{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1700,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.000588235294117647,
      "grad_norm": 86.25792069448507,
      "learning_rate": 5.882352941176471e-08,
      "loss": 4.144,
      "step": 1
    },
    {
      "epoch": 0.0029411764705882353,
      "grad_norm": 92.94334261133154,
      "learning_rate": 2.9411764705882356e-07,
      "loss": 4.2053,
      "step": 5
    },
    {
      "epoch": 0.0058823529411764705,
      "grad_norm": 34.41771480337444,
      "learning_rate": 5.882352941176471e-07,
      "loss": 3.9703,
      "step": 10
    },
    {
      "epoch": 0.008823529411764706,
      "grad_norm": 20.69702610638224,
      "learning_rate": 8.823529411764707e-07,
      "loss": 3.4529,
      "step": 15
    },
    {
      "epoch": 0.011764705882352941,
      "grad_norm": 12.865609762875932,
      "learning_rate": 1.1764705882352942e-06,
      "loss": 3.1192,
      "step": 20
    },
    {
      "epoch": 0.014705882352941176,
      "grad_norm": 14.658994779954883,
      "learning_rate": 1.4705882352941177e-06,
      "loss": 2.9015,
      "step": 25
    },
    {
      "epoch": 0.01764705882352941,
      "grad_norm": 19.64009310214488,
      "learning_rate": 1.7647058823529414e-06,
      "loss": 2.5883,
      "step": 30
    },
    {
      "epoch": 0.020588235294117647,
      "grad_norm": 27.20931711935924,
      "learning_rate": 2.058823529411765e-06,
      "loss": 2.2725,
      "step": 35
    },
    {
      "epoch": 0.023529411764705882,
      "grad_norm": 23.446531565331924,
      "learning_rate": 2.3529411764705885e-06,
      "loss": 1.8073,
      "step": 40
    },
    {
      "epoch": 0.026470588235294117,
      "grad_norm": 13.31815173013865,
      "learning_rate": 2.647058823529412e-06,
      "loss": 1.3518,
      "step": 45
    },
    {
      "epoch": 0.029411764705882353,
      "grad_norm": 5.34093550595349,
      "learning_rate": 2.9411764705882355e-06,
      "loss": 1.1676,
      "step": 50
    },
    {
      "epoch": 0.03235294117647059,
      "grad_norm": 3.3456911840455628,
      "learning_rate": 3.2352941176470594e-06,
      "loss": 1.147,
      "step": 55
    },
    {
      "epoch": 0.03529411764705882,
      "grad_norm": 5.032424322254707,
      "learning_rate": 3.529411764705883e-06,
      "loss": 1.0965,
      "step": 60
    },
    {
      "epoch": 0.03823529411764706,
      "grad_norm": 4.206085507780853,
      "learning_rate": 3.8235294117647055e-06,
      "loss": 1.0636,
      "step": 65
    },
    {
      "epoch": 0.041176470588235294,
      "grad_norm": 8.089055829866213,
      "learning_rate": 4.11764705882353e-06,
      "loss": 1.0582,
      "step": 70
    },
    {
      "epoch": 0.04411764705882353,
      "grad_norm": 3.347364025882152,
      "learning_rate": 4.411764705882353e-06,
      "loss": 1.0539,
      "step": 75
    },
    {
      "epoch": 0.047058823529411764,
      "grad_norm": 6.336557333099317,
      "learning_rate": 4.705882352941177e-06,
      "loss": 1.0379,
      "step": 80
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.978187369479212,
      "learning_rate": 5e-06,
      "loss": 1.0242,
      "step": 85
    },
    {
      "epoch": 0.052941176470588235,
      "grad_norm": 7.818112219971164,
      "learning_rate": 5.294117647058824e-06,
      "loss": 1.0016,
      "step": 90
    },
    {
      "epoch": 0.05588235294117647,
      "grad_norm": 8.433408497312906,
      "learning_rate": 5.588235294117647e-06,
      "loss": 0.9771,
      "step": 95
    },
    {
      "epoch": 0.058823529411764705,
      "grad_norm": 6.212980683823607,
      "learning_rate": 5.882352941176471e-06,
      "loss": 0.9861,
      "step": 100
    },
    {
      "epoch": 0.061764705882352944,
      "grad_norm": 6.647220375875289,
      "learning_rate": 6.176470588235295e-06,
      "loss": 0.9602,
      "step": 105
    },
    {
      "epoch": 0.06470588235294118,
      "grad_norm": 6.527727138992631,
      "learning_rate": 6.470588235294119e-06,
      "loss": 0.9525,
      "step": 110
    },
    {
      "epoch": 0.06764705882352941,
      "grad_norm": 7.509553679384927,
      "learning_rate": 6.764705882352942e-06,
      "loss": 0.9416,
      "step": 115
    },
    {
      "epoch": 0.07058823529411765,
      "grad_norm": 8.103718249384503,
      "learning_rate": 7.058823529411766e-06,
      "loss": 0.9267,
      "step": 120
    },
    {
      "epoch": 0.07352941176470588,
      "grad_norm": 8.520097429021,
      "learning_rate": 7.352941176470589e-06,
      "loss": 0.9396,
      "step": 125
    },
    {
      "epoch": 0.07647058823529412,
      "grad_norm": 9.982219276844697,
      "learning_rate": 7.647058823529411e-06,
      "loss": 0.9259,
      "step": 130
    },
    {
      "epoch": 0.07941176470588235,
      "grad_norm": 10.609383365594049,
      "learning_rate": 7.941176470588236e-06,
      "loss": 0.9262,
      "step": 135
    },
    {
      "epoch": 0.08235294117647059,
      "grad_norm": 6.496638967078595,
      "learning_rate": 8.23529411764706e-06,
      "loss": 0.9169,
      "step": 140
    },
    {
      "epoch": 0.08529411764705883,
      "grad_norm": 5.899471704114852,
      "learning_rate": 8.529411764705883e-06,
      "loss": 0.9172,
      "step": 145
    },
    {
      "epoch": 0.08823529411764706,
      "grad_norm": 19.813714293488992,
      "learning_rate": 8.823529411764707e-06,
      "loss": 0.8899,
      "step": 150
    },
    {
      "epoch": 0.09117647058823529,
      "grad_norm": 3.91249318240706,
      "learning_rate": 9.11764705882353e-06,
      "loss": 0.9074,
      "step": 155
    },
    {
      "epoch": 0.09411764705882353,
      "grad_norm": 10.124669032565054,
      "learning_rate": 9.411764705882354e-06,
      "loss": 0.898,
      "step": 160
    },
    {
      "epoch": 0.09705882352941177,
      "grad_norm": 13.840426312968154,
      "learning_rate": 9.705882352941177e-06,
      "loss": 0.9049,
      "step": 165
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.60421096496902,
      "learning_rate": 1e-05,
      "loss": 0.9154,
      "step": 170
    },
    {
      "epoch": 0.10294117647058823,
      "grad_norm": 6.134422545965741,
      "learning_rate": 9.999736492435867e-06,
      "loss": 0.8836,
      "step": 175
    },
    {
      "epoch": 0.10588235294117647,
      "grad_norm": 11.600519341121485,
      "learning_rate": 9.998945997517957e-06,
      "loss": 0.8904,
      "step": 180
    },
    {
      "epoch": 0.10882352941176471,
      "grad_norm": 11.799379752373861,
      "learning_rate": 9.99762859856683e-06,
      "loss": 0.8941,
      "step": 185
    },
    {
      "epoch": 0.11176470588235295,
      "grad_norm": 6.623810329677796,
      "learning_rate": 9.99578443444032e-06,
      "loss": 0.8729,
      "step": 190
    },
    {
      "epoch": 0.11470588235294117,
      "grad_norm": 6.965926377173582,
      "learning_rate": 9.993413699518906e-06,
      "loss": 0.8733,
      "step": 195
    },
    {
      "epoch": 0.11764705882352941,
      "grad_norm": 4.929345626084727,
      "learning_rate": 9.990516643685222e-06,
      "loss": 0.8559,
      "step": 200
    },
    {
      "epoch": 0.12058823529411765,
      "grad_norm": 4.438658396751935,
      "learning_rate": 9.987093572297716e-06,
      "loss": 0.8693,
      "step": 205
    },
    {
      "epoch": 0.12352941176470589,
      "grad_norm": 9.676542191697266,
      "learning_rate": 9.983144846158472e-06,
