254, "step": 67020 }, { "grad_norm": 1.0461485385894775, "learning_rate": 2.6890409576578235e-05, "loss": 0.024, "step": 67030 }, { "grad_norm": 0.27385687828063965, "learning_rate": 2.687574820114892e-05, "loss": 0.0251, "step": 67040 }, { "grad_norm": 0.38399097323417664, "learning_rate": 2.686108935455305e-05, "loss": 0.0268, "step": 67050 }, { "grad_norm": 0.4159828722476959, "learning_rate": 2.6846433038393603e-05, "loss": 0.0241, "step": 67060 }, { "grad_norm": 0.4873821437358856, "learning_rate": 2.6831779254273482e-05, "loss": 0.0226, "step": 67070 }, { "grad_norm": 0.522074818611145, "learning_rate": 2.6817128003795122e-05, "loss": 0.0249, "step": 67080 }, { "grad_norm": 0.9909592270851135, "learning_rate": 2.680247928856081e-05, "loss": 0.0304, "step": 67090 }, { "grad_norm": 0.4839909076690674, "learning_rate": 2.6787833110172466e-05, "loss": 0.022, "step": 67100 }, { "grad_norm": 1.5140395164489746, "learning_rate": 2.6773189470231807e-05, "loss": 0.0259, "step": 67110 }, { "grad_norm": 0.30012840032577515, "learning_rate": 2.675854837034022e-05, "loss": 0.0199, "step": 67120 }, { "grad_norm": 0.5537835955619812, "learning_rate": 2.6743909812098833e-05, "loss": 0.0267, "step": 67130 }, { "grad_norm": 0.4363349974155426, "learning_rate": 2.67292737971085e-05, "loss": 0.0212, "step": 67140 }, { "grad_norm": 0.40983062982559204, "learning_rate": 2.671464032696982e-05, "loss": 0.0295, "step": 67150 }, { "grad_norm": 0.1986047327518463, "learning_rate": 2.6700009403283055e-05, "loss": 0.0118, "step": 67160 }, { "grad_norm": 0.25854694843292236, "learning_rate": 2.6685381027648214e-05, "loss": 0.0278, "step": 67170 }, { "grad_norm": 0.541623592376709, "learning_rate": 2.6670755201665086e-05, "loss": 0.0283, "step": 67180 }, { "grad_norm": 0.36467689275741577, "learning_rate": 2.6656131926933066e-05, "loss": 0.0258, "step": 67190 }, { "grad_norm": 0.42160922288894653, "learning_rate": 2.664151120505136e-05, "loss": 0.0229, "step": 67200 }, { "grad_norm": 0.39050987362861633, "learning_rate": 2.6626893037618873e-05, "loss": 0.0159, "step": 67210 }, { "grad_norm": 0.4408697485923767, "learning_rate": 2.6612277426234243e-05, "loss": 0.0226, "step": 67220 }, { "grad_norm": 0.37930458784103394, "learning_rate": 2.659766437249577e-05, "loss": 0.0342, "step": 67230 }, { "grad_norm": 0.5695258378982544, "learning_rate": 2.6583053878001557e-05, "loss": 0.0259, "step": 67240 }, { "grad_norm": 0.2339983433485031, "learning_rate": 2.6568445944349323e-05, "loss": 0.0333, "step": 67250 }, { "grad_norm": 0.4337897300720215, "learning_rate": 2.6553840573136657e-05, "loss": 0.0188, "step": 67260 }, { "grad_norm": 0.3236696422100067, "learning_rate": 2.6539237765960706e-05, "loss": 0.0293, "step": 67270 }, { "grad_norm": 0.7024829387664795, "learning_rate": 2.6524637524418465e-05, "loss": 0.0245, "step": 67280 }, { "grad_norm": 0.4355701804161072, "learning_rate": 2.651003985010655e-05, "loss": 0.0236, "step": 67290 }, { "grad_norm": 0.3435835838317871, "learning_rate": 2.649544474462138e-05, "loss": 0.0197, "step": 67300 }, { "grad_norm": 0.25859883427619934, "learning_rate": 2.6480852209559015e-05, "loss": 0.0285, "step": 67310 }, { "grad_norm": 0.41850781440734863, "learning_rate": 2.6466262246515296e-05, "loss": 0.0205, "step": 67320 }, { "grad_norm": 0.47535866498947144, "learning_rate": 2.6451674857085746e-05, "loss": 0.0322, "step": 67330 }, { "grad_norm": 0.5479402542114258, "learning_rate": 2.6437090042865655e-05, "loss": 0.0385, "step": 67340 }, { "grad_norm": 0.3310595452785492, "learning_rate": 2.6422507805449947e-05, "loss": 0.027, "step": 67350 }, { "grad_norm": 0.34457337856292725, "learning_rate": 2.6407928146433357e-05, "loss": 0.0212, "step": 67360 }, { "grad_norm": 0.41556817293167114, "learning_rate": 2.639335106741026e-05, "loss": 0.0225, "step": 67370 }, { "grad_norm": 0.3617505133152008, "learning_rate": 2.6378776569974796e-05, "loss": 0.017, "step": 67380 }, { "grad_norm": 0.533911943435669, "learning_rate": 2.6364204655720814e-05, "loss": 0.0186, "step": 67390 }, { "grad_norm": 0.3610978126525879, "learning_rate": 2.6349635326241894e-05, "loss": 0.0198, "step": 67400 }, { "grad_norm": 0.5083907842636108, "learning_rate": 2.633506858313127e-05, "loss": 0.0273, "step": 67410 }, { "grad_norm": 0.3999113142490387, "learning_rate": 2.6320504427981975e-05, "loss": 0.0178, "step": 67420 }, { "grad_norm": 0.22228966653347015, "learning_rate": 2.6305942862386734e-05, "loss": 0.0222, "step": 67430 }, { "grad_norm": 0.4205690920352936, "learning_rate": 2.629138388793793e-05, "loss": 0.0183, "step": 67440 }, { "grad_norm": 0.2562113404273987, "learning_rate": 2.6276827506227743e-05, "loss": 0.0265, "step": 67450 }, { "grad_norm": 0.31701794266700745, "learning_rate": 2.6262273718848037e-05, "loss": 0.0274, "step": 67460 }, { "grad_norm": 0.3849849998950958, "learning_rate": 2.6247722527390407e-05, "loss": 0.0168, "step": 67470 }, { "grad_norm": 0.20177993178367615, "learning_rate": 2.6233173933446104e-05, "loss": 0.0265, "step": 67480 }, { "grad_norm": 0.502933144569397, "learning_rate": 2.6218627938606188e-05, "loss": 0.0325, "step": 67490 }, { "grad_norm": 0.4331800043582916, "learning_rate": 2.620408454446135e-05, "loss": 0.0205, "step": 67500 }, { "grad_norm": 0.3458752930164337, "learning_rate": 2.6189543752602046e-05, "loss": 0.0238, "step": 67510 }, { "grad_norm": 0.4196033477783203, "learning_rate": 2.6175005564618444e-05, "loss": 0.029, "step": 67520 }, { "grad_norm": 0.43651044368743896, "learning_rate": 2.6160469982100428e-05, "loss": 0.0243, "step": 67530 }, { "grad_norm": 0.3564620316028595, "learning_rate": 2.6145937006637555e-05, "loss": 0.0257, "step": 67540 }, { "grad_norm": 0.6162822842597961, "learning_rate": 2.6131406639819166e-05, "loss": 0.0194, "step": 67550 }, { "grad_norm": 0.37800663709640503, "learning_rate": 2.611687888323424e-05, "loss": 0.0306, "step": 67560 }, { "grad_norm": 0.48112955689430237, "learning_rate": 2.610235373847153e-05, "loss": 0.0236, "step": 67570 }, { "grad_norm": 0.32472193241119385, "learning_rate": 2.6087831207119494e-05, "loss": 0.0285, "step": 67580 }, { "grad_norm": 0.44891273975372314, "learning_rate": 2.60733112907663e-05, "loss": 0.0233, "step": 67590 }, { "grad_norm": 0.5092498064041138, "learning_rate": 2.6058793990999786e-05, "loss": 0.0247, "step": 67600 }, { "grad_norm": 0.380170613527298, "learning_rate": 2.6044279309407593e-05, "loss": 0.0229, "step": 67610 }, { "grad_norm": 0.3140343129634857, "learning_rate": 2.6029767247576976e-05, "loss": 0.0266, "step": 67620 }, { "grad_norm": 0.41703900694847107, "learning_rate": 2.601525780709497e-05, "loss": 0.0194, "step": 67630 }, { "grad_norm": 0.2222364991903305, "learning_rate": 2.6000750989548318e-05, "loss": 0.0214, "step": 67640 }, { "grad_norm": 0.4352421462535858, "learning_rate": 2.5986246796523462e-05, "loss": 0.0282, "step": 67650 }, { "grad_norm": 0.3227146863937378, "learning_rate": 2.5971745229606546e-05, "loss": 0.0201, "step": 67660 }, { "grad_norm": 0.397629052400589, "learning_rate": 2.5957246290383442e-05, "loss": 0.0267, "step": 67670 }, { "grad_norm": 0.43089327216148376, "learning_rate": 2.594274998043975e-05, "loss": 0.0219, "step": 67680 }, { "grad_norm": 0.5647180080413818, "learning_rate": 2.5928256301360733e-05, "loss": 0.0266, "step": 67690 }, { "grad_norm": 0.420197993516922, "learning_rate": 2.591376525473142e-05, "loss": 0.0188, "step": 67700 }, { "grad_norm": 0.48464247584342957, "learning_rate": 2.5899276842136523e-05, "loss": 0.0252, "step": 67710 }, { "grad_norm": 0.5016433000564575, "learning_rate": 2.5884791065160495e-05, "loss": 0.0201, "step": 67720 }, { "grad_norm": 0.42271536588668823, "learning_rate": 2.5870307925387436e-05, "loss": 0.0195, "step": 67730 }, { "grad_norm": 0.19042280316352844, "learning_rate": 2.585582742440125e-05, "loss": 0.0191, "step": 67740 }, { "grad_norm": 0.4235738217830658, "learning_rate": 2.5841349563785455e-05, "loss": 0.0249, "step": 67750 }, { "grad_norm": 0.39210033416748047, "learning_rate": 2.5826874345123355e-05, "loss": 0.0246, "step": 67760 }, { "grad_norm": 0.39193636178970337, "learning_rate": 2.581240176999792e-05, "loss": 0.0173, "step": 67770 }, { "grad_norm": 0.3816409409046173, "learning_rate": 2.579793183999189e-05, "loss": 0.0324, "step": 67780 }, { "grad_norm": 0.52019202709198, "learning_rate": 2.5783464556687627e-05, "loss": 0.0241, "step": 67790 }, { "grad_norm": 0.24954752624034882, "learning_rate": 2.576899992166729e-05, "loss": 0.0153, "step": 67800 }, { "grad_norm": 0.5065709352493286, "learning_rate": 2.5754537936512668e-05, "loss": 0.0253, "step": 67810 }, { "grad_norm": 0.257242888212204, "learning_rate": 2.574007860280535e-05, "loss": 0.0313, "step": 67820 }, { "grad_norm": 0.3953017592430115, "learning_rate": 2.5725621922126518e-05, "loss": 0.046, "step": 67830 }, { "grad_norm": 0.45763054490089417, "learning_rate": 2.571116789605722e-05, "loss": 0.0363, "step": 67840 }, { "grad_norm": 0.37800517678260803, "learning_rate": 2.5696716526178054e-05, "loss": 0.0258, "step": 67850 }, { "grad_norm": 0.3363422155380249, "learning_rate": 2.568226781406946e-05, "loss": 0.0192, "step": 67860 }, { "grad_norm": 0.4071848392486572, "learning_rate": 2.5667821761311474e-05, "loss": 0.0176, "step": 67870 }, { "grad_norm": 0.45148926973342896, "learning_rate": 2.5653378369483917e-05, "loss": 0.017, "step": 67880 }, { "grad_norm": 0.3050122857093811, "learning_rate": 2.5638937640166292e-05, "loss": 0.0508, "step": 67890 }, { "grad_norm": 0.2871628403663635, "learning_rate": 2.5624499574937842e-05, "loss": 0.0246, "step": 67900 }, { "grad_norm": 0.2998866140842438, "learning_rate": 2.5610064175377456e-05, "loss": 0.0234, "step": 67910 }, { "grad_norm": 0.3933001160621643, "learning_rate": 2.5595631443063782e-05, "loss": 0.0204, "step": 67920 }, { "grad_norm": 0.2848595678806305, "learning_rate": 2.5581201379575177e-05, "loss": 0.0176, "step": 67930 }, { "grad_norm": 0.6409688591957092, "learning_rate": 2.5566773986489655e-05, "loss": 0.0242, "step": 67940 }, { "grad_norm": 0.2532139718532562, "learning_rate": 2.5552349265385e-05, "loss": 0.0166, "step": 67950 }, { "grad_norm": 0.36850252747535706, "learning_rate": 2.5537927217838675e-05, "loss": 0.0268, "step": 67960 }, { "grad_norm": 0.22659942507743835, "learning_rate": 2.5523507845427864e-05, "loss": 0.013, "step": 67970 }, { "grad_norm": 0.6347734332084656, "learning_rate": 2.5509091149729413e-05, "loss": 0.0238, "step": 67980 }, { "grad_norm": 0.5045014023780823, "learning_rate": 2.5494677132319962e-05, "loss": 0.0248, "step": 67990 }, { "grad_norm": 0.2504841983318329, "learning_rate": 2.548026579477575e-05, "loss": 0.028, "step": 68000 }, { "grad_norm": 0.47650566697120667, "learning_rate": 2.5465857138672834e-05, "loss": 0.0293, "step": 68010 }, { "grad_norm": 0.7729665637016296, "learning_rate": 2.5451451165586848e-05, "loss": 0.0366, "step": 68020 }, { "grad_norm": 0.5621621608734131, "learning_rate": 2.543704787709329e-05, "loss": 0.02, "step": 68030 }, { "grad_norm": 0.581582248210907, "learning_rate": 2.542264727476723e-05, "loss": 0.0261, "step": 68040 }, { "grad_norm": 0.21955108642578125, "learning_rate": 2.5408249360183535e-05, "loss": 0.0267, "step": 68050 }, { "grad_norm": 0.3723139464855194, "learning_rate": 2.5393854134916694e-05, "loss": 0.0204, "step": 68060 }, { "grad_norm": 0.38595283031463623, "learning_rate": 2.5379461600540993e-05, "loss": 0.0276, "step": 68070 }, { "grad_norm": 0.35972854495048523, "learning_rate": 2.5365071758630317e-05, "loss": 0.0357, "step": 68080 }, { "grad_norm": 0.2608022689819336, "learning_rate": 2.535068461075839e-05, "loss": 0.0292, "step": 68090 }, { "grad_norm": 0.5572091341018677, "learning_rate": 2.5336300158498516e-05, "loss": 0.0239, "step": 68100 }, { "grad_norm": 0.3299266993999481, "learning_rate": 2.5321918403423793e-05, "loss": 0.0172, "step": 68110 }, { "grad_norm": 0.5243204832077026, "learning_rate": 2.5307539347106957e-05, "loss": 0.0258, "step": 68120 }, { "grad_norm": 0.6781449913978577, "learning_rate": 2.529316299112049e-05, "loss": 0.0253, "step": 68130 }, { "grad_norm": 0.3526323139667511, "learning_rate": 2.5278789337036567e-05, "loss": 0.0262, "step": 68140 }, { "grad_norm": 0.4997127652168274, "learning_rate": 2.52644183864271e-05, "loss": 0.0206, "step": 68150 }, { "grad_norm": 0.798185408115387, "learning_rate": 2.525005014086363e-05, "loss": 0.0262, "step": 68160 }, { "grad_norm": 0.3958450257778168, "learning_rate": 2.5235684601917465e-05, "loss": 0.0233, "step": 68170 }, { "grad_norm": 0.583572268486023, "learning_rate": 2.5221321771159618e-05, "loss": 0.0126, "step": 68180 }, { "grad_norm": 0.39029771089553833, "learning_rate": 2.5206961650160743e-05, "loss": 0.0238, "step": 68190 }, { "grad_norm": 0.478127121925354, "learning_rate": 2.519260424049129e-05, "loss": 0.0343, "step": 68200 }, { "grad_norm": 0.7032718062400818, "learning_rate": 2.5178249543721304e-05, "loss": 0.0241, "step": 68210 }, { "grad_norm": 0.4421432614326477, "learning_rate": 2.5163897561420668e-05, "loss": 0.0304, "step": 68220 }, { "grad_norm": 0.37795311212539673, "learning_rate": 2.514954829515883e-05, "loss": 0.0204, "step": 68230 }, { "grad_norm": 0.3658657968044281, "learning_rate": 2.513520174650505e-05, "loss": 0.0235, "step": 68240 }, { "grad_norm": 0.25433459877967834, "learning_rate": 2.5120857917028207e-05, "loss": 0.0241, "step": 68250 }, { "grad_norm": 0.33335602283477783, "learning_rate": 2.5106516808296953e-05, "loss": 0.0292, "step": 68260 }, { "grad_norm": 1.2532718181610107, "learning_rate": 2.5092178421879552e-05, "loss": 0.018, "step": 68270 }, { "grad_norm": 0.6670064330101013, "learning_rate": 2.5077842759344116e-05, "loss": 0.018, "step": 68280 }, { "grad_norm": 0.4212011396884918, "learning_rate": 2.5063509822258304e-05, "loss": 0.0258, "step": 68290 }, { "grad_norm": 0.326571524143219, "learning_rate": 2.5049179612189576e-05, "loss": 0.0255, "step": 68300 }, { "grad_norm": 0.36757713556289673, "learning_rate": 2.5034852130705043e-05, "loss": 0.0179, "step": 68310 }, { "grad_norm": 0.9323000311851501, "learning_rate": 2.5020527379371572e-05, "loss": 0.025, "step": 68320 }, { "grad_norm": 0.6039148569107056, "learning_rate": 2.500620535975564e-05, "loss": 0.0366, "step": 68330 }, { "grad_norm": 0.4638601839542389, "learning_rate": 2.4991886073423515e-05, "loss": 0.0216, "step": 68340 }, { "grad_norm": 0.44695672392845154, "learning_rate": 2.4977569521941134e-05, "loss": 0.0241, "step": 68350 }, { "grad_norm": 0.40048307180404663, "learning_rate": 2.4963255706874143e-05, "loss": 0.0243, "step": 68360 }, { "grad_norm": 0.3967645764350891, "learning_rate": 2.4948944629787853e-05, "loss": 0.0209, "step": 68370 }, { "grad_norm": 0.43391692638397217, "learning_rate": 2.493463629224731e-05, "loss": 0.0284, "step": 68380 }, { "grad_norm": 0.30379870533943176, "learning_rate": 2.492033069581728e-05, "loss": 0.0274, "step": 68390 }, { "grad_norm": 0.2227012813091278, "learning_rate": 2.4906027842062137e-05, "loss": 0.023, "step": 68400 }, { "grad_norm": 0.4998044967651367, "learning_rate": 2.48917277325461e-05, "loss": 0.0212, "step": 68410 }, { "grad_norm": 0.3258229196071625, "learning_rate": 2.487743036883295e-05, "loss": 0.0256, "step": 68420 }, { "grad_norm": 0.34583285450935364, "learning_rate": 2.486313575248626e-05, "loss": 0.0315, "step": 68430 }, { "grad_norm": 0.5066388845443726, "learning_rate": 2.4848843885069235e-05, "loss": 0.021, "step": 68440 }, { "grad_norm": 0.20499898493289948, "learning_rate": 2.483455476814484e-05, "loss": 0.0205, "step": 68450 }, { "grad_norm": 0.6953325271606445, "learning_rate": 2.4820268403275672e-05, "loss": 0.0228, "step": 68460 }, { "grad_norm": 0.8642200231552124, "learning_rate": 2.480598479202413e-05, "loss": 0.0269, "step": 68470 }, { "grad_norm": 0.19716419279575348, "learning_rate": 2.4791703935952193e-05, "loss": 0.0157, "step": 68480 }, { "grad_norm": 1.197429895401001, "learning_rate": 2.477742583662163e-05, "loss": 0.0257, "step": 68490 }, { "grad_norm": 0.37651535868644714, "learning_rate": 2.476315049559384e-05, "loss": 0.0233, "step": 68500 }, { "grad_norm": 0.44858163595199585, "learning_rate": 2.4748877914429987e-05, "loss": 0.0286, "step": 68510 }, { "grad_norm": 0.4148045480251312, "learning_rate": 2.4734608094690857e-05, "loss": 0.0239, "step": 68520 }, { "grad_norm": 0.7222382426261902, "learning_rate": 2.4720341037937006e-05, "loss": 0.0224, "step": 68530 }, { "grad_norm": 0.7221609354019165, "learning_rate": 2.4706076745728645e-05, "loss": 0.0406, "step": 68540 }, { "grad_norm": 0.4250229299068451, "learning_rate": 2.469181521962573e-05, "loss": 0.0327, "step": 68550 }, { "grad_norm": 0.42463964223861694, "learning_rate": 2.467755646118783e-05, "loss": 0.0215, "step": 68560 }, { "grad_norm": 0.39315125346183777, "learning_rate": 2.46633004719743e-05, "loss": 0.021, "step": 68570 }, { "grad_norm": 0.9305281043052673, "learning_rate": 2.4649047253544123e-05, "loss": 0.0272, "step": 68580 }, { "grad_norm": 0.5642587542533875, "learning_rate": 2.4634796807456016e-05, "loss": 0.024, "step": 68590 }, { "grad_norm": 0.25105616450309753, "learning_rate": 2.46205491352684e-05, "loss": 0.0197, "step": 68600 }, { "grad_norm": 0.7545715570449829, "learning_rate": 2.4606304238539375e-05, "loss": 0.0291, "step": 68610 }, { "grad_norm": 0.4039408266544342, "learning_rate": 2.4592062118826753e-05, "loss": 0.0253, "step": 68620 }, { "grad_norm": 0.569352388381958, "learning_rate": 2.4577822777688004e-05, "loss": 0.0207, "step": 68630 }, { "grad_norm": 0.5189698934555054, "learning_rate": 2.4563586216680347e-05, "loss": 0.0223, "step": 68640 }, { "grad_norm": 0.3333075940608978, "learning_rate": 2.4549352437360622e-05, "loss": 0.0211, "step": 68650 }, { "grad_norm": 0.5367621779441833, "learning_rate": 2.4535121441285493e-05, "loss": 0.0243, "step": 68660 }, { "grad_norm": 0.8784202337265015, "learning_rate": 2.4520893230011172e-05, "loss": 0.025, "step": 68670 }, { "grad_norm": 0.5475271344184875, "learning_rate": 2.450666780509368e-05, "loss": 0.0177, "step": 68680 }, { "grad_norm": 0.3060740828514099, "learning_rate": 2.4492445168088656e-05, "loss": 0.0286, "step": 68690 }, { "grad_norm": 1.0726749897003174, "learning_rate": 2.447822532055149e-05, "loss": 0.0303, "step": 68700 }, { "grad_norm": 0.5808847546577454, "learning_rate": 2.4464008264037212e-05, "loss": 0.023, "step": 68710 }, { "grad_norm": 0.5304205417633057, "learning_rate": 2.4449794000100605e-05, "loss": 0.0254, "step": 68720 }, { "grad_norm": 0.5100454092025757, "learning_rate": 2.4435582530296108e-05, "loss": 0.0289, "step": 68730 }, { "grad_norm": 0.6382429599761963, "learning_rate": 2.4421373856177887e-05, "loss": 0.0316, "step": 68740 }, { "grad_norm": 0.5256118178367615, "learning_rate": 2.440716797929975e-05, "loss": 0.0173, "step": 68750 }, { "grad_norm": 0.5362771153450012, "learning_rate": 2.439296490121526e-05, "loss": 0.0281, "step": 68760 }, { "grad_norm": 0.42966529726982117, "learning_rate": 2.4378764623477608e-05, "loss": 0.0226, "step": 68770 }, { "grad_norm": 0.44487228989601135, "learning_rate": 2.436456714763974e-05, "loss": 0.0301, "step": 68780 }, { "grad_norm": 0.2261771410703659, "learning_rate": 2.4350372475254262e-05, "loss": 0.0218, "step": 68790 }, { "grad_norm": 0.476811021566391, "learning_rate": 2.433618060787351e-05, "loss": 0.0292, "step": 68800 }, { "grad_norm": 0.5519216656684875, "learning_rate": 2.432199154704944e-05, "loss": 0.0176, "step": 68810 }, { "grad_norm": 0.4283830523490906, "learning_rate": 2.430780529433377e-05, "loss": 0.0345, "step": 68820 }, { "grad_norm": 0.36968448758125305, "learning_rate": 2.42936218512779e-05, "loss": 0.0264, "step": 68830 }, { "grad_norm": 0.2676880359649658, "learning_rate": 2.427944121943288e-05, "loss": 0.0206, "step": 68840 }, { "grad_norm": 0.553098201751709, "learning_rate": 2.4265263400349497e-05, "loss": 0.0205, "step": 68850 }, { "grad_norm": 0.5524102449417114, "learning_rate": 2.4251088395578213e-05, "loss": 0.0279, "step": 68860 }, { "grad_norm": 0.48515060544013977, "learning_rate": 2.4236916206669214e-05, "loss": 0.0264, "step": 68870 }, { "grad_norm": 0.29564788937568665, "learning_rate": 2.4222746835172304e-05, "loss": 0.0286, "step": 68880 }, { "grad_norm": 0.6834545731544495, "learning_rate": 2.4208580282637065e-05, "loss": 0.0393, "step": 68890 }, { "grad_norm": 0.60982346534729, "learning_rate": 2.4194416550612687e-05, "loss": 0.0286, "step": 68900 }, { "grad_norm": 0.3779238164424896, "learning_rate": 2.418025564064812e-05, "loss": 0.0225, "step": 68910 }, { "grad_norm": 0.38759085536003113, "learning_rate": 2.4166097554291978e-05, "loss": 0.0232, "step": 68920 }, { "grad_norm": 0.2942591607570648, "learning_rate": 2.4151942293092583e-05, "loss": 0.0358, "step": 68930 }, { "grad_norm": 0.411525696516037, "learning_rate": 2.41377898585979e-05, "loss": 0.031, "step": 68940 }, { "grad_norm": 0.6456733345985413, "learning_rate": 2.4123640252355652e-05, "loss": 0.02, "step": 68950 }, { "grad_norm": 0.5956595540046692, "learning_rate": 2.410949347591318e-05, "loss": 0.0256, "step": 68960 }, { "grad_norm": 0.20225246250629425, "learning_rate": 2.4095349530817578e-05, "loss": 0.0329, "step": 68970 }, { "grad_norm": 0.24034827947616577, "learning_rate": 2.4081208418615603e-05, "loss": 0.0229, "step": 68980 }, { "grad_norm": 0.8847797513008118, "learning_rate": 2.4067070140853725e-05, "loss": 0.0277, "step": 68990 }, { "grad_norm": 0.26621297001838684, "learning_rate": 2.405293469907805e-05, "loss": 0.0202, "step": 69000 }, { "grad_norm": 0.352205753326416, "learning_rate": 2.4038802094834446e-05, "loss": 0.0273, "step": 69010 }, { "grad_norm": 0.42567238211631775, "learning_rate": 2.4024672329668395e-05, "loss": 0.0216, "step": 69020 }, { "grad_norm": 0.42084628343582153, "learning_rate": 2.4010545405125125e-05, "loss": 0.0226, "step": 69030 }, { "grad_norm": 0.3501361608505249, "learning_rate": 2.3996421322749528e-05, "loss": 0.0204, "step": 69040 }, { "grad_norm": 0.3863677680492401, "learning_rate": 2.3982300084086224e-05, "loss": 0.0322, "step": 69050 }, { "grad_norm": 0.5369051098823547, "learning_rate": 2.396818169067944e-05, "loss": 0.032, "step": 69060 }, { "grad_norm": 0.6516359448432922, "learning_rate": 2.3954066144073178e-05, "loss": 0.0257, "step": 69070 }, { "grad_norm": 0.6012457013130188, "learning_rate": 2.39399534458111e-05, "loss": 0.0191, "step": 69080 }, { "grad_norm": 0.33933958411216736, "learning_rate": 2.392584359743651e-05, "loss": 0.0184, "step": 69090 }, { "grad_norm": 0.4029417932033539, "learning_rate": 2.3911736600492463e-05, "loss": 0.0171, "step": 69100 }, { "grad_norm": 0.25526344776153564, "learning_rate": 2.3897632456521672e-05, "loss": 0.0182, "step": 69110 }, { "grad_norm": 0.3959799110889435, "learning_rate": 2.388353116706658e-05, "loss": 0.0175, "step": 69120 }, { "grad_norm": 0.3941192030906677, "learning_rate": 2.3869432733669232e-05, "loss": 0.0252, "step": 69130 }, { "grad_norm": 0.329979807138443, "learning_rate": 2.385533715787145e-05, "loss": 0.0236, "step": 69140 }, { "grad_norm": 0.5348583459854126, "learning_rate": 2.3841244441214673e-05, "loss": 0.0298, "step": 69150 }, { "grad_norm": 0.33176615834236145, "learning_rate": 2.3827154585240073e-05, "loss": 0.0227, "step": 69160 }, { "grad_norm": 0.3021968901157379, "learning_rate": 2.3813067591488498e-05, "loss": 0.0291, "step": 69170 }, { "grad_norm": 0.3091140389442444, "learning_rate": 2.3798983461500494e-05, "loss": 0.0235, "step": 69180 }, { "grad_norm": 0.2313879430294037, "learning_rate": 2.3784902196816256e-05, "loss": 0.034, "step": 69190 }, { "grad_norm": 0.4676337242126465, "learning_rate": 2.3770823798975716e-05, "loss": 0.0251, "step": 69200 }, { "grad_norm": 0.4540325105190277, "learning_rate": 2.3756748269518424e-05, "loss": 0.0192, "step": 69210 }, { "grad_norm": 0.37900102138519287, "learning_rate": 2.374267560998371e-05, "loss": 0.0275, "step": 69220 }, { "grad_norm": 0.21951895952224731, "learning_rate": 2.3728605821910477e-05, "loss": 0.0154, "step": 69230 }, { "grad_norm": 0.41060882806777954, "learning_rate": 2.3714538906837453e-05, "loss": 0.0143, "step": 69240 }, { "grad_norm": 0.6800640821456909, "learning_rate": 2.3700474866302913e-05, "loss": 0.0177, "step": 69250 }, { "grad_norm": 0.42867010831832886, "learning_rate": 2.3686413701844917e-05, "loss": 0.0223, "step": 69260 }, { "grad_norm": 0.9993570446968079, "learning_rate": 2.367235541500114e-05, "loss": 0.0279, "step": 69270 }, { "grad_norm": 0.5383478403091431, "learning_rate": 2.3658300007308993e-05, "loss": 0.0296, "step": 69280 }, { "grad_norm": 0.39351680874824524, "learning_rate": 2.364424748030555e-05, "loss": 0.0298, "step": 69290 }, { "grad_norm": 0.3553217351436615, "learning_rate": 2.3630197835527595e-05, "loss": 0.023, "step": 69300 }, { "grad_norm": 0.5240062475204468, "learning_rate": 2.3616151074511546e-05, "loss": 0.0227, "step": 69310 }, { "grad_norm": 0.35089272260665894, "learning_rate": 2.3602107198793545e-05, "loss": 0.0249, "step": 69320 }, { "grad_norm": 0.3555484116077423, "learning_rate": 2.3588066209909427e-05, "loss": 0.0214, "step": 69330 }, { "grad_norm": 0.240113765001297, "learning_rate": 2.3574028109394662e-05, "loss": 0.0229, "step": 69340 }, { "grad_norm": 0.3803713619709015, "learning_rate": 2.3559992898784454e-05, "loss": 0.0277, "step": 69350 }, { "grad_norm": 0.28212830424308777, "learning_rate": 2.354596057961367e-05, "loss": 0.021, "step": 69360 }, { "grad_norm": 0.3426979184150696, "learning_rate": 2.353193115341688e-05, "loss": 0.0265, "step": 69370 }, { "grad_norm": 0.14370505511760712, "learning_rate": 2.3517904621728294e-05, "loss": 0.0245, "step": 69380 }, { "grad_norm": 0.7838943600654602, "learning_rate": 2.3503880986081855e-05, "loss": 0.0303, "step": 69390 }, { "grad_norm": 0.25916871428489685, "learning_rate": 2.3489860248011136e-05, "loss": 0.0243, "step": 69400 }, { "grad_norm": 0.33983680605888367, "learning_rate": 2.3475842409049464e-05, "loss": 0.0226, "step": 69410 }, { "grad_norm": 0.33706411719322205, "learning_rate": 2.346182747072975e-05, "loss": 0.0248, "step": 69420 }, { "grad_norm": 0.44644078612327576, "learning_rate": 2.3447815434584718e-05, "loss": 0.0382, "step": 69430 }, { "grad_norm": 0.41349565982818604, "learning_rate": 2.3433806302146656e-05, "loss": 0.0271, "step": 69440 }, { "grad_norm": 0.326241672039032, "learning_rate": 2.3419800074947612e-05, "loss": 0.0181, "step": 69450 }, { "grad_norm": 0.5128379464149475, "learning_rate": 2.3405796754519248e-05, "loss": 0.0191, "step": 69460 }, { "grad_norm": 0.38693150877952576, "learning_rate": 2.3391796342392992e-05, "loss": 0.0208, "step": 69470 }, { "grad_norm": 0.4073764383792877, "learning_rate": 2.337779884009984e-05, "loss": 0.0186, "step": 69480 }, { "grad_norm": 0.6909039616584778, "learning_rate": 2.3363804249170624e-05, "loss": 0.0197, "step": 69490 }, { "grad_norm": 0.29325783252716064, "learning_rate": 2.3349812571135703e-05, "loss": 0.0209, "step": 69500 }, { "grad_norm": 0.44141116738319397, "learning_rate": 2.333582380752523e-05, "loss": 0.033, "step": 69510 }, { "grad_norm": 0.37834662199020386, "learning_rate": 2.3321837959868963e-05, "loss": 0.0214, "step": 69520 }, { "grad_norm": 0.37838485836982727, "learning_rate": 2.3307855029696383e-05, "loss": 0.0224, "step": 69530 }, { "grad_norm": 0.3183906674385071, "learning_rate": 2.3293875018536648e-05, "loss": 0.0233, "step": 69540 }, { "grad_norm": 0.23630738258361816, "learning_rate": 2.327989792791861e-05, "loss": 0.0478, "step": 69550 }, { "grad_norm": 0.184090718626976, "learning_rate": 2.326592375937074e-05, "loss": 0.0273, "step": 69560 }, { "grad_norm": 0.28182029724121094, "learning_rate": 2.3251952514421248e-05, "loss": 0.022, "step": 69570 }, { "grad_norm": 0.5511471629142761, "learning_rate": 2.323798419459804e-05, "loss": 0.0299, "step": 69580 }, { "grad_norm": 0.23548859357833862, "learning_rate": 2.3224018801428625e-05, "loss": 0.02, "step": 69590 }, { "grad_norm": 0.2742275893688202, "learning_rate": 2.3210056336440273e-05, "loss": 0.0183, "step": 69600 }, { "grad_norm": 0.34129858016967773, "learning_rate": 2.3196096801159844e-05, "loss": 0.0154, "step": 69610 }, { "grad_norm": 0.253487229347229, "learning_rate": 2.3182140197114012e-05, "loss": 0.0205, "step": 69620 }, { "grad_norm": 0.3438258171081543, "learning_rate": 2.3168186525828985e-05, "loss": 0.0254, "step": 69630 }, { "grad_norm": 0.4118873178958893, "learning_rate": 2.3154235788830757e-05, "loss": 0.0165, "step": 69640 }, { "grad_norm": 0.6893437504768372, "learning_rate": 2.3140287987644927e-05, "loss": 0.0256, "step": 69650 }, { "grad_norm": 0.4711707532405853, "learning_rate": 2.3126343123796833e-05, "loss": 0.0228, "step": 69660 }, { "grad_norm": 0.4155538082122803, "learning_rate": 2.3112401198811418e-05, "loss": 0.02, "step": 69670 }, { "grad_norm": 0.44093939661979675, "learning_rate": 2.3098462214213418e-05, "loss": 0.0277, "step": 69680 }, { "grad_norm": 0.3549114763736725, "learning_rate": 2.308452617152712e-05, "loss": 0.0256, "step": 69690 }, { "grad_norm": 0.6224579811096191, "learning_rate": 2.3070593072276603e-05, "loss": 0.0244, "step": 69700 }, { "grad_norm": 0.2916826605796814, "learning_rate": 2.3056662917985518e-05, "loss": 0.0283, "step": 69710 }, { "grad_norm": 0.3814419209957123, "learning_rate": 2.3042735710177283e-05, "loss": 0.0183, "step": 69720 }, { "grad_norm": 0.5127722024917603, "learning_rate": 2.302881145037492e-05, "loss": 0.0315, "step": 69730 }, { "grad_norm": 0.39351847767829895, "learning_rate": 2.3014890140101192e-05, "loss": 0.018, "step": 69740 }, { "grad_norm": 0.5128204226493835, "learning_rate": 2.30009717808785e-05, "loss": 0.0173, "step": 69750 }, { "grad_norm": 0.49041759967803955, "learning_rate": 2.2987056374228967e-05, "loss": 0.0244, "step": 69760 }, { "grad_norm": 0.36907055974006653, "learning_rate": 2.297314392167431e-05, "loss": 0.021, "step": 69770 }, { "grad_norm": 0.6737592220306396, "learning_rate": 2.295923442473601e-05, "loss": 0.034, "step": 69780 }, { "grad_norm": 0.49395138025283813, "learning_rate": 2.294532788493519e-05, "loss": 0.0184, "step": 69790 }, { "grad_norm": 0.3617367446422577, "learning_rate": 2.293142430379262e-05, "loss": 0.0221, "step": 69800 }, { "grad_norm": 0.7645313739776611, "learning_rate": 2.291752368282879e-05, "loss": 0.0291, "step": 69810 }, { "grad_norm": 0.37710148096084595, "learning_rate": 2.290362602356385e-05, "loss": 0.0261, "step": 69820 }, { "grad_norm": 0.697575569152832, "learning_rate": 2.288973132751765e-05, "loss": 0.0224, "step": 69830 }, { "grad_norm": 0.328178346157074, "learning_rate": 2.287583959620965e-05, "loss": 0.0184, "step": 69840 }, { "grad_norm": 0.2543342113494873, "learning_rate": 2.2861950831159074e-05, "loss": 0.0221, "step": 69850 }, { "grad_norm": 0.35585612058639526, "learning_rate": 2.284806503388472e-05, "loss": 0.0159, "step": 69860 }, { "grad_norm": 0.22604194283485413, "learning_rate": 2.283418220590518e-05, "loss": 0.0232, "step": 69870 }, { "grad_norm": 0.45633867383003235, "learning_rate": 2.282030234873862e-05, "loss": 0.0228, "step": 69880 }, { "grad_norm": 0.31232503056526184, "learning_rate": 2.2806425463902948e-05, "loss": 0.0319, "step": 69890 }, { "grad_norm": 0.45435577630996704, "learning_rate": 2.279255155291568e-05, "loss": 0.0249, "step": 69900 }, { "grad_norm": 0.4654986560344696, "learning_rate": 2.277868061729409e-05, "loss": 0.022, "step": 69910 }, { "grad_norm": 0.5385194420814514, "learning_rate": 2.276481265855504e-05, "loss": 0.0271, "step": 69920 }, { "grad_norm": 0.31815284490585327, "learning_rate": 2.2750947678215128e-05, "loss": 0.0252, "step": 69930 }, { "grad_norm": 0.5998975038528442, "learning_rate": 2.2737085677790616e-05, "loss": 0.0357, "step": 69940 }, { "grad_norm": 0.18122906982898712, "learning_rate": 2.2723226658797443e-05, "loss": 0.0217, "step": 69950 }, { "grad_norm": 0.23355089128017426, "learning_rate": 2.270937062275117e-05, "loss": 0.0215, "step": 69960 }, { "grad_norm": 0.5716762542724609, "learning_rate": 2.2695517571167113e-05, "loss": 0.0333, "step": 69970 }, { "grad_norm": 0.42248857021331787, "learning_rate": 2.268166750556019e-05, "loss": 0.0274, "step": 69980 }, { "grad_norm": 0.4041752219200134, "learning_rate": 2.266782042744503e-05, "loss": 0.0222, "step": 69990 }, { "grad_norm": 0.3340100944042206, "learning_rate": 2.2653976338335935e-05, "loss": 0.0262, "step": 70000 }, { "grad_norm": 0.4164508879184723, "learning_rate": 2.2640135239746895e-05, "loss": 0.0301, "step": 70010 }, { "grad_norm": 0.3492295742034912, "learning_rate": 2.2626297133191515e-05, "loss": 0.0245, "step": 70020 }, { "grad_norm": 0.31495633721351624, "learning_rate": 2.261246202018312e-05, "loss": 0.0203, "step": 70030 }, { "grad_norm": 0.6619089841842651, "learning_rate": 2.2598629902234718e-05, "loss": 0.0174, "step": 70040 }, { "grad_norm": 0.6540395617485046, "learning_rate": 2.258480078085894e-05, "loss": 0.0318, "step": 70050 }, { "grad_norm": 0.3889750838279724, "learning_rate": 2.2570974657568126e-05, "loss": 0.0253, "step": 70060 }, { "grad_norm": 0.36313748359680176, "learning_rate": 2.2557151533874286e-05, "loss": 0.0252, "step": 70070 }, { "grad_norm": 0.38051584362983704, "learning_rate": 2.2543331411289116e-05, "loss": 0.0231, "step": 70080 }, { "grad_norm": 0.46411779522895813, "learning_rate": 2.252951429132392e-05, "loss": 0.0225, "step": 70090 }, { "grad_norm": 0.5741090178489685, "learning_rate": 2.2515700175489762e-05, "loss": 0.0233, "step": 70100 }, { "grad_norm": 0.46493151783943176, "learning_rate": 2.2501889065297288e-05, "loss": 0.0249, "step": 70110 }, { "grad_norm": 0.45033249258995056, "learning_rate": 2.2488080962256887e-05, "loss": 0.0209, "step": 70120 }, { "grad_norm": 0.354912668466568, "learning_rate": 2.2474275867878585e-05, "loss": 0.021, "step": 70130 }, { "grad_norm": 0.5158974528312683, "learning_rate": 2.246047378367211e-05, "loss": 0.0257, "step": 70140 }, { "grad_norm": 0.35934045910835266, "learning_rate": 2.2446674711146798e-05, "loss": 0.0261, "step": 70150 }, { "grad_norm": 0.6175347566604614, "learning_rate": 2.2432878651811734e-05, "loss": 0.0298, "step": 70160 }, { "grad_norm": 0.41050803661346436, "learning_rate": 2.2419085607175594e-05, "loss": 0.0288, "step": 70170 }, { "grad_norm": 0.6605318784713745, "learning_rate": 2.240529557874678e-05, "loss": 0.0238, "step": 70180 }, { "grad_norm": 0.33857861161231995, "learning_rate": 2.2391508568033358e-05, "loss": 0.0198, "step": 70190 }, { "grad_norm": 0.4602489769458771, "learning_rate": 2.237772457654307e-05, "loss": 0.0387, "step": 70200 }, { "grad_norm": 0.7330648303031921, "learning_rate": 2.236394360578327e-05, "loss": 0.0245, "step": 70210 }, { "grad_norm": 0.3270484507083893, "learning_rate": 2.2350165657261073e-05, "loss": 0.0271, "step": 70220 }, { "grad_norm": 0.5282891392707825, "learning_rate": 2.2336390732483164e-05, "loss": 0.0288, "step": 70230 }, { "grad_norm": 0.49189648032188416, "learning_rate": 2.2322618832955976e-05, "loss": 0.0234, "step": 70240 }, { "grad_norm": 0.343680202960968, "learning_rate": 2.2308849960185578e-05, "loss": 0.0234, "step": 70250 }, { "grad_norm": 0.43809136748313904, "learning_rate": 2.229508411567774e-05, "loss": 0.0241, "step": 70260 }, { "grad_norm": 0.36273905634880066, "learning_rate": 2.2281321300937823e-05, "loss": 0.0329, "step": 70270 }, { "grad_norm": 0.5297546982765198, "learning_rate": 2.2267561517470937e-05, "loss": 0.0407, "step": 70280 }, { "grad_norm": 0.16913869976997375, "learning_rate": 2.2253804766781845e-05, "loss": 0.0233, "step": 70290 }, { "grad_norm": 0.904032289981842, "learning_rate": 2.2240051050374933e-05, "loss": 0.0248, "step": 70300 }, { "grad_norm": 0.8303642272949219, "learning_rate": 2.22263003697543e-05, "loss": 0.0172, "step": 70310 }, { "grad_norm": 0.37513816356658936, "learning_rate": 2.2212552726423702e-05, "loss": 0.0218, "step": 70320 }, { "grad_norm": 0.2884993851184845, "learning_rate": 2.2198808121886578e-05, "loss": 0.0245, "step": 70330 }, { "grad_norm": 0.2728778123855591, "learning_rate": 2.218506655764598e-05, "loss": 0.0171, "step": 70340 }, { "grad_norm": 0.5558525323867798, "learning_rate": 2.2171328035204704e-05, "loss": 0.0199, "step": 70350 }, { "grad_norm": 0.6307030320167542, "learning_rate": 2.2157592556065133e-05, "loss": 0.0226, "step": 70360 }, { "grad_norm": 0.46460968255996704, "learning_rate": 2.2143860121729382e-05, "loss": 0.0274, "step": 70370 }, { "grad_norm": 0.4177462160587311, "learning_rate": 2.2130130733699206e-05, "loss": 0.0208, "step": 70380 }, { "grad_norm": 0.2508702278137207, "learning_rate": 2.2116404393476052e-05, "loss": 0.0214, "step": 70390 }, { "grad_norm": 0.3379734456539154, "learning_rate": 2.2102681102560967e-05, "loss": 0.0243, "step": 70400 }, { "grad_norm": 0.4596422612667084, "learning_rate": 2.208896086245476e-05, "loss": 0.0367, "step": 70410 }, { "grad_norm": 0.3329307436943054, "learning_rate": 2.2075243674657808e-05, "loss": 0.0413, "step": 70420 }, { "grad_norm": 0.551741361618042, "learning_rate": 2.206152954067024e-05, "loss": 0.0193, "step": 70430 }, { "grad_norm": 0.29083511233329773, "learning_rate": 2.2047818461991758e-05, "loss": 0.0198, "step": 70440 }, { "grad_norm": 0.3753964900970459, "learning_rate": 2.2034110440121864e-05, "loss": 0.03, "step": 70450 }, { "grad_norm": 0.29885584115982056, "learning_rate": 2.2020405476559585e-05, "loss": 0.0205, "step": 70460 }, { "grad_norm": 0.2293282300233841, "learning_rate": 2.2006703572803722e-05, "loss": 0.0253, "step": 70470 }, { "grad_norm": 0.48042744398117065, "learning_rate": 2.1993004730352646e-05, "loss": 0.0221, "step": 70480 }, { "grad_norm": 0.2903940677642822, "learning_rate": 2.1979308950704464e-05, "loss": 0.0159, "step": 70490 }, { "grad_norm": 0.1811990886926651, "learning_rate": 2.1965616235356923e-05, "loss": 0.0249, "step": 70500 }, { "grad_norm": 0.6244199275970459, "learning_rate": 2.1951926585807465e-05, "loss": 0.0194, "step": 70510 }, { "grad_norm": 0.26084956526756287, "learning_rate": 2.193824000355313e-05, "loss": 0.0188, "step": 70520 }, { "grad_norm": 0.28735122084617615, "learning_rate": 2.192455649009067e-05, "loss": 0.022, "step": 70530 }, { "grad_norm": 0.6040366888046265, "learning_rate": 2.1910876046916523e-05, "loss": 0.0289, "step": 70540 }, { "grad_norm": 0.5130309462547302, "learning_rate": 2.1897198675526725e-05, "loss": 0.0353, "step": 70550 }, { "grad_norm": 0.7209383249282837, "learning_rate": 2.1883524377417024e-05, "loss": 0.0338, "step": 70560 }, { "grad_norm": 0.37831515073776245, "learning_rate": 2.1869853154082825e-05, "loss": 0.016, "step": 70570 }, { "grad_norm": 0.34877273440361023, "learning_rate": 2.1856185007019215e-05, "loss": 0.0235, "step": 70580 }, { "grad_norm": 0.4210495948791504, "learning_rate": 2.1842519937720874e-05, "loss": 0.0285, "step": 70590 }, { "grad_norm": 0.541231632232666, "learning_rate": 2.182885794768224e-05, "loss": 0.0209, "step": 70600 }, { "grad_norm": 0.41314461827278137, "learning_rate": 2.1815199038397326e-05, "loss": 0.0231, "step": 70610 }, { "grad_norm": 0.7373329997062683, "learning_rate": 2.1801543211359893e-05, "loss": 0.033, "step": 70620 }, { "grad_norm": 0.38878947496414185, "learning_rate": 2.178789046806326e-05, "loss": 0.0261, "step": 70630 }, { "grad_norm": 0.5061694979667664, "learning_rate": 2.1774240810000547e-05, "loss": 0.0238, "step": 70640 }, { "grad_norm": 0.3418804407119751, "learning_rate": 2.17605942386644e-05, "loss": 0.0258, "step": 70650 }, { "grad_norm": 0.48787590861320496, "learning_rate": 2.174695075554723e-05, "loss": 0.0196, "step": 70660 }, { "grad_norm": 0.2803348898887634, "learning_rate": 2.173331036214103e-05, "loss": 0.0178, "step": 70670 }, { "grad_norm": 0.5952399969100952, "learning_rate": 2.171967305993753e-05, "loss": 0.0199, "step": 70680 }, { "grad_norm": 0.645194411277771, "learning_rate": 2.1706038850428024e-05, "loss": 0.0224, "step": 70690 }, { "grad_norm": 0.43020007014274597, "learning_rate": 2.169240773510361e-05, "loss": 0.0278, "step": 70700 }, { "grad_norm": 0.3070890009403229, "learning_rate": 2.1678779715454906e-05, "loss": 0.016, "step": 70710 }, { "grad_norm": 0.4655442535877228, "learning_rate": 2.166515479297229e-05, "loss": 0.0287, "step": 70720 }, { "grad_norm": 0.3504118323326111, "learning_rate": 2.1651532969145732e-05, "loss": 0.019, "step": 70730 }, { "grad_norm": 0.38021448254585266, "learning_rate": 2.1637914245464903e-05, "loss": 0.0227, "step": 70740 }, { "grad_norm": 0.5919350981712341, "learning_rate": 2.1624298623419133e-05, "loss": 0.0209, "step": 70750 }, { "grad_norm": 0.2574472427368164, "learning_rate": 2.1610686104497413e-05, "loss": 0.0218, "step": 70760 }, { "grad_norm": 0.48006197810173035, "learning_rate": 2.1597076690188366e-05, "loss": 0.0242, "step": 70770 }, { "grad_norm": 0.5180981159210205, "learning_rate": 2.1583470381980315e-05, "loss": 0.0234, "step": 70780 }, { "grad_norm": 0.6464535593986511, "learning_rate": 2.156986718136123e-05, "loss": 0.0288, "step": 70790 }, { "grad_norm": 0.32918253540992737, "learning_rate": 2.155626708981871e-05, "loss": 0.0291, "step": 70800 }, { "grad_norm": 0.519242525100708, "learning_rate": 2.1542670108840075e-05, "loss": 0.0201, "step": 70810 }, { "grad_norm": 0.3140595853328705, "learning_rate": 2.152907623991221e-05, "loss": 0.0261, "step": 70820 }, { "grad_norm": 0.23024532198905945, "learning_rate": 2.1515485484521802e-05, "loss": 0.028, "step": 70830 }, { "grad_norm": 0.32485315203666687, "learning_rate": 2.150189784415506e-05, "loss": 0.024, "step": 70840 }, { "grad_norm": 0.25336238741874695, "learning_rate": 2.148831332029794e-05, "loss": 0.0328, "step": 70850 }, { "grad_norm": 0.7661193013191223, "learning_rate": 2.147473191443599e-05, "loss": 0.0225, "step": 70860 }, { "grad_norm": 0.3384164571762085, "learning_rate": 2.1461153628054492e-05, "loss": 0.023, "step": 70870 }, { "grad_norm": 0.5335525274276733, "learning_rate": 2.1447578462638284e-05, "loss": 0.0224, "step": 70880 }, { "grad_norm": 0.40256553888320923, "learning_rate": 2.143400641967201e-05, "loss": 0.0217, "step": 70890 }, { "grad_norm": 0.3377045691013336, "learning_rate": 2.1420437500639823e-05, "loss": 0.0176, "step": 70900 }, { "grad_norm": 0.47439971566200256, "learning_rate": 2.1406871707025646e-05, "loss": 0.0283, "step": 70910 }, { "grad_norm": 0.3895641565322876, "learning_rate": 2.1393309040312963e-05, "loss": 0.0221, "step": 70920 }, { "grad_norm": 0.18673034012317657, "learning_rate": 2.137974950198502e-05, "loss": 0.0183, "step": 70930 }, { "grad_norm": 0.6715571880340576, "learning_rate": 2.1366193093524617e-05, "loss": 0.0228, "step": 70940 }, { "grad_norm": 0.30254846811294556, "learning_rate": 2.1352639816414288e-05, "loss": 0.0221, "step": 70950 }, { "grad_norm": 0.35419827699661255, "learning_rate": 2.1339089672136202e-05, "loss": 0.0171, "step": 70960 }, { "grad_norm": 0.5608015656471252, "learning_rate": 2.1325542662172193e-05, "loss": 0.0307, "step": 70970 }, { "grad_norm": 0.5462673306465149, "learning_rate": 2.1311998788003712e-05, "loss": 0.0172, "step": 70980 }, { "grad_norm": 0.2542742192745209, "learning_rate": 2.1298458051111907e-05, "loss": 0.0262, "step": 70990 }, { "grad_norm": 0.7723509073257446, "learning_rate": 2.1284920452977602e-05, "loss": 0.0235, "step": 71000 }, { "grad_norm": 0.3639264404773712, "learning_rate": 2.12713859950812e-05, "loss": 0.0182, "step": 71010 }, { "grad_norm": 0.6059115529060364, "learning_rate": 2.1257854678902832e-05, "loss": 0.0265, "step": 71020 }, { "grad_norm": 0.2728951871395111, "learning_rate": 2.124432650592227e-05, "loss": 0.0237, "step": 71030 }, { "grad_norm": 0.370211660861969, "learning_rate": 2.1230801477618937e-05, "loss": 0.0253, "step": 71040 }, { "grad_norm": 0.332888126373291, "learning_rate": 2.121727959547189e-05, "loss": 0.0169, "step": 71050 }, { "grad_norm": 0.5572397708892822, "learning_rate": 2.1203760860959886e-05, "loss": 0.0274, "step": 71060 }, { "grad_norm": 0.43682485818862915, "learning_rate": 2.119024527556127e-05, "loss": 0.022, "step": 71070 }, { "grad_norm": 0.28754472732543945, "learning_rate": 2.117673284075415e-05, "loss": 0.0275, "step": 71080 }, { "grad_norm": 0.37100109457969666, "learning_rate": 2.1163223558016165e-05, "loss": 0.0215, "step": 71090 }, { "grad_norm": 0.24407149851322174, "learning_rate": 2.1149717428824722e-05, "loss": 0.0265, "step": 71100 }, { "grad_norm": 0.5071152448654175, "learning_rate": 2.1136214454656784e-05, "loss": 0.0281, "step": 71110 }, { "grad_norm": 0.4715867340564728, "learning_rate": 2.1122714636989056e-05, "loss": 0.0206, "step": 71120 }, { "grad_norm": 0.3562624454498291, "learning_rate": 2.1109217977297823e-05, "loss": 0.0146, "step": 71130 }, { "grad_norm": 0.36863383650779724, "learning_rate": 2.1095724477059077e-05, "loss": 