{
  "best_metric": 0.9160277022013357,
  "best_model_checkpoint": "./fp32_2e_5/models/qqp-roberta-base/checkpoint-67000",
  "epoch": 5.892181866150734,
  "global_step": 67000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09,
      "learning_rate": 2.9312619082515024e-06,
      "loss": 0.5476,
      "step": 1000
    },
    {
      "epoch": 0.09,
      "eval_accuracy": 0.8186000494682166,
      "eval_combined_score": 0.781458719476822,
      "eval_f1": 0.7443173894854274,
      "eval_loss": 0.3999478816986084,
      "eval_runtime": 100.5489,
      "eval_samples_per_second": 402.093,
      "eval_steps_per_second": 12.571,
      "step": 1000
    },
    {
      "epoch": 0.18,
      "learning_rate": 5.862523816503005e-06,
      "loss": 0.3783,
      "step": 2000
    },
    {
      "epoch": 0.18,
      "eval_accuracy": 0.8439772446203314,
      "eval_combined_score": 0.8249961873510662,
      "eval_f1": 0.8060151300818009,
      "eval_loss": 0.3368555009365082,
      "eval_runtime": 105.3628,
      "eval_samples_per_second": 383.722,
      "eval_steps_per_second": 11.997,
      "step": 2000
    },
    {
      "epoch": 0.26,
      "learning_rate": 8.793785724754508e-06,
      "loss": 0.3465,
      "step": 3000
    },
    {
      "epoch": 0.26,
      "eval_accuracy": 0.8580262181548355,
      "eval_combined_score": 0.8396217180601624,
      "eval_f1": 0.8212172179654894,
      "eval_loss": 0.30989906191825867,
      "eval_runtime": 107.6992,
      "eval_samples_per_second": 375.397,
      "eval_steps_per_second": 11.736,
      "step": 3000
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.172504763300601e-05,
      "loss": 0.3284,
      "step": 4000
    },
    {
      "epoch": 0.35,
      "eval_accuracy": 0.8592134553549344,
      "eval_combined_score": 0.8436663967196221,
      "eval_f1": 0.8281193380843096,
      "eval_loss": 0.30866414308547974,
      "eval_runtime": 104.8651,
      "eval_samples_per_second": 385.543,
      "eval_steps_per_second": 12.054,
      "step": 4000
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.4656309541257511e-05,
      "loss": 0.3244,
      "step": 5000
    },
    {
      "epoch": 0.44,
      "eval_accuracy": 0.8497897600791492,
      "eval_combined_score": 0.835341100584005,
      "eval_f1": 0.8208924410888607,
      "eval_loss": 0.32785937190055847,
      "eval_runtime": 101.6619,
      "eval_samples_per_second": 397.691,
      "eval_steps_per_second": 12.433,
      "step": 5000
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.7587571449509016e-05,
      "loss": 0.3087,
      "step": 6000
    },
    {
      "epoch": 0.53,
      "eval_accuracy": 0.8763047242146921,
      "eval_combined_score": 0.8556630011071151,
      "eval_f1": 0.8350212779995382,
      "eval_loss": 0.2904255986213684,
      "eval_runtime": 100.5764,
      "eval_samples_per_second": 401.983,
      "eval_steps_per_second": 12.568,
      "step": 6000
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.9966880911616945e-05,
      "loss": 0.2958,
      "step": 7000
    },
    {
      "epoch": 0.62,
      "eval_accuracy": 0.8771209497897601,
      "eval_combined_score": 0.8548056701204515,
      "eval_f1": 0.8324903904511429,
      "eval_loss": 0.2824174165725708,
      "eval_runtime": 109.1783,
      "eval_samples_per_second": 370.312,
      "eval_steps_per_second": 11.577,
      "step": 7000
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.9779767417927343e-05,
      "loss": 0.2985,
      "step": 8000
    },
    {
      "epoch": 0.7,
      "eval_accuracy": 0.8826861241652239,
      "eval_combined_score": 0.8620260280680964,
      "eval_f1": 0.8413659319709689,
      "eval_loss": 0.2692563235759735,
      "eval_runtime": 109.9785,
      "eval_samples_per_second": 367.617,
      "eval_steps_per_second": 11.493,
