{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.9920127795527156,
  "eval_steps": 500,
  "global_step": 624,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006389776357827476,
      "grad_norm": 2.055291493195234,
      "learning_rate": 3.125e-06,
      "loss": 1.695,
      "step": 1
    },
    {
      "epoch": 0.012779552715654952,
      "grad_norm": 2.0685233500522586,
      "learning_rate": 6.25e-06,
      "loss": 1.6748,
      "step": 2
    },
    {
      "epoch": 0.025559105431309903,
      "grad_norm": 2.325735299422439,
      "learning_rate": 1.25e-05,
      "loss": 1.6964,
      "step": 4
    },
    {
      "epoch": 0.038338658146964855,
      "grad_norm": 0.4729866673863026,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 1.4325,
      "step": 6
    },
    {
      "epoch": 0.051118210862619806,
      "grad_norm": 0.482620239981458,
      "learning_rate": 2.5e-05,
      "loss": 1.3874,
      "step": 8
    },
    {
      "epoch": 0.06389776357827476,
      "grad_norm": 1.6728433474079003,
      "learning_rate": 3.125e-05,
      "loss": 1.4689,
      "step": 10
    },
    {
      "epoch": 0.07667731629392971,
      "grad_norm": 0.3405987431283081,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 1.3127,
      "step": 12
    },
    {
      "epoch": 0.08945686900958466,
      "grad_norm": 0.2323496464888272,
      "learning_rate": 4.375e-05,
      "loss": 1.2639,
      "step": 14
    },
    {
      "epoch": 0.10223642172523961,
      "grad_norm": 0.18809974511784008,
      "learning_rate": 5e-05,
      "loss": 1.2401,
      "step": 16
    },
    {
      "epoch": 0.11501597444089456,
      "grad_norm": 0.18997340619225084,
      "learning_rate": 5.6250000000000005e-05,
      "loss": 1.2084,
      "step": 18
    },
    {
      "epoch": 0.12779552715654952,
      "grad_norm": 0.15504216343509883,
      "learning_rate": 6.25e-05,
      "loss": 1.1855,
      "step": 20
    },
    {
      "epoch": 0.14057507987220447,
      "grad_norm": 0.12848416587626313,
      "learning_rate": 6.875e-05,
      "loss": 1.146,
      "step": 22
    },
    {
      "epoch": 0.15335463258785942,
      "grad_norm": 0.09889252813730416,
      "learning_rate": 7.500000000000001e-05,
      "loss": 1.1357,
      "step": 24
    },
    {
      "epoch": 0.16613418530351437,
      "grad_norm": 0.09024188902019939,
      "learning_rate": 8.125000000000001e-05,
      "loss": 1.1096,
      "step": 26
    },
    {
      "epoch": 0.17891373801916932,
      "grad_norm": 0.08133676595279006,
      "learning_rate": 8.75e-05,
      "loss": 1.0913,
      "step": 28
    },
    {
      "epoch": 0.19169329073482427,
      "grad_norm": 0.0978463769637292,
      "learning_rate": 9.375e-05,
      "loss": 1.0679,
      "step": 30
    },
    {
      "epoch": 0.20447284345047922,
      "grad_norm": 0.07943889170723487,
      "learning_rate": 0.0001,
      "loss": 1.075,
      "step": 32
    },
    {
      "epoch": 0.21725239616613418,
      "grad_norm": 0.08240884428512509,
      "learning_rate": 9.99971838728789e-05,
      "loss": 1.075,
      "step": 34
    },
    {
      "epoch": 0.23003194888178913,
      "grad_norm": 0.08253986997481327,
      "learning_rate": 9.998873580873848e-05,
      "loss": 1.0652,
      "step": 36
    },
    {
      "epoch": 0.24281150159744408,
      "grad_norm": 0.07954648039103362,
      "learning_rate": 9.997465675921163e-05,
      "loss": 1.0519,
      "step": 38
    },
    {
      "epoch": 0.25559105431309903,
      "grad_norm": 0.0776223200815433,
      "learning_rate": 9.995494831023409e-05,
      "loss": 1.0094,
      "step": 40
    },
    {
      "epoch": 0.268370607028754,
      "grad_norm": 0.08000844411167178,
      "learning_rate": 9.992961268186573e-05,
      "loss": 1.0074,
      "step": 42
    },
    {
      "epoch": 0.28115015974440893,
      "grad_norm": 0.0689657212250583,
      "learning_rate": 9.989865272804063e-05,
      "loss": 1.0087,
      "step": 44
    },
    {
      "epoch": 0.2939297124600639,
      "grad_norm": 0.0722150479128947,
      "learning_rate": 9.986207193624536e-05,
      "loss": 1.0067,
      "step": 46
    },
    {
      "epoch": 0.30670926517571884,
      "grad_norm": 0.06646168454668608,
      "learning_rate": 9.981987442712633e-05,
      "loss": 0.9837,
      "step": 48
    },
    {
      "epoch": 0.3194888178913738,
      "grad_norm": 0.06815852582234988,
      "learning_rate": 9.977206495402554e-05,
      "loss": 1.0024,
      "step": 50
    },
    {
      "epoch": 0.33226837060702874,
      "grad_norm": 0.07469571057420442,
      "learning_rate": 9.971864890244513e-05,
      "loss": 0.9606,
      "step": 52
    },
    {
      "epoch": 0.3450479233226837,
      "grad_norm": 0.07160841663430713,
      "learning_rate": 9.965963228944078e-05,
      "loss": 0.9681,
      "step": 54
    },
    {
      "epoch": 0.35782747603833864,
      "grad_norm": 0.06954866095292117,
      "learning_rate": 9.959502176294383e-05,
      "loss": 0.951,
      "step": 56
    },
    {
      "epoch": 0.3706070287539936,
      "grad_norm": 0.06598684065212063,
      "learning_rate": 9.95248246010126e-05,
      "loss": 0.9501,
      "step": 58
    },
    {
      "epoch": 0.38338658146964855,
      "grad_norm": 0.12103302407814338,
      "learning_rate": 9.944904871101228e-05,
      "loss": 0.9713,
      "step": 60
    },
    {
      "epoch": 0.3961661341853035,
      "grad_norm": 0.07330981053456032,
      "learning_rate": 9.936770262872443e-05,
      "loss": 0.9283,
      "step": 62
    },
    {
      "epoch": 0.40894568690095845,
      "grad_norm": 0.06537535724415816,
      "learning_rate": 9.928079551738543e-05,
      "loss": 0.9118,
      "step": 64
    },
    {
      "epoch": 0.4217252396166134,
      "grad_norm": 0.07457609795137939,
      "learning_rate": 9.918833716665419e-05,
      "loss": 0.9279,
      "step": 66
    },
    {
      "epoch": 0.43450479233226835,
      "grad_norm": 0.07491122165043795,
      "learning_rate": 9.909033799150946e-05,
      "loss": 0.935,
      "step": 68
    },
    {
      "epoch": 0.4472843450479233,
      "grad_norm": 0.06781283989008571,
      "learning_rate": 9.898680903107666e-05,
      "loss": 0.9361,
      "step": 70
    },
    {
      "epoch": 0.46006389776357826,
      "grad_norm": 0.07160916695151898,
      "learning_rate": 9.887776194738432e-05,
      "loss": 0.9159,
      "step": 72
    },
    {
      "epoch": 0.4728434504792332,
      "grad_norm": 0.0681941013678725,
      "learning_rate": 9.876320902405042e-05,
      "loss": 0.8779,
      "step": 74
    },
    {
      "epoch": 0.48562300319488816,
      "grad_norm": 0.07482319269062407,
      "learning_rate": 9.864316316489873e-05,
      "loss": 0.8825,
      "step": 76
    },
    {
      "epoch": 0.4984025559105431,
      "grad_norm": 0.08697975313543096,
