{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.993258426966292,
  "eval_steps": 500,
  "global_step": 333,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008988764044943821,
      "grad_norm": 5.937067934578121,
      "learning_rate": 2.3529411764705885e-06,
      "loss": 0.8925,
      "step": 1
    },
    {
      "epoch": 0.017977528089887642,
      "grad_norm": 5.923591025846198,
      "learning_rate": 4.705882352941177e-06,
      "loss": 0.8908,
      "step": 2
    },
    {
      "epoch": 0.02696629213483146,
      "grad_norm": 5.45957942819648,
      "learning_rate": 7.058823529411766e-06,
      "loss": 0.8716,
      "step": 3
    },
    {
      "epoch": 0.035955056179775284,
      "grad_norm": 3.974971561642825,
      "learning_rate": 9.411764705882354e-06,
      "loss": 0.8395,
      "step": 4
    },
    {
      "epoch": 0.0449438202247191,
      "grad_norm": 2.1772095757152767,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 0.7954,
      "step": 5
    },
    {
      "epoch": 0.05393258426966292,
      "grad_norm": 5.329495741252824,
      "learning_rate": 1.4117647058823532e-05,
      "loss": 0.8391,
      "step": 6
    },
    {
      "epoch": 0.06292134831460675,
      "grad_norm": 7.0082332084809424,
      "learning_rate": 1.647058823529412e-05,
      "loss": 0.8077,
      "step": 7
    },
    {
      "epoch": 0.07191011235955057,
      "grad_norm": 7.179904932078677,
      "learning_rate": 1.8823529411764708e-05,
      "loss": 0.8269,
      "step": 8
    },
    {
      "epoch": 0.08089887640449438,
      "grad_norm": 3.944730501336432,
      "learning_rate": 2.1176470588235296e-05,
      "loss": 0.7655,
      "step": 9
    },
    {
      "epoch": 0.0898876404494382,
      "grad_norm": 3.094924202191644,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 0.7356,
      "step": 10
    },
    {
      "epoch": 0.09887640449438202,
      "grad_norm": 2.1418931101275476,
      "learning_rate": 2.5882352941176475e-05,
      "loss": 0.7015,
      "step": 11
    },
    {
      "epoch": 0.10786516853932585,
      "grad_norm": 1.8786098685363002,
      "learning_rate": 2.8235294117647063e-05,
      "loss": 0.6907,
      "step": 12
    },
    {
      "epoch": 0.11685393258426967,
      "grad_norm": 1.217510887771457,
      "learning_rate": 3.0588235294117644e-05,
      "loss": 0.6657,
      "step": 13
    },
    {
      "epoch": 0.1258426966292135,
      "grad_norm": 1.645912176172542,
      "learning_rate": 3.294117647058824e-05,
      "loss": 0.6501,
      "step": 14
    },
    {
      "epoch": 0.1348314606741573,
      "grad_norm": 1.1214763643282593,
      "learning_rate": 3.529411764705883e-05,
      "loss": 0.6431,
      "step": 15
    },
    {
      "epoch": 0.14382022471910114,
      "grad_norm": 1.3566330254296206,
      "learning_rate": 3.7647058823529415e-05,
      "loss": 0.6305,
      "step": 16
    },
    {
      "epoch": 0.15280898876404495,
      "grad_norm": 1.5612433580633596,
      "learning_rate": 4e-05,
      "loss": 0.6239,
      "step": 17
    },
    {
      "epoch": 0.16179775280898875,
      "grad_norm": 1.6192637419680402,
      "learning_rate": 4.235294117647059e-05,
      "loss": 0.6169,
      "step": 18
    },
    {
      "epoch": 0.1707865168539326,
      "grad_norm": 1.2309268080406381,
      "learning_rate": 4.470588235294118e-05,
      "loss": 0.6097,
      "step": 19
    },
    {
      "epoch": 0.1797752808988764,
      "grad_norm": 1.862538444377319,
      "learning_rate": 4.705882352941177e-05,
      "loss": 0.6134,
      "step": 20
    },
    {
      "epoch": 0.18876404494382024,
      "grad_norm": 1.5958037318215281,
      "learning_rate": 4.941176470588236e-05,
      "loss": 0.6007,
      "step": 21
    },
    {
      "epoch": 0.19775280898876405,
      "grad_norm": 1.357215808802562,
      "learning_rate": 5.176470588235295e-05,
      "loss": 0.5961,
      "step": 22
    },
    {
      "epoch": 0.20674157303370785,
      "grad_norm": 1.3402016135380905,
      "learning_rate": 5.411764705882354e-05,
      "loss": 0.5872,
      "step": 23
    },
    {
      "epoch": 0.2157303370786517,
      "grad_norm": 1.9136536863131055,
      "learning_rate": 5.6470588235294126e-05,
      "loss": 0.5779,
      "step": 24
    },
    {
      "epoch": 0.2247191011235955,
      "grad_norm": 1.0205411195750105,
      "learning_rate": 5.8823529411764714e-05,
      "loss": 0.5728,
      "step": 25
    },
    {
      "epoch": 0.23370786516853934,
      "grad_norm": 2.6452559257739265,
      "learning_rate": 6.117647058823529e-05,
      "loss": 0.5948,
      "step": 26
    },
    {
      "epoch": 0.24269662921348314,
      "grad_norm": 1.7949903848826065,
      "learning_rate": 6.352941176470589e-05,
      "loss": 0.5869,
      "step": 27
    },
    {
      "epoch": 0.251685393258427,
      "grad_norm": 1.9069324360886062,
      "learning_rate": 6.588235294117648e-05,
      "loss": 0.5816,
      "step": 28
    },
    {
      "epoch": 0.2606741573033708,
      "grad_norm": 1.7027526272642033,
      "learning_rate": 6.823529411764707e-05,
      "loss": 0.5674,
      "step": 29
    },
    {
      "epoch": 0.2696629213483146,
      "grad_norm": 1.6647039397669456,
      "learning_rate": 7.058823529411765e-05,
      "loss": 0.5801,
      "step": 30
    },
    {
      "epoch": 0.2786516853932584,
      "grad_norm": 1.2647930520355923,
      "learning_rate": 7.294117647058824e-05,
      "loss": 0.5622,
      "step": 31
    },
    {
      "epoch": 0.2876404494382023,
      "grad_norm": 1.8668167622362088,
      "learning_rate": 7.529411764705883e-05,
      "loss": 0.5628,
      "step": 32
    },
    {
      "epoch": 0.2966292134831461,
      "grad_norm": 1.6151821153814043,
      "learning_rate": 7.764705882352942e-05,
      "loss": 0.569,
      "step": 33
    },
    {
      "epoch": 0.3056179775280899,
      "grad_norm": 1.9972629519815388,
      "learning_rate": 8e-05,
      "loss": 0.5559,
      "step": 34
    },
    {
      "epoch": 0.3146067415730337,
      "grad_norm": 1.930654248915306,
      "learning_rate": 7.999779207981935e-05,
      "loss": 0.5578,
      "step": 35
    },
    {
      "epoch": 0.3235955056179775,
      "grad_norm": 1.1040009369498223,
      "learning_rate": 7.999116856302298e-05,
      "loss": 0.5553,
      "step": 36
    },
    {
      "epoch": 0.3325842696629214,
      "grad_norm": 1.9114787612288138,
      "learning_rate": 7.998013018082072e-05,
      "loss": 0.5518,
      "step": 37
    },
    {
      "epoch": 0.3415730337078652,
      "grad_norm": 2.5196506278998476,
      "learning_rate": 7.996467815180588e-05,
      "loss": 0.5587,
      "step": 38
    },
    {
      "epoch": 0.350561797752809,
      "grad_norm": 1.5471557908374793,
      "learning_rate": 7.994481418182082e-05,
      "loss": 0.5394,
      "step": 39
    },
    {
      "epoch": 0.3595505617977528,
      "grad_norm": 2.747162602890951,
      "learning_rate": 7.992054046376854e-05,
      "loss": 0.5602,
      "step": 40
    },
    {
      "epoch": 0.3685393258426966,
      "grad_norm": 1.9963506460084293,
      "learning_rate": 7.989185967737066e-05,
      "loss": 0.5489,
      "step": 41
    },
    {
      "epoch": 0.3775280898876405,
      "grad_norm": 2.356173456359232,
