{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 40000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.860569953918457, "learning_rate": 2.25e-07, "loss": 1.3243, "step": 10 }, { "grad_norm": 1.8622368574142456, "learning_rate": 4.75e-07, "loss": 1.3239, "step": 20 }, { "grad_norm": 1.9061133861541748, "learning_rate": 7.25e-07, "loss": 1.3229, "step": 30 }, { "grad_norm": 1.584997534751892, "learning_rate": 9.75e-07, "loss": 1.3205, "step": 40 }, { "grad_norm": 1.3556013107299805, "learning_rate": 1.2250000000000001e-06, "loss": 1.3025, "step": 50 }, { "grad_norm": 0.9051105380058289, "learning_rate": 1.475e-06, "loss": 1.268, "step": 60 }, { "grad_norm": 0.647619366645813, "learning_rate": 1.7250000000000002e-06, "loss": 1.2434, "step": 70 }, { "grad_norm": 0.5935598611831665, "learning_rate": 1.975e-06, "loss": 1.2216, "step": 80 }, { "grad_norm": 0.4965655207633972, "learning_rate": 2.225e-06, "loss": 1.2094, "step": 90 }, { "grad_norm": 0.4474756419658661, "learning_rate": 2.4750000000000004e-06, "loss": 1.212, "step": 100 }, { "grad_norm": 0.5706126689910889, "learning_rate": 2.725e-06, "loss": 1.1952, "step": 110 }, { "grad_norm": 0.9454733729362488, "learning_rate": 2.975e-06, "loss": 1.171, "step": 120 }, { "grad_norm": 0.7125617861747742, "learning_rate": 3.225e-06, "loss": 1.13, "step": 130 }, { "grad_norm": 0.47815755009651184, "learning_rate": 3.4750000000000006e-06, "loss": 1.1135, "step": 140 }, { "grad_norm": 0.5524147748947144, "learning_rate": 3.725e-06, "loss": 1.0951, "step": 150 }, { "grad_norm": 0.467182457447052, "learning_rate": 3.975e-06, "loss": 1.0761, "step": 160 }, { "grad_norm": 0.5229169726371765, "learning_rate": 4.225e-06, "loss": 1.0745, "step": 170 }, { "grad_norm": 0.41773730516433716, "learning_rate": 4.475e-06, "loss": 1.0779, "step": 180 }, { "grad_norm": 0.5216078758239746, "learning_rate": 4.7250000000000005e-06, "loss": 1.0637, "step": 190 }, { "grad_norm": 0.6388832330703735, "learning_rate": 4.975000000000001e-06, "loss": 1.0566, "step": 200 }, { "grad_norm": 0.4831826090812683, "learning_rate": 5.225e-06, "loss": 1.0535, "step": 210 }, { "grad_norm": 0.5508063435554504, "learning_rate": 5.475e-06, "loss": 1.051, "step": 220 }, { "grad_norm": 0.4236525595188141, "learning_rate": 5.725e-06, "loss": 1.0414, "step": 230 }, { "grad_norm": 0.4898151755332947, "learning_rate": 5.975e-06, "loss": 1.0382, "step": 240 }, { "grad_norm": 0.4961107075214386, "learning_rate": 6.2250000000000005e-06, "loss": 1.034, "step": 250 }, { "grad_norm": 0.530572772026062, "learning_rate": 6.475000000000001e-06, "loss": 1.0403, "step": 260 }, { "grad_norm": 0.40762758255004883, "learning_rate": 6.725000000000001e-06, "loss": 1.0351, "step": 270 }, { "grad_norm": 0.5200778245925903, "learning_rate": 6.975000000000001e-06, "loss": 1.0332, "step": 280 }, { "grad_norm": 0.5414771437644958, "learning_rate": 7.2249999999999994e-06, "loss": 1.024, "step": 290 }, { "grad_norm": 0.5066478252410889, "learning_rate": 7.4750000000000004e-06, "loss": 1.0267, "step": 300 }, { "grad_norm": 0.6082722544670105, "learning_rate": 7.725e-06, "loss": 1.0252, "step": 310 }, { "grad_norm": 0.4625145196914673, "learning_rate": 7.975e-06, "loss": 1.0269, "step": 320 }, { "grad_norm": 0.5610113739967346, "learning_rate": 8.225e-06, "loss": 1.0143, "step": 330 }, { "grad_norm": 0.6787164211273193, "learning_rate": 8.475000000000001e-06, "loss": 0.9941, "step": 340 }, { "grad_norm": 1.4590646028518677, "learning_rate": 8.725e-06, "loss": 0.9898, "step": 350 }, { "grad_norm": 0.753097653388977, "learning_rate": 8.975e-06, "loss": 0.9615, "step": 360 }, { "grad_norm": 1.3400416374206543, "learning_rate": 9.225e-06, "loss": 0.9387, "step": 370 }, { "grad_norm": 1.0470538139343262, "learning_rate": 9.475e-06, "loss": 0.922, "step": 380 }, { "grad_norm": 1.4087488651275635, "learning_rate": 9.725000000000001e-06, "loss": 0.9037, "step": 390 }, { "grad_norm": 1.8154581785202026, "learning_rate": 9.975e-06, "loss": 0.8595, "step": 400 }, { "grad_norm": 1.5544401407241821, "learning_rate": 1.0225e-05, "loss": 0.8199, "step": 410 }, { "grad_norm": 1.8898271322250366, "learning_rate": 1.0475e-05, "loss": 0.7659, "step": 420 }, { "grad_norm": 1.726564645767212, "learning_rate": 1.0725e-05, "loss": 0.7376, "step": 430 }, { "grad_norm": 1.9561127424240112, "learning_rate": 1.0975e-05, "loss": 0.7104, "step": 440 }, { "grad_norm": 1.3225679397583008, "learning_rate": 1.1225e-05, "loss": 0.6915, "step": 450 }, { "grad_norm": 1.8089001178741455, "learning_rate": 1.1475000000000001e-05, "loss": 0.6547, "step": 460 }, { "grad_norm": 1.7803243398666382, "learning_rate": 1.1725e-05, "loss": 0.6306, "step": 470 }, { "grad_norm": 1.7278836965560913, "learning_rate": 1.1975e-05, "loss": 0.6048, "step": 480 }, { "grad_norm": 1.751409888267517, "learning_rate": 1.2225e-05, "loss": 0.5728, "step": 490 }, { "grad_norm": 2.1508283615112305, "learning_rate": 1.2475e-05, "loss": 0.5533, "step": 500 }, { "grad_norm": 1.9482898712158203, "learning_rate": 1.2725000000000001e-05, "loss": 0.5259, "step": 510 }, { "grad_norm": 1.735801339149475, "learning_rate": 1.2975e-05, "loss": 0.5073, "step": 520 }, { "grad_norm": 2.148799419403076, "learning_rate": 1.3225000000000001e-05, "loss": 0.484, "step": 530 }, { "grad_norm": 1.652020812034607, "learning_rate": 1.3475000000000002e-05, "loss": 0.4659, "step": 540 }, { "grad_norm": 2.0156631469726562, "learning_rate": 1.3725000000000002e-05, "loss": 0.4399, "step": 550 }, { "grad_norm": 1.9321014881134033, "learning_rate": 1.3975000000000003e-05, "loss": 0.4227, "step": 560 }, { "grad_norm": 1.890122413635254, "learning_rate": 1.4225e-05, "loss": 0.4065, "step": 570 }, { "grad_norm": 2.4143149852752686, "learning_rate": 1.4475e-05, "loss": 0.3847, "step": 580 }, { "grad_norm": 1.917590856552124, "learning_rate": 1.4725e-05, "loss": 0.3613, "step": 590 }, { "grad_norm": 1.980224370956421, "learning_rate": 1.4975e-05, "loss": 0.36, "step": 600 }, { "grad_norm": 1.8241479396820068, "learning_rate": 1.5225e-05, "loss": 0.3433, "step": 610 }, { "grad_norm": 1.7297978401184082, "learning_rate": 1.5475e-05, "loss": 0.342, "step": 620 }, { "grad_norm": 1.8881639242172241, "learning_rate": 1.5725e-05, "loss": 0.3309, "step": 630 }, { "grad_norm": 1.6923414468765259, "learning_rate": 1.5975000000000002e-05, "loss": 0.3188, "step": 640 }, { "grad_norm": 2.1035709381103516, "learning_rate": 1.6225e-05, "loss": 0.3199, "step": 650 }, { "grad_norm": 1.8885313272476196, "learning_rate": 1.6475e-05, "loss": 0.3116, "step": 660 }, { "grad_norm": 1.5232712030410767, "learning_rate": 1.6725000000000003e-05, "loss": 0.2904, "step": 670 }, { "grad_norm": 1.6049182415008545, "learning_rate": 1.6975000000000003e-05, "loss": 0.273, "step": 680 }, { "grad_norm": 2.248253345489502, "learning_rate": 1.7225e-05, "loss": 0.2547, "step": 690 }, { "grad_norm": 1.6934330463409424, "learning_rate": 1.7475e-05, "loss": 0.2353, "step": 700 }, { "grad_norm": 2.3303279876708984, "learning_rate": 1.7725e-05, "loss": 0.2313, "step": 710 }, { "grad_norm": 1.9882230758666992, "learning_rate": 1.7975e-05, "loss": 0.2124, "step": 720 }, { "grad_norm": 1.7424615621566772, "learning_rate": 1.8225e-05, "loss": 0.1891, "step": 730 }, { "grad_norm": 1.7147302627563477, "learning_rate": 1.8475000000000002e-05, "loss": 0.1994, "step": 740 }, { "grad_norm": 2.0619044303894043, "learning_rate": 1.8725e-05, "loss": 0.1986, "step": 750 }, { "grad_norm": 1.9220082759857178, "learning_rate": 1.8975e-05, "loss": 0.1941, "step": 760 }, { "grad_norm": 2.1906590461730957, "learning_rate": 1.9225e-05, "loss": 0.1805, "step": 770 }, { "grad_norm": 1.6909537315368652, "learning_rate": 1.9475000000000002e-05, "loss": 0.1669, "step": 780 }, { "grad_norm": 2.0480458736419678, "learning_rate": 1.9725000000000002e-05, "loss": 0.1523, "step": 790 }, { "grad_norm": 1.561774730682373, "learning_rate": 1.9975e-05, "loss": 0.147, "step": 800 }, { "grad_norm": 1.7473176717758179, "learning_rate": 2.0225000000000004e-05, "loss": 0.133, "step": 810 }, { "grad_norm": 1.7309926748275757, "learning_rate": 2.0475e-05, "loss": 0.1301, "step": 820 }, { "grad_norm": 1.4721465110778809, "learning_rate": 2.0725e-05, "loss": 0.1268, "step": 830 }, { "grad_norm": 1.7214860916137695, "learning_rate": 2.0975e-05, "loss": 0.1232, "step": 840 }, { "grad_norm": 2.0637240409851074, "learning_rate": 2.1225e-05, "loss": 0.1248, "step": 850 }, { "grad_norm": 2.212535858154297, "learning_rate": 2.1475e-05, "loss": 0.1311, "step": 860 }, { "grad_norm": 1.6657657623291016, "learning_rate": 2.1725e-05, "loss": 0.1174, "step": 870 }, { "grad_norm": 1.5616034269332886, "learning_rate": 2.1975000000000002e-05, "loss": 0.1247, "step": 880 }, { "grad_norm": 1.9177272319793701, "learning_rate": 2.2225e-05, "loss": 0.1113, "step": 890 }, { "grad_norm": 1.8991684913635254, "learning_rate": 2.2475e-05, "loss": 0.1061, "step": 900 }, { "grad_norm": 1.6902140378952026, "learning_rate": 2.2725000000000003e-05, "loss": 0.1076, "step": 910 }, { "grad_norm": 1.4998904466629028, "learning_rate": 2.2975000000000003e-05, "loss": 0.1054, "step": 920 }, { "grad_norm": 1.7178643941879272, "learning_rate": 2.3225000000000002e-05, "loss": 0.1056, "step": 930 }, { "grad_norm": 1.9046908617019653, "learning_rate": 2.3475e-05, "loss": 0.1018, "step": 940 }, { "grad_norm": 1.459069013595581, "learning_rate": 2.3725e-05, "loss": 0.1076, "step": 950 }, { "grad_norm": 1.5298603773117065, "learning_rate": 2.3975e-05, "loss": 0.1078, "step": 960 }, { "grad_norm": 1.7258670330047607, "learning_rate": 2.4225e-05, "loss": 0.106, "step": 970 }, { "grad_norm": 1.6092653274536133, "learning_rate": 2.4475000000000002e-05, "loss": 0.1046, "step": 980 }, { "grad_norm": 1.8294708728790283, "learning_rate": 2.4725e-05, "loss": 0.1053, "step": 990 }, { "grad_norm": 1.771057367324829, "learning_rate": 2.4975e-05, "loss": 0.1072, "step": 1000 }, { "grad_norm": 1.7862651348114014, "learning_rate": 2.5225e-05, "loss": 0.1002, "step": 1010 }, { "grad_norm": 1.3546617031097412, "learning_rate": 2.5475e-05, "loss": 0.1005, "step": 1020 }, { "grad_norm": 1.6929477453231812, "learning_rate": 2.5725e-05, "loss": 0.1051, "step": 1030 }, { "grad_norm": 1.5154497623443604, "learning_rate": 2.5974999999999998e-05, "loss": 0.0969, "step": 1040 }, { "grad_norm": 1.5828505754470825, "learning_rate": 2.6225e-05, "loss": 0.0909, "step": 1050 }, { "grad_norm": 1.1245936155319214, "learning_rate": 2.6475e-05, "loss": 0.0936, "step": 1060 }, { "grad_norm": 1.324134349822998, "learning_rate": 2.6725e-05, "loss": 0.0921, "step": 1070 }, { "grad_norm": 1.6023107767105103, "learning_rate": 2.6975000000000002e-05, "loss": 0.0864, "step": 1080 }, { "grad_norm": 1.6774718761444092, "learning_rate": 2.7225e-05, "loss": 0.0891, "step": 1090 }, { "grad_norm": 1.6501003503799438, "learning_rate": 2.7475e-05, "loss": 0.0996, "step": 1100 }, { "grad_norm": 1.5846939086914062, "learning_rate": 2.7725e-05, "loss": 0.0833, "step": 1110 }, { "grad_norm": 1.191044569015503, "learning_rate": 2.7975000000000002e-05, "loss": 0.0948, "step": 1120 }, { "grad_norm": 1.3026844263076782, "learning_rate": 2.8225e-05, "loss": 0.0938, "step": 1130 }, { "grad_norm": 1.8582338094711304, "learning_rate": 2.8475e-05, "loss": 0.0906, "step": 1140 }, { "grad_norm": 1.7908824682235718, "learning_rate": 2.8725e-05, "loss": 0.0955, "step": 1150 }, { "grad_norm": 1.6283639669418335, "learning_rate": 2.8975000000000003e-05, "loss": 0.0896, "step": 1160 }, { "grad_norm": 1.2708567380905151, "learning_rate": 2.9225000000000002e-05, "loss": 0.0882, "step": 1170 }, { "grad_norm": 1.616062045097351, "learning_rate": 2.9475e-05, "loss": 0.0897, "step": 1180 }, { "grad_norm": 1.1153043508529663, "learning_rate": 2.9725000000000004e-05, "loss": 0.0836, "step": 1190 }, { "grad_norm": 1.6311445236206055, "learning_rate": 2.9975000000000004e-05, "loss": 0.0912, "step": 1200 }, { "grad_norm": 1.0658538341522217, "learning_rate": 3.0225000000000003e-05, "loss": 0.086, "step": 1210 }, { "grad_norm": 1.196419358253479, "learning_rate": 3.0475000000000002e-05, "loss": 0.0785, "step": 1220 }, { "grad_norm": 1.32984459400177, "learning_rate": 3.0725e-05, "loss": 0.0833, "step": 1230 }, { "grad_norm": 1.2014424800872803, "learning_rate": 3.0975e-05, "loss": 0.0872, "step": 1240 }, { "grad_norm": 1.2646026611328125, "learning_rate": 3.122500000000001e-05, "loss": 0.0852, "step": 1250 }, { "grad_norm": 1.4651663303375244, "learning_rate": 3.1475e-05, "loss": 0.075, "step": 1260 }, { "grad_norm": 1.3455827236175537, "learning_rate": 3.1725e-05, "loss": 0.0918, "step": 1270 }, { "grad_norm": 1.5699114799499512, "learning_rate": 3.1975e-05, "loss": 0.084, "step": 1280 }, { "grad_norm": 1.3420729637145996, "learning_rate": 3.2225e-05, "loss": 0.0876, "step": 1290 }, { "grad_norm": 1.674861192703247, "learning_rate": 3.2474999999999997e-05, "loss": 0.0838, "step": 1300 }, { "grad_norm": 1.3997465372085571, "learning_rate": 3.2725e-05, "loss": 0.0798, "step": 1310 }, { "grad_norm": 1.405192494392395, "learning_rate": 3.2975e-05, "loss": 0.0938, "step": 1320 }, { "grad_norm": 1.1720155477523804, "learning_rate": 3.3225e-05, "loss": 0.0773, "step": 1330 }, { "grad_norm": 1.2635575532913208, "learning_rate": 3.3475e-05, "loss": 0.0697, "step": 1340 }, { "grad_norm": 1.2650095224380493, "learning_rate": 3.3725e-05, "loss": 0.0791, "step": 1350 }, { "grad_norm": 1.062660574913025, "learning_rate": 3.3975e-05, "loss": 0.0859, "step": 1360 }, { "grad_norm": 1.1777929067611694, "learning_rate": 3.4225e-05, "loss": 0.0749, "step": 1370 }, { "grad_norm": 1.1609066724777222, "learning_rate": 3.4475000000000005e-05, "loss": 0.0737, "step": 1380 }, { "grad_norm": 1.085674524307251, "learning_rate": 3.4725000000000004e-05, "loss": 0.0685, "step": 1390 }, { "grad_norm": 1.014859914779663, "learning_rate": 3.4975e-05, "loss": 0.0716, "step": 1400 }, { "grad_norm": 1.6402721405029297, "learning_rate": 3.5225e-05, "loss": 0.0704, "step": 1410 }, { "grad_norm": 1.2711130380630493, "learning_rate": 3.5475e-05, "loss": 0.0741, "step": 1420 }, { "grad_norm": 1.3303911685943604, "learning_rate": 3.5725e-05, "loss": 0.0855, "step": 1430 }, { "grad_norm": 1.6919200420379639, "learning_rate": 3.5975e-05, "loss": 0.0731, "step": 1440 }, { "grad_norm": 1.3457597494125366, "learning_rate": 3.6225000000000006e-05, "loss": 0.0689, "step": 1450 }, { "grad_norm": 1.0700994729995728, "learning_rate": 3.6475000000000006e-05, "loss": 0.0767, "step": 1460 }, { "grad_norm": 1.0146839618682861, "learning_rate": 3.6725000000000005e-05, "loss": 0.068, "step": 1470 }, { "grad_norm": 0.9740791916847229, "learning_rate": 3.6975000000000004e-05, "loss": 0.0657, "step": 1480 }, { "grad_norm": 1.0838881731033325, "learning_rate": 3.7225000000000004e-05, "loss": 0.0742, "step": 1490 }, { "grad_norm": 1.0132752656936646, "learning_rate": 3.7475e-05, "loss": 0.07, "step": 1500 }, { "grad_norm": 1.1348044872283936, "learning_rate": 3.7725e-05, "loss": 0.0749, "step": 1510 }, { "grad_norm": 1.309169888496399, "learning_rate": 3.7975e-05, "loss": 0.0731, "step": 1520 }, { "grad_norm": 1.083322525024414, "learning_rate": 3.8225e-05, "loss": 0.0702, "step": 1530 }, { "grad_norm": 1.2023237943649292, "learning_rate": 3.8475e-05, "loss": 0.0733, "step": 1540 }, { "grad_norm": 1.3348912000656128, "learning_rate": 3.8725e-05, "loss": 0.0692, "step": 1550 }, { "grad_norm": 0.858433187007904, "learning_rate": 3.8975e-05, "loss": 0.071, "step": 1560 }, { "grad_norm": 1.340848684310913, "learning_rate": 3.9225e-05, "loss": 0.0732, "step": 1570 }, { "grad_norm": 1.304295539855957, "learning_rate": 3.9475000000000004e-05, "loss": 0.0647, "step": 1580 }, { "grad_norm": 1.2166643142700195, "learning_rate": 3.9725e-05, "loss": 0.0652, "step": 1590 }, { "grad_norm": 1.0720939636230469, "learning_rate": 3.9975e-05, "loss": 0.0751, "step": 1600 }, { "grad_norm": 0.9998016357421875, "learning_rate": 4.0225e-05, "loss": 0.0669, "step": 1610 }, { "grad_norm": 1.3904045820236206, "learning_rate": 4.0475e-05, "loss": 0.0659, "step": 1620 }, { "grad_norm": 1.0652220249176025, "learning_rate": 4.0725e-05, "loss": 0.0694, "step": 1630 }, { "grad_norm": 1.1695936918258667, "learning_rate": 4.0975e-05, "loss": 0.0708, "step": 1640 }, { "grad_norm": 0.9627931714057922, "learning_rate": 4.1225e-05, "loss": 0.066, "step": 1650 }, { "grad_norm": 0.6591870188713074, "learning_rate": 4.1475000000000005e-05, "loss": 0.0626, "step": 1660 }, { "grad_norm": 1.1707704067230225, "learning_rate": 4.1725000000000005e-05, "loss": 0.069, "step": 1670 }, { "grad_norm": 0.8848738670349121, "learning_rate": 4.1975000000000004e-05, "loss": 0.0655, "step": 1680 }, { "grad_norm": 0.9593561291694641, "learning_rate": 4.2225e-05, "loss": 0.0628, "step": 1690 }, { "grad_norm": 1.2162808179855347, "learning_rate": 4.2475e-05, "loss": 0.0743, "step": 1700 }, { "grad_norm": 0.8655327558517456, "learning_rate": 4.2725e-05, "loss": 0.0678, "step": 1710 }, { "grad_norm": 1.1966687440872192, "learning_rate": 4.2975e-05, "loss": 0.0684, "step": 1720 }, { "grad_norm": 1.2192535400390625, "learning_rate": 4.322500000000001e-05, "loss": 0.0679, "step": 1730 }, { "grad_norm": 1.0060296058654785, "learning_rate": 4.3475000000000006e-05, "loss": 0.0644, "step": 1740 }, { "grad_norm": 1.2788019180297852, "learning_rate": 4.3725000000000006e-05, "loss": 0.0617, "step": 1750 }, { "grad_norm": 0.9235604405403137, "learning_rate": 4.3975e-05, "loss": 0.0586, "step": 1760 }, { "grad_norm": 0.7610540986061096, "learning_rate": 4.4225e-05, "loss": 0.0732, "step": 1770 }, { "grad_norm": 0.771186113357544, "learning_rate": 4.4475e-05, "loss": 0.0672, "step": 1780 }, { "grad_norm": 0.7998961210250854, "learning_rate": 4.4725e-05, "loss": 0.0644, "step": 1790 }, { "grad_norm": 0.9221842288970947, "learning_rate": 4.4975e-05, "loss": 0.0689, "step": 1800 }, { "grad_norm": 0.8497925996780396, "learning_rate": 4.5225e-05, "loss": 0.0637, "step": 1810 }, { "grad_norm": 0.969850480556488, "learning_rate": 4.5475e-05, "loss": 0.0638, "step": 1820 }, { "grad_norm": 1.0014839172363281, "learning_rate": 4.5725e-05, "loss": 0.0607, "step": 1830 }, { "grad_norm": 1.003509283065796, "learning_rate": 4.5975e-05, "loss": 0.0601, "step": 1840 }, { "grad_norm": 1.0910965204238892, "learning_rate": 4.6225e-05, "loss": 0.0607, "step": 1850 }, { "grad_norm": 1.0293738842010498, "learning_rate": 4.6475000000000005e-05, "loss": 0.0599, "step": 1860 }, { "grad_norm": 0.9137545824050903, "learning_rate": 4.6725000000000004e-05, "loss": 0.0616, "step": 1870 }, { "grad_norm": 0.8558566570281982, "learning_rate": 4.6975000000000003e-05, "loss": 0.0581, "step": 1880 }, { "grad_norm": 0.732243537902832, "learning_rate": 4.7225e-05, "loss": 0.0563, "step": 1890 }, { "grad_norm": 0.9701789021492004, "learning_rate": 4.7475e-05, "loss": 0.0603, "step": 1900 }, { "grad_norm": 0.8463003039360046, "learning_rate": 4.7725e-05, "loss": 0.0615, "step": 1910 }, { "grad_norm": 0.6701717972755432, "learning_rate": 4.7975e-05, "loss": 0.0525, "step": 1920 }, { "grad_norm": 1.0384646654129028, "learning_rate": 4.822500000000001e-05, "loss": 0.0575, "step": 1930 }, { "grad_norm": 0.8709242939949036, "learning_rate": 4.8475000000000006e-05, "loss": 0.07, "step": 1940 }, { "grad_norm": 0.8229147791862488, "learning_rate": 4.8725000000000005e-05, "loss": 0.058, "step": 1950 }, { "grad_norm": 0.8545090556144714, "learning_rate": 4.8975000000000005e-05, "loss": 0.0559, "step": 1960 }, { "grad_norm": 0.8478420972824097, "learning_rate": 4.9225000000000004e-05, "loss": 0.059, "step": 1970 }, { "grad_norm": 0.9126168489456177, "learning_rate": 4.9475e-05, "loss": 0.0633, "step": 1980 }, { "grad_norm": 0.7752177715301514, "learning_rate": 4.9725e-05, "loss": 0.0564, "step": 1990 }, { "grad_norm": 0.9906421303749084, "learning_rate": 4.9975e-05, "loss": 0.0568, "step": 2000 }, { "grad_norm": 1.0784865617752075, "learning_rate": 5.0225e-05, "loss": 0.0592, "step": 2010 }, { "grad_norm": 0.9180090427398682, "learning_rate": 5.047500000000001e-05, "loss": 0.0628, "step": 2020 }, { "grad_norm": 0.6478615403175354, "learning_rate": 5.0725e-05, "loss": 0.0554, "step": 2030 }, { "grad_norm": 0.7112268209457397, "learning_rate": 5.0975000000000006e-05, "loss": 0.06, "step": 2040 }, { "grad_norm": 1.2856839895248413, "learning_rate": 5.1225e-05, "loss": 0.0632, "step": 2050 }, { "grad_norm": 1.0430432558059692, "learning_rate": 5.1475000000000004e-05, "loss": 0.0597, "step": 2060 }, { "grad_norm": 1.1008764505386353, "learning_rate": 5.1725000000000004e-05, "loss": 0.0576, "step": 2070 }, { "grad_norm": 1.0061269998550415, "learning_rate": 5.197500000000001e-05, "loss": 0.0605, "step": 2080 }, { "grad_norm": 0.733997642993927, "learning_rate": 5.2225e-05, "loss": 0.062, "step": 2090 }, { "grad_norm": 1.0081887245178223, "learning_rate": 5.247500000000001e-05, "loss": 0.0568, "step": 2100 }, { "grad_norm": 1.0130735635757446, "learning_rate": 5.2725e-05, "loss": 0.0573, "step": 2110 }, { "grad_norm": 0.6643454432487488, "learning_rate": 5.297500000000001e-05, "loss": 0.0544, "step": 2120 }, { "grad_norm": 0.7024112939834595, "learning_rate": 5.3225e-05, "loss": 0.0586, "step": 2130 }, { "grad_norm": 0.7366859316825867, "learning_rate": 5.3475e-05, "loss": 0.0556, "step": 2140 }, { "grad_norm": 0.7533279657363892, "learning_rate": 5.3725000000000005e-05, "loss": 0.0574, "step": 2150 }, { "grad_norm": 0.6134442687034607, "learning_rate": 5.3975e-05, "loss": 0.0469, "step": 2160 }, { "grad_norm": 0.9361967444419861, "learning_rate": 5.4225000000000003e-05, "loss": 0.0559, "step": 2170 }, { "grad_norm": 0.6418201923370361, "learning_rate": 5.4474999999999996e-05, "loss": 0.055, "step": 2180 }, { "grad_norm": 0.7472584247589111, "learning_rate": 5.4725e-05, "loss": 0.0534, "step": 2190 }, { "grad_norm": 0.8877463936805725, "learning_rate": 5.4975e-05, "loss": 0.0501, "step": 2200 }, { "grad_norm": 0.9642484784126282, "learning_rate": 5.522500000000001e-05, "loss": 0.0508, "step": 2210 }, { "grad_norm": 0.8540062308311462, "learning_rate": 5.5475e-05, "loss": 0.0523, "step": 2220 }, { "grad_norm": 0.7555636763572693, "learning_rate": 5.5725000000000006e-05, "loss": 0.0517, "step": 2230 }, { "grad_norm": 0.67600417137146, "learning_rate": 5.5975e-05, "loss": 0.0553, "step": 2240 }, { "grad_norm": 0.9391599297523499, "learning_rate": 5.6225000000000005e-05, "loss": 0.0639, "step": 2250 }, { "grad_norm": 0.6468883156776428, "learning_rate": 5.6475e-05, "loss": 0.065, "step": 2260 }, { "grad_norm": 0.736388087272644, "learning_rate": 5.6725e-05, "loss": 0.0637, "step": 2270 }, { "grad_norm": 0.7386782169342041, "learning_rate": 5.6975e-05, "loss": 0.0563, "step": 2280 }, { "grad_norm": 0.8396263122558594, "learning_rate": 5.722500000000001e-05, "loss": 0.0486, "step": 2290 }, { "grad_norm": 0.7284508943557739, "learning_rate": 5.7475e-05, "loss": 0.0557, "step": 2300 }, { "grad_norm": 0.6834036707878113, "learning_rate": 5.772500000000001e-05, "loss": 0.0546, "step": 2310 }, { "grad_norm": 0.5818115472793579, "learning_rate": 5.7975e-05, "loss": 0.0532, "step": 2320 }, { "grad_norm": 0.7555652856826782, "learning_rate": 5.8225000000000006e-05, "loss": 0.0601, "step": 2330 }, { "grad_norm": 0.7249618172645569, "learning_rate": 5.8475000000000005e-05, "loss": 0.0553, "step": 2340 }, { "grad_norm": 1.051877498626709, "learning_rate": 5.8725000000000004e-05, "loss": 0.0522, "step": 2350 }, { "grad_norm": 0.8893764019012451, "learning_rate": 5.8975000000000004e-05, "loss": 0.0584, "step": 2360 }, { "grad_norm": 0.5572100281715393, "learning_rate": 5.922500000000001e-05, "loss": 0.0566, "step": 2370 }, { "grad_norm": 0.8577748537063599, "learning_rate": 5.9475e-05, "loss": 0.0553, "step": 2380 }, { "grad_norm": 0.7612488865852356, "learning_rate": 5.9724999999999995e-05, "loss": 0.0556, "step": 2390 }, { "grad_norm": 0.7745786309242249, "learning_rate": 5.9975e-05, "loss": 0.0553, "step": 2400 }, { "grad_norm": 0.6258660554885864, "learning_rate": 6.0225e-05, "loss": 0.051, "step": 2410 }, { "grad_norm": 0.5380517840385437, "learning_rate": 6.0475000000000006e-05, "loss": 0.0528, "step": 2420 }, { "grad_norm": 0.8730436563491821, "learning_rate": 6.0725e-05, "loss": 0.0534, "step": 2430 }, { "grad_norm": 0.7553834319114685, "learning_rate": 6.0975000000000005e-05, "loss": 0.0505, "step": 2440 }, { "grad_norm": 0.6756471991539001, "learning_rate": 6.1225e-05, "loss": 0.0519, "step": 2450 }, { "grad_norm": 0.7694554924964905, "learning_rate": 6.1475e-05, "loss": 0.0579, "step": 2460 }, { "grad_norm": 0.9069452285766602, "learning_rate": 6.1725e-05, "loss": 0.0481, "step": 2470 }, { "grad_norm": 0.6306765079498291, "learning_rate": 6.1975e-05, "loss": 0.0575, "step": 2480 }, { "grad_norm": 0.7670426368713379, "learning_rate": 6.2225e-05, "loss": 0.0512, "step": 2490 }, { "grad_norm": 0.7755621671676636, "learning_rate": 6.2475e-05, "loss": 0.0526, "step": 2500 }, { "grad_norm": 0.8366811871528625, "learning_rate": 6.2725e-05, "loss": 0.0538, "step": 2510 }, { "grad_norm": 0.6688365340232849, "learning_rate": 6.297500000000001e-05, "loss": 0.0514, "step": 2520 }, { "grad_norm": 0.7997210025787354, "learning_rate": 6.3225e-05, "loss": 0.0487, "step": 2530 }, { "grad_norm": 0.7413893342018127, "learning_rate": 6.347500000000001e-05, "loss": 0.049, "step": 2540 }, { "grad_norm": 0.740452229976654, "learning_rate": 6.3725e-05, "loss": 0.0568, "step": 2550 }, { "grad_norm": 0.6515949964523315, "learning_rate": 6.397500000000001e-05, "loss": 0.0493, "step": 2560 }, { "grad_norm": 0.6464294791221619, "learning_rate": 6.4225e-05, "loss": 0.049, "step": 2570 }, { "grad_norm": 0.7380954027175903, "learning_rate": 6.447500000000001e-05, "loss": 0.0538, "step": 2580 }, { "grad_norm": 0.7701060771942139, "learning_rate": 6.4725e-05, "loss": 0.0476, "step": 2590 }, { "grad_norm": 0.869131326675415, "learning_rate": 6.497500000000001e-05, "loss": 0.054, "step": 2600 }, { "grad_norm": 0.8203336000442505, "learning_rate": 6.5225e-05, "loss": 0.052, "step": 2610 }, { "grad_norm": 0.5380281209945679, "learning_rate": 6.5475e-05, "loss": 0.0502, "step": 2620 }, { "grad_norm": 0.8151743412017822, "learning_rate": 6.5725e-05, "loss": 0.0561, "step": 2630 }, { "grad_norm": 0.718498945236206, "learning_rate": 6.5975e-05, "loss": 0.0506, "step": 2640 }, { "grad_norm": 0.621749222278595, "learning_rate": 6.6225e-05, "loss": 0.0543, "step": 2650 }, { "grad_norm": 0.7316995859146118, "learning_rate": 6.6475e-05, "loss": 0.0516, "step": 2660 }, { "grad_norm": 0.6590881943702698, "learning_rate": 6.672500000000001e-05, "loss": 0.0536, "step": 2670 }, { "grad_norm": 0.6422820687294006, "learning_rate": 6.6975e-05, "loss": 0.0484, "step": 2680 }, { "grad_norm": 0.7314589023590088, "learning_rate": 6.722500000000001e-05, "loss": 0.0531, "step": 2690 }, { "grad_norm": 0.8417461514472961, "learning_rate": 6.7475e-05, "loss": 0.0509, "step": 2700 }, { "grad_norm": 0.8195475935935974, "learning_rate": 6.7725e-05, "loss": 0.0557, "step": 2710 }, { "grad_norm": 0.5790806412696838, "learning_rate": 6.7975e-05, "loss": 0.0492, "step": 2720 }, { "grad_norm": 0.6212972402572632, "learning_rate": 6.8225e-05, "loss": 0.053, "step": 2730 }, { "grad_norm": 0.6415103673934937, "learning_rate": 6.8475e-05, "loss": 0.0524, "step": 2740 }, { "grad_norm": 0.681734025478363, "learning_rate": 6.8725e-05, "loss": 0.0489, "step": 2750 }, { "grad_norm": 0.6728873252868652, "learning_rate": 6.8975e-05, "loss": 0.0522, "step": 2760 }, { "grad_norm": 0.5486435294151306, "learning_rate": 6.9225e-05, "loss": 0.0448, "step": 2770 }, { "grad_norm": 0.6172059178352356, "learning_rate": 6.9475e-05, "loss": 0.0523, "step": 2780 }, { "grad_norm": 0.514598548412323, "learning_rate": 6.9725e-05, "loss": 0.0514, "step": 2790 }, { "grad_norm": 0.7059509754180908, "learning_rate": 6.997500000000001e-05, "loss": 0.0469, "step": 2800 }, { "grad_norm": 0.7806535959243774, "learning_rate": 7.022500000000001e-05, "loss": 0.0476, "step": 2810 }, { "grad_norm": 0.6727353930473328, "learning_rate": 7.0475e-05, "loss": 0.044, "step": 2820 }, { "grad_norm": 0.5153267979621887, "learning_rate": 7.072500000000001e-05, "loss": 0.0527, "step": 2830 }, { "grad_norm": 0.3580329418182373, "learning_rate": 7.0975e-05, "loss": 0.0509, "step": 2840 }, { "grad_norm": 0.535731852054596, "learning_rate": 7.122500000000001e-05, "loss": 0.0455, "step": 2850 }, { "grad_norm": 0.6909304261207581, "learning_rate": 7.1475e-05, "loss": 0.0468, "step": 2860 }, { "grad_norm": 0.5866256952285767, "learning_rate": 7.172500000000001e-05, "loss": 0.0486, "step": 2870 }, { "grad_norm": 0.5585393905639648, "learning_rate": 7.1975e-05, "loss": 0.0458, "step": 2880 }, { "grad_norm": 0.5972147583961487, "learning_rate": 7.2225e-05, "loss": 0.042, "step": 2890 }, { "grad_norm": 0.8439204096794128, "learning_rate": 7.2475e-05, "loss": 0.0468, "step": 2900 }, { "grad_norm": 0.553803563117981, "learning_rate": 7.272499999999999e-05, "loss": 0.0474, "step": 2910 }, { "grad_norm": 0.6671627759933472, "learning_rate": 7.2975e-05, "loss": 0.0432, "step": 2920 }, { "grad_norm": 0.6662748456001282, "learning_rate": 7.3225e-05, "loss": 0.0499, "step": 2930 }, { "grad_norm": 0.5780026912689209, "learning_rate": 7.347500000000001e-05, "loss": 0.054, "step": 2940 }, { "grad_norm": 0.5941490530967712, "learning_rate": 7.3725e-05, "loss": 0.0518, "step": 2950 }, { "grad_norm": 0.6304528713226318, "learning_rate": 7.397500000000001e-05, "loss": 0.0554, "step": 2960 }, { "grad_norm": 0.6537070870399475, "learning_rate": 7.4225e-05, "loss": 0.0476, "step": 2970 }, { "grad_norm": 0.6715964078903198, "learning_rate": 7.447500000000001e-05, "loss": 0.0479, "step": 2980 }, { "grad_norm": 0.3930942416191101, "learning_rate": 7.4725e-05, "loss": 0.0435, "step": 2990 }, { "grad_norm": 0.44956305623054504, "learning_rate": 7.4975e-05, "loss": 0.0471, "step": 3000 }, { "grad_norm": 0.515681803226471, "learning_rate": 7.5225e-05, "loss": 0.052, "step": 3010 }, { "grad_norm": 0.6683989763259888, "learning_rate": 7.5475e-05, "loss": 0.0472, "step": 3020 }, { "grad_norm": 0.5572440028190613, "learning_rate": 7.5725e-05, "loss": 0.0434, "step": 3030 }, { "grad_norm": 0.7554171681404114, "learning_rate": 7.5975e-05, "loss": 0.0437, "step": 3040 }, { "grad_norm": 0.7334225177764893, "learning_rate": 7.6225e-05, "loss": 0.0471, "step": 3050 }, { "grad_norm": 0.5323683619499207, "learning_rate": 7.6475e-05, "loss": 0.0495, "step": 3060 }, { "grad_norm": 0.6151717305183411, "learning_rate": 7.672500000000001e-05, "loss": 0.0497, "step": 3070 }, { "grad_norm": 0.5227813720703125, "learning_rate": 7.697500000000001e-05, "loss": 0.0469, "step": 3080 }, { "grad_norm": 0.5635611414909363, "learning_rate": 7.722500000000001e-05, "loss": 0.0468, "step": 3090 }, { "grad_norm": 0.6739113926887512, "learning_rate": 7.747500000000001e-05, "loss": 0.0503, "step": 3100 }, { "grad_norm": 0.5643975734710693, "learning_rate": 7.7725e-05, "loss": 0.0531, "step": 3110 }, { "grad_norm": 0.7148908376693726, "learning_rate": 7.797500000000001e-05, "loss": 0.0433, "step": 3120 }, { "grad_norm": 0.6304246187210083, "learning_rate": 7.8225e-05, "loss": 0.0548, "step": 3130 }, { "grad_norm": 0.6709405183792114, "learning_rate": 7.8475e-05, "loss": 0.0438, "step": 3140 }, { "grad_norm": 0.5538259744644165, "learning_rate": 7.8725e-05, "loss": 0.047, "step": 3150 }, { "grad_norm": 0.5186079740524292, "learning_rate": 7.8975e-05, "loss": 0.0397, "step": 3160 }, { "grad_norm": 0.5775082111358643, "learning_rate": 7.9225e-05, "loss": 0.0453, "step": 3170 }, { "grad_norm": 0.572026789188385, "learning_rate": 7.9475e-05, "loss": 0.0472, "step": 3180 }, { "grad_norm": 0.4826521575450897, "learning_rate": 7.9725e-05, "loss": 0.0489, "step": 3190 }, { "grad_norm": 0.6543978452682495, "learning_rate": 7.9975e-05, "loss": 0.0449, "step": 3200 }, { "grad_norm": 0.4963151216506958, "learning_rate": 8.022500000000001e-05, "loss": 0.0422, "step": 3210 }, { "grad_norm": 0.6282826066017151, "learning_rate": 8.0475e-05, "loss": 0.0476, "step": 3220 }, { "grad_norm": 0.5205478668212891, "learning_rate": 8.072500000000001e-05, "loss": 0.0535, "step": 3230 }, { "grad_norm": 0.8019874691963196, "learning_rate": 8.0975e-05, "loss": 0.0482, "step": 3240 }, { "grad_norm": 0.6296453475952148, "learning_rate": 8.122500000000001e-05, "loss": 0.0481, "step": 3250 }, { "grad_norm": 0.6741389036178589, "learning_rate": 8.1475e-05, "loss": 0.0521, "step": 3260 }, { "grad_norm": 0.6566572189331055, "learning_rate": 8.172500000000001e-05, "loss": 0.0499, "step": 3270 }, { "grad_norm": 0.4287855625152588, "learning_rate": 8.1975e-05, "loss": 0.0438, "step": 3280 }, { "grad_norm": 0.5310896635055542, "learning_rate": 8.2225e-05, "loss": 0.0467, "step": 3290 }, { "grad_norm": 0.5373318195343018, "learning_rate": 8.2475e-05, "loss": 0.0471, "step": 3300 }, { "grad_norm": 0.5656524896621704, "learning_rate": 8.2725e-05, "loss": 0.0476, "step": 3310 }, { "grad_norm": 0.5995365977287292, "learning_rate": 8.2975e-05, "loss": 0.0506, "step": 3320 }, { "grad_norm": 0.499523788690567, "learning_rate": 8.3225e-05, "loss": 0.0451, "step": 3330 }, { "grad_norm": 0.5452511310577393, "learning_rate": 8.347500000000001e-05, "loss": 0.0456, "step": 3340 }, { "grad_norm": 0.5301199555397034, "learning_rate": 8.3725e-05, "loss": 0.0464, "step": 3350 }, { "grad_norm": 0.4322058856487274, "learning_rate": 8.397500000000001e-05, "loss": 0.0424, "step": 3360 }, { "grad_norm": 0.460086464881897, "learning_rate": 8.422500000000001e-05, "loss": 0.044, "step": 3370 }, { "grad_norm": 0.4871997535228729, "learning_rate": 8.447500000000001e-05, "loss": 0.0413, "step": 3380 }, { "grad_norm": 0.49588730931282043, "learning_rate": 8.4725e-05, "loss": 0.0454, "step": 3390 }, { "grad_norm": 0.6030573844909668, "learning_rate": 8.4975e-05, "loss": 0.0466, "step": 3400 }, { "grad_norm": 0.6114158630371094, "learning_rate": 8.5225e-05, "loss": 0.0491, "step": 3410 }, { "grad_norm": 0.5132580399513245, "learning_rate": 8.5475e-05, "loss": 0.0437, "step": 3420 }, { "grad_norm": 0.4820050299167633, "learning_rate": 8.5725e-05, "loss": 0.0414, "step": 3430 }, { "grad_norm": 0.6967134475708008, "learning_rate": 8.5975e-05, "loss": 0.0479, "step": 3440 }, { "grad_norm": 0.5232267379760742, "learning_rate": 8.6225e-05, "loss": 0.0409, "step": 3450 }, { "grad_norm": 0.4993670582771301, "learning_rate": 8.6475e-05, "loss": 0.0436, "step": 3460 }, { "grad_norm": 0.5206721425056458, "learning_rate": 8.672500000000001e-05, "loss": 0.0482, "step": 3470 }, { "grad_norm": 0.6051874756813049, "learning_rate": 8.6975e-05, "loss": 0.0439, "step": 3480 }, { "grad_norm": 0.4291444420814514, "learning_rate": 8.7225e-05, "loss": 0.043, "step": 3490 }, { "grad_norm": 0.5004898309707642, "learning_rate": 8.747500000000001e-05, "loss": 0.0396, "step": 3500 }, { "grad_norm": 0.45404934883117676, "learning_rate": 8.7725e-05, "loss": 0.0476, "step": 3510 }, { "grad_norm": 0.5612279772758484, "learning_rate": 8.797500000000001e-05, "loss": 0.0431, "step": 3520 }, { "grad_norm": 0.5149720907211304, "learning_rate": 8.8225e-05, "loss": 0.0453, "step": 3530 }, { "grad_norm": 0.5399808287620544, "learning_rate": 8.847500000000001e-05, "loss": 0.0458, "step": 3540 }, { "grad_norm": 0.6093869805335999, "learning_rate": 8.8725e-05, "loss": 0.0448, "step": 3550 }, { "grad_norm": 0.620223343372345, "learning_rate": 8.897500000000001e-05, "loss": 0.0456, "step": 3560 }, { "grad_norm": 0.4663323163986206, "learning_rate": 8.9225e-05, "loss": 0.0461, "step": 3570 }, { "grad_norm": 0.5784809589385986, "learning_rate": 8.9475e-05, "loss": 0.047, "step": 3580 }, { "grad_norm": 0.4460861384868622, "learning_rate": 8.9725e-05, "loss": 0.042, "step": 3590 }, { "grad_norm": 0.5072846412658691, "learning_rate": 8.9975e-05, "loss": 0.0445, "step": 3600 }, { "grad_norm": 0.5040016770362854, "learning_rate": 9.0225e-05, "loss": 0.0422, "step": 3610 }, { "grad_norm": 0.6363912224769592, "learning_rate": 9.0475e-05, "loss": 0.0418, "step": 3620 }, { "grad_norm": 0.4950333833694458, "learning_rate": 9.072500000000001e-05, "loss": 0.0462, "step": 3630 }, { "grad_norm": 0.49438419938087463, "learning_rate": 9.0975e-05, "loss": 0.0417, "step": 3640 }, { "grad_norm": 0.3908850848674774, "learning_rate": 9.122500000000001e-05, "loss": 0.0391, "step": 3650 }, { "grad_norm": 0.5062656402587891, "learning_rate": 9.1475e-05, "loss": 0.0399, "step": 3660 }, { "grad_norm": 0.6507706642150879, "learning_rate": 9.172500000000001e-05, "loss": 0.0395, "step": 3670 }, { "grad_norm": 0.5512116551399231, "learning_rate": 9.1975e-05, "loss": 0.0398, "step": 3680 }, { "grad_norm": 0.4868468642234802, "learning_rate": 9.2225e-05, "loss": 0.0497, "step": 3690 }, { "grad_norm": 0.5283458232879639, "learning_rate": 9.2475e-05, "loss": 0.0393, "step": 3700 }, { "grad_norm": 0.5997909903526306, "learning_rate": 9.2725e-05, "loss": 0.04, "step": 3710 }, { "grad_norm": 0.4041551947593689, "learning_rate": 9.2975e-05, "loss": 0.0415, "step": 3720 }, { "grad_norm": 0.5316563844680786, "learning_rate": 9.3225e-05, "loss": 0.0444, "step": 3730 }, { "grad_norm": 0.5076853632926941, "learning_rate": 9.3475e-05, "loss": 0.0479, "step": 3740 }, { "grad_norm": 0.6213400363922119, "learning_rate": 9.3725e-05, "loss": 0.0448, "step": 3750 }, { "grad_norm": 0.5178065896034241, "learning_rate": 9.397500000000001e-05, "loss": 0.0454, "step": 3760 }, { "grad_norm": 0.5711163878440857, "learning_rate": 9.422500000000001e-05, "loss": 0.0366, "step": 3770 }, { "grad_norm": 0.6347408294677734, "learning_rate": 9.4475e-05, "loss": 0.0445, "step": 3780 }, { "grad_norm": 0.5168583393096924, "learning_rate": 9.472500000000001e-05, "loss": 0.0407, "step": 3790 }, { "grad_norm": 0.4469931125640869, "learning_rate": 9.4975e-05, "loss": 0.0446, "step": 3800 }, { "grad_norm": 0.5789520144462585, "learning_rate": 9.522500000000001e-05, "loss": 0.0478, "step": 3810 }, { "grad_norm": 0.4918958842754364, "learning_rate": 9.5475e-05, "loss": 0.0501, "step": 3820 }, { "grad_norm": 0.3995557427406311, "learning_rate": 9.572500000000001e-05, "loss": 0.0412, "step": 3830 }, { "grad_norm": 0.4834426939487457, "learning_rate": 9.5975e-05, "loss": 0.0429, "step": 3840 }, { "grad_norm": 0.5953786373138428, "learning_rate": 9.622500000000001e-05, "loss": 0.0449, "step": 3850 }, { "grad_norm": 0.40576398372650146, "learning_rate": 9.6475e-05, "loss": 0.0441, "step": 3860 }, { "grad_norm": 0.5018937587738037, "learning_rate": 9.6725e-05, "loss": 0.0395, "step": 3870 }, { "grad_norm": 0.47828808426856995, "learning_rate": 9.6975e-05, "loss": 0.0404, "step": 3880 }, { "grad_norm": 0.4680037796497345, "learning_rate": 9.7225e-05, "loss": 0.0451, "step": 3890 }, { "grad_norm": 0.4578530788421631, "learning_rate": 9.747500000000001e-05, "loss": 0.0479, "step": 3900 }, { "grad_norm": 0.48216307163238525, "learning_rate": 9.7725e-05, "loss": 0.0446, "step": 3910 }, { "grad_norm": 0.5386159420013428, "learning_rate": 9.797500000000001e-05, "loss": 0.0502, "step": 3920 }, { "grad_norm": 0.5013518929481506, "learning_rate": 9.8225e-05, "loss": 0.0404, "step": 3930 }, { "grad_norm": 0.5942913889884949, "learning_rate": 9.847500000000001e-05, "loss": 0.0445, "step": 3940 }, { "grad_norm": 0.38718295097351074, "learning_rate": 9.8725e-05, "loss": 0.0429, "step": 3950 }, { "grad_norm": 0.3547344207763672, "learning_rate": 9.897500000000001e-05, "loss": 0.0424, "step": 3960 }, { "grad_norm": 0.4197726249694824, "learning_rate": 9.9225e-05, "loss": 0.039, "step": 3970 }, { "grad_norm": 0.33915048837661743, "learning_rate": 9.9475e-05, "loss": 0.0389, "step": 3980 }, { "grad_norm": 0.4036312699317932, "learning_rate": 9.9725e-05, "loss": 0.0417, "step": 3990 }, { "grad_norm": 0.46218857169151306, "learning_rate": 9.9975e-05, "loss": 0.0445, "step": 4000 }, { "grad_norm": 0.47558122873306274, "learning_rate": 9.999999653982884e-05, "loss": 0.0382, "step": 4010 }, { "grad_norm": 0.5080224275588989, "learning_rate": 9.999998457874392e-05, "loss": 0.0384, "step": 4020 }, { "grad_norm": 0.5159898996353149, "learning_rate": 9.999996407402913e-05, "loss": 0.0455, "step": 4030 }, { "grad_norm": 0.4967317581176758, "learning_rate": 9.999993502568801e-05, "loss": 0.0404, "step": 4040 }, { "grad_norm": 0.536630392074585, "learning_rate": 9.999989743372548e-05, "loss": 0.0413, "step": 4050 }, { "grad_norm": 0.3744562864303589, "learning_rate": 9.999985129814798e-05, "loss": 0.041, "step": 4060 }, { "grad_norm": 0.44103363156318665, "learning_rate": 9.99997966189634e-05, "loss": 0.036, "step": 4070 }, { "grad_norm": 0.48081710934638977, "learning_rate": 9.999973339618107e-05, "loss": 0.0403, "step": 4080 }, { "grad_norm": 0.46774008870124817, "learning_rate": 9.999966162981179e-05, "loss": 0.0391, "step": 4090 }, { "grad_norm": 0.4406292140483856, "learning_rate": 9.999958131986784e-05, "loss": 0.037, "step": 4100 }, { "grad_norm": 0.43969884514808655, "learning_rate": 9.999949246636293e-05, "loss": 0.044, "step": 4110 }, { "grad_norm": 0.40806061029434204, "learning_rate": 9.999939506931224e-05, "loss": 0.043, "step": 4120 }, { "grad_norm": 0.48288166522979736, "learning_rate": 9.999928912873243e-05, "loss": 0.0435, "step": 4130 }, { "grad_norm": 0.5041936635971069, "learning_rate": 9.999917464464159e-05, "loss": 0.0392, "step": 4140 }, { "grad_norm": 0.4439799189567566, "learning_rate": 9.999905161705929e-05, "loss": 0.0417, "step": 4150 }, { "grad_norm": 0.4274817407131195, "learning_rate": 9.999892004600653e-05, "loss": 0.0392, "step": 4160 }, { "grad_norm": 0.3829805850982666, "learning_rate": 9.999877993150581e-05, "loss": 0.0409, "step": 4170 }, { "grad_norm": 0.37050098180770874, "learning_rate": 9.999863127358108e-05, "loss": 0.0384, "step": 4180 }, { "grad_norm": 0.6925395727157593, "learning_rate": 9.999847407225773e-05, "loss": 0.0441, "step": 4190 }, { "grad_norm": 0.4985874593257904, "learning_rate": 9.999830832756262e-05, "loss": 0.0397, "step": 4200 }, { "grad_norm": 0.44564226269721985, "learning_rate": 9.999813403952407e-05, "loss": 0.0388, "step": 4210 }, { "grad_norm": 0.5525736808776855, "learning_rate": 9.999795120817187e-05, "loss": 0.0387, "step": 4220 }, { "grad_norm": 0.37908071279525757, "learning_rate": 9.999775983353725e-05, "loss": 0.04, "step": 4230 }, { "grad_norm": 0.4678504765033722, "learning_rate": 9.999755991565292e-05, "loss": 0.0393, "step": 4240 }, { "grad_norm": 0.49166667461395264, "learning_rate": 9.999735145455303e-05, "loss": 0.0392, "step": 4250 }, { "grad_norm": 0.4057224988937378, "learning_rate": 9.99971344502732e-05, "loss": 0.0399, "step": 4260 }, { "grad_norm": 0.3757214844226837, "learning_rate": 9.999690890285053e-05, "loss": 0.0408, "step": 4270 }, { "grad_norm": 0.5076956152915955, "learning_rate": 9.999667481232356e-05, "loss": 0.0435, "step": 4280 }, { "grad_norm": 0.4718482196331024, "learning_rate": 9.999643217873225e-05, "loss": 0.0386, "step": 4290 }, { "grad_norm": 0.4639871418476105, "learning_rate": 9.999618100211809e-05, "loss": 0.0357, "step": 4300 }, { "grad_norm": 0.48065316677093506, "learning_rate": 9.999592128252402e-05, "loss": 0.0372, "step": 4310 }, { "grad_norm": 0.4310382306575775, "learning_rate": 9.999565301999437e-05, "loss": 0.0414, "step": 4320 }, { "grad_norm": 0.3950827419757843, "learning_rate": 9.999537621457502e-05, "loss": 0.0359, "step": 4330 }, { "grad_norm": 0.5074281096458435, "learning_rate": 9.999509086631323e-05, "loss": 0.0395, "step": 4340 }, { "grad_norm": 0.38792669773101807, "learning_rate": 9.99947969752578e-05, "loss": 0.0457, "step": 4350 }, { "grad_norm": 0.39192038774490356, "learning_rate": 9.999449454145891e-05, "loss": 0.0395, "step": 4360 }, { "grad_norm": 0.3484737277030945, "learning_rate": 9.999418356496827e-05, "loss": 0.0409, "step": 4370 }, { "grad_norm": 0.45539599657058716, "learning_rate": 9.999386404583899e-05, "loss": 0.0415, "step": 4380 }, { "grad_norm": 0.4595767855644226, "learning_rate": 9.999353598412568e-05, "loss": 0.0459, "step": 4390 }, { "grad_norm": 0.5508831739425659, "learning_rate": 9.999319937988442e-05, "loss": 0.0382, "step": 4400 }, { "grad_norm": 0.35712650418281555, "learning_rate": 9.999285423317268e-05, "loss": 0.0393, "step": 4410 }, { "grad_norm": 0.41192010045051575, "learning_rate": 9.999250054404947e-05, "loss": 0.0392, "step": 4420 }, { "grad_norm": 0.351890504360199, "learning_rate": 9.99921383125752e-05, "loss": 0.0374, "step": 4430 }, { "grad_norm": 0.6328191161155701, "learning_rate": 9.99917675388118e-05, "loss": 0.0388, "step": 4440 }, { "grad_norm": 0.47150811553001404, "learning_rate": 9.99913882228226e-05, "loss": 0.0413, "step": 4450 }, { "grad_norm": 0.44781795144081116, "learning_rate": 9.999100036467242e-05, "loss": 0.0392, "step": 4460 }, { "grad_norm": 0.3667055070400238, "learning_rate": 9.999060396442753e-05, "loss": 0.0353, "step": 4470 }, { "grad_norm": 0.567460298538208, "learning_rate": 9.999019902215566e-05, "loss": 0.0385, "step": 4480 }, { "grad_norm": 0.4661732614040375, "learning_rate": 9.998978553792602e-05, "loss": 0.0442, "step": 4490 }, { "grad_norm": 0.3055354952812195, "learning_rate": 9.998936351180926e-05, "loss": 0.0366, "step": 4500 }, { "grad_norm": 0.3469107151031494, "learning_rate": 9.998893294387747e-05, "loss": 0.0347, "step": 4510 }, { "grad_norm": 0.3152364194393158, "learning_rate": 9.998849383420426e-05, "loss": 0.0338, "step": 4520 }, { "grad_norm": 0.4181658625602722, "learning_rate": 9.998804618286465e-05, "loss": 0.0395, "step": 4530 }, { "grad_norm": 0.3888331949710846, "learning_rate": 9.99875899899351e-05, "loss": 0.0397, "step": 4540 }, { "grad_norm": 0.3873956799507141, "learning_rate": 9.99871252554936e-05, "loss": 0.0337, "step": 4550 }, { "grad_norm": 0.40072646737098694, "learning_rate": 9.998665197961955e-05, "loss": 0.0361, "step": 4560 }, { "grad_norm": 0.4597735106945038, "learning_rate": 9.998617016239379e-05, "loss": 0.0346, "step": 4570 }, { "grad_norm": 0.39562666416168213, "learning_rate": 9.998567980389869e-05, "loss": 0.0423, "step": 4580 }, { "grad_norm": 0.4199996590614319, "learning_rate": 9.998518090421802e-05, "loss": 0.0421, "step": 4590 }, { "grad_norm": 0.4308481812477112, "learning_rate": 9.998467346343703e-05, "loss": 0.0369, "step": 4600 }, { "grad_norm": 0.5386255383491516, "learning_rate": 9.998415748164243e-05, "loss": 0.0436, "step": 4610 }, { "grad_norm": 0.40902936458587646, "learning_rate": 9.998363295892238e-05, "loss": 0.0372, "step": 4620 }, { "grad_norm": 0.3840460479259491, "learning_rate": 9.998309989536652e-05, "loss": 0.0357, "step": 4630 }, { "grad_norm": 0.48823195695877075, "learning_rate": 9.998255829106593e-05, "loss": 0.0391, "step": 4640 }, { "grad_norm": 0.383924275636673, "learning_rate": 9.998200814611316e-05, "loss": 0.0353, "step": 4650 }, { "grad_norm": 0.38105812668800354, "learning_rate": 9.998144946060219e-05, "loss": 0.0408, "step": 4660 }, { "grad_norm": 0.4539966583251953, "learning_rate": 9.998088223462852e-05, "loss": 0.0372, "step": 4670 }, { "grad_norm": 0.3961423635482788, "learning_rate": 9.998030646828905e-05, "loss": 0.0334, "step": 4680 }, { "grad_norm": 0.3327075242996216, "learning_rate": 9.997972216168217e-05, "loss": 0.0344, "step": 4690 }, { "grad_norm": 0.49045053124427795, "learning_rate": 9.997912931490771e-05, "loss": 0.0449, "step": 4700 }, { "grad_norm": 0.37813153862953186, "learning_rate": 9.9978527928067e-05, "loss": 0.0373, "step": 4710 }, { "grad_norm": 0.3059055507183075, "learning_rate": 9.997791800126277e-05, "loss": 0.0325, "step": 4720 }, { "grad_norm": 0.30632758140563965, "learning_rate": 9.997729953459927e-05, "loss": 0.0346, "step": 4730 }, { "grad_norm": 0.43793514370918274, "learning_rate": 9.997667252818214e-05, "loss": 0.0357, "step": 4740 }, { "grad_norm": 0.44608891010284424, "learning_rate": 9.997603698211855e-05, "loss": 0.0354, "step": 4750 }, { "grad_norm": 0.33075863122940063, "learning_rate": 9.99753928965171e-05, "loss": 0.0392, "step": 4760 }, { "grad_norm": 0.3343106806278229, "learning_rate": 9.997474027148781e-05, "loss": 0.0418, "step": 4770 }, { "grad_norm": 0.40731683373451233, "learning_rate": 9.997407910714223e-05, "loss": 0.0356, "step": 4780 }, { "grad_norm": 0.3214406371116638, "learning_rate": 9.997340940359332e-05, "loss": 0.0394, "step": 4790 }, { "grad_norm": 0.3589399456977844, "learning_rate": 9.997273116095552e-05, "loss": 0.0372, "step": 4800 }, { "grad_norm": 0.517250657081604, "learning_rate": 9.997204437934473e-05, "loss": 0.0325, "step": 4810 }, { "grad_norm": 0.5613711476325989, "learning_rate": 9.997134905887829e-05, "loss": 0.0363, "step": 4820 }, { "grad_norm": 0.41212227940559387, "learning_rate": 9.997064519967501e-05, "loss": 0.0349, "step": 4830 }, { "grad_norm": 0.4087931215763092, "learning_rate": 9.996993280185517e-05, "loss": 0.0358, "step": 4840 }, { "grad_norm": 0.35932064056396484, "learning_rate": 9.99692118655405e-05, "loss": 0.0353, "step": 4850 }, { "grad_norm": 0.40922731161117554, "learning_rate": 9.996848239085417e-05, "loss": 0.033, "step": 4860 }, { "grad_norm": 0.4200252592563629, "learning_rate": 9.996774437792085e-05, "loss": 0.039, "step": 4870 }, { "grad_norm": 0.41649043560028076, "learning_rate": 9.996699782686664e-05, "loss": 0.0355, "step": 4880 }, { "grad_norm": 0.3118925988674164, "learning_rate": 9.996624273781909e-05, "loss": 0.0367, "step": 4890 }, { "grad_norm": 0.4377591907978058, "learning_rate": 9.996547911090725e-05, "loss": 0.0347, "step": 4900 }, { "grad_norm": 0.27006828784942627, "learning_rate": 9.996470694626157e-05, "loss": 0.0335, "step": 4910 }, { "grad_norm": 0.3147871494293213, "learning_rate": 9.996392624401403e-05, "loss": 0.0359, "step": 4920 }, { "grad_norm": 0.3262457251548767, "learning_rate": 9.996313700429801e-05, "loss": 0.0358, "step": 4930 }, { "grad_norm": 0.4495568573474884, "learning_rate": 9.996233922724836e-05, "loss": 0.031, "step": 4940 }, { "grad_norm": 0.42858967185020447, "learning_rate": 9.996153291300141e-05, "loss": 0.0379, "step": 4950 }, { "grad_norm": 0.34447962045669556, "learning_rate": 9.996071806169494e-05, "loss": 0.0372, "step": 4960 }, { "grad_norm": 0.41888663172721863, "learning_rate": 9.995989467346817e-05, "loss": 0.0308, "step": 4970 }, { "grad_norm": 0.42842334508895874, "learning_rate": 9.995906274846183e-05, "loss": 0.0396, "step": 4980 }, { "grad_norm": 0.4065428674221039, "learning_rate": 9.995822228681803e-05, "loss": 0.0381, "step": 4990 }, { "grad_norm": 0.3526872992515564, "learning_rate": 9.99573732886804e-05, "loss": 0.035, "step": 5000 }, { "grad_norm": 0.4071899652481079, "learning_rate": 9.995651575419402e-05, "loss": 0.036, "step": 5010 }, { "grad_norm": 0.5896841883659363, "learning_rate": 9.995564968350541e-05, "loss": 0.0376, "step": 5020 }, { "grad_norm": 0.36095598340034485, "learning_rate": 9.995477507676256e-05, "loss": 0.0333, "step": 5030 }, { "grad_norm": 0.5373207926750183, "learning_rate": 9.995389193411493e-05, "loss": 0.0332, "step": 5040 }, { "grad_norm": 0.45306524634361267, "learning_rate": 9.995300025571339e-05, "loss": 0.0381, "step": 5050 }, { "grad_norm": 0.40149158239364624, "learning_rate": 9.995210004171034e-05, "loss": 0.0374, "step": 5060 }, { "grad_norm": 0.4060264527797699, "learning_rate": 9.995119129225956e-05, "loss": 0.0347, "step": 5070 }, { "grad_norm": 0.4295032024383545, "learning_rate": 9.995027400751637e-05, "loss": 0.0334, "step": 5080 }, { "grad_norm": 0.4219134747982025, "learning_rate": 9.994934818763751e-05, "loss": 0.0363, "step": 5090 }, { "grad_norm": 0.43910256028175354, "learning_rate": 9.994841383278115e-05, "loss": 0.0341, "step": 5100 }, { "grad_norm": 0.3306833505630493, "learning_rate": 9.994747094310695e-05, "loss": 0.0357, "step": 5110 }, { "grad_norm": 0.3618801236152649, "learning_rate": 9.994651951877604e-05, "loss": 0.0317, "step": 5120 }, { "grad_norm": 0.3809983730316162, "learning_rate": 9.994555955995099e-05, "loss": 0.034, "step": 5130 }, { "grad_norm": 0.4732702076435089, "learning_rate": 9.994459106679581e-05, "loss": 0.0364, "step": 5140 }, { "grad_norm": 0.37927693128585815, "learning_rate": 9.994361403947603e-05, "loss": 0.0355, "step": 5150 }, { "grad_norm": 0.46525827050209045, "learning_rate": 9.994262847815854e-05, "loss": 0.0344, "step": 5160 }, { "grad_norm": 0.3434266448020935, "learning_rate": 9.99416343830118e-05, "loss": 0.0316, "step": 5170 }, { "grad_norm": 0.25881528854370117, "learning_rate": 9.994063175420565e-05, "loss": 0.0291, "step": 5180 }, { "grad_norm": 0.32273349165916443, "learning_rate": 9.99396205919114e-05, "loss": 0.0351, "step": 5190 }, { "grad_norm": 0.3977406322956085, "learning_rate": 9.993860089630185e-05, "loss": 0.0347, "step": 5200 }, { "grad_norm": 0.4778915345668793, "learning_rate": 9.993757266755123e-05, "loss": 0.0418, "step": 5210 }, { "grad_norm": 0.3476613461971283, "learning_rate": 9.993653590583522e-05, "loss": 0.0402, "step": 5220 }, { "grad_norm": 0.3140815198421478, "learning_rate": 9.993549061133102e-05, "loss": 0.0356, "step": 5230 }, { "grad_norm": 0.32748013734817505, "learning_rate": 9.993443678421719e-05, "loss": 0.0297, "step": 5240 }, { "grad_norm": 0.3628648519515991, "learning_rate": 9.993337442467384e-05, "loss": 0.0332, "step": 5250 }, { "grad_norm": 0.3664923310279846, "learning_rate": 9.993230353288248e-05, "loss": 0.0344, "step": 5260 }, { "grad_norm": 0.34068167209625244, "learning_rate": 9.993122410902608e-05, "loss": 0.0342, "step": 5270 }, { "grad_norm": 0.4194341003894806, "learning_rate": 9.993013615328912e-05, "loss": 0.0341, "step": 5280 }, { "grad_norm": 0.3868018686771393, "learning_rate": 9.992903966585747e-05, "loss": 0.0338, "step": 5290 }, { "grad_norm": 0.35713067650794983, "learning_rate": 9.992793464691852e-05, "loss": 