{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 18.36210062431142, "eval_steps": 500, "global_step": 50000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.003672420124862284, "grad_norm": 370.0125427246094, "learning_rate": 3.6e-07, "loss": 8.8552, "step": 10 }, { "epoch": 0.007344840249724568, "grad_norm": 279.6959533691406, "learning_rate": 7.6e-07, "loss": 8.4789, "step": 20 }, { "epoch": 0.011017260374586852, "grad_norm": 385.6219482421875, "learning_rate": 1.16e-06, "loss": 7.574, "step": 30 }, { "epoch": 0.014689680499449137, "grad_norm": 246.16920471191406, "learning_rate": 1.56e-06, "loss": 6.4503, "step": 40 }, { "epoch": 0.01836210062431142, "grad_norm": 165.62265014648438, "learning_rate": 1.96e-06, "loss": 5.0014, "step": 50 }, { "epoch": 0.022034520749173704, "grad_norm": 151.15565490722656, "learning_rate": 2.36e-06, "loss": 3.7894, "step": 60 }, { "epoch": 0.02570694087403599, "grad_norm": 104.0170669555664, "learning_rate": 2.7600000000000003e-06, "loss": 2.3899, "step": 70 }, { "epoch": 0.029379360998898273, "grad_norm": 50.899200439453125, "learning_rate": 3.1600000000000007e-06, "loss": 1.4482, "step": 80 }, { "epoch": 0.03305178112376056, "grad_norm": 33.97148895263672, "learning_rate": 3.5600000000000002e-06, "loss": 0.8578, "step": 90 }, { "epoch": 0.03672420124862284, "grad_norm": 11.721701622009277, "learning_rate": 3.96e-06, "loss": 0.5774, "step": 100 }, { "epoch": 0.040396621373485125, "grad_norm": 5.8166422843933105, "learning_rate": 4.360000000000001e-06, "loss": 0.4308, "step": 110 }, { "epoch": 0.04406904149834741, "grad_norm": 2.865495443344116, "learning_rate": 4.76e-06, "loss": 0.3456, "step": 120 }, { "epoch": 0.0477414616232097, "grad_norm": 4.062641620635986, "learning_rate": 5.1600000000000006e-06, "loss": 0.314, "step": 130 }, { "epoch": 0.05141388174807198, "grad_norm": 4.831028938293457, "learning_rate": 5.56e-06, "loss": 0.2996, "step": 140 }, { "epoch": 0.055086301872934264, "grad_norm": 3.04677414894104, "learning_rate": 5.9600000000000005e-06, "loss": 0.2671, "step": 150 }, { "epoch": 0.05875872199779655, "grad_norm": 5.66223669052124, "learning_rate": 6.360000000000001e-06, "loss": 0.2483, "step": 160 }, { "epoch": 0.06243114212265883, "grad_norm": 3.0468716621398926, "learning_rate": 6.76e-06, "loss": 0.2168, "step": 170 }, { "epoch": 0.06610356224752112, "grad_norm": 7.4151530265808105, "learning_rate": 7.16e-06, "loss": 0.208, "step": 180 }, { "epoch": 0.0697759823723834, "grad_norm": 2.8374781608581543, "learning_rate": 7.5600000000000005e-06, "loss": 0.1809, "step": 190 }, { "epoch": 0.07344840249724569, "grad_norm": 3.0849416255950928, "learning_rate": 7.96e-06, "loss": 0.1771, "step": 200 }, { "epoch": 0.07712082262210797, "grad_norm": 1.7636420726776123, "learning_rate": 8.36e-06, "loss": 0.1746, "step": 210 }, { "epoch": 0.08079324274697025, "grad_norm": 3.7576780319213867, "learning_rate": 8.76e-06, "loss": 0.1637, "step": 220 }, { "epoch": 0.08446566287183253, "grad_norm": 2.799579381942749, "learning_rate": 9.16e-06, "loss": 0.1587, "step": 230 }, { "epoch": 0.08813808299669482, "grad_norm": 1.6557139158248901, "learning_rate": 9.560000000000002e-06, "loss": 0.152, "step": 240 }, { "epoch": 0.0918105031215571, "grad_norm": 2.005631685256958, "learning_rate": 9.96e-06, "loss": 0.1477, "step": 250 }, { "epoch": 0.0954829232464194, "grad_norm": 1.5119112730026245, "learning_rate": 1.036e-05, "loss": 0.1414, "step": 260 }, { "epoch": 0.09915534337128168, "grad_norm": 2.2495620250701904, "learning_rate": 1.076e-05, "loss": 0.1453, "step": 270 }, { "epoch": 0.10282776349614396, "grad_norm": 1.4313373565673828, "learning_rate": 1.1160000000000002e-05, "loss": 0.1346, "step": 280 }, { "epoch": 0.10650018362100624, "grad_norm": 1.2735499143600464, "learning_rate": 1.156e-05, "loss": 0.1416, "step": 290 }, { "epoch": 0.11017260374586853, "grad_norm": 1.748020052909851, "learning_rate": 1.196e-05, "loss": 0.1272, "step": 300 }, { "epoch": 0.11384502387073081, "grad_norm": 2.5709664821624756, "learning_rate": 1.236e-05, "loss": 0.1346, "step": 310 }, { "epoch": 0.1175174439955931, "grad_norm": 0.9886828660964966, "learning_rate": 1.276e-05, "loss": 0.1254, "step": 320 }, { "epoch": 0.12118986412045538, "grad_norm": 1.5965126752853394, "learning_rate": 1.316e-05, "loss": 0.1215, "step": 330 }, { "epoch": 0.12486228424531766, "grad_norm": 1.0567289590835571, "learning_rate": 1.356e-05, "loss": 0.1197, "step": 340 }, { "epoch": 0.12853470437017994, "grad_norm": 1.4421463012695312, "learning_rate": 1.396e-05, "loss": 0.1143, "step": 350 }, { "epoch": 0.13220712449504224, "grad_norm": 0.8737083673477173, "learning_rate": 1.4360000000000001e-05, "loss": 0.1121, "step": 360 }, { "epoch": 0.1358795446199045, "grad_norm": 0.9682319760322571, "learning_rate": 1.4760000000000001e-05, "loss": 0.1138, "step": 370 }, { "epoch": 0.1395519647447668, "grad_norm": 0.9975892305374146, "learning_rate": 1.5160000000000002e-05, "loss": 0.1118, "step": 380 }, { "epoch": 0.14322438486962907, "grad_norm": 1.0451033115386963, "learning_rate": 1.556e-05, "loss": 0.1139, "step": 390 }, { "epoch": 0.14689680499449137, "grad_norm": 0.9813297390937805, "learning_rate": 1.596e-05, "loss": 0.1071, "step": 400 }, { "epoch": 0.15056922511935367, "grad_norm": 0.9496484398841858, "learning_rate": 1.636e-05, "loss": 0.1107, "step": 410 }, { "epoch": 0.15424164524421594, "grad_norm": 0.8426526188850403, "learning_rate": 1.6760000000000002e-05, "loss": 0.1019, "step": 420 }, { "epoch": 0.15791406536907823, "grad_norm": 0.9939479231834412, "learning_rate": 1.7160000000000002e-05, "loss": 0.1074, "step": 430 }, { "epoch": 0.1615864854939405, "grad_norm": 0.893804669380188, "learning_rate": 1.756e-05, "loss": 0.1071, "step": 440 }, { "epoch": 0.1652589056188028, "grad_norm": 0.7891087532043457, "learning_rate": 1.796e-05, "loss": 0.1052, "step": 450 }, { "epoch": 0.16893132574366507, "grad_norm": 0.8050212264060974, "learning_rate": 1.8360000000000004e-05, "loss": 0.1077, "step": 460 }, { "epoch": 0.17260374586852736, "grad_norm": 0.6606860756874084, "learning_rate": 1.876e-05, "loss": 0.1061, "step": 470 }, { "epoch": 0.17627616599338963, "grad_norm": 0.5819520354270935, "learning_rate": 1.916e-05, "loss": 0.1004, "step": 480 }, { "epoch": 0.17994858611825193, "grad_norm": 0.5754823088645935, "learning_rate": 1.956e-05, "loss": 0.0977, "step": 490 }, { "epoch": 0.1836210062431142, "grad_norm": 0.6741044521331787, "learning_rate": 1.9960000000000002e-05, "loss": 0.0956, "step": 500 }, { "epoch": 0.1872934263679765, "grad_norm": 0.7368530035018921, "learning_rate": 2.036e-05, "loss": 0.0985, "step": 510 }, { "epoch": 0.1909658464928388, "grad_norm": 1.554929256439209, "learning_rate": 2.076e-05, "loss": 0.0956, "step": 520 }, { "epoch": 0.19463826661770106, "grad_norm": 0.7614655494689941, "learning_rate": 2.116e-05, "loss": 0.0952, "step": 530 }, { "epoch": 0.19831068674256336, "grad_norm": 0.6841129660606384, "learning_rate": 2.1560000000000004e-05, "loss": 0.0954, "step": 540 }, { "epoch": 0.20198310686742563, "grad_norm": 0.567664384841919, "learning_rate": 2.196e-05, "loss": 0.0906, "step": 550 }, { "epoch": 0.20565552699228792, "grad_norm": 0.615355372428894, "learning_rate": 2.236e-05, "loss": 0.0902, "step": 560 }, { "epoch": 0.2093279471171502, "grad_norm": 0.7552409768104553, "learning_rate": 2.2760000000000002e-05, "loss": 0.0891, "step": 570 }, { "epoch": 0.2130003672420125, "grad_norm": 0.5861403346061707, "learning_rate": 2.3160000000000002e-05, "loss": 0.0891, "step": 580 }, { "epoch": 0.21667278736687476, "grad_norm": 0.8447241187095642, "learning_rate": 2.356e-05, "loss": 0.0914, "step": 590 }, { "epoch": 0.22034520749173706, "grad_norm": 0.516206681728363, "learning_rate": 2.396e-05, "loss": 0.0936, "step": 600 }, { "epoch": 0.22401762761659935, "grad_norm": 0.8950227499008179, "learning_rate": 2.4360000000000004e-05, "loss": 0.0934, "step": 610 }, { "epoch": 0.22769004774146162, "grad_norm": 0.5737316608428955, "learning_rate": 2.476e-05, "loss": 0.0832, "step": 620 }, { "epoch": 0.23136246786632392, "grad_norm": 0.6093111634254456, "learning_rate": 2.516e-05, "loss": 0.0843, "step": 630 }, { "epoch": 0.2350348879911862, "grad_norm": 0.8191360831260681, "learning_rate": 2.556e-05, "loss": 0.0896, "step": 640 }, { "epoch": 0.23870730811604848, "grad_norm": 0.7658595442771912, "learning_rate": 2.5960000000000002e-05, "loss": 0.0859, "step": 650 }, { "epoch": 0.24237972824091075, "grad_norm": 0.6534927487373352, "learning_rate": 2.6360000000000002e-05, "loss": 0.0853, "step": 660 }, { "epoch": 0.24605214836577305, "grad_norm": 0.6944789886474609, "learning_rate": 2.676e-05, "loss": 0.0928, "step": 670 }, { "epoch": 0.24972456849063532, "grad_norm": 0.598307728767395, "learning_rate": 2.716e-05, "loss": 0.0892, "step": 680 }, { "epoch": 0.2533969886154976, "grad_norm": 0.8234556913375854, "learning_rate": 2.7560000000000004e-05, "loss": 0.0861, "step": 690 }, { "epoch": 0.2570694087403599, "grad_norm": 0.6253828406333923, "learning_rate": 2.7960000000000003e-05, "loss": 0.0843, "step": 700 }, { "epoch": 0.2607418288652222, "grad_norm": 1.0761011838912964, "learning_rate": 2.8360000000000003e-05, "loss": 0.0888, "step": 710 }, { "epoch": 0.2644142489900845, "grad_norm": 0.7511357069015503, "learning_rate": 2.8760000000000002e-05, "loss": 0.0833, "step": 720 }, { "epoch": 0.2680866691149468, "grad_norm": 0.6557642221450806, "learning_rate": 2.9160000000000005e-05, "loss": 0.0852, "step": 730 }, { "epoch": 0.271759089239809, "grad_norm": 0.43899253010749817, "learning_rate": 2.9559999999999998e-05, "loss": 0.0771, "step": 740 }, { "epoch": 0.2754315093646713, "grad_norm": 0.6799081563949585, "learning_rate": 2.9959999999999998e-05, "loss": 0.0857, "step": 750 }, { "epoch": 0.2791039294895336, "grad_norm": 0.9090906381607056, "learning_rate": 3.036e-05, "loss": 0.0812, "step": 760 }, { "epoch": 0.2827763496143959, "grad_norm": 1.3720818758010864, "learning_rate": 3.076e-05, "loss": 0.0837, "step": 770 }, { "epoch": 0.28644876973925815, "grad_norm": 0.6108842492103577, "learning_rate": 3.116e-05, "loss": 0.0839, "step": 780 }, { "epoch": 0.29012118986412044, "grad_norm": 0.6732390522956848, "learning_rate": 3.156e-05, "loss": 0.0822, "step": 790 }, { "epoch": 0.29379360998898274, "grad_norm": 0.5214940905570984, "learning_rate": 3.196e-05, "loss": 0.0814, "step": 800 }, { "epoch": 0.29746603011384504, "grad_norm": 0.7576974034309387, "learning_rate": 3.236e-05, "loss": 0.0836, "step": 810 }, { "epoch": 0.30113845023870733, "grad_norm": 0.6668350100517273, "learning_rate": 3.2760000000000005e-05, "loss": 0.0832, "step": 820 }, { "epoch": 0.3048108703635696, "grad_norm": 0.9352800250053406, "learning_rate": 3.316e-05, "loss": 0.0842, "step": 830 }, { "epoch": 0.30848329048843187, "grad_norm": 0.6005798578262329, "learning_rate": 3.3560000000000004e-05, "loss": 0.0803, "step": 840 }, { "epoch": 0.31215571061329417, "grad_norm": 0.6044943332672119, "learning_rate": 3.396e-05, "loss": 0.0833, "step": 850 }, { "epoch": 0.31582813073815647, "grad_norm": 0.5753768682479858, "learning_rate": 3.436e-05, "loss": 0.0849, "step": 860 }, { "epoch": 0.3195005508630187, "grad_norm": 0.559956967830658, "learning_rate": 3.4760000000000006e-05, "loss": 0.0788, "step": 870 }, { "epoch": 0.323172970987881, "grad_norm": 0.6040977239608765, "learning_rate": 3.516e-05, "loss": 0.0764, "step": 880 }, { "epoch": 0.3268453911127433, "grad_norm": 0.8212110996246338, "learning_rate": 3.5560000000000005e-05, "loss": 0.0743, "step": 890 }, { "epoch": 0.3305178112376056, "grad_norm": 0.5575911402702332, "learning_rate": 3.596e-05, "loss": 0.0783, "step": 900 }, { "epoch": 0.3341902313624679, "grad_norm": 0.5613875389099121, "learning_rate": 3.636e-05, "loss": 0.0715, "step": 910 }, { "epoch": 0.33786265148733013, "grad_norm": 1.0385085344314575, "learning_rate": 3.676e-05, "loss": 0.0798, "step": 920 }, { "epoch": 0.34153507161219243, "grad_norm": 0.49407443404197693, "learning_rate": 3.716e-05, "loss": 0.0767, "step": 930 }, { "epoch": 0.34520749173705473, "grad_norm": 0.5242489576339722, "learning_rate": 3.756e-05, "loss": 0.0769, "step": 940 }, { "epoch": 0.348879911861917, "grad_norm": 0.6195166110992432, "learning_rate": 3.796e-05, "loss": 0.0805, "step": 950 }, { "epoch": 0.35255233198677927, "grad_norm": 0.5781774520874023, "learning_rate": 3.836e-05, "loss": 0.0707, "step": 960 }, { "epoch": 0.35622475211164156, "grad_norm": 0.6739203929901123, "learning_rate": 3.876e-05, "loss": 0.0789, "step": 970 }, { "epoch": 0.35989717223650386, "grad_norm": 0.6506828665733337, "learning_rate": 3.9160000000000005e-05, "loss": 0.0769, "step": 980 }, { "epoch": 0.36356959236136616, "grad_norm": 0.6613247990608215, "learning_rate": 3.956e-05, "loss": 0.0769, "step": 990 }, { "epoch": 0.3672420124862284, "grad_norm": 0.40287014842033386, "learning_rate": 3.9960000000000004e-05, "loss": 0.0708, "step": 1000 }, { "epoch": 0.3709144326110907, "grad_norm": 0.6262238025665283, "learning_rate": 4.0360000000000007e-05, "loss": 0.0746, "step": 1010 }, { "epoch": 0.374586852735953, "grad_norm": 1.4949642419815063, "learning_rate": 4.076e-05, "loss": 0.0755, "step": 1020 }, { "epoch": 0.3782592728608153, "grad_norm": 0.7433282136917114, "learning_rate": 4.1160000000000006e-05, "loss": 0.0744, "step": 1030 }, { "epoch": 0.3819316929856776, "grad_norm": 0.49738115072250366, "learning_rate": 4.156e-05, "loss": 0.0754, "step": 1040 }, { "epoch": 0.3856041131105398, "grad_norm": 0.6211096048355103, "learning_rate": 4.196e-05, "loss": 0.0699, "step": 1050 }, { "epoch": 0.3892765332354021, "grad_norm": 0.6418870687484741, "learning_rate": 4.236e-05, "loss": 0.0801, "step": 1060 }, { "epoch": 0.3929489533602644, "grad_norm": 0.7288879156112671, "learning_rate": 4.276e-05, "loss": 0.0722, "step": 1070 }, { "epoch": 0.3966213734851267, "grad_norm": 0.6092147827148438, "learning_rate": 4.316e-05, "loss": 0.075, "step": 1080 }, { "epoch": 0.40029379360998896, "grad_norm": 0.633902370929718, "learning_rate": 4.356e-05, "loss": 0.0699, "step": 1090 }, { "epoch": 0.40396621373485125, "grad_norm": 0.44581177830696106, "learning_rate": 4.396e-05, "loss": 0.075, "step": 1100 }, { "epoch": 0.40763863385971355, "grad_norm": 0.5004166960716248, "learning_rate": 4.436e-05, "loss": 0.0752, "step": 1110 }, { "epoch": 0.41131105398457585, "grad_norm": 0.558629035949707, "learning_rate": 4.4760000000000005e-05, "loss": 0.0744, "step": 1120 }, { "epoch": 0.41498347410943814, "grad_norm": 0.5899764895439148, "learning_rate": 4.516e-05, "loss": 0.0736, "step": 1130 }, { "epoch": 0.4186558942343004, "grad_norm": 0.4998798668384552, "learning_rate": 4.5560000000000004e-05, "loss": 0.0762, "step": 1140 }, { "epoch": 0.4223283143591627, "grad_norm": 0.47848886251449585, "learning_rate": 4.596e-05, "loss": 0.0726, "step": 1150 }, { "epoch": 0.426000734484025, "grad_norm": 0.7048205733299255, "learning_rate": 4.636e-05, "loss": 0.0725, "step": 1160 }, { "epoch": 0.4296731546088873, "grad_norm": 0.5359187722206116, "learning_rate": 4.6760000000000006e-05, "loss": 0.073, "step": 1170 }, { "epoch": 0.4333455747337495, "grad_norm": 0.48758548498153687, "learning_rate": 4.716e-05, "loss": 0.0731, "step": 1180 }, { "epoch": 0.4370179948586118, "grad_norm": 0.5292090177536011, "learning_rate": 4.7560000000000005e-05, "loss": 0.073, "step": 1190 }, { "epoch": 0.4406904149834741, "grad_norm": 0.4330541789531708, "learning_rate": 4.796e-05, "loss": 0.071, "step": 1200 }, { "epoch": 0.4443628351083364, "grad_norm": 0.6004024147987366, "learning_rate": 4.836e-05, "loss": 0.0713, "step": 1210 }, { "epoch": 0.4480352552331987, "grad_norm": 0.7866067290306091, "learning_rate": 4.876e-05, "loss": 0.073, "step": 1220 }, { "epoch": 0.45170767535806094, "grad_norm": 0.7243216633796692, "learning_rate": 4.9160000000000004e-05, "loss": 0.0713, "step": 1230 }, { "epoch": 0.45538009548292324, "grad_norm": 1.1026352643966675, "learning_rate": 4.956e-05, "loss": 0.0705, "step": 1240 }, { "epoch": 0.45905251560778554, "grad_norm": 0.9547515511512756, "learning_rate": 4.996e-05, "loss": 0.0777, "step": 1250 }, { "epoch": 0.46272493573264784, "grad_norm": 1.5530067682266235, "learning_rate": 5.0360000000000006e-05, "loss": 0.0732, "step": 1260 }, { "epoch": 0.4663973558575101, "grad_norm": 0.6853833198547363, "learning_rate": 5.076000000000001e-05, "loss": 0.0761, "step": 1270 }, { "epoch": 0.4700697759823724, "grad_norm": 0.5234474539756775, "learning_rate": 5.1160000000000005e-05, "loss": 0.0748, "step": 1280 }, { "epoch": 0.47374219610723467, "grad_norm": 0.5538917779922485, "learning_rate": 5.1559999999999994e-05, "loss": 0.0772, "step": 1290 }, { "epoch": 0.47741461623209697, "grad_norm": 0.7935733199119568, "learning_rate": 5.196e-05, "loss": 0.0694, "step": 1300 }, { "epoch": 0.48108703635695926, "grad_norm": 0.579271674156189, "learning_rate": 5.236e-05, "loss": 0.0665, "step": 1310 }, { "epoch": 0.4847594564818215, "grad_norm": 0.43350937962532043, "learning_rate": 5.2759999999999996e-05, "loss": 0.0733, "step": 1320 }, { "epoch": 0.4884318766066838, "grad_norm": 0.6313692927360535, "learning_rate": 5.316e-05, "loss": 0.067, "step": 1330 }, { "epoch": 0.4921042967315461, "grad_norm": 0.4949837028980255, "learning_rate": 5.356e-05, "loss": 0.0696, "step": 1340 }, { "epoch": 0.4957767168564084, "grad_norm": 0.442827045917511, "learning_rate": 5.396e-05, "loss": 0.0639, "step": 1350 }, { "epoch": 0.49944913698127064, "grad_norm": 0.5162312388420105, "learning_rate": 5.436e-05, "loss": 0.0669, "step": 1360 }, { "epoch": 0.5031215571061329, "grad_norm": 0.53321373462677, "learning_rate": 5.476e-05, "loss": 0.0665, "step": 1370 }, { "epoch": 0.5067939772309952, "grad_norm": 0.4336065649986267, "learning_rate": 5.516e-05, "loss": 0.0673, "step": 1380 }, { "epoch": 0.5104663973558575, "grad_norm": 0.5260268449783325, "learning_rate": 5.556e-05, "loss": 0.062, "step": 1390 }, { "epoch": 0.5141388174807198, "grad_norm": 0.49168410897254944, "learning_rate": 5.596e-05, "loss": 0.0695, "step": 1400 }, { "epoch": 0.5178112376055821, "grad_norm": 0.8377885222434998, "learning_rate": 5.636e-05, "loss": 0.0636, "step": 1410 }, { "epoch": 0.5214836577304444, "grad_norm": 0.49504438042640686, "learning_rate": 5.6760000000000005e-05, "loss": 0.075, "step": 1420 }, { "epoch": 0.5251560778553066, "grad_norm": 0.7498390674591064, "learning_rate": 5.716e-05, "loss": 0.067, "step": 1430 }, { "epoch": 0.528828497980169, "grad_norm": 0.5824487805366516, "learning_rate": 5.7560000000000005e-05, "loss": 0.0682, "step": 1440 }, { "epoch": 0.5325009181050312, "grad_norm": 0.4062468409538269, "learning_rate": 5.796e-05, "loss": 0.0696, "step": 1450 }, { "epoch": 0.5361733382298935, "grad_norm": 0.37125566601753235, "learning_rate": 5.8360000000000004e-05, "loss": 0.0705, "step": 1460 }, { "epoch": 0.5398457583547558, "grad_norm": 0.5642731785774231, "learning_rate": 5.876000000000001e-05, "loss": 0.0696, "step": 1470 }, { "epoch": 0.543518178479618, "grad_norm": 0.529525101184845, "learning_rate": 5.916e-05, "loss": 0.0681, "step": 1480 }, { "epoch": 0.5471905986044804, "grad_norm": 0.3904850482940674, "learning_rate": 5.9560000000000006e-05, "loss": 0.066, "step": 1490 }, { "epoch": 0.5508630187293426, "grad_norm": 0.6852251291275024, "learning_rate": 5.996e-05, "loss": 0.0741, "step": 1500 }, { "epoch": 0.554535438854205, "grad_norm": 0.564313530921936, "learning_rate": 6.0360000000000005e-05, "loss": 0.0683, "step": 1510 }, { "epoch": 0.5582078589790672, "grad_norm": 0.47664201259613037, "learning_rate": 6.076000000000001e-05, "loss": 0.0705, "step": 1520 }, { "epoch": 0.5618802791039295, "grad_norm": 0.45304667949676514, "learning_rate": 6.116e-05, "loss": 0.0668, "step": 1530 }, { "epoch": 0.5655526992287918, "grad_norm": 0.4342691898345947, "learning_rate": 6.156e-05, "loss": 0.072, "step": 1540 }, { "epoch": 0.569225119353654, "grad_norm": 0.7400838732719421, "learning_rate": 6.196000000000001e-05, "loss": 0.0687, "step": 1550 }, { "epoch": 0.5728975394785163, "grad_norm": 0.46706119179725647, "learning_rate": 6.236e-05, "loss": 0.0684, "step": 1560 }, { "epoch": 0.5765699596033786, "grad_norm": 0.4470353126525879, "learning_rate": 6.276e-05, "loss": 0.0627, "step": 1570 }, { "epoch": 0.5802423797282409, "grad_norm": 0.3102373778820038, "learning_rate": 6.316000000000001e-05, "loss": 0.0652, "step": 1580 }, { "epoch": 0.5839147998531032, "grad_norm": 0.437272310256958, "learning_rate": 6.356000000000001e-05, "loss": 0.0675, "step": 1590 }, { "epoch": 0.5875872199779655, "grad_norm": 0.41681209206581116, "learning_rate": 6.396e-05, "loss": 0.0641, "step": 1600 }, { "epoch": 0.5912596401028277, "grad_norm": 0.4436466097831726, "learning_rate": 6.436e-05, "loss": 0.0654, "step": 1610 }, { "epoch": 0.5949320602276901, "grad_norm": 0.6066335439682007, "learning_rate": 6.476e-05, "loss": 0.0641, "step": 1620 }, { "epoch": 0.5986044803525523, "grad_norm": 0.5422555208206177, "learning_rate": 6.515999999999999e-05, "loss": 0.0648, "step": 1630 }, { "epoch": 0.6022769004774147, "grad_norm": 0.5640207529067993, "learning_rate": 6.556e-05, "loss": 0.0666, "step": 1640 }, { "epoch": 0.6059493206022769, "grad_norm": 0.36210939288139343, "learning_rate": 6.596e-05, "loss": 0.0695, "step": 1650 }, { "epoch": 0.6096217407271391, "grad_norm": 0.4840050935745239, "learning_rate": 6.636e-05, "loss": 0.0595, "step": 1660 }, { "epoch": 0.6132941608520015, "grad_norm": 0.38385656476020813, "learning_rate": 6.676e-05, "loss": 0.0636, "step": 1670 }, { "epoch": 0.6169665809768637, "grad_norm": 0.4342021048069, "learning_rate": 6.716e-05, "loss": 0.0672, "step": 1680 }, { "epoch": 0.620639001101726, "grad_norm": 0.4940832555294037, "learning_rate": 6.756e-05, "loss": 0.0655, "step": 1690 }, { "epoch": 0.6243114212265883, "grad_norm": 0.5161705613136292, "learning_rate": 6.796e-05, "loss": 0.0659, "step": 1700 }, { "epoch": 0.6279838413514506, "grad_norm": 0.46779993176460266, "learning_rate": 6.836e-05, "loss": 0.0634, "step": 1710 }, { "epoch": 0.6316562614763129, "grad_norm": 0.3866766691207886, "learning_rate": 6.876e-05, "loss": 0.0671, "step": 1720 }, { "epoch": 0.6353286816011752, "grad_norm": 0.7547004222869873, "learning_rate": 6.916000000000001e-05, "loss": 0.063, "step": 1730 }, { "epoch": 0.6390011017260374, "grad_norm": 0.47922950983047485, "learning_rate": 6.956e-05, "loss": 0.0638, "step": 1740 }, { "epoch": 0.6426735218508998, "grad_norm": 0.35258224606513977, "learning_rate": 6.996e-05, "loss": 0.0647, "step": 1750 }, { "epoch": 0.646345941975762, "grad_norm": 0.395107626914978, "learning_rate": 7.036e-05, "loss": 0.0661, "step": 1760 }, { "epoch": 0.6500183621006244, "grad_norm": 0.4494529068470001, "learning_rate": 7.076000000000001e-05, "loss": 0.0699, "step": 1770 }, { "epoch": 0.6536907822254866, "grad_norm": 0.3747129738330841, "learning_rate": 7.116e-05, "loss": 0.0636, "step": 1780 }, { "epoch": 0.6573632023503488, "grad_norm": 0.3336876928806305, "learning_rate": 7.156e-05, "loss": 0.0723, "step": 1790 }, { "epoch": 0.6610356224752112, "grad_norm": 0.39421841502189636, "learning_rate": 7.196000000000001e-05, "loss": 0.0638, "step": 1800 }, { "epoch": 0.6647080426000734, "grad_norm": 0.41234302520751953, "learning_rate": 7.236e-05, "loss": 0.0641, "step": 1810 }, { "epoch": 0.6683804627249358, "grad_norm": 0.3457406759262085, "learning_rate": 7.276e-05, "loss": 0.0632, "step": 1820 }, { "epoch": 0.672052882849798, "grad_norm": 0.41627633571624756, "learning_rate": 7.316000000000001e-05, "loss": 0.0622, "step": 1830 }, { "epoch": 0.6757253029746603, "grad_norm": 0.44003528356552124, "learning_rate": 7.356000000000001e-05, "loss": 0.0629, "step": 1840 }, { "epoch": 0.6793977230995226, "grad_norm": 0.45657244324684143, "learning_rate": 7.396e-05, "loss": 0.062, "step": 1850 }, { "epoch": 0.6830701432243849, "grad_norm": 0.3303603231906891, "learning_rate": 7.436000000000001e-05, "loss": 0.0653, "step": 1860 }, { "epoch": 0.6867425633492471, "grad_norm": 0.8317999243736267, "learning_rate": 7.476000000000001e-05, "loss": 0.0649, "step": 1870 }, { "epoch": 0.6904149834741095, "grad_norm": 0.9170985221862793, "learning_rate": 7.516e-05, "loss": 0.0648, "step": 1880 }, { "epoch": 0.6940874035989717, "grad_norm": 0.3913446366786957, "learning_rate": 7.556000000000002e-05, "loss": 0.0633, "step": 1890 }, { "epoch": 0.697759823723834, "grad_norm": 0.36745911836624146, "learning_rate": 7.596000000000001e-05, "loss": 0.0687, "step": 1900 }, { "epoch": 0.7014322438486963, "grad_norm": 0.4617859125137329, "learning_rate": 7.636e-05, "loss": 0.0654, "step": 1910 }, { "epoch": 0.7051046639735585, "grad_norm": 0.583976686000824, "learning_rate": 7.676e-05, "loss": 0.0638, "step": 1920 }, { "epoch": 0.7087770840984209, "grad_norm": 0.444866806268692, "learning_rate": 7.716e-05, "loss": 0.0673, "step": 1930 }, { "epoch": 0.7124495042232831, "grad_norm": 0.4594016373157501, "learning_rate": 7.756e-05, "loss": 0.0651, "step": 1940 }, { "epoch": 0.7161219243481455, "grad_norm": 0.4946763515472412, "learning_rate": 7.796e-05, "loss": 0.0645, "step": 1950 }, { "epoch": 0.7197943444730077, "grad_norm": 0.5908123254776001, "learning_rate": 7.836e-05, "loss": 0.0608, "step": 1960 }, { "epoch": 0.72346676459787, "grad_norm": 0.5545368194580078, "learning_rate": 7.876e-05, "loss": 0.0653, "step": 1970 }, { "epoch": 0.7271391847227323, "grad_norm": 0.5174872875213623, "learning_rate": 7.916e-05, "loss": 0.0684, "step": 1980 }, { "epoch": 0.7308116048475946, "grad_norm": 0.38631749153137207, "learning_rate": 7.956e-05, "loss": 0.0616, "step": 1990 }, { "epoch": 0.7344840249724568, "grad_norm": 0.5095217823982239, "learning_rate": 7.996e-05, "loss": 0.0616, "step": 2000 }, { "epoch": 0.7381564450973191, "grad_norm": 0.5413455367088318, "learning_rate": 8.036e-05, "loss": 0.0594, "step": 2010 }, { "epoch": 0.7418288652221814, "grad_norm": 0.5752138495445251, "learning_rate": 8.076e-05, "loss": 0.0632, "step": 2020 }, { "epoch": 0.7455012853470437, "grad_norm": 0.9135099053382874, "learning_rate": 8.116e-05, "loss": 0.0622, "step": 2030 }, { "epoch": 0.749173705471906, "grad_norm": 0.41529956459999084, "learning_rate": 8.156e-05, "loss": 0.0675, "step": 2040 }, { "epoch": 0.7528461255967682, "grad_norm": 0.40829044580459595, "learning_rate": 8.196000000000001e-05, "loss": 0.0637, "step": 2050 }, { "epoch": 0.7565185457216306, "grad_norm": 0.47972410917282104, "learning_rate": 8.236e-05, "loss": 0.0585, "step": 2060 }, { "epoch": 0.7601909658464928, "grad_norm": 0.7207258343696594, "learning_rate": 8.276e-05, "loss": 0.0624, "step": 2070 }, { "epoch": 0.7638633859713552, "grad_norm": 0.42903438210487366, "learning_rate": 8.316000000000001e-05, "loss": 0.0628, "step": 2080 }, { "epoch": 0.7675358060962174, "grad_norm": 0.6668907999992371, "learning_rate": 8.356e-05, "loss": 0.0584, "step": 2090 }, { "epoch": 0.7712082262210797, "grad_norm": 0.5531400442123413, "learning_rate": 8.396e-05, "loss": 0.0582, "step": 2100 }, { "epoch": 0.774880646345942, "grad_norm": 0.4107942581176758, "learning_rate": 8.436000000000001e-05, "loss": 0.064, "step": 2110 }, { "epoch": 0.7785530664708042, "grad_norm": 0.27888375520706177, "learning_rate": 8.476000000000001e-05, "loss": 0.0585, "step": 2120 }, { "epoch": 0.7822254865956666, "grad_norm": 0.47458401322364807, "learning_rate": 8.516e-05, "loss": 0.0606, "step": 2130 }, { "epoch": 0.7858979067205288, "grad_norm": 0.3787247836589813, "learning_rate": 8.556e-05, "loss": 0.0591, "step": 2140 }, { "epoch": 0.7895703268453911, "grad_norm": 0.3813612461090088, "learning_rate": 8.596000000000001e-05, "loss": 0.0631, "step": 2150 }, { "epoch": 0.7932427469702534, "grad_norm": 0.4133201241493225, "learning_rate": 8.636e-05, "loss": 0.059, "step": 2160 }, { "epoch": 0.7969151670951157, "grad_norm": 0.4017906188964844, "learning_rate": 8.676e-05, "loss": 0.0593, "step": 2170 }, { "epoch": 0.8005875872199779, "grad_norm": 0.6113052368164062, "learning_rate": 8.716000000000001e-05, "loss": 0.0624, "step": 2180 }, { "epoch": 0.8042600073448403, "grad_norm": 0.4754387438297272, "learning_rate": 8.756000000000001e-05, "loss": 0.0598, "step": 2190 }, { "epoch": 0.8079324274697025, "grad_norm": 0.4002903401851654, "learning_rate": 8.796e-05, "loss": 0.0614, "step": 2200 }, { "epoch": 0.8116048475945649, "grad_norm": 0.36806583404541016, "learning_rate": 8.836000000000001e-05, "loss": 0.0597, "step": 2210 }, { "epoch": 0.8152772677194271, "grad_norm": 0.29346638917922974, "learning_rate": 8.876e-05, "loss": 0.0574, "step": 2220 }, { "epoch": 0.8189496878442893, "grad_norm": 0.3444065451622009, "learning_rate": 8.916e-05, "loss": 0.061, "step": 2230 }, { "epoch": 0.8226221079691517, "grad_norm": 0.30905765295028687, "learning_rate": 8.956e-05, "loss": 0.0578, "step": 2240 }, { "epoch": 0.8262945280940139, "grad_norm": 0.2858182191848755, "learning_rate": 8.996e-05, "loss": 0.0612, "step": 2250 }, { "epoch": 0.8299669482188763, "grad_norm": 0.44018983840942383, "learning_rate": 9.036e-05, "loss": 0.0613, "step": 2260 }, { "epoch": 0.8336393683437385, "grad_norm": 0.3946186304092407, "learning_rate": 9.076e-05, "loss": 0.0599, "step": 2270 }, { "epoch": 0.8373117884686008, "grad_norm": 0.4004325568675995, "learning_rate": 9.116e-05, "loss": 0.0642, "step": 2280 }, { "epoch": 0.8409842085934631, "grad_norm": 0.775027871131897, "learning_rate": 9.156e-05, "loss": 0.0589, "step": 2290 }, { "epoch": 0.8446566287183254, "grad_norm": 0.5084565281867981, "learning_rate": 9.196000000000001e-05, "loss": 0.0576, "step": 2300 }, { "epoch": 0.8483290488431876, "grad_norm": 0.3214609920978546, "learning_rate": 9.236e-05, "loss": 0.0619, "step": 2310 }, { "epoch": 0.85200146896805, "grad_norm": 0.4125312268733978, "learning_rate": 9.276e-05, "loss": 0.0583, "step": 2320 }, { "epoch": 0.8556738890929122, "grad_norm": 0.4062417149543762, "learning_rate": 9.316000000000001e-05, "loss": 0.0671, "step": 2330 }, { "epoch": 0.8593463092177746, "grad_norm": 0.3425707221031189, "learning_rate": 9.356e-05, "loss": 0.0606, "step": 2340 }, { "epoch": 0.8630187293426368, "grad_norm": 0.48180392384529114, "learning_rate": 9.396e-05, "loss": 0.064, "step": 2350 }, { "epoch": 0.866691149467499, "grad_norm": 0.5295592546463013, "learning_rate": 9.436e-05, "loss": 0.0593, "step": 2360 }, { "epoch": 0.8703635695923614, "grad_norm": 0.40669170022010803, "learning_rate": 9.476000000000001e-05, "loss": 0.0596, "step": 2370 }, { "epoch": 0.8740359897172236, "grad_norm": 0.43409085273742676, "learning_rate": 9.516e-05, "loss": 0.0613, "step": 2380 }, { "epoch": 0.877708409842086, "grad_norm": 0.5296458005905151, "learning_rate": 9.556e-05, "loss": 0.0603, "step": 2390 }, { "epoch": 0.8813808299669482, "grad_norm": 0.34532302618026733, "learning_rate": 9.596000000000001e-05, "loss": 0.0614, "step": 2400 }, { "epoch": 0.8850532500918105, "grad_norm": 0.4645163416862488, "learning_rate": 9.636e-05, "loss": 0.0591, "step": 2410 }, { "epoch": 0.8887256702166728, "grad_norm": 0.3567269444465637, "learning_rate": 9.676e-05, "loss": 0.0569, "step": 2420 }, { "epoch": 0.892398090341535, "grad_norm": 0.4417807459831238, "learning_rate": 9.716000000000001e-05, "loss": 0.0585, "step": 2430 }, { "epoch": 0.8960705104663974, "grad_norm": 0.3409212827682495, "learning_rate": 9.756000000000001e-05, "loss": 0.0583, "step": 2440 }, { "epoch": 0.8997429305912596, "grad_norm": 0.375993937253952, "learning_rate": 9.796e-05, "loss": 0.0574, "step": 2450 }, { "epoch": 0.9034153507161219, "grad_norm": 0.4816308915615082, "learning_rate": 9.836000000000001e-05, "loss": 0.0615, "step": 2460 }, { "epoch": 0.9070877708409842, "grad_norm": 0.35506242513656616, "learning_rate": 9.876000000000001e-05, "loss": 0.0587, "step": 2470 }, { "epoch": 0.9107601909658465, "grad_norm": 0.3360303044319153, "learning_rate": 9.916e-05, "loss": 0.0607, "step": 2480 }, { "epoch": 0.9144326110907087, "grad_norm": 0.6185207366943359, "learning_rate": 9.956e-05, "loss": 0.0552, "step": 2490 }, { "epoch": 0.9181050312155711, "grad_norm": 0.3160017132759094, "learning_rate": 9.996000000000001e-05, "loss": 0.0586, "step": 2500 }, { "epoch": 0.9217774513404333, "grad_norm": 0.8321806788444519, "learning_rate": 9.999999114196196e-05, "loss": 0.0593, "step": 2510 }, { "epoch": 0.9254498714652957, "grad_norm": 0.4675653576850891, "learning_rate": 9.99999605215876e-05, "loss": 0.06, "step": 2520 }, { "epoch": 0.9291222915901579, "grad_norm": 0.4486774206161499, "learning_rate": 9.999990802953179e-05, "loss": 0.0571, "step": 2530 }, { "epoch": 0.9327947117150202, "grad_norm": 0.4508066177368164, "learning_rate": 9.99998336658175e-05, "loss": 0.0605, "step": 2540 }, { "epoch": 0.9364671318398825, "grad_norm": 0.36382028460502625, "learning_rate": 9.999973743047727e-05, "loss": 0.0605, "step": 2550 }, { "epoch": 0.9401395519647447, "grad_norm": 0.45023927092552185, "learning_rate": 9.999961932355319e-05, "loss": 0.0584, "step": 2560 }, { "epoch": 0.9438119720896071, "grad_norm": 0.4253351390361786, "learning_rate": 9.999947934509693e-05, "loss": 0.0604, "step": 2570 }, { "epoch": 0.9474843922144693, "grad_norm": 0.41170790791511536, "learning_rate": 9.999931749516971e-05, "loss": 0.0609, "step": 2580 }, { "epoch": 0.9511568123393316, "grad_norm": 0.42087557911872864, "learning_rate": 9.999913377384233e-05, "loss": 0.0585, "step": 2590 }, { "epoch": 0.9548292324641939, "grad_norm": 0.5364667177200317, "learning_rate": 9.999892818119517e-05, "loss": 0.0607, "step": 2600 }, { "epoch": 0.9585016525890562, "grad_norm": 0.5286917686462402, "learning_rate": 9.999870071731814e-05, "loss": 0.0565, "step": 2610 }, { "epoch": 0.9621740727139185, "grad_norm": 0.3182532787322998, "learning_rate": 9.999845138231076e-05, "loss": 0.0575, "step": 2620 }, { "epoch": 0.9658464928387808, "grad_norm": 0.34501659870147705, "learning_rate": 9.999818017628208e-05, "loss": 0.0554, "step": 2630 }, { "epoch": 0.969518912963643, "grad_norm": 0.43450620770454407, "learning_rate": 9.999788709935078e-05, "loss": 0.0554, "step": 2640 }, { "epoch": 0.9731913330885054, "grad_norm": 0.4685603380203247, "learning_rate": 9.9997572151645e-05, "loss": 0.0602, "step": 2650 }, { "epoch": 0.9768637532133676, "grad_norm": 0.44564858078956604, "learning_rate": 9.999723533330254e-05, "loss": 0.0618, "step": 2660 }, { "epoch": 0.9805361733382298, "grad_norm": 0.3223062753677368, "learning_rate": 9.999687664447074e-05, "loss": 0.0562, "step": 2670 }, { "epoch": 0.9842085934630922, "grad_norm": 0.36951348185539246, "learning_rate": 9.99964960853065e-05, "loss": 0.0614, "step": 2680 }, { "epoch": 0.9878810135879544, "grad_norm": 0.25344952940940857, "learning_rate": 9.999609365597627e-05, "loss": 0.0549, "step": 2690 }, { "epoch": 0.9915534337128168, "grad_norm": 0.453714519739151, "learning_rate": 9.99956693566561e-05, "loss": 0.0619, "step": 2700 }, { "epoch": 0.995225853837679, "grad_norm": 0.32601022720336914, "learning_rate": 9.99952231875316e-05, "loss": 0.0576, "step": 2710 }, { "epoch": 0.9988982739625413, "grad_norm": 0.33036211133003235, "learning_rate": 9.999475514879795e-05, "loss": 0.0573, "step": 2720 }, { "epoch": 1.0025706940874035, "grad_norm": 0.44164949655532837, "learning_rate": 9.999426524065984e-05, "loss": 0.056, "step": 2730 }, { "epoch": 1.0062431142122659, "grad_norm": 0.5584012269973755, "learning_rate": 9.999375346333162e-05, "loss": 0.0587, "step": 2740 }, { "epoch": 1.0099155343371282, "grad_norm": 0.40010857582092285, "learning_rate": 9.999321981703715e-05, "loss": 0.0558, "step": 2750 }, { "epoch": 1.0135879544619903, "grad_norm": 0.33025604486465454, "learning_rate": 9.999266430200985e-05, "loss": 0.0578, "step": 2760 }, { "epoch": 1.0172603745868527, "grad_norm": 0.34739819169044495, "learning_rate": 9.999208691849271e-05, "loss": 0.0562, "step": 2770 }, { "epoch": 1.020932794711715, "grad_norm": 0.3972585201263428, "learning_rate": 9.999148766673832e-05, "loss": 0.0563, "step": 2780 }, { "epoch": 1.0246052148365774, "grad_norm": 0.32456573843955994, "learning_rate": 9.999086654700881e-05, "loss": 0.0556, "step": 2790 }, { "epoch": 1.0282776349614395, "grad_norm": 0.3587065637111664, "learning_rate": 9.999022355957588e-05, "loss": 0.0568, "step": 2800 }, { "epoch": 1.0319500550863019, "grad_norm": 0.4903677999973297, "learning_rate": 9.998955870472079e-05, "loss": 0.0599, "step": 2810 }, { "epoch": 1.0356224752111642, "grad_norm": 0.5346408486366272, "learning_rate": 9.998887198273437e-05, "loss": 0.0553, "step": 2820 }, { "epoch": 1.0392948953360264, "grad_norm": 0.2774538993835449, "learning_rate": 9.998816339391701e-05, "loss": 0.0571, "step": 2830 }, { "epoch": 1.0429673154608887, "grad_norm": 0.4205782413482666, "learning_rate": 9.998743293857868e-05, "loss": 0.0601, "step": 2840 }, { "epoch": 1.046639735585751, "grad_norm": 0.27348756790161133, "learning_rate": 9.998668061703891e-05, "loss": 0.0549, "step": 2850 }, { "epoch": 1.0503121557106132, "grad_norm": 0.2712705135345459, "learning_rate": 9.998590642962679e-05, "loss": 0.0551, "step": 2860 }, { "epoch": 1.0539845758354756, "grad_norm": 0.36297324299812317, "learning_rate": 9.998511037668095e-05, "loss": 0.0581, "step": 2870 }, { "epoch": 1.057656995960338, "grad_norm": 0.2600325047969818, "learning_rate": 9.998429245854964e-05, "loss": 0.0565, "step": 2880 }, { "epoch": 1.0613294160852003, "grad_norm": 0.2767641842365265, "learning_rate": 9.998345267559064e-05, "loss": 0.0546, "step": 2890 }, { "epoch": 1.0650018362100624, "grad_norm": 0.5701701641082764, "learning_rate": 9.998259102817129e-05, "loss": 0.0549, "step": 2900 }, { "epoch": 1.0686742563349247, "grad_norm": 0.4390009641647339, "learning_rate": 9.99817075166685e-05, "loss": 0.0601, "step": 2910 }, { "epoch": 1.072346676459787, "grad_norm": 0.5996479988098145, "learning_rate": 9.998080214146878e-05, "loss": 0.0547, "step": 2920 }, { "epoch": 1.0760190965846492, "grad_norm": 0.2895946502685547, "learning_rate": 9.997987490296813e-05, "loss": 0.0545, "step": 2930 }, { "epoch": 1.0796915167095116, "grad_norm": 0.3272102475166321, "learning_rate": 9.99789258015722e-05, "loss": 0.0561, "step": 2940 }, { "epoch": 1.083363936834374, "grad_norm": 0.3197779059410095, "learning_rate": 9.997795483769611e-05, "loss": 0.0554, "step": 2950 }, { "epoch": 1.087036356959236, "grad_norm": 0.4473246932029724, "learning_rate": 9.997696201176462e-05, "loss": 0.0583, "step": 2960 }, { "epoch": 1.0907087770840984, "grad_norm": 0.3481227159500122, "learning_rate": 9.997594732421203e-05, "loss": 0.0565, "step": 2970 }, { "epoch": 1.0943811972089608, "grad_norm": 0.3280641734600067, "learning_rate": 9.997491077548217e-05, "loss": 0.0531, "step": 2980 }, { "epoch": 1.098053617333823, "grad_norm": 0.5157963037490845, "learning_rate": 9.997385236602851e-05, "loss": 0.0527, "step": 2990 }, { "epoch": 1.1017260374586852, "grad_norm": 0.3513844907283783, "learning_rate": 9.997277209631399e-05, "loss": 0.0543, "step": 3000 }, { "epoch": 1.1053984575835476, "grad_norm": 0.5925564765930176, "learning_rate": 9.997166996681118e-05, "loss": 0.0543, "step": 3010 }, { "epoch": 1.10907087770841, "grad_norm": 0.30370572209358215, "learning_rate": 9.997054597800218e-05, "loss": 0.0575, "step": 3020 }, { "epoch": 1.112743297833272, "grad_norm": 0.3618398904800415, "learning_rate": 9.996940013037866e-05, "loss": 0.0531, "step": 3030 }, { "epoch": 1.1164157179581344, "grad_norm": 0.32652008533477783, "learning_rate": 9.996823242444186e-05, "loss": 0.0549, "step": 3040 }, { "epoch": 1.1200881380829968, "grad_norm": 0.30139821767807007, "learning_rate": 9.996704286070258e-05, "loss": 0.0538, "step": 3050 }, { "epoch": 1.123760558207859, "grad_norm": 0.4269617795944214, "learning_rate": 9.996583143968115e-05, "loss": 0.0517, "step": 3060 }, { "epoch": 1.1274329783327213, "grad_norm": 0.7178385853767395, "learning_rate": 9.99645981619075e-05, "loss": 0.0577, "step": 3070 }, { "epoch": 1.1311053984575836, "grad_norm": 0.584500253200531, "learning_rate": 9.996334302792114e-05, "loss": 0.0538, "step": 3080 }, { "epoch": 1.1347778185824458, "grad_norm": 0.7301107048988342, "learning_rate": 9.996206603827105e-05, "loss": 0.0535, "step": 3090 }, { "epoch": 1.138450238707308, "grad_norm": 0.45780837535858154, "learning_rate": 9.996076719351587e-05, "loss": 0.0527, "step": 3100 }, { "epoch": 1.1421226588321705, "grad_norm": 0.42784926295280457, "learning_rate": 9.995944649422374e-05, "loss": 0.0586, "step": 3110 }, { "epoch": 1.1457950789570326, "grad_norm": 0.42671999335289, "learning_rate": 9.995810394097239e-05, "loss": 0.052, "step": 3120 }, { "epoch": 1.149467499081895, "grad_norm": 0.3212076425552368, "learning_rate": 9.995673953434909e-05, "loss": 0.0549, "step": 3130 }, { "epoch": 1.1531399192067573, "grad_norm": 0.26237380504608154, "learning_rate": 9.995535327495068e-05, "loss": 0.0516, "step": 3140 }, { "epoch": 1.1568123393316196, "grad_norm": 0.2739783525466919, "learning_rate": 9.995394516338355e-05, "loss": 0.0541, "step": 3150 }, { "epoch": 1.1604847594564818, "grad_norm": 0.36502811312675476, "learning_rate": 9.995251520026367e-05, "loss": 0.051, "step": 3160 }, { "epoch": 1.1641571795813441, "grad_norm": 0.2346629649400711, "learning_rate": 9.995106338621656e-05, "loss": 0.0526, "step": 3170 }, { "epoch": 1.1678295997062065, "grad_norm": 0.2725568413734436, "learning_rate": 9.994958972187726e-05, "loss": 0.0507, "step": 3180 }, { "epoch": 1.1715020198310686, "grad_norm": 0.3143811821937561, "learning_rate": 9.994809420789044e-05, "loss": 0.0521, "step": 3190 }, { "epoch": 1.175174439955931, "grad_norm": 0.35082197189331055, "learning_rate": 9.994657684491027e-05, "loss": 0.0562, "step": 3200 }, { "epoch": 1.1788468600807933, "grad_norm": 0.4089239537715912, "learning_rate": 9.994503763360048e-05, "loss": 0.0531, "step": 3210 }, { "epoch": 1.1825192802056554, "grad_norm": 0.7218791246414185, "learning_rate": 9.99434765746344e-05, "loss": 0.0521, "step": 3220 }, { "epoch": 1.1861917003305178, "grad_norm": 0.40440136194229126, "learning_rate": 9.994189366869488e-05, "loss": 0.0507, "step": 3230 }, { "epoch": 1.1898641204553801, "grad_norm": 0.32858529686927795, "learning_rate": 9.994028891647433e-05, "loss": 0.0491, "step": 3240 }, { "epoch": 1.1935365405802423, "grad_norm": 0.3033769726753235, "learning_rate": 9.993866231867475e-05, "loss": 0.0489, "step": 3250 }, { "epoch": 1.1972089607051046, "grad_norm": 0.3382098972797394, "learning_rate": 9.993701387600762e-05, "loss": 0.0564, "step": 3260 }, { "epoch": 1.200881380829967, "grad_norm": 0.2744351923465729, "learning_rate": 9.993534358919408e-05, "loss": 0.0527, "step": 3270 }, { "epoch": 1.2045538009548293, "grad_norm": 0.4571131467819214, "learning_rate": 9.993365145896473e-05, "loss": 0.0502, "step": 3280 }, { "epoch": 1.2082262210796915, "grad_norm": 0.4518074691295624, "learning_rate": 9.993193748605977e-05, "loss": 0.0551, "step": 3290 }, { "epoch": 1.2118986412045538, "grad_norm": 0.3316684067249298, "learning_rate": 9.993020167122898e-05, "loss": 0.0509, "step": 3300 }, { "epoch": 1.2155710613294162, "grad_norm": 0.41578277945518494, "learning_rate": 9.992844401523164e-05, "loss": 0.0501, "step": 3310 }, { "epoch": 1.2192434814542783, "grad_norm": 0.4118627905845642, "learning_rate": 9.992666451883661e-05, "loss": 0.0527, "step": 3320 }, { "epoch": 1.2229159015791407, "grad_norm": 0.38633993268013, "learning_rate": 9.99248631828223e-05, "loss": 0.0515, "step": 3330 }, { "epoch": 1.226588321704003, "grad_norm": 0.3725762367248535, "learning_rate": 9.99230400079767e-05, "loss": 0.0485, "step": 3340 }, { "epoch": 1.2302607418288651, "grad_norm": 0.33291545510292053, "learning_rate": 9.992119499509728e-05, "loss": 0.0546, "step": 3350 }, { "epoch": 1.2339331619537275, "grad_norm": 0.38702869415283203, "learning_rate": 9.991932814499114e-05, "loss": 0.0508, "step": 3360 }, { "epoch": 1.2376055820785898, "grad_norm": 0.5252388715744019, "learning_rate": 9.991743945847493e-05, "loss": 0.0565, "step": 3370 }, { "epoch": 1.241278002203452, "grad_norm": 0.386886864900589, "learning_rate": 9.991552893637478e-05, "loss": 0.0531, "step": 3380 }, { "epoch": 1.2449504223283143, "grad_norm": 0.46932193636894226, "learning_rate": 9.991359657952644e-05, "loss": 0.0521, "step": 3390 }, { "epoch": 1.2486228424531767, "grad_norm": 0.540236234664917, "learning_rate": 9.991164238877519e-05, "loss": 0.0573, "step": 3400 }, { "epoch": 1.252295262578039, "grad_norm": 0.37271228432655334, "learning_rate": 9.990966636497585e-05, "loss": 0.0513, "step": 3410 }, { "epoch": 1.2559676827029012, "grad_norm": 0.27484437823295593, "learning_rate": 9.99076685089928e-05, "loss": 0.0535, "step": 3420 }, { "epoch": 1.2596401028277635, "grad_norm": 0.36095547676086426, "learning_rate": 9.990564882169998e-05, "loss": 0.0504, "step": 3430 }, { "epoch": 1.2633125229526259, "grad_norm": 0.3032003343105316, "learning_rate": 9.990360730398088e-05, "loss": 0.0519, "step": 3440 }, { "epoch": 1.266984943077488, "grad_norm": 0.2606087923049927, "learning_rate": 9.990154395672849e-05, "loss": 0.0481, "step": 3450 }, { "epoch": 1.2706573632023503, "grad_norm": 0.3278205692768097, "learning_rate": 9.989945878084541e-05, "loss": 0.0532, "step": 3460 }, { "epoch": 1.2743297833272127, "grad_norm": 0.6717965006828308, "learning_rate": 9.989735177724378e-05, "loss": 0.051, "step": 3470 }, { "epoch": 1.278002203452075, "grad_norm": 0.2528177499771118, "learning_rate": 9.989522294684526e-05, "loss": 0.0524, "step": 3480 }, { "epoch": 1.2816746235769372, "grad_norm": 0.41772890090942383, "learning_rate": 9.989307229058107e-05, "loss": 0.0515, "step": 3490 }, { "epoch": 1.2853470437017995, "grad_norm": 0.48842498660087585, "learning_rate": 9.989089980939202e-05, "loss": 0.057, "step": 3500 }, { "epoch": 1.2890194638266617, "grad_norm": 0.38779503107070923, "learning_rate": 9.988870550422835e-05, "loss": 0.0507, "step": 3510 }, { "epoch": 1.292691883951524, "grad_norm": 0.5739973783493042, "learning_rate": 9.988648937604999e-05, "loss": 0.0527, "step": 3520 }, { "epoch": 1.2963643040763864, "grad_norm": 0.3562142252922058, "learning_rate": 9.988425142582632e-05, "loss": 0.051, "step": 3530 }, { "epoch": 1.3000367242012487, "grad_norm": 0.30804887413978577, "learning_rate": 9.98819916545363e-05, "loss": 0.0534, "step": 3540 }, { "epoch": 1.3037091443261108, "grad_norm": 0.5179295539855957, "learning_rate": 9.987971006316844e-05, "loss": 0.0541, "step": 3550 }, { "epoch": 1.3073815644509732, "grad_norm": 0.28727003931999207, "learning_rate": 9.987740665272077e-05, "loss": 0.053, "step": 3560 }, { "epoch": 1.3110539845758356, "grad_norm": 0.49559274315834045, "learning_rate": 9.98750814242009e-05, "loss": 0.0527, "step": 3570 }, { "epoch": 1.3147264047006977, "grad_norm": 0.36319488286972046, "learning_rate": 9.987273437862594e-05, "loss": 0.0561, "step": 3580 }, { "epoch": 1.31839882482556, "grad_norm": 0.4091593325138092, "learning_rate": 9.987036551702259e-05, "loss": 0.0506, "step": 3590 }, { "epoch": 1.3220712449504224, "grad_norm": 0.45014336705207825, "learning_rate": 9.986797484042706e-05, "loss": 0.0525, "step": 3600 }, { "epoch": 1.3257436650752847, "grad_norm": 0.2973388433456421, "learning_rate": 9.986556234988512e-05, "loss": 0.047, "step": 3610 }, { "epoch": 1.3294160852001469, "grad_norm": 0.4841061532497406, "learning_rate": 9.986312804645205e-05, "loss": 0.0473, "step": 3620 }, { "epoch": 1.3330885053250092, "grad_norm": 0.31539395451545715, "learning_rate": 9.986067193119273e-05, "loss": 0.0498, "step": 3630 }, { "epoch": 1.3367609254498714, "grad_norm": 0.3809545040130615, "learning_rate": 9.985819400518153e-05, "loss": 0.0528, "step": 3640 }, { "epoch": 1.3404333455747337, "grad_norm": 0.37141409516334534, "learning_rate": 9.985569426950239e-05, "loss": 0.0488, "step": 3650 }, { "epoch": 1.344105765699596, "grad_norm": 0.34278854727745056, "learning_rate": 9.985317272524876e-05, "loss": 0.0487, "step": 3660 }, { "epoch": 1.3477781858244584, "grad_norm": 0.3122859299182892, "learning_rate": 9.985062937352366e-05, "loss": 0.0506, "step": 3670 }, { "epoch": 1.3514506059493205, "grad_norm": 0.32441678643226624, "learning_rate": 9.984806421543966e-05, "loss": 0.049, "step": 3680 }, { "epoch": 1.355123026074183, "grad_norm": 0.3428247570991516, "learning_rate": 9.984547725211881e-05, "loss": 0.0506, "step": 3690 }, { "epoch": 1.3587954461990452, "grad_norm": 0.38374873995780945, "learning_rate": 9.984286848469276e-05, "loss": 0.0472, "step": 3700 }, { "epoch": 1.3624678663239074, "grad_norm": 0.3668331503868103, "learning_rate": 9.984023791430266e-05, "loss": 0.0538, "step": 3710 }, { "epoch": 1.3661402864487697, "grad_norm": 0.25318047404289246, "learning_rate": 9.983758554209924e-05, "loss": 0.0504, "step": 3720 }, { "epoch": 1.369812706573632, "grad_norm": 0.4895252585411072, "learning_rate": 9.983491136924268e-05, "loss": 0.0493, "step": 3730 }, { "epoch": 1.3734851266984944, "grad_norm": 0.2891036868095398, "learning_rate": 9.983221539690282e-05, "loss": 0.0535, "step": 3740 }, { "epoch": 1.3771575468233566, "grad_norm": 0.46826672554016113, "learning_rate": 9.982949762625892e-05, "loss": 0.0513, "step": 3750 }, { "epoch": 1.380829966948219, "grad_norm": 0.3914286196231842, "learning_rate": 9.982675805849986e-05, "loss": 0.0516, "step": 3760 }, { "epoch": 1.384502387073081, "grad_norm": 0.43897801637649536, "learning_rate": 9.982399669482399e-05, "loss": 0.0554, "step": 3770 }, { "epoch": 1.3881748071979434, "grad_norm": 0.350764662027359, "learning_rate": 9.982121353643924e-05, "loss": 0.0511, "step": 3780 }, { "epoch": 1.3918472273228057, "grad_norm": 0.3713981509208679, "learning_rate": 9.981840858456306e-05, "loss": 0.0493, "step": 3790 }, { "epoch": 1.395519647447668, "grad_norm": 0.4496918022632599, "learning_rate": 9.981558184042243e-05, "loss": 0.05, "step": 3800 }, { "epoch": 1.3991920675725302, "grad_norm": 0.33541634678840637, "learning_rate": 9.981273330525387e-05, "loss": 0.0489, "step": 3810 }, { "epoch": 1.4028644876973926, "grad_norm": 0.2583834230899811, "learning_rate": 9.980986298030341e-05, "loss": 0.0523, "step": 3820 }, { "epoch": 1.406536907822255, "grad_norm": 0.46596652269363403, "learning_rate": 9.980697086682662e-05, "loss": 0.0505, "step": 3830 }, { "epoch": 1.410209327947117, "grad_norm": 0.3442924916744232, "learning_rate": 9.980405696608866e-05, "loss": 0.0495, "step": 3840 }, { "epoch": 1.4138817480719794, "grad_norm": 0.40674230456352234, "learning_rate": 9.980112127936409e-05, "loss": 0.0481, "step": 3850 }, { "epoch": 1.4175541681968418, "grad_norm": 0.3338242471218109, "learning_rate": 9.979816380793717e-05, "loss": 0.0515, "step": 3860 }, { "epoch": 1.4212265883217041, "grad_norm": 0.4579961895942688, "learning_rate": 9.979518455310151e-05, "loss": 0.0466, "step": 3870 }, { "epoch": 1.4248990084465663, "grad_norm": 0.48042699694633484, "learning_rate": 9.97921835161604e-05, "loss": 0.0516, "step": 3880 }, { "epoch": 1.4285714285714286, "grad_norm": 0.44948965311050415, "learning_rate": 9.978916069842656e-05, "loss": 0.0497, "step": 3890 }, { "epoch": 1.4322438486962907, "grad_norm": 0.3409099578857422, "learning_rate": 9.97861161012223e-05, "loss": 0.0502, "step": 3900 }, { "epoch": 1.435916268821153, "grad_norm": 0.35010281205177307, "learning_rate": 9.978304972587942e-05, "loss": 0.05, "step": 3910 }, { "epoch": 1.4395886889460154, "grad_norm": 0.2616921067237854, "learning_rate": 9.977996157373925e-05, "loss": 0.0486, "step": 3920 }, { "epoch": 1.4432611090708778, "grad_norm": 0.45934441685676575, "learning_rate": 9.977685164615265e-05, "loss": 0.0473, "step": 3930 }, { "epoch": 1.44693352919574, "grad_norm": 0.4672592282295227, "learning_rate": 9.977371994448002e-05, "loss": 0.0511, "step": 3940 }, { "epoch": 1.4506059493206023, "grad_norm": 0.4792556166648865, "learning_rate": 9.977056647009127e-05, "loss": 0.0521, "step": 3950 }, { "epoch": 1.4542783694454646, "grad_norm": 0.3087407946586609, "learning_rate": 9.976739122436582e-05, "loss": 0.0454, "step": 3960 }, { "epoch": 1.4579507895703268, "grad_norm": 0.5605859160423279, "learning_rate": 9.976419420869265e-05, "loss": 0.0503, "step": 3970 }, { "epoch": 1.461623209695189, "grad_norm": 0.460191011428833, "learning_rate": 9.976097542447025e-05, "loss": 0.0481, "step": 3980 }, { "epoch": 1.4652956298200515, "grad_norm": 0.3959011137485504, "learning_rate": 9.97577348731066e-05, "loss": 0.047, "step": 3990 }, { "epoch": 1.4689680499449138, "grad_norm": 0.43872568011283875, "learning_rate": 9.975447255601927e-05, "loss": 0.0517, "step": 4000 }, { "epoch": 1.472640470069776, "grad_norm": 0.27127742767333984, "learning_rate": 9.975118847463525e-05, "loss": 0.0489, "step": 4010 }, { "epoch": 1.4763128901946383, "grad_norm": 0.2785481810569763, "learning_rate": 9.974788263039114e-05, "loss": 0.0477, "step": 4020 }, { "epoch": 1.4799853103195004, "grad_norm": 0.24817723035812378, "learning_rate": 9.974455502473303e-05, "loss": 0.0466, "step": 4030 }, { "epoch": 1.4836577304443628, "grad_norm": 0.24665164947509766, "learning_rate": 9.974120565911652e-05, "loss": 0.0475, "step": 4040 }, { "epoch": 1.4873301505692251, "grad_norm": 0.36699724197387695, "learning_rate": 9.973783453500674e-05, "loss": 0.0493, "step": 4050 }, { "epoch": 1.4910025706940875, "grad_norm": 0.43315231800079346, "learning_rate": 9.973444165387835e-05, "loss": 0.0512, "step": 4060 }, { "epoch": 1.4946749908189496, "grad_norm": 0.33927804231643677, "learning_rate": 9.973102701721549e-05, "loss": 0.0464, "step": 4070 }, { "epoch": 1.498347410943812, "grad_norm": 0.406687468290329, "learning_rate": 9.972759062651184e-05, "loss": 0.0455, "step": 4080 }, { "epoch": 1.502019831068674, "grad_norm": 0.2971259653568268, "learning_rate": 9.972413248327059e-05, "loss": 0.0468, "step": 4090 }, { "epoch": 1.5056922511935364, "grad_norm": 0.3438482880592346, "learning_rate": 9.972065258900447e-05, "loss": 0.0489, "step": 4100 }, { "epoch": 1.5093646713183988, "grad_norm": 0.36278578639030457, "learning_rate": 9.971715094523569e-05, "loss": 0.0446, "step": 4110 }, { "epoch": 1.5130370914432612, "grad_norm": 0.30859947204589844, "learning_rate": 9.971362755349598e-05, "loss": 0.0466, "step": 4120 }, { "epoch": 1.5167095115681235, "grad_norm": 0.3369341194629669, "learning_rate": 9.971008241532662e-05, "loss": 0.0457, "step": 4130 }, { "epoch": 1.5203819316929856, "grad_norm": 0.5966343879699707, "learning_rate": 9.970651553227835e-05, "loss": 0.0512, "step": 4140 }, { "epoch": 1.524054351817848, "grad_norm": 0.36642828583717346, "learning_rate": 9.970292690591143e-05, "loss": 0.0471, "step": 4150 }, { "epoch": 1.5277267719427101, "grad_norm": 0.471881628036499, "learning_rate": 9.969931653779569e-05, "loss": 0.0446, "step": 4160 }, { "epoch": 1.5313991920675725, "grad_norm": 0.3692777156829834, "learning_rate": 9.969568442951038e-05, "loss": 0.0498, "step": 4170 }, { "epoch": 1.5350716121924348, "grad_norm": 0.3732738196849823, "learning_rate": 9.969203058264436e-05, "loss": 0.0504, "step": 4180 }, { "epoch": 1.5387440323172972, "grad_norm": 0.3958605229854584, "learning_rate": 9.96883549987959e-05, "loss": 0.0521, "step": 4190 }, { "epoch": 1.5424164524421595, "grad_norm": 0.2677345275878906, "learning_rate": 9.968465767957287e-05, "loss": 0.0457, "step": 4200 }, { "epoch": 1.5460888725670217, "grad_norm": 0.45833104848861694, "learning_rate": 9.968093862659256e-05, "loss": 0.047, "step": 4210 }, { "epoch": 1.5497612926918838, "grad_norm": 0.38095584511756897, "learning_rate": 9.967719784148182e-05, "loss": 0.0482, "step": 4220 }, { "epoch": 1.5534337128167461, "grad_norm": 0.3486616313457489, "learning_rate": 9.967343532587702e-05, "loss": 0.0478, "step": 4230 }, { "epoch": 1.5571061329416085, "grad_norm": 0.3854934573173523, "learning_rate": 9.966965108142399e-05, "loss": 0.0482, "step": 4240 }, { "epoch": 1.5607785530664708, "grad_norm": 0.3291497528553009, "learning_rate": 9.96658451097781e-05, "loss": 0.0466, "step": 4250 }, { "epoch": 1.5644509731913332, "grad_norm": 0.3308378756046295, "learning_rate": 9.966201741260419e-05, "loss": 0.051, "step": 4260 }, { "epoch": 1.5681233933161953, "grad_norm": 0.33386945724487305, "learning_rate": 9.965816799157665e-05, "loss": 0.0469, "step": 4270 }, { "epoch": 1.5717958134410577, "grad_norm": 0.41131025552749634, "learning_rate": 9.965429684837935e-05, "loss": 0.0475, "step": 4280 }, { "epoch": 1.5754682335659198, "grad_norm": 0.28767019510269165, "learning_rate": 9.965040398470562e-05, "loss": 0.0452, "step": 4290 }, { "epoch": 1.5791406536907822, "grad_norm": 0.3024898171424866, "learning_rate": 9.964648940225838e-05, "loss": 0.0476, "step": 4300 }, { "epoch": 1.5828130738156445, "grad_norm": 0.4294430911540985, "learning_rate": 9.964255310274997e-05, "loss": 0.0457, "step": 4310 }, { "epoch": 1.5864854939405069, "grad_norm": 0.35421988368034363, "learning_rate": 9.963859508790228e-05, "loss": 0.0437, "step": 4320 }, { "epoch": 1.5901579140653692, "grad_norm": 0.3224738836288452, "learning_rate": 9.963461535944664e-05, "loss": 0.0472, "step": 4330 }, { "epoch": 1.5938303341902313, "grad_norm": 0.2953808009624481, "learning_rate": 9.963061391912399e-05, "loss": 0.0457, "step": 4340 }, { "epoch": 1.5975027543150937, "grad_norm": 0.3281307518482208, "learning_rate": 9.962659076868463e-05, "loss": 0.0464, "step": 4350 }, { "epoch": 1.6011751744399558, "grad_norm": 0.2737230360507965, "learning_rate": 9.962254590988846e-05, "loss": 0.0452, "step": 4360 }, { "epoch": 1.6048475945648182, "grad_norm": 0.2831273674964905, "learning_rate": 9.961847934450481e-05, "loss": 0.0483, "step": 4370 }, { "epoch": 1.6085200146896805, "grad_norm": 0.47394615411758423, "learning_rate": 9.961439107431257e-05, "loss": 0.045, "step": 4380 }, { "epoch": 1.6121924348145429, "grad_norm": 0.2905158996582031, "learning_rate": 9.961028110110006e-05, "loss": 0.0479, "step": 4390 }, { "epoch": 1.615864854939405, "grad_norm": 0.33560141921043396, "learning_rate": 9.960614942666513e-05, "loss": 0.05, "step": 4400 }, { "epoch": 1.6195372750642674, "grad_norm": 0.27083060145378113, "learning_rate": 9.960199605281511e-05, "loss": 0.048, "step": 4410 }, { "epoch": 1.6232096951891295, "grad_norm": 0.326571524143219, "learning_rate": 9.959782098136683e-05, "loss": 0.0496, "step": 4420 }, { "epoch": 1.6268821153139919, "grad_norm": 0.3304294943809509, "learning_rate": 9.959362421414662e-05, "loss": 0.0462, "step": 4430 }, { "epoch": 1.6305545354388542, "grad_norm": 0.39019259810447693, "learning_rate": 9.958940575299027e-05, "loss": 0.0501, "step": 4440 }, { "epoch": 1.6342269555637166, "grad_norm": 0.3958894908428192, "learning_rate": 9.95851655997431e-05, "loss": 0.0441, "step": 4450 }, { "epoch": 1.637899375688579, "grad_norm": 0.40329307317733765, "learning_rate": 9.958090375625986e-05, "loss": 0.046, "step": 4460 }, { "epoch": 1.641571795813441, "grad_norm": 0.33684253692626953, "learning_rate": 9.957662022440486e-05, "loss": 0.0457, "step": 4470 }, { "epoch": 1.6452442159383034, "grad_norm": 0.4812668263912201, "learning_rate": 9.957231500605187e-05, "loss": 0.0445, "step": 4480 }, { "epoch": 1.6489166360631655, "grad_norm": 0.2700076997280121, "learning_rate": 9.95679881030841e-05, "loss": 0.0478, "step": 4490 }, { "epoch": 1.6525890561880279, "grad_norm": 0.31808945536613464, "learning_rate": 9.95636395173943e-05, "loss": 0.0474, "step": 4500 }, { "epoch": 1.6562614763128902, "grad_norm": 0.35798364877700806, "learning_rate": 9.95592692508847e-05, "loss": 0.0508, "step": 4510 }, { "epoch": 1.6599338964377526, "grad_norm": 0.4490469992160797, "learning_rate": 9.9554877305467e-05, "loss": 0.0459, "step": 4520 }, { "epoch": 1.663606316562615, "grad_norm": 0.35144519805908203, "learning_rate": 9.955046368306237e-05, "loss": 0.0438, "step": 4530 }, { "epoch": 1.667278736687477, "grad_norm": 0.3263298273086548, "learning_rate": 9.954602838560153e-05, "loss": 0.0451, "step": 4540 }, { "epoch": 1.6709511568123392, "grad_norm": 0.43081748485565186, "learning_rate": 9.954157141502456e-05, "loss": 0.0442, "step": 4550 }, { "epoch": 1.6746235769372015, "grad_norm": 0.4007404148578644, "learning_rate": 9.953709277328112e-05, "loss": 0.0459, "step": 4560 }, { "epoch": 1.678295997062064, "grad_norm": 0.38571879267692566, "learning_rate": 9.953259246233032e-05, "loss": 0.0438, "step": 4570 }, { "epoch": 1.6819684171869262, "grad_norm": 0.6237848997116089, "learning_rate": 9.952807048414077e-05, "loss": 0.0457, "step": 4580 }, { "epoch": 1.6856408373117886, "grad_norm": 0.31377196311950684, "learning_rate": 9.95235268406905e-05, "loss": 0.0448, "step": 4590 }, { "epoch": 1.6893132574366507, "grad_norm": 0.3646199107170105, "learning_rate": 9.951896153396708e-05, "loss": 0.0477, "step": 4600 }, { "epoch": 1.692985677561513, "grad_norm": 0.28171831369400024, "learning_rate": 9.95143745659675e-05, "loss": 0.0464, "step": 4610 }, { "epoch": 1.6966580976863752, "grad_norm": 0.2675369083881378, "learning_rate": 9.95097659386983e-05, "loss": 0.0475, "step": 4620 }, { "epoch": 1.7003305178112376, "grad_norm": 0.30328625440597534, "learning_rate": 9.950513565417542e-05, "loss": 0.0445, "step": 4630 }, { "epoch": 1.7040029379361, "grad_norm": 0.3386673033237457, "learning_rate": 9.95004837144243e-05, "loss": 0.0459, "step": 4640 }, { "epoch": 1.7076753580609623, "grad_norm": 0.37747883796691895, "learning_rate": 9.949581012147988e-05, "loss": 0.0442, "step": 4650 }, { "epoch": 1.7113477781858246, "grad_norm": 0.3088318407535553, "learning_rate": 9.949111487738653e-05, "loss": 0.0517, "step": 4660 }, { "epoch": 1.7150201983106867, "grad_norm": 0.3810873329639435, "learning_rate": 9.948639798419813e-05, "loss": 0.0451, "step": 4670 }, { "epoch": 1.7186926184355489, "grad_norm": 0.3046751022338867, "learning_rate": 9.948165944397799e-05, "loss": 0.0423, "step": 4680 }, { "epoch": 1.7223650385604112, "grad_norm": 0.26212573051452637, "learning_rate": 9.94768992587989e-05, "loss": 0.0462, "step": 4690 }, { "epoch": 1.7260374586852736, "grad_norm": 0.5924479365348816, "learning_rate": 9.947211743074313e-05, "loss": 0.0508, "step": 4700 }, { "epoch": 1.729709878810136, "grad_norm": 0.4646141827106476, "learning_rate": 9.946731396190246e-05, "loss": 0.0463, "step": 4710 }, { "epoch": 1.7333822989349983, "grad_norm": 0.2696131467819214, "learning_rate": 9.946248885437803e-05, "loss": 0.0468, "step": 4720 }, { "epoch": 1.7370547190598604, "grad_norm": 0.3098793029785156, "learning_rate": 9.945764211028053e-05, "loss": 0.0477, "step": 4730 }, { "epoch": 1.7407271391847228, "grad_norm": 0.3072303831577301, "learning_rate": 9.94527737317301e-05, "loss": 0.0455, "step": 4740 }, { "epoch": 1.744399559309585, "grad_norm": 0.40793246030807495, "learning_rate": 9.944788372085631e-05, "loss": 0.0455, "step": 4750 }, { "epoch": 1.7480719794344473, "grad_norm": 0.5640067458152771, "learning_rate": 9.944297207979825e-05, "loss": 0.0469, "step": 4760 }, { "epoch": 1.7517443995593096, "grad_norm": 0.41774895787239075, "learning_rate": 9.943803881070441e-05, "loss": 0.0461, "step": 4770 }, { "epoch": 1.755416819684172, "grad_norm": 0.3528713881969452, "learning_rate": 9.943308391573278e-05, "loss": 0.0444, "step": 4780 }, { "epoch": 1.7590892398090343, "grad_norm": 0.28451505303382874, "learning_rate": 9.942810739705079e-05, "loss": 0.0476, "step": 4790 }, { "epoch": 1.7627616599338964, "grad_norm": 0.5325042009353638, "learning_rate": 9.942310925683538e-05, "loss": 0.0506, "step": 4800 }, { "epoch": 1.7664340800587586, "grad_norm": 0.4955766797065735, "learning_rate": 9.941808949727285e-05, "loss": 0.0452, "step": 4810 }, { "epoch": 1.770106500183621, "grad_norm": 0.3233533501625061, "learning_rate": 9.941304812055903e-05, "loss": 0.0429, "step": 4820 }, { "epoch": 1.7737789203084833, "grad_norm": 0.3619503676891327, "learning_rate": 9.940798512889921e-05, "loss": 0.0494, "step": 4830 }, { "epoch": 1.7774513404333456, "grad_norm": 0.7412535548210144, "learning_rate": 9.94029005245081e-05, "loss": 0.053, "step": 4840 }, { "epoch": 1.781123760558208, "grad_norm": 0.4123632311820984, "learning_rate": 9.939779430960988e-05, "loss": 0.0467, "step": 4850 }, { "epoch": 1.78479618068307, "grad_norm": 0.33250892162323, "learning_rate": 9.939266648643817e-05, "loss": 0.0448, "step": 4860 }, { "epoch": 1.7884686008079325, "grad_norm": 0.4020741581916809, "learning_rate": 9.938751705723607e-05, "loss": 0.0453, "step": 4870 }, { "epoch": 1.7921410209327946, "grad_norm": 0.26489967107772827, "learning_rate": 9.938234602425613e-05, "loss": 0.0473, "step": 4880 }, { "epoch": 1.795813441057657, "grad_norm": 0.42151492834091187, "learning_rate": 9.93771533897603e-05, "loss": 0.0452, "step": 4890 }, { "epoch": 1.7994858611825193, "grad_norm": 0.257162868976593, "learning_rate": 9.937193915602004e-05, "loss": 0.0451, "step": 4900 }, { "epoch": 1.8031582813073816, "grad_norm": 0.2551344633102417, "learning_rate": 9.936670332531621e-05, "loss": 0.0421, "step": 4910 }, { "epoch": 1.806830701432244, "grad_norm": 0.37055352330207825, "learning_rate": 9.936144589993916e-05, "loss": 0.0436, "step": 4920 }, { "epoch": 1.8105031215571061, "grad_norm": 0.38803768157958984, "learning_rate": 9.935616688218867e-05, "loss": 0.0455, "step": 4930 }, { "epoch": 1.8141755416819683, "grad_norm": 0.22915537655353546, "learning_rate": 9.935086627437395e-05, "loss": 0.0421, "step": 4940 }, { "epoch": 1.8178479618068306, "grad_norm": 0.5804263949394226, "learning_rate": 9.934554407881366e-05, "loss": 0.0437, "step": 4950 }, { "epoch": 1.821520381931693, "grad_norm": 0.31220272183418274, "learning_rate": 9.934020029783593e-05, "loss": 0.0424, "step": 4960 }, { "epoch": 1.8251928020565553, "grad_norm": 0.33828914165496826, "learning_rate": 9.933483493377829e-05, "loss": 0.0476, "step": 4970 }, { "epoch": 1.8288652221814177, "grad_norm": 0.34166786074638367, "learning_rate": 9.932944798898774e-05, "loss": 0.0463, "step": 4980 }, { "epoch": 1.8325376423062798, "grad_norm": 0.30953407287597656, "learning_rate": 9.932403946582072e-05, "loss": 0.044, "step": 4990 }, { "epoch": 1.8362100624311422, "grad_norm": 0.3936258852481842, "learning_rate": 9.93186093666431e-05, "loss": 0.0433, "step": 5000 }, { "epoch": 1.8398824825560043, "grad_norm": 0.47621944546699524, "learning_rate": 9.931315769383018e-05, "loss": 0.045, "step": 5010 }, { "epoch": 1.8435549026808666, "grad_norm": 0.48917990922927856, "learning_rate": 9.930768444976672e-05, "loss": 0.0448, "step": 5020 }, { "epoch": 1.847227322805729, "grad_norm": 0.2793985605239868, "learning_rate": 9.93021896368469e-05, "loss": 0.0416, "step": 5030 }, { "epoch": 1.8508997429305913, "grad_norm": 0.3316487669944763, "learning_rate": 9.929667325747431e-05, "loss": 0.0444, "step": 5040 }, { "epoch": 1.8545721630554537, "grad_norm": 0.3289085030555725, "learning_rate": 9.929113531406205e-05, "loss": 0.0441, "step": 5050 }, { "epoch": 1.8582445831803158, "grad_norm": 0.3271878659725189, "learning_rate": 9.928557580903257e-05, "loss": 0.0434, "step": 5060 }, { "epoch": 1.861917003305178, "grad_norm": 0.4740276038646698, "learning_rate": 9.927999474481779e-05, "loss": 0.0433, "step": 5070 }, { "epoch": 1.8655894234300403, "grad_norm": 0.38253241777420044, "learning_rate": 9.927439212385907e-05, "loss": 0.0424, "step": 5080 }, { "epoch": 1.8692618435549027, "grad_norm": 0.2832046449184418, "learning_rate": 9.926876794860718e-05, "loss": 0.0426, "step": 5090 }, { "epoch": 1.872934263679765, "grad_norm": 0.26664963364601135, "learning_rate": 9.926312222152235e-05, "loss": 0.0443, "step": 5100 }, { "epoch": 1.8766066838046274, "grad_norm": 0.31227919459342957, "learning_rate": 9.925745494507414e-05, "loss": 0.0428, "step": 5110 }, { "epoch": 1.8802791039294895, "grad_norm": 0.48232758045196533, "learning_rate": 9.925176612174169e-05, "loss": 0.0472, "step": 5120 }, { "epoch": 1.8839515240543518, "grad_norm": 0.3002265989780426, "learning_rate": 9.924605575401346e-05, "loss": 0.0426, "step": 5130 }, { "epoch": 1.887623944179214, "grad_norm": 0.36568376421928406, "learning_rate": 9.924032384438733e-05, "loss": 0.0416, "step": 5140 }, { "epoch": 1.8912963643040763, "grad_norm": 0.28417667746543884, "learning_rate": 9.923457039537066e-05, "loss": 0.0426, "step": 5150 }, { "epoch": 1.8949687844289387, "grad_norm": 0.3545460104942322, "learning_rate": 9.92287954094802e-05, "loss": 0.0416, "step": 5160 }, { "epoch": 1.898641204553801, "grad_norm": 0.3380478322505951, "learning_rate": 9.922299888924212e-05, "loss": 0.0409, "step": 5170 }, { "epoch": 1.9023136246786634, "grad_norm": 0.6531864404678345, "learning_rate": 9.921718083719203e-05, "loss": 0.0437, "step": 5180 }, { "epoch": 1.9059860448035255, "grad_norm": 0.4616503119468689, "learning_rate": 9.921134125587491e-05, "loss": 0.045, "step": 5190 }, { "epoch": 1.9096584649283876, "grad_norm": 0.6229126453399658, "learning_rate": 9.920548014784523e-05, "loss": 0.0449, "step": 5200 }, { "epoch": 1.91333088505325, "grad_norm": 0.23759698867797852, "learning_rate": 9.919959751566681e-05, "loss": 0.0455, "step": 5210 }, { "epoch": 1.9170033051781123, "grad_norm": 0.4140455722808838, "learning_rate": 9.919369336191291e-05, "loss": 0.0439, "step": 5220 }, { "epoch": 1.9206757253029747, "grad_norm": 0.22652927041053772, "learning_rate": 9.918776768916625e-05, "loss": 0.0433, "step": 5230 }, { "epoch": 1.924348145427837, "grad_norm": 0.2687198519706726, "learning_rate": 9.918182050001888e-05, "loss": 0.0453, "step": 5240 }, { "epoch": 1.9280205655526992, "grad_norm": 0.4781833589076996, "learning_rate": 9.91758517970723e-05, "loss": 0.0423, "step": 5250 }, { "epoch": 1.9316929856775615, "grad_norm": 0.26637327671051025, "learning_rate": 9.916986158293744e-05, "loss": 0.0456, "step": 5260 }, { "epoch": 1.9353654058024237, "grad_norm": 0.2591918706893921, "learning_rate": 9.916384986023463e-05, "loss": 0.0436, "step": 5270 }, { "epoch": 1.939037825927286, "grad_norm": 0.33919161558151245, "learning_rate": 9.915781663159359e-05, "loss": 0.0429, "step": 5280 }, { "epoch": 1.9427102460521484, "grad_norm": 0.36302268505096436, "learning_rate": 9.915176189965346e-05, "loss": 0.0443, "step": 5290 }, { "epoch": 1.9463826661770107, "grad_norm": 0.8904821276664734, "learning_rate": 9.914568566706279e-05, "loss": 0.0435, "step": 5300 }, { "epoch": 1.950055086301873, "grad_norm": 0.6777756214141846, "learning_rate": 9.913958793647953e-05, "loss": 0.0419, "step": 5310 }, { "epoch": 1.9537275064267352, "grad_norm": 0.2584041655063629, "learning_rate": 9.913346871057102e-05, "loss": 0.0449, "step": 5320 }, { "epoch": 1.9573999265515976, "grad_norm": 0.33154788613319397, "learning_rate": 9.912732799201403e-05, "loss": 0.0435, "step": 5330 }, { "epoch": 1.9610723466764597, "grad_norm": 0.4600938558578491, "learning_rate": 9.912116578349474e-05, "loss": 0.0438, "step": 5340 }, { "epoch": 1.964744766801322, "grad_norm": 0.3837641477584839, "learning_rate": 9.911498208770867e-05, "loss": 0.0443, "step": 5350 }, { "epoch": 1.9684171869261844, "grad_norm": 0.3436868190765381, "learning_rate": 9.910877690736078e-05, "loss": 0.0432, "step": 5360 }, { "epoch": 1.9720896070510467, "grad_norm": 0.516679048538208, "learning_rate": 9.910255024516546e-05, "loss": 0.0467, "step": 5370 }, { "epoch": 1.9757620271759089, "grad_norm": 0.5246255993843079, "learning_rate": 9.909630210384644e-05, "loss": 0.0401, "step": 5380 }, { "epoch": 1.9794344473007712, "grad_norm": 0.43808671832084656, "learning_rate": 9.909003248613688e-05, "loss": 0.0407, "step": 5390 }, { "epoch": 1.9831068674256334, "grad_norm": 0.2624474763870239, "learning_rate": 9.908374139477934e-05, "loss": 0.0414, "step": 5400 }, { "epoch": 1.9867792875504957, "grad_norm": 0.27374669909477234, "learning_rate": 9.907742883252571e-05, "loss": 0.0415, "step": 5410 }, { "epoch": 1.990451707675358, "grad_norm": 0.5663538575172424, "learning_rate": 9.907109480213736e-05, "loss": 0.039, "step": 5420 }, { "epoch": 1.9941241278002204, "grad_norm": 0.4146960973739624, "learning_rate": 9.906473930638498e-05, "loss": 0.0431, "step": 5430 }, { "epoch": 1.9977965479250828, "grad_norm": 0.3634147644042969, "learning_rate": 9.90583623480487e-05, "loss": 0.0425, "step": 5440 }, { "epoch": 2.001468968049945, "grad_norm": 0.48213016986846924, "learning_rate": 9.905196392991802e-05, "loss": 0.0453, "step": 5450 }, { "epoch": 2.005141388174807, "grad_norm": 0.5147107243537903, "learning_rate": 9.904554405479183e-05, "loss": 0.0424, "step": 5460 }, { "epoch": 2.0088138082996694, "grad_norm": 0.41459962725639343, "learning_rate": 9.903910272547838e-05, "loss": 0.0404, "step": 5470 }, { "epoch": 2.0124862284245317, "grad_norm": 0.2907109558582306, "learning_rate": 9.903263994479533e-05, "loss": 0.0424, "step": 5480 }, { "epoch": 2.016158648549394, "grad_norm": 0.32255819439888, "learning_rate": 9.902615571556972e-05, "loss": 0.0406, "step": 5490 }, { "epoch": 2.0198310686742564, "grad_norm": 0.2905977964401245, "learning_rate": 9.9019650040638e-05, "loss": 0.04, "step": 5500 }, { "epoch": 2.023503488799119, "grad_norm": 0.3099583387374878, "learning_rate": 9.901312292284594e-05, "loss": 0.0432, "step": 5510 }, { "epoch": 2.0271759089239807, "grad_norm": 0.40204575657844543, "learning_rate": 9.900657436504875e-05, "loss": 0.0373, "step": 5520 }, { "epoch": 2.030848329048843, "grad_norm": 0.5072874426841736, "learning_rate": 9.900000437011097e-05, "loss": 0.0467, "step": 5530 }, { "epoch": 2.0345207491737054, "grad_norm": 0.40589380264282227, "learning_rate": 9.899341294090653e-05, "loss": 0.0424, "step": 5540 }, { "epoch": 2.0381931692985678, "grad_norm": 0.37515711784362793, "learning_rate": 9.898680008031877e-05, "loss": 0.0423, "step": 5550 }, { "epoch": 2.04186558942343, "grad_norm": 0.3651323616504669, "learning_rate": 9.898016579124037e-05, "loss": 0.0425, "step": 5560 }, { "epoch": 2.0455380095482925, "grad_norm": 0.3262936472892761, "learning_rate": 9.89735100765734e-05, "loss": 0.0406, "step": 5570 }, { "epoch": 2.049210429673155, "grad_norm": 0.4748384952545166, "learning_rate": 9.89668329392293e-05, "loss": 0.0416, "step": 5580 }, { "epoch": 2.0528828497980167, "grad_norm": 0.37978124618530273, "learning_rate": 9.896013438212885e-05, "loss": 0.0443, "step": 5590 }, { "epoch": 2.056555269922879, "grad_norm": 0.5434566140174866, "learning_rate": 9.895341440820225e-05, "loss": 0.0434, "step": 5600 }, { "epoch": 2.0602276900477414, "grad_norm": 0.3358139097690582, "learning_rate": 9.894667302038902e-05, "loss": 0.0433, "step": 5610 }, { "epoch": 2.0639001101726038, "grad_norm": 0.460454523563385, "learning_rate": 9.893991022163812e-05, "loss": 0.0439, "step": 5620 }, { "epoch": 2.067572530297466, "grad_norm": 0.2625895142555237, "learning_rate": 9.893312601490779e-05, "loss": 0.0409, "step": 5630 }, { "epoch": 2.0712449504223285, "grad_norm": 0.5603809952735901, "learning_rate": 9.892632040316568e-05, "loss": 0.0469, "step": 5640 }, { "epoch": 2.0749173705471904, "grad_norm": 0.3060896694660187, "learning_rate": 9.891949338938877e-05, "loss": 0.0375, "step": 5650 }, { "epoch": 2.0785897906720527, "grad_norm": 0.48415789008140564, "learning_rate": 9.89126449765635e-05, "loss": 0.0417, "step": 5660 }, { "epoch": 2.082262210796915, "grad_norm": 0.48877981305122375, "learning_rate": 9.890577516768551e-05, "loss": 0.0411, "step": 5670 }, { "epoch": 2.0859346309217774, "grad_norm": 0.30385127663612366, "learning_rate": 9.889888396575995e-05, "loss": 0.0423, "step": 5680 }, { "epoch": 2.08960705104664, "grad_norm": 0.2835124731063843, "learning_rate": 9.889197137380125e-05, "loss": 0.0429, "step": 5690 }, { "epoch": 2.093279471171502, "grad_norm": 0.28398802876472473, "learning_rate": 9.888503739483321e-05, "loss": 0.0406, "step": 5700 }, { "epoch": 2.0969518912963645, "grad_norm": 0.5105298161506653, "learning_rate": 9.887808203188897e-05, "loss": 0.0444, "step": 5710 }, { "epoch": 2.1006243114212264, "grad_norm": 0.36199072003364563, "learning_rate": 9.887110528801108e-05, "loss": 0.0437, "step": 5720 }, { "epoch": 2.1042967315460888, "grad_norm": 0.3597424626350403, "learning_rate": 9.886410716625137e-05, "loss": 0.042, "step": 5730 }, { "epoch": 2.107969151670951, "grad_norm": 0.5267359018325806, "learning_rate": 9.885708766967108e-05, "loss": 0.0411, "step": 5740 }, { "epoch": 2.1116415717958135, "grad_norm": 0.38790926337242126, "learning_rate": 9.885004680134076e-05, "loss": 0.0421, "step": 5750 }, { "epoch": 2.115313991920676, "grad_norm": 0.3439289629459381, "learning_rate": 9.884298456434033e-05, "loss": 0.0426, "step": 5760 }, { "epoch": 2.118986412045538, "grad_norm": 0.2166810780763626, "learning_rate": 9.883590096175905e-05, "loss": 0.0434, "step": 5770 }, { "epoch": 2.1226588321704005, "grad_norm": 0.3219715356826782, "learning_rate": 9.882879599669554e-05, "loss": 0.044, "step": 5780 }, { "epoch": 2.1263312522952624, "grad_norm": 0.29195109009742737, "learning_rate": 9.882166967225775e-05, "loss": 0.0402, "step": 5790 }, { "epoch": 2.130003672420125, "grad_norm": 0.251016229391098, "learning_rate": 9.881452199156296e-05, "loss": 0.0419, "step": 5800 }, { "epoch": 2.133676092544987, "grad_norm": 0.2889130413532257, "learning_rate": 9.880735295773782e-05, "loss": 0.0403, "step": 5810 }, { "epoch": 2.1373485126698495, "grad_norm": 0.8795928359031677, "learning_rate": 9.88001625739183e-05, "loss": 0.0403, "step": 5820 }, { "epoch": 2.141020932794712, "grad_norm": 0.3470187485218048, "learning_rate": 9.879295084324976e-05, "loss": 0.0449, "step": 5830 }, { "epoch": 2.144693352919574, "grad_norm": 0.7424828410148621, "learning_rate": 9.87857177688868e-05, "loss": 0.0412, "step": 5840 }, { "epoch": 2.148365773044436, "grad_norm": 0.3045026957988739, "learning_rate": 9.877846335399343e-05, "loss": 0.0457, "step": 5850 }, { "epoch": 2.1520381931692985, "grad_norm": 0.35742759704589844, "learning_rate": 9.8771187601743e-05, "loss": 0.0394, "step": 5860 }, { "epoch": 2.155710613294161, "grad_norm": 0.4557603597640991, "learning_rate": 9.876389051531813e-05, "loss": 0.0407, "step": 5870 }, { "epoch": 2.159383033419023, "grad_norm": 0.45971280336380005, "learning_rate": 9.875657209791088e-05, "loss": 0.0476, "step": 5880 }, { "epoch": 2.1630554535438855, "grad_norm": 0.34561944007873535, "learning_rate": 9.874923235272248e-05, "loss": 0.0408, "step": 5890 }, { "epoch": 2.166727873668748, "grad_norm": 0.41513314843177795, "learning_rate": 9.874187128296369e-05, "loss": 0.0423, "step": 5900 }, { "epoch": 2.1704002937936098, "grad_norm": 0.28624337911605835, "learning_rate": 9.873448889185439e-05, "loss": 0.0395, "step": 5910 }, { "epoch": 2.174072713918472, "grad_norm": 0.23978790640830994, "learning_rate": 9.872708518262397e-05, "loss": 0.0415, "step": 5920 }, { "epoch": 2.1777451340433345, "grad_norm": 0.3003111183643341, "learning_rate": 9.871966015851101e-05, "loss": 0.0393, "step": 5930 }, { "epoch": 2.181417554168197, "grad_norm": 0.48093220591545105, "learning_rate": 9.87122138227635e-05, "loss": 0.0415, "step": 5940 }, { "epoch": 2.185089974293059, "grad_norm": 0.2734919488430023, "learning_rate": 9.87047461786387e-05, "loss": 0.0386, "step": 5950 }, { "epoch": 2.1887623944179215, "grad_norm": 0.22532805800437927, "learning_rate": 9.869725722940323e-05, "loss": 0.0382, "step": 5960 }, { "epoch": 2.192434814542784, "grad_norm": 0.25268667936325073, "learning_rate": 9.868974697833299e-05, "loss": 0.0402, "step": 5970 }, { "epoch": 2.196107234667646, "grad_norm": 0.3016376793384552, "learning_rate": 9.868221542871324e-05, "loss": 0.0417, "step": 5980 }, { "epoch": 2.199779654792508, "grad_norm": 0.2930961847305298, "learning_rate": 9.867466258383853e-05, "loss": 0.0414, "step": 5990 }, { "epoch": 2.2034520749173705, "grad_norm": 0.22734208405017853, "learning_rate": 9.866708844701272e-05, "loss": 0.0387, "step": 6000 }, { "epoch": 2.207124495042233, "grad_norm": 0.35137036442756653, "learning_rate": 9.865949302154899e-05, "loss": 0.0395, "step": 6010 }, { "epoch": 2.210796915167095, "grad_norm": 0.3381989002227783, "learning_rate": 9.865187631076987e-05, "loss": 0.0411, "step": 6020 }, { "epoch": 2.2144693352919576, "grad_norm": 0.3781449496746063, "learning_rate": 9.864423831800712e-05, "loss": 0.0374, "step": 6030 }, { "epoch": 2.21814175541682, "grad_norm": 0.31667575240135193, "learning_rate": 9.863657904660191e-05, "loss": 0.0422, "step": 6040 }, { "epoch": 2.221814175541682, "grad_norm": 0.3170611560344696, "learning_rate": 9.862889849990462e-05, "loss": 0.0362, "step": 6050 }, { "epoch": 2.225486595666544, "grad_norm": 0.3473859131336212, "learning_rate": 9.8621196681275e-05, "loss": 0.0397, "step": 6060 }, { "epoch": 2.2291590157914065, "grad_norm": 0.33064505457878113, "learning_rate": 9.861347359408211e-05, "loss": 0.0379, "step": 6070 }, { "epoch": 2.232831435916269, "grad_norm": 0.3282099664211273, "learning_rate": 9.860572924170426e-05, "loss": 0.0438, "step": 6080 }, { "epoch": 2.236503856041131, "grad_norm": 0.23955395817756653, "learning_rate": 9.85979636275291e-05, "loss": 0.0393, "step": 6090 }, { "epoch": 2.2401762761659936, "grad_norm": 0.24031908810138702, "learning_rate": 9.859017675495357e-05, "loss": 0.0402, "step": 6100 }, { "epoch": 2.2438486962908555, "grad_norm": 0.29474779963493347, "learning_rate": 9.858236862738392e-05, "loss": 0.0375, "step": 6110 }, { "epoch": 2.247521116415718, "grad_norm": 0.27669742703437805, "learning_rate": 9.85745392482357e-05, "loss": 0.0432, "step": 6120 }, { "epoch": 2.25119353654058, "grad_norm": 0.30519339442253113, "learning_rate": 9.856668862093372e-05, "loss": 0.0405, "step": 6130 }, { "epoch": 2.2548659566654425, "grad_norm": 0.4417056441307068, "learning_rate": 9.855881674891214e-05, "loss": 0.0397, "step": 6140 }, { "epoch": 2.258538376790305, "grad_norm": 0.5481650233268738, "learning_rate": 9.855092363561437e-05, "loss": 0.0381, "step": 6150 }, { "epoch": 2.2622107969151672, "grad_norm": 0.3048609793186188, "learning_rate": 9.854300928449312e-05, "loss": 0.0419, "step": 6160 }, { "epoch": 2.265883217040029, "grad_norm": 0.4407868981361389, "learning_rate": 9.85350736990104e-05, "loss": 0.0418, "step": 6170 }, { "epoch": 2.2695556371648915, "grad_norm": 0.36908379197120667, "learning_rate": 9.852711688263752e-05, "loss": 0.0404, "step": 6180 }, { "epoch": 2.273228057289754, "grad_norm": 0.3327307105064392, "learning_rate": 9.851913883885503e-05, "loss": 0.0407, "step": 6190 }, { "epoch": 2.276900477414616, "grad_norm": 0.25695696473121643, "learning_rate": 9.851113957115283e-05, "loss": 0.0397, "step": 6200 }, { "epoch": 2.2805728975394786, "grad_norm": 0.3382059633731842, "learning_rate": 9.850311908303005e-05, "loss": 0.0413, "step": 6210 }, { "epoch": 2.284245317664341, "grad_norm": 0.224750354886055, "learning_rate": 9.849507737799514e-05, "loss": 0.0408, "step": 6220 }, { "epoch": 2.2879177377892033, "grad_norm": 0.2615121304988861, "learning_rate": 9.84870144595658e-05, "loss": 0.0387, "step": 6230 }, { "epoch": 2.291590157914065, "grad_norm": 0.3430602252483368, "learning_rate": 9.847893033126903e-05, "loss": 0.0407, "step": 6240 }, { "epoch": 2.2952625780389275, "grad_norm": 0.32987478375434875, "learning_rate": 9.847082499664111e-05, "loss": 0.0392, "step": 6250 }, { "epoch": 2.29893499816379, "grad_norm": 0.3321605920791626, "learning_rate": 9.846269845922758e-05, "loss": 0.0422, "step": 6260 }, { "epoch": 2.3026074182886522, "grad_norm": 0.23884893953800201, "learning_rate": 9.845455072258326e-05, "loss": 0.0403, "step": 6270 }, { "epoch": 2.3062798384135146, "grad_norm": 0.3631207048892975, "learning_rate": 9.844638179027226e-05, "loss": 0.0399, "step": 6280 }, { "epoch": 2.309952258538377, "grad_norm": 0.4131614565849304, "learning_rate": 9.843819166586795e-05, "loss": 0.0433, "step": 6290 }, { "epoch": 2.3136246786632393, "grad_norm": 0.2798607647418976, "learning_rate": 9.842998035295294e-05, "loss": 0.0398, "step": 6300 }, { "epoch": 2.317297098788101, "grad_norm": 0.2824913263320923, "learning_rate": 9.842174785511918e-05, "loss": 0.0418, "step": 6310 }, { "epoch": 2.3209695189129635, "grad_norm": 0.3250288665294647, "learning_rate": 9.841349417596779e-05, "loss": 0.0387, "step": 6320 }, { "epoch": 2.324641939037826, "grad_norm": 0.4131326973438263, "learning_rate": 9.840521931910926e-05, "loss": 0.0393, "step": 6330 }, { "epoch": 2.3283143591626883, "grad_norm": 0.3156285583972931, "learning_rate": 9.839692328816327e-05, "loss": 0.0369, "step": 6340 }, { "epoch": 2.3319867792875506, "grad_norm": 0.44367071986198425, "learning_rate": 9.838860608675879e-05, "loss": 0.0381, "step": 6350 }, { "epoch": 2.335659199412413, "grad_norm": 0.2765277624130249, "learning_rate": 9.838026771853406e-05, "loss": 0.0395, "step": 6360 }, { "epoch": 2.3393316195372753, "grad_norm": 0.38412028551101685, "learning_rate": 9.837190818713655e-05, "loss": 0.0387, "step": 6370 }, { "epoch": 2.343004039662137, "grad_norm": 0.3159184753894806, "learning_rate": 9.836352749622299e-05, "loss": 0.0389, "step": 6380 }, { "epoch": 2.3466764597869996, "grad_norm": 0.3176368176937103, "learning_rate": 9.835512564945941e-05, "loss": 0.0399, "step": 6390 }, { "epoch": 2.350348879911862, "grad_norm": 0.24991914629936218, "learning_rate": 9.834670265052104e-05, "loss": 0.0392, "step": 6400 }, { "epoch": 2.3540213000367243, "grad_norm": 0.39544135332107544, "learning_rate": 9.833825850309239e-05, "loss": 0.0378, "step": 6410 }, { "epoch": 