{
  "best_metric": 0.624123043712898,
  "best_model_checkpoint": "/tmp/test-ner1_/checkpoint-52109",
  "epoch": 120.0,
  "global_step": 58440,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.741183488427583,
      "eval_f1": 0.46964064436183395,
      "eval_loss": 0.8133957386016846,
      "eval_precision": 0.4221430162619737,
      "eval_recall": 0.5291817927953085,
      "eval_runtime": 2.7713,
      "eval_samples_per_second": 401.25,
      "eval_steps_per_second": 25.259,
      "step": 487
    },
    {
      "epoch": 1.03,
      "learning_rate": 2.9743326488706367e-05,
      "loss": 1.0691,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7599856836077309,
      "eval_f1": 0.5087176950661556,
      "eval_loss": 0.7439278960227966,
      "eval_precision": 0.4565024411895251,
      "eval_recall": 0.5744205529181793,
      "eval_runtime": 2.7397,
      "eval_samples_per_second": 405.878,
      "eval_steps_per_second": 25.55,
      "step": 974
    },
    {
      "epoch": 2.05,
      "learning_rate": 2.9486652977412733e-05,
      "loss": 0.6796,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7499642090193271,
      "eval_f1": 0.5268402042595591,
      "eval_loss": 0.8020838499069214,
      "eval_precision": 0.4754946043165468,
      "eval_recall": 0.5906171460485897,
      "eval_runtime": 2.7638,
      "eval_samples_per_second": 402.346,
      "eval_steps_per_second": 25.328,
      "step": 1461
    },
    {
      "epoch": 3.08,
      "learning_rate": 2.9229979466119096e-05,
      "loss": 0.5266,
      "step": 1500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7569315199236459,
      "eval_f1": 0.5452078450721599,
      "eval_loss": 0.826590895652771,
      "eval_precision": 0.4882898806893504,
      "eval_recall": 0.6171460485897794,
      "eval_runtime": 2.7581,
      "eval_samples_per_second": 403.178,
      "eval_steps_per_second": 25.38,
      "step": 1948
    },
    {
      "epoch": 4.11,
      "learning_rate": 2.8973305954825462e-05,
      "loss": 0.4087,
      "step": 2000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7584347411119065,
      "eval_f1": 0.5577331169641744,
      "eval_loss": 0.8819655776023865,
      "eval_precision": 0.5042889390519187,
      "eval_recall": 0.623848087126501,
      "eval_runtime": 2.7608,
      "eval_samples_per_second": 402.775,
      "eval_steps_per_second": 25.355,
      "step": 2435
    },
    {
      "epoch": 5.13,
      "learning_rate": 2.8716632443531828e-05,
      "loss": 0.314,
      "step": 2500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7587687902648532,
      "eval_f1": 0.5619107479572597,
      "eval_loss": 0.8884239196777344,
      "eval_precision": 0.5109739368998628,
      "eval_recall": 0.6241273387321977,
      "eval_runtime": 2.7416,
      "eval_samples_per_second": 405.6,
      "eval_steps_per_second": 25.532,
      "step": 2922
    },
    {
      "epoch": 6.16,
      "learning_rate": 2.8459958932238194e-05,
      "loss": 0.254,
      "step": 3000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7592937246480553,
      "eval_f1": 0.5628216392619556,
      "eval_loss": 0.9709771275520325,
      "eval_precision": 0.511171910624715,
      "eval_recall": 0.6260820999720749,
      "eval_runtime": 2.7411,
      "eval_samples_per_second": 405.673,
      "eval_steps_per_second": 25.537,
      "step": 3409
    },
    {
      "epoch": 7.19,
      "learning_rate": 2.8203285420944557e-05,
      "loss": 0.2096,
      "step": 3500
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7622285850632308,
      "eval_f1": 0.5648183075568968,
      "eval_loss": 1.0743454694747925,
      "eval_precision": 0.5137236962488564,
      "eval_recall": 0.6271991063948618,
      "eval_runtime": 2.7561,
      "eval_samples_per_second": 403.467,
      "eval_steps_per_second": 25.398,
      "step": 3896
    },
    {
      "epoch": 8.21,
      "learning_rate": 2.7946611909650923e-05,
      "loss": 0.1786,
      "step": 4000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.7570985445001193,
      "eval_f1": 0.5668016194331984,
      "eval_loss": 1.1285585165023804,
      "eval_precision": 0.5181586860976174,
      "eval_recall": 0.6255235967606814,
      "eval_runtime": 2.7449,
      "eval_samples_per_second": 405.112,
      "eval_steps_per_second": 25.502,
      "step": 4383
    },
    {
      "epoch": 9.24,
      "learning_rate": 2.7689938398357292e-05,
      "loss": 0.1486,
      "step": 4500
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.7544738725841088,
      "eval_f1": 0.5723700887198986,
      "eval_loss": 1.1630432605743408,
      "eval_precision": 0.5240194940821536,
      "eval_recall": 0.6305501256632225,
      "eval_runtime": 2.8838,
      "eval_samples_per_second": 385.605,
      "eval_steps_per_second": 24.274,
      "step": 4870
    },
    {
      "epoch": 10.27,
      "learning_rate": 2.743326488706366e-05,
      "loss": 0.132,
      "step": 5000
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.7605821999522787,
      "eval_f1": 0.5760409993593851,
      "eval_loss": 1.1933799982070923,
      "eval_precision": 0.5321969696969697,
      "eval_recall": 0.6277576096062553,
      "eval_runtime": 2.7538,
      "eval_samples_per_second": 403.799,
      "eval_steps_per_second": 25.419,
      "step": 5357
    },
    {
      "epoch": 11.29,
      "learning_rate": 2.717659137577002e-05,
      "loss": 0.1098,
      "step": 5500
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.760200429491768,
      "eval_f1": 0.5755844155844156,
      "eval_loss": 1.1861658096313477,
      "eval_precision": 0.5379946588977907,
      "eval_recall": 0.6188215582239598,
      "eval_runtime": 2.7422,
      "eval_samples_per_second": 405.52,
      "eval_steps_per_second": 25.527,
      "step": 5844
    },
    {
      "epoch": 12.32,
      "learning_rate": 2.6919917864476387e-05,
      "loss": 0.094,
      "step": 6000
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.7506084466714388,
      "eval_f1": 0.5764092123679858,
      "eval_loss": 1.3724055290222168,
      "eval_precision": 0.52945301542777,
      "eval_recall": 0.6325048869030997,
      "eval_runtime": 2.7524,
      "eval_samples_per_second": 404.018,
      "eval_steps_per_second": 25.433,
      "step": 6331
    },
    {
      "epoch": 13.35,
      "learning_rate": 2.6663244353182754e-05,
      "loss": 0.084,
      "step": 6500
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.7531615366261035,
      "eval_f1": 0.574173712528824,
      "eval_loss": 1.374582290649414,
      "eval_precision": 0.530414201183432,
      "eval_recall": 0.6258028483663781,
      "eval_runtime": 2.7617,
      "eval_samples_per_second": 402.648,
      "eval_steps_per_second": 25.347,
      "step": 6818
    },
    {
      "epoch": 14.37,
      "learning_rate": 2.640657084188912e-05,
      "loss": 0.0758,
      "step": 7000
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.7581245526127416,
      "eval_f1": 0.5684922922672014,
      "eval_loss": 1.3000197410583496,
      "eval_precision": 0.5156889495225102,
      "eval_recall": 0.6333426417201898,
      "eval_runtime": 2.7658,
      "eval_samples_per_second": 402.06,
      "eval_steps_per_second": 25.31,
      "step": 7305
    },
    {
      "epoch": 15.4,
      "learning_rate": 2.6149897330595482e-05,
      "loss": 0.0694,
      "step": 7500