0.04, "step": 5300 }, { "grad_norm": 0.4750957190990448, "learning_rate": 9.992682109666105e-05, "loss": 0.0397, "step": 5310 }, { "grad_norm": 0.31837981939315796, "learning_rate": 9.992569901527538e-05, "loss": 0.0415, "step": 5320 }, { "grad_norm": 0.4458063840866089, "learning_rate": 9.99245684029532e-05, "loss": 0.0345, "step": 5330 }, { "grad_norm": 0.39007851481437683, "learning_rate": 9.992342925988774e-05, "loss": 0.0401, "step": 5340 }, { "grad_norm": 0.3868679106235504, "learning_rate": 9.992228158627361e-05, "loss": 0.0382, "step": 5350 }, { "grad_norm": 0.3444865643978119, "learning_rate": 9.992112538230693e-05, "loss": 0.0331, "step": 5360 }, { "grad_norm": 0.4024183750152588, "learning_rate": 9.991996064818527e-05, "loss": 0.0331, "step": 5370 }, { "grad_norm": 0.3231894373893738, "learning_rate": 9.991878738410768e-05, "loss": 0.0338, "step": 5380 }, { "grad_norm": 0.28500717878341675, "learning_rate": 9.991760559027457e-05, "loss": 0.0338, "step": 5390 }, { "grad_norm": 0.3753765821456909, "learning_rate": 9.991641526688793e-05, "loss": 0.0302, "step": 5400 }, { "grad_norm": 0.36098191142082214, "learning_rate": 9.991521641415113e-05, "loss": 0.0323, "step": 5410 }, { "grad_norm": 0.2724384069442749, "learning_rate": 9.991400903226904e-05, "loss": 0.031, "step": 5420 }, { "grad_norm": 0.2900634706020355, "learning_rate": 9.991279312144794e-05, "loss": 0.0353, "step": 5430 }, { "grad_norm": 0.29740408062934875, "learning_rate": 9.991156868189564e-05, "loss": 0.0324, "step": 5440 }, { "grad_norm": 0.37989136576652527, "learning_rate": 9.991033571382131e-05, "loss": 0.034, "step": 5450 }, { "grad_norm": 0.41887202858924866, "learning_rate": 9.990909421743569e-05, "loss": 0.0327, "step": 5460 }, { "grad_norm": 0.3076537549495697, "learning_rate": 9.990784419295085e-05, "loss": 0.0303, "step": 5470 }, { "grad_norm": 0.26901566982269287, "learning_rate": 9.990658564058044e-05, "loss": 0.0326, "step": 5480 }, { "grad_norm": 0.3339686393737793, "learning_rate": 9.990531856053948e-05, "loss": 0.033, "step": 5490 }, { "grad_norm": 0.37015560269355774, "learning_rate": 9.99040429530445e-05, "loss": 0.0313, "step": 5500 }, { "grad_norm": 0.3014410138130188, "learning_rate": 9.990275881831346e-05, "loss": 0.0335, "step": 5510 }, { "grad_norm": 0.37858060002326965, "learning_rate": 9.990146615656577e-05, "loss": 0.0347, "step": 5520 }, { "grad_norm": 0.32660192251205444, "learning_rate": 9.990016496802233e-05, "loss": 0.0332, "step": 5530 }, { "grad_norm": 0.34484946727752686, "learning_rate": 9.989885525290548e-05, "loss": 0.0319, "step": 5540 }, { "grad_norm": 0.312542200088501, "learning_rate": 9.989753701143897e-05, "loss": 0.033, "step": 5550 }, { "grad_norm": 0.479170560836792, "learning_rate": 9.989621024384812e-05, "loss": 0.0322, "step": 5560 }, { "grad_norm": 0.4115579128265381, "learning_rate": 9.989487495035959e-05, "loss": 0.0342, "step": 5570 }, { "grad_norm": 0.3931794762611389, "learning_rate": 9.989353113120156e-05, "loss": 0.0296, "step": 5580 }, { "grad_norm": 0.2526790201663971, "learning_rate": 9.989217878660366e-05, "loss": 0.0326, "step": 5590 }, { "grad_norm": 0.32855290174484253, "learning_rate": 9.989081791679695e-05, "loss": 0.0306, "step": 5600 }, { "grad_norm": 0.3531206250190735, "learning_rate": 9.988944852201397e-05, "loss": 0.0314, "step": 5610 }, { "grad_norm": 0.41100025177001953, "learning_rate": 9.988807060248873e-05, "loss": 0.0323, "step": 5620 }, { "grad_norm": 0.35145309567451477, "learning_rate": 9.988668415845665e-05, "loss": 0.0298, "step": 5630 }, { "grad_norm": 0.2989272177219391, "learning_rate": 9.988528919015466e-05, "loss": 0.0286, "step": 5640 }, { "grad_norm": 0.31739282608032227, "learning_rate": 9.988388569782112e-05, "loss": 0.0301, "step": 5650 }, { "grad_norm": 0.37526148557662964, "learning_rate": 9.988247368169583e-05, "loss": 0.0373, "step": 5660 }, { "grad_norm": 0.29414454102516174, "learning_rate": 9.988105314202007e-05, "loss": 0.0306, "step": 5670 }, { "grad_norm": 0.36253559589385986, "learning_rate": 9.987962407903659e-05, "loss": 0.0317, "step": 5680 }, { "grad_norm": 0.4175386130809784, "learning_rate": 9.987818649298957e-05, "loss": 0.0303, "step": 5690 }, { "grad_norm": 0.436500608921051, "learning_rate": 9.987674038412465e-05, "loss": 0.0353, "step": 5700 }, { "grad_norm": 0.42382943630218506, "learning_rate": 9.987528575268891e-05, "loss": 0.032, "step": 5710 }, { "grad_norm": 0.2757951021194458, "learning_rate": 9.987382259893095e-05, "loss": 0.0321, "step": 5720 }, { "grad_norm": 0.41561776399612427, "learning_rate": 9.987235092310074e-05, "loss": 0.0318, "step": 5730 }, { "grad_norm": 0.3394024968147278, "learning_rate": 9.987087072544978e-05, "loss": 0.0343, "step": 5740 }, { "grad_norm": 0.44215551018714905, "learning_rate": 9.9869382006231e-05, "loss": 0.0341, "step": 5750 }, { "grad_norm": 0.3137679100036621, "learning_rate": 9.986788476569875e-05, "loss": 0.0315, "step": 5760 }, { "grad_norm": 0.3578633666038513, "learning_rate": 9.986637900410887e-05, "loss": 0.0352, "step": 5770 }, { "grad_norm": 0.3281468451023102, "learning_rate": 9.986486472171869e-05, "loss": 0.0321, "step": 5780 }, { "grad_norm": 0.45072659850120544, "learning_rate": 9.986334191878692e-05, "loss": 0.0341, "step": 5790 }, { "grad_norm": 0.3654766380786896, "learning_rate": 9.986181059557378e-05, "loss": 0.0319, "step": 5800 }, { "grad_norm": 0.3343709707260132, "learning_rate": 9.986027075234094e-05, "loss": 0.0304, "step": 5810 }, { "grad_norm": 0.34941601753234863, "learning_rate": 9.985872238935152e-05, "loss": 0.0312, "step": 5820 }, { "grad_norm": 0.34704017639160156, "learning_rate": 9.985716550687008e-05, "loss": 0.034, "step": 5830 }, { "grad_norm": 0.4856829047203064, "learning_rate": 9.985560010516264e-05, "loss": 0.0332, "step": 5840 }, { "grad_norm": 0.4019835591316223, "learning_rate": 9.985402618449668e-05, "loss": 0.0336, "step": 5850 }, { "grad_norm": 0.3479267656803131, "learning_rate": 9.985244374514118e-05, "loss": 0.0337, "step": 5860 }, { "grad_norm": 0.48318013548851013, "learning_rate": 9.985085278736651e-05, "loss": 0.0373, "step": 5870 }, { "grad_norm": 0.42336684465408325, "learning_rate": 9.984925331144452e-05, "loss": 0.0356, "step": 5880 }, { "grad_norm": 0.3215020000934601, "learning_rate": 9.984764531764851e-05, "loss": 0.0311, "step": 5890 }, { "grad_norm": 0.4494810104370117, "learning_rate": 9.984602880625326e-05, "loss": 0.0321, "step": 5900 }, { "grad_norm": 0.2652421295642853, "learning_rate": 9.9844403777535e-05, "loss": 0.0336, "step": 5910 }, { "grad_norm": 0.44637423753738403, "learning_rate": 9.984277023177135e-05, "loss": 0.0312, "step": 5920 }, { "grad_norm": 0.29258304834365845, "learning_rate": 9.984112816924148e-05, "loss": 0.0292, "step": 5930 }, { "grad_norm": 0.3191801607608795, "learning_rate": 9.983947759022596e-05, "loss": 0.0336, "step": 5940 }, { "grad_norm": 0.2908138036727905, "learning_rate": 9.983781849500682e-05, "loss": 0.0298, "step": 5950 }, { "grad_norm": 0.34816813468933105, "learning_rate": 9.98361508838676e-05, "loss": 0.0294, "step": 5960 }, { "grad_norm": 0.338552325963974, "learning_rate": 9.98344747570932e-05, "loss": 0.0336, "step": 5970 }, { "grad_norm": 0.2623736262321472, "learning_rate": 9.983279011497004e-05, "loss": 0.0299, "step": 5980 }, { "grad_norm": 0.34975093603134155, "learning_rate": 9.983109695778596e-05, "loss": 0.0329, "step": 5990 }, { "grad_norm": 0.3581708073616028, "learning_rate": 9.982939528583032e-05, "loss": 0.0333, "step": 6000 }, { "grad_norm": 0.27011409401893616, "learning_rate": 9.982768509939385e-05, "loss": 0.0278, "step": 6010 }, { "grad_norm": 0.3321669399738312, "learning_rate": 9.982596639876879e-05, "loss": 0.0272, "step": 6020 }, { "grad_norm": 0.3262157440185547, "learning_rate": 9.982423918424881e-05, "loss": 0.0282, "step": 6030 }, { "grad_norm": 0.3092035949230194, "learning_rate": 9.982250345612908e-05, "loss": 0.0319, "step": 6040 }, { "grad_norm": 0.3108018934726715, "learning_rate": 9.982075921470611e-05, "loss": 0.0319, "step": 6050 }, { "grad_norm": 0.32161515951156616, "learning_rate": 9.981900646027802e-05, "loss": 0.0286, "step": 6060 }, { "grad_norm": 0.2808590531349182, "learning_rate": 9.981724519314425e-05, "loss": 0.0319, "step": 6070 }, { "grad_norm": 0.4124588370323181, "learning_rate": 9.981547541360581e-05, "loss": 0.0334, "step": 6080 }, { "grad_norm": 0.29417183995246887, "learning_rate": 9.981369712196508e-05, "loss": 0.0287, "step": 6090 }, { "grad_norm": 0.44982674717903137, "learning_rate": 9.981191031852592e-05, "loss": 0.0372, "step": 6100 }, { "grad_norm": 0.23906557261943817, "learning_rate": 9.981011500359362e-05, "loss": 0.0316, "step": 6110 }, { "grad_norm": 0.40655723214149475, "learning_rate": 9.9808311177475e-05, "loss": 0.036, "step": 6120 }, { "grad_norm": 0.32039526104927063, "learning_rate": 9.980649884047826e-05, "loss": 0.0302, "step": 6130 }, { "grad_norm": 0.4179061949253082, "learning_rate": 9.980467799291307e-05, "loss": 0.0293, "step": 6140 }, { "grad_norm": 0.3827386796474457, "learning_rate": 9.980284863509058e-05, "loss": 0.0349, "step": 6150 }, { "grad_norm": 0.33617115020751953, "learning_rate": 9.980101076732334e-05, "loss": 0.0296, "step": 6160 }, { "grad_norm": 0.2995837330818176, "learning_rate": 9.979916438992544e-05, "loss": 0.0317, "step": 6170 }, { "grad_norm": 0.3518136143684387, "learning_rate": 9.979730950321237e-05, "loss": 0.0307, "step": 6180 }, { "grad_norm": 0.3361510932445526, "learning_rate": 9.979544610750104e-05, "loss": 0.0286, "step": 6190 }, { "grad_norm": 0.3469744026660919, "learning_rate": 9.97935742031099e-05, "loss": 0.0301, "step": 6200 }, { "grad_norm": 0.28316450119018555, "learning_rate": 9.979169379035878e-05, "loss": 0.0302, "step": 6210 }, { "grad_norm": 0.3428787887096405, "learning_rate": 9.978980486956899e-05, "loss": 0.0346, "step": 6220 }, { "grad_norm": 0.36163434386253357, "learning_rate": 9.978790744106332e-05, "loss": 0.0309, "step": 6230 }, { "grad_norm": 0.40377575159072876, "learning_rate": 9.978600150516594e-05, "loss": 0.0302, "step": 6240 }, { "grad_norm": 0.2866468131542206, "learning_rate": 9.978408706220259e-05, "loss": 0.0342, "step": 6250 }, { "grad_norm": 0.29650208353996277, "learning_rate": 9.978216411250032e-05, "loss": 0.0347, "step": 6260 }, { "grad_norm": 0.30374664068222046, "learning_rate": 9.978023265638778e-05, "loss": 0.0283, "step": 6270 }, { "grad_norm": 0.32150813937187195, "learning_rate": 9.977829269419495e-05, "loss": 0.0309, "step": 6280 }, { "grad_norm": 0.41207990050315857, "learning_rate": 9.977634422625335e-05, "loss": 0.0288, "step": 6290 }, { "grad_norm": 0.3754229247570038, "learning_rate": 9.97743872528959e-05, "loss": 0.0372, "step": 6300 }, { "grad_norm": 0.39992260932922363, "learning_rate": 9.9772421774457e-05, "loss": 0.0347, "step": 6310 }, { "grad_norm": 0.2779526114463806, "learning_rate": 9.977044779127252e-05, "loss": 0.0314, "step": 6320 }, { "grad_norm": 0.3349473178386688, "learning_rate": 9.976846530367971e-05, "loss": 0.0327, "step": 6330 }, { "grad_norm": 0.3913077414035797, "learning_rate": 9.976647431201735e-05, "loss": 0.0348, "step": 6340 }, { "grad_norm": 0.31271082162857056, "learning_rate": 9.976447481662568e-05, "loss": 0.0334, "step": 6350 }, { "grad_norm": 0.2740461528301239, "learning_rate": 9.976246681784629e-05, "loss": 0.0353, "step": 6360 }, { "grad_norm": 0.2991318702697754, "learning_rate": 9.976045031602234e-05, "loss": 0.0327, "step": 6370 }, { "grad_norm": 0.3986910879611969, "learning_rate": 9.975842531149837e-05, "loss": 0.0349, "step": 6380 }, { "grad_norm": 0.293093204498291, "learning_rate": 9.975639180462043e-05, "loss": 0.033, "step": 6390 }, { "grad_norm": 0.41637787222862244, "learning_rate": 9.975434979573596e-05, "loss": 0.0295, "step": 6400 }, { "grad_norm": 0.378795325756073, "learning_rate": 9.97522992851939e-05, "loss": 0.0368, "step": 6410 }, { "grad_norm": 0.40567880868911743, "learning_rate": 9.975024027334461e-05, "loss": 0.0343, "step": 6420 }, { "grad_norm": 0.32505759596824646, "learning_rate": 9.974817276053993e-05, "loss": 0.0316, "step": 6430 }, { "grad_norm": 0.2743367850780487, "learning_rate": 9.974609674713315e-05, "loss": 0.0307, "step": 6440 }, { "grad_norm": 0.37759843468666077, "learning_rate": 9.9744012233479e-05, "loss": 0.0302, "step": 6450 }, { "grad_norm": 0.45010432600975037, "learning_rate": 9.974191921993366e-05, "loss": 0.031, "step": 6460 }, { "grad_norm": 0.3921460211277008, "learning_rate": 9.973981770685474e-05, "loss": 0.0309, "step": 6470 }, { "grad_norm": 0.313876748085022, "learning_rate": 9.97377076946014e-05, "loss": 0.0334, "step": 6480 }, { "grad_norm": 0.3535693287849426, "learning_rate": 9.973558918353412e-05, "loss": 0.033, "step": 6490 }, { "grad_norm": 0.3872658908367157, "learning_rate": 9.973346217401494e-05, "loss": 0.0288, "step": 6500 }, { "grad_norm": 0.40003201365470886, "learning_rate": 9.973132666640726e-05, "loss": 0.0303, "step": 6510 }, { "grad_norm": 0.3949401080608368, "learning_rate": 9.972918266107602e-05, "loss": 0.0298, "step": 6520 }, { "grad_norm": 0.3031045198440552, "learning_rate": 9.972703015838756e-05, "loss": 0.0309, "step": 6530 }, { "grad_norm": 0.3342360854148865, "learning_rate": 9.97248691587097e-05, "loss": 0.0324, "step": 6540 }, { "grad_norm": 0.42212340235710144, "learning_rate": 9.972269966241166e-05, "loss": 0.032, "step": 6550 }, { "grad_norm": 0.2800045311450958, "learning_rate": 9.972052166986417e-05, "loss": 0.0326, "step": 6560 }, { "grad_norm": 0.33253800868988037, "learning_rate": 9.971833518143938e-05, "loss": 0.0342, "step": 6570 }, { "grad_norm": 0.3413967192173004, "learning_rate": 9.971614019751093e-05, "loss": 0.0293, "step": 6580 }, { "grad_norm": 0.423807293176651, "learning_rate": 9.971393671845383e-05, "loss": 0.032, "step": 6590 }, { "grad_norm": 0.37101051211357117, "learning_rate": 9.971172474464464e-05, "loss": 0.0344, "step": 6600 }, { "grad_norm": 0.26471230387687683, "learning_rate": 9.97095042764613e-05, "loss": 0.0346, "step": 6610 }, { "grad_norm": 0.27418580651283264, "learning_rate": 9.970727531428324e-05, "loss": 0.0265, "step": 6620 }, { "grad_norm": 0.35305094718933105, "learning_rate": 9.970503785849132e-05, "loss": 0.0351, "step": 6630 }, { "grad_norm": 0.31228193640708923, "learning_rate": 9.970279190946788e-05, "loss": 0.0322, "step": 6640 }, { "grad_norm": 0.300090491771698, "learning_rate": 9.970053746759667e-05, "loss": 0.0298, "step": 6650 }, { "grad_norm": 0.29818427562713623, "learning_rate": 9.969827453326292e-05, "loss": 0.0323, "step": 6660 }, { "grad_norm": 0.31473416090011597, "learning_rate": 9.969600310685332e-05, "loss": 0.0334, "step": 6670 }, { "grad_norm": 0.31777480244636536, "learning_rate": 9.969372318875596e-05, "loss": 0.0318, "step": 6680 }, { "grad_norm": 0.39961791038513184, "learning_rate": 9.969143477936043e-05, "loss": 0.0316, "step": 6690 }, { "grad_norm": 0.2868131101131439, "learning_rate": 9.968913787905775e-05, "loss": 0.0317, "step": 6700 }, { "grad_norm": 0.3242379128932953, "learning_rate": 9.968683248824045e-05, "loss": 0.0305, "step": 6710 }, { "grad_norm": 0.29403743147850037, "learning_rate": 9.968451860730238e-05, "loss": 0.0313, "step": 6720 }, { "grad_norm": 0.3361237049102783, "learning_rate": 9.968219623663896e-05, "loss": 0.0278, "step": 6730 }, { "grad_norm": 0.33603182435035706, "learning_rate": 9.967986537664702e-05, "loss": 0.0292, "step": 6740 }, { "grad_norm": 0.36707979440689087, "learning_rate": 9.967752602772483e-05, "loss": 0.0281, "step": 6750 }, { "grad_norm": 0.36621546745300293, "learning_rate": 9.967517819027212e-05, "loss": 0.0286, "step": 6760 }, { "grad_norm": 0.3524555265903473, "learning_rate": 9.967282186469009e-05, "loss": 0.0295, "step": 6770 }, { "grad_norm": 0.3717699944972992, "learning_rate": 9.967045705138135e-05, "loss": 0.0304, "step": 6780 }, { "grad_norm": 0.29584020376205444, "learning_rate": 9.966808375074998e-05, "loss": 0.0313, "step": 6790 }, { "grad_norm": 0.3446604013442993, "learning_rate": 9.966570196320154e-05, "loss": 0.0291, "step": 6800 }, { "grad_norm": 0.3009924590587616, "learning_rate": 9.966331168914299e-05, "loss": 0.0298, "step": 6810 }, { "grad_norm": 0.3252083659172058, "learning_rate": 9.966091292898277e-05, "loss": 0.032, "step": 6820 }, { "grad_norm": 0.31732451915740967, "learning_rate": 9.965850568313076e-05, "loss": 0.0285, "step": 6830 }, { "grad_norm": 0.28692251443862915, "learning_rate": 9.965608995199827e-05, "loss": 0.0303, "step": 6840 }, { "grad_norm": 0.31890904903411865, "learning_rate": 9.965366573599812e-05, "loss": 0.0321, "step": 6850 }, { "grad_norm": 0.394978404045105, "learning_rate": 9.965123303554453e-05, "loss": 0.0337, "step": 6860 }, { "grad_norm": 0.30974698066711426, "learning_rate": 9.964879185105317e-05, "loss": 0.0265, "step": 6870 }, { "grad_norm": 0.47844287753105164, "learning_rate": 9.964634218294119e-05, "loss": 0.0342, "step": 6880 }, { "grad_norm": 0.25579068064689636, "learning_rate": 9.964388403162714e-05, "loss": 0.0269, "step": 6890 }, { "grad_norm": 0.30806058645248413, "learning_rate": 9.96414173975311e-05, "loss": 0.0269, "step": 6900 }, { "grad_norm": 0.3879929780960083, "learning_rate": 9.963894228107451e-05, "loss": 0.027, "step": 6910 }, { "grad_norm": 0.3306339681148529, "learning_rate": 9.963645868268032e-05, "loss": 0.0298, "step": 6920 }, { "grad_norm": 0.322408527135849, "learning_rate": 9.963396660277289e-05, "loss": 0.0298, "step": 6930 }, { "grad_norm": 0.34294936060905457, "learning_rate": 9.963146604177807e-05, "loss": 0.0266, "step": 6940 }, { "grad_norm": 0.3528563678264618, "learning_rate": 9.962895700012311e-05, "loss": 0.0274, "step": 6950 }, { "grad_norm": 0.33402037620544434, "learning_rate": 9.962643947823677e-05, "loss": 0.0293, "step": 6960 }, { "grad_norm": 0.30564388632774353, "learning_rate": 9.962391347654921e-05, "loss": 0.0264, "step": 6970 }, { "grad_norm": 0.26794636249542236, "learning_rate": 9.962137899549204e-05, "loss": 0.0301, "step": 6980 }, { "grad_norm": 0.3118613660335541, "learning_rate": 9.961883603549835e-05, "loss": 0.0275, "step": 6990 }, { "grad_norm": 0.24929030239582062, "learning_rate": 9.961628459700267e-05, "loss": 0.0277, "step": 7000 }, { "grad_norm": 0.28984713554382324, "learning_rate": 9.961372468044095e-05, "loss": 0.0326, "step": 7010 }, { "grad_norm": 0.3789452910423279, "learning_rate": 9.961115628625062e-05, "loss": 0.0315, "step": 7020 }, { "grad_norm": 0.3071320056915283, "learning_rate": 9.960857941487056e-05, "loss": 0.0267, "step": 7030 }, { "grad_norm": 0.2810409367084503, "learning_rate": 9.960599406674106e-05, "loss": 0.0277, "step": 7040 }, { "grad_norm": 0.33025237917900085, "learning_rate": 9.960340024230393e-05, "loss": 0.0275, "step": 7050 }, { "grad_norm": 0.3062329888343811, "learning_rate": 9.960079794200232e-05, "loss": 0.0295, "step": 7060 }, { "grad_norm": 0.30280786752700806, "learning_rate": 9.959818716628096e-05, "loss": 0.0268, "step": 7070 }, { "grad_norm": 0.2836000919342041, "learning_rate": 9.95955679155859e-05, "loss": 0.0287, "step": 7080 }, { "grad_norm": 0.21535572409629822, "learning_rate": 9.959294019036472e-05, "loss": 0.0287, "step": 7090 }, { "grad_norm": 0.3657655715942383, "learning_rate": 9.959030399106646e-05, "loss": 0.0256, "step": 7100 }, { "grad_norm": 0.3094343841075897, "learning_rate": 9.958765931814153e-05, "loss": 0.0293, "step": 7110 }, { "grad_norm": 0.2588033676147461, "learning_rate": 9.958500617204184e-05, "loss": 0.028, "step": 7120 }, { "grad_norm": 0.32083967328071594, "learning_rate": 9.958234455322075e-05, "loss": 0.0235, "step": 7130 }, { "grad_norm": 0.31645241379737854, "learning_rate": 9.957967446213308e-05, "loss": 0.0294, "step": 7140 }, { "grad_norm": 0.3182238042354584, "learning_rate": 9.957699589923501e-05, "loss": 0.0259, "step": 7150 }, { "grad_norm": 0.4327462613582611, "learning_rate": 9.957430886498431e-05, "loss": 0.0287, "step": 7160 }, { "grad_norm": 0.3378325402736664, "learning_rate": 9.957161335984008e-05, "loss": 0.0271, "step": 7170 }, { "grad_norm": 0.2914672791957855, "learning_rate": 9.956890938426291e-05, "loss": 0.0307, "step": 7180 }, { "grad_norm": 0.36374273896217346, "learning_rate": 9.956619693871482e-05, "loss": 0.0275, "step": 7190 }, { "grad_norm": 0.2872345447540283, "learning_rate": 9.956347602365934e-05, "loss": 0.0282, "step": 7200 }, { "grad_norm": 0.26603156328201294, "learning_rate": 9.956074663956135e-05, "loss": 0.031, "step": 7210 }, { "grad_norm": 0.2327250987291336, "learning_rate": 9.955800878688726e-05, "loss": 0.0285, "step": 7220 }, { "grad_norm": 0.36189329624176025, "learning_rate": 9.955526246610489e-05, "loss": 0.0307, "step": 7230 }, { "grad_norm": 0.3818337917327881, "learning_rate": 9.955250767768349e-05, "loss": 0.0282, "step": 7240 }, { "grad_norm": 0.37623628973960876, "learning_rate": 9.95497444220938e-05, "loss": 0.0272, "step": 7250 }, { "grad_norm": 0.28843870759010315, "learning_rate": 9.954697269980797e-05, "loss": 0.0284, "step": 7260 }, { "grad_norm": 0.2297140210866928, "learning_rate": 9.954419251129962e-05, "loss": 0.0273, "step": 7270 }, { "grad_norm": 0.2455778270959854, "learning_rate": 9.95414038570438e-05, "loss": 0.0245, "step": 7280 }, { "grad_norm": 0.26574674248695374, "learning_rate": 9.953860673751703e-05, "loss": 0.03, "step": 7290 }, { "grad_norm": 0.24484989047050476, "learning_rate": 9.953580115319725e-05, "loss": 0.0265, "step": 7300 }, { "grad_norm": 0.4486551582813263, "learning_rate": 9.953298710456387e-05, "loss": 0.0269, "step": 7310 }, { "grad_norm": 0.27044323086738586, "learning_rate": 9.953016459209771e-05, "loss": 0.0269, "step": 7320 }, { "grad_norm": 0.21435406804084778, "learning_rate": 9.952733361628108e-05, "loss": 0.0303, "step": 7330 }, { "grad_norm": 0.31195732951164246, "learning_rate": 9.952449417759772e-05, "loss": 0.0269, "step": 7340 }, { "grad_norm": 0.29248130321502686, "learning_rate": 9.952164627653279e-05, "loss": 0.0303, "step": 7350 }, { "grad_norm": 0.26608806848526, "learning_rate": 9.951878991357292e-05, "loss": 0.0259, "step": 7360 }, { "grad_norm": 0.3730471432209015, "learning_rate": 9.951592508920622e-05, "loss": 0.028, "step": 7370 }, { "grad_norm": 0.40321242809295654, "learning_rate": 9.951305180392219e-05, "loss": 0.0287, "step": 7380 }, { "grad_norm": 0.3581785261631012, "learning_rate": 9.951017005821178e-05, "loss": 0.0319, "step": 7390 }, { "grad_norm": 0.3257274031639099, "learning_rate": 9.95072798525674e-05, "loss": 0.0286, "step": 7400 }, { "grad_norm": 0.326824426651001, "learning_rate": 9.950438118748293e-05, "loss": 0.0333, "step": 7410 }, { "grad_norm": 0.3922901749610901, "learning_rate": 9.950147406345366e-05, "loss": 0.029, "step": 7420 }, { "grad_norm": 0.42798489332199097, "learning_rate": 9.949855848097635e-05, "loss": 0.0377, "step": 7430 }, { "grad_norm": 0.33772584795951843, "learning_rate": 9.949563444054916e-05, "loss": 0.0316, "step": 7440 }, { "grad_norm": 0.3607120215892792, "learning_rate": 9.949270194267178e-05, "loss": 0.0311, "step": 7450 }, { "grad_norm": 0.31016337871551514, "learning_rate": 9.948976098784526e-05, "loss": 0.0307, "step": 7460 }, { "grad_norm": 0.28036895394325256, "learning_rate": 9.948681157657213e-05, "loss": 0.0358, "step": 7470 }, { "grad_norm": 0.3305390775203705, "learning_rate": 9.948385370935638e-05, "loss": 0.0286, "step": 7480 }, { "grad_norm": 0.33335158228874207, "learning_rate": 9.94808873867034e-05, "loss": 0.0307, "step": 7490 }, { "grad_norm": 0.2813812494277954, "learning_rate": 9.947791260912009e-05, "loss": 0.0312, "step": 7500 }, { "grad_norm": 0.34753185510635376, "learning_rate": 9.947492937711474e-05, "loss": 0.0272, "step": 7510 }, { "grad_norm": 0.2816055417060852, "learning_rate": 9.947193769119707e-05, "loss": 0.0262, "step": 7520 }, { "grad_norm": 0.33268776535987854, "learning_rate": 9.946893755187834e-05, "loss": 0.0286, "step": 7530 }, { "grad_norm": 0.2870080769062042, "learning_rate": 9.946592895967115e-05, "loss": 0.0276, "step": 7540 }, { "grad_norm": 0.28646236658096313, "learning_rate": 9.94629119150896e-05, "loss": 0.0283, "step": 7550 }, { "grad_norm": 0.29341915249824524, "learning_rate": 9.94598864186492e-05, "loss": 0.0301, "step": 7560 }, { "grad_norm": 0.3601512908935547, "learning_rate": 9.945685247086696e-05, "loss": 0.027, "step": 7570 }, { "grad_norm": 0.32693928480148315, "learning_rate": 9.945381007226129e-05, "loss": 0.0315, "step": 7580 }, { "grad_norm": 0.2766807973384857, "learning_rate": 9.945075922335203e-05, "loss": 0.0249, "step": 7590 }, { "grad_norm": 0.2775833308696747, "learning_rate": 9.944769992466049e-05, "loss": 0.0268, "step": 7600 }, { "grad_norm": 0.25141218304634094, "learning_rate": 9.944463217670945e-05, "loss": 0.0271, "step": 7610 }, { "grad_norm": 0.38559865951538086, "learning_rate": 9.944155598002307e-05, "loss": 0.0289, "step": 7620 }, { "grad_norm": 0.34874987602233887, "learning_rate": 9.943847133512701e-05, "loss": 0.0259, "step": 7630 }, { "grad_norm": 0.35255441069602966, "learning_rate": 9.943537824254834e-05, "loss": 0.0325, "step": 7640 }, { "grad_norm": 0.30447474122047424, "learning_rate": 9.943227670281559e-05, "loss": 0.0316, "step": 7650 }, { "grad_norm": 0.25805339217185974, "learning_rate": 9.942916671645873e-05, "loss": 0.0234, "step": 7660 }, { "grad_norm": 0.29041576385498047, "learning_rate": 9.942604828400916e-05, "loss": 0.027, "step": 7670 }, { "grad_norm": 0.21350136399269104, "learning_rate": 9.942292140599975e-05, "loss": 0.0294, "step": 7680 }, { "grad_norm": 0.3106301724910736, "learning_rate": 9.94197860829648e-05, "loss": 0.0287, "step": 7690 }, { "grad_norm": 0.2821715474128723, "learning_rate": 9.941664231544004e-05, "loss": 0.0274, "step": 7700 }, { "grad_norm": 0.21744681894779205, "learning_rate": 9.941349010396264e-05, "loss": 0.0292, "step": 7710 }, { "grad_norm": 0.2489222139120102, "learning_rate": 9.941032944907125e-05, "loss": 0.0234, "step": 7720 }, { "grad_norm": 0.2925085425376892, "learning_rate": 9.940716035130596e-05, "loss": 0.0274, "step": 7730 }, { "grad_norm": 0.24520546197891235, "learning_rate": 9.940398281120821e-05, "loss": 0.0255, "step": 7740 }, { "grad_norm": 0.27954229712486267, "learning_rate": 9.940079682932102e-05, "loss": 0.0263, "step": 7750 }, { "grad_norm": 0.31528937816619873, "learning_rate": 9.939760240618877e-05, "loss": 0.0276, "step": 7760 }, { "grad_norm": 0.23356424272060394, "learning_rate": 9.939439954235729e-05, "loss": 0.0246, "step": 7770 }, { "grad_norm": 0.32574254274368286, "learning_rate": 9.939118823837387e-05, "loss": 0.0269, "step": 7780 }, { "grad_norm": 0.2662278413772583, "learning_rate": 9.938796849478725e-05, "loss": 0.0241, "step": 7790 }, { "grad_norm": 0.2841043770313263, "learning_rate": 9.938474031214755e-05, "loss": 0.028, "step": 7800 }, { "grad_norm": 0.24374398589134216, "learning_rate": 9.938150369100643e-05, "loss": 0.0272, "step": 7810 }, { "grad_norm": 0.3137780725955963, "learning_rate": 9.93782586319169e-05, "loss": 0.0252, "step": 7820 }, { "grad_norm": 0.2915097773075104, "learning_rate": 9.937500513543348e-05, "loss": 0.0259, "step": 7830 }, { "grad_norm": 0.28134915232658386, "learning_rate": 9.937174320211207e-05, "loss": 0.0287, "step": 7840 }, { "grad_norm": 0.2655693590641022, "learning_rate": 9.936847283251009e-05, "loss": 0.0244, "step": 7850 }, { "grad_norm": 0.22679224610328674, "learning_rate": 9.936519402718632e-05, "loss": 0.0266, "step": 7860 }, { "grad_norm": 0.23549294471740723, "learning_rate": 9.936190678670102e-05, "loss": 0.0248, "step": 7870 }, { "grad_norm": 0.32267647981643677, "learning_rate": 9.935861111161593e-05, "loss": 0.03, "step": 7880 }, { "grad_norm": 0.28885260224342346, "learning_rate": 9.935530700249416e-05, "loss": 0.0249, "step": 7890 }, { "grad_norm": 0.26346561312675476, "learning_rate": 9.935199445990028e-05, "loss": 0.0269, "step": 7900 }, { "grad_norm": 0.3071261942386627, "learning_rate": 9.934867348440033e-05, "loss": 0.03, "step": 7910 }, { "grad_norm": 0.3374512791633606, "learning_rate": 9.934534407656176e-05, "loss": 0.0266, "step": 7920 }, { "grad_norm": 0.24586933851242065, "learning_rate": 9.93420062369535e-05, "loss": 0.0276, "step": 7930 }, { "grad_norm": 0.2350643426179886, "learning_rate": 9.933865996614589e-05, "loss": 0.0284, "step": 7940 }, { "grad_norm": 0.1695854216814041, "learning_rate": 9.933530526471068e-05, "loss": 0.0239, "step": 7950 }, { "grad_norm": 0.3198727071285248, "learning_rate": 9.933194213322114e-05, "loss": 0.026, "step": 7960 }, { "grad_norm": 0.3804194927215576, "learning_rate": 9.932857057225192e-05, "loss": 0.026, "step": 7970 }, { "grad_norm": 0.3200033903121948, "learning_rate": 9.932519058237912e-05, "loss": 0.0282, "step": 7980 }, { "grad_norm": 0.28231024742126465, "learning_rate": 9.932180216418032e-05, "loss": 0.0252, "step": 7990 }, { "grad_norm": 0.280191034078598, "learning_rate": 9.931840531823446e-05, "loss": 0.0235, "step": 8000 }, { "grad_norm": 0.3032490313053131, "learning_rate": 9.9315000045122e-05, "loss": 0.0272, "step": 8010 }, { "grad_norm": 0.27908647060394287, "learning_rate": 9.931158634542481e-05, "loss": 0.0304, "step": 8020 }, { "grad_norm": 0.21226905286312103, "learning_rate": 9.930816421972617e-05, "loss": 0.0254, "step": 8030 }, { "grad_norm": 0.2952476739883423, "learning_rate": 9.930473366861086e-05, "loss": 0.0283, "step": 8040 }, { "grad_norm": 0.2578623592853546, "learning_rate": 9.930129469266505e-05, "loss": 0.0245, "step": 8050 }, { "grad_norm": 0.2718468904495239, "learning_rate": 9.929784729247638e-05, "loss": 0.0239, "step": 8060 }, { "grad_norm": 0.2990884482860565, "learning_rate": 9.929439146863389e-05, "loss": 0.0257, "step": 8070 }, { "grad_norm": 0.26199761033058167, "learning_rate": 9.92909272217281e-05, "loss": 0.0269, "step": 8080 }, { "grad_norm": 0.30154699087142944, "learning_rate": 9.928745455235097e-05, "loss": 0.0237, "step": 8090 }, { "grad_norm": 0.3083197772502899, "learning_rate": 9.928397346109588e-05, "loss": 0.0257, "step": 8100 }, { "grad_norm": 0.21810069680213928, "learning_rate": 9.928048394855762e-05, "loss": 0.0286, "step": 8110 }, { "grad_norm": 0.2537144422531128, "learning_rate": 9.92769860153325e-05, "loss": 0.0272, "step": 8120 }, { "grad_norm": 0.2547464966773987, "learning_rate": 9.927347966201819e-05, "loss": 0.0256, "step": 8130 }, { "grad_norm": 0.48991501331329346, "learning_rate": 9.926996488921383e-05, "loss": 0.0282, "step": 8140 }, { "grad_norm": 0.2648126780986786, "learning_rate": 9.926644169752001e-05, "loss": 0.0269, "step": 8150 }, { "grad_norm": 0.24916383624076843, "learning_rate": 9.926291008753875e-05, "loss": 0.0272, "step": 8160 }, { "grad_norm": 0.25709614157676697, "learning_rate": 9.92593700598735e-05, "loss": 0.0271, "step": 8170 }, { "grad_norm": 0.34661492705345154, "learning_rate": 9.925582161512915e-05, "loss": 0.0283, "step": 8180 }, { "grad_norm": 0.28424912691116333, "learning_rate": 9.925226475391205e-05, "loss": 0.0267, "step": 8190 }, { "grad_norm": 0.2545764446258545, "learning_rate": 9.924869947682993e-05, "loss": 0.0232, "step": 8200 }, { "grad_norm": 0.31617555022239685, "learning_rate": 9.924512578449204e-05, "loss": 0.0254, "step": 8210 }, { "grad_norm": 0.2793187201023102, "learning_rate": 9.924154367750901e-05, "loss": 0.0235, "step": 8220 }, { "grad_norm": 0.3159618377685547, "learning_rate": 9.923795315649293e-05, "loss": 0.026, "step": 8230 }, { "grad_norm": 0.2319050431251526, "learning_rate": 9.92343542220573e-05, "loss": 0.0243, "step": 8240 }, { "grad_norm": 0.23076334595680237, "learning_rate": 9.92307468748171e-05, "loss": 0.0245, "step": 8250 }, { "grad_norm": 0.2562018036842346, "learning_rate": 9.922713111538873e-05, "loss": 0.0259, "step": 8260 }, { "grad_norm": 0.21568480134010315, "learning_rate": 9.922350694439003e-05, "loss": 0.0288, "step": 8270 }, { "grad_norm": 0.32604074478149414, "learning_rate": 9.921987436244024e-05, "loss": 0.0278, "step": 8280 }, { "grad_norm": 0.2643720805644989, "learning_rate": 9.921623337016008e-05, "loss": 0.0263, "step": 8290 }, { "grad_norm": 0.23866379261016846, "learning_rate": 9.921258396817172e-05, "loss": 0.0231, "step": 8300 }, { "grad_norm": 0.30316346883773804, "learning_rate": 9.920892615709874e-05, "loss": 0.027, "step": 8310 }, { "grad_norm": 0.2918998599052429, "learning_rate": 9.920525993756612e-05, "loss": 0.0277, "step": 8320 }, { "grad_norm": 0.33733484148979187, "learning_rate": 9.920158531020036e-05, "loss": 0.0252, "step": 8330 }, { "grad_norm": 0.3290717601776123, "learning_rate": 9.919790227562933e-05, "loss": 0.0238, "step": 8340 }, { "grad_norm": 0.25003930926322937, "learning_rate": 9.919421083448237e-05, "loss": 0.0243, "step": 8350 }, { "grad_norm": 0.2824684977531433, "learning_rate": 9.919051098739022e-05, "loss": 0.0251, "step": 8360 }, { "grad_norm": 0.2395760416984558, "learning_rate": 9.918680273498514e-05, "loss": 0.0258, "step": 8370 }, { "grad_norm": 0.30845609307289124, "learning_rate": 9.918308607790072e-05, "loss": 0.0254, "step": 8380 }, { "grad_norm": 0.2919270098209381, "learning_rate": 9.917936101677205e-05, "loss": 0.0251, "step": 8390 }, { "grad_norm": 0.2610682547092438, "learning_rate": 9.917562755223564e-05, "loss": 0.0243, "step": 8400 }, { "grad_norm": 0.29407235980033875, "learning_rate": 9.917188568492944e-05, "loss": 0.0248, "step": 8410 }, { "grad_norm": 0.2445041388273239, "learning_rate": 9.916813541549283e-05, "loss": 0.0251, "step": 8420 }, { "grad_norm": 0.34061282873153687, "learning_rate": 9.916437674456663e-05, "loss": 0.023, "step": 8430 }, { "grad_norm": 0.2743150293827057, "learning_rate": 9.916060967279308e-05, "loss": 0.0253, "step": 8440 }, { "grad_norm": 0.3331696689128876, "learning_rate": 9.91568342008159e-05, "loss": 0.0264, "step": 8450 }, { "grad_norm": 0.33407220244407654, "learning_rate": 9.915305032928019e-05, "loss": 0.0275, "step": 8460 }, { "grad_norm": 0.29055577516555786, "learning_rate": 9.914925805883253e-05, "loss": 0.0294, "step": 8470 }, { "grad_norm": 0.214312344789505, "learning_rate": 9.914545739012088e-05, "loss": 0.025, "step": 8480 }, { "grad_norm": 0.25376027822494507, "learning_rate": 9.91416483237947e-05, "loss": 0.0242, "step": 8490 }, { "grad_norm": 0.36510488390922546, "learning_rate": 9.913783086050485e-05, "loss": 0.0249, "step": 8500 }, { "grad_norm": 0.34537383913993835, "learning_rate": 9.913400500090364e-05, "loss": 0.0255, "step": 8510 }, { "grad_norm": 0.338864266872406, "learning_rate": 9.913017074564479e-05, "loss": 0.0268, "step": 8520 }, { "grad_norm": 0.23767967522144318, "learning_rate": 9.912632809538348e-05, "loss": 0.0288, "step": 8530 }, { "grad_norm": 0.41507768630981445, "learning_rate": 9.912247705077629e-05, "loss": 0.0284, "step": 8540 }, { "grad_norm": 0.28101661801338196, "learning_rate": 9.911861761248127e-05, "loss": 0.0264, "step": 8550 }, { "grad_norm": 0.26730823516845703, "learning_rate": 9.91147497811579e-05, "loss": 0.0274, "step": 8560 }, { "grad_norm": 0.3304431736469269, "learning_rate": 9.911087355746709e-05, "loss": 0.0231, "step": 8570 }, { "grad_norm": 0.26209837198257446, "learning_rate": 9.910698894207117e-05, "loss": 0.0272, "step": 8580 }, { "grad_norm": 0.2598782479763031, "learning_rate": 9.910309593563392e-05, "loss": 0.0241, "step": 8590 }, { "grad_norm": 0.2352330982685089, "learning_rate": 9.909919453882057e-05, "loss": 0.0227, "step": 8600 }, { "grad_norm": 0.3026887774467468, "learning_rate": 9.90952847522977e-05, "loss": 0.0246, "step": 8610 }, { "grad_norm": 0.310516357421875, "learning_rate": 9.909136657673346e-05, "loss": 0.0255, "step": 8620 }, { "grad_norm": 0.2757624685764313, "learning_rate": 9.908744001279731e-05, "loss": 0.0203, "step": 8630 }, { "grad_norm": 0.24725466966629028, "learning_rate": 9.90835050611602e-05, "loss": 0.023, "step": 8640 }, { "grad_norm": 0.2419414520263672, "learning_rate": 9.90795617224945e-05, "loss": 0.0255, "step": 8650 }, { "grad_norm": 0.21239009499549866, "learning_rate": 9.907560999747405e-05, "loss": 0.0268, "step": 8660 }, { "grad_norm": 0.24071963131427765, "learning_rate": 9.907164988677408e-05, "loss": 0.0274, "step": 8670 }, { "grad_norm": 0.2513822615146637, "learning_rate": 9.906768139107124e-05, "loss": 0.0259, "step": 8680 }, { "grad_norm": 0.2775571942329407, "learning_rate": 9.906370451104367e-05, "loss": 0.0242, "step": 8690 }, { "grad_norm": 0.27476558089256287, "learning_rate": 9.905971924737088e-05, "loss": 0.0215, "step": 8700 }, { "grad_norm": 0.29474565386772156, "learning_rate": 9.905572560073387e-05, "loss": 0.0233, "step": 8710 }, { "grad_norm": 0.36263975501060486, "learning_rate": 9.905172357181501e-05, "loss": 0.0219, "step": 8720 }, { "grad_norm": 0.4072614014148712, "learning_rate": 9.904771316129817e-05, "loss": 0.0264, "step": 8730 }, { "grad_norm": 0.24039621651172638, "learning_rate": 9.904369436986862e-05, "loss": 0.0236, "step": 8740 }, { "grad_norm": 0.3096477687358856, "learning_rate": 9.903966719821303e-05, "loss": 0.0272, "step": 8750 }, { "grad_norm": 0.2595920264720917, "learning_rate": 9.903563164701956e-05, "loss": 0.025, "step": 8760 }, { "grad_norm": 0.24862787127494812, "learning_rate": 9.903158771697778e-05, "loss": 0.0227, "step": 8770 }, { "grad_norm": 0.2633056640625, "learning_rate": 9.902753540877867e-05, "loss": 0.0231, "step": 8780 }, { "grad_norm": 0.3052651584148407, "learning_rate": 9.902347472311466e-05, "loss": 0.0231, "step": 8790 }, { "grad_norm": 0.21607303619384766, "learning_rate": 9.901940566067962e-05, "loss": 0.026, "step": 8800 }, { "grad_norm": 0.30814751982688904, "learning_rate": 9.901532822216883e-05, "loss": 0.025, "step": 8810 }, { "grad_norm": 0.268695592880249, "learning_rate": 9.901124240827904e-05, "loss": 0.0268, "step": 8820 }, { "grad_norm": 0.29058170318603516, "learning_rate": 9.900714821970835e-05, "loss": 0.0296, "step": 8830 }, { "grad_norm": 0.240178182721138, "learning_rate": 9.900304565715641e-05, "loss": 0.0283, "step": 8840 }, { "grad_norm": 0.29668721556663513, "learning_rate": 9.899893472132419e-05, "loss": 0.0294, "step": 8850 }, { "grad_norm": 0.2329792082309723, "learning_rate": 9.899481541291415e-05, "loss": 0.0253, "step": 8860 }, { "grad_norm": 0.26130080223083496, "learning_rate": 9.899068773263016e-05, "loss": 0.0263, "step": 8870 }, { "grad_norm": 0.2958475649356842, "learning_rate": 9.898655168117754e-05, "loss": 0.0222, "step": 8880 }, { "grad_norm": 0.3059578537940979, "learning_rate": 9.898240725926302e-05, "loss": 0.0267, "step": 8890 }, { "grad_norm": 0.27021804451942444, "learning_rate": 9.897825446759478e-05, "loss": 0.0203, "step": 8900 }, { "grad_norm": 0.21097029745578766, "learning_rate": 9.897409330688241e-05, "loss": 0.0234, "step": 8910 }, { "grad_norm": 0.22245940566062927, "learning_rate": 9.896992377783692e-05, "loss": 0.0219, "step": 8920 }, { "grad_norm": 0.2661912441253662, "learning_rate": 9.89657458811708e-05, "loss": 0.0226, "step": 8930 }, { "grad_norm": 0.33075660467147827, "learning_rate": 9.896155961759792e-05, "loss": 0.0227, "step": 8940 }, { "grad_norm": 0.3743683397769928, "learning_rate": 9.895736498783361e-05, "loss": 0.025, "step": 8950 }, { "grad_norm": 0.3052465617656708, "learning_rate": 9.895316199259462e-05, "loss": 0.0295, "step": 8960 }, { "grad_norm": 0.2352268248796463, "learning_rate": 9.894895063259909e-05, "loss": 0.0243, "step": 8970 }, { "grad_norm": 0.24770447611808777, "learning_rate": 9.894473090856667e-05, "loss": 0.0272, "step": 8980 }, { "grad_norm": 0.31868740916252136, "learning_rate": 9.894050282121839e-05, "loss": 0.0256, "step": 8990 }, { "grad_norm": 0.2137341946363449, "learning_rate": 9.893626637127668e-05, "loss": 0.0286, "step": 9000 }, { "grad_norm": 0.429611474275589, "learning_rate": 9.893202155946546e-05, "loss": 0.0272, "step": 9010 }, { "grad_norm": 0.31601178646087646, "learning_rate": 9.892776838651006e-05, "loss": 0.0219, "step": 9020 }, { "grad_norm": 0.28447648882865906, "learning_rate": 9.892350685313722e-05, "loss": 0.0282, "step": 9030 }, { "grad_norm": 0.28270047903060913, "learning_rate": 9.891923696007513e-05, "loss": 0.0276, "step": 9040 }, { "grad_norm": 0.23434831202030182, "learning_rate": 9.891495870805336e-05, "loss": 0.0238, "step": 9050 }, { "grad_norm": 0.24227569997310638, "learning_rate": 9.891067209780298e-05, "loss": 0.023, "step": 9060 }, { "grad_norm": 0.3394296169281006, "learning_rate": 9.890637713005646e-05, "loss": 0.0252, "step": 9070 }, { "grad_norm": 0.2490614801645279, "learning_rate": 9.890207380554767e-05, "loss": 0.026, "step": 9080 }, { "grad_norm": 0.3193151354789734, "learning_rate": 9.889776212501196e-05, "loss": 0.0254, "step": 9090 }, { "grad_norm": 0.27913698554039, "learning_rate": 9.889344208918605e-05, "loss": 0.0253, "step": 9100 }, { "grad_norm": 0.33102068305015564, "learning_rate": 9.888911369880812e-05, "loss": 0.0259, "step": 9110 }, { "grad_norm": 0.3185352087020874, "learning_rate": 9.888477695461777e-05, "loss": 0.0256, "step": 9120 }, { "grad_norm": 0.23022976517677307, "learning_rate": 9.888043185735607e-05, "loss": 0.0226, "step": 9130 }, { "grad_norm": 0.2743668556213379, "learning_rate": 9.887607840776542e-05, "loss": 0.024, "step": 9140 }, { "grad_norm": 0.25833383202552795, "learning_rate": 9.887171660658975e-05, "loss": 0.0211, "step": 9150 }, { "grad_norm": 0.2832155227661133, "learning_rate": 9.886734645457435e-05, "loss": 0.0248, "step": 9160 }, { "grad_norm": 0.28538742661476135, "learning_rate": 9.886296795246597e-05, "loss": 0.0261, "step": 9170 }, { "grad_norm": 0.3523280620574951, "learning_rate": 9.885858110101276e-05, "loss": 0.0234, "step": 9180 }, { "grad_norm": 0.2724619209766388, "learning_rate": 9.885418590096434e-05, "loss": 0.0243, "step": 9190 }, { "grad_norm": 0.23917672038078308, "learning_rate": 9.88497823530717e-05, "loss": 0.0246, "step": 9200 }, { "grad_norm": 0.28724125027656555, "learning_rate": 9.884537045808732e-05, "loss": 0.0249, "step": 9210 }, { "grad_norm": 0.2570827305316925, "learning_rate": 9.884095021676502e-05, "loss": 0.0266, "step": 9220 }, { "grad_norm": 0.2290259748697281, "learning_rate": 9.883652162986017e-05, "loss": 0.0246, "step": 9230 }, { "grad_norm": 0.26286888122558594, "learning_rate": 9.883208469812943e-05, "loss": 0.0259, "step": 9240 }, { "grad_norm": 0.24233274161815643, "learning_rate": 9.882763942233098e-05, "loss": 0.0223, "step": 9250 }, { "grad_norm": 0.28220298886299133, "learning_rate": 9.882318580322441e-05, "loss": 0.0235, "step": 9260 }, { "grad_norm": 0.2292555272579193, "learning_rate": 9.881872384157067e-05, "loss": 0.0237, "step": 9270 }, { "grad_norm": 0.3243931531906128, "learning_rate": 9.881425353813225e-05, "loss": 0.0247, "step": 9280 }, { "grad_norm": 0.26601341366767883, "learning_rate": 9.880977489367296e-05, "loss": 0.0231, "step": 9290 }, { "grad_norm": 0.2321728616952896, "learning_rate": 9.88052879089581e-05, "loss": 0.0252, "step": 9300 }, { "grad_norm": 0.2889944314956665, "learning_rate": 9.880079258475434e-05, "loss": 0.0217, "step": 9310 }, { "grad_norm": 0.32944631576538086, "learning_rate": 9.879628892182985e-05, "loss": 0.022, "step": 9320 }, { "grad_norm": 0.2711389362812042, "learning_rate": 9.879177692095416e-05, "loss": 0.0236, "step": 9330 }, { "grad_norm": 0.2711578905582428, "learning_rate": 9.878725658289825e-05, "loss": 0.0221, "step": 9340 }, { "grad_norm": 0.22206059098243713, "learning_rate": 9.878272790843454e-05, "loss": 0.0232, "step": 9350 }, { "grad_norm": 0.36544111371040344, "learning_rate": 9.877819089833682e-05, "loss": 0.0225, "step": 9360 }, { "grad_norm": 0.3333769142627716, "learning_rate": 9.877364555338038e-05, "loss": 0.0205, "step": 9370 }, { "grad_norm": 0.2839515209197998, "learning_rate": 9.876909187434186e-05, "loss": 0.0241, "step": 9380 }, { "grad_norm": 0.33025306463241577, "learning_rate": 9.876452986199939e-05, "loss": 0.0214, "step": 9390 }, { "grad_norm": 0.2807713747024536, "learning_rate": 9.875995951713248e-05, "loss": 0.0237, "step": 9400 }, { "grad_norm": 0.30168190598487854, "learning_rate": 9.875538084052207e-05, "loss": 0.0194, "step": 9410 }, { "grad_norm": 0.2349560260772705, "learning_rate": 9.875079383295053e-05, "loss": 0.0235, "step": 9420 }, { "grad_norm": 0.2764713764190674, "learning_rate": 9.874619849520167e-05, "loss": 0.0249, "step": 9430 }, { "grad_norm": 0.28886398673057556, "learning_rate": 9.874159482806069e-05, "loss": 0.021, "step": 9440 }, { "grad_norm": 0.24910812079906464, "learning_rate": 9.873698283231426e-05, "loss": 0.024, "step": 9450 }, { "grad_norm": 0.2765171527862549, "learning_rate": 9.87323625087504e-05, "loss": 0.0198, "step": 9460 }, { "grad_norm": 0.2401961237192154, "learning_rate": 9.872773385815863e-05, "loss": 0.0217, "step": 9470 }, { "grad_norm": 0.1781603991985321, "learning_rate": 9.872309688132986e-05, "loss": 0.0242, "step": 9480 }, { "grad_norm": 0.2552567720413208, "learning_rate": 9.871845157905639e-05, "loss": 0.0231, "step": 9490 }, { "grad_norm": 0.3151342570781708, "learning_rate": 9.871379795213201e-05, "loss": 0.0226, "step": 9500 }, { "grad_norm": 0.19547970592975616, "learning_rate": 9.87091360013519e-05, "loss": 0.0229, "step": 9510 }, { "grad_norm": 0.2735896706581116, "learning_rate": 9.870446572751262e-05, "loss": 0.0196, "step": 9520 }, { "grad_norm": 0.23493467271327972, "learning_rate": 9.869978713141224e-05, "loss": 0.0168, "step": 9530 }, { "grad_norm": 0.1945289820432663, "learning_rate": 9.869510021385016e-05, "loss": 0.0205, "step": 9540 }, { "grad_norm": 0.2845538854598999, "learning_rate": 9.869040497562727e-05, "loss": 0.0265, "step": 9550 }, { "grad_norm": 0.30696040391921997, "learning_rate": 9.868570141754587e-05, "loss": 0.0251, "step": 9560 }, { "grad_norm": 0.24865157902240753, "learning_rate": 9.868098954040965e-05, "loss": 0.0215, "step": 9570 }, { "grad_norm": 0.27224642038345337, "learning_rate": 9.867626934502374e-05, "loss": 0.0242, "step": 9580 }, { "grad_norm": 0.2913452386856079, "learning_rate": 9.86715408321947e-05, "loss": 0.0257, "step": 9590 }, { "grad_norm": 0.2980773448944092, "learning_rate": 9.86668040027305e-05, "loss": 0.0208, "step": 9600 }, { "grad_norm": 0.24255603551864624, "learning_rate": 9.866205885744053e-05, "loss": 0.0228, "step": 9610 }, { "grad_norm": 0.28803715109825134, "learning_rate": 9.865730539713563e-05, "loss": 0.0207, "step": 9620 }, { "grad_norm": 0.35573193430900574, "learning_rate": 9.8652543622628e-05, "loss": 0.0213, "step": 9630 }, { "grad_norm": 0.3282439112663269, "learning_rate": 9.864777353473132e-05, "loss": 0.0267, "step": 9640 }, { "grad_norm": 0.24854782223701477, "learning_rate": 9.864299513426068e-05, "loss": 0.0266, "step": 9650 }, { "grad_norm": 0.17916548252105713, "learning_rate": 9.863820842203254e-05, "loss": 0.0244, "step": 9660 }, { "grad_norm": 0.26183488965034485, "learning_rate": 9.863341339886483e-05, "loss": 0.0197, "step": 9670 }, { "grad_norm": 0.2783297896385193, "learning_rate": 9.86286100655769e-05, "loss": 0.0199, "step": 9680 }, { "grad_norm": 0.26491281390190125, "learning_rate": 9.862379842298953e-05, "loss": 0.0252, "step": 9690 }, { "grad_norm": 0.28414246439933777, "learning_rate": 9.861897847192485e-05, "loss": 0.0224, "step": 9700 }, { "grad_norm": 0.2297244518995285, "learning_rate": 9.86141502132065e-05, "loss": 0.0215, "step": 9710 }, { "grad_norm": 0.18601740896701813, "learning_rate": 9.860931364765946e-05, "loss": 0.0217, "step": 9720 }, { "grad_norm": 0.26643380522727966, "learning_rate": 9.860446877611021e-05, "loss": 0.0207, "step": 9730 }, { "grad_norm": 0.21844923496246338, "learning_rate": 9.859961559938655e-05, "loss": 0.022, "step": 9740 }, { "grad_norm": 0.3454916477203369, "learning_rate": 9.85947541183178e-05, "loss": 0.0236, "step": 9750 }, { "grad_norm": 0.24942243099212646, "learning_rate": 9.858988433373463e-05, "loss": 0.022, "step": 9760 }, { "grad_norm": 0.2719470262527466, "learning_rate": 9.858500624646918e-05, "loss": 0.0202, "step": 9770 }, { "grad_norm": 0.25101467967033386, "learning_rate": 9.858011985735497e-05, "loss": 0.0183, "step": 9780 }, { "grad_norm": 0.24541154503822327, "learning_rate": 9.857522516722693e-05, "loss": 0.0196, "step": 9790 }, { "grad_norm": 0.25851181149482727, "learning_rate": 9.857032217692145e-05, "loss": 0.0211, "step": 9800 }, { "grad_norm": 0.28646916151046753, "learning_rate": 9.856541088727631e-05, "loss": 0.0307, "step": 9810 }, { "grad_norm": 0.2808128297328949, "learning_rate": 9.856049129913072e-05, "loss": 0.0252, "step": 9820 }, { "grad_norm": 0.19208292663097382, "learning_rate": 9.85555634133253e-05, "loss": 0.0261, "step": 9830 }, { "grad_norm": 0.22904756665229797, "learning_rate": 9.855062723070208e-05, "loss": 0.0245, "step": 9840 }, { "grad_norm": 0.3469700515270233, "learning_rate": 9.854568275210454e-05, "loss": 0.0255, "step": 9850 }, { "grad_norm": 0.2743058502674103, "learning_rate": 9.854072997837754e-05, "loss": 0.0216, "step": 9860 }, { "grad_norm": 0.20936082303524017, "learning_rate": 9.853576891036737e-05, "loss": 0.0182, "step": 9870 }, { "grad_norm": 0.2674473524093628, "learning_rate": 9.853079954892177e-05, "loss": 0.0199, "step": 9880 }, { "grad_norm": 0.22296008467674255, "learning_rate": 9.852582189488983e-05, "loss": 0.0201, "step": 9890 }, { "grad_norm": 0.2653500735759735, "learning_rate": 9.852083594912212e-05, "loss": 0.0219, "step": 9900 }, { "grad_norm": 0.28917205333709717, "learning_rate": 9.851584171247058e-05, "loss": 0.0221, "step": 9910 }, { "grad_norm": 0.2129373997449875, "learning_rate": 9.851083918578863e-05, "loss": 0.0199, "step": 9920 }, { "grad_norm": 0.1836032271385193, "learning_rate": 9.850582836993103e-05, "loss": 0.0208, "step": 9930 }, { "grad_norm": 0.2520897090435028, "learning_rate": 9.850080926575397e-05, "loss": 0.0178, "step": 9940 }, { "grad_norm": 0.22127720713615417, "learning_rate": 9.849578187411515e-05, "loss": 0.0264, "step": 9950 }, { "grad_norm": 0.3581167757511139, "learning_rate": 9.849074619587354e-05, "loss": 0.0214, "step": 9960 }, { "grad_norm": 0.22527644038200378, "learning_rate": 9.848570223188964e-05, "loss": 0.0205, "step": 9970 }, { "grad_norm": 0.22000327706336975, "learning_rate": 9.848064998302531e-05, "loss": 0.0183, "step": 9980 }, { "grad_norm": 0.2742737829685211, "learning_rate": 9.847558945014386e-05, "loss": 0.0207, "step": 9990 }, { "grad_norm": 0.31411808729171753, "learning_rate": 9.847052063410996e-05, "loss": 0.0211, "step": 10000 }, { "grad_norm": 0.25457775592803955, "learning_rate": 9.846544353578977e-05, "loss": 0.0184, "step": 10010 }, { "grad_norm": 0.2662210166454315, "learning_rate": 9.846035815605081e-05, "loss": 0.0213, "step": 10020 }, { "grad_norm": 0.2701549828052521, "learning_rate": 9.845526449576204e-05, "loss": 0.0224, "step": 10030 }, { "grad_norm": 0.27346497774124146, "learning_rate": 9.845016255579383e-05, "loss": 0.0206, "step": 10040 }, { "grad_norm": 0.23092889785766602, "learning_rate": 9.844505233701794e-05, "loss": 0.0196, "step": 10050 }, { "grad_norm": 0.24691258370876312, "learning_rate": 9.843993384030757e-05, "loss": 0.022, "step": 10060 }, { "grad_norm": 0.20926912128925323, "learning_rate": 9.843480706653737e-05, "loss": 0.0198, "step": 10070 }, { "grad_norm": 0.2821975350379944, "learning_rate": 9.84296720165833e-05, "loss": 0.0277, "step": 10080 }, { "grad_norm": 0.29660147428512573, "learning_rate": 9.842452869132286e-05, "loss": 0.0209, "step": 10090 }, { "grad_norm": 0.2709963917732239, "learning_rate": 9.841937709163489e-05, "loss": 0.0212, "step": 10100 }, { "grad_norm": 0.2755948007106781, "learning_rate": 9.841421721839962e-05, "loss": 0.0176, "step": 10110 }, { "grad_norm": 0.30169475078582764, "learning_rate": 9.840904907249879e-05, "loss": 0.0206, "step": 10120 }, { "grad_norm": 0.26872122287750244, "learning_rate": 9.840387265481545e-05, "loss": 0.0234, "step": 10130 }, { "grad_norm": 0.23733337223529816, "learning_rate": 9.839868796623411e-05, "loss": 0.0208, "step": 10140 }, { "grad_norm": 0.2779839038848877, "learning_rate": 9.839349500764072e-05, "loss": 0.0197, "step": 10150 }, { "grad_norm": 0.2559874951839447, "learning_rate": 9.83882937799226e-05, "loss": 0.0224, "step": 10160 }, { "grad_norm": 0.20768356323242188, "learning_rate": 9.838308428396849e-05, "loss": 0.0224, "step": 10170 }, { "grad_norm": 0.24775773286819458, "learning_rate": 9.837786652066854e-05, "loss": 0.0221, "step": 10180 }, { "grad_norm": 0.2966967821121216, "learning_rate": 9.837264049091437e-05, "loss": 0.0218, "step": 10190 }, { "grad_norm": 0.363496869802475, "learning_rate": 9.836740619559893e-05, "loss": 0.0216, "step": 10200 }, { "grad_norm": 0.29772040247917175, "learning_rate": 9.836216363561659e-05, "loss": 0.0249, "step": 10210 }, { "grad_norm": 0.31085264682769775, "learning_rate": 9.835691281186322e-05, "loss": 0.0252, "step": 10220 }, { "grad_norm": 0.32387977838516235, "learning_rate": 9.8351653725236e-05, "loss": 0.0225, "step": 10230 }, { "grad_norm": 0.20584265887737274, "learning_rate": 9.83463863766336e-05, "loss": 0.02, "step": 10240 }, { "grad_norm": 0.18892347812652588, "learning_rate": 9.834111076695602e-05, "loss": 0.0179, "step": 10250 }, { "grad_norm": 0.2631441652774811, "learning_rate": 9.833582689710477e-05, "loss": 0.0245, "step": 10260 }, { "grad_norm": 0.21795813739299774, "learning_rate": 9.833053476798268e-05, "loss": 0.0229, "step": 10270 }, { "grad_norm": 0.2746921479701996, "learning_rate": 9.832523438049404e-05, "loss": 0.0227, "step": 10280 }, { "grad_norm": 0.3197917640209198, "learning_rate": 9.831992573554454e-05, "loss": 0.0194, "step": 10290 }, { "grad_norm": 0.19161240756511688, "learning_rate": 9.831460883404128e-05, "loss": 0.0213, "step": 10300 }, { "grad_norm": 