{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9896907216494846,
  "eval_steps": 500,
  "global_step": 435,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006872852233676976,
      "grad_norm": 0.9486322019687681,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 1.3163,
      "step": 1
    },
    {
      "epoch": 0.013745704467353952,
      "grad_norm": 0.9581819297739079,
      "learning_rate": 9.090909090909091e-06,
      "loss": 1.3039,
      "step": 2
    },
    {
      "epoch": 0.020618556701030927,
      "grad_norm": 1.0054838528858523,
      "learning_rate": 1.3636363636363637e-05,
      "loss": 1.3912,
      "step": 3
    },
    {
      "epoch": 0.027491408934707903,
      "grad_norm": 0.973313664446148,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 1.3176,
      "step": 4
    },
    {
      "epoch": 0.03436426116838488,
      "grad_norm": 0.9112881771285355,
      "learning_rate": 2.272727272727273e-05,
      "loss": 1.2813,
      "step": 5
    },
    {
      "epoch": 0.041237113402061855,
      "grad_norm": 0.868853801792406,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 1.2778,
      "step": 6
    },
    {
      "epoch": 0.048109965635738834,
      "grad_norm": 0.7999023078891502,
      "learning_rate": 3.181818181818182e-05,
      "loss": 1.2184,
      "step": 7
    },
    {
      "epoch": 0.054982817869415807,
      "grad_norm": 0.6175826046713091,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 1.1459,
      "step": 8
    },
    {
      "epoch": 0.061855670103092786,
      "grad_norm": 0.5099365471572147,
      "learning_rate": 4.0909090909090915e-05,
      "loss": 1.0493,
      "step": 9
    },
    {
      "epoch": 0.06872852233676977,
      "grad_norm": 0.5444792063302974,
      "learning_rate": 4.545454545454546e-05,
      "loss": 1.0206,
      "step": 10
    },
    {
      "epoch": 0.07560137457044673,
      "grad_norm": 0.612097075909489,
      "learning_rate": 5e-05,
      "loss": 0.9556,
      "step": 11
    },
    {
      "epoch": 0.08247422680412371,
      "grad_norm": 0.5965134047415941,
      "learning_rate": 5.4545454545454546e-05,
      "loss": 0.8832,
      "step": 12
    },
    {
      "epoch": 0.08934707903780069,
      "grad_norm": 0.6321048435692244,
      "learning_rate": 5.90909090909091e-05,
      "loss": 0.8113,
      "step": 13
    },
    {
      "epoch": 0.09621993127147767,
      "grad_norm": 0.5433713534011196,
      "learning_rate": 6.363636363636364e-05,
      "loss": 0.7699,
      "step": 14
    },
    {
      "epoch": 0.10309278350515463,
      "grad_norm": 0.5641051016325467,
      "learning_rate": 6.818181818181818e-05,
      "loss": 0.6987,
      "step": 15
    },
    {
      "epoch": 0.10996563573883161,
      "grad_norm": 0.4741800529459741,
      "learning_rate": 7.272727272727273e-05,
      "loss": 0.6418,
      "step": 16
    },
    {
      "epoch": 0.11683848797250859,
      "grad_norm": 0.31919675330271313,
      "learning_rate": 7.727272727272727e-05,
      "loss": 0.5883,
      "step": 17
    },
    {
      "epoch": 0.12371134020618557,
      "grad_norm": 0.2704850432116006,
      "learning_rate": 8.181818181818183e-05,
      "loss": 0.576,
      "step": 18
    },
    {
      "epoch": 0.13058419243986255,
      "grad_norm": 0.26062797028012374,
      "learning_rate": 8.636363636363637e-05,
      "loss": 0.5267,
      "step": 19
    },
    {
      "epoch": 0.13745704467353953,
      "grad_norm": 0.18740805347521763,
      "learning_rate": 9.090909090909092e-05,
      "loss": 0.5469,
      "step": 20
    },
    {
      "epoch": 0.14432989690721648,
      "grad_norm": 0.22017175307824505,
      "learning_rate": 9.545454545454546e-05,
      "loss": 0.5226,
      "step": 21
    },
    {
      "epoch": 0.15120274914089346,
      "grad_norm": 0.1887039063728806,
      "learning_rate": 0.0001,
      "loss": 0.4873,
      "step": 22
    },
    {
      "epoch": 0.15807560137457044,
      "grad_norm": 0.1793474824943474,
      "learning_rate": 0.00010454545454545455,
      "loss": 0.4978,
      "step": 23
    },
    {
      "epoch": 0.16494845360824742,
      "grad_norm": 0.23697858977104094,
      "learning_rate": 0.00010909090909090909,
      "loss": 0.4814,
      "step": 24
    },
    {
      "epoch": 0.1718213058419244,
      "grad_norm": 0.17636901545890651,
      "learning_rate": 0.00011363636363636365,
      "loss": 0.5103,
      "step": 25
    },
    {
      "epoch": 0.17869415807560138,
      "grad_norm": 0.16142348168311232,
      "learning_rate": 0.0001181818181818182,
      "loss": 0.4773,
      "step": 26
    },
    {
      "epoch": 0.18556701030927836,
      "grad_norm": 0.1425385061824693,
      "learning_rate": 0.00012272727272727272,
      "loss": 0.492,
      "step": 27
    },
    {
      "epoch": 0.19243986254295534,
      "grad_norm": 0.1265888154640839,
      "learning_rate": 0.00012727272727272728,
      "loss": 0.4854,
      "step": 28
    },
    {
      "epoch": 0.19931271477663232,
      "grad_norm": 0.11449596495934233,
      "learning_rate": 0.0001318181818181818,
      "loss": 0.4615,
      "step": 29
    },
    {
      "epoch": 0.20618556701030927,
      "grad_norm": 0.12461604075605497,
      "learning_rate": 0.00013636363636363637,
      "loss": 0.4703,
      "step": 30
    },
    {
      "epoch": 0.21305841924398625,
      "grad_norm": 0.12169420338653658,
      "learning_rate": 0.00014090909090909093,
      "loss": 0.4433,
      "step": 31
    },
    {
      "epoch": 0.21993127147766323,
      "grad_norm": 0.10581158171701577,
      "learning_rate": 0.00014545454545454546,
      "loss": 0.4397,
      "step": 32
    },
    {
      "epoch": 0.2268041237113402,
      "grad_norm": 0.11163612722377497,
      "learning_rate": 0.00015000000000000001,
      "loss": 0.4605,
      "step": 33
    },
    {
      "epoch": 0.23367697594501718,
      "grad_norm": 0.11309459248751377,
      "learning_rate": 0.00015454545454545454,
      "loss": 0.4495,
      "step": 34
    },
    {
      "epoch": 0.24054982817869416,
      "grad_norm": 0.11367215299047806,
      "learning_rate": 0.0001590909090909091,
      "loss": 0.4308,
      "step": 35
    },
    {
      "epoch": 0.24742268041237114,
      "grad_norm": 0.10809464118167192,
      "learning_rate": 0.00016363636363636366,
      "loss": 0.4368,
      "step": 36
    },
    {
      "epoch": 0.2542955326460481,
      "grad_norm": 0.10502574406388546,
      "learning_rate": 0.0001681818181818182,
      "loss": 0.4196,
      "step": 37
    },
    {
      "epoch": 0.2611683848797251,
      "grad_norm": 0.10320182075757336,
      "learning_rate": 0.00017272727272727275,
      "loss": 0.42,
      "step": 38
    },
    {
      "epoch": 0.26804123711340205,
      "grad_norm": 0.11138418873828733,
      "learning_rate": 0.00017727272727272728,
      "loss": 0.4514,
      "step": 39
    },
    {
      "epoch": 0.27491408934707906,
      "grad_norm": 0.11111081000023773,
      "learning_rate": 0.00018181818181818183,
      "loss": 0.4162,
      "step": 40
    },
    {
      "epoch": 0.281786941580756,
      "grad_norm": 0.10773385531295475,
      "learning_rate": 0.00018636363636363636,
      "loss": 0.4058,
      "step": 41
    },
    {
      "epoch": 0.28865979381443296,
      "grad_norm": 0.10631032605354059,
      "learning_rate": 0.00019090909090909092,
      "loss": 0.4029,
      "step": 42
    },
    {
      "epoch": 0.29553264604810997,
      "grad_norm": 0.10836378456827221,
      "learning_rate": 0.00019545454545454548,
      "loss": 0.4096,
      "step": 43
    },
    {
      "epoch": 0.3024054982817869,
      "grad_norm": 0.10880423378093391,
      "learning_rate": 0.0002,
      "loss": 0.4196,
      "step": 44
    },
    {
      "epoch": 0.30927835051546393,
      "grad_norm": 0.1066915228080529,
      "learning_rate": 0.00019999677214588312,
      "loss": 0.4049,
      "step": 45
    },
    {
      "epoch": 0.3161512027491409,
      "grad_norm": 0.1009132049783852,
      "learning_rate": 0.00019998708879191335,
      "loss": 0.4015,
      "step": 46
    },
    {
      "epoch": 0.3230240549828179,
      "grad_norm": 0.10423644166095698,
      "learning_rate": 0.00019997095056321971,
      "loss": 0.3989,
      "step": 47
    },
    {
      "epoch": 0.32989690721649484,
      "grad_norm": 0.10222126638135792,
      "learning_rate": 0.00019994835850163924,
      "loss": 0.3911,
      "step": 48
    },
    {
      "epoch": 0.33676975945017185,
      "grad_norm": 0.09539411299428839,
      "learning_rate": 0.00019991931406564944,
      "loss": 0.3861,
      "step": 49
    },
    {
      "epoch": 0.3436426116838488,
      "grad_norm": 0.10040146458411044,
      "learning_rate": 0.00019988381913027442,
      "loss": 0.3932,
      "step": 50
    },
    {
      "epoch": 0.35051546391752575,
      "grad_norm": 0.10573179360993629,
      "learning_rate": 0.00019984187598696363,
      "loss": 0.3936,
      "step": 51
    },
    {
      "epoch": 0.35738831615120276,
      "grad_norm": 0.10564577769908695,
      "learning_rate": 0.00019979348734344398,
      "loss": 0.4071,
      "step": 52
    },
    {
      "epoch": 0.3642611683848797,
      "grad_norm": 0.10695511977009023,
      "learning_rate": 0.00019973865632354516,
      "loss": 0.3882,
      "step": 53
    },
    {
      "epoch": 0.3711340206185567,
      "grad_norm": 0.10064589165597464,
      "learning_rate": 0.0001996773864669978,
      "loss": 0.3785,
      "step": 54
    },
    {
      "epoch": 0.37800687285223367,
      "grad_norm": 0.09791457914129995,
      "learning_rate": 0.00019960968172920516,
