{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.916590284142988,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0018331805682859762,
      "grad_norm": 0.07204689830541611,
      "learning_rate": 4e-05,
      "loss": 0.02,
      "step": 1
    },
    {
      "epoch": 0.0036663611365719525,
      "grad_norm": 0.0971730649471283,
      "learning_rate": 8e-05,
      "loss": 0.037,
      "step": 2
    },
    {
      "epoch": 0.005499541704857928,
      "grad_norm": 0.0824146643280983,
      "learning_rate": 0.00012,
      "loss": 0.0258,
      "step": 3
    },
    {
      "epoch": 0.007332722273143905,
      "grad_norm": 0.08816614001989365,
      "learning_rate": 0.00016,
      "loss": 0.0266,
      "step": 4
    },
    {
      "epoch": 0.00916590284142988,
      "grad_norm": 0.09286794066429138,
      "learning_rate": 0.0002,
      "loss": 0.0329,
      "step": 5
    },
    {
      "epoch": 0.010999083409715857,
      "grad_norm": 0.10284189879894257,
      "learning_rate": 0.00019981566820276498,
      "loss": 0.0338,
      "step": 6
    },
    {
      "epoch": 0.012832263978001834,
      "grad_norm": 0.07772455364465714,
      "learning_rate": 0.00019963133640552995,
      "loss": 0.0238,
      "step": 7
    },
    {
      "epoch": 0.01466544454628781,
      "grad_norm": 0.10870955139398575,
      "learning_rate": 0.00019944700460829492,
      "loss": 0.0302,
      "step": 8
    },
    {
      "epoch": 0.016498625114573784,
      "grad_norm": 0.10848188400268555,
      "learning_rate": 0.00019926267281105992,
      "loss": 0.0314,
      "step": 9
    },
    {
      "epoch": 0.01833180568285976,
      "grad_norm": 0.10071466863155365,
      "learning_rate": 0.0001990783410138249,
      "loss": 0.0281,
      "step": 10
    },
    {
      "epoch": 0.02016498625114574,
      "grad_norm": 0.08322206884622574,
      "learning_rate": 0.00019889400921658986,
      "loss": 0.0236,
      "step": 11
    },
    {
      "epoch": 0.021998166819431713,
      "grad_norm": 0.12429114431142807,
      "learning_rate": 0.00019870967741935483,
      "loss": 0.0488,
      "step": 12
    },
    {
      "epoch": 0.02383134738771769,
      "grad_norm": 0.08638562262058258,
      "learning_rate": 0.0001985253456221198,
      "loss": 0.0217,
      "step": 13
    },
    {
      "epoch": 0.025664527956003668,
      "grad_norm": 0.07736796885728836,
      "learning_rate": 0.0001983410138248848,
      "loss": 0.0223,
      "step": 14
    },
    {
      "epoch": 0.027497708524289642,
      "grad_norm": 0.08814552426338196,
      "learning_rate": 0.00019815668202764977,
      "loss": 0.0314,
      "step": 15
    },
    {
      "epoch": 0.02933088909257562,
      "grad_norm": 0.09797844290733337,
      "learning_rate": 0.00019797235023041477,
      "loss": 0.0341,
      "step": 16
    },
    {
      "epoch": 0.031164069660861594,
      "grad_norm": 0.11390747874975204,
      "learning_rate": 0.00019778801843317974,
      "loss": 0.0329,
      "step": 17
    },
    {
      "epoch": 0.03299725022914757,
      "grad_norm": 0.12596647441387177,
      "learning_rate": 0.00019760368663594472,
      "loss": 0.0479,
      "step": 18
    },
    {
      "epoch": 0.034830430797433545,
      "grad_norm": 0.08672292530536652,
      "learning_rate": 0.00019741935483870969,
      "loss": 0.0242,
      "step": 19
    },
    {
      "epoch": 0.03666361136571952,
      "grad_norm": 0.07240811735391617,
      "learning_rate": 0.00019723502304147466,
      "loss": 0.023,
      "step": 20
    },
    {
      "epoch": 0.0384967919340055,
      "grad_norm": 0.09740838408470154,
      "learning_rate": 0.00019705069124423966,
      "loss": 0.0325,
      "step": 21
    },
    {
      "epoch": 0.04032997250229148,
      "grad_norm": 0.10614212602376938,
      "learning_rate": 0.00019686635944700463,
      "loss": 0.0415,
      "step": 22
    },
    {
      "epoch": 0.04216315307057745,
      "grad_norm": 0.06703123450279236,
      "learning_rate": 0.0001966820276497696,
      "loss": 0.0201,
      "step": 23
    },
    {
      "epoch": 0.043996333638863426,
      "grad_norm": 0.11590456962585449,
      "learning_rate": 0.00019649769585253457,
      "loss": 0.0444,
      "step": 24
    },
    {
      "epoch": 0.045829514207149404,
      "grad_norm": 0.08663053065538406,
      "learning_rate": 0.00019631336405529954,
      "loss": 0.0331,
      "step": 25
    },
    {
      "epoch": 0.04766269477543538,
      "grad_norm": 0.0977555587887764,
      "learning_rate": 0.0001961290322580645,
      "loss": 0.0262,
      "step": 26
    },
    {
      "epoch": 0.04949587534372136,
      "grad_norm": 0.07877817749977112,
      "learning_rate": 0.0001959447004608295,
      "loss": 0.0309,
      "step": 27
    },
    {
      "epoch": 0.051329055912007336,
      "grad_norm": 0.09205110371112823,
      "learning_rate": 0.00019576036866359448,
      "loss": 0.0363,
      "step": 28
    },
    {
      "epoch": 0.05316223648029331,
      "grad_norm": 0.07161393016576767,
      "learning_rate": 0.00019557603686635945,
      "loss": 0.0231,
      "step": 29
    },
    {
      "epoch": 0.054995417048579284,
      "grad_norm": 0.1110185980796814,
      "learning_rate": 0.00019539170506912442,
      "loss": 0.0414,
      "step": 30
    },
    {
      "epoch": 0.05682859761686526,
      "grad_norm": 0.08546338975429535,
      "learning_rate": 0.0001952073732718894,
      "loss": 0.0297,
      "step": 31
    },
    {
      "epoch": 0.05866177818515124,
      "grad_norm": 0.08454198390245438,
      "learning_rate": 0.00019502304147465436,
      "loss": 0.0244,
      "step": 32
    },
    {
      "epoch": 0.06049495875343722,
      "grad_norm": 0.09889410436153412,
      "learning_rate": 0.00019483870967741936,
      "loss": 0.0334,
      "step": 33
    },
    {
      "epoch": 0.06232813932172319,
      "grad_norm": 0.10116388648748398,
      "learning_rate": 0.00019465437788018433,
      "loss": 0.0309,
      "step": 34
    },
    {
      "epoch": 0.06416131989000917,
      "grad_norm": 0.12611277401447296,
      "learning_rate": 0.00019447004608294933,
      "loss": 0.0511,
      "step": 35
    },
    {
      "epoch": 0.06599450045829514,
      "grad_norm": 0.11593794077634811,
      "learning_rate": 0.0001942857142857143,
      "loss": 0.046,
      "step": 36
    },
    {
      "epoch": 0.06782768102658111,
      "grad_norm": 0.09269952028989792,
      "learning_rate": 0.00019410138248847927,
      "loss": 0.0269,
      "step": 37
    },
    {
      "epoch": 0.06966086159486709,
      "grad_norm": 0.08128319680690765,
      "learning_rate": 0.00019391705069124425,
      "loss": 0.0267,
      "step": 38
    },
    {
      "epoch": 0.07149404216315307,
      "grad_norm": 0.16776973009109497,
      "learning_rate": 0.00019373271889400924,
      "loss": 0.0333,
      "step": 39
    },
    {
      "epoch": 0.07332722273143905,
      "grad_norm": 0.09216058254241943,
      "learning_rate": 0.00019354838709677422,
      "loss": 0.03,
      "step": 40
    },
    {
      "epoch": 0.07516040329972502,
      "grad_norm": 0.0869499146938324,
      "learning_rate": 0.00019336405529953919,
      "loss": 0.0227,
      "step": 41
    },
    {
      "epoch": 0.076993583868011,
      "grad_norm": 0.08851771056652069,
      "learning_rate": 0.00019317972350230416,
      "loss": 0.0269,
      "step": 42
    },
    {
      "epoch": 0.07882676443629698,
      "grad_norm": 0.07213577628135681,
      "learning_rate": 0.00019299539170506913,
      "loss": 0.0271,
      "step": 43
    },
    {
      "epoch": 0.08065994500458296,
      "grad_norm": 0.08532075583934784,
      "learning_rate": 0.0001928110599078341,
      "loss": 0.0343,
      "step": 44
    },
    {
      "epoch": 0.08249312557286893,
      "grad_norm": 0.1002303957939148,
      "learning_rate": 0.0001926267281105991,
      "loss": 0.0421,
      "step": 45
    },
    {
      "epoch": 0.0843263061411549,
      "grad_norm": 0.10179378092288971,
      "learning_rate": 0.00019244239631336407,
      "loss": 0.0472,
      "step": 46
    },
    {
      "epoch": 0.08615948670944087,
      "grad_norm": 0.08771562576293945,
      "learning_rate": 0.00019225806451612904,
      "loss": 0.0329,
      "step": 47
    },
    {
      "epoch": 0.08799266727772685,
      "grad_norm": 0.0852821096777916,
      "learning_rate": 0.000192073732718894,
      "loss": 0.0271,
      "step": 48
    },
    {
      "epoch": 0.08982584784601283,
      "grad_norm": 0.0855298638343811,
      "learning_rate": 0.00019188940092165898,
      "loss": 0.0366,
      "step": 49
    },
    {
      "epoch": 0.09165902841429881,
      "grad_norm": 0.10242363810539246,
      "learning_rate": 0.00019170506912442395,
      "loss": 0.0409,
      "step": 50
    },
    {
      "epoch": 0.09349220898258478,
      "grad_norm": 0.07894504070281982,
      "learning_rate": 0.00019152073732718895,
      "loss": 0.0261,
      "step": 51
    },
    {
      "epoch": 0.09532538955087076,
      "grad_norm": 0.0891820639371872,
      "learning_rate": 0.00019133640552995392,
      "loss": 0.0371,
      "step": 52
    },
    {
      "epoch": 0.09715857011915674,
      "grad_norm": 0.09067510813474655,
      "learning_rate": 0.0001911520737327189,
      "loss": 0.0369,
      "step": 53
    },
    {
      "epoch": 0.09899175068744272,
      "grad_norm": 0.07762423902750015,
      "learning_rate": 0.0001909677419354839,
      "loss": 0.0267,
      "step": 54
    },
    {
      "epoch": 0.1008249312557287,
      "grad_norm": 0.09637603908777237,
      "learning_rate": 0.00019078341013824886,
      "loss": 0.0384,
      "step": 55
    },
    {
      "epoch": 0.10265811182401467,
      "grad_norm": 0.08040502667427063,
      "learning_rate": 0.00019059907834101383,
      "loss": 0.0291,
      "step": 56
    },
    {
      "epoch": 0.10449129239230064,
      "grad_norm": 0.07859490811824799,
      "learning_rate": 0.00019041474654377883,
      "loss": 0.0307,
      "step": 57
    },
    {
      "epoch": 0.10632447296058661,
      "grad_norm": 0.08587613701820374,
      "learning_rate": 0.0001902304147465438,
      "loss": 0.0327,
      "step": 58
    },
    {
      "epoch": 0.10815765352887259,
      "grad_norm": 0.09821908921003342,
      "learning_rate": 0.00019004608294930877,
      "loss": 0.0302,
      "step": 59
    },
    {
      "epoch": 0.10999083409715857,
      "grad_norm": 0.09804884344339371,
      "learning_rate": 0.00018986175115207375,
      "loss": 0.0397,
      "step": 60
    },
    {
      "epoch": 0.11182401466544455,
      "grad_norm": 0.1171504408121109,
      "learning_rate": 0.00018967741935483872,
      "loss": 0.0406,
      "step": 61
    },
    {
      "epoch": 0.11365719523373052,
      "grad_norm": 0.09860570728778839,
      "learning_rate": 0.0001894930875576037,
      "loss": 0.0381,
      "step": 62
    },
    {
      "epoch": 0.1154903758020165,
      "grad_norm": 0.10370708256959915,