0.2753503620624542, "learning_rate": 9.830928367689278e-05, "loss": 0.019, "step": 10310 }, { "grad_norm": 0.3212698698043823, "learning_rate": 9.830395026500896e-05, "loss": 0.0195, "step": 10320 }, { "grad_norm": 0.23608526587486267, "learning_rate": 9.829860859930115e-05, "loss": 0.023, "step": 10330 }, { "grad_norm": 0.20227892696857452, "learning_rate": 9.829325868068212e-05, "loss": 0.0198, "step": 10340 }, { "grad_norm": 0.19421106576919556, "learning_rate": 9.8287900510066e-05, "loss": 0.0245, "step": 10350 }, { "grad_norm": 0.23872700333595276, "learning_rate": 9.828253408836834e-05, "loss": 0.0226, "step": 10360 }, { "grad_norm": 0.26046982407569885, "learning_rate": 9.827715941650615e-05, "loss": 0.021, "step": 10370 }, { "grad_norm": 0.3336348235607147, "learning_rate": 9.82717764953978e-05, "loss": 0.0191, "step": 10380 }, { "grad_norm": 0.280892550945282, "learning_rate": 9.826638532596308e-05, "loss": 0.0193, "step": 10390 }, { "grad_norm": 0.19596044719219208, "learning_rate": 9.82609859091232e-05, "loss": 0.0203, "step": 10400 }, { "grad_norm": 0.27427351474761963, "learning_rate": 9.825557824580076e-05, "loss": 0.0198, "step": 10410 }, { "grad_norm": 0.2156284749507904, "learning_rate": 9.82501623369198e-05, "loss": 0.0183, "step": 10420 }, { "grad_norm": 0.24881967902183533, "learning_rate": 9.824473818340574e-05, "loss": 0.0213, "step": 10430 }, { "grad_norm": 0.24838240444660187, "learning_rate": 9.823930578618541e-05, "loss": 0.019, "step": 10440 }, { "grad_norm": 0.3010011315345764, "learning_rate": 9.823386514618709e-05, "loss": 0.0186, "step": 10450 }, { "grad_norm": 0.23860502243041992, "learning_rate": 9.82284162643404e-05, "loss": 0.0195, "step": 10460 }, { "grad_norm": 0.23005107045173645, "learning_rate": 9.822295914157642e-05, "loss": 0.0208, "step": 10470 }, { "grad_norm": 0.2035345882177353, "learning_rate": 9.821749377882763e-05, "loss": 0.0188, "step": 10480 }, { "grad_norm": 0.251631498336792, "learning_rate": 9.821202017702791e-05, "loss": 0.0176, "step": 10490 }, { "grad_norm": 0.29367268085479736, "learning_rate": 9.820653833711253e-05, "loss": 0.0188, "step": 10500 }, { "grad_norm": 0.2673150599002838, "learning_rate": 9.820104826001822e-05, "loss": 0.021, "step": 10510 }, { "grad_norm": 0.2669656276702881, "learning_rate": 9.819554994668305e-05, "loss": 0.0204, "step": 10520 }, { "grad_norm": 0.27234187722206116, "learning_rate": 9.819004339804654e-05, "loss": 0.0208, "step": 10530 }, { "grad_norm": 0.24865807592868805, "learning_rate": 9.818452861504961e-05, "loss": 0.0213, "step": 10540 }, { "grad_norm": 0.251247763633728, "learning_rate": 9.81790055986346e-05, "loss": 0.0219, "step": 10550 }, { "grad_norm": 0.20467518270015717, "learning_rate": 9.817347434974523e-05, "loss": 0.0191, "step": 10560 }, { "grad_norm": 0.24677444994449615, "learning_rate": 9.816793486932664e-05, "loss": 0.02, "step": 10570 }, { "grad_norm": 0.26145970821380615, "learning_rate": 9.816238715832538e-05, "loss": 0.0183, "step": 10580 }, { "grad_norm": 0.24970807135105133, "learning_rate": 9.815683121768939e-05, "loss": 0.022, "step": 10590 }, { "grad_norm": 0.3170432448387146, "learning_rate": 9.815126704836804e-05, "loss": 0.0218, "step": 10600 }, { "grad_norm": 0.20162895321846008, "learning_rate": 9.81456946513121e-05, "loss": 0.0228, "step": 10610 }, { "grad_norm": 0.3139369487762451, "learning_rate": 9.814011402747373e-05, "loss": 0.0251, "step": 10620 }, { "grad_norm": 0.27631875872612, "learning_rate": 9.813452517780651e-05, "loss": 0.0193, "step": 10630 }, { "grad_norm": 0.3597125709056854, "learning_rate": 9.81289281032654e-05, "loss": 0.0233, "step": 10640 }, { "grad_norm": 0.19995814561843872, "learning_rate": 9.812332280480683e-05, "loss": 0.0207, "step": 10650 }, { "grad_norm": 0.24928607046604156, "learning_rate": 9.811770928338854e-05, "loss": 0.02, "step": 10660 }, { "grad_norm": 0.2163933962583542, "learning_rate": 9.811208753996979e-05, "loss": 0.0185, "step": 10670 }, { "grad_norm": 0.21756291389465332, "learning_rate": 9.810645757551113e-05, "loss": 0.0164, "step": 10680 }, { "grad_norm": 0.1823148876428604, "learning_rate": 9.810081939097459e-05, "loss": 0.0182, "step": 10690 }, { "grad_norm": 0.21835565567016602, "learning_rate": 9.809517298732356e-05, "loss": 0.0201, "step": 10700 }, { "grad_norm": 0.23560181260108948, "learning_rate": 9.80895183655229e-05, "loss": 0.0168, "step": 10710 }, { "grad_norm": 0.30936095118522644, "learning_rate": 9.808385552653877e-05, "loss": 0.0206, "step": 10720 }, { "grad_norm": 0.2589513957500458, "learning_rate": 9.807818447133886e-05, "loss": 0.0211, "step": 10730 }, { "grad_norm": 0.2318219393491745, "learning_rate": 9.807250520089215e-05, "loss": 0.0187, "step": 10740 }, { "grad_norm": 0.24855804443359375, "learning_rate": 9.806681771616908e-05, "loss": 0.0181, "step": 10750 }, { "grad_norm": 0.2865772247314453, "learning_rate": 9.80611220181415e-05, "loss": 0.0193, "step": 10760 }, { "grad_norm": 0.18095450103282928, "learning_rate": 9.805541810778264e-05, "loss": 0.0212, "step": 10770 }, { "grad_norm": 0.2831066846847534, "learning_rate": 9.804970598606716e-05, "loss": 0.0183, "step": 10780 }, { "grad_norm": 0.31310126185417175, "learning_rate": 9.804398565397106e-05, "loss": 0.0227, "step": 10790 }, { "grad_norm": 0.2982061505317688, "learning_rate": 9.803825711247183e-05, "loss": 0.0208, "step": 10800 }, { "grad_norm": 0.22761954367160797, "learning_rate": 9.803252036254831e-05, "loss": 0.0162, "step": 10810 }, { "grad_norm": 0.24059349298477173, "learning_rate": 9.802677540518076e-05, "loss": 0.0192, "step": 10820 }, { "grad_norm": 0.24163252115249634, "learning_rate": 9.802102224135081e-05, "loss": 0.0212, "step": 10830 }, { "grad_norm": 0.32080215215682983, "learning_rate": 9.801526087204155e-05, "loss": 0.0192, "step": 10840 }, { "grad_norm": 0.2431286722421646, "learning_rate": 9.800949129823743e-05, "loss": 0.0186, "step": 10850 }, { "grad_norm": 0.21801026165485382, "learning_rate": 9.80037135209243e-05, "loss": 0.0203, "step": 10860 }, { "grad_norm": 0.19707898795604706, "learning_rate": 9.799792754108946e-05, "loss": 0.02, "step": 10870 }, { "grad_norm": 0.23175683617591858, "learning_rate": 9.799213335972152e-05, "loss": 0.018, "step": 10880 }, { "grad_norm": 0.32507050037384033, "learning_rate": 9.798633097781058e-05, "loss": 0.0217, "step": 10890 }, { "grad_norm": 0.19469788670539856, "learning_rate": 9.79805203963481e-05, "loss": 0.0184, "step": 10900 }, { "grad_norm": 0.23286043107509613, "learning_rate": 9.797470161632697e-05, "loss": 0.0203, "step": 10910 }, { "grad_norm": 0.19037559628486633, "learning_rate": 9.796887463874145e-05, "loss": 0.0216, "step": 10920 }, { "grad_norm": 0.29935377836227417, "learning_rate": 9.796303946458718e-05, "loss": 0.0199, "step": 10930 }, { "grad_norm": 0.27110710740089417, "learning_rate": 9.795719609486127e-05, "loss": 0.0202, "step": 10940 }, { "grad_norm": 0.20094704627990723, "learning_rate": 9.795134453056219e-05, "loss": 0.0166, "step": 10950 }, { "grad_norm": 0.31566593050956726, "learning_rate": 9.794548477268979e-05, "loss": 0.0167, "step": 10960 }, { "grad_norm": 0.22683922946453094, "learning_rate": 9.793961682224537e-05, "loss": 0.0167, "step": 10970 }, { "grad_norm": 0.2967931628227234, "learning_rate": 9.793374068023156e-05, "loss": 0.0193, "step": 10980 }, { "grad_norm": 0.24770449101924896, "learning_rate": 9.792785634765247e-05, "loss": 0.0219, "step": 10990 }, { "grad_norm": 0.20860859751701355, "learning_rate": 9.792196382551357e-05, "loss": 0.0189, "step": 11000 }, { "grad_norm": 0.2426098734140396, "learning_rate": 9.791606311482171e-05, "loss": 0.0167, "step": 11010 }, { "grad_norm": 0.17215624451637268, "learning_rate": 9.791015421658518e-05, "loss": 0.0189, "step": 11020 }, { "grad_norm": 0.2391616255044937, "learning_rate": 9.790423713181362e-05, "loss": 0.0208, "step": 11030 }, { "grad_norm": 0.2591468095779419, "learning_rate": 9.789831186151814e-05, "loss": 0.0185, "step": 11040 }, { "grad_norm": 0.24873405694961548, "learning_rate": 9.789237840671118e-05, "loss": 0.0201, "step": 11050 }, { "grad_norm": 0.22591346502304077, "learning_rate": 9.78864367684066e-05, "loss": 0.0181, "step": 11060 }, { "grad_norm": 0.19802020490169525, "learning_rate": 9.788048694761968e-05, "loss": 0.0189, "step": 11070 }, { "grad_norm": 0.21087238192558289, "learning_rate": 9.787452894536709e-05, "loss": 0.0198, "step": 11080 }, { "grad_norm": 0.2462492734193802, "learning_rate": 9.786856276266685e-05, "loss": 0.0169, "step": 11090 }, { "grad_norm": 0.26824137568473816, "learning_rate": 9.786258840053845e-05, "loss": 0.0191, "step": 11100 }, { "grad_norm": 0.26113706827163696, "learning_rate": 9.785660586000273e-05, "loss": 0.0185, "step": 11110 }, { "grad_norm": 0.24330054223537445, "learning_rate": 9.785061514208196e-05, "loss": 0.0178, "step": 11120 }, { "grad_norm": 0.17363354563713074, "learning_rate": 9.784461624779977e-05, "loss": 0.0191, "step": 11130 }, { "grad_norm": 0.23355449736118317, "learning_rate": 9.783860917818123e-05, "loss": 0.0194, "step": 11140 }, { "grad_norm": 0.2067982405424118, "learning_rate": 9.783259393425277e-05, "loss": 0.0205, "step": 11150 }, { "grad_norm": 0.20034822821617126, "learning_rate": 9.782657051704221e-05, "loss": 0.015, "step": 11160 }, { "grad_norm": 0.21181023120880127, "learning_rate": 9.782053892757883e-05, "loss": 0.0173, "step": 11170 }, { "grad_norm": 0.2412755936384201, "learning_rate": 9.781449916689324e-05, "loss": 0.0201, "step": 11180 }, { "grad_norm": 0.302108496427536, "learning_rate": 9.780845123601746e-05, "loss": 0.0178, "step": 11190 }, { "grad_norm": 0.2077556997537613, "learning_rate": 9.780239513598492e-05, "loss": 0.0183, "step": 11200 }, { "grad_norm": 0.25392580032348633, "learning_rate": 9.779633086783047e-05, "loss": 0.0215, "step": 11210 }, { "grad_norm": 0.2572821378707886, "learning_rate": 9.779025843259031e-05, "loss": 0.0181, "step": 11220 }, { "grad_norm": 0.22984322905540466, "learning_rate": 9.778417783130204e-05, "loss": 0.0176, "step": 11230 }, { "grad_norm": 0.3258267343044281, "learning_rate": 9.777808906500468e-05, "loss": 0.0173, "step": 11240 }, { "grad_norm": 0.29706907272338867, "learning_rate": 9.777199213473862e-05, "loss": 0.0197, "step": 11250 }, { "grad_norm": 0.20368540287017822, "learning_rate": 9.77658870415457e-05, "loss": 0.0214, "step": 11260 }, { "grad_norm": 0.28699803352355957, "learning_rate": 9.775977378646906e-05, "loss": 0.0236, "step": 11270 }, { "grad_norm": 0.15464606881141663, "learning_rate": 9.775365237055331e-05, "loss": 0.018, "step": 11280 }, { "grad_norm": 0.18403440713882446, "learning_rate": 9.774752279484445e-05, "loss": 0.0167, "step": 11290 }, { "grad_norm": 0.21173976361751556, "learning_rate": 9.774138506038984e-05, "loss": 0.018, "step": 11300 }, { "grad_norm": 0.36451637744903564, "learning_rate": 9.773523916823826e-05, "loss": 0.0232, "step": 11310 }, { "grad_norm": 0.24131257832050323, "learning_rate": 9.772908511943986e-05, "loss": 0.017, "step": 11320 }, { "grad_norm": 0.15370824933052063, "learning_rate": 9.77229229150462e-05, "loss": 0.0184, "step": 11330 }, { "grad_norm": 0.2703579366207123, "learning_rate": 9.771675255611024e-05, "loss": 0.0209, "step": 11340 }, { "grad_norm": 0.23680193722248077, "learning_rate": 9.771057404368632e-05, "loss": 0.0213, "step": 11350 }, { "grad_norm": 0.1818649172782898, "learning_rate": 9.770438737883018e-05, "loss": 0.0159, "step": 11360 }, { "grad_norm": 0.23410730063915253, "learning_rate": 9.769819256259898e-05, "loss": 0.0186, "step": 11370 }, { "grad_norm": 0.24684174358844757, "learning_rate": 9.769198959605119e-05, "loss": 0.0208, "step": 11380 }, { "grad_norm": 0.2125740647315979, "learning_rate": 9.768577848024678e-05, "loss": 0.0191, "step": 11390 }, { "grad_norm": 0.2055560052394867, "learning_rate": 9.767955921624702e-05, "loss": 0.0175, "step": 11400 }, { "grad_norm": 0.28393498063087463, "learning_rate": 9.767333180511465e-05, "loss": 0.0215, "step": 11410 }, { "grad_norm": 0.29249680042266846, "learning_rate": 9.766709624791373e-05, "loss": 0.0231, "step": 11420 }, { "grad_norm": 0.23336024582386017, "learning_rate": 9.766085254570975e-05, "loss": 0.0216, "step": 11430 }, { "grad_norm": 0.26228955388069153, "learning_rate": 9.76546006995696e-05, "loss": 0.0193, "step": 11440 }, { "grad_norm": 0.276697039604187, "learning_rate": 9.764834071056155e-05, "loss": 0.017, "step": 11450 }, { "grad_norm": 0.2436208575963974, "learning_rate": 9.764207257975526e-05, "loss": 0.0193, "step": 11460 }, { "grad_norm": 0.21613559126853943, "learning_rate": 9.763579630822179e-05, "loss": 0.0193, "step": 11470 }, { "grad_norm": 0.24003471434116364, "learning_rate": 9.762951189703356e-05, "loss": 0.0185, "step": 11480 }, { "grad_norm": 0.16236068308353424, "learning_rate": 9.762321934726442e-05, "loss": 0.017, "step": 11490 }, { "grad_norm": 0.22664526104927063, "learning_rate": 9.761691865998959e-05, "loss": 0.0216, "step": 11500 }, { "grad_norm": 0.19641518592834473, "learning_rate": 9.76106098362857e-05, "loss": 0.0184, "step": 11510 }, { "grad_norm": 0.22687800228595734, "learning_rate": 9.760429287723072e-05, "loss": 0.0196, "step": 11520 }, { "grad_norm": 0.2506125867366791, "learning_rate": 9.759796778390406e-05, "loss": 0.0187, "step": 11530 }, { "grad_norm": 0.3021716773509979, "learning_rate": 9.759163455738653e-05, "loss": 0.0214, "step": 11540 }, { "grad_norm": 0.25747621059417725, "learning_rate": 9.75852931987603e-05, "loss": 0.0165, "step": 11550 }, { "grad_norm": 0.20939120650291443, "learning_rate": 9.757894370910891e-05, "loss": 0.0217, "step": 11560 }, { "grad_norm": 0.31193143129348755, "learning_rate": 9.757258608951733e-05, "loss": 0.0216, "step": 11570 }, { "grad_norm": 0.2590436637401581, "learning_rate": 9.75662203410719e-05, "loss": 0.0175, "step": 11580 }, { "grad_norm": 0.24751071631908417, "learning_rate": 9.755984646486034e-05, "loss": 0.016, "step": 11590 }, { "grad_norm": 0.26308026909828186, "learning_rate": 9.75534644619718e-05, "loss": 0.0151, "step": 11600 }, { "grad_norm": 0.2311978042125702, "learning_rate": 9.754707433349676e-05, "loss": 0.0175, "step": 11610 }, { "grad_norm": 0.21577808260917664, "learning_rate": 9.754067608052715e-05, "loss": 0.0212, "step": 11620 }, { "grad_norm": 0.24116291105747223, "learning_rate": 9.753426970415622e-05, "loss": 0.0166, "step": 11630 }, { "grad_norm": 0.2259247601032257, "learning_rate": 9.752785520547868e-05, "loss": 0.0183, "step": 11640 }, { "grad_norm": 0.2909344434738159, "learning_rate": 9.752143258559056e-05, "loss": 0.0183, "step": 11650 }, { "grad_norm": 0.23327824473381042, "learning_rate": 9.751500184558933e-05, "loss": 0.0185, "step": 11660 }, { "grad_norm": 0.3358984589576721, "learning_rate": 9.750856298657383e-05, "loss": 0.0164, "step": 11670 }, { "grad_norm": 0.24789085984230042, "learning_rate": 9.750211600964428e-05, "loss": 0.0168, "step": 11680 }, { "grad_norm": 0.27223631739616394, "learning_rate": 9.749566091590226e-05, "loss": 0.0207, "step": 11690 }, { "grad_norm": 0.1782216876745224, "learning_rate": 9.748919770645083e-05, "loss": 0.021, "step": 11700 }, { "grad_norm": 0.2937195897102356, "learning_rate": 9.748272638239432e-05, "loss": 0.0185, "step": 11710 }, { "grad_norm": 0.2690843343734741, "learning_rate": 9.747624694483855e-05, "loss": 0.0214, "step": 11720 }, { "grad_norm": 0.21338774263858795, "learning_rate": 9.746975939489065e-05, "loss": 0.0152, "step": 11730 }, { "grad_norm": 0.23127780854701996, "learning_rate": 9.746326373365918e-05, "loss": 0.0161, "step": 11740 }, { "grad_norm": 0.30655670166015625, "learning_rate": 9.745675996225403e-05, "loss": 0.017, "step": 11750 }, { "grad_norm": 0.2714972198009491, "learning_rate": 9.745024808178657e-05, "loss": 0.0188, "step": 11760 }, { "grad_norm": 0.26539796590805054, "learning_rate": 9.744372809336947e-05, "loss": 0.0179, "step": 11770 }, { "grad_norm": 0.19609051942825317, "learning_rate": 9.743719999811682e-05, "loss": 0.022, "step": 11780 }, { "grad_norm": 0.2119135856628418, "learning_rate": 9.743066379714412e-05, "loss": 0.0171, "step": 11790 }, { "grad_norm": 0.2161322385072708, "learning_rate": 9.74241194915682e-05, "loss": 0.0219, "step": 11800 }, { "grad_norm": 0.20984263718128204, "learning_rate": 9.741756708250731e-05, "loss": 0.0181, "step": 11810 }, { "grad_norm": 0.2146335393190384, "learning_rate": 9.741100657108109e-05, "loss": 0.0214, "step": 11820 }, { "grad_norm": 0.23377656936645508, "learning_rate": 9.740443795841054e-05, "loss": 0.0193, "step": 11830 }, { "grad_norm": 0.2549448609352112, "learning_rate": 9.739786124561805e-05, "loss": 0.0184, "step": 11840 }, { "grad_norm": 0.1891292929649353, "learning_rate": 9.73912764338274e-05, "loss": 0.0206, "step": 11850 }, { "grad_norm": 0.2631802558898926, "learning_rate": 9.738468352416377e-05, "loss": 0.0215, "step": 11860 }, { "grad_norm": 0.21549108624458313, "learning_rate": 9.737808251775369e-05, "loss": 0.0225, "step": 11870 }, { "grad_norm": 0.24498775601387024, "learning_rate": 9.737147341572512e-05, "loss": 0.0185, "step": 11880 }, { "grad_norm": 0.15566003322601318, "learning_rate": 9.736485621920735e-05, "loss": 0.0188, "step": 11890 }, { "grad_norm": 0.2207133173942566, "learning_rate": 9.735823092933108e-05, "loss": 0.0269, "step": 11900 }, { "grad_norm": 0.21394386887550354, "learning_rate": 9.735159754722838e-05, "loss": 0.0168, "step": 11910 }, { "grad_norm": 0.34019774198532104, "learning_rate": 9.734495607403275e-05, "loss": 0.0191, "step": 11920 }, { "grad_norm": 0.24572747945785522, "learning_rate": 9.733830651087901e-05, "loss": 0.0208, "step": 11930 }, { "grad_norm": 0.24684973061084747, "learning_rate": 9.733164885890338e-05, "loss": 0.0169, "step": 11940 }, { "grad_norm": 0.2332763522863388, "learning_rate": 9.732498311924349e-05, "loss": 0.0226, "step": 11950 }, { "grad_norm": 0.19186647236347198, "learning_rate": 9.731830929303833e-05, "loss": 0.021, "step": 11960 }, { "grad_norm": 0.2521184980869293, "learning_rate": 9.731162738142827e-05, "loss": 0.0191, "step": 11970 }, { "grad_norm": 0.2190186232328415, "learning_rate": 9.730493738555506e-05, "loss": 0.0219, "step": 11980 }, { "grad_norm": 0.2570900619029999, "learning_rate": 9.729823930656186e-05, "loss": 0.0198, "step": 11990 }, { "grad_norm": 0.23784680664539337, "learning_rate": 9.729153314559316e-05, "loss": 0.0194, "step": 12000 }, { "grad_norm": 0.2848218083381653, "learning_rate": 9.728481890379486e-05, "loss": 0.0239, "step": 12010 }, { "grad_norm": 0.2184322327375412, "learning_rate": 9.727809658231428e-05, "loss": 0.0206, "step": 12020 }, { "grad_norm": 0.2308783382177353, "learning_rate": 9.727136618230003e-05, "loss": 0.0211, "step": 12030 }, { "grad_norm": 0.20286741852760315, "learning_rate": 9.726462770490219e-05, "loss": 0.0215, "step": 12040 }, { "grad_norm": 0.21771752834320068, "learning_rate": 9.725788115127214e-05, "loss": 0.0176, "step": 12050 }, { "grad_norm": 0.2841379642486572, "learning_rate": 9.725112652256274e-05, "loss": 0.0165, "step": 12060 }, { "grad_norm": 0.23236247897148132, "learning_rate": 9.724436381992812e-05, "loss": 0.0189, "step": 12070 }, { "grad_norm": 0.2598341107368469, "learning_rate": 9.723759304452387e-05, "loss": 0.018, "step": 12080 }, { "grad_norm": 0.22638548910617828, "learning_rate": 9.72308141975069e-05, "loss": 0.0174, "step": 12090 }, { "grad_norm": 0.23349322378635406, "learning_rate": 9.722402728003557e-05, "loss": 0.0191, "step": 12100 }, { "grad_norm": 0.24527104198932648, "learning_rate": 9.721723229326953e-05, "loss": 0.0175, "step": 12110 }, { "grad_norm": 0.2088807225227356, "learning_rate": 9.721042923836992e-05, "loss": 0.0206, "step": 12120 }, { "grad_norm": 0.19337712228298187, "learning_rate": 9.720361811649914e-05, "loss": 0.0173, "step": 12130 }, { "grad_norm": 0.2278745472431183, "learning_rate": 9.719679892882106e-05, "loss": 0.0144, "step": 12140 }, { "grad_norm": 0.2557029128074646, "learning_rate": 9.718997167650085e-05, "loss": 0.0166, "step": 12150 }, { "grad_norm": 0.2709615230560303, "learning_rate": 9.718313636070515e-05, "loss": 0.0212, "step": 12160 }, { "grad_norm": 0.2005307674407959, "learning_rate": 9.717629298260192e-05, "loss": 0.02, "step": 12170 }, { "grad_norm": 0.2316996157169342, "learning_rate": 9.716944154336047e-05, "loss": 0.0151, "step": 12180 }, { "grad_norm": 0.24413275718688965, "learning_rate": 9.716258204415157e-05, "loss": 0.0156, "step": 12190 }, { "grad_norm": 0.2638744115829468, "learning_rate": 9.715571448614728e-05, "loss": 0.0198, "step": 12200 }, { "grad_norm": 0.22032560408115387, "learning_rate": 9.71488388705211e-05, "loss": 0.0163, "step": 12210 }, { "grad_norm": 0.2596608102321625, "learning_rate": 9.714195519844788e-05, "loss": 0.0183, "step": 12220 }, { "grad_norm": 0.23607812821865082, "learning_rate": 9.713506347110386e-05, "loss": 0.0164, "step": 12230 }, { "grad_norm": 0.20091059803962708, "learning_rate": 9.712816368966662e-05, "loss": 0.0172, "step": 12240 }, { "grad_norm": 0.19193263351917267, "learning_rate": 9.712125585531517e-05, "loss": 0.019, "step": 12250 }, { "grad_norm": 0.2338545322418213, "learning_rate": 9.711433996922988e-05, "loss": 0.0178, "step": 12260 }, { "grad_norm": 0.2549067735671997, "learning_rate": 9.710741603259245e-05, "loss": 0.0212, "step": 12270 }, { "grad_norm": 0.18616418540477753, "learning_rate": 9.710048404658603e-05, "loss": 0.0183, "step": 12280 }, { "grad_norm": 0.21029910445213318, "learning_rate": 9.709354401239508e-05, "loss": 0.0173, "step": 12290 }, { "grad_norm": 0.2043541669845581, "learning_rate": 9.708659593120546e-05, "loss": 0.0195, "step": 12300 }, { "grad_norm": 0.22288228571414948, "learning_rate": 9.707963980420443e-05, "loss": 0.0202, "step": 12310 }, { "grad_norm": 0.22377672791481018, "learning_rate": 9.707267563258058e-05, "loss": 0.0151, "step": 12320 }, { "grad_norm": 0.2050267606973648, "learning_rate": 9.70657034175239e-05, "loss": 0.0186, "step": 12330 }, { "grad_norm": 0.24282202124595642, "learning_rate": 9.705872316022577e-05, "loss": 0.0231, "step": 12340 }, { "grad_norm": 0.19775891304016113, "learning_rate": 9.705173486187891e-05, "loss": 0.0185, "step": 12350 }, { "grad_norm": 0.3107026517391205, "learning_rate": 9.704473852367741e-05, "loss": 0.0233, "step": 12360 }, { "grad_norm": 0.21506814658641815, "learning_rate": 9.70377341468168e-05, "loss": 0.0195, "step": 12370 }, { "grad_norm": 0.26368188858032227, "learning_rate": 9.703072173249389e-05, "loss": 0.0183, "step": 12380 }, { "grad_norm": 0.2302297055721283, "learning_rate": 9.702370128190693e-05, "loss": 0.0174, "step": 12390 }, { "grad_norm": 0.23353789746761322, "learning_rate": 9.701667279625552e-05, "loss": 0.0198, "step": 12400 }, { "grad_norm": 0.24629352986812592, "learning_rate": 9.700963627674065e-05, "loss": 0.0166, "step": 12410 }, { "grad_norm": 0.1963203400373459, "learning_rate": 9.700259172456466e-05, "loss": 0.0169, "step": 12420 }, { "grad_norm": 0.15178027749061584, "learning_rate": 9.699553914093124e-05, "loss": 0.0143, "step": 12430 }, { "grad_norm": 0.21788498759269714, "learning_rate": 9.698847852704553e-05, "loss": 0.0157, "step": 12440 }, { "grad_norm": 0.2335553914308548, "learning_rate": 9.6981409884114e-05, "loss": 0.0246, "step": 12450 }, { "grad_norm": 0.26294779777526855, "learning_rate": 9.697433321334443e-05, "loss": 0.0175, "step": 12460 }, { "grad_norm": 0.27945592999458313, "learning_rate": 9.696724851594607e-05, "loss": 0.0217, "step": 12470 }, { "grad_norm": 0.2171793282032013, "learning_rate": 9.696015579312952e-05, "loss": 0.0161, "step": 12480 }, { "grad_norm": 0.23566308617591858, "learning_rate": 9.695305504610668e-05, "loss": 0.016, "step": 12490 }, { "grad_norm": 0.2201966494321823, "learning_rate": 9.694594627609092e-05, "loss": 0.0149, "step": 12500 }, { "grad_norm": 0.24564073979854584, "learning_rate": 9.693882948429691e-05, "loss": 0.0166, "step": 12510 }, { "grad_norm": 0.31598761677742004, "learning_rate": 9.693170467194071e-05, "loss": 0.0184, "step": 12520 }, { "grad_norm": 0.2479761391878128, "learning_rate": 9.692457184023977e-05, "loss": 0.0193, "step": 12530 }, { "grad_norm": 0.18146267533302307, "learning_rate": 9.691743099041291e-05, "loss": 0.0174, "step": 12540 }, { "grad_norm": 0.20328675210475922, "learning_rate": 9.691028212368027e-05, "loss": 0.0171, "step": 12550 }, { "grad_norm": 0.20155413448810577, "learning_rate": 9.690312524126342e-05, "loss": 0.0152, "step": 12560 }, { "grad_norm": 0.16934765875339508, "learning_rate": 9.689596034438527e-05, "loss": 0.016, "step": 12570 }, { "grad_norm": 0.19516955316066742, "learning_rate": 9.688878743427012e-05, "loss": 0.0157, "step": 12580 }, { "grad_norm": 0.17517024278640747, "learning_rate": 9.688160651214359e-05, "loss": 0.0168, "step": 12590 }, { "grad_norm": 0.18813739717006683, "learning_rate": 9.687441757923273e-05, "loss": 0.0175, "step": 12600 }, { "grad_norm": 0.22786124050617218, "learning_rate": 9.68672206367659e-05, "loss": 0.0174, "step": 12610 }, { "grad_norm": 0.2076239436864853, "learning_rate": 9.686001568597291e-05, "loss": 0.0203, "step": 12620 }, { "grad_norm": 0.2590121626853943, "learning_rate": 9.685280272808486e-05, "loss": 0.0177, "step": 12630 }, { "grad_norm": 0.294260710477829, "learning_rate": 9.684558176433424e-05, "loss": 0.019, "step": 12640 }, { "grad_norm": 0.22256337106227875, "learning_rate": 9.683835279595495e-05, "loss": 0.0178, "step": 12650 }, { "grad_norm": 0.23459775745868683, "learning_rate": 9.683111582418216e-05, "loss": 0.0139, "step": 12660 }, { "grad_norm": 0.20583802461624146, "learning_rate": 9.682387085025254e-05, "loss": 0.0139, "step": 12670 }, { "grad_norm": 0.1962287724018097, "learning_rate": 9.681661787540401e-05, "loss": 0.0165, "step": 12680 }, { "grad_norm": 0.18124492466449738, "learning_rate": 9.680935690087593e-05, "loss": 0.018, "step": 12690 }, { "grad_norm": 0.26906752586364746, "learning_rate": 9.680208792790901e-05, "loss": 0.0148, "step": 12700 }, { "grad_norm": 0.17397117614746094, "learning_rate": 9.679481095774529e-05, "loss": 0.0151, "step": 12710 }, { "grad_norm": 0.20076322555541992, "learning_rate": 9.678752599162822e-05, "loss": 0.0168, "step": 12720 }, { "grad_norm": 0.21285657584667206, "learning_rate": 9.678023303080259e-05, "loss": 0.0153, "step": 12730 }, { "grad_norm": 0.20520640909671783, "learning_rate": 9.677293207651459e-05, "loss": 0.0152, "step": 12740 }, { "grad_norm": 0.15296684205532074, "learning_rate": 9.676562313001173e-05, "loss": 0.0153, "step": 12750 }, { "grad_norm": 0.1760982871055603, "learning_rate": 9.675830619254293e-05, "loss": 0.0163, "step": 12760 }, { "grad_norm": 0.2528170645236969, "learning_rate": 9.675098126535843e-05, "loss": 0.0172, "step": 12770 }, { "grad_norm": 0.2755984663963318, "learning_rate": 9.674364834970988e-05, "loss": 0.0201, "step": 12780 }, { "grad_norm": 0.23145456612110138, "learning_rate": 9.673630744685028e-05, "loss": 0.0149, "step": 12790 }, { "grad_norm": 0.2552284002304077, "learning_rate": 9.672895855803397e-05, "loss": 0.0165, "step": 12800 }, { "grad_norm": 0.2438223958015442, "learning_rate": 9.672160168451667e-05, "loss": 0.0152, "step": 12810 }, { "grad_norm": 0.2698279321193695, "learning_rate": 9.671423682755549e-05, "loss": 0.0181, "step": 12820 }, { "grad_norm": 0.184822678565979, "learning_rate": 9.670686398840888e-05, "loss": 0.016, "step": 12830 }, { "grad_norm": 0.2210543155670166, "learning_rate": 9.669948316833664e-05, "loss": 0.0169, "step": 12840 }, { "grad_norm": 0.25965413451194763, "learning_rate": 9.669209436859997e-05, "loss": 0.0164, "step": 12850 }, { "grad_norm": 0.19715526700019836, "learning_rate": 9.66846975904614e-05, "loss": 0.0193, "step": 12860 }, { "grad_norm": 0.26969072222709656, "learning_rate": 9.667729283518483e-05, "loss": 0.0172, "step": 12870 }, { "grad_norm": 0.22404585778713226, "learning_rate": 9.666988010403557e-05, "loss": 0.0157, "step": 12880 }, { "grad_norm": 0.21284960210323334, "learning_rate": 9.66624593982802e-05, "loss": 0.0174, "step": 12890 }, { "grad_norm": 0.17549870908260345, "learning_rate": 9.665503071918675e-05, "loss": 0.0178, "step": 12900 }, { "grad_norm": 0.26669925451278687, "learning_rate": 9.664759406802456e-05, "loss": 0.0157, "step": 12910 }, { "grad_norm": 0.23461879789829254, "learning_rate": 9.664014944606437e-05, "loss": 0.0174, "step": 12920 }, { "grad_norm": 0.19523027539253235, "learning_rate": 9.663269685457822e-05, "loss": 0.0171, "step": 12930 }, { "grad_norm": 0.270212322473526, "learning_rate": 9.662523629483962e-05, "loss": 0.0179, "step": 12940 }, { "grad_norm": 0.2024407833814621, "learning_rate": 9.661776776812333e-05, "loss": 0.0132, "step": 12950 }, { "grad_norm": 0.2795463800430298, "learning_rate": 9.661029127570553e-05, "loss": 0.0176, "step": 12960 }, { "grad_norm": 0.27173709869384766, "learning_rate": 9.660280681886373e-05, "loss": 0.0161, "step": 12970 }, { "grad_norm": 0.2642972767353058, "learning_rate": 9.659531439887685e-05, "loss": 0.0186, "step": 12980 }, { "grad_norm": 0.2702217400074005, "learning_rate": 9.658781401702511e-05, "loss": 0.0235, "step": 12990 }, { "grad_norm": 0.16606765985488892, "learning_rate": 9.658030567459015e-05, "loss": 0.0154, "step": 13000 }, { "grad_norm": 0.1944122314453125, "learning_rate": 9.65727893728549e-05, "loss": 0.0183, "step": 13010 }, { "grad_norm": 0.22805163264274597, "learning_rate": 9.656526511310375e-05, "loss": 0.0166, "step": 13020 }, { "grad_norm": 0.16560675203800201, "learning_rate": 9.655773289662233e-05, "loss": 0.0162, "step": 13030 }, { "grad_norm": 0.2495838850736618, "learning_rate": 9.655019272469772e-05, "loss": 0.022, "step": 13040 }, { "grad_norm": 0.19928640127182007, "learning_rate": 9.654264459861832e-05, "loss": 0.0189, "step": 13050 }, { "grad_norm": 0.21007664501667023, "learning_rate": 9.653508851967391e-05, "loss": 0.0164, "step": 13060 }, { "grad_norm": 0.21982063353061676, "learning_rate": 9.65275244891556e-05, "loss": 0.0147, "step": 13070 }, { "grad_norm": 0.1934698224067688, "learning_rate": 9.651995250835591e-05, "loss": 0.0139, "step": 13080 }, { "grad_norm": 0.2871226370334625, "learning_rate": 9.651237257856862e-05, "loss": 0.0162, "step": 13090 }, { "grad_norm": 0.21497440338134766, "learning_rate": 9.6504784701089e-05, "loss": 0.0142, "step": 13100 }, { "grad_norm": 0.24172694981098175, "learning_rate": 9.649718887721357e-05, "loss": 0.0188, "step": 13110 }, { "grad_norm": 0.18929420411586761, "learning_rate": 9.648958510824028e-05, "loss": 0.0161, "step": 13120 }, { "grad_norm": 0.14485952258110046, "learning_rate": 9.648197339546837e-05, "loss": 0.0172, "step": 13130 }, { "grad_norm": 0.1769542694091797, "learning_rate": 9.647435374019851e-05, "loss": 0.0172, "step": 13140 }, { "grad_norm": 0.23840214312076569, "learning_rate": 9.646672614373266e-05, "loss": 0.0187, "step": 13150 }, { "grad_norm": 0.2101353257894516, "learning_rate": 9.645909060737418e-05, "loss": 0.016, "step": 13160 }, { "grad_norm": 0.27099448442459106, "learning_rate": 9.645144713242778e-05, "loss": 0.0198, "step": 13170 }, { "grad_norm": 0.3861684501171112, "learning_rate": 9.64437957201995e-05, "loss": 0.0172, "step": 13180 }, { "grad_norm": 0.2505514323711395, "learning_rate": 9.643613637199678e-05, "loss": 0.0185, "step": 13190 }, { "grad_norm": 0.1709376871585846, "learning_rate": 9.642846908912839e-05, "loss": 0.0175, "step": 13200 }, { "grad_norm": 0.20759426057338715, "learning_rate": 9.642079387290444e-05, "loss": 0.0158, "step": 13210 }, { "grad_norm": 0.19094319641590118, "learning_rate": 9.641311072463644e-05, "loss": 0.0184, "step": 13220 }, { "grad_norm": 0.2737497389316559, "learning_rate": 9.640541964563722e-05, "loss": 0.0186, "step": 13230 }, { "grad_norm": 0.2968781888484955, "learning_rate": 9.639772063722096e-05, "loss": 0.0191, "step": 13240 }, { "grad_norm": 0.1703050583600998, "learning_rate": 9.639001370070324e-05, "loss": 0.0177, "step": 13250 }, { "grad_norm": 0.26694026589393616, "learning_rate": 9.638229883740095e-05, "loss": 0.0171, "step": 13260 }, { "grad_norm": 0.15137510001659393, "learning_rate": 9.637457604863233e-05, "loss": 0.0157, "step": 13270 }, { "grad_norm": 0.2010483741760254, "learning_rate": 9.636684533571703e-05, "loss": 0.0167, "step": 13280 }, { "grad_norm": 0.20618289709091187, "learning_rate": 9.635910669997599e-05, "loss": 0.0128, "step": 13290 }, { "grad_norm": 0.22684600949287415, "learning_rate": 9.635136014273154e-05, "loss": 0.0168, "step": 13300 }, { "grad_norm": 0.24777112901210785, "learning_rate": 9.634360566530735e-05, "loss": 0.016, "step": 13310 }, { "grad_norm": 0.23307088017463684, "learning_rate": 9.633584326902845e-05, "loss": 0.0163, "step": 13320 }, { "grad_norm": 0.20759014785289764, "learning_rate": 9.632807295522124e-05, "loss": 0.0171, "step": 13330 }, { "grad_norm": 0.24985843896865845, "learning_rate": 9.632029472521342e-05, "loss": 0.0157, "step": 13340 }, { "grad_norm": 0.18707413971424103, "learning_rate": 9.631250858033409e-05, "loss": 0.0169, "step": 13350 }, { "grad_norm": 0.2065788209438324, "learning_rate": 9.630471452191371e-05, "loss": 0.0175, "step": 13360 }, { "grad_norm": 0.25150319933891296, "learning_rate": 9.629691255128405e-05, "loss": 0.0184, "step": 13370 }, { "grad_norm": 0.17933592200279236, "learning_rate": 9.628910266977825e-05, "loss": 0.0161, "step": 13380 }, { "grad_norm": 0.19932271540164948, "learning_rate": 9.628128487873083e-05, "loss": 0.0142, "step": 13390 }, { "grad_norm": 0.20389987528324127, "learning_rate": 9.627345917947761e-05, "loss": 0.0169, "step": 13400 }, { "grad_norm": 0.16544349491596222, "learning_rate": 9.626562557335579e-05, "loss": 0.0145, "step": 13410 }, { "grad_norm": 0.21174520254135132, "learning_rate": 9.625778406170393e-05, "loss": 0.0143, "step": 13420 }, { "grad_norm": 0.1846555471420288, "learning_rate": 9.624993464586193e-05, "loss": 0.0176, "step": 13430 }, { "grad_norm": 0.23791874945163727, "learning_rate": 9.624207732717105e-05, "loss": 0.0188, "step": 13440 }, { "grad_norm": 0.19176696240901947, "learning_rate": 9.623421210697386e-05, "loss": 0.0168, "step": 13450 }, { "grad_norm": 0.18734455108642578, "learning_rate": 9.622633898661434e-05, "loss": 0.0177, "step": 13460 }, { "grad_norm": 0.22047671675682068, "learning_rate": 9.621845796743778e-05, "loss": 0.0204, "step": 13470 }, { "grad_norm": 0.30438095331192017, "learning_rate": 9.621056905079082e-05, "loss": 0.0164, "step": 13480 }, { "grad_norm": 0.22318394482135773, "learning_rate": 9.620267223802149e-05, "loss": 0.0153, "step": 13490 }, { "grad_norm": 0.1950947493314743, "learning_rate": 9.619476753047911e-05, "loss": 0.0152, "step": 13500 }, { "grad_norm": 0.20401926338672638, "learning_rate": 9.618685492951438e-05, "loss": 0.0178, "step": 13510 }, { "grad_norm": 0.2219652682542801, "learning_rate": 9.617893443647938e-05, "loss": 0.0138, "step": 13520 }, { "grad_norm": 0.25310707092285156, "learning_rate": 9.617100605272746e-05, "loss": 0.0138, "step": 13530 }, { "grad_norm": 0.2261883020401001, "learning_rate": 9.616306977961338e-05, "loss": 0.0132, "step": 13540 }, { "grad_norm": 0.1948072910308838, "learning_rate": 9.615512561849326e-05, "loss": 0.0143, "step": 13550 }, { "grad_norm": 0.16016468405723572, "learning_rate": 9.61471735707245e-05, "loss": 0.0192, "step": 13560 }, { "grad_norm": 0.24007196724414825, "learning_rate": 9.613921363766592e-05, "loss": 0.0164, "step": 13570 }, { "grad_norm": 0.2264889031648636, "learning_rate": 9.613124582067763e-05, "loss": 0.0147, "step": 13580 }, { "grad_norm": 0.2072652131319046, "learning_rate": 9.612327012112112e-05, "loss": 0.0171, "step": 13590 }, { "grad_norm": 0.21765077114105225, "learning_rate": 9.611528654035921e-05, "loss": 0.0147, "step": 13600 }, { "grad_norm": 0.18195794522762299, "learning_rate": 9.610729507975611e-05, "loss": 0.0168, "step": 13610 }, { "grad_norm": 0.2185433954000473, "learning_rate": 9.609929574067731e-05, "loss": 0.0148, "step": 13620 }, { "grad_norm": 0.17825539410114288, "learning_rate": 9.609128852448967e-05, "loss": 0.0165, "step": 13630 }, { "grad_norm": 0.19283010065555573, "learning_rate": 9.608327343256143e-05, "loss": 0.0164, "step": 13640 }, { "grad_norm": 0.22244344651699066, "learning_rate": 9.607525046626216e-05, "loss": 0.0168, "step": 13650 }, { "grad_norm": 0.20195305347442627, "learning_rate": 9.606721962696272e-05, "loss": 0.0137, "step": 13660 }, { "grad_norm": 0.20519553124904633, "learning_rate": 9.60591809160354e-05, "loss": 0.0163, "step": 13670 }, { "grad_norm": 0.30077341198921204, "learning_rate": 9.605113433485378e-05, "loss": 0.0164, "step": 13680 }, { "grad_norm": 0.2296438366174698, "learning_rate": 9.604307988479279e-05, "loss": 0.0173, "step": 13690 }, { "grad_norm": 0.2209903746843338, "learning_rate": 9.603501756722876e-05, "loss": 0.0157, "step": 13700 }, { "grad_norm": 0.21444061398506165, "learning_rate": 9.602694738353927e-05, "loss": 0.0179, "step": 13710 }, { "grad_norm": 0.24035565555095673, "learning_rate": 9.601886933510331e-05, "loss": 0.0174, "step": 13720 }, { "grad_norm": 0.29391804337501526, "learning_rate": 9.60107834233012e-05, "loss": 0.0161, "step": 13730 }, { "grad_norm": 0.26144537329673767, "learning_rate": 9.60026896495146e-05, "loss": 0.0188, "step": 13740 }, { "grad_norm": 0.20588329434394836, "learning_rate": 9.599458801512652e-05, "loss": 0.0144, "step": 13750 }, { "grad_norm": 0.23460988700389862, "learning_rate": 9.598647852152129e-05, "loss": 0.0139, "step": 13760 }, { "grad_norm": 0.19748848676681519, "learning_rate": 9.597836117008462e-05, "loss": 0.0165, "step": 13770 }, { "grad_norm": 0.2136508822441101, "learning_rate": 9.597023596220356e-05, "loss": 0.0186, "step": 13780 }, { "grad_norm": 0.20651255548000336, "learning_rate": 9.596210289926643e-05, "loss": 0.015, "step": 13790 }, { "grad_norm": 0.20346438884735107, "learning_rate": 9.5953961982663e-05, "loss": 0.0165, "step": 13800 }, { "grad_norm": 0.2181222289800644, "learning_rate": 9.594581321378431e-05, "loss": 0.0144, "step": 13810 }, { "grad_norm": 0.2647300958633423, "learning_rate": 9.593765659402276e-05, "loss": 0.0161, "step": 13820 }, { "grad_norm": 0.24854819476604462, "learning_rate": 9.59294921247721e-05, "loss": 0.0138, "step": 13830 }, { "grad_norm": 0.2798340916633606, "learning_rate": 9.59213198074274e-05, "loss": 0.0147, "step": 13840 }, { "grad_norm": 0.240288645029068, "learning_rate": 9.59131396433851e-05, "loss": 0.0165, "step": 13850 }, { "grad_norm": 0.149499773979187, "learning_rate": 9.590495163404297e-05, "loss": 0.0178, "step": 13860 }, { "grad_norm": 0.24500694870948792, "learning_rate": 9.589675578080009e-05, "loss": 0.0161, "step": 13870 }, { "grad_norm": 0.18840570747852325, "learning_rate": 9.588855208505694e-05, "loss": 0.0161, "step": 13880 }, { "grad_norm": 0.17274071276187897, "learning_rate": 9.588034054821529e-05, "loss": 0.0154, "step": 13890 }, { "grad_norm": 0.23773428797721863, "learning_rate": 9.587212117167826e-05, "loss": 0.0176, "step": 13900 }, { "grad_norm": 0.30661019682884216, "learning_rate": 9.586389395685033e-05, "loss": 0.0136, "step": 13910 }, { "grad_norm": 0.25267305970191956, "learning_rate": 9.585565890513733e-05, "loss": 0.0139, "step": 13920 }, { "grad_norm": 0.2802104949951172, "learning_rate": 9.584741601794636e-05, "loss": 0.015, "step": 13930 }, { "grad_norm": 0.1682552546262741, "learning_rate": 9.58391652966859e-05, "loss": 0.0148, "step": 13940 }, { "grad_norm": 0.22949470579624176, "learning_rate": 9.583090674276583e-05, "loss": 0.0158, "step": 13950 }, { "grad_norm": 0.2965390086174011, "learning_rate": 9.582264035759726e-05, "loss": 0.0154, "step": 13960 }, { "grad_norm": 0.3363208472728729, "learning_rate": 9.58143661425927e-05, "loss": 0.0172, "step": 13970 }, { "grad_norm": 0.2185737043619156, "learning_rate": 9.580608409916601e-05, "loss": 0.0173, "step": 13980 }, { "grad_norm": 0.2641846835613251, "learning_rate": 9.579779422873233e-05, "loss": 0.0178, "step": 13990 }, { "grad_norm": 0.19930650293827057, "learning_rate": 9.578949653270819e-05, "loss": 0.022, "step": 14000 }, { "grad_norm": 0.2061692774295807, "learning_rate": 9.578119101251144e-05, "loss": 0.0139, "step": 14010 }, { "grad_norm": 0.2510366439819336, "learning_rate": 9.577287766956127e-05, "loss": 0.0156, "step": 14020 }, { "grad_norm": 0.3139680325984955, "learning_rate": 9.57645565052782e-05, "loss": 0.0146, "step": 14030 }, { "grad_norm": 0.2194836437702179, "learning_rate": 9.575622752108407e-05, "loss": 0.018, "step": 14040 }, { "grad_norm": 0.21642881631851196, "learning_rate": 9.57478907184021e-05, "loss": 0.0182, "step": 14050 }, { "grad_norm": 0.2174355387687683, "learning_rate": 9.573954609865681e-05, "loss": 0.0168, "step": 14060 }, { "grad_norm": 0.2727245092391968, "learning_rate": 9.573119366327408e-05, "loss": 0.0167, "step": 14070 }, { "grad_norm": 0.2756000757217407, "learning_rate": 9.57228334136811e-05, "loss": 0.0152, "step": 14080 }, { "grad_norm": 0.25948530435562134, "learning_rate": 9.571446535130641e-05, "loss": 0.0165, "step": 14090 }, { "grad_norm": 0.2280368208885193, "learning_rate": 9.570608947757988e-05, "loss": 0.0162, "step": 14100 }, { "grad_norm": 0.23621384799480438, "learning_rate": 9.569770579393274e-05, "loss": 0.0147, "step": 14110 }, { "grad_norm": 0.168595552444458, "learning_rate": 9.56893143017975e-05, "loss": 0.0159, "step": 14120 }, { "grad_norm": 0.26961126923561096, "learning_rate": 9.568091500260806e-05, "loss": 0.015, "step": 14130 }, { "grad_norm": 0.31040436029434204, "learning_rate": 9.567250789779961e-05, "loss": 0.0148, "step": 14140 }, { "grad_norm": 0.22717849910259247, "learning_rate": 9.566409298880872e-05, "loss": 0.0168, "step": 14150 }, { "grad_norm": 0.24022771418094635, "learning_rate": 9.565567027707326e-05, "loss": 0.0177, "step": 14160 }, { "grad_norm": 0.27063602209091187, "learning_rate": 9.56472397640324e-05, "loss": 0.0153, "step": 14170 }, { "grad_norm": 0.21138276159763336, "learning_rate": 9.563880145112675e-05, "loss": 0.0209, "step": 14180 }, { "grad_norm": 0.18731264770030975, "learning_rate": 9.563035533979814e-05, "loss": 0.0174, "step": 14190 }, { "grad_norm": 0.17954090237617493, "learning_rate": 9.562190143148981e-05, "loss": 0.019, "step": 14200 }, { "grad_norm": 0.16324183344841003, "learning_rate": 9.561343972764627e-05, "loss": 0.0131, "step": 14210 }, { "grad_norm": 0.2430698424577713, "learning_rate": 9.560497022971343e-05, "loss": 0.0151, "step": 14220 }, { "grad_norm": 0.2567923367023468, "learning_rate": 9.559649293913847e-05, "loss": 0.0153, "step": 14230 }, { "grad_norm": 0.23903347551822662, "learning_rate": 9.558800785736993e-05, "loss": 0.0178, "step": 14240 }, { "grad_norm": 0.18131917715072632, "learning_rate": 9.557951498585767e-05, "loss": 0.0171, "step": 14250 }, { "grad_norm": 0.2233499437570572, "learning_rate": 9.557101432605293e-05, "loss": 0.0149, "step": 14260 }, { "grad_norm": 0.2769075632095337, "learning_rate": 9.556250587940818e-05, "loss": 0.0157, "step": 14270 }, { "grad_norm": 0.17980851233005524, "learning_rate": 9.555398964737734e-05, "loss": 0.0165, "step": 14280 }, { "grad_norm": 0.2187618464231491, "learning_rate": 9.554546563141555e-05, "loss": 0.0164, "step": 14290 }, { "grad_norm": 0.24520431458950043, "learning_rate": 9.553693383297937e-05, "loss": 0.019, "step": 14300 }, { "grad_norm": 0.2781008780002594, "learning_rate": 9.552839425352663e-05, "loss": 0.0164, "step": 14310 }, { "grad_norm": 0.3100804090499878, "learning_rate": 9.551984689451652e-05, "loss": 0.0172, "step": 14320 }, { "grad_norm": 0.21813325583934784, "learning_rate": 9.551129175740953e-05, "loss": 0.0172, "step": 14330 }, { "grad_norm": 0.15607230365276337, "learning_rate": 9.550272884366754e-05, "loss": 0.0132, "step": 14340 }, { "grad_norm": 0.13316132128238678, "learning_rate": 9.549415815475369e-05, "loss": 0.0161, "step": 14350 }, { "grad_norm": 0.19185705482959747, "learning_rate": 9.548557969213247e-05, "loss": 0.0137, "step": 14360 }, { "grad_norm": 0.17517995834350586, "learning_rate": 9.547699345726972e-05, "loss": 0.0121, "step": 14370 }, { "grad_norm": 0.16909581422805786, "learning_rate": 9.546839945163257e-05, "loss": 0.0143, "step": 14380 }, { "grad_norm": 0.23495493829250336, "learning_rate": 9.545979767668953e-05, "loss": 0.0185, "step": 14390 }, { "grad_norm": 0.17915064096450806, "learning_rate": 9.54511881339104e-05, "loss": 0.0165, "step": 14400 }, { "grad_norm": 0.21922999620437622, "learning_rate": 9.54425708247663e-05, "loss": 0.0174, "step": 14410 }, { "grad_norm": 0.19437918066978455, "learning_rate": 9.543394575072972e-05, "loss": 0.0148, "step": 14420 }, { "grad_norm": 0.17015407979488373, "learning_rate": 9.542531291327441e-05, "loss": 0.0136, "step": 14430 }, { "grad_norm": 0.1874106377363205, "learning_rate": 9.541667231387552e-05, "loss": 0.014, "step": 14440 }, { "grad_norm": 0.15396079421043396, "learning_rate": 9.540802395400949e-05, "loss": 0.015, "step": 14450 }, { "grad_norm": 0.21720708906650543, "learning_rate": 9.539936783515406e-05, "loss": 0.0157, "step": 14460 }, { "grad_norm": 0.22563029825687408, "learning_rate": 9.539070395878835e-05, "loss": 0.017, "step": 14470 }, { "grad_norm": 0.271232008934021, "learning_rate": 9.538203232639277e-05, "loss": 0.0143, "step": 14480 }, { "grad_norm": 0.1767921894788742, "learning_rate": 9.537335293944907e-05, "loss": 0.0172, "step": 14490 }, { "grad_norm": 0.23264245688915253, "learning_rate": 9.536466579944032e-05, "loss": 0.015, "step": 14500 }, { "grad_norm": 0.17412585020065308, "learning_rate": 9.535597090785091e-05, "loss": 0.0171, "step": 14510 }, { "grad_norm": 0.20157469809055328, "learning_rate": 9.534726826616656e-05, "loss": 0.0141, "step": 14520 }, { "grad_norm": 0.2104669213294983, "learning_rate": 9.53385578758743e-05, "loss": 0.015, "step": 14530 }, { "grad_norm": 0.20775416493415833, "learning_rate": 9.532983973846252e-05, "loss": 0.0176, "step": 14540 }, { "grad_norm": 0.21362438797950745, "learning_rate": 9.53211138554209e-05, "loss": 0.0141, "step": 14550 }, { "grad_norm": 0.19457371532917023, "learning_rate": 9.531238022824047e-05, "loss": 0.0153, "step": 14560 }, { "grad_norm": 0.24912115931510925, "learning_rate": 9.530363885841355e-05, "loss": 0.0169, "step": 14570 }, { "grad_norm": 0.22012904286384583, "learning_rate": 9.52948897474338e-05, "loss": 0.0165, "step": 14580 }, { "grad_norm": 0.19237808883190155, "learning_rate": 9.528613289679622e-05, "loss": 0.0156, "step": 14590 }, { "grad_norm": 0.19322167336940765, "learning_rate": 9.52773683079971e-05, "loss": 0.0146, "step": 14600 }, { "grad_norm": 0.1895631104707718, "learning_rate": 9.526859598253407e-05, "loss": 0.0149, "step": 14610 }, { "grad_norm": 0.25511929392814636, "learning_rate": 9.525981592190609e-05, "loss": 0.0168, "step": 14620 }, { "grad_norm": 0.2641356587409973, "learning_rate": 9.525102812761342e-05, "loss": 0.0171, "step": 14630 }, { "grad_norm": 0.2364606410264969, "learning_rate": 9.524223260115768e-05, "loss": 0.0152, "step": 14640 }, { "grad_norm": 0.21393972635269165, "learning_rate": 9.523342934404175e-05, "loss": 0.0133, "step": 14650 }, { "grad_norm": 0.18665675818920135, "learning_rate": 9.522461835776989e-05, "loss": 0.0195, "step": 14660 }, { "grad_norm": 0.1894092857837677, "learning_rate": 9.521579964384764e-05, "loss": 0.015, "step": 14670 }, { "grad_norm": 0.15082916617393494, "learning_rate": 9.52069732037819e-05, "loss": 0.016, "step": 14680 }, { "grad_norm": 0.1565917283296585, "learning_rate": 9.519813903908083e-05, "loss": 0.0141, "step": 14690 }, { "grad_norm": 0.18193398416042328, "learning_rate": 9.5189297151254e-05, "loss": 0.014, "step": 14700 }, { "grad_norm": 0.16322758793830872, "learning_rate": 9.518044754181218e-05, "loss": 0.0122, "step": 14710 }, { "grad_norm": 0.24516046047210693, "learning_rate": 9.51715902122676e-05, "loss": 0.0154, "step": 14720 }, { "grad_norm": 0.18170224130153656, "learning_rate": 9.516272516413368e-05, "loss": 0.0148, "step": 14730 }, { "grad_norm": 0.21977677941322327, "learning_rate": 9.515385239892525e-05, "loss": 0.0193, "step": 14740 }, { "grad_norm": 0.19066490232944489, "learning_rate": 9.514497191815839e-05, "loss": 0.0149, "step": 14750 }, { "grad_norm": 0.1829857975244522, "learning_rate": 9.513608372335055e-05, "loss": 0.0138, "step": 14760 }, { "grad_norm": 0.17810222506523132, "learning_rate": 9.512718781602045e-05, "loss": 0.0146, "step": 14770 }, { "grad_norm": 0.18926112353801727, "learning_rate": 9.511828419768823e-05, "loss": 0.0136, "step": 14780 }, { "grad_norm": 0.21961013972759247, "learning_rate": 9.510937286987521e-05, "loss": 0.0151, "step": 14790 }, { "grad_norm": 0.2028304636478424, "learning_rate": 9.510045383410408e-05, "loss": 0.0138, "step": 14800 }, { "grad_norm": 0.21292144060134888, "learning_rate": 9.509152709189892e-05, "loss": 0.0137, "step": 14810 }, { "grad_norm": 0.20804092288017273, "learning_rate": 9.508259264478504e-05, "loss": 0.0159, "step": 14820 }, { "grad_norm": 0.17790494859218597, "learning_rate": 9.507365049428909e-05, "loss": 0.0157, "step": 14830 }, { "grad_norm": 0.29386353492736816, "learning_rate": 9.506470064193902e-05, "loss": 0.0167, "step": 14840 }, { "grad_norm": 0.2873755991458893, "learning_rate": 9.505574308926414e-05, "loss": 0.0155, "step": 14850 }, { "grad_norm": 0.2048511505126953, "learning_rate": 9.504677783779505e-05, "loss": 0.0125, "step": 14860 }, { "grad_norm": 0.15592798590660095, "learning_rate": 9.503780488906365e-05, "loss": 0.0135, "step": 14870 }, { "grad_norm": 0.1945038139820099, "learning_rate": 9.502882424460319e-05, "loss": 0.0163, "step": 14880 }, { "grad_norm": 0.15814238786697388, "learning_rate": 9.501983590594821e-05, "loss": 0.0155, "step": 14890 }, { "grad_norm": 0.1729760318994522, "learning_rate": 9.501083987463455e-05, "loss": 0.0151, "step": 14900 }, { "grad_norm": 0.22398172318935394, "learning_rate": 9.500183615219942e-05, "loss": 0.0142, "step": 14910 }, { "grad_norm": 0.21553607285022736, "learning_rate": 9.49928247401813e-05, "loss": 0.0181, "step": 14920 }, { "grad_norm": 0.22412176430225372, "learning_rate": 9.498380564011997e-05, "loss": 0.0161, "step": 14930 }, { "grad_norm": 0.25078025460243225, "learning_rate": 9.497477885355656e-05, "loss": 0.0176, "step": 14940 }, { "grad_norm": 0.19521233439445496, "learning_rate": 9.496574438203353e-05, "loss": 0.0133, "step": 14950 }, { "grad_norm": 0.21532173454761505, "learning_rate": 9.495670222709459e-05, "loss": 0.0131, "step": 14960 }, { "grad_norm": 0.22771164774894714, "learning_rate": 9.494765239028483e-05, "loss": 0.02, "step": 14970 }, { "grad_norm": 0.17312800884246826, "learning_rate": 9.493859487315057e-05, "loss": 0.0147, "step": 14980 }, { "grad_norm": 0.15833616256713867, "learning_rate": 9.492952967723953e-05, "loss": 0.0138, "step": 14990 }, { "grad_norm": 0.22707541286945343, "learning_rate": 9.492045680410068e-05, "loss": 0.0157, "step": 15000 }, { "grad_norm": 0.17792391777038574, "learning_rate": 9.491137625528436e-05, "loss": 0.0134, "step": 15010 }, { "grad_norm": 0.24224473536014557, "learning_rate": 9.490228803234215e-05, "loss": 0.016, "step": 15020 }, { "grad_norm": 0.2889551520347595, "learning_rate": 9.489319213682701e-05, "loss": 0.0138, "step": 15030 }, { "grad_norm": 0.1903514564037323, "learning_rate": 9.488408857029316e-05, "loss": 0.0122, "step": 15040 }, { "grad_norm": 0.18838542699813843, "learning_rate": 9.487497733429616e-05, "loss": 0.013, "step": 15050 }, { "grad_norm": 0.21955768764019012, "learning_rate": 9.486585843039286e-05, "loss": 0.012, "step": 15060 }, { "grad_norm": 0.20124860107898712, "learning_rate": 9.485673186014143e-05, "loss": 0.013, "step": 15070 }, { "grad_norm": 0.16698910295963287, "learning_rate": 9.484759762510137e-05, "loss": 0.0148, "step": 15080 }, { "grad_norm": 0.2219776213169098, "learning_rate": 9.483845572683346e-05, "loss": 0.0141, "step": 15090 }, { "grad_norm": 0.20417623221874237, "learning_rate": 9.48293061668998e-05, "loss": 0.0143, "step": 15100 }, { "grad_norm": 0.20680339634418488, "learning_rate": 9.48201489468638e-05, "loss": 0.0155, "step": 15110 }, { "grad_norm": 0.1396108716726303, "learning_rate": 9.481098406829016e-05, "loss": 0.0133, "step": 15120 }, { "grad_norm": 0.20173166692256927, "learning_rate": 9.480181153274495e-05, "loss": 0.015, "step": 15130 }, { "grad_norm": 0.23895032703876495, "learning_rate": 9.479263134179548e-05, "loss": 0.0189, "step": 15140 }, { "grad_norm": 0.1985754370689392, "learning_rate": 9.478344349701039e-05, "loss": 0.0183, "step": 15150 }, { "grad_norm": 0.17960470914840698, "learning_rate": 9.477424799995964e-05, "loss": 0.0175, "step": 15160 }, { "grad_norm": 0.22012966871261597, "learning_rate": 9.476504485221448e-05, "loss": 0.0171, "step": 15170 }, { "grad_norm": 0.189297154545784, "learning_rate": 9.475583405534748e-05, "loss": 0.0158, "step": 15180 }, { "grad_norm": 0.19875064492225647, "learning_rate": 9.474661561093251e-05, "loss": 0.0177, "step": 15190 }, { "grad_norm": 0.23701705038547516, "learning_rate": 9.473738952054478e-05, "loss": 0.0152, "step": 15200 }, { "grad_norm": 0.28980177640914917, "learning_rate": 9.472815578576073e-05, "loss": 0.0169, "step": 15210 }, { "grad_norm": 0.21594515442848206, "learning_rate": 9.471891440815817e-05, "loss": 0.0153, "step": 15220 }, { "grad_norm": 0.18157224357128143, "learning_rate": 9.470966538931621e-05, "loss": 0.0139, "step": 15230 }, { "grad_norm": 0.2092215120792389, "learning_rate": 9.470040873081525e-05, "loss": 0.0152, "step": 15240 }, { "grad_norm": 0.19965815544128418, "learning_rate": 9.469114443423698e-05, "loss": 0.0151, "step": 15250 }, { "grad_norm": 0.18764428794384003, "learning_rate": 9.468187250116445e-05, "loss": 0.0166, "step": 