{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.3328340822100183,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006656681644200366,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 1.9001,
      "step": 1
    },
    {
      "epoch": 0.0013313363288400732,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 1.1836,
      "step": 2
    },
    {
      "epoch": 0.00199700449326011,
      "grad_norm": Infinity,
      "learning_rate": 0.0,
      "loss": 2.0432,
      "step": 3
    },
    {
      "epoch": 0.0026626726576801465,
      "grad_norm": Infinity,
      "learning_rate": 0.0,
      "loss": 2.7364,
      "step": 4
    },
    {
      "epoch": 0.003328340822100183,
      "grad_norm": 305.05010986328125,
      "learning_rate": 0.0001,
      "loss": 2.7278,
      "step": 5
    },
    {
      "epoch": 0.00399400898652022,
      "grad_norm": 195.2991180419922,
      "learning_rate": 0.0002,
      "loss": 1.5935,
      "step": 6
    },
    {
      "epoch": 0.004659677150940256,
      "grad_norm": 135.4779815673828,
      "learning_rate": 0.00019995559502664298,
      "loss": 1.558,
      "step": 7
    },
    {
      "epoch": 0.005325345315360293,
      "grad_norm": 175.23486328125,
      "learning_rate": 0.00019991119005328598,
      "loss": 1.8278,
      "step": 8
    },
    {
      "epoch": 0.005991013479780329,
      "grad_norm": 70.18768310546875,
      "learning_rate": 0.00019986678507992895,
      "loss": 1.2821,
      "step": 9
    },
    {
      "epoch": 0.006656681644200366,
      "grad_norm": 91.9547348022461,
      "learning_rate": 0.00019982238010657195,
      "loss": 1.3009,
      "step": 10
    },
    {
      "epoch": 0.007322349808620403,
      "grad_norm": 140.80667114257812,
      "learning_rate": 0.00019977797513321492,
      "loss": 1.4739,
      "step": 11
    },
    {
      "epoch": 0.00798801797304044,
      "grad_norm": 117.13510131835938,
      "learning_rate": 0.00019973357015985792,
      "loss": 1.0922,
      "step": 12
    },
    {
      "epoch": 0.008653686137460476,
      "grad_norm": 87.54082489013672,
      "learning_rate": 0.0001996891651865009,
      "loss": 1.6633,
      "step": 13
    },
    {
      "epoch": 0.009319354301880512,
      "grad_norm": 178.7975616455078,
      "learning_rate": 0.0001996447602131439,
      "loss": 1.7614,
      "step": 14
    },
    {
      "epoch": 0.00998502246630055,
      "grad_norm": 199.108154296875,
      "learning_rate": 0.00019960035523978686,
      "loss": 1.4873,
      "step": 15
    },
    {
      "epoch": 0.010650690630720586,
      "grad_norm": 97.36341857910156,
      "learning_rate": 0.00019955595026642986,
      "loss": 1.2921,
      "step": 16
    },
    {
      "epoch": 0.011316358795140622,
      "grad_norm": 145.5021209716797,
      "learning_rate": 0.00019951154529307283,
      "loss": 1.4667,
      "step": 17
    },
    {
      "epoch": 0.011982026959560658,
      "grad_norm": 163.09628295898438,
      "learning_rate": 0.00019946714031971583,
      "loss": 1.4959,
      "step": 18
    },
    {
      "epoch": 0.012647695123980696,
      "grad_norm": 181.53123474121094,
      "learning_rate": 0.0001994227353463588,
      "loss": 1.4931,
      "step": 19
    },
    {
      "epoch": 0.013313363288400732,
      "grad_norm": 134.01087951660156,
      "learning_rate": 0.00019937833037300177,
      "loss": 1.1797,
      "step": 20
    },
    {
      "epoch": 0.013979031452820768,
      "grad_norm": 138.4213409423828,
      "learning_rate": 0.00019933392539964477,
      "loss": 1.3021,
      "step": 21
    },
    {
      "epoch": 0.014644699617240806,
      "grad_norm": 94.12645721435547,
      "learning_rate": 0.00019928952042628774,
      "loss": 1.2099,
      "step": 22
    },
    {
      "epoch": 0.015310367781660842,
      "grad_norm": 68.36018371582031,
      "learning_rate": 0.00019924511545293074,
      "loss": 1.0593,
      "step": 23
    },
    {
      "epoch": 0.01597603594608088,
      "grad_norm": 83.23944854736328,
      "learning_rate": 0.0001992007104795737,
      "loss": 1.1152,
      "step": 24
    },
    {
      "epoch": 0.016641704110500914,
      "grad_norm": 145.29953002929688,
      "learning_rate": 0.0001991563055062167,
      "loss": 1.194,
      "step": 25
    },
    {
      "epoch": 0.017307372274920952,
      "grad_norm": 71.01487731933594,
      "learning_rate": 0.00019911190053285968,
      "loss": 1.0201,
      "step": 26
    },
    {
      "epoch": 0.01797304043934099,
      "grad_norm": 82.99989318847656,
      "learning_rate": 0.00019906749555950268,
      "loss": 1.3319,
      "step": 27
    },
    {
      "epoch": 0.018638708603761024,
      "grad_norm": 51.65462112426758,
      "learning_rate": 0.00019902309058614565,
      "loss": 0.8622,
      "step": 28
    },
    {
      "epoch": 0.019304376768181062,
      "grad_norm": 106.69747161865234,
      "learning_rate": 0.00019897868561278865,
      "loss": 1.0051,
      "step": 29
    },
    {
      "epoch": 0.0199700449326011,
      "grad_norm": 85.53784942626953,
      "learning_rate": 0.00019893428063943162,
      "loss": 0.7687,
      "step": 30
    },
    {
      "epoch": 0.020635713097021134,
      "grad_norm": 141.33482360839844,
      "learning_rate": 0.00019888987566607462,
      "loss": 0.9697,
      "step": 31
    },
    {
      "epoch": 0.021301381261441172,
      "grad_norm": 110.15571594238281,
      "learning_rate": 0.0001988454706927176,
      "loss": 0.6955,
      "step": 32
    },
    {
      "epoch": 0.02196704942586121,
      "grad_norm": 123.28990173339844,
      "learning_rate": 0.00019880106571936056,
      "loss": 0.9405,
      "step": 33
    },
    {
      "epoch": 0.022632717590281244,
      "grad_norm": 104.15376281738281,
      "learning_rate": 0.00019875666074600356,
      "loss": 1.1353,
      "step": 34
    },
    {
      "epoch": 0.023298385754701282,
      "grad_norm": 122.83297729492188,
      "learning_rate": 0.00019871225577264653,
      "loss": 0.9912,
      "step": 35
    },
    {
      "epoch": 0.023964053919121316,
      "grad_norm": 137.1550750732422,
      "learning_rate": 0.00019866785079928953,
      "loss": 1.0156,
      "step": 36
    },
    {
      "epoch": 0.024629722083541354,
      "grad_norm": 95.7591781616211,
      "learning_rate": 0.0001986234458259325,
      "loss": 0.9295,
      "step": 37
    },
    {
      "epoch": 0.025295390247961392,
      "grad_norm": 163.35743713378906,
      "learning_rate": 0.0001985790408525755,
      "loss": 0.9697,
      "step": 38
    },
    {
      "epoch": 0.025961058412381426,
      "grad_norm": 42.57448959350586,
      "learning_rate": 0.00019853463587921847,
      "loss": 0.8329,
      "step": 39
    },
    {
      "epoch": 0.026626726576801464,
      "grad_norm": 161.05844116210938,
      "learning_rate": 0.00019849023090586147,
      "loss": 0.8588,
      "step": 40
    },
    {
      "epoch": 0.027292394741221502,
      "grad_norm": 123.78165435791016,
      "learning_rate": 0.00019844582593250444,
      "loss": 0.8082,
      "step": 41
    },
    {
      "epoch": 0.027958062905641536,
      "grad_norm": 117.62382507324219,
      "learning_rate": 0.00019840142095914744,
      "loss": 1.0692,
      "step": 42
    },
    {
      "epoch": 0.028623731070061574,
      "grad_norm": NaN,
      "learning_rate": 0.00019840142095914744,
      "loss": 1.316,
      "step": 43
    },
    {
      "epoch": 0.029289399234481612,
      "grad_norm": 63.35919952392578,
      "learning_rate": 0.0001983570159857904,
      "loss": 0.6028,
      "step": 44
    },
    {
      "epoch": 0.029955067398901646,
      "grad_norm": 277.1460266113281,
      "learning_rate": 0.00019831261101243338,
      "loss": 1.4349,
      "step": 45
    },
    {
      "epoch": 0.030620735563321684,
      "grad_norm": NaN,
      "learning_rate": 0.00019831261101243338,
      "loss": 1.816,
      "step": 46
    },
    {
      "epoch": 0.03128640372774172,
      "grad_norm": 206.0849151611328,
      "learning_rate": 0.00019826820603907638,
      "loss": 1.7035,
      "step": 47
    },
    {
      "epoch": 0.03195207189216176,
      "grad_norm": 154.10183715820312,
      "learning_rate": 0.00019822380106571935,
      "loss": 0.7897,
      "step": 48
    },
    {
      "epoch": 0.032617740056581794,
      "grad_norm": 182.32650756835938,
      "learning_rate": 0.00019817939609236235,
      "loss": 1.2611,
      "step": 49
    },
    {
      "epoch": 0.03328340822100183,
      "grad_norm": 100.30379486083984,
      "learning_rate": 0.00019813499111900532,
      "loss": 0.8746,
      "step": 50
    },
    {
      "epoch": 0.03394907638542187,
      "grad_norm": 99.05520629882812,
      "learning_rate": 0.00019809058614564832,
      "loss": 0.6544,
      "step": 51
    },
    {
      "epoch": 0.034614744549841904,
      "grad_norm": 75.6047592163086,
      "learning_rate": 0.0001980461811722913,
      "loss": 0.5308,
      "step": 52
    },
    {
      "epoch": 0.03528041271426194,
      "grad_norm": 81.81549835205078,
      "learning_rate": 0.0001980017761989343,
      "loss": 1.0103,
      "step": 53
    },
    {
      "epoch": 0.03594608087868198,
      "grad_norm": 305.5655517578125,
      "learning_rate": 0.00019795737122557726,
      "loss": 2.0811,
      "step": 54
    },
    {
      "epoch": 0.036611749043102014,
      "grad_norm": 254.49461364746094,
      "learning_rate": 0.00019791296625222026,
      "loss": 1.5316,
      "step": 55
    },
    {
      "epoch": 0.03727741720752205,
      "grad_norm": 249.2161407470703,
      "learning_rate": 0.00019786856127886323,
      "loss": 1.4851,
      "step": 56
    },
    {
      "epoch": 0.03794308537194209,
      "grad_norm": 187.69805908203125,
      "learning_rate": 0.00019782415630550623,
      "loss": 1.0635,
      "step": 57
    },
    {
      "epoch": 0.038608753536362124,
      "grad_norm": 135.42141723632812,
      "learning_rate": 0.0001977797513321492,
      "loss": 0.8954,
      "step": 58
    },
    {
      "epoch": 0.03927442170078216,
      "grad_norm": 198.7463836669922,
      "learning_rate": 0.0001977353463587922,
      "loss": 1.2765,
      "step": 59
    },
    {
      "epoch": 0.0399400898652022,
      "grad_norm": 256.67279052734375,
      "learning_rate": 0.00019769094138543517,
      "loss": 1.4927,
      "step": 60
    },
    {
      "epoch": 0.040605758029622234,
      "grad_norm": 64.07623291015625,
      "learning_rate": 0.00019764653641207817,
      "loss": 0.7874,
      "step": 61
    },
    {
      "epoch": 0.04127142619404227,
      "grad_norm": 147.7261962890625,
      "learning_rate": 0.00019760213143872114,
      "loss": 0.7924,
      "step": 62
    },
    {
      "epoch": 0.04193709435846231,
      "grad_norm": 122.39730072021484,
      "learning_rate": 0.00019755772646536413,
      "loss": 0.7656,
      "step": 63
    },
    {