      "step": 8000
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.959265392423775e-05,
      "loss": 0.2854,
      "step": 9000
    },
    {
      "epoch": 0.79,
      "eval_accuracy": 0.8790749443482563,
      "eval_combined_score": 0.8634582963003614,
      "eval_f1": 0.8478416482524664,
      "eval_loss": 0.2772486209869385,
      "eval_runtime": 104.6366,
      "eval_samples_per_second": 386.385,
      "eval_steps_per_second": 12.08,
      "step": 9000
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.940554043054815e-05,
      "loss": 0.279,
      "step": 10000
    },
    {
      "epoch": 0.88,
      "eval_accuracy": 0.8851842691070987,
      "eval_combined_score": 0.8687501900384056,
      "eval_f1": 0.8523161109697124,
      "eval_loss": 0.26354455947875977,
      "eval_runtime": 102.343,
      "eval_samples_per_second": 395.044,
      "eval_steps_per_second": 12.351,
      "step": 10000
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.9218426936858555e-05,
      "loss": 0.2728,
      "step": 11000
    },
    {
      "epoch": 0.97,
      "eval_accuracy": 0.8893643334157804,
      "eval_combined_score": 0.8714862904192304,
      "eval_f1": 0.8536082474226804,
      "eval_loss": 0.2556052803993225,
      "eval_runtime": 100.7949,
      "eval_samples_per_second": 401.112,
      "eval_steps_per_second": 12.54,
      "step": 11000
    },
    {
      "epoch": 1.06,
      "learning_rate": 1.9031313443168953e-05,
      "loss": 0.2474,
      "step": 12000
    },
    {
      "epoch": 1.06,
      "eval_accuracy": 0.8884986396240415,
      "eval_combined_score": 0.8716473743117953,
      "eval_f1": 0.854796108999549,
      "eval_loss": 0.26520249247550964,
      "eval_runtime": 102.8555,
      "eval_samples_per_second": 393.076,
      "eval_steps_per_second": 12.289,
      "step": 12000
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.884419994947936e-05,
      "loss": 0.2382,
      "step": 13000
    },
    {
      "epoch": 1.14,
      "eval_accuracy": 0.8925302992827109,
      "eval_combined_score": 0.8776649412356591,
      "eval_f1": 0.8627995831886073,
      "eval_loss": 0.25566014647483826,
      "eval_runtime": 99.8698,
      "eval_samples_per_second": 404.827,
      "eval_steps_per_second": 12.656,
      "step": 13000
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.865708645578976e-05,
      "loss": 0.2437,
      "step": 14000
    },
    {
      "epoch": 1.23,
      "eval_accuracy": 0.8940143457828346,
      "eval_combined_score": 0.8742345089194055,
      "eval_f1": 0.8544546720559764,
      "eval_loss": 0.25206559896469116,
      "eval_runtime": 98.7673,
      "eval_samples_per_second": 409.346,
      "eval_steps_per_second": 12.798,
      "step": 14000
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.8469972962100165e-05,
      "loss": 0.2314,
      "step": 15000
    },
    {
      "epoch": 1.32,
      "eval_accuracy": 0.8962651496413554,
      "eval_combined_score": 0.8779659416144943,
      "eval_f1": 0.859666733587633,
      "eval_loss": 0.2604176104068756,
      "eval_runtime": 99.873,
      "eval_samples_per_second": 404.814,
      "eval_steps_per_second": 12.656,
      "step": 15000
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.8282859468410567e-05,
      "loss": 0.2347,
      "step": 16000
    },
    {
      "epoch": 1.41,
      "eval_accuracy": 0.897872866683156,
      "eval_combined_score": 0.8811078037131652,
      "eval_f1": 0.8643427407431743,
      "eval_loss": 0.2466404139995575,
      "eval_runtime": 99.5334,
      "eval_samples_per_second": 406.195,
      "eval_steps_per_second": 12.699,
      "step": 16000
    },
    {
      "epoch": 1.5,
      "learning_rate": 1.809574597472097e-05,
      "loss": 0.2322,
      "step": 17000
    },
    {