15260 }, { "grad_norm": 0.21896563470363617, "learning_rate": 9.467259293318197e-05, "loss": 0.0146, "step": 15270 }, { "grad_norm": 0.2820788323879242, "learning_rate": 9.466330573187514e-05, "loss": 0.0161, "step": 15280 }, { "grad_norm": 0.1909375786781311, "learning_rate": 9.46540108988309e-05, "loss": 0.0147, "step": 15290 }, { "grad_norm": 0.2037779688835144, "learning_rate": 9.46447084356375e-05, "loss": 0.0141, "step": 15300 }, { "grad_norm": 0.18636658787727356, "learning_rate": 9.463539834388447e-05, "loss": 0.0136, "step": 15310 }, { "grad_norm": 0.18482594192028046, "learning_rate": 9.462608062516263e-05, "loss": 0.0149, "step": 15320 }, { "grad_norm": 0.175292506814003, "learning_rate": 9.461675528106413e-05, "loss": 0.0135, "step": 15330 }, { "grad_norm": 0.24913290143013, "learning_rate": 9.460742231318244e-05, "loss": 0.0158, "step": 15340 }, { "grad_norm": 0.18674449622631073, "learning_rate": 9.459808172311229e-05, "loss": 0.0156, "step": 15350 }, { "grad_norm": 0.21122854948043823, "learning_rate": 9.458873351244972e-05, "loss": 0.0152, "step": 15360 }, { "grad_norm": 0.19809463620185852, "learning_rate": 9.457937768279211e-05, "loss": 0.0184, "step": 15370 }, { "grad_norm": 0.283115029335022, "learning_rate": 9.45700142357381e-05, "loss": 0.0181, "step": 15380 }, { "grad_norm": 0.20120052993297577, "learning_rate": 9.456064317288765e-05, "loss": 0.0162, "step": 15390 }, { "grad_norm": 0.23138999938964844, "learning_rate": 9.455126449584201e-05, "loss": 0.0166, "step": 15400 }, { "grad_norm": 0.21239376068115234, "learning_rate": 9.454187820620375e-05, "loss": 0.0156, "step": 15410 }, { "grad_norm": 0.21859639883041382, "learning_rate": 9.453248430557673e-05, "loss": 0.0152, "step": 15420 }, { "grad_norm": 0.1894707828760147, "learning_rate": 9.452308279556611e-05, "loss": 0.0158, "step": 15430 }, { "grad_norm": 0.18262776732444763, "learning_rate": 9.451367367777835e-05, "loss": 0.0174, "step": 15440 }, { "grad_norm": 0.22224976122379303, "learning_rate": 9.450425695382122e-05, "loss": 0.0142, "step": 15450 }, { "grad_norm": 0.16037744283676147, "learning_rate": 9.449483262530375e-05, "loss": 0.0149, "step": 15460 }, { "grad_norm": 0.17149855196475983, "learning_rate": 9.448540069383633e-05, "loss": 0.0168, "step": 15470 }, { "grad_norm": 0.2612302303314209, "learning_rate": 9.447596116103061e-05, "loss": 0.0169, "step": 15480 }, { "grad_norm": 0.22736352682113647, "learning_rate": 9.446651402849955e-05, "loss": 0.0226, "step": 15490 }, { "grad_norm": 0.22456373274326324, "learning_rate": 9.44570592978574e-05, "loss": 0.0184, "step": 15500 }, { "grad_norm": 0.24114446341991425, "learning_rate": 9.444759697071972e-05, "loss": 0.0162, "step": 15510 }, { "grad_norm": 0.20375441014766693, "learning_rate": 9.443812704870336e-05, "loss": 0.0157, "step": 15520 }, { "grad_norm": 0.17585515975952148, "learning_rate": 9.442864953342649e-05, "loss": 0.0156, "step": 15530 }, { "grad_norm": 0.1715581864118576, "learning_rate": 9.441916442650852e-05, "loss": 0.014, "step": 15540 }, { "grad_norm": 0.24562612175941467, "learning_rate": 9.440967172957023e-05, "loss": 0.0161, "step": 15550 }, { "grad_norm": 0.22953324019908905, "learning_rate": 9.440017144423364e-05, "loss": 0.0143, "step": 15560 }, { "grad_norm": 0.16200122237205505, "learning_rate": 9.439066357212209e-05, "loss": 0.0138, "step": 15570 }, { "grad_norm": 0.1936037689447403, "learning_rate": 9.438114811486022e-05, "loss": 0.012, "step": 15580 }, { "grad_norm": 0.2164386659860611, "learning_rate": 9.4371625074074e-05, "loss": 0.015, "step": 15590 }, { "grad_norm": 0.1690523326396942, "learning_rate": 9.436209445139059e-05, "loss": 0.0155, "step": 15600 }, { "grad_norm": 0.19450794160366058, "learning_rate": 9.435255624843855e-05, "loss": 0.0148, "step": 15610 }, { "grad_norm": 0.2043362408876419, "learning_rate": 9.43430104668477e-05, "loss": 0.0129, "step": 15620 }, { "grad_norm": 0.13433201611042023, "learning_rate": 9.433345710824914e-05, "loss": 0.0143, "step": 15630 }, { "grad_norm": 0.22227294743061066, "learning_rate": 9.432389617427529e-05, "loss": 0.015, "step": 15640 }, { "grad_norm": 0.25457146763801575, "learning_rate": 9.431432766655984e-05, "loss": 0.0154, "step": 15650 }, { "grad_norm": 0.2934326231479645, "learning_rate": 9.430475158673778e-05, "loss": 0.0174, "step": 15660 }, { "grad_norm": 0.27890437841415405, "learning_rate": 9.429516793644542e-05, "loss": 0.0152, "step": 15670 }, { "grad_norm": 0.15750284492969513, "learning_rate": 9.428557671732034e-05, "loss": 0.0153, "step": 15680 }, { "grad_norm": 0.16807059943675995, "learning_rate": 9.42759779310014e-05, "loss": 0.0134, "step": 15690 }, { "grad_norm": 0.1845616102218628, "learning_rate": 9.426637157912879e-05, "loss": 0.0124, "step": 15700 }, { "grad_norm": 0.20896375179290771, "learning_rate": 9.425675766334397e-05, "loss": 0.0147, "step": 15710 }, { "grad_norm": 0.218370258808136, "learning_rate": 9.424713618528968e-05, "loss": 0.0141, "step": 15720 }, { "grad_norm": 0.27523383498191833, "learning_rate": 9.423750714661e-05, "loss": 0.016, "step": 15730 }, { "grad_norm": 0.18659274280071259, "learning_rate": 9.422787054895022e-05, "loss": 0.0151, "step": 15740 }, { "grad_norm": 0.221063032746315, "learning_rate": 9.4218226393957e-05, "loss": 0.016, "step": 15750 }, { "grad_norm": 0.25589483976364136, "learning_rate": 9.420857468327828e-05, "loss": 0.0168, "step": 15760 }, { "grad_norm": 0.17920751869678497, "learning_rate": 9.419891541856323e-05, "loss": 0.0154, "step": 15770 }, { "grad_norm": 0.23211568593978882, "learning_rate": 9.41892486014624e-05, "loss": 0.0155, "step": 15780 }, { "grad_norm": 0.21368321776390076, "learning_rate": 9.417957423362756e-05, "loss": 0.0149, "step": 15790 }, { "grad_norm": 0.2551528811454773, "learning_rate": 9.416989231671178e-05, "loss": 0.0148, "step": 15800 }, { "grad_norm": 0.2671786844730377, "learning_rate": 9.416020285236946e-05, "loss": 0.0144, "step": 15810 }, { "grad_norm": 0.2014499306678772, "learning_rate": 9.415050584225626e-05, "loss": 0.0188, "step": 15820 }, { "grad_norm": 0.282433420419693, "learning_rate": 9.414080128802914e-05, "loss": 0.0136, "step": 15830 }, { "grad_norm": 0.20739005506038666, "learning_rate": 9.413108919134632e-05, "loss": 0.0189, "step": 15840 }, { "grad_norm": 0.21819499135017395, "learning_rate": 9.412136955386734e-05, "loss": 0.0156, "step": 15850 }, { "grad_norm": 0.23596498370170593, "learning_rate": 9.411164237725303e-05, "loss": 0.0157, "step": 15860 }, { "grad_norm": 0.2241153120994568, "learning_rate": 9.41019076631655e-05, "loss": 0.0138, "step": 15870 }, { "grad_norm": 0.2127571552991867, "learning_rate": 9.409216541326815e-05, "loss": 0.0172, "step": 15880 }, { "grad_norm": 0.17850786447525024, "learning_rate": 9.408241562922564e-05, "loss": 0.016, "step": 15890 }, { "grad_norm": 0.25764110684394836, "learning_rate": 9.407265831270395e-05, "loss": 0.0183, "step": 15900 }, { "grad_norm": 0.21994079649448395, "learning_rate": 9.406289346537035e-05, "loss": 0.0168, "step": 15910 }, { "grad_norm": 0.22193308174610138, "learning_rate": 9.405312108889339e-05, "loss": 0.0204, "step": 15920 }, { "grad_norm": 0.2323351949453354, "learning_rate": 9.404334118494288e-05, "loss": 0.0144, "step": 15930 }, { "grad_norm": 0.20963378250598907, "learning_rate": 9.403355375518995e-05, "loss": 0.0144, "step": 15940 }, { "grad_norm": 0.2274738848209381, "learning_rate": 9.4023758801307e-05, "loss": 0.0169, "step": 15950 }, { "grad_norm": 0.16944725811481476, "learning_rate": 9.401395632496774e-05, "loss": 0.0136, "step": 15960 }, { "grad_norm": 0.22099831700325012, "learning_rate": 9.400414632784711e-05, "loss": 0.0146, "step": 15970 }, { "grad_norm": 0.2635050415992737, "learning_rate": 9.39943288116214e-05, "loss": 0.0145, "step": 15980 }, { "grad_norm": 0.17473898828029633, "learning_rate": 9.398450377796815e-05, "loss": 0.0148, "step": 15990 }, { "grad_norm": 0.15729351341724396, "learning_rate": 9.397467122856616e-05, "loss": 0.0178, "step": 16000 }, { "grad_norm": 0.24251219630241394, "learning_rate": 9.396483116509558e-05, "loss": 0.0147, "step": 16010 }, { "grad_norm": 0.19058552384376526, "learning_rate": 9.39549835892378e-05, "loss": 0.0162, "step": 16020 }, { "grad_norm": 0.16026179492473602, "learning_rate": 9.39451285026755e-05, "loss": 0.0135, "step": 16030 }, { "grad_norm": 0.17643272876739502, "learning_rate": 9.393526590709262e-05, "loss": 0.0129, "step": 16040 }, { "grad_norm": 0.44451668858528137, "learning_rate": 9.392539580417444e-05, "loss": 0.0144, "step": 16050 }, { "grad_norm": 0.2260991781949997, "learning_rate": 9.391551819560747e-05, "loss": 0.0158, "step": 16060 }, { "grad_norm": 0.1883980929851532, "learning_rate": 9.390563308307955e-05, "loss": 0.013, "step": 16070 }, { "grad_norm": 0.19482481479644775, "learning_rate": 9.389574046827974e-05, "loss": 0.0163, "step": 16080 }, { "grad_norm": 0.22698457539081573, "learning_rate": 9.388584035289845e-05, "loss": 0.0139, "step": 16090 }, { "grad_norm": 0.20428594946861267, "learning_rate": 9.387593273862732e-05, "loss": 0.0142, "step": 16100 }, { "grad_norm": 0.20300623774528503, "learning_rate": 9.386601762715929e-05, "loss": 0.0143, "step": 16110 }, { "grad_norm": 0.1584065854549408, "learning_rate": 9.38560950201886e-05, "loss": 0.0144, "step": 16120 }, { "grad_norm": 0.24244751036167145, "learning_rate": 9.384616491941071e-05, "loss": 0.0149, "step": 16130 }, { "grad_norm": 0.19458402693271637, "learning_rate": 9.383622732652245e-05, "loss": 0.0145, "step": 16140 }, { "grad_norm": 0.19291695952415466, "learning_rate": 9.382628224322187e-05, "loss": 0.0141, "step": 16150 }, { "grad_norm": 0.2169528305530548, "learning_rate": 9.381632967120829e-05, "loss": 0.0155, "step": 16160 }, { "grad_norm": 0.20131489634513855, "learning_rate": 9.380636961218235e-05, "loss": 0.0137, "step": 16170 }, { "grad_norm": 0.16708463430404663, "learning_rate": 9.379640206784597e-05, "loss": 0.0148, "step": 16180 }, { "grad_norm": 0.17231105268001556, "learning_rate": 9.378642703990229e-05, "loss": 0.0143, "step": 16190 }, { "grad_norm": 0.20470812916755676, "learning_rate": 9.37764445300558e-05, "loss": 0.0142, "step": 16200 }, { "grad_norm": 0.202308788895607, "learning_rate": 9.376645454001222e-05, "loss": 0.0143, "step": 16210 }, { "grad_norm": 0.2071443498134613, "learning_rate": 9.375645707147858e-05, "loss": 0.0157, "step": 16220 }, { "grad_norm": 0.1626971811056137, "learning_rate": 9.374645212616316e-05, "loss": 0.0143, "step": 16230 }, { "grad_norm": 0.25165072083473206, "learning_rate": 9.373643970577555e-05, "loss": 0.0165, "step": 16240 }, { "grad_norm": 0.18346302211284637, "learning_rate": 9.372641981202659e-05, "loss": 0.0121, "step": 16250 }, { "grad_norm": 0.15054887533187866, "learning_rate": 9.37163924466284e-05, "loss": 0.0168, "step": 16260 }, { "grad_norm": 0.1829509288072586, "learning_rate": 9.370635761129438e-05, "loss": 0.014, "step": 16270 }, { "grad_norm": 0.2632943093776703, "learning_rate": 9.36963153077392e-05, "loss": 0.0175, "step": 16280 }, { "grad_norm": 0.21063534915447235, "learning_rate": 9.368626553767888e-05, "loss": 0.0128, "step": 16290 }, { "grad_norm": 0.23867347836494446, "learning_rate": 9.367620830283057e-05, "loss": 0.0149, "step": 16300 }, { "grad_norm": 0.20908594131469727, "learning_rate": 9.366614360491281e-05, "loss": 0.0158, "step": 16310 }, { "grad_norm": 0.1716577708721161, "learning_rate": 9.365607144564539e-05, "loss": 0.0144, "step": 16320 }, { "grad_norm": 0.19288307428359985, "learning_rate": 9.364599182674934e-05, "loss": 0.0125, "step": 16330 }, { "grad_norm": 0.22355541586875916, "learning_rate": 9.3635904749947e-05, "loss": 0.0117, "step": 16340 }, { "grad_norm": 0.18235385417938232, "learning_rate": 9.362581021696202e-05, "loss": 0.0153, "step": 16350 }, { "grad_norm": 0.1709153950214386, "learning_rate": 9.361570822951921e-05, "loss": 0.0157, "step": 16360 }, { "grad_norm": 0.16989582777023315, "learning_rate": 9.360559878934476e-05, "loss": 0.0115, "step": 16370 }, { "grad_norm": 0.19979673624038696, "learning_rate": 9.359548189816611e-05, "loss": 0.0117, "step": 16380 }, { "grad_norm": 0.2721458971500397, "learning_rate": 9.358535755771193e-05, "loss": 0.0142, "step": 16390 }, { "grad_norm": 0.1853140890598297, "learning_rate": 9.357522576971221e-05, "loss": 0.0168, "step": 16400 }, { "grad_norm": 0.27469924092292786, "learning_rate": 9.356508653589819e-05, "loss": 0.0159, "step": 16410 }, { "grad_norm": 0.18201424181461334, "learning_rate": 9.355493985800237e-05, "loss": 0.0164, "step": 16420 }, { "grad_norm": 0.2516048550605774, "learning_rate": 9.354478573775857e-05, "loss": 0.0128, "step": 16430 }, { "grad_norm": 0.21669533848762512, "learning_rate": 9.353462417690186e-05, "loss": 0.0125, "step": 16440 }, { "grad_norm": 0.17233873903751373, "learning_rate": 9.352445517716853e-05, "loss": 0.0137, "step": 16450 }, { "grad_norm": 0.23146270215511322, "learning_rate": 9.351427874029621e-05, "loss": 0.0161, "step": 16460 }, { "grad_norm": 0.15821228921413422, "learning_rate": 9.350409486802379e-05, "loss": 0.0138, "step": 16470 }, { "grad_norm": 0.28589949011802673, "learning_rate": 9.349390356209138e-05, "loss": 0.0147, "step": 16480 }, { "grad_norm": 0.15311972796916962, "learning_rate": 9.348370482424042e-05, "loss": 0.0128, "step": 16490 }, { "grad_norm": 0.14729881286621094, "learning_rate": 9.347349865621357e-05, "loss": 0.0134, "step": 16500 }, { "grad_norm": 0.22944141924381256, "learning_rate": 9.346328505975481e-05, "loss": 0.0112, "step": 16510 }, { "grad_norm": 0.1315709948539734, "learning_rate": 9.345306403660936e-05, "loss": 0.0129, "step": 16520 }, { "grad_norm": 0.17331179976463318, "learning_rate": 9.344283558852371e-05, "loss": 0.014, "step": 16530 }, { "grad_norm": 0.17075039446353912, "learning_rate": 9.343259971724563e-05, "loss": 0.0125, "step": 16540 }, { "grad_norm": 0.17244505882263184, "learning_rate": 9.342235642452413e-05, "loss": 0.0106, "step": 16550 }, { "grad_norm": 0.17025811970233917, "learning_rate": 9.341210571210954e-05, "loss": 0.0139, "step": 16560 }, { "grad_norm": 0.19089414179325104, "learning_rate": 9.340184758175338e-05, "loss": 0.016, "step": 16570 }, { "grad_norm": 0.17980900406837463, "learning_rate": 9.339158203520854e-05, "loss": 0.0125, "step": 16580 }, { "grad_norm": 0.1702941656112671, "learning_rate": 9.338130907422908e-05, "loss": 0.0131, "step": 16590 }, { "grad_norm": 0.2055627703666687, "learning_rate": 9.337102870057037e-05, "loss": 0.0127, "step": 16600 }, { "grad_norm": 0.1566259264945984, "learning_rate": 9.336074091598907e-05, "loss": 0.0118, "step": 16610 }, { "grad_norm": 0.18759667873382568, "learning_rate": 9.335044572224306e-05, "loss": 0.0141, "step": 16620 }, { "grad_norm": 0.17636196315288544, "learning_rate": 9.334014312109151e-05, "loss": 0.0163, "step": 16630 }, { "grad_norm": 0.20377306640148163, "learning_rate": 9.332983311429486e-05, "loss": 0.0145, "step": 16640 }, { "grad_norm": 0.17281439900398254, "learning_rate": 9.33195157036148e-05, "loss": 0.0195, "step": 16650 }, { "grad_norm": 0.2420060783624649, "learning_rate": 9.330919089081432e-05, "loss": 0.0129, "step": 16660 }, { "grad_norm": 0.2264830321073532, "learning_rate": 9.32988586776576e-05, "loss": 0.0161, "step": 16670 }, { "grad_norm": 0.17585420608520508, "learning_rate": 9.328851906591016e-05, "loss": 0.013, "step": 16680 }, { "grad_norm": 0.16332072019577026, "learning_rate": 9.327817205733875e-05, "loss": 0.0121, "step": 16690 }, { "grad_norm": 0.13611899316310883, "learning_rate": 9.326781765371142e-05, "loss": 0.0131, "step": 16700 }, { "grad_norm": 0.13810484111309052, "learning_rate": 9.325745585679741e-05, "loss": 0.0139, "step": 16710 }, { "grad_norm": 0.18998828530311584, "learning_rate": 9.32470866683673e-05, "loss": 0.0124, "step": 16720 }, { "grad_norm": 0.20481348037719727, "learning_rate": 9.323671009019288e-05, "loss": 0.0132, "step": 16730 }, { "grad_norm": 0.24451670050621033, "learning_rate": 9.322632612404725e-05, "loss": 0.0166, "step": 16740 }, { "grad_norm": 0.19329074025154114, "learning_rate": 9.321593477170471e-05, "loss": 0.0142, "step": 16750 }, { "grad_norm": 0.2028050273656845, "learning_rate": 9.320553603494088e-05, "loss": 0.0138, "step": 16760 }, { "grad_norm": 0.2445106953382492, "learning_rate": 9.319512991553261e-05, "loss": 0.0158, "step": 16770 }, { "grad_norm": 0.26800277829170227, "learning_rate": 9.318471641525803e-05, "loss": 0.0148, "step": 16780 }, { "grad_norm": 0.21553301811218262, "learning_rate": 9.317429553589652e-05, "loss": 0.0128, "step": 16790 }, { "grad_norm": 0.2209736853837967, "learning_rate": 9.316386727922873e-05, "loss": 0.0141, "step": 16800 }, { "grad_norm": 0.2248503714799881, "learning_rate": 9.315343164703656e-05, "loss": 0.0103, "step": 16810 }, { "grad_norm": 0.2546813189983368, "learning_rate": 9.314298864110316e-05, "loss": 0.0129, "step": 16820 }, { "grad_norm": 0.22441382706165314, "learning_rate": 9.313253826321295e-05, "loss": 0.0141, "step": 16830 }, { "grad_norm": 0.23682861030101776, "learning_rate": 9.312208051515165e-05, "loss": 0.0128, "step": 16840 }, { "grad_norm": 0.19214722514152527, "learning_rate": 9.311161539870618e-05, "loss": 0.0157, "step": 16850 }, { "grad_norm": 0.2195443958044052, "learning_rate": 9.310114291566474e-05, "loss": 0.0131, "step": 16860 }, { "grad_norm": 0.2332598865032196, "learning_rate": 9.309066306781679e-05, "loss": 0.0143, "step": 16870 }, { "grad_norm": 0.17662639915943146, "learning_rate": 9.308017585695306e-05, "loss": 0.0143, "step": 16880 }, { "grad_norm": 0.18776196241378784, "learning_rate": 9.306968128486552e-05, "loss": 0.0128, "step": 16890 }, { "grad_norm": 0.1858181655406952, "learning_rate": 9.30591793533474e-05, "loss": 0.0148, "step": 16900 }, { "grad_norm": 0.17203634977340698, "learning_rate": 9.304867006419321e-05, "loss": 0.0125, "step": 16910 }, { "grad_norm": 0.2152835875749588, "learning_rate": 9.303815341919868e-05, "loss": 0.0161, "step": 16920 }, { "grad_norm": 0.17682258784770966, "learning_rate": 9.302762942016084e-05, "loss": 0.0116, "step": 16930 }, { "grad_norm": 0.19837455451488495, "learning_rate": 9.301709806887792e-05, "loss": 0.0127, "step": 16940 }, { "grad_norm": 0.2138434201478958, "learning_rate": 9.300655936714948e-05, "loss": 0.015, "step": 16950 }, { "grad_norm": 0.17846663296222687, "learning_rate": 9.299601331677627e-05, "loss": 0.0156, "step": 16960 }, { "grad_norm": 0.22215530276298523, "learning_rate": 9.298545991956033e-05, "loss": 0.0151, "step": 16970 }, { "grad_norm": 0.24003225564956665, "learning_rate": 9.297489917730493e-05, "loss": 0.0141, "step": 16980 }, { "grad_norm": 0.20125454664230347, "learning_rate": 9.296433109181464e-05, "loss": 0.0124, "step": 16990 }, { "grad_norm": 0.15877866744995117, "learning_rate": 9.295375566489523e-05, "loss": 0.0143, "step": 17000 }, { "grad_norm": 0.24383875727653503, "learning_rate": 9.294317289835379e-05, "loss": 0.0176, "step": 17010 }, { "grad_norm": 0.2392447590827942, "learning_rate": 9.293258279399859e-05, "loss": 0.0159, "step": 17020 }, { "grad_norm": 0.25246503949165344, "learning_rate": 9.292198535363919e-05, "loss": 0.014, "step": 17030 }, { "grad_norm": 0.24258491396903992, "learning_rate": 9.291138057908641e-05, "loss": 0.0158, "step": 17040 }, { "grad_norm": 0.24967226386070251, "learning_rate": 9.290076847215234e-05, "loss": 0.0163, "step": 17050 }, { "grad_norm": 0.2210652232170105, "learning_rate": 9.289014903465025e-05, "loss": 0.0153, "step": 17060 }, { "grad_norm": 0.21248972415924072, "learning_rate": 9.287952226839475e-05, "loss": 0.0151, "step": 17070 }, { "grad_norm": 0.14314502477645874, "learning_rate": 9.286888817520164e-05, "loss": 0.0116, "step": 17080 }, { "grad_norm": 0.22830408811569214, "learning_rate": 9.285824675688803e-05, "loss": 0.016, "step": 17090 }, { "grad_norm": 0.18147626519203186, "learning_rate": 9.28475980152722e-05, "loss": 0.0119, "step": 17100 }, { "grad_norm": 0.15433046221733093, "learning_rate": 9.283694195217379e-05, "loss": 0.0125, "step": 17110 }, { "grad_norm": 0.1743672788143158, "learning_rate": 9.282627856941356e-05, "loss": 0.0135, "step": 17120 }, { "grad_norm": 0.21597051620483398, "learning_rate": 9.281560786881363e-05, "loss": 0.0135, "step": 17130 }, { "grad_norm": 0.1819896697998047, "learning_rate": 9.280492985219733e-05, "loss": 0.0146, "step": 17140 }, { "grad_norm": 0.21282856166362762, "learning_rate": 9.279424452138924e-05, "loss": 0.013, "step": 17150 }, { "grad_norm": 0.15203578770160675, "learning_rate": 9.278355187821517e-05, "loss": 0.0137, "step": 17160 }, { "grad_norm": 0.20708979666233063, "learning_rate": 9.277285192450224e-05, "loss": 0.0122, "step": 17170 }, { "grad_norm": 0.2199411690235138, "learning_rate": 9.276214466207875e-05, "loss": 0.0122, "step": 17180 }, { "grad_norm": 0.23344135284423828, "learning_rate": 9.275143009277427e-05, "loss": 0.0164, "step": 17190 }, { "grad_norm": 0.24107620120048523, "learning_rate": 9.274070821841964e-05, "loss": 0.0118, "step": 17200 }, { "grad_norm": 0.19071821868419647, "learning_rate": 9.272997904084696e-05, "loss": 0.0112, "step": 17210 }, { "grad_norm": 0.21350201964378357, "learning_rate": 9.271924256188951e-05, "loss": 0.0127, "step": 17220 }, { "grad_norm": 0.19189412891864777, "learning_rate": 9.270849878338189e-05, "loss": 0.016, "step": 17230 }, { "grad_norm": 0.1699749231338501, "learning_rate": 9.269774770715991e-05, "loss": 0.0154, "step": 17240 }, { "grad_norm": 0.22636796534061432, "learning_rate": 9.268698933506061e-05, "loss": 0.0123, "step": 17250 }, { "grad_norm": 0.1410752236843109, "learning_rate": 9.267622366892235e-05, "loss": 0.0151, "step": 17260 }, { "grad_norm": 0.2108774334192276, "learning_rate": 9.266545071058465e-05, "loss": 0.0135, "step": 17270 }, { "grad_norm": 0.15542387962341309, "learning_rate": 9.265467046188833e-05, "loss": 0.014, "step": 17280 }, { "grad_norm": 0.18995369970798492, "learning_rate": 9.264388292467543e-05, "loss": 0.0115, "step": 17290 }, { "grad_norm": 0.19943377375602722, "learning_rate": 9.263308810078926e-05, "loss": 0.0132, "step": 17300 }, { "grad_norm": 0.1743374764919281, "learning_rate": 9.262228599207434e-05, "loss": 0.0123, "step": 17310 }, { "grad_norm": 0.22246171534061432, "learning_rate": 9.261147660037647e-05, "loss": 0.0129, "step": 17320 }, { "grad_norm": 0.17776615917682648, "learning_rate": 9.26006599275427e-05, "loss": 0.0164, "step": 17330 }, { "grad_norm": 0.1749865561723709, "learning_rate": 9.258983597542124e-05, "loss": 0.0131, "step": 17340 }, { "grad_norm": 0.16338396072387695, "learning_rate": 9.257900474586167e-05, "loss": 0.0149, "step": 17350 }, { "grad_norm": 0.1723996251821518, "learning_rate": 9.256816624071471e-05, "loss": 0.0134, "step": 17360 }, { "grad_norm": 0.14805234968662262, "learning_rate": 9.25573204618324e-05, "loss": 0.0122, "step": 17370 }, { "grad_norm": 0.17443765699863434, "learning_rate": 9.254646741106796e-05, "loss": 0.0138, "step": 17380 }, { "grad_norm": 0.17497475445270538, "learning_rate": 9.253560709027589e-05, "loss": 0.014, "step": 17390 }, { "grad_norm": 0.23810270428657532, "learning_rate": 9.252473950131192e-05, "loss": 0.0154, "step": 17400 }, { "grad_norm": 0.15730918943881989, "learning_rate": 9.251386464603302e-05, "loss": 0.0148, "step": 17410 }, { "grad_norm": 0.18589729070663452, "learning_rate": 9.250298252629741e-05, "loss": 0.0131, "step": 17420 }, { "grad_norm": 0.17964768409729004, "learning_rate": 9.249209314396454e-05, "loss": 0.0135, "step": 17430 }, { "grad_norm": 0.20012836158275604, "learning_rate": 9.248119650089513e-05, "loss": 0.0136, "step": 17440 }, { "grad_norm": 0.2277182936668396, "learning_rate": 9.247029259895108e-05, "loss": 0.0155, "step": 17450 }, { "grad_norm": 0.12044143676757812, "learning_rate": 9.24593814399956e-05, "loss": 0.0111, "step": 17460 }, { "grad_norm": 0.2101229876279831, "learning_rate": 9.244846302589309e-05, "loss": 0.0139, "step": 17470 }, { "grad_norm": 0.16428367793560028, "learning_rate": 9.243753735850923e-05, "loss": 0.0136, "step": 17480 }, { "grad_norm": 0.1505555361509323, "learning_rate": 9.24266044397109e-05, "loss": 0.0133, "step": 17490 }, { "grad_norm": 0.14988477528095245, "learning_rate": 9.241566427136624e-05, "loss": 0.0139, "step": 17500 }, { "grad_norm": 0.1770629584789276, "learning_rate": 9.240471685534463e-05, "loss": 0.0145, "step": 17510 }, { "grad_norm": 0.21696192026138306, "learning_rate": 9.239376219351667e-05, "loss": 0.0108, "step": 17520 }, { "grad_norm": 0.17195232212543488, "learning_rate": 9.238280028775425e-05, "loss": 0.0139, "step": 17530 }, { "grad_norm": 0.22402401268482208, "learning_rate": 9.237183113993041e-05, "loss": 0.0136, "step": 17540 }, { "grad_norm": 0.2215157002210617, "learning_rate": 9.236085475191952e-05, "loss": 0.0122, "step": 17550 }, { "grad_norm": 0.24035313725471497, "learning_rate": 9.234987112559709e-05, "loss": 0.014, "step": 17560 }, { "grad_norm": 0.15574344992637634, "learning_rate": 9.233888026283999e-05, "loss": 0.0139, "step": 17570 }, { "grad_norm": 0.1950378268957138, "learning_rate": 9.232788216552619e-05, "loss": 0.0137, "step": 17580 }, { "grad_norm": 0.16145354509353638, "learning_rate": 9.231687683553502e-05, "loss": 0.0142, "step": 17590 }, { "grad_norm": 0.19705677032470703, "learning_rate": 9.230586427474698e-05, "loss": 0.0122, "step": 17600 }, { "grad_norm": 0.2366867959499359, "learning_rate": 9.229484448504379e-05, "loss": 0.0135, "step": 17610 }, { "grad_norm": 0.1753239780664444, "learning_rate": 9.228381746830843e-05, "loss": 0.0126, "step": 17620 }, { "grad_norm": 0.15550604462623596, "learning_rate": 9.227278322642514e-05, "loss": 0.0147, "step": 17630 }, { "grad_norm": 0.19015583395957947, "learning_rate": 9.226174176127937e-05, "loss": 0.0142, "step": 17640 }, { "grad_norm": 0.26237231492996216, "learning_rate": 9.22506930747578e-05, "loss": 0.0106, "step": 17650 }, { "grad_norm": 0.16041891276836395, "learning_rate": 9.223963716874831e-05, "loss": 0.0109, "step": 17660 }, { "grad_norm": 0.17696517705917358, "learning_rate": 9.222857404514012e-05, "loss": 0.0119, "step": 17670 }, { "grad_norm": 0.21645700931549072, "learning_rate": 9.221750370582355e-05, "loss": 0.0166, "step": 17680 }, { "grad_norm": 0.16444453597068787, "learning_rate": 9.220642615269028e-05, "loss": 0.0131, "step": 17690 }, { "grad_norm": 0.2460213154554367, "learning_rate": 9.219534138763311e-05, "loss": 0.0137, "step": 17700 }, { "grad_norm": 0.18964852392673492, "learning_rate": 9.218424941254613e-05, "loss": 0.0127, "step": 17710 }, { "grad_norm": 0.1613643765449524, "learning_rate": 9.217315022932468e-05, "loss": 0.0151, "step": 17720 }, { "grad_norm": 0.1725803166627884, "learning_rate": 9.216204383986528e-05, "loss": 0.013, "step": 17730 }, { "grad_norm": 0.20526035130023956, "learning_rate": 9.215093024606574e-05, "loss": 0.0168, "step": 17740 }, { "grad_norm": 0.19584161043167114, "learning_rate": 9.213980944982506e-05, "loss": 0.0116, "step": 17750 }, { "grad_norm": 0.2473888099193573, "learning_rate": 9.212868145304346e-05, "loss": 0.0191, "step": 17760 }, { "grad_norm": 0.21274705231189728, "learning_rate": 9.211754625762241e-05, "loss": 0.0142, "step": 17770 }, { "grad_norm": 0.16856501996517181, "learning_rate": 9.210640386546463e-05, "loss": 0.013, "step": 17780 }, { "grad_norm": 0.1928894966840744, "learning_rate": 9.209525427847405e-05, "loss": 0.0143, "step": 17790 }, { "grad_norm": 0.22166326642036438, "learning_rate": 9.208409749855583e-05, "loss": 0.0118, "step": 17800 }, { "grad_norm": 0.21438342332839966, "learning_rate": 9.207293352761633e-05, "loss": 0.0172, "step": 17810 }, { "grad_norm": 0.20039872825145721, "learning_rate": 9.206176236756319e-05, "loss": 0.0138, "step": 17820 }, { "grad_norm": 0.19992221891880035, "learning_rate": 9.205058402030525e-05, "loss": 0.0122, "step": 17830 }, { "grad_norm": 0.17921507358551025, "learning_rate": 9.203939848775259e-05, "loss": 0.0129, "step": 17840 }, { "grad_norm": 0.1891181617975235, "learning_rate": 9.202820577181652e-05, "loss": 0.0124, "step": 17850 }, { "grad_norm": 0.20886464416980743, "learning_rate": 9.201700587440953e-05, "loss": 0.0133, "step": 17860 }, { "grad_norm": 0.22125442326068878, "learning_rate": 9.200579879744544e-05, "loss": 0.0165, "step": 17870 }, { "grad_norm": 0.28288981318473816, "learning_rate": 9.199458454283918e-05, "loss": 0.0126, "step": 17880 }, { "grad_norm": 0.23540043830871582, "learning_rate": 9.198336311250697e-05, "loss": 0.0159, "step": 17890 }, { "grad_norm": 0.2269206941127777, "learning_rate": 9.197213450836626e-05, "loss": 0.0141, "step": 17900 }, { "grad_norm": 0.16351130604743958, "learning_rate": 9.19608987323357e-05, "loss": 0.0157, "step": 17910 }, { "grad_norm": 0.20740322768688202, "learning_rate": 9.194965578633517e-05, "loss": 0.0163, "step": 17920 }, { "grad_norm": 0.19839566946029663, "learning_rate": 9.193840567228582e-05, "loss": 0.013, "step": 17930 }, { "grad_norm": 0.17743068933486938, "learning_rate": 9.192714839210994e-05, "loss": 0.0151, "step": 17940 }, { "grad_norm": 0.20228804647922516, "learning_rate": 9.19158839477311e-05, "loss": 0.0115, "step": 17950 }, { "grad_norm": 0.22186242043972015, "learning_rate": 9.190461234107411e-05, "loss": 0.0174, "step": 17960 }, { "grad_norm": 0.18722623586654663, "learning_rate": 9.189333357406496e-05, "loss": 0.0143, "step": 17970 }, { "grad_norm": 0.16136790812015533, "learning_rate": 9.188204764863089e-05, "loss": 0.0122, "step": 17980 }, { "grad_norm": 0.18374672532081604, "learning_rate": 9.187075456670033e-05, "loss": 0.0149, "step": 17990 }, { "grad_norm": 0.16200052201747894, "learning_rate": 9.1859454330203e-05, "loss": 0.0132, "step": 18000 }, { "grad_norm": 0.23650895059108734, "learning_rate": 9.18481469410698e-05, "loss": 0.0151, "step": 18010 }, { "grad_norm": 0.1517019420862198, "learning_rate": 9.183683240123281e-05, "loss": 0.0132, "step": 18020 }, { "grad_norm": 0.1549462229013443, "learning_rate": 9.182551071262541e-05, "loss": 0.0119, "step": 18030 }, { "grad_norm": 0.15861834585666656, "learning_rate": 9.181418187718218e-05, "loss": 0.0128, "step": 18040 }, { "grad_norm": 0.20243056118488312, "learning_rate": 9.180284589683888e-05, "loss": 0.0136, "step": 18050 }, { "grad_norm": 0.1826222836971283, "learning_rate": 9.17915027735325e-05, "loss": 0.0159, "step": 18060 }, { "grad_norm": 0.22005526721477509, "learning_rate": 9.178015250920133e-05, "loss": 0.0162, "step": 18070 }, { "grad_norm": 0.2231910079717636, "learning_rate": 9.176879510578477e-05, "loss": 0.0122, "step": 18080 }, { "grad_norm": 0.2409885674715042, "learning_rate": 9.17574305652235e-05, "loss": 0.0145, "step": 18090 }, { "grad_norm": 0.24543628096580505, "learning_rate": 9.174605888945942e-05, "loss": 0.0144, "step": 18100 }, { "grad_norm": 0.15734244883060455, "learning_rate": 9.173468008043564e-05, "loss": 0.0122, "step": 18110 }, { "grad_norm": 0.2120562195777893, "learning_rate": 9.172329414009648e-05, "loss": 0.0181, "step": 18120 }, { "grad_norm": 0.22510062158107758, "learning_rate": 9.171190107038747e-05, "loss": 0.0145, "step": 18130 }, { "grad_norm": 0.19996482133865356, "learning_rate": 9.170050087325541e-05, "loss": 0.0132, "step": 18140 }, { "grad_norm": 0.1946832239627838, "learning_rate": 9.168909355064824e-05, "loss": 0.0117, "step": 18150 }, { "grad_norm": 0.23889774084091187, "learning_rate": 9.167767910451519e-05, "loss": 0.0133, "step": 18160 }, { "grad_norm": 0.21417541801929474, "learning_rate": 9.166625753680669e-05, "loss": 0.0134, "step": 18170 }, { "grad_norm": 0.1514919549226761, "learning_rate": 9.165482884947431e-05, "loss": 0.0137, "step": 18180 }, { "grad_norm": 0.17841129004955292, "learning_rate": 9.164339304447098e-05, "loss": 0.0141, "step": 18190 }, { "grad_norm": 0.2573177218437195, "learning_rate": 9.163195012375072e-05, "loss": 0.0149, "step": 18200 }, { "grad_norm": 0.14912240207195282, "learning_rate": 9.16205000892688e-05, "loss": 0.0122, "step": 18210 }, { "grad_norm": 0.19713225960731506, "learning_rate": 9.160904294298175e-05, "loss": 0.0127, "step": 18220 }, { "grad_norm": 0.20363280177116394, "learning_rate": 9.159757868684727e-05, "loss": 0.0161, "step": 18230 }, { "grad_norm": 0.19096125662326813, "learning_rate": 9.15861073228243e-05, "loss": 0.0166, "step": 18240 }, { "grad_norm": 0.17602857947349548, "learning_rate": 9.157462885287296e-05, "loss": 0.0149, "step": 18250 }, { "grad_norm": 0.21382023394107819, "learning_rate": 9.156314327895461e-05, "loss": 0.0142, "step": 18260 }, { "grad_norm": 0.16911453008651733, "learning_rate": 9.155165060303185e-05, "loss": 0.0152, "step": 18270 }, { "grad_norm": 0.33933597803115845, "learning_rate": 9.154015082706841e-05, "loss": 0.014, "step": 18280 }, { "grad_norm": 0.19784002006053925, "learning_rate": 9.152864395302936e-05, "loss": 0.0119, "step": 18290 }, { "grad_norm": 0.2724061608314514, "learning_rate": 9.151712998288085e-05, "loss": 0.0126, "step": 18300 }, { "grad_norm": 0.21507041156291962, "learning_rate": 9.150560891859031e-05, "loss": 0.0176, "step": 18310 }, { "grad_norm": 0.1790425330400467, "learning_rate": 9.14940807621264e-05, "loss": 0.0118, "step": 18320 }, { "grad_norm": 0.20529644191265106, "learning_rate": 9.148254551545894e-05, "loss": 0.0163, "step": 18330 }, { "grad_norm": 0.24837712943553925, "learning_rate": 9.147100318055901e-05, "loss": 0.0138, "step": 18340 }, { "grad_norm": 0.218406081199646, "learning_rate": 9.145945375939888e-05, "loss": 0.014, "step": 18350 }, { "grad_norm": 0.21016927063465118, "learning_rate": 9.144789725395203e-05, "loss": 0.0132, "step": 18360 }, { "grad_norm": 0.16028355062007904, "learning_rate": 9.14363336661931e-05, "loss": 0.0122, "step": 18370 }, { "grad_norm": 0.17350605130195618, "learning_rate": 9.142476299809806e-05, "loss": 0.0144, "step": 18380 }, { "grad_norm": 0.18095062673091888, "learning_rate": 9.1413185251644e-05, "loss": 0.0109, "step": 18390 }, { "grad_norm": 0.15203708410263062, "learning_rate": 9.140160042880923e-05, "loss": 0.0124, "step": 18400 }, { "grad_norm": 0.2144121676683426, "learning_rate": 9.139000853157327e-05, "loss": 0.0123, "step": 18410 }, { "grad_norm": 0.29982665181159973, "learning_rate": 9.137840956191688e-05, "loss": 0.0122, "step": 18420 }, { "grad_norm": 0.184938445687294, "learning_rate": 9.136680352182199e-05, "loss": 0.0152, "step": 18430 }, { "grad_norm": 0.1835850030183792, "learning_rate": 9.135519041327177e-05, "loss": 0.0151, "step": 18440 }, { "grad_norm": 0.24095039069652557, "learning_rate": 9.134357023825058e-05, "loss": 0.0152, "step": 18450 }, { "grad_norm": 0.26438823342323303, "learning_rate": 9.133194299874398e-05, "loss": 0.0116, "step": 18460 }, { "grad_norm": 0.14029064774513245, "learning_rate": 9.132030869673876e-05, "loss": 0.0133, "step": 18470 }, { "grad_norm": 0.13778282701969147, "learning_rate": 9.130866733422288e-05, "loss": 0.0118, "step": 18480 }, { "grad_norm": 0.18640293180942535, "learning_rate": 9.129701891318556e-05, "loss": 0.0156, "step": 18490 }, { "grad_norm": 0.21208779513835907, "learning_rate": 9.128536343561718e-05, "loss": 0.0133, "step": 18500 }, { "grad_norm": 0.20185844600200653, "learning_rate": 9.127370090350934e-05, "loss": 0.0143, "step": 18510 }, { "grad_norm": 0.171695277094841, "learning_rate": 9.126203131885487e-05, "loss": 0.0131, "step": 18520 }, { "grad_norm": 0.3312089443206787, "learning_rate": 9.125035468364775e-05, "loss": 0.0127, "step": 18530 }, { "grad_norm": 0.23836182057857513, "learning_rate": 9.123867099988322e-05, "loss": 0.0153, "step": 18540 }, { "grad_norm": 0.20346926152706146, "learning_rate": 9.122698026955769e-05, "loss": 0.0151, "step": 18550 }, { "grad_norm": 0.19699165225028992, "learning_rate": 9.12152824946688e-05, "loss": 0.0124, "step": 18560 }, { "grad_norm": 0.18771366775035858, "learning_rate": 9.120357767721538e-05, "loss": 0.0115, "step": 18570 }, { "grad_norm": 0.17075955867767334, "learning_rate": 9.119186581919745e-05, "loss": 0.011, "step": 18580 }, { "grad_norm": 0.19065038859844208, "learning_rate": 9.118014692261624e-05, "loss": 0.0135, "step": 18590 }, { "grad_norm": 0.16554714739322662, "learning_rate": 9.116842098947422e-05, "loss": 0.0119, "step": 18600 }, { "grad_norm": 0.19793666899204254, "learning_rate": 9.115668802177499e-05, "loss": 0.0144, "step": 18610 }, { "grad_norm": 0.22393065690994263, "learning_rate": 9.114494802152342e-05, "loss": 0.0151, "step": 18620 }, { "grad_norm": 0.23331038653850555, "learning_rate": 9.113320099072555e-05, "loss": 0.0146, "step": 18630 }, { "grad_norm": 0.1782078742980957, "learning_rate": 9.112144693138864e-05, "loss": 0.0113, "step": 18640 }, { "grad_norm": 0.14894218742847443, "learning_rate": 9.110968584552111e-05, "loss": 0.0115, "step": 18650 }, { "grad_norm": 0.19811050593852997, "learning_rate": 9.109791773513264e-05, "loss": 0.0127, "step": 18660 }, { "grad_norm": 0.1920621246099472, "learning_rate": 9.108614260223403e-05, "loss": 0.0115, "step": 18670 }, { "grad_norm": 0.23308110237121582, "learning_rate": 9.107436044883738e-05, "loss": 0.0153, "step": 18680 }, { "grad_norm": 0.19370634853839874, "learning_rate": 9.10625712769559e-05, "loss": 0.0141, "step": 18690 }, { "grad_norm": 0.15249225497245789, "learning_rate": 9.105077508860406e-05, "loss": 0.0162, "step": 18700 }, { "grad_norm": 0.1839636117219925, "learning_rate": 9.103897188579751e-05, "loss": 0.0122, "step": 18710 }, { "grad_norm": 0.14599676430225372, "learning_rate": 9.102716167055308e-05, "loss": 0.0126, "step": 18720 }, { "grad_norm": 0.2068442851305008, "learning_rate": 9.10153444448888e-05, "loss": 0.0146, "step": 18730 }, { "grad_norm": 0.195174440741539, "learning_rate": 9.100352021082393e-05, "loss": 0.0123, "step": 18740 }, { "grad_norm": 0.2407478243112564, "learning_rate": 9.099168897037891e-05, "loss": 0.0124, "step": 18750 }, { "grad_norm": 0.181065171957016, "learning_rate": 9.097985072557538e-05, "loss": 0.0104, "step": 18760 }, { "grad_norm": 0.23571264743804932, "learning_rate": 9.096800547843615e-05, "loss": 0.0117, "step": 18770 }, { "grad_norm": 0.1971934288740158, "learning_rate": 9.095615323098526e-05, "loss": 0.0129, "step": 18780 }, { "grad_norm": 0.12422408163547516, "learning_rate": 9.094429398524795e-05, "loss": 0.0111, "step": 18790 }, { "grad_norm": 0.23461951315402985, "learning_rate": 9.093242774325061e-05, "loss": 0.013, "step": 18800 }, { "grad_norm": 0.10859572142362595, "learning_rate": 9.092055450702088e-05, "loss": 0.0117, "step": 18810 }, { "grad_norm": 0.18824456632137299, "learning_rate": 9.090867427858756e-05, "loss": 0.0162, "step": 18820 }, { "grad_norm": 0.20775724947452545, "learning_rate": 9.089678705998066e-05, "loss": 0.0129, "step": 18830 }, { "grad_norm": 0.15437346696853638, "learning_rate": 9.088489285323139e-05, "loss": 0.0132, "step": 18840 }, { "grad_norm": 0.17958448827266693, "learning_rate": 9.087299166037212e-05, "loss": 0.0127, "step": 18850 }, { "grad_norm": 0.18387262523174286, "learning_rate": 9.086108348343647e-05, "loss": 0.0125, "step": 18860 }, { "grad_norm": 0.1510317176580429, "learning_rate": 9.08491683244592e-05, "loss": 0.011, "step": 18870 }, { "grad_norm": 0.19256602227687836, "learning_rate": 9.08372461854763e-05, "loss": 0.0136, "step": 18880 }, { "grad_norm": 0.17824296653270721, "learning_rate": 9.082531706852492e-05, "loss": 0.0148, "step": 18890 }, { "grad_norm": 0.14880792796611786, "learning_rate": 9.081338097564342e-05, "loss": 0.0119, "step": 18900 }, { "grad_norm": 0.2163858562707901, "learning_rate": 9.080143790887137e-05, "loss": 0.0133, "step": 18910 }, { "grad_norm": 0.19323782622814178, "learning_rate": 9.07894878702495e-05, "loss": 0.0125, "step": 18920 }, { "grad_norm": 0.15614891052246094, "learning_rate": 9.077753086181974e-05, "loss": 0.0134, "step": 18930 }, { "grad_norm": 0.15938818454742432, "learning_rate": 9.076556688562524e-05, "loss": 0.0125, "step": 18940 }, { "grad_norm": 0.19265899062156677, "learning_rate": 9.075359594371029e-05, "loss": 0.0132, "step": 18950 }, { "grad_norm": 0.17136149108409882, "learning_rate": 9.07416180381204e-05, "loss": 0.0141, "step": 18960 }, { "grad_norm": 0.20749300718307495, "learning_rate": 9.072963317090228e-05, "loss": 0.0129, "step": 18970 }, { "grad_norm": 0.212813138961792, "learning_rate": 9.071764134410382e-05, "loss": 0.0133, "step": 18980 }, { "grad_norm": 0.14468073844909668, "learning_rate": 9.070564255977407e-05, "loss": 0.0132, "step": 18990 }, { "grad_norm": 0.189195916056633, "learning_rate": 9.06936368199633e-05, "loss": 0.0134, "step": 19000 }, { "grad_norm": 0.18951648473739624, "learning_rate": 9.0681624126723e-05, "loss": 0.0126, "step": 19010 }, { "grad_norm": 0.15426264703273773, "learning_rate": 9.066960448210576e-05, "loss": 0.013, "step": 19020 }, { "grad_norm": 0.15015970170497894, "learning_rate": 9.065757788816543e-05, "loss": 0.0147, "step": 19030 }, { "grad_norm": 0.18159504234790802, "learning_rate": 9.064554434695705e-05, "loss": 0.011, "step": 19040 }, { "grad_norm": 0.19993504881858826, "learning_rate": 9.063350386053677e-05, "loss": 0.0118, "step": 19050 }, { "grad_norm": 0.18363803625106812, "learning_rate": 9.062145643096202e-05, "loss": 0.0111, "step": 19060 }, { "grad_norm": 0.21207058429718018, "learning_rate": 9.060940206029136e-05, "loss": 0.0132, "step": 19070 }, { "grad_norm": 0.1648569107055664, "learning_rate": 9.059734075058457e-05, "loss": 0.0156, "step": 19080 }, { "grad_norm": 0.17429585754871368, "learning_rate": 9.058527250390257e-05, "loss": 0.0126, "step": 19090 }, { "grad_norm": 0.16477568447589874, "learning_rate": 9.057319732230752e-05, "loss": 0.0122, "step": 19100 }, { "grad_norm": 0.1417025327682495, "learning_rate": 9.056111520786273e-05, "loss": 0.0113, "step": 19110 }, { "grad_norm": 0.1806042343378067, "learning_rate": 9.054902616263268e-05, "loss": 0.0137, "step": 19120 }, { "grad_norm": 0.16554681956768036, "learning_rate": 9.05369301886831e-05, "loss": 0.0116, "step": 19130 }, { "grad_norm": 0.15505123138427734, "learning_rate": 9.052482728808083e-05, "loss": 0.0115, "step": 19140 }, { "grad_norm": 0.17102472484111786, "learning_rate": 9.051271746289391e-05, "loss": 0.0119, "step": 19150 }, { "grad_norm": 0.13587075471878052, "learning_rate": 9.050060071519162e-05, "loss": 0.0106, "step": 19160 }, { "grad_norm": 0.2036549597978592, "learning_rate": 9.048847704704437e-05, "loss": 0.0115, "step": 19170 }, { "grad_norm": 0.20469102263450623, "learning_rate": 9.047634646052376e-05, "loss": 0.0141, "step": 19180 }, { "grad_norm": 0.1712925285100937, "learning_rate": 9.046420895770256e-05, "loss": 0.0129, "step": 19190 }, { "grad_norm": 0.21492059528827667, "learning_rate": 9.045206454065473e-05, "loss": 0.0119, "step": 19200 }, { "grad_norm": 0.1727629005908966, "learning_rate": 9.043991321145546e-05, "loss": 0.0118, "step": 19210 }, { "grad_norm": 0.12963104248046875, "learning_rate": 9.042775497218105e-05, "loss": 0.0107, "step": 19220 }, { "grad_norm": 0.18781951069831848, "learning_rate": 9.041558982490901e-05, "loss": 0.0122, "step": 19230 }, { "grad_norm": 0.21126677095890045, "learning_rate": 9.040341777171805e-05, "loss": 0.0125, "step": 19240 }, { "grad_norm": 0.16456206142902374, "learning_rate": 9.039123881468802e-05, "loss": 0.0146, "step": 19250 }, { "grad_norm": 0.16611948609352112, "learning_rate": 9.037905295589998e-05, "loss": 0.0114, "step": 19260 }, { "grad_norm": 0.15447163581848145, "learning_rate": 9.036686019743617e-05, "loss": 0.0115, "step": 19270 }, { "grad_norm": 0.25507643818855286, "learning_rate": 9.035466054137997e-05, "loss": 0.0132, "step": 19280 }, { "grad_norm": 0.24597585201263428, "learning_rate": 9.0342453989816e-05, "loss": 0.0129, "step": 19290 }, { "grad_norm": 0.15553338825702667, "learning_rate": 9.033024054483e-05, "loss": 0.0106, "step": 19300 }, { "grad_norm": 0.15428820252418518, "learning_rate": 9.031802020850894e-05, "loss": 0.0102, "step": 19310 }, { "grad_norm": 0.169927716255188, "learning_rate": 9.030579298294092e-05, "loss": 0.0123, "step": 19320 }, { "grad_norm": 0.22080400586128235, "learning_rate": 9.029355887021524e-05, "loss": 0.0142, "step": 19330 }, { "grad_norm": 0.16756249964237213, "learning_rate": 9.028131787242238e-05, "loss": 0.0143, "step": 19340 }, { "grad_norm": 0.13718251883983612, "learning_rate": 9.026906999165399e-05, "loss": 0.0116, "step": 19350 }, { "grad_norm": 0.16382631659507751, "learning_rate": 9.025681523000291e-05, "loss": 0.0128, "step": 19360 }, { "grad_norm": 0.22259511053562164, "learning_rate": 9.024455358956315e-05, "loss": 0.0118, "step": 19370 }, { "grad_norm": 0.228518545627594, "learning_rate": 9.023228507242984e-05, "loss": 0.0132, "step": 19380 }, { "grad_norm": 0.17146626114845276, "learning_rate": 9.022000968069937e-05, "loss": 0.0136, "step": 19390 }, { "grad_norm": 0.20421798527240753, "learning_rate": 9.020772741646928e-05, "loss": 0.0125, "step": 19400 }, { "grad_norm": 0.2708723843097687, "learning_rate": 9.019543828183826e-05, "loss": 0.0153, "step": 19410 }, { "grad_norm": 0.18332992494106293, "learning_rate": 9.018314227890616e-05, "loss": 0.0113, "step": 19420 }, { "grad_norm": 0.2386111468076706, "learning_rate": 9.017083940977408e-05, "loss": 0.0127, "step": 19430 }, { "grad_norm": 0.20144227147102356, "learning_rate": 9.015852967654422e-05, "loss": 0.0127, "step": 19440 }, { "grad_norm": 0.23339514434337616, "learning_rate": 9.014621308131996e-05, "loss": 0.0141, "step": 19450 }, { "grad_norm": 0.18909837305545807, "learning_rate": 9.01338896262059e-05, "loss": 0.0112, "step": 19460 }, { "grad_norm": 0.1814464032649994, "learning_rate": 9.012155931330777e-05, "loss": 0.0135, "step": 19470 }, { "grad_norm": 0.15173906087875366, "learning_rate": 9.010922214473246e-05, "loss": 0.0137, "step": 19480 }, { "grad_norm": 0.17886030673980713, "learning_rate": 9.009687812258808e-05, "loss": 0.0108, "step": 19490 }, { "grad_norm": 0.1411319226026535, "learning_rate": 9.00845272489839e-05, "loss": 0.0098, "step": 19500 }, { "grad_norm": 0.15544065833091736, "learning_rate": 9.007216952603031e-05, "loss": 0.014, "step": 19510 }, { "grad_norm": 0.14678438007831573, "learning_rate": 9.005980495583894e-05, "loss": 0.011, "step": 19520 }, { "grad_norm": 0.14610466361045837, "learning_rate": 9.004743354052252e-05, "loss": 0.0123, "step": 19530 }, { "grad_norm": 0.20351964235305786, "learning_rate": 9.003505528219503e-05, "loss": 0.0111, "step": 19540 }, { "grad_norm": 0.22129780054092407, "learning_rate": 9.002267018297154e-05, "loss": 0.0136, "step": 19550 }, { "grad_norm": 0.19861210882663727, "learning_rate": 9.001027824496834e-05, "loss": 0.0122, "step": 19560 }, { "grad_norm": 0.2008909434080124, "learning_rate": 8.999787947030287e-05, "loss": 0.0143, "step": 19570 }, { "grad_norm": 0.1577673852443695, "learning_rate": 8.998547386109376e-05, "loss": 0.0095, "step": 19580 }, { "grad_norm": 0.1669655442237854, "learning_rate": 8.997306141946073e-05, "loss": 0.0127, "step": 19590 }, { "grad_norm": 0.25557249784469604, "learning_rate": 8.996064214752481e-05, "loss": 0.0138, "step": 19600 }, { "grad_norm": 0.18797093629837036, "learning_rate": 8.994821604740806e-05, "loss": 0.0111, "step": 19610 }, { "grad_norm": 0.16497628390789032, "learning_rate": 8.993578312123377e-05, "loss": 0.0123, "step": 19620 }, { "grad_norm": 0.2098742127418518, "learning_rate": 8.992334337112639e-05, "loss": 0.0108, "step": 19630 }, { "grad_norm": 0.20337410271167755, "learning_rate": 8.991089679921154e-05, "loss": 0.0143, "step": 19640 }, { "grad_norm": 0.13102540373802185, "learning_rate": 8.989844340761599e-05, "loss": 0.0111, "step": 19650 }, { "grad_norm": 0.18353909254074097, "learning_rate": 8.988598319846768e-05, "loss": 0.0108, "step": 19660 }, { "grad_norm": 0.14911077916622162, "learning_rate": 8.987351617389574e-05, "loss": 0.0126, "step": 19670 }, { "grad_norm": 0.17060181498527527, "learning_rate": 8.98610423360304e-05, "loss": 0.0134, "step": 19680 }, { "grad_norm": 0.21845252811908722, "learning_rate": 8.984856168700317e-05, "loss": 0.0112, "step": 19690 }, { "grad_norm": 0.133164182305336, "learning_rate": 8.983607422894658e-05, "loss": 0.0102, "step": 19700 }, { "grad_norm": 0.1486656367778778, "learning_rate": 8.982357996399442e-05, "loss": 0.0127, "step": 19710 }, { "grad_norm": 0.15950413048267365, "learning_rate": 8.981107889428164e-05, "loss": 0.011, "step": 19720 }, { "grad_norm": 0.18453404307365417, "learning_rate": 8.979857102194428e-05, "loss": 0.0149, "step": 19730 }, { "grad_norm": 0.17544421553611755, "learning_rate": 8.978605634911968e-05, "loss": 0.012, "step": 19740 }, { "grad_norm": 0.14410410821437836, "learning_rate": 8.977353487794616e-05, "loss": 0.0105, "step": 19750 }, { "grad_norm": 0.1578279733657837, "learning_rate": 8.976100661056334e-05, "loss": 0.01, "step": 19760 }, { "grad_norm": 0.2828732132911682, "learning_rate": 8.974847154911197e-05, "loss": 0.014, "step": 19770 }, { "grad_norm": 0.24007229506969452, "learning_rate": 8.973592969573393e-05, "loss": 0.0118, "step": 19780 }, { "grad_norm": 0.1512778103351593, "learning_rate": 8.972338105257228e-05, "loss": 0.0128, "step": 19790 }, { "grad_norm": 0.1501401960849762, "learning_rate": 8.971082562177125e-05, "loss": 0.0102, "step": 19800 }, { "grad_norm": 0.20458421111106873, "learning_rate": 8.96982634054762e-05, "loss": 0.0123, "step": 19810 }, { "grad_norm": 0.2077050656080246, "learning_rate": 8.96856944058337e-05, "loss": 0.0122, "step": 19820 }, { "grad_norm": 0.2462027221918106, "learning_rate": 8.967311862499144e-05, "loss": 0.0144, "step": 19830 }, { "grad_norm": 0.21804365515708923, "learning_rate": 8.966053606509825e-05, "loss": 0.0125, "step": 19840 }, { "grad_norm": 0.16984902322292328, "learning_rate": 8.964794672830417e-05, "loss": 0.0119, "step": 19850 }, { "grad_norm": 0.20671623945236206, "learning_rate": 8.963535061676038e-05, "loss": 0.0146, "step": 19860 }, { "grad_norm": 0.1293109953403473, "learning_rate": 8.962274773261918e-05, "loss": 0.0144, "step": 19870 }, { "grad_norm": 0.19944562017917633, "learning_rate": 8.961013807803409e-05, "loss": 0.0139, "step": 19880 }, { "grad_norm": 0.17458152770996094, "learning_rate": 8.959752165515973e-05, "loss": 0.0117, "step": 19890 }, { "grad_norm": 0.1712852120399475, "learning_rate": 8.958489846615193e-05, "loss": 0.0129, "step": 19900 }, { "grad_norm": 0.2058737426996231, "learning_rate": 8.957226851316762e-05, "loss": 0.0151, "step": 19910 }, { "grad_norm": 0.21653218567371368, "learning_rate": 8.955963179836493e-05, "loss": 0.0131, "step": 19920 }, { "grad_norm": 0.1633061170578003, "learning_rate": 8.954698832390312e-05, "loss": 0.0115, "step": 19930 }, { "grad_norm": 0.1969883143901825, "learning_rate": 8.953433809194263e-05, "loss": 0.0126, "step": 19940 }, { "grad_norm": 0.14125077426433563, "learning_rate": 8.9521681104645e-05, "loss": 0.0094, "step": 19950 }, { "grad_norm": 0.15786442160606384, "learning_rate": 8.9509017364173e-05, "loss": 0.0127, "step": 19960 }, { "grad_norm": 0.19302570819854736, "learning_rate": 8.949634687269052e-05, "loss": 0.0142, "step": 19970 }, { "grad_norm": 0.22654831409454346, "learning_rate": 8.948366963236259e-05, "loss": 0.0111, "step": 19980 }, { "grad_norm": 0.18191757798194885, "learning_rate": 8.947098564535538e-05, "loss": 0.011, "step": 19990 }, { "grad_norm": 0.1614924967288971, "learning_rate": 8.945829491383627e-05, "loss": 0.0121, "step": 20000 }, { "grad_norm": 0.2052549570798874, "learning_rate": 8.944559743997374e-05, "loss": 0.016, "step": 20010 }, { "grad_norm": 0.16096705198287964, "learning_rate": 8.943289322593746e-05, "loss": 0.0122, "step": 20020 }, { "grad_norm": 0.15936493873596191, "learning_rate": 8.942018227389821e-05, "loss": 0.0119, "step": 20030 }, { "grad_norm": 0.194100484251976, "learning_rate": 8.940746458602795e-05, "loss": 0.0114, "step": 20040 }, { "grad_norm": 0.18453969061374664, "learning_rate": 8.939474016449979e-05, "loss": 0.011, "step": 20050 }, { "grad_norm": 0.19046039879322052, "learning_rate": 8.938200901148799e-05, "loss": 0.0124, "step": 20060 }, { "grad_norm": 0.22977358102798462, "learning_rate": 8.936927112916795e-05, "loss": 0.0132, "step": 20070 }, { "grad_norm": 0.15774136781692505, "learning_rate": 8.935652651971622e-05, "loss": 0.0115, "step": 20080 }, { "grad_norm": 0.21610625088214874, "learning_rate": 8.934377518531052e-05, "loss": 0.0143, "step": 20090 }, { "grad_norm": 0.21794386208057404, "learning_rate": 8.933101712812967e-05, "loss": 0.0168, "step": 20100 }, { "grad_norm": 0.192006915807724, "learning_rate": 8.931825235035374e-05, "loss": 0.0151, "step": 20110 }, { "grad_norm": 0.19524814188480377, "learning_rate": 8.930548085416382e-05, "loss": 0.0153, "step": 20120 }, { "grad_norm": 0.18854393064975739, "learning_rate": 8.92927026417422e-05, "loss": 0.0133, "step": 20130 }, { "grad_norm": 0.2829955220222473, "learning_rate": 8.92799177152724e-05, "loss": 0.0113, "step": 20140 }, { "grad_norm": 0.1650342345237732, "learning_rate": 8.926712607693895e-05, "loss": 0.0152, "step": 20150 }, { "grad_norm": 0.22833721339702606, "learning_rate": 8.925432772892762e-05, "loss": 0.0112, "step": 20160 }, { "grad_norm": 0.19418616592884064, "learning_rate": 8.924152267342529e-05, "loss": 0.0112, "step": 20170 }, { "grad_norm": 0.1623813658952713, "learning_rate": 8.922871091261998e-05, "loss": 0.0113, "step": 20180 }, { "grad_norm": 0.17073604464530945, "learning_rate": 8.92158924487009e-05, "loss": 0.0118, "step": 20190 }, { "grad_norm": 0.17282451689243317, "learning_rate": 8.920306728385834e-05, "loss": 0.0111, "step": 20200 }, { "grad_norm": 0.24892738461494446, "learning_rate": 8.919023542028379e-05, "loss": 0.012, "step": 20210 }, { "grad_norm": 