      "learning_rate": 7.985877498887149e-05,
      "loss": 0.554,
      "step": 42
    },
    {
      "epoch": 0.3865168539325843,
      "grad_norm": 1.9651033697846167,
      "learning_rate": 7.982129005068865e-05,
      "loss": 0.5432,
      "step": 43
    },
    {
      "epoch": 0.3955056179775281,
      "grad_norm": 1.7900792925575555,
      "learning_rate": 7.977940900100967e-05,
      "loss": 0.5476,
      "step": 44
    },
    {
      "epoch": 0.4044943820224719,
      "grad_norm": 1.8559704498293903,
      "learning_rate": 7.973313646333532e-05,
      "loss": 0.535,
      "step": 45
    },
    {
      "epoch": 0.4134831460674157,
      "grad_norm": 1.4415920297319014,
      "learning_rate": 7.968247754596908e-05,
      "loss": 0.5339,
      "step": 46
    },
    {
      "epoch": 0.42247191011235957,
      "grad_norm": 1.4135208728964568,
      "learning_rate": 7.962743784145323e-05,
      "loss": 0.5381,
      "step": 47
    },
    {
      "epoch": 0.4314606741573034,
      "grad_norm": 1.6688537146906361,
      "learning_rate": 7.956802342595152e-05,
      "loss": 0.5328,
      "step": 48
    },
    {
      "epoch": 0.4404494382022472,
      "grad_norm": 1.3978532275006461,
      "learning_rate": 7.950424085857827e-05,
      "loss": 0.5265,
      "step": 49
    },
    {
      "epoch": 0.449438202247191,
      "grad_norm": 1.7334227968766784,
      "learning_rate": 7.943609718067437e-05,
      "loss": 0.5245,
      "step": 50
    },
    {
      "epoch": 0.4584269662921348,
      "grad_norm": 0.7953199977805888,
      "learning_rate": 7.936359991502993e-05,
      "loss": 0.5167,
      "step": 51
    },
    {
      "epoch": 0.46741573033707867,
      "grad_norm": 1.6871487499435203,
      "learning_rate": 7.92867570650537e-05,
      "loss": 0.5325,
      "step": 52
    },
    {
      "epoch": 0.4764044943820225,
      "grad_norm": 1.304827694741159,
      "learning_rate": 7.920557711388967e-05,
      "loss": 0.5371,
      "step": 53
    },
    {
      "epoch": 0.4853932584269663,
      "grad_norm": 1.6036182086909745,
      "learning_rate": 7.912006902348045e-05,
      "loss": 0.5245,
      "step": 54
    },
    {
      "epoch": 0.4943820224719101,
      "grad_norm": 1.0607757354012914,
      "learning_rate": 7.903024223357797e-05,
      "loss": 0.5289,
      "step": 55
    },
    {
      "epoch": 0.503370786516854,
      "grad_norm": 2.0583378114050728,
      "learning_rate": 7.893610666070134e-05,
      "loss": 0.5346,
      "step": 56
    },
    {
      "epoch": 0.5123595505617977,
      "grad_norm": 1.4305584900288937,
      "learning_rate": 7.883767269704209e-05,
      "loss": 0.5251,
      "step": 57
    },
    {
      "epoch": 0.5213483146067416,
      "grad_norm": 1.5466083404792839,
      "learning_rate": 7.873495120931697e-05,
      "loss": 0.5242,
      "step": 58
    },
    {
      "epoch": 0.5303370786516854,
      "grad_norm": 1.3322841661439688,
      "learning_rate": 7.86279535375683e-05,
      "loss": 0.5266,
      "step": 59
    },
    {
      "epoch": 0.5393258426966292,
      "grad_norm": 1.2220900027281376,
      "learning_rate": 7.851669149391198e-05,
      "loss": 0.5228,
      "step": 60
    },
    {
      "epoch": 0.5483146067415731,
      "grad_norm": 1.5982554174133687,
      "learning_rate": 7.84011773612336e-05,
      "loss": 0.5231,
      "step": 61
    },
    {
      "epoch": 0.5573033707865168,
      "grad_norm": 1.146536447272186,
      "learning_rate": 7.828142389183239e-05,
      "loss": 0.5195,
      "step": 62
    },
    {
      "epoch": 0.5662921348314607,
      "grad_norm": 1.0879405901836858,
      "learning_rate": 7.815744430601344e-05,
      "loss": 0.5129,
      "step": 63
    },
    {
      "epoch": 0.5752808988764045,
      "grad_norm": 1.6921788651579457,
      "learning_rate": 7.802925229062823e-05,
      "loss": 0.521,
      "step": 64
    },
    {
      "epoch": 0.5842696629213483,
      "grad_norm": 1.0795726495535796,
      "learning_rate": 7.789686199756365e-05,
      "loss": 0.5172,
      "step": 65
    },
    {
      "epoch": 0.5932584269662922,
      "grad_norm": 1.1065195406196553,
      "learning_rate": 7.776028804217968e-05,
      "loss": 0.5134,
      "step": 66
    },
    {
      "epoch": 0.6022471910112359,
      "grad_norm": 1.2749574533792616,
      "learning_rate": 7.761954550169593e-05,
      "loss": 0.5161,
      "step": 67
    },
    {
      "epoch": 0.6112359550561798,
      "grad_norm": 1.3329895959967077,
      "learning_rate": 7.74746499135272e-05,
      "loss": 0.5153,
      "step": 68
    },
    {
      "epoch": 0.6202247191011236,
      "grad_norm": 1.233521082461672,
      "learning_rate": 7.732561727356811e-05,
      "loss": 0.5086,
      "step": 69
    },
    {
      "epoch": 0.6292134831460674,
      "grad_norm": 1.1760347925547088,
      "learning_rate": 7.717246403442735e-05,
      "loss": 0.5137,
      "step": 70
    },
    {
      "epoch": 0.6382022471910113,
      "grad_norm": 1.1780036452883795,
      "learning_rate": 7.701520710361129e-05,
      "loss": 0.4999,
      "step": 71
    },
    {
      "epoch": 0.647191011235955,
      "grad_norm": 0.948808130697301,
      "learning_rate": 7.685386384165748e-05,
      "loss": 0.5036,
      "step": 72
    },
    {
      "epoch": 0.6561797752808989,
      "grad_norm": 1.7675174539861906,
      "learning_rate": 7.668845206021812e-05,
      "loss": 0.5132,
      "step": 73
    },
    {
      "epoch": 0.6651685393258427,
      "grad_norm": 0.7501534484404624,
      "learning_rate": 7.651899002009375e-05,
      "loss": 0.5044,
      "step": 74
    },
    {
      "epoch": 0.6741573033707865,
      "grad_norm": 1.2580366817929822,
      "learning_rate": 7.634549642921725e-05,
      "loss": 0.5158,
      "step": 75
    },
    {
      "epoch": 0.6831460674157304,
      "grad_norm": 1.2116010404784998,
      "learning_rate": 7.616799044058867e-05,
      "loss": 0.5132,
      "step": 76
    },
    {
      "epoch": 0.6921348314606741,
      "grad_norm": 1.504394845229427,
      "learning_rate": 7.598649165016073e-05,
      "loss": 0.5051,
      "step": 77
    },
    {
      "epoch": 0.701123595505618,
      "grad_norm": 0.8286673874695757,
      "learning_rate": 7.58010200946755e-05,
      "loss": 0.5,
      "step": 78
    },
    {
      "epoch": 0.7101123595505618,
      "grad_norm": 1.6610859144234769,
      "learning_rate": 7.561159624945257e-05,
      "loss": 0.5094,
      "step": 79
    },
    {
      "epoch": 0.7191011235955056,
      "grad_norm": 0.8748352440236109,
      "learning_rate": 7.541824102612839e-05,
      "loss": 0.5009,
      "step": 80
    },
    {
      "epoch": 0.7280898876404495,
      "grad_norm": 1.4412943546400492,
      "learning_rate": 7.5220975770348e-05,
      "loss": 0.501,
      "step": 81
    },
    {
      "epoch": 0.7370786516853932,
      "grad_norm": 1.017225970339285,
      "learning_rate": 7.501982225940833e-05,
      "loss": 0.5009,
      "step": 82
    },
    {
      "epoch": 0.7460674157303371,
      "grad_norm": 1.3512165456613465,
      "learning_rate": 7.48148026998542e-05,
      "loss": 0.503,
      "step": 83
    },
    {
      "epoch": 0.755056179775281,
      "grad_norm": 1.0549525315803745,
      "learning_rate": 7.460593972502674e-05,
      "loss": 0.4953,