0.018, "step": 71140 }, { "grad_norm": 0.3803473114967346, "learning_rate": 2.1082234137748434e-05, "loss": 0.0236, "step": 71150 }, { "grad_norm": 0.1786855012178421, "learning_rate": 2.106874696084122e-05, "loss": 0.0194, "step": 71160 }, { "grad_norm": 0.39269742369651794, "learning_rate": 2.1055262947812315e-05, "loss": 0.0257, "step": 71170 }, { "grad_norm": 0.7461766004562378, "learning_rate": 2.104178210013636e-05, "loss": 0.03, "step": 71180 }, { "grad_norm": 0.37197625637054443, "learning_rate": 2.1028304419287554e-05, "loss": 0.0234, "step": 71190 }, { "grad_norm": 0.31721335649490356, "learning_rate": 2.101482990673982e-05, "loss": 0.0476, "step": 71200 }, { "grad_norm": 0.2248842567205429, "learning_rate": 2.1001358563966707e-05, "loss": 0.0168, "step": 71210 }, { "grad_norm": 0.307364821434021, "learning_rate": 2.098789039244144e-05, "loss": 0.0209, "step": 71220 }, { "grad_norm": 0.36909374594688416, "learning_rate": 2.097442539363685e-05, "loss": 0.0195, "step": 71230 }, { "grad_norm": 0.19196294248104095, "learning_rate": 2.0960963569025448e-05, "loss": 0.0129, "step": 71240 }, { "grad_norm": 0.2366277277469635, "learning_rate": 2.0947504920079435e-05, "loss": 0.0188, "step": 71250 }, { "grad_norm": 0.36989524960517883, "learning_rate": 2.093404944827059e-05, "loss": 0.0227, "step": 71260 }, { "grad_norm": 0.5774921774864197, "learning_rate": 2.0920597155070394e-05, "loss": 0.0203, "step": 71270 }, { "grad_norm": 0.5566659569740295, "learning_rate": 2.090714804194997e-05, "loss": 0.0259, "step": 71280 }, { "grad_norm": 1.1010245084762573, "learning_rate": 2.0893702110380114e-05, "loss": 0.0243, "step": 71290 }, { "grad_norm": 0.48572272062301636, "learning_rate": 2.0880259361831216e-05, "loss": 0.0194, "step": 71300 }, { "grad_norm": 0.38724249601364136, "learning_rate": 2.0866819797773397e-05, "loss": 0.0252, "step": 71310 }, { "grad_norm": 0.32664984464645386, "learning_rate": 2.085338341967634e-05, "loss": 0.0252, "step": 71320 }, { "grad_norm": 0.6993916034698486, "learning_rate": 2.083995022900946e-05, "loss": 0.0275, "step": 71330 }, { "grad_norm": 0.46621373295783997, "learning_rate": 2.0826520227241768e-05, "loss": 0.0249, "step": 71340 }, { "grad_norm": 0.4285570979118347, "learning_rate": 2.081309341584199e-05, "loss": 0.017, "step": 71350 }, { "grad_norm": 0.35291174054145813, "learning_rate": 2.0799669796278413e-05, "loss": 0.0183, "step": 71360 }, { "grad_norm": 0.918084442615509, "learning_rate": 2.078624937001906e-05, "loss": 0.021, "step": 71370 }, { "grad_norm": 0.36297252774238586, "learning_rate": 2.0772832138531544e-05, "loss": 0.0244, "step": 71380 }, { "grad_norm": 0.33872851729393005, "learning_rate": 2.0759418103283158e-05, "loss": 0.031, "step": 71390 }, { "grad_norm": 0.4365693926811218, "learning_rate": 2.0746007265740846e-05, "loss": 0.0243, "step": 71400 }, { "grad_norm": 0.22990582883358002, "learning_rate": 2.0732599627371215e-05, "loss": 0.0155, "step": 71410 }, { "grad_norm": 0.5423010587692261, "learning_rate": 2.0719195189640468e-05, "loss": 0.018, "step": 71420 }, { "grad_norm": 0.42005348205566406, "learning_rate": 2.0705793954014528e-05, "loss": 0.0256, "step": 71430 }, { "grad_norm": 0.2334100753068924, "learning_rate": 2.06923959219589e-05, "loss": 0.0151, "step": 71440 }, { "grad_norm": 0.37441155314445496, "learning_rate": 2.0679001094938793e-05, "loss": 0.0176, "step": 71450 }, { "grad_norm": 0.2863101065158844, "learning_rate": 2.066560947441904e-05, "loss": 0.025, "step": 71460 }, { "grad_norm": 0.41220754384994507, "learning_rate": 2.0652221061864152e-05, "loss": 0.0181, "step": 71470 }, { "grad_norm": 0.5847857594490051, "learning_rate": 2.0638835858738226e-05, "loss": 0.0233, "step": 71480 }, { "grad_norm": 0.2475959062576294, "learning_rate": 2.0625453866505068e-05, "loss": 0.0214, "step": 71490 }, { "grad_norm": 0.36301833391189575, "learning_rate": 2.0612075086628123e-05, "loss": 0.0193, "step": 71500 }, { "grad_norm": 0.2820294499397278, "learning_rate": 2.059869952057045e-05, "loss": 0.0249, "step": 71510 }, { "grad_norm": 0.4409404695034027, "learning_rate": 2.0585327169794793e-05, "loss": 0.0338, "step": 71520 }, { "grad_norm": 0.39652198553085327, "learning_rate": 2.0571958035763534e-05, "loss": 0.0178, "step": 71530 }, { "grad_norm": 0.45811623334884644, "learning_rate": 2.0558592119938715e-05, "loss": 0.0266, "step": 71540 }, { "grad_norm": 0.23974621295928955, "learning_rate": 2.054522942378198e-05, "loss": 0.028, "step": 71550 }, { "grad_norm": 0.20677450299263, "learning_rate": 2.053186994875469e-05, "loss": 0.0211, "step": 71560 }, { "grad_norm": 0.45440223813056946, "learning_rate": 2.0518513696317775e-05, "loss": 0.0177, "step": 71570 }, { "grad_norm": 0.23374053835868835, "learning_rate": 2.050516066793188e-05, "loss": 0.0175, "step": 71580 }, { "grad_norm": 0.19167035818099976, "learning_rate": 2.0491810865057276e-05, "loss": 0.0271, "step": 71590 }, { "grad_norm": 0.34675395488739014, "learning_rate": 2.0478464289153882e-05, "loss": 0.0298, "step": 71600 }, { "grad_norm": 0.4181620180606842, "learning_rate": 2.0465120941681232e-05, "loss": 0.0302, "step": 71610 }, { "grad_norm": 0.2705581784248352, "learning_rate": 2.0451780824098566e-05, "loss": 0.0174, "step": 71620 }, { "grad_norm": 0.2585398256778717, "learning_rate": 2.043844393786471e-05, "loss": 0.0232, "step": 71630 }, { "grad_norm": 0.5433647632598877, "learning_rate": 2.0425110284438193e-05, "loss": 0.0233, "step": 71640 }, { "grad_norm": 0.3706977665424347, "learning_rate": 2.041177986527712e-05, "loss": 0.0233, "step": 71650 }, { "grad_norm": 0.7403846979141235, "learning_rate": 2.0398452681839347e-05, "loss": 0.0278, "step": 71660 }, { "grad_norm": 0.5213766098022461, "learning_rate": 2.0385128735582272e-05, "loss": 0.0279, "step": 71670 }, { "grad_norm": 0.17745628952980042, "learning_rate": 2.0371808027963003e-05, "loss": 0.0224, "step": 71680 }, { "grad_norm": 0.47046810388565063, "learning_rate": 2.035849056043825e-05, "loss": 0.0241, "step": 71690 }, { "grad_norm": 0.6008168458938599, "learning_rate": 2.03451763344644e-05, "loss": 0.0204, "step": 71700 }, { "grad_norm": 0.313703715801239, "learning_rate": 2.0331865351497482e-05, "loss": 0.0267, "step": 71710 }, { "grad_norm": 0.6941224336624146, "learning_rate": 2.0318557612993185e-05, "loss": 0.0267, "step": 71720 }, { "grad_norm": 0.3541021943092346, "learning_rate": 2.0305253120406782e-05, "loss": 0.0221, "step": 71730 }, { "grad_norm": 0.30677729845046997, "learning_rate": 2.029195187519325e-05, "loss": 0.0193, "step": 71740 }, { "grad_norm": 0.3538563549518585, "learning_rate": 2.0278653878807217e-05, "loss": 0.0205, "step": 71750 }, { "grad_norm": 0.6775062680244446, "learning_rate": 2.0265359132702892e-05, "loss": 0.0242, "step": 71760 }, { "grad_norm": 0.4263536036014557, "learning_rate": 2.0252067638334182e-05, "loss": 0.0146, "step": 71770 }, { "grad_norm": 0.2881096601486206, "learning_rate": 2.023877939715464e-05, "loss": 0.022, "step": 71780 }, { "grad_norm": 0.20301203429698944, "learning_rate": 2.0225494410617448e-05, "loss": 0.0271, "step": 71790 }, { "grad_norm": 0.41536760330200195, "learning_rate": 2.02122126801754e-05, "loss": 0.0272, "step": 71800 }, { "grad_norm": 0.3354295492172241, "learning_rate": 2.0198934207281018e-05, "loss": 0.0458, "step": 71810 }, { "grad_norm": 0.4532906711101532, "learning_rate": 2.0185658993386357e-05, "loss": 0.0236, "step": 71820 }, { "grad_norm": 0.7884275913238525, "learning_rate": 2.0172387039943226e-05, "loss": 0.0337, "step": 71830 }, { "grad_norm": 0.2038765251636505, "learning_rate": 2.0159118348402968e-05, "loss": 0.0188, "step": 71840 }, { "grad_norm": 0.6205840706825256, "learning_rate": 2.0145852920216697e-05, "loss": 0.0155, "step": 71850 }, { "grad_norm": 0.4176652431488037, "learning_rate": 2.0132590756835045e-05, "loss": 0.0259, "step": 71860 }, { "grad_norm": 0.5893437266349792, "learning_rate": 2.0119331859708386e-05, "loss": 0.0266, "step": 71870 }, { "grad_norm": 0.4101518988609314, "learning_rate": 2.0106076230286654e-05, "loss": 0.0277, "step": 71880 }, { "grad_norm": 0.615633487701416, "learning_rate": 2.0092823870019488e-05, "loss": 0.0203, "step": 71890 }, { "grad_norm": 0.5218396782875061, "learning_rate": 2.0079574780356115e-05, "loss": 0.0272, "step": 71900 }, { "grad_norm": 0.7477638721466064, "learning_rate": 2.0066328962745486e-05, "loss": 0.0129, "step": 71910 }, { "grad_norm": 0.20506946742534637, "learning_rate": 2.00530864186361e-05, "loss": 0.0306, "step": 71920 }, { "grad_norm": 0.47635385394096375, "learning_rate": 2.003984714947618e-05, "loss": 0.0283, "step": 71930 }, { "grad_norm": 0.29314929246902466, "learning_rate": 2.0026611156713515e-05, "loss": 0.0202, "step": 71940 }, { "grad_norm": 0.4001759886741638, "learning_rate": 2.0013378441795584e-05, "loss": 0.0179, "step": 71950 }, { "grad_norm": 0.5462442636489868, "learning_rate": 2.0000149006169506e-05, "loss": 0.0282, "step": 71960 }, { "grad_norm": 0.33482152223587036, "learning_rate": 1.9986922851282048e-05, "loss": 0.0293, "step": 71970 }, { "grad_norm": 0.5994817614555359, "learning_rate": 1.997369997857956e-05, "loss": 0.0202, "step": 71980 }, { "grad_norm": 0.17560215294361115, "learning_rate": 1.9960480389508103e-05, "loss": 0.0222, "step": 71990 }, { "grad_norm": 0.3026907742023468, "learning_rate": 1.9947264085513363e-05, "loss": 0.0416, "step": 72000 }, { "grad_norm": 0.5135044455528259, "learning_rate": 1.993405106804062e-05, "loss": 0.0182, "step": 72010 }, { "grad_norm": 0.550743579864502, "learning_rate": 1.9920841338534867e-05, "loss": 0.0222, "step": 72020 }, { "grad_norm": 0.235332652926445, "learning_rate": 1.9907634898440646e-05, "loss": 0.0235, "step": 72030 }, { "grad_norm": 0.5280531644821167, "learning_rate": 1.989443174920227e-05, "loss": 0.0283, "step": 72040 }, { "grad_norm": 0.33364424109458923, "learning_rate": 1.988123189226355e-05, "loss": 0.023, "step": 72050 }, { "grad_norm": 0.28117045760154724, "learning_rate": 1.986803532906805e-05, "loss": 0.0272, "step": 72060 }, { "grad_norm": 0.4350893795490265, "learning_rate": 1.985484206105889e-05, "loss": 0.029, "step": 72070 }, { "grad_norm": 0.345319539308548, "learning_rate": 1.9841652089678887e-05, "loss": 0.0173, "step": 72080 }, { "grad_norm": 0.30217134952545166, "learning_rate": 1.9828465416370435e-05, "loss": 0.0317, "step": 72090 }, { "grad_norm": 0.49732348322868347, "learning_rate": 1.9815282042575685e-05, "loss": 0.0185, "step": 72100 }, { "grad_norm": 0.47946229577064514, "learning_rate": 1.9802101969736285e-05, "loss": 0.0294, "step": 72110 }, { "grad_norm": 0.40884071588516235, "learning_rate": 1.9788925199293633e-05, "loss": 0.0174, "step": 72120 }, { "grad_norm": 0.42001309990882874, "learning_rate": 1.977575173268868e-05, "loss": 0.0195, "step": 72130 }, { "grad_norm": 0.40364471077919006, "learning_rate": 1.9762581571362094e-05, "loss": 0.0235, "step": 72140 }, { "grad_norm": 0.4475158452987671, "learning_rate": 1.9749414716754112e-05, "loss": 0.0163, "step": 72150 }, { "grad_norm": 0.37344130873680115, "learning_rate": 1.973625117030465e-05, "loss": 0.0226, "step": 72160 }, { "grad_norm": 0.4904366135597229, "learning_rate": 1.972309093345326e-05, "loss": 0.0196, "step": 72170 }, { "grad_norm": 0.21609093248844147, "learning_rate": 1.970993400763915e-05, "loss": 0.0239, "step": 72180 }, { "grad_norm": 0.567291796207428, "learning_rate": 1.969678039430109e-05, "loss": 0.0234, "step": 72190 }, { "grad_norm": 0.5382118821144104, "learning_rate": 1.9683630094877564e-05, "loss": 0.0295, "step": 72200 }, { "grad_norm": 0.5116124749183655, "learning_rate": 1.9670483110806696e-05, "loss": 0.0202, "step": 72210 }, { "grad_norm": 0.3317338824272156, "learning_rate": 1.9657339443526172e-05, "loss": 0.0147, "step": 72220 }, { "grad_norm": 0.4859132468700409, "learning_rate": 1.964419909447339e-05, "loss": 0.0155, "step": 72230 }, { "grad_norm": 0.2306305468082428, "learning_rate": 1.963106206508536e-05, "loss": 0.0148, "step": 72240 }, { "grad_norm": 0.6727184057235718, "learning_rate": 1.9617928356798738e-05, "loss": 0.0229, "step": 72250 }, { "grad_norm": 0.3208141326904297, "learning_rate": 1.960479797104977e-05, "loss": 0.0242, "step": 72260 }, { "grad_norm": 0.3672705292701721, "learning_rate": 1.9591670909274418e-05, "loss": 0.023, "step": 72270 }, { "grad_norm": 0.4703079164028168, "learning_rate": 1.9578547172908184e-05, "loss": 0.0224, "step": 72280 }, { "grad_norm": 0.500444769859314, "learning_rate": 1.9565426763386325e-05, "loss": 0.0188, "step": 72290 }, { "grad_norm": 0.2772858738899231, "learning_rate": 1.9552309682143622e-05, "loss": 0.0168, "step": 72300 }, { "grad_norm": 0.42281094193458557, "learning_rate": 1.9539195930614574e-05, "loss": 0.0212, "step": 72310 }, { "grad_norm": 0.252285361289978, "learning_rate": 1.952608551023324e-05, "loss": 0.0162, "step": 72320 }, { "grad_norm": 0.27013999223709106, "learning_rate": 1.9512978422433397e-05, "loss": 0.0122, "step": 72330 }, { "grad_norm": 0.47992172837257385, "learning_rate": 1.9499874668648376e-05, "loss": 0.0159, "step": 72340 }, { "grad_norm": 0.2650616466999054, "learning_rate": 1.9486774250311213e-05, "loss": 0.027, "step": 72350 }, { "grad_norm": 0.3331201672554016, "learning_rate": 1.9473677168854537e-05, "loss": 0.0273, "step": 72360 }, { "grad_norm": 0.3649253845214844, "learning_rate": 1.946058342571065e-05, "loss": 0.0332, "step": 72370 }, { "grad_norm": 0.5257813930511475, "learning_rate": 1.9447493022311424e-05, "loss": 0.0179, "step": 72380 }, { "grad_norm": 0.4564738869667053, "learning_rate": 1.9434405960088442e-05, "loss": 0.0263, "step": 72390 }, { "grad_norm": 0.5557138919830322, "learning_rate": 1.9421322240472857e-05, "loss": 0.0201, "step": 72400 }, { "grad_norm": 0.46194180846214294, "learning_rate": 1.9408241864895494e-05, "loss": 0.0211, "step": 72410 }, { "grad_norm": 0.7613824009895325, "learning_rate": 1.939516483478681e-05, "loss": 0.0263, "step": 72420 }, { "grad_norm": 0.41902244091033936, "learning_rate": 1.9382091151576902e-05, "loss": 0.0404, "step": 72430 }, { "grad_norm": 0.5297175049781799, "learning_rate": 1.9369020816695454e-05, "loss": 0.0221, "step": 72440 }, { "grad_norm": 0.49808335304260254, "learning_rate": 1.935595383157184e-05, "loss": 0.0238, "step": 72450 }, { "grad_norm": 0.34028196334838867, "learning_rate": 1.9342890197635057e-05, "loss": 0.0223, "step": 72460 }, { "grad_norm": 0.2618710696697235, "learning_rate": 1.9329829916313685e-05, "loss": 0.0222, "step": 72470 }, { "grad_norm": 0.269652783870697, "learning_rate": 1.9316772989036013e-05, "loss": 0.022, "step": 72480 }, { "grad_norm": 0.41948556900024414, "learning_rate": 1.930371941722991e-05, "loss": 0.0174, "step": 72490 }, { "grad_norm": 0.30059710144996643, "learning_rate": 1.929066920232292e-05, "loss": 0.0138, "step": 72500 }, { "grad_norm": 0.44219282269477844, "learning_rate": 1.9277622345742152e-05, "loss": 0.0263, "step": 72510 }, { "grad_norm": 0.18432942032814026, "learning_rate": 1.926457884891444e-05, "loss": 0.0168, "step": 72520 }, { "grad_norm": 0.3754384219646454, "learning_rate": 1.9251538713266147e-05, "loss": 0.021, "step": 72530 }, { "grad_norm": 0.40148454904556274, "learning_rate": 1.9238501940223358e-05, "loss": 0.0251, "step": 72540 }, { "grad_norm": 0.2848987877368927, "learning_rate": 1.922546853121175e-05, "loss": 0.0246, "step": 72550 }, { "grad_norm": 0.5017551779747009, "learning_rate": 1.9212438487656647e-05, "loss": 0.0311, "step": 72560 }, { "grad_norm": 0.5553677082061768, "learning_rate": 1.9199411810982965e-05, "loss": 0.041, "step": 72570 }, { "grad_norm": 0.4736349880695343, "learning_rate": 1.918638850261532e-05, "loss": 0.0309, "step": 72580 }, { "grad_norm": 0.21973998844623566, "learning_rate": 1.9173368563977885e-05, "loss": 0.0145, "step": 72590 }, { "grad_norm": 0.24923451244831085, "learning_rate": 1.9160351996494513e-05, "loss": 0.0178, "step": 72600 }, { "grad_norm": 0.32418838143348694, "learning_rate": 1.9147338801588677e-05, "loss": 0.0184, "step": 72610 }, { "grad_norm": 0.4427131116390228, "learning_rate": 1.913432898068351e-05, "loss": 0.0201, "step": 72620 }, { "grad_norm": 0.42614737153053284, "learning_rate": 1.912132253520169e-05, "loss": 0.0157, "step": 72630 }, { "grad_norm": 0.28983166813850403, "learning_rate": 1.910831946656564e-05, "loss": 0.0198, "step": 72640 }, { "grad_norm": 0.36730873584747314, "learning_rate": 1.909531977619731e-05, "loss": 0.0266, "step": 72650 }, { "grad_norm": 0.3768622875213623, "learning_rate": 1.908232346551834e-05, "loss": 0.0194, "step": 72660 }, { "grad_norm": 0.4152430593967438, "learning_rate": 1.9069330535949993e-05, "loss": 0.0148, "step": 72670 }, { "grad_norm": 0.21986046433448792, "learning_rate": 1.9056340988913167e-05, "loss": 0.0205, "step": 72680 }, { "grad_norm": 0.3242240846157074, "learning_rate": 1.9043354825828357e-05, "loss": 0.0197, "step": 72690 }, { "grad_norm": 0.4311985373497009, "learning_rate": 1.9030372048115724e-05, "loss": 0.0292, "step": 72700 }, { "grad_norm": 0.7509191632270813, "learning_rate": 1.9017392657195056e-05, "loss": 0.0358, "step": 72710 }, { "grad_norm": 0.37245258688926697, "learning_rate": 1.9004416654485724e-05, "loss": 0.019, "step": 72720 }, { "grad_norm": 0.5008540749549866, "learning_rate": 1.8991444041406785e-05, "loss": 0.0238, "step": 72730 }, { "grad_norm": 0.3377772569656372, "learning_rate": 1.8978474819376912e-05, "loss": 0.0179, "step": 72740 }, { "grad_norm": 0.6088671088218689, "learning_rate": 1.8965508989814406e-05, "loss": 0.0218, "step": 72750 }, { "grad_norm": 0.5316018462181091, "learning_rate": 1.8952546554137158e-05, "loss": 0.0405, "step": 72760 }, { "grad_norm": 0.40826383233070374, "learning_rate": 1.8939587513762757e-05, "loss": 0.0198, "step": 72770 }, { "grad_norm": 0.6339783668518066, "learning_rate": 1.892663187010835e-05, "loss": 0.0326, "step": 72780 }, { "grad_norm": 0.3704231083393097, "learning_rate": 1.891367962459077e-05, "loss": 0.0232, "step": 72790 }, { "grad_norm": 0.7266966700553894, "learning_rate": 1.890073077862644e-05, "loss": 0.0179, "step": 72800 }, { "grad_norm": 0.5757479071617126, "learning_rate": 1.888778533363146e-05, "loss": 0.0317, "step": 72810 }, { "grad_norm": 0.40993988513946533, "learning_rate": 1.8874843291021478e-05, "loss": 0.0162, "step": 72820 }, { "grad_norm": 0.2954520285129547, "learning_rate": 1.8861904652211853e-05, "loss": 0.0172, "step": 72830 }, { "grad_norm": 0.7163534164428711, "learning_rate": 1.8848969418617507e-05, "loss": 0.0196, "step": 72840 }, { "grad_norm": 0.1977420598268509, "learning_rate": 1.8836037591653044e-05, "loss": 0.0131, "step": 72850 }, { "grad_norm": 0.7050755023956299, "learning_rate": 1.882310917273262e-05, "loss": 0.028, "step": 72860 }, { "grad_norm": 0.6757055521011353, "learning_rate": 1.881018416327014e-05, "loss": 0.0376, "step": 72870 }, { "grad_norm": 0.20171897113323212, "learning_rate": 1.8797262564679014e-05, "loss": 0.0171, "step": 72880 }, { "grad_norm": 0.31403639912605286, "learning_rate": 1.8784344378372355e-05, "loss": 0.0199, "step": 72890 }, { "grad_norm": 0.5026949048042297, "learning_rate": 1.8771429605762843e-05, "loss": 0.0277, "step": 72900 }, { "grad_norm": 0.4381386637687683, "learning_rate": 1.8758518248262835e-05, "loss": 0.0218, "step": 72910 }, { "grad_norm": 0.24392199516296387, "learning_rate": 1.8745610307284305e-05, "loss": 0.0187, "step": 72920 }, { "grad_norm": 0.3818354606628418, "learning_rate": 1.8732705784238862e-05, "loss": 0.0158, "step": 72930 }, { "grad_norm": 0.9026896357536316, "learning_rate": 1.8719804680537682e-05, "loss": 0.0181, "step": 72940 }, { "grad_norm": 0.21124187111854553, "learning_rate": 1.870690699759164e-05, "loss": 0.0214, "step": 72950 }, { "grad_norm": 0.5984336137771606, "learning_rate": 1.8694012736811213e-05, "loss": 0.0214, "step": 72960 }, { "grad_norm": 0.14020492136478424, "learning_rate": 1.8681121899606463e-05, "loss": 0.0113, "step": 72970 }, { "grad_norm": 0.2720642685890198, "learning_rate": 1.8668234487387144e-05, "loss": 0.0182, "step": 72980 }, { "grad_norm": 0.5265174508094788, "learning_rate": 1.8655350501562592e-05, "loss": 0.0193, "step": 72990 }, { "grad_norm": 0.2637721300125122, "learning_rate": 1.8642469943541796e-05, "loss": 0.0268, "step": 73000 }, { "grad_norm": 0.48933491110801697, "learning_rate": 1.8629592814733328e-05, "loss": 0.0203, "step": 73010 }, { "grad_norm": 0.327149897813797, "learning_rate": 1.8616719116545438e-05, "loss": 0.0164, "step": 73020 }, { "grad_norm": 0.41361114382743835, "learning_rate": 1.8603848850385952e-05, "loss": 0.015, "step": 73030 }, { "grad_norm": 0.19817452132701874, "learning_rate": 1.859098201766236e-05, "loss": 0.0145, "step": 73040 }, { "grad_norm": 0.4752572476863861, "learning_rate": 1.8578118619781726e-05, "loss": 0.0228, "step": 73050 }, { "grad_norm": 0.5314309597015381, "learning_rate": 1.856525865815083e-05, "loss": 0.021, "step": 73060 }, { "grad_norm": 0.34527671337127686, "learning_rate": 1.8552402134175966e-05, "loss": 0.0193, "step": 73070 }, { "grad_norm": 0.37968868017196655, "learning_rate": 1.8539549049263143e-05, "loss": 0.019, "step": 73080 }, { "grad_norm": 0.16050852835178375, "learning_rate": 1.8526699404817917e-05, "loss": 0.0246, "step": 73090 }, { "grad_norm": 0.38820138573646545, "learning_rate": 1.8513853202245546e-05, "loss": 0.0147, "step": 73100 }, { "grad_norm": 0.5988773107528687, "learning_rate": 1.850101044295081e-05, "loss": 0.0279, "step": 73110 }, { "grad_norm": 0.42590320110321045, "learning_rate": 1.848817112833825e-05, "loss": 0.0277, "step": 73120 }, { "grad_norm": 0.3202652037143707, "learning_rate": 1.8475335259811905e-05, "loss": 0.0157, "step": 73130 }, { "grad_norm": 0.8854885697364807, "learning_rate": 1.8462502838775508e-05, "loss": 0.0194, "step": 73140 }, { "grad_norm": 0.36071187257766724, "learning_rate": 1.8449673866632378e-05, "loss": 0.0184, "step": 73150 }, { "grad_norm": 0.2496882826089859, "learning_rate": 1.8436848344785473e-05, "loss": 0.0491, "step": 73160 }, { "grad_norm": 0.27814754843711853, "learning_rate": 1.8424026274637378e-05, "loss": 0.0259, "step": 73170 }, { "grad_norm": 0.5853844285011292, "learning_rate": 1.8411207657590312e-05, "loss": 0.0294, "step": 73180 }, { "grad_norm": 0.3174135088920593, "learning_rate": 1.839839249504607e-05, "loss": 0.0196, "step": 73190 }, { "grad_norm": 0.1281084269285202, "learning_rate": 1.8385580788406108e-05, "loss": 0.0273, "step": 73200 }, { "grad_norm": 0.4289776086807251, "learning_rate": 1.837277253907152e-05, "loss": 0.0281, "step": 73210 }, { "grad_norm": 0.4143698811531067, "learning_rate": 1.8359967748442957e-05, "loss": 0.0198, "step": 73220 }, { "grad_norm": 0.44050148129463196, "learning_rate": 1.8347166417920774e-05, "loss": 0.019, "step": 73230 }, { "grad_norm": 0.2834152579307556, "learning_rate": 1.833436854890484e-05, "loss": 0.0203, "step": 73240 }, { "grad_norm": 0.936368465423584, "learning_rate": 1.8321574142794796e-05, "loss": 0.0239, "step": 73250 }, { "grad_norm": 1.2873964309692383, "learning_rate": 1.8308783200989748e-05, "loss": 0.0252, "step": 73260 }, { "grad_norm": 0.27182093262672424, "learning_rate": 1.829599572488855e-05, "loss": 0.0136, "step": 73270 }, { "grad_norm": 0.16579239070415497, "learning_rate": 1.828321171588957e-05, "loss": 0.0287, "step": 73280 }, { "grad_norm": 0.43200936913490295, "learning_rate": 1.827043117539089e-05, "loss": 0.0239, "step": 73290 }, { "grad_norm": 0.3820010721683502, "learning_rate": 1.8257654104790116e-05, "loss": 0.0293, "step": 73300 }, { "grad_norm": 0.8098986744880676, "learning_rate": 1.82448805054846e-05, "loss": 0.0237, "step": 73310 }, { "grad_norm": 0.2964829206466675, "learning_rate": 1.823211037887119e-05, "loss": 0.0187, "step": 73320 }, { "grad_norm": 0.26064279675483704, "learning_rate": 1.821934372634645e-05, "loss": 0.0253, "step": 73330 }, { "grad_norm": 0.5799574255943298, "learning_rate": 1.8206580549306472e-05, "loss": 0.024, "step": 73340 }, { "grad_norm": 0.5335740447044373, "learning_rate": 1.819382084914707e-05, "loss": 0.0207, "step": 73350 }, { "grad_norm": 0.6539322137832642, "learning_rate": 1.8181064627263577e-05, "loss": 0.0354, "step": 73360 }, { "grad_norm": 0.25824567675590515, "learning_rate": 1.8168311885051022e-05, "loss": 0.0277, "step": 73370 }, { "grad_norm": 0.1437930315732956, "learning_rate": 1.815556262390402e-05, "loss": 0.024, "step": 73380 }, { "grad_norm": 0.2575647830963135, "learning_rate": 1.8142816845216826e-05, "loss": 0.0169, "step": 73390 }, { "grad_norm": 0.7229675650596619, "learning_rate": 1.8130074550383276e-05, "loss": 0.016, "step": 73400 }, { "grad_norm": 0.4102534055709839, "learning_rate": 1.8117335740796848e-05, "loss": 0.0164, "step": 73410 }, { "grad_norm": 0.4302271008491516, "learning_rate": 1.810460041785067e-05, "loss": 0.033, "step": 73420 }, { "grad_norm": 0.37701615691185, "learning_rate": 1.809186858293742e-05, "loss": 0.0194, "step": 73430 }, { "grad_norm": 0.5865994095802307, "learning_rate": 1.8079140237449448e-05, "loss": 0.0246, "step": 73440 }, { "grad_norm": 0.27314087748527527, "learning_rate": 1.8066415382778707e-05, "loss": 0.0191, "step": 73450 }, { "grad_norm": 0.5318112969398499, "learning_rate": 1.8053694020316786e-05, "loss": 0.0197, "step": 73460 }, { "grad_norm": 0.32712891697883606, "learning_rate": 1.8040976151454837e-05, "loss": 0.0257, "step": 73470 }, { "grad_norm": 0.35901641845703125, "learning_rate": 1.8028261777583704e-05, "loss": 0.0208, "step": 73480 }, { "grad_norm": 0.2655530869960785, "learning_rate": 1.8015550900093758e-05, "loss": 0.0192, "step": 73490 }, { "grad_norm": 0.4801729917526245, "learning_rate": 1.800284352037512e-05, "loss": 0.0196, "step": 73500 }, { "grad_norm": 0.47219642996788025, "learning_rate": 1.799013963981738e-05, "loss": 0.0282, "step": 73510 }, { "grad_norm": 0.35204869508743286, "learning_rate": 1.797743925980987e-05, "loss": 0.0153, "step": 73520 }, { "grad_norm": 0.31404635310173035, "learning_rate": 1.7964742381741435e-05, "loss": 0.0171, "step": 73530 }, { "grad_norm": 0.40872272849082947, "learning_rate": 1.7952049007000627e-05, "loss": 0.0229, "step": 73540 }, { "grad_norm": 0.5811820030212402, "learning_rate": 1.7939359136975535e-05, "loss": 0.0254, "step": 73550 }, { "grad_norm": 0.2694295048713684, "learning_rate": 1.7926672773053932e-05, "loss": 0.031, "step": 73560 }, { "grad_norm": 0.4266238808631897, "learning_rate": 1.791398991662317e-05, "loss": 0.0222, "step": 73570 }, { "grad_norm": 0.49774494767189026, "learning_rate": 1.7901310569070244e-05, "loss": 0.0243, "step": 73580 }, { "grad_norm": 0.5022046566009521, "learning_rate": 1.7888634731781724e-05, "loss": 0.023, "step": 73590 }, { "grad_norm": 0.8359641432762146, "learning_rate": 1.7875962406143844e-05, "loss": 0.0171, "step": 73600 }, { "grad_norm": 0.2681678235530853, "learning_rate": 1.7863293593542403e-05, "loss": 0.0193, "step": 73610 }, { "grad_norm": 0.48492512106895447, "learning_rate": 1.7850628295362853e-05, "loss": 0.0251, "step": 73620 }, { "grad_norm": 0.2682124078273773, "learning_rate": 1.7837966512990263e-05, "loss": 0.0151, "step": 73630 }, { "grad_norm": 0.4039389193058014, "learning_rate": 1.782530824780931e-05, "loss": 0.0221, "step": 73640 }, { "grad_norm": 0.30222854018211365, "learning_rate": 1.7812653501204263e-05, "loss": 0.0259, "step": 73650 }, { "grad_norm": 0.44855421781539917, "learning_rate": 1.780000227455903e-05, "loss": 0.0217, "step": 73660 }, { "grad_norm": 0.2459663599729538, "learning_rate": 1.778735456925716e-05, "loss": 0.017, "step": 73670 }, { "grad_norm": 0.28833839297294617, "learning_rate": 1.7774710386681737e-05, "loss": 0.0211, "step": 73680 }, { "grad_norm": 0.22283929586410522, "learning_rate": 1.7762069728215545e-05, "loss": 0.0272, "step": 73690 }, { "grad_norm": 0.4717601239681244, "learning_rate": 1.7749432595240932e-05, "loss": 0.02, "step": 73700 }, { "grad_norm": 0.24781818687915802, "learning_rate": 1.773679898913991e-05, "loss": 0.0299, "step": 73710 }, { "grad_norm": 0.41854673624038696, "learning_rate": 1.7724168911294015e-05, "loss": 0.0219, "step": 73720 }, { "grad_norm": 0.8246659636497498, "learning_rate": 1.7711542363084504e-05, "loss": 0.0223, "step": 73730 }, { "grad_norm": 0.5053256154060364, "learning_rate": 1.7698919345892158e-05, "loss": 0.0235, "step": 73740 }, { "grad_norm": 0.23415860533714294, "learning_rate": 1.768629986109743e-05, "loss": 0.0177, "step": 73750 }, { "grad_norm": 0.36211588978767395, "learning_rate": 1.7673683910080357e-05, "loss": 0.0334, "step": 73760 }, { "grad_norm": 0.3844231367111206, "learning_rate": 1.7661071494220638e-05, "loss": 0.0221, "step": 73770 }, { "grad_norm": 0.7433190941810608, "learning_rate": 1.7648462614897498e-05, "loss": 0.0231, "step": 73780 }, { "grad_norm": 0.3651314079761505, "learning_rate": 1.7635857273489863e-05, "loss": 0.0288, "step": 73790 }, { "grad_norm": 0.2994169294834137, "learning_rate": 1.7623255471376198e-05, "loss": 0.0232, "step": 73800 }, { "grad_norm": 0.4372546672821045, "learning_rate": 1.7610657209934646e-05, "loss": 0.0204, "step": 73810 }, { "grad_norm": 0.29625219106674194, "learning_rate": 1.7598062490542915e-05, "loss": 0.0228, "step": 73820 }, { "grad_norm": 0.3012363314628601, "learning_rate": 1.7585471314578377e-05, "loss": 0.0165, "step": 73830 }, { "grad_norm": 0.252103716135025, "learning_rate": 1.7572883683417945e-05, "loss": 0.0219, "step": 73840 }, { "grad_norm": 1.5750705003738403, "learning_rate": 1.7560299598438212e-05, "loss": 0.0222, "step": 73850 }, { "grad_norm": 0.9532939195632935, "learning_rate": 1.7547719061015334e-05, "loss": 0.0279, "step": 73860 }, { "grad_norm": 0.32404085993766785, "learning_rate": 1.75351420725251e-05, "loss": 0.0247, "step": 73870 }, { "grad_norm": 0.3731617331504822, "learning_rate": 1.7522568634342923e-05, "loss": 0.0296, "step": 73880 }, { "grad_norm": 0.7639005780220032, "learning_rate": 1.7509998747843825e-05, "loss": 0.0193, "step": 73890 }, { "grad_norm": 0.4246194064617157, "learning_rate": 1.7497432414402403e-05, "loss": 0.0182, "step": 73900 }, { "grad_norm": 0.40604734420776367, "learning_rate": 1.7484869635392903e-05, "loss": 0.0215, "step": 73910 }, { "grad_norm": 0.788607656955719, "learning_rate": 1.747231041218919e-05, "loss": 0.0256, "step": 73920 }, { "grad_norm": 0.34801843762397766, "learning_rate": 1.7459754746164692e-05, "loss": 0.0167, "step": 73930 }, { "grad_norm": 0.22550931572914124, "learning_rate": 1.744720263869249e-05, "loss": 0.0176, "step": 73940 }, { "grad_norm": 0.565658450126648, "learning_rate": 1.7434654091145263e-05, "loss": 0.038, "step": 73950 }, { "grad_norm": 0.27616041898727417, "learning_rate": 1.742210910489532e-05, "loss": 0.0261, "step": 73960 }, { "grad_norm": 0.2587544620037079, "learning_rate": 1.7409567681314532e-05, "loss": 0.0195, "step": 73970 }, { "grad_norm": 0.2827124297618866, "learning_rate": 1.7397029821774434e-05, "loss": 0.017, "step": 73980 }, { "grad_norm": 0.41247689723968506, "learning_rate": 1.7384495527646126e-05, "loss": 0.0215, "step": 73990 }, { "grad_norm": 0.2944759428501129, "learning_rate": 1.737196480030035e-05, "loss": 0.022, "step": 74000 }, { "grad_norm": 0.42936840653419495, "learning_rate": 1.7359437641107447e-05, "loss": 0.0208, "step": 74010 }, { "grad_norm": 0.1789734810590744, "learning_rate": 1.7346914051437387e-05, "loss": 0.021, "step": 74020 }, { "grad_norm": 0.8498557209968567, "learning_rate": 1.733439403265969e-05, "loss": 0.0201, "step": 74030 }, { "grad_norm": 0.3696625232696533, "learning_rate": 1.732187758614357e-05, "loss": 0.0177, "step": 74040 }, { "grad_norm": 0.3655829429626465, "learning_rate": 1.730936471325777e-05, "loss": 0.0214, "step": 74050 }, { "grad_norm": 0.37493574619293213, "learning_rate": 1.7296855415370718e-05, "loss": 0.0182, "step": 74060 }, { "grad_norm": 0.3057175576686859, "learning_rate": 1.728434969385035e-05, "loss": 0.022, "step": 74070 }, { "grad_norm": 0.5262454152107239, "learning_rate": 1.7271847550064346e-05, "loss": 0.0173, "step": 74080 }, { "grad_norm": 0.2440367192029953, "learning_rate": 1.725934898537987e-05, "loss": 0.031, "step": 74090 }, { "grad_norm": 0.21126773953437805, "learning_rate": 1.7246854001163783e-05, "loss": 0.0282, "step": 74100 }, { "grad_norm": 0.3911682963371277, "learning_rate": 1.7234362598782476e-05, "loss": 0.0158, "step": 74110 }, { "grad_norm": 0.37648889422416687, "learning_rate": 1.722187477960202e-05, "loss": 0.0307, "step": 74120 }, { "grad_norm": 0.31335049867630005, "learning_rate": 1.7209390544988057e-05, "loss": 0.0197, "step": 74130 }, { "grad_norm": 0.3542163074016571, "learning_rate": 1.7196909896305856e-05, "loss": 0.0188, "step": 74140 }, { "grad_norm": 0.5228849649429321, "learning_rate": 1.718443283492026e-05, "loss": 0.0222, "step": 74150 }, { "grad_norm": 0.3533417880535126, "learning_rate": 1.7171959362195746e-05, "loss": 0.0211, "step": 74160 }, { "grad_norm": 0.24476639926433563, "learning_rate": 1.715948947949642e-05, "loss": 0.014, "step": 74170 }, { "grad_norm": 0.3099297881126404, "learning_rate": 1.714702318818593e-05, "loss": 0.0166, "step": 74180 }, { "grad_norm": 0.33257341384887695, "learning_rate": 1.7134560489627592e-05, "loss": 0.0203, "step": 74190 }, { "grad_norm": 0.8749219179153442, "learning_rate": 1.712210138518431e-05, "loss": 0.0197, "step": 74200 }, { "grad_norm": 0.3660975694656372, "learning_rate": 1.71096458762186e-05, "loss": 0.0182, "step": 74210 }, { "grad_norm": 0.2539460361003876, "learning_rate": 1.7097193964092552e-05, "loss": 0.0175, "step": 74220 }, { "grad_norm": 0.4735027849674225, "learning_rate": 1.708474565016792e-05, "loss": 0.0264, "step": 74230 }, { "grad_norm": 0.34311443567276, "learning_rate": 1.7072300935806e-05, "loss": 0.018, "step": 74240 }, { "grad_norm": 0.3334561586380005, "learning_rate": 1.705985982236776e-05, "loss": 0.0174, "step": 74250 }, { "grad_norm": 0.2270449548959732, "learning_rate": 1.7047422311213697e-05, "loss": 0.0236, "step": 74260 }, { "grad_norm": 0.47107812762260437, "learning_rate": 1.703498840370401e-05, "loss": 0.0336, "step": 74270 }, { "grad_norm": 1.6846566200256348, "learning_rate": 1.702255810119842e-05, "loss": 0.0226, "step": 74280 }, { "grad_norm": 0.3672572076320648, "learning_rate": 1.7010131405056305e-05, "loss": 0.027, "step": 74290 }, { "grad_norm": 0.4539070725440979, "learning_rate": 1.6997708316636613e-05, "loss": 0.0185, "step": 74300 }, { "grad_norm": 0.5002179145812988, "learning_rate": 1.6985288837297932e-05, "loss": 0.0294, "step": 74310 }, { "grad_norm": 0.4099394381046295, "learning_rate": 1.6972872968398394e-05, "loss": 0.0239, "step": 74320 }, { "grad_norm": 0.48859903216362, "learning_rate": 1.6960460711295845e-05, "loss": 0.0234, "step": 74330 }, { "grad_norm": 0.46983271837234497, "learning_rate": 1.6948052067347623e-05, "loss": 0.018, "step": 74340 }, { "grad_norm": 0.38857993483543396, "learning_rate": 1.693564703791075e-05, "loss": 0.0289, "step": 74350 }, { "grad_norm": 0.37195444107055664, "learning_rate": 1.6923245624341793e-05, "loss": 0.04, "step": 74360 }, { "grad_norm": 0.4402916133403778, "learning_rate": 1.6910847827996962e-05, "loss": 0.0193, "step": 74370 }, { "grad_norm": 0.7035643458366394, "learning_rate": 1.689845365023206e-05, "loss": 0.0245, "step": 74380 }, { "grad_norm": 0.37612903118133545, "learning_rate": 1.6886063092402505e-05, "loss": 0.0151, "step": 74390 }, { "grad_norm": 0.1945626139640808, "learning_rate": 1.6873676155863317e-05, "loss": 0.0196, "step": 74400 }, { "grad_norm": 0.3717817962169647, "learning_rate": 1.6861292841969085e-05, "loss": 0.0168, "step": 74410 }, { "grad_norm": 0.2741537094116211, "learning_rate": 1.684891315207406e-05, "loss": 0.0163, "step": 74420 }, { "grad_norm": 0.26927125453948975, "learning_rate": 1.6836537087532034e-05, "loss": 0.0157, "step": 74430 }, { "grad_norm": 0.19109275937080383, "learning_rate": 1.6824164649696465e-05, "loss": 0.0215, "step": 74440 }, { "grad_norm": 0.40108340978622437, "learning_rate": 1.681179583992034e-05, "loss": 0.0273, "step": 74450 }, { "grad_norm": 0.4376758933067322, "learning_rate": 1.679943065955636e-05, "loss": 0.0138, "step": 74460 }, { "grad_norm": 0.2811950445175171, "learning_rate": 1.678706910995671e-05, "loss": 0.0145, "step": 74470 }, { "grad_norm": 0.549422025680542, "learning_rate": 1.6774711192473262e-05, "loss": 0.0241, "step": 74480 }, { "grad_norm": 0.3924335241317749, "learning_rate": 1.676235690845743e-05, "loss": 0.0199, "step": 74490 }, { "grad_norm": 0.273155152797699, "learning_rate": 1.6750006259260285e-05, "loss": 0.0258, "step": 74500 }, { "grad_norm": 0.5100836753845215, "learning_rate": 1.673765924623244e-05, "loss": 0.021, "step": 74510 }, { "grad_norm": 0.34829437732696533, "learning_rate": 1.67253158707242e-05, "loss": 0.0174, "step": 74520 }, { "grad_norm": 0.4139965772628784, "learning_rate": 1.671297613408537e-05, "loss": 0.0191, "step": 74530 }, { "grad_norm": 0.4839545488357544, "learning_rate": 1.6700640037665444e-05, "loss": 0.0163, "step": 74540 }, { "grad_norm": 0.4826609790325165, "learning_rate": 1.668830758281344e-05, "loss": 0.0364, "step": 74550 }, { "grad_norm": 0.18814848363399506, "learning_rate": 1.6675978770878052e-05, "loss": 0.0294, "step": 74560 }, { "grad_norm": 0.23031243681907654, "learning_rate": 1.6663653603207506e-05, "loss": 0.0172, "step": 74570 }, { "grad_norm": 0.3203931748867035, "learning_rate": 1.6651332081149685e-05, "loss": 0.0313, "step": 74580 }, { "grad_norm": 0.22282058000564575, "learning_rate": 1.6639014206052046e-05, "loss": 0.0199, "step": 74590 }, { "grad_norm": 0.4600750505924225, "learning_rate": 1.6626699979261657e-05, "loss": 0.0177, "step": 74600 }, { "grad_norm": 0.5391849875450134, "learning_rate": 1.6614389402125197e-05, "loss": 0.0207, "step": 74610 }, { "grad_norm": 0.9869552850723267, "learning_rate": 1.66020824759889e-05, "loss": 0.024, "step": 74620 }, { "grad_norm": 0.39794304966926575, "learning_rate": 1.6589779202198675e-05, "loss": 0.022, "step": 74630 }, { "grad_norm": 0.47258996963500977, "learning_rate": 1.6577479582099924e-05, "loss": 0.0316, "step": 74640 }, { "grad_norm": 0.9820029139518738, "learning_rate": 1.6565183617037788e-05, "loss": 0.026, "step": 74650 }, { "grad_norm": 0.6535501480102539, "learning_rate": 1.6552891308356894e-05, "loss": 0.0201, "step": 74660 }, { "grad_norm": 0.4846273362636566, "learning_rate": 1.6540602657401526e-05, "loss": 0.0252, "step": 74670 }, { "grad_norm": 0.3597504198551178, "learning_rate": 1.6528317665515537e-05, "loss": 0.022, "step": 74680 }, { "grad_norm": 0.7187293767929077, "learning_rate": 1.6516036334042416e-05, "loss": 0.0175, "step": 74690 }, { "grad_norm": 0.3026805818080902, "learning_rate": 1.650375866432519e-05, "loss": 0.0204, "step": 74700 }, { "grad_norm": 0.5906566977500916, "learning_rate": 1.649148465770659e-05, "loss": 0.0344, "step": 74710 }, { "grad_norm": 0.22585229575634003, "learning_rate": 1.647921431552883e-05, "loss": 0.0191, "step": 74720 }, { "grad_norm": 0.16968132555484772, "learning_rate": 1.646694763913381e-05, "loss": 0.0229, "step": 74730 }, { "grad_norm": 0.3391079008579254, "learning_rate": 1.6454684629862966e-05, "loss": 0.0179, "step": 74740 }, { "grad_norm": 0.41546446084976196, "learning_rate": 1.6442425289057388e-05, "loss": 0.0253, "step": 74750 }, { "grad_norm": 0.366638720035553, "learning_rate": 1.6430169618057718e-05, "loss": 0.018, "step": 74760 }, { "grad_norm": 0.1613597720861435, "learning_rate": 1.641791761820422e-05, "loss": 0.0174, "step": 74770 }, { "grad_norm": 0.583036482334137, "learning_rate": 1.640566929083676e-05, "loss": 0.019, "step": 74780 }, { "grad_norm": 0.6178756952285767, "learning_rate": 1.6393424637294823e-05, "loss": 0.0185, "step": 74790 }, { "grad_norm": 0.2854067087173462, "learning_rate": 1.638118365891742e-05, "loss": 0.0212, "step": 74800 }, { "grad_norm": 0.3254298269748688, "learning_rate": 1.636894635704323e-05, "loss": 0.0191, "step": 74810 }, { "grad_norm": 0.49114829301834106, "learning_rate": 1.6356712733010517e-05, "loss": 0.0244, "step": 74820 }, { "grad_norm": 0.9060371518135071, "learning_rate": 1.6344482788157096e-05, "loss": 0.0217, "step": 74830 }, { "grad_norm": 0.4373464584350586, "learning_rate": 1.6332256523820444e-05, "loss": 0.0185, "step": 74840 }, { "grad_norm": 0.35338565707206726, "learning_rate": 1.6320033941337597e-05, "loss": 0.0243, "step": 74850 }, { "grad_norm": 0.3212287724018097, "learning_rate": 1.6307815042045216e-05, "loss": 0.0161, "step": 74860 }, { "grad_norm": 0.5483728051185608, "learning_rate": 1.6295599827279513e-05, "loss": 0.0436, "step": 74870 }, { "grad_norm": 0.20949766039848328, "learning_rate": 1.6283388298376356e-05, "loss": 0.0229, "step": 74880 }, { "grad_norm": 0.4155563712120056, "learning_rate": 1.6271180456671125e-05, "loss": 0.0256, "step": 74890 }, { "grad_norm": 0.3870023190975189, "learning_rate": 1.625897630349893e-05, "loss": 0.034, "step": 74900 }, { "grad_norm": 0.4245546758174896, "learning_rate": 1.6246775840194335e-05, "loss": 0.0227, "step": 74910 }, { "grad_norm": 0.42345619201660156, "learning_rate": 1.6234579068091604e-05, "loss": 0.017, "step": 74920 }, { "grad_norm": 0.4890103340148926, "learning_rate": 1.6222385988524524e-05, "loss": 0.0414, "step": 74930 }, { "grad_norm": 0.4363817572593689, "learning_rate": 1.6210196602826544e-05, "loss": 0.0337, "step": 74940 }, { "grad_norm": 0.7508004903793335, "learning_rate": 1.6198010912330646e-05, "loss": 0.0184, "step": 74950 }, { "grad_norm": 0.4921696186065674, "learning_rate": 1.6185828918369456e-05, "loss": 0.0221, "step": 74960 }, { "grad_norm": 0.5289285778999329, "learning_rate": 1.617365062227517e-05, "loss": 0.0297, "step": 74970 }, { "grad_norm": 0.6878465414047241, "learning_rate": 1.616147602537962e-05, "loss": 0.0283, "step": 74980 }, { "grad_norm": 0.4422570466995239, "learning_rate": 1.614930512901415e-05, "loss": 0.0212, "step": 74990 }, { "grad_norm": 0.5762587189674377, "learning_rate": 1.6137137934509804e-05, "loss": 0.0224, "step": 75000 }, { "grad_norm": 0.30514827370643616, "learning_rate": 1.6124974443197123e-05, "loss": 0.0267, "step": 75010 }, { "grad_norm": 0.7109375596046448, "learning_rate": 1.6112814656406306e-05, "loss": 0.0288, "step": 75020 }, { "grad_norm": 0.3936718702316284, "learning_rate": 1.6100658575467125e-05, "loss": 0.0271, "step": 75030 }, { "grad_norm": 0.47593367099761963, "learning_rate": 1.6088506201708974e-05, "loss": 0.0296, "step": 75040 }, { "grad_norm": 0.24084968864917755, "learning_rate": 1.6076357536460785e-05, "loss": 0.0189, "step": 75050 }, { "grad_norm": 0.4355616271495819, "learning_rate": 1.606421258105113e-05, "loss": 0.0236, "step": 75060 }, { "grad_norm": 0.4018464684486389, "learning_rate": 1.6052071336808183e-05, "loss": 0.0242, "step": 75070 }, { "grad_norm": 0.4228500425815582, "learning_rate": 1.6039933805059658e-05, "loss": 0.0228, "step": 75080 }, { "grad_norm": 0.22093696892261505, "learning_rate": 1.6027799987132915e-05, "loss": 0.0281, "step": 75090 }, { "grad_norm": 0.30493104457855225, "learning_rate": 1.6015669884354888e-05, "loss": 0.0263, "step": 75100 }, { "grad_norm": 0.2856813967227936, "learning_rate": 1.6003543498052125e-05, "loss": 0.0231, "step": 75110 }, { "grad_norm": 0.954839289188385, "learning_rate": 1.5991420829550715e-05, "loss": 0.0384, "step": 75120 }, { "grad_norm": 0.6883605122566223, "learning_rate": 1.5979301880176405e-05, "loss": 0.0256, "step": 75130 }, { "grad_norm": 0.3266308009624481, "learning_rate": 1.5967186651254485e-05, "loss": 0.0237, "step": 75140 }, { "grad_norm": 0.2866412103176117, "learning_rate": 1.5955075144109853e-05, "loss": 0.0221, "step": 75150 }, { "grad_norm": 0.1273774355649948, "learning_rate": 1.594296736006703e-05, "loss": 0.0148, "step": 75160 }, { "grad_norm": 0.8139450550079346, "learning_rate": 1.59308633004501e-05, "loss": 0.0169, "step": 75170 }, { "grad_norm": 0.43833306431770325, "learning_rate": 1.5918762966582724e-05, "loss": 0.0297, "step": 75180 }, { "grad_norm": 0.2425450086593628, "learning_rate": 1.5906666359788203e-05, "loss": 0.0183, "step": 75190 }, { "grad_norm": 0.5167868733406067, "learning_rate": 1.589457348138938e-05, "loss": 0.0202, "step": 75200 }, { "grad_norm": 0.45177939534187317, "learning_rate": 1.588248433270872e-05, "loss": 0.0195, "step": 75210 }, { "grad_norm": 0.2635093331336975, "learning_rate": 1.587039891506828e-05, "loss": 0.0188, "step": 75220 }, { "grad_norm": 0.2738029658794403, "learning_rate": 1.5858317229789715e-05, "loss": 0.0214, "step": 75230 }, { "grad_norm": 0.3261612355709076, "learning_rate": 1.584623927819423e-05, "loss": 0.0199, "step": 75240 }, { "grad_norm": 0.3253129720687866, "learning_rate": 1.583416506160269e-05, "loss": 0.0161, "step": 75250 }, { "grad_norm": 0.5171202421188354, "learning_rate": 1.5822094581335472e-05, "loss": 