0.2302129715681076, "learning_rate": 8.917739686016988e-05, "loss": 0.0148, "step": 20220 }, { "grad_norm": 0.17293429374694824, "learning_rate": 8.916455160571033e-05, "loss": 0.0102, "step": 20230 }, { "grad_norm": 0.2572401165962219, "learning_rate": 8.915169965910008e-05, "loss": 0.0116, "step": 20240 }, { "grad_norm": 0.1979701668024063, "learning_rate": 8.913884102253514e-05, "loss": 0.0137, "step": 20250 }, { "grad_norm": 0.207551047205925, "learning_rate": 8.912597569821273e-05, "loss": 0.0144, "step": 20260 }, { "grad_norm": 0.18062080442905426, "learning_rate": 8.911310368833118e-05, "loss": 0.0111, "step": 20270 }, { "grad_norm": 0.2221459150314331, "learning_rate": 8.910022499508994e-05, "loss": 0.0119, "step": 20280 }, { "grad_norm": 0.18595586717128754, "learning_rate": 8.908733962068965e-05, "loss": 0.0121, "step": 20290 }, { "grad_norm": 0.17996752262115479, "learning_rate": 8.907444756733207e-05, "loss": 0.0107, "step": 20300 }, { "grad_norm": 0.1355288028717041, "learning_rate": 8.906154883722006e-05, "loss": 0.0162, "step": 20310 }, { "grad_norm": 0.15123143792152405, "learning_rate": 8.904864343255773e-05, "loss": 0.0112, "step": 20320 }, { "grad_norm": 0.13376747071743011, "learning_rate": 8.90357313555502e-05, "loss": 0.0116, "step": 20330 }, { "grad_norm": 0.16011349856853485, "learning_rate": 8.90228126084038e-05, "loss": 0.011, "step": 20340 }, { "grad_norm": 0.16490861773490906, "learning_rate": 8.900988719332601e-05, "loss": 0.0132, "step": 20350 }, { "grad_norm": 0.17034660279750824, "learning_rate": 8.899695511252542e-05, "loss": 0.0127, "step": 20360 }, { "grad_norm": 0.16047410666942596, "learning_rate": 8.898401636821176e-05, "loss": 0.0126, "step": 20370 }, { "grad_norm": 0.17646487057209015, "learning_rate": 8.897107096259593e-05, "loss": 0.0133, "step": 20380 }, { "grad_norm": 0.11809487640857697, "learning_rate": 8.895811889788994e-05, "loss": 0.0145, "step": 20390 }, { "grad_norm": 0.1535244584083557, "learning_rate": 8.894516017630692e-05, "loss": 0.0138, "step": 20400 }, { "grad_norm": 0.12109021097421646, "learning_rate": 8.893219480006118e-05, "loss": 0.0123, "step": 20410 }, { "grad_norm": 0.10847415775060654, "learning_rate": 8.891922277136817e-05, "loss": 0.0088, "step": 20420 }, { "grad_norm": 0.205900177359581, "learning_rate": 8.890624409244441e-05, "loss": 0.0093, "step": 20430 }, { "grad_norm": 0.19330404698848724, "learning_rate": 8.889325876550763e-05, "loss": 0.0124, "step": 20440 }, { "grad_norm": 0.16080093383789062, "learning_rate": 8.888026679277666e-05, "loss": 0.0098, "step": 20450 }, { "grad_norm": 0.221133753657341, "learning_rate": 8.886726817647147e-05, "loss": 0.015, "step": 20460 }, { "grad_norm": 0.21844372153282166, "learning_rate": 8.885426291881319e-05, "loss": 0.0135, "step": 20470 }, { "grad_norm": 0.22213339805603027, "learning_rate": 8.884125102202401e-05, "loss": 0.0148, "step": 20480 }, { "grad_norm": 0.2356795072555542, "learning_rate": 8.882823248832736e-05, "loss": 0.0151, "step": 20490 }, { "grad_norm": 0.16801230609416962, "learning_rate": 8.881520731994772e-05, "loss": 0.0135, "step": 20500 }, { "grad_norm": 0.18454387784004211, "learning_rate": 8.880217551911077e-05, "loss": 0.0114, "step": 20510 }, { "grad_norm": 0.20877902209758759, "learning_rate": 8.878913708804323e-05, "loss": 0.0167, "step": 20520 }, { "grad_norm": 0.20192131400108337, "learning_rate": 8.877609202897308e-05, "loss": 0.0153, "step": 20530 }, { "grad_norm": 0.11031981557607651, "learning_rate": 8.876304034412933e-05, "loss": 0.01, "step": 20540 }, { "grad_norm": 0.18772821128368378, "learning_rate": 8.874998203574214e-05, "loss": 0.0136, "step": 20550 }, { "grad_norm": 0.1628429889678955, "learning_rate": 8.873691710604284e-05, "loss": 0.0149, "step": 20560 }, { "grad_norm": 0.18291835486888885, "learning_rate": 8.872384555726387e-05, "loss": 0.0135, "step": 20570 }, { "grad_norm": 0.26137998700141907, "learning_rate": 8.871076739163878e-05, "loss": 0.0155, "step": 20580 }, { "grad_norm": 0.16331322491168976, "learning_rate": 8.86976826114023e-05, "loss": 0.0117, "step": 20590 }, { "grad_norm": 0.16924898326396942, "learning_rate": 8.868459121879023e-05, "loss": 0.0111, "step": 20600 }, { "grad_norm": 0.18220050632953644, "learning_rate": 8.867149321603956e-05, "loss": 0.0115, "step": 20610 }, { "grad_norm": 0.1777925342321396, "learning_rate": 8.865838860538835e-05, "loss": 0.0093, "step": 20620 }, { "grad_norm": 0.1656009554862976, "learning_rate": 8.864527738907585e-05, "loss": 0.0145, "step": 20630 }, { "grad_norm": 0.15868161618709564, "learning_rate": 8.863215956934239e-05, "loss": 0.0123, "step": 20640 }, { "grad_norm": 0.1554887294769287, "learning_rate": 8.861903514842947e-05, "loss": 0.0134, "step": 20650 }, { "grad_norm": 0.16340544819831848, "learning_rate": 8.860590412857966e-05, "loss": 0.015, "step": 20660 }, { "grad_norm": 0.19506396353244781, "learning_rate": 8.85927665120367e-05, "loss": 0.0152, "step": 20670 }, { "grad_norm": 0.19622743129730225, "learning_rate": 8.857962230104546e-05, "loss": 0.0128, "step": 20680 }, { "grad_norm": 0.15882591903209686, "learning_rate": 8.856647149785193e-05, "loss": 0.0122, "step": 20690 }, { "grad_norm": 0.1935979127883911, "learning_rate": 8.855331410470322e-05, "loss": 0.0134, "step": 20700 }, { "grad_norm": 0.19893115758895874, "learning_rate": 8.854015012384756e-05, "loss": 0.014, "step": 20710 }, { "grad_norm": 0.20435203611850739, "learning_rate": 8.852697955753433e-05, "loss": 0.0125, "step": 20720 }, { "grad_norm": 0.1716882288455963, "learning_rate": 8.851380240801399e-05, "loss": 0.0143, "step": 20730 }, { "grad_norm": 0.2167215496301651, "learning_rate": 8.850061867753818e-05, "loss": 0.0148, "step": 20740 }, { "grad_norm": 0.1908976435661316, "learning_rate": 8.848742836835963e-05, "loss": 0.0124, "step": 20750 }, { "grad_norm": 0.19660796225070953, "learning_rate": 8.847423148273221e-05, "loss": 0.0133, "step": 20760 }, { "grad_norm": 0.17097246646881104, "learning_rate": 8.846102802291092e-05, "loss": 0.0153, "step": 20770 }, { "grad_norm": 0.14276312291622162, "learning_rate": 8.844781799115183e-05, "loss": 0.0103, "step": 20780 }, { "grad_norm": 0.1550275683403015, "learning_rate": 8.84346013897122e-05, "loss": 0.0115, "step": 20790 }, { "grad_norm": 0.23897235095500946, "learning_rate": 8.842137822085038e-05, "loss": 0.0108, "step": 20800 }, { "grad_norm": 0.22728176414966583, "learning_rate": 8.840814848682585e-05, "loss": 0.0102, "step": 20810 }, { "grad_norm": 0.1892882138490677, "learning_rate": 8.83949121898992e-05, "loss": 0.0114, "step": 20820 }, { "grad_norm": 0.13951954245567322, "learning_rate": 8.838166933233217e-05, "loss": 0.0122, "step": 20830 }, { "grad_norm": 0.18027573823928833, "learning_rate": 8.83684199163876e-05, "loss": 0.0107, "step": 20840 }, { "grad_norm": 0.161132350564003, "learning_rate": 8.835516394432943e-05, "loss": 0.0107, "step": 20850 }, { "grad_norm": 0.1802576333284378, "learning_rate": 8.834190141842276e-05, "loss": 0.0123, "step": 20860 }, { "grad_norm": 0.1426866501569748, "learning_rate": 8.83286323409338e-05, "loss": 0.0114, "step": 20870 }, { "grad_norm": 0.2029677778482437, "learning_rate": 8.831535671412986e-05, "loss": 0.0121, "step": 20880 }, { "grad_norm": 0.18391336500644684, "learning_rate": 8.830207454027938e-05, "loss": 0.0122, "step": 20890 }, { "grad_norm": 0.15209537744522095, "learning_rate": 8.828878582165192e-05, "loss": 0.0104, "step": 20900 }, { "grad_norm": 0.19344407320022583, "learning_rate": 8.827549056051818e-05, "loss": 0.0125, "step": 20910 }, { "grad_norm": 0.15876570343971252, "learning_rate": 8.826218875914993e-05, "loss": 0.0113, "step": 20920 }, { "grad_norm": 0.15806300938129425, "learning_rate": 8.82488804198201e-05, "loss": 0.0114, "step": 20930 }, { "grad_norm": 0.15650320053100586, "learning_rate": 8.82355655448027e-05, "loss": 0.0153, "step": 20940 }, { "grad_norm": 0.17902247607707977, "learning_rate": 8.822224413637293e-05, "loss": 0.0101, "step": 20950 }, { "grad_norm": 0.1479729264974594, "learning_rate": 8.820891619680697e-05, "loss": 0.0115, "step": 20960 }, { "grad_norm": 0.21312180161476135, "learning_rate": 8.819558172838227e-05, "loss": 0.0116, "step": 20970 }, { "grad_norm": 0.11714068800210953, "learning_rate": 8.818224073337731e-05, "loss": 0.0122, "step": 20980 }, { "grad_norm": 0.18330101668834686, "learning_rate": 8.816889321407169e-05, "loss": 0.0098, "step": 20990 }, { "grad_norm": 0.13438500463962555, "learning_rate": 8.815553917274615e-05, "loss": 0.0144, "step": 21000 }, { "grad_norm": 0.18965138494968414, "learning_rate": 8.81421786116825e-05, "loss": 0.0114, "step": 21010 }, { "grad_norm": 0.2047502100467682, "learning_rate": 8.812881153316373e-05, "loss": 0.0134, "step": 21020 }, { "grad_norm": 0.14362786710262299, "learning_rate": 8.81154379394739e-05, "loss": 0.0117, "step": 21030 }, { "grad_norm": 0.19027529656887054, "learning_rate": 8.810205783289818e-05, "loss": 0.0123, "step": 21040 }, { "grad_norm": 0.26225072145462036, "learning_rate": 8.808867121572286e-05, "loss": 0.0152, "step": 21050 }, { "grad_norm": 0.19097627699375153, "learning_rate": 8.807527809023537e-05, "loss": 0.0109, "step": 21060 }, { "grad_norm": 0.16103079915046692, "learning_rate": 8.80618784587242e-05, "loss": 0.0157, "step": 21070 }, { "grad_norm": 0.1791510134935379, "learning_rate": 8.804847232347902e-05, "loss": 0.0128, "step": 21080 }, { "grad_norm": 0.24483193457126617, "learning_rate": 8.803505968679054e-05, "loss": 0.0124, "step": 21090 }, { "grad_norm": 0.16821818053722382, "learning_rate": 8.802164055095061e-05, "loss": 0.0122, "step": 21100 }, { "grad_norm": 0.17985224723815918, "learning_rate": 8.80082149182522e-05, "loss": 0.0147, "step": 21110 }, { "grad_norm": 0.20496881008148193, "learning_rate": 8.79947827909894e-05, "loss": 0.0107, "step": 21120 }, { "grad_norm": 0.15553079545497894, "learning_rate": 8.798134417145738e-05, "loss": 0.0133, "step": 21130 }, { "grad_norm": 0.2669253647327423, "learning_rate": 8.796789906195243e-05, "loss": 0.0137, "step": 21140 }, { "grad_norm": 0.17205627262592316, "learning_rate": 8.795444746477195e-05, "loss": 0.0139, "step": 21150 }, { "grad_norm": 0.15072250366210938, "learning_rate": 8.794098938221446e-05, "loss": 0.0121, "step": 21160 }, { "grad_norm": 0.11938054114580154, "learning_rate": 8.792752481657957e-05, "loss": 0.0119, "step": 21170 }, { "grad_norm": 0.2036229372024536, "learning_rate": 8.791405377016802e-05, "loss": 0.0123, "step": 21180 }, { "grad_norm": 0.22304382920265198, "learning_rate": 8.790057624528163e-05, "loss": 0.0138, "step": 21190 }, { "grad_norm": 0.1627958118915558, "learning_rate": 8.788709224422333e-05, "loss": 0.0129, "step": 21200 }, { "grad_norm": 0.18219949305057526, "learning_rate": 8.787360176929717e-05, "loss": 0.0112, "step": 21210 }, { "grad_norm": 0.15502268075942993, "learning_rate": 8.786010482280834e-05, "loss": 0.0122, "step": 21220 }, { "grad_norm": 0.17360511422157288, "learning_rate": 8.784660140706306e-05, "loss": 0.0129, "step": 21230 }, { "grad_norm": 0.1540069282054901, "learning_rate": 8.783309152436872e-05, "loss": 0.0117, "step": 21240 }, { "grad_norm": 0.15816989541053772, "learning_rate": 8.781957517703375e-05, "loss": 0.011, "step": 21250 }, { "grad_norm": 0.13983385264873505, "learning_rate": 8.780605236736776e-05, "loss": 0.0108, "step": 21260 }, { "grad_norm": 0.16931375861167908, "learning_rate": 8.779252309768142e-05, "loss": 0.0137, "step": 21270 }, { "grad_norm": 0.16367901861667633, "learning_rate": 8.777898737028652e-05, "loss": 0.0129, "step": 21280 }, { "grad_norm": 0.18163257837295532, "learning_rate": 8.776544518749591e-05, "loss": 0.0121, "step": 21290 }, { "grad_norm": 0.17724500596523285, "learning_rate": 8.775189655162364e-05, "loss": 0.0102, "step": 21300 }, { "grad_norm": 0.23525984585285187, "learning_rate": 8.773834146498474e-05, "loss": 0.0137, "step": 21310 }, { "grad_norm": 0.1470358818769455, "learning_rate": 8.772477992989545e-05, "loss": 0.0121, "step": 21320 }, { "grad_norm": 0.1652807593345642, "learning_rate": 8.771121194867304e-05, "loss": 0.0085, "step": 21330 }, { "grad_norm": 0.15392349660396576, "learning_rate": 8.769763752363589e-05, "loss": 0.0124, "step": 21340 }, { "grad_norm": 0.16219785809516907, "learning_rate": 8.768405665710352e-05, "loss": 0.0132, "step": 21350 }, { "grad_norm": 0.18207646906375885, "learning_rate": 8.767046935139655e-05, "loss": 0.0117, "step": 21360 }, { "grad_norm": 0.18561969697475433, "learning_rate": 8.765687560883666e-05, "loss": 0.0117, "step": 21370 }, { "grad_norm": 0.18936818838119507, "learning_rate": 8.764327543174664e-05, "loss": 0.0106, "step": 21380 }, { "grad_norm": 0.16137167811393738, "learning_rate": 8.762966882245038e-05, "loss": 0.0139, "step": 21390 }, { "grad_norm": 0.18472124636173248, "learning_rate": 8.761605578327291e-05, "loss": 0.0128, "step": 21400 }, { "grad_norm": 0.16547517478466034, "learning_rate": 8.76024363165403e-05, "loss": 0.0138, "step": 21410 }, { "grad_norm": 0.21528850495815277, "learning_rate": 8.758881042457976e-05, "loss": 0.0128, "step": 21420 }, { "grad_norm": 0.17191541194915771, "learning_rate": 8.757517810971957e-05, "loss": 0.0124, "step": 21430 }, { "grad_norm": 0.17188243567943573, "learning_rate": 8.756153937428913e-05, "loss": 0.0138, "step": 21440 }, { "grad_norm": 0.15001462399959564, "learning_rate": 8.754789422061889e-05, "loss": 0.0151, "step": 21450 }, { "grad_norm": 0.1504221111536026, "learning_rate": 8.753424265104052e-05, "loss": 0.0109, "step": 21460 }, { "grad_norm": 0.1742183119058609, "learning_rate": 8.752058466788659e-05, "loss": 0.0099, "step": 21470 }, { "grad_norm": 0.1738952100276947, "learning_rate": 8.750692027349097e-05, "loss": 0.011, "step": 21480 }, { "grad_norm": 0.13260243833065033, "learning_rate": 8.749324947018847e-05, "loss": 0.012, "step": 21490 }, { "grad_norm": 0.14603668451309204, "learning_rate": 8.747957226031507e-05, "loss": 0.0134, "step": 21500 }, { "grad_norm": 0.2202417105436325, "learning_rate": 8.746588864620787e-05, "loss": 0.0099, "step": 21510 }, { "grad_norm": 0.21268969774246216, "learning_rate": 8.745219863020498e-05, "loss": 0.0116, "step": 21520 }, { "grad_norm": 0.15866747498512268, "learning_rate": 8.743850221464564e-05, "loss": 0.0121, "step": 21530 }, { "grad_norm": 0.18437452614307404, "learning_rate": 8.742479940187026e-05, "loss": 0.0134, "step": 21540 }, { "grad_norm": 0.17186127603054047, "learning_rate": 8.74110901942202e-05, "loss": 0.0112, "step": 21550 }, { "grad_norm": 0.20806816220283508, "learning_rate": 8.739737459403803e-05, "loss": 0.0127, "step": 21560 }, { "grad_norm": 0.1715961992740631, "learning_rate": 8.738365260366737e-05, "loss": 0.0108, "step": 21570 }, { "grad_norm": 0.1411212682723999, "learning_rate": 8.736992422545292e-05, "loss": 0.0143, "step": 21580 }, { "grad_norm": 0.25738608837127686, "learning_rate": 8.73561894617405e-05, "loss": 0.0127, "step": 21590 }, { "grad_norm": 0.18163205683231354, "learning_rate": 8.734244831487697e-05, "loss": 0.0142, "step": 21600 }, { "grad_norm": 0.16528554260730743, "learning_rate": 8.732870078721035e-05, "loss": 0.0114, "step": 21610 }, { "grad_norm": 0.16956190764904022, "learning_rate": 8.731494688108972e-05, "loss": 0.0122, "step": 21620 }, { "grad_norm": 0.17834685742855072, "learning_rate": 8.730118659886523e-05, "loss": 0.0126, "step": 21630 }, { "grad_norm": 0.20271369814872742, "learning_rate": 8.728741994288814e-05, "loss": 0.0121, "step": 21640 }, { "grad_norm": 0.1539432555437088, "learning_rate": 8.727364691551079e-05, "loss": 0.0101, "step": 21650 }, { "grad_norm": 0.19114767014980316, "learning_rate": 8.725986751908661e-05, "loss": 0.012, "step": 21660 }, { "grad_norm": 0.16144928336143494, "learning_rate": 8.724608175597016e-05, "loss": 0.0119, "step": 21670 }, { "grad_norm": 0.18769580125808716, "learning_rate": 8.723228962851699e-05, "loss": 0.0116, "step": 21680 }, { "grad_norm": 0.17372167110443115, "learning_rate": 8.721849113908385e-05, "loss": 0.0124, "step": 21690 }, { "grad_norm": 0.15132175385951996, "learning_rate": 8.720468629002848e-05, "loss": 0.0123, "step": 21700 }, { "grad_norm": 0.14092087745666504, "learning_rate": 8.719087508370978e-05, "loss": 0.0123, "step": 21710 }, { "grad_norm": 0.2128206342458725, "learning_rate": 8.717705752248772e-05, "loss": 0.017, "step": 21720 }, { "grad_norm": 0.16288234293460846, "learning_rate": 8.71632336087233e-05, "loss": 0.0139, "step": 21730 }, { "grad_norm": 0.1330074667930603, "learning_rate": 8.71494033447787e-05, "loss": 0.0102, "step": 21740 }, { "grad_norm": 0.16808564960956573, "learning_rate": 8.713556673301708e-05, "loss": 0.0111, "step": 21750 }, { "grad_norm": 0.17865179479122162, "learning_rate": 8.712172377580278e-05, "loss": 0.012, "step": 21760 }, { "grad_norm": 0.14618156850337982, "learning_rate": 8.710787447550114e-05, "loss": 0.0138, "step": 21770 }, { "grad_norm": 0.22812049090862274, "learning_rate": 8.70940188344787e-05, "loss": 0.0114, "step": 21780 }, { "grad_norm": 0.1902359575033188, "learning_rate": 8.708015685510293e-05, "loss": 0.0098, "step": 21790 }, { "grad_norm": 0.21198885142803192, "learning_rate": 8.706628853974252e-05, "loss": 0.0104, "step": 21800 }, { "grad_norm": 0.17521412670612335, "learning_rate": 8.705241389076715e-05, "loss": 0.0129, "step": 21810 }, { "grad_norm": 0.14795149862766266, "learning_rate": 8.703853291054764e-05, "loss": 0.0099, "step": 21820 }, { "grad_norm": 0.1830134093761444, "learning_rate": 8.702464560145587e-05, "loss": 0.0114, "step": 21830 }, { "grad_norm": 0.19202940165996552, "learning_rate": 8.701075196586476e-05, "loss": 0.0105, "step": 21840 }, { "grad_norm": 0.19199828803539276, "learning_rate": 8.699685200614842e-05, "loss": 0.0132, "step": 21850 }, { "grad_norm": 0.11773224174976349, "learning_rate": 8.698294572468193e-05, "loss": 0.0107, "step": 21860 }, { "grad_norm": 0.18651705980300903, "learning_rate": 8.696903312384148e-05, "loss": 0.0126, "step": 21870 }, { "grad_norm": 0.20121489465236664, "learning_rate": 8.695511420600439e-05, "loss": 0.0109, "step": 21880 }, { "grad_norm": 0.1596883088350296, "learning_rate": 8.694118897354901e-05, "loss": 0.0118, "step": 21890 }, { "grad_norm": 0.1554563045501709, "learning_rate": 8.692725742885478e-05, "loss": 0.0121, "step": 21900 }, { "grad_norm": 0.17439478635787964, "learning_rate": 8.691331957430221e-05, "loss": 0.0113, "step": 21910 }, { "grad_norm": 0.19215288758277893, "learning_rate": 8.68993754122729e-05, "loss": 0.0104, "step": 21920 }, { "grad_norm": 0.18130524456501007, "learning_rate": 8.688542494514955e-05, "loss": 0.0115, "step": 21930 }, { "grad_norm": 0.13448356091976166, "learning_rate": 8.68714681753159e-05, "loss": 0.0116, "step": 21940 }, { "grad_norm": 0.1495244801044464, "learning_rate": 8.685750510515676e-05, "loss": 0.0089, "step": 21950 }, { "grad_norm": 0.1547326296567917, "learning_rate": 8.684353573705805e-05, "loss": 0.0103, "step": 21960 }, { "grad_norm": 0.14756140112876892, "learning_rate": 8.682956007340677e-05, "loss": 0.0136, "step": 21970 }, { "grad_norm": 0.1505957543849945, "learning_rate": 8.681557811659095e-05, "loss": 0.013, "step": 21980 }, { "grad_norm": 0.15310151875019073, "learning_rate": 8.680158986899974e-05, "loss": 0.012, "step": 21990 }, { "grad_norm": 0.15834444761276245, "learning_rate": 8.678759533302335e-05, "loss": 0.0088, "step": 22000 }, { "grad_norm": 0.1580670326948166, "learning_rate": 8.677359451105308e-05, "loss": 0.0118, "step": 22010 }, { "grad_norm": 0.1631334125995636, "learning_rate": 8.675958740548123e-05, "loss": 0.0113, "step": 22020 }, { "grad_norm": 0.2120421975851059, "learning_rate": 8.674557401870129e-05, "loss": 0.0115, "step": 22030 }, { "grad_norm": 0.20874981582164764, "learning_rate": 8.673155435310775e-05, "loss": 0.0128, "step": 22040 }, { "grad_norm": 0.17149226367473602, "learning_rate": 8.671752841109617e-05, "loss": 0.0102, "step": 22050 }, { "grad_norm": 0.15252189338207245, "learning_rate": 8.670349619506321e-05, "loss": 0.0103, "step": 22060 }, { "grad_norm": 0.22473566234111786, "learning_rate": 8.66894577074066e-05, "loss": 0.0117, "step": 22070 }, { "grad_norm": 0.17181867361068726, "learning_rate": 8.667541295052513e-05, "loss": 0.0105, "step": 22080 }, { "grad_norm": 0.24931295216083527, "learning_rate": 8.666136192681865e-05, "loss": 0.0114, "step": 22090 }, { "grad_norm": 0.1478455364704132, "learning_rate": 8.664730463868811e-05, "loss": 0.014, "step": 22100 }, { "grad_norm": 0.1785571128129959, "learning_rate": 8.663324108853552e-05, "loss": 0.0102, "step": 22110 }, { "grad_norm": 0.1793355792760849, "learning_rate": 8.661917127876395e-05, "loss": 0.0109, "step": 22120 }, { "grad_norm": 0.233780637383461, "learning_rate": 8.660509521177754e-05, "loss": 0.0134, "step": 22130 }, { "grad_norm": 0.1368517428636551, "learning_rate": 8.65910128899815e-05, "loss": 0.0122, "step": 22140 }, { "grad_norm": 0.18203873932361603, "learning_rate": 8.657692431578214e-05, "loss": 0.0114, "step": 22150 }, { "grad_norm": 0.1858668476343155, "learning_rate": 8.656282949158679e-05, "loss": 0.0103, "step": 22160 }, { "grad_norm": 0.16793303191661835, "learning_rate": 8.654872841980388e-05, "loss": 0.0113, "step": 22170 }, { "grad_norm": 0.14359214901924133, "learning_rate": 8.653462110284289e-05, "loss": 0.0125, "step": 22180 }, { "grad_norm": 0.22780129313468933, "learning_rate": 8.652050754311437e-05, "loss": 0.0109, "step": 22190 }, { "grad_norm": 0.23086465895175934, "learning_rate": 8.650638774302995e-05, "loss": 0.0121, "step": 22200 }, { "grad_norm": 0.1767311990261078, "learning_rate": 8.649226170500233e-05, "loss": 0.0115, "step": 22210 }, { "grad_norm": 0.2072315365076065, "learning_rate": 8.647812943144524e-05, "loss": 0.0141, "step": 22220 }, { "grad_norm": 0.19007259607315063, "learning_rate": 8.646399092477351e-05, "loss": 0.0113, "step": 22230 }, { "grad_norm": 0.1681956946849823, "learning_rate": 8.644984618740301e-05, "loss": 0.012, "step": 22240 }, { "grad_norm": 0.1733224093914032, "learning_rate": 8.643569522175073e-05, "loss": 0.0169, "step": 22250 }, { "grad_norm": 0.13537786900997162, "learning_rate": 8.642153803023463e-05, "loss": 0.0135, "step": 22260 }, { "grad_norm": 0.15664394199848175, "learning_rate": 8.640737461527383e-05, "loss": 0.0099, "step": 22270 }, { "grad_norm": 0.1840028017759323, "learning_rate": 8.639320497928845e-05, "loss": 0.0111, "step": 22280 }, { "grad_norm": 0.18565016984939575, "learning_rate": 8.637902912469969e-05, "loss": 0.0115, "step": 22290 }, { "grad_norm": 0.15818729996681213, "learning_rate": 8.636484705392982e-05, "loss": 0.0111, "step": 22300 }, { "grad_norm": 0.12230047583580017, "learning_rate": 8.635065876940216e-05, "loss": 0.0113, "step": 22310 }, { "grad_norm": 0.1726868748664856, "learning_rate": 8.633646427354112e-05, "loss": 0.0109, "step": 22320 }, { "grad_norm": 0.17817550897598267, "learning_rate": 8.632226356877213e-05, "loss": 0.0142, "step": 22330 }, { "grad_norm": 0.1810556799173355, "learning_rate": 8.630805665752173e-05, "loss": 0.0115, "step": 22340 }, { "grad_norm": 0.20075325667858124, "learning_rate": 8.629384354221748e-05, "loss": 0.0107, "step": 22350 }, { "grad_norm": 0.1456870436668396, "learning_rate": 8.627962422528797e-05, "loss": 0.0114, "step": 22360 }, { "grad_norm": 0.2009436935186386, "learning_rate": 8.626539870916296e-05, "loss": 0.0117, "step": 22370 }, { "grad_norm": 0.15488146245479584, "learning_rate": 8.625116699627317e-05, "loss": 0.011, "step": 22380 }, { "grad_norm": 0.28004157543182373, "learning_rate": 8.623692908905041e-05, "loss": 0.0141, "step": 22390 }, { "grad_norm": 0.2291289120912552, "learning_rate": 8.622268498992755e-05, "loss": 0.0134, "step": 22400 }, { "grad_norm": 0.1856473684310913, "learning_rate": 8.620843470133851e-05, "loss": 0.011, "step": 22410 }, { "grad_norm": 0.15377479791641235, "learning_rate": 8.619417822571829e-05, "loss": 0.0137, "step": 22420 }, { "grad_norm": 0.22604672610759735, "learning_rate": 8.617991556550292e-05, "loss": 0.0134, "step": 22430 }, { "grad_norm": 0.15820124745368958, "learning_rate": 8.616564672312952e-05, "loss": 0.0135, "step": 22440 }, { "grad_norm": 0.22734926640987396, "learning_rate": 8.61513717010362e-05, "loss": 0.0108, "step": 22450 }, { "grad_norm": 0.16584305465221405, "learning_rate": 8.613709050166221e-05, "loss": 0.0102, "step": 22460 }, { "grad_norm": 0.12389175593852997, "learning_rate": 8.61228031274478e-05, "loss": 0.0104, "step": 22470 }, { "grad_norm": 0.20284563302993774, "learning_rate": 8.610850958083431e-05, "loss": 0.0109, "step": 22480 }, { "grad_norm": 0.1440453678369522, "learning_rate": 8.609420986426409e-05, "loss": 0.0103, "step": 22490 }, { "grad_norm": 0.17537270486354828, "learning_rate": 8.60799039801806e-05, "loss": 0.0109, "step": 22500 }, { "grad_norm": 0.1924111396074295, "learning_rate": 8.606559193102828e-05, "loss": 0.011, "step": 22510 }, { "grad_norm": 0.11461764574050903, "learning_rate": 8.605127371925273e-05, "loss": 0.0102, "step": 22520 }, { "grad_norm": 0.12611624598503113, "learning_rate": 8.603694934730047e-05, "loss": 0.0113, "step": 22530 }, { "grad_norm": 0.1884607970714569, "learning_rate": 8.602261881761919e-05, "loss": 0.0122, "step": 22540 }, { "grad_norm": 0.1911512315273285, "learning_rate": 8.600828213265759e-05, "loss": 0.0106, "step": 22550 }, { "grad_norm": 0.1983567327260971, "learning_rate": 8.599393929486539e-05, "loss": 0.0106, "step": 22560 }, { "grad_norm": 0.13314223289489746, "learning_rate": 8.59795903066934e-05, "loss": 0.0102, "step": 22570 }, { "grad_norm": 0.15824981033802032, "learning_rate": 8.596523517059347e-05, "loss": 0.009, "step": 22580 }, { "grad_norm": 0.16493076086044312, "learning_rate": 8.59508738890185e-05, "loss": 0.0097, "step": 22590 }, { "grad_norm": 0.20265845954418182, "learning_rate": 8.593650646442246e-05, "loss": 0.0111, "step": 22600 }, { "grad_norm": 0.16707660257816315, "learning_rate": 8.59221328992603e-05, "loss": 0.0099, "step": 22610 }, { "grad_norm": 0.15929026901721954, "learning_rate": 8.590775319598813e-05, "loss": 0.0094, "step": 22620 }, { "grad_norm": 0.1718435287475586, "learning_rate": 8.589336735706301e-05, "loss": 0.0099, "step": 22630 }, { "grad_norm": 0.1809978038072586, "learning_rate": 8.587897538494307e-05, "loss": 0.0103, "step": 22640 }, { "grad_norm": 0.15212875604629517, "learning_rate": 8.586457728208756e-05, "loss": 0.0127, "step": 22650 }, { "grad_norm": 0.1901150643825531, "learning_rate": 8.585017305095667e-05, "loss": 0.0115, "step": 22660 }, { "grad_norm": 0.20352181792259216, "learning_rate": 8.583576269401173e-05, "loss": 0.0164, "step": 22670 }, { "grad_norm": 0.19211697578430176, "learning_rate": 8.582134621371504e-05, "loss": 0.0121, "step": 22680 }, { "grad_norm": 0.16774946451187134, "learning_rate": 8.580692361253e-05, "loss": 0.0121, "step": 22690 }, { "grad_norm": 0.1957998275756836, "learning_rate": 8.579249489292104e-05, "loss": 0.0116, "step": 22700 }, { "grad_norm": 0.19476599991321564, "learning_rate": 8.577806005735363e-05, "loss": 0.0128, "step": 22710 }, { "grad_norm": 0.1546115279197693, "learning_rate": 8.576361910829429e-05, "loss": 0.0112, "step": 22720 }, { "grad_norm": 0.2026883214712143, "learning_rate": 8.574917204821057e-05, "loss": 0.0113, "step": 22730 }, { "grad_norm": 0.2536403238773346, "learning_rate": 8.57347188795711e-05, "loss": 0.0116, "step": 22740 }, { "grad_norm": 0.1406298577785492, "learning_rate": 8.572025960484551e-05, "loss": 0.0114, "step": 22750 }, { "grad_norm": 0.1688297986984253, "learning_rate": 8.57057942265045e-05, "loss": 0.0103, "step": 22760 }, { "grad_norm": 0.15677902102470398, "learning_rate": 8.569132274701984e-05, "loss": 0.0123, "step": 22770 }, { "grad_norm": 0.20779938995838165, "learning_rate": 8.567684516886427e-05, "loss": 0.0116, "step": 22780 }, { "grad_norm": 0.1554947942495346, "learning_rate": 8.56623614945116e-05, "loss": 0.012, "step": 22790 }, { "grad_norm": 0.15010464191436768, "learning_rate": 8.564787172643675e-05, "loss": 0.011, "step": 22800 }, { "grad_norm": 0.2211492955684662, "learning_rate": 8.563337586711559e-05, "loss": 0.0152, "step": 22810 }, { "grad_norm": 0.19498886168003082, "learning_rate": 8.561887391902506e-05, "loss": 0.0095, "step": 22820 }, { "grad_norm": 0.24572578072547913, "learning_rate": 8.560436588464316e-05, "loss": 0.0131, "step": 22830 }, { "grad_norm": 0.14142805337905884, "learning_rate": 8.55898517664489e-05, "loss": 0.0105, "step": 22840 }, { "grad_norm": 0.19077281653881073, "learning_rate": 8.557533156692236e-05, "loss": 0.0129, "step": 22850 }, { "grad_norm": 0.18561753630638123, "learning_rate": 8.556080528854467e-05, "loss": 0.0118, "step": 22860 }, { "grad_norm": 0.15746891498565674, "learning_rate": 8.554627293379791e-05, "loss": 0.0116, "step": 22870 }, { "grad_norm": 0.13767066597938538, "learning_rate": 8.553173450516531e-05, "loss": 0.0106, "step": 22880 }, { "grad_norm": 0.17230068147182465, "learning_rate": 8.551719000513108e-05, "loss": 0.0104, "step": 22890 }, { "grad_norm": 0.2923627197742462, "learning_rate": 8.550263943618049e-05, "loss": 0.0133, "step": 22900 }, { "grad_norm": 0.14344632625579834, "learning_rate": 8.54880828007998e-05, "loss": 0.0124, "step": 22910 }, { "grad_norm": 0.1616484373807907, "learning_rate": 8.547352010147637e-05, "loss": 0.0114, "step": 22920 }, { "grad_norm": 0.1854427456855774, "learning_rate": 8.545895134069855e-05, "loss": 0.011, "step": 22930 }, { "grad_norm": 0.1503603607416153, "learning_rate": 8.544437652095576e-05, "loss": 0.0117, "step": 22940 }, { "grad_norm": 0.18264548480510712, "learning_rate": 8.542979564473843e-05, "loss": 0.0128, "step": 22950 }, { "grad_norm": 0.251026451587677, "learning_rate": 8.541520871453802e-05, "loss": 0.0126, "step": 22960 }, { "grad_norm": 0.15979565680027008, "learning_rate": 8.540061573284705e-05, "loss": 0.0122, "step": 22970 }, { "grad_norm": 0.16734836995601654, "learning_rate": 8.538601670215906e-05, "loss": 0.0104, "step": 22980 }, { "grad_norm": 0.13384944200515747, "learning_rate": 8.537141162496864e-05, "loss": 0.0096, "step": 22990 }, { "grad_norm": 0.14679943025112152, "learning_rate": 8.535680050377137e-05, "loss": 0.0114, "step": 23000 }, { "grad_norm": 0.14889559149742126, "learning_rate": 8.534218334106391e-05, "loss": 0.0102, "step": 23010 }, { "grad_norm": 0.1612592190504074, "learning_rate": 8.532756013934393e-05, "loss": 0.0096, "step": 23020 }, { "grad_norm": 0.15890130400657654, "learning_rate": 8.531293090111012e-05, "loss": 0.01, "step": 23030 }, { "grad_norm": 0.16539372503757477, "learning_rate": 8.529829562886225e-05, "loss": 0.0115, "step": 23040 }, { "grad_norm": 0.1926967352628708, "learning_rate": 8.528365432510105e-05, "loss": 0.0102, "step": 23050 }, { "grad_norm": 0.176364004611969, "learning_rate": 8.526900699232833e-05, "loss": 0.0131, "step": 23060 }, { "grad_norm": 0.17542652785778046, "learning_rate": 8.525435363304695e-05, "loss": 0.0127, "step": 23070 }, { "grad_norm": 0.2122577726840973, "learning_rate": 8.523969424976072e-05, "loss": 0.0137, "step": 23080 }, { "grad_norm": 0.23133480548858643, "learning_rate": 8.522502884497457e-05, "loss": 0.0117, "step": 23090 }, { "grad_norm": 0.2058950960636139, "learning_rate": 8.521035742119437e-05, "loss": 0.0125, "step": 23100 }, { "grad_norm": 0.2068047970533371, "learning_rate": 8.519567998092712e-05, "loss": 0.0109, "step": 23110 }, { "grad_norm": 0.1812582165002823, "learning_rate": 8.518099652668075e-05, "loss": 0.0095, "step": 23120 }, { "grad_norm": 0.1492076814174652, "learning_rate": 8.516630706096429e-05, "loss": 0.0106, "step": 23130 }, { "grad_norm": 0.17563295364379883, "learning_rate": 8.515161158628773e-05, "loss": 0.0092, "step": 23140 }, { "grad_norm": 0.12060334533452988, "learning_rate": 8.513691010516216e-05, "loss": 0.0115, "step": 23150 }, { "grad_norm": 0.16632069647312164, "learning_rate": 8.512220262009966e-05, "loss": 0.0097, "step": 23160 }, { "grad_norm": 0.11744379997253418, "learning_rate": 8.510748913361332e-05, "loss": 0.0098, "step": 23170 }, { "grad_norm": 0.15728989243507385, "learning_rate": 8.509276964821726e-05, "loss": 0.0098, "step": 23180 }, { "grad_norm": 0.1730661690235138, "learning_rate": 8.507804416642669e-05, "loss": 0.0127, "step": 23190 }, { "grad_norm": 0.26725634932518005, "learning_rate": 8.506331269075774e-05, "loss": 0.0153, "step": 23200 }, { "grad_norm": 0.23186005651950836, "learning_rate": 8.504857522372765e-05, "loss": 0.0102, "step": 23210 }, { "grad_norm": 0.22802473604679108, "learning_rate": 8.503383176785461e-05, "loss": 0.0094, "step": 23220 }, { "grad_norm": 0.17197832465171814, "learning_rate": 8.501908232565792e-05, "loss": 0.0116, "step": 23230 }, { "grad_norm": 0.15673279762268066, "learning_rate": 8.50043268996578e-05, "loss": 0.0105, "step": 23240 }, { "grad_norm": 0.16538448631763458, "learning_rate": 8.498956549237562e-05, "loss": 0.0087, "step": 23250 }, { "grad_norm": 0.14455200731754303, "learning_rate": 8.497479810633366e-05, "loss": 0.0094, "step": 23260 }, { "grad_norm": 0.1388026922941208, "learning_rate": 8.496002474405525e-05, "loss": 0.0104, "step": 23270 }, { "grad_norm": 0.17085573077201843, "learning_rate": 8.494524540806478e-05, "loss": 0.0118, "step": 23280 }, { "grad_norm": 0.17517629265785217, "learning_rate": 8.493046010088761e-05, "loss": 0.0108, "step": 23290 }, { "grad_norm": 0.1542561799287796, "learning_rate": 8.491566882505018e-05, "loss": 0.0106, "step": 23300 }, { "grad_norm": 0.14408113062381744, "learning_rate": 8.490087158307988e-05, "loss": 0.0109, "step": 23310 }, { "grad_norm": 0.15876217186450958, "learning_rate": 8.488606837750518e-05, "loss": 0.0118, "step": 23320 }, { "grad_norm": 0.16063153743743896, "learning_rate": 8.487125921085552e-05, "loss": 0.0107, "step": 23330 }, { "grad_norm": 0.16772745549678802, "learning_rate": 8.485644408566141e-05, "loss": 0.0098, "step": 23340 }, { "grad_norm": 0.22331467270851135, "learning_rate": 8.484162300445431e-05, "loss": 0.0113, "step": 23350 }, { "grad_norm": 0.20619319379329681, "learning_rate": 8.482679596976676e-05, "loss": 0.0122, "step": 23360 }, { "grad_norm": 0.14231230318546295, "learning_rate": 8.48119629841323e-05, "loss": 0.0111, "step": 23370 }, { "grad_norm": 0.19259950518608093, "learning_rate": 8.479712405008547e-05, "loss": 0.0149, "step": 23380 }, { "grad_norm": 0.14577007293701172, "learning_rate": 8.478227917016184e-05, "loss": 0.0116, "step": 23390 }, { "grad_norm": 0.18833357095718384, "learning_rate": 8.476742834689801e-05, "loss": 0.0121, "step": 23400 }, { "grad_norm": 0.1454228013753891, "learning_rate": 8.475257158283157e-05, "loss": 0.0115, "step": 23410 }, { "grad_norm": 0.17820146679878235, "learning_rate": 8.473770888050112e-05, "loss": 0.0133, "step": 23420 }, { "grad_norm": 0.2220805287361145, "learning_rate": 8.47228402424463e-05, "loss": 0.0135, "step": 23430 }, { "grad_norm": 0.16164423525333405, "learning_rate": 8.470796567120775e-05, "loss": 0.0106, "step": 23440 }, { "grad_norm": 0.16835325956344604, "learning_rate": 8.469308516932714e-05, "loss": 0.0118, "step": 23450 }, { "grad_norm": 0.16889476776123047, "learning_rate": 8.467819873934714e-05, "loss": 0.0121, "step": 23460 }, { "grad_norm": 0.2116212695837021, "learning_rate": 8.466330638381143e-05, "loss": 0.0121, "step": 23470 }, { "grad_norm": 0.20569856464862823, "learning_rate": 8.464840810526469e-05, "loss": 0.0103, "step": 23480 }, { "grad_norm": 0.20824982225894928, "learning_rate": 8.463350390625264e-05, "loss": 0.0082, "step": 23490 }, { "grad_norm": 0.17924469709396362, "learning_rate": 8.4618593789322e-05, "loss": 0.0107, "step": 23500 }, { "grad_norm": 0.15527191758155823, "learning_rate": 8.46036777570205e-05, "loss": 0.0128, "step": 23510 }, { "grad_norm": 0.21046370267868042, "learning_rate": 8.458875581189688e-05, "loss": 0.0123, "step": 23520 }, { "grad_norm": 0.17567460238933563, "learning_rate": 8.457382795650092e-05, "loss": 0.0115, "step": 23530 }, { "grad_norm": 0.16019892692565918, "learning_rate": 8.455889419338335e-05, "loss": 0.0126, "step": 23540 }, { "grad_norm": 0.1488616019487381, "learning_rate": 8.454395452509593e-05, "loss": 0.0107, "step": 23550 }, { "grad_norm": 0.12600074708461761, "learning_rate": 8.452900895419146e-05, "loss": 0.0093, "step": 23560 }, { "grad_norm": 0.1645817756652832, "learning_rate": 8.451405748322376e-05, "loss": 0.0113, "step": 23570 }, { "grad_norm": 0.23329342901706696, "learning_rate": 8.449910011474759e-05, "loss": 0.0148, "step": 23580 }, { "grad_norm": 0.11419038474559784, "learning_rate": 8.448413685131876e-05, "loss": 0.0106, "step": 23590 }, { "grad_norm": 0.14498770236968994, "learning_rate": 8.446916769549407e-05, "loss": 0.0104, "step": 23600 }, { "grad_norm": 0.13731376826763153, "learning_rate": 8.445419264983136e-05, "loss": 0.0102, "step": 23610 }, { "grad_norm": 0.12229356169700623, "learning_rate": 8.443921171688947e-05, "loss": 0.0094, "step": 23620 }, { "grad_norm": 0.1615619659423828, "learning_rate": 8.442422489922819e-05, "loss": 0.0108, "step": 23630 }, { "grad_norm": 0.1763504147529602, "learning_rate": 8.440923219940838e-05, "loss": 0.0093, "step": 23640 }, { "grad_norm": 0.14945580065250397, "learning_rate": 8.439423361999189e-05, "loss": 0.0094, "step": 23650 }, { "grad_norm": 0.1977214366197586, "learning_rate": 8.437922916354155e-05, "loss": 0.0089, "step": 23660 }, { "grad_norm": 0.1581323891878128, "learning_rate": 8.436421883262123e-05, "loss": 0.0109, "step": 23670 }, { "grad_norm": 0.17789240181446075, "learning_rate": 8.434920262979577e-05, "loss": 0.0085, "step": 23680 }, { "grad_norm": 0.24832327663898468, "learning_rate": 8.433418055763104e-05, "loss": 0.0111, "step": 23690 }, { "grad_norm": 0.14845257997512817, "learning_rate": 8.431915261869389e-05, "loss": 0.0109, "step": 23700 }, { "grad_norm": 0.1570783108472824, "learning_rate": 8.43041188155522e-05, "loss": 0.0088, "step": 23710 }, { "grad_norm": 0.17776094377040863, "learning_rate": 8.428907915077481e-05, "loss": 0.0128, "step": 23720 }, { "grad_norm": 0.19605042040348053, "learning_rate": 8.42740336269316e-05, "loss": 0.0104, "step": 23730 }, { "grad_norm": 0.1313427835702896, "learning_rate": 8.425898224659345e-05, "loss": 0.0117, "step": 23740 }, { "grad_norm": 0.1372910439968109, "learning_rate": 8.42439250123322e-05, "loss": 0.0089, "step": 23750 }, { "grad_norm": 0.1719067394733429, "learning_rate": 8.422886192672076e-05, "loss": 0.0113, "step": 23760 }, { "grad_norm": 0.14942771196365356, "learning_rate": 8.421379299233297e-05, "loss": 0.0103, "step": 23770 }, { "grad_norm": 0.19460433721542358, "learning_rate": 8.419871821174371e-05, "loss": 0.011, "step": 23780 }, { "grad_norm": 0.18079499900341034, "learning_rate": 8.418363758752884e-05, "loss": 0.0115, "step": 23790 }, { "grad_norm": 0.15080396831035614, "learning_rate": 8.416855112226523e-05, "loss": 0.0122, "step": 23800 }, { "grad_norm": 0.19557367265224457, "learning_rate": 8.415345881853075e-05, "loss": 0.0104, "step": 23810 }, { "grad_norm": 0.1802893429994583, "learning_rate": 8.413836067890426e-05, "loss": 0.0146, "step": 23820 }, { "grad_norm": 0.15868273377418518, "learning_rate": 8.41232567059656e-05, "loss": 0.009, "step": 23830 }, { "grad_norm": 0.09781438112258911, "learning_rate": 8.410814690229565e-05, "loss": 0.0116, "step": 23840 }, { "grad_norm": 0.15066425502300262, "learning_rate": 8.409303127047626e-05, "loss": 0.0101, "step": 23850 }, { "grad_norm": 0.2178894430398941, "learning_rate": 8.407790981309028e-05, "loss": 0.0096, "step": 23860 }, { "grad_norm": 0.15759237110614777, "learning_rate": 8.406278253272153e-05, "loss": 0.0111, "step": 23870 }, { "grad_norm": 0.1472923457622528, "learning_rate": 8.404764943195487e-05, "loss": 0.0097, "step": 23880 }, { "grad_norm": 0.16229593753814697, "learning_rate": 8.403251051337613e-05, "loss": 0.0117, "step": 23890 }, { "grad_norm": 0.17590780556201935, "learning_rate": 8.401736577957214e-05, "loss": 0.0113, "step": 23900 }, { "grad_norm": 0.17587822675704956, "learning_rate": 8.40022152331307e-05, "loss": 0.0101, "step": 23910 }, { "grad_norm": 0.1344776749610901, "learning_rate": 8.398705887664064e-05, "loss": 0.0098, "step": 23920 }, { "grad_norm": 0.17602264881134033, "learning_rate": 8.397189671269177e-05, "loss": 0.0093, "step": 23930 }, { "grad_norm": 0.14407986402511597, "learning_rate": 8.395672874387488e-05, "loss": 0.0111, "step": 23940 }, { "grad_norm": 0.22279976308345795, "learning_rate": 8.394155497278177e-05, "loss": 0.0116, "step": 23950 }, { "grad_norm": 0.13037890195846558, "learning_rate": 8.392637540200523e-05, "loss": 0.0122, "step": 23960 }, { "grad_norm": 0.17852269113063812, "learning_rate": 8.391119003413902e-05, "loss": 0.0085, "step": 23970 }, { "grad_norm": 0.16656631231307983, "learning_rate": 8.38959988717779e-05, "loss": 0.0083, "step": 23980 }, { "grad_norm": 0.16915909945964813, "learning_rate": 8.388080191751764e-05, "loss": 0.0106, "step": 23990 }, { "grad_norm": 0.15743836760520935, "learning_rate": 8.386559917395496e-05, "loss": 0.0105, "step": 24000 }, { "grad_norm": 0.12927772104740143, "learning_rate": 8.385039064368761e-05, "loss": 0.0102, "step": 24010 }, { "grad_norm": 0.14714080095291138, "learning_rate": 8.383517632931431e-05, "loss": 0.0109, "step": 24020 }, { "grad_norm": 0.1398756504058838, "learning_rate": 8.381995623343477e-05, "loss": 0.0097, "step": 24030 }, { "grad_norm": 0.1917419135570526, "learning_rate": 8.380473035864968e-05, "loss": 0.011, "step": 24040 }, { "grad_norm": 0.15769138932228088, "learning_rate": 8.378949870756076e-05, "loss": 0.0111, "step": 24050 }, { "grad_norm": 0.18541298806667328, "learning_rate": 8.377426128277063e-05, "loss": 0.0123, "step": 24060 }, { "grad_norm": 0.1649526208639145, "learning_rate": 8.375901808688298e-05, "loss": 0.0101, "step": 24070 }, { "grad_norm": 0.15239505469799042, "learning_rate": 8.374376912250246e-05, "loss": 0.008, "step": 24080 }, { "grad_norm": 0.15988491475582123, "learning_rate": 8.372851439223468e-05, "loss": 0.0116, "step": 24090 }, { "grad_norm": 0.25723856687545776, "learning_rate": 8.371325389868627e-05, "loss": 0.0103, "step": 24100 }, { "grad_norm": 0.16710491478443146, "learning_rate": 8.369798764446482e-05, "loss": 0.0105, "step": 24110 }, { "grad_norm": 0.15474209189414978, "learning_rate": 8.368271563217893e-05, "loss": 0.0108, "step": 24120 }, { "grad_norm": 0.14825904369354248, "learning_rate": 8.366743786443817e-05, "loss": 0.0098, "step": 24130 }, { "grad_norm": 0.12993289530277252, "learning_rate": 8.365215434385309e-05, "loss": 0.0101, "step": 24140 }, { "grad_norm": 0.11262226849794388, "learning_rate": 8.36368650730352e-05, "loss": 0.0092, "step": 24150 }, { "grad_norm": 0.1331234723329544, "learning_rate": 8.362157005459705e-05, "loss": 0.0108, "step": 24160 }, { "grad_norm": 0.15510597825050354, "learning_rate": 8.360626929115213e-05, "loss": 0.0095, "step": 24170 }, { "grad_norm": 0.19491632282733917, "learning_rate": 8.359096278531492e-05, "loss": 0.0143, "step": 24180 }, { "grad_norm": 0.15635652840137482, "learning_rate": 8.357565053970088e-05, "loss": 0.0139, "step": 24190 }, { "grad_norm": 0.19012346863746643, "learning_rate": 8.356033255692647e-05, "loss": 0.0098, "step": 24200 }, { "grad_norm": 0.14673206210136414, "learning_rate": 8.354500883960911e-05, "loss": 0.0095, "step": 24210 }, { "grad_norm": 0.15664273500442505, "learning_rate": 8.352967939036717e-05, "loss": 0.0134, "step": 24220 }, { "grad_norm": 0.16675183176994324, "learning_rate": 8.35143442118201e-05, "loss": 0.0122, "step": 24230 }, { "grad_norm": 0.17623554170131683, "learning_rate": 8.349900330658819e-05, "loss": 0.0109, "step": 24240 }, { "grad_norm": 0.1888611614704132, "learning_rate": 8.348365667729284e-05, "loss": 0.0113, "step": 24250 }, { "grad_norm": 0.18289506435394287, "learning_rate": 8.346830432655633e-05, "loss": 0.0104, "step": 24260 }, { "grad_norm": 0.17709991335868835, "learning_rate": 8.345294625700195e-05, "loss": 0.0091, "step": 24270 }, { "grad_norm": 0.2187710702419281, "learning_rate": 8.343758247125402e-05, "loss": 0.0105, "step": 24280 }, { "grad_norm": 0.1878693401813507, "learning_rate": 8.342221297193776e-05, "loss": 0.0142, "step": 24290 }, { "grad_norm": 0.18010495603084564, "learning_rate": 8.34068377616794e-05, "loss": 0.0111, "step": 24300 }, { "grad_norm": 0.1802467256784439, "learning_rate": 8.339145684310615e-05, "loss": 0.0112, "step": 24310 }, { "grad_norm": 0.13666614890098572, "learning_rate": 8.337607021884618e-05, "loss": 0.01, "step": 24320 }, { "grad_norm": 0.16156360507011414, "learning_rate": 8.336067789152867e-05, "loss": 0.01, "step": 24330 }, { "grad_norm": 0.18044811487197876, "learning_rate": 8.334527986378369e-05, "loss": 0.0114, "step": 24340 }, { "grad_norm": 0.1496366709470749, "learning_rate": 8.332987613824239e-05, "loss": 0.0102, "step": 24350 }, { "grad_norm": 0.18646863102912903, "learning_rate": 8.331446671753685e-05, "loss": 0.0123, "step": 24360 }, { "grad_norm": 0.1259274184703827, "learning_rate": 8.329905160430007e-05, "loss": 0.0119, "step": 24370 }, { "grad_norm": 0.1868606060743332, "learning_rate": 8.328363080116611e-05, "loss": 0.0112, "step": 24380 }, { "grad_norm": 0.19441388547420502, "learning_rate": 8.326820431076997e-05, "loss": 0.0112, "step": 24390 }, { "grad_norm": 0.1196989119052887, "learning_rate": 8.325277213574759e-05, "loss": 0.0093, "step": 24400 }, { "grad_norm": 0.19461920857429504, "learning_rate": 8.32373342787359e-05, "loss": 0.0121, "step": 24410 }, { "grad_norm": 0.17258450388908386, "learning_rate": 8.322189074237285e-05, "loss": 0.0102, "step": 24420 }, { "grad_norm": 0.19837674498558044, "learning_rate": 8.32064415292973e-05, "loss": 0.0122, "step": 24430 }, { "grad_norm": 0.20383796095848083, "learning_rate": 8.319098664214907e-05, "loss": 0.0128, "step": 24440 }, { "grad_norm": 0.16266584396362305, "learning_rate": 8.3175526083569e-05, "loss": 0.0109, "step": 24450 }, { "grad_norm": 0.14719462394714355, "learning_rate": 8.316005985619889e-05, "loss": 0.0094, "step": 24460 }, { "grad_norm": 0.17792566120624542, "learning_rate": 8.314458796268147e-05, "loss": 0.0101, "step": 24470 }, { "grad_norm": 0.14468854665756226, "learning_rate": 8.312911040566047e-05, "loss": 0.0098, "step": 24480 }, { "grad_norm": 0.2370949536561966, "learning_rate": 8.31136271877806e-05, "loss": 0.0125, "step": 24490 }, { "grad_norm": 0.18740904331207275, "learning_rate": 8.309813831168748e-05, "loss": 0.0136, "step": 24500 }, { "grad_norm": 0.11523115634918213, "learning_rate": 8.308264378002777e-05, "loss": 0.0108, "step": 24510 }, { "grad_norm": 0.14972567558288574, "learning_rate": 8.306714359544906e-05, "loss": 0.0107, "step": 24520 }, { "grad_norm": 0.14256836473941803, "learning_rate": 8.30516377605999e-05, "loss": 0.0127, "step": 24530 }, { "grad_norm": 0.15585178136825562, "learning_rate": 8.30361262781298e-05, "loss": 0.0137, "step": 24540 }, { "grad_norm": 0.14251749217510223, "learning_rate": 8.302060915068924e-05, "loss": 0.0119, "step": 24550 }, { "grad_norm": 0.14389845728874207, "learning_rate": 8.300508638092972e-05, "loss": 0.0115, "step": 24560 }, { "grad_norm": 0.12390919774770737, "learning_rate": 8.298955797150361e-05, "loss": 0.0135, "step": 24570 }, { "grad_norm": 0.15541012585163116, "learning_rate": 8.297402392506433e-05, "loss": 0.0099, "step": 24580 }, { "grad_norm": 0.14819714426994324, "learning_rate": 8.295848424426617e-05, "loss": 0.0092, "step": 24590 }, { "grad_norm": 0.17523135244846344, "learning_rate": 8.29429389317645e-05, "loss": 0.012, "step": 24600 }, { "grad_norm": 0.15500542521476746, "learning_rate": 8.292738799021556e-05, "loss": 0.0108, "step": 24610 }, { "grad_norm": 0.17895722389221191, "learning_rate": 8.291183142227656e-05, "loss": 0.0119, "step": 24620 }, { "grad_norm": 0.23062652349472046, "learning_rate": 8.289626923060572e-05, "loss": 0.01, "step": 24630 }, { "grad_norm": 0.2854451537132263, "learning_rate": 8.288070141786218e-05, "loss": 0.0101, "step": 24640 }, { "grad_norm": 0.13133594393730164, "learning_rate": 8.286512798670605e-05, "loss": 0.0103, "step": 24650 }, { "grad_norm": 0.17015039920806885, "learning_rate": 8.284954893979842e-05, "loss": 0.0114, "step": 24660 }, { "grad_norm": 0.18312180042266846, "learning_rate": 8.283396427980131e-05, "loss": 0.0113, "step": 24670 }, { "grad_norm": 0.17046575248241425, "learning_rate": 8.281837400937771e-05, "loss": 0.0123, "step": 24680 }, { "grad_norm": 0.18530330061912537, "learning_rate": 8.28027781311916e-05, "loss": 0.012, "step": 24690 }, { "grad_norm": 0.1575993150472641, "learning_rate": 8.278717664790785e-05, "loss": 0.0101, "step": 24700 }, { "grad_norm": 0.19595900177955627, "learning_rate": 8.277156956219234e-05, "loss": 0.0104, "step": 24710 }, { "grad_norm": 0.18413618206977844, "learning_rate": 8.275595687671189e-05, "loss": 0.0137, "step": 24720 }, { "grad_norm": 0.1813150942325592, "learning_rate": 8.27403385941343e-05, "loss": 0.0116, "step": 24730 }, { "grad_norm": 0.19083410501480103, "learning_rate": 8.272471471712828e-05, "loss": 0.0115, "step": 24740 }, { "grad_norm": 0.1753520518541336, "learning_rate": 8.270908524836355e-05, "loss": 0.0101, "step": 24750 }, { "grad_norm": 0.1570303738117218, "learning_rate": 8.269345019051074e-05, "loss": 0.0125, "step": 24760 }, { "grad_norm": 0.16283442080020905, "learning_rate": 8.267780954624147e-05, "loss": 0.0097, "step": 24770 }, { "grad_norm": 0.15268723666667938, "learning_rate": 8.266216331822827e-05, "loss": 0.0095, "step": 24780 }, { "grad_norm": 0.16223274171352386, "learning_rate": 8.264651150914469e-05, "loss": 0.0121, "step": 24790 }, { "grad_norm": 0.22260740399360657, "learning_rate": 8.263085412166517e-05, "loss": 0.0105, "step": 24800 }, { "grad_norm": 0.22069260478019714, "learning_rate": 8.261519115846514e-05, "loss": 0.0125, "step": 24810 }, { "grad_norm": 0.14479634165763855, "learning_rate": 8.259952262222096e-05, "loss": 0.0091, "step": 24820 }, { "grad_norm": 0.19923828542232513, "learning_rate": 8.258384851560997e-05, "loss": 0.01, "step": 24830 }, { "grad_norm": 0.1573273241519928, "learning_rate": 8.256816884131044e-05, "loss": 0.0096, "step": 24840 }, { "grad_norm": 0.18439778685569763, "learning_rate": 8.255248360200159e-05, "loss": 0.0106, "step": 24850 }, { "grad_norm": 0.1257765293121338, "learning_rate": 8.253679280036359e-05, "loss": 0.0087, "step": 24860 }, { "grad_norm": 0.1701350063085556, "learning_rate": 8.252109643907762e-05, "loss": 0.0097, "step": 24870 }, { "grad_norm": 0.1254681497812271, "learning_rate": 8.250539452082569e-05, "loss": 0.0105, "step": 24880 }, { "grad_norm": 0.16689161956310272, "learning_rate": 8.248968704829087e-05, "loss": 0.0091, "step": 24890 }, { "grad_norm": 0.17415861785411835, "learning_rate": 8.247397402415714e-05, "loss": 0.0102, "step": 24900 }, { "grad_norm": 0.154610276222229, "learning_rate": 8.24582554511094e-05, "loss": 0.0103, "step": 24910 }, { "grad_norm": 0.16022755205631256, "learning_rate": 8.244253133183355e-05, "loss": 0.0128, "step": 24920 }, { "grad_norm": 0.177652046084404, "learning_rate": 8.24268016690164e-05, "loss": 0.011, "step": 24930 }, { "grad_norm": 0.20152294635772705, "learning_rate": 8.241106646534571e-05, "loss": 0.0097, "step": 24940 }, { "grad_norm": 0.1297825425863266, "learning_rate": 8.23953257235102e-05, "loss": 0.0104, "step": 24950 }, { "grad_norm": 0.13554944097995758, "learning_rate": 8.237957944619956e-05, "loss": 0.0109, "step": 24960 }, { "grad_norm": 0.21353651583194733, "learning_rate": 8.236382763610437e-05, "loss": 0.0091, "step": 24970 }, { "grad_norm": 0.2359066754579544, "learning_rate": 8.234807029591619e-05, "loss": 0.013, "step": 24980 }, { "grad_norm": 0.1372891515493393, "learning_rate": 8.233230742832752e-05, "loss": 0.0113, "step": 24990 }, { "grad_norm": 0.14441533386707306, "learning_rate": 8.231653903603178e-05, "loss": 0.0111, "step": 25000 }, { "grad_norm": 0.16387659311294556, "learning_rate": 8.23007651217234e-05, "loss": 0.0102, "step": 25010 }, { "grad_norm": 0.17226769030094147, "learning_rate": 8.228498568809769e-05, "loss": 0.0104, "step": 25020 }, { "grad_norm": 0.16898013651371002, "learning_rate": 8.22692007378509e-05, "loss": 0.0107, "step": 25030 }, { "grad_norm": 0.16116800904273987, "learning_rate": 8.225341027368028e-05, "loss": 0.009, "step": 25040 }, { "grad_norm": 0.16911356151103973, "learning_rate": 8.223761429828399e-05, "loss": 0.0108, "step": 25050 }, { "grad_norm": 0.12489544600248337, "learning_rate": 8.22218128143611e-05, "loss": 0.0099, "step": 25060 }, { "grad_norm": 0.19216127693653107, "learning_rate": 8.220600582461166e-05, "loss": 0.013, "step": 25070 }, { "grad_norm": 0.19040898978710175, "learning_rate": 8.219019333173668e-05, "loss": 0.01, "step": 25080 }, { "grad_norm": 0.15267576277256012, "learning_rate": 8.217437533843805e-05, "loss": 0.0098, "step": 25090 }, { "grad_norm": 0.11686119437217712, "learning_rate": 8.215855184741867e-05, "loss": 0.011, "step": 25100 }, { "grad_norm": 0.12950199842453003, "learning_rate": 8.21427228613823e-05, "loss": 0.0101, "step": 25110 }, { "grad_norm": 0.14649388194084167, "learning_rate": 8.21268883830337e-05, "loss": 0.0122, "step": 25120 }, { "grad_norm": 0.16210870444774628, "learning_rate": 8.211104841507855e-05, "loss": 0.0106, "step": 25130 }, { "grad_norm": 0.19812645018100739, "learning_rate": 8.209520296022346e-05, "loss": 0.0131, "step": 25140 }, { "grad_norm": 0.20859886705875397, "learning_rate": 8.207935202117599e-05, "loss": 0.0105, "step": 25150 }, { "grad_norm": 0.15099141001701355, "learning_rate": 8.206349560064463e-05, "loss": 0.0096, "step": 25160 }, { "grad_norm": 0.18789678812026978, "learning_rate": 