      "learning_rate": 9.851763789250525e-05,
      "loss": 0.922,
      "step": 78
    },
    {
      "epoch": 0.5111821086261981,
      "grad_norm": 0.09978612068745818,
      "learning_rate": 9.838664734667495e-05,
      "loss": 0.8894,
      "step": 80
    },
    {
      "epoch": 0.5239616613418531,
      "grad_norm": 0.09384667638421956,
      "learning_rate": 9.825020628284896e-05,
      "loss": 0.8593,
      "step": 82
    },
    {
      "epoch": 0.536741214057508,
      "grad_norm": 0.06932081799385038,
      "learning_rate": 9.810833007044247e-05,
      "loss": 0.8662,
      "step": 84
    },
    {
      "epoch": 0.549520766773163,
      "grad_norm": 0.10358699944795004,
      "learning_rate": 9.796103469111351e-05,
      "loss": 0.8723,
      "step": 86
    },
    {
      "epoch": 0.5623003194888179,
      "grad_norm": 0.07169243369499742,
      "learning_rate": 9.780833673696254e-05,
      "loss": 0.8482,
      "step": 88
    },
    {
      "epoch": 0.5750798722044729,
      "grad_norm": 0.1050406308556227,
      "learning_rate": 9.76502534086636e-05,
      "loss": 0.8496,
      "step": 90
    },
    {
      "epoch": 0.5878594249201278,
      "grad_norm": 0.07201905690967678,
      "learning_rate": 9.74868025135266e-05,
      "loss": 0.8291,
      "step": 92
    },
    {
      "epoch": 0.6006389776357828,
      "grad_norm": 1.2625349021090781,
      "learning_rate": 9.731800246349148e-05,
      "loss": 0.8503,
      "step": 94
    },
    {
      "epoch": 0.6134185303514377,
      "grad_norm": 0.17981258022070712,
      "learning_rate": 9.714387227305422e-05,
      "loss": 0.8231,
      "step": 96
    },
    {
      "epoch": 0.6261980830670927,
      "grad_norm": 0.07561478832740967,
      "learning_rate": 9.696443155712486e-05,
      "loss": 0.8119,
      "step": 98
    },
    {
      "epoch": 0.6389776357827476,
      "grad_norm": 0.08195686915168865,
      "learning_rate": 9.67797005288181e-05,
      "loss": 0.7926,
      "step": 100
    },
    {
      "epoch": 0.6517571884984026,
      "grad_norm": 0.0890476280007116,
      "learning_rate": 9.65896999971763e-05,
      "loss": 0.8039,
      "step": 102
    },
    {
      "epoch": 0.6645367412140575,
      "grad_norm": 0.07738578891457887,
      "learning_rate": 9.639445136482548e-05,
      "loss": 0.7721,
      "step": 104
    },
    {
      "epoch": 0.6773162939297125,
      "grad_norm": 0.0743037172920425,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.794,
      "step": 106
    },
    {
      "epoch": 0.6900958466453674,
      "grad_norm": 0.08803835897602165,
      "learning_rate": 9.598829836188694e-05,
      "loss": 0.7721,
      "step": 108
    },
    {
      "epoch": 0.7028753993610224,
      "grad_norm": 0.07702819696223887,
      "learning_rate": 9.577743974243874e-05,
      "loss": 0.7765,
      "step": 110
    },
    {
      "epoch": 0.7156549520766773,
      "grad_norm": 0.07473535070111323,
      "learning_rate": 9.55614245194068e-05,
      "loss": 0.7598,
      "step": 112
    },
    {
      "epoch": 0.7284345047923323,
      "grad_norm": 0.08433756541496004,
      "learning_rate": 9.534027702584425e-05,
      "loss": 0.7727,
      "step": 114
    },
    {
      "epoch": 0.7412140575079872,
      "grad_norm": 0.07483257658817612,
      "learning_rate": 9.511402217292926e-05,
      "loss": 0.7465,
      "step": 116
    },
    {
      "epoch": 0.7539936102236422,
      "grad_norm": 0.0880318685591304,
      "learning_rate": 9.488268544715896e-05,
      "loss": 0.7321,
      "step": 118
    },
    {
      "epoch": 0.7667731629392971,
      "grad_norm": 0.07719604899450865,
      "learning_rate": 9.464629290747842e-05,
      "loss": 0.7624,
      "step": 120
    },
    {
      "epoch": 0.7795527156549521,
      "grad_norm": 0.0733176421376437,
      "learning_rate": 9.440487118234535e-05,
      "loss": 0.6975,
      "step": 122
    },
    {
      "epoch": 0.792332268370607,
      "grad_norm": 0.07051701385784455,
      "learning_rate": 9.415844746673047e-05,
      "loss": 0.7127,
      "step": 124
    },
    {
      "epoch": 0.805111821086262,
      "grad_norm": 0.0729787562181977,
      "learning_rate": 9.390704951905411e-05,
      "loss": 0.7503,
      "step": 126
    },
    {
      "epoch": 0.8178913738019169,
      "grad_norm": 0.07128874732953779,
      "learning_rate": 9.365070565805941e-05,
      "loss": 0.6941,
      "step": 128
    },
    {
      "epoch": 0.8306709265175719,
      "grad_norm": 0.07804844381711577,
      "learning_rate": 9.338944475962237e-05,
      "loss": 0.7197,
      "step": 130
    },
    {
      "epoch": 0.8434504792332268,
      "grad_norm": 0.08207580744924538,
      "learning_rate": 9.312329625349902e-05,
      "loss": 0.7134,
      "step": 132
    },
    {
      "epoch": 0.8562300319488818,
      "grad_norm": 0.10268159904999394,
      "learning_rate": 9.285229012001047e-05,
      "loss": 0.705,
      "step": 134
    },
    {
      "epoch": 0.8690095846645367,
      "grad_norm": 0.07097527094154266,
      "learning_rate": 9.257645688666556e-05,
      "loss": 0.7036,
      "step": 136
    },
    {
      "epoch": 0.8817891373801917,
      "grad_norm": 0.07284443178958877,
      "learning_rate": 9.22958276247223e-05,
      "loss": 0.7313,
      "step": 138
    },
    {
      "epoch": 0.8945686900958466,
      "grad_norm": 0.07294697279525543,
      "learning_rate": 9.201043394568773e-05,
      "loss": 0.6847,
      "step": 140
    },
    {
      "epoch": 0.9073482428115016,
      "grad_norm": 0.0725032039002937,
      "learning_rate": 9.172030799775699e-05,
      "loss": 0.6877,
      "step": 142
    },
    {
      "epoch": 0.9201277955271565,
      "grad_norm": 0.06708836437156662,
      "learning_rate": 9.142548246219212e-05,
      "loss": 0.6837,
      "step": 144
    },
    {
      "epoch": 0.9329073482428115,
      "grad_norm": 0.07361178534656698,
      "learning_rate": 9.112599054964057e-05,
      "loss": 0.6522,
      "step": 146
    },
    {
      "epoch": 0.9456869009584664,
      "grad_norm": 0.06961060997060975,
      "learning_rate": 9.082186599639428e-05,
      "loss": 0.6732,
      "step": 148
    },
    {
      "epoch": 0.9584664536741214,
      "grad_norm": 0.06369267112915664,
      "learning_rate": 9.051314306058933e-05,
      "loss": 0.6615,
      "step": 150
    },
    {
      "epoch": 0.9712460063897763,
      "grad_norm": 0.06667729772792583,
      "learning_rate": 9.019985651834703e-05,
      "loss": 0.6742,
      "step": 152
    },
    {
      "epoch": 0.9840255591054313,
      "grad_norm": 0.07052786453330319,
      "learning_rate": 8.988204165985649e-05,
      "loss": 0.6365,
      "step": 154
    },
    {
      "epoch": 0.9968051118210862,
      "grad_norm": 0.06352217971127558,
      "learning_rate": 8.955973428539944e-05,
      "loss": 0.6531,