0.0241, "step": 75260 }, { "grad_norm": 0.34994596242904663, "learning_rate": 1.5810027838712604e-05, "loss": 0.019, "step": 75270 }, { "grad_norm": 0.3151814937591553, "learning_rate": 1.5797964835053687e-05, "loss": 0.0189, "step": 75280 }, { "grad_norm": 0.4448288381099701, "learning_rate": 1.5785905571677933e-05, "loss": 0.0215, "step": 75290 }, { "grad_norm": 0.3059892952442169, "learning_rate": 1.5773850049904086e-05, "loss": 0.0176, "step": 75300 }, { "grad_norm": 0.8547743558883667, "learning_rate": 1.5761798271050543e-05, "loss": 0.0371, "step": 75310 }, { "grad_norm": 0.4914465546607971, "learning_rate": 1.5749750236435277e-05, "loss": 0.0238, "step": 75320 }, { "grad_norm": 0.27626603841781616, "learning_rate": 1.5737705947375815e-05, "loss": 0.0166, "step": 75330 }, { "grad_norm": 0.45444250106811523, "learning_rate": 1.5725665405189322e-05, "loss": 0.0188, "step": 75340 }, { "grad_norm": 0.2712768018245697, "learning_rate": 1.571362861119253e-05, "loss": 0.0247, "step": 75350 }, { "grad_norm": 0.4931786358356476, "learning_rate": 1.570159556670177e-05, "loss": 0.0251, "step": 75360 }, { "grad_norm": 0.3251335024833679, "learning_rate": 1.5689566273032946e-05, "loss": 0.0281, "step": 75370 }, { "grad_norm": 0.47240597009658813, "learning_rate": 1.5677540731501583e-05, "loss": 0.0231, "step": 75380 }, { "grad_norm": 0.2737060785293579, "learning_rate": 1.5665518943422746e-05, "loss": 0.0185, "step": 75390 }, { "grad_norm": 0.2914453148841858, "learning_rate": 1.5653500910111134e-05, "loss": 0.0188, "step": 75400 }, { "grad_norm": 0.19116416573524475, "learning_rate": 1.5641486632881023e-05, "loss": 0.0235, "step": 75410 }, { "grad_norm": 0.46979790925979614, "learning_rate": 1.56294761130463e-05, "loss": 0.0194, "step": 75420 }, { "grad_norm": 0.231531023979187, "learning_rate": 1.5617469351920373e-05, "loss": 0.024, "step": 75430 }, { "grad_norm": 0.6397873163223267, "learning_rate": 1.5605466350816315e-05, "loss": 0.0243, "step": 75440 }, { "grad_norm": 0.2535801827907562, "learning_rate": 1.559346711104674e-05, "loss": 0.0187, "step": 75450 }, { "grad_norm": 0.23525185883045197, "learning_rate": 1.5581471633923883e-05, "loss": 0.0199, "step": 75460 }, { "grad_norm": 0.38195013999938965, "learning_rate": 1.5569479920759505e-05, "loss": 0.0234, "step": 75470 }, { "grad_norm": 0.2355632334947586, "learning_rate": 1.5557491972865072e-05, "loss": 0.0163, "step": 75480 }, { "grad_norm": 0.28340619802474976, "learning_rate": 1.554550779155152e-05, "loss": 0.025, "step": 75490 }, { "grad_norm": 0.388464093208313, "learning_rate": 1.5533527378129453e-05, "loss": 0.0244, "step": 75500 }, { "grad_norm": 0.6359161138534546, "learning_rate": 1.552155073390899e-05, "loss": 0.0443, "step": 75510 }, { "grad_norm": 0.7216911911964417, "learning_rate": 1.5509577860199903e-05, "loss": 0.0235, "step": 75520 }, { "grad_norm": 0.3296290338039398, "learning_rate": 1.5497608758311525e-05, "loss": 0.0209, "step": 75530 }, { "grad_norm": 0.6884995102882385, "learning_rate": 1.5485643429552797e-05, "loss": 0.0194, "step": 75540 }, { "grad_norm": 0.21889181435108185, "learning_rate": 1.5473681875232197e-05, "loss": 0.014, "step": 75550 }, { "grad_norm": 0.5101158618927002, "learning_rate": 1.5461724096657837e-05, "loss": 0.0172, "step": 75560 }, { "grad_norm": 0.43569791316986084, "learning_rate": 1.5449770095137416e-05, "loss": 0.018, "step": 75570 }, { "grad_norm": 0.32186761498451233, "learning_rate": 1.5437819871978178e-05, "loss": 0.0206, "step": 75580 }, { "grad_norm": 0.3398561179637909, "learning_rate": 1.5425873428486992e-05, "loss": 0.0174, "step": 75590 }, { "grad_norm": 0.15487462282180786, "learning_rate": 1.5413930765970312e-05, "loss": 0.0169, "step": 75600 }, { "grad_norm": 0.5438726544380188, "learning_rate": 1.5401991885734175e-05, "loss": 0.0293, "step": 75610 }, { "grad_norm": 0.4104127883911133, "learning_rate": 1.5390056789084174e-05, "loss": 0.0226, "step": 75620 }, { "grad_norm": 0.5696563124656677, "learning_rate": 1.5378125477325545e-05, "loss": 0.0267, "step": 75630 }, { "grad_norm": 0.4168838858604431, "learning_rate": 1.5366197951763046e-05, "loss": 0.0289, "step": 75640 }, { "grad_norm": 0.17015990614891052, "learning_rate": 1.5354274213701075e-05, "loss": 0.0166, "step": 75650 }, { "grad_norm": 0.3679541349411011, "learning_rate": 1.5342354264443566e-05, "loss": 0.0165, "step": 75660 }, { "grad_norm": 0.32388365268707275, "learning_rate": 1.533043810529411e-05, "loss": 0.0331, "step": 75670 }, { "grad_norm": 0.5720617175102234, "learning_rate": 1.5318525737555807e-05, "loss": 0.0281, "step": 75680 }, { "grad_norm": 0.2911914885044098, "learning_rate": 1.53066171625314e-05, "loss": 0.0227, "step": 75690 }, { "grad_norm": 0.2455240786075592, "learning_rate": 1.529471238152317e-05, "loss": 0.0153, "step": 75700 }, { "grad_norm": 0.8499547839164734, "learning_rate": 1.5282811395833023e-05, "loss": 0.0223, "step": 75710 }, { "grad_norm": 0.6139523386955261, "learning_rate": 1.5270914206762397e-05, "loss": 0.0193, "step": 75720 }, { "grad_norm": 0.2005351036787033, "learning_rate": 1.525902081561241e-05, "loss": 0.0162, "step": 75730 }, { "grad_norm": 0.43424344062805176, "learning_rate": 1.5247131223683664e-05, "loss": 0.0201, "step": 75740 }, { "grad_norm": 0.18632932007312775, "learning_rate": 1.5235245432276408e-05, "loss": 0.031, "step": 75750 }, { "grad_norm": 0.44852232933044434, "learning_rate": 1.5223363442690424e-05, "loss": 0.02, "step": 75760 }, { "grad_norm": 0.3954218924045563, "learning_rate": 1.5211485256225133e-05, "loss": 0.0188, "step": 75770 }, { "grad_norm": 0.20815524458885193, "learning_rate": 1.519961087417951e-05, "loss": 0.021, "step": 75780 }, { "grad_norm": 0.292512446641922, "learning_rate": 1.5187740297852137e-05, "loss": 0.0159, "step": 75790 }, { "grad_norm": 0.39369529485702515, "learning_rate": 1.517587352854113e-05, "loss": 0.0217, "step": 75800 }, { "grad_norm": 0.2324799746274948, "learning_rate": 1.516401056754423e-05, "loss": 0.0245, "step": 75810 }, { "grad_norm": 0.21704283356666565, "learning_rate": 1.5152151416158783e-05, "loss": 0.0145, "step": 75820 }, { "grad_norm": 0.31121116876602173, "learning_rate": 1.5140296075681642e-05, "loss": 0.0154, "step": 75830 }, { "grad_norm": 0.1770056039094925, "learning_rate": 1.5128444547409331e-05, "loss": 0.0191, "step": 75840 }, { "grad_norm": 0.7376454472541809, "learning_rate": 1.5116596832637864e-05, "loss": 0.0229, "step": 75850 }, { "grad_norm": 0.21798136830329895, "learning_rate": 1.5104752932662952e-05, "loss": 0.0251, "step": 75860 }, { "grad_norm": 0.36114197969436646, "learning_rate": 1.5092912848779778e-05, "loss": 0.02, "step": 75870 }, { "grad_norm": 0.2910798490047455, "learning_rate": 1.5081076582283194e-05, "loss": 0.0215, "step": 75880 }, { "grad_norm": 0.5539753437042236, "learning_rate": 1.5069244134467553e-05, "loss": 0.0232, "step": 75890 }, { "grad_norm": 0.464028000831604, "learning_rate": 1.5057415506626876e-05, "loss": 0.0283, "step": 75900 }, { "grad_norm": 0.4476892650127411, "learning_rate": 1.5045590700054668e-05, "loss": 0.0159, "step": 75910 }, { "grad_norm": 0.431387335062027, "learning_rate": 1.5033769716044144e-05, "loss": 0.0224, "step": 75920 }, { "grad_norm": 0.5450536012649536, "learning_rate": 1.5021952555887975e-05, "loss": 0.0269, "step": 75930 }, { "grad_norm": 0.5016350746154785, "learning_rate": 1.50101392208785e-05, "loss": 0.0175, "step": 75940 }, { "grad_norm": 0.3189755976200104, "learning_rate": 1.4998329712307579e-05, "loss": 0.0191, "step": 75950 }, { "grad_norm": 0.4879720211029053, "learning_rate": 1.4986524031466708e-05, "loss": 0.0264, "step": 75960 }, { "grad_norm": 0.49647948145866394, "learning_rate": 1.4974722179646916e-05, "loss": 0.0166, "step": 75970 }, { "grad_norm": 0.7842943668365479, "learning_rate": 1.4962924158138847e-05, "loss": 0.0175, "step": 75980 }, { "grad_norm": 0.5301392674446106, "learning_rate": 1.4951129968232714e-05, "loss": 0.0371, "step": 75990 }, { "grad_norm": 0.24311420321464539, "learning_rate": 1.4939339611218329e-05, "loss": 0.0254, "step": 76000 }, { "grad_norm": 0.3020879924297333, "learning_rate": 1.4927553088385032e-05, "loss": 0.0211, "step": 76010 }, { "grad_norm": 0.31915709376335144, "learning_rate": 1.4915770401021801e-05, "loss": 0.0282, "step": 76020 }, { "grad_norm": 0.49870720505714417, "learning_rate": 1.4903991550417185e-05, "loss": 0.019, "step": 76030 }, { "grad_norm": 0.1881127953529358, "learning_rate": 1.4892216537859277e-05, "loss": 0.0239, "step": 76040 }, { "grad_norm": 0.1957886964082718, "learning_rate": 1.4880445364635775e-05, "loss": 0.0304, "step": 76050 }, { "grad_norm": 0.4076918661594391, "learning_rate": 1.4868678032033973e-05, "loss": 0.0177, "step": 76060 }, { "grad_norm": 0.35097458958625793, "learning_rate": 1.4856914541340739e-05, "loss": 0.0186, "step": 76070 }, { "grad_norm": 0.23600223660469055, "learning_rate": 1.4845154893842472e-05, "loss": 0.0173, "step": 76080 }, { "grad_norm": 0.2386808693408966, "learning_rate": 1.4833399090825234e-05, "loss": 0.0437, "step": 76090 }, { "grad_norm": 0.274315744638443, "learning_rate": 1.4821647133574562e-05, "loss": 0.0246, "step": 76100 }, { "grad_norm": 0.3421327769756317, "learning_rate": 1.4809899023375706e-05, "loss": 0.021, "step": 76110 }, { "grad_norm": 0.4125916659832001, "learning_rate": 1.4798154761513366e-05, "loss": 0.0219, "step": 76120 }, { "grad_norm": 0.2718330919742584, "learning_rate": 1.4786414349271916e-05, "loss": 0.0158, "step": 76130 }, { "grad_norm": 0.40705788135528564, "learning_rate": 1.4774677787935226e-05, "loss": 0.0184, "step": 76140 }, { "grad_norm": 0.3358761966228485, "learning_rate": 1.4762945078786838e-05, "loss": 0.0267, "step": 76150 }, { "grad_norm": 0.26043930649757385, "learning_rate": 1.4751216223109777e-05, "loss": 0.0195, "step": 76160 }, { "grad_norm": 0.5617281198501587, "learning_rate": 1.4739491222186714e-05, "loss": 0.0149, "step": 76170 }, { "grad_norm": 0.45079153776168823, "learning_rate": 1.4727770077299879e-05, "loss": 0.0227, "step": 76180 }, { "grad_norm": 0.2823069989681244, "learning_rate": 1.4716052789731094e-05, "loss": 0.026, "step": 76190 }, { "grad_norm": 0.2854142189025879, "learning_rate": 1.470433936076171e-05, "loss": 0.0239, "step": 76200 }, { "grad_norm": 0.18138782680034637, "learning_rate": 1.4692629791672719e-05, "loss": 0.015, "step": 76210 }, { "grad_norm": 0.5466282963752747, "learning_rate": 1.468092408374464e-05, "loss": 0.0262, "step": 76220 }, { "grad_norm": 0.3807937502861023, "learning_rate": 1.4669222238257602e-05, "loss": 0.0234, "step": 76230 }, { "grad_norm": 0.2373700588941574, "learning_rate": 1.4657524256491301e-05, "loss": 0.0207, "step": 76240 }, { "grad_norm": 0.5481625199317932, "learning_rate": 1.464583013972502e-05, "loss": 0.019, "step": 76250 }, { "grad_norm": 0.4443565011024475, "learning_rate": 1.4634139889237591e-05, "loss": 0.0352, "step": 76260 }, { "grad_norm": 0.6386887431144714, "learning_rate": 1.4622453506307448e-05, "loss": 0.0266, "step": 76270 }, { "grad_norm": 0.8637030720710754, "learning_rate": 1.461077099221262e-05, "loss": 0.038, "step": 76280 }, { "grad_norm": 0.18240152299404144, "learning_rate": 1.4599092348230652e-05, "loss": 0.0172, "step": 76290 }, { "grad_norm": 0.41240590810775757, "learning_rate": 1.4587417575638724e-05, "loss": 0.0156, "step": 76300 }, { "grad_norm": 0.3565059304237366, "learning_rate": 1.4575746675713563e-05, "loss": 0.0296, "step": 76310 }, { "grad_norm": 0.48642757534980774, "learning_rate": 1.4564079649731504e-05, "loss": 0.0187, "step": 76320 }, { "grad_norm": 0.19585999846458435, "learning_rate": 1.4552416498968402e-05, "loss": 0.0256, "step": 76330 }, { "grad_norm": 0.5441451072692871, "learning_rate": 1.4540757224699747e-05, "loss": 0.0199, "step": 76340 }, { "grad_norm": 0.4099462330341339, "learning_rate": 1.452910182820056e-05, "loss": 0.0255, "step": 76350 }, { "grad_norm": 0.32798922061920166, "learning_rate": 1.4517450310745468e-05, "loss": 0.019, "step": 76360 }, { "grad_norm": 0.34445878863334656, "learning_rate": 1.450580267360866e-05, "loss": 0.0338, "step": 76370 }, { "grad_norm": 0.33478498458862305, "learning_rate": 1.4494158918063922e-05, "loss": 0.021, "step": 76380 }, { "grad_norm": 0.22781863808631897, "learning_rate": 1.4482519045384568e-05, "loss": 0.0173, "step": 76390 }, { "grad_norm": 0.501727819442749, "learning_rate": 1.4470883056843548e-05, "loss": 0.0219, "step": 76400 }, { "grad_norm": 0.22177636623382568, "learning_rate": 1.4459250953713322e-05, "loss": 0.0179, "step": 76410 }, { "grad_norm": 0.5431949496269226, "learning_rate": 1.444762273726598e-05, "loss": 0.0226, "step": 76420 }, { "grad_norm": 0.46480870246887207, "learning_rate": 1.443599840877316e-05, "loss": 0.0172, "step": 76430 }, { "grad_norm": 0.36512258648872375, "learning_rate": 1.4424377969506103e-05, "loss": 0.0201, "step": 76440 }, { "grad_norm": 0.2794596254825592, "learning_rate": 1.4412761420735565e-05, "loss": 0.0185, "step": 76450 }, { "grad_norm": 0.9180408716201782, "learning_rate": 1.4401148763731953e-05, "loss": 0.0476, "step": 76460 }, { "grad_norm": 0.31190165877342224, "learning_rate": 1.4389539999765173e-05, "loss": 0.0167, "step": 76470 }, { "grad_norm": 0.34678763151168823, "learning_rate": 1.4377935130104758e-05, "loss": 0.0232, "step": 76480 }, { "grad_norm": 0.3556163012981415, "learning_rate": 1.4366334156019801e-05, "loss": 0.0188, "step": 76490 }, { "grad_norm": 0.2628171145915985, "learning_rate": 1.435473707877898e-05, "loss": 0.0147, "step": 76500 }, { "grad_norm": 0.30020853877067566, "learning_rate": 1.4343143899650508e-05, "loss": 0.0152, "step": 76510 }, { "grad_norm": 0.6009678840637207, "learning_rate": 1.4331554619902204e-05, "loss": 0.0248, "step": 76520 }, { "grad_norm": 0.18434885144233704, "learning_rate": 1.4319969240801484e-05, "loss": 0.0211, "step": 76530 }, { "grad_norm": 0.261759877204895, "learning_rate": 1.430838776361526e-05, "loss": 0.0202, "step": 76540 }, { "grad_norm": 0.4175730049610138, "learning_rate": 1.4296810189610093e-05, "loss": 0.0144, "step": 76550 }, { "grad_norm": 0.5347710847854614, "learning_rate": 1.4285236520052087e-05, "loss": 0.0292, "step": 76560 }, { "grad_norm": 0.31205853819847107, "learning_rate": 1.4273666756206933e-05, "loss": 0.0165, "step": 76570 }, { "grad_norm": 0.2698727250099182, "learning_rate": 1.4262100899339848e-05, "loss": 0.0185, "step": 76580 }, { "grad_norm": 0.6013822555541992, "learning_rate": 1.4250538950715696e-05, "loss": 0.0257, "step": 76590 }, { "grad_norm": 0.570339560508728, "learning_rate": 1.4238980911598842e-05, "loss": 0.0118, "step": 76600 }, { "grad_norm": 0.5763193964958191, "learning_rate": 1.4227426783253267e-05, "loss": 0.0198, "step": 76610 }, { "grad_norm": 0.9932509660720825, "learning_rate": 1.4215876566942521e-05, "loss": 0.0268, "step": 76620 }, { "grad_norm": 0.36569809913635254, "learning_rate": 1.4204330263929721e-05, "loss": 0.0309, "step": 76630 }, { "grad_norm": 0.2703154683113098, "learning_rate": 1.4192787875477537e-05, "loss": 0.0127, "step": 76640 }, { "grad_norm": 0.3786506652832031, "learning_rate": 1.4181249402848246e-05, "loss": 0.0132, "step": 76650 }, { "grad_norm": 0.5263581871986389, "learning_rate": 1.4169714847303656e-05, "loss": 0.0203, "step": 76660 }, { "grad_norm": 0.6342873573303223, "learning_rate": 1.4158184210105197e-05, "loss": 0.0248, "step": 76670 }, { "grad_norm": 0.3388962745666504, "learning_rate": 1.4146657492513787e-05, "loss": 0.0207, "step": 76680 }, { "grad_norm": 0.6855401992797852, "learning_rate": 1.4135134695790047e-05, "loss": 0.0171, "step": 76690 }, { "grad_norm": 0.5881219506263733, "learning_rate": 1.4123615821194037e-05, "loss": 0.0165, "step": 76700 }, { "grad_norm": 0.6139965057373047, "learning_rate": 1.4112100869985473e-05, "loss": 0.0198, "step": 76710 }, { "grad_norm": 0.40300920605659485, "learning_rate": 1.4100589843423589e-05, "loss": 0.0155, "step": 76720 }, { "grad_norm": 0.3750699758529663, "learning_rate": 1.408908274276723e-05, "loss": 0.0211, "step": 76730 }, { "grad_norm": 0.30525344610214233, "learning_rate": 1.4077579569274786e-05, "loss": 0.0092, "step": 76740 }, { "grad_norm": 0.14637978374958038, "learning_rate": 1.4066080324204245e-05, "loss": 0.0245, "step": 76750 }, { "grad_norm": 0.40306490659713745, "learning_rate": 1.405458500881312e-05, "loss": 0.0198, "step": 76760 }, { "grad_norm": 0.254830539226532, "learning_rate": 1.4043093624358538e-05, "loss": 0.0222, "step": 76770 }, { "grad_norm": 0.33593055605888367, "learning_rate": 1.4031606172097189e-05, "loss": 0.0239, "step": 76780 }, { "grad_norm": 0.471671462059021, "learning_rate": 1.40201226532853e-05, "loss": 0.0275, "step": 76790 }, { "grad_norm": 0.4567040801048279, "learning_rate": 1.4008643069178701e-05, "loss": 0.0249, "step": 76800 }, { "grad_norm": 0.3419703543186188, "learning_rate": 1.399716742103278e-05, "loss": 0.0199, "step": 76810 }, { "grad_norm": 0.42674800753593445, "learning_rate": 1.398569571010252e-05, "loss": 0.0191, "step": 76820 }, { "grad_norm": 0.41016340255737305, "learning_rate": 1.3974227937642414e-05, "loss": 0.0319, "step": 76830 }, { "grad_norm": 0.3375098705291748, "learning_rate": 1.3962764104906595e-05, "loss": 0.0244, "step": 76840 }, { "grad_norm": 0.135409876704216, "learning_rate": 1.3951304213148696e-05, "loss": 0.0239, "step": 76850 }, { "grad_norm": 0.35079339146614075, "learning_rate": 1.3939848263621979e-05, "loss": 0.0279, "step": 76860 }, { "grad_norm": 0.2604023814201355, "learning_rate": 1.3928396257579219e-05, "loss": 0.0147, "step": 76870 }, { "grad_norm": 0.5392560362815857, "learning_rate": 1.3916948196272833e-05, "loss": 0.0336, "step": 76880 }, { "grad_norm": 0.28073030710220337, "learning_rate": 1.3905504080954734e-05, "loss": 0.0146, "step": 76890 }, { "grad_norm": 0.4863203167915344, "learning_rate": 1.3894063912876454e-05, "loss": 0.0265, "step": 76900 }, { "grad_norm": 0.2848518490791321, "learning_rate": 1.3882627693289046e-05, "loss": 0.0169, "step": 76910 }, { "grad_norm": 0.30679309368133545, "learning_rate": 1.3871195423443183e-05, "loss": 0.0171, "step": 76920 }, { "grad_norm": 0.23842759430408478, "learning_rate": 1.3859767104589044e-05, "loss": 0.0185, "step": 76930 }, { "grad_norm": 0.42856040596961975, "learning_rate": 1.3848342737976467e-05, "loss": 0.0287, "step": 76940 }, { "grad_norm": 0.6657617092132568, "learning_rate": 1.3836922324854757e-05, "loss": 0.0209, "step": 76950 }, { "grad_norm": 0.34231045842170715, "learning_rate": 1.382550586647287e-05, "loss": 0.0252, "step": 76960 }, { "grad_norm": 0.8186041116714478, "learning_rate": 1.3814093364079257e-05, "loss": 0.044, "step": 76970 }, { "grad_norm": 0.27587318420410156, "learning_rate": 1.3802684818921991e-05, "loss": 0.0158, "step": 76980 }, { "grad_norm": 0.19592070579528809, "learning_rate": 1.3791280232248687e-05, "loss": 0.0176, "step": 76990 }, { "grad_norm": 0.3814437687397003, "learning_rate": 1.3779879605306555e-05, "loss": 0.0151, "step": 77000 }, { "grad_norm": 0.4325672686100006, "learning_rate": 1.376848293934232e-05, "loss": 0.0218, "step": 77010 }, { "grad_norm": 0.42788049578666687, "learning_rate": 1.3757090235602316e-05, "loss": 0.0219, "step": 77020 }, { "grad_norm": 0.5910475254058838, "learning_rate": 1.3745701495332447e-05, "loss": 0.0278, "step": 77030 }, { "grad_norm": 0.3363134562969208, "learning_rate": 1.3734316719778145e-05, "loss": 0.0209, "step": 77040 }, { "grad_norm": 0.2514426112174988, "learning_rate": 1.3722935910184454e-05, "loss": 0.0205, "step": 77050 }, { "grad_norm": 0.47120964527130127, "learning_rate": 1.3711559067795921e-05, "loss": 0.0178, "step": 77060 }, { "grad_norm": 0.3369300067424774, "learning_rate": 1.3700186193856756e-05, "loss": 0.0228, "step": 77070 }, { "grad_norm": 0.3112722337245941, "learning_rate": 1.368881728961064e-05, "loss": 0.0212, "step": 77080 }, { "grad_norm": 0.48185276985168457, "learning_rate": 1.3677452356300885e-05, "loss": 0.0171, "step": 77090 }, { "grad_norm": 0.4859413504600525, "learning_rate": 1.3666091395170321e-05, "loss": 0.0229, "step": 77100 }, { "grad_norm": 0.4369521737098694, "learning_rate": 1.3654734407461384e-05, "loss": 0.0185, "step": 77110 }, { "grad_norm": 0.5760061740875244, "learning_rate": 1.3643381394416022e-05, "loss": 0.0199, "step": 77120 }, { "grad_norm": 0.5186681747436523, "learning_rate": 1.3632032357275838e-05, "loss": 0.0199, "step": 77130 }, { "grad_norm": 0.36620205640792847, "learning_rate": 1.3620687297281898e-05, "loss": 0.0224, "step": 77140 }, { "grad_norm": 0.29937678575515747, "learning_rate": 1.3609346215674917e-05, "loss": 0.0189, "step": 77150 }, { "grad_norm": 0.561477541923523, "learning_rate": 1.3598009113695109e-05, "loss": 0.0256, "step": 77160 }, { "grad_norm": 0.3231104910373688, "learning_rate": 1.358667599258231e-05, "loss": 0.0205, "step": 77170 }, { "grad_norm": 0.3714790940284729, "learning_rate": 1.357534685357586e-05, "loss": 0.016, "step": 77180 }, { "grad_norm": 0.2335057556629181, "learning_rate": 1.3564021697914715e-05, "loss": 0.0173, "step": 77190 }, { "grad_norm": 0.34248074889183044, "learning_rate": 1.3552700526837381e-05, "loss": 0.0323, "step": 77200 }, { "grad_norm": 0.364173948764801, "learning_rate": 1.3541383341581925e-05, "loss": 0.0191, "step": 77210 }, { "grad_norm": 0.35015374422073364, "learning_rate": 1.3530070143385965e-05, "loss": 0.0246, "step": 77220 }, { "grad_norm": 0.44022274017333984, "learning_rate": 1.3518760933486702e-05, "loss": 0.0171, "step": 77230 }, { "grad_norm": 0.3265122175216675, "learning_rate": 1.3507455713120908e-05, "loss": 0.0215, "step": 77240 }, { "grad_norm": 0.3577278256416321, "learning_rate": 1.3496154483524876e-05, "loss": 0.0274, "step": 77250 }, { "grad_norm": 0.5423455238342285, "learning_rate": 1.3484857245934506e-05, "loss": 0.0232, "step": 77260 }, { "grad_norm": 0.8486266136169434, "learning_rate": 1.3473564001585248e-05, "loss": 0.0194, "step": 77270 }, { "grad_norm": 0.5676456093788147, "learning_rate": 1.3462274751712129e-05, "loss": 0.0183, "step": 77280 }, { "grad_norm": 0.4080842435359955, "learning_rate": 1.3450989497549698e-05, "loss": 0.0214, "step": 77290 }, { "grad_norm": 0.728697657585144, "learning_rate": 1.3439708240332115e-05, "loss": 0.0215, "step": 77300 }, { "grad_norm": 0.2945285141468048, "learning_rate": 1.3428430981293043e-05, "loss": 0.0246, "step": 77310 }, { "grad_norm": 0.16181036829948425, "learning_rate": 1.3417157721665808e-05, "loss": 0.0155, "step": 77320 }, { "grad_norm": 0.1885177046060562, "learning_rate": 1.3405888462683186e-05, "loss": 0.0245, "step": 77330 }, { "grad_norm": 0.21580864489078522, "learning_rate": 1.3394623205577595e-05, "loss": 0.0136, "step": 77340 }, { "grad_norm": 0.33045607805252075, "learning_rate": 1.3383361951580963e-05, "loss": 0.0227, "step": 77350 }, { "grad_norm": 0.5871177911758423, "learning_rate": 1.337210470192483e-05, "loss": 0.0254, "step": 77360 }, { "grad_norm": 0.27916011214256287, "learning_rate": 1.3360851457840245e-05, "loss": 0.0197, "step": 77370 }, { "grad_norm": 0.35461312532424927, "learning_rate": 1.3349602220557856e-05, "loss": 0.0235, "step": 77380 }, { "grad_norm": 0.4757864475250244, "learning_rate": 1.3338356991307865e-05, "loss": 0.0208, "step": 77390 }, { "grad_norm": 0.40887215733528137, "learning_rate": 1.3327115771320048e-05, "loss": 0.0214, "step": 77400 }, { "grad_norm": 0.33535534143447876, "learning_rate": 1.3315878561823697e-05, "loss": 0.0334, "step": 77410 }, { "grad_norm": 0.2447085976600647, "learning_rate": 1.330464536404773e-05, "loss": 0.0176, "step": 77420 }, { "grad_norm": 0.5118458867073059, "learning_rate": 1.3293416179220563e-05, "loss": 0.0209, "step": 77430 }, { "grad_norm": 0.785585880279541, "learning_rate": 1.328219100857021e-05, "loss": 0.0293, "step": 77440 }, { "grad_norm": 0.2576414942741394, "learning_rate": 1.327096985332425e-05, "loss": 0.012, "step": 77450 }, { "grad_norm": 0.3058946132659912, "learning_rate": 1.3259752714709816e-05, "loss": 0.0211, "step": 77460 }, { "grad_norm": 0.15825454890727997, "learning_rate": 1.3248539593953569e-05, "loss": 0.0145, "step": 77470 }, { "grad_norm": 0.37638673186302185, "learning_rate": 1.323733049228178e-05, "loss": 0.0164, "step": 77480 }, { "grad_norm": 0.6318296194076538, "learning_rate": 1.3226125410920276e-05, "loss": 0.0277, "step": 77490 }, { "grad_norm": 0.5641838908195496, "learning_rate": 1.3214924351094388e-05, "loss": 0.0188, "step": 77500 }, { "grad_norm": 0.4992891252040863, "learning_rate": 1.3203727314029069e-05, "loss": 0.0177, "step": 77510 }, { "grad_norm": 0.5685753226280212, "learning_rate": 1.3192534300948811e-05, "loss": 0.0193, "step": 77520 }, { "grad_norm": 0.3567071557044983, "learning_rate": 1.3181345313077675e-05, "loss": 0.0179, "step": 77530 }, { "grad_norm": 0.3835587203502655, "learning_rate": 1.3170160351639243e-05, "loss": 0.0206, "step": 77540 }, { "grad_norm": 0.2030416578054428, "learning_rate": 1.3158979417856721e-05, "loss": 0.0189, "step": 77550 }, { "grad_norm": 0.537874698638916, "learning_rate": 1.31478025129528e-05, "loss": 0.0216, "step": 77560 }, { "grad_norm": 0.36476656794548035, "learning_rate": 1.3136629638149794e-05, "loss": 0.0221, "step": 77570 }, { "grad_norm": 0.7432723641395569, "learning_rate": 1.3125460794669547e-05, "loss": 0.0167, "step": 77580 }, { "grad_norm": 0.323512464761734, "learning_rate": 1.3114295983733483e-05, "loss": 0.0217, "step": 77590 }, { "grad_norm": 0.757703959941864, "learning_rate": 1.3103135206562534e-05, "loss": 0.0182, "step": 77600 }, { "grad_norm": 0.4212322235107422, "learning_rate": 1.3091978464377264e-05, "loss": 0.033, "step": 77610 }, { "grad_norm": 0.3493383824825287, "learning_rate": 1.3080825758397724e-05, "loss": 0.0277, "step": 77620 }, { "grad_norm": 0.4179883897304535, "learning_rate": 1.3069677089843568e-05, "loss": 0.024, "step": 77630 }, { "grad_norm": 0.3801622986793518, "learning_rate": 1.305853245993401e-05, "loss": 0.0169, "step": 77640 }, { "grad_norm": 0.5174313187599182, "learning_rate": 1.3047391869887809e-05, "loss": 0.0144, "step": 77650 }, { "grad_norm": 0.29799458384513855, "learning_rate": 1.3036255320923263e-05, "loss": 0.033, "step": 77660 }, { "grad_norm": 0.7017481327056885, "learning_rate": 1.3025122814258273e-05, "loss": 0.0208, "step": 77670 }, { "grad_norm": 0.6296066641807556, "learning_rate": 1.3013994351110249e-05, "loss": 0.0187, "step": 77680 }, { "grad_norm": 0.15473034977912903, "learning_rate": 1.3002869932696194e-05, "loss": 0.0266, "step": 77690 }, { "grad_norm": 0.18139459192752838, "learning_rate": 1.2991749560232664e-05, "loss": 0.0162, "step": 77700 }, { "grad_norm": 0.3124525249004364, "learning_rate": 1.2980633234935768e-05, "loss": 0.0172, "step": 77710 }, { "grad_norm": 0.3234005868434906, "learning_rate": 1.2969520958021153e-05, "loss": 0.0141, "step": 77720 }, { "grad_norm": 0.29796138405799866, "learning_rate": 1.295841273070405e-05, "loss": 0.0152, "step": 77730 }, { "grad_norm": 0.6470608115196228, "learning_rate": 1.2947308554199255e-05, "loss": 0.0158, "step": 77740 }, { "grad_norm": 0.1916796714067459, "learning_rate": 1.2936208429721075e-05, "loss": 0.0225, "step": 77750 }, { "grad_norm": 0.9753814935684204, "learning_rate": 1.2925112358483416e-05, "loss": 0.0255, "step": 77760 }, { "grad_norm": 0.32403454184532166, "learning_rate": 1.2914020341699728e-05, "loss": 0.0179, "step": 77770 }, { "grad_norm": 0.44676312804222107, "learning_rate": 1.2902932380583032e-05, "loss": 0.0222, "step": 77780 }, { "grad_norm": 0.35236310958862305, "learning_rate": 1.2891848476345863e-05, "loss": 0.0231, "step": 77790 }, { "grad_norm": 0.553010880947113, "learning_rate": 1.2880768630200368e-05, "loss": 0.0156, "step": 77800 }, { "grad_norm": 0.39810502529144287, "learning_rate": 1.286969284335819e-05, "loss": 0.0173, "step": 77810 }, { "grad_norm": 0.5754091143608093, "learning_rate": 1.2858621117030578e-05, "loss": 0.029, "step": 77820 }, { "grad_norm": 0.21575455367565155, "learning_rate": 1.2847553452428324e-05, "loss": 0.0256, "step": 77830 }, { "grad_norm": 0.6679108142852783, "learning_rate": 1.283648985076178e-05, "loss": 0.0282, "step": 77840 }, { "grad_norm": 0.3178909718990326, "learning_rate": 1.2825430313240817e-05, "loss": 0.0194, "step": 77850 }, { "grad_norm": 0.3297904431819916, "learning_rate": 1.281437484107491e-05, "loss": 0.0218, "step": 77860 }, { "grad_norm": 1.0061191320419312, "learning_rate": 1.2803323435473058e-05, "loss": 0.0275, "step": 77870 }, { "grad_norm": 0.2960337996482849, "learning_rate": 1.2792276097643841e-05, "loss": 0.0235, "step": 77880 }, { "grad_norm": 0.5644716024398804, "learning_rate": 1.2781232828795342e-05, "loss": 0.0167, "step": 77890 }, { "grad_norm": 0.2734161913394928, "learning_rate": 1.2770193630135292e-05, "loss": 0.0192, "step": 77900 }, { "grad_norm": 0.1329525262117386, "learning_rate": 1.2759158502870877e-05, "loss": 0.0142, "step": 77910 }, { "grad_norm": 0.3126201033592224, "learning_rate": 1.2748127448208913e-05, "loss": 0.0196, "step": 77920 }, { "grad_norm": 0.12061071395874023, "learning_rate": 1.2737100467355706e-05, "loss": 0.016, "step": 77930 }, { "grad_norm": 0.5508826971054077, "learning_rate": 1.2726077561517169e-05, "loss": 0.0194, "step": 77940 }, { "grad_norm": 0.27108389139175415, "learning_rate": 1.2715058731898755e-05, "loss": 0.0204, "step": 77950 }, { "grad_norm": 0.1892518848180771, "learning_rate": 1.2704043979705471e-05, "loss": 0.0271, "step": 77960 }, { "grad_norm": 0.26074299216270447, "learning_rate": 1.269303330614185e-05, "loss": 0.0156, "step": 77970 }, { "grad_norm": 0.31929340958595276, "learning_rate": 1.2682026712412015e-05, "loss": 0.0287, "step": 77980 }, { "grad_norm": 0.6067538857460022, "learning_rate": 1.2671024199719644e-05, "loss": 0.019, "step": 77990 }, { "grad_norm": 0.35124778747558594, "learning_rate": 1.2660025769267936e-05, "loss": 0.0153, "step": 78000 }, { "grad_norm": 0.21045655012130737, "learning_rate": 1.2649031422259667e-05, "loss": 0.0218, "step": 78010 }, { "grad_norm": 0.39967793226242065, "learning_rate": 1.2638041159897163e-05, "loss": 0.0165, "step": 78020 }, { "grad_norm": 0.44190576672554016, "learning_rate": 1.262705498338232e-05, "loss": 0.0192, "step": 78030 }, { "grad_norm": 0.2536466717720032, "learning_rate": 1.2616072893916542e-05, "loss": 0.0139, "step": 78040 }, { "grad_norm": 0.39814966917037964, "learning_rate": 1.2605094892700836e-05, "loss": 0.0189, "step": 78050 }, { "grad_norm": 0.23474954068660736, "learning_rate": 1.259412098093572e-05, "loss": 0.0152, "step": 78060 }, { "grad_norm": 0.22723489999771118, "learning_rate": 1.258315115982131e-05, "loss": 0.0156, "step": 78070 }, { "grad_norm": 0.5951916575431824, "learning_rate": 1.2572185430557203e-05, "loss": 0.0176, "step": 78080 }, { "grad_norm": 0.39535167813301086, "learning_rate": 1.2561223794342659e-05, "loss": 0.014, "step": 78090 }, { "grad_norm": 0.5315536260604858, "learning_rate": 1.255026625237637e-05, "loss": 0.0214, "step": 78100 }, { "grad_norm": 0.5424441695213318, "learning_rate": 1.2539312805856678e-05, "loss": 0.0171, "step": 78110 }, { "grad_norm": 0.2693302035331726, "learning_rate": 1.25283634559814e-05, "loss": 0.0146, "step": 78120 }, { "grad_norm": 0.1372811198234558, "learning_rate": 1.2517418203947972e-05, "loss": 0.0184, "step": 78130 }, { "grad_norm": 0.6825462579727173, "learning_rate": 1.2506477050953308e-05, "loss": 0.0229, "step": 78140 }, { "grad_norm": 0.43616604804992676, "learning_rate": 1.2495539998193972e-05, "loss": 0.029, "step": 78150 }, { "grad_norm": 0.5153859853744507, "learning_rate": 1.2484607046865975e-05, "loss": 0.0255, "step": 78160 }, { "grad_norm": 0.7412675619125366, "learning_rate": 1.2473678198164968e-05, "loss": 0.0186, "step": 78170 }, { "grad_norm": 0.3487262427806854, "learning_rate": 1.2462753453286075e-05, "loss": 0.017, "step": 78180 }, { "grad_norm": 0.2194344699382782, "learning_rate": 1.2451832813424031e-05, "loss": 0.0182, "step": 78190 }, { "grad_norm": 0.45859813690185547, "learning_rate": 1.2440916279773095e-05, "loss": 0.0128, "step": 78200 }, { "grad_norm": 0.15832366049289703, "learning_rate": 1.2430003853527101e-05, "loss": 0.0132, "step": 78210 }, { "grad_norm": 0.34825143218040466, "learning_rate": 1.2419095535879383e-05, "loss": 0.024, "step": 78220 }, { "grad_norm": 0.13218119740486145, "learning_rate": 1.2408191328022878e-05, "loss": 0.0156, "step": 78230 }, { "grad_norm": 0.38730236887931824, "learning_rate": 1.2397291231150066e-05, "loss": 0.015, "step": 78240 }, { "grad_norm": 0.3876360356807709, "learning_rate": 1.2386395246452937e-05, "loss": 0.0176, "step": 78250 }, { "grad_norm": 0.3404397964477539, "learning_rate": 1.2375503375123083e-05, "loss": 0.0163, "step": 78260 }, { "grad_norm": 0.6221644282341003, "learning_rate": 1.2364615618351583e-05, "loss": 0.0165, "step": 78270 }, { "grad_norm": 0.26144692301750183, "learning_rate": 1.2353731977329169e-05, "loss": 0.0181, "step": 78280 }, { "grad_norm": 0.12882180511951447, "learning_rate": 1.2342852453246007e-05, "loss": 0.018, "step": 78290 }, { "grad_norm": 0.6461440324783325, "learning_rate": 1.2331977047291898e-05, "loss": 0.0401, "step": 78300 }, { "grad_norm": 0.2334769070148468, "learning_rate": 1.2321105760656137e-05, "loss": 0.0174, "step": 78310 }, { "grad_norm": 0.39064252376556396, "learning_rate": 1.2310238594527613e-05, "loss": 0.0195, "step": 78320 }, { "grad_norm": 0.37948551774024963, "learning_rate": 1.2299375550094693e-05, "loss": 0.0223, "step": 78330 }, { "grad_norm": 0.3005928099155426, "learning_rate": 1.2288516628545421e-05, "loss": 0.017, "step": 78340 }, { "grad_norm": 0.20733213424682617, "learning_rate": 1.2277661831067255e-05, "loss": 0.0198, "step": 78350 }, { "grad_norm": 0.09122855216264725, "learning_rate": 1.2266811158847286e-05, "loss": 0.0196, "step": 78360 }, { "grad_norm": 0.18027563393115997, "learning_rate": 1.2255964613072107e-05, "loss": 0.0171, "step": 78370 }, { "grad_norm": 0.4676530361175537, "learning_rate": 1.2245122194927905e-05, "loss": 0.0242, "step": 78380 }, { "grad_norm": 0.43401721119880676, "learning_rate": 1.2234283905600364e-05, "loss": 0.0143, "step": 78390 }, { "grad_norm": 0.3733733296394348, "learning_rate": 1.222344974627475e-05, "loss": 0.014, "step": 78400 }, { "grad_norm": 0.2858322262763977, "learning_rate": 1.2212619718135875e-05, "loss": 0.0237, "step": 78410 }, { "grad_norm": 0.29773348569869995, "learning_rate": 1.2201793822368118e-05, "loss": 0.022, "step": 78420 }, { "grad_norm": 0.3111523985862732, "learning_rate": 1.2190972060155337e-05, "loss": 0.0146, "step": 78430 }, { "grad_norm": 1.483578085899353, "learning_rate": 1.2180154432681013e-05, "loss": 0.0218, "step": 78440 }, { "grad_norm": 0.5334858298301697, "learning_rate": 1.2169340941128148e-05, "loss": 0.0193, "step": 78450 }, { "grad_norm": 0.26885899901390076, "learning_rate": 1.2158531586679267e-05, "loss": 0.02, "step": 78460 }, { "grad_norm": 0.39609295129776, "learning_rate": 1.2147726370516477e-05, "loss": 0.0272, "step": 78470 }, { "grad_norm": 0.3417681157588959, "learning_rate": 1.2136925293821422e-05, "loss": 0.0218, "step": 78480 }, { "grad_norm": 0.3253991901874542, "learning_rate": 1.2126128357775306e-05, "loss": 0.0185, "step": 78490 }, { "grad_norm": 0.4824027121067047, "learning_rate": 1.2115335563558838e-05, "loss": 0.0137, "step": 78500 }, { "grad_norm": 0.41974076628685, "learning_rate": 1.2104546912352327e-05, "loss": 0.0197, "step": 78510 }, { "grad_norm": 0.510625958442688, "learning_rate": 1.2093762405335557e-05, "loss": 0.0138, "step": 78520 }, { "grad_norm": 0.28014957904815674, "learning_rate": 1.2082982043687974e-05, "loss": 0.0225, "step": 78530 }, { "grad_norm": 0.4920631945133209, "learning_rate": 1.207220582858845e-05, "loss": 0.0141, "step": 78540 }, { "grad_norm": 0.4227845072746277, "learning_rate": 1.206143376121549e-05, "loss": 0.0195, "step": 78550 }, { "grad_norm": 0.5218507647514343, "learning_rate": 1.2050665842747078e-05, "loss": 0.0218, "step": 78560 }, { "grad_norm": 0.45696237683296204, "learning_rate": 1.2039902074360804e-05, "loss": 0.0335, "step": 78570 }, { "grad_norm": 0.3423129916191101, "learning_rate": 1.2029142457233755e-05, "loss": 0.0328, "step": 78580 }, { "grad_norm": 0.4328921437263489, "learning_rate": 1.2018386992542601e-05, "loss": 0.0189, "step": 78590 }, { "grad_norm": 0.481669157743454, "learning_rate": 1.2007635681463542e-05, "loss": 0.0186, "step": 78600 }, { "grad_norm": 0.3719334900379181, "learning_rate": 1.1996888525172334e-05, "loss": 0.0298, "step": 78610 }, { "grad_norm": 0.2556893527507782, "learning_rate": 1.198614552484425e-05, "loss": 0.0124, "step": 78620 }, { "grad_norm": 0.29728469252586365, "learning_rate": 1.197540668165415e-05, "loss": 0.0187, "step": 78630 }, { "grad_norm": 0.2992728650569916, "learning_rate": 1.1964671996776395e-05, "loss": 0.0155, "step": 78640 }, { "grad_norm": 0.1968095302581787, "learning_rate": 1.1953941471384922e-05, "loss": 0.0233, "step": 78650 }, { "grad_norm": 0.1806981861591339, "learning_rate": 1.1943215106653205e-05, "loss": 0.0143, "step": 78660 }, { "grad_norm": 0.5165706872940063, "learning_rate": 1.193249290375429e-05, "loss": 0.0184, "step": 78670 }, { "grad_norm": 0.33454909920692444, "learning_rate": 1.1921774863860696e-05, "loss": 0.0247, "step": 78680 }, { "grad_norm": 1.0701699256896973, "learning_rate": 1.1911060988144556e-05, "loss": 0.0235, "step": 78690 }, { "grad_norm": 0.32153019309043884, "learning_rate": 1.1900351277777533e-05, "loss": 0.0196, "step": 78700 }, { "grad_norm": 0.23919960856437683, "learning_rate": 1.1889645733930798e-05, "loss": 0.0155, "step": 78710 }, { "grad_norm": 0.23241287469863892, "learning_rate": 1.1878944357775107e-05, "loss": 0.0201, "step": 78720 }, { "grad_norm": 0.26057764887809753, "learning_rate": 1.1868247150480749e-05, "loss": 0.0156, "step": 78730 }, { "grad_norm": 0.240279421210289, "learning_rate": 1.1857554113217567e-05, "loss": 0.0184, "step": 78740 }, { "grad_norm": 0.4407442510128021, "learning_rate": 1.1846865247154903e-05, "loss": 0.0192, "step": 78750 }, { "grad_norm": 0.5652039051055908, "learning_rate": 1.1836180553461711e-05, "loss": 0.018, "step": 78760 }, { "grad_norm": 0.5264766216278076, "learning_rate": 1.1825500033306419e-05, "loss": 0.02, "step": 78770 }, { "grad_norm": 0.19627052545547485, "learning_rate": 1.1814823687857052e-05, "loss": 0.0198, "step": 78780 }, { "grad_norm": 0.4165132939815521, "learning_rate": 1.1804151518281153e-05, "loss": 0.032, "step": 78790 }, { "grad_norm": 0.20299841463565826, "learning_rate": 1.179348352574583e-05, "loss": 0.021, "step": 78800 }, { "grad_norm": 0.3566228449344635, "learning_rate": 1.1782819711417697e-05, "loss": 0.0206, "step": 78810 }, { "grad_norm": 0.6661624312400818, "learning_rate": 1.1772160076462951e-05, "loss": 0.0364, "step": 78820 }, { "grad_norm": 0.40140071511268616, "learning_rate": 1.1761504622047293e-05, "loss": 0.018, "step": 78830 }, { "grad_norm": 0.20076896250247955, "learning_rate": 1.1750853349335999e-05, "loss": 0.0153, "step": 78840 }, { "grad_norm": 0.4206864535808563, "learning_rate": 1.1740206259493874e-05, "loss": 0.0224, "step": 78850 }, { "grad_norm": 0.5281702280044556, "learning_rate": 1.1729563353685286e-05, "loss": 0.0193, "step": 78860 }, { "grad_norm": 0.337077260017395, "learning_rate": 1.1718924633074091e-05, "loss": 0.0223, "step": 78870 }, { "grad_norm": 0.33451899886131287, "learning_rate": 1.1708290098823765e-05, "loss": 0.0222, "step": 78880 }, { "grad_norm": 0.2220848947763443, "learning_rate": 1.1697659752097245e-05, "loss": 0.0137, "step": 78890 }, { "grad_norm": 0.21410849690437317, "learning_rate": 1.1687033594057068e-05, "loss": 0.0195, "step": 78900 }, { "grad_norm": 0.31419920921325684, "learning_rate": 1.1676411625865297e-05, "loss": 0.019, "step": 78910 }, { "grad_norm": 0.36978161334991455, "learning_rate": 1.1665793848683542e-05, "loss": 0.0237, "step": 78920 }, { "grad_norm": 0.3256838023662567, "learning_rate": 1.1655180263672927e-05, "loss": 0.0163, "step": 78930 }, { "grad_norm": 0.4387842118740082, "learning_rate": 1.1644570871994142e-05, "loss": 0.0137, "step": 78940 }, { "grad_norm": 0.28051450848579407, "learning_rate": 1.163396567480744e-05, "loss": 0.0251, "step": 78950 }, { "grad_norm": 0.2573156952857971, "learning_rate": 1.1623364673272552e-05, "loss": 0.0184, "step": 78960 }, { "grad_norm": 0.45529499650001526, "learning_rate": 1.1612767868548807e-05, "loss": 0.0218, "step": 78970 }, { "grad_norm": 0.49816131591796875, "learning_rate": 1.1602175261795057e-05, "loss": 0.0294, "step": 78980 }, { "grad_norm": 0.3897166848182678, "learning_rate": 1.1591586854169696e-05, "loss": 0.0181, "step": 78990 }, { "grad_norm": 0.5821906328201294, "learning_rate": 1.158100264683064e-05, "loss": 0.0203, "step": 79000 }, { "grad_norm": 0.3755244314670563, "learning_rate": 1.1570422640935386e-05, "loss": 0.0193, "step": 79010 }, { "grad_norm": 0.3686812222003937, "learning_rate": 1.155984683764092e-05, "loss": 0.022, "step": 79020 }, { "grad_norm": 0.5066416263580322, "learning_rate": 1.1549275238103812e-05, "loss": 0.0216, "step": 79030 }, { "grad_norm": 0.4465009272098541, "learning_rate": 1.1538707843480152e-05, "loss": 0.0229, "step": 79040 }, { "grad_norm": 0.37067094445228577, "learning_rate": 1.1528144654925593e-05, "loss": 0.0174, "step": 79050 }, { "grad_norm": 0.4879986345767975, "learning_rate": 1.1517585673595277e-05, "loss": 0.014, "step": 79060 }, { "grad_norm": 0.4977380335330963, "learning_rate": 1.150703090064395e-05, "loss": 0.0172, "step": 79070 }, { "grad_norm": 0.18786682188510895, "learning_rate": 1.1496480337225835e-05, "loss": 0.0139, "step": 79080 }, { "grad_norm": 0.5058441162109375, "learning_rate": 1.1485933984494751e-05, "loss": 0.0217, "step": 79090 }, { "grad_norm": 0.6983460187911987, "learning_rate": 1.1475391843604e-05, "loss": 0.0415, "step": 79100 }, { "grad_norm": 0.46326059103012085, "learning_rate": 1.1464853915706503e-05, "loss": 0.0219, "step": 79110 }, { "grad_norm": 0.5631691813468933, "learning_rate": 1.1454320201954626e-05, "loss": 0.0267, "step": 79120 }, { "grad_norm": 0.2111225426197052, "learning_rate": 1.1443790703500356e-05, "loss": 0.0188, "step": 79130 }, { "grad_norm": 0.31651467084884644, "learning_rate": 1.1433265421495154e-05, "loss": 0.0145, "step": 79140 }, { "grad_norm": 0.6649762392044067, "learning_rate": 1.1422744357090065e-05, "loss": 0.0145, "step": 79150 }, { "grad_norm": 0.484017938375473, "learning_rate": 1.1412227511435647e-05, "loss": 0.0308, "step": 79160 }, { "grad_norm": 0.26393455266952515, "learning_rate": 1.1401714885682025e-05, "loss": 0.0318, "step": 79170 }, { "grad_norm": 1.017138957977295, "learning_rate": 1.1391206480978823e-05, "loss": 0.0283, "step": 79180 }, { "grad_norm": 0.4650023579597473, "learning_rate": 1.1380702298475227e-05, "loss": 0.024, "step": 79190 }, { "grad_norm": 0.5536490082740784, "learning_rate": 1.1370202339319985e-05, "loss": 0.0224, "step": 79200 }, { "grad_norm": 0.4409794807434082, "learning_rate": 1.1359706604661319e-05, "loss": 0.0163, "step": 79210 }, { "grad_norm": 0.47086647152900696, "learning_rate": 1.1349215095647042e-05, "loss": 0.0193, "step": 79220 }, { "grad_norm": 0.5200304388999939, "learning_rate": 1.1338727813424493e-05, "loss": 0.0176, "step": 79230 }, { "grad_norm": 0.34769386053085327, "learning_rate": 1.132824475914056e-05, "loss": 0.0165, "step": 79240 }, { "grad_norm": 0.21317444741725922, "learning_rate": 1.131776593394162e-05, "loss": 0.0287, "step": 79250 }, { "grad_norm": 0.3632715940475464, "learning_rate": 1.1307291338973657e-05, "loss": 0.0134, "step": 79260 }, { "grad_norm": 0.28732848167419434, "learning_rate": 1.1296820975382121e-05, "loss": 0.0174, "step": 79270 }, { "grad_norm": 0.23667021095752716, "learning_rate": 1.1286354844312074e-05, "loss": 0.0456, "step": 79280 }, { "grad_norm": 0.8037078976631165, "learning_rate": 1.1275892946908023e-05, "loss": 0.0464, "step": 79290 }, { "grad_norm": 0.4791507422924042, "learning_rate": 1.126543528431413e-05, "loss": 0.024, "step": 79300 }, { "grad_norm": 0.40379786491394043, "learning_rate": 1.125498185767398e-05, "loss": 0.026, "step": 79310 }, { "grad_norm": 0.29821938276290894, "learning_rate": 1.1244532668130781e-05, "loss": 0.0152, "step": 79320 }, { "grad_norm": 0.30913662910461426, "learning_rate": 1.1234087716827203e-05, "loss": 0.0185, "step": 79330 }, { "grad_norm": 0.2762138545513153, "learning_rate": 1.1223647004905524e-05, "loss": 0.0214, "step": 79340 }, { "grad_norm": 0.3470464050769806, "learning_rate": 1.1213210533507485e-05, "loss": 0.0166, "step": 79350 }, { "grad_norm": 0.7240222096443176, "learning_rate": 1.1202778303774447e-05, "loss": 0.0184, "step": 79360 }, { "grad_norm": 0.5310588479042053, "learning_rate": 1.1192350316847234e-05, "loss": 0.0165, "step": 79370 }, { "grad_norm": 0.6782034039497375, "learning_rate": 1.1181926573866258e-05, "loss": 0.0205, "step": 79380 }, { "grad_norm": 0.3430876135826111, "learning_rate": 1.1171507075971416e-05, "loss": 0.0123, "step": 79390 }, { "grad_norm": 0.23548872768878937, "learning_rate": 1.1161091824302177e-05, "loss": 0.0122, "step": 79400 }, { "grad_norm": 0.571090042591095, "learning_rate": 1.115068081999755e-05, "loss": 0.0223, "step": 79410 }, { "grad_norm": 0.4050534963607788, "learning_rate": 1.1140274064196071e-05, "loss": 0.0146, "step": 79420 }, { "grad_norm": 0.5914576053619385, "learning_rate": 1.112987155803578e-05, "loss": 0.0165, "step": 79430 }, { "grad_norm": 0.625018835067749, "learning_rate": 1.1119473302654298e-05, "loss": 0.0198, "step": 79440 }, { "grad_norm": 0.685663104057312, "learning_rate": 1.1109079299188769e-05, "loss": 0.0415, "step": 79450 }, { "grad_norm": 0.3929814398288727, "learning_rate": 1.1098689548775847e-05, "loss": 0.0272, "step": 79460 }, { "grad_norm": 0.34267911314964294, "learning_rate": 1.1088304052551762e-05, "loss": 0.0213, "step": 79470 }, { "grad_norm": 0.22264955937862396, "learning_rate": 1.1077922811652215e-05, "loss": 0.0142, "step": 79480 }, { "grad_norm": 0.3596348166465759, "learning_rate": 1.1067545827212534e-05, "loss": 0.0254, "step": 79490 }, { "grad_norm": 0.25239285826683044, "learning_rate": 1.1057173100367496e-05, "loss": 0.0171, "step": 79500 }, { "grad_norm": 0.3329136073589325, "learning_rate": 1.1046804632251474e-05, "loss": 0.0204, "step": 79510 }, { "grad_norm": 0.39055076241493225, "learning_rate": 1.1036440423998318e-05, "loss": 0.0178, "step": 79520 }, { "grad_norm": 0.5608648061752319, "learning_rate": 1.1026080476741469e-05, "loss": 0.0219, "step": 79530 }, { "grad_norm": 0.24143089354038239, "learning_rate": 1.1015724791613836e-05, "loss": 0.02, "step": 79540 }, { "grad_norm": 0.39241012930870056, "learning_rate": 1.1005373369747951e-05, "loss": 0.0223, "step": 79550 }, { "grad_norm": 0.2769375741481781, "learning_rate": 1.09950262122758e-05, "loss": 0.0208, "step": 79560 }, { "grad_norm": 0.4201788604259491, "learning_rate": 1.0984683320328948e-05, "loss": 0.0285, "step": 79570 }, { "grad_norm": 0.2664512097835541, "learning_rate": 1.0974344695038458e-05, "loss": 0.0167, "step": 79580 }, { "grad_norm": 0.44948697090148926, "learning_rate": 1.0964010337534968e-05, "loss": 0.0186, "step": 79590 }, { "grad_norm": 0.447620689868927, "learning_rate": 1.0953680248948611e-05, "loss": 0.0306, "step": 79600 }, { "grad_norm": 0.23532137274742126, "learning_rate": 1.0943354430409069e-05, "loss": 0.0265, "step": 79610 }, { "grad_norm": 0.47819191217422485, "learning_rate": 1.0933032883045574e-05, "loss": 0.0201, "step": 79620 }, { "grad_norm": 0.14422613382339478, "learning_rate": 1.0922715607986872e-05, "loss": 0.0206, "step": 79630 }, { "grad_norm": 0.2624410390853882, "learning_rate": 1.0912402606361234e-05, "loss": 0.0259, "step": 79640 }, { "grad_norm": 0.34819450974464417, "learning_rate": 1.090209387929647e-05, "loss": 0.0231, "step": 79650 }, { "grad_norm": 0.2693547308444977, "learning_rate": 1.0891789427919957e-05, "loss": 0.0153, "step": 79660 }, { "grad_norm": 0.37348172068595886, "learning_rate": 1.088148925335854e-05, "loss": 0.0224, "step": 79670 }, { "grad_norm": 0.3062489330768585, "learning_rate": 1.087119335673864e-05, "loss": 0.0204, "step": 79680 }, { "grad_norm": 0.36265480518341064, "learning_rate": 1.0860901739186208e-05, "loss": 0.0166, "step": 79690 }, { "grad_norm": 0.3660309314727783, "learning_rate": 1.085061440182673e-05, "loss": 0.0126, "step": 79700 }, { "grad_norm": 0.4251755177974701, "learning_rate": 1.0840331345785193e-05, "loss": 0.0198, "step": 79710 }, { "grad_norm": 0.8738268613815308, "learning_rate": 1.083005257218615e-05, "loss": 0.0214, "step": 79720 }, { "grad_norm": 0.21681657433509827, "learning_rate": 1.0819778082153642e-05, "loss": 0.0169, "step": 79730 }, { "grad_norm": 0.39208945631980896, "learning_rate": 1.0809507876811325e-05, "loss": 0.0234, "step": 79740 }, { "grad_norm": 0.2000851035118103, "learning_rate": 1.0799241957282292e-05, "loss": 0.0176, "step": 79750 }, { "grad_norm": 0.538141667842865, "learning_rate": 1.078898032468923e-05, "loss": 0.0191, "step": 79760 }, { "grad_norm": 0.2961901128292084, "learning_rate": 1.0778722980154315e-05, "loss": 0.0204, "step": 79770 }, { "grad_norm": 0.35002976655960083, "learning_rate": 1.0768469924799302e-05, "loss": 0.0168, "step": 79780 }, { "grad_norm": 0.5652275085449219, "learning_rate": 1.0758221159745419e-05, "loss": 0.0146, "step": 79790 }, { "grad_norm": 0.31633272767066956, "learning_rate": 1.074797668611347e-05, "loss": 0.0134, "step": 79800 }, { "grad_norm": 0.313637375831604, "learning_rate": 1.0737736505023777e-05, "loss": 0.0244, "step": 79810 }, { "grad_norm": 0.2585194408893585, "learning_rate": 1.0727500617596203e-05, "loss": 0.0138, "step": 79820 }, { "grad_norm": 0.3133571743965149, "learning_rate": 1.0717269024950104e-05, "loss": 0.0146, "step": 79830 }, { "grad_norm": 0.5003575682640076, "learning_rate": 1.070704172820442e-05, "loss": 0.0224, "step": 79840 }, { "grad_norm": 0.9014260768890381, "learning_rate": 1.0696818728477559e-05, "loss": 0.0311, "step": 79850 }, { "grad_norm": 0.2287522405385971, "learning_rate": 1.0686600026887511e-05, "loss": 0.0223, "step": 79860 }, { "grad_norm": 0.34220993518829346, "learning_rate": 1.0676385624551777e-05, "loss": 0.017, "step": 79870 }, { "grad_norm": 0.5188735723495483, "learning_rate": 1.0666175522587402e-05, "loss": 0.0187, "step": 79880 }, { "grad_norm": 0.30998387932777405, "learning_rate": 1.0655969722110925e-05, "loss": 0.0206, "step": 79890 }, { "grad_norm": 0.42260149121284485, "learning_rate": 1.0645768224238439e-05, "loss": 0.0234, "step": 79900 }, { "grad_norm": 0.3644494116306305, "learning_rate": 1.063557103008559e-05, "loss": 0.0209, "step": 79910 }, { "grad_norm": 0.6437186002731323, "learning_rate": 1.0625378140767494e-05, "loss": 0.0179, "step": 79920 }, { "grad_norm": 0.8118435144424438, "learning_rate": 1.0615189557398846e-05, "loss": 0.0288, "step": 79930 }, { "grad_norm": 0.400800496339798, "learning_rate": 1.060500528109386e-05, "loss": 0.0197, "step": 79940 }, { "grad_norm": 0.1601584255695343, "learning_rate": 1.0594825312966272e-05, "loss": 0.021, "step": 79950 }, { "grad_norm": 0.41903015971183777, "learning_rate": 1.0584649654129336e-05, "loss": 0.0228, "step": 79960 }, { "grad_norm": 0.24308140575885773, "learning_rate": 1.0574478305695863e-05, "loss": 0.0175, "step": 79970 }, { "grad_norm": 0.3113025426864624, "learning_rate": 1.0564311268778154e-05, "loss": 0.0163, "step": 79980 }, { "grad_norm": 0.3065667152404785, "learning_rate": 1.0554148544488073e-05, "loss": 0.0247, "step": 79990 }, { "grad_norm": 0.592487096786499, "learning_rate": 1.0543990133937004e-05, "loss": 0.0224, "step": 80000 }, { "grad_norm": 0.6561332941055298, "learning_rate": 1.0533836038235868e-05, "loss": 0.0157, "step": 80010 }, { "grad_norm": 0.38410091400146484, "learning_rate": 1.052368625849507e-05, "loss": 0.0205, "step": 80020 }, { "grad_norm": 0.48958709836006165, "learning_rate": 1.05135407958246e-05, "loss": 0.0229, "step": 80030 }, { "grad_norm": 0.25764888525009155, "learning_rate": 1.0503399651333934e-05, "loss": 0.0117, "step": 80040 }, { "grad_norm": 0.25718194246292114, "learning_rate": 1.04932628261321e-05, "loss": 0.0165, "step": 80050 }, { "grad_norm": 0.4451533854007721, "learning_rate": 1.0483130321327644e-05, "loss": 0.0293, "step": 80060 }, { "grad_norm": 0.35769662261009216, "learning_rate": 1.0473002138028654e-05, "loss": 0.0121, "step": 80070 }, { "grad_norm": 0.23048147559165955, "learning_rate": 1.0462878277342713e-05, "loss": 0.023, "step": 80080 }, { "grad_norm": 0.2229100614786148, "learning_rate": 1.0452758740376972e-05, "loss": 0.0183, "step": 80090 }, { "grad_norm": 0.27980467677116394, "learning_rate": 1.0442643528238067e-05, "loss": 0.0213, "step": 80100 }, { "grad_norm": 0.6723426580429077, "learning_rate": 1.0432532642032188e-05, "loss": 0.0221, "step": 80110 }, { "grad_norm": 0.7237527966499329, "learning_rate": 1.042242608286505e-05, "loss": 0.0193, "step": 80120 }, { "grad_norm": 0.28656336665153503, "learning_rate": 1.041232385184191e-05, "loss": 0.021, "step": 80130 }, { "grad_norm": 0.472636878490448, "learning_rate": 1.0402225950067495e-05, "loss": 0.028, "step": 80140 }, { "grad_norm": 1.0249416828155518, "learning_rate": 1.0392132378646119e-05, "loss": 0.0173, "step": 80150 }, { "grad_norm": 0.5041460990905762, "learning_rate": 1.0382043138681607e-05, "loss": 0.0222, "step": 80160 }, { "grad_norm": 0.7068042159080505, "learning_rate": 1.0371958231277284e-05, "loss": 0.0161, "step": 80170 }, { "grad_norm": 0.5026992559432983, "learning_rate": 1.0361877657536024e-05, "loss": 0.0282, "step": 80180 }, { "grad_norm": 0.3103882968425751, "learning_rate": 1.035180141856023e-05, "loss": 0.0238, "step": 80190 }, { "grad_norm": 0.24373583495616913, "learning_rate": 1.0341729515451836e-05, "loss": 0.0198, "step": 80200 }, { "grad_norm": 0.4272271394729614, "learning_rate": 1.0331661949312265e-05, "loss": 0.013, "step": 80210 }, { "grad_norm": 0.21423867344856262, "learning_rate": 1.0321598721242514e-05, "loss": 0.015, "step": 80220 }, { "grad_norm": 0.565525233745575, "learning_rate": 1.0311539832343064e-05, "loss": 0.0148, "step": 80230 }, { "grad_norm": 0.2386138141155243, "learning_rate": 1.0301485283713942e-05, "loss": 0.0197, "step": 80240 }, { "grad_norm": 0.21505317091941833, "learning_rate": 1.0291435076454703e-05, "loss": 0.0192, "step": 80250 }, { "grad_norm": 0.510369062423706, "learning_rate": 1.028138921166444e-05, "loss": 0.0264, "step": 80260 }, { "grad_norm": 0.49157825112342834, "learning_rate": 1.0271347690441719e-05, "loss": 0.0223, "step": 80270 }, { "grad_norm": 0.37106040120124817, "learning_rate": 1.0261310513884698e-05, "loss": 0.0261, "step": 80280 }, { "grad_norm": 0.43571388721466064, "learning_rate": 1.0251277683090998e-05, "loss": 0.0287, "step": 80290 }, { "grad_norm": 0.8186731338500977, "learning_rate": 1.0241249199157826e-05, "loss": 0.0335, "step": 80300 }, { "grad_norm": 0.2348324954509735, "learning_rate": 1.0231225063181837e-05, "loss": 0.0191, "step": 80310 }, { "grad_norm": 0.2996349334716797, "learning_rate": 1.0221205276259316e-05, "loss": 0.0276, "step": 80320 }, { "grad_norm": 0.5067605376243591, "learning_rate": 1.0211189839485958e-05, "loss": 0.0154, "step": 80330 }, { "grad_norm": 0.48707062005996704, "learning_rate": 1.0201178753957075e-05, "loss": 0.0123, "step": 80340 }, { "grad_norm": 0.4406358599662781, "learning_rate": 1.0191172020767432e-05, "loss": 0.0173, "step": 80350 }, { "grad_norm": 0.29160359501838684, "learning_rate": 1.0181169641011362e-05, "loss": 0.0193, "step": 80360 }, { "grad_norm": 0.4085141718387604, "learning_rate": 1.0171171615782721e-05, "loss": 0.0291, "step": 80370 }, { "grad_norm": 0.30955377221107483, "learning_rate": 1.0161177946174866e-05, "loss": 0.0182, "step": 80380 }, { "grad_norm": 0.3905322253704071, "learning_rate": 1.0151188633280706e-05, "loss": 0.0173, "step": 80390 }, { "grad_norm": 0.3476172387599945, "learning_rate": 1.0141203678192635e-05, "loss": 0.0179, "step": 80400 }, { "grad_norm": 0.3315299451351166, "learning_rate": 1.0131223082002612e-05, "loss": 0.0221, "step": 80410 }, { "grad_norm": 0.22803843021392822, "learning_rate": 1.012124684580208e-05, "loss": 0.0217, "step": 80420 }, { "grad_norm": 0.2015766054391861, "learning_rate": 1.0111274970682039e-05, "loss": 0.021, "step": 80430 }, { "grad_norm": 0.2637553811073303, "learning_rate": 1.010130745773299e-05, "loss": 0.0181, "step": 80440 }, { "grad_norm": 0.22874484956264496, "learning_rate": 1.0091344308044986e-05, "loss": 0.0321, "step": 80450 }, { "grad_norm": 0.7532521486282349, "learning_rate": 1.0081385522707554e-05, "loss": 0.0183, "step": 80460 }, { "grad_norm": 0.29036808013916016, "learning_rate": 1.0071431102809803e-05, "loss": 0.0249, "step": 80470 }, { "grad_norm": 0.47372928261756897, "learning_rate": 1.0061481049440297e-05, "loss": 0.0186, "step": 80480 }, { "grad_norm": 0.531481921672821, "learning_rate": 1.0051535363687187e-05, "loss": 0.0153, "step": 80490 }, { "grad_norm": 0.4245957136154175, "learning_rate": 1.0041594046638087e-05, "loss": 0.0185, "step": 80500 }, { "grad_norm": 0.6437625288963318, "learning_rate": 1.0031657099380204e-05, "loss": 0.0178, "step": 80510 }, { "grad_norm": 0.6307567358016968, "learning_rate": 1.0021724523000204e-05, "loss": 0.0214, "step": 80520 }, { "grad_norm": 0.5899002552032471, "learning_rate": 1.001179631858431e-05, "loss": 0.0429, "step": 80530 }, { "grad_norm": 0.38456469774246216, "learning_rate": 1.0001872487218245e-05, "loss": 0.0196, "step": 80540 }, { "grad_norm": 0.47251513600349426, "learning_rate": 9.991953029987272e-06, "loss": 0.0205, "step": 80550 }, { "grad_norm": 0.38149943947792053, "learning_rate": 9.982037947976147e-06, "loss": 0.0185, "step": 80560 }, { "grad_norm": 0.2988046109676361, "learning_rate": 9.972127242269208e-06, "loss": 0.0296, "step": 80570 }, { "grad_norm": 0.21913689374923706, "learning_rate": 9.962220913950238e-06, "loss": 0.0121, "step": 80580 }, { "grad_norm": 0.5785266160964966, "learning_rate": 9.952318964102591e-06, "loss": 0.0278, "step": 80590 }, { "grad_norm": 0.40672770142555237, "learning_rate": 9.942421393809148e-06, "loss": 0.0233, "step": 80600 }, { "grad_norm": 0.653597354888916, "learning_rate": 9.932528204152264e-06, "loss": 0.0234, "step": 80610 }, { "grad_norm": 0.393196702003479, "learning_rate": 9.922639396213856e-06, "loss": 0.026, "step": 80620 }, { "grad_norm": 0.40727221965789795, "learning_rate": 9.912754971075344e-06, "loss": 0.021, "step": 80630 }, { "grad_norm": 0.45251455903053284, "learning_rate": 9.902874929817696e-06, "loss": 0.0156, "step": 80640 }, { "grad_norm": 0.42337071895599365, "learning_rate": 9.892999273521342e-06, "loss": 0.0218, "step": 80650 }, { "grad_norm": 0.5249279737472534, "learning_rate": 9.883128003266302e-06, "loss": 0.0215, "step": 80660 }, { "grad_norm": 0.6119744777679443, "learning_rate": 9.87326112013206e-06, "loss": 0.0184, "step": 80670 }, { "grad_norm": 0.35622939467430115, "learning_rate": 9.863398625197662e-06, "loss": 0.0165, "step": 80680 }, { "grad_norm": 0.3021387457847595, "learning_rate": 9.853540519541615e-06, "loss": 0.0254, "step": 80690 }, { "grad_norm": 0.4740382134914398, "learning_rate": 9.843686804242053e-06, "loss": 0.0172, "step": 80700 }, { "grad_norm": 0.2703644037246704, "learning_rate": 9.833837480376506e-06, "loss": 0.0165, "step": 80710 }, { "grad_norm": 0.21999134123325348, "learning_rate": 9.82399254902212e-06, "loss": 0.0136, "step": 80720 }, { "grad_norm": 0.27865028381347656, "learning_rate": 9.814152011255494e-06, "loss": 0.0147, "step": 80730 }, { "grad_norm": 0.4626317620277405, "learning_rate": 9.804315868152797e-06, "loss": 0.0242, "step": 80740 }, { "grad_norm": 0.4247804880142212, "learning_rate": 9.79448412078966e-06, "loss": 0.0176, "step": 80750 }, { "grad_norm": 0.4025416076183319, "learning_rate": 9.784656770241312e-06, "loss": 0.0203, "step": 80760 }, { "grad_norm": 0.2992628216743469, "learning_rate": 9.77483381758243e-06, "loss": 0.0168, "step": 80770 }, { "grad_norm": 0.4524829387664795, "learning_rate": 9.76501526388725e-06, "loss": 0.0183, "step": 80780 }, { "grad_norm": 0.3523436486721039, "learning_rate": 9.7552011102295e-06, "loss": 0.0157, "step": 80790 }, { "grad_norm": 0.9123333692550659, "learning_rate": 9.745391357682449e-06, "loss": 0.0304, "step": 80800 }, { "grad_norm": 0.8104836940765381, "learning_rate": 9.735586007318875e-06, "loss": 0.0162, "step": 80810 }, { "grad_norm": 0.4977695047855377, "learning_rate": 9.725785060211096e-06, "loss": 0.0187, "step": 80820 }, { "grad_norm": 0.5330108404159546, "learning_rate": 9.715988517430896e-06, "loss": 0.019, "step": 80830 }, { "grad_norm": 0.21305100619792938, "learning_rate": 9.706196380049625e-06, "loss": 0.0224, "step": 80840 }, { "grad_norm": 0.5865403413772583, "learning_rate": 9.69640864913815e-06, "loss": 0.0263, "step": 80850 }, { "grad_norm": 0.39984503388404846, "learning_rate": 9.686625325766818e-06, "loss": 0.0166, "step": 80860 }, { "grad_norm": 0.3280591070652008, "learning_rate": 9.67684641100554e-06, "loss": 0.0163, "step": 80870 }, { "grad_norm": 0.33382681012153625, "learning_rate": 9.667071905923692e-06, "loss": 0.0192, "step": 80880 }, { "grad_norm": 0.27062854170799255, "learning_rate": 9.657301811590242e-06, "loss": 0.0173, "step": 80890 }, { "grad_norm": 0.5131201148033142, "learning_rate": 9.647536129073603e-06, "loss": 0.0158, "step": 80900 }, { "grad_norm": 0.46748143434524536, "learning_rate": 9.637774859441751e-06, "loss": 0.021, "step": 80910 }, { "grad_norm": 0.2392088621854782, "learning_rate": 9.628018003762152e-06, "loss": 0.0286, "step": 80920 }, { "grad_norm": 0.25832998752593994, "learning_rate": 9.618265563101813e-06, "loss": 0.0158, "step": 80930 }, { "grad_norm": 0.454210102558136, "learning_rate": 9.608517538527211e-06, "loss": 0.0167, "step": 80940 }, { "grad_norm": 0.38053515553474426, "learning_rate": 9.598773931104431e-06, "loss": 0.0222, "step": 80950 }, { "grad_norm": 0.24347297847270966, "learning_rate": 9.589034741898983e-06, "loss": 0.0202, "step": 80960 }, { "grad_norm": 0.2557190954685211, "learning_rate": 9.579299971975947e-06, "loss": 0.0222, "step": 80970 }, { "grad_norm": 0.3825829327106476, "learning_rate": 9.569569622399882e-06, "loss": 0.0166, "step": 80980 }, { "grad_norm": 0.30950596928596497, "learning_rate": 9.559843694234916e-06, "loss": 0.0168, "step": 80990 }, { "grad_norm": 0.46091321110725403, "learning_rate": 9.550122188544625e-06, "loss": 0.0226, "step": 81000 }, { "grad_norm": 0.48821619153022766, "learning_rate": 9.540405106392152e-06, "loss": 0.0165, "step": 81010 }, { "grad_norm": 0.4740523099899292, "learning_rate": 9.53069244884015e-06, "loss": 0.0192, "step": 81020 }, { "grad_norm": 0.48804569244384766, "learning_rate": 9.52098421695079e-06, "loss": 0.0158, "step": 81030 }, { "grad_norm": 0.35602954030036926, "learning_rate": 9.511280411785728e-06, "loss": 0.0171, "step": 81040 }, { "grad_norm": 0.5264070630073547, "learning_rate": 9.50158103440616e-06, "loss": 0.0275, "step": 81050 }, { "grad_norm": 0.35762059688568115, "learning_rate": 9.491886085872815e-06, "loss": 0.0167, "step": 81060 }, { "grad_norm": 0.3634323179721832, "learning_rate": 9.482195567245895e-06, "loss": 0.0159, "step": 81070 }, { "grad_norm": 0.22680196166038513, "learning_rate": 9.472509479585151e-06, "loss": 0.016, "step": 81080 }, { "grad_norm": 0.5370268225669861, "learning_rate": 9.462827823949833e-06, "loss": 0.0276, "step": 81090 }, { "grad_norm": 0.40011975169181824, "learning_rate": 9.453150601398735e-06, "loss": 0.0153, "step": 81100 }, { "grad_norm": 0.48843705654144287, "learning_rate": 9.44347781299011e-06, "loss": 0.0157, "step": 81110 }, { "grad_norm": 0.23420365154743195, "learning_rate": 9.43380945978179e-06, "loss": 0.0165, "step": 81120 }, { "grad_norm": 0.7572217583656311, "learning_rate": 9.424145542831047e-06, "loss": 0.0185, "step": 81130 }, { "grad_norm": 0.3844521939754486, "learning_rate": 9.414486063194778e-06, "loss": 0.0196, "step": 81140 }, { "grad_norm": 0.4445623755455017, "learning_rate": 9.404831021929273e-06, "loss": 0.02, "step": 81150 }, { "grad_norm": 0.259189248085022, "learning_rate": 9.395180420090427e-06, "loss": 0.0149, "step": 81160 }, { "grad_norm": 0.6379006505012512, "learning_rate": 9.385534258733591e-06, "loss": 0.0228, "step": 81170 }, { "grad_norm": 0.33789777755737305, "learning_rate": 9.375892538913677e-06, "loss": 0.0213, "step": 81180 }, { "grad_norm": 0.3142228424549103, "learning_rate": 9.36625526168507e-06, "loss": 0.0263, "step": 81190 }, { "grad_norm": 0.39534252882003784, "learning_rate": 9.356622428101686e-06, "loss": 0.0179, "step": 81200 }, { "grad_norm": 0.2560976445674896, "learning_rate": 9.346994039216972e-06, "loss": 0.0174, "step": 81210 }, { "grad_norm": 0.5037955045700073, "learning_rate": 9.337370096083886e-06, "loss": 0.0198, "step": 81220 }, { "grad_norm": 0.4082530438899994, "learning_rate": 9.327750599754854e-06, "loss": 0.0154, "step": 81230 }, { "grad_norm": 0.753730297088623, "learning_rate": 9.318135551281875e-06, "loss": 0.0148, "step": 81240 }, { "grad_norm": 0.6877368092536926, "learning_rate": 9.308524951716424e-06, "loss": 0.0183, "step": 81250 }, { "grad_norm": 0.31590721011161804, "learning_rate": 9.298918802109497e-06, "loss": 0.019, "step": 81260 }, { "grad_norm": 0.36478766798973083, "learning_rate": 9.289317103511625e-06, "loss": 0.0221, "step": 81270 }, { "grad_norm": 0.38613811135292053, "learning_rate": 9.279719856972835e-06, "loss": 0.0162, "step": 81280 }, { "grad_norm": 0.574663519859314, "learning_rate": 9.270127063542645e-06, "loss": 0.0228, "step": 81290 }, { "grad_norm": 0.13408860564231873, "learning_rate": 9.260538724270124e-06, "loss": 0.0127, "step": 81300 }, { "grad_norm": 0.7165133953094482, "learning_rate": 9.250954840203852e-06, "loss": 0.0325, "step": 81310 }, { "grad_norm": 0.43502986431121826, "learning_rate": 9.241375412391878e-06, "loss": 0.0299, "step": 81320 }, { "grad_norm": 0.19271205365657806, "learning_rate": 9.23180044188181e-06, "loss": 0.0234, "step": 81330 }, { "grad_norm": 0.5807420611381531, "learning_rate": 9.222229929720755e-06, "loss": 0.03, "step": 81340 }, { "grad_norm": 0.5852134823799133, "learning_rate": 9.212663876955336e-06, "loss": 0.0148, "step": 81350 }, { "grad_norm": 0.23211927711963654, "learning_rate": 9.20310228463166e-06, "loss": 0.0149, "step": 81360 }, { "grad_norm": 0.1935117244720459, "learning_rate": 9.19354515379539e-06, "loss": 0.0195, "step": 81370 }, { "grad_norm": 0.4310298264026642, "learning_rate": 9.183992485491666e-06, "loss": 0.0157, "step": 81380 }, { "grad_norm": 0.37084293365478516, "learning_rate": 9.174444280765154e-06, "loss": 0.018, "step": 81390 }, { "grad_norm": 0.2159697413444519, "learning_rate": 9.164900540660032e-06, "loss": 0.0181, "step": 81400 }, { "grad_norm": 0.5926898121833801, "learning_rate": 9.155361266220008e-06, "loss": 0.0108, "step": 81410 }, { "grad_norm": 0.3299044370651245, "learning_rate": 9.145826458488254e-06, "loss": 0.0214, "step": 81420 }, { "grad_norm": 0.4536178410053253, "learning_rate": 9.136296118507509e-06, "loss": 0.0269, "step": 81430 }, { "grad_norm": 0.4761102795600891, "learning_rate": 9.126770247319966e-06, "loss": 0.0201, "step": 81440 }, { "grad_norm": 0.5075562596321106, "learning_rate": 9.11724884596738e-06, "loss": 0.0179, "step": 81450 }, { "grad_norm": 0.8082340359687805, "learning_rate": 9.107731915490996e-06, "loss": 0.0151, "step": 81460 }, { "grad_norm": 0.3626270890235901, "learning_rate": 9.098219456931579e-06, "loss": 0.0173, "step": 81470 }, { "grad_norm": 0.36152148246765137, "learning_rate": 9.088711471329376e-06, "loss": 0.0179, "step": 81480 }, { "grad_norm": 0.6323955059051514, "learning_rate": 9.079207959724184e-06, "loss": 0.0276, "step": 81490 }, { "grad_norm": 0.27127450704574585, "learning_rate": 9.069708923155279e-06, "loss": 0.0187, "step": 81500 }, { "grad_norm": 0.21985426545143127, "learning_rate": 9.060214362661468e-06, "loss": 0.015, "step": 81510 }, { "grad_norm": 0.587078869342804, "learning_rate": 9.050724279281064e-06, "loss": 0.0142, "step": 81520 }, { "grad_norm": 0.4361487627029419, "learning_rate": 9.041238674051894e-06, "loss": 0.0209, "step": 81530 }, { "grad_norm": 0.22763191163539886, "learning_rate": 9.031757548011277e-06, "loss": 0.0145, "step": 81540 }, { "grad_norm": 0.42635491490364075, "learning_rate": 9.022280902196051e-06, "loss": 0.0139, "step": 81550 }, { "grad_norm": 0.37279975414276123, "learning_rate": 9.012808737642592e-06, "loss": 0.0337, "step": 81560 }, { "grad_norm": 0.5059149861335754, "learning_rate": 9.003341055386739e-06, "loss": 0.0233, "step": 81570 }, { "grad_norm": 0.4995949864387512, "learning_rate": 8.993877856463862e-06, "loss": 0.0198, "step": 81580 }, { "grad_norm": 0.3893004357814789, "learning_rate": 8.984419141908856e-06, "loss": 0.0152, "step": 81590 }, { "grad_norm": 0.26133373379707336, "learning_rate": 8.974964912756117e-06, "loss": 0.0149, "step": 81600 }, { "grad_norm": 0.6930065751075745, "learning_rate": 8.965515170039523e-06, "loss": 0.0295, "step": 81610 }, { "grad_norm": 0.5566492676734924, "learning_rate": 8.956069914792514e-06, "loss": 0.014, "step": 81620 }, { "grad_norm": 0.2708582282066345, "learning_rate": 8.946629148047969e-06, "loss": 0.0164, "step": 81630 }, { "grad_norm": 0.29091086983680725, "learning_rate": 8.93719287083834e-06, "loss": 0.0273, "step": 81640 }, { "grad_norm": 0.4272240996360779, "learning_rate": 8.927761084195569e-06, "loss": 0.0145, "step": 81650 }, { "grad_norm": 0.17443832755088806, "learning_rate": 8.918333789151096e-06, "loss": 0.0122, "step": 81660 }, { "grad_norm": 0.5261386632919312, "learning_rate": 8.90891098673587e-06, "loss": 0.0242, "step": 81670 }, { "grad_norm": 0.2064923793077469, "learning_rate": 8.89949267798037e-06, "loss": 0.0209, "step": 81680 }, { "grad_norm": 0.21771834790706635, "learning_rate": 8.890078863914542e-06, "loss": 0.018, "step": 81690 }, { "grad_norm": 1.2573920488357544, "learning_rate": 8.880669545567894e-06, "loss": 0.0104, "step": 81700 }, { "grad_norm": 0.19495564699172974, "learning_rate": 8.871264723969381e-06, "loss": 0.0122, "step": 81710 }, { "grad_norm": 0.32097136974334717, "learning_rate": 8.86186440014754e-06, "loss": 0.0155, "step": 81720 }, { "grad_norm": 0.4259285628795624, "learning_rate": 8.852468575130351e-06, "loss": 0.0219, "step": 81730 }, { "grad_norm": 0.48390746116638184, "learning_rate": 8.843077249945347e-06, "loss": 0.0158, "step": 81740 }, { "grad_norm": 0.46400755643844604, "learning_rate": 8.833690425619523e-06, "loss": 0.015, "step": 81750 }, { "grad_norm": 0.4804389178752899, "learning_rate": 8.82430810317942e-06, "loss": 0.0207, "step": 81760 }, { "grad_norm": 0.24360840022563934, "learning_rate": 8.814930283651074e-06, "loss": 0.016, "step": 81770 }, { "grad_norm": 0.8142675161361694, "learning_rate": 8.805556968060048e-06, "loss": 0.0232, "step": 81780 }, { "grad_norm": 0.450227290391922, "learning_rate": 8.796188157431362e-06, "loss": 0.0133, "step": 81790 }, { "grad_norm": 0.1980019062757492, "learning_rate": 8.786823852789595e-06, "loss": 0.0186, "step": 81800 }, { "grad_norm": 0.21305806934833527, "learning_rate": 8.777464055158813e-06, "loss": 0.0234, "step": 81810 }, { "grad_norm": 0.38283970952033997, "learning_rate": 8.768108765562578e-06, "loss": 0.0189, "step": 81820 }, { "grad_norm": 0.20516863465309143, "learning_rate": 8.758757985023975e-06, "loss": 0.0152, "step": 81830 }, { "grad_norm": 0.3786364793777466, "learning_rate": 8.749411714565598e-06, "loss": 0.0133, "step": 81840 }, { "grad_norm": 0.35529953241348267, "learning_rate": 8.74006995520954e-06, "loss": 0.0172, "step": 81850 }, { "grad_norm": 1.159397840499878, "learning_rate": 8.730732707977397e-06, "loss": 0.0185, "step": 81860 }, { "grad_norm": 0.1701740324497223, "learning_rate": 8.721399973890281e-06, "loss": 0.0238, "step": 81870 }, { "grad_norm": 0.3085642158985138, "learning_rate": 8.712071753968792e-06, "loss": 0.0187, "step": 81880 }, { "grad_norm": 0.26935282349586487, "learning_rate": 8.702748049233072e-06, "loss": 0.0172, "step": 81890 }, { "grad_norm": 0.25757211446762085, "learning_rate": 8.693428860702718e-06, "loss": 0.0188, "step": 81900 }, { "grad_norm": 0.1399451047182083, "learning_rate": 8.684114189396898e-06, "loss": 0.0144, "step": 81910 }, { "grad_norm": 0.6031689047813416, "learning_rate": 8.674804036334222e-06, "loss": 0.0255, "step": 81920 }, { "grad_norm": 0.32613304257392883, "learning_rate": 8.665498402532857e-06, "loss": 0.0187, "step": 81930 }, { "grad_norm": 0.5585517883300781, "learning_rate": 8.656197289010432e-06, "loss": 0.0216, "step": 81940 }, { "grad_norm": 0.49648770689964294, "learning_rate": 8.646900696784122e-06, "loss": 0.0178, "step": 81950 }, { "grad_norm": 0.4198974668979645, "learning_rate": 8.637608626870553e-06, "loss": 0.0196, "step": 81960 }, { "grad_norm": 0.22006350755691528, "learning_rate": 8.628321080285944e-06, "loss": 0.0144, "step": 81970 }, { "grad_norm": 0.4320030212402344, "learning_rate": 8.619038058045925e-06, "loss": 0.0244, "step": 81980 }, { "grad_norm": 1.0523948669433594, "learning_rate": 8.609759561165704e-06, "loss": 0.0199, "step": 81990 }, { "grad_norm": 0.3720702826976776, "learning_rate": 8.600485590659934e-06, "loss": 0.0213, "step": 82000 }, { "grad_norm": 0.39018720388412476, "learning_rate": 8.591216147542813e-06, "loss": 0.0173, "step": 82010 }, { "grad_norm": 0.16874828934669495, "learning_rate": 8.581951232828045e-06, "loss": 0.0214, "step": 82020 }, { "grad_norm": 0.3563283383846283, "learning_rate": 8.572690847528825e-06, "loss": 0.018, "step": 82030 }, { "grad_norm": 0.34117549657821655, "learning_rate": 8.563434992657838e-06, "loss": 0.0126, "step": 82040 }, { "grad_norm": 0.22025592625141144, "learning_rate": 8.554183669227307e-06, "loss": 0.0139, "step": 82050 }, { "grad_norm": 0.8516334295272827, "learning_rate": 8.544936878248937e-06, "loss": 0.0317, "step": 82060 }, { "grad_norm": 0.39474353194236755, "learning_rate": 8.535694620733942e-06, "loss": 0.0293, "step": 82070 }, { "grad_norm": 0.40455690026283264, "learning_rate": 8.526456897693053e-06, "loss": 0.0144, "step": 82080 }, { "grad_norm": 0.18975238502025604, "learning_rate": 8.517223710136456e-06, "loss": 0.0246, "step": 82090 }, { "grad_norm": 0.2122606486082077, "learning_rate": 8.507995059073931e-06, "loss": 0.0155, "step": 82100 }, { "grad_norm": 0.3742935061454773, "learning_rate": 8.498770945514673e-06, "loss": 0.0207, "step": 82110 }, { "grad_norm": 0.36751967668533325, "learning_rate": 8.489551370467436e-06, "loss": 0.0186, "step": 82120 }, { "grad_norm": 0.2733854353427887, "learning_rate": 8.480336334940442e-06, "loss": 0.0288, "step": 82130 }, { "grad_norm": 0.21653041243553162, "learning_rate": 8.471125839941458e-06, "loss": 0.023, "step": 82140 }, { "grad_norm": 0.37230271100997925, "learning_rate": 8.461919886477682e-06, "loss": 0.0194, "step": 82150 }, { "grad_norm": 0.41445788741111755, "learning_rate": 8.452718475555927e-06, "loss": 0.02, "step": 82160 }, { "grad_norm": 0.2962915301322937, "learning_rate": 8.443521608182398e-06, "loss": 0.0244, "step": 82170 }, { "grad_norm": 0.5541384220123291, "learning_rate": 8.43432928536288e-06, "loss": 0.022, "step": 82180 }, { "grad_norm": 0.21979452669620514, "learning_rate": 8.425141508102607e-06, "loss": 0.0161, "step": 82190 }, { "grad_norm": 0.18811146914958954, "learning_rate": 8.415958277406365e-06, "loss": 0.02, "step": 82200 }, { "grad_norm": 0.4021033048629761, "learning_rate": 8.406779594278392e-06, "loss": 0.0178, "step": 82210 }, { "grad_norm": 0.22892995178699493, "learning_rate": 8.397605459722468e-06, "loss": 0.0174, "step": 82220 }, { "grad_norm": 0.4223823547363281, "learning_rate": 8.388435874741863e-06, "loss": 0.0178, "step": 82230 }, { "grad_norm": 0.25480467081069946, "learning_rate": 8.37927084033936e-06, "loss": 0.0181, "step": 82240 }, { "grad_norm": 0.2900639772415161, "learning_rate": 8.370110357517208e-06, "loss": 0.0198, "step": 82250 }, { "grad_norm": 0.40199026465415955, "learning_rate": 8.360954427277201e-06, "loss": 0.0263, "step": 82260 }, { "grad_norm": 0.2622136175632477, "learning_rate": 8.351803050620627e-06, "loss": 0.0176, "step": 82270 }, { "grad_norm": 0.5413219332695007, "learning_rate": 8.34265622854824e-06, "loss": 0.0182, "step": 82280 }, { "grad_norm": 0.4590255618095398, "learning_rate": 8.333513962060335e-06, "loss": 0.0149, "step": 82290 }, { "grad_norm": 0.2886175513267517, "learning_rate": 8.324376252156701e-06, "loss": 0.0142, "step": 82300 }, { "grad_norm": 0.4338666498661041, "learning_rate": 8.315243099836633e-06, "loss": 0.0244, "step": 82310 }, { "grad_norm": 0.2707754671573639, "learning_rate": 8.3061145060989e-06, "loss": 0.0205, "step": 82320 }, { "grad_norm": 0.25228652358055115, "learning_rate": 8.296990471941806e-06, "loss": 0.0129, "step": 82330 }, { "grad_norm": 0.42368510365486145, "learning_rate": 8.287870998363112e-06, "loss": 0.0249, "step": 82340 }, { "grad_norm": 0.2689739167690277, "learning_rate": 8.278756086360156e-06, "loss": 0.014, "step": 82350 }, { "grad_norm": 0.4017695486545563, "learning_rate": 8.269645736929693e-06, "loss": 0.0149, "step": 82360 }, { "grad_norm": 0.22992739081382751, "learning_rate": 8.260539951068048e-06, "loss": 0.0157, "step": 82370 }, { "grad_norm": 0.3277720510959625, "learning_rate": 8.251438729770983e-06, "loss": 0.0184, "step": 82380 }, { "grad_norm": 0.5646988153457642, "learning_rate": 8.242342074033826e-06, "loss": 0.0193, "step": 82390 }, { "grad_norm": 0.4961875379085541, "learning_rate": 8.233249984851344e-06, "loss": 0.0151, "step": 82400 }, { "grad_norm": 0.17732395231723785, "learning_rate": 8.224162463217849e-06, "loss": 0.0207, "step": 82410 }, { "grad_norm": 0.1846473515033722, "learning_rate": 8.215079510127139e-06, "loss": 0.011, "step": 82420 }, { "grad_norm": 0.17064301669597626, "learning_rate": 8.206001126572521e-06, "loss": 0.0197, "step": 82430 }, { "grad_norm": 1.324641227722168, "learning_rate": 8.196927313546776e-06, "loss": 0.0166, "step": 82440 }, { "grad_norm": 0.3671327531337738, "learning_rate": 8.18785807204222e-06, "loss": 0.0176, "step": 82450 }, { "grad_norm": 0.31007489562034607, "learning_rate": 8.178793403050627e-06, "loss": 0.0292, "step": 82460 }, { "grad_norm": 0.49493759870529175, "learning_rate": 8.16973330756331e-06, "loss": 0.0169, "step": 82470 }, { "grad_norm": 0.38672029972076416, "learning_rate": 8.16067778657107e-06, "loss": 0.0125, "step": 82480 }, { "grad_norm": 0.2933151125907898, "learning_rate": 8.151626841064214e-06, "loss": 0.0166, "step": 82490 }, { "grad_norm": 0.13380703330039978, "learning_rate": 8.142580472032518e-06, "loss": 0.0201, "step": 82500 }, { "grad_norm": 0.3213702440261841, "learning_rate": 8.133538680465285e-06, "loss": 0.0315, "step": 82510 }, { "grad_norm": 0.16234825551509857, "learning_rate": 8.124501467351336e-06, "loss": 0.0187, "step": 82520 }, { "grad_norm": 0.33102691173553467, "learning_rate": 8.11546883367893e-06, "loss": 0.0143, "step": 82530 }, { "grad_norm": 0.23688246309757233, "learning_rate": 8.106440780435882e-06, "loss": 0.0241, "step": 82540 }, { "grad_norm": 0.2913426160812378, "learning_rate": 8.097417308609485e-06, "loss": 0.019, "step": 82550 }, { "grad_norm": 0.2468080073595047, "learning_rate": 8.088398419186544e-06, "loss": 0.0168, "step": 82560 }, { "grad_norm": 0.22161516547203064, "learning_rate": 8.079384113153321e-06, "loss": 0.0137, "step": 82570 }, { "grad_norm": 0.3702455163002014, "learning_rate": 8.070374391495644e-06, "loss": 0.0157, "step": 82580 }, { "grad_norm": 0.3094208836555481, "learning_rate": 8.061369255198764e-06, "loss": 0.0178, "step": 82590 }, { "grad_norm": 0.43711167573928833, "learning_rate": 8.052368705247493e-06, "loss": 0.0177, "step": 82600 }, { "grad_norm": 0.3036373555660248, "learning_rate": 8.043372742626114e-06, "loss": 0.0189, "step": 82610 }, { "grad_norm": 0.28569677472114563, "learning_rate": 8.034381368318422e-06, "loss": 0.0177, "step": 82620 }, { "grad_norm": 0.22613786160945892, "learning_rate": 8.025394583307677e-06, "loss": 0.0142, "step": 82630 }, { "grad_norm": 0.20774874091148376, "learning_rate": 8.016412388576683e-06, "loss": 0.0148, "step": 82640 }, { "grad_norm": 0.4113980233669281, "learning_rate": 8.007434785107699e-06, "loss": 0.0178, "step": 82650 }, { "grad_norm": 0.6335445046424866, "learning_rate": 7.998461773882509e-06, "loss": 0.0172, "step": 82660 }, { "grad_norm": 0.17033645510673523, "learning_rate": 7.989493355882393e-06, "loss": 0.0177, "step": 82670 }, { "grad_norm": 0.47843074798583984, "learning_rate": 7.980529532088137e-06, "loss": 0.0197, "step": 82680 }, { "grad_norm": 0.25584155321121216, "learning_rate": 7.97157030347998e-06, "loss": 0.0197, "step": 82690 }, { "grad_norm": 0.4737471640110016, "learning_rate": 7.96261567103772e-06, "loss": 0.0198, "step": 82700 }, { "grad_norm": 0.1777038723230362, "learning_rate": 7.953665635740593e-06, "loss": 0.0153, "step": 82710 }, { "grad_norm": 0.33167093992233276, "learning_rate": 7.94472019856738e-06, "loss": 0.015, "step": 82720 }, { "grad_norm": 0.6530258655548096, "learning_rate": 7.935779360496337e-06, "loss": 0.0164, "step": 82730 }, { "grad_norm": 0.43859538435935974, "learning_rate": 7.92684312250523e-06, "loss": 0.0252, "step": 82740 }, { "grad_norm": 0.2515918016433716, "learning_rate": 7.917911485571289e-06, "loss": 0.0138, "step": 82750 }, { "grad_norm": 0.5285636782646179, "learning_rate": 7.90898445067128e-06, "loss": 0.0161, "step": 82760 }, { "grad_norm": 0.46955618262290955, "learning_rate": 7.900062018781462e-06, "loss": 0.0162, "step": 82770 }, { "grad_norm": 0.5287226438522339, "learning_rate": 7.891144190877553e-06, "loss": 0.0147, "step": 82780 }, { "grad_norm": 0.30242639780044556, "learning_rate": 7.882230967934806e-06, "loss": 0.0143, "step": 82790 }, { "grad_norm": 0.38148534297943115, "learning_rate": 7.873322350927959e-06, "loss": 0.0144, "step": 82800 }, { "grad_norm": 0.2959122955799103, "learning_rate": 7.864418340831253e-06, "loss": 0.0126, "step": 82810 }, { "grad_norm": 0.3164287209510803, "learning_rate": 7.855518938618395e-06, "loss": 0.0161, "step": 82820 }, { "grad_norm": 0.5934767127037048, "learning_rate": 7.846624145262633e-06, "loss": 0.021, "step": 82830 }, { "grad_norm": 0.21353067457675934, "learning_rate": 7.837733961736672e-06, "loss": 0.0133, "step": 82840 }, { "grad_norm": 0.41100260615348816, "learning_rate": 7.828848389012727e-06, "loss": 0.0175, "step": 82850 }, { "grad_norm": 0.2030002474784851, "learning_rate": 7.819967428062524e-06, "loss": 0.0186, "step": 82860 }, { "grad_norm": 0.2792662978172302, "learning_rate": 7.811091079857274e-06, "loss": 0.0173, "step": 82870 }, { "grad_norm": 0.27338820695877075, "learning_rate": 7.802219345367667e-06, "loss": 0.0097, "step": 82880 }, { "grad_norm": 0.6088329553604126, "learning_rate": 7.793352225563916e-06, "loss": 0.0217, "step": 82890 }, { "grad_norm": 0.2905958890914917, "learning_rate": 7.7844897214157e-06, "loss": 0.0189, "step": 82900 }, { "grad_norm": 0.3635396957397461, "learning_rate": 7.775631833892228e-06, "loss": 0.0224, "step": 82910 }, { "grad_norm": 0.2891480028629303, "learning_rate": 7.76677856396215e-06, "loss": 0.0195, "step": 82920 }, { "grad_norm": 0.4383794069290161, "learning_rate": 7.757929912593697e-06, "loss": 0.0171, "step": 82930 }, { "grad_norm": 0.44226914644241333, "learning_rate": 7.749085880754509e-06, "loss": 0.0155, "step": 82940 }, { "grad_norm": 0.3450576663017273, "learning_rate": 7.740246469411777e-06, "loss": 0.0186, "step": 82950 }, { "grad_norm": 4.11019229888916, "learning_rate": 7.731411679532142e-06, "loss": 0.0249, "step": 82960 }, { "grad_norm": 0.332914263010025, "learning_rate": 7.722581512081779e-06, "loss": 0.0205, "step": 82970 }, { "grad_norm": 0.5750504732131958, "learning_rate": 7.713755968026331e-06, "loss": 0.0149, "step": 82980 }, { "grad_norm": 0.41936272382736206, "learning_rate": 7.704935048330969e-06, "loss": 0.0151, "step": 82990 }, { "grad_norm": 0.23691734671592712, "learning_rate": 7.696118753960313e-06, "loss": 0.0167, "step": 83000 }, { "grad_norm": 0.3793327212333679, "learning_rate": 7.687307085878503e-06, "loss": 0.0172, "step": 83010 }, { "grad_norm": 0.48285961151123047, "learning_rate": 7.678500045049186e-06, "loss": 0.0216, "step": 83020 }, { "grad_norm": 0.4541430175304413, "learning_rate": 7.669697632435463e-06, "loss": 0.0237, "step": 83030 }, { "grad_norm": 0.5513138771057129, "learning_rate": 7.660899848999963e-06, "loss": 0.0206, "step": 83040 }, { "grad_norm": 0.4212020933628082, "learning_rate": 7.652106695704803e-06, "loss": 0.0294, "step": 83050 }, { "grad_norm": 0.2549968957901001, "learning_rate": 7.643318173511598e-06, "loss": 0.0196, "step": 83060 }, { "grad_norm": 0.37593749165534973, "learning_rate": 7.634534283381423e-06, "loss": 0.0133, "step": 83070 }, { "grad_norm": 0.5259774327278137, "learning_rate": 7.625755026274889e-06, "loss": 0.0134, "step": 83080 }, { "grad_norm": 0.23887960612773895, "learning_rate": 7.616980403152069e-06, "loss": 0.0145, "step": 83090 }, { "grad_norm": 0.11587696522474289, "learning_rate": 7.608210414972561e-06, "loss": 0.0133, "step": 83100 }, { "grad_norm": 0.30173882842063904, "learning_rate": 7.599445062695404e-06, "loss": 0.0269, "step": 83110 }, { "grad_norm": 0.18753066658973694, "learning_rate": 7.590684347279214e-06, "loss": 0.0148, "step": 83120 }, { "grad_norm": 0.688625693321228, "learning_rate": 7.581928269682004e-06, "loss": 0.0165, "step": 83130 }, { "grad_norm": 0.727269172668457, "learning_rate": 7.573176830861361e-06, "loss": 0.0322, "step": 83140 }, { "grad_norm": 0.43513184785842896, "learning_rate": 7.5644300317743e-06, "loss": 0.0178, "step": 83150 }, { "grad_norm": 0.35655874013900757, "learning_rate": 7.555687873377376e-06, "loss": 0.0258, "step": 83160 }, { "grad_norm": 0.13966715335845947, "learning_rate": 7.546950356626592e-06, "loss": 0.0182, "step": 83170 }, { "grad_norm": 0.5107025504112244, "learning_rate": 7.538217482477516e-06, "loss": 0.0215, "step": 83180 }, { "grad_norm": 0.4071812629699707, "learning_rate": 7.529489251885119e-06, "loss": 0.0205, "step": 83190 }, { "grad_norm": 0.2406563013792038, "learning_rate": 7.520765665803941e-06, "loss": 0.0165, "step": 83200 }, { "grad_norm": 0.3869665265083313, "learning_rate": 7.512046725187949e-06, "loss": 0.033, "step": 83210 }, { "grad_norm": 0.25986337661743164, "learning_rate": 7.503332430990645e-06, "loss": 0.0133, "step": 83220 }, { "grad_norm": 0.24636077880859375, "learning_rate": 7.494622784165017e-06, "loss": 0.0163, "step": 83230 }, { "grad_norm": 0.3044112026691437, "learning_rate": 7.485917785663543e-06, "loss": 0.0206, "step": 83240 }, { "grad_norm": 0.5854325890541077, "learning_rate": 7.477217436438172e-06, "loss": 0.0186, "step": 83250 }, { "grad_norm": 0.24625423550605774, "learning_rate": 7.46852173744037e-06, "loss": 0.0168, "step": 83260 }, { "grad_norm": 0.2710133492946625, "learning_rate": 7.4598306896211004e-06, "loss": 0.0138, "step": 83270 }, { "grad_norm": 0.3733016848564148, "learning_rate": 7.451144293930773e-06, "loss": 0.0122, "step": 83280 }, { "grad_norm": 0.3453444242477417, "learning_rate": 7.4424625513193455e-06, "loss": 0.0134, "step": 83290 }, { "grad_norm": 0.7077244520187378, "learning_rate": 7.433785462736209e-06, "loss": 0.0207, "step": 83300 }, { "grad_norm": 0.3236692547798157, "learning_rate": 7.425113029130315e-06, "loss": 0.0137, "step": 83310 }, { "grad_norm": 0.29735493659973145, "learning_rate": 7.416445251450044e-06, "loss": 0.0248, "step": 83320 }, { "grad_norm": 0.24901644885540009, "learning_rate": 7.4077821306433035e-06, "loss": 0.014, "step": 83330 }, { "grad_norm": 0.22623249888420105, "learning_rate": 7.399123667657465e-06, "loss": 0.018, "step": 83340 }, { "grad_norm": 0.3609360158443451, "learning_rate": 7.390469863439426e-06, "loss": 0.0156, "step": 83350 }, { "grad_norm": 0.09559573978185654, "learning_rate": 7.381820718935512e-06, "loss": 0.0212, "step": 83360 }, { "grad_norm": 0.22043322026729584, "learning_rate": 7.373176235091633e-06, "loss": 0.0189, "step": 83370 }, { "grad_norm": 0.21829275786876678, "learning_rate": 7.364536412853101e-06, "loss": 0.015, "step": 83380 }, { "grad_norm": 0.7890047430992126, "learning_rate": 7.355901253164776e-06, "loss": 0.0157, "step": 83390 }, { "grad_norm": 0.280167818069458, "learning_rate": 7.34727075697097e-06, "loss": 0.0154, "step": 83400 }, { "grad_norm": 0.3074631094932556, "learning_rate": 7.3386449252155095e-06, "loss": 0.0163, "step": 83410 }, { "grad_norm": 0.6830593943595886, "learning_rate": 7.330023758841692e-06, "loss": 0.0157, "step": 83420 }, { "grad_norm": 0.5750285983085632, "learning_rate": 7.321407258792323e-06, "loss": 0.0203, "step": 83430 }, { "grad_norm": 0.7055575251579285, "learning_rate": 7.312795426009694e-06, "loss": 0.0216, "step": 83440 }, { "grad_norm": 0.4409504234790802, "learning_rate": 7.304188261435585e-06, "loss": 0.0287, "step": 83450 }, { "grad_norm": 0.24296267330646515, "learning_rate": 7.295585766011242e-06, "loss": 0.0219, "step": 83460 }, { "grad_norm": 0.39794090390205383, "learning_rate": 7.286987940677436e-06, "loss": 0.0173, "step": 83470 }, { "grad_norm": 0.39228832721710205, "learning_rate": 7.278394786374426e-06, "loss": 0.0188, "step": 83480 }, { "grad_norm": 0.5276981592178345, "learning_rate": 7.269806304041915e-06, "loss": 0.0174, "step": 83490 }, { "grad_norm": 0.4905109107494354, "learning_rate": 7.261222494619147e-06, "loss": 0.0168, "step": 83500 }, { "grad_norm": 0.38359472155570984, "learning_rate": 