{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 32.03642272949219,
      "learning_rate": 1.9904e-05,
      "loss": 2.5441,
      "step": 25
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.376370429992676,
      "learning_rate": 1.9804000000000002e-05,
      "loss": 2.3283,
      "step": 50
    },
    {
      "epoch": 0.03,
      "grad_norm": 16.60682487487793,
      "learning_rate": 1.9704e-05,
      "loss": 2.246,
      "step": 75
    },
    {
      "epoch": 0.04,
      "grad_norm": 15.391746520996094,
      "learning_rate": 1.9604e-05,
      "loss": 2.2186,
      "step": 100
    },
    {
      "epoch": 0.05,
      "grad_norm": 12.73359203338623,
      "learning_rate": 1.9504e-05,
      "loss": 2.3115,
      "step": 125
    },
    {
      "epoch": 0.06,
      "grad_norm": 12.478874206542969,
      "learning_rate": 1.9404e-05,
      "loss": 2.3705,
      "step": 150
    },
    {
      "epoch": 0.07,
      "grad_norm": 14.907727241516113,
      "learning_rate": 1.9304e-05,
      "loss": 2.2694,
      "step": 175
    },
    {
      "epoch": 0.08,
      "grad_norm": 19.148876190185547,
      "learning_rate": 1.9204e-05,
      "loss": 2.2233,
      "step": 200
    },
    {
      "epoch": 0.09,
      "grad_norm": 11.990699768066406,
      "learning_rate": 1.9104000000000002e-05,
      "loss": 2.2177,
      "step": 225
    },
    {
      "epoch": 0.1,
      "grad_norm": 11.667814254760742,
      "learning_rate": 1.9004000000000003e-05,
      "loss": 2.2965,
      "step": 250
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.88569164276123,
      "learning_rate": 1.8904000000000004e-05,
      "loss": 2.2782,
      "step": 275
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.427560806274414,
      "learning_rate": 1.8804e-05,
      "loss": 2.319,
      "step": 300
    },
    {
      "epoch": 0.13,
      "grad_norm": 16.95577049255371,
      "learning_rate": 1.8704000000000003e-05,
      "loss": 2.2264,
      "step": 325
    },
    {
      "epoch": 0.14,
      "grad_norm": 13.226414680480957,
      "learning_rate": 1.8604000000000003e-05,
      "loss": 2.2142,
      "step": 350
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.91146183013916,
      "learning_rate": 1.8504e-05,
      "loss": 2.2503,
      "step": 375
    },
    {
      "epoch": 0.16,
      "grad_norm": 12.8374605178833,
      "learning_rate": 1.8404000000000002e-05,
      "loss": 2.179,
      "step": 400
    },
    {
      "epoch": 0.17,
      "grad_norm": 12.140827178955078,
      "learning_rate": 1.8304000000000003e-05,
      "loss": 2.4125,
      "step": 425
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.9592924118042,
      "learning_rate": 1.8204e-05,
      "loss": 2.166,
      "step": 450
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.604344367980957,
      "learning_rate": 1.8104e-05,
      "loss": 2.2307,
      "step": 475
    },
    {
      "epoch": 0.2,
      "grad_norm": 13.164838790893555,
      "learning_rate": 1.8004000000000002e-05,
      "loss": 2.3573,
      "step": 500
    },
    {
      "epoch": 0.21,
      "grad_norm": 12.212058067321777,
      "learning_rate": 1.7904e-05,
      "loss": 2.2179,
      "step": 525
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.600943565368652,
      "learning_rate": 1.7804e-05,
      "loss": 2.2763,
      "step": 550
    },
    {
      "epoch": 0.23,
      "grad_norm": 12.947696685791016,
      "learning_rate": 1.7704000000000002e-05,
      "loss": 2.0638,
      "step": 575
    },
    {
      "epoch": 0.24,
      "grad_norm": 17.80840301513672,
      "learning_rate": 1.7604e-05,
      "loss": 2.2055,
      "step": 600
    },
    {
      "epoch": 0.25,
      "grad_norm": 12.0412015914917,
      "learning_rate": 1.7504e-05,
      "loss": 2.1173,
      "step": 625
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.081609725952148,