      "step": 156
    },
    {
      "epoch": 1.011182108626198,
      "grad_norm": 0.0907023898699884,
      "learning_rate": 8.923297070131737e-05,
      "loss": 0.6986,
      "step": 158
    },
    {
      "epoch": 1.023961661341853,
      "grad_norm": 0.06588723514264389,
      "learning_rate": 8.890178771592199e-05,
      "loss": 0.4221,
      "step": 160
    },
    {
      "epoch": 1.036741214057508,
      "grad_norm": 0.07457104912562523,
      "learning_rate": 8.856622263534875e-05,
      "loss": 0.4375,
      "step": 162
    },
    {
      "epoch": 1.049520766773163,
      "grad_norm": 0.08716030746078077,
      "learning_rate": 8.822631325935463e-05,
      "loss": 0.4633,
      "step": 164
    },
    {
      "epoch": 1.0623003194888179,
      "grad_norm": 0.07564657660605784,
      "learning_rate": 8.788209787706015e-05,
      "loss": 0.4149,
      "step": 166
    },
    {
      "epoch": 1.0750798722044728,
      "grad_norm": 0.2601478494309565,
      "learning_rate": 8.753361526263621e-05,
      "loss": 0.4644,
      "step": 168
    },
    {
      "epoch": 1.0878594249201279,
      "grad_norm": 0.07236244361689621,
      "learning_rate": 8.718090467093654e-05,
      "loss": 0.445,
      "step": 170
    },
    {
      "epoch": 1.1006389776357828,
      "grad_norm": 0.07360308087849284,
      "learning_rate": 8.682400583307562e-05,
      "loss": 0.4189,
      "step": 172
    },
    {
      "epoch": 1.1134185303514377,
      "grad_norm": 0.06934965586236702,
      "learning_rate": 8.646295895195333e-05,
      "loss": 0.4168,
      "step": 174
    },
    {
      "epoch": 1.1261980830670926,
      "grad_norm": 0.06652725595095291,
      "learning_rate": 8.609780469772623e-05,
      "loss": 0.4332,
      "step": 176
    },
    {
      "epoch": 1.1389776357827477,
      "grad_norm": 0.06493423808775205,
      "learning_rate": 8.572858420322627e-05,
      "loss": 0.4126,
      "step": 178
    },
    {
      "epoch": 1.1517571884984026,
      "grad_norm": 0.07224306242862681,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.4639,
      "step": 180
    },
    {
      "epoch": 1.1645367412140575,
      "grad_norm": 0.06325420247080109,
      "learning_rate": 8.497811131026046e-05,
      "loss": 0.4097,
      "step": 182
    },
    {
      "epoch": 1.1773162939297124,
      "grad_norm": 0.05960690196531746,
      "learning_rate": 8.459694344887732e-05,
      "loss": 0.4258,
      "step": 184
    },
    {
      "epoch": 1.1900958466453675,
      "grad_norm": 0.06526403248406679,
      "learning_rate": 8.421187841186402e-05,
      "loss": 0.4453,
      "step": 186
    },
    {
      "epoch": 1.2028753993610224,
      "grad_norm": 0.06754636177295095,
      "learning_rate": 8.382295957490436e-05,
      "loss": 0.4277,
      "step": 188
    },
    {
      "epoch": 1.2156549520766773,
      "grad_norm": 0.11883404710840821,
      "learning_rate": 8.343023074779368e-05,
      "loss": 0.4386,
      "step": 190
    },
    {
      "epoch": 1.2284345047923322,
      "grad_norm": 0.07793571463351197,
      "learning_rate": 8.303373616950408e-05,
      "loss": 0.4072,
      "step": 192
    },
    {
      "epoch": 1.2412140575079873,
      "grad_norm": 0.06518657342856102,
      "learning_rate": 8.263352050320094e-05,
      "loss": 0.4396,
      "step": 194
    },
    {
      "epoch": 1.2539936102236422,
      "grad_norm": 0.05974282037032855,
      "learning_rate": 8.222962883121196e-05,
      "loss": 0.4016,
      "step": 196
    },
    {
      "epoch": 1.266773162939297,
      "grad_norm": 0.0693639502217822,
      "learning_rate": 8.182210664994878e-05,
      "loss": 0.3808,
      "step": 198
    },
    {
      "epoch": 1.279552715654952,
      "grad_norm": 0.06127831754623801,
      "learning_rate": 8.141099986478212e-05,
      "loss": 0.3961,
      "step": 200
    },
    {
      "epoch": 1.292332268370607,
      "grad_norm": 0.06755312065722066,
      "learning_rate": 8.099635478487064e-05,
      "loss": 0.3894,
      "step": 202
    },
    {
      "epoch": 1.305111821086262,
      "grad_norm": 0.0584212869146413,
      "learning_rate": 8.057821811794458e-05,
      "loss": 0.414,
      "step": 204
    },
    {
      "epoch": 1.317891373801917,
      "grad_norm": 0.05983512956529008,
      "learning_rate": 8.015663696504422e-05,
      "loss": 0.3634,
      "step": 206
    },
    {
      "epoch": 1.330670926517572,
      "grad_norm": 0.05778218969166584,
      "learning_rate": 7.973165881521434e-05,
      "loss": 0.4233,
      "step": 208
    },
    {
      "epoch": 1.343450479233227,
      "grad_norm": 0.058310021079803646,
      "learning_rate": 7.930333154015466e-05,
      "loss": 0.4061,
      "step": 210
    },
    {
      "epoch": 1.3562300319488818,
      "grad_norm": 0.0642143238679532,
      "learning_rate": 7.88717033888274e-05,
      "loss": 0.4083,
      "step": 212
    },
    {
      "epoch": 1.3690095846645367,
      "grad_norm": 0.05656381877721736,
      "learning_rate": 7.843682298202235e-05,
      "loss": 0.4033,
      "step": 214
    },
    {
      "epoch": 1.3817891373801916,
      "grad_norm": 0.05518190162844295,
      "learning_rate": 7.799873930687978e-05,
      "loss": 0.3953,
      "step": 216
    },
    {
      "epoch": 1.3945686900958467,
      "grad_norm": 0.05903661851778338,
      "learning_rate": 7.755750171137246e-05,
      "loss": 0.4096,
      "step": 218
    },
    {
      "epoch": 1.4073482428115016,
      "grad_norm": 0.05833074145436464,
      "learning_rate": 7.711315989874677e-05,
      "loss": 0.4151,
      "step": 220
    },
    {
      "epoch": 1.4201277955271565,
      "grad_norm": 0.05919878363690307,
      "learning_rate": 7.666576392192389e-05,
      "loss": 0.39,
      "step": 222
    },
    {
      "epoch": 1.4329073482428116,
      "grad_norm": 0.05913664327254173,
      "learning_rate": 7.621536417786159e-05,
      "loss": 0.4005,
      "step": 224
    },
    {
      "epoch": 1.4456869009584665,
      "grad_norm": 0.0640842931075253,
      "learning_rate": 7.576201140187727e-05,
      "loss": 0.4165,
      "step": 226
    },
    {
      "epoch": 1.4584664536741214,
      "grad_norm": 0.062131879810909965,
      "learning_rate": 7.530575666193283e-05,
      "loss": 0.3891,
      "step": 228
    },
    {
      "epoch": 1.4712460063897763,
      "grad_norm": 0.06992276137309804,
      "learning_rate": 7.484665135288213e-05,
      "loss": 0.3971,
      "step": 230
    },
    {
      "epoch": 1.4840255591054312,
      "grad_norm": 0.06078790664861669,
      "learning_rate": 7.438474719068173e-05,
      "loss": 0.3961,
      "step": 232
    },
    {
      "epoch": 1.4968051118210863,
      "grad_norm": 0.06922648734675908,
      "learning_rate": 7.392009620656513e-05,
      "loss": 0.4331,
      "step": 234
    },
    {