      "step": 84
    },
    {
      "epoch": 0.7640449438202247,
      "grad_norm": 1.1196254022912724,
      "learning_rate": 7.439325639256483e-05,
      "loss": 0.4897,
      "step": 85
    },
    {
      "epoch": 0.7730337078651686,
      "grad_norm": 1.1317585655215339,
      "learning_rate": 7.417677618185955e-05,
      "loss": 0.4969,
      "step": 86
    },
    {
      "epoch": 0.7820224719101123,
      "grad_norm": 1.2763979132088588,
      "learning_rate": 7.39565229914622e-05,
      "loss": 0.5046,
      "step": 87
    },
    {
      "epoch": 0.7910112359550562,
      "grad_norm": 1.2131663136230482,
      "learning_rate": 7.373252113644596e-05,
      "loss": 0.5006,
      "step": 88
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.121494784629588,
      "learning_rate": 7.350479534572166e-05,
      "loss": 0.4949,
      "step": 89
    },
    {
      "epoch": 0.8089887640449438,
      "grad_norm": 0.9241654008148638,
      "learning_rate": 7.327337075930775e-05,
      "loss": 0.5003,
      "step": 90
    },
    {
      "epoch": 0.8179775280898877,
      "grad_norm": 1.4368488374102792,
      "learning_rate": 7.303827292555495e-05,
      "loss": 0.4907,
      "step": 91
    },
    {
      "epoch": 0.8269662921348314,
      "grad_norm": 1.2319104612634353,
      "learning_rate": 7.279952779832584e-05,
      "loss": 0.5043,
      "step": 92
    },
    {
      "epoch": 0.8359550561797753,
      "grad_norm": 0.8671468092587438,
      "learning_rate": 7.255716173412966e-05,
      "loss": 0.486,
      "step": 93
    },
    {
      "epoch": 0.8449438202247191,
      "grad_norm": 1.3364076805186555,
      "learning_rate": 7.23112014892126e-05,
      "loss": 0.4932,
      "step": 94
    },
    {
      "epoch": 0.8539325842696629,
      "grad_norm": 1.0720511804076949,
      "learning_rate": 7.20616742166041e-05,
      "loss": 0.49,
      "step": 95
    },
    {
      "epoch": 0.8629213483146068,
      "grad_norm": 0.8784137232709633,
      "learning_rate": 7.180860746311917e-05,
      "loss": 0.4911,
      "step": 96
    },
    {
      "epoch": 0.8719101123595505,
      "grad_norm": 1.0650244396836273,
      "learning_rate": 7.155202916631743e-05,
      "loss": 0.4922,
      "step": 97
    },
    {
      "epoch": 0.8808988764044944,
      "grad_norm": 1.15396517981509,
      "learning_rate": 7.129196765141886e-05,
      "loss": 0.4899,
      "step": 98
    },
    {
      "epoch": 0.8898876404494382,
      "grad_norm": 0.8409463006756092,
      "learning_rate": 7.10284516281768e-05,
      "loss": 0.4935,
      "step": 99
    },
    {
      "epoch": 0.898876404494382,
      "grad_norm": 0.8987420745372244,
      "learning_rate": 7.076151018770854e-05,
      "loss": 0.4933,
      "step": 100
    },
    {
      "epoch": 0.9078651685393259,
      "grad_norm": 1.3476943698616615,
      "learning_rate": 7.049117279928374e-05,
      "loss": 0.4941,
      "step": 101
    },
    {
      "epoch": 0.9168539325842696,
      "grad_norm": 1.0470417246613994,
      "learning_rate": 7.021746930707117e-05,
      "loss": 0.4878,
      "step": 102
    },
    {
      "epoch": 0.9258426966292135,
      "grad_norm": 1.222327504764566,
      "learning_rate": 6.994042992684406e-05,
      "loss": 0.478,
      "step": 103
    },
    {
      "epoch": 0.9348314606741573,
      "grad_norm": 0.8143177540157983,
      "learning_rate": 6.966008524264429e-05,
      "loss": 0.4868,
      "step": 104
    },
    {
      "epoch": 0.9438202247191011,
      "grad_norm": 1.3154017535872904,
      "learning_rate": 6.937646620340618e-05,
      "loss": 0.4921,
      "step": 105
    },
    {
      "epoch": 0.952808988764045,
      "grad_norm": 1.2828038188846893,
      "learning_rate": 6.908960411953973e-05,
      "loss": 0.4871,
      "step": 106
    },
    {
      "epoch": 0.9617977528089887,
      "grad_norm": 0.7484103692531322,
      "learning_rate": 6.879953065947416e-05,
      "loss": 0.4856,
      "step": 107
    },
    {
      "epoch": 0.9707865168539326,
      "grad_norm": 1.393234551386416,
      "learning_rate": 6.850627784616178e-05,
      "loss": 0.4884,
      "step": 108
    },
    {
      "epoch": 0.9797752808988764,
      "grad_norm": 0.6778085483847683,
      "learning_rate": 6.82098780535428e-05,
      "loss": 0.4844,
      "step": 109
    },
    {
      "epoch": 0.9887640449438202,
      "grad_norm": 0.95356400631233,
      "learning_rate": 6.791036400297142e-05,
      "loss": 0.4766,
      "step": 110
    },
    {
      "epoch": 0.9977528089887641,
      "grad_norm": 0.9616360385718218,
      "learning_rate": 6.760776875960347e-05,
      "loss": 0.4796,
      "step": 111
    },
    {
      "epoch": 1.006741573033708,
      "grad_norm": 1.8748279594605497,
      "learning_rate": 6.730212572874618e-05,
      "loss": 0.8205,
      "step": 112
    },
    {
      "epoch": 1.0157303370786517,
      "grad_norm": 0.7712220221008119,
      "learning_rate": 6.699346865217031e-05,
      "loss": 0.4547,
      "step": 113
    },
    {
      "epoch": 1.0247191011235954,
      "grad_norm": 1.1476060733360436,
      "learning_rate": 6.668183160438531e-05,
      "loss": 0.4598,
      "step": 114
    },
    {
      "epoch": 1.0337078651685394,
      "grad_norm": 1.091336171209942,
      "learning_rate": 6.636724898887751e-05,
      "loss": 0.4566,
      "step": 115
    },
    {
      "epoch": 1.0426966292134832,
      "grad_norm": 1.356832256792843,
      "learning_rate": 6.604975553431219e-05,
      "loss": 0.4636,
      "step": 116
    },
    {
      "epoch": 1.051685393258427,
      "grad_norm": 0.8230787378003628,
      "learning_rate": 6.572938629069959e-05,
      "loss": 0.4539,
      "step": 117
    },
    {
      "epoch": 1.060674157303371,
      "grad_norm": 0.7577254786527922,
      "learning_rate": 6.540617662552565e-05,
      "loss": 0.4626,
      "step": 118
    },
    {
      "epoch": 1.0696629213483146,
      "grad_norm": 0.9663188263281904,
      "learning_rate": 6.508016221984747e-05,
      "loss": 0.4507,
      "step": 119
    },
    {
      "epoch": 1.0786516853932584,
      "grad_norm": 1.4024310383300067,
      "learning_rate": 6.475137906435435e-05,
      "loss": 0.4613,
      "step": 120
    },
    {
      "epoch": 1.0876404494382022,
      "grad_norm": 0.7663106953730127,
      "learning_rate": 6.441986345539446e-05,
      "loss": 0.4656,
      "step": 121
    },
    {
      "epoch": 1.0966292134831461,
      "grad_norm": 1.1216710559366627,
      "learning_rate": 6.408565199096798e-05,
      "loss": 0.4529,
      "step": 122
    },
    {
      "epoch": 1.1056179775280899,
      "grad_norm": 0.8823684903577822,
      "learning_rate": 6.374878156668676e-05,
      "loss": 0.4588,
      "step": 123
    },
    {
      "epoch": 1.1146067415730336,
      "grad_norm": 0.7542533792179187,
      "learning_rate": 6.340928937170118e-05,
      "loss": 0.4551,
      "step": 124
    },
    {
      "epoch": 1.1235955056179776,
      "grad_norm": 1.1622893113229344,
      "learning_rate": 6.30672128845947e-05,
      "loss": 0.4587,
      "step": 125
    },
    {
      "epoch": 1.1325842696629214,
      "grad_norm": 0.6982046729289911,
      "learning_rate": 6.272258986924624e-05,
      "loss": 0.4527,
      "step": 126
    },
    {
