{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0257510729613735,
  "eval_steps": 500,
  "global_step": 240,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004291845493562232,
      "grad_norm": 0.48923006653785706,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 3.0401,
      "step": 1
    },
    {
      "epoch": 0.008583690987124463,
      "grad_norm": 0.4999135434627533,
      "learning_rate": 4.800000000000001e-06,
      "loss": 3.0045,
      "step": 2
    },
    {
      "epoch": 0.012875536480686695,
      "grad_norm": 0.5719347596168518,
      "learning_rate": 7.2e-06,
      "loss": 3.233,
      "step": 3
    },
    {
      "epoch": 0.017167381974248927,
      "grad_norm": 0.4990224242210388,
      "learning_rate": 9.600000000000001e-06,
      "loss": 3.0813,
      "step": 4
    },
    {
      "epoch": 0.02145922746781116,
      "grad_norm": 0.5448071360588074,
      "learning_rate": 1.2e-05,
      "loss": 3.1513,
      "step": 5
    },
    {
      "epoch": 0.02575107296137339,
      "grad_norm": 0.4316995441913605,
      "learning_rate": 1.44e-05,
      "loss": 2.9247,
      "step": 6
    },
    {
      "epoch": 0.030042918454935622,
      "grad_norm": 0.39892056584358215,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 3.0529,
      "step": 7
    },
    {
      "epoch": 0.034334763948497854,
      "grad_norm": 0.3437541723251343,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 2.742,
      "step": 8
    },
    {
      "epoch": 0.03862660944206009,
      "grad_norm": 0.3232952356338501,
      "learning_rate": 2.16e-05,
      "loss": 2.8807,
      "step": 9
    },
    {
      "epoch": 0.04291845493562232,
      "grad_norm": 0.2316465526819229,
      "learning_rate": 2.4e-05,
      "loss": 2.8284,
      "step": 10
    },
    {
      "epoch": 0.04721030042918455,
      "grad_norm": 0.1895979940891266,
      "learning_rate": 2.64e-05,
      "loss": 2.846,
      "step": 11
    },
    {
      "epoch": 0.05150214592274678,
      "grad_norm": 0.18949951231479645,
      "learning_rate": 2.88e-05,
      "loss": 2.6129,
      "step": 12
    },
    {
      "epoch": 0.055793991416309016,
      "grad_norm": 0.29111406207084656,
      "learning_rate": 3.12e-05,
      "loss": 2.8553,
      "step": 13
    },
    {
      "epoch": 0.060085836909871244,
      "grad_norm": 0.3171645998954773,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 2.7663,
      "step": 14
    },
    {
      "epoch": 0.06437768240343347,
      "grad_norm": 0.3853575587272644,
      "learning_rate": 3.6e-05,
      "loss": 2.8468,
      "step": 15
    },
    {
      "epoch": 0.06866952789699571,
      "grad_norm": 0.33584126830101013,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 2.8286,
      "step": 16
    },
    {
      "epoch": 0.07296137339055794,
      "grad_norm": 0.22323380410671234,
      "learning_rate": 4.08e-05,
      "loss": 2.9605,
      "step": 17
    },
    {
      "epoch": 0.07725321888412018,
      "grad_norm": 0.20761723816394806,
      "learning_rate": 4.32e-05,
      "loss": 2.5953,
      "step": 18
    },
    {
      "epoch": 0.0815450643776824,
      "grad_norm": 0.18277060985565186,
      "learning_rate": 4.5600000000000004e-05,
      "loss": 2.5327,
      "step": 19
    },
    {
      "epoch": 0.08583690987124463,
      "grad_norm": 0.25170719623565674,
      "learning_rate": 4.8e-05,
      "loss": 2.7127,
      "step": 20
    },
    {
      "epoch": 0.09012875536480687,
      "grad_norm": 0.2634769678115845,
      "learning_rate": 5.04e-05,
      "loss": 2.7035,
      "step": 21
    },
    {
      "epoch": 0.0944206008583691,
      "grad_norm": 0.1749686598777771,
      "learning_rate": 5.28e-05,
      "loss": 2.7313,
      "step": 22
    },
    {
      "epoch": 0.09871244635193133,
      "grad_norm": 0.19000954926013947,
      "learning_rate": 5.520000000000001e-05,
      "loss": 2.6551,
      "step": 23
    },
    {
      "epoch": 0.10300429184549356,
      "grad_norm": 0.16048619151115417,
      "learning_rate": 5.76e-05,
      "loss": 2.6901,
      "step": 24
    },
    {
      "epoch": 0.1072961373390558,
      "grad_norm": 0.1346961259841919,
      "learning_rate": 6e-05,
      "loss": 2.6762,
      "step": 25
    },
    {
      "epoch": 0.11158798283261803,
      "grad_norm": 0.10837673395872116,
      "learning_rate": 5.9999238776847435e-05,
      "loss": 2.5797,
      "step": 26
    },
    {
      "epoch": 0.11587982832618025,
      "grad_norm": 0.12566141784191132,
      "learning_rate": 5.9996955146020456e-05,
      "loss": 2.6367,
      "step": 27
    },
    {
      "epoch": 0.12017167381974249,
      "grad_norm": 0.10796438157558441,
      "learning_rate": 5.999314922340924e-05,
      "loss": 2.6986,
      "step": 28
    },
    {
      "epoch": 0.12446351931330472,
      "grad_norm": 0.1174619048833847,
      "learning_rate": 5.9987821202157545e-05,
      "loss": 2.4532,
      "step": 29
    },
    {
      "epoch": 0.12875536480686695,
      "grad_norm": 0.12774387001991272,
      "learning_rate": 5.998097135265291e-05,
      "loss": 2.7154,
      "step": 30
    },
    {
      "epoch": 0.13304721030042918,
      "grad_norm": 0.12038439512252808,
      "learning_rate": 5.9972600022512946e-05,