      "epoch": 1.5095846645367412,
      "grad_norm": 0.05766139832102871,
      "learning_rate": 7.345275074118185e-05,
      "loss": 0.4182,
      "step": 236
    },
    {
      "epoch": 1.5223642172523961,
      "grad_norm": 0.06292873231888371,
      "learning_rate": 7.298276343870151e-05,
      "loss": 0.4061,
      "step": 238
    },
    {
      "epoch": 1.5351437699680512,
      "grad_norm": 0.06000860844713537,
      "learning_rate": 7.251018724088367e-05,
      "loss": 0.4023,
      "step": 240
    },
    {
      "epoch": 1.547923322683706,
      "grad_norm": 0.0585777107714916,
      "learning_rate": 7.203507538111423e-05,
      "loss": 0.3855,
      "step": 242
    },
    {
      "epoch": 1.560702875399361,
      "grad_norm": 0.0571671995255021,
      "learning_rate": 7.155748137840892e-05,
      "loss": 0.3951,
      "step": 244
    },
    {
      "epoch": 1.573482428115016,
      "grad_norm": 0.053447175708899994,
      "learning_rate": 7.107745903138472e-05,
      "loss": 0.3745,
      "step": 246
    },
    {
      "epoch": 1.5862619808306708,
      "grad_norm": 0.055736902711725635,
      "learning_rate": 7.059506241219965e-05,
      "loss": 0.3911,
      "step": 248
    },
    {
      "epoch": 1.599041533546326,
      "grad_norm": 0.05715355824554817,
      "learning_rate": 7.011034586046176e-05,
      "loss": 0.4043,
      "step": 250
    },
    {
      "epoch": 1.6118210862619808,
      "grad_norm": 0.06030447320081754,
      "learning_rate": 6.962336397710819e-05,
      "loss": 0.3899,
      "step": 252
    },
    {
      "epoch": 1.6246006389776357,
      "grad_norm": 0.061239135474291606,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.4246,
      "step": 254
    },
    {
      "epoch": 1.6373801916932909,
      "grad_norm": 0.05695235071864785,
      "learning_rate": 6.864282388901544e-05,
      "loss": 0.3953,
      "step": 256
    },
    {
      "epoch": 1.6501597444089455,
      "grad_norm": 0.05308868251491366,
      "learning_rate": 6.814937613729766e-05,
      "loss": 0.4103,
      "step": 258
    },
    {
      "epoch": 1.6629392971246006,
      "grad_norm": 0.054046791633493914,
      "learning_rate": 6.765388394756504e-05,
      "loss": 0.4059,
      "step": 260
    },
    {
      "epoch": 1.6757188498402555,
      "grad_norm": 0.05148697040730548,
      "learning_rate": 6.715640313457733e-05,
      "loss": 0.3767,
      "step": 262
    },
    {
      "epoch": 1.6884984025559104,
      "grad_norm": 0.05318569591896447,
      "learning_rate": 6.665698973710288e-05,
      "loss": 0.3708,
      "step": 264
    },
    {
      "epoch": 1.7012779552715656,
      "grad_norm": 0.05196719070381999,
      "learning_rate": 6.615570001160626e-05,
      "loss": 0.4042,
      "step": 266
    },
    {
      "epoch": 1.7140575079872205,
      "grad_norm": 0.05632881769869459,
      "learning_rate": 6.565259042591113e-05,
      "loss": 0.3987,
      "step": 268
    },
    {
      "epoch": 1.7268370607028753,
      "grad_norm": 0.05470059818193366,
      "learning_rate": 6.514771765283942e-05,
      "loss": 0.3973,
      "step": 270
    },
    {
      "epoch": 1.7396166134185305,
      "grad_norm": 0.056351811449582394,
      "learning_rate": 6.464113856382752e-05,
      "loss": 0.3864,
      "step": 272
    },
    {
      "epoch": 1.7523961661341851,
      "grad_norm": 0.05831258279981057,
      "learning_rate": 6.413291022251989e-05,
      "loss": 0.4041,
      "step": 274
    },
    {
      "epoch": 1.7651757188498403,
      "grad_norm": 0.053467450310740065,
      "learning_rate": 6.362308987834115e-05,
      "loss": 0.3814,
      "step": 276
    },
    {
      "epoch": 1.7779552715654952,
      "grad_norm": 0.051287152623381335,
      "learning_rate": 6.311173496004723e-05,
      "loss": 0.395,
      "step": 278
    },
    {
      "epoch": 1.79073482428115,
      "grad_norm": 0.05429714498773308,
      "learning_rate": 6.259890306925627e-05,
      "loss": 0.3821,
      "step": 280
    },
    {
      "epoch": 1.8035143769968052,
      "grad_norm": 0.057523653580626326,
      "learning_rate": 6.208465197396013e-05,
      "loss": 0.3984,
      "step": 282
    },
    {
      "epoch": 1.81629392971246,
      "grad_norm": 0.05724842136937287,
      "learning_rate": 6.156903960201709e-05,
      "loss": 0.4181,
      "step": 284
    },
    {
      "epoch": 1.829073482428115,
      "grad_norm": 0.052227309043480996,
      "learning_rate": 6.105212403462651e-05,
      "loss": 0.4049,
      "step": 286
    },
    {
      "epoch": 1.84185303514377,
      "grad_norm": 0.04967908325326877,
      "learning_rate": 6.0533963499786314e-05,
      "loss": 0.4117,
      "step": 288
    },
    {
      "epoch": 1.854632587859425,
      "grad_norm": 0.05539898234566285,
      "learning_rate": 6.001461636573397e-05,
      "loss": 0.4006,
      "step": 290
    },
    {
      "epoch": 1.8674121405750799,
      "grad_norm": 0.05795414669880149,
      "learning_rate": 5.949414113437142e-05,
      "loss": 0.386,
      "step": 292
    },
    {
      "epoch": 1.880191693290735,
      "grad_norm": 0.050446841270231885,
      "learning_rate": 5.897259643467527e-05,
      "loss": 0.3842,
      "step": 294
    },
    {
      "epoch": 1.8929712460063897,
      "grad_norm": 0.052453051506198604,
      "learning_rate": 5.8450041016092464e-05,
      "loss": 0.3525,
      "step": 296
    },
    {
      "epoch": 1.9057507987220448,
      "grad_norm": 0.052803823491155276,
      "learning_rate": 5.792653374192245e-05,
      "loss": 0.3963,
      "step": 298
    },
    {
      "epoch": 1.9185303514376997,
      "grad_norm": 0.05180901601155745,
      "learning_rate": 5.7402133582686576e-05,
      "loss": 0.3798,
      "step": 300
    },
    {
      "epoch": 1.9313099041533546,
      "grad_norm": 0.05166645429890597,
      "learning_rate": 5.6876899609485256e-05,
      "loss": 0.3838,
      "step": 302
    },
    {
      "epoch": 1.9440894568690097,
      "grad_norm": 0.05306354741968808,
      "learning_rate": 5.6350890987343944e-05,
      "loss": 0.4165,
      "step": 304
    },
    {
      "epoch": 1.9568690095846646,
      "grad_norm": 0.0860975722690725,
      "learning_rate": 5.582416696854853e-05,
      "loss": 0.3737,
      "step": 306
    },
    {
      "epoch": 1.9696485623003195,
      "grad_norm": 0.05323286133666828,
      "learning_rate": 5.5296786885970805e-05,
      "loss": 0.3889,
      "step": 308
    },
    {
      "epoch": 1.9824281150159746,
      "grad_norm": 0.05299665331057226,
      "learning_rate": 5.476881014638491e-05,
      "loss": 0.3896,
      "step": 310
    },
    {
      "epoch": 1.9952076677316293,
      "grad_norm": 0.05157945275339266,
      "learning_rate": 5.4240296223775465e-05,
      "loss": 0.3637,
      "step": 312
    },
    {
      "epoch": 2.009584664536741,
