{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.948051948051948,
  "eval_steps": 500,
  "global_step": 190,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.025974025974025976,
      "grad_norm": 7.932461880382458,
      "learning_rate": 2.105263157894737e-06,
      "loss": 1.3022,
      "step": 1
    },
    {
      "epoch": 0.05194805194805195,
      "grad_norm": 7.914768259228103,
      "learning_rate": 4.210526315789474e-06,
      "loss": 1.2794,
      "step": 2
    },
    {
      "epoch": 0.07792207792207792,
      "grad_norm": 7.484077281744907,
      "learning_rate": 6.31578947368421e-06,
      "loss": 1.2625,
      "step": 3
    },
    {
      "epoch": 0.1038961038961039,
      "grad_norm": 5.714449349246834,
      "learning_rate": 8.421052631578948e-06,
      "loss": 1.2443,
      "step": 4
    },
    {
      "epoch": 0.12987012987012986,
      "grad_norm": 3.1210181292904724,
      "learning_rate": 1.0526315789473684e-05,
      "loss": 1.1532,
      "step": 5
    },
    {
      "epoch": 0.15584415584415584,
      "grad_norm": 5.253391384757655,
      "learning_rate": 1.263157894736842e-05,
      "loss": 1.0944,
      "step": 6
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 5.463591834581839,
      "learning_rate": 1.4736842105263159e-05,
      "loss": 1.1138,
      "step": 7
    },
    {
      "epoch": 0.2077922077922078,
      "grad_norm": 7.387714938198011,
      "learning_rate": 1.6842105263157896e-05,
      "loss": 1.1008,
      "step": 8
    },
    {
      "epoch": 0.23376623376623376,
      "grad_norm": 6.205307398999116,
      "learning_rate": 1.894736842105263e-05,
      "loss": 1.0759,
      "step": 9
    },
    {
      "epoch": 0.2597402597402597,
      "grad_norm": 4.626405073408399,
      "learning_rate": 2.105263157894737e-05,
      "loss": 1.0411,
      "step": 10
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 3.5062975875923623,
      "learning_rate": 2.3157894736842107e-05,
      "loss": 0.9892,
      "step": 11
    },
    {
      "epoch": 0.3116883116883117,
      "grad_norm": 3.095922849196239,
      "learning_rate": 2.526315789473684e-05,
      "loss": 0.981,
      "step": 12
    },
    {
      "epoch": 0.33766233766233766,
      "grad_norm": 2.620362610628679,
      "learning_rate": 2.7368421052631583e-05,
      "loss": 0.9321,
      "step": 13
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 2.095917606876017,
      "learning_rate": 2.9473684210526317e-05,
      "loss": 0.9431,
      "step": 14
    },
    {
      "epoch": 0.38961038961038963,
      "grad_norm": 1.6957008846140307,
      "learning_rate": 3.157894736842106e-05,
      "loss": 0.937,
      "step": 15
    },
    {
      "epoch": 0.4155844155844156,
      "grad_norm": 1.7112898298596606,
      "learning_rate": 3.368421052631579e-05,
      "loss": 0.9237,
      "step": 16
    },
    {
      "epoch": 0.44155844155844154,
      "grad_norm": 1.7675227036340526,
      "learning_rate": 3.578947368421053e-05,
      "loss": 0.913,
      "step": 17
    },
    {
      "epoch": 0.4675324675324675,
      "grad_norm": 1.6383266641991525,
      "learning_rate": 3.789473684210526e-05,
      "loss": 0.9288,
      "step": 18
    },
    {
      "epoch": 0.4935064935064935,
      "grad_norm": 1.294665470376389,
      "learning_rate": 4e-05,
      "loss": 0.9351,
      "step": 19
    },
    {
      "epoch": 0.5194805194805194,
      "grad_norm": 1.604347276742893,
      "learning_rate": 3.999662483266646e-05,
      "loss": 0.8908,
      "step": 20
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.4355986610502338,
      "learning_rate": 3.9986500469841275e-05,
      "loss": 0.9129,
      "step": 21
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 630.6282393505426,
      "learning_rate": 3.996963032866633e-05,
      "loss": 1.137,
      "step": 22
    },
    {
      "epoch": 0.5974025974025974,
      "grad_norm": 64.18229097051045,
      "learning_rate": 3.994602010309655e-05,
      "loss": 0.9911,
      "step": 23
    },
    {
      "epoch": 0.6233766233766234,
      "grad_norm": 3.1233891543931755,
      "learning_rate": 3.991567776197815e-05,