      "epoch": 1.1415730337078651,
      "grad_norm": 0.5854116152358985,
      "learning_rate": 6.237545837066133e-05,
      "loss": 0.4518,
      "step": 127
    },
    {
      "epoch": 1.1505617977528089,
      "grad_norm": 0.47197050837846644,
      "learning_rate": 6.202585671077204e-05,
      "loss": 0.4474,
      "step": 128
    },
    {
      "epoch": 1.1595505617977528,
      "grad_norm": 0.4631791319657276,
      "learning_rate": 6.167382348420637e-05,
      "loss": 0.4573,
      "step": 129
    },
    {
      "epoch": 1.1685393258426966,
      "grad_norm": 0.6398622174196904,
      "learning_rate": 6.131939755402755e-05,
      "loss": 0.4551,
      "step": 130
    },
    {
      "epoch": 1.1775280898876406,
      "grad_norm": 1.110637394887812,
      "learning_rate": 6.09626180474438e-05,
      "loss": 0.4524,
      "step": 131
    },
    {
      "epoch": 1.1865168539325843,
      "grad_norm": 1.377350137912211,
      "learning_rate": 6.060352435148874e-05,
      "loss": 0.4559,
      "step": 132
    },
    {
      "epoch": 1.195505617977528,
      "grad_norm": 0.5283491845810788,
      "learning_rate": 6.024215610867327e-05,
      "loss": 0.4603,
      "step": 133
    },
    {
      "epoch": 1.2044943820224718,
      "grad_norm": 0.9966520169090612,
      "learning_rate": 5.9878553212609184e-05,
      "loss": 0.4516,
      "step": 134
    },
    {
      "epoch": 1.2134831460674158,
      "grad_norm": 1.4576011324058094,
      "learning_rate": 5.95127558036051e-05,
      "loss": 0.4596,
      "step": 135
    },
    {
      "epoch": 1.2224719101123596,
      "grad_norm": 0.697821502136658,
      "learning_rate": 5.9144804264235066e-05,
      "loss": 0.4576,
      "step": 136
    },
    {
      "epoch": 1.2314606741573033,
      "grad_norm": 1.5590501658526488,
      "learning_rate": 5.8774739214880554e-05,
      "loss": 0.4602,
      "step": 137
    },
    {
      "epoch": 1.2404494382022473,
      "grad_norm": 0.6045793781300405,
      "learning_rate": 5.840260150924609e-05,
      "loss": 0.45,
      "step": 138
    },
    {
      "epoch": 1.249438202247191,
      "grad_norm": 1.417376379462992,
      "learning_rate": 5.802843222984919e-05,
      "loss": 0.4564,
      "step": 139
    },
    {
      "epoch": 1.2584269662921348,
      "grad_norm": 0.7383419101529723,
      "learning_rate": 5.765227268348501e-05,
      "loss": 0.4533,
      "step": 140
    },
    {
      "epoch": 1.2674157303370785,
      "grad_norm": 1.4264362126630619,
      "learning_rate": 5.727416439666622e-05,
      "loss": 0.457,
      "step": 141
    },
    {
      "epoch": 1.2764044943820225,
      "grad_norm": 0.8638040394017062,
      "learning_rate": 5.689414911103867e-05,
      "loss": 0.4516,
      "step": 142
    },
    {
      "epoch": 1.2853932584269663,
      "grad_norm": 1.42829306450921,
      "learning_rate": 5.651226877877326e-05,
      "loss": 0.4556,
      "step": 143
    },
    {
      "epoch": 1.29438202247191,
      "grad_norm": 1.0514839616322087,
      "learning_rate": 5.612856555793459e-05,
      "loss": 0.4554,
      "step": 144
    },
    {
      "epoch": 1.303370786516854,
      "grad_norm": 1.2796676473517508,
      "learning_rate": 5.574308180782693e-05,
      "loss": 0.4491,
      "step": 145
    },
    {
      "epoch": 1.3123595505617978,
      "grad_norm": 1.0932581128919143,
      "learning_rate": 5.5355860084317787e-05,
      "loss": 0.4575,
      "step": 146
    },
    {
      "epoch": 1.3213483146067415,
      "grad_norm": 0.8711585447910877,
      "learning_rate": 5.496694313514009e-05,
      "loss": 0.4536,
      "step": 147
    },
    {
      "epoch": 1.3303370786516853,
      "grad_norm": 0.976493438245937,
      "learning_rate": 5.457637389517285e-05,
      "loss": 0.4616,
      "step": 148
    },
    {
      "epoch": 1.3393258426966292,
      "grad_norm": 0.600498943325478,
      "learning_rate": 5.4184195481701425e-05,
      "loss": 0.4474,
      "step": 149
    },
    {
      "epoch": 1.348314606741573,
      "grad_norm": 0.7706578711873172,
      "learning_rate": 5.3790451189657486e-05,
      "loss": 0.4511,
      "step": 150
    },
    {
      "epoch": 1.357303370786517,
      "grad_norm": 0.5178771915217835,
      "learning_rate": 5.339518448683945e-05,
      "loss": 0.4514,
      "step": 151
    },
    {
      "epoch": 1.3662921348314607,
      "grad_norm": 0.7334914676346533,
      "learning_rate": 5.2998439009113814e-05,
      "loss": 0.4522,
      "step": 152
    },
    {
      "epoch": 1.3752808988764045,
      "grad_norm": 0.5701908032527899,
      "learning_rate": 5.260025855559792e-05,
      "loss": 0.457,
      "step": 153
    },
    {
      "epoch": 1.3842696629213482,
      "grad_norm": 0.45968308679963693,
      "learning_rate": 5.2200687083824706e-05,
      "loss": 0.4501,
      "step": 154
    },
    {
      "epoch": 1.3932584269662922,
      "grad_norm": 0.5556354013625224,
      "learning_rate": 5.179976870488999e-05,
      "loss": 0.4508,
      "step": 155
    },
    {
      "epoch": 1.402247191011236,
      "grad_norm": 0.4445793572850935,
      "learning_rate": 5.1397547678582745e-05,
      "loss": 0.4507,
      "step": 156
    },
    {
      "epoch": 1.4112359550561797,
      "grad_norm": 0.5201992901363155,
      "learning_rate": 5.099406840849902e-05,
      "loss": 0.4535,
      "step": 157
    },
    {
      "epoch": 1.4202247191011237,
      "grad_norm": 0.5297850125662515,
      "learning_rate": 5.058937543713999e-05,
      "loss": 0.4436,
      "step": 158
    },
    {
      "epoch": 1.4292134831460674,
      "grad_norm": 0.3969459180847379,
      "learning_rate": 5.018351344099453e-05,
      "loss": 0.4532,
      "step": 159
    },
    {
      "epoch": 1.4382022471910112,
      "grad_norm": 0.47432246499785363,
      "learning_rate": 4.9776527225607274e-05,
      "loss": 0.4511,
      "step": 160
    },
    {
      "epoch": 1.447191011235955,
      "grad_norm": 0.31975394190454653,
      "learning_rate": 4.93684617206321e-05,
      "loss": 0.4485,
      "step": 161
    },
    {
      "epoch": 1.456179775280899,
      "grad_norm": 0.4164002143535295,
      "learning_rate": 4.89593619748722e-05,
      "loss": 0.4481,
      "step": 162
    },
    {
      "epoch": 1.4651685393258427,
      "grad_norm": 0.367528177243067,
      "learning_rate": 4.8549273151306795e-05,
      "loss": 0.4446,
      "step": 163
    },
    {
      "epoch": 1.4741573033707867,
      "grad_norm": 0.36494718836358464,
      "learning_rate": 4.8138240522105365e-05,
      "loss": 0.4516,
      "step": 164
    },
    {
      "epoch": 1.4831460674157304,
      "grad_norm": 0.2981123577566662,
      "learning_rate": 4.7726309463629733e-05,
      "loss": 0.439,
      "step": 165
    },
    {
      "epoch": 1.4921348314606742,
      "grad_norm": 0.31072790704633446,
      "learning_rate": 4.731352545142478e-05,
      "loss": 0.449,
      "step": 166
    },
    {
      "epoch": 1.501123595505618,
      "grad_norm": 0.3992742056784971,
      "learning_rate": 4.689993405519802e-05,
      "loss": 0.4433,
      "step": 167
    },
    {
      "epoch": 1.5101123595505617,
      "grad_norm": 0.2594367011131309,
      "learning_rate": 4.648558093378899e-05,
      "loss": 0.4501,
      "step": 168
    },
    {
