{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2023706273489448,
  "eval_steps": 500,
  "global_step": 700,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0002891008962127783,
      "grad_norm": 10.526118368353698,
      "learning_rate": 5.714285714285715e-07,
      "loss": 1.8059,
      "step": 1
    },
    {
      "epoch": 0.0005782017924255566,
      "grad_norm": 10.546832449734852,
      "learning_rate": 1.142857142857143e-06,
      "loss": 1.8355,
      "step": 2
    },
    {
      "epoch": 0.0008673026886383347,
      "grad_norm": 9.678077753237705,
      "learning_rate": 1.7142857142857145e-06,
      "loss": 1.8885,
      "step": 3
    },
    {
      "epoch": 0.0011564035848511131,
      "grad_norm": 7.133178553946601,
      "learning_rate": 2.285714285714286e-06,
      "loss": 1.9111,
      "step": 4
    },
    {
      "epoch": 0.0014455044810638912,
      "grad_norm": 4.135673123106036,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 1.7868,
      "step": 5
    },
    {
      "epoch": 0.0017346053772766695,
      "grad_norm": 3.0971146670050285,
      "learning_rate": 3.428571428571429e-06,
      "loss": 1.8928,
      "step": 6
    },
    {
      "epoch": 0.002023706273489448,
      "grad_norm": 2.5830650058945004,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.8522,
      "step": 7
    },
    {
      "epoch": 0.0023128071697022263,
      "grad_norm": 2.3055953581941364,
      "learning_rate": 4.571428571428572e-06,
      "loss": 1.7831,
      "step": 8
    },
    {
      "epoch": 0.0026019080659150044,
      "grad_norm": 1.8810933130960161,
      "learning_rate": 5.142857142857142e-06,
      "loss": 1.8335,
      "step": 9
    },
    {
      "epoch": 0.0028910089621277824,
      "grad_norm": 2.308546357873222,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 1.7858,
      "step": 10
    },
    {
      "epoch": 0.003180109858340561,
      "grad_norm": 1.9146380317808562,
      "learning_rate": 6.285714285714286e-06,
      "loss": 1.8005,
      "step": 11
    },
    {
      "epoch": 0.003469210754553339,
      "grad_norm": 1.7733597852169312,
      "learning_rate": 6.857142857142858e-06,
      "loss": 1.7313,
      "step": 12
    },
    {
      "epoch": 0.0037583116507661175,
      "grad_norm": 1.9163487932648622,
      "learning_rate": 7.428571428571429e-06,
      "loss": 1.8676,
      "step": 13
    },
    {
      "epoch": 0.004047412546978896,
      "grad_norm": 1.81837480862002,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.7696,
      "step": 14
    },
    {
      "epoch": 0.004336513443191674,
      "grad_norm": 1.7872590059217233,
      "learning_rate": 8.571428571428571e-06,
      "loss": 1.804,
      "step": 15
    },
    {
      "epoch": 0.004625614339404453,
      "grad_norm": 1.875901273963478,
      "learning_rate": 9.142857142857144e-06,
      "loss": 1.7379,
      "step": 16
    },
    {
      "epoch": 0.004914715235617231,
      "grad_norm": 1.6902352819283006,
      "learning_rate": 9.714285714285715e-06,
      "loss": 1.7908,
      "step": 17
    },
    {
      "epoch": 0.005203816131830009,
      "grad_norm": 1.673145216450599,
      "learning_rate": 1.0285714285714285e-05,
      "loss": 1.8525,
      "step": 18
    },
    {
      "epoch": 0.005492917028042787,
      "grad_norm": 1.6917810585273594,
      "learning_rate": 1.0857142857142858e-05,
      "loss": 1.7779,
      "step": 19
    },
    {
      "epoch": 0.005782017924255565,
      "grad_norm": 1.5858495861042294,
      "learning_rate": 1.1428571428571429e-05,
      "loss": 1.6791,
      "step": 20
    },
    {
      "epoch": 0.006071118820468344,
      "grad_norm": 2.004987370634698,
      "learning_rate": 1.2e-05,
      "loss": 1.8019,
      "step": 21
    },
    {
      "epoch": 0.006360219716681122,
      "grad_norm": 1.5348568469917172,
      "learning_rate": 1.2571428571428572e-05,
      "loss": 1.8565,
      "step": 22
    },
    {
      "epoch": 0.0066493206128939,
      "grad_norm": 1.8857981149629766,
      "learning_rate": 1.3142857142857145e-05,
      "loss": 1.7292,
      "step": 23
    },
    {
      "epoch": 0.006938421509106678,
      "grad_norm": 1.4949439821061543,
      "learning_rate": 1.3714285714285716e-05,
      "loss": 1.7357,
      "step": 24
    },
    {
      "epoch": 0.007227522405319456,
      "grad_norm": 1.693165357859836,
      "learning_rate": 1.4285714285714287e-05,
      "loss": 1.8103,
      "step": 25
    },
    {
      "epoch": 0.007516623301532235,
      "grad_norm": 1.6138034582127923,
      "learning_rate": 1.4857142857142858e-05,
      "loss": 1.7634,
      "step": 26
    },
    {
      "epoch": 0.007805724197745013,
      "grad_norm": 1.6357966310241434,
      "learning_rate": 1.542857142857143e-05,
      "loss": 1.7699,
      "step": 27
    },
    {
      "epoch": 0.008094825093957791,
      "grad_norm": 1.5038647382371504,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 1.7824,
      "step": 28
    },
    {
      "epoch": 0.00838392599017057,
      "grad_norm": 1.4835637524801124,
      "learning_rate": 1.6571428571428574e-05,
      "loss": 1.685,
      "step": 29
    },
    {
      "epoch": 0.008673026886383347,
      "grad_norm": 1.5071883679399785,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 1.8,
      "step": 30
    },
    {
      "epoch": 0.008962127782596125,
      "grad_norm": 1.7942262887809508,
      "learning_rate": 1.7714285714285717e-05,
      "loss": 1.8129,
      "step": 31
    },
    {
      "epoch": 0.009251228678808905,
      "grad_norm": 1.6582976562001455,
      "learning_rate": 1.8285714285714288e-05,
      "loss": 1.8308,
      "step": 32
    },
    {
      "epoch": 0.009540329575021683,
      "grad_norm": 1.5337861742601142,
      "learning_rate": 1.885714285714286e-05,
      "loss": 1.7456,
      "step": 33
    },
    {
      "epoch": 0.009829430471234461,
      "grad_norm": 1.8002402808530826,
      "learning_rate": 1.942857142857143e-05,
      "loss": 1.9268,
      "step": 34
    },
    {
      "epoch": 0.01011853136744724,
      "grad_norm": 1.7514076269246104,
      "learning_rate": 2e-05,
      "loss": 1.652,
      "step": 35
    },
    {
      "epoch": 0.010407632263660017,
      "grad_norm": 1.8491103433261384,
      "learning_rate": 1.9999995790774362e-05,
      "loss": 1.8142,
      "step": 36
    },
    {
      "epoch": 0.010696733159872795,
      "grad_norm": 1.8365212131962536,
      "learning_rate": 1.9999983163100983e-05,
      "loss": 1.7107,
      "step": 37
    },
    {
      "epoch": 0.010985834056085574,
      "grad_norm": 1.5873275727358707,
      "learning_rate": 1.9999962116990498e-05,
      "loss": 1.6696,
      "step": 38
    },
    {
      "epoch": 0.011274934952298352,
      "grad_norm": 1.7256221512542549,
      "learning_rate": 1.999993265246062e-05,
      "loss": 1.6901,
      "step": 39
    },
    {
      "epoch": 0.01156403584851113,
      "grad_norm": 1.8002146086893285,
      "learning_rate": 1.9999894769536163e-05,
      "loss": 1.8886,
      "step": 40
    },
    {
      "epoch": 0.01185313674472391,
      "grad_norm": 1.8367868066313766,
      "learning_rate": 1.999984846824901e-05,
      "loss": 1.7876,
      "step": 41
    },
    {
      "epoch": 0.012142237640936688,
      "grad_norm": 1.5841989203084583,
      "learning_rate": 1.999979374863814e-05,
      "loss": 1.9172,
      "step": 42
    },
    {
      "epoch": 0.012431338537149466,
      "grad_norm": 1.5522062286750515,
      "learning_rate": 1.9999730610749623e-05,
      "loss": 1.8207,
      "step": 43
    },
    {
      "epoch": 0.012720439433362244,
      "grad_norm": 1.6373562793041219,
      "learning_rate": 1.9999659054636612e-05,
      "loss": 1.6358,
      "step": 44
    },
    {
      "epoch": 0.013009540329575022,
      "grad_norm": 1.4771807978853968,
      "learning_rate": 1.9999579080359337e-05,
      "loss": 1.6684,
      "step": 45
    },
    {
      "epoch": 0.0132986412257878,
      "grad_norm": 1.6255295191073982,
      "learning_rate": 1.9999490687985134e-05,
      "loss": 1.7113,
      "step": 46
    },
    {
      "epoch": 0.013587742122000578,
      "grad_norm": 1.4924171597100626,
      "learning_rate": 1.999939387758841e-05,
      "loss": 1.7272,
      "step": 47
    },
    {
      "epoch": 0.013876843018213356,
      "grad_norm": 1.536593102227256,
      "learning_rate": 1.9999288649250667e-05,
      "loss": 1.7532,
      "step": 48
    },
    {
      "epoch": 0.014165943914426134,
      "grad_norm": 1.510648288143349,
      "learning_rate": 1.9999175003060487e-05,
      "loss": 1.8358,
      "step": 49
    },
    {
      "epoch": 0.014455044810638912,
      "grad_norm": 1.4755302580340564,
      "learning_rate": 1.9999052939113552e-05,
      "loss": 1.8305,
      "step": 50
    },
    {
      "epoch": 0.014744145706851692,
      "grad_norm": 1.659035076331237,
      "learning_rate": 1.9998922457512608e-05,
      "loss": 1.7916,
      "step": 51
    },
    {
      "epoch": 0.01503324660306447,
      "grad_norm": 1.4194234922554345,
      "learning_rate": 1.9998783558367506e-05,
      "loss": 1.6693,
      "step": 52
    },
    {
      "epoch": 0.015322347499277248,
      "grad_norm": 1.5709438854610118,
      "learning_rate": 1.9998636241795184e-05,
      "loss": 1.7873,
      "step": 53
    },
    {
      "epoch": 0.015611448395490026,
      "grad_norm": 1.520579238616479,
      "learning_rate": 1.999848050791965e-05,
      "loss": 1.6949,
      "step": 54
    },
    {
      "epoch": 0.015900549291702804,
      "grad_norm": 1.6871179175362505,
      "learning_rate": 1.9998316356872015e-05,
      "loss": 1.7697,
      "step": 55
    },
    {
      "epoch": 0.016189650187915582,
      "grad_norm": 1.3949302668328034,
      "learning_rate": 1.9998143788790462e-05,
      "loss": 1.6176,
      "step": 56
    },
    {
      "epoch": 0.01647875108412836,
      "grad_norm": 1.76393196989828,
      "learning_rate": 1.9997962803820274e-05,
      "loss": 1.7379,
      "step": 57
    },
    {
      "epoch": 0.01676785198034114,
      "grad_norm": 2.3952825919806724,
      "learning_rate": 1.9997773402113804e-05,
      "loss": 1.8324,
      "step": 58
    },
    {
      "epoch": 0.017056952876553916,
      "grad_norm": 1.5638827679672842,
      "learning_rate": 1.9997575583830508e-05,
      "loss": 1.783,
      "step": 59
    },
    {
      "epoch": 0.017346053772766695,
      "grad_norm": 1.5321008746402438,
      "learning_rate": 1.999736934913691e-05,
      "loss": 1.7136,
      "step": 60
    },
    {
      "epoch": 0.017635154668979473,
      "grad_norm": 1.46754278560031,
      "learning_rate": 1.9997154698206636e-05,
      "loss": 1.7567,
      "step": 61
    },
    {
      "epoch": 0.01792425556519225,
      "grad_norm": 1.5842416386679679,
      "learning_rate": 1.999693163122038e-05,
      "loss": 1.757,
      "step": 62
    },
    {
      "epoch": 0.01821335646140503,
      "grad_norm": 1.5267999996929678,
      "learning_rate": 1.9996700148365936e-05,
      "loss": 1.768,
      "step": 63
    },
    {
      "epoch": 0.01850245735761781,
      "grad_norm": 1.6371904649667028,
      "learning_rate": 1.9996460249838176e-05,
      "loss": 1.8574,
      "step": 64
    },
    {
      "epoch": 0.01879155825383059,
      "grad_norm": 1.4572549097707117,
      "learning_rate": 1.9996211935839053e-05,
      "loss": 1.8134,
      "step": 65
    },
    {
      "epoch": 0.019080659150043366,
      "grad_norm": 1.6026953718941006,
      "learning_rate": 1.9995955206577617e-05,
      "loss": 1.7718,
      "step": 66
    },
    {
      "epoch": 0.019369760046256145,
      "grad_norm": 1.4731386293380266,
      "learning_rate": 1.9995690062269985e-05,
      "loss": 1.7598,
      "step": 67
    },
    {
      "epoch": 0.019658860942468923,
      "grad_norm": 1.4280673736098164,
      "learning_rate": 1.9995416503139372e-05,
      "loss": 1.7469,
      "step": 68
    },
    {
      "epoch": 0.0199479618386817,
      "grad_norm": 1.4976095361519832,
      "learning_rate": 1.9995134529416077e-05,
      "loss": 1.7258,
      "step": 69
    },
    {
      "epoch": 0.02023706273489448,
      "grad_norm": 1.3734046074406898,
      "learning_rate": 1.9994844141337467e-05,
      "loss": 1.6921,
      "step": 70
    },
    {
      "epoch": 0.020526163631107257,
      "grad_norm": 1.5285691045400083,
      "learning_rate": 1.9994545339148017e-05,
      "loss": 1.755,
      "step": 71
    },
    {
      "epoch": 0.020815264527320035,
      "grad_norm": 1.7237565941986015,
      "learning_rate": 1.9994238123099264e-05,
      "loss": 1.7254,
      "step": 72
    },
    {
      "epoch": 0.021104365423532813,
      "grad_norm": 1.4511698572347005,
      "learning_rate": 1.9993922493449833e-05,
      "loss": 1.7339,
      "step": 73
    },
    {
      "epoch": 0.02139346631974559,
      "grad_norm": 1.3034115810336904,
      "learning_rate": 1.9993598450465446e-05,
      "loss": 1.8015,
      "step": 74
    },
    {
      "epoch": 0.02168256721595837,
      "grad_norm": 1.3220719840165054,
      "learning_rate": 1.9993265994418887e-05,
      "loss": 1.7794,
      "step": 75
    },
    {
      "epoch": 0.021971668112171147,
      "grad_norm": 1.3514726546077938,
      "learning_rate": 1.999292512559004e-05,
      "loss": 1.781,
      "step": 76
    },
    {
      "epoch": 0.022260769008383925,
      "grad_norm": 1.4854967526025784,
      "learning_rate": 1.9992575844265857e-05,
      "loss": 1.8188,
      "step": 77
    },
    {
      "epoch": 0.022549869904596703,
      "grad_norm": 1.42527411671424,
      "learning_rate": 1.9992218150740386e-05,
      "loss": 1.7858,
      "step": 78
    },
    {
      "epoch": 0.02283897080080948,
      "grad_norm": 1.3690019880914763,
      "learning_rate": 1.9991852045314744e-05,
      "loss": 1.69,
      "step": 79
    },
    {
      "epoch": 0.02312807169702226,
      "grad_norm": 1.4796987548843166,
      "learning_rate": 1.9991477528297136e-05,
      "loss": 1.7501,
      "step": 80
    },
    {
      "epoch": 0.023417172593235037,
      "grad_norm": 1.4237249543510602,
      "learning_rate": 1.999109460000285e-05,
      "loss": 1.7634,
      "step": 81
    },
    {
      "epoch": 0.02370627348944782,
      "grad_norm": 1.3087042024971207,
      "learning_rate": 1.9990703260754248e-05,
      "loss": 1.8106,
      "step": 82
    },
    {
      "epoch": 0.023995374385660597,
      "grad_norm": 1.4158584649264259,
      "learning_rate": 1.999030351088078e-05,
      "loss": 1.7503,
      "step": 83
    },
    {
      "epoch": 0.024284475281873375,
      "grad_norm": 1.4266437173978106,
      "learning_rate": 1.9989895350718972e-05,
      "loss": 1.6435,
      "step": 84
    },
    {
      "epoch": 0.024573576178086153,
      "grad_norm": 1.519614177980851,
      "learning_rate": 1.9989478780612434e-05,
      "loss": 1.7155,
      "step": 85
    },
    {
      "epoch": 0.02486267707429893,
      "grad_norm": 1.4829974498829237,
      "learning_rate": 1.9989053800911854e-05,
      "loss": 1.7406,
      "step": 86
    },
    {
      "epoch": 0.02515177797051171,
      "grad_norm": 1.5340468112410737,
      "learning_rate": 1.9988620411974994e-05,
      "loss": 1.6487,
      "step": 87
    },
    {
      "epoch": 0.025440878866724487,
      "grad_norm": 1.3928928308705317,