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.7593175853018372,
      "eval_f1": 0.5867220995192932,
      "eval_loss": 1.4194592237472534,
      "eval_precision": 0.5485908649173955,
      "eval_recall": 0.6305501256632225,
      "eval_runtime": 2.7493,
      "eval_samples_per_second": 404.466,
      "eval_steps_per_second": 25.461,
      "step": 7792
    },
    {
      "epoch": 16.43,
      "learning_rate": 2.589322381930185e-05,
      "loss": 0.062,
      "step": 8000
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.7465521355285135,
      "eval_f1": 0.5717906475731847,
      "eval_loss": 1.4974385499954224,
      "eval_precision": 0.5234338747099768,
      "eval_recall": 0.629991622451829,
      "eval_runtime": 2.7551,
      "eval_samples_per_second": 403.608,
      "eval_steps_per_second": 25.407,
      "step": 8279
    },
    {
      "epoch": 17.45,
      "learning_rate": 2.5636550308008215e-05,
      "loss": 0.0543,
      "step": 8500
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.7567644953471725,
      "eval_f1": 0.5741626794258373,
      "eval_loss": 1.5013597011566162,
      "eval_precision": 0.5346820809248555,
      "eval_recall": 0.6199385646467467,
      "eval_runtime": 2.7536,
      "eval_samples_per_second": 403.832,
      "eval_steps_per_second": 25.421,
      "step": 8766
    },
    {
      "epoch": 18.48,
      "learning_rate": 2.537987679671458e-05,
      "loss": 0.0471,
      "step": 9000
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.7546408971605822,
      "eval_f1": 0.5768982020437201,
      "eval_loss": 1.5164920091629028,
      "eval_precision": 0.5373493975903615,
      "eval_recall": 0.6227310807037141,
      "eval_runtime": 2.7556,
      "eval_samples_per_second": 403.547,
      "eval_steps_per_second": 25.403,
      "step": 9253
    },
    {
      "epoch": 19.51,
      "learning_rate": 2.5123203285420943e-05,
      "loss": 0.0449,
      "step": 9500
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.7568122166547363,
      "eval_f1": 0.5733962504782553,
      "eval_loss": 1.5718824863433838,
      "eval_precision": 0.5276995305164319,
      "eval_recall": 0.6277576096062553,
      "eval_runtime": 2.7667,
      "eval_samples_per_second": 401.921,
      "eval_steps_per_second": 25.301,
      "step": 9740
    },
    {
      "epoch": 20.53,
      "learning_rate": 2.486652977412731e-05,
      "loss": 0.0451,
      "step": 10000
    },
    {
      "epoch": 21.0,
      "eval_accuracy": 0.7607253638749701,
      "eval_f1": 0.5917858548746885,
      "eval_loss": 1.5306912660598755,
      "eval_precision": 0.5581683168316832,
      "eval_recall": 0.6297123708461324,
      "eval_runtime": 2.7618,
      "eval_samples_per_second": 402.633,
      "eval_steps_per_second": 25.346,
      "step": 10227
    },
    {
      "epoch": 21.56,
      "learning_rate": 2.4609856262833676e-05,
      "loss": 0.039,
      "step": 10500
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.7572417084228108,
      "eval_f1": 0.5844206174912802,
      "eval_loss": 1.5783377885818481,
      "eval_precision": 0.54375,
      "eval_recall": 0.6316671320860094,
      "eval_runtime": 2.7632,
      "eval_samples_per_second": 402.427,
      "eval_steps_per_second": 25.333,
      "step": 10714
    },
    {
      "epoch": 22.59,
      "learning_rate": 2.4353182751540042e-05,
      "loss": 0.0363,
      "step": 11000
    },
    {
      "epoch": 23.0,
      "eval_accuracy": 0.7541875447387258,
      "eval_f1": 0.5802802416763081,
      "eval_loss": 1.6342318058013916,
      "eval_precision": 0.5376369699857074,
      "eval_recall": 0.6302708740575258,
      "eval_runtime": 2.7447,
      "eval_samples_per_second": 405.151,
      "eval_steps_per_second": 25.504,
      "step": 11201
    },
    {
      "epoch": 23.61,
      "learning_rate": 2.4096509240246405e-05,
      "loss": 0.0326,
      "step": 11500
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.7596993557623479,
      "eval_f1": 0.5911304118963022,
      "eval_loss": 1.641722321510315,
      "eval_precision": 0.5589845694375312,
      "eval_recall": 0.6271991063948618,
      "eval_runtime": 2.7546,
      "eval_samples_per_second": 403.689,
      "eval_steps_per_second": 25.412,
      "step": 11688
    },
    {
      "epoch": 24.64,
      "learning_rate": 2.383983572895277e-05,
      "loss": 0.0296,
      "step": 12000
    },
    {
      "epoch": 25.0,
      "eval_accuracy": 0.7586733476497256,
      "eval_f1": 0.5861406430126809,
      "eval_loss": 1.6684845685958862,
      "eval_precision": 0.5414103170847137,
      "eval_recall": 0.6389276738341245,
      "eval_runtime": 2.8726,
      "eval_samples_per_second": 387.108,
      "eval_steps_per_second": 24.368,
      "step": 12175
    },
    {
      "epoch": 25.67,
      "learning_rate": 2.3583162217659137e-05,
      "loss": 0.0283,
      "step": 12500
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.760200429491768,
      "eval_f1": 0.5926797385620914,
      "eval_loss": 1.7346807718276978,
      "eval_precision": 0.5571393462767265,
      "eval_recall": 0.6330633901144932,
      "eval_runtime": 2.7525,
      "eval_samples_per_second": 404.0,
      "eval_steps_per_second": 25.432,
      "step": 12662
    },
    {
      "epoch": 26.69,
      "learning_rate": 2.3326488706365506e-05,
      "loss": 0.0277,
      "step": 13000
    },
    {
      "epoch": 27.0,
      "eval_accuracy": 0.7632068718682892,
      "eval_f1": 0.6025674613570867,
      "eval_loss": 1.6559849977493286,
      "eval_precision": 0.5674808783617074,
      "eval_recall": 0.6422786931024853,
      "eval_runtime": 2.7331,
      "eval_samples_per_second": 406.857,
      "eval_steps_per_second": 25.612,
      "step": 13149
    },
    {
      "epoch": 27.72,
      "learning_rate": 2.306981519507187e-05,
      "loss": 0.025,
      "step": 13500
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.7613934621808638,
      "eval_f1": 0.6024861147844485,
      "eval_loss": 1.7496830224990845,
      "eval_precision": 0.5722180356694297,
      "eval_recall": 0.6361351577771572,
      "eval_runtime": 2.7227,
      "eval_samples_per_second": 408.425,
      "eval_steps_per_second": 25.71,
      "step": 13636
    },
    {
      "epoch": 28.75,
      "learning_rate": 2.2813141683778235e-05,
      "loss": 0.0241,
      "step": 14000
    },
    {
      "epoch": 29.0,
      "eval_accuracy": 0.7637795275590551,
      "eval_f1": 0.5988181221273802,
      "eval_loss": 1.710971474647522,
      "eval_precision": 0.5651958353991076,
      "eval_recall": 0.6366936609885507,
      "eval_runtime": 2.7599,
      "eval_samples_per_second": 402.917,
      "eval_steps_per_second": 25.363,
      "step": 14123
    },
    {
      "epoch": 29.77,
      "learning_rate": 2.25564681724846e-05,
      "loss": 0.0242,
      "step": 14500
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.7647339537103317,
      "eval_f1": 0.5951438374241224,
      "eval_loss": 1.794677734375,
      "eval_precision": 0.5641731298473855,
      "eval_recall": 0.6297123708461324,
      "eval_runtime": 2.7628,
      "eval_samples_per_second": 402.492,
      "eval_steps_per_second": 25.337,
      "step": 14610
    },
    {
      "epoch": 30.8,
      "learning_rate": 2.2299794661190967e-05,
      "loss": 0.0219,
      "step": 15000
    },
    {
      "epoch": 31.0,