      "loss": 2.6464,
      "step": 31
    },
    {
      "epoch": 0.13733905579399142,
      "grad_norm": 0.1207851693034172,
      "learning_rate": 5.996270763656767e-05,
      "loss": 2.745,
      "step": 32
    },
    {
      "epoch": 0.14163090128755365,
      "grad_norm": 0.12519803643226624,
      "learning_rate": 5.9951294696837966e-05,
      "loss": 2.4953,
      "step": 33
    },
    {
      "epoch": 0.1459227467811159,
      "grad_norm": 0.11793860048055649,
      "learning_rate": 5.993836178251009e-05,
      "loss": 2.5615,
      "step": 34
    },
    {
      "epoch": 0.15021459227467812,
      "grad_norm": 0.10079298913478851,
      "learning_rate": 5.99239095499063e-05,
      "loss": 2.5886,
      "step": 35
    },
    {
      "epoch": 0.15450643776824036,
      "grad_norm": 0.1022179052233696,
      "learning_rate": 5.990793873245154e-05,
      "loss": 2.4649,
      "step": 36
    },
    {
      "epoch": 0.15879828326180256,
      "grad_norm": 0.1409706175327301,
      "learning_rate": 5.989045014063621e-05,
      "loss": 2.7143,
      "step": 37
    },
    {
      "epoch": 0.1630901287553648,
      "grad_norm": 0.11522237211465836,
      "learning_rate": 5.9871444661975037e-05,
      "loss": 2.6908,
      "step": 38
    },
    {
      "epoch": 0.16738197424892703,
      "grad_norm": 0.11969179660081863,
      "learning_rate": 5.9850923260962045e-05,
      "loss": 2.6429,
      "step": 39
    },
    {
      "epoch": 0.17167381974248927,
      "grad_norm": 0.13146190345287323,
      "learning_rate": 5.982888697902161e-05,
      "loss": 2.5398,
      "step": 40
    },
    {
      "epoch": 0.1759656652360515,
      "grad_norm": 0.08940872550010681,
      "learning_rate": 5.98053369344556e-05,
      "loss": 2.6048,
      "step": 41
    },
    {
      "epoch": 0.18025751072961374,
      "grad_norm": 0.0957597866654396,
      "learning_rate": 5.978027432238662e-05,
      "loss": 2.758,
      "step": 42
    },
    {
      "epoch": 0.18454935622317598,
      "grad_norm": 0.12192820012569427,
      "learning_rate": 5.975370041469738e-05,
      "loss": 2.6317,
      "step": 43
    },
    {
      "epoch": 0.1888412017167382,
      "grad_norm": 0.11058337986469269,
      "learning_rate": 5.972561655996614e-05,
      "loss": 2.631,
      "step": 44
    },
    {
      "epoch": 0.19313304721030042,
      "grad_norm": 0.0976683646440506,
      "learning_rate": 5.969602418339825e-05,
      "loss": 2.5953,
      "step": 45
    },
    {
      "epoch": 0.19742489270386265,
      "grad_norm": 0.07907278090715408,
      "learning_rate": 5.966492478675385e-05,
      "loss": 2.5996,
      "step": 46
    },
    {
      "epoch": 0.2017167381974249,
      "grad_norm": 0.0848410427570343,
      "learning_rate": 5.963231994827169e-05,
      "loss": 2.5325,
      "step": 47
    },
    {
      "epoch": 0.20600858369098712,
      "grad_norm": 0.0901293084025383,
      "learning_rate": 5.9598211322588925e-05,
      "loss": 2.4015,
      "step": 48
    },
    {
      "epoch": 0.21030042918454936,
      "grad_norm": 0.09382897615432739,
      "learning_rate": 5.956260064065728e-05,
      "loss": 2.6818,
      "step": 49
    },
    {
      "epoch": 0.2145922746781116,
      "grad_norm": 0.10944745689630508,
      "learning_rate": 5.952548970965513e-05,
      "loss": 2.4901,
      "step": 50
    },
    {
      "epoch": 0.21888412017167383,
      "grad_norm": 0.1057049036026001,
      "learning_rate": 5.948688041289578e-05,
      "loss": 2.5959,
      "step": 51
    },
    {
      "epoch": 0.22317596566523606,
      "grad_norm": 0.10103785246610641,
      "learning_rate": 5.944677470973196e-05,
      "loss": 2.493,
      "step": 52
    },
    {
      "epoch": 0.22746781115879827,
      "grad_norm": 0.11647158861160278,
      "learning_rate": 5.9405174635456315e-05,
      "loss": 2.6496,
      "step": 53
    },
    {
      "epoch": 0.2317596566523605,
      "grad_norm": 0.09211481362581253,
      "learning_rate": 5.9362082301198156e-05,
      "loss": 2.6669,
      "step": 54
    },
    {
      "epoch": 0.23605150214592274,
      "grad_norm": 0.09014479070901871,
      "learning_rate": 5.931749989381632e-05,
      "loss": 2.6098,
      "step": 55
    },
    {
      "epoch": 0.24034334763948498,
      "grad_norm": 0.16046330332756042,
      "learning_rate": 5.9271429675788184e-05,
      "loss": 2.4713,
      "step": 56
    },
    {
      "epoch": 0.2446351931330472,
      "grad_norm": 0.1069481372833252,
      "learning_rate": 5.9223873985094866e-05,
      "loss": 2.3664,
      "step": 57
    },
    {
      "epoch": 0.24892703862660945,
      "grad_norm": 0.10389809310436249,
      "learning_rate": 5.9174835235102536e-05,
      "loss": 2.5278,
      "step": 58
    },
    {
      "epoch": 0.2532188841201717,
      "grad_norm": 0.11560378223657608,
      "learning_rate": 5.912431591443999e-05,
      "loss": 2.4859,
      "step": 59
    },
    {
      "epoch": 0.2575107296137339,
      "grad_norm": 0.09909158945083618,
      "learning_rate": 5.9072318586872344e-05,
      "loss": 2.4819,
      "step": 60
    },
    {
      "epoch": 0.26180257510729615,
      "grad_norm": 0.08729315549135208,
      "learning_rate": 5.901884589117089e-05,
      "loss": 2.5432,
