{
  "best_metric": 94.22774869109946,
  "best_model_checkpoint": "outputs/bitfit/t5-base/qqp/checkpoint-8600",
  "epoch": 3.0,
  "global_step": 34017,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "eval_accuracy": 93.60000000000001,
      "eval_average_metrics": 92.41643835616439,
      "eval_f1": 91.23287671232877,
      "eval_loss": 0.06451932340860367,
      "eval_runtime": 4.5218,
      "eval_samples_per_second": 221.15,
      "step": 200
    },
    {
      "epoch": 0.04,
      "eval_accuracy": 90.2,
      "eval_average_metrics": 89.15339805825244,
      "eval_f1": 88.10679611650487,
      "eval_loss": 0.08826350420713425,
      "eval_runtime": 4.5364,
      "eval_samples_per_second": 220.438,
      "step": 400
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.00029559044007408056,
      "loss": 0.1448,
      "step": 500
    },
    {
      "epoch": 0.05,
      "eval_accuracy": 93.7,
      "eval_average_metrics": 92.76439688715953,
      "eval_f1": 91.82879377431907,
      "eval_loss": 0.061929114162921906,
      "eval_runtime": 4.5766,
      "eval_samples_per_second": 218.503,
      "step": 600
    },
    {
      "epoch": 0.07,
      "eval_accuracy": 93.30000000000001,
      "eval_average_metrics": 92.3385456885457,
      "eval_f1": 91.37709137709138,
      "eval_loss": 0.0669504851102829,
      "eval_runtime": 4.5237,
      "eval_samples_per_second": 221.059,
      "step": 800
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002911808801481612,
      "loss": 0.0782,
      "step": 1000
    },
    {
      "epoch": 0.09,
      "eval_accuracy": 93.7,
      "eval_average_metrics": 92.74308996088658,
      "eval_f1": 91.78617992177314,
      "eval_loss": 0.06603064388036728,
      "eval_runtime": 4.727,
      "eval_samples_per_second": 211.549,
      "step": 1000
    },
    {
      "epoch": 0.11,
      "eval_accuracy": 93.30000000000001,
      "eval_average_metrics": 92.14125168236878,
      "eval_f1": 90.98250336473754,
      "eval_loss": 0.06394415348768234,
      "eval_runtime": 4.6179,
      "eval_samples_per_second": 216.548,
      "step": 1200
    },
    {
      "epoch": 0.12,
      "eval_accuracy": 93.30000000000001,
      "eval_average_metrics": 92.07974079126876,
      "eval_f1": 90.85948158253751,
      "eval_loss": 0.059915054589509964,
      "eval_runtime": 4.4912,
      "eval_samples_per_second": 222.655,
      "step": 1400
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002867713202222418,
      "loss": 0.0773,
      "step": 1500
    },
    {
      "epoch": 0.14,
      "eval_accuracy": 93.0,
      "eval_average_metrics": 92.0242966751918,
      "eval_f1": 91.04859335038363,
      "eval_loss": 0.06896140426397324,
      "eval_runtime": 4.614,
      "eval_samples_per_second": 216.732,
      "step": 1600
    },
    {
      "epoch": 0.16,
      "eval_accuracy": 93.60000000000001,
      "eval_average_metrics": 92.61151832460735,
      "eval_f1": 91.62303664921467,
      "eval_loss": 0.06234096363186836,
      "eval_runtime": 4.6818,
      "eval_samples_per_second": 213.593,
      "step": 1800
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002823617602963224,
      "loss": 0.0746,
      "step": 2000
    },
    {
      "epoch": 0.18,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.23333333333332,
      "eval_f1": 92.26666666666667,
      "eval_loss": 0.0571160614490509,
      "eval_runtime": 4.7897,
      "eval_samples_per_second": 208.783,
      "step": 2000
    },
    {
      "epoch": 0.19,
      "eval_accuracy": 93.60000000000001,
      "eval_average_metrics": 92.54468085106384,
      "eval_f1": 91.48936170212767,
      "eval_loss": 0.059621669352054596,
      "eval_runtime": 4.6087,
      "eval_samples_per_second": 216.979,
      "step": 2200
    },
    {
      "epoch": 0.21,
      "eval_accuracy": 93.60000000000001,
      "eval_average_metrics": 92.36786703601109,
      "eval_f1": 91.13573407202216,
      "eval_loss": 0.06162749230861664,
      "eval_runtime": 4.7312,
      "eval_samples_per_second": 211.361,
      "step": 2400
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00027795220037040303,
      "loss": 0.0771,
      "step": 2500
    },
    {
      "epoch": 0.23,
      "eval_accuracy": 93.10000000000001,
      "eval_average_metrics": 91.91912751677853,
      "eval_f1": 90.73825503355705,
      "eval_loss": 0.06196223199367523,
      "eval_runtime": 4.4778,
      "eval_samples_per_second": 223.322,
      "step": 2600
    },
    {
      "epoch": 0.25,
      "eval_accuracy": 93.7,
      "eval_average_metrics": 92.72155963302754,
      "eval_f1": 91.74311926605506,
      "eval_loss": 0.059104129672050476,
      "eval_runtime": 4.462,
      "eval_samples_per_second": 224.113,
      "step": 2800
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002735426404444836,
      "loss": 0.0764,
      "step": 3000
    },
    {
      "epoch": 0.26,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.1633069828722,
      "eval_f1": 92.2266139657444,
      "eval_loss": 0.059123676270246506,
      "eval_runtime": 4.5787,
      "eval_samples_per_second": 218.4,
      "step": 3000
    },
    {
      "epoch": 0.28,
      "eval_accuracy": 91.5,
      "eval_average_metrics": 90.49660074165637,
      "eval_f1": 89.49320148331273,
      "eval_loss": 0.07883985340595245,
      "eval_runtime": 4.5148,
      "eval_samples_per_second": 221.496,
      "step": 3200
    },
    {
      "epoch": 0.3,
      "eval_accuracy": 93.10000000000001,
      "eval_average_metrics": 92.02837483617301,
      "eval_f1": 90.956749672346,
      "eval_loss": 0.06356123834848404,
      "eval_runtime": 4.5836,
      "eval_samples_per_second": 218.171,
      "step": 3400
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002691330805185642,
      "loss": 0.0732,
      "step": 3500
    },
    {
      "epoch": 0.32,
      "eval_accuracy": 93.7,
      "eval_average_metrics": 92.598987854251,
      "eval_f1": 91.49797570850201,
      "eval_loss": 0.058661118149757385,
      "eval_runtime": 4.581,
      "eval_samples_per_second": 218.292,
      "step": 3600
    },
    {
      "epoch": 0.34,
      "eval_accuracy": 93.10000000000001,
      "eval_average_metrics": 92.00454545454545,
      "eval_f1": 90.90909090909089,
      "eval_loss": 0.06345341354608536,
      "eval_runtime": 4.6337,
      "eval_samples_per_second": 215.808,
      "step": 3800
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00026472352059264486,
      "loss": 0.0745,
      "step": 4000
    },
    {
      "epoch": 0.35,
      "eval_accuracy": 92.7,
      "eval_average_metrics": 91.72389100126742,
      "eval_f1": 90.74778200253485,
      "eval_loss": 0.07112478464841843,
      "eval_runtime": 4.5897,
      "eval_samples_per_second": 217.881,
      "step": 4000
    },
    {
      "epoch": 0.37,
      "eval_accuracy": 92.4,
      "eval_average_metrics": 91.38987341772153,
      "eval_f1": 90.37974683544304,
      "eval_loss": 0.06926184892654419,
      "eval_runtime": 4.5334,