      "learning_rate": 1.9988178614166706e-05,
      "loss": 1.7535,
      "step": 88
    },
    {
      "epoch": 0.025729979762937265,
      "grad_norm": 1.4648291742355268,
      "learning_rate": 1.9987728407858912e-05,
      "loss": 1.6977,
      "step": 89
    },
    {
      "epoch": 0.026019080659150044,
      "grad_norm": 1.5330205874026324,
      "learning_rate": 1.9987269793430618e-05,
      "loss": 1.8134,
      "step": 90
    },
    {
      "epoch": 0.02630818155536282,
      "grad_norm": 1.4322119486638492,
      "learning_rate": 1.9986802771267902e-05,
      "loss": 1.7493,
      "step": 91
    },
    {
      "epoch": 0.0265972824515756,
      "grad_norm": 1.3741952459535092,
      "learning_rate": 1.9986327341763933e-05,
      "loss": 1.7498,
      "step": 92
    },
    {
      "epoch": 0.026886383347788378,
      "grad_norm": 1.4439173869710846,
      "learning_rate": 1.998584350531894e-05,
      "loss": 1.8364,
      "step": 93
    },
    {
      "epoch": 0.027175484244001156,
      "grad_norm": 1.5238329839004898,
      "learning_rate": 1.9985351262340242e-05,
      "loss": 1.7025,
      "step": 94
    },
    {
      "epoch": 0.027464585140213934,
      "grad_norm": 1.4211403871339836,
      "learning_rate": 1.998485061324223e-05,
      "loss": 1.8061,
      "step": 95
    },
    {
      "epoch": 0.027753686036426712,
      "grad_norm": 1.4891220326396055,
      "learning_rate": 1.998434155844638e-05,
      "loss": 1.7756,
      "step": 96
    },
    {
      "epoch": 0.02804278693263949,
      "grad_norm": 1.4672105325874756,
      "learning_rate": 1.9983824098381226e-05,
      "loss": 1.7642,
      "step": 97
    },
    {
      "epoch": 0.028331887828852268,
      "grad_norm": 1.4102584299249543,
      "learning_rate": 1.9983298233482396e-05,
      "loss": 1.7302,
      "step": 98
    },
    {
      "epoch": 0.028620988725065046,
      "grad_norm": 1.4570496244711266,
      "learning_rate": 1.9982763964192586e-05,
      "loss": 1.7299,
      "step": 99
    },
    {
      "epoch": 0.028910089621277824,
      "grad_norm": 1.4536709099623053,
      "learning_rate": 1.9982221290961568e-05,
      "loss": 1.7647,
      "step": 100
    },
    {
      "epoch": 0.029199190517490606,
      "grad_norm": 1.427202594686143,
      "learning_rate": 1.9981670214246186e-05,
      "loss": 1.7361,
      "step": 101
    },
    {
      "epoch": 0.029488291413703384,
      "grad_norm": 1.4166737515686485,
      "learning_rate": 1.9981110734510368e-05,
      "loss": 1.7107,
      "step": 102
    },
    {
      "epoch": 0.029777392309916162,
      "grad_norm": 1.4291514974937183,
      "learning_rate": 1.99805428522251e-05,
      "loss": 1.7162,
      "step": 103
    },
    {
      "epoch": 0.03006649320612894,
      "grad_norm": 1.3691457808724619,
      "learning_rate": 1.997996656786846e-05,
      "loss": 1.8823,
      "step": 104
    },
    {
      "epoch": 0.030355594102341718,
      "grad_norm": 1.38193263647762,
      "learning_rate": 1.9979381881925587e-05,
      "loss": 1.6658,
      "step": 105
    },
    {
      "epoch": 0.030644694998554496,
      "grad_norm": 1.4331634763444776,
      "learning_rate": 1.997878879488869e-05,
      "loss": 1.7695,
      "step": 106
    },
    {
      "epoch": 0.030933795894767274,
      "grad_norm": 1.4389434938173666,
      "learning_rate": 1.9978187307257064e-05,
      "loss": 1.7356,
      "step": 107
    },
    {
      "epoch": 0.031222896790980052,
      "grad_norm": 1.4057807430118254,
      "learning_rate": 1.9977577419537066e-05,
      "loss": 1.7191,
      "step": 108
    },
    {
      "epoch": 0.03151199768719283,
      "grad_norm": 1.4476225746298625,
      "learning_rate": 1.9976959132242128e-05,
      "loss": 1.8028,
      "step": 109
    },
    {
      "epoch": 0.03180109858340561,
      "grad_norm": 1.3595624610249801,
      "learning_rate": 1.997633244589275e-05,
      "loss": 1.708,
      "step": 110
    },
    {
      "epoch": 0.032090199479618386,
      "grad_norm": 1.5622440441344578,
      "learning_rate": 1.997569736101651e-05,
      "loss": 1.7138,
      "step": 111
    },
    {
      "epoch": 0.032379300375831165,
      "grad_norm": 1.4451837327532462,
      "learning_rate": 1.9975053878148045e-05,
      "loss": 1.745,
      "step": 112
    },
    {
      "epoch": 0.03266840127204394,
      "grad_norm": 1.3461945134165731,
      "learning_rate": 1.9974401997829063e-05,
      "loss": 1.792,
      "step": 113
    },
    {
      "epoch": 0.03295750216825672,
      "grad_norm": 1.4916956994354813,
      "learning_rate": 1.997374172060836e-05,
      "loss": 1.7312,
      "step": 114
    },
    {
      "epoch": 0.0332466030644695,
      "grad_norm": 1.3473635350533286,
      "learning_rate": 1.997307304704178e-05,
      "loss": 1.7257,
      "step": 115
    },
    {
      "epoch": 0.03353570396068228,
      "grad_norm": 1.5308835567804706,
      "learning_rate": 1.9972395977692243e-05,
      "loss": 1.8833,
      "step": 116
    },
    {
      "epoch": 0.033824804856895055,
      "grad_norm": 1.4933799858376915,
      "learning_rate": 1.9971710513129735e-05,
      "loss": 1.8031,
      "step": 117
    },
    {
      "epoch": 0.03411390575310783,
      "grad_norm": 1.4666676384345598,
      "learning_rate": 1.9971016653931315e-05,
      "loss": 1.699,
      "step": 118
    },
    {
      "epoch": 0.03440300664932061,
      "grad_norm": 1.376396676563015,
      "learning_rate": 1.9970314400681098e-05,
      "loss": 1.7561,
      "step": 119
    },
    {
      "epoch": 0.03469210754553339,
      "grad_norm": 1.4309738039308064,
      "learning_rate": 1.9969603753970285e-05,
      "loss": 1.7812,
      "step": 120
    },
    {
      "epoch": 0.03498120844174617,
      "grad_norm": 1.4395837496367765,
      "learning_rate": 1.9968884714397116e-05,
      "loss": 1.6863,
      "step": 121
    },
    {
      "epoch": 0.035270309337958945,
      "grad_norm": 1.3440476522906626,
      "learning_rate": 1.9968157282566917e-05,
      "loss": 1.7026,
      "step": 122
    },
    {
      "epoch": 0.03555941023417172,
      "grad_norm": 1.3253598540978417,
      "learning_rate": 1.9967421459092076e-05,
      "loss": 1.6773,
      "step": 123
    },
    {
      "epoch": 0.0358485111303845,
      "grad_norm": 1.4654991309929606,
      "learning_rate": 1.996667724459204e-05,
      "loss": 1.6995,
      "step": 124
    },
    {
      "epoch": 0.03613761202659728,
      "grad_norm": 1.3695366677193352,
      "learning_rate": 1.9965924639693323e-05,
      "loss": 1.8028,
      "step": 125
    },
    {
      "epoch": 0.03642671292281006,
      "grad_norm": 1.5996594030035125,
      "learning_rate": 1.9965163645029496e-05,
      "loss": 1.6509,
      "step": 126
    },
    {
      "epoch": 0.036715813819022836,
      "grad_norm": 1.3638136094831714,
      "learning_rate": 1.9964394261241205e-05,
      "loss": 1.7974,
      "step": 127
    },
    {
      "epoch": 0.03700491471523562,
      "grad_norm": 1.5950089807963124,
      "learning_rate": 1.9963616488976156e-05,
      "loss": 1.7817,
      "step": 128
    },
    {
      "epoch": 0.0372940156114484,
      "grad_norm": 1.3874690940551964,
      "learning_rate": 1.9962830328889104e-05,
      "loss": 1.7727,
      "step": 129
    },
    {
      "epoch": 0.03758311650766118,
      "grad_norm": 1.428332455733304,
      "learning_rate": 1.9962035781641877e-05,
      "loss": 1.8162,
      "step": 130
    },
    {
      "epoch": 0.037872217403873955,
      "grad_norm": 1.413377966510162,
      "learning_rate": 1.996123284790336e-05,
      "loss": 1.6724,
      "step": 131
    },
    {
      "epoch": 0.03816131830008673,
      "grad_norm": 1.4114781128468488,
      "learning_rate": 1.9960421528349503e-05,
      "loss": 1.6688,
      "step": 132
    },
    {
      "epoch": 0.03845041919629951,
      "grad_norm": 1.4667937866255885,
      "learning_rate": 1.9959601823663305e-05,
      "loss": 1.7404,
      "step": 133
    },
    {
      "epoch": 0.03873952009251229,
      "grad_norm": 1.3864312896377629,
      "learning_rate": 1.9958773734534834e-05,
      "loss": 1.6797,
      "step": 134
    },
    {
      "epoch": 0.03902862098872507,
      "grad_norm": 1.5863275696219379,
      "learning_rate": 1.9957937261661215e-05,
      "loss": 1.7093,
      "step": 135
    },
    {
      "epoch": 0.039317721884937845,
      "grad_norm": 1.3943132521368478,
      "learning_rate": 1.9957092405746628e-05,
      "loss": 1.7466,
      "step": 136
    },
    {
      "epoch": 0.03960682278115062,
      "grad_norm": 1.4418545458171168,
      "learning_rate": 1.9956239167502304e-05,
      "loss": 1.8363,
      "step": 137
    },
    {
      "epoch": 0.0398959236773634,
      "grad_norm": 1.6597022379228474,
      "learning_rate": 1.9955377547646546e-05,
      "loss": 1.7726,
      "step": 138
    },
    {
      "epoch": 0.04018502457357618,
      "grad_norm": 1.374590779273713,
      "learning_rate": 1.9954507546904697e-05,
      "loss": 1.7559,
      "step": 139
    },
    {
      "epoch": 0.04047412546978896,
      "grad_norm": 1.2866499565497842,
      "learning_rate": 1.995362916600917e-05,
      "loss": 1.7478,
      "step": 140
    },
    {
      "epoch": 0.040763226366001735,
      "grad_norm": 1.3549533524362687,
      "learning_rate": 1.9952742405699425e-05,
      "loss": 1.8389,
      "step": 141
    },
    {
      "epoch": 0.041052327262214514,
      "grad_norm": 1.3512745611701633,
      "learning_rate": 1.995184726672197e-05,
      "loss": 1.7975,
      "step": 142
    },
    {
      "epoch": 0.04134142815842729,
      "grad_norm": 1.3460191452855619,
      "learning_rate": 1.995094374983038e-05,
      "loss": 1.7245,
      "step": 143
    },
    {
      "epoch": 0.04163052905464007,
      "grad_norm": 1.3305167255129993,
      "learning_rate": 1.9950031855785276e-05,
      "loss": 1.6294,
      "step": 144
    },
    {
      "epoch": 0.04191962995085285,
      "grad_norm": 1.34284871707125,
      "learning_rate": 1.9949111585354328e-05,
      "loss": 1.7124,
      "step": 145
    },
    {
      "epoch": 0.042208730847065626,
      "grad_norm": 1.3772925926259127,
      "learning_rate": 1.9948182939312258e-05,
      "loss": 1.7095,
      "step": 146
    },
    {
      "epoch": 0.042497831743278404,
      "grad_norm": 1.384298315882825,
      "learning_rate": 1.994724591844085e-05,
      "loss": 1.8031,
      "step": 147
    },
    {
      "epoch": 0.04278693263949118,
      "grad_norm": 1.3740208297445895,
      "learning_rate": 1.994630052352893e-05,
      "loss": 1.7814,
      "step": 148
    },
    {
      "epoch": 0.04307603353570396,
      "grad_norm": 1.3187990568089156,
      "learning_rate": 1.9945346755372367e-05,
      "loss": 1.7486,
      "step": 149
    },
    {
      "epoch": 0.04336513443191674,
      "grad_norm": 1.3347832866557865,
      "learning_rate": 1.9944384614774095e-05,
      "loss": 1.7225,
      "step": 150
    },
    {
      "epoch": 0.043654235328129516,
      "grad_norm": 1.3850913268385951,
      "learning_rate": 1.9943414102544083e-05,
      "loss": 1.7608,
      "step": 151
    },
    {
      "epoch": 0.043943336224342294,
      "grad_norm": 1.3180540802523992,
      "learning_rate": 1.994243521949935e-05,
      "loss": 1.7534,
      "step": 152
    },
    {
      "epoch": 0.04423243712055507,
      "grad_norm": 1.3989923351107776,
      "learning_rate": 1.9941447966463966e-05,
      "loss": 1.7734,
      "step": 153
    },
    {
      "epoch": 0.04452153801676785,
      "grad_norm": 1.2709853505702844,
      "learning_rate": 1.9940452344269045e-05,
      "loss": 1.7463,
      "step": 154
    },
    {
      "epoch": 0.04481063891298063,
      "grad_norm": 1.3117490865184025,
      "learning_rate": 1.9939448353752745e-05,
      "loss": 1.7484,
      "step": 155
    },
    {
      "epoch": 0.045099739809193407,
      "grad_norm": 1.3068520843456393,
      "learning_rate": 1.9938435995760275e-05,
      "loss": 1.7793,
      "step": 156
    },
    {
      "epoch": 0.045388840705406185,
      "grad_norm": 1.2903723753260385,
      "learning_rate": 1.993741527114388e-05,
      "loss": 1.7654,
      "step": 157
    },
    {
      "epoch": 0.04567794160161896,
      "grad_norm": 1.3281351334689,
      "learning_rate": 1.993638618076285e-05,
      "loss": 1.7574,
      "step": 158
    },
    {
      "epoch": 0.04596704249783174,
      "grad_norm": 1.331538167128145,
      "learning_rate": 1.9935348725483526e-05,
      "loss": 1.7431,
      "step": 159
    },
    {
      "epoch": 0.04625614339404452,
      "grad_norm": 1.4516409761685403,
      "learning_rate": 1.9934302906179285e-05,
      "loss": 1.7334,
      "step": 160
    },
    {
      "epoch": 0.0465452442902573,
      "grad_norm": 1.2977322772074695,
      "learning_rate": 1.9933248723730536e-05,
      "loss": 1.6306,
      "step": 161
    },
    {
      "epoch": 0.046834345186470075,
      "grad_norm": 1.357062910781076,
      "learning_rate": 1.9932186179024744e-05,
      "loss": 1.668,
      "step": 162
    },
    {
      "epoch": 0.04712344608268285,
      "grad_norm": 1.3803460903193827,
      "learning_rate": 1.9931115272956405e-05,
      "loss": 1.7275,
      "step": 163
    },
    {
      "epoch": 0.04741254697889564,
      "grad_norm": 1.315843862445011,
      "learning_rate": 1.9930036006427058e-05,
      "loss": 1.6444,
      "step": 164
    },
    {
      "epoch": 0.047701647875108416,
      "grad_norm": 1.4617283691426304,
      "learning_rate": 1.9928948380345274e-05,
      "loss": 1.8109,
      "step": 165
    },
    {
      "epoch": 0.047990748771321194,
      "grad_norm": 1.38383885394047,
      "learning_rate": 1.9927852395626672e-05,
      "loss": 1.7686,
      "step": 166
    },
    {
      "epoch": 0.04827984966753397,
      "grad_norm": 1.458724964408903,
      "learning_rate": 1.9926748053193898e-05,
      "loss": 1.6962,
      "step": 167
    },
    {
      "epoch": 0.04856895056374675,
      "grad_norm": 1.3411049270177544,
      "learning_rate": 1.9925635353976634e-05,
      "loss": 1.6548,
      "step": 168
    },
    {
      "epoch": 0.04885805145995953,
      "grad_norm": 1.4316515117975288,
      "learning_rate": 1.992451429891161e-05,
      "loss": 1.7439,
      "step": 169
    },
    {
      "epoch": 0.049147152356172306,
      "grad_norm": 1.3310049050553332,
      "learning_rate": 1.9923384888942568e-05,
      "loss": 1.7042,
      "step": 170
    },
    {
      "epoch": 0.049436253252385085,
      "grad_norm": 1.411118381246596,
      "learning_rate": 1.9922247125020307e-05,
      "loss": 1.7172,
      "step": 171
    },
    {
      "epoch": 0.04972535414859786,
      "grad_norm": 1.2882521449870807,
      "learning_rate": 1.992110100810264e-05,
      "loss": 1.7679,
      "step": 172
    },
    {
      "epoch": 0.05001445504481064,
      "grad_norm": 1.3412263735063734,
      "learning_rate": 1.991994653915443e-05,
      "loss": 1.6375,
      "step": 173
    },
    {
      "epoch": 0.05030355594102342,
      "grad_norm": 1.2875141274832036,
      "learning_rate": 1.991878371914755e-05,
      "loss": 1.6816,
      "step": 174
    },
    {
      "epoch": 0.0505926568372362,
      "grad_norm": 1.2707786543811448,
      "learning_rate": 1.991761254906092e-05,
      "loss": 1.7382,