2.3576937201615866, "grad_norm": 0.24084070324897766, "learning_rate": 9.832979321086723e-05, "loss": 0.0407, "step": 6420 }, { "epoch": 2.3613661402864485, "grad_norm": 0.5084562301635742, "learning_rate": 9.832130677754854e-05, "loss": 0.0367, "step": 6430 }, { "epoch": 2.365038560411311, "grad_norm": 0.41990840435028076, "learning_rate": 9.831279920684861e-05, "loss": 0.0383, "step": 6440 }, { "epoch": 2.3687109805361732, "grad_norm": 0.37860527634620667, "learning_rate": 9.830427050248891e-05, "loss": 0.0371, "step": 6450 }, { "epoch": 2.3723834006610356, "grad_norm": 0.2337464690208435, "learning_rate": 9.829572066820021e-05, "loss": 0.0391, "step": 6460 }, { "epoch": 2.376055820785898, "grad_norm": 0.26003801822662354, "learning_rate": 9.828714970772247e-05, "loss": 0.0393, "step": 6470 }, { "epoch": 2.3797282409107603, "grad_norm": 0.22383326292037964, "learning_rate": 9.827855762480493e-05, "loss": 0.039, "step": 6480 }, { "epoch": 2.3834006610356226, "grad_norm": 0.2611485421657562, "learning_rate": 9.826994442320608e-05, "loss": 0.0382, "step": 6490 }, { "epoch": 2.3870730811604846, "grad_norm": 0.3399944007396698, "learning_rate": 9.82613101066936e-05, "loss": 0.0386, "step": 6500 }, { "epoch": 2.390745501285347, "grad_norm": 0.30107349157333374, "learning_rate": 9.825265467904445e-05, "loss": 0.0385, "step": 6510 }, { "epoch": 2.3944179214102093, "grad_norm": 0.31851184368133545, "learning_rate": 9.82439781440448e-05, "loss": 0.0405, "step": 6520 }, { "epoch": 2.3980903415350716, "grad_norm": 0.45066890120506287, "learning_rate": 9.823528050549006e-05, "loss": 0.0403, "step": 6530 }, { "epoch": 2.401762761659934, "grad_norm": 0.4221578538417816, "learning_rate": 9.822656176718487e-05, "loss": 0.0398, "step": 6540 }, { "epoch": 2.4054351817847963, "grad_norm": 0.37969520688056946, "learning_rate": 9.82178219329431e-05, "loss": 0.0393, "step": 6550 }, { "epoch": 2.4091076019096587, "grad_norm": 0.29835575819015503, "learning_rate": 9.820906100658789e-05, "loss": 0.0396, "step": 6560 }, { "epoch": 2.4127800220345206, "grad_norm": 0.3965451717376709, "learning_rate": 9.82002789919515e-05, "loss": 0.0426, "step": 6570 }, { "epoch": 2.416452442159383, "grad_norm": 0.34618157148361206, "learning_rate": 9.819147589287554e-05, "loss": 0.044, "step": 6580 }, { "epoch": 2.4201248622842453, "grad_norm": 0.3138091266155243, "learning_rate": 9.818265171321074e-05, "loss": 0.0397, "step": 6590 }, { "epoch": 2.4237972824091076, "grad_norm": 0.3529050648212433, "learning_rate": 9.817380645681714e-05, "loss": 0.0404, "step": 6600 }, { "epoch": 2.42746970253397, "grad_norm": 0.2973199486732483, "learning_rate": 9.816494012756392e-05, "loss": 0.0411, "step": 6610 }, { "epoch": 2.4311421226588323, "grad_norm": 0.2867656946182251, "learning_rate": 9.815605272932953e-05, "loss": 0.0391, "step": 6620 }, { "epoch": 2.4348145427836947, "grad_norm": 0.3809772729873657, "learning_rate": 9.814714426600162e-05, "loss": 0.0435, "step": 6630 }, { "epoch": 2.4384869629085566, "grad_norm": 0.33727070689201355, "learning_rate": 9.813821474147704e-05, "loss": 0.0392, "step": 6640 }, { "epoch": 2.442159383033419, "grad_norm": 0.2759070098400116, "learning_rate": 9.812926415966189e-05, "loss": 0.0376, "step": 6650 }, { "epoch": 2.4458318031582813, "grad_norm": 0.40307384729385376, "learning_rate": 9.812029252447143e-05, "loss": 0.0401, "step": 6660 }, { "epoch": 2.4495042232831437, "grad_norm": 0.29866182804107666, "learning_rate": 9.811129983983018e-05, "loss": 0.0397, "step": 6670 }, { "epoch": 2.453176643408006, "grad_norm": 0.36344560980796814, "learning_rate": 9.810228610967185e-05, "loss": 0.0416, "step": 6680 }, { "epoch": 2.456849063532868, "grad_norm": 0.37325891852378845, "learning_rate": 9.809325133793934e-05, "loss": 0.0379, "step": 6690 }, { "epoch": 2.4605214836577303, "grad_norm": 0.26771870255470276, "learning_rate": 9.808419552858477e-05, "loss": 0.0398, "step": 6700 }, { "epoch": 2.4641939037825926, "grad_norm": 0.29054006934165955, "learning_rate": 9.807511868556944e-05, "loss": 0.0381, "step": 6710 }, { "epoch": 2.467866323907455, "grad_norm": 0.25668230652809143, "learning_rate": 9.806602081286392e-05, "loss": 0.0407, "step": 6720 }, { "epoch": 2.4715387440323173, "grad_norm": 0.42105695605278015, "learning_rate": 9.805690191444785e-05, "loss": 0.0395, "step": 6730 }, { "epoch": 2.4752111641571797, "grad_norm": 0.3615626394748688, "learning_rate": 9.804776199431022e-05, "loss": 0.0421, "step": 6740 }, { "epoch": 2.478883584282042, "grad_norm": 0.2102382481098175, "learning_rate": 9.80386010564491e-05, "loss": 0.0369, "step": 6750 }, { "epoch": 2.482556004406904, "grad_norm": 0.32103022933006287, "learning_rate": 9.802941910487183e-05, "loss": 0.0374, "step": 6760 }, { "epoch": 2.4862284245317663, "grad_norm": 0.28580087423324585, "learning_rate": 9.802021614359487e-05, "loss": 0.0395, "step": 6770 }, { "epoch": 2.4899008446566286, "grad_norm": 0.2996770143508911, "learning_rate": 9.801099217664394e-05, "loss": 0.0373, "step": 6780 }, { "epoch": 2.493573264781491, "grad_norm": 0.3614078462123871, "learning_rate": 9.800174720805388e-05, "loss": 0.037, "step": 6790 }, { "epoch": 2.4972456849063533, "grad_norm": 0.5022622346878052, "learning_rate": 9.799248124186878e-05, "loss": 0.0431, "step": 6800 }, { "epoch": 2.5009181050312157, "grad_norm": 0.393021821975708, "learning_rate": 9.798319428214187e-05, "loss": 0.0365, "step": 6810 }, { "epoch": 2.504590525156078, "grad_norm": 0.25063011050224304, "learning_rate": 9.79738863329356e-05, "loss": 0.0377, "step": 6820 }, { "epoch": 2.50826294528094, "grad_norm": 0.3842640519142151, "learning_rate": 9.79645573983216e-05, "loss": 0.0353, "step": 6830 }, { "epoch": 2.5119353654058023, "grad_norm": 0.32475659251213074, "learning_rate": 9.795520748238062e-05, "loss": 0.0371, "step": 6840 }, { "epoch": 2.5156077855306647, "grad_norm": 0.4488423764705658, "learning_rate": 9.794583658920264e-05, "loss": 0.035, "step": 6850 }, { "epoch": 2.519280205655527, "grad_norm": 0.38933780789375305, "learning_rate": 9.793644472288683e-05, "loss": 0.0413, "step": 6860 }, { "epoch": 2.5229526257803894, "grad_norm": 0.4778420925140381, "learning_rate": 9.79270318875415e-05, "loss": 0.0402, "step": 6870 }, { "epoch": 2.5266250459052517, "grad_norm": 0.34345772862434387, "learning_rate": 9.791759808728416e-05, "loss": 0.0383, "step": 6880 }, { "epoch": 2.530297466030114, "grad_norm": 0.2596583664417267, "learning_rate": 9.790814332624143e-05, "loss": 0.0378, "step": 6890 }, { "epoch": 2.533969886154976, "grad_norm": 0.6046827435493469, "learning_rate": 9.789866760854919e-05, "loss": 0.0402, "step": 6900 }, { "epoch": 2.5376423062798383, "grad_norm": 0.3346410393714905, "learning_rate": 9.788917093835242e-05, "loss": 0.0402, "step": 6910 }, { "epoch": 2.5413147264047007, "grad_norm": 0.40887513756752014, "learning_rate": 9.787965331980528e-05, "loss": 0.042, "step": 6920 }, { "epoch": 2.544987146529563, "grad_norm": 0.23225101828575134, "learning_rate": 9.787011475707111e-05, "loss": 0.0379, "step": 6930 }, { "epoch": 2.5486595666544254, "grad_norm": 0.20746177434921265, "learning_rate": 9.786055525432241e-05, "loss": 0.0371, "step": 6940 }, { "epoch": 2.5523319867792873, "grad_norm": 0.3574647009372711, "learning_rate": 9.78509748157408e-05, "loss": 0.0422, "step": 6950 }, { "epoch": 2.55600440690415, "grad_norm": 0.32719355821609497, "learning_rate": 9.784137344551713e-05, "loss": 0.0373, "step": 6960 }, { "epoch": 2.559676827029012, "grad_norm": 0.24420630931854248, "learning_rate": 9.783175114785134e-05, "loss": 0.0397, "step": 6970 }, { "epoch": 2.5633492471538744, "grad_norm": 0.2932735085487366, "learning_rate": 9.782210792695254e-05, "loss": 0.0388, "step": 6980 }, { "epoch": 2.5670216672787367, "grad_norm": 0.4242243766784668, "learning_rate": 9.781244378703904e-05, "loss": 0.0411, "step": 6990 }, { "epoch": 2.570694087403599, "grad_norm": 0.29554054141044617, "learning_rate": 9.780275873233824e-05, "loss": 0.0386, "step": 7000 }, { "epoch": 2.5743665075284614, "grad_norm": 0.3440375030040741, "learning_rate": 9.77930527670867e-05, "loss": 0.0372, "step": 7010 }, { "epoch": 2.5780389276533233, "grad_norm": 0.344284325838089, "learning_rate": 9.778332589553016e-05, "loss": 0.0367, "step": 7020 }, { "epoch": 2.5817113477781857, "grad_norm": 0.20114487409591675, "learning_rate": 9.777357812192349e-05, "loss": 0.0392, "step": 7030 }, { "epoch": 2.585383767903048, "grad_norm": 0.2918180227279663, "learning_rate": 9.776380945053066e-05, "loss": 0.0385, "step": 7040 }, { "epoch": 2.5890561880279104, "grad_norm": 0.2653922438621521, "learning_rate": 9.775401988562486e-05, "loss": 0.0406, "step": 7050 }, { "epoch": 2.5927286081527727, "grad_norm": 0.2555534541606903, "learning_rate": 9.774420943148835e-05, "loss": 0.0364, "step": 7060 }, { "epoch": 2.596401028277635, "grad_norm": 0.30115899443626404, "learning_rate": 9.773437809241257e-05, "loss": 0.0394, "step": 7070 }, { "epoch": 2.6000734484024974, "grad_norm": 0.40862637758255005, "learning_rate": 9.772452587269808e-05, "loss": 0.0362, "step": 7080 }, { "epoch": 2.6037458685273593, "grad_norm": 0.3946945071220398, "learning_rate": 9.771465277665457e-05, "loss": 0.0399, "step": 7090 }, { "epoch": 2.6074182886522217, "grad_norm": 0.38057777285575867, "learning_rate": 9.770475880860089e-05, "loss": 0.0399, "step": 7100 }, { "epoch": 2.611090708777084, "grad_norm": 0.2868947982788086, "learning_rate": 9.769484397286495e-05, "loss": 0.0366, "step": 7110 }, { "epoch": 2.6147631289019464, "grad_norm": 0.3568795621395111, "learning_rate": 9.768490827378388e-05, "loss": 0.0395, "step": 7120 }, { "epoch": 2.6184355490268088, "grad_norm": 0.7191362380981445, "learning_rate": 9.767495171570387e-05, "loss": 0.0392, "step": 7130 }, { "epoch": 2.622107969151671, "grad_norm": 0.3663073778152466, "learning_rate": 9.766497430298027e-05, "loss": 0.0394, "step": 7140 }, { "epoch": 2.6257803892765335, "grad_norm": 0.26784226298332214, "learning_rate": 9.765497603997754e-05, "loss": 0.041, "step": 7150 }, { "epoch": 2.6294528094013954, "grad_norm": 0.34962716698646545, "learning_rate": 9.764495693106926e-05, "loss": 0.0403, "step": 7160 }, { "epoch": 2.6331252295262577, "grad_norm": 0.26591208577156067, "learning_rate": 9.76349169806381e-05, "loss": 0.0375, "step": 7170 }, { "epoch": 2.63679764965112, "grad_norm": 0.28928476572036743, "learning_rate": 9.762485619307591e-05, "loss": 0.0386, "step": 7180 }, { "epoch": 2.6404700697759824, "grad_norm": 0.39672791957855225, "learning_rate": 9.76147745727836e-05, "loss": 0.043, "step": 7190 }, { "epoch": 2.6441424899008448, "grad_norm": 0.2678619921207428, "learning_rate": 9.760467212417124e-05, "loss": 0.0383, "step": 7200 }, { "epoch": 2.6478149100257067, "grad_norm": 0.31879428029060364, "learning_rate": 9.759454885165796e-05, "loss": 0.0368, "step": 7210 }, { "epoch": 2.6514873301505695, "grad_norm": 0.40957731008529663, "learning_rate": 9.7584404759672e-05, "loss": 0.0407, "step": 7220 }, { "epoch": 2.6551597502754314, "grad_norm": 0.2298721969127655, "learning_rate": 9.75742398526508e-05, "loss": 0.0368, "step": 7230 }, { "epoch": 2.6588321704002937, "grad_norm": 0.23018202185630798, "learning_rate": 9.756405413504077e-05, "loss": 0.0374, "step": 7240 }, { "epoch": 2.662504590525156, "grad_norm": 0.460349440574646, "learning_rate": 9.755384761129752e-05, "loss": 0.0354, "step": 7250 }, { "epoch": 2.6661770106500184, "grad_norm": 0.33253908157348633, "learning_rate": 9.754362028588572e-05, "loss": 0.0374, "step": 7260 }, { "epoch": 2.669849430774881, "grad_norm": 0.36773982644081116, "learning_rate": 9.753337216327917e-05, "loss": 0.0363, "step": 7270 }, { "epoch": 2.6735218508997427, "grad_norm": 0.27943527698516846, "learning_rate": 9.752310324796073e-05, "loss": 0.0382, "step": 7280 }, { "epoch": 2.677194271024605, "grad_norm": 0.30617910623550415, "learning_rate": 9.751281354442235e-05, "loss": 0.0375, "step": 7290 }, { "epoch": 2.6808666911494674, "grad_norm": 0.24900229275226593, "learning_rate": 9.750250305716513e-05, "loss": 0.0387, "step": 7300 }, { "epoch": 2.6845391112743298, "grad_norm": 0.32763010263442993, "learning_rate": 9.749217179069923e-05, "loss": 0.0394, "step": 7310 }, { "epoch": 2.688211531399192, "grad_norm": 0.4690152704715729, "learning_rate": 9.748181974954386e-05, "loss": 0.0372, "step": 7320 }, { "epoch": 2.6918839515240545, "grad_norm": 0.2899266481399536, "learning_rate": 9.74714469382274e-05, "loss": 0.0397, "step": 7330 }, { "epoch": 2.695556371648917, "grad_norm": 0.30799975991249084, "learning_rate": 9.746105336128724e-05, "loss": 0.0349, "step": 7340 }, { "epoch": 2.6992287917737787, "grad_norm": 0.26266714930534363, "learning_rate": 9.745063902326988e-05, "loss": 0.0366, "step": 7350 }, { "epoch": 2.702901211898641, "grad_norm": 0.31730198860168457, "learning_rate": 9.744020392873093e-05, "loss": 0.036, "step": 7360 }, { "epoch": 2.7065736320235034, "grad_norm": 0.38839370012283325, "learning_rate": 9.742974808223504e-05, "loss": 0.037, "step": 7370 }, { "epoch": 2.710246052148366, "grad_norm": 0.30720117688179016, "learning_rate": 9.741927148835596e-05, "loss": 0.0397, "step": 7380 }, { "epoch": 2.713918472273228, "grad_norm": 0.21320083737373352, "learning_rate": 9.740877415167648e-05, "loss": 0.0388, "step": 7390 }, { "epoch": 2.7175908923980905, "grad_norm": 0.3118077218532562, "learning_rate": 9.739825607678854e-05, "loss": 0.0379, "step": 7400 }, { "epoch": 2.721263312522953, "grad_norm": 0.31821393966674805, "learning_rate": 9.738771726829308e-05, "loss": 0.0377, "step": 7410 }, { "epoch": 2.7249357326478147, "grad_norm": 0.3700689375400543, "learning_rate": 9.737715773080011e-05, "loss": 0.0403, "step": 7420 }, { "epoch": 2.728608152772677, "grad_norm": 0.27187633514404297, "learning_rate": 9.736657746892877e-05, "loss": 0.038, "step": 7430 }, { "epoch": 2.7322805728975395, "grad_norm": 0.241797536611557, "learning_rate": 9.735597648730718e-05, "loss": 0.041, "step": 7440 }, { "epoch": 2.735952993022402, "grad_norm": 0.43704837560653687, "learning_rate": 9.734535479057262e-05, "loss": 0.0383, "step": 7450 }, { "epoch": 2.739625413147264, "grad_norm": 0.36168622970581055, "learning_rate": 9.733471238337136e-05, "loss": 0.0383, "step": 7460 }, { "epoch": 2.743297833272126, "grad_norm": 0.35653817653656006, "learning_rate": 9.732404927035871e-05, "loss": 0.0384, "step": 7470 }, { "epoch": 2.746970253396989, "grad_norm": 0.37300729751586914, "learning_rate": 9.731336545619915e-05, "loss": 0.0364, "step": 7480 }, { "epoch": 2.7506426735218508, "grad_norm": 0.2261621654033661, "learning_rate": 9.73026609455661e-05, "loss": 0.0399, "step": 7490 }, { "epoch": 2.754315093646713, "grad_norm": 0.23663759231567383, "learning_rate": 9.72919357431421e-05, "loss": 0.0364, "step": 7500 }, { "epoch": 2.7579875137715755, "grad_norm": 0.30120882391929626, "learning_rate": 9.72811898536187e-05, "loss": 0.0361, "step": 7510 }, { "epoch": 2.761659933896438, "grad_norm": 0.27787405252456665, "learning_rate": 9.72704232816965e-05, "loss": 0.0417, "step": 7520 }, { "epoch": 2.7653323540213, "grad_norm": 0.32935959100723267, "learning_rate": 9.725963603208519e-05, "loss": 0.0393, "step": 7530 }, { "epoch": 2.769004774146162, "grad_norm": 0.3514164388179779, "learning_rate": 9.724882810950348e-05, "loss": 0.0353, "step": 7540 }, { "epoch": 2.772677194271025, "grad_norm": 0.3228343427181244, "learning_rate": 9.723799951867913e-05, "loss": 0.039, "step": 7550 }, { "epoch": 2.776349614395887, "grad_norm": 0.3400120735168457, "learning_rate": 9.722715026434889e-05, "loss": 0.0356, "step": 7560 }, { "epoch": 2.780022034520749, "grad_norm": 0.30100712180137634, "learning_rate": 9.721628035125866e-05, "loss": 0.038, "step": 7570 }, { "epoch": 2.7836944546456115, "grad_norm": 0.28605931997299194, "learning_rate": 9.720538978416325e-05, "loss": 0.0381, "step": 7580 }, { "epoch": 2.787366874770474, "grad_norm": 0.37283098697662354, "learning_rate": 9.71944785678266e-05, "loss": 0.0379, "step": 7590 }, { "epoch": 2.791039294895336, "grad_norm": 0.22697977721691132, "learning_rate": 9.718354670702161e-05, "loss": 0.0391, "step": 7600 }, { "epoch": 2.794711715020198, "grad_norm": 0.35865554213523865, "learning_rate": 9.717259420653029e-05, "loss": 0.0374, "step": 7610 }, { "epoch": 2.7983841351450605, "grad_norm": 0.28434979915618896, "learning_rate": 9.716162107114361e-05, "loss": 0.0377, "step": 7620 }, { "epoch": 2.802056555269923, "grad_norm": 0.9750490188598633, "learning_rate": 9.71506273056616e-05, "loss": 0.0363, "step": 7630 }, { "epoch": 2.805728975394785, "grad_norm": 0.35271456837654114, "learning_rate": 9.713961291489331e-05, "loss": 0.0369, "step": 7640 }, { "epoch": 2.8094013955196475, "grad_norm": 0.34873703122138977, "learning_rate": 9.71285779036568e-05, "loss": 0.0393, "step": 7650 }, { "epoch": 2.81307381564451, "grad_norm": 0.38175368309020996, "learning_rate": 9.711752227677916e-05, "loss": 0.036, "step": 7660 }, { "epoch": 2.816746235769372, "grad_norm": 0.3082677721977234, "learning_rate": 9.710644603909649e-05, "loss": 0.0359, "step": 7670 }, { "epoch": 2.820418655894234, "grad_norm": 0.29651573300361633, "learning_rate": 9.709534919545393e-05, "loss": 0.0367, "step": 7680 }, { "epoch": 2.8240910760190965, "grad_norm": 0.2999531030654907, "learning_rate": 9.708423175070563e-05, "loss": 0.0377, "step": 7690 }, { "epoch": 2.827763496143959, "grad_norm": 0.3805941045284271, "learning_rate": 9.707309370971468e-05, "loss": 0.0379, "step": 7700 }, { "epoch": 2.831435916268821, "grad_norm": 0.22950077056884766, "learning_rate": 9.70619350773533e-05, "loss": 0.0405, "step": 7710 }, { "epoch": 2.8351083363936835, "grad_norm": 0.30463188886642456, "learning_rate": 9.70507558585026e-05, "loss": 0.0371, "step": 7720 }, { "epoch": 2.8387807565185454, "grad_norm": 0.20810405910015106, "learning_rate": 9.703955605805281e-05, "loss": 0.0389, "step": 7730 }, { "epoch": 2.8424531766434082, "grad_norm": 0.4523184299468994, "learning_rate": 9.702833568090306e-05, "loss": 0.0374, "step": 7740 }, { "epoch": 2.84612559676827, "grad_norm": 0.34106430411338806, "learning_rate": 9.701709473196154e-05, "loss": 0.0373, "step": 7750 }, { "epoch": 2.8497980168931325, "grad_norm": 0.46462154388427734, "learning_rate": 9.700583321614541e-05, "loss": 0.0375, "step": 7760 }, { "epoch": 2.853470437017995, "grad_norm": 0.33843153715133667, "learning_rate": 9.699455113838085e-05, "loss": 0.0365, "step": 7770 }, { "epoch": 2.857142857142857, "grad_norm": 0.42227113246917725, "learning_rate": 9.698324850360303e-05, "loss": 0.0394, "step": 7780 }, { "epoch": 2.8608152772677196, "grad_norm": 0.22156544029712677, "learning_rate": 9.69719253167561e-05, "loss": 0.037, "step": 7790 }, { "epoch": 2.8644876973925815, "grad_norm": 0.18959300220012665, "learning_rate": 9.69605815827932e-05, "loss": 0.0378, "step": 7800 }, { "epoch": 2.8681601175174443, "grad_norm": 0.3028106391429901, "learning_rate": 9.694921730667647e-05, "loss": 0.0362, "step": 7810 }, { "epoch": 2.871832537642306, "grad_norm": 0.30171793699264526, "learning_rate": 9.693783249337705e-05, "loss": 0.04, "step": 7820 }, { "epoch": 2.8755049577671685, "grad_norm": 0.28248101472854614, "learning_rate": 9.692642714787501e-05, "loss": 0.0398, "step": 7830 }, { "epoch": 2.879177377892031, "grad_norm": 0.29408887028694153, "learning_rate": 9.691500127515945e-05, "loss": 0.038, "step": 7840 }, { "epoch": 2.8828497980168932, "grad_norm": 0.3004066050052643, "learning_rate": 9.690355488022844e-05, "loss": 0.037, "step": 7850 }, { "epoch": 2.8865222181417556, "grad_norm": 0.40725618600845337, "learning_rate": 9.689208796808902e-05, "loss": 0.0338, "step": 7860 }, { "epoch": 2.8901946382666175, "grad_norm": 0.34566235542297363, "learning_rate": 9.68806005437572e-05, "loss": 0.0367, "step": 7870 }, { "epoch": 2.89386705839148, "grad_norm": 0.29805320501327515, "learning_rate": 9.686909261225796e-05, "loss": 0.0365, "step": 7880 }, { "epoch": 2.897539478516342, "grad_norm": 0.34649136662483215, "learning_rate": 9.685756417862531e-05, "loss": 0.0367, "step": 7890 }, { "epoch": 2.9012118986412045, "grad_norm": 0.2619168162345886, "learning_rate": 9.684601524790212e-05, "loss": 0.0365, "step": 7900 }, { "epoch": 2.904884318766067, "grad_norm": 0.2785470485687256, "learning_rate": 9.68344458251403e-05, "loss": 0.0379, "step": 7910 }, { "epoch": 2.9085567388909292, "grad_norm": 0.2577872574329376, "learning_rate": 9.682285591540072e-05, "loss": 0.0359, "step": 7920 }, { "epoch": 2.9122291590157916, "grad_norm": 0.46411389112472534, "learning_rate": 9.681124552375322e-05, "loss": 0.0373, "step": 7930 }, { "epoch": 2.9159015791406535, "grad_norm": 0.31439897418022156, "learning_rate": 9.679961465527654e-05, "loss": 0.0353, "step": 7940 }, { "epoch": 2.919573999265516, "grad_norm": 0.4401150047779083, "learning_rate": 9.678796331505843e-05, "loss": 0.0394, "step": 7950 }, { "epoch": 2.923246419390378, "grad_norm": 0.3058633804321289, "learning_rate": 9.677629150819558e-05, "loss": 0.0403, "step": 7960 }, { "epoch": 2.9269188395152406, "grad_norm": 0.31861600279808044, "learning_rate": 9.676459923979366e-05, "loss": 0.0372, "step": 7970 }, { "epoch": 2.930591259640103, "grad_norm": 0.3158496022224426, "learning_rate": 9.675288651496723e-05, "loss": 0.0429, "step": 7980 }, { "epoch": 2.9342636797649653, "grad_norm": 0.3432671129703522, "learning_rate": 9.674115333883986e-05, "loss": 0.0376, "step": 7990 }, { "epoch": 2.9379360998898276, "grad_norm": 0.4189271628856659, "learning_rate": 9.672939971654402e-05, "loss": 0.0356, "step": 8000 }, { "epoch": 2.9416085200146895, "grad_norm": 0.47407209873199463, "learning_rate": 9.671762565322117e-05, "loss": 0.0382, "step": 8010 }, { "epoch": 2.945280940139552, "grad_norm": 0.2090795487165451, "learning_rate": 9.670583115402166e-05, "loss": 0.0364, "step": 8020 }, { "epoch": 2.9489533602644142, "grad_norm": 0.2671118974685669, "learning_rate": 9.669401622410482e-05, "loss": 0.0365, "step": 8030 }, { "epoch": 2.9526257803892766, "grad_norm": 0.2975383400917053, "learning_rate": 9.668218086863887e-05, "loss": 0.0353, "step": 8040 }, { "epoch": 2.956298200514139, "grad_norm": 0.4209805727005005, "learning_rate": 9.667032509280103e-05, "loss": 0.0357, "step": 8050 }, { "epoch": 2.959970620639001, "grad_norm": 0.24457359313964844, "learning_rate": 9.665844890177743e-05, "loss": 0.0367, "step": 8060 }, { "epoch": 2.9636430407638636, "grad_norm": 0.27309420704841614, "learning_rate": 9.664655230076309e-05, "loss": 0.0369, "step": 8070 }, { "epoch": 2.9673154608887256, "grad_norm": 0.47591349482536316, "learning_rate": 9.663463529496199e-05, "loss": 0.0364, "step": 8080 }, { "epoch": 2.970987881013588, "grad_norm": 0.32411056756973267, "learning_rate": 9.662269788958705e-05, "loss": 0.0367, "step": 8090 }, { "epoch": 2.9746603011384503, "grad_norm": 0.3758200407028198, "learning_rate": 9.661074008986008e-05, "loss": 0.0357, "step": 8100 }, { "epoch": 2.9783327212633126, "grad_norm": 0.3297320306301117, "learning_rate": 9.659876190101184e-05, "loss": 0.0361, "step": 8110 }, { "epoch": 2.982005141388175, "grad_norm": 0.38448846340179443, "learning_rate": 9.6586763328282e-05, "loss": 0.0358, "step": 8120 }, { "epoch": 2.985677561513037, "grad_norm": 0.2986893653869629, "learning_rate": 9.657474437691913e-05, "loss": 0.0362, "step": 8130 }, { "epoch": 2.9893499816378992, "grad_norm": 0.38175132870674133, "learning_rate": 9.656270505218073e-05, "loss": 0.0365, "step": 8140 }, { "epoch": 2.9930224017627616, "grad_norm": 0.3190513551235199, "learning_rate": 9.655064535933321e-05, "loss": 0.0378, "step": 8150 }, { "epoch": 2.996694821887624, "grad_norm": 0.2510859966278076, "learning_rate": 9.653856530365189e-05, "loss": 0.0359, "step": 8160 }, { "epoch": 3.0003672420124863, "grad_norm": 0.3045142889022827, "learning_rate": 9.6526464890421e-05, "loss": 0.0364, "step": 8170 }, { "epoch": 3.0040396621373486, "grad_norm": 0.3480423092842102, "learning_rate": 9.651434412493367e-05, "loss": 0.0372, "step": 8180 }, { "epoch": 3.007712082262211, "grad_norm": 0.31054338812828064, "learning_rate": 9.650220301249195e-05, "loss": 0.0353, "step": 8190 }, { "epoch": 3.011384502387073, "grad_norm": 0.28488242626190186, "learning_rate": 9.649004155840675e-05, "loss": 0.0349, "step": 8200 }, { "epoch": 3.0150569225119352, "grad_norm": 0.21094581484794617, "learning_rate": 9.647785976799792e-05, "loss": 0.0376, "step": 8210 }, { "epoch": 3.0187293426367976, "grad_norm": 0.2984619140625, "learning_rate": 9.646565764659417e-05, "loss": 0.0359, "step": 8220 }, { "epoch": 3.02240176276166, "grad_norm": 0.3793299198150635, "learning_rate": 9.645343519953314e-05, "loss": 0.0374, "step": 8230 }, { "epoch": 3.0260741828865223, "grad_norm": 0.27605247497558594, "learning_rate": 9.644119243216135e-05, "loss": 0.0339, "step": 8240 }, { "epoch": 3.0297466030113847, "grad_norm": 0.3164132833480835, "learning_rate": 9.642892934983417e-05, "loss": 0.034, "step": 8250 }, { "epoch": 3.033419023136247, "grad_norm": 0.337787389755249, "learning_rate": 9.641664595791593e-05, "loss": 0.0352, "step": 8260 }, { "epoch": 3.037091443261109, "grad_norm": 0.4062500298023224, "learning_rate": 9.640434226177977e-05, "loss": 0.0418, "step": 8270 }, { "epoch": 3.0407638633859713, "grad_norm": 0.2267233431339264, "learning_rate": 9.639201826680777e-05, "loss": 0.0366, "step": 8280 }, { "epoch": 3.0444362835108336, "grad_norm": 0.2844814658164978, "learning_rate": 9.637967397839083e-05, "loss": 0.036, "step": 8290 }, { "epoch": 3.048108703635696, "grad_norm": 0.28603658080101013, "learning_rate": 9.63673094019288e-05, "loss": 0.0326, "step": 8300 }, { "epoch": 3.0517811237605583, "grad_norm": 0.3627729117870331, "learning_rate": 9.635492454283035e-05, "loss": 0.0393, "step": 8310 }, { "epoch": 3.0554535438854207, "grad_norm": 0.3649597465991974, "learning_rate": 9.634251940651302e-05, "loss": 0.0377, "step": 8320 }, { "epoch": 3.0591259640102826, "grad_norm": 0.452258825302124, "learning_rate": 9.633009399840327e-05, "loss": 0.0369, "step": 8330 }, { "epoch": 3.062798384135145, "grad_norm": 0.3374057412147522, "learning_rate": 9.631764832393639e-05, "loss": 0.0346, "step": 8340 }, { "epoch": 3.0664708042600073, "grad_norm": 0.38793671131134033, "learning_rate": 9.630518238855651e-05, "loss": 0.038, "step": 8350 }, { "epoch": 3.0701432243848696, "grad_norm": 0.32766595482826233, "learning_rate": 9.629269619771668e-05, "loss": 0.0357, "step": 8360 }, { "epoch": 3.073815644509732, "grad_norm": 0.24933235347270966, "learning_rate": 9.628018975687878e-05, "loss": 0.0335, "step": 8370 }, { "epoch": 3.0774880646345943, "grad_norm": 0.30373960733413696, "learning_rate": 9.626766307151355e-05, "loss": 0.0357, "step": 8380 }, { "epoch": 3.0811604847594563, "grad_norm": 0.32706519961357117, "learning_rate": 9.62551161471006e-05, "loss": 0.0364, "step": 8390 }, { "epoch": 3.0848329048843186, "grad_norm": 0.3051760196685791, "learning_rate": 9.624254898912837e-05, "loss": 0.0354, "step": 8400 }, { "epoch": 3.088505325009181, "grad_norm": 0.21841676533222198, "learning_rate": 9.622996160309414e-05, "loss": 0.0352, "step": 8410 }, { "epoch": 3.0921777451340433, "grad_norm": 0.37986311316490173, "learning_rate": 9.62173539945041e-05, "loss": 0.0359, "step": 8420 }, { "epoch": 3.0958501652589057, "grad_norm": 0.21062558889389038, "learning_rate": 9.620472616887323e-05, "loss": 0.0347, "step": 8430 }, { "epoch": 3.099522585383768, "grad_norm": 0.42726781964302063, "learning_rate": 9.619207813172536e-05, "loss": 0.0344, "step": 8440 }, { "epoch": 3.1031950055086304, "grad_norm": 0.36414772272109985, "learning_rate": 9.617940988859318e-05, "loss": 0.0349, "step": 8450 }, { "epoch": 3.1068674256334923, "grad_norm": 0.31926479935646057, "learning_rate": 9.616672144501821e-05, "loss": 0.0345, "step": 8460 }, { "epoch": 3.1105398457583546, "grad_norm": 0.2954327166080475, "learning_rate": 9.615401280655081e-05, "loss": 0.0331, "step": 8470 }, { "epoch": 3.114212265883217, "grad_norm": 0.21420814096927643, "learning_rate": 9.614128397875017e-05, "loss": 0.0396, "step": 8480 }, { "epoch": 3.1178846860080793, "grad_norm": 0.2617756426334381, "learning_rate": 9.612853496718429e-05, "loss": 0.0368, "step": 8490 }, { "epoch": 3.1215571061329417, "grad_norm": 0.3212507367134094, "learning_rate": 9.611576577743004e-05, "loss": 0.0332, "step": 8500 }, { "epoch": 3.125229526257804, "grad_norm": 0.2711309790611267, "learning_rate": 9.610297641507313e-05, "loss": 0.0347, "step": 8510 }, { "epoch": 3.1289019463826664, "grad_norm": 0.3175702691078186, "learning_rate": 9.6090166885708e-05, "loss": 0.0359, "step": 8520 }, { "epoch": 3.1325743665075283, "grad_norm": 0.3962799906730652, "learning_rate": 9.607733719493798e-05, "loss": 0.0325, "step": 8530 }, { "epoch": 3.1362467866323906, "grad_norm": 0.44438639283180237, "learning_rate": 9.606448734837527e-05, "loss": 0.0327, "step": 8540 }, { "epoch": 3.139919206757253, "grad_norm": 0.23799948394298553, "learning_rate": 9.605161735164079e-05, "loss": 0.0364, "step": 8550 }, { "epoch": 3.1435916268821154, "grad_norm": 0.29699450731277466, "learning_rate": 9.60387272103643e-05, "loss": 0.0339, "step": 8560 }, { "epoch": 3.1472640470069777, "grad_norm": 0.35715433955192566, "learning_rate": 9.60258169301844e-05, "loss": 0.0357, "step": 8570 }, { "epoch": 3.15093646713184, "grad_norm": 0.29290297627449036, "learning_rate": 9.601288651674851e-05, "loss": 0.0338, "step": 8580 }, { "epoch": 3.154608887256702, "grad_norm": 0.43383854627609253, "learning_rate": 9.59999359757128e-05, "loss": 0.0331, "step": 8590 }, { "epoch": 3.1582813073815643, "grad_norm": 0.258991003036499, "learning_rate": 9.598696531274227e-05, "loss": 0.0384, "step": 8600 }, { "epoch": 3.1619537275064267, "grad_norm": 0.3273131251335144, "learning_rate": 9.597397453351076e-05, "loss": 0.0329, "step": 8610 }, { "epoch": 3.165626147631289, "grad_norm": 0.4264838695526123, "learning_rate": 9.596096364370087e-05, "loss": 0.0364, "step": 8620 }, { "epoch": 3.1692985677561514, "grad_norm": 0.2913195490837097, "learning_rate": 9.594793264900399e-05, "loss": 0.0343, "step": 8630 }, { "epoch": 3.1729709878810137, "grad_norm": 0.2959482669830322, "learning_rate": 9.593488155512032e-05, "loss": 0.035, "step": 8640 }, { "epoch": 3.176643408005876, "grad_norm": 0.5134719014167786, "learning_rate": 9.592181036775886e-05, "loss": 0.0361, "step": 8650 }, { "epoch": 3.180315828130738, "grad_norm": 0.44891229271888733, "learning_rate": 9.590871909263741e-05, "loss": 0.0367, "step": 8660 }, { "epoch": 3.1839882482556003, "grad_norm": 0.24575890600681305, "learning_rate": 9.589560773548252e-05, "loss": 0.0354, "step": 8670 }, { "epoch": 3.1876606683804627, "grad_norm": 0.276115357875824, "learning_rate": 9.588247630202952e-05, "loss": 0.0372, "step": 8680 }, { "epoch": 3.191333088505325, "grad_norm": 0.19513128697872162, "learning_rate": 9.586932479802258e-05, "loss": 0.036, "step": 8690 }, { "epoch": 3.1950055086301874, "grad_norm": 0.32807445526123047, "learning_rate": 9.585615322921462e-05, "loss": 0.0359, "step": 8700 }, { "epoch": 3.1986779287550497, "grad_norm": 0.32943686842918396, "learning_rate": 9.58429616013673e-05, "loss": 0.0355, "step": 8710 }, { "epoch": 3.2023503488799117, "grad_norm": 0.29824909567832947, "learning_rate": 9.58297499202511e-05, "loss": 0.035, "step": 8720 }, { "epoch": 3.206022769004774, "grad_norm": 0.40529265999794006, "learning_rate": 9.581651819164525e-05, "loss": 0.0355, "step": 8730 }, { "epoch": 3.2096951891296364, "grad_norm": 0.2553085386753082, "learning_rate": 9.580326642133777e-05, "loss": 0.0357, "step": 8740 }, { "epoch": 3.2133676092544987, "grad_norm": 0.25720739364624023, "learning_rate": 9.578999461512544e-05, "loss": 0.0352, "step": 8750 }, { "epoch": 3.217040029379361, "grad_norm": 0.3737129867076874, "learning_rate": 9.577670277881378e-05, "loss": 0.0352, "step": 8760 }, { "epoch": 3.2207124495042234, "grad_norm": 0.21639184653759003, "learning_rate": 9.57633909182171e-05, "loss": 0.0335, "step": 8770 }, { "epoch": 3.2243848696290858, "grad_norm": 0.3890397846698761, "learning_rate": 9.575005903915847e-05, "loss": 0.0361, "step": 8780 }, { "epoch": 3.2280572897539477, "grad_norm": 0.3864012062549591, "learning_rate": 9.573670714746972e-05, "loss": 0.0341, "step": 8790 }, { "epoch": 3.23172970987881, "grad_norm": 0.3489084243774414, "learning_rate": 9.572333524899136e-05, "loss": 0.0344, "step": 8800 }, { "epoch": 3.2354021300036724, "grad_norm": 0.42477545142173767, "learning_rate": 9.570994334957278e-05, "loss": 0.0378, "step": 8810 }, { "epoch": 3.2390745501285347, "grad_norm": 0.4064060151576996, "learning_rate": 9.569653145507201e-05, "loss": 0.0368, "step": 8820 }, { "epoch": 3.242746970253397, "grad_norm": 0.24311788380146027, "learning_rate": 9.568309957135587e-05, "loss": 0.034, "step": 8830 }, { "epoch": 3.2464193903782594, "grad_norm": 0.23521679639816284, "learning_rate": 9.566964770429995e-05, "loss": 0.0323, "step": 8840 }, { "epoch": 3.250091810503122, "grad_norm": 0.34361347556114197, "learning_rate": 9.565617585978853e-05, "loss": 0.0359, "step": 8850 }, { "epoch": 3.2537642306279837, "grad_norm": 0.2619020342826843, "learning_rate": 9.564268404371466e-05, "loss": 0.0361, "step": 8860 }, { "epoch": 3.257436650752846, "grad_norm": 0.35044288635253906, "learning_rate": 9.562917226198013e-05, "loss": 0.0324, "step": 8870 }, { "epoch": 3.2611090708777084, "grad_norm": 0.41983088850975037, "learning_rate": 9.561564052049543e-05, "loss": 0.0359, "step": 8880 }, { "epoch": 3.2647814910025708, "grad_norm": 0.33293649554252625, "learning_rate": 9.560208882517982e-05, "loss": 0.0347, "step": 8890 }, { "epoch": 3.268453911127433, "grad_norm": 0.3261009454727173, "learning_rate": 9.558851718196127e-05, "loss": 0.037, "step": 8900 }, { "epoch": 3.272126331252295, "grad_norm": 0.38644254207611084, "learning_rate": 9.557492559677646e-05, "loss": 0.0325, "step": 8910 }, { "epoch": 3.2757987513771574, "grad_norm": 0.3807685971260071, "learning_rate": 9.556131407557082e-05, "loss": 0.0332, "step": 8920 }, { "epoch": 3.2794711715020197, "grad_norm": 0.3081607222557068, "learning_rate": 9.554768262429853e-05, "loss": 0.0332, "step": 8930 }, { "epoch": 3.283143591626882, "grad_norm": 0.30905380845069885, "learning_rate": 9.553403124892239e-05, "loss": 0.0342, "step": 8940 }, { "epoch": 3.2868160117517444, "grad_norm": 0.30069848895072937, "learning_rate": 9.552035995541402e-05, "loss": 0.0365, "step": 8950 }, { "epoch": 3.290488431876607, "grad_norm": 0.33492860198020935, "learning_rate": 9.550666874975368e-05, "loss": 0.0357, "step": 8960 }, { "epoch": 3.294160852001469, "grad_norm": 0.4662661850452423, "learning_rate": 9.549295763793038e-05, "loss": 0.0329, "step": 8970 }, { "epoch": 3.297833272126331, "grad_norm": 0.5918603539466858, "learning_rate": 9.547922662594183e-05, "loss": 0.0382, "step": 8980 }, { "epoch": 3.3015056922511934, "grad_norm": 0.5100317597389221, "learning_rate": 9.546547571979443e-05, "loss": 0.0367, "step": 8990 }, { "epoch": 3.3051781123760557, "grad_norm": 0.30494627356529236, "learning_rate": 9.545170492550331e-05, "loss": 0.0379, "step": 9000 }, { "epoch": 3.308850532500918, "grad_norm": 0.320484459400177, "learning_rate": 9.543791424909226e-05, "loss": 0.0361, "step": 9010 }, { "epoch": 3.3125229526257804, "grad_norm": 0.3920677900314331, "learning_rate": 9.542410369659382e-05, "loss": 0.0343, "step": 9020 }, { "epoch": 3.316195372750643, "grad_norm": 0.2486322671175003, "learning_rate": 9.541027327404916e-05, "loss": 0.0363, "step": 9030 }, { "epoch": 3.319867792875505, "grad_norm": 0.40499022603034973, "learning_rate": 9.539642298750821e-05, "loss": 0.0341, "step": 9040 }, { "epoch": 3.323540213000367, "grad_norm": 0.2580760419368744, "learning_rate": 9.538255284302954e-05, "loss": 0.0351, "step": 9050 }, { "epoch": 3.3272126331252294, "grad_norm": 0.31785231828689575, "learning_rate": 9.536866284668043e-05, "loss": 0.0346, "step": 9060 }, { "epoch": 3.3308850532500918, "grad_norm": 0.22526153922080994, "learning_rate": 9.53547530045368e-05, "loss": 0.0342, "step": 9070 }, { "epoch": 3.334557473374954, "grad_norm": 0.3060578405857086, "learning_rate": 9.534082332268335e-05, "loss": 0.0338, "step": 9080 }, { "epoch": 3.3382298934998165, "grad_norm": 0.352847158908844, "learning_rate": 9.532687380721334e-05, "loss": 0.0336, "step": 9090 }, { "epoch": 3.341902313624679, "grad_norm": 0.3081820607185364, "learning_rate": 9.53129044642288e-05, "loss": 0.035, "step": 9100 }, { "epoch": 3.345574733749541, "grad_norm": 0.2533964514732361, "learning_rate": 9.529891529984039e-05, "loss": 0.0368, "step": 9110 }, { "epoch": 3.349247153874403, "grad_norm": 0.23000222444534302, "learning_rate": 9.528490632016743e-05, "loss": 0.0339, "step": 9120 }, { "epoch": 3.3529195739992654, "grad_norm": 0.27939897775650024, "learning_rate": 9.52708775313379e-05, "loss": 0.0307, "step": 9130 }, { "epoch": 3.356591994124128, "grad_norm": 0.304876446723938, "learning_rate": 9.525682893948853e-05, "loss": 0.0341, "step": 9140 }, { "epoch": 3.36026441424899, "grad_norm": 0.2773711085319519, "learning_rate": 9.524276055076463e-05, "loss": 0.0351, "step": 9150 }, { "epoch": 3.3639368343738525, "grad_norm": 0.2964112162590027, "learning_rate": 9.522867237132016e-05, "loss": 0.0339, "step": 9160 }, { "epoch": 3.367609254498715, "grad_norm": 0.24285945296287537, "learning_rate": 9.52145644073178e-05, "loss": 0.0333, "step": 9170 }, { "epoch": 3.3712816746235768, "grad_norm": 0.30059248208999634, "learning_rate": 9.520043666492884e-05, "loss": 0.0362, "step": 9180 }, { "epoch": 3.374954094748439, "grad_norm": 0.20395362377166748, "learning_rate": 9.518628915033323e-05, "loss": 0.0345, "step": 9190 }, { "epoch": 3.3786265148733015, "grad_norm": 0.3463868200778961, "learning_rate": 9.517212186971957e-05, "loss": 0.0349, "step": 9200 }, { "epoch": 3.382298934998164, "grad_norm": 0.3333044946193695, "learning_rate": 9.515793482928515e-05, "loss": 0.035, "step": 9210 }, { "epoch": 3.385971355123026, "grad_norm": 0.3012828230857849, "learning_rate": 9.514372803523582e-05, "loss": 0.0366, "step": 9220 }, { "epoch": 3.3896437752478885, "grad_norm": 0.27191999554634094, "learning_rate": 9.512950149378613e-05, "loss": 0.0355, "step": 9230 }, { "epoch": 3.3933161953727504, "grad_norm": 0.2752731144428253, "learning_rate": 9.511525521115925e-05, "loss": 0.0318, "step": 9240 }, { "epoch": 3.3969886154976128, "grad_norm": 0.25991326570510864, "learning_rate": 9.510098919358698e-05, "loss": 0.0353, "step": 9250 }, { "epoch": 3.400661035622475, "grad_norm": 0.24396944046020508, "learning_rate": 9.508670344730979e-05, "loss": 0.0347, "step": 9260 }, { "epoch": 3.4043334557473375, "grad_norm": 0.2863084375858307, "learning_rate": 9.507239797857672e-05, "loss": 0.0357, "step": 9270 }, { "epoch": 3.4080058758722, "grad_norm": 0.3717808723449707, "learning_rate": 9.505807279364548e-05, "loss": 0.0323, "step": 9280 }, { "epoch": 3.411678295997062, "grad_norm": 0.28595080971717834, "learning_rate": 9.504372789878239e-05, "loss": 0.0337, "step": 9290 }, { "epoch": 3.4153507161219245, "grad_norm": 0.28523609042167664, "learning_rate": 9.502936330026239e-05, "loss": 0.0346, "step": 9300 }, { "epoch": 3.4190231362467864, "grad_norm": 0.26236382126808167, "learning_rate": 9.501497900436906e-05, "loss": 0.035, "step": 9310 }, { "epoch": 3.422695556371649, "grad_norm": 0.3088955581188202, "learning_rate": 9.500057501739457e-05, "loss": 0.0338, "step": 9320 }, { "epoch": 3.426367976496511, "grad_norm": 0.3072747588157654, "learning_rate": 9.49861513456397e-05, "loss": 0.0372, "step": 9330 }, { "epoch": 3.4300403966213735, "grad_norm": 0.32916784286499023, "learning_rate": 9.497170799541388e-05, "loss": 0.0334, "step": 9340 }, { "epoch": 3.433712816746236, "grad_norm": 0.2740839123725891, "learning_rate": 9.495724497303508e-05, "loss": 0.0334, "step": 9350 }, { "epoch": 3.437385236871098, "grad_norm": 0.3106094002723694, "learning_rate": 9.494276228482998e-05, "loss": 0.0368, "step": 9360 }, { "epoch": 3.4410576569959606, "grad_norm": 0.277332603931427, "learning_rate": 9.492825993713374e-05, "loss": 0.0352, "step": 9370 }, { "epoch": 3.4447300771208225, "grad_norm": 0.40084198117256165, "learning_rate": 9.491373793629023e-05, "loss": 0.0352, "step": 9380 }, { "epoch": 3.448402497245685, "grad_norm": 0.23994986712932587, "learning_rate": 9.489919628865182e-05, "loss": 0.0362, "step": 9390 }, { "epoch": 3.452074917370547, "grad_norm": 0.4292678236961365, "learning_rate": 9.488463500057955e-05, "loss": 0.0318, "step": 9400 }, { "epoch": 3.4557473374954095, "grad_norm": 0.3179846704006195, "learning_rate": 9.487005407844302e-05, "loss": 0.0325, "step": 9410 }, { "epoch": 3.459419757620272, "grad_norm": 0.3521129786968231, "learning_rate": 9.485545352862039e-05, "loss": 0.0345, "step": 9420 }, { "epoch": 3.4630921777451342, "grad_norm": 0.4470025897026062, "learning_rate": 9.48408333574985e-05, "loss": 0.0339, "step": 9430 }, { "epoch": 3.466764597869996, "grad_norm": 0.37135693430900574, "learning_rate": 9.482619357147264e-05, "loss": 0.0369, "step": 9440 }, { "epoch": 3.4704370179948585, "grad_norm": 0.26431307196617126, "learning_rate": 9.481153417694679e-05, "loss": 0.0341, "step": 9450 }, { "epoch": 3.474109438119721, "grad_norm": 0.34543299674987793, "learning_rate": 9.479685518033347e-05, "loss": 0.0372, "step": 9460 }, { "epoch": 3.477781858244583, "grad_norm": 0.30119797587394714, "learning_rate": 9.478215658805375e-05, "loss": 0.0329, "step": 9470 }, { "epoch": 3.4814542783694455, "grad_norm": 0.22209973633289337, "learning_rate": 9.47674384065373e-05, "loss": 0.0328, "step": 9480 }, { "epoch": 3.485126698494308, "grad_norm": 0.24418999254703522, "learning_rate": 9.475270064222237e-05, "loss": 0.0377, "step": 9490 }, { "epoch": 3.48879911861917, "grad_norm": 0.2994679808616638, "learning_rate": 9.473794330155572e-05, "loss": 0.0358, "step": 9500 }, { "epoch": 3.492471538744032, "grad_norm": 0.3008648157119751, "learning_rate": 9.472316639099276e-05, "loss": 0.0343, "step": 9510 }, { "epoch": 3.4961439588688945, "grad_norm": 0.30228397250175476, "learning_rate": 9.470836991699739e-05, "loss": 0.0359, "step": 9520 }, { "epoch": 3.499816378993757, "grad_norm": 0.3494626581668854, "learning_rate": 9.469355388604208e-05, "loss": 0.0336, "step": 9530 }, { "epoch": 3.503488799118619, "grad_norm": 0.3412083387374878, "learning_rate": 9.467871830460787e-05, "loss": 0.0374, "step": 9540 }, { "epoch": 3.5071612192434816, "grad_norm": 0.27382680773735046, "learning_rate": 9.466386317918436e-05, "loss": 0.0348, "step": 9550 }, { "epoch": 3.510833639368344, "grad_norm": 0.22382964193820953, "learning_rate": 9.464898851626969e-05, "loss": 0.0329, "step": 9560 }, { "epoch": 3.514506059493206, "grad_norm": 0.36148494482040405, "learning_rate": 9.463409432237051e-05, "loss": 0.0335, "step": 9570 }, { "epoch": 3.518178479618068, "grad_norm": 0.32502710819244385, "learning_rate": 9.461918060400209e-05, "loss": 0.0363, "step": 9580 }, { "epoch": 3.5218508997429305, "grad_norm": 0.30727410316467285, "learning_rate": 9.460424736768816e-05, "loss": 0.0344, "step": 9590 }, { "epoch": 3.525523319867793, "grad_norm": 0.2876565456390381, "learning_rate": 9.458929461996105e-05, "loss": 0.0323, "step": 9600 }, { "epoch": 3.5291957399926552, "grad_norm": 0.25542327761650085, "learning_rate": 9.457432236736158e-05, "loss": 0.0358, "step": 9610 }, { "epoch": 3.5328681601175176, "grad_norm": 0.301963746547699, "learning_rate": 9.455933061643916e-05, "loss": 0.033, "step": 9620 }, { "epoch": 3.53654058024238, "grad_norm": 0.299420028924942, "learning_rate": 9.454431937375164e-05, "loss": 0.0355, "step": 9630 }, { "epoch": 3.540213000367242, "grad_norm": 0.2275063693523407, "learning_rate": 9.452928864586547e-05, "loss": 0.0327, "step": 9640 }, { "epoch": 3.543885420492104, "grad_norm": 0.26451683044433594, "learning_rate": 9.451423843935563e-05, "loss": 0.0347, "step": 9650 }, { "epoch": 3.5475578406169666, "grad_norm": 0.27868202328681946, "learning_rate": 9.449916876080553e-05, "loss": 0.0347, "step": 9660 }, { "epoch": 3.551230260741829, "grad_norm": 0.2305579036474228, "learning_rate": 9.44840796168072e-05, "loss": 0.0343, "step": 9670 }, { "epoch": 3.5549026808666913, "grad_norm": 0.3831007778644562, "learning_rate": 9.446897101396115e-05, "loss": 0.0342, "step": 9680 }, { "epoch": 3.558575100991553, "grad_norm": 0.2755167782306671, "learning_rate": 9.445384295887638e-05, "loss": 0.0315, "step": 9690 }, { "epoch": 3.562247521116416, "grad_norm": 0.253224641084671, "learning_rate": 9.443869545817043e-05, "loss": 0.0333, "step": 9700 }, { "epoch": 3.565919941241278, "grad_norm": 0.29231148958206177, "learning_rate": 9.442352851846929e-05, "loss": 0.034, "step": 9710 }, { "epoch": 3.56959236136614, "grad_norm": 0.2720223665237427, "learning_rate": 9.440834214640755e-05, "loss": 0.0344, "step": 9720 }, { "epoch": 3.5732647814910026, "grad_norm": 0.3216252624988556, "learning_rate": 9.439313634862823e-05, "loss": 0.0342, "step": 9730 }, { "epoch": 3.576937201615865, "grad_norm": 0.29116368293762207, "learning_rate": 9.437791113178282e-05, "loss": 0.0316, "step": 9740 }, { "epoch": 3.5806096217407273, "grad_norm": 0.3095034956932068, "learning_rate": 9.43626665025314e-05, "loss": 0.0344, "step": 9750 }, { "epoch": 3.584282041865589, "grad_norm": 0.31522494554519653, "learning_rate": 9.434740246754248e-05, "loss": 0.0331, "step": 9760 }, { "epoch": 3.587954461990452, "grad_norm": 0.3946789503097534, "learning_rate": 9.433211903349304e-05, "loss": 0.0328, "step": 9770 }, { "epoch": 3.591626882115314, "grad_norm": 0.24425417184829712, "learning_rate": 9.431681620706858e-05, "loss": 0.0326, "step": 9780 }, { "epoch": 3.5952993022401762, "grad_norm": 0.31765979528427124, "learning_rate": 9.43014939949631e-05, "loss": 0.0346, "step": 9790 }, { "epoch": 3.5989717223650386, "grad_norm": 0.3255269229412079, "learning_rate": 9.428615240387904e-05, "loss": 0.0322, "step": 9800 }, { "epoch": 3.602644142489901, "grad_norm": 0.36538735032081604, "learning_rate": 9.427079144052732e-05, "loss": 0.0339, "step": 9810 }, { "epoch": 3.6063165626147633, "grad_norm": 0.3626943826675415, "learning_rate": 9.425541111162739e-05, "loss": 0.0328, "step": 9820 }, { "epoch": 3.609988982739625, "grad_norm": 0.22024662792682648, "learning_rate": 9.424001142390709e-05, "loss": 0.035, "step": 9830 }, { "epoch": 3.6136614028644876, "grad_norm": 0.2863999009132385, "learning_rate": 9.422459238410277e-05, "loss": 0.0325, "step": 9840 }, { "epoch": 3.61733382298935, "grad_norm": 0.4217221140861511, "learning_rate": 9.420915399895926e-05, "loss": 0.0355, "step": 9850 }, { "epoch": 3.6210062431142123, "grad_norm": 0.2839545011520386, "learning_rate": 9.419369627522981e-05, "loss": 0.0336, "step": 9860 }, { "epoch": 3.6246786632390746, "grad_norm": 0.28867989778518677, "learning_rate": 9.417821921967618e-05, "loss": 0.0318, "step": 9870 }, { "epoch": 3.628351083363937, "grad_norm": 0.22029875218868256, "learning_rate": 9.416272283906855e-05, "loss": 0.034, "step": 9880 }, { "epoch": 3.6320235034887993, "grad_norm": 0.3098442852497101, "learning_rate": 9.414720714018554e-05, "loss": 0.0342, "step": 9890 }, { "epoch": 3.6356959236136612, "grad_norm": 0.4068317413330078, "learning_rate": 9.413167212981427e-05, "loss": 0.0339, "step": 9900 }, { "epoch": 3.6393683437385236, "grad_norm": 0.33098751306533813, "learning_rate": 9.41161178147503e-05, "loss": 0.0363, "step": 9910 }, { "epoch": 3.643040763863386, "grad_norm": 0.33062615990638733, "learning_rate": 9.410054420179755e-05, "loss": 0.037, "step": 9920 }, { "epoch": 3.6467131839882483, "grad_norm": 0.38571590185165405, "learning_rate": 9.408495129776851e-05, "loss": 0.0321, "step": 9930 }, { "epoch": 3.6503856041131106, "grad_norm": 0.32717379927635193, "learning_rate": 9.406933910948403e-05, "loss": 0.0384, "step": 9940 }, { "epoch": 3.654058024237973, "grad_norm": 0.2586788237094879, "learning_rate": 9.40537076437734e-05, "loss": 0.035, "step": 9950 }, { "epoch": 3.6577304443628353, "grad_norm": 0.3420557379722595, "learning_rate": 9.403805690747436e-05, "loss": 0.0345, "step": 9960 }, { "epoch": 3.6614028644876973, "grad_norm": 0.2628384232521057, "learning_rate": 9.402238690743308e-05, "loss": 0.0326, "step": 9970 }, { "epoch": 3.6650752846125596, "grad_norm": 0.43909358978271484, "learning_rate": 9.400669765050413e-05, "loss": 0.0316, "step": 9980 }, { "epoch": 3.668747704737422, "grad_norm": 0.3353974223136902, "learning_rate": 9.399098914355055e-05, "loss": 0.0344, "step": 9990 }, { "epoch": 3.6724201248622843, "grad_norm": 0.3450661599636078, "learning_rate": 9.397526139344378e-05, "loss": 0.0347, "step": 10000 }, { "epoch": 3.6760925449871467, "grad_norm": 0.4128000736236572, "learning_rate": 9.395951440706362e-05, "loss": 0.0343, "step": 10010 }, { "epoch": 3.6797649651120086, "grad_norm": 0.31701838970184326, "learning_rate": 9.394374819129839e-05, "loss": 0.0323, "step": 10020 }, { "epoch": 3.6834373852368714, "grad_norm": 0.3130081593990326, "learning_rate": 9.392796275304474e-05, "loss": 0.0337, "step": 10030 }, { "epoch": 3.6871098053617333, "grad_norm": 0.3103707432746887, "learning_rate": 9.391215809920778e-05, "loss": 0.035, "step": 10040 }, { "epoch": 3.6907822254865956, "grad_norm": 0.18323861062526703, "learning_rate": 9.389633423670099e-05, "loss": 0.0319, "step": 10050 }, { "epoch": 3.694454645611458, "grad_norm": 0.27198588848114014, "learning_rate": 9.388049117244626e-05, "loss": 0.0338, "step": 10060 }, { "epoch": 3.6981270657363203, "grad_norm": 0.2456187754869461, "learning_rate": 9.386462891337389e-05, "loss": 0.0329, "step": 10070 }, { "epoch": 3.7017994858611827, "grad_norm": 0.36419156193733215, "learning_rate": 9.384874746642257e-05, "loss": 0.0333, "step": 10080 }, { "epoch": 3.7054719059860446, "grad_norm": 0.22404156625270844, "learning_rate": 9.383284683853937e-05, "loss": 0.0303, "step": 10090 }, { "epoch": 3.709144326110907, "grad_norm": 0.34323275089263916, "learning_rate": 9.381692703667981e-05, "loss": 0.0323, "step": 10100 }, { "epoch": 3.7128167462357693, "grad_norm": 0.43188896775245667, "learning_rate": 9.380098806780771e-05, "loss": 0.0343, "step": 10110 }, { "epoch": 3.7164891663606316, "grad_norm": 0.2691934108734131, "learning_rate": 9.378502993889533e-05, "loss": 0.0333, "step": 10120 }, { "epoch": 3.720161586485494, "grad_norm": 0.30337610840797424, "learning_rate": 9.376905265692329e-05, "loss": 0.031, "step": 10130 }, { "epoch": 3.7238340066103564, "grad_norm": 0.26906317472457886, "learning_rate": 9.37530562288806e-05, "loss": 0.0319, "step": 10140 }, { "epoch": 3.7275064267352187, "grad_norm": 0.2767673432826996, "learning_rate": 9.373704066176465e-05, "loss": 0.0334, "step": 10150 }, { "epoch": 3.7311788468600806, "grad_norm": 0.3395156264305115, "learning_rate": 9.372100596258118e-05, "loss": 0.0322, "step": 10160 }, { "epoch": 3.734851266984943, "grad_norm": 0.34775060415267944, "learning_rate": 9.370495213834433e-05, "loss": 0.0334, "step": 10170 }, { "epoch": 3.7385236871098053, "grad_norm": 0.31151798367500305, "learning_rate": 9.368887919607657e-05, "loss": 0.0331, "step": 10180 }, { "epoch": 3.7421961072346677, "grad_norm": 0.33851516246795654, "learning_rate": 9.367278714280876e-05, "loss": 0.0338, "step": 10190 }, { "epoch": 3.74586852735953, "grad_norm": 0.25468820333480835, "learning_rate": 9.36566759855801e-05, "loss": 0.0324, "step": 10200 }, { "epoch": 3.7495409474843924, "grad_norm": 0.4918968975543976, "learning_rate": 9.36405457314382e-05, "loss": 0.0359, "step": 10210 }, { "epoch": 3.7532133676092547, "grad_norm": 0.2589881122112274, "learning_rate": 9.36243963874389e-05, "loss": 0.0334, "step": 10220 }, { "epoch": 3.7568857877341166, "grad_norm": 0.29881298542022705, "learning_rate": 9.360822796064655e-05, "loss": 0.0335, "step": 10230 }, { "epoch": 3.760558207858979, "grad_norm": 0.487813800573349, "learning_rate": 9.359204045813372e-05, "loss": 0.0336, "step": 10240 }, { "epoch": 3.7642306279838413, "grad_norm": 0.28629425168037415, "learning_rate": 9.357583388698141e-05, "loss": 0.0326, "step": 10250 }, { "epoch": 3.7679030481087037, "grad_norm": 0.2826322913169861, "learning_rate": 9.35596082542789e-05, "loss": 0.0331, "step": 10260 }, { "epoch": 3.771575468233566, "grad_norm": 0.23700781166553497, "learning_rate": 9.354336356712383e-05, "loss": 0.0319, "step": 10270 }, { "epoch": 3.775247888358428, "grad_norm": 0.41486677527427673, "learning_rate": 9.35270998326222e-05, "loss": 0.0376, "step": 10280 }, { "epoch": 3.7789203084832907, "grad_norm": 0.3048959970474243, "learning_rate": 9.351081705788831e-05, "loss": 0.0341, "step": 10290 }, { "epoch": 3.7825927286081527, "grad_norm": 0.4117546081542969, "learning_rate": 9.349451525004477e-05, "loss": 0.0325, "step": 10300 }, { "epoch": 3.786265148733015, "grad_norm": 0.2823678255081177, "learning_rate": 9.347819441622261e-05, "loss": 0.0316, "step": 10310 }, { "epoch": 3.7899375688578774, "grad_norm": 0.2955907881259918, "learning_rate": 9.346185456356105e-05, "loss": 0.0345, "step": 10320 }, { "epoch": 3.7936099889827397, "grad_norm": 0.27884286642074585, "learning_rate": 9.344549569920774e-05, "loss": 0.0302, "step": 10330 }, { "epoch": 3.797282409107602, "grad_norm": 0.2839728593826294, "learning_rate": 9.342911783031858e-05, "loss": 0.0313, "step": 10340 }, { "epoch": 3.800954829232464, "grad_norm": 0.28371694684028625, "learning_rate": 9.341272096405782e-05, "loss": 0.031, "step": 10350 }, { "epoch": 3.8046272493573263, "grad_norm": 0.2409438043832779, "learning_rate": 9.3396305107598e-05, "loss": 0.0288, "step": 10360 }, { "epoch": 3.8082996694821887, "grad_norm": 0.31802961230278015, "learning_rate": 9.337987026811998e-05, "loss": 0.0335, "step": 10370 }, { "epoch": 3.811972089607051, "grad_norm": 0.40102407336235046, "learning_rate": 9.33634164528129e-05, "loss": 0.0314, "step": 10380 }, { "epoch": 3.8156445097319134, "grad_norm": 0.4040955901145935, "learning_rate": 9.334694366887424e-05, "loss": 0.0304, "step": 10390 }, { "epoch": 3.8193169298567757, "grad_norm": 0.40451350808143616, "learning_rate": 9.333045192350973e-05, "loss": 0.034, "step": 10400 }, { "epoch": 3.822989349981638, "grad_norm": 0.34920138120651245, "learning_rate": 9.331394122393345e-05, "loss": 0.0329, "step": 10410 }, { "epoch": 3.8266617701065, "grad_norm": 0.2457280308008194, "learning_rate": 9.329741157736771e-05, "loss": 0.0336, "step": 10420 }, { "epoch": 3.8303341902313623, "grad_norm": 0.22831042110919952, "learning_rate": 9.328086299104317e-05, "loss": 0.0331, "step": 10430 }, { "epoch": 3.8340066103562247, "grad_norm": 0.2484305053949356, "learning_rate": 9.326429547219872e-05, "loss": 0.0328, "step": 10440 }, { "epoch": 3.837679030481087, "grad_norm": 0.31765320897102356, "learning_rate": 9.324770902808155e-05, "loss": 0.0318, "step": 10450 }, { "epoch": 3.8413514506059494, "grad_norm": 0.28799790143966675, "learning_rate": 9.323110366594717e-05, "loss": 0.0305, "step": 10460 }, { "epoch": 3.8450238707308118, "grad_norm": 0.2695273458957672, "learning_rate": 9.32144793930593e-05, "loss": 0.0332, "step": 10470 }, { "epoch": 3.848696290855674, "grad_norm": 0.22238586843013763, "learning_rate": 9.319783621668996e-05, "loss": 0.03, "step": 10480 }, { "epoch": 3.852368710980536, "grad_norm": 0.1946725845336914, "learning_rate": 9.318117414411947e-05, "loss": 0.0319, "step": 10490 }, { "epoch": 3.8560411311053984, "grad_norm": 0.1860978901386261, "learning_rate": 9.316449318263635e-05, "loss": 0.0311, "step": 10500 }, { "epoch": 3.8597135512302607, "grad_norm": 0.2564229965209961, "learning_rate": 9.314779333953744e-05, "loss": 0.0361, "step": 10510 }, { "epoch": 3.863385971355123, "grad_norm": 0.3502614498138428, "learning_rate": 9.313107462212781e-05, "loss": 0.0339, "step": 10520 }, { "epoch": 3.8670583914799854, "grad_norm": 0.34937888383865356, "learning_rate": 9.311433703772082e-05, "loss": 0.032, "step": 10530 }, { "epoch": 3.8707308116048473, "grad_norm": 0.1965707540512085, "learning_rate": 9.3097580593638e-05, "loss": 0.0313, "step": 10540 }, { "epoch": 3.87440323172971, "grad_norm": 0.3482417166233063, "learning_rate": 9.308080529720926e-05, "loss": 0.0307, "step": 10550 }, { "epoch": 3.878075651854572, "grad_norm": 0.27428075671195984, "learning_rate": 9.306401115577264e-05, "loss": 0.0311, "step": 10560 }, { "epoch": 3.8817480719794344, "grad_norm": 0.35161811113357544, "learning_rate": 9.304719817667447e-05, "loss": 0.0306, "step": 10570 }, { "epoch": 3.8854204921042967, "grad_norm": 0.26628801226615906, "learning_rate": 9.303036636726934e-05, "loss": 0.0291, "step": 10580 }, { "epoch": 3.889092912229159, "grad_norm": 0.5572242736816406, "learning_rate": 9.301351573492003e-05, "loss": 0.0309, "step": 10590 }, { "epoch": 3.8927653323540214, "grad_norm": 0.24504514038562775, "learning_rate": 9.299664628699758e-05, "loss": 0.0297, "step": 10600 }, { "epoch": 3.8964377524788834, "grad_norm": 0.2871670722961426, "learning_rate": 9.297975803088129e-05, "loss": 0.0307, "step": 10610 }, { "epoch": 3.9001101726037457, "grad_norm": 0.3640875220298767, "learning_rate": 9.296285097395864e-05, "loss": 0.0298, "step": 10620 }, { "epoch": 3.903782592728608, "grad_norm": 0.23202000558376312, "learning_rate": 9.294592512362533e-05, "loss": 0.0333, "step": 10630 }, { "epoch": 3.9074550128534704, "grad_norm": 0.35989153385162354, "learning_rate": 9.292898048728533e-05, "loss": 0.031, "step": 10640 }, { "epoch": 3.9111274329783328, "grad_norm": 0.4782460927963257, "learning_rate": 9.29120170723508e-05, "loss": 0.0328, "step": 10650 }, { "epoch": 3.914799853103195, "grad_norm": 0.21137700974941254, "learning_rate": 9.28950348862421e-05, "loss": 0.0313, "step": 10660 }, { "epoch": 3.9184722732280575, "grad_norm": 0.38560354709625244, "learning_rate": 9.287803393638781e-05, "loss": 0.0309, "step": 10670 }, { "epoch": 3.9221446933529194, "grad_norm": 0.29609158635139465, "learning_rate": 9.286101423022474e-05, "loss": 0.0344, "step": 10680 }, { "epoch": 3.9258171134777817, "grad_norm": 0.2964165210723877, "learning_rate": 9.284397577519788e-05, "loss": 0.0319, "step": 10690 }, { "epoch": 3.929489533602644, "grad_norm": 0.2241217941045761, "learning_rate": 9.282691857876043e-05, "loss": 0.0308, "step": 10700 }, { "epoch": 3.9331619537275064, "grad_norm": 0.38399192690849304, "learning_rate": 9.280984264837377e-05, "loss": 0.0347, "step": 10710 }, { "epoch": 3.936834373852369, "grad_norm": 0.24236664175987244, "learning_rate": 9.279274799150752e-05, "loss": 0.0289, "step": 10720 }, { "epoch": 3.940506793977231, "grad_norm": 0.23392292857170105, "learning_rate": 9.277563461563945e-05, "loss": 0.0292, "step": 10730 }, { "epoch": 3.9441792141020935, "grad_norm": 0.30663758516311646, "learning_rate": 9.275850252825555e-05, "loss": 0.0326, "step": 10740 }, { "epoch": 3.9478516342269554, "grad_norm": 0.2960768938064575, "learning_rate": 9.274135173684994e-05, "loss": 0.0318, "step": 10750 }, { "epoch": 3.9515240543518177, "grad_norm": 0.28218236565589905, "learning_rate": 9.272418224892498e-05, "loss": 0.0318, "step": 10760 }, { "epoch": 3.95519647447668, "grad_norm": 0.22266888618469238, "learning_rate": 9.27069940719912e-05, "loss": 0.0318, "step": 10770 }, { "epoch": 3.9588688946015425, "grad_norm": 0.2852974534034729, "learning_rate": 9.268978721356727e-05, "loss": 0.0311, "step": 10780 }, { "epoch": 3.962541314726405, "grad_norm": 0.4975876808166504, "learning_rate": 9.267256168118008e-05, "loss": 0.034, "step": 10790 }, { "epoch": 3.9662137348512667, "grad_norm": 0.21327440440654755, "learning_rate": 9.265531748236463e-05, "loss": 0.0321, "step": 10800 }, { "epoch": 3.9698861549761295, "grad_norm": 0.42730358242988586, "learning_rate": 9.263805462466416e-05, "loss": 0.0324, "step": 10810 }, { "epoch": 3.9735585751009914, "grad_norm": 0.32748615741729736, "learning_rate": 9.262077311562998e-05, "loss": 0.0338, "step": 10820 }, { "epoch": 3.9772309952258538, "grad_norm": 0.334220826625824, "learning_rate": 9.260347296282165e-05, "loss": 0.0323, "step": 10830 }, { "epoch": 3.980903415350716, "grad_norm": 0.38051626086235046, "learning_rate": 9.258615417380683e-05, "loss": 0.0309, "step": 10840 }, { "epoch": 3.9845758354755785, "grad_norm": 0.3165581226348877, "learning_rate": 9.256881675616133e-05, "loss": 0.032, "step": 10850 }, { "epoch": 3.988248255600441, "grad_norm": 0.2257159948348999, "learning_rate": 9.255146071746917e-05, "loss": 0.0304, "step": 10860 }, { "epoch": 3.9919206757253027, "grad_norm": 0.39684996008872986, "learning_rate": 9.253408606532241e-05, "loss": 0.0329, "step": 10870 }, { "epoch": 3.9955930958501655, "grad_norm": 0.37098637223243713, "learning_rate": 9.251669280732137e-05, "loss": 0.0318, "step": 10880 }, { "epoch": 3.9992655159750274, "grad_norm": 0.25810930132865906, "learning_rate": 9.249928095107441e-05, "loss": 0.0322, "step": 10890 }, { "epoch": 4.00293793609989, "grad_norm": 0.40813973546028137, "learning_rate": 9.248185050419811e-05, "loss": 0.0308, "step": 10900 }, { "epoch": 4.006610356224752, "grad_norm": 0.3213004171848297, "learning_rate": 9.24644014743171e-05, "loss": 0.0322, "step": 10910 }, { "epoch": 4.010282776349614, "grad_norm": 0.29025721549987793, "learning_rate": 9.24469338690642e-05, "loss": 0.029, "step": 10920 }, { "epoch": 4.013955196474477, "grad_norm": 0.3865046203136444, "learning_rate": 9.242944769608033e-05, "loss": 0.0332, "step": 10930 }, { "epoch": 4.017627616599339, "grad_norm": 0.31081679463386536, "learning_rate": 9.241194296301454e-05, "loss": 0.0316, "step": 10940 }, { "epoch": 4.0213000367242016, "grad_norm": 0.25276973843574524, "learning_rate": 9.239441967752397e-05, "loss": 0.0334, "step": 10950 }, { "epoch": 4.0249724568490635, "grad_norm": 0.31508252024650574, "learning_rate": 9.237687784727393e-05, "loss": 0.0315, "step": 10960 }, { "epoch": 4.028644876973926, "grad_norm": 0.4580555558204651, "learning_rate": 9.235931747993781e-05, "loss": 0.0312, "step": 10970 }, { "epoch": 4.032317297098788, "grad_norm": 0.3188674747943878, "learning_rate": 9.234173858319707e-05, "loss": 0.0312, "step": 10980 }, { "epoch": 4.03598971722365, "grad_norm": 0.47070664167404175, "learning_rate": 9.23241411647414e-05, "loss": 0.0334, "step": 10990 }, { "epoch": 4.039662137348513, "grad_norm": 0.23283740878105164, "learning_rate": 9.230652523226841e-05, "loss": 0.0345, "step": 11000 }, { "epoch": 4.043334557473375, "grad_norm": 0.2963833510875702, "learning_rate": 9.2288890793484e-05, "loss": 0.0316, "step": 11010 }, { "epoch": 4.047006977598238, "grad_norm": 0.330451101064682, "learning_rate": 9.227123785610199e-05, "loss": 0.0296, "step": 11020 }, { "epoch": 4.0506793977230995, "grad_norm": 0.23626618087291718, "learning_rate": 9.225356642784443e-05, "loss": 0.033, "step": 11030 }, { "epoch": 4.054351817847961, "grad_norm": 0.2573946714401245, "learning_rate": 9.223587651644139e-05, "loss": 0.0308, "step": 11040 }, { "epoch": 4.058024237972824, "grad_norm": 0.200330451130867, "learning_rate": 9.221816812963104e-05, "loss": 0.0312, "step": 11050 }, { "epoch": 4.061696658097686, "grad_norm": 0.2327907830476761, "learning_rate": 9.22004412751596e-05, "loss": 0.0292, "step": 11060 }, { "epoch": 4.065369078222549, "grad_norm": 0.30124786496162415, "learning_rate": 9.218269596078146e-05, "loss": 0.0325, "step": 11070 }, { "epoch": 4.069041498347411, "grad_norm": 0.2670992910861969, "learning_rate": 9.216493219425895e-05, "loss": 0.033, "step": 11080 }, { "epoch": 4.072713918472274, "grad_norm": 0.20509515702724457, "learning_rate": 9.21471499833626e-05, "loss": 0.0308, "step": 11090 }, { "epoch": 4.0763863385971355, "grad_norm": 0.3577521741390228, "learning_rate": 9.212934933587094e-05, "loss": 0.0327, "step": 11100 }, { "epoch": 4.080058758721997, "grad_norm": 0.39029011130332947, "learning_rate": 9.211153025957056e-05, "loss": 0.0347, "step": 11110 }, { "epoch": 4.08373117884686, "grad_norm": 0.3291454017162323, "learning_rate": 9.209369276225614e-05, "loss": 0.0322, "step": 11120 }, { "epoch": 4.087403598971722, "grad_norm": 0.25012457370758057, "learning_rate": 9.207583685173042e-05, "loss": 0.0304, "step": 11130 }, { "epoch": 4.091076019096585, "grad_norm": 0.3227810263633728, "learning_rate": 9.205796253580417e-05, "loss": 0.0347, "step": 11140 }, { "epoch": 4.094748439221447, "grad_norm": 0.2839024066925049, "learning_rate": 9.204006982229621e-05, "loss": 0.0355, "step": 11150 }, { "epoch": 4.09842085934631, "grad_norm": 0.30125293135643005, "learning_rate": 9.202215871903346e-05, "loss": 0.0347, "step": 11160 }, { "epoch": 4.1020932794711715, "grad_norm": 0.42964670062065125, "learning_rate": 9.20042292338508e-05, "loss": 0.0321, "step": 11170 }, { "epoch": 4.105765699596033, "grad_norm": 0.2958247661590576, "learning_rate": 9.198628137459123e-05, "loss": 0.0317, "step": 11180 }, { "epoch": 4.109438119720896, "grad_norm": 0.36162176728248596, "learning_rate": 9.196831514910572e-05, "loss": 0.0351, "step": 11190 }, { "epoch": 4.113110539845758, "grad_norm": 0.30073267221450806, "learning_rate": 9.195033056525332e-05, "loss": 0.0331, "step": 11200 }, { "epoch": 4.116782959970621, "grad_norm": 0.2352515161037445, "learning_rate": 9.193232763090114e-05, "loss": 0.0371, "step": 11210 }, { "epoch": 4.120455380095483, "grad_norm": 0.3546741306781769, "learning_rate": 9.191430635392422e-05, "loss": 0.0315, "step": 11220 }, { "epoch": 4.124127800220346, "grad_norm": 0.21312949061393738, "learning_rate": 9.18962667422057e-05, "loss": 0.0349, "step": 11230 }, { "epoch": 4.1278002203452075, "grad_norm": 0.42275333404541016, "learning_rate": 9.187820880363671e-05, "loss": 0.0309, "step": 11240 }, { "epoch": 4.1314726404700695, "grad_norm": 0.3079930245876312, "learning_rate": 9.186013254611643e-05, "loss": 0.0349, "step": 11250 }, { "epoch": 4.135145060594932, "grad_norm": 0.2920031249523163, "learning_rate": 9.1842037977552e-05, "loss": 0.0308, "step": 11260 }, { "epoch": 4.138817480719794, "grad_norm": 0.2583586573600769, "learning_rate": 9.182392510585862e-05, "loss": 0.0303, "step": 11270 }, { "epoch": 4.142489900844657, "grad_norm": 0.2799769639968872, "learning_rate": 9.180579393895946e-05, "loss": 0.0326, "step": 11280 }, { "epoch": 4.146162320969519, "grad_norm": 0.26204654574394226, "learning_rate": 9.178764448478572e-05, "loss": 0.0299, "step": 11290 }, { "epoch": 4.149834741094381, "grad_norm": 0.23167608678340912, "learning_rate": 9.176947675127658e-05, "loss": 0.0325, "step": 11300 }, { "epoch": 4.153507161219244, "grad_norm": 0.3393232822418213, "learning_rate": 9.175129074637924e-05, "loss": 0.0296, "step": 11310 }, { "epoch": 4.1571795813441055, "grad_norm": 0.2415020912885666, "learning_rate": 9.173308647804884e-05, "loss": 0.0304, "step": 11320 }, { "epoch": 4.160852001468968, "grad_norm": 0.17748233675956726, "learning_rate": 9.171486395424859e-05, "loss": 0.0306, "step": 11330 }, { "epoch": 4.16452442159383, "grad_norm": 0.22911405563354492, "learning_rate": 9.16966231829496e-05, "loss": 0.031, "step": 11340 }, { "epoch": 4.168196841718693, "grad_norm": 0.2508696913719177, "learning_rate": 9.167836417213105e-05, "loss": 0.0292, "step": 11350 }, { "epoch": 4.171869261843555, "grad_norm": 0.3049671947956085, "learning_rate": 9.166008692978001e-05, "loss": 0.03, "step": 11360 }, { "epoch": 4.175541681968417, "grad_norm": 0.39494433999061584, "learning_rate": 9.164179146389158e-05, "loss": 0.0337, "step": 11370 }, { "epoch": 4.17921410209328, "grad_norm": 0.23440274596214294, "learning_rate": 9.162347778246882e-05, "loss": 0.0302, "step": 11380 }, { "epoch": 4.1828865222181415, "grad_norm": 0.28843268752098083, "learning_rate": 9.160514589352276e-05, "loss": 0.0314, "step": 11390 }, { "epoch": 4.186558942343004, "grad_norm": 0.3144293427467346, "learning_rate": 9.15867958050724e-05, "loss": 0.0274, "step": 11400 }, { "epoch": 4.190231362467866, "grad_norm": 0.25710299611091614, "learning_rate": 9.156842752514466e-05, "loss": 0.0297, "step": 11410 }, { "epoch": 4.193903782592729, "grad_norm": 0.2964564263820648, "learning_rate": 9.155004106177447e-05, "loss": 0.0319, "step": 11420 }, { "epoch": 4.197576202717591, "grad_norm": 0.17780107259750366, "learning_rate": 9.153163642300471e-05, "loss": 0.0305, "step": 11430 }, { "epoch": 4.201248622842453, "grad_norm": 0.24660834670066833, "learning_rate": 9.151321361688616e-05, "loss": 0.0304, "step": 11440 }, { "epoch": 4.204921042967316, "grad_norm": 0.27001047134399414, "learning_rate": 9.149477265147762e-05, "loss": 0.0295, "step": 11450 }, { "epoch": 4.2085934630921775, "grad_norm": 0.35138311982154846, "learning_rate": 9.147631353484574e-05, "loss": 0.0354, "step": 11460 }, { "epoch": 4.21226588321704, "grad_norm": 0.375591516494751, "learning_rate": 9.145783627506522e-05, "loss": 0.0309, "step": 11470 }, { "epoch": 4.215938303341902, "grad_norm": 0.22167237102985382, "learning_rate": 9.143934088021861e-05, "loss": 0.0303, "step": 11480 }, { "epoch": 4.219610723466765, "grad_norm": 0.20374342799186707, "learning_rate": 9.142082735839645e-05, "loss": 0.0304, "step": 11490 }, { "epoch": 4.223283143591627, "grad_norm": 0.24607843160629272, "learning_rate": 9.140229571769715e-05, "loss": 0.0315, "step": 11500 }, { "epoch": 4.226955563716489, "grad_norm": 0.3041512072086334, "learning_rate": 9.138374596622709e-05, "loss": 0.0294, "step": 11510 }, { "epoch": 4.230627983841352, "grad_norm": 0.2847870886325836, "learning_rate": 9.136517811210059e-05, "loss": 0.0305, "step": 11520 }, { "epoch": 4.2343004039662135, "grad_norm": 0.3542657196521759, "learning_rate": 9.134659216343984e-05, "loss": 0.0318, "step": 11530 }, { "epoch": 4.237972824091076, "grad_norm": 0.3207806944847107, "learning_rate": 9.132798812837494e-05, "loss": 0.0292, "step": 11540 }, { "epoch": 4.241645244215938, "grad_norm": 0.26237455010414124, "learning_rate": 9.130936601504396e-05, "loss": 0.0314, "step": 11550 }, { "epoch": 4.245317664340801, "grad_norm": 0.4435146152973175, "learning_rate": 9.129072583159284e-05, "loss": 0.03, "step": 11560 }, { "epoch": 4.248990084465663, "grad_norm": 0.2189512997865677, "learning_rate": 9.127206758617542e-05, "loss": 0.0298, "step": 11570 }, { "epoch": 4.252662504590525, "grad_norm": 0.33551299571990967, "learning_rate": 9.125339128695346e-05, "loss": 0.0296, "step": 11580 }, { "epoch": 4.256334924715388, "grad_norm": 0.3184555768966675, "learning_rate": 9.123469694209659e-05, "loss": 0.0317, "step": 11590 }, { "epoch": 4.26000734484025, "grad_norm": 0.3075922727584839, "learning_rate": 9.121598455978239e-05, "loss": 0.0291, "step": 11600 }, { "epoch": 4.263679764965112, "grad_norm": 0.34011855721473694, "learning_rate": 9.119725414819624e-05, "loss": 0.0299, "step": 11610 }, { "epoch": 4.267352185089974, "grad_norm": 0.2225658893585205, "learning_rate": 9.117850571553149e-05, "loss": 0.0313, "step": 11620 }, { "epoch": 4.271024605214836, "grad_norm": 0.4850257337093353, "learning_rate": 9.115973926998935e-05, "loss": 0.0283, "step": 11630 }, { "epoch": 4.274697025339699, "grad_norm": 0.24065685272216797, "learning_rate": 9.114095481977888e-05, "loss": 0.0327, "step": 11640 }, { "epoch": 4.278369445464561, "grad_norm": 0.3132742941379547, "learning_rate": 9.112215237311703e-05, "loss": 0.0288, "step": 11650 }, { "epoch": 4.282041865589424, "grad_norm": 0.31925293803215027, "learning_rate": 9.110333193822867e-05, "loss": 0.0305, "step": 11660 }, { "epoch": 4.285714285714286, "grad_norm": 0.2746545970439911, "learning_rate": 9.108449352334645e-05, "loss": 0.0332, "step": 11670 }, { "epoch": 4.289386705839148, "grad_norm": 0.2361440509557724, "learning_rate": 9.106563713671094e-05, "loss": 0.0291, "step": 11680 }, { "epoch": 4.29305912596401, "grad_norm": 0.3369962275028229, "learning_rate": 9.104676278657061e-05, "loss": 0.0317, "step": 11690 }, { "epoch": 4.296731546088872, "grad_norm": 0.2537284791469574, "learning_rate": 9.102787048118169e-05, "loss": 0.0313, "step": 11700 }, { "epoch": 4.300403966213735, "grad_norm": 0.29814767837524414, "learning_rate": 9.100896022880834e-05, "loss": 0.0342, "step": 11710 }, { "epoch": 4.304076386338597, "grad_norm": 0.25758126378059387, "learning_rate": 9.099003203772254e-05, "loss": 0.0291, "step": 11720 }, { "epoch": 4.30774880646346, "grad_norm": 0.315684050321579, "learning_rate": 9.097108591620413e-05, "loss": 0.0309, "step": 11730 }, { "epoch": 4.311421226588322, "grad_norm": 0.2481468766927719, "learning_rate": 9.095212187254078e-05, "loss": 0.0283, "step": 11740 }, { "epoch": 4.3150936467131835, "grad_norm": 0.44850897789001465, "learning_rate": 9.093313991502801e-05, "loss": 0.0312, "step": 11750 }, { "epoch": 4.318766066838046, "grad_norm": 0.27853184938430786, "learning_rate": 9.091414005196917e-05, "loss": 0.03, "step": 11760 }, { "epoch": 4.322438486962908, "grad_norm": 0.36502325534820557, "learning_rate": 9.089512229167545e-05, "loss": 0.0298, "step": 11770 }, { "epoch": 4.326110907087771, "grad_norm": 0.24253767728805542, "learning_rate": 9.087608664246587e-05, "loss": 0.028, "step": 11780 }, { "epoch": 4.329783327212633, "grad_norm": 0.21808089315891266, "learning_rate": 9.085703311266727e-05, "loss": 0.0306, "step": 11790 }, { "epoch": 4.333455747337496, "grad_norm": 0.29164963960647583, "learning_rate": 9.083796171061429e-05, "loss": 0.0298, "step": 11800 }, { "epoch": 4.337128167462358, "grad_norm": 0.3426056504249573, "learning_rate": 9.081887244464941e-05, "loss": 0.0294, "step": 11810 }, { "epoch": 4.3408005875872195, "grad_norm": 0.3579704761505127, "learning_rate": 9.079976532312297e-05, "loss": 0.0312, "step": 11820 }, { "epoch": 4.344473007712082, "grad_norm": 0.33658382296562195, "learning_rate": 9.078064035439301e-05, "loss": 0.0314, "step": 11830 }, { "epoch": 4.348145427836944, "grad_norm": 0.26844558119773865, "learning_rate": 9.07614975468255e-05, "loss": 0.0296, "step": 11840 }, { "epoch": 4.351817847961807, "grad_norm": 0.3030480444431305, "learning_rate": 9.074233690879412e-05, "loss": 0.0322, "step": 11850 }, { "epoch": 4.355490268086669, "grad_norm": 0.29617616534233093, "learning_rate": 9.072315844868038e-05, "loss": 0.033, "step": 11860 }, { "epoch": 4.359162688211532, "grad_norm": 0.34553584456443787, "learning_rate": 9.07039621748736e-05, "loss": 0.0325, "step": 11870 }, { "epoch": 4.362835108336394, "grad_norm": 0.3040260672569275, "learning_rate": 9.06847480957709e-05, "loss": 0.0291, "step": 11880 }, { "epoch": 4.366507528461256, "grad_norm": 0.2587890625, "learning_rate": 9.066551621977713e-05, "loss": 0.0314, "step": 11890 }, { "epoch": 4.370179948586118, "grad_norm": 1.3108793497085571, "learning_rate": 9.064626655530501e-05, "loss": 0.0341, "step": 11900 }, { "epoch": 4.37385236871098, "grad_norm": 0.3058304190635681, "learning_rate": 9.062699911077497e-05, "loss": 0.0322, "step": 11910 }, { "epoch": 4.377524788835843, "grad_norm": 0.5916448831558228, "learning_rate": 9.060771389461524e-05, "loss": 0.0321, "step": 11920 }, { "epoch": 4.381197208960705, "grad_norm": 0.3289046883583069, "learning_rate": 9.058841091526187e-05, "loss": 0.0294, "step": 11930 }, { "epoch": 4.384869629085568, "grad_norm": 0.2879277169704437, "learning_rate": 9.056909018115858e-05, "loss": 0.0323, "step": 11940 }, { "epoch": 4.38854204921043, "grad_norm": 0.4995669424533844, "learning_rate": 9.054975170075697e-05, "loss": 0.0313, "step": 11950 }, { "epoch": 4.392214469335292, "grad_norm": 0.1730368435382843, "learning_rate": 9.053039548251631e-05, "loss": 0.0299, "step": 11960 }, { "epoch": 4.395886889460154, "grad_norm": 0.27123725414276123, "learning_rate": 9.051102153490368e-05, "loss": 0.0334, "step": 11970 }, { "epoch": 4.399559309585016, "grad_norm": 0.3242649435997009, "learning_rate": 9.04916298663939e-05, "loss": 0.0301, "step": 11980 }, { "epoch": 4.403231729709879, "grad_norm": 0.2654187083244324, "learning_rate": 9.047222048546955e-05, "loss": 0.0307, "step": 11990 }, { "epoch": 4.406904149834741, "grad_norm": 0.41929516196250916, "learning_rate": 9.045279340062097e-05, "loss": 0.0292, "step": 12000 }, { "epoch": 4.410576569959604, "grad_norm": 0.226300910115242, "learning_rate": 9.043334862034618e-05, "loss": 0.0334, "step": 12010 }, { "epoch": 4.414248990084466, "grad_norm": 0.5617692470550537, "learning_rate": 9.041388615315102e-05, "loss": 0.0344, "step": 12020 }, { "epoch": 4.417921410209328, "grad_norm": 0.29709917306900024, "learning_rate": 9.039440600754905e-05, "loss": 0.0326, "step": 12030 }, { "epoch": 4.42159383033419, "grad_norm": 0.4672183394432068, "learning_rate": 9.037490819206151e-05, "loss": 0.0286, "step": 12040 }, { "epoch": 4.425266250459052, "grad_norm": 0.2926776707172394, "learning_rate": 9.035539271521744e-05, "loss": 0.0318, "step": 12050 }, { "epoch": 4.428938670583915, "grad_norm": 0.4260300397872925, "learning_rate": 9.033585958555356e-05, "loss": 0.0322, "step": 12060 }, { "epoch": 4.432611090708777, "grad_norm": 0.3386729061603546, "learning_rate": 9.031630881161431e-05, "loss": 0.0303, "step": 12070 }, { "epoch": 4.43628351083364, "grad_norm": 0.6141470670700073, "learning_rate": 9.029674040195186e-05, "loss": 0.0312, "step": 12080 }, { "epoch": 4.439955930958502, "grad_norm": 0.4446175694465637, "learning_rate": 9.027715436512613e-05, "loss": 0.0301, "step": 12090 }, { "epoch": 4.443628351083364, "grad_norm": 0.31750914454460144, "learning_rate": 9.02575507097047e-05, "loss": 0.0318, "step": 12100 }, { "epoch": 4.447300771208226, "grad_norm": 0.317455917596817, "learning_rate": 9.023792944426286e-05, "loss": 0.0312, "step": 12110 }, { "epoch": 4.450973191333088, "grad_norm": 0.2708892524242401, "learning_rate": 9.021829057738364e-05, "loss": 0.0304, "step": 12120 }, { "epoch": 4.454645611457951, "grad_norm": 0.34287962317466736, "learning_rate": 9.019863411765775e-05, "loss": 0.029, "step": 12130 }, { "epoch": 4.458318031582813, "grad_norm": 0.32432615756988525, "learning_rate": 9.017896007368357e-05, "loss": 0.0316, "step": 12140 }, { "epoch": 4.461990451707675, "grad_norm": 0.3623322546482086, "learning_rate": 9.015926845406722e-05, "loss": 0.0321, "step": 12150 }, { "epoch": 4.465662871832538, "grad_norm": 0.36757123470306396, "learning_rate": 9.013955926742245e-05, "loss": 0.0372, "step": 12160 }, { "epoch": 4.4693352919574, "grad_norm": 0.2233571708202362, "learning_rate": 9.011983252237077e-05, "loss": 0.0316, "step": 12170 }, { "epoch": 4.473007712082262, "grad_norm": 0.3674604892730713, "learning_rate": 9.01000882275413e-05, "loss": 0.0329, "step": 12180 }, { "epoch": 4.476680132207124, "grad_norm": 0.23020006716251373, "learning_rate": 9.008032639157088e-05, "loss": 0.032, "step": 12190 }, { "epoch": 4.480352552331987, "grad_norm": 0.2717619836330414, "learning_rate": 9.006054702310401e-05, "loss": 0.0265, "step": 12200 }, { "epoch": 4.484024972456849, "grad_norm": 0.3001994788646698, "learning_rate": 9.004075013079283e-05, "loss": 0.031, "step": 12210 }, { "epoch": 4.487697392581711, "grad_norm": 0.20221830904483795, "learning_rate": 9.00209357232972e-05, "loss": 0.0304, "step": 12220 }, { "epoch": 4.491369812706574, "grad_norm": 0.23727606236934662, "learning_rate": 9.000110380928461e-05, "loss": 0.0329, "step": 12230 }, { "epoch": 4.495042232831436, "grad_norm": 0.38708096742630005, "learning_rate": 8.998125439743021e-05, "loss": 0.0287, "step": 12240 }, { "epoch": 4.4987146529562985, "grad_norm": 0.3765375018119812, "learning_rate": 8.996138749641682e-05, "loss": 0.0299, "step": 12250 }, { "epoch": 4.50238707308116, "grad_norm": 0.33746087551116943, "learning_rate": 8.994150311493488e-05, "loss": 0.0302, "step": 12260 }, { "epoch": 4.506059493206022, "grad_norm": 0.30909106135368347, "learning_rate": 8.992160126168247e-05, "loss": 0.0313, "step": 12270 }, { "epoch": 4.509731913330885, "grad_norm": 0.2086952179670334, "learning_rate": 8.99016819453654e-05, "loss": 0.03, "step": 12280 }, { "epoch": 4.513404333455747, "grad_norm": 0.24405395984649658, "learning_rate": 8.988174517469702e-05, "loss": 0.0284, "step": 12290 }, { "epoch": 4.51707675358061, "grad_norm": 0.2548918128013611, "learning_rate": 8.986179095839835e-05, "loss": 0.0273, "step": 12300 }, { "epoch": 4.520749173705472, "grad_norm": 0.2461901307106018, "learning_rate": 8.984181930519804e-05, "loss": 0.0324, "step": 12310 }, { "epoch": 4.5244215938303345, "grad_norm": 0.3803139626979828, "learning_rate": 8.982183022383237e-05, "loss": 0.0282, "step": 12320 }, { "epoch": 4.528094013955196, "grad_norm": 0.5196132063865662, "learning_rate": 8.980182372304525e-05, "loss": 0.0276, "step": 12330 }, { "epoch": 4.531766434080058, "grad_norm": 0.31406131386756897, "learning_rate": 8.97817998115882e-05, "loss": 0.0306, "step": 12340 }, { "epoch": 4.535438854204921, "grad_norm": 0.2608523964881897, "learning_rate": 8.976175849822038e-05, "loss": 0.0308, "step": 12350 }, { "epoch": 4.539111274329783, "grad_norm": 0.2682661712169647, "learning_rate": 8.97416997917085e-05, "loss": 0.029, "step": 12360 }, { "epoch": 4.542783694454646, "grad_norm": 0.24757790565490723, "learning_rate": 8.972162370082695e-05, "loss": 0.0279, "step": 12370 }, { "epoch": 4.546456114579508, "grad_norm": 0.5627750158309937, "learning_rate": 8.97015302343577e-05, "loss": 0.0333, "step": 12380 }, { "epoch": 4.5501285347043705, "grad_norm": 0.4019037187099457, "learning_rate": 8.968141940109027e-05, "loss": 0.0296, "step": 12390 }, { "epoch": 4.553800954829232, "grad_norm": 0.2652541697025299, "learning_rate": 8.966129120982188e-05, "loss": 0.0325, "step": 12400 }, { "epoch": 4.557473374954094, "grad_norm": 0.4507718086242676, "learning_rate": 8.964114566935724e-05, "loss": 0.0307, "step": 12410 }, { "epoch": 4.561145795078957, "grad_norm": 0.3059037923812866, "learning_rate": 8.962098278850871e-05, "loss": 0.0307, "step": 12420 }, { "epoch": 4.564818215203819, "grad_norm": 0.28851738572120667, "learning_rate": 8.960080257609622e-05, "loss": 0.0321, "step": 12430 }, { "epoch": 4.568490635328682, "grad_norm": 0.18401303887367249, "learning_rate": 8.95806050409473e-05, "loss": 0.0302, "step": 12440 }, { "epoch": 4.572163055453544, "grad_norm": 0.2997156083583832, "learning_rate": 8.9560390191897e-05, "loss": 0.0293, "step": 12450 }, { "epoch": 4.5758354755784065, "grad_norm": 0.30969882011413574, "learning_rate": 8.9540158037788e-05, "loss": 0.0306, "step": 12460 }, { "epoch": 4.579507895703268, "grad_norm": 0.31957536935806274, "learning_rate": 8.951990858747054e-05, "loss": 0.0298, "step": 12470 }, { "epoch": 4.58318031582813, "grad_norm": 0.43467971682548523, "learning_rate": 8.94996418498024e-05, "loss": 0.0332, "step": 12480 }, { "epoch": 4.586852735952993, "grad_norm": 0.22954688966274261, "learning_rate": 8.947935783364896e-05, "loss": 0.0294, "step": 12490 }, { "epoch": 4.590525156077855, "grad_norm": 0.20307645201683044, "learning_rate": 8.945905654788311e-05, "loss": 0.0301, "step": 12500 }, { "epoch": 4.594197576202718, "grad_norm": 0.2599152624607086, "learning_rate": 8.943873800138535e-05, "loss": 0.0286, "step": 12510 }, { "epoch": 4.59786999632758, "grad_norm": 0.20548658072948456, "learning_rate": 8.94184022030437e-05, "loss": 0.0289, "step": 12520 }, { "epoch": 4.6015424164524426, "grad_norm": 0.2530631124973297, "learning_rate": 8.939804916175372e-05, "loss": 0.028, "step": 12530 }, { "epoch": 4.6052148365773045, "grad_norm": 0.2256859689950943, "learning_rate": 8.93776788864185e-05, "loss": 0.0299, "step": 12540 }, { "epoch": 4.608887256702166, "grad_norm": 0.3371962010860443, "learning_rate": 8.935729138594873e-05, "loss": 0.0308, "step": 12550 }, { "epoch": 4.612559676827029, "grad_norm": 0.2189456820487976, "learning_rate": 8.933688666926258e-05, "loss": 0.026, "step": 12560 }, { "epoch": 4.616232096951891, "grad_norm": 0.2957622706890106, "learning_rate": 8.931646474528575e-05, "loss": 0.0309, "step": 12570 }, { "epoch": 4.619904517076754, "grad_norm": 0.21797290444374084, "learning_rate": 8.929602562295151e-05, "loss": 0.028, "step": 12580 }, { "epoch": 4.623576937201616, "grad_norm": 0.3231044411659241, "learning_rate": 8.92755693112006e-05, "loss": 0.0289, "step": 12590 }, { "epoch": 4.627249357326479, "grad_norm": 0.2797386944293976, "learning_rate": 8.925509581898136e-05, "loss": 0.0289, "step": 12600 }, { "epoch": 4.6309217774513405, "grad_norm": 0.2552315890789032, "learning_rate": 8.923460515524951e-05, "loss": 0.028, "step": 12610 }, { "epoch": 4.634594197576202, "grad_norm": 0.19685098528862, "learning_rate": 8.921409732896842e-05, "loss": 0.0289, "step": 12620 }, { "epoch": 4.638266617701065, "grad_norm": 0.34121012687683105, "learning_rate": 8.919357234910887e-05, "loss": 0.0314, "step": 12630 }, { "epoch": 4.641939037825927, "grad_norm": 0.359062522649765, "learning_rate": 8.917303022464923e-05, "loss": 0.0293, "step": 12640 }, { "epoch": 4.64561145795079, "grad_norm": 0.2422504425048828, "learning_rate": 8.915247096457531e-05, "loss": 0.0293, "step": 12650 }, { "epoch": 4.649283878075652, "grad_norm": 0.2831946313381195, "learning_rate": 8.91318945778804e-05, "loss": 0.0312, "step": 12660 }, { "epoch": 4.652956298200515, "grad_norm": 0.3987691104412079, "learning_rate": 8.911130107356534e-05, "loss": 0.0312, "step": 12670 }, { "epoch": 4.6566287183253765, "grad_norm": 0.3077241778373718, "learning_rate": 8.90906904606384e-05, "loss": 0.029, "step": 12680 }, { "epoch": 4.660301138450238, "grad_norm": 0.4278874099254608, "learning_rate": 8.90700627481154e-05, "loss": 0.0313, "step": 12690 }, { "epoch": 4.663973558575101, "grad_norm": 0.19896098971366882, "learning_rate": 8.904941794501957e-05, "loss": 0.0307, "step": 12700 }, { "epoch": 4.667645978699963, "grad_norm": 0.4436928331851959, "learning_rate": 8.902875606038166e-05, "loss": 0.0282, "step": 12710 }, { "epoch": 4.671318398824826, "grad_norm": 0.30717217922210693, "learning_rate": 8.900807710323989e-05, "loss": 0.0299, "step": 12720 }, { "epoch": 4.674990818949688, "grad_norm": 0.21817220747470856, "learning_rate": 8.898738108263993e-05, "loss": 0.0292, "step": 12730 }, { "epoch": 4.678663239074551, "grad_norm": 0.29704633355140686, "learning_rate": 8.896666800763491e-05, "loss": 0.0288, "step": 12740 }, { "epoch": 4.6823356591994125, "grad_norm": 0.30308791995048523, "learning_rate": 8.894593788728546e-05, "loss": 0.0301, "step": 12750 }, { "epoch": 4.686008079324274, "grad_norm": 0.36495861411094666, "learning_rate": 8.892519073065961e-05, "loss": 0.0296, "step": 12760 }, { "epoch": 4.689680499449137, "grad_norm": 0.20812030136585236, "learning_rate": 8.89044265468329e-05, "loss": 0.0293, "step": 12770 }, { "epoch": 4.693352919573999, "grad_norm": 0.2994144856929779, "learning_rate": 8.888364534488827e-05, "loss": 0.0329, "step": 12780 }, { "epoch": 4.697025339698862, "grad_norm": 0.2647581398487091, "learning_rate": 8.886284713391613e-05, "loss": 0.0329, "step": 12790 }, { "epoch": 4.700697759823724, "grad_norm": 0.3222537636756897, "learning_rate": 8.884203192301431e-05, "loss": 0.0318, "step": 12800 }, { "epoch": 4.704370179948586, "grad_norm": 0.38024404644966125, "learning_rate": 8.88211997212881e-05, "loss": 0.0309, "step": 12810 }, { "epoch": 4.7080426000734485, "grad_norm": 0.28931647539138794, "learning_rate": 8.880035053785023e-05, "loss": 0.0303, "step": 12820 }, { "epoch": 4.7117150201983105, "grad_norm": 0.4664151072502136, "learning_rate": 8.877948438182083e-05, "loss": 0.0312, "step": 12830 }, { "epoch": 4.715387440323173, "grad_norm": 0.24017931520938873, "learning_rate": 8.875860126232745e-05, "loss": 0.0298, "step": 12840 }, { "epoch": 4.719059860448035, "grad_norm": 0.3569366931915283, "learning_rate": 8.87377011885051e-05, "loss": 0.0292, "step": 12850 }, { "epoch": 4.722732280572897, "grad_norm": 0.3294772803783417, "learning_rate": 8.871678416949617e-05, "loss": 0.0294, "step": 12860 }, { "epoch": 4.72640470069776, "grad_norm": 0.3218986988067627, "learning_rate": 8.869585021445046e-05, "loss": 0.0303, "step": 12870 }, { "epoch": 4.730077120822622, "grad_norm": 0.345134437084198, "learning_rate": 8.867489933252521e-05, "loss": 0.0302, "step": 12880 }, { "epoch": 4.733749540947485, "grad_norm": 0.2812615633010864, "learning_rate": 8.865393153288504e-05, "loss": 0.0289, "step": 12890 }, { "epoch": 4.7374219610723465, "grad_norm": 0.3268434703350067, "learning_rate": 8.8632946824702e-05, "loss": 0.0277, "step": 12900 }, { "epoch": 4.741094381197209, "grad_norm": 0.32242587208747864, "learning_rate": 8.86119452171555e-05, "loss": 0.0269, "step": 12910 }, { "epoch": 4.744766801322071, "grad_norm": 0.3042134642601013, "learning_rate": 8.859092671943234e-05, "loss": 0.0296, "step": 12920 }, { "epoch": 4.748439221446933, "grad_norm": 0.3386726975440979, "learning_rate": 8.856989134072676e-05, "loss": 0.0305, "step": 12930 }, { "epoch": 4.752111641571796, "grad_norm": 0.26832830905914307, "learning_rate": 8.85488390902403e-05, "loss": 0.0278, "step": 12940 }, { "epoch": 4.755784061696658, "grad_norm": 0.2194317728281021, "learning_rate": 8.852776997718199e-05, "loss": 0.028, "step": 12950 }, { "epoch": 4.759456481821521, "grad_norm": 0.3843802213668823, "learning_rate": 8.850668401076812e-05, "loss": 0.0314, "step": 12960 }, { "epoch": 4.7631289019463825, "grad_norm": 0.33294108510017395, "learning_rate": 8.848558120022246e-05, "loss": 0.031, "step": 12970 }, { "epoch": 4.766801322071245, "grad_norm": 0.20293930172920227, "learning_rate": 8.846446155477603e-05, "loss": 0.0293, "step": 12980 }, { "epoch": 4.770473742196107, "grad_norm": 0.31210893392562866, "learning_rate": 8.844332508366735e-05, "loss": 0.0282, "step": 12990 }, { "epoch": 4.774146162320969, "grad_norm": 0.2929299473762512, "learning_rate": 8.84221717961422e-05, "loss": 0.0331, "step": 13000 }, { "epoch": 4.777818582445832, "grad_norm": 0.22757446765899658, "learning_rate": 8.840100170145374e-05, "loss": 0.0307, "step": 13010 }, { "epoch": 4.781491002570694, "grad_norm": 0.3484274744987488, "learning_rate": 8.837981480886249e-05, "loss": 0.0291, "step": 13020 }, { "epoch": 4.785163422695557, "grad_norm": 0.4061032235622406, "learning_rate": 8.835861112763633e-05, "loss": 0.0303, "step": 13030 }, { "epoch": 4.7888358428204185, "grad_norm": 0.2843060791492462, "learning_rate": 8.833739066705044e-05, "loss": 0.0286, "step": 13040 }, { "epoch": 4.792508262945281, "grad_norm": 0.36971181631088257, "learning_rate": 8.831615343638742e-05, "loss": 0.0291, "step": 13050 }, { "epoch": 4.796180683070143, "grad_norm": 0.2603151798248291, "learning_rate": 8.829489944493711e-05, "loss": 0.0299, "step": 13060 }, { "epoch": 4.799853103195005, "grad_norm": 0.2255980670452118, "learning_rate": 8.827362870199675e-05, "loss": 0.0269, "step": 13070 }, { "epoch": 4.803525523319868, "grad_norm": 0.242426335811615, "learning_rate": 8.825234121687089e-05, "loss": 0.0276, "step": 13080 }, { "epoch": 4.80719794344473, "grad_norm": 0.5236140489578247, "learning_rate": 8.823103699887139e-05, "loss": 0.0311, "step": 13090 }, { "epoch": 4.810870363569593, "grad_norm": 0.27965953946113586, "learning_rate": 8.820971605731745e-05, "loss": 0.0294, "step": 13100 }, { "epoch": 4.8145427836944545, "grad_norm": 0.18627698719501495, "learning_rate": 8.818837840153556e-05, "loss": 0.0273, "step": 13110 }, { "epoch": 4.818215203819317, "grad_norm": 0.3338632881641388, "learning_rate": 8.816702404085952e-05, "loss": 0.0318, "step": 13120 }, { "epoch": 4.821887623944179, "grad_norm": 0.21443797647953033, "learning_rate": 8.814565298463048e-05, "loss": 0.0313, "step": 13130 }, { "epoch": 4.825560044069041, "grad_norm": 0.30508503317832947, "learning_rate": 8.812426524219688e-05, "loss": 0.0284, "step": 13140 }, { "epoch": 4.829232464193904, "grad_norm": 0.21384692192077637, "learning_rate": 8.81028608229144e-05, "loss": 0.0286, "step": 13150 }, { "epoch": 4.832904884318766, "grad_norm": 0.3014053404331207, "learning_rate": 8.808143973614611e-05, "loss": 0.027, "step": 13160 }, { "epoch": 4.836577304443629, "grad_norm": 0.544240415096283, "learning_rate": 8.806000199126228e-05, "loss": 0.0294, "step": 13170 }, { "epoch": 4.840249724568491, "grad_norm": 0.235159769654274, "learning_rate": 8.803854759764052e-05, "loss": 0.0294, "step": 13180 }, { "epoch": 4.843922144693353, "grad_norm": 0.3317994177341461, "learning_rate": 8.801707656466572e-05, "loss": 0.0273, "step": 13190 }, { "epoch": 4.847594564818215, "grad_norm": 0.25029268860816956, "learning_rate": 8.799558890173003e-05, "loss": 0.0307, "step": 13200 }, { "epoch": 4.851266984943077, "grad_norm": 0.2527422606945038, "learning_rate": 8.79740846182329e-05, "loss": 0.0299, "step": 13210 }, { "epoch": 4.85493940506794, "grad_norm": 0.21002531051635742, "learning_rate": 8.7952563723581e-05, "loss": 0.0292, "step": 13220 }, { "epoch": 4.858611825192802, "grad_norm": 0.28536301851272583, "learning_rate": 8.793102622718834e-05, "loss": 0.0303, "step": 13230 }, { "epoch": 4.862284245317665, "grad_norm": 0.2482774257659912, "learning_rate": 8.790947213847613e-05, "loss": 0.0309, "step": 13240 }, { "epoch": 4.865956665442527, "grad_norm": 0.4032038152217865, "learning_rate": 8.788790146687286e-05, "loss": 0.0303, "step": 13250 }, { "epoch": 4.869629085567389, "grad_norm": 0.22761279344558716, "learning_rate": 8.786631422181429e-05, "loss": 0.0314, "step": 13260 }, { "epoch": 4.873301505692251, "grad_norm": 0.22703050076961517, "learning_rate": 8.78447104127434e-05, "loss": 0.0316, "step": 13270 }, { "epoch": 4.876973925817113, "grad_norm": 0.24038803577423096, "learning_rate": 8.782309004911042e-05, "loss": 0.0299, "step": 13280 }, { "epoch": 4.880646345941976, "grad_norm": 0.37212130427360535, "learning_rate": 8.780145314037286e-05, "loss": 0.0298, "step": 13290 }, { "epoch": 4.884318766066838, "grad_norm": 0.3619602620601654, "learning_rate": 8.777979969599542e-05, "loss": 0.0309, "step": 13300 }, { "epoch": 4.887991186191701, "grad_norm": 0.23806971311569214, "learning_rate": 8.775812972545006e-05, "loss": 0.0253, "step": 13310 }, { "epoch": 4.891663606316563, "grad_norm": 0.2585442066192627, "learning_rate": 8.773644323821596e-05, "loss": 0.0287, "step": 13320 }, { "epoch": 4.8953360264414245, "grad_norm": 0.23996718227863312, "learning_rate": 8.771474024377953e-05, "loss": 0.0298, "step": 13330 }, { "epoch": 4.899008446566287, "grad_norm": 0.3103991746902466, "learning_rate": 8.769302075163438e-05, "loss": 0.0268, "step": 13340 }, { "epoch": 4.902680866691149, "grad_norm": 0.5350465178489685, "learning_rate": 8.767128477128137e-05, "loss": 0.0276, "step": 13350 }, { "epoch": 4.906353286816012, "grad_norm": 0.3212592899799347, "learning_rate": 8.764953231222854e-05, "loss": 0.0305, "step": 13360 }, { "epoch": 4.910025706940874, "grad_norm": 0.2798633873462677, "learning_rate": 8.762776338399119e-05, "loss": 0.0293, "step": 13370 }, { "epoch": 4.913698127065736, "grad_norm": 0.5674859881401062, "learning_rate": 8.760597799609176e-05, "loss": 0.0297, "step": 13380 }, { "epoch": 4.917370547190599, "grad_norm": 0.41166120767593384, "learning_rate": 8.758417615805992e-05, "loss": 0.0291, "step": 13390 }, { "epoch": 4.9210429673154605, "grad_norm": 0.26505473256111145, "learning_rate": 8.756235787943254e-05, "loss": 0.026, "step": 13400 }, { "epoch": 4.924715387440323, "grad_norm": 0.3068337142467499, "learning_rate": 8.754052316975367e-05, "loss": 0.0299, "step": 13410 }, { "epoch": 4.928387807565185, "grad_norm": 0.30054402351379395, "learning_rate": 8.751867203857455e-05, "loss": 0.0299, "step": 13420 }, { "epoch": 4.932060227690048, "grad_norm": 0.4350398778915405, "learning_rate": 8.749680449545363e-05, "loss": 0.0311, "step": 13430 }, { "epoch": 4.93573264781491, "grad_norm": 0.5344095230102539, "learning_rate": 8.747492054995649e-05, "loss": 0.0322, "step": 13440 }, { "epoch": 4.939405067939772, "grad_norm": 0.1881568282842636, "learning_rate": 8.745302021165595e-05, "loss": 0.0292, "step": 13450 }, { "epoch": 4.943077488064635, "grad_norm": 0.30106082558631897, "learning_rate": 8.743110349013192e-05, "loss": 0.0314, "step": 13460 }, { "epoch": 4.946749908189497, "grad_norm": 0.2785680592060089, "learning_rate": 8.740917039497153e-05, "loss": 0.0303, "step": 13470 }, { "epoch": 4.950422328314359, "grad_norm": 0.2391214370727539, "learning_rate": 8.738722093576906e-05, "loss": 0.0264, "step": 13480 }, { "epoch": 4.954094748439221, "grad_norm": 0.34809255599975586, "learning_rate": 8.736525512212597e-05, "loss": 0.0334, "step": 13490 }, { "epoch": 4.957767168564084, "grad_norm": 0.42494654655456543, "learning_rate": 8.734327296365084e-05, "loss": 0.0311, "step": 13500 }, { "epoch": 4.961439588688946, "grad_norm": 0.292166531085968, "learning_rate": 8.732127446995939e-05, "loss": 0.0328, "step": 13510 }, { "epoch": 4.965112008813808, "grad_norm": 0.31973591446876526, "learning_rate": 8.729925965067454e-05, "loss": 0.0284, "step": 13520 }, { "epoch": 4.968784428938671, "grad_norm": 0.33787208795547485, "learning_rate": 8.72772285154263e-05, "loss": 0.0294, "step": 13530 }, { "epoch": 4.972456849063533, "grad_norm": 0.32420578598976135, "learning_rate": 8.725518107385187e-05, "loss": 0.0337, "step": 13540 }, { "epoch": 4.976129269188395, "grad_norm": 0.24357233941555023, "learning_rate": 8.72331173355955e-05, "loss": 0.0283, "step": 13550 }, { "epoch": 4.979801689313257, "grad_norm": 0.3369728624820709, "learning_rate": 8.721103731030867e-05, "loss": 0.0296, "step": 13560 }, { "epoch": 4.98347410943812, "grad_norm": 0.34278592467308044, "learning_rate": 8.718894100764989e-05, "loss": 0.0275, "step": 13570 }, { "epoch": 4.987146529562982, "grad_norm": 0.25273656845092773, "learning_rate": 8.716682843728485e-05, "loss": 0.0311, "step": 13580 }, { "epoch": 4.990818949687844, "grad_norm": 0.36297500133514404, "learning_rate": 8.714469960888634e-05, "loss": 0.0292, "step": 13590 }, { "epoch": 4.994491369812707, "grad_norm": 0.23331964015960693, "learning_rate": 8.712255453213427e-05, "loss": 0.0308, "step": 13600 }, { "epoch": 4.998163789937569, "grad_norm": 0.26855674386024475, "learning_rate": 8.710039321671563e-05, "loss": 0.0292, "step": 13610 }, { "epoch": 5.001836210062431, "grad_norm": 0.2753860056400299, "learning_rate": 8.707821567232456e-05, "loss": 0.0292, "step": 13620 }, { "epoch": 5.005508630187293, "grad_norm": 0.2758404016494751, "learning_rate": 8.705602190866225e-05, "loss": 0.0311, "step": 13630 }, { "epoch": 5.009181050312156, "grad_norm": 0.2256908416748047, "learning_rate": 8.703381193543701e-05, "loss": 0.0286, "step": 13640 }, { "epoch": 5.012853470437018, "grad_norm": 0.3057275712490082, "learning_rate": 8.701158576236423e-05, "loss": 0.0275, "step": 13650 }, { "epoch": 5.01652589056188, "grad_norm": 0.2725174129009247, "learning_rate": 8.69893433991664e-05, "loss": 0.0293, "step": 13660 }, { "epoch": 5.020198310686743, "grad_norm": 0.22623220086097717, "learning_rate": 8.69670848555731e-05, "loss": 0.0265, "step": 13670 }, { "epoch": 5.023870730811605, "grad_norm": 0.316595196723938, "learning_rate": 8.694481014132096e-05, "loss": 0.0307, "step": 13680 }, { "epoch": 5.027543150936467, "grad_norm": 0.20911888778209686, "learning_rate": 8.69225192661537e-05, "loss": 0.0259, "step": 13690 }, { "epoch": 5.031215571061329, "grad_norm": 0.41656339168548584, "learning_rate": 8.690021223982208e-05, "loss": 0.0285, "step": 13700 }, { "epoch": 5.034887991186192, "grad_norm": 0.25245392322540283, "learning_rate": 8.687788907208398e-05, "loss": 0.0295, "step": 13710 }, { "epoch": 5.038560411311054, "grad_norm": 0.2974954843521118, "learning_rate": 8.685554977270431e-05, "loss": 0.0284, "step": 13720 }, { "epoch": 5.042232831435916, "grad_norm": 0.24956844747066498, "learning_rate": 8.683319435145503e-05, "loss": 0.0289, "step": 13730 }, { "epoch": 5.045905251560779, "grad_norm": 0.2695448696613312, "learning_rate": 8.681082281811517e-05, "loss": 0.0279, "step": 13740 }, { "epoch": 5.049577671685641, "grad_norm": 0.3032446801662445, "learning_rate": 8.67884351824708e-05, "loss": 0.031, "step": 13750 }, { "epoch": 5.053250091810503, "grad_norm": 0.2690153121948242, "learning_rate": 8.676603145431501e-05, "loss": 0.0294, "step": 13760 }, { "epoch": 5.056922511935365, "grad_norm": 0.30473148822784424, "learning_rate": 8.674361164344799e-05, "loss": 0.0282, "step": 13770 }, { "epoch": 5.060594932060227, "grad_norm": 0.3975018262863159, "learning_rate": 8.672117575967688e-05, "loss": 0.0282, "step": 13780 }, { "epoch": 5.06426735218509, "grad_norm": 0.3588855266571045, "learning_rate": 8.669872381281595e-05, "loss": 0.0298, "step": 13790 }, { "epoch": 5.067939772309952, "grad_norm": 0.2649274468421936, "learning_rate": 8.667625581268639e-05, "loss": 0.0296, "step": 13800 }, { "epoch": 5.071612192434815, "grad_norm": 0.21977584064006805, "learning_rate": 8.665377176911651e-05, "loss": 0.031, "step": 13810 }, { "epoch": 5.075284612559677, "grad_norm": 0.2546766996383667, "learning_rate": 8.663127169194159e-05, "loss": 0.0275, "step": 13820 }, { "epoch": 5.0789570326845395, "grad_norm": 0.2584453523159027, "learning_rate": 8.660875559100389e-05, "loss": 0.0294, "step": 13830 }, { "epoch": 5.082629452809401, "grad_norm": 0.3204551935195923, "learning_rate": 8.658622347615274e-05, "loss": 0.0283, "step": 13840 }, { "epoch": 5.086301872934263, "grad_norm": 0.21930430829524994, "learning_rate": 8.656367535724448e-05, "loss": 0.0275, "step": 13850 }, { "epoch": 5.089974293059126, "grad_norm": 0.3362864851951599, "learning_rate": 8.65411112441424e-05, "loss": 0.031, "step": 13860 }, { "epoch": 5.093646713183988, "grad_norm": 0.2723229229450226, "learning_rate": 8.651853114671679e-05, "loss": 0.0271, "step": 13870 }, { "epoch": 5.097319133308851, "grad_norm": 0.2778928875923157, "learning_rate": 8.649593507484499e-05, "loss": 0.0281, "step": 13880 }, { "epoch": 5.100991553433713, "grad_norm": 0.26327401399612427, "learning_rate": 8.647332303841126e-05, "loss": 0.0283, "step": 13890 }, { "epoch": 5.1046639735585755, "grad_norm": 0.21511173248291016, "learning_rate": 8.645069504730689e-05, "loss": 0.0278, "step": 13900 }, { "epoch": 5.108336393683437, "grad_norm": 0.3428310453891754, "learning_rate": 8.64280511114301e-05, "loss": 0.0275, "step": 13910 }, { "epoch": 5.112008813808299, "grad_norm": 0.2693440318107605, "learning_rate": 8.640539124068617e-05, "loss": 0.0287, "step": 13920 }, { "epoch": 5.115681233933162, "grad_norm": 0.2965971827507019, "learning_rate": 8.638271544498727e-05, "loss": 0.0285, "step": 13930 }, { "epoch": 5.119353654058024, "grad_norm": 0.26893937587738037, "learning_rate": 8.636002373425257e-05, "loss": 0.0286, "step": 13940 }, { "epoch": 5.123026074182887, "grad_norm": 0.35257402062416077, "learning_rate": 8.633731611840817e-05, "loss": 0.0293, "step": 13950 }, { "epoch": 5.126698494307749, "grad_norm": 0.24392551183700562, "learning_rate": 8.631459260738717e-05, "loss": 0.0272, "step": 13960 }, { "epoch": 5.1303709144326115, "grad_norm": 0.26036423444747925, "learning_rate": 8.62918532111296e-05, "loss": 0.0261, "step": 13970 }, { "epoch": 5.134043334557473, "grad_norm": 0.3489587903022766, "learning_rate": 8.626909793958248e-05, "loss": 0.0258, "step": 13980 }, { "epoch": 5.137715754682335, "grad_norm": 0.27549928426742554, "learning_rate": 8.624632680269969e-05, "loss": 0.0268, "step": 13990 }, { "epoch": 5.141388174807198, "grad_norm": 0.3008381128311157, "learning_rate": 8.622353981044212e-05, "loss": 0.0284, "step": 14000 }, { "epoch": 5.14506059493206, "grad_norm": 0.20696213841438293, "learning_rate": 8.620073697277757e-05, "loss": 0.0281, "step": 14010 }, { "epoch": 5.148733015056923, "grad_norm": 0.2428346872329712, "learning_rate": 8.617791829968079e-05, "loss": 0.0293, "step": 14020 }, { "epoch": 5.152405435181785, "grad_norm": 0.18842558562755585, "learning_rate": 8.615508380113344e-05, "loss": 0.0272, "step": 14030 }, { "epoch": 5.156077855306647, "grad_norm": 0.20758944749832153, "learning_rate": 8.613223348712408e-05, "loss": 0.0273, "step": 14040 }, { "epoch": 5.159750275431509, "grad_norm": 0.2064233124256134, "learning_rate": 8.610936736764824e-05, "loss": 0.0249, "step": 14050 }, { "epoch": 5.163422695556371, "grad_norm": 0.24117974936962128, "learning_rate": 8.608648545270833e-05, "loss": 0.0255, "step": 14060 }, { "epoch": 5.167095115681234, "grad_norm": 0.25220787525177, "learning_rate": 8.606358775231366e-05, "loss": 0.0268, "step": 14070 }, { "epoch": 5.170767535806096, "grad_norm": 0.2605541944503784, "learning_rate": 8.60406742764805e-05, "loss": 0.0268, "step": 14080 }, { "epoch": 5.174439955930959, "grad_norm": 0.25376275181770325, "learning_rate": 8.601774503523195e-05, "loss": 0.027, "step": 14090 }, { "epoch": 5.178112376055821, "grad_norm": 0.311712384223938, "learning_rate": 8.599480003859805e-05, "loss": 0.0288, "step": 14100 }, { "epoch": 5.181784796180683, "grad_norm": 0.2947756350040436, "learning_rate": 8.597183929661573e-05, "loss": 0.0284, "step": 14110 }, { "epoch": 5.1854572163055455, "grad_norm": 0.3609215021133423, "learning_rate": 8.594886281932879e-05, "loss": 0.025, "step": 14120 }, { "epoch": 5.189129636430407, "grad_norm": 0.2714504897594452, "learning_rate": 8.59258706167879e-05, "loss": 0.0283, "step": 14130 }, { "epoch": 5.19280205655527, "grad_norm": 0.2215922772884369, "learning_rate": 8.590286269905068e-05, "loss": 0.0299, "step": 14140 }, { "epoch": 5.196474476680132, "grad_norm": 0.2586687207221985, "learning_rate": 8.587983907618154e-05, "loss": 0.0302, "step": 14150 }, { "epoch": 5.200146896804995, "grad_norm": 0.3093457520008087, "learning_rate": 8.585679975825178e-05, "loss": 0.0297, "step": 14160 }, { "epoch": 5.203819316929857, "grad_norm": 0.3444720208644867, "learning_rate": 8.583374475533962e-05, "loss": 0.0292, "step": 14170 }, { "epoch": 5.207491737054719, "grad_norm": 0.2680359184741974, "learning_rate": 8.581067407753009e-05, "loss": 0.0297, "step": 14180 }, { "epoch": 5.2111641571795815, "grad_norm": 0.2609495222568512, "learning_rate": 8.578758773491507e-05, "loss": 0.0282, "step": 14190 }, { "epoch": 5.214836577304443, "grad_norm": 0.3387976586818695, "learning_rate": 8.576448573759332e-05, "loss": 0.0276, "step": 14200 }, { "epoch": 5.218508997429306, "grad_norm": 0.25863122940063477, "learning_rate": 8.574136809567044e-05, "loss": 0.0282, "step": 14210 }, { "epoch": 5.222181417554168, "grad_norm": 0.22368746995925903, "learning_rate": 8.57182348192589e-05, "loss": 0.0282, "step": 14220 }, { "epoch": 5.225853837679031, "grad_norm": 0.2694986164569855, "learning_rate": 8.569508591847792e-05, "loss": 0.0277, "step": 14230 }, { "epoch": 5.229526257803893, "grad_norm": 0.22656206786632538, "learning_rate": 8.567192140345367e-05, "loss": 0.0259, "step": 14240 }, { "epoch": 5.233198677928755, "grad_norm": 0.34101617336273193, "learning_rate": 8.564874128431906e-05, "loss": 0.0294, "step": 14250 }, { "epoch": 5.2368710980536175, "grad_norm": 0.27664121985435486, "learning_rate": 8.562554557121389e-05, "loss": 0.0303, "step": 14260 }, { "epoch": 5.240543518178479, "grad_norm": 0.2548114061355591, "learning_rate": 8.560233427428475e-05, "loss": 0.027, "step": 14270 }, { "epoch": 5.244215938303342, "grad_norm": 0.2637856900691986, "learning_rate": 8.557910740368503e-05, "loss": 0.0278, "step": 14280 }, { "epoch": 5.247888358428204, "grad_norm": 0.32461482286453247, "learning_rate": 8.555586496957495e-05, "loss": 0.0275, "step": 14290 }, { "epoch": 5.251560778553067, "grad_norm": 0.22362478077411652, "learning_rate": 8.553260698212155e-05, "loss": 0.0287, "step": 14300 }, { "epoch": 5.255233198677929, "grad_norm": 0.29319748282432556, "learning_rate": 8.550933345149868e-05, "loss": 0.0274, "step": 14310 }, { "epoch": 5.258905618802791, "grad_norm": 0.4031248986721039, "learning_rate": 8.548604438788696e-05, "loss": 0.0298, "step": 14320 }, { "epoch": 5.2625780389276535, "grad_norm": 0.2980773448944092, "learning_rate": 8.546273980147383e-05, "loss": 0.0308, "step": 14330 }, { "epoch": 5.266250459052515, "grad_norm": 0.31330785155296326, "learning_rate": 8.543941970245348e-05, "loss": 0.0294, "step": 14340 }, { "epoch": 5.269922879177378, "grad_norm": 0.30465853214263916, "learning_rate": 8.541608410102693e-05, "loss": 0.0267, "step": 14350 }, { "epoch": 5.27359529930224, "grad_norm": 0.3249189257621765, "learning_rate": 8.539273300740195e-05, "loss": 0.0298, "step": 14360 }, { "epoch": 5.277267719427103, "grad_norm": 0.19223812222480774, "learning_rate": 8.536936643179313e-05, "loss": 0.028, "step": 14370 }, { "epoch": 5.280940139551965, "grad_norm": 0.20636962354183197, "learning_rate": 8.534598438442179e-05, "loss": 0.0277, "step": 14380 }, { "epoch": 5.284612559676827, "grad_norm": 0.286969393491745, "learning_rate": 8.532258687551603e-05, "loss": 0.0296, "step": 14390 }, { "epoch": 5.2882849798016895, "grad_norm": 0.508809506893158, "learning_rate": 8.529917391531071e-05, "loss": 0.0285, "step": 14400 }, { "epoch": 5.2919573999265515, "grad_norm": 0.46257010102272034, "learning_rate": 8.527574551404747e-05, "loss": 0.0296, "step": 14410 }, { "epoch": 5.295629820051414, "grad_norm": 0.33848631381988525, "learning_rate": 8.525230168197468e-05, "loss": 0.0273, "step": 14420 }, { "epoch": 5.299302240176276, "grad_norm": 0.3503470718860626, "learning_rate": 8.522884242934745e-05, "loss": 0.0272, "step": 14430 }, { "epoch": 5.302974660301138, "grad_norm": 0.298374742269516, "learning_rate": 8.520536776642768e-05, "loss": 0.0277, "step": 14440 }, { "epoch": 5.306647080426001, "grad_norm": 0.232881098985672, "learning_rate": 8.5181877703484e-05, "loss": 0.0325, "step": 14450 }, { "epoch": 5.310319500550863, "grad_norm": 0.2890375852584839, "learning_rate": 8.51583722507917e-05, "loss": 0.0276, "step": 14460 }, { "epoch": 5.313991920675726, "grad_norm": 0.26234373450279236, "learning_rate": 8.513485141863293e-05, "loss": 0.0297, "step": 14470 }, { "epoch": 5.3176643408005875, "grad_norm": 0.3937322199344635, "learning_rate": 8.511131521729647e-05, "loss": 0.0313, "step": 14480 }, { "epoch": 5.32133676092545, "grad_norm": 0.26279449462890625, "learning_rate": 8.508776365707787e-05, "loss": 0.0274, "step": 14490 }, { "epoch": 5.325009181050312, "grad_norm": 0.3422848582267761, "learning_rate": 8.506419674827934e-05, "loss": 0.0268, "step": 14500 }, { "epoch": 5.328681601175174, "grad_norm": 0.21661919355392456, "learning_rate": 8.50406145012099e-05, "loss": 0.0258, "step": 14510 }, { "epoch": 5.332354021300037, "grad_norm": 0.22451700270175934, "learning_rate": 8.501701692618519e-05, "loss": 0.0251, "step": 14520 }, { "epoch": 5.336026441424899, "grad_norm": 0.27322515845298767, "learning_rate": 8.499340403352761e-05, "loss": 0.0275, "step": 14530 }, { "epoch": 5.339698861549762, "grad_norm": 0.20184707641601562, "learning_rate": 8.496977583356623e-05, "loss": 0.0265, "step": 14540 }, { "epoch": 5.3433712816746235, "grad_norm": 0.3038698434829712, "learning_rate": 8.494613233663684e-05, "loss": 0.0287, "step": 14550 }, { "epoch": 5.347043701799485, "grad_norm": 0.22549887001514435, "learning_rate": 8.492247355308189e-05, "loss": 0.027, "step": 14560 }, { "epoch": 5.350716121924348, "grad_norm": 0.34337007999420166, "learning_rate": 8.489879949325056e-05, "loss": 0.0271, "step": 14570 }, { "epoch": 5.35438854204921, "grad_norm": 0.24280047416687012, "learning_rate": 8.487511016749868e-05, "loss": 0.0274, "step": 14580 }, { "epoch": 5.358060962174073, "grad_norm": 0.2682701349258423, "learning_rate": 8.485140558618874e-05, "loss": 0.0279, "step": 14590 }, { "epoch": 5.361733382298935, "grad_norm": 0.28286120295524597, "learning_rate": 8.482768575968995e-05, "loss": 0.0265, "step": 14600 }, { "epoch": 5.365405802423798, "grad_norm": 0.5607072114944458, "learning_rate": 8.480395069837818e-05, "loss": 0.0291, "step": 14610 }, { "epoch": 5.3690782225486595, "grad_norm": 0.2594565451145172, "learning_rate": 8.478020041263595e-05, "loss": 0.0263, "step": 14620 }, { "epoch": 5.372750642673521, "grad_norm": 0.22953617572784424, "learning_rate": 8.475643491285242e-05, "loss": 0.0304, "step": 14630 }, { "epoch": 5.376423062798384, "grad_norm": 0.2842211425304413, "learning_rate": 8.473265420942345e-05, "loss": 0.0281, "step": 14640 }, { "epoch": 5.380095482923246, "grad_norm": 0.32914119958877563, "learning_rate": 8.470885831275151e-05, "loss": 0.0292, "step": 14650 }, { "epoch": 5.383767903048109, "grad_norm": 0.26332926750183105, "learning_rate": 8.468504723324574e-05, "loss": 0.0264, "step": 14660 }, { "epoch": 5.387440323172971, "grad_norm": 0.23071512579917908, "learning_rate": 8.466122098132193e-05, "loss": 0.0265, "step": 14670 }, { "epoch": 5.391112743297834, "grad_norm": 0.26753708720207214, "learning_rate": 8.463737956740245e-05, "loss": 0.0275, "step": 14680 }, { "epoch": 5.3947851634226955, "grad_norm": 0.24299344420433044, "learning_rate": 8.461352300191639e-05, "loss": 0.0253, "step": 14690 }, { "epoch": 5.3984575835475574, "grad_norm": 0.16719795763492584, "learning_rate": 8.45896512952994e-05, "loss": 0.026, "step": 14700 }, { "epoch": 5.40213000367242, "grad_norm": 0.3418828845024109, "learning_rate": 8.456576445799377e-05, "loss": 0.0349, "step": 14710 }, { "epoch": 5.405802423797282, "grad_norm": 0.21807077527046204, "learning_rate": 8.454186250044844e-05, "loss": 0.0299, "step": 14720 }, { "epoch": 5.409474843922145, "grad_norm": 0.27975994348526, "learning_rate": 8.451794543311892e-05, "loss": 0.0293, "step": 14730 }, { "epoch": 5.413147264047007, "grad_norm": 0.3666967749595642, "learning_rate": 8.449401326646736e-05, "loss": 0.0305, "step": 14740 }, { "epoch": 5.41681968417187, "grad_norm": 0.3488795757293701, "learning_rate": 8.447006601096248e-05, "loss": 0.0292, "step": 14750 }, { "epoch": 5.420492104296732, "grad_norm": 0.3235775828361511, "learning_rate": 8.444610367707964e-05, "loss": 0.0275, "step": 14760 }, { "epoch": 5.4241645244215935, "grad_norm": 0.3662128746509552, "learning_rate": 8.442212627530078e-05, "loss": 0.0295, "step": 14770 }, { "epoch": 5.427836944546456, "grad_norm": 0.3006584346294403, "learning_rate": 8.439813381611441e-05, "loss": 0.028, "step": 14780 }, { "epoch": 5.431509364671318, "grad_norm": 0.26064079999923706, "learning_rate": 8.437412631001567e-05, "loss": 0.0273, "step": 14790 }, { "epoch": 5.435181784796181, "grad_norm": 0.2606194317340851, "learning_rate": 8.435010376750626e-05, "loss": 0.0265, "step": 14800 }, { "epoch": 5.438854204921043, "grad_norm": 0.31250524520874023, "learning_rate": 8.432606619909442e-05, "loss": 0.0268, "step": 14810 }, { "epoch": 5.442526625045906, "grad_norm": 0.5294637680053711, "learning_rate": 8.430201361529506e-05, "loss": 0.0273, "step": 14820 }, { "epoch": 5.446199045170768, "grad_norm": 0.3313214182853699, "learning_rate": 8.427794602662954e-05, "loss": 0.0287, "step": 14830 }, { "epoch": 5.4498714652956295, "grad_norm": 0.18907128274440765, "learning_rate": 8.425386344362586e-05, "loss": 0.0283, "step": 14840 }, { "epoch": 5.453543885420492, "grad_norm": 0.20969264209270477, "learning_rate": 8.422976587681859e-05, "loss": 0.0257, "step": 14850 }, { "epoch": 5.457216305545354, "grad_norm": 0.29356250166893005, "learning_rate": 8.42056533367488e-05, "loss": 0.0268, "step": 14860 }, { "epoch": 5.460888725670217, "grad_norm": 0.5306440591812134, "learning_rate": 8.41815258339641e-05, "loss": 0.0292, "step": 14870 }, { "epoch": 5.464561145795079, "grad_norm": 0.3136846125125885, "learning_rate": 8.415738337901874e-05, "loss": 0.0268, "step": 14880 }, { "epoch": 5.468233565919942, "grad_norm": 0.38210126757621765, "learning_rate": 8.413322598247342e-05, "loss": 0.029, "step": 14890 }, { "epoch": 5.471905986044804, "grad_norm": 0.24425244331359863, "learning_rate": 8.41090536548954e-05, "loss": 0.0265, "step": 14900 }, { "epoch": 5.4755784061696655, "grad_norm": 0.23529069125652313, "learning_rate": 8.408486640685849e-05, "loss": 0.0274, "step": 14910 }, { "epoch": 5.479250826294528, "grad_norm": 0.29114192724227905, "learning_rate": 8.4060664248943e-05, "loss": 0.03, "step": 14920 }, { "epoch": 5.48292324641939, "grad_norm": 0.24060024321079254, "learning_rate": 8.40364471917358e-05, "loss": 0.0274, "step": 14930 }, { "epoch": 5.486595666544253, "grad_norm": 0.23532554507255554, "learning_rate": 8.401221524583024e-05, "loss": 0.0284, "step": 14940 }, { "epoch": 5.490268086669115, "grad_norm": 0.49932795763015747, "learning_rate": 8.398796842182619e-05, "loss": 0.0274, "step": 14950 }, { "epoch": 5.493940506793977, "grad_norm": 0.2219206690788269, "learning_rate": 8.396370673033006e-05, "loss": 0.0266, "step": 14960 }, { "epoch": 5.49761292691884, "grad_norm": 0.2558097243309021, "learning_rate": 8.39394301819547e-05, "loss": 0.0262, "step": 14970 }, { "epoch": 5.5012853470437015, "grad_norm": 0.30366185307502747, "learning_rate": 8.391513878731949e-05, "loss": 0.0268, "step": 14980 }, { "epoch": 5.504957767168564, "grad_norm": 0.7028218507766724, "learning_rate": 8.389083255705037e-05, "loss": 0.0299, "step": 14990 }, { "epoch": 5.508630187293426, "grad_norm": 0.19970464706420898, "learning_rate": 8.386651150177968e-05, "loss": 0.0268, "step": 15000 }, { "epoch": 5.512302607418289, "grad_norm": 0.3568069636821747, "learning_rate": 8.384217563214627e-05, "loss": 0.0264, "step": 15010 }, { "epoch": 5.515975027543151, "grad_norm": 0.20001636445522308, "learning_rate": 8.381782495879546e-05, "loss": 0.0255, "step": 15020 }, { "epoch": 5.519647447668013, "grad_norm": 0.2319098860025406, "learning_rate": 8.37934594923791e-05, "loss": 0.0297, "step": 15030 }, { "epoch": 5.523319867792876, "grad_norm": 0.3085505962371826, "learning_rate": 8.376907924355546e-05, "loss": 0.0286, "step": 15040 }, { "epoch": 5.526992287917738, "grad_norm": 0.3115941882133484, "learning_rate": 8.374468422298926e-05, "loss": 0.0265, "step": 15050 }, { "epoch": 5.5306647080426, "grad_norm": 0.3271495997905731, "learning_rate": 8.372027444135176e-05, "loss": 0.0265, "step": 15060 }, { "epoch": 5.534337128167462, "grad_norm": 0.2846648097038269, "learning_rate": 8.36958499093206e-05, "loss": 0.0285, "step": 15070 }, { "epoch": 5.538009548292324, "grad_norm": 0.25691351294517517, "learning_rate": 8.367141063757988e-05, "loss": 0.0274, "step": 15080 }, { "epoch": 5.541681968417187, "grad_norm": 0.2832016050815582, "learning_rate": 8.364695663682022e-05, "loss": 0.0256, "step": 15090 }, { "epoch": 5.545354388542049, "grad_norm": 0.24204683303833008, "learning_rate": 8.362248791773858e-05, "loss": 0.0266, "step": 15100 }, { "epoch": 5.549026808666912, "grad_norm": 0.18906079232692719, "learning_rate": 8.359800449103842e-05, "loss": 0.0302, "step": 15110 }, { "epoch": 5.552699228791774, "grad_norm": 0.36823171377182007, "learning_rate": 8.357350636742967e-05, "loss": 0.0309, "step": 15120 }, { "epoch": 5.556371648916636, "grad_norm": 0.36088818311691284, "learning_rate": 8.354899355762858e-05, "loss": 0.0278, "step": 15130 }, { "epoch": 5.560044069041498, "grad_norm": 0.2381884902715683, "learning_rate": 8.352446607235791e-05, "loss": 0.0262, "step": 15140 }, { "epoch": 5.56371648916636, "grad_norm": 0.2779700756072998, "learning_rate": 8.349992392234683e-05, "loss": 0.0283, "step": 15150 }, { "epoch": 5.567388909291223, "grad_norm": 0.3365420401096344, "learning_rate": 8.347536711833088e-05, "loss": 0.0271, "step": 15160 }, { "epoch": 5.571061329416085, "grad_norm": 0.51047283411026, "learning_rate": 8.345079567105206e-05, "loss": 0.0281, "step": 15170 }, { "epoch": 5.574733749540948, "grad_norm": 0.39344915747642517, "learning_rate": 8.342620959125875e-05, "loss": 0.0246, "step": 15180 }, { "epoch": 5.57840616966581, "grad_norm": 0.29328587651252747, "learning_rate": 8.340160888970574e-05, "loss": 0.0285, "step": 15190 }, { "epoch": 5.582078589790672, "grad_norm": 0.23527337610721588, "learning_rate": 8.337699357715422e-05, "loss": 0.0281, "step": 15200 }, { "epoch": 5.585751009915534, "grad_norm": 0.32805800437927246, "learning_rate": 8.335236366437174e-05, "loss": 0.0261, "step": 15210 }, { "epoch": 5.589423430040396, "grad_norm": 0.3249019682407379, "learning_rate": 8.332771916213227e-05, "loss": 0.0279, "step": 15220 }, { "epoch": 5.593095850165259, "grad_norm": 0.261152982711792, "learning_rate": 8.330306008121618e-05, "loss": 0.0263, "step": 15230 }, { "epoch": 5.596768270290121, "grad_norm": 0.33417201042175293, "learning_rate": 8.327838643241014e-05, "loss": 0.0285, "step": 15240 }, { "epoch": 5.600440690414984, "grad_norm": 0.24773003160953522, "learning_rate": 8.32536982265073e-05, "loss": 0.0285, "step": 15250 }, { "epoch": 5.604113110539846, "grad_norm": 0.21377398073673248, "learning_rate": 8.32289954743071e-05, "loss": 0.0292, "step": 15260 }, { "epoch": 5.607785530664708, "grad_norm": 0.22758187353610992, "learning_rate": 8.320427818661532e-05, "loss": 0.0277, "step": 15270 }, { "epoch": 5.61145795078957, "grad_norm": 0.2253362387418747, "learning_rate": 8.317954637424419e-05, "loss": 0.0252, "step": 15280 }, { "epoch": 5.615130370914432, "grad_norm": 0.3164371848106384, "learning_rate": 8.315480004801221e-05, "loss": 0.0291, "step": 15290 }, { "epoch": 5.618802791039295, "grad_norm": 0.27193430066108704, "learning_rate": 8.31300392187443e-05, "loss": 0.0265, "step": 15300 }, { "epoch": 5.622475211164157, "grad_norm": 0.3251718580722809, "learning_rate": 8.31052638972717e-05, "loss": 0.0278, "step": 15310 }, { "epoch": 5.62614763128902, "grad_norm": 0.23559047281742096, "learning_rate": 8.30804740944319e-05, "loss": 0.0305, "step": 15320 }, { "epoch": 5.629820051413882, "grad_norm": 0.3535640835762024, "learning_rate": 8.305566982106887e-05, "loss": 0.028, "step": 15330 }, { "epoch": 5.633492471538744, "grad_norm": 0.3575272262096405, "learning_rate": 8.303085108803282e-05, "loss": 0.0275, "step": 15340 }, { "epoch": 5.637164891663606, "grad_norm": 0.17910179495811462, "learning_rate": 8.300601790618032e-05, "loss": 0.026, "step": 15350 }, { "epoch": 5.640837311788468, "grad_norm": 0.19077420234680176, "learning_rate": 8.298117028637423e-05, "loss": 0.0272, "step": 15360 }, { "epoch": 5.644509731913331, "grad_norm": 0.2765766382217407, "learning_rate": 8.295630823948375e-05, "loss": 0.0253, "step": 15370 }, { "epoch": 5.648182152038193, "grad_norm": 0.2338743805885315, "learning_rate": 8.293143177638438e-05, "loss": 0.0277, "step": 15380 }, { "epoch": 5.651854572163056, "grad_norm": 0.3438621759414673, "learning_rate": 8.290654090795795e-05, "loss": 0.0306, "step": 15390 }, { "epoch": 5.655526992287918, "grad_norm": 0.31008654832839966, "learning_rate": 8.288163564509253e-05, "loss": 0.0292, "step": 15400 }, { "epoch": 5.6591994124127805, "grad_norm": 0.2687542736530304, "learning_rate": 8.285671599868257e-05, "loss": 0.0269, "step": 15410 }, { "epoch": 5.662871832537642, "grad_norm": 0.2393391877412796, "learning_rate": 8.283178197962875e-05, "loss": 0.0294, "step": 15420 }, { "epoch": 5.666544252662504, "grad_norm": 0.2466728538274765, "learning_rate": 8.280683359883808e-05, "loss": 0.0263, "step": 15430 }, { "epoch": 5.670216672787367, "grad_norm": 0.22812581062316895, "learning_rate": 8.278187086722377e-05, "loss": 0.0268, "step": 15440 }, { "epoch": 5.673889092912229, "grad_norm": 0.31420400738716125, "learning_rate": 8.275689379570544e-05, "loss": 0.0277, "step": 15450 }, { "epoch": 5.677561513037092, "grad_norm": 0.3672104775905609, "learning_rate": 8.273190239520887e-05, "loss": 0.026, "step": 15460 }, { "epoch": 5.681233933161954, "grad_norm": 0.2599397301673889, "learning_rate": 8.270689667666616e-05, "loss": 0.0264, "step": 15470 }, { "epoch": 5.6849063532868165, "grad_norm": 0.35963746905326843, "learning_rate": 8.268187665101563e-05, "loss": 0.0254, "step": 15480 }, { "epoch": 5.688578773411678, "grad_norm": 0.2287723869085312, "learning_rate": 8.265684232920194e-05, "loss": 0.0277, "step": 15490 }, { "epoch": 5.69225119353654, "grad_norm": 0.20844940841197968, "learning_rate": 8.263179372217592e-05, "loss": 0.0257, "step": 15500 }, { "epoch": 5.695923613661403, "grad_norm": 0.32612791657447815, "learning_rate": 8.260673084089468e-05, "loss": 0.0259, "step": 15510 }, { "epoch": 5.699596033786265, "grad_norm": 0.47398146986961365, "learning_rate": 8.25816536963216e-05, "loss": 0.0261, "step": 15520 }, { "epoch": 5.703268453911128, "grad_norm": 0.3000003695487976, "learning_rate": 8.255656229942626e-05, "loss": 0.0295, "step": 15530 }, { "epoch": 5.70694087403599, "grad_norm": 0.23204860091209412, "learning_rate": 8.253145666118447e-05, "loss": 0.0289, "step": 15540 }, { "epoch": 5.710613294160852, "grad_norm": 0.23990896344184875, "learning_rate": 8.250633679257834e-05, "loss": 0.0278, "step": 15550 }, { "epoch": 5.714285714285714, "grad_norm": 0.23072388768196106, "learning_rate": 8.24812027045961e-05, "loss": 0.0268, "step": 15560 }, { "epoch": 5.717958134410576, "grad_norm": 0.3232477307319641, "learning_rate": 8.245605440823226e-05, "loss": 0.0257, "step": 15570 }, { "epoch": 5.721630554535439, "grad_norm": 0.23074749112129211, "learning_rate": 8.243089191448759e-05, "loss": 0.0256, "step": 15580 }, { "epoch": 5.725302974660301, "grad_norm": 0.2482587844133377, "learning_rate": 8.240571523436896e-05, "loss": 0.0269, "step": 15590 }, { "epoch": 5.728975394785163, "grad_norm": 0.2419169694185257, "learning_rate": 8.238052437888955e-05, "loss": 0.0264, "step": 15600 }, { "epoch": 5.732647814910026, "grad_norm": 0.27324992418289185, "learning_rate": 8.235531935906866e-05, "loss": 0.0271, "step": 15610 }, { "epoch": 5.736320235034888, "grad_norm": 0.24288307130336761, "learning_rate": 8.233010018593185e-05, "loss": 0.0263, "step": 15620 }, { "epoch": 5.73999265515975, "grad_norm": 0.33449679613113403, "learning_rate": 8.230486687051082e-05, "loss": 0.0244, "step": 15630 }, { "epoch": 5.743665075284612, "grad_norm": 0.17074866592884064, "learning_rate": 8.22796194238435e-05, "loss": 0.0252, "step": 15640 }, { "epoch": 5.747337495409475, "grad_norm": 0.1802920401096344, "learning_rate": 8.225435785697397e-05, "loss": 0.0285, "step": 15650 }, { "epoch": 5.751009915534337, "grad_norm": 0.3665565848350525, "learning_rate": 8.222908218095252e-05, "loss": 0.0295, "step": 15660 }, { "epoch": 5.754682335659199, "grad_norm": 0.2837558388710022, "learning_rate": 8.220379240683553e-05, "loss": 0.0264, "step": 15670 }, { "epoch": 5.758354755784062, "grad_norm": 0.39863279461860657, "learning_rate": 8.217848854568566e-05, "loss": 0.0287, "step": 15680 }, { "epoch": 5.762027175908924, "grad_norm": 0.22928264737129211, "learning_rate": 8.215317060857167e-05, "loss": 0.0292, "step": 15690 }, { "epoch": 5.7656995960337865, "grad_norm": 0.3062025308609009, "learning_rate": 8.212783860656847e-05, "loss": 0.0245, "step": 15700 }, { "epoch": 5.769372016158648, "grad_norm": 0.2448819875717163, "learning_rate": 8.210249255075717e-05, "loss": 0.0254, "step": 15710 }, { "epoch": 5.773044436283511, "grad_norm": 0.19364775717258453, "learning_rate": 8.207713245222496e-05, "loss": 0.0307, "step": 15720 }, { "epoch": 5.776716856408373, "grad_norm": 0.35228782892227173, "learning_rate": 8.205175832206522e-05, "loss": 0.0267, "step": 15730 }, { "epoch": 5.780389276533235, "grad_norm": 0.3398289680480957, "learning_rate": 8.202637017137747e-05, "loss": 0.028, "step": 15740 }, { "epoch": 5.784061696658098, "grad_norm": 0.34570029377937317, "learning_rate": 8.200096801126734e-05, "loss": 0.0262, "step": 15750 }, { "epoch": 5.78773411678296, "grad_norm": 0.3339861035346985, "learning_rate": 8.197555185284658e-05, "loss": 0.0272, "step": 15760 }, { "epoch": 5.7914065369078225, "grad_norm": 0.28348180651664734, "learning_rate": 8.19501217072331e-05, "loss": 0.0283, "step": 15770 }, { "epoch": 5.795078957032684, "grad_norm": 0.37971943616867065, "learning_rate": 8.192467758555091e-05, "loss": 0.0284, "step": 15780 }, { "epoch": 5.798751377157547, "grad_norm": 0.3097408711910248, "learning_rate": 8.189921949893014e-05, "loss": 0.0273, "step": 15790 }, { "epoch": 5.802423797282409, "grad_norm": 0.3693545460700989, "learning_rate": 8.187374745850702e-05, "loss": 0.0268, "step": 15800 }, { "epoch": 5.806096217407271, "grad_norm": 0.23655907809734344, "learning_rate": 8.184826147542387e-05, "loss": 0.0274, "step": 15810 }, { "epoch": 5.809768637532134, "grad_norm": 0.23658664524555206, "learning_rate": 8.18227615608291e-05, "loss": 0.0276, "step": 15820 }, { "epoch": 5.813441057656996, "grad_norm": 0.6487385034561157, "learning_rate": 8.179724772587732e-05, "loss": 0.0263, "step": 15830 }, { "epoch": 5.8171134777818585, "grad_norm": 0.17754749953746796, "learning_rate": 8.177171998172908e-05, "loss": 0.025, "step": 15840 }, { "epoch": 5.82078589790672, "grad_norm": 0.3911205530166626, "learning_rate": 8.174617833955112e-05, "loss": 0.0303, "step": 15850 }, { "epoch": 5.824458318031583, "grad_norm": 0.226066455245018, "learning_rate": 8.172062281051618e-05, "loss": 0.0262, "step": 15860 }, { "epoch": 5.828130738156445, "grad_norm": 0.30284616351127625, "learning_rate": 8.169505340580316e-05, "loss": 0.0269, "step": 15870 }, { "epoch": 5.831803158281307, "grad_norm": 0.4542354941368103, "learning_rate": 8.166947013659696e-05, "loss": 0.0301, "step": 15880 }, { "epoch": 5.83547557840617, "grad_norm": 0.3379444479942322, "learning_rate": 8.164387301408858e-05, "loss": 0.0274, "step": 15890 }, { "epoch": 5.839147998531032, "grad_norm": 0.34488385915756226, "learning_rate": 8.161826204947507e-05, "loss": 0.0279, "step": 15900 }, { "epoch": 5.8428204186558945, "grad_norm": 0.210555300116539, "learning_rate": 8.159263725395954e-05, "loss": 0.0268, "step": 15910 }, { "epoch": 5.846492838780756, "grad_norm": 0.2945752441883087, "learning_rate": 8.156699863875115e-05, "loss": 0.0258, "step": 15920 }, { "epoch": 5.850165258905619, "grad_norm": 0.2617398798465729, "learning_rate": 8.154134621506508e-05, "loss": 0.0315, "step": 15930 }, { "epoch": 5.853837679030481, "grad_norm": 0.24061667919158936, "learning_rate": 8.15156799941226e-05, "loss": 0.0258, "step": 15940 }, { "epoch": 5.857510099155343, "grad_norm": 0.27907148003578186, "learning_rate": 8.148999998715096e-05, "loss": 0.0265, "step": 15950 }, { "epoch": 5.861182519280206, "grad_norm": 0.2478230744600296, "learning_rate": 8.146430620538348e-05, "loss": 0.0296, "step": 15960 }, { "epoch": 5.864854939405068, "grad_norm": 0.24928437173366547, "learning_rate": 8.143859866005949e-05, "loss": 0.0263, "step": 15970 }, { "epoch": 5.8685273595299305, "grad_norm": 0.4266970157623291, "learning_rate": 8.141287736242438e-05, "loss": 0.0256, "step": 15980 }, { "epoch": 5.8721997796547925, "grad_norm": 0.5335562825202942, "learning_rate": 8.138714232372945e-05, "loss": 0.0256, "step": 15990 }, { "epoch": 5.875872199779655, "grad_norm": 0.28212305903434753, "learning_rate": 8.136139355523213e-05, "loss": 0.0257, "step": 16000 }, { "epoch": 5.879544619904517, "grad_norm": 0.2802443206310272, "learning_rate": 8.13356310681958e-05, "loss": 0.0244, "step": 16010 }, { "epoch": 5.883217040029379, "grad_norm": 0.24303601682186127, "learning_rate": 8.130985487388982e-05, "loss": 0.0241, "step": 16020 }, { "epoch": 5.886889460154242, "grad_norm": 0.2848217785358429, "learning_rate": 8.12840649835896e-05, "loss": 0.0237, "step": 16030 }, { "epoch": 5.890561880279104, "grad_norm": 0.3111328184604645, "learning_rate": 8.125826140857653e-05, "loss": 0.0269, "step": 16040 }, { "epoch": 5.894234300403967, "grad_norm": 0.24348457157611847, "learning_rate": 8.123244416013792e-05, "loss": 0.0267, "step": 16050 }, { "epoch": 5.8979067205288285, "grad_norm": 0.32522520422935486, "learning_rate": 8.120661324956716e-05, "loss": 0.0255, "step": 16060 }, { "epoch": 5.901579140653691, "grad_norm": 0.7859708070755005, "learning_rate": 8.118076868816353e-05, "loss": 0.0272, "step": 16070 }, { "epoch": 5.905251560778553, "grad_norm": 0.2856409549713135, "learning_rate": 8.115491048723234e-05, "loss": 0.0275, "step": 16080 }, { "epoch": 5.908923980903415, "grad_norm": 0.31613701581954956, "learning_rate": 8.112903865808486e-05, "loss": 0.0268, "step": 16090 }, { "epoch": 5.912596401028278, "grad_norm": 0.21131475269794464, "learning_rate": 8.110315321203826e-05, "loss": 0.027, "step": 16100 }, { "epoch": 5.91626882115314, "grad_norm": 0.2575575113296509, "learning_rate": 8.107725416041575e-05, "loss": 0.0255, "step": 16110 }, { "epoch": 5.919941241278003, "grad_norm": 0.23423214256763458, "learning_rate": 8.105134151454645e-05, "loss": 0.0261, "step": 16120 }, { "epoch": 5.9236136614028645, "grad_norm": 0.1917841136455536, "learning_rate": 8.102541528576544e-05, "loss": 0.0255, "step": 16130 }, { "epoch": 5.927286081527726, "grad_norm": 0.3423001170158386, "learning_rate": 8.099947548541371e-05, "loss": 0.0266, "step": 16140 }, { "epoch": 5.930958501652589, "grad_norm": 0.2979225814342499, "learning_rate": 8.097352212483824e-05, "loss": 0.0279, "step": 16150 }, { "epoch": 5.934630921777451, "grad_norm": 0.40145862102508545, "learning_rate": 8.094755521539187e-05, "loss": 0.029, "step": 16160 }, { "epoch": 5.938303341902314, "grad_norm": 0.3083479404449463, "learning_rate": 8.092157476843346e-05, "loss": 0.0251, "step": 16170 }, { "epoch": 5.941975762027176, "grad_norm": 0.2704997956752777, "learning_rate": 8.089558079532769e-05, "loss": 0.0272, "step": 16180 }, { "epoch": 5.945648182152038, "grad_norm": 0.4113396406173706, "learning_rate": 8.086957330744522e-05, "loss": 0.0262, "step": 16190 }, { "epoch": 5.9493206022769005, "grad_norm": 0.3293769955635071, "learning_rate": 8.084355231616265e-05, "loss": 0.0277, "step": 16200 }, { "epoch": 5.952993022401762, "grad_norm": 0.4017903804779053, "learning_rate": 8.08175178328624e-05, "loss": 0.0258, "step": 16210 }, { "epoch": 5.956665442526625, "grad_norm": 0.2710065543651581, "learning_rate": 8.079146986893284e-05, "loss": 0.0277, "step": 16220 }, { "epoch": 5.960337862651487, "grad_norm": 0.25977689027786255, "learning_rate": 8.076540843576826e-05, "loss": 0.0264, "step": 16230 }, { "epoch": 5.96401028277635, "grad_norm": 0.5660577416419983, "learning_rate": 8.073933354476879e-05, "loss": 0.028, "step": 16240 }, { "epoch": 5.967682702901212, "grad_norm": 0.28542956709861755, "learning_rate": 8.07132452073405e-05, "loss": 0.0263, "step": 16250 }, { "epoch": 5.971355123026074, "grad_norm": 0.25229206681251526, "learning_rate": 8.068714343489529e-05, "loss": 0.0291, "step": 16260 }, { "epoch": 5.9750275431509365, "grad_norm": 0.26937511563301086, "learning_rate": 8.066102823885097e-05, "loss": 0.0256, "step": 16270 }, { "epoch": 5.9786999632757984, "grad_norm": 0.30744028091430664, "learning_rate": 8.063489963063125e-05, "loss": 0.0254, "step": 16280 }, { "epoch": 5.982372383400661, "grad_norm": 0.2005740851163864, "learning_rate": 8.06087576216656e-05, "loss": 0.0275, "step": 16290 }, { "epoch": 5.986044803525523, "grad_norm": 0.26591527462005615, "learning_rate": 8.058260222338946e-05, "loss": 0.0257, "step": 16300 }, { "epoch": 5.989717223650386, "grad_norm": 0.2714082896709442, "learning_rate": 8.05564334472441e-05, "loss": 0.0271, "step": 16310 }, { "epoch": 5.993389643775248, "grad_norm": 0.2964874804019928, "learning_rate": 8.053025130467664e-05, "loss": 0.0259, "step": 16320 }, { "epoch": 5.99706206390011, "grad_norm": 0.2155727595090866, "learning_rate": 8.050405580714e-05, "loss": 0.0267, "step": 16330 }, { "epoch": 6.000734484024973, "grad_norm": 0.17627450823783875, "learning_rate": 8.047784696609301e-05, "loss": 0.0261, "step": 16340 }, { "epoch": 6.0044069041498345, "grad_norm": 0.3035691976547241, "learning_rate": 8.045162479300029e-05, "loss": 0.0261, "step": 16350 }, { "epoch": 6.008079324274697, "grad_norm": 0.3048657178878784, "learning_rate": 8.042538929933232e-05, "loss": 0.0278, "step": 16360 }, { "epoch": 6.011751744399559, "grad_norm": 0.2936152219772339, "learning_rate": 8.039914049656539e-05, "loss": 0.0266, "step": 16370 }, { "epoch": 6.015424164524422, "grad_norm": 0.19345930218696594, "learning_rate": 8.037287839618163e-05, "loss": 0.0286, "step": 16380 }, { "epoch": 6.019096584649284, "grad_norm": 0.47823426127433777, "learning_rate": 8.034660300966898e-05, "loss": 0.0278, "step": 16390 }, { "epoch": 6.022769004774146, "grad_norm": 0.7171221971511841, "learning_rate": 8.032031434852115e-05, "loss": 0.0266, "step": 16400 }, { "epoch": 6.026441424899009, "grad_norm": 0.3027423322200775, "learning_rate": 8.029401242423773e-05, "loss": 0.0269, "step": 16410 }, { "epoch": 6.0301138450238705, "grad_norm": 0.24017687141895294, "learning_rate": 8.026769724832406e-05, "loss": 0.0256, "step": 16420 }, { "epoch": 6.033786265148733, "grad_norm": 0.30915915966033936, "learning_rate": 8.02413688322913e-05, "loss": 0.025, "step": 16430 }, { "epoch": 6.037458685273595, "grad_norm": 0.5190786123275757, "learning_rate": 8.021502718765639e-05, "loss": 0.0276, "step": 16440 }, { "epoch": 6.041131105398458, "grad_norm": 0.2239983081817627, "learning_rate": 8.018867232594204e-05, "loss": 0.0258, "step": 16450 }, { "epoch": 6.04480352552332, "grad_norm": 0.26356491446495056, "learning_rate": 8.016230425867679e-05, "loss": 0.0269, "step": 16460 }, { "epoch": 6.048475945648182, "grad_norm": 0.24632760882377625, "learning_rate": 8.013592299739494e-05, "loss": 0.0256, "step": 16470 }, { "epoch": 6.052148365773045, "grad_norm": 0.2521224021911621, "learning_rate": 8.01095285536365e-05, "loss": 0.0265, "step": 16480 }, { "epoch": 6.0558207858979065, "grad_norm": 0.5043326616287231, "learning_rate": 8.008312093894735e-05, "loss": 0.0297, "step": 16490 }, { "epoch": 6.059493206022769, "grad_norm": 0.2777562439441681, "learning_rate": 8.005670016487903e-05, "loss": 0.0264, "step": 16500 }, { "epoch": 6.063165626147631, "grad_norm": 0.3385476768016815, "learning_rate": 8.003026624298892e-05, "loss": 0.0256, "step": 16510 }, { "epoch": 6.066838046272494, "grad_norm": 0.2560729384422302, "learning_rate": 8.00038191848401e-05, "loss": 0.0256, "step": 16520 }, { "epoch": 6.070510466397356, "grad_norm": 0.2971794605255127, "learning_rate": 7.997735900200142e-05, "loss": 0.0251, "step": 16530 }, { "epoch": 6.074182886522218, "grad_norm": 0.28747957944869995, "learning_rate": 7.995088570604745e-05, "loss": 0.0255, "step": 16540 }, { "epoch": 6.077855306647081, "grad_norm": 0.28661152720451355, "learning_rate": 7.992439930855852e-05, "loss": 0.0266, "step": 16550 }, { "epoch": 6.0815277267719425, "grad_norm": 0.4556141793727875, "learning_rate": 7.989789982112067e-05, "loss": 0.0258, "step": 16560 }, { "epoch": 6.085200146896805, "grad_norm": 0.39197084307670593, "learning_rate": 7.987138725532572e-05, "loss": 0.0259, "step": 16570 }, { "epoch": 6.088872567021667, "grad_norm": 0.5955106616020203, "learning_rate": 7.98448616227711e-05, "loss": 0.0278, "step": 16580 }, { "epoch": 6.092544987146529, "grad_norm": 0.23845720291137695, "learning_rate": 7.981832293506004e-05, "loss": 0.0265, "step": 16590 }, { "epoch": 6.096217407271392, "grad_norm": 0.2184152603149414, "learning_rate": 7.97917712038015e-05, "loss": 0.0283, "step": 16600 }, { "epoch": 6.099889827396254, "grad_norm": 0.24574178457260132, "learning_rate": 7.976520644061007e-05, "loss": 0.0263, "step": 16610 }, { "epoch": 6.103562247521117, "grad_norm": 0.18981799483299255, "learning_rate": 7.973862865710611e-05, "loss": 0.0257, "step": 16620 }, { "epoch": 6.1072346676459786, "grad_norm": 0.23752327263355255, "learning_rate": 7.971203786491563e-05, "loss": 0.0259, "step": 16630 }, { "epoch": 6.110907087770841, "grad_norm": 0.24541959166526794, "learning_rate": 7.968543407567034e-05, "loss": 0.0238, "step": 16640 }, { "epoch": 6.114579507895703, "grad_norm": 0.27735692262649536, "learning_rate": 7.965881730100765e-05, "loss": 0.0266, "step": 16650 }, { "epoch": 6.118251928020565, "grad_norm": 0.2705981433391571, "learning_rate": 7.963218755257064e-05, "loss": 0.0266, "step": 16660 }, { "epoch": 6.121924348145428, "grad_norm": 0.28529787063598633, "learning_rate": 7.960554484200807e-05, "loss": 0.0261, "step": 16670 }, { "epoch": 6.12559676827029, "grad_norm": 0.37840908765792847, "learning_rate": 7.957888918097436e-05, "loss": 0.0246, "step": 16680 }, { "epoch": 6.129269188395153, "grad_norm": 0.3042815923690796, "learning_rate": 7.955222058112962e-05, "loss": 0.0261, "step": 16690 }, { "epoch": 6.132941608520015, "grad_norm": 0.2811054289340973, "learning_rate": 7.952553905413955e-05, "loss": 0.0289, "step": 16700 }, { "epoch": 6.136614028644877, "grad_norm": 0.20674030482769012, "learning_rate": 7.949884461167563e-05, "loss": 0.0229, "step": 16710 }, { "epoch": 6.140286448769739, "grad_norm": 0.22342047095298767, "learning_rate": 7.947213726541484e-05, "loss": 0.0292, "step": 16720 }, { "epoch": 6.143958868894601, "grad_norm": 0.2844748795032501, "learning_rate": 7.944541702703996e-05, "loss": 0.0255, "step": 16730 }, { "epoch": 6.147631289019464, "grad_norm": 0.3257623612880707, "learning_rate": 7.941868390823928e-05, "loss": 0.0248, "step": 16740 }, { "epoch": 6.151303709144326, "grad_norm": 0.21609681844711304, "learning_rate": 7.939193792070677e-05, "loss": 0.0248, "step": 16750 }, { "epoch": 6.154976129269189, "grad_norm": 0.22773194313049316, "learning_rate": 7.936517907614207e-05, "loss": 0.026, "step": 16760 }, { "epoch": 6.158648549394051, "grad_norm": 0.23891940712928772, "learning_rate": 7.933840738625036e-05, "loss": 0.0272, "step": 16770 }, { "epoch": 6.1623209695189125, "grad_norm": 0.23902584612369537, "learning_rate": 7.931162286274252e-05, "loss": 0.024, "step": 16780 }, { "epoch": 6.165993389643775, "grad_norm": 0.2842901945114136, "learning_rate": 7.928482551733503e-05, "loss": 0.0256, "step": 16790 }, { "epoch": 6.169665809768637, "grad_norm": 0.20796018838882446, "learning_rate": 7.925801536174991e-05, "loss": 0.0263, "step": 16800 }, { "epoch": 6.1733382298935, "grad_norm": 0.3443889617919922, "learning_rate": 7.923119240771484e-05, "loss": 0.0276, "step": 16810 }, { "epoch": 6.177010650018362, "grad_norm": 0.2537263333797455, "learning_rate": 7.920435666696313e-05, "loss": 0.0257, "step": 16820 }, { "epoch": 6.180683070143225, "grad_norm": 0.18848197162151337, "learning_rate": 7.917750815123362e-05, "loss": 0.0258, "step": 16830 }, { "epoch": 6.184355490268087, "grad_norm": 0.2884879410266876, "learning_rate": 7.915064687227075e-05, "loss": 0.0259, "step": 16840 }, { "epoch": 6.1880279103929485, "grad_norm": 0.24009858071804047, "learning_rate": 7.912377284182458e-05, "loss": 0.0239, "step": 16850 }, { "epoch": 6.191700330517811, "grad_norm": 0.24533338844776154, "learning_rate": 7.90968860716507e-05, "loss": 0.0254, "step": 16860 }, { "epoch": 6.195372750642673, "grad_norm": 0.16708090901374817, "learning_rate": 7.906998657351033e-05, "loss": 0.0239, "step": 16870 }, { "epoch": 6.199045170767536, "grad_norm": 0.23856595158576965, "learning_rate": 7.90430743591702e-05, "loss": 0.0276, "step": 16880 }, { "epoch": 6.202717590892398, "grad_norm": 0.21859735250473022, "learning_rate": 7.901614944040263e-05, "loss": 0.0282, "step": 16890 }, { "epoch": 6.206390011017261, "grad_norm": 0.23779389262199402, "learning_rate": 7.89892118289855e-05, "loss": 0.0286, "step": 16900 }, { "epoch": 6.210062431142123, "grad_norm": 0.24191193282604218, "learning_rate": 7.896226153670223e-05, "loss": 0.0263, "step": 16910 }, { "epoch": 6.2137348512669845, "grad_norm": 0.2584352195262909, "learning_rate": 7.89352985753418e-05, "loss": 0.0248, "step": 16920 }, { "epoch": 6.217407271391847, "grad_norm": 0.3180558979511261, "learning_rate": 7.890832295669871e-05, "loss": 0.0256, "step": 16930 }, { "epoch": 6.221079691516709, "grad_norm": 0.2964029312133789, "learning_rate": 7.888133469257303e-05, "loss": 0.0254, "step": 16940 }, { "epoch": 6.224752111641572, "grad_norm": 0.2578714191913605, "learning_rate": 7.885433379477034e-05, "loss": 0.0251, "step": 16950 }, { "epoch": 6.228424531766434, "grad_norm": 0.20122241973876953, "learning_rate": 7.882732027510174e-05, "loss": 0.0248, "step": 16960 }, { "epoch": 6.232096951891297, "grad_norm": 0.19247481226921082, "learning_rate": 7.880029414538386e-05, "loss": 0.0278, "step": 16970 }, { "epoch": 6.235769372016159, "grad_norm": 0.3269324004650116, "learning_rate": 7.877325541743888e-05, "loss": 0.0248, "step": 16980 }, { "epoch": 6.239441792141021, "grad_norm": 0.3291674256324768, "learning_rate": 7.874620410309443e-05, "loss": 0.028, "step": 16990 }, { "epoch": 6.243114212265883, "grad_norm": 0.28135740756988525, "learning_rate": 7.871914021418368e-05, "loss": 0.027, "step": 17000 }, { "epoch": 6.246786632390745, "grad_norm": 0.2292659729719162, "learning_rate": 7.869206376254528e-05, "loss": 0.0244, "step": 17010 }, { "epoch": 6.250459052515608, "grad_norm": 0.2261676788330078, "learning_rate": 7.86649747600234e-05, "loss": 0.0264, "step": 17020 }, { "epoch": 6.25413147264047, "grad_norm": 0.28833240270614624, "learning_rate": 7.86378732184677e-05, "loss": 0.0256, "step": 17030 }, { "epoch": 6.257803892765333, "grad_norm": 0.2911396026611328, "learning_rate": 7.86107591497333e-05, "loss": 0.0239, "step": 17040 }, { "epoch": 6.261476312890195, "grad_norm": 0.22285683453083038, "learning_rate": 7.858363256568083e-05, "loss": 0.0262, "step": 17050 }, { "epoch": 6.265148733015057, "grad_norm": 0.2722022831439972, "learning_rate": 7.855649347817637e-05, "loss": 0.0246, "step": 17060 }, { "epoch": 6.268821153139919, "grad_norm": 0.22358299791812897, "learning_rate": 7.85293418990915e-05, "loss": 0.0262, "step": 17070 }, { "epoch": 6.272493573264781, "grad_norm": 0.3516102135181427, "learning_rate": 7.85021778403032e-05, "loss": 0.0258, "step": 17080 }, { "epoch": 6.276165993389644, "grad_norm": 0.2956371009349823, "learning_rate": 7.847500131369397e-05, "loss": 0.0237, "step": 17090 }, { "epoch": 6.279838413514506, "grad_norm": 0.2802915573120117, "learning_rate": 7.844781233115176e-05, "loss": 0.0263, "step": 17100 }, { "epoch": 6.283510833639369, "grad_norm": 0.2721591889858246, "learning_rate": 7.842061090456996e-05, "loss": 0.0271, "step": 17110 }, { "epoch": 6.287183253764231, "grad_norm": 0.2807040214538574, "learning_rate": 7.839339704584736e-05, "loss": 0.0265, "step": 17120 }, { "epoch": 6.290855673889093, "grad_norm": 0.22694487869739532, "learning_rate": 7.836617076688828e-05, "loss": 0.0265, "step": 17130 }, { "epoch": 6.294528094013955, "grad_norm": 0.2746659517288208, "learning_rate": 7.833893207960236e-05, "loss": 0.0272, "step": 17140 }, { "epoch": 6.298200514138817, "grad_norm": 0.17486731708049774, "learning_rate": 7.831168099590478e-05, "loss": 0.0255, "step": 17150 }, { "epoch": 6.30187293426368, "grad_norm": 0.2619394063949585, "learning_rate": 7.828441752771607e-05, "loss": 0.0264, "step": 17160 }, { "epoch": 6.305545354388542, "grad_norm": 0.36318153142929077, "learning_rate": 7.825714168696219e-05, "loss": 0.0262, "step": 17170 }, { "epoch": 6.309217774513404, "grad_norm": 0.2108909785747528, "learning_rate": 7.82298534855745e-05, "loss": 0.0258, "step": 17180 }, { "epoch": 6.312890194638267, "grad_norm": 0.22828520834445953, "learning_rate": 7.820255293548984e-05, "loss": 0.0247, "step": 17190 }, { "epoch": 6.316562614763129, "grad_norm": 0.33185988664627075, "learning_rate": 7.817524004865035e-05, "loss": 0.0254, "step": 17200 }, { "epoch": 6.320235034887991, "grad_norm": 0.33359295129776, "learning_rate": 7.814791483700366e-05, "loss": 0.0266, "step": 17210 }, { "epoch": 6.323907455012853, "grad_norm": 0.3209976553916931, "learning_rate": 7.81205773125027e-05, "loss": 0.0254, "step": 17220 }, { "epoch": 6.327579875137716, "grad_norm": 0.2773836553096771, "learning_rate": 7.809322748710585e-05, "loss": 0.0242, "step": 17230 }, { "epoch": 6.331252295262578, "grad_norm": 0.373184472322464, "learning_rate": 7.806586537277685e-05, "loss": 0.0247, "step": 17240 }, { "epoch": 6.33492471538744, "grad_norm": 0.27269694209098816, "learning_rate": 7.803849098148484e-05, "loss": 0.0272, "step": 17250 }, { "epoch": 6.338597135512303, "grad_norm": 0.21994420886039734, "learning_rate": 7.801110432520427e-05, "loss": 0.0246, "step": 17260 }, { "epoch": 6.342269555637165, "grad_norm": 0.27050894498825073, "learning_rate": 7.798370541591504e-05, "loss": 0.025, "step": 17270 }, { "epoch": 6.3459419757620275, "grad_norm": 0.360270619392395, "learning_rate": 7.795629426560232e-05, "loss": 0.0256, "step": 17280 }, { "epoch": 6.349614395886889, "grad_norm": 0.25018852949142456, "learning_rate": 7.792887088625671e-05, "loss": 0.0251, "step": 17290 }, { "epoch": 6.353286816011752, "grad_norm": 0.3290656805038452, "learning_rate": 7.790143528987412e-05, "loss": 0.0255, "step": 17300 }, { "epoch": 6.356959236136614, "grad_norm": 0.2600035071372986, "learning_rate": 7.787398748845578e-05, "loss": 0.0274, "step": 17310 }, { "epoch": 6.360631656261476, "grad_norm": 0.29802054166793823, "learning_rate": 7.784652749400834e-05, "loss": 0.0239, "step": 17320 }, { "epoch": 6.364304076386339, "grad_norm": 0.23743006587028503, "learning_rate": 7.78190553185437e-05, "loss": 0.0253, "step": 17330 }, { "epoch": 6.367976496511201, "grad_norm": 0.23196040093898773, "learning_rate": 7.779157097407915e-05, "loss": 0.025, "step": 17340 }, { "epoch": 6.3716489166360635, "grad_norm": 0.291549414396286, "learning_rate": 7.776407447263725e-05, "loss": 0.0277, "step": 17350 }, { "epoch": 6.375321336760925, "grad_norm": 0.26166868209838867, "learning_rate": 7.773656582624593e-05, "loss": 0.0267, "step": 17360 }, { "epoch": 6.378993756885787, "grad_norm": 0.30293503403663635, "learning_rate": 7.770904504693837e-05, "loss": 0.0251, "step": 17370 }, { "epoch": 6.38266617701065, "grad_norm": 0.25013166666030884, "learning_rate": 7.768151214675314e-05, "loss": 0.024, "step": 17380 }, { "epoch": 6.386338597135512, "grad_norm": 0.29565730690956116, "learning_rate": 7.765396713773402e-05, "loss": 0.0242, "step": 17390 }, { "epoch": 6.390011017260375, "grad_norm": 0.3785666227340698, "learning_rate": 7.762641003193017e-05, "loss": 0.0262, "step": 17400 }, { "epoch": 6.393683437385237, "grad_norm": 0.5067731142044067, "learning_rate": 7.759884084139597e-05, "loss": 0.024, "step": 17410 }, { "epoch": 6.3973558575100995, "grad_norm": 0.3036097288131714, "learning_rate": 7.757125957819116e-05, "loss": 0.0246, "step": 17420 }, { "epoch": 6.401028277634961, "grad_norm": 0.19774475693702698, "learning_rate": 7.754366625438069e-05, "loss": 0.0284, "step": 17430 }, { "epoch": 6.404700697759823, "grad_norm": 0.22531741857528687, "learning_rate": 7.751606088203483e-05, "loss": 0.0247, "step": 17440 }, { "epoch": 6.408373117884686, "grad_norm": 0.39068350195884705, "learning_rate": 7.74884434732291e-05, "loss": 0.0248, "step": 17450 }, { "epoch": 6.412045538009548, "grad_norm": 0.2911527454853058, "learning_rate": 7.746081404004431e-05, "loss": 0.0267, "step": 17460 }, { "epoch": 6.415717958134411, "grad_norm": 0.2690383791923523, "learning_rate": 7.743317259456649e-05, "loss": 0.0248, "step": 17470 }, { "epoch": 6.419390378259273, "grad_norm": 0.25592443346977234, "learning_rate": 7.740551914888696e-05, "loss": 0.0249, "step": 17480 }, { "epoch": 6.4230627983841355, "grad_norm": 0.2512260377407074, "learning_rate": 7.737785371510229e-05, "loss": 0.025, "step": 17490 }, { "epoch": 6.426735218508997, "grad_norm": 0.37453627586364746, "learning_rate": 7.735017630531426e-05, "loss": 0.0241, "step": 17500 }, { "epoch": 6.430407638633859, "grad_norm": 0.29776811599731445, "learning_rate": 7.732248693162991e-05, "loss": 0.023, "step": 17510 }, { "epoch": 6.434080058758722, "grad_norm": 0.4291761815547943, "learning_rate": 7.729478560616153e-05, "loss": 0.0261, "step": 17520 }, { "epoch": 6.437752478883584, "grad_norm": 0.2821456491947174, "learning_rate": 7.726707234102659e-05, "loss": 0.0256, "step": 17530 }, { "epoch": 6.441424899008447, "grad_norm": 0.22105146944522858, "learning_rate": 7.723934714834785e-05, "loss": 0.027, "step": 17540 }, { "epoch": 6.445097319133309, "grad_norm": 0.258132666349411, "learning_rate": 7.721161004025323e-05, "loss": 0.0259, "step": 17550 }, { "epoch": 6.4487697392581715, "grad_norm": 0.2922881841659546, "learning_rate": 7.718386102887588e-05, "loss": 0.0271, "step": 17560 }, { "epoch": 6.4524421593830334, "grad_norm": 0.23068848252296448, "learning_rate": 7.715610012635418e-05, "loss": 0.0239, "step": 17570 }, { "epoch": 6.456114579507895, "grad_norm": 0.6113669276237488, "learning_rate": 7.712832734483166e-05, "loss": 0.0256, "step": 17580 }, { "epoch": 6.459786999632758, "grad_norm": 0.23495689034461975, "learning_rate": 7.710054269645713e-05, "loss": 0.0234, "step": 17590 }, { "epoch": 6.46345941975762, "grad_norm": 0.22314700484275818, "learning_rate": 7.707274619338449e-05, "loss": 0.0234, "step": 17600 }, { "epoch": 6.467131839882483, "grad_norm": 0.38118523359298706, "learning_rate": 7.704493784777289e-05, "loss": 0.027, "step": 17610 }, { "epoch": 6.470804260007345, "grad_norm": 0.40617066621780396, "learning_rate": 7.701711767178665e-05, "loss": 0.025, "step": 17620 }, { "epoch": 6.474476680132208, "grad_norm": 0.837655782699585, "learning_rate": 7.698928567759525e-05, "loss": 0.0252, "step": 17630 }, { "epoch": 6.4781491002570695, "grad_norm": 0.26843515038490295, "learning_rate": 7.696144187737335e-05, "loss": 0.0244, "step": 17640 }, { "epoch": 6.481821520381931, "grad_norm": 0.31379246711730957, "learning_rate": 7.69335862833008e-05, "loss": 0.0235, "step": 17650 }, { "epoch": 6.485493940506794, "grad_norm": 0.26383331418037415, "learning_rate": 7.690571890756256e-05, "loss": 0.0256, "step": 17660 }, { "epoch": 6.489166360631656, "grad_norm": 0.28988972306251526, "learning_rate": 7.687783976234877e-05, "loss": 0.0273, "step": 17670 }, { "epoch": 6.492838780756519, "grad_norm": 1.267217755317688, "learning_rate": 7.684994885985472e-05, "loss": 0.0256, "step": 17680 }, { "epoch": 6.496511200881381, "grad_norm": 0.3377976417541504, "learning_rate": 7.682204621228083e-05, "loss": 0.0245, "step": 17690 }, { "epoch": 6.500183621006244, "grad_norm": 0.38558200001716614, "learning_rate": 7.679413183183267e-05, "loss": 0.0283, "step": 17700 }, { "epoch": 6.5038560411311055, "grad_norm": 0.21954388916492462, "learning_rate": 7.676620573072093e-05, "loss": 0.0245, "step": 17710 }, { "epoch": 6.507528461255967, "grad_norm": 0.3029584288597107, "learning_rate": 7.673826792116145e-05, "loss": 0.0251, "step": 17720 }, { "epoch": 6.51120088138083, "grad_norm": 0.2441740334033966, "learning_rate": 7.671031841537519e-05, "loss": 0.026, "step": 17730 }, { "epoch": 6.514873301505692, "grad_norm": 0.2475913017988205, "learning_rate": 7.668235722558819e-05, "loss": 0.0238, "step": 17740 }, { "epoch": 6.518545721630555, "grad_norm": 0.2721603214740753, "learning_rate": 7.665438436403161e-05, "loss": 0.0247, "step": 17750 }, { "epoch": 6.522218141755417, "grad_norm": 0.274267315864563, "learning_rate": 7.662639984294178e-05, "loss": 0.0273, "step": 17760 }, { "epoch": 6.525890561880279, "grad_norm": 0.2797848880290985, "learning_rate": 7.659840367456002e-05, "loss": 0.0263, "step": 17770 }, { "epoch": 6.5295629820051415, "grad_norm": 0.2364100068807602, "learning_rate": 7.657039587113287e-05, "loss": 0.0258, "step": 17780 }, { "epoch": 6.533235402130003, "grad_norm": 0.24898307025432587, "learning_rate": 7.654237644491182e-05, "loss": 0.0253, "step": 17790 }, { "epoch": 6.536907822254866, "grad_norm": 0.34796562790870667, "learning_rate": 7.651434540815358e-05, "loss": 0.029, "step": 17800 }, { "epoch": 6.540580242379728, "grad_norm": 0.25948721170425415, "learning_rate": 7.648630277311986e-05, "loss": 0.0267, "step": 17810 }, { "epoch": 6.54425266250459, "grad_norm": 0.25831809639930725, "learning_rate": 7.645824855207744e-05, "loss": 0.0252, "step": 17820 }, { "epoch": 6.547925082629453, "grad_norm": 0.3658599257469177, "learning_rate": 7.643018275729821e-05, "loss": 0.0252, "step": 17830 }, { "epoch": 6.551597502754315, "grad_norm": 0.3040847182273865, "learning_rate": 7.640210540105911e-05, "loss": 0.028, "step": 17840 }, { "epoch": 6.5552699228791775, "grad_norm": 0.2564866840839386, "learning_rate": 7.637401649564213e-05, "loss": 0.0297, "step": 17850 }, { "epoch": 6.558942343004039, "grad_norm": 0.2520524561405182, "learning_rate": 7.63459160533343e-05, "loss": 0.0249, "step": 17860 }, { "epoch": 6.562614763128902, "grad_norm": 0.23316426575183868, "learning_rate": 7.631780408642769e-05, "loss": 0.0235, "step": 17870 }, { "epoch": 6.566287183253764, "grad_norm": 0.33941829204559326, "learning_rate": 7.628968060721946e-05, "loss": 0.0276, "step": 17880 }, { "epoch": 6.569959603378626, "grad_norm": 0.23393668234348297, "learning_rate": 7.626154562801176e-05, "loss": 0.0241, "step": 17890 }, { "epoch": 6.573632023503489, "grad_norm": 0.27457574009895325, "learning_rate": 7.623339916111177e-05, "loss": 0.024, "step": 17900 }, { "epoch": 6.577304443628351, "grad_norm": 0.24049244821071625, "learning_rate": 7.620524121883174e-05, "loss": 0.0238, "step": 17910 }, { "epoch": 6.580976863753214, "grad_norm": 0.4396432042121887, "learning_rate": 7.617707181348892e-05, "loss": 0.0286, "step": 17920 }, { "epoch": 6.5846492838780755, "grad_norm": 0.27878227829933167, "learning_rate": 7.614889095740554e-05, "loss": 0.0265, "step": 17930 }, { "epoch": 6.588321704002938, "grad_norm": 0.38756808638572693, "learning_rate": 7.612069866290885e-05, "loss": 0.0265, "step": 17940 }, { "epoch": 6.5919941241278, "grad_norm": 0.3063761293888092, "learning_rate": 7.609249494233115e-05, "loss": 0.0241, "step": 17950 }, { "epoch": 6.595666544252662, "grad_norm": 0.36861011385917664, "learning_rate": 7.60642798080097e-05, "loss": 0.026, "step": 17960 }, { "epoch": 6.599338964377525, "grad_norm": 0.4309022128582001, "learning_rate": 7.603605327228673e-05, "loss": 0.0286, "step": 17970 }, { "epoch": 6.603011384502387, "grad_norm": 0.26363444328308105, "learning_rate": 7.600781534750953e-05, "loss": 0.0259, "step": 17980 }, { "epoch": 6.60668380462725, "grad_norm": 0.3192136585712433, "learning_rate": 7.597956604603029e-05, "loss": 0.027, "step": 17990 }, { "epoch": 6.6103562247521115, "grad_norm": 0.2639593183994293, "learning_rate": 7.595130538020622e-05, "loss": 0.0244, "step": 18000 }, { "epoch": 6.614028644876974, "grad_norm": 0.2539912164211273, "learning_rate": 7.592303336239952e-05, "loss": 0.0244, "step": 18010 }, { "epoch": 6.617701065001836, "grad_norm": 0.3127405345439911, "learning_rate": 7.589475000497733e-05, "loss": 0.0268, "step": 18020 }, { "epoch": 6.621373485126698, "grad_norm": 0.33425408601760864, "learning_rate": 7.586645532031171e-05, "loss": 0.0263, "step": 18030 }, { "epoch": 6.625045905251561, "grad_norm": 0.23870836198329926, "learning_rate": 7.583814932077975e-05, "loss": 0.0263, "step": 18040 }, { "epoch": 6.628718325376423, "grad_norm": 0.39090579748153687, "learning_rate": 7.580983201876343e-05, "loss": 0.0257, "step": 18050 }, { "epoch": 6.632390745501286, "grad_norm": 0.2173571139574051, "learning_rate": 7.578150342664973e-05, "loss": 0.0257, "step": 18060 }, { "epoch": 6.6360631656261475, "grad_norm": 0.2703176438808441, "learning_rate": 7.575316355683055e-05, "loss": 0.0233, "step": 18070 }, { "epoch": 6.63973558575101, "grad_norm": 0.5194708704948425, "learning_rate": 7.572481242170266e-05, "loss": 0.0246, "step": 18080 }, { "epoch": 6.643408005875872, "grad_norm": 0.200648695230484, "learning_rate": 7.569645003366786e-05, "loss": 0.0244, "step": 18090 }, { "epoch": 6.647080426000734, "grad_norm": 0.3147595524787903, "learning_rate": 7.566807640513278e-05, "loss": 0.0259, "step": 18100 }, { "epoch": 6.650752846125597, "grad_norm": 0.29303455352783203, "learning_rate": 7.563969154850905e-05, "loss": 0.0274, "step": 18110 }, { "epoch": 6.654425266250459, "grad_norm": 0.3610703945159912, "learning_rate": 7.561129547621314e-05, "loss": 0.0244, "step": 18120 }, { "epoch": 6.658097686375322, "grad_norm": 0.23346447944641113, "learning_rate": 7.558288820066648e-05, "loss": 0.024, "step": 18130 }, { "epoch": 6.6617701065001835, "grad_norm": 0.23068122565746307, "learning_rate": 7.555446973429534e-05, "loss": 0.0266, "step": 18140 }, { "epoch": 6.665442526625046, "grad_norm": 0.2561209201812744, "learning_rate": 7.552604008953097e-05, "loss": 0.0229, "step": 18150 }, { "epoch": 6.669114946749908, "grad_norm": 0.21910661458969116, "learning_rate": 7.549759927880942e-05, "loss": 0.0252, "step": 18160 }, { "epoch": 6.67278736687477, "grad_norm": 0.5016894340515137, "learning_rate": 7.54691473145717e-05, "loss": 0.0232, "step": 18170 }, { "epoch": 6.676459786999633, "grad_norm": 0.34080591797828674, "learning_rate": 7.544068420926365e-05, "loss": 0.0264, "step": 18180 }, { "epoch": 6.680132207124495, "grad_norm": 0.33807337284088135, "learning_rate": 7.5412209975336e-05, "loss": 0.0242, "step": 18190 }, { "epoch": 6.683804627249358, "grad_norm": 0.21636052429676056, "learning_rate": 7.538372462524435e-05, "loss": 0.0227, "step": 18200 }, { "epoch": 6.6874770473742196, "grad_norm": 0.23577287793159485, "learning_rate": 7.535522817144915e-05, "loss": 0.0226, "step": 18210 }, { "epoch": 6.691149467499082, "grad_norm": 0.26754578948020935, "learning_rate": 7.532672062641575e-05, "loss": 0.0245, "step": 18220 }, { "epoch": 6.694821887623944, "grad_norm": 0.28858762979507446, "learning_rate": 7.529820200261427e-05, "loss": 0.0252, "step": 18230 }, { "epoch": 6.698494307748806, "grad_norm": 0.2593161463737488, "learning_rate": 7.526967231251977e-05, "loss": 0.0239, "step": 18240 }, { "epoch": 6.702166727873669, "grad_norm": 0.258903443813324, "learning_rate": 7.524113156861206e-05, "loss": 0.025, "step": 18250 }, { "epoch": 6.705839147998531, "grad_norm": 0.24723251163959503, "learning_rate": 7.521257978337587e-05, "loss": 0.0235, "step": 18260 }, { "epoch": 6.709511568123394, "grad_norm": 0.27340930700302124, "learning_rate": 7.518401696930072e-05, "loss": 0.0246, "step": 18270 }, { "epoch": 6.713183988248256, "grad_norm": 0.28165650367736816, "learning_rate": 7.515544313888092e-05, "loss": 0.0242, "step": 18280 }, { "epoch": 6.716856408373118, "grad_norm": 0.4514661431312561, "learning_rate": 7.512685830461568e-05, "loss": 0.0269, "step": 18290 }, { "epoch": 6.72052882849798, "grad_norm": 0.26077163219451904, "learning_rate": 7.509826247900892e-05, "loss": 0.024, "step": 18300 }, { "epoch": 6.724201248622842, "grad_norm": 0.21385765075683594, "learning_rate": 7.50696556745695e-05, "loss": 0.0247, "step": 18310 }, { "epoch": 6.727873668747705, "grad_norm": 0.2793295979499817, "learning_rate": 7.504103790381095e-05, "loss": 0.0241, "step": 18320 }, { "epoch": 6.731546088872567, "grad_norm": 0.2903467118740082, "learning_rate": 7.501240917925167e-05, "loss": 0.0256, "step": 18330 }, { "epoch": 6.73521850899743, "grad_norm": 0.25302767753601074, "learning_rate": 7.498376951341483e-05, "loss": 0.0244, "step": 18340 }, { "epoch": 6.738890929122292, "grad_norm": 0.33145415782928467, "learning_rate": 7.495511891882844e-05, "loss": 0.0233, "step": 18350 }, { "epoch": 6.7425633492471535, "grad_norm": 0.27580106258392334, "learning_rate": 7.492645740802517e-05, "loss": 0.024, "step": 18360 }, { "epoch": 6.746235769372016, "grad_norm": 0.2548539936542511, "learning_rate": 7.489778499354259e-05, "loss": 0.0263, "step": 18370 }, { "epoch": 6.749908189496878, "grad_norm": 0.24162432551383972, "learning_rate": 7.486910168792298e-05, "loss": 0.0249, "step": 18380 }, { "epoch": 6.753580609621741, "grad_norm": 0.32745614647865295, "learning_rate": 7.484040750371337e-05, "loss": 0.0237, "step": 18390 }, { "epoch": 6.757253029746603, "grad_norm": 0.28757965564727783, "learning_rate": 7.48117024534656e-05, "loss": 0.0272, "step": 18400 }, { "epoch": 6.760925449871465, "grad_norm": 0.21921999752521515, "learning_rate": 7.478298654973621e-05, "loss": 0.027, "step": 18410 }, { "epoch": 6.764597869996328, "grad_norm": 0.2995629608631134, "learning_rate": 7.475425980508654e-05, "loss": 0.0251, "step": 18420 }, { "epoch": 6.7682702901211895, "grad_norm": 0.28757062554359436, "learning_rate": 7.47255222320826e-05, "loss": 0.0241, "step": 18430 }, { "epoch": 6.771942710246052, "grad_norm": 0.3808482885360718, "learning_rate": 7.469677384329522e-05, "loss": 0.0254, "step": 18440 }, { "epoch": 6.775615130370914, "grad_norm": 0.23034216463565826, "learning_rate": 7.46680146512999e-05, "loss": 0.0243, "step": 18450 }, { "epoch": 6.779287550495777, "grad_norm": 0.3388243317604065, "learning_rate": 7.463924466867692e-05, "loss": 0.0246, "step": 18460 }, { "epoch": 6.782959970620639, "grad_norm": 0.26025858521461487, "learning_rate": 7.461046390801121e-05, "loss": 0.0229, "step": 18470 }, { "epoch": 6.786632390745501, "grad_norm": 0.34925851225852966, "learning_rate": 7.458167238189248e-05, "loss": 0.0228, "step": 18480 }, { "epoch": 6.790304810870364, "grad_norm": 0.2630910575389862, "learning_rate": 7.455287010291509e-05, "loss": 0.0243, "step": 18490 }, { "epoch": 6.7939772309952255, "grad_norm": 0.24883128702640533, "learning_rate": 7.452405708367818e-05, "loss": 0.0242, "step": 18500 }, { "epoch": 6.797649651120088, "grad_norm": 0.2828759253025055, "learning_rate": 7.449523333678549e-05, "loss": 0.0279, "step": 18510 }, { "epoch": 6.80132207124495, "grad_norm": 0.2376924604177475, "learning_rate": 7.446639887484555e-05, "loss": 0.0247, "step": 18520 }, { "epoch": 6.804994491369813, "grad_norm": 0.22318704426288605, "learning_rate": 7.443755371047152e-05, "loss": 0.0236, "step": 18530 }, { "epoch": 6.808666911494675, "grad_norm": 0.2925713062286377, "learning_rate": 7.440869785628124e-05, "loss": 0.0292, "step": 18540 }, { "epoch": 6.812339331619537, "grad_norm": 0.4346839189529419, "learning_rate": 7.437983132489727e-05, "loss": 0.0235, "step": 18550 }, { "epoch": 6.8160117517444, "grad_norm": 0.33776119351387024, "learning_rate": 7.435095412894678e-05, "loss": 0.0264, "step": 18560 }, { "epoch": 6.819684171869262, "grad_norm": 0.20958948135375977, "learning_rate": 7.432206628106165e-05, "loss": 0.0257, "step": 18570 }, { "epoch": 6.823356591994124, "grad_norm": 0.25248658657073975, "learning_rate": 7.429316779387842e-05, "loss": 0.0243, "step": 18580 }, { "epoch": 6.827029012118986, "grad_norm": 0.19831442832946777, "learning_rate": 7.426425868003825e-05, "loss": 0.0256, "step": 18590 }, { "epoch": 6.830701432243849, "grad_norm": 0.2248951494693756, "learning_rate": 7.423533895218698e-05, "loss": 0.0257, "step": 18600 }, { "epoch": 6.834373852368711, "grad_norm": 0.345965713262558, "learning_rate": 7.420640862297507e-05, "loss": 0.0256, "step": 18610 }, { "epoch": 6.838046272493573, "grad_norm": 0.2906971871852875, "learning_rate": 7.417746770505766e-05, "loss": 0.0233, "step": 18620 }, { "epoch": 6.841718692618436, "grad_norm": 0.2124578058719635, "learning_rate": 7.414851621109444e-05, "loss": 0.0239, "step": 18630 }, { "epoch": 6.845391112743298, "grad_norm": 0.2596817910671234, "learning_rate": 7.411955415374982e-05, "loss": 0.0224, "step": 18640 }, { "epoch": 6.84906353286816, "grad_norm": 0.34875473380088806, "learning_rate": 7.409058154569278e-05, "loss": 0.0229, "step": 18650 }, { "epoch": 6.852735952993022, "grad_norm": 0.22864218056201935, "learning_rate": 7.406159839959689e-05, "loss": 0.0238, "step": 18660 }, { "epoch": 6.856408373117885, "grad_norm": 0.2970447540283203, "learning_rate": 7.403260472814039e-05, "loss": 0.0299, "step": 18670 }, { "epoch": 6.860080793242747, "grad_norm": 0.252895712852478, "learning_rate": 7.400360054400611e-05, "loss": 0.0252, "step": 18680 }, { "epoch": 6.863753213367609, "grad_norm": 0.4425031840801239, "learning_rate": 7.397458585988143e-05, "loss": 0.0247, "step": 18690 }, { "epoch": 6.867425633492472, "grad_norm": 0.32822439074516296, "learning_rate": 7.394556068845841e-05, "loss": 0.0244, "step": 18700 }, { "epoch": 6.871098053617334, "grad_norm": 0.3025503158569336, "learning_rate": 7.39165250424336e-05, "loss": 0.0257, "step": 18710 }, { "epoch": 6.874770473742196, "grad_norm": 0.35377734899520874, "learning_rate": 7.38874789345082e-05, "loss": 0.0254, "step": 18720 }, { "epoch": 6.878442893867058, "grad_norm": 0.3955232501029968, "learning_rate": 7.385842237738794e-05, "loss": 0.0244, "step": 18730 }, { "epoch": 6.882115313991921, "grad_norm": 0.30009129643440247, "learning_rate": 7.382935538378318e-05, "loss": 0.0246, "step": 18740 }, { "epoch": 6.885787734116783, "grad_norm": 0.24840591847896576, "learning_rate": 7.38002779664088e-05, "loss": 0.0233, "step": 18750 }, { "epoch": 6.889460154241645, "grad_norm": 0.2076328843832016, "learning_rate": 7.377119013798422e-05, "loss": 0.0237, "step": 18760 }, { "epoch": 6.893132574366508, "grad_norm": 0.2167355716228485, "learning_rate": 7.374209191123349e-05, "loss": 0.0242, "step": 18770 }, { "epoch": 6.89680499449137, "grad_norm": 0.275301069021225, "learning_rate": 7.371298329888512e-05, "loss": 0.0237, "step": 18780 }, { "epoch": 6.900477414616232, "grad_norm": 0.3452235162258148, "learning_rate": 7.368386431367223e-05, "loss": 0.0237, "step": 18790 }, { "epoch": 6.904149834741094, "grad_norm": 0.2739337086677551, "learning_rate": 7.365473496833246e-05, "loss": 0.0251, "step": 18800 }, { "epoch": 6.907822254865957, "grad_norm": 0.2185799479484558, "learning_rate": 7.362559527560797e-05, "loss": 0.0246, "step": 18810 }, { "epoch": 6.911494674990819, "grad_norm": 0.1969984769821167, "learning_rate": 7.359644524824545e-05, "loss": 0.0243, "step": 18820 }, { "epoch": 6.915167095115681, "grad_norm": 0.33408477902412415, "learning_rate": 7.35672848989961e-05, "loss": 0.0227, "step": 18830 }, { "epoch": 6.918839515240544, "grad_norm": 0.2532242238521576, "learning_rate": 7.353811424061565e-05, "loss": 0.0224, "step": 18840 }, { "epoch": 6.922511935365406, "grad_norm": 0.24991093575954437, "learning_rate": 7.350893328586435e-05, "loss": 0.0246, "step": 18850 }, { "epoch": 6.9261843554902685, "grad_norm": 0.3234580159187317, "learning_rate": 7.347974204750696e-05, "loss": 0.0246, "step": 18860 }, { "epoch": 6.92985677561513, "grad_norm": 0.29731762409210205, "learning_rate": 7.345054053831266e-05, "loss": 0.0238, "step": 18870 }, { "epoch": 6.933529195739992, "grad_norm": 0.3058214485645294, "learning_rate": 7.342132877105523e-05, "loss": 0.0246, "step": 18880 }, { "epoch": 6.937201615864855, "grad_norm": 0.3570636808872223, "learning_rate": 7.339210675851288e-05, "loss": 0.0229, "step": 18890 }, { "epoch": 6.940874035989717, "grad_norm": 0.2664746046066284, "learning_rate": 7.336287451346831e-05, "loss": 0.0246, "step": 18900 }, { "epoch": 6.94454645611458, "grad_norm": 0.33516252040863037, "learning_rate": 7.333363204870868e-05, "loss": 0.0232, "step": 18910 }, { "epoch": 6.948218876239442, "grad_norm": 0.312897652387619, "learning_rate": 7.330437937702566e-05, "loss": 0.0254, "step": 18920 }, { "epoch": 6.951891296364304, "grad_norm": 0.2103230059146881, "learning_rate": 7.327511651121534e-05, "loss": 0.024, "step": 18930 }, { "epoch": 6.955563716489166, "grad_norm": 0.18755221366882324, "learning_rate": 7.324584346407833e-05, "loss": 