8.204763370133881e-05, "loss": 0.0091, "step": 25170 }, { "grad_norm": 0.17070446908473969, "learning_rate": 8.203176632596892e-05, "loss": 0.0115, "step": 25180 }, { "grad_norm": 0.13094031810760498, "learning_rate": 8.20158934772462e-05, "loss": 0.0104, "step": 25190 }, { "grad_norm": 0.14794090390205383, "learning_rate": 8.200001515788294e-05, "loss": 0.0083, "step": 25200 }, { "grad_norm": 0.271310418844223, "learning_rate": 8.198413137059228e-05, "loss": 0.0089, "step": 25210 }, { "grad_norm": 0.12980802357196808, "learning_rate": 8.196824211808835e-05, "loss": 0.013, "step": 25220 }, { "grad_norm": 0.18455937504768372, "learning_rate": 8.195234740308617e-05, "loss": 0.0117, "step": 25230 }, { "grad_norm": 0.16567717492580414, "learning_rate": 8.193644722830171e-05, "loss": 0.011, "step": 25240 }, { "grad_norm": 0.15492984652519226, "learning_rate": 8.19205415964519e-05, "loss": 0.0091, "step": 25250 }, { "grad_norm": 0.1374175250530243, "learning_rate": 8.190463051025456e-05, "loss": 0.0099, "step": 25260 }, { "grad_norm": 0.1330735683441162, "learning_rate": 8.188871397242843e-05, "loss": 0.0106, "step": 25270 }, { "grad_norm": 0.18909671902656555, "learning_rate": 8.187279198569326e-05, "loss": 0.0101, "step": 25280 }, { "grad_norm": 0.17271322011947632, "learning_rate": 8.185686455276966e-05, "loss": 0.0123, "step": 25290 }, { "grad_norm": 0.13830183446407318, "learning_rate": 8.184093167637921e-05, "loss": 0.011, "step": 25300 }, { "grad_norm": 0.18048594892024994, "learning_rate": 8.182499335924437e-05, "loss": 0.0129, "step": 25310 }, { "grad_norm": 0.1149212121963501, "learning_rate": 8.18090496040886e-05, "loss": 0.0113, "step": 25320 }, { "grad_norm": 0.13008993864059448, "learning_rate": 8.179310041363621e-05, "loss": 0.0095, "step": 25330 }, { "grad_norm": 0.16615988314151764, "learning_rate": 8.17771457906125e-05, "loss": 0.0096, "step": 25340 }, { "grad_norm": 0.12351275980472565, "learning_rate": 8.176118573774371e-05, "loss": 0.0086, "step": 25350 }, { "grad_norm": 0.20811353623867035, "learning_rate": 8.174522025775692e-05, "loss": 0.0093, "step": 25360 }, { "grad_norm": 0.2091580182313919, "learning_rate": 8.172924935338022e-05, "loss": 0.0113, "step": 25370 }, { "grad_norm": 0.137334406375885, "learning_rate": 8.171327302734262e-05, "loss": 0.0108, "step": 25380 }, { "grad_norm": 0.15934376418590546, "learning_rate": 8.169729128237401e-05, "loss": 0.0115, "step": 25390 }, { "grad_norm": 0.17944253981113434, "learning_rate": 8.168130412120525e-05, "loss": 0.012, "step": 25400 }, { "grad_norm": 0.14882883429527283, "learning_rate": 8.16653115465681e-05, "loss": 0.0139, "step": 25410 }, { "grad_norm": 0.10838498175144196, "learning_rate": 8.164931356119526e-05, "loss": 0.0117, "step": 25420 }, { "grad_norm": 0.11456838250160217, "learning_rate": 8.163331016782032e-05, "loss": 0.01, "step": 25430 }, { "grad_norm": 0.13456197082996368, "learning_rate": 8.161730136917785e-05, "loss": 0.0126, "step": 25440 }, { "grad_norm": 0.13117820024490356, "learning_rate": 8.160128716800333e-05, "loss": 0.0089, "step": 25450 }, { "grad_norm": 0.14348308742046356, "learning_rate": 8.158526756703313e-05, "loss": 0.0125, "step": 25460 }, { "grad_norm": 0.1328696608543396, "learning_rate": 8.156924256900455e-05, "loss": 0.0147, "step": 25470 }, { "grad_norm": 0.2183344066143036, "learning_rate": 8.155321217665584e-05, "loss": 0.013, "step": 25480 }, { "grad_norm": 0.17520758509635925, "learning_rate": 8.153717639272614e-05, "loss": 0.0117, "step": 25490 }, { "grad_norm": 0.2103191465139389, "learning_rate": 8.152113521995555e-05, "loss": 0.011, "step": 25500 }, { "grad_norm": 0.13210029900074005, "learning_rate": 8.150508866108505e-05, "loss": 0.0107, "step": 25510 }, { "grad_norm": 0.1530006378889084, "learning_rate": 8.148903671885657e-05, "loss": 0.0133, "step": 25520 }, { "grad_norm": 0.21882420778274536, "learning_rate": 8.147297939601292e-05, "loss": 0.0125, "step": 25530 }, { "grad_norm": 0.19180887937545776, "learning_rate": 8.145691669529792e-05, "loss": 0.0117, "step": 25540 }, { "grad_norm": 0.13815638422966003, "learning_rate": 8.144084861945618e-05, "loss": 0.0078, "step": 25550 }, { "grad_norm": 0.16677965223789215, "learning_rate": 8.142477517123333e-05, "loss": 0.0086, "step": 25560 }, { "grad_norm": 0.1630145013332367, "learning_rate": 8.140869635337586e-05, "loss": 0.0099, "step": 25570 }, { "grad_norm": 0.1365882307291031, "learning_rate": 8.139261216863123e-05, "loss": 0.0113, "step": 25580 }, { "grad_norm": 0.19484016299247742, "learning_rate": 8.137652261974776e-05, "loss": 0.0126, "step": 25590 }, { "grad_norm": 0.1859152913093567, "learning_rate": 8.136042770947472e-05, "loss": 0.0124, "step": 25600 }, { "grad_norm": 0.12776127457618713, "learning_rate": 8.134432744056228e-05, "loss": 0.0097, "step": 25610 }, { "grad_norm": 0.1673325002193451, "learning_rate": 8.132822181576158e-05, "loss": 0.0113, "step": 25620 }, { "grad_norm": 0.13655546307563782, "learning_rate": 8.131211083782459e-05, "loss": 0.0118, "step": 25630 }, { "grad_norm": 0.19023285806179047, "learning_rate": 8.129599450950424e-05, "loss": 0.0098, "step": 25640 }, { "grad_norm": 0.129611998796463, "learning_rate": 8.127987283355438e-05, "loss": 0.0099, "step": 25650 }, { "grad_norm": 0.21673418581485748, "learning_rate": 8.126374581272976e-05, "loss": 0.0103, "step": 25660 }, { "grad_norm": 0.18912957608699799, "learning_rate": 8.124761344978605e-05, "loss": 0.0149, "step": 25670 }, { "grad_norm": 0.13640494644641876, "learning_rate": 8.12314757474798e-05, "loss": 0.0114, "step": 25680 }, { "grad_norm": 0.12902577221393585, "learning_rate": 8.121533270856856e-05, "loss": 0.0078, "step": 25690 }, { "grad_norm": 0.23684151470661163, "learning_rate": 8.119918433581069e-05, "loss": 0.0123, "step": 25700 }, { "grad_norm": 0.1317010372877121, "learning_rate": 8.118303063196551e-05, "loss": 0.0123, "step": 25710 }, { "grad_norm": 0.1434342861175537, "learning_rate": 8.116687159979326e-05, "loss": 0.0088, "step": 25720 }, { "grad_norm": 0.15365226566791534, "learning_rate": 8.115070724205508e-05, "loss": 0.0109, "step": 25730 }, { "grad_norm": 0.1488017439842224, "learning_rate": 8.113453756151296e-05, "loss": 0.0119, "step": 25740 }, { "grad_norm": 0.15770530700683594, "learning_rate": 8.111836256092995e-05, "loss": 0.0113, "step": 25750 }, { "grad_norm": 0.21537093818187714, "learning_rate": 8.110218224306985e-05, "loss": 0.0112, "step": 25760 }, { "grad_norm": 0.16866947710514069, "learning_rate": 8.108599661069745e-05, "loss": 0.0095, "step": 25770 }, { "grad_norm": 0.14893020689487457, "learning_rate": 8.106980566657845e-05, "loss": 0.0103, "step": 25780 }, { "grad_norm": 0.15027672052383423, "learning_rate": 8.10536094134794e-05, "loss": 0.0106, "step": 25790 }, { "grad_norm": 0.14314571022987366, "learning_rate": 8.103740785416783e-05, "loss": 0.0087, "step": 25800 }, { "grad_norm": 0.13166293501853943, "learning_rate": 8.102120099141212e-05, "loss": 0.0093, "step": 25810 }, { "grad_norm": 0.13270245492458344, "learning_rate": 8.100498882798163e-05, "loss": 0.0097, "step": 25820 }, { "grad_norm": 0.12775948643684387, "learning_rate": 8.09887713666465e-05, "loss": 0.0115, "step": 25830 }, { "grad_norm": 0.17733155190944672, "learning_rate": 8.09725486101779e-05, "loss": 0.0109, "step": 25840 }, { "grad_norm": 0.09690903127193451, "learning_rate": 8.095632056134784e-05, "loss": 0.0103, "step": 25850 }, { "grad_norm": 0.1976139098405838, "learning_rate": 8.094008722292925e-05, "loss": 0.0143, "step": 25860 }, { "grad_norm": 0.1316927820444107, "learning_rate": 8.092384859769598e-05, "loss": 0.0084, "step": 25870 }, { "grad_norm": 0.15632256865501404, "learning_rate": 8.090760468842275e-05, "loss": 0.0134, "step": 25880 }, { "grad_norm": 0.2513068616390228, "learning_rate": 8.089135549788521e-05, "loss": 0.0107, "step": 25890 }, { "grad_norm": 0.1492447555065155, "learning_rate": 8.087510102885987e-05, "loss": 0.0092, "step": 25900 }, { "grad_norm": 0.1685609370470047, "learning_rate": 8.085884128412422e-05, "loss": 0.0114, "step": 25910 }, { "grad_norm": 0.16826803982257843, "learning_rate": 8.084257626645659e-05, "loss": 0.0116, "step": 25920 }, { "grad_norm": 0.12500765919685364, "learning_rate": 8.08263059786362e-05, "loss": 0.0082, "step": 25930 }, { "grad_norm": 0.17554651200771332, "learning_rate": 8.081003042344325e-05, "loss": 0.0084, "step": 25940 }, { "grad_norm": 0.18273064494132996, "learning_rate": 8.079374960365872e-05, "loss": 0.0094, "step": 25950 }, { "grad_norm": 0.14482815563678741, "learning_rate": 8.077746352206463e-05, "loss": 0.0087, "step": 25960 }, { "grad_norm": 0.14739841222763062, "learning_rate": 8.076117218144377e-05, "loss": 0.0078, "step": 25970 }, { "grad_norm": 0.11198671162128448, "learning_rate": 8.074487558457991e-05, "loss": 0.009, "step": 25980 }, { "grad_norm": 0.15139040350914001, "learning_rate": 8.072857373425768e-05, "loss": 0.0092, "step": 25990 }, { "grad_norm": 0.15111152827739716, "learning_rate": 8.071226663326264e-05, "loss": 0.0094, "step": 26000 }, { "grad_norm": 0.21506521105766296, "learning_rate": 8.069595428438121e-05, "loss": 0.0141, "step": 26010 }, { "grad_norm": 0.11468123644590378, "learning_rate": 8.067963669040072e-05, "loss": 0.0086, "step": 26020 }, { "grad_norm": 0.16388724744319916, "learning_rate": 8.066331385410942e-05, "loss": 0.0122, "step": 26030 }, { "grad_norm": 0.1311924010515213, "learning_rate": 8.064698577829641e-05, "loss": 0.0091, "step": 26040 }, { "grad_norm": 0.14343015849590302, "learning_rate": 8.063065246575175e-05, "loss": 0.0082, "step": 26050 }, { "grad_norm": 0.15852238237857819, "learning_rate": 8.061431391926631e-05, "loss": 0.0095, "step": 26060 }, { "grad_norm": 0.20577386021614075, "learning_rate": 8.059797014163195e-05, "loss": 0.0108, "step": 26070 }, { "grad_norm": 0.16684582829475403, "learning_rate": 8.058162113564133e-05, "loss": 0.0106, "step": 26080 }, { "grad_norm": 0.1456877738237381, "learning_rate": 8.056526690408806e-05, "loss": 0.0112, "step": 26090 }, { "grad_norm": 0.15318775177001953, "learning_rate": 8.054890744976666e-05, "loss": 0.0107, "step": 26100 }, { "grad_norm": 0.16085605323314667, "learning_rate": 8.053254277547248e-05, "loss": 0.0094, "step": 26110 }, { "grad_norm": 0.17107321321964264, "learning_rate": 8.051617288400182e-05, "loss": 0.0095, "step": 26120 }, { "grad_norm": 0.24641656875610352, "learning_rate": 8.049979777815182e-05, "loss": 0.0111, "step": 26130 }, { "grad_norm": 0.12765535712242126, "learning_rate": 8.048341746072054e-05, "loss": 0.0095, "step": 26140 }, { "grad_norm": 0.16884730756282806, "learning_rate": 8.046703193450696e-05, "loss": 0.0115, "step": 26150 }, { "grad_norm": 0.16540493071079254, "learning_rate": 8.04506412023109e-05, "loss": 0.0097, "step": 26160 }, { "grad_norm": 0.18053190410137177, "learning_rate": 8.043424526693306e-05, "loss": 0.0117, "step": 26170 }, { "grad_norm": 0.19664686918258667, "learning_rate": 8.04178441311751e-05, "loss": 0.0084, "step": 26180 }, { "grad_norm": 0.13554072380065918, "learning_rate": 8.04014377978395e-05, "loss": 0.0088, "step": 26190 }, { "grad_norm": 0.22413648664951324, "learning_rate": 8.038502626972967e-05, "loss": 0.0123, "step": 26200 }, { "grad_norm": 0.18914878368377686, "learning_rate": 8.036860954964989e-05, "loss": 0.0122, "step": 26210 }, { "grad_norm": 0.1446836292743683, "learning_rate": 8.035218764040531e-05, "loss": 0.0104, "step": 26220 }, { "grad_norm": 0.2668229341506958, "learning_rate": 8.033576054480199e-05, "loss": 0.0137, "step": 26230 }, { "grad_norm": 0.1980150043964386, "learning_rate": 8.031932826564688e-05, "loss": 0.0097, "step": 26240 }, { "grad_norm": 0.18819984793663025, "learning_rate": 8.030289080574782e-05, "loss": 0.0084, "step": 26250 }, { "grad_norm": 0.1375066041946411, "learning_rate": 8.028644816791349e-05, "loss": 0.0104, "step": 26260 }, { "grad_norm": 0.13853423297405243, "learning_rate": 8.027000035495351e-05, "loss": 0.0095, "step": 26270 }, { "grad_norm": 0.13661785423755646, "learning_rate": 8.025354736967836e-05, "loss": 0.0076, "step": 26280 }, { "grad_norm": 0.16988739371299744, "learning_rate": 8.023708921489941e-05, "loss": 0.0082, "step": 26290 }, { "grad_norm": 0.15165358781814575, "learning_rate": 8.022062589342887e-05, "loss": 0.011, "step": 26300 }, { "grad_norm": 0.18853320181369781, "learning_rate": 8.020415740807993e-05, "loss": 0.0095, "step": 26310 }, { "grad_norm": 0.12828172743320465, "learning_rate": 8.018768376166656e-05, "loss": 0.0112, "step": 26320 }, { "grad_norm": 0.1675451099872589, "learning_rate": 8.017120495700368e-05, "loss": 0.0127, "step": 26330 }, { "grad_norm": 0.17760153114795685, "learning_rate": 8.015472099690704e-05, "loss": 0.0114, "step": 26340 }, { "grad_norm": 0.17856347560882568, "learning_rate": 8.013823188419332e-05, "loss": 0.0136, "step": 26350 }, { "grad_norm": 0.17806866765022278, "learning_rate": 8.012173762168006e-05, "loss": 0.0097, "step": 26360 }, { "grad_norm": 0.17697571218013763, "learning_rate": 8.010523821218567e-05, "loss": 0.0091, "step": 26370 }, { "grad_norm": 0.1894480586051941, "learning_rate": 8.008873365852945e-05, "loss": 0.0112, "step": 26380 }, { "grad_norm": 0.14456579089164734, "learning_rate": 8.007222396353157e-05, "loss": 0.0111, "step": 26390 }, { "grad_norm": 0.14229360222816467, "learning_rate": 8.00557091300131e-05, "loss": 0.0092, "step": 26400 }, { "grad_norm": 0.13332301378250122, "learning_rate": 8.003918916079597e-05, "loss": 0.0102, "step": 26410 }, { "grad_norm": 0.1391783207654953, "learning_rate": 8.002266405870298e-05, "loss": 0.0102, "step": 26420 }, { "grad_norm": 0.15172533690929413, "learning_rate": 8.000613382655782e-05, "loss": 0.0093, "step": 26430 }, { "grad_norm": 0.14691630005836487, "learning_rate": 7.998959846718505e-05, "loss": 0.0116, "step": 26440 }, { "grad_norm": 0.17552585899829865, "learning_rate": 7.997305798341012e-05, "loss": 0.0096, "step": 26450 }, { "grad_norm": 0.18267646431922913, "learning_rate": 7.995651237805937e-05, "loss": 0.0109, "step": 26460 }, { "grad_norm": 0.21531760692596436, "learning_rate": 7.993996165395996e-05, "loss": 0.0102, "step": 26470 }, { "grad_norm": 0.1535194218158722, "learning_rate": 7.992340581393996e-05, "loss": 0.0099, "step": 26480 }, { "grad_norm": 0.15116167068481445, "learning_rate": 7.990684486082831e-05, "loss": 0.0114, "step": 26490 }, { "grad_norm": 0.1205342635512352, "learning_rate": 7.989027879745482e-05, "loss": 0.0089, "step": 26500 }, { "grad_norm": 0.1255778819322586, "learning_rate": 7.98737076266502e-05, "loss": 0.0101, "step": 26510 }, { "grad_norm": 0.14533032476902008, "learning_rate": 7.985713135124598e-05, "loss": 0.0102, "step": 26520 }, { "grad_norm": 0.1631283164024353, "learning_rate": 7.98405499740746e-05, "loss": 0.0101, "step": 26530 }, { "grad_norm": 0.20324401557445526, "learning_rate": 7.98239634979694e-05, "loss": 0.013, "step": 26540 }, { "grad_norm": 0.15549248456954956, "learning_rate": 7.98073719257645e-05, "loss": 0.01, "step": 26550 }, { "grad_norm": 0.12492267042398453, "learning_rate": 7.979077526029499e-05, "loss": 0.0099, "step": 26560 }, { "grad_norm": 0.16917277872562408, "learning_rate": 7.977417350439675e-05, "loss": 0.0116, "step": 26570 }, { "grad_norm": 0.14697982370853424, "learning_rate": 7.97575666609066e-05, "loss": 0.0118, "step": 26580 }, { "grad_norm": 0.14052802324295044, "learning_rate": 7.974095473266216e-05, "loss": 0.0088, "step": 26590 }, { "grad_norm": 0.14315804839134216, "learning_rate": 7.972433772250198e-05, "loss": 0.009, "step": 26600 }, { "grad_norm": 0.16357555985450745, "learning_rate": 7.970771563326544e-05, "loss": 0.0118, "step": 26610 }, { "grad_norm": 0.15304772555828094, "learning_rate": 7.96910884677928e-05, "loss": 0.0124, "step": 26620 }, { "grad_norm": 0.16730095446109772, "learning_rate": 7.967445622892523e-05, "loss": 0.0107, "step": 26630 }, { "grad_norm": 0.1094360426068306, "learning_rate": 7.965781891950465e-05, "loss": 0.0094, "step": 26640 }, { "grad_norm": 0.10614850372076035, "learning_rate": 7.964117654237397e-05, "loss": 0.0085, "step": 26650 }, { "grad_norm": 0.11715391278266907, "learning_rate": 7.962452910037692e-05, "loss": 0.011, "step": 26660 }, { "grad_norm": 0.1681603491306305, "learning_rate": 7.96078765963581e-05, "loss": 0.0102, "step": 26670 }, { "grad_norm": 0.14217253029346466, "learning_rate": 7.95912190331629e-05, "loss": 0.0089, "step": 26680 }, { "grad_norm": 0.14238956570625305, "learning_rate": 7.957455641363772e-05, "loss": 0.0088, "step": 26690 }, { "grad_norm": 0.20526757836341858, "learning_rate": 7.955788874062968e-05, "loss": 0.0098, "step": 26700 }, { "grad_norm": 0.1836182028055191, "learning_rate": 7.95412160169869e-05, "loss": 0.0118, "step": 26710 }, { "grad_norm": 0.1745455265045166, "learning_rate": 7.952453824555824e-05, "loss": 0.0097, "step": 26720 }, { "grad_norm": 0.13854140043258667, "learning_rate": 7.95078554291935e-05, "loss": 0.0086, "step": 26730 }, { "grad_norm": 0.2012222856283188, "learning_rate": 7.94911675707433e-05, "loss": 0.0073, "step": 26740 }, { "grad_norm": 0.14428403973579407, "learning_rate": 7.947447467305915e-05, "loss": 0.0088, "step": 26750 }, { "grad_norm": 0.12447347491979599, "learning_rate": 7.94577767389934e-05, "loss": 0.0104, "step": 26760 }, { "grad_norm": 0.14462696015834808, "learning_rate": 7.944107377139928e-05, "loss": 0.0085, "step": 26770 }, { "grad_norm": 0.18903137743473053, "learning_rate": 7.942436577313088e-05, "loss": 0.012, "step": 26780 }, { "grad_norm": 0.1223364919424057, "learning_rate": 7.940765274704312e-05, "loss": 0.0074, "step": 26790 }, { "grad_norm": 0.12827405333518982, "learning_rate": 7.939093469599181e-05, "loss": 0.01, "step": 26800 }, { "grad_norm": 0.1563674509525299, "learning_rate": 7.93742116228336e-05, "loss": 0.0098, "step": 26810 }, { "grad_norm": 0.14927369356155396, "learning_rate": 7.935748353042602e-05, "loss": 0.0107, "step": 26820 }, { "grad_norm": 0.18325869739055634, "learning_rate": 7.934075042162744e-05, "loss": 0.0113, "step": 26830 }, { "grad_norm": 0.16639180481433868, "learning_rate": 7.932401229929705e-05, "loss": 0.0098, "step": 26840 }, { "grad_norm": 0.15420763194561005, "learning_rate": 7.9307269166295e-05, "loss": 0.0092, "step": 26850 }, { "grad_norm": 0.17195941507816315, "learning_rate": 7.92905210254822e-05, "loss": 0.0088, "step": 26860 }, { "grad_norm": 0.21547460556030273, "learning_rate": 7.927376787972045e-05, "loss": 0.012, "step": 26870 }, { "grad_norm": 0.22627969086170197, "learning_rate": 7.92570097318724e-05, "loss": 0.011, "step": 26880 }, { "grad_norm": 0.17127268016338348, "learning_rate": 7.924024658480158e-05, "loss": 0.0128, "step": 26890 }, { "grad_norm": 0.1629418581724167, "learning_rate": 7.922347844137233e-05, "loss": 0.011, "step": 26900 }, { "grad_norm": 0.2030329555273056, "learning_rate": 7.92067053044499e-05, "loss": 0.0104, "step": 26910 }, { "grad_norm": 0.13653546571731567, "learning_rate": 7.918992717690031e-05, "loss": 0.0096, "step": 26920 }, { "grad_norm": 0.22349093854427338, "learning_rate": 7.917314406159053e-05, "loss": 0.0097, "step": 26930 }, { "grad_norm": 0.14002832770347595, "learning_rate": 7.915635596138832e-05, "loss": 0.0101, "step": 26940 }, { "grad_norm": 0.20054443180561066, "learning_rate": 7.913956287916228e-05, "loss": 0.0132, "step": 26950 }, { "grad_norm": 0.13079479336738586, "learning_rate": 7.912276481778193e-05, "loss": 0.0111, "step": 26960 }, { "grad_norm": 0.18170174956321716, "learning_rate": 7.910596178011759e-05, "loss": 0.0104, "step": 26970 }, { "grad_norm": 0.17446684837341309, "learning_rate": 7.908915376904043e-05, "loss": 0.0094, "step": 26980 }, { "grad_norm": 0.15166132152080536, "learning_rate": 7.907234078742247e-05, "loss": 0.0102, "step": 26990 }, { "grad_norm": 0.1538044661283493, "learning_rate": 7.90555228381366e-05, "loss": 0.0089, "step": 27000 }, { "grad_norm": 0.14284966886043549, "learning_rate": 7.903869992405656e-05, "loss": 0.0102, "step": 27010 }, { "grad_norm": 0.18402571976184845, "learning_rate": 7.902187204805691e-05, "loss": 0.0099, "step": 27020 }, { "grad_norm": 0.147378608584404, "learning_rate": 7.900503921301308e-05, "loss": 0.01, "step": 27030 }, { "grad_norm": 0.15903964638710022, "learning_rate": 7.898820142180133e-05, "loss": 0.0089, "step": 27040 }, { "grad_norm": 0.14309611916542053, "learning_rate": 7.897135867729879e-05, "loss": 0.0088, "step": 27050 }, { "grad_norm": 0.14443650841712952, "learning_rate": 7.89545109823834e-05, "loss": 0.0117, "step": 27060 }, { "grad_norm": 0.15741685032844543, "learning_rate": 7.8937658339934e-05, "loss": 0.0102, "step": 27070 }, { "grad_norm": 0.14861077070236206, "learning_rate": 7.892080075283026e-05, "loss": 0.0107, "step": 27080 }, { "grad_norm": 0.15862765908241272, "learning_rate": 7.890393822395263e-05, "loss": 0.0077, "step": 27090 }, { "grad_norm": 0.17642293870449066, "learning_rate": 7.88870707561825e-05, "loss": 0.0091, "step": 27100 }, { "grad_norm": 0.10745935142040253, "learning_rate": 7.887019835240203e-05, "loss": 0.0084, "step": 27110 }, { "grad_norm": 0.19046007096767426, "learning_rate": 7.885332101549427e-05, "loss": 0.0101, "step": 27120 }, { "grad_norm": 0.1755060851573944, "learning_rate": 7.883643874834308e-05, "loss": 0.0098, "step": 27130 }, { "grad_norm": 0.11983124166727066, "learning_rate": 7.881955155383321e-05, "loss": 0.0089, "step": 27140 }, { "grad_norm": 0.16970190405845642, "learning_rate": 7.880265943485017e-05, "loss": 0.0109, "step": 27150 }, { "grad_norm": 0.15588033199310303, "learning_rate": 7.878576239428038e-05, "loss": 0.0107, "step": 27160 }, { "grad_norm": 0.16581614315509796, "learning_rate": 7.87688604350111e-05, "loss": 0.0125, "step": 27170 }, { "grad_norm": 0.12425563484430313, "learning_rate": 7.875195355993042e-05, "loss": 0.0083, "step": 27180 }, { "grad_norm": 0.1694648414850235, "learning_rate": 7.873504177192724e-05, "loss": 0.0113, "step": 27190 }, { "grad_norm": 0.12487969547510147, "learning_rate": 7.87181250738913e-05, "loss": 0.0089, "step": 27200 }, { "grad_norm": 0.11288503557443619, "learning_rate": 7.870120346871324e-05, "loss": 0.0091, "step": 27210 }, { "grad_norm": 0.1503376066684723, "learning_rate": 7.86842769592845e-05, "loss": 0.0087, "step": 27220 }, { "grad_norm": 0.16553464531898499, "learning_rate": 7.866734554849732e-05, "loss": 0.0092, "step": 27230 }, { "grad_norm": 0.15760046243667603, "learning_rate": 7.865040923924486e-05, "loss": 0.0087, "step": 27240 }, { "grad_norm": 0.12871424853801727, "learning_rate": 7.863346803442104e-05, "loss": 0.0098, "step": 27250 }, { "grad_norm": 0.21078161895275116, "learning_rate": 7.861652193692067e-05, "loss": 0.0097, "step": 27260 }, { "grad_norm": 0.16074393689632416, "learning_rate": 7.859957094963937e-05, "loss": 0.0093, "step": 27270 }, { "grad_norm": 0.14700263738632202, "learning_rate": 7.858261507547357e-05, "loss": 0.0092, "step": 27280 }, { "grad_norm": 0.10009203851222992, "learning_rate": 7.856565431732061e-05, "loss": 0.0097, "step": 27290 }, { "grad_norm": 0.19860097765922546, "learning_rate": 7.854868867807859e-05, "loss": 0.0122, "step": 27300 }, { "grad_norm": 0.1522388756275177, "learning_rate": 7.85317181606465e-05, "loss": 0.011, "step": 27310 }, { "grad_norm": 0.20385514199733734, "learning_rate": 7.85147427679241e-05, "loss": 0.0098, "step": 27320 }, { "grad_norm": 0.14279615879058838, "learning_rate": 7.849776250281205e-05, "loss": 0.0082, "step": 27330 }, { "grad_norm": 0.12492938339710236, "learning_rate": 7.84807773682118e-05, "loss": 0.0082, "step": 27340 }, { "grad_norm": 0.15642377734184265, "learning_rate": 7.846378736702565e-05, "loss": 0.0083, "step": 27350 }, { "grad_norm": 0.12537647783756256, "learning_rate": 7.844679250215671e-05, "loss": 0.0077, "step": 27360 }, { "grad_norm": 0.13903522491455078, "learning_rate": 7.842979277650898e-05, "loss": 0.0108, "step": 27370 }, { "grad_norm": 0.16409163177013397, "learning_rate": 7.84127881929872e-05, "loss": 0.0101, "step": 27380 }, { "grad_norm": 0.16232307255268097, "learning_rate": 7.839577875449704e-05, "loss": 0.0083, "step": 27390 }, { "grad_norm": 0.1540490686893463, "learning_rate": 7.837876446394489e-05, "loss": 0.0091, "step": 27400 }, { "grad_norm": 0.1442699432373047, "learning_rate": 7.836174532423805e-05, "loss": 0.0092, "step": 27410 }, { "grad_norm": 0.1324843466281891, "learning_rate": 7.834472133828466e-05, "loss": 0.0098, "step": 27420 }, { "grad_norm": 0.15961694717407227, "learning_rate": 7.832769250899359e-05, "loss": 0.0112, "step": 27430 }, { "grad_norm": 0.16334563493728638, "learning_rate": 7.831065883927464e-05, "loss": 0.0098, "step": 27440 }, { "grad_norm": 0.1690838485956192, "learning_rate": 7.829362033203841e-05, "loss": 0.0126, "step": 27450 }, { "grad_norm": 0.15704502165317535, "learning_rate": 7.827657699019628e-05, "loss": 0.0092, "step": 27460 }, { "grad_norm": 0.13823239505290985, "learning_rate": 7.825952881666052e-05, "loss": 0.011, "step": 27470 }, { "grad_norm": 0.1179407387971878, "learning_rate": 7.824247581434418e-05, "loss": 0.0096, "step": 27480 }, { "grad_norm": 0.148721382021904, "learning_rate": 7.822541798616116e-05, "loss": 0.0074, "step": 27490 }, { "grad_norm": 0.12807200849056244, "learning_rate": 7.820835533502617e-05, "loss": 0.0095, "step": 27500 }, { "grad_norm": 0.12594644725322723, "learning_rate": 7.819128786385475e-05, "loss": 0.008, "step": 27510 }, { "grad_norm": 0.12028370797634125, "learning_rate": 7.817421557556329e-05, "loss": 0.0091, "step": 27520 }, { "grad_norm": 0.1367892026901245, "learning_rate": 7.815713847306893e-05, "loss": 0.0073, "step": 27530 }, { "grad_norm": 0.13103905320167542, "learning_rate": 7.81400565592897e-05, "loss": 0.0103, "step": 27540 }, { "grad_norm": 0.1330389827489853, "learning_rate": 7.812296983714444e-05, "loss": 0.0097, "step": 27550 }, { "grad_norm": 0.15002410113811493, "learning_rate": 7.810587830955281e-05, "loss": 0.0088, "step": 27560 }, { "grad_norm": 0.1515215039253235, "learning_rate": 7.808878197943528e-05, "loss": 0.011, "step": 27570 }, { "grad_norm": 0.19766691327095032, "learning_rate": 7.807168084971312e-05, "loss": 0.0101, "step": 27580 }, { "grad_norm": 0.11189950257539749, "learning_rate": 7.805457492330849e-05, "loss": 0.0113, "step": 27590 }, { "grad_norm": 0.1278807818889618, "learning_rate": 7.803746420314428e-05, "loss": 0.0096, "step": 27600 }, { "grad_norm": 0.10957106202840805, "learning_rate": 7.802034869214428e-05, "loss": 0.0108, "step": 27610 }, { "grad_norm": 0.17288289964199066, "learning_rate": 7.800322839323303e-05, "loss": 0.0136, "step": 27620 }, { "grad_norm": 0.19713440537452698, "learning_rate": 7.798610330933593e-05, "loss": 0.0116, "step": 27630 }, { "grad_norm": 0.22201089560985565, "learning_rate": 7.796897344337922e-05, "loss": 0.0099, "step": 27640 }, { "grad_norm": 0.17908824980258942, "learning_rate": 7.795183879828989e-05, "loss": 0.0104, "step": 27650 }, { "grad_norm": 0.13215936720371246, "learning_rate": 7.793469937699579e-05, "loss": 0.0089, "step": 27660 }, { "grad_norm": 0.136248379945755, "learning_rate": 7.791755518242558e-05, "loss": 0.0078, "step": 27670 }, { "grad_norm": 0.16348205506801605, "learning_rate": 7.790040621750876e-05, "loss": 0.008, "step": 27680 }, { "grad_norm": 0.17583787441253662, "learning_rate": 7.788325248517558e-05, "loss": 0.0094, "step": 27690 }, { "grad_norm": 0.1715930998325348, "learning_rate": 7.786609398835715e-05, "loss": 0.0102, "step": 27700 }, { "grad_norm": 0.17954421043395996, "learning_rate": 7.784893072998541e-05, "loss": 0.0092, "step": 27710 }, { "grad_norm": 0.1552717387676239, "learning_rate": 7.783176271299306e-05, "loss": 0.0084, "step": 27720 }, { "grad_norm": 0.1305704563856125, "learning_rate": 7.781458994031368e-05, "loss": 0.0083, "step": 27730 }, { "grad_norm": 0.18370085954666138, "learning_rate": 7.779741241488161e-05, "loss": 0.0091, "step": 27740 }, { "grad_norm": 0.17149722576141357, "learning_rate": 7.7780230139632e-05, "loss": 0.0106, "step": 27750 }, { "grad_norm": 0.12037265300750732, "learning_rate": 7.776304311750087e-05, "loss": 0.0099, "step": 27760 }, { "grad_norm": 0.13709820806980133, "learning_rate": 7.7745851351425e-05, "loss": 0.0113, "step": 27770 }, { "grad_norm": 0.15448758006095886, "learning_rate": 7.772865484434197e-05, "loss": 0.0079, "step": 27780 }, { "grad_norm": 0.1781730055809021, "learning_rate": 7.77114535991902e-05, "loss": 0.0101, "step": 27790 }, { "grad_norm": 0.16256670653820038, "learning_rate": 7.769424761890893e-05, "loss": 0.0099, "step": 27800 }, { "grad_norm": 0.17894384264945984, "learning_rate": 7.767703690643817e-05, "loss": 0.0096, "step": 27810 }, { "grad_norm": 0.20631687343120575, "learning_rate": 7.76598214647188e-05, "loss": 0.0117, "step": 27820 }, { "grad_norm": 0.15590515732765198, "learning_rate": 7.764260129669241e-05, "loss": 0.009, "step": 27830 }, { "grad_norm": 0.13920389115810394, "learning_rate": 7.76253764053015e-05, "loss": 0.0086, "step": 27840 }, { "grad_norm": 0.1718839704990387, "learning_rate": 7.760814679348932e-05, "loss": 0.0094, "step": 27850 }, { "grad_norm": 0.16635701060295105, "learning_rate": 7.759091246419992e-05, "loss": 0.0101, "step": 27860 }, { "grad_norm": 0.18659982085227966, "learning_rate": 7.75736734203782e-05, "loss": 0.0111, "step": 27870 }, { "grad_norm": 0.13899144530296326, "learning_rate": 7.755642966496985e-05, "loss": 0.0107, "step": 27880 }, { "grad_norm": 0.16076381504535675, "learning_rate": 7.753918120092132e-05, "loss": 0.0119, "step": 27890 }, { "grad_norm": 0.15479114651679993, "learning_rate": 7.752192803117993e-05, "loss": 0.0093, "step": 27900 }, { "grad_norm": 0.1437397003173828, "learning_rate": 7.750467015869377e-05, "loss": 0.0122, "step": 27910 }, { "grad_norm": 0.13467542827129364, "learning_rate": 7.748740758641174e-05, "loss": 0.0088, "step": 27920 }, { "grad_norm": 0.14495247602462769, "learning_rate": 7.74701403172835e-05, "loss": 0.0117, "step": 27930 }, { "grad_norm": 0.11812850087881088, "learning_rate": 7.745286835425962e-05, "loss": 0.0091, "step": 27940 }, { "grad_norm": 0.14496389031410217, "learning_rate": 7.743559170029138e-05, "loss": 0.0126, "step": 27950 }, { "grad_norm": 0.17605189979076385, "learning_rate": 7.741831035833087e-05, "loss": 0.0104, "step": 27960 }, { "grad_norm": 0.12659016251564026, "learning_rate": 7.740102433133102e-05, "loss": 0.0108, "step": 27970 }, { "grad_norm": 0.1604958325624466, "learning_rate": 7.738373362224553e-05, "loss": 0.0095, "step": 27980 }, { "grad_norm": 0.13193969428539276, "learning_rate": 7.73664382340289e-05, "loss": 0.0109, "step": 27990 }, { "grad_norm": 0.15988412499427795, "learning_rate": 7.734913816963647e-05, "loss": 0.0088, "step": 28000 }, { "grad_norm": 0.1368294656276703, "learning_rate": 7.73318334320243e-05, "loss": 0.011, "step": 28010 }, { "grad_norm": 0.1857517510652542, "learning_rate": 7.731452402414934e-05, "loss": 0.0091, "step": 28020 }, { "grad_norm": 0.1629507690668106, "learning_rate": 7.729720994896928e-05, "loss": 0.0093, "step": 28030 }, { "grad_norm": 0.12049230933189392, "learning_rate": 7.727989120944262e-05, "loss": 0.0088, "step": 28040 }, { "grad_norm": 0.17895740270614624, "learning_rate": 7.726256780852865e-05, "loss": 0.013, "step": 28050 }, { "grad_norm": 0.19973163306713104, "learning_rate": 7.724523974918749e-05, "loss": 0.0096, "step": 28060 }, { "grad_norm": 0.1708507239818573, "learning_rate": 7.722790703438002e-05, "loss": 0.0084, "step": 28070 }, { "grad_norm": 0.2012304961681366, "learning_rate": 7.72105696670679e-05, "loss": 0.0088, "step": 28080 }, { "grad_norm": 0.20106936991214752, "learning_rate": 7.719322765021364e-05, "loss": 0.0069, "step": 28090 }, { "grad_norm": 0.13924764096736908, "learning_rate": 7.717588098678051e-05, "loss": 0.0119, "step": 28100 }, { "grad_norm": 0.16621798276901245, "learning_rate": 7.715852967973258e-05, "loss": 0.0078, "step": 28110 }, { "grad_norm": 0.18373218178749084, "learning_rate": 7.714117373203474e-05, "loss": 0.0079, "step": 28120 }, { "grad_norm": 0.12372680753469467, "learning_rate": 7.712381314665259e-05, "loss": 0.0089, "step": 28130 }, { "grad_norm": 0.12013468146324158, "learning_rate": 7.710644792655261e-05, "loss": 0.0098, "step": 28140 }, { "grad_norm": 0.12757471203804016, "learning_rate": 7.708907807470207e-05, "loss": 0.0069, "step": 28150 }, { "grad_norm": 0.09091275185346603, "learning_rate": 7.707170359406896e-05, "loss": 0.0089, "step": 28160 }, { "grad_norm": 0.15850400924682617, "learning_rate": 7.705432448762213e-05, "loss": 0.011, "step": 28170 }, { "grad_norm": 0.18885312974452972, "learning_rate": 7.703694075833117e-05, "loss": 0.0085, "step": 28180 }, { "grad_norm": 0.14243149757385254, "learning_rate": 7.70195524091665e-05, "loss": 0.0113, "step": 28190 }, { "grad_norm": 0.13210445642471313, "learning_rate": 7.70021594430993e-05, "loss": 0.009, "step": 28200 }, { "grad_norm": 0.10508514940738678, "learning_rate": 7.698476186310157e-05, "loss": 0.0091, "step": 28210 }, { "grad_norm": 0.15583287179470062, "learning_rate": 7.696735967214608e-05, "loss": 0.0091, "step": 28220 }, { "grad_norm": 0.14397333562374115, "learning_rate": 7.694995287320636e-05, "loss": 0.0085, "step": 28230 }, { "grad_norm": 0.12577369809150696, "learning_rate": 7.693254146925679e-05, "loss": 0.0091, "step": 28240 }, { "grad_norm": 0.19527952373027802, "learning_rate": 7.691512546327251e-05, "loss": 0.0088, "step": 28250 }, { "grad_norm": 0.10872604697942734, "learning_rate": 7.689770485822939e-05, "loss": 0.0093, "step": 28260 }, { "grad_norm": 0.1477864682674408, "learning_rate": 7.688027965710416e-05, "loss": 0.0082, "step": 28270 }, { "grad_norm": 0.11826980113983154, "learning_rate": 7.686284986287433e-05, "loss": 0.0101, "step": 28280 }, { "grad_norm": 0.16669824719429016, "learning_rate": 7.684541547851817e-05, "loss": 0.0085, "step": 28290 }, { "grad_norm": 0.13230767846107483, "learning_rate": 7.68279765070147e-05, "loss": 0.0098, "step": 28300 }, { "grad_norm": 0.12384877353906631, "learning_rate": 7.68105329513438e-05, "loss": 0.0102, "step": 28310 }, { "grad_norm": 0.1491764783859253, "learning_rate": 7.67930848144861e-05, "loss": 0.0109, "step": 28320 }, { "grad_norm": 0.18160147964954376, "learning_rate": 7.6775632099423e-05, "loss": 0.0087, "step": 28330 }, { "grad_norm": 0.1410718709230423, "learning_rate": 7.675817480913667e-05, "loss": 0.0092, "step": 28340 }, { "grad_norm": 0.1696975976228714, "learning_rate": 7.674071294661011e-05, "loss": 0.0085, "step": 28350 }, { "grad_norm": 0.17209608852863312, "learning_rate": 7.672324651482707e-05, "loss": 0.0089, "step": 28360 }, { "grad_norm": 0.2528674304485321, "learning_rate": 7.670577551677209e-05, "loss": 0.0091, "step": 28370 }, { "grad_norm": 0.2023579478263855, "learning_rate": 7.668829995543047e-05, "loss": 0.0098, "step": 28380 }, { "grad_norm": 0.14989234507083893, "learning_rate": 7.667081983378832e-05, "loss": 0.0098, "step": 28390 }, { "grad_norm": 0.12037703394889832, "learning_rate": 7.66533351548325e-05, "loss": 0.0073, "step": 28400 }, { "grad_norm": 0.14752274751663208, "learning_rate": 7.663584592155069e-05, "loss": 0.0115, "step": 28410 }, { "grad_norm": 0.1362091302871704, "learning_rate": 7.661835213693129e-05, "loss": 0.0084, "step": 28420 }, { "grad_norm": 0.14786559343338013, "learning_rate": 7.660085380396353e-05, "loss": 0.0091, "step": 28430 }, { "grad_norm": 0.13841243088245392, "learning_rate": 7.658335092563738e-05, "loss": 0.0088, "step": 28440 }, { "grad_norm": 0.1801440268754959, "learning_rate": 7.656584350494362e-05, "loss": 0.0096, "step": 28450 }, { "grad_norm": 0.2033512443304062, "learning_rate": 7.654833154487378e-05, "loss": 0.0095, "step": 28460 }, { "grad_norm": 0.1762855350971222, "learning_rate": 7.653081504842017e-05, "loss": 0.0109, "step": 28470 }, { "grad_norm": 0.17419801652431488, "learning_rate": 7.65132940185759e-05, "loss": 0.0107, "step": 28480 }, { "grad_norm": 0.1308082640171051, "learning_rate": 7.649576845833481e-05, "loss": 0.0086, "step": 28490 }, { "grad_norm": 0.15253648161888123, "learning_rate": 7.647823837069156e-05, "loss": 0.0096, "step": 28500 }, { "grad_norm": 0.12214966118335724, "learning_rate": 7.646070375864156e-05, "loss": 0.0071, "step": 28510 }, { "grad_norm": 0.1765369027853012, "learning_rate": 7.644316462518099e-05, "loss": 0.0082, "step": 28520 }, { "grad_norm": 0.1538756638765335, "learning_rate": 7.642562097330679e-05, "loss": 0.0097, "step": 28530 }, { "grad_norm": 0.1352345049381256, "learning_rate": 7.640807280601671e-05, "loss": 0.0091, "step": 28540 }, { "grad_norm": 0.1675070971250534, "learning_rate": 7.639052012630927e-05, "loss": 0.0096, "step": 28550 }, { "grad_norm": 0.14616337418556213, "learning_rate": 7.63729629371837e-05, "loss": 0.0099, "step": 28560 }, { "grad_norm": 0.13287366926670074, "learning_rate": 7.635540124164009e-05, "loss": 0.0101, "step": 28570 }, { "grad_norm": 0.09504611045122147, "learning_rate": 7.633783504267922e-05, "loss": 0.0093, "step": 28580 }, { "grad_norm": 0.18526355922222137, "learning_rate": 7.632026434330269e-05, "loss": 0.0089, "step": 28590 }, { "grad_norm": 0.12713584303855896, "learning_rate": 7.630268914651282e-05, "loss": 0.0087, "step": 28600 }, { "grad_norm": 0.22180411219596863, "learning_rate": 7.628510945531278e-05, "loss": 0.0097, "step": 28610 }, { "grad_norm": 0.13537707924842834, "learning_rate": 7.626752527270641e-05, "loss": 0.0095, "step": 28620 }, { "grad_norm": 0.15199492871761322, "learning_rate": 7.62499366016984e-05, "loss": 0.0115, "step": 28630 }, { "grad_norm": 0.14082209765911102, "learning_rate": 7.623234344529416e-05, "loss": 0.0078, "step": 28640 }, { "grad_norm": 0.196713387966156, "learning_rate": 7.62147458064999e-05, "loss": 0.0107, "step": 28650 }, { "grad_norm": 0.15024565160274506, "learning_rate": 7.619714368832254e-05, "loss": 0.0081, "step": 28660 }, { "grad_norm": 0.11645779013633728, "learning_rate": 7.61795370937698e-05, "loss": 0.0087, "step": 28670 }, { "grad_norm": 0.17162072658538818, "learning_rate": 7.61619260258502e-05, "loss": 0.0102, "step": 28680 }, { "grad_norm": 0.15654055774211884, "learning_rate": 7.614431048757298e-05, "loss": 0.0103, "step": 28690 }, { "grad_norm": 0.16216720640659332, "learning_rate": 7.612669048194814e-05, "loss": 0.0101, "step": 28700 }, { "grad_norm": 0.14586612582206726, "learning_rate": 7.610906601198646e-05, "loss": 0.0106, "step": 28710 }, { "grad_norm": 0.0970478430390358, "learning_rate": 7.60914370806995e-05, "loss": 0.0073, "step": 28720 }, { "grad_norm": 0.124961256980896, "learning_rate": 7.607380369109953e-05, "loss": 0.01, "step": 28730 }, { "grad_norm": 0.19023914635181427, "learning_rate": 7.605616584619961e-05, "loss": 0.0079, "step": 28740 }, { "grad_norm": 0.17457818984985352, "learning_rate": 7.603852354901362e-05, "loss": 0.0101, "step": 28750 }, { "grad_norm": 0.14488853514194489, "learning_rate": 7.602087680255609e-05, "loss": 0.0094, "step": 28760 }, { "grad_norm": 0.151014506816864, "learning_rate": 7.600322560984238e-05, "loss": 0.0099, "step": 28770 }, { "grad_norm": 0.16092370450496674, "learning_rate": 7.598556997388863e-05, "loss": 0.0102, "step": 28780 }, { "grad_norm": 0.12269136309623718, "learning_rate": 7.596790989771166e-05, "loss": 0.008, "step": 28790 }, { "grad_norm": 0.13091275095939636, "learning_rate": 7.595024538432914e-05, "loss": 0.0081, "step": 28800 }, { "grad_norm": 0.09798149019479752, "learning_rate": 7.59325764367594e-05, "loss": 0.0083, "step": 28810 }, { "grad_norm": 0.11454281210899353, "learning_rate": 7.59149030580216e-05, "loss": 0.0118, "step": 28820 }, { "grad_norm": 0.09296300262212753, "learning_rate": 7.589722525113562e-05, "loss": 0.0073, "step": 28830 }, { "grad_norm": 0.16627521812915802, "learning_rate": 7.587954301912216e-05, "loss": 0.0095, "step": 28840 }, { "grad_norm": 0.13852596282958984, "learning_rate": 7.586185636500263e-05, "loss": 0.0081, "step": 28850 }, { "grad_norm": 0.14847271144390106, "learning_rate": 7.584416529179914e-05, "loss": 0.0082, "step": 28860 }, { "grad_norm": 0.1280481517314911, "learning_rate": 7.582646980253465e-05, "loss": 0.0086, "step": 28870 }, { "grad_norm": 0.13948512077331543, "learning_rate": 7.580876990023282e-05, "loss": 0.0087, "step": 28880 }, { "grad_norm": 0.11039379984140396, "learning_rate": 7.579106558791809e-05, "loss": 0.0076, "step": 28890 }, { "grad_norm": 0.14888307452201843, "learning_rate": 7.577335686861565e-05, "loss": 0.0098, "step": 28900 }, { "grad_norm": 0.20173980295658112, "learning_rate": 7.575564374535141e-05, "loss": 0.008, "step": 28910 }, { "grad_norm": 0.15542829036712646, "learning_rate": 7.573792622115207e-05, "loss": 0.0074, "step": 28920 }, { "grad_norm": 0.15418048202991486, "learning_rate": 7.572020429904507e-05, "loss": 0.0109, "step": 28930 }, { "grad_norm": 0.1630503088235855, "learning_rate": 7.570247798205861e-05, "loss": 0.0085, "step": 28940 }, { "grad_norm": 0.16644644737243652, "learning_rate": 7.568474727322164e-05, "loss": 0.009, "step": 28950 }, { "grad_norm": 0.1703784316778183, "learning_rate": 7.566701217556384e-05, "loss": 0.0121, "step": 28960 }, { "grad_norm": 0.13248780369758606, "learning_rate": 7.564927269211564e-05, "loss": 0.0098, "step": 28970 }, { "grad_norm": 0.10057955980300903, "learning_rate": 7.563152882590824e-05, "loss": 0.0088, "step": 28980 }, { "grad_norm": 0.22589875757694244, "learning_rate": 7.56137805799736e-05, "loss": 0.0107, "step": 28990 }, { "grad_norm": 0.15861548483371735, "learning_rate": 7.559602795734439e-05, "loss": 0.0099, "step": 29000 }, { "grad_norm": 0.1206212043762207, "learning_rate": 7.557827096105408e-05, "loss": 0.0094, "step": 29010 }, { "grad_norm": 0.1427171379327774, "learning_rate": 7.55605095941368e-05, "loss": 0.01, "step": 29020 }, { "grad_norm": 0.13868676126003265, "learning_rate": 7.554274385962752e-05, "loss": 0.0086, "step": 29030 }, { "grad_norm": 0.14070408046245575, "learning_rate": 7.552497376056191e-05, "loss": 0.012, "step": 29040 }, { "grad_norm": 0.13755278289318085, "learning_rate": 7.550719929997639e-05, "loss": 0.0071, "step": 29050 }, { "grad_norm": 0.13083961606025696, "learning_rate": 7.548942048090813e-05, "loss": 0.0083, "step": 29060 }, { "grad_norm": 0.17375311255455017, "learning_rate": 7.547163730639506e-05, "loss": 0.0099, "step": 29070 }, { "grad_norm": 0.11778810620307922, "learning_rate": 7.545384977947583e-05, "loss": 0.0087, "step": 29080 }, { "grad_norm": 0.1170424148440361, "learning_rate": 7.543605790318981e-05, "loss": 0.0091, "step": 29090 }, { "grad_norm": 0.19610176980495453, "learning_rate": 7.54182616805772e-05, "loss": 0.0092, "step": 29100 }, { "grad_norm": 0.19876962900161743, "learning_rate": 7.540046111467885e-05, "loss": 0.008, "step": 29110 }, { "grad_norm": 0.21486397087574005, "learning_rate": 7.53826562085364e-05, "loss": 0.0089, "step": 29120 }, { "grad_norm": 0.14913812279701233, "learning_rate": 7.536484696519221e-05, "loss": 0.0093, "step": 29130 }, { "grad_norm": 0.12779377400875092, "learning_rate": 7.534703338768942e-05, "loss": 0.0116, "step": 29140 }, { "grad_norm": 0.16591724753379822, "learning_rate": 7.532921547907185e-05, "loss": 0.01, "step": 29150 }, { "grad_norm": 0.14631251990795135, "learning_rate": 7.531139324238412e-05, "loss": 0.0103, "step": 29160 }, { "grad_norm": 0.1724787801504135, "learning_rate": 7.529356668067157e-05, "loss": 0.0097, "step": 29170 }, { "grad_norm": 0.1673765480518341, "learning_rate": 7.527573579698023e-05, "loss": 0.0095, "step": 29180 }, { "grad_norm": 0.14547638595104218, "learning_rate": 7.525790059435693e-05, "loss": 0.0097, "step": 29190 }, { "grad_norm": 0.091932013630867, "learning_rate": 7.524006107584926e-05, "loss": 0.0078, "step": 29200 }, { "grad_norm": 0.14194415509700775, "learning_rate": 7.522221724450544e-05, "loss": 0.0077, "step": 29210 }, { "grad_norm": 0.1671827733516693, "learning_rate": 7.520436910337451e-05, "loss": 0.0094, "step": 29220 }, { "grad_norm": 0.18675172328948975, "learning_rate": 7.518651665550627e-05, "loss": 0.0095, "step": 29230 }, { "grad_norm": 0.21546170115470886, "learning_rate": 7.516865990395117e-05, "loss": 0.0116, "step": 29240 }, { "grad_norm": 0.1564439982175827, "learning_rate": 7.515079885176047e-05, "loss": 0.0098, "step": 29250 }, { "grad_norm": 0.1345544308423996, "learning_rate": 7.513293350198612e-05, "loss": 0.0078, "step": 29260 }, { "grad_norm": 0.1027713268995285, "learning_rate": 7.511506385768081e-05, "loss": 0.0087, "step": 29270 }, { "grad_norm": 0.12549838423728943, "learning_rate": 7.509718992189801e-05, "loss": 0.0078, "step": 29280 }, { "grad_norm": 0.15048140287399292, "learning_rate": 7.507931169769182e-05, "loss": 0.0089, "step": 29290 }, { "grad_norm": 0.1582394242286682, "learning_rate": 7.506142918811722e-05, "loss": 0.0099, "step": 29300 }, { "grad_norm": 0.1679166704416275, "learning_rate": 7.504354239622978e-05, "loss": 0.0083, "step": 29310 }, { "grad_norm": 0.15460707247257233, "learning_rate": 7.50256513250859e-05, "loss": 0.0085, "step": 29320 }, { "grad_norm": 0.09950435161590576, "learning_rate": 7.500775597774265e-05, "loss": 0.0095, "step": 29330 }, { "grad_norm": 0.12094581127166748, "learning_rate": 7.498985635725788e-05, "loss": 0.0072, "step": 29340 }, { "grad_norm": 0.13730162382125854, "learning_rate": 7.497195246669012e-05, "loss": 0.0091, "step": 29350 }, { "grad_norm": 0.13735538721084595, "learning_rate": 7.495404430909868e-05, "loss": 0.0093, "step": 29360 }, { "grad_norm": 0.09191171079874039, "learning_rate": 7.493613188754356e-05, "loss": 0.0092, "step": 29370 }, { "grad_norm": 0.14467813074588776, "learning_rate": 7.49182152050855e-05, "loss": 0.0077, "step": 29380 }, { "grad_norm": 0.15029028058052063, "learning_rate": 7.490029426478598e-05, "loss": 0.0081, "step": 29390 }, { "grad_norm": 0.1542588472366333, "learning_rate": 7.488236906970719e-05, "loss": 0.0083, "step": 29400 }, { "grad_norm": 0.14004909992218018, "learning_rate": 7.486443962291207e-05, "loss": 0.0072, "step": 29410 }, { "grad_norm": 0.15191933512687683, "learning_rate": 7.484650592746424e-05, "loss": 0.008, "step": 29420 }, { "grad_norm": 0.13507653772830963, "learning_rate": 7.482856798642811e-05, "loss": 0.0091, "step": 29430 }, { "grad_norm": 0.12626972794532776, "learning_rate": 7.481062580286878e-05, "loss": 0.009, "step": 29440 }, { "grad_norm": 0.11723260581493378, "learning_rate": 7.479267937985208e-05, "loss": 0.009, "step": 29450 }, { "grad_norm": 0.12205787003040314, "learning_rate": 7.477472872044456e-05, "loss": 0.0093, "step": 29460 }, { "grad_norm": 0.1653633713722229, "learning_rate": 7.475677382771347e-05, "loss": 0.0072, "step": 29470 }, { "grad_norm": 0.16143366694450378, "learning_rate": 7.473881470472683e-05, "loss": 0.0088, "step": 29480 }, { "grad_norm": 0.12395932525396347, "learning_rate": 7.472085135455337e-05, "loss": 0.0075, "step": 29490 }, { "grad_norm": 0.16442151367664337, "learning_rate": 7.470288378026256e-05, "loss": 0.009, "step": 29500 }, { "grad_norm": 0.12769626080989838, "learning_rate": 7.468491198492451e-05, "loss": 0.0098, "step": 29510 }, { "grad_norm": 0.15546245872974396, "learning_rate": 7.466693597161013e-05, "loss": 0.0097, "step": 29520 }, { "grad_norm": 0.16635747253894806, "learning_rate": 7.464895574339104e-05, "loss": 0.0073, "step": 29530 }, { "grad_norm": 0.14186394214630127, "learning_rate": 7.463097130333958e-05, "loss": 0.0112, "step": 29540 }, { "grad_norm": 0.17688588798046112, "learning_rate": 7.461298265452876e-05, "loss": 0.0078, "step": 29550 }, { "grad_norm": 0.1637892872095108, "learning_rate": 7.459498980003239e-05, "loss": 0.0095, "step": 29560 }, { "grad_norm": 0.11147934198379517, "learning_rate": 7.457699274292493e-05, "loss": 0.0077, "step": 29570 }, { "grad_norm": 0.1117839366197586, "learning_rate": 7.455899148628159e-05, "loss": 0.0074, "step": 29580 }, { "grad_norm": 0.11833705753087997, "learning_rate": 7.45409860331783e-05, "loss": 0.0073, "step": 29590 }, { "grad_norm": 0.23692169785499573, "learning_rate": 7.452297638669169e-05, "loss": 0.0109, "step": 29600 }, { "grad_norm": 0.210484117269516, "learning_rate": 7.450496254989911e-05, "loss": 0.012, "step": 29610 }, { "grad_norm": 0.13378816843032837, "learning_rate": 7.448694452587866e-05, "loss": 0.0113, "step": 29620 }, { "grad_norm": 0.14953383803367615, "learning_rate": 7.44689223177091e-05, "loss": 0.009, "step": 29630 }, { "grad_norm": 0.13016103208065033, "learning_rate": 7.445089592846994e-05, "loss": 0.0085, "step": 29640 }, { "grad_norm": 0.10241067409515381, "learning_rate": 7.443286536124141e-05, "loss": 0.0089, "step": 29650 }, { "grad_norm": 0.16326507925987244, "learning_rate": 7.441483061910443e-05, "loss": 0.0089, "step": 29660 }, { "grad_norm": 0.1268124282360077, "learning_rate": 7.439679170514064e-05, "loss": 0.0079, "step": 29670 }, { "grad_norm": 0.13118094205856323, "learning_rate": 7.43787486224324e-05, "loss": 0.0079, "step": 29680 }, { "grad_norm": 0.15019313991069794, "learning_rate": 7.436070137406276e-05, "loss": 0.0087, "step": 29690 }, { "grad_norm": 0.1826428771018982, "learning_rate": 7.434264996311556e-05, "loss": 0.0104, "step": 29700 }, { "grad_norm": 0.18195563554763794, "learning_rate": 7.432459439267525e-05, "loss": 0.0089, "step": 29710 }, { "grad_norm": 0.14729633927345276, "learning_rate": 7.430653466582701e-05, "loss": 0.0092, "step": 29720 }, { "grad_norm": 0.12516872584819794, "learning_rate": 7.42884707856568e-05, "loss": 0.0126, "step": 29730 }, { "grad_norm": 0.1253841072320938, "learning_rate": 7.427040275525122e-05, "loss": 0.0068, "step": 29740 }, { "grad_norm": 0.14429941773414612, "learning_rate": 7.42523305776976e-05, "loss": 0.0102, "step": 29750 }, { "grad_norm": 0.12620623409748077, "learning_rate": 7.4234254256084e-05, "loss": 0.0089, "step": 29760 }, { "grad_norm": 0.1257653832435608, "learning_rate": 7.421617379349915e-05, "loss": 0.0102, "step": 29770 }, { "grad_norm": 0.13234011828899384, "learning_rate": 7.41980891930325e-05, "loss": 0.0096, "step": 29780 }, { "grad_norm": 0.15901681780815125, "learning_rate": 7.418000045777425e-05, "loss": 0.0085, "step": 29790 }, { "grad_norm": 0.16309338808059692, "learning_rate": 7.416190759081523e-05, "loss": 0.0125, "step": 29800 }, { "grad_norm": 0.18695950508117676, "learning_rate": 7.414381059524704e-05, "loss": 0.0073, "step": 29810 }, { "grad_norm": 0.19431999325752258, "learning_rate": 7.412570947416195e-05, "loss": 0.0092, "step": 29820 }, { "grad_norm": 0.20234908163547516, "learning_rate": 7.410760423065295e-05, "loss": 0.0104, "step": 29830 }, { "grad_norm": 0.18681958317756653, "learning_rate": 7.408949486781372e-05, "loss": 0.0088, "step": 29840 }, { "grad_norm": 0.19801069796085358, "learning_rate": 7.407138138873868e-05, "loss": 0.0109, "step": 29850 }, { "grad_norm": 0.13264842331409454, "learning_rate": 7.405326379652292e-05, "loss": 0.0116, "step": 29860 }, { "grad_norm": 0.16374462842941284, "learning_rate": 7.403514209426222e-05, "loss": 0.0099, "step": 29870 }, { "grad_norm": 0.13161960244178772, "learning_rate": 7.40170162850531e-05, "loss": 0.0118, "step": 29880 }, { "grad_norm": 0.1531296819448471, "learning_rate": 7.399888637199278e-05, "loss": 0.0125, "step": 29890 }, { "grad_norm": 0.1409519463777542, "learning_rate": 7.398075235817914e-05, "loss": 0.0093, "step": 29900 }, { "grad_norm": 0.2197212278842926, "learning_rate": 7.39626142467108e-05, "loss": 0.0106, "step": 29910 }, { "grad_norm": 0.08194482326507568, "learning_rate": 7.394447204068706e-05, "loss": 0.0089, "step": 29920 }, { "grad_norm": 0.15485535562038422, "learning_rate": 7.392632574320793e-05, "loss": 0.0117, "step": 29930 }, { "grad_norm": 0.10918758809566498, "learning_rate": 7.390817535737411e-05, "loss": 0.0102, "step": 29940 }, { "grad_norm": 0.147957906126976, "learning_rate": 7.389002088628703e-05, "loss": 0.0098, "step": 29950 }, { "grad_norm": 0.12990622222423553, "learning_rate": 7.387186233304877e-05, "loss": 0.008, "step": 29960 }, { "grad_norm": 0.15864302217960358, "learning_rate": 7.385369970076212e-05, "loss": 0.0081, "step": 29970 }, { "grad_norm": 0.11586452275514603, "learning_rate": 7.38355329925306e-05, "loss": 0.0072, "step": 29980 }, { "grad_norm": 0.13457463681697845, "learning_rate": 7.381736221145838e-05, "loss": 0.0081, "step": 29990 }, { "grad_norm": 0.12805552780628204, "learning_rate": 7.37991873606504e-05, "loss": 0.0086, "step": 30000 }, { "grad_norm": 0.16539442539215088, "learning_rate": 7.378100844321218e-05, "loss": 0.0078, "step": 30010 }, { "grad_norm": 0.11866694688796997, "learning_rate": 7.376282546225004e-05, "loss": 0.0123, "step": 30020 }, { "grad_norm": 0.20622456073760986, "learning_rate": 7.374463842087094e-05, "loss": 0.0097, "step": 30030 }, { "grad_norm": 0.08855980634689331, "learning_rate": 7.372644732218254e-05, "loss": 0.0075, "step": 30040 }, { "grad_norm": 0.15500979125499725, "learning_rate": 7.370825216929322e-05, "loss": 0.0108, "step": 30050 }, { "grad_norm": 0.15543141961097717, "learning_rate": 7.369005296531205e-05, "loss": 0.0103, "step": 30060 }, { "grad_norm": 0.1071658805012703, "learning_rate": 7.367184971334873e-05, "loss": 0.008, "step": 30070 }, { "grad_norm": 0.09441720694303513, "learning_rate": 7.365364241651371e-05, "loss": 0.0079, "step": 30080 }, { "grad_norm": 0.10495318472385406, "learning_rate": 7.363543107791815e-05, "loss": 0.0094, "step": 30090 }, { "grad_norm": 0.146455317735672, "learning_rate": 7.361721570067384e-05, "loss": 0.0106, "step": 30100 }, { "grad_norm": 0.1375695914030075, "learning_rate": 7.359899628789331e-05, "loss": 0.0099, "step": 30110 }, { "grad_norm": 0.10393136739730835, "learning_rate": 7.358077284268974e-05, "loss": 