      "step": 175
    },
    {
      "epoch": 0.050881757733448975,
      "grad_norm": 1.2815480979153002,
      "learning_rate": 1.9916433029880485e-05,
      "loss": 1.7519,
      "step": 176
    },
    {
      "epoch": 0.05117085862966175,
      "grad_norm": 1.358987922639111,
      "learning_rate": 1.991524516259921e-05,
      "loss": 1.7895,
      "step": 177
    },
    {
      "epoch": 0.05145995952587453,
      "grad_norm": 1.2688791830287245,
      "learning_rate": 1.9914048948217105e-05,
      "loss": 1.776,
      "step": 178
    },
    {
      "epoch": 0.05174906042208731,
      "grad_norm": 1.2813810920416182,
      "learning_rate": 1.9912844387741194e-05,
      "loss": 1.7085,
      "step": 179
    },
    {
      "epoch": 0.05203816131830009,
      "grad_norm": 1.2916088597948596,
      "learning_rate": 1.991163148218553e-05,
      "loss": 1.7124,
      "step": 180
    },
    {
      "epoch": 0.052327262214512865,
      "grad_norm": 1.310527847083729,
      "learning_rate": 1.9910410232571187e-05,
      "loss": 1.6333,
      "step": 181
    },
    {
      "epoch": 0.05261636311072564,
      "grad_norm": 1.3292526744086122,
      "learning_rate": 1.9909180639926274e-05,
      "loss": 1.663,
      "step": 182
    },
    {
      "epoch": 0.05290546400693842,
      "grad_norm": 1.4107465041661023,
      "learning_rate": 1.990794270528592e-05,
      "loss": 1.6939,
      "step": 183
    },
    {
      "epoch": 0.0531945649031512,
      "grad_norm": 1.5697948856294517,
      "learning_rate": 1.9906696429692265e-05,
      "loss": 1.79,
      "step": 184
    },
    {
      "epoch": 0.05348366579936398,
      "grad_norm": 1.359687672105227,
      "learning_rate": 1.9905441814194482e-05,
      "loss": 1.6024,
      "step": 185
    },
    {
      "epoch": 0.053772766695576756,
      "grad_norm": 1.435220148856358,
      "learning_rate": 1.990417885984877e-05,
      "loss": 1.6539,
      "step": 186
    },
    {
      "epoch": 0.054061867591789534,
      "grad_norm": 1.4702625299572514,
      "learning_rate": 1.990290756771834e-05,
      "loss": 1.7306,
      "step": 187
    },
    {
      "epoch": 0.05435096848800231,
      "grad_norm": 1.5247866204720824,
      "learning_rate": 1.9901627938873416e-05,
      "loss": 1.7609,
      "step": 188
    },
    {
      "epoch": 0.05464006938421509,
      "grad_norm": 1.3031717794857007,
      "learning_rate": 1.9900339974391252e-05,
      "loss": 1.7389,
      "step": 189
    },
    {
      "epoch": 0.05492917028042787,
      "grad_norm": 1.3511696476383066,
      "learning_rate": 1.9899043675356114e-05,
      "loss": 1.7297,
      "step": 190
    },
    {
      "epoch": 0.055218271176640646,
      "grad_norm": 1.4509301331653846,
      "learning_rate": 1.9897739042859285e-05,
      "loss": 1.737,
      "step": 191
    },
    {
      "epoch": 0.055507372072853424,
      "grad_norm": 1.3380070550795329,
      "learning_rate": 1.9896426077999062e-05,
      "loss": 1.6662,
      "step": 192
    },
    {
      "epoch": 0.0557964729690662,
      "grad_norm": 1.3684181423939343,
      "learning_rate": 1.989510478188076e-05,
      "loss": 1.7297,
      "step": 193
    },
    {
      "epoch": 0.05608557386527898,
      "grad_norm": 1.4948293977771248,
      "learning_rate": 1.9893775155616704e-05,
      "loss": 1.7733,
      "step": 194
    },
    {
      "epoch": 0.05637467476149176,
      "grad_norm": 6.056711246922273,
      "learning_rate": 1.989243720032624e-05,
      "loss": 1.8127,
      "step": 195
    },
    {
      "epoch": 0.056663775657704536,
      "grad_norm": 1.4031162820625223,
      "learning_rate": 1.989109091713571e-05,
      "loss": 1.7476,
      "step": 196
    },
    {
      "epoch": 0.056952876553917314,
      "grad_norm": 1.388743000455743,
      "learning_rate": 1.988973630717848e-05,
      "loss": 1.7256,
      "step": 197
    },
    {
      "epoch": 0.05724197745013009,
      "grad_norm": 1.3228310217145096,
      "learning_rate": 1.9888373371594923e-05,
      "loss": 1.7153,
      "step": 198
    },
    {
      "epoch": 0.05753107834634287,
      "grad_norm": 1.349390586328493,
      "learning_rate": 1.9887002111532413e-05,
      "loss": 1.7913,
      "step": 199
    },
    {
      "epoch": 0.05782017924255565,
      "grad_norm": 1.3218835310967434,
      "learning_rate": 1.9885622528145346e-05,
      "loss": 1.893,
      "step": 200
    },
    {
      "epoch": 0.058109280138768434,
      "grad_norm": 1.3066393320266039,
      "learning_rate": 1.9884234622595117e-05,
      "loss": 1.7445,
      "step": 201
    },
    {
      "epoch": 0.05839838103498121,
      "grad_norm": 1.3431846081510803,
      "learning_rate": 1.988283839605013e-05,
      "loss": 1.764,
      "step": 202
    },
    {
      "epoch": 0.05868748193119399,
      "grad_norm": 1.2716836951252142,
      "learning_rate": 1.988143384968578e-05,
      "loss": 1.689,
      "step": 203
    },
    {
      "epoch": 0.05897658282740677,
      "grad_norm": 1.3463201818447268,
      "learning_rate": 1.9880020984684486e-05,
      "loss": 1.6615,
      "step": 204
    },
    {
      "epoch": 0.059265683723619546,
      "grad_norm": 1.3067391536442128,
      "learning_rate": 1.9878599802235662e-05,
      "loss": 1.683,
      "step": 205
    },
    {
      "epoch": 0.059554784619832324,
      "grad_norm": 1.407390031033922,
      "learning_rate": 1.987717030353572e-05,
      "loss": 1.7617,
      "step": 206
    },
    {
      "epoch": 0.0598438855160451,
      "grad_norm": 1.2785798130045682,
      "learning_rate": 1.9875732489788082e-05,
      "loss": 1.6828,
      "step": 207
    },
    {
      "epoch": 0.06013298641225788,
      "grad_norm": 1.581333677513783,
      "learning_rate": 1.9874286362203162e-05,
      "loss": 1.776,
      "step": 208
    },
    {
      "epoch": 0.06042208730847066,
      "grad_norm": 1.3170624490173282,
      "learning_rate": 1.987283192199837e-05,
      "loss": 1.7548,
      "step": 209
    },
    {
      "epoch": 0.060711188204683436,
      "grad_norm": 1.6773819921976063,
      "learning_rate": 1.9871369170398126e-05,
      "loss": 1.7935,
      "step": 210
    },
    {
      "epoch": 0.061000289100896214,
      "grad_norm": 1.4063651245701678,
      "learning_rate": 1.9869898108633834e-05,
      "loss": 1.7806,
      "step": 211
    },
    {
      "epoch": 0.06128938999710899,
      "grad_norm": 1.3241937165367996,
      "learning_rate": 1.986841873794391e-05,
      "loss": 1.7262,
      "step": 212
    },
    {
      "epoch": 0.06157849089332177,
      "grad_norm": 1.49389208627189,
      "learning_rate": 1.9866931059573748e-05,
      "loss": 1.7103,
      "step": 213
    },
    {
      "epoch": 0.06186759178953455,
      "grad_norm": 1.219442276366396,
      "learning_rate": 1.9865435074775744e-05,
      "loss": 1.6194,
      "step": 214
    },
    {
      "epoch": 0.062156692685747326,
      "grad_norm": 1.3453821292169061,
      "learning_rate": 1.9863930784809284e-05,
      "loss": 1.8086,
      "step": 215
    },
    {
      "epoch": 0.062445793581960105,
      "grad_norm": 1.2961743317843437,
      "learning_rate": 1.986241819094075e-05,
      "loss": 1.6352,
      "step": 216
    },
    {
      "epoch": 0.06273489447817288,
      "grad_norm": 1.3729904036841547,
      "learning_rate": 1.986089729444351e-05,
      "loss": 1.7041,
      "step": 217
    },
    {
      "epoch": 0.06302399537438566,
      "grad_norm": 1.3443809779979903,
      "learning_rate": 1.9859368096597925e-05,
      "loss": 1.7332,
      "step": 218
    },
    {
      "epoch": 0.06331309627059845,
      "grad_norm": 1.3367070791840931,
      "learning_rate": 1.9857830598691337e-05,
      "loss": 1.639,
      "step": 219
    },
    {
      "epoch": 0.06360219716681122,
      "grad_norm": 1.3799415387024292,
      "learning_rate": 1.9856284802018087e-05,
      "loss": 1.7675,
      "step": 220
    },
    {
      "epoch": 0.063891298063024,
      "grad_norm": 1.3464389753341524,
      "learning_rate": 1.9854730707879502e-05,
      "loss": 1.6702,
      "step": 221
    },
    {
      "epoch": 0.06418039895923677,
      "grad_norm": 1.3867842033565003,
      "learning_rate": 1.9853168317583874e-05,
      "loss": 1.7216,
      "step": 222
    },
    {
      "epoch": 0.06446949985544956,
      "grad_norm": 1.2892516322419894,
      "learning_rate": 1.9851597632446508e-05,
      "loss": 1.6755,
      "step": 223
    },
    {
      "epoch": 0.06475860075166233,
      "grad_norm": 1.4307425363074875,
      "learning_rate": 1.9850018653789665e-05,
      "loss": 1.7618,
      "step": 224
    },
    {
      "epoch": 0.06504770164787511,
      "grad_norm": 1.288672937098248,
      "learning_rate": 1.984843138294261e-05,
      "loss": 1.756,
      "step": 225
    },
    {
      "epoch": 0.06533680254408789,
      "grad_norm": 1.3540206108809039,
      "learning_rate": 1.9846835821241573e-05,
      "loss": 1.7373,
      "step": 226
    },
    {
      "epoch": 0.06562590344030067,
      "grad_norm": 1.3108759943656667,
      "learning_rate": 1.9845231970029774e-05,
      "loss": 1.6429,
      "step": 227
    },
    {
      "epoch": 0.06591500433651344,
      "grad_norm": 1.3549691917787245,
      "learning_rate": 1.9843619830657404e-05,
      "loss": 1.6692,
      "step": 228
    },
    {
      "epoch": 0.06620410523272623,
      "grad_norm": 1.350566839669386,
      "learning_rate": 1.984199940448164e-05,
      "loss": 1.6499,
      "step": 229
    },
    {
      "epoch": 0.066493206128939,
      "grad_norm": 1.3992430914185707,
      "learning_rate": 1.9840370692866624e-05,
      "loss": 1.7279,
      "step": 230
    },
    {
      "epoch": 0.06678230702515178,
      "grad_norm": 1.336273448437935,
      "learning_rate": 1.9838733697183482e-05,
      "loss": 1.7378,
      "step": 231
    },
    {
      "epoch": 0.06707140792136455,
      "grad_norm": 1.3714334108906678,
      "learning_rate": 1.9837088418810312e-05,
      "loss": 1.6845,
      "step": 232
    },
    {
      "epoch": 0.06736050881757734,
      "grad_norm": 1.331192814515942,
      "learning_rate": 1.983543485913218e-05,
      "loss": 1.7605,
      "step": 233
    },
    {
      "epoch": 0.06764960971379011,
      "grad_norm": 1.3590047127242209,
      "learning_rate": 1.983377301954113e-05,
      "loss": 1.7027,
      "step": 234
    },
    {
      "epoch": 0.0679387106100029,
      "grad_norm": 1.2628103031697302,
      "learning_rate": 1.983210290143617e-05,
      "loss": 1.6682,
      "step": 235
    },
    {
      "epoch": 0.06822781150621567,
      "grad_norm": 1.31345700201241,
      "learning_rate": 1.9830424506223286e-05,
      "loss": 1.7787,
      "step": 236
    },
    {
      "epoch": 0.06851691240242845,
      "grad_norm": 1.367333629142093,
      "learning_rate": 1.9828737835315422e-05,
      "loss": 1.7182,
      "step": 237
    },
    {
      "epoch": 0.06880601329864122,
      "grad_norm": 1.3814327986016506,
      "learning_rate": 1.9827042890132498e-05,
      "loss": 1.8058,
      "step": 238
    },
    {
      "epoch": 0.06909511419485401,
      "grad_norm": 1.2616372891530419,
      "learning_rate": 1.982533967210139e-05,
      "loss": 1.7131,
      "step": 239
    },
    {
      "epoch": 0.06938421509106678,
      "grad_norm": 1.3695087479555836,
      "learning_rate": 1.982362818265595e-05,
      "loss": 1.7163,
      "step": 240
    },
    {
      "epoch": 0.06967331598727956,
      "grad_norm": 1.3557868876434347,
      "learning_rate": 1.9821908423236983e-05,
      "loss": 1.7421,
      "step": 241
    },
    {
      "epoch": 0.06996241688349233,
      "grad_norm": 1.392442344450402,
      "learning_rate": 1.982018039529226e-05,
      "loss": 1.747,
      "step": 242
    },
    {
      "epoch": 0.07025151777970512,
      "grad_norm": 1.4471159007355705,
      "learning_rate": 1.9818444100276517e-05,
      "loss": 1.6452,
      "step": 243
    },
    {
      "epoch": 0.07054061867591789,
      "grad_norm": 1.323270395385573,
      "learning_rate": 1.981669953965144e-05,
      "loss": 1.795,
      "step": 244
    },
    {
      "epoch": 0.07082971957213068,
      "grad_norm": 1.32959668728154,
      "learning_rate": 1.9814946714885682e-05,
      "loss": 1.8027,
      "step": 245
    },
    {
      "epoch": 0.07111882046834345,
      "grad_norm": 1.3027587139671195,
      "learning_rate": 1.981318562745485e-05,
      "loss": 1.7326,
      "step": 246
    },
    {
      "epoch": 0.07140792136455623,
      "grad_norm": 1.328453385919723,
      "learning_rate": 1.9811416278841507e-05,
      "loss": 1.6839,
      "step": 247
    },
    {
      "epoch": 0.071697022260769,
      "grad_norm": 1.2271866822510065,
      "learning_rate": 1.980963867053517e-05,
      "loss": 1.6021,
      "step": 248
    },
    {
      "epoch": 0.07198612315698179,
      "grad_norm": 1.2908284006992323,
      "learning_rate": 1.9807852804032306e-05,
      "loss": 1.6339,
      "step": 249
    },
    {
      "epoch": 0.07227522405319456,
      "grad_norm": 1.373522694157661,
      "learning_rate": 1.9806058680836343e-05,
      "loss": 1.6902,
      "step": 250
    },
    {
      "epoch": 0.07256432494940734,
      "grad_norm": 1.2634235256940958,
      "learning_rate": 1.9804256302457653e-05,
      "loss": 1.6517,
      "step": 251
    },
    {
      "epoch": 0.07285342584562012,
      "grad_norm": 1.3278784508700443,
      "learning_rate": 1.9802445670413562e-05,
      "loss": 1.7972,
      "step": 252
    },
    {
      "epoch": 0.0731425267418329,
      "grad_norm": 1.3552386657179452,
      "learning_rate": 1.980062678622834e-05,
      "loss": 1.726,
      "step": 253
    },
    {
      "epoch": 0.07343162763804567,
      "grad_norm": 1.3742079147545718,
      "learning_rate": 1.9798799651433204e-05,
      "loss": 1.7317,
      "step": 254
    },
    {
      "epoch": 0.07372072853425846,
      "grad_norm": 1.4074238143735376,
      "learning_rate": 1.9796964267566322e-05,
      "loss": 1.7275,
      "step": 255
    },
    {
      "epoch": 0.07400982943047124,
      "grad_norm": 1.286876584367334,
      "learning_rate": 1.9795120636172804e-05,
      "loss": 1.6061,
      "step": 256
    },
    {
      "epoch": 0.07429893032668401,
      "grad_norm": 1.3521349118231711,
      "learning_rate": 1.9793268758804693e-05,
      "loss": 1.7237,
      "step": 257
    },
    {
      "epoch": 0.0745880312228968,
      "grad_norm": 1.2816583167182087,
      "learning_rate": 1.979140863702099e-05,
      "loss": 1.6783,
      "step": 258
    },
    {
      "epoch": 0.07487713211910957,
      "grad_norm": 1.3634926357253074,
      "learning_rate": 1.978954027238763e-05,
      "loss": 1.6784,
      "step": 259
    },
    {
      "epoch": 0.07516623301532235,
      "grad_norm": 1.4038229521295733,
      "learning_rate": 1.9787663666477486e-05,
      "loss": 1.6535,
      "step": 260
    },
    {
      "epoch": 0.07545533391153512,
      "grad_norm": 1.2842218606303095,
      "learning_rate": 1.9785778820870366e-05,
      "loss": 1.7058,
      "step": 261
    },
    {
      "epoch": 0.07574443480774791,
      "grad_norm": 1.347020075001843,
      "learning_rate": 1.9783885737153022e-05,
      "loss": 1.7832,
      "step": 262
    },
    {
      "epoch": 0.07603353570396068,
      "grad_norm": 1.26053283819741,