0.0235, "step": 18940 }, { "epoch": 6.959236136614028, "grad_norm": 0.2294468879699707, "learning_rate": 7.321656024841962e-05, "loss": 0.0229, "step": 18950 }, { "epoch": 6.962908556738891, "grad_norm": 0.3177627623081207, "learning_rate": 7.31872668770487e-05, "loss": 0.0264, "step": 18960 }, { "epoch": 6.966580976863753, "grad_norm": 0.3190106749534607, "learning_rate": 7.315796336277949e-05, "loss": 0.0231, "step": 18970 }, { "epoch": 6.970253396988616, "grad_norm": 0.1920298933982849, "learning_rate": 7.312864971843031e-05, "loss": 0.0242, "step": 18980 }, { "epoch": 6.973925817113478, "grad_norm": 0.2524753510951996, "learning_rate": 7.309932595682398e-05, "loss": 0.0223, "step": 18990 }, { "epoch": 6.97759823723834, "grad_norm": 0.19331850111484528, "learning_rate": 7.30699920907877e-05, "loss": 0.0216, "step": 19000 }, { "epoch": 6.981270657363202, "grad_norm": 0.28113821148872375, "learning_rate": 7.304064813315308e-05, "loss": 0.0231, "step": 19010 }, { "epoch": 6.984943077488064, "grad_norm": 0.3853086233139038, "learning_rate": 7.30112940967562e-05, "loss": 0.0257, "step": 19020 }, { "epoch": 6.988615497612927, "grad_norm": 0.3919420540332794, "learning_rate": 7.298192999443747e-05, "loss": 0.0249, "step": 19030 }, { "epoch": 6.992287917737789, "grad_norm": 0.3101566731929779, "learning_rate": 7.295255583904179e-05, "loss": 0.0247, "step": 19040 }, { "epoch": 6.995960337862652, "grad_norm": 0.3007933497428894, "learning_rate": 7.29231716434184e-05, "loss": 0.0232, "step": 19050 }, { "epoch": 6.999632757987514, "grad_norm": 0.26404547691345215, "learning_rate": 7.289377742042092e-05, "loss": 0.0237, "step": 19060 }, { "epoch": 7.003305178112376, "grad_norm": 0.2756626605987549, "learning_rate": 7.286437318290741e-05, "loss": 0.0238, "step": 19070 }, { "epoch": 7.006977598237238, "grad_norm": 0.21925777196884155, "learning_rate": 7.283495894374027e-05, "loss": 0.0242, "step": 19080 }, { "epoch": 7.0106500183621, "grad_norm": 0.25105178356170654, "learning_rate": 7.280553471578629e-05, "loss": 0.0229, "step": 19090 }, { "epoch": 7.014322438486963, "grad_norm": 0.24110648036003113, "learning_rate": 7.277610051191663e-05, "loss": 0.0242, "step": 19100 }, { "epoch": 7.017994858611825, "grad_norm": 0.32990381121635437, "learning_rate": 7.27466563450068e-05, "loss": 0.0236, "step": 19110 }, { "epoch": 7.021667278736688, "grad_norm": 0.22163504362106323, "learning_rate": 7.271720222793668e-05, "loss": 0.0236, "step": 19120 }, { "epoch": 7.02533969886155, "grad_norm": 0.29912275075912476, "learning_rate": 7.268773817359054e-05, "loss": 0.0243, "step": 19130 }, { "epoch": 7.029012118986412, "grad_norm": 0.4709335267543793, "learning_rate": 7.265826419485689e-05, "loss": 0.0237, "step": 19140 }, { "epoch": 7.0326845391112744, "grad_norm": 0.2103959023952484, "learning_rate": 7.26287803046287e-05, "loss": 0.0248, "step": 19150 }, { "epoch": 7.036356959236136, "grad_norm": 0.17416225373744965, "learning_rate": 7.259928651580322e-05, "loss": 0.0239, "step": 19160 }, { "epoch": 7.040029379360999, "grad_norm": 0.3385475277900696, "learning_rate": 7.256978284128202e-05, "loss": 0.0263, "step": 19170 }, { "epoch": 7.043701799485861, "grad_norm": 0.33724334836006165, "learning_rate": 7.254026929397101e-05, "loss": 0.024, "step": 19180 }, { "epoch": 7.047374219610724, "grad_norm": 0.2819097638130188, "learning_rate": 7.251074588678044e-05, "loss": 0.0234, "step": 19190 }, { "epoch": 7.051046639735586, "grad_norm": 0.24482068419456482, "learning_rate": 7.248121263262482e-05, "loss": 0.0246, "step": 19200 }, { "epoch": 7.054719059860448, "grad_norm": 0.2852606475353241, "learning_rate": 7.245166954442304e-05, "loss": 0.0247, "step": 19210 }, { "epoch": 7.0583914799853105, "grad_norm": 0.2067844271659851, "learning_rate": 7.24221166350982e-05, "loss": 0.0237, "step": 19220 }, { "epoch": 7.062063900110172, "grad_norm": 0.22818298637866974, "learning_rate": 7.239255391757777e-05, "loss": 0.0252, "step": 19230 }, { "epoch": 7.065736320235035, "grad_norm": 0.22788619995117188, "learning_rate": 7.236298140479351e-05, "loss": 0.0232, "step": 19240 }, { "epoch": 7.069408740359897, "grad_norm": 0.22036877274513245, "learning_rate": 7.233339910968141e-05, "loss": 0.0233, "step": 19250 }, { "epoch": 7.07308116048476, "grad_norm": 0.1677493155002594, "learning_rate": 7.230380704518178e-05, "loss": 0.0232, "step": 19260 }, { "epoch": 7.076753580609622, "grad_norm": 0.2709800601005554, "learning_rate": 7.227420522423919e-05, "loss": 0.0232, "step": 19270 }, { "epoch": 7.080426000734484, "grad_norm": 0.26131728291511536, "learning_rate": 7.224459365980249e-05, "loss": 0.0239, "step": 19280 }, { "epoch": 7.0840984208593465, "grad_norm": 0.24337291717529297, "learning_rate": 7.221497236482479e-05, "loss": 0.0256, "step": 19290 }, { "epoch": 7.087770840984208, "grad_norm": 0.22795744240283966, "learning_rate": 7.218534135226345e-05, "loss": 0.023, "step": 19300 }, { "epoch": 7.091443261109071, "grad_norm": 0.22708086669445038, "learning_rate": 7.215570063508005e-05, "loss": 0.0231, "step": 19310 }, { "epoch": 7.095115681233933, "grad_norm": 0.17634761333465576, "learning_rate": 7.21260502262405e-05, "loss": 0.0235, "step": 19320 }, { "epoch": 7.098788101358795, "grad_norm": 0.296440452337265, "learning_rate": 7.209639013871487e-05, "loss": 0.0246, "step": 19330 }, { "epoch": 7.102460521483658, "grad_norm": 0.18457594513893127, "learning_rate": 7.206672038547746e-05, "loss": 0.024, "step": 19340 }, { "epoch": 7.10613294160852, "grad_norm": 0.36054396629333496, "learning_rate": 7.203704097950687e-05, "loss": 0.0243, "step": 19350 }, { "epoch": 7.1098053617333825, "grad_norm": 0.21658645570278168, "learning_rate": 7.200735193378587e-05, "loss": 0.0224, "step": 19360 }, { "epoch": 7.113477781858244, "grad_norm": 0.28834423422813416, "learning_rate": 7.197765326130145e-05, "loss": 0.0229, "step": 19370 }, { "epoch": 7.117150201983107, "grad_norm": 0.33797070384025574, "learning_rate": 7.194794497504481e-05, "loss": 0.0255, "step": 19380 }, { "epoch": 7.120822622107969, "grad_norm": 0.3455098271369934, "learning_rate": 7.19182270880114e-05, "loss": 0.0241, "step": 19390 }, { "epoch": 7.124495042232831, "grad_norm": 0.2992563247680664, "learning_rate": 7.188849961320079e-05, "loss": 0.0221, "step": 19400 }, { "epoch": 7.128167462357694, "grad_norm": 0.3127216398715973, "learning_rate": 7.185876256361681e-05, "loss": 0.0242, "step": 19410 }, { "epoch": 7.131839882482556, "grad_norm": 0.23859848082065582, "learning_rate": 7.182901595226745e-05, "loss": 0.0227, "step": 19420 }, { "epoch": 7.1355123026074185, "grad_norm": 0.3399483859539032, "learning_rate": 7.179925979216491e-05, "loss": 0.0235, "step": 19430 }, { "epoch": 7.13918472273228, "grad_norm": 0.24369919300079346, "learning_rate": 7.17694940963255e-05, "loss": 0.0238, "step": 19440 }, { "epoch": 7.142857142857143, "grad_norm": 0.35449546575546265, "learning_rate": 7.17397188777698e-05, "loss": 0.0245, "step": 19450 }, { "epoch": 7.146529562982005, "grad_norm": 0.28249233961105347, "learning_rate": 7.170993414952245e-05, "loss": 0.0219, "step": 19460 }, { "epoch": 7.150201983106867, "grad_norm": 0.23378999531269073, "learning_rate": 7.168013992461233e-05, "loss": 0.0247, "step": 19470 }, { "epoch": 7.15387440323173, "grad_norm": 0.24294811487197876, "learning_rate": 7.165033621607245e-05, "loss": 0.025, "step": 19480 }, { "epoch": 7.157546823356592, "grad_norm": 0.32504594326019287, "learning_rate": 7.162052303693995e-05, "loss": 0.0252, "step": 19490 }, { "epoch": 7.1612192434814546, "grad_norm": 0.28217148780822754, "learning_rate": 7.159070040025614e-05, "loss": 0.0227, "step": 19500 }, { "epoch": 7.1648916636063165, "grad_norm": 0.2982415556907654, "learning_rate": 7.156086831906646e-05, "loss": 0.0239, "step": 19510 }, { "epoch": 7.168564083731179, "grad_norm": 0.41211023926734924, "learning_rate": 7.153102680642044e-05, "loss": 0.0236, "step": 19520 }, { "epoch": 7.172236503856041, "grad_norm": 0.2792886793613434, "learning_rate": 7.150117587537182e-05, "loss": 0.0244, "step": 19530 }, { "epoch": 7.175908923980903, "grad_norm": 0.29016828536987305, "learning_rate": 7.147131553897838e-05, "loss": 0.0233, "step": 19540 }, { "epoch": 7.179581344105766, "grad_norm": 0.20811864733695984, "learning_rate": 7.144144581030205e-05, "loss": 0.0271, "step": 19550 }, { "epoch": 7.183253764230628, "grad_norm": 0.24074284732341766, "learning_rate": 7.141156670240889e-05, "loss": 0.0255, "step": 19560 }, { "epoch": 7.186926184355491, "grad_norm": 0.36868786811828613, "learning_rate": 7.138167822836899e-05, "loss": 0.0254, "step": 19570 }, { "epoch": 7.1905986044803525, "grad_norm": 0.282387375831604, "learning_rate": 7.135178040125661e-05, "loss": 0.0236, "step": 19580 }, { "epoch": 7.194271024605214, "grad_norm": 0.27685320377349854, "learning_rate": 7.13218732341501e-05, "loss": 0.0224, "step": 19590 }, { "epoch": 7.197943444730077, "grad_norm": 0.24771401286125183, "learning_rate": 7.129195674013181e-05, "loss": 0.0234, "step": 19600 }, { "epoch": 7.201615864854939, "grad_norm": 0.237092986702919, "learning_rate": 7.126203093228829e-05, "loss": 0.0222, "step": 19610 }, { "epoch": 7.205288284979802, "grad_norm": 0.27141255140304565, "learning_rate": 7.123209582371006e-05, "loss": 0.0237, "step": 19620 }, { "epoch": 7.208960705104664, "grad_norm": 0.24271436035633087, "learning_rate": 7.120215142749179e-05, "loss": 0.0243, "step": 19630 }, { "epoch": 7.212633125229527, "grad_norm": 0.2096133828163147, "learning_rate": 7.117219775673218e-05, "loss": 0.0211, "step": 19640 }, { "epoch": 7.2163055453543885, "grad_norm": 0.20547133684158325, "learning_rate": 7.114223482453394e-05, "loss": 0.0214, "step": 19650 }, { "epoch": 7.21997796547925, "grad_norm": 0.2049928903579712, "learning_rate": 7.11122626440039e-05, "loss": 0.0234, "step": 19660 }, { "epoch": 7.223650385604113, "grad_norm": 0.21589058637619019, "learning_rate": 7.10822812282529e-05, "loss": 0.0241, "step": 19670 }, { "epoch": 7.227322805728975, "grad_norm": 0.2764926552772522, "learning_rate": 7.105229059039586e-05, "loss": 0.0208, "step": 19680 }, { "epoch": 7.230995225853838, "grad_norm": 0.3007257282733917, "learning_rate": 7.102229074355168e-05, "loss": 0.0224, "step": 19690 }, { "epoch": 7.2346676459787, "grad_norm": 0.4293968975543976, "learning_rate": 7.099228170084332e-05, "loss": 0.0289, "step": 19700 }, { "epoch": 7.238340066103563, "grad_norm": 0.2741640508174896, "learning_rate": 7.096226347539771e-05, "loss": 0.0238, "step": 19710 }, { "epoch": 7.2420124862284245, "grad_norm": 0.23259775340557098, "learning_rate": 7.093223608034592e-05, "loss": 0.0226, "step": 19720 }, { "epoch": 7.245684906353286, "grad_norm": 0.23166123032569885, "learning_rate": 7.090219952882291e-05, "loss": 0.0244, "step": 19730 }, { "epoch": 7.249357326478149, "grad_norm": 0.31682878732681274, "learning_rate": 7.087215383396768e-05, "loss": 0.0258, "step": 19740 }, { "epoch": 7.253029746603011, "grad_norm": 0.2735385000705719, "learning_rate": 7.084209900892325e-05, "loss": 0.0249, "step": 19750 }, { "epoch": 7.256702166727874, "grad_norm": 0.35036981105804443, "learning_rate": 7.081203506683663e-05, "loss": 0.023, "step": 19760 }, { "epoch": 7.260374586852736, "grad_norm": 0.3614453375339508, "learning_rate": 7.078196202085879e-05, "loss": 0.024, "step": 19770 }, { "epoch": 7.264047006977599, "grad_norm": 0.42518866062164307, "learning_rate": 7.075187988414473e-05, "loss": 0.0261, "step": 19780 }, { "epoch": 7.2677194271024605, "grad_norm": 0.3203870356082916, "learning_rate": 7.072178866985334e-05, "loss": 0.026, "step": 19790 }, { "epoch": 7.2713918472273225, "grad_norm": 0.26991716027259827, "learning_rate": 7.069168839114764e-05, "loss": 0.0228, "step": 19800 }, { "epoch": 7.275064267352185, "grad_norm": 0.22805632650852203, "learning_rate": 7.066157906119441e-05, "loss": 0.0235, "step": 19810 }, { "epoch": 7.278736687477047, "grad_norm": 0.38321346044540405, "learning_rate": 7.063146069316456e-05, "loss": 0.0229, "step": 19820 }, { "epoch": 7.28240910760191, "grad_norm": 0.2272404134273529, "learning_rate": 7.060133330023287e-05, "loss": 0.0259, "step": 19830 }, { "epoch": 7.286081527726772, "grad_norm": 0.24065947532653809, "learning_rate": 7.057119689557807e-05, "loss": 0.0262, "step": 19840 }, { "epoch": 7.289753947851635, "grad_norm": 0.27987563610076904, "learning_rate": 7.054105149238287e-05, "loss": 0.0245, "step": 19850 }, { "epoch": 7.293426367976497, "grad_norm": 0.25636592507362366, "learning_rate": 7.051089710383387e-05, "loss": 0.0251, "step": 19860 }, { "epoch": 7.2970987881013585, "grad_norm": 0.22888168692588806, "learning_rate": 7.048073374312166e-05, "loss": 0.0233, "step": 19870 }, { "epoch": 7.300771208226221, "grad_norm": 0.6648494005203247, "learning_rate": 7.045056142344067e-05, "loss": 0.0234, "step": 19880 }, { "epoch": 7.304443628351083, "grad_norm": 0.2699631154537201, "learning_rate": 7.042038015798934e-05, "loss": 0.0251, "step": 19890 }, { "epoch": 7.308116048475946, "grad_norm": 0.2802621126174927, "learning_rate": 7.039018995996995e-05, "loss": 0.0248, "step": 19900 }, { "epoch": 7.311788468600808, "grad_norm": 0.3043128550052643, "learning_rate": 7.035999084258876e-05, "loss": 0.0233, "step": 19910 }, { "epoch": 7.315460888725671, "grad_norm": 0.43914154171943665, "learning_rate": 7.032978281905583e-05, "loss": 0.0247, "step": 19920 }, { "epoch": 7.319133308850533, "grad_norm": 0.36438852548599243, "learning_rate": 7.029956590258522e-05, "loss": 0.0225, "step": 19930 }, { "epoch": 7.3228057289753945, "grad_norm": 0.1729118973016739, "learning_rate": 7.026934010639483e-05, "loss": 0.0224, "step": 19940 }, { "epoch": 7.326478149100257, "grad_norm": 0.3582545816898346, "learning_rate": 7.023910544370645e-05, "loss": 0.0236, "step": 19950 }, { "epoch": 7.330150569225119, "grad_norm": 0.2552391588687897, "learning_rate": 7.020886192774573e-05, "loss": 0.0226, "step": 19960 }, { "epoch": 7.333822989349982, "grad_norm": 0.3199057877063751, "learning_rate": 7.017860957174222e-05, "loss": 0.0235, "step": 19970 }, { "epoch": 7.337495409474844, "grad_norm": 0.27000829577445984, "learning_rate": 7.014834838892934e-05, "loss": 0.0223, "step": 19980 }, { "epoch": 7.341167829599706, "grad_norm": 0.3316521644592285, "learning_rate": 7.011807839254435e-05, "loss": 0.0241, "step": 19990 }, { "epoch": 7.344840249724569, "grad_norm": 0.2719564139842987, "learning_rate": 7.008779959582837e-05, "loss": 0.0233, "step": 20000 }, { "epoch": 7.3485126698494305, "grad_norm": 0.28684529662132263, "learning_rate": 7.005751201202639e-05, "loss": 0.0229, "step": 20010 }, { "epoch": 7.352185089974293, "grad_norm": 0.1989067792892456, "learning_rate": 7.002721565438724e-05, "loss": 0.0215, "step": 20020 }, { "epoch": 7.355857510099155, "grad_norm": 0.3633812665939331, "learning_rate": 6.999691053616353e-05, "loss": 0.0228, "step": 20030 }, { "epoch": 7.359529930224018, "grad_norm": 0.21956096589565277, "learning_rate": 6.996659667061178e-05, "loss": 0.0241, "step": 20040 }, { "epoch": 7.36320235034888, "grad_norm": 0.32092544436454773, "learning_rate": 6.993627407099229e-05, "loss": 0.0218, "step": 20050 }, { "epoch": 7.366874770473742, "grad_norm": 0.19738249480724335, "learning_rate": 6.990594275056921e-05, "loss": 0.0226, "step": 20060 }, { "epoch": 7.370547190598605, "grad_norm": 0.20780077576637268, "learning_rate": 6.987560272261049e-05, "loss": 0.0223, "step": 20070 }, { "epoch": 7.3742196107234665, "grad_norm": 0.28896698355674744, "learning_rate": 6.984525400038788e-05, "loss": 0.0233, "step": 20080 }, { "epoch": 7.377892030848329, "grad_norm": 0.22728373110294342, "learning_rate": 6.981489659717693e-05, "loss": 0.0225, "step": 20090 }, { "epoch": 7.381564450973191, "grad_norm": 0.3325047492980957, "learning_rate": 6.978453052625703e-05, "loss": 0.0231, "step": 20100 }, { "epoch": 7.385236871098053, "grad_norm": 0.1985635906457901, "learning_rate": 6.975415580091131e-05, "loss": 0.0222, "step": 20110 }, { "epoch": 7.388909291222916, "grad_norm": 0.39916351437568665, "learning_rate": 6.972377243442672e-05, "loss": 0.023, "step": 20120 }, { "epoch": 7.392581711347778, "grad_norm": 0.2921762764453888, "learning_rate": 6.969338044009396e-05, "loss": 0.0212, "step": 20130 }, { "epoch": 7.396254131472641, "grad_norm": 0.4117771089076996, "learning_rate": 6.966297983120753e-05, "loss": 0.0237, "step": 20140 }, { "epoch": 7.399926551597503, "grad_norm": 0.2839895486831665, "learning_rate": 6.963257062106571e-05, "loss": 0.0254, "step": 20150 }, { "epoch": 7.403598971722365, "grad_norm": 0.2856266498565674, "learning_rate": 6.960215282297048e-05, "loss": 0.0258, "step": 20160 }, { "epoch": 7.407271391847227, "grad_norm": 0.27133846282958984, "learning_rate": 6.957172645022765e-05, "loss": 0.022, "step": 20170 }, { "epoch": 7.410943811972089, "grad_norm": 0.3183497190475464, "learning_rate": 6.954129151614673e-05, "loss": 0.026, "step": 20180 }, { "epoch": 7.414616232096952, "grad_norm": 0.24349236488342285, "learning_rate": 6.9510848034041e-05, "loss": 0.0226, "step": 20190 }, { "epoch": 7.418288652221814, "grad_norm": 0.29143327474594116, "learning_rate": 6.94803960172275e-05, "loss": 0.0281, "step": 20200 }, { "epoch": 7.421961072346677, "grad_norm": 0.32525190711021423, "learning_rate": 6.944993547902692e-05, "loss": 0.0231, "step": 20210 }, { "epoch": 7.425633492471539, "grad_norm": 0.25474879145622253, "learning_rate": 6.941946643276379e-05, "loss": 0.0244, "step": 20220 }, { "epoch": 7.429305912596401, "grad_norm": 0.2858433723449707, "learning_rate": 6.938898889176626e-05, "loss": 0.0219, "step": 20230 }, { "epoch": 7.432978332721263, "grad_norm": 0.23640015721321106, "learning_rate": 6.935850286936627e-05, "loss": 0.0236, "step": 20240 }, { "epoch": 7.436650752846125, "grad_norm": 0.19163291156291962, "learning_rate": 6.932800837889943e-05, "loss": 0.0212, "step": 20250 }, { "epoch": 7.440323172970988, "grad_norm": 0.48487725853919983, "learning_rate": 6.92975054337051e-05, "loss": 0.0238, "step": 20260 }, { "epoch": 7.44399559309585, "grad_norm": 0.4067818224430084, "learning_rate": 6.926699404712627e-05, "loss": 0.0234, "step": 20270 }, { "epoch": 7.447668013220713, "grad_norm": 0.22451215982437134, "learning_rate": 6.923647423250967e-05, "loss": 0.0233, "step": 20280 }, { "epoch": 7.451340433345575, "grad_norm": 0.2707313597202301, "learning_rate": 6.920594600320567e-05, "loss": 0.0218, "step": 20290 }, { "epoch": 7.455012853470437, "grad_norm": 0.26224592328071594, "learning_rate": 6.91754093725684e-05, "loss": 0.0218, "step": 20300 }, { "epoch": 7.458685273595299, "grad_norm": 0.20538577437400818, "learning_rate": 6.914486435395561e-05, "loss": 0.0222, "step": 20310 }, { "epoch": 7.462357693720161, "grad_norm": 0.35103920102119446, "learning_rate": 6.911431096072871e-05, "loss": 0.0222, "step": 20320 }, { "epoch": 7.466030113845024, "grad_norm": 0.19676339626312256, "learning_rate": 6.90837492062528e-05, "loss": 0.0233, "step": 20330 }, { "epoch": 7.469702533969886, "grad_norm": 0.25322043895721436, "learning_rate": 6.905317910389664e-05, "loss": 0.0257, "step": 20340 }, { "epoch": 7.473374954094749, "grad_norm": 0.33606213331222534, "learning_rate": 6.902260066703262e-05, "loss": 0.0234, "step": 20350 }, { "epoch": 7.477047374219611, "grad_norm": 0.2975016236305237, "learning_rate": 6.89920139090368e-05, "loss": 0.0237, "step": 20360 }, { "epoch": 7.480719794344473, "grad_norm": 0.23959165811538696, "learning_rate": 6.896141884328886e-05, "loss": 0.0218, "step": 20370 }, { "epoch": 7.484392214469335, "grad_norm": 0.3019363582134247, "learning_rate": 6.893081548317211e-05, "loss": 0.0224, "step": 20380 }, { "epoch": 7.488064634594197, "grad_norm": 0.20276497304439545, "learning_rate": 6.890020384207353e-05, "loss": 0.0234, "step": 20390 }, { "epoch": 7.49173705471906, "grad_norm": 0.3031065762042999, "learning_rate": 6.886958393338366e-05, "loss": 0.022, "step": 20400 }, { "epoch": 7.495409474843922, "grad_norm": 0.30225080251693726, "learning_rate": 6.883895577049668e-05, "loss": 0.0224, "step": 20410 }, { "epoch": 7.499081894968785, "grad_norm": 0.2078598588705063, "learning_rate": 6.880831936681045e-05, "loss": 0.0231, "step": 20420 }, { "epoch": 7.502754315093647, "grad_norm": 0.3368227183818817, "learning_rate": 6.87776747357263e-05, "loss": 0.0232, "step": 20430 }, { "epoch": 7.5064267352185094, "grad_norm": 0.28301480412483215, "learning_rate": 6.874702189064927e-05, "loss": 0.0238, "step": 20440 }, { "epoch": 7.510099155343371, "grad_norm": 0.4064921736717224, "learning_rate": 6.871636084498796e-05, "loss": 0.0214, "step": 20450 }, { "epoch": 7.513771575468233, "grad_norm": 0.29180648922920227, "learning_rate": 6.868569161215453e-05, "loss": 0.0222, "step": 20460 }, { "epoch": 7.517443995593096, "grad_norm": 0.28162384033203125, "learning_rate": 6.865501420556477e-05, "loss": 0.0232, "step": 20470 }, { "epoch": 7.521116415717958, "grad_norm": 0.21322816610336304, "learning_rate": 6.862432863863802e-05, "loss": 0.0236, "step": 20480 }, { "epoch": 7.524788835842821, "grad_norm": 0.7625221014022827, "learning_rate": 6.859363492479716e-05, "loss": 0.0211, "step": 20490 }, { "epoch": 7.528461255967683, "grad_norm": 0.2348807007074356, "learning_rate": 6.856293307746868e-05, "loss": 0.0223, "step": 20500 }, { "epoch": 7.5321336760925455, "grad_norm": 0.18983571231365204, "learning_rate": 6.853222311008263e-05, "loss": 0.0252, "step": 20510 }, { "epoch": 7.535806096217407, "grad_norm": 0.3857493996620178, "learning_rate": 6.850150503607256e-05, "loss": 0.0256, "step": 20520 }, { "epoch": 7.539478516342269, "grad_norm": 0.23346176743507385, "learning_rate": 6.847077886887562e-05, "loss": 0.0223, "step": 20530 }, { "epoch": 7.543150936467132, "grad_norm": 0.20439255237579346, "learning_rate": 6.844004462193248e-05, "loss": 0.0226, "step": 20540 }, { "epoch": 7.546823356591994, "grad_norm": 0.2787385880947113, "learning_rate": 6.840930230868734e-05, "loss": 0.0229, "step": 20550 }, { "epoch": 7.550495776716857, "grad_norm": 0.2940802276134491, "learning_rate": 6.837855194258791e-05, "loss": 0.0245, "step": 20560 }, { "epoch": 7.554168196841719, "grad_norm": 0.2658807337284088, "learning_rate": 6.834779353708548e-05, "loss": 0.0249, "step": 20570 }, { "epoch": 7.557840616966581, "grad_norm": 0.3004235327243805, "learning_rate": 6.831702710563478e-05, "loss": 0.0244, "step": 20580 }, { "epoch": 7.561513037091443, "grad_norm": 0.20013582706451416, "learning_rate": 6.828625266169416e-05, "loss": 0.0222, "step": 20590 }, { "epoch": 7.565185457216305, "grad_norm": 0.3064676523208618, "learning_rate": 6.825547021872534e-05, "loss": 0.0244, "step": 20600 }, { "epoch": 7.568857877341168, "grad_norm": 0.373770534992218, "learning_rate": 6.822467979019365e-05, "loss": 0.023, "step": 20610 }, { "epoch": 7.57253029746603, "grad_norm": 0.343040406703949, "learning_rate": 6.819388138956783e-05, "loss": 0.0239, "step": 20620 }, { "epoch": 7.576202717590892, "grad_norm": 0.20805391669273376, "learning_rate": 6.816307503032019e-05, "loss": 0.0205, "step": 20630 }, { "epoch": 7.579875137715755, "grad_norm": 0.2298242747783661, "learning_rate": 6.813226072592645e-05, "loss": 0.0224, "step": 20640 }, { "epoch": 7.583547557840617, "grad_norm": 0.2583193778991699, "learning_rate": 6.810143848986583e-05, "loss": 0.024, "step": 20650 }, { "epoch": 7.587219977965479, "grad_norm": 0.23380088806152344, "learning_rate": 6.807060833562106e-05, "loss": 0.0247, "step": 20660 }, { "epoch": 7.590892398090341, "grad_norm": 0.8300031423568726, "learning_rate": 6.803977027667825e-05, "loss": 0.0231, "step": 20670 }, { "epoch": 7.594564818215204, "grad_norm": 0.2994946837425232, "learning_rate": 6.800892432652704e-05, "loss": 0.0224, "step": 20680 }, { "epoch": 7.598237238340066, "grad_norm": 0.3425619602203369, "learning_rate": 6.797807049866049e-05, "loss": 0.0238, "step": 20690 }, { "epoch": 7.601909658464928, "grad_norm": 0.3337707817554474, "learning_rate": 6.794720880657512e-05, "loss": 0.0234, "step": 20700 }, { "epoch": 7.605582078589791, "grad_norm": 0.3433572053909302, "learning_rate": 6.791633926377085e-05, "loss": 0.0222, "step": 20710 }, { "epoch": 7.609254498714653, "grad_norm": 0.2567949593067169, "learning_rate": 6.788546188375114e-05, "loss": 0.0228, "step": 20720 }, { "epoch": 7.6129269188395154, "grad_norm": 0.2115497589111328, "learning_rate": 6.785457668002274e-05, "loss": 0.0215, "step": 20730 }, { "epoch": 7.616599338964377, "grad_norm": 0.4027356505393982, "learning_rate": 6.78236836660959e-05, "loss": 0.022, "step": 20740 }, { "epoch": 7.62027175908924, "grad_norm": 0.3281591534614563, "learning_rate": 6.779278285548427e-05, "loss": 0.021, "step": 20750 }, { "epoch": 7.623944179214102, "grad_norm": 0.2381872534751892, "learning_rate": 6.776187426170493e-05, "loss": 0.0221, "step": 20760 }, { "epoch": 7.627616599338964, "grad_norm": 0.28863319754600525, "learning_rate": 6.773095789827837e-05, "loss": 0.023, "step": 20770 }, { "epoch": 7.631289019463827, "grad_norm": 0.21547682583332062, "learning_rate": 6.770003377872841e-05, "loss": 0.0232, "step": 20780 }, { "epoch": 7.634961439588689, "grad_norm": 0.252474844455719, "learning_rate": 6.766910191658233e-05, "loss": 0.0226, "step": 20790 }, { "epoch": 7.6386338597135515, "grad_norm": 0.24135065078735352, "learning_rate": 6.763816232537079e-05, "loss": 0.0249, "step": 20800 }, { "epoch": 7.642306279838413, "grad_norm": 0.2977295219898224, "learning_rate": 6.76072150186278e-05, "loss": 0.0229, "step": 20810 }, { "epoch": 7.645978699963276, "grad_norm": 0.41215887665748596, "learning_rate": 6.757626000989077e-05, "loss": 0.0229, "step": 20820 }, { "epoch": 7.649651120088138, "grad_norm": 0.19402863085269928, "learning_rate": 6.754529731270048e-05, "loss": 0.0242, "step": 20830 }, { "epoch": 7.653323540213, "grad_norm": 0.21384087204933167, "learning_rate": 6.751432694060105e-05, "loss": 0.023, "step": 20840 }, { "epoch": 7.656995960337863, "grad_norm": 0.1175934225320816, "learning_rate": 6.748334890713999e-05, "loss": 0.0239, "step": 20850 }, { "epoch": 7.660668380462725, "grad_norm": 0.16107866168022156, "learning_rate": 6.745236322586813e-05, "loss": 0.0222, "step": 20860 }, { "epoch": 7.6643408005875875, "grad_norm": 0.2966820001602173, "learning_rate": 6.742136991033967e-05, "loss": 0.021, "step": 20870 }, { "epoch": 7.668013220712449, "grad_norm": 0.2539168894290924, "learning_rate": 6.739036897411215e-05, "loss": 0.0214, "step": 20880 }, { "epoch": 7.671685640837312, "grad_norm": 0.22076316177845, "learning_rate": 6.73593604307464e-05, "loss": 0.0213, "step": 20890 }, { "epoch": 7.675358060962174, "grad_norm": 0.21446983516216278, "learning_rate": 6.732834429380663e-05, "loss": 0.0222, "step": 20900 }, { "epoch": 7.679030481087036, "grad_norm": 0.2904009521007538, "learning_rate": 6.729732057686036e-05, "loss": 0.0225, "step": 20910 }, { "epoch": 7.682702901211899, "grad_norm": 0.29972001910209656, "learning_rate": 6.726628929347839e-05, "loss": 0.0232, "step": 20920 }, { "epoch": 7.686375321336761, "grad_norm": 0.24770474433898926, "learning_rate": 6.723525045723488e-05, "loss": 0.0234, "step": 20930 }, { "epoch": 7.6900477414616235, "grad_norm": 0.35356631875038147, "learning_rate": 6.720420408170728e-05, "loss": 0.0239, "step": 20940 }, { "epoch": 7.693720161586485, "grad_norm": 0.5497469305992126, "learning_rate": 6.71731501804763e-05, "loss": 0.0224, "step": 20950 }, { "epoch": 7.697392581711348, "grad_norm": 0.2800517976284027, "learning_rate": 6.714208876712602e-05, "loss": 0.0222, "step": 20960 }, { "epoch": 7.70106500183621, "grad_norm": 0.2550998330116272, "learning_rate": 6.71110198552437e-05, "loss": 0.0243, "step": 20970 }, { "epoch": 7.704737421961072, "grad_norm": 0.3304532766342163, "learning_rate": 6.707994345841996e-05, "loss": 0.0223, "step": 20980 }, { "epoch": 7.708409842085935, "grad_norm": 0.4161221981048584, "learning_rate": 6.70488595902487e-05, "loss": 0.0247, "step": 20990 }, { "epoch": 7.712082262210797, "grad_norm": 0.27734386920928955, "learning_rate": 6.7017768264327e-05, "loss": 0.0272, "step": 21000 }, { "epoch": 7.7157546823356595, "grad_norm": 0.3363996148109436, "learning_rate": 6.698666949425532e-05, "loss": 0.0219, "step": 21010 }, { "epoch": 7.719427102460521, "grad_norm": 0.42687222361564636, "learning_rate": 6.69555632936373e-05, "loss": 0.0223, "step": 21020 }, { "epoch": 7.723099522585384, "grad_norm": 0.4138833284378052, "learning_rate": 6.692444967607983e-05, "loss": 0.0229, "step": 21030 }, { "epoch": 7.726771942710246, "grad_norm": 0.38989534974098206, "learning_rate": 6.689332865519309e-05, "loss": 0.0233, "step": 21040 }, { "epoch": 7.730444362835108, "grad_norm": 0.3303651809692383, "learning_rate": 6.686220024459045e-05, "loss": 0.0228, "step": 21050 }, { "epoch": 7.734116782959971, "grad_norm": 0.37063032388687134, "learning_rate": 6.683106445788856e-05, "loss": 0.0246, "step": 21060 }, { "epoch": 7.737789203084833, "grad_norm": 0.22112131118774414, "learning_rate": 6.679992130870723e-05, "loss": 0.0221, "step": 21070 }, { "epoch": 7.7414616232096956, "grad_norm": 0.2276989072561264, "learning_rate": 6.676877081066957e-05, "loss": 0.0248, "step": 21080 }, { "epoch": 7.7451340433345575, "grad_norm": 0.20651568472385406, "learning_rate": 6.673761297740186e-05, "loss": 0.0221, "step": 21090 }, { "epoch": 7.748806463459419, "grad_norm": 0.3001308739185333, "learning_rate": 6.670644782253358e-05, "loss": 0.0236, "step": 21100 }, { "epoch": 7.752478883584282, "grad_norm": 0.2585291266441345, "learning_rate": 6.667527535969744e-05, "loss": 0.0228, "step": 21110 }, { "epoch": 7.756151303709144, "grad_norm": 0.32218700647354126, "learning_rate": 6.664409560252933e-05, "loss": 0.0231, "step": 21120 }, { "epoch": 7.759823723834007, "grad_norm": 0.2402542531490326, "learning_rate": 6.661290856466833e-05, "loss": 0.0251, "step": 21130 }, { "epoch": 7.763496143958869, "grad_norm": 0.4651947021484375, "learning_rate": 6.658171425975672e-05, "loss": 0.0245, "step": 21140 }, { "epoch": 7.767168564083731, "grad_norm": 0.31865862011909485, "learning_rate": 6.655051270143994e-05, "loss": 0.022, "step": 21150 }, { "epoch": 7.7708409842085935, "grad_norm": 0.28430283069610596, "learning_rate": 6.651930390336662e-05, "loss": 0.0221, "step": 21160 }, { "epoch": 7.774513404333455, "grad_norm": 0.2574547231197357, "learning_rate": 6.648808787918854e-05, "loss": 0.0244, "step": 21170 }, { "epoch": 7.778185824458318, "grad_norm": 0.2673279941082001, "learning_rate": 6.645686464256065e-05, "loss": 0.0222, "step": 21180 }, { "epoch": 7.78185824458318, "grad_norm": 0.24267394840717316, "learning_rate": 6.642563420714106e-05, "loss": 0.022, "step": 21190 }, { "epoch": 7.785530664708043, "grad_norm": 0.2427055388689041, "learning_rate": 6.639439658659102e-05, "loss": 0.0225, "step": 21200 }, { "epoch": 7.789203084832905, "grad_norm": 0.21865634620189667, "learning_rate": 6.636315179457492e-05, "loss": 0.0223, "step": 21210 }, { "epoch": 7.792875504957767, "grad_norm": 0.23387901484966278, "learning_rate": 6.633189984476031e-05, "loss": 0.0222, "step": 21220 }, { "epoch": 7.7965479250826295, "grad_norm": 0.23529238998889923, "learning_rate": 6.630064075081785e-05, "loss": 0.0231, "step": 21230 }, { "epoch": 7.800220345207491, "grad_norm": 0.3093707263469696, "learning_rate": 6.626937452642132e-05, "loss": 0.0221, "step": 21240 }, { "epoch": 7.803892765332354, "grad_norm": 0.3196214735507965, "learning_rate": 6.623810118524765e-05, "loss": 0.0221, "step": 21250 }, { "epoch": 7.807565185457216, "grad_norm": 0.39034005999565125, "learning_rate": 6.620682074097682e-05, "loss": 0.0237, "step": 21260 }, { "epoch": 7.811237605582079, "grad_norm": 0.3521046042442322, "learning_rate": 6.6175533207292e-05, "loss": 0.0247, "step": 21270 }, { "epoch": 7.814910025706941, "grad_norm": 0.2573798596858978, "learning_rate": 6.61442385978794e-05, "loss": 0.0233, "step": 21280 }, { "epoch": 7.818582445831803, "grad_norm": 0.3091355264186859, "learning_rate": 6.611293692642836e-05, "loss": 0.0237, "step": 21290 }, { "epoch": 7.8222548659566655, "grad_norm": 0.2847461998462677, "learning_rate": 6.608162820663128e-05, "loss": 0.0247, "step": 21300 }, { "epoch": 7.825927286081527, "grad_norm": 0.26362815499305725, "learning_rate": 6.605031245218365e-05, "loss": 0.0213, "step": 21310 }, { "epoch": 7.82959970620639, "grad_norm": 0.2771645188331604, "learning_rate": 6.601898967678408e-05, "loss": 0.0238, "step": 21320 }, { "epoch": 7.833272126331252, "grad_norm": 0.4259653389453888, "learning_rate": 6.598765989413419e-05, "loss": 0.0233, "step": 21330 }, { "epoch": 7.836944546456115, "grad_norm": 0.5030953884124756, "learning_rate": 6.595632311793867e-05, "loss": 0.0229, "step": 21340 }, { "epoch": 7.840616966580977, "grad_norm": 0.282998263835907, "learning_rate": 6.592497936190534e-05, "loss": 0.0235, "step": 21350 }, { "epoch": 7.844289386705839, "grad_norm": 0.26816126704216003, "learning_rate": 6.5893628639745e-05, "loss": 0.0246, "step": 21360 }, { "epoch": 7.8479618068307015, "grad_norm": 0.3465459644794464, "learning_rate": 6.586227096517151e-05, "loss": 0.0222, "step": 21370 }, { "epoch": 7.8516342269555635, "grad_norm": 0.44157376885414124, "learning_rate": 6.58309063519018e-05, "loss": 0.022, "step": 21380 }, { "epoch": 7.855306647080426, "grad_norm": 0.25050434470176697, "learning_rate": 6.579953481365582e-05, "loss": 0.0223, "step": 21390 }, { "epoch": 7.858979067205288, "grad_norm": 0.3419983386993408, "learning_rate": 6.576815636415652e-05, "loss": 0.0227, "step": 21400 }, { "epoch": 7.862651487330151, "grad_norm": 0.2795390188694, "learning_rate": 6.573677101712991e-05, "loss": 0.0238, "step": 21410 }, { "epoch": 7.866323907455013, "grad_norm": 0.23786762356758118, "learning_rate": 6.570537878630502e-05, "loss": 0.023, "step": 21420 }, { "epoch": 7.869996327579875, "grad_norm": 0.28167012333869934, "learning_rate": 6.56739796854139e-05, "loss": 0.0222, "step": 21430 }, { "epoch": 7.873668747704738, "grad_norm": 0.3349510133266449, "learning_rate": 6.564257372819153e-05, "loss": 0.0239, "step": 21440 }, { "epoch": 7.8773411678295995, "grad_norm": 0.206067755818367, "learning_rate": 6.561116092837596e-05, "loss": 0.0239, "step": 21450 }, { "epoch": 7.881013587954462, "grad_norm": 0.20309433341026306, "learning_rate": 6.557974129970825e-05, "loss": 0.0215, "step": 21460 }, { "epoch": 7.884686008079324, "grad_norm": 0.2442062497138977, "learning_rate": 6.554831485593239e-05, "loss": 0.0227, "step": 21470 }, { "epoch": 7.888358428204187, "grad_norm": 0.2849159836769104, "learning_rate": 6.551688161079534e-05, "loss": 0.0208, "step": 21480 }, { "epoch": 7.892030848329049, "grad_norm": 0.22789937257766724, "learning_rate": 6.548544157804712e-05, "loss": 0.0222, "step": 21490 }, { "epoch": 7.895703268453911, "grad_norm": 0.6462578177452087, "learning_rate": 6.545399477144065e-05, "loss": 0.0211, "step": 21500 }, { "epoch": 7.899375688578774, "grad_norm": 0.2550191581249237, "learning_rate": 6.542254120473184e-05, "loss": 0.0213, "step": 21510 }, { "epoch": 7.9030481087036355, "grad_norm": 0.2639574110507965, "learning_rate": 6.539108089167953e-05, "loss": 0.0241, "step": 21520 }, { "epoch": 7.906720528828498, "grad_norm": 0.25723329186439514, "learning_rate": 6.535961384604554e-05, "loss": 0.0216, "step": 21530 }, { "epoch": 7.91039294895336, "grad_norm": 0.2388553023338318, "learning_rate": 6.532814008159461e-05, "loss": 0.0224, "step": 21540 }, { "epoch": 7.914065369078223, "grad_norm": 0.25545281171798706, "learning_rate": 6.529665961209446e-05, "loss": 0.0217, "step": 21550 }, { "epoch": 7.917737789203085, "grad_norm": 0.20402707159519196, "learning_rate": 6.52651724513157e-05, "loss": 0.0201, "step": 21560 }, { "epoch": 7.921410209327947, "grad_norm": 0.2019907534122467, "learning_rate": 6.523367861303189e-05, "loss": 0.0217, "step": 21570 }, { "epoch": 7.92508262945281, "grad_norm": 0.22669656574726105, "learning_rate": 6.520217811101951e-05, "loss": 0.0222, "step": 21580 }, { "epoch": 7.9287550495776715, "grad_norm": 0.23559004068374634, "learning_rate": 6.517067095905793e-05, "loss": 0.024, "step": 21590 }, { "epoch": 7.932427469702534, "grad_norm": 0.28043410181999207, "learning_rate": 6.513915717092948e-05, "loss": 0.0219, "step": 21600 }, { "epoch": 7.936099889827396, "grad_norm": 0.2749064266681671, "learning_rate": 6.510763676041935e-05, "loss": 0.0231, "step": 21610 }, { "epoch": 7.939772309952259, "grad_norm": 0.3767941892147064, "learning_rate": 6.50761097413156e-05, "loss": 0.023, "step": 21620 }, { "epoch": 7.943444730077121, "grad_norm": 0.29293355345726013, "learning_rate": 6.504457612740928e-05, "loss": 0.0234, "step": 21630 }, { "epoch": 7.947117150201983, "grad_norm": 0.3386143743991852, "learning_rate": 6.501303593249423e-05, "loss": 0.0222, "step": 21640 }, { "epoch": 7.950789570326846, "grad_norm": 0.2933361828327179, "learning_rate": 6.49814891703672e-05, "loss": 0.0204, "step": 21650 }, { "epoch": 7.9544619904517075, "grad_norm": 0.17368318140506744, "learning_rate": 6.494993585482783e-05, "loss": 0.0216, "step": 21660 }, { "epoch": 7.95813441057657, "grad_norm": 0.33431240916252136, "learning_rate": 6.49183759996786e-05, "loss": 0.022, "step": 21670 }, { "epoch": 7.961806830701432, "grad_norm": 0.22633367776870728, "learning_rate": 6.488680961872486e-05, "loss": 0.0224, "step": 21680 }, { "epoch": 7.965479250826294, "grad_norm": 0.35623255372047424, "learning_rate": 6.485523672577485e-05, "loss": 0.0231, "step": 21690 }, { "epoch": 7.969151670951157, "grad_norm": 0.29127731919288635, "learning_rate": 6.482365733463959e-05, "loss": 0.0249, "step": 21700 }, { "epoch": 7.972824091076019, "grad_norm": 0.19998469948768616, "learning_rate": 6.4792071459133e-05, "loss": 0.0217, "step": 21710 }, { "epoch": 7.976496511200882, "grad_norm": 0.3375009596347809, "learning_rate": 6.476047911307179e-05, "loss": 0.0205, "step": 21720 }, { "epoch": 7.980168931325744, "grad_norm": 0.3243757486343384, "learning_rate": 6.472888031027556e-05, "loss": 0.0215, "step": 21730 }, { "epoch": 7.9838413514506055, "grad_norm": 0.24846196174621582, "learning_rate": 6.469727506456666e-05, "loss": 0.0208, "step": 21740 }, { "epoch": 7.987513771575468, "grad_norm": 0.222214013338089, "learning_rate": 6.466566338977034e-05, "loss": 0.0227, "step": 21750 }, { "epoch": 7.99118619170033, "grad_norm": 0.435680091381073, "learning_rate": 6.46340452997146e-05, "loss": 0.0229, "step": 21760 }, { "epoch": 7.994858611825193, "grad_norm": 0.16867439448833466, "learning_rate": 6.460242080823025e-05, "loss": 0.0219, "step": 21770 }, { "epoch": 7.998531031950055, "grad_norm": 0.4333229660987854, "learning_rate": 6.457078992915093e-05, "loss": 0.0233, "step": 21780 }, { "epoch": 8.002203452074918, "grad_norm": 0.2600307762622833, "learning_rate": 6.453915267631308e-05, "loss": 0.0224, "step": 21790 }, { "epoch": 8.00587587219978, "grad_norm": 0.2531704008579254, "learning_rate": 6.450750906355589e-05, "loss": 0.0223, "step": 21800 }, { "epoch": 8.009548292324641, "grad_norm": 0.25449785590171814, "learning_rate": 6.447585910472134e-05, "loss": 0.0232, "step": 21810 }, { "epoch": 8.013220712449504, "grad_norm": 0.2808874845504761, "learning_rate": 6.444420281365424e-05, "loss": 0.0219, "step": 21820 }, { "epoch": 8.016893132574367, "grad_norm": 0.21463026106357574, "learning_rate": 6.441254020420208e-05, "loss": 0.0223, "step": 21830 }, { "epoch": 8.020565552699228, "grad_norm": 0.3119121491909027, "learning_rate": 6.438087129021522e-05, "loss": 0.0218, "step": 21840 }, { "epoch": 8.024237972824091, "grad_norm": 0.2741183042526245, "learning_rate": 6.434919608554663e-05, "loss": 0.023, "step": 21850 }, { "epoch": 8.027910392948954, "grad_norm": 0.2594101130962372, "learning_rate": 6.43175146040522e-05, "loss": 0.0207, "step": 21860 }, { "epoch": 8.031582813073816, "grad_norm": 0.3102363646030426, "learning_rate": 6.428582685959045e-05, "loss": 0.022, "step": 21870 }, { "epoch": 8.035255233198678, "grad_norm": 0.32198962569236755, "learning_rate": 6.42541328660227e-05, "loss": 0.0208, "step": 21880 }, { "epoch": 8.03892765332354, "grad_norm": 0.20729795098304749, "learning_rate": 6.422243263721293e-05, "loss": 0.0217, "step": 21890 }, { "epoch": 8.042600073448403, "grad_norm": 0.22619496285915375, "learning_rate": 6.419072618702794e-05, "loss": 0.0216, "step": 21900 }, { "epoch": 8.046272493573264, "grad_norm": 0.23390859365463257, "learning_rate": 6.41590135293372e-05, "loss": 0.021, "step": 21910 }, { "epoch": 8.049944913698127, "grad_norm": 0.21926778554916382, "learning_rate": 6.41272946780129e-05, "loss": 0.0211, "step": 21920 }, { "epoch": 8.05361733382299, "grad_norm": 0.3703407347202301, "learning_rate": 6.409556964692996e-05, "loss": 0.0245, "step": 21930 }, { "epoch": 8.057289753947853, "grad_norm": 0.3779028654098511, "learning_rate": 6.406383844996595e-05, "loss": 0.0238, "step": 21940 }, { "epoch": 8.060962174072714, "grad_norm": 0.28195127844810486, "learning_rate": 6.403210110100121e-05, "loss": 0.022, "step": 21950 }, { "epoch": 8.064634594197576, "grad_norm": 0.3152956962585449, "learning_rate": 6.400035761391871e-05, "loss": 0.0228, "step": 21960 }, { "epoch": 8.06830701432244, "grad_norm": 0.5359178185462952, "learning_rate": 6.396860800260416e-05, "loss": 0.0208, "step": 21970 }, { "epoch": 8.0719794344473, "grad_norm": 0.1945299655199051, "learning_rate": 6.393685228094588e-05, "loss": 0.026, "step": 21980 }, { "epoch": 8.075651854572163, "grad_norm": 0.16965384781360626, "learning_rate": 6.390509046283493e-05, "loss": 0.022, "step": 21990 }, { "epoch": 8.079324274697026, "grad_norm": 0.18780316412448883, "learning_rate": 6.387332256216498e-05, "loss": 0.0215, "step": 22000 }, { "epoch": 8.082996694821887, "grad_norm": 0.35356661677360535, "learning_rate": 6.384154859283243e-05, "loss": 0.0221, "step": 22010 }, { "epoch": 8.08666911494675, "grad_norm": 0.173456609249115, "learning_rate": 6.380976856873627e-05, "loss": 0.0232, "step": 22020 }, { "epoch": 8.090341535071612, "grad_norm": 0.26372966170310974, "learning_rate": 6.377798250377815e-05, "loss": 0.0234, "step": 22030 }, { "epoch": 8.094013955196475, "grad_norm": 0.23183104395866394, "learning_rate": 6.374619041186238e-05, "loss": 0.023, "step": 22040 }, { "epoch": 8.097686375321336, "grad_norm": 0.45477232336997986, "learning_rate": 6.371439230689592e-05, "loss": 0.0222, "step": 22050 }, { "epoch": 8.101358795446199, "grad_norm": 0.2154809981584549, "learning_rate": 6.368258820278833e-05, "loss": 0.0197, "step": 22060 }, { "epoch": 8.105031215571062, "grad_norm": 0.2500705122947693, "learning_rate": 6.365077811345182e-05, "loss": 0.0196, "step": 22070 }, { "epoch": 8.108703635695923, "grad_norm": 0.15424180030822754, "learning_rate": 6.361896205280117e-05, "loss": 0.0217, "step": 22080 }, { "epoch": 8.112376055820786, "grad_norm": 0.21619805693626404, "learning_rate": 6.358714003475384e-05, "loss": 0.0218, "step": 22090 }, { "epoch": 8.116048475945648, "grad_norm": 0.2803816497325897, "learning_rate": 6.355531207322983e-05, "loss": 0.0238, "step": 22100 }, { "epoch": 8.119720896070511, "grad_norm": 0.3344622254371643, "learning_rate": 6.35234781821518e-05, "loss": 0.0244, "step": 22110 }, { "epoch": 8.123393316195372, "grad_norm": 0.24622249603271484, "learning_rate": 6.349163837544497e-05, "loss": 0.0221, "step": 22120 }, { "epoch": 8.127065736320235, "grad_norm": 0.255928635597229, "learning_rate": 6.345979266703714e-05, "loss": 0.0214, "step": 22130 }, { "epoch": 8.130738156445098, "grad_norm": 0.3699726164340973, "learning_rate": 6.342794107085872e-05, "loss": 0.0222, "step": 22140 }, { "epoch": 8.134410576569959, "grad_norm": 0.16245658695697784, "learning_rate": 6.339608360084265e-05, "loss": 0.0217, "step": 22150 }, { "epoch": 8.138082996694822, "grad_norm": 0.15435796976089478, "learning_rate": 6.336422027092452e-05, "loss": 0.0208, "step": 22160 }, { "epoch": 8.141755416819684, "grad_norm": 0.23334839940071106, "learning_rate": 6.333235109504241e-05, "loss": 0.0224, "step": 22170 }, { "epoch": 8.145427836944547, "grad_norm": 0.2648986279964447, "learning_rate": 6.330047608713695e-05, "loss": 0.0211, "step": 22180 }, { "epoch": 8.149100257069408, "grad_norm": 0.2974124252796173, "learning_rate": 6.32685952611514e-05, "loss": 0.0206, "step": 22190 }, { "epoch": 8.152772677194271, "grad_norm": 0.24592211842536926, "learning_rate": 6.323670863103149e-05, "loss": 0.0206, "step": 22200 }, { "epoch": 8.156445097319134, "grad_norm": 0.3990066349506378, "learning_rate": 6.32048162107255e-05, "loss": 0.02, "step": 22210 }, { "epoch": 8.160117517443995, "grad_norm": 0.22319337725639343, "learning_rate": 6.317291801418431e-05, "loss": 0.022, "step": 22220 }, { "epoch": 8.163789937568858, "grad_norm": 0.3226301074028015, "learning_rate": 6.31410140553612e-05, "loss": 0.0217, "step": 22230 }, { "epoch": 8.16746235769372, "grad_norm": 0.28364017605781555, "learning_rate": 6.310910434821211e-05, "loss": 0.0228, "step": 22240 }, { "epoch": 8.171134777818583, "grad_norm": 0.26624810695648193, "learning_rate": 6.307718890669541e-05, "loss": 0.0217, "step": 22250 }, { "epoch": 8.174807197943444, "grad_norm": 0.2358824610710144, "learning_rate": 6.304526774477198e-05, "loss": 0.0231, "step": 22260 }, { "epoch": 8.178479618068307, "grad_norm": 0.2958988845348358, "learning_rate": 6.301334087640526e-05, "loss": 0.021, "step": 22270 }, { "epoch": 8.18215203819317, "grad_norm": 0.333046019077301, "learning_rate": 6.298140831556112e-05, "loss": 0.0227, "step": 22280 }, { "epoch": 8.18582445831803, "grad_norm": 0.37359997630119324, "learning_rate": 6.294947007620794e-05, "loss": 0.0196, "step": 22290 }, { "epoch": 8.189496878442894, "grad_norm": 0.48238706588745117, "learning_rate": 6.291752617231661e-05, "loss": 0.0201, "step": 22300 }, { "epoch": 8.193169298567756, "grad_norm": 0.2640591263771057, "learning_rate": 6.288557661786047e-05, "loss": 0.0204, "step": 22310 }, { "epoch": 8.19684171869262, "grad_norm": 0.2997620701789856, "learning_rate": 6.285362142681534e-05, "loss": 0.0247, "step": 22320 }, { "epoch": 8.20051413881748, "grad_norm": 0.35877302289009094, "learning_rate": 6.28216606131595e-05, "loss": 0.0224, "step": 22330 }, { "epoch": 8.204186558942343, "grad_norm": 0.3724140226840973, "learning_rate": 6.27896941908737e-05, "loss": 0.0221, "step": 22340 }, { "epoch": 8.207858979067206, "grad_norm": 0.36579951643943787, "learning_rate": 6.275772217394115e-05, "loss": 0.0222, "step": 22350 }, { "epoch": 8.211531399192067, "grad_norm": 0.2645737826824188, "learning_rate": 6.27257445763475e-05, "loss": 0.0208, "step": 22360 }, { "epoch": 8.21520381931693, "grad_norm": 0.26770734786987305, "learning_rate": 6.269376141208082e-05, "loss": 0.021, "step": 22370 }, { "epoch": 8.218876239441792, "grad_norm": 0.20772282779216766, "learning_rate": 6.266177269513162e-05, "loss": 0.0242, "step": 22380 }, { "epoch": 8.222548659566655, "grad_norm": 0.27000439167022705, "learning_rate": 6.262977843949289e-05, "loss": 0.0219, "step": 22390 }, { "epoch": 8.226221079691516, "grad_norm": 0.23805762827396393, "learning_rate": 6.259777865915997e-05, "loss": 0.0224, "step": 22400 }, { "epoch": 8.229893499816379, "grad_norm": 0.24306952953338623, "learning_rate": 6.256577336813069e-05, "loss": 0.0231, "step": 22410 }, { "epoch": 8.233565919941242, "grad_norm": 0.2985019385814667, "learning_rate": 6.253376258040521e-05, "loss": 0.0209, "step": 22420 }, { "epoch": 8.237238340066103, "grad_norm": 0.26374953985214233, "learning_rate": 6.250174630998616e-05, "loss": 0.0224, "step": 22430 }, { "epoch": 8.240910760190966, "grad_norm": 0.25928178429603577, "learning_rate": 6.246972457087854e-05, "loss": 0.0217, "step": 22440 }, { "epoch": 8.244583180315828, "grad_norm": 0.22988545894622803, "learning_rate": 6.243769737708974e-05, "loss": 0.0265, "step": 22450 }, { "epoch": 8.248255600440691, "grad_norm": 0.6141746044158936, "learning_rate": 6.240566474262955e-05, "loss": 0.0244, "step": 22460 }, { "epoch": 8.251928020565552, "grad_norm": 0.18861693143844604, "learning_rate": 6.237362668151012e-05, "loss": 0.0206, "step": 22470 }, { "epoch": 8.255600440690415, "grad_norm": 0.27976590394973755, "learning_rate": 6.2341583207746e-05, "loss": 0.0226, "step": 22480 }, { "epoch": 8.259272860815278, "grad_norm": 0.2589653730392456, "learning_rate": 6.23095343353541e-05, "loss": 0.0227, "step": 22490 }, { "epoch": 8.262945280940139, "grad_norm": 0.345901221036911, "learning_rate": 6.227748007835366e-05, "loss": 0.0211, "step": 22500 }, { "epoch": 8.266617701065002, "grad_norm": 0.3715842366218567, "learning_rate": 6.224542045076632e-05, "loss": 0.0212, "step": 22510 }, { "epoch": 8.270290121189865, "grad_norm": 0.3535839915275574, "learning_rate": 6.221335546661606e-05, "loss": 0.0208, "step": 22520 }, { "epoch": 8.273962541314727, "grad_norm": 0.35591447353363037, "learning_rate": 6.218128513992917e-05, "loss": 0.0213, "step": 22530 }, { "epoch": 8.277634961439588, "grad_norm": 0.506365180015564, "learning_rate": 6.214920948473433e-05, "loss": 0.0196, "step": 22540 }, { "epoch": 8.281307381564451, "grad_norm": 0.20687313377857208, "learning_rate": 6.211712851506249e-05, "loss": 0.02, "step": 22550 }, { "epoch": 8.284979801689314, "grad_norm": 0.2891630530357361, "learning_rate": 6.208504224494696e-05, "loss": 0.0229, "step": 22560 }, { "epoch": 8.288652221814175, "grad_norm": 0.2819647490978241, "learning_rate": 6.205295068842338e-05, "loss": 0.0208, "step": 22570 }, { "epoch": 8.292324641939038, "grad_norm": 0.25595635175704956, "learning_rate": 6.202085385952968e-05, "loss": 0.0212, "step": 22580 }, { "epoch": 8.2959970620639, "grad_norm": 0.30583956837654114, "learning_rate": 6.19887517723061e-05, "loss": 0.0211, "step": 22590 }, { "epoch": 8.299669482188762, "grad_norm": 0.6246015429496765, "learning_rate": 6.19566444407952e-05, "loss": 0.0215, "step": 22600 }, { "epoch": 8.303341902313624, "grad_norm": 0.3007830083370209, "learning_rate": 6.192453187904178e-05, "loss": 0.0224, "step": 22610 }, { "epoch": 8.307014322438487, "grad_norm": 0.22327104210853577, "learning_rate": 6.189241410109303e-05, "loss": 0.0226, "step": 22620 }, { "epoch": 8.31068674256335, "grad_norm": 0.28829634189605713, "learning_rate": 6.186029112099829e-05, "loss": 0.0202, "step": 22630 }, { "epoch": 8.314359162688211, "grad_norm": 0.3860694468021393, "learning_rate": 6.182816295280926e-05, "loss": 0.022, "step": 22640 }, { "epoch": 8.318031582813074, "grad_norm": 0.4609093964099884, "learning_rate": 6.179602961057993e-05, "loss": 0.022, "step": 22650 }, { "epoch": 8.321704002937937, "grad_norm": 0.22446991503238678, "learning_rate": 6.176389110836647e-05, "loss": 0.0219, "step": 22660 }, { "epoch": 8.325376423062798, "grad_norm": 0.3885171115398407, "learning_rate": 6.173174746022738e-05, "loss": 0.0215, "step": 22670 }, { "epoch": 8.32904884318766, "grad_norm": 0.25333908200263977, "learning_rate": 6.169959868022338e-05, "loss": 0.0227, "step": 22680 }, { "epoch": 8.332721263312523, "grad_norm": 0.1946304738521576, "learning_rate": 6.166744478241742e-05, "loss": 0.0222, "step": 22690 }, { "epoch": 8.336393683437386, "grad_norm": 0.3429334759712219, "learning_rate": 6.163528578087473e-05, "loss": 0.0228, "step": 22700 }, { "epoch": 8.340066103562247, "grad_norm": 0.1768101304769516, "learning_rate": 6.160312168966274e-05, "loss": 0.0209, "step": 22710 }, { "epoch": 8.34373852368711, "grad_norm": 0.22272206842899323, "learning_rate": 6.157095252285112e-05, "loss": 0.0227, "step": 22720 }, { "epoch": 8.347410943811973, "grad_norm": 0.27020394802093506, "learning_rate": 6.153877829451174e-05, "loss": 0.0224, "step": 22730 }, { "epoch": 8.351083363936834, "grad_norm": 0.25993847846984863, "learning_rate": 6.150659901871872e-05, "loss": 0.0218, "step": 22740 }, { "epoch": 8.354755784061696, "grad_norm": 0.21658560633659363, "learning_rate": 6.147441470954836e-05, "loss": 0.0212, "step": 22750 }, { "epoch": 8.35842820418656, "grad_norm": 0.32566729187965393, "learning_rate": 6.144222538107918e-05, "loss": 0.0223, "step": 22760 }, { "epoch": 8.362100624311422, "grad_norm": 0.27099356055259705, "learning_rate": 6.141003104739188e-05, "loss": 0.0217, "step": 22770 }, { "epoch": 8.365773044436283, "grad_norm": 0.2576679289340973, "learning_rate": 6.137783172256935e-05, "loss": 0.021, "step": 22780 }, { "epoch": 8.369445464561146, "grad_norm": 0.3468053638935089, "learning_rate": 6.134562742069669e-05, "loss": 0.0218, "step": 22790 }, { "epoch": 8.373117884686009, "grad_norm": 0.2937867343425751, "learning_rate": 6.131341815586111e-05, "loss": 0.0216, "step": 22800 }, { "epoch": 8.37679030481087, "grad_norm": 0.3091747760772705, "learning_rate": 6.128120394215209e-05, "loss": 0.0232, "step": 22810 }, { "epoch": 8.380462724935732, "grad_norm": 0.3281193673610687, "learning_rate": 6.124898479366116e-05, "loss": 0.0214, "step": 22820 }, { "epoch": 8.384135145060595, "grad_norm": 0.20103760063648224, "learning_rate": 6.121676072448214e-05, "loss": 0.0203, "step": 22830 }, { "epoch": 8.387807565185458, "grad_norm": 0.2934650480747223, "learning_rate": 6.118453174871089e-05, "loss": 0.0228, "step": 22840 }, { "epoch": 8.391479985310319, "grad_norm": 0.2551950216293335, "learning_rate": 6.115229788044548e-05, "loss": 0.0199, "step": 22850 }, { "epoch": 8.395152405435182, "grad_norm": 0.14587852358818054, "learning_rate": 6.112005913378607e-05, "loss": 0.0209, "step": 22860 }, { "epoch": 8.398824825560045, "grad_norm": 0.2002585232257843, "learning_rate": 6.1087815522835e-05, "loss": 0.0203, "step": 22870 }, { "epoch": 8.402497245684906, "grad_norm": 0.26049745082855225, "learning_rate": 6.105556706169673e-05, "loss": 0.0213, "step": 22880 }, { "epoch": 8.406169665809768, "grad_norm": 0.28098076581954956, "learning_rate": 6.102331376447782e-05, "loss": 0.0235, "step": 22890 }, { "epoch": 8.409842085934631, "grad_norm": 0.2562839686870575, "learning_rate": 6.099105564528697e-05, "loss": 0.0211, "step": 22900 }, { "epoch": 8.413514506059494, "grad_norm": 0.36292293667793274, "learning_rate": 6.095879271823495e-05, "loss": 0.022, "step": 22910 }, { "epoch": 8.417186926184355, "grad_norm": 0.460105836391449, "learning_rate": 6.09265249974347e-05, "loss": 0.0217, "step": 22920 }, { "epoch": 8.420859346309218, "grad_norm": 0.33331790566444397, "learning_rate": 6.089425249700118e-05, "loss": 0.0214, "step": 22930 }, { "epoch": 8.42453176643408, "grad_norm": 0.41474878787994385, "learning_rate": 6.0861975231051505e-05, "loss": 0.0202, "step": 22940 }, { "epoch": 8.428204186558942, "grad_norm": 0.3121233880519867, "learning_rate": 6.082969321370484e-05, "loss": 0.0222, "step": 22950 }, { "epoch": 8.431876606683804, "grad_norm": 0.3271666169166565, "learning_rate": 6.079740645908243e-05, "loss": 0.0197, "step": 22960 }, { "epoch": 8.435549026808667, "grad_norm": 0.17104566097259521, "learning_rate": 6.07651149813076e-05, "loss": 0.0205, "step": 22970 }, { "epoch": 8.43922144693353, "grad_norm": 0.2867182493209839, "learning_rate": 6.073281879450574e-05, "loss": 0.022, "step": 22980 }, { "epoch": 8.442893867058391, "grad_norm": 0.3994315564632416, "learning_rate": 6.070051791280431e-05, "loss": 0.0201, "step": 22990 }, { "epoch": 8.446566287183254, "grad_norm": 0.29605525732040405, "learning_rate": 6.066821235033281e-05, "loss": 0.0223, "step": 23000 }, { "epoch": 8.450238707308117, "grad_norm": 0.25658178329467773, "learning_rate": 6.063590212122277e-05, "loss": 0.0211, "step": 23010 }, { "epoch": 8.453911127432978, "grad_norm": 0.2106522023677826, "learning_rate": 6.0603587239607794e-05, "loss": 0.021, "step": 23020 }, { "epoch": 8.45758354755784, "grad_norm": 0.20072536170482635, "learning_rate": 6.057126771962353e-05, "loss": 0.0209, "step": 23030 }, { "epoch": 8.461255967682703, "grad_norm": 0.15866127610206604, "learning_rate": 6.05389435754076e-05, "loss": 0.0223, "step": 23040 }, { "epoch": 8.464928387807564, "grad_norm": 0.43218058347702026, "learning_rate": 6.0506614821099705e-05, "loss": 0.0232, "step": 23050 }, { "epoch": 8.468600807932427, "grad_norm": 0.33084848523139954, "learning_rate": 6.047428147084153e-05, "loss": 0.0222, "step": 23060 }, { "epoch": 8.47227322805729, "grad_norm": 0.22684340178966522, "learning_rate": 6.044194353877679e-05, "loss": 0.0246, "step": 23070 }, { "epoch": 8.475945648182153, "grad_norm": 0.4021429717540741, "learning_rate": 6.04096010390512e-05, "loss": 0.0233, "step": 23080 }, { "epoch": 8.479618068307014, "grad_norm": 0.3398418724536896, "learning_rate": 6.0377253985812445e-05, "loss": 0.0198, "step": 23090 }, { "epoch": 8.483290488431876, "grad_norm": 0.25414150953292847, "learning_rate": 6.0344902393210235e-05, "loss": 0.0223, "step": 23100 }, { "epoch": 8.48696290855674, "grad_norm": 0.2525053918361664, "learning_rate": 6.031254627539625e-05, "loss": 0.0212, "step": 23110 }, { "epoch": 8.490635328681602, "grad_norm": 0.2593008279800415, "learning_rate": 6.028018564652418e-05, "loss": 0.022, "step": 23120 }, { "epoch": 8.494307748806463, "grad_norm": 0.6220697164535522, "learning_rate": 6.024782052074964e-05, "loss": 0.0209, "step": 23130 }, { "epoch": 8.497980168931326, "grad_norm": 0.21791021525859833, "learning_rate": 6.021545091223026e-05, "loss": 0.0206, "step": 23140 }, { "epoch": 8.501652589056189, "grad_norm": 0.2936094105243683, "learning_rate": 6.0183076835125586e-05, "loss": 0.021, "step": 23150 }, { "epoch": 8.50532500918105, "grad_norm": 0.29949575662612915, "learning_rate": 6.015069830359714e-05, "loss": 0.0208, "step": 23160 }, { "epoch": 8.508997429305913, "grad_norm": 0.3058389127254486, "learning_rate": 6.0118315331808384e-05, "loss": 0.0201, "step": 23170 }, { "epoch": 8.512669849430775, "grad_norm": 0.2164880931377411, "learning_rate": 6.0085927933924745e-05, "loss": 0.0209, "step": 23180 }, { "epoch": 8.516342269555636, "grad_norm": 0.3078678250312805, "learning_rate": 6.0053536124113585e-05, "loss": 0.0197, "step": 23190 }, { "epoch": 8.5200146896805, "grad_norm": 0.2662976384162903, "learning_rate": 6.0021139916544144e-05, "loss": 0.022, "step": 23200 }, { "epoch": 8.523687109805362, "grad_norm": 0.1792815774679184, "learning_rate": 5.998873932538764e-05, "loss": 0.0232, "step": 23210 }, { "epoch": 8.527359529930225, "grad_norm": 0.3469696640968323, "learning_rate": 5.995633436481721e-05, "loss": 0.021, "step": 23220 }, { "epoch": 8.531031950055086, "grad_norm": 0.38666555285453796, "learning_rate": 5.9923925049007855e-05, "loss": 0.0237, "step": 23230 }, { "epoch": 8.534704370179949, "grad_norm": 0.2989456355571747, "learning_rate": 5.989151139213656e-05, "loss": 0.0221, "step": 23240 }, { "epoch": 8.538376790304811, "grad_norm": 0.2695375084877014, "learning_rate": 5.9859093408382094e-05, "loss": 0.0247, "step": 23250 }, { "epoch": 8.542049210429672, "grad_norm": 0.28233951330184937, "learning_rate": 5.982667111192523e-05, "loss": 0.0225, "step": 23260 }, { "epoch": 8.545721630554535, "grad_norm": 0.3792947232723236, "learning_rate": 5.979424451694857e-05, "loss": 0.0204, "step": 23270 }, { "epoch": 8.549394050679398, "grad_norm": 0.2029707133769989, "learning_rate": 5.97618136376366e-05, "loss": 0.0228, "step": 23280 }, { "epoch": 8.55306647080426, "grad_norm": 0.29392915964126587, "learning_rate": 5.97293784881757e-05, "loss": 0.0234, "step": 23290 }, { "epoch": 8.556738890929122, "grad_norm": 0.18912437558174133, "learning_rate": 5.969693908275411e-05, "loss": 0.0205, "step": 23300 }, { "epoch": 8.560411311053985, "grad_norm": 0.23312829434871674, "learning_rate": 5.9664495435561916e-05, "loss": 0.0196, "step": 23310 }, { "epoch": 8.564083731178847, "grad_norm": 0.23886315524578094, "learning_rate": 5.963204756079108e-05, "loss": 0.0211, "step": 23320 }, { "epoch": 8.567756151303708, "grad_norm": 0.20118562877178192, "learning_rate": 5.959959547263538e-05, "loss": 0.021, "step": 23330 }, { "epoch": 8.571428571428571, "grad_norm": 0.3700241148471832, "learning_rate": 5.95671391852905e-05, "loss": 0.0213, "step": 23340 }, { "epoch": 8.575100991553434, "grad_norm": 0.2630314528942108, "learning_rate": 5.95346787129539e-05, "loss": 0.0212, "step": 23350 }, { "epoch": 8.578773411678297, "grad_norm": 0.29674065113067627, "learning_rate": 5.950221406982489e-05, "loss": 0.0184, "step": 23360 }, { "epoch": 8.582445831803158, "grad_norm": 0.23874764144420624, "learning_rate": 5.9469745270104625e-05, "loss": 0.0217, "step": 23370 }, { "epoch": 8.58611825192802, "grad_norm": 0.2552565932273865, "learning_rate": 5.9437272327996064e-05, "loss": 0.0217, "step": 23380 }, { "epoch": 8.589790672052883, "grad_norm": 0.3038073480129242, "learning_rate": 5.940479525770395e-05, "loss": 0.0227, "step": 23390 }, { "epoch": 8.593463092177744, "grad_norm": 0.44706934690475464, "learning_rate": 5.9372314073434876e-05, "loss": 0.0216, "step": 23400 }, { "epoch": 8.597135512302607, "grad_norm": 0.41026243567466736, "learning_rate": 5.93398287893972e-05, "loss": 0.0198, "step": 23410 }, { "epoch": 8.60080793242747, "grad_norm": 0.33435264229774475, "learning_rate": 5.930733941980111e-05, "loss": 0.0206, "step": 23420 }, { "epoch": 8.604480352552333, "grad_norm": 0.324104368686676, "learning_rate": 5.927484597885855e-05, "loss": 0.0212, "step": 23430 }, { "epoch": 8.608152772677194, "grad_norm": 0.1922387331724167, "learning_rate": 5.924234848078325e-05, "loss": 0.0207, "step": 23440 }, { "epoch": 8.611825192802057, "grad_norm": 0.22549094259738922, "learning_rate": 5.9209846939790715e-05, "loss": 0.0203, "step": 23450 }, { "epoch": 8.61549761292692, "grad_norm": 0.2613135278224945, "learning_rate": 5.917734137009824e-05, "loss": 0.0217, "step": 23460 }, { "epoch": 8.61917003305178, "grad_norm": 0.29223528504371643, "learning_rate": 5.914483178592485e-05, "loss": 0.0203, "step": 23470 }, { "epoch": 8.622842453176643, "grad_norm": 0.3479973375797272, "learning_rate": 5.911231820149136e-05, "loss": 0.0217, "step": 23480 }, { "epoch": 8.626514873301506, "grad_norm": 0.26826074719429016, "learning_rate": 5.907980063102031e-05, "loss": 0.022, "step": 23490 }, { "epoch": 8.630187293426367, "grad_norm": 0.3708033859729767, "learning_rate": 5.904727908873597e-05, "loss": 0.02, "step": 23500 }, { "epoch": 8.63385971355123, "grad_norm": 0.20577815175056458, "learning_rate": 5.90147535888644e-05, "loss": 0.0199, "step": 23510 }, { "epoch": 8.637532133676093, "grad_norm": 0.25992634892463684, "learning_rate": 5.898222414563333e-05, "loss": 0.0225, "step": 23520 }, { "epoch": 8.641204553800955, "grad_norm": 0.27004730701446533, "learning_rate": 5.8949690773272256e-05, "loss": 0.0211, "step": 23530 }, { "epoch": 8.644876973925816, "grad_norm": 0.22595427930355072, "learning_rate": 5.891715348601239e-05, "loss": 0.0202, "step": 23540 }, { "epoch": 8.64854939405068, "grad_norm": 0.24601294100284576, "learning_rate": 5.888461229808663e-05, "loss": 0.0212, "step": 23550 }, { "epoch": 8.652221814175542, "grad_norm": 0.24931611120700836, "learning_rate": 5.885206722372959e-05, "loss": 0.0209, "step": 23560 }, { "epoch": 8.655894234300405, "grad_norm": 0.4533168375492096, "learning_rate": 5.881951827717763e-05, "loss": 0.0217, "step": 23570 }, { "epoch": 8.659566654425266, "grad_norm": 0.252851665019989, "learning_rate": 5.878696547266872e-05, "loss": 0.0225, "step": 23580 }, { "epoch": 8.663239074550129, "grad_norm": 0.43297845125198364, "learning_rate": 5.875440882444259e-05, "loss": 0.0209, "step": 23590 }, { "epoch": 8.666911494674991, "grad_norm": 0.23770716786384583, "learning_rate": 5.8721848346740617e-05, "loss": 0.0204, "step": 23600 }, { "epoch": 8.670583914799852, "grad_norm": 0.3319176435470581, "learning_rate": 5.868928405380585e-05, "loss": 0.0216, "step": 23610 }, { "epoch": 8.674256334924715, "grad_norm": 0.3448919653892517, "learning_rate": 5.865671595988304e-05, "loss": 0.0224, "step": 23620 }, { "epoch": 8.677928755049578, "grad_norm": 0.22621499001979828, "learning_rate": 5.862414407921857e-05, "loss": 0.0213, "step": 23630 }, { "epoch": 8.681601175174439, "grad_norm": 0.4871378540992737, "learning_rate": 5.8591568426060475e-05, "loss": 0.0231, "step": 23640 }, { "epoch": 8.685273595299302, "grad_norm": 0.33242088556289673, "learning_rate": 5.8558989014658464e-05, "loss": 0.0203, "step": 23650 }, { "epoch": 8.688946015424165, "grad_norm": 0.4342893362045288, "learning_rate": 5.8526405859263855e-05, "loss": 0.0227, "step": 23660 }, { "epoch": 8.692618435549027, "grad_norm": 0.35137301683425903, "learning_rate": 5.8493818974129667e-05, "loss": 0.022, "step": 23670 }, { "epoch": 8.696290855673888, "grad_norm": 0.28146129846572876, "learning_rate": 5.846122837351048e-05, "loss": 0.0205, "step": 23680 }, { "epoch": 8.699963275798751, "grad_norm": 0.310761421918869, "learning_rate": 5.842863407166255e-05, "loss": 0.0209, "step": 23690 }, { "epoch": 8.703635695923614, "grad_norm": 0.22751587629318237, "learning_rate": 5.83960360828437e-05, "loss": 0.0198, "step": 23700 }, { "epoch": 8.707308116048477, "grad_norm": 0.22768250107765198, "learning_rate": 5.836343442131343e-05, "loss": 0.0201, "step": 23710 }, { "epoch": 8.710980536173338, "grad_norm": 0.3153688311576843, "learning_rate": 5.833082910133281e-05, "loss": 0.0234, "step": 23720 }, { "epoch": 8.7146529562982, "grad_norm": 0.23718152940273285, "learning_rate": 5.82982201371645e-05, "loss": 0.0253, "step": 23730 }, { "epoch": 8.718325376423063, "grad_norm": 0.4390016496181488, "learning_rate": 5.826560754307279e-05, "loss": 0.0203, "step": 23740 }, { "epoch": 8.721997796547925, "grad_norm": 0.2525496184825897, "learning_rate": 5.8232991333323516e-05, "loss": 0.0225, "step": 23750 }, { "epoch": 8.725670216672787, "grad_norm": 0.2356886863708496, "learning_rate": 5.8200371522184116e-05, "loss": 0.0225, "step": 23760 }, { "epoch": 8.72934263679765, "grad_norm": 0.23890654742717743, "learning_rate": 5.816774812392362e-05, "loss": 0.0216, "step": 23770 }, { "epoch": 8.733015056922511, "grad_norm": 0.21368370950222015, "learning_rate": 5.8135121152812614e-05, "loss": 0.0219, "step": 23780 }, { "epoch": 8.736687477047374, "grad_norm": 0.4586635231971741, "learning_rate": 5.8102490623123204e-05, "loss": 0.023, "step": 23790 }, { "epoch": 8.740359897172237, "grad_norm": 0.39429670572280884, "learning_rate": 5.806985654912914e-05, "loss": 0.0197, "step": 23800 }, { "epoch": 8.7440323172971, "grad_norm": 0.297670841217041, "learning_rate": 5.8037218945105644e-05, "loss": 0.0214, "step": 23810 }, { "epoch": 8.74770473742196, "grad_norm": 0.2017538845539093, "learning_rate": 5.800457782532953e-05, "loss": 0.0202, "step": 23820 }, { "epoch": 8.751377157546823, "grad_norm": 0.2566623389720917, "learning_rate": 5.797193320407912e-05, "loss": 0.0226, "step": 23830 }, { "epoch": 8.755049577671686, "grad_norm": 0.33547624945640564, "learning_rate": 5.793928509563431e-05, "loss": 0.0209, "step": 23840 }, { "epoch": 8.758721997796547, "grad_norm": 0.2641615569591522, "learning_rate": 5.790663351427644e-05, "loss": 0.0244, "step": 23850 }, { "epoch": 8.76239441792141, "grad_norm": 0.2577139437198639, "learning_rate": 5.7873978474288484e-05, "loss": 0.0228, "step": 23860 }, { "epoch": 8.766066838046273, "grad_norm": 0.30250027775764465, "learning_rate": 5.78413199899548e-05, "loss": 0.022, "step": 23870 }, { "epoch": 8.769739258171136, "grad_norm": 0.2821713089942932, "learning_rate": 5.780865807556136e-05, "loss": 0.0228, "step": 23880 }, { "epoch": 8.773411678295997, "grad_norm": 0.3239734172821045, "learning_rate": 5.777599274539559e-05, "loss": 0.0232, "step": 23890 }, { "epoch": 8.77708409842086, "grad_norm": 0.29194730520248413, "learning_rate": 5.77433240137464e-05, "loss": 0.0194, "step": 23900 }, { "epoch": 8.780756518545722, "grad_norm": 0.29677170515060425, "learning_rate": 5.771065189490422e-05, "loss": 0.0226, "step": 23910 }, { "epoch": 8.784428938670583, "grad_norm": 0.34870994091033936, "learning_rate": 5.767797640316095e-05, "loss": 0.0229, "step": 23920 }, { "epoch": 8.788101358795446, "grad_norm": 0.2059013694524765, "learning_rate": 5.764529755280994e-05, "loss": 0.0199, "step": 23930 }, { "epoch": 8.791773778920309, "grad_norm": 0.22742438316345215, "learning_rate": 5.761261535814603e-05, "loss": 0.0205, "step": 23940 }, { "epoch": 8.795446199045172, "grad_norm": 0.28239721059799194, "learning_rate": 5.757992983346554e-05, "loss": 0.0217, "step": 23950 }, { "epoch": 8.799118619170033, "grad_norm": 0.25425079464912415, "learning_rate": 5.754724099306621e-05, "loss": 0.0211, "step": 23960 }, { "epoch": 8.802791039294895, "grad_norm": 0.26586371660232544, "learning_rate": 5.751454885124727e-05, "loss": 0.0198, "step": 23970 }, { "epoch": 8.806463459419758, "grad_norm": 0.27789855003356934, "learning_rate": 5.748185342230935e-05, "loss": 0.0217, "step": 23980 }, { "epoch": 8.81013587954462, "grad_norm": 0.3087098002433777, "learning_rate": 5.744915472055456e-05, "loss": 0.0218, "step": 23990 }, { "epoch": 8.813808299669482, "grad_norm": 0.30610495805740356, "learning_rate": 5.741645276028642e-05, "loss": 0.0218, "step": 24000 }, { "epoch": 8.817480719794345, "grad_norm": 0.3658243715763092, "learning_rate": 5.7383747555809873e-05, "loss": 0.0208, "step": 24010 }, { "epoch": 8.821153139919208, "grad_norm": 0.2873465418815613, "learning_rate": 5.7351039121431294e-05, "loss": 0.0221, "step": 24020 }, { "epoch": 8.824825560044069, "grad_norm": 0.2562883496284485, "learning_rate": 5.7318327471458454e-05, "loss": 0.0214, "step": 24030 }, { "epoch": 8.828497980168931, "grad_norm": 0.3090607821941376, "learning_rate": 5.728561262020054e-05, "loss": 0.0213, "step": 24040 }, { "epoch": 8.832170400293794, "grad_norm": 0.39869314432144165, "learning_rate": 5.725289458196814e-05, "loss": 0.0206, "step": 24050 }, { "epoch": 8.835842820418655, "grad_norm": 0.2934393286705017, "learning_rate": 5.722017337107324e-05, "loss": 0.0214, "step": 24060 }, { "epoch": 8.839515240543518, "grad_norm": 0.27143633365631104, "learning_rate": 5.7187449001829206e-05, "loss": 0.0222, "step": 24070 }, { "epoch": 8.84318766066838, "grad_norm": 0.25234031677246094, "learning_rate": 5.715472148855081e-05, "loss": 0.02, "step": 24080 }, { "epoch": 8.846860080793242, "grad_norm": 0.3015248775482178, "learning_rate": 5.712199084555414e-05, "loss": 0.0211, "step": 24090 }, { "epoch": 8.850532500918105, "grad_norm": 0.3809743821620941, "learning_rate": 5.7089257087156734e-05, "loss": 0.02, "step": 24100 }, { "epoch": 8.854204921042967, "grad_norm": 0.3187362253665924, "learning_rate": 5.70565202276774e-05, "loss": 0.0205, "step": 24110 }, { "epoch": 8.85787734116783, "grad_norm": 0.2382960319519043, "learning_rate": 5.70237802814364e-05, "loss": 0.0217, "step": 24120 }, { "epoch": 8.861549761292691, "grad_norm": 0.28791624307632446, "learning_rate": 5.6991037262755277e-05, "loss": 0.0203, "step": 24130 }, { "epoch": 8.865222181417554, "grad_norm": 0.18645505607128143, "learning_rate": 5.6958291185956924e-05, "loss": 0.02, "step": 24140 }, { "epoch": 8.868894601542417, "grad_norm": 0.17993779480457306, "learning_rate": 5.692554206536561e-05, "loss": 0.0199, "step": 24150 }, { "epoch": 8.87256702166728, "grad_norm": 0.2336183488368988, "learning_rate": 5.689278991530692e-05, "loss": 0.0195, "step": 24160 }, { "epoch": 8.87623944179214, "grad_norm": 0.2602829337120056, "learning_rate": 5.686003475010773e-05, "loss": 0.0193, "step": 24170 }, { "epoch": 8.879911861917003, "grad_norm": 0.25139978528022766, "learning_rate": 5.682727658409628e-05, "loss": 0.0197, "step": 24180 }, { "epoch": 8.883584282041866, "grad_norm": 0.2695843577384949, "learning_rate": 5.67945154316021e-05, "loss": 0.0211, "step": 24190 }, { "epoch": 8.887256702166727, "grad_norm": 0.4636688828468323, "learning_rate": 5.676175130695602e-05, "loss": 0.0184, "step": 24200 }, { "epoch": 8.89092912229159, "grad_norm": 0.41085657477378845, "learning_rate": 5.672898422449022e-05, "loss": 0.0219, "step": 24210 }, { "epoch": 8.894601542416453, "grad_norm": 0.3450271189212799, "learning_rate": 5.66962141985381e-05, "loss": 0.0203, "step": 24220 }, { "epoch": 8.898273962541314, "grad_norm": 0.2249874770641327, "learning_rate": 5.666344124343439e-05, "loss": 0.0206, "step": 24230 }, { "epoch": 8.901946382666177, "grad_norm": 0.27293264865875244, "learning_rate": 5.66306653735151e-05, "loss": 0.0217, "step": 24240 }, { "epoch": 8.90561880279104, "grad_norm": 0.36036962270736694, "learning_rate": 5.659788660311749e-05, "loss": 0.0193, "step": 24250 }, { "epoch": 8.909291222915902, "grad_norm": 0.479300856590271, "learning_rate": 5.656510494658014e-05, "loss": 0.0213, "step": 24260 }, { "epoch": 8.912963643040763, "grad_norm": 0.3312857747077942, "learning_rate": 5.6532320418242835e-05, "loss": 0.0206, "step": 24270 }, { "epoch": 8.916636063165626, "grad_norm": 0.2836154103279114, "learning_rate": 5.649953303244665e-05, "loss": 0.0217, "step": 24280 }, { "epoch": 8.920308483290489, "grad_norm": 0.3876383304595947, "learning_rate": 5.646674280353389e-05, "loss": 0.0197, "step": 24290 }, { "epoch": 8.92398090341535, "grad_norm": 0.37179481983184814, "learning_rate": 5.6433949745848135e-05, "loss": 0.0214, "step": 24300 }, { "epoch": 8.927653323540213, "grad_norm": 0.37350448966026306, "learning_rate": 5.6401153873734166e-05, "loss": 0.0201, "step": 24310 }, { "epoch": 8.931325743665075, "grad_norm": 0.3587366044521332, "learning_rate": 5.6368355201538036e-05, "loss": 0.0237, "step": 24320 }, { "epoch": 8.934998163789938, "grad_norm": 0.3167526423931122, "learning_rate": 5.633555374360697e-05, "loss": 0.0215, "step": 24330 }, { "epoch": 8.9386705839148, "grad_norm": 0.2097128927707672, "learning_rate": 5.630274951428944e-05, "loss": 0.0189, "step": 24340 }, { "epoch": 8.942343004039662, "grad_norm": 0.15032558143138885, "learning_rate": 5.626994252793517e-05, "loss": 0.0199, "step": 24350 }, { "epoch": 8.946015424164525, "grad_norm": 0.3290087878704071, "learning_rate": 5.6237132798895006e-05, "loss": 0.0202, "step": 24360 }, { "epoch": 8.949687844289386, "grad_norm": 0.3506488502025604, "learning_rate": 5.620432034152106e-05, "loss": 0.018, "step": 24370 }, { "epoch": 8.953360264414249, "grad_norm": 0.6044869422912598, "learning_rate": 5.6171505170166604e-05, "loss": 0.0206, "step": 24380 }, { "epoch": 8.957032684539112, "grad_norm": 0.2402845025062561, "learning_rate": 5.613868729918612e-05, "loss": 0.0235, "step": 24390 }, { "epoch": 8.960705104663974, "grad_norm": 0.6367641091346741, "learning_rate": 5.6105866742935256e-05, "loss": 0.0213, "step": 24400 }, { "epoch": 8.964377524788835, "grad_norm": 0.29127413034439087, "learning_rate": 5.607304351577083e-05, "loss": 0.0206, "step": 24410 }, { "epoch": 8.968049944913698, "grad_norm": 0.4302724599838257, "learning_rate": 5.604021763205085e-05, "loss": 0.0242, "step": 24420 }, { "epoch": 8.971722365038561, "grad_norm": 0.2874281704425812, "learning_rate": 5.600738910613445e-05, "loss": 0.0204, "step": 24430 }, { "epoch": 8.975394785163422, "grad_norm": 0.23400433361530304, "learning_rate": 5.597455795238198e-05, "loss": 0.0194, "step": 24440 }, { "epoch": 8.979067205288285, "grad_norm": 0.46693935990333557, "learning_rate": 5.5941724185154876e-05, "loss": 0.0212, "step": 24450 }, { "epoch": 8.982739625413148, "grad_norm": 0.301058292388916, "learning_rate": 5.5908887818815746e-05, "loss": 0.0197, "step": 24460 }, { "epoch": 8.98641204553801, "grad_norm": 0.32560279965400696, "learning_rate": 5.587604886772834e-05, "loss": 0.0237, "step": 24470 }, { "epoch": 8.990084465662871, "grad_norm": 0.19776298105716705, "learning_rate": 5.584320734625752e-05, "loss": 0.0193, "step": 24480 }, { "epoch": 8.993756885787734, "grad_norm": 0.326419472694397, "learning_rate": 5.581036326876928e-05, "loss": 0.0199, "step": 24490 }, { "epoch": 8.997429305912597, "grad_norm": 0.26663631200790405, "learning_rate": 5.5777516649630756e-05, "loss": 0.0195, "step": 24500 }, { "epoch": 9.001101726037458, "grad_norm": 0.5080954432487488, "learning_rate": 5.574466750321017e-05, "loss": 0.0202, "step": 24510 }, { "epoch": 9.00477414616232, "grad_norm": 0.2545473873615265, "learning_rate": 5.571181584387683e-05, "loss": 0.0209, "step": 24520 }, { "epoch": 9.008446566287184, "grad_norm": 0.21688315272331238, "learning_rate": 5.5678961686001196e-05, "loss": 0.021, "step": 24530 }, { "epoch": 9.012118986412046, "grad_norm": 0.2758161425590515, "learning_rate": 5.5646105043954785e-05, "loss": 0.0206, "step": 24540 }, { "epoch": 9.015791406536907, "grad_norm": 0.270038366317749, "learning_rate": 5.5613245932110205e-05, "loss": 0.0192, "step": 24550 }, { "epoch": 9.01946382666177, "grad_norm": 0.5274330377578735, "learning_rate": 5.5580384364841157e-05, "loss": 0.0198, "step": 24560 }, { "epoch": 9.023136246786633, "grad_norm": 0.4133802652359009, "learning_rate": 5.554752035652237e-05, "loss": 0.0202, "step": 24570 }, { "epoch": 9.026808666911494, "grad_norm": 0.3794413208961487, "learning_rate": 5.5514653921529736e-05, "loss": 0.018, "step": 24580 }, { "epoch": 9.030481087036357, "grad_norm": 0.21639277040958405, "learning_rate": 5.5481785074240114e-05, "loss": 0.0222, "step": 24590 }, { "epoch": 9.03415350716122, "grad_norm": 0.3661118447780609, "learning_rate": 5.544891382903146e-05, "loss": 0.0216, "step": 24600 }, { "epoch": 9.037825927286082, "grad_norm": 0.5167155265808105, "learning_rate": 5.541604020028277e-05, "loss": 0.0219, "step": 24610 }, { "epoch": 9.041498347410943, "grad_norm": 0.2936228811740875, "learning_rate": 5.5383164202374105e-05, "loss": 0.0214, "step": 24620 }, { "epoch": 9.045170767535806, "grad_norm": 0.2691381871700287, "learning_rate": 5.5350285849686515e-05, "loss": 0.0204, "step": 24630 }, { "epoch": 9.048843187660669, "grad_norm": 0.2958510220050812, "learning_rate": 5.531740515660213e-05, "loss": 0.0193, "step": 24640 }, { "epoch": 9.05251560778553, "grad_norm": 0.24733902513980865, "learning_rate": 5.528452213750408e-05, "loss": 0.0208, "step": 24650 }, { "epoch": 9.056188027910393, "grad_norm": 0.23016498982906342, "learning_rate": 5.52516368067765e-05, "loss": 0.019, "step": 24660 }, { "epoch": 9.059860448035256, "grad_norm": 0.20219121873378754, "learning_rate": 5.5218749178804563e-05, "loss": 0.0199, "step": 24670 }, { "epoch": 9.063532868160118, "grad_norm": 0.31176766753196716, "learning_rate": 5.518585926797443e-05, "loss": 0.0211, "step": 24680 }, { "epoch": 9.06720528828498, "grad_norm": 0.35760387778282166, "learning_rate": 5.5152967088673265e-05, "loss": 0.0224, "step": 24690 }, { "epoch": 9.070877708409842, "grad_norm": 0.22837099432945251, "learning_rate": 5.512007265528924e-05, "loss": 0.0205, "step": 24700 }, { "epoch": 9.074550128534705, "grad_norm": 0.3384171426296234, "learning_rate": 5.508717598221148e-05, "loss": 0.0192, "step": 24710 }, { "epoch": 9.078222548659566, "grad_norm": 0.2688662111759186, "learning_rate": 5.505427708383012e-05, "loss": 0.0224, "step": 24720 }, { "epoch": 9.081894968784429, "grad_norm": 0.2550411522388458, "learning_rate": 5.502137597453624e-05, "loss": 0.0195, "step": 24730 }, { "epoch": 9.085567388909292, "grad_norm": 0.21340779960155487, "learning_rate": 5.4988472668721914e-05, "loss": 0.0197, "step": 24740 }, { "epoch": 9.089239809034153, "grad_norm": 0.365145742893219, "learning_rate": 5.4955567180780164e-05, "loss": 0.02, "step": 24750 }, { "epoch": 9.092912229159015, "grad_norm": 0.32367151975631714, "learning_rate": 5.492265952510497e-05, "loss": 0.0189, "step": 24760 }, { "epoch": 9.096584649283878, "grad_norm": 0.2544712722301483, "learning_rate": 5.4889749716091254e-05, "loss": 0.019, "step": 24770 }, { "epoch": 9.100257069408741, "grad_norm": 0.45618513226509094, "learning_rate": 5.4856837768134885e-05, "loss": 0.0202, "step": 24780 }, { "epoch": 9.103929489533602, "grad_norm": 0.2742355465888977, "learning_rate": 5.482392369563267e-05, "loss": 0.0198, "step": 24790 }, { "epoch": 9.107601909658465, "grad_norm": 0.3458285629749298, "learning_rate": 5.4791007512982354e-05, "loss": 0.02, "step": 24800 }, { "epoch": 9.111274329783328, "grad_norm": 0.26018816232681274, "learning_rate": 5.4758089234582566e-05, "loss": 0.0189, "step": 24810 }, { "epoch": 9.114946749908189, "grad_norm": 0.25516214966773987, "learning_rate": 5.47251688748329e-05, "loss": 0.0196, "step": 24820 }, { "epoch": 9.118619170033051, "grad_norm": 0.2412457913160324, "learning_rate": 5.4692246448133843e-05, "loss": 0.0201, "step": 24830 }, { "epoch": 9.122291590157914, "grad_norm": 0.21462798118591309, "learning_rate": 5.465932196888676e-05, "loss": 0.0195, "step": 24840 }, { "epoch": 9.125964010282777, "grad_norm": 0.20324991643428802, "learning_rate": 5.462639545149395e-05, "loss": 0.0209, "step": 24850 }, { "epoch": 9.129636430407638, "grad_norm": 0.2637501060962677, "learning_rate": 5.459346691035861e-05, "loss": 0.0208, "step": 24860 }, { "epoch": 9.1333088505325, "grad_norm": 0.2729634642601013, "learning_rate": 5.4560536359884776e-05, "loss": 0.0175, "step": 24870 }, { "epoch": 9.136981270657364, "grad_norm": 0.27392005920410156, "learning_rate": 5.45276038144774e-05, "loss": 0.0212, "step": 24880 }, { "epoch": 9.140653690782225, "grad_norm": 0.3111582100391388, "learning_rate": 5.449466928854229e-05, "loss": 0.0207, "step": 24890 }, { "epoch": 9.144326110907087, "grad_norm": 0.29444774985313416, "learning_rate": 5.446173279648613e-05, "loss": 0.0204, "step": 24900 }, { "epoch": 9.14799853103195, "grad_norm": 0.4614615738391876, "learning_rate": 5.442879435271646e-05, "loss": 0.0189, "step": 24910 }, { "epoch": 9.151670951156813, "grad_norm": 0.3315463662147522, "learning_rate": 5.4395853971641675e-05, "loss": 0.0189, "step": 24920 }, { "epoch": 9.155343371281674, "grad_norm": 0.2162683755159378, "learning_rate": 5.4362911667671015e-05, "loss": 0.0195, "step": 24930 }, { "epoch": 9.159015791406537, "grad_norm": 0.24691951274871826, "learning_rate": 5.432996745521458e-05, "loss": 0.0211, "step": 24940 }, { "epoch": 9.1626882115314, "grad_norm": 0.27696219086647034, "learning_rate": 5.4297021348683264e-05, "loss": 0.0188, "step": 24950 }, { "epoch": 9.16636063165626, "grad_norm": 0.22690629959106445, "learning_rate": 5.426407336248882e-05, "loss": 0.0192, "step": 24960 }, { "epoch": 9.170033051781123, "grad_norm": 0.2376561015844345, "learning_rate": 5.423112351104382e-05, "loss": 0.0224, "step": 24970 }, { "epoch": 9.173705471905986, "grad_norm": 0.2512870728969574, "learning_rate": 5.4198171808761654e-05, "loss": 0.02, "step": 24980 }, { "epoch": 9.177377892030849, "grad_norm": 0.18819032609462738, "learning_rate": 5.41652182700565e-05, "loss": 0.019, "step": 24990 }, { "epoch": 9.18105031215571, "grad_norm": 0.21973921358585358, "learning_rate": 5.413226290934338e-05, "loss": 0.0202, "step": 25000 }, { "epoch": 9.184722732280573, "grad_norm": 0.252090185880661, "learning_rate": 5.4099305741038056e-05, "loss": 0.0228, "step": 25010 }, { "epoch": 9.188395152405436, "grad_norm": 0.48793134093284607, "learning_rate": 5.406634677955713e-05, "loss": 0.0199, "step": 25020 }, { "epoch": 9.192067572530297, "grad_norm": 0.22793832421302795, "learning_rate": 5.403338603931798e-05, "loss": 0.0198, "step": 25030 }, { "epoch": 9.19573999265516, "grad_norm": 0.2642729878425598, "learning_rate": 5.4000423534738745e-05, "loss": 0.018, "step": 25040 }, { "epoch": 9.199412412780022, "grad_norm": 0.389852911233902, "learning_rate": 5.396745928023835e-05, "loss": 0.0208, "step": 25050 }, { "epoch": 9.203084832904885, "grad_norm": 0.6172726154327393, "learning_rate": 5.3934493290236474e-05, "loss": 0.0202, "step": 25060 }, { "epoch": 9.206757253029746, "grad_norm": 0.2510662376880646, "learning_rate": 5.390152557915357e-05, "loss": 0.0201, "step": 25070 }, { "epoch": 9.210429673154609, "grad_norm": 0.35934996604919434, "learning_rate": 5.3868556161410796e-05, "loss": 0.0197, "step": 25080 }, { "epoch": 9.214102093279472, "grad_norm": 0.3355807065963745, "learning_rate": 5.383558505143015e-05, "loss": 0.0192, "step": 25090 }, { "epoch": 9.217774513404333, "grad_norm": 0.3174123764038086, "learning_rate": 5.38026122636343e-05, "loss": 0.0209, "step": 25100 }, { "epoch": 9.221446933529196, "grad_norm": 0.24266518652439117, "learning_rate": 5.3769637812446636e-05, "loss": 0.0215, "step": 25110 }, { "epoch": 9.225119353654058, "grad_norm": 0.3607197701931, "learning_rate": 5.373666171229133e-05, "loss": 0.0213, "step": 25120 }, { "epoch": 9.228791773778921, "grad_norm": 0.2315870076417923, "learning_rate": 5.370368397759324e-05, "loss": 0.0183, "step": 25130 }, { "epoch": 9.232464193903782, "grad_norm": 0.38699090480804443, "learning_rate": 5.3670704622777944e-05, "loss": 0.0221, "step": 25140 }, { "epoch": 9.236136614028645, "grad_norm": 0.23462806642055511, "learning_rate": 5.3637723662271744e-05, "loss": 0.0221, "step": 25150 }, { "epoch": 9.239809034153508, "grad_norm": 0.21523568034172058, "learning_rate": 5.360474111050162e-05, "loss": 0.0209, "step": 25160 }, { "epoch": 9.243481454278369, "grad_norm": 0.24683761596679688, "learning_rate": 5.357175698189527e-05, "loss": 0.02, "step": 25170 }, { "epoch": 9.247153874403232, "grad_norm": 0.39213529229164124, "learning_rate": 5.353877129088107e-05, "loss": 0.0189, "step": 25180 }, { "epoch": 9.250826294528094, "grad_norm": 0.2655245363712311, "learning_rate": 5.350578405188809e-05, "loss": 0.0194, "step": 25190 }, { "epoch": 9.254498714652957, "grad_norm": 0.3413146436214447, "learning_rate": 5.347279527934603e-05, "loss": 