      "learning_rate": 0.00018930875576036869,
      "loss": 0.0419,
      "step": 63
    },
    {
      "epoch": 0.11732355637030248,
      "grad_norm": 0.06204281374812126,
      "learning_rate": 0.00018912442396313366,
      "loss": 0.0219,
      "step": 64
    },
    {
      "epoch": 0.11915673693858846,
      "grad_norm": 0.09292633086442947,
      "learning_rate": 0.00018894009216589863,
      "loss": 0.0293,
      "step": 65
    },
    {
      "epoch": 0.12098991750687443,
      "grad_norm": 0.10509534925222397,
      "learning_rate": 0.0001887557603686636,
      "loss": 0.04,
      "step": 66
    },
    {
      "epoch": 0.1228230980751604,
      "grad_norm": 0.08757118135690689,
      "learning_rate": 0.00018857142857142857,
      "loss": 0.0295,
      "step": 67
    },
    {
      "epoch": 0.12465627864344637,
      "grad_norm": 0.08099905401468277,
      "learning_rate": 0.00018838709677419354,
      "loss": 0.027,
      "step": 68
    },
    {
      "epoch": 0.12648945921173235,
      "grad_norm": 0.0924796611070633,
      "learning_rate": 0.0001882027649769585,
      "loss": 0.0289,
      "step": 69
    },
    {
      "epoch": 0.12832263978001834,
      "grad_norm": 0.09552083164453506,
      "learning_rate": 0.0001880184331797235,
      "loss": 0.0322,
      "step": 70
    },
    {
      "epoch": 0.1301558203483043,
      "grad_norm": 0.09000714868307114,
      "learning_rate": 0.00018783410138248848,
      "loss": 0.0304,
      "step": 71
    },
    {
      "epoch": 0.13198900091659027,
      "grad_norm": 0.08881785720586777,
      "learning_rate": 0.00018764976958525345,
      "loss": 0.0305,
      "step": 72
    },
    {
      "epoch": 0.13382218148487626,
      "grad_norm": 0.10535095632076263,
      "learning_rate": 0.00018746543778801845,
      "loss": 0.0435,
      "step": 73
    },
    {
      "epoch": 0.13565536205316223,
      "grad_norm": 0.097115658223629,
      "learning_rate": 0.00018728110599078342,
      "loss": 0.0501,
      "step": 74
    },
    {
      "epoch": 0.13748854262144822,
      "grad_norm": 0.10467157512903214,
      "learning_rate": 0.0001870967741935484,
      "loss": 0.0523,
      "step": 75
    },
    {
      "epoch": 0.13932172318973418,
      "grad_norm": 0.08117009699344635,
      "learning_rate": 0.0001869124423963134,
      "loss": 0.0272,
      "step": 76
    },
    {
      "epoch": 0.14115490375802017,
      "grad_norm": 0.09257414937019348,
      "learning_rate": 0.00018672811059907836,
      "loss": 0.0375,
      "step": 77
    },
    {
      "epoch": 0.14298808432630614,
      "grad_norm": 0.07603290677070618,
      "learning_rate": 0.00018654377880184333,
      "loss": 0.0245,
      "step": 78
    },
    {
      "epoch": 0.14482126489459213,
      "grad_norm": 0.08395121246576309,
      "learning_rate": 0.0001863594470046083,
      "loss": 0.0379,
      "step": 79
    },
    {
      "epoch": 0.1466544454628781,
      "grad_norm": 0.09991391003131866,
      "learning_rate": 0.00018617511520737328,
      "loss": 0.0326,
      "step": 80
    },
    {
      "epoch": 0.14848762603116408,
      "grad_norm": 0.0780223086476326,
      "learning_rate": 0.00018599078341013825,
      "loss": 0.0332,
      "step": 81
    },
    {
      "epoch": 0.15032080659945005,
      "grad_norm": 0.07637890428304672,
      "learning_rate": 0.00018580645161290325,
      "loss": 0.0271,
      "step": 82
    },
    {
      "epoch": 0.152153987167736,
      "grad_norm": 0.0815526694059372,
      "learning_rate": 0.00018562211981566822,
      "loss": 0.0289,
      "step": 83
    },
    {
      "epoch": 0.153987167736022,
      "grad_norm": 0.07798799872398376,
      "learning_rate": 0.0001854377880184332,
      "loss": 0.032,
      "step": 84
    },
    {
      "epoch": 0.15582034830430797,
      "grad_norm": 0.06180557608604431,
      "learning_rate": 0.00018525345622119816,
      "loss": 0.0207,
      "step": 85
    },
    {
      "epoch": 0.15765352887259396,
      "grad_norm": 0.07907485961914062,
      "learning_rate": 0.00018506912442396313,
      "loss": 0.0281,
      "step": 86
    },
    {
      "epoch": 0.15948670944087992,
      "grad_norm": 0.0909823551774025,
      "learning_rate": 0.0001848847926267281,
      "loss": 0.032,
      "step": 87
    },
    {
      "epoch": 0.1613198900091659,
      "grad_norm": 0.09836460649967194,
      "learning_rate": 0.0001847004608294931,
      "loss": 0.0409,
      "step": 88
    },
    {
      "epoch": 0.16315307057745188,
      "grad_norm": 0.05404837429523468,
      "learning_rate": 0.00018451612903225807,
      "loss": 0.02,
      "step": 89
    },
    {
      "epoch": 0.16498625114573787,
      "grad_norm": 0.09892542660236359,
      "learning_rate": 0.00018433179723502304,
      "loss": 0.0321,
      "step": 90
    },
    {
      "epoch": 0.16681943171402383,
      "grad_norm": 0.08493707329034805,
      "learning_rate": 0.000184147465437788,
      "loss": 0.0343,
      "step": 91
    },
    {
      "epoch": 0.1686526122823098,
      "grad_norm": 0.10508857667446136,
      "learning_rate": 0.000183963133640553,
      "loss": 0.0397,
      "step": 92
    },
    {
      "epoch": 0.17048579285059579,
      "grad_norm": 0.08235018700361252,
      "learning_rate": 0.00018377880184331798,
      "loss": 0.0262,
      "step": 93
    },
    {
      "epoch": 0.17231897341888175,
      "grad_norm": 0.06784114986658096,
      "learning_rate": 0.00018359447004608298,
      "loss": 0.0247,
      "step": 94
    },
    {
      "epoch": 0.17415215398716774,
      "grad_norm": 0.10621548444032669,
      "learning_rate": 0.00018341013824884795,
      "loss": 0.035,
      "step": 95
    },
    {
      "epoch": 0.1759853345554537,
      "grad_norm": 0.10951874405145645,
      "learning_rate": 0.00018322580645161292,
      "loss": 0.0409,
      "step": 96
    },
    {
      "epoch": 0.1778185151237397,
      "grad_norm": 0.08758855611085892,
      "learning_rate": 0.0001830414746543779,
      "loss": 0.0389,
      "step": 97
    },
    {
      "epoch": 0.17965169569202566,
      "grad_norm": 0.10536627471446991,
      "learning_rate": 0.00018285714285714286,
      "loss": 0.0372,
      "step": 98
    },
    {
      "epoch": 0.18148487626031165,
      "grad_norm": 0.10224272310733795,
      "learning_rate": 0.00018267281105990784,
      "loss": 0.0394,
      "step": 99
    },
    {
      "epoch": 0.18331805682859761,
      "grad_norm": 0.07795912027359009,
      "learning_rate": 0.00018248847926267283,
      "loss": 0.0271,
      "step": 100
    },
    {
      "epoch": 0.1851512373968836,
      "grad_norm": 0.0965060293674469,
      "learning_rate": 0.0001823041474654378,
      "loss": 0.0357,
      "step": 101
    },
    {
      "epoch": 0.18698441796516957,
      "grad_norm": 0.061280328780412674,
      "learning_rate": 0.00018211981566820278,
      "loss": 0.0263,
      "step": 102
    },
    {
      "epoch": 0.18881759853345553,
      "grad_norm": 0.08007590472698212,
      "learning_rate": 0.00018193548387096775,
      "loss": 0.0324,
      "step": 103
    },
    {
      "epoch": 0.19065077910174152,
      "grad_norm": 0.08586332201957703,
      "learning_rate": 0.00018175115207373272,
      "loss": 0.0362,
      "step": 104
    },
    {
      "epoch": 0.1924839596700275,
      "grad_norm": 0.08350996673107147,
      "learning_rate": 0.0001815668202764977,
      "loss": 0.029,
      "step": 105
    },
    {
      "epoch": 0.19431714023831348,
      "grad_norm": 0.07932569086551666,
      "learning_rate": 0.0001813824884792627,
      "loss": 0.0262,
      "step": 106
    },
    {
      "epoch": 0.19615032080659944,
      "grad_norm": 0.08850853145122528,
      "learning_rate": 0.00018119815668202766,
      "loss": 0.0376,
      "step": 107
    },
    {
      "epoch": 0.19798350137488543,
      "grad_norm": 0.10090523213148117,
      "learning_rate": 0.00018101382488479263,
      "loss": 0.0437,
      "step": 108
    },
    {
      "epoch": 0.1998166819431714,
      "grad_norm": 0.09374283999204636,
      "learning_rate": 0.0001808294930875576,
      "loss": 0.0348,
      "step": 109
    },
    {
      "epoch": 0.2016498625114574,
      "grad_norm": 0.09550321102142334,
      "learning_rate": 0.00018064516129032257,
      "loss": 0.0415,
      "step": 110
    },
    {
      "epoch": 0.20348304307974335,
      "grad_norm": 0.09793347865343094,
      "learning_rate": 0.00018046082949308757,
      "loss": 0.0363,
      "step": 111
    },
    {
      "epoch": 0.20531622364802934,
      "grad_norm": 0.05917959660291672,
      "learning_rate": 0.00018027649769585254,
      "loss": 0.0233,
      "step": 112
    },
    {
      "epoch": 0.2071494042163153,
      "grad_norm": 0.08095124363899231,
      "learning_rate": 0.00018009216589861754,
      "loss": 0.0273,
      "step": 113
    },
    {
      "epoch": 0.20898258478460127,
      "grad_norm": 0.09870299696922302,
      "learning_rate": 0.0001799078341013825,
      "loss": 0.0341,
      "step": 114
    },
    {
      "epoch": 0.21081576535288726,
      "grad_norm": 0.06982927024364471,
      "learning_rate": 0.00017972350230414748,
      "loss": 0.0263,
      "step": 115
    },
    {
      "epoch": 0.21264894592117323,
      "grad_norm": 0.10992158949375153,
      "learning_rate": 0.00017953917050691245,
      "loss": 0.0444,
      "step": 116
    },
    {
      "epoch": 0.21448212648945922,
      "grad_norm": 0.11806368082761765,
      "learning_rate": 0.00017935483870967742,
      "loss": 0.045,
      "step": 117
    },
    {
      "epoch": 0.21631530705774518,
      "grad_norm": 0.061528291553258896,
      "learning_rate": 0.00017917050691244242,
      "loss": 0.0194,
      "step": 118
    },
    {
      "epoch": 0.21814848762603117,
      "grad_norm": 0.08814897388219833,
      "learning_rate": 0.0001789861751152074,
      "loss": 0.0335,
      "step": 119
    },
    {
      "epoch": 0.21998166819431714,
      "grad_norm": 0.07923831045627594,
      "learning_rate": 0.00017880184331797236,
      "loss": 0.0304,
      "step": 120
    },
    {
      "epoch": 0.22181484876260313,
      "grad_norm": 0.06676612794399261,
      "learning_rate": 0.00017861751152073734,
      "loss": 0.0261,
      "step": 121
    },
    {
      "epoch": 0.2236480293308891,
      "grad_norm": 0.10044591873884201,
      "learning_rate": 0.0001784331797235023,
      "loss": 0.0352,
      "step": 122
    },
    {
      "epoch": 0.22548120989917506,
      "grad_norm": 0.09440556168556213,
      "learning_rate": 0.00017824884792626728,
      "loss": 0.0394,
      "step": 123
    },
    {
      "epoch": 0.22731439046746105,
      "grad_norm": 0.07944708317518234,
      "learning_rate": 0.00017806451612903228,
      "loss": 0.0292,
      "step": 124
    },
    {
      "epoch": 0.229147571035747,
      "grad_norm": 0.11380550265312195,
      "learning_rate": 0.00017788018433179725,
      "loss": 0.0479,
      "step": 125
    },
    {
      "epoch": 0.230980751604033,