      "loss": 0.9639,
      "step": 24
    },
    {
      "epoch": 0.6493506493506493,
      "grad_norm": 2.077319867927905,
      "learning_rate": 3.9878613546358996e-05,
      "loss": 0.9394,
      "step": 25
    },
    {
      "epoch": 0.6753246753246753,
      "grad_norm": 6.491527923131102,
      "learning_rate": 3.983483996603205e-05,
      "loss": 0.8881,
      "step": 26
    },
    {
      "epoch": 0.7012987012987013,
      "grad_norm": 39.42852606734072,
      "learning_rate": 3.978437179531316e-05,
      "loss": 0.9145,
      "step": 27
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 3.4103842126583013,
      "learning_rate": 3.972722606805445e-05,
      "loss": 0.9565,
      "step": 28
    },
    {
      "epoch": 0.7532467532467533,
      "grad_norm": 2.603677290738666,
      "learning_rate": 3.9663422071895103e-05,
      "loss": 0.9527,
      "step": 29
    },
    {
      "epoch": 0.7792207792207793,
      "grad_norm": 1.5693962270477466,
      "learning_rate": 3.959298134175148e-05,
      "loss": 0.9261,
      "step": 30
    },
    {
      "epoch": 0.8051948051948052,
      "grad_norm": 1.9966129861478277,
      "learning_rate": 3.9515927652548714e-05,
      "loss": 0.8946,
      "step": 31
    },
    {
      "epoch": 0.8311688311688312,
      "grad_norm": 2.6605968358600744,
      "learning_rate": 3.943228701119628e-05,
      "loss": 0.9076,
      "step": 32
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 2.1460676505845138,
      "learning_rate": 3.934208764781022e-05,
      "loss": 0.8956,
      "step": 33
    },
    {
      "epoch": 0.8831168831168831,
      "grad_norm": 1.5905762959039638,
      "learning_rate": 3.924536000618501e-05,
      "loss": 0.8986,
      "step": 34
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 1.53645783721347,
      "learning_rate": 3.9142136733518285e-05,
      "loss": 0.8862,
      "step": 35
    },
    {
      "epoch": 0.935064935064935,
      "grad_norm": 1.68321757769248,
      "learning_rate": 3.903245266939184e-05,
      "loss": 0.9017,
      "step": 36
    },
    {
      "epoch": 0.961038961038961,
      "grad_norm": 1.30815074965975,
      "learning_rate": 3.8916344834012695e-05,
      "loss": 0.8698,
      "step": 37
    },
    {
      "epoch": 0.987012987012987,
      "grad_norm": 1.165747960130663,
      "learning_rate": 3.879385241571817e-05,
      "loss": 0.8618,
      "step": 38
    },
    {
      "epoch": 1.0162337662337662,
      "grad_norm": 1.182349191382658,
      "learning_rate": 3.866501675774914e-05,
      "loss": 0.7754,
      "step": 39
    },
    {
      "epoch": 1.0422077922077921,
      "grad_norm": 1.1769523292955406,
      "learning_rate": 3.8529881344296037e-05,
      "loss": 0.7007,
      "step": 40
    },
    {
      "epoch": 1.0681818181818181,
      "grad_norm": 1.0627768684949523,
      "learning_rate": 3.8388491785822154e-05,
      "loss": 0.6888,
      "step": 41
    },
    {
      "epoch": 1.094155844155844,
      "grad_norm": 1.0474113890412635,
      "learning_rate": 3.8240895803669415e-05,
      "loss": 0.666,
      "step": 42
    },
    {
      "epoch": 1.12012987012987,
      "grad_norm": 1.1513266436057166,
      "learning_rate": 3.808714321395155e-05,
      "loss": 0.6711,
      "step": 43
    },
    {
      "epoch": 1.146103896103896,
      "grad_norm": 1.0575614450294766,
      "learning_rate": 3.792728591074041e-05,
      "loss": 0.6675,
      "step": 44
    },
    {
      "epoch": 1.172077922077922,
      "grad_norm": 1.0417037432847753,
      "learning_rate": 3.776137784855076e-05,
      "loss": 0.6505,
      "step": 45
    },
    {
      "epoch": 1.198051948051948,
      "grad_norm": 1.0718429071818538,
      "learning_rate": 3.758947502412978e-05,
      "loss": 0.6398,
      "step": 46
    },
    {
      "epoch": 1.224025974025974,
      "grad_norm": 1.035859051149055,
      "learning_rate": 3.741163545755725e-05,
      "loss": 0.6344,
      "step": 47
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.2059393425085077,
      "learning_rate": 3.722791917266273e-05,
      "loss": 0.6576,
      "step": 48
    },
    {
      "epoch": 1.275974025974026,