      "grad_norm": 0.09139947660133817,
      "learning_rate": 5.3711304652638126e-05,
      "loss": 0.3775,
      "step": 314
    },
    {
      "epoch": 2.022364217252396,
      "grad_norm": 0.10130414532724454,
      "learning_rate": 5.318189502127332e-05,
      "loss": 0.2112,
      "step": 316
    },
    {
      "epoch": 2.0351437699680512,
      "grad_norm": 0.0633333619180165,
      "learning_rate": 5.265212696507387e-05,
      "loss": 0.2004,
      "step": 318
    },
    {
      "epoch": 2.047923322683706,
      "grad_norm": 0.0668276114954086,
      "learning_rate": 5.212206015980742e-05,
      "loss": 0.2019,
      "step": 320
    },
    {
      "epoch": 2.060702875399361,
      "grad_norm": 0.05942503367303514,
      "learning_rate": 5.159175431489424e-05,
      "loss": 0.1978,
      "step": 322
    },
    {
      "epoch": 2.073482428115016,
      "grad_norm": 0.07284145764738766,
      "learning_rate": 5.1061269166681183e-05,
      "loss": 0.1935,
      "step": 324
    },
    {
      "epoch": 2.086261980830671,
      "grad_norm": 0.052260140697323494,
      "learning_rate": 5.053066447171282e-05,
      "loss": 0.1854,
      "step": 326
    },
    {
      "epoch": 2.099041533546326,
      "grad_norm": 0.05754923159453662,
      "learning_rate": 5e-05,
      "loss": 0.1965,
      "step": 328
    },
    {
      "epoch": 2.1118210862619806,
      "grad_norm": 0.05500397186780569,
      "learning_rate": 4.94693355282872e-05,
      "loss": 0.1827,
      "step": 330
    },
    {
      "epoch": 2.1246006389776357,
      "grad_norm": 0.061606661346763424,
      "learning_rate": 4.893873083331882e-05,
      "loss": 0.2008,
      "step": 332
    },
    {
      "epoch": 2.137380191693291,
      "grad_norm": 0.05678242709297541,
      "learning_rate": 4.840824568510579e-05,
      "loss": 0.1853,
      "step": 334
    },
    {
      "epoch": 2.1501597444089455,
      "grad_norm": 0.054080318070508115,
      "learning_rate": 4.78779398401926e-05,
      "loss": 0.1952,
      "step": 336
    },
    {
      "epoch": 2.1629392971246006,
      "grad_norm": 0.057204881343756786,
      "learning_rate": 4.734787303492615e-05,
      "loss": 0.1778,
      "step": 338
    },
    {
      "epoch": 2.1757188498402558,
      "grad_norm": 0.6941487667655994,
      "learning_rate": 4.6818104978726685e-05,
      "loss": 0.219,
      "step": 340
    },
    {
      "epoch": 2.1884984025559104,
      "grad_norm": 0.06999590590614403,
      "learning_rate": 4.628869534736187e-05,
      "loss": 0.181,
      "step": 342
    },
    {
      "epoch": 2.2012779552715656,
      "grad_norm": 0.07558854262088241,
      "learning_rate": 4.575970377622456e-05,
      "loss": 0.2349,
      "step": 344
    },
    {
      "epoch": 2.2140575079872207,
      "grad_norm": 0.07120027160683609,
      "learning_rate": 4.52311898536151e-05,
      "loss": 0.1993,
      "step": 346
    },
    {
      "epoch": 2.2268370607028753,
      "grad_norm": 0.05697032990090494,
      "learning_rate": 4.47032131140292e-05,
      "loss": 0.1739,
      "step": 348
    },
    {
      "epoch": 2.2396166134185305,
      "grad_norm": 0.06092977319118132,
      "learning_rate": 4.4175833031451473e-05,
      "loss": 0.188,
      "step": 350
    },
    {
      "epoch": 2.252396166134185,
      "grad_norm": 0.05900721095602371,
      "learning_rate": 4.364910901265606e-05,
      "loss": 0.1778,
      "step": 352
    },
    {
      "epoch": 2.2651757188498403,
      "grad_norm": 0.08992850669862418,
      "learning_rate": 4.3123100390514756e-05,
      "loss": 0.1838,
      "step": 354
    },
    {
      "epoch": 2.2779552715654954,
      "grad_norm": 0.059213794143429914,
      "learning_rate": 4.2597866417313436e-05,
      "loss": 0.1902,
      "step": 356
    },
    {
      "epoch": 2.29073482428115,
      "grad_norm": 0.051525349318871976,
      "learning_rate": 4.207346625807756e-05,
      "loss": 0.1784,
      "step": 358
    },
    {
      "epoch": 2.303514376996805,
      "grad_norm": 0.055922862481655594,
      "learning_rate": 4.1549958983907555e-05,
      "loss": 0.1827,
      "step": 360
    },
    {
      "epoch": 2.31629392971246,
      "grad_norm": 0.054189632126131766,
      "learning_rate": 4.102740356532473e-05,
      "loss": 0.186,
      "step": 362
    },
    {
      "epoch": 2.329073482428115,
      "grad_norm": 0.06298745746452741,
      "learning_rate": 4.050585886562858e-05,
      "loss": 0.1854,
      "step": 364
    },
    {
      "epoch": 2.34185303514377,
      "grad_norm": 0.06476475169367538,
      "learning_rate": 3.998538363426605e-05,
      "loss": 0.1794,
      "step": 366
    },
    {
      "epoch": 2.3546325878594248,
      "grad_norm": 0.05187178001518817,
      "learning_rate": 3.94660365002137e-05,
      "loss": 0.1817,
      "step": 368
    },
    {
      "epoch": 2.36741214057508,
      "grad_norm": 0.05110076217610542,
      "learning_rate": 3.894787596537352e-05,
      "loss": 0.1757,
      "step": 370
    },
    {
      "epoch": 2.380191693290735,
      "grad_norm": 0.061027606854849537,
      "learning_rate": 3.843096039798293e-05,
      "loss": 0.1888,
      "step": 372
    },
    {
      "epoch": 2.3929712460063897,
      "grad_norm": 0.05689282057128392,
      "learning_rate": 3.791534802603988e-05,
      "loss": 0.1972,
      "step": 374
    },
    {
      "epoch": 2.405750798722045,
      "grad_norm": 0.05144327012401281,
      "learning_rate": 3.740109693074375e-05,
      "loss": 0.1975,
      "step": 376
    },
    {
      "epoch": 2.4185303514377,
      "grad_norm": 0.07243681779987425,
      "learning_rate": 3.68882650399528e-05,
      "loss": 0.1865,
      "step": 378
    },
    {
      "epoch": 2.4313099041533546,
      "grad_norm": 0.11601839655528177,
      "learning_rate": 3.637691012165886e-05,
      "loss": 0.1977,
      "step": 380
    },
    {
      "epoch": 2.4440894568690097,
      "grad_norm": 0.05323975029748036,
      "learning_rate": 3.586708977748012e-05,
      "loss": 0.1873,
      "step": 382
    },
    {
      "epoch": 2.4568690095846644,
      "grad_norm": 0.0499469664737551,
      "learning_rate": 3.5358861436172485e-05,
      "loss": 0.1832,
      "step": 384
    },
    {
      "epoch": 2.4696485623003195,
      "grad_norm": 0.05043024991533826,
      "learning_rate": 3.485228234716058e-05,
      "loss": 0.1821,
      "step": 386
    },
    {
      "epoch": 2.4824281150159746,
      "grad_norm": 0.054685112352780986,
      "learning_rate": 3.434740957408889e-05,
      "loss": 0.1816,
      "step": 388
    },
    {
      "epoch": 2.4952076677316293,
      "grad_norm": 0.057237969167094144,
      "learning_rate": 3.3844299988393755e-05,
      "loss": 0.1909,
      "step": 390
    },
    {
      "epoch": 2.5079872204472844,
      "grad_norm": 0.05134273506646416,
      "learning_rate": 3.334301026289712e-05,