0.021, "step": 25200 }, { "epoch": 9.258171134777818, "grad_norm": 0.31547337770462036, "learning_rate": 5.343980498768536e-05, "loss": 0.0178, "step": 25210 }, { "epoch": 9.261843554902681, "grad_norm": 0.2581675052642822, "learning_rate": 5.340681319133711e-05, "loss": 0.0185, "step": 25220 }, { "epoch": 9.265515975027544, "grad_norm": 0.2761007845401764, "learning_rate": 5.337381990473305e-05, "loss": 0.0195, "step": 25230 }, { "epoch": 9.269188395152405, "grad_norm": 0.36455559730529785, "learning_rate": 5.3340825142305504e-05, "loss": 0.0189, "step": 25240 }, { "epoch": 9.272860815277268, "grad_norm": 0.32733970880508423, "learning_rate": 5.330782891848756e-05, "loss": 0.0208, "step": 25250 }, { "epoch": 9.27653323540213, "grad_norm": 0.33040568232536316, "learning_rate": 5.327483124771285e-05, "loss": 0.018, "step": 25260 }, { "epoch": 9.280205655526991, "grad_norm": 0.273468554019928, "learning_rate": 5.3241832144415706e-05, "loss": 0.0208, "step": 25270 }, { "epoch": 9.283878075651854, "grad_norm": 0.2774815857410431, "learning_rate": 5.320883162303104e-05, "loss": 0.0202, "step": 25280 }, { "epoch": 9.287550495776717, "grad_norm": 0.2728939652442932, "learning_rate": 5.3175829697994416e-05, "loss": 0.019, "step": 25290 }, { "epoch": 9.29122291590158, "grad_norm": 0.40283146500587463, "learning_rate": 5.3142826383741975e-05, "loss": 0.0202, "step": 25300 }, { "epoch": 9.29489533602644, "grad_norm": 0.3411037027835846, "learning_rate": 5.3109821694710506e-05, "loss": 0.0195, "step": 25310 }, { "epoch": 9.298567756151304, "grad_norm": 0.7733019590377808, "learning_rate": 5.307681564533736e-05, "loss": 0.0202, "step": 25320 }, { "epoch": 9.302240176276166, "grad_norm": 0.3102809488773346, "learning_rate": 5.3043808250060535e-05, "loss": 0.0222, "step": 25330 }, { "epoch": 9.30591259640103, "grad_norm": 0.2978496849536896, "learning_rate": 5.3010799523318566e-05, "loss": 0.0203, "step": 25340 }, { "epoch": 9.30958501652589, "grad_norm": 0.2556949853897095, "learning_rate": 5.29777894795506e-05, "loss": 0.0209, "step": 25350 }, { "epoch": 9.313257436650753, "grad_norm": 0.21390381455421448, "learning_rate": 5.294477813319634e-05, "loss": 0.0203, "step": 25360 }, { "epoch": 9.316929856775616, "grad_norm": 0.16539081931114197, "learning_rate": 5.291176549869608e-05, "loss": 0.0222, "step": 25370 }, { "epoch": 9.320602276900477, "grad_norm": 0.20727470517158508, "learning_rate": 5.287875159049065e-05, "loss": 0.0205, "step": 25380 }, { "epoch": 9.32427469702534, "grad_norm": 0.335224986076355, "learning_rate": 5.284573642302149e-05, "loss": 0.0206, "step": 25390 }, { "epoch": 9.327947117150202, "grad_norm": 0.3934212028980255, "learning_rate": 5.2812720010730543e-05, "loss": 0.0182, "step": 25400 }, { "epoch": 9.331619537275063, "grad_norm": 0.340194970369339, "learning_rate": 5.277970236806029e-05, "loss": 0.019, "step": 25410 }, { "epoch": 9.335291957399926, "grad_norm": 0.46698781847953796, "learning_rate": 5.2746683509453795e-05, "loss": 0.0201, "step": 25420 }, { "epoch": 9.338964377524789, "grad_norm": 0.28705939650535583, "learning_rate": 5.271366344935461e-05, "loss": 0.0204, "step": 25430 }, { "epoch": 9.342636797649652, "grad_norm": 0.21002072095870972, "learning_rate": 5.268064220220683e-05, "loss": 0.0219, "step": 25440 }, { "epoch": 9.346309217774513, "grad_norm": 0.33719968795776367, "learning_rate": 5.2647619782455094e-05, "loss": 0.021, "step": 25450 }, { "epoch": 9.349981637899376, "grad_norm": 0.3174508213996887, "learning_rate": 5.261459620454451e-05, "loss": 0.0216, "step": 25460 }, { "epoch": 9.353654058024238, "grad_norm": 0.3475327491760254, "learning_rate": 5.258157148292071e-05, "loss": 0.0199, "step": 25470 }, { "epoch": 9.3573264781491, "grad_norm": 0.3417593240737915, "learning_rate": 5.2548545632029846e-05, "loss": 0.0199, "step": 25480 }, { "epoch": 9.360998898273962, "grad_norm": 0.44467559456825256, "learning_rate": 5.2515518666318534e-05, "loss": 0.0197, "step": 25490 }, { "epoch": 9.364671318398825, "grad_norm": 0.21415676176548004, "learning_rate": 5.248249060023389e-05, "loss": 0.0203, "step": 25500 }, { "epoch": 9.368343738523688, "grad_norm": 0.33646389842033386, "learning_rate": 5.2449461448223517e-05, "loss": 0.0202, "step": 25510 }, { "epoch": 9.372016158648549, "grad_norm": 0.2814105153083801, "learning_rate": 5.2416431224735486e-05, "loss": 0.0209, "step": 25520 }, { "epoch": 9.375688578773412, "grad_norm": 0.1938990354537964, "learning_rate": 5.238339994421836e-05, "loss": 0.022, "step": 25530 }, { "epoch": 9.379360998898274, "grad_norm": 0.40154242515563965, "learning_rate": 5.235036762112111e-05, "loss": 0.0211, "step": 25540 }, { "epoch": 9.383033419023135, "grad_norm": 0.3178681433200836, "learning_rate": 5.2317334269893206e-05, "loss": 0.0191, "step": 25550 }, { "epoch": 9.386705839147998, "grad_norm": 0.29764115810394287, "learning_rate": 5.228429990498458e-05, "loss": 0.0203, "step": 25560 }, { "epoch": 9.390378259272861, "grad_norm": 0.24724306166172028, "learning_rate": 5.225126454084556e-05, "loss": 0.0222, "step": 25570 }, { "epoch": 9.394050679397724, "grad_norm": 0.23831430077552795, "learning_rate": 5.221822819192694e-05, "loss": 0.0192, "step": 25580 }, { "epoch": 9.397723099522585, "grad_norm": 0.29261648654937744, "learning_rate": 5.218519087267995e-05, "loss": 0.0196, "step": 25590 }, { "epoch": 9.401395519647448, "grad_norm": 0.20895501971244812, "learning_rate": 5.2152152597556225e-05, "loss": 0.0195, "step": 25600 }, { "epoch": 9.40506793977231, "grad_norm": 0.2901381254196167, "learning_rate": 5.211911338100784e-05, "loss": 0.0201, "step": 25610 }, { "epoch": 9.408740359897172, "grad_norm": 0.19661365449428558, "learning_rate": 5.2086073237487264e-05, "loss": 0.0192, "step": 25620 }, { "epoch": 9.412412780022034, "grad_norm": 0.2744089663028717, "learning_rate": 5.205303218144739e-05, "loss": 0.0206, "step": 25630 }, { "epoch": 9.416085200146897, "grad_norm": 0.48694053292274475, "learning_rate": 5.2019990227341494e-05, "loss": 0.0203, "step": 25640 }, { "epoch": 9.41975762027176, "grad_norm": 0.3656236231327057, "learning_rate": 5.198694738962324e-05, "loss": 0.0203, "step": 25650 }, { "epoch": 9.423430040396621, "grad_norm": 0.5480469465255737, "learning_rate": 5.19539036827467e-05, "loss": 0.0198, "step": 25660 }, { "epoch": 9.427102460521484, "grad_norm": 0.38665977120399475, "learning_rate": 5.192085912116629e-05, "loss": 0.0206, "step": 25670 }, { "epoch": 9.430774880646347, "grad_norm": 0.2154540717601776, "learning_rate": 5.188781371933685e-05, "loss": 0.0206, "step": 25680 }, { "epoch": 9.434447300771208, "grad_norm": 0.40523627400398254, "learning_rate": 5.185476749171356e-05, "loss": 0.0212, "step": 25690 }, { "epoch": 9.43811972089607, "grad_norm": 0.18971672654151917, "learning_rate": 5.1821720452751945e-05, "loss": 0.0216, "step": 25700 }, { "epoch": 9.441792141020933, "grad_norm": 0.3119276463985443, "learning_rate": 5.178867261690791e-05, "loss": 0.0193, "step": 25710 }, { "epoch": 9.445464561145796, "grad_norm": 0.21814584732055664, "learning_rate": 5.1755623998637715e-05, "loss": 0.0193, "step": 25720 }, { "epoch": 9.449136981270657, "grad_norm": 0.29684045910835266, "learning_rate": 5.172257461239794e-05, "loss": 0.0209, "step": 25730 }, { "epoch": 9.45280940139552, "grad_norm": 0.2594817876815796, "learning_rate": 5.168952447264548e-05, "loss": 0.0181, "step": 25740 }, { "epoch": 9.456481821520383, "grad_norm": 0.36297371983528137, "learning_rate": 5.165647359383763e-05, "loss": 0.018, "step": 25750 }, { "epoch": 9.460154241645244, "grad_norm": 0.2346714586019516, "learning_rate": 5.1623421990431954e-05, "loss": 0.0201, "step": 25760 }, { "epoch": 9.463826661770106, "grad_norm": 0.18134082853794098, "learning_rate": 5.159036967688635e-05, "loss": 0.018, "step": 25770 }, { "epoch": 9.46749908189497, "grad_norm": 0.4358888566493988, "learning_rate": 5.1557316667659015e-05, "loss": 0.0228, "step": 25780 }, { "epoch": 9.471171502019832, "grad_norm": 0.24110078811645508, "learning_rate": 5.152426297720846e-05, "loss": 0.0186, "step": 25790 }, { "epoch": 9.474843922144693, "grad_norm": 0.3054088354110718, "learning_rate": 5.1491208619993505e-05, "loss": 0.0206, "step": 25800 }, { "epoch": 9.478516342269556, "grad_norm": 0.24741823971271515, "learning_rate": 5.145815361047321e-05, "loss": 0.0179, "step": 25810 }, { "epoch": 9.482188762394419, "grad_norm": 0.31639012694358826, "learning_rate": 5.1425097963107003e-05, "loss": 0.0186, "step": 25820 }, { "epoch": 9.48586118251928, "grad_norm": 0.2452494502067566, "learning_rate": 5.1392041692354544e-05, "loss": 0.0199, "step": 25830 }, { "epoch": 9.489533602644142, "grad_norm": 0.26967060565948486, "learning_rate": 5.1358984812675736e-05, "loss": 0.0213, "step": 25840 }, { "epoch": 9.493206022769005, "grad_norm": 0.22898012399673462, "learning_rate": 5.13259273385308e-05, "loss": 0.0205, "step": 25850 }, { "epoch": 9.496878442893866, "grad_norm": 0.28029099106788635, "learning_rate": 5.12928692843802e-05, "loss": 0.0192, "step": 25860 }, { "epoch": 9.500550863018729, "grad_norm": 0.21216775476932526, "learning_rate": 5.125981066468465e-05, "loss": 0.0191, "step": 25870 }, { "epoch": 9.504223283143592, "grad_norm": 0.22466832399368286, "learning_rate": 5.122675149390514e-05, "loss": 0.0196, "step": 25880 }, { "epoch": 9.507895703268455, "grad_norm": 0.36697179079055786, "learning_rate": 5.119369178650282e-05, "loss": 0.0197, "step": 25890 }, { "epoch": 9.511568123393316, "grad_norm": 0.42883169651031494, "learning_rate": 5.1160631556939156e-05, "loss": 0.019, "step": 25900 }, { "epoch": 9.515240543518178, "grad_norm": 0.3379577696323395, "learning_rate": 5.112757081967584e-05, "loss": 0.0191, "step": 25910 }, { "epoch": 9.518912963643041, "grad_norm": 0.2560490369796753, "learning_rate": 5.109450958917472e-05, "loss": 0.0202, "step": 25920 }, { "epoch": 9.522585383767904, "grad_norm": 0.171842560172081, "learning_rate": 5.106144787989794e-05, "loss": 0.0197, "step": 25930 }, { "epoch": 9.526257803892765, "grad_norm": 0.32757270336151123, "learning_rate": 5.102838570630778e-05, "loss": 0.0204, "step": 25940 }, { "epoch": 9.529930224017628, "grad_norm": 0.4419039785861969, "learning_rate": 5.099532308286678e-05, "loss": 0.0198, "step": 25950 }, { "epoch": 9.53360264414249, "grad_norm": 0.22048123180866241, "learning_rate": 5.0962260024037645e-05, "loss": 0.0192, "step": 25960 }, { "epoch": 9.537275064267352, "grad_norm": 0.22219276428222656, "learning_rate": 5.0929196544283276e-05, "loss": 0.0213, "step": 25970 }, { "epoch": 9.540947484392214, "grad_norm": 0.27537015080451965, "learning_rate": 5.089613265806678e-05, "loss": 0.0181, "step": 25980 }, { "epoch": 9.544619904517077, "grad_norm": 0.2206609547138214, "learning_rate": 5.086306837985141e-05, "loss": 0.0209, "step": 25990 }, { "epoch": 9.548292324641938, "grad_norm": 0.27892428636550903, "learning_rate": 5.0830003724100604e-05, "loss": 0.0211, "step": 26000 }, { "epoch": 9.551964744766801, "grad_norm": 0.2572023570537567, "learning_rate": 5.0796938705277984e-05, "loss": 0.0193, "step": 26010 }, { "epoch": 9.555637164891664, "grad_norm": 0.3462391197681427, "learning_rate": 5.076387333784729e-05, "loss": 0.0187, "step": 26020 }, { "epoch": 9.559309585016527, "grad_norm": 0.2813008725643158, "learning_rate": 5.0730807636272445e-05, "loss": 0.0217, "step": 26030 }, { "epoch": 9.562982005141388, "grad_norm": 0.2183527797460556, "learning_rate": 5.0697741615017524e-05, "loss": 0.0198, "step": 26040 }, { "epoch": 9.56665442526625, "grad_norm": 0.23335257172584534, "learning_rate": 5.06646752885467e-05, "loss": 0.0198, "step": 26050 }, { "epoch": 9.570326845391113, "grad_norm": 0.2782585024833679, "learning_rate": 5.0631608671324324e-05, "loss": 0.0224, "step": 26060 }, { "epoch": 9.573999265515974, "grad_norm": 0.28420060873031616, "learning_rate": 5.059854177781487e-05, "loss": 0.0174, "step": 26070 }, { "epoch": 9.577671685640837, "grad_norm": 0.3724050223827362, "learning_rate": 5.05654746224829e-05, "loss": 0.021, "step": 26080 }, { "epoch": 9.5813441057657, "grad_norm": 0.241680309176445, "learning_rate": 5.053240721979311e-05, "loss": 0.0201, "step": 26090 }, { "epoch": 9.585016525890563, "grad_norm": 0.24949222803115845, "learning_rate": 5.049933958421031e-05, "loss": 0.0229, "step": 26100 }, { "epoch": 9.588688946015424, "grad_norm": 0.41031089425086975, "learning_rate": 5.046627173019943e-05, "loss": 0.0199, "step": 26110 }, { "epoch": 9.592361366140286, "grad_norm": 0.348268061876297, "learning_rate": 5.043320367222545e-05, "loss": 0.0192, "step": 26120 }, { "epoch": 9.59603378626515, "grad_norm": 0.22938024997711182, "learning_rate": 5.040013542475346e-05, "loss": 0.0202, "step": 26130 }, { "epoch": 9.59970620639001, "grad_norm": 0.2744646370410919, "learning_rate": 5.0367067002248646e-05, "loss": 0.0184, "step": 26140 }, { "epoch": 9.603378626514873, "grad_norm": 0.19532617926597595, "learning_rate": 5.0333998419176254e-05, "loss": 0.0216, "step": 26150 }, { "epoch": 9.607051046639736, "grad_norm": 0.32102346420288086, "learning_rate": 5.030092969000161e-05, "loss": 0.0204, "step": 26160 }, { "epoch": 9.610723466764599, "grad_norm": 0.3674446940422058, "learning_rate": 5.026786082919008e-05, "loss": 0.018, "step": 26170 }, { "epoch": 9.61439588688946, "grad_norm": 0.27250736951828003, "learning_rate": 5.0234791851207144e-05, "loss": 0.0194, "step": 26180 }, { "epoch": 9.618068307014322, "grad_norm": 0.2372112274169922, "learning_rate": 5.0201722770518256e-05, "loss": 0.0192, "step": 26190 }, { "epoch": 9.621740727139185, "grad_norm": 0.23337644338607788, "learning_rate": 5.0168653601589e-05, "loss": 0.02, "step": 26200 }, { "epoch": 9.625413147264046, "grad_norm": 0.36065781116485596, "learning_rate": 5.01355843588849e-05, "loss": 0.0208, "step": 26210 }, { "epoch": 9.629085567388909, "grad_norm": 0.6539337038993835, "learning_rate": 5.010251505687161e-05, "loss": 0.0195, "step": 26220 }, { "epoch": 9.632757987513772, "grad_norm": 0.20147530734539032, "learning_rate": 5.006944571001474e-05, "loss": 0.0195, "step": 26230 }, { "epoch": 9.636430407638635, "grad_norm": 0.21751704812049866, "learning_rate": 5.003637633277997e-05, "loss": 0.0191, "step": 26240 }, { "epoch": 9.640102827763496, "grad_norm": 0.24781446158885956, "learning_rate": 5.0003306939632946e-05, "loss": 0.0202, "step": 26250 }, { "epoch": 9.643775247888358, "grad_norm": 0.38428929448127747, "learning_rate": 4.997023754503937e-05, "loss": 0.0202, "step": 26260 }, { "epoch": 9.647447668013221, "grad_norm": 0.28729602694511414, "learning_rate": 4.9937168163464897e-05, "loss": 0.0202, "step": 26270 }, { "epoch": 9.651120088138082, "grad_norm": 0.2720409333705902, "learning_rate": 4.990409880937519e-05, "loss": 0.019, "step": 26280 }, { "epoch": 9.654792508262945, "grad_norm": 0.42715904116630554, "learning_rate": 4.9871029497235936e-05, "loss": 0.0206, "step": 26290 }, { "epoch": 9.658464928387808, "grad_norm": 0.21683259308338165, "learning_rate": 4.9837960241512766e-05, "loss": 0.0197, "step": 26300 }, { "epoch": 9.662137348512669, "grad_norm": 0.29459336400032043, "learning_rate": 4.9804891056671265e-05, "loss": 0.019, "step": 26310 }, { "epoch": 9.665809768637532, "grad_norm": 0.3164699375629425, "learning_rate": 4.9771821957177074e-05, "loss": 0.018, "step": 26320 }, { "epoch": 9.669482188762395, "grad_norm": 0.3513029217720032, "learning_rate": 4.9738752957495694e-05, "loss": 0.02, "step": 26330 }, { "epoch": 9.673154608887257, "grad_norm": 0.2484755665063858, "learning_rate": 4.970568407209265e-05, "loss": 0.0181, "step": 26340 }, { "epoch": 9.676827029012118, "grad_norm": 0.30611053109169006, "learning_rate": 4.967261531543339e-05, "loss": 0.0205, "step": 26350 }, { "epoch": 9.680499449136981, "grad_norm": 0.2256113737821579, "learning_rate": 4.963954670198332e-05, "loss": 0.0179, "step": 26360 }, { "epoch": 9.684171869261844, "grad_norm": 0.33197128772735596, "learning_rate": 4.960647824620776e-05, "loss": 0.0194, "step": 26370 }, { "epoch": 9.687844289386707, "grad_norm": 0.2777130603790283, "learning_rate": 4.9573409962571995e-05, "loss": 0.0213, "step": 26380 }, { "epoch": 9.691516709511568, "grad_norm": 0.24268947541713715, "learning_rate": 4.954034186554121e-05, "loss": 0.0174, "step": 26390 }, { "epoch": 9.69518912963643, "grad_norm": 0.24325235188007355, "learning_rate": 4.950727396958051e-05, "loss": 0.0183, "step": 26400 }, { "epoch": 9.698861549761293, "grad_norm": 0.3254251182079315, "learning_rate": 4.947420628915489e-05, "loss": 0.0194, "step": 26410 }, { "epoch": 9.702533969886154, "grad_norm": 0.20934663712978363, "learning_rate": 4.944113883872932e-05, "loss": 0.0174, "step": 26420 }, { "epoch": 9.706206390011017, "grad_norm": 0.22115367650985718, "learning_rate": 4.9408071632768624e-05, "loss": 0.0188, "step": 26430 }, { "epoch": 9.70987881013588, "grad_norm": 0.26688212156295776, "learning_rate": 4.937500468573747e-05, "loss": 0.0195, "step": 26440 }, { "epoch": 9.713551230260741, "grad_norm": 0.16121028363704681, "learning_rate": 4.934193801210053e-05, "loss": 0.0201, "step": 26450 }, { "epoch": 9.717223650385604, "grad_norm": 0.1825864017009735, "learning_rate": 4.930887162632224e-05, "loss": 0.0201, "step": 26460 }, { "epoch": 9.720896070510467, "grad_norm": 0.22363626956939697, "learning_rate": 4.9275805542866994e-05, "loss": 0.0176, "step": 26470 }, { "epoch": 9.72456849063533, "grad_norm": 0.22302275896072388, "learning_rate": 4.9242739776199006e-05, "loss": 0.0185, "step": 26480 }, { "epoch": 9.72824091076019, "grad_norm": 0.20900410413742065, "learning_rate": 4.920967434078237e-05, "loss": 0.0178, "step": 26490 }, { "epoch": 9.731913330885053, "grad_norm": 0.2319834679365158, "learning_rate": 4.917660925108101e-05, "loss": 0.0206, "step": 26500 }, { "epoch": 9.735585751009916, "grad_norm": 0.25404730439186096, "learning_rate": 4.914354452155876e-05, "loss": 0.0191, "step": 26510 }, { "epoch": 9.739258171134777, "grad_norm": 0.27197280526161194, "learning_rate": 4.911048016667923e-05, "loss": 0.0203, "step": 26520 }, { "epoch": 9.74293059125964, "grad_norm": 0.28384214639663696, "learning_rate": 4.907741620090588e-05, "loss": 0.0179, "step": 26530 }, { "epoch": 9.746603011384503, "grad_norm": 0.19956260919570923, "learning_rate": 4.9044352638702046e-05, "loss": 0.019, "step": 26540 }, { "epoch": 9.750275431509365, "grad_norm": 0.319550096988678, "learning_rate": 4.9011289494530834e-05, "loss": 0.021, "step": 26550 }, { "epoch": 9.753947851634226, "grad_norm": 0.23120824992656708, "learning_rate": 4.897822678285516e-05, "loss": 0.0186, "step": 26560 }, { "epoch": 9.75762027175909, "grad_norm": 0.23956599831581116, "learning_rate": 4.8945164518137835e-05, "loss": 0.0196, "step": 26570 }, { "epoch": 9.761292691883952, "grad_norm": 0.2387133091688156, "learning_rate": 4.891210271484138e-05, "loss": 0.0206, "step": 26580 }, { "epoch": 9.764965112008813, "grad_norm": 0.27635693550109863, "learning_rate": 4.8879041387428164e-05, "loss": 0.0192, "step": 26590 }, { "epoch": 9.768637532133676, "grad_norm": 0.21461892127990723, "learning_rate": 4.8845980550360306e-05, "loss": 0.0196, "step": 26600 }, { "epoch": 9.772309952258539, "grad_norm": 0.40779930353164673, "learning_rate": 4.8812920218099786e-05, "loss": 0.0208, "step": 26610 }, { "epoch": 9.775982372383401, "grad_norm": 0.3521687984466553, "learning_rate": 4.8779860405108294e-05, "loss": 0.0189, "step": 26620 }, { "epoch": 9.779654792508262, "grad_norm": 0.29101869463920593, "learning_rate": 4.8746801125847285e-05, "loss": 0.0201, "step": 26630 }, { "epoch": 9.783327212633125, "grad_norm": 0.3944571912288666, "learning_rate": 4.8713742394778064e-05, "loss": 0.0231, "step": 26640 }, { "epoch": 9.786999632757988, "grad_norm": 0.4579140543937683, "learning_rate": 4.868068422636162e-05, "loss": 0.0216, "step": 26650 }, { "epoch": 9.790672052882849, "grad_norm": 0.2574957013130188, "learning_rate": 4.86476266350587e-05, "loss": 0.0201, "step": 26660 }, { "epoch": 9.794344473007712, "grad_norm": 0.28458502888679504, "learning_rate": 4.861456963532985e-05, "loss": 0.0194, "step": 26670 }, { "epoch": 9.798016893132575, "grad_norm": 0.3878946006298065, "learning_rate": 4.8581513241635316e-05, "loss": 0.0188, "step": 26680 }, { "epoch": 9.801689313257437, "grad_norm": 0.1330614686012268, "learning_rate": 4.8548457468435065e-05, "loss": 0.0186, "step": 26690 }, { "epoch": 9.805361733382298, "grad_norm": 0.3820025622844696, "learning_rate": 4.851540233018884e-05, "loss": 0.0188, "step": 26700 }, { "epoch": 9.809034153507161, "grad_norm": 0.1897294670343399, "learning_rate": 4.848234784135608e-05, "loss": 0.0208, "step": 26710 }, { "epoch": 9.812706573632024, "grad_norm": 0.21062450110912323, "learning_rate": 4.8449294016395916e-05, "loss": 0.02, "step": 26720 }, { "epoch": 9.816378993756885, "grad_norm": 0.4511318802833557, "learning_rate": 4.8416240869767246e-05, "loss": 0.018, "step": 26730 }, { "epoch": 9.820051413881748, "grad_norm": 0.2621796131134033, "learning_rate": 4.838318841592864e-05, "loss": 0.0201, "step": 26740 }, { "epoch": 9.82372383400661, "grad_norm": 0.27163153886795044, "learning_rate": 4.835013666933834e-05, "loss": 0.0192, "step": 26750 }, { "epoch": 9.827396254131473, "grad_norm": 0.24845406413078308, "learning_rate": 4.83170856444543e-05, "loss": 0.0193, "step": 26760 }, { "epoch": 9.831068674256334, "grad_norm": 0.23590105772018433, "learning_rate": 4.828403535573419e-05, "loss": 0.0198, "step": 26770 }, { "epoch": 9.834741094381197, "grad_norm": 0.23702068626880646, "learning_rate": 4.825098581763532e-05, "loss": 0.0191, "step": 26780 }, { "epoch": 9.83841351450606, "grad_norm": 0.36168810725212097, "learning_rate": 4.821793704461465e-05, "loss": 0.0202, "step": 26790 }, { "epoch": 9.842085934630921, "grad_norm": 0.22400982677936554, "learning_rate": 4.8184889051128884e-05, "loss": 0.0192, "step": 26800 }, { "epoch": 9.845758354755784, "grad_norm": 0.28203949332237244, "learning_rate": 4.81518418516343e-05, "loss": 0.0191, "step": 26810 }, { "epoch": 9.849430774880647, "grad_norm": 0.390813410282135, "learning_rate": 4.811879546058688e-05, "loss": 0.0189, "step": 26820 }, { "epoch": 9.85310319500551, "grad_norm": 0.3316987156867981, "learning_rate": 4.808574989244224e-05, "loss": 0.0194, "step": 26830 }, { "epoch": 9.85677561513037, "grad_norm": 0.2979455590248108, "learning_rate": 4.8052705161655644e-05, "loss": 0.0197, "step": 26840 }, { "epoch": 9.860448035255233, "grad_norm": 0.3666852116584778, "learning_rate": 4.8019661282681936e-05, "loss": 0.0195, "step": 26850 }, { "epoch": 9.864120455380096, "grad_norm": 0.310859352350235, "learning_rate": 4.79866182699757e-05, "loss": 0.0222, "step": 26860 }, { "epoch": 9.867792875504957, "grad_norm": 0.2666810154914856, "learning_rate": 4.795357613799103e-05, "loss": 0.0186, "step": 26870 }, { "epoch": 9.87146529562982, "grad_norm": 0.3397727906703949, "learning_rate": 4.792053490118166e-05, "loss": 0.0205, "step": 26880 }, { "epoch": 9.875137715754683, "grad_norm": 0.2004554122686386, "learning_rate": 4.788749457400099e-05, "loss": 0.018, "step": 26890 }, { "epoch": 9.878810135879544, "grad_norm": 0.15223811566829681, "learning_rate": 4.7854455170901966e-05, "loss": 0.0199, "step": 26900 }, { "epoch": 9.882482556004407, "grad_norm": 0.20749245584011078, "learning_rate": 4.7821416706337136e-05, "loss": 0.0195, "step": 26910 }, { "epoch": 9.88615497612927, "grad_norm": 0.3082486093044281, "learning_rate": 4.7788379194758645e-05, "loss": 0.0174, "step": 26920 }, { "epoch": 9.889827396254132, "grad_norm": 0.42015862464904785, "learning_rate": 4.7755342650618246e-05, "loss": 0.0186, "step": 26930 }, { "epoch": 9.893499816378993, "grad_norm": 0.3639332950115204, "learning_rate": 4.7722307088367235e-05, "loss": 0.0217, "step": 26940 }, { "epoch": 9.897172236503856, "grad_norm": 0.28031742572784424, "learning_rate": 4.768927252245646e-05, "loss": 0.0189, "step": 26950 }, { "epoch": 9.900844656628719, "grad_norm": 0.3103528618812561, "learning_rate": 4.765623896733642e-05, "loss": 0.0192, "step": 26960 }, { "epoch": 9.904517076753582, "grad_norm": 0.500171959400177, "learning_rate": 4.762320643745708e-05, "loss": 0.0182, "step": 26970 }, { "epoch": 9.908189496878443, "grad_norm": 0.2734226584434509, "learning_rate": 4.7590174947267966e-05, "loss": 0.0182, "step": 26980 }, { "epoch": 9.911861917003305, "grad_norm": 0.24641767144203186, "learning_rate": 4.755714451121823e-05, "loss": 0.0213, "step": 26990 }, { "epoch": 9.915534337128168, "grad_norm": 0.21376626193523407, "learning_rate": 4.7524115143756476e-05, "loss": 0.0178, "step": 27000 }, { "epoch": 9.91920675725303, "grad_norm": 0.22290736436843872, "learning_rate": 4.7491086859330854e-05, "loss": 0.0195, "step": 27010 }, { "epoch": 9.922879177377892, "grad_norm": 0.27350568771362305, "learning_rate": 4.74580596723891e-05, "loss": 0.0187, "step": 27020 }, { "epoch": 9.926551597502755, "grad_norm": 0.2541103959083557, "learning_rate": 4.742503359737841e-05, "loss": 0.0193, "step": 27030 }, { "epoch": 9.930224017627616, "grad_norm": 0.27475637197494507, "learning_rate": 4.739200864874547e-05, "loss": 0.0192, "step": 27040 }, { "epoch": 9.933896437752479, "grad_norm": 0.256971538066864, "learning_rate": 4.735898484093658e-05, "loss": 0.0187, "step": 27050 }, { "epoch": 9.937568857877341, "grad_norm": 0.31392404437065125, "learning_rate": 4.7325962188397444e-05, "loss": 0.0171, "step": 27060 }, { "epoch": 9.941241278002204, "grad_norm": 0.24793167412281036, "learning_rate": 4.729294070557329e-05, "loss": 0.0188, "step": 27070 }, { "epoch": 9.944913698127065, "grad_norm": 0.4545385539531708, "learning_rate": 4.7259920406908833e-05, "loss": 0.0206, "step": 27080 }, { "epoch": 9.948586118251928, "grad_norm": 0.304248183965683, "learning_rate": 4.7226901306848306e-05, "loss": 0.0213, "step": 27090 }, { "epoch": 9.95225853837679, "grad_norm": 0.24821165204048157, "learning_rate": 4.719388341983535e-05, "loss": 0.0202, "step": 27100 }, { "epoch": 9.955930958501652, "grad_norm": 0.28063222765922546, "learning_rate": 4.71608667603131e-05, "loss": 0.0206, "step": 27110 }, { "epoch": 9.959603378626515, "grad_norm": 0.5596874952316284, "learning_rate": 4.712785134272422e-05, "loss": 0.0188, "step": 27120 }, { "epoch": 9.963275798751377, "grad_norm": 0.18386408686637878, "learning_rate": 4.7094837181510734e-05, "loss": 0.0209, "step": 27130 }, { "epoch": 9.96694821887624, "grad_norm": 0.3313400447368622, "learning_rate": 4.706182429111414e-05, "loss": 0.0189, "step": 27140 }, { "epoch": 9.970620639001101, "grad_norm": 0.16718442738056183, "learning_rate": 4.702881268597545e-05, "loss": 0.0196, "step": 27150 }, { "epoch": 9.974293059125964, "grad_norm": 0.2779512107372284, "learning_rate": 4.699580238053503e-05, "loss": 0.0192, "step": 27160 }, { "epoch": 9.977965479250827, "grad_norm": 0.2804941236972809, "learning_rate": 4.69627933892327e-05, "loss": 0.0178, "step": 27170 }, { "epoch": 9.981637899375688, "grad_norm": 0.21919390559196472, "learning_rate": 4.692978572650773e-05, "loss": 0.0225, "step": 27180 }, { "epoch": 9.98531031950055, "grad_norm": 0.26205500960350037, "learning_rate": 4.689677940679881e-05, "loss": 0.0198, "step": 27190 }, { "epoch": 9.988982739625413, "grad_norm": 0.27378037571907043, "learning_rate": 4.6863774444543984e-05, "loss": 0.0185, "step": 27200 }, { "epoch": 9.992655159750276, "grad_norm": 0.31205710768699646, "learning_rate": 4.6830770854180784e-05, "loss": 0.0191, "step": 27210 }, { "epoch": 9.996327579875137, "grad_norm": 0.21455001831054688, "learning_rate": 4.679776865014608e-05, "loss": 0.0185, "step": 27220 }, { "epoch": 10.0, "grad_norm": 0.35200220346450806, "learning_rate": 4.676476784687615e-05, "loss": 0.0184, "step": 27230 }, { "epoch": 10.003672420124863, "grad_norm": 0.2574729025363922, "learning_rate": 4.673176845880669e-05, "loss": 0.0185, "step": 27240 }, { "epoch": 10.007344840249724, "grad_norm": 0.22513633966445923, "learning_rate": 4.669877050037275e-05, "loss": 0.0191, "step": 27250 }, { "epoch": 10.011017260374587, "grad_norm": 0.22502222657203674, "learning_rate": 4.666577398600874e-05, "loss": 0.0191, "step": 27260 }, { "epoch": 10.01468968049945, "grad_norm": 0.17053654789924622, "learning_rate": 4.663277893014846e-05, "loss": 0.0182, "step": 27270 }, { "epoch": 10.018362100624312, "grad_norm": 0.2576551139354706, "learning_rate": 4.659978534722508e-05, "loss": 0.0186, "step": 27280 }, { "epoch": 10.022034520749173, "grad_norm": 0.2316872775554657, "learning_rate": 4.6566793251671104e-05, "loss": 0.0196, "step": 27290 }, { "epoch": 10.025706940874036, "grad_norm": 0.46977826952934265, "learning_rate": 4.65338026579184e-05, "loss": 0.0199, "step": 27300 }, { "epoch": 10.029379360998899, "grad_norm": 0.34212595224380493, "learning_rate": 4.650081358039817e-05, "loss": 0.0182, "step": 27310 }, { "epoch": 10.03305178112376, "grad_norm": 0.22159533202648163, "learning_rate": 4.646782603354098e-05, "loss": 0.0189, "step": 27320 }, { "epoch": 10.036724201248623, "grad_norm": 0.30063432455062866, "learning_rate": 4.643484003177666e-05, "loss": 0.019, "step": 27330 }, { "epoch": 10.040396621373485, "grad_norm": 0.170504629611969, "learning_rate": 4.6401855589534454e-05, "loss": 0.0199, "step": 27340 }, { "epoch": 10.044069041498348, "grad_norm": 0.2754218876361847, "learning_rate": 4.636887272124286e-05, "loss": 0.0199, "step": 27350 }, { "epoch": 10.04774146162321, "grad_norm": 0.2913573384284973, "learning_rate": 4.6335891441329685e-05, "loss": 0.0187, "step": 27360 }, { "epoch": 10.051413881748072, "grad_norm": 0.37481948733329773, "learning_rate": 4.630291176422209e-05, "loss": 0.0204, "step": 27370 }, { "epoch": 10.055086301872935, "grad_norm": 0.33804211020469666, "learning_rate": 4.62699337043465e-05, "loss": 0.0186, "step": 27380 }, { "epoch": 10.058758721997796, "grad_norm": 0.20603089034557343, "learning_rate": 4.623695727612861e-05, "loss": 0.0195, "step": 27390 }, { "epoch": 10.062431142122659, "grad_norm": 0.29974791407585144, "learning_rate": 4.620398249399347e-05, "loss": 0.0186, "step": 27400 }, { "epoch": 10.066103562247521, "grad_norm": 0.30096620321273804, "learning_rate": 4.617100937236535e-05, "loss": 0.0193, "step": 27410 }, { "epoch": 10.069775982372384, "grad_norm": 0.218524768948555, "learning_rate": 4.613803792566781e-05, "loss": 0.0198, "step": 27420 }, { "epoch": 10.073448402497245, "grad_norm": 0.3332307040691376, "learning_rate": 4.61050681683237e-05, "loss": 0.0176, "step": 27430 }, { "epoch": 10.077120822622108, "grad_norm": 0.28056344389915466, "learning_rate": 4.6072100114755104e-05, "loss": 0.0182, "step": 27440 }, { "epoch": 10.08079324274697, "grad_norm": 0.20256339013576508, "learning_rate": 4.603913377938336e-05, "loss": 0.0195, "step": 27450 }, { "epoch": 10.084465662871832, "grad_norm": 0.23719756305217743, "learning_rate": 4.600616917662905e-05, "loss": 0.0222, "step": 27460 }, { "epoch": 10.088138082996695, "grad_norm": 0.26492586731910706, "learning_rate": 4.597320632091206e-05, "loss": 0.0183, "step": 27470 }, { "epoch": 10.091810503121557, "grad_norm": 0.22299693524837494, "learning_rate": 4.594024522665142e-05, "loss": 0.0191, "step": 27480 }, { "epoch": 10.09548292324642, "grad_norm": 0.25159159302711487, "learning_rate": 4.590728590826543e-05, "loss": 0.0203, "step": 27490 }, { "epoch": 10.099155343371281, "grad_norm": 0.232728511095047, "learning_rate": 4.587432838017165e-05, "loss": 0.02, "step": 27500 }, { "epoch": 10.102827763496144, "grad_norm": 0.1660882532596588, "learning_rate": 4.58413726567868e-05, "loss": 0.018, "step": 27510 }, { "epoch": 10.106500183621007, "grad_norm": 0.1855497509241104, "learning_rate": 4.5808418752526814e-05, "loss": 0.0198, "step": 27520 }, { "epoch": 10.110172603745868, "grad_norm": 0.1847846955060959, "learning_rate": 4.5775466681806895e-05, "loss": 0.016, "step": 27530 }, { "epoch": 10.11384502387073, "grad_norm": 0.23802553117275238, "learning_rate": 4.574251645904136e-05, "loss": 0.0182, "step": 27540 }, { "epoch": 10.117517443995593, "grad_norm": 0.2904793918132782, "learning_rate": 4.5709568098643776e-05, "loss": 0.0176, "step": 27550 }, { "epoch": 10.121189864120455, "grad_norm": 0.39061984419822693, "learning_rate": 4.567662161502687e-05, "loss": 0.0181, "step": 27560 }, { "epoch": 10.124862284245317, "grad_norm": 0.23047636449337006, "learning_rate": 4.5643677022602554e-05, "loss": 0.0187, "step": 27570 }, { "epoch": 10.12853470437018, "grad_norm": 0.34967347979545593, "learning_rate": 4.56107343357819e-05, "loss": 0.0178, "step": 27580 }, { "epoch": 10.132207124495043, "grad_norm": 0.3047192096710205, "learning_rate": 4.557779356897518e-05, "loss": 0.0168, "step": 27590 }, { "epoch": 10.135879544619904, "grad_norm": 0.29713696241378784, "learning_rate": 4.55448547365918e-05, "loss": 0.0195, "step": 27600 }, { "epoch": 10.139551964744767, "grad_norm": 0.19994090497493744, "learning_rate": 4.5511917853040315e-05, "loss": 0.0186, "step": 27610 }, { "epoch": 10.14322438486963, "grad_norm": 0.22952978312969208, "learning_rate": 4.547898293272842e-05, "loss": 0.0182, "step": 27620 }, { "epoch": 10.14689680499449, "grad_norm": 0.30646875500679016, "learning_rate": 4.544604999006301e-05, "loss": 0.0183, "step": 27630 }, { "epoch": 10.150569225119353, "grad_norm": 0.2957550585269928, "learning_rate": 4.541311903945004e-05, "loss": 0.0183, "step": 27640 }, { "epoch": 10.154241645244216, "grad_norm": 0.312610000371933, "learning_rate": 4.538019009529463e-05, "loss": 0.0199, "step": 27650 }, { "epoch": 10.157914065369079, "grad_norm": 0.6569544076919556, "learning_rate": 4.5347263172001035e-05, "loss": 0.0182, "step": 27660 }, { "epoch": 10.16158648549394, "grad_norm": 0.2970186471939087, "learning_rate": 4.531433828397261e-05, "loss": 0.0193, "step": 27670 }, { "epoch": 10.165258905618803, "grad_norm": 0.30370742082595825, "learning_rate": 4.528141544561179e-05, "loss": 0.019, "step": 27680 }, { "epoch": 10.168931325743666, "grad_norm": 0.32576340436935425, "learning_rate": 4.5248494671320186e-05, "loss": 0.0178, "step": 27690 }, { "epoch": 10.172603745868527, "grad_norm": 0.2863040864467621, "learning_rate": 4.521557597549844e-05, "loss": 0.0187, "step": 27700 }, { "epoch": 10.17627616599339, "grad_norm": 0.1747882217168808, "learning_rate": 4.5182659372546284e-05, "loss": 0.0179, "step": 27710 }, { "epoch": 10.179948586118252, "grad_norm": 0.2321597933769226, "learning_rate": 4.514974487686261e-05, "loss": 0.0182, "step": 27720 }, { "epoch": 10.183621006243115, "grad_norm": 0.18374888598918915, "learning_rate": 4.5116832502845306e-05, "loss": 0.0206, "step": 27730 }, { "epoch": 10.187293426367976, "grad_norm": 0.37515339255332947, "learning_rate": 4.508392226489136e-05, "loss": 0.0176, "step": 27740 }, { "epoch": 10.190965846492839, "grad_norm": 0.30109432339668274, "learning_rate": 4.505101417739683e-05, "loss": 0.0186, "step": 27750 }, { "epoch": 10.194638266617702, "grad_norm": 0.3563247323036194, "learning_rate": 4.501810825475685e-05, "loss": 0.0192, "step": 27760 }, { "epoch": 10.198310686742563, "grad_norm": 0.23846521973609924, "learning_rate": 4.4985204511365556e-05, "loss": 0.0192, "step": 27770 }, { "epoch": 10.201983106867425, "grad_norm": 0.24678164720535278, "learning_rate": 4.49523029616162e-05, "loss": 0.0187, "step": 27780 }, { "epoch": 10.205655526992288, "grad_norm": 0.264370322227478, "learning_rate": 4.491940361990101e-05, "loss": 0.0183, "step": 27790 }, { "epoch": 10.209327947117151, "grad_norm": 0.37711089849472046, "learning_rate": 4.48865065006113e-05, "loss": 0.0171, "step": 27800 }, { "epoch": 10.213000367242012, "grad_norm": 0.318307101726532, "learning_rate": 4.485361161813735e-05, "loss": 0.0193, "step": 27810 }, { "epoch": 10.216672787366875, "grad_norm": 0.24768127501010895, "learning_rate": 4.482071898686854e-05, "loss": 0.019, "step": 27820 }, { "epoch": 10.220345207491738, "grad_norm": 0.2212926596403122, "learning_rate": 4.478782862119321e-05, "loss": 0.0184, "step": 27830 }, { "epoch": 10.224017627616599, "grad_norm": 0.17821410298347473, "learning_rate": 4.475494053549869e-05, "loss": 0.019, "step": 27840 }, { "epoch": 10.227690047741461, "grad_norm": 0.3248125910758972, "learning_rate": 4.4722054744171405e-05, "loss": 0.0173, "step": 27850 }, { "epoch": 10.231362467866324, "grad_norm": 0.2679917812347412, "learning_rate": 4.468917126159669e-05, "loss": 0.0173, "step": 27860 }, { "epoch": 10.235034887991187, "grad_norm": 0.2998839318752289, "learning_rate": 4.465629010215887e-05, "loss": 0.019, "step": 27870 }, { "epoch": 10.238707308116048, "grad_norm": 0.28288203477859497, "learning_rate": 4.462341128024133e-05, "loss": 0.0186, "step": 27880 }, { "epoch": 10.24237972824091, "grad_norm": 0.36060529947280884, "learning_rate": 4.4590534810226356e-05, "loss": 0.0186, "step": 27890 }, { "epoch": 10.246052148365774, "grad_norm": 0.22151124477386475, "learning_rate": 4.455766070649523e-05, "loss": 0.0184, "step": 27900 }, { "epoch": 10.249724568490635, "grad_norm": 0.3787975609302521, "learning_rate": 4.4524788983428215e-05, "loss": 0.0188, "step": 27910 }, { "epoch": 10.253396988615497, "grad_norm": 0.23865239322185516, "learning_rate": 4.4491919655404515e-05, "loss": 0.0184, "step": 27920 }, { "epoch": 10.25706940874036, "grad_norm": 0.3483262062072754, "learning_rate": 4.445905273680228e-05, "loss": 0.0199, "step": 27930 }, { "epoch": 10.260741828865223, "grad_norm": 0.33129385113716125, "learning_rate": 4.442618824199865e-05, "loss": 0.0195, "step": 27940 }, { "epoch": 10.264414248990084, "grad_norm": 0.2626710534095764, "learning_rate": 4.4393326185369635e-05, "loss": 0.0192, "step": 27950 }, { "epoch": 10.268086669114947, "grad_norm": 0.24617516994476318, "learning_rate": 4.436046658129024e-05, "loss": 0.018, "step": 27960 }, { "epoch": 10.27175908923981, "grad_norm": 0.3842035233974457, "learning_rate": 4.432760944413433e-05, "loss": 0.0175, "step": 27970 }, { "epoch": 10.27543150936467, "grad_norm": 0.21065862476825714, "learning_rate": 4.4294754788274796e-05, "loss": 0.0178, "step": 27980 }, { "epoch": 10.279103929489533, "grad_norm": 0.22702984511852264, "learning_rate": 4.426190262808334e-05, "loss": 0.0172, "step": 27990 }, { "epoch": 10.282776349614396, "grad_norm": 0.18850652873516083, "learning_rate": 4.42290529779306e-05, "loss": 0.0185, "step": 28000 }, { "epoch": 10.286448769739259, "grad_norm": 0.2558286786079407, "learning_rate": 4.4196205852186176e-05, "loss": 0.0179, "step": 28010 }, { "epoch": 10.29012118986412, "grad_norm": 0.29564592242240906, "learning_rate": 4.416336126521848e-05, "loss": 0.0215, "step": 28020 }, { "epoch": 10.293793609988983, "grad_norm": 0.21667353808879852, "learning_rate": 4.4130519231394865e-05, "loss": 0.0192, "step": 28030 }, { "epoch": 10.297466030113846, "grad_norm": 0.2990483343601227, "learning_rate": 4.409767976508154e-05, "loss": 0.0189, "step": 28040 }, { "epoch": 10.301138450238707, "grad_norm": 0.2748870849609375, "learning_rate": 4.406484288064363e-05, "loss": 0.0187, "step": 28050 }, { "epoch": 10.30481087036357, "grad_norm": 0.43740734457969666, "learning_rate": 4.4032008592445055e-05, "loss": 0.019, "step": 28060 }, { "epoch": 10.308483290488432, "grad_norm": 0.2923041880130768, "learning_rate": 4.3999176914848714e-05, "loss": 0.0182, "step": 28070 }, { "epoch": 10.312155710613293, "grad_norm": 0.42060649394989014, "learning_rate": 4.396634786221625e-05, "loss": 0.0176, "step": 28080 }, { "epoch": 10.315828130738156, "grad_norm": 0.28482386469841003, "learning_rate": 4.3933521448908203e-05, "loss": 0.0194, "step": 28090 }, { "epoch": 10.319500550863019, "grad_norm": 0.21711628139019012, "learning_rate": 4.3900697689284e-05, "loss": 0.0199, "step": 28100 }, { "epoch": 10.323172970987882, "grad_norm": 0.24537959694862366, "learning_rate": 4.386787659770185e-05, "loss": 0.0175, "step": 28110 }, { "epoch": 10.326845391112743, "grad_norm": 0.2260241061449051, "learning_rate": 4.383505818851878e-05, "loss": 0.0186, "step": 28120 }, { "epoch": 10.330517811237605, "grad_norm": 0.3111761808395386, "learning_rate": 4.380224247609073e-05, "loss": 0.0171, "step": 28130 }, { "epoch": 10.334190231362468, "grad_norm": 0.4362003803253174, "learning_rate": 4.3769429474772376e-05, "loss": 0.0165, "step": 28140 }, { "epoch": 10.33786265148733, "grad_norm": 0.42787373065948486, "learning_rate": 4.373661919891726e-05, "loss": 0.0171, "step": 28150 }, { "epoch": 10.341535071612192, "grad_norm": 0.4129069149494171, "learning_rate": 4.3703811662877674e-05, "loss": 0.0168, "step": 28160 }, { "epoch": 10.345207491737055, "grad_norm": 0.23690001666545868, "learning_rate": 4.36710068810048e-05, "loss": 0.0185, "step": 28170 }, { "epoch": 10.348879911861918, "grad_norm": 0.330162912607193, "learning_rate": 4.363820486764853e-05, "loss": 0.0173, "step": 28180 }, { "epoch": 10.352552331986779, "grad_norm": 0.37791067361831665, "learning_rate": 4.360540563715758e-05, "loss": 0.0163, "step": 28190 }, { "epoch": 10.356224752111642, "grad_norm": 0.2761913537979126, "learning_rate": 4.357260920387946e-05, "loss": 0.0196, "step": 28200 }, { "epoch": 10.359897172236504, "grad_norm": 0.24142572283744812, "learning_rate": 4.353981558216045e-05, "loss": 0.0179, "step": 28210 }, { "epoch": 10.363569592361365, "grad_norm": 0.26954394578933716, "learning_rate": 4.3507024786345565e-05, "loss": 0.0185, "step": 28220 }, { "epoch": 10.367242012486228, "grad_norm": 0.3127514123916626, "learning_rate": 4.3474236830778646e-05, "loss": 0.0187, "step": 28230 }, { "epoch": 10.370914432611091, "grad_norm": 0.4675455391407013, "learning_rate": 4.344145172980223e-05, "loss": 0.0181, "step": 28240 }, { "epoch": 10.374586852735954, "grad_norm": 0.4532056748867035, "learning_rate": 4.3408669497757656e-05, "loss": 0.0187, "step": 28250 }, { "epoch": 10.378259272860815, "grad_norm": 0.2476673722267151, "learning_rate": 4.3375890148984976e-05, "loss": 0.0185, "step": 28260 }, { "epoch": 10.381931692985678, "grad_norm": 0.2520010471343994, "learning_rate": 4.334311369782301e-05, "loss": 0.0185, "step": 28270 }, { "epoch": 10.38560411311054, "grad_norm": 0.27636536955833435, "learning_rate": 4.3310340158609244e-05, "loss": 0.0171, "step": 28280 }, { "epoch": 10.389276533235401, "grad_norm": 0.31686878204345703, "learning_rate": 4.327756954568e-05, "loss": 0.0182, "step": 28290 }, { "epoch": 10.392948953360264, "grad_norm": 0.38625141978263855, "learning_rate": 4.324480187337021e-05, "loss": 0.0168, "step": 28300 }, { "epoch": 10.396621373485127, "grad_norm": 0.2923319935798645, "learning_rate": 4.3212037156013596e-05, "loss": 0.0173, "step": 28310 }, { "epoch": 10.40029379360999, "grad_norm": 0.2921278178691864, "learning_rate": 4.317927540794251e-05, "loss": 0.0191, "step": 28320 }, { "epoch": 10.40396621373485, "grad_norm": 0.35856980085372925, "learning_rate": 4.314651664348812e-05, "loss": 0.0195, "step": 28330 }, { "epoch": 10.407638633859714, "grad_norm": 0.26759716868400574, "learning_rate": 4.3113760876980195e-05, "loss": 0.0179, "step": 28340 }, { "epoch": 10.411311053984576, "grad_norm": 0.2451809197664261, "learning_rate": 4.308100812274719e-05, "loss": 0.02, "step": 28350 }, { "epoch": 10.414983474109437, "grad_norm": 0.34193816781044006, "learning_rate": 4.304825839511632e-05, "loss": 0.0177, "step": 28360 }, { "epoch": 10.4186558942343, "grad_norm": 0.22469192743301392, "learning_rate": 4.3015511708413395e-05, "loss": 0.0192, "step": 28370 }, { "epoch": 10.422328314359163, "grad_norm": 0.3873807489871979, "learning_rate": 4.298276807696294e-05, "loss": 0.019, "step": 28380 }, { "epoch": 10.426000734484026, "grad_norm": 0.3300996720790863, "learning_rate": 4.2950027515088136e-05, "loss": 0.0192, "step": 28390 }, { "epoch": 10.429673154608887, "grad_norm": 0.29065442085266113, "learning_rate": 4.291729003711083e-05, "loss": 0.0181, "step": 28400 }, { "epoch": 10.43334557473375, "grad_norm": 0.21022994816303253, "learning_rate": 4.2884555657351455e-05, "loss": 0.02, "step": 28410 }, { "epoch": 10.437017994858612, "grad_norm": 0.2063482701778412, "learning_rate": 4.285182439012921e-05, "loss": 0.0182, "step": 28420 }, { "epoch": 10.440690414983473, "grad_norm": 0.1741226762533188, "learning_rate": 4.281909624976183e-05, "loss": 0.0158, "step": 28430 }, { "epoch": 10.444362835108336, "grad_norm": 0.31840693950653076, "learning_rate": 4.278637125056568e-05, "loss": 0.019, "step": 28440 }, { "epoch": 10.448035255233199, "grad_norm": 0.34049680829048157, "learning_rate": 4.2753649406855855e-05, "loss": 0.0181, "step": 28450 }, { "epoch": 10.451707675358062, "grad_norm": 0.33190980553627014, "learning_rate": 4.2720930732945966e-05, "loss": 0.0206, "step": 28460 }, { "epoch": 10.455380095482923, "grad_norm": 0.25460025668144226, "learning_rate": 4.268821524314826e-05, "loss": 0.0171, "step": 28470 }, { "epoch": 10.459052515607786, "grad_norm": 0.3041885495185852, "learning_rate": 4.2655502951773623e-05, "loss": 0.0179, "step": 28480 }, { "epoch": 10.462724935732648, "grad_norm": 0.23357631266117096, "learning_rate": 4.262279387313152e-05, "loss": 0.019, "step": 28490 }, { "epoch": 10.46639735585751, "grad_norm": 0.4557451009750366, "learning_rate": 4.259008802153001e-05, "loss": 0.0178, "step": 28500 }, { "epoch": 10.470069775982372, "grad_norm": 0.25521495938301086, "learning_rate": 4.2557385411275735e-05, "loss": 0.0175, "step": 28510 }, { "epoch": 10.473742196107235, "grad_norm": 0.24986565113067627, "learning_rate": 4.252468605667395e-05, "loss": 0.0193, "step": 28520 }, { "epoch": 10.477414616232098, "grad_norm": 0.1887599676847458, "learning_rate": 4.2491989972028456e-05, "loss": 0.0173, "step": 28530 }, { "epoch": 10.481087036356959, "grad_norm": 0.2808990180492401, "learning_rate": 4.245929717164161e-05, "loss": 0.0194, "step": 28540 }, { "epoch": 10.484759456481822, "grad_norm": 0.24932603538036346, "learning_rate": 4.242660766981438e-05, "loss": 0.0184, "step": 28550 }, { "epoch": 10.488431876606684, "grad_norm": 0.24500110745429993, "learning_rate": 4.2393921480846264e-05, "loss": 0.0164, "step": 28560 }, { "epoch": 10.492104296731545, "grad_norm": 0.20467375218868256, "learning_rate": 4.236123861903528e-05, "loss": 0.0169, "step": 28570 }, { "epoch": 10.495776716856408, "grad_norm": 0.43701794743537903, "learning_rate": 4.232855909867806e-05, "loss": 0.0169, "step": 28580 }, { "epoch": 10.499449136981271, "grad_norm": 0.31340256333351135, "learning_rate": 4.229588293406972e-05, "loss": 0.0173, "step": 28590 }, { "epoch": 10.503121557106134, "grad_norm": 0.2242688536643982, "learning_rate": 4.226321013950392e-05, "loss": 0.0191, "step": 28600 }, { "epoch": 10.506793977230995, "grad_norm": 0.25731876492500305, "learning_rate": 4.223054072927286e-05, "loss": 0.0169, "step": 28610 }, { "epoch": 10.510466397355858, "grad_norm": 0.28213298320770264, "learning_rate": 4.219787471766724e-05, "loss": 0.0205, "step": 28620 }, { "epoch": 10.51413881748072, "grad_norm": 0.22972629964351654, "learning_rate": 4.216521211897629e-05, "loss": 0.0177, "step": 28630 }, { "epoch": 10.517811237605581, "grad_norm": 0.3978835642337799, "learning_rate": 4.2132552947487734e-05, "loss": 0.0182, "step": 28640 }, { "epoch": 10.521483657730444, "grad_norm": 0.18055924773216248, "learning_rate": 4.209989721748783e-05, "loss": 0.0189, "step": 28650 }, { "epoch": 10.525156077855307, "grad_norm": 0.25926193594932556, "learning_rate": 4.206724494326127e-05, "loss": 0.0159, "step": 28660 }, { "epoch": 10.528828497980168, "grad_norm": 0.23521395027637482, "learning_rate": 4.203459613909126e-05, "loss": 0.0184, "step": 28670 }, { "epoch": 10.53250091810503, "grad_norm": 0.4085614085197449, "learning_rate": 4.200195081925955e-05, "loss": 0.0189, "step": 28680 }, { "epoch": 10.536173338229894, "grad_norm": 0.4989596903324127, "learning_rate": 4.196930899804627e-05, "loss": 0.0182, "step": 28690 }, { "epoch": 10.539845758354756, "grad_norm": 0.1506185233592987, "learning_rate": 4.1936670689730044e-05, "loss": 0.0168, "step": 28700 }, { "epoch": 10.543518178479617, "grad_norm": 0.29143664240837097, "learning_rate": 4.190403590858804e-05, "loss": 0.0168, "step": 28710 }, { "epoch": 10.54719059860448, "grad_norm": 0.3090100586414337, "learning_rate": 4.187140466889575e-05, "loss": 0.0175, "step": 28720 }, { "epoch": 10.550863018729343, "grad_norm": 0.21326079964637756, "learning_rate": 4.183877698492724e-05, "loss": 0.0178, "step": 28730 }, { "epoch": 10.554535438854206, "grad_norm": 0.22118543088436127, "learning_rate": 4.1806152870954935e-05, "loss": 0.0183, "step": 28740 }, { "epoch": 10.558207858979067, "grad_norm": 0.3368147909641266, "learning_rate": 4.177353234124976e-05, "loss": 0.0165, "step": 28750 }, { "epoch": 10.56188027910393, "grad_norm": 0.19751261174678802, "learning_rate": 4.1740915410081e-05, "loss": 0.0195, "step": 28760 }, { "epoch": 10.565552699228792, "grad_norm": 0.22031502425670624, "learning_rate": 4.170830209171646e-05, "loss": 0.0187, "step": 28770 }, { "epoch": 10.569225119353653, "grad_norm": 0.23732203245162964, "learning_rate": 4.1675692400422286e-05, "loss": 0.0187, "step": 28780 }, { "epoch": 10.572897539478516, "grad_norm": 0.2587411403656006, "learning_rate": 4.164308635046304e-05, "loss": 0.0179, "step": 28790 }, { "epoch": 10.576569959603379, "grad_norm": 0.3485160768032074, "learning_rate": 4.161048395610177e-05, "loss": 0.0172, "step": 28800 }, { "epoch": 10.58024237972824, "grad_norm": 0.305112361907959, "learning_rate": 4.157788523159985e-05, "loss": 0.0181, "step": 28810 }, { "epoch": 10.583914799853103, "grad_norm": 0.26900506019592285, "learning_rate": 4.154529019121706e-05, "loss": 0.0184, "step": 28820 }, { "epoch": 10.587587219977966, "grad_norm": 0.3031077980995178, "learning_rate": 4.1512698849211564e-05, "loss": 0.0161, "step": 28830 }, { "epoch": 10.591259640102828, "grad_norm": 0.25834232568740845, "learning_rate": 4.1480111219839954e-05, "loss": 0.017, "step": 28840 }, { "epoch": 10.59493206022769, "grad_norm": 0.1861768513917923, "learning_rate": 4.144752731735714e-05, "loss": 0.0182, "step": 28850 }, { "epoch": 10.598604480352552, "grad_norm": 0.22138343751430511, "learning_rate": 4.1414947156016446e-05, "loss": 0.0178, "step": 28860 }, { "epoch": 10.602276900477415, "grad_norm": 0.2321632206439972, "learning_rate": 4.138237075006953e-05, "loss": 0.0182, "step": 28870 }, { "epoch": 10.605949320602276, "grad_norm": 0.31264063715934753, "learning_rate": 4.134979811376645e-05, "loss": 0.0181, "step": 28880 }, { "epoch": 10.609621740727139, "grad_norm": 0.20999574661254883, "learning_rate": 4.131722926135552e-05, "loss": 0.0168, "step": 28890 }, { "epoch": 10.613294160852002, "grad_norm": 0.2222721129655838, "learning_rate": 4.1284664207083526e-05, "loss": 0.0167, "step": 28900 }, { "epoch": 10.616966580976865, "grad_norm": 0.3608798086643219, "learning_rate": 4.125210296519551e-05, "loss": 0.018, "step": 28910 }, { "epoch": 10.620639001101726, "grad_norm": 0.40586933493614197, "learning_rate": 4.121954554993484e-05, "loss": 0.0182, "step": 28920 }, { "epoch": 10.624311421226588, "grad_norm": 0.2532545328140259, "learning_rate": 4.118699197554327e-05, "loss": 0.0201, "step": 28930 }, { "epoch": 10.627983841351451, "grad_norm": 0.3116741478443146, "learning_rate": 4.1154442256260814e-05, "loss": 0.0169, "step": 28940 }, { "epoch": 10.631656261476312, "grad_norm": 0.297575980424881, "learning_rate": 4.1121896406325824e-05, "loss": 0.0168, "step": 28950 }, { "epoch": 10.635328681601175, "grad_norm": 0.2481306493282318, "learning_rate": 4.108935443997499e-05, "loss": 0.0165, "step": 28960 }, { "epoch": 10.639001101726038, "grad_norm": 0.2964721620082855, "learning_rate": 4.105681637144325e-05, "loss": 0.0181, "step": 28970 }, { "epoch": 10.6426735218509, "grad_norm": 0.24028247594833374, "learning_rate": 4.102428221496387e-05, "loss": 0.0182, "step": 28980 }, { "epoch": 10.646345941975762, "grad_norm": 0.22107400000095367, "learning_rate": 4.099175198476838e-05, "loss": 0.0209, "step": 28990 }, { "epoch": 10.650018362100624, "grad_norm": 0.23830416798591614, "learning_rate": 4.0959225695086635e-05, "loss": 0.0189, "step": 29000 }, { "epoch": 10.653690782225487, "grad_norm": 0.29962581396102905, "learning_rate": 4.0926703360146724e-05, "loss": 0.0188, "step": 29010 }, { "epoch": 10.657363202350348, "grad_norm": 0.5922450423240662, "learning_rate": 4.0894184994175e-05, "loss": 0.0197, "step": 29020 }, { "epoch": 10.661035622475211, "grad_norm": 0.19404666125774384, "learning_rate": 4.086167061139614e-05, "loss": 0.0167, "step": 29030 }, { "epoch": 10.664708042600074, "grad_norm": 0.2829257845878601, "learning_rate": 4.082916022603303e-05, "loss": 0.0183, "step": 29040 }, { "epoch": 10.668380462724937, "grad_norm": 0.25342968106269836, "learning_rate": 4.079665385230678e-05, "loss": 0.0169, "step": 29050 }, { "epoch": 10.672052882849798, "grad_norm": 0.4339969754219055, "learning_rate": 4.076415150443683e-05, "loss": 0.0187, "step": 29060 }, { "epoch": 10.67572530297466, "grad_norm": 0.2575914263725281, "learning_rate": 4.073165319664079e-05, "loss": 0.0177, "step": 29070 }, { "epoch": 10.679397723099523, "grad_norm": 0.24392718076705933, "learning_rate": 4.069915894313451e-05, "loss": 0.0174, "step": 29080 }, { "epoch": 10.683070143224384, "grad_norm": 0.23307208716869354, "learning_rate": 4.066666875813212e-05, "loss": 0.0193, "step": 29090 }, { "epoch": 10.686742563349247, "grad_norm": 0.2878267765045166, "learning_rate": 4.063418265584589e-05, "loss": 0.0176, "step": 29100 }, { "epoch": 10.69041498347411, "grad_norm": 0.2375117689371109, "learning_rate": 4.060170065048636e-05, "loss": 0.02, "step": 29110 }, { "epoch": 10.69408740359897, "grad_norm": 0.254926860332489, "learning_rate": 4.056922275626227e-05, "loss": 0.0169, "step": 29120 }, { "epoch": 10.697759823723834, "grad_norm": 0.3183646500110626, "learning_rate": 4.053674898738056e-05, "loss": 0.0174, "step": 29130 }, { "epoch": 10.701432243848696, "grad_norm": 0.32828477025032043, "learning_rate": 4.050427935804633e-05, "loss": 0.0175, "step": 29140 }, { "epoch": 10.70510466397356, "grad_norm": 0.286025732755661, "learning_rate": 4.047181388246293e-05, "loss": 0.0165, "step": 29150 }, { "epoch": 10.70877708409842, "grad_norm": 0.27847495675086975, "learning_rate": 4.043935257483185e-05, "loss": 0.0187, "step": 29160 }, { "epoch": 10.712449504223283, "grad_norm": 0.21999497711658478, "learning_rate": 4.0406895449352775e-05, "loss": 0.0185, "step": 29170 }, { "epoch": 10.716121924348146, "grad_norm": 0.5174596309661865, "learning_rate": 4.0374442520223525e-05, "loss": 0.019, "step": 29180 }, { "epoch": 10.719794344473009, "grad_norm": 0.34792667627334595, "learning_rate": 4.0341993801640155e-05, "loss": 0.017, "step": 29190 }, { "epoch": 10.72346676459787, "grad_norm": 0.38226234912872314, "learning_rate": 4.030954930779681e-05, "loss": 0.0184, "step": 29200 }, { "epoch": 10.727139184722732, "grad_norm": 0.2632199823856354, "learning_rate": 4.027710905288582e-05, "loss": 0.0181, "step": 29210 }, { "epoch": 10.730811604847595, "grad_norm": 0.5874730944633484, "learning_rate": 4.0244673051097666e-05, "loss": 0.0193, "step": 29220 }, { "epoch": 10.734484024972456, "grad_norm": 0.2673512399196625, "learning_rate": 4.021224131662095e-05, "loss": 0.0185, "step": 29230 }, { "epoch": 10.738156445097319, "grad_norm": 0.19703102111816406, "learning_rate": 4.0179813863642396e-05, "loss": 0.0173, "step": 29240 }, { "epoch": 10.741828865222182, "grad_norm": 0.4040406048297882, "learning_rate": 4.014739070634691e-05, "loss": 0.0163, "step": 29250 }, { "epoch": 10.745501285347043, "grad_norm": 0.38027235865592957, "learning_rate": 4.011497185891746e-05, "loss": 0.0185, "step": 29260 }, { "epoch": 10.749173705471906, "grad_norm": 0.40633243322372437, "learning_rate": 4.008255733553513e-05, "loss": 0.019, "step": 29270 }, { "epoch": 10.752846125596768, "grad_norm": 0.2077779918909073, "learning_rate": 4.0050147150379194e-05, "loss": 0.0181, "step": 29280 }, { "epoch": 10.756518545721631, "grad_norm": 0.2587442100048065, "learning_rate": 4.001774131762692e-05, "loss": 0.0174, "step": 29290 }, { "epoch": 10.760190965846492, "grad_norm": 0.44945192337036133, "learning_rate": 3.9985339851453705e-05, "loss": 0.0175, "step": 29300 }, { "epoch": 10.763863385971355, "grad_norm": 0.32615166902542114, "learning_rate": 3.9952942766033114e-05, "loss": 0.0189, "step": 29310 }, { "epoch": 10.767535806096218, "grad_norm": 0.17920078337192535, "learning_rate": 3.9920550075536686e-05, "loss": 0.016, "step": 29320 }, { "epoch": 10.771208226221079, "grad_norm": 0.22978457808494568, "learning_rate": 3.988816179413408e-05, "loss": 0.0186, "step": 29330 }, { "epoch": 10.774880646345942, "grad_norm": 0.2628881633281708, "learning_rate": 3.985577793599307e-05, "loss": 0.0174, "step": 29340 }, { "epoch": 10.778553066470804, "grad_norm": 0.33043134212493896, "learning_rate": 3.982339851527942e-05, "loss": 0.0176, "step": 29350 }, { "epoch": 10.782225486595667, "grad_norm": 0.39924705028533936, "learning_rate": 3.979102354615701e-05, "loss": 0.0169, "step": 29360 }, { "epoch": 10.785897906720528, "grad_norm": 0.3019220232963562, "learning_rate": 3.975865304278772e-05, "loss": 0.0163, "step": 29370 }, { "epoch": 10.789570326845391, "grad_norm": 0.2149813175201416, "learning_rate": 3.9726287019331546e-05, "loss": 0.0188, "step": 29380 }, { "epoch": 10.793242746970254, "grad_norm": 0.6655194163322449, "learning_rate": 3.9693925489946474e-05, "loss": 0.0167, "step": 29390 }, { "epoch": 10.796915167095115, "grad_norm": 0.3212476074695587, "learning_rate": 3.966156846878852e-05, "loss": 0.0183, "step": 29400 }, { "epoch": 10.800587587219978, "grad_norm": 0.26396405696868896, "learning_rate": 3.962921597001177e-05, "loss": 0.0167, "step": 29410 }, { "epoch": 10.80426000734484, "grad_norm": 0.36970824003219604, "learning_rate": 3.9596868007768294e-05, "loss": 0.0163, "step": 29420 }, { "epoch": 10.807932427469703, "grad_norm": 0.8731439709663391, "learning_rate": 3.9564524596208165e-05, "loss": 0.0187, "step": 29430 }, { "epoch": 10.811604847594564, "grad_norm": 0.26976972818374634, "learning_rate": 3.953218574947954e-05, "loss": 0.0165, "step": 29440 }, { "epoch": 10.815277267719427, "grad_norm": 0.3003588914871216, "learning_rate": 3.949985148172849e-05, "loss": 0.0166, "step": 29450 }, { "epoch": 10.81894968784429, "grad_norm": 0.2916804254055023, "learning_rate": 3.946752180709914e-05, "loss": 0.0185, "step": 29460 }, { "epoch": 10.822622107969151, "grad_norm": 0.521664559841156, "learning_rate": 3.943519673973357e-05, "loss": 0.0176, "step": 29470 }, { "epoch": 10.826294528094014, "grad_norm": 0.35168710350990295, "learning_rate": 3.940287629377189e-05, "loss": 0.0175, "step": 29480 }, { "epoch": 10.829966948218877, "grad_norm": 0.6382197737693787, "learning_rate": 3.937056048335214e-05, "loss": 0.0176, "step": 29490 }, { "epoch": 10.83363936834374, "grad_norm": 0.32534632086753845, "learning_rate": 3.933824932261037e-05, "loss": 0.0178, "step": 29500 }, { "epoch": 10.8373117884686, "grad_norm": 0.2071135938167572, "learning_rate": 3.9305942825680576e-05, "loss": 0.0186, "step": 29510 }, { "epoch": 10.840984208593463, "grad_norm": 0.3139011859893799, "learning_rate": 3.9273641006694706e-05, "loss": 0.0181, "step": 29520 }, { "epoch": 10.844656628718326, "grad_norm": 0.23381240665912628, "learning_rate": 3.924134387978266e-05, "loss": 0.0171, "step": 29530 }, { "epoch": 10.848329048843187, "grad_norm": 0.3943120539188385, "learning_rate": 3.920905145907234e-05, "loss": 0.0161, "step": 29540 }, { "epoch": 10.85200146896805, "grad_norm": 0.3313505947589874, "learning_rate": 3.9176763758689525e-05, "loss": 0.0165, "step": 29550 }, { "epoch": 10.855673889092913, "grad_norm": 0.3046749532222748, "learning_rate": 3.9144480792757925e-05, "loss": 0.0179, "step": 29560 }, { "epoch": 10.859346309217775, "grad_norm": 0.2643473148345947, "learning_rate": 3.911220257539926e-05, "loss": 0.0165, "step": 29570 }, { "epoch": 10.863018729342636, "grad_norm": 0.30565497279167175, "learning_rate": 3.907992912073308e-05, "loss": 0.0182, "step": 29580 }, { "epoch": 10.8666911494675, "grad_norm": 0.38567689061164856, "learning_rate": 3.90476604428769e-05, "loss": 0.0189, "step": 29590 }, { "epoch": 10.870363569592362, "grad_norm": 0.2648216784000397, "learning_rate": 3.901539655594614e-05, "loss": 0.0195, "step": 29600 }, { "epoch": 10.874035989717223, "grad_norm": 0.3455870449542999, "learning_rate": 3.898313747405413e-05, "loss": 0.0182, "step": 29610 }, { "epoch": 10.877708409842086, "grad_norm": 0.19517509639263153, "learning_rate": 3.895088321131205e-05, "loss": 0.0162, "step": 29620 }, { "epoch": 10.881380829966949, "grad_norm": 0.6585201621055603, "learning_rate": 3.891863378182907e-05, "loss": 0.0171, "step": 29630 }, { "epoch": 10.885053250091811, "grad_norm": 0.2763679623603821, "learning_rate": 3.8886389199712145e-05, "loss": 0.0169, "step": 29640 }, { "epoch": 10.888725670216672, "grad_norm": 0.2349582016468048, "learning_rate": 3.885414947906614e-05, "loss": 0.0179, "step": 29650 }, { "epoch": 10.892398090341535, "grad_norm": 0.3203451335430145, "learning_rate": 3.882191463399385e-05, "loss": 0.0182, "step": 29660 }, { "epoch": 10.896070510466398, "grad_norm": 0.3383379578590393, "learning_rate": 3.878968467859585e-05, "loss": 0.0176, "step": 29670 }, { "epoch": 10.899742930591259, "grad_norm": 0.20553506910800934, "learning_rate": 3.875745962697063e-05, "loss": 0.0178, "step": 29680 }, { "epoch": 10.903415350716122, "grad_norm": 0.23806388676166534, "learning_rate": 3.872523949321454e-05, "loss": 0.0168, "step": 29690 }, { "epoch": 10.907087770840985, "grad_norm": 0.16010917723178864, "learning_rate": 3.869302429142173e-05, "loss": 0.0161, "step": 29700 }, { "epoch": 10.910760190965846, "grad_norm": 0.25719353556632996, "learning_rate": 3.866081403568425e-05, "loss": 0.0169, "step": 29710 }, { "epoch": 10.914432611090708, "grad_norm": 0.35939398407936096, "learning_rate": 3.8628608740091934e-05, "loss": 0.0183, "step": 29720 }, { "epoch": 10.918105031215571, "grad_norm": 0.19458772242069244, "learning_rate": 3.85964084187325e-05, "loss": 0.017, "step": 29730 }, { "epoch": 10.921777451340434, "grad_norm": 0.2065858244895935, "learning_rate": 3.856421308569146e-05, "loss": 0.0176, "step": 29740 }, { "epoch": 10.925449871465295, "grad_norm": 0.32664379477500916, "learning_rate": 3.853202275505212e-05, "loss": 0.018, "step": 29750 }, { "epoch": 10.929122291590158, "grad_norm": 0.284491628408432, "learning_rate": 3.849983744089565e-05, "loss": 0.019, "step": 29760 }, { "epoch": 10.93279471171502, "grad_norm": 0.22081634402275085, "learning_rate": 3.8467657157301e-05, "loss": 0.018, "step": 29770 }, { "epoch": 10.936467131839883, "grad_norm": 0.2378135770559311, "learning_rate": 3.8435481918344884e-05, "loss": 0.0171, "step": 29780 }, { "epoch": 10.940139551964744, "grad_norm": 0.29393741488456726, "learning_rate": 3.84033117381019e-05, "loss": 0.0172, "step": 29790 }, { "epoch": 10.943811972089607, "grad_norm": 0.21167677640914917, "learning_rate": 3.8371146630644335e-05, "loss": 0.0189, "step": 29800 }, { "epoch": 10.94748439221447, "grad_norm": 0.34783023595809937, "learning_rate": 3.833898661004231e-05, "loss": 0.0176, "step": 29810 }, { "epoch": 10.951156812339331, "grad_norm": 0.22845807671546936, "learning_rate": 3.830683169036372e-05, "loss": 0.0183, "step": 29820 }, { "epoch": 10.954829232464194, "grad_norm": 0.32073497772216797, "learning_rate": 3.827468188567422e-05, "loss": 0.0172, "step": 29830 }, { "epoch": 10.958501652589057, "grad_norm": 0.28694984316825867, "learning_rate": 3.82425372100372e-05, "loss": 0.0199, "step": 29840 }, { "epoch": 10.962174072713918, "grad_norm": 0.2614661753177643, "learning_rate": 3.821039767751387e-05, "loss": 0.0187, "step": 29850 }, { "epoch": 10.96584649283878, "grad_norm": 0.4236642122268677, "learning_rate": 3.817826330216315e-05, "loss": 0.0192, "step": 29860 }, { "epoch": 10.969518912963643, "grad_norm": 0.45395350456237793, "learning_rate": 3.814613409804169e-05, "loss": 0.0187, "step": 29870 }, { "epoch": 10.973191333088506, "grad_norm": 0.21339669823646545, "learning_rate": 3.811401007920388e-05, "loss": 0.0175, "step": 29880 }, { "epoch": 10.976863753213367, "grad_norm": 0.240875244140625, "learning_rate": 3.8081891259701895e-05, "loss": 0.0161, "step": 29890 }, { "epoch": 10.98053617333823, "grad_norm": 0.3237466812133789, "learning_rate": 3.8049777653585584e-05, "loss": 0.0181, "step": 29900 }, { "epoch": 10.984208593463093, "grad_norm": 0.4806617498397827, "learning_rate": 3.80176692749025e-05, "loss": 0.0172, "step": 29910 }, { "epoch": 10.987881013587954, "grad_norm": 0.2686917185783386, "learning_rate": 3.798556613769799e-05, "loss": 0.0162, "step": 29920 }, { "epoch": 10.991553433712816, "grad_norm": 0.2210841327905655, "learning_rate": 3.795346825601502e-05, "loss": 0.0177, "step": 29930 }, { "epoch": 10.99522585383768, "grad_norm": 0.35858628153800964, "learning_rate": 3.7921375643894305e-05, "loss": 0.0171, "step": 29940 }, { "epoch": 10.998898273962542, "grad_norm": 0.2971234917640686, "learning_rate": 3.788928831537424e-05, "loss": 0.018, "step": 29950 }, { "epoch": 11.002570694087403, "grad_norm": 0.27050384879112244, "learning_rate": 3.7857206284490934e-05, "loss": 0.0177, "step": 29960 }, { "epoch": 11.006243114212266, "grad_norm": 0.18105681240558624, "learning_rate": 3.7825129565278114e-05, "loss": 0.0181, "step": 29970 }, { "epoch": 11.009915534337129, "grad_norm": 0.21155673265457153, "learning_rate": 3.779305817176728e-05, "loss": 0.0167, "step": 29980 }, { "epoch": 11.01358795446199, "grad_norm": 0.2310904860496521, "learning_rate": 3.776099211798751e-05, "loss": 0.0172, "step": 29990 }, { "epoch": 11.017260374586852, "grad_norm": 0.22527939081192017, "learning_rate": 3.7728931417965584e-05, "loss": 0.0178, "step": 30000 }, { "epoch": 11.020932794711715, "grad_norm": 0.2975407838821411, "learning_rate": 3.769687608572598e-05, "loss": 0.0199, "step": 30010 }, { "epoch": 11.024605214836578, "grad_norm": 0.2984549105167389, "learning_rate": 3.766482613529076e-05, "loss": 0.0177, "step": 30020 }, { "epoch": 11.028277634961439, "grad_norm": 0.3633682131767273, "learning_rate": 3.763278158067964e-05, "loss": 0.0171, "step": 30030 }, { "epoch": 11.031950055086302, "grad_norm": 0.21590697765350342, "learning_rate": 3.7600742435910045e-05, "loss": 0.0195, "step": 30040 }, { "epoch": 11.035622475211165, "grad_norm": 0.5518885254859924, "learning_rate": 3.756870871499696e-05, "loss": 0.0173, "step": 30050 }, { "epoch": 11.039294895336026, "grad_norm": 0.3477107584476471, "learning_rate": 3.7536680431953025e-05, "loss": 0.0182, "step": 30060 }, { "epoch": 11.042967315460888, "grad_norm": 0.19859236478805542, "learning_rate": 3.7504657600788485e-05, "loss": 0.0179, "step": 30070 }, { "epoch": 11.046639735585751, "grad_norm": 0.24750880897045135, "learning_rate": 3.747264023551125e-05, "loss": 0.0166, "step": 30080 }, { "epoch": 11.050312155710614, "grad_norm": 0.2585533559322357, "learning_rate": 3.744062835012678e-05, "loss": 0.018, "step": 30090 }, { "epoch": 11.053984575835475, "grad_norm": 0.2510918080806732, "learning_rate": 3.740862195863814e-05, "loss": 0.0187, "step": 30100 }, { "epoch": 11.057656995960338, "grad_norm": 0.23708362877368927, "learning_rate": 3.737662107504606e-05, "loss": 0.0183, "step": 30110 }, { "epoch": 11.0613294160852, "grad_norm": 0.26383817195892334, "learning_rate": 3.734462571334879e-05, "loss": 0.0176, "step": 30120 }, { "epoch": 11.065001836210062, "grad_norm": 0.39843109250068665, "learning_rate": 3.731263588754217e-05, "loss": 0.0185, "step": 30130 }, { "epoch": 11.068674256334925, "grad_norm": 0.270707905292511, "learning_rate": 3.7280651611619665e-05, "loss": 0.0196, "step": 30140 }, { "epoch": 11.072346676459787, "grad_norm": 0.544576108455658, "learning_rate": 3.724867289957229e-05, "loss": 0.0189, "step": 30150 }, { "epoch": 11.07601909658465, "grad_norm": 0.26770052313804626, "learning_rate": 3.7216699765388576e-05, "loss": 0.0176, "step": 30160 }, { "epoch": 11.079691516709511, "grad_norm": 0.17117998003959656, "learning_rate": 3.71847322230547e-05, "loss": 0.0175, "step": 30170 }, { "epoch": 11.083363936834374, "grad_norm": 0.3760446310043335, "learning_rate": 3.7152770286554335e-05, "loss": 0.0204, "step": 30180 }, { "epoch": 11.087036356959237, "grad_norm": 0.22709454596042633, "learning_rate": 3.712081396986872e-05, "loss": 0.0174, "step": 30190 }, { "epoch": 11.090708777084098, "grad_norm": 0.22699420154094696, "learning_rate": 3.708886328697663e-05, "loss": 0.0184, "step": 30200 }, { "epoch": 11.09438119720896, "grad_norm": 0.3546546399593353, "learning_rate": 3.7056918251854395e-05, "loss": 0.0178, "step": 30210 }, { "epoch": 11.098053617333823, "grad_norm": 0.2972567677497864, "learning_rate": 3.7024978878475846e-05, "loss": 0.0194, "step": 30220 }, { "epoch": 11.101726037458686, "grad_norm": 0.32714906334877014, "learning_rate": 3.699304518081233e-05, "loss": 0.017, "step": 30230 }, { "epoch": 11.105398457583547, "grad_norm": 0.2149687260389328, "learning_rate": 3.696111717283277e-05, "loss": 0.0166, "step": 30240 }, { "epoch": 11.10907087770841, "grad_norm": 0.5443338751792908, "learning_rate": 3.692919486850355e-05, "loss": 0.0166, "step": 30250 }, { "epoch": 11.112743297833273, "grad_norm": 0.2889995276927948, "learning_rate": 3.689727828178854e-05, "loss": 0.0161, "step": 30260 }, { "epoch": 11.116415717958134, "grad_norm": 0.279969722032547, "learning_rate": 3.6865367426649185e-05, "loss": 0.0186, "step": 30270 }, { "epoch": 11.120088138082997, "grad_norm": 0.1946660429239273, "learning_rate": 3.6833462317044346e-05, "loss": 0.0179, "step": 30280 }, { "epoch": 11.12376055820786, "grad_norm": 0.19751077890396118, "learning_rate": 3.6801562966930406e-05, "loss": 0.0164, "step": 30290 }, { "epoch": 11.12743297833272, "grad_norm": 0.5398383140563965, "learning_rate": 3.676966939026123e-05, "loss": 0.0167, "step": 30300 }, { "epoch": 11.131105398457583, "grad_norm": 0.23436270654201508, "learning_rate": 3.6737781600988166e-05, "loss": 0.0172, "step": 30310 }, { "epoch": 11.134777818582446, "grad_norm": 0.28069567680358887, "learning_rate": 3.670589961305998e-05, "loss": 0.016, "step": 30320 }, { "epoch": 11.138450238707309, "grad_norm": 0.23715856671333313, "learning_rate": 3.667402344042298e-05, "loss": 0.0184, "step": 30330 }, { "epoch": 11.14212265883217, "grad_norm": 0.22114722430706024, "learning_rate": 3.664215309702086e-05, "loss": 0.0177, "step": 30340 }, { "epoch": 11.145795078957033, "grad_norm": 0.27572157979011536, "learning_rate": 3.661028859679478e-05, "loss": 0.016, "step": 30350 }, { "epoch": 11.149467499081895, "grad_norm": 0.3118865489959717, "learning_rate": 3.657842995368339e-05, "loss": 0.0167, "step": 30360 }, { "epoch": 11.153139919206756, "grad_norm": 0.31214258074760437, "learning_rate": 3.654657718162273e-05, "loss": 0.0165, "step": 30370 }, { "epoch": 11.15681233933162, "grad_norm": 0.4255649745464325, "learning_rate": 3.6514730294546275e-05, "loss": 0.0183, "step": 30380 }, { "epoch": 11.160484759456482, "grad_norm": 0.380187451839447, "learning_rate": 3.648288930638492e-05, "loss": 0.0171, "step": 30390 }, { "epoch": 11.164157179581345, "grad_norm": 0.253572940826416, "learning_rate": 3.645105423106704e-05, "loss": 0.0172, "step": 30400 }, { "epoch": 11.167829599706206, "grad_norm": 0.323063462972641, "learning_rate": 3.6419225082518336e-05, "loss": 0.0164, "step": 30410 }, { "epoch": 11.171502019831069, "grad_norm": 0.2938285768032074, "learning_rate": 3.6387401874661975e-05, "loss": 0.0177, "step": 30420 }, { "epoch": 11.175174439955931, "grad_norm": 0.18445171415805817, "learning_rate": 3.6355584621418516e-05, "loss": 0.0175, "step": 30430 }, { "epoch": 11.178846860080792, "grad_norm": 0.23936812579631805, "learning_rate": 3.632377333670591e-05, "loss": 0.0178, "step": 30440 }, { "epoch": 11.182519280205655, "grad_norm": 0.22177420556545258, "learning_rate": 3.629196803443946e-05, "loss": 0.0175, "step": 30450 }, { "epoch": 11.186191700330518, "grad_norm": 0.283186137676239, "learning_rate": 3.626016872853193e-05, "loss": 0.017, "step": 30460 }, { "epoch": 11.18986412045538, "grad_norm": 0.26658543944358826, "learning_rate": 3.6228375432893405e-05, "loss": 0.0183, "step": 30470 }, { "epoch": 11.193536540580242, "grad_norm": 0.2969624400138855, "learning_rate": 3.619658816143131e-05, "loss": 0.0168, "step": 30480 }, { "epoch": 11.197208960705105, "grad_norm": 0.22323006391525269, "learning_rate": 3.6164806928050544e-05, "loss": 0.0182, "step": 30490 }, { "epoch": 11.200881380829967, "grad_norm": 0.3082468509674072, "learning_rate": 3.613303174665327e-05, "loss": 0.0172, "step": 30500 }, { "epoch": 11.204553800954828, "grad_norm": 0.16378255188465118, "learning_rate": 3.6101262631138996e-05, "loss": 0.017, "step": 30510 }, { "epoch": 11.208226221079691, "grad_norm": 0.3095082938671112, "learning_rate": 3.606949959540467e-05, "loss": 0.0171, "step": 30520 }, { "epoch": 11.211898641204554, "grad_norm": 0.3269926905632019, "learning_rate": 3.603774265334449e-05, "loss": 0.0188, "step": 30530 }, { "epoch": 11.215571061329417, "grad_norm": 0.2626449167728424, "learning_rate": 3.600599181885003e-05, "loss": 0.0161, "step": 30540 }, { "epoch": 11.219243481454278, "grad_norm": 0.3361591100692749, "learning_rate": 3.597424710581018e-05, "loss": 0.0181, "step": 30550 }, { "epoch": 11.22291590157914, "grad_norm": 0.22118239104747772, "learning_rate": 3.5942508528111164e-05, "loss": 0.0165, "step": 30560 }, { "epoch": 11.226588321704003, "grad_norm": 0.3341403305530548, "learning_rate": 3.591077609963651e-05, "loss": 0.017, "step": 30570 }, { "epoch": 11.230260741828864, "grad_norm": 0.5389223098754883, "learning_rate": 3.587904983426704e-05, "loss": 0.0167, "step": 30580 }, { "epoch": 11.233933161953727, "grad_norm": 0.24478165805339813, "learning_rate": 3.584732974588094e-05, "loss": 0.0181, "step": 30590 }, { "epoch": 11.23760558207859, "grad_norm": 0.30873435735702515, "learning_rate": 3.5815615848353645e-05, "loss": 0.0181, "step": 30600 }, { "epoch": 11.241278002203453, "grad_norm": 0.2375093251466751, "learning_rate": 3.578390815555784e-05, "loss": 0.0173, "step": 30610 }, { "epoch": 11.244950422328314, "grad_norm": 0.15187951922416687, "learning_rate": 3.5752206681363626e-05, "loss": 0.0153, "step": 30620 }, { "epoch": 11.248622842453177, "grad_norm": 0.23065589368343353, "learning_rate": 3.572051143963826e-05, "loss": 0.0166, "step": 30630 }, { "epoch": 11.25229526257804, "grad_norm": 1.0516819953918457, "learning_rate": 3.5688822444246294e-05, "loss": 0.0149, "step": 30640 }, { "epoch": 11.2559676827029, "grad_norm": 0.23905658721923828, "learning_rate": 3.565713970904963e-05, "loss": 0.0165, "step": 30650 }, { "epoch": 11.259640102827763, "grad_norm": 0.49958163499832153, "learning_rate": 3.562546324790733e-05, "loss": 0.0181, "step": 30660 }, { "epoch": 11.263312522952626, "grad_norm": 0.4414450526237488, "learning_rate": 3.559379307467577e-05, "loss": 0.0166, "step": 30670 }, { "epoch": 11.266984943077489, "grad_norm": 0.2674228250980377, "learning_rate": 3.556212920320856e-05, "loss": 0.0175, "step": 30680 }, { "epoch": 11.27065736320235, "grad_norm": 0.2931108772754669, "learning_rate": 3.553047164735655e-05, "loss": 0.0167, "step": 30690 }, { "epoch": 11.274329783327213, "grad_norm": 0.323974609375, "learning_rate": 3.549882042096783e-05, "loss": 0.017, "step": 30700 }, { "epoch": 11.278002203452075, "grad_norm": 0.25204187631607056, "learning_rate": 3.546717553788772e-05, "loss": 0.0166, "step": 30710 }, { "epoch": 11.281674623576937, "grad_norm": 0.22417280077934265, "learning_rate": 3.543553701195879e-05, "loss": 0.0177, "step": 30720 }, { "epoch": 11.2853470437018, "grad_norm": 0.4621465504169464, "learning_rate": 3.540390485702077e-05, "loss": 0.0194, "step": 30730 }, { "epoch": 11.289019463826662, "grad_norm": 0.29419127106666565, "learning_rate": 3.5372279086910645e-05, "loss": 0.0172, "step": 30740 }, { "epoch": 11.292691883951525, "grad_norm": 0.29491880536079407, "learning_rate": 3.5340659715462635e-05, "loss": 0.0181, "step": 30750 }, { "epoch": 11.296364304076386, "grad_norm": 0.2096215784549713, "learning_rate": 3.530904675650808e-05, "loss": 0.0161, "step": 30760 }, { "epoch": 11.300036724201249, "grad_norm": 0.23987899720668793, "learning_rate": 3.527744022387559e-05, "loss": 0.0204, "step": 30770 }, { "epoch": 11.303709144326112, "grad_norm": 0.2857324481010437, "learning_rate": 3.524584013139094e-05, "loss": 0.018, "step": 30780 }, { "epoch": 11.307381564450973, "grad_norm": 0.3287353813648224, "learning_rate": 3.521424649287708e-05, "loss": 0.0196, "step": 30790 }, { "epoch": 11.311053984575835, "grad_norm": 0.21246999502182007, "learning_rate": 3.518265932215412e-05, "loss": 0.0166, "step": 30800 }, { "epoch": 11.314726404700698, "grad_norm": 0.36162933707237244, "learning_rate": 3.51510786330394e-05, "loss": 0.0162, "step": 30810 }, { "epoch": 11.318398824825561, "grad_norm": 0.2500694692134857, "learning_rate": 3.511950443934736e-05, "loss": 0.0164, "step": 30820 }, { "epoch": 11.322071244950422, "grad_norm": 0.39491626620292664, "learning_rate": 3.508793675488961e-05, "loss": 0.0197, "step": 30830 }, { "epoch": 11.325743665075285, "grad_norm": 0.3527678847312927, "learning_rate": 3.505637559347498e-05, "loss": 0.0193, "step": 30840 }, { "epoch": 11.329416085200148, "grad_norm": 0.38018742203712463, "learning_rate": 3.5024820968909345e-05, "loss": 0.0171, "step": 30850 }, { "epoch": 11.333088505325009, "grad_norm": 0.5603693127632141, "learning_rate": 3.4993272894995774e-05, "loss": 0.0156, "step": 30860 }, { "epoch": 11.336760925449871, "grad_norm": 0.3248864710330963, "learning_rate": 3.49617313855345e-05, "loss": 0.0191, "step": 30870 }, { "epoch": 11.340433345574734, "grad_norm": 0.2089538723230362, "learning_rate": 3.4930196454322834e-05, "loss": 0.0166, "step": 30880 }, { "epoch": 11.344105765699595, "grad_norm": 0.42003682255744934, "learning_rate": 3.48986681151552e-05, "loss": 0.0172, "step": 30890 }, { "epoch": 11.347778185824458, "grad_norm": 0.37269100546836853, "learning_rate": 3.4867146381823204e-05, "loss": 0.0161, "step": 30900 }, { "epoch": 11.35145060594932, "grad_norm": 0.5144019722938538, "learning_rate": 3.48356312681155e-05, "loss": 0.0166, "step": 30910 }, { "epoch": 11.355123026074184, "grad_norm": 0.1873856484889984, "learning_rate": 3.480412278781788e-05, "loss": 0.0163, "step": 30920 }, { "epoch": 11.358795446199045, "grad_norm": 0.4087476134300232, "learning_rate": 3.47726209547132e-05, "loss": 0.0155, "step": 30930 }, { "epoch": 11.362467866323907, "grad_norm": 0.31624770164489746, "learning_rate": 3.474112578258147e-05, "loss": 0.0147, "step": 30940 }, { "epoch": 11.36614028644877, "grad_norm": 0.1864752322435379, "learning_rate": 3.470963728519974e-05, "loss": 0.0195, "step": 30950 }, { "epoch": 11.369812706573631, "grad_norm": 0.27404889464378357, "learning_rate": 3.467815547634211e-05, "loss": 0.0165, "step": 30960 }, { "epoch": 11.373485126698494, "grad_norm": 0.5025901794433594, "learning_rate": 3.4646680369779836e-05, "loss": 0.018, "step": 30970 }, { "epoch": 11.377157546823357, "grad_norm": 0.3378005027770996, "learning_rate": 3.46152119792812e-05, "loss": 0.017, "step": 30980 }, { "epoch": 11.38082996694822, "grad_norm": 0.16188101470470428, "learning_rate": 3.45837503186115e-05, "loss": 0.0188, "step": 30990 }, { "epoch": 11.38450238707308, "grad_norm": 0.24022524058818817, "learning_rate": 3.455229540153318e-05, "loss": 0.0175, "step": 31000 }, { "epoch": 11.388174807197943, "grad_norm": 0.2624332010746002, "learning_rate": 3.4520847241805665e-05, "loss": 0.0168, "step": 31010 }, { "epoch": 11.391847227322806, "grad_norm": 0.2556271255016327, "learning_rate": 3.4489405853185465e-05, "loss": 0.0162, "step": 31020 }, { "epoch": 11.395519647447667, "grad_norm": 0.26845988631248474, "learning_rate": 3.44579712494261e-05, "loss": 0.0172, "step": 31030 }, { "epoch": 11.39919206757253, "grad_norm": 0.4306483268737793, "learning_rate": 3.442654344427814e-05, "loss": 0.0174, "step": 31040 }, { "epoch": 11.402864487697393, "grad_norm": 0.3866557776927948, "learning_rate": 3.439512245148916e-05, "loss": 0.0167, "step": 31050 }, { "epoch": 11.406536907822256, "grad_norm": 0.3291987180709839, "learning_rate": 3.4363708284803796e-05, "loss": 0.0172, "step": 31060 }, { "epoch": 11.410209327947117, "grad_norm": 0.2251775711774826, "learning_rate": 3.433230095796367e-05, "loss": 0.0173, "step": 31070 }, { "epoch": 11.41388174807198, "grad_norm": 0.3117924630641937, "learning_rate": 3.430090048470739e-05, "loss": 0.0163, "step": 31080 }, { "epoch": 11.417554168196842, "grad_norm": 0.30334073305130005, "learning_rate": 3.4269506878770585e-05, "loss": 0.0185, "step": 31090 }, { "epoch": 11.421226588321703, "grad_norm": 0.17817312479019165, "learning_rate": 3.4238120153885925e-05, "loss": 0.0158, "step": 31100 }, { "epoch": 11.424899008446566, "grad_norm": 0.296842485666275, "learning_rate": 3.4206740323782994e-05, "loss": 0.0204, "step": 31110 }, { "epoch": 11.428571428571429, "grad_norm": 0.2999389171600342, "learning_rate": 3.4175367402188395e-05, "loss": 0.017, "step": 31120 }, { "epoch": 11.432243848696292, "grad_norm": 0.20100966095924377, "learning_rate": 3.414400140282574e-05, "loss": 0.0169, "step": 31130 }, { "epoch": 11.435916268821153, "grad_norm": 0.35031720995903015, "learning_rate": 3.411264233941556e-05, "loss": 0.0176, "step": 31140 }, { "epoch": 11.439588688946015, "grad_norm": 0.25864195823669434, "learning_rate": 3.408129022567537e-05, "loss": 0.0184, "step": 31150 }, { "epoch": 11.443261109070878, "grad_norm": 0.24337375164031982, "learning_rate": 3.404994507531968e-05, "loss": 0.0157, "step": 31160 }, { "epoch": 11.44693352919574, "grad_norm": 0.18845468759536743, "learning_rate": 3.40186069020599e-05, "loss": 0.0166, "step": 31170 }, { "epoch": 11.450605949320602, "grad_norm": 0.2660517990589142, "learning_rate": 3.3987275719604386e-05, "loss": 0.016, "step": 31180 }, { "epoch": 11.454278369445465, "grad_norm": 0.3155783414840698, "learning_rate": 3.3955951541658526e-05, "loss": 0.0151, "step": 31190 }, { "epoch": 11.457950789570328, "grad_norm": 0.3940323293209076, "learning_rate": 3.392463438192454e-05, "loss": 0.016, "step": 31200 }, { "epoch": 11.461623209695189, "grad_norm": 0.3502984642982483, "learning_rate": 3.38933242541016e-05, "loss": 0.0165, "step": 31210 }, { "epoch": 11.465295629820051, "grad_norm": 0.3672579526901245, "learning_rate": 3.386202117188586e-05, "loss": 0.0173, "step": 31220 }, { "epoch": 11.468968049944914, "grad_norm": 0.24269354343414307, "learning_rate": 3.383072514897033e-05, "loss": 0.0167, "step": 31230 }, { "epoch": 11.472640470069775, "grad_norm": 0.24964560568332672, "learning_rate": 3.379943619904494e-05, "loss": 0.018, "step": 31240 }, { "epoch": 11.476312890194638, "grad_norm": 0.2918854355812073, "learning_rate": 3.3768154335796583e-05, "loss": 0.017, "step": 31250 }, { "epoch": 11.4799853103195, "grad_norm": 0.27354854345321655, "learning_rate": 3.3736879572908963e-05, "loss": 0.0156, "step": 31260 }, { "epoch": 11.483657730444364, "grad_norm": 0.4532611668109894, "learning_rate": 3.370561192406276e-05, "loss": 0.0166, "step": 31270 }, { "epoch": 11.487330150569225, "grad_norm": 0.23028971254825592, "learning_rate": 3.367435140293547e-05, "loss": 0.0174, "step": 31280 }, { "epoch": 11.491002570694087, "grad_norm": 0.2347053587436676, "learning_rate": 3.3643098023201545e-05, "loss": 0.0168, "step": 31290 }, { "epoch": 11.49467499081895, "grad_norm": 0.2015911042690277, "learning_rate": 3.361185179853225e-05, "loss": 0.0158, "step": 31300 }, { "epoch": 11.498347410943811, "grad_norm": 0.24248504638671875, "learning_rate": 3.358061274259574e-05, "loss": 0.015, "step": 31310 }, { "epoch": 11.502019831068674, "grad_norm": 0.32432061433792114, "learning_rate": 3.3549380869057064e-05, "loss": 0.0173, "step": 31320 }, { "epoch": 11.505692251193537, "grad_norm": 0.2951544523239136, "learning_rate": 3.351815619157809e-05, "loss": 0.0163, "step": 31330 }, { "epoch": 11.509364671318398, "grad_norm": 0.347642183303833, "learning_rate": 3.348693872381754e-05, "loss": 0.0147, "step": 31340 }, { "epoch": 11.51303709144326, "grad_norm": 0.3069751262664795, "learning_rate": 3.3455728479431025e-05, "loss": 0.0152, "step": 31350 }, { "epoch": 11.516709511568124, "grad_norm": 0.2588239908218384, "learning_rate": 3.342452547207094e-05, "loss": 0.0172, "step": 31360 }, { "epoch": 11.520381931692986, "grad_norm": 0.24304816126823425, "learning_rate": 3.3393329715386555e-05, "loss": 0.0178, "step": 31370 }, { "epoch": 11.524054351817847, "grad_norm": 0.18194478750228882, "learning_rate": 3.336214122302395e-05, "loss": 0.0155, "step": 31380 }, { "epoch": 11.52772677194271, "grad_norm": 0.1471649557352066, "learning_rate": 3.333096000862604e-05, "loss": 0.0157, "step": 31390 }, { "epoch": 11.531399192067573, "grad_norm": 0.25844284892082214, "learning_rate": 3.3299786085832516e-05, "loss": 0.0156, "step": 31400 }, { "epoch": 11.535071612192436, "grad_norm": 0.4087688624858856, "learning_rate": 3.3268619468279956e-05, "loss": 0.0198, "step": 31410 }, { "epoch": 11.538744032317297, "grad_norm": 0.3007652461528778, "learning_rate": 3.323746016960167e-05, "loss": 