      "learning_rate": 1.7404e-05,
      "loss": 2.2154,
      "step": 650
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.901471138000488,
      "learning_rate": 1.7304000000000002e-05,
      "loss": 2.0941,
      "step": 675
    },
    {
      "epoch": 0.28,
      "grad_norm": 11.670356750488281,
      "learning_rate": 1.7204e-05,
      "loss": 2.1167,
      "step": 700
    },
    {
      "epoch": 0.29,
      "grad_norm": 14.8759126663208,
      "learning_rate": 1.7104e-05,
      "loss": 2.1551,
      "step": 725
    },
    {
      "epoch": 0.3,
      "grad_norm": 11.987970352172852,
      "learning_rate": 1.7004000000000002e-05,
      "loss": 2.2229,
      "step": 750
    },
    {
      "epoch": 0.31,
      "grad_norm": 13.350611686706543,
      "learning_rate": 1.6904e-05,
      "loss": 2.1713,
      "step": 775
    },
    {
      "epoch": 0.32,
      "grad_norm": 10.762746810913086,
      "learning_rate": 1.6804e-05,
      "loss": 2.0882,
      "step": 800
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.588376998901367,
      "learning_rate": 1.6704e-05,
      "loss": 2.0899,
      "step": 825
    },
    {
      "epoch": 0.34,
      "grad_norm": 12.84415054321289,
      "learning_rate": 1.6604000000000002e-05,
      "loss": 2.1902,
      "step": 850
    },
    {
      "epoch": 0.35,
      "grad_norm": 11.016281127929688,
      "learning_rate": 1.6504000000000003e-05,
      "loss": 1.961,
      "step": 875
    },
    {
      "epoch": 0.36,
      "grad_norm": 11.485596656799316,
      "learning_rate": 1.6404e-05,
      "loss": 2.0134,
      "step": 900
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.29987907409668,
      "learning_rate": 1.6304000000000002e-05,
      "loss": 2.2747,
      "step": 925
    },
    {
      "epoch": 0.38,
      "grad_norm": 12.311944961547852,
      "learning_rate": 1.6204000000000003e-05,
      "loss": 2.0809,
      "step": 950
    },
    {
      "epoch": 0.39,
      "grad_norm": 14.393691062927246,
      "learning_rate": 1.6104e-05,
      "loss": 2.0551,
      "step": 975
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.741480827331543,
      "learning_rate": 1.6004e-05,
      "loss": 1.8158,
      "step": 1000
    },
    {
      "epoch": 0.41,
      "grad_norm": 15.93017864227295,
      "learning_rate": 1.5904000000000002e-05,
      "loss": 2.1485,
      "step": 1025
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.9660062789917,
      "learning_rate": 1.5804000000000003e-05,
      "loss": 2.1187,
      "step": 1050
    },
    {
      "epoch": 0.43,
      "grad_norm": 12.61098575592041,
      "learning_rate": 1.5704e-05,
      "loss": 1.9836,
      "step": 1075
    },
    {
      "epoch": 0.44,
      "grad_norm": 11.482502937316895,
      "learning_rate": 1.5604000000000002e-05,
      "loss": 2.0814,
      "step": 1100
    },
    {
      "epoch": 0.45,
      "grad_norm": 11.588091850280762,
      "learning_rate": 1.5504000000000003e-05,
      "loss": 2.0365,
      "step": 1125
    },
    {
      "epoch": 0.46,
      "grad_norm": 12.21688175201416,
      "learning_rate": 1.5404e-05,
      "loss": 2.0287,
      "step": 1150
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.9583740234375,
      "learning_rate": 1.5304e-05,
      "loss": 1.9626,
      "step": 1175
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.103066444396973,
      "learning_rate": 1.5204e-05,
      "loss": 2.0527,
      "step": 1200
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.230585098266602,
      "learning_rate": 1.5104000000000001e-05,
      "loss": 1.9494,
      "step": 1225
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.652959823608398,
      "learning_rate": 1.5004e-05,
      "loss": 1.9135,
      "step": 1250
    },
    {
      "epoch": 0.51,
      "grad_norm": 11.546061515808105,
      "learning_rate": 1.4904e-05,
      "loss": 2.0405,