      "epoch": 1.5191011235955056,
      "grad_norm": 0.2830467959001486,
      "learning_rate": 4.607051183012862e-05,
      "loss": 0.4422,
      "step": 169
    },
    {
      "epoch": 1.5280898876404494,
      "grad_norm": 0.28137593410743117,
      "learning_rate": 4.5654772566189415e-05,
      "loss": 0.4439,
      "step": 170
    },
    {
      "epoch": 1.5370786516853934,
      "grad_norm": 0.30222361671491843,
      "learning_rate": 4.5238409037926905e-05,
      "loss": 0.4464,
      "step": 171
    },
    {
      "epoch": 1.5460674157303371,
      "grad_norm": 0.2984953677611448,
      "learning_rate": 4.4821467210212924e-05,
      "loss": 0.4515,
      "step": 172
    },
    {
      "epoch": 1.5550561797752809,
      "grad_norm": 0.29477762612841285,
      "learning_rate": 4.4403993111761265e-05,
      "loss": 0.4487,
      "step": 173
    },
    {
      "epoch": 1.5640449438202246,
      "grad_norm": 0.27738094537303265,
      "learning_rate": 4.398603283004626e-05,
      "loss": 0.4349,
      "step": 174
    },
    {
      "epoch": 1.5730337078651684,
      "grad_norm": 0.27179761514817713,
      "learning_rate": 4.356763250621496e-05,
      "loss": 0.439,
      "step": 175
    },
    {
      "epoch": 1.5820224719101124,
      "grad_norm": 0.30145150102869067,
      "learning_rate": 4.314883832999326e-05,
      "loss": 0.4399,
      "step": 176
    },
    {
      "epoch": 1.5910112359550563,
      "grad_norm": 0.27593285844535975,
      "learning_rate": 4.272969653458685e-05,
      "loss": 0.449,
      "step": 177
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.30712340293439133,
      "learning_rate": 4.231025339157714e-05,
      "loss": 0.4368,
      "step": 178
    },
    {
      "epoch": 1.6089887640449438,
      "grad_norm": 0.26832779433391346,
      "learning_rate": 4.189055520581315e-05,
      "loss": 0.4447,
      "step": 179
    },
    {
      "epoch": 1.6179775280898876,
      "grad_norm": 0.2800675843384406,
      "learning_rate": 4.147064831029959e-05,
      "loss": 0.445,
      "step": 180
    },
    {
      "epoch": 1.6269662921348313,
      "grad_norm": 0.3067111758990106,
      "learning_rate": 4.105057906108189e-05,
      "loss": 0.4452,
      "step": 181
    },
    {
      "epoch": 1.6359550561797753,
      "grad_norm": 0.24971049417224553,
      "learning_rate": 4.063039383212866e-05,
      "loss": 0.4425,
      "step": 182
    },
    {
      "epoch": 1.644943820224719,
      "grad_norm": 0.2020310973856206,
      "learning_rate": 4.021013901021225e-05,
      "loss": 0.4411,
      "step": 183
    },
    {
      "epoch": 1.653932584269663,
      "grad_norm": 0.2574200192037008,
      "learning_rate": 3.978986098978777e-05,
      "loss": 0.4517,
      "step": 184
    },
    {
      "epoch": 1.6629213483146068,
      "grad_norm": 0.2736980359380189,
      "learning_rate": 3.936960616787135e-05,
      "loss": 0.4446,
      "step": 185
    },
    {
      "epoch": 1.6719101123595506,
      "grad_norm": 0.24658217170903196,
      "learning_rate": 3.8949420938918124e-05,
      "loss": 0.4364,
      "step": 186
    },
    {
      "epoch": 1.6808988764044943,
      "grad_norm": 0.2767599168622161,
      "learning_rate": 3.852935168970042e-05,
      "loss": 0.4357,
      "step": 187
    },
    {
      "epoch": 1.689887640449438,
      "grad_norm": 0.2548602216487942,
      "learning_rate": 3.810944479418686e-05,
      "loss": 0.4399,
      "step": 188
    },
    {
      "epoch": 1.698876404494382,
      "grad_norm": 0.2640517946445997,
      "learning_rate": 3.768974660842287e-05,
      "loss": 0.4475,
      "step": 189
    },
    {
      "epoch": 1.7078651685393258,
      "grad_norm": 0.25913290546087026,
      "learning_rate": 3.727030346541317e-05,
      "loss": 0.437,
      "step": 190
    },
    {
      "epoch": 1.7168539325842698,
      "grad_norm": 0.21742230610782412,
      "learning_rate": 3.685116167000675e-05,
      "loss": 0.4376,
      "step": 191
    },
    {
      "epoch": 1.7258426966292135,
      "grad_norm": 0.2076811227564877,
      "learning_rate": 3.6432367493785056e-05,
      "loss": 0.438,
      "step": 192
    },
    {
      "epoch": 1.7348314606741573,
      "grad_norm": 0.3321278391017944,
      "learning_rate": 3.601396716995375e-05,
      "loss": 0.4388,
      "step": 193
    },
    {
      "epoch": 1.743820224719101,
      "grad_norm": 0.2478941601472815,
      "learning_rate": 3.559600688823875e-05,
      "loss": 0.4452,
      "step": 194
    },
    {
      "epoch": 1.7528089887640448,
      "grad_norm": 0.28877121684016926,
      "learning_rate": 3.517853278978708e-05,
      "loss": 0.442,
      "step": 195
    },
    {
      "epoch": 1.7617977528089888,
      "grad_norm": 0.2393813520807247,
      "learning_rate": 3.4761590962073115e-05,
      "loss": 0.4437,
      "step": 196
    },
    {
      "epoch": 1.7707865168539327,
      "grad_norm": 0.26771357483054375,
      "learning_rate": 3.434522743381061e-05,
      "loss": 0.4496,
      "step": 197
    },
    {
      "epoch": 1.7797752808988765,
      "grad_norm": 0.2506533050953217,
      "learning_rate": 3.39294881698714e-05,
      "loss": 0.4443,
      "step": 198
    },
    {
      "epoch": 1.7887640449438202,
      "grad_norm": 0.21252098175250356,
      "learning_rate": 3.3514419066211025e-05,
      "loss": 0.4383,
      "step": 199
    },
    {
      "epoch": 1.797752808988764,
      "grad_norm": 0.3227331899101581,
      "learning_rate": 3.310006594480199e-05,
      "loss": 0.4373,
      "step": 200
    },
    {
      "epoch": 1.8067415730337077,
      "grad_norm": 0.2595288336777384,
      "learning_rate": 3.268647454857524e-05,
      "loss": 0.4432,
      "step": 201
    },
    {
      "epoch": 1.8157303370786517,
      "grad_norm": 0.22248085130320008,
      "learning_rate": 3.227369053637028e-05,
      "loss": 0.4378,
      "step": 202
    },
    {
      "epoch": 1.8247191011235955,
      "grad_norm": 0.26022739350380103,
      "learning_rate": 3.1861759477894656e-05,
      "loss": 0.4511,
      "step": 203
    },
    {
      "epoch": 1.8337078651685395,
      "grad_norm": 0.2525774955154011,
      "learning_rate": 3.145072684869322e-05,
      "loss": 0.4408,
      "step": 204
    },
    {
      "epoch": 1.8426966292134832,
      "grad_norm": 0.19933336701791182,
      "learning_rate": 3.104063802512782e-05,
      "loss": 0.4428,
      "step": 205
    },
    {
      "epoch": 1.851685393258427,
      "grad_norm": 0.2618315402075057,
      "learning_rate": 3.063153827936792e-05,
      "loss": 0.4387,
      "step": 206
    },
    {
      "epoch": 1.8606741573033707,
      "grad_norm": 0.2612229535126854,
      "learning_rate": 3.0223472774392753e-05,
      "loss": 0.4373,
      "step": 207
    },
    {
      "epoch": 1.8696629213483145,
      "grad_norm": 0.23175557173041528,
      "learning_rate": 2.9816486559005482e-05,
      "loss": 0.4428,
      "step": 208
    },
    {
      "epoch": 1.8786516853932584,
      "grad_norm": 0.2659832710112894,
      "learning_rate": 2.9410624562860026e-05,
      "loss": 0.4374,
      "step": 209
    },
    {
      "epoch": 1.8876404494382022,
      "grad_norm": 0.2394604361123856,
      "learning_rate": 2.9005931591500974e-05,
      "loss": 0.4377,
      "step": 210
    },
    {