      "loss": 0.8788,
      "step": 210
    },
    {
      "epoch": 0.1264705882352941,
      "grad_norm": 10.638416959011728,
      "learning_rate": 9.978670881475173e-06,
      "loss": 0.8517,
      "step": 215
    },
    {
      "epoch": 0.12941176470588237,
      "grad_norm": 15.793160052452487,
      "learning_rate": 9.973672149817232e-06,
      "loss": 0.8362,
      "step": 220
    },
    {
      "epoch": 0.1323529411764706,
      "grad_norm": 14.96288140935782,
      "learning_rate": 9.96814917806609e-06,
      "loss": 0.8517,
      "step": 225
    },
    {
      "epoch": 0.13529411764705881,
      "grad_norm": 17.94048430648359,
      "learning_rate": 9.96210254835968e-06,
      "loss": 0.8452,
      "step": 230
    },
    {
      "epoch": 0.13823529411764707,
      "grad_norm": 10.072819918084187,
      "learning_rate": 9.955532898031069e-06,
      "loss": 0.8391,
      "step": 235
    },
    {
      "epoch": 0.1411764705882353,
      "grad_norm": 7.684632649087268,
      "learning_rate": 9.948440919541277e-06,
      "loss": 0.8289,
      "step": 240
    },
    {
      "epoch": 0.14411764705882352,
      "grad_norm": 17.954771803869264,
      "learning_rate": 9.940827360406297e-06,
      "loss": 0.8067,
      "step": 245
    },
    {
      "epoch": 0.14705882352941177,
      "grad_norm": 12.023797137924742,
      "learning_rate": 9.932693023118299e-06,
      "loss": 0.8175,
      "step": 250
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.948930140888852,
      "learning_rate": 9.924038765061042e-06,
      "loss": 0.8226,
      "step": 255
    },
    {
      "epoch": 0.15294117647058825,
      "grad_norm": 17.86031699503285,
      "learning_rate": 9.91486549841951e-06,
      "loss": 0.8168,
      "step": 260
    },
    {
      "epoch": 0.15588235294117647,
      "grad_norm": 10.990033307622202,
      "learning_rate": 9.905174190083763e-06,
      "loss": 0.8047,
      "step": 265
    },
    {
      "epoch": 0.1588235294117647,
      "grad_norm": 8.715291883213558,
      "learning_rate": 9.894965861547023e-06,
      "loss": 0.8007,
      "step": 270
    },
    {
      "epoch": 0.16176470588235295,
      "grad_norm": 13.856071733371282,
      "learning_rate": 9.884241588798004e-06,
      "loss": 0.8017,
      "step": 275
    },
    {
      "epoch": 0.16470588235294117,
      "grad_norm": 15.297710097108334,
      "learning_rate": 9.873002502207502e-06,
      "loss": 0.8055,
      "step": 280
    },
    {
      "epoch": 0.1676470588235294,
      "grad_norm": 10.725176474324847,
      "learning_rate": 9.861249786409248e-06,
      "loss": 0.7972,
      "step": 285
    },
    {
      "epoch": 0.17058823529411765,
      "grad_norm": 12.860465116141345,
      "learning_rate": 9.848984680175049e-06,
      "loss": 0.8145,
      "step": 290
    },
    {
      "epoch": 0.17352941176470588,
      "grad_norm": 13.411817152907572,
      "learning_rate": 9.836208476284208e-06,
      "loss": 0.7992,
      "step": 295
    },
    {
      "epoch": 0.17647058823529413,
      "grad_norm": 10.317196889908503,
      "learning_rate": 9.822922521387277e-06,
      "loss": 0.7846,
      "step": 300
    },
    {
      "epoch": 0.17941176470588235,
      "grad_norm": 4.179704998479472,
      "learning_rate": 9.809128215864096e-06,
      "loss": 0.8025,
      "step": 305
    },
    {
      "epoch": 0.18235294117647058,
      "grad_norm": 6.430799430688913,
      "learning_rate": 9.794827013676206e-06,
      "loss": 0.7991,
      "step": 310
    },
    {
      "epoch": 0.18529411764705883,
      "grad_norm": 5.850743250582622,
      "learning_rate": 9.78002042221359e-06,
      "loss": 0.8157,
      "step": 315
    },
    {
      "epoch": 0.18823529411764706,
      "grad_norm": 7.981852850027179,
      "learning_rate": 9.764710002135784e-06,
      "loss": 0.8315,
      "step": 320
    },
    {
      "epoch": 0.19117647058823528,
      "grad_norm": 4.246718673048229,
      "learning_rate": 9.748897367207391e-06,
      "loss": 0.8126,
      "step": 325
    },
    {
      "epoch": 0.19411764705882353,
      "grad_norm": 3.197412855424511,
      "learning_rate": 9.732584184127973e-06,
      "loss": 0.8054,
      "step": 330
    },
    {
      "epoch": 0.19705882352941176,
      "grad_norm": 3.866629151755055,
      "learning_rate": 9.715772172356388e-06,
      "loss": 0.8106,
      "step": 335
    },
    {
      "epoch": 0.2,
      "grad_norm": 3.4312289366659603,
      "learning_rate": 9.698463103929542e-06,
      "loss": 0.803,
      "step": 340
    },
    {
      "epoch": 0.20294117647058824,
      "grad_norm": 4.537035689074838,
      "learning_rate": 9.68065880327562e-06,
      "loss": 0.8197,
      "step": 345
    },
    {
      "epoch": 0.20588235294117646,
      "grad_norm": 4.167020748172961,
      "learning_rate": 9.66236114702178e-06,
      "loss": 0.8128,
      "step": 350
    },
    {
      "epoch": 0.2088235294117647,
      "grad_norm": 7.084719945270455,
      "learning_rate": 9.643572063796352e-06,
      "loss": 0.7983,
      "step": 355
    },
    {
      "epoch": 0.21176470588235294,
      "grad_norm": 4.345120253918932,
      "learning_rate": 9.62429353402556e-06,
      "loss": 0.7932,
      "step": 360
    },
    {
      "epoch": 0.21470588235294116,
      "grad_norm": 3.8354401768943536,
      "learning_rate": 9.60452758972477e-06,
      "loss": 0.7911,
      "step": 365
    },
    {
      "epoch": 0.21764705882352942,
      "grad_norm": 3.895028761361664,
      "learning_rate": 9.584276314284316e-06,
      "loss": 0.7903,
      "step": 370
    },
    {
      "epoch": 0.22058823529411764,
      "grad_norm": 4.768654905615113,
      "learning_rate": 9.563541842249903e-06,
      "loss": 0.814,
      "step": 375
    },
    {
      "epoch": 0.2235294117647059,
      "grad_norm": 4.001695620575136,
      "learning_rate": 9.542326359097619e-06,
      "loss": 0.7983,
      "step": 380
    },
    {
      "epoch": 0.22647058823529412,
      "grad_norm": 5.464979383463447,
      "learning_rate": 9.520632101003579e-06,
      "loss": 0.8113,
      "step": 385
    },
    {
      "epoch": 0.22941176470588234,
      "grad_norm": 4.410052168842928,
      "learning_rate": 9.498461354608228e-06,
      "loss": 0.7871,
      "step": 390
    },
    {
      "epoch": 0.2323529411764706,
      "grad_norm": 10.214130104526346,
      "learning_rate": 9.475816456775313e-06,
      "loss": 0.7673,
      "step": 395
    },
    {
      "epoch": 0.23529411764705882,
      "grad_norm": 7.775696918112043,
      "learning_rate": 9.452699794345583e-06,
      "loss": 0.8504,
      "step": 400
    },
    {
      "epoch": 0.23823529411764705,
      "grad_norm": 4.793572966156681,
      "learning_rate": 9.429113803885199e-06,
      "loss": 0.7983,
      "step": 405
    },
    {
      "epoch": 0.2411764705882353,
      "grad_norm": 6.2733719119866755,
      "learning_rate": 9.405060971428924e-06,
      "loss": 0.7545,
      "step": 410
    },
    {
      "epoch": 0.24411764705882352,
      "grad_norm": 5.665044173485091,
      "learning_rate": 9.380543832218069e-06,
      "loss": 0.7694,
      "step": 415
    },
    {
      "epoch": 0.24705882352941178,
      "grad_norm": 6.667717620194777,
      "learning_rate": 9.355564970433288e-06,
      "loss": 0.7749,
      "step": 420
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.3946898220456827,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.7857,