      "step": 61
    },
    {
      "epoch": 0.26609442060085836,
      "grad_norm": 0.11550580710172653,
      "learning_rate": 5.896390054097922e-05,
      "loss": 2.592,
      "step": 62
    },
    {
      "epoch": 0.2703862660944206,
      "grad_norm": 0.10853075236082077,
      "learning_rate": 5.8907485324675545e-05,
      "loss": 2.3256,
      "step": 63
    },
    {
      "epoch": 0.27467811158798283,
      "grad_norm": 0.09386483579874039,
      "learning_rate": 5.884960310523109e-05,
      "loss": 2.676,
      "step": 64
    },
    {
      "epoch": 0.27896995708154504,
      "grad_norm": 0.12031152844429016,
      "learning_rate": 5.879025682006491e-05,
      "loss": 2.5849,
      "step": 65
    },
    {
      "epoch": 0.2832618025751073,
      "grad_norm": 0.09846463054418564,
      "learning_rate": 5.872944948089474e-05,
      "loss": 2.542,
      "step": 66
    },
    {
      "epoch": 0.2875536480686695,
      "grad_norm": 0.09881763160228729,
      "learning_rate": 5.8667184173584226e-05,
      "loss": 2.6389,
      "step": 67
    },
    {
      "epoch": 0.2918454935622318,
      "grad_norm": 0.12872399389743805,
      "learning_rate": 5.860346405798625e-05,
      "loss": 2.5906,
      "step": 68
    },
    {
      "epoch": 0.296137339055794,
      "grad_norm": 0.0927918404340744,
      "learning_rate": 5.853829236778266e-05,
      "loss": 2.3233,
      "step": 69
    },
    {
      "epoch": 0.30042918454935624,
      "grad_norm": 0.1097516193985939,
      "learning_rate": 5.847167241032006e-05,
      "loss": 2.3085,
      "step": 70
    },
    {
      "epoch": 0.30472103004291845,
      "grad_norm": 0.12274881452322006,
      "learning_rate": 5.8403607566442066e-05,
      "loss": 2.4758,
      "step": 71
    },
    {
      "epoch": 0.3090128755364807,
      "grad_norm": 0.10112845152616501,
      "learning_rate": 5.833410129031768e-05,
      "loss": 2.6385,
      "step": 72
    },
    {
      "epoch": 0.3133047210300429,
      "grad_norm": 0.107538603246212,
      "learning_rate": 5.8263157109266e-05,
      "loss": 2.4394,
      "step": 73
    },
    {
      "epoch": 0.31759656652360513,
      "grad_norm": 0.15851671993732452,
      "learning_rate": 5.819077862357725e-05,
      "loss": 2.348,
      "step": 74
    },
    {
      "epoch": 0.3218884120171674,
      "grad_norm": 0.09074151515960693,
      "learning_rate": 5.811696950633003e-05,
      "loss": 2.4884,
      "step": 75
    },
    {
      "epoch": 0.3261802575107296,
      "grad_norm": 0.10217540711164474,
      "learning_rate": 5.8041733503204934e-05,
      "loss": 2.5566,
      "step": 76
    },
    {
      "epoch": 0.33047210300429186,
      "grad_norm": 0.12177087366580963,
      "learning_rate": 5.796507443229445e-05,
      "loss": 2.4852,
      "step": 77
    },
    {
      "epoch": 0.33476394849785407,
      "grad_norm": 0.10884562134742737,
      "learning_rate": 5.788699618390924e-05,
      "loss": 2.6596,
      "step": 78
    },
    {
      "epoch": 0.33905579399141633,
      "grad_norm": 0.09898590296506882,
      "learning_rate": 5.7807502720380655e-05,
      "loss": 2.5299,
      "step": 79
    },
    {
      "epoch": 0.34334763948497854,
      "grad_norm": 0.16023777425289154,
      "learning_rate": 5.772659807585968e-05,
      "loss": 2.5859,
      "step": 80
    },
    {
      "epoch": 0.34763948497854075,
      "grad_norm": 0.1152929812669754,
      "learning_rate": 5.764428635611223e-05,
      "loss": 2.6133,
      "step": 81
    },
    {
      "epoch": 0.351931330472103,
      "grad_norm": 0.11540824174880981,
      "learning_rate": 5.756057173831075e-05,
      "loss": 2.4024,
      "step": 82
    },
    {
      "epoch": 0.3562231759656652,
      "grad_norm": 0.10451588034629822,
      "learning_rate": 5.7475458470822275e-05,
      "loss": 2.4877,
      "step": 83
    },
    {
      "epoch": 0.3605150214592275,
      "grad_norm": 0.09626314043998718,
      "learning_rate": 5.7388950872992764e-05,
      "loss": 2.67,
      "step": 84
    },
    {
      "epoch": 0.3648068669527897,
      "grad_norm": 0.11973975598812103,
      "learning_rate": 5.7301053334928e-05,
      "loss": 2.3961,
      "step": 85
    },
    {
      "epoch": 0.36909871244635195,
      "grad_norm": 0.09656506031751633,
      "learning_rate": 5.7211770317270696e-05,
      "loss": 2.5181,
      "step": 86
    },
    {
      "epoch": 0.37339055793991416,
      "grad_norm": 0.09045439958572388,
      "learning_rate": 5.712110635097422e-05,
      "loss": 2.6467,
      "step": 87
    },
    {
      "epoch": 0.3776824034334764,
      "grad_norm": 0.11997085064649582,
      "learning_rate": 5.702906603707257e-05,
      "loss": 2.6225,
      "step": 88
    },
    {
      "epoch": 0.38197424892703863,
      "grad_norm": 0.10198812931776047,
      "learning_rate": 5.6935654046446955e-05,
      "loss": 2.5756,
      "step": 89
    },
    {
      "epoch": 0.38626609442060084,
      "grad_norm": 0.12353700399398804,
      "learning_rate": 5.684087511958869e-05,
      "loss": 2.5561,
      "step": 90
    },
    {
      "epoch": 0.3905579399141631,
      "grad_norm": 0.09727780520915985,
      "learning_rate": 5.674473406635869e-05,
      "loss": 2.3674,
      "step": 91