      "grad_norm": 0.8647464266057819,
      "learning_rate": 3.703838817676654e-05,
      "loss": 0.634,
      "step": 49
    },
    {
      "epoch": 1.301948051948052,
      "grad_norm": 0.9794755119362144,
      "learning_rate": 3.684310643975132e-05,
      "loss": 0.6587,
      "step": 50
    },
    {
      "epoch": 1.327922077922078,
      "grad_norm": 0.8945933571618422,
      "learning_rate": 3.6642139872471006e-05,
      "loss": 0.6489,
      "step": 51
    },
    {
      "epoch": 1.353896103896104,
      "grad_norm": 0.7218628490321455,
      "learning_rate": 3.64355563045049e-05,
      "loss": 0.6499,
      "step": 52
    },
    {
      "epoch": 1.37987012987013,
      "grad_norm": 0.9164999911062107,
      "learning_rate": 3.622342546126405e-05,
      "loss": 0.6396,
      "step": 53
    },
    {
      "epoch": 1.405844155844156,
      "grad_norm": 0.8475563662080481,
      "learning_rate": 3.600581894045768e-05,
      "loss": 0.6728,
      "step": 54
    },
    {
      "epoch": 1.4318181818181819,
      "grad_norm": 0.9702631527667891,
      "learning_rate": 3.578281018792788e-05,
      "loss": 0.6644,
      "step": 55
    },
    {
      "epoch": 1.4577922077922079,
      "grad_norm": 0.9538702840059207,
      "learning_rate": 3.555447447286028e-05,
      "loss": 0.643,
      "step": 56
    },
    {
      "epoch": 1.4837662337662338,
      "grad_norm": 0.8910634080576203,
      "learning_rate": 3.532088886237956e-05,
      "loss": 0.6628,
      "step": 57
    },
    {
      "epoch": 1.5097402597402598,
      "grad_norm": 0.9694684889310542,
      "learning_rate": 3.508213219553793e-05,
      "loss": 0.6267,
      "step": 58
    },
    {
      "epoch": 1.5357142857142856,
      "grad_norm": 0.7645363444588047,
      "learning_rate": 3.483828505670563e-05,
      "loss": 0.6621,
      "step": 59
    },
    {
      "epoch": 1.5616883116883118,
      "grad_norm": 1.0003412015299207,
      "learning_rate": 3.458942974837242e-05,
      "loss": 0.6523,
      "step": 60
    },
    {
      "epoch": 1.5876623376623376,
      "grad_norm": 0.9130736095702605,
      "learning_rate": 3.433565026336903e-05,
      "loss": 0.6607,
      "step": 61
    },
    {
      "epoch": 1.6136363636363638,
      "grad_norm": 0.9845762925311329,
      "learning_rate": 3.4077032256518236e-05,
      "loss": 0.6396,
      "step": 62
    },
    {
      "epoch": 1.6396103896103895,
      "grad_norm": 0.8116185410937748,
      "learning_rate": 3.381366301572489e-05,
      "loss": 0.6413,
      "step": 63
    },
    {
      "epoch": 1.6655844155844157,
      "grad_norm": 0.86437358110291,
      "learning_rate": 3.354563143251483e-05,
      "loss": 0.6711,
      "step": 64
    },
    {
      "epoch": 1.6915584415584415,
      "grad_norm": 0.8420048228007022,
      "learning_rate": 3.327302797203243e-05,
      "loss": 0.6604,
      "step": 65
    },
    {
      "epoch": 1.7175324675324677,
      "grad_norm": 0.7816899345223021,
      "learning_rate": 3.29959446425072e-05,
      "loss": 0.651,
      "step": 66
    },
    {
      "epoch": 1.7435064935064934,
      "grad_norm": 0.9303056765525075,
      "learning_rate": 3.2714474964199365e-05,
      "loss": 0.6289,
      "step": 67
    },
    {
      "epoch": 1.7694805194805194,
      "grad_norm": 0.7805801403650556,
      "learning_rate": 3.24287139378353e-05,
      "loss": 0.6411,
      "step": 68
    },
    {
      "epoch": 1.7954545454545454,
      "grad_norm": 0.753331126461648,
      "learning_rate": 3.213875801254314e-05,
      "loss": 0.6646,
      "step": 69
    },
    {
      "epoch": 1.8214285714285714,
      "grad_norm": 0.7358002312013492,
      "learning_rate": 3.1844705053299606e-05,
      "loss": 0.6293,
      "step": 70
    },
    {
      "epoch": 1.8474025974025974,
      "grad_norm": 0.8070564650099749,
      "learning_rate": 3.154665430789893e-05,
      "loss": 0.6596,
      "step": 71
    },
    {
      "epoch": 1.8733766233766234,
      "grad_norm": 0.7226075974955157,
      "learning_rate": 3.1244706373455084e-05,
      "loss": 0.667,
      "step": 72
    },
    {
      "epoch": 1.8993506493506493,
      "grad_norm": 0.7791333934794239,