      "learning_rate": 1.9781984416919138e-05,
      "loss": 1.644,
      "step": 263
    },
    {
      "epoch": 0.07632263660017347,
      "grad_norm": 1.2495419382455348,
      "learning_rate": 1.9780074861769328e-05,
      "loss": 1.6807,
      "step": 264
    },
    {
      "epoch": 0.07661173749638624,
      "grad_norm": 1.7139411191235945,
      "learning_rate": 1.977815707331114e-05,
      "loss": 1.7222,
      "step": 265
    },
    {
      "epoch": 0.07690083839259902,
      "grad_norm": 1.4547863480275938,
      "learning_rate": 1.9776231053159063e-05,
      "loss": 1.7774,
      "step": 266
    },
    {
      "epoch": 0.07718993928881179,
      "grad_norm": 1.3437347227693044,
      "learning_rate": 1.97742968029345e-05,
      "loss": 1.7208,
      "step": 267
    },
    {
      "epoch": 0.07747904018502458,
      "grad_norm": 1.354739938343438,
      "learning_rate": 1.9772354324265796e-05,
      "loss": 1.6918,
      "step": 268
    },
    {
      "epoch": 0.07776814108123735,
      "grad_norm": 1.2544728177655773,
      "learning_rate": 1.977040361878821e-05,
      "loss": 1.6351,
      "step": 269
    },
    {
      "epoch": 0.07805724197745013,
      "grad_norm": 1.3417761983959458,
      "learning_rate": 1.9768444688143938e-05,
      "loss": 1.6564,
      "step": 270
    },
    {
      "epoch": 0.0783463428736629,
      "grad_norm": 1.3422116229415715,
      "learning_rate": 1.9766477533982094e-05,
      "loss": 1.8033,
      "step": 271
    },
    {
      "epoch": 0.07863544376987569,
      "grad_norm": 1.2706457122008528,
      "learning_rate": 1.976450215795872e-05,
      "loss": 1.6505,
      "step": 272
    },
    {
      "epoch": 0.07892454466608846,
      "grad_norm": 1.267352186352676,
      "learning_rate": 1.9762518561736782e-05,
      "loss": 1.6591,
      "step": 273
    },
    {
      "epoch": 0.07921364556230125,
      "grad_norm": 1.4466694988557338,
      "learning_rate": 1.976052674698615e-05,
      "loss": 1.7952,
      "step": 274
    },
    {
      "epoch": 0.07950274645851402,
      "grad_norm": 1.321832538613218,
      "learning_rate": 1.975852671538363e-05,
      "loss": 1.7061,
      "step": 275
    },
    {
      "epoch": 0.0797918473547268,
      "grad_norm": 1.2829050605849048,
      "learning_rate": 1.9756518468612934e-05,
      "loss": 1.6605,
      "step": 276
    },
    {
      "epoch": 0.08008094825093957,
      "grad_norm": 1.2987764926867447,
      "learning_rate": 1.9754502008364702e-05,
      "loss": 1.7047,
      "step": 277
    },
    {
      "epoch": 0.08037004914715236,
      "grad_norm": 1.2552473311885344,
      "learning_rate": 1.9752477336336473e-05,
      "loss": 1.7031,
      "step": 278
    },
    {
      "epoch": 0.08065915004336513,
      "grad_norm": 1.1985068633897615,
      "learning_rate": 1.9750444454232714e-05,
      "loss": 1.7382,
      "step": 279
    },
    {
      "epoch": 0.08094825093957791,
      "grad_norm": 1.314553485624065,
      "learning_rate": 1.9748403363764795e-05,
      "loss": 1.6491,
      "step": 280
    },
    {
      "epoch": 0.08123735183579069,
      "grad_norm": 1.294998222736541,
      "learning_rate": 1.9746354066650995e-05,
      "loss": 1.7044,
      "step": 281
    },
    {
      "epoch": 0.08152645273200347,
      "grad_norm": 1.3584436916798586,
      "learning_rate": 1.974429656461651e-05,
      "loss": 1.6672,
      "step": 282
    },
    {
      "epoch": 0.08181555362821624,
      "grad_norm": 1.3102650044903212,
      "learning_rate": 1.9742230859393432e-05,
      "loss": 1.6954,
      "step": 283
    },
    {
      "epoch": 0.08210465452442903,
      "grad_norm": 2.693221641429579,
      "learning_rate": 1.974015695272077e-05,
      "loss": 1.6296,
      "step": 284
    },
    {
      "epoch": 0.0823937554206418,
      "grad_norm": 1.3351424022555927,
      "learning_rate": 1.9738074846344427e-05,
      "loss": 1.6594,
      "step": 285
    },
    {
      "epoch": 0.08268285631685458,
      "grad_norm": 1.2766910430053877,
      "learning_rate": 1.973598454201722e-05,
      "loss": 1.6983,
      "step": 286
    },
    {
      "epoch": 0.08297195721306735,
      "grad_norm": 1.4668406259400848,
      "learning_rate": 1.973388604149886e-05,
      "loss": 1.6698,
      "step": 287
    },
    {
      "epoch": 0.08326105810928014,
      "grad_norm": 1.3104546778761357,
      "learning_rate": 1.9731779346555957e-05,
      "loss": 1.6723,
      "step": 288
    },
    {
      "epoch": 0.08355015900549291,
      "grad_norm": 1.531005942756331,
      "learning_rate": 1.9729664458962024e-05,
      "loss": 1.7863,
      "step": 289
    },
    {
      "epoch": 0.0838392599017057,
      "grad_norm": 1.3653225817314174,
      "learning_rate": 1.9727541380497465e-05,
      "loss": 1.7107,
      "step": 290
    },
    {
      "epoch": 0.08412836079791847,
      "grad_norm": 1.3892591267410797,
      "learning_rate": 1.972541011294959e-05,
      "loss": 1.6403,
      "step": 291
    },
    {
      "epoch": 0.08441746169413125,
      "grad_norm": 1.4118648874854935,
      "learning_rate": 1.972327065811259e-05,
      "loss": 1.6871,
      "step": 292
    },
    {
      "epoch": 0.08470656259034404,
      "grad_norm": 1.4198103761024161,
      "learning_rate": 1.972112301778756e-05,
      "loss": 1.7095,
      "step": 293
    },
    {
      "epoch": 0.08499566348655681,
      "grad_norm": 1.3896634687986935,
      "learning_rate": 1.9718967193782475e-05,
      "loss": 1.6347,
      "step": 294
    },
    {
      "epoch": 0.08528476438276959,
      "grad_norm": 1.3431516951928908,
      "learning_rate": 1.971680318791221e-05,
      "loss": 1.7874,
      "step": 295
    },
    {
      "epoch": 0.08557386527898236,
      "grad_norm": 1.3768126854675709,
      "learning_rate": 1.9714631001998517e-05,
      "loss": 1.6698,
      "step": 296
    },
    {
      "epoch": 0.08586296617519515,
      "grad_norm": 1.452732246792645,
      "learning_rate": 1.9712450637870048e-05,
      "loss": 1.7896,
      "step": 297
    },
    {
      "epoch": 0.08615206707140792,
      "grad_norm": 1.3268182615670008,
      "learning_rate": 1.9710262097362323e-05,
      "loss": 1.7162,
      "step": 298
    },
    {
      "epoch": 0.0864411679676207,
      "grad_norm": 1.3451587005617647,
      "learning_rate": 1.9708065382317763e-05,
      "loss": 1.6132,
      "step": 299
    },
    {
      "epoch": 0.08673026886383348,
      "grad_norm": 1.3567905514483718,
      "learning_rate": 1.9705860494585653e-05,
      "loss": 1.7181,
      "step": 300
    },
    {
      "epoch": 0.08701936976004626,
      "grad_norm": 1.4178400258624095,
      "learning_rate": 1.9703647436022174e-05,
      "loss": 1.7197,
      "step": 301
    },
    {
      "epoch": 0.08730847065625903,
      "grad_norm": 1.3176335169067535,
      "learning_rate": 1.970142620849038e-05,
      "loss": 1.6433,
      "step": 302
    },
    {
      "epoch": 0.08759757155247182,
      "grad_norm": 1.3848318848405352,
      "learning_rate": 1.9699196813860188e-05,
      "loss": 1.7975,
      "step": 303
    },
    {
      "epoch": 0.08788667244868459,
      "grad_norm": 1.4792413426843254,
      "learning_rate": 1.9696959254008416e-05,
      "loss": 1.7221,
      "step": 304
    },
    {
      "epoch": 0.08817577334489737,
      "grad_norm": 1.3289471852066919,
      "learning_rate": 1.9694713530818738e-05,
      "loss": 1.7631,
      "step": 305
    },
    {
      "epoch": 0.08846487424111014,
      "grad_norm": 1.4455763431259683,
      "learning_rate": 1.9692459646181706e-05,
      "loss": 1.758,
      "step": 306
    },
    {
      "epoch": 0.08875397513732293,
      "grad_norm": 1.4156278432006244,
      "learning_rate": 1.969019760199474e-05,
      "loss": 1.7106,
      "step": 307
    },
    {
      "epoch": 0.0890430760335357,
      "grad_norm": 1.335012985289393,
      "learning_rate": 1.9687927400162134e-05,
      "loss": 1.6303,
      "step": 308
    },
    {
      "epoch": 0.08933217692974849,
      "grad_norm": 1.3515063736435737,
      "learning_rate": 1.9685649042595045e-05,
      "loss": 1.6907,
      "step": 309
    },
    {
      "epoch": 0.08962127782596126,
      "grad_norm": 1.2906659764446389,
      "learning_rate": 1.96833625312115e-05,
      "loss": 1.7601,
      "step": 310
    },
    {
      "epoch": 0.08991037872217404,
      "grad_norm": 1.4128720604920837,
      "learning_rate": 1.968106786793638e-05,
      "loss": 1.685,
      "step": 311
    },
    {
      "epoch": 0.09019947961838681,
      "grad_norm": 1.3417671224975973,
      "learning_rate": 1.967876505470144e-05,
      "loss": 1.8098,
      "step": 312
    },
    {
      "epoch": 0.0904885805145996,
      "grad_norm": 1.2524980838386586,
      "learning_rate": 1.9676454093445293e-05,
      "loss": 1.7347,
      "step": 313
    },
    {
      "epoch": 0.09077768141081237,
      "grad_norm": 1.2903112596909276,
      "learning_rate": 1.9674134986113407e-05,
      "loss": 1.7183,
      "step": 314
    },
    {
      "epoch": 0.09106678230702515,
      "grad_norm": 1.315346772447713,
      "learning_rate": 1.9671807734658122e-05,
      "loss": 1.666,
      "step": 315
    },
    {
      "epoch": 0.09135588320323793,
      "grad_norm": 1.2815003343295048,
      "learning_rate": 1.966947234103861e-05,
      "loss": 1.6706,
      "step": 316
    },
    {
      "epoch": 0.09164498409945071,
      "grad_norm": 1.3032057492518452,
      "learning_rate": 1.9667128807220918e-05,
      "loss": 1.6852,
      "step": 317
    },
    {
      "epoch": 0.09193408499566348,
      "grad_norm": 1.399838923906023,
      "learning_rate": 1.9664777135177936e-05,
      "loss": 1.6786,
      "step": 318
    },
    {
      "epoch": 0.09222318589187627,
      "grad_norm": 1.3183521495611081,
      "learning_rate": 1.9662417326889408e-05,
      "loss": 1.7656,
      "step": 319
    },
    {
      "epoch": 0.09251228678808904,
      "grad_norm": 1.2919964291273878,
      "learning_rate": 1.966004938434193e-05,
      "loss": 1.7994,
      "step": 320
    },
    {
      "epoch": 0.09280138768430182,
      "grad_norm": 1.331071824643648,
      "learning_rate": 1.965767330952894e-05,
      "loss": 1.7315,
      "step": 321
    },
    {
      "epoch": 0.0930904885805146,
      "grad_norm": 1.3127780153887916,
      "learning_rate": 1.9655289104450724e-05,
      "loss": 1.58,
      "step": 322
    },
    {
      "epoch": 0.09337958947672738,
      "grad_norm": 1.427592788785929,
      "learning_rate": 1.9652896771114416e-05,
      "loss": 1.6926,
      "step": 323
    },
    {
      "epoch": 0.09366869037294015,
      "grad_norm": 1.3536283403103335,
      "learning_rate": 1.9650496311533985e-05,
      "loss": 1.6208,
      "step": 324
    },
    {
      "epoch": 0.09395779126915293,
      "grad_norm": 1.3049249531971474,
      "learning_rate": 1.9648087727730254e-05,
      "loss": 1.7007,
      "step": 325
    },
    {
      "epoch": 0.0942468921653657,
      "grad_norm": 8.074839924042823,
      "learning_rate": 1.9645671021730876e-05,
      "loss": 1.7401,
      "step": 326
    },
    {
      "epoch": 0.09453599306157849,
      "grad_norm": 1.689958269781092,
      "learning_rate": 1.9643246195570337e-05,
      "loss": 1.7612,
      "step": 327
    },
    {
      "epoch": 0.09482509395779128,
      "grad_norm": 1.3759192019092044,
      "learning_rate": 1.9640813251289968e-05,
      "loss": 1.6256,
      "step": 328
    },
    {
      "epoch": 0.09511419485400405,
      "grad_norm": 1.2852662388113483,
      "learning_rate": 1.9638372190937938e-05,
      "loss": 1.5961,
      "step": 329
    },
    {
      "epoch": 0.09540329575021683,
      "grad_norm": 1.3397296142416237,
      "learning_rate": 1.963592301656923e-05,
      "loss": 1.7228,
      "step": 330
    },
    {
      "epoch": 0.0956923966464296,
      "grad_norm": 1.3050235554989664,
      "learning_rate": 1.963346573024568e-05,
      "loss": 1.631,
      "step": 331
    },
    {
      "epoch": 0.09598149754264239,
      "grad_norm": 1.2687361679426568,
      "learning_rate": 1.9631000334035936e-05,
      "loss": 1.727,
      "step": 332
    },
    {
      "epoch": 0.09627059843885516,
      "grad_norm": 1.4434478962876411,
      "learning_rate": 1.962852683001548e-05,
      "loss": 1.6118,
      "step": 333
    },
    {
      "epoch": 0.09655969933506794,
      "grad_norm": 1.3835831592110455,
      "learning_rate": 1.9626045220266626e-05,
      "loss": 1.7184,
      "step": 334
    },
    {
      "epoch": 0.09684880023128072,
      "grad_norm": 1.285345165652217,
      "learning_rate": 1.96235555068785e-05,
      "loss": 1.7396,
      "step": 335
    },
    {
      "epoch": 0.0971379011274935,
      "grad_norm": 1.2741876410509978,
      "learning_rate": 1.9621057691947052e-05,
      "loss": 1.6827,
      "step": 336
    },
    {
      "epoch": 0.09742700202370627,
      "grad_norm": 1.3414110861817643,
      "learning_rate": 1.961855177757506e-05,
      "loss": 1.7649,
      "step": 337
    },
    {
      "epoch": 0.09771610291991906,
      "grad_norm": 1.3155148016487777,
      "learning_rate": 1.9616037765872115e-05,
      "loss": 1.6339,
      "step": 338
    },
    {
      "epoch": 0.09800520381613183,
      "grad_norm": 1.3422815166211728,
      "learning_rate": 1.9613515658954627e-05,
      "loss": 1.6248,
      "step": 339
    },
    {
      "epoch": 0.09829430471234461,
      "grad_norm": 1.2928851291446535,
      "learning_rate": 1.9610985458945813e-05,
      "loss": 1.6883,
      "step": 340
    },
    {
      "epoch": 0.09858340560855738,
      "grad_norm": 1.310011271967797,
      "learning_rate": 1.9608447167975716e-05,
      "loss": 1.5808,
      "step": 341
    },
    {
      "epoch": 0.09887250650477017,
      "grad_norm": 1.3065951865270455,
      "learning_rate": 1.960590078818118e-05,
      "loss": 1.6558,
      "step": 342
    },
    {
      "epoch": 0.09916160740098294,
      "grad_norm": 1.3349828850162693,
      "learning_rate": 1.960334632170587e-05,
      "loss": 1.7289,
      "step": 343
    },
    {
      "epoch": 0.09945070829719573,
      "grad_norm": 1.2943452424584643,
      "learning_rate": 1.960078377070024e-05,
      "loss": 1.613,
      "step": 344
    },
    {
      "epoch": 0.0997398091934085,
      "grad_norm": 1.4861917307646695,
      "learning_rate": 1.9598213137321568e-05,
      "loss": 1.7105,
      "step": 345
    },
    {
      "epoch": 0.10002891008962128,
      "grad_norm": 1.3722730686933504,
      "learning_rate": 1.959563442373393e-05,
      "loss": 1.7553,
      "step": 346
    },
    {
      "epoch": 0.10031801098583405,
      "grad_norm": 1.3453966326217088,
      "learning_rate": 1.95930476321082e-05,
      "loss": 1.737,
      "step": 347
    },
    {
      "epoch": 0.10060711188204684,
      "grad_norm": 1.3523784458893544,
      "learning_rate": 1.9590452764622056e-05,
      "loss": 1.6319,
      "step": 348
    },
    {
      "epoch": 0.10089621277825961,
      "grad_norm": 1.3502487223737636,
      "learning_rate": 1.9587849823459976e-05,
      "loss": 1.6941,
      "step": 349
    },
    {
      "epoch": 0.1011853136744724,
      "grad_norm": 1.398518128752155,
      "learning_rate": 1.9585238810813233e-05,
      "loss": 1.6475,
      "step": 350
    },
    {
      "epoch": 0.10147441457068516,