      "eval_accuracy": 0.7565258888093533,
      "eval_f1": 0.592573084013695,
      "eval_loss": 1.8282643556594849,
      "eval_precision": 0.5606777971592325,
      "eval_recall": 0.6283161128176487,
      "eval_runtime": 2.7362,
      "eval_samples_per_second": 406.407,
      "eval_steps_per_second": 25.583,
      "step": 15097
    },
    {
      "epoch": 31.83,
      "learning_rate": 2.204312114989733e-05,
      "loss": 0.0193,
      "step": 15500
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.7648293963254593,
      "eval_f1": 0.5969198088157196,
      "eval_loss": 1.8161470890045166,
      "eval_precision": 0.568969881042774,
      "eval_recall": 0.6277576096062553,
      "eval_runtime": 2.7545,
      "eval_samples_per_second": 403.709,
      "eval_steps_per_second": 25.413,
      "step": 15584
    },
    {
      "epoch": 32.85,
      "learning_rate": 2.1786447638603696e-05,
      "loss": 0.0185,
      "step": 16000
    },
    {
      "epoch": 33.0,
      "eval_accuracy": 0.7608685277976617,
      "eval_f1": 0.5930080876597965,
      "eval_loss": 1.846176028251648,
      "eval_precision": 0.5564259485924112,
      "eval_recall": 0.6347388997486736,
      "eval_runtime": 2.7276,
      "eval_samples_per_second": 407.69,
      "eval_steps_per_second": 25.664,
      "step": 16071
    },
    {
      "epoch": 33.88,
      "learning_rate": 2.1529774127310062e-05,
      "loss": 0.0195,
      "step": 16500
    },
    {
      "epoch": 34.0,
      "eval_accuracy": 0.755810069195896,
      "eval_f1": 0.5868997912317329,
      "eval_loss": 1.9018374681472778,
      "eval_precision": 0.5508204751408278,
      "eval_recall": 0.628036861211952,
      "eval_runtime": 2.7537,
      "eval_samples_per_second": 403.819,
      "eval_steps_per_second": 25.42,
      "step": 16558
    },
    {
      "epoch": 34.91,
      "learning_rate": 2.127310061601643e-05,
      "loss": 0.0181,
      "step": 17000
    },
    {
      "epoch": 35.0,
      "eval_accuracy": 0.7597470770699117,
      "eval_f1": 0.5975321606720924,
      "eval_loss": 1.8523436784744263,
      "eval_precision": 0.5637849888531088,
      "eval_recall": 0.6355766545657637,
      "eval_runtime": 2.7278,
      "eval_samples_per_second": 407.651,
      "eval_steps_per_second": 25.661,
      "step": 17045
    },
    {
      "epoch": 35.93,
      "learning_rate": 2.1016427104722795e-05,
      "loss": 0.0182,
      "step": 17500
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.7611309949892627,
      "eval_f1": 0.6036228023441662,
      "eval_loss": 1.8343547582626343,
      "eval_precision": 0.5770308123249299,
      "eval_recall": 0.6327841385087964,
      "eval_runtime": 2.7612,
      "eval_samples_per_second": 402.717,
      "eval_steps_per_second": 25.351,
      "step": 17532
    },
    {
      "epoch": 36.96,
      "learning_rate": 2.0759753593429157e-05,
      "loss": 0.0153,
      "step": 18000
    },
    {
      "epoch": 37.0,
      "eval_accuracy": 0.7668814125507039,
      "eval_f1": 0.603166156711454,
      "eval_loss": 1.8464767932891846,
      "eval_precision": 0.5759654471544715,
      "eval_recall": 0.6330633901144932,
      "eval_runtime": 2.7392,
      "eval_samples_per_second": 405.958,
      "eval_steps_per_second": 25.555,
      "step": 18019
    },
    {
      "epoch": 37.99,
      "learning_rate": 2.0503080082135523e-05,
      "loss": 0.0142,
      "step": 18500
    },
    {
      "epoch": 38.0,
      "eval_accuracy": 0.7632068718682892,
      "eval_f1": 0.5945442448436461,
      "eval_loss": 1.8911150693893433,
      "eval_precision": 0.5678698525673614,
      "eval_recall": 0.623848087126501,
      "eval_runtime": 2.7453,
      "eval_samples_per_second": 405.06,
      "eval_steps_per_second": 25.498,
      "step": 18506
    },
    {
      "epoch": 39.0,
      "eval_accuracy": 0.7622763063707946,
      "eval_f1": 0.6007257089100927,
      "eval_loss": 1.8849174976348877,
      "eval_precision": 0.5790155440414507,
      "eval_recall": 0.6241273387321977,
      "eval_runtime": 2.7696,
      "eval_samples_per_second": 401.495,
      "eval_steps_per_second": 25.274,
      "step": 18993
    },
    {
      "epoch": 39.01,
      "learning_rate": 2.024640657084189e-05,
      "loss": 0.0151,
      "step": 19000
    },
    {
      "epoch": 40.0,
      "eval_accuracy": 0.7665235027439752,
      "eval_f1": 0.5976520811099254,
      "eval_loss": 1.8399417400360107,
      "eval_precision": 0.5721583652618135,
      "eval_recall": 0.6255235967606814,
      "eval_runtime": 2.8345,
      "eval_samples_per_second": 392.315,
      "eval_steps_per_second": 24.696,
      "step": 19480
    },
    {
      "epoch": 40.04,
      "learning_rate": 1.9989733059548256e-05,
      "loss": 0.0148,
      "step": 19500
    },
    {
      "epoch": 41.0,
      "eval_accuracy": 0.7649248389405869,
      "eval_f1": 0.596647742633144,
      "eval_loss": 1.843032956123352,
      "eval_precision": 0.5782027770500393,
      "eval_recall": 0.6163082937726891,
      "eval_runtime": 2.7539,
      "eval_samples_per_second": 403.796,
      "eval_steps_per_second": 25.419,
      "step": 19967
    },
    {
      "epoch": 41.07,
      "learning_rate": 1.973305954825462e-05,
      "loss": 0.0138,
      "step": 20000
    },
    {
      "epoch": 42.0,
      "eval_accuracy": 0.7691481746599856,
      "eval_f1": 0.5887899423782085,
      "eval_loss": 1.8764090538024902,
      "eval_precision": 0.5543773119605425,
      "eval_recall": 0.6277576096062553,
      "eval_runtime": 2.7704,
      "eval_samples_per_second": 401.388,
      "eval_steps_per_second": 25.267,
      "step": 20454
    },
    {
      "epoch": 42.09,
      "learning_rate": 1.9476386036960984e-05,
      "loss": 0.0147,
      "step": 20500
    },
    {
      "epoch": 43.0,
      "eval_accuracy": 0.7666428060128847,
      "eval_f1": 0.6014559894109861,
      "eval_loss": 1.9270243644714355,
      "eval_precision": 0.571716155007549,
      "eval_recall": 0.6344596481429768,
      "eval_runtime": 2.7281,
      "eval_samples_per_second": 407.612,
      "eval_steps_per_second": 25.659,
      "step": 20941
    },
    {
      "epoch": 43.12,
      "learning_rate": 1.921971252566735e-05,
      "loss": 0.0148,
      "step": 21000
    },
    {
      "epoch": 44.0,
      "eval_accuracy": 0.7710570269625387,
      "eval_f1": 0.5908850026497084,
      "eval_loss": 1.8888484239578247,
      "eval_precision": 0.5621376354928157,
      "eval_recall": 0.6227310807037141,
      "eval_runtime": 2.775,
      "eval_samples_per_second": 400.718,
      "eval_steps_per_second": 25.225,
      "step": 21428
    },
    {
      "epoch": 44.15,
      "learning_rate": 1.896303901437372e-05,
      "loss": 0.0123,
      "step": 21500
    },
    {
      "epoch": 45.0,
      "eval_accuracy": 0.7653066094010976,
      "eval_f1": 0.5868878357030016,
      "eval_loss": 1.8992524147033691,
      "eval_precision": 0.5551681195516812,
      "eval_recall": 0.6224518290980173,
      "eval_runtime": 2.7417,
      "eval_samples_per_second": 405.586,
      "eval_steps_per_second": 25.531,
      "step": 21915
    },
    {
      "epoch": 45.17,
      "learning_rate": 1.8706365503080083e-05,
      "loss": 0.0115,
      "step": 22000
    },
    {
      "epoch": 46.0,
      "eval_accuracy": 0.7644714865187306,
      "eval_f1": 0.597897503285151,
      "eval_loss": 1.947526216506958,