    },
    {
      "epoch": 0.3948497854077253,
      "grad_norm": 0.13741353154182434,
      "learning_rate": 5.664723576574332e-05,
      "loss": 2.4249,
      "step": 92
    },
    {
      "epoch": 0.39914163090128757,
      "grad_norm": 0.09821099787950516,
      "learning_rate": 5.6548385165606835e-05,
      "loss": 2.5469,
      "step": 93
    },
    {
      "epoch": 0.4034334763948498,
      "grad_norm": 0.12704302370548248,
      "learning_rate": 5.644818728244027e-05,
      "loss": 2.5243,
      "step": 94
    },
    {
      "epoch": 0.40772532188841204,
      "grad_norm": 0.14363928139209747,
      "learning_rate": 5.634664720110686e-05,
      "loss": 2.3621,
      "step": 95
    },
    {
      "epoch": 0.41201716738197425,
      "grad_norm": 0.09331246465444565,
      "learning_rate": 5.6243770074583985e-05,
      "loss": 2.4098,
      "step": 96
    },
    {
      "epoch": 0.41630901287553645,
      "grad_norm": 0.13661201298236847,
      "learning_rate": 5.613956112370168e-05,
      "loss": 2.3998,
      "step": 97
    },
    {
      "epoch": 0.4206008583690987,
      "grad_norm": 0.15692031383514404,
      "learning_rate": 5.60340256368777e-05,
      "loss": 2.6108,
      "step": 98
    },
    {
      "epoch": 0.4248927038626609,
      "grad_norm": 0.1194879561662674,
      "learning_rate": 5.59271689698491e-05,
      "loss": 2.609,
      "step": 99
    },
    {
      "epoch": 0.4291845493562232,
      "grad_norm": 0.14594995975494385,
      "learning_rate": 5.581899654540049e-05,
      "loss": 2.3956,
      "step": 100
    },
    {
      "epoch": 0.4334763948497854,
      "grad_norm": 0.11137986928224564,
      "learning_rate": 5.570951385308879e-05,
      "loss": 2.6656,
      "step": 101
    },
    {
      "epoch": 0.43776824034334766,
      "grad_norm": 0.16587291657924652,
      "learning_rate": 5.559872644896467e-05,
      "loss": 2.4815,
      "step": 102
    },
    {
      "epoch": 0.44206008583690987,
      "grad_norm": 0.1582924723625183,
      "learning_rate": 5.548663995529062e-05,
      "loss": 2.4112,
      "step": 103
    },
    {
      "epoch": 0.44635193133047213,
      "grad_norm": 0.12027094513177872,
      "learning_rate": 5.5373260060255563e-05,
      "loss": 2.4611,
      "step": 104
    },
    {
      "epoch": 0.45064377682403434,
      "grad_norm": 0.17147387564182281,
      "learning_rate": 5.525859251768625e-05,
      "loss": 2.4084,
      "step": 105
    },
    {
      "epoch": 0.45493562231759654,
      "grad_norm": 0.138417050242424,
      "learning_rate": 5.5142643146755215e-05,
      "loss": 2.655,
      "step": 106
    },
    {
      "epoch": 0.4592274678111588,
      "grad_norm": 0.15900184214115143,
      "learning_rate": 5.5025417831685533e-05,
      "loss": 2.4029,
      "step": 107
    },
    {
      "epoch": 0.463519313304721,
      "grad_norm": 0.1893249750137329,
      "learning_rate": 5.4906922521452105e-05,
      "loss": 2.367,
      "step": 108
    },
    {
      "epoch": 0.4678111587982833,
      "grad_norm": 0.13178469240665436,
      "learning_rate": 5.478716322947985e-05,
      "loss": 2.4976,
      "step": 109
    },
    {
      "epoch": 0.4721030042918455,
      "grad_norm": 0.22058036923408508,
      "learning_rate": 5.466614603333848e-05,
      "loss": 2.4154,
      "step": 110
    },
    {
      "epoch": 0.47639484978540775,
      "grad_norm": 0.10874748975038528,
      "learning_rate": 5.4543877074434106e-05,
      "loss": 2.3568,
      "step": 111
    },
    {
      "epoch": 0.48068669527896996,
      "grad_norm": 0.1645105481147766,
      "learning_rate": 5.4420362557697546e-05,
      "loss": 2.5891,
      "step": 112
    },
    {
      "epoch": 0.48497854077253216,
      "grad_norm": 0.10352698713541031,
      "learning_rate": 5.429560875126946e-05,
      "loss": 2.2487,
      "step": 113
    },
    {
      "epoch": 0.4892703862660944,
      "grad_norm": 0.10524651408195496,
      "learning_rate": 5.4169621986182234e-05,
      "loss": 2.5208,
      "step": 114
    },
    {
      "epoch": 0.49356223175965663,
      "grad_norm": 0.10389380156993866,
      "learning_rate": 5.40424086560387e-05,
      "loss": 2.3373,
      "step": 115
    },
    {
      "epoch": 0.4978540772532189,
      "grad_norm": 0.11282224953174591,
      "learning_rate": 5.3913975216687675e-05,
      "loss": 2.4336,
      "step": 116
    },
    {
      "epoch": 0.5021459227467812,
      "grad_norm": 0.14653658866882324,
      "learning_rate": 5.378432818589633e-05,
      "loss": 2.5201,
      "step": 117
    },
    {
      "epoch": 0.5064377682403434,
      "grad_norm": 0.10697951167821884,
      "learning_rate": 5.365347414301943e-05,
      "loss": 2.5,
      "step": 118
    },
    {
      "epoch": 0.5107296137339056,
      "grad_norm": 0.08726029098033905,
      "learning_rate": 5.352141972866545e-05,
      "loss": 2.4833,
      "step": 119
    },
    {
      "epoch": 0.5150214592274678,
      "grad_norm": 0.1065029427409172,
      "learning_rate": 5.3388171644359565e-05,
      "loss": 2.3921,
      "step": 120
    },
    {
      "epoch": 0.51931330472103,
      "grad_norm": 0.12069697678089142,
      "learning_rate": 5.325373665220356e-05,
      "loss": 2.4335,
      "step": 121
    },