      "learning_rate": 3.093896316244855e-05,
      "loss": 0.6465,
      "step": 73
    },
    {
      "epoch": 1.9253246753246753,
      "grad_norm": 0.7754676549638453,
      "learning_rate": 3.062952786832912e-05,
      "loss": 0.6554,
      "step": 74
    },
    {
      "epoch": 1.9512987012987013,
      "grad_norm": 0.6557877212038498,
      "learning_rate": 3.0316504930686485e-05,
      "loss": 0.6566,
      "step": 75
    },
    {
      "epoch": 1.9772727272727273,
      "grad_norm": 0.8194426729485009,
      "learning_rate": 3.0000000000000004e-05,
      "loss": 0.6627,
      "step": 76
    },
    {
      "epoch": 2.0064935064935066,
      "grad_norm": 0.7334083955009458,
      "learning_rate": 2.9680119901979984e-05,
      "loss": 0.6111,
      "step": 77
    },
    {
      "epoch": 2.0324675324675323,
      "grad_norm": 1.1895868881255909,
      "learning_rate": 2.935697260151216e-05,
      "loss": 0.3871,
      "step": 78
    },
    {
      "epoch": 2.0584415584415585,
      "grad_norm": 0.9604279263135902,
      "learning_rate": 2.903066716621779e-05,
      "loss": 0.4077,
      "step": 79
    },
    {
      "epoch": 2.0844155844155843,
      "grad_norm": 1.1393762686141848,
      "learning_rate": 2.8701313729641467e-05,
      "loss": 0.3866,
      "step": 80
    },
    {
      "epoch": 2.1103896103896105,
      "grad_norm": 0.7106851428516716,
      "learning_rate": 2.8369023454079223e-05,
      "loss": 0.3825,
      "step": 81
    },
    {
      "epoch": 2.1363636363636362,
      "grad_norm": 0.8224703275776866,
      "learning_rate": 2.8033908493059394e-05,
      "loss": 0.3714,
      "step": 82
    },
    {
      "epoch": 2.1623376623376624,
      "grad_norm": 0.6836375654772662,
      "learning_rate": 2.7696081953488917e-05,
      "loss": 0.3888,
      "step": 83
    },
    {
      "epoch": 2.188311688311688,
      "grad_norm": 0.7313622114325852,
      "learning_rate": 2.735565785747787e-05,
      "loss": 0.3642,
      "step": 84
    },
    {
      "epoch": 2.2142857142857144,
      "grad_norm": 0.6822815355692488,
      "learning_rate": 2.7012751103855092e-05,
      "loss": 0.3783,
      "step": 85
    },
    {
      "epoch": 2.24025974025974,
      "grad_norm": 0.6553172875405961,
      "learning_rate": 2.6667477429387915e-05,
      "loss": 0.3754,
      "step": 86
    },
    {
      "epoch": 2.2662337662337664,
      "grad_norm": 0.6074920244076911,
      "learning_rate": 2.6319953369719057e-05,
      "loss": 0.3473,
      "step": 87
    },
    {
      "epoch": 2.292207792207792,
      "grad_norm": 0.6021128235097986,
      "learning_rate": 2.5970296220033894e-05,
      "loss": 0.3471,
      "step": 88
    },
    {
      "epoch": 2.3181818181818183,
      "grad_norm": 0.6940589972604451,
      "learning_rate": 2.5618623995471394e-05,
      "loss": 0.3837,
      "step": 89
    },
    {
      "epoch": 2.344155844155844,
      "grad_norm": 0.6033654970817208,
      "learning_rate": 2.5265055391291986e-05,
      "loss": 0.3528,
      "step": 90
    },
    {
      "epoch": 2.3701298701298703,
      "grad_norm": 0.6333023948866248,
      "learning_rate": 2.4909709742815986e-05,
      "loss": 0.3649,
      "step": 91
    },
    {
      "epoch": 2.396103896103896,
      "grad_norm": 0.6001340560802214,
      "learning_rate": 2.4552706985145873e-05,
      "loss": 0.3669,
      "step": 92
    },
    {
      "epoch": 2.4220779220779223,
      "grad_norm": 0.6197546210826222,
      "learning_rate": 2.4194167612686208e-05,
      "loss": 0.3532,
      "step": 93
    },
    {
      "epoch": 2.448051948051948,
      "grad_norm": 0.565080044721377,
      "learning_rate": 2.3834212638474773e-05,
      "loss": 0.3657,
      "step": 94
    },
    {
      "epoch": 2.474025974025974,
      "grad_norm": 0.5902739460347666,
      "learning_rate": 2.3472963553338614e-05,
      "loss": 0.3621,
      "step": 95
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.5749686649380882,
      "learning_rate": 2.3110542284888866e-05,
      "loss": 0.3467,
      "step": 96
    },
    {
      "epoch": 2.525974025974026,
      "grad_norm": 0.5450163127235501,