      "eval_precision": 0.5646562422437329,
      "eval_recall": 0.635297402960067,
      "eval_runtime": 2.7396,
      "eval_samples_per_second": 405.892,
      "eval_steps_per_second": 25.551,
      "step": 22402
    },
    {
      "epoch": 46.2,
      "learning_rate": 1.844969199178645e-05,
      "loss": 0.0107,
      "step": 22500
    },
    {
      "epoch": 47.0,
      "eval_accuracy": 0.7674302075876879,
      "eval_f1": 0.605424089337942,
      "eval_loss": 1.994935393333435,
      "eval_precision": 0.5777721390510023,
      "eval_recall": 0.6358559061714605,
      "eval_runtime": 2.7286,
      "eval_samples_per_second": 407.535,
      "eval_steps_per_second": 25.654,
      "step": 22889
    },
    {
      "epoch": 47.23,
      "learning_rate": 1.8193018480492815e-05,
      "loss": 0.0098,
      "step": 23000
    },
    {
      "epoch": 48.0,
      "eval_accuracy": 0.7680744452397996,
      "eval_f1": 0.5976063829787234,
      "eval_loss": 1.9606735706329346,
      "eval_precision": 0.5704493526275705,
      "eval_recall": 0.6274783580005585,
      "eval_runtime": 2.772,
      "eval_samples_per_second": 401.157,
      "eval_steps_per_second": 25.253,
      "step": 23376
    },
    {
      "epoch": 48.25,
      "learning_rate": 1.793634496919918e-05,
      "loss": 0.012,
      "step": 23500
    },
    {
      "epoch": 49.0,
      "eval_accuracy": 0.7675733715103794,
      "eval_f1": 0.6134034165571617,
      "eval_loss": 1.918538212776184,
      "eval_precision": 0.5793000744601638,
      "eval_recall": 0.6517732476961743,
      "eval_runtime": 2.7319,
      "eval_samples_per_second": 407.049,
      "eval_steps_per_second": 25.624,
      "step": 23863
    },
    {
      "epoch": 49.28,
      "learning_rate": 1.7679671457905544e-05,
      "loss": 0.0117,
      "step": 24000
    },
    {
      "epoch": 50.0,
      "eval_accuracy": 0.7698401336196612,
      "eval_f1": 0.6049822064056939,
      "eval_loss": 1.9814343452453613,
      "eval_precision": 0.572890664003994,
      "eval_recall": 0.6408824350740017,
      "eval_runtime": 2.761,
      "eval_samples_per_second": 402.752,
      "eval_steps_per_second": 25.353,
      "step": 24350
    },
    {
      "epoch": 50.31,
      "learning_rate": 1.742299794661191e-05,
      "loss": 0.0093,
      "step": 24500
    },
    {
      "epoch": 51.0,
      "eval_accuracy": 0.7662371748985922,
      "eval_f1": 0.6067415730337078,
      "eval_loss": 2.035374164581299,
      "eval_precision": 0.5760542168674698,
      "eval_recall": 0.6408824350740017,
      "eval_runtime": 2.7628,
      "eval_samples_per_second": 402.494,
      "eval_steps_per_second": 25.337,
      "step": 24837
    },
    {
      "epoch": 51.33,
      "learning_rate": 1.7166324435318276e-05,
      "loss": 0.0082,
      "step": 25000
    },
    {
      "epoch": 52.0,
      "eval_accuracy": 0.7683130517776187,
      "eval_f1": 0.6179188429087987,
      "eval_loss": 1.9876421689987183,
      "eval_precision": 0.5936695831188883,
      "eval_recall": 0.6442334543423625,
      "eval_runtime": 2.751,
      "eval_samples_per_second": 404.224,
      "eval_steps_per_second": 25.446,
      "step": 25324
    },
    {
      "epoch": 52.36,
      "learning_rate": 1.6909650924024642e-05,
      "loss": 0.0077,
      "step": 25500
    },
    {
      "epoch": 53.0,
      "eval_accuracy": 0.7691481746599856,
      "eval_f1": 0.6208498428747096,
      "eval_loss": 2.0615577697753906,
      "eval_precision": 0.6078116639914393,
      "eval_recall": 0.6344596481429768,
      "eval_runtime": 2.7544,
      "eval_samples_per_second": 403.721,
      "eval_steps_per_second": 25.414,
      "step": 25811
    },
    {
      "epoch": 53.39,
      "learning_rate": 1.6652977412731005e-05,
      "loss": 0.0087,
      "step": 26000
    },
    {
      "epoch": 54.0,
      "eval_accuracy": 0.7652588880935338,
      "eval_f1": 0.5977975878342947,
      "eval_loss": 1.9790315628051758,
      "eval_precision": 0.5633802816901409,
      "eval_recall": 0.6366936609885507,
      "eval_runtime": 2.7387,
      "eval_samples_per_second": 406.036,
      "eval_steps_per_second": 25.56,
      "step": 26298
    },
    {
      "epoch": 54.41,
      "learning_rate": 1.639630390143737e-05,
      "loss": 0.0102,
      "step": 26500
    },
    {
      "epoch": 55.0,
      "eval_accuracy": 0.7677642567406346,
      "eval_f1": 0.6056356660933986,
      "eval_loss": 2.068793773651123,
      "eval_precision": 0.5754147812971342,
      "eval_recall": 0.6392069254398213,
      "eval_runtime": 2.8605,
      "eval_samples_per_second": 388.746,
      "eval_steps_per_second": 24.471,
      "step": 26785
    },
    {
      "epoch": 55.44,
      "learning_rate": 1.6139630390143737e-05,
      "loss": 0.0073,
      "step": 27000
    },
    {
      "epoch": 56.0,
      "eval_accuracy": 0.7678596993557624,
      "eval_f1": 0.6073495759860008,
      "eval_loss": 1.960081696510315,
      "eval_precision": 0.5862785862785863,
      "eval_recall": 0.629991622451829,
      "eval_runtime": 2.7429,
      "eval_samples_per_second": 405.41,
      "eval_steps_per_second": 25.52,
      "step": 27272
    },
    {
      "epoch": 56.47,
      "learning_rate": 1.5882956878850103e-05,
      "loss": 0.0087,
      "step": 27500
    },
    {
      "epoch": 57.0,
      "eval_accuracy": 0.7683369124314006,
      "eval_f1": 0.6085343228200372,
      "eval_loss": 2.0414817333221436,
      "eval_precision": 0.5790668348045397,
      "eval_recall": 0.6411616866796984,
      "eval_runtime": 2.7325,
      "eval_samples_per_second": 406.95,
      "eval_steps_per_second": 25.617,
      "step": 27759
    },
    {
      "epoch": 57.49,
      "learning_rate": 1.5626283367556466e-05,
      "loss": 0.0082,
      "step": 28000
    },
    {
      "epoch": 58.0,
      "eval_accuracy": 0.766571224051539,
      "eval_f1": 0.6019978969505784,
      "eval_loss": 2.077375888824463,
      "eval_precision": 0.5686615346411721,
      "eval_recall": 0.639486177045518,
      "eval_runtime": 2.7299,
      "eval_samples_per_second": 407.338,
      "eval_steps_per_second": 25.642,
      "step": 28246
    },
    {
      "epoch": 58.52,
      "learning_rate": 1.5369609856262832e-05,
      "loss": 0.0056,
      "step": 28500
    },
    {
      "epoch": 59.0,
      "eval_accuracy": 0.7637318062514913,
      "eval_f1": 0.6061579651941098,
      "eval_loss": 2.077326536178589,
      "eval_precision": 0.5821547955772692,
      "eval_recall": 0.6322256352974029,
      "eval_runtime": 2.7596,
      "eval_samples_per_second": 402.96,
      "eval_steps_per_second": 25.366,
      "step": 28733
    },
    {
      "epoch": 59.55,
      "learning_rate": 1.5112936344969198e-05,
      "loss": 0.0076,
      "step": 29000
    },
    {
      "epoch": 60.0,
      "eval_accuracy": 0.7695299451204963,
      "eval_f1": 0.6170241467691892,
      "eval_loss": 2.104527235031128,
      "eval_precision": 0.5968162839248434,
      "eval_recall": 0.6386484222284278,
      "eval_runtime": 2.73,
      "eval_samples_per_second": 407.333,
      "eval_steps_per_second": 25.641,
      "step": 29220
    },
    {
      "epoch": 60.57,
      "learning_rate": 1.4856262833675564e-05,
      "loss": 0.0071,
      "step": 29500
    },
    {
      "epoch": 61.0,
      "eval_accuracy": 0.768241469816273,
      "eval_f1": 0.6094618408567168,