      "step": 1275
    },
    {
      "epoch": 0.52,
      "grad_norm": 12.885430335998535,
      "learning_rate": 1.4804000000000001e-05,
      "loss": 2.1517,
      "step": 1300
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.972395896911621,
      "learning_rate": 1.4704e-05,
      "loss": 2.1219,
      "step": 1325
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.801010131835938,
      "learning_rate": 1.4604000000000001e-05,
      "loss": 1.9599,
      "step": 1350
    },
    {
      "epoch": 0.55,
      "grad_norm": 10.134483337402344,
      "learning_rate": 1.4504e-05,
      "loss": 2.0494,
      "step": 1375
    },
    {
      "epoch": 0.56,
      "grad_norm": 13.115943908691406,
      "learning_rate": 1.4404e-05,
      "loss": 1.9955,
      "step": 1400
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.545869827270508,
      "learning_rate": 1.4304e-05,
      "loss": 2.0235,
      "step": 1425
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.695992469787598,
      "learning_rate": 1.4204000000000002e-05,
      "loss": 1.8495,
      "step": 1450
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.978080749511719,
      "learning_rate": 1.4104000000000003e-05,
      "loss": 1.8959,
      "step": 1475
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.295175552368164,
      "learning_rate": 1.4004000000000002e-05,
      "loss": 1.9118,
      "step": 1500
    },
    {
      "epoch": 0.61,
      "grad_norm": 13.270401954650879,
      "learning_rate": 1.3904000000000003e-05,
      "loss": 2.1106,
      "step": 1525
    },
    {
      "epoch": 0.62,
      "grad_norm": 13.775677680969238,
      "learning_rate": 1.3804000000000002e-05,
      "loss": 1.8277,
      "step": 1550
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.527593612670898,
      "learning_rate": 1.3704000000000001e-05,
      "loss": 1.7639,
      "step": 1575
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.629280090332031,
      "learning_rate": 1.3604000000000002e-05,
      "loss": 1.9064,
      "step": 1600
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.989888191223145,
      "learning_rate": 1.3504000000000001e-05,
      "loss": 1.9594,
      "step": 1625
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.411887168884277,
      "learning_rate": 1.3404e-05,
      "loss": 2.1262,
      "step": 1650
    },
    {
      "epoch": 0.67,
      "grad_norm": 12.676020622253418,
      "learning_rate": 1.3304000000000002e-05,
      "loss": 1.676,
      "step": 1675
    },
    {
      "epoch": 0.68,
      "grad_norm": 11.182994842529297,
      "learning_rate": 1.3204000000000001e-05,
      "loss": 1.9182,
      "step": 1700
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.866786003112793,
      "learning_rate": 1.3104000000000002e-05,
      "loss": 1.9195,
      "step": 1725
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.576918601989746,
      "learning_rate": 1.3004000000000001e-05,
      "loss": 1.9578,
      "step": 1750
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.5067138671875,
      "learning_rate": 1.2904e-05,
      "loss": 1.8355,
      "step": 1775
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.727161407470703,
      "learning_rate": 1.2804000000000001e-05,
      "loss": 1.9025,
      "step": 1800
    },
    {
      "epoch": 0.73,
      "grad_norm": 13.11409854888916,
      "learning_rate": 1.2704e-05,
      "loss": 1.7382,
      "step": 1825
    },
    {
      "epoch": 0.74,
      "grad_norm": 11.595176696777344,
      "learning_rate": 1.2604e-05,
      "loss": 1.9146,
      "step": 1850
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.093597412109375,
      "learning_rate": 1.2504000000000001e-05,
      "loss": 1.8007,
      "step": 1875
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.586053848266602,
      "learning_rate": 1.2404e-05,
      "loss": 1.845,