      "eval_samples_per_second": 220.585,
      "step": 4200
    },
    {
      "epoch": 0.39,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.2021505376344,
      "eval_f1": 92.20430107526882,
      "eval_loss": 0.06072888895869255,
      "eval_runtime": 4.7545,
      "eval_samples_per_second": 210.329,
      "step": 4400
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.00026031396066672545,
      "loss": 0.0766,
      "step": 4500
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 93.89999999999999,
      "eval_average_metrics": 92.8779038718291,
      "eval_f1": 91.85580774365822,
      "eval_loss": 0.06117413192987442,
      "eval_runtime": 4.5554,
      "eval_samples_per_second": 219.52,
      "step": 4600
    },
    {
      "epoch": 0.42,
      "eval_accuracy": 94.0,
      "eval_average_metrics": 93.13402061855669,
      "eval_f1": 92.2680412371134,
      "eval_loss": 0.060421667993068695,
      "eval_runtime": 4.5444,
      "eval_samples_per_second": 220.049,
      "step": 4800
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00025590440074080604,
      "loss": 0.0729,
      "step": 5000
    },
    {
      "epoch": 0.44,
      "eval_accuracy": 93.0,
      "eval_average_metrics": 91.95454545454547,
      "eval_f1": 90.90909090909092,
      "eval_loss": 0.06169410049915314,
      "eval_runtime": 4.6688,
      "eval_samples_per_second": 214.188,
      "step": 5000
    },
    {
      "epoch": 0.46,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.3244966442953,
      "eval_f1": 92.34899328859059,
      "eval_loss": 0.0613214485347271,
      "eval_runtime": 4.5049,
      "eval_samples_per_second": 221.979,
      "step": 5200
    },
    {
      "epoch": 0.48,
      "eval_accuracy": 93.7,
      "eval_average_metrics": 92.71070959264125,
      "eval_f1": 91.72141918528251,
      "eval_loss": 0.06161003187298775,
      "eval_runtime": 4.5685,
      "eval_samples_per_second": 218.891,
      "step": 5400
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002514948408148867,
      "loss": 0.071,
      "step": 5500
    },
    {
      "epoch": 0.49,
      "eval_accuracy": 93.89999999999999,
      "eval_average_metrics": 92.95262123197902,
      "eval_f1": 92.00524246395806,
      "eval_loss": 0.06118384748697281,
      "eval_runtime": 4.564,
      "eval_samples_per_second": 219.107,
      "step": 5600
    },
    {
      "epoch": 0.51,
      "eval_accuracy": 93.0,
      "eval_average_metrics": 92.0470737913486,
      "eval_f1": 91.09414758269719,
      "eval_loss": 0.06946446746587753,
      "eval_runtime": 4.5787,
      "eval_samples_per_second": 218.4,
      "step": 5800
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002470852808889673,
      "loss": 0.0748,
      "step": 6000
    },
    {
      "epoch": 0.53,
      "eval_accuracy": 94.6,
      "eval_average_metrics": 93.7095744680851,
      "eval_f1": 92.81914893617021,
      "eval_loss": 0.05765092372894287,
      "eval_runtime": 4.5272,
      "eval_samples_per_second": 220.889,
      "step": 6000
    },
    {
      "epoch": 0.55,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.3244966442953,
      "eval_f1": 92.34899328859059,
      "eval_loss": 0.05894589051604271,
      "eval_runtime": 4.6099,
      "eval_samples_per_second": 216.924,
      "step": 6200
    },
    {
      "epoch": 0.56,
      "eval_accuracy": 93.7,
      "eval_average_metrics": 92.598987854251,
      "eval_f1": 91.49797570850201,
      "eval_loss": 0.061102479696273804,
      "eval_runtime": 4.6948,
      "eval_samples_per_second": 213.001,
      "step": 6400
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.00024267572096304786,
      "loss": 0.074,
      "step": 6500
    },
    {
      "epoch": 0.58,
      "eval_accuracy": 93.0,
      "eval_average_metrics": 92.03571428571429,
      "eval_f1": 91.07142857142858,
      "eval_loss": 0.06452207267284393,
      "eval_runtime": 4.6106,
      "eval_samples_per_second": 216.891,
      "step": 6600
    },
    {
      "epoch": 0.6,
      "eval_accuracy": 93.5,
      "eval_average_metrics": 92.55645161290323,
      "eval_f1": 91.61290322580645,
      "eval_loss": 0.05938281863927841,
      "eval_runtime": 4.5228,
      "eval_samples_per_second": 221.102,
      "step": 6800
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.00023826616103712848,
      "loss": 0.0738,
      "step": 7000
    },
    {
      "epoch": 0.62,
      "eval_accuracy": 94.6,
      "eval_average_metrics": 93.77519582245431,
      "eval_f1": 92.95039164490862,
      "eval_loss": 0.057858582586050034,
      "eval_runtime": 4.5704,
      "eval_samples_per_second": 218.797,
      "step": 7000
    },
    {
      "epoch": 0.63,
      "eval_accuracy": 94.6,
      "eval_average_metrics": 93.71909814323607,
      "eval_f1": 92.83819628647215,
      "eval_loss": 0.05671229586005211,
      "eval_runtime": 4.4966,
      "eval_samples_per_second": 222.39,
      "step": 7200
    },
    {
      "epoch": 0.65,
      "eval_accuracy": 94.5,
      "eval_average_metrics": 93.64580602883355,
      "eval_f1": 92.7916120576671,
      "eval_loss": 0.059491805732250214,
      "eval_runtime": 4.5973,
      "eval_samples_per_second": 217.521,
      "step": 7400
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00023385660111120907,
      "loss": 0.0746,
      "step": 7500
    },
    {
      "epoch": 0.67,
      "eval_accuracy": 94.5,
      "eval_average_metrics": 93.53879892037787,
      "eval_f1": 92.57759784075573,
      "eval_loss": 0.057486891746520996,
      "eval_runtime": 4.6372,
      "eval_samples_per_second": 215.649,
      "step": 7600
    },
    {
      "epoch": 0.69,
      "eval_accuracy": 94.69999999999999,
      "eval_average_metrics": 93.81194926568757,
      "eval_f1": 92.92389853137516,
      "eval_loss": 0.05628298968076706,
      "eval_runtime": 4.6937,
      "eval_samples_per_second": 213.051,
      "step": 7800
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002294470411852897,
      "loss": 0.0762,
      "step": 8000
    },
    {
      "epoch": 0.71,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.4566844919786,
      "eval_f1": 92.51336898395722,
      "eval_loss": 0.05849047377705574,
      "eval_runtime": 4.6139,
      "eval_samples_per_second": 216.737,
      "step": 8000
    },
    {
      "epoch": 0.72,
      "eval_accuracy": 94.69999999999999,
      "eval_average_metrics": 93.86773981603153,
      "eval_f1": 93.03547963206307,
      "eval_loss": 0.056792281568050385,
      "eval_runtime": 4.6916,
      "eval_samples_per_second": 213.147,
      "step": 8200
    },
    {
      "epoch": 0.74,
      "eval_accuracy": 94.5,
      "eval_average_metrics": 93.59794156706508,
      "eval_f1": 92.69588313413014,
      "eval_loss": 0.05638590082526207,
      "eval_runtime": 4.6952,
      "eval_samples_per_second": 212.982,
      "step": 8400
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002250374812593703,
      "loss": 0.0726,
      "step": 8500
    },
    {
      "epoch": 0.76,