      "epoch": 1.5,
      "eval_accuracy": 0.8983675488498639,
      "eval_combined_score": 0.8828389557469845,
      "eval_f1": 0.867310362644105,
      "eval_loss": 0.24636423587799072,
      "eval_runtime": 100.203,
      "eval_samples_per_second": 403.481,
      "eval_steps_per_second": 12.614,
      "step": 17000
    },
    {
      "epoch": 1.58,
      "learning_rate": 1.790863248103137e-05,
      "loss": 0.2373,
      "step": 18000
    },
    {
      "epoch": 1.58,
      "eval_accuracy": 0.8971061093247589,
      "eval_combined_score": 0.883107204810262,
      "eval_f1": 0.8691083002957649,
      "eval_loss": 0.2558729350566864,
      "eval_runtime": 102.812,
      "eval_samples_per_second": 393.242,
      "eval_steps_per_second": 12.294,
      "step": 18000
    },
    {
      "epoch": 1.67,
      "learning_rate": 1.7721518987341772e-05,
      "loss": 0.2295,
      "step": 19000
    },
    {
      "epoch": 1.67,
      "eval_accuracy": 0.8999752658916647,
      "eval_combined_score": 0.8847323563693139,
      "eval_f1": 0.8694894468469633,
      "eval_loss": 0.25413957238197327,
      "eval_runtime": 101.1101,
      "eval_samples_per_second": 399.861,
      "eval_steps_per_second": 12.501,
      "step": 19000
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.7534405493652177e-05,
      "loss": 0.2295,
      "step": 20000
    },
    {
      "epoch": 1.76,
      "eval_accuracy": 0.9001978728666832,
      "eval_combined_score": 0.8849152590077334,
      "eval_f1": 0.8696326451487837,
      "eval_loss": 0.24725091457366943,
      "eval_runtime": 102.8761,
      "eval_samples_per_second": 392.997,
      "eval_steps_per_second": 12.287,
      "step": 20000
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.734729199996258e-05,
      "loss": 0.2281,
      "step": 21000
    },
    {
      "epoch": 1.85,
      "eval_accuracy": 0.8977491961414791,
      "eval_combined_score": 0.8776085657993287,
      "eval_f1": 0.8574679354571784,
      "eval_loss": 0.2688974142074585,
      "eval_runtime": 99.2566,
      "eval_samples_per_second": 407.328,
      "eval_steps_per_second": 12.735,
      "step": 21000
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.716017850627298e-05,
      "loss": 0.228,
      "step": 22000
    },
    {
      "epoch": 1.93,
      "eval_accuracy": 0.9038090526836507,
      "eval_combined_score": 0.888468647308778,
      "eval_f1": 0.8731282419339053,
      "eval_loss": 0.23130430281162262,
      "eval_runtime": 99.2642,
      "eval_samples_per_second": 407.297,
      "eval_steps_per_second": 12.734,
      "step": 22000
    },
    {
      "epoch": 2.02,
      "learning_rate": 1.6973065012583382e-05,
      "loss": 0.2116,
      "step": 23000
    },
    {
      "epoch": 2.02,
      "eval_accuracy": 0.903067029433589,
      "eval_combined_score": 0.8880802586693867,
      "eval_f1": 0.8730934879051844,
      "eval_loss": 0.24339179694652557,
      "eval_runtime": 99.8059,
      "eval_samples_per_second": 405.086,
      "eval_steps_per_second": 12.665,
      "step": 23000
    },
    {
      "epoch": 2.11,
      "learning_rate": 1.6785951518893787e-05,
      "loss": 0.179,
      "step": 24000
    },
    {
      "epoch": 2.11,
      "eval_accuracy": 0.9035617116002969,
      "eval_combined_score": 0.8877347307631893,
      "eval_f1": 0.8719077499260817,
      "eval_loss": 0.2653352916240692,
      "eval_runtime": 98.8508,
      "eval_samples_per_second": 409.0,
      "eval_steps_per_second": 12.787,
      "step": 24000
    },
    {
      "epoch": 2.2,
      "learning_rate": 1.659883802520419e-05,
      "loss": 0.1752,
      "step": 25000
    },
    {
      "epoch": 2.2,
      "eval_accuracy": 0.9018303240168192,
      "eval_combined_score": 0.8860091930795462,
      "eval_f1": 0.8701880621422732,