7.252643359044825e-06, "loss": 0.0158, "step": 83510 }, { "grad_norm": 0.20327551662921906, "learning_rate": 7.244068898257172e-06, "loss": 0.0221, "step": 83520 }, { "grad_norm": 0.24218183755874634, "learning_rate": 7.2354991131938495e-06, "loss": 0.0161, "step": 83530 }, { "grad_norm": 0.6948850154876709, "learning_rate": 7.2269340047920574e-06, "loss": 0.0183, "step": 83540 }, { "grad_norm": 0.4231569468975067, "learning_rate": 7.218373573988435e-06, "loss": 0.0169, "step": 83550 }, { "grad_norm": 0.30552050471305847, "learning_rate": 7.209817821719178e-06, "loss": 0.0205, "step": 83560 }, { "grad_norm": 0.1734088510274887, "learning_rate": 7.201266748919894e-06, "loss": 0.0189, "step": 83570 }, { "grad_norm": 0.4464634656906128, "learning_rate": 7.19272035652575e-06, "loss": 0.0132, "step": 83580 }, { "grad_norm": 0.5407547950744629, "learning_rate": 7.184178645471334e-06, "loss": 0.0183, "step": 83590 }, { "grad_norm": 0.4172717034816742, "learning_rate": 7.175641616690775e-06, "loss": 0.0219, "step": 83600 }, { "grad_norm": 0.35742250084877014, "learning_rate": 7.167109271117661e-06, "loss": 0.0198, "step": 83610 }, { "grad_norm": 0.4416399896144867, "learning_rate": 7.158581609685072e-06, "loss": 0.0171, "step": 83620 }, { "grad_norm": 0.27325108647346497, "learning_rate": 7.150058633325591e-06, "loss": 0.0142, "step": 83630 }, { "grad_norm": 0.3337517976760864, "learning_rate": 7.141540342971287e-06, "loss": 0.0165, "step": 83640 }, { "grad_norm": 0.40213367342948914, "learning_rate": 7.133026739553677e-06, "loss": 0.0222, "step": 83650 }, { "grad_norm": 0.3695783019065857, "learning_rate": 7.124517824003829e-06, "loss": 0.0171, "step": 83660 }, { "grad_norm": 0.4077398180961609, "learning_rate": 7.116013597252236e-06, "loss": 0.0187, "step": 83670 }, { "grad_norm": 0.3741157650947571, "learning_rate": 7.107514060228921e-06, "loss": 0.0174, "step": 83680 }, { "grad_norm": 0.5080595016479492, "learning_rate": 7.099019213863378e-06, "loss": 0.0175, "step": 83690 }, { "grad_norm": 0.41049718856811523, "learning_rate": 7.090529059084611e-06, "loss": 0.0171, "step": 83700 }, { "grad_norm": 0.2800188958644867, "learning_rate": 7.082043596821058e-06, "loss": 0.016, "step": 83710 }, { "grad_norm": 0.3446486294269562, "learning_rate": 7.07356282800069e-06, "loss": 0.0325, "step": 83720 }, { "grad_norm": 0.27530285716056824, "learning_rate": 7.0650867535509645e-06, "loss": 0.0173, "step": 83730 }, { "grad_norm": 0.23758862912654877, "learning_rate": 7.0566153743987895e-06, "loss": 0.0172, "step": 83740 }, { "grad_norm": 0.1278369128704071, "learning_rate": 7.04814869147059e-06, "loss": 0.014, "step": 83750 }, { "grad_norm": 0.1578359156847, "learning_rate": 7.0396867056922725e-06, "loss": 0.0197, "step": 83760 }, { "grad_norm": 0.2530297040939331, "learning_rate": 7.031229417989244e-06, "loss": 0.0218, "step": 83770 }, { "grad_norm": 0.20963412523269653, "learning_rate": 7.022776829286348e-06, "loss": 0.0089, "step": 83780 }, { "grad_norm": 0.21617984771728516, "learning_rate": 7.014328940507975e-06, "loss": 0.022, "step": 83790 }, { "grad_norm": 0.17961972951889038, "learning_rate": 7.005885752577945e-06, "loss": 0.0185, "step": 83800 }, { "grad_norm": 0.43331143260002136, "learning_rate": 6.997447266419615e-06, "loss": 0.0194, "step": 83810 }, { "grad_norm": 0.3845294415950775, "learning_rate": 6.989013482955797e-06, "loss": 0.015, "step": 83820 }, { "grad_norm": 0.30166715383529663, "learning_rate": 6.980584403108808e-06, "loss": 0.0146, "step": 83830 }, { "grad_norm": 0.15260402858257294, "learning_rate": 6.972160027800417e-06, "loss": 0.0135, "step": 83840 }, { "grad_norm": 0.2320057451725006, "learning_rate": 6.963740357951931e-06, "loss": 0.0171, "step": 83850 }, { "grad_norm": 0.45864126086235046, "learning_rate": 6.95532539448408e-06, "loss": 0.0203, "step": 83860 }, { "grad_norm": 0.2964134216308594, "learning_rate": 6.946915138317128e-06, "loss": 0.0137, "step": 83870 }, { "grad_norm": 0.5526636838912964, "learning_rate": 6.9385095903708065e-06, "loss": 0.0157, "step": 83880 }, { "grad_norm": 0.36936795711517334, "learning_rate": 6.930108751564346e-06, "loss": 0.0205, "step": 83890 }, { "grad_norm": 0.32574206590652466, "learning_rate": 6.921712622816428e-06, "loss": 0.0275, "step": 83900 }, { "grad_norm": 0.48973193764686584, "learning_rate": 6.913321205045259e-06, "loss": 0.0184, "step": 83910 }, { "grad_norm": 0.2166285365819931, "learning_rate": 6.904934499168491e-06, "loss": 0.015, "step": 83920 }, { "grad_norm": 0.5361771583557129, "learning_rate": 6.896552506103299e-06, "loss": 0.0197, "step": 83930 }, { "grad_norm": 0.21975015103816986, "learning_rate": 6.888175226766313e-06, "loss": 0.0159, "step": 83940 }, { "grad_norm": 0.8262494802474976, "learning_rate": 6.879802662073686e-06, "loss": 0.0166, "step": 83950 }, { "grad_norm": 0.4060753583908081, "learning_rate": 6.871434812940991e-06, "loss": 0.0222, "step": 83960 }, { "grad_norm": 0.43171027302742004, "learning_rate": 6.863071680283345e-06, "loss": 0.0211, "step": 83970 }, { "grad_norm": 0.4648923873901367, "learning_rate": 6.854713265015339e-06, "loss": 0.0182, "step": 83980 }, { "grad_norm": 0.3057655692100525, "learning_rate": 6.8463595680510025e-06, "loss": 0.0203, "step": 83990 }, { "grad_norm": 0.2909715473651886, "learning_rate": 6.838010590303911e-06, "loss": 0.015, "step": 84000 }, { "grad_norm": 0.2549618184566498, "learning_rate": 6.829666332687085e-06, "loss": 0.015, "step": 84010 }, { "grad_norm": 0.18321993947029114, "learning_rate": 6.821326796113059e-06, "loss": 0.0112, "step": 84020 }, { "grad_norm": 0.18217992782592773, "learning_rate": 6.812991981493799e-06, "loss": 0.0149, "step": 84030 }, { "grad_norm": 0.4452316164970398, "learning_rate": 6.804661889740821e-06, "loss": 0.0252, "step": 84040 }, { "grad_norm": 0.3378782868385315, "learning_rate": 6.7963365217650625e-06, "loss": 0.0176, "step": 84050 }, { "grad_norm": 0.3519519865512848, "learning_rate": 6.788015878476983e-06, "loss": 0.0179, "step": 84060 }, { "grad_norm": 0.3528059720993042, "learning_rate": 6.779699960786523e-06, "loss": 0.0178, "step": 84070 }, { "grad_norm": 0.16591417789459229, "learning_rate": 6.771388769603104e-06, "loss": 0.0344, "step": 84080 }, { "grad_norm": 0.7370012998580933, "learning_rate": 6.763082305835605e-06, "loss": 0.0196, "step": 84090 }, { "grad_norm": 0.6211649179458618, "learning_rate": 6.754780570392433e-06, "loss": 0.0259, "step": 84100 }, { "grad_norm": 0.17310982942581177, "learning_rate": 6.7464835641814215e-06, "loss": 0.0147, "step": 84110 }, { "grad_norm": 0.4135955572128296, "learning_rate": 6.7381912881099515e-06, "loss": 0.0193, "step": 84120 }, { "grad_norm": 0.8606869578361511, "learning_rate": 6.72990374308482e-06, "loss": 0.019, "step": 84130 }, { "grad_norm": 0.5035006403923035, "learning_rate": 6.721620930012379e-06, "loss": 0.0181, "step": 84140 }, { "grad_norm": 0.5110532641410828, "learning_rate": 6.713342849798393e-06, "loss": 0.0142, "step": 84150 }, { "grad_norm": 0.5711697936058044, "learning_rate": 6.705069503348166e-06, "loss": 0.0239, "step": 84160 }, { "grad_norm": 0.49647456407546997, "learning_rate": 6.696800891566435e-06, "loss": 0.0173, "step": 84170 }, { "grad_norm": 0.34789490699768066, "learning_rate": 6.688537015357449e-06, "loss": 0.0162, "step": 84180 }, { "grad_norm": 0.16023142635822296, "learning_rate": 6.680277875624941e-06, "loss": 0.0204, "step": 84190 }, { "grad_norm": 0.40836164355278015, "learning_rate": 6.672023473272121e-06, "loss": 0.0258, "step": 84200 }, { "grad_norm": 0.1755717247724533, "learning_rate": 6.663773809201667e-06, "loss": 0.0203, "step": 84210 }, { "grad_norm": 0.34888720512390137, "learning_rate": 6.655528884315748e-06, "loss": 0.0202, "step": 84220 }, { "grad_norm": 1.0360769033432007, "learning_rate": 6.647288699516036e-06, "loss": 0.0251, "step": 84230 }, { "grad_norm": 0.30854153633117676, "learning_rate": 6.639053255703642e-06, "loss": 0.0164, "step": 84240 }, { "grad_norm": 0.3647730052471161, "learning_rate": 6.6308225537791925e-06, "loss": 0.0172, "step": 84250 }, { "grad_norm": 0.40470173954963684, "learning_rate": 6.622596594642777e-06, "loss": 0.0175, "step": 84260 }, { "grad_norm": 0.23830001056194305, "learning_rate": 6.614375379193999e-06, "loss": 0.0184, "step": 84270 }, { "grad_norm": 0.3777003884315491, "learning_rate": 6.606158908331889e-06, "loss": 0.0179, "step": 84280 }, { "grad_norm": 0.7815060615539551, "learning_rate": 6.597947182955005e-06, "loss": 0.0221, "step": 84290 }, { "grad_norm": 0.550084114074707, "learning_rate": 6.589740203961359e-06, "loss": 0.0233, "step": 84300 }, { "grad_norm": 0.6013808250427246, "learning_rate": 6.581537972248464e-06, "loss": 0.0147, "step": 84310 }, { "grad_norm": 0.41576847434043884, "learning_rate": 6.573340488713275e-06, "loss": 0.0159, "step": 84320 }, { "grad_norm": 0.5933488607406616, "learning_rate": 6.565147754252304e-06, "loss": 0.0209, "step": 84330 }, { "grad_norm": 0.5611593127250671, "learning_rate": 6.556959769761462e-06, "loss": 0.0202, "step": 84340 }, { "grad_norm": 0.27296626567840576, "learning_rate": 6.548776536136192e-06, "loss": 0.012, "step": 84350 }, { "grad_norm": 0.7327494025230408, "learning_rate": 6.54059805427138e-06, "loss": 0.012, "step": 84360 }, { "grad_norm": 0.3100244104862213, "learning_rate": 6.532424325061437e-06, "loss": 0.0155, "step": 84370 }, { "grad_norm": 0.34817954897880554, "learning_rate": 6.5242553494001935e-06, "loss": 0.0148, "step": 84380 }, { "grad_norm": 0.41798993945121765, "learning_rate": 6.516091128181046e-06, "loss": 0.0148, "step": 84390 }, { "grad_norm": 0.39035382866859436, "learning_rate": 6.507931662296785e-06, "loss": 0.0192, "step": 84400 }, { "grad_norm": 0.3225371837615967, "learning_rate": 6.4997769526397354e-06, "loss": 0.0179, "step": 84410 }, { "grad_norm": 0.30043911933898926, "learning_rate": 6.491627000101663e-06, "loss": 0.0173, "step": 84420 }, { "grad_norm": 0.3655754327774048, "learning_rate": 6.4834818055738545e-06, "loss": 0.0186, "step": 84430 }, { "grad_norm": 0.659666895866394, "learning_rate": 6.4753413699470465e-06, "loss": 0.0223, "step": 84440 }, { "grad_norm": 0.3993251323699951, "learning_rate": 6.467205694111478e-06, "loss": 0.013, "step": 84450 }, { "grad_norm": 0.17757418751716614, "learning_rate": 6.459074778956836e-06, "loss": 0.0194, "step": 84460 }, { "grad_norm": 0.4360688030719757, "learning_rate": 6.450948625372311e-06, "loss": 0.0181, "step": 84470 }, { "grad_norm": 0.35064470767974854, "learning_rate": 6.442827234246585e-06, "loss": 0.0195, "step": 84480 }, { "grad_norm": 0.23687158524990082, "learning_rate": 6.43471060646777e-06, "loss": 0.0156, "step": 84490 }, { "grad_norm": 0.3417745530605316, "learning_rate": 6.426598742923512e-06, "loss": 0.0167, "step": 84500 }, { "grad_norm": 0.33084815740585327, "learning_rate": 6.418491644500885e-06, "loss": 0.0253, "step": 84510 }, { "grad_norm": 0.2555789649486542, "learning_rate": 6.410389312086512e-06, "loss": 0.0183, "step": 84520 }, { "grad_norm": 0.21798287332057953, "learning_rate": 6.4022917465664135e-06, "loss": 0.0161, "step": 84530 }, { "grad_norm": 0.5473187565803528, "learning_rate": 6.3941989488261575e-06, "loss": 0.0263, "step": 84540 }, { "grad_norm": 0.28310874104499817, "learning_rate": 6.386110919750726e-06, "loss": 0.0197, "step": 84550 }, { "grad_norm": 1.4149423837661743, "learning_rate": 6.378027660224651e-06, "loss": 0.0169, "step": 84560 }, { "grad_norm": 0.15941885113716125, "learning_rate": 6.369949171131856e-06, "loss": 0.0194, "step": 84570 }, { "grad_norm": 0.3699958920478821, "learning_rate": 6.361875453355848e-06, "loss": 0.0203, "step": 84580 }, { "grad_norm": 0.2881360650062561, "learning_rate": 6.353806507779525e-06, "loss": 0.0182, "step": 84590 }, { "grad_norm": 0.23527003824710846, "learning_rate": 6.3457423352853044e-06, "loss": 0.014, "step": 84600 }, { "grad_norm": 0.25324028730392456, "learning_rate": 6.337682936755062e-06, "loss": 0.0141, "step": 84610 }, { "grad_norm": 0.510818362236023, "learning_rate": 6.329628313070179e-06, "loss": 0.0129, "step": 84620 }, { "grad_norm": 0.35066792368888855, "learning_rate": 6.321578465111477e-06, "loss": 0.0146, "step": 84630 }, { "grad_norm": 0.27480536699295044, "learning_rate": 6.31353339375928e-06, "loss": 0.0175, "step": 84640 }, { "grad_norm": 0.3355214297771454, "learning_rate": 6.305493099893395e-06, "loss": 0.0155, "step": 84650 }, { "grad_norm": 0.3578614890575409, "learning_rate": 6.297457584393096e-06, "loss": 0.0145, "step": 84660 }, { "grad_norm": 0.45387962460517883, "learning_rate": 6.289426848137126e-06, "loss": 0.0179, "step": 84670 }, { "grad_norm": 0.33916693925857544, "learning_rate": 6.281400892003713e-06, "loss": 0.0288, "step": 84680 }, { "grad_norm": 0.2932150959968567, "learning_rate": 6.273379716870581e-06, "loss": 0.0166, "step": 84690 }, { "grad_norm": 0.4278799891471863, "learning_rate": 6.265363323614887e-06, "loss": 0.0197, "step": 84700 }, { "grad_norm": 0.3548992872238159, "learning_rate": 6.257351713113308e-06, "loss": 0.0136, "step": 84710 }, { "grad_norm": 0.21469393372535706, "learning_rate": 6.249344886241981e-06, "loss": 0.0197, "step": 84720 }, { "grad_norm": 0.285558819770813, "learning_rate": 6.2413428438765295e-06, "loss": 0.0188, "step": 84730 }, { "grad_norm": 0.2717277705669403, "learning_rate": 6.233345586892026e-06, "loss": 0.018, "step": 84740 }, { "grad_norm": 0.3255700170993805, "learning_rate": 6.225353116163052e-06, "loss": 0.0218, "step": 84750 }, { "grad_norm": 0.9160884618759155, "learning_rate": 6.217365432563627e-06, "loss": 0.0182, "step": 84760 }, { "grad_norm": 0.710090160369873, "learning_rate": 6.209382536967318e-06, "loss": 0.0213, "step": 84770 }, { "grad_norm": 0.44724220037460327, "learning_rate": 6.201404430247082e-06, "loss": 0.0233, "step": 84780 }, { "grad_norm": 0.2106294184923172, "learning_rate": 6.1934311132754165e-06, "loss": 0.0184, "step": 84790 }, { "grad_norm": 0.3049359619617462, "learning_rate": 6.185462586924257e-06, "loss": 0.0173, "step": 84800 }, { "grad_norm": 0.39648693799972534, "learning_rate": 6.177498852065039e-06, "loss": 0.0159, "step": 84810 }, { "grad_norm": 0.2291363626718521, "learning_rate": 6.169539909568655e-06, "loss": 0.0215, "step": 84820 }, { "grad_norm": 0.8077457547187805, "learning_rate": 6.161585760305488e-06, "loss": 0.0184, "step": 84830 }, { "grad_norm": 0.36294490098953247, "learning_rate": 6.153636405145391e-06, "loss": 0.0147, "step": 84840 }, { "grad_norm": 0.28293079137802124, "learning_rate": 6.145691844957702e-06, "loss": 0.0227, "step": 84850 }, { "grad_norm": 0.33587726950645447, "learning_rate": 6.1377520806112145e-06, "loss": 0.0186, "step": 84860 }, { "grad_norm": 0.4116518199443817, "learning_rate": 6.129817112974218e-06, "loss": 0.0168, "step": 84870 }, { "grad_norm": 0.507637083530426, "learning_rate": 6.121886942914457e-06, "loss": 0.016, "step": 84880 }, { "grad_norm": 0.32126864790916443, "learning_rate": 6.11396157129917e-06, "loss": 0.0145, "step": 84890 }, { "grad_norm": 0.3388391137123108, "learning_rate": 6.106040998995066e-06, "loss": 0.0155, "step": 84900 }, { "grad_norm": 0.4414542019367218, "learning_rate": 6.098125226868329e-06, "loss": 0.0171, "step": 84910 }, { "grad_norm": 0.3233920931816101, "learning_rate": 6.090214255784604e-06, "loss": 0.0131, "step": 84920 }, { "grad_norm": 0.5185689926147461, "learning_rate": 6.0823080866090346e-06, "loss": 0.0147, "step": 84930 }, { "grad_norm": 0.3042111396789551, "learning_rate": 6.074406720206233e-06, "loss": 0.0171, "step": 84940 }, { "grad_norm": 0.3066103160381317, "learning_rate": 6.066510157440259e-06, "loss": 0.0151, "step": 84950 }, { "grad_norm": 0.23178806900978088, "learning_rate": 6.058618399174676e-06, "loss": 0.0092, "step": 84960 }, { "grad_norm": 0.36071136593818665, "learning_rate": 6.050731446272528e-06, "loss": 0.0303, "step": 84970 }, { "grad_norm": 0.3830828070640564, "learning_rate": 6.042849299596321e-06, "loss": 0.0185, "step": 84980 }, { "grad_norm": 0.5024453401565552, "learning_rate": 6.034971960008013e-06, "loss": 0.0197, "step": 84990 }, { "grad_norm": 0.43573760986328125, "learning_rate": 6.027099428369082e-06, "loss": 0.0204, "step": 85000 }, { "grad_norm": 0.33988064527511597, "learning_rate": 6.019231705540435e-06, "loss": 0.0179, "step": 85010 }, { "grad_norm": 0.28673216700553894, "learning_rate": 6.011368792382488e-06, "loss": 0.0167, "step": 85020 }, { "grad_norm": 0.5893987417221069, "learning_rate": 6.00351068975511e-06, "loss": 0.0188, "step": 85030 }, { "grad_norm": 0.26483139395713806, "learning_rate": 5.9956573985176674e-06, "loss": 0.0209, "step": 85040 }, { "grad_norm": 0.2784838378429413, "learning_rate": 5.987808919528959e-06, "loss": 0.0147, "step": 85050 }, { "grad_norm": 0.2985662519931793, "learning_rate": 5.979965253647307e-06, "loss": 0.0197, "step": 85060 }, { "grad_norm": 0.2498435229063034, "learning_rate": 5.9721264017304655e-06, "loss": 0.0214, "step": 85070 }, { "grad_norm": 0.1773691624403, "learning_rate": 5.964292364635682e-06, "loss": 0.0151, "step": 85080 }, { "grad_norm": 0.22456738352775574, "learning_rate": 5.956463143219682e-06, "loss": 0.015, "step": 85090 }, { "grad_norm": 0.4361017644405365, "learning_rate": 5.948638738338663e-06, "loss": 0.02, "step": 85100 }, { "grad_norm": 0.43141382932662964, "learning_rate": 5.940819150848271e-06, "loss": 0.016, "step": 85110 }, { "grad_norm": 0.3464985191822052, "learning_rate": 5.933004381603663e-06, "loss": 0.0124, "step": 85120 }, { "grad_norm": 0.40672001242637634, "learning_rate": 5.92519443145943e-06, "loss": 0.0117, "step": 85130 }, { "grad_norm": 0.43182751536369324, "learning_rate": 5.917389301269671e-06, "loss": 0.0204, "step": 85140 }, { "grad_norm": 0.1960853934288025, "learning_rate": 5.9095889918879435e-06, "loss": 0.0091, "step": 85150 }, { "grad_norm": 0.5817102789878845, "learning_rate": 5.901793504167286e-06, "loss": 0.017, "step": 85160 }, { "grad_norm": 0.5955003499984741, "learning_rate": 5.894002838960172e-06, "loss": 0.0152, "step": 85170 }, { "grad_norm": 0.48293256759643555, "learning_rate": 5.886216997118604e-06, "loss": 0.0309, "step": 85180 }, { "grad_norm": 0.18979008495807648, "learning_rate": 5.878435979494023e-06, "loss": 0.0153, "step": 85190 }, { "grad_norm": 0.3966205418109894, "learning_rate": 5.870659786937344e-06, "loss": 0.0127, "step": 85200 }, { "grad_norm": 0.35891419649124146, "learning_rate": 5.862888420298962e-06, "loss": 0.0157, "step": 85210 }, { "grad_norm": 0.4686201214790344, "learning_rate": 5.855121880428743e-06, "loss": 0.0166, "step": 85220 }, { "grad_norm": 0.3458445370197296, "learning_rate": 5.84736016817603e-06, "loss": 0.0317, "step": 85230 }, { "grad_norm": 0.6305473446846008, "learning_rate": 5.839603284389622e-06, "loss": 0.0145, "step": 85240 }, { "grad_norm": 0.3101019859313965, "learning_rate": 5.831851229917812e-06, "loss": 0.0144, "step": 85250 }, { "grad_norm": 0.35330337285995483, "learning_rate": 5.824104005608338e-06, "loss": 0.0145, "step": 85260 }, { "grad_norm": 0.24609826505184174, "learning_rate": 5.81636161230843e-06, "loss": 0.016, "step": 85270 }, { "grad_norm": 0.1850053369998932, "learning_rate": 5.808624050864791e-06, "loss": 0.0166, "step": 85280 }, { "grad_norm": 0.5592384338378906, "learning_rate": 5.800891322123592e-06, "loss": 0.0195, "step": 85290 }, { "grad_norm": 0.4050705134868622, "learning_rate": 5.793163426930454e-06, "loss": 0.017, "step": 85300 }, { "grad_norm": 0.2958143949508667, "learning_rate": 5.785440366130512e-06, "loss": 0.0252, "step": 85310 }, { "grad_norm": 0.3580373227596283, "learning_rate": 5.777722140568326e-06, "loss": 0.014, "step": 85320 }, { "grad_norm": 0.43195030093193054, "learning_rate": 5.7700087510879685e-06, "loss": 0.0147, "step": 85330 }, { "grad_norm": 0.2418360859155655, "learning_rate": 5.762300198532933e-06, "loss": 0.0098, "step": 85340 }, { "grad_norm": 0.4885574281215668, "learning_rate": 5.7545964837462606e-06, "loss": 0.0182, "step": 85350 }, { "grad_norm": 0.4774245619773865, "learning_rate": 5.746897607570384e-06, "loss": 0.0198, "step": 85360 }, { "grad_norm": 0.28681978583335876, "learning_rate": 5.7392035708472626e-06, "loss": 0.026, "step": 85370 }, { "grad_norm": 0.2868284285068512, "learning_rate": 5.731514374418278e-06, "loss": 0.015, "step": 85380 }, { "grad_norm": 0.581110954284668, "learning_rate": 5.7238300191243325e-06, "loss": 0.0188, "step": 85390 }, { "grad_norm": 0.18868549168109894, "learning_rate": 5.716150505805762e-06, "loss": 0.0197, "step": 85400 }, { "grad_norm": 0.5784862041473389, "learning_rate": 5.708475835302407e-06, "loss": 0.0228, "step": 85410 }, { "grad_norm": 0.5091700553894043, "learning_rate": 5.70080600845353e-06, "loss": 0.0189, "step": 85420 }, { "grad_norm": 0.422913521528244, "learning_rate": 5.693141026097909e-06, "loss": 0.0405, "step": 85430 }, { "grad_norm": 0.818600594997406, "learning_rate": 5.6854808890737745e-06, "loss": 0.0135, "step": 85440 }, { "grad_norm": 0.1629083901643753, "learning_rate": 5.6778255982188225e-06, "loss": 0.0153, "step": 85450 }, { "grad_norm": 0.30166056752204895, "learning_rate": 5.670175154370217e-06, "loss": 0.0177, "step": 85460 }, { "grad_norm": 0.4042821526527405, "learning_rate": 5.662529558364616e-06, "loss": 0.0167, "step": 85470 }, { "grad_norm": 0.37978261709213257, "learning_rate": 5.654888811038128e-06, "loss": 0.0136, "step": 85480 }, { "grad_norm": 0.4508659839630127, "learning_rate": 5.647252913226314e-06, "loss": 0.0186, "step": 85490 }, { "grad_norm": 0.3520616590976715, "learning_rate": 5.639621865764255e-06, "loss": 0.0175, "step": 85500 }, { "grad_norm": 0.2386196255683899, "learning_rate": 5.63199566948644e-06, "loss": 0.0129, "step": 85510 }, { "grad_norm": 0.30583611130714417, "learning_rate": 5.6243743252268835e-06, "loss": 0.0221, "step": 85520 }, { "grad_norm": 0.16178178787231445, "learning_rate": 5.616757833819009e-06, "loss": 0.0174, "step": 85530 }, { "grad_norm": 1.0129274129867554, "learning_rate": 5.609146196095794e-06, "loss": 0.0151, "step": 85540 }, { "grad_norm": 0.6284330487251282, "learning_rate": 5.6015394128896005e-06, "loss": 0.0189, "step": 85550 }, { "grad_norm": 0.22225333750247955, "learning_rate": 5.593937485032308e-06, "loss": 0.0116, "step": 85560 }, { "grad_norm": 0.3448484539985657, "learning_rate": 5.586340413355246e-06, "loss": 0.0146, "step": 85570 }, { "grad_norm": 0.22384610772132874, "learning_rate": 5.578748198689227e-06, "loss": 0.0211, "step": 85580 }, { "grad_norm": 0.39241915941238403, "learning_rate": 5.571160841864504e-06, "loss": 0.0158, "step": 85590 }, { "grad_norm": 0.1437344253063202, "learning_rate": 5.563578343710846e-06, "loss": 0.0123, "step": 85600 }, { "grad_norm": 0.34130650758743286, "learning_rate": 5.556000705057446e-06, "loss": 0.0138, "step": 85610 }, { "grad_norm": 0.4191395938396454, "learning_rate": 5.548427926732996e-06, "loss": 0.0202, "step": 85620 }, { "grad_norm": 0.5204575657844543, "learning_rate": 5.540860009565629e-06, "loss": 0.0188, "step": 85630 }, { "grad_norm": 0.6325485110282898, "learning_rate": 5.533296954382966e-06, "loss": 0.0157, "step": 85640 }, { "grad_norm": 0.12775884568691254, "learning_rate": 5.525738762012095e-06, "loss": 0.015, "step": 85650 }, { "grad_norm": 0.4764302670955658, "learning_rate": 5.518185433279577e-06, "loss": 0.0239, "step": 85660 }, { "grad_norm": 0.12566934525966644, "learning_rate": 5.510636969011407e-06, "loss": 0.0203, "step": 85670 }, { "grad_norm": 0.7547056674957275, "learning_rate": 5.503093370033096e-06, "loss": 0.0295, "step": 85680 }, { "grad_norm": 0.36119064688682556, "learning_rate": 5.495554637169603e-06, "loss": 0.0168, "step": 85690 }, { "grad_norm": 0.27408313751220703, "learning_rate": 5.488020771245328e-06, "loss": 0.0164, "step": 85700 }, { "grad_norm": 0.2029024213552475, "learning_rate": 5.480491773084195e-06, "loss": 0.0167, "step": 85710 }, { "grad_norm": 0.2824437618255615, "learning_rate": 5.472967643509525e-06, "loss": 0.0159, "step": 85720 }, { "grad_norm": 0.35525521636009216, "learning_rate": 5.46544838334419e-06, "loss": 0.0169, "step": 85730 }, { "grad_norm": 0.25449806451797485, "learning_rate": 5.4579339934104525e-06, "loss": 0.0159, "step": 85740 }, { "grad_norm": 0.16051362454891205, "learning_rate": 5.450424474530091e-06, "loss": 0.0188, "step": 85750 }, { "grad_norm": 0.28302356600761414, "learning_rate": 5.442919827524323e-06, "loss": 0.0127, "step": 85760 }, { "grad_norm": 0.4911472499370575, "learning_rate": 5.435420053213863e-06, "loss": 0.0192, "step": 85770 }, { "grad_norm": 0.4227895140647888, "learning_rate": 5.427925152418845e-06, "loss": 0.0185, "step": 85780 }, { "grad_norm": 0.333441823720932, "learning_rate": 5.420435125958934e-06, "loss": 0.0215, "step": 85790 }, { "grad_norm": 0.23497970402240753, "learning_rate": 5.412949974653209e-06, "loss": 0.0138, "step": 85800 }, { "grad_norm": 0.254851371049881, "learning_rate": 5.405469699320248e-06, "loss": 0.0144, "step": 85810 }, { "grad_norm": 0.3901461958885193, "learning_rate": 5.397994300778064e-06, "loss": 0.0133, "step": 85820 }, { "grad_norm": 0.26427116990089417, "learning_rate": 5.39052377984417e-06, "loss": 0.019, "step": 85830 }, { "grad_norm": 0.25820669531822205, "learning_rate": 5.383058137335523e-06, "loss": 0.0154, "step": 85840 }, { "grad_norm": 0.19129163026809692, "learning_rate": 5.375597374068553e-06, "loss": 0.0243, "step": 85850 }, { "grad_norm": 0.3436492681503296, "learning_rate": 5.368141490859163e-06, "loss": 0.0189, "step": 85860 }, { "grad_norm": 0.5721080303192139, "learning_rate": 5.360690488522729e-06, "loss": 0.0191, "step": 85870 }, { "grad_norm": 0.6291126012802124, "learning_rate": 5.353244367874055e-06, "loss": 0.0201, "step": 85880 }, { "grad_norm": 0.13548272848129272, "learning_rate": 5.345803129727456e-06, "loss": 0.0153, "step": 85890 }, { "grad_norm": 0.2877775728702545, "learning_rate": 5.338366774896697e-06, "loss": 0.0149, "step": 85900 }, { "grad_norm": 0.2710290849208832, "learning_rate": 5.330935304194995e-06, "loss": 0.0182, "step": 85910 }, { "grad_norm": 0.2080307900905609, "learning_rate": 5.323508718435044e-06, "loss": 0.0174, "step": 85920 }, { "grad_norm": 0.3394048511981964, "learning_rate": 5.316087018429011e-06, "loss": 0.0137, "step": 85930 }, { "grad_norm": 0.29222777485847473, "learning_rate": 5.30867020498853e-06, "loss": 0.022, "step": 85940 }, { "grad_norm": 0.20330306887626648, "learning_rate": 5.30125827892467e-06, "loss": 0.0189, "step": 85950 }, { "grad_norm": 0.324893981218338, "learning_rate": 5.293851241048015e-06, "loss": 0.0152, "step": 85960 }, { "grad_norm": 0.3860172629356384, "learning_rate": 5.286449092168555e-06, "loss": 0.0178, "step": 85970 }, { "grad_norm": 0.6130490899085999, "learning_rate": 5.2790518330958105e-06, "loss": 0.0249, "step": 85980 }, { "grad_norm": 0.2352464497089386, "learning_rate": 5.271659464638717e-06, "loss": 0.0141, "step": 85990 }, { "grad_norm": 0.42358046770095825, "learning_rate": 5.264271987605701e-06, "loss": 0.0226, "step": 86000 }, { "grad_norm": 0.3826656937599182, "learning_rate": 5.256889402804632e-06, "loss": 0.0186, "step": 86010 }, { "grad_norm": 0.3306424915790558, "learning_rate": 5.249511711042881e-06, "loss": 0.0121, "step": 86020 }, { "grad_norm": 0.8767844438552856, "learning_rate": 5.242138913127237e-06, "loss": 0.0187, "step": 86030 }, { "grad_norm": 0.40130189061164856, "learning_rate": 5.234771009863987e-06, "loss": 0.0183, "step": 86040 }, { "grad_norm": 0.4366467297077179, "learning_rate": 5.227408002058881e-06, "loss": 0.021, "step": 86050 }, { "grad_norm": 0.3140180706977844, "learning_rate": 5.220049890517126e-06, "loss": 0.0253, "step": 86060 }, { "grad_norm": 0.22681446373462677, "learning_rate": 5.212696676043377e-06, "loss": 0.0194, "step": 86070 }, { "grad_norm": 0.3686400353908539, "learning_rate": 5.205348359441798e-06, "loss": 0.0169, "step": 86080 }, { "grad_norm": 0.44170665740966797, "learning_rate": 5.198004941515966e-06, "loss": 0.0189, "step": 86090 }, { "grad_norm": 0.387338250875473, "learning_rate": 5.190666423068952e-06, "loss": 0.0158, "step": 86100 }, { "grad_norm": 0.25552892684936523, "learning_rate": 5.183332804903284e-06, "loss": 0.0118, "step": 86110 }, { "grad_norm": 0.24259567260742188, "learning_rate": 5.176004087820979e-06, "loss": 0.0257, "step": 86120 }, { "grad_norm": 1.2937977313995361, "learning_rate": 5.168680272623455e-06, "loss": 0.0214, "step": 86130 }, { "grad_norm": 0.28279852867126465, "learning_rate": 5.161361360111661e-06, "loss": 0.0137, "step": 86140 }, { "grad_norm": 0.4332692623138428, "learning_rate": 5.154047351085984e-06, "loss": 0.021, "step": 86150 }, { "grad_norm": 0.14400048553943634, "learning_rate": 5.1467382463462385e-06, "loss": 0.0154, "step": 86160 }, { "grad_norm": 0.4670834541320801, "learning_rate": 5.139434046691788e-06, "loss": 0.0209, "step": 86170 }, { "grad_norm": 0.3481953740119934, "learning_rate": 5.13213475292137e-06, "loss": 0.0135, "step": 86180 }, { "grad_norm": 0.1590672731399536, "learning_rate": 5.12484036583325e-06, "loss": 0.0134, "step": 86190 }, { "grad_norm": 0.8060299158096313, "learning_rate": 5.117550886225109e-06, "loss": 0.0179, "step": 86200 }, { "grad_norm": 1.1822481155395508, "learning_rate": 5.11026631489413e-06, "loss": 0.0202, "step": 86210 }, { "grad_norm": 0.5437448024749756, "learning_rate": 5.102986652636926e-06, "loss": 0.0222, "step": 86220 }, { "grad_norm": 0.3859602212905884, "learning_rate": 5.095711900249606e-06, "loss": 0.0185, "step": 86230 }, { "grad_norm": 0.34091222286224365, "learning_rate": 5.088442058527715e-06, "loss": 0.0188, "step": 86240 }, { "grad_norm": 1.004588007926941, "learning_rate": 5.08117712826629e-06, "loss": 0.0229, "step": 86250 }, { "grad_norm": 0.48042771220207214, "learning_rate": 5.073917110259785e-06, "loss": 0.0151, "step": 86260 }, { "grad_norm": 0.329707533121109, "learning_rate": 5.066662005302175e-06, "loss": 0.0226, "step": 86270 }, { "grad_norm": 0.3063031733036041, "learning_rate": 5.059411814186838e-06, "loss": 0.0135, "step": 86280 }, { "grad_norm": 0.4964352250099182, "learning_rate": 5.05216653770666e-06, "loss": 0.0152, "step": 86290 }, { "grad_norm": 0.3802935779094696, "learning_rate": 5.044926176653969e-06, "loss": 0.0168, "step": 86300 }, { "grad_norm": 0.2697899341583252, "learning_rate": 5.037690731820571e-06, "loss": 0.0448, "step": 86310 }, { "grad_norm": 0.6630591154098511, "learning_rate": 5.030460203997711e-06, "loss": 0.0156, "step": 86320 }, { "grad_norm": 0.18909643590450287, "learning_rate": 5.0232345939761174e-06, "loss": 0.0167, "step": 86330 }, { "grad_norm": 0.15876679122447968, "learning_rate": 5.016013902545957e-06, "loss": 0.0216, "step": 86340 }, { "grad_norm": 0.34433433413505554, "learning_rate": 5.008798130496889e-06, "loss": 0.017, "step": 86350 }, { "grad_norm": 0.4508744180202484, "learning_rate": 5.001587278618008e-06, "loss": 0.0195, "step": 86360 }, { "grad_norm": 0.38774746656417847, "learning_rate": 4.994381347697891e-06, "loss": 0.0171, "step": 86370 }, { "grad_norm": 0.40903785824775696, "learning_rate": 4.987180338524572e-06, "loss": 0.015, "step": 86380 }, { "grad_norm": 0.27744388580322266, "learning_rate": 4.9799842518855275e-06, "loss": 0.0155, "step": 86390 }, { "grad_norm": 2.333630084991455, "learning_rate": 4.9727930885677285e-06, "loss": 0.0264, "step": 86400 }, { "grad_norm": 0.43180176615715027, "learning_rate": 4.965606849357573e-06, "loss": 0.019, "step": 86410 }, { "grad_norm": 0.437872976064682, "learning_rate": 4.9584255350409445e-06, "loss": 0.0236, "step": 86420 }, { "grad_norm": 0.19857218861579895, "learning_rate": 4.951249146403175e-06, "loss": 0.0118, "step": 86430 }, { "grad_norm": 0.6457153558731079, "learning_rate": 4.9440776842290815e-06, "loss": 0.0132, "step": 86440 }, { "grad_norm": 0.4364836812019348, "learning_rate": 4.936911149302903e-06, "loss": 0.028, "step": 86450 }, { "grad_norm": 0.42645636200904846, "learning_rate": 4.929749542408379e-06, "loss": 0.0137, "step": 86460 }, { "grad_norm": 1.0261989831924438, "learning_rate": 4.922592864328673e-06, "loss": 0.0241, "step": 86470 }, { "grad_norm": 0.39634183049201965, "learning_rate": 4.915441115846436e-06, "loss": 0.0134, "step": 86480 }, { "grad_norm": 0.3097441494464874, "learning_rate": 4.908294297743782e-06, "loss": 0.0222, "step": 86490 }, { "grad_norm": 0.18919888138771057, "learning_rate": 4.901152410802273e-06, "loss": 0.03, "step": 86500 }, { "grad_norm": 0.3823181092739105, "learning_rate": 4.894015455802919e-06, "loss": 0.0229, "step": 86510 }, { "grad_norm": 0.945938766002655, "learning_rate": 4.8868834335262346e-06, "loss": 0.0195, "step": 86520 }, { "grad_norm": 0.23323440551757812, "learning_rate": 4.87975634475214e-06, "loss": 0.0189, "step": 86530 }, { "grad_norm": 0.3940768837928772, "learning_rate": 4.8726341902600614e-06, "loss": 0.0125, "step": 86540 }, { "grad_norm": 0.2587323784828186, "learning_rate": 4.865516970828843e-06, "loss": 0.0198, "step": 86550 }, { "grad_norm": 0.30997535586357117, "learning_rate": 4.85840468723685e-06, "loss": 0.0142, "step": 86560 }, { "grad_norm": 0.33892837166786194, "learning_rate": 4.85129734026184e-06, "loss": 0.0165, "step": 86570 }, { "grad_norm": 0.2591162919998169, "learning_rate": 4.844194930681073e-06, "loss": 0.0234, "step": 86580 }, { "grad_norm": 0.29741334915161133, "learning_rate": 4.837097459271267e-06, "loss": 0.012, "step": 86590 }, { "grad_norm": 0.11878369748592377, "learning_rate": 4.830004926808573e-06, "loss": 0.015, "step": 86600 }, { "grad_norm": 0.4137403666973114, "learning_rate": 4.822917334068627e-06, "loss": 0.0121, "step": 86610 }, { "grad_norm": 0.44669121503829956, "learning_rate": 4.81583468182652e-06, "loss": 0.0195, "step": 86620 }, { "grad_norm": 0.1856248825788498, "learning_rate": 4.808756970856809e-06, "loss": 0.0174, "step": 86630 }, { "grad_norm": 0.18278935551643372, "learning_rate": 4.801684201933476e-06, "loss": 0.0132, "step": 86640 }, { "grad_norm": 0.3192473351955414, "learning_rate": 4.794616375830019e-06, "loss": 0.0215, "step": 86650 }, { "grad_norm": 0.22168143093585968, "learning_rate": 4.787553493319341e-06, "loss": 0.0198, "step": 86660 }, { "grad_norm": 0.28544172644615173, "learning_rate": 4.780495555173836e-06, "loss": 0.0173, "step": 86670 }, { "grad_norm": 0.35903698205947876, "learning_rate": 4.7734425621653475e-06, "loss": 0.0192, "step": 86680 }, { "grad_norm": 0.4797835946083069, "learning_rate": 4.766394515065198e-06, "loss": 0.0189, "step": 86690 }, { "grad_norm": 0.271666944026947, "learning_rate": 4.75935141464412e-06, "loss": 0.0182, "step": 86700 }, { "grad_norm": 0.3938177525997162, "learning_rate": 4.75231326167237e-06, "loss": 0.0193, "step": 86710 }, { "grad_norm": 0.4065687358379364, "learning_rate": 4.745280056919599e-06, "loss": 0.0137, "step": 86720 }, { "grad_norm": 0.18554528057575226, "learning_rate": 4.738251801154975e-06, "loss": 0.0256, "step": 86730 }, { "grad_norm": 0.40688735246658325, "learning_rate": 4.731228495147061e-06, "loss": 0.0156, "step": 86740 }, { "grad_norm": 0.11819042265415192, "learning_rate": 4.7242101396639595e-06, "loss": 0.0143, "step": 86750 }, { "grad_norm": 0.6579478979110718, "learning_rate": 4.717196735473151e-06, "loss": 0.0213, "step": 86760 }, { "grad_norm": 0.26383253931999207, "learning_rate": 4.710188283341643e-06, "loss": 0.0148, "step": 86770 }, { "grad_norm": 0.5538223385810852, "learning_rate": 4.703184784035835e-06, "loss": 0.0203, "step": 86780 }, { "grad_norm": 0.2577878534793854, "learning_rate": 4.696186238321642e-06, "loss": 0.0159, "step": 86790 }, { "grad_norm": 0.23513367772102356, "learning_rate": 4.689192646964408e-06, "loss": 0.0111, "step": 86800 }, { "grad_norm": 0.685358464717865, "learning_rate": 4.682204010728952e-06, "loss": 0.0179, "step": 86810 }, { "grad_norm": 0.4357569217681885, "learning_rate": 4.67522033037952e-06, "loss": 0.0155, "step": 86820 }, { "grad_norm": 0.24400411546230316, "learning_rate": 4.668241606679846e-06, "loss": 0.0164, "step": 86830 }, { "grad_norm": 0.39171314239501953, "learning_rate": 4.66126784039313e-06, "loss": 0.0099, "step": 86840 }, { "grad_norm": 0.1444828361272812, "learning_rate": 4.6542990322819836e-06, "loss": 0.0148, "step": 86850 }, { "grad_norm": 0.2195301651954651, "learning_rate": 4.64733518310852e-06, "loss": 0.0168, "step": 86860 }, { "grad_norm": 0.4373542368412018, "learning_rate": 4.640376293634291e-06, "loss": 0.0282, "step": 86870 }, { "grad_norm": 0.43738412857055664, "learning_rate": 4.633422364620327e-06, "loss": 0.0133, "step": 86880 }, { "grad_norm": 0.39740562438964844, "learning_rate": 4.6264733968270745e-06, "loss": 0.0163, "step": 86890 }, { "grad_norm": 0.7563769221305847, "learning_rate": 4.619529391014482e-06, "loss": 0.0287, "step": 86900 }, { "grad_norm": 0.3172760307788849, "learning_rate": 4.61259034794192e-06, "loss": 0.0129, "step": 86910 }, { "grad_norm": 0.3619784414768219, "learning_rate": 4.605656268368247e-06, "loss": 0.0203, "step": 86920 }, { "grad_norm": 0.4399626851081848, "learning_rate": 4.598727153051735e-06, "loss": 0.0173, "step": 86930 }, { "grad_norm": 0.515919029712677, "learning_rate": 4.591803002750178e-06, "loss": 0.0145, "step": 86940 }, { "grad_norm": 0.14489801228046417, "learning_rate": 4.58488381822077e-06, "loss": 0.0092, "step": 86950 }, { "grad_norm": 0.3566429615020752, "learning_rate": 4.5779696002201946e-06, "loss": 0.0144, "step": 86960 }, { "grad_norm": 0.27153828740119934, "learning_rate": 4.571060349504563e-06, "loss": 0.0152, "step": 86970 }, { "grad_norm": 0.25500306487083435, "learning_rate": 4.564156066829478e-06, "loss": 0.0121, "step": 86980 }, { "grad_norm": 0.45980286598205566, "learning_rate": 4.557256752949957e-06, "loss": 0.0149, "step": 86990 }, { "grad_norm": 0.4973592460155487, "learning_rate": 4.550362408620534e-06, "loss": 0.0213, "step": 87000 }, { "grad_norm": 0.5302615761756897, "learning_rate": 4.543473034595136e-06, "loss": 0.0216, "step": 87010 }, { "grad_norm": 0.5866948962211609, "learning_rate": 4.536588631627192e-06, "loss": 0.0204, "step": 87020 }, { "grad_norm": 0.2189672589302063, "learning_rate": 4.5297092004695565e-06, "loss": 0.0147, "step": 87030 }, { "grad_norm": 0.5303723216056824, "learning_rate": 4.522834741874555e-06, "loss": 0.0178, "step": 87040 }, { "grad_norm": 0.3750758767127991, "learning_rate": 4.515965256593974e-06, "loss": 0.0143, "step": 87050 }, { "grad_norm": 0.2883361876010895, "learning_rate": 4.509100745379058e-06, "loss": 0.0174, "step": 87060 }, { "grad_norm": 1.0504058599472046, "learning_rate": 4.502241208980479e-06, "loss": 0.0181, "step": 87070 }, { "grad_norm": 0.3516044616699219, "learning_rate": 4.495386648148398e-06, "loss": 0.0116, "step": 87080 }, { "grad_norm": 0.24302616715431213, "learning_rate": 4.488537063632431e-06, "loss": 0.0157, "step": 87090 }, { "grad_norm": 0.35010209679603577, "learning_rate": 4.481692456181608e-06, "loss": 0.0188, "step": 87100 }, { "grad_norm": 0.2756747901439667, "learning_rate": 4.474852826544479e-06, "loss": 0.0194, "step": 87110 }, { "grad_norm": 0.2120513916015625, "learning_rate": 4.468018175468974e-06, "loss": 0.0212, "step": 87120 }, { "grad_norm": 0.25933587551116943, "learning_rate": 4.461188503702568e-06, "loss": 0.0149, "step": 87130 }, { "grad_norm": 0.36648842692375183, "learning_rate": 4.454363811992107e-06, "loss": 0.0219, "step": 87140 }, { "grad_norm": 0.3076574206352234, "learning_rate": 4.447544101083956e-06, "loss": 0.0209, "step": 87150 }, { "grad_norm": 0.19096910953521729, "learning_rate": 4.440729371723884e-06, "loss": 0.0139, "step": 87160 }, { "grad_norm": 0.16354815661907196, "learning_rate": 4.433919624657162e-06, "loss": 0.0125, "step": 87170 }, { "grad_norm": 1.062768816947937, "learning_rate": 4.427114860628462e-06, "loss": 0.018, "step": 87180 }, { "grad_norm": 0.9096742868423462, "learning_rate": 4.420315080381982e-06, "loss": 0.0166, "step": 87190 }, { "grad_norm": 0.4330659508705139, "learning_rate": 4.4135202846613045e-06, "loss": 0.0201, "step": 87200 }, { "grad_norm": 0.21025292575359344, "learning_rate": 4.406730474209525e-06, "loss": 0.0143, "step": 87210 }, { "grad_norm": 0.2688120901584625, "learning_rate": 4.399945649769138e-06, "loss": 0.0167, "step": 87220 }, { "grad_norm": 0.14942075312137604, "learning_rate": 4.393165812082151e-06, "loss": 0.0128, "step": 87230 }, { "grad_norm": 0.4138387441635132, "learning_rate": 4.386390961889969e-06, "loss": 0.0161, "step": 87240 }, { "grad_norm": 0.3212096393108368, "learning_rate": 4.3796210999334895e-06, "loss": 0.0197, "step": 87250 }, { "grad_norm": 0.21089862287044525, "learning_rate": 4.372856226953065e-06, "loss": 0.0113, "step": 87260 }, { "grad_norm": 0.31144747138023376, "learning_rate": 4.366096343688486e-06, "loss": 0.0168, "step": 87270 }, { "grad_norm": 0.28240785002708435, "learning_rate": 4.359341450878995e-06, "loss": 0.0183, "step": 87280 }, { "grad_norm": 0.2513195276260376, "learning_rate": 4.352591549263302e-06, "loss": 0.0135, "step": 87290 }, { "grad_norm": 0.32654789090156555, "learning_rate": 4.345846639579582e-06, "loss": 0.0173, "step": 87300 }, { "grad_norm": 0.31686165928840637, "learning_rate": 4.339106722565417e-06, "loss": 0.0114, "step": 87310 }, { "grad_norm": 0.20584797859191895, "learning_rate": 4.332371798957896e-06, "loss": 0.0173, "step": 87320 }, { "grad_norm": 0.29090994596481323, "learning_rate": 4.3256418694935295e-06, "loss": 0.0183, "step": 87330 }, { "grad_norm": 0.40765923261642456, "learning_rate": 4.3189169349083125e-06, "loss": 0.0201, "step": 87340 }, { "grad_norm": 0.43064889311790466, "learning_rate": 4.312196995937645e-06, "loss": 0.0175, "step": 87350 }, { "grad_norm": 0.5546737909317017, "learning_rate": 4.305482053316434e-06, "loss": 0.016, "step": 87360 }, { "grad_norm": 0.402923583984375, "learning_rate": 4.2987721077789925e-06, "loss": 0.0273, "step": 87370 }, { "grad_norm": 0.1892174333333969, "learning_rate": 4.2920671600591325e-06, "loss": 0.0162, "step": 87380 }, { "grad_norm": 0.6080377697944641, "learning_rate": 4.285367210890084e-06, "loss": 0.0184, "step": 87390 }, { "grad_norm": 0.67659991979599, "learning_rate": 4.278672261004551e-06, "loss": 0.0251, "step": 87400 }, { "grad_norm": 0.22309766709804535, "learning_rate": 4.2719823111346704e-06, "loss": 0.0126, "step": 87410 }, { "grad_norm": 1.243634819984436, "learning_rate": 4.265297362012061e-06, "loss": 0.0262, "step": 87420 }, { "grad_norm": 0.844592809677124, "learning_rate": 4.258617414367766e-06, "loss": 0.0136, "step": 87430 }, { "grad_norm": 0.585285484790802, "learning_rate": 4.251942468932296e-06, "loss": 0.0161, "step": 87440 }, { "grad_norm": 0.1614425927400589, "learning_rate": 4.2452725264356175e-06, "loss": 0.0113, "step": 87450 }, { "grad_norm": 0.4644177556037903, "learning_rate": 4.2386075876071575e-06, "loss": 0.0155, "step": 87460 }, { "grad_norm": 0.2105187475681305, "learning_rate": 4.231947653175755e-06, "loss": 0.0245, "step": 87470 }, { "grad_norm": 0.2644543945789337, "learning_rate": 4.225292723869761e-06, "loss": 0.0219, "step": 87480 }, { "grad_norm": 0.5273299813270569, "learning_rate": 4.218642800416916e-06, "loss": 0.0165, "step": 87490 }, { "grad_norm": 1.0866353511810303, "learning_rate": 4.211997883544472e-06, "loss": 0.0206, "step": 87500 }, { "grad_norm": 0.6532799601554871, "learning_rate": 4.205357973979096e-06, "loss": 0.0144, "step": 87510 }, { "grad_norm": 0.2968400716781616, "learning_rate": 4.198723072446925e-06, "loss": 0.0145, "step": 87520 }, { "grad_norm": 0.5418610572814941, "learning_rate": 4.192093179673534e-06, "loss": 0.0214, "step": 87530 }, { "grad_norm": 0.5096808671951294, "learning_rate": 4.185468296383959e-06, "loss": 0.0215, "step": 87540 }, { "grad_norm": 0.1536695659160614, "learning_rate": 4.1788484233027005e-06, "loss": 0.0189, "step": 87550 }, { "grad_norm": 0.30618178844451904, "learning_rate": 4.172233561153677e-06, "loss": 0.0168, "step": 87560 }, { "grad_norm": 0.26087525486946106, "learning_rate": 4.165623710660288e-06, "loss": 0.0165, "step": 87570 }, { "grad_norm": 0.49692854285240173, "learning_rate": 4.1590188725453835e-06, "loss": 0.0173, "step": 87580 }, { "grad_norm": 0.32023245096206665, "learning_rate": 4.152419047531259e-06, "loss": 0.0289, "step": 87590 }, { "grad_norm": 0.36358222365379333, "learning_rate": 4.145824236339651e-06, "loss": 0.0155, "step": 87600 }, { "grad_norm": 0.40542304515838623, "learning_rate": 4.1392344396917695e-06, "loss": 0.0126, "step": 87610 }, { "grad_norm": 0.5925392508506775, "learning_rate": 4.132649658308252e-06, "loss": 0.0188, "step": 87620 }, { "grad_norm": 0.41171014308929443, "learning_rate": 4.126069892909213e-06, "loss": 0.0159, "step": 87630 }, { "grad_norm": 0.6193239092826843, "learning_rate": 4.119495144214191e-06, "loss": 0.0156, "step": 87640 }, { "grad_norm": 0.23509418964385986, "learning_rate": 