    {
      "epoch": 0.5236051502145923,
      "grad_norm": 0.11373002082109451,
      "learning_rate": 5.311812157453266e-05,
      "loss": 2.5685,
      "step": 122
    },
    {
      "epoch": 0.5278969957081545,
      "grad_norm": 0.13439607620239258,
      "learning_rate": 5.298133329356934e-05,
      "loss": 2.4198,
      "step": 123
    },
    {
      "epoch": 0.5321888412017167,
      "grad_norm": 0.10372010618448257,
      "learning_rate": 5.284337875107403e-05,
      "loss": 2.4413,
      "step": 124
    },
    {
      "epoch": 0.5364806866952789,
      "grad_norm": 0.1336512565612793,
      "learning_rate": 5.2704264947992855e-05,
      "loss": 2.5762,
      "step": 125
    },
    {
      "epoch": 0.5407725321888412,
      "grad_norm": 0.12671151757240295,
      "learning_rate": 5.256399894410232e-05,
      "loss": 2.5616,
      "step": 126
    },
    {
      "epoch": 0.5450643776824035,
      "grad_norm": 0.10133809596300125,
      "learning_rate": 5.242258785765106e-05,
      "loss": 2.3512,
      "step": 127
    },
    {
      "epoch": 0.5493562231759657,
      "grad_norm": 0.10729261487722397,
      "learning_rate": 5.228003886499863e-05,
      "loss": 2.4487,
      "step": 128
    },
    {
      "epoch": 0.5536480686695279,
      "grad_norm": 0.10313385725021362,
      "learning_rate": 5.213635920025127e-05,
      "loss": 2.3446,
      "step": 129
    },
    {
      "epoch": 0.5579399141630901,
      "grad_norm": 0.11684712022542953,
      "learning_rate": 5.1991556154894786e-05,
      "loss": 2.4377,
      "step": 130
    },
    {
      "epoch": 0.5622317596566524,
      "grad_norm": 0.12217995524406433,
      "learning_rate": 5.1845637077424576e-05,
      "loss": 2.4721,
      "step": 131
    },
    {
      "epoch": 0.5665236051502146,
      "grad_norm": 0.1254771202802658,
      "learning_rate": 5.169860937297264e-05,
      "loss": 2.3565,
      "step": 132
    },
    {
      "epoch": 0.5708154506437768,
      "grad_norm": 0.10808940976858139,
      "learning_rate": 5.155048050293183e-05,
      "loss": 2.464,
      "step": 133
    },
    {
      "epoch": 0.575107296137339,
      "grad_norm": 0.1136014387011528,
      "learning_rate": 5.140125798457716e-05,
      "loss": 2.4179,
      "step": 134
    },
    {
      "epoch": 0.5793991416309013,
      "grad_norm": 0.12022742629051208,
      "learning_rate": 5.125094939068439e-05,
      "loss": 2.6821,
      "step": 135
    },
    {
      "epoch": 0.5836909871244635,
      "grad_norm": 0.11005590111017227,
      "learning_rate": 5.109956234914558e-05,
      "loss": 2.5086,
      "step": 136
    },
    {
      "epoch": 0.5879828326180258,
      "grad_norm": 0.14693163335323334,
      "learning_rate": 5.0947104542582184e-05,
      "loss": 2.5458,
      "step": 137
    },
    {
      "epoch": 0.592274678111588,
      "grad_norm": 0.09363128244876862,
      "learning_rate": 5.0793583707954984e-05,
      "loss": 2.5145,
      "step": 138
    },
    {
      "epoch": 0.5965665236051502,
      "grad_norm": 0.19561366736888885,
      "learning_rate": 5.063900763617156e-05,
      "loss": 2.5379,
      "step": 139
    },
    {
      "epoch": 0.6008583690987125,
      "grad_norm": 0.10757976770401001,
      "learning_rate": 5.04833841716909e-05,
      "loss": 2.5076,
      "step": 140
    },
    {
      "epoch": 0.6051502145922747,
      "grad_norm": 0.13970068097114563,
      "learning_rate": 5.032672121212529e-05,
      "loss": 2.472,
      "step": 141
    },
    {
      "epoch": 0.6094420600858369,
      "grad_norm": 0.13635942339897156,
      "learning_rate": 5.0169026707839506e-05,
      "loss": 2.4703,
      "step": 142
    },
    {
      "epoch": 0.6137339055793991,
      "grad_norm": 0.10295607149600983,
      "learning_rate": 5.001030866154741e-05,
      "loss": 2.3987,
      "step": 143
    },
    {
      "epoch": 0.6180257510729614,
      "grad_norm": 0.1324918419122696,
      "learning_rate": 4.985057512790579e-05,
      "loss": 2.4456,
      "step": 144
    },
    {
      "epoch": 0.6223175965665236,
      "grad_norm": 0.10353874415159225,
      "learning_rate": 4.968983421310555e-05,
      "loss": 2.3827,
      "step": 145
    },
    {
      "epoch": 0.6266094420600858,
      "grad_norm": 0.10886164009571075,
      "learning_rate": 4.952809407446043e-05,
      "loss": 2.4664,
      "step": 146
    },
    {
      "epoch": 0.630901287553648,
      "grad_norm": 0.1082598865032196,
      "learning_rate": 4.9365362919992994e-05,
      "loss": 2.423,
      "step": 147
    },
    {
      "epoch": 0.6351931330472103,
      "grad_norm": 0.14654968678951263,
      "learning_rate": 4.9201649008018055e-05,
      "loss": 2.3452,
      "step": 148
    },
    {
      "epoch": 0.6394849785407726,
      "grad_norm": 0.10624393820762634,
      "learning_rate": 4.9036960646723617e-05,
      "loss": 2.4337,
      "step": 149
    },
    {
      "epoch": 0.6437768240343348,
      "grad_norm": 0.10619154572486877,
      "learning_rate": 4.887130619374927e-05,
      "loss": 2.2317,
      "step": 150
    },
    {
      "epoch": 0.648068669527897,
      "grad_norm": 0.1240013837814331,
      "learning_rate": 4.870469405576201e-05,
      "loss": 2.4641,
      "step": 151
    },
    {