      "loss": 0.1782,
      "step": 392
    },
    {
      "epoch": 2.520766773162939,
      "grad_norm": 0.049993934417102925,
      "learning_rate": 3.284359686542269e-05,
      "loss": 0.1928,
      "step": 394
    },
    {
      "epoch": 2.533546325878594,
      "grad_norm": 0.06457823051474779,
      "learning_rate": 3.234611605243496e-05,
      "loss": 0.196,
      "step": 396
    },
    {
      "epoch": 2.5463258785942493,
      "grad_norm": 0.051805062617152425,
      "learning_rate": 3.1850623862702344e-05,
      "loss": 0.1881,
      "step": 398
    },
    {
      "epoch": 2.559105431309904,
      "grad_norm": 0.049188541484928724,
      "learning_rate": 3.135717611098458e-05,
      "loss": 0.1806,
      "step": 400
    },
    {
      "epoch": 2.571884984025559,
      "grad_norm": 0.05687592017078177,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.1784,
      "step": 402
    },
    {
      "epoch": 2.584664536741214,
      "grad_norm": 0.05098573657706369,
      "learning_rate": 3.0376636022891812e-05,
      "loss": 0.1932,
      "step": 404
    },
    {
      "epoch": 2.597444089456869,
      "grad_norm": 0.052376381772842893,
      "learning_rate": 2.9889654139538246e-05,
      "loss": 0.1889,
      "step": 406
    },
    {
      "epoch": 2.610223642172524,
      "grad_norm": 0.05031660077056393,
      "learning_rate": 2.9404937587800375e-05,
      "loss": 0.1769,
      "step": 408
    },
    {
      "epoch": 2.623003194888179,
      "grad_norm": 0.04930354808056054,
      "learning_rate": 2.8922540968615286e-05,
      "loss": 0.1685,
      "step": 410
    },
    {
      "epoch": 2.635782747603834,
      "grad_norm": 0.06709139465230578,
      "learning_rate": 2.8442518621591086e-05,
      "loss": 0.1785,
      "step": 412
    },
    {
      "epoch": 2.648562300319489,
      "grad_norm": 0.0503489735828908,
      "learning_rate": 2.7964924618885778e-05,
      "loss": 0.1689,
      "step": 414
    },
    {
      "epoch": 2.661341853035144,
      "grad_norm": 0.05047783892143097,
      "learning_rate": 2.748981275911633e-05,
      "loss": 0.1808,
      "step": 416
    },
    {
      "epoch": 2.6741214057507987,
      "grad_norm": 0.04955419672921838,
      "learning_rate": 2.701723656129851e-05,
      "loss": 0.1727,
      "step": 418
    },
    {
      "epoch": 2.686900958466454,
      "grad_norm": 0.04769759775271665,
      "learning_rate": 2.6547249258818164e-05,
      "loss": 0.1708,
      "step": 420
    },
    {
      "epoch": 2.6996805111821085,
      "grad_norm": 0.050324696099913684,
      "learning_rate": 2.607990379343489e-05,
      "loss": 0.1817,
      "step": 422
    },
    {
      "epoch": 2.7124600638977636,
      "grad_norm": 0.05249496210993974,
      "learning_rate": 2.5615252809318284e-05,
      "loss": 0.1836,
      "step": 424
    },
    {
      "epoch": 2.7252396166134183,
      "grad_norm": 0.0472378188955872,
      "learning_rate": 2.5153348647117857e-05,
      "loss": 0.1736,
      "step": 426
    },
    {
      "epoch": 2.7380191693290734,
      "grad_norm": 0.049243154928981264,
      "learning_rate": 2.469424333806718e-05,
      "loss": 0.1675,
      "step": 428
    },
    {
      "epoch": 2.7507987220447285,
      "grad_norm": 0.05096273109137321,
      "learning_rate": 2.4237988598122752e-05,
      "loss": 0.1658,
      "step": 430
    },
    {
      "epoch": 2.763578274760383,
      "grad_norm": 0.0514806212844811,
      "learning_rate": 2.3784635822138424e-05,
      "loss": 0.1922,
      "step": 432
    },
    {
      "epoch": 2.7763578274760383,
      "grad_norm": 0.05006269553229606,
      "learning_rate": 2.333423607807613e-05,
      "loss": 0.1887,
      "step": 434
    },
    {
      "epoch": 2.7891373801916934,
      "grad_norm": 0.04935551516167026,
      "learning_rate": 2.288684010125325e-05,
      "loss": 0.1763,
      "step": 436
    },
    {
      "epoch": 2.801916932907348,
      "grad_norm": 0.05353903496894845,
      "learning_rate": 2.2442498288627556e-05,
      "loss": 0.1944,
      "step": 438
    },
    {
      "epoch": 2.8146964856230032,
      "grad_norm": 0.04697149845887787,
      "learning_rate": 2.2001260693120233e-05,
      "loss": 0.1672,
      "step": 440
    },
    {
      "epoch": 2.8274760383386583,
      "grad_norm": 0.054384654770629585,
      "learning_rate": 2.156317701797766e-05,
      "loss": 0.1807,
      "step": 442
    },
    {
      "epoch": 2.840255591054313,
      "grad_norm": 0.04684823569442938,
      "learning_rate": 2.1128296611172593e-05,
      "loss": 0.171,
      "step": 444
    },
    {
      "epoch": 2.853035143769968,
      "grad_norm": 0.0498371244165766,
      "learning_rate": 2.0696668459845355e-05,
      "loss": 0.1827,
      "step": 446
    },
    {
      "epoch": 2.8658146964856233,
      "grad_norm": 0.04969475724913098,
      "learning_rate": 2.026834118478567e-05,
      "loss": 0.1749,
      "step": 448
    },
    {
      "epoch": 2.878594249201278,
      "grad_norm": 0.051902756416916496,
      "learning_rate": 1.98433630349558e-05,
      "loss": 0.1891,
      "step": 450
    },
    {
      "epoch": 2.891373801916933,
      "grad_norm": 0.05102564026340021,
      "learning_rate": 1.9421781882055444e-05,
      "loss": 0.1849,
      "step": 452
    },
    {
      "epoch": 2.9041533546325877,
      "grad_norm": 0.05200929870376942,
      "learning_rate": 1.9003645215129355e-05,
      "loss": 0.1891,
      "step": 454
    },
    {
      "epoch": 2.916932907348243,
      "grad_norm": 0.05083154953396676,
      "learning_rate": 1.858900013521788e-05,
      "loss": 0.179,
      "step": 456
    },
    {
      "epoch": 2.9297124600638975,
      "grad_norm": 0.049127219472404525,
      "learning_rate": 1.817789335005121e-05,
      "loss": 0.17,
      "step": 458
    },
    {
      "epoch": 2.9424920127795526,
      "grad_norm": 0.049677004679461886,
      "learning_rate": 1.777037116878804e-05,
      "loss": 0.1831,
      "step": 460
    },
    {
      "epoch": 2.9552715654952078,
      "grad_norm": 0.054496479788018075,
      "learning_rate": 1.7366479496799077e-05,
      "loss": 0.1843,
      "step": 462
    },
    {
      "epoch": 2.9680511182108624,
      "grad_norm": 0.04820092295738451,
      "learning_rate": 1.6966263830495936e-05,
      "loss": 0.1685,
      "step": 464
    },
    {
      "epoch": 2.9808306709265175,
      "grad_norm": 0.04915420841884947,
      "learning_rate": 1.656976925220633e-05,
      "loss": 0.1875,
      "step": 466
    },
    {
      "epoch": 2.9936102236421727,
      "grad_norm": 0.07661474807504913,
      "learning_rate": 1.6177040425095662e-05,
      "loss": 0.1891,
      "step": 468
    },
    {
      "epoch": 3.0079872204472844,
      "grad_norm": 0.07655695803476695,
      "learning_rate": 1.5788121588135975e-05,