      "grad_norm": 1.200679760577495,
      "learning_rate": 1.9582619728879898e-05,
      "loss": 1.7337,
      "step": 351
    },
    {
      "epoch": 0.10176351546689795,
      "grad_norm": 1.3585311474803605,
      "learning_rate": 1.9579992579864827e-05,
      "loss": 1.7192,
      "step": 352
    },
    {
      "epoch": 0.10205261636311072,
      "grad_norm": 1.4501829885257667,
      "learning_rate": 1.9577357365979673e-05,
      "loss": 1.7227,
      "step": 353
    },
    {
      "epoch": 0.1023417172593235,
      "grad_norm": 1.2783896023412555,
      "learning_rate": 1.9574714089442884e-05,
      "loss": 1.7873,
      "step": 354
    },
    {
      "epoch": 0.10263081815553628,
      "grad_norm": 1.2457865420249044,
      "learning_rate": 1.9572062752479684e-05,
      "loss": 1.6078,
      "step": 355
    },
    {
      "epoch": 0.10291991905174906,
      "grad_norm": 1.298867064640292,
      "learning_rate": 1.956940335732209e-05,
      "loss": 1.7645,
      "step": 356
    },
    {
      "epoch": 0.10320901994796183,
      "grad_norm": 1.2599727613590195,
      "learning_rate": 1.95667359062089e-05,
      "loss": 1.6698,
      "step": 357
    },
    {
      "epoch": 0.10349812084417462,
      "grad_norm": 1.2508214829214046,
      "learning_rate": 1.9564060401385697e-05,
      "loss": 1.6279,
      "step": 358
    },
    {
      "epoch": 0.10378722174038739,
      "grad_norm": 1.1832707796953907,
      "learning_rate": 1.956137684510484e-05,
      "loss": 1.5799,
      "step": 359
    },
    {
      "epoch": 0.10407632263660017,
      "grad_norm": 1.3001383458399352,
      "learning_rate": 1.9558685239625467e-05,
      "loss": 1.6116,
      "step": 360
    },
    {
      "epoch": 0.10436542353281295,
      "grad_norm": 1.2428470322542244,
      "learning_rate": 1.9555985587213495e-05,
      "loss": 1.6716,
      "step": 361
    },
    {
      "epoch": 0.10465452442902573,
      "grad_norm": 1.3093773655524557,
      "learning_rate": 1.955327789014161e-05,
      "loss": 1.6915,
      "step": 362
    },
    {
      "epoch": 0.1049436253252385,
      "grad_norm": 1.383021031949293,
      "learning_rate": 1.9550562150689277e-05,
      "loss": 1.7317,
      "step": 363
    },
    {
      "epoch": 0.10523272622145129,
      "grad_norm": 1.329573981449402,
      "learning_rate": 1.9547838371142726e-05,
      "loss": 1.5913,
      "step": 364
    },
    {
      "epoch": 0.10552182711766407,
      "grad_norm": 1.4166951368370655,
      "learning_rate": 1.9545106553794962e-05,
      "loss": 1.6004,
      "step": 365
    },
    {
      "epoch": 0.10581092801387684,
      "grad_norm": 1.3377620749480776,
      "learning_rate": 1.9542366700945746e-05,
      "loss": 1.7907,
      "step": 366
    },
    {
      "epoch": 0.10610002891008963,
      "grad_norm": 1.3388694333443414,
      "learning_rate": 1.953961881490161e-05,
      "loss": 1.6979,
      "step": 367
    },
    {
      "epoch": 0.1063891298063024,
      "grad_norm": 1.3581005844418126,
      "learning_rate": 1.9536862897975852e-05,
      "loss": 1.6602,
      "step": 368
    },
    {
      "epoch": 0.10667823070251518,
      "grad_norm": 1.3242684005167475,
      "learning_rate": 1.9534098952488527e-05,
      "loss": 1.7131,
      "step": 369
    },
    {
      "epoch": 0.10696733159872795,
      "grad_norm": 1.2765833725967193,
      "learning_rate": 1.9531326980766444e-05,
      "loss": 1.7917,
      "step": 370
    },
    {
      "epoch": 0.10725643249494074,
      "grad_norm": 1.2957317184999853,
      "learning_rate": 1.952854698514318e-05,
      "loss": 1.6569,
      "step": 371
    },
    {
      "epoch": 0.10754553339115351,
      "grad_norm": 1.3050908294406351,
      "learning_rate": 1.952575896795906e-05,
      "loss": 1.5815,
      "step": 372
    },
    {
      "epoch": 0.1078346342873663,
      "grad_norm": 1.4007267378768093,
      "learning_rate": 1.952296293156116e-05,
      "loss": 1.6437,
      "step": 373
    },
    {
      "epoch": 0.10812373518357907,
      "grad_norm": 1.288112071905893,
      "learning_rate": 1.952015887830331e-05,
      "loss": 1.7148,
      "step": 374
    },
    {
      "epoch": 0.10841283607979185,
      "grad_norm": 1.4132340648150645,
      "learning_rate": 1.9517346810546094e-05,
      "loss": 1.7236,
      "step": 375
    },
    {
      "epoch": 0.10870193697600462,
      "grad_norm": 1.3520649123502502,
      "learning_rate": 1.951452673065683e-05,
      "loss": 1.7027,
      "step": 376
    },
    {
      "epoch": 0.10899103787221741,
      "grad_norm": 1.344289195938231,
      "learning_rate": 1.951169864100959e-05,
      "loss": 1.6753,
      "step": 377
    },
    {
      "epoch": 0.10928013876843018,
      "grad_norm": 1.2727655897790644,
      "learning_rate": 1.950886254398519e-05,
      "loss": 1.7595,
      "step": 378
    },
    {
      "epoch": 0.10956923966464296,
      "grad_norm": 1.3488163579726409,
      "learning_rate": 1.9506018441971186e-05,
      "loss": 1.6687,
      "step": 379
    },
    {
      "epoch": 0.10985834056085574,
      "grad_norm": 1.3474523993403258,
      "learning_rate": 1.9503166337361863e-05,
      "loss": 1.6676,
      "step": 380
    },
    {
      "epoch": 0.11014744145706852,
      "grad_norm": 1.2758765009101956,
      "learning_rate": 1.9500306232558263e-05,
      "loss": 1.6571,
      "step": 381
    },
    {
      "epoch": 0.11043654235328129,
      "grad_norm": 1.2196542542467945,
      "learning_rate": 1.9497438129968144e-05,
      "loss": 1.6551,
      "step": 382
    },
    {
      "epoch": 0.11072564324949408,
      "grad_norm": 1.301338171513902,
      "learning_rate": 1.9494562032006e-05,
      "loss": 1.6548,
      "step": 383
    },
    {
      "epoch": 0.11101474414570685,
      "grad_norm": 1.3203283874481997,
      "learning_rate": 1.9491677941093075e-05,
      "loss": 1.7338,
      "step": 384
    },
    {
      "epoch": 0.11130384504191963,
      "grad_norm": 1.3024684784620577,
      "learning_rate": 1.9488785859657314e-05,
      "loss": 1.5528,
      "step": 385
    },
    {
      "epoch": 0.1115929459381324,
      "grad_norm": 1.2625026147962017,
      "learning_rate": 1.9485885790133402e-05,
      "loss": 1.6967,
      "step": 386
    },
    {
      "epoch": 0.11188204683434519,
      "grad_norm": 1.49068356270609,
      "learning_rate": 1.9482977734962753e-05,
      "loss": 1.7096,
      "step": 387
    },
    {
      "epoch": 0.11217114773055796,
      "grad_norm": 1.3130185709588313,
      "learning_rate": 1.9480061696593502e-05,
      "loss": 1.7174,
      "step": 388
    },
    {
      "epoch": 0.11246024862677075,
      "grad_norm": 1.2742915647508892,
      "learning_rate": 1.9477137677480493e-05,
      "loss": 1.6696,
      "step": 389
    },
    {
      "epoch": 0.11274934952298352,
      "grad_norm": 1.3627324489936457,
      "learning_rate": 1.9474205680085302e-05,
      "loss": 1.6542,
      "step": 390
    },
    {
      "epoch": 0.1130384504191963,
      "grad_norm": 1.3437349437643114,
      "learning_rate": 1.947126570687622e-05,
      "loss": 1.6747,
      "step": 391
    },
    {
      "epoch": 0.11332755131540907,
      "grad_norm": 1.2678997121287945,
      "learning_rate": 1.9468317760328244e-05,
      "loss": 1.6918,
      "step": 392
    },
    {
      "epoch": 0.11361665221162186,
      "grad_norm": 1.2929245484043537,
      "learning_rate": 1.9465361842923087e-05,
      "loss": 1.7217,
      "step": 393
    },
    {
      "epoch": 0.11390575310783463,
      "grad_norm": 1.2499932038294317,
      "learning_rate": 1.9462397957149182e-05,
      "loss": 1.6128,
      "step": 394
    },
    {
      "epoch": 0.11419485400404741,
      "grad_norm": 1.221557019735438,
      "learning_rate": 1.9459426105501653e-05,
      "loss": 1.6467,
      "step": 395
    },
    {
      "epoch": 0.11448395490026018,
      "grad_norm": 1.2693539163839065,
      "learning_rate": 1.945644629048234e-05,
      "loss": 1.7283,
      "step": 396
    },
    {
      "epoch": 0.11477305579647297,
      "grad_norm": 1.21274399309353,
      "learning_rate": 1.9453458514599794e-05,
      "loss": 1.7008,
      "step": 397
    },
    {
      "epoch": 0.11506215669268574,
      "grad_norm": 1.2218147838259905,
      "learning_rate": 1.945046278036925e-05,
      "loss": 1.7437,
      "step": 398
    },
    {
      "epoch": 0.11535125758889853,
      "grad_norm": 1.2968685896469103,
      "learning_rate": 1.9447459090312656e-05,
      "loss": 1.6983,
      "step": 399
    },
    {
      "epoch": 0.1156403584851113,
      "grad_norm": 1.2334993392446711,
      "learning_rate": 1.9444447446958658e-05,
      "loss": 1.7061,
      "step": 400
    },
    {
      "epoch": 0.11592945938132408,
      "grad_norm": 1.2984129611641557,
      "learning_rate": 1.9441427852842584e-05,
      "loss": 1.663,
      "step": 401
    },
    {
      "epoch": 0.11621856027753687,
      "grad_norm": 1.2475965365995239,
      "learning_rate": 1.943840031050647e-05,
      "loss": 1.6427,
      "step": 402
    },
    {
      "epoch": 0.11650766117374964,
      "grad_norm": 1.2694332854835886,
      "learning_rate": 1.943536482249904e-05,
      "loss": 1.7047,
      "step": 403
    },
    {
      "epoch": 0.11679676206996242,
      "grad_norm": 1.2646968431962804,
      "learning_rate": 1.94323213913757e-05,
      "loss": 1.6885,
      "step": 404
    },
    {
      "epoch": 0.1170858629661752,
      "grad_norm": 1.2293159012021218,
      "learning_rate": 1.9429270019698553e-05,
      "loss": 1.757,
      "step": 405
    },
    {
      "epoch": 0.11737496386238798,
      "grad_norm": 1.2484085036943657,
      "learning_rate": 1.9426210710036374e-05,
      "loss": 1.6741,
      "step": 406
    },
    {
      "epoch": 0.11766406475860075,
      "grad_norm": 1.43839701132394,
      "learning_rate": 1.9423143464964633e-05,
      "loss": 1.6768,
      "step": 407
    },
    {
      "epoch": 0.11795316565481354,
      "grad_norm": 1.292908155403034,
      "learning_rate": 1.9420068287065476e-05,
      "loss": 1.6126,
      "step": 408
    },
    {
      "epoch": 0.1182422665510263,
      "grad_norm": 1.370089960450461,
      "learning_rate": 1.9416985178927724e-05,
      "loss": 1.731,
      "step": 409
    },
    {
      "epoch": 0.11853136744723909,
      "grad_norm": 1.3262176752311388,
      "learning_rate": 1.941389414314688e-05,
      "loss": 1.6579,
      "step": 410
    },
    {
      "epoch": 0.11882046834345186,
      "grad_norm": 1.339305591546624,
      "learning_rate": 1.9410795182325113e-05,
      "loss": 1.7495,
      "step": 411
    },
    {
      "epoch": 0.11910956923966465,
      "grad_norm": 1.341821261680916,
      "learning_rate": 1.940768829907127e-05,
      "loss": 1.7005,
      "step": 412
    },
    {
      "epoch": 0.11939867013587742,
      "grad_norm": 1.276085151807024,
      "learning_rate": 1.9404573496000867e-05,
      "loss": 1.6043,
      "step": 413
    },
    {
      "epoch": 0.1196877710320902,
      "grad_norm": 1.302570146214278,
      "learning_rate": 1.940145077573608e-05,
      "loss": 1.6722,
      "step": 414
    },
    {
      "epoch": 0.11997687192830297,
      "grad_norm": 1.2582811301580585,
      "learning_rate": 1.9398320140905765e-05,
      "loss": 1.6383,
      "step": 415
    },
    {
      "epoch": 0.12026597282451576,
      "grad_norm": 1.258733310784144,
      "learning_rate": 1.9395181594145428e-05,
      "loss": 1.5806,
      "step": 416
    },
    {
      "epoch": 0.12055507372072853,
      "grad_norm": 1.2988890119500425,
      "learning_rate": 1.9392035138097235e-05,
      "loss": 1.7048,
      "step": 417
    },
    {
      "epoch": 0.12084417461694132,
      "grad_norm": 1.29122939937491,
      "learning_rate": 1.938888077541002e-05,
      "loss": 1.5844,
      "step": 418
    },
    {
      "epoch": 0.12113327551315409,
      "grad_norm": 1.2756402426612268,
      "learning_rate": 1.9385718508739263e-05,
      "loss": 1.6935,
      "step": 419
    },
    {
      "epoch": 0.12142237640936687,
      "grad_norm": 1.3542547094512076,
      "learning_rate": 1.9382548340747105e-05,
      "loss": 1.6781,
      "step": 420
    },
    {
      "epoch": 0.12171147730557964,
      "grad_norm": 1.3120014515081435,
      "learning_rate": 1.937937027410234e-05,
      "loss": 1.6601,
      "step": 421
    },
    {
      "epoch": 0.12200057820179243,
      "grad_norm": 1.2955651895370563,
      "learning_rate": 1.9376184311480404e-05,
      "loss": 1.59,
      "step": 422
    },
    {
      "epoch": 0.1222896790980052,
      "grad_norm": 1.3192882851086605,
      "learning_rate": 1.9372990455563384e-05,
      "loss": 1.6291,
      "step": 423
    },
    {
      "epoch": 0.12257877999421798,
      "grad_norm": 1.4080370752052034,
      "learning_rate": 1.9369788709040014e-05,
      "loss": 1.5843,
      "step": 424
    },
    {
      "epoch": 0.12286788089043076,
      "grad_norm": 1.309890861391924,
      "learning_rate": 1.936657907460567e-05,
      "loss": 1.7576,
      "step": 425
    },
    {
      "epoch": 0.12315698178664354,
      "grad_norm": 1.3296661949737203,
      "learning_rate": 1.936336155496236e-05,
      "loss": 1.6412,
      "step": 426
    },
    {
      "epoch": 0.12344608268285631,
      "grad_norm": 1.2137542022109604,
      "learning_rate": 1.9360136152818746e-05,
      "loss": 1.6677,
      "step": 427
    },
    {
      "epoch": 0.1237351835790691,
      "grad_norm": 1.304478996513997,
      "learning_rate": 1.9356902870890114e-05,
      "loss": 1.6157,
      "step": 428
    },
    {
      "epoch": 0.12402428447528187,
      "grad_norm": 1.2362598408529617,
      "learning_rate": 1.9353661711898382e-05,
      "loss": 1.6493,
      "step": 429
    },
    {
      "epoch": 0.12431338537149465,
      "grad_norm": 1.3868990501411094,
      "learning_rate": 1.9350412678572114e-05,
      "loss": 1.7838,
      "step": 430
    },
    {
      "epoch": 0.12460248626770742,
      "grad_norm": 1.3811144656508765,
      "learning_rate": 1.934715577364648e-05,
      "loss": 1.6164,
      "step": 431
    },
    {
      "epoch": 0.12489158716392021,
      "grad_norm": 1.1886802269263785,
      "learning_rate": 1.93438909998633e-05,
      "loss": 1.6917,
      "step": 432
    },
    {
      "epoch": 0.125180688060133,
      "grad_norm": 1.373861158324828,
      "learning_rate": 1.9340618359971003e-05,
      "loss": 1.7273,
      "step": 433
    },
    {
      "epoch": 0.12546978895634575,
      "grad_norm": 1.360887411910304,
      "learning_rate": 1.9337337856724647e-05,
      "loss": 1.6129,
      "step": 434
    },
    {
      "epoch": 0.12575888985255854,
      "grad_norm": 1.2310796622388105,
      "learning_rate": 1.9334049492885906e-05,
      "loss": 1.686,
      "step": 435
    },
    {
      "epoch": 0.12604799074877132,
      "grad_norm": 1.2150119003782842,
      "learning_rate": 1.9330753271223073e-05,
      "loss": 1.7097,
      "step": 436
    },
    {
      "epoch": 0.1263370916449841,
      "grad_norm": 1.3137735986745411,
      "learning_rate": 1.932744919451106e-05,
      "loss": 1.7312,
      "step": 437
    },
    {
      "epoch": 0.1266261925411969,
      "grad_norm": 1.19581010090534,
      "learning_rate": 1.9324137265531383e-05,
      "loss": 1.7073,
      "step": 438