      "grad_norm": 0.09983845055103302,
      "learning_rate": 0.00017769585253456222,
      "loss": 0.043,
      "step": 126
    },
    {
      "epoch": 0.23281393217231897,
      "grad_norm": 0.09731481224298477,
      "learning_rate": 0.0001775115207373272,
      "loss": 0.0346,
      "step": 127
    },
    {
      "epoch": 0.23464711274060496,
      "grad_norm": 0.08189846575260162,
      "learning_rate": 0.00017732718894009216,
      "loss": 0.0314,
      "step": 128
    },
    {
      "epoch": 0.23648029330889092,
      "grad_norm": 0.11007854342460632,
      "learning_rate": 0.00017714285714285713,
      "loss": 0.0395,
      "step": 129
    },
    {
      "epoch": 0.2383134738771769,
      "grad_norm": 0.06934443861246109,
      "learning_rate": 0.00017695852534562213,
      "loss": 0.0289,
      "step": 130
    },
    {
      "epoch": 0.24014665444546288,
      "grad_norm": 0.0965190976858139,
      "learning_rate": 0.0001767741935483871,
      "loss": 0.0341,
      "step": 131
    },
    {
      "epoch": 0.24197983501374887,
      "grad_norm": 0.11201060563325882,
      "learning_rate": 0.0001765898617511521,
      "loss": 0.0404,
      "step": 132
    },
    {
      "epoch": 0.24381301558203483,
      "grad_norm": 0.07356410473585129,
      "learning_rate": 0.00017640552995391707,
      "loss": 0.0225,
      "step": 133
    },
    {
      "epoch": 0.2456461961503208,
      "grad_norm": 0.09507370740175247,
      "learning_rate": 0.00017622119815668204,
      "loss": 0.0409,
      "step": 134
    },
    {
      "epoch": 0.24747937671860679,
      "grad_norm": 0.09385097771883011,
      "learning_rate": 0.000176036866359447,
      "loss": 0.0314,
      "step": 135
    },
    {
      "epoch": 0.24931255728689275,
      "grad_norm": 0.08933474868535995,
      "learning_rate": 0.00017585253456221198,
      "loss": 0.0308,
      "step": 136
    },
    {
      "epoch": 0.25114573785517874,
      "grad_norm": 0.07607486099004745,
      "learning_rate": 0.00017566820276497698,
      "loss": 0.0272,
      "step": 137
    },
    {
      "epoch": 0.2529789184234647,
      "grad_norm": 0.060726869851350784,
      "learning_rate": 0.00017548387096774195,
      "loss": 0.0254,
      "step": 138
    },
    {
      "epoch": 0.25481209899175067,
      "grad_norm": 0.08471496403217316,
      "learning_rate": 0.00017529953917050692,
      "loss": 0.0299,
      "step": 139
    },
    {
      "epoch": 0.2566452795600367,
      "grad_norm": 0.06967601180076599,
      "learning_rate": 0.0001751152073732719,
      "loss": 0.0234,
      "step": 140
    },
    {
      "epoch": 0.25847846012832265,
      "grad_norm": 0.08925054222345352,
      "learning_rate": 0.00017493087557603687,
      "loss": 0.0375,
      "step": 141
    },
    {
      "epoch": 0.2603116406966086,
      "grad_norm": 0.07857096195220947,
      "learning_rate": 0.00017474654377880184,
      "loss": 0.0294,
      "step": 142
    },
    {
      "epoch": 0.2621448212648946,
      "grad_norm": 0.10110893100500107,
      "learning_rate": 0.00017456221198156684,
      "loss": 0.0388,
      "step": 143
    },
    {
      "epoch": 0.26397800183318054,
      "grad_norm": 0.10537184774875641,
      "learning_rate": 0.0001743778801843318,
      "loss": 0.0426,
      "step": 144
    },
    {
      "epoch": 0.26581118240146656,
      "grad_norm": 0.10398052632808685,
      "learning_rate": 0.00017419354838709678,
      "loss": 0.0362,
      "step": 145
    },
    {
      "epoch": 0.2676443629697525,
      "grad_norm": 0.06848938763141632,
      "learning_rate": 0.00017400921658986175,
      "loss": 0.028,
      "step": 146
    },
    {
      "epoch": 0.2694775435380385,
      "grad_norm": 0.08947031199932098,
      "learning_rate": 0.00017382488479262672,
      "loss": 0.0272,
      "step": 147
    },
    {
      "epoch": 0.27131072410632445,
      "grad_norm": 0.09318867325782776,
      "learning_rate": 0.0001736405529953917,
      "loss": 0.0328,
      "step": 148
    },
    {
      "epoch": 0.27314390467461047,
      "grad_norm": 0.0872950330376625,
      "learning_rate": 0.0001734562211981567,
      "loss": 0.0304,
      "step": 149
    },
    {
      "epoch": 0.27497708524289644,
      "grad_norm": 0.12327979505062103,
      "learning_rate": 0.00017327188940092166,
      "loss": 0.0542,
      "step": 150
    },
    {
      "epoch": 0.2768102658111824,
      "grad_norm": 0.08264505118131638,
      "learning_rate": 0.00017308755760368666,
      "loss": 0.0262,
      "step": 151
    },
    {
      "epoch": 0.27864344637946836,
      "grad_norm": 0.09241585433483124,
      "learning_rate": 0.00017290322580645163,
      "loss": 0.0323,
      "step": 152
    },
    {
      "epoch": 0.2804766269477543,
      "grad_norm": 0.09120775014162064,
      "learning_rate": 0.0001727188940092166,
      "loss": 0.031,
      "step": 153
    },
    {
      "epoch": 0.28230980751604035,
      "grad_norm": 0.05967549607157707,
      "learning_rate": 0.00017253456221198157,
      "loss": 0.02,
      "step": 154
    },
    {
      "epoch": 0.2841429880843263,
      "grad_norm": 0.08845420181751251,
      "learning_rate": 0.00017235023041474657,
      "loss": 0.0359,
      "step": 155
    },
    {
      "epoch": 0.2859761686526123,
      "grad_norm": 0.10303748399019241,
      "learning_rate": 0.00017216589861751154,
      "loss": 0.0336,
      "step": 156
    },
    {
      "epoch": 0.28780934922089824,
      "grad_norm": 0.11286526173353195,
      "learning_rate": 0.0001719815668202765,
      "loss": 0.0433,
      "step": 157
    },
    {
      "epoch": 0.28964252978918426,
      "grad_norm": 0.05800803378224373,
      "learning_rate": 0.00017179723502304148,
      "loss": 0.0191,
      "step": 158
    },
    {
      "epoch": 0.2914757103574702,
      "grad_norm": 0.06295135617256165,
      "learning_rate": 0.00017161290322580645,
      "loss": 0.0211,
      "step": 159
    },
    {
      "epoch": 0.2933088909257562,
      "grad_norm": 0.061198145151138306,
      "learning_rate": 0.00017142857142857143,
      "loss": 0.0222,
      "step": 160
    },
    {
      "epoch": 0.29514207149404215,
      "grad_norm": 0.08423091471195221,
      "learning_rate": 0.00017124423963133642,
      "loss": 0.029,
      "step": 161
    },
    {
      "epoch": 0.29697525206232817,
      "grad_norm": 0.07596798241138458,
      "learning_rate": 0.0001710599078341014,
      "loss": 0.0257,
      "step": 162
    },
    {
      "epoch": 0.29880843263061413,
      "grad_norm": 0.12243133038282394,
      "learning_rate": 0.00017087557603686637,
      "loss": 0.0488,
      "step": 163
    },
    {
      "epoch": 0.3006416131989001,
      "grad_norm": 0.10156381130218506,
      "learning_rate": 0.00017069124423963134,
      "loss": 0.0309,
      "step": 164
    },
    {
      "epoch": 0.30247479376718606,
      "grad_norm": 0.05903761461377144,
      "learning_rate": 0.0001705069124423963,
      "loss": 0.0179,
      "step": 165
    },
    {
      "epoch": 0.304307974335472,
      "grad_norm": 0.08366727083921432,
      "learning_rate": 0.00017032258064516128,
      "loss": 0.0297,
      "step": 166
    },
    {
      "epoch": 0.30614115490375804,
      "grad_norm": 0.09768462926149368,
      "learning_rate": 0.00017013824884792628,
      "loss": 0.0388,
      "step": 167
    },
    {
      "epoch": 0.307974335472044,
      "grad_norm": 0.07999719679355621,
      "learning_rate": 0.00016995391705069125,
      "loss": 0.0285,
      "step": 168
    },
    {
      "epoch": 0.30980751604032997,
      "grad_norm": 0.10129693150520325,
      "learning_rate": 0.00016976958525345622,
      "loss": 0.0437,
      "step": 169
    },
    {
      "epoch": 0.31164069660861593,
      "grad_norm": 0.07942084223031998,
      "learning_rate": 0.00016958525345622122,
      "loss": 0.028,
      "step": 170
    },
    {
      "epoch": 0.31347387717690195,
      "grad_norm": 0.09509172290563583,
      "learning_rate": 0.0001694009216589862,
      "loss": 0.035,
      "step": 171
    },
    {
      "epoch": 0.3153070577451879,
      "grad_norm": 0.10119883716106415,
      "learning_rate": 0.00016921658986175116,
      "loss": 0.0407,
      "step": 172
    },
    {
      "epoch": 0.3171402383134739,
      "grad_norm": 0.10474774241447449,
      "learning_rate": 0.00016903225806451616,
      "loss": 0.0382,
      "step": 173
    },
    {
      "epoch": 0.31897341888175984,
      "grad_norm": 0.08802273869514465,
      "learning_rate": 0.00016884792626728113,
      "loss": 0.0311,
      "step": 174
    },
    {
      "epoch": 0.3208065994500458,
      "grad_norm": 0.06499020010232925,
      "learning_rate": 0.0001686635944700461,
      "loss": 0.0247,
      "step": 175
    },
    {
      "epoch": 0.3226397800183318,
      "grad_norm": 0.09561455249786377,
      "learning_rate": 0.00016847926267281107,
      "loss": 0.0376,
      "step": 176
    },
    {
      "epoch": 0.3244729605866178,
      "grad_norm": 0.051068369299173355,
      "learning_rate": 0.00016829493087557604,
      "loss": 0.0194,
      "step": 177
    },
    {
      "epoch": 0.32630614115490375,
      "grad_norm": 0.09048140794038773,
      "learning_rate": 0.000168110599078341,
      "loss": 0.0344,
      "step": 178
    },
    {
      "epoch": 0.3281393217231897,
      "grad_norm": 0.08035707473754883,
      "learning_rate": 0.000167926267281106,
      "loss": 0.0323,
      "step": 179
    },
    {
      "epoch": 0.32997250229147573,
      "grad_norm": 0.10091862827539444,
      "learning_rate": 0.00016774193548387098,
      "loss": 0.0406,
      "step": 180
    },
    {
      "epoch": 0.3318056828597617,
      "grad_norm": 0.09862423688173294,
      "learning_rate": 0.00016755760368663595,
      "loss": 0.0475,
      "step": 181
    },
    {
      "epoch": 0.33363886342804766,
      "grad_norm": 0.10344900190830231,
      "learning_rate": 0.00016737327188940092,
      "loss": 0.0389,
      "step": 182
    },
    {
      "epoch": 0.3354720439963336,
      "grad_norm": 0.10084162652492523,
      "learning_rate": 0.0001671889400921659,
      "loss": 0.0387,
      "step": 183
    },
    {
      "epoch": 0.3373052245646196,
      "grad_norm": 0.08419749140739441,
      "learning_rate": 0.00016700460829493087,
      "loss": 0.0298,
      "step": 184
    },
    {
      "epoch": 0.3391384051329056,
      "grad_norm": 0.09623179584741592,
      "learning_rate": 0.00016682027649769587,
      "loss": 0.0291,
      "step": 185
    },
    {
      "epoch": 0.34097158570119157,
      "grad_norm": 0.0974535197019577,
      "learning_rate": 0.00016663594470046084,
      "loss": 0.0337,
      "step": 186
    },
    {
      "epoch": 0.34280476626947753,
      "grad_norm": 0.10164261609315872,
      "learning_rate": 0.0001664516129032258,
      "loss": 0.0466,
      "step": 187
    },
    {
      "epoch": 0.3446379468377635,
      "grad_norm": 0.0877864882349968,
      "learning_rate": 0.00016626728110599078,