      "step": 425
    },
    {
      "epoch": 0.2529411764705882,
      "grad_norm": 7.472860005366773,
      "learning_rate": 9.30423265892184e-06,
      "loss": 0.7711,
      "step": 430
    },
    {
      "epoch": 0.25588235294117645,
      "grad_norm": 4.70932174586482,
      "learning_rate": 9.277884619776116e-06,
      "loss": 0.7765,
      "step": 435
    },
    {
      "epoch": 0.25882352941176473,
      "grad_norm": 9.57060451964989,
      "learning_rate": 9.251085678648072e-06,
      "loss": 0.7604,
      "step": 440
    },
    {
      "epoch": 0.26176470588235295,
      "grad_norm": 4.681414682219643,
      "learning_rate": 9.223838660227183e-06,
      "loss": 0.7747,
      "step": 445
    },
    {
      "epoch": 0.2647058823529412,
      "grad_norm": 7.367138755832612,
      "learning_rate": 9.196146436431635e-06,
      "loss": 0.7634,
      "step": 450
    },
    {
      "epoch": 0.2676470588235294,
      "grad_norm": 5.627007459776944,
      "learning_rate": 9.168011926105598e-06,
      "loss": 0.7641,
      "step": 455
    },
    {
      "epoch": 0.27058823529411763,
      "grad_norm": 3.5740204640033078,
      "learning_rate": 9.13943809471159e-06,
      "loss": 0.7565,
      "step": 460
    },
    {
      "epoch": 0.2735294117647059,
      "grad_norm": 7.69019676925858,
      "learning_rate": 9.110427954017891e-06,
      "loss": 0.7733,
      "step": 465
    },
    {
      "epoch": 0.27647058823529413,
      "grad_norm": 3.5962732922894975,
      "learning_rate": 9.08098456178111e-06,
      "loss": 0.765,
      "step": 470
    },
    {
      "epoch": 0.27941176470588236,
      "grad_norm": 4.241563630767514,
      "learning_rate": 9.051111021423868e-06,
      "loss": 0.7754,
      "step": 475
    },
    {
      "epoch": 0.2823529411764706,
      "grad_norm": 7.8000912530998745,
      "learning_rate": 9.020810481707709e-06,
      "loss": 0.7585,
      "step": 480
    },
    {
      "epoch": 0.2852941176470588,
      "grad_norm": 3.186234693569658,
      "learning_rate": 8.990086136401199e-06,
      "loss": 0.7547,
      "step": 485
    },
    {
      "epoch": 0.28823529411764703,
      "grad_norm": 7.586644694603505,
      "learning_rate": 8.958941223943292e-06,
      "loss": 0.7427,
      "step": 490
    },
    {
      "epoch": 0.2911764705882353,
      "grad_norm": 3.258615795641642,
      "learning_rate": 8.927379027101994e-06,
      "loss": 0.7399,
      "step": 495
    },
    {
      "epoch": 0.29411764705882354,
      "grad_norm": 3.731634633993737,
      "learning_rate": 8.895402872628352e-06,
      "loss": 0.7391,
      "step": 500
    },
    {
      "epoch": 0.29705882352941176,
      "grad_norm": 4.803677561739918,
      "learning_rate": 8.863016130905795e-06,
      "loss": 0.7407,
      "step": 505
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.655699535414225,
      "learning_rate": 8.83022221559489e-06,
      "loss": 0.7506,
      "step": 510
    },
    {
      "epoch": 0.3029411764705882,
      "grad_norm": 4.373831528386331,
      "learning_rate": 8.797024583273536e-06,
      "loss": 0.7355,
      "step": 515
    },
    {
      "epoch": 0.3058823529411765,
      "grad_norm": 3.9385838728737266,
      "learning_rate": 8.763426733072624e-06,
      "loss": 0.735,
      "step": 520
    },
    {
      "epoch": 0.3088235294117647,
      "grad_norm": 4.929470086337327,
      "learning_rate": 8.729432206307218e-06,
      "loss": 0.7407,
      "step": 525
    },
    {
      "epoch": 0.31176470588235294,
      "grad_norm": 4.985252344557763,
      "learning_rate": 8.695044586103297e-06,
      "loss": 0.7368,
      "step": 530
    },
    {
      "epoch": 0.31470588235294117,
      "grad_norm": 4.402532806236826,
      "learning_rate": 8.660267497020074e-06,
      "loss": 0.7378,
      "step": 535
    },
    {
      "epoch": 0.3176470588235294,
      "grad_norm": 6.632318764297617,
      "learning_rate": 8.625104604667965e-06,
      "loss": 0.7182,
      "step": 540
    },
    {
      "epoch": 0.3205882352941177,
      "grad_norm": 3.0548060644071566,
      "learning_rate": 8.58955961532221e-06,
      "loss": 0.7198,
      "step": 545
    },
    {
      "epoch": 0.3235294117647059,
      "grad_norm": 7.1426568719242445,
      "learning_rate": 8.553636275532236e-06,
      "loss": 0.7336,
      "step": 550
    },
    {
      "epoch": 0.3264705882352941,
      "grad_norm": 10.88738661858406,
      "learning_rate": 8.51733837172675e-06,
      "loss": 0.7368,
      "step": 555
    },
    {
      "epoch": 0.32941176470588235,
      "grad_norm": 6.282693925795995,
      "learning_rate": 8.480669729814635e-06,
      "loss": 0.7276,
      "step": 560
    },
    {
      "epoch": 0.3323529411764706,
      "grad_norm": 9.555955539230844,
      "learning_rate": 8.443634214781693e-06,
      "loss": 0.702,
      "step": 565
    },
    {
      "epoch": 0.3352941176470588,
      "grad_norm": 4.506254396871987,
      "learning_rate": 8.40623573028327e-06,
      "loss": 0.7129,
      "step": 570
    },
    {
      "epoch": 0.3382352941176471,
      "grad_norm": 3.63704291073538,
      "learning_rate": 8.368478218232787e-06,
      "loss": 0.719,
      "step": 575
    },
    {
      "epoch": 0.3411764705882353,
      "grad_norm": 2.9414631231401462,
      "learning_rate": 8.330365658386252e-06,
      "loss": 0.7119,
      "step": 580
    },
    {
      "epoch": 0.34411764705882353,
      "grad_norm": 3.8962634035076675,
      "learning_rate": 8.291902067922791e-06,
      "loss": 0.7044,
      "step": 585
    },
    {
      "epoch": 0.34705882352941175,
      "grad_norm": 2.996619696992914,
      "learning_rate": 8.25309150102121e-06,
      "loss": 0.7221,
      "step": 590
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.6918987952391653,
      "learning_rate": 8.213938048432697e-06,
      "loss": 0.7152,
      "step": 595
    },
    {
      "epoch": 0.35294117647058826,
      "grad_norm": 5.1406824761237475,
      "learning_rate": 8.174445837049614e-06,
      "loss": 0.7122,
      "step": 600
    },
    {
      "epoch": 0.3558823529411765,
      "grad_norm": 3.334206362728506,
      "learning_rate": 8.134619029470535e-06,
      "loss": 0.7021,
      "step": 605
    },
    {
      "epoch": 0.3588235294117647,
      "grad_norm": 4.968392821499114,
      "learning_rate": 8.094461823561473e-06,
      "loss": 0.7053,
      "step": 610
    },
    {
      "epoch": 0.36176470588235293,
      "grad_norm": 3.1842109451175165,
      "learning_rate": 8.05397845201344e-06,
      "loss": 0.7142,
      "step": 615
    },
    {
      "epoch": 0.36470588235294116,
      "grad_norm": 7.122667502997686,
      "learning_rate": 8.013173181896283e-06,
      "loss": 0.7158,
      "step": 620
    },
    {
      "epoch": 0.36764705882352944,
      "grad_norm": 3.7490095954612563,
      "learning_rate": 7.972050314208934e-06,
      "loss": 0.7073,
      "step": 625
    },
    {
      "epoch": 0.37058823529411766,
      "grad_norm": 3.418137694246427,
      "learning_rate": 7.930614183426074e-06,
      "loss": 0.7098,
      "step": 630
    },
    {
      "epoch": 0.3735294117647059,
      "grad_norm": 3.8866651266730092,
      "learning_rate": 7.888869157041257e-06,
      "loss": 0.6827,
      "step": 635
    },
    {
      "epoch": 0.3764705882352941,
      "grad_norm": 3.5194202567335466,
      "learning_rate": 7.846819635106569e-06,
      "loss": 0.6891,
      "step": 640
    },
    {