      "eval_accuracy": 95.0,
      "eval_average_metrics": 94.22774869109946,
      "eval_f1": 93.45549738219894,
      "eval_loss": 0.055720701813697815,
      "eval_runtime": 4.5004,
      "eval_samples_per_second": 222.204,
      "step": 8600
    },
    {
      "epoch": 0.78,
      "eval_accuracy": 94.0,
      "eval_average_metrics": 93.08355091383812,
      "eval_f1": 92.16710182767625,
      "eval_loss": 0.06084197014570236,
      "eval_runtime": 4.5822,
      "eval_samples_per_second": 218.238,
      "step": 8800
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002206279213334509,
      "loss": 0.0734,
      "step": 9000
    },
    {
      "epoch": 0.79,
      "eval_accuracy": 93.10000000000001,
      "eval_average_metrics": 92.14386973180078,
      "eval_f1": 91.18773946360155,
      "eval_loss": 0.06530317664146423,
      "eval_runtime": 4.5035,
      "eval_samples_per_second": 222.05,
      "step": 9000
    },
    {
      "epoch": 0.81,
      "eval_accuracy": 93.8,
      "eval_average_metrics": 92.76666666666665,
      "eval_f1": 91.73333333333332,
      "eval_loss": 0.05946441367268562,
      "eval_runtime": 4.8109,
      "eval_samples_per_second": 207.861,
      "step": 9200
    },
    {
      "epoch": 0.83,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.4566844919786,
      "eval_f1": 92.51336898395722,
      "eval_loss": 0.059339020401239395,
      "eval_runtime": 4.5265,
      "eval_samples_per_second": 220.922,
      "step": 9400
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00021621836140753152,
      "loss": 0.0731,
      "step": 9500
    },
    {
      "epoch": 0.85,
      "eval_accuracy": 92.60000000000001,
      "eval_average_metrics": 91.64005037783375,
      "eval_f1": 90.6801007556675,
      "eval_loss": 0.07186109572649002,
      "eval_runtime": 4.5431,
      "eval_samples_per_second": 220.114,
      "step": 9600
    },
    {
      "epoch": 0.86,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.1633069828722,
      "eval_f1": 92.2266139657444,
      "eval_loss": 0.05946135148406029,
      "eval_runtime": 4.5655,
      "eval_samples_per_second": 219.036,
      "step": 9800
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002118088014816121,
      "loss": 0.0733,
      "step": 10000
    },
    {
      "epoch": 0.88,
      "eval_accuracy": 93.89999999999999,
      "eval_average_metrics": 92.95262123197902,
      "eval_f1": 92.00524246395806,
      "eval_loss": 0.06076710671186447,
      "eval_runtime": 4.5407,
      "eval_samples_per_second": 220.229,
      "step": 10000
    },
    {
      "epoch": 0.9,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.34492656875835,
      "eval_f1": 92.3898531375167,
      "eval_loss": 0.055939000099897385,
      "eval_runtime": 4.4996,
      "eval_samples_per_second": 222.24,
      "step": 10200
    },
    {
      "epoch": 0.92,
      "eval_accuracy": 94.8,
      "eval_average_metrics": 93.96084656084656,
      "eval_f1": 93.12169312169313,
      "eval_loss": 0.05636580288410187,
      "eval_runtime": 4.4643,
      "eval_samples_per_second": 223.998,
      "step": 10400
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.00020739924155569272,
      "loss": 0.0738,
      "step": 10500
    },
    {
      "epoch": 0.93,
      "eval_accuracy": 93.8,
      "eval_average_metrics": 92.71081081081081,
      "eval_f1": 91.62162162162161,
      "eval_loss": 0.059175312519073486,
      "eval_runtime": 4.7877,
      "eval_samples_per_second": 208.867,
      "step": 10600
    },
    {
      "epoch": 0.95,
      "eval_accuracy": 93.89999999999999,
      "eval_average_metrics": 92.91026490066224,
      "eval_f1": 91.9205298013245,
      "eval_loss": 0.059644319117069244,
      "eval_runtime": 4.6781,
      "eval_samples_per_second": 213.761,
      "step": 10800
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002029896816297733,
      "loss": 0.0752,
      "step": 11000
    },
    {
      "epoch": 0.97,
      "eval_accuracy": 93.60000000000001,
      "eval_average_metrics": 92.622454308094,
      "eval_f1": 91.644908616188,
      "eval_loss": 0.061212606728076935,
      "eval_runtime": 4.4874,
      "eval_samples_per_second": 222.845,
      "step": 11000
    },
    {
      "epoch": 0.99,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.41474442988203,
      "eval_f1": 92.52948885976409,
      "eval_loss": 0.059587035328149796,
      "eval_runtime": 4.5784,
      "eval_samples_per_second": 218.418,
      "step": 11200
    },
    {
      "epoch": 1.01,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.2533462033462,
      "eval_f1": 92.4066924066924,
      "eval_loss": 0.060919877141714096,
      "eval_runtime": 4.5416,
      "eval_samples_per_second": 220.185,
      "step": 11400
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00019858012170385393,
      "loss": 0.0716,
      "step": 11500
    },
    {
      "epoch": 1.02,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.11141522029372,
      "eval_f1": 92.12283044058745,
      "eval_loss": 0.05961688980460167,
      "eval_runtime": 4.5998,
      "eval_samples_per_second": 217.402,
      "step": 11600
    },
    {
      "epoch": 1.04,
      "eval_accuracy": 94.0,
      "eval_average_metrics": 93.0212201591512,
      "eval_f1": 92.04244031830238,
      "eval_loss": 0.06122226640582085,
      "eval_runtime": 4.6213,
      "eval_samples_per_second": 216.391,
      "step": 11800
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.00019417056177793455,
      "loss": 0.0713,
      "step": 12000
    },
    {
      "epoch": 1.06,
      "eval_accuracy": 94.0,
      "eval_average_metrics": 92.96774193548387,
      "eval_f1": 91.93548387096774,
      "eval_loss": 0.06119931861758232,
      "eval_runtime": 4.5888,
      "eval_samples_per_second": 217.92,
      "step": 12000
    },
    {
      "epoch": 1.08,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.31419919246298,
      "eval_f1": 92.32839838492598,
      "eval_loss": 0.05847727879881859,
      "eval_runtime": 4.4633,
      "eval_samples_per_second": 224.05,
      "step": 12200
    },
    {
      "epoch": 1.09,
      "eval_accuracy": 93.4,
      "eval_average_metrics": 92.39190600522193,
      "eval_f1": 91.38381201044386,
      "eval_loss": 0.06247144192457199,
      "eval_runtime": 4.5667,
      "eval_samples_per_second": 218.978,
      "step": 12400
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00018976100185201514,
      "loss": 0.0687,
      "step": 12500
    },
    {
      "epoch": 1.11,
      "eval_accuracy": 93.89999999999999,
      "eval_average_metrics": 93.04475032010242,
      "eval_f1": 92.18950064020484,
      "eval_loss": 0.0635332465171814,
      "eval_runtime": 4.5944,
      "eval_samples_per_second": 217.654,
      "step": 12600
    },
    {
      "epoch": 1.13,
      "eval_accuracy": 94.0,
      "eval_average_metrics": 93.05263157894737,
      "eval_f1": 92.10526315789474,
      "eval_loss": 0.06063272804021835,
      "eval_runtime": 4.5058,