      "epoch": 0.042602762522882344,
      "grad_norm": 94.13729095458984,
      "learning_rate": 0.0001975133214920071,
      "loss": 0.9051,
      "step": 64
    },
    {
      "epoch": 0.04326843068730238,
      "grad_norm": 48.220184326171875,
      "learning_rate": 0.0001974689165186501,
      "loss": 0.6262,
      "step": 65
    },
    {
      "epoch": 0.04393409885172242,
      "grad_norm": 32.33949661254883,
      "learning_rate": 0.00019742451154529308,
      "loss": 0.544,
      "step": 66
    },
    {
      "epoch": 0.044599767016142454,
      "grad_norm": 133.78097534179688,
      "learning_rate": 0.00019738010657193607,
      "loss": 0.6472,
      "step": 67
    },
    {
      "epoch": 0.04526543518056249,
      "grad_norm": 105.52882385253906,
      "learning_rate": 0.00019733570159857907,
      "loss": 0.7272,
      "step": 68
    },
    {
      "epoch": 0.04593110334498253,
      "grad_norm": 40.91939926147461,
      "learning_rate": 0.00019729129662522204,
      "loss": 0.6232,
      "step": 69
    },
    {
      "epoch": 0.046596771509402564,
      "grad_norm": 48.693763732910156,
      "learning_rate": 0.00019724689165186504,
      "loss": 0.3698,
      "step": 70
    },
    {
      "epoch": 0.0472624396738226,
      "grad_norm": 98.95121765136719,
      "learning_rate": 0.00019720248667850801,
      "loss": 0.4914,
      "step": 71
    },
    {
      "epoch": 0.04792810783824263,
      "grad_norm": 217.00401306152344,
      "learning_rate": 0.00019715808170515098,
      "loss": 1.3837,
      "step": 72
    },
    {
      "epoch": 0.048593776002662674,
      "grad_norm": 97.31861877441406,
      "learning_rate": 0.00019711367673179398,
      "loss": 0.7037,
      "step": 73
    },
    {
      "epoch": 0.04925944416708271,
      "grad_norm": 101.28430938720703,
      "learning_rate": 0.00019706927175843695,
      "loss": 1.3999,
      "step": 74
    },
    {
      "epoch": 0.04992511233150274,
      "grad_norm": 86.35800170898438,
      "learning_rate": 0.00019702486678507995,
      "loss": 0.8045,
      "step": 75
    },
    {
      "epoch": 0.050590780495922784,
      "grad_norm": 102.23155212402344,
      "learning_rate": 0.00019698046181172292,
      "loss": 0.7521,
      "step": 76
    },
    {
      "epoch": 0.05125644866034282,
      "grad_norm": 125.80261993408203,
      "learning_rate": 0.00019693605683836592,
      "loss": 0.7933,
      "step": 77
    },
    {
      "epoch": 0.05192211682476285,
      "grad_norm": 160.28123474121094,
      "learning_rate": 0.0001968916518650089,
      "loss": 0.6355,
      "step": 78
    },
    {
      "epoch": 0.052587784989182894,
      "grad_norm": 42.919464111328125,
      "learning_rate": 0.0001968472468916519,
      "loss": 0.2693,
      "step": 79
    },
    {
      "epoch": 0.05325345315360293,
      "grad_norm": 56.61810302734375,
      "learning_rate": 0.00019680284191829486,
      "loss": 0.3141,
      "step": 80
    },
    {
      "epoch": 0.05391912131802296,
      "grad_norm": 299.3166198730469,
      "learning_rate": 0.00019675843694493786,
      "loss": 1.7119,
      "step": 81
    },
    {
      "epoch": 0.054584789482443004,
      "grad_norm": 133.17459106445312,
      "learning_rate": 0.00019671403197158083,
      "loss": 1.2885,
      "step": 82
    },
    {
      "epoch": 0.05525045764686304,
      "grad_norm": 131.6356964111328,
      "learning_rate": 0.00019666962699822383,
      "loss": 0.7976,
      "step": 83
    },
    {
      "epoch": 0.05591612581128307,
      "grad_norm": 122.43846130371094,
      "learning_rate": 0.0001966252220248668,
      "loss": 0.7031,
      "step": 84
    },
    {
      "epoch": 0.056581793975703114,
      "grad_norm": 51.365901947021484,
      "learning_rate": 0.00019658081705150977,
      "loss": 0.6684,
      "step": 85
    },
    {
      "epoch": 0.05724746214012315,
      "grad_norm": 251.70664978027344,
      "learning_rate": 0.00019653641207815277,
      "loss": 1.2329,
      "step": 86
    },
    {
      "epoch": 0.05791313030454318,
      "grad_norm": 94.94196319580078,
      "learning_rate": 0.00019649200710479574,
      "loss": 0.746,
      "step": 87
    },
    {
      "epoch": 0.058578798468963224,
      "grad_norm": 106.89917755126953,
      "learning_rate": 0.00019644760213143874,
      "loss": 0.8735,
      "step": 88
    },
    {
      "epoch": 0.05924446663338326,
      "grad_norm": 130.65577697753906,
      "learning_rate": 0.00019640319715808171,
      "loss": 1.0905,
      "step": 89
    },
    {
      "epoch": 0.05991013479780329,
      "grad_norm": 37.6524543762207,
      "learning_rate": 0.0001963587921847247,
      "loss": 0.447,
      "step": 90
    },
    {
      "epoch": 0.060575802962223334,
      "grad_norm": 55.81675720214844,
      "learning_rate": 0.00019631438721136768,
      "loss": 0.8305,
      "step": 91
    },
    {
      "epoch": 0.06124147112664337,
      "grad_norm": 56.32381057739258,
      "learning_rate": 0.00019626998223801068,
      "loss": 0.6437,
      "step": 92
    },
    {
      "epoch": 0.0619071392910634,
      "grad_norm": 107.72248840332031,
      "learning_rate": 0.00019622557726465365,
      "loss": 0.7331,
      "step": 93
    },
    {
      "epoch": 0.06257280745548344,
      "grad_norm": 60.965084075927734,
      "learning_rate": 0.00019618117229129665,
      "loss": 0.1975,
      "step": 94
    },
    {
      "epoch": 0.06323847561990348,
      "grad_norm": 236.83071899414062,
      "learning_rate": 0.00019613676731793962,
      "loss": 1.3465,
      "step": 95
    },
    {
      "epoch": 0.06390414378432352,
      "grad_norm": 32.12284851074219,
      "learning_rate": 0.0001960923623445826,
      "loss": 0.4818,
      "step": 96
    },
    {
      "epoch": 0.06456981194874355,
      "grad_norm": 139.45631408691406,
      "learning_rate": 0.0001960479573712256,
      "loss": 0.7907,
      "step": 97
    },
    {
      "epoch": 0.06523548011316359,
      "grad_norm": 85.41361999511719,
      "learning_rate": 0.00019600355239786856,
      "loss": 0.5648,
      "step": 98
    },
    {
      "epoch": 0.06590114827758363,
      "grad_norm": 79.85189056396484,
      "learning_rate": 0.00019595914742451156,
      "loss": 0.7735,
      "step": 99
    },
    {
      "epoch": 0.06656681644200366,
      "grad_norm": 64.68962860107422,
      "learning_rate": 0.00019591474245115453,
      "loss": 0.4128,
      "step": 100
    },
    {
      "epoch": 0.0672324846064237,
      "grad_norm": 58.707054138183594,
      "learning_rate": 0.00019587033747779753,
      "loss": 0.8143,
      "step": 101
    },
    {
      "epoch": 0.06789815277084374,
      "grad_norm": 107.83698272705078,
      "learning_rate": 0.0001958259325044405,
      "loss": 0.8662,
      "step": 102
    },
    {
      "epoch": 0.06856382093526377,
      "grad_norm": 221.64857482910156,
      "learning_rate": 0.0001957815275310835,
      "loss": 1.2667,
      "step": 103
    },
    {
      "epoch": 0.06922948909968381,
      "grad_norm": 75.43016815185547,
      "learning_rate": 0.00019573712255772647,
      "loss": 0.6056,
      "step": 104
    },
    {
      "epoch": 0.06989515726410385,
      "grad_norm": 81.48358917236328,
      "learning_rate": 0.00019569271758436947,
      "loss": 0.4954,
      "step": 105
    },
    {
      "epoch": 0.07056082542852388,
      "grad_norm": 100.88373565673828,
      "learning_rate": 0.00019564831261101244,
      "loss": 0.8129,
      "step": 106
    },
    {
      "epoch": 0.07122649359294392,
      "grad_norm": 56.06926727294922,
      "learning_rate": 0.00019560390763765544,
      "loss": 0.5694,
      "step": 107
    },
    {
      "epoch": 0.07189216175736396,
      "grad_norm": 119.89936828613281,
      "learning_rate": 0.0001955595026642984,
      "loss": 0.8492,
      "step": 108
    },
    {
      "epoch": 0.07255782992178399,
      "grad_norm": 226.14344787597656,
      "learning_rate": 0.00019551509769094138,
      "loss": 1.5915,
      "step": 109
    },
    {
      "epoch": 0.07322349808620403,
      "grad_norm": 167.6849365234375,
      "learning_rate": 0.00019547069271758438,
      "loss": 0.8734,
      "step": 110
    },
    {
      "epoch": 0.07388916625062407,
      "grad_norm": 191.29388427734375,
      "learning_rate": 0.00019542628774422735,
      "loss": 1.1915,
      "step": 111
    },
    {
      "epoch": 0.0745548344150441,
      "grad_norm": 74.29916381835938,
      "learning_rate": 0.00019538188277087035,
      "loss": 0.399,
      "step": 112
    },
    {
      "epoch": 0.07522050257946414,
      "grad_norm": 193.20899963378906,
      "learning_rate": 0.00019533747779751332,
      "loss": 1.0812,
      "step": 113
    },
    {
      "epoch": 0.07588617074388418,
      "grad_norm": 104.37834930419922,
      "learning_rate": 0.00019529307282415632,
      "loss": 0.7821,
      "step": 114
    },
    {
      "epoch": 0.0765518389083042,
      "grad_norm": 63.434146881103516,
      "learning_rate": 0.0001952486678507993,
      "loss": 0.5636,
      "step": 115
    },
    {
      "epoch": 0.07721750707272425,
      "grad_norm": 163.28578186035156,
      "learning_rate": 0.0001952042628774423,
      "loss": 0.8423,
      "step": 116
    },
    {
      "epoch": 0.07788317523714429,
      "grad_norm": 94.21780395507812,
      "learning_rate": 0.00019515985790408526,
      "loss": 0.6704,
      "step": 117
    },
    {
      "epoch": 0.07854884340156432,
      "grad_norm": 97.0186767578125,
      "learning_rate": 0.00019511545293072826,
      "loss": 0.4906,
      "step": 118
    },
    {
      "epoch": 0.07921451156598436,
      "grad_norm": 67.47974395751953,
      "learning_rate": 0.00019507104795737123,
      "loss": 0.5328,
      "step": 119
    },
    {
      "epoch": 0.0798801797304044,
      "grad_norm": 113.68756103515625,
      "learning_rate": 0.00019502664298401423,
      "loss": 0.7459,
      "step": 120
    },
    {
      "epoch": 0.08054584789482443,
      "grad_norm": 71.1500244140625,
      "learning_rate": 0.0001949822380106572,
      "loss": 0.6177,
      "step": 121
    },
    {
      "epoch": 0.08121151605924447,
      "grad_norm": 131.37168884277344,
      "learning_rate": 0.00019493783303730017,
      "loss": 0.7955,
      "step": 122
    },
    {
      "epoch": 0.08187718422366451,
      "grad_norm": 72.7484359741211,
      "learning_rate": 0.00019489342806394317,
      "loss": 0.7026,
      "step": 123
    },
    {
      "epoch": 0.08254285238808454,
      "grad_norm": 34.95873260498047,
      "learning_rate": 0.00019484902309058614,
      "loss": 0.3607,
      "step": 124
    },
    {
      "epoch": 0.08320852055250458,
      "grad_norm": 23.322694778442383,
      "learning_rate": 0.00019480461811722914,
      "loss": 0.4757,
      "step": 125
    },
    {
      "epoch": 0.08387418871692462,
      "grad_norm": 102.11334228515625,