      "eval_loss": 0.2552090883255005,
      "eval_runtime": 99.1983,
      "eval_samples_per_second": 407.568,
      "eval_steps_per_second": 12.742,
      "step": 25000
    },
    {
      "epoch": 2.29,
      "learning_rate": 1.6411724531514594e-05,
      "loss": 0.1822,
      "step": 26000
    },
    {
      "epoch": 2.29,
      "eval_accuracy": 0.9064803363838734,
      "eval_combined_score": 0.8916663806141756,
      "eval_f1": 0.8768524248444777,
      "eval_loss": 0.2510685920715332,
      "eval_runtime": 98.8214,
      "eval_samples_per_second": 409.122,
      "eval_steps_per_second": 12.791,
      "step": 26000
    },
    {
      "epoch": 2.37,
      "learning_rate": 1.6224611037824992e-05,
      "loss": 0.1826,
      "step": 27000
    },
    {
      "epoch": 2.37,
      "eval_accuracy": 0.9051199604254266,
      "eval_combined_score": 0.8899457206149071,
      "eval_f1": 0.8747714808043876,
      "eval_loss": 0.24112574756145477,
      "eval_runtime": 100.4418,
      "eval_samples_per_second": 402.521,
      "eval_steps_per_second": 12.584,
      "step": 27000
    },
    {
      "epoch": 2.46,
      "learning_rate": 1.6037497544135397e-05,
      "loss": 0.1807,
      "step": 28000
    },
    {
      "epoch": 2.46,
      "eval_accuracy": 0.9052683650754391,
      "eval_combined_score": 0.8889197827506554,
      "eval_f1": 0.8725712004258717,
      "eval_loss": 0.28508177399635315,
      "eval_runtime": 99.6721,
      "eval_samples_per_second": 405.63,
      "eval_steps_per_second": 12.682,
      "step": 28000
    },
    {
      "epoch": 2.55,
      "learning_rate": 1.58503840504458e-05,
      "loss": 0.1829,
      "step": 29000
    },
    {
      "epoch": 2.55,
      "eval_accuracy": 0.903512243383626,
      "eval_combined_score": 0.888178430111253,
      "eval_f1": 0.8728446168388799,
      "eval_loss": 0.24632664024829865,
      "eval_runtime": 98.7974,
      "eval_samples_per_second": 409.221,
      "eval_steps_per_second": 12.794,
      "step": 29000
    },
    {
      "epoch": 2.64,
      "learning_rate": 1.5663270556756204e-05,
      "loss": 0.1832,
      "step": 30000
    },
    {
      "epoch": 2.64,
      "eval_accuracy": 0.9063071976255256,
      "eval_combined_score": 0.8893352822146093,
      "eval_f1": 0.8723633668036931,
      "eval_loss": 0.2510886788368225,
      "eval_runtime": 99.841,
      "eval_samples_per_second": 404.944,
      "eval_steps_per_second": 12.66,
      "step": 30000
    },
    {
      "epoch": 2.73,
      "learning_rate": 1.5476157063066602e-05,
      "loss": 0.1752,
      "step": 31000
    },
    {
      "epoch": 2.73,
      "eval_accuracy": 0.9047984170170665,
      "eval_combined_score": 0.8909587334582507,
      "eval_f1": 0.8771190498994349,
      "eval_loss": 0.25542038679122925,
      "eval_runtime": 98.5241,
      "eval_samples_per_second": 410.356,
      "eval_steps_per_second": 12.829,
      "step": 31000
    },
    {
      "epoch": 2.81,
      "learning_rate": 1.5289043569377007e-05,
      "loss": 0.1822,
      "step": 32000
    },
    {
      "epoch": 2.81,
      "eval_accuracy": 0.9059361860004946,
      "eval_combined_score": 0.8898770813542074,
      "eval_f1": 0.87381797670792,
      "eval_loss": 0.2384674996137619,
      "eval_runtime": 104.6367,
      "eval_samples_per_second": 386.384,
      "eval_steps_per_second": 12.08,
      "step": 32000
    },
    {
      "epoch": 2.9,
      "learning_rate": 1.510193007568741e-05,
      "loss": 0.1834,
      "step": 33000
    },
    {
      "epoch": 2.9,
      "eval_accuracy": 0.9094484293841207,
      "eval_combined_score": 0.895065771906153,
      "eval_f1": 0.8806831144281851,
      "eval_loss": 0.25289300084114075,
      "eval_runtime": 106.9571,
      "eval_samples_per_second": 378.002,