0.0156, "step": 31420 }, { "epoch": 11.54241645244216, "grad_norm": 0.45737653970718384, "learning_rate": 3.3206308203427806e-05, "loss": 0.0155, "step": 31430 }, { "epoch": 11.546088872567022, "grad_norm": 0.46497464179992676, "learning_rate": 3.3175163583385254e-05, "loss": 0.0175, "step": 31440 }, { "epoch": 11.549761292691883, "grad_norm": 0.17459511756896973, "learning_rate": 3.314402632309779e-05, "loss": 0.019, "step": 31450 }, { "epoch": 11.553433712816746, "grad_norm": 0.31115496158599854, "learning_rate": 3.311289643618587e-05, "loss": 0.0178, "step": 31460 }, { "epoch": 11.557106132941609, "grad_norm": 0.17750532925128937, "learning_rate": 3.308177393626676e-05, "loss": 0.0172, "step": 31470 }, { "epoch": 11.56077855306647, "grad_norm": 0.4564632773399353, "learning_rate": 3.305065883695453e-05, "loss": 0.0157, "step": 31480 }, { "epoch": 11.564450973191333, "grad_norm": 0.20430761575698853, "learning_rate": 3.301955115185996e-05, "loss": 0.0182, "step": 31490 }, { "epoch": 11.568123393316196, "grad_norm": 0.2155899852514267, "learning_rate": 3.29884508945906e-05, "loss": 0.0173, "step": 31500 }, { "epoch": 11.571795813441058, "grad_norm": 0.313524454832077, "learning_rate": 3.295735807875079e-05, "loss": 0.0168, "step": 31510 }, { "epoch": 11.57546823356592, "grad_norm": 0.427131712436676, "learning_rate": 3.292627271794156e-05, "loss": 0.0157, "step": 31520 }, { "epoch": 11.579140653690782, "grad_norm": 0.4271935522556305, "learning_rate": 3.289519482576069e-05, "loss": 0.0173, "step": 31530 }, { "epoch": 11.582813073815645, "grad_norm": 0.22171035408973694, "learning_rate": 3.286412441580274e-05, "loss": 0.0166, "step": 31540 }, { "epoch": 11.586485493940506, "grad_norm": 0.3162977993488312, "learning_rate": 3.283306150165896e-05, "loss": 0.0165, "step": 31550 }, { "epoch": 11.590157914065369, "grad_norm": 0.4270020127296448, "learning_rate": 3.2802006096917285e-05, "loss": 0.0176, "step": 31560 }, { "epoch": 11.593830334190232, "grad_norm": 0.21385885775089264, "learning_rate": 3.277095821516245e-05, "loss": 0.0174, "step": 31570 }, { "epoch": 11.597502754315094, "grad_norm": 0.2643318474292755, "learning_rate": 3.273991786997583e-05, "loss": 0.0173, "step": 31580 }, { "epoch": 11.601175174439955, "grad_norm": 0.2960383892059326, "learning_rate": 3.2708885074935514e-05, "loss": 0.0151, "step": 31590 }, { "epoch": 11.604847594564818, "grad_norm": 0.22150734066963196, "learning_rate": 3.267785984361633e-05, "loss": 0.0206, "step": 31600 }, { "epoch": 11.608520014689681, "grad_norm": 0.22767767310142517, "learning_rate": 3.2646842189589754e-05, "loss": 0.0173, "step": 31610 }, { "epoch": 11.612192434814542, "grad_norm": 0.46445226669311523, "learning_rate": 3.261583212642397e-05, "loss": 0.0152, "step": 31620 }, { "epoch": 11.615864854939405, "grad_norm": 0.35689836740493774, "learning_rate": 3.25848296676838e-05, "loss": 0.0164, "step": 31630 }, { "epoch": 11.619537275064268, "grad_norm": 0.22833020985126495, "learning_rate": 3.255383482693083e-05, "loss": 0.0169, "step": 31640 }, { "epoch": 11.62320969518913, "grad_norm": 0.48306721448898315, "learning_rate": 3.252284761772322e-05, "loss": 0.0168, "step": 31650 }, { "epoch": 11.626882115313991, "grad_norm": 0.2352346032857895, "learning_rate": 3.249186805361581e-05, "loss": 0.0182, "step": 31660 }, { "epoch": 11.630554535438854, "grad_norm": 0.2147291600704193, "learning_rate": 3.246089614816016e-05, "loss": 0.0171, "step": 31670 }, { "epoch": 11.634226955563717, "grad_norm": 0.2811402380466461, "learning_rate": 3.242993191490441e-05, "loss": 0.016, "step": 31680 }, { "epoch": 11.637899375688578, "grad_norm": 0.27692508697509766, "learning_rate": 3.239897536739335e-05, "loss": 0.0189, "step": 31690 }, { "epoch": 11.64157179581344, "grad_norm": 0.3477288484573364, "learning_rate": 3.236802651916845e-05, "loss": 0.0173, "step": 31700 }, { "epoch": 11.645244215938304, "grad_norm": 0.3445194661617279, "learning_rate": 3.23370853837678e-05, "loss": 0.0176, "step": 31710 }, { "epoch": 11.648916636063166, "grad_norm": 0.2437494695186615, "learning_rate": 3.2306151974726056e-05, "loss": 0.0189, "step": 31720 }, { "epoch": 11.652589056188027, "grad_norm": 0.4255701005458832, "learning_rate": 3.227522630557459e-05, "loss": 0.0172, "step": 31730 }, { "epoch": 11.65626147631289, "grad_norm": 0.4896620512008667, "learning_rate": 3.224430838984132e-05, "loss": 0.017, "step": 31740 }, { "epoch": 11.659933896437753, "grad_norm": 0.32304495573043823, "learning_rate": 3.2213398241050784e-05, "loss": 0.0169, "step": 31750 }, { "epoch": 11.663606316562614, "grad_norm": 0.21025493741035461, "learning_rate": 3.218249587272415e-05, "loss": 0.017, "step": 31760 }, { "epoch": 11.667278736687477, "grad_norm": 0.5320831537246704, "learning_rate": 3.215160129837916e-05, "loss": 0.0155, "step": 31770 }, { "epoch": 11.67095115681234, "grad_norm": 0.2771301865577698, "learning_rate": 3.2120714531530146e-05, "loss": 0.0179, "step": 31780 }, { "epoch": 11.674623576937202, "grad_norm": 0.428880512714386, "learning_rate": 3.2089835585688014e-05, "loss": 0.0169, "step": 31790 }, { "epoch": 11.678295997062063, "grad_norm": 0.3549007773399353, "learning_rate": 3.20589644743603e-05, "loss": 0.0178, "step": 31800 }, { "epoch": 11.681968417186926, "grad_norm": 0.3179112374782562, "learning_rate": 3.2028101211051054e-05, "loss": 0.0148, "step": 31810 }, { "epoch": 11.685640837311789, "grad_norm": 0.21402715146541595, "learning_rate": 3.19972458092609e-05, "loss": 0.017, "step": 31820 }, { "epoch": 11.68931325743665, "grad_norm": 0.2526879608631134, "learning_rate": 3.196639828248707e-05, "loss": 0.017, "step": 31830 }, { "epoch": 11.692985677561513, "grad_norm": 0.32158491015434265, "learning_rate": 3.193555864422329e-05, "loss": 0.0163, "step": 31840 }, { "epoch": 11.696658097686376, "grad_norm": 0.2070973664522171, "learning_rate": 3.19047269079599e-05, "loss": 0.0156, "step": 31850 }, { "epoch": 11.700330517811238, "grad_norm": 0.2868635654449463, "learning_rate": 3.1873903087183734e-05, "loss": 0.0165, "step": 31860 }, { "epoch": 11.7040029379361, "grad_norm": 0.344813734292984, "learning_rate": 3.1843087195378176e-05, "loss": 0.0158, "step": 31870 }, { "epoch": 11.707675358060962, "grad_norm": 0.27750879526138306, "learning_rate": 3.181227924602312e-05, "loss": 0.015, "step": 31880 }, { "epoch": 11.711347778185825, "grad_norm": 0.32841169834136963, "learning_rate": 3.178147925259506e-05, "loss": 0.0178, "step": 31890 }, { "epoch": 11.715020198310686, "grad_norm": 0.25167933106422424, "learning_rate": 3.175068722856693e-05, "loss": 0.0181, "step": 31900 }, { "epoch": 11.718692618435549, "grad_norm": 0.2819907069206238, "learning_rate": 3.17199031874082e-05, "loss": 0.0158, "step": 31910 }, { "epoch": 11.722365038560412, "grad_norm": 0.25643834471702576, "learning_rate": 3.168912714258487e-05, "loss": 0.0156, "step": 31920 }, { "epoch": 11.726037458685273, "grad_norm": 0.2665044665336609, "learning_rate": 3.165835910755944e-05, "loss": 0.0171, "step": 31930 }, { "epoch": 11.729709878810135, "grad_norm": 0.2391519546508789, "learning_rate": 3.1627599095790874e-05, "loss": 0.0179, "step": 31940 }, { "epoch": 11.733382298934998, "grad_norm": 0.2022399753332138, "learning_rate": 3.159684712073462e-05, "loss": 0.0162, "step": 31950 }, { "epoch": 11.737054719059861, "grad_norm": 0.2000601589679718, "learning_rate": 3.1566103195842695e-05, "loss": 0.0182, "step": 31960 }, { "epoch": 11.740727139184722, "grad_norm": 0.14808028936386108, "learning_rate": 3.153536733456349e-05, "loss": 0.015, "step": 31970 }, { "epoch": 11.744399559309585, "grad_norm": 0.45121148228645325, "learning_rate": 3.150463955034194e-05, "loss": 0.0174, "step": 31980 }, { "epoch": 11.748071979434448, "grad_norm": 0.26166489720344543, "learning_rate": 3.147391985661942e-05, "loss": 0.016, "step": 31990 }, { "epoch": 11.75174439955931, "grad_norm": 0.2845264971256256, "learning_rate": 3.1443208266833755e-05, "loss": 0.0186, "step": 32000 }, { "epoch": 11.755416819684172, "grad_norm": 0.2647551894187927, "learning_rate": 3.141250479441924e-05, "loss": 0.0155, "step": 32010 }, { "epoch": 11.759089239809034, "grad_norm": 0.42569899559020996, "learning_rate": 3.138180945280664e-05, "loss": 0.0173, "step": 32020 }, { "epoch": 11.762761659933897, "grad_norm": 0.34750041365623474, "learning_rate": 3.1351122255423116e-05, "loss": 0.018, "step": 32030 }, { "epoch": 11.766434080058758, "grad_norm": 0.27443209290504456, "learning_rate": 3.132044321569228e-05, "loss": 0.0165, "step": 32040 }, { "epoch": 11.770106500183621, "grad_norm": 0.2697645425796509, "learning_rate": 3.1289772347034226e-05, "loss": 0.0155, "step": 32050 }, { "epoch": 11.773778920308484, "grad_norm": 0.3106454908847809, "learning_rate": 3.1259109662865404e-05, "loss": 0.0178, "step": 32060 }, { "epoch": 11.777451340433345, "grad_norm": 0.24719050526618958, "learning_rate": 3.122845517659871e-05, "loss": 0.0156, "step": 32070 }, { "epoch": 11.781123760558208, "grad_norm": 0.17255783081054688, "learning_rate": 3.119780890164348e-05, "loss": 0.0171, "step": 32080 }, { "epoch": 11.78479618068307, "grad_norm": 0.1939501315355301, "learning_rate": 3.116717085140542e-05, "loss": 0.0159, "step": 32090 }, { "epoch": 11.788468600807933, "grad_norm": 0.2648431956768036, "learning_rate": 3.1136541039286657e-05, "loss": 0.0158, "step": 32100 }, { "epoch": 11.792141020932794, "grad_norm": 0.6245795488357544, "learning_rate": 3.110591947868571e-05, "loss": 0.0159, "step": 32110 }, { "epoch": 11.795813441057657, "grad_norm": 0.4422670304775238, "learning_rate": 3.107530618299751e-05, "loss": 0.0188, "step": 32120 }, { "epoch": 11.79948586118252, "grad_norm": 0.20239774882793427, "learning_rate": 3.104470116561333e-05, "loss": 0.0151, "step": 32130 }, { "epoch": 11.80315828130738, "grad_norm": 0.5380593538284302, "learning_rate": 3.1014104439920835e-05, "loss": 0.0156, "step": 32140 }, { "epoch": 11.806830701432244, "grad_norm": 0.46244651079177856, "learning_rate": 3.098351601930411e-05, "loss": 0.016, "step": 32150 }, { "epoch": 11.810503121557106, "grad_norm": 0.41733288764953613, "learning_rate": 3.0952935917143533e-05, "loss": 0.0166, "step": 32160 }, { "epoch": 11.81417554168197, "grad_norm": 0.4223516881465912, "learning_rate": 3.092236414681588e-05, "loss": 0.0184, "step": 32170 }, { "epoch": 11.81784796180683, "grad_norm": 0.29653552174568176, "learning_rate": 3.089180072169431e-05, "loss": 0.0168, "step": 32180 }, { "epoch": 11.821520381931693, "grad_norm": 0.3168022036552429, "learning_rate": 3.08612456551483e-05, "loss": 0.0154, "step": 32190 }, { "epoch": 11.825192802056556, "grad_norm": 0.211436465382576, "learning_rate": 3.083069896054363e-05, "loss": 0.0147, "step": 32200 }, { "epoch": 11.828865222181417, "grad_norm": 0.2635735869407654, "learning_rate": 3.080016065124253e-05, "loss": 0.0169, "step": 32210 }, { "epoch": 11.83253764230628, "grad_norm": 0.2937183976173401, "learning_rate": 3.076963074060344e-05, "loss": 0.0166, "step": 32220 }, { "epoch": 11.836210062431142, "grad_norm": 0.1962141990661621, "learning_rate": 3.07391092419812e-05, "loss": 0.0157, "step": 32230 }, { "epoch": 11.839882482556005, "grad_norm": 0.201870396733284, "learning_rate": 3.070859616872695e-05, "loss": 0.0161, "step": 32240 }, { "epoch": 11.843554902680866, "grad_norm": 0.297465056180954, "learning_rate": 3.067809153418816e-05, "loss": 0.0158, "step": 32250 }, { "epoch": 11.847227322805729, "grad_norm": 0.27343377470970154, "learning_rate": 3.0647595351708545e-05, "loss": 0.0145, "step": 32260 }, { "epoch": 11.850899742930592, "grad_norm": 0.19026316702365875, "learning_rate": 3.061710763462824e-05, "loss": 0.0166, "step": 32270 }, { "epoch": 11.854572163055453, "grad_norm": 0.2355101853609085, "learning_rate": 3.0586628396283564e-05, "loss": 0.0161, "step": 32280 }, { "epoch": 11.858244583180316, "grad_norm": 0.39042171835899353, "learning_rate": 3.0556157650007174e-05, "loss": 0.0162, "step": 32290 }, { "epoch": 11.861917003305178, "grad_norm": 0.1934501826763153, "learning_rate": 3.052569540912801e-05, "loss": 0.0158, "step": 32300 }, { "epoch": 11.865589423430041, "grad_norm": 0.3503546416759491, "learning_rate": 3.049524168697131e-05, "loss": 0.0171, "step": 32310 }, { "epoch": 11.869261843554902, "grad_norm": 0.23445938527584076, "learning_rate": 3.0464796496858538e-05, "loss": 0.0151, "step": 32320 }, { "epoch": 11.872934263679765, "grad_norm": 0.3237612843513489, "learning_rate": 3.0434359852107473e-05, "loss": 0.0161, "step": 32330 }, { "epoch": 11.876606683804628, "grad_norm": 0.32687458395957947, "learning_rate": 3.0403931766032135e-05, "loss": 0.0169, "step": 32340 }, { "epoch": 11.880279103929489, "grad_norm": 0.29768863320350647, "learning_rate": 3.0373512251942814e-05, "loss": 0.0159, "step": 32350 }, { "epoch": 11.883951524054352, "grad_norm": 0.23446573317050934, "learning_rate": 3.0343101323146007e-05, "loss": 0.0164, "step": 32360 }, { "epoch": 11.887623944179214, "grad_norm": 0.26786139607429504, "learning_rate": 3.0312698992944523e-05, "loss": 0.0162, "step": 32370 }, { "epoch": 11.891296364304077, "grad_norm": 0.41480788588523865, "learning_rate": 3.028230527463736e-05, "loss": 0.0165, "step": 32380 }, { "epoch": 11.894968784428938, "grad_norm": 0.2627407908439636, "learning_rate": 3.0251920181519746e-05, "loss": 0.0175, "step": 32390 }, { "epoch": 11.898641204553801, "grad_norm": 0.41425275802612305, "learning_rate": 3.0221543726883183e-05, "loss": 0.0175, "step": 32400 }, { "epoch": 11.902313624678664, "grad_norm": 0.28346824645996094, "learning_rate": 3.019117592401535e-05, "loss": 0.0164, "step": 32410 }, { "epoch": 11.905986044803525, "grad_norm": 0.2761649191379547, "learning_rate": 3.0160816786200137e-05, "loss": 0.0164, "step": 32420 }, { "epoch": 11.909658464928388, "grad_norm": 0.4557639956474304, "learning_rate": 3.01304663267177e-05, "loss": 0.0166, "step": 32430 }, { "epoch": 11.91333088505325, "grad_norm": 0.2568313479423523, "learning_rate": 3.0100124558844338e-05, "loss": 0.0172, "step": 32440 }, { "epoch": 11.917003305178113, "grad_norm": 0.3246898353099823, "learning_rate": 3.0069791495852562e-05, "loss": 0.0159, "step": 32450 }, { "epoch": 11.920675725302974, "grad_norm": 0.19839294254779816, "learning_rate": 3.0039467151011113e-05, "loss": 0.0157, "step": 32460 }, { "epoch": 11.924348145427837, "grad_norm": 0.4784209728240967, "learning_rate": 3.000915153758487e-05, "loss": 0.017, "step": 32470 }, { "epoch": 11.9280205655527, "grad_norm": 0.2501758933067322, "learning_rate": 2.9978844668834917e-05, "loss": 0.0144, "step": 32480 }, { "epoch": 11.93169298567756, "grad_norm": 0.23134668171405792, "learning_rate": 2.994854655801849e-05, "loss": 0.0184, "step": 32490 }, { "epoch": 11.935365405802424, "grad_norm": 0.43715110421180725, "learning_rate": 2.991825721838905e-05, "loss": 0.0166, "step": 32500 }, { "epoch": 11.939037825927286, "grad_norm": 0.2836112678050995, "learning_rate": 2.9887976663196166e-05, "loss": 0.0165, "step": 32510 }, { "epoch": 11.942710246052147, "grad_norm": 0.5069801211357117, "learning_rate": 2.985770490568556e-05, "loss": 0.0165, "step": 32520 }, { "epoch": 11.94638266617701, "grad_norm": 0.2152481973171234, "learning_rate": 2.9827441959099157e-05, "loss": 0.0164, "step": 32530 }, { "epoch": 11.950055086301873, "grad_norm": 0.33101287484169006, "learning_rate": 2.979718783667499e-05, "loss": 0.0169, "step": 32540 }, { "epoch": 11.953727506426736, "grad_norm": 0.6424102187156677, "learning_rate": 2.9766942551647215e-05, "loss": 0.0161, "step": 32550 }, { "epoch": 11.957399926551597, "grad_norm": 0.40683117508888245, "learning_rate": 2.9736706117246188e-05, "loss": 0.0173, "step": 32560 }, { "epoch": 11.96107234667646, "grad_norm": 0.5226064324378967, "learning_rate": 2.9706478546698324e-05, "loss": 0.0179, "step": 32570 }, { "epoch": 11.964744766801322, "grad_norm": 0.28999724984169006, "learning_rate": 2.9676259853226195e-05, "loss": 0.0196, "step": 32580 }, { "epoch": 11.968417186926185, "grad_norm": 0.4292665123939514, "learning_rate": 2.964605005004849e-05, "loss": 0.0171, "step": 32590 }, { "epoch": 11.972089607051046, "grad_norm": 0.38284945487976074, "learning_rate": 2.9615849150380016e-05, "loss": 0.0186, "step": 32600 }, { "epoch": 11.975762027175909, "grad_norm": 0.28423529863357544, "learning_rate": 2.9585657167431637e-05, "loss": 0.0153, "step": 32610 }, { "epoch": 11.979434447300772, "grad_norm": 0.2261793166399002, "learning_rate": 2.9555474114410388e-05, "loss": 0.0158, "step": 32620 }, { "epoch": 11.983106867425633, "grad_norm": 0.6571604609489441, "learning_rate": 2.9525300004519348e-05, "loss": 0.0147, "step": 32630 }, { "epoch": 11.986779287550496, "grad_norm": 0.38381046056747437, "learning_rate": 2.9495134850957696e-05, "loss": 0.0169, "step": 32640 }, { "epoch": 11.990451707675359, "grad_norm": 0.2974524199962616, "learning_rate": 2.9464978666920674e-05, "loss": 0.0167, "step": 32650 }, { "epoch": 11.99412412780022, "grad_norm": 0.9762736558914185, "learning_rate": 2.9434831465599662e-05, "loss": 0.0159, "step": 32660 }, { "epoch": 11.997796547925082, "grad_norm": 0.38168323040008545, "learning_rate": 2.9404693260182047e-05, "loss": 0.0164, "step": 32670 }, { "epoch": 12.001468968049945, "grad_norm": 0.25761398673057556, "learning_rate": 2.9374564063851295e-05, "loss": 0.0164, "step": 32680 }, { "epoch": 12.005141388174808, "grad_norm": 0.28610989451408386, "learning_rate": 2.9344443889786967e-05, "loss": 0.0166, "step": 32690 }, { "epoch": 12.008813808299669, "grad_norm": 0.2423606663942337, "learning_rate": 2.931433275116463e-05, "loss": 0.0156, "step": 32700 }, { "epoch": 12.012486228424532, "grad_norm": 0.24019883573055267, "learning_rate": 2.928423066115592e-05, "loss": 0.0171, "step": 32710 }, { "epoch": 12.016158648549395, "grad_norm": 0.25684818625450134, "learning_rate": 2.9254137632928537e-05, "loss": 0.0156, "step": 32720 }, { "epoch": 12.019831068674256, "grad_norm": 0.3629138171672821, "learning_rate": 2.9224053679646167e-05, "loss": 0.0147, "step": 32730 }, { "epoch": 12.023503488799118, "grad_norm": 0.4546966254711151, "learning_rate": 2.9193978814468547e-05, "loss": 0.0172, "step": 32740 }, { "epoch": 12.027175908923981, "grad_norm": 0.6860809326171875, "learning_rate": 2.9163913050551485e-05, "loss": 0.0146, "step": 32750 }, { "epoch": 12.030848329048844, "grad_norm": 0.30580854415893555, "learning_rate": 2.9133856401046735e-05, "loss": 0.0149, "step": 32760 }, { "epoch": 12.034520749173705, "grad_norm": 0.4638478457927704, "learning_rate": 2.910380887910209e-05, "loss": 0.0178, "step": 32770 }, { "epoch": 12.038193169298568, "grad_norm": 0.35784104466438293, "learning_rate": 2.9073770497861386e-05, "loss": 0.0169, "step": 32780 }, { "epoch": 12.04186558942343, "grad_norm": 0.5175175666809082, "learning_rate": 2.904374127046442e-05, "loss": 0.0154, "step": 32790 }, { "epoch": 12.045538009548292, "grad_norm": 0.35212454199790955, "learning_rate": 2.9013721210046973e-05, "loss": 0.0167, "step": 32800 }, { "epoch": 12.049210429673154, "grad_norm": 0.323920339345932, "learning_rate": 2.8983710329740882e-05, "loss": 0.0161, "step": 32810 }, { "epoch": 12.052882849798017, "grad_norm": 0.35317906737327576, "learning_rate": 2.8953708642673898e-05, "loss": 0.0166, "step": 32820 }, { "epoch": 12.05655526992288, "grad_norm": 0.20462118089199066, "learning_rate": 2.8923716161969782e-05, "loss": 0.0152, "step": 32830 }, { "epoch": 12.060227690047741, "grad_norm": 0.2833590805530548, "learning_rate": 2.8893732900748248e-05, "loss": 0.0161, "step": 32840 }, { "epoch": 12.063900110172604, "grad_norm": 0.7664385437965393, "learning_rate": 2.8863758872125036e-05, "loss": 0.0178, "step": 32850 }, { "epoch": 12.067572530297467, "grad_norm": 0.443113774061203, "learning_rate": 2.883379408921175e-05, "loss": 0.0157, "step": 32860 }, { "epoch": 12.071244950422328, "grad_norm": 0.5517141819000244, "learning_rate": 2.880383856511606e-05, "loss": 0.0162, "step": 32870 }, { "epoch": 12.07491737054719, "grad_norm": 0.2957908511161804, "learning_rate": 2.8773892312941476e-05, "loss": 0.0154, "step": 32880 }, { "epoch": 12.078589790672053, "grad_norm": 0.21582695841789246, "learning_rate": 2.8743955345787555e-05, "loss": 0.0161, "step": 32890 }, { "epoch": 12.082262210796916, "grad_norm": 0.27707818150520325, "learning_rate": 2.87140276767497e-05, "loss": 0.0159, "step": 32900 }, { "epoch": 12.085934630921777, "grad_norm": 0.2075410932302475, "learning_rate": 2.868410931891934e-05, "loss": 0.0158, "step": 32910 }, { "epoch": 12.08960705104664, "grad_norm": 0.24083204567432404, "learning_rate": 2.8654200285383748e-05, "loss": 0.0147, "step": 32920 }, { "epoch": 12.093279471171503, "grad_norm": 0.32776957750320435, "learning_rate": 2.8624300589226126e-05, "loss": 0.0157, "step": 32930 }, { "epoch": 12.096951891296364, "grad_norm": 0.2532942593097687, "learning_rate": 2.859441024352568e-05, "loss": 0.0161, "step": 32940 }, { "epoch": 12.100624311421226, "grad_norm": 0.41727474331855774, "learning_rate": 2.8564529261357427e-05, "loss": 0.0168, "step": 32950 }, { "epoch": 12.10429673154609, "grad_norm": 0.28578317165374756, "learning_rate": 2.8534657655792307e-05, "loss": 0.0149, "step": 32960 }, { "epoch": 12.107969151670952, "grad_norm": 0.38002926111221313, "learning_rate": 2.8504795439897208e-05, "loss": 0.0164, "step": 32970 }, { "epoch": 12.111641571795813, "grad_norm": 0.2852552533149719, "learning_rate": 2.847494262673487e-05, "loss": 0.0159, "step": 32980 }, { "epoch": 12.115313991920676, "grad_norm": 0.3167131245136261, "learning_rate": 2.8445099229363925e-05, "loss": 0.014, "step": 32990 }, { "epoch": 12.118986412045539, "grad_norm": 0.7688689231872559, "learning_rate": 2.841526526083886e-05, "loss": 0.0153, "step": 33000 }, { "epoch": 12.1226588321704, "grad_norm": 0.6423482298851013, "learning_rate": 2.838544073421011e-05, "loss": 0.0167, "step": 33010 }, { "epoch": 12.126331252295262, "grad_norm": 0.24916529655456543, "learning_rate": 2.8355625662523928e-05, "loss": 0.0149, "step": 33020 }, { "epoch": 12.130003672420125, "grad_norm": 0.4721095860004425, "learning_rate": 2.8325820058822407e-05, "loss": 0.0157, "step": 33030 }, { "epoch": 12.133676092544988, "grad_norm": 0.2689950466156006, "learning_rate": 2.829602393614357e-05, "loss": 0.0155, "step": 33040 }, { "epoch": 12.137348512669849, "grad_norm": 0.3175390362739563, "learning_rate": 2.8266237307521236e-05, "loss": 0.0155, "step": 33050 }, { "epoch": 12.141020932794712, "grad_norm": 0.249390110373497, "learning_rate": 2.8236460185985074e-05, "loss": 0.0163, "step": 33060 }, { "epoch": 12.144693352919575, "grad_norm": 0.1677497774362564, "learning_rate": 2.8206692584560646e-05, "loss": 0.0154, "step": 33070 }, { "epoch": 12.148365773044436, "grad_norm": 0.31918835639953613, "learning_rate": 2.8176934516269283e-05, "loss": 0.0148, "step": 33080 }, { "epoch": 12.152038193169298, "grad_norm": 0.21614351868629456, "learning_rate": 2.8147185994128167e-05, "loss": 0.0145, "step": 33090 }, { "epoch": 12.155710613294161, "grad_norm": 0.33083680272102356, "learning_rate": 2.8117447031150346e-05, "loss": 0.0162, "step": 33100 }, { "epoch": 12.159383033419022, "grad_norm": 0.3012891709804535, "learning_rate": 2.8087717640344618e-05, "loss": 0.0166, "step": 33110 }, { "epoch": 12.163055453543885, "grad_norm": 0.37713637948036194, "learning_rate": 2.805799783471566e-05, "loss": 0.0146, "step": 33120 }, { "epoch": 12.166727873668748, "grad_norm": 0.369090735912323, "learning_rate": 2.8028287627263895e-05, "loss": 0.015, "step": 33130 }, { "epoch": 12.17040029379361, "grad_norm": 0.3599450886249542, "learning_rate": 2.7998587030985608e-05, "loss": 0.0147, "step": 33140 }, { "epoch": 12.174072713918472, "grad_norm": 0.4571481943130493, "learning_rate": 2.7968896058872806e-05, "loss": 0.0179, "step": 33150 }, { "epoch": 12.177745134043334, "grad_norm": 0.3430076837539673, "learning_rate": 2.7939214723913388e-05, "loss": 0.0151, "step": 33160 }, { "epoch": 12.181417554168197, "grad_norm": 0.49062302708625793, "learning_rate": 2.7909543039090935e-05, "loss": 0.0171, "step": 33170 }, { "epoch": 12.185089974293058, "grad_norm": 0.31482452154159546, "learning_rate": 2.787988101738487e-05, "loss": 0.0144, "step": 33180 }, { "epoch": 12.188762394417921, "grad_norm": 0.258498877286911, "learning_rate": 2.7850228671770334e-05, "loss": 0.0177, "step": 33190 }, { "epoch": 12.192434814542784, "grad_norm": 0.3286862075328827, "learning_rate": 2.782058601521832e-05, "loss": 0.0174, "step": 33200 }, { "epoch": 12.196107234667647, "grad_norm": 0.2738645374774933, "learning_rate": 2.77909530606955e-05, "loss": 0.0165, "step": 33210 }, { "epoch": 12.199779654792508, "grad_norm": 0.5848024487495422, "learning_rate": 2.7761329821164335e-05, "loss": 0.0158, "step": 33220 }, { "epoch": 12.20345207491737, "grad_norm": 0.24946407973766327, "learning_rate": 2.7731716309583062e-05, "loss": 0.0143, "step": 33230 }, { "epoch": 12.207124495042233, "grad_norm": 0.19923532009124756, "learning_rate": 2.7702112538905624e-05, "loss": 0.0153, "step": 33240 }, { "epoch": 12.210796915167094, "grad_norm": 0.27707722783088684, "learning_rate": 2.7672518522081682e-05, "loss": 0.0149, "step": 33250 }, { "epoch": 12.214469335291957, "grad_norm": 0.19650298357009888, "learning_rate": 2.7642934272056726e-05, "loss": 0.016, "step": 33260 }, { "epoch": 12.21814175541682, "grad_norm": 0.23646868765354156, "learning_rate": 2.7613359801771877e-05, "loss": 0.0186, "step": 33270 }, { "epoch": 12.221814175541683, "grad_norm": 0.253185510635376, "learning_rate": 2.7583795124163996e-05, "loss": 0.0166, "step": 33280 }, { "epoch": 12.225486595666544, "grad_norm": 0.26900896430015564, "learning_rate": 2.755424025216572e-05, "loss": 0.0148, "step": 33290 }, { "epoch": 12.229159015791407, "grad_norm": 0.3235369026660919, "learning_rate": 2.7524695198705334e-05, "loss": 0.0155, "step": 33300 }, { "epoch": 12.23283143591627, "grad_norm": 0.1663653403520584, "learning_rate": 2.749515997670683e-05, "loss": 0.0156, "step": 33310 }, { "epoch": 12.23650385604113, "grad_norm": 0.4641658067703247, "learning_rate": 2.746563459908995e-05, "loss": 0.0168, "step": 33320 }, { "epoch": 12.240176276165993, "grad_norm": 0.29573947191238403, "learning_rate": 2.7436119078770085e-05, "loss": 0.0172, "step": 33330 }, { "epoch": 12.243848696290856, "grad_norm": 0.2958330512046814, "learning_rate": 2.74066134286583e-05, "loss": 0.0157, "step": 33340 }, { "epoch": 12.247521116415719, "grad_norm": 0.24507924914360046, "learning_rate": 2.7377117661661412e-05, "loss": 0.0147, "step": 33350 }, { "epoch": 12.25119353654058, "grad_norm": 0.312961608171463, "learning_rate": 2.734763179068183e-05, "loss": 0.0166, "step": 33360 }, { "epoch": 12.254865956665443, "grad_norm": 0.24200661480426788, "learning_rate": 2.7318155828617718e-05, "loss": 0.0193, "step": 33370 }, { "epoch": 12.258538376790305, "grad_norm": 0.42118608951568604, "learning_rate": 2.7288689788362832e-05, "loss": 0.0168, "step": 33380 }, { "epoch": 12.262210796915166, "grad_norm": 0.17750610411167145, "learning_rate": 2.725923368280664e-05, "loss": 0.0169, "step": 33390 }, { "epoch": 12.26588321704003, "grad_norm": 0.4147227108478546, "learning_rate": 2.7229787524834237e-05, "loss": 0.0162, "step": 33400 }, { "epoch": 12.269555637164892, "grad_norm": 0.2932412028312683, "learning_rate": 2.7200351327326345e-05, "loss": 0.0145, "step": 33410 }, { "epoch": 12.273228057289755, "grad_norm": 0.2920820713043213, "learning_rate": 2.7170925103159407e-05, "loss": 0.0162, "step": 33420 }, { "epoch": 12.276900477414616, "grad_norm": 0.3003472685813904, "learning_rate": 2.714150886520542e-05, "loss": 0.0155, "step": 33430 }, { "epoch": 12.280572897539479, "grad_norm": 0.3032567799091339, "learning_rate": 2.7112102626332038e-05, "loss": 0.0157, "step": 33440 }, { "epoch": 12.284245317664341, "grad_norm": 0.3440392315387726, "learning_rate": 2.7082706399402574e-05, "loss": 0.0157, "step": 33450 }, { "epoch": 12.287917737789202, "grad_norm": 0.3725508451461792, "learning_rate": 2.7053320197275934e-05, "loss": 0.0163, "step": 33460 }, { "epoch": 12.291590157914065, "grad_norm": 0.2636512219905853, "learning_rate": 2.7023944032806614e-05, "loss": 0.0153, "step": 33470 }, { "epoch": 12.295262578038928, "grad_norm": 0.39059945940971375, "learning_rate": 2.699457791884478e-05, "loss": 0.015, "step": 33480 }, { "epoch": 12.29893499816379, "grad_norm": 0.28625422716140747, "learning_rate": 2.6965221868236155e-05, "loss": 0.0165, "step": 33490 }, { "epoch": 12.302607418288652, "grad_norm": 0.3085009455680847, "learning_rate": 2.6935875893822077e-05, "loss": 0.0174, "step": 33500 }, { "epoch": 12.306279838413515, "grad_norm": 0.22158990800380707, "learning_rate": 2.690654000843944e-05, "loss": 0.0167, "step": 33510 }, { "epoch": 12.309952258538377, "grad_norm": 0.2567075192928314, "learning_rate": 2.687721422492081e-05, "loss": 0.0157, "step": 33520 }, { "epoch": 12.313624678663238, "grad_norm": 0.23316046595573425, "learning_rate": 2.6847898556094252e-05, "loss": 0.0153, "step": 33530 }, { "epoch": 12.317297098788101, "grad_norm": 0.26986247301101685, "learning_rate": 2.6818593014783417e-05, "loss": 0.0168, "step": 33540 }, { "epoch": 12.320969518912964, "grad_norm": 0.28260132670402527, "learning_rate": 2.6789297613807585e-05, "loss": 0.0155, "step": 33550 }, { "epoch": 12.324641939037825, "grad_norm": 0.4078384041786194, "learning_rate": 2.676001236598154e-05, "loss": 0.0159, "step": 33560 }, { "epoch": 12.328314359162688, "grad_norm": 0.25777682662010193, "learning_rate": 2.6730737284115626e-05, "loss": 0.0135, "step": 33570 }, { "epoch": 12.33198677928755, "grad_norm": 0.6066299080848694, "learning_rate": 2.670147238101579e-05, "loss": 0.0167, "step": 33580 }, { "epoch": 12.335659199412413, "grad_norm": 0.5224542617797852, "learning_rate": 2.6672217669483473e-05, "loss": 0.0163, "step": 33590 }, { "epoch": 12.339331619537274, "grad_norm": 0.28351712226867676, "learning_rate": 2.6642973162315706e-05, "loss": 0.0163, "step": 33600 }, { "epoch": 12.343004039662137, "grad_norm": 0.2257867157459259, "learning_rate": 2.6613738872304988e-05, "loss": 0.0157, "step": 33610 }, { "epoch": 12.346676459787, "grad_norm": 0.4025912582874298, "learning_rate": 2.6584514812239437e-05, "loss": 0.0187, "step": 33620 }, { "epoch": 12.350348879911863, "grad_norm": 0.2212330847978592, "learning_rate": 2.655530099490261e-05, "loss": 0.016, "step": 33630 }, { "epoch": 12.354021300036724, "grad_norm": 0.5971361994743347, "learning_rate": 2.6526097433073653e-05, "loss": 0.0168, "step": 33640 }, { "epoch": 12.357693720161587, "grad_norm": 0.4255036413669586, "learning_rate": 2.649690413952719e-05, "loss": 0.017, "step": 33650 }, { "epoch": 12.36136614028645, "grad_norm": 0.22518384456634521, "learning_rate": 2.646772112703334e-05, "loss": 0.0158, "step": 33660 }, { "epoch": 12.36503856041131, "grad_norm": 0.4037148654460907, "learning_rate": 2.6438548408357768e-05, "loss": 0.0155, "step": 33670 }, { "epoch": 12.368710980536173, "grad_norm": 0.3306407928466797, "learning_rate": 2.6409385996261603e-05, "loss": 0.0144, "step": 33680 }, { "epoch": 12.372383400661036, "grad_norm": 0.17200376093387604, "learning_rate": 2.638023390350148e-05, "loss": 0.0153, "step": 33690 }, { "epoch": 12.376055820785897, "grad_norm": 0.19159568846225739, "learning_rate": 2.635109214282948e-05, "loss": 0.0154, "step": 33700 }, { "epoch": 12.37972824091076, "grad_norm": 0.3173816502094269, "learning_rate": 2.6321960726993254e-05, "loss": 0.016, "step": 33710 }, { "epoch": 12.383400661035623, "grad_norm": 0.39592814445495605, "learning_rate": 2.6292839668735846e-05, "loss": 0.0147, "step": 33720 }, { "epoch": 12.387073081160485, "grad_norm": 0.28968778252601624, "learning_rate": 2.6263728980795778e-05, "loss": 0.0136, "step": 33730 }, { "epoch": 12.390745501285346, "grad_norm": 0.34755170345306396, "learning_rate": 2.6234628675907092e-05, "loss": 0.0188, "step": 33740 }, { "epoch": 12.39441792141021, "grad_norm": 0.29712530970573425, "learning_rate": 2.620553876679923e-05, "loss": 0.0156, "step": 33750 }, { "epoch": 12.398090341535072, "grad_norm": 2.1680197715759277, "learning_rate": 2.617645926619709e-05, "loss": 0.0151, "step": 33760 }, { "epoch": 12.401762761659933, "grad_norm": 0.31979629397392273, "learning_rate": 2.6147390186821064e-05, "loss": 0.0158, "step": 33770 }, { "epoch": 12.405435181784796, "grad_norm": 0.31512120366096497, "learning_rate": 2.6118331541386955e-05, "loss": 0.0153, "step": 33780 }, { "epoch": 12.409107601909659, "grad_norm": 0.4413827657699585, "learning_rate": 2.6089283342605953e-05, "loss": 0.0154, "step": 33790 }, { "epoch": 12.412780022034521, "grad_norm": 0.4031507074832916, "learning_rate": 2.6060245603184797e-05, "loss": 0.0173, "step": 33800 }, { "epoch": 12.416452442159382, "grad_norm": 0.21301841735839844, "learning_rate": 2.6031218335825536e-05, "loss": 0.0153, "step": 33810 }, { "epoch": 12.420124862284245, "grad_norm": 0.4746973514556885, "learning_rate": 2.600220155322568e-05, "loss": 0.015, "step": 33820 }, { "epoch": 12.423797282409108, "grad_norm": 0.4583825469017029, "learning_rate": 2.5973195268078186e-05, "loss": 0.015, "step": 33830 }, { "epoch": 12.427469702533969, "grad_norm": 0.15710577368736267, "learning_rate": 2.5944199493071352e-05, "loss": 0.0164, "step": 33840 }, { "epoch": 12.431142122658832, "grad_norm": 0.2697913944721222, "learning_rate": 2.591521424088895e-05, "loss": 0.0151, "step": 33850 }, { "epoch": 12.434814542783695, "grad_norm": 0.452675461769104, "learning_rate": 2.5886239524210086e-05, "loss": 0.0153, "step": 33860 }, { "epoch": 12.438486962908557, "grad_norm": 0.22732141613960266, "learning_rate": 2.5857275355709317e-05, "loss": 0.0136, "step": 33870 }, { "epoch": 12.442159383033419, "grad_norm": 0.4630001187324524, "learning_rate": 2.5828321748056533e-05, "loss": 0.014, "step": 33880 }, { "epoch": 12.445831803158281, "grad_norm": 0.23199018836021423, "learning_rate": 2.5799378713917012e-05, "loss": 0.015, "step": 33890 }, { "epoch": 12.449504223283144, "grad_norm": 0.24975349009037018, "learning_rate": 2.5770446265951455e-05, "loss": 0.0155, "step": 33900 }, { "epoch": 12.453176643408005, "grad_norm": 0.43197494745254517, "learning_rate": 2.5741524416815877e-05, "loss": 0.0147, "step": 33910 }, { "epoch": 12.456849063532868, "grad_norm": 0.21276280283927917, "learning_rate": 2.571261317916166e-05, "loss": 0.0134, "step": 33920 }, { "epoch": 12.46052148365773, "grad_norm": 0.2733558118343353, "learning_rate": 2.5683712565635597e-05, "loss": 0.0145, "step": 33930 }, { "epoch": 12.464193903782594, "grad_norm": 0.3530008792877197, "learning_rate": 2.565482258887979e-05, "loss": 0.0141, "step": 33940 }, { "epoch": 12.467866323907455, "grad_norm": 0.31567949056625366, "learning_rate": 2.5625943261531664e-05, "loss": 0.0166, "step": 33950 }, { "epoch": 12.471538744032317, "grad_norm": 0.22105753421783447, "learning_rate": 2.5597074596224067e-05, "loss": 0.0147, "step": 33960 }, { "epoch": 12.47521116415718, "grad_norm": 0.20938605070114136, "learning_rate": 2.5568216605585105e-05, "loss": 0.019, "step": 33970 }, { "epoch": 12.478883584282041, "grad_norm": 0.36708128452301025, "learning_rate": 2.5539369302238235e-05, "loss": 0.0144, "step": 33980 }, { "epoch": 12.482556004406904, "grad_norm": 0.2748314440250397, "learning_rate": 2.5510532698802276e-05, "loss": 0.014, "step": 33990 }, { "epoch": 12.486228424531767, "grad_norm": 0.25404495000839233, "learning_rate": 2.5481706807891324e-05, "loss": 0.0177, "step": 34000 }, { "epoch": 12.48990084465663, "grad_norm": 0.4007192850112915, "learning_rate": 2.5452891642114783e-05, "loss": 0.016, "step": 34010 }, { "epoch": 12.49357326478149, "grad_norm": 0.2972203195095062, "learning_rate": 2.542408721407742e-05, "loss": 0.016, "step": 34020 }, { "epoch": 12.497245684906353, "grad_norm": 0.40836766362190247, "learning_rate": 2.539529353637926e-05, "loss": 0.0152, "step": 34030 }, { "epoch": 12.500918105031216, "grad_norm": 0.403413861989975, "learning_rate": 2.5366510621615626e-05, "loss": 0.015, "step": 34040 }, { "epoch": 12.504590525156077, "grad_norm": 0.8778652548789978, "learning_rate": 2.5337738482377132e-05, "loss": 0.0149, "step": 34050 }, { "epoch": 12.50826294528094, "grad_norm": 0.22821387648582458, "learning_rate": 2.5308977131249722e-05, "loss": 0.0148, "step": 34060 }, { "epoch": 12.511935365405803, "grad_norm": 0.24913495779037476, "learning_rate": 2.5280226580814554e-05, "loss": 0.0156, "step": 34070 }, { "epoch": 12.515607785530666, "grad_norm": 0.21460683643817902, "learning_rate": 2.525148684364813e-05, "loss": 0.0142, "step": 34080 }, { "epoch": 12.519280205655527, "grad_norm": 0.45453062653541565, "learning_rate": 2.522275793232215e-05, "loss": 0.0156, "step": 34090 }, { "epoch": 12.52295262578039, "grad_norm": 0.33554327487945557, "learning_rate": 2.519403985940365e-05, "loss": 0.0151, "step": 34100 }, { "epoch": 12.526625045905252, "grad_norm": 0.2237386256456375, "learning_rate": 2.5165332637454863e-05, "loss": 0.0161, "step": 34110 }, { "epoch": 12.530297466030113, "grad_norm": 0.28172844648361206, "learning_rate": 2.513663627903334e-05, "loss": 0.0165, "step": 34120 }, { "epoch": 12.533969886154976, "grad_norm": 0.29675373435020447, "learning_rate": 2.5107950796691815e-05, "loss": 0.0144, "step": 34130 }, { "epoch": 12.537642306279839, "grad_norm": 0.21867059171199799, "learning_rate": 2.5079276202978276e-05, "loss": 0.0158, "step": 34140 }, { "epoch": 12.5413147264047, "grad_norm": 0.2168935388326645, "learning_rate": 2.505061251043601e-05, "loss": 0.0157, "step": 34150 }, { "epoch": 12.544987146529563, "grad_norm": 0.3662051558494568, "learning_rate": 2.502195973160346e-05, "loss": 0.0154, "step": 34160 }, { "epoch": 12.548659566654425, "grad_norm": 0.26960718631744385, "learning_rate": 2.4993317879014316e-05, "loss": 0.0154, "step": 34170 }, { "epoch": 12.552331986779288, "grad_norm": 0.17184503376483917, "learning_rate": 2.4964686965197526e-05, "loss": 0.016, "step": 34180 }, { "epoch": 12.55600440690415, "grad_norm": 0.21315425634384155, "learning_rate": 2.4936067002677217e-05, "loss": 0.0145, "step": 34190 }, { "epoch": 12.559676827029012, "grad_norm": 0.357282429933548, "learning_rate": 2.4907458003972724e-05, "loss": 0.0152, "step": 34200 }, { "epoch": 12.563349247153875, "grad_norm": 0.24709755182266235, "learning_rate": 2.4878859981598572e-05, "loss": 0.0171, "step": 34210 }, { "epoch": 12.567021667278738, "grad_norm": 0.30786287784576416, "learning_rate": 2.4850272948064555e-05, "loss": 0.0159, "step": 34220 }, { "epoch": 12.570694087403599, "grad_norm": 0.23071883618831635, "learning_rate": 2.4821696915875596e-05, "loss": 0.0146, "step": 34230 }, { "epoch": 12.574366507528461, "grad_norm": 0.21398021280765533, "learning_rate": 2.4793131897531798e-05, "loss": 0.0143, "step": 34240 }, { "epoch": 12.578038927653324, "grad_norm": 0.46571266651153564, "learning_rate": 2.47645779055285e-05, "loss": 0.0151, "step": 34250 }, { "epoch": 12.581711347778185, "grad_norm": 0.30343303084373474, "learning_rate": 2.4736034952356186e-05, "loss": 0.0153, "step": 34260 }, { "epoch": 12.585383767903048, "grad_norm": 0.3826904892921448, "learning_rate": 2.470750305050048e-05, "loss": 0.0163, "step": 34270 }, { "epoch": 12.58905618802791, "grad_norm": 0.31948938965797424, "learning_rate": 2.4678982212442253e-05, "loss": 0.0147, "step": 34280 }, { "epoch": 12.592728608152772, "grad_norm": 0.28474411368370056, "learning_rate": 2.4650472450657463e-05, "loss": 0.0156, "step": 34290 }, { "epoch": 12.596401028277635, "grad_norm": 0.40385740995407104, "learning_rate": 2.4621973777617237e-05, "loss": 0.0147, "step": 34300 }, { "epoch": 12.600073448402497, "grad_norm": 0.26678112149238586, "learning_rate": 2.459348620578789e-05, "loss": 0.0147, "step": 34310 }, { "epoch": 12.60374586852736, "grad_norm": 0.34175533056259155, "learning_rate": 2.456500974763083e-05, "loss": 0.0157, "step": 34320 }, { "epoch": 12.607418288652221, "grad_norm": 0.350985050201416, "learning_rate": 2.4536544415602653e-05, "loss": 0.0154, "step": 34330 }, { "epoch": 12.611090708777084, "grad_norm": 0.19146466255187988, "learning_rate": 2.4508090222155027e-05, "loss": 0.015, "step": 34340 }, { "epoch": 12.614763128901947, "grad_norm": 0.2711258828639984, "learning_rate": 2.4479647179734822e-05, "loss": 0.0159, "step": 34350 }, { "epoch": 12.618435549026808, "grad_norm": 0.3528446555137634, "learning_rate": 2.4451215300783953e-05, "loss": 0.0159, "step": 34360 }, { "epoch": 12.62210796915167, "grad_norm": 0.46365389227867126, "learning_rate": 2.442279459773953e-05, "loss": 0.0147, "step": 34370 }, { "epoch": 12.625780389276533, "grad_norm": 0.33463016152381897, "learning_rate": 2.4394385083033712e-05, "loss": 0.0167, "step": 34380 }, { "epoch": 12.629452809401396, "grad_norm": 0.43895596265792847, "learning_rate": 2.4365986769093792e-05, "loss": 0.0148, "step": 34390 }, { "epoch": 12.633125229526257, "grad_norm": 0.25063642859458923, "learning_rate": 2.4337599668342127e-05, "loss": 0.0153, "step": 34400 }, { "epoch": 12.63679764965112, "grad_norm": 0.34906288981437683, "learning_rate": 2.4309223793196245e-05, "loss": 0.0152, "step": 34410 }, { "epoch": 12.640470069775983, "grad_norm": 0.28395092487335205, "learning_rate": 2.4280859156068698e-05, "loss": 0.0161, "step": 34420 }, { "epoch": 12.644142489900844, "grad_norm": 0.3897046446800232, "learning_rate": 2.4252505769367123e-05, "loss": 0.0171, "step": 34430 }, { "epoch": 12.647814910025707, "grad_norm": 0.2336229681968689, "learning_rate": 2.422416364549429e-05, "loss": 0.0159, "step": 34440 }, { "epoch": 12.65148733015057, "grad_norm": 0.24921707808971405, "learning_rate": 2.4195832796847994e-05, "loss": 0.0156, "step": 34450 }, { "epoch": 12.655159750275432, "grad_norm": 0.41680708527565, "learning_rate": 2.416751323582109e-05, "loss": 0.0165, "step": 34460 }, { "epoch": 12.658832170400293, "grad_norm": 0.4249974191188812, "learning_rate": 2.4139204974801544e-05, "loss": 0.016, "step": 34470 }, { "epoch": 12.662504590525156, "grad_norm": 0.29693031311035156, "learning_rate": 2.4110908026172347e-05, "loss": 0.0153, "step": 34480 }, { "epoch": 12.666177010650019, "grad_norm": 0.31344377994537354, "learning_rate": 2.4082622402311518e-05, "loss": 0.0154, "step": 34490 }, { "epoch": 12.66984943077488, "grad_norm": 0.42241552472114563, "learning_rate": 2.4054348115592183e-05, "loss": 0.0146, "step": 34500 }, { "epoch": 12.673521850899743, "grad_norm": 0.2133871167898178, "learning_rate": 2.4026085178382475e-05, "loss": 0.0158, "step": 34510 }, { "epoch": 12.677194271024605, "grad_norm": 0.31146812438964844, "learning_rate": 2.3997833603045523e-05, "loss": 0.0153, "step": 34520 }, { "epoch": 12.680866691149468, "grad_norm": 0.40909436345100403, "learning_rate": 2.396959340193958e-05, "loss": 0.0152, "step": 34530 }, { "epoch": 12.68453911127433, "grad_norm": 0.19893622398376465, "learning_rate": 2.3941364587417836e-05, "loss": 0.0138, "step": 34540 }, { "epoch": 12.688211531399192, "grad_norm": 0.4407423734664917, "learning_rate": 2.3913147171828526e-05, "loss": 0.0154, "step": 34550 }, { "epoch": 12.691883951524055, "grad_norm": 0.38656216859817505, "learning_rate": 2.388494116751494e-05, "loss": 0.0155, "step": 34560 }, { "epoch": 12.695556371648916, "grad_norm": 0.28976407647132874, "learning_rate": 2.3856746586815302e-05, "loss": 0.0131, "step": 34570 }, { "epoch": 12.699228791773779, "grad_norm": 0.6876988410949707, "learning_rate": 2.3828563442062924e-05, "loss": 0.0148, "step": 34580 }, { "epoch": 12.702901211898642, "grad_norm": 0.2740987241268158, "learning_rate": 2.3800391745586026e-05, "loss": 0.0153, "step": 34590 }, { "epoch": 12.706573632023504, "grad_norm": 0.2273719757795334, "learning_rate": 2.377223150970791e-05, "loss": 0.0154, "step": 34600 }, { "epoch": 12.710246052148365, "grad_norm": 0.25800663232803345, "learning_rate": 2.3744082746746793e-05, "loss": 0.0159, "step": 34610 }, { "epoch": 12.713918472273228, "grad_norm": 0.4347565472126007, "learning_rate": 2.3715945469015893e-05, "loss": 0.0177, "step": 34620 }, { "epoch": 12.717590892398091, "grad_norm": 0.26956552267074585, "learning_rate": 2.368781968882343e-05, "loss": 0.0158, "step": 34630 }, { "epoch": 12.721263312522952, "grad_norm": 0.29907768964767456, "learning_rate": 2.365970541847258e-05, "loss": 0.016, "step": 34640 }, { "epoch": 12.724935732647815, "grad_norm": 0.6437905430793762, "learning_rate": 2.3631602670261445e-05, "loss": 0.0159, "step": 34650 }, { "epoch": 12.728608152772678, "grad_norm": 0.5176235437393188, "learning_rate": 2.3603511456483164e-05, "loss": 0.0158, "step": 34660 }, { "epoch": 12.73228057289754, "grad_norm": 0.35360968112945557, "learning_rate": 2.357543178942577e-05, "loss": 0.0149, "step": 34670 }, { "epoch": 12.735952993022401, "grad_norm": 0.6499298214912415, "learning_rate": 2.3547363681372242e-05, "loss": 0.0152, "step": 34680 }, { "epoch": 12.739625413147264, "grad_norm": 0.3020074665546417, "learning_rate": 2.3519307144600565e-05, "loss": 0.015, "step": 34690 }, { "epoch": 12.743297833272127, "grad_norm": 0.3044690489768982, "learning_rate": 2.3491262191383607e-05, "loss": 0.0154, "step": 34700 }, { "epoch": 12.746970253396988, "grad_norm": 0.21966351568698883, "learning_rate": 2.3463228833989154e-05, "loss": 0.0135, "step": 34710 }, { "epoch": 12.75064267352185, "grad_norm": 0.297707200050354, "learning_rate": 2.343520708467999e-05, "loss": 0.0166, "step": 34720 }, { "epoch": 12.754315093646714, "grad_norm": 0.5178436040878296, "learning_rate": 2.3407196955713767e-05, "loss": 0.0158, "step": 34730 }, { "epoch": 12.757987513771575, "grad_norm": 0.3509904444217682, "learning_rate": 2.337919845934306e-05, "loss": 0.0155, "step": 34740 }, { "epoch": 12.761659933896437, "grad_norm": 0.3331945240497589, "learning_rate": 2.335121160781535e-05, "loss": 0.0153, "step": 34750 }, { "epoch": 12.7653323540213, "grad_norm": 0.38518428802490234, "learning_rate": 2.3323236413373063e-05, "loss": 0.0149, "step": 34760 }, { "epoch": 12.769004774146163, "grad_norm": 0.22318141162395477, "learning_rate": 2.3295272888253494e-05, "loss": 0.014, "step": 34770 }, { "epoch": 12.772677194271024, "grad_norm": 0.4164704978466034, "learning_rate": 2.3267321044688815e-05, "loss": 0.019, "step": 34780 }, { "epoch": 12.776349614395887, "grad_norm": 0.21842379868030548, "learning_rate": 2.323938089490615e-05, "loss": 0.0153, "step": 34790 }, { "epoch": 12.78002203452075, "grad_norm": 0.31045931577682495, "learning_rate": 2.321145245112743e-05, "loss": 0.019, "step": 34800 }, { "epoch": 12.783694454645612, "grad_norm": 0.21786439418792725, "learning_rate": 2.318353572556953e-05, "loss": 0.0143, "step": 34810 }, { "epoch": 12.787366874770473, "grad_norm": 0.3415124714374542, "learning_rate": 2.3155630730444183e-05, "loss": 0.0148, "step": 34820 }, { "epoch": 12.791039294895336, "grad_norm": 0.42987334728240967, "learning_rate": 2.3127737477957968e-05, "loss": 0.014, "step": 34830 }, { "epoch": 12.794711715020199, "grad_norm": 0.4478113055229187, "learning_rate": 2.3099855980312324e-05, "loss": 0.0157, "step": 34840 }, { "epoch": 12.79838413514506, "grad_norm": 0.3270617425441742, "learning_rate": 2.3071986249703602e-05, "loss": 0.0148, "step": 34850 }, { "epoch": 12.802056555269923, "grad_norm": 0.31638801097869873, "learning_rate": 2.3044128298322953e-05, "loss": 0.0149, "step": 34860 }, { "epoch": 12.805728975394786, "grad_norm": 0.17323608696460724, "learning_rate": 2.301628213835637e-05, "loss": 0.0132, "step": 34870 }, { "epoch": 12.809401395519647, "grad_norm": 0.3389813303947449, "learning_rate": 2.2988447781984744e-05, "loss": 0.0149, "step": 34880 }, { "epoch": 12.81307381564451, "grad_norm": 0.38380125164985657, "learning_rate": 2.2960625241383753e-05, "loss": 0.0144, "step": 34890 }, { "epoch": 12.816746235769372, "grad_norm": 0.27457162737846375, "learning_rate": 2.2932814528723922e-05, "loss": 0.0141, "step": 34900 }, { "epoch": 12.820418655894235, "grad_norm": 0.41553303599357605, "learning_rate": 2.2905015656170582e-05, "loss": 0.0151, "step": 34910 }, { "epoch": 12.824091076019096, "grad_norm": 0.37753960490226746, "learning_rate": 2.287722863588394e-05, "loss": 0.0165, "step": 34920 }, { "epoch": 12.827763496143959, "grad_norm": 0.19209443032741547, "learning_rate": 2.2849453480018962e-05, "loss": 0.0157, "step": 34930 }, { "epoch": 12.831435916268822, "grad_norm": 0.3273376226425171, "learning_rate": 2.282169020072543e-05, "loss": 0.0147, "step": 34940 }, { "epoch": 12.835108336393683, "grad_norm": 0.39646244049072266, "learning_rate": 2.2793938810147986e-05, "loss": 0.0145, "step": 34950 }, { "epoch": 12.838780756518545, "grad_norm": 0.34786662459373474, "learning_rate": 2.2766199320426007e-05, "loss": 0.0158, "step": 34960 }, { "epoch": 12.842453176643408, "grad_norm": 0.49341094493865967, "learning_rate": 2.2738471743693674e-05, "loss": 0.0141, "step": 34970 }, { "epoch": 12.846125596768271, "grad_norm": 0.268868088722229, "learning_rate": 2.2710756092080006e-05, "loss": 0.0161, "step": 34980 }, { "epoch": 12.849798016893132, "grad_norm": 0.5842325687408447, "learning_rate": 2.2683052377708747e-05, "loss": 0.0142, "step": 34990 }, { "epoch": 12.853470437017995, "grad_norm": 0.45567575097084045, "learning_rate": 2.265536061269843e-05, "loss": 0.0152, "step": 35000 }, { "epoch": 12.857142857142858, "grad_norm": 0.37380191683769226, "learning_rate": 2.262768080916241e-05, "loss": 0.0136, "step": 35010 }, { "epoch": 12.860815277267719, "grad_norm": 0.2046617865562439, "learning_rate": 2.260001297920874e-05, "loss": 0.0148, "step": 35020 }, { "epoch": 12.864487697392581, "grad_norm": 0.39988017082214355, "learning_rate": 2.2572357134940297e-05, "loss": 0.0163, "step": 35030 }, { "epoch": 12.868160117517444, "grad_norm": 0.2732398808002472, "learning_rate": 2.2544713288454655e-05, "loss": 0.015, "step": 35040 }, { "epoch": 12.871832537642307, "grad_norm": 0.44685760140419006, "learning_rate": 2.2517081451844213e-05, "loss": 0.0167, "step": 35050 }, { "epoch": 12.875504957767168, "grad_norm": 0.5927951335906982, "learning_rate": 2.248946163719605e-05, "loss": 0.0149, "step": 35060 }, { "epoch": 12.87917737789203, "grad_norm": 0.33471155166625977, "learning_rate": 2.2461853856591996e-05, "loss": 0.0166, "step": 35070 }, { "epoch": 12.882849798016894, "grad_norm": 0.41232576966285706, "learning_rate": 2.2434258122108664e-05, "loss": 0.0147, "step": 35080 }, { "epoch": 12.886522218141755, "grad_norm": 0.6075072884559631, "learning_rate": 2.2406674445817356e-05, "loss": 0.0134, "step": 35090 }, { "epoch": 12.890194638266617, "grad_norm": 0.4814360439777374, "learning_rate": 2.2379102839784083e-05, "loss": 0.0157, "step": 35100 }, { "epoch": 12.89386705839148, "grad_norm": 0.3889501690864563, "learning_rate": 2.2351543316069645e-05, "loss": 0.0148, "step": 35110 }, { "epoch": 12.897539478516343, "grad_norm": 0.2575279474258423, "learning_rate": 2.2323995886729493e-05, "loss": 0.0143, "step": 35120 }, { "epoch": 12.901211898641204, "grad_norm": 0.38412779569625854, "learning_rate": 2.2296460563813786e-05, "loss": 0.0167, "step": 35130 }, { "epoch": 12.904884318766067, "grad_norm": 0.3882034420967102, "learning_rate": 2.2268937359367457e-05, "loss": 0.0158, "step": 35140 }, { "epoch": 12.90855673889093, "grad_norm": 0.49269309639930725, "learning_rate": 2.224142628543006e-05, "loss": 0.0162, "step": 35150 }, { "epoch": 12.91222915901579, "grad_norm": 0.4041212201118469, "learning_rate": 2.2213927354035863e-05, "loss": 0.0151, "step": 35160 }, { "epoch": 12.915901579140654, "grad_norm": 0.3334115445613861, "learning_rate": 2.218644057721387e-05, "loss": 0.0166, "step": 35170 }, { "epoch": 12.919573999265516, "grad_norm": 0.35506221652030945, "learning_rate": 2.2158965966987703e-05, "loss": 0.0146, "step": 35180 }, { "epoch": 12.923246419390377, "grad_norm": 0.4296961724758148, "learning_rate": 2.2131503535375687e-05, "loss": 0.0169, "step": 35190 }, { "epoch": 12.92691883951524, "grad_norm": 0.5114132761955261, "learning_rate": 2.2104053294390846e-05, "loss": 0.016, "step": 35200 }, { "epoch": 12.930591259640103, "grad_norm": 0.3089686632156372, "learning_rate": 2.2076615256040834e-05, "loss": 0.0154, "step": 35210 }, { "epoch": 12.934263679764966, "grad_norm": 0.42510345578193665, "learning_rate": 2.2049189432327966e-05, "loss": 0.0152, "step": 35220 }, { "epoch": 12.937936099889827, "grad_norm": 0.3407425880432129, "learning_rate": 2.202177583524926e-05, "loss": 0.0137, "step": 35230 }, { "epoch": 12.94160852001469, "grad_norm": 0.40440723299980164, "learning_rate": 2.1994374476796343e-05, "loss": 0.0149, "step": 35240 }, { "epoch": 12.945280940139552, "grad_norm": 0.2209589034318924, "learning_rate": 2.1966985368955477e-05, "loss": 0.0168, "step": 35250 }, { "epoch": 12.948953360264415, "grad_norm": 0.369227796792984, "learning_rate": 2.193960852370762e-05, "loss": 0.015, "step": 35260 }, { "epoch": 12.952625780389276, "grad_norm": 0.34515994787216187, "learning_rate": 2.1912243953028317e-05, "loss": 0.0155, "step": 35270 }, { "epoch": 12.956298200514139, "grad_norm": 0.26712924242019653, "learning_rate": 2.1884891668887776e-05, "loss": 0.0163, "step": 35280 }, { "epoch": 12.959970620639002, "grad_norm": 0.3419017493724823, "learning_rate": 2.1857551683250787e-05, "loss": 0.0157, "step": 35290 }, { "epoch": 12.963643040763863, "grad_norm": 0.31367117166519165, "learning_rate": 2.1830224008076827e-05, "loss": 0.0157, "step": 35300 }, { "epoch": 12.967315460888726, "grad_norm": 0.2464066594839096, "learning_rate": 2.1802908655319924e-05, "loss": 0.0154, "step": 35310 }, { "epoch": 12.970987881013588, "grad_norm": 0.37261107563972473, "learning_rate": 2.1775605636928738e-05, "loss": 0.015, "step": 35320 }, { "epoch": 12.97466030113845, "grad_norm": 0.2980431616306305, "learning_rate": 2.1748314964846555e-05, "loss": 0.0138, "step": 35330 }, { "epoch": 12.978332721263312, "grad_norm": 0.2736537456512451, "learning_rate": 2.172103665101124e-05, "loss": 0.014, "step": 35340 }, { "epoch": 12.982005141388175, "grad_norm": 0.2310410887002945, "learning_rate": 2.1693770707355226e-05, "loss": 0.0134, "step": 35350 }, { "epoch": 12.985677561513038, "grad_norm": 1.1802501678466797, "learning_rate": 2.1666517145805605e-05, "loss": 0.0149, "step": 35360 }, { "epoch": 12.989349981637899, "grad_norm": 0.7383949756622314, "learning_rate": 2.1639275978283985e-05, "loss": 0.017, "step": 35370 }, { "epoch": 12.993022401762762, "grad_norm": 0.24470564723014832, "learning_rate": 2.1612047216706567e-05, "loss": 0.0162, "step": 35380 }, { "epoch": 12.996694821887624, "grad_norm": 0.36271047592163086, "learning_rate": 2.158483087298417e-05, "loss": 0.0146, "step": 35390 }, { "epoch": 13.000367242012485, "grad_norm": 0.29421111941337585, "learning_rate": 2.155762695902213e-05, "loss": 0.0153, "step": 35400 }, { "epoch": 13.004039662137348, "grad_norm": 0.2874378263950348, "learning_rate": 2.1530435486720363e-05, "loss": 0.0154, "step": 35410 }, { "epoch": 13.007712082262211, "grad_norm": 0.5342977046966553, "learning_rate": 2.1503256467973327e-05, "loss": 0.0136, "step": 35420 }, { "epoch": 13.011384502387074, "grad_norm": 0.19406980276107788, "learning_rate": 2.147608991467008e-05, "loss": 0.0144, "step": 35430 }, { "epoch": 13.015056922511935, "grad_norm": 0.21319445967674255, "learning_rate": 2.144893583869418e-05, "loss": 0.0143, "step": 35440 }, { "epoch": 13.018729342636798, "grad_norm": 0.49144893884658813, "learning_rate": 2.142179425192372e-05, "loss": 0.0155, "step": 35450 }, { "epoch": 13.02240176276166, "grad_norm": 0.693650484085083, "learning_rate": 2.1394665166231393e-05, "loss": 0.0148, "step": 35460 }, { "epoch": 13.026074182886521, "grad_norm": 0.5058463215827942, "learning_rate": 2.1367548593484364e-05, "loss": 0.0133, "step": 35470 }, { "epoch": 13.029746603011384, "grad_norm": 0.2814042568206787, "learning_rate": 2.134044454554433e-05, "loss": 0.014, "step": 35480 }, { "epoch": 13.033419023136247, "grad_norm": 0.4067586660385132, "learning_rate": 2.1313353034267548e-05, "loss": 0.014, "step": 35490 }, { "epoch": 13.03709144326111, "grad_norm": 0.49709174036979675, "learning_rate": 2.1286274071504735e-05, "loss": 0.0159, "step": 35500 }, { "epoch": 13.04076386338597, "grad_norm": 0.24868018925189972, "learning_rate": 2.125920766910119e-05, "loss": 0.0156, "step": 35510 }, { "epoch": 13.044436283510834, "grad_norm": 0.2205243706703186, "learning_rate": 2.1232153838896628e-05, "loss": 0.0139, "step": 35520 }, { "epoch": 13.048108703635696, "grad_norm": 0.41244959831237793, "learning_rate": 2.1205112592725366e-05, "loss": 0.0137, "step": 35530 }, { "epoch": 13.051781123760557, "grad_norm": 0.2932889461517334, "learning_rate": 2.1178083942416114e-05, "loss": 0.0147, "step": 35540 }, { "epoch": 13.05545354388542, "grad_norm": 0.3212268650531769, "learning_rate": 2.1151067899792164e-05, "loss": 0.0144, "step": 35550 }, { "epoch": 13.059125964010283, "grad_norm": 0.4146861732006073, "learning_rate": 2.1124064476671235e-05, "loss": 0.0151, "step": 35560 }, { "epoch": 13.062798384135146, "grad_norm": 0.24683205783367157, "learning_rate": 2.1097073684865516e-05, "loss": 0.0142, "step": 35570 }, { "epoch": 13.066470804260007, "grad_norm": 0.29249340295791626, "learning_rate": 2.107009553618174e-05, "loss": 0.0146, "step": 35580 }, { "epoch": 13.07014322438487, "grad_norm": 0.2131553590297699, "learning_rate": 2.1043130042421044e-05, "loss": 0.014, "step": 35590 }, { "epoch": 13.073815644509732, "grad_norm": 0.31204408407211304, "learning_rate": 2.1016177215379053e-05, "loss": 0.0143, "step": 35600 }, { "epoch": 13.077488064634593, "grad_norm": 0.23189732432365417, "learning_rate": 2.0989237066845824e-05, "loss": 0.0162, "step": 35610 }, { "epoch": 13.081160484759456, "grad_norm": 0.39200782775878906, "learning_rate": 2.096230960860594e-05, "loss": 0.0147, "step": 35620 }, { "epoch": 13.084832904884319, "grad_norm": 0.22994302213191986, "learning_rate": 2.0935394852438366e-05, "loss": 0.0143, "step": 35630 }, { "epoch": 13.088505325009182, "grad_norm": 0.7777165770530701, "learning_rate": 2.090849281011651e-05, "loss": 0.0162, "step": 35640 }, { "epoch": 13.092177745134043, "grad_norm": 0.39792555570602417, "learning_rate": 2.088160349340827e-05, "loss": 0.0154, "step": 35650 }, { "epoch": 13.095850165258906, "grad_norm": 0.477901428937912, "learning_rate": 2.085472691407594e-05, "loss": 0.0142, "step": 35660 }, { "epoch": 13.099522585383768, "grad_norm": 0.18243440985679626, "learning_rate": 2.082786308387622e-05, "loss": 0.0152, "step": 35670 }, { "epoch": 13.10319500550863, "grad_norm": 0.39385345578193665, "learning_rate": 2.0801012014560306e-05, "loss": 0.0137, "step": 35680 }, { "epoch": 13.106867425633492, "grad_norm": 0.23532018065452576, "learning_rate": 2.0774173717873746e-05, "loss": 0.0153, "step": 35690 }, { "epoch": 13.110539845758355, "grad_norm": 0.23563073575496674, "learning_rate": 2.074734820555651e-05, "loss": 0.0149, "step": 35700 }, { "epoch": 13.114212265883218, "grad_norm": 0.30490410327911377, "learning_rate": 2.0720535489343014e-05, "loss": 0.0142, "step": 35710 }, { "epoch": 13.117884686008079, "grad_norm": 0.3045545220375061, "learning_rate": 2.0693735580962044e-05, "loss": 0.0145, "step": 35720 }, { "epoch": 13.121557106132942, "grad_norm": 0.30425870418548584, "learning_rate": 2.0666948492136766e-05, "loss": 0.0158, "step": 35730 }, { "epoch": 13.125229526257804, "grad_norm": 0.3149678409099579, "learning_rate": 2.0640174234584797e-05, "loss": 0.0132, "step": 35740 }, { "epoch": 13.128901946382665, "grad_norm": 0.28925973176956177, "learning_rate": 2.061341282001808e-05, "loss": 0.013, "step": 35750 }, { "epoch": 13.132574366507528, "grad_norm": 0.43941378593444824, "learning_rate": 2.0586664260142997e-05, "loss": 0.0149, "step": 35760 }, { "epoch": 13.136246786632391, "grad_norm": 0.24036455154418945, "learning_rate": 2.0559928566660237e-05, "loss": 0.0135, "step": 35770 }, { "epoch": 13.139919206757254, "grad_norm": 0.2871135473251343, "learning_rate": 2.0533205751264945e-05, "loss": 0.0143, "step": 35780 }, { "epoch": 13.143591626882115, "grad_norm": 0.21900121867656708, "learning_rate": 2.050649582564656e-05, "loss": 0.0146, "step": 35790 }, { "epoch": 13.147264047006978, "grad_norm": 0.3898184299468994, "learning_rate": 2.04797988014889e-05, "loss": 0.0144, "step": 35800 }, { "epoch": 13.15093646713184, "grad_norm": 0.2591882646083832, "learning_rate": 2.0453114690470194e-05, "loss": 0.014, "step": 35810 }, { "epoch": 13.154608887256702, "grad_norm": 0.5478537678718567, "learning_rate": 2.0426443504262944e-05, "loss": 0.0141, "step": 35820 }, { "epoch": 13.158281307381564, "grad_norm": 0.20427218079566956, "learning_rate": 2.0399785254534027e-05, "loss": 0.0154, "step": 35830 }, { "epoch": 13.161953727506427, "grad_norm": 0.3516172766685486, "learning_rate": 2.0373139952944697e-05, "loss": 0.0147, "step": 35840 }, { "epoch": 13.16562614763129, "grad_norm": 0.25382745265960693, "learning_rate": 2.03465076111505e-05, "loss": 0.0152, "step": 35850 }, { "epoch": 13.169298567756151, "grad_norm": 0.3802040219306946, "learning_rate": 2.0319888240801305e-05, "loss": 0.0142, "step": 35860 }, { "epoch": 13.172970987881014, "grad_norm": 0.27666032314300537, "learning_rate": 2.0293281853541375e-05, "loss": 0.0136, "step": 35870 }, { "epoch": 13.176643408005877, "grad_norm": 0.24439214169979095, "learning_rate": 2.0266688461009216e-05, "loss": 0.0145, "step": 35880 }, { "epoch": 13.180315828130738, "grad_norm": 0.2354615479707718, "learning_rate": 2.0240108074837666e-05, "loss": 0.0146, "step": 35890 }, { "epoch": 13.1839882482556, "grad_norm": 0.4612107276916504, "learning_rate": 2.021354070665393e-05, "loss": 0.0151, "step": 35900 }, { "epoch": 13.187660668380463, "grad_norm": 0.295195072889328, "learning_rate": 2.0186986368079465e-05, "loss": 0.0136, "step": 35910 }, { "epoch": 13.191333088505324, "grad_norm": 0.3479766547679901, "learning_rate": 2.016044507073001e-05, "loss": 0.0145, "step": 35920 }, { "epoch": 13.195005508630187, "grad_norm": 0.4168682098388672, "learning_rate": 2.013391682621567e-05, "loss": 0.015, "step": 35930 }, { "epoch": 13.19867792875505, "grad_norm": 0.2463943511247635, "learning_rate": 2.0107401646140787e-05, "loss": 0.0161, "step": 35940 }, { "epoch": 13.202350348879913, "grad_norm": 0.22763493657112122, "learning_rate": 2.0080899542104003e-05, "loss": 0.0138, "step": 35950 }, { "epoch": 13.206022769004774, "grad_norm": 0.5128545761108398, "learning_rate": 2.0054410525698215e-05, "loss": 0.0132, "step": 35960 }, { "epoch": 13.209695189129636, "grad_norm": 0.22371938824653625, "learning_rate": 2.0027934608510653e-05, "loss": 0.0138, "step": 35970 }, { "epoch": 13.2133676092545, "grad_norm": 0.3839286267757416, "learning_rate": 2.000147180212275e-05, "loss": 0.0148, "step": 35980 }, { "epoch": 13.21704002937936, "grad_norm": 0.3480253219604492, "learning_rate": 1.9975022118110275e-05, "loss": 0.0139, "step": 35990 }, { "epoch": 13.220712449504223, "grad_norm": 0.17980600893497467, "learning_rate": 1.994858556804318e-05, "loss": 0.0151, "step": 36000 }, { "epoch": 13.224384869629086, "grad_norm": 0.5066483020782471, "learning_rate": 1.9922162163485748e-05, "loss": 0.0152, "step": 36010 }, { "epoch": 13.228057289753949, "grad_norm": 0.3301886320114136, "learning_rate": 1.989575191599643e-05, "loss": 0.0133, "step": 36020 }, { "epoch": 13.23172970987881, "grad_norm": 0.34610602259635925, "learning_rate": 1.9869354837128012e-05, "loss": 0.0144, "step": 36030 }, { "epoch": 13.235402130003672, "grad_norm": 0.36284780502319336, "learning_rate": 1.984297093842745e-05, "loss": 0.0147, "step": 36040 }, { "epoch": 13.239074550128535, "grad_norm": 0.3017236590385437, "learning_rate": 1.981660023143594e-05, "loss": 0.0142, "step": 36050 }, { "epoch": 13.242746970253396, "grad_norm": 0.3595093786716461, "learning_rate": 1.9790242727688963e-05, "loss": 0.0146, "step": 36060 }, { "epoch": 13.246419390378259, "grad_norm": 0.364146888256073, "learning_rate": 1.9763898438716165e-05, "loss": 0.0145, "step": 36070 }, { "epoch": 13.250091810503122, "grad_norm": 0.40095481276512146, "learning_rate": 1.973756737604142e-05, "loss": 0.0145, "step": 36080 }, { "epoch": 13.253764230627985, "grad_norm": 0.2995758056640625, "learning_rate": 1.9711249551182858e-05, "loss": 0.016, "step": 36090 }, { "epoch": 13.257436650752846, "grad_norm": 0.3215252161026001, "learning_rate": 1.968494497565278e-05, "loss": 0.0136, "step": 36100 }, { "epoch": 13.261109070877708, "grad_norm": 0.2896016538143158, "learning_rate": 1.9658653660957698e-05, "loss": 0.014, "step": 36110 }, { "epoch": 13.264781491002571, "grad_norm": 0.24769969284534454, "learning_rate": 1.9632375618598302e-05, "loss": 0.0137, "step": 36120 }, { "epoch": 13.268453911127432, "grad_norm": 0.2804526686668396, "learning_rate": 1.960611086006955e-05, "loss": 0.0148, "step": 36130 }, { "epoch": 13.272126331252295, "grad_norm": 0.28383588790893555, "learning_rate": 1.9579859396860513e-05, "loss": 0.014, "step": 36140 }, { "epoch": 13.275798751377158, "grad_norm": 0.3772199749946594, "learning_rate": 1.9553621240454452e-05, "loss": 0.0146, "step": 36150 }, { "epoch": 13.27947117150202, "grad_norm": 0.27752912044525146, "learning_rate": 1.9527396402328864e-05, "loss": 0.0138, "step": 36160 }, { "epoch": 13.283143591626882, "grad_norm": 0.2577204704284668, "learning_rate": 1.9501184893955376e-05, "loss": 0.0146, "step": 36170 }, { "epoch": 13.286816011751744, "grad_norm": 0.41662290692329407, "learning_rate": 1.9474986726799765e-05, "loss": 0.0159, "step": 36180 }, { "epoch": 13.290488431876607, "grad_norm": 0.5972603559494019, "learning_rate": 1.944880191232204e-05, "loss": 0.0146, "step": 36190 }, { "epoch": 13.294160852001468, "grad_norm": 0.6618220806121826, "learning_rate": 1.9422630461976304e-05, "loss": 0.0136, "step": 36200 }, { "epoch": 13.297833272126331, "grad_norm": 0.3349192440509796, "learning_rate": 1.9396472387210835e-05, "loss": 0.0148, "step": 36210 }, { "epoch": 13.301505692251194, "grad_norm": 0.40544557571411133, "learning_rate": 1.9370327699468082e-05, "loss": 0.0155, "step": 36220 }, { "epoch": 13.305178112376057, "grad_norm": 0.58525151014328, "learning_rate": 1.9344196410184594e-05, "loss": 0.0145, "step": 36230 }, { "epoch": 13.308850532500918, "grad_norm": 0.2808391749858856, "learning_rate": 1.9318078530791123e-05, "loss": 0.016, "step": 36240 }, { "epoch": 13.31252295262578, "grad_norm": 0.3852009177207947, "learning_rate": 1.9291974072712473e-05, "loss": 0.0127, "step": 36250 }, { "epoch": 13.316195372750643, "grad_norm": 0.4364300072193146, "learning_rate": 1.9265883047367656e-05, "loss": 0.0156, "step": 36260 }, { "epoch": 13.319867792875504, "grad_norm": 0.24180085957050323, "learning_rate": 1.9239805466169748e-05, "loss": 0.0159, "step": 36270 }, { "epoch": 13.323540213000367, "grad_norm": 0.2929805517196655, "learning_rate": 1.9213741340525992e-05, "loss": 0.0168, "step": 36280 }, { "epoch": 13.32721263312523, "grad_norm": 0.4169134795665741, "learning_rate": 1.9187690681837713e-05, "loss": 0.0138, "step": 36290 }, { "epoch": 13.330885053250093, "grad_norm": 0.28001320362091064, "learning_rate": 1.916165350150035e-05, "loss": 0.0136, "step": 36300 }, { "epoch": 13.334557473374954, "grad_norm": 0.3391796946525574, "learning_rate": 1.9135629810903434e-05, "loss": 0.0132, "step": 36310 }, { "epoch": 13.338229893499816, "grad_norm": 0.42951205372810364, "learning_rate": 1.9109619621430648e-05, "loss": 0.0136, "step": 36320 }, { "epoch": 13.34190231362468, "grad_norm": 0.3768037259578705, "learning_rate": 1.9083622944459712e-05, "loss": 0.0166, "step": 36330 }, { "epoch": 13.34557473374954, "grad_norm": 0.2521744966506958, "learning_rate": 1.9057639791362437e-05, "loss": 0.0159, "step": 36340 }, { "epoch": 13.349247153874403, "grad_norm": 0.3298259377479553, "learning_rate": 1.903167017350478e-05, "loss": 0.0144, "step": 36350 }, { "epoch": 13.352919573999266, "grad_norm": 0.3061855733394623, "learning_rate": 1.9005714102246707e-05, "loss": 0.0138, "step": 36360 }, { "epoch": 13.356591994124127, "grad_norm": 0.4665680527687073, "learning_rate": 1.897977158894227e-05, "loss": 0.0144, "step": 36370 }, { "epoch": 13.36026441424899, "grad_norm": 0.2904871702194214, "learning_rate": 1.8953842644939644e-05, "loss": 0.0135, "step": 36380 }, { "epoch": 13.363936834373852, "grad_norm": 0.2973885238170624, "learning_rate": 1.8927927281581015e-05, "loss": 0.0142, "step": 36390 }, { "epoch": 13.367609254498715, "grad_norm": 0.32912567257881165, "learning_rate": 1.890202551020262e-05, "loss": 0.0124, "step": 36400 }, { "epoch": 13.371281674623576, "grad_norm": 0.6835106015205383, "learning_rate": 1.8876137342134813e-05, "loss": 0.0145, "step": 36410 }, { "epoch": 13.374954094748439, "grad_norm": 0.269431471824646, "learning_rate": 1.885026278870194e-05, "loss": 0.0143, "step": 36420 }, { "epoch": 13.378626514873302, "grad_norm": 0.28154101967811584, "learning_rate": 1.8824401861222384e-05, "loss": 0.0163, "step": 36430 }, { "epoch": 13.382298934998165, "grad_norm": 0.8558709621429443, "learning_rate": 1.879855457100864e-05, "loss": 0.0151, "step": 36440 }, { "epoch": 13.385971355123026, "grad_norm": 0.23666691780090332, "learning_rate": 1.877272092936717e-05, "loss": 0.0135, "step": 36450 }, { "epoch": 13.389643775247889, "grad_norm": 0.4549174904823303, "learning_rate": 1.8746900947598467e-05, "loss": 0.0148, "step": 36460 }, { "epoch": 13.393316195372751, "grad_norm": 0.5238533616065979, "learning_rate": 1.87210946369971e-05, "loss": 0.0141, "step": 36470 }, { "epoch": 13.396988615497612, "grad_norm": 0.31599244475364685, "learning_rate": 1.8695302008851602e-05, "loss": 0.0146, "step": 36480 }, { "epoch": 13.400661035622475, "grad_norm": 0.25897443294525146, "learning_rate": 1.8669523074444568e-05, "loss": 0.0148, "step": 36490 }, { "epoch": 13.404333455747338, "grad_norm": 0.22797627747058868, "learning_rate": 1.8643757845052545e-05, "loss": 0.0139, "step": 36500 }, { "epoch": 13.408005875872199, "grad_norm": 0.5216183066368103, "learning_rate": 1.8618006331946164e-05, "loss": 0.0141, "step": 36510 }, { "epoch": 13.411678295997062, "grad_norm": 0.2890298664569855, "learning_rate": 1.859226854638999e-05, "loss": 0.0146, "step": 36520 }, { "epoch": 13.415350716121925, "grad_norm": 0.6770146489143372, "learning_rate": 1.8566544499642587e-05, "loss": 0.0142, "step": 36530 }, { "epoch": 13.419023136246787, "grad_norm": 0.5270805954933167, "learning_rate": 1.854083420295656e-05, "loss": 0.0154, "step": 36540 }, { "epoch": 13.422695556371648, "grad_norm": 0.22554203867912292, "learning_rate": 1.8515137667578453e-05, "loss": 0.0164, "step": 36550 }, { "epoch": 13.426367976496511, "grad_norm": 0.31098899245262146, "learning_rate": 1.8489454904748784e-05, "loss": 0.0149, "step": 36560 }, { "epoch": 13.430040396621374, "grad_norm": 0.16465453803539276, "learning_rate": 1.8463785925702094e-05, "loss": 0.0142, "step": 36570 }, { "epoch": 13.433712816746235, "grad_norm": 0.559220016002655, "learning_rate": 1.843813074166686e-05, "loss": 0.0154, "step": 36580 }, { "epoch": 13.437385236871098, "grad_norm": 0.2894880771636963, "learning_rate": 1.841248936386551e-05, "loss": 0.015, "step": 36590 }, { "epoch": 13.44105765699596, "grad_norm": 0.20582440495491028, "learning_rate": 1.8386861803514483e-05, "loss": 0.0129, "step": 36600 }, { "epoch": 13.444730077120823, "grad_norm": 0.28797802329063416, "learning_rate": 1.836124807182414e-05, "loss": 0.0141, "step": 36610 }, { "epoch": 13.448402497245684, "grad_norm": 0.4344905912876129, "learning_rate": 1.8335648179998783e-05, "loss": 0.0159, "step": 36620 }, { "epoch": 13.452074917370547, "grad_norm": 0.2414349913597107, "learning_rate": 1.8310062139236667e-05, "loss": 0.0138, "step": 36630 }, { "epoch": 13.45574733749541, "grad_norm": 0.3187009394168854, "learning_rate": 1.828448996073002e-05, "loss": 0.0138, "step": 36640 }, { "epoch": 13.459419757620271, "grad_norm": 0.46531903743743896, "learning_rate": 1.825893165566498e-05, "loss": 0.0151, "step": 36650 }, { "epoch": 13.463092177745134, "grad_norm": 0.24045974016189575, "learning_rate": 1.823338723522159e-05, "loss": 0.0154, "step": 36660 }, { "epoch": 13.466764597869997, "grad_norm": 0.26864519715309143, "learning_rate": 1.8207856710573878e-05, "loss": 0.014, "step": 36670 }, { "epoch": 13.47043701799486, "grad_norm": 0.2871227264404297, "learning_rate": 1.8182340092889756e-05, "loss": 0.0146, "step": 36680 }, { "epoch": 13.47410943811972, "grad_norm": 0.32427528500556946, "learning_rate": 1.8156837393331038e-05, "loss": 0.0148, "step": 36690 }, { "epoch": 13.477781858244583, "grad_norm": 0.23830345273017883, "learning_rate": 1.81313486230535e-05, "loss": 0.0129, "step": 36700 }, { "epoch": 13.481454278369446, "grad_norm": 0.49395039677619934, "learning_rate": 1.8105873793206773e-05, "loss": 0.0152, "step": 36710 }, { "epoch": 13.485126698494307, "grad_norm": 0.46493270993232727, "learning_rate": 1.8080412914934436e-05, "loss": 0.0137, "step": 36720 }, { "epoch": 13.48879911861917, "grad_norm": 0.2985070049762726, "learning_rate": 1.805496599937392e-05, "loss": 0.0154, "step": 36730 }, { "epoch": 13.492471538744033, "grad_norm": 0.39043861627578735, "learning_rate": 1.802953305765659e-05, "loss": 0.0136, "step": 36740 }, { "epoch": 13.496143958868895, "grad_norm": 0.31767502427101135, "learning_rate": 1.8004114100907653e-05, "loss": 0.0153, "step": 36750 }, { "epoch": 13.499816378993756, "grad_norm": 0.5619130730628967, "learning_rate": 1.7978709140246263e-05, "loss": 0.0135, "step": 36760 }, { "epoch": 13.50348879911862, "grad_norm": 0.30027347803115845, "learning_rate": 1.795331818678539e-05, "loss": 0.0139, "step": 36770 }, { "epoch": 13.507161219243482, "grad_norm": 0.2665104568004608, "learning_rate": 1.792794125163188e-05, "loss": 0.0148, "step": 36780 }, { "epoch": 13.510833639368343, "grad_norm": 0.3058805763721466, "learning_rate": 1.7902578345886507e-05, "loss": 0.0133, "step": 36790 }, { "epoch": 13.514506059493206, "grad_norm": 0.23810869455337524, "learning_rate": 1.7877229480643843e-05, "loss": 0.0141, "step": 36800 }, { "epoch": 13.518178479618069, "grad_norm": 0.3016137182712555, "learning_rate": 1.785189466699235e-05, "loss": 0.0163, "step": 36810 }, { "epoch": 13.521850899742931, "grad_norm": 0.24063102900981903, "learning_rate": 1.7826573916014317e-05, "loss": 0.0161, "step": 36820 }, { "epoch": 13.525523319867792, "grad_norm": 0.3597909212112427, "learning_rate": 1.7801267238785935e-05, "loss": 0.0136, "step": 36830 }, { "epoch": 13.529195739992655, "grad_norm": 0.4426382780075073, "learning_rate": 1.7775974646377187e-05, "loss": 0.0143, "step": 36840 }, { "epoch": 13.532868160117518, "grad_norm": 0.2965446412563324, "learning_rate": 1.775069614985189e-05, "loss": 0.016, "step": 36850 }, { "epoch": 13.536540580242379, "grad_norm": 0.35719409584999084, "learning_rate": 1.7725431760267757e-05, "loss": 0.0147, "step": 36860 }, { "epoch": 13.540213000367242, "grad_norm": 0.41682887077331543, "learning_rate": 1.7700181488676277e-05, "loss": 0.0131, "step": 36870 }, { "epoch": 13.543885420492105, "grad_norm": 0.37600088119506836, "learning_rate": 1.767494534612275e-05, "loss": 0.0139, "step": 36880 }, { "epoch": 13.547557840616967, "grad_norm": 0.6192750930786133, "learning_rate": 1.7649723343646367e-05, "loss": 0.0157, "step": 36890 }, { "epoch": 13.551230260741828, "grad_norm": 0.36920005083084106, "learning_rate": 1.7624515492280063e-05, "loss": 0.0138, "step": 36900 }, { "epoch": 13.554902680866691, "grad_norm": 0.302630215883255, "learning_rate": 1.7599321803050596e-05, "loss": 0.0139, "step": 36910 }, { "epoch": 13.558575100991554, "grad_norm": 0.35073789954185486, "learning_rate": 1.7574142286978574e-05, "loss": 0.0145, "step": 36920 }, { "epoch": 13.562247521116415, "grad_norm": 0.3785172998905182, "learning_rate": 1.7548976955078354e-05, "loss": 0.0151, "step": 36930 }, { "epoch": 13.565919941241278, "grad_norm": 0.41213706135749817, "learning_rate": 1.7523825818358092e-05, "loss": 0.0133, "step": 36940 }, { "epoch": 13.56959236136614, "grad_norm": 0.2185063362121582, "learning_rate": 1.7498688887819785e-05, "loss": 0.0145, "step": 36950 }, { "epoch": 13.573264781491002, "grad_norm": 0.39725300669670105, "learning_rate": 1.7473566174459143e-05, "loss": 0.0141, "step": 36960 }, { "epoch": 13.576937201615864, "grad_norm": 0.21721425652503967, "learning_rate": 1.7448457689265735e-05, "loss": 0.0135, "step": 36970 }, { "epoch": 13.580609621740727, "grad_norm": 0.37085726857185364, "learning_rate": 1.7423363443222823e-05, "loss": 0.0144, "step": 36980 }, { "epoch": 13.58428204186559, "grad_norm": 0.31604787707328796, "learning_rate": 1.7398283447307524e-05, "loss": 0.0154, "step": 36990 }, { "epoch": 13.587954461990451, "grad_norm": 0.7014448046684265, "learning_rate": 1.737321771249066e-05, "loss": 0.0135, "step": 37000 }, { "epoch": 13.591626882115314, "grad_norm": 0.24107137322425842, "learning_rate": 1.7348166249736814e-05, "loss": 0.0146, "step": 37010 }, { "epoch": 13.595299302240177, "grad_norm": 0.42208632826805115, "learning_rate": 1.732312907000439e-05, "loss": 0.0149, "step": 37020 }, { "epoch": 13.59897172236504, "grad_norm": 0.4825209379196167, "learning_rate": 1.7298106184245478e-05, "loss": 0.0146, "step": 37030 }, { "epoch": 13.6026441424899, "grad_norm": 0.29736220836639404, "learning_rate": 1.727309760340592e-05, "loss": 0.0145, "step": 37040 }, { "epoch": 13.606316562614763, "grad_norm": 0.36665403842926025, "learning_rate": 1.724810333842536e-05, "loss": 0.0137, "step": 37050 }, { "epoch": 13.609988982739626, "grad_norm": 0.37044188380241394, "learning_rate": 1.7223123400237112e-05, "loss": 0.0137, "step": 37060 }, { "epoch": 13.613661402864487, "grad_norm": 0.218028262257576, "learning_rate": 1.7198157799768238e-05, "loss": 0.014, "step": 37070 }, { "epoch": 13.61733382298935, "grad_norm": 0.27187827229499817, "learning_rate": 1.7173206547939576e-05, "loss": 0.0159, "step": 37080 }, { "epoch": 13.621006243114213, "grad_norm": 0.3837171196937561, "learning_rate": 1.7148269655665628e-05, "loss": 0.0144, "step": 37090 }, { "epoch": 13.624678663239074, "grad_norm": 0.33232951164245605, "learning_rate": 1.712334713385463e-05, "loss": 0.0147, "step": 37100 }, { "epoch": 13.628351083363937, "grad_norm": 0.30304399132728577, "learning_rate": 1.709843899340856e-05, "loss": 0.0148, "step": 37110 }, { "epoch": 13.6320235034888, "grad_norm": 0.2202390730381012, "learning_rate": 1.7073545245223084e-05, "loss": 0.0136, "step": 37120 }, { "epoch": 13.635695923613662, "grad_norm": 0.28125157952308655, "learning_rate": 1.704866590018755e-05, "loss": 0.0142, "step": 37130 }, { "epoch": 13.639368343738523, "grad_norm": 0.31244710087776184, "learning_rate": 1.7023800969185054e-05, "loss": 0.0172, "step": 37140 }, { "epoch": 13.643040763863386, "grad_norm": 1.334671974182129, "learning_rate": 1.6998950463092354e-05, "loss": 0.0147, "step": 37150 }, { "epoch": 13.646713183988249, "grad_norm": 0.3426568806171417, "learning_rate": 1.6974114392779907e-05, "loss": 0.0132, "step": 37160 }, { "epoch": 13.65038560411311, "grad_norm": 0.5581358075141907, "learning_rate": 1.694929276911183e-05, "loss": 0.0161, "step": 37170 }, { "epoch": 13.654058024237973, "grad_norm": 0.24856582283973694, "learning_rate": 1.6924485602945983e-05, "loss": 0.0146, "step": 37180 }, { "epoch": 13.657730444362835, "grad_norm": 0.25657224655151367, "learning_rate": 1.6899692905133825e-05, "loss": 0.0158, "step": 37190 }, { "epoch": 13.661402864487698, "grad_norm": 0.34695500135421753, "learning_rate": 1.6874914686520565e-05, "loss": 0.0147, "step": 37200 }, { "epoch": 13.66507528461256, "grad_norm": 0.1720697283744812, "learning_rate": 1.6850150957945e-05, "loss": 0.0157, "step": 37210 }, { "epoch": 13.668747704737422, "grad_norm": 0.3240295648574829, "learning_rate": 1.682540173023966e-05, "loss": 0.0137, "step": 37220 }, { "epoch": 13.672420124862285, "grad_norm": 0.2800706624984741, "learning_rate": 1.6800667014230665e-05, "loss": 0.0141, "step": 37230 }, { "epoch": 13.676092544987146, "grad_norm": 0.21652650833129883, "learning_rate": 1.677594682073785e-05, "loss": 0.0138, "step": 37240 }, { "epoch": 13.679764965112009, "grad_norm": 0.38232874870300293, "learning_rate": 1.6751241160574653e-05, "loss": 0.0142, "step": 37250 }, { "epoch": 13.683437385236871, "grad_norm": 0.37647193670272827, "learning_rate": 1.6726550044548156e-05, "loss": 0.0136, "step": 37260 }, { "epoch": 13.687109805361734, "grad_norm": 0.42604345083236694, "learning_rate": 1.6701873483459125e-05, "loss": 0.0131, "step": 37270 }, { "epoch": 13.690782225486595, "grad_norm": 0.5083785057067871, "learning_rate": 1.6677211488101906e-05, "loss": 0.0148, "step": 37280 }, { "epoch": 13.694454645611458, "grad_norm": 0.40500348806381226, "learning_rate": 1.6652564069264475e-05, "loss": 0.0135, "step": 37290 }, { "epoch": 13.69812706573632, "grad_norm": 0.2792101204395294, "learning_rate": 1.662793123772849e-05, "loss": 0.0135, "step": 37300 }, { "epoch": 13.701799485861182, "grad_norm": 0.24303200840950012, "learning_rate": 1.6603313004269168e-05, "loss": 0.0137, "step": 37310 }, { "epoch": 13.705471905986045, "grad_norm": 0.5943309664726257, "learning_rate": 1.657870937965536e-05, "loss": 0.0125, "step": 37320 }, { "epoch": 13.709144326110907, "grad_norm": 0.47958454489707947, "learning_rate": 1.655412037464952e-05, "loss": 0.0133, "step": 37330 }, { "epoch": 13.71281674623577, "grad_norm": 0.3554556965827942, "learning_rate": 1.652954600000773e-05, "loss": 0.0131, "step": 37340 }, { "epoch": 13.716489166360631, "grad_norm": 0.3739229738712311, "learning_rate": 1.650498626647965e-05, "loss": 0.0146, "step": 37350 }, { "epoch": 13.720161586485494, "grad_norm": 0.28917989134788513, "learning_rate": 1.6480441184808526e-05, "loss": 0.0145, "step": 37360 }, { "epoch": 13.723834006610357, "grad_norm": 0.3376738429069519, "learning_rate": 1.6455910765731242e-05, "loss": 0.0132, "step": 37370 }, { "epoch": 13.727506426735218, "grad_norm": 0.23709024488925934, "learning_rate": 1.6431395019978217e-05, "loss": 0.0149, "step": 37380 }, { "epoch": 13.73117884686008, "grad_norm": 0.8620518445968628, "learning_rate": 1.6406893958273467e-05, "loss": 0.0133, "step": 37390 }, { "epoch": 13.734851266984943, "grad_norm": 0.18566206097602844, "learning_rate": 1.6382407591334602e-05, "loss": 0.0154, "step": 37400 }, { "epoch": 13.738523687109804, "grad_norm": 0.18809960782527924, "learning_rate": 1.635793592987279e-05, "loss": 0.0128, "step": 37410 }, { "epoch": 13.742196107234667, "grad_norm": 0.37167060375213623, "learning_rate": 1.633347898459275e-05, "loss": 0.0131, "step": 37420 }, { "epoch": 13.74586852735953, "grad_norm": 0.2723248600959778, "learning_rate": 1.6309036766192805e-05, "loss": 0.0132, "step": 37430 }, { "epoch": 13.749540947484393, "grad_norm": 0.3223777115345001, "learning_rate": 1.628460928536478e-05, "loss": 0.0149, "step": 37440 }, { "epoch": 13.753213367609254, "grad_norm": 0.3420261740684509, "learning_rate": 1.6260196552794128e-05, "loss": 0.0141, "step": 37450 }, { "epoch": 13.756885787734117, "grad_norm": 0.49428796768188477, "learning_rate": 1.6235798579159766e-05, "loss": 0.0137, "step": 37460 }, { "epoch": 13.76055820785898, "grad_norm": 0.3615309000015259, "learning_rate": 1.6211415375134233e-05, "loss": 0.0173, "step": 37470 }, { "epoch": 13.764230627983842, "grad_norm": 0.4770021140575409, "learning_rate": 1.618704695138353e-05, "loss": 0.0145, "step": 37480 }, { "epoch": 13.767903048108703, "grad_norm": 0.3008681833744049, "learning_rate": 1.6162693318567273e-05, "loss": 0.0147, "step": 37490 }, { "epoch": 13.771575468233566, "grad_norm": 0.3640727698802948, "learning_rate": 1.613835448733856e-05, "loss": 0.0147, "step": 37500 }, { "epoch": 13.775247888358429, "grad_norm": 0.37107810378074646, "learning_rate": 1.6114030468344006e-05, "loss": 0.0143, "step": 37510 }, { "epoch": 13.77892030848329, "grad_norm": 0.8358626365661621, "learning_rate": 1.6089721272223755e-05, "loss": 0.0137, "step": 37520 }, { "epoch": 13.782592728608153, "grad_norm": 0.324090838432312, "learning_rate": 1.606542690961151e-05, "loss": 0.014, "step": 37530 }, { "epoch": 13.786265148733015, "grad_norm": 0.36580169200897217, "learning_rate": 1.6041147391134438e-05, "loss": 0.0139, "step": 37540 }, { "epoch": 13.789937568857876, "grad_norm": 0.27166616916656494, "learning_rate": 1.6016882727413196e-05, "loss": 0.0145, "step": 37550 }, { "epoch": 13.79360998898274, "grad_norm": 0.2043985277414322, "learning_rate": 1.5992632929062014e-05, "loss": 0.013, "step": 37560 }, { "epoch": 13.797282409107602, "grad_norm": 0.5370633602142334, "learning_rate": 1.5968398006688557e-05, "loss": 0.0165, "step": 37570 }, { "epoch": 13.800954829232465, "grad_norm": 0.29029950499534607, "learning_rate": 1.594417797089399e-05, "loss": 0.0133, "step": 37580 }, { "epoch": 13.804627249357326, "grad_norm": 0.3782190680503845, "learning_rate": 1.5919972832273015e-05, "loss": 0.0138, "step": 37590 }, { "epoch": 13.808299669482189, "grad_norm": 0.2526024281978607, "learning_rate": 1.5895782601413767e-05, "loss": 0.0157, "step": 37600 }, { "epoch": 13.811972089607051, "grad_norm": 0.5265287160873413, "learning_rate": 