      "step": 1900
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.637762069702148,
      "learning_rate": 1.2304000000000001e-05,
      "loss": 2.0743,
      "step": 1925
    },
    {
      "epoch": 0.78,
      "grad_norm": 13.100274085998535,
      "learning_rate": 1.2204e-05,
      "loss": 2.0598,
      "step": 1950
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.020994186401367,
      "learning_rate": 1.2104e-05,
      "loss": 2.0117,
      "step": 1975
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.455347061157227,
      "learning_rate": 1.2004e-05,
      "loss": 1.6429,
      "step": 2000
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.668612480163574,
      "learning_rate": 1.1904e-05,
      "loss": 2.0177,
      "step": 2025
    },
    {
      "epoch": 0.82,
      "grad_norm": 13.308269500732422,
      "learning_rate": 1.1803999999999999e-05,
      "loss": 1.8674,
      "step": 2050
    },
    {
      "epoch": 0.83,
      "grad_norm": 12.84695053100586,
      "learning_rate": 1.1704000000000002e-05,
      "loss": 1.8027,
      "step": 2075
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.068670272827148,
      "learning_rate": 1.1604000000000003e-05,
      "loss": 1.7137,
      "step": 2100
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.16296672821045,
      "learning_rate": 1.1504000000000002e-05,
      "loss": 1.9627,
      "step": 2125
    },
    {
      "epoch": 0.86,
      "grad_norm": 15.622565269470215,
      "learning_rate": 1.1404000000000001e-05,
      "loss": 2.005,
      "step": 2150
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.689506530761719,
      "learning_rate": 1.1304000000000002e-05,
      "loss": 2.0241,
      "step": 2175
    },
    {
      "epoch": 0.88,
      "grad_norm": 18.542600631713867,
      "learning_rate": 1.1204000000000001e-05,
      "loss": 1.8033,
      "step": 2200
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.676534652709961,
      "learning_rate": 1.1104e-05,
      "loss": 1.8776,
      "step": 2225
    },
    {
      "epoch": 0.9,
      "grad_norm": 15.894271850585938,
      "learning_rate": 1.1004000000000002e-05,
      "loss": 1.5818,
      "step": 2250
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.65552043914795,
      "learning_rate": 1.0904000000000001e-05,
      "loss": 1.9338,
      "step": 2275
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.770137786865234,
      "learning_rate": 1.0804000000000002e-05,
      "loss": 1.7558,
      "step": 2300
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.38111400604248,
      "learning_rate": 1.0704000000000001e-05,
      "loss": 1.7643,
      "step": 2325
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.971551895141602,
      "learning_rate": 1.0604e-05,
      "loss": 1.6931,
      "step": 2350
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.95477294921875,
      "learning_rate": 1.0504000000000001e-05,
      "loss": 1.8131,
      "step": 2375
    },
    {
      "epoch": 0.96,
      "grad_norm": 11.37536334991455,
      "learning_rate": 1.0404e-05,
      "loss": 1.7392,
      "step": 2400
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.042652130126953,
      "learning_rate": 1.0304e-05,
      "loss": 1.7464,
      "step": 2425
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.678910255432129,
      "learning_rate": 1.0204000000000001e-05,
      "loss": 1.6348,
      "step": 2450
    },
    {
      "epoch": 0.99,
      "grad_norm": 12.561325073242188,
      "learning_rate": 1.0104e-05,
      "loss": 1.5169,
      "step": 2475
    },
    {
      "epoch": 1.0,
      "grad_norm": 14.051156044006348,
      "learning_rate": 1.0004000000000001e-05,
      "loss": 1.7342,
      "step": 2500
    },
    {
      "epoch": 1.01,
      "grad_norm": 12.802934646606445,
      "learning_rate": 9.904e-06,
      "loss": 1.1948,