      "epoch": 0.6523605150214592,
      "grad_norm": 0.11749018728733063,
      "learning_rate": 4.853713268802962e-05,
      "loss": 2.6394,
      "step": 152
    },
    {
      "epoch": 0.6566523605150214,
      "grad_norm": 0.11559685319662094,
      "learning_rate": 4.836863059399161e-05,
      "loss": 2.6346,
      "step": 153
    },
    {
      "epoch": 0.6609442060085837,
      "grad_norm": 0.12158619612455368,
      "learning_rate": 4.819919632482766e-05,
      "loss": 2.5024,
      "step": 154
    },
    {
      "epoch": 0.6652360515021459,
      "grad_norm": 0.10826972126960754,
      "learning_rate": 4.802883847902368e-05,
      "loss": 2.5089,
      "step": 155
    },
    {
      "epoch": 0.6695278969957081,
      "grad_norm": 0.12029755115509033,
      "learning_rate": 4.785756570193543e-05,
      "loss": 2.5017,
      "step": 156
    },
    {
      "epoch": 0.6738197424892703,
      "grad_norm": 0.11460699141025543,
      "learning_rate": 4.76853866853498e-05,
      "loss": 2.4295,
      "step": 157
    },
    {
      "epoch": 0.6781115879828327,
      "grad_norm": 0.10718906670808792,
      "learning_rate": 4.75123101670437e-05,
      "loss": 2.4886,
      "step": 158
    },
    {
      "epoch": 0.6824034334763949,
      "grad_norm": 0.11268241703510284,
      "learning_rate": 4.733834493034066e-05,
      "loss": 2.5896,
      "step": 159
    },
    {
      "epoch": 0.6866952789699571,
      "grad_norm": 0.12579713761806488,
      "learning_rate": 4.716349980366509e-05,
      "loss": 2.3132,
      "step": 160
    },
    {
      "epoch": 0.6909871244635193,
      "grad_norm": 0.11618609726428986,
      "learning_rate": 4.698778366009421e-05,
      "loss": 2.4478,
      "step": 161
    },
    {
      "epoch": 0.6952789699570815,
      "grad_norm": 0.12247326225042343,
      "learning_rate": 4.681120541690781e-05,
      "loss": 2.4638,
      "step": 162
    },
    {
      "epoch": 0.6995708154506438,
      "grad_norm": 0.11064168810844421,
      "learning_rate": 4.663377403513569e-05,
      "loss": 2.6396,
      "step": 163
    },
    {
      "epoch": 0.703862660944206,
      "grad_norm": 0.1138993427157402,
      "learning_rate": 4.64554985191029e-05,
      "loss": 2.3691,
      "step": 164
    },
    {
      "epoch": 0.7081545064377682,
      "grad_norm": 0.11352220177650452,
      "learning_rate": 4.6276387915972783e-05,
      "loss": 2.5472,
      "step": 165
    },
    {
      "epoch": 0.7124463519313304,
      "grad_norm": 0.13823212683200836,
      "learning_rate": 4.609645131528789e-05,
      "loss": 2.3696,
      "step": 166
    },
    {
      "epoch": 0.7167381974248928,
      "grad_norm": 0.16149644553661346,
      "learning_rate": 4.5915697848508645e-05,
      "loss": 2.3184,
      "step": 167
    },
    {
      "epoch": 0.721030042918455,
      "grad_norm": 0.10547155141830444,
      "learning_rate": 4.5734136688549964e-05,
      "loss": 2.4221,
      "step": 168
    },
    {
      "epoch": 0.7253218884120172,
      "grad_norm": 0.1176508441567421,
      "learning_rate": 4.555177704931576e-05,
      "loss": 2.4223,
      "step": 169
    },
    {
      "epoch": 0.7296137339055794,
      "grad_norm": 0.12838046252727509,
      "learning_rate": 4.5368628185231314e-05,
      "loss": 2.5376,
      "step": 170
    },
    {
      "epoch": 0.7339055793991416,
      "grad_norm": 0.1296073943376541,
      "learning_rate": 4.518469939077369e-05,
      "loss": 2.4453,
      "step": 171
    },
    {
      "epoch": 0.7381974248927039,
      "grad_norm": 0.13209278881549835,
      "learning_rate": 4.5e-05,
      "loss": 2.4902,
      "step": 172
    },
    {
      "epoch": 0.7424892703862661,
      "grad_norm": 0.13168583810329437,
      "learning_rate": 4.4814539386073744e-05,
      "loss": 2.2446,
      "step": 173
    },
    {
      "epoch": 0.7467811158798283,
      "grad_norm": 0.12592634558677673,
      "learning_rate": 4.462832696078915e-05,
      "loss": 2.5433,
      "step": 174
    },
    {
      "epoch": 0.7510729613733905,
      "grad_norm": 0.12793345749378204,
      "learning_rate": 4.4441372174093495e-05,
      "loss": 2.4155,
      "step": 175
    },
    {
      "epoch": 0.7553648068669528,
      "grad_norm": 0.12817440927028656,
      "learning_rate": 4.4253684513607585e-05,
      "loss": 2.4488,
      "step": 176
    },
    {
      "epoch": 0.759656652360515,
      "grad_norm": 0.12905055284500122,
      "learning_rate": 4.406527350414427e-05,
      "loss": 2.3792,
      "step": 177
    },
    {
      "epoch": 0.7639484978540773,
      "grad_norm": 0.11602860689163208,
      "learning_rate": 4.3876148707225067e-05,
      "loss": 2.2747,
      "step": 178
    },
    {
      "epoch": 0.7682403433476395,
      "grad_norm": 0.1343417763710022,
      "learning_rate": 4.368631972059489e-05,
      "loss": 2.3509,
      "step": 179
    },
    {
      "epoch": 0.7725321888412017,
      "grad_norm": 0.17558203637599945,
      "learning_rate": 4.349579617773507e-05,
      "loss": 2.2545,
      "step": 180
    },
    {
      "epoch": 0.776824034334764,
      "grad_norm": 0.11374380439519882,
      "learning_rate": 4.33045877473744e-05,
      "loss": 2.4592,
      "step": 181
    },
    {
      "epoch": 0.7811158798283262,