      "loss": 0.3811,
      "step": 55
    },
    {
      "epoch": 0.3848797250859107,
      "grad_norm": 0.1118566682283669,
      "learning_rate": 0.00019953554648098748,
      "loss": 0.3938,
      "step": 56
    },
    {
      "epoch": 0.3917525773195876,
      "grad_norm": 0.10214824698553887,
      "learning_rate": 0.0001994549855083001,
      "loss": 0.3925,
      "step": 57
    },
    {
      "epoch": 0.39862542955326463,
      "grad_norm": 0.1014103487175917,
      "learning_rate": 0.0001993680040119244,
      "loss": 0.3802,
      "step": 58
    },
    {
      "epoch": 0.4054982817869416,
      "grad_norm": 0.0993499098682824,
      "learning_rate": 0.00019927460760713197,
      "loss": 0.381,
      "step": 59
    },
    {
      "epoch": 0.41237113402061853,
      "grad_norm": 0.10206073076536111,
      "learning_rate": 0.00019917480232332224,
      "loss": 0.3657,
      "step": 60
    },
    {
      "epoch": 0.41924398625429554,
      "grad_norm": 0.09828556302343387,
      "learning_rate": 0.00019906859460363307,
      "loss": 0.3761,
      "step": 61
    },
    {
      "epoch": 0.4261168384879725,
      "grad_norm": 0.10139688359355604,
      "learning_rate": 0.00019895599130452505,
      "loss": 0.3749,
      "step": 62
    },
    {
      "epoch": 0.4329896907216495,
      "grad_norm": 0.10711125411772042,
      "learning_rate": 0.0001988369996953386,
      "loss": 0.3706,
      "step": 63
    },
    {
      "epoch": 0.43986254295532645,
      "grad_norm": 0.09724017597982272,
      "learning_rate": 0.00019871162745782478,
      "loss": 0.3705,
      "step": 64
    },
    {
      "epoch": 0.44673539518900346,
      "grad_norm": 0.1012279008931452,
      "learning_rate": 0.00019857988268564953,
      "loss": 0.3712,
      "step": 65
    },
    {
      "epoch": 0.4536082474226804,
      "grad_norm": 0.09806011793458508,
      "learning_rate": 0.0001984417738838709,
      "loss": 0.355,
      "step": 66
    },
    {
      "epoch": 0.46048109965635736,
      "grad_norm": 0.10463912089089605,
      "learning_rate": 0.0001982973099683902,
      "loss": 0.3812,
      "step": 67
    },
    {
      "epoch": 0.46735395189003437,
      "grad_norm": 0.10616651630870087,
      "learning_rate": 0.0001981465002653763,
      "loss": 0.3671,
      "step": 68
    },
    {
      "epoch": 0.4742268041237113,
      "grad_norm": 0.11658375430444518,
      "learning_rate": 0.00019798935451066361,
      "loss": 0.3822,
      "step": 69
    },
    {
      "epoch": 0.48109965635738833,
      "grad_norm": 0.10686011983880458,
      "learning_rate": 0.0001978258828491236,
      "loss": 0.3564,
      "step": 70
    },
    {
      "epoch": 0.4879725085910653,
      "grad_norm": 0.10160970818733071,
      "learning_rate": 0.00019765609583400977,
      "loss": 0.3656,
      "step": 71
    },
    {
      "epoch": 0.4948453608247423,
      "grad_norm": 0.10845572916815471,
      "learning_rate": 0.0001974800044262764,
      "loss": 0.3623,
      "step": 72
    },
    {
      "epoch": 0.5017182130584192,
      "grad_norm": 0.10898867605088586,
      "learning_rate": 0.00019729761999387103,
      "loss": 0.3689,
      "step": 73
    },
    {
      "epoch": 0.5085910652920962,
      "grad_norm": 0.10030807679917116,
      "learning_rate": 0.00019710895431100046,
      "loss": 0.3743,
      "step": 74
    },
    {
      "epoch": 0.5154639175257731,
      "grad_norm": 0.10640353541210322,
      "learning_rate": 0.00019691401955737072,
      "loss": 0.3664,
      "step": 75
    },
    {
      "epoch": 0.5223367697594502,
      "grad_norm": 0.10268077724904813,
      "learning_rate": 0.00019671282831740076,
      "loss": 0.3446,
      "step": 76
    },
    {
      "epoch": 0.5292096219931272,
      "grad_norm": 0.1026968446802754,
      "learning_rate": 0.00019650539357941003,
      "loss": 0.3524,
      "step": 77
    },
    {
      "epoch": 0.5360824742268041,
      "grad_norm": 0.10265145337121352,
      "learning_rate": 0.00019629172873477995,
      "loss": 0.3615,
      "step": 78
    },
    {
      "epoch": 0.5429553264604811,
      "grad_norm": 0.10234137978544197,
      "learning_rate": 0.00019607184757708951,
      "loss": 0.348,
      "step": 79
    },
    {
      "epoch": 0.5498281786941581,
      "grad_norm": 0.10414003831964998,
      "learning_rate": 0.00019584576430122473,
      "loss": 0.3455,
      "step": 80
    },
    {
      "epoch": 0.5567010309278351,
      "grad_norm": 0.10371223386491027,
      "learning_rate": 0.00019561349350246226,
      "loss": 0.3378,
      "step": 81
    },
    {
      "epoch": 0.563573883161512,
      "grad_norm": 0.10200261115662425,
      "learning_rate": 0.00019537505017552716,
      "loss": 0.3424,
      "step": 82
    },
    {
      "epoch": 0.570446735395189,
      "grad_norm": 0.10679690224389146,
      "learning_rate": 0.00019513044971362494,
      "loss": 0.3381,
      "step": 83
    },
    {
      "epoch": 0.5773195876288659,
      "grad_norm": 0.11747209193015294,
      "learning_rate": 0.00019487970790744774,
      "loss": 0.3608,
      "step": 84
    },
    {
      "epoch": 0.584192439862543,
      "grad_norm": 0.10455385108750616,
      "learning_rate": 0.000194622840944155,
      "loss": 0.3513,
      "step": 85
    },
    {
      "epoch": 0.5910652920962199,
      "grad_norm": 0.11103687968369377,
      "learning_rate": 0.00019435986540632843,
      "loss": 0.3322,
      "step": 86
    },
    {
      "epoch": 0.5979381443298969,
      "grad_norm": 0.1044628007637211,
      "learning_rate": 0.00019409079827090145,
      "loss": 0.3452,
      "step": 87
    },
    {
      "epoch": 0.6048109965635738,
      "grad_norm": 0.1062512518470538,
      "learning_rate": 0.00019381565690806328,
      "loss": 0.3295,
      "step": 88
    },
    {
      "epoch": 0.6116838487972509,
      "grad_norm": 0.10737195453497204,
      "learning_rate": 0.00019353445908013755,
      "loss": 0.3322,
      "step": 89
    },
    {
      "epoch": 0.6185567010309279,
      "grad_norm": 0.10702988631259788,
      "learning_rate": 0.00019324722294043558,
      "loss": 0.3436,
      "step": 90
    },
    {
      "epoch": 0.6254295532646048,
      "grad_norm": 0.10765269601948493,
      "learning_rate": 0.00019295396703208453,
      "loss": 0.3628,
      "step": 91
    },
    {
      "epoch": 0.6323024054982818,
      "grad_norm": 0.1002631772076248,
      "learning_rate": 0.00019265471028683014,
      "loss": 0.337,
      "step": 92
    },
    {
      "epoch": 0.6391752577319587,
      "grad_norm": 0.10607703170572558,
      "learning_rate": 0.00019234947202381486,
      "loss": 0.3312,
      "step": 93
    },
    {
      "epoch": 0.6460481099656358,
      "grad_norm": 0.10229636934864904,
      "learning_rate": 0.00019203827194833026,
      "loss": 0.3613,
      "step": 94
    },
    {
      "epoch": 0.6529209621993127,
      "grad_norm": 0.10916337452983911,
      "learning_rate": 0.00019172113015054532,
      "loss": 0.3402,
      "step": 95
    },
    {
      "epoch": 0.6597938144329897,
      "grad_norm": 0.10094159423939311,
      "learning_rate": 0.00019139806710420914,
      "loss": 0.3308,
      "step": 96
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.10420707581199606,
      "learning_rate": 0.00019106910366532942,
      "loss": 0.3297,
      "step": 97
    },
    {
      "epoch": 0.6735395189003437,
      "grad_norm": 0.11033214518450266,
      "learning_rate": 0.000190734261070826,
      "loss": 0.3368,
      "step": 98
    },
    {
      "epoch": 0.6804123711340206,
      "grad_norm": 0.10648307368306249,
      "learning_rate": 0.00019039356093715975,
      "loss": 0.3436,
      "step": 99
    },
    {
      "epoch": 0.6872852233676976,
      "grad_norm": 0.1023047486718038,
      "learning_rate": 0.00019004702525893732,
      "loss": 0.3389,
      "step": 100
    },
    {
      "epoch": 0.6941580756013745,
      "grad_norm": 0.10884134971992637,
      "learning_rate": 0.000189694676407491,
      "loss": 0.3343,
      "step": 101
    },
    {
      "epoch": 0.7010309278350515,
      "grad_norm": 0.10624545430750702,
      "learning_rate": 0.0001893365371294346,
      "loss": 0.334,
      "step": 102
    },
    {
      "epoch": 0.7079037800687286,
      "grad_norm": 0.10418100794965049,
      "learning_rate": 0.00018897263054519498,
      "loss": 0.3332,
      "step": 103
    },
    {
      "epoch": 0.7147766323024055,
      "grad_norm": 0.10442261089390925,
      "learning_rate": 0.00018860298014751944,
      "loss": 0.3231,
      "step": 104
    },
    {
      "epoch": 0.7216494845360825,
      "grad_norm": 0.11510206196708461,
      "learning_rate": 0.0001882276097999592,
      "loss": 0.3331,
      "step": 105
    },
    {
      "epoch": 0.7285223367697594,
      "grad_norm": 0.10447026389489032,
      "learning_rate": 0.00018784654373532866,
      "loss": 0.3454,
      "step": 106
    },
    {
      "epoch": 0.7353951890034365,
      "grad_norm": 0.10590015702307645,
      "learning_rate": 0.00018745980655414114,
      "loss": 0.3283,
      "step": 107
    },
    {
      "epoch": 0.7422680412371134,
      "grad_norm": 0.1019688644295514,
      "learning_rate": 0.00018706742322302064,
      "loss": 0.3368,
      "step": 108
    },
    {
      "epoch": 0.7491408934707904,
      "grad_norm": 0.11013306581811545,
      "learning_rate": 0.00018666941907309026,
      "loss": 0.3371,
      "step": 109
    },
    {
      "epoch": 0.7560137457044673,
      "grad_norm": 0.09950125567962896,
      "learning_rate": 0.0001862658197983366,
      "loss": 0.3206,
      "step": 110
    },