1.587160728889785e-05, "loss": 0.0135, "step": 37610 }, { "epoch": 13.815644509731914, "grad_norm": 0.44395965337753296, "learning_rate": 1.58474469053004e-05, "loss": 0.0128, "step": 37620 }, { "epoch": 13.819316929856775, "grad_norm": 0.37654435634613037, "learning_rate": 1.5823301461189976e-05, "loss": 0.0131, "step": 37630 }, { "epoch": 13.822989349981638, "grad_norm": 0.5041285753250122, "learning_rate": 1.5799170967128606e-05, "loss": 0.0147, "step": 37640 }, { "epoch": 13.8266617701065, "grad_norm": 0.24114301800727844, "learning_rate": 1.5775055433671814e-05, "loss": 0.0134, "step": 37650 }, { "epoch": 13.830334190231362, "grad_norm": 0.5460918545722961, "learning_rate": 1.5750954871368535e-05, "loss": 0.0148, "step": 37660 }, { "epoch": 13.834006610356225, "grad_norm": 0.22952085733413696, "learning_rate": 1.5726869290761158e-05, "loss": 0.0138, "step": 37670 }, { "epoch": 13.837679030481087, "grad_norm": 0.35708028078079224, "learning_rate": 1.5702798702385574e-05, "loss": 0.0145, "step": 37680 }, { "epoch": 13.841351450605949, "grad_norm": 0.4838898479938507, "learning_rate": 1.5678743116771038e-05, "loss": 0.013, "step": 37690 }, { "epoch": 13.845023870730811, "grad_norm": 0.3441845178604126, "learning_rate": 1.5654702544440313e-05, "loss": 0.015, "step": 37700 }, { "epoch": 13.848696290855674, "grad_norm": 0.38133496046066284, "learning_rate": 1.5630676995909532e-05, "loss": 0.0129, "step": 37710 }, { "epoch": 13.852368710980537, "grad_norm": 0.2713313102722168, "learning_rate": 1.560666648168832e-05, "loss": 0.0138, "step": 37720 }, { "epoch": 13.856041131105398, "grad_norm": 0.3270692229270935, "learning_rate": 1.5582671012279663e-05, "loss": 0.0131, "step": 37730 }, { "epoch": 13.85971355123026, "grad_norm": 0.40233054757118225, "learning_rate": 1.5558690598179997e-05, "loss": 0.0134, "step": 37740 }, { "epoch": 13.863385971355124, "grad_norm": 0.7971767783164978, "learning_rate": 1.553472524987919e-05, "loss": 0.0133, "step": 37750 }, { "epoch": 13.867058391479985, "grad_norm": 0.8247472643852234, "learning_rate": 1.5510774977860485e-05, "loss": 0.0131, "step": 37760 }, { "epoch": 13.870730811604847, "grad_norm": 0.45054587721824646, "learning_rate": 1.5486839792600532e-05, "loss": 0.014, "step": 37770 }, { "epoch": 13.87440323172971, "grad_norm": 0.28405439853668213, "learning_rate": 1.546291970456942e-05, "loss": 0.0129, "step": 37780 }, { "epoch": 13.878075651854573, "grad_norm": 0.529670238494873, "learning_rate": 1.5439014724230606e-05, "loss": 0.0139, "step": 37790 }, { "epoch": 13.881748071979434, "grad_norm": 0.2535316050052643, "learning_rate": 1.5415124862040907e-05, "loss": 0.0125, "step": 37800 }, { "epoch": 13.885420492104297, "grad_norm": 0.32091477513313293, "learning_rate": 1.53912501284506e-05, "loss": 0.0118, "step": 37810 }, { "epoch": 13.88909291222916, "grad_norm": 0.4116326570510864, "learning_rate": 1.5367390533903292e-05, "loss": 0.0137, "step": 37820 }, { "epoch": 13.89276533235402, "grad_norm": 0.29422566294670105, "learning_rate": 1.5343546088835964e-05, "loss": 0.0144, "step": 37830 }, { "epoch": 13.896437752478883, "grad_norm": 0.23721200227737427, "learning_rate": 1.531971680367901e-05, "loss": 0.0154, "step": 37840 }, { "epoch": 13.900110172603746, "grad_norm": 0.3039915859699249, "learning_rate": 1.529590268885616e-05, "loss": 0.0136, "step": 37850 }, { "epoch": 13.903782592728609, "grad_norm": 0.3524664640426636, "learning_rate": 1.5272103754784517e-05, "loss": 0.0143, "step": 37860 }, { "epoch": 13.90745501285347, "grad_norm": 0.2780155539512634, "learning_rate": 1.5248320011874523e-05, "loss": 0.0135, "step": 37870 }, { "epoch": 13.911127432978333, "grad_norm": 0.1866299957036972, "learning_rate": 1.5224551470530035e-05, "loss": 0.0122, "step": 37880 }, { "epoch": 13.914799853103196, "grad_norm": 0.297381192445755, "learning_rate": 1.5200798141148193e-05, "loss": 0.0136, "step": 37890 }, { "epoch": 13.918472273228057, "grad_norm": 0.490366131067276, "learning_rate": 1.5177060034119505e-05, "loss": 0.014, "step": 37900 }, { "epoch": 13.92214469335292, "grad_norm": 0.2518513798713684, "learning_rate": 1.5153337159827835e-05, "loss": 0.0128, "step": 37910 }, { "epoch": 13.925817113477782, "grad_norm": 0.17269141972064972, "learning_rate": 1.5129629528650397e-05, "loss": 0.0143, "step": 37920 }, { "epoch": 13.929489533602645, "grad_norm": 0.5658406615257263, "learning_rate": 1.5105937150957671e-05, "loss": 0.015, "step": 37930 }, { "epoch": 13.933161953727506, "grad_norm": 0.7431902289390564, "learning_rate": 1.5082260037113548e-05, "loss": 0.0151, "step": 37940 }, { "epoch": 13.936834373852369, "grad_norm": 0.8515739440917969, "learning_rate": 1.5058598197475187e-05, "loss": 0.0133, "step": 37950 }, { "epoch": 13.940506793977232, "grad_norm": 0.25353890657424927, "learning_rate": 1.5034951642393064e-05, "loss": 0.0148, "step": 37960 }, { "epoch": 13.944179214102093, "grad_norm": 0.19928693771362305, "learning_rate": 1.5011320382211013e-05, "loss": 0.0144, "step": 37970 }, { "epoch": 13.947851634226955, "grad_norm": 0.42640116810798645, "learning_rate": 1.4987704427266136e-05, "loss": 0.0143, "step": 37980 }, { "epoch": 13.951524054351818, "grad_norm": 0.5517075061798096, "learning_rate": 1.4964103787888839e-05, "loss": 0.0146, "step": 37990 }, { "epoch": 13.95519647447668, "grad_norm": 0.26018184423446655, "learning_rate": 1.4940518474402882e-05, "loss": 0.0145, "step": 38000 }, { "epoch": 13.958868894601542, "grad_norm": 0.3966052532196045, "learning_rate": 1.4916948497125249e-05, "loss": 0.0134, "step": 38010 }, { "epoch": 13.962541314726405, "grad_norm": 0.6212257742881775, "learning_rate": 1.4893393866366267e-05, "loss": 0.0137, "step": 38020 }, { "epoch": 13.966213734851268, "grad_norm": 0.2933609187602997, "learning_rate": 1.4869854592429506e-05, "loss": 0.0134, "step": 38030 }, { "epoch": 13.969886154976129, "grad_norm": 0.22227206826210022, "learning_rate": 1.4846330685611875e-05, "loss": 0.014, "step": 38040 }, { "epoch": 13.973558575100991, "grad_norm": 0.33709725737571716, "learning_rate": 1.482282215620352e-05, "loss": 0.0139, "step": 38050 }, { "epoch": 13.977230995225854, "grad_norm": 0.21460217237472534, "learning_rate": 1.4799329014487857e-05, "loss": 0.0144, "step": 38060 }, { "epoch": 13.980903415350717, "grad_norm": 0.23501786589622498, "learning_rate": 1.4775851270741603e-05, "loss": 0.0137, "step": 38070 }, { "epoch": 13.984575835475578, "grad_norm": 0.2783341705799103, "learning_rate": 1.4752388935234719e-05, "loss": 0.0146, "step": 38080 }, { "epoch": 13.98824825560044, "grad_norm": 0.2872925102710724, "learning_rate": 1.4728942018230401e-05, "loss": 0.0142, "step": 38090 }, { "epoch": 13.991920675725304, "grad_norm": 0.649976909160614, "learning_rate": 1.4705510529985167e-05, "loss": 0.0139, "step": 38100 }, { "epoch": 13.995593095850165, "grad_norm": 0.25524836778640747, "learning_rate": 1.4682094480748726e-05, "loss": 0.0152, "step": 38110 }, { "epoch": 13.999265515975027, "grad_norm": 0.5180684328079224, "learning_rate": 1.4658693880764035e-05, "loss": 0.0142, "step": 38120 }, { "epoch": 14.00293793609989, "grad_norm": 0.3017578125, "learning_rate": 1.463530874026735e-05, "loss": 0.0149, "step": 38130 }, { "epoch": 14.006610356224751, "grad_norm": 0.31603744626045227, "learning_rate": 1.461193906948809e-05, "loss": 0.0138, "step": 38140 }, { "epoch": 14.010282776349614, "grad_norm": 0.41736969351768494, "learning_rate": 1.4588584878648975e-05, "loss": 0.0155, "step": 38150 }, { "epoch": 14.013955196474477, "grad_norm": 0.2279786467552185, "learning_rate": 1.4565246177965896e-05, "loss": 0.0133, "step": 38160 }, { "epoch": 14.01762761659934, "grad_norm": 0.4158843755722046, "learning_rate": 1.4541922977648015e-05, "loss": 0.0133, "step": 38170 }, { "epoch": 14.0213000367242, "grad_norm": 0.7454878091812134, "learning_rate": 1.451861528789768e-05, "loss": 0.015, "step": 38180 }, { "epoch": 14.024972456849063, "grad_norm": 0.37612447142601013, "learning_rate": 1.4495323118910454e-05, "loss": 0.0133, "step": 38190 }, { "epoch": 14.028644876973926, "grad_norm": 0.3737999200820923, "learning_rate": 1.4472046480875145e-05, "loss": 0.016, "step": 38200 }, { "epoch": 14.032317297098787, "grad_norm": 0.28642934560775757, "learning_rate": 1.4448785383973745e-05, "loss": 0.0172, "step": 38210 }, { "epoch": 14.03598971722365, "grad_norm": 0.46233806014060974, "learning_rate": 1.4425539838381418e-05, "loss": 0.0146, "step": 38220 }, { "epoch": 14.039662137348513, "grad_norm": 0.2808571457862854, "learning_rate": 1.4402309854266594e-05, "loss": 0.0155, "step": 38230 }, { "epoch": 14.043334557473376, "grad_norm": 0.5514389276504517, "learning_rate": 1.4379095441790846e-05, "loss": 0.014, "step": 38240 }, { "epoch": 14.047006977598237, "grad_norm": 0.20194880664348602, "learning_rate": 1.4355896611108921e-05, "loss": 0.0136, "step": 38250 }, { "epoch": 14.0506793977231, "grad_norm": 0.36744070053100586, "learning_rate": 1.433271337236881e-05, "loss": 0.015, "step": 38260 }, { "epoch": 14.054351817847962, "grad_norm": 0.3773752748966217, "learning_rate": 1.4309545735711638e-05, "loss": 0.0136, "step": 38270 }, { "epoch": 14.058024237972823, "grad_norm": 0.35903629660606384, "learning_rate": 1.4286393711271696e-05, "loss": 0.0155, "step": 38280 }, { "epoch": 14.061696658097686, "grad_norm": 0.44978365302085876, "learning_rate": 1.4263257309176497e-05, "loss": 0.0152, "step": 38290 }, { "epoch": 14.065369078222549, "grad_norm": 0.22562438249588013, "learning_rate": 1.4240136539546678e-05, "loss": 0.0145, "step": 38300 }, { "epoch": 14.069041498347412, "grad_norm": 0.36663973331451416, "learning_rate": 1.4217031412496029e-05, "loss": 0.0146, "step": 38310 }, { "epoch": 14.072713918472273, "grad_norm": 0.30117806792259216, "learning_rate": 1.4193941938131554e-05, "loss": 0.0131, "step": 38320 }, { "epoch": 14.076386338597136, "grad_norm": 0.2551809847354889, "learning_rate": 1.4170868126553356e-05, "loss": 0.0132, "step": 38330 }, { "epoch": 14.080058758721998, "grad_norm": 0.21929650008678436, "learning_rate": 1.4147809987854682e-05, "loss": 0.0129, "step": 38340 }, { "epoch": 14.08373117884686, "grad_norm": 0.29999813437461853, "learning_rate": 1.4124767532121985e-05, "loss": 0.0146, "step": 38350 }, { "epoch": 14.087403598971722, "grad_norm": 0.288796067237854, "learning_rate": 1.41017407694348e-05, "loss": 0.0131, "step": 38360 }, { "epoch": 14.091076019096585, "grad_norm": 0.5720213651657104, "learning_rate": 1.4078729709865801e-05, "loss": 0.0146, "step": 38370 }, { "epoch": 14.094748439221448, "grad_norm": 0.3777710497379303, "learning_rate": 1.4055734363480833e-05, "loss": 0.0145, "step": 38380 }, { "epoch": 14.098420859346309, "grad_norm": 0.3949754238128662, "learning_rate": 1.4032754740338815e-05, "loss": 0.0155, "step": 38390 }, { "epoch": 14.102093279471172, "grad_norm": 0.3546322286128998, "learning_rate": 1.4009790850491844e-05, "loss": 0.0136, "step": 38400 }, { "epoch": 14.105765699596034, "grad_norm": 1.2262784242630005, "learning_rate": 1.3986842703985076e-05, "loss": 0.0121, "step": 38410 }, { "epoch": 14.109438119720895, "grad_norm": 0.2724156975746155, "learning_rate": 1.3963910310856842e-05, "loss": 0.0141, "step": 38420 }, { "epoch": 14.113110539845758, "grad_norm": 0.5569440126419067, "learning_rate": 1.394099368113853e-05, "loss": 0.0141, "step": 38430 }, { "epoch": 14.116782959970621, "grad_norm": 0.7246431112289429, "learning_rate": 1.3918092824854644e-05, "loss": 0.0153, "step": 38440 }, { "epoch": 14.120455380095484, "grad_norm": 0.7813596129417419, "learning_rate": 1.3895207752022821e-05, "loss": 0.0146, "step": 38450 }, { "epoch": 14.124127800220345, "grad_norm": 0.24793660640716553, "learning_rate": 1.3872338472653756e-05, "loss": 0.0131, "step": 38460 }, { "epoch": 14.127800220345208, "grad_norm": 0.2127167284488678, "learning_rate": 1.3849484996751233e-05, "loss": 0.0133, "step": 38470 }, { "epoch": 14.13147264047007, "grad_norm": 0.44198790192604065, "learning_rate": 1.382664733431217e-05, "loss": 0.0146, "step": 38480 }, { "epoch": 14.135145060594931, "grad_norm": 0.2515034079551697, "learning_rate": 1.3803825495326522e-05, "loss": 0.0131, "step": 38490 }, { "epoch": 14.138817480719794, "grad_norm": 0.31173163652420044, "learning_rate": 1.3781019489777324e-05, "loss": 0.0165, "step": 38500 }, { "epoch": 14.142489900844657, "grad_norm": 0.520566999912262, "learning_rate": 1.375822932764072e-05, "loss": 0.014, "step": 38510 }, { "epoch": 14.14616232096952, "grad_norm": 0.24845318496227264, "learning_rate": 1.373545501888589e-05, "loss": 0.0142, "step": 38520 }, { "epoch": 14.14983474109438, "grad_norm": 0.5307636857032776, "learning_rate": 1.3712696573475092e-05, "loss": 0.0131, "step": 38530 }, { "epoch": 14.153507161219244, "grad_norm": 0.3813515305519104, "learning_rate": 1.368995400136363e-05, "loss": 0.0124, "step": 38540 }, { "epoch": 14.157179581344106, "grad_norm": 0.47795242071151733, "learning_rate": 1.3667227312499903e-05, "loss": 0.0141, "step": 38550 }, { "epoch": 14.160852001468967, "grad_norm": 0.5456012487411499, "learning_rate": 1.3644516516825323e-05, "loss": 0.0144, "step": 38560 }, { "epoch": 14.16452442159383, "grad_norm": 0.4732775390148163, "learning_rate": 1.3621821624274356e-05, "loss": 0.0138, "step": 38570 }, { "epoch": 14.168196841718693, "grad_norm": 0.7935214638710022, "learning_rate": 1.3599142644774537e-05, "loss": 0.0129, "step": 38580 }, { "epoch": 14.171869261843556, "grad_norm": 0.9855293035507202, "learning_rate": 1.357647958824641e-05, "loss": 0.014, "step": 38590 }, { "epoch": 14.175541681968417, "grad_norm": 0.3295113742351532, "learning_rate": 1.3553832464603555e-05, "loss": 0.0146, "step": 38600 }, { "epoch": 14.17921410209328, "grad_norm": 0.2658267915248871, "learning_rate": 1.3531201283752625e-05, "loss": 0.0144, "step": 38610 }, { "epoch": 14.182886522218142, "grad_norm": 0.39297574758529663, "learning_rate": 1.350858605559323e-05, "loss": 0.0137, "step": 38620 }, { "epoch": 14.186558942343003, "grad_norm": 0.34640777111053467, "learning_rate": 1.348598679001808e-05, "loss": 0.0144, "step": 38630 }, { "epoch": 14.190231362467866, "grad_norm": 0.24880944192409515, "learning_rate": 1.3463403496912819e-05, "loss": 0.0153, "step": 38640 }, { "epoch": 14.193903782592729, "grad_norm": 0.33360761404037476, "learning_rate": 1.3440836186156187e-05, "loss": 0.014, "step": 38650 }, { "epoch": 14.19757620271759, "grad_norm": 0.2533024847507477, "learning_rate": 1.3418284867619852e-05, "loss": 0.0132, "step": 38660 }, { "epoch": 14.201248622842453, "grad_norm": 0.4352372884750366, "learning_rate": 1.3395749551168567e-05, "loss": 0.0135, "step": 38670 }, { "epoch": 14.204921042967316, "grad_norm": 0.4686737060546875, "learning_rate": 1.3373230246660024e-05, "loss": 0.0127, "step": 38680 }, { "epoch": 14.208593463092178, "grad_norm": 0.29947370290756226, "learning_rate": 1.335072696394492e-05, "loss": 0.0131, "step": 38690 }, { "epoch": 14.21226588321704, "grad_norm": 0.43862804770469666, "learning_rate": 1.3328239712866986e-05, "loss": 0.013, "step": 38700 }, { "epoch": 14.215938303341902, "grad_norm": 0.5899583101272583, "learning_rate": 1.3305768503262889e-05, "loss": 0.014, "step": 38710 }, { "epoch": 14.219610723466765, "grad_norm": 0.30283305048942566, "learning_rate": 1.3283313344962305e-05, "loss": 0.0133, "step": 38720 }, { "epoch": 14.223283143591626, "grad_norm": 0.4743841886520386, "learning_rate": 1.3260874247787863e-05, "loss": 0.0141, "step": 38730 }, { "epoch": 14.226955563716489, "grad_norm": 0.6692870855331421, "learning_rate": 1.3238451221555226e-05, "loss": 0.0153, "step": 38740 }, { "epoch": 14.230627983841352, "grad_norm": 0.20351898670196533, "learning_rate": 1.3216044276072965e-05, "loss": 0.0131, "step": 38750 }, { "epoch": 14.234300403966214, "grad_norm": 0.39161956310272217, "learning_rate": 1.3193653421142622e-05, "loss": 0.0132, "step": 38760 }, { "epoch": 14.237972824091075, "grad_norm": 0.8555998802185059, "learning_rate": 1.3171278666558756e-05, "loss": 0.015, "step": 38770 }, { "epoch": 14.241645244215938, "grad_norm": 0.5901001691818237, "learning_rate": 1.3148920022108819e-05, "loss": 0.0161, "step": 38780 }, { "epoch": 14.245317664340801, "grad_norm": 0.30731308460235596, "learning_rate": 1.312657749757324e-05, "loss": 0.0141, "step": 38790 }, { "epoch": 14.248990084465662, "grad_norm": 0.396282434463501, "learning_rate": 1.3104251102725412e-05, "loss": 0.0133, "step": 38800 }, { "epoch": 14.252662504590525, "grad_norm": 0.2674926519393921, "learning_rate": 1.3081940847331659e-05, "loss": 0.0156, "step": 38810 }, { "epoch": 14.256334924715388, "grad_norm": 0.36089974641799927, "learning_rate": 1.3059646741151222e-05, "loss": 0.0135, "step": 38820 }, { "epoch": 14.26000734484025, "grad_norm": 0.3677332103252411, "learning_rate": 1.3037368793936334e-05, "loss": 0.0121, "step": 38830 }, { "epoch": 14.263679764965111, "grad_norm": 0.20985287427902222, "learning_rate": 1.3015107015432104e-05, "loss": 0.0142, "step": 38840 }, { "epoch": 14.267352185089974, "grad_norm": 0.4543467164039612, "learning_rate": 1.2992861415376584e-05, "loss": 0.0127, "step": 38850 }, { "epoch": 14.271024605214837, "grad_norm": 0.4605911672115326, "learning_rate": 1.2970632003500782e-05, "loss": 0.0126, "step": 38860 }, { "epoch": 14.274697025339698, "grad_norm": 0.24123938381671906, "learning_rate": 1.294841878952856e-05, "loss": 0.0167, "step": 38870 }, { "epoch": 14.27836944546456, "grad_norm": 0.47361651062965393, "learning_rate": 1.2926221783176779e-05, "loss": 0.0135, "step": 38880 }, { "epoch": 14.282041865589424, "grad_norm": 0.3866773843765259, "learning_rate": 1.290404099415512e-05, "loss": 0.0132, "step": 38890 }, { "epoch": 14.285714285714286, "grad_norm": 0.6802390813827515, "learning_rate": 1.2881876432166246e-05, "loss": 0.0129, "step": 38900 }, { "epoch": 14.289386705839147, "grad_norm": 0.31160587072372437, "learning_rate": 1.2859728106905678e-05, "loss": 0.0133, "step": 38910 }, { "epoch": 14.29305912596401, "grad_norm": 0.47678545117378235, "learning_rate": 1.2837596028061832e-05, "loss": 0.0135, "step": 38920 }, { "epoch": 14.296731546088873, "grad_norm": 0.2788827419281006, "learning_rate": 1.2815480205316055e-05, "loss": 0.0128, "step": 38930 }, { "epoch": 14.300403966213734, "grad_norm": 0.3542990982532501, "learning_rate": 1.279338064834255e-05, "loss": 0.0139, "step": 38940 }, { "epoch": 14.304076386338597, "grad_norm": 0.49600130319595337, "learning_rate": 1.27712973668084e-05, "loss": 0.0138, "step": 38950 }, { "epoch": 14.30774880646346, "grad_norm": 0.3025256097316742, "learning_rate": 1.2749230370373605e-05, "loss": 0.0136, "step": 38960 }, { "epoch": 14.311421226588322, "grad_norm": 0.4994334578514099, "learning_rate": 1.2727179668691009e-05, "loss": 0.0125, "step": 38970 }, { "epoch": 14.315093646713184, "grad_norm": 0.3284779489040375, "learning_rate": 1.2705145271406327e-05, "loss": 0.0139, "step": 38980 }, { "epoch": 14.318766066838046, "grad_norm": 0.2539371848106384, "learning_rate": 1.2683127188158172e-05, "loss": 0.0129, "step": 38990 }, { "epoch": 14.322438486962909, "grad_norm": 0.4002991020679474, "learning_rate": 1.2661125428577997e-05, "loss": 0.0137, "step": 39000 }, { "epoch": 14.32611090708777, "grad_norm": 0.3732140362262726, "learning_rate": 1.2639140002290101e-05, "loss": 0.0162, "step": 39010 }, { "epoch": 14.329783327212633, "grad_norm": 0.430143803358078, "learning_rate": 1.261717091891168e-05, "loss": 0.0137, "step": 39020 }, { "epoch": 14.333455747337496, "grad_norm": 0.3491341769695282, "learning_rate": 1.2595218188052755e-05, "loss": 0.0122, "step": 39030 }, { "epoch": 14.337128167462359, "grad_norm": 0.28331536054611206, "learning_rate": 1.2573281819316174e-05, "loss": 0.0134, "step": 39040 }, { "epoch": 14.34080058758722, "grad_norm": 1.3033314943313599, "learning_rate": 1.2551361822297674e-05, "loss": 0.0146, "step": 39050 }, { "epoch": 14.344473007712082, "grad_norm": 0.28641819953918457, "learning_rate": 1.2529458206585804e-05, "loss": 0.0135, "step": 39060 }, { "epoch": 14.348145427836945, "grad_norm": 0.21426913142204285, "learning_rate": 1.250757098176194e-05, "loss": 0.013, "step": 39070 }, { "epoch": 14.351817847961806, "grad_norm": 0.27954113483428955, "learning_rate": 1.2485700157400294e-05, "loss": 0.012, "step": 39080 }, { "epoch": 14.355490268086669, "grad_norm": 0.4907306134700775, "learning_rate": 1.246384574306792e-05, "loss": 0.0131, "step": 39090 }, { "epoch": 14.359162688211532, "grad_norm": 0.43154245615005493, "learning_rate": 1.2442007748324664e-05, "loss": 0.0123, "step": 39100 }, { "epoch": 14.362835108336395, "grad_norm": 0.17270593345165253, "learning_rate": 1.2420186182723237e-05, "loss": 0.0138, "step": 39110 }, { "epoch": 14.366507528461256, "grad_norm": 0.4330189526081085, "learning_rate": 1.23983810558091e-05, "loss": 0.0127, "step": 39120 }, { "epoch": 14.370179948586118, "grad_norm": 0.6385335922241211, "learning_rate": 1.2376592377120582e-05, "loss": 0.0122, "step": 39130 }, { "epoch": 14.373852368710981, "grad_norm": 0.37965869903564453, "learning_rate": 1.2354820156188768e-05, "loss": 0.0156, "step": 39140 }, { "epoch": 14.377524788835842, "grad_norm": 0.2337585985660553, "learning_rate": 1.2333064402537597e-05, "loss": 0.0139, "step": 39150 }, { "epoch": 14.381197208960705, "grad_norm": 0.35615095496177673, "learning_rate": 1.2311325125683754e-05, "loss": 0.0134, "step": 39160 }, { "epoch": 14.384869629085568, "grad_norm": 0.28997182846069336, "learning_rate": 1.228960233513673e-05, "loss": 0.0126, "step": 39170 }, { "epoch": 14.388542049210429, "grad_norm": 0.3565957248210907, "learning_rate": 1.2267896040398836e-05, "loss": 0.0123, "step": 39180 }, { "epoch": 14.392214469335292, "grad_norm": 0.15636369585990906, "learning_rate": 1.2246206250965125e-05, "loss": 0.0133, "step": 39190 }, { "epoch": 14.395886889460154, "grad_norm": 0.5812007784843445, "learning_rate": 1.2224532976323444e-05, "loss": 0.0138, "step": 39200 }, { "epoch": 14.399559309585017, "grad_norm": 0.2651045620441437, "learning_rate": 1.220287622595444e-05, "loss": 0.0123, "step": 39210 }, { "epoch": 14.403231729709878, "grad_norm": 0.518202543258667, "learning_rate": 1.2181236009331497e-05, "loss": 0.0142, "step": 39220 }, { "epoch": 14.406904149834741, "grad_norm": 0.5359904170036316, "learning_rate": 1.2159612335920784e-05, "loss": 0.0132, "step": 39230 }, { "epoch": 14.410576569959604, "grad_norm": 0.47535037994384766, "learning_rate": 1.2138005215181219e-05, "loss": 0.0136, "step": 39240 }, { "epoch": 14.414248990084465, "grad_norm": 0.2418341040611267, "learning_rate": 1.2116414656564512e-05, "loss": 0.0159, "step": 39250 }, { "epoch": 14.417921410209328, "grad_norm": 0.24922525882720947, "learning_rate": 1.2094840669515095e-05, "loss": 0.0136, "step": 39260 }, { "epoch": 14.42159383033419, "grad_norm": 0.3072936534881592, "learning_rate": 1.2073283263470152e-05, "loss": 0.0122, "step": 39270 }, { "epoch": 14.425266250459053, "grad_norm": 0.24613860249519348, "learning_rate": 1.2051742447859653e-05, "loss": 0.0135, "step": 39280 }, { "epoch": 14.428938670583914, "grad_norm": 0.2304149866104126, "learning_rate": 1.203021823210626e-05, "loss": 0.0147, "step": 39290 }, { "epoch": 14.432611090708777, "grad_norm": 0.40547046065330505, "learning_rate": 1.2008710625625397e-05, "loss": 0.0129, "step": 39300 }, { "epoch": 14.43628351083364, "grad_norm": 0.4483988285064697, "learning_rate": 1.1987219637825236e-05, "loss": 0.0143, "step": 39310 }, { "epoch": 14.4399559309585, "grad_norm": 0.6252856254577637, "learning_rate": 1.1965745278106656e-05, "loss": 0.0147, "step": 39320 }, { "epoch": 14.443628351083364, "grad_norm": 0.26412534713745117, "learning_rate": 1.1944287555863259e-05, "loss": 0.0123, "step": 39330 }, { "epoch": 14.447300771208226, "grad_norm": 0.3253084123134613, "learning_rate": 1.1922846480481409e-05, "loss": 0.0159, "step": 39340 }, { "epoch": 14.45097319133309, "grad_norm": 0.3737611174583435, "learning_rate": 1.1901422061340133e-05, "loss": 0.0133, "step": 39350 }, { "epoch": 14.45464561145795, "grad_norm": 0.3126097619533539, "learning_rate": 1.1880014307811227e-05, "loss": 0.0126, "step": 39360 }, { "epoch": 14.458318031582813, "grad_norm": 0.36380869150161743, "learning_rate": 1.1858623229259142e-05, "loss": 0.0118, "step": 39370 }, { "epoch": 14.461990451707676, "grad_norm": 0.37916967272758484, "learning_rate": 1.1837248835041092e-05, "loss": 0.0137, "step": 39380 }, { "epoch": 14.465662871832537, "grad_norm": 0.3882027268409729, "learning_rate": 1.1815891134506935e-05, "loss": 0.0139, "step": 39390 }, { "epoch": 14.4693352919574, "grad_norm": 0.43715837597846985, "learning_rate": 1.1794550136999277e-05, "loss": 0.0128, "step": 39400 }, { "epoch": 14.473007712082262, "grad_norm": 0.33004775643348694, "learning_rate": 1.1773225851853387e-05, "loss": 0.0138, "step": 39410 }, { "epoch": 14.476680132207125, "grad_norm": 0.25838950276374817, "learning_rate": 1.1751918288397235e-05, "loss": 0.0125, "step": 39420 }, { "epoch": 14.480352552331986, "grad_norm": 0.6296628713607788, "learning_rate": 1.1730627455951449e-05, "loss": 0.0123, "step": 39430 }, { "epoch": 14.484024972456849, "grad_norm": 0.7656412124633789, "learning_rate": 1.17093533638294e-05, "loss": 0.0128, "step": 39440 }, { "epoch": 14.487697392581712, "grad_norm": 0.2921547293663025, "learning_rate": 1.168809602133708e-05, "loss": 0.0131, "step": 39450 }, { "epoch": 14.491369812706573, "grad_norm": 0.31485995650291443, "learning_rate": 1.1666855437773162e-05, "loss": 0.0132, "step": 39460 }, { "epoch": 14.495042232831436, "grad_norm": 0.28736722469329834, "learning_rate": 1.1645631622429026e-05, "loss": 0.0132, "step": 39470 }, { "epoch": 14.498714652956298, "grad_norm": 0.3613273799419403, "learning_rate": 1.1624424584588673e-05, "loss": 0.0134, "step": 39480 }, { "epoch": 14.502387073081161, "grad_norm": 0.3618992865085602, "learning_rate": 1.1603234333528768e-05, "loss": 0.0139, "step": 39490 }, { "epoch": 14.506059493206022, "grad_norm": 0.5778965353965759, "learning_rate": 1.1582060878518674e-05, "loss": 0.0129, "step": 39500 }, { "epoch": 14.509731913330885, "grad_norm": 0.2557634711265564, "learning_rate": 1.1560904228820375e-05, "loss": 0.0126, "step": 39510 }, { "epoch": 14.513404333455748, "grad_norm": 0.38748571276664734, "learning_rate": 1.1539764393688478e-05, "loss": 0.0155, "step": 39520 }, { "epoch": 14.517076753580609, "grad_norm": 0.6983728408813477, "learning_rate": 1.151864138237031e-05, "loss": 0.0126, "step": 39530 }, { "epoch": 14.520749173705472, "grad_norm": 0.5171712040901184, "learning_rate": 1.1497535204105769e-05, "loss": 0.0143, "step": 39540 }, { "epoch": 14.524421593830334, "grad_norm": 0.2955350875854492, "learning_rate": 1.14764458681274e-05, "loss": 0.0124, "step": 39550 }, { "epoch": 14.528094013955197, "grad_norm": 0.21496661007404327, "learning_rate": 1.1455373383660434e-05, "loss": 0.0127, "step": 39560 }, { "epoch": 14.531766434080058, "grad_norm": 0.6918035745620728, "learning_rate": 1.1434317759922664e-05, "loss": 0.0144, "step": 39570 }, { "epoch": 14.535438854204921, "grad_norm": 0.778753936290741, "learning_rate": 1.1413279006124528e-05, "loss": 0.0135, "step": 39580 }, { "epoch": 14.539111274329784, "grad_norm": 0.2958500385284424, "learning_rate": 1.1392257131469119e-05, "loss": 0.0136, "step": 39590 }, { "epoch": 14.542783694454645, "grad_norm": 0.415815144777298, "learning_rate": 1.1371252145152095e-05, "loss": 0.0132, "step": 39600 }, { "epoch": 14.546456114579508, "grad_norm": 0.4018492102622986, "learning_rate": 1.1350264056361776e-05, "loss": 0.0144, "step": 39610 }, { "epoch": 14.55012853470437, "grad_norm": 0.38691210746765137, "learning_rate": 1.1329292874279029e-05, "loss": 0.0134, "step": 39620 }, { "epoch": 14.553800954829232, "grad_norm": 0.2947126626968384, "learning_rate": 1.1308338608077401e-05, "loss": 0.0137, "step": 39630 }, { "epoch": 14.557473374954094, "grad_norm": 0.4076423943042755, "learning_rate": 1.1287401266922981e-05, "loss": 0.0127, "step": 39640 }, { "epoch": 14.561145795078957, "grad_norm": 0.491349458694458, "learning_rate": 1.126648085997446e-05, "loss": 0.0129, "step": 39650 }, { "epoch": 14.56481821520382, "grad_norm": 0.4068666100502014, "learning_rate": 1.124557739638316e-05, "loss": 0.0147, "step": 39660 }, { "epoch": 14.568490635328681, "grad_norm": 0.2801954746246338, "learning_rate": 1.1224690885292955e-05, "loss": 0.013, "step": 39670 }, { "epoch": 14.572163055453544, "grad_norm": 0.3649669289588928, "learning_rate": 1.1203821335840303e-05, "loss": 0.0134, "step": 39680 }, { "epoch": 14.575835475578407, "grad_norm": 1.002479910850525, "learning_rate": 1.1182968757154278e-05, "loss": 0.013, "step": 39690 }, { "epoch": 14.57950789570327, "grad_norm": 0.8622420430183411, "learning_rate": 1.1162133158356492e-05, "loss": 0.0133, "step": 39700 }, { "epoch": 14.58318031582813, "grad_norm": 0.4199959337711334, "learning_rate": 1.1141314548561133e-05, "loss": 0.015, "step": 39710 }, { "epoch": 14.586852735952993, "grad_norm": 0.6464262008666992, "learning_rate": 1.1120512936874994e-05, "loss": 0.0129, "step": 39720 }, { "epoch": 14.590525156077856, "grad_norm": 1.363634467124939, "learning_rate": 1.109972833239739e-05, "loss": 0.0134, "step": 39730 }, { "epoch": 14.594197576202717, "grad_norm": 0.7857040762901306, "learning_rate": 1.1078960744220218e-05, "loss": 0.0131, "step": 39740 }, { "epoch": 14.59786999632758, "grad_norm": 0.9173161387443542, "learning_rate": 1.1058210181427914e-05, "loss": 0.0135, "step": 39750 }, { "epoch": 14.601542416452443, "grad_norm": 0.4166286885738373, "learning_rate": 1.10374766530975e-05, "loss": 0.0134, "step": 39760 }, { "epoch": 14.605214836577304, "grad_norm": 0.3938673734664917, "learning_rate": 1.1016760168298518e-05, "loss": 0.0132, "step": 39770 }, { "epoch": 14.608887256702166, "grad_norm": 0.6108233332633972, "learning_rate": 1.0996060736093044e-05, "loss": 0.0124, "step": 39780 }, { "epoch": 14.61255967682703, "grad_norm": 0.29816582798957825, "learning_rate": 1.0975378365535749e-05, "loss": 0.0141, "step": 39790 }, { "epoch": 14.616232096951892, "grad_norm": 0.5763866305351257, "learning_rate": 1.0954713065673777e-05, "loss": 0.0127, "step": 39800 }, { "epoch": 14.619904517076753, "grad_norm": 0.20731930434703827, "learning_rate": 1.0934064845546832e-05, "loss": 0.0154, "step": 39810 }, { "epoch": 14.623576937201616, "grad_norm": 0.2814958989620209, "learning_rate": 1.0913433714187166e-05, "loss": 0.014, "step": 39820 }, { "epoch": 14.627249357326479, "grad_norm": 0.6222380995750427, "learning_rate": 1.089281968061951e-05, "loss": 0.0133, "step": 39830 }, { "epoch": 14.630921777451341, "grad_norm": 0.37603825330734253, "learning_rate": 1.0872222753861172e-05, "loss": 0.0133, "step": 39840 }, { "epoch": 14.634594197576202, "grad_norm": 0.16740481555461884, "learning_rate": 1.0851642942921925e-05, "loss": 0.0146, "step": 39850 }, { "epoch": 14.638266617701065, "grad_norm": 0.9885563254356384, "learning_rate": 1.0831080256804093e-05, "loss": 0.015, "step": 39860 }, { "epoch": 14.641939037825928, "grad_norm": 1.018775224685669, "learning_rate": 1.0810534704502478e-05, "loss": 0.0147, "step": 39870 }, { "epoch": 14.645611457950789, "grad_norm": 0.2838963270187378, "learning_rate": 1.0790006295004423e-05, "loss": 0.0121, "step": 39880 }, { "epoch": 14.649283878075652, "grad_norm": 0.46677166223526, "learning_rate": 1.076949503728974e-05, "loss": 0.0142, "step": 39890 }, { "epoch": 14.652956298200515, "grad_norm": 0.20928342640399933, "learning_rate": 1.074900094033074e-05, "loss": 0.013, "step": 39900 }, { "epoch": 14.656628718325376, "grad_norm": 0.2516467571258545, "learning_rate": 1.0728524013092256e-05, "loss": 0.0174, "step": 39910 }, { "epoch": 14.660301138450238, "grad_norm": 0.3698095977306366, "learning_rate": 1.0708064264531581e-05, "loss": 0.0157, "step": 39920 }, { "epoch": 14.663973558575101, "grad_norm": 0.3436869978904724, "learning_rate": 1.068762170359851e-05, "loss": 0.0126, "step": 39930 }, { "epoch": 14.667645978699964, "grad_norm": 0.44795310497283936, "learning_rate": 1.0667196339235297e-05, "loss": 0.0128, "step": 39940 }, { "epoch": 14.671318398824825, "grad_norm": 0.2559700906276703, "learning_rate": 1.0646788180376716e-05, "loss": 0.0134, "step": 39950 }, { "epoch": 14.674990818949688, "grad_norm": 0.39888831973075867, "learning_rate": 1.0626397235949976e-05, "loss": 0.0129, "step": 39960 }, { "epoch": 14.67866323907455, "grad_norm": 0.38910362124443054, "learning_rate": 1.0606023514874751e-05, "loss": 0.0119, "step": 39970 }, { "epoch": 14.682335659199412, "grad_norm": 0.3876059353351593, "learning_rate": 1.0585667026063233e-05, "loss": 0.013, "step": 39980 }, { "epoch": 14.686008079324274, "grad_norm": 0.4659295380115509, "learning_rate": 1.0565327778420025e-05, "loss": 0.0135, "step": 39990 }, { "epoch": 14.689680499449137, "grad_norm": 0.6414388418197632, "learning_rate": 1.0545005780842193e-05, "loss": 0.0131, "step": 40000 }, { "epoch": 14.693352919574, "grad_norm": 0.32320013642311096, "learning_rate": 1.0524701042219299e-05, "loss": 0.0116, "step": 40010 }, { "epoch": 14.697025339698861, "grad_norm": 0.9034081101417542, "learning_rate": 1.0504413571433303e-05, "loss": 0.0127, "step": 40020 }, { "epoch": 14.700697759823724, "grad_norm": 0.47725406289100647, "learning_rate": 1.0484143377358624e-05, "loss": 0.0138, "step": 40030 }, { "epoch": 14.704370179948587, "grad_norm": 0.18332085013389587, "learning_rate": 1.046389046886217e-05, "loss": 0.0124, "step": 40040 }, { "epoch": 14.708042600073448, "grad_norm": 0.5009816288948059, "learning_rate": 1.044365485480322e-05, "loss": 0.0135, "step": 40050 }, { "epoch": 14.71171502019831, "grad_norm": 0.3282582461833954, "learning_rate": 1.0423436544033522e-05, "loss": 0.0143, "step": 40060 }, { "epoch": 14.715387440323173, "grad_norm": 0.24682745337486267, "learning_rate": 1.0403235545397272e-05, "loss": 0.0135, "step": 40070 }, { "epoch": 14.719059860448036, "grad_norm": 0.23580805957317352, "learning_rate": 1.0383051867731043e-05, "loss": 0.0127, "step": 40080 }, { "epoch": 14.722732280572897, "grad_norm": 0.3264089524745941, "learning_rate": 1.0362885519863896e-05, "loss": 0.0127, "step": 40090 }, { "epoch": 14.72640470069776, "grad_norm": 0.9135966300964355, "learning_rate": 1.034273651061724e-05, "loss": 0.0122, "step": 40100 }, { "epoch": 14.730077120822623, "grad_norm": 0.6126454472541809, "learning_rate": 1.0322604848804973e-05, "loss": 0.0124, "step": 40110 }, { "epoch": 14.733749540947484, "grad_norm": 0.44618549942970276, "learning_rate": 1.030249054323335e-05, "loss": 0.013, "step": 40120 }, { "epoch": 14.737421961072346, "grad_norm": 0.3077215254306793, "learning_rate": 1.0282393602701041e-05, "loss": 0.0121, "step": 40130 }, { "epoch": 14.74109438119721, "grad_norm": 0.42048782110214233, "learning_rate": 1.026231403599915e-05, "loss": 0.0134, "step": 40140 }, { "epoch": 14.744766801322072, "grad_norm": 0.3773375153541565, "learning_rate": 1.0242251851911155e-05, "loss": 0.012, "step": 40150 }, { "epoch": 14.748439221446933, "grad_norm": 0.468683123588562, "learning_rate": 1.0222207059212918e-05, "loss": 0.0143, "step": 40160 }, { "epoch": 14.752111641571796, "grad_norm": 0.5739783644676208, "learning_rate": 1.0202179666672746e-05, "loss": 0.0136, "step": 40170 }, { "epoch": 14.755784061696659, "grad_norm": 0.25812655687332153, "learning_rate": 1.0182169683051284e-05, "loss": 0.0119, "step": 40180 }, { "epoch": 14.75945648182152, "grad_norm": 0.23383644223213196, "learning_rate": 1.0162177117101567e-05, "loss": 0.0137, "step": 40190 }, { "epoch": 14.763128901946382, "grad_norm": 0.29809606075286865, "learning_rate": 1.0142201977569044e-05, "loss": 0.0136, "step": 40200 }, { "epoch": 14.766801322071245, "grad_norm": 0.29434990882873535, "learning_rate": 1.0122244273191517e-05, "loss": 0.0121, "step": 40210 }, { "epoch": 14.770473742196106, "grad_norm": 1.1037476062774658, "learning_rate": 1.0102304012699137e-05, "loss": 0.0133, "step": 40220 }, { "epoch": 14.774146162320969, "grad_norm": 0.8061330318450928, "learning_rate": 1.0082381204814484e-05, "loss": 0.0133, "step": 40230 }, { "epoch": 14.777818582445832, "grad_norm": 0.206083744764328, "learning_rate": 1.0062475858252463e-05, "loss": 0.0127, "step": 40240 }, { "epoch": 14.781491002570695, "grad_norm": 0.35948818922042847, "learning_rate": 1.0042587981720326e-05, "loss": 0.0133, "step": 40250 }, { "epoch": 14.785163422695556, "grad_norm": 0.6029041409492493, "learning_rate": 1.002271758391774e-05, "loss": 0.0132, "step": 40260 }, { "epoch": 14.788835842820419, "grad_norm": 0.5854074954986572, "learning_rate": 1.0002864673536676e-05, "loss": 0.0129, "step": 40270 }, { "epoch": 14.792508262945281, "grad_norm": 0.18727679550647736, "learning_rate": 9.983029259261473e-06, "loss": 0.0123, "step": 40280 }, { "epoch": 14.796180683070144, "grad_norm": 0.30616679787635803, "learning_rate": 9.963211349768798e-06, "loss": 0.0124, "step": 40290 }, { "epoch": 14.799853103195005, "grad_norm": 0.367549866437912, "learning_rate": 9.94341095372771e-06, "loss": 0.0137, "step": 40300 }, { "epoch": 14.803525523319868, "grad_norm": 0.5112230777740479, "learning_rate": 9.923628079799546e-06, "loss": 0.0138, "step": 40310 }, { "epoch": 14.80719794344473, "grad_norm": 0.2646074891090393, "learning_rate": 9.903862736638031e-06, "loss": 0.0137, "step": 40320 }, { "epoch": 14.810870363569592, "grad_norm": 0.4764687716960907, "learning_rate": 9.884114932889171e-06, "loss": 0.0124, "step": 40330 }, { "epoch": 14.814542783694455, "grad_norm": 0.24572104215621948, "learning_rate": 9.864384677191357e-06, "loss": 0.0142, "step": 40340 }, { "epoch": 14.818215203819317, "grad_norm": 0.4057663381099701, "learning_rate": 9.844671978175247e-06, "loss": 0.0135, "step": 40350 }, { "epoch": 14.821887623944178, "grad_norm": 0.3723219633102417, "learning_rate": 9.824976844463862e-06, "loss": 0.0126, "step": 40360 }, { "epoch": 14.825560044069041, "grad_norm": 0.6219672560691833, "learning_rate": 9.805299284672525e-06, "loss": 0.0131, "step": 40370 }, { "epoch": 14.829232464193904, "grad_norm": 0.6842615008354187, "learning_rate": 9.785639307408845e-06, "loss": 0.0132, "step": 40380 }, { "epoch": 14.832904884318767, "grad_norm": 0.6611681580543518, "learning_rate": 9.765996921272792e-06, "loss": 0.0129, "step": 40390 }, { "epoch": 14.836577304443628, "grad_norm": 0.21533028781414032, "learning_rate": 9.746372134856595e-06, "loss": 0.013, "step": 40400 }, { "epoch": 14.84024972456849, "grad_norm": 0.25044262409210205, "learning_rate": 9.726764956744805e-06, "loss": 0.0132, "step": 40410 }, { "epoch": 14.843922144693353, "grad_norm": 0.6031060218811035, "learning_rate": 9.707175395514278e-06, "loss": 0.0136, "step": 40420 }, { "epoch": 14.847594564818214, "grad_norm": 0.2398698329925537, "learning_rate": 9.68760345973415e-06, "loss": 0.0121, "step": 40430 }, { "epoch": 14.851266984943077, "grad_norm": 0.2753085792064667, "learning_rate": 9.668049157965848e-06, "loss": 0.0142, "step": 40440 }, { "epoch": 14.85493940506794, "grad_norm": 0.2707692086696625, "learning_rate": 9.648512498763074e-06, "loss": 0.0124, "step": 40450 }, { "epoch": 14.858611825192803, "grad_norm": 0.6872156262397766, "learning_rate": 9.62899349067185e-06, "loss": 0.0127, "step": 40460 }, { "epoch": 14.862284245317664, "grad_norm": 0.2195097953081131, "learning_rate": 9.609492142230442e-06, "loss": 0.0128, "step": 40470 }, { "epoch": 14.865956665442527, "grad_norm": 0.2903066873550415, "learning_rate": 9.59000846196939e-06, "loss": 0.0139, "step": 40480 }, { "epoch": 14.86962908556739, "grad_norm": 0.22594718635082245, "learning_rate": 9.570542458411552e-06, "loss": 0.0123, "step": 40490 }, { "epoch": 14.87330150569225, "grad_norm": 0.31264159083366394, "learning_rate": 9.551094140071993e-06, "loss": 0.0126, "step": 40500 }, { "epoch": 14.876973925817113, "grad_norm": 0.932300329208374, "learning_rate": 9.53166351545806e-06, "loss": 0.0129, "step": 40510 }, { "epoch": 14.880646345941976, "grad_norm": 0.27230289578437805, "learning_rate": 9.512250593069393e-06, "loss": 0.0131, "step": 40520 }, { "epoch": 14.884318766066839, "grad_norm": 0.5630219578742981, "learning_rate": 9.492855381397859e-06, "loss": 0.013, "step": 40530 }, { "epoch": 14.8879911861917, "grad_norm": 0.4918246567249298, "learning_rate": 9.473477888927556e-06, "loss": 0.012, "step": 40540 }, { "epoch": 14.891663606316563, "grad_norm": 0.28147995471954346, "learning_rate": 9.454118124134892e-06, "loss": 0.0123, "step": 40550 }, { "epoch": 14.895336026441425, "grad_norm": 0.6947166323661804, "learning_rate": 9.434776095488463e-06, "loss": 0.0127, "step": 40560 }, { "epoch": 14.899008446566286, "grad_norm": 0.5053012371063232, "learning_rate": 9.41545181144915e-06, "loss": 0.0126, "step": 40570 }, { "epoch": 14.90268086669115, "grad_norm": 0.5217415690422058, "learning_rate": 9.396145280470025e-06, "loss": 0.0129, "step": 40580 }, { "epoch": 14.906353286816012, "grad_norm": 0.4806729555130005, "learning_rate": 9.376856510996445e-06, "loss": 0.0138, "step": 40590 }, { "epoch": 14.910025706940875, "grad_norm": 0.42067238688468933, "learning_rate": 9.357585511465955e-06, "loss": 0.0142, "step": 40600 }, { "epoch": 14.913698127065736, "grad_norm": 0.3396258056163788, "learning_rate": 9.338332290308367e-06, "loss": 0.0125, "step": 40610 }, { "epoch": 14.917370547190599, "grad_norm": 0.606428325176239, "learning_rate": 9.319096855945675e-06, "loss": 0.0135, "step": 40620 }, { "epoch": 14.921042967315461, "grad_norm": 0.26980793476104736, "learning_rate": 9.299879216792118e-06, "loss": 0.012, "step": 40630 }, { "epoch": 14.924715387440322, "grad_norm": 0.6051700711250305, "learning_rate": 9.280679381254132e-06, "loss": 0.0123, "step": 40640 }, { "epoch": 14.928387807565185, "grad_norm": 0.6438549757003784, "learning_rate": 9.261497357730397e-06, "loss": 0.0135, "step": 40650 }, { "epoch": 14.932060227690048, "grad_norm": 0.6630073189735413, "learning_rate": 9.242333154611777e-06, "loss": 0.0129, "step": 40660 }, { "epoch": 14.93573264781491, "grad_norm": 0.5277788639068604, "learning_rate": 9.223186780281323e-06, "loss": 0.0133, "step": 40670 }, { "epoch": 14.939405067939772, "grad_norm": 0.25674968957901, "learning_rate": 9.20405824311435e-06, "loss": 0.0127, "step": 40680 }, { "epoch": 14.943077488064635, "grad_norm": 0.40851497650146484, "learning_rate": 9.184947551478301e-06, "loss": 0.0132, "step": 40690 }, { "epoch": 14.946749908189497, "grad_norm": 0.7694573998451233, "learning_rate": 9.165854713732835e-06, "loss": 0.0122, "step": 40700 }, { "epoch": 14.950422328314358, "grad_norm": 0.26155924797058105, "learning_rate": 9.146779738229838e-06, "loss": 0.0117, "step": 40710 }, { "epoch": 14.954094748439221, "grad_norm": 0.26289281249046326, "learning_rate": 9.12772263331334e-06, "loss": 0.0151, "step": 40720 }, { "epoch": 14.957767168564084, "grad_norm": 0.5008358359336853, "learning_rate": 9.10868340731954e-06, "loss": 0.0146, "step": 40730 }, { "epoch": 14.961439588688947, "grad_norm": 0.7254243493080139, "learning_rate": 9.089662068576887e-06, "loss": 0.0138, "step": 40740 }, { "epoch": 14.965112008813808, "grad_norm": 0.5093281269073486, "learning_rate": 9.07065862540593e-06, "loss": 0.0144, "step": 40750 }, { "epoch": 14.96878442893867, "grad_norm": 0.2828194499015808, "learning_rate": 9.051673086119422e-06, "loss": 0.0149, "step": 40760 }, { "epoch": 14.972456849063533, "grad_norm": 0.384288489818573, "learning_rate": 9.0327054590223e-06, "loss": 0.0124, "step": 40770 }, { "epoch": 14.976129269188394, "grad_norm": 0.2089531123638153, "learning_rate": 9.01375575241164e-06, "loss": 0.0123, "step": 40780 }, { "epoch": 14.979801689313257, "grad_norm": 0.37399908900260925, "learning_rate": 8.994823974576672e-06, "loss": 0.0138, "step": 40790 }, { "epoch": 14.98347410943812, "grad_norm": 0.3908240795135498, "learning_rate": 8.975910133798816e-06, "loss": 0.012, "step": 40800 }, { "epoch": 14.987146529562981, "grad_norm": 0.33313634991645813, "learning_rate": 8.957014238351635e-06, "loss": 0.0134, "step": 40810 }, { "epoch": 14.990818949687844, "grad_norm": 0.36305487155914307, "learning_rate": 8.938136296500826e-06, "loss": 0.0121, "step": 40820 }, { "epoch": 14.994491369812707, "grad_norm": 0.4679270386695862, "learning_rate": 8.919276316504227e-06, "loss": 0.0122, "step": 40830 }, { "epoch": 14.99816378993757, "grad_norm": 0.4502190053462982, "learning_rate": 8.900434306611866e-06, "loss": 0.0118, "step": 40840 }, { "epoch": 15.00183621006243, "grad_norm": 0.4238332509994507, "learning_rate": 8.881610275065861e-06, "loss": 0.0124, "step": 40850 }, { "epoch": 15.005508630187293, "grad_norm": 0.5289883017539978, "learning_rate": 8.862804230100463e-06, "loss": 0.013, "step": 40860 }, { "epoch": 15.009181050312156, "grad_norm": 0.5898679494857788, "learning_rate": 8.844016179942116e-06, "loss": 0.0132, "step": 40870 }, { "epoch": 15.012853470437017, "grad_norm": 0.2492133527994156, "learning_rate": 8.825246132809329e-06, "loss": 0.0123, "step": 40880 }, { "epoch": 15.01652589056188, "grad_norm": 0.36119765043258667, "learning_rate": 8.80649409691275e-06, "loss": 0.0131, "step": 40890 }, { "epoch": 15.020198310686743, "grad_norm": 0.47246578335762024, "learning_rate": 8.78776008045517e-06, "loss": 0.0126, "step": 40900 }, { "epoch": 15.023870730811606, "grad_norm": 0.6308493614196777, "learning_rate": 8.76904409163149e-06, "loss": 0.0136, "step": 40910 }, { "epoch": 15.027543150936467, "grad_norm": 0.3906067907810211, "learning_rate": 8.750346138628696e-06, "loss": 0.0114, "step": 40920 }, { "epoch": 15.03121557106133, "grad_norm": 0.45201608538627625, "learning_rate": 8.73166622962593e-06, "loss": 0.0137, "step": 40930 }, { "epoch": 15.034887991186192, "grad_norm": 0.4194927513599396, "learning_rate": 8.713004372794415e-06, "loss": 0.0122, "step": 40940 }, { "epoch": 15.038560411311053, "grad_norm": 0.22281323373317719, "learning_rate": 8.694360576297461e-06, "loss": 0.0155, "step": 40950 }, { "epoch": 15.042232831435916, "grad_norm": 0.2816654443740845, "learning_rate": 8.675734848290529e-06, "loss": 0.0137, "step": 40960 }, { "epoch": 15.045905251560779, "grad_norm": 0.3222253620624542, "learning_rate": 8.657127196921133e-06, "loss": 0.0131, "step": 40970 }, { "epoch": 15.049577671685642, "grad_norm": 0.48781388998031616, "learning_rate": 8.638537630328885e-06, "loss": 0.013, "step": 40980 }, { "epoch": 15.053250091810503, "grad_norm": 0.4643940031528473, "learning_rate": 8.619966156645487e-06, "loss": 0.0126, "step": 40990 }, { "epoch": 15.056922511935365, "grad_norm": 0.455735445022583, "learning_rate": 8.601412783994756e-06, "loss": 0.0125, "step": 41000 }, { "epoch": 15.060594932060228, "grad_norm": 0.720184326171875, "learning_rate": 8.582877520492538e-06, "loss": 0.0122, "step": 41010 }, { "epoch": 15.06426735218509, "grad_norm": 0.4399794638156891, "learning_rate": 8.564360374246821e-06, "loss": 0.013, "step": 41020 }, { "epoch": 15.067939772309952, "grad_norm": 0.5547994375228882, "learning_rate": 8.545861353357597e-06, "loss": 0.0123, "step": 41030 }, { "epoch": 15.071612192434815, "grad_norm": 0.4786963164806366, "learning_rate": 8.527380465917001e-06, "loss": 0.0113, "step": 41040 }, { "epoch": 15.075284612559678, "grad_norm": 0.41571858525276184, "learning_rate": 8.508917720009175e-06, "loss": 0.0123, "step": 41050 }, { "epoch": 15.078957032684539, "grad_norm": 0.18228942155838013, "learning_rate": 8.490473123710374e-06, "loss": 0.0132, "step": 41060 }, { "epoch": 15.082629452809401, "grad_norm": 0.2957160770893097, "learning_rate": 8.472046685088874e-06, "loss": 0.0137, "step": 41070 }, { "epoch": 15.086301872934264, "grad_norm": 0.6806185245513916, "learning_rate": 8.453638412205023e-06, "loss": 0.0126, "step": 41080 }, { "epoch": 15.089974293059125, "grad_norm": 0.5150038003921509, "learning_rate": 8.435248313111243e-06, "loss": 0.0122, "step": 41090 }, { "epoch": 15.093646713183988, "grad_norm": 0.40975069999694824, "learning_rate": 8.416876395851974e-06, "loss": 0.013, "step": 41100 }, { "epoch": 15.09731913330885, "grad_norm": 0.21555039286613464, "learning_rate": 8.39852266846371e-06, "loss": 0.0132, "step": 41110 }, { "epoch": 15.100991553433714, "grad_norm": 0.2784751355648041, "learning_rate": 8.380187138975021e-06, "loss": 0.0125, "step": 41120 }, { "epoch": 15.104663973558575, "grad_norm": 0.3403794765472412, "learning_rate": 8.36186981540647e-06, "loss": 0.0117, "step": 41130 }, { "epoch": 15.108336393683437, "grad_norm": 0.4864414930343628, "learning_rate": 8.343570705770681e-06, "loss": 0.0124, "step": 41140 }, { "epoch": 15.1120088138083, "grad_norm": 0.22339269518852234, "learning_rate": 8.325289818072296e-06, "loss": 0.0132, "step": 41150 }, { "epoch": 15.115681233933161, "grad_norm": 0.7445751428604126, "learning_rate": 8.307027160308018e-06, "loss": 0.0143, "step": 41160 }, { "epoch": 15.119353654058024, "grad_norm": 0.3841477036476135, "learning_rate": 8.288782740466538e-06, "loss": 0.0131, "step": 41170 }, { "epoch": 15.123026074182887, "grad_norm": 0.44478553533554077, "learning_rate": 8.270556566528581e-06, "loss": 0.0129, "step": 41180 }, { "epoch": 15.12669849430775, "grad_norm": 0.21269172430038452, "learning_rate": 8.252348646466918e-06, "loss": 0.0119, "step": 41190 }, { "epoch": 15.13037091443261, "grad_norm": 0.5228449106216431, "learning_rate": 8.234158988246294e-06, "loss": 0.0133, "step": 41200 }, { "epoch": 15.134043334557473, "grad_norm": 0.34729233384132385, "learning_rate": 8.215987599823476e-06, "loss": 0.0128, "step": 41210 }, { "epoch": 15.137715754682336, "grad_norm": 0.3725246489048004, "learning_rate": 8.19783448914727e-06, "loss": 0.0122, "step": 41220 }, { "epoch": 15.141388174807197, "grad_norm": 0.4989231526851654, "learning_rate": 8.179699664158452e-06, "loss": 0.0131, "step": 41230 }, { "epoch": 15.14506059493206, "grad_norm": 0.545247495174408, "learning_rate": 8.16158313278979e-06, "loss": 0.0121, "step": 41240 }, { "epoch": 15.148733015056923, "grad_norm": 0.2389480173587799, "learning_rate": 8.14348490296611e-06, "loss": 0.0127, "step": 41250 }, { "epoch": 15.152405435181786, "grad_norm": 0.4379352927207947, "learning_rate": 8.125404982604156e-06, "loss": 0.0122, "step": 41260 }, { "epoch": 15.156077855306647, "grad_norm": 0.49356913566589355, "learning_rate": 8.107343379612725e-06, "loss": 0.0118, "step": 41270 }, { "epoch": 15.15975027543151, "grad_norm": 0.49790826439857483, "learning_rate": 8.08930010189256e-06, "loss": 0.0126, "step": 41280 }, { "epoch": 15.163422695556372, "grad_norm": 0.2778591513633728, "learning_rate": 8.071275157336423e-06, "loss": 0.0114, "step": 41290 }, { "epoch": 15.167095115681233, "grad_norm": 0.9498281478881836, "learning_rate": 8.053268553829018e-06, "loss": 0.0125, "step": 41300 }, { "epoch": 15.170767535806096, "grad_norm": 0.5931099057197571, "learning_rate": 8.03528029924704e-06, "loss": 0.0124, "step": 41310 }, { "epoch": 15.174439955930959, "grad_norm": 0.28100138902664185, "learning_rate": 8.017310401459188e-06, "loss": 0.013, "step": 41320 }, { "epoch": 15.178112376055822, "grad_norm": 0.2683504819869995, "learning_rate": 7.999358868326095e-06, "loss": 0.0128, "step": 41330 }, { "epoch": 15.181784796180683, "grad_norm": 1.1404277086257935, "learning_rate": 7.981425707700352e-06, "loss": 0.0126, "step": 41340 }, { "epoch": 15.185457216305545, "grad_norm": 0.2273070067167282, "learning_rate": 7.96351092742656e-06, "loss": 0.0122, "step": 41350 }, { "epoch": 15.189129636430408, "grad_norm": 0.3487933278083801, "learning_rate": 7.945614535341246e-06, "loss": 0.0121, "step": 41360 }, { "epoch": 15.19280205655527, "grad_norm": 0.3557739555835724, "learning_rate": 7.92773653927288e-06, "loss": 0.0126, "step": 41370 }, { "epoch": 15.196474476680132, "grad_norm": 0.3370782434940338, "learning_rate": 7.909876947041934e-06, "loss": 0.013, "step": 41380 }, { "epoch": 15.200146896804995, "grad_norm": 0.3271489143371582, "learning_rate": 7.892035766460782e-06, "loss": 0.0141, "step": 41390 }, { "epoch": 15.203819316929856, "grad_norm": 0.7287963032722473, "learning_rate": 7.874213005333764e-06, "loss": 0.0126, "step": 41400 }, { "epoch": 15.207491737054719, "grad_norm": 0.6184098124504089, "learning_rate": 7.856408671457166e-06, "loss": 0.014, "step": 41410 }, { "epoch": 15.211164157179581, "grad_norm": 0.59895259141922, "learning_rate": 7.838622772619214e-06, "loss": 0.0129, "step": 41420 }, { "epoch": 15.214836577304444, "grad_norm": 0.20099912583827972, "learning_rate": 7.820855316600045e-06, "loss": 0.0141, "step": 41430 }, { "epoch": 15.218508997429305, "grad_norm": 0.33290335536003113, "learning_rate": 7.803106311171771e-06, "loss": 0.014, "step": 41440 }, { "epoch": 15.222181417554168, "grad_norm": 0.6037889122962952, "learning_rate": 7.785375764098401e-06, "loss": 0.0128, "step": 41450 }, { "epoch": 15.225853837679031, "grad_norm": 0.4383578896522522, "learning_rate": 7.767663683135862e-06, "loss": 0.0121, "step": 41460 }, { "epoch": 15.229526257803892, "grad_norm": 0.9814091920852661, "learning_rate": 7.749970076032049e-06, "loss": 0.0118, "step": 41470 }, { "epoch": 15.233198677928755, "grad_norm": 0.9090442061424255, "learning_rate": 7.73229495052673e-06, "loss": 0.0129, "step": 41480 }, { "epoch": 15.236871098053618, "grad_norm": 0.713481605052948, "learning_rate": 7.714638314351597e-06, "loss": 0.0129, "step": 41490 }, { "epoch": 15.24054351817848, "grad_norm": 0.23420444130897522, "learning_rate": 7.69700017523029e-06, "loss": 0.0124, "step": 41500 }, { "epoch": 15.244215938303341, "grad_norm": 0.8015528321266174, "learning_rate": 7.679380540878294e-06, "loss": 0.012, "step": 41510 }, { "epoch": 15.247888358428204, "grad_norm": 0.6186612248420715, "learning_rate": 7.661779419003073e-06, "loss": 0.0116, "step": 41520 }, { "epoch": 15.251560778553067, "grad_norm": 0.6316471099853516, "learning_rate": 7.644196817303928e-06, "loss": 0.0128, "step": 41530 }, { "epoch": 15.255233198677928, "grad_norm": 0.1883460432291031, "learning_rate": 7.626632743472106e-06, "loss": 0.0117, "step": 41540 }, { "epoch": 15.25890561880279, "grad_norm": 0.33955278992652893, "learning_rate": 7.609087205190724e-06, "loss": 0.0127, "step": 41550 }, { "epoch": 15.262578038927654, "grad_norm": 0.7664459943771362, "learning_rate": 7.591560210134774e-06, "loss": 0.0123, "step": 41560 }, { "epoch": 15.266250459052516, "grad_norm": 0.6617947220802307, "learning_rate": 7.574051765971196e-06, "loss": 0.0136, "step": 41570 }, { "epoch": 15.269922879177377, "grad_norm": 0.43158528208732605, "learning_rate": 7.556561880358754e-06, "loss": 0.0116, "step": 41580 }, { "epoch": 15.27359529930224, "grad_norm": 0.39456436038017273, "learning_rate": 7.5390905609481196e-06, "loss": 0.0122, "step": 41590 }, { "epoch": 15.277267719427103, "grad_norm": 0.31324470043182373, "learning_rate": 7.521637815381855e-06, "loss": 0.0121, "step": 41600 }, { "epoch": 15.280940139551964, "grad_norm": 0.18594348430633545, "learning_rate": 7.504203651294373e-06, "loss": 0.0117, "step": 41610 }, { "epoch": 15.284612559676827, "grad_norm": 0.5119040012359619, "learning_rate": 7.486788076311963e-06, "loss": 0.0143, "step": 41620 }, { "epoch": 15.28828497980169, "grad_norm": 0.41083478927612305, "learning_rate": 7.469391098052808e-06, "loss": 0.0118, "step": 41630 }, { "epoch": 15.291957399926552, "grad_norm": 0.46773701906204224, "learning_rate": 7.452012724126928e-06, "loss": 0.0139, "step": 41640 }, { "epoch": 15.295629820051413, "grad_norm": 0.5484057664871216, "learning_rate": 7.434652962136213e-06, "loss": 0.0129, "step": 41650 }, { "epoch": 15.299302240176276, "grad_norm": 0.43479496240615845, "learning_rate": 7.4173118196744e-06, "loss": 0.0126, "step": 41660 }, { "epoch": 15.302974660301139, "grad_norm": 0.1387973576784134, "learning_rate": 7.39998930432712e-06, "loss": 0.0129, "step": 41670 }, { "epoch": 15.306647080426, "grad_norm": 0.5543040633201599, "learning_rate": 7.38268542367182e-06, "loss": 0.013, "step": 41680 }, { "epoch": 15.310319500550863, "grad_norm": 0.4046419858932495, "learning_rate": 7.365400185277782e-06, "loss": 0.0143, "step": 41690 }, { "epoch": 15.313991920675726, "grad_norm": 0.32718250155448914, "learning_rate": 7.348133596706197e-06, "loss": 0.0114, "step": 41700 }, { "epoch": 15.317664340800588, "grad_norm": 0.27791687846183777, "learning_rate": 7.3308856655100286e-06, "loss": 0.0125, "step": 41710 }, { "epoch": 15.32133676092545, "grad_norm": 0.47859928011894226, "learning_rate": 7.313656399234109e-06, "loss": 0.0129, "step": 41720 }, { "epoch": 15.325009181050312, "grad_norm": 0.22650903463363647, "learning_rate": 7.296445805415125e-06, "loss": 0.0138, "step": 41730 }, { "epoch": 15.328681601175175, "grad_norm": 0.554569661617279, "learning_rate": 7.2792538915815504e-06, "loss": 0.0124, "step": 41740 }, { "epoch": 15.332354021300036, "grad_norm": 0.4875098168849945, "learning_rate": 7.262080665253729e-06, "loss": 0.0113, "step": 41750 }, { "epoch": 15.336026441424899, "grad_norm": 0.38607800006866455, "learning_rate": 7.244926133943803e-06, "loss": 0.0118, "step": 41760 }, { "epoch": 15.339698861549762, "grad_norm": 0.29802456498146057, "learning_rate": 7.227790305155763e-06, "loss": 0.0132, "step": 41770 }, { "epoch": 15.343371281674624, "grad_norm": 0.4519979953765869, "learning_rate": 7.210673186385381e-06, "loss": 0.013, "step": 41780 }, { "epoch": 15.347043701799485, "grad_norm": 0.40514394640922546, "learning_rate": 7.193574785120288e-06, "loss": 0.0128, "step": 41790 }, { "epoch": 15.350716121924348, "grad_norm": 0.19093473255634308, "learning_rate": 7.176495108839893e-06, "loss": 0.0125, "step": 41800 }, { "epoch": 15.354388542049211, "grad_norm": 0.3769567906856537, "learning_rate": 7.159434165015417e-06, "loss": 0.0125, "step": 41810 }, { "epoch": 15.358060962174072, "grad_norm": 0.2893868386745453, "learning_rate": 7.142391961109923e-06, "loss": 0.0122, "step": 41820 }, { "epoch": 15.361733382298935, "grad_norm": 0.3813830614089966, "learning_rate": 7.1253685045782335e-06, "loss": 0.0134, "step": 41830 }, { "epoch": 15.365405802423798, "grad_norm": 0.5802186727523804, "learning_rate": 7.108363802866991e-06, "loss": 0.013, "step": 41840 }, { "epoch": 15.36907822254866, "grad_norm": 0.3538472056388855, "learning_rate": 7.091377863414611e-06, "loss": 0.0112, "step": 41850 }, { "epoch": 15.372750642673521, "grad_norm": 0.4832002818584442, "learning_rate": 7.074410693651351e-06, "loss": 0.0118, "step": 41860 }, { "epoch": 15.376423062798384, "grad_norm": 0.22720593214035034, "learning_rate": 7.057462300999212e-06, "loss": 0.0163, "step": 41870 }, { "epoch": 15.380095482923247, "grad_norm": 0.28376200795173645, "learning_rate": 7.0405326928719876e-06, "loss": 0.0126, "step": 41880 }, { "epoch": 15.383767903048108, "grad_norm": 0.5085828304290771, "learning_rate": 7.0236218766752844e-06, "loss": 0.012, "step": 41890 }, { "epoch": 15.38744032317297, "grad_norm": 0.2924444377422333, "learning_rate": 7.006729859806454e-06, "loss": 0.0139, "step": 41900 }, { "epoch": 15.391112743297834, "grad_norm": 0.22077758610248566, "learning_rate": 6.989856649654636e-06, "loss": 0.0125, "step": 41910 }, { "epoch": 15.394785163422696, "grad_norm": 0.38548386096954346, "learning_rate": 6.973002253600764e-06, "loss": 0.0138, "step": 41920 }, { "epoch": 15.398457583547557, "grad_norm": 0.251054048538208, "learning_rate": 6.956166679017506e-06, "loss": 0.0132, "step": 41930 }, { "epoch": 15.40213000367242, "grad_norm": 0.33113107085227966, "learning_rate": 6.939349933269312e-06, "loss": 0.012, "step": 41940 }, { "epoch": 15.405802423797283, "grad_norm": 0.22787286341190338, "learning_rate": 6.9225520237124146e-06, "loss": 0.0129, "step": 41950 }, { "epoch": 15.409474843922144, "grad_norm": 0.6346112489700317, "learning_rate": 6.905772957694784e-06, "loss": 0.0131, "step": 41960 }, { "epoch": 15.413147264047007, "grad_norm": 0.2729949653148651, "learning_rate": 6.889012742556133e-06, "loss": 0.0126, "step": 41970 }, { "epoch": 15.41681968417187, "grad_norm": 0.513242244720459, "learning_rate": 6.872271385627982e-06, "loss": 0.0126, "step": 41980 }, { "epoch": 15.42049210429673, "grad_norm": 1.0823336839675903, "learning_rate": 6.8555488942335435e-06, "loss": 0.0125, "step": 41990 }, { "epoch": 15.424164524421593, "grad_norm": 0.4270120859146118, "learning_rate": 6.838845275687822e-06, "loss": 0.0129, "step": 42000 }, { "epoch": 15.427836944546456, "grad_norm": 0.42236271500587463, "learning_rate": 6.822160537297534e-06, "loss": 0.0133, "step": 42010 }, { "epoch": 15.431509364671319, "grad_norm": 0.33801838755607605, "learning_rate": 6.805494686361164e-06, "loss": 0.0122, "step": 42020 }, { "epoch": 15.43518178479618, "grad_norm": 0.31863877177238464, "learning_rate": 6.788847730168912e-06, "loss": 0.0134, "step": 42030 }, { "epoch": 15.438854204921043, "grad_norm": 0.6518105864524841, "learning_rate": 6.772219676002716e-06, "loss": 0.0126, "step": 42040 }, { "epoch": 15.442526625045906, "grad_norm": 0.6589422225952148, "learning_rate": 6.755610531136269e-06, "loss": 0.0113, "step": 42050 }, { "epoch": 15.446199045170767, "grad_norm": 0.2629208266735077, "learning_rate": 6.73902030283497e-06, "loss": 0.0124, "step": 42060 }, { "epoch": 15.44987146529563, "grad_norm": 0.41090428829193115, "learning_rate": 6.7224489983559295e-06, "loss": 0.0146, "step": 42070 }, { "epoch": 15.453543885420492, "grad_norm": 0.4214426875114441, "learning_rate": 6.705896624948027e-06, "loss": 0.0119, "step": 42080 }, { "epoch": 15.457216305545355, "grad_norm": 0.26380905508995056, "learning_rate": 6.689363189851816e-06, "loss": 0.0114, "step": 42090 }, { "epoch": 15.460888725670216, "grad_norm": 0.7930306196212769, "learning_rate": 6.672848700299583e-06, "loss": 0.0126, "step": 42100 }, { "epoch": 15.464561145795079, "grad_norm": 0.21166999638080597, "learning_rate": 6.656353163515339e-06, "loss": 0.0112, "step": 42110 }, { "epoch": 15.468233565919942, "grad_norm": 0.556329071521759, "learning_rate": 6.6398765867147824e-06, "loss": 0.0136, "step": 42120 }, { "epoch": 15.471905986044803, "grad_norm": 0.3007883131504059, "learning_rate": 6.623418977105322e-06, "loss": 0.0112, "step": 42130 }, { "epoch": 15.475578406169666, "grad_norm": 0.36618751287460327, "learning_rate": 6.606980341886087e-06, "loss": 0.0122, "step": 42140 }, { "epoch": 15.479250826294528, "grad_norm": 0.22935909032821655, "learning_rate": 6.590560688247898e-06, "loss": 0.0138, "step": 42150 }, { "epoch": 15.482923246419391, "grad_norm": 0.36263057589530945, "learning_rate": 6.574160023373249e-06, "loss": 0.0117, "step": 42160 }, { "epoch": 15.486595666544252, "grad_norm": 0.31276625394821167, "learning_rate": 6.557778354436367e-06, "loss": 0.0128, "step": 42170 }, { "epoch": 15.490268086669115, "grad_norm": 0.23300305008888245, "learning_rate": 6.5414156886031475e-06, "loss": 0.0119, "step": 42180 }, { "epoch": 15.493940506793978, "grad_norm": 0.38510820269584656, "learning_rate": 6.525072033031177e-06, "loss": 0.0119, "step": 42190 }, { "epoch": 15.497612926918839, "grad_norm": 0.6302034258842468, "learning_rate": 6.508747394869708e-06, "loss": 0.0123, "step": 42200 }, { "epoch": 15.501285347043702, "grad_norm": 0.3784047067165375, "learning_rate": 6.492441781259717e-06, "loss": 0.0124, "step": 42210 }, { "epoch": 15.504957767168564, "grad_norm": 0.6220857501029968, "learning_rate": 6.476155199333811e-06, "loss": 0.0114, "step": 42220 }, { "epoch": 15.508630187293427, "grad_norm": 0.313406378030777, "learning_rate": 6.459887656216318e-06, "loss": 0.0102, "step": 42230 }, { "epoch": 15.512302607418288, "grad_norm": 0.4719286859035492, "learning_rate": 6.443639159023191e-06, "loss": 0.0124, "step": 42240 }, { "epoch": 15.515975027543151, "grad_norm": 0.211216002702713, "learning_rate": 6.427409714862092e-06, "loss": 0.0131, "step": 42250 }, { "epoch": 15.519647447668014, "grad_norm": 0.6083997488021851, "learning_rate": 6.411199330832318e-06, "loss": 0.0134, "step": 42260 }, { "epoch": 15.523319867792875, "grad_norm": 0.3440993130207062, "learning_rate": 6.3950080140248556e-06, "loss": 0.0115, "step": 42270 }, { "epoch": 15.526992287917738, "grad_norm": 0.6272560358047485, "learning_rate": 6.378835771522335e-06, "loss": 0.0124, "step": 42280 }, { "epoch": 15.5306647080426, "grad_norm": 0.24399635195732117, "learning_rate": 6.362682610399023e-06, "loss": 0.0126, "step": 42290 }, { "epoch": 15.534337128167463, "grad_norm": 0.2835855782032013, "learning_rate": 6.346548537720886e-06, "loss": 0.012, "step": 42300 }, { "epoch": 15.538009548292324, "grad_norm": 0.47891202569007874, "learning_rate": 6.330433560545512e-06, "loss": 0.013, "step": 42310 }, { "epoch": 15.541681968417187, "grad_norm": 0.47730693221092224, "learning_rate": 6.3143376859221225e-06, "loss": 0.012, "step": 42320 }, { "epoch": 15.54535438854205, "grad_norm": 0.5078564286231995, "learning_rate": 6.298260920891624e-06, "loss": 0.0117, "step": 42330 }, { "epoch": 15.54902680866691, "grad_norm": 0.30232125520706177, "learning_rate": 6.282203272486525e-06, "loss": 0.0128, "step": 42340 }, { "epoch": 15.552699228791774, "grad_norm": 0.4072986841201782, "learning_rate": 6.266164747730985e-06, "loss": 0.0121, "step": 42350 }, { "epoch": 15.556371648916636, "grad_norm": 0.382743775844574, "learning_rate": 6.250145353640796e-06, "loss": 0.013, "step": 42360 }, { "epoch": 15.5600440690415, "grad_norm": 0.4873083233833313, "learning_rate": 6.234145097223404e-06, "loss": 0.0111, "step": 42370 }, { "epoch": 15.56371648916636, "grad_norm": 0.2457607239484787, "learning_rate": 6.218163985477854e-06, "loss": 0.0132, "step": 42380 }, { "epoch": 15.567388909291223, "grad_norm": 0.3998853266239166, "learning_rate": 6.202202025394815e-06, "loss": 0.0125, "step": 42390 }, { "epoch": 15.571061329416086, "grad_norm": 0.3438202142715454, "learning_rate": 6.186259223956614e-06, "loss": 0.0117, "step": 42400 }, { "epoch": 15.574733749540947, "grad_norm": 0.3501660227775574, "learning_rate": 6.170335588137161e-06, "loss": 0.0108, "step": 42410 }, { "epoch": 15.57840616966581, "grad_norm": 0.38015246391296387, "learning_rate": 6.154431124901982e-06, "loss": 0.0123, "step": 42420 }, { "epoch": 15.582078589790672, "grad_norm": 0.5692043304443359, "learning_rate": 6.1385458412082595e-06, "loss": 0.012, "step": 42430 }, { "epoch": 15.585751009915533, "grad_norm": 0.2597409784793854, "learning_rate": 6.122679744004739e-06, "loss": 0.0138, "step": 42440 }, { "epoch": 15.589423430040396, "grad_norm": 0.5192714333534241, "learning_rate": 6.106832840231786e-06, "loss": 0.0118, "step": 42450 }, { "epoch": 15.593095850165259, "grad_norm": 0.42487478256225586, "learning_rate": 6.091005136821387e-06, "loss": 0.0134, "step": 42460 }, { "epoch": 15.596768270290122, "grad_norm": 0.7663887739181519, "learning_rate": 6.075196640697101e-06, "loss": 0.0119, "step": 42470 }, { "epoch": 15.600440690414983, "grad_norm": 0.29738420248031616, "learning_rate": 6.059407358774122e-06, "loss": 0.0133, "step": 42480 }, { "epoch": 15.604113110539846, "grad_norm": 0.37382999062538147, "learning_rate": 6.043637297959198e-06, "loss": 0.0127, "step": 42490 }, { "epoch": 15.607785530664708, "grad_norm": 0.37527212500572205, "learning_rate": 6.0278864651507065e-06, "loss": 0.0115, "step": 42500 }, { "epoch": 15.611457950789571, "grad_norm": 0.5314360857009888, "learning_rate": 6.012154867238584e-06, "loss": 0.0112, "step": 42510 }, { "epoch": 15.615130370914432, "grad_norm": 0.22723910212516785, "learning_rate": 5.996442511104378e-06, "loss": 0.0114, "step": 42520 }, { "epoch": 15.618802791039295, "grad_norm": 0.480822890996933, "learning_rate": 5.9807494036211995e-06, "loss": 0.0129, "step": 42530 }, { "epoch": 15.622475211164158, "grad_norm": 0.776154100894928, "learning_rate": 5.965075551653748e-06, "loss": 0.0112, "step": 42540 }, { "epoch": 15.626147631289019, "grad_norm": 0.24543921649456024, "learning_rate": 5.949420962058294e-06, "loss": 0.0128, "step": 42550 }, { "epoch": 15.629820051413882, "grad_norm": 0.6737300753593445, "learning_rate": 5.933785641682693e-06, "loss": 0.0112, "step": 42560 }, { "epoch": 15.633492471538744, "grad_norm": 0.40863800048828125, "learning_rate": 5.918169597366369e-06, "loss": 0.0119, "step": 42570 }, { "epoch": 15.637164891663605, "grad_norm": 0.6501052975654602, "learning_rate": 5.902572835940295e-06, "loss": 0.0128, "step": 42580 }, { "epoch": 15.640837311788468, "grad_norm": 0.3952637314796448, "learning_rate": 5.886995364227044e-06, "loss": 0.0136, "step": 42590 }, { "epoch": 15.644509731913331, "grad_norm": 0.46428972482681274, "learning_rate": 5.871437189040718e-06, "loss": 0.0114, "step": 42600 }, { "epoch": 15.648182152038194, "grad_norm": 0.36749401688575745, "learning_rate": 5.855898317186992e-06, "loss": 0.0146, "step": 42610 }, { "epoch": 15.651854572163055, "grad_norm": 0.24360784888267517, "learning_rate": 5.840378755463105e-06, "loss": 0.0127, "step": 42620 }, { "epoch": 15.655526992287918, "grad_norm": 0.26326048374176025, "learning_rate": 5.824878510657833e-06, "loss": 0.0125, "step": 42630 }, { "epoch": 15.65919941241278, "grad_norm": 0.3889860212802887, "learning_rate": 5.809397589551502e-06, "loss": 0.0138, "step": 42640 }, { "epoch": 15.662871832537641, "grad_norm": 0.7260791659355164, "learning_rate": 5.793935998916017e-06, "loss": 0.0124, "step": 42650 }, { "epoch": 15.666544252662504, "grad_norm": 0.33343738317489624, "learning_rate": 5.778493745514785e-06, "loss": 0.0112, "step": 42660 }, { "epoch": 15.670216672787367, "grad_norm": 0.30944013595581055, "learning_rate": 5.763070836102768e-06, "loss": 0.0119, "step": 42670 }, { "epoch": 15.67388909291223, "grad_norm": 0.9446799755096436, "learning_rate": 5.7476672774264885e-06, "loss": 0.0131, "step": 42680 }, { "epoch": 15.677561513037091, "grad_norm": 0.35107988119125366, "learning_rate": 5.732283076223977e-06, "loss": 0.012, "step": 42690 }, { "epoch": 15.681233933161954, "grad_norm": 0.27760183811187744, "learning_rate": 5.716918239224794e-06, "loss": 0.0136, "step": 42700 }, { "epoch": 15.684906353286816, "grad_norm": 0.7873547673225403, "learning_rate": 5.70157277315006e-06, "loss": 0.0115, "step": 42710 }, { "epoch": 15.688578773411678, "grad_norm": 0.22844810783863068, "learning_rate": 5.686246684712387e-06, "loss": 0.0141, "step": 42720 }, { "epoch": 15.69225119353654, "grad_norm": 0.28162941336631775, "learning_rate": 5.670939980615947e-06, "loss": 0.0112, "step": 42730 }, { "epoch": 15.695923613661403, "grad_norm": 0.49828556180000305, "learning_rate": 5.655652667556388e-06, "loss": 0.0122, "step": 42740 }, { "epoch": 15.699596033786266, "grad_norm": 0.29718017578125, "learning_rate": 5.640384752220928e-06, "loss": 0.0124, "step": 42750 }, { "epoch": 15.703268453911127, "grad_norm": 0.2749581038951874, "learning_rate": 5.625136241288253e-06, "loss": 0.0113, "step": 42760 }, { "epoch": 15.70694087403599, "grad_norm": 0.5608986616134644, "learning_rate": 5.609907141428572e-06, "loss": 0.0119, "step": 42770 }, { "epoch": 15.710613294160853, "grad_norm": 0.5183524489402771, "learning_rate": 5.594697459303633e-06, "loss": 0.0128, "step": 42780 }, { "epoch": 15.714285714285714, "grad_norm": 0.2383880615234375, "learning_rate": 5.579507201566658e-06, "loss": 0.0133, "step": 42790 }, { "epoch": 15.717958134410576, "grad_norm": 0.4254871606826782, "learning_rate": 5.564336374862372e-06, "loss": 0.012, "step": 42800 }, { "epoch": 15.72163055453544, "grad_norm": 0.3366397023200989, "learning_rate": 5.549184985827027e-06, "loss": 0.012, "step": 42810 }, { "epoch": 15.725302974660302, "grad_norm": 0.6127275824546814, "learning_rate": 5.534053041088349e-06, "loss": 0.0123, "step": 42820 }, { "epoch": 15.728975394785163, "grad_norm": 0.3889101445674896, "learning_rate": 5.5189405472655495e-06, "loss": 0.0124, "step": 42830 }, { "epoch": 15.732647814910026, "grad_norm": 0.21839603781700134, "learning_rate": 5.503847510969374e-06, "loss": 0.0133, "step": 42840 }, { "epoch": 15.736320235034889, "grad_norm": 1.0763449668884277, "learning_rate": 5.488773938802011e-06, "loss": 0.0122, "step": 42850 }, { "epoch": 15.73999265515975, "grad_norm": 0.3978995680809021, "learning_rate": 5.473719837357155e-06, "loss": 0.0123, "step": 42860 }, { "epoch": 15.743665075284612, "grad_norm": 0.3353121876716614, "learning_rate": 5.4586852132199825e-06, "loss": 0.0119, "step": 42870 }, { "epoch": 15.747337495409475, "grad_norm": 0.23313014209270477, "learning_rate": 5.443670072967155e-06, "loss": 0.0118, "step": 42880 }, { "epoch": 15.751009915534338, "grad_norm": 0.9431841373443604, "learning_rate": 5.428674423166796e-06, "loss": 0.0123, "step": 42890 }, { "epoch": 15.754682335659199, "grad_norm": 0.3425047695636749, "learning_rate": 5.413698270378515e-06, "loss": 0.0119, "step": 42900 }, { "epoch": 15.758354755784062, "grad_norm": 0.5224376320838928, "learning_rate": 5.398741621153397e-06, "loss": 0.0127, "step": 42910 }, { "epoch": 15.762027175908925, "grad_norm": 0.3850327432155609, "learning_rate": 5.383804482033983e-06, "loss": 0.0133, "step": 42920 }, { "epoch": 15.765699596033786, "grad_norm": 0.20191818475723267, "learning_rate": 5.368886859554273e-06, "loss": 0.0113, "step": 42930 }, { "epoch": 15.769372016158648, "grad_norm": 0.26564887166023254, "learning_rate": 5.353988760239764e-06, "loss": 0.0143, "step": 42940 }, { "epoch": 15.773044436283511, "grad_norm": 0.4575659930706024, "learning_rate": 5.339110190607377e-06, "loss": 0.0126, "step": 42950 }, { "epoch": 15.776716856408374, "grad_norm": 0.5545980930328369, "learning_rate": 5.324251157165511e-06, "loss": 0.0118, "step": 42960 }, { "epoch": 15.780389276533235, "grad_norm": 0.5295005440711975, "learning_rate": 5.309411666413999e-06, "loss": 0.0123, "step": 42970 }, { "epoch": 15.784061696658098, "grad_norm": 0.324951171875, "learning_rate": 5.294591724844161e-06, "loss": 0.0116, "step": 42980 }, { "epoch": 15.78773411678296, "grad_norm": 1.0378642082214355, "learning_rate": 5.279791338938716e-06, "loss": 0.0108, "step": 42990 }, { "epoch": 15.791406536907822, "grad_norm": 0.1766442060470581, "learning_rate": 5.265010515171886e-06, "loss": 0.0125, "step": 43000 }, { "epoch": 15.795078957032684, "grad_norm": 0.2825714349746704, "learning_rate": 5.250249260009282e-06, "loss": 0.0125, "step": 43010 }, { "epoch": 15.798751377157547, "grad_norm": 0.5986965298652649, "learning_rate": 5.2355075799079835e-06, "loss": 0.0121, "step": 43020 }, { "epoch": 15.802423797282408, "grad_norm": 0.6424899101257324, "learning_rate": 5.220785481316509e-06, "loss": 0.0114, "step": 43030 }, { "epoch": 15.806096217407271, "grad_norm": 0.29990607500076294, "learning_rate": 5.206082970674797e-06, "loss": 0.0128, "step": 43040 }, { "epoch": 15.809768637532134, "grad_norm": 0.4210345447063446, "learning_rate": 5.191400054414231e-06, "loss": 0.0115, "step": 43050 }, { "epoch": 15.813441057656997, "grad_norm": 1.0016779899597168, "learning_rate": 5.176736738957605e-06, "loss": 0.0127, "step": 43060 }, { "epoch": 15.817113477781858, "grad_norm": 0.4781007766723633, "learning_rate": 5.162093030719173e-06, "loss": 0.0112, "step": 43070 }, { "epoch": 15.82078589790672, "grad_norm": 0.3607788681983948, "learning_rate": 5.1474689361045715e-06, "loss": 0.0119, "step": 43080 }, { "epoch": 15.824458318031583, "grad_norm": 0.3095433712005615, "learning_rate": 5.132864461510878e-06, "loss": 0.0128, "step": 43090 }, { "epoch": 15.828130738156446, "grad_norm": 0.9192935824394226, "learning_rate": 5.118279613326604e-06, "loss": 0.0119, "step": 43100 }, { "epoch": 15.831803158281307, "grad_norm": 0.5190499424934387, "learning_rate": 5.103714397931636e-06, "loss": 0.0124, "step": 43110 }, { "epoch": 15.83547557840617, "grad_norm": 0.7711858153343201, "learning_rate": 5.0891688216972985e-06, "loss": 0.013, "step": 43120 }, { "epoch": 15.839147998531033, "grad_norm": 0.651675283908844, "learning_rate": 5.074642890986336e-06, "loss": 0.0123, "step": 43130 }, { "epoch": 15.842820418655894, "grad_norm": 0.4521139860153198, "learning_rate": 5.060136612152871e-06, "loss": 0.0112, "step": 43140 }, { "epoch": 15.846492838780756, "grad_norm": 0.6170340180397034, "learning_rate": 5.045649991542434e-06, "loss": 0.013, "step": 43150 }, { "epoch": 15.85016525890562, "grad_norm": 0.42955324053764343, "learning_rate": 5.031183035491982e-06, "loss": 0.0121, "step": 43160 }, { "epoch": 15.85383767903048, "grad_norm": 0.526716411113739, "learning_rate": 5.016735750329854e-06, "loss": 0.0123, "step": 43170 }, { "epoch": 15.857510099155343, "grad_norm": 0.5597043037414551, "learning_rate": 5.002308142375761e-06, "loss": 0.0119, "step": 43180 }, { "epoch": 15.861182519280206, "grad_norm": 0.6054921746253967, "learning_rate": 4.987900217940855e-06, "loss": 0.0116, "step": 43190 }, { "epoch": 15.864854939405069, "grad_norm": 1.106483817100525, "learning_rate": 4.973511983327633e-06, "loss": 0.0125, "step": 43200 }, { "epoch": 15.86852735952993, "grad_norm": 0.53074049949646, "learning_rate": 4.959143444830011e-06, "loss": 0.0114, "step": 43210 }, { "epoch": 15.872199779654792, "grad_norm": 0.4090287685394287, "learning_rate": 4.944794608733267e-06, "loss": 0.0145, "step": 43220 }, { "epoch": 15.875872199779655, "grad_norm": 0.4503961205482483, "learning_rate": 4.930465481314078e-06, "loss": 0.0125, "step": 43230 }, { "epoch": 15.879544619904516, "grad_norm": 0.4629295766353607, "learning_rate": 4.916156068840483e-06, "loss": 0.0116, "step": 43240 }, { "epoch": 15.883217040029379, "grad_norm": 0.2266702651977539, "learning_rate": 4.901866377571901e-06, "loss": 0.0137, "step": 43250 }, { "epoch": 15.886889460154242, "grad_norm": 0.30713847279548645, "learning_rate": 4.887596413759143e-06, "loss": 0.0111, "step": 43260 }, { "epoch": 15.890561880279105, "grad_norm": 0.4764055013656616, "learning_rate": 4.873346183644373e-06, "loss": 0.0117, "step": 43270 }, { "epoch": 15.894234300403966, "grad_norm": 0.6510208249092102, "learning_rate": 4.8591156934611085e-06, "loss": 0.0118, "step": 43280 }, { "epoch": 15.897906720528828, "grad_norm": 0.20087336003780365, "learning_rate": 4.8449049494342704e-06, "loss": 0.0116, "step": 43290 }, { "epoch": 15.901579140653691, "grad_norm": 0.4276371896266937, "learning_rate": 4.830713957780114e-06, "loss": 0.0131, "step": 43300 }, { "epoch": 15.905251560778552, "grad_norm": 0.3770741820335388, "learning_rate": 4.816542724706253e-06, "loss": 0.0121, "step": 43310 }, { "epoch": 15.908923980903415, "grad_norm": 0.3781507909297943, "learning_rate": 4.802391256411675e-06, "loss": 0.0117, "step": 43320 }, { "epoch": 15.912596401028278, "grad_norm": 0.34452691674232483, "learning_rate": 4.788259559086717e-06, "loss": 0.012, "step": 43330 }, { "epoch": 15.91626882115314, "grad_norm": 0.3128071129322052, "learning_rate": 4.774147638913046e-06, "loss": 0.0117, "step": 43340 }, { "epoch": 15.919941241278002, "grad_norm": 0.7655140161514282, "learning_rate": 4.760055502063721e-06, "loss": 0.012, "step": 43350 }, { "epoch": 15.923613661402864, "grad_norm": 0.2808550000190735, "learning_rate": 4.745983154703099e-06, "loss": 0.0122, "step": 43360 }, { "epoch": 15.927286081527727, "grad_norm": 0.3975486755371094, "learning_rate": 4.731930602986906e-06, "loss": 0.0112, "step": 43370 }, { "epoch": 15.930958501652588, "grad_norm": 0.43511050939559937, "learning_rate": 4.7178978530622196e-06, "loss": 0.0118, "step": 43380 }, { "epoch": 15.934630921777451, "grad_norm": 0.2594851851463318, "learning_rate": 4.70388491106743e-06, "loss": 0.0133, "step": 43390 }, { "epoch": 15.938303341902314, "grad_norm": 0.30430713295936584, "learning_rate": 4.689891783132278e-06, "loss": 0.0122, "step": 43400 }, { "epoch": 15.941975762027177, "grad_norm": 0.21705839037895203, "learning_rate": 4.675918475377822e-06, "loss": 0.0116, "step": 43410 }, { "epoch": 15.945648182152038, "grad_norm": 0.4011084735393524, "learning_rate": 4.6619649939164845e-06, "loss": 0.0122, "step": 43420 }, { "epoch": 15.9493206022769, "grad_norm": 0.23547744750976562, "learning_rate": 4.6480313448519665e-06, "loss": 0.0121, "step": 43430 }, { "epoch": 15.952993022401763, "grad_norm": 0.2511739432811737, "learning_rate": 4.634117534279337e-06, "loss": 0.0124, "step": 43440 }, { "epoch": 15.956665442526624, "grad_norm": 0.42172762751579285, "learning_rate": 4.620223568284959e-06, "loss": 0.0126, "step": 43450 }, { "epoch": 15.960337862651487, "grad_norm": 0.2782103717327118, "learning_rate": 4.60634945294654e-06, "loss": 0.0125, "step": 43460 }, { "epoch": 15.96401028277635, "grad_norm": 0.6994455456733704, "learning_rate": 4.592495194333074e-06, "loss": 0.0125, "step": 43470 }, { "epoch": 15.967682702901213, "grad_norm": 0.31718477606773376, "learning_rate": 4.578660798504897e-06, "loss": 0.0115, "step": 43480 }, { "epoch": 15.971355123026074, "grad_norm": 0.30019426345825195, "learning_rate": 4.5648462715136346e-06, "loss": 0.0142, "step": 43490 }, { "epoch": 15.975027543150937, "grad_norm": 0.4694027304649353, "learning_rate": 4.551051619402225e-06, "loss": 0.0116, "step": 43500 }, { "epoch": 15.9786999632758, "grad_norm": 0.4341231882572174, "learning_rate": 4.537276848204936e-06, "loss": 0.0115, "step": 43510 }, { "epoch": 15.98237238340066, "grad_norm": 0.2620559632778168, "learning_rate": 4.523521963947308e-06, "loss": 0.0122, "step": 43520 }, { "epoch": 15.986044803525523, "grad_norm": 0.5312332510948181, "learning_rate": 4.509786972646191e-06, "loss": 0.0122, "step": 43530 }, { "epoch": 15.989717223650386, "grad_norm": 0.5015466213226318, "learning_rate": 4.496071880309749e-06, "loss": 0.012, "step": 43540 }, { "epoch": 15.993389643775249, "grad_norm": 0.5018495917320251, "learning_rate": 4.4823766929374234e-06, "loss": 0.0136, "step": 43550 }, { "epoch": 15.99706206390011, "grad_norm": 1.1066985130310059, "learning_rate": 4.468701416519955e-06, "loss": 0.0119, "step": 43560 }, { "epoch": 16.00073448402497, "grad_norm": 0.22754031419754028, "learning_rate": 4.4550460570393635e-06, "loss": 0.011, "step": 43570 }, { "epoch": 16.004406904149835, "grad_norm": 0.4297010004520416, "learning_rate": 4.441410620468989e-06, "loss": 0.0117, "step": 43580 }, { "epoch": 16.008079324274696, "grad_norm": 0.4575624167919159, "learning_rate": 4.427795112773419e-06, "loss": 0.0118, "step": 43590 }, { "epoch": 16.01175174439956, "grad_norm": 0.6750297546386719, "learning_rate": 4.414199539908531e-06, "loss": 0.0121, "step": 43600 }, { "epoch": 16.015424164524422, "grad_norm": 0.3915020227432251, "learning_rate": 4.400623907821516e-06, "loss": 0.0122, "step": 43610 }, { "epoch": 16.019096584649283, "grad_norm": 0.2755237817764282, "learning_rate": 4.387068222450802e-06, "loss": 0.0124, "step": 43620 }, { "epoch": 16.022769004774148, "grad_norm": 0.4317629933357239, "learning_rate": 4.3735324897260954e-06, "loss": 0.0123, "step": 43630 }, { "epoch": 16.02644142489901, "grad_norm": 0.3794620931148529, "learning_rate": 4.36001671556841e-06, "loss": 0.0114, "step": 43640 }, { "epoch": 16.03011384502387, "grad_norm": 0.6561622619628906, "learning_rate": 4.346520905889989e-06, "loss": 0.0127, "step": 43650 }, { "epoch": 16.033786265148734, "grad_norm": 0.31085124611854553, "learning_rate": 4.333045066594354e-06, "loss": 0.0117, "step": 43660 }, { "epoch": 16.037458685273595, "grad_norm": 0.36253803968429565, "learning_rate": 4.319589203576313e-06, "loss": 0.0123, "step": 43670 }, { "epoch": 16.041131105398456, "grad_norm": 0.3097110986709595, "learning_rate": 4.306153322721895e-06, "loss": 0.0129, "step": 43680 }, { "epoch": 16.04480352552332, "grad_norm": 0.2577579617500305, "learning_rate": 4.292737429908422e-06, "loss": 0.0125, "step": 43690 }, { "epoch": 16.048475945648182, "grad_norm": 0.8095298409461975, "learning_rate": 4.279341531004471e-06, "loss": 0.0129, "step": 43700 }, { "epoch": 16.052148365773043, "grad_norm": 0.6441011428833008, "learning_rate": 4.26596563186985e-06, "loss": 0.0123, "step": 43710 }, { "epoch": 16.055820785897907, "grad_norm": 0.33668097853660583, "learning_rate": 4.252609738355634e-06, "loss": 0.0126, "step": 43720 }, { "epoch": 16.05949320602277, "grad_norm": 0.5831376910209656, "learning_rate": 4.239273856304149e-06, "loss": 0.013, "step": 43730 }, { "epoch": 16.063165626147633, "grad_norm": 0.6246250867843628, "learning_rate": 4.2259579915489625e-06, "loss": 0.0121, "step": 43740 }, { "epoch": 16.066838046272494, "grad_norm": 0.6513746976852417, "learning_rate": 4.212662149914886e-06, "loss": 0.012, "step": 43750 }, { "epoch": 16.070510466397355, "grad_norm": 0.39878642559051514, "learning_rate": 4.199386337217958e-06, "loss": 0.0115, "step": 43760 }, { "epoch": 16.07418288652222, "grad_norm": 0.26175206899642944, "learning_rate": 4.186130559265494e-06, "loss": 0.012, "step": 43770 }, { "epoch": 16.07785530664708, "grad_norm": 0.3586254119873047, "learning_rate": 4.172894821856011e-06, "loss": 0.0116, "step": 43780 }, { "epoch": 16.08152772677194, "grad_norm": 0.7389575242996216, "learning_rate": 4.159679130779254e-06, "loss": 0.0113, "step": 43790 }, { "epoch": 16.085200146896806, "grad_norm": 0.9141965508460999, "learning_rate": 4.1464834918162445e-06, "loss": 0.0117, "step": 43800 }, { "epoch": 16.088872567021667, "grad_norm": 0.30677732825279236, "learning_rate": 4.133307910739187e-06, "loss": 0.012, "step": 43810 }, { "epoch": 16.09254498714653, "grad_norm": 0.9344452023506165, "learning_rate": 