      "learning_rate": 2.2747071156368166e-05,
      "loss": 0.3624,
      "step": 97
    },
    {
      "epoch": 2.551948051948052,
      "grad_norm": 0.6232436601785027,
      "learning_rate": 2.2382672845364474e-05,
      "loss": 0.3564,
      "step": 98
    },
    {
      "epoch": 2.5779220779220777,
      "grad_norm": 0.5515744279861114,
      "learning_rate": 2.201747034240537e-05,
      "loss": 0.3658,
      "step": 99
    },
    {
      "epoch": 2.603896103896104,
      "grad_norm": 0.5362907375585828,
      "learning_rate": 2.165158690944665e-05,
      "loss": 0.3473,
      "step": 100
    },
    {
      "epoch": 2.62987012987013,
      "grad_norm": 0.6161976462395182,
      "learning_rate": 2.1285146038269406e-05,
      "loss": 0.3416,
      "step": 101
    },
    {
      "epoch": 2.655844155844156,
      "grad_norm": 0.5199919758117065,
      "learning_rate": 2.091827140879944e-05,
      "loss": 0.3448,
      "step": 102
    },
    {
      "epoch": 2.6818181818181817,
      "grad_norm": 0.5858886775846843,
      "learning_rate": 2.0551086847363245e-05,
      "loss": 0.3433,
      "step": 103
    },
    {
      "epoch": 2.707792207792208,
      "grad_norm": 0.5676263304852041,
      "learning_rate": 2.0183716284894533e-05,
      "loss": 0.3467,
      "step": 104
    },
    {
      "epoch": 2.7337662337662336,
      "grad_norm": 0.5586462671917939,
      "learning_rate": 1.9816283715105474e-05,
      "loss": 0.3482,
      "step": 105
    },
    {
      "epoch": 2.75974025974026,
      "grad_norm": 0.6276307789744222,
      "learning_rate": 1.9448913152636765e-05,
      "loss": 0.3545,
      "step": 106
    },
    {
      "epoch": 2.7857142857142856,
      "grad_norm": 0.5484342032145305,
      "learning_rate": 1.9081728591200565e-05,
      "loss": 0.3666,
      "step": 107
    },
    {
      "epoch": 2.811688311688312,
      "grad_norm": 0.5688020278596353,
      "learning_rate": 1.87148539617306e-05,
      "loss": 0.344,
      "step": 108
    },
    {
      "epoch": 2.8376623376623376,
      "grad_norm": 0.539614144085222,
      "learning_rate": 1.8348413090553356e-05,
      "loss": 0.3387,
      "step": 109
    },
    {
      "epoch": 2.8636363636363638,
      "grad_norm": 0.5698805550812537,
      "learning_rate": 1.7982529657594637e-05,
      "loss": 0.3604,
      "step": 110
    },
    {
      "epoch": 2.8896103896103895,
      "grad_norm": 0.5212709930299778,
      "learning_rate": 1.761732715463553e-05,
      "loss": 0.3363,
      "step": 111
    },
    {
      "epoch": 2.9155844155844157,
      "grad_norm": 0.5284822211250491,
      "learning_rate": 1.7252928843631838e-05,
      "loss": 0.3546,
      "step": 112
    },
    {
      "epoch": 2.9415584415584415,
      "grad_norm": 0.5401001882166652,
      "learning_rate": 1.6889457715111144e-05,
      "loss": 0.3486,
      "step": 113
    },
    {
      "epoch": 2.9675324675324677,
      "grad_norm": 0.5229863215386347,
      "learning_rate": 1.6527036446661396e-05,
      "loss": 0.3352,
      "step": 114
    },
    {
      "epoch": 2.9935064935064934,
      "grad_norm": 0.5347563548020706,
      "learning_rate": 1.6165787361525237e-05,
      "loss": 0.3593,
      "step": 115
    },
    {
      "epoch": 3.022727272727273,
      "grad_norm": 0.672978181037337,
      "learning_rate": 1.5805832387313795e-05,
      "loss": 0.2407,
      "step": 116
    },
    {
      "epoch": 3.0487012987012987,
      "grad_norm": 0.679162551176236,
      "learning_rate": 1.544729301485414e-05,
      "loss": 0.1911,
      "step": 117
    },
    {
      "epoch": 3.074675324675325,
      "grad_norm": 0.48652138994921784,
      "learning_rate": 1.5090290257184019e-05,
      "loss": 0.1896,
      "step": 118
    },
    {
      "epoch": 3.1006493506493507,
      "grad_norm": 0.5783075834024586,
      "learning_rate": 1.4734944608708022e-05,
      "loss": 0.2093,
      "step": 119
    },
    {
      "epoch": 3.1266233766233764,
      "grad_norm": 0.8020276248905834,
      "learning_rate": 1.4381376004528616e-05,
      "loss": 0.2057,
      "step": 120
    },
    {
      "epoch": 3.1525974025974026,
      "grad_norm": 0.667374545311432,