      "loss": 0.0244,
      "step": 188
    },
    {
      "epoch": 0.3464711274060495,
      "grad_norm": 0.08111972361803055,
      "learning_rate": 0.00016608294930875578,
      "loss": 0.0285,
      "step": 189
    },
    {
      "epoch": 0.3483043079743355,
      "grad_norm": 0.0656951442360878,
      "learning_rate": 0.00016589861751152075,
      "loss": 0.0235,
      "step": 190
    },
    {
      "epoch": 0.35013748854262144,
      "grad_norm": 0.08031731843948364,
      "learning_rate": 0.00016571428571428575,
      "loss": 0.0321,
      "step": 191
    },
    {
      "epoch": 0.3519706691109074,
      "grad_norm": 0.1022307500243187,
      "learning_rate": 0.00016552995391705072,
      "loss": 0.0437,
      "step": 192
    },
    {
      "epoch": 0.3538038496791934,
      "grad_norm": 0.07790978997945786,
      "learning_rate": 0.0001653456221198157,
      "loss": 0.0305,
      "step": 193
    },
    {
      "epoch": 0.3556370302474794,
      "grad_norm": 0.06656166166067123,
      "learning_rate": 0.00016516129032258066,
      "loss": 0.027,
      "step": 194
    },
    {
      "epoch": 0.35747021081576535,
      "grad_norm": 0.08255946636199951,
      "learning_rate": 0.00016497695852534563,
      "loss": 0.0294,
      "step": 195
    },
    {
      "epoch": 0.3593033913840513,
      "grad_norm": 0.0771537646651268,
      "learning_rate": 0.0001647926267281106,
      "loss": 0.027,
      "step": 196
    },
    {
      "epoch": 0.3611365719523373,
      "grad_norm": 0.08929789811372757,
      "learning_rate": 0.00016460829493087557,
      "loss": 0.0435,
      "step": 197
    },
    {
      "epoch": 0.3629697525206233,
      "grad_norm": 0.07332134991884232,
      "learning_rate": 0.00016442396313364057,
      "loss": 0.0254,
      "step": 198
    },
    {
      "epoch": 0.36480293308890926,
      "grad_norm": 0.09681219607591629,
      "learning_rate": 0.00016423963133640554,
      "loss": 0.0404,
      "step": 199
    },
    {
      "epoch": 0.36663611365719523,
      "grad_norm": 0.0849548950791359,
      "learning_rate": 0.0001640552995391705,
      "loss": 0.0337,
      "step": 200
    },
    {
      "epoch": 0.3684692942254812,
      "grad_norm": 0.10003779083490372,
      "learning_rate": 0.00016387096774193548,
      "loss": 0.034,
      "step": 201
    },
    {
      "epoch": 0.3703024747937672,
      "grad_norm": 0.09228444844484329,
      "learning_rate": 0.00016368663594470046,
      "loss": 0.0362,
      "step": 202
    },
    {
      "epoch": 0.3721356553620532,
      "grad_norm": 0.06095033884048462,
      "learning_rate": 0.00016350230414746543,
      "loss": 0.0197,
      "step": 203
    },
    {
      "epoch": 0.37396883593033914,
      "grad_norm": 0.09504009038209915,
      "learning_rate": 0.00016331797235023042,
      "loss": 0.0356,
      "step": 204
    },
    {
      "epoch": 0.3758020164986251,
      "grad_norm": 0.07681944966316223,
      "learning_rate": 0.0001631336405529954,
      "loss": 0.0329,
      "step": 205
    },
    {
      "epoch": 0.37763519706691107,
      "grad_norm": 0.0894550010561943,
      "learning_rate": 0.00016294930875576037,
      "loss": 0.0327,
      "step": 206
    },
    {
      "epoch": 0.3794683776351971,
      "grad_norm": 0.09541459381580353,
      "learning_rate": 0.00016276497695852534,
      "loss": 0.0368,
      "step": 207
    },
    {
      "epoch": 0.38130155820348305,
      "grad_norm": 0.07463543862104416,
      "learning_rate": 0.00016258064516129034,
      "loss": 0.0272,
      "step": 208
    },
    {
      "epoch": 0.383134738771769,
      "grad_norm": 0.08396443724632263,
      "learning_rate": 0.0001623963133640553,
      "loss": 0.0316,
      "step": 209
    },
    {
      "epoch": 0.384967919340055,
      "grad_norm": 0.07742145657539368,
      "learning_rate": 0.0001622119815668203,
      "loss": 0.0276,
      "step": 210
    },
    {
      "epoch": 0.386801099908341,
      "grad_norm": 0.07310166209936142,
      "learning_rate": 0.00016202764976958528,
      "loss": 0.0241,
      "step": 211
    },
    {
      "epoch": 0.38863428047662696,
      "grad_norm": 0.09384534507989883,
      "learning_rate": 0.00016184331797235025,
      "loss": 0.0386,
      "step": 212
    },
    {
      "epoch": 0.3904674610449129,
      "grad_norm": 0.09084580093622208,
      "learning_rate": 0.00016165898617511522,
      "loss": 0.0293,
      "step": 213
    },
    {
      "epoch": 0.3923006416131989,
      "grad_norm": 0.10701391100883484,
      "learning_rate": 0.0001614746543778802,
      "loss": 0.046,
      "step": 214
    },
    {
      "epoch": 0.39413382218148485,
      "grad_norm": 0.07608213275671005,
      "learning_rate": 0.00016129032258064516,
      "loss": 0.031,
      "step": 215
    },
    {
      "epoch": 0.39596700274977087,
      "grad_norm": 0.09281232208013535,
      "learning_rate": 0.00016110599078341016,
      "loss": 0.0311,
      "step": 216
    },
    {
      "epoch": 0.39780018331805683,
      "grad_norm": 0.07548707723617554,
      "learning_rate": 0.00016092165898617513,
      "loss": 0.0244,
      "step": 217
    },
    {
      "epoch": 0.3996333638863428,
      "grad_norm": 0.08597145974636078,
      "learning_rate": 0.0001607373271889401,
      "loss": 0.0296,
      "step": 218
    },
    {
      "epoch": 0.40146654445462876,
      "grad_norm": 0.090579092502594,
      "learning_rate": 0.00016055299539170507,
      "loss": 0.0298,
      "step": 219
    },
    {
      "epoch": 0.4032997250229148,
      "grad_norm": 0.07005604356527328,
      "learning_rate": 0.00016036866359447004,
      "loss": 0.0254,
      "step": 220
    },
    {
      "epoch": 0.40513290559120074,
      "grad_norm": 0.09800952672958374,
      "learning_rate": 0.00016018433179723501,
      "loss": 0.0353,
      "step": 221
    },
    {
      "epoch": 0.4069660861594867,
      "grad_norm": 0.09682459384202957,
      "learning_rate": 0.00016,
      "loss": 0.0342,
      "step": 222
    },
    {
      "epoch": 0.40879926672777267,
      "grad_norm": 0.08852767199277878,
      "learning_rate": 0.00015981566820276498,
      "loss": 0.0356,
      "step": 223
    },
    {
      "epoch": 0.4106324472960587,
      "grad_norm": 0.07314983755350113,
      "learning_rate": 0.00015963133640552996,
      "loss": 0.0266,
      "step": 224
    },
    {
      "epoch": 0.41246562786434465,
      "grad_norm": 0.08612879365682602,
      "learning_rate": 0.00015944700460829493,
      "loss": 0.0314,
      "step": 225
    },
    {
      "epoch": 0.4142988084326306,
      "grad_norm": 0.09286980330944061,
      "learning_rate": 0.0001592626728110599,
      "loss": 0.0359,
      "step": 226
    },
    {
      "epoch": 0.4161319890009166,
      "grad_norm": 0.0787709653377533,
      "learning_rate": 0.0001590783410138249,
      "loss": 0.0299,
      "step": 227
    },
    {
      "epoch": 0.41796516956920254,
      "grad_norm": 0.08612968027591705,
      "learning_rate": 0.00015889400921658987,
      "loss": 0.0299,
      "step": 228
    },
    {
      "epoch": 0.41979835013748856,
      "grad_norm": 0.06926131248474121,
      "learning_rate": 0.00015870967741935487,
      "loss": 0.0257,
      "step": 229
    },
    {
      "epoch": 0.4216315307057745,
      "grad_norm": 0.09236182272434235,
      "learning_rate": 0.00015852534562211984,
      "loss": 0.0284,
      "step": 230
    },
    {
      "epoch": 0.4234647112740605,
      "grad_norm": 0.07634787261486053,
      "learning_rate": 0.0001583410138248848,
      "loss": 0.0335,
      "step": 231
    },
    {
      "epoch": 0.42529789184234645,
      "grad_norm": 0.10165869444608688,
      "learning_rate": 0.00015815668202764978,
      "loss": 0.0407,
      "step": 232
    },
    {
      "epoch": 0.4271310724106325,
      "grad_norm": 0.09464540332555771,
      "learning_rate": 0.00015797235023041475,
      "loss": 0.0389,
      "step": 233
    },
    {
      "epoch": 0.42896425297891844,
      "grad_norm": 0.0773908719420433,
      "learning_rate": 0.00015778801843317975,
      "loss": 0.0329,
      "step": 234
    },
    {
      "epoch": 0.4307974335472044,
      "grad_norm": 0.08985087275505066,
      "learning_rate": 0.00015760368663594472,
      "loss": 0.0328,
      "step": 235
    },
    {
      "epoch": 0.43263061411549036,
      "grad_norm": 0.06766209751367569,
      "learning_rate": 0.0001574193548387097,
      "loss": 0.0227,
      "step": 236
    },
    {
      "epoch": 0.4344637946837763,
      "grad_norm": 0.07593253999948502,
      "learning_rate": 0.00015723502304147466,
      "loss": 0.0252,
      "step": 237
    },
    {
      "epoch": 0.43629697525206235,
      "grad_norm": 0.0670701190829277,
      "learning_rate": 0.00015705069124423963,
      "loss": 0.022,
      "step": 238
    },
    {
      "epoch": 0.4381301558203483,
      "grad_norm": 0.08073533326387405,
      "learning_rate": 0.0001568663594470046,
      "loss": 0.0314,
      "step": 239
    },
    {
      "epoch": 0.4399633363886343,
      "grad_norm": 0.09111002087593079,
      "learning_rate": 0.0001566820276497696,
      "loss": 0.0298,
      "step": 240
    },
    {
      "epoch": 0.44179651695692024,
      "grad_norm": 0.07969270646572113,
      "learning_rate": 0.00015649769585253457,
      "loss": 0.0295,
      "step": 241
    },
    {
      "epoch": 0.44362969752520626,
      "grad_norm": 0.05996888503432274,
      "learning_rate": 0.00015631336405529954,
      "loss": 0.0233,
      "step": 242
    },
    {
      "epoch": 0.4454628780934922,
      "grad_norm": 0.0899052619934082,
      "learning_rate": 0.00015612903225806451,
      "loss": 0.0328,
      "step": 243
    },
    {
      "epoch": 0.4472960586617782,
      "grad_norm": 0.11184020340442657,
      "learning_rate": 0.00015594470046082949,
      "loss": 0.0456,
      "step": 244
    },
    {
      "epoch": 0.44912923923006415,
      "grad_norm": 0.08015838265419006,
      "learning_rate": 0.00015576036866359446,
      "loss": 0.0244,
      "step": 245
    },
    {
      "epoch": 0.4509624197983501,
      "grad_norm": 0.10306031256914139,
      "learning_rate": 0.00015557603686635946,
      "loss": 0.0397,
      "step": 246
    },
    {
      "epoch": 0.45279560036663613,
      "grad_norm": 0.08525974303483963,
      "learning_rate": 0.00015539170506912443,
      "loss": 0.0317,
      "step": 247
    },
    {
      "epoch": 0.4546287809349221,
      "grad_norm": 0.10056892782449722,
      "learning_rate": 0.00015520737327188942,
      "loss": 0.0456,
      "step": 248
    },
    {
      "epoch": 0.45646196150320806,
      "grad_norm": 0.09121023863554001,
      "learning_rate": 0.0001550230414746544,
      "loss": 0.0308,
      "step": 249
    },
    {
      "epoch": 0.458295142071494,
      "grad_norm": 0.06088346242904663,
      "learning_rate": 0.00015483870967741937,
      "loss": 0.0235,
      "step": 250
    },
    {
      "epoch": 0.46012832263978004,