      "eval_steps_per_second": 11.818,
      "step": 33000
    },
    {
      "epoch": 2.99,
      "learning_rate": 1.4914816581997812e-05,
      "loss": 0.1829,
      "step": 34000
    },
    {
      "epoch": 2.99,
      "eval_accuracy": 0.9079643828839971,
      "eval_combined_score": 0.8944023406718317,
      "eval_f1": 0.8808402984596663,
      "eval_loss": 0.25022605061531067,
      "eval_runtime": 99.5394,
      "eval_samples_per_second": 406.171,
      "eval_steps_per_second": 12.698,
      "step": 34000
    },
    {
      "epoch": 3.08,
      "learning_rate": 1.4727703088308214e-05,
      "loss": 0.1477,
      "step": 35000
    },
    {
      "epoch": 3.08,
      "eval_accuracy": 0.9075686371506307,
      "eval_combined_score": 0.892660304050797,
      "eval_f1": 0.8777519709509634,
      "eval_loss": 0.27075424790382385,
      "eval_runtime": 99.0291,
      "eval_samples_per_second": 408.264,
      "eval_steps_per_second": 12.764,
      "step": 35000
    },
    {
      "epoch": 3.17,
      "learning_rate": 1.4540589594618617e-05,
      "loss": 0.1429,
      "step": 36000
    },
    {
      "epoch": 3.17,
      "eval_accuracy": 0.908780608459065,
      "eval_combined_score": 0.8936844838450184,
      "eval_f1": 0.8785883592309719,
      "eval_loss": 0.2728239893913269,
      "eval_runtime": 99.8083,
      "eval_samples_per_second": 405.076,
      "eval_steps_per_second": 12.664,
      "step": 36000
    },
    {
      "epoch": 3.25,
      "learning_rate": 1.435347610092902e-05,
      "loss": 0.1424,
      "step": 37000
    },
    {
      "epoch": 3.25,
      "eval_accuracy": 0.9095226317091268,
      "eval_combined_score": 0.8944181947724124,
      "eval_f1": 0.8793137578356978,
      "eval_loss": 0.3181060254573822,
      "eval_runtime": 99.1709,
      "eval_samples_per_second": 407.68,
      "eval_steps_per_second": 12.746,
      "step": 37000
    },
    {
      "epoch": 3.34,
      "learning_rate": 1.416636260723942e-05,
      "loss": 0.1421,
      "step": 38000
    },
    {
      "epoch": 3.34,
      "eval_accuracy": 0.9079891169923324,
      "eval_combined_score": 0.8935460697083859,
      "eval_f1": 0.8791030224244393,
      "eval_loss": 0.25459086894989014,
      "eval_runtime": 101.0313,
      "eval_samples_per_second": 400.173,
      "eval_steps_per_second": 12.511,
      "step": 38000
    },
    {
      "epoch": 3.43,
      "learning_rate": 1.3979249113549824e-05,
      "loss": 0.1453,
      "step": 39000
    },
    {
      "epoch": 3.43,
      "eval_accuracy": 0.9072965619589414,
      "eval_combined_score": 0.8937684241287549,
      "eval_f1": 0.8802402862985685,
      "eval_loss": 0.2652507722377777,
      "eval_runtime": 98.7666,
      "eval_samples_per_second": 409.349,
      "eval_steps_per_second": 12.798,
      "step": 39000
    },
    {
      "epoch": 3.52,
      "learning_rate": 1.3792135619860227e-05,
      "loss": 0.1442,
      "step": 40000
    },
    {
      "epoch": 3.52,
      "eval_accuracy": 0.9106109324758842,
      "eval_combined_score": 0.896522122060974,
      "eval_f1": 0.8824333116460638,
      "eval_loss": 0.27689746022224426,
      "eval_runtime": 98.8771,
      "eval_samples_per_second": 408.891,
      "eval_steps_per_second": 12.784,
      "step": 40000
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.360502212617063e-05,
      "loss": 0.1485,
      "step": 41000
    },
    {
      "epoch": 3.61,
      "eval_accuracy": 0.911031412317586,
      "eval_combined_score": 0.8962277549473212,
      "eval_f1": 0.8814240975770563,
      "eval_loss": 0.2877097427845001,
      "eval_runtime": 99.1266,
      "eval_samples_per_second": 407.862,
      "eval_steps_per_second": 12.751,
      "step": 41000
    },
    {
      "epoch": 3.69,
      "learning_rate": 1.341790863248103e-05,
      "loss": 0.1453,