      "eval_samples_per_second": 221.934,
      "step": 12800
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.00018535144192609576,
      "loss": 0.0711,
      "step": 13000
    },
    {
      "epoch": 1.15,
      "eval_accuracy": 94.0,
      "eval_average_metrics": 93.12403100775194,
      "eval_f1": 92.24806201550389,
      "eval_loss": 0.06045162305235863,
      "eval_runtime": 4.6598,
      "eval_samples_per_second": 214.601,
      "step": 13000
    },
    {
      "epoch": 1.16,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.50606860158311,
      "eval_f1": 92.61213720316623,
      "eval_loss": 0.06117108836770058,
      "eval_runtime": 4.501,
      "eval_samples_per_second": 222.172,
      "step": 13200
    },
    {
      "epoch": 1.18,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.4962962962963,
      "eval_f1": 92.5925925925926,
      "eval_loss": 0.05846463143825531,
      "eval_runtime": 4.4931,
      "eval_samples_per_second": 222.561,
      "step": 13400
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.00018094188200017637,
      "loss": 0.0694,
      "step": 13500
    },
    {
      "epoch": 1.2,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.06889338731443,
      "eval_f1": 92.03778677462888,
      "eval_loss": 0.05917409434914589,
      "eval_runtime": 4.6082,
      "eval_samples_per_second": 217.003,
      "step": 13600
    },
    {
      "epoch": 1.22,
      "eval_accuracy": 93.8,
      "eval_average_metrics": 92.85300261096606,
      "eval_f1": 91.9060052219321,
      "eval_loss": 0.06280769407749176,
      "eval_runtime": 4.5282,
      "eval_samples_per_second": 220.836,
      "step": 13800
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.00017653232207425696,
      "loss": 0.0741,
      "step": 14000
    },
    {
      "epoch": 1.23,
      "eval_accuracy": 93.60000000000001,
      "eval_average_metrics": 92.65492227979274,
      "eval_f1": 91.70984455958549,
      "eval_loss": 0.06333824247121811,
      "eval_runtime": 4.4743,
      "eval_samples_per_second": 223.497,
      "step": 14000
    },
    {
      "epoch": 1.25,
      "eval_accuracy": 93.89999999999999,
      "eval_average_metrics": 92.94211563731932,
      "eval_f1": 91.98423127463865,
      "eval_loss": 0.06064913421869278,
      "eval_runtime": 4.6765,
      "eval_samples_per_second": 213.836,
      "step": 14200
    },
    {
      "epoch": 1.27,
      "eval_accuracy": 92.60000000000001,
      "eval_average_metrics": 91.62828282828283,
      "eval_f1": 90.65656565656566,
      "eval_loss": 0.07161322236061096,
      "eval_runtime": 4.5138,
      "eval_samples_per_second": 221.545,
      "step": 14400
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.00017212276214833758,
      "loss": 0.0715,
      "step": 14500
    },
    {
      "epoch": 1.29,
      "eval_accuracy": 93.7,
      "eval_average_metrics": 92.6998023715415,
      "eval_f1": 91.699604743083,
      "eval_loss": 0.06242042034864426,
      "eval_runtime": 4.764,
      "eval_samples_per_second": 209.909,
      "step": 14600
    },
    {
      "epoch": 1.31,
      "eval_accuracy": 93.7,
      "eval_average_metrics": 92.73235294117647,
      "eval_f1": 91.76470588235294,
      "eval_loss": 0.0626644566655159,
      "eval_runtime": 4.4732,
      "eval_samples_per_second": 223.555,
      "step": 14800
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0001677132022224182,
      "loss": 0.0714,
      "step": 15000
    },
    {
      "epoch": 1.32,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.54464751958224,
      "eval_f1": 92.68929503916449,
      "eval_loss": 0.05990656465291977,
      "eval_runtime": 4.5922,
      "eval_samples_per_second": 217.762,
      "step": 15000
    },
    {
      "epoch": 1.34,
      "eval_accuracy": 94.6,
      "eval_average_metrics": 93.73799472295514,
      "eval_f1": 92.87598944591029,
      "eval_loss": 0.060957495123147964,
      "eval_runtime": 4.4536,
      "eval_samples_per_second": 224.537,
      "step": 15200
    },
    {
      "epoch": 1.36,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.51578947368421,
      "eval_f1": 92.63157894736842,
      "eval_loss": 0.06167261675000191,
      "eval_runtime": 4.4865,
      "eval_samples_per_second": 222.89,
      "step": 15400
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0001633036422964988,
      "loss": 0.0707,
      "step": 15500
    },
    {
      "epoch": 1.38,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.51578947368421,
      "eval_f1": 92.63157894736842,
      "eval_loss": 0.061066027730703354,
      "eval_runtime": 4.5716,
      "eval_samples_per_second": 218.74,
      "step": 15600
    },
    {
      "epoch": 1.39,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.1633069828722,
      "eval_f1": 92.2266139657444,
      "eval_loss": 0.06235107034444809,
      "eval_runtime": 4.499,
      "eval_samples_per_second": 222.27,
      "step": 15800
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00015889408237057938,
      "loss": 0.0709,
      "step": 16000
    },
    {
      "epoch": 1.41,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.47659574468085,
      "eval_f1": 92.55319148936171,
      "eval_loss": 0.06194847822189331,
      "eval_runtime": 4.5231,
      "eval_samples_per_second": 221.086,
      "step": 16000
    },
    {
      "epoch": 1.43,
      "eval_accuracy": 94.6,
      "eval_average_metrics": 93.76596858638743,
      "eval_f1": 92.93193717277488,
      "eval_loss": 0.05966123938560486,
      "eval_runtime": 4.637,
      "eval_samples_per_second": 215.656,
      "step": 16200
    },
    {
      "epoch": 1.45,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.44664879356569,
      "eval_f1": 92.49329758713138,
      "eval_loss": 0.06104936823248863,
      "eval_runtime": 4.5374,
      "eval_samples_per_second": 220.391,
      "step": 16400
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.00015448452244466002,
      "loss": 0.0729,
      "step": 16500
    },
    {
      "epoch": 1.46,
      "eval_accuracy": 94.69999999999999,
      "eval_average_metrics": 93.87686762778506,
      "eval_f1": 93.05373525557013,
      "eval_loss": 0.06205834820866585,
      "eval_runtime": 4.4838,
      "eval_samples_per_second": 223.023,
      "step": 16600
    },
    {
      "epoch": 1.48,
      "eval_accuracy": 94.6,
      "eval_average_metrics": 93.78437500000001,
      "eval_f1": 92.96875000000001,
      "eval_loss": 0.06089754402637482,
      "eval_runtime": 4.5038,
      "eval_samples_per_second": 222.035,
      "step": 16800
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.00015007496251874061,
      "loss": 0.07,
      "step": 17000
    },
    {
      "epoch": 1.5,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.36288659793814,
      "eval_f1": 92.52577319587628,
      "eval_loss": 0.06112566590309143,
      "eval_runtime": 4.6026,
      "eval_samples_per_second": 217.269,
      "step": 17000
    },
    {