      "eval_loss": 2.0994060039520264,
      "eval_precision": 0.5922023182297155,
      "eval_recall": 0.6277576096062553,
      "eval_runtime": 2.7589,
      "eval_samples_per_second": 403.058,
      "eval_steps_per_second": 25.372,
      "step": 29707
    },
    {
      "epoch": 61.6,
      "learning_rate": 1.459958932238193e-05,
      "loss": 0.0076,
      "step": 30000
    },
    {
      "epoch": 62.0,
      "eval_accuracy": 0.7649725602481507,
      "eval_f1": 0.609375,
      "eval_loss": 2.0936837196350098,
      "eval_precision": 0.5794510198942332,
      "eval_recall": 0.6425579447081821,
      "eval_runtime": 2.7348,
      "eval_samples_per_second": 406.615,
      "eval_steps_per_second": 25.596,
      "step": 30194
    },
    {
      "epoch": 62.63,
      "learning_rate": 1.4342915811088295e-05,
      "loss": 0.0082,
      "step": 30500
    },
    {
      "epoch": 63.0,
      "eval_accuracy": 0.7682653304700549,
      "eval_f1": 0.6062616078535421,
      "eval_loss": 2.0307247638702393,
      "eval_precision": 0.5774576699519838,
      "eval_recall": 0.6380899190170344,
      "eval_runtime": 2.7517,
      "eval_samples_per_second": 404.118,
      "eval_steps_per_second": 25.439,
      "step": 30681
    },
    {
      "epoch": 63.66,
      "learning_rate": 1.4086242299794661e-05,
      "loss": 0.0068,
      "step": 31000
    },
    {
      "epoch": 64.0,
      "eval_accuracy": 0.7597232164161298,
      "eval_f1": 0.6074766355140186,
      "eval_loss": 2.1657230854034424,
      "eval_precision": 0.581990278843694,
      "eval_recall": 0.635297402960067,
      "eval_runtime": 2.7205,
      "eval_samples_per_second": 408.743,
      "eval_steps_per_second": 25.73,
      "step": 31168
    },
    {
      "epoch": 64.68,
      "learning_rate": 1.3829568788501027e-05,
      "loss": 0.0065,
      "step": 31500
    },
    {
      "epoch": 65.0,
      "eval_accuracy": 0.7691720353137675,
      "eval_f1": 0.6134431455897981,
      "eval_loss": 2.0141701698303223,
      "eval_precision": 0.5850012667848999,
      "eval_recall": 0.644791957553756,
      "eval_runtime": 2.7583,
      "eval_samples_per_second": 403.142,
      "eval_steps_per_second": 25.378,
      "step": 31655
    },
    {
      "epoch": 65.71,
      "learning_rate": 1.3572895277207393e-05,
      "loss": 0.0062,
      "step": 32000
    },
    {
      "epoch": 66.0,
      "eval_accuracy": 0.76022429014555,
      "eval_f1": 0.6064225053078556,
      "eval_loss": 2.1378581523895264,
      "eval_precision": 0.5777496839443742,
      "eval_recall": 0.6380899190170344,
      "eval_runtime": 2.7502,
      "eval_samples_per_second": 404.334,
      "eval_steps_per_second": 25.453,
      "step": 32142
    },
    {
      "epoch": 66.74,
      "learning_rate": 1.331622176591376e-05,
      "loss": 0.0059,
      "step": 32500
    },
    {
      "epoch": 67.0,
      "eval_accuracy": 0.7631352899069435,
      "eval_f1": 0.6117240462581417,
      "eval_loss": 2.1318540573120117,
      "eval_precision": 0.5837138508371386,
      "eval_recall": 0.6425579447081821,
      "eval_runtime": 2.7357,
      "eval_samples_per_second": 406.471,
      "eval_steps_per_second": 25.587,
      "step": 32629
    },
    {
      "epoch": 67.76,
      "learning_rate": 1.3059548254620124e-05,
      "loss": 0.0053,
      "step": 33000
    },
    {
      "epoch": 68.0,
      "eval_accuracy": 0.768217609162491,
      "eval_f1": 0.6046449900464498,
      "eval_loss": 2.1246144771575928,
      "eval_precision": 0.5761254425897825,
      "eval_recall": 0.6361351577771572,
      "eval_runtime": 2.7525,
      "eval_samples_per_second": 404.001,
      "eval_steps_per_second": 25.432,
      "step": 33116
    },
    {
      "epoch": 68.79,
      "learning_rate": 1.280287474332649e-05,
      "loss": 0.0049,
      "step": 33500
    },
    {
      "epoch": 69.0,
      "eval_accuracy": 0.765736101169172,
      "eval_f1": 0.60803618946248,
      "eval_loss": 2.151386022567749,
      "eval_precision": 0.5806861499364676,
      "eval_recall": 0.6380899190170344,
      "eval_runtime": 2.7251,
      "eval_samples_per_second": 408.061,
      "eval_steps_per_second": 25.687,
      "step": 33603
    },
    {
      "epoch": 69.82,
      "learning_rate": 1.2546201232032854e-05,
      "loss": 0.0037,
      "step": 34000
    },
    {
      "epoch": 70.0,
      "eval_accuracy": 0.7679790026246719,
      "eval_f1": 0.6107114308553158,
      "eval_loss": 2.163627862930298,
      "eval_precision": 0.5839490445859873,
      "eval_recall": 0.6400446802569115,
      "eval_runtime": 2.8617,
      "eval_samples_per_second": 388.576,
      "eval_steps_per_second": 24.461,
      "step": 34090
    },
    {
      "epoch": 70.84,
      "learning_rate": 1.228952772073922e-05,
      "loss": 0.0053,
      "step": 34500
    },
    {
      "epoch": 71.0,
      "eval_accuracy": 0.7639465521355285,
      "eval_f1": 0.6052596089008766,
      "eval_loss": 2.1477560997009277,
      "eval_precision": 0.5852895148669797,
      "eval_recall": 0.6266406031834683,
      "eval_runtime": 2.752,
      "eval_samples_per_second": 404.075,
      "eval_steps_per_second": 25.436,
      "step": 34577
    },
    {
      "epoch": 71.87,
      "learning_rate": 1.2032854209445585e-05,
      "loss": 0.0051,
      "step": 35000
    },
    {
      "epoch": 72.0,
      "eval_accuracy": 0.7688141255070389,
      "eval_f1": 0.6074976818121606,
      "eval_loss": 2.1522202491760254,
      "eval_precision": 0.5778729838709677,
      "eval_recall": 0.6403239318626082,
      "eval_runtime": 2.7541,
      "eval_samples_per_second": 403.762,
      "eval_steps_per_second": 25.417,
      "step": 35064
    },
    {
      "epoch": 72.9,
      "learning_rate": 1.1776180698151951e-05,
      "loss": 0.0047,
      "step": 35500
    },
    {
      "epoch": 73.0,
      "eval_accuracy": 0.767120019088523,
      "eval_f1": 0.6093333333333334,
      "eval_loss": 2.1609299182891846,
      "eval_precision": 0.5830569022709875,
      "eval_recall": 0.6380899190170344,
      "eval_runtime": 2.741,
      "eval_samples_per_second": 405.695,
      "eval_steps_per_second": 25.538,
      "step": 35551
    },
    {
      "epoch": 73.92,
      "learning_rate": 1.1519507186858315e-05,
      "loss": 0.0036,
      "step": 36000
    },
    {
      "epoch": 74.0,
      "eval_accuracy": 0.7705559532331185,
      "eval_f1": 0.6200566878121204,
      "eval_loss": 2.175739288330078,
      "eval_precision": 0.6000522466039707,
      "eval_recall": 0.6414409382853952,
      "eval_runtime": 2.7261,
      "eval_samples_per_second": 407.915,
      "eval_steps_per_second": 25.678,
      "step": 36038
    },
    {
      "epoch": 74.95,
      "learning_rate": 1.1262833675564683e-05,
      "loss": 0.004,
      "step": 36500
    },
    {
      "epoch": 75.0,
      "eval_accuracy": 0.7661894535910284,
      "eval_f1": 0.6165353279016962,
      "eval_loss": 2.2280185222625732,
      "eval_precision": 0.5908858166922683,
      "eval_recall": 0.6445127059480592,
      "eval_runtime": 2.75,
      "eval_samples_per_second": 404.367,
      "eval_steps_per_second": 25.455,
      "step": 36525
    },
    {
      "epoch": 75.98,
      "learning_rate": 1.1006160164271048e-05,
      "loss": 0.0036,
      "step": 37000
    },
    {
      "epoch": 76.0,
      "eval_accuracy": 0.770985445001193,
      "eval_f1": 0.6190347071583514,