      "learning_rate": 0.0001947602131438721,
      "loss": 0.9689,
      "step": 126
    },
    {
      "epoch": 0.08453985688134465,
      "grad_norm": 70.20091247558594,
      "learning_rate": 0.0001947158081705151,
      "loss": 0.5523,
      "step": 127
    },
    {
      "epoch": 0.08520552504576469,
      "grad_norm": 108.84575653076172,
      "learning_rate": 0.00019467140319715808,
      "loss": 0.7221,
      "step": 128
    },
    {
      "epoch": 0.08587119321018473,
      "grad_norm": 74.28465270996094,
      "learning_rate": 0.00019462699822380108,
      "loss": 0.3408,
      "step": 129
    },
    {
      "epoch": 0.08653686137460476,
      "grad_norm": 45.54611587524414,
      "learning_rate": 0.00019458259325044405,
      "loss": 0.3064,
      "step": 130
    },
    {
      "epoch": 0.0872025295390248,
      "grad_norm": 107.5720443725586,
      "learning_rate": 0.00019453818827708705,
      "loss": 0.4285,
      "step": 131
    },
    {
      "epoch": 0.08786819770344484,
      "grad_norm": 113.14404296875,
      "learning_rate": 0.00019449378330373002,
      "loss": 0.7893,
      "step": 132
    },
    {
      "epoch": 0.08853386586786487,
      "grad_norm": 41.36758041381836,
      "learning_rate": 0.00019444937833037302,
      "loss": 0.2135,
      "step": 133
    },
    {
      "epoch": 0.08919953403228491,
      "grad_norm": 111.0009765625,
      "learning_rate": 0.000194404973357016,
      "loss": 0.4909,
      "step": 134
    },
    {
      "epoch": 0.08986520219670495,
      "grad_norm": 119.50690460205078,
      "learning_rate": 0.00019436056838365896,
      "loss": 0.8744,
      "step": 135
    },
    {
      "epoch": 0.09053087036112498,
      "grad_norm": 83.16583251953125,
      "learning_rate": 0.00019431616341030196,
      "loss": 0.5229,
      "step": 136
    },
    {
      "epoch": 0.09119653852554502,
      "grad_norm": 96.94984436035156,
      "learning_rate": 0.00019427175843694493,
      "loss": 0.6746,
      "step": 137
    },
    {
      "epoch": 0.09186220668996506,
      "grad_norm": 65.90393829345703,
      "learning_rate": 0.00019422735346358793,
      "loss": 0.3112,
      "step": 138
    },
    {
      "epoch": 0.09252787485438509,
      "grad_norm": 92.0575180053711,
      "learning_rate": 0.0001941829484902309,
      "loss": 0.5264,
      "step": 139
    },
    {
      "epoch": 0.09319354301880513,
      "grad_norm": 63.846744537353516,
      "learning_rate": 0.0001941385435168739,
      "loss": 0.8869,
      "step": 140
    },
    {
      "epoch": 0.09385921118322517,
      "grad_norm": 129.04086303710938,
      "learning_rate": 0.00019409413854351687,
      "loss": 0.7077,
      "step": 141
    },
    {
      "epoch": 0.0945248793476452,
      "grad_norm": 38.49999237060547,
      "learning_rate": 0.00019404973357015987,
      "loss": 0.1854,
      "step": 142
    },
    {
      "epoch": 0.09519054751206524,
      "grad_norm": 142.83619689941406,
      "learning_rate": 0.00019400532859680284,
      "loss": 1.3549,
      "step": 143
    },
    {
      "epoch": 0.09585621567648527,
      "grad_norm": 58.420223236083984,
      "learning_rate": 0.00019396092362344584,
      "loss": 0.3978,
      "step": 144
    },
    {
      "epoch": 0.0965218838409053,
      "grad_norm": 82.57951354980469,
      "learning_rate": 0.0001939165186500888,
      "loss": 0.9771,
      "step": 145
    },
    {
      "epoch": 0.09718755200532535,
      "grad_norm": 52.69244384765625,
      "learning_rate": 0.00019387211367673178,
      "loss": 0.389,
      "step": 146
    },
    {
      "epoch": 0.09785322016974538,
      "grad_norm": 99.57772827148438,
      "learning_rate": 0.00019382770870337478,
      "loss": 0.938,
      "step": 147
    },
    {
      "epoch": 0.09851888833416542,
      "grad_norm": 66.55644226074219,
      "learning_rate": 0.00019378330373001775,
      "loss": 0.4984,
      "step": 148
    },
    {
      "epoch": 0.09918455649858546,
      "grad_norm": 64.8500747680664,
      "learning_rate": 0.00019373889875666075,
      "loss": 0.4189,
      "step": 149
    },
    {
      "epoch": 0.09985022466300549,
      "grad_norm": 113.6292495727539,
      "learning_rate": 0.00019369449378330372,
      "loss": 0.486,
      "step": 150
    },
    {
      "epoch": 0.10051589282742553,
      "grad_norm": 83.64993286132812,
      "learning_rate": 0.00019365008880994672,
      "loss": 0.3288,
      "step": 151
    },
    {
      "epoch": 0.10118156099184557,
      "grad_norm": 70.0993423461914,
      "learning_rate": 0.0001936056838365897,
      "loss": 0.7637,
      "step": 152
    },
    {
      "epoch": 0.1018472291562656,
      "grad_norm": 68.05122375488281,
      "learning_rate": 0.0001935612788632327,
      "loss": 0.3255,
      "step": 153
    },
    {
      "epoch": 0.10251289732068564,
      "grad_norm": 56.31193923950195,
      "learning_rate": 0.00019351687388987566,
      "loss": 0.1978,
      "step": 154
    },
    {
      "epoch": 0.10317856548510568,
      "grad_norm": 131.5350799560547,
      "learning_rate": 0.00019347246891651866,
      "loss": 0.7034,
      "step": 155
    },
    {
      "epoch": 0.1038442336495257,
      "grad_norm": 143.9272003173828,
      "learning_rate": 0.00019342806394316163,
      "loss": 0.6122,
      "step": 156
    },
    {
      "epoch": 0.10450990181394575,
      "grad_norm": 56.10287094116211,
      "learning_rate": 0.00019338365896980463,
      "loss": 0.5982,
      "step": 157
    },
    {
      "epoch": 0.10517556997836579,
      "grad_norm": 113.03327941894531,
      "learning_rate": 0.0001933392539964476,
      "loss": 0.5578,
      "step": 158
    },
    {
      "epoch": 0.10584123814278582,
      "grad_norm": 156.54730224609375,
      "learning_rate": 0.0001932948490230906,
      "loss": 0.9944,
      "step": 159
    },
    {
      "epoch": 0.10650690630720586,
      "grad_norm": 117.35420989990234,
      "learning_rate": 0.00019325044404973357,
      "loss": 0.9612,
      "step": 160
    },
    {
      "epoch": 0.1071725744716259,
      "grad_norm": 137.27517700195312,
      "learning_rate": 0.00019320603907637657,
      "loss": 0.8892,
      "step": 161
    },
    {
      "epoch": 0.10783824263604593,
      "grad_norm": 121.9662094116211,
      "learning_rate": 0.00019316163410301954,
      "loss": 0.8541,
      "step": 162
    },
    {
      "epoch": 0.10850391080046597,
      "grad_norm": 50.96897506713867,
      "learning_rate": 0.00019311722912966254,
      "loss": 0.4626,
      "step": 163
    },
    {
      "epoch": 0.10916957896488601,
      "grad_norm": 86.26678466796875,
      "learning_rate": 0.0001930728241563055,
      "loss": 0.4941,
      "step": 164
    },
    {
      "epoch": 0.10983524712930604,
      "grad_norm": 92.44398498535156,
      "learning_rate": 0.0001930284191829485,
      "loss": 0.6016,
      "step": 165
    },
    {
      "epoch": 0.11050091529372608,
      "grad_norm": 76.84557342529297,
      "learning_rate": 0.00019298401420959148,
      "loss": 0.465,
      "step": 166
    },
    {
      "epoch": 0.11116658345814612,
      "grad_norm": 120.3193588256836,
      "learning_rate": 0.00019293960923623448,
      "loss": 0.7597,
      "step": 167
    },
    {
      "epoch": 0.11183225162256615,
      "grad_norm": 98.53234100341797,
      "learning_rate": 0.00019289520426287745,
      "loss": 0.4161,
      "step": 168
    },
    {
      "epoch": 0.11249791978698619,
      "grad_norm": 136.8874053955078,
      "learning_rate": 0.00019285079928952045,
      "loss": 1.0431,
      "step": 169
    },
    {
      "epoch": 0.11316358795140623,
      "grad_norm": 145.96075439453125,
      "learning_rate": 0.00019280639431616342,
      "loss": 0.8223,
      "step": 170
    },
    {
      "epoch": 0.11382925611582626,
      "grad_norm": 95.09919738769531,
      "learning_rate": 0.00019276198934280642,
      "loss": 0.4007,
      "step": 171
    },
    {
      "epoch": 0.1144949242802463,
      "grad_norm": 95.6208267211914,
      "learning_rate": 0.0001927175843694494,
      "loss": 0.4496,
      "step": 172
    },
    {
      "epoch": 0.11516059244466634,
      "grad_norm": 131.4598846435547,
      "learning_rate": 0.00019267317939609239,
      "loss": 0.7203,
      "step": 173
    },
    {
      "epoch": 0.11582626060908637,
      "grad_norm": 183.2200469970703,
      "learning_rate": 0.00019262877442273536,
      "loss": 0.8879,
      "step": 174
    },
    {
      "epoch": 0.1164919287735064,
      "grad_norm": 65.30191802978516,
      "learning_rate": 0.00019258436944937836,
      "loss": 0.2161,
      "step": 175
    },
    {
      "epoch": 0.11715759693792645,
      "grad_norm": 100.97784423828125,
      "learning_rate": 0.00019253996447602133,
      "loss": 0.8095,
      "step": 176
    },
    {
      "epoch": 0.11782326510234647,
      "grad_norm": 57.705963134765625,
      "learning_rate": 0.00019249555950266432,
      "loss": 0.3389,
      "step": 177
    },
    {
      "epoch": 0.11848893326676652,
      "grad_norm": 119.22857666015625,
      "learning_rate": 0.0001924511545293073,
      "loss": 0.5287,
      "step": 178
    },
    {
      "epoch": 0.11915460143118656,
      "grad_norm": 97.59465026855469,
      "learning_rate": 0.0001924067495559503,
      "loss": 0.4567,
      "step": 179
    },
    {
      "epoch": 0.11982026959560658,
      "grad_norm": 47.80794906616211,
      "learning_rate": 0.00019236234458259327,
      "loss": 0.3795,
      "step": 180
    },
    {
      "epoch": 0.12048593776002663,
      "grad_norm": 93.19955444335938,
      "learning_rate": 0.00019231793960923626,
      "loss": 1.0376,
      "step": 181
    },
    {
      "epoch": 0.12115160592444667,
      "grad_norm": 57.24619674682617,
      "learning_rate": 0.00019227353463587924,
      "loss": 0.3167,
      "step": 182
    },
    {
      "epoch": 0.1218172740888667,
      "grad_norm": 71.7004165649414,
      "learning_rate": 0.00019222912966252223,
      "loss": 0.3515,
      "step": 183
    },
    {
      "epoch": 0.12248294225328674,
      "grad_norm": 42.93682098388672,
      "learning_rate": 0.0001921847246891652,
      "loss": 0.3272,
      "step": 184
    },
    {
      "epoch": 0.12314861041770678,
      "grad_norm": 88.22802734375,
      "learning_rate": 0.00019214031971580818,
      "loss": 0.5751,
      "step": 185
    },
    {
      "epoch": 0.1238142785821268,
      "grad_norm": 88.80204772949219,
      "learning_rate": 0.00019209591474245117,
      "loss": 0.5306,
      "step": 186
    },
    {
      "epoch": 0.12447994674654685,
      "grad_norm": 92.46751403808594,
      "learning_rate": 0.00019205150976909415,
      "loss": 0.4262,
      "step": 187
    },
    {
      "epoch": 0.12514561491096687,
      "grad_norm": 55.72347640991211,
      "learning_rate": 0.00019200710479573714,
      "loss": 0.3939,
      "step": 188
    },
    {