      "step": 42000
    },
    {
      "epoch": 3.69,
      "eval_accuracy": 0.9098194410091516,
      "eval_combined_score": 0.895775167500813,
      "eval_f1": 0.8817308939924743,
      "eval_loss": 0.27044087648391724,
      "eval_runtime": 99.2706,
      "eval_samples_per_second": 407.271,
      "eval_steps_per_second": 12.733,
      "step": 42000
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.3230795138791434e-05,
      "loss": 0.1492,
      "step": 43000
    },
    {
      "epoch": 3.78,
      "eval_accuracy": 0.91348008904279,
      "eval_combined_score": 0.898925685140197,
      "eval_f1": 0.8843712812376041,
      "eval_loss": 0.25266170501708984,
      "eval_runtime": 100.9281,
      "eval_samples_per_second": 400.582,
      "eval_steps_per_second": 12.524,
      "step": 43000
    },
    {
      "epoch": 3.87,
      "learning_rate": 1.3043681645101837e-05,
      "loss": 0.1444,
      "step": 44000
    },
    {
      "epoch": 3.87,
      "eval_accuracy": 0.911847637892654,
      "eval_combined_score": 0.8970186225956724,
      "eval_f1": 0.882189607298691,
      "eval_loss": 0.2621923089027405,
      "eval_runtime": 98.9247,
      "eval_samples_per_second": 408.695,
      "eval_steps_per_second": 12.777,
      "step": 44000
    },
    {
      "epoch": 3.96,
      "learning_rate": 1.285656815141224e-05,
      "loss": 0.1433,
      "step": 45000
    },
    {
      "epoch": 3.96,
      "eval_accuracy": 0.9120949789760079,
      "eval_combined_score": 0.8979680019732961,
      "eval_f1": 0.8838410249705844,
      "eval_loss": 0.27234378457069397,
      "eval_runtime": 102.783,
      "eval_samples_per_second": 393.353,
      "eval_steps_per_second": 12.298,
      "step": 45000
    },
    {
      "epoch": 4.05,
      "learning_rate": 1.2669454657722643e-05,
      "loss": 0.1315,
      "step": 46000
    },
    {
      "epoch": 4.05,
      "eval_accuracy": 0.9134306208261193,
      "eval_combined_score": 0.8995033318593784,
      "eval_f1": 0.8855760428926376,
      "eval_loss": 0.29698267579078674,
      "eval_runtime": 104.3977,
      "eval_samples_per_second": 387.269,
      "eval_steps_per_second": 12.108,
      "step": 46000
    },
    {
      "epoch": 4.13,
      "learning_rate": 1.2482341164033044e-05,
      "loss": 0.1135,
      "step": 47000
    },
    {
      "epoch": 4.13,
      "eval_accuracy": 0.9112540192926045,
      "eval_combined_score": 0.8968766821639864,
      "eval_f1": 0.8824993450353682,
      "eval_loss": 0.30147695541381836,
      "eval_runtime": 100.9744,
      "eval_samples_per_second": 400.398,
      "eval_steps_per_second": 12.518,
      "step": 47000
    },
    {
      "epoch": 4.22,
      "learning_rate": 1.2295227670343448e-05,
      "loss": 0.1178,
      "step": 48000
    },
    {
      "epoch": 4.22,
      "eval_accuracy": 0.9088548107840713,
      "eval_combined_score": 0.8948400699242306,
      "eval_f1": 0.8808253290643899,
      "eval_loss": 0.30377110838890076,
      "eval_runtime": 101.3426,
      "eval_samples_per_second": 398.944,
      "eval_steps_per_second": 12.473,
      "step": 48000
    },
    {
      "epoch": 4.31,
      "learning_rate": 1.2108114176653851e-05,
      "loss": 0.1192,
      "step": 49000
    },
    {
      "epoch": 4.31,
      "eval_accuracy": 0.9117487014593124,
      "eval_combined_score": 0.8974361955933878,
      "eval_f1": 0.8831236897274632,
      "eval_loss": 0.3054348826408386,
      "eval_runtime": 102.6949,
      "eval_samples_per_second": 393.691,
      "eval_steps_per_second": 12.308,
      "step": 49000
    },
    {
      "epoch": 4.4,
      "learning_rate": 1.1921000682964253e-05,
      "loss": 0.1113,
      "step": 50000
    },
    {
      "epoch": 4.4,
      "eval_accuracy": 0.910784071234232,
      "eval_combined_score": 0.8965060853118298,