      "grad_norm": 0.16170468926429749,
      "learning_rate": 4.31127041329985e-05,
      "loss": 2.5294,
      "step": 182
    },
    {
      "epoch": 0.7854077253218884,
      "grad_norm": 0.1064993217587471,
      "learning_rate": 4.2920155072357335e-05,
      "loss": 2.4469,
      "step": 183
    },
    {
      "epoch": 0.7896995708154506,
      "grad_norm": 0.14201977849006653,
      "learning_rate": 4.2726950336971115e-05,
      "loss": 2.4416,
      "step": 184
    },
    {
      "epoch": 0.7939914163090128,
      "grad_norm": 0.12264000624418259,
      "learning_rate": 4.2533099731634376e-05,
      "loss": 2.5181,
      "step": 185
    },
    {
      "epoch": 0.7982832618025751,
      "grad_norm": 0.12580525875091553,
      "learning_rate": 4.233861309391835e-05,
      "loss": 2.3767,
      "step": 186
    },
    {
      "epoch": 0.8025751072961373,
      "grad_norm": 0.15109586715698242,
      "learning_rate": 4.214350029367181e-05,
      "loss": 2.3987,
      "step": 187
    },
    {
      "epoch": 0.8068669527896996,
      "grad_norm": 0.15967005491256714,
      "learning_rate": 4.1947771232520165e-05,
      "loss": 2.4359,
      "step": 188
    },
    {
      "epoch": 0.8111587982832618,
      "grad_norm": 0.10725586116313934,
      "learning_rate": 4.175143584336295e-05,
      "loss": 2.4294,
      "step": 189
    },
    {
      "epoch": 0.8154506437768241,
      "grad_norm": 0.12302330881357193,
      "learning_rate": 4.155450408986972e-05,
      "loss": 2.4109,
      "step": 190
    },
    {
      "epoch": 0.8197424892703863,
      "grad_norm": 0.13229899108409882,
      "learning_rate": 4.1356985965974536e-05,
      "loss": 2.5981,
      "step": 191
    },
    {
      "epoch": 0.8240343347639485,
      "grad_norm": 0.17274914681911469,
      "learning_rate": 4.115889149536863e-05,
      "loss": 2.3754,
      "step": 192
    },
    {
      "epoch": 0.8283261802575107,
      "grad_norm": 0.10760139673948288,
      "learning_rate": 4.0960230730991856e-05,
      "loss": 2.3867,
      "step": 193
    },
    {
      "epoch": 0.8326180257510729,
      "grad_norm": 0.17432305216789246,
      "learning_rate": 4.076101375452241e-05,
      "loss": 2.4265,
      "step": 194
    },
    {
      "epoch": 0.8369098712446352,
      "grad_norm": 0.15838249027729034,
      "learning_rate": 4.05612506758653e-05,
      "loss": 2.4993,
      "step": 195
    },
    {
      "epoch": 0.8412017167381974,
      "grad_norm": 0.12234227359294891,
      "learning_rate": 4.0360951632639226e-05,
      "loss": 2.3877,
      "step": 196
    },
    {
      "epoch": 0.8454935622317596,
      "grad_norm": 0.1838800013065338,
      "learning_rate": 4.016012678966213e-05,
      "loss": 2.2853,
      "step": 197
    },
    {
      "epoch": 0.8497854077253219,
      "grad_norm": 0.15433116257190704,
      "learning_rate": 3.995878633843535e-05,
      "loss": 2.4563,
      "step": 198
    },
    {
      "epoch": 0.8540772532188842,
      "grad_norm": 0.1247561052441597,
      "learning_rate": 3.9756940496626416e-05,
      "loss": 2.4479,
      "step": 199
    },
    {
      "epoch": 0.8583690987124464,
      "grad_norm": 0.1395588368177414,
      "learning_rate": 3.955459950755054e-05,
      "loss": 2.5016,
      "step": 200
    },
    {
      "epoch": 0.8626609442060086,
      "grad_norm": 0.1456160992383957,
      "learning_rate": 3.9351773639650745e-05,
      "loss": 2.3222,
      "step": 201
    },
    {
      "epoch": 0.8669527896995708,
      "grad_norm": 0.1138000562787056,
      "learning_rate": 3.914847318597682e-05,
      "loss": 2.503,
      "step": 202
    },
    {
      "epoch": 0.871244635193133,
      "grad_norm": 0.18854430317878723,
      "learning_rate": 3.894470846366291e-05,
      "loss": 2.3799,
      "step": 203
    },
    {
      "epoch": 0.8755364806866953,
      "grad_norm": 0.14914338290691376,
      "learning_rate": 3.874048981340397e-05,
      "loss": 2.4057,
      "step": 204
    },
    {
      "epoch": 0.8798283261802575,
      "grad_norm": 0.1462559849023819,
      "learning_rate": 3.853582759893097e-05,
      "loss": 2.3917,
      "step": 205
    },
    {
      "epoch": 0.8841201716738197,
      "grad_norm": 0.13625341653823853,
      "learning_rate": 3.833073220648501e-05,
      "loss": 2.3863,
      "step": 206
    },
    {
      "epoch": 0.8884120171673819,
      "grad_norm": 0.1319255530834198,
      "learning_rate": 3.812521404429016e-05,
      "loss": 2.3733,
      "step": 207
    },
    {
      "epoch": 0.8927038626609443,
      "grad_norm": 0.12159706652164459,
      "learning_rate": 3.7919283542025295e-05,
      "loss": 2.402,
      "step": 208
    },
    {
      "epoch": 0.8969957081545065,
      "grad_norm": 0.12551629543304443,
      "learning_rate": 3.7712951150294845e-05,
      "loss": 2.3653,
      "step": 209
    },
    {
      "epoch": 0.9012875536480687,
      "grad_norm": 0.12352804839611053,
      "learning_rate": 3.7506227340098376e-05,
      "loss": 2.2787,
      "step": 210
    },
    {
      "epoch": 0.9055793991416309,
      "grad_norm": 0.1033705621957779,
      "learning_rate": 3.729912260229926e-05,
      "loss": 2.2657,
      "step": 211
    },
    {
      "epoch": 0.9098712446351931,