      "step": 2525
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.442876815795898,
      "learning_rate": 9.804000000000001e-06,
      "loss": 1.2588,
      "step": 2550
    },
    {
      "epoch": 1.03,
      "grad_norm": 11.199936866760254,
      "learning_rate": 9.704e-06,
      "loss": 1.1248,
      "step": 2575
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.727993965148926,
      "learning_rate": 9.604000000000002e-06,
      "loss": 1.1481,
      "step": 2600
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.783479690551758,
      "learning_rate": 9.504e-06,
      "loss": 1.0526,
      "step": 2625
    },
    {
      "epoch": 1.06,
      "grad_norm": 12.407696723937988,
      "learning_rate": 9.404e-06,
      "loss": 1.1352,
      "step": 2650
    },
    {
      "epoch": 1.07,
      "grad_norm": 14.621116638183594,
      "learning_rate": 9.304000000000001e-06,
      "loss": 1.1035,
      "step": 2675
    },
    {
      "epoch": 1.08,
      "grad_norm": 13.569993019104004,
      "learning_rate": 9.204e-06,
      "loss": 1.2436,
      "step": 2700
    },
    {
      "epoch": 1.09,
      "grad_norm": 13.364871978759766,
      "learning_rate": 9.104000000000001e-06,
      "loss": 1.186,
      "step": 2725
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.892740249633789,
      "learning_rate": 9.004e-06,
      "loss": 1.1549,
      "step": 2750
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.509928703308105,
      "learning_rate": 8.904e-06,
      "loss": 1.1382,
      "step": 2775
    },
    {
      "epoch": 1.12,
      "grad_norm": 11.294106483459473,
      "learning_rate": 8.804e-06,
      "loss": 1.1838,
      "step": 2800
    },
    {
      "epoch": 1.13,
      "grad_norm": 17.69602394104004,
      "learning_rate": 8.704e-06,
      "loss": 1.0854,
      "step": 2825
    },
    {
      "epoch": 1.1400000000000001,
      "grad_norm": 11.137506484985352,
      "learning_rate": 8.604000000000001e-06,
      "loss": 1.1801,
      "step": 2850
    },
    {
      "epoch": 1.15,
      "grad_norm": 8.684813499450684,
      "learning_rate": 8.504000000000002e-06,
      "loss": 1.1309,
      "step": 2875
    },
    {
      "epoch": 1.16,
      "grad_norm": 18.216867446899414,
      "learning_rate": 8.404000000000001e-06,
      "loss": 1.1336,
      "step": 2900
    },
    {
      "epoch": 1.17,
      "grad_norm": 13.37553882598877,
      "learning_rate": 8.304e-06,
      "loss": 1.0348,
      "step": 2925
    },
    {
      "epoch": 1.18,
      "grad_norm": 12.126663208007812,
      "learning_rate": 8.204000000000001e-06,
      "loss": 1.0356,
      "step": 2950
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.19808292388916,
      "learning_rate": 8.104e-06,
      "loss": 1.1679,
      "step": 2975
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.421396255493164,
      "learning_rate": 8.004e-06,
      "loss": 1.1297,
      "step": 3000
    },
    {
      "epoch": 1.21,
      "grad_norm": 13.185827255249023,
      "learning_rate": 7.904000000000001e-06,
      "loss": 1.2146,
      "step": 3025
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.685110092163086,
      "learning_rate": 7.804e-06,
      "loss": 1.2583,
      "step": 3050
    },
    {
      "epoch": 1.23,
      "grad_norm": 12.375425338745117,
      "learning_rate": 7.704000000000001e-06,
      "loss": 1.1996,
      "step": 3075
    },
    {
      "epoch": 1.24,
      "grad_norm": 14.291858673095703,
      "learning_rate": 7.604e-06,
      "loss": 1.2209,
      "step": 3100
    },
    {
      "epoch": 1.25,
      "grad_norm": 15.820944786071777,
      "learning_rate": 7.5040000000000005e-06,
      "loss": 1.3115,
      "step": 3125
    },
    {
      "epoch": 1.26,
      "grad_norm": 11.021650314331055,
      "learning_rate": 7.404e-06,
      "loss": 1.025,
      "step": 3150