      "epoch": 0.37941176470588234,
      "grad_norm": 3.847874692639519,
      "learning_rate": 7.80447004976885e-06,
      "loss": 0.667,
      "step": 645
    },
    {
      "epoch": 0.38235294117647056,
      "grad_norm": 5.871281951965833,
      "learning_rate": 7.76182486480253e-06,
      "loss": 0.6792,
      "step": 650
    },
    {
      "epoch": 0.38529411764705884,
      "grad_norm": 2.9461535013380344,
      "learning_rate": 7.718888575139134e-06,
      "loss": 0.6821,
      "step": 655
    },
    {
      "epoch": 0.38823529411764707,
      "grad_norm": 3.545429320117265,
      "learning_rate": 7.675665706393502e-06,
      "loss": 0.6837,
      "step": 660
    },
    {
      "epoch": 0.3911764705882353,
      "grad_norm": 3.9453651020162406,
      "learning_rate": 7.63216081438678e-06,
      "loss": 0.6944,
      "step": 665
    },
    {
      "epoch": 0.3941176470588235,
      "grad_norm": 4.157385411843414,
      "learning_rate": 7.588378484666214e-06,
      "loss": 0.6803,
      "step": 670
    },
    {
      "epoch": 0.39705882352941174,
      "grad_norm": 2.9879188800031287,
      "learning_rate": 7.544323332021826e-06,
      "loss": 0.6716,
      "step": 675
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.065231322987666,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.6899,
      "step": 680
    },
    {
      "epoch": 0.40294117647058825,
      "grad_norm": 5.937109080515486,
      "learning_rate": 7.4554131604140425e-06,
      "loss": 0.6858,
      "step": 685
    },
    {
      "epoch": 0.40588235294117647,
      "grad_norm": 3.589250867729836,
      "learning_rate": 7.4105675128517456e-06,
      "loss": 0.6911,
      "step": 690
    },
    {
      "epoch": 0.4088235294117647,
      "grad_norm": 3.459912742752029,
      "learning_rate": 7.365467784180051e-06,
      "loss": 0.6931,
      "step": 695
    },
    {
      "epoch": 0.4117647058823529,
      "grad_norm": 7.570102416020596,
      "learning_rate": 7.320118728046818e-06,
      "loss": 0.6919,
      "step": 700
    },
    {
      "epoch": 0.4147058823529412,
      "grad_norm": 5.7516115714170235,
      "learning_rate": 7.274525124379773e-06,
      "loss": 0.6613,
      "step": 705
    },
    {
      "epoch": 0.4176470588235294,
      "grad_norm": 6.937823731044612,
      "learning_rate": 7.2286917788826926e-06,
      "loss": 0.6681,
      "step": 710
    },
    {
      "epoch": 0.42058823529411765,
      "grad_norm": 2.702492838292246,
      "learning_rate": 7.182623522528866e-06,
      "loss": 0.6621,
      "step": 715
    },
    {
      "epoch": 0.4235294117647059,
      "grad_norm": 4.510429847780033,
      "learning_rate": 7.136325211051905e-06,
      "loss": 0.6577,
      "step": 720
    },
    {
      "epoch": 0.4264705882352941,
      "grad_norm": 3.6330047784117503,
      "learning_rate": 7.089801724433918e-06,
      "loss": 0.6682,
      "step": 725
    },
    {
      "epoch": 0.4294117647058823,
      "grad_norm": 3.105718892497887,
      "learning_rate": 7.043057966391158e-06,
      "loss": 0.6632,
      "step": 730
    },
    {
      "epoch": 0.4323529411764706,
      "grad_norm": 5.111900502997939,
      "learning_rate": 6.996098863857155e-06,
      "loss": 0.6486,
      "step": 735
    },
    {
      "epoch": 0.43529411764705883,
      "grad_norm": 2.5268899583639315,
      "learning_rate": 6.948929366463397e-06,
      "loss": 0.6523,
      "step": 740
    },
    {
      "epoch": 0.43823529411764706,
      "grad_norm": 3.2780721096181407,
      "learning_rate": 6.9015544460176296e-06,
      "loss": 0.655,
      "step": 745
    },
    {
      "epoch": 0.4411764705882353,
      "grad_norm": 3.3381264245743596,
      "learning_rate": 6.8539790959798045e-06,
      "loss": 0.654,
      "step": 750
    },
    {
      "epoch": 0.4441176470588235,
      "grad_norm": 4.44121574928845,
      "learning_rate": 6.806208330935766e-06,
      "loss": 0.6372,
      "step": 755
    },
    {
      "epoch": 0.4470588235294118,
      "grad_norm": 2.685752664996779,
      "learning_rate": 6.758247186068684e-06,
      "loss": 0.6413,
      "step": 760
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.955512012069018,
      "learning_rate": 6.710100716628345e-06,
      "loss": 0.6516,
      "step": 765
    },
    {
      "epoch": 0.45294117647058824,
      "grad_norm": 3.1611374536081454,
      "learning_rate": 6.6617739973982985e-06,
      "loss": 0.6649,
      "step": 770
    },
    {
      "epoch": 0.45588235294117646,
      "grad_norm": 4.369165428910403,
      "learning_rate": 6.613272122160975e-06,
      "loss": 0.6622,
      "step": 775
    },
    {
      "epoch": 0.4588235294117647,
      "grad_norm": 3.614068134809545,
      "learning_rate": 6.5646002031607726e-06,
      "loss": 0.6469,
      "step": 780
    },
    {
      "epoch": 0.46176470588235297,
      "grad_norm": 2.902322450613806,
      "learning_rate": 6.515763370565218e-06,
      "loss": 0.631,
      "step": 785
    },
    {
      "epoch": 0.4647058823529412,
      "grad_norm": 5.870110269988736,
      "learning_rate": 6.466766771924231e-06,
      "loss": 0.652,
      "step": 790
    },
    {
      "epoch": 0.4676470588235294,
      "grad_norm": 3.9059862931450313,
      "learning_rate": 6.417615571627555e-06,
      "loss": 0.6318,
      "step": 795
    },
    {
      "epoch": 0.47058823529411764,
      "grad_norm": 2.8197589495167192,
      "learning_rate": 6.368314950360416e-06,
      "loss": 0.6322,
      "step": 800
    },
    {
      "epoch": 0.47352941176470587,
      "grad_norm": 3.418950733265166,
      "learning_rate": 6.318870104557459e-06,
      "loss": 0.6187,
      "step": 805
    },
    {
      "epoch": 0.4764705882352941,
      "grad_norm": 4.432452311916624,
      "learning_rate": 6.269286245855039e-06,
      "loss": 0.602,
      "step": 810
    },
    {
      "epoch": 0.47941176470588237,
      "grad_norm": 2.5812197112565887,
      "learning_rate": 6.219568600541886e-06,
      "loss": 0.6174,
      "step": 815
    },
    {
      "epoch": 0.4823529411764706,
      "grad_norm": 3.937227710483061,
      "learning_rate": 6.169722409008244e-06,
      "loss": 0.619,
      "step": 820
    },
    {
      "epoch": 0.4852941176470588,
      "grad_norm": 3.0605504256904443,
      "learning_rate": 6.119752925193516e-06,
      "loss": 0.6334,
      "step": 825
    },
    {
      "epoch": 0.48823529411764705,
      "grad_norm": 4.037595617734812,
      "learning_rate": 6.0696654160324875e-06,
      "loss": 0.6279,
      "step": 830
    },
    {
      "epoch": 0.49117647058823527,
      "grad_norm": 5.106015487464359,
      "learning_rate": 6.019465160900173e-06,
      "loss": 0.6354,
      "step": 835
    },
    {
      "epoch": 0.49411764705882355,
      "grad_norm": 3.845547749417895,
      "learning_rate": 5.9691574510553505e-06,
      "loss": 0.6415,
      "step": 840
    },
    {
      "epoch": 0.4970588235294118,
      "grad_norm": 3.1754711810999394,
      "learning_rate": 5.918747589082853e-06,
      "loss": 0.6243,
      "step": 845
    },
    {
      "epoch": 0.5,
      "grad_norm": 3.8222639828081775,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.6065,
      "step": 850
    },
    {
      "epoch": 0.5029411764705882,
      "grad_norm": 3.5634408734197054,
      "learning_rate": 5.817642672369825e-06,
      "loss": 0.6008,
      "step": 855
    },
    {
      "epoch": 0.5058823529411764,