0.0084, "step": 30120 }, { "grad_norm": 0.17772094905376434, "learning_rate": 7.356254536817702e-05, "loss": 0.0086, "step": 30130 }, { "grad_norm": 0.18132798373699188, "learning_rate": 7.354431386746973e-05, "loss": 0.0131, "step": 30140 }, { "grad_norm": 0.1447572112083435, "learning_rate": 7.352607834368316e-05, "loss": 0.0105, "step": 30150 }, { "grad_norm": 0.15443536639213562, "learning_rate": 7.350783879993324e-05, "loss": 0.0098, "step": 30160 }, { "grad_norm": 0.13144628703594208, "learning_rate": 7.348959523933658e-05, "loss": 0.0111, "step": 30170 }, { "grad_norm": 0.275363564491272, "learning_rate": 7.347134766501057e-05, "loss": 0.0097, "step": 30180 }, { "grad_norm": 0.1628565788269043, "learning_rate": 7.345309608007315e-05, "loss": 0.0093, "step": 30190 }, { "grad_norm": 0.21232491731643677, "learning_rate": 7.343484048764308e-05, "loss": 0.0113, "step": 30200 }, { "grad_norm": 0.129865825176239, "learning_rate": 7.341658089083972e-05, "loss": 0.0106, "step": 30210 }, { "grad_norm": 0.18057383596897125, "learning_rate": 7.339831729278313e-05, "loss": 0.0122, "step": 30220 }, { "grad_norm": 0.1658496856689453, "learning_rate": 7.338004969659404e-05, "loss": 0.0085, "step": 30230 }, { "grad_norm": 0.14341852068901062, "learning_rate": 7.336177810539391e-05, "loss": 0.01, "step": 30240 }, { "grad_norm": 0.17179885506629944, "learning_rate": 7.334350252230485e-05, "loss": 0.012, "step": 30250 }, { "grad_norm": 0.15223847329616547, "learning_rate": 7.332522295044965e-05, "loss": 0.0082, "step": 30260 }, { "grad_norm": 0.16302621364593506, "learning_rate": 7.33069393929518e-05, "loss": 0.0099, "step": 30270 }, { "grad_norm": 0.18123076856136322, "learning_rate": 7.328865185293545e-05, "loss": 0.0118, "step": 30280 }, { "grad_norm": 0.17295357584953308, "learning_rate": 7.327036033352546e-05, "loss": 0.0101, "step": 30290 }, { "grad_norm": 0.18173766136169434, "learning_rate": 7.325206483784733e-05, "loss": 0.009, "step": 30300 }, { "grad_norm": 0.0980154424905777, "learning_rate": 7.323376536902724e-05, "loss": 0.0081, "step": 30310 }, { "grad_norm": 0.14705625176429749, "learning_rate": 7.321546193019213e-05, "loss": 0.0102, "step": 30320 }, { "grad_norm": 0.12620466947555542, "learning_rate": 7.31971545244695e-05, "loss": 0.0093, "step": 30330 }, { "grad_norm": 0.15772458910942078, "learning_rate": 7.31788431549876e-05, "loss": 0.0105, "step": 30340 }, { "grad_norm": 0.14908091723918915, "learning_rate": 7.316052782487534e-05, "loss": 0.0092, "step": 30350 }, { "grad_norm": 0.2581122815608978, "learning_rate": 7.314220853726234e-05, "loss": 0.01, "step": 30360 }, { "grad_norm": 0.196310356259346, "learning_rate": 7.312388529527884e-05, "loss": 0.0117, "step": 30370 }, { "grad_norm": 0.21913619339466095, "learning_rate": 7.310555810205577e-05, "loss": 0.0115, "step": 30380 }, { "grad_norm": 0.20016834139823914, "learning_rate": 7.308722696072476e-05, "loss": 0.0114, "step": 30390 }, { "grad_norm": 0.12298766523599625, "learning_rate": 7.306889187441811e-05, "loss": 0.0089, "step": 30400 }, { "grad_norm": 0.1669459044933319, "learning_rate": 7.305055284626876e-05, "loss": 0.0099, "step": 30410 }, { "grad_norm": 0.12455475330352783, "learning_rate": 7.303220987941037e-05, "loss": 0.0095, "step": 30420 }, { "grad_norm": 0.1309415102005005, "learning_rate": 7.301386297697726e-05, "loss": 0.0068, "step": 30430 }, { "grad_norm": 0.14928773045539856, "learning_rate": 7.299551214210438e-05, "loss": 0.0073, "step": 30440 }, { "grad_norm": 0.19866709411144257, "learning_rate": 7.297715737792738e-05, "loss": 0.009, "step": 30450 }, { "grad_norm": 0.1083252876996994, "learning_rate": 7.295879868758265e-05, "loss": 0.0081, "step": 30460 }, { "grad_norm": 0.13570040464401245, "learning_rate": 7.294043607420713e-05, "loss": 0.009, "step": 30470 }, { "grad_norm": 0.14112524688243866, "learning_rate": 7.292206954093852e-05, "loss": 0.0083, "step": 30480 }, { "grad_norm": 0.21329064667224884, "learning_rate": 7.290369909091515e-05, "loss": 0.0092, "step": 30490 }, { "grad_norm": 0.14372555911540985, "learning_rate": 7.2885324727276e-05, "loss": 0.0099, "step": 30500 }, { "grad_norm": 0.1731569766998291, "learning_rate": 7.286694645316076e-05, "loss": 0.0089, "step": 30510 }, { "grad_norm": 0.16546759009361267, "learning_rate": 7.284856427170982e-05, "loss": 0.0094, "step": 30520 }, { "grad_norm": 0.10861140489578247, "learning_rate": 7.283017818606414e-05, "loss": 0.0103, "step": 30530 }, { "grad_norm": 0.10322161018848419, "learning_rate": 7.28117881993654e-05, "loss": 0.0084, "step": 30540 }, { "grad_norm": 0.13114246726036072, "learning_rate": 7.279339431475598e-05, "loss": 0.0077, "step": 30550 }, { "grad_norm": 0.14509578049182892, "learning_rate": 7.277499653537887e-05, "loss": 0.0114, "step": 30560 }, { "grad_norm": 0.17001815140247345, "learning_rate": 7.275659486437776e-05, "loss": 0.0129, "step": 30570 }, { "grad_norm": 0.12943527102470398, "learning_rate": 7.273818930489695e-05, "loss": 0.0117, "step": 30580 }, { "grad_norm": 0.12709373235702515, "learning_rate": 7.271977986008151e-05, "loss": 0.0083, "step": 30590 }, { "grad_norm": 0.14567671716213226, "learning_rate": 7.270136653307705e-05, "loss": 0.0084, "step": 30600 }, { "grad_norm": 0.12682101130485535, "learning_rate": 7.268294932702994e-05, "loss": 0.0104, "step": 30610 }, { "grad_norm": 0.15878300368785858, "learning_rate": 7.266452824508719e-05, "loss": 0.0072, "step": 30620 }, { "grad_norm": 0.1270492672920227, "learning_rate": 7.264610329039643e-05, "loss": 0.009, "step": 30630 }, { "grad_norm": 0.14777502417564392, "learning_rate": 7.262767446610599e-05, "loss": 0.0093, "step": 30640 }, { "grad_norm": 0.11728434264659882, "learning_rate": 7.260924177536485e-05, "loss": 0.0082, "step": 30650 }, { "grad_norm": 0.13728183507919312, "learning_rate": 7.259080522132265e-05, "loss": 0.0105, "step": 30660 }, { "grad_norm": 0.18678951263427734, "learning_rate": 7.257236480712972e-05, "loss": 0.0084, "step": 30670 }, { "grad_norm": 0.1154007762670517, "learning_rate": 7.255392053593697e-05, "loss": 0.008, "step": 30680 }, { "grad_norm": 0.16899190843105316, "learning_rate": 7.253547241089607e-05, "loss": 0.0094, "step": 30690 }, { "grad_norm": 0.1554338037967682, "learning_rate": 7.251702043515927e-05, "loss": 0.0095, "step": 30700 }, { "grad_norm": 0.1558094024658203, "learning_rate": 7.249856461187952e-05, "loss": 0.0095, "step": 30710 }, { "grad_norm": 0.12984062731266022, "learning_rate": 7.248010494421042e-05, "loss": 0.0098, "step": 30720 }, { "grad_norm": 0.14702460169792175, "learning_rate": 7.246164143530622e-05, "loss": 0.011, "step": 30730 }, { "grad_norm": 0.18946778774261475, "learning_rate": 7.244317408832181e-05, "loss": 0.0102, "step": 30740 }, { "grad_norm": 0.12440568953752518, "learning_rate": 7.242470290641279e-05, "loss": 0.0076, "step": 30750 }, { "grad_norm": 0.16675060987472534, "learning_rate": 7.240622789273536e-05, "loss": 0.0092, "step": 30760 }, { "grad_norm": 0.16421838104724884, "learning_rate": 7.238774905044638e-05, "loss": 0.0093, "step": 30770 }, { "grad_norm": 0.1384352445602417, "learning_rate": 7.236926638270341e-05, "loss": 0.0103, "step": 30780 }, { "grad_norm": 0.1820966750383377, "learning_rate": 7.23507798926646e-05, "loss": 0.0075, "step": 30790 }, { "grad_norm": 0.18712766468524933, "learning_rate": 7.23322895834888e-05, "loss": 0.011, "step": 30800 }, { "grad_norm": 0.15846551954746246, "learning_rate": 7.231379545833552e-05, "loss": 0.0102, "step": 30810 }, { "grad_norm": 0.16206394135951996, "learning_rate": 7.229529752036487e-05, "loss": 0.0106, "step": 30820 }, { "grad_norm": 0.1068892553448677, "learning_rate": 7.227679577273765e-05, "loss": 0.0097, "step": 30830 }, { "grad_norm": 0.1835152953863144, "learning_rate": 7.225829021861529e-05, "loss": 0.0074, "step": 30840 }, { "grad_norm": 0.11058258265256882, "learning_rate": 7.223978086115992e-05, "loss": 0.0076, "step": 30850 }, { "grad_norm": 0.1257384717464447, "learning_rate": 7.222126770353425e-05, "loss": 0.0103, "step": 30860 }, { "grad_norm": 0.12493560463190079, "learning_rate": 7.22027507489017e-05, "loss": 0.0116, "step": 30870 }, { "grad_norm": 0.1770094484090805, "learning_rate": 7.218423000042627e-05, "loss": 0.0106, "step": 30880 }, { "grad_norm": 0.1459859311580658, "learning_rate": 7.216570546127268e-05, "loss": 0.0083, "step": 30890 }, { "grad_norm": 0.1298588365316391, "learning_rate": 7.214717713460626e-05, "loss": 0.0098, "step": 30900 }, { "grad_norm": 0.1565939486026764, "learning_rate": 7.2128645023593e-05, "loss": 0.0092, "step": 30910 }, { "grad_norm": 0.13840548694133759, "learning_rate": 7.211010913139951e-05, "loss": 0.0098, "step": 30920 }, { "grad_norm": 0.14677579700946808, "learning_rate": 7.209156946119308e-05, "loss": 0.0083, "step": 30930 }, { "grad_norm": 0.1836932748556137, "learning_rate": 7.207302601614166e-05, "loss": 0.007, "step": 30940 }, { "grad_norm": 0.13588035106658936, "learning_rate": 7.205447879941378e-05, "loss": 0.0083, "step": 30950 }, { "grad_norm": 0.1201007068157196, "learning_rate": 7.203592781417866e-05, "loss": 0.0097, "step": 30960 }, { "grad_norm": 0.11161544919013977, "learning_rate": 7.201737306360617e-05, "loss": 0.0079, "step": 30970 }, { "grad_norm": 0.12282886356115341, "learning_rate": 7.19988145508668e-05, "loss": 0.0078, "step": 30980 }, { "grad_norm": 0.12214528769254684, "learning_rate": 7.198025227913168e-05, "loss": 0.0109, "step": 30990 }, { "grad_norm": 0.10702121257781982, "learning_rate": 7.196168625157261e-05, "loss": 0.0082, "step": 31000 }, { "grad_norm": 0.157353013753891, "learning_rate": 7.194311647136201e-05, "loss": 0.0098, "step": 31010 }, { "grad_norm": 0.1356114000082016, "learning_rate": 7.192454294167297e-05, "loss": 0.0099, "step": 31020 }, { "grad_norm": 0.10226580500602722, "learning_rate": 7.190596566567917e-05, "loss": 0.0074, "step": 31030 }, { "grad_norm": 0.16528508067131042, "learning_rate": 7.188738464655496e-05, "loss": 0.0116, "step": 31040 }, { "grad_norm": 0.13403509557247162, "learning_rate": 7.186879988747533e-05, "loss": 0.0083, "step": 31050 }, { "grad_norm": 0.11590493470430374, "learning_rate": 7.185021139161592e-05, "loss": 0.0082, "step": 31060 }, { "grad_norm": 0.09759602695703506, "learning_rate": 7.1831619162153e-05, "loss": 0.0089, "step": 31070 }, { "grad_norm": 0.1344018429517746, "learning_rate": 7.181302320226345e-05, "loss": 0.0093, "step": 31080 }, { "grad_norm": 0.13993613421916962, "learning_rate": 7.179442351512482e-05, "loss": 0.0075, "step": 31090 }, { "grad_norm": 0.15968771278858185, "learning_rate": 7.177582010391528e-05, "loss": 0.0091, "step": 31100 }, { "grad_norm": 0.1613226979970932, "learning_rate": 7.175721297181366e-05, "loss": 0.0085, "step": 31110 }, { "grad_norm": 0.17614947259426117, "learning_rate": 7.173860212199942e-05, "loss": 0.0096, "step": 31120 }, { "grad_norm": 0.18608230352401733, "learning_rate": 7.171998755765263e-05, "loss": 0.0083, "step": 31130 }, { "grad_norm": 0.13221345841884613, "learning_rate": 7.170136928195398e-05, "loss": 0.0085, "step": 31140 }, { "grad_norm": 0.176734060049057, "learning_rate": 7.168274729808489e-05, "loss": 0.0084, "step": 31150 }, { "grad_norm": 0.16883458197116852, "learning_rate": 7.166412160922728e-05, "loss": 0.0095, "step": 31160 }, { "grad_norm": 0.12945421040058136, "learning_rate": 7.164549221856382e-05, "loss": 0.0082, "step": 31170 }, { "grad_norm": 0.16304557025432587, "learning_rate": 7.162685912927775e-05, "loss": 0.0087, "step": 31180 }, { "grad_norm": 0.1554166078567505, "learning_rate": 7.160822234455294e-05, "loss": 0.0079, "step": 31190 }, { "grad_norm": 0.15588803589344025, "learning_rate": 7.158958186757391e-05, "loss": 0.0078, "step": 31200 }, { "grad_norm": 0.1536329984664917, "learning_rate": 7.157093770152582e-05, "loss": 0.0087, "step": 31210 }, { "grad_norm": 0.15679559111595154, "learning_rate": 7.155228984959446e-05, "loss": 0.0104, "step": 31220 }, { "grad_norm": 0.1416853368282318, "learning_rate": 7.153363831496621e-05, "loss": 0.008, "step": 31230 }, { "grad_norm": 0.13983581960201263, "learning_rate": 7.151498310082811e-05, "loss": 0.0113, "step": 31240 }, { "grad_norm": 0.11861738562583923, "learning_rate": 7.149632421036784e-05, "loss": 0.008, "step": 31250 }, { "grad_norm": 0.10681671649217606, "learning_rate": 7.147766164677369e-05, "loss": 0.0074, "step": 31260 }, { "grad_norm": 0.13818933069705963, "learning_rate": 7.145899541323459e-05, "loss": 0.0089, "step": 31270 }, { "grad_norm": 0.1470550000667572, "learning_rate": 7.144032551294007e-05, "loss": 0.0115, "step": 31280 }, { "grad_norm": 0.16973626613616943, "learning_rate": 7.14216519490803e-05, "loss": 0.0104, "step": 31290 }, { "grad_norm": 0.2165132313966751, "learning_rate": 7.140297472484609e-05, "loss": 0.0116, "step": 31300 }, { "grad_norm": 0.18586932122707367, "learning_rate": 7.138429384342891e-05, "loss": 0.0073, "step": 31310 }, { "grad_norm": 0.16259658336639404, "learning_rate": 7.136560930802074e-05, "loss": 0.0081, "step": 31320 }, { "grad_norm": 0.14399394392967224, "learning_rate": 7.134692112181431e-05, "loss": 0.0083, "step": 31330 }, { "grad_norm": 0.14753317832946777, "learning_rate": 7.13282292880029e-05, "loss": 0.0085, "step": 31340 }, { "grad_norm": 0.21801277995109558, "learning_rate": 7.130953380978043e-05, "loss": 0.0104, "step": 31350 }, { "grad_norm": 0.1504327952861786, "learning_rate": 7.129083469034144e-05, "loss": 0.0101, "step": 31360 }, { "grad_norm": 0.11399159580469131, "learning_rate": 7.127213193288112e-05, "loss": 0.0099, "step": 31370 }, { "grad_norm": 0.14965586364269257, "learning_rate": 7.125342554059522e-05, "loss": 0.0096, "step": 31380 }, { "grad_norm": 0.1626797467470169, "learning_rate": 7.12347155166802e-05, "loss": 0.0128, "step": 31390 }, { "grad_norm": 0.15877826511859894, "learning_rate": 7.121600186433306e-05, "loss": 0.01, "step": 31400 }, { "grad_norm": 0.15658330917358398, "learning_rate": 7.119728458675148e-05, "loss": 0.01, "step": 31410 }, { "grad_norm": 0.16116900742053986, "learning_rate": 7.117856368713369e-05, "loss": 0.0117, "step": 31420 }, { "grad_norm": 0.11924699693918228, "learning_rate": 7.115983916867861e-05, "loss": 0.0084, "step": 31430 }, { "grad_norm": 0.18154065310955048, "learning_rate": 7.114111103458574e-05, "loss": 0.0081, "step": 31440 }, { "grad_norm": 0.18134067952632904, "learning_rate": 7.11223792880552e-05, "loss": 0.0082, "step": 31450 }, { "grad_norm": 0.14576318860054016, "learning_rate": 7.110364393228773e-05, "loss": 0.009, "step": 31460 }, { "grad_norm": 0.13816680014133453, "learning_rate": 7.108490497048471e-05, "loss": 0.008, "step": 31470 }, { "grad_norm": 0.12281075865030289, "learning_rate": 7.10661624058481e-05, "loss": 0.0079, "step": 31480 }, { "grad_norm": 0.16298815608024597, "learning_rate": 7.10474162415805e-05, "loss": 0.009, "step": 31490 }, { "grad_norm": 0.12227577716112137, "learning_rate": 7.102866648088511e-05, "loss": 0.0092, "step": 31500 }, { "grad_norm": 0.12328702956438065, "learning_rate": 7.100991312696576e-05, "loss": 0.0074, "step": 31510 }, { "grad_norm": 0.12428238987922668, "learning_rate": 7.099115618302686e-05, "loss": 0.0073, "step": 31520 }, { "grad_norm": 0.15193286538124084, "learning_rate": 7.097239565227349e-05, "loss": 0.0075, "step": 31530 }, { "grad_norm": 0.1420663446187973, "learning_rate": 7.09536315379113e-05, "loss": 0.0101, "step": 31540 }, { "grad_norm": 0.1107199639081955, "learning_rate": 7.093486384314656e-05, "loss": 0.0121, "step": 31550 }, { "grad_norm": 0.11755646020174026, "learning_rate": 7.091609257118616e-05, "loss": 0.008, "step": 31560 }, { "grad_norm": 0.15830770134925842, "learning_rate": 7.08973177252376e-05, "loss": 0.0066, "step": 31570 }, { "grad_norm": 0.19403523206710815, "learning_rate": 7.087853930850898e-05, "loss": 0.0077, "step": 31580 }, { "grad_norm": 0.164156973361969, "learning_rate": 7.085975732420903e-05, "loss": 0.0082, "step": 31590 }, { "grad_norm": 0.14530552923679352, "learning_rate": 7.084097177554706e-05, "loss": 0.0089, "step": 31600 }, { "grad_norm": 0.13645032048225403, "learning_rate": 7.082218266573301e-05, "loss": 0.0078, "step": 31610 }, { "grad_norm": 0.20190219581127167, "learning_rate": 7.080338999797743e-05, "loss": 0.011, "step": 31620 }, { "grad_norm": 0.15219254791736603, "learning_rate": 7.07845937754915e-05, "loss": 0.0063, "step": 31630 }, { "grad_norm": 0.1467476338148117, "learning_rate": 7.076579400148693e-05, "loss": 0.0121, "step": 31640 }, { "grad_norm": 0.1616474837064743, "learning_rate": 7.074699067917611e-05, "loss": 0.0087, "step": 31650 }, { "grad_norm": 0.16109874844551086, "learning_rate": 7.072818381177201e-05, "loss": 0.0109, "step": 31660 }, { "grad_norm": 0.16443203389644623, "learning_rate": 7.070937340248823e-05, "loss": 0.0104, "step": 31670 }, { "grad_norm": 0.12128221243619919, "learning_rate": 7.069055945453893e-05, "loss": 0.0087, "step": 31680 }, { "grad_norm": 0.11078846454620361, "learning_rate": 7.067174197113892e-05, "loss": 0.0072, "step": 31690 }, { "grad_norm": 0.14475885033607483, "learning_rate": 7.065292095550355e-05, "loss": 0.0098, "step": 31700 }, { "grad_norm": 0.15980264544487, "learning_rate": 7.063409641084887e-05, "loss": 0.008, "step": 31710 }, { "grad_norm": 0.1365363746881485, "learning_rate": 7.061526834039145e-05, "loss": 0.0074, "step": 31720 }, { "grad_norm": 0.15054841339588165, "learning_rate": 7.05964367473485e-05, "loss": 0.0087, "step": 31730 }, { "grad_norm": 0.12994693219661713, "learning_rate": 7.057760163493783e-05, "loss": 0.0088, "step": 31740 }, { "grad_norm": 0.1529729962348938, "learning_rate": 7.055876300637783e-05, "loss": 0.0095, "step": 31750 }, { "grad_norm": 0.15852883458137512, "learning_rate": 7.053992086488753e-05, "loss": 0.0084, "step": 31760 }, { "grad_norm": 0.15262183547019958, "learning_rate": 7.052107521368651e-05, "loss": 0.0088, "step": 31770 }, { "grad_norm": 0.11503352224826813, "learning_rate": 7.0502226055995e-05, "loss": 0.0074, "step": 31780 }, { "grad_norm": 0.18802830576896667, "learning_rate": 7.048337339503379e-05, "loss": 0.0095, "step": 31790 }, { "grad_norm": 0.146648570895195, "learning_rate": 7.046451723402427e-05, "loss": 0.0093, "step": 31800 }, { "grad_norm": 0.10533671826124191, "learning_rate": 7.044565757618848e-05, "loss": 0.0088, "step": 31810 }, { "grad_norm": 0.12674690783023834, "learning_rate": 7.042679442474899e-05, "loss": 0.0076, "step": 31820 }, { "grad_norm": 0.10322245210409164, "learning_rate": 7.040792778292902e-05, "loss": 0.0088, "step": 31830 }, { "grad_norm": 0.1352733075618744, "learning_rate": 7.038905765395234e-05, "loss": 0.0084, "step": 31840 }, { "grad_norm": 0.14991238713264465, "learning_rate": 7.037018404104334e-05, "loss": 0.0083, "step": 31850 }, { "grad_norm": 0.15229302644729614, "learning_rate": 7.035130694742702e-05, "loss": 0.0082, "step": 31860 }, { "grad_norm": 0.17020025849342346, "learning_rate": 7.033242637632897e-05, "loss": 0.0073, "step": 31870 }, { "grad_norm": 0.10069385915994644, "learning_rate": 7.031354233097534e-05, "loss": 0.0077, "step": 31880 }, { "grad_norm": 0.11921391636133194, "learning_rate": 7.029465481459289e-05, "loss": 0.0067, "step": 31890 }, { "grad_norm": 0.1437440663576126, "learning_rate": 7.027576383040898e-05, "loss": 0.0071, "step": 31900 }, { "grad_norm": 0.13285881280899048, "learning_rate": 7.025686938165159e-05, "loss": 0.0086, "step": 31910 }, { "grad_norm": 0.19097834825515747, "learning_rate": 7.023797147154924e-05, "loss": 0.0075, "step": 31920 }, { "grad_norm": 0.12939687073230743, "learning_rate": 7.021907010333111e-05, "loss": 0.0077, "step": 31930 }, { "grad_norm": 0.13124065101146698, "learning_rate": 7.020016528022685e-05, "loss": 0.0113, "step": 31940 }, { "grad_norm": 0.17052163183689117, "learning_rate": 7.018125700546683e-05, "loss": 0.0082, "step": 31950 }, { "grad_norm": 0.17850501835346222, "learning_rate": 7.016234528228196e-05, "loss": 0.0112, "step": 31960 }, { "grad_norm": 0.14820900559425354, "learning_rate": 7.014343011390372e-05, "loss": 0.0079, "step": 31970 }, { "grad_norm": 0.1535813957452774, "learning_rate": 7.01245115035642e-05, "loss": 0.0104, "step": 31980 }, { "grad_norm": 0.13808734714984894, "learning_rate": 7.010558945449606e-05, "loss": 0.0076, "step": 31990 }, { "grad_norm": 0.18216945230960846, "learning_rate": 7.008666396993258e-05, "loss": 0.0089, "step": 32000 }, { "grad_norm": 0.15085163712501526, "learning_rate": 7.006773505310759e-05, "loss": 0.0104, "step": 32010 }, { "grad_norm": 0.17270028591156006, "learning_rate": 7.004880270725553e-05, "loss": 0.0082, "step": 32020 }, { "grad_norm": 0.13857445120811462, "learning_rate": 7.002986693561144e-05, "loss": 0.008, "step": 32030 }, { "grad_norm": 0.1306232213973999, "learning_rate": 7.001092774141089e-05, "loss": 0.0074, "step": 32040 }, { "grad_norm": 0.16537417471408844, "learning_rate": 6.999198512789009e-05, "loss": 0.0067, "step": 32050 }, { "grad_norm": 0.11492355167865753, "learning_rate": 6.997303909828584e-05, "loss": 0.0089, "step": 32060 }, { "grad_norm": 0.15644226968288422, "learning_rate": 6.995408965583544e-05, "loss": 0.0069, "step": 32070 }, { "grad_norm": 0.16976261138916016, "learning_rate": 6.993513680377688e-05, "loss": 0.0096, "step": 32080 }, { "grad_norm": 0.11952625960111618, "learning_rate": 6.991618054534868e-05, "loss": 0.0085, "step": 32090 }, { "grad_norm": 0.15230508148670197, "learning_rate": 6.989722088378991e-05, "loss": 0.0072, "step": 32100 }, { "grad_norm": 0.08681576699018478, "learning_rate": 6.987825782234027e-05, "loss": 0.0071, "step": 32110 }, { "grad_norm": 0.15302248299121857, "learning_rate": 6.985929136424006e-05, "loss": 0.0087, "step": 32120 }, { "grad_norm": 0.17234550416469574, "learning_rate": 6.984032151273012e-05, "loss": 0.0085, "step": 32130 }, { "grad_norm": 0.11639297008514404, "learning_rate": 6.982134827105186e-05, "loss": 0.0083, "step": 32140 }, { "grad_norm": 0.09888231009244919, "learning_rate": 6.980237164244729e-05, "loss": 0.0098, "step": 32150 }, { "grad_norm": 0.12418515980243683, "learning_rate": 6.9783391630159e-05, "loss": 0.0074, "step": 32160 }, { "grad_norm": 0.2151973992586136, "learning_rate": 6.976440823743015e-05, "loss": 0.0076, "step": 32170 }, { "grad_norm": 0.16489416360855103, "learning_rate": 6.974542146750451e-05, "loss": 0.0082, "step": 32180 }, { "grad_norm": 0.11645040661096573, "learning_rate": 6.972643132362637e-05, "loss": 0.0065, "step": 32190 }, { "grad_norm": 0.11984441429376602, "learning_rate": 6.970743780904064e-05, "loss": 0.0068, "step": 32200 }, { "grad_norm": 0.14740554988384247, "learning_rate": 6.968844092699277e-05, "loss": 0.0115, "step": 32210 }, { "grad_norm": 0.14697711169719696, "learning_rate": 6.966944068072883e-05, "loss": 0.0078, "step": 32220 }, { "grad_norm": 0.1599210798740387, "learning_rate": 6.965043707349545e-05, "loss": 0.0086, "step": 32230 }, { "grad_norm": 0.1639052927494049, "learning_rate": 6.963143010853982e-05, "loss": 0.0088, "step": 32240 }, { "grad_norm": 0.1256527304649353, "learning_rate": 6.961241978910971e-05, "loss": 0.0065, "step": 32250 }, { "grad_norm": 0.11513933539390564, "learning_rate": 6.959340611845344e-05, "loss": 0.01, "step": 32260 }, { "grad_norm": 0.08908367156982422, "learning_rate": 6.957438909981995e-05, "loss": 0.0078, "step": 32270 }, { "grad_norm": 0.1688232570886612, "learning_rate": 6.955536873645872e-05, "loss": 0.0106, "step": 32280 }, { "grad_norm": 0.16037923097610474, "learning_rate": 6.953634503161982e-05, "loss": 0.0095, "step": 32290 }, { "grad_norm": 0.12209872156381607, "learning_rate": 6.951731798855387e-05, "loss": 0.0088, "step": 32300 }, { "grad_norm": 0.1964162290096283, "learning_rate": 6.949828761051208e-05, "loss": 0.0088, "step": 32310 }, { "grad_norm": 0.11742933094501495, "learning_rate": 6.947925390074622e-05, "loss": 0.0066, "step": 32320 }, { "grad_norm": 0.0956679955124855, "learning_rate": 6.946021686250863e-05, "loss": 0.0084, "step": 32330 }, { "grad_norm": 0.1350691318511963, "learning_rate": 6.94411764990522e-05, "loss": 0.008, "step": 32340 }, { "grad_norm": 0.07240062952041626, "learning_rate": 6.942213281363044e-05, "loss": 0.0084, "step": 32350 }, { "grad_norm": 0.10744220018386841, "learning_rate": 6.940308580949737e-05, "loss": 0.0087, "step": 32360 }, { "grad_norm": 0.1774333417415619, "learning_rate": 6.93840354899076e-05, "loss": 0.0082, "step": 32370 }, { "grad_norm": 0.14605821669101715, "learning_rate": 6.936498185811633e-05, "loss": 0.0076, "step": 32380 }, { "grad_norm": 0.19644032418727875, "learning_rate": 6.93459249173793e-05, "loss": 0.0105, "step": 32390 }, { "grad_norm": 0.21440988779067993, "learning_rate": 6.932686467095279e-05, "loss": 0.0101, "step": 32400 }, { "grad_norm": 0.15256451070308685, "learning_rate": 6.930780112209373e-05, "loss": 0.0109, "step": 32410 }, { "grad_norm": 0.1319105625152588, "learning_rate": 6.92887342740595e-05, "loss": 0.0086, "step": 32420 }, { "grad_norm": 0.161131352186203, "learning_rate": 6.926966413010816e-05, "loss": 0.0086, "step": 32430 }, { "grad_norm": 0.1231161504983902, "learning_rate": 6.925059069349824e-05, "loss": 0.0072, "step": 32440 }, { "grad_norm": 0.2245924174785614, "learning_rate": 6.923151396748886e-05, "loss": 0.0079, "step": 32450 }, { "grad_norm": 0.1680801659822464, "learning_rate": 6.921243395533974e-05, "loss": 0.0094, "step": 32460 }, { "grad_norm": 0.18674038350582123, "learning_rate": 6.919335066031109e-05, "loss": 0.0097, "step": 32470 }, { "grad_norm": 0.14734460413455963, "learning_rate": 6.917426408566379e-05, "loss": 0.0096, "step": 32480 }, { "grad_norm": 0.19291238486766815, "learning_rate": 6.915517423465916e-05, "loss": 0.008, "step": 32490 }, { "grad_norm": 0.10975504666566849, "learning_rate": 6.913608111055914e-05, "loss": 0.0066, "step": 32500 }, { "grad_norm": 0.11944965273141861, "learning_rate": 6.911698471662623e-05, "loss": 0.0076, "step": 32510 }, { "grad_norm": 0.18459999561309814, "learning_rate": 6.90978850561235e-05, "loss": 0.0084, "step": 32520 }, { "grad_norm": 0.1360398828983307, "learning_rate": 6.907878213231454e-05, "loss": 0.0084, "step": 32530 }, { "grad_norm": 0.20554789900779724, "learning_rate": 6.90596759484635e-05, "loss": 0.0137, "step": 32540 }, { "grad_norm": 0.14989115297794342, "learning_rate": 6.904056650783514e-05, "loss": 0.0086, "step": 32550 }, { "grad_norm": 0.14685577154159546, "learning_rate": 6.902145381369471e-05, "loss": 0.0076, "step": 32560 }, { "grad_norm": 0.14059598743915558, "learning_rate": 6.900233786930808e-05, "loss": 0.0087, "step": 32570 }, { "grad_norm": 0.15717975795269012, "learning_rate": 6.898321867794161e-05, "loss": 0.0108, "step": 32580 }, { "grad_norm": 0.1550162136554718, "learning_rate": 6.896409624286226e-05, "loss": 0.0128, "step": 32590 }, { "grad_norm": 0.12110990285873413, "learning_rate": 6.894497056733754e-05, "loss": 0.0117, "step": 32600 }, { "grad_norm": 0.16021274030208588, "learning_rate": 6.89258416546355e-05, "loss": 0.0074, "step": 32610 }, { "grad_norm": 0.13303402066230774, "learning_rate": 6.890670950802474e-05, "loss": 0.0063, "step": 32620 }, { "grad_norm": 0.14643645286560059, "learning_rate": 6.88875741307744e-05, "loss": 0.0093, "step": 32630 }, { "grad_norm": 0.1446468085050583, "learning_rate": 6.886843552615425e-05, "loss": 0.0073, "step": 32640 }, { "grad_norm": 0.1598459929227829, "learning_rate": 6.884929369743451e-05, "loss": 0.0063, "step": 32650 }, { "grad_norm": 0.1703338474035263, "learning_rate": 6.8830148647886e-05, "loss": 0.0088, "step": 32660 }, { "grad_norm": 0.12738193571567535, "learning_rate": 6.88110003807801e-05, "loss": 0.0083, "step": 32670 }, { "grad_norm": 0.11869776248931885, "learning_rate": 6.87918488993887e-05, "loss": 0.0102, "step": 32680 }, { "grad_norm": 0.11334115266799927, "learning_rate": 6.877269420698431e-05, "loss": 0.0077, "step": 32690 }, { "grad_norm": 0.1331024020910263, "learning_rate": 6.875353630683989e-05, "loss": 0.0091, "step": 32700 }, { "grad_norm": 0.13480211794376373, "learning_rate": 6.873437520222905e-05, "loss": 0.0065, "step": 32710 }, { "grad_norm": 0.140361025929451, "learning_rate": 6.871521089642585e-05, "loss": 0.008, "step": 32720 }, { "grad_norm": 0.20002250373363495, "learning_rate": 6.869604339270498e-05, "loss": 0.0097, "step": 32730 }, { "grad_norm": 0.11561625450849533, "learning_rate": 6.867687269434164e-05, "loss": 0.0065, "step": 32740 }, { "grad_norm": 0.17131252586841583, "learning_rate": 6.865769880461156e-05, "loss": 0.0091, "step": 32750 }, { "grad_norm": 0.14304794371128082, "learning_rate": 6.863852172679104e-05, "loss": 0.0079, "step": 32760 }, { "grad_norm": 0.12803415954113007, "learning_rate": 6.861934146415693e-05, "loss": 0.009, "step": 32770 }, { "grad_norm": 0.2068672776222229, "learning_rate": 6.86001580199866e-05, "loss": 0.0079, "step": 32780 }, { "grad_norm": 0.17090588808059692, "learning_rate": 6.858097139755798e-05, "loss": 0.0077, "step": 32790 }, { "grad_norm": 0.12518583238124847, "learning_rate": 6.856178160014955e-05, "loss": 0.0075, "step": 32800 }, { "grad_norm": 0.15149961411952972, "learning_rate": 6.85425886310403e-05, "loss": 0.0083, "step": 32810 }, { "grad_norm": 0.12272909283638, "learning_rate": 6.852339249350979e-05, "loss": 0.0074, "step": 32820 }, { "grad_norm": 0.19023331999778748, "learning_rate": 6.850419319083812e-05, "loss": 0.0114, "step": 32830 }, { "grad_norm": 0.16820324957370758, "learning_rate": 6.848499072630592e-05, "loss": 0.0081, "step": 32840 }, { "grad_norm": 0.0953526720404625, "learning_rate": 6.846578510319439e-05, "loss": 0.0069, "step": 32850 }, { "grad_norm": 0.1326930820941925, "learning_rate": 6.844657632478519e-05, "loss": 0.0075, "step": 32860 }, { "grad_norm": 0.10172086209058762, "learning_rate": 6.842736439436063e-05, "loss": 0.0071, "step": 32870 }, { "grad_norm": 0.10170914977788925, "learning_rate": 6.84081493152035e-05, "loss": 0.0098, "step": 32880 }, { "grad_norm": 0.08228331804275513, "learning_rate": 6.83889310905971e-05, "loss": 0.0064, "step": 32890 }, { "grad_norm": 0.10679800808429718, "learning_rate": 6.836970972382533e-05, "loss": 0.0082, "step": 32900 }, { "grad_norm": 0.12591010332107544, "learning_rate": 6.835048521817257e-05, "loss": 0.0061, "step": 32910 }, { "grad_norm": 0.13396301865577698, "learning_rate": 6.833125757692379e-05, "loss": 0.0081, "step": 32920 }, { "grad_norm": 0.17452271282672882, "learning_rate": 6.831202680336441e-05, "loss": 0.0081, "step": 32930 }, { "grad_norm": 0.14542323350906372, "learning_rate": 6.829279290078052e-05, "loss": 0.0102, "step": 32940 }, { "grad_norm": 0.11466647684574127, "learning_rate": 6.827355587245863e-05, "loss": 0.0089, "step": 32950 }, { "grad_norm": 0.10411026328802109, "learning_rate": 6.82543157216858e-05, "loss": 0.0075, "step": 32960 }, { "grad_norm": 0.14680570363998413, "learning_rate": 6.823507245174969e-05, "loss": 0.0072, "step": 32970 }, { "grad_norm": 0.1382770985364914, "learning_rate": 6.821582606593841e-05, "loss": 0.0082, "step": 32980 }, { "grad_norm": 0.10834760218858719, "learning_rate": 6.81965765675407e-05, "loss": 0.009, "step": 32990 }, { "grad_norm": 0.1219145730137825, "learning_rate": 6.81773239598457e-05, "loss": 0.0086, "step": 33000 }, { "grad_norm": 0.16610310971736908, "learning_rate": 6.815806824614319e-05, "loss": 0.0071, "step": 33010 }, { "grad_norm": 0.13592968881130219, "learning_rate": 6.813880942972343e-05, "loss": 0.0086, "step": 33020 }, { "grad_norm": 0.10922008752822876, "learning_rate": 6.811954751387726e-05, "loss": 0.0104, "step": 33030 }, { "grad_norm": 0.11993951350450516, "learning_rate": 6.810028250189598e-05, "loss": 0.0094, "step": 33040 }, { "grad_norm": 0.16851741075515747, "learning_rate": 6.808101439707147e-05, "loss": 0.0089, "step": 33050 }, { "grad_norm": 0.13773298263549805, "learning_rate": 6.806174320269609e-05, "loss": 0.0093, "step": 33060 }, { "grad_norm": 0.20038247108459473, "learning_rate": 6.804246892206281e-05, "loss": 0.008, "step": 33070 }, { "grad_norm": 0.15292686223983765, "learning_rate": 6.802319155846506e-05, "loss": 0.0113, "step": 33080 }, { "grad_norm": 0.22102907299995422, "learning_rate": 6.800391111519679e-05, "loss": 0.0106, "step": 33090 }, { "grad_norm": 0.15599924325942993, "learning_rate": 6.798462759555253e-05, "loss": 0.0087, "step": 33100 }, { "grad_norm": 0.11608188599348068, "learning_rate": 6.79653410028273e-05, "loss": 0.0088, "step": 33110 }, { "grad_norm": 0.1501965969800949, "learning_rate": 6.794605134031663e-05, "loss": 0.0082, "step": 33120 }, { "grad_norm": 0.14032134413719177, "learning_rate": 6.792675861131661e-05, "loss": 0.0086, "step": 33130 }, { "grad_norm": 0.17009703814983368, "learning_rate": 6.790746281912386e-05, "loss": 0.0123, "step": 33140 }, { "grad_norm": 0.11932583153247833, "learning_rate": 6.788816396703546e-05, "loss": 0.0094, "step": 33150 }, { "grad_norm": 0.1173878014087677, "learning_rate": 6.78688620583491e-05, "loss": 0.0085, "step": 33160 }, { "grad_norm": 0.14663872122764587, "learning_rate": 6.784955709636292e-05, "loss": 0.0099, "step": 33170 }, { "grad_norm": 0.14689749479293823, "learning_rate": 6.783024908437564e-05, "loss": 0.0098, "step": 33180 }, { "grad_norm": 0.1899154633283615, "learning_rate": 6.781093802568641e-05, "loss": 0.0073, "step": 33190 }, { "grad_norm": 0.1509283185005188, "learning_rate": 6.779162392359504e-05, "loss": 0.0088, "step": 33200 }, { "grad_norm": 0.14020274579524994, "learning_rate": 6.777230678140172e-05, "loss": 0.0112, "step": 33210 }, { "grad_norm": 0.1235385313630104, "learning_rate": 6.775298660240726e-05, "loss": 0.0088, "step": 33220 }, { "grad_norm": 0.14761707186698914, "learning_rate": 6.773366338991292e-05, "loss": 0.0091, "step": 33230 }, { "grad_norm": 0.13415832817554474, "learning_rate": 6.771433714722052e-05, "loss": 0.0079, "step": 33240 }, { "grad_norm": 0.09840031713247299, "learning_rate": 6.769500787763239e-05, "loss": 0.0078, "step": 33250 }, { "grad_norm": 0.1650964915752411, "learning_rate": 6.76756755844514e-05, "loss": 0.0113, "step": 33260 }, { "grad_norm": 0.11704685539007187, "learning_rate": 6.765634027098087e-05, "loss": 0.0084, "step": 33270 }, { "grad_norm": 0.11310939490795135, "learning_rate": 6.763700194052468e-05, "loss": 0.0087, "step": 33280 }, { "grad_norm": 0.10786750167608261, "learning_rate": 6.761766059638723e-05, "loss": 0.0084, "step": 33290 }, { "grad_norm": 0.12362904101610184, "learning_rate": 6.759831624187345e-05, "loss": 0.0074, "step": 33300 }, { "grad_norm": 0.13154150545597076, "learning_rate": 6.757896888028871e-05, "loss": 0.0088, "step": 33310 }, { "grad_norm": 0.14739373326301575, "learning_rate": 6.7559618514939e-05, "loss": 0.0067, "step": 33320 }, { "grad_norm": 0.19947510957717896, "learning_rate": 6.754026514913073e-05, "loss": 0.0111, "step": 33330 }, { "grad_norm": 0.13052494823932648, "learning_rate": 6.752090878617087e-05, "loss": 0.0081, "step": 33340 }, { "grad_norm": 0.17678038775920868, "learning_rate": 6.75015494293669e-05, "loss": 0.0075, "step": 33350 }, { "grad_norm": 0.2011823058128357, "learning_rate": 6.74821870820268e-05, "loss": 0.008, "step": 33360 }, { "grad_norm": 0.12389396131038666, "learning_rate": 6.746282174745907e-05, "loss": 0.009, "step": 33370 }, { "grad_norm": 0.13580425083637238, "learning_rate": 6.744345342897271e-05, "loss": 0.0092, "step": 33380 }, { "grad_norm": 0.12768591940402985, "learning_rate": 6.742408212987724e-05, "loss": 0.0093, "step": 33390 }, { "grad_norm": 0.10393647104501724, "learning_rate": 6.740470785348269e-05, "loss": 0.0084, "step": 33400 }, { "grad_norm": 0.19341689348220825, "learning_rate": 6.738533060309958e-05, "loss": 0.0094, "step": 33410 }, { "grad_norm": 0.154168501496315, "learning_rate": 6.736595038203894e-05, "loss": 0.0071, "step": 33420 }, { "grad_norm": 0.1339930146932602, "learning_rate": 6.734656719361236e-05, "loss": 0.0074, "step": 33430 }, { "grad_norm": 0.19339291751384735, "learning_rate": 6.732718104113189e-05, "loss": 0.0125, "step": 33440 }, { "grad_norm": 0.11494791507720947, "learning_rate": 6.730779192791006e-05, "loss": 0.0069, "step": 33450 }, { "grad_norm": 0.0926760882139206, "learning_rate": 6.728839985725997e-05, "loss": 0.0059, "step": 33460 }, { "grad_norm": 0.16152919828891754, "learning_rate": 6.726900483249517e-05, "loss": 0.0094, "step": 33470 }, { "grad_norm": 0.1551598459482193, "learning_rate": 6.724960685692976e-05, "loss": 0.0109, "step": 33480 }, { "grad_norm": 0.11676707118749619, "learning_rate": 6.723020593387833e-05, "loss": 0.0082, "step": 33490 }, { "grad_norm": 0.08458415418863297, "learning_rate": 6.721080206665593e-05, "loss": 0.0076, "step": 33500 }, { "grad_norm": 0.16724786162376404, "learning_rate": 6.719139525857819e-05, "loss": 0.0116, "step": 33510 }, { "grad_norm": 0.14397817850112915, "learning_rate": 6.717198551296117e-05, "loss": 0.0067, "step": 33520 }, { "grad_norm": 0.18888762593269348, "learning_rate": 6.715257283312148e-05, "loss": 0.0099, "step": 33530 }, { "grad_norm": 0.12088195234537125, "learning_rate": 6.713315722237623e-05, "loss": 0.0071, "step": 33540 }, { "grad_norm": 0.09778212755918503, "learning_rate": 6.7113738684043e-05, "loss": 0.0078, "step": 33550 }, { "grad_norm": 0.12330593913793564, "learning_rate": 6.709431722143989e-05, "loss": 0.0077, "step": 33560 }, { "grad_norm": 0.14763663709163666, "learning_rate": 6.70748928378855e-05, "loss": 0.0074, "step": 33570 }, { "grad_norm": 0.15277694165706635, "learning_rate": 6.705546553669891e-05, "loss": 0.0083, "step": 33580 }, { "grad_norm": 0.1538081020116806, "learning_rate": 6.703603532119974e-05, "loss": 0.0083, "step": 33590 }, { "grad_norm": 0.11588240414857864, "learning_rate": 6.701660219470808e-05, "loss": 0.0069, "step": 33600 }, { "grad_norm": 0.13287188112735748, "learning_rate": 6.69971661605445e-05, "loss": 0.0089, "step": 33610 }, { "grad_norm": 0.10464531183242798, "learning_rate": 6.697772722203008e-05, "loss": 0.0076, "step": 33620 }, { "grad_norm": 0.1266394406557083, "learning_rate": 6.695828538248643e-05, "loss": 0.0086, "step": 33630 }, { "grad_norm": 0.12615756690502167, "learning_rate": 6.693884064523563e-05, "loss": 0.0071, "step": 33640 }, { "grad_norm": 0.1631901115179062, "learning_rate": 6.691939301360023e-05, "loss": 0.0092, "step": 33650 }, { "grad_norm": 0.12408995628356934, "learning_rate": 6.689994249090333e-05, "loss": 0.0079, "step": 33660 }, { "grad_norm": 0.1639450043439865, "learning_rate": 6.688048908046845e-05, "loss": 0.0074, "step": 33670 }, { "grad_norm": 0.12207282334566116, "learning_rate": 6.686103278561969e-05, "loss": 0.0073, "step": 33680 }, { "grad_norm": 0.13263297080993652, "learning_rate": 6.684157360968156e-05, "loss": 0.0097, "step": 33690 }, { "grad_norm": 0.12755775451660156, "learning_rate": 6.682211155597911e-05, "loss": 0.0064, "step": 33700 }, { "grad_norm": 0.12503470480442047, "learning_rate": 6.680264662783789e-05, "loss": 0.0084, "step": 33710 }, { "grad_norm": 0.11157228797674179, "learning_rate": 6.678317882858391e-05, "loss": 0.008, "step": 33720 }, { "grad_norm": 0.14319902658462524, "learning_rate": 6.67637081615437e-05, "loss": 0.0073, "step": 33730 }, { "grad_norm": 0.1059616431593895, "learning_rate": 6.674423463004427e-05, "loss": 0.006, "step": 33740 }, { "grad_norm": 0.09393222630023956, "learning_rate": 6.672475823741308e-05, "loss": 0.009, "step": 33750 }, { "grad_norm": 0.07562635838985443, "learning_rate": 6.670527898697811e-05, "loss": 0.0074, "step": 33760 }, { "grad_norm": 0.12565308809280396, "learning_rate": 6.668579688206788e-05, "loss": 0.0082, "step": 33770 }, { "grad_norm": 0.18099023401737213, "learning_rate": 6.666631192601131e-05, "loss": 0.0088, "step": 33780 }, { "grad_norm": 0.13143399357795715, "learning_rate": 6.664682412213785e-05, "loss": 0.0086, "step": 33790 }, { "grad_norm": 0.11795210838317871, "learning_rate": 6.662733347377745e-05, "loss": 0.0094, "step": 33800 }, { "grad_norm": 0.1214912086725235, "learning_rate": 6.660783998426051e-05, "loss": 0.008, "step": 33810 }, { "grad_norm": 0.1258729249238968, "learning_rate": 6.658834365691794e-05, "loss": 0.0062, "step": 33820 }, { "grad_norm": 0.11914307624101639, "learning_rate": 6.656884449508115e-05, "loss": 0.0088, "step": 33830 }, { "grad_norm": 0.1089019626379013, "learning_rate": 6.654934250208198e-05, "loss": 0.0069, "step": 33840 }, { "grad_norm": 0.14513400197029114, "learning_rate": 6.65298376812528e-05, "loss": 0.0082, "step": 33850 }, { "grad_norm": 0.12203316390514374, "learning_rate": 6.651033003592646e-05, "loss": 0.0063, "step": 33860 }, { "grad_norm": 0.1188744381070137, "learning_rate": 6.649081956943626e-05, "loss": 0.007, "step": 33870 }, { "grad_norm": 0.1343322992324829, "learning_rate": 6.647130628511604e-05, "loss": 0.0085, "step": 33880 }, { "grad_norm": 0.19197200238704681, "learning_rate": 6.645179018630005e-05, "loss": 0.009, "step": 33890 }, { "grad_norm": 0.15222088992595673, "learning_rate": 6.643227127632309e-05, "loss": 0.0096, "step": 33900 }, { "grad_norm": 0.09353234618902206, "learning_rate": 6.641274955852038e-05, "loss": 0.0057, "step": 33910 }, { "grad_norm": 0.12669892609119415, "learning_rate": 6.639322503622768e-05, "loss": 0.0075, "step": 33920 }, { "grad_norm": 0.16405576467514038, "learning_rate": 6.637369771278116e-05, "loss": 0.0117, "step": 33930 }, { "grad_norm": 0.11495303362607956, "learning_rate": 6.635416759151751e-05, "loss": 0.008, "step": 33940 }, { "grad_norm": 0.10970637202262878, "learning_rate": 6.633463467577394e-05, "loss": 0.0068, "step": 33950 }, { "grad_norm": 0.1395941972732544, "learning_rate": 6.631509896888803e-05, "loss": 0.0087, "step": 33960 }, { "grad_norm": 0.1702391505241394, "learning_rate": 6.629556047419794e-05, "loss": 0.0094, "step": 33970 }, { "grad_norm": 0.13460762798786163, "learning_rate": 6.627601919504223e-05, "loss": 0.0103, "step": 33980 }, { "grad_norm": 0.15455617010593414, "learning_rate": 6.625647513476001e-05, "loss": 0.0089, "step": 33990 }, { "grad_norm": 0.10049626231193542, "learning_rate": 6.62369282966908e-05, "loss": 0.0097, "step": 34000 }, { "grad_norm": 0.12410616874694824, "learning_rate": 6.621737868417464e-05, "loss": 0.0069, "step": 34010 }, { "grad_norm": 0.11054138839244843, "learning_rate": 6.619782630055198e-05, "loss": 0.0079, "step": 34020 }, { "grad_norm": 0.12894786894321442, "learning_rate": 6.617827114916382e-05, "loss": 0.0104, "step": 34030 }, { "grad_norm": 0.07669106870889664, "learning_rate": 6.615871323335161e-05, "loss": 0.0075, "step": 34040 }, { "grad_norm": 0.130892813205719, "learning_rate": 6.613915255645725e-05, "loss": 0.0076, "step": 34050 }, { "grad_norm": 0.15470564365386963, "learning_rate": 6.611958912182312e-05, "loss": 0.0103, "step": 34060 }, { "grad_norm": 0.17576290667057037, "learning_rate": 6.610002293279207e-05, "loss": 0.0086, "step": 34070 }, { "grad_norm": 0.1503688097000122, "learning_rate": 6.608045399270746e-05, "loss": 0.0082, "step": 34080 }, { "grad_norm": 0.1409715861082077, "learning_rate": 6.606088230491304e-05, "loss": 0.0089, "step": 34090 }, { "grad_norm": 0.10758540779352188, "learning_rate": 6.604130787275312e-05, "loss": 0.0074, "step": 34100 }, { "grad_norm": 0.13329842686653137, "learning_rate": 6.602173069957242e-05, "loss": 0.0103, "step": 34110 }, { "grad_norm": 0.1142427995800972, "learning_rate": 6.600215078871612e-05, "loss": 0.0094, "step": 34120 }, { "grad_norm": 0.17950217425823212, "learning_rate": 6.598256814352992e-05, "loss": 0.0102, "step": 34130 }, { "grad_norm": 0.11051057279109955, "learning_rate": 6.596298276735995e-05, "loss": 0.0068, "step": 34140 }, { "grad_norm": 0.11655031889677048, "learning_rate": 6.594339466355282e-05, "loss": 0.0084, "step": 34150 }, { "grad_norm": 0.1346159130334854, "learning_rate": 6.592380383545558e-05, "loss": 0.0075, "step": 34160 }, { "grad_norm": 0.1359252631664276, "learning_rate": 6.590421028641577e-05, "loss": 0.0081, "step": 34170 }, { "grad_norm": 0.12661495804786682, "learning_rate": 6.588461401978143e-05, "loss": 0.0092, "step": 34180 }, { "grad_norm": 0.18413595855236053, "learning_rate": 6.586501503890099e-05, "loss": 0.0083, "step": 34190 }, { "grad_norm": 0.16191889345645905, "learning_rate": 6.584541334712338e-05, "loss": 0.0072, "step": 34200 }, { "grad_norm": 0.10708226263523102, "learning_rate": 6.582580894779802e-05, "loss": 0.0066, "step": 34210 }, { "grad_norm": 0.12560313940048218, "learning_rate": 6.580620184427473e-05, "loss": 0.0072, "step": 34220 }, { "grad_norm": 0.13632453978061676, "learning_rate": 6.578659203990385e-05, "loss": 0.0078, "step": 34230 }, { "grad_norm": 0.1303594410419464, "learning_rate": 6.576697953803615e-05, "loss": 0.0088, "step": 34240 }, { "grad_norm": 0.10883106291294098, "learning_rate": 6.57473643420229e-05, "loss": 0.01, "step": 34250 }, { "grad_norm": 0.16012032330036163, "learning_rate": 6.572774645521574e-05, "loss": 0.0118, "step": 34260 }, { "grad_norm": 0.23579327762126923, "learning_rate": 6.570812588096688e-05, "loss": 0.0195, "step": 34270 }, { "grad_norm": 0.17480404675006866, "learning_rate": 6.568850262262893e-05, "loss": 0.0164, "step": 34280 }, { "grad_norm": 0.16559819877147675, "learning_rate": 6.566887668355497e-05, "loss": 0.0138, "step": 34290 }, { "grad_norm": 0.17291830480098724, "learning_rate": 6.564924806709851e-05, "loss": 0.01, "step": 34300 }, { "grad_norm": 0.16073286533355713, "learning_rate": 6.562961677661359e-05, "loss": 0.0084, "step": 34310 }, { "grad_norm": 0.15448562800884247, "learning_rate": 6.56099828154546e-05, "loss": 0.0114, "step": 34320 }, { "grad_norm": 0.09027348458766937, "learning_rate": 6.55903461869765e-05, "loss": 0.0089, "step": 34330 }, { "grad_norm": 0.09428020566701889, "learning_rate": 6.557070689453465e-05, "loss": 0.0083, "step": 34340 }, { "grad_norm": 0.09834796190261841, "learning_rate": 6.555106494148482e-05, "loss": 0.0096, "step": 34350 }, { "grad_norm": 0.12319780141115189, "learning_rate": 6.553142033118333e-05, "loss": 0.0103, "step": 34360 }, { "grad_norm": 0.1423141062259674, "learning_rate": 6.551177306698688e-05, "loss": 0.0079, "step": 34370 }, { "grad_norm": 0.11658324301242828, "learning_rate": 6.549212315225267e-05, "loss": 0.0086, "step": 34380 }, { "grad_norm": 0.27242159843444824, "learning_rate": 6.547247059033833e-05, "loss": 0.0129, "step": 34390 }, { "grad_norm": 0.15961402654647827, "learning_rate": 6.545281538460193e-05, "loss": 0.0076, "step": 34400 }, { "grad_norm": 0.12970970571041107, "learning_rate": 6.543315753840202e-05, "loss": 0.0074, "step": 34410 }, { "grad_norm": 0.15194140374660492, "learning_rate": 6.541349705509758e-05, "loss": 0.0071, "step": 34420 }, { "grad_norm": 0.13511380553245544, "learning_rate": 6.539383393804805e-05, "loss": 0.008, "step": 34430 }, { "grad_norm": 0.12726247310638428, "learning_rate": 6.537416819061333e-05, "loss": 0.0066, "step": 34440 }, { "grad_norm": 0.20021207630634308, "learning_rate": 6.535449981615375e-05, "loss": 0.0092, "step": 34450 }, { "grad_norm": 0.12689882516860962, "learning_rate": 6.53348288180301e-05, "loss": 0.0077, "step": 34460 }, { "grad_norm": 0.1597723364830017, "learning_rate": 6.531515519960361e-05, "loss": 0.0079, "step": 34470 }, { "grad_norm": 0.1258612424135208, "learning_rate": 6.529547896423597e-05, "loss": 0.0088, "step": 34480 }, { "grad_norm": 0.1409263163805008, "learning_rate": 6.52758001152893e-05, "loss": 0.0079, "step": 34490 }, { "grad_norm": 0.10763351619243622, "learning_rate": 6.525611865612618e-05, "loss": 0.0077, "step": 34500 }, { "grad_norm": 0.13143858313560486, "learning_rate": 6.523643459010966e-05, "loss": 0.0086, "step": 34510 }, { "grad_norm": 0.15674860775470734, "learning_rate": 6.521674792060317e-05, "loss": 0.0089, "step": 34520 }, { "grad_norm": 0.12402408570051193, "learning_rate": 6.519705865097063e-05, "loss": 0.0084, "step": 34530 }, { "grad_norm": 0.14260783791542053, "learning_rate": 6.517736678457641e-05, "loss": 0.0077, "step": 34540 }, { "grad_norm": 0.1154111847281456, "learning_rate": 6.515767232478534e-05, "loss": 0.0075, "step": 34550 }, { "grad_norm": 0.1135363057255745, "learning_rate": 6.51379752749626e-05, "loss": 0.0077, "step": 34560 }, { "grad_norm": 0.14258970320224762, "learning_rate": 6.511827563847393e-05, "loss": 0.0077, "step": 34570 }, { "grad_norm": 0.09845367074012756, "learning_rate": 6.509857341868542e-05, "loss": 0.009, "step": 34580 }, { "grad_norm": 0.1448487639427185, "learning_rate": 6.507886861896367e-05, "loss": 0.0072, "step": 34590 }, { "grad_norm": 0.18339650332927704, "learning_rate": 6.505916124267567e-05, "loss": 0.007, "step": 34600 }, { "grad_norm": 0.13172481954097748, "learning_rate": 6.503945129318891e-05, "loss": 0.008, "step": 34610 }, { "grad_norm": 0.19343580305576324, "learning_rate": 6.501973877387122e-05, "loss": 0.0088, "step": 34620 }, { "grad_norm": 0.111544668674469, "learning_rate": 6.500002368809098e-05, "loss": 0.0096, "step": 34630 }, { "grad_norm": 0.11975830048322678, "learning_rate": 6.498030603921694e-05, "loss": 0.0094, "step": 34640 }, { "grad_norm": 0.1800435334444046, "learning_rate": 6.496058583061832e-05, "loss": 0.0114, "step": 34650 }, { "grad_norm": 0.11055704206228256, "learning_rate": 6.494086306566475e-05, "loss": 0.0067, "step": 34660 }, { "grad_norm": 0.1530979722738266, "learning_rate": 6.492113774772632e-05, "loss": 0.0074, "step": 34670 }, { "grad_norm": 0.11414182186126709, "learning_rate": 6.490140988017354e-05, "loss": 0.0094, "step": 34680 }, { "grad_norm": 0.11158569902181625, "learning_rate": 6.488167946637736e-05, "loss": 0.0077, "step": 34690 }, { "grad_norm": 0.1600719392299652, "learning_rate": 6.486194650970915e-05, "loss": 0.0099, "step": 34700 }, { "grad_norm": 0.12302014231681824, "learning_rate": 6.48422110135408e-05, "loss": 0.0078, "step": 34710 }, { "grad_norm": 0.17525628209114075, "learning_rate": 6.482247298124451e-05, "loss": 0.0083, "step": 34720 }, { "grad_norm": 0.20036344230175018, "learning_rate": 6.480273241619297e-05, "loss": 0.0077, "step": 34730 }, { "grad_norm": 0.10894613713026047, "learning_rate": 6.478298932175933e-05, "loss": 0.0087, "step": 34740 }, { "grad_norm": 0.10414829105138779, "learning_rate": 6.476324370131712e-05, "loss": 0.0082, "step": 34750 }, { "grad_norm": 0.14105546474456787, "learning_rate": 6.474349555824036e-05, "loss": 0.0069, "step": 34760 }, { "grad_norm": 0.14050520956516266, "learning_rate": 6.472374489590342e-05, "loss": 0.0076, "step": 34770 }, { "grad_norm": 0.11526966094970703, "learning_rate": 6.470399171768118e-05, "loss": 0.0064, "step": 34780 }, { "grad_norm": 0.1355915665626526, "learning_rate": 6.468423602694891e-05, "loss": 0.0071, "step": 34790 }, { "grad_norm": 0.17989778518676758, "learning_rate": 6.466447782708232e-05, "loss": 0.007, "step": 34800 }, { "grad_norm": 0.18212249875068665, "learning_rate": 6.464471712145754e-05, "loss": 0.0095, "step": 34810 }, { "grad_norm": 0.1380433589220047, "learning_rate": 6.462495391345114e-05, "loss": 0.0073, "step": 34820 }, { "grad_norm": 0.10168581455945969, "learning_rate": 6.46051882064401e-05, "loss": 0.0097, "step": 34830 }, { "grad_norm": 0.13514554500579834, "learning_rate": 6.458542000380186e-05, "loss": 0.0071, "step": 34840 }, { "grad_norm": 0.11787349730730057, "learning_rate": 6.456564930891424e-05, "loss": 0.0063, "step": 34850 }, { "grad_norm": 0.13857290148735046, "learning_rate": 6.454587612515555e-05, "loss": 0.0064, "step": 34860 }, { "grad_norm": 0.12243583798408508, "learning_rate": 6.452610045590444e-05, "loss": 0.01, "step": 34870 }, { "grad_norm": 0.10792921483516693, "learning_rate": 6.450632230454005e-05, "loss": 0.0071, "step": 34880 }, { "grad_norm": 0.1346009224653244, "learning_rate": 6.448654167444195e-05, "loss": 0.0091, "step": 34890 }, { "grad_norm": 0.1370830237865448, "learning_rate": 6.446675856899005e-05, "loss": 0.0066, "step": 34900 }, { "grad_norm": 0.11229541897773743, "learning_rate": 6.444697299156481e-05, "loss": 0.0069, "step": 34910 }, { "grad_norm": 0.1493556797504425, "learning_rate": 6.442718494554701e-05, "loss": 0.0071, "step": 34920 }, { "grad_norm": 0.1409921497106552, "learning_rate": 6.440739443431787e-05, "loss": 0.0075, "step": 34930 }, { "grad_norm": 0.1251692920923233, "learning_rate": 6.438760146125906e-05, "loss": 0.0088, "step": 34940 }, { "grad_norm": 0.1332203447818756, "learning_rate": 6.436780602975267e-05, "loss": 0.0073, "step": 34950 }, { "grad_norm": 0.14162562787532806, "learning_rate": 6.43480081431812e-05, "loss": 0.0076, "step": 34960 }, { "grad_norm": 0.09807407855987549, "learning_rate": 6.432820780492756e-05, "loss": 0.0073, "step": 34970 }, { "grad_norm": 0.1224493756890297, "learning_rate": 6.430840501837506e-05, "loss": 0.0102, "step": 34980 }, { "grad_norm": 0.15207824110984802, "learning_rate": 6.428859978690748e-05, "loss": 0.0083, "step": 34990 }, { "grad_norm": 0.13872461020946503, "learning_rate": 6.426879211390901e-05, "loss": 0.0079, "step": 35000 }, { "grad_norm": 0.18197603523731232, "learning_rate": 6.424898200276422e-05, "loss": 0.0123, "step": 35010 }, { "grad_norm": 0.11821135133504868, "learning_rate": 6.42291694568581e-05, "loss": 0.0081, "step": 35020 }, { "grad_norm": 0.14522455632686615, "learning_rate": 6.42093544795761e-05, "loss": 0.0096, "step": 35030 }, { "grad_norm": 0.13562551140785217, "learning_rate": 6.418953707430403e-05, "loss": 0.0083, "step": 35040 }, { "grad_norm": 0.14672738313674927, "learning_rate": 6.416971724442819e-05, "loss": 0.0073, "step": 35050 }, { "grad_norm": 0.13162986934185028, "learning_rate": 6.414989499333519e-05, "loss": 0.0067, "step": 35060 }, { "grad_norm": 0.1522529274225235, "learning_rate": 6.413007032441214e-05, "loss": 0.0088, "step": 35070 }, { "grad_norm": 0.1317203938961029, "learning_rate": 6.411024324104653e-05, "loss": 0.0127, "step": 35080 }, { "grad_norm": 0.15635725855827332, "learning_rate": 6.409041374662628e-05, "loss": 0.0088, "step": 35090 }, { "grad_norm": 0.12998318672180176, "learning_rate": 6.407058184453967e-05, "loss": 0.0079, "step": 35100 }, { "grad_norm": 0.12526699900627136, "learning_rate": 6.405074753817548e-05, "loss": 0.0108, "step": 35110 }, { "grad_norm": 0.13709823787212372, "learning_rate": 6.40309108309228e-05, "loss": 0.0084, "step": 35120 }, { "grad_norm": 0.13476768136024475, "learning_rate": 6.401107172617122e-05, "loss": 0.0073, "step": 35130 }, { "grad_norm": 0.17803218960762024, "learning_rate": 6.399123022731068e-05, "loss": 0.0065, "step": 35140 }, { "grad_norm": 0.09720846265554428, "learning_rate": 6.397138633773157e-05, "loss": 0.0066, "step": 35150 }, { "grad_norm": 0.15216463804244995, "learning_rate": 6.395154006082463e-05, "loss": 0.007, "step": 35160 }, { "grad_norm": 0.12791191041469574, "learning_rate": 6.393169139998109e-05, "loss": 0.0069, "step": 35170 }, { "grad_norm": 0.17580130696296692, "learning_rate": 6.39118403585925e-05, "loss": 0.0098, "step": 35180 }, { "grad_norm": 0.18983303010463715, "learning_rate": 6.38919869400509e-05, "loss": 0.0097, "step": 35190 }, { "grad_norm": 0.12641766667366028, "learning_rate": 6.387213114774865e-05, "loss": 0.0115, "step": 35200 }, { "grad_norm": 0.1283017098903656, "learning_rate": 6.385227298507863e-05, "loss": 0.0076, "step": 35210 }, { "grad_norm": 0.16578704118728638, "learning_rate": 6.3832412455434e-05, "loss": 0.0068, "step": 35220 }, { "grad_norm": 0.1569463610649109, "learning_rate": 6.381254956220841e-05, "loss": 0.0074, "step": 35230 }, { "grad_norm": 0.13891254365444183, "learning_rate": 6.379268430879586e-05, "loss": 0.008, "step": 35240 }, { "grad_norm": 0.12737861275672913, "learning_rate": 6.37728166985908e-05, "loss": 0.0082, "step": 35250 }, { "grad_norm": 0.11597830802202225, "learning_rate": 6.375294673498804e-05, "loss": 0.0074, "step": 35260 }, { "grad_norm": 0.1288936883211136, "learning_rate": 6.373307442138284e-05, "loss": 0.0073, "step": 35270 }, { "grad_norm": 0.10936114192008972, "learning_rate": 6.371319976117081e-05, "loss": 0.006, "step": 35280 }, { "grad_norm": 0.14796271920204163, "learning_rate": 6.3693322757748e-05, "loss": 0.0079, "step": 35290 }, { "grad_norm": 0.10115615278482437, "learning_rate": 6.367344341451086e-05, "loss": 0.009, "step": 35300 }, { "grad_norm": 0.20717009902000427, "learning_rate": 6.36535617348562e-05, "loss": 0.0084, "step": 35310 }, { "grad_norm": 0.15781031548976898, "learning_rate": 6.363367772218128e-05, "loss": 0.0082, "step": 35320 }, { "grad_norm": 0.1299697756767273, "learning_rate": 6.36137913798837e-05, "loss": 0.0091, "step": 35330 }, { "grad_norm": 0.11986713856458664, "learning_rate": 6.359390271136151e-05, "loss": 0.0083, "step": 35340 }, { "grad_norm": 0.14540724456310272, "learning_rate": 6.357401172001314e-05, "loss": 0.008, "step": 35350 }, { "grad_norm": 0.11903982609510422, "learning_rate": 6.355411840923742e-05, "loss": 0.0065, "step": 35360 }, { "grad_norm": 0.11483602225780487, "learning_rate": 6.353422278243358e-05, "loss": 0.0111, "step": 35370 }, { "grad_norm": 0.11424829065799713, "learning_rate": 6.351432484300121e-05, "loss": 0.006, "step": 35380 }, { "grad_norm": 0.1584799736738205, "learning_rate": 6.349442459434036e-05, "loss": 0.0088, "step": 35390 }, { "grad_norm": 0.22271579504013062, "learning_rate": 6.34745220398514e-05, "loss": 0.0085, "step": 35400 }, { "grad_norm": 0.10319247841835022, "learning_rate": 6.345461718293518e-05, "loss": 0.0072, "step": 35410 }, { "grad_norm": 0.15211515128612518, "learning_rate": 6.343471002699286e-05, "loss": 0.0099, "step": 35420 }, { "grad_norm": 0.09769623726606369, "learning_rate": 6.341480057542602e-05, "loss": 0.0087, "step": 35430 }, { "grad_norm": 0.16141816973686218, "learning_rate": 6.339488883163667e-05, "loss": 0.0114, "step": 35440 }, { "grad_norm": 0.18128006160259247, "learning_rate": 6.337497479902716e-05, "loss": 0.0081, "step": 35450 }, { "grad_norm": 0.10958285629749298, "learning_rate": 6.335505848100027e-05, "loss": 0.0092, "step": 35460 }, { "grad_norm": 0.1486629992723465, "learning_rate": 6.333513988095915e-05, "loss": 0.0095, "step": 35470 }, { "grad_norm": 0.1128963977098465, "learning_rate": 6.331521900230735e-05, "loss": 0.0079, "step": 35480 }, { "grad_norm": 0.1143358126282692, "learning_rate": 6.329529584844878e-05, "loss": 0.0055, "step": 35490 }, { "grad_norm": 0.11177734285593033, "learning_rate": 6.327537042278777e-05, "loss": 0.0101, "step": 35500 }, { "grad_norm": 0.13435278832912445, "learning_rate": 6.325544272872905e-05, "loss": 0.011, "step": 35510 }, { "grad_norm": 0.09336785227060318, "learning_rate": 6.323551276967771e-05, "loss": 0.0085, "step": 35520 }, { "grad_norm": 0.14038006961345673, "learning_rate": 6.321558054903922e-05, "loss": 0.0086, "step": 35530 }, { "grad_norm": 0.13704364001750946, "learning_rate": 6.319564607021947e-05, "loss": 0.0089, "step": 35540 }, { "grad_norm": 0.17503447830677032, "learning_rate": 6.31757093366247e-05, "loss": 0.0067, "step": 35550 }, { "grad_norm": 0.08992141485214233, "learning_rate": 6.315577035166154e-05, "loss": 0.0078, "step": 35560 }, { "grad_norm": 0.09808928519487381, "learning_rate": 6.313582911873708e-05, "loss": 0.0077, "step": 35570 }, { "grad_norm": 0.1363820731639862, "learning_rate": 6.311588564125865e-05, "loss": 0.007, "step": 35580 }, { "grad_norm": 0.19073723256587982, "learning_rate": 6.30959399226341e-05, "loss": 0.0077, "step": 35590 }, { "grad_norm": 0.15367601811885834, "learning_rate": 6.30759919662716e-05, "loss": 0.0072, "step": 35600 }, { "grad_norm": 0.12378322333097458, "learning_rate": 6.30560417755797e-05, "loss": 0.0076, "step": 35610 }, { "grad_norm": 0.15511362254619598, "learning_rate": 6.303608935396735e-05, "loss": 0.0089, "step": 35620 }, { "grad_norm": 0.16040068864822388, "learning_rate": 6.301613470484386e-05, "loss": 0.0104, "step": 35630 }, { "grad_norm": 0.2022494524717331, "learning_rate": 6.299617783161893e-05, "loss": 0.011, "step": 35640 }, { "grad_norm": 0.14070463180541992, "learning_rate": 6.297621873770266e-05, "loss": 0.0083, "step": 35650 }, { "grad_norm": 0.11973851174116135, "learning_rate": 6.29562574265055e-05, "loss": 0.0086, "step": 35660 }, { "grad_norm": 0.09656040370464325, "learning_rate": 6.293629390143834e-05, "loss": 0.0084, "step": 35670 }, { "grad_norm": 0.14017488062381744, "learning_rate": 6.291632816591232e-05, "loss": 0.0084, "step": 35680 }, { "grad_norm": 0.12101903557777405, "learning_rate": 6.28963602233391e-05, "loss": 0.0099, "step": 35690 }, { "grad_norm": 0.14457204937934875, "learning_rate": 6.287639007713062e-05, "loss": 0.008, "step": 35700 }, { "grad_norm": 0.1120874360203743, "learning_rate": 6.285641773069926e-05, "loss": 0.0069, "step": 35710 }, { "grad_norm": 0.1358618438243866, "learning_rate": 6.283644318745773e-05, "loss": 0.0099, "step": 35720 }, { "grad_norm": 0.23412728309631348, "learning_rate": 6.281646645081912e-05, "loss": 0.0088, "step": 35730 }, { "grad_norm": 0.12175772339105606, "learning_rate": 6.279648752419693e-05, "loss": 0.008, "step": 35740 }, { "grad_norm": 0.13314326107501984, "learning_rate": 6.2776506411005e-05, "loss": 0.008, "step": 35750 }, { "grad_norm": 0.13438980281352997, "learning_rate": 6.275652311465758e-05, "loss": 0.0071, "step": 35760 }, { "grad_norm": 0.11495932191610336, "learning_rate": 6.273653763856926e-05, "loss": 0.0094, "step": 35770 }, { "grad_norm": 0.12027505785226822, "learning_rate": 6.271654998615501e-05, "loss": 0.006, "step": 35780 }, { "grad_norm": 0.1085178405046463, "learning_rate": 6.269656016083013e-05, "loss": 0.0062, "step": 35790 }, { "grad_norm": 0.1339692622423172, "learning_rate": 6.267656816601038e-05, "loss": 0.0086, "step": 35800 }, { "grad_norm": 0.16006630659103394, "learning_rate": 6.265657400511185e-05, "loss": 0.0093, "step": 35810 }, { "grad_norm": 0.1875039041042328, "learning_rate": 6.263657768155098e-05, "loss": 0.0076, "step": 35820 }, { "grad_norm": 0.09980420768260956, "learning_rate": 6.261657919874457e-05, "loss": 0.007, "step": 35830 }, { "grad_norm": 0.1092684417963028, "learning_rate": 6.259657856010986e-05, "loss": 0.0081, "step": 35840 }, { "grad_norm": 0.10019497573375702, "learning_rate": 6.257657576906439e-05, "loss": 0.0069, "step": 35850 }, { "grad_norm": 0.0974956750869751, "learning_rate": 6.255657082902609e-05, "loss": 0.0056, "step": 35860 }, { "grad_norm": 0.09535941481590271, "learning_rate": 6.253656374341325e-05, "loss": 0.0099, "step": 35870 }, { "grad_norm": 0.11164573580026627, "learning_rate": 6.251655451564457e-05, "loss": 0.0058, "step": 35880 }, { "grad_norm": 0.15643352270126343, "learning_rate": 6.249654314913902e-05, "loss": 0.0061, "step": 35890 }, { "grad_norm": 0.12867553532123566, "learning_rate": 6.247652964731604e-05, "loss": 0.0076, "step": 35900 }, { "grad_norm": 0.15712870657444, "learning_rate": 6.245651401359537e-05, "loss": 0.0106, "step": 35910 }, { "grad_norm": 0.15920855104923248, "learning_rate": 6.243649625139715e-05, "loss": 0.0057, "step": 35920 }, { "grad_norm": 0.12472416460514069, "learning_rate": 6.241647636414185e-05, "loss": 0.007, "step": 35930 }, { "grad_norm": 0.0991515964269638, "learning_rate": 6.239645435525034e-05, "loss": 0.0066, "step": 35940 }, { "grad_norm": 0.0695868507027626, "learning_rate": 6.237643022814381e-05, "loss": 0.0078, "step": 35950 }, { "grad_norm": 0.09875107556581497, "learning_rate": 6.235640398624386e-05, "loss": 0.0064, "step": 35960 }, { "grad_norm": 0.14324969053268433, "learning_rate": 6.233637563297243e-05, "loss": 0.0077, "step": 35970 }, { "grad_norm": 0.14387571811676025, "learning_rate": 6.23163451717518e-05, "loss": 0.008, "step": 35980 }, { "grad_norm": 0.14419931173324585, "learning_rate": 6.229631260600463e-05, "loss": 0.0083, "step": 35990 }, { "grad_norm": 0.10316380858421326, "learning_rate": 6.227627793915392e-05, "loss": 0.0088, "step": 36000 }, { "grad_norm": 0.11364123970270157, "learning_rate": 6.225624117462309e-05, "loss": 0.0067, "step": 36010 }, { "grad_norm": 0.13903690874576569, "learning_rate": 6.223620231583586e-05, "loss": 0.0068, "step": 36020 }, { "grad_norm": 0.1193193793296814, "learning_rate": 6.221616136621629e-05, "loss": 0.0095, "step": 36030 }, { "grad_norm": 0.11623455584049225, "learning_rate": 6.219611832918887e-05, "loss": 0.0079, "step": 36040 }, { "grad_norm": 0.09492585062980652, "learning_rate": 6.217607320817838e-05, "loss": 0.007, "step": 36050 }, { "grad_norm": 0.10347652435302734, "learning_rate": 6.215602600661001e-05, "loss": 0.0081, "step": 36060 }, { "grad_norm": 0.15734298527240753, "learning_rate": 6.213597672790925e-05, "loss": 0.0082, "step": 36070 }, { "grad_norm": 0.153029203414917, "learning_rate": 6.2115925375502e-05, "loss": 0.0079, "step": 36080 }, { "grad_norm": 0.1413414031267166, "learning_rate": 6.209587195281447e-05, "loss": 0.0091, "step": 36090 }, { "grad_norm": 0.13748276233673096, "learning_rate": 6.207581646327324e-05, "loss": 0.0105, "step": 36100 }, { "grad_norm": 0.1319030225276947, "learning_rate": 6.205575891030526e-05, "loss": 0.0081, "step": 36110 }, { "grad_norm": 0.14951293170452118, "learning_rate": 6.203569929733781e-05, "loss": 0.0077, "step": 36120 }, { "grad_norm": 0.1790236383676529, "learning_rate": 6.201563762779852e-05, "loss": 0.0078, "step": 36130 }, { "grad_norm": 0.10429598391056061, "learning_rate": 6.199557390511538e-05, "loss": 0.0061, "step": 36140 }, { "grad_norm": 0.16533230245113373, "learning_rate": 6.197550813271675e-05, "loss": 0.0076, "step": 36150 }, { "grad_norm": 0.11061668395996094, "learning_rate": 6.195544031403131e-05, "loss": 0.0068, "step": 36160 }, { "grad_norm": 0.1496538519859314, "learning_rate": 6.19353704524881e-05, "loss": 0.0093, "step": 36170 }, { "grad_norm": 0.10838852822780609, "learning_rate": 6.191529855151652e-05, "loss": 0.0078, "step": 36180 }, { "grad_norm": 0.10057618468999863, "learning_rate": 6.189522461454629e-05, "loss": 0.0067, "step": 36190 }, { "grad_norm": 0.1342383623123169, "learning_rate": 6.187514864500752e-05, "loss": 0.0062, "step": 36200 }, { "grad_norm": 0.0978202372789383, "learning_rate": 6.185507064633062e-05, "loss": 0.0058, "step": 36210 }, { "grad_norm": 0.11833624541759491, "learning_rate": 6.18349906219464e-05, "loss": 0.0063, "step": 36220 }, { "grad_norm": 0.11408492922782898, "learning_rate": 6.181490857528596e-05, "loss": 0.0097, "step": 36230 }, { "grad_norm": 0.1929982453584671, "learning_rate": 6.179482450978077e-05, "loss": 0.0077, "step": 36240 }, { "grad_norm": 0.0975540280342102, "learning_rate": 6.177473842886269e-05, "loss": 0.0072, "step": 36250 }, { "grad_norm": 0.17548248171806335, "learning_rate": 6.175465033596382e-05, "loss": 0.0077, "step": 36260 }, { "grad_norm": 0.13926559686660767, "learning_rate": 6.173456023451671e-05, "loss": 0.0076, "step": 36270 }, { "grad_norm": 0.14489039778709412, "learning_rate": 6.171446812795422e-05, "loss": 0.0081, "step": 36280 }, { "grad_norm": 0.17293605208396912, "learning_rate": 6.169437401970949e-05, "loss": 0.0075, "step": 36290 }, { "grad_norm": 0.1380087435245514, "learning_rate": 6.16742779132161e-05, "loss": 0.0063, "step": 36300 }, { "grad_norm": 0.13214096426963806, "learning_rate": 6.165417981190789e-05, "loss": 0.0069, "step": 36310 }, { "grad_norm": 0.11000916361808777, "learning_rate": 6.16340797192191e-05, "loss": 0.0064, "step": 36320 }, { "grad_norm": 0.1315373182296753, "learning_rate": 6.161397763858427e-05, "loss": 0.0089, "step": 36330 }, { "grad_norm": 0.1653812974691391, "learning_rate": 6.159387357343834e-05, "loss": 0.009, "step": 36340 }, { "grad_norm": 0.11932937055826187, "learning_rate": 6.157376752721648e-05, "loss": 0.0077, "step": 36350 }, { "grad_norm": 0.0990162044763565, "learning_rate": 6.155365950335428e-05, "loss": 0.006, "step": 36360 }, { "grad_norm": 0.13154244422912598, "learning_rate": 6.153354950528768e-05, "loss": 0.0077, "step": 36370 }, { "grad_norm": 0.16161425411701202, "learning_rate": 6.151343753645293e-05, "loss": 0.0086, "step": 36380 }, { "grad_norm": 0.15171529352664948, "learning_rate": 6.149332360028657e-05, "loss": 0.0104, "step": 36390 }, { "grad_norm": 0.12541580200195312, "learning_rate": 6.147320770022555e-05, "loss": 0.0071, "step": 36400 }, { "grad_norm": 0.17150886356830597, "learning_rate": 6.145308983970715e-05, "loss": 0.007, "step": 36410 }, { "grad_norm": 0.1292881816625595, "learning_rate": 6.143297002216892e-05, "loss": 0.0068, "step": 36420 }, { "grad_norm": 0.10176168382167816, "learning_rate": 6.141284825104882e-05, "loss": 0.0078, "step": 36430 }, { "grad_norm": 0.09045305848121643, "learning_rate": 6.13927245297851e-05, "loss": 0.0072, "step": 36440 }, { "grad_norm": 0.19968080520629883, "learning_rate": 6.137259886181633e-05, "loss": 0.0078, "step": 36450 }, { "grad_norm": 0.1384812444448471, "learning_rate": 6.135247125058145e-05, "loss": 0.007, "step": 36460 }, { "grad_norm": 0.1310977041721344, "learning_rate": 6.133234169951974e-05, "loss": 0.0071, "step": 36470 }, { "grad_norm": 0.10718825459480286, "learning_rate": 6.131221021207078e-05, "loss": 0.0068, "step": 36480 }, { "grad_norm": 0.11993750184774399, "learning_rate": 6.129207679167448e-05, "loss": 0.007, "step": 36490 }, { "grad_norm": 0.12756210565567017, "learning_rate": 6.127194144177109e-05, "loss": 0.0064, "step": 36500 }, { "grad_norm": 0.10590154677629471, "learning_rate": 6.125180416580118e-05, "loss": 0.0049, "step": 36510 }, { "grad_norm": 0.0973580852150917, "learning_rate": 6.123166496720571e-05, "loss": 0.0052, "step": 36520 }, { "grad_norm": 0.06894566118717194, "learning_rate": 6.121152384942588e-05, "loss": 0.0054, "step": 36530 }, { "grad_norm": 0.14590319991111755, "learning_rate": 6.119138081590324e-05, "loss": 0.0082, "step": 36540 }, { "grad_norm": 0.15572570264339447, "learning_rate": 6.117123587007971e-05, "loss": 0.0072, "step": 36550 }, { "grad_norm": 0.12252950668334961, "learning_rate": 6.11510890153975e-05, "loss": 0.0057, "step": 36560 }, { "grad_norm": 0.11785105615854263, "learning_rate": 6.113094025529916e-05, "loss": 0.0063, "step": 36570 }, { "grad_norm": 0.11254092305898666, "learning_rate": 6.111078959322757e-05, "loss": 0.0094, "step": 36580 }, { "grad_norm": 0.1746598333120346, "learning_rate": 6.109063703262592e-05, "loss": 0.0096, "step": 36590 }, { "grad_norm": 0.14269186556339264, "learning_rate": 6.107048257693772e-05, "loss": 0.0086, "step": 36600 }, { "grad_norm": 0.11104366183280945, "learning_rate": 6.105032622960683e-05, "loss": 0.0098, "step": 36610 }, { "grad_norm": 0.10274547338485718, "learning_rate": 6.103016799407743e-05, "loss": 0.008, "step": 36620 }, { "grad_norm": 0.14764785766601562, "learning_rate": 6.1010007873793984e-05, "loss": 0.0059, "step": 36630 }, { "grad_norm": 0.14869540929794312, "learning_rate": 6.098984587220131e-05, "loss": 0.0066, "step": 36640 }, { "grad_norm": 0.11990802735090256, "learning_rate": 6.096968199274456e-05, "loss": 0.0094, "step": 36650 }, { "grad_norm": 0.13425850868225098, "learning_rate": 6.0949516238869166e-05, "loss": 0.006, "step": 36660 }, { "grad_norm": 0.1322871446609497, "learning_rate": 6.092934861402092e-05, "loss": 0.0063, "step": 36670 }, { "grad_norm": 0.11599282175302505, "learning_rate": 6.0909179121645924e-05, "loss": 0.0074, "step": 36680 }, { "grad_norm": 0.1362949162721634, "learning_rate": 6.0889007765190576e-05, "loss": 0.0085, "step": 36690 }, { "grad_norm": 0.1305612325668335, "learning_rate": 6.086883454810162e-05, "loss": 0.0075, "step": 36700 }, { "grad_norm": 0.09843941777944565, "learning_rate": 6.0848659473826084e-05, "loss": 0.0072, "step": 36710 }, { "grad_norm": 0.16259488463401794, "learning_rate": 6.082848254581138e-05, "loss": 0.0084, "step": 36720 }, { "grad_norm": 0.10037059336900711, "learning_rate": 6.080830376750517e-05, "loss": 0.0067, "step": 36730 }, { "grad_norm": 0.10632017999887466, "learning_rate": 6.0788123142355445e-05, "loss": 0.0065, "step": 36740 }, { "grad_norm": 0.09347204118967056, "learning_rate": 6.076794067381052e-05, "loss": 0.0094, "step": 36750 }, { "grad_norm": 0.16442175209522247, "learning_rate": 6.074775636531905e-05, "loss": 0.0088, "step": 36760 }, { "grad_norm": 0.14991290867328644, "learning_rate": 6.072757022032997e-05, "loss": 0.0053, "step": 36770 }, { "grad_norm": 0.14823561906814575, "learning_rate": 6.070738224229253e-05, "loss": 0.0077, "step": 36780 }, { "grad_norm": 0.13643482327461243, "learning_rate": 6.0687192434656314e-05, "loss": 0.0079, "step": 36790 }, { "grad_norm": 0.15892213582992554, "learning_rate": 6.066700080087121e-05, "loss": 0.0088, "step": 36800 }, { "grad_norm": 0.11238997429609299, "learning_rate": 6.0646807344387424e-05, "loss": 0.0063, "step": 36810 }, { "grad_norm": 0.13008810579776764, "learning_rate": 6.062661206865543e-05, "loss": 0.0086, "step": 36820 }, { "grad_norm": 0.14392341673374176, "learning_rate": 6.06064149771261e-05, "loss": 0.0102, "step": 36830 }, { "grad_norm": 0.15272140502929688, "learning_rate": 6.058621607325051e-05, "loss": 0.0072, "step": 36840 }, { "grad_norm": 0.15202999114990234, "learning_rate": 6.056601536048014e-05, "loss": 0.0087, "step": 36850 }, { "grad_norm": 0.17835773527622223, "learning_rate": 6.0545812842266725e-05, "loss": 0.0084, "step": 36860 }, { "grad_norm": 0.1294049620628357, "learning_rate": 6.052560852206232e-05, "loss": 0.0074, "step": 36870 }, { "grad_norm": 0.08915605396032333, "learning_rate": 6.05054024033193e-05, "loss": 0.008, "step": 36880 }, { "grad_norm": 0.13705593347549438, "learning_rate": 6.048519448949032e-05, "loss": 0.0065, "step": 36890 }, { "grad_norm": 0.14662252366542816, "learning_rate": 6.046498478402839e-05, "loss": 0.0068, "step": 36900 }, { "grad_norm": 0.18151946365833282, "learning_rate": 6.044477329038677e-05, "loss": 0.0107, "step": 36910 }, { "grad_norm": 0.13824544847011566, "learning_rate": 6.042456001201906e-05, "loss": 0.0069, "step": 36920 }, { "grad_norm": 0.09207116812467575, "learning_rate": 6.040434495237917e-05, "loss": 0.0061, "step": 36930 }, { "grad_norm": 0.15199212729930878, "learning_rate": 6.0384128114921256e-05, "loss": 0.0083, "step": 36940 }, { "grad_norm": 0.15959152579307556, "learning_rate": 6.036390950309987e-05, "loss": 0.0106, "step": 36950 }, { "grad_norm": 0.12229810655117035, "learning_rate": 6.0343689120369805e-05, "loss": 0.0112, "step": 36960 }, { "grad_norm": 0.1223808005452156, "learning_rate": 6.032346697018616e-05, "loss": 0.008, "step": 36970 }, { "grad_norm": 0.11547254770994186, "learning_rate": 6.0303243056004375e-05, "loss": 0.0065, "step": 36980 }, { "grad_norm": 0.1396402269601822, "learning_rate": 6.0283017381280136e-05, "loss": 0.0062, "step": 36990 }, { "grad_norm": 0.10467445850372314, "learning_rate": 6.026278994946948e-05, "loss": 0.0091, "step": 37000 }, { "grad_norm": 0.16552890837192535, "learning_rate": 6.02425607640287e-05, "loss": 0.0078, "step": 37010 }, { "grad_norm": 0.12794490158557892, "learning_rate": 6.022232982841441e-05, "loss": 0.0064, "step": 37020 }, { "grad_norm": 0.1327289491891861, "learning_rate": 6.020209714608355e-05, "loss": 0.0067, "step": 37030 }, { "grad_norm": 0.10871978849172592, "learning_rate": 6.018186272049331e-05, "loss": 0.0089, "step": 37040 }, { "grad_norm": 0.17000271379947662, "learning_rate": 6.01616265551012e-05, "loss": 0.0077, "step": 37050 }, { "grad_norm": 0.11039507389068604, "learning_rate": 6.014138865336503e-05, "loss": 0.0091, "step": 37060 }, { "grad_norm": 0.12659093737602234, "learning_rate": 6.0121149018742905e-05, "loss": 0.0068, "step": 37070 }, { "grad_norm": 0.16015033423900604, "learning_rate": 6.010090765469325e-05, "loss": 0.0066, "step": 37080 }, { "grad_norm": 0.15344054996967316, "learning_rate": 6.008066456467473e-05, "loss": 0.0093, "step": 37090 }, { "grad_norm": 0.13679683208465576, "learning_rate": 6.0060419752146335e-05, "loss": 0.0073, "step": 37100 }, { "grad_norm": 0.15039269626140594, "learning_rate": 6.0040173220567353e-05, "loss": 0.0065, "step": 37110 }, { "grad_norm": 0.1824645698070526, "learning_rate": 6.001992497339737e-05, "loss": 0.0057, "step": 37120 }, { "grad_norm": 0.11167758703231812, "learning_rate": 5.999967501409626e-05, "loss": 0.0065, "step": 37130 }, { "grad_norm": 0.10827397555112839, "learning_rate": 5.997942334612418e-05, "loss": 0.0061, "step": 37140 }, { "grad_norm": 0.16931606829166412, "learning_rate": 5.995916997294158e-05, "loss": 0.0081, "step": 37150 }, { "grad_norm": 0.14775338768959045, "learning_rate": 5.9938914898009214e-05, "loss": 0.0072, "step": 37160 }, { "grad_norm": 0.1554899960756302, "learning_rate": 5.991865812478813e-05, "loss": 0.0085, "step": 37170 }, { "grad_norm": 0.12990830838680267, "learning_rate": 5.989839965673964e-05, "loss": 0.0084, "step": 37180 }, { "grad_norm": 0.1507844775915146, "learning_rate": 5.987813949732539e-05, "loss": 0.0108, "step": 37190 }, { "grad_norm": 0.1250305473804474, "learning_rate": 5.9857877650007255e-05, "loss": 0.008, "step": 37200 }, { "grad_norm": 0.1582815945148468, "learning_rate": 5.983761411824744e-05, "loss": 0.01, "step": 37210 }, { "grad_norm": 0.13178424537181854, "learning_rate": 5.981734890550844e-05, "loss": 0.0083, "step": 37220 }, { "grad_norm": 0.13112834095954895, "learning_rate": 5.979708201525301e-05, "loss": 0.0093, "step": 37230 }, { "grad_norm": 0.1353292316198349, "learning_rate": 5.977681345094422e-05, "loss": 0.0063, "step": 37240 }, { "grad_norm": 0.1301397979259491, "learning_rate": 5.97565432160454e-05, "loss": 0.008, "step": 37250 }, { "grad_norm": 0.10830657929182053, "learning_rate": 5.9736271314020186e-05, "loss": 0.0071, "step": 37260 }, { "grad_norm": 0.14870749413967133, "learning_rate": 5.971599774833251e-05, "loss": 0.0067, "step": 37270 }, { "grad_norm": 0.15640395879745483, "learning_rate": 5.9695722522446525e-05, "loss": 0.0086, "step": 37280 }, { "grad_norm": 0.16708868741989136, "learning_rate": 5.9675445639826765e-05, "loss": 0.0063, "step": 37290 }, { "grad_norm": 0.13161326944828033, "learning_rate": 5.965516710393796e-05, "loss": 0.0082, "step": 37300 }, { "grad_norm": 0.1011207327246666, "learning_rate": 5.963488691824516e-05, "loss": 0.0062, "step": 37310 }, { "grad_norm": 0.14924605190753937, "learning_rate": 5.96146050862137e-05, "loss": 0.0103, "step": 37320 }, { "grad_norm": 0.17238543927669525, "learning_rate": 5.959432161130919e-05, "loss": 0.0088, "step": 37330 }, { "grad_norm": 0.13060766458511353, "learning_rate": 5.9574036496997545e-05, "loss": 0.0089, "step": 37340 }, { "grad_norm": 0.10235485434532166, "learning_rate": 5.955374974674488e-05, "loss": 0.0086, "step": 37350 }, { "grad_norm": 0.11578378826379776, "learning_rate": 5.9533461364017696e-05, "loss": 0.0084, "step": 37360 }, { "grad_norm": 0.09585435688495636, "learning_rate": 5.9513171352282716e-05, "loss": 0.0065, "step": 37370 }, { "grad_norm": 0.13662710785865784, "learning_rate": 5.949287971500692e-05, "loss": 0.0074, "step": 37380 }, { "grad_norm": 0.09699229896068573, "learning_rate": 5.947258645565762e-05, "loss": 0.0089, "step": 37390 }, { "grad_norm": 0.11854325979948044, "learning_rate": 5.945229157770237e-05, "loss": 0.0079, "step": 37400 }, { "grad_norm": 0.12347825616598129, "learning_rate": 5.9431995084609006e-05, "loss": 0.0104, "step": 37410 }, { "grad_norm": 0.1272701621055603, "learning_rate": 5.941169697984564e-05, "loss": 0.0067, "step": 37420 }, { "grad_norm": 0.10762027651071548, "learning_rate": 5.9391397266880675e-05, "loss": 0.006, "step": 37430 }, { "grad_norm": 0.13583026826381683, "learning_rate": 5.937109594918279e-05, "loss": 0.0104, "step": 37440 }, { "grad_norm": 0.10638024657964706, "learning_rate": 5.9350793030220884e-05, "loss": 0.0075, "step": 37450 }, { "grad_norm": 0.1479272097349167, "learning_rate": 5.933048851346421e-05, "loss": 0.0098, "step": 37460 }, { "grad_norm": 0.13106603920459747, "learning_rate": 5.931018240238224e-05, "loss": 0.0089, "step": 37470 }, { "grad_norm": 0.17446422576904297, "learning_rate": 5.928987470044471e-05, "loss": 0.0079, "step": 37480 }, { "grad_norm": 0.12251490354537964, "learning_rate": 5.9269565411121695e-05, "loss": 0.008, "step": 37490 }, { "grad_norm": 0.14778321981430054, "learning_rate": 5.924925453788347e-05, "loss": 0.0088, "step": 37500 }, { "grad_norm": 0.13061058521270752, "learning_rate": 5.92289420842006e-05, "loss": 0.007, "step": 37510 }, { "grad_norm": 0.12779982388019562, "learning_rate": 5.9208628053543945e-05, "loss": 0.0102, "step": 37520 }, { "grad_norm": 0.18988314270973206, "learning_rate": 5.918831244938462e-05, "loss": 0.0086, "step": 37530 }, { "grad_norm": 0.15187084674835205, "learning_rate": 5.9167995275194e-05, "loss": 0.0083, "step": 37540 }, { "grad_norm": 0.10284318029880524, "learning_rate": 5.914767653444373e-05, "loss": 0.0082, "step": 37550 }, { "grad_norm": 0.11086934059858322, "learning_rate": 5.912735623060572e-05, "loss": 0.0077, "step": 37560 }, { "grad_norm": 0.14998981356620789, "learning_rate": 5.910703436715217e-05, "loss": 0.0107, "step": 37570 }, { "grad_norm": 0.15151384472846985, "learning_rate": 5.908671094755552e-05, "loss": 0.007, "step": 37580 }, { "grad_norm": 0.16839799284934998, "learning_rate": 5.906638597528851e-05, "loss": 0.0072, "step": 37590 }, { "grad_norm": 0.110945925116539, "learning_rate": 5.9046059453824076e-05, "loss": 0.0051, "step": 37600 }, { "grad_norm": 0.14397044479846954, "learning_rate": 5.9025731386635505e-05, "loss": 0.0069, "step": 37610 }, { "grad_norm": 0.11567030847072601, "learning_rate": 5.900540177719629e-05, "loss": 0.006, "step": 37620 }, { "grad_norm": 0.11322491616010666, "learning_rate": 5.898507062898021e-05, "loss": 0.0056, "step": 37630 }, { "grad_norm": 0.19257237017154694, "learning_rate": 5.8964737945461316e-05, "loss": 0.0066, "step": 37640 }, { "grad_norm": 0.1656814068555832, "learning_rate": 5.8944403730113885e-05, "loss": 0.0087, "step": 37650 }, { "grad_norm": 0.0944715216755867, "learning_rate": 5.892406798641248e-05, "loss": 0.0078, "step": 37660 }, { "grad_norm": 0.14352117478847504, "learning_rate": 5.890373071783193e-05, "loss": 0.0093, "step": 37670 }, { "grad_norm": 0.14057058095932007, "learning_rate": 5.888339192784732e-05, "loss": 0.0108, "step": 37680 }, { "grad_norm": 0.09522232413291931, "learning_rate": 5.8863051619934003e-05, "loss": 0.0092, "step": 37690 }, { "grad_norm": 0.0789249911904335, "learning_rate": 5.8842709797567554e-05, "loss": 0.0064, "step": 37700 }, { "grad_norm": 0.10420020669698715, "learning_rate": 5.8822366464223855e-05, "loss": 0.0077, "step": 37710 }, { "grad_norm": 0.12516510486602783, "learning_rate": 5.880202162337901e-05, "loss": 0.0064, "step": 37720 }, { "grad_norm": 0.08892136067152023, "learning_rate": 5.8781675278509405e-05, "loss": 0.006, "step": 37730 }, { "grad_norm": 0.16961108148097992, "learning_rate": 5.8761327433091696e-05, "loss": 0.0068, "step": 37740 }, { "grad_norm": 0.11805272102355957, "learning_rate": 5.874097809060275e-05, "loss": 0.0073, "step": 37750 }, { "grad_norm": 0.10439689457416534, "learning_rate": 5.87206272545197e-05, "loss": 0.0071, "step": 37760 }, { "grad_norm": 0.10039620101451874, "learning_rate": 5.8700274928319955e-05, "loss": 0.0054, "step": 37770 }, { "grad_norm": 0.11434631794691086, "learning_rate": 5.867992111548118e-05, "loss": 0.0092, "step": 37780 }, { "grad_norm": 0.12152652442455292, "learning_rate": 5.865956581948131e-05, "loss": 0.0051, "step": 37790 }, { "grad_norm": 0.11234888434410095, "learning_rate": 5.863920904379845e-05, "loss": 0.0061, "step": 37800 }, { "grad_norm": 0.08668126910924911, "learning_rate": 5.8618850791911064e-05, "loss": 0.0074, "step": 37810 }, { "grad_norm": 0.14091083407402039, "learning_rate": 5.859849106729779e-05, "loss": 0.0079, "step": 37820 }, { "grad_norm": 0.1624642312526703, "learning_rate": 5.857812987343758e-05, "loss": 0.0059, "step": 37830 }, { "grad_norm": 0.16918078064918518, "learning_rate": 5.855776721380957e-05, "loss": 0.0068, "step": 37840 }, { "grad_norm": 0.13278011977672577, "learning_rate": 5.8537403091893217e-05, "loss": 0.0091, "step": 37850 }, { "grad_norm": 0.08827786892652512, "learning_rate": 5.851703751116816e-05, "loss": 0.0063, "step": 37860 }, { "grad_norm": 0.1410876363515854, "learning_rate": 5.8496670475114336e-05, "loss": 0.0058, "step": 37870 }, { "grad_norm": 0.12181023508310318, "learning_rate": 5.84763019872119e-05, "loss": 0.0068, "step": 37880 }, { "grad_norm": 0.1253756731748581, "learning_rate": 5.845593205094131e-05, "loss": 0.0067, "step": 37890 }, { "grad_norm": 0.13713902235031128, "learning_rate": 5.843556066978318e-05, "loss": 0.0066, "step": 37900 }, { "grad_norm": 0.09199566394090652, "learning_rate": 5.8415187847218455e-05, "loss": 0.0065, "step": 37910 }, { "grad_norm": 0.11449813097715378, "learning_rate": 5.839481358672827e-05, "loss": 0.0064, "step": 37920 }, { "grad_norm": 0.12389139831066132, "learning_rate": 5.837443789179407e-05, "loss": 0.0069, "step": 37930 }, { "grad_norm": 0.09707915782928467, "learning_rate": 5.8354060765897445e-05, "loss": 0.0066, "step": 37940 }, { "grad_norm": 0.11164036393165588, "learning_rate": 5.8333682212520334e-05, "loss": 0.0062, "step": 37950 }, { "grad_norm": 0.10708779096603394, "learning_rate": 5.831330223514486e-05, "loss": 0.0069, "step": 37960 }, { "grad_norm": 0.11435877531766891, "learning_rate": 5.8292920837253396e-05, "loss": 0.0074, "step": 37970 }, { "grad_norm": 0.16130001842975616, "learning_rate": 5.827253802232857e-05, "loss": 0.0076, "step": 37980 }, { "grad_norm": 0.08964280039072037, "learning_rate": 5.825215379385327e-05, "loss": 0.0057, "step": 37990 }, { "grad_norm": 0.15512287616729736, "learning_rate": 5.823176815531057e-05, "loss": 0.01, "step": 38000 }, { "grad_norm": 0.1412704735994339, "learning_rate": 5.8211381110183826e-05, "loss": 0.0073, "step": 38010 }, { "grad_norm": 0.11398015171289444, "learning_rate": 5.8190992661956645e-05, "loss": 0.0065, "step": 38020 }, { "grad_norm": 0.1264055371284485, "learning_rate": 5.817060281411284e-05, "loss": 0.0052, "step": 38030 }, { "grad_norm": 0.07599341869354248, "learning_rate": 5.815021157013647e-05, "loss": 0.0058, "step": 38040 }, { "grad_norm": 0.10798081010580063, "learning_rate": 5.8129818933511856e-05, "loss": 0.008, "step": 38050 }, { "grad_norm": 0.08873172849416733, "learning_rate": 5.8109424907723544e-05, "loss": 0.0072, "step": 38060 }, { "grad_norm": 0.0971498042345047, "learning_rate": 5.80890294962563e-05, "loss": 0.0053, "step": 38070 }, { "grad_norm": 0.1462305635213852, "learning_rate": 5.806863270259515e-05, "loss": 0.0072, "step": 38080 }, { "grad_norm": 0.1271641105413437, "learning_rate": 5.804823453022536e-05, "loss": 0.009, "step": 38090 }, { "grad_norm": 0.137285515666008, "learning_rate": 5.80278349826324e-05, "loss": 0.0088, "step": 38100 }, { "grad_norm": 0.10468050092458725, "learning_rate": 5.8007434063302014e-05, "loss": 0.0069, "step": 38110 }, { "grad_norm": 0.1801498979330063, "learning_rate": 5.7987031775720136e-05, "loss": 0.0073, "step": 38120 }, { "grad_norm": 0.1509336531162262, "learning_rate": 5.7966628123372976e-05, "loss": 0.0093, "step": 38130 }, { "grad_norm": 0.1348246932029724, "learning_rate": 5.7946223109746956e-05, "loss": 0.0057, "step": 38140 }, { "grad_norm": 0.16058558225631714, "learning_rate": 5.7925816738328754e-05, "loss": 0.0057, "step": 38150 }, { "grad_norm": 0.16233009099960327, "learning_rate": 5.790540901260521e-05, "loss": 0.0073, "step": 38160 }, { "grad_norm": 0.13998833298683167, "learning_rate": 5.788499993606351e-05, "loss": 0.0086, "step": 38170 }, { "grad_norm": 0.10389234125614166, "learning_rate": 5.786458951219096e-05, "loss": 0.0058, "step": 38180 }, { "grad_norm": 0.11676081269979477, "learning_rate": 5.784417774447517e-05, "loss": 0.0076, "step": 38190 }, { "grad_norm": 0.12798045575618744, "learning_rate": 5.782376463640393e-05, "loss": 0.0064, "step": 38200 }, { "grad_norm": 0.12107692658901215, "learning_rate": 5.780335019146531e-05, "loss": 0.0072, "step": 38210 }, { "grad_norm": 0.15204107761383057, "learning_rate": 5.778293441314755e-05, "loss": 0.007, "step": 38220 }, { "grad_norm": 0.1230846643447876, "learning_rate": 5.776251730493917e-05, "loss": 0.0097, "step": 38230 }, { "grad_norm": 0.13633349537849426, "learning_rate": 5.774209887032887e-05, "loss": 0.0069, "step": 38240 }, { "grad_norm": 0.13354764878749847, "learning_rate": 5.772167911280565e-05, "loss": 0.0084, "step": 38250 }, { "grad_norm": 0.1257258653640747, "learning_rate": 5.770125803585864e-05, "loss": 0.0078, "step": 38260 }, { "grad_norm": 0.16245517134666443, "learning_rate": 5.768083564297726e-05, "loss": 0.0068, "step": 38270 }, { "grad_norm": 0.16946512460708618, "learning_rate": 5.766041193765114e-05, "loss": 0.0076, "step": 38280 }, { "grad_norm": 0.11052614450454712, "learning_rate": 5.763998692337015e-05, "loss": 0.0076, "step": 38290 }, { "grad_norm": 0.08453895896673203, "learning_rate": 5.761956060362433e-05, "loss": 0.0064, "step": 38300 }, { "grad_norm": 0.11126245558261871, "learning_rate": 5.7599132981904e-05, "loss": 0.0071, "step": 38310 }, { "grad_norm": 0.09462876617908478, "learning_rate": 5.75787040616997e-05, "loss": 0.0066, "step": 38320 }, { "grad_norm": 0.09298515319824219, "learning_rate": 5.755827384650212e-05, "loss": 0.0082, "step": 38330 }, { "grad_norm": 0.12764480710029602, "learning_rate": 5.753784233980228e-05, "loss": 0.0109, "step": 38340 }, { "grad_norm": 0.1335211545228958, "learning_rate": 5.751740954509135e-05, "loss": 0.0068, "step": 38350 }, { "grad_norm": 0.14106208086013794, "learning_rate": 5.7496975465860715e-05, "loss": 0.0071, "step": 38360 }, { "grad_norm": 0.17424680292606354, "learning_rate": 5.747654010560202e-05, "loss": 0.0086, "step": 38370 }, { "grad_norm": 0.17857657372951508, "learning_rate": 5.7456103467807097e-05, "loss": 0.0083, "step": 38380 }, { "grad_norm": 0.16724185645580292, "learning_rate": 5.7435665555968046e-05, "loss": 0.0069, "step": 38390 }, { "grad_norm": 0.11005294322967529, "learning_rate": 5.74152263735771e-05, "loss": 0.0065, "step": 38400 }, { "grad_norm": 0.17869268357753754, "learning_rate": 5.739478592412677e-05, "loss": 0.0103, "step": 38410 }, { "grad_norm": 0.16518272459506989, "learning_rate": 5.7374344211109766e-05, "loss": 0.0083, "step": 38420 }, { "grad_norm": 0.1347580850124359, "learning_rate": 5.735390123801904e-05, "loss": 0.008, "step": 38430 }, { "grad_norm": 0.1529112458229065, "learning_rate": 5.7333457008347704e-05, "loss": 0.0066, "step": 38440 }, { "grad_norm": 0.15355005860328674, "learning_rate": 5.7313011525589156e-05, "loss": 0.0073, "step": 38450 }, { "grad_norm": 0.11846549063920975, "learning_rate": 5.729256479323694e-05, "loss": 0.0091, "step": 38460 }, { "grad_norm": 0.11830013245344162, "learning_rate": 5.727211681478485e-05, "loss": 0.008, "step": 38470 }, { "grad_norm": 0.11352997273206711, "learning_rate": 5.7251667593726886e-05, "loss": 0.0072, "step": 38480 }, { "grad_norm": 0.11413177847862244, "learning_rate": 5.723121713355728e-05, "loss": 0.0082, "step": 38490 }, { "grad_norm": 0.09753549844026566, "learning_rate": 5.721076543777044e-05, "loss": 0.0088, "step": 38500 }, { "grad_norm": 0.11313440650701523, "learning_rate": 5.7190312509860986e-05, "loss": 0.0089, "step": 38510 }, { "grad_norm": 0.1857863813638687, "learning_rate": 5.716985835332379e-05, "loss": 0.0064, "step": 38520 }, { "grad_norm": 0.13062992691993713, "learning_rate": 5.714940297165389e-05, "loss": 0.0073, "step": 38530 }, { "grad_norm": 0.12577232718467712, "learning_rate": 5.712894636834656e-05, "loss": 0.0088, "step": 38540 }, { "grad_norm": 0.13537685573101044, "learning_rate": 5.7108488546897276e-05, "loss": 0.0102, "step": 38550 }, { "grad_norm": 0.11794325709342957, "learning_rate": 5.708802951080172e-05, "loss": 0.0064, "step": 38560 }, { "grad_norm": 0.11457855999469757, "learning_rate": 5.706756926355576e-05, "loss": 0.0071, "step": 38570 }, { "grad_norm": 0.116780124604702, "learning_rate": 5.704710780865554e-05, "loss": 0.007, "step": 38580 }, { "grad_norm": 0.15223824977874756, "learning_rate": 5.7026645149597325e-05, "loss": 0.0095, "step": 38590 }, { "grad_norm": 0.0813203677535057, "learning_rate": 5.700618128987764e-05, "loss": 0.0068, "step": 38600 }, { "grad_norm": 0.13912521302700043, "learning_rate": 5.698571623299317e-05, "loss": 0.0076, "step": 38610 }, { "grad_norm": 0.1265251338481903, "learning_rate": 5.696524998244086e-05, "loss": 0.0095, "step": 38620 }, { "grad_norm": 0.09124810993671417, "learning_rate": 5.6944782541717836e-05, "loss": 0.0066, "step": 38630 }, { "grad_norm": 0.12590093910694122, "learning_rate": 5.69243139143214e-05, "loss": 0.0074, "step": 38640 }, { "grad_norm": 0.09537876397371292, "learning_rate": 5.6903844103749125e-05, "loss": 0.0079, "step": 38650 }, { "grad_norm": 0.19654592871665955, "learning_rate": 5.688337311349869e-05, "loss": 0.0076, "step": 38660 }, { "grad_norm": 0.1423349529504776, "learning_rate": 5.6862900947068074e-05, "loss": 0.0068, "step": 38670 }, { "grad_norm": 0.11860872805118561, "learning_rate": 5.6842427607955374e-05, "loss": 0.0074, "step": 38680 }, { "grad_norm": 0.13342951238155365, "learning_rate": 5.682195309965893e-05, "loss": 0.0069, "step": 38690 }, { "grad_norm": 0.10939036309719086, "learning_rate": 5.6801477425677294e-05, "loss": 0.0066, "step": 38700 }, { "grad_norm": 0.135727658867836, "learning_rate": 5.678100058950917e-05, "loss": 0.0075, "step": 38710 }, { "grad_norm": 0.10338586568832397, "learning_rate": 5.676052259465352e-05, "loss": 0.0069, "step": 38720 }, { "grad_norm": 0.19994457066059113, "learning_rate": 5.674004344460945e-05, "loss": 0.0079, "step": 38730 }, { "grad_norm": 0.11549139767885208, "learning_rate": 5.6719563142876295e-05, "loss": 0.0079, "step": 38740 }, { "grad_norm": 0.12380542606115341, "learning_rate": 5.669908169295359e-05, "loss": 0.0093, "step": 38750 }, { "grad_norm": 0.12488492578268051, "learning_rate": 5.667859909834105e-05, "loss": 0.0066, "step": 38760 }, { "grad_norm": 0.1484716534614563, "learning_rate": 5.6658115362538565e-05, "loss": 0.0086, "step": 38770 }, { "grad_norm": 0.09634076058864594, "learning_rate": 5.663763048904628e-05, "loss": 0.0062, "step": 38780 }, { "grad_norm": 0.12292508780956268, "learning_rate": 5.661714448136447e-05, "loss": 0.0069, "step": 38790 }, { "grad_norm": 0.12250132113695145, "learning_rate": 5.659665734299366e-05, "loss": 0.0071, "step": 38800 }, { "grad_norm": 0.15847128629684448, "learning_rate": 5.6576169077434516e-05, "loss": 0.0078, "step": 38810 }, { "grad_norm": 0.0862194374203682, "learning_rate": 5.6555679688187944e-05, "loss": 0.0082, "step": 38820 }, { "grad_norm": 0.15212494134902954, "learning_rate": 5.6535189178755e-05, "loss": 0.0076, "step": 38830 }, { "grad_norm": 0.08725537359714508, "learning_rate": 5.651469755263698e-05, "loss": 0.0066, "step": 38840 }, { "grad_norm": 0.09724313020706177, "learning_rate": 5.6494204813335316e-05, "loss": 0.007, "step": 38850 }, { "grad_norm": 0.1137147769331932, "learning_rate": 5.647371096435168e-05, "loss": 0.0096, "step": 38860 }, { "grad_norm": 0.15767325460910797, "learning_rate": 5.645321600918788e-05, "loss": 0.0073, "step": 38870 }, { "grad_norm": 0.17278572916984558, "learning_rate": 5.643271995134597e-05, "loss": 0.0072, "step": 38880 }, { "grad_norm": 0.15469150245189667, "learning_rate": 5.641222279432814e-05, "loss": 0.0071, "step": 38890 }, { "grad_norm": 0.1362851858139038, "learning_rate": 5.6391724541636834e-05, "loss": 0.0086, "step": 38900 }, { "grad_norm": 0.11588861048221588, "learning_rate": 5.6371225196774605e-05, "loss": 0.0083, "step": 38910 }, { "grad_norm": 0.11044156551361084, "learning_rate": 5.635072476324423e-05, "loss": 0.0074, "step": 38920 }, { "grad_norm": 0.1004352942109108, "learning_rate": 5.63302232445487e-05, "loss": 0.0059, "step": 38930 }, { "grad_norm": 0.11056917905807495, "learning_rate": 5.6309720644191144e-05, "loss": 0.0068, "step": 38940 }, { "grad_norm": 0.1376808136701584, "learning_rate": 5.628921696567491e-05, "loss": 0.0081, "step": 38950 }, { "grad_norm": 0.13740475475788116, "learning_rate": 5.62687122125035e-05, "loss": 0.0073, "step": 38960 }, { "grad_norm": 0.13507848978042603, "learning_rate": 5.624820638818062e-05, "loss": 0.0063, "step": 38970 }, { "grad_norm": 0.12441527843475342, "learning_rate": 5.6227699496210164e-05, "loss": 0.0072, "step": 38980 }, { "grad_norm": 0.10036002099514008, "learning_rate": 5.6207191540096195e-05, "loss": 0.008, "step": 38990 }, { "grad_norm": 0.1645498275756836, "learning_rate": 5.618668252334296e-05, "loss": 0.0068, "step": 39000 }, { "grad_norm": 0.1555056869983673, "learning_rate": 5.616617244945488e-05, "loss": 0.0086, "step": 39010 }, { "grad_norm": 0.11654243618249893, "learning_rate": 5.614566132193656e-05, "loss": 0.0061, "step": 39020 }, { "grad_norm": 0.09186206758022308, "learning_rate": 5.612514914429282e-05, "loss": 0.008, "step": 39030 }, { "grad_norm": 0.17855021357536316, "learning_rate": 5.610463592002863e-05, "loss": 0.0088, "step": 39040 }, { "grad_norm": 0.12224022299051285, "learning_rate": 5.608412165264909e-05, "loss": 0.0057, "step": 39050 }, { "grad_norm": 0.09213852882385254, "learning_rate": 5.606360634565959e-05, "loss": 0.0089, "step": 39060 }, { "grad_norm": 0.12418083101511002, "learning_rate": 5.604309000256559e-05, "loss": 0.0083, "step": 39070 }, { "grad_norm": 0.15000513195991516, "learning_rate": 5.6022572626872785e-05, "loss": 0.0058, "step": 39080 }, { "grad_norm": 0.14065614342689514, "learning_rate": 5.600205422208704e-05, "loss": 0.0081, "step": 39090 }, { "grad_norm": 0.08968483656644821, "learning_rate": 5.59815347917144e-05, "loss": 0.0079, "step": 39100 }, { "grad_norm": 0.16774830222129822, "learning_rate": 5.596101433926103e-05, "loss": 0.0053, "step": 39110 }, { "grad_norm": 0.1608010083436966, "learning_rate": 5.5940492868233364e-05, "loss": 0.0096, "step": 39120 }, { "grad_norm": 0.11264584958553314, "learning_rate": 5.591997038213793e-05, "loss": 0.0089, "step": 39130 }, { "grad_norm": 0.1571396440267563, "learning_rate": 5.5899446884481475e-05, "loss": 0.0076, "step": 39140 }, { "grad_norm": 0.11817991733551025, "learning_rate": 5.5878922378770906e-05, "loss": 0.0061, "step": 39150 }, { "grad_norm": 0.13368062674999237, "learning_rate": 5.5858396868513285e-05, "loss": 0.0064, "step": 39160 }, { "grad_norm": 0.0940900519490242, "learning_rate": 5.583787035721586e-05, "loss": 0.0058, "step": 39170 }, { "grad_norm": 0.12954066693782806, "learning_rate": 5.581734284838606e-05, "loss": 0.0065, "step": 39180 }, { "grad_norm": 0.12350958585739136, "learning_rate": 5.579681434553147e-05, "loss": 0.0077, "step": 39190 }, { "grad_norm": 0.1286369413137436, "learning_rate": 5.5776284852159854e-05, "loss": 0.0065, "step": 39200 }, { "grad_norm": 0.13656601309776306, "learning_rate": 5.575575437177913e-05, "loss": 0.0075, "step": 39210 }, { "grad_norm": 0.10017666220664978, "learning_rate": 5.573522290789742e-05, "loss": 0.0077, "step": 39220 }, { "grad_norm": 0.14611008763313293, "learning_rate": 5.571469046402298e-05, "loss": 0.0072, "step": 39230 }, { "grad_norm": 0.214033842086792, "learning_rate": 5.5694157043664205e-05, "loss": 0.0106, "step": 39240 }, { "grad_norm": 0.14352916181087494, "learning_rate": 5.567362265032975e-05, "loss": 0.0052, "step": 39250 }, { "grad_norm": 0.1576082408428192, "learning_rate": 5.565308728752836e-05, "loss": 0.0056, "step": 39260 }, { "grad_norm": 0.1561266928911209, "learning_rate": 5.5632550958768956e-05, "loss": 0.0059, "step": 39270 }, { "grad_norm": 0.13988861441612244, "learning_rate": 5.5612013667560636e-05, "loss": 0.0098, "step": 39280 }, { "grad_norm": 0.12857849895954132, "learning_rate": 5.5591475417412676e-05, "loss": 0.0095, "step": 39290 }, { "grad_norm": 0.12722617387771606, "learning_rate": 5.557093621183451e-05, "loss": 0.0067, "step": 39300 }, { "grad_norm": 0.13284501433372498, "learning_rate": 5.55503960543357e-05, "loss": 0.0075, "step": 39310 }, { "grad_norm": 0.16253340244293213, "learning_rate": 5.552985494842601e-05, "loss": 0.0071, "step": 39320 }, { "grad_norm": 0.09734006971120834, "learning_rate": 5.550931289761534e-05, "loss": 0.008, "step": 39330 }, { "grad_norm": 0.1373463124036789, "learning_rate": 5.548876990541378e-05, "loss": 0.0062, "step": 39340 }, { "grad_norm": 0.1223529726266861, "learning_rate": 5.5468225975331564e-05, "loss": 0.0091, "step": 39350 }, { "grad_norm": 0.10870279371738434, "learning_rate": 5.544768111087909e-05, "loss": 0.007, "step": 39360 }, { "grad_norm": 0.13021421432495117, "learning_rate": 5.54271353155669e-05, "loss": 0.0063, "step": 39370 }, { "grad_norm": 0.12717847526073456, "learning_rate": 5.5406588592905715e-05, "loss": 0.0064, "step": 39380 }, { "grad_norm": 0.11888089776039124, "learning_rate": 5.5386040946406416e-05, "loss": 0.007, "step": 39390 }, { "grad_norm": 0.1199902668595314, "learning_rate": 5.536549237958004e-05, "loss": 0.0063, "step": 39400 }, { "grad_norm": 0.1114746481180191, "learning_rate": 5.5344942895937744e-05, "loss": 0.0065, "step": 39410 }, { "grad_norm": 0.0935436338186264, "learning_rate": 5.5324392498990904e-05, "loss": 0.0069, "step": 39420 }, { "grad_norm": 0.1311984360218048, "learning_rate": 5.5303841192251e-05, "loss": 0.0075, "step": 39430 }, { "grad_norm": 0.1251956969499588, "learning_rate": 5.52832889792297e-05, "loss": 0.0092, "step": 39440 }, { "grad_norm": 0.07190483808517456, "learning_rate": 5.526273586343881e-05, "loss": 0.0075, "step": 39450 }, { "grad_norm": 0.14243662357330322, "learning_rate": 5.5242181848390306e-05, "loss": 0.0067, "step": 39460 }, { "grad_norm": 0.17193613946437836, "learning_rate": 5.5221626937596285e-05, "loss": 0.0069, "step": 39470 }, { "grad_norm": 0.14958330988883972, "learning_rate": 5.520107113456903e-05, "loss": 0.0065, "step": 39480 }, { "grad_norm": 0.16020077466964722, "learning_rate": 5.5180514442820974e-05, "loss": 0.007, "step": 39490 }, { "grad_norm": 0.13901139795780182, "learning_rate": 5.515995686586469e-05, "loss": 0.0066, "step": 39500 }, { "grad_norm": 0.18909774720668793, "learning_rate": 5.5139398407212916e-05, "loss": 0.0112, "step": 39510 }, { "grad_norm": 0.11538078635931015, "learning_rate": 5.511883907037849e-05, "loss": 0.0081, "step": 39520 }, { "grad_norm": 0.10198120027780533, "learning_rate": 5.509827885887449e-05, "loss": 0.0057, "step": 39530 }, { "grad_norm": 0.11390295624732971, "learning_rate": 5.507771777621406e-05, "loss": 0.0072, "step": 39540 }, { "grad_norm": 0.11063650995492935, "learning_rate": 5.505715582591052e-05, "loss": 0.0066, "step": 39550 }, { "grad_norm": 0.06188827380537987, "learning_rate": 5.50365930114774e-05, "loss": 0.0057, "step": 39560 }, { "grad_norm": 0.11397481709718704, "learning_rate": 5.5016029336428255e-05, "loss": 0.0059, "step": 39570 }, { "grad_norm": 0.1480642706155777, "learning_rate": 5.49954648042769e-05, "loss": 0.0079, "step": 39580 }, { "grad_norm": 0.1944265514612198, "learning_rate": 5.4974899418537226e-05, "loss": 0.0079, "step": 39590 }, { "grad_norm": 0.12102872878313065, "learning_rate": 5.4954333182723316e-05, "loss": 0.0066, "step": 39600 }, { "grad_norm": 0.10728129744529724, "learning_rate": 5.493376610034937e-05, "loss": 0.0074, "step": 39610 }, { "grad_norm": 0.1318259835243225, "learning_rate": 5.4913198174929735e-05, "loss": 0.0051, "step": 39620 }, { "grad_norm": 0.1006425991654396, "learning_rate": 5.489262940997889e-05, "loss": 0.0058, "step": 39630 }, { "grad_norm": 0.0973823070526123, "learning_rate": 5.487205980901151e-05, "loss": 0.0059, "step": 39640 }, { "grad_norm": 0.09465688467025757, "learning_rate": 5.485148937554234e-05, "loss": 0.0055, "step": 39650 }, { "grad_norm": 0.11045150458812714, "learning_rate": 5.483091811308635e-05, "loss": 0.0069, "step": 39660 }, { "grad_norm": 0.13162922859191895, "learning_rate": 5.4810346025158564e-05, "loss": 0.0072, "step": 39670 }, { "grad_norm": 0.12503549456596375, "learning_rate": 5.478977311527421e-05, "loss": 0.0074, "step": 39680 }, { "grad_norm": 0.11626826971769333, "learning_rate": 5.476919938694863e-05, "loss": 0.0075, "step": 39690 }, { "grad_norm": 0.12564995884895325, "learning_rate": 5.474862484369733e-05, "loss": 0.0073, "step": 39700 }, { "grad_norm": 0.1452188640832901, "learning_rate": 5.472804948903589e-05, "loss": 0.0083, "step": 39710 }, { "grad_norm": 0.10972950607538223, "learning_rate": 5.470747332648013e-05, "loss": 0.0073, "step": 39720 }, { "grad_norm": 0.1934821903705597, "learning_rate": 5.468689635954591e-05, "loss": 0.0083, "step": 39730 }, { "grad_norm": 0.14601697027683258, "learning_rate": 5.46663185917493e-05, "loss": 0.0102, "step": 39740 }, { "grad_norm": 0.11509350687265396, "learning_rate": 5.464574002660645e-05, "loss": 0.0056, "step": 39750 }, { "grad_norm": 0.15425072610378265, "learning_rate": 5.4625160667633724e-05, "loss": 0.0071, "step": 39760 }, { "grad_norm": 0.11145369708538055, "learning_rate": 5.4604580518347505e-05, "loss": 0.0064, "step": 39770 }, { "grad_norm": 0.10378362238407135, "learning_rate": 5.458399958226442e-05, "loss": 0.0062, "step": 39780 }, { "grad_norm": 0.12144382297992706, "learning_rate": 5.456341786290119e-05, "loss": 0.0084, "step": 39790 }, { "grad_norm": 0.1140793040394783, "learning_rate": 5.454283536377465e-05, "loss": 0.0069, "step": 39800 }, { "grad_norm": 0.09944137930870056, "learning_rate": 5.452225208840179e-05, "loss": 0.0071, "step": 39810 }, { "grad_norm": 0.1056007444858551, "learning_rate": 5.450166804029975e-05, "loss": 0.0081, "step": 39820 }, { "grad_norm": 0.07626016438007355, "learning_rate": 5.448108322298574e-05, "loss": 0.0067, "step": 39830 }, { "grad_norm": 0.15056106448173523, "learning_rate": 5.446049763997717e-05, "loss": 0.0071, "step": 39840 }, { "grad_norm": 0.11650989949703217, "learning_rate": 5.4439911294791546e-05, "loss": 0.0088, "step": 39850 }, { "grad_norm": 0.12630526721477509, "learning_rate": 5.441932419094652e-05, "loss": 0.0072, "step": 39860 }, { "grad_norm": 0.13953931629657745, "learning_rate": 5.439873633195985e-05, "loss": 0.0068, "step": 39870 }, { "grad_norm": 0.14868222177028656, "learning_rate": 5.437814772134947e-05, "loss": 0.0067, "step": 39880 }, { "grad_norm": 0.11189553141593933, "learning_rate": 5.4357558362633366e-05, "loss": 0.0097, "step": 39890 }, { "grad_norm": 0.12150534242391586, "learning_rate": 5.4336968259329726e-05, "loss": 0.0087, "step": 39900 }, { "grad_norm": 0.15681466460227966, "learning_rate": 5.431637741495681e-05, "loss": 0.0069, "step": 39910 }, { "grad_norm": 0.09876108169555664, "learning_rate": 5.429578583303307e-05, "loss": 0.0085, "step": 39920 }, { "grad_norm": 0.09725185483694077, "learning_rate": 5.427519351707701e-05, "loss": 0.0091, "step": 39930 }, { "grad_norm": 0.10035663098096848, "learning_rate": 5.42546004706073e-05, "loss": 0.0064, "step": 39940 }, { "grad_norm": 0.06573490053415298, "learning_rate": 5.4234006697142735e-05, "loss": 0.0047, "step": 39950 }, { "grad_norm": 0.11373064666986465, "learning_rate": 5.421341220020224e-05, "loss": 0.0075, "step": 39960 }, { "grad_norm": 0.12338899821043015, "learning_rate": 5.419281698330482e-05, "loss": 0.0055, "step": 39970 }, { "grad_norm": 0.08223327994346619, "learning_rate": 5.4172221049969665e-05, "loss": 0.0067, "step": 39980 }, { "grad_norm": 0.1321699172258377, "learning_rate": 5.415162440371604e-05, "loss": 0.0053, "step": 39990 }, { "grad_norm": 0.11506293714046478, "learning_rate": 5.413102704806334e-05, "loss": 0.0059, "step": 40000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 48, "trial_name": null, "trial_params": null }