      "epoch": 1.52,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.48647214854111,
      "eval_f1": 92.57294429708222,
      "eval_loss": 0.06089947372674942,
      "eval_runtime": 4.5389,
      "eval_samples_per_second": 220.318,
      "step": 17200
    },
    {
      "epoch": 1.53,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.24354838709678,
      "eval_f1": 92.38709677419355,
      "eval_loss": 0.06110972911119461,
      "eval_runtime": 4.6859,
      "eval_samples_per_second": 213.405,
      "step": 17400
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.00014566540259282123,
      "loss": 0.0669,
      "step": 17500
    },
    {
      "epoch": 1.55,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.38513870541613,
      "eval_f1": 92.47027741083225,
      "eval_loss": 0.06174538657069206,
      "eval_runtime": 4.6674,
      "eval_samples_per_second": 214.254,
      "step": 17600
    },
    {
      "epoch": 1.57,
      "eval_accuracy": 94.6,
      "eval_average_metrics": 93.7095744680851,
      "eval_f1": 92.81914893617021,
      "eval_loss": 0.059681929647922516,
      "eval_runtime": 4.6196,
      "eval_samples_per_second": 216.471,
      "step": 17800
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.00014125584266690182,
      "loss": 0.07,
      "step": 18000
    },
    {
      "epoch": 1.59,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.29422572178477,
      "eval_f1": 92.38845144356955,
      "eval_loss": 0.061346184462308884,
      "eval_runtime": 4.4984,
      "eval_samples_per_second": 222.301,
      "step": 18000
    },
    {
      "epoch": 1.61,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.28421052631577,
      "eval_f1": 92.36842105263158,
      "eval_loss": 0.06077203154563904,
      "eval_runtime": 4.454,
      "eval_samples_per_second": 224.518,
      "step": 18200
    },
    {
      "epoch": 1.62,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.13233731739707,
      "eval_f1": 92.16467463479415,
      "eval_loss": 0.05959217995405197,
      "eval_runtime": 4.588,
      "eval_samples_per_second": 217.959,
      "step": 18400
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.00013684628274098244,
      "loss": 0.069,
      "step": 18500
    },
    {
      "epoch": 1.64,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.46666666666665,
      "eval_f1": 92.53333333333332,
      "eval_loss": 0.06017257645726204,
      "eval_runtime": 4.483,
      "eval_samples_per_second": 223.066,
      "step": 18600
    },
    {
      "epoch": 1.66,
      "eval_accuracy": 94.5,
      "eval_average_metrics": 93.58821571238349,
      "eval_f1": 92.67643142476697,
      "eval_loss": 0.058851905167102814,
      "eval_runtime": 4.5048,
      "eval_samples_per_second": 221.985,
      "step": 18800
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.00013243672281506306,
      "loss": 0.0713,
      "step": 19000
    },
    {
      "epoch": 1.68,
      "eval_accuracy": 93.89999999999999,
      "eval_average_metrics": 92.97346805736636,
      "eval_f1": 92.04693611473273,
      "eval_loss": 0.06167756766080856,
      "eval_runtime": 4.4859,
      "eval_samples_per_second": 222.921,
      "step": 19000
    },
    {
      "epoch": 1.69,
      "eval_accuracy": 93.7,
      "eval_average_metrics": 92.6998023715415,
      "eval_f1": 91.699604743083,
      "eval_loss": 0.06253690272569656,
      "eval_runtime": 4.582,
      "eval_samples_per_second": 218.244,
      "step": 19200
    },
    {
      "epoch": 1.71,
      "eval_accuracy": 93.89999999999999,
      "eval_average_metrics": 93.00433376455368,
      "eval_f1": 92.10866752910736,
      "eval_loss": 0.06255872547626495,
      "eval_runtime": 4.5188,
      "eval_samples_per_second": 221.296,
      "step": 19400
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.00012802716288914365,
      "loss": 0.0699,
      "step": 19500
    },
    {
      "epoch": 1.73,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.40492772667542,
      "eval_f1": 92.50985545335085,
      "eval_loss": 0.062451381236314774,
      "eval_runtime": 4.5919,
      "eval_samples_per_second": 217.773,
      "step": 19600
    },
    {
      "epoch": 1.75,
      "eval_accuracy": 94.0,
      "eval_average_metrics": 93.01063829787235,
      "eval_f1": 92.0212765957447,
      "eval_loss": 0.06319490820169449,
      "eval_runtime": 4.591,
      "eval_samples_per_second": 217.817,
      "step": 19800
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00012361760296322426,
      "loss": 0.0698,
      "step": 20000
    },
    {
      "epoch": 1.76,
      "eval_accuracy": 93.5,
      "eval_average_metrics": 92.51271186440678,
      "eval_f1": 91.52542372881356,
      "eval_loss": 0.06364640593528748,
      "eval_runtime": 4.5171,
      "eval_samples_per_second": 221.382,
      "step": 20000
    },
    {
      "epoch": 1.78,
      "eval_accuracy": 93.89999999999999,
      "eval_average_metrics": 92.99409857328145,
      "eval_f1": 92.0881971465629,
      "eval_loss": 0.06635148823261261,
      "eval_runtime": 4.6206,
      "eval_samples_per_second": 216.422,
      "step": 20200
    },
    {
      "epoch": 1.8,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.11141522029372,
      "eval_f1": 92.12283044058745,
      "eval_loss": 0.0606299452483654,
      "eval_runtime": 4.4605,
      "eval_samples_per_second": 224.19,
      "step": 20400
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00011920804303730487,
      "loss": 0.0703,
      "step": 20500
    },
    {
      "epoch": 1.82,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.32395833333332,
      "eval_f1": 92.44791666666666,
      "eval_loss": 0.060722097754478455,
      "eval_runtime": 4.5249,
      "eval_samples_per_second": 221.001,
      "step": 20600
    },
    {
      "epoch": 1.83,
      "eval_accuracy": 93.8,
      "eval_average_metrics": 92.76666666666665,
      "eval_f1": 91.73333333333332,
      "eval_loss": 0.05862819775938988,
      "eval_runtime": 4.5187,
      "eval_samples_per_second": 221.304,
      "step": 20800
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.00011479848311138547,
      "loss": 0.0698,
      "step": 21000
    },
    {
      "epoch": 1.85,
      "eval_accuracy": 93.8,
      "eval_average_metrics": 92.79947089947089,
      "eval_f1": 91.7989417989418,
      "eval_loss": 0.06128830835223198,
      "eval_runtime": 4.521,
      "eval_samples_per_second": 221.188,
      "step": 21000
    },
    {
      "epoch": 1.87,
      "eval_accuracy": 93.8,
      "eval_average_metrics": 92.87402597402597,
      "eval_f1": 91.94805194805194,
      "eval_loss": 0.06374780088663101,
      "eval_runtime": 4.4879,
      "eval_samples_per_second": 222.822,
      "step": 21200
    },
    {
      "epoch": 1.89,
      "eval_accuracy": 93.89999999999999,
      "eval_average_metrics": 92.88874833555259,
      "eval_f1": 91.87749667110519,
      "eval_loss": 0.06154455617070198,
      "eval_runtime": 4.6011,
      "eval_samples_per_second": 217.337,