      "grad_norm": 0.07533244043588638,
      "learning_rate": 0.00015465437788018434,
      "loss": 0.0309,
      "step": 251
    },
    {
      "epoch": 0.461961503208066,
      "grad_norm": 0.08711180835962296,
      "learning_rate": 0.00015447004608294934,
      "loss": 0.038,
      "step": 252
    },
    {
      "epoch": 0.46379468377635197,
      "grad_norm": 0.08756791800260544,
      "learning_rate": 0.0001542857142857143,
      "loss": 0.0332,
      "step": 253
    },
    {
      "epoch": 0.46562786434463793,
      "grad_norm": 0.09424729645252228,
      "learning_rate": 0.00015410138248847928,
      "loss": 0.0282,
      "step": 254
    },
    {
      "epoch": 0.4674610449129239,
      "grad_norm": 0.10178559273481369,
      "learning_rate": 0.00015391705069124425,
      "loss": 0.0319,
      "step": 255
    },
    {
      "epoch": 0.4692942254812099,
      "grad_norm": 0.09317290782928467,
      "learning_rate": 0.00015373271889400922,
      "loss": 0.0294,
      "step": 256
    },
    {
      "epoch": 0.4711274060494959,
      "grad_norm": 0.11380328983068466,
      "learning_rate": 0.0001535483870967742,
      "loss": 0.0474,
      "step": 257
    },
    {
      "epoch": 0.47296058661778184,
      "grad_norm": 0.0751282274723053,
      "learning_rate": 0.0001533640552995392,
      "loss": 0.0328,
      "step": 258
    },
    {
      "epoch": 0.4747937671860678,
      "grad_norm": 0.07097076624631882,
      "learning_rate": 0.00015317972350230416,
      "loss": 0.0314,
      "step": 259
    },
    {
      "epoch": 0.4766269477543538,
      "grad_norm": 0.0817055031657219,
      "learning_rate": 0.00015299539170506913,
      "loss": 0.0348,
      "step": 260
    },
    {
      "epoch": 0.4784601283226398,
      "grad_norm": 0.08542328327894211,
      "learning_rate": 0.0001528110599078341,
      "loss": 0.0314,
      "step": 261
    },
    {
      "epoch": 0.48029330889092575,
      "grad_norm": 0.08346550911664963,
      "learning_rate": 0.00015262672811059907,
      "loss": 0.0342,
      "step": 262
    },
    {
      "epoch": 0.4821264894592117,
      "grad_norm": 0.06232753023505211,
      "learning_rate": 0.00015244239631336405,
      "loss": 0.024,
      "step": 263
    },
    {
      "epoch": 0.48395967002749773,
      "grad_norm": 0.08413577824831009,
      "learning_rate": 0.00015225806451612902,
      "loss": 0.0289,
      "step": 264
    },
    {
      "epoch": 0.4857928505957837,
      "grad_norm": 0.1072312667965889,
      "learning_rate": 0.00015207373271889401,
      "loss": 0.0436,
      "step": 265
    },
    {
      "epoch": 0.48762603116406966,
      "grad_norm": 0.09056718647480011,
      "learning_rate": 0.00015188940092165899,
      "loss": 0.0299,
      "step": 266
    },
    {
      "epoch": 0.4894592117323556,
      "grad_norm": 0.07195229828357697,
      "learning_rate": 0.00015170506912442398,
      "loss": 0.0243,
      "step": 267
    },
    {
      "epoch": 0.4912923923006416,
      "grad_norm": 0.09091556072235107,
      "learning_rate": 0.00015152073732718895,
      "loss": 0.0371,
      "step": 268
    },
    {
      "epoch": 0.4931255728689276,
      "grad_norm": 0.08215435594320297,
      "learning_rate": 0.00015133640552995393,
      "loss": 0.0269,
      "step": 269
    },
    {
      "epoch": 0.49495875343721357,
      "grad_norm": 0.07215309143066406,
      "learning_rate": 0.0001511520737327189,
      "loss": 0.0253,
      "step": 270
    },
    {
      "epoch": 0.49679193400549954,
      "grad_norm": 0.08347906917333603,
      "learning_rate": 0.0001509677419354839,
      "loss": 0.0273,
      "step": 271
    },
    {
      "epoch": 0.4986251145737855,
      "grad_norm": 0.061136987060308456,
      "learning_rate": 0.00015078341013824887,
      "loss": 0.0231,
      "step": 272
    },
    {
      "epoch": 0.5004582951420715,
      "grad_norm": 0.10640832036733627,
      "learning_rate": 0.00015059907834101384,
      "loss": 0.0455,
      "step": 273
    },
    {
      "epoch": 0.5022914757103575,
      "grad_norm": 0.08627432584762573,
      "learning_rate": 0.0001504147465437788,
      "loss": 0.0243,
      "step": 274
    },
    {
      "epoch": 0.5041246562786434,
      "grad_norm": 0.08695763349533081,
      "learning_rate": 0.00015023041474654378,
      "loss": 0.034,
      "step": 275
    },
    {
      "epoch": 0.5059578368469294,
      "grad_norm": 0.07717976719141006,
      "learning_rate": 0.00015004608294930875,
      "loss": 0.0275,
      "step": 276
    },
    {
      "epoch": 0.5077910174152154,
      "grad_norm": 0.0763443112373352,
      "learning_rate": 0.00014986175115207375,
      "loss": 0.0274,
      "step": 277
    },
    {
      "epoch": 0.5096241979835013,
      "grad_norm": 0.06165534630417824,
      "learning_rate": 0.00014967741935483872,
      "loss": 0.0234,
      "step": 278
    },
    {
      "epoch": 0.5114573785517873,
      "grad_norm": 0.08029788732528687,
      "learning_rate": 0.0001494930875576037,
      "loss": 0.0272,
      "step": 279
    },
    {
      "epoch": 0.5132905591200734,
      "grad_norm": 0.08326773345470428,
      "learning_rate": 0.00014930875576036866,
      "loss": 0.0262,
      "step": 280
    },
    {
      "epoch": 0.5151237396883593,
      "grad_norm": 0.08208340406417847,
      "learning_rate": 0.00014912442396313363,
      "loss": 0.0264,
      "step": 281
    },
    {
      "epoch": 0.5169569202566453,
      "grad_norm": 0.0779559388756752,
      "learning_rate": 0.0001489400921658986,
      "loss": 0.0211,
      "step": 282
    },
    {
      "epoch": 0.5187901008249313,
      "grad_norm": 0.09281262755393982,
      "learning_rate": 0.0001487557603686636,
      "loss": 0.0346,
      "step": 283
    },
    {
      "epoch": 0.5206232813932172,
      "grad_norm": 0.09531724452972412,
      "learning_rate": 0.00014857142857142857,
      "loss": 0.0374,
      "step": 284
    },
    {
      "epoch": 0.5224564619615032,
      "grad_norm": 0.10953988879919052,
      "learning_rate": 0.00014838709677419355,
      "loss": 0.038,
      "step": 285
    },
    {
      "epoch": 0.5242896425297892,
      "grad_norm": 0.09352127462625504,
      "learning_rate": 0.00014820276497695854,
      "loss": 0.0378,
      "step": 286
    },
    {
      "epoch": 0.5261228230980751,
      "grad_norm": 0.05571591481566429,
      "learning_rate": 0.00014801843317972351,
      "loss": 0.02,
      "step": 287
    },
    {
      "epoch": 0.5279560036663611,
      "grad_norm": 0.09128769487142563,
      "learning_rate": 0.00014783410138248849,
      "loss": 0.0327,
      "step": 288
    },
    {
      "epoch": 0.5297891842346472,
      "grad_norm": 0.09302745014429092,
      "learning_rate": 0.00014764976958525348,
      "loss": 0.029,
      "step": 289
    },
    {
      "epoch": 0.5316223648029331,
      "grad_norm": 0.05895543098449707,
      "learning_rate": 0.00014746543778801845,
      "loss": 0.0203,
      "step": 290
    },
    {
      "epoch": 0.5334555453712191,
      "grad_norm": 0.08418423682451248,
      "learning_rate": 0.00014728110599078343,
      "loss": 0.0329,
      "step": 291
    },
    {
      "epoch": 0.535288725939505,
      "grad_norm": 0.09295564144849777,
      "learning_rate": 0.0001470967741935484,
      "loss": 0.0401,
      "step": 292
    },
    {
      "epoch": 0.537121906507791,
      "grad_norm": 0.08649936318397522,
      "learning_rate": 0.00014691244239631337,
      "loss": 0.0343,
      "step": 293
    },
    {
      "epoch": 0.538955087076077,
      "grad_norm": 0.08355950564146042,
      "learning_rate": 0.00014672811059907834,
      "loss": 0.0315,
      "step": 294
    },
    {
      "epoch": 0.5407882676443629,
      "grad_norm": 0.06531458348035812,
      "learning_rate": 0.00014654377880184334,
      "loss": 0.0217,
      "step": 295
    },
    {
      "epoch": 0.5426214482126489,
      "grad_norm": 0.09049852937459946,
      "learning_rate": 0.0001463594470046083,
      "loss": 0.0298,
      "step": 296
    },
    {
      "epoch": 0.5444546287809349,
      "grad_norm": 0.08500348031520844,
      "learning_rate": 0.00014617511520737328,
      "loss": 0.0253,
      "step": 297
    },
    {
      "epoch": 0.5462878093492209,
      "grad_norm": 0.09162382036447525,
      "learning_rate": 0.00014599078341013825,
      "loss": 0.0299,
      "step": 298
    },
    {
      "epoch": 0.5481209899175069,
      "grad_norm": 0.07611165195703506,
      "learning_rate": 0.00014580645161290322,
      "loss": 0.0256,
      "step": 299
    },
    {
      "epoch": 0.5499541704857929,
      "grad_norm": 0.09361441433429718,
      "learning_rate": 0.0001456221198156682,
      "loss": 0.029,
      "step": 300
    },
    {
      "epoch": 0.5517873510540788,
      "grad_norm": 0.09738872200250626,
      "learning_rate": 0.0001454377880184332,
      "loss": 0.0288,
      "step": 301
    },
    {
      "epoch": 0.5536205316223648,
      "grad_norm": 0.08112243562936783,
      "learning_rate": 0.00014525345622119816,
      "loss": 0.0311,
      "step": 302
    },
    {
      "epoch": 0.5554537121906508,
      "grad_norm": 0.08341687172651291,
      "learning_rate": 0.00014506912442396313,
      "loss": 0.0316,
      "step": 303
    },
    {
      "epoch": 0.5572868927589367,
      "grad_norm": 0.07399066537618637,
      "learning_rate": 0.0001448847926267281,
      "loss": 0.0257,
      "step": 304
    },
    {
      "epoch": 0.5591200733272227,
      "grad_norm": 0.07534675300121307,
      "learning_rate": 0.0001447004608294931,
      "loss": 0.0259,
      "step": 305
    },
    {
      "epoch": 0.5609532538955087,
      "grad_norm": 0.08384109288454056,
      "learning_rate": 0.00014451612903225807,
      "loss": 0.0277,
      "step": 306
    },
    {
      "epoch": 0.5627864344637947,
      "grad_norm": 0.08456786721944809,
      "learning_rate": 0.00014433179723502307,
      "loss": 0.0375,
      "step": 307
    },
    {
      "epoch": 0.5646196150320807,
      "grad_norm": 0.07096688449382782,
      "learning_rate": 0.00014414746543778804,
      "loss": 0.0241,
      "step": 308
    },
    {
      "epoch": 0.5664527956003667,
      "grad_norm": 0.09720040112733841,
      "learning_rate": 0.00014396313364055301,
      "loss": 0.0333,
      "step": 309
    },
    {
      "epoch": 0.5682859761686526,
      "grad_norm": 0.11616980284452438,
      "learning_rate": 0.00014377880184331799,
      "loss": 0.0431,
      "step": 310
    },
    {
      "epoch": 0.5701191567369386,
      "grad_norm": 0.08847475051879883,
      "learning_rate": 0.00014359447004608296,
      "loss": 0.0366,
      "step": 311
    },
    {
      "epoch": 0.5719523373052245,
      "grad_norm": 0.09937264025211334,
      "learning_rate": 0.00014341013824884793,
      "loss": 0.0371,
      "step": 312
    },
    {
      "epoch": 0.5737855178735105,
      "grad_norm": 0.059645283967256546,
      "learning_rate": 0.00014322580645161293,
      "loss": 0.0232,