      "grad_norm": 2.923823486420052,
      "learning_rate": 5.766958274393428e-06,
      "loss": 0.6266,
      "step": 860
    },
    {
      "epoch": 0.5088235294117647,
      "grad_norm": 3.4223776040897373,
      "learning_rate": 5.716193036694359e-06,
      "loss": 0.6141,
      "step": 865
    },
    {
      "epoch": 0.5117647058823529,
      "grad_norm": 3.036605144513134,
      "learning_rate": 5.66535231008227e-06,
      "loss": 0.598,
      "step": 870
    },
    {
      "epoch": 0.5147058823529411,
      "grad_norm": 2.9215046859208695,
      "learning_rate": 5.614441453323571e-06,
      "loss": 0.6074,
      "step": 875
    },
    {
      "epoch": 0.5176470588235295,
      "grad_norm": 3.5327811692579387,
      "learning_rate": 5.5634658325766066e-06,
      "loss": 0.6052,
      "step": 880
    },
    {
      "epoch": 0.5205882352941177,
      "grad_norm": 4.232585058220642,
      "learning_rate": 5.512430820826035e-06,
      "loss": 0.6165,
      "step": 885
    },
    {
      "epoch": 0.5235294117647059,
      "grad_norm": 7.156188965411479,
      "learning_rate": 5.46134179731651e-06,
      "loss": 0.6012,
      "step": 890
    },
    {
      "epoch": 0.5264705882352941,
      "grad_norm": 3.0787975216808356,
      "learning_rate": 5.41020414698569e-06,
      "loss": 0.5919,
      "step": 895
    },
    {
      "epoch": 0.5294117647058824,
      "grad_norm": 2.641117254170942,
      "learning_rate": 5.359023259896638e-06,
      "loss": 0.5921,
      "step": 900
    },
    {
      "epoch": 0.5323529411764706,
      "grad_norm": 5.7492713146889205,
      "learning_rate": 5.3078045306697154e-06,
      "loss": 0.6009,
      "step": 905
    },
    {
      "epoch": 0.5352941176470588,
      "grad_norm": 3.0420972441444984,
      "learning_rate": 5.2565533579139484e-06,
      "loss": 0.5803,
      "step": 910
    },
    {
      "epoch": 0.538235294117647,
      "grad_norm": 4.94141259844524,
      "learning_rate": 5.205275143658018e-06,
      "loss": 0.5777,
      "step": 915
    },
    {
      "epoch": 0.5411764705882353,
      "grad_norm": 2.8294470597059362,
      "learning_rate": 5.153975292780852e-06,
      "loss": 0.5853,
      "step": 920
    },
    {
      "epoch": 0.5441176470588235,
      "grad_norm": 2.657834885061226,
      "learning_rate": 5.102659212441953e-06,
      "loss": 0.5753,
      "step": 925
    },
    {
      "epoch": 0.5470588235294118,
      "grad_norm": 2.5027366282176264,
      "learning_rate": 5.05133231151145e-06,
      "loss": 0.5979,
      "step": 930
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.972259109016088,
      "learning_rate": 5e-06,
      "loss": 0.5934,
      "step": 935
    },
    {
      "epoch": 0.5529411764705883,
      "grad_norm": 3.7344170961635235,
      "learning_rate": 4.948667688488552e-06,
      "loss": 0.5765,
      "step": 940
    },
    {
      "epoch": 0.5558823529411765,
      "grad_norm": 3.6094582616510187,
      "learning_rate": 4.8973407875580485e-06,
      "loss": 0.5763,
      "step": 945
    },
    {
      "epoch": 0.5588235294117647,
      "grad_norm": 3.0365592437357183,
      "learning_rate": 4.846024707219149e-06,
      "loss": 0.5637,
      "step": 950
    },
    {
      "epoch": 0.5617647058823529,
      "grad_norm": 2.689099476865793,
      "learning_rate": 4.794724856341985e-06,
      "loss": 0.5702,
      "step": 955
    },
    {
      "epoch": 0.5647058823529412,
      "grad_norm": 3.1936877032969977,
      "learning_rate": 4.7434466420860515e-06,
      "loss": 0.566,
      "step": 960
    },
    {
      "epoch": 0.5676470588235294,
      "grad_norm": 3.3687980239848976,
      "learning_rate": 4.692195469330286e-06,
      "loss": 0.5855,
      "step": 965
    },
    {
      "epoch": 0.5705882352941176,
      "grad_norm": 3.652904628883288,
      "learning_rate": 4.640976740103363e-06,
      "loss": 0.5712,
      "step": 970
    },
    {
      "epoch": 0.5735294117647058,
      "grad_norm": 2.6471377767162116,
      "learning_rate": 4.589795853014313e-06,
      "loss": 0.5764,
      "step": 975
    },
    {
      "epoch": 0.5764705882352941,
      "grad_norm": 3.4072798307585845,
      "learning_rate": 4.53865820268349e-06,
      "loss": 0.5584,
      "step": 980
    },
    {
      "epoch": 0.5794117647058824,
      "grad_norm": 2.9495397115764717,
      "learning_rate": 4.4875691791739655e-06,
      "loss": 0.5603,
      "step": 985
    },
    {
      "epoch": 0.5823529411764706,
      "grad_norm": 2.967359329887194,
      "learning_rate": 4.436534167423395e-06,
      "loss": 0.5606,
      "step": 990
    },
    {
      "epoch": 0.5852941176470589,
      "grad_norm": 3.6733313275854824,
      "learning_rate": 4.3855585466764305e-06,
      "loss": 0.5753,
      "step": 995
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 3.3017502365781692,
      "learning_rate": 4.334647689917734e-06,
      "loss": 0.562,
      "step": 1000
    },
    {
      "epoch": 0.5911764705882353,
      "grad_norm": 2.60212577629923,
      "learning_rate": 4.283806963305644e-06,
      "loss": 0.5614,
      "step": 1005
    },
    {
      "epoch": 0.5941176470588235,
      "grad_norm": 3.2117623019305164,
      "learning_rate": 4.233041725606573e-06,
      "loss": 0.5561,
      "step": 1010
    },
    {
      "epoch": 0.5970588235294118,
      "grad_norm": 2.5342533645318697,
      "learning_rate": 4.182357327630175e-06,
      "loss": 0.5456,
      "step": 1015
    },
    {
      "epoch": 0.6,
      "grad_norm": 3.491419616001698,
      "learning_rate": 4.131759111665349e-06,
      "loss": 0.5458,
      "step": 1020
    },
    {
      "epoch": 0.6029411764705882,
      "grad_norm": 2.5355347312480108,
      "learning_rate": 4.081252410917148e-06,
      "loss": 0.5639,
      "step": 1025
    },
    {
      "epoch": 0.6058823529411764,
      "grad_norm": 3.9415364552013177,
      "learning_rate": 4.03084254894465e-06,
      "loss": 0.5532,
      "step": 1030
    },
    {
      "epoch": 0.6088235294117647,
      "grad_norm": 4.101652297241306,
      "learning_rate": 3.980534839099829e-06,
      "loss": 0.5513,
      "step": 1035
    },
    {
      "epoch": 0.611764705882353,
      "grad_norm": 2.8200195019787944,
      "learning_rate": 3.930334583967514e-06,
      "loss": 0.543,
      "step": 1040
    },
    {
      "epoch": 0.6147058823529412,
      "grad_norm": 5.052470209996225,
      "learning_rate": 3.8802470748064855e-06,
      "loss": 0.5317,
      "step": 1045
    },
    {
      "epoch": 0.6176470588235294,
      "grad_norm": 3.0712588067545545,
      "learning_rate": 3.8302775909917585e-06,
      "loss": 0.5276,
      "step": 1050
    },
    {
      "epoch": 0.6205882352941177,
      "grad_norm": 2.8843549116949676,
      "learning_rate": 3.7804313994581143e-06,
      "loss": 0.5436,
      "step": 1055
    },
    {
      "epoch": 0.6235294117647059,
      "grad_norm": 3.71731186768865,
      "learning_rate": 3.730713754144961e-06,
      "loss": 0.542,
      "step": 1060
    },
    {
      "epoch": 0.6264705882352941,
      "grad_norm": 3.195039960104732,
      "learning_rate": 3.68112989544254e-06,
      "loss": 0.5569,
      "step": 1065
    },
    {
      "epoch": 0.6294117647058823,
      "grad_norm": 3.0760143966442426,
      "learning_rate": 3.6316850496395863e-06,
      "loss": 0.5279,
      "step": 1070
    },
    {
      "epoch": 0.6323529411764706,
      "grad_norm": 2.706150933189529,