    {
      "epoch": 0.7628865979381443,
      "grad_norm": 0.09961853222606298,
      "learning_rate": 0.0001858566514539513,
      "loss": 0.3245,
      "step": 111
    },
    {
      "epoch": 0.7697594501718213,
      "grad_norm": 0.10822873754218582,
      "learning_rate": 0.00018544194045464886,
      "loss": 0.336,
      "step": 112
    },
    {
      "epoch": 0.7766323024054983,
      "grad_norm": 0.10256746162360054,
      "learning_rate": 0.00018502171357296144,
      "loss": 0.3279,
      "step": 113
    },
    {
      "epoch": 0.7835051546391752,
      "grad_norm": 0.09750927062180448,
      "learning_rate": 0.0001845959979375104,
      "loss": 0.3388,
      "step": 114
    },
    {
      "epoch": 0.7903780068728522,
      "grad_norm": 0.09793726368683904,
      "learning_rate": 0.00018416482103125506,
      "loss": 0.3303,
      "step": 115
    },
    {
      "epoch": 0.7972508591065293,
      "grad_norm": 0.10598098121402388,
      "learning_rate": 0.0001837282106897185,
      "loss": 0.3216,
      "step": 116
    },
    {
      "epoch": 0.8041237113402062,
      "grad_norm": 0.10184730892936761,
      "learning_rate": 0.00018328619509919044,
      "loss": 0.3341,
      "step": 117
    },
    {
      "epoch": 0.8109965635738832,
      "grad_norm": 0.10057188453324177,
      "learning_rate": 0.0001828388027949078,
      "loss": 0.3377,
      "step": 118
    },
    {
      "epoch": 0.8178694158075601,
      "grad_norm": 0.10702981460753135,
      "learning_rate": 0.00018238606265921238,
      "loss": 0.3408,
      "step": 119
    },
    {
      "epoch": 0.8247422680412371,
      "grad_norm": 0.10861206391221787,
      "learning_rate": 0.00018192800391968642,
      "loss": 0.3413,
      "step": 120
    },
    {
      "epoch": 0.8316151202749141,
      "grad_norm": 0.10665429392095056,
      "learning_rate": 0.00018146465614726567,
      "loss": 0.3234,
      "step": 121
    },
    {
      "epoch": 0.8384879725085911,
      "grad_norm": 0.10180559686021746,
      "learning_rate": 0.00018099604925433043,
      "loss": 0.3263,
      "step": 122
    },
    {
      "epoch": 0.845360824742268,
      "grad_norm": 0.11025637152512564,
      "learning_rate": 0.00018052221349277442,
      "loss": 0.3295,
      "step": 123
    },
    {
      "epoch": 0.852233676975945,
      "grad_norm": 0.10266709615387361,
      "learning_rate": 0.00018004317945205197,
      "loss": 0.3275,
      "step": 124
    },
    {
      "epoch": 0.8591065292096219,
      "grad_norm": 0.10117796525027944,
      "learning_rate": 0.0001795589780572031,
      "loss": 0.3541,
      "step": 125
    },
    {
      "epoch": 0.865979381443299,
      "grad_norm": 0.10526178050004703,
      "learning_rate": 0.00017906964056685706,
      "loss": 0.3312,
      "step": 126
    },
    {
      "epoch": 0.872852233676976,
      "grad_norm": 0.1063992770321544,
      "learning_rate": 0.00017857519857121458,
      "loss": 0.3312,
      "step": 127
    },
    {
      "epoch": 0.8797250859106529,
      "grad_norm": 0.10996856445056626,
      "learning_rate": 0.00017807568399000822,
      "loss": 0.3527,
      "step": 128
    },
    {
      "epoch": 0.8865979381443299,
      "grad_norm": 0.09753881684843836,
      "learning_rate": 0.000177571129070442,
      "loss": 0.3117,
      "step": 129
    },
    {
      "epoch": 0.8934707903780069,
      "grad_norm": 0.10008693889485847,
      "learning_rate": 0.0001770615663851093,
      "loss": 0.3188,
      "step": 130
    },
    {
      "epoch": 0.9003436426116839,
      "grad_norm": 0.10344071862676044,
      "learning_rate": 0.0001765470288298905,
      "loss": 0.3252,
      "step": 131
    },
    {
      "epoch": 0.9072164948453608,
      "grad_norm": 0.09310502474075653,
      "learning_rate": 0.0001760275496218288,
      "loss": 0.3058,
      "step": 132
    },
    {
      "epoch": 0.9140893470790378,
      "grad_norm": 0.10039282316295665,
      "learning_rate": 0.0001755031622969862,
      "loss": 0.3468,
      "step": 133
    },
    {
      "epoch": 0.9209621993127147,
      "grad_norm": 0.09871302822261965,
      "learning_rate": 0.00017497390070827848,
      "loss": 0.3241,
      "step": 134
    },
    {
      "epoch": 0.9278350515463918,
      "grad_norm": 0.10019109474446979,
      "learning_rate": 0.00017443979902328956,
      "loss": 0.3247,
      "step": 135
    },
    {
      "epoch": 0.9347079037800687,
      "grad_norm": 0.09742365798894928,
      "learning_rate": 0.00017390089172206592,
      "loss": 0.3268,
      "step": 136
    },
    {
      "epoch": 0.9415807560137457,
      "grad_norm": 0.09613939400204266,
      "learning_rate": 0.00017335721359489057,
      "loss": 0.3017,
      "step": 137
    },
    {
      "epoch": 0.9484536082474226,
      "grad_norm": 0.11015286314690642,
      "learning_rate": 0.00017280879974003707,
      "loss": 0.3369,
      "step": 138
    },
    {
      "epoch": 0.9553264604810997,
      "grad_norm": 0.09848269873542065,
      "learning_rate": 0.0001722556855615039,
      "loss": 0.3215,
      "step": 139
    },
    {
      "epoch": 0.9621993127147767,
      "grad_norm": 0.10356185439767568,
      "learning_rate": 0.00017169790676672858,
      "loss": 0.3209,
      "step": 140
    },
    {
      "epoch": 0.9690721649484536,
      "grad_norm": 0.10200587363931662,
      "learning_rate": 0.0001711354993642827,
      "loss": 0.3156,
      "step": 141
    },
    {
      "epoch": 0.9759450171821306,
      "grad_norm": 0.10631936681232042,
      "learning_rate": 0.0001705684996615472,
      "loss": 0.3139,
      "step": 142
    },
    {
      "epoch": 0.9828178694158075,
      "grad_norm": 0.09891246233686207,
      "learning_rate": 0.0001699969442623686,
      "loss": 0.3252,
      "step": 143
    },
    {
      "epoch": 0.9896907216494846,
      "grad_norm": 0.09631426655006751,
      "learning_rate": 0.00016942087006469592,
      "loss": 0.3092,
      "step": 144
    },
    {
      "epoch": 0.9965635738831615,
      "grad_norm": 0.1007256123012449,
      "learning_rate": 0.00016884031425819853,
      "loss": 0.3214,
      "step": 145
    },
    {
      "epoch": 0.9965635738831615,
      "eval_loss": 0.31896448135375977,
      "eval_runtime": 31.3575,
      "eval_samples_per_second": 31.157,
      "eval_steps_per_second": 0.989,
      "step": 145
    },
    {
      "epoch": 1.0034364261168385,
      "grad_norm": 0.09742513564214555,
      "learning_rate": 0.00016825531432186543,
      "loss": 0.3131,
      "step": 146
    },
    {
      "epoch": 1.0103092783505154,
      "grad_norm": 0.1000843163672815,
      "learning_rate": 0.00016766590802158566,
      "loss": 0.308,
      "step": 147
    },
    {
      "epoch": 1.0171821305841924,
      "grad_norm": 0.1018182338551042,
      "learning_rate": 0.0001670721334077103,
      "loss": 0.3077,
      "step": 148
    },
    {
      "epoch": 1.0240549828178693,
      "grad_norm": 0.10928234134401811,
      "learning_rate": 0.00016647402881259598,
      "loss": 0.3137,
      "step": 149
    },
    {
      "epoch": 1.0309278350515463,
      "grad_norm": 0.10482761572794695,
      "learning_rate": 0.00016587163284813032,
      "loss": 0.2969,
      "step": 150
    },
    {
      "epoch": 1.0378006872852235,
      "grad_norm": 0.10755010011949057,
      "learning_rate": 0.00016526498440323914,
      "loss": 0.3027,
      "step": 151
    },
    {
      "epoch": 1.0446735395189004,
      "grad_norm": 0.09961868507020075,
      "learning_rate": 0.0001646541226413761,
      "loss": 0.2918,
      "step": 152
    },
    {
      "epoch": 1.0515463917525774,
      "grad_norm": 0.10275553359018949,
      "learning_rate": 0.00016403908699799425,
      "loss": 0.2935,
      "step": 153
    },
    {
      "epoch": 1.0584192439862543,
      "grad_norm": 0.10314122954609171,
      "learning_rate": 0.00016341991717800023,
      "loss": 0.3107,
      "step": 154
    },
    {
      "epoch": 1.0652920962199313,
      "grad_norm": 0.10201304020136209,
      "learning_rate": 0.00016279665315319114,
      "loss": 0.2925,
      "step": 155
    },
    {
      "epoch": 1.0721649484536082,
      "grad_norm": 0.11295490175270977,
      "learning_rate": 0.0001621693351596739,
      "loss": 0.3165,
      "step": 156
    },
    {
      "epoch": 1.0790378006872852,
      "grad_norm": 0.11187236691946645,
      "learning_rate": 0.00016153800369526788,
      "loss": 0.3008,
      "step": 157
    },
    {
      "epoch": 1.0859106529209621,
      "grad_norm": 0.10649522892041895,
      "learning_rate": 0.0001609026995168904,
      "loss": 0.3072,
      "step": 158
    },
    {
      "epoch": 1.0927835051546393,
      "grad_norm": 0.10138792178292377,
      "learning_rate": 0.00016026346363792567,
      "loss": 0.2968,
      "step": 159
    },
    {
      "epoch": 1.0996563573883162,
      "grad_norm": 0.10490079005967201,
      "learning_rate": 0.00015962033732557686,
      "loss": 0.3074,
      "step": 160
    },
    {
      "epoch": 1.1065292096219932,
      "grad_norm": 0.10339230680715779,
      "learning_rate": 0.00015897336209820239,
      "loss": 0.2904,
      "step": 161
    },
    {
      "epoch": 1.1134020618556701,
      "grad_norm": 0.11480233498020057,
      "learning_rate": 0.00015832257972263523,
      "loss": 0.31,
      "step": 162
    },
    {
      "epoch": 1.120274914089347,
      "grad_norm": 0.10620125747115913,
      "learning_rate": 0.00015766803221148673,
      "loss": 0.3041,
      "step": 163
    },
    {
      "epoch": 1.127147766323024,
      "grad_norm": 0.10787426446881714,
      "learning_rate": 0.0001570097618204345,