      "step": 313
    },
    {
      "epoch": 0.5756186984417965,
      "grad_norm": 0.06614042073488235,
      "learning_rate": 0.0001430414746543779,
      "loss": 0.0234,
      "step": 314
    },
    {
      "epoch": 0.5774518790100825,
      "grad_norm": 0.10562101751565933,
      "learning_rate": 0.00014285714285714287,
      "loss": 0.0387,
      "step": 315
    },
    {
      "epoch": 0.5792850595783685,
      "grad_norm": 0.10281278938055038,
      "learning_rate": 0.00014267281105990784,
      "loss": 0.0339,
      "step": 316
    },
    {
      "epoch": 0.5811182401466545,
      "grad_norm": 0.10205813497304916,
      "learning_rate": 0.0001424884792626728,
      "loss": 0.0426,
      "step": 317
    },
    {
      "epoch": 0.5829514207149404,
      "grad_norm": 0.07571630924940109,
      "learning_rate": 0.00014230414746543778,
      "loss": 0.0311,
      "step": 318
    },
    {
      "epoch": 0.5847846012832264,
      "grad_norm": 0.08494407683610916,
      "learning_rate": 0.00014211981566820278,
      "loss": 0.0301,
      "step": 319
    },
    {
      "epoch": 0.5866177818515124,
      "grad_norm": 0.05080355703830719,
      "learning_rate": 0.00014193548387096775,
      "loss": 0.017,
      "step": 320
    },
    {
      "epoch": 0.5884509624197983,
      "grad_norm": 0.08271925151348114,
      "learning_rate": 0.00014175115207373272,
      "loss": 0.0304,
      "step": 321
    },
    {
      "epoch": 0.5902841429880843,
      "grad_norm": 0.0785074308514595,
      "learning_rate": 0.0001415668202764977,
      "loss": 0.0315,
      "step": 322
    },
    {
      "epoch": 0.5921173235563703,
      "grad_norm": 0.08633995056152344,
      "learning_rate": 0.00014138248847926266,
      "loss": 0.0366,
      "step": 323
    },
    {
      "epoch": 0.5939505041246563,
      "grad_norm": 0.07227237522602081,
      "learning_rate": 0.00014119815668202766,
      "loss": 0.0256,
      "step": 324
    },
    {
      "epoch": 0.5957836846929423,
      "grad_norm": 0.07901857793331146,
      "learning_rate": 0.00014101382488479263,
      "loss": 0.0224,
      "step": 325
    },
    {
      "epoch": 0.5976168652612283,
      "grad_norm": 0.0696311965584755,
      "learning_rate": 0.00014082949308755763,
      "loss": 0.0245,
      "step": 326
    },
    {
      "epoch": 0.5994500458295142,
      "grad_norm": 0.0882924422621727,
      "learning_rate": 0.0001406451612903226,
      "loss": 0.0349,
      "step": 327
    },
    {
      "epoch": 0.6012832263978002,
      "grad_norm": 0.07235971093177795,
      "learning_rate": 0.00014046082949308757,
      "loss": 0.0224,
      "step": 328
    },
    {
      "epoch": 0.6031164069660861,
      "grad_norm": 0.059238139539957047,
      "learning_rate": 0.00014027649769585254,
      "loss": 0.0204,
      "step": 329
    },
    {
      "epoch": 0.6049495875343721,
      "grad_norm": 0.07901135087013245,
      "learning_rate": 0.00014009216589861752,
      "loss": 0.0273,
      "step": 330
    },
    {
      "epoch": 0.6067827681026581,
      "grad_norm": 0.07229477912187576,
      "learning_rate": 0.0001399078341013825,
      "loss": 0.029,
      "step": 331
    },
    {
      "epoch": 0.608615948670944,
      "grad_norm": 0.07699091732501984,
      "learning_rate": 0.00013972350230414749,
      "loss": 0.026,
      "step": 332
    },
    {
      "epoch": 0.6104491292392301,
      "grad_norm": 0.08664306253194809,
      "learning_rate": 0.00013953917050691246,
      "loss": 0.032,
      "step": 333
    },
    {
      "epoch": 0.6122823098075161,
      "grad_norm": 0.09364963322877884,
      "learning_rate": 0.00013935483870967743,
      "loss": 0.0395,
      "step": 334
    },
    {
      "epoch": 0.614115490375802,
      "grad_norm": 0.0945357009768486,
      "learning_rate": 0.0001391705069124424,
      "loss": 0.0326,
      "step": 335
    },
    {
      "epoch": 0.615948670944088,
      "grad_norm": 0.08954328298568726,
      "learning_rate": 0.00013898617511520737,
      "loss": 0.0326,
      "step": 336
    },
    {
      "epoch": 0.617781851512374,
      "grad_norm": 0.09588358551263809,
      "learning_rate": 0.00013880184331797234,
      "loss": 0.0335,
      "step": 337
    },
    {
      "epoch": 0.6196150320806599,
      "grad_norm": 0.0669722706079483,
      "learning_rate": 0.00013861751152073734,
      "loss": 0.0217,
      "step": 338
    },
    {
      "epoch": 0.6214482126489459,
      "grad_norm": 0.06408808380365372,
      "learning_rate": 0.0001384331797235023,
      "loss": 0.0211,
      "step": 339
    },
    {
      "epoch": 0.6232813932172319,
      "grad_norm": 0.07464352995157242,
      "learning_rate": 0.00013824884792626728,
      "loss": 0.0314,
      "step": 340
    },
    {
      "epoch": 0.6251145737855178,
      "grad_norm": 0.08413645625114441,
      "learning_rate": 0.00013806451612903225,
      "loss": 0.024,
      "step": 341
    },
    {
      "epoch": 0.6269477543538039,
      "grad_norm": 0.08873338252305984,
      "learning_rate": 0.00013788018433179722,
      "loss": 0.0281,
      "step": 342
    },
    {
      "epoch": 0.6287809349220899,
      "grad_norm": 0.07131095975637436,
      "learning_rate": 0.00013769585253456222,
      "loss": 0.0288,
      "step": 343
    },
    {
      "epoch": 0.6306141154903758,
      "grad_norm": 0.05225027725100517,
      "learning_rate": 0.0001375115207373272,
      "loss": 0.0178,
      "step": 344
    },
    {
      "epoch": 0.6324472960586618,
      "grad_norm": 0.06744900345802307,
      "learning_rate": 0.0001373271889400922,
      "loss": 0.0217,
      "step": 345
    },
    {
      "epoch": 0.6342804766269478,
      "grad_norm": 0.07560716569423676,
      "learning_rate": 0.00013714285714285716,
      "loss": 0.0272,
      "step": 346
    },
    {
      "epoch": 0.6361136571952337,
      "grad_norm": 0.0856630727648735,
      "learning_rate": 0.00013695852534562213,
      "loss": 0.0296,
      "step": 347
    },
    {
      "epoch": 0.6379468377635197,
      "grad_norm": 0.07206695526838303,
      "learning_rate": 0.0001367741935483871,
      "loss": 0.0212,
      "step": 348
    },
    {
      "epoch": 0.6397800183318056,
      "grad_norm": 0.08413973450660706,
      "learning_rate": 0.00013658986175115208,
      "loss": 0.0309,
      "step": 349
    },
    {
      "epoch": 0.6416131989000916,
      "grad_norm": 0.08946281671524048,
      "learning_rate": 0.00013640552995391707,
      "loss": 0.0341,
      "step": 350
    },
    {
      "epoch": 0.6434463794683777,
      "grad_norm": 0.09534583985805511,
      "learning_rate": 0.00013622119815668204,
      "loss": 0.0301,
      "step": 351
    },
    {
      "epoch": 0.6452795600366636,
      "grad_norm": 0.11210379004478455,
      "learning_rate": 0.00013603686635944702,
      "loss": 0.0546,
      "step": 352
    },
    {
      "epoch": 0.6471127406049496,
      "grad_norm": 0.08674897998571396,
      "learning_rate": 0.000135852534562212,
      "loss": 0.034,
      "step": 353
    },
    {
      "epoch": 0.6489459211732356,
      "grad_norm": 0.0715477392077446,
      "learning_rate": 0.00013566820276497696,
      "loss": 0.0294,
      "step": 354
    },
    {
      "epoch": 0.6507791017415215,
      "grad_norm": 0.06651375442743301,
      "learning_rate": 0.00013548387096774193,
      "loss": 0.0246,
      "step": 355
    },
    {
      "epoch": 0.6526122823098075,
      "grad_norm": 0.103563591837883,
      "learning_rate": 0.00013529953917050693,
      "loss": 0.048,
      "step": 356
    },
    {
      "epoch": 0.6544454628780935,
      "grad_norm": 0.11457951366901398,
      "learning_rate": 0.0001351152073732719,
      "loss": 0.0287,
      "step": 357
    },
    {
      "epoch": 0.6562786434463794,
      "grad_norm": 0.060861390084028244,
      "learning_rate": 0.00013493087557603687,
      "loss": 0.025,
      "step": 358
    },
    {
      "epoch": 0.6581118240146654,
      "grad_norm": 0.058961618691682816,
      "learning_rate": 0.00013474654377880184,
      "loss": 0.0208,
      "step": 359
    },
    {
      "epoch": 0.6599450045829515,
      "grad_norm": 0.10199262201786041,
      "learning_rate": 0.0001345622119815668,
      "loss": 0.0319,
      "step": 360
    },
    {
      "epoch": 0.6617781851512374,
      "grad_norm": 0.0710933730006218,
      "learning_rate": 0.00013437788018433178,
      "loss": 0.0245,
      "step": 361
    },
    {
      "epoch": 0.6636113657195234,
      "grad_norm": 0.06901897490024567,
      "learning_rate": 0.00013419354838709678,
      "loss": 0.0189,
      "step": 362
    },
    {
      "epoch": 0.6654445462878094,
      "grad_norm": 0.08602811396121979,
      "learning_rate": 0.00013400921658986175,
      "loss": 0.0344,
      "step": 363
    },
    {
      "epoch": 0.6672777268560953,
      "grad_norm": 0.052524056285619736,
      "learning_rate": 0.00013382488479262675,
      "loss": 0.0211,
      "step": 364
    },
    {
      "epoch": 0.6691109074243813,
      "grad_norm": 0.07917725294828415,
      "learning_rate": 0.00013364055299539172,
      "loss": 0.0285,
      "step": 365
    },
    {
      "epoch": 0.6709440879926672,
      "grad_norm": 0.0799289420247078,
      "learning_rate": 0.0001334562211981567,
      "loss": 0.0296,
      "step": 366
    },
    {
      "epoch": 0.6727772685609532,
      "grad_norm": 0.08307263255119324,
      "learning_rate": 0.00013327188940092166,
      "loss": 0.0259,
      "step": 367
    },
    {
      "epoch": 0.6746104491292392,
      "grad_norm": 0.08490724861621857,
      "learning_rate": 0.00013308755760368666,
      "loss": 0.0317,
      "step": 368
    },
    {
      "epoch": 0.6764436296975253,
      "grad_norm": 0.07663150131702423,
      "learning_rate": 0.00013290322580645163,
      "loss": 0.0279,
      "step": 369
    },
    {
      "epoch": 0.6782768102658112,
      "grad_norm": 0.08253347873687744,
      "learning_rate": 0.0001327188940092166,
      "loss": 0.0323,
      "step": 370
    },
    {
      "epoch": 0.6801099908340972,
      "grad_norm": 0.06597882509231567,
      "learning_rate": 0.00013253456221198157,
      "loss": 0.0275,
      "step": 371
    },
    {
      "epoch": 0.6819431714023831,
      "grad_norm": 0.08167731761932373,
      "learning_rate": 0.00013235023041474655,
      "loss": 0.0256,
      "step": 372
    },
    {
      "epoch": 0.6837763519706691,
      "grad_norm": 0.1000673770904541,
      "learning_rate": 0.00013216589861751152,
      "loss": 0.0428,
      "step": 373
    },
    {
      "epoch": 0.6856095325389551,
      "grad_norm": 0.08605007827281952,
      "learning_rate": 0.00013198156682027652,
      "loss": 0.0365,
      "step": 374
    },
    {
      "epoch": 0.687442713107241,
      "grad_norm": 0.06620384752750397,
      "learning_rate": 0.0001317972350230415,
      "loss": 0.0212,
      "step": 375
    },
    {
      "epoch": 0.689275893675527,
      "grad_norm": 0.08665503561496735,