      "epoch": 1.8966292134831462,
      "grad_norm": 0.21085090202042212,
      "learning_rate": 2.860245232141726e-05,
      "loss": 0.4446,
      "step": 211
    },
    {
      "epoch": 1.90561797752809,
      "grad_norm": 0.24320843039780787,
      "learning_rate": 2.8200231295110012e-05,
      "loss": 0.4449,
      "step": 212
    },
    {
      "epoch": 1.9146067415730337,
      "grad_norm": 0.18825252205454282,
      "learning_rate": 2.7799312916175294e-05,
      "loss": 0.4474,
      "step": 213
    },
    {
      "epoch": 1.9235955056179774,
      "grad_norm": 0.21020622489218485,
      "learning_rate": 2.7399741444402087e-05,
      "loss": 0.4396,
      "step": 214
    },
    {
      "epoch": 1.9325842696629212,
      "grad_norm": 0.19967691714665464,
      "learning_rate": 2.7001560990886196e-05,
      "loss": 0.432,
      "step": 215
    },
    {
      "epoch": 1.9415730337078652,
      "grad_norm": 0.1723917394030639,
      "learning_rate": 2.6604815513160556e-05,
      "loss": 0.4347,
      "step": 216
    },
    {
      "epoch": 1.9505617977528091,
      "grad_norm": 0.2286024751356636,
      "learning_rate": 2.6209548810342517e-05,
      "loss": 0.4272,
      "step": 217
    },
    {
      "epoch": 1.9595505617977529,
      "grad_norm": 0.1609850927229707,
      "learning_rate": 2.5815804518298575e-05,
      "loss": 0.4373,
      "step": 218
    },
    {
      "epoch": 1.9685393258426966,
      "grad_norm": 0.17225954372932897,
      "learning_rate": 2.542362610482715e-05,
      "loss": 0.4346,
      "step": 219
    },
    {
      "epoch": 1.9775280898876404,
      "grad_norm": 0.14538818432484127,
      "learning_rate": 2.503305686485991e-05,
      "loss": 0.444,
      "step": 220
    },
    {
      "epoch": 1.9865168539325841,
      "grad_norm": 0.17741790417981537,
      "learning_rate": 2.464413991568222e-05,
      "loss": 0.4413,
      "step": 221
    },
    {
      "epoch": 1.9955056179775281,
      "grad_norm": 0.19631454210187727,
      "learning_rate": 2.4256918192173088e-05,
      "loss": 0.4383,
      "step": 222
    },
    {
      "epoch": 2.004494382022472,
      "grad_norm": 0.2777440257511273,
      "learning_rate": 2.3871434442065414e-05,
      "loss": 0.7545,
      "step": 223
    },
    {
      "epoch": 2.013483146067416,
      "grad_norm": 0.2471458608915711,
      "learning_rate": 2.3487731221226754e-05,
      "loss": 0.4154,
      "step": 224
    },
    {
      "epoch": 2.0224719101123596,
      "grad_norm": 0.1821697922847147,
      "learning_rate": 2.3105850888961348e-05,
      "loss": 0.4165,
      "step": 225
    },
    {
      "epoch": 2.0314606741573034,
      "grad_norm": 0.24585959006825425,
      "learning_rate": 2.272583560333379e-05,
      "loss": 0.4141,
      "step": 226
    },
    {
      "epoch": 2.040449438202247,
      "grad_norm": 0.1702964012359119,
      "learning_rate": 2.2347727316515e-05,
      "loss": 0.4173,
      "step": 227
    },
    {
      "epoch": 2.049438202247191,
      "grad_norm": 0.24547374442583564,
      "learning_rate": 2.1971567770150814e-05,
      "loss": 0.4039,
      "step": 228
    },
    {
      "epoch": 2.0584269662921346,
      "grad_norm": 0.19448387676230083,
      "learning_rate": 2.1597398490753917e-05,
      "loss": 0.4126,
      "step": 229
    },
    {
      "epoch": 2.067415730337079,
      "grad_norm": 0.1711476507488459,
      "learning_rate": 2.1225260785119456e-05,
      "loss": 0.4112,
      "step": 230
    },
    {
      "epoch": 2.0764044943820226,
      "grad_norm": 0.21251938925846103,
      "learning_rate": 2.0855195735764947e-05,
      "loss": 0.4136,
      "step": 231
    },
    {
      "epoch": 2.0853932584269663,
      "grad_norm": 0.19044392751354922,
      "learning_rate": 2.0487244196394912e-05,
      "loss": 0.4158,
      "step": 232
    },
    {
      "epoch": 2.09438202247191,
      "grad_norm": 0.1815645592240795,
      "learning_rate": 2.0121446787390822e-05,
      "loss": 0.414,
      "step": 233
    },
    {
      "epoch": 2.103370786516854,
      "grad_norm": 0.2175948364317461,
      "learning_rate": 1.9757843891326736e-05,
      "loss": 0.409,
      "step": 234
    },
    {
      "epoch": 2.1123595505617976,
      "grad_norm": 0.16627200046068014,
      "learning_rate": 1.939647564851127e-05,
      "loss": 0.4118,
      "step": 235
    },
    {
      "epoch": 2.121348314606742,
      "grad_norm": 0.18089353116200013,
      "learning_rate": 1.9037381952556217e-05,
      "loss": 0.4082,
      "step": 236
    },
    {
      "epoch": 2.1303370786516855,
      "grad_norm": 0.1938341058307494,
      "learning_rate": 1.8680602445972463e-05,
      "loss": 0.4208,
      "step": 237
    },
    {
      "epoch": 2.1393258426966293,
      "grad_norm": 0.15296391756902566,
      "learning_rate": 1.832617651579365e-05,
      "loss": 0.4143,
      "step": 238
    },
    {
      "epoch": 2.148314606741573,
      "grad_norm": 0.2007977943202684,
      "learning_rate": 1.797414328922797e-05,
      "loss": 0.4126,
      "step": 239
    },
    {
      "epoch": 2.157303370786517,
      "grad_norm": 0.14549535548231118,
      "learning_rate": 1.7624541629338676e-05,
      "loss": 0.4107,
      "step": 240
    },
    {
      "epoch": 2.1662921348314605,
      "grad_norm": 0.1643469642395592,
      "learning_rate": 1.7277410130753775e-05,
      "loss": 0.4147,
      "step": 241
    },
    {
      "epoch": 2.1752808988764043,
      "grad_norm": 0.1743166147174567,
      "learning_rate": 1.6932787115405318e-05,
      "loss": 0.4079,
      "step": 242
    },
    {
      "epoch": 2.1842696629213485,
      "grad_norm": 0.15594946749382455,
      "learning_rate": 1.6590710628298826e-05,
      "loss": 0.41,
      "step": 243
    },
    {
      "epoch": 2.1932584269662923,
      "grad_norm": 0.16475084403286017,
      "learning_rate": 1.6251218433313254e-05,
      "loss": 0.4091,
      "step": 244
    },
    {
      "epoch": 2.202247191011236,
      "grad_norm": 0.13895524394350253,
      "learning_rate": 1.591434800903203e-05,
      "loss": 0.408,
      "step": 245
    },
    {
      "epoch": 2.2112359550561798,
      "grad_norm": 0.14999957095600605,
      "learning_rate": 1.558013654460555e-05,
      "loss": 0.409,
      "step": 246
    },
    {
      "epoch": 2.2202247191011235,
      "grad_norm": 0.1411289442089014,
      "learning_rate": 1.5248620935645666e-05,
      "loss": 0.4108,
      "step": 247
    },
    {
      "epoch": 2.2292134831460673,
      "grad_norm": 0.13130064830617885,
      "learning_rate": 1.4919837780152544e-05,
      "loss": 0.4086,
      "step": 248
    },
    {
      "epoch": 2.238202247191011,
      "grad_norm": 0.1362901361388582,
      "learning_rate": 1.4593823374474374e-05,
      "loss": 0.4134,
      "step": 249
    },
    {
      "epoch": 2.247191011235955,
      "grad_norm": 0.1641330954517184,
      "learning_rate": 1.4270613709300429e-05,
      "loss": 0.4105,
      "step": 250
    },
    {
      "epoch": 2.256179775280899,
      "grad_norm": 0.11707105305267158,
      "learning_rate": 1.3950244465687833e-05,
      "loss": 0.4113,
      "step": 251
    },
    {
      "epoch": 2.2651685393258427,
      "grad_norm": 0.14883912799444574,
      "learning_rate": 1.3632751011122497e-05,