      "step": 21400
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00011038892318546609,
      "loss": 0.0709,
      "step": 21500
    },
    {
      "epoch": 1.9,
      "eval_accuracy": 94.5,
      "eval_average_metrics": 93.57843791722297,
      "eval_f1": 92.65687583444593,
      "eval_loss": 0.060043178498744965,
      "eval_runtime": 4.46,
      "eval_samples_per_second": 224.215,
      "step": 21600
    },
    {
      "epoch": 1.92,
      "eval_accuracy": 93.89999999999999,
      "eval_average_metrics": 92.93155467720685,
      "eval_f1": 91.9631093544137,
      "eval_loss": 0.061132512986660004,
      "eval_runtime": 4.4987,
      "eval_samples_per_second": 222.287,
      "step": 21800
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.00010597936325954669,
      "loss": 0.0695,
      "step": 22000
    },
    {
      "epoch": 1.94,
      "eval_accuracy": 93.5,
      "eval_average_metrics": 92.56724581724582,
      "eval_f1": 91.63449163449164,
      "eval_loss": 0.06395059078931808,
      "eval_runtime": 4.6548,
      "eval_samples_per_second": 214.832,
      "step": 22000
    },
    {
      "epoch": 1.96,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.29422572178477,
      "eval_f1": 92.38845144356955,
      "eval_loss": 0.06141780689358711,
      "eval_runtime": 4.4836,
      "eval_samples_per_second": 223.034,
      "step": 22200
    },
    {
      "epoch": 1.98,
      "eval_accuracy": 94.5,
      "eval_average_metrics": 93.65522875816993,
      "eval_f1": 92.81045751633987,
      "eval_loss": 0.058759015053510666,
      "eval_runtime": 4.5162,
      "eval_samples_per_second": 221.426,
      "step": 22400
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0001015698033336273,
      "loss": 0.0715,
      "step": 22500
    },
    {
      "epoch": 1.99,
      "eval_accuracy": 93.89999999999999,
      "eval_average_metrics": 92.97346805736636,
      "eval_f1": 92.04693611473273,
      "eval_loss": 0.06228160858154297,
      "eval_runtime": 4.4726,
      "eval_samples_per_second": 223.582,
      "step": 22600
    },
    {
      "epoch": 2.01,
      "eval_accuracy": 94.6,
      "eval_average_metrics": 93.73799472295514,
      "eval_f1": 92.87598944591029,
      "eval_loss": 0.05991463363170624,
      "eval_runtime": 4.5003,
      "eval_samples_per_second": 222.206,
      "step": 22800
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.71602434077079e-05,
      "loss": 0.0682,
      "step": 23000
    },
    {
      "epoch": 2.03,
      "eval_accuracy": 94.0,
      "eval_average_metrics": 93.1038961038961,
      "eval_f1": 92.20779220779221,
      "eval_loss": 0.061682794243097305,
      "eval_runtime": 4.611,
      "eval_samples_per_second": 216.874,
      "step": 23000
    },
    {
      "epoch": 2.05,
      "eval_accuracy": 93.5,
      "eval_average_metrics": 92.55645161290323,
      "eval_f1": 91.61290322580645,
      "eval_loss": 0.06373216211795807,
      "eval_runtime": 4.6044,
      "eval_samples_per_second": 217.186,
      "step": 23200
    },
    {
      "epoch": 2.06,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.34492656875835,
      "eval_f1": 92.3898531375167,
      "eval_loss": 0.05869932472705841,
      "eval_runtime": 4.4706,
      "eval_samples_per_second": 223.684,
      "step": 23400
    },
    {
      "epoch": 2.07,
      "learning_rate": 9.27506834817885e-05,
      "loss": 0.0652,
      "step": 23500
    },
    {
      "epoch": 2.08,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.50606860158311,
      "eval_f1": 92.61213720316623,
      "eval_loss": 0.06166525185108185,
      "eval_runtime": 4.5725,
      "eval_samples_per_second": 218.699,
      "step": 23600
    },
    {
      "epoch": 2.1,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.33376623376623,
      "eval_f1": 92.46753246753246,
      "eval_loss": 0.06055561453104019,
      "eval_runtime": 4.5492,
      "eval_samples_per_second": 219.818,
      "step": 23800
    },
    {
      "epoch": 2.12,
      "learning_rate": 8.834112355586911e-05,
      "loss": 0.0691,
      "step": 24000
    },
    {
      "epoch": 2.12,
      "eval_accuracy": 93.7,
      "eval_average_metrics": 92.75377113133939,
      "eval_f1": 91.8075422626788,
      "eval_loss": 0.06339309364557266,
      "eval_runtime": 4.6365,
      "eval_samples_per_second": 215.678,
      "step": 24000
    },
    {
      "epoch": 2.13,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.1633069828722,
      "eval_f1": 92.2266139657444,
      "eval_loss": 0.06319531798362732,
      "eval_runtime": 4.5722,
      "eval_samples_per_second": 218.712,
      "step": 24200
    },
    {
      "epoch": 2.15,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.14271523178809,
      "eval_f1": 92.18543046357617,
      "eval_loss": 0.060979247093200684,
      "eval_runtime": 4.4363,
      "eval_samples_per_second": 225.412,
      "step": 24400
    },
    {
      "epoch": 2.16,
      "learning_rate": 8.393156362994973e-05,
      "loss": 0.0679,
      "step": 24500
    },
    {
      "epoch": 2.17,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.38513870541613,
      "eval_f1": 92.47027741083225,
      "eval_loss": 0.061841148883104324,
      "eval_runtime": 4.4945,
      "eval_samples_per_second": 222.493,
      "step": 24600
    },
    {
      "epoch": 2.19,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.51578947368421,
      "eval_f1": 92.63157894736842,
      "eval_loss": 0.06020021066069603,
      "eval_runtime": 4.6482,
      "eval_samples_per_second": 215.136,
      "step": 24800
    },
    {
      "epoch": 2.2,
      "learning_rate": 7.952200370403033e-05,
      "loss": 0.0678,
      "step": 25000
    },
    {
      "epoch": 2.2,
      "eval_accuracy": 94.69999999999999,
      "eval_average_metrics": 93.86773981603153,
      "eval_f1": 93.03547963206307,
      "eval_loss": 0.061626460403203964,
      "eval_runtime": 4.4847,
      "eval_samples_per_second": 222.982,
      "step": 25000
    },
    {
      "epoch": 2.22,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.35505992010653,
      "eval_f1": 92.41011984021304,
      "eval_loss": 0.05932234972715378,
      "eval_runtime": 4.444,
      "eval_samples_per_second": 225.02,
      "step": 25200
    },
    {
      "epoch": 2.24,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.35505992010653,
      "eval_f1": 92.41011984021304,
      "eval_loss": 0.05860959738492966,
      "eval_runtime": 4.4729,
      "eval_samples_per_second": 223.568,
      "step": 25400
    },
    {
      "epoch": 2.25,
      "learning_rate": 7.511244377811093e-05,
      "loss": 0.0687,
      "step": 25500
    },
    {
      "epoch": 2.26,
      "eval_accuracy": 94.6,
      "eval_average_metrics": 93.74736842105261,
      "eval_f1": 92.89473684210525,
      "eval_loss": 0.05995591729879379,
      "eval_runtime": 4.6311,
      "eval_samples_per_second": 215.933,
      "step": 25600
    },
    {