      "eval_loss": 2.219874382019043,
      "eval_precision": 0.6015810276679842,
      "eval_recall": 0.6375314158056409,
      "eval_runtime": 2.7612,
      "eval_samples_per_second": 402.72,
      "eval_steps_per_second": 25.351,
      "step": 37012
    },
    {
      "epoch": 77.0,
      "eval_accuracy": 0.7684800763540921,
      "eval_f1": 0.6117552978808477,
      "eval_loss": 2.1809566020965576,
      "eval_precision": 0.5851606323304437,
      "eval_recall": 0.6408824350740017,
      "eval_runtime": 2.7598,
      "eval_samples_per_second": 402.922,
      "eval_steps_per_second": 25.364,
      "step": 37499
    },
    {
      "epoch": 77.0,
      "learning_rate": 1.0749486652977414e-05,
      "loss": 0.0043,
      "step": 37500
    },
    {
      "epoch": 78.0,
      "eval_accuracy": 0.7688857074683846,
      "eval_f1": 0.6095212623696176,
      "eval_loss": 2.2160749435424805,
      "eval_precision": 0.5848088273030536,
      "eval_recall": 0.636414409382854,
      "eval_runtime": 2.7615,
      "eval_samples_per_second": 402.681,
      "eval_steps_per_second": 25.349,
      "step": 37986
    },
    {
      "epoch": 78.03,
      "learning_rate": 1.0492813141683778e-05,
      "loss": 0.0039,
      "step": 38000
    },
    {
      "epoch": 79.0,
      "eval_accuracy": 0.7693867811978048,
      "eval_f1": 0.6086727989487517,
      "eval_loss": 2.187803030014038,
      "eval_precision": 0.5748324646314222,
      "eval_recall": 0.6467467187936331,
      "eval_runtime": 2.7479,
      "eval_samples_per_second": 404.677,
      "eval_steps_per_second": 25.474,
      "step": 38473
    },
    {
      "epoch": 79.06,
      "learning_rate": 1.0236139630390144e-05,
      "loss": 0.0052,
      "step": 38500
    },
    {
      "epoch": 80.0,
      "eval_accuracy": 0.7653066094010976,
      "eval_f1": 0.6083209909788608,
      "eval_loss": 2.271164894104004,
      "eval_precision": 0.5873634945397815,
      "eval_recall": 0.6308293772689193,
      "eval_runtime": 2.7444,
      "eval_samples_per_second": 405.19,
      "eval_steps_per_second": 25.507,
      "step": 38960
    },
    {
      "epoch": 80.08,
      "learning_rate": 9.979466119096509e-06,
      "loss": 0.0034,
      "step": 39000
    },
    {
      "epoch": 81.0,
      "eval_accuracy": 0.7658076831305177,
      "eval_f1": 0.6129723934601984,
      "eval_loss": 2.264491319656372,
      "eval_precision": 0.5892811131151765,
      "eval_recall": 0.6386484222284278,
      "eval_runtime": 2.7444,
      "eval_samples_per_second": 405.192,
      "eval_steps_per_second": 25.507,
      "step": 39447
    },
    {
      "epoch": 81.11,
      "learning_rate": 9.722792607802875e-06,
      "loss": 0.0027,
      "step": 39500
    },
    {
      "epoch": 82.0,
      "eval_accuracy": 0.7651157241708423,
      "eval_f1": 0.6160738528373609,
      "eval_loss": 2.2353475093841553,
      "eval_precision": 0.5994715984147952,
      "eval_recall": 0.6336218933258866,
      "eval_runtime": 2.7335,
      "eval_samples_per_second": 406.808,
      "eval_steps_per_second": 25.608,
      "step": 39934
    },
    {
      "epoch": 82.14,
      "learning_rate": 9.46611909650924e-06,
      "loss": 0.0026,
      "step": 40000
    },
    {
      "epoch": 83.0,
      "eval_accuracy": 0.7630398472918158,
      "eval_f1": 0.6092892517735242,
      "eval_loss": 2.3130922317504883,
      "eval_precision": 0.5850899742930591,
      "eval_recall": 0.6355766545657637,
      "eval_runtime": 2.7731,
      "eval_samples_per_second": 400.996,
      "eval_steps_per_second": 25.243,
      "step": 40421
    },
    {
      "epoch": 83.16,
      "learning_rate": 9.209445585215607e-06,
      "loss": 0.0017,
      "step": 40500
    },
    {
      "epoch": 84.0,
      "eval_accuracy": 0.766022429014555,
      "eval_f1": 0.6101919258769026,
      "eval_loss": 2.279829978942871,
      "eval_precision": 0.5800201308505284,
      "eval_recall": 0.643674951130969,
      "eval_runtime": 2.8746,
      "eval_samples_per_second": 386.835,
      "eval_steps_per_second": 24.351,
      "step": 40908
    },
    {
      "epoch": 84.19,
      "learning_rate": 8.952772073921972e-06,
      "loss": 0.0022,
      "step": 41000
    },
    {
      "epoch": 85.0,
      "eval_accuracy": 0.7636840849439275,
      "eval_f1": 0.6126270733012306,
      "eval_loss": 2.3181393146514893,
      "eval_precision": 0.5879332477535302,
      "eval_recall": 0.639486177045518,
      "eval_runtime": 2.7182,
      "eval_samples_per_second": 409.097,
      "eval_steps_per_second": 25.752,
      "step": 41395
    },
    {
      "epoch": 85.22,
      "learning_rate": 8.696098562628338e-06,
      "loss": 0.0032,
      "step": 41500
    },
    {
      "epoch": 86.0,
      "eval_accuracy": 0.769625387735624,
      "eval_f1": 0.6169463995668651,
      "eval_loss": 2.296386480331421,
      "eval_precision": 0.5986340950879958,
      "eval_recall": 0.636414409382854,
      "eval_runtime": 2.7321,
      "eval_samples_per_second": 407.016,
      "eval_steps_per_second": 25.622,
      "step": 41882
    },
    {
      "epoch": 86.24,
      "learning_rate": 8.439425051334702e-06,
      "loss": 0.003,
      "step": 42000
    },
    {
      "epoch": 87.0,
      "eval_accuracy": 0.7664996420901933,
      "eval_f1": 0.6199271942834029,
      "eval_loss": 2.250883102416992,
      "eval_precision": 0.5993222106360793,
      "eval_recall": 0.6419994414967886,
      "eval_runtime": 2.7205,
      "eval_samples_per_second": 408.748,
      "eval_steps_per_second": 25.731,
      "step": 42369
    },
    {
      "epoch": 87.27,
      "learning_rate": 8.182751540041068e-06,
      "loss": 0.003,
      "step": 42500
    },
    {
      "epoch": 88.0,
      "eval_accuracy": 0.7704843712717728,
      "eval_f1": 0.6209611729568286,
      "eval_loss": 2.2512495517730713,
      "eval_precision": 0.604227212681638,
      "eval_recall": 0.6386484222284278,
      "eval_runtime": 2.7574,
      "eval_samples_per_second": 403.273,
      "eval_steps_per_second": 25.386,
      "step": 42856
    },
    {
      "epoch": 88.3,
      "learning_rate": 7.926078028747433e-06,
      "loss": 0.0027,
      "step": 43000
    },
    {
      "epoch": 89.0,
      "eval_accuracy": 0.7695299451204963,
      "eval_f1": 0.612212529738303,
      "eval_loss": 2.2786777019500732,
      "eval_precision": 0.5811794228356336,
      "eval_recall": 0.6467467187936331,
      "eval_runtime": 2.7267,
      "eval_samples_per_second": 407.814,
      "eval_steps_per_second": 25.672,
      "step": 43343
    },
    {
      "epoch": 89.32,
      "learning_rate": 7.669404517453799e-06,
      "loss": 0.0016,
      "step": 43500
    },
    {
      "epoch": 90.0,
      "eval_accuracy": 0.7652588880935338,
      "eval_f1": 0.6130278406820301,
      "eval_loss": 2.2572543621063232,
      "eval_precision": 0.5860927152317881,
      "eval_recall": 0.6425579447081821,
      "eval_runtime": 2.7593,
      "eval_samples_per_second": 403.005,
      "eval_steps_per_second": 25.369,
      "step": 43830
    },
    {
      "epoch": 90.35,
      "learning_rate": 7.412731006160165e-06,
      "loss": 0.0028,
      "step": 44000
    },
    {
      "epoch": 91.0,
      "eval_accuracy": 0.7694106418515867,
      "eval_f1": 0.6204956463496316,
      "eval_loss": 2.247675895690918,
      "eval_precision": 0.596292481977343,