      "learning_rate": 3.5823844283724464e-06,
      "loss": 0.5325,
      "step": 1075
    },
    {
      "epoch": 0.6352941176470588,
      "grad_norm": 2.8909978900989683,
      "learning_rate": 3.5332332280757706e-06,
      "loss": 0.5275,
      "step": 1080
    },
    {
      "epoch": 0.638235294117647,
      "grad_norm": 2.4810301000022346,
      "learning_rate": 3.484236629434783e-06,
      "loss": 0.5449,
      "step": 1085
    },
    {
      "epoch": 0.6411764705882353,
      "grad_norm": 2.6110146697831036,
      "learning_rate": 3.4353997968392295e-06,
      "loss": 0.5348,
      "step": 1090
    },
    {
      "epoch": 0.6441176470588236,
      "grad_norm": 2.6207736282199208,
      "learning_rate": 3.386727877839027e-06,
      "loss": 0.5362,
      "step": 1095
    },
    {
      "epoch": 0.6470588235294118,
      "grad_norm": 2.81407811382054,
      "learning_rate": 3.3382260026017027e-06,
      "loss": 0.5238,
      "step": 1100
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.572013835784688,
      "learning_rate": 3.289899283371657e-06,
      "loss": 0.5251,
      "step": 1105
    },
    {
      "epoch": 0.6529411764705882,
      "grad_norm": 2.587794908048521,
      "learning_rate": 3.241752813931316e-06,
      "loss": 0.5362,
      "step": 1110
    },
    {
      "epoch": 0.6558823529411765,
      "grad_norm": 2.6632456475884423,
      "learning_rate": 3.1937916690642356e-06,
      "loss": 0.524,
      "step": 1115
    },
    {
      "epoch": 0.6588235294117647,
      "grad_norm": 3.2994948697776407,
      "learning_rate": 3.1460209040201967e-06,
      "loss": 0.5246,
      "step": 1120
    },
    {
      "epoch": 0.6617647058823529,
      "grad_norm": 2.5014355845754426,
      "learning_rate": 3.098445553982372e-06,
      "loss": 0.5201,
      "step": 1125
    },
    {
      "epoch": 0.6647058823529411,
      "grad_norm": 3.615311051852331,
      "learning_rate": 3.0510706335366034e-06,
      "loss": 0.5328,
      "step": 1130
    },
    {
      "epoch": 0.6676470588235294,
      "grad_norm": 2.8016358462604716,
      "learning_rate": 3.0039011361428466e-06,
      "loss": 0.5307,
      "step": 1135
    },
    {
      "epoch": 0.6705882352941176,
      "grad_norm": 2.5797528822198896,
      "learning_rate": 2.956942033608843e-06,
      "loss": 0.5086,
      "step": 1140
    },
    {
      "epoch": 0.6735294117647059,
      "grad_norm": 3.703697916863966,
      "learning_rate": 2.910198275566085e-06,
      "loss": 0.5119,
      "step": 1145
    },
    {
      "epoch": 0.6764705882352942,
      "grad_norm": 2.4989182080083716,
      "learning_rate": 2.863674788948097e-06,
      "loss": 0.506,
      "step": 1150
    },
    {
      "epoch": 0.6794117647058824,
      "grad_norm": 2.926008906071238,
      "learning_rate": 2.817376477471132e-06,
      "loss": 0.5061,
      "step": 1155
    },
    {
      "epoch": 0.6823529411764706,
      "grad_norm": 2.5057586009739117,
      "learning_rate": 2.771308221117309e-06,
      "loss": 0.5158,
      "step": 1160
    },
    {
      "epoch": 0.6852941176470588,
      "grad_norm": 2.556924814964562,
      "learning_rate": 2.725474875620228e-06,
      "loss": 0.5241,
      "step": 1165
    },
    {
      "epoch": 0.6882352941176471,
      "grad_norm": 2.6209251376261586,
      "learning_rate": 2.6798812719531843e-06,
      "loss": 0.5139,
      "step": 1170
    },
    {
      "epoch": 0.6911764705882353,
      "grad_norm": 3.565884155626241,
      "learning_rate": 2.6345322158199503e-06,
      "loss": 0.5196,
      "step": 1175
    },
    {
      "epoch": 0.6941176470588235,
      "grad_norm": 2.711017441488018,
      "learning_rate": 2.5894324871482557e-06,
      "loss": 0.4967,
      "step": 1180
    },
    {
      "epoch": 0.6970588235294117,
      "grad_norm": 2.4420021545344355,
      "learning_rate": 2.544586839585961e-06,
      "loss": 0.4988,
      "step": 1185
    },
    {
      "epoch": 0.7,
      "grad_norm": 3.318545670339647,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.4935,
      "step": 1190
    },
    {
      "epoch": 0.7029411764705882,
      "grad_norm": 2.7341852078487987,
      "learning_rate": 2.4556766679781763e-06,
      "loss": 0.4996,
      "step": 1195
    },
    {
      "epoch": 0.7058823529411765,
      "grad_norm": 2.5448910575508994,
      "learning_rate": 2.411621515333788e-06,
      "loss": 0.5095,
      "step": 1200
    },
    {
      "epoch": 0.7088235294117647,
      "grad_norm": 2.3716682194974448,
      "learning_rate": 2.3678391856132203e-06,
      "loss": 0.4869,
      "step": 1205
    },
    {
      "epoch": 0.711764705882353,
      "grad_norm": 2.5986125491232017,
      "learning_rate": 2.324334293606499e-06,
      "loss": 0.4964,
      "step": 1210
    },
    {
      "epoch": 0.7147058823529412,
      "grad_norm": 2.6704869293015565,
      "learning_rate": 2.2811114248608675e-06,
      "loss": 0.4851,
      "step": 1215
    },
    {
      "epoch": 0.7176470588235294,
      "grad_norm": 2.434069159222151,
      "learning_rate": 2.238175135197471e-06,
      "loss": 0.5144,
      "step": 1220
    },
    {
      "epoch": 0.7205882352941176,
      "grad_norm": 2.8751039146064996,
      "learning_rate": 2.1955299502311523e-06,
      "loss": 0.4893,
      "step": 1225
    },
    {
      "epoch": 0.7235294117647059,
      "grad_norm": 2.8716211792398103,
      "learning_rate": 2.1531803648934333e-06,
      "loss": 0.4998,
      "step": 1230
    },
    {
      "epoch": 0.7264705882352941,
      "grad_norm": 2.5538647251200417,
      "learning_rate": 2.1111308429587446e-06,
      "loss": 0.4968,
      "step": 1235
    },
    {
      "epoch": 0.7294117647058823,
      "grad_norm": 2.3552149327326406,
      "learning_rate": 2.069385816573928e-06,
      "loss": 0.4751,
      "step": 1240
    },
    {
      "epoch": 0.7323529411764705,
      "grad_norm": 2.9092840839893994,
      "learning_rate": 2.0279496857910667e-06,
      "loss": 0.4892,
      "step": 1245
    },
    {
      "epoch": 0.7352941176470589,
      "grad_norm": 2.7140627954551797,
      "learning_rate": 1.9868268181037186e-06,
      "loss": 0.5007,
      "step": 1250
    },
    {
      "epoch": 0.7382352941176471,
      "grad_norm": 2.494839501975104,
      "learning_rate": 1.9460215479865613e-06,
      "loss": 0.4845,
      "step": 1255
    },
    {
      "epoch": 0.7411764705882353,
      "grad_norm": 2.3723796356846556,
      "learning_rate": 1.9055381764385272e-06,
      "loss": 0.4829,
      "step": 1260
    },
    {
      "epoch": 0.7441176470588236,
      "grad_norm": 2.3842886824345113,
      "learning_rate": 1.865380970529469e-06,
      "loss": 0.4853,
      "step": 1265
    },
    {
      "epoch": 0.7470588235294118,
      "grad_norm": 3.0004523133733523,
      "learning_rate": 1.8255541629503865e-06,
      "loss": 0.4721,
      "step": 1270
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.17555574871206,
      "learning_rate": 1.7860619515673034e-06,
      "loss": 0.4744,
      "step": 1275
    },
    {
      "epoch": 0.7529411764705882,
      "grad_norm": 2.4034252951946233,
      "learning_rate": 1.746908498978791e-06,
      "loss": 0.4848,
      "step": 1280
    },
    {
      "epoch": 0.7558823529411764,
      "grad_norm": 2.57638925801068,
      "learning_rate": 1.708097932077213e-06,
      "loss": 0.4771,
      "step": 1285
    },
    {
      "epoch": 0.7588235294117647,
      "grad_norm": 2.648859686704048,