      "learning_rate": 1.4029703779966116e-05,
      "loss": 0.1884,
      "step": 121
    },
    {
      "epoch": 3.1785714285714284,
      "grad_norm": 0.5056105500696159,
      "learning_rate": 1.3680046630280952e-05,
      "loss": 0.1977,
      "step": 122
    },
    {
      "epoch": 3.2045454545454546,
      "grad_norm": 0.49060521809385876,
      "learning_rate": 1.3332522570612097e-05,
      "loss": 0.1891,
      "step": 123
    },
    {
      "epoch": 3.2305194805194803,
      "grad_norm": 0.5481519220939601,
      "learning_rate": 1.2987248896144915e-05,
      "loss": 0.1912,
      "step": 124
    },
    {
      "epoch": 3.2564935064935066,
      "grad_norm": 0.5257555666517691,
      "learning_rate": 1.2644342142522142e-05,
      "loss": 0.1982,
      "step": 125
    },
    {
      "epoch": 3.2824675324675323,
      "grad_norm": 0.48753413095850306,
      "learning_rate": 1.230391804651109e-05,
      "loss": 0.1825,
      "step": 126
    },
    {
      "epoch": 3.3084415584415585,
      "grad_norm": 0.4612445791577537,
      "learning_rate": 1.1966091506940616e-05,
      "loss": 0.1963,
      "step": 127
    },
    {
      "epoch": 3.3344155844155843,
      "grad_norm": 0.44552190602064307,
      "learning_rate": 1.1630976545920777e-05,
      "loss": 0.1706,
      "step": 128
    },
    {
      "epoch": 3.3603896103896105,
      "grad_norm": 0.48659675686631854,
      "learning_rate": 1.1298686270358542e-05,
      "loss": 0.1913,
      "step": 129
    },
    {
      "epoch": 3.3863636363636362,
      "grad_norm": 0.4764396048263091,
      "learning_rate": 1.0969332833782217e-05,
      "loss": 0.1671,
      "step": 130
    },
    {
      "epoch": 3.4123376623376624,
      "grad_norm": 0.4740155487267979,
      "learning_rate": 1.0643027398487848e-05,
      "loss": 0.1854,
      "step": 131
    },
    {
      "epoch": 3.438311688311688,
      "grad_norm": 0.4551923057514232,
      "learning_rate": 1.031988009802003e-05,
      "loss": 0.1709,
      "step": 132
    },
    {
      "epoch": 3.4642857142857144,
      "grad_norm": 0.43070005789183097,
      "learning_rate": 1.0000000000000006e-05,
      "loss": 0.1975,
      "step": 133
    },
    {
      "epoch": 3.49025974025974,
      "grad_norm": 0.383014073709059,
      "learning_rate": 9.683495069313527e-06,
      "loss": 0.1787,
      "step": 134
    },
    {
      "epoch": 3.5162337662337664,
      "grad_norm": 0.4286879659999333,
      "learning_rate": 9.370472131670887e-06,
      "loss": 0.1979,
      "step": 135
    },
    {
      "epoch": 3.542207792207792,
      "grad_norm": 0.41592759389422845,
      "learning_rate": 9.061036837551467e-06,
      "loss": 0.1672,
      "step": 136
    },
    {
      "epoch": 3.5681818181818183,
      "grad_norm": 0.4101654305939979,
      "learning_rate": 8.755293626544921e-06,
      "loss": 0.1578,
      "step": 137
    },
    {
      "epoch": 3.594155844155844,
      "grad_norm": 0.41025853609303076,
      "learning_rate": 8.453345692101076e-06,
      "loss": 0.1732,
      "step": 138
    },
    {
      "epoch": 3.62012987012987,
      "grad_norm": 0.4099487747185865,
      "learning_rate": 8.155294946700402e-06,
      "loss": 0.1868,
      "step": 139
    },
    {
      "epoch": 3.646103896103896,
      "grad_norm": 0.39174988018880874,
      "learning_rate": 7.861241987456869e-06,
      "loss": 0.1921,
      "step": 140
    },
    {
      "epoch": 3.6720779220779223,
      "grad_norm": 0.3960964308201075,
      "learning_rate": 7.571286062164709e-06,
      "loss": 0.1902,
      "step": 141
    },
    {
      "epoch": 3.698051948051948,
      "grad_norm": 0.41695917345370964,
      "learning_rate": 7.285525035800645e-06,
      "loss": 0.1885,
      "step": 142
    },
    {
      "epoch": 3.724025974025974,
      "grad_norm": 0.3946638754267475,
      "learning_rate": 7.0040553574928115e-06,
      "loss": 0.1646,
      "step": 143
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.3843896631591302,
      "learning_rate": 6.7269720279675755e-06,
      "loss": 0.1769,
      "step": 144
    },
    {
      "epoch": 3.775974025974026,
      "grad_norm": 0.39754823944143786,