4.112925412942214e-06, "loss": 0.0193, "step": 87650 }, { "grad_norm": 0.3950125575065613, "learning_rate": 4.106360699811712e-06, "loss": 0.0192, "step": 87660 }, { "grad_norm": 0.27648279070854187, "learning_rate": 4.099801005540616e-06, "loss": 0.0182, "step": 87670 }, { "grad_norm": 0.3481249213218689, "learning_rate": 4.093246330846256e-06, "loss": 0.0179, "step": 87680 }, { "grad_norm": 0.13831621408462524, "learning_rate": 4.086696676445467e-06, "loss": 0.0153, "step": 87690 }, { "grad_norm": 0.2567022740840912, "learning_rate": 4.0801520430544914e-06, "loss": 0.0218, "step": 87700 }, { "grad_norm": 0.08003012090921402, "learning_rate": 4.073612431389062e-06, "loss": 0.0134, "step": 87710 }, { "grad_norm": 0.507911205291748, "learning_rate": 4.067077842164319e-06, "loss": 0.0197, "step": 87720 }, { "grad_norm": 0.3642892837524414, "learning_rate": 4.0605482760948945e-06, "loss": 0.0185, "step": 87730 }, { "grad_norm": 0.5438726544380188, "learning_rate": 4.054023733894829e-06, "loss": 0.0117, "step": 87740 }, { "grad_norm": 0.25754091143608093, "learning_rate": 4.047504216277653e-06, "loss": 0.0127, "step": 87750 }, { "grad_norm": 0.20811870694160461, "learning_rate": 4.040989723956323e-06, "loss": 0.0179, "step": 87760 }, { "grad_norm": 0.6912152767181396, "learning_rate": 4.03448025764327e-06, "loss": 0.0212, "step": 87770 }, { "grad_norm": 0.44592922925949097, "learning_rate": 4.027975818050339e-06, "loss": 0.0166, "step": 87780 }, { "grad_norm": 0.15037164092063904, "learning_rate": 4.021476405888858e-06, "loss": 0.0145, "step": 87790 }, { "grad_norm": 0.32311105728149414, "learning_rate": 4.01498202186959e-06, "loss": 0.0165, "step": 87800 }, { "grad_norm": 0.25911056995391846, "learning_rate": 4.008492666702751e-06, "loss": 0.0155, "step": 87810 }, { "grad_norm": 0.2121693640947342, "learning_rate": 4.002008341098001e-06, "loss": 0.0131, "step": 87820 }, { "grad_norm": 0.3355141878128052, "learning_rate": 3.995529045764462e-06, "loss": 0.0115, "step": 87830 }, { "grad_norm": 0.311200350522995, "learning_rate": 3.989054781410712e-06, "loss": 0.0156, "step": 87840 }, { "grad_norm": 0.33179396390914917, "learning_rate": 3.982585548744744e-06, "loss": 0.0135, "step": 87850 }, { "grad_norm": 0.3552244305610657, "learning_rate": 3.9761213484740435e-06, "loss": 0.0175, "step": 87860 }, { "grad_norm": 0.38167107105255127, "learning_rate": 3.9696621813055055e-06, "loss": 0.0126, "step": 87870 }, { "grad_norm": 0.2874738872051239, "learning_rate": 3.9632080479455044e-06, "loss": 0.016, "step": 87880 }, { "grad_norm": 0.28730782866477966, "learning_rate": 3.956758949099859e-06, "loss": 0.0234, "step": 87890 }, { "grad_norm": 0.1882532238960266, "learning_rate": 3.9503148854738385e-06, "loss": 0.0126, "step": 87900 }, { "grad_norm": 0.29804515838623047, "learning_rate": 3.943875857772139e-06, "loss": 0.0244, "step": 87910 }, { "grad_norm": 0.4350285232067108, "learning_rate": 3.9374418666989375e-06, "loss": 0.0199, "step": 87920 }, { "grad_norm": 0.7281711101531982, "learning_rate": 3.931012912957838e-06, "loss": 0.0149, "step": 87930 }, { "grad_norm": 0.44652384519577026, "learning_rate": 3.9245889972519045e-06, "loss": 0.0226, "step": 87940 }, { "grad_norm": 0.2678075134754181, "learning_rate": 3.9181701202836265e-06, "loss": 0.0167, "step": 87950 }, { "grad_norm": 0.5207602381706238, "learning_rate": 3.911756282755003e-06, "loss": 0.0229, "step": 87960 }, { "grad_norm": 0.32758182287216187, "learning_rate": 3.905347485367411e-06, "loss": 0.0174, "step": 87970 }, { "grad_norm": 0.5849791765213013, "learning_rate": 3.898943728821725e-06, "loss": 0.0328, "step": 87980 }, { "grad_norm": 0.38777267932891846, "learning_rate": 3.892545013818227e-06, "loss": 0.0172, "step": 87990 }, { "grad_norm": 0.21149875223636627, "learning_rate": 3.886151341056687e-06, "loss": 0.0184, "step": 88000 }, { "grad_norm": 0.09219079464673996, "learning_rate": 3.8797627112363066e-06, "loss": 0.011, "step": 88010 }, { "grad_norm": 0.5144064426422119, "learning_rate": 3.873379125055743e-06, "loss": 0.0168, "step": 88020 }, { "grad_norm": 0.4449234902858734, "learning_rate": 3.867000583213082e-06, "loss": 0.0163, "step": 88030 }, { "grad_norm": 0.26386067271232605, "learning_rate": 3.860627086405871e-06, "loss": 0.0144, "step": 88040 }, { "grad_norm": 0.26997870206832886, "learning_rate": 3.854258635331126e-06, "loss": 0.0137, "step": 88050 }, { "grad_norm": 0.603722333908081, "learning_rate": 3.8478952306852665e-06, "loss": 0.0218, "step": 88060 }, { "grad_norm": 0.44096699357032776, "learning_rate": 3.841536873164198e-06, "loss": 0.0125, "step": 88070 }, { "grad_norm": 0.3188951909542084, "learning_rate": 3.8351835634632626e-06, "loss": 0.0137, "step": 88080 }, { "grad_norm": 0.3135114014148712, "learning_rate": 3.828835302277245e-06, "loss": 0.0196, "step": 88090 }, { "grad_norm": 0.4936029016971588, "learning_rate": 3.822492090300378e-06, "loss": 0.0177, "step": 88100 }, { "grad_norm": 0.32877036929130554, "learning_rate": 3.8161539282263566e-06, "loss": 0.019, "step": 88110 }, { "grad_norm": 1.0729758739471436, "learning_rate": 3.8098208167482995e-06, "loss": 0.0228, "step": 88120 }, { "grad_norm": 0.5120702981948853, "learning_rate": 3.803492756558796e-06, "loss": 0.0149, "step": 88130 }, { "grad_norm": 0.26151013374328613, "learning_rate": 3.797169748349855e-06, "loss": 0.0192, "step": 88140 }, { "grad_norm": 0.4684126377105713, "learning_rate": 3.790851792812977e-06, "loss": 0.0159, "step": 88150 }, { "grad_norm": 0.4829845130443573, "learning_rate": 3.784538890639072e-06, "loss": 0.0165, "step": 88160 }, { "grad_norm": 0.2992253005504608, "learning_rate": 3.7782310425185153e-06, "loss": 0.0121, "step": 88170 }, { "grad_norm": 0.1893201768398285, "learning_rate": 3.77192824914111e-06, "loss": 0.016, "step": 88180 }, { "grad_norm": 0.447679340839386, "learning_rate": 3.7656305111961333e-06, "loss": 0.0215, "step": 88190 }, { "grad_norm": 0.2498890608549118, "learning_rate": 3.7593378293722782e-06, "loss": 0.0203, "step": 88200 }, { "grad_norm": 0.44764286279678345, "learning_rate": 3.7530502043577333e-06, "loss": 0.0249, "step": 88210 }, { "grad_norm": 0.12854669988155365, "learning_rate": 3.746767636840076e-06, "loss": 0.0112, "step": 88220 }, { "grad_norm": 0.21695999801158905, "learning_rate": 3.74049012750638e-06, "loss": 0.0146, "step": 88230 }, { "grad_norm": 0.2684004306793213, "learning_rate": 3.734217677043128e-06, "loss": 0.0162, "step": 88240 }, { "grad_norm": 0.3706987500190735, "learning_rate": 3.727950286136267e-06, "loss": 0.0162, "step": 88250 }, { "grad_norm": 0.2893572747707367, "learning_rate": 3.7216879554711925e-06, "loss": 0.0119, "step": 88260 }, { "grad_norm": 0.11257193982601166, "learning_rate": 3.715430685732757e-06, "loss": 0.0111, "step": 88270 }, { "grad_norm": 0.2718622386455536, "learning_rate": 3.7091784776052243e-06, "loss": 0.0183, "step": 88280 }, { "grad_norm": 0.33958274126052856, "learning_rate": 3.7029313317723356e-06, "loss": 0.0198, "step": 88290 }, { "grad_norm": 0.14823095500469208, "learning_rate": 3.6966892489172787e-06, "loss": 0.0144, "step": 88300 }, { "grad_norm": 0.5605679750442505, "learning_rate": 3.6904522297226575e-06, "loss": 0.0197, "step": 88310 }, { "grad_norm": 0.24687857925891876, "learning_rate": 3.6842202748705647e-06, "loss": 0.0132, "step": 88320 }, { "grad_norm": 0.34230098128318787, "learning_rate": 3.677993385042494e-06, "loss": 0.0187, "step": 88330 }, { "grad_norm": 0.3112460970878601, "learning_rate": 3.67177156091944e-06, "loss": 0.0134, "step": 88340 }, { "grad_norm": 0.6376904249191284, "learning_rate": 3.6655548031817866e-06, "loss": 0.0161, "step": 88350 }, { "grad_norm": 0.20194974541664124, "learning_rate": 3.6593431125094e-06, "loss": 0.0207, "step": 88360 }, { "grad_norm": 0.14916293323040009, "learning_rate": 3.6531364895815766e-06, "loss": 0.0199, "step": 88370 }, { "grad_norm": 0.6723553538322449, "learning_rate": 3.6469349350770677e-06, "loss": 0.0209, "step": 88380 }, { "grad_norm": 0.23784247040748596, "learning_rate": 3.6407384496740525e-06, "loss": 0.0186, "step": 88390 }, { "grad_norm": 0.34709426760673523, "learning_rate": 3.6345470340501954e-06, "loss": 0.0168, "step": 88400 }, { "grad_norm": 0.2566344141960144, "learning_rate": 3.6283606888825537e-06, "loss": 0.0151, "step": 88410 }, { "grad_norm": 0.41008028388023376, "learning_rate": 3.622179414847676e-06, "loss": 0.0111, "step": 88420 }, { "grad_norm": 0.2926207482814789, "learning_rate": 3.616003212621527e-06, "loss": 0.0201, "step": 88430 }, { "grad_norm": 0.5383453369140625, "learning_rate": 3.609832082879533e-06, "loss": 0.0174, "step": 88440 }, { "grad_norm": 0.9103260636329651, "learning_rate": 3.6036660262965485e-06, "loss": 0.0142, "step": 88450 }, { "grad_norm": 0.29326122999191284, "learning_rate": 3.597505043546895e-06, "loss": 0.0129, "step": 88460 }, { "grad_norm": 0.18211336433887482, "learning_rate": 3.5913491353043227e-06, "loss": 0.0163, "step": 88470 }, { "grad_norm": 0.432207316160202, "learning_rate": 3.5851983022420375e-06, "loss": 0.0166, "step": 88480 }, { "grad_norm": 0.2298343926668167, "learning_rate": 3.5790525450326797e-06, "loss": 0.0236, "step": 88490 }, { "grad_norm": 0.8260915875434875, "learning_rate": 3.572911864348344e-06, "loss": 0.0168, "step": 88500 }, { "grad_norm": 0.902348518371582, "learning_rate": 3.5667762608605717e-06, "loss": 0.0212, "step": 88510 }, { "grad_norm": 0.12740691006183624, "learning_rate": 3.5606457352403255e-06, "loss": 0.0124, "step": 88520 }, { "grad_norm": 0.22671306133270264, "learning_rate": 3.554520288158042e-06, "loss": 0.0201, "step": 88530 }, { "grad_norm": 0.3248681128025055, "learning_rate": 3.5483999202835903e-06, "loss": 0.0133, "step": 88540 }, { "grad_norm": 0.2937164306640625, "learning_rate": 3.5422846322862913e-06, "loss": 0.0174, "step": 88550 }, { "grad_norm": 0.3689746856689453, "learning_rate": 3.5361744248348925e-06, "loss": 0.0204, "step": 88560 }, { "grad_norm": 0.533344030380249, "learning_rate": 3.5300692985976047e-06, "loss": 0.0252, "step": 88570 }, { "grad_norm": 0.7252365946769714, "learning_rate": 3.52396925424206e-06, "loss": 0.0242, "step": 88580 }, { "grad_norm": 0.1620216816663742, "learning_rate": 3.5178742924353747e-06, "loss": 0.0122, "step": 88590 }, { "grad_norm": 0.20263198018074036, "learning_rate": 3.511784413844066e-06, "loss": 0.0214, "step": 88600 }, { "grad_norm": 0.4478342831134796, "learning_rate": 3.5056996191341287e-06, "loss": 0.0194, "step": 88610 }, { "grad_norm": 0.44442296028137207, "learning_rate": 3.4996199089709692e-06, "loss": 0.0104, "step": 88620 }, { "grad_norm": 0.24258463084697723, "learning_rate": 3.4935452840194725e-06, "loss": 0.0143, "step": 88630 }, { "grad_norm": 0.13627256453037262, "learning_rate": 3.48747574494393e-06, "loss": 0.0162, "step": 88640 }, { "grad_norm": 0.341626912355423, "learning_rate": 3.4814112924081154e-06, "loss": 0.0158, "step": 88650 }, { "grad_norm": 0.2521745264530182, "learning_rate": 3.475351927075221e-06, "loss": 0.0164, "step": 88660 }, { "grad_norm": 0.30430424213409424, "learning_rate": 3.469297649607894e-06, "loss": 0.0168, "step": 88670 }, { "grad_norm": 0.7098873257637024, "learning_rate": 3.463248460668217e-06, "loss": 0.0217, "step": 88680 }, { "grad_norm": 0.20241451263427734, "learning_rate": 3.457204360917726e-06, "loss": 0.013, "step": 88690 }, { "grad_norm": 0.49370959401130676, "learning_rate": 3.4511653510173827e-06, "loss": 0.0201, "step": 88700 }, { "grad_norm": 0.4519508481025696, "learning_rate": 3.4451314316276074e-06, "loss": 0.0219, "step": 88710 }, { "grad_norm": 0.15082918107509613, "learning_rate": 3.4391026034082674e-06, "loss": 0.0143, "step": 88720 }, { "grad_norm": 0.41023942828178406, "learning_rate": 3.433078867018674e-06, "loss": 0.0135, "step": 88730 }, { "grad_norm": 0.4408782422542572, "learning_rate": 3.427060223117551e-06, "loss": 0.0146, "step": 88740 }, { "grad_norm": 0.19654586911201477, "learning_rate": 3.4210466723631042e-06, "loss": 0.0207, "step": 88750 }, { "grad_norm": 1.0157469511032104, "learning_rate": 3.415038215412969e-06, "loss": 0.0181, "step": 88760 }, { "grad_norm": 0.3224535286426544, "learning_rate": 3.4090348529242088e-06, "loss": 0.0196, "step": 88770 }, { "grad_norm": 0.41500920057296753, "learning_rate": 3.4030365855533487e-06, "loss": 0.0142, "step": 88780 }, { "grad_norm": 0.12499409914016724, "learning_rate": 3.397043413956347e-06, "loss": 0.0183, "step": 88790 }, { "grad_norm": 0.7306079268455505, "learning_rate": 3.3910553387886237e-06, "loss": 0.0154, "step": 88800 }, { "grad_norm": 0.3071562945842743, "learning_rate": 3.3850723607049996e-06, "loss": 0.0383, "step": 88810 }, { "grad_norm": 0.5173028111457825, "learning_rate": 3.3790944803597902e-06, "loss": 0.0217, "step": 88820 }, { "grad_norm": 0.359541118144989, "learning_rate": 3.373121698406706e-06, "loss": 0.0164, "step": 88830 }, { "grad_norm": 0.27628380060195923, "learning_rate": 3.3671540154989302e-06, "loss": 0.0203, "step": 88840 }, { "grad_norm": 0.34809771180152893, "learning_rate": 3.361191432289079e-06, "loss": 0.0181, "step": 88850 }, { "grad_norm": 0.3193211257457733, "learning_rate": 3.3552339494292207e-06, "loss": 0.0154, "step": 88860 }, { "grad_norm": 0.30900293588638306, "learning_rate": 3.349281567570839e-06, "loss": 0.0132, "step": 88870 }, { "grad_norm": 0.4777858555316925, "learning_rate": 3.3433342873648965e-06, "loss": 0.0193, "step": 88880 }, { "grad_norm": 0.6287406086921692, "learning_rate": 3.337392109461762e-06, "loss": 0.0188, "step": 88890 }, { "grad_norm": 0.30077609419822693, "learning_rate": 3.331455034511266e-06, "loss": 0.0199, "step": 88900 }, { "grad_norm": 0.5550068616867065, "learning_rate": 3.325523063162689e-06, "loss": 0.0185, "step": 88910 }, { "grad_norm": 0.17094899713993073, "learning_rate": 3.3195961960647393e-06, "loss": 0.0227, "step": 88920 }, { "grad_norm": 0.3783712685108185, "learning_rate": 3.31367443386556e-06, "loss": 0.0188, "step": 88930 }, { "grad_norm": 0.24192021787166595, "learning_rate": 3.3077577772127556e-06, "loss": 0.0235, "step": 88940 }, { "grad_norm": 0.3128982186317444, "learning_rate": 3.301846226753358e-06, "loss": 0.0176, "step": 88950 }, { "grad_norm": 0.3448425531387329, "learning_rate": 3.2959397831338444e-06, "loss": 0.0233, "step": 88960 }, { "grad_norm": 0.34073615074157715, "learning_rate": 3.2900384470001432e-06, "loss": 0.013, "step": 88970 }, { "grad_norm": 0.3097648024559021, "learning_rate": 3.28414221899761e-06, "loss": 0.0138, "step": 88980 }, { "grad_norm": 0.2821953594684601, "learning_rate": 3.278251099771046e-06, "loss": 0.0161, "step": 88990 }, { "grad_norm": 0.18241344392299652, "learning_rate": 3.2723650899646906e-06, "loss": 0.0138, "step": 89000 }, { "grad_norm": 0.41723087430000305, "learning_rate": 3.2664841902222466e-06, "loss": 0.0157, "step": 89010 }, { "grad_norm": 0.39363205432891846, "learning_rate": 3.2606084011868208e-06, "loss": 0.0173, "step": 89020 }, { "grad_norm": 0.4331623613834381, "learning_rate": 3.254737723500989e-06, "loss": 0.0158, "step": 89030 }, { "grad_norm": 0.5753498077392578, "learning_rate": 3.2488721578067595e-06, "loss": 0.0198, "step": 89040 }, { "grad_norm": 0.550233781337738, "learning_rate": 3.243011704745591e-06, "loss": 0.018, "step": 89050 }, { "grad_norm": 0.43317854404449463, "learning_rate": 3.237156364958355e-06, "loss": 0.0181, "step": 89060 }, { "grad_norm": 0.500079870223999, "learning_rate": 3.231306139085394e-06, "loss": 0.0264, "step": 89070 }, { "grad_norm": 0.3330298960208893, "learning_rate": 3.2254610277664797e-06, "loss": 0.0119, "step": 89080 }, { "grad_norm": 0.3738524913787842, "learning_rate": 3.2196210316408183e-06, "loss": 0.0318, "step": 89090 }, { "grad_norm": 0.701526939868927, "learning_rate": 3.2137861513470644e-06, "loss": 0.0215, "step": 89100 }, { "grad_norm": 0.2704058289527893, "learning_rate": 3.2079563875233252e-06, "loss": 0.016, "step": 89110 }, { "grad_norm": 0.1790900081396103, "learning_rate": 3.2021317408071184e-06, "loss": 0.0195, "step": 89120 }, { "grad_norm": 0.3679516911506653, "learning_rate": 3.196312211835434e-06, "loss": 0.0231, "step": 89130 }, { "grad_norm": 0.14882655441761017, "learning_rate": 3.1904978012446686e-06, "loss": 0.0177, "step": 89140 }, { "grad_norm": 0.2627478539943695, "learning_rate": 3.184688509670691e-06, "loss": 0.0425, "step": 89150 }, { "grad_norm": 0.41314780712127686, "learning_rate": 3.178884337748783e-06, "loss": 0.0126, "step": 89160 }, { "grad_norm": 0.34141233563423157, "learning_rate": 3.173085286113697e-06, "loss": 0.0146, "step": 89170 }, { "grad_norm": 0.4085231423377991, "learning_rate": 3.1672913553995986e-06, "loss": 0.0179, "step": 89180 }, { "grad_norm": 0.19635295867919922, "learning_rate": 3.1615025462401138e-06, "loss": 0.0182, "step": 89190 }, { "grad_norm": 0.2343170940876007, "learning_rate": 3.1557188592682808e-06, "loss": 0.011, "step": 89200 }, { "grad_norm": 0.3423258066177368, "learning_rate": 3.1499402951166045e-06, "loss": 0.016, "step": 89210 }, { "grad_norm": 0.8457340598106384, "learning_rate": 3.144166854417019e-06, "loss": 0.0174, "step": 89220 }, { "grad_norm": 0.49774643778800964, "learning_rate": 3.1383985378009073e-06, "loss": 0.0106, "step": 89230 }, { "grad_norm": 0.4559340178966522, "learning_rate": 3.1326353458990654e-06, "loss": 0.0179, "step": 89240 }, { "grad_norm": 0.3867040276527405, "learning_rate": 3.126877279341761e-06, "loss": 0.0185, "step": 89250 }, { "grad_norm": 0.570643961429596, "learning_rate": 3.1211243387586964e-06, "loss": 0.0211, "step": 89260 }, { "grad_norm": 0.30013033747673035, "learning_rate": 3.115376524778979e-06, "loss": 0.0168, "step": 89270 }, { "grad_norm": 0.42807888984680176, "learning_rate": 3.109633838031195e-06, "loss": 0.02, "step": 89280 }, { "grad_norm": 0.5010614395141602, "learning_rate": 3.103896279143359e-06, "loss": 0.013, "step": 89290 }, { "grad_norm": 0.29365646839141846, "learning_rate": 3.098163848742924e-06, "loss": 0.0189, "step": 89300 }, { "grad_norm": 0.25472506880760193, "learning_rate": 3.0924365474567675e-06, "loss": 0.0159, "step": 89310 }, { "grad_norm": 0.36772364377975464, "learning_rate": 3.086714375911237e-06, "loss": 0.019, "step": 89320 }, { "grad_norm": 0.2452625334262848, "learning_rate": 3.0809973347320775e-06, "loss": 0.021, "step": 89330 }, { "grad_norm": 0.2933729290962219, "learning_rate": 3.075285424544516e-06, "loss": 0.0175, "step": 89340 }, { "grad_norm": 0.23686033487319946, "learning_rate": 3.069578645973176e-06, "loss": 0.0164, "step": 89350 }, { "grad_norm": 0.2127314954996109, "learning_rate": 3.0638769996421735e-06, "loss": 0.019, "step": 89360 }, { "grad_norm": 0.3371378481388092, "learning_rate": 3.058180486175005e-06, "loss": 0.0209, "step": 89370 }, { "grad_norm": 0.1936669945716858, "learning_rate": 3.052489106194645e-06, "loss": 0.0139, "step": 89380 }, { "grad_norm": 0.45988598465919495, "learning_rate": 3.0468028603234946e-06, "loss": 0.0175, "step": 89390 }, { "grad_norm": 0.4337140619754791, "learning_rate": 3.04112174918339e-06, "loss": 0.0256, "step": 89400 }, { "grad_norm": 0.13195937871932983, "learning_rate": 3.0354457733955954e-06, "loss": 0.013, "step": 89410 }, { "grad_norm": 0.45195192098617554, "learning_rate": 3.029774933580859e-06, "loss": 0.0175, "step": 89420 }, { "grad_norm": 0.7527593970298767, "learning_rate": 3.0241092303593065e-06, "loss": 0.0128, "step": 89430 }, { "grad_norm": 0.22068074345588684, "learning_rate": 3.018448664350554e-06, "loss": 0.0141, "step": 89440 }, { "grad_norm": 0.21646058559417725, "learning_rate": 3.0127932361736055e-06, "loss": 0.0152, "step": 89450 }, { "grad_norm": 0.2591570317745209, "learning_rate": 3.00714294644695e-06, "loss": 0.0172, "step": 89460 }, { "grad_norm": 0.241019144654274, "learning_rate": 3.0014977957884828e-06, "loss": 0.0364, "step": 89470 }, { "grad_norm": 0.24578380584716797, "learning_rate": 2.99585778481557e-06, "loss": 0.0135, "step": 89480 }, { "grad_norm": 0.502297043800354, "learning_rate": 2.990222914144969e-06, "loss": 0.0187, "step": 89490 }, { "grad_norm": 0.27181780338287354, "learning_rate": 2.984593184392914e-06, "loss": 0.024, "step": 89500 }, { "grad_norm": 0.6810826659202576, "learning_rate": 2.978968596175069e-06, "loss": 0.0212, "step": 89510 }, { "grad_norm": 0.3955575227737427, "learning_rate": 2.9733491501065137e-06, "loss": 0.0219, "step": 89520 }, { "grad_norm": 0.32859110832214355, "learning_rate": 2.9677348468018005e-06, "loss": 0.0194, "step": 89530 }, { "grad_norm": 0.2574378252029419, "learning_rate": 2.9621256868748726e-06, "loss": 0.0179, "step": 89540 }, { "grad_norm": 0.4643511176109314, "learning_rate": 2.956521670939183e-06, "loss": 0.0212, "step": 89550 }, { "grad_norm": 0.2778986990451813, "learning_rate": 2.950922799607542e-06, "loss": 0.0174, "step": 89560 }, { "grad_norm": 0.49089035391807556, "learning_rate": 2.9453290734922535e-06, "loss": 0.0418, "step": 89570 }, { "grad_norm": 0.2020704448223114, "learning_rate": 2.939740493205029e-06, "loss": 0.015, "step": 89580 }, { "grad_norm": 0.4701421856880188, "learning_rate": 2.9341570593570345e-06, "loss": 0.0232, "step": 89590 }, { "grad_norm": 0.3223499655723572, "learning_rate": 2.928578772558843e-06, "loss": 0.0159, "step": 89600 }, { "grad_norm": 0.5606140494346619, "learning_rate": 2.923005633420528e-06, "loss": 0.0156, "step": 89610 }, { "grad_norm": 0.642084002494812, "learning_rate": 2.9174376425515237e-06, "loss": 0.0231, "step": 89620 }, { "grad_norm": 0.21378576755523682, "learning_rate": 2.91187480056076e-06, "loss": 0.0169, "step": 89630 }, { "grad_norm": 0.16297361254692078, "learning_rate": 2.9063171080565664e-06, "loss": 0.0314, "step": 89640 }, { "grad_norm": 0.29772239923477173, "learning_rate": 2.900764565646741e-06, "loss": 0.0167, "step": 89650 }, { "grad_norm": 0.1718413382768631, "learning_rate": 2.895217173938475e-06, "loss": 0.0273, "step": 89660 }, { "grad_norm": 0.14619764685630798, "learning_rate": 2.8896749335384445e-06, "loss": 0.0107, "step": 89670 }, { "grad_norm": 0.4102941155433655, "learning_rate": 2.884137845052731e-06, "loss": 0.017, "step": 89680 }, { "grad_norm": 0.51876300573349, "learning_rate": 2.8786059090868734e-06, "loss": 0.0197, "step": 89690 }, { "grad_norm": 0.32729652523994446, "learning_rate": 2.87307912624582e-06, "loss": 0.0161, "step": 89700 }, { "grad_norm": 0.3328390121459961, "learning_rate": 2.867557497133977e-06, "loss": 0.0109, "step": 89710 }, { "grad_norm": 0.46945807337760925, "learning_rate": 2.862041022355194e-06, "loss": 0.0127, "step": 89720 }, { "grad_norm": 0.7398070693016052, "learning_rate": 2.8565297025127226e-06, "loss": 0.0247, "step": 89730 }, { "grad_norm": 0.2504716217517853, "learning_rate": 2.8510235382092853e-06, "loss": 0.0219, "step": 89740 }, { "grad_norm": 0.5541175007820129, "learning_rate": 2.845522530047029e-06, "loss": 0.014, "step": 89750 }, { "grad_norm": 0.4287507236003876, "learning_rate": 2.8400266786275387e-06, "loss": 0.0151, "step": 89760 }, { "grad_norm": 0.39290812611579895, "learning_rate": 2.834535984551817e-06, "loss": 0.0182, "step": 89770 }, { "grad_norm": 0.4720712900161743, "learning_rate": 2.8290504484203397e-06, "loss": 0.0214, "step": 89780 }, { "grad_norm": 0.227116197347641, "learning_rate": 2.8235700708329705e-06, "loss": 0.0145, "step": 89790 }, { "grad_norm": 0.4443991780281067, "learning_rate": 2.818094852389058e-06, "loss": 0.0187, "step": 89800 }, { "grad_norm": 0.08002691715955734, "learning_rate": 2.8126247936873516e-06, "loss": 0.0149, "step": 89810 }, { "grad_norm": 0.18433523178100586, "learning_rate": 2.8071598953260614e-06, "loss": 0.0183, "step": 89820 }, { "grad_norm": 0.39983147382736206, "learning_rate": 2.8017001579028035e-06, "loss": 0.0181, "step": 89830 }, { "grad_norm": 0.6816833019256592, "learning_rate": 2.7962455820146617e-06, "loss": 0.0224, "step": 89840 }, { "grad_norm": 0.2570696175098419, "learning_rate": 2.7907961682581253e-06, "loss": 0.0186, "step": 89850 }, { "grad_norm": 0.14872920513153076, "learning_rate": 2.7853519172291453e-06, "loss": 0.0087, "step": 89860 }, { "grad_norm": 0.685863196849823, "learning_rate": 2.7799128295230904e-06, "loss": 0.0207, "step": 89870 }, { "grad_norm": 0.4843114912509918, "learning_rate": 2.774478905734784e-06, "loss": 0.0182, "step": 89880 }, { "grad_norm": 0.5410098433494568, "learning_rate": 2.7690501464584516e-06, "loss": 0.022, "step": 89890 }, { "grad_norm": 0.5221163630485535, "learning_rate": 2.763626552287796e-06, "loss": 0.0121, "step": 89900 }, { "grad_norm": 0.5579643249511719, "learning_rate": 2.7582081238159083e-06, "loss": 0.0177, "step": 89910 }, { "grad_norm": 0.35701730847358704, "learning_rate": 2.7527948616353548e-06, "loss": 0.0152, "step": 89920 }, { "grad_norm": 0.28765231370925903, "learning_rate": 2.7473867663381226e-06, "loss": 0.013, "step": 89930 }, { "grad_norm": 0.5870562791824341, "learning_rate": 2.7419838385156383e-06, "loss": 0.0263, "step": 89940 }, { "grad_norm": 0.5282362699508667, "learning_rate": 2.7365860787587407e-06, "loss": 0.017, "step": 89950 }, { "grad_norm": 0.23781616985797882, "learning_rate": 2.7311934876577295e-06, "loss": 0.0227, "step": 89960 }, { "grad_norm": 0.3299820125102997, "learning_rate": 2.725806065802339e-06, "loss": 0.0126, "step": 89970 }, { "grad_norm": 0.22869417071342468, "learning_rate": 2.720423813781714e-06, "loss": 0.018, "step": 89980 }, { "grad_norm": 0.37389785051345825, "learning_rate": 2.715046732184462e-06, "loss": 0.0132, "step": 89990 }, { "grad_norm": 0.36535441875457764, "learning_rate": 2.7096748215986013e-06, "loss": 0.0244, "step": 90000 }, { "grad_norm": 0.3753446638584137, "learning_rate": 2.704308082611612e-06, "loss": 0.0104, "step": 90010 }, { "grad_norm": 0.4895082116127014, "learning_rate": 2.698946515810374e-06, "loss": 0.0153, "step": 90020 }, { "grad_norm": 0.5107182860374451, "learning_rate": 2.6935901217812367e-06, "loss": 0.0168, "step": 90030 }, { "grad_norm": 0.31799793243408203, "learning_rate": 2.688238901109952e-06, "loss": 0.0268, "step": 90040 }, { "grad_norm": 0.2931841313838959, "learning_rate": 2.68289285438173e-06, "loss": 0.0131, "step": 90050 }, { "grad_norm": 0.40916523337364197, "learning_rate": 2.6775519821812033e-06, "loss": 0.0161, "step": 90060 }, { "grad_norm": 0.33037927746772766, "learning_rate": 2.672216285092449e-06, "loss": 0.0171, "step": 90070 }, { "grad_norm": 0.24939702451229095, "learning_rate": 2.666885763698962e-06, "loss": 0.0142, "step": 90080 }, { "grad_norm": 0.34969982504844666, "learning_rate": 2.661560418583692e-06, "loss": 0.0113, "step": 90090 }, { "grad_norm": 0.3116450607776642, "learning_rate": 2.656240250328995e-06, "loss": 0.0175, "step": 90100 }, { "grad_norm": 0.40267354249954224, "learning_rate": 2.650925259516679e-06, "loss": 0.0211, "step": 90110 }, { "grad_norm": 0.4051710367202759, "learning_rate": 2.645615446727995e-06, "loss": 0.0149, "step": 90120 }, { "grad_norm": 0.41301846504211426, "learning_rate": 2.6403108125436115e-06, "loss": 0.019, "step": 90130 }, { "grad_norm": 0.3338923156261444, "learning_rate": 2.6350113575436264e-06, "loss": 0.0206, "step": 90140 }, { "grad_norm": 0.3054888844490051, "learning_rate": 2.629717082307598e-06, "loss": 0.0115, "step": 90150 }, { "grad_norm": 0.378085732460022, "learning_rate": 2.6244279874144793e-06, "loss": 0.0139, "step": 90160 }, { "grad_norm": 0.25594690442085266, "learning_rate": 2.6191440734426862e-06, "loss": 0.0165, "step": 90170 }, { "grad_norm": 0.32119476795196533, "learning_rate": 2.613865340970062e-06, "loss": 0.0146, "step": 90180 }, { "grad_norm": 0.4887717366218567, "learning_rate": 2.608591790573889e-06, "loss": 0.0172, "step": 90190 }, { "grad_norm": 0.46660417318344116, "learning_rate": 2.603323422830861e-06, "loss": 0.0148, "step": 90200 }, { "grad_norm": 0.6275385618209839, "learning_rate": 2.5980602383171183e-06, "loss": 0.019, "step": 90210 }, { "grad_norm": 0.18716979026794434, "learning_rate": 2.5928022376082496e-06, "loss": 0.0153, "step": 90220 }, { "grad_norm": 0.6245133876800537, "learning_rate": 2.5875494212792505e-06, "loss": 0.0165, "step": 90230 }, { "grad_norm": 0.27798670530319214, "learning_rate": 2.5823017899045565e-06, "loss": 0.0178, "step": 90240 }, { "grad_norm": 0.41356363892555237, "learning_rate": 2.577059344058047e-06, "loss": 0.0152, "step": 90250 }, { "grad_norm": 0.3207850456237793, "learning_rate": 2.5718220843130415e-06, "loss": 0.0126, "step": 90260 }, { "grad_norm": 0.17209647595882416, "learning_rate": 2.5665900112422535e-06, "loss": 0.0166, "step": 90270 }, { "grad_norm": 0.3390037417411804, "learning_rate": 2.561363125417876e-06, "loss": 0.0098, "step": 90280 }, { "grad_norm": 0.318011999130249, "learning_rate": 2.556141427411496e-06, "loss": 0.0111, "step": 90290 }, { "grad_norm": 0.22426065802574158, "learning_rate": 2.550924917794156e-06, "loss": 0.0141, "step": 90300 }, { "grad_norm": 0.3398592174053192, "learning_rate": 2.545713597136329e-06, "loss": 0.0151, "step": 90310 }, { "grad_norm": 0.4041862189769745, "learning_rate": 2.5405074660079298e-06, "loss": 0.0351, "step": 90320 }, { "grad_norm": 1.3889405727386475, "learning_rate": 2.5353065249782647e-06, "loss": 0.0211, "step": 90330 }, { "grad_norm": 0.43818220496177673, "learning_rate": 2.530110774616129e-06, "loss": 0.0217, "step": 90340 }, { "grad_norm": 0.6463175415992737, "learning_rate": 2.524920215489701e-06, "loss": 0.0162, "step": 90350 }, { "grad_norm": 0.14018653333187103, "learning_rate": 2.519734848166627e-06, "loss": 0.017, "step": 90360 }, { "grad_norm": 0.6982580423355103, "learning_rate": 2.514554673213948e-06, "loss": 0.0237, "step": 90370 }, { "grad_norm": 0.7365248203277588, "learning_rate": 2.5093796911981936e-06, "loss": 0.0169, "step": 90380 }, { "grad_norm": 0.3202887773513794, "learning_rate": 2.5042099026852672e-06, "loss": 0.0166, "step": 90390 }, { "grad_norm": 0.9171807169914246, "learning_rate": 2.499045308240544e-06, "loss": 0.0143, "step": 90400 }, { "grad_norm": 0.4257568418979645, "learning_rate": 2.4938859084288003e-06, "loss": 0.0178, "step": 90410 }, { "grad_norm": 0.8381681442260742, "learning_rate": 2.4887317038142734e-06, "loss": 0.0169, "step": 90420 }, { "grad_norm": 0.4120793342590332, "learning_rate": 2.4835826949606124e-06, "loss": 0.0129, "step": 90430 }, { "grad_norm": 0.5279366374015808, "learning_rate": 2.4784388824309167e-06, "loss": 0.0186, "step": 90440 }, { "grad_norm": 0.1617823839187622, "learning_rate": 2.473300266787687e-06, "loss": 0.014, "step": 90450 }, { "grad_norm": 0.44073498249053955, "learning_rate": 2.46816684859289e-06, "loss": 0.0141, "step": 90460 }, { "grad_norm": 0.2200775444507599, "learning_rate": 2.4630386284079043e-06, "loss": 0.0112, "step": 90470 }, { "grad_norm": 0.18731774389743805, "learning_rate": 2.4579156067935427e-06, "loss": 0.0219, "step": 90480 }, { "grad_norm": 0.49405044317245483, "learning_rate": 2.452797784310046e-06, "loss": 0.0172, "step": 90490 }, { "grad_norm": 0.22341640293598175, "learning_rate": 2.447685161517105e-06, "loss": 0.0142, "step": 90500 }, { "grad_norm": 0.3555142879486084, "learning_rate": 2.4425777389738224e-06, "loss": 0.012, "step": 90510 }, { "grad_norm": 0.225869819521904, "learning_rate": 2.43747551723873e-06, "loss": 0.0176, "step": 90520 }, { "grad_norm": 0.3292277455329895, "learning_rate": 2.4323784968698136e-06, "loss": 0.02, "step": 90530 }, { "grad_norm": 0.23987576365470886, "learning_rate": 2.4272866784244565e-06, "loss": 0.0105, "step": 90540 }, { "grad_norm": 0.22756914794445038, "learning_rate": 2.422200062459518e-06, "loss": 0.0152, "step": 90550 }, { "grad_norm": 0.3674112856388092, "learning_rate": 2.41711864953123e-06, "loss": 0.016, "step": 90560 }, { "grad_norm": 0.1374695599079132, "learning_rate": 2.412042440195322e-06, "loss": 0.0182, "step": 90570 }, { "grad_norm": 0.12650443613529205, "learning_rate": 2.4069714350068993e-06, "loss": 0.0206, "step": 90580 }, { "grad_norm": 0.38892418146133423, "learning_rate": 2.4019056345205294e-06, "loss": 0.0302, "step": 90590 }, { "grad_norm": 0.5100664496421814, "learning_rate": 2.396845039290191e-06, "loss": 0.018, "step": 90600 }, { "grad_norm": 0.31323662400245667, "learning_rate": 2.3917896498693194e-06, "loss": 0.019, "step": 90610 }, { "grad_norm": 0.3607114851474762, "learning_rate": 2.3867394668107335e-06, "loss": 0.0305, "step": 90620 }, { "grad_norm": 0.3054001033306122, "learning_rate": 2.3816944906667527e-06, "loss": 0.0128, "step": 90630 }, { "grad_norm": 0.20722946524620056, "learning_rate": 2.3766547219890634e-06, "loss": 0.0111, "step": 90640 }, { "grad_norm": 0.38382744789123535, "learning_rate": 2.371620161328819e-06, "loss": 0.0166, "step": 90650 }, { "grad_norm": 0.3487527370452881, "learning_rate": 2.366590809236574e-06, "loss": 0.0119, "step": 90660 }, { "grad_norm": 0.22471581399440765, "learning_rate": 2.3615666662623494e-06, "loss": 0.0185, "step": 90670 }, { "grad_norm": 0.3399878442287445, "learning_rate": 2.3565477329555664e-06, "loss": 0.0162, "step": 90680 }, { "grad_norm": 0.5006915926933289, "learning_rate": 2.3515340098650975e-06, "loss": 0.0228, "step": 90690 }, { "grad_norm": 0.19484947621822357, "learning_rate": 2.3465254975392258e-06, "loss": 0.0146, "step": 90700 }, { "grad_norm": 0.38179659843444824, "learning_rate": 2.3415221965256805e-06, "loss": 0.029, "step": 90710 }, { "grad_norm": 0.5192692279815674, "learning_rate": 2.3365241073716237e-06, "loss": 0.0278, "step": 90720 }, { "grad_norm": 0.18287469446659088, "learning_rate": 2.3315312306236183e-06, "loss": 0.0158, "step": 90730 }, { "grad_norm": 0.20332852005958557, "learning_rate": 2.326543566827699e-06, "loss": 0.0147, "step": 90740 }, { "grad_norm": 0.4183402955532074, "learning_rate": 2.3215611165292862e-06, "loss": 0.0131, "step": 90750 }, { "grad_norm": 0.41866087913513184, "learning_rate": 2.3165838802732765e-06, "loss": 0.0128, "step": 90760 }, { "grad_norm": 0.5981538891792297, "learning_rate": 2.311611858603957e-06, "loss": 0.0151, "step": 90770 }, { "grad_norm": 0.7624903917312622, "learning_rate": 2.3066450520650763e-06, "loss": 0.0214, "step": 90780 }, { "grad_norm": 0.29705238342285156, "learning_rate": 2.3016834611997773e-06, "loss": 0.0135, "step": 90790 }, { "grad_norm": 0.119473896920681, "learning_rate": 2.296727086550665e-06, "loss": 0.0098, "step": 90800 }, { "grad_norm": 0.21056969463825226, "learning_rate": 2.2917759286597507e-06, "loss": 0.0157, "step": 90810 }, { "grad_norm": 0.1335892230272293, "learning_rate": 2.2868299880685006e-06, "loss": 0.0197, "step": 90820 }, { "grad_norm": 0.26428037881851196, "learning_rate": 2.2818892653177827e-06, "loss": 0.0156, "step": 90830 }, { "grad_norm": 0.6081991791725159, "learning_rate": 2.2769537609479196e-06, "loss": 0.0122, "step": 90840 }, { "grad_norm": 0.3389163315296173, "learning_rate": 2.2720234754986357e-06, "loss": 0.0116, "step": 90850 }, { "grad_norm": 0.3839074671268463, "learning_rate": 2.2670984095091108e-06, "loss": 0.0228, "step": 90860 }, { "grad_norm": 0.8397493362426758, "learning_rate": 2.262178563517936e-06, "loss": 0.0254, "step": 90870 }, { "grad_norm": 0.39552560448646545, "learning_rate": 2.257263938063137e-06, "loss": 0.0135, "step": 90880 }, { "grad_norm": 0.23107635974884033, "learning_rate": 2.2523545336821715e-06, "loss": 0.0222, "step": 90890 }, { "grad_norm": 0.3468995690345764, "learning_rate": 2.247450350911939e-06, "loss": 0.0235, "step": 90900 }, { "grad_norm": 0.6331632733345032, "learning_rate": 2.2425513902887328e-06, "loss": 0.0215, "step": 90910 }, { "grad_norm": 0.4198603630065918, "learning_rate": 2.2376576523483015e-06, "loss": 0.0183, "step": 90920 }, { "grad_norm": 0.39329469203948975, "learning_rate": 2.232769137625829e-06, "loss": 0.0159, "step": 90930 }, { "grad_norm": 0.2615829110145569, "learning_rate": 2.227885846655903e-06, "loss": 0.0201, "step": 90940 }, { "grad_norm": 0.4802420735359192, "learning_rate": 2.2230077799725523e-06, "loss": 0.0195, "step": 90950 }, { "grad_norm": 0.3997189998626709, "learning_rate": 2.2181349381092396e-06, "loss": 0.0117, "step": 90960 }, { "grad_norm": 0.4968153238296509, "learning_rate": 2.2132673215988552e-06, "loss": 0.0202, "step": 90970 }, { "grad_norm": 0.550561249256134, "learning_rate": 2.2084049309737065e-06, "loss": 0.0169, "step": 90980 }, { "grad_norm": 0.2942996621131897, "learning_rate": 2.2035477667655513e-06, "loss": 0.0118, "step": 90990 }, { "grad_norm": 0.4502967298030853, "learning_rate": 2.1986958295055315e-06, "loss": 0.0223, "step": 91000 }, { "grad_norm": 1.6313443183898926, "learning_rate": 2.193849119724284e-06, "loss": 0.0168, "step": 91010 }, { "grad_norm": 0.4257736802101135, "learning_rate": 2.1890076379518175e-06, "loss": 0.0177, "step": 91020 }, { "grad_norm": 0.15422841906547546, "learning_rate": 2.1841713847175925e-06, "loss": 0.016, "step": 91030 }, { "grad_norm": 0.1294679343700409, "learning_rate": 2.1793403605504968e-06, "loss": 0.0127, "step": 91040 }, { "grad_norm": 0.2675154209136963, "learning_rate": 2.1745145659788413e-06, "loss": 0.0162, "step": 91050 }, { "grad_norm": 0.5536401271820068, "learning_rate": 2.1696940015303645e-06, "loss": 0.0131, "step": 91060 }, { "grad_norm": 0.17969389259815216, "learning_rate": 2.1648786677322397e-06, "loss": 0.0153, "step": 91070 }, { "grad_norm": 0.2262173593044281, "learning_rate": 2.1600685651110673e-06, "loss": 0.0161, "step": 91080 }, { "grad_norm": 0.5616562366485596, "learning_rate": 2.1552636941928717e-06, "loss": 0.0214, "step": 91090 }, { "grad_norm": 0.2293108105659485, "learning_rate": 2.1504640555031043e-06, "loss": 0.0227, "step": 91100 }, { "grad_norm": 0.41645917296409607, "learning_rate": 2.1456696495666506e-06, "loss": 0.0182, "step": 91110 }, { "grad_norm": 0.24010051786899567, "learning_rate": 2.1408804769078084e-06, "loss": 0.0181, "step": 91120 }, { "grad_norm": 0.08543181419372559, "learning_rate": 2.136096538050325e-06, "loss": 0.0141, "step": 91130 }, { "grad_norm": 0.21009555459022522, "learning_rate": 2.13131783351736e-06, "loss": 0.0149, "step": 91140 }, { "grad_norm": 0.4446159601211548, "learning_rate": 2.126544363831512e-06, "loss": 0.0229, "step": 91150 }, { "grad_norm": 0.3597552478313446, "learning_rate": 2.121776129514791e-06, "loss": 0.016, "step": 91160 }, { "grad_norm": 0.17807568609714508, "learning_rate": 2.1170131310886523e-06, "loss": 0.0125, "step": 91170 }, { "grad_norm": 0.4421550929546356, "learning_rate": 2.112255369073968e-06, "loss": 0.0138, "step": 91180 }, { "grad_norm": 0.24736887216567993, "learning_rate": 2.1075028439910328e-06, "loss": 0.0152, "step": 91190 }, { "grad_norm": 0.48639506101608276, "learning_rate": 2.1027555563595813e-06, "loss": 0.0168, "step": 91200 }, { "grad_norm": 0.4120703935623169, "learning_rate": 2.0980135066987706e-06, "loss": 0.0124, "step": 91210 }, { "grad_norm": 0.33227667212486267, "learning_rate": 2.093276695527191e-06, "loss": 0.0163, "step": 91220 }, { "grad_norm": 0.5416762232780457, "learning_rate": 2.0885451233628396e-06, "loss": 0.0223, "step": 91230 }, { "grad_norm": 0.3614982068538666, "learning_rate": 2.0838187907231633e-06, "loss": 0.0192, "step": 91240 }, { "grad_norm": 0.44354385137557983, "learning_rate": 2.0790976981250153e-06, "loss": 0.0095, "step": 91250 }, { "grad_norm": 0.5009803175926208, "learning_rate": 2.0743818460847e-06, "loss": 0.0168, "step": 91260 }, { "grad_norm": 0.18841303884983063, "learning_rate": 2.069671235117926e-06, "loss": 0.0176, "step": 91270 }, { "grad_norm": 0.1915600448846817, "learning_rate": 2.0649658657398484e-06, "loss": 0.0149, "step": 91280 }, { "grad_norm": 0.7537316679954529, "learning_rate": 2.060265738465034e-06, "loss": 0.0174, "step": 91290 }, { "grad_norm": 0.18634450435638428, "learning_rate": 2.055570853807487e-06, "loss": 0.0167, "step": 91300 }, { "grad_norm": 0.1667473167181015, "learning_rate": 2.0508812122806196e-06, "loss": 0.0142, "step": 91310 }, { "grad_norm": 0.35225915908813477, "learning_rate": 2.046196814397294e-06, "loss": 0.0162, "step": 91320 }, { "grad_norm": 0.2734921872615814, "learning_rate": 2.0415176606697896e-06, "loss": 0.018, "step": 91330 }, { "grad_norm": 0.3238128125667572, "learning_rate": 2.0368437516098126e-06, "loss": 0.0204, "step": 91340 }, { "grad_norm": 0.18402501940727234, "learning_rate": 2.032175087728483e-06, "loss": 0.0183, "step": 91350 }, { "grad_norm": 0.21726228296756744, "learning_rate": 2.027511669536375e-06, "loss": 0.016, "step": 91360 }, { "grad_norm": 0.43689286708831787, "learning_rate": 2.0228534975434643e-06, "loss": 0.0231, "step": 91370 }, { "grad_norm": 0.2676829695701599, "learning_rate": 2.0182005722591548e-06, "loss": 0.0178, "step": 91380 }, { "grad_norm": 0.2510375380516052, "learning_rate": 2.0135528941922944e-06, "loss": 0.0155, "step": 91390 }, { "grad_norm": 0.24843469262123108, "learning_rate": 2.008910463851149e-06, "loss": 0.0201, "step": 91400 }, { "grad_norm": 0.6224273443222046, "learning_rate": 2.0042732817433963e-06, "loss": 0.0216, "step": 91410 }, { "grad_norm": 0.24292628467082977, "learning_rate": 1.999641348376158e-06, "loss": 0.0209, "step": 91420 }, { "grad_norm": 0.24729035794734955, "learning_rate": 1.9950146642559787e-06, "loss": 0.0185, "step": 91430 }, { "grad_norm": 0.40631335973739624, "learning_rate": 1.990393229888815e-06, "loss": 0.0193, "step": 91440 }, { "grad_norm": 0.29540151357650757, "learning_rate": 1.9857770457800684e-06, "loss": 0.0177, "step": 91450 }, { "grad_norm": 0.15458935499191284, "learning_rate": 1.9811661124345517e-06, "loss": 0.0149, "step": 91460 }, { "grad_norm": 0.4686659574508667, "learning_rate": 1.976560430356522e-06, "loss": 0.0221, "step": 91470 }, { "grad_norm": 0.12230949848890305, "learning_rate": 1.9719600000496383e-06, "loss": 0.012, "step": 91480 }, { "grad_norm": 0.34891122579574585, "learning_rate": 1.967364822017004e-06, "loss": 0.0148, "step": 91490 }, { "grad_norm": 0.2036523073911667, "learning_rate": 1.962774896761127e-06, "loss": 0.016, "step": 91500 }, { "grad_norm": 0.2015821784734726, "learning_rate": 1.958190224783968e-06, "loss": 0.0186, "step": 91510 }, { "grad_norm": 0.19166256487369537, "learning_rate": 1.9536108065868975e-06, "loss": 0.0165, "step": 91520 }, { "grad_norm": 0.2142874002456665, "learning_rate": 1.949036642670715e-06, "loss": 0.0097, "step": 91530 }, { "grad_norm": 0.37735286355018616, "learning_rate": 1.9444677335356377e-06, "loss": 0.0145, "step": 91540 }, { "grad_norm": 0.26161980628967285, "learning_rate": 1.939904079681326e-06, "loss": 0.0155, "step": 91550 }, { "grad_norm": 0.093620166182518, "learning_rate": 1.935345681606837e-06, "loss": 0.0257, "step": 91560 }, { "grad_norm": 0.4421501159667969, "learning_rate": 1.930792539810694e-06, "loss": 0.0245, "step": 91570 }, { "grad_norm": 0.5213766098022461, "learning_rate": 1.9262446547907877e-06, "loss": 0.0188, "step": 91580 }, { "grad_norm": 0.3142337501049042, "learning_rate": 1.921702027044503e-06, "loss": 0.0186, "step": 91590 }, { "grad_norm": 0.1871718168258667, "learning_rate": 1.917164657068593e-06, "loss": 0.0194, "step": 91600 }, { "grad_norm": 0.3411063849925995, "learning_rate": 1.912632545359272e-06, "loss": 0.0186, "step": 91610 }, { "grad_norm": 0.6679843664169312, "learning_rate": 1.9081056924121488e-06, "loss": 0.0325, "step": 91620 }, { "grad_norm": 0.346617728471756, "learning_rate": 1.9035840987222887e-06, "loss": 0.0148, "step": 91630 }, { "grad_norm": 0.44516822695732117, "learning_rate": 1.8990677647841515e-06, "loss": 0.014, "step": 91640 }, { "grad_norm": 0.14822064340114594, "learning_rate": 1.8945566910916534e-06, "loss": 0.0294, "step": 91650 }, { "grad_norm": 0.1572169065475464, "learning_rate": 1.8900508781381054e-06, "loss": 0.0155, "step": 91660 }, { "grad_norm": 0.43587878346443176, "learning_rate": 1.8855503264162633e-06, "loss": 0.0189, "step": 91670 }, { "grad_norm": 0.20873017609119415, "learning_rate": 1.8810550364182998e-06, "loss": 0.0304, "step": 91680 }, { "grad_norm": 0.7365474104881287, "learning_rate": 1.8765650086358112e-06, "loss": 0.0212, "step": 91690 }, { "grad_norm": 0.45476123690605164, "learning_rate": 1.8720802435598151e-06, "loss": 0.0282, "step": 91700 }, { "grad_norm": 0.36935943365097046, "learning_rate": 1.8676007416807694e-06, "loss": 0.0193, "step": 91710 }, { "grad_norm": 0.2961321771144867, "learning_rate": 1.8631265034885437e-06, "loss": 0.0093, "step": 91720 }, { "grad_norm": 0.5616695880889893, "learning_rate": 1.858657529472424e-06, "loss": 0.0218, "step": 91730 }, { "grad_norm": 0.5336782932281494, "learning_rate": 1.8541938201211473e-06, "loss": 0.0163, "step": 91740 }, { "grad_norm": 0.21022379398345947, "learning_rate": 1.8497353759228398e-06, "loss": 0.0242, "step": 91750 }, { "grad_norm": 0.319830983877182, "learning_rate": 1.8452821973650835e-06, "loss": 0.0186, "step": 91760 }, { "grad_norm": 0.20518681406974792, "learning_rate": 