      "learning_rate": 0.00013161290322580646,
      "loss": 0.0311,
      "step": 376
    },
    {
      "epoch": 0.6911090742438131,
      "grad_norm": 0.06734751164913177,
      "learning_rate": 0.00013142857142857143,
      "loss": 0.0252,
      "step": 377
    },
    {
      "epoch": 0.692942254812099,
      "grad_norm": 0.06160259246826172,
      "learning_rate": 0.0001312442396313364,
      "loss": 0.0238,
      "step": 378
    },
    {
      "epoch": 0.694775435380385,
      "grad_norm": 0.0831260159611702,
      "learning_rate": 0.00013105990783410137,
      "loss": 0.0322,
      "step": 379
    },
    {
      "epoch": 0.696608615948671,
      "grad_norm": 0.07255002856254578,
      "learning_rate": 0.00013087557603686637,
      "loss": 0.0265,
      "step": 380
    },
    {
      "epoch": 0.6984417965169569,
      "grad_norm": 0.08250346034765244,
      "learning_rate": 0.00013069124423963134,
      "loss": 0.031,
      "step": 381
    },
    {
      "epoch": 0.7002749770852429,
      "grad_norm": 0.06267958134412766,
      "learning_rate": 0.0001305069124423963,
      "loss": 0.023,
      "step": 382
    },
    {
      "epoch": 0.7021081576535289,
      "grad_norm": 0.08702743053436279,
      "learning_rate": 0.0001303225806451613,
      "loss": 0.0294,
      "step": 383
    },
    {
      "epoch": 0.7039413382218148,
      "grad_norm": 0.08447282761335373,
      "learning_rate": 0.00013013824884792628,
      "loss": 0.0282,
      "step": 384
    },
    {
      "epoch": 0.7057745187901008,
      "grad_norm": 0.0854048877954483,
      "learning_rate": 0.00012995391705069125,
      "loss": 0.0301,
      "step": 385
    },
    {
      "epoch": 0.7076076993583869,
      "grad_norm": 0.08276678621768951,
      "learning_rate": 0.00012976958525345625,
      "loss": 0.0273,
      "step": 386
    },
    {
      "epoch": 0.7094408799266728,
      "grad_norm": 0.11089053750038147,
      "learning_rate": 0.00012958525345622122,
      "loss": 0.0385,
      "step": 387
    },
    {
      "epoch": 0.7112740604949588,
      "grad_norm": 0.08636972308158875,
      "learning_rate": 0.0001294009216589862,
      "loss": 0.0272,
      "step": 388
    },
    {
      "epoch": 0.7131072410632447,
      "grad_norm": 0.09062766283750534,
      "learning_rate": 0.00012921658986175116,
      "loss": 0.0327,
      "step": 389
    },
    {
      "epoch": 0.7149404216315307,
      "grad_norm": 0.0708693340420723,
      "learning_rate": 0.00012903225806451613,
      "loss": 0.0249,
      "step": 390
    },
    {
      "epoch": 0.7167736021998167,
      "grad_norm": 0.07402048259973526,
      "learning_rate": 0.0001288479262672811,
      "loss": 0.0184,
      "step": 391
    },
    {
      "epoch": 0.7186067827681026,
      "grad_norm": 0.13229697942733765,
      "learning_rate": 0.00012866359447004608,
      "loss": 0.0489,
      "step": 392
    },
    {
      "epoch": 0.7204399633363886,
      "grad_norm": 0.06817866861820221,
      "learning_rate": 0.00012847926267281107,
      "loss": 0.0233,
      "step": 393
    },
    {
      "epoch": 0.7222731439046746,
      "grad_norm": 0.09490000456571579,
      "learning_rate": 0.00012829493087557605,
      "loss": 0.0303,
      "step": 394
    },
    {
      "epoch": 0.7241063244729606,
      "grad_norm": 0.06846782565116882,
      "learning_rate": 0.00012811059907834102,
      "loss": 0.0238,
      "step": 395
    },
    {
      "epoch": 0.7259395050412466,
      "grad_norm": 0.09812495112419128,
      "learning_rate": 0.000127926267281106,
      "loss": 0.0324,
      "step": 396
    },
    {
      "epoch": 0.7277726856095326,
      "grad_norm": 0.08136089891195297,
      "learning_rate": 0.00012774193548387096,
      "loss": 0.0309,
      "step": 397
    },
    {
      "epoch": 0.7296058661778185,
      "grad_norm": 0.07089602202177048,
      "learning_rate": 0.00012755760368663593,
      "loss": 0.0251,
      "step": 398
    },
    {
      "epoch": 0.7314390467461045,
      "grad_norm": 0.10192608833312988,
      "learning_rate": 0.00012737327188940093,
      "loss": 0.0345,
      "step": 399
    },
    {
      "epoch": 0.7332722273143905,
      "grad_norm": 0.07810863107442856,
      "learning_rate": 0.0001271889400921659,
      "loss": 0.0264,
      "step": 400
    },
    {
      "epoch": 0.7351054078826764,
      "grad_norm": 0.0839025229215622,
      "learning_rate": 0.00012700460829493087,
      "loss": 0.0279,
      "step": 401
    },
    {
      "epoch": 0.7369385884509624,
      "grad_norm": 0.08430635184049606,
      "learning_rate": 0.00012682027649769587,
      "loss": 0.0345,
      "step": 402
    },
    {
      "epoch": 0.7387717690192483,
      "grad_norm": 0.05288069695234299,
      "learning_rate": 0.00012663594470046084,
      "loss": 0.0181,
      "step": 403
    },
    {
      "epoch": 0.7406049495875344,
      "grad_norm": 0.06337739527225494,
      "learning_rate": 0.0001264516129032258,
      "loss": 0.024,
      "step": 404
    },
    {
      "epoch": 0.7424381301558204,
      "grad_norm": 0.08434322476387024,
      "learning_rate": 0.0001262672811059908,
      "loss": 0.0277,
      "step": 405
    },
    {
      "epoch": 0.7442713107241063,
      "grad_norm": 0.07922643423080444,
      "learning_rate": 0.00012608294930875578,
      "loss": 0.0277,
      "step": 406
    },
    {
      "epoch": 0.7461044912923923,
      "grad_norm": 0.0888214185833931,
      "learning_rate": 0.00012589861751152075,
      "loss": 0.0405,
      "step": 407
    },
    {
      "epoch": 0.7479376718606783,
      "grad_norm": 0.08394885063171387,
      "learning_rate": 0.00012571428571428572,
      "loss": 0.0268,
      "step": 408
    },
    {
      "epoch": 0.7497708524289642,
      "grad_norm": 0.08741919696331024,
      "learning_rate": 0.0001255299539170507,
      "loss": 0.034,
      "step": 409
    },
    {
      "epoch": 0.7516040329972502,
      "grad_norm": 0.08985461294651031,
      "learning_rate": 0.00012534562211981566,
      "loss": 0.0349,
      "step": 410
    },
    {
      "epoch": 0.7534372135655362,
      "grad_norm": 0.055699512362480164,
      "learning_rate": 0.00012516129032258066,
      "loss": 0.0205,
      "step": 411
    },
    {
      "epoch": 0.7552703941338221,
      "grad_norm": 0.09050854295492172,
      "learning_rate": 0.00012497695852534563,
      "loss": 0.0283,
      "step": 412
    },
    {
      "epoch": 0.7571035747021082,
      "grad_norm": 0.11931566148996353,
      "learning_rate": 0.0001247926267281106,
      "loss": 0.0614,
      "step": 413
    },
    {
      "epoch": 0.7589367552703942,
      "grad_norm": 0.06575947254896164,
      "learning_rate": 0.00012460829493087558,
      "loss": 0.0214,
      "step": 414
    },
    {
      "epoch": 0.7607699358386801,
      "grad_norm": 0.0591006726026535,
      "learning_rate": 0.00012442396313364055,
      "loss": 0.0198,
      "step": 415
    },
    {
      "epoch": 0.7626031164069661,
      "grad_norm": 0.12222932279109955,
      "learning_rate": 0.00012423963133640552,
      "loss": 0.0453,
      "step": 416
    },
    {
      "epoch": 0.7644362969752521,
      "grad_norm": 0.07346849143505096,
      "learning_rate": 0.00012405529953917052,
      "loss": 0.027,
      "step": 417
    },
    {
      "epoch": 0.766269477543538,
      "grad_norm": 0.05644283443689346,
      "learning_rate": 0.0001238709677419355,
      "loss": 0.0225,
      "step": 418
    },
    {
      "epoch": 0.768102658111824,
      "grad_norm": 0.07721511274576187,
      "learning_rate": 0.00012368663594470046,
      "loss": 0.0279,
      "step": 419
    },
    {
      "epoch": 0.76993583868011,
      "grad_norm": 0.08552254736423492,
      "learning_rate": 0.00012350230414746543,
      "loss": 0.0294,
      "step": 420
    },
    {
      "epoch": 0.7717690192483959,
      "grad_norm": 0.06860620528459549,
      "learning_rate": 0.00012331797235023043,
      "loss": 0.0287,
      "step": 421
    },
    {
      "epoch": 0.773602199816682,
      "grad_norm": 0.09986478835344315,
      "learning_rate": 0.0001231336405529954,
      "loss": 0.0396,
      "step": 422
    },
    {
      "epoch": 0.775435380384968,
      "grad_norm": 0.08369968086481094,
      "learning_rate": 0.00012294930875576037,
      "loss": 0.0251,
      "step": 423
    },
    {
      "epoch": 0.7772685609532539,
      "grad_norm": 0.06935148686170578,
      "learning_rate": 0.00012276497695852537,
      "loss": 0.0261,
      "step": 424
    },
    {
      "epoch": 0.7791017415215399,
      "grad_norm": 0.08505766093730927,
      "learning_rate": 0.00012258064516129034,
      "loss": 0.0318,
      "step": 425
    },
    {
      "epoch": 0.7809349220898258,
      "grad_norm": 0.07806456089019775,
      "learning_rate": 0.0001223963133640553,
      "loss": 0.0264,
      "step": 426
    },
    {
      "epoch": 0.7827681026581118,
      "grad_norm": 0.10398893058300018,
      "learning_rate": 0.00012221198156682028,
      "loss": 0.0397,
      "step": 427
    },
    {
      "epoch": 0.7846012832263978,
      "grad_norm": 0.07210515439510345,
      "learning_rate": 0.00012202764976958525,
      "loss": 0.0292,
      "step": 428
    },
    {
      "epoch": 0.7864344637946837,
      "grad_norm": 0.06596438586711884,
      "learning_rate": 0.00012184331797235025,
      "loss": 0.0233,
      "step": 429
    },
    {
      "epoch": 0.7882676443629697,
      "grad_norm": 0.11197281628847122,
      "learning_rate": 0.00012165898617511522,
      "loss": 0.0362,
      "step": 430
    },
    {
      "epoch": 0.7901008249312558,
      "grad_norm": 0.10142118483781815,
      "learning_rate": 0.0001214746543778802,
      "loss": 0.0384,
      "step": 431
    },
    {
      "epoch": 0.7919340054995417,
      "grad_norm": 0.08676854521036148,
      "learning_rate": 0.00012129032258064516,
      "loss": 0.0287,
      "step": 432
    },
    {
      "epoch": 0.7937671860678277,
      "grad_norm": 0.08402638882398605,
      "learning_rate": 0.00012110599078341014,
      "loss": 0.03,
      "step": 433
    },
    {
      "epoch": 0.7956003666361137,
      "grad_norm": 0.07576748728752136,
      "learning_rate": 0.00012092165898617511,
      "loss": 0.027,
      "step": 434
    },
    {
      "epoch": 0.7974335472043996,
      "grad_norm": 0.08877792209386826,
      "learning_rate": 0.0001207373271889401,
      "loss": 0.0294,
      "step": 435
    },
    {
      "epoch": 0.7992667277726856,
      "grad_norm": 0.09211436659097672,
      "learning_rate": 0.00012055299539170508,
      "loss": 0.0306,
      "step": 436
    },
    {
      "epoch": 0.8010999083409716,
      "grad_norm": 0.0953056812286377,
      "learning_rate": 0.00012036866359447006,
      "loss": 0.0343,
      "step": 437
    },
    {
      "epoch": 0.8029330889092575,
      "grad_norm": 0.08642080426216125,
      "learning_rate": 0.00012018433179723503,
      "loss": 0.0343,
      "step": 438
    },
    {
      "epoch": 0.8047662694775435,