    },
    {
      "epoch": 0.12691529343740965,
      "grad_norm": 1.1864958252257658,
      "learning_rate": 1.9320817487072174e-05,
      "loss": 1.6634,
      "step": 439
    },
    {
      "epoch": 0.12720439433362243,
      "grad_norm": 1.2519130079679632,
      "learning_rate": 1.9317489861928173e-05,
      "loss": 1.6686,
      "step": 440
    },
    {
      "epoch": 0.12749349522983522,
      "grad_norm": 1.2098081719210536,
      "learning_rate": 1.9314154392900728e-05,
      "loss": 1.697,
      "step": 441
    },
    {
      "epoch": 0.127782596126048,
      "grad_norm": 1.297112805486722,
      "learning_rate": 1.9310811082797784e-05,
      "loss": 1.5979,
      "step": 442
    },
    {
      "epoch": 0.12807169702226076,
      "grad_norm": 1.3006902970607843,
      "learning_rate": 1.930745993443389e-05,
      "loss": 1.5987,
      "step": 443
    },
    {
      "epoch": 0.12836079791847355,
      "grad_norm": 1.3888684660127295,
      "learning_rate": 1.9304100950630197e-05,
      "loss": 1.6047,
      "step": 444
    },
    {
      "epoch": 0.12864989881468633,
      "grad_norm": 1.3325945840074427,
      "learning_rate": 1.9300734134214443e-05,
      "loss": 1.6813,
      "step": 445
    },
    {
      "epoch": 0.12893899971089912,
      "grad_norm": 1.3053169239838074,
      "learning_rate": 1.9297359488020974e-05,
      "loss": 1.8416,
      "step": 446
    },
    {
      "epoch": 0.12922810060711187,
      "grad_norm": 1.1934876374596244,
      "learning_rate": 1.9293977014890716e-05,
      "loss": 1.7005,
      "step": 447
    },
    {
      "epoch": 0.12951720150332466,
      "grad_norm": 1.2640391776060138,
      "learning_rate": 1.929058671767119e-05,
      "loss": 1.6578,
      "step": 448
    },
    {
      "epoch": 0.12980630239953744,
      "grad_norm": 1.2646711203792806,
      "learning_rate": 1.9287188599216492e-05,
      "loss": 1.6699,
      "step": 449
    },
    {
      "epoch": 0.13009540329575023,
      "grad_norm": 1.2320011875752572,
      "learning_rate": 1.928378266238732e-05,
      "loss": 1.6525,
      "step": 450
    },
    {
      "epoch": 0.13038450419196299,
      "grad_norm": 1.34573408732214,
      "learning_rate": 1.9280368910050943e-05,
      "loss": 1.7015,
      "step": 451
    },
    {
      "epoch": 0.13067360508817577,
      "grad_norm": 1.2353671099343633,
      "learning_rate": 1.9276947345081213e-05,
      "loss": 1.6397,
      "step": 452
    },
    {
      "epoch": 0.13096270598438856,
      "grad_norm": 1.2197831661100311,
      "learning_rate": 1.9273517970358557e-05,
      "loss": 1.6836,
      "step": 453
    },
    {
      "epoch": 0.13125180688060134,
      "grad_norm": 1.328367741401711,
      "learning_rate": 1.9270080788769978e-05,
      "loss": 1.7178,
      "step": 454
    },
    {
      "epoch": 0.1315409077768141,
      "grad_norm": 1.2668567068996344,
      "learning_rate": 1.9266635803209047e-05,
      "loss": 1.7255,
      "step": 455
    },
    {
      "epoch": 0.13183000867302688,
      "grad_norm": 1.2322035718739746,
      "learning_rate": 1.9263183016575912e-05,
      "loss": 1.612,
      "step": 456
    },
    {
      "epoch": 0.13211910956923967,
      "grad_norm": 1.244606536596135,
      "learning_rate": 1.9259722431777286e-05,
      "loss": 1.7994,
      "step": 457
    },
    {
      "epoch": 0.13240821046545245,
      "grad_norm": 1.2384126698619529,
      "learning_rate": 1.9256254051726443e-05,
      "loss": 1.6401,
      "step": 458
    },
    {
      "epoch": 0.1326973113616652,
      "grad_norm": 1.2780953635907069,
      "learning_rate": 1.925277787934322e-05,
      "loss": 1.5826,
      "step": 459
    },
    {
      "epoch": 0.132986412257878,
      "grad_norm": 1.2798396535904,
      "learning_rate": 1.9249293917554017e-05,
      "loss": 1.674,
      "step": 460
    },
    {
      "epoch": 0.13327551315409078,
      "grad_norm": 1.2799203698146353,
      "learning_rate": 1.924580216929179e-05,
      "loss": 1.6331,
      "step": 461
    },
    {
      "epoch": 0.13356461405030357,
      "grad_norm": 1.2111633406900368,
      "learning_rate": 1.9242302637496057e-05,
      "loss": 1.6682,
      "step": 462
    },
    {
      "epoch": 0.13385371494651632,
      "grad_norm": 1.2112859487688281,
      "learning_rate": 1.9238795325112867e-05,
      "loss": 1.6484,
      "step": 463
    },
    {
      "epoch": 0.1341428158427291,
      "grad_norm": 1.288135140665184,
      "learning_rate": 1.9235280235094852e-05,
      "loss": 1.626,
      "step": 464
    },
    {
      "epoch": 0.1344319167389419,
      "grad_norm": 1.308513915152716,
      "learning_rate": 1.923175737040116e-05,
      "loss": 1.6597,
      "step": 465
    },
    {
      "epoch": 0.13472101763515468,
      "grad_norm": 1.3262569041157928,
      "learning_rate": 1.9228226733997503e-05,
      "loss": 1.7642,
      "step": 466
    },
    {
      "epoch": 0.13501011853136743,
      "grad_norm": 1.3424290380477188,
      "learning_rate": 1.9224688328856127e-05,
      "loss": 1.7465,
      "step": 467
    },
    {
      "epoch": 0.13529921942758022,
      "grad_norm": 1.2651279914803906,
      "learning_rate": 1.9221142157955825e-05,
      "loss": 1.6126,
      "step": 468
    },
    {
      "epoch": 0.135588320323793,
      "grad_norm": 1.3558655043001309,
      "learning_rate": 1.921758822428192e-05,
      "loss": 1.7556,
      "step": 469
    },
    {
      "epoch": 0.1358774212200058,
      "grad_norm": 1.402001055732072,
      "learning_rate": 1.921402653082628e-05,
      "loss": 1.7992,
      "step": 470
    },
    {
      "epoch": 0.13616652211621855,
      "grad_norm": 1.3182689558852307,
      "learning_rate": 1.9210457080587285e-05,
      "loss": 1.6262,
      "step": 471
    },
    {
      "epoch": 0.13645562301243133,
      "grad_norm": 1.2723760551262082,
      "learning_rate": 1.9206879876569877e-05,
      "loss": 1.7064,
      "step": 472
    },
    {
      "epoch": 0.13674472390864412,
      "grad_norm": 1.2954400935028445,
      "learning_rate": 1.92032949217855e-05,
      "loss": 1.6374,
      "step": 473
    },
    {
      "epoch": 0.1370338248048569,
      "grad_norm": 1.2402948267042866,
      "learning_rate": 1.9199702219252128e-05,
      "loss": 1.5951,
      "step": 474
    },
    {
      "epoch": 0.1373229257010697,
      "grad_norm": 1.2739630063945762,
      "learning_rate": 1.919610177199426e-05,
      "loss": 1.7167,
      "step": 475
    },
    {
      "epoch": 0.13761202659728244,
      "grad_norm": 1.4113213475938031,
      "learning_rate": 1.9192493583042922e-05,
      "loss": 1.5751,
      "step": 476
    },
    {
      "epoch": 0.13790112749349523,
      "grad_norm": 1.2872821986905296,
      "learning_rate": 1.9188877655435645e-05,
      "loss": 1.65,
      "step": 477
    },
    {
      "epoch": 0.13819022838970801,
      "grad_norm": 1.2480655610937714,
      "learning_rate": 1.918525399221648e-05,
      "loss": 1.6882,
      "step": 478
    },
    {
      "epoch": 0.1384793292859208,
      "grad_norm": 1.2681694886992483,
      "learning_rate": 1.9181622596435993e-05,
      "loss": 1.7528,
      "step": 479
    },
    {
      "epoch": 0.13876843018213356,
      "grad_norm": 1.226404564427387,
      "learning_rate": 1.917798347115125e-05,
      "loss": 1.5918,
      "step": 480
    },
    {
      "epoch": 0.13905753107834634,
      "grad_norm": 1.2687390154564733,
      "learning_rate": 1.9174336619425842e-05,
      "loss": 1.7533,
      "step": 481
    },
    {
      "epoch": 0.13934663197455913,
      "grad_norm": 1.2254053559809934,
      "learning_rate": 1.9170682044329845e-05,
      "loss": 1.5968,
      "step": 482
    },
    {
      "epoch": 0.1396357328707719,
      "grad_norm": 1.2732869875336201,
      "learning_rate": 1.9167019748939847e-05,
      "loss": 1.6944,
      "step": 483
    },
    {
      "epoch": 0.13992483376698467,
      "grad_norm": 1.2541243190490785,
      "learning_rate": 1.9163349736338935e-05,
      "loss": 1.6237,
      "step": 484
    },
    {
      "epoch": 0.14021393466319745,
      "grad_norm": 1.3291823171118335,
      "learning_rate": 1.915967200961669e-05,
      "loss": 1.5639,
      "step": 485
    },
    {
      "epoch": 0.14050303555941024,
      "grad_norm": 1.2528167731427482,
      "learning_rate": 1.9155986571869184e-05,
      "loss": 1.7458,
      "step": 486
    },
    {
      "epoch": 0.14079213645562302,
      "grad_norm": 1.2107095174484985,
      "learning_rate": 1.9152293426198997e-05,
      "loss": 1.6838,
      "step": 487
    },
    {
      "epoch": 0.14108123735183578,
      "grad_norm": 1.2070210238330805,
      "learning_rate": 1.9148592575715175e-05,
      "loss": 1.6957,
      "step": 488
    },
    {
      "epoch": 0.14137033824804857,
      "grad_norm": 1.3501466427051918,
      "learning_rate": 1.9144884023533262e-05,
      "loss": 1.6848,
      "step": 489
    },
    {
      "epoch": 0.14165943914426135,
      "grad_norm": 1.289837942023481,
      "learning_rate": 1.914116777277529e-05,
      "loss": 1.6398,
      "step": 490
    },
    {
      "epoch": 0.14194854004047414,
      "grad_norm": 1.2072886210991958,
      "learning_rate": 1.9137443826569758e-05,
      "loss": 1.6803,
      "step": 491
    },
    {
      "epoch": 0.1422376409366869,
      "grad_norm": 1.2261256594899907,
      "learning_rate": 1.9133712188051664e-05,
      "loss": 1.8143,
      "step": 492
    },
    {
      "epoch": 0.14252674183289968,
      "grad_norm": 1.191555412915695,
      "learning_rate": 1.912997286036246e-05,
      "loss": 1.7273,
      "step": 493
    },
    {
      "epoch": 0.14281584272911246,
      "grad_norm": 1.2668769555610104,
      "learning_rate": 1.9126225846650083e-05,
      "loss": 1.6729,
      "step": 494
    },
    {
      "epoch": 0.14310494362532525,
      "grad_norm": 1.2310680992910144,
      "learning_rate": 1.912247115006894e-05,
      "loss": 1.5509,
      "step": 495
    },
    {
      "epoch": 0.143394044521538,
      "grad_norm": 1.2540456207457256,
      "learning_rate": 1.91187087737799e-05,
      "loss": 1.6512,
      "step": 496
    },
    {
      "epoch": 0.1436831454177508,
      "grad_norm": 1.2983726556931725,
      "learning_rate": 1.9114938720950307e-05,
      "loss": 1.6434,
      "step": 497
    },
    {
      "epoch": 0.14397224631396358,
      "grad_norm": 1.2424942795398506,
      "learning_rate": 1.9111160994753957e-05,
      "loss": 1.712,
      "step": 498
    },
    {
      "epoch": 0.14426134721017636,
      "grad_norm": 1.2549250776078968,
      "learning_rate": 1.9107375598371112e-05,
      "loss": 1.5818,
      "step": 499
    },
    {
      "epoch": 0.14455044810638912,
      "grad_norm": 1.171699859687026,
      "learning_rate": 1.910358253498849e-05,
      "loss": 1.6075,
      "step": 500
    },
    {
      "epoch": 0.1448395490026019,
      "grad_norm": 1.2660287905231888,
      "learning_rate": 1.9099781807799263e-05,
      "loss": 1.6403,
      "step": 501
    },
    {
      "epoch": 0.1451286498988147,
      "grad_norm": 1.2138215200444276,
      "learning_rate": 1.9095973420003056e-05,
      "loss": 1.6214,
      "step": 502
    },
    {
      "epoch": 0.14541775079502747,
      "grad_norm": 1.2379532082411215,
      "learning_rate": 1.909215737480594e-05,
      "loss": 1.5915,
      "step": 503
    },
    {
      "epoch": 0.14570685169124023,
      "grad_norm": 1.2691642883216934,
      "learning_rate": 1.9088333675420432e-05,
      "loss": 1.6913,
      "step": 504
    },
    {
      "epoch": 0.14599595258745302,
      "grad_norm": 1.3365845490222084,
      "learning_rate": 1.9084502325065494e-05,
      "loss": 1.7174,
      "step": 505
    },
    {
      "epoch": 0.1462850534836658,
      "grad_norm": 1.2933988368966916,
      "learning_rate": 1.9080663326966536e-05,
      "loss": 1.6947,
      "step": 506
    },
    {
      "epoch": 0.14657415437987859,
      "grad_norm": 1.1997391314057113,
      "learning_rate": 1.9076816684355398e-05,
      "loss": 1.6454,
      "step": 507
    },
    {
      "epoch": 0.14686325527609134,
      "grad_norm": 1.2253205782269874,
      "learning_rate": 1.9072962400470348e-05,
      "loss": 1.6475,
      "step": 508
    },
    {
      "epoch": 0.14715235617230413,
      "grad_norm": 1.149822466017807,
      "learning_rate": 1.9069100478556112e-05,
      "loss": 1.5374,
      "step": 509
    },
    {
      "epoch": 0.1474414570685169,
      "grad_norm": 1.2247546090753016,
      "learning_rate": 1.906523092186382e-05,
      "loss": 1.7436,
      "step": 510
    },
    {
      "epoch": 0.1477305579647297,
      "grad_norm": 1.1999484360154251,
      "learning_rate": 1.9061353733651036e-05,
      "loss": 1.6747,
      "step": 511
    },
    {
      "epoch": 0.14801965886094248,
      "grad_norm": 1.1947780812062576,
      "learning_rate": 1.905746891718176e-05,
      "loss": 1.59,
      "step": 512
    },
    {
      "epoch": 0.14830875975715524,
      "grad_norm": 1.2483481364125735,
      "learning_rate": 1.9053576475726406e-05,
      "loss": 1.6343,
      "step": 513
    },
    {
      "epoch": 0.14859786065336802,
      "grad_norm": 1.2303596001336339,
      "learning_rate": 1.9049676412561803e-05,
      "loss": 1.701,
      "step": 514
    },
    {
      "epoch": 0.1488869615495808,
      "grad_norm": 1.2116845736311206,
      "learning_rate": 1.9045768730971198e-05,
      "loss": 1.7507,
      "step": 515
    },
    {
      "epoch": 0.1491760624457936,
      "grad_norm": 1.299089146099033,
      "learning_rate": 1.904185343424426e-05,
      "loss": 1.7314,
      "step": 516
    },
    {
      "epoch": 0.14946516334200635,
      "grad_norm": 1.2754215966625346,
      "learning_rate": 1.9037930525677058e-05,
      "loss": 1.6955,
      "step": 517
    },
    {
      "epoch": 0.14975426423821914,
      "grad_norm": 1.2864098329743727,
      "learning_rate": 1.9034000008572073e-05,
      "loss": 1.6926,
      "step": 518
    },
    {
      "epoch": 0.15004336513443192,
      "grad_norm": 1.2920675731077655,
      "learning_rate": 1.9030061886238195e-05,
      "loss": 1.6528,
      "step": 519
    },
    {
      "epoch": 0.1503324660306447,
      "grad_norm": 1.2265090425552296,
      "learning_rate": 1.9026116161990712e-05,
      "loss": 1.647,
      "step": 520
    },
    {
      "epoch": 0.15062156692685746,
      "grad_norm": 1.305532335335852,
      "learning_rate": 1.902216283915131e-05,
      "loss": 1.6085,
      "step": 521
    },
    {
      "epoch": 0.15091066782307025,
      "grad_norm": 1.2996647218299158,
      "learning_rate": 1.901820192104808e-05,
      "loss": 1.6858,
      "step": 522
    },
    {
      "epoch": 0.15119976871928303,
      "grad_norm": 1.2728823004232528,
      "learning_rate": 1.9014233411015496e-05,
      "loss": 1.6152,
      "step": 523
    },
    {
      "epoch": 0.15148886961549582,
      "grad_norm": 1.237846119035289,
      "learning_rate": 1.901025731239443e-05,
      "loss": 1.7852,
      "step": 524
    },
    {
      "epoch": 0.15177797051170858,
      "grad_norm": 1.263777772165666,
      "learning_rate": 1.9006273628532146e-05,
      "loss": 1.6714,
      "step": 525
    },
    {
      "epoch": 0.15206707140792136,
      "grad_norm": 1.1966029829151998,
      "learning_rate": 1.9002282362782284e-05,