1.840834284934856e-06, "loss": 0.0324, "step": 91770 }, { "grad_norm": 0.12137588113546371, "learning_rate": 1.8363916391185954e-06, "loss": 0.0278, "step": 91780 }, { "grad_norm": 0.24682874977588654, "learning_rate": 1.8319542604021245e-06, "loss": 0.0126, "step": 91790 }, { "grad_norm": 0.2168087512254715, "learning_rate": 1.8275221492707273e-06, "loss": 0.0158, "step": 91800 }, { "grad_norm": 0.1703246533870697, "learning_rate": 1.8230953062090717e-06, "loss": 0.0229, "step": 91810 }, { "grad_norm": 0.3250802457332611, "learning_rate": 1.8186737317012926e-06, "loss": 0.0153, "step": 91820 }, { "grad_norm": 0.5059216618537903, "learning_rate": 1.8142574262308977e-06, "loss": 0.0128, "step": 91830 }, { "grad_norm": 0.27637961506843567, "learning_rate": 1.8098463902808727e-06, "loss": 0.0146, "step": 91840 }, { "grad_norm": 0.18188291788101196, "learning_rate": 1.805440624333593e-06, "loss": 0.0123, "step": 91850 }, { "grad_norm": 0.31634795665740967, "learning_rate": 1.8010401288708734e-06, "loss": 0.0145, "step": 91860 }, { "grad_norm": 0.15470027923583984, "learning_rate": 1.7966449043739343e-06, "loss": 0.0165, "step": 91870 }, { "grad_norm": 0.3738331198692322, "learning_rate": 1.79225495132343e-06, "loss": 0.0122, "step": 91880 }, { "grad_norm": 0.497901976108551, "learning_rate": 1.7878702701994487e-06, "loss": 0.0184, "step": 91890 }, { "grad_norm": 0.34303879737854004, "learning_rate": 1.783490861481496e-06, "loss": 0.0208, "step": 91900 }, { "grad_norm": 0.24469414353370667, "learning_rate": 1.7791167256484832e-06, "loss": 0.0162, "step": 91910 }, { "grad_norm": 0.24351249635219574, "learning_rate": 1.774747863178766e-06, "loss": 0.0098, "step": 91920 }, { "grad_norm": 0.28695616126060486, "learning_rate": 1.7703842745501186e-06, "loss": 0.0132, "step": 91930 }, { "grad_norm": 0.2853090763092041, "learning_rate": 1.7660259602397366e-06, "loss": 0.0137, "step": 91940 }, { "grad_norm": 0.3180369734764099, "learning_rate": 1.7616729207242388e-06, "loss": 0.0111, "step": 91950 }, { "grad_norm": 0.3961564302444458, "learning_rate": 1.7573251564796556e-06, "loss": 0.0143, "step": 91960 }, { "grad_norm": 0.5045382976531982, "learning_rate": 1.7529826679814676e-06, "loss": 0.0344, "step": 91970 }, { "grad_norm": 0.2790881097316742, "learning_rate": 1.7486454557045562e-06, "loss": 0.0143, "step": 91980 }, { "grad_norm": 0.3443445563316345, "learning_rate": 1.7443135201232364e-06, "loss": 0.0118, "step": 91990 }, { "grad_norm": 0.21733932197093964, "learning_rate": 1.7399868617112347e-06, "loss": 0.0214, "step": 92000 }, { "grad_norm": 0.15410104393959045, "learning_rate": 1.7356654809417171e-06, "loss": 0.0114, "step": 92010 }, { "grad_norm": 0.249522402882576, "learning_rate": 1.7313493782872448e-06, "loss": 0.0127, "step": 92020 }, { "grad_norm": 0.2531716227531433, "learning_rate": 1.7270385542198453e-06, "loss": 0.0151, "step": 92030 }, { "grad_norm": 0.20713093876838684, "learning_rate": 1.7227330092109307e-06, "loss": 0.02, "step": 92040 }, { "grad_norm": 0.15949976444244385, "learning_rate": 1.718432743731352e-06, "loss": 0.012, "step": 92050 }, { "grad_norm": 0.5217360854148865, "learning_rate": 1.7141377582513718e-06, "loss": 0.0305, "step": 92060 }, { "grad_norm": 0.4716062843799591, "learning_rate": 1.709848053240698e-06, "loss": 0.0187, "step": 92070 }, { "grad_norm": 0.308423787355423, "learning_rate": 1.7055636291684328e-06, "loss": 0.0104, "step": 92080 }, { "grad_norm": 0.17244939506053925, "learning_rate": 1.7012844865031185e-06, "loss": 0.0146, "step": 92090 }, { "grad_norm": 0.3455533981323242, "learning_rate": 1.6970106257127138e-06, "loss": 0.0177, "step": 92100 }, { "grad_norm": 0.15508030354976654, "learning_rate": 1.6927420472646117e-06, "loss": 0.0143, "step": 92110 }, { "grad_norm": 0.1760643571615219, "learning_rate": 1.6884787516256106e-06, "loss": 0.0124, "step": 92120 }, { "grad_norm": 0.3014078438282013, "learning_rate": 1.6842207392619325e-06, "loss": 0.0132, "step": 92130 }, { "grad_norm": 0.13171471655368805, "learning_rate": 1.679968010639238e-06, "loss": 0.0369, "step": 92140 }, { "grad_norm": 0.14403842389583588, "learning_rate": 1.6757205662225829e-06, "loss": 0.027, "step": 92150 }, { "grad_norm": 0.1641928106546402, "learning_rate": 1.6714784064764844e-06, "loss": 0.0125, "step": 92160 }, { "grad_norm": 0.45680662989616394, "learning_rate": 1.6672415318648383e-06, "loss": 0.0171, "step": 92170 }, { "grad_norm": 0.5444208383560181, "learning_rate": 1.6630099428510016e-06, "loss": 0.0284, "step": 92180 }, { "grad_norm": 0.15909838676452637, "learning_rate": 1.6587836398977153e-06, "loss": 0.012, "step": 92190 }, { "grad_norm": 0.13717669248580933, "learning_rate": 1.6545626234671762e-06, "loss": 0.0184, "step": 92200 }, { "grad_norm": 0.6461944580078125, "learning_rate": 1.650346894020971e-06, "loss": 0.0207, "step": 92210 }, { "grad_norm": 0.46369650959968567, "learning_rate": 1.6461364520201472e-06, "loss": 0.0249, "step": 92220 }, { "grad_norm": 0.24665702879428864, "learning_rate": 1.6419312979251366e-06, "loss": 0.02, "step": 92230 }, { "grad_norm": 0.40572336316108704, "learning_rate": 1.6377314321958214e-06, "loss": 0.0132, "step": 92240 }, { "grad_norm": 0.21334026753902435, "learning_rate": 1.6335368552914787e-06, "loss": 0.0142, "step": 92250 }, { "grad_norm": 0.20232175290584564, "learning_rate": 1.6293475676708358e-06, "loss": 0.0133, "step": 92260 }, { "grad_norm": 0.23248697817325592, "learning_rate": 1.6251635697920097e-06, "loss": 0.0119, "step": 92270 }, { "grad_norm": 0.5800694227218628, "learning_rate": 1.6209848621125624e-06, "loss": 0.0149, "step": 92280 }, { "grad_norm": 0.4608134627342224, "learning_rate": 1.6168114450894778e-06, "loss": 0.0139, "step": 92290 }, { "grad_norm": 0.18814098834991455, "learning_rate": 1.612643319179158e-06, "loss": 0.015, "step": 92300 }, { "grad_norm": 0.4022114872932434, "learning_rate": 1.6084804848374046e-06, "loss": 0.0143, "step": 92310 }, { "grad_norm": 0.43714210391044617, "learning_rate": 1.6043229425194817e-06, "loss": 0.0231, "step": 92320 }, { "grad_norm": 0.3192604184150696, "learning_rate": 1.6001706926800308e-06, "loss": 0.0151, "step": 92330 }, { "grad_norm": 0.16570258140563965, "learning_rate": 1.5960237357731444e-06, "loss": 0.0223, "step": 92340 }, { "grad_norm": 0.13992032408714294, "learning_rate": 1.5918820722523264e-06, "loss": 0.0116, "step": 92350 }, { "grad_norm": 0.5606909394264221, "learning_rate": 1.587745702570509e-06, "loss": 0.0223, "step": 92360 }, { "grad_norm": 0.2752937376499176, "learning_rate": 1.5836146271800356e-06, "loss": 0.0129, "step": 92370 }, { "grad_norm": 0.3909408450126648, "learning_rate": 1.5794888465326728e-06, "loss": 0.0191, "step": 92380 }, { "grad_norm": 0.3590382933616638, "learning_rate": 1.5753683610796155e-06, "loss": 0.0158, "step": 92390 }, { "grad_norm": 0.6067187190055847, "learning_rate": 1.5712531712714528e-06, "loss": 0.0236, "step": 92400 }, { "grad_norm": 0.3295194208621979, "learning_rate": 1.5671432775582474e-06, "loss": 0.0162, "step": 92410 }, { "grad_norm": 0.3146228492259979, "learning_rate": 1.563038680389428e-06, "loss": 0.0151, "step": 92420 }, { "grad_norm": 0.25235897302627563, "learning_rate": 1.5589393802138808e-06, "loss": 0.0211, "step": 92430 }, { "grad_norm": 0.3214209973812103, "learning_rate": 1.554845377479891e-06, "loss": 0.0134, "step": 92440 }, { "grad_norm": 0.3716595470905304, "learning_rate": 1.550756672635173e-06, "loss": 0.0135, "step": 92450 }, { "grad_norm": 0.25352802872657776, "learning_rate": 1.5466732661268635e-06, "loss": 0.0141, "step": 92460 }, { "grad_norm": 0.3346051871776581, "learning_rate": 1.5425951584015163e-06, "loss": 0.0258, "step": 92470 }, { "grad_norm": 0.18610742688179016, "learning_rate": 1.5385223499051026e-06, "loss": 0.0158, "step": 92480 }, { "grad_norm": 0.22022795677185059, "learning_rate": 1.5344548410830328e-06, "loss": 0.0118, "step": 92490 }, { "grad_norm": 0.43924564123153687, "learning_rate": 1.5303926323801121e-06, "loss": 0.0183, "step": 92500 }, { "grad_norm": 0.1740570366382599, "learning_rate": 1.5263357242405795e-06, "loss": 0.02, "step": 92510 }, { "grad_norm": 0.2786971628665924, "learning_rate": 1.5222841171080915e-06, "loss": 0.0188, "step": 92520 }, { "grad_norm": 0.2999890446662903, "learning_rate": 1.5182378114257268e-06, "loss": 0.0122, "step": 92530 }, { "grad_norm": 0.21061845123767853, "learning_rate": 1.5141968076359869e-06, "loss": 0.0158, "step": 92540 }, { "grad_norm": 0.21513354778289795, "learning_rate": 1.5101611061807908e-06, "loss": 0.0199, "step": 92550 }, { "grad_norm": 0.18420085310935974, "learning_rate": 1.506130707501463e-06, "loss": 0.0124, "step": 92560 }, { "grad_norm": 0.2692342698574066, "learning_rate": 1.5021056120387734e-06, "loss": 0.0141, "step": 92570 }, { "grad_norm": 0.6335331201553345, "learning_rate": 1.4980858202329084e-06, "loss": 0.0196, "step": 92580 }, { "grad_norm": 0.6022046208381653, "learning_rate": 1.4940713325234445e-06, "loss": 0.017, "step": 92590 }, { "grad_norm": 0.2583552598953247, "learning_rate": 1.4900621493494137e-06, "loss": 0.018, "step": 92600 }, { "grad_norm": 0.5975597500801086, "learning_rate": 1.4860582711492544e-06, "loss": 0.0183, "step": 92610 }, { "grad_norm": 0.3369710445404053, "learning_rate": 1.4820596983608325e-06, "loss": 0.0185, "step": 92620 }, { "grad_norm": 0.4336947798728943, "learning_rate": 1.4780664314214044e-06, "loss": 0.0121, "step": 92630 }, { "grad_norm": 0.3061152398586273, "learning_rate": 1.474078470767687e-06, "loss": 0.0126, "step": 92640 }, { "grad_norm": 0.27442094683647156, "learning_rate": 1.470095816835787e-06, "loss": 0.012, "step": 92650 }, { "grad_norm": 0.5313802361488342, "learning_rate": 1.4661184700612395e-06, "loss": 0.0152, "step": 92660 }, { "grad_norm": 0.30142951011657715, "learning_rate": 1.4621464308790133e-06, "loss": 0.0205, "step": 92670 }, { "grad_norm": 0.9423466920852661, "learning_rate": 1.4581796997234775e-06, "loss": 0.0256, "step": 92680 }, { "grad_norm": 0.45158445835113525, "learning_rate": 1.4542182770284297e-06, "loss": 0.0132, "step": 92690 }, { "grad_norm": 0.18836764991283417, "learning_rate": 1.450262163227084e-06, "loss": 0.0143, "step": 92700 }, { "grad_norm": 0.2893284261226654, "learning_rate": 1.4463113587520671e-06, "loss": 0.0148, "step": 92710 }, { "grad_norm": 0.10853894799947739, "learning_rate": 1.4423658640354443e-06, "loss": 0.0182, "step": 92720 }, { "grad_norm": 0.42676419019699097, "learning_rate": 1.438425679508687e-06, "loss": 0.0153, "step": 92730 }, { "grad_norm": 0.20837600529193878, "learning_rate": 1.4344908056026895e-06, "loss": 0.0143, "step": 92740 }, { "grad_norm": 0.5437180995941162, "learning_rate": 1.4305612427477521e-06, "loss": 0.0137, "step": 92750 }, { "grad_norm": 0.29885512590408325, "learning_rate": 1.42663699137362e-06, "loss": 0.0208, "step": 92760 }, { "grad_norm": 0.31589803099632263, "learning_rate": 1.4227180519094386e-06, "loss": 0.0287, "step": 92770 }, { "grad_norm": 0.3602531850337982, "learning_rate": 1.4188044247837706e-06, "loss": 0.0252, "step": 92780 }, { "grad_norm": 0.3671284019947052, "learning_rate": 1.4148961104246128e-06, "loss": 0.0186, "step": 92790 }, { "grad_norm": 0.23289842903614044, "learning_rate": 1.4109931092593731e-06, "loss": 0.0101, "step": 92800 }, { "grad_norm": 0.12978805601596832, "learning_rate": 1.4070954217148658e-06, "loss": 0.0184, "step": 92810 }, { "grad_norm": 0.3659723401069641, "learning_rate": 1.4032030482173497e-06, "loss": 0.0139, "step": 92820 }, { "grad_norm": 0.2812803387641907, "learning_rate": 1.3993159891924901e-06, "loss": 0.0165, "step": 92830 }, { "grad_norm": 0.6571155190467834, "learning_rate": 1.3954342450653524e-06, "loss": 0.0165, "step": 92840 }, { "grad_norm": 0.4565369188785553, "learning_rate": 1.3915578162604582e-06, "loss": 0.0137, "step": 92850 }, { "grad_norm": 0.4555296301841736, "learning_rate": 1.3876867032017126e-06, "loss": 0.0127, "step": 92860 }, { "grad_norm": 0.18010997772216797, "learning_rate": 1.383820906312472e-06, "loss": 0.0181, "step": 92870 }, { "grad_norm": 0.4105060398578644, "learning_rate": 1.3799604260154752e-06, "loss": 0.0106, "step": 92880 }, { "grad_norm": 0.4281782805919647, "learning_rate": 1.376105262732913e-06, "loss": 0.0153, "step": 92890 }, { "grad_norm": 0.41062071919441223, "learning_rate": 1.37225541688637e-06, "loss": 0.0178, "step": 92900 }, { "grad_norm": 0.4944627583026886, "learning_rate": 1.3684108888968594e-06, "loss": 0.0138, "step": 92910 }, { "grad_norm": 0.45407113432884216, "learning_rate": 1.364571679184823e-06, "loss": 0.0145, "step": 92920 }, { "grad_norm": 0.29205840826034546, "learning_rate": 1.3607377881701077e-06, "loss": 0.0163, "step": 92930 }, { "grad_norm": 0.2199356108903885, "learning_rate": 1.3569092162719787e-06, "loss": 0.0155, "step": 92940 }, { "grad_norm": 0.46386751532554626, "learning_rate": 1.353085963909123e-06, "loss": 0.0228, "step": 92950 }, { "grad_norm": 0.3081638514995575, "learning_rate": 1.3492680314996454e-06, "loss": 0.0188, "step": 92960 }, { "grad_norm": 0.47358983755111694, "learning_rate": 1.345455419461078e-06, "loss": 0.0168, "step": 92970 }, { "grad_norm": 0.4170495271682739, "learning_rate": 1.3416481282103431e-06, "loss": 0.0112, "step": 92980 }, { "grad_norm": 0.278951495885849, "learning_rate": 1.337846158163819e-06, "loss": 0.0187, "step": 92990 }, { "grad_norm": 0.20357808470726013, "learning_rate": 1.334049509737273e-06, "loss": 0.0125, "step": 93000 }, { "grad_norm": 0.2707585096359253, "learning_rate": 1.3302581833459116e-06, "loss": 0.0203, "step": 93010 }, { "grad_norm": 0.27519404888153076, "learning_rate": 1.3264721794043366e-06, "loss": 0.0169, "step": 93020 }, { "grad_norm": 0.2292216271162033, "learning_rate": 1.3226914983265836e-06, "loss": 0.0127, "step": 93030 }, { "grad_norm": 0.3777380883693695, "learning_rate": 1.3189161405260997e-06, "loss": 0.0123, "step": 93040 }, { "grad_norm": 0.4506937861442566, "learning_rate": 1.3151461064157656e-06, "loss": 0.0325, "step": 93050 }, { "grad_norm": 0.4491133391857147, "learning_rate": 1.3113813964078459e-06, "loss": 0.0333, "step": 93060 }, { "grad_norm": 0.32961133122444153, "learning_rate": 1.3076220109140614e-06, "loss": 0.0131, "step": 93070 }, { "grad_norm": 0.2229115515947342, "learning_rate": 1.3038679503455275e-06, "loss": 0.0193, "step": 93080 }, { "grad_norm": 0.2019539624452591, "learning_rate": 1.3001192151127773e-06, "loss": 0.011, "step": 93090 }, { "grad_norm": 1.1032137870788574, "learning_rate": 1.2963758056257657e-06, "loss": 0.0164, "step": 93100 }, { "grad_norm": 0.17605462670326233, "learning_rate": 1.2926377222938768e-06, "loss": 0.0087, "step": 93110 }, { "grad_norm": 0.28966712951660156, "learning_rate": 1.2889049655259e-06, "loss": 0.0119, "step": 93120 }, { "grad_norm": 0.29220452904701233, "learning_rate": 1.2851775357300312e-06, "loss": 0.0142, "step": 93130 }, { "grad_norm": 0.23513825237751007, "learning_rate": 1.281455433313916e-06, "loss": 0.016, "step": 93140 }, { "grad_norm": 0.2647441029548645, "learning_rate": 1.2777386586845796e-06, "loss": 0.0173, "step": 93150 }, { "grad_norm": 0.224985733628273, "learning_rate": 1.2740272122484964e-06, "loss": 0.0158, "step": 93160 }, { "grad_norm": 0.25568073987960815, "learning_rate": 1.2703210944115308e-06, "loss": 0.0115, "step": 93170 }, { "grad_norm": 0.3868709206581116, "learning_rate": 1.2666203055789916e-06, "loss": 0.0165, "step": 93180 }, { "grad_norm": 0.17526182532310486, "learning_rate": 1.2629248461555832e-06, "loss": 0.0156, "step": 93190 }, { "grad_norm": 0.17605751752853394, "learning_rate": 1.2592347165454431e-06, "loss": 0.0152, "step": 93200 }, { "grad_norm": 0.28728818893432617, "learning_rate": 1.2555499171521102e-06, "loss": 0.0151, "step": 93210 }, { "grad_norm": 0.17537911236286163, "learning_rate": 1.251870448378556e-06, "loss": 0.0127, "step": 93220 }, { "grad_norm": 0.2976645827293396, "learning_rate": 1.2481963106271478e-06, "loss": 0.0177, "step": 93230 }, { "grad_norm": 0.36981984972953796, "learning_rate": 1.2445275042997084e-06, "loss": 0.0153, "step": 93240 }, { "grad_norm": 0.37756994366645813, "learning_rate": 1.2408640297974283e-06, "loss": 0.0144, "step": 93250 }, { "grad_norm": 0.32272377610206604, "learning_rate": 1.2372058875209535e-06, "loss": 0.019, "step": 93260 }, { "grad_norm": 0.28250643610954285, "learning_rate": 1.2335530778703252e-06, "loss": 0.0149, "step": 93270 }, { "grad_norm": 0.2374131828546524, "learning_rate": 1.2299056012450182e-06, "loss": 0.0124, "step": 93280 }, { "grad_norm": 0.2547603249549866, "learning_rate": 1.2262634580439026e-06, "loss": 0.0173, "step": 93290 }, { "grad_norm": 0.18054884672164917, "learning_rate": 1.2226266486652926e-06, "loss": 0.0204, "step": 93300 }, { "grad_norm": 0.8973629474639893, "learning_rate": 1.2189951735068927e-06, "loss": 0.0143, "step": 93310 }, { "grad_norm": 0.3236842155456543, "learning_rate": 1.2153690329658407e-06, "loss": 0.0238, "step": 93320 }, { "grad_norm": 0.2715357840061188, "learning_rate": 1.2117482274386859e-06, "loss": 0.0101, "step": 93330 }, { "grad_norm": 0.16218021512031555, "learning_rate": 1.2081327573213953e-06, "loss": 0.0195, "step": 93340 }, { "grad_norm": 0.14830955862998962, "learning_rate": 1.2045226230093465e-06, "loss": 0.014, "step": 93350 }, { "grad_norm": 0.1915161907672882, "learning_rate": 1.2009178248973296e-06, "loss": 0.0207, "step": 93360 }, { "grad_norm": 0.27609169483184814, "learning_rate": 1.1973183633795849e-06, "loss": 0.014, "step": 93370 }, { "grad_norm": 0.656401515007019, "learning_rate": 1.193724238849725e-06, "loss": 0.0172, "step": 93380 }, { "grad_norm": 0.3057362139225006, "learning_rate": 1.190135451700808e-06, "loss": 0.0139, "step": 93390 }, { "grad_norm": 0.17613670229911804, "learning_rate": 1.186552002325292e-06, "loss": 0.0156, "step": 93400 }, { "grad_norm": 0.36340978741645813, "learning_rate": 1.1829738911150578e-06, "loss": 0.0184, "step": 93410 }, { "grad_norm": 0.8157735466957092, "learning_rate": 1.1794011184613985e-06, "loss": 0.0204, "step": 93420 }, { "grad_norm": 0.4289504587650299, "learning_rate": 1.17583368475504e-06, "loss": 0.0194, "step": 93430 }, { "grad_norm": 0.7040900588035583, "learning_rate": 1.1722715903860982e-06, "loss": 0.0124, "step": 93440 }, { "grad_norm": 0.24875321984291077, "learning_rate": 1.168714835744128e-06, "loss": 0.0169, "step": 93450 }, { "grad_norm": 0.32603204250335693, "learning_rate": 1.165163421218085e-06, "loss": 0.0271, "step": 93460 }, { "grad_norm": 0.18666213750839233, "learning_rate": 1.1616173471963477e-06, "loss": 0.0129, "step": 93470 }, { "grad_norm": 0.46979066729545593, "learning_rate": 1.1580766140667109e-06, "loss": 0.0161, "step": 93480 }, { "grad_norm": 1.574563980102539, "learning_rate": 1.1545412222163764e-06, "loss": 0.0216, "step": 93490 }, { "grad_norm": 0.1834293156862259, "learning_rate": 1.1510111720319795e-06, "loss": 0.0193, "step": 93500 }, { "grad_norm": 0.6565107107162476, "learning_rate": 1.1474864638995663e-06, "loss": 0.0141, "step": 93510 }, { "grad_norm": 0.31691890954971313, "learning_rate": 1.1439670982045735e-06, "loss": 0.0283, "step": 93520 }, { "grad_norm": 0.302063912153244, "learning_rate": 1.1404530753318877e-06, "loss": 0.0188, "step": 93530 }, { "grad_norm": 0.2875958979129791, "learning_rate": 1.136944395665801e-06, "loss": 0.0152, "step": 93540 }, { "grad_norm": 0.5635239481925964, "learning_rate": 1.1334410595900014e-06, "loss": 0.0192, "step": 93550 }, { "grad_norm": 0.43689805269241333, "learning_rate": 1.129943067487621e-06, "loss": 0.0181, "step": 93560 }, { "grad_norm": 0.42409223318099976, "learning_rate": 1.126450419741193e-06, "loss": 0.0083, "step": 93570 }, { "grad_norm": 0.26355960965156555, "learning_rate": 1.1229631167326671e-06, "loss": 0.0183, "step": 93580 }, { "grad_norm": 0.9021810293197632, "learning_rate": 1.1194811588434052e-06, "loss": 0.0158, "step": 93590 }, { "grad_norm": 0.28793463110923767, "learning_rate": 1.1160045464542023e-06, "loss": 0.0226, "step": 93600 }, { "grad_norm": 0.25576838850975037, "learning_rate": 1.1125332799452325e-06, "loss": 0.0326, "step": 93610 }, { "grad_norm": 0.42466863989830017, "learning_rate": 1.109067359696131e-06, "loss": 0.0234, "step": 93620 }, { "grad_norm": 0.46838808059692383, "learning_rate": 1.1056067860859166e-06, "loss": 0.0262, "step": 93630 }, { "grad_norm": 0.2540540397167206, "learning_rate": 1.102151559493031e-06, "loss": 0.0172, "step": 93640 }, { "grad_norm": 0.5676891207695007, "learning_rate": 1.0987016802953332e-06, "loss": 0.0186, "step": 93650 }, { "grad_norm": 0.2311188280582428, "learning_rate": 1.0952571488701048e-06, "loss": 0.0171, "step": 93660 }, { "grad_norm": 0.30591994524002075, "learning_rate": 1.0918179655940163e-06, "loss": 0.0185, "step": 93670 }, { "grad_norm": 0.27511656284332275, "learning_rate": 1.088384130843184e-06, "loss": 0.0141, "step": 93680 }, { "grad_norm": 0.6805049777030945, "learning_rate": 1.0849556449931296e-06, "loss": 0.0146, "step": 93690 }, { "grad_norm": 0.32199543714523315, "learning_rate": 1.0815325084187866e-06, "loss": 0.0105, "step": 93700 }, { "grad_norm": 0.2577578127384186, "learning_rate": 1.0781147214944943e-06, "loss": 0.0115, "step": 93710 }, { "grad_norm": 0.4861302971839905, "learning_rate": 1.0747022845940313e-06, "loss": 0.022, "step": 93720 }, { "grad_norm": 0.2048708200454712, "learning_rate": 1.0712951980905605e-06, "loss": 0.0139, "step": 93730 }, { "grad_norm": 0.23060333728790283, "learning_rate": 1.0678934623566839e-06, "loss": 0.0206, "step": 93740 }, { "grad_norm": 0.14254580438137054, "learning_rate": 1.0644970777644091e-06, "loss": 0.0147, "step": 93750 }, { "grad_norm": 0.15231722593307495, "learning_rate": 1.0611060446851673e-06, "loss": 0.0107, "step": 93760 }, { "grad_norm": 0.19646500051021576, "learning_rate": 1.0577203634897836e-06, "loss": 0.0177, "step": 93770 }, { "grad_norm": 0.33213743567466736, "learning_rate": 1.0543400345485178e-06, "loss": 0.0124, "step": 93780 }, { "grad_norm": 0.3543619215488434, "learning_rate": 1.0509650582310404e-06, "loss": 0.011, "step": 93790 }, { "grad_norm": 0.22447240352630615, "learning_rate": 1.0475954349064287e-06, "loss": 0.0166, "step": 93800 }, { "grad_norm": 0.5102598667144775, "learning_rate": 1.0442311649431825e-06, "loss": 0.0121, "step": 93810 }, { "grad_norm": 0.6689732074737549, "learning_rate": 1.0408722487092071e-06, "loss": 0.0123, "step": 93820 }, { "grad_norm": 0.25608283281326294, "learning_rate": 1.0375186865718424e-06, "loss": 0.0241, "step": 93830 }, { "grad_norm": 1.1256662607192993, "learning_rate": 1.0341704788978113e-06, "loss": 0.0182, "step": 93840 }, { "grad_norm": 0.3320954144001007, "learning_rate": 1.0308276260532879e-06, "loss": 0.0182, "step": 93850 }, { "grad_norm": 0.45244792103767395, "learning_rate": 1.0274901284038241e-06, "loss": 0.0132, "step": 93860 }, { "grad_norm": 0.20930561423301697, "learning_rate": 1.0241579863144114e-06, "loss": 0.0188, "step": 93870 }, { "grad_norm": 0.19667580723762512, "learning_rate": 1.0208312001494413e-06, "loss": 0.0225, "step": 93880 }, { "grad_norm": 0.2110365629196167, "learning_rate": 1.0175097702727453e-06, "loss": 0.0174, "step": 93890 }, { "grad_norm": 0.22353330254554749, "learning_rate": 1.0141936970475275e-06, "loss": 0.0118, "step": 93900 }, { "grad_norm": 0.3291237950325012, "learning_rate": 1.0108829808364417e-06, "loss": 0.014, "step": 93910 }, { "grad_norm": 0.36732587218284607, "learning_rate": 1.0075776220015376e-06, "loss": 0.013, "step": 93920 }, { "grad_norm": 0.3058498799800873, "learning_rate": 1.0042776209042816e-06, "loss": 0.019, "step": 93930 }, { "grad_norm": 0.1279076784849167, "learning_rate": 1.000982977905568e-06, "loss": 0.0113, "step": 93940 }, { "grad_norm": 0.2089197039604187, "learning_rate": 9.976936933656867e-07, "loss": 0.021, "step": 93950 }, { "grad_norm": 0.16059373319149017, "learning_rate": 9.944097676443442e-07, "loss": 0.0115, "step": 93960 }, { "grad_norm": 0.23030219972133636, "learning_rate": 9.911312011006757e-07, "loss": 0.0112, "step": 93970 }, { "grad_norm": 0.30097317695617676, "learning_rate": 9.878579940932109e-07, "loss": 0.0218, "step": 93980 }, { "grad_norm": 0.14561809599399567, "learning_rate": 9.84590146979908e-07, "loss": 0.0204, "step": 93990 }, { "grad_norm": 0.32802408933639526, "learning_rate": 9.813276601181365e-07, "loss": 0.0164, "step": 94000 }, { "grad_norm": 0.40043142437934875, "learning_rate": 9.780705338646724e-07, "loss": 0.0182, "step": 94010 }, { "grad_norm": 0.6335753798484802, "learning_rate": 9.748187685757083e-07, "loss": 0.0172, "step": 94020 }, { "grad_norm": 0.15758116543293, "learning_rate": 9.715723646068543e-07, "loss": 0.0175, "step": 94030 }, { "grad_norm": 0.29287105798721313, "learning_rate": 9.68331322313143e-07, "loss": 0.0211, "step": 94040 }, { "grad_norm": 0.3077143728733063, "learning_rate": 9.650956420489909e-07, "loss": 0.0213, "step": 94050 }, { "grad_norm": 0.3050597906112671, "learning_rate": 9.618653241682595e-07, "loss": 0.0112, "step": 94060 }, { "grad_norm": 0.184483602643013, "learning_rate": 9.586403690242053e-07, "loss": 0.015, "step": 94070 }, { "grad_norm": 0.12397976219654083, "learning_rate": 9.554207769695122e-07, "loss": 0.0236, "step": 94080 }, { "grad_norm": 0.3000670075416565, "learning_rate": 9.522065483562659e-07, "loss": 0.02, "step": 94090 }, { "grad_norm": 0.2051064819097519, "learning_rate": 9.489976835359681e-07, "loss": 0.0207, "step": 94100 }, { "grad_norm": 0.19391153752803802, "learning_rate": 9.457941828595329e-07, "loss": 0.0122, "step": 94110 }, { "grad_norm": 0.455122172832489, "learning_rate": 9.425960466772965e-07, "loss": 0.0109, "step": 94120 }, { "grad_norm": 0.7409496903419495, "learning_rate": 9.394032753390014e-07, "loss": 0.0174, "step": 94130 }, { "grad_norm": 0.21140162646770477, "learning_rate": 9.362158691938017e-07, "loss": 0.0175, "step": 94140 }, { "grad_norm": 0.3671686053276062, "learning_rate": 9.330338285902685e-07, "loss": 0.0212, "step": 94150 }, { "grad_norm": 0.30264729261398315, "learning_rate": 9.298571538763845e-07, "loss": 0.013, "step": 94160 }, { "grad_norm": 0.2516545057296753, "learning_rate": 9.266858453995441e-07, "loss": 0.0123, "step": 94170 }, { "grad_norm": 0.2931627929210663, "learning_rate": 9.235199035065645e-07, "loss": 0.0134, "step": 94180 }, { "grad_norm": 0.4387148320674896, "learning_rate": 9.203593285436574e-07, "loss": 0.0154, "step": 94190 }, { "grad_norm": 0.382893443107605, "learning_rate": 9.172041208564686e-07, "loss": 0.0158, "step": 94200 }, { "grad_norm": 1.3914735317230225, "learning_rate": 9.140542807900332e-07, "loss": 0.0189, "step": 94210 }, { "grad_norm": 0.5439120531082153, "learning_rate": 9.109098086888313e-07, "loss": 0.0191, "step": 94220 }, { "grad_norm": 0.21477189660072327, "learning_rate": 9.077707048967265e-07, "loss": 0.0096, "step": 94230 }, { "grad_norm": 0.4005109965801239, "learning_rate": 9.046369697570057e-07, "loss": 0.0182, "step": 94240 }, { "grad_norm": 0.19234389066696167, "learning_rate": 9.015086036123721e-07, "loss": 0.0147, "step": 94250 }, { "grad_norm": 0.41554397344589233, "learning_rate": 8.983856068049413e-07, "loss": 0.0165, "step": 94260 }, { "grad_norm": 0.33849549293518066, "learning_rate": 8.952679796762398e-07, "loss": 0.0169, "step": 94270 }, { "grad_norm": 0.21758049726486206, "learning_rate": 8.921557225671951e-07, "loss": 0.0193, "step": 94280 }, { "grad_norm": 0.32630205154418945, "learning_rate": 8.890488358181792e-07, "loss": 0.0132, "step": 94290 }, { "grad_norm": 0.6897929310798645, "learning_rate": 8.859473197689372e-07, "loss": 0.0251, "step": 94300 }, { "grad_norm": 0.26766344904899597, "learning_rate": 8.828511747586588e-07, "loss": 0.0176, "step": 94310 }, { "grad_norm": 0.5168921947479248, "learning_rate": 8.797604011259287e-07, "loss": 0.0149, "step": 94320 }, { "grad_norm": 0.20449958741664886, "learning_rate": 8.766749992087487e-07, "loss": 0.018, "step": 94330 }, { "grad_norm": 0.3151025176048279, "learning_rate": 8.735949693445378e-07, "loss": 0.0202, "step": 94340 }, { "grad_norm": 0.4726533889770508, "learning_rate": 8.705203118701211e-07, "loss": 0.0152, "step": 94350 }, { "grad_norm": 0.29299017786979675, "learning_rate": 8.674510271217295e-07, "loss": 0.0168, "step": 94360 }, { "grad_norm": 0.17770561575889587, "learning_rate": 8.643871154350336e-07, "loss": 0.0132, "step": 94370 }, { "grad_norm": 0.3181646764278412, "learning_rate": 8.613285771450818e-07, "loss": 0.0182, "step": 94380 }, { "grad_norm": 0.32079267501831055, "learning_rate": 8.582754125863623e-07, "loss": 0.0175, "step": 94390 }, { "grad_norm": 0.2732751667499542, "learning_rate": 8.55227622092758e-07, "loss": 0.0129, "step": 94400 }, { "grad_norm": 0.21957524120807648, "learning_rate": 8.521852059975743e-07, "loss": 0.0124, "step": 94410 }, { "grad_norm": 0.3645058572292328, "learning_rate": 8.491481646335231e-07, "loss": 0.0134, "step": 94420 }, { "grad_norm": 0.24139370024204254, "learning_rate": 8.46116498332733e-07, "loss": 0.0212, "step": 94430 }, { "grad_norm": 0.5399819612503052, "learning_rate": 8.430902074267333e-07, "loss": 0.0206, "step": 94440 }, { "grad_norm": 0.34226134419441223, "learning_rate": 8.400692922464925e-07, "loss": 0.0145, "step": 94450 }, { "grad_norm": 1.1652839183807373, "learning_rate": 8.370537531223577e-07, "loss": 0.0206, "step": 94460 }, { "grad_norm": 0.32156750559806824, "learning_rate": 8.340435903841092e-07, "loss": 0.0194, "step": 94470 }, { "grad_norm": 0.2814391851425171, "learning_rate": 8.310388043609341e-07, "loss": 0.0165, "step": 94480 }, { "grad_norm": 0.34332728385925293, "learning_rate": 8.280393953814303e-07, "loss": 0.0185, "step": 94490 }, { "grad_norm": 0.32978737354278564, "learning_rate": 8.250453637736077e-07, "loss": 0.0166, "step": 94500 }, { "grad_norm": 0.43061742186546326, "learning_rate": 8.22056709864899e-07, "loss": 0.0249, "step": 94510 }, { "grad_norm": 0.6483331918716431, "learning_rate": 8.190734339821204e-07, "loss": 0.0265, "step": 94520 }, { "grad_norm": 0.31708773970603943, "learning_rate": 8.160955364515333e-07, "loss": 0.0132, "step": 94530 }, { "grad_norm": 0.18521419167518616, "learning_rate": 8.131230175987936e-07, "loss": 0.0158, "step": 94540 }, { "grad_norm": 0.29226166009902954, "learning_rate": 8.101558777489693e-07, "loss": 0.0178, "step": 94550 }, { "grad_norm": 0.1706431806087494, "learning_rate": 8.071941172265396e-07, "loss": 0.0125, "step": 94560 }, { "grad_norm": 0.46003443002700806, "learning_rate": 8.042377363553955e-07, "loss": 0.0125, "step": 94570 }, { "grad_norm": 0.1878625452518463, "learning_rate": 8.01286735458856e-07, "loss": 0.0181, "step": 94580 }, { "grad_norm": 0.18925470113754272, "learning_rate": 7.983411148596298e-07, "loss": 0.0147, "step": 94590 }, { "grad_norm": 0.287891149520874, "learning_rate": 7.954008748798425e-07, "loss": 0.0185, "step": 94600 }, { "grad_norm": 0.4830593168735504, "learning_rate": 7.924660158410369e-07, "loss": 0.0186, "step": 94610 }, { "grad_norm": 0.5191847085952759, "learning_rate": 7.895365380641728e-07, "loss": 0.0183, "step": 94620 }, { "grad_norm": 0.13147419691085815, "learning_rate": 7.866124418695941e-07, "loss": 0.0118, "step": 94630 }, { "grad_norm": 0.10795942693948746, "learning_rate": 7.83693727577095e-07, "loss": 0.0127, "step": 94640 }, { "grad_norm": 0.6333268880844116, "learning_rate": 7.807803955058535e-07, "loss": 0.0165, "step": 94650 }, { "grad_norm": 0.293373167514801, "learning_rate": 7.778724459744702e-07, "loss": 0.0142, "step": 94660 }, { "grad_norm": 0.3859555125236511, "learning_rate": 7.749698793009464e-07, "loss": 0.0126, "step": 94670 }, { "grad_norm": 0.26699331402778625, "learning_rate": 7.720726958027169e-07, "loss": 0.0144, "step": 94680 }, { "grad_norm": 0.0751027911901474, "learning_rate": 7.691808957965951e-07, "loss": 0.0089, "step": 94690 }, { "grad_norm": 0.29533055424690247, "learning_rate": 7.662944795988336e-07, "loss": 0.0131, "step": 94700 }, { "grad_norm": 0.2347302883863449, "learning_rate": 7.634134475250854e-07, "loss": 0.0193, "step": 94710 }, { "grad_norm": 0.5131544470787048, "learning_rate": 7.605377998904262e-07, "loss": 0.0152, "step": 94720 }, { "grad_norm": 0.46208325028419495, "learning_rate": 7.5766753700931e-07, "loss": 0.0102, "step": 94730 }, { "grad_norm": 0.5808555483818054, "learning_rate": 7.54802659195647e-07, "loss": 0.0317, "step": 94740 }, { "grad_norm": 0.15992426872253418, "learning_rate": 7.519431667627253e-07, "loss": 0.0166, "step": 94750 }, { "grad_norm": 0.34281960129737854, "learning_rate": 7.490890600232503e-07, "loss": 0.0124, "step": 94760 }, { "grad_norm": 0.19695886969566345, "learning_rate": 7.462403392893502e-07, "loss": 0.0166, "step": 94770 }, { "grad_norm": 0.6233435273170471, "learning_rate": 7.433970048725536e-07, "loss": 0.0177, "step": 94780 }, { "grad_norm": 0.5672996640205383, "learning_rate": 7.405590570838062e-07, "loss": 0.0244, "step": 94790 }, { "grad_norm": 0.37397676706314087, "learning_rate": 7.377264962334596e-07, "loss": 0.0197, "step": 94800 }, { "grad_norm": 0.3196555972099304, "learning_rate": 7.348993226312773e-07, "loss": 0.0111, "step": 94810 }, { "grad_norm": 0.29027894139289856, "learning_rate": 7.32077536586434e-07, "loss": 0.017, "step": 94820 }, { "grad_norm": 0.37653183937072754, "learning_rate": 7.292611384075221e-07, "loss": 0.0149, "step": 94830 }, { "grad_norm": 0.2630762457847595, "learning_rate": 7.264501284025338e-07, "loss": 0.0219, "step": 94840 }, { "grad_norm": 0.44348767399787903, "learning_rate": 7.236445068788789e-07, "loss": 0.0171, "step": 94850 }, { "grad_norm": 0.2225724309682846, "learning_rate": 7.20844274143373e-07, "loss": 0.0116, "step": 94860 }, { "grad_norm": 0.3400211036205292, "learning_rate": 7.180494305022489e-07, "loss": 0.0423, "step": 94870 }, { "grad_norm": 0.5070186853408813, "learning_rate": 7.152599762611456e-07, "loss": 0.0104, "step": 94880 }, { "grad_norm": 0.3736516535282135, "learning_rate": 7.124759117251079e-07, "loss": 0.0218, "step": 94890 }, { "grad_norm": 0.1515795737504959, "learning_rate": 7.096972371986033e-07, "loss": 0.0153, "step": 94900 }, { "grad_norm": 0.7653791308403015, "learning_rate": 7.069239529855054e-07, "loss": 0.0325, "step": 94910 }, { "grad_norm": 1.159102201461792, "learning_rate": 7.041560593890939e-07, "loss": 0.0186, "step": 94920 }, { "grad_norm": 0.4011605381965637, "learning_rate": 7.013935567120599e-07, "loss": 0.0198, "step": 94930 }, { "grad_norm": 0.15051090717315674, "learning_rate": 6.986364452565009e-07, "loss": 0.0184, "step": 94940 }, { "grad_norm": 0.5378780364990234, "learning_rate": 6.95884725323942e-07, "loss": 0.0188, "step": 94950 }, { "grad_norm": 0.6258465051651001, "learning_rate": 6.931383972152982e-07, "loss": 0.0163, "step": 94960 }, { "grad_norm": 0.3141649663448334, "learning_rate": 6.903974612309127e-07, "loss": 0.0114, "step": 94970 }, { "grad_norm": 0.21245039999485016, "learning_rate": 6.876619176705179e-07, "loss": 0.0113, "step": 94980 }, { "grad_norm": 0.42756417393684387, "learning_rate": 6.8493176683328e-07, "loss": 0.0131, "step": 94990 }, { "grad_norm": 0.2165088951587677, "learning_rate": 6.822070090177657e-07, "loss": 0.0233, "step": 95000 }, { "grad_norm": 0.22357361018657684, "learning_rate": 6.794876445219367e-07, "loss": 0.0146, "step": 95010 }, { "grad_norm": 0.2795836329460144, "learning_rate": 6.767736736431885e-07, "loss": 0.0124, "step": 95020 }, { "grad_norm": 0.10469665378332138, "learning_rate": 6.740650966783113e-07, "loss": 0.0145, "step": 95030 }, { "grad_norm": 0.16668327152729034, "learning_rate": 6.713619139235239e-07, "loss": 0.0182, "step": 95040 }, { "grad_norm": 0.4549807608127594, "learning_rate": 6.686641256744286e-07, "loss": 0.0167, "step": 95050 }, { "grad_norm": 0.678885281085968, "learning_rate": 6.659717322260562e-07, "loss": 0.018, "step": 95060 }, { "grad_norm": 0.3054860532283783, "learning_rate": 6.632847338728377e-07, "loss": 0.0164, "step": 95070 }, { "grad_norm": 0.21050377190113068, "learning_rate": 6.60603130908627e-07, "loss": 0.0135, "step": 95080 }, { "grad_norm": 0.6635273694992065, "learning_rate": 6.579269236266783e-07, "loss": 0.0164, "step": 95090 }, { "grad_norm": 0.20751433074474335, "learning_rate": 6.552561123196632e-07, "loss": 0.0096, "step": 95100 }, { "grad_norm": 0.2914988696575165, "learning_rate": 6.525906972796425e-07, "loss": 0.0251, "step": 95110 }, { "grad_norm": 0.5143693089485168, "learning_rate": 6.499306787981108e-07, "loss": 0.0201, "step": 95120 }, { "grad_norm": 0.5375809669494629, "learning_rate": 6.472760571659631e-07, "loss": 0.0193, "step": 95130 }, { "grad_norm": 0.2829038202762604, "learning_rate": 6.446268326735061e-07, "loss": 0.0158, "step": 95140 }, { "grad_norm": 0.35649484395980835, "learning_rate": 6.41983005610458e-07, "loss": 0.0217, "step": 95150 }, { "grad_norm": 0.5339862108230591, "learning_rate": 6.393445762659378e-07, "loss": 0.0176, "step": 95160 }, { "grad_norm": 0.13977384567260742, "learning_rate": 6.36711544928481e-07, "loss": 0.0143, "step": 95170 }, { "grad_norm": 0.5419162511825562, "learning_rate": 6.340839118860298e-07, "loss": 0.0145, "step": 95180 }, { "grad_norm": 0.4015948474407196, "learning_rate": 6.314616774259485e-07, "loss": 0.0134, "step": 95190 }, { "grad_norm": 0.853746771812439, "learning_rate": 6.288448418349858e-07, "loss": 0.0375, "step": 95200 }, { "grad_norm": 0.40980958938598633, "learning_rate": 6.262334053993235e-07, "loss": 0.0175, "step": 95210 }, { "grad_norm": 0.2735436260700226, "learning_rate": 6.236273684045501e-07, "loss": 0.0147, "step": 95220 }, { "grad_norm": 0.3890054225921631, "learning_rate": 6.210267311356488e-07, "loss": 0.0159, "step": 95230 }, { "grad_norm": 0.3480057120323181, "learning_rate": 6.184314938770197e-07, "loss": 0.0172, "step": 95240 }, { "grad_norm": 0.39994654059410095, "learning_rate": 6.158416569124858e-07, "loss": 0.0343, "step": 95250 }, { "grad_norm": 0.23124654591083527, "learning_rate": 6.132572205252541e-07, "loss": 0.0167, "step": 95260 }, { "grad_norm": 1.219089150428772, "learning_rate": 6.106781849979648e-07, "loss": 0.0217, "step": 95270 }, { "grad_norm": 0.53741455078125, "learning_rate": 6.081045506126482e-07, "loss": 0.0099, "step": 95280 }, { "grad_norm": 0.2733381390571594, "learning_rate": 6.055363176507678e-07, "loss": 0.0145, "step": 95290 }, { "grad_norm": 0.3514198660850525, "learning_rate": 6.029734863931602e-07, "loss": 0.0169, "step": 95300 }, { "grad_norm": 0.28653013706207275, "learning_rate": 6.004160571201179e-07, "loss": 0.0136, "step": 95310 }, { "grad_norm": 0.3318502902984619, "learning_rate": 5.978640301112947e-07, "loss": 0.0149, "step": 95320 }, { "grad_norm": 0.16622360050678253, "learning_rate": 5.9531740564579e-07, "loss": 0.0212, "step": 95330 }, { "grad_norm": 0.3770703673362732, "learning_rate": 5.927761840020918e-07, "loss": 0.0108, "step": 95340 }, { "grad_norm": 0.3900185823440552, "learning_rate": 5.902403654581113e-07, "loss": 0.0149, "step": 95350 }, { "grad_norm": 0.3731933832168579, "learning_rate": 5.877099502911542e-07, "loss": 0.0183, "step": 95360 }, { "grad_norm": 0.5930110216140747, "learning_rate": 5.851849387779496e-07, "loss": 0.0155, "step": 95370 }, { "grad_norm": 0.17918910086154938, "learning_rate": 5.826653311946262e-07, "loss": 0.0112, "step": 95380 }, { "grad_norm": 0.3319344222545624, "learning_rate": 5.801511278167249e-07, "loss": 0.0144, "step": 95390 }, { "grad_norm": 0.11133327335119247, "learning_rate": 5.776423289191868e-07, "loss": 0.015, "step": 95400 }, { "grad_norm": 0.5367357134819031, "learning_rate": 5.751389347763869e-07, "loss": 0.0137, "step": 95410 }, { "grad_norm": 0.9266070127487183, "learning_rate": 5.726409456620841e-07, "loss": 0.0118, "step": 95420 }, { "grad_norm": 0.17351749539375305, "learning_rate": 5.701483618494541e-07, "loss": 0.0146, "step": 95430 }, { "grad_norm": 0.3956795930862427, "learning_rate": 5.67661183611079e-07, "loss": 0.0129, "step": 95440 }, { "grad_norm": 0.20637279748916626, "learning_rate": 5.651794112189579e-07, "loss": 0.0166, "step": 95450 }, { "grad_norm": 0.3746580481529236, "learning_rate": 5.627030449444903e-07, "loss": 0.0176, "step": 95460 }, { "grad_norm": 0.6528658866882324, "learning_rate": 5.60232085058493e-07, "loss": 0.0315, "step": 95470 }, { "grad_norm": 0.28855833411216736, "learning_rate": 5.577665318311776e-07, "loss": 0.0146, "step": 95480 }, { "grad_norm": 0.5637621879577637, "learning_rate": 5.553063855321838e-07, "loss": 0.0157, "step": 95490 }, { "grad_norm": 0.2046506255865097, "learning_rate": 5.528516464305467e-07, "loss": 0.0144, "step": 95500 }, { "grad_norm": 0.23016054928302765, "learning_rate": 5.504023147947068e-07, "loss": 0.0171, "step": 95510 }, { "grad_norm": 0.21064940094947815, "learning_rate": 5.479583908925279e-07, "loss": 0.0152, "step": 95520 }, { "grad_norm": 0.14792025089263916, "learning_rate": 5.455198749912626e-07, "loss": 0.0175, "step": 95530 }, { "grad_norm": 0.916067898273468, "learning_rate": 5.430867673575979e-07, "loss": 0.0135, "step": 95540 }, { "grad_norm": 0.30944234132766724, "learning_rate": 5.406590682575985e-07, "loss": 0.0132, "step": 95550 }, { "grad_norm": 0.12487044185400009, "learning_rate": 5.382367779567687e-07, "loss": 0.0308, "step": 95560 }, { "grad_norm": 0.22398532927036285, "learning_rate": 5.358198967200023e-07, "loss": 0.0144, "step": 95570 }, { "grad_norm": 0.2701584994792938, "learning_rate": 5.334084248115989e-07, "loss": 0.0126, "step": 95580 }, { "grad_norm": 0.2458311915397644, "learning_rate": 5.310023624952754e-07, "loss": 0.0124, "step": 95590 }, { "grad_norm": 0.12994785606861115, "learning_rate": 5.286017100341656e-07, "loss": 0.0122, "step": 95600 }, { "grad_norm": 0.2795974016189575, "learning_rate": 5.262064676907874e-07, "loss": 0.0167, "step": 95610 }, { "grad_norm": 0.24426142871379852, "learning_rate": 5.238166357270924e-07, "loss": 0.0124, "step": 95620 }, { "grad_norm": 0.6621856093406677, "learning_rate": 5.214322144044215e-07, "loss": 0.0161, "step": 95630 }, { "grad_norm": 0.6367968320846558, "learning_rate": 5.190532039835327e-07, "loss": 0.0219, "step": 95640 }, { "grad_norm": 0.7216346859931946, "learning_rate": 5.166796047245903e-07, "loss": 0.0313, "step": 95650 }, { "grad_norm": 0.38259243965148926, "learning_rate": 5.143114168871699e-07, "loss": 0.0171, "step": 95660 }, { "grad_norm": 0.21845731139183044, "learning_rate": 5.119486407302532e-07, "loss": 0.0104, "step": 95670 }, { "grad_norm": 0.16956914961338043, "learning_rate": 5.095912765122335e-07, "loss": 0.0174, "step": 95680 }, { "grad_norm": 0.46832263469696045, "learning_rate": 5.07239324490899e-07, "loss": 0.0413, "step": 95690 }, { "grad_norm": 0.4933759868144989, "learning_rate": 5.048927849234608e-07, "loss": 0.0234, "step": 95700 }, { "grad_norm": 0.22018380463123322, "learning_rate": 5.025516580665358e-07, "loss": 0.0126, "step": 95710 }, { "grad_norm": 0.3690893054008484, "learning_rate": 5.002159441761412e-07, "loss": 0.0115, "step": 95720 }, { "grad_norm": 0.46032479405403137, "learning_rate": 4.978856435077062e-07, "loss": 0.0156, "step": 95730 }, { "grad_norm": 0.2544989287853241, "learning_rate": 4.955607563160769e-07, "loss": 0.0138, "step": 95740 }, { "grad_norm": 0.1241331696510315, "learning_rate": 4.932412828554944e-07, "loss": 0.0139, "step": 95750 }, { "grad_norm": 0.29552486538887024, "learning_rate": 4.909272233796114e-07, "loss": 0.0129, "step": 95760 }, { "grad_norm": 0.2577337324619293, "learning_rate": 4.88618578141492e-07, "loss": 0.0222, "step": 95770 }, { "grad_norm": 0.3023649752140045, "learning_rate": 4.863153473936011e-07, "loss": 0.017, "step": 95780 }, { "grad_norm": 0.36346685886383057, "learning_rate": 4.840175313878259e-07, "loss": 0.0266, "step": 95790 }, { "grad_norm": 0.21344831585884094, "learning_rate": 4.817251303754433e-07, "loss": 0.0167, "step": 95800 }, { "grad_norm": 0.3552452027797699, "learning_rate": 4.794381446071583e-07, "loss": 0.0144, "step": 95810 }, { "grad_norm": 0.3393254578113556, "learning_rate": 4.771565743330597e-07, "loss": 0.0157, "step": 95820 }, { "grad_norm": 0.30456414818763733, "learning_rate": 4.7488041980266463e-07, "loss": 0.0141, "step": 95830 }, { "grad_norm": 0.32633471488952637, "learning_rate": 4.726096812648795e-07, "loss": 0.02, "step": 95840 }, { "grad_norm": 0.36898019909858704, "learning_rate": 4.703443589680445e-07, "loss": 0.0162, "step": 95850 }, { "grad_norm": 0.1255306452512741, "learning_rate": 4.6808445315987826e-07, "loss": 0.0167, "step": 95860 }, { "grad_norm": 0.6296116709709167, "learning_rate": 4.658299640875274e-07, "loss": 0.0199, "step": 95870 }, { "grad_norm": 0.23169681429862976, "learning_rate": 4.635808919975393e-07, "loss": 0.0104, "step": 95880 }, { "grad_norm": 0.18652865290641785, "learning_rate": 4.6133723713586706e-07, "loss": 0.0143, "step": 