      "learning_rate": 1.6696343416137495e-06,
      "loss": 0.4977,
      "step": 1290
    },
    {
      "epoch": 0.7617647058823529,
      "grad_norm": 2.695298184877126,
      "learning_rate": 1.6315217817672142e-06,
      "loss": 0.4785,
      "step": 1295
    },
    {
      "epoch": 0.7647058823529411,
      "grad_norm": 2.5055321455184485,
      "learning_rate": 1.5937642697167288e-06,
      "loss": 0.4734,
      "step": 1300
    },
    {
      "epoch": 0.7676470588235295,
      "grad_norm": 2.635547159269979,
      "learning_rate": 1.5563657852183072e-06,
      "loss": 0.4824,
      "step": 1305
    },
    {
      "epoch": 0.7705882352941177,
      "grad_norm": 2.3664290913138757,
      "learning_rate": 1.5193302701853674e-06,
      "loss": 0.4622,
      "step": 1310
    },
    {
      "epoch": 0.7735294117647059,
      "grad_norm": 3.119885061171242,
      "learning_rate": 1.4826616282732509e-06,
      "loss": 0.4718,
      "step": 1315
    },
    {
      "epoch": 0.7764705882352941,
      "grad_norm": 2.490972462452176,
      "learning_rate": 1.4463637244677648e-06,
      "loss": 0.4669,
      "step": 1320
    },
    {
      "epoch": 0.7794117647058824,
      "grad_norm": 2.458731892804391,
      "learning_rate": 1.410440384677791e-06,
      "loss": 0.4617,
      "step": 1325
    },
    {
      "epoch": 0.7823529411764706,
      "grad_norm": 2.958025742420681,
      "learning_rate": 1.374895395332037e-06,
      "loss": 0.4696,
      "step": 1330
    },
    {
      "epoch": 0.7852941176470588,
      "grad_norm": 2.9915004376321104,
      "learning_rate": 1.339732502979928e-06,
      "loss": 0.4636,
      "step": 1335
    },
    {
      "epoch": 0.788235294117647,
      "grad_norm": 2.966122147337875,
      "learning_rate": 1.3049554138967052e-06,
      "loss": 0.4506,
      "step": 1340
    },
    {
      "epoch": 0.7911764705882353,
      "grad_norm": 2.37703783512199,
      "learning_rate": 1.2705677936927841e-06,
      "loss": 0.47,
      "step": 1345
    },
    {
      "epoch": 0.7941176470588235,
      "grad_norm": 2.378564585605087,
      "learning_rate": 1.2365732669273778e-06,
      "loss": 0.4494,
      "step": 1350
    },
    {
      "epoch": 0.7970588235294118,
      "grad_norm": 2.9848625505080766,
      "learning_rate": 1.202975416726464e-06,
      "loss": 0.4545,
      "step": 1355
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.30742224730397,
      "learning_rate": 1.1697777844051105e-06,
      "loss": 0.4661,
      "step": 1360
    },
    {
      "epoch": 0.8029411764705883,
      "grad_norm": 2.6109048720465386,
      "learning_rate": 1.1369838690942059e-06,
      "loss": 0.4533,
      "step": 1365
    },
    {
      "epoch": 0.8058823529411765,
      "grad_norm": 2.9803797967996633,
      "learning_rate": 1.1045971273716476e-06,
      "loss": 0.4437,
      "step": 1370
    },
    {
      "epoch": 0.8088235294117647,
      "grad_norm": 2.7563794002126416,
      "learning_rate": 1.072620972898007e-06,
      "loss": 0.4524,
      "step": 1375
    },
    {
      "epoch": 0.8117647058823529,
      "grad_norm": 2.397027701415243,
      "learning_rate": 1.0410587760567104e-06,
      "loss": 0.4648,
      "step": 1380
    },
    {
      "epoch": 0.8147058823529412,
      "grad_norm": 2.862646785242907,
      "learning_rate": 1.0099138635988026e-06,
      "loss": 0.4592,
      "step": 1385
    },
    {
      "epoch": 0.8176470588235294,
      "grad_norm": 2.703458227745046,
      "learning_rate": 9.791895182922911e-07,
      "loss": 0.4656,
      "step": 1390
    },
    {
      "epoch": 0.8205882352941176,
      "grad_norm": 2.502457085827656,
      "learning_rate": 9.488889785761324e-07,
      "loss": 0.4487,
      "step": 1395
    },
    {
      "epoch": 0.8235294117647058,
      "grad_norm": 2.6423546848684167,
      "learning_rate": 9.190154382188921e-07,
      "loss": 0.4565,
      "step": 1400
    },
    {
      "epoch": 0.8264705882352941,
      "grad_norm": 2.532810863987098,
      "learning_rate": 8.895720459821089e-07,
      "loss": 0.4603,
      "step": 1405
    },
    {
      "epoch": 0.8294117647058824,
      "grad_norm": 2.24073062154865,
      "learning_rate": 8.605619052884106e-07,
      "loss": 0.455,
      "step": 1410
    },
    {
      "epoch": 0.8323529411764706,
      "grad_norm": 2.3347312069958477,
      "learning_rate": 8.31988073894403e-07,
      "loss": 0.4542,
      "step": 1415
    },
    {
      "epoch": 0.8352941176470589,
      "grad_norm": 2.2510986140412723,
      "learning_rate": 8.03853563568367e-07,
      "loss": 0.4422,
      "step": 1420
    },
    {
      "epoch": 0.8382352941176471,
      "grad_norm": 2.466059819140582,
      "learning_rate": 7.761613397728174e-07,
      "loss": 0.4436,
      "step": 1425
    },
    {
      "epoch": 0.8411764705882353,
      "grad_norm": 2.314552647197002,
      "learning_rate": 7.489143213519301e-07,
      "loss": 0.4526,
      "step": 1430
    },
    {
      "epoch": 0.8441176470588235,
      "grad_norm": 2.2167187458342186,
      "learning_rate": 7.221153802238845e-07,
      "loss": 0.4396,
      "step": 1435
    },
    {
      "epoch": 0.8470588235294118,
      "grad_norm": 2.4410281100307536,
      "learning_rate": 6.957673410781617e-07,
      "loss": 0.4493,
      "step": 1440
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.391369587902256,
      "learning_rate": 6.698729810778065e-07,
      "loss": 0.4559,
      "step": 1445
    },
    {
      "epoch": 0.8529411764705882,
      "grad_norm": 2.147240414992554,
      "learning_rate": 6.444350295667112e-07,
      "loss": 0.4423,
      "step": 1450
    },
    {
      "epoch": 0.8558823529411764,
      "grad_norm": 2.425220228026074,
      "learning_rate": 6.194561677819327e-07,
      "loss": 0.454,
      "step": 1455
    },
    {
      "epoch": 0.8588235294117647,
      "grad_norm": 2.478981866538809,
      "learning_rate": 5.949390285710777e-07,
      "loss": 0.4565,
      "step": 1460
    },
    {
      "epoch": 0.861764705882353,
      "grad_norm": 2.4324955757388484,
      "learning_rate": 5.708861961148004e-07,
      "loss": 0.4485,
      "step": 1465
    },
    {
      "epoch": 0.8647058823529412,
      "grad_norm": 2.518242605502862,
      "learning_rate": 5.473002056544191e-07,
      "loss": 0.4465,
      "step": 1470
    },
    {
      "epoch": 0.8676470588235294,
      "grad_norm": 2.4153603903758767,
      "learning_rate": 5.241835432246888e-07,
      "loss": 0.4449,
      "step": 1475
    },
    {
      "epoch": 0.8705882352941177,
      "grad_norm": 2.5154960957655463,
      "learning_rate": 5.015386453917742e-07,
      "loss": 0.4416,
      "step": 1480
    },
    {
      "epoch": 0.8735294117647059,
      "grad_norm": 2.5868546835610897,
      "learning_rate": 4.793678989964207e-07,
      "loss": 0.4343,
      "step": 1485
    },
    {
      "epoch": 0.8764705882352941,
      "grad_norm": 2.483503936880007,
      "learning_rate": 4.576736409023813e-07,
      "loss": 0.4375,
      "step": 1490
    },
    {
      "epoch": 0.8794117647058823,
      "grad_norm": 2.354484991161176,
      "learning_rate": 4.364581577500987e-07,
      "loss": 0.4396,
      "step": 1495
    },
    {
      "epoch": 0.8823529411764706,
      "grad_norm": 2.32433065376712,
      "learning_rate": 4.15723685715686e-07,
      "loss": 0.4417,
      "step": 1500
    },
    {
      "epoch": 0.8852941176470588,