      "loss": 1.5772,
      "step": 526
    },
    {
      "epoch": 0.15235617230413415,
      "grad_norm": 1.2573815442286682,
      "learning_rate": 1.8998283518504874e-05,
      "loss": 1.6586,
      "step": 527
    },
    {
      "epoch": 0.15264527320034693,
      "grad_norm": 1.3315550643097303,
      "learning_rate": 1.899427709906632e-05,
      "loss": 1.5776,
      "step": 528
    },
    {
      "epoch": 0.1529343740965597,
      "grad_norm": 1.27455167369078,
      "learning_rate": 1.899026310783941e-05,
      "loss": 1.6383,
      "step": 529
    },
    {
      "epoch": 0.15322347499277247,
      "grad_norm": 1.2177716057481967,
      "learning_rate": 1.8986241548203305e-05,
      "loss": 1.7259,
      "step": 530
    },
    {
      "epoch": 0.15351257588898526,
      "grad_norm": 1.4172388350158833,
      "learning_rate": 1.898221242354353e-05,
      "loss": 1.6084,
      "step": 531
    },
    {
      "epoch": 0.15380167678519804,
      "grad_norm": 1.2977973632114062,
      "learning_rate": 1.8978175737251984e-05,
      "loss": 1.7148,
      "step": 532
    },
    {
      "epoch": 0.1540907776814108,
      "grad_norm": 1.3288320048402222,
      "learning_rate": 1.897413149272694e-05,
      "loss": 1.7092,
      "step": 533
    },
    {
      "epoch": 0.15437987857762359,
      "grad_norm": 1.362022094342137,
      "learning_rate": 1.8970079693373014e-05,
      "loss": 1.6727,
      "step": 534
    },
    {
      "epoch": 0.15466897947383637,
      "grad_norm": 1.2868416836773502,
      "learning_rate": 1.8966020342601204e-05,
      "loss": 1.6544,
      "step": 535
    },
    {
      "epoch": 0.15495808037004916,
      "grad_norm": 1.2255959793529978,
      "learning_rate": 1.8961953443828853e-05,
      "loss": 1.595,
      "step": 536
    },
    {
      "epoch": 0.1552471812662619,
      "grad_norm": 1.2265187293683357,
      "learning_rate": 1.8957879000479647e-05,
      "loss": 1.6067,
      "step": 537
    },
    {
      "epoch": 0.1555362821624747,
      "grad_norm": 1.198940426728034,
      "learning_rate": 1.8953797015983654e-05,
      "loss": 1.6146,
      "step": 538
    },
    {
      "epoch": 0.15582538305868748,
      "grad_norm": 1.190556858810557,
      "learning_rate": 1.8949707493777262e-05,
      "loss": 1.699,
      "step": 539
    },
    {
      "epoch": 0.15611448395490027,
      "grad_norm": 1.1336765732790504,
      "learning_rate": 1.894561043730322e-05,
      "loss": 1.5966,
      "step": 540
    },
    {
      "epoch": 0.15640358485111303,
      "grad_norm": 1.1667137603744793,
      "learning_rate": 1.8941505850010612e-05,
      "loss": 1.6644,
      "step": 541
    },
    {
      "epoch": 0.1566926857473258,
      "grad_norm": 1.3524401243741684,
      "learning_rate": 1.8937393735354865e-05,
      "loss": 1.6905,
      "step": 542
    },
    {
      "epoch": 0.1569817866435386,
      "grad_norm": 1.2665686572396084,
      "learning_rate": 1.8933274096797745e-05,
      "loss": 1.5214,
      "step": 543
    },
    {
      "epoch": 0.15727088753975138,
      "grad_norm": 1.2333567103961114,
      "learning_rate": 1.892914693780735e-05,
      "loss": 1.6044,
      "step": 544
    },
    {
      "epoch": 0.15755998843596414,
      "grad_norm": 1.2700672132910338,
      "learning_rate": 1.8925012261858103e-05,
      "loss": 1.6295,
      "step": 545
    },
    {
      "epoch": 0.15784908933217692,
      "grad_norm": 1.3269513720335189,
      "learning_rate": 1.892087007243077e-05,
      "loss": 1.5688,
      "step": 546
    },
    {
      "epoch": 0.1581381902283897,
      "grad_norm": 1.444704489940057,
      "learning_rate": 1.8916720373012425e-05,
      "loss": 1.6048,
      "step": 547
    },
    {
      "epoch": 0.1584272911246025,
      "grad_norm": 1.2362771720222352,
      "learning_rate": 1.8912563167096478e-05,
      "loss": 1.6548,
      "step": 548
    },
    {
      "epoch": 0.15871639202081528,
      "grad_norm": 1.3841466139534755,
      "learning_rate": 1.8908398458182647e-05,
      "loss": 1.5571,
      "step": 549
    },
    {
      "epoch": 0.15900549291702804,
      "grad_norm": 1.3571467399932051,
      "learning_rate": 1.8904226249776977e-05,
      "loss": 1.741,
      "step": 550
    },
    {
      "epoch": 0.15929459381324082,
      "grad_norm": 1.3013296148852647,
      "learning_rate": 1.8900046545391817e-05,
      "loss": 1.6925,
      "step": 551
    },
    {
      "epoch": 0.1595836947094536,
      "grad_norm": 1.4245052366468887,
      "learning_rate": 1.8895859348545832e-05,
      "loss": 1.6373,
      "step": 552
    },
    {
      "epoch": 0.1598727956056664,
      "grad_norm": 1.4862334656509957,
      "learning_rate": 1.8891664662763998e-05,
      "loss": 1.7684,
      "step": 553
    },
    {
      "epoch": 0.16016189650187915,
      "grad_norm": 1.4591040164361495,
      "learning_rate": 1.8887462491577585e-05,
      "loss": 1.6048,
      "step": 554
    },
    {
      "epoch": 0.16045099739809193,
      "grad_norm": 1.3905790801269649,
      "learning_rate": 1.8883252838524167e-05,
      "loss": 1.7057,
      "step": 555
    },
    {
      "epoch": 0.16074009829430472,
      "grad_norm": 1.296590703552253,
      "learning_rate": 1.8879035707147634e-05,
      "loss": 1.5708,
      "step": 556
    },
    {
      "epoch": 0.1610291991905175,
      "grad_norm": 1.4224940976099907,
      "learning_rate": 1.887481110099814e-05,
      "loss": 1.6613,
      "step": 557
    },
    {
      "epoch": 0.16131830008673026,
      "grad_norm": 1.209416335942298,
      "learning_rate": 1.8870579023632163e-05,
      "loss": 1.6734,
      "step": 558
    },
    {
      "epoch": 0.16160740098294304,
      "grad_norm": 1.3856482467534985,
      "learning_rate": 1.886633947861245e-05,
      "loss": 1.6075,
      "step": 559
    },
    {
      "epoch": 0.16189650187915583,
      "grad_norm": 1.298197860303536,
      "learning_rate": 1.8862092469508043e-05,
      "loss": 1.7309,
      "step": 560
    },
    {
      "epoch": 0.16218560277536861,
      "grad_norm": 1.2884008702670706,
      "learning_rate": 1.8857837999894265e-05,
      "loss": 1.7367,
      "step": 561
    },
    {
      "epoch": 0.16247470367158137,
      "grad_norm": 1.3295149830237742,
      "learning_rate": 1.8853576073352724e-05,
      "loss": 1.6373,
      "step": 562
    },
    {
      "epoch": 0.16276380456779416,
      "grad_norm": 1.244943878603924,
      "learning_rate": 1.8849306693471297e-05,
      "loss": 1.7586,
      "step": 563
    },
    {
      "epoch": 0.16305290546400694,
      "grad_norm": 1.4571917293027963,
      "learning_rate": 1.884502986384414e-05,
      "loss": 1.6308,
      "step": 564
    },
    {
      "epoch": 0.16334200636021973,
      "grad_norm": 1.2273508652400458,
      "learning_rate": 1.884074558807169e-05,
      "loss": 1.7033,
      "step": 565
    },
    {
      "epoch": 0.16363110725643248,
      "grad_norm": 1.3596367181968156,
      "learning_rate": 1.8836453869760633e-05,
      "loss": 1.7688,
      "step": 566
    },
    {
      "epoch": 0.16392020815264527,
      "grad_norm": 1.1782222979069559,
      "learning_rate": 1.883215471252394e-05,
      "loss": 1.6864,
      "step": 567
    },
    {
      "epoch": 0.16420930904885805,
      "grad_norm": 1.1876724035966852,
      "learning_rate": 1.882784811998083e-05,
      "loss": 1.651,
      "step": 568
    },
    {
      "epoch": 0.16449840994507084,
      "grad_norm": 1.2807588292559342,
      "learning_rate": 1.8823534095756787e-05,
      "loss": 1.7127,
      "step": 569
    },
    {
      "epoch": 0.1647875108412836,
      "grad_norm": 1.1959776867333827,
      "learning_rate": 1.881921264348355e-05,
      "loss": 1.6117,
      "step": 570
    },
    {
      "epoch": 0.16507661173749638,
      "grad_norm": 1.1977710735486093,
      "learning_rate": 1.881488376679912e-05,
      "loss": 1.6682,
      "step": 571
    },
    {
      "epoch": 0.16536571263370917,
      "grad_norm": 1.2133712471676765,
      "learning_rate": 1.8810547469347737e-05,
      "loss": 1.701,
      "step": 572
    },
    {
      "epoch": 0.16565481352992195,
      "grad_norm": 1.1615194296116027,
      "learning_rate": 1.880620375477989e-05,
      "loss": 1.6957,
      "step": 573
    },
    {
      "epoch": 0.1659439144261347,
      "grad_norm": 1.2434508263741582,
      "learning_rate": 1.880185262675231e-05,
      "loss": 1.6334,
      "step": 574
    },
    {
      "epoch": 0.1662330153223475,
      "grad_norm": 1.1996625528866651,
      "learning_rate": 1.8797494088927984e-05,
      "loss": 1.7306,
      "step": 575
    },
    {
      "epoch": 0.16652211621856028,
      "grad_norm": 1.25369862209721,
      "learning_rate": 1.8793128144976114e-05,
      "loss": 1.6611,
      "step": 576
    },
    {
      "epoch": 0.16681121711477306,
      "grad_norm": 1.2478485036785194,
      "learning_rate": 1.878875479857216e-05,
      "loss": 1.6418,
      "step": 577
    },
    {
      "epoch": 0.16710031801098582,
      "grad_norm": 1.3329254058696818,
      "learning_rate": 1.8784374053397792e-05,
      "loss": 1.6929,
      "step": 578
    },
    {
      "epoch": 0.1673894189071986,
      "grad_norm": 1.3690716864028842,
      "learning_rate": 1.8779985913140927e-05,
      "loss": 1.6979,
      "step": 579
    },
    {
      "epoch": 0.1676785198034114,
      "grad_norm": 1.2637256897353144,
      "learning_rate": 1.877559038149569e-05,
      "loss": 1.6229,
      "step": 580
    },
    {
      "epoch": 0.16796762069962418,
      "grad_norm": 1.3196301095363738,
      "learning_rate": 1.8771187462162452e-05,
      "loss": 1.6806,
      "step": 581
    },
    {
      "epoch": 0.16825672159583693,
      "grad_norm": 1.2751675322567242,
      "learning_rate": 1.8766777158847777e-05,
      "loss": 1.7331,
      "step": 582
    },
    {
      "epoch": 0.16854582249204972,
      "grad_norm": 1.2217555466284777,
      "learning_rate": 1.8762359475264464e-05,
      "loss": 1.6439,
      "step": 583
    },
    {
      "epoch": 0.1688349233882625,
      "grad_norm": 1.2138119129265534,
      "learning_rate": 1.8757934415131513e-05,
      "loss": 1.6229,
      "step": 584
    },
    {
      "epoch": 0.1691240242844753,
      "grad_norm": 1.2481504467799762,
      "learning_rate": 1.8753501982174142e-05,
      "loss": 1.6527,
      "step": 585
    },
    {
      "epoch": 0.16941312518068807,
      "grad_norm": 1.1551738364013855,
      "learning_rate": 1.8749062180123778e-05,
      "loss": 1.6202,
      "step": 586
    },
    {
      "epoch": 0.16970222607690083,
      "grad_norm": 1.221158314927715,
      "learning_rate": 1.8744615012718038e-05,
      "loss": 1.674,
      "step": 587
    },
    {
      "epoch": 0.16999132697311362,
      "grad_norm": 8.33538007172163,
      "learning_rate": 1.8740160483700754e-05,
      "loss": 1.7745,
      "step": 588
    },
    {
      "epoch": 0.1702804278693264,
      "grad_norm": 1.2594866391024326,
      "learning_rate": 1.8735698596821948e-05,
      "loss": 1.7114,
      "step": 589
    },
    {
      "epoch": 0.17056952876553919,
      "grad_norm": 1.1762832459510764,
      "learning_rate": 1.8731229355837834e-05,
      "loss": 1.6011,
      "step": 590
    },
    {
      "epoch": 0.17085862966175194,
      "grad_norm": 1.2410876448625578,
      "learning_rate": 1.8726752764510827e-05,
      "loss": 1.5289,
      "step": 591
    },
    {
      "epoch": 0.17114773055796473,
      "grad_norm": 1.2509478918932246,
      "learning_rate": 1.8722268826609524e-05,
      "loss": 1.6921,
      "step": 592
    },
    {
      "epoch": 0.1714368314541775,
      "grad_norm": 1.2876270619453627,
      "learning_rate": 1.87177775459087e-05,
      "loss": 1.6632,
      "step": 593
    },
    {
      "epoch": 0.1717259323503903,
      "grad_norm": 1.236444222374534,
      "learning_rate": 1.8713278926189324e-05,
      "loss": 1.628,
      "step": 594
    },
    {
      "epoch": 0.17201503324660306,
      "grad_norm": 1.3325085992804107,
      "learning_rate": 1.8708772971238528e-05,
      "loss": 1.7139,
      "step": 595
    },
    {
      "epoch": 0.17230413414281584,
      "grad_norm": 1.3054803605234377,
      "learning_rate": 1.8704259684849636e-05,
      "loss": 1.6715,
      "step": 596
    },
    {
      "epoch": 0.17259323503902863,
      "grad_norm": 1.2114622828234745,
      "learning_rate": 1.869973907082214e-05,
      "loss": 1.6289,
      "step": 597
    },
    {
      "epoch": 0.1728823359352414,
      "grad_norm": 1.3255943035899682,
      "learning_rate": 1.8695211132961686e-05,
      "loss": 1.5875,
      "step": 598
    },
    {
      "epoch": 0.17317143683145417,
      "grad_norm": 1.2436093130787584,
      "learning_rate": 1.86906758750801e-05,
      "loss": 1.6891,
      "step": 599
    },
    {
      "epoch": 0.17346053772766695,
      "grad_norm": 1.1789555197737465,
      "learning_rate": 1.8686133300995374e-05,
      "loss": 1.7097,
      "step": 600
    },
    {
      "epoch": 0.17374963862387974,
      "grad_norm": 1.3205013238390153,
      "learning_rate": 1.8681583414531644e-05,
      "loss": 1.6392,
      "step": 601
    },
    {
      "epoch": 0.17403873952009252,
      "grad_norm": 1.1648007546672043,
      "learning_rate": 1.8677026219519212e-05,
      "loss": 1.675,
      "step": 602
    },
    {
      "epoch": 0.17432784041630528,
      "grad_norm": 1.2633263020507286,
      "learning_rate": 1.8672461719794532e-05,
      "loss": 1.5957,
      "step": 603
    },
    {
      "epoch": 0.17461694131251806,
      "grad_norm": 1.2570430438589553,
      "learning_rate": 1.8667889919200203e-05,
      "loss": 1.7026,
      "step": 604
    },
    {
      "epoch": 0.17490604220873085,
      "grad_norm": 1.203370562367749,
      "learning_rate": 1.8663310821584976e-05,
      "loss": 1.7054,
      "step": 605
    },
    {
      "epoch": 0.17519514310494363,
      "grad_norm": 1.2387868298918792,
      "learning_rate": 1.865872443080374e-05,
      "loss": 1.7107,
      "step": 606
    },
    {
      "epoch": 0.1754842440011564,
      "grad_norm": 1.197249289822212,
      "learning_rate": 1.8654130750717526e-05,
      "loss": 1.5961,
      "step": 607
    },
    {
      "epoch": 0.17577334489736918,
      "grad_norm": 1.1922137775006185,
      "learning_rate": 1.8649529785193502e-05,
      "loss": 1.6349,
      "step": 608
    },
    {
      "epoch": 0.17606244579358196,
      "grad_norm": 1.238013689830228,
      "learning_rate": 1.8644921538104967e-05,
      "loss": 1.6487,
      "step": 609
    },
    {
      "epoch": 0.17635154668979475,
      "grad_norm": 1.198711220073165,
      "learning_rate": 1.864030601333135e-05,
      "loss": 1.6152,
      "step": 610
    },
    {
      "epoch": 0.1766406475860075,
      "grad_norm": 1.1608845614417607,
      "learning_rate": 1.8635683214758213e-05,
      "loss": 1.6114,
      "step": 611
    },
    {
      "epoch": 0.1769297484822203,
      "grad_norm": 1.223527743047107,
      "learning_rate": 1.8631053146277233e-05,
      "loss": 1.5324,
      "step": 612
    },
    {
      "epoch": 0.17721884937843307,
      "grad_norm": 1.2244849270886753,
      "learning_rate": 1.8626415811786215e-05,
      "loss": 1.7019,
      "step": 613
    },
    {
      "epoch": 0.17750795027464586,
      "grad_norm": 1.2422110644907864,