      "grad_norm": 0.13322466611862183,
      "learning_rate": 3.7091647447092265e-05,
      "loss": 2.4372,
      "step": 212
    },
    {
      "epoch": 0.9141630901287554,
      "grad_norm": 0.1130763366818428,
      "learning_rate": 3.6883812403470165e-05,
      "loss": 2.3048,
      "step": 213
    },
    {
      "epoch": 0.9184549356223176,
      "grad_norm": 0.13779044151306152,
      "learning_rate": 3.667562801868943e-05,
      "loss": 2.4289,
      "step": 214
    },
    {
      "epoch": 0.9227467811158798,
      "grad_norm": 0.11905843019485474,
      "learning_rate": 3.646710485773499e-05,
      "loss": 2.513,
      "step": 215
    },
    {
      "epoch": 0.927038626609442,
      "grad_norm": 0.14390388131141663,
      "learning_rate": 3.625825350278403e-05,
      "loss": 2.3498,
      "step": 216
    },
    {
      "epoch": 0.9313304721030042,
      "grad_norm": 0.11969032138586044,
      "learning_rate": 3.6049084552669e-05,
      "loss": 2.3459,
      "step": 217
    },
    {
      "epoch": 0.9356223175965666,
      "grad_norm": 0.12738005816936493,
      "learning_rate": 3.5839608622339755e-05,
      "loss": 2.2709,
      "step": 218
    },
    {
      "epoch": 0.9399141630901288,
      "grad_norm": 0.12393856048583984,
      "learning_rate": 3.562983634232483e-05,
      "loss": 2.3483,
      "step": 219
    },
    {
      "epoch": 0.944206008583691,
      "grad_norm": 0.12841208279132843,
      "learning_rate": 3.541977835819197e-05,
      "loss": 2.3226,
      "step": 220
    },
    {
      "epoch": 0.9484978540772532,
      "grad_norm": 0.12215293943881989,
      "learning_rate": 3.520944533000792e-05,
      "loss": 2.4929,
      "step": 221
    },
    {
      "epoch": 0.9527896995708155,
      "grad_norm": 0.10832389444112778,
      "learning_rate": 3.4998847931797374e-05,
      "loss": 2.4507,
      "step": 222
    },
    {
      "epoch": 0.9570815450643777,
      "grad_norm": 0.10891756415367126,
      "learning_rate": 3.478799685100138e-05,
      "loss": 2.4425,
      "step": 223
    },
    {
      "epoch": 0.9613733905579399,
      "grad_norm": 0.124311164021492,
      "learning_rate": 3.457690278793489e-05,
      "loss": 2.5339,
      "step": 224
    },
    {
      "epoch": 0.9656652360515021,
      "grad_norm": 0.17225196957588196,
      "learning_rate": 3.436557645524379e-05,
      "loss": 2.4238,
      "step": 225
    },
    {
      "epoch": 0.9699570815450643,
      "grad_norm": 0.12012173235416412,
      "learning_rate": 3.415402857736122e-05,
      "loss": 2.5209,
      "step": 226
    },
    {
      "epoch": 0.9742489270386266,
      "grad_norm": 0.17129479348659515,
      "learning_rate": 3.394226988996336e-05,
      "loss": 2.5539,
      "step": 227
    },
    {
      "epoch": 0.9785407725321889,
      "grad_norm": 0.13525094091892242,
      "learning_rate": 3.373031113942456e-05,
      "loss": 2.2249,
      "step": 228
    },
    {
      "epoch": 0.9828326180257511,
      "grad_norm": 0.11054141074419022,
      "learning_rate": 3.351816308227206e-05,
      "loss": 2.4088,
      "step": 229
    },
    {
      "epoch": 0.9871244635193133,
      "grad_norm": 0.11343766748905182,
      "learning_rate": 3.330583648464004e-05,
      "loss": 2.0978,
      "step": 230
    },
    {
      "epoch": 0.9914163090128756,
      "grad_norm": 0.22153422236442566,
      "learning_rate": 3.309334212172331e-05,
      "loss": 2.5402,
      "step": 231
    },
    {
      "epoch": 0.9957081545064378,
      "grad_norm": 0.11208418756723404,
      "learning_rate": 3.288069077723045e-05,
      "loss": 2.3747,
      "step": 232
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1252536177635193,
      "learning_rate": 3.26678932428366e-05,
      "loss": 2.4297,
      "step": 233
    },
    {
      "epoch": 1.0042918454935623,
      "grad_norm": 0.23914779722690582,
      "learning_rate": 3.2454960317635747e-05,
      "loss": 2.5602,
      "step": 234
    },
    {
      "epoch": 1.0042918454935623,
      "grad_norm": 0.15806083381175995,
      "learning_rate": 3.2241902807592734e-05,
      "loss": 2.3144,
      "step": 235
    },
    {
      "epoch": 1.0085836909871244,
      "grad_norm": 0.11224311590194702,
      "learning_rate": 3.202873152499485e-05,
      "loss": 2.3083,
      "step": 236
    },
    {
      "epoch": 1.0128755364806867,
      "grad_norm": 0.12876267731189728,
      "learning_rate": 3.181545728790317e-05,
      "loss": 2.2468,
      "step": 237
    },
    {
      "epoch": 1.0171673819742488,
      "grad_norm": 0.16172175109386444,
      "learning_rate": 3.160209091960347e-05,
      "loss": 2.4206,
      "step": 238
    },
    {
      "epoch": 1.0214592274678111,
      "grad_norm": 0.10915794968605042,
      "learning_rate": 3.138864324805707e-05,
      "loss": 2.3789,
      "step": 239
    },
    {
      "epoch": 1.0257510729613735,
      "grad_norm": 0.18216541409492493,
      "learning_rate": 3.117512510535128e-05,
      "loss": 2.3726,
      "step": 240
    }
  ],
  "logging_steps": 1,
  "max_steps": 466,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 24,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.3380404450615624e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}