95890 }, { "grad_norm": 0.20340195298194885, "learning_rate": 4.590989997478756e-07, "loss": 0.016, "step": 95900 }, { "grad_norm": 0.3885968327522278, "learning_rate": 4.5686618007833027e-07, "loss": 0.0227, "step": 95910 }, { "grad_norm": 0.3822466731071472, "learning_rate": 4.54638778371419e-07, "loss": 0.0172, "step": 95920 }, { "grad_norm": 0.30065494775772095, "learning_rate": 4.524167948707192e-07, "loss": 0.0141, "step": 95930 }, { "grad_norm": 0.31133607029914856, "learning_rate": 4.5020022981921986e-07, "loss": 0.0187, "step": 95940 }, { "grad_norm": 0.5236334800720215, "learning_rate": 4.47989083459327e-07, "loss": 0.0206, "step": 95950 }, { "grad_norm": 0.37006399035453796, "learning_rate": 4.457833560328528e-07, "loss": 0.0167, "step": 95960 }, { "grad_norm": 0.24191926419734955, "learning_rate": 4.4358304778100434e-07, "loss": 0.019, "step": 95970 }, { "grad_norm": 0.24475036561489105, "learning_rate": 4.413881589444058e-07, "loss": 0.0153, "step": 95980 }, { "grad_norm": 0.20513637363910675, "learning_rate": 4.391986897630873e-07, "loss": 0.0127, "step": 95990 }, { "grad_norm": 0.24068696796894073, "learning_rate": 4.370146404764908e-07, "loss": 0.0129, "step": 96000 }, { "grad_norm": 0.37861427664756775, "learning_rate": 4.3483601132345285e-07, "loss": 0.0156, "step": 96010 }, { "grad_norm": 0.2911907732486725, "learning_rate": 4.32662802542233e-07, "loss": 0.0132, "step": 96020 }, { "grad_norm": 0.14401789009571075, "learning_rate": 4.3049501437047444e-07, "loss": 0.0168, "step": 96030 }, { "grad_norm": 0.23966355621814728, "learning_rate": 4.283326470452653e-07, "loss": 0.0175, "step": 96040 }, { "grad_norm": 0.26941564679145813, "learning_rate": 4.2617570080306647e-07, "loss": 0.0156, "step": 96050 }, { "grad_norm": 0.6531180143356323, "learning_rate": 4.2402417587975585e-07, "loss": 0.0304, "step": 96060 }, { "grad_norm": 0.38250860571861267, "learning_rate": 4.218780725106286e-07, "loss": 0.019, "step": 96070 }, { "grad_norm": 0.40119266510009766, "learning_rate": 4.197373909303748e-07, "loss": 0.0133, "step": 96080 }, { "grad_norm": 0.27399879693984985, "learning_rate": 4.17602131373096e-07, "loss": 0.0182, "step": 96090 }, { "grad_norm": 1.8034274578094482, "learning_rate": 4.154722940723055e-07, "loss": 0.0191, "step": 96100 }, { "grad_norm": 0.5493558049201965, "learning_rate": 4.1334787926091136e-07, "loss": 0.0214, "step": 96110 }, { "grad_norm": 0.29586338996887207, "learning_rate": 4.1122888717124444e-07, "loss": 0.0153, "step": 96120 }, { "grad_norm": 0.24425876140594482, "learning_rate": 4.091153180350249e-07, "loss": 0.0283, "step": 96130 }, { "grad_norm": 0.4954598844051361, "learning_rate": 4.0700717208340125e-07, "loss": 0.0297, "step": 96140 }, { "grad_norm": 0.28053218126296997, "learning_rate": 4.049044495469112e-07, "loss": 0.0111, "step": 96150 }, { "grad_norm": 0.40320590138435364, "learning_rate": 4.0280715065549866e-07, "loss": 0.0172, "step": 96160 }, { "grad_norm": 0.5795013308525085, "learning_rate": 4.0071527563853575e-07, "loss": 0.0192, "step": 96170 }, { "grad_norm": 0.4654296934604645, "learning_rate": 3.9862882472477823e-07, "loss": 0.0204, "step": 96180 }, { "grad_norm": 0.2453782856464386, "learning_rate": 3.9654779814239375e-07, "loss": 0.0147, "step": 96190 }, { "grad_norm": 0.42569634318351746, "learning_rate": 3.944721961189668e-07, "loss": 0.0126, "step": 96200 }, { "grad_norm": 0.7622908353805542, "learning_rate": 3.9240201888147697e-07, "loss": 0.0195, "step": 96210 }, { "grad_norm": 0.22847440838813782, "learning_rate": 3.9033726665632096e-07, "loss": 0.0123, "step": 96220 }, { "grad_norm": 0.2597765624523163, "learning_rate": 3.882779396692959e-07, "loss": 0.0146, "step": 96230 }, { "grad_norm": 0.32182109355926514, "learning_rate": 3.8622403814559947e-07, "loss": 0.0151, "step": 96240 }, { "grad_norm": 0.35955706238746643, "learning_rate": 3.841755623098575e-07, "loss": 0.018, "step": 96250 }, { "grad_norm": 0.19422778487205505, "learning_rate": 3.821325123860742e-07, "loss": 0.0141, "step": 96260 }, { "grad_norm": 0.20616456866264343, "learning_rate": 3.800948885976818e-07, "loss": 0.0178, "step": 96270 }, { "grad_norm": 0.179453507065773, "learning_rate": 3.780626911675078e-07, "loss": 0.009, "step": 96280 }, { "grad_norm": 0.6293070912361145, "learning_rate": 3.76035920317791e-07, "loss": 0.0162, "step": 96290 }, { "grad_norm": 0.7504730820655823, "learning_rate": 3.740145762701819e-07, "loss": 0.0215, "step": 96300 }, { "grad_norm": 0.36462119221687317, "learning_rate": 3.7199865924572585e-07, "loss": 0.0179, "step": 96310 }, { "grad_norm": 0.15769809484481812, "learning_rate": 3.699881694648799e-07, "loss": 0.0228, "step": 96320 }, { "grad_norm": 0.5726707577705383, "learning_rate": 3.679831071475126e-07, "loss": 0.0198, "step": 96330 }, { "grad_norm": 0.22178217768669128, "learning_rate": 3.65983472512893e-07, "loss": 0.0169, "step": 96340 }, { "grad_norm": 0.48145806789398193, "learning_rate": 3.6398926577969616e-07, "loss": 0.0196, "step": 96350 }, { "grad_norm": 0.4568769931793213, "learning_rate": 3.620004871660032e-07, "loss": 0.0205, "step": 96360 }, { "grad_norm": 0.43601006269454956, "learning_rate": 3.600171368893124e-07, "loss": 0.0146, "step": 96370 }, { "grad_norm": 0.2973664402961731, "learning_rate": 3.580392151665113e-07, "loss": 0.0122, "step": 96380 }, { "grad_norm": 0.3464581072330475, "learning_rate": 3.5606672221391024e-07, "loss": 0.0169, "step": 96390 }, { "grad_norm": 0.49045178294181824, "learning_rate": 3.540996582472089e-07, "loss": 0.0229, "step": 96400 }, { "grad_norm": 0.16998860239982605, "learning_rate": 3.521380234815297e-07, "loss": 0.0246, "step": 96410 }, { "grad_norm": 0.3787582814693451, "learning_rate": 3.5018181813138985e-07, "loss": 0.0103, "step": 96420 }, { "grad_norm": 0.3388907015323639, "learning_rate": 3.482310424107238e-07, "loss": 0.0103, "step": 96430 }, { "grad_norm": 0.18077504634857178, "learning_rate": 3.4628569653285537e-07, "loss": 0.0117, "step": 96440 }, { "grad_norm": 0.30294427275657654, "learning_rate": 3.44345780710531e-07, "loss": 0.0205, "step": 96450 }, { "grad_norm": 0.2866649031639099, "learning_rate": 3.424112951558978e-07, "loss": 0.0248, "step": 96460 }, { "grad_norm": 0.4254210591316223, "learning_rate": 3.40482240080503e-07, "loss": 0.016, "step": 96470 }, { "grad_norm": 0.8153874278068542, "learning_rate": 3.3855861569531133e-07, "loss": 0.035, "step": 96480 }, { "grad_norm": 0.2099212110042572, "learning_rate": 3.3664042221068226e-07, "loss": 0.0154, "step": 96490 }, { "grad_norm": 0.3152700960636139, "learning_rate": 3.347276598363924e-07, "loss": 0.0248, "step": 96500 }, { "grad_norm": 0.24617339670658112, "learning_rate": 3.328203287816134e-07, "loss": 0.0178, "step": 96510 }, { "grad_norm": 0.3888247609138489, "learning_rate": 3.3091842925493386e-07, "loss": 0.0167, "step": 96520 }, { "grad_norm": 0.42491400241851807, "learning_rate": 3.2902196146433196e-07, "loss": 0.0203, "step": 96530 }, { "grad_norm": 0.2379842847585678, "learning_rate": 3.2713092561720836e-07, "loss": 0.013, "step": 96540 }, { "grad_norm": 0.22773049771785736, "learning_rate": 3.2524532192036996e-07, "loss": 0.0114, "step": 96550 }, { "grad_norm": 0.41342446208000183, "learning_rate": 3.2336515058001836e-07, "loss": 0.0154, "step": 96560 }, { "grad_norm": 0.4184502959251404, "learning_rate": 3.2149041180176143e-07, "loss": 0.0167, "step": 96570 }, { "grad_norm": 0.4477643072605133, "learning_rate": 3.1962110579062954e-07, "loss": 0.013, "step": 96580 }, { "grad_norm": 0.37442412972450256, "learning_rate": 3.177572327510314e-07, "loss": 0.015, "step": 96590 }, { "grad_norm": 0.19472070038318634, "learning_rate": 3.158987928868151e-07, "loss": 0.0194, "step": 96600 }, { "grad_norm": 0.4014747738838196, "learning_rate": 3.1404578640120144e-07, "loss": 0.024, "step": 96610 }, { "grad_norm": 0.24745821952819824, "learning_rate": 3.1219821349683933e-07, "loss": 0.011, "step": 96620 }, { "grad_norm": 0.2837756276130676, "learning_rate": 3.103560743757783e-07, "loss": 0.0135, "step": 96630 }, { "grad_norm": 0.3723759353160858, "learning_rate": 3.0851936923946834e-07, "loss": 0.0188, "step": 96640 }, { "grad_norm": 0.43276315927505493, "learning_rate": 3.06688098288771e-07, "loss": 0.0187, "step": 96650 }, { "grad_norm": 0.16988161206245422, "learning_rate": 3.0486226172394825e-07, "loss": 0.0137, "step": 96660 }, { "grad_norm": 0.44102296233177185, "learning_rate": 3.030418597446738e-07, "loss": 0.0145, "step": 96670 }, { "grad_norm": 0.3362569510936737, "learning_rate": 3.012268925500272e-07, "loss": 0.0154, "step": 96680 }, { "grad_norm": 0.2921390235424042, "learning_rate": 2.994173603384831e-07, "loss": 0.0198, "step": 96690 }, { "grad_norm": 0.4994562864303589, "learning_rate": 2.9761326330793317e-07, "loss": 0.0154, "step": 96700 }, { "grad_norm": 0.26024556159973145, "learning_rate": 2.958146016556751e-07, "loss": 0.0098, "step": 96710 }, { "grad_norm": 0.7425609827041626, "learning_rate": 2.94021375578396e-07, "loss": 0.0166, "step": 96720 }, { "grad_norm": 0.5548093318939209, "learning_rate": 2.9223358527221135e-07, "loss": 0.0182, "step": 96730 }, { "grad_norm": 0.4852122962474823, "learning_rate": 2.904512309326257e-07, "loss": 0.024, "step": 96740 }, { "grad_norm": 0.3403434455394745, "learning_rate": 2.8867431275455545e-07, "loss": 0.014, "step": 96750 }, { "grad_norm": 0.24552299082279205, "learning_rate": 2.86902830932323e-07, "loss": 0.0199, "step": 96760 }, { "grad_norm": 0.22092178463935852, "learning_rate": 2.851367856596565e-07, "loss": 0.0158, "step": 96770 }, { "grad_norm": 0.4489211142063141, "learning_rate": 2.83376177129685e-07, "loss": 0.0128, "step": 96780 }, { "grad_norm": 0.4839756488800049, "learning_rate": 2.8162100553494884e-07, "loss": 0.0209, "step": 96790 }, { "grad_norm": 0.2432449460029602, "learning_rate": 2.798712710673834e-07, "loss": 0.0144, "step": 96800 }, { "grad_norm": 0.7086372971534729, "learning_rate": 2.7812697391834676e-07, "loss": 0.0178, "step": 96810 }, { "grad_norm": 0.2272385209798813, "learning_rate": 2.7638811427858627e-07, "loss": 0.0119, "step": 96820 }, { "grad_norm": 0.43628251552581787, "learning_rate": 2.746546923382665e-07, "loss": 0.0156, "step": 96830 }, { "grad_norm": 0.3523271083831787, "learning_rate": 2.729267082869469e-07, "loss": 0.0172, "step": 96840 }, { "grad_norm": 0.5129662156105042, "learning_rate": 2.712041623135986e-07, "loss": 0.0167, "step": 96850 }, { "grad_norm": 0.2818509340286255, "learning_rate": 2.694870546065931e-07, "loss": 0.0301, "step": 96860 }, { "grad_norm": 0.18995553255081177, "learning_rate": 2.677753853537246e-07, "loss": 0.0121, "step": 96870 }, { "grad_norm": 0.3587227463722229, "learning_rate": 2.660691547421601e-07, "loss": 0.0098, "step": 96880 }, { "grad_norm": 0.2951185703277588, "learning_rate": 2.6436836295850033e-07, "loss": 0.0136, "step": 96890 }, { "grad_norm": 0.4662298560142517, "learning_rate": 2.6267301018874093e-07, "loss": 0.0208, "step": 96900 }, { "grad_norm": 0.5728116035461426, "learning_rate": 2.6098309661828355e-07, "loss": 0.0145, "step": 96910 }, { "grad_norm": 0.14214468002319336, "learning_rate": 2.592986224319305e-07, "loss": 0.0201, "step": 96920 }, { "grad_norm": 0.20137202739715576, "learning_rate": 2.576195878139009e-07, "loss": 0.0151, "step": 96930 }, { "grad_norm": 0.267902672290802, "learning_rate": 2.5594599294780364e-07, "loss": 0.0199, "step": 96940 }, { "grad_norm": 0.5189889669418335, "learning_rate": 2.542778380166644e-07, "loss": 0.0139, "step": 96950 }, { "grad_norm": 0.5043095946311951, "learning_rate": 2.526151232029095e-07, "loss": 0.0178, "step": 96960 }, { "grad_norm": 0.14987215399742126, "learning_rate": 2.5095784868836567e-07, "loss": 0.0195, "step": 96970 }, { "grad_norm": 0.25480419397354126, "learning_rate": 2.493060146542825e-07, "loss": 0.0213, "step": 96980 }, { "grad_norm": 2.289076089859009, "learning_rate": 2.476596212812876e-07, "loss": 0.0214, "step": 96990 }, { "grad_norm": 0.15089339017868042, "learning_rate": 2.4601866874943703e-07, "loss": 0.0134, "step": 97000 }, { "grad_norm": 0.3407367169857025, "learning_rate": 2.4438315723818163e-07, "loss": 0.0165, "step": 97010 }, { "grad_norm": 0.25027212500572205, "learning_rate": 2.4275308692637833e-07, "loss": 0.0119, "step": 97020 }, { "grad_norm": 0.19609221816062927, "learning_rate": 2.411284579922901e-07, "loss": 0.0169, "step": 97030 }, { "grad_norm": 0.46933889389038086, "learning_rate": 2.395092706135804e-07, "loss": 0.0162, "step": 97040 }, { "grad_norm": 0.34830448031425476, "learning_rate": 2.3789552496731872e-07, "loss": 0.0128, "step": 97050 }, { "grad_norm": 0.47506117820739746, "learning_rate": 2.3628722122999158e-07, "loss": 0.0198, "step": 97060 }, { "grad_norm": 0.1778150349855423, "learning_rate": 2.3468435957747503e-07, "loss": 0.016, "step": 97070 }, { "grad_norm": 0.09750748425722122, "learning_rate": 2.330869401850566e-07, "loss": 0.0146, "step": 97080 }, { "grad_norm": 1.5726466178894043, "learning_rate": 2.3149496322742437e-07, "loss": 0.0196, "step": 97090 }, { "grad_norm": 0.16707392036914825, "learning_rate": 2.2990842887868346e-07, "loss": 0.0124, "step": 97100 }, { "grad_norm": 0.6750462055206299, "learning_rate": 2.2832733731232292e-07, "loss": 0.0166, "step": 97110 }, { "grad_norm": 0.3247049152851105, "learning_rate": 2.2675168870124886e-07, "loss": 0.0125, "step": 97120 }, { "grad_norm": 0.8073732852935791, "learning_rate": 2.2518148321778455e-07, "loss": 0.0275, "step": 97130 }, { "grad_norm": 0.5244402289390564, "learning_rate": 2.2361672103363706e-07, "loss": 0.0151, "step": 97140 }, { "grad_norm": 0.5869646072387695, "learning_rate": 2.220574023199251e-07, "loss": 0.0185, "step": 97150 }, { "grad_norm": 0.2772614657878876, "learning_rate": 2.2050352724717894e-07, "loss": 0.0251, "step": 97160 }, { "grad_norm": 0.3005613386631012, "learning_rate": 2.1895509598532372e-07, "loss": 0.013, "step": 97170 }, { "grad_norm": 0.20774120092391968, "learning_rate": 2.1741210870369066e-07, "loss": 0.0116, "step": 97180 }, { "grad_norm": 0.3852846622467041, "learning_rate": 2.1587456557102814e-07, "loss": 0.0116, "step": 97190 }, { "grad_norm": 0.32411184906959534, "learning_rate": 2.143424667554683e-07, "loss": 0.0111, "step": 97200 }, { "grad_norm": 0.3956531882286072, "learning_rate": 2.128158124245716e-07, "loss": 0.0421, "step": 97210 }, { "grad_norm": 0.5740496516227722, "learning_rate": 2.1129460274527668e-07, "loss": 0.06, "step": 97220 }, { "grad_norm": 0.3840274214744568, "learning_rate": 2.0977883788395046e-07, "loss": 0.0173, "step": 97230 }, { "grad_norm": 1.0808916091918945, "learning_rate": 2.0826851800634928e-07, "loss": 0.0145, "step": 97240 }, { "grad_norm": 0.47674447298049927, "learning_rate": 2.0676364327764653e-07, "loss": 0.0158, "step": 97250 }, { "grad_norm": 0.4731227457523346, "learning_rate": 2.0526421386240502e-07, "loss": 0.0136, "step": 97260 }, { "grad_norm": 0.34993976354599, "learning_rate": 2.0377022992461025e-07, "loss": 0.015, "step": 97270 }, { "grad_norm": 0.397387832403183, "learning_rate": 2.0228169162763156e-07, "loss": 0.0128, "step": 97280 }, { "grad_norm": 0.44809237122535706, "learning_rate": 2.0079859913425536e-07, "loss": 0.0152, "step": 97290 }, { "grad_norm": 0.3872731924057007, "learning_rate": 1.9932095260667417e-07, "loss": 0.023, "step": 97300 }, { "grad_norm": 0.16061462461948395, "learning_rate": 1.9784875220648092e-07, "loss": 0.0133, "step": 97310 }, { "grad_norm": 0.2980593740940094, "learning_rate": 1.9638199809466907e-07, "loss": 0.0173, "step": 97320 }, { "grad_norm": 0.6059490442276001, "learning_rate": 1.9492069043164364e-07, "loss": 0.0176, "step": 97330 }, { "grad_norm": 0.5554118156433105, "learning_rate": 1.9346482937721567e-07, "loss": 0.0209, "step": 97340 }, { "grad_norm": 0.3372240364551544, "learning_rate": 1.920144150905856e-07, "loss": 0.0102, "step": 97350 }, { "grad_norm": 0.3084697127342224, "learning_rate": 1.9056944773037655e-07, "loss": 0.0128, "step": 97360 }, { "grad_norm": 0.19856664538383484, "learning_rate": 1.8912992745460102e-07, "loss": 0.0128, "step": 97370 }, { "grad_norm": 0.36733558773994446, "learning_rate": 1.876958544206886e-07, "loss": 0.0153, "step": 97380 }, { "grad_norm": 0.29876306653022766, "learning_rate": 1.8626722878546942e-07, "loss": 0.0185, "step": 97390 }, { "grad_norm": 0.17545658349990845, "learning_rate": 1.8484405070516851e-07, "loss": 0.0083, "step": 97400 }, { "grad_norm": 0.5186609625816345, "learning_rate": 1.8342632033542805e-07, "loss": 0.0178, "step": 97410 }, { "grad_norm": 0.21242287755012512, "learning_rate": 1.8201403783129066e-07, "loss": 0.0183, "step": 97420 }, { "grad_norm": 0.26530638337135315, "learning_rate": 1.806072033471884e-07, "loss": 0.0176, "step": 97430 }, { "grad_norm": 0.3023824095726013, "learning_rate": 1.79205817036987e-07, "loss": 0.0128, "step": 97440 }, { "grad_norm": 0.5836938619613647, "learning_rate": 1.7780987905393065e-07, "loss": 0.0178, "step": 97450 }, { "grad_norm": 0.3286891579627991, "learning_rate": 1.764193895506805e-07, "loss": 0.0134, "step": 97460 }, { "grad_norm": 0.33907097578048706, "learning_rate": 1.7503434867929824e-07, "loss": 0.015, "step": 97470 }, { "grad_norm": 0.3254438042640686, "learning_rate": 1.7365475659125165e-07, "loss": 0.0179, "step": 97480 }, { "grad_norm": 0.31755971908569336, "learning_rate": 1.7228061343740332e-07, "loss": 0.0111, "step": 97490 }, { "grad_norm": 0.4055589735507965, "learning_rate": 1.7091191936803308e-07, "loss": 0.0128, "step": 97500 }, { "grad_norm": 0.5794509053230286, "learning_rate": 1.6954867453281563e-07, "loss": 0.0171, "step": 97510 }, { "grad_norm": 0.30632883310317993, "learning_rate": 1.6819087908084285e-07, "loss": 0.0186, "step": 97520 }, { "grad_norm": 0.44087985157966614, "learning_rate": 1.6683853316058485e-07, "loss": 0.0145, "step": 97530 }, { "grad_norm": 0.3243960440158844, "learning_rate": 1.6549163691995106e-07, "loss": 0.0117, "step": 97540 }, { "grad_norm": 0.4328264594078064, "learning_rate": 1.641501905062237e-07, "loss": 0.0199, "step": 97550 }, { "grad_norm": 0.4390639066696167, "learning_rate": 1.628141940661021e-07, "loss": 0.0177, "step": 97560 }, { "grad_norm": 0.26281508803367615, "learning_rate": 1.6148364774569158e-07, "loss": 0.0151, "step": 97570 }, { "grad_norm": 0.4421093761920929, "learning_rate": 1.6015855169050355e-07, "loss": 0.0157, "step": 97580 }, { "grad_norm": 0.2131255716085434, "learning_rate": 1.5883890604543873e-07, "loss": 0.0105, "step": 97590 }, { "grad_norm": 0.20361872017383575, "learning_rate": 1.5752471095482057e-07, "loss": 0.0107, "step": 97600 }, { "grad_norm": 0.369059681892395, "learning_rate": 1.5621596656235638e-07, "loss": 0.0145, "step": 97610 }, { "grad_norm": 0.7399066090583801, "learning_rate": 1.5491267301117608e-07, "loss": 0.022, "step": 97620 }, { "grad_norm": 0.40379598736763, "learning_rate": 1.536148304438101e-07, "loss": 0.0195, "step": 97630 }, { "grad_norm": 0.4313834309577942, "learning_rate": 1.5232243900217823e-07, "loss": 0.0213, "step": 97640 }, { "grad_norm": 0.4490598440170288, "learning_rate": 1.51035498827623e-07, "loss": 0.0187, "step": 97650 }, { "grad_norm": 0.2921987771987915, "learning_rate": 1.497540100608763e-07, "loss": 0.0168, "step": 97660 }, { "grad_norm": 0.1723998486995697, "learning_rate": 1.4847797284208153e-07, "loss": 0.0153, "step": 97670 }, { "grad_norm": 0.21790407598018646, "learning_rate": 1.472073873107882e-07, "loss": 0.0215, "step": 97680 }, { "grad_norm": 0.486499160528183, "learning_rate": 1.459422536059407e-07, "loss": 0.0156, "step": 97690 }, { "grad_norm": 0.43528422713279724, "learning_rate": 1.4468257186589508e-07, "loss": 0.0187, "step": 97700 }, { "grad_norm": 0.4928846061229706, "learning_rate": 1.4342834222840217e-07, "loss": 0.018, "step": 97710 }, { "grad_norm": 0.2599274218082428, "learning_rate": 1.4217956483063566e-07, "loss": 0.0146, "step": 97720 }, { "grad_norm": 0.5386921763420105, "learning_rate": 1.409362398091474e-07, "loss": 0.0125, "step": 97730 }, { "grad_norm": 0.2984285354614258, "learning_rate": 1.3969836729990638e-07, "loss": 0.0194, "step": 97740 }, { "grad_norm": 0.22100837528705597, "learning_rate": 1.384659474382932e-07, "loss": 0.0287, "step": 97750 }, { "grad_norm": 0.25803354382514954, "learning_rate": 1.372389803590779e-07, "loss": 0.0131, "step": 97760 }, { "grad_norm": 0.35474643111228943, "learning_rate": 1.360174661964364e-07, "loss": 0.0162, "step": 97770 }, { "grad_norm": 0.29070693254470825, "learning_rate": 1.3480140508396188e-07, "loss": 0.0228, "step": 97780 }, { "grad_norm": 0.4553469717502594, "learning_rate": 1.335907971546313e-07, "loss": 0.0197, "step": 97790 }, { "grad_norm": 0.3768312633037567, "learning_rate": 1.3238564254083875e-07, "loss": 0.0154, "step": 97800 }, { "grad_norm": 0.28359830379486084, "learning_rate": 1.3118594137437323e-07, "loss": 0.0131, "step": 97810 }, { "grad_norm": 0.6299105286598206, "learning_rate": 1.2999169378644094e-07, "loss": 0.022, "step": 97820 }, { "grad_norm": 0.1478554755449295, "learning_rate": 1.2880289990763738e-07, "loss": 0.0133, "step": 97830 }, { "grad_norm": 0.5339816808700562, "learning_rate": 1.2761955986796968e-07, "loss": 0.0195, "step": 97840 }, { "grad_norm": 0.23771561682224274, "learning_rate": 1.264416737968399e-07, "loss": 0.0117, "step": 97850 }, { "grad_norm": 0.08835217356681824, "learning_rate": 1.2526924182307275e-07, "loss": 0.0116, "step": 97860 }, { "grad_norm": 0.20771361887454987, "learning_rate": 1.2410226407487124e-07, "loss": 0.0133, "step": 97870 }, { "grad_norm": 0.613595187664032, "learning_rate": 1.229407406798555e-07, "loss": 0.014, "step": 97880 }, { "grad_norm": 0.2737872302532196, "learning_rate": 1.2178467176505726e-07, "loss": 0.0098, "step": 97890 }, { "grad_norm": 0.4290003776550293, "learning_rate": 1.2063405745689205e-07, "loss": 0.0208, "step": 97900 }, { "grad_norm": 0.1919228881597519, "learning_rate": 1.1948889788119812e-07, "loss": 0.0107, "step": 97910 }, { "grad_norm": 0.27763041853904724, "learning_rate": 1.1834919316320303e-07, "loss": 0.016, "step": 97920 }, { "grad_norm": 0.4198698401451111, "learning_rate": 1.1721494342754048e-07, "loss": 0.015, "step": 97930 }, { "grad_norm": 0.24732796847820282, "learning_rate": 1.1608614879825563e-07, "loss": 0.0129, "step": 97940 }, { "grad_norm": 0.1122916042804718, "learning_rate": 1.1496280939879422e-07, "loss": 0.0099, "step": 97950 }, { "grad_norm": 0.4086534082889557, "learning_rate": 1.1384492535199687e-07, "loss": 0.0149, "step": 97960 }, { "grad_norm": 0.179981991648674, "learning_rate": 1.1273249678011578e-07, "loss": 0.0114, "step": 97970 }, { "grad_norm": 0.1706319898366928, "learning_rate": 1.1162552380480362e-07, "loss": 0.0178, "step": 97980 }, { "grad_norm": 0.2674238085746765, "learning_rate": 1.1052400654711359e-07, "loss": 0.0117, "step": 97990 }, { "grad_norm": 0.39754176139831543, "learning_rate": 1.0942794512751597e-07, "loss": 0.0248, "step": 98000 }, { "grad_norm": 0.27277132868766785, "learning_rate": 1.0833733966587045e-07, "loss": 0.0163, "step": 98010 }, { "grad_norm": 0.4034135043621063, "learning_rate": 1.0725219028143718e-07, "loss": 0.026, "step": 98020 }, { "grad_norm": 0.31097719073295593, "learning_rate": 1.0617249709289345e-07, "loss": 0.0183, "step": 98030 }, { "grad_norm": 0.395986407995224, "learning_rate": 1.0509826021831703e-07, "loss": 0.0128, "step": 98040 }, { "grad_norm": 0.31627246737480164, "learning_rate": 1.0402947977517508e-07, "loss": 0.0125, "step": 98050 }, { "grad_norm": 0.3914482593536377, "learning_rate": 1.0296615588035185e-07, "loss": 0.0165, "step": 98060 }, { "grad_norm": 0.3232404887676239, "learning_rate": 1.0190828865012659e-07, "loss": 0.0193, "step": 98070 }, { "grad_norm": 0.20973415672779083, "learning_rate": 1.008558782001956e-07, "loss": 0.0239, "step": 98080 }, { "grad_norm": 0.20541690289974213, "learning_rate": 9.980892464563906e-08, "loss": 0.0194, "step": 98090 }, { "grad_norm": 0.45635896921157837, "learning_rate": 9.876742810095985e-08, "loss": 0.0224, "step": 98100 }, { "grad_norm": 0.3541387617588043, "learning_rate": 9.77313886800446e-08, "loss": 0.0128, "step": 98110 }, { "grad_norm": 0.14249972999095917, "learning_rate": 9.670080649619717e-08, "loss": 0.0229, "step": 98120 }, { "grad_norm": 0.5747587084770203, "learning_rate": 9.567568166212737e-08, "loss": 0.0268, "step": 98130 }, { "grad_norm": 0.20219570398330688, "learning_rate": 9.465601428992888e-08, "loss": 0.0162, "step": 98140 }, { "grad_norm": 0.21568743884563446, "learning_rate": 9.364180449111803e-08, "loss": 0.015, "step": 98150 }, { "grad_norm": 0.27849265933036804, "learning_rate": 9.263305237661169e-08, "loss": 0.0191, "step": 98160 }, { "grad_norm": 0.5992693305015564, "learning_rate": 9.162975805671603e-08, "loss": 0.012, "step": 98170 }, { "grad_norm": 0.2716355323791504, "learning_rate": 9.063192164115442e-08, "loss": 0.0202, "step": 98180 }, { "grad_norm": 0.36774569749832153, "learning_rate": 8.963954323904511e-08, "loss": 0.0239, "step": 98190 }, { "grad_norm": 0.41739732027053833, "learning_rate": 8.865262295891796e-08, "loss": 0.0112, "step": 98200 }, { "grad_norm": 0.4967678189277649, "learning_rate": 8.767116090870331e-08, "loss": 0.0169, "step": 98210 }, { "grad_norm": 0.4705413281917572, "learning_rate": 8.669515719572086e-08, "loss": 0.0148, "step": 98220 }, { "grad_norm": 0.27817046642303467, "learning_rate": 8.572461192671855e-08, "loss": 0.016, "step": 98230 }, { "grad_norm": 0.5962724089622498, "learning_rate": 8.475952520782815e-08, "loss": 0.0249, "step": 98240 }, { "grad_norm": 0.5093677043914795, "learning_rate": 8.379989714458747e-08, "loss": 0.0148, "step": 98250 }, { "grad_norm": 0.1844964176416397, "learning_rate": 8.284572784194034e-08, "loss": 0.0165, "step": 98260 }, { "grad_norm": 0.2912473976612091, "learning_rate": 8.189701740424216e-08, "loss": 0.0125, "step": 98270 }, { "grad_norm": 0.24619469046592712, "learning_rate": 8.095376593522663e-08, "loss": 0.0168, "step": 98280 }, { "grad_norm": 0.11283271759748459, "learning_rate": 8.001597353806123e-08, "loss": 0.0153, "step": 98290 }, { "grad_norm": 0.2907082438468933, "learning_rate": 7.90836403152917e-08, "loss": 0.0111, "step": 98300 }, { "grad_norm": 0.3843969702720642, "learning_rate": 7.815676636888092e-08, "loss": 0.0135, "step": 98310 }, { "grad_norm": 0.5154813528060913, "learning_rate": 7.723535180019226e-08, "loss": 0.0146, "step": 98320 }, { "grad_norm": 0.5143082141876221, "learning_rate": 7.6319396709984e-08, "loss": 0.0157, "step": 98330 }, { "grad_norm": 0.7793170809745789, "learning_rate": 7.540890119843158e-08, "loss": 0.0175, "step": 98340 }, { "grad_norm": 0.34333306550979614, "learning_rate": 7.450386536509979e-08, "loss": 0.0111, "step": 98350 }, { "grad_norm": 0.36885708570480347, "learning_rate": 7.360428930895947e-08, "loss": 0.0146, "step": 98360 }, { "grad_norm": 0.26740309596061707, "learning_rate": 7.271017312839302e-08, "loss": 0.0143, "step": 98370 }, { "grad_norm": 0.2552538216114044, "learning_rate": 7.182151692118333e-08, "loss": 0.0155, "step": 98380 }, { "grad_norm": 0.17905747890472412, "learning_rate": 7.093832078449713e-08, "loss": 0.0112, "step": 98390 }, { "grad_norm": 0.2096264660358429, "learning_rate": 7.006058481493493e-08, "loss": 0.0172, "step": 98400 }, { "grad_norm": 0.6393709182739258, "learning_rate": 6.918830910847552e-08, "loss": 0.0194, "step": 98410 }, { "grad_norm": 0.27317702770233154, "learning_rate": 6.832149376052033e-08, "loss": 0.0143, "step": 98420 }, { "grad_norm": 0.11328810453414917, "learning_rate": 6.746013886584912e-08, "loss": 0.0127, "step": 98430 }, { "grad_norm": 0.3155882656574249, "learning_rate": 6.660424451866431e-08, "loss": 0.0189, "step": 98440 }, { "grad_norm": 0.14672298729419708, "learning_rate": 6.575381081256327e-08, "loss": 0.0115, "step": 98450 }, { "grad_norm": 0.6814141869544983, "learning_rate": 6.49088378405549e-08, "loss": 0.0197, "step": 98460 }, { "grad_norm": 0.4028204083442688, "learning_rate": 6.406932569503755e-08, "loss": 0.0213, "step": 98470 }, { "grad_norm": 0.34427690505981445, "learning_rate": 6.323527446782662e-08, "loss": 0.0176, "step": 98480 }, { "grad_norm": 0.14326712489128113, "learning_rate": 6.240668425012142e-08, "loss": 0.0175, "step": 98490 }, { "grad_norm": 0.3654484152793884, "learning_rate": 6.158355513254388e-08, "loss": 0.011, "step": 98500 }, { "grad_norm": 0.36187806725502014, "learning_rate": 6.076588720510534e-08, "loss": 0.0182, "step": 98510 }, { "grad_norm": 0.3411220610141754, "learning_rate": 5.995368055722316e-08, "loss": 0.0155, "step": 98520 }, { "grad_norm": 0.5392399430274963, "learning_rate": 5.914693527773185e-08, "loss": 0.0119, "step": 98530 }, { "grad_norm": 0.29452747106552124, "learning_rate": 5.834565145483861e-08, "loss": 0.0182, "step": 98540 }, { "grad_norm": 0.32921290397644043, "learning_rate": 5.754982917618446e-08, "loss": 0.0193, "step": 98550 }, { "grad_norm": 0.18671530485153198, "learning_rate": 5.675946852878866e-08, "loss": 0.0136, "step": 98560 }, { "grad_norm": 0.2912750840187073, "learning_rate": 5.59745695990932e-08, "loss": 0.0174, "step": 98570 }, { "grad_norm": 0.8421001434326172, "learning_rate": 5.51951324729294e-08, "loss": 0.0199, "step": 98580 }, { "grad_norm": 0.48905354738235474, "learning_rate": 5.4421157235529093e-08, "loss": 0.0129, "step": 98590 }, { "grad_norm": 0.4057646095752716, "learning_rate": 5.365264397154679e-08, "loss": 0.015, "step": 98600 }, { "grad_norm": 0.17166711390018463, "learning_rate": 5.288959276501526e-08, "loss": 0.0136, "step": 98610 }, { "grad_norm": 0.4759710133075714, "learning_rate": 5.2132003699378876e-08, "loss": 0.0156, "step": 98620 }, { "grad_norm": 0.9489527940750122, "learning_rate": 5.137987685749912e-08, "loss": 0.0132, "step": 98630 }, { "grad_norm": 0.2727106511592865, "learning_rate": 5.0633212321610227e-08, "loss": 0.0175, "step": 98640 }, { "grad_norm": 0.4447094798088074, "learning_rate": 4.989201017338019e-08, "loss": 0.0148, "step": 98650 }, { "grad_norm": 0.3952520191669464, "learning_rate": 4.9156270493860846e-08, "loss": 0.0147, "step": 98660 }, { "grad_norm": 0.23740091919898987, "learning_rate": 4.842599336351561e-08, "loss": 0.0213, "step": 98670 }, { "grad_norm": 0.32079818844795227, "learning_rate": 4.7701178862197274e-08, "loss": 0.0202, "step": 98680 }, { "grad_norm": 0.6034818291664124, "learning_rate": 4.698182706918131e-08, "loss": 0.0192, "step": 98690 }, { "grad_norm": 0.38954076170921326, "learning_rate": 4.6267938063121466e-08, "loss": 0.0112, "step": 98700 }, { "grad_norm": 0.5157155394554138, "learning_rate": 4.555951192209973e-08, "loss": 0.0168, "step": 98710 }, { "grad_norm": 0.26583582162857056, "learning_rate": 4.485654872358747e-08, "loss": 0.0184, "step": 98720 }, { "grad_norm": 0.24547988176345825, "learning_rate": 4.415904854445097e-08, "loss": 0.0175, "step": 98730 }, { "grad_norm": 0.3231584429740906, "learning_rate": 4.34670114609792e-08, "loss": 0.0161, "step": 98740 }, { "grad_norm": 0.4529385566711426, "learning_rate": 4.278043754884498e-08, "loss": 0.0161, "step": 98750 }, { "grad_norm": 0.3490164875984192, "learning_rate": 4.209932688312712e-08, "loss": 0.0142, "step": 98760 }, { "grad_norm": 0.43202659487724304, "learning_rate": 4.14236795383216e-08, "loss": 0.0128, "step": 98770 }, { "grad_norm": 0.38210275769233704, "learning_rate": 4.075349558830821e-08, "loss": 0.0293, "step": 98780 }, { "grad_norm": 0.1720951348543167, "learning_rate": 4.008877510638387e-08, "loss": 0.0123, "step": 98790 }, { "grad_norm": 0.3520962595939636, "learning_rate": 3.942951816523488e-08, "loss": 0.0127, "step": 98800 }, { "grad_norm": 0.39785459637641907, "learning_rate": 3.8775724836959125e-08, "loss": 0.0138, "step": 98810 }, { "grad_norm": 0.28143125772476196, "learning_rate": 3.812739519305497e-08, "loss": 0.0104, "step": 98820 }, { "grad_norm": 0.3978322148323059, "learning_rate": 3.748452930442126e-08, "loss": 0.0218, "step": 98830 }, { "grad_norm": 0.29584866762161255, "learning_rate": 3.6847127241362855e-08, "loss": 0.0126, "step": 98840 }, { "grad_norm": 0.35982754826545715, "learning_rate": 3.621518907358512e-08, "loss": 0.0128, "step": 98850 }, { "grad_norm": 0.7130780220031738, "learning_rate": 3.558871487019388e-08, "loss": 0.0166, "step": 98860 }, { "grad_norm": 0.9649338722229004, "learning_rate": 3.4967704699701007e-08, "loss": 0.016, "step": 98870 }, { "grad_norm": 0.2755472958087921, "learning_rate": 3.435215863001884e-08, "loss": 0.014, "step": 98880 }, { "grad_norm": 0.2545510530471802, "learning_rate": 3.374207672846019e-08, "loss": 0.0167, "step": 98890 }, { "grad_norm": 0.7164398431777954, "learning_rate": 3.313745906174392e-08, "loss": 0.0151, "step": 98900 }, { "grad_norm": 0.3395403027534485, "learning_rate": 3.253830569599492e-08, "loss": 0.0145, "step": 98910 }, { "grad_norm": 0.3702397346496582, "learning_rate": 3.1944616696727436e-08, "loss": 0.0122, "step": 98920 }, { "grad_norm": 0.3331102430820465, "learning_rate": 3.135639212887287e-08, "loss": 0.011, "step": 98930 }, { "grad_norm": 0.29544591903686523, "learning_rate": 3.077363205675754e-08, "loss": 0.01, "step": 98940 }, { "grad_norm": 0.13773047924041748, "learning_rate": 3.0196336544113804e-08, "loss": 0.0141, "step": 98950 }, { "grad_norm": 0.3135790526866913, "learning_rate": 2.9624505654063383e-08, "loss": 0.0161, "step": 98960 }, { "grad_norm": 0.592689573764801, "learning_rate": 2.90581394491507e-08, "loss": 0.016, "step": 98970 }, { "grad_norm": 0.3597588837146759, "learning_rate": 2.8497237991309545e-08, "loss": 0.0123, "step": 98980 }, { "grad_norm": 0.1917456090450287, "learning_rate": 2.7941801341879735e-08, "loss": 0.0258, "step": 98990 }, { "grad_norm": 0.14756110310554504, "learning_rate": 2.7391829561601578e-08, "loss": 0.0192, "step": 99000 }, { "grad_norm": 0.5038615465164185, "learning_rate": 2.6847322710621403e-08, "loss": 0.0192, "step": 99010 }, { "grad_norm": 0.2883153259754181, "learning_rate": 2.6308280848486022e-08, "loss": 0.0271, "step": 99020 }, { "grad_norm": 0.428434282541275, "learning_rate": 2.5774704034137177e-08, "loss": 0.0132, "step": 99030 }, { "grad_norm": 0.2624412477016449, "learning_rate": 2.524659232593929e-08, "loss": 0.0175, "step": 99040 }, { "grad_norm": 0.5317081809043884, "learning_rate": 2.472394578163506e-08, "loss": 0.0219, "step": 99050 }, { "grad_norm": 0.26963526010513306, "learning_rate": 2.4206764458378772e-08, "loss": 0.0154, "step": 99060 }, { "grad_norm": 0.2566838562488556, "learning_rate": 2.3695048412736285e-08, "loss": 0.0159, "step": 99070 }, { "grad_norm": 0.2773998975753784, "learning_rate": 2.3188797700668395e-08, "loss": 0.0157, "step": 99080 }, { "grad_norm": 0.31440892815589905, "learning_rate": 2.268801237753082e-08, "loss": 0.0124, "step": 99090 }, { "grad_norm": 0.2517511546611786, "learning_rate": 2.2192692498090862e-08, "loss": 0.0206, "step": 99100 }, { "grad_norm": 0.1914127767086029, "learning_rate": 2.170283811652185e-08, "loss": 0.017, "step": 99110 }, { "grad_norm": 0.25203460454940796, "learning_rate": 2.12184492863865e-08, "loss": 0.0122, "step": 99120 }, { "grad_norm": 0.3888562023639679, "learning_rate": 2.073952606066465e-08, "loss": 0.0143, "step": 99130 }, { "grad_norm": 0.45740756392478943, "learning_rate": 2.026606849171997e-08, "loss": 0.0208, "step": 99140 }, { "grad_norm": 0.4658254384994507, "learning_rate": 1.9798076631333262e-08, "loss": 0.0171, "step": 99150 }, { "grad_norm": 0.3553817570209503, "learning_rate": 1.933555053069136e-08, "loss": 0.0197, "step": 99160 }, { "grad_norm": 0.6487941741943359, "learning_rate": 1.887849024036492e-08, "loss": 0.0281, "step": 99170 }, { "grad_norm": 0.34305885434150696, "learning_rate": 1.8426895810341736e-08, "loss": 0.0142, "step": 99180 }, { "grad_norm": 0.4659314751625061, "learning_rate": 1.798076729000453e-08, "loss": 0.0154, "step": 99190 }, { "grad_norm": 0.4170774817466736, "learning_rate": 1.75401047281476e-08, "loss": 0.0177, "step": 99200 }, { "grad_norm": 0.44780245423316956, "learning_rate": 1.7104908172954626e-08, "loss": 0.0185, "step": 99210 }, { "grad_norm": 0.5822029113769531, "learning_rate": 1.6675177672020868e-08, "loss": 0.0248, "step": 99220 }, { "grad_norm": 0.25369808077812195, "learning_rate": 1.6250913272342073e-08, "loss": 0.018, "step": 99230 }, { "grad_norm": 0.49796029925346375, "learning_rate": 1.5832115020314453e-08, "loss": 0.0156, "step": 99240 }, { "grad_norm": 0.41383248567581177, "learning_rate": 1.5418782961734712e-08, "loss": 0.0177, "step": 99250 }, { "grad_norm": 0.37814775109291077, "learning_rate": 1.501091714181113e-08, "loss": 0.0179, "step": 99260 }, { "grad_norm": 0.23542365431785583, "learning_rate": 1.460851760513582e-08, "loss": 0.0148, "step": 99270 }, { "grad_norm": 0.3305877149105072, "learning_rate": 1.4211584395723566e-08, "loss": 0.0148, "step": 99280 }, { "grad_norm": 0.35437795519828796, "learning_rate": 1.382011755697854e-08, "loss": 0.0121, "step": 99290 }, { "grad_norm": 0.19865572452545166, "learning_rate": 1.3434117131716494e-08, "loss": 0.0121, "step": 99300 }, { "grad_norm": 0.34387803077697754, "learning_rate": 1.3053583162142557e-08, "loss": 0.0173, "step": 99310 }, { "grad_norm": 0.12797753512859344, "learning_rate": 1.2678515689873438e-08, "loss": 0.014, "step": 99320 }, { "grad_norm": 0.18021519482135773, "learning_rate": 1.2308914755931877e-08, "loss": 0.0142, "step": 99330 }, { "grad_norm": 0.5654650926589966, "learning_rate": 1.1944780400729993e-08, "loss": 0.0212, "step": 99340 }, { "grad_norm": 0.5232794880867004, "learning_rate": 1.1586112664085935e-08, "loss": 0.0134, "step": 99350 }, { "grad_norm": 0.308339387178421, "learning_rate": 1.1232911585234984e-08, "loss": 0.012, "step": 99360 }, { "grad_norm": 0.3068558871746063, "learning_rate": 1.0885177202790697e-08, "loss": 0.0176, "step": 99370 }, { "grad_norm": 0.7003617286682129, "learning_rate": 1.0542909554789315e-08, "loss": 0.0172, "step": 99380 }, { "grad_norm": 0.4298933148384094, "learning_rate": 1.0206108678656456e-08, "loss": 0.023, "step": 99390 }, { "grad_norm": 0.2905767261981964, "learning_rate": 9.874774611223769e-09, "loss": 0.0084, "step": 99400 }, { "grad_norm": 0.1583160012960434, "learning_rate": 9.548907388728933e-09, "loss": 0.0181, "step": 99410 }, { "grad_norm": 0.2848087549209595, "learning_rate": 9.228507046804557e-09, "loss": 0.0151, "step": 99420 }, { "grad_norm": 0.4366418421268463, "learning_rate": 8.91357362048928e-09, "loss": 0.0119, "step": 99430 }, { "grad_norm": 0.3121880292892456, "learning_rate": 8.604107144227769e-09, "loss": 0.0285, "step": 99440 }, { "grad_norm": 0.11551947891712189, "learning_rate": 8.300107651859623e-09, "loss": 0.0172, "step": 99450 }, { "grad_norm": 0.18855120241641998, "learning_rate": 8.001575176630472e-09, "loss": 0.0129, "step": 99460 }, { "grad_norm": 0.5641087293624878, "learning_rate": 7.708509751186422e-09, "loss": 0.0192, "step": 99470 }, { "grad_norm": 0.5328437685966492, "learning_rate": 7.420911407579611e-09, "loss": 0.0185, "step": 99480 }, { "grad_norm": 0.36338064074516296, "learning_rate": 7.13878017725711e-09, "loss": 0.0174, "step": 99490 }, { "grad_norm": 0.30861231684684753, "learning_rate": 6.8621160910720125e-09, "loss": 0.0161, "step": 99500 }, { "grad_norm": 0.38982367515563965, "learning_rate": 6.59091917928345e-09, "loss": 0.0174, "step": 99510 }, { "grad_norm": 0.508311927318573, "learning_rate": 6.32518947155103e-09, "loss": 0.0109, "step": 99520 }, { "grad_norm": 0.4426022469997406, "learning_rate": 6.064926996929288e-09, "loss": 0.0133, "step": 99530 }, { "grad_norm": 0.15114620327949524, "learning_rate": 5.810131783884343e-09, "loss": 0.0201, "step": 99540 }, { "grad_norm": 0.4053741991519928, "learning_rate": 5.5608038602772415e-09, "loss": 0.0141, "step": 99550 }, { "grad_norm": 0.4873444736003876, "learning_rate": 5.316943253375062e-09, "loss": 0.0112, "step": 99560 }, { "grad_norm": 0.183027982711792, "learning_rate": 5.07854998984536e-09, "loss": 0.0144, "step": 99570 }, { "grad_norm": 0.32817113399505615, "learning_rate": 4.845624095756174e-09, "loss": 0.0126, "step": 99580 }, { "grad_norm": 0.9769222140312195, "learning_rate": 4.618165596587121e-09, "loss": 0.0253, "step": 99590 }, { "grad_norm": 0.15335343778133392, "learning_rate": 4.396174517207197e-09, "loss": 0.0146, "step": 99600 }, { "grad_norm": 0.2033723145723343, "learning_rate": 4.179650881896979e-09, "loss": 0.0207, "step": 99610 }, { "grad_norm": 0.2810041606426239, "learning_rate": 3.968594714331975e-09, "loss": 0.0131, "step": 99620 }, { "grad_norm": 0.3412679135799408, "learning_rate": 3.763006037593719e-09, "loss": 0.0193, "step": 99630 }, { "grad_norm": 0.18657104671001434, "learning_rate": 3.5628848741586786e-09, "loss": 0.0135, "step": 99640 }, { "grad_norm": 0.4070751667022705, "learning_rate": 3.368231245926001e-09, "loss": 0.021, "step": 99650 }, { "grad_norm": 0.3465459942817688, "learning_rate": 3.179045174167561e-09, "loss": 0.0128, "step": 99660 }, { "grad_norm": 0.2612697184085846, "learning_rate": 2.9953266795834657e-09, "loss": 0.016, "step": 99670 }, { "grad_norm": 0.23063728213310242, "learning_rate": 2.817075782263201e-09, "loss": 0.0239, "step": 99680 }, { "grad_norm": 0.26045501232147217, "learning_rate": 2.6442925016967325e-09, "loss": 0.0245, "step": 99690 }, { "grad_norm": 0.3955804705619812, "learning_rate": 2.4769768567800555e-09, "loss": 0.0119, "step": 99700 }, { "grad_norm": 0.32010993361473083, "learning_rate": 2.315128865809646e-09, "loss": 0.0167, "step": 99710 }, { "grad_norm": 0.6490411162376404, "learning_rate": 2.1587485464880097e-09, "loss": 0.0125, "step": 99720 }, { "grad_norm": 0.28152358531951904, "learning_rate": 2.0078359159125813e-09, "loss": 0.0162, "step": 99730 }, { "grad_norm": 0.5732853412628174, "learning_rate": 1.8623909905923776e-09, "loss": 0.0166, "step": 99740 }, { "grad_norm": 0.27047908306121826, "learning_rate": 1.7224137864257916e-09, "loss": 0.0156, "step": 99750 }, { "grad_norm": 0.3840849697589874, "learning_rate": 1.5879043187283505e-09, "loss": 0.0213, "step": 99760 }, { "grad_norm": 0.16109110414981842, "learning_rate": 1.458862602204958e-09, "loss": 0.021, "step": 99770 }, { "grad_norm": 0.38918817043304443, "learning_rate": 1.3352886509720996e-09, "loss": 0.0181, "step": 99780 }, { "grad_norm": 0.49997612833976746, "learning_rate": 1.2171824785356389e-09, "loss": 0.0151, "step": 99790 }, { "grad_norm": 0.26328080892562866, "learning_rate": 1.1045440978185718e-09, "loss": 0.0171, "step": 99800 }, { "grad_norm": 0.44085079431533813, "learning_rate": 9.973735211332714e-10, "loss": 0.016, "step": 99810 }, { "grad_norm": 0.3360311686992645, "learning_rate": 8.956707602036929e-10, "loss": 0.012, "step": 99820 }, { "grad_norm": 0.49286797642707825, "learning_rate": 7.994358261542712e-10, "loss": 0.0168, "step": 99830 }, { "grad_norm": 0.3818642199039459, "learning_rate": 7.086687295043692e-10, "loss": 0.018, "step": 99840 }, { "grad_norm": 0.2695784568786621, "learning_rate": 6.23369480179381e-10, "loss": 0.0115, "step": 99850 }, { "grad_norm": 0.13492344319820404, "learning_rate": 5.43538087510731e-10, "loss": 0.0122, "step": 99860 }, { "grad_norm": 0.5490757822990417, "learning_rate": 4.691745602303233e-10, "loss": 0.0219, "step": 99870 }, { "grad_norm": 0.2996133863925934, "learning_rate": 4.0027890646499034e-10, "loss": 0.0123, "step": 99880 }, { "grad_norm": 0.2603457570075989, "learning_rate": 3.368511337531466e-10, "loss": 0.0188, "step": 99890 }, { "grad_norm": 0.665877103805542, "learning_rate": 2.788912490281348e-10, "loss": 0.0165, "step": 99900 }, { "grad_norm": 0.4178089201450348, "learning_rate": 2.2639925863487954e-10, "loss": 0.0166, "step": 99910 }, { "grad_norm": 0.32004544138908386, "learning_rate": 1.7937516830213163e-10, "loss": 0.0192, "step": 99920 }, { "grad_norm": 0.9397129416465759, "learning_rate": 1.3781898318687702e-10, "loss": 0.0166, "step": 99930 }, { "grad_norm": 0.41126611828804016, "learning_rate": 1.0173070781882566e-10, "loss": 0.0132, "step": 99940 }, { "grad_norm": 0.3971182405948639, "learning_rate": 7.111034615592261e-11, "loss": 0.0197, "step": 99950 }, { "grad_norm": 0.5176187753677368, "learning_rate": 4.595790153993918e-11, "loss": 0.0222, "step": 99960 }, { "grad_norm": 0.4477447271347046, "learning_rate": 2.6273376729779585e-11, "loss": 0.0138, "step": 99970 }, { "grad_norm": 0.3359496593475342, "learning_rate": 1.2056773868174276e-11, "loss": 0.0141, "step": 99980 }, { "grad_norm": 0.3740485608577728, "learning_rate": 3.3080945149865926e-12, "loss": 0.0125, "step": 99990 }, { "grad_norm": 0.2738246023654938, "learning_rate": 2.7339625008337977e-14, "loss": 0.0172, "step": 100000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 5000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 4, "trial_name": null, "trial_params": null }