      "learning_rate": 1.8621771215189065e-05,
      "loss": 1.6713,
      "step": 614
    },
    {
      "epoch": 0.17779705117085862,
      "grad_norm": 1.210470089961754,
      "learning_rate": 1.861711936039583e-05,
      "loss": 1.6569,
      "step": 615
    },
    {
      "epoch": 0.1780861520670714,
      "grad_norm": 1.207776683125208,
      "learning_rate": 1.8612460251322637e-05,
      "loss": 1.5186,
      "step": 616
    },
    {
      "epoch": 0.1783752529632842,
      "grad_norm": 1.2462777938743828,
      "learning_rate": 1.8607793891891744e-05,
      "loss": 1.7027,
      "step": 617
    },
    {
      "epoch": 0.17866435385949697,
      "grad_norm": 1.2187274782133524,
      "learning_rate": 1.8603120286031496e-05,
      "loss": 1.7697,
      "step": 618
    },
    {
      "epoch": 0.17895345475570976,
      "grad_norm": 1.2542678538432044,
      "learning_rate": 1.859843943767635e-05,
      "loss": 1.7017,
      "step": 619
    },
    {
      "epoch": 0.1792425556519225,
      "grad_norm": 1.1778318765224127,
      "learning_rate": 1.8593751350766854e-05,
      "loss": 1.6475,
      "step": 620
    },
    {
      "epoch": 0.1795316565481353,
      "grad_norm": 1.2711298413207304,
      "learning_rate": 1.858905602924965e-05,
      "loss": 1.6394,
      "step": 621
    },
    {
      "epoch": 0.17982075744434808,
      "grad_norm": 1.232207257276827,
      "learning_rate": 1.8584353477077476e-05,
      "loss": 1.6074,
      "step": 622
    },
    {
      "epoch": 0.18010985834056087,
      "grad_norm": 1.2540297946919436,
      "learning_rate": 1.857964369820915e-05,
      "loss": 1.6683,
      "step": 623
    },
    {
      "epoch": 0.18039895923677363,
      "grad_norm": 1.2340677553615835,
      "learning_rate": 1.857492669660957e-05,
      "loss": 1.6716,
      "step": 624
    },
    {
      "epoch": 0.1806880601329864,
      "grad_norm": 1.2065806455145902,
      "learning_rate": 1.857020247624973e-05,
      "loss": 1.5365,
      "step": 625
    },
    {
      "epoch": 0.1809771610291992,
      "grad_norm": 1.2263400735274184,
      "learning_rate": 1.856547104110669e-05,
      "loss": 1.5683,
      "step": 626
    },
    {
      "epoch": 0.18126626192541198,
      "grad_norm": 1.347386958745164,
      "learning_rate": 1.8560732395163585e-05,
      "loss": 1.7436,
      "step": 627
    },
    {
      "epoch": 0.18155536282162474,
      "grad_norm": 1.3051300929582412,
      "learning_rate": 1.8555986542409615e-05,
      "loss": 1.7476,
      "step": 628
    },
    {
      "epoch": 0.18184446371783752,
      "grad_norm": 1.3479071566939564,
      "learning_rate": 1.855123348684006e-05,
      "loss": 1.6623,
      "step": 629
    },
    {
      "epoch": 0.1821335646140503,
      "grad_norm": 1.2568804583760171,
      "learning_rate": 1.8546473232456255e-05,
      "loss": 1.6444,
      "step": 630
    },
    {
      "epoch": 0.1824226655102631,
      "grad_norm": 1.2426342867282067,
      "learning_rate": 1.85417057832656e-05,
      "loss": 1.5964,
      "step": 631
    },
    {
      "epoch": 0.18271176640647585,
      "grad_norm": 1.2638177775731838,
      "learning_rate": 1.853693114328154e-05,
      "loss": 1.5322,
      "step": 632
    },
    {
      "epoch": 0.18300086730268864,
      "grad_norm": 1.164069058129512,
      "learning_rate": 1.8532149316523587e-05,
      "loss": 1.5849,
      "step": 633
    },
    {
      "epoch": 0.18328996819890142,
      "grad_norm": 1.297248418393498,
      "learning_rate": 1.8527360307017306e-05,
      "loss": 1.7388,
      "step": 634
    },
    {
      "epoch": 0.1835790690951142,
      "grad_norm": 1.1341179063142641,
      "learning_rate": 1.852256411879429e-05,
      "loss": 1.7354,
      "step": 635
    },
    {
      "epoch": 0.18386816999132696,
      "grad_norm": 1.215237098350405,
      "learning_rate": 1.8517760755892197e-05,
      "loss": 1.6552,
      "step": 636
    },
    {
      "epoch": 0.18415727088753975,
      "grad_norm": 1.1982757446916235,
      "learning_rate": 1.8512950222354704e-05,
      "loss": 1.6313,
      "step": 637
    },
    {
      "epoch": 0.18444637178375253,
      "grad_norm": 1.2063061953012788,
      "learning_rate": 1.850813252223154e-05,
      "loss": 1.6108,
      "step": 638
    },
    {
      "epoch": 0.18473547267996532,
      "grad_norm": 1.2120232512101812,
      "learning_rate": 1.8503307659578466e-05,
      "loss": 1.6357,
      "step": 639
    },
    {
      "epoch": 0.18502457357617808,
      "grad_norm": 1.2479587047666476,
      "learning_rate": 1.8498475638457265e-05,
      "loss": 1.6444,
      "step": 640
    },
    {
      "epoch": 0.18531367447239086,
      "grad_norm": 1.2758952934748575,
      "learning_rate": 1.849363646293575e-05,
      "loss": 1.632,
      "step": 641
    },
    {
      "epoch": 0.18560277536860365,
      "grad_norm": 1.1490779391882489,
      "learning_rate": 1.8488790137087764e-05,
      "loss": 1.6316,
      "step": 642
    },
    {
      "epoch": 0.18589187626481643,
      "grad_norm": 1.195110289142274,
      "learning_rate": 1.8483936664993152e-05,
      "loss": 1.6018,
      "step": 643
    },
    {
      "epoch": 0.1861809771610292,
      "grad_norm": 1.283174200034784,
      "learning_rate": 1.847907605073779e-05,
      "loss": 1.6488,
      "step": 644
    },
    {
      "epoch": 0.18647007805724197,
      "grad_norm": 1.2662030686633299,
      "learning_rate": 1.847420829841357e-05,
      "loss": 1.6538,
      "step": 645
    },
    {
      "epoch": 0.18675917895345476,
      "grad_norm": 1.2237194592018654,
      "learning_rate": 1.8469333412118375e-05,
      "loss": 1.6171,
      "step": 646
    },
    {
      "epoch": 0.18704827984966754,
      "grad_norm": 1.155988769874379,
      "learning_rate": 1.846445139595611e-05,
      "loss": 1.5969,
      "step": 647
    },
    {
      "epoch": 0.1873373807458803,
      "grad_norm": 1.2416911437703613,
      "learning_rate": 1.845956225403667e-05,
      "loss": 1.6744,
      "step": 648
    },
    {
      "epoch": 0.18762648164209308,
      "grad_norm": 1.1875358817766466,
      "learning_rate": 1.8454665990475967e-05,
      "loss": 1.5515,
      "step": 649
    },
    {
      "epoch": 0.18791558253830587,
      "grad_norm": 1.1818241708803934,
      "learning_rate": 1.844976260939589e-05,
      "loss": 1.5487,
      "step": 650
    },
    {
      "epoch": 0.18820468343451865,
      "grad_norm": 1.1840647734438647,
      "learning_rate": 1.8444852114924325e-05,
      "loss": 1.6148,
      "step": 651
    },
    {
      "epoch": 0.1884937843307314,
      "grad_norm": 1.2706827502060694,
      "learning_rate": 1.8439934511195147e-05,
      "loss": 1.6112,
      "step": 652
    },
    {
      "epoch": 0.1887828852269442,
      "grad_norm": 1.220131649175792,
      "learning_rate": 1.843500980234822e-05,
      "loss": 1.6449,
      "step": 653
    },
    {
      "epoch": 0.18907198612315698,
      "grad_norm": 1.103922527658955,
      "learning_rate": 1.8430077992529388e-05,
      "loss": 1.6444,
      "step": 654
    },
    {
      "epoch": 0.18936108701936977,
      "grad_norm": 1.1899809619524633,
      "learning_rate": 1.8425139085890466e-05,
      "loss": 1.5967,
      "step": 655
    },
    {
      "epoch": 0.18965018791558255,
      "grad_norm": 1.2244806367070153,
      "learning_rate": 1.8420193086589252e-05,
      "loss": 1.6504,
      "step": 656
    },
    {
      "epoch": 0.1899392888117953,
      "grad_norm": 1.1912067775204556,
      "learning_rate": 1.841523999878951e-05,
      "loss": 1.718,
      "step": 657
    },
    {
      "epoch": 0.1902283897080081,
      "grad_norm": 1.2117478146838403,
      "learning_rate": 1.8410279826660976e-05,
      "loss": 1.5804,
      "step": 658
    },
    {
      "epoch": 0.19051749060422088,
      "grad_norm": 1.1341524351911334,
      "learning_rate": 1.840531257437934e-05,
      "loss": 1.6287,
      "step": 659
    },
    {
      "epoch": 0.19080659150043366,
      "grad_norm": 1.1914356537636899,
      "learning_rate": 1.8400338246126265e-05,
      "loss": 1.6563,
      "step": 660
    },
    {
      "epoch": 0.19109569239664642,
      "grad_norm": 1.4105624776944174,
      "learning_rate": 1.839535684608936e-05,
      "loss": 1.7139,
      "step": 661
    },
    {
      "epoch": 0.1913847932928592,
      "grad_norm": 1.1645723203046823,
      "learning_rate": 1.8390368378462197e-05,
      "loss": 1.7166,
      "step": 662
    },
    {
      "epoch": 0.191673894189072,
      "grad_norm": 1.2936969076152187,
      "learning_rate": 1.8385372847444296e-05,
      "loss": 1.6482,
      "step": 663
    },
    {
      "epoch": 0.19196299508528478,
      "grad_norm": 1.1569033962294648,
      "learning_rate": 1.838037025724111e-05,
      "loss": 1.6843,
      "step": 664
    },
    {
      "epoch": 0.19225209598149753,
      "grad_norm": 1.1405851915704683,
      "learning_rate": 1.837536061206405e-05,
      "loss": 1.6033,
      "step": 665
    },
    {
      "epoch": 0.19254119687771032,
      "grad_norm": 1.2542382381366322,
      "learning_rate": 1.8370343916130467e-05,
      "loss": 1.6706,
      "step": 666
    },
    {
      "epoch": 0.1928302977739231,
      "grad_norm": 1.2159546224332614,
      "learning_rate": 1.8365320173663638e-05,
      "loss": 1.6918,
      "step": 667
    },
    {
      "epoch": 0.1931193986701359,
      "grad_norm": 1.2532817994557603,
      "learning_rate": 1.8360289388892773e-05,
      "loss": 1.6683,
      "step": 668
    },
    {
      "epoch": 0.19340849956634865,
      "grad_norm": 1.2395321203333016,
      "learning_rate": 1.8355251566053016e-05,
      "loss": 1.629,
      "step": 669
    },
    {
      "epoch": 0.19369760046256143,
      "grad_norm": 1.1689572668191017,
      "learning_rate": 1.8350206709385437e-05,
      "loss": 1.6462,
      "step": 670
    },
    {
      "epoch": 0.19398670135877422,
      "grad_norm": 1.1981131507977363,
      "learning_rate": 1.8345154823137015e-05,
      "loss": 1.5874,
      "step": 671
    },
    {
      "epoch": 0.194275802254987,
      "grad_norm": 1.1854854067275211,
      "learning_rate": 1.834009591156067e-05,
      "loss": 1.5968,
      "step": 672
    },
    {
      "epoch": 0.19456490315119976,
      "grad_norm": 1.1769715429764516,
      "learning_rate": 1.8335029978915207e-05,
      "loss": 1.5856,
      "step": 673
    },
    {
      "epoch": 0.19485400404741254,
      "grad_norm": 1.1759193873525724,
      "learning_rate": 1.8329957029465367e-05,
      "loss": 1.676,
      "step": 674
    },
    {
      "epoch": 0.19514310494362533,
      "grad_norm": 1.2034916407402159,
      "learning_rate": 1.8324877067481782e-05,
      "loss": 1.6385,
      "step": 675
    },
    {
      "epoch": 0.1954322058398381,
      "grad_norm": 1.1498736100928149,
      "learning_rate": 1.8319790097240998e-05,
      "loss": 1.5622,
      "step": 676
    },
    {
      "epoch": 0.19572130673605087,
      "grad_norm": 1.1833854630281373,
      "learning_rate": 1.8314696123025456e-05,
      "loss": 1.5906,
      "step": 677
    },
    {
      "epoch": 0.19601040763226366,
      "grad_norm": 1.3064848599083938,
      "learning_rate": 1.8309595149123486e-05,
      "loss": 1.5784,
      "step": 678
    },
    {
      "epoch": 0.19629950852847644,
      "grad_norm": 1.207598354497605,
      "learning_rate": 1.8304487179829325e-05,
      "loss": 1.6724,
      "step": 679
    },
    {
      "epoch": 0.19658860942468923,
      "grad_norm": 1.3657596862796557,
      "learning_rate": 1.829937221944309e-05,
      "loss": 1.6808,
      "step": 680
    },
    {
      "epoch": 0.19687771032090198,
      "grad_norm": 1.2300109610049414,
      "learning_rate": 1.8294250272270787e-05,
      "loss": 1.701,
      "step": 681
    },
    {
      "epoch": 0.19716681121711477,
      "grad_norm": 1.1932414325288254,
      "learning_rate": 1.82891213426243e-05,
      "loss": 1.6844,
      "step": 682
    },
    {
      "epoch": 0.19745591211332755,
      "grad_norm": 1.2040899317538816,
      "learning_rate": 1.8283985434821394e-05,
      "loss": 1.6674,
      "step": 683
    },
    {
      "epoch": 0.19774501300954034,
      "grad_norm": 1.20051829187703,
      "learning_rate": 1.827884255318571e-05,
      "loss": 1.6988,
      "step": 684
    },
    {
      "epoch": 0.1980341139057531,
      "grad_norm": 1.1808862977245562,
      "learning_rate": 1.8273692702046754e-05,
      "loss": 1.5823,
      "step": 685
    },
    {
      "epoch": 0.19832321480196588,
      "grad_norm": 1.226726509340714,
      "learning_rate": 1.8268535885739905e-05,
      "loss": 1.6371,
      "step": 686
    },
    {
      "epoch": 0.19861231569817867,
      "grad_norm": 1.254783885869565,
      "learning_rate": 1.8263372108606404e-05,
      "loss": 1.5849,
      "step": 687
    },
    {
      "epoch": 0.19890141659439145,
      "grad_norm": 1.2210257028134839,
      "learning_rate": 1.8258201374993355e-05,
      "loss": 1.7316,
      "step": 688
    },
    {
      "epoch": 0.1991905174906042,
      "grad_norm": 1.27099587011664,
      "learning_rate": 1.8253023689253707e-05,
      "loss": 1.592,
      "step": 689
    },
    {
      "epoch": 0.199479618386817,
      "grad_norm": 1.366689582202721,
      "learning_rate": 1.8247839055746276e-05,
      "loss": 1.5789,
      "step": 690
    },
    {
      "epoch": 0.19976871928302978,
      "grad_norm": 1.1755442721898772,
      "learning_rate": 1.8242647478835717e-05,
      "loss": 1.6069,
      "step": 691
    },
    {
      "epoch": 0.20005782017924256,
      "grad_norm": 1.2600721970141062,
      "learning_rate": 1.8237448962892537e-05,
      "loss": 1.7384,
      "step": 692
    },
    {
      "epoch": 0.20034692107545535,
      "grad_norm": 1.22621954725729,
      "learning_rate": 1.823224351229308e-05,
      "loss": 1.6226,
      "step": 693
    },
    {
      "epoch": 0.2006360219716681,
      "grad_norm": 1.2182097639154552,
      "learning_rate": 1.822703113141953e-05,
      "loss": 1.613,
      "step": 694
    },
    {
      "epoch": 0.2009251228678809,
      "grad_norm": 1.2785953541313604,
      "learning_rate": 1.82218118246599e-05,
      "loss": 1.7014,
      "step": 695
    },
    {
      "epoch": 0.20121422376409367,
      "grad_norm": 1.2338324878833378,
      "learning_rate": 1.821658559640804e-05,
      "loss": 1.6298,
      "step": 696
    },
    {
      "epoch": 0.20150332466030646,
      "grad_norm": 1.1768453406356614,
      "learning_rate": 1.821135245106363e-05,
      "loss": 1.5619,
      "step": 697
    },
    {
      "epoch": 0.20179242555651922,
      "grad_norm": 1.1771764186726315,
      "learning_rate": 1.820611239303216e-05,
      "loss": 1.68,
      "step": 698
    },
    {
      "epoch": 0.202081526452732,
      "grad_norm": 1.1751259370580003,
      "learning_rate": 1.8200865426724955e-05,
      "loss": 1.6353,
      "step": 699
    },
    {
      "epoch": 0.2023706273489448,
      "grad_norm": 1.234717870608629,
      "learning_rate": 1.819561155655914e-05,
      "loss": 1.6222,
      "step": 700
    }
  ],
  "logging_steps": 1,
  "max_steps": 3459,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 700,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 73282879488000.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}