      "epoch": 0.12581128307538691,
      "grad_norm": 72.36625671386719,
      "learning_rate": 0.00019196269982238012,
      "loss": 0.3739,
      "step": 189
    },
    {
      "epoch": 0.12647695123980696,
      "grad_norm": 96.56300354003906,
      "learning_rate": 0.00019191829484902311,
      "loss": 0.4574,
      "step": 190
    },
    {
      "epoch": 0.127142619404227,
      "grad_norm": 122.78588104248047,
      "learning_rate": 0.00019187388987566609,
      "loss": 0.7646,
      "step": 191
    },
    {
      "epoch": 0.12780828756864704,
      "grad_norm": 74.9114990234375,
      "learning_rate": 0.00019182948490230908,
      "loss": 0.2156,
      "step": 192
    },
    {
      "epoch": 0.12847395573306705,
      "grad_norm": 61.17184829711914,
      "learning_rate": 0.00019178507992895206,
      "loss": 0.7067,
      "step": 193
    },
    {
      "epoch": 0.1291396238974871,
      "grad_norm": 120.3477783203125,
      "learning_rate": 0.00019174067495559505,
      "loss": 0.7562,
      "step": 194
    },
    {
      "epoch": 0.12980529206190713,
      "grad_norm": 204.0528564453125,
      "learning_rate": 0.00019169626998223802,
      "loss": 1.2893,
      "step": 195
    },
    {
      "epoch": 0.13047096022632718,
      "grad_norm": 91.55975341796875,
      "learning_rate": 0.000191651865008881,
      "loss": 1.025,
      "step": 196
    },
    {
      "epoch": 0.13113662839074722,
      "grad_norm": 111.15852355957031,
      "learning_rate": 0.000191607460035524,
      "loss": 0.4063,
      "step": 197
    },
    {
      "epoch": 0.13180229655516726,
      "grad_norm": 90.34394073486328,
      "learning_rate": 0.00019156305506216697,
      "loss": 0.6954,
      "step": 198
    },
    {
      "epoch": 0.13246796471958727,
      "grad_norm": 95.49404907226562,
      "learning_rate": 0.00019151865008880996,
      "loss": 0.6267,
      "step": 199
    },
    {
      "epoch": 0.1331336328840073,
      "grad_norm": 41.02389144897461,
      "learning_rate": 0.00019147424511545294,
      "loss": 0.2385,
      "step": 200
    },
    {
      "epoch": 0.13379930104842735,
      "grad_norm": 91.21604919433594,
      "learning_rate": 0.00019142984014209593,
      "loss": 0.4531,
      "step": 201
    },
    {
      "epoch": 0.1344649692128474,
      "grad_norm": 49.900184631347656,
      "learning_rate": 0.0001913854351687389,
      "loss": 0.1657,
      "step": 202
    },
    {
      "epoch": 0.13513063737726744,
      "grad_norm": 144.78623962402344,
      "learning_rate": 0.0001913410301953819,
      "loss": 0.7326,
      "step": 203
    },
    {
      "epoch": 0.13579630554168748,
      "grad_norm": 143.92132568359375,
      "learning_rate": 0.00019129662522202487,
      "loss": 1.0342,
      "step": 204
    },
    {
      "epoch": 0.1364619737061075,
      "grad_norm": 107.93486022949219,
      "learning_rate": 0.00019125222024866787,
      "loss": 0.3643,
      "step": 205
    },
    {
      "epoch": 0.13712764187052753,
      "grad_norm": 69.4767074584961,
      "learning_rate": 0.00019120781527531084,
      "loss": 0.4302,
      "step": 206
    },
    {
      "epoch": 0.13779331003494757,
      "grad_norm": 102.22624206542969,
      "learning_rate": 0.00019116341030195384,
      "loss": 0.7252,
      "step": 207
    },
    {
      "epoch": 0.13845897819936762,
      "grad_norm": 39.15114974975586,
      "learning_rate": 0.00019111900532859681,
      "loss": 0.376,
      "step": 208
    },
    {
      "epoch": 0.13912464636378766,
      "grad_norm": 65.12492370605469,
      "learning_rate": 0.00019107460035523979,
      "loss": 0.6005,
      "step": 209
    },
    {
      "epoch": 0.1397903145282077,
      "grad_norm": 95.04318237304688,
      "learning_rate": 0.00019103019538188278,
      "loss": 0.8371,
      "step": 210
    },
    {
      "epoch": 0.1404559826926277,
      "grad_norm": 119.12702941894531,
      "learning_rate": 0.00019098579040852576,
      "loss": 0.5701,
      "step": 211
    },
    {
      "epoch": 0.14112165085704775,
      "grad_norm": 56.38259506225586,
      "learning_rate": 0.00019094138543516875,
      "loss": 0.6285,
      "step": 212
    },
    {
      "epoch": 0.1417873190214678,
      "grad_norm": 29.10639762878418,
      "learning_rate": 0.00019089698046181172,
      "loss": 0.4639,
      "step": 213
    },
    {
      "epoch": 0.14245298718588784,
      "grad_norm": 60.445865631103516,
      "learning_rate": 0.00019085257548845472,
      "loss": 0.6099,
      "step": 214
    },
    {
      "epoch": 0.14311865535030788,
      "grad_norm": 83.33357238769531,
      "learning_rate": 0.0001908081705150977,
      "loss": 0.6992,
      "step": 215
    },
    {
      "epoch": 0.14378432351472792,
      "grad_norm": 50.539451599121094,
      "learning_rate": 0.0001907637655417407,
      "loss": 0.673,
      "step": 216
    },
    {
      "epoch": 0.14444999167914793,
      "grad_norm": 108.4279556274414,
      "learning_rate": 0.00019071936056838366,
      "loss": 0.7867,
      "step": 217
    },
    {
      "epoch": 0.14511565984356797,
      "grad_norm": 71.25292205810547,
      "learning_rate": 0.00019067495559502666,
      "loss": 0.2414,
      "step": 218
    },
    {
      "epoch": 0.14578132800798801,
      "grad_norm": 96.37142944335938,
      "learning_rate": 0.00019063055062166963,
      "loss": 0.4556,
      "step": 219
    },
    {
      "epoch": 0.14644699617240806,
      "grad_norm": 105.84346008300781,
      "learning_rate": 0.00019058614564831263,
      "loss": 0.5288,
      "step": 220
    },
    {
      "epoch": 0.1471126643368281,
      "grad_norm": 53.66656494140625,
      "learning_rate": 0.0001905417406749556,
      "loss": 0.2665,
      "step": 221
    },
    {
      "epoch": 0.14777833250124814,
      "grad_norm": 73.43135070800781,
      "learning_rate": 0.00019049733570159857,
      "loss": 0.465,
      "step": 222
    },
    {
      "epoch": 0.14844400066566815,
      "grad_norm": 108.94127655029297,
      "learning_rate": 0.00019045293072824157,
      "loss": 0.6516,
      "step": 223
    },
    {
      "epoch": 0.1491096688300882,
      "grad_norm": 23.10163688659668,
      "learning_rate": 0.00019040852575488454,
      "loss": 0.1401,
      "step": 224
    },
    {
      "epoch": 0.14977533699450823,
      "grad_norm": 182.62322998046875,
      "learning_rate": 0.00019036412078152754,
      "loss": 0.7953,
      "step": 225
    },
    {
      "epoch": 0.15044100515892828,
      "grad_norm": 104.31300354003906,
      "learning_rate": 0.00019031971580817051,
      "loss": 0.5117,
      "step": 226
    },
    {
      "epoch": 0.15110667332334832,
      "grad_norm": 122.88026428222656,
      "learning_rate": 0.0001902753108348135,
      "loss": 0.8029,
      "step": 227
    },
    {
      "epoch": 0.15177234148776836,
      "grad_norm": 72.17605590820312,
      "learning_rate": 0.00019023090586145648,
      "loss": 0.3403,
      "step": 228
    },
    {
      "epoch": 0.15243800965218837,
      "grad_norm": 55.70381164550781,
      "learning_rate": 0.00019018650088809948,
      "loss": 0.2811,
      "step": 229
    },
    {
      "epoch": 0.1531036778166084,
      "grad_norm": 80.8816146850586,
      "learning_rate": 0.00019014209591474245,
      "loss": 1.0904,
      "step": 230
    },
    {
      "epoch": 0.15376934598102845,
      "grad_norm": 77.0992431640625,
      "learning_rate": 0.00019009769094138545,
      "loss": 0.8617,
      "step": 231
    },
    {
      "epoch": 0.1544350141454485,
      "grad_norm": 47.77476119995117,
      "learning_rate": 0.00019005328596802842,
      "loss": 0.3185,
      "step": 232
    },
    {
      "epoch": 0.15510068230986854,
      "grad_norm": 77.98711395263672,
      "learning_rate": 0.00019000888099467142,
      "loss": 0.2655,
      "step": 233
    },
    {
      "epoch": 0.15576635047428858,
      "grad_norm": 63.54255294799805,
      "learning_rate": 0.0001899644760213144,
      "loss": 0.3955,
      "step": 234
    },
    {
      "epoch": 0.1564320186387086,
      "grad_norm": 190.26271057128906,
      "learning_rate": 0.00018992007104795736,
      "loss": 1.2886,
      "step": 235
    },
    {
      "epoch": 0.15709768680312863,
      "grad_norm": 117.44766998291016,
      "learning_rate": 0.00018987566607460036,
      "loss": 0.7142,
      "step": 236
    },
    {
      "epoch": 0.15776335496754867,
      "grad_norm": 133.50717163085938,
      "learning_rate": 0.00018983126110124333,
      "loss": 0.8898,
      "step": 237
    },
    {
      "epoch": 0.15842902313196872,
      "grad_norm": 62.771507263183594,
      "learning_rate": 0.00018978685612788633,
      "loss": 0.3062,
      "step": 238
    },
    {
      "epoch": 0.15909469129638876,
      "grad_norm": 74.36737060546875,
      "learning_rate": 0.0001897424511545293,
      "loss": 0.5336,
      "step": 239
    },
    {
      "epoch": 0.1597603594608088,
      "grad_norm": 137.45458984375,
      "learning_rate": 0.0001896980461811723,
      "loss": 0.7025,
      "step": 240
    },
    {
      "epoch": 0.1604260276252288,
      "grad_norm": 54.84712600708008,
      "learning_rate": 0.00018965364120781527,
      "loss": 0.338,
      "step": 241
    },
    {
      "epoch": 0.16109169578964885,
      "grad_norm": 127.16575622558594,
      "learning_rate": 0.00018960923623445827,
      "loss": 1.0382,
      "step": 242
    },
    {
      "epoch": 0.1617573639540689,
      "grad_norm": 93.72176361083984,
      "learning_rate": 0.00018956483126110124,
      "loss": 1.0084,
      "step": 243
    },
    {
      "epoch": 0.16242303211848894,
      "grad_norm": 56.54580307006836,
      "learning_rate": 0.00018952042628774424,
      "loss": 0.5353,
      "step": 244
    },
    {
      "epoch": 0.16308870028290898,
      "grad_norm": 76.96385955810547,
      "learning_rate": 0.0001894760213143872,
      "loss": 0.397,
      "step": 245
    },
    {
      "epoch": 0.16375436844732902,
      "grad_norm": 91.15630340576172,
      "learning_rate": 0.00018943161634103018,
      "loss": 0.4533,
      "step": 246
    },
    {
      "epoch": 0.16442003661174903,
      "grad_norm": 103.4432373046875,
      "learning_rate": 0.00018938721136767318,
      "loss": 0.6514,
      "step": 247
    },
    {
      "epoch": 0.16508570477616907,
      "grad_norm": 60.81359100341797,
      "learning_rate": 0.00018934280639431615,
      "loss": 0.4884,
      "step": 248
    },
    {
      "epoch": 0.16575137294058911,
      "grad_norm": 46.649139404296875,
      "learning_rate": 0.00018929840142095915,
      "loss": 0.271,
      "step": 249
    },
    {
      "epoch": 0.16641704110500916,
      "grad_norm": 58.7072868347168,
      "learning_rate": 0.00018925399644760212,
      "loss": 0.27,
      "step": 250
    },
    {
      "epoch": 0.1670827092694292,
      "grad_norm": 79.08338928222656,