      "learning_rate": 6.4543685674851834e-06,
      "loss": 0.1927,
      "step": 145
    },
    {
      "epoch": 3.801948051948052,
      "grad_norm": 0.39430307748603793,
      "learning_rate": 6.1863369842751145e-06,
      "loss": 0.1835,
      "step": 146
    },
    {
      "epoch": 3.8279220779220777,
      "grad_norm": 0.3925790082334243,
      "learning_rate": 5.922967743481773e-06,
      "loss": 0.1912,
      "step": 147
    },
    {
      "epoch": 3.853896103896104,
      "grad_norm": 0.4061383035436692,
      "learning_rate": 5.664349736630979e-06,
      "loss": 0.1688,
      "step": 148
    },
    {
      "epoch": 3.87987012987013,
      "grad_norm": 0.3887498420237992,
      "learning_rate": 5.410570251627587e-06,
      "loss": 0.202,
      "step": 149
    },
    {
      "epoch": 3.905844155844156,
      "grad_norm": 0.38629968407857657,
      "learning_rate": 5.161714943294372e-06,
      "loss": 0.1841,
      "step": 150
    },
    {
      "epoch": 3.9318181818181817,
      "grad_norm": 0.38720595837879046,
      "learning_rate": 4.917867804462077e-06,
      "loss": 0.1819,
      "step": 151
    },
    {
      "epoch": 3.957792207792208,
      "grad_norm": 0.38121444486474165,
      "learning_rate": 4.679111137620442e-06,
      "loss": 0.187,
      "step": 152
    },
    {
      "epoch": 3.9837662337662336,
      "grad_norm": 0.3830372856152854,
      "learning_rate": 4.445525527139725e-06,
      "loss": 0.187,
      "step": 153
    },
    {
      "epoch": 4.012987012987013,
      "grad_norm": 0.3791401299823691,
      "learning_rate": 4.217189812072131e-06,
      "loss": 0.1686,
      "step": 154
    },
    {
      "epoch": 4.038961038961039,
      "grad_norm": 0.4148718569874963,
      "learning_rate": 3.994181059542321e-06,
      "loss": 0.1176,
      "step": 155
    },
    {
      "epoch": 4.064935064935065,
      "grad_norm": 0.35673107519822433,
      "learning_rate": 3.7765745387359574e-06,
      "loss": 0.1154,
      "step": 156
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.3402341389305598,
      "learning_rate": 3.564443695495099e-06,
      "loss": 0.1369,
      "step": 157
    },
    {
      "epoch": 4.116883116883117,
      "grad_norm": 0.33628395582708454,
      "learning_rate": 3.357860127529e-06,
      "loss": 0.1167,
      "step": 158
    },
    {
      "epoch": 4.142857142857143,
      "grad_norm": 0.30102455569288517,
      "learning_rate": 3.156893560248688e-06,
      "loss": 0.1242,
      "step": 159
    },
    {
      "epoch": 4.1688311688311686,
      "grad_norm": 0.2771068438895251,
      "learning_rate": 2.9616118232334613e-06,
      "loss": 0.1187,
      "step": 160
    },
    {
      "epoch": 4.194805194805195,
      "grad_norm": 0.294541202996176,
      "learning_rate": 2.7720808273372823e-06,
      "loss": 0.1143,
      "step": 161
    },
    {
      "epoch": 4.220779220779221,
      "grad_norm": 0.3243600487046826,
      "learning_rate": 2.588364542442754e-06,
      "loss": 0.1267,
      "step": 162
    },
    {
      "epoch": 4.246753246753247,
      "grad_norm": 0.3313486959659834,
      "learning_rate": 2.410524975870221e-06,
      "loss": 0.127,
      "step": 163
    },
    {
      "epoch": 4.2727272727272725,
      "grad_norm": 0.35233311448800625,
      "learning_rate": 2.2386221514492502e-06,
      "loss": 0.132,
      "step": 164
    },
    {
      "epoch": 4.298701298701299,
      "grad_norm": 0.34772021499799033,
      "learning_rate": 2.0727140892595998e-06,
      "loss": 0.1249,
      "step": 165
    },
    {
      "epoch": 4.324675324675325,
      "grad_norm": 0.31722684001225576,
      "learning_rate": 1.9128567860484516e-06,
      "loss": 0.1297,
      "step": 166
    },
    {
      "epoch": 4.35064935064935,
      "grad_norm": 0.31190729931884087,
      "learning_rate": 1.759104196330592e-06,
      "loss": 0.1163,
      "step": 167
    },
    {
      "epoch": 4.376623376623376,
      "grad_norm": 0.29649306847713347,
      "learning_rate": 1.6115082141778459e-06,
      "loss": 0.1219,
      "step": 168
    },
    {
      "epoch": 4.402597402597403,
      "grad_norm": 0.28001769998158305,