      "loss": 0.1837,
      "step": 470
    },
    {
      "epoch": 3.0207667731629395,
      "grad_norm": 0.060916330302725,
      "learning_rate": 1.5403056551122697e-05,
      "loss": 0.0872,
      "step": 472
    },
    {
      "epoch": 3.033546325878594,
      "grad_norm": 0.052542395235648506,
      "learning_rate": 1.5021888689739549e-05,
      "loss": 0.0778,
      "step": 474
    },
    {
      "epoch": 3.0463258785942493,
      "grad_norm": 0.20368087770560855,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.102,
      "step": 476
    },
    {
      "epoch": 3.059105431309904,
      "grad_norm": 0.10396707161226072,
      "learning_rate": 1.427141579677374e-05,
      "loss": 0.083,
      "step": 478
    },
    {
      "epoch": 3.071884984025559,
      "grad_norm": 0.04599720220665865,
      "learning_rate": 1.3902195302273779e-05,
      "loss": 0.0757,
      "step": 480
    },
    {
      "epoch": 3.084664536741214,
      "grad_norm": 0.056109340867354925,
      "learning_rate": 1.3537041048046695e-05,
      "loss": 0.081,
      "step": 482
    },
    {
      "epoch": 3.097444089456869,
      "grad_norm": 0.048015102375770044,
      "learning_rate": 1.3175994166924394e-05,
      "loss": 0.0802,
      "step": 484
    },
    {
      "epoch": 3.110223642172524,
      "grad_norm": 0.04645228076024571,
      "learning_rate": 1.2819095329063469e-05,
      "loss": 0.0787,
      "step": 486
    },
    {
      "epoch": 3.123003194888179,
      "grad_norm": 0.04637085498651796,
      "learning_rate": 1.246638473736378e-05,
      "loss": 0.0839,
      "step": 488
    },
    {
      "epoch": 3.135782747603834,
      "grad_norm": 0.05039074009256794,
      "learning_rate": 1.2117902122939861e-05,
      "loss": 0.0812,
      "step": 490
    },
    {
      "epoch": 3.148562300319489,
      "grad_norm": 0.05079569512274489,
      "learning_rate": 1.1773686740645384e-05,
      "loss": 0.0797,
      "step": 492
    },
    {
      "epoch": 3.1613418530351436,
      "grad_norm": 0.04286375870307716,
      "learning_rate": 1.1433777364651271e-05,
      "loss": 0.0737,
      "step": 494
    },
    {
      "epoch": 3.1741214057507987,
      "grad_norm": 0.03982951021947898,
      "learning_rate": 1.1098212284078036e-05,
      "loss": 0.0722,
      "step": 496
    },
    {
      "epoch": 3.186900958466454,
      "grad_norm": 0.0446624849328897,
      "learning_rate": 1.076702929868264e-05,
      "loss": 0.079,
      "step": 498
    },
    {
      "epoch": 3.1996805111821085,
      "grad_norm": 0.04376807908723891,
      "learning_rate": 1.0440265714600572e-05,
      "loss": 0.0837,
      "step": 500
    },
    {
      "epoch": 3.2124600638977636,
      "grad_norm": 0.04087367539850916,
      "learning_rate": 1.0117958340143507e-05,
      "loss": 0.076,
      "step": 502
    },
    {
      "epoch": 3.2252396166134187,
      "grad_norm": 0.04066584417219993,
      "learning_rate": 9.800143481652979e-06,
      "loss": 0.0701,
      "step": 504
    },
    {
      "epoch": 3.2380191693290734,
      "grad_norm": 0.08215263649470263,
      "learning_rate": 9.48685693941067e-06,
      "loss": 0.0776,
      "step": 506
    },
    {
      "epoch": 3.2507987220447285,
      "grad_norm": 0.0437601284673361,
      "learning_rate": 9.17813400360572e-06,
      "loss": 0.0764,
      "step": 508
    },
    {
      "epoch": 3.263578274760383,
      "grad_norm": 0.04382435518426366,
      "learning_rate": 8.874009450359427e-06,
      "loss": 0.0826,
      "step": 510
    },
    {
      "epoch": 3.2763578274760383,
      "grad_norm": 0.04095610913441161,
      "learning_rate": 8.574517537807897e-06,
      "loss": 0.0753,
      "step": 512
    },
    {
      "epoch": 3.2891373801916934,
      "grad_norm": 0.040525949300480126,
      "learning_rate": 8.279692002243027e-06,
      "loss": 0.0764,
      "step": 514
    },
    {
      "epoch": 3.301916932907348,
      "grad_norm": 0.043675586209021296,
      "learning_rate": 7.989566054312287e-06,
      "loss": 0.0817,
      "step": 516
    },
    {
      "epoch": 3.3146964856230032,
      "grad_norm": 0.04319420448553361,
      "learning_rate": 7.704172375277691e-06,
      "loss": 0.0759,
      "step": 518
    },
    {
      "epoch": 3.3274760383386583,
      "grad_norm": 0.044446852802239034,
      "learning_rate": 7.423543113334436e-06,
      "loss": 0.0813,
      "step": 520
    },
    {
      "epoch": 3.340255591054313,
      "grad_norm": 0.09121973616663154,
      "learning_rate": 7.14770987998954e-06,
      "loss": 0.0838,
      "step": 522
    },
    {
      "epoch": 3.353035143769968,
      "grad_norm": 0.05879997473879583,
      "learning_rate": 6.876703746500984e-06,
      "loss": 0.0738,
      "step": 524
    },
    {
      "epoch": 3.365814696485623,
      "grad_norm": 0.04667273388126841,
      "learning_rate": 6.610555240377652e-06,
      "loss": 0.0787,
      "step": 526
    },
    {
      "epoch": 3.378594249201278,
      "grad_norm": 0.042105033545020404,
      "learning_rate": 6.349294341940593e-06,
      "loss": 0.0801,
      "step": 528
    },
    {
      "epoch": 3.391373801916933,
      "grad_norm": 0.0407975465413022,
      "learning_rate": 6.092950480945897e-06,
      "loss": 0.0735,
      "step": 530
    },
    {
      "epoch": 3.4041533546325877,
      "grad_norm": 0.04234912863253251,
      "learning_rate": 5.841552533269534e-06,
      "loss": 0.0772,
      "step": 532
    },
    {
      "epoch": 3.416932907348243,
      "grad_norm": 0.04032120711392374,
      "learning_rate": 5.595128817654638e-06,
      "loss": 0.0749,
      "step": 534
    },
    {
      "epoch": 3.4297124600638975,
      "grad_norm": 0.041050930036482094,
      "learning_rate": 5.353707092521582e-06,
      "loss": 0.0769,
      "step": 536
    },
    {
      "epoch": 3.4424920127795526,
      "grad_norm": 0.043382176933190755,
      "learning_rate": 5.117314552841052e-06,
      "loss": 0.0767,
      "step": 538
    },
    {
      "epoch": 3.4552715654952078,
      "grad_norm": 0.039240502138117625,
      "learning_rate": 4.885977827070748e-06,
      "loss": 0.0721,
      "step": 540
    },
    {
      "epoch": 3.4680511182108624,
      "grad_norm": 0.040812347040587296,
      "learning_rate": 4.659722974155767e-06,
      "loss": 0.1114,
      "step": 542
    },
    {
      "epoch": 3.4808306709265175,
      "grad_norm": 0.0423787622918925,
      "learning_rate": 4.43857548059321e-06,
      "loss": 0.0778,
      "step": 544
    },
    {
      "epoch": 3.4936102236421727,
      "grad_norm": 0.042228923687598445,
      "learning_rate": 4.2225602575612755e-06,
      "loss": 0.0814,
      "step": 546
    },
    {
      "epoch": 3.5063897763578273,
      "grad_norm": 0.0407267289339222,
      "learning_rate": 4.011701638113063e-06,
      "loss": 0.0782,
      "step": 548
    },