      "grad_norm": 2.3785384367167173,
      "learning_rate": 3.9547241027523164e-07,
      "loss": 0.4372,
      "step": 1505
    },
    {
      "epoch": 0.888235294117647,
      "grad_norm": 2.2739855313885484,
      "learning_rate": 3.7570646597444196e-07,
      "loss": 0.4382,
      "step": 1510
    },
    {
      "epoch": 0.8911764705882353,
      "grad_norm": 2.4248149464102324,
      "learning_rate": 3.564279362036488e-07,
      "loss": 0.4504,
      "step": 1515
    },
    {
      "epoch": 0.8941176470588236,
      "grad_norm": 2.5206865046215623,
      "learning_rate": 3.3763885297822153e-07,
      "loss": 0.4425,
      "step": 1520
    },
    {
      "epoch": 0.8970588235294118,
      "grad_norm": 2.321436598632796,
      "learning_rate": 3.1934119672438093e-07,
      "loss": 0.4508,
      "step": 1525
    },
    {
      "epoch": 0.9,
      "grad_norm": 2.2956489839218253,
      "learning_rate": 3.015368960704584e-07,
      "loss": 0.4218,
      "step": 1530
    },
    {
      "epoch": 0.9029411764705882,
      "grad_norm": 2.6380831075744045,
      "learning_rate": 2.842278276436128e-07,
      "loss": 0.4312,
      "step": 1535
    },
    {
      "epoch": 0.9058823529411765,
      "grad_norm": 2.353068804554375,
      "learning_rate": 2.6741581587202747e-07,
      "loss": 0.4572,
      "step": 1540
    },
    {
      "epoch": 0.9088235294117647,
      "grad_norm": 2.3504938166434286,
      "learning_rate": 2.511026327926114e-07,
      "loss": 0.4415,
      "step": 1545
    },
    {
      "epoch": 0.9117647058823529,
      "grad_norm": 2.4033670936853864,
      "learning_rate": 2.3528999786421758e-07,
      "loss": 0.4383,
      "step": 1550
    },
    {
      "epoch": 0.9147058823529411,
      "grad_norm": 2.5018773276624557,
      "learning_rate": 2.1997957778641166e-07,
      "loss": 0.448,
      "step": 1555
    },
    {
      "epoch": 0.9176470588235294,
      "grad_norm": 2.3367055632905753,
      "learning_rate": 2.0517298632379445e-07,
      "loss": 0.4497,
      "step": 1560
    },
    {
      "epoch": 0.9205882352941176,
      "grad_norm": 2.765125262874024,
      "learning_rate": 1.908717841359048e-07,
      "loss": 0.4507,
      "step": 1565
    },
    {
      "epoch": 0.9235294117647059,
      "grad_norm": 2.376902313703198,
      "learning_rate": 1.770774786127244e-07,
      "loss": 0.4183,
      "step": 1570
    },
    {
      "epoch": 0.9264705882352942,
      "grad_norm": 2.413442304678235,
      "learning_rate": 1.6379152371579277e-07,
      "loss": 0.4334,
      "step": 1575
    },
    {
      "epoch": 0.9294117647058824,
      "grad_norm": 2.3143531531697596,
      "learning_rate": 1.510153198249531e-07,
      "loss": 0.436,
      "step": 1580
    },
    {
      "epoch": 0.9323529411764706,
      "grad_norm": 2.34608810626235,
      "learning_rate": 1.3875021359075257e-07,
      "loss": 0.44,
      "step": 1585
    },
    {
      "epoch": 0.9352941176470588,
      "grad_norm": 2.35355433201976,
      "learning_rate": 1.2699749779249926e-07,
      "loss": 0.4314,
      "step": 1590
    },
    {
      "epoch": 0.9382352941176471,
      "grad_norm": 2.218100203321306,
      "learning_rate": 1.157584112019966e-07,
      "loss": 0.4384,
      "step": 1595
    },
    {
      "epoch": 0.9411764705882353,
      "grad_norm": 2.3877266166874676,
      "learning_rate": 1.0503413845297739e-07,
      "loss": 0.4357,
      "step": 1600
    },
    {
      "epoch": 0.9441176470588235,
      "grad_norm": 2.3187529416036936,
      "learning_rate": 9.482580991623747e-08,
      "loss": 0.4323,
      "step": 1605
    },
    {
      "epoch": 0.9470588235294117,
      "grad_norm": 2.2887581674114648,
      "learning_rate": 8.513450158049109e-08,
      "loss": 0.4448,
      "step": 1610
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.3836470257770945,
      "learning_rate": 7.59612349389599e-08,
      "loss": 0.4269,
      "step": 1615
    },
    {
      "epoch": 0.9529411764705882,
      "grad_norm": 2.233426183677112,
      "learning_rate": 6.730697688170251e-08,
      "loss": 0.4351,
      "step": 1620
    },
    {
      "epoch": 0.9558823529411765,
      "grad_norm": 2.2576790941626284,
      "learning_rate": 5.917263959370312e-08,
      "loss": 0.4355,
      "step": 1625
    },
    {
      "epoch": 0.9588235294117647,
      "grad_norm": 2.2984294403136163,
      "learning_rate": 5.155908045872349e-08,
      "loss": 0.4406,
      "step": 1630
    },
    {
      "epoch": 0.961764705882353,
      "grad_norm": 2.3707792344113945,
      "learning_rate": 4.446710196893245e-08,
      "loss": 0.4385,
      "step": 1635
    },
    {
      "epoch": 0.9647058823529412,
      "grad_norm": 2.2246371108164262,
      "learning_rate": 3.7897451640321326e-08,
      "loss": 0.4246,
      "step": 1640
    },
    {
      "epoch": 0.9676470588235294,
      "grad_norm": 2.622284137214347,
      "learning_rate": 3.185082193391143e-08,
      "loss": 0.4279,
      "step": 1645
    },
    {
      "epoch": 0.9705882352941176,
      "grad_norm": 2.3946023382142383,
      "learning_rate": 2.6327850182769065e-08,
      "loss": 0.4573,
      "step": 1650
    },
    {
      "epoch": 0.9735294117647059,
      "grad_norm": 2.354547522725208,
      "learning_rate": 2.1329118524827662e-08,
      "loss": 0.4405,
      "step": 1655
    },
    {
      "epoch": 0.9764705882352941,
      "grad_norm": 2.3346382733247815,
      "learning_rate": 1.6855153841527915e-08,
      "loss": 0.4386,
      "step": 1660
    },
    {
      "epoch": 0.9794117647058823,
      "grad_norm": 2.3205008178624764,
      "learning_rate": 1.2906427702284452e-08,
      "loss": 0.4304,
      "step": 1665
    },
    {
      "epoch": 0.9823529411764705,
      "grad_norm": 2.4270314026839714,
      "learning_rate": 9.48335631477948e-09,
      "loss": 0.4415,
      "step": 1670
    },
    {
      "epoch": 0.9852941176470589,
      "grad_norm": 2.6174654383893214,
      "learning_rate": 6.586300481095098e-09,
      "loss": 0.4272,
      "step": 1675
    },
    {
      "epoch": 0.9882352941176471,
      "grad_norm": 2.2960002654994973,
      "learning_rate": 4.2155655596809455e-09,
      "loss": 0.4334,
      "step": 1680
    },
    {
      "epoch": 0.9911764705882353,
      "grad_norm": 2.443491425156529,
      "learning_rate": 2.371401433170495e-09,
      "loss": 0.4249,
      "step": 1685
    },
    {
      "epoch": 0.9941176470588236,
      "grad_norm": 2.2442264144591926,
      "learning_rate": 1.054002482043237e-09,
      "loss": 0.428,
      "step": 1690
    },
    {
      "epoch": 0.9970588235294118,
      "grad_norm": 2.4769821745544953,
      "learning_rate": 2.6350756413440203e-10,
      "loss": 0.4334,
      "step": 1695
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.3764816774266806,
      "learning_rate": 0.0,
      "loss": 0.4299,
      "step": 1700
    },
    {
      "epoch": 1.0,
      "eval_runtime": 4.4513,
      "eval_samples_per_second": 2.247,
      "eval_steps_per_second": 0.674,
      "step": 1700
    },
    {
      "epoch": 1.0,
      "step": 1700,
      "total_flos": 117051280588800.0,
      "train_loss": 0.691351400122923,
      "train_runtime": 20676.98,
      "train_samples_per_second": 1.315,
      "train_steps_per_second": 0.082
    }
  ],
  "logging_steps": 5,
  "max_steps": 1700,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 117051280588800.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}