      "learning_rate": 0.00018920959147424512,
      "loss": 0.4671,
      "step": 251
    },
    {
      "epoch": 0.16774837743384924,
      "grad_norm": 98.57723236083984,
      "learning_rate": 0.0001891651865008881,
      "loss": 0.3235,
      "step": 252
    },
    {
      "epoch": 0.16841404559826925,
      "grad_norm": 131.836181640625,
      "learning_rate": 0.0001891207815275311,
      "loss": 0.4688,
      "step": 253
    },
    {
      "epoch": 0.1690797137626893,
      "grad_norm": 67.51493835449219,
      "learning_rate": 0.00018907637655417406,
      "loss": 0.6911,
      "step": 254
    },
    {
      "epoch": 0.16974538192710933,
      "grad_norm": 51.76738357543945,
      "learning_rate": 0.00018903197158081706,
      "loss": 0.1837,
      "step": 255
    },
    {
      "epoch": 0.17041105009152938,
      "grad_norm": 52.23995590209961,
      "learning_rate": 0.00018898756660746003,
      "loss": 0.71,
      "step": 256
    },
    {
      "epoch": 0.17107671825594942,
      "grad_norm": 49.984336853027344,
      "learning_rate": 0.00018894316163410303,
      "loss": 0.3221,
      "step": 257
    },
    {
      "epoch": 0.17174238642036946,
      "grad_norm": 48.08108901977539,
      "learning_rate": 0.000188898756660746,
      "loss": 0.3334,
      "step": 258
    },
    {
      "epoch": 0.17240805458478947,
      "grad_norm": 24.143003463745117,
      "learning_rate": 0.00018885435168738897,
      "loss": 0.2096,
      "step": 259
    },
    {
      "epoch": 0.1730737227492095,
      "grad_norm": 116.61510467529297,
      "learning_rate": 0.00018880994671403197,
      "loss": 0.8852,
      "step": 260
    },
    {
      "epoch": 0.17373939091362955,
      "grad_norm": 41.168052673339844,
      "learning_rate": 0.00018876554174067494,
      "loss": 0.2541,
      "step": 261
    },
    {
      "epoch": 0.1744050590780496,
      "grad_norm": 19.256343841552734,
      "learning_rate": 0.00018872113676731794,
      "loss": 0.2278,
      "step": 262
    },
    {
      "epoch": 0.17507072724246964,
      "grad_norm": 51.543418884277344,
      "learning_rate": 0.0001886767317939609,
      "loss": 0.2494,
      "step": 263
    },
    {
      "epoch": 0.17573639540688968,
      "grad_norm": 44.67826461791992,
      "learning_rate": 0.0001886323268206039,
      "loss": 0.3112,
      "step": 264
    },
    {
      "epoch": 0.1764020635713097,
      "grad_norm": 41.30339431762695,
      "learning_rate": 0.00018858792184724688,
      "loss": 0.5114,
      "step": 265
    },
    {
      "epoch": 0.17706773173572973,
      "grad_norm": 12.575431823730469,
      "learning_rate": 0.00018854351687388988,
      "loss": 0.0375,
      "step": 266
    },
    {
      "epoch": 0.17773339990014977,
      "grad_norm": 71.91178894042969,
      "learning_rate": 0.00018849911190053285,
      "loss": 0.5989,
      "step": 267
    },
    {
      "epoch": 0.17839906806456982,
      "grad_norm": 83.31620788574219,
      "learning_rate": 0.00018845470692717585,
      "loss": 0.51,
      "step": 268
    },
    {
      "epoch": 0.17906473622898986,
      "grad_norm": 83.13764190673828,
      "learning_rate": 0.00018841030195381885,
      "loss": 1.1146,
      "step": 269
    },
    {
      "epoch": 0.1797304043934099,
      "grad_norm": 151.9849853515625,
      "learning_rate": 0.00018836589698046182,
      "loss": 1.1219,
      "step": 270
    },
    {
      "epoch": 0.1803960725578299,
      "grad_norm": 121.70146179199219,
      "learning_rate": 0.00018832149200710482,
      "loss": 0.9397,
      "step": 271
    },
    {
      "epoch": 0.18106174072224995,
      "grad_norm": 51.62641906738281,
      "learning_rate": 0.0001882770870337478,
      "loss": 0.2968,
      "step": 272
    },
    {
      "epoch": 0.18172740888667,
      "grad_norm": 66.82881164550781,
      "learning_rate": 0.0001882326820603908,
      "loss": 0.5365,
      "step": 273
    },
    {
      "epoch": 0.18239307705109004,
      "grad_norm": 78.71537017822266,
      "learning_rate": 0.00018818827708703376,
      "loss": 0.8149,
      "step": 274
    },
    {
      "epoch": 0.18305874521551008,
      "grad_norm": 137.4581756591797,
      "learning_rate": 0.00018814387211367676,
      "loss": 0.6777,
      "step": 275
    },
    {
      "epoch": 0.18372441337993012,
      "grad_norm": 64.81407165527344,
      "learning_rate": 0.00018809946714031973,
      "loss": 0.7677,
      "step": 276
    },
    {
      "epoch": 0.18439008154435013,
      "grad_norm": 96.37416076660156,
      "learning_rate": 0.00018805506216696273,
      "loss": 1.332,
      "step": 277
    },
    {
      "epoch": 0.18505574970877017,
      "grad_norm": 47.11737823486328,
      "learning_rate": 0.0001880106571936057,
      "loss": 0.29,
      "step": 278
    },
    {
      "epoch": 0.18572141787319021,
      "grad_norm": 111.5100326538086,
      "learning_rate": 0.0001879662522202487,
      "loss": 0.5466,
      "step": 279
    },
    {
      "epoch": 0.18638708603761026,
      "grad_norm": 123.91747283935547,
      "learning_rate": 0.00018792184724689167,
      "loss": 0.6733,
      "step": 280
    },
    {
      "epoch": 0.1870527542020303,
      "grad_norm": 107.9887924194336,
      "learning_rate": 0.00018787744227353467,
      "loss": 0.4885,
      "step": 281
    },
    {
      "epoch": 0.18771842236645034,
      "grad_norm": 76.66865539550781,
      "learning_rate": 0.00018783303730017764,
      "loss": 0.6393,
      "step": 282
    },
    {
      "epoch": 0.18838409053087035,
      "grad_norm": 67.36553192138672,
      "learning_rate": 0.00018778863232682064,
      "loss": 0.4087,
      "step": 283
    },
    {
      "epoch": 0.1890497586952904,
      "grad_norm": 50.0120849609375,
      "learning_rate": 0.0001877442273534636,
      "loss": 0.3911,
      "step": 284
    },
    {
      "epoch": 0.18971542685971043,
      "grad_norm": 109.88864135742188,
      "learning_rate": 0.00018769982238010658,
      "loss": 0.6727,
      "step": 285
    },
    {
      "epoch": 0.19038109502413048,
      "grad_norm": 114.64707946777344,
      "learning_rate": 0.00018765541740674958,
      "loss": 0.8044,
      "step": 286
    },
    {
      "epoch": 0.19104676318855052,
      "grad_norm": 130.69219970703125,
      "learning_rate": 0.00018761101243339255,
      "loss": 0.6119,
      "step": 287
    },
    {
      "epoch": 0.19171243135297053,
      "grad_norm": 50.73808670043945,
      "learning_rate": 0.00018756660746003555,
      "loss": 0.4887,
      "step": 288
    },
    {
      "epoch": 0.19237809951739057,
      "grad_norm": 143.2826385498047,
      "learning_rate": 0.00018752220248667852,
      "loss": 0.8228,
      "step": 289
    },
    {
      "epoch": 0.1930437676818106,
      "grad_norm": 132.3501739501953,
      "learning_rate": 0.00018747779751332152,
      "loss": 0.808,
      "step": 290
    },
    {
      "epoch": 0.19370943584623065,
      "grad_norm": 74.29531860351562,
      "learning_rate": 0.0001874333925399645,
      "loss": 0.7815,
      "step": 291
    },
    {
      "epoch": 0.1943751040106507,
      "grad_norm": 81.61593627929688,
      "learning_rate": 0.00018738898756660749,
      "loss": 0.5732,
      "step": 292
    },
    {
      "epoch": 0.19504077217507074,
      "grad_norm": 86.1830825805664,
      "learning_rate": 0.00018734458259325046,
      "loss": 0.5226,
      "step": 293
    },
    {
      "epoch": 0.19570644033949075,
      "grad_norm": 23.495864868164062,
      "learning_rate": 0.00018730017761989346,
      "loss": 0.1476,
      "step": 294
    },
    {
      "epoch": 0.1963721085039108,
      "grad_norm": 45.22040939331055,
      "learning_rate": 0.00018725577264653643,
      "loss": 0.4335,
      "step": 295
    },
    {
      "epoch": 0.19703777666833083,
      "grad_norm": 102.33810424804688,
      "learning_rate": 0.0001872113676731794,
      "loss": 0.6778,
      "step": 296
    },
    {
      "epoch": 0.19770344483275087,
      "grad_norm": 90.49777221679688,
      "learning_rate": 0.0001871669626998224,
      "loss": 0.7652,
      "step": 297
    },
    {
      "epoch": 0.19836911299717092,
      "grad_norm": 14.38054084777832,
      "learning_rate": 0.00018712255772646537,
      "loss": 0.0598,
      "step": 298
    },
    {
      "epoch": 0.19903478116159096,
      "grad_norm": 24.415199279785156,
      "learning_rate": 0.00018707815275310837,
      "loss": 0.3666,
      "step": 299
    },
    {
      "epoch": 0.19970044932601097,
      "grad_norm": 37.869869232177734,
      "learning_rate": 0.00018703374777975134,
      "loss": 0.1511,
      "step": 300
    },
    {
      "epoch": 0.200366117490431,
      "grad_norm": 115.4464111328125,
      "learning_rate": 0.00018698934280639434,
      "loss": 0.7441,
      "step": 301
    },
    {
      "epoch": 0.20103178565485105,
      "grad_norm": 124.49742889404297,
      "learning_rate": 0.0001869449378330373,
      "loss": 0.6898,
      "step": 302
    },
    {
      "epoch": 0.2016974538192711,
      "grad_norm": 45.529197692871094,
      "learning_rate": 0.0001869005328596803,
      "loss": 0.2891,
      "step": 303
    },
    {
      "epoch": 0.20236312198369114,
      "grad_norm": 107.60458374023438,
      "learning_rate": 0.00018685612788632328,
      "loss": 0.6037,
      "step": 304
    },
    {
      "epoch": 0.20302879014811118,
      "grad_norm": 72.92047119140625,
      "learning_rate": 0.00018681172291296628,
      "loss": 0.8364,
      "step": 305
    },
    {
      "epoch": 0.2036944583125312,
      "grad_norm": 80.35264587402344,
      "learning_rate": 0.00018676731793960925,
      "loss": 0.6218,
      "step": 306
    },
    {
      "epoch": 0.20436012647695123,
      "grad_norm": 181.0091552734375,
      "learning_rate": 0.00018672291296625225,
      "loss": 0.8943,
      "step": 307
    },
    {
      "epoch": 0.20502579464137127,
      "grad_norm": 54.508758544921875,
      "learning_rate": 0.00018667850799289522,
      "loss": 0.3581,
      "step": 308
    },
    {
      "epoch": 0.20569146280579131,
      "grad_norm": 22.76518440246582,
      "learning_rate": 0.0001866341030195382,
      "loss": 0.0707,
      "step": 309
    },
    {
      "epoch": 0.20635713097021136,
      "grad_norm": 144.45138549804688,
      "learning_rate": 0.00018658969804618119,
      "loss": 1.199,
      "step": 310
    },
    {
      "epoch": 0.2070227991346314,
      "grad_norm": 51.15176010131836,
      "learning_rate": 0.00018654529307282416,
      "loss": 0.2966,
      "step": 311
    },
    {
      "epoch": 0.2076884672990514,
      "grad_norm": 84.0188980102539,
      "learning_rate": 0.00018650088809946716,
      "loss": 0.3478,
      "step": 312
    },
    {
      "epoch": 0.20835413546347145,
      "grad_norm": 74.65904235839844,
      "learning_rate": 0.00018645648312611013,
      "loss": 0.5906,
      "step": 313
    },
    {