    {
      "epoch": 3.5191693290734825,
      "grad_norm": 0.0389855165359938,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0789,
      "step": 550
    },
    {
      "epoch": 3.5319488817891376,
      "grad_norm": 0.040904703617676376,
      "learning_rate": 3.605548635174533e-06,
      "loss": 0.078,
      "step": 552
    },
    {
      "epoch": 3.5447284345047922,
      "grad_norm": 0.04093280012624571,
      "learning_rate": 3.410300002823691e-06,
      "loss": 0.0777,
      "step": 554
    },
    {
      "epoch": 3.5575079872204474,
      "grad_norm": 0.042904856841507744,
      "learning_rate": 3.220299471181898e-06,
      "loss": 0.0757,
      "step": 556
    },
    {
      "epoch": 3.5702875399361025,
      "grad_norm": 0.0436449067886704,
      "learning_rate": 3.035568442875136e-06,
      "loss": 0.0798,
      "step": 558
    },
    {
      "epoch": 3.583067092651757,
      "grad_norm": 0.035664776931118955,
      "learning_rate": 2.85612772694579e-06,
      "loss": 0.0632,
      "step": 560
    },
    {
      "epoch": 3.5958466453674123,
      "grad_norm": 0.03847526723825484,
      "learning_rate": 2.6819975365085237e-06,
      "loss": 0.0744,
      "step": 562
    },
    {
      "epoch": 3.608626198083067,
      "grad_norm": 0.039939236612970476,
      "learning_rate": 2.5131974864734066e-06,
      "loss": 0.0794,
      "step": 564
    },
    {
      "epoch": 3.621405750798722,
      "grad_norm": 0.040388305870748,
      "learning_rate": 2.349746591336405e-06,
      "loss": 0.0718,
      "step": 566
    },
    {
      "epoch": 3.6341853035143767,
      "grad_norm": 0.04232813426430434,
      "learning_rate": 2.191663263037458e-06,
      "loss": 0.0769,
      "step": 568
    },
    {
      "epoch": 3.646964856230032,
      "grad_norm": 0.04213845492527589,
      "learning_rate": 2.0389653088865036e-06,
      "loss": 0.0728,
      "step": 570
    },
    {
      "epoch": 3.659744408945687,
      "grad_norm": 0.04098999517730541,
      "learning_rate": 1.8916699295575324e-06,
      "loss": 0.0724,
      "step": 572
    },
    {
      "epoch": 3.6725239616613417,
      "grad_norm": 0.037533240934183365,
      "learning_rate": 1.7497937171510547e-06,
      "loss": 0.0709,
      "step": 574
    },
    {
      "epoch": 3.6853035143769968,
      "grad_norm": 0.039040304607963414,
      "learning_rate": 1.6133526533250565e-06,
      "loss": 0.0756,
      "step": 576
    },
    {
      "epoch": 3.698083067092652,
      "grad_norm": 0.04065729024121047,
      "learning_rate": 1.4823621074947503e-06,
      "loss": 0.0774,
      "step": 578
    },
    {
      "epoch": 3.7108626198083066,
      "grad_norm": 0.04252602887603373,
      "learning_rate": 1.3568368351012717e-06,
      "loss": 0.0824,
      "step": 580
    },
    {
      "epoch": 3.7236421725239617,
      "grad_norm": 0.04343672134882273,
      "learning_rate": 1.236790975949592e-06,
      "loss": 0.074,
      "step": 582
    },
    {
      "epoch": 3.736421725239617,
      "grad_norm": 0.0403766223584342,
      "learning_rate": 1.1222380526156928e-06,
      "loss": 0.0755,
      "step": 584
    },
    {
      "epoch": 3.7492012779552715,
      "grad_norm": 0.04234625541762105,
      "learning_rate": 1.0131909689233442e-06,
      "loss": 0.0814,
      "step": 586
    },
    {
      "epoch": 3.7619808306709266,
      "grad_norm": 0.03861507912847567,
      "learning_rate": 9.096620084905472e-07,
      "loss": 0.0664,
      "step": 588
    },
    {
      "epoch": 3.7747603833865817,
      "grad_norm": 0.041892733843973705,
      "learning_rate": 8.11662833345822e-07,
      "loss": 0.0832,
      "step": 590
    },
    {
      "epoch": 3.7875399361022364,
      "grad_norm": 0.05413889236863839,
      "learning_rate": 7.192044826145771e-07,
      "loss": 0.0921,
      "step": 592
    },
    {
      "epoch": 3.8003194888178915,
      "grad_norm": 0.04010280150213322,
      "learning_rate": 6.322973712755697e-07,
      "loss": 0.0752,
      "step": 594
    },
    {
      "epoch": 3.813099041533546,
      "grad_norm": 0.04321423418162425,
      "learning_rate": 5.509512889877333e-07,
      "loss": 0.0781,
      "step": 596
    },
    {
      "epoch": 3.8258785942492013,
      "grad_norm": 0.04049679761598481,
      "learning_rate": 4.7517539898741524e-07,
      "loss": 0.0694,
      "step": 598
    },
    {
      "epoch": 3.838658146964856,
      "grad_norm": 0.04258434666712487,
      "learning_rate": 4.049782370561583e-07,
      "loss": 0.0756,
      "step": 600
    },
    {
      "epoch": 3.851437699680511,
      "grad_norm": 0.03927978960342531,
      "learning_rate": 3.4036771055923066e-07,
      "loss": 0.075,
      "step": 602
    },
    {
      "epoch": 3.864217252396166,
      "grad_norm": 0.04093422273122725,
      "learning_rate": 2.813510975548772e-07,
      "loss": 0.0793,
      "step": 604
    },
    {
      "epoch": 3.876996805111821,
      "grad_norm": 0.0433141394014271,
      "learning_rate": 2.2793504597447002e-07,
      "loss": 0.0796,
      "step": 606
    },
    {
      "epoch": 3.889776357827476,
      "grad_norm": 0.04198937065288365,
      "learning_rate": 1.8012557287367392e-07,
      "loss": 0.0753,
      "step": 608
    },
    {
      "epoch": 3.902555910543131,
      "grad_norm": 0.043002763720086865,
      "learning_rate": 1.379280637546443e-07,
      "loss": 0.0917,
      "step": 610
    },
    {
      "epoch": 3.915335463258786,
      "grad_norm": 0.042376326172823574,
      "learning_rate": 1.0134727195937333e-07,
      "loss": 0.0747,
      "step": 612
    },
    {
      "epoch": 3.928115015974441,
      "grad_norm": 0.040600294489722695,
      "learning_rate": 7.038731813426291e-08,
      "loss": 0.0714,
      "step": 614
    },
    {
      "epoch": 3.940894568690096,
      "grad_norm": 0.0395711578920217,
      "learning_rate": 4.5051689765929214e-08,
      "loss": 0.0801,
      "step": 616
    },
    {
      "epoch": 3.9536741214057507,
      "grad_norm": 0.03943811103107873,
      "learning_rate": 2.534324078837802e-08,
      "loss": 0.074,
      "step": 618
    },
    {
      "epoch": 3.966453674121406,
      "grad_norm": 0.041905340316324986,
      "learning_rate": 1.1264191261528557e-08,
      "loss": 0.0771,
      "step": 620
    },
    {
      "epoch": 3.979233226837061,
      "grad_norm": 0.06686218682304546,
      "learning_rate": 2.8161271211024633e-09,
      "loss": 0.0811,
      "step": 622
    },
    {
      "epoch": 3.9920127795527156,
      "grad_norm": 0.03876030687059135,
      "learning_rate": 0.0,
      "loss": 0.076,
      "step": 624
    }
  ],
  "logging_steps": 2,
  "max_steps": 624,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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": 2.5262683933881926e+19,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}