      "loss": 0.4117,
      "step": 252
    },
    {
      "epoch": 2.2741573033707865,
      "grad_norm": 0.14576734750428819,
      "learning_rate": 1.3318168395614697e-05,
      "loss": 0.4086,
      "step": 253
    },
    {
      "epoch": 2.2831460674157302,
      "grad_norm": 0.1325863043302826,
      "learning_rate": 1.3006531347829699e-05,
      "loss": 0.4075,
      "step": 254
    },
    {
      "epoch": 2.292134831460674,
      "grad_norm": 0.14803632152992002,
      "learning_rate": 1.2697874271253844e-05,
      "loss": 0.4092,
      "step": 255
    },
    {
      "epoch": 2.3011235955056177,
      "grad_norm": 0.1484465193910235,
      "learning_rate": 1.2392231240396542e-05,
      "loss": 0.4103,
      "step": 256
    },
    {
      "epoch": 2.310112359550562,
      "grad_norm": 0.13481968569964842,
      "learning_rate": 1.2089635997028592e-05,
      "loss": 0.4165,
      "step": 257
    },
    {
      "epoch": 2.3191011235955057,
      "grad_norm": 0.14559693501049936,
      "learning_rate": 1.1790121946457212e-05,
      "loss": 0.4166,
      "step": 258
    },
    {
      "epoch": 2.3280898876404494,
      "grad_norm": 0.12726242281070635,
      "learning_rate": 1.1493722153838239e-05,
      "loss": 0.407,
      "step": 259
    },
    {
      "epoch": 2.337078651685393,
      "grad_norm": 0.13666677240818353,
      "learning_rate": 1.120046934052585e-05,
      "loss": 0.413,
      "step": 260
    },
    {
      "epoch": 2.346067415730337,
      "grad_norm": 0.1267204820522753,
      "learning_rate": 1.0910395880460274e-05,
      "loss": 0.4112,
      "step": 261
    },
    {
      "epoch": 2.355056179775281,
      "grad_norm": 0.13846022363503335,
      "learning_rate": 1.062353379659383e-05,
      "loss": 0.4182,
      "step": 262
    },
    {
      "epoch": 2.364044943820225,
      "grad_norm": 0.13931951845911503,
      "learning_rate": 1.0339914757355718e-05,
      "loss": 0.4058,
      "step": 263
    },
    {
      "epoch": 2.3730337078651687,
      "grad_norm": 0.14081173124556376,
      "learning_rate": 1.0059570073155953e-05,
      "loss": 0.4224,
      "step": 264
    },
    {
      "epoch": 2.3820224719101124,
      "grad_norm": 0.11818135028408412,
      "learning_rate": 9.782530692928832e-06,
      "loss": 0.4175,
      "step": 265
    },
    {
      "epoch": 2.391011235955056,
      "grad_norm": 0.12675733200213124,
      "learning_rate": 9.508827200716273e-06,
      "loss": 0.4087,
      "step": 266
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.12595759152751707,
      "learning_rate": 9.238489812291469e-06,
      "loss": 0.4081,
      "step": 267
    },
    {
      "epoch": 2.4089887640449437,
      "grad_norm": 0.12527623811173608,
      "learning_rate": 8.971548371823205e-06,
      "loss": 0.4082,
      "step": 268
    },
    {
      "epoch": 2.417977528089888,
      "grad_norm": 0.122998316643011,
      "learning_rate": 8.708032348581144e-06,
      "loss": 0.4073,
      "step": 269
    },
    {
      "epoch": 2.4269662921348316,
      "grad_norm": 0.10564845207022606,
      "learning_rate": 8.447970833682584e-06,
      "loss": 0.4113,
      "step": 270
    },
    {
      "epoch": 2.4359550561797754,
      "grad_norm": 0.10685572250478023,
      "learning_rate": 8.191392536880852e-06,
      "loss": 0.4071,
      "step": 271
    },
    {
      "epoch": 2.444943820224719,
      "grad_norm": 0.11554548264247637,
      "learning_rate": 7.938325783395924e-06,
      "loss": 0.4019,
      "step": 272
    },
    {
      "epoch": 2.453932584269663,
      "grad_norm": 0.11306318481457633,
      "learning_rate": 7.68879851078741e-06,
      "loss": 0.4116,
      "step": 273
    },
    {
      "epoch": 2.4629213483146066,
      "grad_norm": 0.10283238222894554,
      "learning_rate": 7.442838265870347e-06,
      "loss": 0.4097,
      "step": 274
    },
    {
      "epoch": 2.4719101123595504,
      "grad_norm": 0.11264537827542116,
      "learning_rate": 7.2004722016741605e-06,
      "loss": 0.4086,
      "step": 275
    },
    {
      "epoch": 2.4808988764044946,
      "grad_norm": 0.1184428903128185,
      "learning_rate": 6.961727074445055e-06,
      "loss": 0.4048,
      "step": 276
    },
    {
      "epoch": 2.4898876404494383,
      "grad_norm": 0.10381057720149338,
      "learning_rate": 6.726629240692255e-06,
      "loss": 0.406,
      "step": 277
    },
    {
      "epoch": 2.498876404494382,
      "grad_norm": 0.11411996855242165,
      "learning_rate": 6.4952046542783395e-06,
      "loss": 0.4084,
      "step": 278
    },
    {
      "epoch": 2.507865168539326,
      "grad_norm": 0.11085772018759905,
      "learning_rate": 6.2674788635540415e-06,
      "loss": 0.4088,
      "step": 279
    },
    {
      "epoch": 2.5168539325842696,
      "grad_norm": 0.11581754129424132,
      "learning_rate": 6.04347700853781e-06,
      "loss": 0.4091,
      "step": 280
    },
    {
      "epoch": 2.5258426966292133,
      "grad_norm": 0.10915675557144788,
      "learning_rate": 5.823223818140458e-06,
      "loss": 0.4135,
      "step": 281
    },
    {
      "epoch": 2.534831460674157,
      "grad_norm": 0.10472009125683931,
      "learning_rate": 5.606743607435183e-06,
      "loss": 0.4192,
      "step": 282
    },
    {
      "epoch": 2.5438202247191013,
      "grad_norm": 0.10196405389390528,
      "learning_rate": 5.394060274973267e-06,
      "loss": 0.4143,
      "step": 283
    },
    {
      "epoch": 2.552808988764045,
      "grad_norm": 0.09299288202132956,
      "learning_rate": 5.185197300145817e-06,
      "loss": 0.4073,
      "step": 284
    },
    {
      "epoch": 2.561797752808989,
      "grad_norm": 0.10859218890312015,
      "learning_rate": 4.980177740591678e-06,
      "loss": 0.408,
      "step": 285
    },
    {
      "epoch": 2.5707865168539326,
      "grad_norm": 0.1030110328672013,
      "learning_rate": 4.779024229652005e-06,
      "loss": 0.4086,
      "step": 286
    },
    {
      "epoch": 2.5797752808988763,
      "grad_norm": 0.09705963975610957,
      "learning_rate": 4.581758973871609e-06,
      "loss": 0.4053,
      "step": 287
    },
    {
      "epoch": 2.58876404494382,
      "grad_norm": 0.08893242924745905,
      "learning_rate": 4.3884037505474455e-06,
      "loss": 0.4051,
      "step": 288
    },
    {
      "epoch": 2.597752808988764,
      "grad_norm": 0.09476330135285733,
      "learning_rate": 4.198979905324496e-06,
      "loss": 0.4062,
      "step": 289
    },
    {
      "epoch": 2.606741573033708,
      "grad_norm": 0.1125307729702412,
      "learning_rate": 4.0135083498392905e-06,
      "loss": 0.4143,
      "step": 290
    },
    {
      "epoch": 2.6157303370786518,
      "grad_norm": 0.10546064232164497,
      "learning_rate": 3.832009559411338e-06,
      "loss": 0.407,
      "step": 291
    },
    {
      "epoch": 2.6247191011235955,
      "grad_norm": 0.09832745095728368,
      "learning_rate": 3.654503570782755e-06,
      "loss": 0.4094,
      "step": 292
    },
    {
      "epoch": 2.6337078651685393,
      "grad_norm": 0.09906441503157463,
      "learning_rate": 3.481009979906258e-06,
      "loss": 0.408,
      "step": 293
    },
    {
      "epoch": 2.642696629213483,
      "grad_norm": 0.09565058554084314,
      "learning_rate": 3.311547939781887e-06,