      "epoch": 0.2090198036278915,
      "grad_norm": 76.40025329589844,
      "learning_rate": 0.00018641207815275313,
      "loss": 0.5313,
      "step": 314
    },
    {
      "epoch": 0.20968547179231153,
      "grad_norm": 67.46442413330078,
      "learning_rate": 0.0001863676731793961,
      "loss": 0.5431,
      "step": 315
    },
    {
      "epoch": 0.21035113995673158,
      "grad_norm": 153.51947021484375,
      "learning_rate": 0.0001863232682060391,
      "loss": 1.4613,
      "step": 316
    },
    {
      "epoch": 0.21101680812115162,
      "grad_norm": 141.46417236328125,
      "learning_rate": 0.00018627886323268207,
      "loss": 0.6173,
      "step": 317
    },
    {
      "epoch": 0.21168247628557163,
      "grad_norm": 117.94355773925781,
      "learning_rate": 0.00018623445825932506,
      "loss": 0.703,
      "step": 318
    },
    {
      "epoch": 0.21234814444999167,
      "grad_norm": 72.00027465820312,
      "learning_rate": 0.00018619005328596804,
      "loss": 0.3786,
      "step": 319
    },
    {
      "epoch": 0.2130138126144117,
      "grad_norm": 54.18503189086914,
      "learning_rate": 0.00018614564831261103,
      "loss": 0.3633,
      "step": 320
    },
    {
      "epoch": 0.21367948077883175,
      "grad_norm": 121.49830627441406,
      "learning_rate": 0.000186101243339254,
      "loss": 0.7459,
      "step": 321
    },
    {
      "epoch": 0.2143451489432518,
      "grad_norm": 176.50872802734375,
      "learning_rate": 0.00018605683836589698,
      "loss": 0.8166,
      "step": 322
    },
    {
      "epoch": 0.21501081710767184,
      "grad_norm": 151.5113067626953,
      "learning_rate": 0.00018601243339253998,
      "loss": 0.7919,
      "step": 323
    },
    {
      "epoch": 0.21567648527209185,
      "grad_norm": 84.63894653320312,
      "learning_rate": 0.00018596802841918295,
      "loss": 0.7106,
      "step": 324
    },
    {
      "epoch": 0.2163421534365119,
      "grad_norm": 102.90968322753906,
      "learning_rate": 0.00018592362344582595,
      "loss": 0.689,
      "step": 325
    },
    {
      "epoch": 0.21700782160093193,
      "grad_norm": 61.89885330200195,
      "learning_rate": 0.00018587921847246892,
      "loss": 0.3393,
      "step": 326
    },
    {
      "epoch": 0.21767348976535197,
      "grad_norm": 106.49290466308594,
      "learning_rate": 0.00018583481349911191,
      "loss": 0.3928,
      "step": 327
    },
    {
      "epoch": 0.21833915792977202,
      "grad_norm": 43.59565353393555,
      "learning_rate": 0.00018579040852575489,
      "loss": 0.3696,
      "step": 328
    },
    {
      "epoch": 0.21900482609419206,
      "grad_norm": 67.05834197998047,
      "learning_rate": 0.00018574600355239788,
      "loss": 0.8861,
      "step": 329
    },
    {
      "epoch": 0.21967049425861207,
      "grad_norm": 108.7596435546875,
      "learning_rate": 0.00018570159857904086,
      "loss": 1.1147,
      "step": 330
    },
    {
      "epoch": 0.2203361624230321,
      "grad_norm": 45.871212005615234,
      "learning_rate": 0.00018565719360568385,
      "loss": 0.2634,
      "step": 331
    },
    {
      "epoch": 0.22100183058745215,
      "grad_norm": 97.05673217773438,
      "learning_rate": 0.00018561278863232683,
      "loss": 0.7176,
      "step": 332
    },
    {
      "epoch": 0.2216674987518722,
      "grad_norm": 37.64607238769531,
      "learning_rate": 0.00018556838365896982,
      "loss": 0.405,
      "step": 333
    },
    {
      "epoch": 0.22233316691629224,
      "grad_norm": 82.15957641601562,
      "learning_rate": 0.0001855239786856128,
      "loss": 0.3589,
      "step": 334
    },
    {
      "epoch": 0.22299883508071228,
      "grad_norm": 48.32238006591797,
      "learning_rate": 0.00018547957371225577,
      "loss": 0.2946,
      "step": 335
    },
    {
      "epoch": 0.2236645032451323,
      "grad_norm": 33.3898811340332,
      "learning_rate": 0.00018543516873889876,
      "loss": 0.388,
      "step": 336
    },
    {
      "epoch": 0.22433017140955233,
      "grad_norm": 42.87260818481445,
      "learning_rate": 0.00018539076376554174,
      "loss": 0.2302,
      "step": 337
    },
    {
      "epoch": 0.22499583957397237,
      "grad_norm": 58.45392608642578,
      "learning_rate": 0.00018534635879218473,
      "loss": 0.6616,
      "step": 338
    },
    {
      "epoch": 0.22566150773839241,
      "grad_norm": 28.114885330200195,
      "learning_rate": 0.0001853019538188277,
      "loss": 0.1042,
      "step": 339
    },
    {
      "epoch": 0.22632717590281246,
      "grad_norm": 51.35744857788086,
      "learning_rate": 0.0001852575488454707,
      "loss": 0.4782,
      "step": 340
    },
    {
      "epoch": 0.2269928440672325,
      "grad_norm": 82.15337371826172,
      "learning_rate": 0.00018521314387211368,
      "loss": 0.3746,
      "step": 341
    },
    {
      "epoch": 0.2276585122316525,
      "grad_norm": 109.32012176513672,
      "learning_rate": 0.00018516873889875667,
      "loss": 0.6589,
      "step": 342
    },
    {
      "epoch": 0.22832418039607255,
      "grad_norm": 155.4207000732422,
      "learning_rate": 0.00018512433392539965,
      "loss": 1.1968,
      "step": 343
    },
    {
      "epoch": 0.2289898485604926,
      "grad_norm": 33.03417205810547,
      "learning_rate": 0.00018507992895204264,
      "loss": 0.3845,
      "step": 344
    },
    {
      "epoch": 0.22965551672491263,
      "grad_norm": 37.060585021972656,
      "learning_rate": 0.00018503552397868561,
      "loss": 0.1341,
      "step": 345
    },
    {
      "epoch": 0.23032118488933268,
      "grad_norm": 106.52214050292969,
      "learning_rate": 0.00018499111900532859,
      "loss": 0.7823,
      "step": 346
    },
    {
      "epoch": 0.23098685305375272,
      "grad_norm": 48.00297164916992,
      "learning_rate": 0.00018494671403197158,
      "loss": 0.6375,
      "step": 347
    },
    {
      "epoch": 0.23165252121817273,
      "grad_norm": 51.31806564331055,
      "learning_rate": 0.00018490230905861456,
      "loss": 0.4176,
      "step": 348
    },
    {
      "epoch": 0.23231818938259277,
      "grad_norm": 57.65751647949219,
      "learning_rate": 0.00018485790408525755,
      "loss": 0.2368,
      "step": 349
    },
    {
      "epoch": 0.2329838575470128,
      "grad_norm": 46.041908264160156,
      "learning_rate": 0.00018481349911190053,
      "loss": 0.4219,
      "step": 350
    },
    {
      "epoch": 0.23364952571143285,
      "grad_norm": 97.23726654052734,
      "learning_rate": 0.00018476909413854352,
      "loss": 0.8618,
      "step": 351
    },
    {
      "epoch": 0.2343151938758529,
      "grad_norm": 39.60933303833008,
      "learning_rate": 0.0001847246891651865,
      "loss": 0.2446,
      "step": 352
    },
    {
      "epoch": 0.23498086204027294,
      "grad_norm": 83.39789581298828,
      "learning_rate": 0.0001846802841918295,
      "loss": 0.5283,
      "step": 353
    },
    {
      "epoch": 0.23564653020469295,
      "grad_norm": 23.28626823425293,
      "learning_rate": 0.00018463587921847246,
      "loss": 0.2269,
      "step": 354
    },
    {
      "epoch": 0.236312198369113,
      "grad_norm": 75.43800354003906,
      "learning_rate": 0.00018459147424511546,
      "loss": 0.5959,
      "step": 355
    },
    {
      "epoch": 0.23697786653353303,
      "grad_norm": 33.72787094116211,
      "learning_rate": 0.00018454706927175843,
      "loss": 0.2795,
      "step": 356
    },
    {
      "epoch": 0.23764353469795307,
      "grad_norm": 56.4429817199707,
      "learning_rate": 0.00018450266429840143,
      "loss": 0.6191,
      "step": 357
    },
    {
      "epoch": 0.23830920286237312,
      "grad_norm": 97.20005798339844,
      "learning_rate": 0.0001844582593250444,
      "loss": 0.6252,
      "step": 358
    },
    {
      "epoch": 0.23897487102679316,
      "grad_norm": NaN,
      "learning_rate": 0.0001844582593250444,
      "loss": 1.1109,
      "step": 359
    },
    {
      "epoch": 0.23964053919121317,
      "grad_norm": 103.0977783203125,
      "learning_rate": 0.00018441385435168738,
      "loss": 0.5326,
      "step": 360
    },
    {
      "epoch": 0.2403062073556332,
      "grad_norm": 33.12345504760742,
      "learning_rate": 0.00018436944937833037,
      "loss": 0.4465,
      "step": 361
    },
    {
      "epoch": 0.24097187552005325,
      "grad_norm": 29.757606506347656,
      "learning_rate": 0.00018432504440497335,
      "loss": 0.3853,
      "step": 362
    },
    {
      "epoch": 0.2416375436844733,
      "grad_norm": 52.76167678833008,
      "learning_rate": 0.00018428063943161634,
      "loss": 0.3521,
      "step": 363
    },
    {
      "epoch": 0.24230321184889334,
      "grad_norm": 80.51199340820312,
      "learning_rate": 0.00018423623445825931,
      "loss": 0.4551,
      "step": 364
    },
    {
      "epoch": 0.24296888001331338,
      "grad_norm": 47.3682861328125,
      "learning_rate": 0.0001841918294849023,
      "loss": 0.2393,
      "step": 365
    },
    {
      "epoch": 0.2436345481777334,
      "grad_norm": 80.18375396728516,
      "learning_rate": 0.00018414742451154528,
      "loss": 0.9122,
      "step": 366
    },
    {
      "epoch": 0.24430021634215343,
      "grad_norm": 65.6572265625,
      "learning_rate": 0.00018410301953818828,
      "loss": 0.3345,
      "step": 367
    },
    {
      "epoch": 0.24496588450657347,
      "grad_norm": 108.21941375732422,
      "learning_rate": 0.00018405861456483125,
      "loss": 0.5405,
      "step": 368
    },
    {
      "epoch": 0.2456315526709935,
      "grad_norm": 57.63541030883789,
      "learning_rate": 0.00018401420959147425,
      "loss": 0.2502,
      "step": 369
    },
    {
      "epoch": 0.24629722083541356,
      "grad_norm": 35.54652786254883,
      "learning_rate": 0.00018396980461811722,
      "loss": 0.5846,
      "step": 370
    },
    {
      "epoch": 0.2469628889998336,
      "grad_norm": 55.58504867553711,
      "learning_rate": 0.00018392539964476022,
      "loss": 0.2154,
      "step": 371
    },
    {
      "epoch": 0.2476285571642536,
      "grad_norm": 55.62077331542969,
      "learning_rate": 0.0001838809946714032,
      "loss": 0.4984,
      "step": 372
    },
    {
      "epoch": 0.24829422532867365,
      "grad_norm": 87.00872039794922,
      "learning_rate": 0.0001838365896980462,
      "loss": 0.4538,
      "step": 373
    },
    {
      "epoch": 0.2489598934930937,
      "grad_norm": 46.597068786621094,
      "learning_rate": 0.00018379218472468916,
      "loss": 0.648,
      "step": 374
    },
    {
      "epoch": 0.24962556165751373,
      "grad_norm": 99.84630584716797,
      "learning_rate": 0.00018374777975133216,
      "loss": 0.7931,
      "step": 375
    },
    {
      "epoch": 0.25029122982193375,
      "grad_norm": 54.02254104614258,
      "learning_rate": 0.00018370337477797513,