      "eval_recall": 0.6467467187936331,
      "eval_runtime": 2.7505,
      "eval_samples_per_second": 404.29,
      "eval_steps_per_second": 25.45,
      "step": 44317
    },
    {
      "epoch": 91.38,
      "learning_rate": 7.15605749486653e-06,
      "loss": 0.0022,
      "step": 44500
    },
    {
      "epoch": 92.0,
      "eval_accuracy": 0.765187306132188,
      "eval_f1": 0.6163021868787275,
      "eval_loss": 2.2445809841156006,
      "eval_precision": 0.5865287588294652,
      "eval_recall": 0.6492599832449036,
      "eval_runtime": 2.7626,
      "eval_samples_per_second": 402.524,
      "eval_steps_per_second": 25.339,
      "step": 44804
    },
    {
      "epoch": 92.4,
      "learning_rate": 6.899383983572895e-06,
      "loss": 0.0017,
      "step": 45000
    },
    {
      "epoch": 93.0,
      "eval_accuracy": 0.7661417322834646,
      "eval_f1": 0.6177255739455418,
      "eval_loss": 2.25286602973938,
      "eval_precision": 0.591664535924316,
      "eval_recall": 0.6461882155822396,
      "eval_runtime": 2.7411,
      "eval_samples_per_second": 405.68,
      "eval_steps_per_second": 25.537,
      "step": 45291
    },
    {
      "epoch": 93.43,
      "learning_rate": 6.642710472279261e-06,
      "loss": 0.0017,
      "step": 45500
    },
    {
      "epoch": 94.0,
      "eval_accuracy": 0.7649725602481507,
      "eval_f1": 0.6157979580870501,
      "eval_loss": 2.2623953819274902,
      "eval_precision": 0.5933212529122444,
      "eval_recall": 0.6400446802569115,
      "eval_runtime": 2.7635,
      "eval_samples_per_second": 402.391,
      "eval_steps_per_second": 25.33,
      "step": 45778
    },
    {
      "epoch": 94.46,
      "learning_rate": 6.386036960985627e-06,
      "loss": 0.0015,
      "step": 46000
    },
    {
      "epoch": 95.0,
      "eval_accuracy": 0.7649964209019328,
      "eval_f1": 0.6160291931342073,
      "eval_loss": 2.2783970832824707,
      "eval_precision": 0.5969093766369827,
      "eval_recall": 0.636414409382854,
      "eval_runtime": 2.7635,
      "eval_samples_per_second": 402.387,
      "eval_steps_per_second": 25.33,
      "step": 46265
    },
    {
      "epoch": 95.48,
      "learning_rate": 6.129363449691992e-06,
      "loss": 0.0012,
      "step": 46500
    },
    {
      "epoch": 96.0,
      "eval_accuracy": 0.7629444046766881,
      "eval_f1": 0.6143217749435366,
      "eval_loss": 2.303816318511963,
      "eval_precision": 0.58590978205778,
      "eval_recall": 0.6456297123708461,
      "eval_runtime": 2.7486,
      "eval_samples_per_second": 404.571,
      "eval_steps_per_second": 25.468,
      "step": 46752
    },
    {
      "epoch": 96.51,
      "learning_rate": 5.872689938398357e-06,
      "loss": 0.0019,
      "step": 47000
    },
    {
      "epoch": 97.0,
      "eval_accuracy": 0.764900978286805,
      "eval_f1": 0.6164437971666887,
      "eval_loss": 2.3128607273101807,
      "eval_precision": 0.5861027190332326,
      "eval_recall": 0.6500977380619939,
      "eval_runtime": 2.7259,
      "eval_samples_per_second": 407.943,
      "eval_steps_per_second": 25.68,
      "step": 47239
    },
    {
      "epoch": 97.54,
      "learning_rate": 5.6160164271047226e-06,
      "loss": 0.001,
      "step": 47500
    },
    {
      "epoch": 98.0,
      "eval_accuracy": 0.7681698878549272,
      "eval_f1": 0.615528781793842,
      "eval_loss": 2.3077099323272705,
      "eval_precision": 0.5911545384417588,
      "eval_recall": 0.6419994414967886,
      "eval_runtime": 2.7427,
      "eval_samples_per_second": 405.435,
      "eval_steps_per_second": 25.522,
      "step": 47726
    },
    {
      "epoch": 98.56,
      "learning_rate": 5.359342915811089e-06,
      "loss": 0.0009,
      "step": 48000
    },
    {
      "epoch": 99.0,
      "eval_accuracy": 0.7633023144834169,
      "eval_f1": 0.6161656646626587,
      "eval_loss": 2.3493497371673584,
      "eval_precision": 0.5906762295081968,
      "eval_recall": 0.6439542027366657,
      "eval_runtime": 2.7482,
      "eval_samples_per_second": 404.633,
      "eval_steps_per_second": 25.471,
      "step": 48213
    },
    {
      "epoch": 99.59,
      "learning_rate": 5.102669404517454e-06,
      "loss": 0.0015,
      "step": 48500
    },
    {
      "epoch": 100.0,
      "eval_accuracy": 0.7701264614650442,
      "eval_f1": 0.62121007950411,
      "eval_loss": 2.319547414779663,
      "eval_precision": 0.6002604166666666,
      "eval_recall": 0.643674951130969,
      "eval_runtime": 2.8329,
      "eval_samples_per_second": 392.534,
      "eval_steps_per_second": 24.71,
      "step": 48700
    },
    {
      "epoch": 100.62,
      "learning_rate": 4.845995893223819e-06,
      "loss": 0.001,
      "step": 49000
    },
    {
      "epoch": 101.0,
      "eval_accuracy": 0.7711286089238845,
      "eval_f1": 0.6214266631044617,
      "eval_loss": 2.3443996906280518,
      "eval_precision": 0.5956466069142126,
      "eval_recall": 0.6495392348506004,
      "eval_runtime": 2.7763,
      "eval_samples_per_second": 400.528,
      "eval_steps_per_second": 25.213,
      "step": 49187
    },
    {
      "epoch": 101.64,
      "learning_rate": 4.5893223819301845e-06,
      "loss": 0.0008,
      "step": 49500
    },
    {
      "epoch": 102.0,
      "eval_accuracy": 0.7638749701741828,
      "eval_f1": 0.6155906777390839,
      "eval_loss": 2.404651403427124,
      "eval_precision": 0.5915057915057915,
      "eval_recall": 0.6417201898910919,
      "eval_runtime": 2.7517,
      "eval_samples_per_second": 404.111,
      "eval_steps_per_second": 25.439,
      "step": 49674
    },
    {
      "epoch": 102.67,
      "learning_rate": 4.332648870636551e-06,
      "loss": 0.0011,
      "step": 50000
    },
    {
      "epoch": 103.0,
      "eval_accuracy": 0.7672393223574326,
      "eval_f1": 0.6098464796188459,
      "eval_loss": 2.344151258468628,
      "eval_precision": 0.579622641509434,
      "eval_recall": 0.6433956995252723,
      "eval_runtime": 2.7536,
      "eval_samples_per_second": 403.83,
      "eval_steps_per_second": 25.421,
      "step": 50161
    },
    {
      "epoch": 103.7,
      "learning_rate": 4.075975359342916e-06,
      "loss": 0.0009,
      "step": 50500
    },
    {
      "epoch": 104.0,
      "eval_accuracy": 0.768217609162491,
      "eval_f1": 0.6160439266104192,
      "eval_loss": 2.3377583026885986,
      "eval_precision": 0.5918682449819866,
      "eval_recall": 0.6422786931024853,
      "eval_runtime": 2.718,
      "eval_samples_per_second": 409.123,
      "eval_steps_per_second": 25.754,
      "step": 50648
    },
    {
      "epoch": 104.72,
      "learning_rate": 3.819301848049281e-06,
      "loss": 0.0011,
      "step": 51000
    },
    {
      "epoch": 105.0,
      "eval_accuracy": 0.7702934860415175,
      "eval_f1": 0.6217602591792657,
      "eval_loss": 2.319119691848755,
      "eval_precision": 0.6017768487065587,
      "eval_recall": 0.6431164479195756,
      "eval_runtime": 2.7509,
      "eval_samples_per_second": 404.228,
      "eval_steps_per_second": 25.446,
      "step": 51135
    },
    {
      "epoch": 105.75,
      "learning_rate": 3.5626283367556473e-06,
      "loss": 0.0007,
      "step": 51500
    },
    {
      "epoch": 106.0,
      "eval_accuracy": 0.7682891911238368,
      "eval_f1": 0.6160821442859047,
      "eval_loss": 2.376638412475586,
      "eval_precision": 0.5895865237366003,