      "loss": 0.4064,
      "step": 294
    },
    {
      "epoch": 2.6516853932584272,
      "grad_norm": 0.09922131509032601,
      "learning_rate": 3.14613615834253e-06,
      "loss": 0.4044,
      "step": 295
    },
    {
      "epoch": 2.6606741573033705,
      "grad_norm": 0.10360258898579061,
      "learning_rate": 2.9847928963887198e-06,
      "loss": 0.4054,
      "step": 296
    },
    {
      "epoch": 2.6696629213483147,
      "grad_norm": 0.09576331652157899,
      "learning_rate": 2.8275359655726586e-06,
      "loss": 0.4065,
      "step": 297
    },
    {
      "epoch": 2.6786516853932585,
      "grad_norm": 0.08804843115482651,
      "learning_rate": 2.6743827264319012e-06,
      "loss": 0.4027,
      "step": 298
    },
    {
      "epoch": 2.6876404494382022,
      "grad_norm": 0.0940651708680553,
      "learning_rate": 2.5253500864728155e-06,
      "loss": 0.4048,
      "step": 299
    },
    {
      "epoch": 2.696629213483146,
      "grad_norm": 0.09087393252432437,
      "learning_rate": 2.3804544983040724e-06,
      "loss": 0.4025,
      "step": 300
    },
    {
      "epoch": 2.7056179775280897,
      "grad_norm": 0.09939517236482039,
      "learning_rate": 2.23971195782033e-06,
      "loss": 0.406,
      "step": 301
    },
    {
      "epoch": 2.714606741573034,
      "grad_norm": 0.15328626125954675,
      "learning_rate": 2.1031380024363645e-06,
      "loss": 0.4121,
      "step": 302
    },
    {
      "epoch": 2.7235955056179773,
      "grad_norm": 0.08802675816187339,
      "learning_rate": 1.9707477093717786e-06,
      "loss": 0.4083,
      "step": 303
    },
    {
      "epoch": 2.7325842696629215,
      "grad_norm": 0.08803502781546772,
      "learning_rate": 1.8425556939865696e-06,
      "loss": 0.4146,
      "step": 304
    },
    {
      "epoch": 2.741573033707865,
      "grad_norm": 0.08953466563820135,
      "learning_rate": 1.7185761081676222e-06,
      "loss": 0.4076,
      "step": 305
    },
    {
      "epoch": 2.750561797752809,
      "grad_norm": 0.08902482730609233,
      "learning_rate": 1.5988226387664151e-06,
      "loss": 0.4072,
      "step": 306
    },
    {
      "epoch": 2.7595505617977527,
      "grad_norm": 0.08914869211011264,
      "learning_rate": 1.4833085060880349e-06,
      "loss": 0.4116,
      "step": 307
    },
    {
      "epoch": 2.7685393258426965,
      "grad_norm": 0.0839377567882791,
      "learning_rate": 1.3720464624317108e-06,
      "loss": 0.4028,
      "step": 308
    },
    {
      "epoch": 2.7775280898876407,
      "grad_norm": 0.0860599340838286,
      "learning_rate": 1.2650487906830234e-06,
      "loss": 0.408,
      "step": 309
    },
    {
      "epoch": 2.7865168539325844,
      "grad_norm": 0.08469381729317074,
      "learning_rate": 1.1623273029579195e-06,
      "loss": 0.4124,
      "step": 310
    },
    {
      "epoch": 2.795505617977528,
      "grad_norm": 0.08378981417352441,
      "learning_rate": 1.063893339298674e-06,
      "loss": 0.4153,
      "step": 311
    },
    {
      "epoch": 2.804494382022472,
      "grad_norm": 0.08792748870981947,
      "learning_rate": 9.697577664220303e-07,
      "loss": 0.4128,
      "step": 312
    },
    {
      "epoch": 2.8134831460674157,
      "grad_norm": 0.08428338920338808,
      "learning_rate": 8.799309765195452e-07,
      "loss": 0.4047,
      "step": 313
    },
    {
      "epoch": 2.8224719101123594,
      "grad_norm": 0.08341919328157262,
      "learning_rate": 7.944228861103264e-07,
      "loss": 0.4035,
      "step": 314
    },
    {
      "epoch": 2.831460674157303,
      "grad_norm": 0.08371596387722409,
      "learning_rate": 7.132429349463011e-07,
      "loss": 0.4059,
      "step": 315
    },
    {
      "epoch": 2.8404494382022474,
      "grad_norm": 0.0792325512831226,
      "learning_rate": 6.364000849700791e-07,
      "loss": 0.4039,
      "step": 316
    },
    {
      "epoch": 2.849438202247191,
      "grad_norm": 0.08312852243276637,
      "learning_rate": 5.639028193256257e-07,
      "loss": 0.4067,
      "step": 317
    },
    {
      "epoch": 2.858426966292135,
      "grad_norm": 0.08209771251775913,
      "learning_rate": 4.957591414217344e-07,
      "loss": 0.4077,
      "step": 318
    },
    {
      "epoch": 2.8674157303370786,
      "grad_norm": 0.08673033811521172,
      "learning_rate": 4.3197657404848935e-07,
      "loss": 0.4163,
      "step": 319
    },
    {
      "epoch": 2.8764044943820224,
      "grad_norm": 0.07889124751001583,
      "learning_rate": 3.725621585467698e-07,
      "loss": 0.4074,
      "step": 320
    },
    {
      "epoch": 2.885393258426966,
      "grad_norm": 0.08351100735021297,
      "learning_rate": 3.1752245403092963e-07,
      "loss": 0.4047,
      "step": 321
    },
    {
      "epoch": 2.89438202247191,
      "grad_norm": 0.0841384102489667,
      "learning_rate": 2.6686353666468323e-07,
      "loss": 0.4015,
      "step": 322
    },
    {
      "epoch": 2.903370786516854,
      "grad_norm": 0.08172285090606024,
      "learning_rate": 2.2059099899033098e-07,
      "loss": 0.4007,
      "step": 323
    },
    {
      "epoch": 2.912359550561798,
      "grad_norm": 0.08042881118540746,
      "learning_rate": 1.7870994931135977e-07,
      "loss": 0.4088,
      "step": 324
    },
    {
      "epoch": 2.9213483146067416,
      "grad_norm": 0.08582343232361876,
      "learning_rate": 1.412250111285074e-07,
      "loss": 0.4081,
      "step": 325
    },
    {
      "epoch": 2.9303370786516854,
      "grad_norm": 0.0810315622092791,
      "learning_rate": 1.0814032262935315e-07,
      "loss": 0.4079,
      "step": 326
    },
    {
      "epoch": 2.939325842696629,
      "grad_norm": 0.07928028852132449,
      "learning_rate": 7.945953623146096e-08,
      "loss": 0.4076,
      "step": 327
    },
    {
      "epoch": 2.9483146067415733,
      "grad_norm": 0.08125928562829822,
      "learning_rate": 5.518581817918645e-08,
      "loss": 0.4127,
      "step": 328
    },
    {
      "epoch": 2.9573033707865166,
      "grad_norm": 0.07827386715600733,
      "learning_rate": 3.532184819412532e-08,
      "loss": 0.4074,
      "step": 329
    },
    {
      "epoch": 2.966292134831461,
      "grad_norm": 0.08205445804803274,
      "learning_rate": 1.9869819179292315e-08,
      "loss": 0.4099,
      "step": 330
    },
    {
      "epoch": 2.9752808988764046,
      "grad_norm": 0.08077759984225538,
      "learning_rate": 8.83143697702149e-09,
      "loss": 0.4092,
      "step": 331
    },
    {
      "epoch": 2.9842696629213483,
      "grad_norm": 0.08245700392469553,
      "learning_rate": 2.2079201806501916e-09,
      "loss": 0.4106,
      "step": 332
    },
    {
      "epoch": 2.993258426966292,
      "grad_norm": 0.08229489202929206,
      "learning_rate": 0.0,
      "loss": 0.4064,
      "step": 333
    },
    {
      "epoch": 2.993258426966292,
      "step": 333,
      "total_flos": 7.998180122580484e+18,
      "train_loss": 0.4748834180939305,
      "train_runtime": 32625.0257,
      "train_samples_per_second": 5.234,
      "train_steps_per_second": 0.01
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 333,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.998180122580484e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}