      "loss": 0.3142,
      "step": 164
    },
    {
      "epoch": 1.134020618556701,
      "grad_norm": 0.10375711742292576,
      "learning_rate": 0.00015634781104549442,
      "loss": 0.2916,
      "step": 165
    },
    {
      "epoch": 1.140893470790378,
      "grad_norm": 0.1085251854270091,
      "learning_rate": 0.00015568222262027717,
      "loss": 0.2969,
      "step": 166
    },
    {
      "epoch": 1.147766323024055,
      "grad_norm": 0.10620158303898176,
      "learning_rate": 0.00015501303951322943,
      "loss": 0.3069,
      "step": 167
    },
    {
      "epoch": 1.1546391752577319,
      "grad_norm": 0.10221091343876472,
      "learning_rate": 0.00015434030492486023,
      "loss": 0.3015,
      "step": 168
    },
    {
      "epoch": 1.161512027491409,
      "grad_norm": 0.11025720771873336,
      "learning_rate": 0.00015366406228495172,
      "loss": 0.2923,
      "step": 169
    },
    {
      "epoch": 1.168384879725086,
      "grad_norm": 0.11019089511672239,
      "learning_rate": 0.00015298435524975572,
      "loss": 0.3088,
      "step": 170
    },
    {
      "epoch": 1.175257731958763,
      "grad_norm": 0.1066893326930603,
      "learning_rate": 0.00015230122769917527,
      "loss": 0.3006,
      "step": 171
    },
    {
      "epoch": 1.1821305841924399,
      "grad_norm": 0.11028825839961288,
      "learning_rate": 0.00015161472373393186,
      "loss": 0.3028,
      "step": 172
    },
    {
      "epoch": 1.1890034364261168,
      "grad_norm": 0.10136590795578818,
      "learning_rate": 0.00015092488767271857,
      "loss": 0.2789,
      "step": 173
    },
    {
      "epoch": 1.1958762886597938,
      "grad_norm": 0.1032678246834342,
      "learning_rate": 0.00015023176404933874,
      "loss": 0.2975,
      "step": 174
    },
    {
      "epoch": 1.2027491408934707,
      "grad_norm": 0.11056648590020815,
      "learning_rate": 0.00014953539760983122,
      "loss": 0.3013,
      "step": 175
    },
    {
      "epoch": 1.2096219931271477,
      "grad_norm": 0.09978579785299047,
      "learning_rate": 0.0001488358333095816,
      "loss": 0.2955,
      "step": 176
    },
    {
      "epoch": 1.2164948453608249,
      "grad_norm": 0.10699681452783022,
      "learning_rate": 0.00014813311631041995,
      "loss": 0.2993,
      "step": 177
    },
    {
      "epoch": 1.2233676975945018,
      "grad_norm": 0.10408363973129214,
      "learning_rate": 0.00014742729197770552,
      "loss": 0.2954,
      "step": 178
    },
    {
      "epoch": 1.2302405498281788,
      "grad_norm": 0.1016249945819849,
      "learning_rate": 0.00014671840587739783,
      "loss": 0.2904,
      "step": 179
    },
    {
      "epoch": 1.2371134020618557,
      "grad_norm": 0.10519072387735487,
      "learning_rate": 0.00014600650377311522,
      "loss": 0.2916,
      "step": 180
    },
    {
      "epoch": 1.2439862542955327,
      "grad_norm": 0.1058597383518849,
      "learning_rate": 0.0001452916316231805,
      "loss": 0.301,
      "step": 181
    },
    {
      "epoch": 1.2508591065292096,
      "grad_norm": 0.10402401073301179,
      "learning_rate": 0.00014457383557765386,
      "loss": 0.2875,
      "step": 182
    },
    {
      "epoch": 1.2577319587628866,
      "grad_norm": 0.11052274366476848,
      "learning_rate": 0.00014385316197535372,
      "loss": 0.3033,
      "step": 183
    },
    {
      "epoch": 1.2646048109965635,
      "grad_norm": 0.11220133909982184,
      "learning_rate": 0.00014312965734086518,
      "loss": 0.3048,
      "step": 184
    },
    {
      "epoch": 1.2714776632302405,
      "grad_norm": 0.10111031103821645,
      "learning_rate": 0.0001424033683815365,
      "loss": 0.2883,
      "step": 185
    },
    {
      "epoch": 1.2783505154639174,
      "grad_norm": 0.108844384730101,
      "learning_rate": 0.00014167434198446383,
      "loss": 0.3156,
      "step": 186
    },
    {
      "epoch": 1.2852233676975944,
      "grad_norm": 0.11772254803070373,
      "learning_rate": 0.00014094262521346427,
      "loss": 0.2894,
      "step": 187
    },
    {
      "epoch": 1.2920962199312716,
      "grad_norm": 0.10084875638276648,
      "learning_rate": 0.00014020826530603776,
      "loss": 0.2941,
      "step": 188
    },
    {
      "epoch": 1.2989690721649485,
      "grad_norm": 0.10560762521845199,
      "learning_rate": 0.00013947130967031717,
      "loss": 0.2974,
      "step": 189
    },
    {
      "epoch": 1.3058419243986255,
      "grad_norm": 0.1112230814008033,
      "learning_rate": 0.00013873180588200827,
      "loss": 0.3051,
      "step": 190
    },
    {
      "epoch": 1.3127147766323024,
      "grad_norm": 0.10899405854558633,
      "learning_rate": 0.00013798980168131794,
      "loss": 0.3167,
      "step": 191
    },
    {
      "epoch": 1.3195876288659794,
      "grad_norm": 0.10459030666604682,
      "learning_rate": 0.00013724534496987247,
      "loss": 0.2944,
      "step": 192
    },
    {
      "epoch": 1.3264604810996563,
      "grad_norm": 0.10751960916175171,
      "learning_rate": 0.00013649848380762513,
      "loss": 0.2749,
      "step": 193
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.10402996897559527,
      "learning_rate": 0.0001357492664097534,
      "loss": 0.2972,
      "step": 194
    },
    {
      "epoch": 1.3402061855670104,
      "grad_norm": 0.10280900537193306,
      "learning_rate": 0.00013499774114354655,
      "loss": 0.3093,
      "step": 195
    },
    {
      "epoch": 1.3470790378006874,
      "grad_norm": 0.10262166065058195,
      "learning_rate": 0.0001342439565252831,
      "loss": 0.3023,
      "step": 196
    },
    {
      "epoch": 1.3539518900343643,
      "grad_norm": 0.11035322484519891,
      "learning_rate": 0.00013348796121709862,
      "loss": 0.283,
      "step": 197
    },
    {
      "epoch": 1.3608247422680413,
      "grad_norm": 0.10183546777840774,
      "learning_rate": 0.0001327298040238446,
      "loss": 0.2895,
      "step": 198
    },
    {
      "epoch": 1.3676975945017182,
      "grad_norm": 0.09846452794645016,
      "learning_rate": 0.00013196953388993726,
      "loss": 0.2775,
      "step": 199
    },
    {
      "epoch": 1.3745704467353952,
      "grad_norm": 0.10082546961742879,
      "learning_rate": 0.00013120719989619833,
      "loss": 0.2927,
      "step": 200
    },
    {
      "epoch": 1.3814432989690721,
      "grad_norm": 0.10338270125162533,
      "learning_rate": 0.00013044285125668614,
      "loss": 0.3022,
      "step": 201
    },
    {
      "epoch": 1.388316151202749,
      "grad_norm": 0.10489686085507173,
      "learning_rate": 0.0001296765373155188,
      "loss": 0.2886,
      "step": 202
    },
    {
      "epoch": 1.395189003436426,
      "grad_norm": 0.10826965142841156,
      "learning_rate": 0.00012890830754368855,
      "loss": 0.2839,
      "step": 203
    },
    {
      "epoch": 1.402061855670103,
      "grad_norm": 0.1095657642687174,
      "learning_rate": 0.0001281382115358679,
      "loss": 0.2974,
      "step": 204
    },
    {
      "epoch": 1.40893470790378,
      "grad_norm": 0.10290443785837475,
      "learning_rate": 0.0001273662990072083,
      "loss": 0.2916,
      "step": 205
    },
    {
      "epoch": 1.4158075601374571,
      "grad_norm": 0.10488572942462458,
      "learning_rate": 0.00012659261979013043,
      "loss": 0.2896,
      "step": 206
    },
    {
      "epoch": 1.422680412371134,
      "grad_norm": 0.0991511111540952,
      "learning_rate": 0.00012581722383110718,
      "loss": 0.2815,
      "step": 207
    },
    {
      "epoch": 1.429553264604811,
      "grad_norm": 0.11084616014817153,
      "learning_rate": 0.00012504016118743935,
      "loss": 0.2857,
      "step": 208
    },
    {
      "epoch": 1.436426116838488,
      "grad_norm": 0.10755359121093663,
      "learning_rate": 0.00012426148202402404,
      "loss": 0.2784,
      "step": 209
    },
    {
      "epoch": 1.443298969072165,
      "grad_norm": 0.10341376660828379,
      "learning_rate": 0.00012348123661011601,
      "loss": 0.2947,
      "step": 210
    },
    {
      "epoch": 1.4501718213058419,
      "grad_norm": 0.10964970839444545,
      "learning_rate": 0.00012269947531608276,
      "loss": 0.3056,
      "step": 211
    },
    {
      "epoch": 1.4570446735395188,
      "grad_norm": 0.10241977399698814,
      "learning_rate": 0.00012191624861015254,
      "loss": 0.3102,
      "step": 212
    },
    {
      "epoch": 1.463917525773196,
      "grad_norm": 0.10337164620385511,
      "learning_rate": 0.00012113160705515625,
      "loss": 0.3122,
      "step": 213
    },
    {
      "epoch": 1.470790378006873,
      "grad_norm": 0.10002116602053045,
      "learning_rate": 0.0001203456013052634,
      "loss": 0.2905,
      "step": 214
    },
    {
      "epoch": 1.47766323024055,
      "grad_norm": 0.1005828345016487,
      "learning_rate": 0.00011955828210271187,
      "loss": 0.2897,
      "step": 215
    },
    {
      "epoch": 1.4845360824742269,
      "grad_norm": 0.10315190703705018,
      "learning_rate": 0.00011876970027453222,
      "loss": 0.2896,
      "step": 216
    },
    {
      "epoch": 1.4914089347079038,
      "grad_norm": 0.10209266642811851,
      "learning_rate": 0.00011797990672926652,
      "loss": 0.2862,
      "step": 217
    },
    {
      "epoch": 1.4982817869415808,
      "grad_norm": 0.10350382211989749,
      "learning_rate": 0.00011718895245368167,
      "loss": 0.2993,
      "step": 218
    },
    {
      "epoch": 1.5051546391752577,
      "grad_norm": 0.09959069681120952,
      "learning_rate": 0.00011639688850947799,
      "loss": 0.279,