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.467336654663086,
      "learning_rate": 7.304000000000001e-06,
      "loss": 1.1548,
      "step": 3175
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.907764434814453,
      "learning_rate": 7.204000000000001e-06,
      "loss": 1.2011,
      "step": 3200
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.38372802734375,
      "learning_rate": 7.104000000000001e-06,
      "loss": 1.1483,
      "step": 3225
    },
    {
      "epoch": 1.3,
      "grad_norm": 14.970967292785645,
      "learning_rate": 7.004000000000001e-06,
      "loss": 1.1487,
      "step": 3250
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.194636344909668,
      "learning_rate": 6.904e-06,
      "loss": 1.1764,
      "step": 3275
    },
    {
      "epoch": 1.32,
      "grad_norm": 16.364320755004883,
      "learning_rate": 6.804e-06,
      "loss": 1.2492,
      "step": 3300
    },
    {
      "epoch": 1.33,
      "grad_norm": 15.269165992736816,
      "learning_rate": 6.7040000000000005e-06,
      "loss": 1.0329,
      "step": 3325
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.500872611999512,
      "learning_rate": 6.604000000000001e-06,
      "loss": 1.092,
      "step": 3350
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.786423683166504,
      "learning_rate": 6.504e-06,
      "loss": 1.2249,
      "step": 3375
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 13.731120109558105,
      "learning_rate": 6.404e-06,
      "loss": 1.1118,
      "step": 3400
    },
    {
      "epoch": 1.37,
      "grad_norm": 16.703580856323242,
      "learning_rate": 6.304e-06,
      "loss": 1.0879,
      "step": 3425
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.39274787902832,
      "learning_rate": 6.204e-06,
      "loss": 1.0874,
      "step": 3450
    },
    {
      "epoch": 1.3900000000000001,
      "grad_norm": 14.778815269470215,
      "learning_rate": 6.104000000000001e-06,
      "loss": 1.0982,
      "step": 3475
    },
    {
      "epoch": 1.4,
      "grad_norm": 15.760618209838867,
      "learning_rate": 6.004000000000001e-06,
      "loss": 0.9911,
      "step": 3500
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.676987648010254,
      "learning_rate": 5.9040000000000006e-06,
      "loss": 0.9779,
      "step": 3525
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.369353294372559,
      "learning_rate": 5.804000000000001e-06,
      "loss": 0.9832,
      "step": 3550
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.173103332519531,
      "learning_rate": 5.704000000000001e-06,
      "loss": 1.1413,
      "step": 3575
    },
    {
      "epoch": 1.44,
      "grad_norm": 14.497014999389648,
      "learning_rate": 5.604000000000001e-06,
      "loss": 1.0769,
      "step": 3600
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.242537498474121,
      "learning_rate": 5.504e-06,
      "loss": 1.077,
      "step": 3625
    },
    {
      "epoch": 1.46,
      "grad_norm": 14.58755874633789,
      "learning_rate": 5.404e-06,
      "loss": 1.0362,
      "step": 3650
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.268484115600586,
      "learning_rate": 5.304e-06,
      "loss": 1.1701,
      "step": 3675
    },
    {
      "epoch": 1.48,
      "grad_norm": 12.811514854431152,
      "learning_rate": 5.2040000000000005e-06,
      "loss": 1.1419,
      "step": 3700
    },
    {
      "epoch": 1.49,
      "grad_norm": 13.174732208251953,
      "learning_rate": 5.104e-06,
      "loss": 0.9834,
      "step": 3725
    },
    {
      "epoch": 1.5,
      "grad_norm": 14.920330047607422,
      "learning_rate": 5.004e-06,
      "loss": 1.1624,
      "step": 3750
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.778399467468262,
      "learning_rate": 4.904000000000001e-06,
      "loss": 1.0573,
      "step": 3775
    },
    {
      "epoch": 1.52,