      "eval_recall": 0.6450712091594527,
      "eval_runtime": 2.7518,
      "eval_samples_per_second": 404.103,
      "eval_steps_per_second": 25.438,
      "step": 51622
    },
    {
      "epoch": 106.78,
      "learning_rate": 3.3059548254620125e-06,
      "loss": 0.0004,
      "step": 52000
    },
    {
      "epoch": 107.0,
      "eval_accuracy": 0.77568599379623,
      "eval_f1": 0.624123043712898,
      "eval_loss": 2.3492467403411865,
      "eval_precision": 0.6037588097102584,
      "eval_recall": 0.6459089639765428,
      "eval_runtime": 2.7841,
      "eval_samples_per_second": 399.413,
      "eval_steps_per_second": 25.143,
      "step": 52109
    },
    {
      "epoch": 107.8,
      "learning_rate": 3.049281314168378e-06,
      "loss": 0.0008,
      "step": 52500
    },
    {
      "epoch": 108.0,
      "eval_accuracy": 0.7681221665473634,
      "eval_f1": 0.620874698148645,
      "eval_loss": 2.3653135299682617,
      "eval_precision": 0.5974696617609089,
      "eval_recall": 0.6461882155822396,
      "eval_runtime": 2.742,
      "eval_samples_per_second": 405.547,
      "eval_steps_per_second": 25.529,
      "step": 52596
    },
    {
      "epoch": 108.83,
      "learning_rate": 2.7926078028747435e-06,
      "loss": 0.0005,
      "step": 53000
    },
    {
      "epoch": 109.0,
      "eval_accuracy": 0.7691720353137675,
      "eval_f1": 0.6206246634356489,
      "eval_loss": 2.3852195739746094,
      "eval_precision": 0.5991681829997401,
      "eval_recall": 0.643674951130969,
      "eval_runtime": 2.756,
      "eval_samples_per_second": 403.488,
      "eval_steps_per_second": 25.399,
      "step": 53083
    },
    {
      "epoch": 109.86,
      "learning_rate": 2.5359342915811088e-06,
      "loss": 0.0005,
      "step": 53500
    },
    {
      "epoch": 110.0,
      "eval_accuracy": 0.7685277976616559,
      "eval_f1": 0.6224392891059558,
      "eval_loss": 2.406304359436035,
      "eval_precision": 0.6052770448548813,
      "eval_recall": 0.6406031834683049,
      "eval_runtime": 2.7427,
      "eval_samples_per_second": 405.436,
      "eval_steps_per_second": 25.522,
      "step": 53570
    },
    {
      "epoch": 110.88,
      "learning_rate": 2.2792607802874745e-06,
      "loss": 0.0008,
      "step": 54000
    },
    {
      "epoch": 111.0,
      "eval_accuracy": 0.7683369124314006,
      "eval_f1": 0.6195049371026647,
      "eval_loss": 2.4256536960601807,
      "eval_precision": 0.6007345225603358,
      "eval_recall": 0.639486177045518,
      "eval_runtime": 2.7562,
      "eval_samples_per_second": 403.45,
      "eval_steps_per_second": 25.397,
      "step": 54057
    },
    {
      "epoch": 111.91,
      "learning_rate": 2.0225872689938397e-06,
      "loss": 0.0009,
      "step": 54500
    },
    {
      "epoch": 112.0,
      "eval_accuracy": 0.7700071581961345,
      "eval_f1": 0.6207082267402719,
      "eval_loss": 2.40315842628479,
      "eval_precision": 0.5993239729589184,
      "eval_recall": 0.643674951130969,
      "eval_runtime": 2.7393,
      "eval_samples_per_second": 405.937,
      "eval_steps_per_second": 25.554,
      "step": 54544
    },
    {
      "epoch": 112.94,
      "learning_rate": 1.7659137577002054e-06,
      "loss": 0.0006,
      "step": 55000
    },
    {
      "epoch": 113.0,
      "eval_accuracy": 0.770722977809592,
      "eval_f1": 0.6195783536994763,
      "eval_loss": 2.3877792358398438,
      "eval_precision": 0.5967408173823073,
      "eval_recall": 0.6442334543423625,
      "eval_runtime": 2.7437,
      "eval_samples_per_second": 405.288,
      "eval_steps_per_second": 25.513,
      "step": 55031
    },
    {
      "epoch": 113.96,
      "learning_rate": 1.509240246406571e-06,
      "loss": 0.0003,
      "step": 55500
    },
    {
      "epoch": 114.0,
      "eval_accuracy": 0.771271772846576,
      "eval_f1": 0.6211180124223602,
      "eval_loss": 2.3938703536987305,
      "eval_precision": 0.6013071895424836,
      "eval_recall": 0.6422786931024853,
      "eval_runtime": 2.7378,
      "eval_samples_per_second": 406.165,
      "eval_steps_per_second": 25.568,
      "step": 55518
    },
    {
      "epoch": 114.99,
      "learning_rate": 1.2525667351129362e-06,
      "loss": 0.0003,
      "step": 56000
    },
    {
      "epoch": 115.0,
      "eval_accuracy": 0.7702934860415175,
      "eval_f1": 0.618289722147289,
      "eval_loss": 2.412465810775757,
      "eval_precision": 0.597965040438299,
      "eval_recall": 0.6400446802569115,
      "eval_runtime": 2.8471,
      "eval_samples_per_second": 390.575,
      "eval_steps_per_second": 24.587,
      "step": 56005
    },
    {
      "epoch": 116.0,
      "eval_accuracy": 0.7705798138869006,
      "eval_f1": 0.6196730099169123,
      "eval_loss": 2.420255661010742,
      "eval_precision": 0.5957227518680752,
      "eval_recall": 0.6456297123708461,
      "eval_runtime": 2.7618,
      "eval_samples_per_second": 402.642,
      "eval_steps_per_second": 25.346,
      "step": 56492
    },
    {
      "epoch": 116.02,
      "learning_rate": 9.958932238193019e-07,
      "loss": 0.0003,
      "step": 56500
    },
    {
      "epoch": 117.0,
      "eval_accuracy": 0.770722977809592,
      "eval_f1": 0.6205501618122977,
      "eval_loss": 2.410393476486206,
      "eval_precision": 0.6,
      "eval_recall": 0.6425579447081821,
      "eval_runtime": 2.7288,
      "eval_samples_per_second": 407.51,
      "eval_steps_per_second": 25.653,
      "step": 56979
    },
    {
      "epoch": 117.04,
      "learning_rate": 7.392197125256674e-07,
      "loss": 0.0004,
      "step": 57000
    },
    {
      "epoch": 118.0,
      "eval_accuracy": 0.7695538057742782,
      "eval_f1": 0.6216835016835018,
      "eval_loss": 2.4210033416748047,
      "eval_precision": 0.6004162330905307,
      "eval_recall": 0.6445127059480592,
      "eval_runtime": 2.7601,
      "eval_samples_per_second": 402.887,
      "eval_steps_per_second": 25.362,
      "step": 57466
    },
    {
      "epoch": 118.07,
      "learning_rate": 4.82546201232033e-07,
      "loss": 0.0004,
      "step": 57500
    },
    {
      "epoch": 119.0,
      "eval_accuracy": 0.7691720353137675,
      "eval_f1": 0.6201508620689655,
      "eval_loss": 2.4213058948516846,
      "eval_precision": 0.5990111891751236,
      "eval_recall": 0.6428371963138788,
      "eval_runtime": 2.776,
      "eval_samples_per_second": 400.579,
      "eval_steps_per_second": 25.216,
      "step": 57953
    },
    {
      "epoch": 119.1,
      "learning_rate": 2.2587268993839835e-07,
      "loss": 0.0004,
      "step": 58000
    },
    {
      "epoch": 120.0,
      "eval_accuracy": 0.7694106418515867,
      "eval_f1": 0.6200296535921282,
      "eval_loss": 2.4216408729553223,
      "eval_precision": 0.5992704533611256,
      "eval_recall": 0.6422786931024853,
      "eval_runtime": 2.7416,
      "eval_samples_per_second": 405.598,
      "eval_steps_per_second": 25.532,
      "step": 58440
    },
    {
      "epoch": 120.0,
      "step": 58440,
      "total_flos": 1.220726808511488e+17,
      "train_loss": 0.04618847079620959,
      "train_runtime": 7270.5358,
      "train_samples_per_second": 128.491,
      "train_steps_per_second": 8.038
    }
  ],
  "max_steps": 58440,
  "num_train_epochs": 120,
  "total_flos": 1.220726808511488e+17,
  "trial_name": null,
  "trial_params": null
}