      "step": 219
    },
    {
      "epoch": 1.5120274914089347,
      "grad_norm": 0.1017699606831777,
      "learning_rate": 0.00011560376602999272,
      "loss": 0.2924,
      "step": 220
    },
    {
      "epoch": 1.5189003436426116,
      "grad_norm": 0.10199010748060407,
      "learning_rate": 0.00011480963621689905,
      "loss": 0.3007,
      "step": 221
    },
    {
      "epoch": 1.5257731958762886,
      "grad_norm": 0.10438569087355797,
      "learning_rate": 0.00011401455033690076,
      "loss": 0.2828,
      "step": 222
    },
    {
      "epoch": 1.5326460481099655,
      "grad_norm": 0.106590652554821,
      "learning_rate": 0.00011321855971842243,
      "loss": 0.299,
      "step": 223
    },
    {
      "epoch": 1.5395189003436425,
      "grad_norm": 0.10067806688032917,
      "learning_rate": 0.00011242171574829599,
      "loss": 0.2897,
      "step": 224
    },
    {
      "epoch": 1.5463917525773194,
      "grad_norm": 0.11078489209589217,
      "learning_rate": 0.00011162406986844323,
      "loss": 0.2791,
      "step": 225
    },
    {
      "epoch": 1.5532646048109966,
      "grad_norm": 0.10748644329087202,
      "learning_rate": 0.00011082567357255484,
      "loss": 0.3004,
      "step": 226
    },
    {
      "epoch": 1.5601374570446735,
      "grad_norm": 0.09232201373607851,
      "learning_rate": 0.00011002657840276627,
      "loss": 0.2647,
      "step": 227
    },
    {
      "epoch": 1.5670103092783505,
      "grad_norm": 0.10419820524514652,
      "learning_rate": 0.00010922683594633021,
      "loss": 0.2897,
      "step": 228
    },
    {
      "epoch": 1.5738831615120275,
      "grad_norm": 0.11387366232726741,
      "learning_rate": 0.00010842649783228624,
      "loss": 0.3071,
      "step": 229
    },
    {
      "epoch": 1.5807560137457046,
      "grad_norm": 0.1112209148078402,
      "learning_rate": 0.00010762561572812788,
      "loss": 0.2791,
      "step": 230
    },
    {
      "epoch": 1.5876288659793816,
      "grad_norm": 0.1027329268180551,
      "learning_rate": 0.0001068242413364671,
      "loss": 0.2962,
      "step": 231
    },
    {
      "epoch": 1.5945017182130585,
      "grad_norm": 0.10377377307338428,
      "learning_rate": 0.00010602242639169648,
      "loss": 0.2905,
      "step": 232
    },
    {
      "epoch": 1.6013745704467355,
      "grad_norm": 0.10333591660725541,
      "learning_rate": 0.0001052202226566494,
      "loss": 0.294,
      "step": 233
    },
    {
      "epoch": 1.6082474226804124,
      "grad_norm": 0.10653830908113968,
      "learning_rate": 0.00010441768191925847,
      "loss": 0.3023,
      "step": 234
    },
    {
      "epoch": 1.6151202749140894,
      "grad_norm": 0.10203788732177198,
      "learning_rate": 0.00010361485598921212,
      "loss": 0.2961,
      "step": 235
    },
    {
      "epoch": 1.6219931271477663,
      "grad_norm": 0.105467450842672,
      "learning_rate": 0.00010281179669461005,
      "loss": 0.2992,
      "step": 236
    },
    {
      "epoch": 1.6288659793814433,
      "grad_norm": 0.10232600415297333,
      "learning_rate": 0.00010200855587861724,
      "loss": 0.2894,
      "step": 237
    },
    {
      "epoch": 1.6357388316151202,
      "grad_norm": 0.1000078228085237,
      "learning_rate": 0.0001012051853961172,
      "loss": 0.2924,
      "step": 238
    },
    {
      "epoch": 1.6426116838487972,
      "grad_norm": 0.10443130272760445,
      "learning_rate": 0.00010040173711036431,
      "loss": 0.2987,
      "step": 239
    },
    {
      "epoch": 1.6494845360824741,
      "grad_norm": 0.10490297286528327,
      "learning_rate": 9.959826288963571e-05,
      "loss": 0.2859,
      "step": 240
    },
    {
      "epoch": 1.656357388316151,
      "grad_norm": 0.10350833328466466,
      "learning_rate": 9.879481460388282e-05,
      "loss": 0.2931,
      "step": 241
    },
    {
      "epoch": 1.663230240549828,
      "grad_norm": 0.10225492666499335,
      "learning_rate": 9.799144412138275e-05,
      "loss": 0.2878,
      "step": 242
    },
    {
      "epoch": 1.670103092783505,
      "grad_norm": 0.10508397146202163,
      "learning_rate": 9.718820330538998e-05,
      "loss": 0.3004,
      "step": 243
    },
    {
      "epoch": 1.6769759450171822,
      "grad_norm": 0.10283137687490299,
      "learning_rate": 9.638514401078788e-05,
      "loss": 0.2714,
      "step": 244
    },
    {
      "epoch": 1.6838487972508591,
      "grad_norm": 0.10397776989480932,
      "learning_rate": 9.558231808074156e-05,
      "loss": 0.2771,
      "step": 245
    },
    {
      "epoch": 1.690721649484536,
      "grad_norm": 0.10092959577494873,
      "learning_rate": 9.477977734335061e-05,
      "loss": 0.2899,
      "step": 246
    },
    {
      "epoch": 1.697594501718213,
      "grad_norm": 0.09948236816567041,
      "learning_rate": 9.397757360830353e-05,
      "loss": 0.2805,
      "step": 247
    },
    {
      "epoch": 1.7044673539518902,
      "grad_norm": 0.10097655488635103,
      "learning_rate": 9.317575866353292e-05,
      "loss": 0.296,
      "step": 248
    },
    {
      "epoch": 1.7113402061855671,
      "grad_norm": 0.09936217630195925,
      "learning_rate": 9.23743842718721e-05,
      "loss": 0.2973,
      "step": 249
    },
    {
      "epoch": 1.718213058419244,
      "grad_norm": 0.09913343698008316,
      "learning_rate": 9.157350216771378e-05,
      "loss": 0.2899,
      "step": 250
    },
    {
      "epoch": 1.725085910652921,
      "grad_norm": 0.09873019613022356,
      "learning_rate": 9.077316405366981e-05,
      "loss": 0.2827,
      "step": 251
    },
    {
      "epoch": 1.731958762886598,
      "grad_norm": 0.09795703366595764,
      "learning_rate": 8.997342159723371e-05,
      "loss": 0.2833,
      "step": 252
    },
    {
      "epoch": 1.738831615120275,
      "grad_norm": 0.0971884033549596,
      "learning_rate": 8.917432642744518e-05,
      "loss": 0.2813,
      "step": 253
    },
    {
      "epoch": 1.745704467353952,
      "grad_norm": 0.10306201837901838,
      "learning_rate": 8.83759301315568e-05,
      "loss": 0.2924,
      "step": 254
    },
    {
      "epoch": 1.7525773195876289,
      "grad_norm": 0.10374979571562502,
      "learning_rate": 8.757828425170404e-05,
      "loss": 0.2838,
      "step": 255
    },
    {
      "epoch": 1.7594501718213058,
      "grad_norm": 0.1048687957670358,
      "learning_rate": 8.678144028157759e-05,
      "loss": 0.2886,
      "step": 256
    },
    {
      "epoch": 1.7663230240549828,
      "grad_norm": 0.10256152550776314,
      "learning_rate": 8.598544966309925e-05,
      "loss": 0.2879,
      "step": 257
    },
    {
      "epoch": 1.7731958762886597,
      "grad_norm": 0.10130883507858145,
      "learning_rate": 8.519036378310096e-05,
      "loss": 0.2909,
      "step": 258
    },
    {
      "epoch": 1.7800687285223367,
      "grad_norm": 0.10467386983378474,
      "learning_rate": 8.43962339700073e-05,
      "loss": 0.2828,
      "step": 259
    },
    {
      "epoch": 1.7869415807560136,
      "grad_norm": 0.09820302182298171,
      "learning_rate": 8.360311149052205e-05,
      "loss": 0.2844,
      "step": 260
    },
    {
      "epoch": 1.7938144329896906,
      "grad_norm": 0.10102934299035139,
      "learning_rate": 8.281104754631835e-05,
      "loss": 0.2898,
      "step": 261
    },
    {
      "epoch": 1.8006872852233677,
      "grad_norm": 0.0958354586875355,
      "learning_rate": 8.20200932707335e-05,
      "loss": 0.28,
      "step": 262
    },
    {
      "epoch": 1.8075601374570447,
      "grad_norm": 0.09957495231488761,
      "learning_rate": 8.123029972546781e-05,
      "loss": 0.284,
      "step": 263
    },
    {
      "epoch": 1.8144329896907216,
      "grad_norm": 0.10693182349883178,
      "learning_rate": 8.044171789728816e-05,
      "loss": 0.2892,
      "step": 264
    },
    {
      "epoch": 1.8213058419243986,
      "grad_norm": 0.10646064359160178,
      "learning_rate": 7.965439869473664e-05,
      "loss": 0.2958,
      "step": 265
    },
    {
      "epoch": 1.8281786941580758,
      "grad_norm": 0.10085406622127646,
      "learning_rate": 7.886839294484377e-05,
      "loss": 0.2818,
      "step": 266
    },
    {
      "epoch": 1.8350515463917527,
      "grad_norm": 0.10500151086048896,
      "learning_rate": 7.808375138984745e-05,
      "loss": 0.2733,
      "step": 267
    },
    {
      "epoch": 1.8419243986254297,
      "grad_norm": 0.1002635668475348,
      "learning_rate": 7.730052468391725e-05,
      "loss": 0.285,
      "step": 268
    },
    {
      "epoch": 1.8487972508591066,
      "grad_norm": 0.09642728830623505,
      "learning_rate": 7.6518763389884e-05,
      "loss": 0.2728,
      "step": 269
    },
    {
      "epoch": 1.8556701030927836,
      "grad_norm": 0.10268576835416232,
      "learning_rate": 7.573851797597602e-05,
      "loss": 0.2795,
      "step": 270
    },
    {
      "epoch": 1.8625429553264605,
      "grad_norm": 0.10795408269996427,
      "learning_rate": 7.495983881256067e-05,
      "loss": 0.295,
      "step": 271
    },
    {
      "epoch": 1.8694158075601375,
      "grad_norm": 0.10121463339523644,
      "learning_rate": 7.418277616889282e-05,
      "loss": 0.2787,
      "step": 272
    },
    {
      "epoch": 1.8762886597938144,
      "grad_norm": 0.0982041211411212,
      "learning_rate": 7.340738020986961e-05,
      "loss": 0.2913,
      "step": 273
    },
    {
      "epoch": 1.8831615120274914,
      "grad_norm": 0.09689663299295423,
      "learning_rate": 7.263370099279172e-05,
      "loss": 0.2678,
      "step": 274
    },
    {