      "grad_norm": 0.09826899319887161,
      "learning_rate": 0.00012,
      "loss": 0.0301,
      "step": 439
    },
    {
      "epoch": 0.8065994500458296,
      "grad_norm": 0.09221632778644562,
      "learning_rate": 0.00011981566820276497,
      "loss": 0.0342,
      "step": 440
    },
    {
      "epoch": 0.8084326306141155,
      "grad_norm": 0.091212198138237,
      "learning_rate": 0.00011963133640552997,
      "loss": 0.0388,
      "step": 441
    },
    {
      "epoch": 0.8102658111824015,
      "grad_norm": 0.07262887060642242,
      "learning_rate": 0.00011944700460829494,
      "loss": 0.0269,
      "step": 442
    },
    {
      "epoch": 0.8120989917506874,
      "grad_norm": 0.09429402649402618,
      "learning_rate": 0.00011926267281105991,
      "loss": 0.0344,
      "step": 443
    },
    {
      "epoch": 0.8139321723189734,
      "grad_norm": 0.07602915167808533,
      "learning_rate": 0.00011907834101382489,
      "loss": 0.0229,
      "step": 444
    },
    {
      "epoch": 0.8157653528872594,
      "grad_norm": 0.08622591942548752,
      "learning_rate": 0.00011889400921658986,
      "loss": 0.0299,
      "step": 445
    },
    {
      "epoch": 0.8175985334555453,
      "grad_norm": 0.07046458125114441,
      "learning_rate": 0.00011870967741935484,
      "loss": 0.0283,
      "step": 446
    },
    {
      "epoch": 0.8194317140238313,
      "grad_norm": 0.06315992772579193,
      "learning_rate": 0.00011852534562211983,
      "loss": 0.0231,
      "step": 447
    },
    {
      "epoch": 0.8212648945921174,
      "grad_norm": 0.11382705718278885,
      "learning_rate": 0.00011834101382488481,
      "loss": 0.0292,
      "step": 448
    },
    {
      "epoch": 0.8230980751604033,
      "grad_norm": 0.0543043278157711,
      "learning_rate": 0.00011815668202764978,
      "loss": 0.021,
      "step": 449
    },
    {
      "epoch": 0.8249312557286893,
      "grad_norm": 0.08020134270191193,
      "learning_rate": 0.00011797235023041475,
      "loss": 0.024,
      "step": 450
    },
    {
      "epoch": 0.8267644362969753,
      "grad_norm": 0.08001308888196945,
      "learning_rate": 0.00011778801843317972,
      "loss": 0.026,
      "step": 451
    },
    {
      "epoch": 0.8285976168652612,
      "grad_norm": 0.07345568388700485,
      "learning_rate": 0.0001176036866359447,
      "loss": 0.0252,
      "step": 452
    },
    {
      "epoch": 0.8304307974335472,
      "grad_norm": 0.08047681301832199,
      "learning_rate": 0.00011741935483870967,
      "loss": 0.0382,
      "step": 453
    },
    {
      "epoch": 0.8322639780018332,
      "grad_norm": 0.07006672024726868,
      "learning_rate": 0.00011723502304147466,
      "loss": 0.0282,
      "step": 454
    },
    {
      "epoch": 0.8340971585701191,
      "grad_norm": 0.08989156782627106,
      "learning_rate": 0.00011705069124423964,
      "loss": 0.0358,
      "step": 455
    },
    {
      "epoch": 0.8359303391384051,
      "grad_norm": 0.07250788062810898,
      "learning_rate": 0.00011686635944700462,
      "loss": 0.0282,
      "step": 456
    },
    {
      "epoch": 0.8377635197066912,
      "grad_norm": 0.07556509971618652,
      "learning_rate": 0.00011668202764976959,
      "loss": 0.0285,
      "step": 457
    },
    {
      "epoch": 0.8395967002749771,
      "grad_norm": 0.07886163890361786,
      "learning_rate": 0.00011649769585253456,
      "loss": 0.0245,
      "step": 458
    },
    {
      "epoch": 0.8414298808432631,
      "grad_norm": 0.10710224509239197,
      "learning_rate": 0.00011631336405529953,
      "loss": 0.0448,
      "step": 459
    },
    {
      "epoch": 0.843263061411549,
      "grad_norm": 0.092331163585186,
      "learning_rate": 0.00011612903225806453,
      "loss": 0.0319,
      "step": 460
    },
    {
      "epoch": 0.845096241979835,
      "grad_norm": 0.07957134395837784,
      "learning_rate": 0.0001159447004608295,
      "loss": 0.0308,
      "step": 461
    },
    {
      "epoch": 0.846929422548121,
      "grad_norm": 0.06576257944107056,
      "learning_rate": 0.00011576036866359447,
      "loss": 0.0244,
      "step": 462
    },
    {
      "epoch": 0.8487626031164069,
      "grad_norm": 0.060584548860788345,
      "learning_rate": 0.00011557603686635945,
      "loss": 0.0211,
      "step": 463
    },
    {
      "epoch": 0.8505957836846929,
      "grad_norm": 0.0770929604768753,
      "learning_rate": 0.00011539170506912442,
      "loss": 0.0283,
      "step": 464
    },
    {
      "epoch": 0.8524289642529789,
      "grad_norm": 0.10374728590250015,
      "learning_rate": 0.0001152073732718894,
      "loss": 0.031,
      "step": 465
    },
    {
      "epoch": 0.854262144821265,
      "grad_norm": 0.08729080855846405,
      "learning_rate": 0.00011502304147465439,
      "loss": 0.0292,
      "step": 466
    },
    {
      "epoch": 0.8560953253895509,
      "grad_norm": 0.08986352384090424,
      "learning_rate": 0.00011483870967741937,
      "loss": 0.0416,
      "step": 467
    },
    {
      "epoch": 0.8579285059578369,
      "grad_norm": 0.09187289327383041,
      "learning_rate": 0.00011465437788018434,
      "loss": 0.0379,
      "step": 468
    },
    {
      "epoch": 0.8597616865261228,
      "grad_norm": 0.07304208725690842,
      "learning_rate": 0.00011447004608294931,
      "loss": 0.0233,
      "step": 469
    },
    {
      "epoch": 0.8615948670944088,
      "grad_norm": 0.08240114897489548,
      "learning_rate": 0.00011428571428571428,
      "loss": 0.0231,
      "step": 470
    },
    {
      "epoch": 0.8634280476626948,
      "grad_norm": 0.08332082629203796,
      "learning_rate": 0.00011410138248847925,
      "loss": 0.0256,
      "step": 471
    },
    {
      "epoch": 0.8652612282309807,
      "grad_norm": 0.09715988487005234,
      "learning_rate": 0.00011391705069124425,
      "loss": 0.0389,
      "step": 472
    },
    {
      "epoch": 0.8670944087992667,
      "grad_norm": 0.0702865719795227,
      "learning_rate": 0.00011373271889400922,
      "loss": 0.0241,
      "step": 473
    },
    {
      "epoch": 0.8689275893675527,
      "grad_norm": 0.07741749286651611,
      "learning_rate": 0.0001135483870967742,
      "loss": 0.0294,
      "step": 474
    },
    {
      "epoch": 0.8707607699358387,
      "grad_norm": 0.08457162976264954,
      "learning_rate": 0.00011336405529953918,
      "loss": 0.0273,
      "step": 475
    },
    {
      "epoch": 0.8725939505041247,
      "grad_norm": 0.08684583753347397,
      "learning_rate": 0.00011317972350230415,
      "loss": 0.0301,
      "step": 476
    },
    {
      "epoch": 0.8744271310724107,
      "grad_norm": 0.07238451391458511,
      "learning_rate": 0.00011299539170506912,
      "loss": 0.0299,
      "step": 477
    },
    {
      "epoch": 0.8762603116406966,
      "grad_norm": 0.06936534494161606,
      "learning_rate": 0.00011281105990783412,
      "loss": 0.0255,
      "step": 478
    },
    {
      "epoch": 0.8780934922089826,
      "grad_norm": 0.0781572014093399,
      "learning_rate": 0.00011262672811059909,
      "loss": 0.0272,
      "step": 479
    },
    {
      "epoch": 0.8799266727772685,
      "grad_norm": 0.05346061289310455,
      "learning_rate": 0.00011244239631336406,
      "loss": 0.0186,
      "step": 480
    },
    {
      "epoch": 0.8817598533455545,
      "grad_norm": 0.10047460347414017,
      "learning_rate": 0.00011225806451612903,
      "loss": 0.0261,
      "step": 481
    },
    {
      "epoch": 0.8835930339138405,
      "grad_norm": 0.09505181759595871,
      "learning_rate": 0.000112073732718894,
      "loss": 0.0293,
      "step": 482
    },
    {
      "epoch": 0.8854262144821264,
      "grad_norm": 0.08039534837007523,
      "learning_rate": 0.00011188940092165898,
      "loss": 0.0339,
      "step": 483
    },
    {
      "epoch": 0.8872593950504125,
      "grad_norm": 0.0761064887046814,
      "learning_rate": 0.00011170506912442397,
      "loss": 0.032,
      "step": 484
    },
    {
      "epoch": 0.8890925756186985,
      "grad_norm": 0.0830230712890625,
      "learning_rate": 0.00011152073732718894,
      "loss": 0.0282,
      "step": 485
    },
    {
      "epoch": 0.8909257561869844,
      "grad_norm": 0.0726420059800148,
      "learning_rate": 0.00011133640552995393,
      "loss": 0.0293,
      "step": 486
    },
    {
      "epoch": 0.8927589367552704,
      "grad_norm": 0.07475108653306961,
      "learning_rate": 0.0001111520737327189,
      "loss": 0.0311,
      "step": 487
    },
    {
      "epoch": 0.8945921173235564,
      "grad_norm": 0.07357434183359146,
      "learning_rate": 0.00011096774193548387,
      "loss": 0.0255,
      "step": 488
    },
    {
      "epoch": 0.8964252978918423,
      "grad_norm": 0.09986915439367294,
      "learning_rate": 0.00011078341013824884,
      "loss": 0.0312,
      "step": 489
    },
    {
      "epoch": 0.8982584784601283,
      "grad_norm": 0.10107365250587463,
      "learning_rate": 0.00011059907834101384,
      "loss": 0.037,
      "step": 490
    },
    {
      "epoch": 0.9000916590284143,
      "grad_norm": 0.08120604604482651,
      "learning_rate": 0.00011041474654377881,
      "loss": 0.035,
      "step": 491
    },
    {
      "epoch": 0.9019248395967002,
      "grad_norm": 0.05490780994296074,
      "learning_rate": 0.00011023041474654378,
      "loss": 0.0202,
      "step": 492
    },
    {
      "epoch": 0.9037580201649863,
      "grad_norm": 0.10805200785398483,
      "learning_rate": 0.00011004608294930875,
      "loss": 0.045,
      "step": 493
    },
    {
      "epoch": 0.9055912007332723,
      "grad_norm": 0.05866795405745506,
      "learning_rate": 0.00010986175115207374,
      "loss": 0.0203,
      "step": 494
    },
    {
      "epoch": 0.9074243813015582,
      "grad_norm": 0.09001267701387405,
      "learning_rate": 0.00010967741935483871,
      "loss": 0.0286,
      "step": 495
    },
    {
      "epoch": 0.9092575618698442,
      "grad_norm": 0.0694073960185051,
      "learning_rate": 0.00010949308755760371,
      "loss": 0.0318,
      "step": 496
    },
    {
      "epoch": 0.9110907424381302,
      "grad_norm": 0.06593529880046844,
      "learning_rate": 0.00010930875576036868,
      "loss": 0.0227,
      "step": 497
    },
    {
      "epoch": 0.9129239230064161,
      "grad_norm": 0.06799976527690887,
      "learning_rate": 0.00010912442396313365,
      "loss": 0.024,
      "step": 498
    },
    {
      "epoch": 0.9147571035747021,
      "grad_norm": 0.07797495275735855,
      "learning_rate": 0.00010894009216589862,
      "loss": 0.0268,
      "step": 499
    },
    {
      "epoch": 0.916590284142988,
      "grad_norm": 0.07280347496271133,
      "learning_rate": 0.00010875576036866359,
      "loss": 0.0252,
      "step": 500
    }
  ],
  "logging_steps": 1,
  "max_steps": 1090,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.5010830413889536e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}