      "learning_rate": 1.4701186557039648e-06,
      "loss": 0.1096,
      "step": 169
    },
    {
      "epoch": 4.428571428571429,
      "grad_norm": 0.2810796360606971,
      "learning_rate": 1.334983242250858e-06,
      "loss": 0.1097,
      "step": 170
    },
    {
      "epoch": 4.454545454545454,
      "grad_norm": 0.2833719049954468,
      "learning_rate": 1.2061475842818337e-06,
      "loss": 0.1279,
      "step": 171
    },
    {
      "epoch": 4.48051948051948,
      "grad_norm": 0.2634347576438535,
      "learning_rate": 1.0836551659873073e-06,
      "loss": 0.123,
      "step": 172
    },
    {
      "epoch": 4.5064935064935066,
      "grad_norm": 0.2792887436301362,
      "learning_rate": 9.67547330608165e-07,
      "loss": 0.1189,
      "step": 173
    },
    {
      "epoch": 4.532467532467533,
      "grad_norm": 0.2615347102434456,
      "learning_rate": 8.578632664817177e-07,
      "loss": 0.1249,
      "step": 174
    },
    {
      "epoch": 4.558441558441558,
      "grad_norm": 0.2723001428445289,
      "learning_rate": 7.546399938149918e-07,
      "loss": 0.1137,
      "step": 175
    },
    {
      "epoch": 4.584415584415584,
      "grad_norm": 0.2612744190545831,
      "learning_rate": 6.579123521897867e-07,
      "loss": 0.1015,
      "step": 176
    },
    {
      "epoch": 4.6103896103896105,
      "grad_norm": 0.2689642699590416,
      "learning_rate": 5.677129888037236e-07,
      "loss": 0.1126,
      "step": 177
    },
    {
      "epoch": 4.636363636363637,
      "grad_norm": 0.269452974192897,
      "learning_rate": 4.840723474512876e-07,
      "loss": 0.123,
      "step": 178
    },
    {
      "epoch": 4.662337662337662,
      "grad_norm": 0.26461030349016573,
      "learning_rate": 4.070186582485214e-07,
      "loss": 0.1227,
      "step": 179
    },
    {
      "epoch": 4.688311688311688,
      "grad_norm": 0.26195993538778606,
      "learning_rate": 3.3657792810489975e-07,
      "loss": 0.1265,
      "step": 180
    },
    {
      "epoch": 4.714285714285714,
      "grad_norm": 0.2629718231282119,
      "learning_rate": 2.7277393194555357e-07,
      "loss": 0.0993,
      "step": 181
    },
    {
      "epoch": 4.740259740259741,
      "grad_norm": 0.2748041688236983,
      "learning_rate": 2.1562820468684187e-07,
      "loss": 0.1271,
      "step": 182
    },
    {
      "epoch": 4.766233766233766,
      "grad_norm": 0.28368676223609396,
      "learning_rate": 1.6516003396795489e-07,
      "loss": 0.132,
      "step": 183
    },
    {
      "epoch": 4.792207792207792,
      "grad_norm": 0.2691168035289099,
      "learning_rate": 1.2138645364101032e-07,
      "loss": 0.099,
      "step": 184
    },
    {
      "epoch": 4.818181818181818,
      "grad_norm": 0.26074215375130616,
      "learning_rate": 8.432223802185002e-08,
      "loss": 0.1111,
      "step": 185
    },
    {
      "epoch": 4.8441558441558445,
      "grad_norm": 0.27028450763035056,
      "learning_rate": 5.3979896903453287e-08,
      "loss": 0.1229,
      "step": 186
    },
    {
      "epoch": 4.87012987012987,
      "grad_norm": 0.27411525111121277,
      "learning_rate": 3.036967133367652e-08,
      "loss": 0.1303,
      "step": 187
    },
    {
      "epoch": 4.896103896103896,
      "grad_norm": 0.25069357083928406,
      "learning_rate": 1.349953015872707e-08,
      "loss": 0.1212,
      "step": 188
    },
    {
      "epoch": 4.922077922077922,
      "grad_norm": 0.2630680904345287,
      "learning_rate": 3.3751673335458147e-09,
      "loss": 0.1268,
      "step": 189
    },
    {
      "epoch": 4.948051948051948,
      "grad_norm": 0.2705412617221253,
      "learning_rate": 0.0,
      "loss": 0.1421,
      "step": 190
    },
    {
      "epoch": 4.948051948051948,
      "step": 190,
      "total_flos": 1.8291450201715507e+17,
      "train_loss": 0.4666951280283301,
      "train_runtime": 4594.7358,
      "train_samples_per_second": 5.357,
      "train_steps_per_second": 0.041
    }
  ],
  "logging_steps": 1,
  "max_steps": 190,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 1.8291450201715507e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}