      "epoch": 2.28,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.51578947368421,
      "eval_f1": 92.63157894736842,
      "eval_loss": 0.06067919358611107,
      "eval_runtime": 4.4705,
      "eval_samples_per_second": 223.69,
      "step": 25800
    },
    {
      "epoch": 2.29,
      "learning_rate": 7.070288385219154e-05,
      "loss": 0.0665,
      "step": 26000
    },
    {
      "epoch": 2.29,
      "eval_accuracy": 94.6,
      "eval_average_metrics": 93.74736842105261,
      "eval_f1": 92.89473684210525,
      "eval_loss": 0.06090604141354561,
      "eval_runtime": 4.4777,
      "eval_samples_per_second": 223.33,
      "step": 26000
    },
    {
      "epoch": 2.31,
      "eval_accuracy": 94.5,
      "eval_average_metrics": 93.63633377135348,
      "eval_f1": 92.77266754270696,
      "eval_loss": 0.06175965070724487,
      "eval_runtime": 4.5456,
      "eval_samples_per_second": 219.993,
      "step": 26200
    },
    {
      "epoch": 2.33,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.1937908496732,
      "eval_f1": 92.2875816993464,
      "eval_loss": 0.062108419835567474,
      "eval_runtime": 4.5414,
      "eval_samples_per_second": 220.196,
      "step": 26400
    },
    {
      "epoch": 2.34,
      "learning_rate": 6.629332392627216e-05,
      "loss": 0.0681,
      "step": 26500
    },
    {
      "epoch": 2.35,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.48647214854111,
      "eval_f1": 92.57294429708222,
      "eval_loss": 0.060741446912288666,
      "eval_runtime": 4.4624,
      "eval_samples_per_second": 224.096,
      "step": 26600
    },
    {
      "epoch": 2.36,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.4962962962963,
      "eval_f1": 92.5925925925926,
      "eval_loss": 0.06029416620731354,
      "eval_runtime": 4.5256,
      "eval_samples_per_second": 220.966,
      "step": 26800
    },
    {
      "epoch": 2.38,
      "learning_rate": 6.188376400035276e-05,
      "loss": 0.0667,
      "step": 27000
    },
    {
      "epoch": 2.38,
      "eval_accuracy": 94.69999999999999,
      "eval_average_metrics": 93.84933949801848,
      "eval_f1": 92.99867899603699,
      "eval_loss": 0.059210509061813354,
      "eval_runtime": 4.8741,
      "eval_samples_per_second": 205.167,
      "step": 27000
    },
    {
      "epoch": 2.4,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.41474442988203,
      "eval_f1": 92.52948885976409,
      "eval_loss": 0.0605180561542511,
      "eval_runtime": 4.5293,
      "eval_samples_per_second": 220.783,
      "step": 27200
    },
    {
      "epoch": 2.42,
      "eval_accuracy": 94.5,
      "eval_average_metrics": 93.64580602883355,
      "eval_f1": 92.7916120576671,
      "eval_loss": 0.060811493545770645,
      "eval_runtime": 4.5424,
      "eval_samples_per_second": 220.147,
      "step": 27400
    },
    {
      "epoch": 2.43,
      "learning_rate": 5.747420407443336e-05,
      "loss": 0.0685,
      "step": 27500
    },
    {
      "epoch": 2.43,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.11141522029372,
      "eval_f1": 92.12283044058745,
      "eval_loss": 0.05978462100028992,
      "eval_runtime": 4.4831,
      "eval_samples_per_second": 223.06,
      "step": 27600
    },
    {
      "epoch": 2.45,
      "eval_accuracy": 93.8,
      "eval_average_metrics": 92.87402597402597,
      "eval_f1": 91.94805194805194,
      "eval_loss": 0.06267183274030685,
      "eval_runtime": 4.4576,
      "eval_samples_per_second": 224.334,
      "step": 27800
    },
    {
      "epoch": 2.47,
      "learning_rate": 5.3064644148513973e-05,
      "loss": 0.0672,
      "step": 28000
    },
    {
      "epoch": 2.47,
      "eval_accuracy": 94.0,
      "eval_average_metrics": 93.06299212598425,
      "eval_f1": 92.1259842519685,
      "eval_loss": 0.061355073004961014,
      "eval_runtime": 4.5194,
      "eval_samples_per_second": 221.27,
      "step": 28000
    },
    {
      "epoch": 2.49,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.29422572178477,
      "eval_f1": 92.38845144356955,
      "eval_loss": 0.06131287291646004,
      "eval_runtime": 4.5837,
      "eval_samples_per_second": 218.165,
      "step": 28200
    },
    {
      "epoch": 2.5,
      "eval_accuracy": 94.0,
      "eval_average_metrics": 93.04221635883906,
      "eval_f1": 92.0844327176781,
      "eval_loss": 0.06105473265051842,
      "eval_runtime": 4.5101,
      "eval_samples_per_second": 221.726,
      "step": 28400
    },
    {
      "epoch": 2.51,
      "learning_rate": 4.8655084222594584e-05,
      "loss": 0.0656,
      "step": 28500
    },
    {
      "epoch": 2.52,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.25384615384615,
      "eval_f1": 92.3076923076923,
      "eval_loss": 0.06093791127204895,
      "eval_runtime": 4.6588,
      "eval_samples_per_second": 214.647,
      "step": 28600
    },
    {
      "epoch": 2.54,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.40492772667542,
      "eval_f1": 92.50985545335085,
      "eval_loss": 0.061501096934080124,
      "eval_runtime": 4.5262,
      "eval_samples_per_second": 220.936,
      "step": 28800
    },
    {
      "epoch": 2.56,
      "learning_rate": 4.424552429667519e-05,
      "loss": 0.067,
      "step": 29000
    },
    {
      "epoch": 2.56,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.27414248021108,
      "eval_f1": 92.34828496042218,
      "eval_loss": 0.05971948057413101,
      "eval_runtime": 4.5243,
      "eval_samples_per_second": 221.027,
      "step": 29000
    },
    {
      "epoch": 2.58,
      "eval_accuracy": 93.60000000000001,
      "eval_average_metrics": 92.65492227979274,
      "eval_f1": 91.70984455958549,
      "eval_loss": 0.063376285135746,
      "eval_runtime": 4.6334,
      "eval_samples_per_second": 215.825,
      "step": 29200
    },
    {
      "epoch": 2.59,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.26402116402116,
      "eval_f1": 92.32804232804234,
      "eval_loss": 0.06081530451774597,
      "eval_runtime": 4.7045,
      "eval_samples_per_second": 212.561,
      "step": 29400
    },
    {
      "epoch": 2.6,
      "learning_rate": 3.98359643707558e-05,
      "loss": 0.0675,
      "step": 29500
    },
    {
      "epoch": 2.61,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.18368283093054,
      "eval_f1": 92.26736566186108,
      "eval_loss": 0.062262628227472305,
      "eval_runtime": 4.6273,
      "eval_samples_per_second": 216.108,
      "step": 29600
    },
    {
      "epoch": 2.63,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.3751655629139,
      "eval_f1": 92.45033112582782,
      "eval_loss": 0.06007382273674011,
      "eval_runtime": 4.5698,
      "eval_samples_per_second": 218.83,
      "step": 29800
    },
    {
      "epoch": 2.65,
      "learning_rate": 3.54264044448364e-05,
      "loss": 0.0682,
      "step": 30000
    },
    {
      "epoch": 2.65,
      "eval_accuracy": 94.1,
      "eval_average_metrics": 93.1633069828722,
      "eval_f1": 92.2266139657444,