      "loss": 0.2499,
      "step": 376
    },
    {
      "epoch": 0.2509568979863538,
      "grad_norm": 29.319000244140625,
      "learning_rate": 0.00018365896980461813,
      "loss": 0.4007,
      "step": 377
    },
    {
      "epoch": 0.25162256615077383,
      "grad_norm": 113.27352905273438,
      "learning_rate": 0.0001836145648312611,
      "loss": 0.7379,
      "step": 378
    },
    {
      "epoch": 0.25228823431519387,
      "grad_norm": 112.12903594970703,
      "learning_rate": 0.0001835701598579041,
      "loss": 0.6326,
      "step": 379
    },
    {
      "epoch": 0.2529539024796139,
      "grad_norm": 123.96865844726562,
      "learning_rate": 0.00018352575488454707,
      "loss": 0.5521,
      "step": 380
    },
    {
      "epoch": 0.25361957064403395,
      "grad_norm": 42.07058334350586,
      "learning_rate": 0.00018348134991119007,
      "loss": 0.4259,
      "step": 381
    },
    {
      "epoch": 0.254285238808454,
      "grad_norm": 55.002960205078125,
      "learning_rate": 0.00018343694493783304,
      "loss": 0.4797,
      "step": 382
    },
    {
      "epoch": 0.25495090697287404,
      "grad_norm": 33.028038024902344,
      "learning_rate": 0.00018339253996447604,
      "loss": 0.2893,
      "step": 383
    },
    {
      "epoch": 0.2556165751372941,
      "grad_norm": 54.003570556640625,
      "learning_rate": 0.000183348134991119,
      "loss": 0.4897,
      "step": 384
    },
    {
      "epoch": 0.2562822433017141,
      "grad_norm": 80.12110137939453,
      "learning_rate": 0.000183303730017762,
      "loss": 0.2902,
      "step": 385
    },
    {
      "epoch": 0.2569479114661341,
      "grad_norm": 46.586753845214844,
      "learning_rate": 0.00018325932504440498,
      "loss": 0.3721,
      "step": 386
    },
    {
      "epoch": 0.25761357963055415,
      "grad_norm": 45.596553802490234,
      "learning_rate": 0.00018321492007104798,
      "loss": 0.2719,
      "step": 387
    },
    {
      "epoch": 0.2582792477949742,
      "grad_norm": 90.06937408447266,
      "learning_rate": 0.00018317051509769095,
      "loss": 0.6424,
      "step": 388
    },
    {
      "epoch": 0.25894491595939423,
      "grad_norm": 44.34123611450195,
      "learning_rate": 0.00018312611012433395,
      "loss": 0.2926,
      "step": 389
    },
    {
      "epoch": 0.25961058412381427,
      "grad_norm": 123.05774688720703,
      "learning_rate": 0.00018308170515097692,
      "loss": 0.8703,
      "step": 390
    },
    {
      "epoch": 0.2602762522882343,
      "grad_norm": 132.67784118652344,
      "learning_rate": 0.00018303730017761992,
      "loss": 0.492,
      "step": 391
    },
    {
      "epoch": 0.26094192045265435,
      "grad_norm": 65.75527954101562,
      "learning_rate": 0.0001829928952042629,
      "loss": 0.2371,
      "step": 392
    },
    {
      "epoch": 0.2616075886170744,
      "grad_norm": 70.54750061035156,
      "learning_rate": 0.0001829484902309059,
      "loss": 0.5192,
      "step": 393
    },
    {
      "epoch": 0.26227325678149443,
      "grad_norm": 85.00897216796875,
      "learning_rate": 0.00018290408525754886,
      "loss": 0.7114,
      "step": 394
    },
    {
      "epoch": 0.2629389249459145,
      "grad_norm": 96.78849792480469,
      "learning_rate": 0.00018285968028419186,
      "loss": 0.4515,
      "step": 395
    },
    {
      "epoch": 0.2636045931103345,
      "grad_norm": 21.13523292541504,
      "learning_rate": 0.00018281527531083483,
      "loss": 0.3526,
      "step": 396
    },
    {
      "epoch": 0.26427026127475456,
      "grad_norm": 29.27351188659668,
      "learning_rate": 0.0001827708703374778,
      "loss": 0.5151,
      "step": 397
    },
    {
      "epoch": 0.26493592943917454,
      "grad_norm": 35.63692855834961,
      "learning_rate": 0.0001827264653641208,
      "loss": 0.2374,
      "step": 398
    },
    {
      "epoch": 0.2656015976035946,
      "grad_norm": 66.08139038085938,
      "learning_rate": 0.00018268206039076377,
      "loss": 0.4133,
      "step": 399
    },
    {
      "epoch": 0.2662672657680146,
      "grad_norm": 34.3465461730957,
      "learning_rate": 0.00018263765541740677,
      "loss": 0.3082,
      "step": 400
    },
    {
      "epoch": 0.26693293393243467,
      "grad_norm": 100.62875366210938,
      "learning_rate": 0.00018259325044404974,
      "loss": 0.6884,
      "step": 401
    },
    {
      "epoch": 0.2675986020968547,
      "grad_norm": 139.31932067871094,
      "learning_rate": 0.00018254884547069274,
      "loss": 1.0101,
      "step": 402
    },
    {
      "epoch": 0.26826427026127475,
      "grad_norm": 118.92940521240234,
      "learning_rate": 0.0001825044404973357,
      "loss": 0.9706,
      "step": 403
    },
    {
      "epoch": 0.2689299384256948,
      "grad_norm": 151.63421630859375,
      "learning_rate": 0.0001824600355239787,
      "loss": 0.961,
      "step": 404
    },
    {
      "epoch": 0.26959560659011483,
      "grad_norm": 27.14082145690918,
      "learning_rate": 0.00018241563055062168,
      "loss": 0.227,
      "step": 405
    },
    {
      "epoch": 0.2702612747545349,
      "grad_norm": 80.59782409667969,
      "learning_rate": 0.00018237122557726468,
      "loss": 0.6828,
      "step": 406
    },
    {
      "epoch": 0.2709269429189549,
      "grad_norm": 49.46958541870117,
      "learning_rate": 0.00018232682060390765,
      "loss": 0.3184,
      "step": 407
    },
    {
      "epoch": 0.27159261108337496,
      "grad_norm": 73.98738098144531,
      "learning_rate": 0.00018228241563055065,
      "loss": 0.6858,
      "step": 408
    },
    {
      "epoch": 0.272258279247795,
      "grad_norm": 86.27637481689453,
      "learning_rate": 0.00018223801065719362,
      "loss": 0.5358,
      "step": 409
    },
    {
      "epoch": 0.272923947412215,
      "grad_norm": 145.42340087890625,
      "learning_rate": 0.0001821936056838366,
      "loss": 1.0992,
      "step": 410
    },
    {
      "epoch": 0.273589615576635,
      "grad_norm": 41.17599105834961,
      "learning_rate": 0.0001821492007104796,
      "loss": 0.369,
      "step": 411
    },
    {
      "epoch": 0.27425528374105507,
      "grad_norm": 55.49694061279297,
      "learning_rate": 0.00018210479573712256,
      "loss": 0.2842,
      "step": 412
    },
    {
      "epoch": 0.2749209519054751,
      "grad_norm": 47.01858139038086,
      "learning_rate": 0.00018206039076376556,
      "loss": 0.3368,
      "step": 413
    },
    {
      "epoch": 0.27558662006989515,
      "grad_norm": 40.550880432128906,
      "learning_rate": 0.00018201598579040853,
      "loss": 0.2099,
      "step": 414
    },
    {
      "epoch": 0.2762522882343152,
      "grad_norm": 38.81869125366211,
      "learning_rate": 0.00018197158081705153,
      "loss": 0.28,
      "step": 415
    },
    {
      "epoch": 0.27691795639873523,
      "grad_norm": 65.78675842285156,
      "learning_rate": 0.0001819271758436945,
      "loss": 0.521,
      "step": 416
    },
    {
      "epoch": 0.2775836245631553,
      "grad_norm": 73.13219451904297,
      "learning_rate": 0.0001818827708703375,
      "loss": 0.7937,
      "step": 417
    },
    {
      "epoch": 0.2782492927275753,
      "grad_norm": 65.54695129394531,
      "learning_rate": 0.00018183836589698047,
      "loss": 0.4079,
      "step": 418
    },
    {
      "epoch": 0.27891496089199536,
      "grad_norm": 34.30622863769531,
      "learning_rate": 0.00018179396092362347,
      "loss": 0.0964,
      "step": 419
    },
    {
      "epoch": 0.2795806290564154,
      "grad_norm": 67.12702941894531,
      "learning_rate": 0.00018174955595026644,
      "loss": 1.4374,
      "step": 420
    },
    {
      "epoch": 0.28024629722083544,
      "grad_norm": 33.8336181640625,
      "learning_rate": 0.00018170515097690944,
      "loss": 0.1509,
      "step": 421
    },
    {
      "epoch": 0.2809119653852554,
      "grad_norm": 66.7548599243164,
      "learning_rate": 0.0001816607460035524,
      "loss": 0.2623,
      "step": 422
    },
    {
      "epoch": 0.28157763354967547,
      "grad_norm": 57.463253021240234,
      "learning_rate": 0.00018161634103019538,
      "loss": 0.3713,
      "step": 423
    },
    {
      "epoch": 0.2822433017140955,
      "grad_norm": 32.50680923461914,
      "learning_rate": 0.00018157193605683838,
      "loss": 0.1945,
      "step": 424
    },
    {
      "epoch": 0.28290896987851555,
      "grad_norm": 72.0036849975586,
      "learning_rate": 0.00018152753108348135,
      "loss": 0.3324,
      "step": 425
    },
    {
      "epoch": 0.2835746380429356,
      "grad_norm": 38.52781295776367,
      "learning_rate": 0.00018148312611012435,
      "loss": 0.185,
      "step": 426
    },
    {
      "epoch": 0.28424030620735563,
      "grad_norm": 74.13616943359375,
      "learning_rate": 0.00018143872113676732,
      "loss": 0.2684,
      "step": 427
    },
    {
      "epoch": 0.28490597437177567,
      "grad_norm": 76.92218017578125,
      "learning_rate": 0.00018139431616341032,
      "loss": 0.1456,
      "step": 428
    },
    {
      "epoch": 0.2855716425361957,
      "grad_norm": 131.1004638671875,
      "learning_rate": 0.0001813499111900533,
      "loss": 0.6252,
      "step": 429
    },
    {
      "epoch": 0.28623731070061575,
      "grad_norm": 112.32926940917969,
      "learning_rate": 0.0001813055062166963,
      "loss": 0.8337,
      "step": 430
    },
    {
      "epoch": 0.2869029788650358,
      "grad_norm": 125.7912368774414,
      "learning_rate": 0.00018126110124333926,
      "loss": 0.3741,
      "step": 431
    },
    {
      "epoch": 0.28756864702945584,
      "grad_norm": 164.3785858154297,
      "learning_rate": 0.00018121669626998226,
      "loss": 0.9616,
      "step": 432
    },
    {
      "epoch": 0.2882343151938759,
      "grad_norm": 191.42367553710938,
      "learning_rate": 0.00018117229129662523,
      "loss": 0.8257,
      "step": 433
    },
    {
      "epoch": 0.28889998335829586,
      "grad_norm": 79.08208465576172,
      "learning_rate": 0.00018112788632326823,
      "loss": 0.5202,
      "step": 434
    },
    {
      "epoch": 0.2895656515227159,
      "grad_norm": 109.46979522705078,
      "learning_rate": 0.0001810834813499112,
      "loss": 0.941,
      "step": 435
    },
    {
      "epoch": 0.29023131968713595,
      "grad_norm": 74.53866577148438,
      "learning_rate": 0.00018103907637655417,
      "loss": 0.7279,
      "step": 436
    },
    {
      "epoch": 0.290896987851556,
      "grad_norm": 137.80848693847656,
      "learning_rate": 0.00018099467140319717,
      "loss": 1.4923,
      "step": 437
    },
    {
      "epoch": 0.29156265601597603,
      "grad_norm": 126.38627624511719,
      "learning_rate": 0.00018095026642984014,
      "loss": 0.721,
      "step": 438
    },
    {
      "epoch": 0.29222832418039607,