4.120152393311521e-06, "loss": 0.0129, "step": 43820 }, { "epoch": 16.096217407271393, "grad_norm": 0.33869099617004395, "learning_rate": 4.107016945287933e-06, "loss": 0.0119, "step": 43830 }, { "epoch": 16.099889827396254, "grad_norm": 0.28631317615509033, "learning_rate": 4.09390157241431e-06, "loss": 0.0117, "step": 43840 }, { "epoch": 16.103562247521115, "grad_norm": 0.3459970951080322, "learning_rate": 4.0808062804277515e-06, "loss": 0.0115, "step": 43850 }, { "epoch": 16.10723466764598, "grad_norm": 0.5360576510429382, "learning_rate": 4.067731075056596e-06, "loss": 0.0127, "step": 43860 }, { "epoch": 16.11090708777084, "grad_norm": 0.2649495601654053, "learning_rate": 4.054675962020376e-06, "loss": 0.0121, "step": 43870 }, { "epoch": 16.114579507895705, "grad_norm": 0.3189980685710907, "learning_rate": 4.04164094702984e-06, "loss": 0.0112, "step": 43880 }, { "epoch": 16.118251928020566, "grad_norm": 0.428073525428772, "learning_rate": 4.028626035786953e-06, "loss": 0.0117, "step": 43890 }, { "epoch": 16.121924348145427, "grad_norm": 0.44529518485069275, "learning_rate": 4.0156312339848675e-06, "loss": 0.0115, "step": 43900 }, { "epoch": 16.12559676827029, "grad_norm": 0.3974951505661011, "learning_rate": 4.00265654730797e-06, "loss": 0.0109, "step": 43910 }, { "epoch": 16.129269188395153, "grad_norm": 1.0217186212539673, "learning_rate": 3.989701981431809e-06, "loss": 0.0116, "step": 43920 }, { "epoch": 16.132941608520014, "grad_norm": 0.4579240381717682, "learning_rate": 3.976767542023169e-06, "loss": 0.0119, "step": 43930 }, { "epoch": 16.13661402864488, "grad_norm": 0.35783952474594116, "learning_rate": 3.963853234740006e-06, "loss": 0.0125, "step": 43940 }, { "epoch": 16.14028644876974, "grad_norm": 0.49181923270225525, "learning_rate": 3.950959065231464e-06, "loss": 0.0128, "step": 43950 }, { "epoch": 16.1439588688946, "grad_norm": 0.7841059565544128, "learning_rate": 3.938085039137917e-06, "loss": 0.0119, "step": 43960 }, { "epoch": 16.147631289019465, "grad_norm": 0.8212094902992249, "learning_rate": 3.925231162090881e-06, "loss": 0.0121, "step": 43970 }, { "epoch": 16.151303709144326, "grad_norm": 0.45490798354148865, "learning_rate": 3.912397439713084e-06, "loss": 0.0115, "step": 43980 }, { "epoch": 16.154976129269187, "grad_norm": 0.1820978969335556, "learning_rate": 3.899583877618435e-06, "loss": 0.013, "step": 43990 }, { "epoch": 16.15864854939405, "grad_norm": 0.4016929268836975, "learning_rate": 3.8867904814120225e-06, "loss": 0.0125, "step": 44000 }, { "epoch": 16.162320969518913, "grad_norm": 0.28113821148872375, "learning_rate": 3.8740172566901044e-06, "loss": 0.01, "step": 44010 }, { "epoch": 16.165993389643774, "grad_norm": 0.4948899745941162, "learning_rate": 3.861264209040133e-06, "loss": 0.0123, "step": 44020 }, { "epoch": 16.169665809768638, "grad_norm": 0.5366498231887817, "learning_rate": 3.848531344040718e-06, "loss": 0.011, "step": 44030 }, { "epoch": 16.1733382298935, "grad_norm": 0.255174458026886, "learning_rate": 3.835818667261642e-06, "loss": 0.0127, "step": 44040 }, { "epoch": 16.177010650018364, "grad_norm": 0.3073102831840515, "learning_rate": 3.823126184263881e-06, "loss": 0.0115, "step": 44050 }, { "epoch": 16.180683070143225, "grad_norm": 0.3650033175945282, "learning_rate": 3.810453900599542e-06, "loss": 0.0126, "step": 44060 }, { "epoch": 16.184355490268086, "grad_norm": 0.45238158106803894, "learning_rate": 3.7978018218119137e-06, "loss": 0.012, "step": 44070 }, { "epoch": 16.18802791039295, "grad_norm": 0.32607927918434143, "learning_rate": 3.7851699534354524e-06, "loss": 0.0107, "step": 44080 }, { "epoch": 16.19170033051781, "grad_norm": 0.3519165515899658, "learning_rate": 3.77255830099576e-06, "loss": 0.0122, "step": 44090 }, { "epoch": 16.195372750642672, "grad_norm": 0.27475506067276, "learning_rate": 3.759966870009607e-06, "loss": 0.0117, "step": 44100 }, { "epoch": 16.199045170767537, "grad_norm": 0.4023715555667877, "learning_rate": 3.7473956659848987e-06, "loss": 0.0111, "step": 44110 }, { "epoch": 16.202717590892398, "grad_norm": 0.5588219165802002, "learning_rate": 3.7348446944207305e-06, "loss": 0.011, "step": 44120 }, { "epoch": 16.20639001101726, "grad_norm": 0.1692427694797516, "learning_rate": 3.7223139608073e-06, "loss": 0.0116, "step": 44130 }, { "epoch": 16.210062431142124, "grad_norm": 0.4286520779132843, "learning_rate": 3.7098034706259943e-06, "loss": 0.0123, "step": 44140 }, { "epoch": 16.213734851266985, "grad_norm": 0.4520436227321625, "learning_rate": 3.697313229349314e-06, "loss": 0.0114, "step": 44150 }, { "epoch": 16.217407271391846, "grad_norm": 0.22037000954151154, "learning_rate": 3.684843242440933e-06, "loss": 0.0124, "step": 44160 }, { "epoch": 16.22107969151671, "grad_norm": 0.3304981589317322, "learning_rate": 3.672393515355621e-06, "loss": 0.0119, "step": 44170 }, { "epoch": 16.22475211164157, "grad_norm": 0.19008778035640717, "learning_rate": 3.659964053539333e-06, "loss": 0.0117, "step": 44180 }, { "epoch": 16.228424531766436, "grad_norm": 0.43182072043418884, "learning_rate": 3.647554862429131e-06, "loss": 0.0108, "step": 44190 }, { "epoch": 16.232096951891297, "grad_norm": 0.40210360288619995, "learning_rate": 3.635165947453212e-06, "loss": 0.0121, "step": 44200 }, { "epoch": 16.235769372016158, "grad_norm": 0.45101824402809143, "learning_rate": 3.622797314030918e-06, "loss": 0.0125, "step": 44210 }, { "epoch": 16.239441792141022, "grad_norm": 0.3776438534259796, "learning_rate": 3.610448967572705e-06, "loss": 0.0129, "step": 44220 }, { "epoch": 16.243114212265883, "grad_norm": 0.35141435265541077, "learning_rate": 3.5981209134801464e-06, "loss": 0.0104, "step": 44230 }, { "epoch": 16.246786632390744, "grad_norm": 0.33341696858406067, "learning_rate": 3.585813157145973e-06, "loss": 0.0111, "step": 44240 }, { "epoch": 16.25045905251561, "grad_norm": 0.510565996170044, "learning_rate": 3.5735257039540114e-06, "loss": 0.0121, "step": 44250 }, { "epoch": 16.25413147264047, "grad_norm": 0.5499540567398071, "learning_rate": 3.5612585592792014e-06, "loss": 0.0131, "step": 44260 }, { "epoch": 16.25780389276533, "grad_norm": 0.6026186943054199, "learning_rate": 3.5490117284876055e-06, "loss": 0.0112, "step": 44270 }, { "epoch": 16.261476312890196, "grad_norm": 0.4519186019897461, "learning_rate": 3.5367852169364224e-06, "loss": 0.0128, "step": 44280 }, { "epoch": 16.265148733015057, "grad_norm": 0.6195192337036133, "learning_rate": 3.524579029973929e-06, "loss": 0.0116, "step": 44290 }, { "epoch": 16.268821153139918, "grad_norm": 0.4895787239074707, "learning_rate": 3.5123931729395267e-06, "loss": 0.0119, "step": 44300 }, { "epoch": 16.272493573264782, "grad_norm": 0.5370087623596191, "learning_rate": 3.500227651163729e-06, "loss": 0.013, "step": 44310 }, { "epoch": 16.276165993389643, "grad_norm": 0.5002564191818237, "learning_rate": 3.488082469968146e-06, "loss": 0.0119, "step": 44320 }, { "epoch": 16.279838413514508, "grad_norm": 0.2292511761188507, "learning_rate": 3.475957634665489e-06, "loss": 0.0118, "step": 44330 }, { "epoch": 16.28351083363937, "grad_norm": 0.30124619603157043, "learning_rate": 3.463853150559582e-06, "loss": 0.0125, "step": 44340 }, { "epoch": 16.28718325376423, "grad_norm": 0.7932311296463013, "learning_rate": 3.451769022945334e-06, "loss": 0.0123, "step": 44350 }, { "epoch": 16.290855673889094, "grad_norm": 0.3638269901275635, "learning_rate": 3.4397052571087385e-06, "loss": 0.0123, "step": 44360 }, { "epoch": 16.294528094013955, "grad_norm": 0.3075825572013855, "learning_rate": 3.427661858326925e-06, "loss": 0.01, "step": 44370 }, { "epoch": 16.298200514138816, "grad_norm": 0.5725074410438538, "learning_rate": 3.4156388318680566e-06, "loss": 0.0117, "step": 44380 }, { "epoch": 16.30187293426368, "grad_norm": 0.3570789396762848, "learning_rate": 3.4036361829914322e-06, "loss": 0.012, "step": 44390 }, { "epoch": 16.305545354388542, "grad_norm": 0.41983917355537415, "learning_rate": 3.391653916947407e-06, "loss": 0.0134, "step": 44400 }, { "epoch": 16.309217774513403, "grad_norm": 0.27175775170326233, "learning_rate": 3.3796920389774444e-06, "loss": 0.0119, "step": 44410 }, { "epoch": 16.312890194638268, "grad_norm": 0.18464480340480804, "learning_rate": 3.3677505543140643e-06, "loss": 0.0142, "step": 44420 }, { "epoch": 16.31656261476313, "grad_norm": 0.3230923116207123, "learning_rate": 3.355829468180882e-06, "loss": 0.0126, "step": 44430 }, { "epoch": 16.32023503488799, "grad_norm": 0.28958332538604736, "learning_rate": 3.343928785792588e-06, "loss": 0.0106, "step": 44440 }, { "epoch": 16.323907455012854, "grad_norm": 0.6693487167358398, "learning_rate": 3.3320485123549405e-06, "loss": 0.0124, "step": 44450 }, { "epoch": 16.327579875137715, "grad_norm": 0.259894460439682, "learning_rate": 3.3201886530647707e-06, "loss": 0.0114, "step": 44460 }, { "epoch": 16.33125229526258, "grad_norm": 0.2609195113182068, "learning_rate": 3.3083492131100015e-06, "loss": 0.0122, "step": 44470 }, { "epoch": 16.33492471538744, "grad_norm": 0.6525461077690125, "learning_rate": 3.2965301976695896e-06, "loss": 0.0099, "step": 44480 }, { "epoch": 16.338597135512302, "grad_norm": 1.6380215883255005, "learning_rate": 3.284731611913577e-06, "loss": 0.0119, "step": 44490 }, { "epoch": 16.342269555637166, "grad_norm": 0.6572830080986023, "learning_rate": 3.27295346100307e-06, "loss": 0.0115, "step": 44500 }, { "epoch": 16.345941975762027, "grad_norm": 0.7539956569671631, "learning_rate": 3.2611957500902347e-06, "loss": 0.0116, "step": 44510 }, { "epoch": 16.34961439588689, "grad_norm": 0.48943454027175903, "learning_rate": 3.2494584843182796e-06, "loss": 0.0118, "step": 44520 }, { "epoch": 16.353286816011753, "grad_norm": 0.3604115843772888, "learning_rate": 3.2377416688214978e-06, "loss": 0.0119, "step": 44530 }, { "epoch": 16.356959236136614, "grad_norm": 1.0015807151794434, "learning_rate": 3.2260453087252175e-06, "loss": 0.012, "step": 44540 }, { "epoch": 16.360631656261475, "grad_norm": 0.5346447229385376, "learning_rate": 3.2143694091458175e-06, "loss": 0.0127, "step": 44550 }, { "epoch": 16.36430407638634, "grad_norm": 0.42131495475769043, "learning_rate": 3.2027139751907466e-06, "loss": 0.0126, "step": 44560 }, { "epoch": 16.3679764965112, "grad_norm": 0.4850321114063263, "learning_rate": 3.191079011958481e-06, "loss": 0.0104, "step": 44570 }, { "epoch": 16.37164891663606, "grad_norm": 0.31082895398139954, "learning_rate": 3.1794645245385337e-06, "loss": 0.0121, "step": 44580 }, { "epoch": 16.375321336760926, "grad_norm": 0.4567244052886963, "learning_rate": 3.1678705180115063e-06, "loss": 0.0158, "step": 44590 }, { "epoch": 16.378993756885787, "grad_norm": 0.3978477120399475, "learning_rate": 3.1562969974489933e-06, "loss": 0.0111, "step": 44600 }, { "epoch": 16.38266617701065, "grad_norm": 0.6529583930969238, "learning_rate": 3.144743967913638e-06, "loss": 0.0116, "step": 44610 }, { "epoch": 16.386338597135513, "grad_norm": 0.4269106090068817, "learning_rate": 3.1332114344591436e-06, "loss": 0.0123, "step": 44620 }, { "epoch": 16.390011017260374, "grad_norm": 0.3995301127433777, "learning_rate": 3.1216994021302235e-06, "loss": 0.0118, "step": 44630 }, { "epoch": 16.39368343738524, "grad_norm": 0.2721642255783081, "learning_rate": 3.1102078759626396e-06, "loss": 0.0117, "step": 44640 }, { "epoch": 16.3973558575101, "grad_norm": 0.38013511896133423, "learning_rate": 3.098736860983159e-06, "loss": 0.0124, "step": 44650 }, { "epoch": 16.40102827763496, "grad_norm": 0.6748245358467102, "learning_rate": 3.087286362209618e-06, "loss": 0.012, "step": 44660 }, { "epoch": 16.404700697759825, "grad_norm": 0.2123679518699646, "learning_rate": 3.0758563846508383e-06, "loss": 0.0121, "step": 44670 }, { "epoch": 16.408373117884686, "grad_norm": 0.5994270443916321, "learning_rate": 3.0644469333066748e-06, "loss": 0.0117, "step": 44680 }, { "epoch": 16.412045538009547, "grad_norm": 0.36831948161125183, "learning_rate": 3.0530580131680307e-06, "loss": 0.0125, "step": 44690 }, { "epoch": 16.41571795813441, "grad_norm": 0.8136029243469238, "learning_rate": 3.0416896292167875e-06, "loss": 0.0132, "step": 44700 }, { "epoch": 16.419390378259273, "grad_norm": 0.5642306208610535, "learning_rate": 3.030341786425861e-06, "loss": 0.0113, "step": 44710 }, { "epoch": 16.423062798384134, "grad_norm": 0.5069273710250854, "learning_rate": 3.0190144897591975e-06, "loss": 0.0113, "step": 44720 }, { "epoch": 16.426735218509, "grad_norm": 1.2932811975479126, "learning_rate": 3.0077077441717395e-06, "loss": 0.0124, "step": 44730 }, { "epoch": 16.43040763863386, "grad_norm": 0.27322813868522644, "learning_rate": 2.9964215546094244e-06, "loss": 0.0124, "step": 44740 }, { "epoch": 16.43408005875872, "grad_norm": 0.5487871766090393, "learning_rate": 2.9851559260092367e-06, "loss": 0.0131, "step": 44750 }, { "epoch": 16.437752478883585, "grad_norm": 0.3195132613182068, "learning_rate": 2.973910863299134e-06, "loss": 0.0114, "step": 44760 }, { "epoch": 16.441424899008446, "grad_norm": 0.2900964021682739, "learning_rate": 2.962686371398088e-06, "loss": 0.0116, "step": 44770 }, { "epoch": 16.44509731913331, "grad_norm": 0.39177632331848145, "learning_rate": 2.95148245521607e-06, "loss": 0.0134, "step": 44780 }, { "epoch": 16.44876973925817, "grad_norm": 0.3513185679912567, "learning_rate": 2.9402991196540665e-06, "loss": 0.0123, "step": 44790 }, { "epoch": 16.452442159383033, "grad_norm": 0.56032395362854, "learning_rate": 2.929136369604041e-06, "loss": 0.012, "step": 44800 }, { "epoch": 16.456114579507897, "grad_norm": 0.5423532724380493, "learning_rate": 2.9179942099489487e-06, "loss": 0.011, "step": 44810 }, { "epoch": 16.459786999632758, "grad_norm": 0.46859705448150635, "learning_rate": 2.9068726455627694e-06, "loss": 0.0142, "step": 44820 }, { "epoch": 16.46345941975762, "grad_norm": 0.6500265002250671, "learning_rate": 2.8957716813104384e-06, "loss": 0.0129, "step": 44830 }, { "epoch": 16.467131839882484, "grad_norm": 0.2066100388765335, "learning_rate": 2.8846913220479e-06, "loss": 0.0124, "step": 44840 }, { "epoch": 16.470804260007345, "grad_norm": 0.27692294120788574, "learning_rate": 2.873631572622082e-06, "loss": 0.0119, "step": 44850 }, { "epoch": 16.474476680132206, "grad_norm": 0.41772258281707764, "learning_rate": 2.862592437870881e-06, "loss": 0.0123, "step": 44860 }, { "epoch": 16.47814910025707, "grad_norm": 0.34859880805015564, "learning_rate": 2.8515739226232075e-06, "loss": 0.0128, "step": 44870 }, { "epoch": 16.48182152038193, "grad_norm": 0.4385228753089905, "learning_rate": 2.840576031698927e-06, "loss": 0.0111, "step": 44880 }, { "epoch": 16.485493940506792, "grad_norm": 0.5324676036834717, "learning_rate": 2.8295987699088923e-06, "loss": 0.0129, "step": 44890 }, { "epoch": 16.489166360631657, "grad_norm": 0.3176063001155853, "learning_rate": 2.8186421420549227e-06, "loss": 0.0119, "step": 44900 }, { "epoch": 16.492838780756518, "grad_norm": 0.558587908744812, "learning_rate": 2.8077061529298343e-06, "loss": 0.0131, "step": 44910 }, { "epoch": 16.496511200881383, "grad_norm": 0.3235524892807007, "learning_rate": 2.79679080731739e-06, "loss": 0.0117, "step": 44920 }, { "epoch": 16.500183621006244, "grad_norm": 0.339927613735199, "learning_rate": 2.7858961099923254e-06, "loss": 0.0113, "step": 44930 }, { "epoch": 16.503856041131105, "grad_norm": 0.3374999463558197, "learning_rate": 2.7750220657203727e-06, "loss": 0.0119, "step": 44940 }, { "epoch": 16.50752846125597, "grad_norm": 0.3878500163555145, "learning_rate": 2.764168679258189e-06, "loss": 0.011, "step": 44950 }, { "epoch": 16.51120088138083, "grad_norm": 0.3302108347415924, "learning_rate": 2.7533359553534254e-06, "loss": 0.0105, "step": 44960 }, { "epoch": 16.51487330150569, "grad_norm": 0.33180662989616394, "learning_rate": 2.7425238987446754e-06, "loss": 0.0102, "step": 44970 }, { "epoch": 16.518545721630556, "grad_norm": 0.3175395429134369, "learning_rate": 2.731732514161506e-06, "loss": 0.0118, "step": 44980 }, { "epoch": 16.522218141755417, "grad_norm": 0.3310856223106384, "learning_rate": 2.720961806324435e-06, "loss": 0.0125, "step": 44990 }, { "epoch": 16.525890561880278, "grad_norm": 0.2926732003688812, "learning_rate": 2.7102117799449266e-06, "loss": 0.0114, "step": 45000 }, { "epoch": 16.529562982005142, "grad_norm": 0.2863992750644684, "learning_rate": 2.6994824397254258e-06, "loss": 0.0113, "step": 45010 }, { "epoch": 16.533235402130003, "grad_norm": 0.2172027975320816, "learning_rate": 2.688773790359295e-06, "loss": 0.0132, "step": 45020 }, { "epoch": 16.536907822254864, "grad_norm": 0.2520172894001007, "learning_rate": 2.678085836530858e-06, "loss": 0.0128, "step": 45030 }, { "epoch": 16.54058024237973, "grad_norm": 0.4439863860607147, "learning_rate": 2.6674185829154096e-06, "loss": 0.0125, "step": 45040 }, { "epoch": 16.54425266250459, "grad_norm": 0.512855589389801, "learning_rate": 2.6567720341791556e-06, "loss": 0.0107, "step": 45050 }, { "epoch": 16.547925082629455, "grad_norm": 0.20047996938228607, "learning_rate": 2.646146194979249e-06, "loss": 0.0115, "step": 45060 }, { "epoch": 16.551597502754316, "grad_norm": 0.3539259731769562, "learning_rate": 2.6355410699638094e-06, "loss": 0.014, "step": 45070 }, { "epoch": 16.555269922879177, "grad_norm": 0.30438628792762756, "learning_rate": 2.6249566637718714e-06, "loss": 0.0115, "step": 45080 }, { "epoch": 16.55894234300404, "grad_norm": 0.3601897954940796, "learning_rate": 2.614392981033409e-06, "loss": 0.013, "step": 45090 }, { "epoch": 16.562614763128902, "grad_norm": 0.1510007232427597, "learning_rate": 2.603850026369348e-06, "loss": 0.0117, "step": 45100 }, { "epoch": 16.566287183253763, "grad_norm": 0.37204286456108093, "learning_rate": 2.593327804391521e-06, "loss": 0.0118, "step": 45110 }, { "epoch": 16.569959603378628, "grad_norm": 0.425546258687973, "learning_rate": 2.5828263197027247e-06, "loss": 0.012, "step": 45120 }, { "epoch": 16.57363202350349, "grad_norm": 0.3243373930454254, "learning_rate": 2.57234557689664e-06, "loss": 0.012, "step": 45130 }, { "epoch": 16.57730444362835, "grad_norm": 1.3374818563461304, "learning_rate": 2.5618855805579213e-06, "loss": 0.012, "step": 45140 }, { "epoch": 16.580976863753214, "grad_norm": 0.4180919826030731, "learning_rate": 2.5514463352621197e-06, "loss": 0.0125, "step": 45150 }, { "epoch": 16.584649283878075, "grad_norm": 0.44017672538757324, "learning_rate": 2.541027845575711e-06, "loss": 0.0123, "step": 45160 }, { "epoch": 16.588321704002936, "grad_norm": 0.24431483447551727, "learning_rate": 2.530630116056104e-06, "loss": 0.0112, "step": 45170 }, { "epoch": 16.5919941241278, "grad_norm": 0.6673574447631836, "learning_rate": 2.5202531512516114e-06, "loss": 0.0111, "step": 45180 }, { "epoch": 16.595666544252662, "grad_norm": 0.5039991140365601, "learning_rate": 2.509896955701474e-06, "loss": 0.0118, "step": 45190 }, { "epoch": 16.599338964377523, "grad_norm": 0.38430166244506836, "learning_rate": 2.4995615339358404e-06, "loss": 0.0111, "step": 45200 }, { "epoch": 16.603011384502388, "grad_norm": 0.3773319125175476, "learning_rate": 2.489246890475788e-06, "loss": 0.0148, "step": 45210 }, { "epoch": 16.60668380462725, "grad_norm": 0.8590237498283386, "learning_rate": 2.4789530298332696e-06, "loss": 0.0115, "step": 45220 }, { "epoch": 16.610356224752113, "grad_norm": 0.4909283220767975, "learning_rate": 2.4686799565111885e-06, "loss": 0.0115, "step": 45230 }, { "epoch": 16.614028644876974, "grad_norm": 0.11452154815196991, "learning_rate": 2.4584276750033332e-06, "loss": 0.0118, "step": 45240 }, { "epoch": 16.617701065001835, "grad_norm": 0.31974101066589355, "learning_rate": 2.448196189794383e-06, "loss": 0.0137, "step": 45250 }, { "epoch": 16.6213734851267, "grad_norm": 0.2594102919101715, "learning_rate": 2.437985505359963e-06, "loss": 0.0119, "step": 45260 }, { "epoch": 16.62504590525156, "grad_norm": 0.23858410120010376, "learning_rate": 2.4277956261665626e-06, "loss": 0.0121, "step": 45270 }, { "epoch": 16.628718325376422, "grad_norm": 0.43161576986312866, "learning_rate": 2.4176265566715706e-06, "loss": 0.0109, "step": 45280 }, { "epoch": 16.632390745501286, "grad_norm": 0.47175684571266174, "learning_rate": 2.407478301323296e-06, "loss": 0.0122, "step": 45290 }, { "epoch": 16.636063165626148, "grad_norm": 0.4156486392021179, "learning_rate": 2.397350864560932e-06, "loss": 0.0119, "step": 45300 }, { "epoch": 16.63973558575101, "grad_norm": 0.6373164057731628, "learning_rate": 2.387244250814552e-06, "loss": 0.0127, "step": 45310 }, { "epoch": 16.643408005875873, "grad_norm": 0.416446328163147, "learning_rate": 2.377158464505136e-06, "loss": 0.0123, "step": 45320 }, { "epoch": 16.647080426000734, "grad_norm": 0.30029964447021484, "learning_rate": 2.36709351004456e-06, "loss": 0.0134, "step": 45330 }, { "epoch": 16.650752846125595, "grad_norm": 1.0722664594650269, "learning_rate": 2.357049391835564e-06, "loss": 0.0123, "step": 45340 }, { "epoch": 16.65442526625046, "grad_norm": 0.7485013008117676, "learning_rate": 2.347026114271794e-06, "loss": 0.0112, "step": 45350 }, { "epoch": 16.65809768637532, "grad_norm": 0.24788062274456024, "learning_rate": 2.3370236817377653e-06, "loss": 0.0129, "step": 45360 }, { "epoch": 16.661770106500185, "grad_norm": 0.24952131509780884, "learning_rate": 2.3270420986088947e-06, "loss": 0.0119, "step": 45370 }, { "epoch": 16.665442526625046, "grad_norm": 0.2828926742076874, "learning_rate": 2.3170813692514505e-06, "loss": 0.0136, "step": 45380 }, { "epoch": 16.669114946749907, "grad_norm": 0.40258729457855225, "learning_rate": 2.3071414980226025e-06, "loss": 0.0128, "step": 45390 }, { "epoch": 16.672787366874772, "grad_norm": 0.5970770716667175, "learning_rate": 2.29722248927039e-06, "loss": 0.012, "step": 45400 }, { "epoch": 16.676459786999633, "grad_norm": 0.45576661825180054, "learning_rate": 2.287324347333719e-06, "loss": 0.0142, "step": 45410 }, { "epoch": 16.680132207124494, "grad_norm": 1.1527488231658936, "learning_rate": 2.2774470765423773e-06, "loss": 0.0122, "step": 45420 }, { "epoch": 16.68380462724936, "grad_norm": 0.17176927626132965, "learning_rate": 2.2675906812170133e-06, "loss": 0.0121, "step": 45430 }, { "epoch": 16.68747704737422, "grad_norm": 0.5813843011856079, "learning_rate": 2.257755165669151e-06, "loss": 0.012, "step": 45440 }, { "epoch": 16.69114946749908, "grad_norm": 0.5641176700592041, "learning_rate": 2.2479405342011872e-06, "loss": 0.0116, "step": 45450 }, { "epoch": 16.694821887623945, "grad_norm": 0.17775243520736694, "learning_rate": 2.238146791106366e-06, "loss": 0.0109, "step": 45460 }, { "epoch": 16.698494307748806, "grad_norm": 0.3337056636810303, "learning_rate": 2.2283739406688096e-06, "loss": 0.0119, "step": 45470 }, { "epoch": 16.702166727873667, "grad_norm": 0.46698302030563354, "learning_rate": 2.2186219871634774e-06, "loss": 0.0115, "step": 45480 }, { "epoch": 16.70583914799853, "grad_norm": 0.2723940312862396, "learning_rate": 2.2088909348562335e-06, "loss": 0.0119, "step": 45490 }, { "epoch": 16.709511568123393, "grad_norm": 0.39962470531463623, "learning_rate": 2.1991807880037517e-06, "loss": 0.0118, "step": 45500 }, { "epoch": 16.713183988248254, "grad_norm": 0.41131529211997986, "learning_rate": 2.1894915508535784e-06, "loss": 0.0138, "step": 45510 }, { "epoch": 16.71685640837312, "grad_norm": 0.9090644121170044, "learning_rate": 2.179823227644123e-06, "loss": 0.0124, "step": 45520 }, { "epoch": 16.72052882849798, "grad_norm": 0.34142741560935974, "learning_rate": 2.170175822604642e-06, "loss": 0.0116, "step": 45530 }, { "epoch": 16.724201248622844, "grad_norm": 0.4949135482311249, "learning_rate": 2.1605493399552256e-06, "loss": 0.0114, "step": 45540 }, { "epoch": 16.727873668747705, "grad_norm": 0.373651385307312, "learning_rate": 2.150943783906839e-06, "loss": 0.0135, "step": 45550 }, { "epoch": 16.731546088872566, "grad_norm": 0.570288360118866, "learning_rate": 2.1413591586612712e-06, "loss": 0.012, "step": 45560 }, { "epoch": 16.73521850899743, "grad_norm": 0.41897210478782654, "learning_rate": 2.1317954684111576e-06, "loss": 0.0123, "step": 45570 }, { "epoch": 16.73889092912229, "grad_norm": 0.6477263569831848, "learning_rate": 2.1222527173399896e-06, "loss": 0.0113, "step": 45580 }, { "epoch": 16.742563349247153, "grad_norm": 0.5197526812553406, "learning_rate": 2.112730909622085e-06, "loss": 0.0108, "step": 45590 }, { "epoch": 16.746235769372017, "grad_norm": 0.7105072736740112, "learning_rate": 2.103230049422611e-06, "loss": 0.011, "step": 45600 }, { "epoch": 16.749908189496878, "grad_norm": 0.26469749212265015, "learning_rate": 2.0937501408975603e-06, "loss": 0.0117, "step": 45610 }, { "epoch": 16.75358060962174, "grad_norm": 0.5737707614898682, "learning_rate": 2.084291188193782e-06, "loss": 0.0124, "step": 45620 }, { "epoch": 16.757253029746604, "grad_norm": 0.2694288194179535, "learning_rate": 2.0748531954489224e-06, "loss": 0.0124, "step": 45630 }, { "epoch": 16.760925449871465, "grad_norm": 0.5413709878921509, "learning_rate": 2.065436166791496e-06, "loss": 0.0112, "step": 45640 }, { "epoch": 16.76459786999633, "grad_norm": 0.5219104886054993, "learning_rate": 2.056040106340823e-06, "loss": 0.011, "step": 45650 }, { "epoch": 16.76827029012119, "grad_norm": 0.4761609733104706, "learning_rate": 2.0466650182070667e-06, "loss": 0.0111, "step": 45660 }, { "epoch": 16.77194271024605, "grad_norm": 0.27096837759017944, "learning_rate": 2.0373109064911964e-06, "loss": 0.0128, "step": 45670 }, { "epoch": 16.775615130370916, "grad_norm": 0.38568368554115295, "learning_rate": 2.027977775285028e-06, "loss": 0.0112, "step": 45680 }, { "epoch": 16.779287550495777, "grad_norm": 0.6279948949813843, "learning_rate": 2.0186656286711958e-06, "loss": 0.0113, "step": 45690 }, { "epoch": 16.782959970620638, "grad_norm": 0.5017985701560974, "learning_rate": 2.00937447072313e-06, "loss": 0.012, "step": 45700 }, { "epoch": 16.786632390745503, "grad_norm": 0.17287065088748932, "learning_rate": 2.0001043055051183e-06, "loss": 0.0114, "step": 45710 }, { "epoch": 16.790304810870364, "grad_norm": 0.8097580075263977, "learning_rate": 1.9908551370722393e-06, "loss": 0.0111, "step": 45720 }, { "epoch": 16.793977230995225, "grad_norm": 0.6212011575698853, "learning_rate": 1.981626969470385e-06, "loss": 0.0116, "step": 45730 }, { "epoch": 16.79764965112009, "grad_norm": 0.38698068261146545, "learning_rate": 1.9724198067362864e-06, "loss": 0.0114, "step": 45740 }, { "epoch": 16.80132207124495, "grad_norm": 0.49199503660202026, "learning_rate": 1.963233652897456e-06, "loss": 0.0123, "step": 45750 }, { "epoch": 16.80499449136981, "grad_norm": 0.2186126410961151, "learning_rate": 1.9540685119722234e-06, "loss": 0.0111, "step": 45760 }, { "epoch": 16.808666911494676, "grad_norm": 0.27022412419319153, "learning_rate": 1.9449243879697543e-06, "loss": 0.0117, "step": 45770 }, { "epoch": 16.812339331619537, "grad_norm": 0.2839871048927307, "learning_rate": 1.9358012848899874e-06, "loss": 0.011, "step": 45780 }, { "epoch": 16.816011751744398, "grad_norm": 0.41862040758132935, "learning_rate": 1.92669920672367e-06, "loss": 0.0102, "step": 45790 }, { "epoch": 16.819684171869262, "grad_norm": 0.8877990245819092, "learning_rate": 1.917618157452372e-06, "loss": 0.0126, "step": 45800 }, { "epoch": 16.823356591994123, "grad_norm": 0.5642653703689575, "learning_rate": 1.9085581410484498e-06, "loss": 0.0124, "step": 45810 }, { "epoch": 16.827029012118988, "grad_norm": 0.5637781023979187, "learning_rate": 1.8995191614750607e-06, "loss": 0.0116, "step": 45820 }, { "epoch": 16.83070143224385, "grad_norm": 0.5596258044242859, "learning_rate": 1.890501222686164e-06, "loss": 0.0119, "step": 45830 }, { "epoch": 16.83437385236871, "grad_norm": 0.5899092555046082, "learning_rate": 1.8815043286265045e-06, "loss": 0.0129, "step": 45840 }, { "epoch": 16.838046272493575, "grad_norm": 0.6302478313446045, "learning_rate": 1.8725284832316448e-06, "loss": 0.0123, "step": 45850 }, { "epoch": 16.841718692618436, "grad_norm": 0.6414161324501038, "learning_rate": 1.8635736904279055e-06, "loss": 0.0128, "step": 45860 }, { "epoch": 16.845391112743297, "grad_norm": 0.24956342577934265, "learning_rate": 1.8546399541324367e-06, "loss": 0.0125, "step": 45870 }, { "epoch": 16.84906353286816, "grad_norm": 0.8933922052383423, "learning_rate": 1.8457272782531453e-06, "loss": 0.011, "step": 45880 }, { "epoch": 16.852735952993022, "grad_norm": 0.45291489362716675, "learning_rate": 1.836835666688741e-06, "loss": 0.0106, "step": 45890 }, { "epoch": 16.856408373117883, "grad_norm": 0.42069363594055176, "learning_rate": 1.8279651233287175e-06, "loss": 0.011, "step": 45900 }, { "epoch": 16.860080793242748, "grad_norm": 0.24887536466121674, "learning_rate": 1.8191156520533547e-06, "loss": 0.0112, "step": 45910 }, { "epoch": 16.86375321336761, "grad_norm": 0.25529706478118896, "learning_rate": 1.8102872567337059e-06, "loss": 0.0122, "step": 45920 }, { "epoch": 16.86742563349247, "grad_norm": 0.4686102867126465, "learning_rate": 1.8014799412316152e-06, "loss": 0.0102, "step": 45930 }, { "epoch": 16.871098053617335, "grad_norm": 0.3092283308506012, "learning_rate": 1.7926937093997065e-06, "loss": 0.0119, "step": 45940 }, { "epoch": 16.874770473742196, "grad_norm": 0.6091191172599792, "learning_rate": 1.7839285650813608e-06, "loss": 0.0122, "step": 45950 }, { "epoch": 16.87844289386706, "grad_norm": 0.7222652435302734, "learning_rate": 1.7751845121107668e-06, "loss": 0.012, "step": 45960 }, { "epoch": 16.88211531399192, "grad_norm": 0.33253833651542664, "learning_rate": 1.7664615543128648e-06, "loss": 0.0127, "step": 45970 }, { "epoch": 16.885787734116782, "grad_norm": 0.5926063656806946, "learning_rate": 1.7577596955033692e-06, "loss": 0.011, "step": 45980 }, { "epoch": 16.889460154241647, "grad_norm": 0.39026811718940735, "learning_rate": 1.7490789394887686e-06, "loss": 0.0117, "step": 45990 }, { "epoch": 16.893132574366508, "grad_norm": 0.31570881605148315, "learning_rate": 1.7404192900663252e-06, "loss": 0.0106, "step": 46000 }, { "epoch": 16.89680499449137, "grad_norm": 0.8343237638473511, "learning_rate": 1.7317807510240647e-06, "loss": 0.0114, "step": 46010 }, { "epoch": 16.900477414616233, "grad_norm": 0.18860608339309692, "learning_rate": 1.7231633261407642e-06, "loss": 0.0118, "step": 46020 }, { "epoch": 16.904149834741094, "grad_norm": 0.18526193499565125, "learning_rate": 1.7145670191859974e-06, "loss": 0.0127, "step": 46030 }, { "epoch": 16.907822254865955, "grad_norm": 1.0886540412902832, "learning_rate": 1.7059918339200732e-06, "loss": 0.0112, "step": 46040 }, { "epoch": 16.91149467499082, "grad_norm": 0.2286754995584488, "learning_rate": 1.6974377740940629e-06, "loss": 0.0118, "step": 46050 }, { "epoch": 16.91516709511568, "grad_norm": 0.8012502789497375, "learning_rate": 1.6889048434498066e-06, "loss": 0.011, "step": 46060 }, { "epoch": 16.918839515240542, "grad_norm": 0.5218085646629333, "learning_rate": 1.6803930457199023e-06, "loss": 0.0113, "step": 46070 }, { "epoch": 16.922511935365407, "grad_norm": 0.26714006066322327, "learning_rate": 1.6719023846276938e-06, "loss": 0.0123, "step": 46080 }, { "epoch": 16.926184355490268, "grad_norm": 0.4378867447376251, "learning_rate": 1.6634328638872877e-06, "loss": 0.0133, "step": 46090 }, { "epoch": 16.92985677561513, "grad_norm": 0.8500576019287109, "learning_rate": 1.6549844872035424e-06, "loss": 0.0118, "step": 46100 }, { "epoch": 16.933529195739993, "grad_norm": 0.24715076386928558, "learning_rate": 1.6465572582720579e-06, "loss": 0.0119, "step": 46110 }, { "epoch": 16.937201615864854, "grad_norm": 0.780332624912262, "learning_rate": 1.6381511807792017e-06, "loss": 0.0109, "step": 46120 }, { "epoch": 16.94087403598972, "grad_norm": 0.346129834651947, "learning_rate": 1.6297662584020656e-06, "loss": 0.0108, "step": 46130 }, { "epoch": 16.94454645611458, "grad_norm": 0.3214295208454132, "learning_rate": 1.621402494808505e-06, "loss": 0.0117, "step": 46140 }, { "epoch": 16.94821887623944, "grad_norm": 0.3102918565273285, "learning_rate": 1.613059893657115e-06, "loss": 0.0106, "step": 46150 }, { "epoch": 16.951891296364305, "grad_norm": 0.5419324040412903, "learning_rate": 1.604738458597238e-06, "loss": 0.0124, "step": 46160 }, { "epoch": 16.955563716489166, "grad_norm": 0.4093334972858429, "learning_rate": 1.5964381932689398e-06, "loss": 0.0123, "step": 46170 }, { "epoch": 16.959236136614027, "grad_norm": 0.5575114488601685, "learning_rate": 1.5881591013030384e-06, "loss": 0.0102, "step": 46180 }, { "epoch": 16.962908556738892, "grad_norm": 0.5316087007522583, "learning_rate": 1.5799011863211089e-06, "loss": 0.0123, "step": 46190 }, { "epoch": 16.966580976863753, "grad_norm": 0.390379399061203, "learning_rate": 1.5716644519354285e-06, "loss": 0.0126, "step": 46200 }, { "epoch": 16.970253396988614, "grad_norm": 0.6060245633125305, "learning_rate": 1.5634489017490151e-06, "loss": 0.0129, "step": 46210 }, { "epoch": 16.97392581711348, "grad_norm": 0.3961685597896576, "learning_rate": 1.555254539355655e-06, "loss": 0.0115, "step": 46220 }, { "epoch": 16.97759823723834, "grad_norm": 0.24187251925468445, "learning_rate": 1.5470813683398255e-06, "loss": 0.0121, "step": 46230 }, { "epoch": 16.981270657363204, "grad_norm": 0.7870013117790222, "learning_rate": 1.5389293922767444e-06, "loss": 0.0107, "step": 46240 }, { "epoch": 16.984943077488065, "grad_norm": 0.5158800482749939, "learning_rate": 1.530798614732376e-06, "loss": 0.0128, "step": 46250 }, { "epoch": 16.988615497612926, "grad_norm": 0.2775740623474121, "learning_rate": 1.5226890392633919e-06, "loss": 0.0111, "step": 46260 }, { "epoch": 16.99228791773779, "grad_norm": 0.4039190411567688, "learning_rate": 1.5146006694171932e-06, "loss": 0.0117, "step": 46270 }, { "epoch": 16.995960337862652, "grad_norm": 0.2889106869697571, "learning_rate": 1.5065335087319166e-06, "loss": 0.0105, "step": 46280 }, { "epoch": 16.999632757987513, "grad_norm": 0.24839378893375397, "learning_rate": 1.4984875607364057e-06, "loss": 0.0126, "step": 46290 }, { "epoch": 17.003305178112377, "grad_norm": 0.2107834666967392, "learning_rate": 1.4904628289502287e-06, "loss": 0.0128, "step": 46300 }, { "epoch": 17.00697759823724, "grad_norm": 1.070357322692871, "learning_rate": 1.482459316883683e-06, "loss": 0.0129, "step": 46310 }, { "epoch": 17.0106500183621, "grad_norm": 0.37857386469841003, "learning_rate": 1.474477028037763e-06, "loss": 0.0119, "step": 46320 }, { "epoch": 17.014322438486964, "grad_norm": 0.36337292194366455, "learning_rate": 1.4665159659042139e-06, "loss": 0.0108, "step": 46330 }, { "epoch": 17.017994858611825, "grad_norm": 0.527871310710907, "learning_rate": 1.4585761339654613e-06, "loss": 0.0119, "step": 46340 }, { "epoch": 17.021667278736686, "grad_norm": 0.21579504013061523, "learning_rate": 1.4506575356946606e-06, "loss": 0.012, "step": 46350 }, { "epoch": 17.02533969886155, "grad_norm": 0.4609515964984894, "learning_rate": 1.4427601745556796e-06, "loss": 0.011, "step": 46360 }, { "epoch": 17.02901211898641, "grad_norm": 0.27999019622802734, "learning_rate": 1.4348840540030828e-06, "loss": 0.0112, "step": 46370 }, { "epoch": 17.032684539111273, "grad_norm": 0.40856266021728516, "learning_rate": 1.4270291774821586e-06, "loss": 0.011, "step": 46380 }, { "epoch": 17.036356959236137, "grad_norm": 0.6804297566413879, "learning_rate": 1.4191955484289032e-06, "loss": 0.0124, "step": 46390 }, { "epoch": 17.040029379361, "grad_norm": 0.3391474187374115, "learning_rate": 1.4113831702699976e-06, "loss": 0.0112, "step": 46400 }, { "epoch": 17.043701799485863, "grad_norm": 0.31149476766586304, "learning_rate": 1.4035920464228526e-06, "loss": 0.0113, "step": 46410 }, { "epoch": 17.047374219610724, "grad_norm": 0.4281553626060486, "learning_rate": 1.3958221802955696e-06, "loss": 0.011, "step": 46420 }, { "epoch": 17.051046639735585, "grad_norm": 0.5978056788444519, "learning_rate": 1.3880735752869412e-06, "loss": 0.0125, "step": 46430 }, { "epoch": 17.05471905986045, "grad_norm": 0.23938612639904022, "learning_rate": 1.3803462347864893e-06, "loss": 0.0126, "step": 46440 }, { "epoch": 17.05839147998531, "grad_norm": 0.6316798329353333, "learning_rate": 1.372640162174399e-06, "loss": 0.0112, "step": 46450 }, { "epoch": 17.06206390011017, "grad_norm": 0.5478540062904358, "learning_rate": 1.3649553608215738e-06, "loss": 0.0125, "step": 46460 }, { "epoch": 17.065736320235036, "grad_norm": 0.27728110551834106, "learning_rate": 1.357291834089608e-06, "loss": 0.0114, "step": 46470 }, { "epoch": 17.069408740359897, "grad_norm": 0.2962017059326172, "learning_rate": 1.3496495853307921e-06, "loss": 0.0121, "step": 46480 }, { "epoch": 17.073081160484758, "grad_norm": 0.4493582844734192, "learning_rate": 1.3420286178880914e-06, "loss": 0.0117, "step": 46490 }, { "epoch": 17.076753580609623, "grad_norm": 0.3536125719547272, "learning_rate": 1.3344289350951943e-06, "loss": 0.0113, "step": 46500 }, { "epoch": 17.080426000734484, "grad_norm": 0.7260836362838745, "learning_rate": 1.3268505402764475e-06, "loss": 0.0119, "step": 46510 }, { "epoch": 17.084098420859345, "grad_norm": 0.30191725492477417, "learning_rate": 1.3192934367469045e-06, "loss": 0.0127, "step": 46520 }, { "epoch": 17.08777084098421, "grad_norm": 0.46771353483200073, "learning_rate": 1.3117576278122934e-06, "loss": 0.0115, "step": 46530 }, { "epoch": 17.09144326110907, "grad_norm": 0.2857609689235687, "learning_rate": 1.3042431167690382e-06, "loss": 0.0111, "step": 46540 }, { "epoch": 17.095115681233935, "grad_norm": 0.311416357755661, "learning_rate": 1.2967499069042377e-06, "loss": 0.0118, "step": 46550 }, { "epoch": 17.098788101358796, "grad_norm": 0.24022413790225983, "learning_rate": 1.2892780014956862e-06, "loss": 0.0132, "step": 46560 }, { "epoch": 17.102460521483657, "grad_norm": 0.23293761909008026, "learning_rate": 1.2818274038118306e-06, "loss": 0.0114, "step": 46570 }, { "epoch": 17.10613294160852, "grad_norm": 0.8018966913223267, "learning_rate": 1.2743981171118357e-06, "loss": 0.0109, "step": 46580 }, { "epoch": 17.109805361733383, "grad_norm": 0.5813621282577515, "learning_rate": 1.2669901446455024e-06, "loss": 0.0113, "step": 46590 }, { "epoch": 17.113477781858244, "grad_norm": 0.2728043496608734, "learning_rate": 1.2596034896533549e-06, "loss": 0.0112, "step": 46600 }, { "epoch": 17.117150201983108, "grad_norm": 0.5702455639839172, "learning_rate": 1.2522381553665474e-06, "loss": 0.0125, "step": 46610 }, { "epoch": 17.12082262210797, "grad_norm": 0.37612229585647583, "learning_rate": 1.2448941450069306e-06, "loss": 0.0112, "step": 46620 }, { "epoch": 17.12449504223283, "grad_norm": 0.2927033007144928, "learning_rate": 1.2375714617870293e-06, "loss": 0.011, "step": 46630 }, { "epoch": 17.128167462357695, "grad_norm": 0.25561293959617615, "learning_rate": 1.2302701089100366e-06, "loss": 0.0113, "step": 46640 }, { "epoch": 17.131839882482556, "grad_norm": 0.4314945936203003, "learning_rate": 1.2229900895698032e-06, "loss": 0.0115, "step": 46650 }, { "epoch": 17.135512302607417, "grad_norm": 0.6750689148902893, "learning_rate": 1.2157314069508596e-06, "loss": 0.0109, "step": 46660 }, { "epoch": 17.13918472273228, "grad_norm": 0.3816036880016327, "learning_rate": 1.2084940642284048e-06, "loss": 0.0133, "step": 46670 }, { "epoch": 17.142857142857142, "grad_norm": 0.41293197870254517, "learning_rate": 1.2012780645682952e-06, "loss": 0.0124, "step": 46680 }, { "epoch": 17.146529562982003, "grad_norm": 0.4046745300292969, "learning_rate": 1.1940834111270505e-06, "loss": 0.011, "step": 46690 }, { "epoch": 17.150201983106868, "grad_norm": 0.5718973278999329, "learning_rate": 1.1869101070518695e-06, "loss": 0.0104, "step": 46700 }, { "epoch": 17.15387440323173, "grad_norm": 0.2652706205844879, "learning_rate": 1.1797581554805813e-06, "loss": 0.0114, "step": 46710 }, { "epoch": 17.157546823356594, "grad_norm": 0.6551404595375061, "learning_rate": 1.1726275595417001e-06, "loss": 0.0115, "step": 46720 }, { "epoch": 17.161219243481455, "grad_norm": 0.3068683445453644, "learning_rate": 1.1655183223543976e-06, "loss": 0.0113, "step": 46730 }, { "epoch": 17.164891663606316, "grad_norm": 0.5296006798744202, "learning_rate": 1.1584304470284912e-06, "loss": 0.0118, "step": 46740 }, { "epoch": 17.16856408373118, "grad_norm": 0.18874691426753998, "learning_rate": 1.1513639366644513e-06, "loss": 0.0118, "step": 46750 }, { "epoch": 17.17223650385604, "grad_norm": 0.3515603840351105, "learning_rate": 1.1443187943534162e-06, "loss": 0.0112, "step": 46760 }, { "epoch": 17.175908923980902, "grad_norm": 0.2395690679550171, "learning_rate": 1.1372950231771707e-06, "loss": 0.0121, "step": 46770 }, { "epoch": 17.179581344105767, "grad_norm": 0.47318702936172485, "learning_rate": 1.1302926262081515e-06, "loss": 0.012, "step": 46780 }, { "epoch": 17.183253764230628, "grad_norm": 0.4315923750400543, "learning_rate": 1.1233116065094362e-06, "loss": 0.0113, "step": 46790 }, { "epoch": 17.18692618435549, "grad_norm": 1.1393027305603027, "learning_rate": 1.1163519671347767e-06, "loss": 0.0109, "step": 46800 }, { "epoch": 17.190598604480353, "grad_norm": 0.3862130343914032, "learning_rate": 1.1094137111285375e-06, "loss": 0.0113, "step": 46810 }, { "epoch": 17.194271024605214, "grad_norm": 0.4523804783821106, "learning_rate": 1.102496841525763e-06, "loss": 0.0105, "step": 46820 }, { "epoch": 17.197943444730075, "grad_norm": 0.3344443142414093, "learning_rate": 1.095601361352122e-06, "loss": 0.0143, "step": 46830 }, { "epoch": 17.20161586485494, "grad_norm": 0.45364874601364136, "learning_rate": 1.088727273623924e-06, "loss": 0.0123, "step": 46840 }, { "epoch": 17.2052882849798, "grad_norm": 0.2890119254589081, "learning_rate": 1.0818745813481356e-06, "loss": 0.0124, "step": 46850 }, { "epoch": 17.208960705104666, "grad_norm": 0.21076935529708862, "learning_rate": 1.0750432875223594e-06, "loss": 0.0118, "step": 46860 }, { "epoch": 17.212633125229527, "grad_norm": 0.19490967690944672, "learning_rate": 1.0682333951348323e-06, "loss": 0.0109, "step": 46870 }, { "epoch": 17.216305545354388, "grad_norm": 0.5655698180198669, "learning_rate": 1.0614449071644272e-06, "loss": 0.012, "step": 46880 }, { "epoch": 17.219977965479252, "grad_norm": 0.27291083335876465, "learning_rate": 1.054677826580669e-06, "loss": 0.011, "step": 46890 }, { "epoch": 17.223650385604113, "grad_norm": 0.4745767116546631, "learning_rate": 1.0479321563437062e-06, "loss": 0.0119, "step": 46900 }, { "epoch": 17.227322805728974, "grad_norm": 0.20383788645267487, "learning_rate": 1.0412078994043172e-06, "loss": 0.0107, "step": 46910 }, { "epoch": 17.23099522585384, "grad_norm": 0.4481552541255951, "learning_rate": 1.0345050587039273e-06, "loss": 0.015, "step": 46920 }, { "epoch": 17.2346676459787, "grad_norm": 0.35481125116348267, "learning_rate": 1.0278236371745854e-06, "loss": 0.0123, "step": 46930 }, { "epoch": 17.23834006610356, "grad_norm": 0.34159260988235474, "learning_rate": 1.02116363773897e-06, "loss": 0.0129, "step": 46940 }, { "epoch": 17.242012486228425, "grad_norm": 0.3125540614128113, "learning_rate": 1.0145250633103954e-06, "loss": 0.0124, "step": 46950 }, { "epoch": 17.245684906353286, "grad_norm": 0.8621304035186768, "learning_rate": 1.007907916792794e-06, "loss": 0.0115, "step": 46960 }, { "epoch": 17.249357326478147, "grad_norm": 0.4102981984615326, "learning_rate": 1.001312201080734e-06, "loss": 0.0111, "step": 46970 }, { "epoch": 17.253029746603012, "grad_norm": 0.9469910264015198, "learning_rate": 9.947379190594075e-07, "loss": 0.0106, "step": 46980 }, { "epoch": 17.256702166727873, "grad_norm": 0.34846627712249756, "learning_rate": 9.88185073604625e-07, "loss": 0.0119, "step": 46990 }, { "epoch": 17.260374586852738, "grad_norm": 0.20796088874340057, "learning_rate": 9.816536675828214e-07, "loss": 0.0124, "step": 47000 }, { "epoch": 17.2640470069776, "grad_norm": 0.34051769971847534, "learning_rate": 9.751437038510615e-07, "loss": 0.0114, "step": 47010 }, { "epoch": 17.26771942710246, "grad_norm": 0.6502571105957031, "learning_rate": 9.686551852570225e-07, "loss": 0.0121, "step": 47020 }, { "epoch": 17.271391847227324, "grad_norm": 0.4530430734157562, "learning_rate": 9.621881146390067e-07, "loss": 0.0107, "step": 47030 }, { "epoch": 17.275064267352185, "grad_norm": 0.3674444854259491, "learning_rate": 9.557424948259175e-07, "loss": 0.0121, "step": 47040 }, { "epoch": 17.278736687477046, "grad_norm": 0.450968861579895, "learning_rate": 9.493183286373053e-07, "loss": 0.0118, "step": 47050 }, { "epoch": 17.28240910760191, "grad_norm": 0.21523422002792358, "learning_rate": 9.429156188833111e-07, "loss": 0.0113, "step": 47060 }, { "epoch": 17.286081527726772, "grad_norm": 0.3301832675933838, "learning_rate": 9.365343683646944e-07, "loss": 0.0127, "step": 47070 }, { "epoch": 17.289753947851633, "grad_norm": 0.373295396566391, "learning_rate": 9.301745798728334e-07, "loss": 0.013, "step": 47080 }, { "epoch": 17.293426367976497, "grad_norm": 0.4803923964500427, "learning_rate": 9.238362561897141e-07, "loss": 0.0133, "step": 47090 }, { "epoch": 17.29709878810136, "grad_norm": 0.2948378622531891, "learning_rate": 9.17519400087935e-07, "loss": 0.0113, "step": 47100 }, { "epoch": 17.30077120822622, "grad_norm": 0.5565650463104248, "learning_rate": 9.112240143307138e-07, "loss": 0.0134, "step": 47110 }, { "epoch": 17.304443628351084, "grad_norm": 0.16711224615573883, "learning_rate": 9.049501016718475e-07, "loss": 0.0124, "step": 47120 }, { "epoch": 17.308116048475945, "grad_norm": 0.5382400155067444, "learning_rate": 8.986976648557743e-07, "loss": 0.0137, "step": 47130 }, { "epoch": 17.31178846860081, "grad_norm": 0.6725248098373413, "learning_rate": 8.924667066175119e-07, "loss": 0.0118, "step": 47140 }, { "epoch": 17.31546088872567, "grad_norm": 1.0623077154159546, "learning_rate": 8.86257229682702e-07, "loss": 0.0112, "step": 47150 }, { "epoch": 17.31913330885053, "grad_norm": 0.3887804448604584, "learning_rate": 8.800692367675722e-07, "loss": 0.0116, "step": 47160 }, { "epoch": 17.322805728975396, "grad_norm": 0.17788612842559814, "learning_rate": 8.739027305789683e-07, "loss": 0.0116, "step": 47170 }, { "epoch": 17.326478149100257, "grad_norm": 0.3781226873397827, "learning_rate": 8.677577138143267e-07, "loss": 0.0116, "step": 47180 }, { "epoch": 17.33015056922512, "grad_norm": 0.35088321566581726, "learning_rate": 8.616341891616864e-07, "loss": 0.0119, "step": 47190 }, { "epoch": 17.333822989349983, "grad_norm": 0.28273236751556396, "learning_rate": 8.555321592996824e-07, "loss": 0.0105, "step": 47200 }, { "epoch": 17.337495409474844, "grad_norm": 0.5173104405403137, "learning_rate": 8.494516268975572e-07, "loss": 0.0111, "step": 47210 }, { "epoch": 17.341167829599705, "grad_norm": 0.23601381480693817, "learning_rate": 8.433925946151333e-07, "loss": 0.0111, "step": 47220 }, { "epoch": 17.34484024972457, "grad_norm": 0.6396370530128479, "learning_rate": 8.373550651028406e-07, "loss": 0.0127, "step": 47230 }, { "epoch": 17.34851266984943, "grad_norm": 0.40217119455337524, "learning_rate": 8.313390410017052e-07, "loss": 0.0109, "step": 47240 }, { "epoch": 17.35218508997429, "grad_norm": 0.7920993566513062, "learning_rate": 8.253445249433334e-07, "loss": 0.011, "step": 47250 }, { "epoch": 17.355857510099156, "grad_norm": 0.27022066712379456, "learning_rate": 8.193715195499329e-07, "loss": 0.0111, "step": 47260 }, { "epoch": 17.359529930224017, "grad_norm": 0.3249916136264801, "learning_rate": 8.134200274342918e-07, "loss": 0.0112, "step": 47270 }, { "epoch": 17.363202350348878, "grad_norm": 0.30095410346984863, "learning_rate": 8.074900511998107e-07, "loss": 0.0112, "step": 47280 }, { "epoch": 17.366874770473743, "grad_norm": 0.25780966877937317, "learning_rate": 8.01581593440448e-07, "loss": 0.0107, "step": 47290 }, { "epoch": 17.370547190598604, "grad_norm": 0.2514718472957611, "learning_rate": 7.956946567407753e-07, "loss": 0.0117, "step": 47300 }, { "epoch": 17.37421961072347, "grad_norm": 0.29706400632858276, "learning_rate": 7.898292436759324e-07, "loss": 0.0128, "step": 47310 }, { "epoch": 17.37789203084833, "grad_norm": 0.3193608224391937, "learning_rate": 7.839853568116506e-07, "loss": 0.0144, "step": 47320 }, { "epoch": 17.38156445097319, "grad_norm": 0.3209393322467804, "learning_rate": 7.78162998704246e-07, "loss": 0.0121, "step": 47330 }, { "epoch": 17.385236871098055, "grad_norm": 0.6927816271781921, "learning_rate": 7.723621719006147e-07, "loss": 0.0113, "step": 47340 }, { "epoch": 17.388909291222916, "grad_norm": 0.5102827548980713, "learning_rate": 7.665828789382323e-07, "loss": 0.0106, "step": 47350 }, { "epoch": 17.392581711347777, "grad_norm": 0.23443740606307983, "learning_rate": 7.608251223451602e-07, "loss": 0.0105, "step": 47360 }, { "epoch": 17.39625413147264, "grad_norm": 0.39586880803108215, "learning_rate": 7.550889046400444e-07, "loss": 0.0108, "step": 47370 }, { "epoch": 17.399926551597503, "grad_norm": 0.4259014427661896, "learning_rate": 7.493742283320893e-07, "loss": 0.0112, "step": 47380 }, { "epoch": 17.403598971722364, "grad_norm": 0.4238590896129608, "learning_rate": 7.43681095921095e-07, "loss": 0.0108, "step": 47390 }, { "epoch": 17.407271391847228, "grad_norm": 0.36507558822631836, "learning_rate": 7.380095098974305e-07, "loss": 0.0121, "step": 47400 }, { "epoch": 17.41094381197209, "grad_norm": 0.38603997230529785, "learning_rate": 7.323594727420446e-07, "loss": 0.0109, "step": 47410 }, { "epoch": 17.41461623209695, "grad_norm": 0.5605124831199646, "learning_rate": 7.267309869264438e-07, "loss": 0.0114, "step": 47420 }, { "epoch": 17.418288652221815, "grad_norm": 0.3762330412864685, "learning_rate": 7.211240549127252e-07, "loss": 0.0153, "step": 47430 }, { "epoch": 17.421961072346676, "grad_norm": 0.38411539793014526, "learning_rate": 7.155386791535546e-07, "loss": 0.0105, "step": 47440 }, { "epoch": 17.42563349247154, "grad_norm": 0.3806089162826538, "learning_rate": 7.09974862092161e-07, "loss": 0.0113, "step": 47450 }, { "epoch": 17.4293059125964, "grad_norm": 0.7976626753807068, "learning_rate": 7.044326061623529e-07, "loss": 0.0128, "step": 47460 }, { "epoch": 17.432978332721262, "grad_norm": 0.5019456148147583, "learning_rate": 6.989119137884914e-07, "loss": 0.0119, "step": 47470 }, { "epoch": 17.436650752846127, "grad_norm": 0.3740284740924835, "learning_rate": 6.934127873855223e-07, "loss": 0.0118, "step": 47480 }, { "epoch": 17.440323172970988, "grad_norm": 0.2638052999973297, "learning_rate": 6.879352293589492e-07, "loss": 0.0115, "step": 47490 }, { "epoch": 17.44399559309585, "grad_norm": 0.5289780497550964, "learning_rate": 6.824792421048388e-07, "loss": 0.0109, "step": 47500 }, { "epoch": 17.447668013220714, "grad_norm": 0.3746584951877594, "learning_rate": 6.770448280098319e-07, "loss": 0.0137, "step": 47510 }, { "epoch": 17.451340433345575, "grad_norm": 0.3989318013191223, "learning_rate": 6.716319894511158e-07, "loss": 0.0121, "step": 47520 }, { "epoch": 17.455012853470436, "grad_norm": 0.402434766292572, "learning_rate": 6.662407287964634e-07, "loss": 0.0103, "step": 47530 }, { "epoch": 17.4586852735953, "grad_norm": 0.2768719494342804, "learning_rate": 6.608710484041824e-07, "loss": 0.0117, "step": 47540 }, { "epoch": 17.46235769372016, "grad_norm": 0.3377125561237335, "learning_rate": 6.555229506231608e-07, "loss": 0.0126, "step": 47550 }, { "epoch": 17.466030113845022, "grad_norm": 0.5194752216339111, "learning_rate": 6.501964377928382e-07, "loss": 0.0113, "step": 47560 }, { "epoch": 17.469702533969887, "grad_norm": 0.5857775807380676, "learning_rate": 6.448915122432064e-07, "loss": 0.012, "step": 47570 }, { "epoch": 17.473374954094748, "grad_norm": 0.4614880681037903, "learning_rate": 6.396081762948258e-07, "loss": 0.0113, "step": 47580 }, { "epoch": 17.477047374219612, "grad_norm": 0.3030819296836853, "learning_rate": 6.343464322588033e-07, "loss": 0.0111, "step": 47590 }, { "epoch": 17.480719794344473, "grad_norm": 0.41073542833328247, "learning_rate": 6.291062824368143e-07, "loss": 0.0117, "step": 47600 }, { "epoch": 17.484392214469334, "grad_norm": 0.20317105948925018, "learning_rate": 6.238877291210587e-07, "loss": 0.0112, "step": 47610 }, { "epoch": 17.4880646345942, "grad_norm": 0.43552255630493164, "learning_rate": 6.186907745943271e-07, "loss": 0.013, "step": 47620 }, { "epoch": 17.49173705471906, "grad_norm": 0.7154546976089478, "learning_rate": 6.135154211299399e-07, "loss": 0.0121, "step": 47630 }, { "epoch": 17.49540947484392, "grad_norm": 0.5263808965682983, "learning_rate": 6.083616709917639e-07, "loss": 0.0113, "step": 47640 }, { "epoch": 17.499081894968786, "grad_norm": 0.3062903583049774, "learning_rate": 6.032295264342403e-07, "loss": 0.0121, "step": 47650 }, { "epoch": 17.502754315093647, "grad_norm": 1.3024773597717285, "learning_rate": 5.981189897023287e-07, "loss": 0.0113, "step": 47660 }, { "epoch": 17.506426735218508, "grad_norm": 0.48348864912986755, "learning_rate": 5.930300630315522e-07, "loss": 0.0114, "step": 47670 }, { "epoch": 17.510099155343372, "grad_norm": 0.5254379510879517, "learning_rate": 5.879627486479911e-07, "loss": 0.0109, "step": 47680 }, { "epoch": 17.513771575468233, "grad_norm": 0.3207124173641205, "learning_rate": 5.829170487682556e-07, "loss": 0.0112, "step": 47690 }, { "epoch": 17.517443995593094, "grad_norm": 0.3249039649963379, "learning_rate": 5.778929655995025e-07, "loss": 0.0123, "step": 47700 }, { "epoch": 17.52111641571796, "grad_norm": 0.39842185378074646, "learning_rate": 5.728905013394403e-07, "loss": 0.0134, "step": 47710 }, { "epoch": 17.52478883584282, "grad_norm": 0.3675837218761444, "learning_rate": 5.679096581763188e-07, "loss": 0.0108, "step": 47720 }, { "epoch": 17.528461255967684, "grad_norm": 0.3927013576030731, "learning_rate": 5.629504382889173e-07, "loss": 0.0112, "step": 47730 }, { "epoch": 17.532133676092545, "grad_norm": 0.23430463671684265, "learning_rate": 5.580128438465837e-07, "loss": 0.0116, "step": 47740 }, { "epoch": 17.535806096217406, "grad_norm": 0.6624082326889038, "learning_rate": 5.530968770091737e-07, "loss": 0.0115, "step": 47750 }, { "epoch": 17.53947851634227, "grad_norm": 0.5279398560523987, "learning_rate": 5.482025399271063e-07, "loss": 0.0108, "step": 47760 }, { "epoch": 17.543150936467132, "grad_norm": 0.48203837871551514, "learning_rate": 5.433298347413296e-07, "loss": 0.011, "step": 47770 }, { "epoch": 17.546823356591993, "grad_norm": 0.38618311285972595, "learning_rate": 5.384787635833332e-07, "loss": 0.011, "step": 47780 }, { "epoch": 17.550495776716858, "grad_norm": 1.2961318492889404, "learning_rate": 5.33649328575142e-07, "loss": 0.0121, "step": 47790 }, { "epoch": 17.55416819684172, "grad_norm": 0.5386044979095459, "learning_rate": 5.288415318292994e-07, "loss": 0.0121, "step": 47800 }, { "epoch": 17.55784061696658, "grad_norm": 0.4462156295776367, "learning_rate": 5.24055375448923e-07, "loss": 0.0109, "step": 47810 }, { "epoch": 17.561513037091444, "grad_norm": 0.6235929131507874, "learning_rate": 5.192908615276215e-07, "loss": 0.0108, "step": 47820 }, { "epoch": 17.565185457216305, "grad_norm": 0.17957504093647003, "learning_rate": 5.145479921495556e-07, "loss": 0.011, "step": 47830 }, { "epoch": 17.568857877341166, "grad_norm": 0.16772963106632233, "learning_rate": 5.098267693894321e-07, "loss": 0.0106, "step": 47840 }, { "epoch": 17.57253029746603, "grad_norm": 0.35352200269699097, "learning_rate": 5.051271953124603e-07, "loss": 0.0113, "step": 47850 }, { "epoch": 17.576202717590892, "grad_norm": 0.2060849368572235, "learning_rate": 5.004492719744014e-07, "loss": 0.012, "step": 47860 }, { "epoch": 17.579875137715753, "grad_norm": 0.4183388650417328, "learning_rate": 4.957930014215351e-07, "loss": 0.0111, "step": 47870 }, { "epoch": 17.583547557840618, "grad_norm": 0.3116723895072937, "learning_rate": 4.91158385690671e-07, "loss": 0.0122, "step": 47880 }, { "epoch": 17.58721997796548, "grad_norm": 0.2460365891456604, "learning_rate": 4.865454268091485e-07, "loss": 0.0115, "step": 47890 }, { "epoch": 17.590892398090343, "grad_norm": 0.4565396308898926, "learning_rate": 4.819541267948313e-07, "loss": 0.0124, "step": 47900 }, { "epoch": 17.594564818215204, "grad_norm": 0.6383718252182007, "learning_rate": 4.773844876561129e-07, "loss": 0.0124, "step": 47910 }, { "epoch": 17.598237238340065, "grad_norm": 0.30401813983917236, "learning_rate": 4.7283651139191086e-07, "loss": 0.0112, "step": 47920 }, { "epoch": 17.60190965846493, "grad_norm": 0.47268274426460266, "learning_rate": 4.6831019999165617e-07, "loss": 0.0142, "step": 47930 }, { "epoch": 17.60558207858979, "grad_norm": 0.16912227869033813, "learning_rate": 4.6380555543532045e-07, "loss": 0.0115, "step": 47940 }, { "epoch": 17.60925449871465, "grad_norm": 0.5298471450805664, "learning_rate": 4.5932257969338313e-07, "loss": 0.0116, "step": 47950 }, { "epoch": 17.612926918839516, "grad_norm": 0.2988956868648529, "learning_rate": 4.5486127472684215e-07, "loss": 0.0107, "step": 47960 }, { "epoch": 17.616599338964377, "grad_norm": 0.537384033203125, "learning_rate": 4.5042164248724205e-07, "loss": 0.0127, "step": 47970 }, { "epoch": 17.62027175908924, "grad_norm": 0.7278279066085815, "learning_rate": 4.4600368491661826e-07, "loss": 0.0103, "step": 47980 }, { "epoch": 17.623944179214103, "grad_norm": 0.33417242765426636, "learning_rate": 4.4160740394753595e-07, "loss": 0.0113, "step": 47990 }, { "epoch": 17.627616599338964, "grad_norm": 0.2935248017311096, "learning_rate": 4.3723280150307357e-07, "loss": 0.0114, "step": 48000 }, { "epoch": 17.631289019463825, "grad_norm": 0.4484615623950958, "learning_rate": 4.3287987949683917e-07, "loss": 0.0118, "step": 48010 }, { "epoch": 17.63496143958869, "grad_norm": 0.5291367173194885, "learning_rate": 4.285486398329486e-07, "loss": 0.0109, "step": 48020 }, { "epoch": 17.63863385971355, "grad_norm": 0.30211564898490906, "learning_rate": 4.242390844060251e-07, "loss": 0.0124, "step": 48030 }, { "epoch": 17.642306279838415, "grad_norm": 0.23004834353923798, "learning_rate": 4.1995121510122193e-07, "loss": 0.0115, "step": 48040 }, { "epoch": 17.645978699963276, "grad_norm": 0.48562973737716675, "learning_rate": 4.156850337941942e-07, "loss": 0.0106, "step": 48050 }, { "epoch": 17.649651120088137, "grad_norm": 0.38103076815605164, "learning_rate": 4.114405423511214e-07, "loss": 0.0126, "step": 48060 }, { "epoch": 17.653323540213, "grad_norm": 0.6181902289390564, "learning_rate": 4.072177426286794e-07, "loss": 0.0109, "step": 48070 }, { "epoch": 17.656995960337863, "grad_norm": 0.4015485644340515, "learning_rate": 4.0301663647407396e-07, "loss": 0.0119, "step": 48080 }, { "epoch": 17.660668380462724, "grad_norm": 0.2604920268058777, "learning_rate": 3.988372257249962e-07, "loss": 0.0129, "step": 48090 }, { "epoch": 17.66434080058759, "grad_norm": 0.661689281463623, "learning_rate": 3.946795122096836e-07, "loss": 0.0103, "step": 48100 }, { "epoch": 17.66801322071245, "grad_norm": 0.6005699634552002, "learning_rate": 3.9054349774684254e-07, "loss": 0.0114, "step": 48110 }, { "epoch": 17.67168564083731, "grad_norm": 0.5450016856193542, "learning_rate": 3.864291841457146e-07, "loss": 0.0116, "step": 48120 }, { "epoch": 17.675358060962175, "grad_norm": 1.344825267791748, "learning_rate": 3.8233657320604355e-07, "loss": 0.013, "step": 48130 }, { "epoch": 17.679030481087036, "grad_norm": 0.46780064702033997, "learning_rate": 3.782656667180695e-07, "loss": 0.0107, "step": 48140 }, { "epoch": 17.682702901211897, "grad_norm": 0.5434199571609497, "learning_rate": 3.742164664625458e-07, "loss": 0.0114, "step": 48150 }, { "epoch": 17.68637532133676, "grad_norm": 0.3929862678050995, "learning_rate": 3.7018897421073337e-07, "loss": 0.0114, "step": 48160 }, { "epoch": 17.690047741461623, "grad_norm": 0.21990276873111725, "learning_rate": 3.6618319172439517e-07, "loss": 0.0114, "step": 48170 }, { "epoch": 17.693720161586487, "grad_norm": 0.2652556598186493, "learning_rate": 3.6219912075579065e-07, "loss": 0.011, "step": 48180 }, { "epoch": 17.697392581711348, "grad_norm": 0.3685949742794037, "learning_rate": 3.5823676304769793e-07, "loss": 0.0117, "step": 48190 }, { "epoch": 17.70106500183621, "grad_norm": 0.7510524988174438, "learning_rate": 3.5429612033338057e-07, "loss": 0.0136, "step": 48200 }, { "epoch": 17.704737421961074, "grad_norm": 0.41572314500808716, "learning_rate": 3.503771943366041e-07, "loss": 0.0101, "step": 48210 }, { "epoch": 17.708409842085935, "grad_norm": 0.7415062189102173, "learning_rate": 3.464799867716473e-07, "loss": 0.014, "step": 48220 }, { "epoch": 17.712082262210796, "grad_norm": 0.2650798261165619, "learning_rate": 3.4260449934327977e-07, "loss": 0.0111, "step": 48230 }, { "epoch": 17.71575468233566, "grad_norm": 0.8469932079315186, "learning_rate": 3.3875073374677326e-07, "loss": 0.0123, "step": 48240 }, { "epoch": 17.71942710246052, "grad_norm": 0.5207273960113525, "learning_rate": 3.3491869166789035e-07, "loss": 0.0112, "step": 48250 }, { "epoch": 17.723099522585382, "grad_norm": 0.6603397130966187, "learning_rate": 3.3110837478290137e-07, "loss": 0.0112, "step": 48260 }, { "epoch": 17.726771942710247, "grad_norm": 0.5760461091995239, "learning_rate": 3.273197847585674e-07, "loss": 0.012, "step": 48270 }, { "epoch": 17.730444362835108, "grad_norm": 1.0855072736740112, "learning_rate": 3.2355292325214617e-07, "loss": 0.0109, "step": 48280 }, { "epoch": 17.73411678295997, "grad_norm": 0.3107012212276459, "learning_rate": 3.198077919113862e-07, "loss": 0.011, "step": 48290 }, { "epoch": 17.737789203084834, "grad_norm": 0.2176312953233719, "learning_rate": 3.160843923745438e-07, "loss": 0.0108, "step": 48300 }, { "epoch": 17.741461623209695, "grad_norm": 0.5574002861976624, "learning_rate": 3.123827262703549e-07, "loss": 0.014, "step": 48310 }, { "epoch": 17.74513404333456, "grad_norm": 0.465326189994812, "learning_rate": 3.0870279521805765e-07, "loss": 0.0116, "step": 48320 }, { "epoch": 17.74880646345942, "grad_norm": 0.16656765341758728, "learning_rate": 3.0504460082737553e-07, "loss": 0.0108, "step": 48330 }, { "epoch": 17.75247888358428, "grad_norm": 0.39942610263824463, "learning_rate": 3.0140814469852286e-07, "loss": 0.0121, "step": 48340 }, { "epoch": 17.756151303709146, "grad_norm": 0.25381606817245483, "learning_rate": 2.9779342842221614e-07, "loss": 0.0113, "step": 48350 }, { "epoch": 17.759823723834007, "grad_norm": 0.629497230052948, "learning_rate": 2.942004535796572e-07, "loss": 0.0131, "step": 48360 }, { "epoch": 17.763496143958868, "grad_norm": 0.34194454550743103, "learning_rate": 2.906292217425277e-07, "loss": 0.0108, "step": 48370 }, { "epoch": 17.767168564083732, "grad_norm": 0.3199201822280884, "learning_rate": 2.8707973447300584e-07, "loss": 0.0104, "step": 48380 }, { "epoch": 17.770840984208593, "grad_norm": 0.7033660411834717, "learning_rate": 2.8355199332376627e-07, "loss": 0.0126, "step": 48390 }, { "epoch": 17.774513404333455, "grad_norm": 0.39527493715286255, "learning_rate": 2.8004599983795785e-07, "loss": 0.0133, "step": 48400 }, { "epoch": 17.77818582445832, "grad_norm": 0.41424864530563354, "learning_rate": 2.765617555492206e-07, "loss": 0.0114, "step": 48410 }, { "epoch": 17.78185824458318, "grad_norm": 0.1838262379169464, "learning_rate": 2.7309926198168524e-07, "loss": 0.0124, "step": 48420 }, { "epoch": 17.78553066470804, "grad_norm": 0.4973922073841095, "learning_rate": 2.696585206499569e-07, "loss": 0.0113, "step": 48430 }, { "epoch": 17.789203084832906, "grad_norm": 0.4492335319519043, "learning_rate": 2.662395330591372e-07, "loss": 0.0119, "step": 48440 }, { "epoch": 17.792875504957767, "grad_norm": 0.31726738810539246, "learning_rate": 2.628423007048075e-07, "loss": 0.0112, "step": 48450 }, { "epoch": 17.796547925082628, "grad_norm": 0.21399500966072083, "learning_rate": 2.594668250730292e-07, "loss": 0.0111, "step": 48460 }, { "epoch": 17.800220345207492, "grad_norm": 0.4278280735015869, "learning_rate": 2.5611310764035445e-07, "loss": 0.011, "step": 48470 }, { "epoch": 17.803892765332353, "grad_norm": 0.45058658719062805, "learning_rate": 2.5278114987380973e-07, "loss": 0.0108, "step": 48480 }, { "epoch": 17.807565185457218, "grad_norm": 0.41450396180152893, "learning_rate": 2.494709532309125e-07, "loss": 0.0113, "step": 48490 }, { "epoch": 17.81123760558208, "grad_norm": 0.5740877389907837, "learning_rate": 2.4618251915964896e-07, "loss": 0.0119, "step": 48500 }, { "epoch": 17.81491002570694, "grad_norm": 0.2288929969072342, "learning_rate": 2.429158490984962e-07, "loss": 0.0108, "step": 48510 }, { "epoch": 17.818582445831805, "grad_norm": 0.2492036074399948, "learning_rate": 2.396709444764056e-07, "loss": 0.0109, "step": 48520 }, { "epoch": 17.822254865956666, "grad_norm": 0.48701199889183044, "learning_rate": 2.364478067128084e-07, "loss": 0.0112, "step": 48530 }, { "epoch": 17.825927286081527, "grad_norm": 0.29829442501068115, "learning_rate": 2.332464372176102e-07, "loss": 0.0116, "step": 48540 }, { "epoch": 17.82959970620639, "grad_norm": 0.38674208521842957, "learning_rate": 2.300668373912074e-07, "loss": 0.0113, "step": 48550 }, { "epoch": 17.833272126331252, "grad_norm": 0.7775603532791138, "learning_rate": 2.269090086244541e-07, "loss": 0.0112, "step": 48560 }, { "epoch": 17.836944546456113, "grad_norm": 0.2602197527885437, "learning_rate": 2.237729522987009e-07, "loss": 0.0115, "step": 48570 }, { "epoch": 17.840616966580978, "grad_norm": 0.4185846745967865, "learning_rate": 2.2065866978576156e-07, "loss": 0.0119, "step": 48580 }, { "epoch": 17.84428938670584, "grad_norm": 0.20782782137393951, "learning_rate": 2.1756616244792972e-07, "loss": 0.0128, "step": 48590 }, { "epoch": 17.8479618068307, "grad_norm": 0.43341580033302307, "learning_rate": 2.1449543163797324e-07, "loss": 0.0115, "step": 48600 }, { "epoch": 17.851634226955564, "grad_norm": 0.29304590821266174, "learning_rate": 2.1144647869913992e-07, "loss": 0.0138, "step": 48610 }, { "epoch": 17.855306647080425, "grad_norm": 0.38838231563568115, "learning_rate": 2.0841930496513507e-07, "loss": 0.0117, "step": 48620 }, { "epoch": 17.85897906720529, "grad_norm": 0.31232959032058716, "learning_rate": 2.0541391176014945e-07, "loss": 0.0115, "step": 48630 }, { "epoch": 17.86265148733015, "grad_norm": 0.363089382648468, "learning_rate": 2.0243030039884814e-07, "loss": 0.0115, "step": 48640 }, { "epoch": 17.866323907455012, "grad_norm": 0.7331089973449707, "learning_rate": 1.994684721863649e-07, "loss": 0.0109, "step": 48650 }, { "epoch": 17.869996327579877, "grad_norm": 0.2235995978116989, "learning_rate": 1.9652842841829667e-07, "loss": 0.0112, "step": 48660 }, { "epoch": 17.873668747704738, "grad_norm": 0.3580150604248047, "learning_rate": 1.9361017038072583e-07, "loss": 0.0114, "step": 48670 }, { "epoch": 17.8773411678296, "grad_norm": 0.4178679883480072, "learning_rate": 1.907136993501979e-07, "loss": 0.0122, "step": 48680 }, { "epoch": 17.881013587954463, "grad_norm": 0.3092726767063141, "learning_rate": 1.878390165937216e-07, "loss": 0.0115, "step": 48690 }, { "epoch": 17.884686008079324, "grad_norm": 0.7169584631919861, "learning_rate": 1.849861233687855e-07, "loss": 0.0121, "step": 48700 }, { "epoch": 17.888358428204185, "grad_norm": 0.2648726999759674, "learning_rate": 1.8215502092334135e-07, "loss": 0.0114, "step": 48710 }, { "epoch": 17.89203084832905, "grad_norm": 1.1296603679656982, "learning_rate": 1.7934571049580963e-07, "loss": 0.0121, "step": 48720 }, { "epoch": 17.89570326845391, "grad_norm": 0.22433724999427795, "learning_rate": 1.7655819331507394e-07, "loss": 0.0109, "step": 48730 }, { "epoch": 17.899375688578772, "grad_norm": 0.3536888062953949, "learning_rate": 1.7379247060049785e-07, "loss": 0.0114, "step": 48740 }, { "epoch": 17.903048108703636, "grad_norm": 0.5598531365394592, "learning_rate": 1.7104854356189693e-07, "loss": 0.0108, "step": 48750 }, { "epoch": 17.906720528828497, "grad_norm": 0.2476821392774582, "learning_rate": 1.6832641339955546e-07, "loss": 0.0115, "step": 48760 }, { "epoch": 17.910392948953362, "grad_norm": 0.3895569443702698, "learning_rate": 1.6562608130422653e-07, "loss": 0.0106, "step": 48770 }, { "epoch": 17.914065369078223, "grad_norm": 0.34442007541656494, "learning_rate": 1.6294754845713745e-07, "loss": 0.0117, "step": 48780 }, { "epoch": 17.917737789203084, "grad_norm": 0.40573424100875854, "learning_rate": 1.60290816029951e-07, "loss": 0.011, "step": 48790 }, { "epoch": 17.92141020932795, "grad_norm": 0.47384464740753174, "learning_rate": 1.5765588518482643e-07, "loss": 0.0133, "step": 48800 }, { "epoch": 17.92508262945281, "grad_norm": 0.4135729968547821, "learning_rate": 1.5504275707436956e-07, "loss": 0.0134, "step": 48810 }, { "epoch": 17.92875504957767, "grad_norm": 0.22279681265354156, "learning_rate": 1.524514328416493e-07, "loss": 0.0121, "step": 48820 }, { "epoch": 17.932427469702535, "grad_norm": 0.2515946626663208, "learning_rate": 1.498819136201979e-07, "loss": 0.0113, "step": 48830 }, { "epoch": 17.936099889827396, "grad_norm": 0.4338081181049347, "learning_rate": 1.4733420053401616e-07, "loss": 0.0109, "step": 48840 }, { "epoch": 17.939772309952257, "grad_norm": 0.9003645181655884, "learning_rate": 1.4480829469755153e-07, "loss": 0.011, "step": 48850 }, { "epoch": 17.943444730077122, "grad_norm": 1.2310829162597656, "learning_rate": 1.4230419721572574e-07, "loss": 0.0122, "step": 48860 }, { "epoch": 17.947117150201983, "grad_norm": 0.4803580343723297, "learning_rate": 1.3982190918391812e-07, "loss": 0.0127, "step": 48870 }, { "epoch": 17.950789570326844, "grad_norm": 0.612623929977417, "learning_rate": 1.3736143168796012e-07, "loss": 0.0109, "step": 48880 }, { "epoch": 17.95446199045171, "grad_norm": 0.3081841766834259, "learning_rate": 1.349227658041574e-07, "loss": 0.0139, "step": 48890 }, { "epoch": 17.95813441057657, "grad_norm": 0.5545953512191772, "learning_rate": 1.325059125992567e-07, "loss": 0.0121, "step": 48900 }, { "epoch": 17.961806830701434, "grad_norm": 0.39176151156425476, "learning_rate": 1.3011087313047344e-07, "loss": 0.011, "step": 48910 }, { "epoch": 17.965479250826295, "grad_norm": 0.6753301620483398, "learning_rate": 1.2773764844548064e-07, "loss": 0.0111, "step": 48920 }, { "epoch": 17.969151670951156, "grad_norm": 0.5216216444969177, "learning_rate": 1.2538623958240903e-07, "loss": 0.0113, "step": 48930 }, { "epoch": 17.97282409107602, "grad_norm": 0.4013872444629669, "learning_rate": 1.2305664756984137e-07, "loss": 0.0118, "step": 48940 }, { "epoch": 17.97649651120088, "grad_norm": 0.8564138412475586, "learning_rate": 1.2074887342681806e-07, "loss": 0.0114, "step": 48950 }, { "epoch": 17.980168931325743, "grad_norm": 0.3486960828304291, "learning_rate": 1.1846291816284827e-07, "loss": 0.0114, "step": 48960 }, { "epoch": 17.983841351450607, "grad_norm": 0.599341094493866, "learning_rate": 1.1619878277787654e-07, "loss": 0.0105, "step": 48970 }, { "epoch": 17.98751377157547, "grad_norm": 0.2919934093952179, "learning_rate": 1.1395646826231621e-07, "loss": 0.0112, "step": 48980 }, { "epoch": 17.99118619170033, "grad_norm": 0.5119036436080933, "learning_rate": 1.1173597559703263e-07, "loss": 0.0126, "step": 48990 }, { "epoch": 17.994858611825194, "grad_norm": 0.5569196343421936, "learning_rate": 1.0953730575334331e-07, "loss": 0.0113, "step": 49000 }, { "epoch": 17.998531031950055, "grad_norm": 0.35803431272506714, "learning_rate": 1.0736045969301778e-07, "loss": 0.0124, "step": 49010 }, { "epoch": 18.002203452074916, "grad_norm": 0.36809083819389343, "learning_rate": 1.0520543836829433e-07, "loss": 0.0111, "step": 49020 }, { "epoch": 18.00587587219978, "grad_norm": 0.2993401288986206, "learning_rate": 1.0307224272183558e-07, "loss": 0.0128, "step": 49030 }, { "epoch": 18.00954829232464, "grad_norm": 0.5783461332321167, "learning_rate": 1.0096087368678952e-07, "loss": 0.0108, "step": 49040 }, { "epoch": 18.013220712449503, "grad_norm": 0.24067340791225433, "learning_rate": 9.88713321867285e-08, "loss": 0.0121, "step": 49050 }, { "epoch": 18.016893132574367, "grad_norm": 0.8678693175315857, "learning_rate": 9.680361913569913e-08, "loss": 0.0106, "step": 49060 }, { "epoch": 18.020565552699228, "grad_norm": 0.38622793555259705, "learning_rate": 9.475773543818344e-08, "loss": 0.0107, "step": 49070 }, { "epoch": 18.024237972824093, "grad_norm": 0.4137016236782074, "learning_rate": 9.273368198911559e-08, "loss": 0.0114, "step": 49080 }, { "epoch": 18.027910392948954, "grad_norm": 0.380018025636673, "learning_rate": 9.07314596738984e-08, "loss": 0.0113, "step": 49090 }, { "epoch": 18.031582813073815, "grad_norm": 0.3101377487182617, "learning_rate": 8.875106936836463e-08, "loss": 0.0119, "step": 49100 }, { "epoch": 18.03525523319868, "grad_norm": 0.5387400388717651, "learning_rate": 8.679251193879911e-08, "loss": 0.0108, "step": 49110 }, { "epoch": 18.03892765332354, "grad_norm": 0.42667528986930847, "learning_rate": 8.485578824194984e-08, "loss": 0.0136, "step": 49120 }, { "epoch": 18.0426000734484, "grad_norm": 1.3308360576629639, "learning_rate": 8.294089912500024e-08, "loss": 0.0113, "step": 49130 }, { "epoch": 18.046272493573266, "grad_norm": 0.30933406949043274, "learning_rate": 8.10478454255914e-08, "loss": 0.0115, "step": 49140 }, { "epoch": 18.049944913698127, "grad_norm": 0.34455522894859314, "learning_rate": 7.917662797181091e-08, "loss": 0.0124, "step": 49150 }, { "epoch": 18.053617333822988, "grad_norm": 0.41845959424972534, "learning_rate": 7.732724758218734e-08, "loss": 0.0118, "step": 49160 }, { "epoch": 18.057289753947853, "grad_norm": 0.3027036488056183, "learning_rate": 7.549970506570691e-08, "loss": 0.0116, "step": 49170 }, { "epoch": 18.060962174072714, "grad_norm": 0.5531248450279236, "learning_rate": 7.369400122179126e-08, "loss": 0.0102, "step": 49180 }, { "epoch": 18.064634594197575, "grad_norm": 0.28578945994377136, "learning_rate": 7.191013684032522e-08, "loss": 0.0113, "step": 49190 }, { "epoch": 18.06830701432244, "grad_norm": 0.24093808233737946, "learning_rate": 7.014811270163458e-08, "loss": 0.0113, "step": 49200 }, { "epoch": 18.0719794344473, "grad_norm": 0.28217625617980957, "learning_rate": 6.840792957647501e-08, "loss": 0.0111, "step": 49210 }, { "epoch": 18.075651854572165, "grad_norm": 0.2497086524963379, "learning_rate": 6.668958822607096e-08, "loss": 0.0114, "step": 49220 }, { "epoch": 18.079324274697026, "grad_norm": 0.5161713361740112, "learning_rate": 6.499308940208226e-08, "loss": 0.0106, "step": 49230 }, { "epoch": 18.082996694821887, "grad_norm": 0.347530335187912, "learning_rate": 6.331843384661529e-08, "loss": 0.013, "step": 49240 }, { "epoch": 18.08666911494675, "grad_norm": 0.41419899463653564, "learning_rate": 6.166562229222295e-08, "loss": 0.0116, "step": 49250 }, { "epoch": 18.090341535071612, "grad_norm": 0.44905632734298706, "learning_rate": 6.003465546189358e-08, "loss": 0.011, "step": 49260 }, { "epoch": 18.094013955196473, "grad_norm": 0.3904307186603546, "learning_rate": 5.8425534069078694e-08, "loss": 0.0118, "step": 49270 }, { "epoch": 18.097686375321338, "grad_norm": 0.8581438660621643, "learning_rate": 5.6838258817654144e-08, "loss": 0.0117, "step": 49280 }, { "epoch": 18.1013587954462, "grad_norm": 0.3785218894481659, "learning_rate": 5.527283040195341e-08, "loss": 0.0103, "step": 49290 }, { "epoch": 18.10503121557106, "grad_norm": 0.23719461262226105, "learning_rate": 5.372924950673986e-08, "loss": 0.0109, "step": 49300 }, { "epoch": 18.108703635695925, "grad_norm": 0.4926145076751709, "learning_rate": 5.2207516807240006e-08, "loss": 0.0122, "step": 49310 }, { "epoch": 18.112376055820786, "grad_norm": 0.2867356538772583, "learning_rate": 5.07076329690992e-08, "loss": 0.0108, "step": 49320 }, { "epoch": 18.116048475945647, "grad_norm": 0.2572305500507355, "learning_rate": 4.922959864842591e-08, "loss": 0.0122, "step": 49330 }, { "epoch": 18.11972089607051, "grad_norm": 0.7594588994979858, "learning_rate": 4.7773414491758537e-08, "loss": 0.0106, "step": 49340 }, { "epoch": 18.123393316195372, "grad_norm": 0.4704957902431488, "learning_rate": 4.6339081136087534e-08, "loss": 0.0109, "step": 49350 }, { "epoch": 18.127065736320237, "grad_norm": 0.24250628054141998, "learning_rate": 4.492659920882769e-08, "loss": 0.0099, "step": 49360 }, { "epoch": 18.130738156445098, "grad_norm": 0.45374003052711487, "learning_rate": 4.3535969327862525e-08, "loss": 0.0111, "step": 49370 }, { "epoch": 18.13441057656996, "grad_norm": 0.2207803875207901, "learning_rate": 4.2167192101483234e-08, "loss": 0.0112, "step": 49380 }, { "epoch": 18.138082996694823, "grad_norm": 0.5262560844421387, "learning_rate": 4.0820268128449744e-08, "loss": 0.0115, "step": 49390 }, { "epoch": 18.141755416819684, "grad_norm": 0.4983327090740204, "learning_rate": 3.949519799795742e-08, "loss": 0.0114, "step": 49400 }, { "epoch": 18.145427836944545, "grad_norm": 0.2241242378950119, "learning_rate": 3.819198228962595e-08, "loss": 0.0133, "step": 49410 }, { "epoch": 18.14910025706941, "grad_norm": 0.3369518220424652, "learning_rate": 3.6910621573527094e-08, "loss": 0.0111, "step": 49420 }, { "epoch": 18.15277267719427, "grad_norm": 0.3014081120491028, "learning_rate": 3.5651116410179156e-08, "loss": 0.0114, "step": 49430 }, { "epoch": 18.156445097319132, "grad_norm": 0.3181496262550354, "learning_rate": 3.441346735052475e-08, "loss": 0.0108, "step": 49440 }, { "epoch": 18.160117517443997, "grad_norm": 0.729159951210022, "learning_rate": 3.31976749359586e-08, "loss": 0.0104, "step": 49450 }, { "epoch": 18.163789937568858, "grad_norm": 0.25764000415802, "learning_rate": 3.200373969831083e-08, "loss": 0.0113, "step": 49460 }, { "epoch": 18.16746235769372, "grad_norm": 0.40761879086494446, "learning_rate": 3.0831662159841456e-08, "loss": 0.0108, "step": 49470 }, { "epoch": 18.171134777818583, "grad_norm": 0.24509263038635254, "learning_rate": 2.9681442833268124e-08, "loss": 0.0112, "step": 49480 }, { "epoch": 18.174807197943444, "grad_norm": 0.33092355728149414, "learning_rate": 2.8553082221721704e-08, "loss": 0.0121, "step": 49490 }, { "epoch": 18.178479618068305, "grad_norm": 0.3478018343448639, "learning_rate": 2.7446580818801805e-08, "loss": 0.0117, "step": 49500 }, { "epoch": 18.18215203819317, "grad_norm": 0.33603477478027344, "learning_rate": 2.63619391085157e-08, "loss": 0.0113, "step": 49510 }, { "epoch": 18.18582445831803, "grad_norm": 0.9729706048965454, "learning_rate": 2.5299157565328302e-08, "loss": 0.0107, "step": 49520 }, { "epoch": 18.189496878442895, "grad_norm": 0.718342125415802, "learning_rate": 2.4258236654139954e-08, "loss": 0.0115, "step": 49530 }, { "epoch": 18.193169298567756, "grad_norm": 0.43532466888427734, "learning_rate": 2.3239176830280873e-08, "loss": 0.0127, "step": 49540 }, { "epoch": 18.196841718692617, "grad_norm": 0.48677146434783936, "learning_rate": 2.2241978539522256e-08, "loss": 0.012, "step": 49550 }, { "epoch": 18.200514138817482, "grad_norm": 0.3458465039730072, "learning_rate": 2.1266642218070733e-08, "loss": 0.0127, "step": 49560 }, { "epoch": 18.204186558942343, "grad_norm": 0.5192819237709045, "learning_rate": 2.0313168292579456e-08, "loss": 0.0114, "step": 49570 }, { "epoch": 18.207858979067204, "grad_norm": 0.5143216848373413, "learning_rate": 1.9381557180120357e-08, "loss": 0.0128, "step": 49580 }, { "epoch": 18.21153139919207, "grad_norm": 0.48798421025276184, "learning_rate": 1.8471809288217456e-08, "loss": 0.0105, "step": 49590 }, { "epoch": 18.21520381931693, "grad_norm": 0.2510438561439514, "learning_rate": 1.758392501482464e-08, "loss": 0.0113, "step": 49600 }, { "epoch": 18.21887623944179, "grad_norm": 0.17344647645950317, "learning_rate": 1.6717904748331236e-08, "loss": 0.0112, "step": 49610 }, { "epoch": 18.222548659566655, "grad_norm": 0.43451452255249023, "learning_rate": 1.5873748867561988e-08, "loss": 0.0125, "step": 49620 }, { "epoch": 18.226221079691516, "grad_norm": 1.462277889251709, "learning_rate": 1.5051457741788177e-08, "loss": 0.0115, "step": 49630 }, { "epoch": 18.229893499816377, "grad_norm": 0.5700339078903198, "learning_rate": 1.425103173069986e-08, "loss": 0.0121, "step": 49640 }, { "epoch": 18.233565919941242, "grad_norm": 1.199957251548767, "learning_rate": 1.347247118442807e-08, "loss": 0.0115, "step": 49650 }, { "epoch": 18.237238340066103, "grad_norm": 0.26961514353752136, "learning_rate": 1.2715776443550375e-08, "loss": 0.0122, "step": 49660 }, { "epoch": 18.240910760190967, "grad_norm": 0.5301100611686707, "learning_rate": 1.1980947839063116e-08, "loss": 0.0114, "step": 49670 }, { "epoch": 18.24458318031583, "grad_norm": 0.2207854986190796, "learning_rate": 1.1267985692414718e-08, "loss": 0.0109, "step": 49680 }, { "epoch": 18.24825560044069, "grad_norm": 0.47040778398513794, "learning_rate": 1.0576890315466826e-08, "loss": 0.012, "step": 49690 }, { "epoch": 18.251928020565554, "grad_norm": 0.32215383648872375, "learning_rate": 9.907662010538721e-09, "loss": 0.0105, "step": 49700 }, { "epoch": 18.255600440690415, "grad_norm": 0.22148683667182922, "learning_rate": 9.26030107036846e-09, "loss": 0.0136, "step": 49710 }, { "epoch": 18.259272860815276, "grad_norm": 0.27567052841186523, "learning_rate": 8.63480777813508e-09, "loss": 0.0118, "step": 49720 }, { "epoch": 18.26294528094014, "grad_norm": 0.556927502155304, "learning_rate": 8.03118240744749e-09, "loss": 0.0108, "step": 49730 }, { "epoch": 18.266617701065, "grad_norm": 0.33938997983932495, "learning_rate": 7.4494252223555885e-09, "loss": 0.0128, "step": 49740 }, { "epoch": 18.270290121189863, "grad_norm": 0.3842446804046631, "learning_rate": 6.8895364773391426e-09, "loss": 0.0106, "step": 49750 }, { "epoch": 18.273962541314727, "grad_norm": 0.25517737865448, "learning_rate": 6.3515164173078015e-09, "loss": 0.0132, "step": 49760 }, { "epoch": 18.27763496143959, "grad_norm": 0.4568251967430115, "learning_rate": 5.835365277617744e-09, "loss": 0.0121, "step": 49770 }, { "epoch": 18.28130738156445, "grad_norm": 0.6074317693710327, "learning_rate": 5.341083284049475e-09, "loss": 0.011, "step": 49780 }, { "epoch": 18.284979801689314, "grad_norm": 0.2638219892978668, "learning_rate": 4.868670652813378e-09, "loss": 0.0118, "step": 49790 }, { "epoch": 18.288652221814175, "grad_norm": 0.3173985183238983, "learning_rate": 4.418127590560817e-09, "loss": 0.0107, "step": 49800 }, { "epoch": 18.29232464193904, "grad_norm": 0.316808819770813, "learning_rate": 3.989454294378581e-09, "loss": 0.0133, "step": 49810 }, { "epoch": 18.2959970620639, "grad_norm": 0.30985206365585327, "learning_rate": 3.5826509517777883e-09, "loss": 0.0116, "step": 49820 }, { "epoch": 18.29966948218876, "grad_norm": 0.554371178150177, "learning_rate": 3.1977177407105372e-09, "loss": 0.0117, "step": 49830 }, { "epoch": 18.303341902313626, "grad_norm": 0.26113972067832947, "learning_rate": 2.8346548295588027e-09, "loss": 0.0123, "step": 49840 }, { "epoch": 18.307014322438487, "grad_norm": 0.3784499168395996, "learning_rate": 2.493462377139988e-09, "loss": 0.012, "step": 49850 }, { "epoch": 18.310686742563348, "grad_norm": 0.607413649559021, "learning_rate": 2.1741405327013744e-09, "loss": 0.0113, "step": 49860 }, { "epoch": 18.314359162688213, "grad_norm": 0.26114439964294434, "learning_rate": 1.8766894359312227e-09, "loss": 0.0117, "step": 49870 }, { "epoch": 18.318031582813074, "grad_norm": 0.910749614238739, "learning_rate": 1.6011092169310182e-09, "loss": 0.0137, "step": 49880 }, { "epoch": 18.321704002937935, "grad_norm": 0.21557731926441193, "learning_rate": 1.3473999962654304e-09, "loss": 0.0121, "step": 49890 }, { "epoch": 18.3253764230628, "grad_norm": 0.32687726616859436, "learning_rate": 1.1155618849068016e-09, "loss": 0.0104, "step": 49900 }, { "epoch": 18.32904884318766, "grad_norm": 0.29352477192878723, "learning_rate": 9.055949842684541e-10, "loss": 0.0116, "step": 49910 }, { "epoch": 18.33272126331252, "grad_norm": 0.3012467622756958, "learning_rate": 7.174993861991386e-10, "loss": 0.012, "step": 49920 }, { "epoch": 18.336393683437386, "grad_norm": 0.4071049392223358, "learning_rate": 5.512751729774834e-10, "loss": 0.0116, "step": 49930 }, { "epoch": 18.340066103562247, "grad_norm": 0.46300023794174194, "learning_rate": 4.0692241731199456e-10, "loss": 0.011, "step": 49940 }, { "epoch": 18.34373852368711, "grad_norm": 0.4106196463108063, "learning_rate": 2.8444118235770865e-10, "loss": 0.0112, "step": 49950 }, { "epoch": 18.347410943811973, "grad_norm": 0.3666224181652069, "learning_rate": 1.8383152168288675e-10, "loss": 0.0113, "step": 49960 }, { "epoch": 18.351083363936834, "grad_norm": 0.3779792785644531, "learning_rate": 1.050934793023206e-10, "loss": 0.0117, "step": 49970 }, { "epoch": 18.354755784061698, "grad_norm": 0.5685268640518188, "learning_rate": 4.822708965512845e-11, "loss": 0.0106, "step": 49980 }, { "epoch": 18.35842820418656, "grad_norm": 0.2597259283065796, "learning_rate": 1.323237761585716e-11, "loss": 0.015, "step": 49990 }, { "epoch": 18.36210062431142, "grad_norm": 0.2930794954299927, "learning_rate": 1.0935850003335191e-13, "loss": 0.0122, "step": 50000 }, { "epoch": 18.36210062431142, "step": 50000, "total_flos": 0.0, "train_loss": 0.03463119722574949, "train_runtime": 82226.4807, "train_samples_per_second": 155.668, "train_steps_per_second": 0.608 } ], "logging_steps": 10, "max_steps": 50000, "num_input_tokens_seen": 0, "num_train_epochs": 19, "save_steps": 25000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 32, "trial_name": null, "trial_params": null }