      "eval_loss": 0.0607917495071888,
      "eval_runtime": 4.6423,
      "eval_samples_per_second": 215.411,
      "step": 30000
    },
    {
      "epoch": 2.66,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.5254593175853,
      "eval_f1": 92.6509186351706,
      "eval_loss": 0.06171978637576103,
      "eval_runtime": 4.4956,
      "eval_samples_per_second": 222.439,
      "step": 30200
    },
    {
      "epoch": 2.68,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.4566844919786,
      "eval_f1": 92.51336898395722,
      "eval_loss": 0.05954898148775101,
      "eval_runtime": 4.5069,
      "eval_samples_per_second": 221.881,
      "step": 30400
    },
    {
      "epoch": 2.69,
      "learning_rate": 3.1016844518917006e-05,
      "loss": 0.0684,
      "step": 30500
    },
    {
      "epoch": 2.7,
      "eval_accuracy": 94.5,
      "eval_average_metrics": 93.64580602883355,
      "eval_f1": 92.7916120576671,
      "eval_loss": 0.06073066592216492,
      "eval_runtime": 4.5568,
      "eval_samples_per_second": 219.452,
      "step": 30600
    },
    {
      "epoch": 2.72,
      "eval_accuracy": 94.5,
      "eval_average_metrics": 93.64580602883355,
      "eval_f1": 92.7916120576671,
      "eval_loss": 0.06212097778916359,
      "eval_runtime": 4.4991,
      "eval_samples_per_second": 222.265,
      "step": 30800
    },
    {
      "epoch": 2.73,
      "learning_rate": 2.6607284592997617e-05,
      "loss": 0.0644,
      "step": 31000
    },
    {
      "epoch": 2.73,
      "eval_accuracy": 94.5,
      "eval_average_metrics": 93.64580602883355,
      "eval_f1": 92.7916120576671,
      "eval_loss": 0.061464857310056686,
      "eval_runtime": 4.6313,
      "eval_samples_per_second": 215.924,
      "step": 31000
    },
    {
      "epoch": 2.75,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.28421052631577,
      "eval_f1": 92.36842105263158,
      "eval_loss": 0.06165764480829239,
      "eval_runtime": 4.4772,
      "eval_samples_per_second": 223.356,
      "step": 31200
    },
    {
      "epoch": 2.77,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.27414248021108,
      "eval_f1": 92.34828496042218,
      "eval_loss": 0.061222758144140244,
      "eval_runtime": 4.485,
      "eval_samples_per_second": 222.965,
      "step": 31400
    },
    {
      "epoch": 2.78,
      "learning_rate": 2.219772466707822e-05,
      "loss": 0.0656,
      "step": 31500
    },
    {
      "epoch": 2.79,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.40492772667542,
      "eval_f1": 92.50985545335085,
      "eval_loss": 0.06175553798675537,
      "eval_runtime": 4.4473,
      "eval_samples_per_second": 224.857,
      "step": 31600
    },
    {
      "epoch": 2.8,
      "eval_accuracy": 94.0,
      "eval_average_metrics": 93.04221635883906,
      "eval_f1": 92.0844327176781,
      "eval_loss": 0.06141304597258568,
      "eval_runtime": 4.5384,
      "eval_samples_per_second": 220.341,
      "step": 31800
    },
    {
      "epoch": 2.82,
      "learning_rate": 1.778816474115883e-05,
      "loss": 0.0682,
      "step": 32000
    },
    {
      "epoch": 2.82,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.39505928853755,
      "eval_f1": 92.49011857707511,
      "eval_loss": 0.06122256815433502,
      "eval_runtime": 4.532,
      "eval_samples_per_second": 220.652,
      "step": 32000
    },
    {
      "epoch": 2.84,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.5254593175853,
      "eval_f1": 92.6509186351706,
      "eval_loss": 0.06179660186171532,
      "eval_runtime": 4.5432,
      "eval_samples_per_second": 220.11,
      "step": 32200
    },
    {
      "epoch": 2.86,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.26402116402116,
      "eval_f1": 92.32804232804234,
      "eval_loss": 0.060935478657484055,
      "eval_runtime": 4.5308,
      "eval_samples_per_second": 220.712,
      "step": 32400
    },
    {
      "epoch": 2.87,
      "learning_rate": 1.3378604815239437e-05,
      "loss": 0.0628,
      "step": 32500
    },
    {
      "epoch": 2.88,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.51578947368421,
      "eval_f1": 92.63157894736842,
      "eval_loss": 0.06167520210146904,
      "eval_runtime": 4.5363,
      "eval_samples_per_second": 220.442,
      "step": 32600
    },
    {
      "epoch": 2.89,
      "eval_accuracy": 94.39999999999999,
      "eval_average_metrics": 93.51578947368421,
      "eval_f1": 92.63157894736842,
      "eval_loss": 0.061225228011608124,
      "eval_runtime": 4.5208,
      "eval_samples_per_second": 221.199,
      "step": 32800
    },
    {
      "epoch": 2.91,
      "learning_rate": 8.969044889320046e-06,
      "loss": 0.0659,
      "step": 33000
    },
    {
      "epoch": 2.91,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.3751655629139,
      "eval_f1": 92.45033112582782,
      "eval_loss": 0.06039771810173988,
      "eval_runtime": 4.5643,
      "eval_samples_per_second": 219.093,
      "step": 33000
    },
    {
      "epoch": 2.93,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.40492772667542,
      "eval_f1": 92.50985545335085,
      "eval_loss": 0.06096240133047104,
      "eval_runtime": 4.5827,
      "eval_samples_per_second": 218.214,
      "step": 33200
    },
    {
      "epoch": 2.95,
      "eval_accuracy": 94.3,
      "eval_average_metrics": 93.38513870541613,
      "eval_f1": 92.47027741083225,
      "eval_loss": 0.060673393309116364,
      "eval_runtime": 4.9126,
      "eval_samples_per_second": 203.559,
      "step": 33400
    },
    {
      "epoch": 2.95,
      "learning_rate": 4.559484963400652e-06,
      "loss": 0.0692,
      "step": 33500
    },
    {
      "epoch": 2.96,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.26402116402116,
      "eval_f1": 92.32804232804234,
      "eval_loss": 0.06072871759533882,
      "eval_runtime": 4.5081,
      "eval_samples_per_second": 221.824,
      "step": 33600
    },
    {
      "epoch": 2.98,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.26402116402116,
      "eval_f1": 92.32804232804234,
      "eval_loss": 0.06088118627667427,
      "eval_runtime": 4.513,
      "eval_samples_per_second": 221.581,
      "step": 33800
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.4992503748125936e-07,
      "loss": 0.0654,
      "step": 34000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 94.19999999999999,
      "eval_average_metrics": 93.26402116402116,
      "eval_f1": 92.32804232804234,
      "eval_loss": 0.060787323862314224,
      "eval_runtime": 4.526,
      "eval_samples_per_second": 220.947,
      "step": 34000
    },
    {
      "epoch": 3.0,
      "step": 34017,
      "total_flos": 1.0629344517601075e+17,
      "train_loss": 0.07169761398949699,
      "train_runtime": 13428.6442,
      "train_samples_per_second": 81.061,
      "train_steps_per_second": 2.533
    }
  ],
  "max_steps": 34017,
  "num_train_epochs": 3,
  "total_flos": 1.0629344517601075e+17,
  "trial_name": null,
  "trial_params": null
}