      "grad_norm": 98.57434844970703,
      "learning_rate": 0.00018090586145648314,
      "loss": 0.5846,
      "step": 439
    },
    {
      "epoch": 0.2928939923448161,
      "grad_norm": 33.60892868041992,
      "learning_rate": 0.0001808614564831261,
      "loss": 0.4007,
      "step": 440
    },
    {
      "epoch": 0.29355966050923615,
      "grad_norm": 84.4782943725586,
      "learning_rate": 0.0001808170515097691,
      "loss": 0.6742,
      "step": 441
    },
    {
      "epoch": 0.2942253286736562,
      "grad_norm": 78.77970886230469,
      "learning_rate": 0.00018077264653641208,
      "loss": 0.5823,
      "step": 442
    },
    {
      "epoch": 0.29489099683807624,
      "grad_norm": 157.4933319091797,
      "learning_rate": 0.00018072824156305508,
      "loss": 0.7795,
      "step": 443
    },
    {
      "epoch": 0.2955566650024963,
      "grad_norm": 18.716060638427734,
      "learning_rate": 0.00018068383658969805,
      "loss": 0.3715,
      "step": 444
    },
    {
      "epoch": 0.2962223331669163,
      "grad_norm": 85.15080261230469,
      "learning_rate": 0.00018063943161634105,
      "loss": 0.606,
      "step": 445
    },
    {
      "epoch": 0.2968880013313363,
      "grad_norm": 83.07491302490234,
      "learning_rate": 0.00018059502664298402,
      "loss": 0.4614,
      "step": 446
    },
    {
      "epoch": 0.29755366949575635,
      "grad_norm": 52.511085510253906,
      "learning_rate": 0.000180550621669627,
      "loss": 0.4219,
      "step": 447
    },
    {
      "epoch": 0.2982193376601764,
      "grad_norm": 126.2627944946289,
      "learning_rate": 0.00018050621669627,
      "loss": 0.6889,
      "step": 448
    },
    {
      "epoch": 0.29888500582459643,
      "grad_norm": 73.53312683105469,
      "learning_rate": 0.00018046181172291296,
      "loss": 0.577,
      "step": 449
    },
    {
      "epoch": 0.29955067398901647,
      "grad_norm": 103.77356719970703,
      "learning_rate": 0.00018041740674955596,
      "loss": 0.7924,
      "step": 450
    },
    {
      "epoch": 0.3002163421534365,
      "grad_norm": 89.16803741455078,
      "learning_rate": 0.00018037300177619893,
      "loss": 0.5117,
      "step": 451
    },
    {
      "epoch": 0.30088201031785655,
      "grad_norm": 116.20187377929688,
      "learning_rate": 0.00018032859680284193,
      "loss": 0.6893,
      "step": 452
    },
    {
      "epoch": 0.3015476784822766,
      "grad_norm": 130.5649871826172,
      "learning_rate": 0.0001802841918294849,
      "loss": 0.6782,
      "step": 453
    },
    {
      "epoch": 0.30221334664669663,
      "grad_norm": 71.50609588623047,
      "learning_rate": 0.0001802397868561279,
      "loss": 0.4608,
      "step": 454
    },
    {
      "epoch": 0.3028790148111167,
      "grad_norm": 101.877197265625,
      "learning_rate": 0.00018019538188277087,
      "loss": 0.5512,
      "step": 455
    },
    {
      "epoch": 0.3035446829755367,
      "grad_norm": 135.80545043945312,
      "learning_rate": 0.00018015097690941387,
      "loss": 0.6684,
      "step": 456
    },
    {
      "epoch": 0.30421035113995676,
      "grad_norm": 83.38865661621094,
      "learning_rate": 0.00018010657193605684,
      "loss": 0.4582,
      "step": 457
    },
    {
      "epoch": 0.30487601930437674,
      "grad_norm": 93.57701110839844,
      "learning_rate": 0.00018006216696269984,
      "loss": 0.7277,
      "step": 458
    },
    {
      "epoch": 0.3055416874687968,
      "grad_norm": 72.99111938476562,
      "learning_rate": 0.0001800177619893428,
      "loss": 0.5737,
      "step": 459
    },
    {
      "epoch": 0.3062073556332168,
      "grad_norm": 64.9205551147461,
      "learning_rate": 0.00017997335701598578,
      "loss": 0.3831,
      "step": 460
    },
    {
      "epoch": 0.30687302379763687,
      "grad_norm": 36.749935150146484,
      "learning_rate": 0.00017992895204262878,
      "loss": 0.4547,
      "step": 461
    },
    {
      "epoch": 0.3075386919620569,
      "grad_norm": 77.0653076171875,
      "learning_rate": 0.00017988454706927175,
      "loss": 0.8629,
      "step": 462
    },
    {
      "epoch": 0.30820436012647695,
      "grad_norm": 48.22278594970703,
      "learning_rate": 0.00017984014209591475,
      "loss": 0.3634,
      "step": 463
    },
    {
      "epoch": 0.308870028290897,
      "grad_norm": 39.0225715637207,
      "learning_rate": 0.00017979573712255772,
      "loss": 0.317,
      "step": 464
    },
    {
      "epoch": 0.30953569645531703,
      "grad_norm": 68.747314453125,
      "learning_rate": 0.00017975133214920072,
      "loss": 0.3624,
      "step": 465
    },
    {
      "epoch": 0.3102013646197371,
      "grad_norm": 71.55371856689453,
      "learning_rate": 0.0001797069271758437,
      "loss": 0.3992,
      "step": 466
    },
    {
      "epoch": 0.3108670327841571,
      "grad_norm": 36.87495422363281,
      "learning_rate": 0.00017966252220248669,
      "loss": 0.2098,
      "step": 467
    },
    {
      "epoch": 0.31153270094857716,
      "grad_norm": 10.664894104003906,
      "learning_rate": 0.00017961811722912966,
      "loss": 0.0542,
      "step": 468
    },
    {
      "epoch": 0.3121983691129972,
      "grad_norm": 34.32514953613281,
      "learning_rate": 0.00017957371225577265,
      "loss": 0.3147,
      "step": 469
    },
    {
      "epoch": 0.3128640372774172,
      "grad_norm": 24.026748657226562,
      "learning_rate": 0.00017952930728241563,
      "loss": 0.0797,
      "step": 470
    },
    {
      "epoch": 0.3135297054418372,
      "grad_norm": 23.764530181884766,
      "learning_rate": 0.00017948490230905862,
      "loss": 0.0965,
      "step": 471
    },
    {
      "epoch": 0.31419537360625727,
      "grad_norm": 5.237178325653076,
      "learning_rate": 0.0001794404973357016,
      "loss": 0.0106,
      "step": 472
    },
    {
      "epoch": 0.3148610417706773,
      "grad_norm": 20.264196395874023,
      "learning_rate": 0.0001793960923623446,
      "loss": 0.0546,
      "step": 473
    },
    {
      "epoch": 0.31552670993509735,
      "grad_norm": 79.56272888183594,
      "learning_rate": 0.00017935168738898757,
      "loss": 0.5564,
      "step": 474
    },
    {
      "epoch": 0.3161923780995174,
      "grad_norm": 37.59955596923828,
      "learning_rate": 0.00017930728241563056,
      "loss": 0.3443,
      "step": 475
    },
    {
      "epoch": 0.31685804626393743,
      "grad_norm": 49.22467803955078,
      "learning_rate": 0.00017926287744227354,
      "loss": 0.3749,
      "step": 476
    },
    {
      "epoch": 0.3175237144283575,
      "grad_norm": 175.10850524902344,
      "learning_rate": 0.00017921847246891653,
      "loss": 1.7856,
      "step": 477
    },
    {
      "epoch": 0.3181893825927775,
      "grad_norm": 47.71009826660156,
      "learning_rate": 0.0001791740674955595,
      "loss": 0.7082,
      "step": 478
    },
    {
      "epoch": 0.31885505075719756,
      "grad_norm": 82.27182006835938,
      "learning_rate": 0.0001791296625222025,
      "loss": 0.9351,
      "step": 479
    },
    {
      "epoch": 0.3195207189216176,
      "grad_norm": 79.33952331542969,
      "learning_rate": 0.00017908525754884547,
      "loss": 0.3707,
      "step": 480
    },
    {
      "epoch": 0.3201863870860376,
      "grad_norm": 88.69058227539062,
      "learning_rate": 0.00017904085257548847,
      "loss": 0.85,
      "step": 481
    },
    {
      "epoch": 0.3208520552504576,
      "grad_norm": 110.95427703857422,
      "learning_rate": 0.00017899644760213144,
      "loss": 0.3507,
      "step": 482
    },
    {
      "epoch": 0.32151772341487767,
      "grad_norm": 101.72648620605469,
      "learning_rate": 0.00017895204262877444,
      "loss": 0.8123,
      "step": 483
    },
    {
      "epoch": 0.3221833915792977,
      "grad_norm": 172.9845733642578,
      "learning_rate": 0.00017890763765541741,
      "loss": 1.1193,
      "step": 484
    },
    {
      "epoch": 0.32284905974371775,
      "grad_norm": 40.887081146240234,
      "learning_rate": 0.0001788632326820604,
      "loss": 0.1791,
      "step": 485
    },
    {
      "epoch": 0.3235147279081378,
      "grad_norm": 89.45999145507812,
      "learning_rate": 0.00017881882770870338,
      "loss": 0.6104,
      "step": 486
    },
    {
      "epoch": 0.32418039607255783,
      "grad_norm": 101.5306167602539,
      "learning_rate": 0.00017877442273534638,
      "loss": 0.5197,
      "step": 487
    },
    {
      "epoch": 0.32484606423697787,
      "grad_norm": 25.401044845581055,
      "learning_rate": 0.00017873001776198935,
      "loss": 0.2328,
      "step": 488
    },
    {
      "epoch": 0.3255117324013979,
      "grad_norm": 122.61588287353516,
      "learning_rate": 0.00017868561278863235,
      "loss": 0.9357,
      "step": 489
    },
    {
      "epoch": 0.32617740056581795,
      "grad_norm": 51.50986099243164,
      "learning_rate": 0.00017864120781527532,
      "loss": 0.506,
      "step": 490
    },
    {
      "epoch": 0.326843068730238,
      "grad_norm": 64.19971466064453,
      "learning_rate": 0.00017859680284191832,
      "loss": 0.2713,
      "step": 491
    },
    {
      "epoch": 0.32750873689465804,
      "grad_norm": 19.61275863647461,
      "learning_rate": 0.0001785523978685613,
      "loss": 0.1013,
      "step": 492
    },
    {
      "epoch": 0.328174405059078,
      "grad_norm": 66.36785125732422,
      "learning_rate": 0.0001785079928952043,
      "loss": 0.3726,
      "step": 493
    },
    {
      "epoch": 0.32884007322349806,
      "grad_norm": 87.43013763427734,
      "learning_rate": 0.00017846358792184726,
      "loss": 0.4851,
      "step": 494
    },
    {
      "epoch": 0.3295057413879181,
      "grad_norm": 63.96004867553711,
      "learning_rate": 0.00017841918294849026,
      "loss": 0.2111,
      "step": 495
    },
    {
      "epoch": 0.33017140955233815,
      "grad_norm": 81.01036834716797,
      "learning_rate": 0.00017837477797513323,
      "loss": 0.6494,
      "step": 496
    },
    {
      "epoch": 0.3308370777167582,
      "grad_norm": 63.75148010253906,
      "learning_rate": 0.0001783303730017762,
      "loss": 0.3041,
      "step": 497
    },
    {
      "epoch": 0.33150274588117823,
      "grad_norm": 91.487060546875,
      "learning_rate": 0.0001782859680284192,
      "loss": 0.689,
      "step": 498
    },
    {
      "epoch": 0.33216841404559827,
      "grad_norm": 58.02032470703125,
      "learning_rate": 0.00017824156305506217,
      "loss": 0.5346,
      "step": 499
    },
    {
      "epoch": 0.3328340822100183,
      "grad_norm": 73.02722930908203,
      "learning_rate": 0.00017819715808170517,
      "loss": 0.541,
      "step": 500
    }
  ],
  "logging_steps": 1,
  "max_steps": 4506,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.22820987830272e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}