      "grad_norm": 13.889476776123047,
      "learning_rate": 4.804e-06,
      "loss": 1.1256,
      "step": 3800
    },
    {
      "epoch": 1.53,
      "grad_norm": 14.520740509033203,
      "learning_rate": 4.704e-06,
      "loss": 1.0834,
      "step": 3825
    },
    {
      "epoch": 1.54,
      "grad_norm": 12.798078536987305,
      "learning_rate": 4.604e-06,
      "loss": 1.0705,
      "step": 3850
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.737931251525879,
      "learning_rate": 4.504e-06,
      "loss": 1.1076,
      "step": 3875
    },
    {
      "epoch": 1.56,
      "grad_norm": 12.435004234313965,
      "learning_rate": 4.4040000000000005e-06,
      "loss": 1.0675,
      "step": 3900
    },
    {
      "epoch": 1.5699999999999998,
      "grad_norm": 11.721802711486816,
      "learning_rate": 4.304000000000001e-06,
      "loss": 1.1715,
      "step": 3925
    },
    {
      "epoch": 1.58,
      "grad_norm": 16.310001373291016,
      "learning_rate": 4.204e-06,
      "loss": 1.0577,
      "step": 3950
    },
    {
      "epoch": 1.5899999999999999,
      "grad_norm": 9.448426246643066,
      "learning_rate": 4.104e-06,
      "loss": 0.9881,
      "step": 3975
    },
    {
      "epoch": 1.6,
      "grad_norm": 15.235527992248535,
      "learning_rate": 4.004e-06,
      "loss": 0.9509,
      "step": 4000
    },
    {
      "epoch": 1.6099999999999999,
      "grad_norm": 11.758926391601562,
      "learning_rate": 3.904e-06,
      "loss": 1.017,
      "step": 4025
    },
    {
      "epoch": 1.62,
      "grad_norm": 12.346951484680176,
      "learning_rate": 3.8040000000000003e-06,
      "loss": 1.1987,
      "step": 4050
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.503011703491211,
      "learning_rate": 3.7040000000000005e-06,
      "loss": 1.0665,
      "step": 4075
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 9.963557243347168,
      "learning_rate": 3.604e-06,
      "loss": 1.1667,
      "step": 4100
    },
    {
      "epoch": 1.65,
      "grad_norm": 15.800553321838379,
      "learning_rate": 3.5040000000000002e-06,
      "loss": 1.0871,
      "step": 4125
    },
    {
      "epoch": 1.6600000000000001,
      "grad_norm": 14.133501052856445,
      "learning_rate": 3.404e-06,
      "loss": 1.0588,
      "step": 4150
    },
    {
      "epoch": 1.67,
      "grad_norm": 15.049243927001953,
      "learning_rate": 3.3040000000000005e-06,
      "loss": 1.1755,
      "step": 4175
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 11.534449577331543,
      "learning_rate": 3.2040000000000006e-06,
      "loss": 1.0809,
      "step": 4200
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.926283836364746,
      "learning_rate": 3.1040000000000003e-06,
      "loss": 1.1184,
      "step": 4225
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.016489028930664,
      "learning_rate": 3.0040000000000004e-06,
      "loss": 0.9284,
      "step": 4250
    },
    {
      "epoch": 1.71,
      "grad_norm": 16.352996826171875,
      "learning_rate": 2.904e-06,
      "loss": 0.8954,
      "step": 4275
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.342592239379883,
      "learning_rate": 2.804e-06,
      "loss": 1.0614,
      "step": 4300
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.883009910583496,
      "learning_rate": 2.704e-06,
      "loss": 1.006,
      "step": 4325
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.741077423095703,
      "learning_rate": 2.6040000000000004e-06,
      "loss": 1.1116,
      "step": 4350
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.227665901184082,
      "learning_rate": 2.5040000000000005e-06,
      "loss": 1.0802,
      "step": 4375
    },
    {
      "epoch": 1.76,
      "grad_norm": 12.50661849975586,
      "learning_rate": 2.404e-06,
      "loss": 0.9816,
      "step": 4400
    },
    {