      "epoch": 1.8900343642611683,
      "grad_norm": 0.09890680768599139,
      "learning_rate": 7.186178846413214e-05,
      "loss": 0.2622,
      "step": 275
    },
    {
      "epoch": 1.8969072164948453,
      "grad_norm": 0.10218531402065698,
      "learning_rate": 7.109169245631149e-05,
      "loss": 0.2948,
      "step": 276
    },
    {
      "epoch": 1.9037800687285222,
      "grad_norm": 0.09970764344361568,
      "learning_rate": 7.032346268448118e-05,
      "loss": 0.2936,
      "step": 277
    },
    {
      "epoch": 1.9106529209621992,
      "grad_norm": 0.10646182283668124,
      "learning_rate": 6.955714874331387e-05,
      "loss": 0.2841,
      "step": 278
    },
    {
      "epoch": 1.9175257731958761,
      "grad_norm": 0.1005308262146528,
      "learning_rate": 6.87928001038017e-05,
      "loss": 0.2741,
      "step": 279
    },
    {
      "epoch": 1.9243986254295533,
      "grad_norm": 0.09801297561347891,
      "learning_rate": 6.803046611006278e-05,
      "loss": 0.2812,
      "step": 280
    },
    {
      "epoch": 1.9312714776632303,
      "grad_norm": 0.10072370231430908,
      "learning_rate": 6.727019597615545e-05,
      "loss": 0.2861,
      "step": 281
    },
    {
      "epoch": 1.9381443298969072,
      "grad_norm": 0.10002866089004976,
      "learning_rate": 6.651203878290139e-05,
      "loss": 0.2747,
      "step": 282
    },
    {
      "epoch": 1.9450171821305842,
      "grad_norm": 0.09520289893088804,
      "learning_rate": 6.575604347471695e-05,
      "loss": 0.275,
      "step": 283
    },
    {
      "epoch": 1.9518900343642611,
      "grad_norm": 0.10328917600068767,
      "learning_rate": 6.500225885645346e-05,
      "loss": 0.2836,
      "step": 284
    },
    {
      "epoch": 1.9587628865979383,
      "grad_norm": 0.09800249344342209,
      "learning_rate": 6.425073359024663e-05,
      "loss": 0.2904,
      "step": 285
    },
    {
      "epoch": 1.9656357388316152,
      "grad_norm": 0.09952065394630734,
      "learning_rate": 6.350151619237488e-05,
      "loss": 0.2804,
      "step": 286
    },
    {
      "epoch": 1.9725085910652922,
      "grad_norm": 0.10493099048201263,
      "learning_rate": 6.275465503012751e-05,
      "loss": 0.2818,
      "step": 287
    },
    {
      "epoch": 1.9793814432989691,
      "grad_norm": 0.09722127539298636,
      "learning_rate": 6.201019831868208e-05,
      "loss": 0.2839,
      "step": 288
    },
    {
      "epoch": 1.986254295532646,
      "grad_norm": 0.09609058787958422,
      "learning_rate": 6.126819411799175e-05,
      "loss": 0.2725,
      "step": 289
    },
    {
      "epoch": 1.993127147766323,
      "grad_norm": 0.09821072360835045,
      "learning_rate": 6.052869032968285e-05,
      "loss": 0.2811,
      "step": 290
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.10288780904807039,
      "learning_rate": 5.979173469396227e-05,
      "loss": 0.2761,
      "step": 291
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.2949652373790741,
      "eval_runtime": 27.5209,
      "eval_samples_per_second": 35.5,
      "eval_steps_per_second": 1.126,
      "step": 291
    },
    {
      "epoch": 2.006872852233677,
      "grad_norm": 0.097538336458536,
      "learning_rate": 5.905737478653572e-05,
      "loss": 0.2679,
      "step": 292
    },
    {
      "epoch": 2.013745704467354,
      "grad_norm": 0.09988881705573123,
      "learning_rate": 5.83256580155362e-05,
      "loss": 0.2525,
      "step": 293
    },
    {
      "epoch": 2.020618556701031,
      "grad_norm": 0.0964482303899947,
      "learning_rate": 5.7596631618463514e-05,
      "loss": 0.2425,
      "step": 294
    },
    {
      "epoch": 2.027491408934708,
      "grad_norm": 0.10864263032849888,
      "learning_rate": 5.687034265913485e-05,
      "loss": 0.27,
      "step": 295
    },
    {
      "epoch": 2.0343642611683848,
      "grad_norm": 0.11734835455675438,
      "learning_rate": 5.614683802464631e-05,
      "loss": 0.2624,
      "step": 296
    },
    {
      "epoch": 2.0412371134020617,
      "grad_norm": 0.11289186335789302,
      "learning_rate": 5.542616442234618e-05,
      "loss": 0.259,
      "step": 297
    },
    {
      "epoch": 2.0481099656357387,
      "grad_norm": 0.11224803307996767,
      "learning_rate": 5.470836837681954e-05,
      "loss": 0.2607,
      "step": 298
    },
    {
      "epoch": 2.0549828178694156,
      "grad_norm": 0.10867092848715432,
      "learning_rate": 5.399349622688479e-05,
      "loss": 0.2703,
      "step": 299
    },
    {
      "epoch": 2.0618556701030926,
      "grad_norm": 0.10624834548230039,
      "learning_rate": 5.32815941226022e-05,
      "loss": 0.2654,
      "step": 300
    },
    {
      "epoch": 2.06872852233677,
      "grad_norm": 0.10758062626317398,
      "learning_rate": 5.2572708022294504e-05,
      "loss": 0.2698,
      "step": 301
    },
    {
      "epoch": 2.075601374570447,
      "grad_norm": 0.1071415029446744,
      "learning_rate": 5.1866883689580056e-05,
      "loss": 0.2632,
      "step": 302
    },
    {
      "epoch": 2.082474226804124,
      "grad_norm": 0.10125455614344606,
      "learning_rate": 5.116416669041843e-05,
      "loss": 0.2538,
      "step": 303
    },
    {
      "epoch": 2.089347079037801,
      "grad_norm": 0.10522800373863611,
      "learning_rate": 5.046460239016879e-05,
      "loss": 0.2693,
      "step": 304
    },
    {
      "epoch": 2.0962199312714778,
      "grad_norm": 0.10672396619233848,
      "learning_rate": 4.976823595066128e-05,
      "loss": 0.2575,
      "step": 305
    },
    {
      "epoch": 2.1030927835051547,
      "grad_norm": 0.10982835933289055,
      "learning_rate": 4.907511232728145e-05,
      "loss": 0.2684,
      "step": 306
    },
    {
      "epoch": 2.1099656357388317,
      "grad_norm": 0.10787287180280963,
      "learning_rate": 4.8385276266068146e-05,
      "loss": 0.258,
      "step": 307
    },
    {
      "epoch": 2.1168384879725086,
      "grad_norm": 0.11168112585189982,
      "learning_rate": 4.7698772300824756e-05,
      "loss": 0.2665,
      "step": 308
    },
    {
      "epoch": 2.1237113402061856,
      "grad_norm": 0.10887273811279173,
      "learning_rate": 4.7015644750244306e-05,
      "loss": 0.2609,
      "step": 309
    },
    {
      "epoch": 2.1305841924398625,
      "grad_norm": 0.10561230206084926,
      "learning_rate": 4.6335937715048306e-05,
      "loss": 0.2569,
      "step": 310
    },
    {
      "epoch": 2.1374570446735395,
      "grad_norm": 0.1085500386167761,
      "learning_rate": 4.565969507513981e-05,
      "loss": 0.2636,
      "step": 311
    },
    {
      "epoch": 2.1443298969072164,
      "grad_norm": 0.1012240249829896,
      "learning_rate": 4.498696048677059e-05,
      "loss": 0.2547,
      "step": 312
    },
    {
      "epoch": 2.1512027491408934,
      "grad_norm": 0.1030596147725605,
      "learning_rate": 4.4317777379722866e-05,
      "loss": 0.2531,
      "step": 313
    },
    {
      "epoch": 2.1580756013745703,
      "grad_norm": 0.10613673774187961,
      "learning_rate": 4.365218895450558e-05,
      "loss": 0.2608,
      "step": 314
    },
    {
      "epoch": 2.1649484536082473,
      "grad_norm": 0.10496672072120719,
      "learning_rate": 4.29902381795655e-05,
      "loss": 0.2711,
      "step": 315
    },
    {
      "epoch": 2.1718213058419242,
      "grad_norm": 0.1017243256571039,
      "learning_rate": 4.2331967788513295e-05,
      "loss": 0.2506,
      "step": 316
    },
    {
      "epoch": 2.178694158075601,
      "grad_norm": 0.1083774959415722,
      "learning_rate": 4.167742027736482e-05,
      "loss": 0.2706,
      "step": 317
    },
    {
      "epoch": 2.1855670103092786,
      "grad_norm": 0.10375985752479969,
      "learning_rate": 4.102663790179764e-05,
      "loss": 0.2446,
      "step": 318
    },
    {
      "epoch": 2.1924398625429555,
      "grad_norm": 0.10866264016891114,
      "learning_rate": 4.037966267442315e-05,
      "loss": 0.2782,
      "step": 319
    },
    {
      "epoch": 2.1993127147766325,
      "grad_norm": 0.10829606781249333,
      "learning_rate": 3.973653636207437e-05,
      "loss": 0.2628,
      "step": 320
    },
    {
      "epoch": 2.2061855670103094,
      "grad_norm": 0.11273324743023601,
      "learning_rate": 3.909730048310962e-05,
      "loss": 0.2661,
      "step": 321
    },
    {
      "epoch": 2.2130584192439864,
      "grad_norm": 0.10783971891161509,
      "learning_rate": 3.846199630473216e-05,
      "loss": 0.257,
      "step": 322
    },
    {
      "epoch": 2.2199312714776633,
      "grad_norm": 0.10744882559462787,
      "learning_rate": 3.7830664840326145e-05,
      "loss": 0.2431,
      "step": 323
    },
    {
      "epoch": 2.2268041237113403,
      "grad_norm": 0.10847966130290622,
      "learning_rate": 3.720334684680889e-05,
      "loss": 0.2645,
      "step": 324
    },
    {
      "epoch": 2.2336769759450172,
      "grad_norm": 0.10687099898167607,
      "learning_rate": 3.6580082821999786e-05,
      "loss": 0.2558,
      "step": 325
    },
    {
      "epoch": 2.240549828178694,
      "grad_norm": 0.10744277727601985,
      "learning_rate": 3.596091300200578e-05,
      "loss": 0.2648,
      "step": 326
    },
    {
      "epoch": 2.247422680412371,
      "grad_norm": 0.1075527766804997,
      "learning_rate": 3.534587735862391e-05,
      "loss": 0.2662,
      "step": 327
    },
    {
      "epoch": 2.254295532646048,
      "grad_norm": 0.10693314058206296,
      "learning_rate": 3.473501559676088e-05,
      "loss": 0.2691,
      "step": 328
    },
    {
      "epoch": 2.261168384879725,