      "epoch": 1.77,
      "grad_norm": 13.356459617614746,
      "learning_rate": 2.3040000000000003e-06,
      "loss": 1.0844,
      "step": 4425
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.1768159866333,
      "learning_rate": 2.2040000000000004e-06,
      "loss": 0.9909,
      "step": 4450
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.657759666442871,
      "learning_rate": 2.104e-06,
      "loss": 0.9877,
      "step": 4475
    },
    {
      "epoch": 1.8,
      "grad_norm": 14.901777267456055,
      "learning_rate": 2.004e-06,
      "loss": 0.9987,
      "step": 4500
    },
    {
      "epoch": 1.81,
      "grad_norm": 17.132478713989258,
      "learning_rate": 1.9040000000000003e-06,
      "loss": 1.0573,
      "step": 4525
    },
    {
      "epoch": 1.8199999999999998,
      "grad_norm": 13.090239524841309,
      "learning_rate": 1.8040000000000002e-06,
      "loss": 0.9955,
      "step": 4550
    },
    {
      "epoch": 1.83,
      "grad_norm": 7.524806022644043,
      "learning_rate": 1.7040000000000001e-06,
      "loss": 1.0188,
      "step": 4575
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 14.137269973754883,
      "learning_rate": 1.604e-06,
      "loss": 0.9507,
      "step": 4600
    },
    {
      "epoch": 1.85,
      "grad_norm": 15.662611961364746,
      "learning_rate": 1.5040000000000001e-06,
      "loss": 1.0512,
      "step": 4625
    },
    {
      "epoch": 1.8599999999999999,
      "grad_norm": 13.104068756103516,
      "learning_rate": 1.404e-06,
      "loss": 1.0257,
      "step": 4650
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.180632591247559,
      "learning_rate": 1.304e-06,
      "loss": 1.1014,
      "step": 4675
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.715453147888184,
      "learning_rate": 1.204e-06,
      "loss": 0.9764,
      "step": 4700
    },
    {
      "epoch": 1.8900000000000001,
      "grad_norm": 10.140015602111816,
      "learning_rate": 1.1040000000000001e-06,
      "loss": 1.1436,
      "step": 4725
    },
    {
      "epoch": 1.9,
      "grad_norm": 13.225520133972168,
      "learning_rate": 1.004e-06,
      "loss": 1.021,
      "step": 4750
    },
    {
      "epoch": 1.9100000000000001,
      "grad_norm": 11.760268211364746,
      "learning_rate": 9.04e-07,
      "loss": 1.0674,
      "step": 4775
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.296557426452637,
      "learning_rate": 8.04e-07,
      "loss": 1.044,
      "step": 4800
    },
    {
      "epoch": 1.9300000000000002,
      "grad_norm": 11.622244834899902,
      "learning_rate": 7.040000000000001e-07,
      "loss": 1.0317,
      "step": 4825
    },
    {
      "epoch": 1.94,
      "grad_norm": 14.334859848022461,
      "learning_rate": 6.040000000000001e-07,
      "loss": 1.1214,
      "step": 4850
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.96068286895752,
      "learning_rate": 5.040000000000001e-07,
      "loss": 1.1726,
      "step": 4875
    },
    {
      "epoch": 1.96,
      "grad_norm": 14.571761131286621,
      "learning_rate": 4.04e-07,
      "loss": 0.9106,
      "step": 4900
    },
    {
      "epoch": 1.97,
      "grad_norm": 11.797457695007324,
      "learning_rate": 3.04e-07,
      "loss": 1.1328,
      "step": 4925
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.585701942443848,
      "learning_rate": 2.0400000000000003e-07,
      "loss": 1.0677,
      "step": 4950
    },
    {
      "epoch": 1.99,
      "grad_norm": 15.56652545928955,
      "learning_rate": 1.04e-07,
      "loss": 1.0312,
      "step": 4975
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.480283737182617,
      "learning_rate": 4e-09,
      "loss": 0.8741,
      "step": 5000
    }
  ],
  "logging_steps": 25,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.1142437888e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}