      "grad_norm": 0.10977366989631175,
      "learning_rate": 3.4128367151869714e-05,
      "loss": 0.2721,
      "step": 329
    },
    {
      "epoch": 2.268041237113402,
      "grad_norm": 0.10791580217848652,
      "learning_rate": 3.352597118740404e-05,
      "loss": 0.2574,
      "step": 330
    },
    {
      "epoch": 2.274914089347079,
      "grad_norm": 0.1117741340523688,
      "learning_rate": 3.292786659228973e-05,
      "loss": 0.2701,
      "step": 331
    },
    {
      "epoch": 2.281786941580756,
      "grad_norm": 0.10832840646361684,
      "learning_rate": 3.233409197841437e-05,
      "loss": 0.2533,
      "step": 332
    },
    {
      "epoch": 2.288659793814433,
      "grad_norm": 0.11021992972158987,
      "learning_rate": 3.174468567813461e-05,
      "loss": 0.2506,
      "step": 333
    },
    {
      "epoch": 2.29553264604811,
      "grad_norm": 0.11244737426877117,
      "learning_rate": 3.115968574180149e-05,
      "loss": 0.2741,
      "step": 334
    },
    {
      "epoch": 2.3024054982817868,
      "grad_norm": 0.10439400399730109,
      "learning_rate": 3.0579129935304066e-05,
      "loss": 0.2423,
      "step": 335
    },
    {
      "epoch": 2.3092783505154637,
      "grad_norm": 0.11154784823612464,
      "learning_rate": 3.0003055737631403e-05,
      "loss": 0.2606,
      "step": 336
    },
    {
      "epoch": 2.3161512027491407,
      "grad_norm": 0.11618867786187727,
      "learning_rate": 2.9431500338452832e-05,
      "loss": 0.2705,
      "step": 337
    },
    {
      "epoch": 2.323024054982818,
      "grad_norm": 0.10953255344278685,
      "learning_rate": 2.886450063571735e-05,
      "loss": 0.2597,
      "step": 338
    },
    {
      "epoch": 2.329896907216495,
      "grad_norm": 0.10670783300766909,
      "learning_rate": 2.8302093233271453e-05,
      "loss": 0.2451,
      "step": 339
    },
    {
      "epoch": 2.336769759450172,
      "grad_norm": 0.10795627839159605,
      "learning_rate": 2.7744314438496088e-05,
      "loss": 0.2571,
      "step": 340
    },
    {
      "epoch": 2.343642611683849,
      "grad_norm": 0.10876660642511277,
      "learning_rate": 2.7191200259962934e-05,
      "loss": 0.2646,
      "step": 341
    },
    {
      "epoch": 2.350515463917526,
      "grad_norm": 0.10862804831060525,
      "learning_rate": 2.6642786405109475e-05,
      "loss": 0.2679,
      "step": 342
    },
    {
      "epoch": 2.357388316151203,
      "grad_norm": 0.1068062333293375,
      "learning_rate": 2.6099108277934103e-05,
      "loss": 0.2676,
      "step": 343
    },
    {
      "epoch": 2.3642611683848798,
      "grad_norm": 0.10420269958200333,
      "learning_rate": 2.556020097671046e-05,
      "loss": 0.2669,
      "step": 344
    },
    {
      "epoch": 2.3711340206185567,
      "grad_norm": 0.10951228936587228,
      "learning_rate": 2.5026099291721516e-05,
      "loss": 0.2519,
      "step": 345
    },
    {
      "epoch": 2.3780068728522337,
      "grad_norm": 0.11129513461474132,
      "learning_rate": 2.449683770301382e-05,
      "loss": 0.2571,
      "step": 346
    },
    {
      "epoch": 2.3848797250859106,
      "grad_norm": 0.10763529170641503,
      "learning_rate": 2.397245037817125e-05,
      "loss": 0.2567,
      "step": 347
    },
    {
      "epoch": 2.3917525773195876,
      "grad_norm": 0.10496278836242602,
      "learning_rate": 2.345297117010954e-05,
      "loss": 0.2459,
      "step": 348
    },
    {
      "epoch": 2.3986254295532645,
      "grad_norm": 0.11015993294501161,
      "learning_rate": 2.2938433614890697e-05,
      "loss": 0.2702,
      "step": 349
    },
    {
      "epoch": 2.4054982817869415,
      "grad_norm": 0.10738471004067172,
      "learning_rate": 2.242887092955801e-05,
      "loss": 0.2585,
      "step": 350
    },
    {
      "epoch": 2.4123711340206184,
      "grad_norm": 0.11218254036958039,
      "learning_rate": 2.1924316009991787e-05,
      "loss": 0.2607,
      "step": 351
    },
    {
      "epoch": 2.4192439862542954,
      "grad_norm": 0.10844975577221563,
      "learning_rate": 2.1424801428785447e-05,
      "loss": 0.2561,
      "step": 352
    },
    {
      "epoch": 2.4261168384879723,
      "grad_norm": 0.10599271281183413,
      "learning_rate": 2.0930359433142932e-05,
      "loss": 0.2501,
      "step": 353
    },
    {
      "epoch": 2.4329896907216497,
      "grad_norm": 0.11017719350663994,
      "learning_rate": 2.0441021942796944e-05,
      "loss": 0.2618,
      "step": 354
    },
    {
      "epoch": 2.4398625429553267,
      "grad_norm": 0.10668690539385935,
      "learning_rate": 1.995682054794803e-05,
      "loss": 0.264,
      "step": 355
    },
    {
      "epoch": 2.4467353951890036,
      "grad_norm": 0.10450005811019492,
      "learning_rate": 1.9477786507225616e-05,
      "loss": 0.2553,
      "step": 356
    },
    {
      "epoch": 2.4536082474226806,
      "grad_norm": 0.11220780349903246,
      "learning_rate": 1.900395074566962e-05,
      "loss": 0.2654,
      "step": 357
    },
    {
      "epoch": 2.4604810996563575,
      "grad_norm": 0.108477409950911,
      "learning_rate": 1.8535343852734332e-05,
      "loss": 0.2569,
      "step": 358
    },
    {
      "epoch": 2.4673539518900345,
      "grad_norm": 0.10460682785615502,
      "learning_rate": 1.8071996080313602e-05,
      "loss": 0.2535,
      "step": 359
    },
    {
      "epoch": 2.4742268041237114,
      "grad_norm": 0.1076833252048245,
      "learning_rate": 1.76139373407876e-05,
      "loss": 0.2651,
      "step": 360
    },
    {
      "epoch": 2.4810996563573884,
      "grad_norm": 0.10826152402956611,
      "learning_rate": 1.7161197205092216e-05,
      "loss": 0.2676,
      "step": 361
    },
    {
      "epoch": 2.4879725085910653,
      "grad_norm": 0.10750093502253717,
      "learning_rate": 1.6713804900809582e-05,
      "loss": 0.2559,
      "step": 362
    },
    {
      "epoch": 2.4948453608247423,
      "grad_norm": 0.1038001972682153,
      "learning_rate": 1.6271789310281517e-05,
      "loss": 0.2423,
      "step": 363
    },
    {
      "epoch": 2.5017182130584192,
      "grad_norm": 0.10672842152395061,
      "learning_rate": 1.583517896874498e-05,
      "loss": 0.2593,
      "step": 364
    },
    {
      "epoch": 2.508591065292096,
      "grad_norm": 0.10761120219562789,
      "learning_rate": 1.540400206248963e-05,
      "loss": 0.2595,
      "step": 365
    },
    {
      "epoch": 2.515463917525773,
      "grad_norm": 0.10816060382518869,
      "learning_rate": 1.4978286427038601e-05,
      "loss": 0.2539,
      "step": 366
    },
    {
      "epoch": 2.52233676975945,
      "grad_norm": 0.10455085413886135,
      "learning_rate": 1.4558059545351143e-05,
      "loss": 0.2564,
      "step": 367
    },
    {
      "epoch": 2.529209621993127,
      "grad_norm": 0.10969390724809498,
      "learning_rate": 1.4143348546048707e-05,
      "loss": 0.2552,
      "step": 368
    },
    {
      "epoch": 2.536082474226804,
      "grad_norm": 0.11070585821467833,
      "learning_rate": 1.3734180201663439e-05,
      "loss": 0.2561,
      "step": 369
    },
    {
      "epoch": 2.542955326460481,
      "grad_norm": 0.11065650967955783,
      "learning_rate": 1.3330580926909763e-05,
      "loss": 0.2552,
      "step": 370
    },
    {
      "epoch": 2.549828178694158,
      "grad_norm": 0.10753458918108932,
      "learning_rate": 1.2932576776979377e-05,
      "loss": 0.2599,
      "step": 371
    },
    {
      "epoch": 2.556701030927835,
      "grad_norm": 0.11215812047440889,
      "learning_rate": 1.2540193445858883e-05,
      "loss": 0.26,
      "step": 372
    },
    {
      "epoch": 2.563573883161512,
      "grad_norm": 0.11029227457181179,
      "learning_rate": 1.2153456264671337e-05,
      "loss": 0.2529,
      "step": 373
    },
    {
      "epoch": 2.5704467353951888,
      "grad_norm": 0.1095036155268631,
      "learning_rate": 1.1772390200040817e-05,
      "loss": 0.2503,
      "step": 374
    },
    {
      "epoch": 2.5773195876288657,
      "grad_norm": 0.10487179029197675,
      "learning_rate": 1.139701985248055e-05,
      "loss": 0.2421,
      "step": 375
    },
    {
      "epoch": 2.584192439862543,
      "grad_norm": 0.10299175954283599,
      "learning_rate": 1.1027369454805058e-05,
      "loss": 0.2502,
      "step": 376
    },
    {
      "epoch": 2.59106529209622,
      "grad_norm": 0.10855354801856847,
      "learning_rate": 1.0663462870565411e-05,
      "loss": 0.2608,
      "step": 377
    },
    {
      "epoch": 2.597938144329897,
      "grad_norm": 0.10315061729781284,
      "learning_rate": 1.0305323592509009e-05,
      "loss": 0.2528,
      "step": 378
    },
    {
      "epoch": 2.604810996563574,
      "grad_norm": 0.10626039326093263,
      "learning_rate": 9.952974741062703e-06,
      "loss": 0.264,
      "step": 379
    },
    {
      "epoch": 2.611683848797251,
      "grad_norm": 0.10745271675212126,
      "learning_rate": 9.606439062840256e-06,
      "loss": 0.2594,
      "step": 380
    },
    {
      "epoch": 2.618556701030928,
      "grad_norm": 0.10974406713413064,
      "learning_rate": 9.265738929174051e-06,
      "loss": 0.2582,
      "step": 381
    },
    {
      "epoch": 2.625429553264605,
      "grad_norm": 0.10537263197622973,
      "learning_rate": 8.93089633467058e-06,
      "loss": 0.2513,
      "step": 382
    },
    {
      "epoch": 2.6323024054982818,
      "grad_norm": 0.10396472858049301,
      "learning_rate": 8.601932895790877e-06,
      "loss": 0.2583,
      "step": 383
    },
    {
      "epoch": 2.6391752577319587,
      "grad_norm": 0.10276278086397593,
      "learning_rate": 8.278869849454718e-06,
      "loss": 0.2529,
      "step": 384
    },
    {
      "epoch": 2.6460481099656357,
      "grad_norm": 0.10899885081403418,
      "learning_rate": 7.961728051669737e-06,
      "loss": 0.2637,
      "step": 385
    },
    {
      "epoch": 2.6529209621993126,
      "grad_norm": 0.10582666731745977,
      "learning_rate": 7.650527976185173e-06,
      "loss": 0.2535,
      "step": 386
    },
    {
      "epoch": 2.6597938144329896,
      "grad_norm": 0.10725397582281598,
      "learning_rate": 7.3452897131698564e-06,
      "loss": 0.2603,
      "step": 387
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.10749509280278831,
      "learning_rate": 7.046032967915483e-06,
      "loss": 0.2489,
      "step": 388
    },
    {
      "epoch": 2.673539518900344,
      "grad_norm": 0.1087201328812968,
      "learning_rate": 6.75277705956443e-06,
      "loss": 0.2604,
      "step": 389
    },
    {
      "epoch": 2.680412371134021,
      "grad_norm": 0.10783575017518213,
      "learning_rate": 6.465540919862456e-06,
      "loss": 0.2557,
      "step": 390
    },
    {
      "epoch": 2.687285223367698,
      "grad_norm": 0.10758942906584705,
      "learning_rate": 6.184343091936751e-06,
      "loss": 0.2526,
      "step": 391
    },
    {
      "epoch": 2.6941580756013748,
      "grad_norm": 0.10528296367950161,
      "learning_rate": 5.909201729098579e-06,
      "loss": 0.2487,
      "step": 392
    },
    {
      "epoch": 2.7010309278350517,
      "grad_norm": 0.11256201944946768,
      "learning_rate": 5.640134593671598e-06,
      "loss": 0.2714,
      "step": 393
    },
    {
      "epoch": 2.7079037800687287,
      "grad_norm": 0.10863241059875908,
      "learning_rate": 5.3771590558450265e-06,
      "loss": 0.2574,
      "step": 394
    },
    {
      "epoch": 2.7147766323024056,
      "grad_norm": 0.10549349184272533,
      "learning_rate": 5.12029209255227e-06,
      "loss": 0.2554,
      "step": 395
    },
    {
      "epoch": 2.7216494845360826,
      "grad_norm": 0.10847174602209737,
      "learning_rate": 4.869550286375091e-06,
      "loss": 0.268,
      "step": 396
    },
    {
      "epoch": 2.7285223367697595,
      "grad_norm": 0.10488356613015705,
      "learning_rate": 4.624949824472858e-06,
      "loss": 0.2671,
      "step": 397
    },
    {
      "epoch": 2.7353951890034365,
      "grad_norm": 0.10475860935600055,
      "learning_rate": 4.386506497537757e-06,
      "loss": 0.2514,
      "step": 398
    },
    {
      "epoch": 2.7422680412371134,
      "grad_norm": 0.10597713121898167,
      "learning_rate": 4.154235698775277e-06,
      "loss": 0.2597,
      "step": 399
    },
    {
      "epoch": 2.7491408934707904,
      "grad_norm": 0.10595800588015507,
      "learning_rate": 3.928152422910491e-06,
      "loss": 0.2602,
      "step": 400
    },
    {
      "epoch": 2.7560137457044673,
      "grad_norm": 0.11154644583398933,
      "learning_rate": 3.7082712652200867e-06,
      "loss": 0.264,
      "step": 401
    },
    {
      "epoch": 2.7628865979381443,
      "grad_norm": 0.10694150624625907,
      "learning_rate": 3.4946064205899965e-06,
      "loss": 0.25,
      "step": 402
    },
    {
      "epoch": 2.7697594501718212,
      "grad_norm": 0.10318772901644747,
      "learning_rate": 3.287171682599255e-06,
      "loss": 0.2493,
      "step": 403
    },
    {
      "epoch": 2.776632302405498,
      "grad_norm": 0.1061914310952031,
      "learning_rate": 3.085980442629288e-06,
      "loss": 0.2591,
      "step": 404
    },
    {
      "epoch": 2.783505154639175,
      "grad_norm": 0.10446042630271005,
      "learning_rate": 2.8910456889995498e-06,
      "loss": 0.254,
      "step": 405
    },
    {
      "epoch": 2.790378006872852,
      "grad_norm": 0.10518845803888338,
      "learning_rate": 2.7023800061289907e-06,
      "loss": 0.2533,
      "step": 406
    },
    {
      "epoch": 2.797250859106529,
      "grad_norm": 0.10867396165520546,
      "learning_rate": 2.5199955737236104e-06,
      "loss": 0.2724,
      "step": 407
    },
    {
      "epoch": 2.804123711340206,
      "grad_norm": 0.10694114528768675,
      "learning_rate": 2.3439041659902407e-06,
      "loss": 0.2564,
      "step": 408
    },
    {
      "epoch": 2.810996563573883,
      "grad_norm": 0.10518421192748285,
      "learning_rate": 2.174117150876398e-06,
      "loss": 0.2609,
      "step": 409
    },
    {
      "epoch": 2.81786941580756,
      "grad_norm": 0.10729647762632918,
      "learning_rate": 2.010645489336382e-06,
      "loss": 0.2587,
      "step": 410
    },
    {
      "epoch": 2.824742268041237,
      "grad_norm": 0.10754314349393188,
      "learning_rate": 1.8534997346237093e-06,
      "loss": 0.2632,
      "step": 411
    },
    {
      "epoch": 2.8316151202749142,
      "grad_norm": 0.10802289125290783,
      "learning_rate": 1.7026900316098215e-06,
      "loss": 0.2566,
      "step": 412
    },
    {
      "epoch": 2.838487972508591,
      "grad_norm": 0.10931075718242897,
      "learning_rate": 1.5582261161291245e-06,
      "loss": 0.2626,
      "step": 413
    },
    {
      "epoch": 2.845360824742268,
      "grad_norm": 0.10501355229769045,
      "learning_rate": 1.4201173143504888e-06,
      "loss": 0.2621,
      "step": 414
    },
    {
      "epoch": 2.852233676975945,
      "grad_norm": 0.10753436065183695,
      "learning_rate": 1.2883725421752201e-06,
      "loss": 0.2475,
      "step": 415
    },
    {
      "epoch": 2.859106529209622,
      "grad_norm": 0.10463768922989454,
      "learning_rate": 1.1630003046614323e-06,
      "loss": 0.2409,
      "step": 416
    },
    {
      "epoch": 2.865979381443299,
      "grad_norm": 0.10543898222201153,
      "learning_rate": 1.0440086954749517e-06,
      "loss": 0.2505,
      "step": 417
    },
    {
      "epoch": 2.872852233676976,
      "grad_norm": 0.10892369128650758,
      "learning_rate": 9.314053963669245e-07,
      "loss": 0.2456,
      "step": 418
    },
    {
      "epoch": 2.879725085910653,
      "grad_norm": 0.1062255349453702,
      "learning_rate": 8.251976766777913e-07,
      "loss": 0.2649,
      "step": 419
    },
    {
      "epoch": 2.88659793814433,
      "grad_norm": 0.10869941184854195,
      "learning_rate": 7.253923928680406e-07,
      "loss": 0.2559,
      "step": 420
    },
    {
      "epoch": 2.893470790378007,
      "grad_norm": 0.1109190039810706,
      "learning_rate": 6.319959880756177e-07,
      "loss": 0.2529,
      "step": 421
    },
    {
      "epoch": 2.9003436426116838,
      "grad_norm": 0.10656654333503314,
      "learning_rate": 5.450144916999134e-07,
      "loss": 0.2566,
      "step": 422
    },
    {
      "epoch": 2.9072164948453607,
      "grad_norm": 0.10681708087523223,
      "learning_rate": 4.644535190125421e-07,
      "loss": 0.2526,
      "step": 423
    },
    {
      "epoch": 2.9140893470790377,
      "grad_norm": 0.11101267026333346,
      "learning_rate": 3.903182707948649e-07,
      "loss": 0.2545,
      "step": 424
    },
    {
      "epoch": 2.9209621993127146,
      "grad_norm": 0.10449251320574354,
      "learning_rate": 3.2261353300219176e-07,
      "loss": 0.2397,
      "step": 425
    },
    {
      "epoch": 2.927835051546392,
      "grad_norm": 0.10614426653162999,
      "learning_rate": 2.613436764548505e-07,
      "loss": 0.2451,
      "step": 426
    },
    {
      "epoch": 2.934707903780069,
      "grad_norm": 0.10344408619271417,
      "learning_rate": 2.0651265655603492e-07,
      "loss": 0.231,
      "step": 427
    },
    {
      "epoch": 2.941580756013746,
      "grad_norm": 0.10751538816500814,
      "learning_rate": 1.5812401303639813e-07,
      "loss": 0.253,
      "step": 428
    },
    {
      "epoch": 2.948453608247423,
      "grad_norm": 0.10902510633946422,
      "learning_rate": 1.1618086972559062e-07,
      "loss": 0.2534,
      "step": 429
    },
    {
      "epoch": 2.9553264604811,
      "grad_norm": 0.10961740281781805,
      "learning_rate": 8.068593435055505e-08,
      "loss": 0.2611,
      "step": 430
    },
    {
      "epoch": 2.9621993127147768,
      "grad_norm": 0.10667412761426893,
      "learning_rate": 5.164149836077714e-08,
      "loss": 0.2566,
      "step": 431
    },
    {
      "epoch": 2.9690721649484537,
      "grad_norm": 0.10469145515742652,
      "learning_rate": 2.9049436780281825e-08,
      "loss": 0.245,
      "step": 432
    },
    {
      "epoch": 2.9759450171821307,
      "grad_norm": 0.10866002746514136,
      "learning_rate": 1.2911208086663351e-08,
      "loss": 0.258,
      "step": 433
    },
    {
      "epoch": 2.9828178694158076,
      "grad_norm": 0.10779386109678486,
      "learning_rate": 3.2278541168717646e-09,
      "loss": 0.2668,
      "step": 434
    },
    {
      "epoch": 2.9896907216494846,
      "grad_norm": 0.11009984251752782,
      "learning_rate": 0.0,
      "loss": 0.2592,
      "step": 435
    },
    {
      "epoch": 2.9896907216494846,
      "eval_loss": 0.29278308153152466,
      "eval_runtime": 27.5656,
      "eval_samples_per_second": 35.443,
      "eval_steps_per_second": 1.125,
      "step": 435
    },
    {
      "epoch": 2.9896907216494846,
      "step": 435,
      "total_flos": 1.330526714897367e+17,
      "train_loss": 0.3336494640714821,
      "train_runtime": 4477.4048,
      "train_samples_per_second": 12.436,
      "train_steps_per_second": 0.097
    }
  ],
  "logging_steps": 1,
  "max_steps": 435,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.330526714897367e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}