      "eval_f1": 0.8822280993894277,
      "eval_loss": 0.31508392095565796,
      "eval_runtime": 100.8416,
      "eval_samples_per_second": 400.926,
      "eval_steps_per_second": 12.535,
      "step": 50000
    },
    {
      "epoch": 4.49,
      "learning_rate": 1.1733887189274656e-05,
      "loss": 0.1176,
      "step": 51000
    },
    {
      "epoch": 4.49,
      "eval_accuracy": 0.9117487014593124,
      "eval_combined_score": 0.8964195750181319,
      "eval_f1": 0.8810904485769512,
      "eval_loss": 0.3410942554473877,
      "eval_runtime": 101.1296,
      "eval_samples_per_second": 399.784,
      "eval_steps_per_second": 12.499,
      "step": 51000
    },
    {
      "epoch": 4.57,
      "learning_rate": 1.1546773695585058e-05,
      "loss": 0.1245,
      "step": 52000
    },
    {
      "epoch": 4.57,
      "eval_accuracy": 0.9097947069008162,
      "eval_combined_score": 0.8954127669229248,
      "eval_f1": 0.8810308269450335,
      "eval_loss": 0.3030332028865814,
      "eval_runtime": 100.675,
      "eval_samples_per_second": 401.589,
      "eval_steps_per_second": 12.555,
      "step": 52000
    },
    {
      "epoch": 4.66,
      "learning_rate": 1.1359660201895461e-05,
      "loss": 0.1197,
      "step": 53000
    },
    {
      "epoch": 4.66,
      "eval_accuracy": 0.9111056146425921,
      "eval_combined_score": 0.8957604335940053,
      "eval_f1": 0.8804152525454183,
      "eval_loss": 0.3057613670825958,
      "eval_runtime": 102.3073,
      "eval_samples_per_second": 395.182,
      "eval_steps_per_second": 12.355,
      "step": 53000
    },
    {
      "epoch": 4.75,
      "learning_rate": 1.1172546708205863e-05,
      "loss": 0.1226,
      "step": 54000
    },
    {
      "epoch": 4.75,
      "eval_accuracy": 0.9141479099678457,
      "eval_combined_score": 0.8996355965526801,
      "eval_f1": 0.8851232831375145,
      "eval_loss": 0.31216293573379517,
      "eval_runtime": 101.2062,
      "eval_samples_per_second": 399.481,
      "eval_steps_per_second": 12.489,
      "step": 54000
    },
    {
      "epoch": 4.84,
      "learning_rate": 1.0985433214516266e-05,
      "loss": 0.1177,
      "step": 55000
    },
    {
      "epoch": 4.84,
      "eval_accuracy": 0.9131338115260945,
      "eval_combined_score": 0.8982512840839825,
      "eval_f1": 0.8833687566418704,
      "eval_loss": 0.32634615898132324,
      "eval_runtime": 101.5822,
      "eval_samples_per_second": 398.003,
      "eval_steps_per_second": 12.443,
      "step": 55000
    },
    {
      "epoch": 4.92,
      "learning_rate": 1.079831972082667e-05,
      "loss": 0.1172,
      "step": 56000
    },
    {
      "epoch": 4.92,
      "eval_accuracy": 0.912663863467722,
      "eval_combined_score": 0.8966402794364677,
      "eval_f1": 0.8806166954052135,
      "eval_loss": 0.3014249801635742,
      "eval_runtime": 101.6983,
      "eval_samples_per_second": 397.549,
      "eval_steps_per_second": 12.429,
      "step": 56000
    },
    {
      "epoch": 5.01,
      "learning_rate": 1.0611206227137073e-05,
      "loss": 0.1123,
      "step": 57000
    },
    {
      "epoch": 5.01,
      "eval_accuracy": 0.914296314617858,
      "eval_combined_score": 0.8996530522745814,
      "eval_f1": 0.8850097899313046,
      "eval_loss": 0.3819635510444641,
      "eval_runtime": 101.5331,
      "eval_samples_per_second": 398.195,
      "eval_steps_per_second": 12.449,
      "step": 57000
    },
    {
      "epoch": 5.1,
      "learning_rate": 1.0424092733447473e-05,
      "loss": 0.0951,
      "step": 58000
    },
    {
      "epoch": 5.1,
      "eval_accuracy": 0.9105614642592135,
      "eval_combined_score": 0.8972352717469176,
      "eval_f1": 0.8839090792346218,
      "eval_loss": 0.3715880215167999,
      "eval_runtime": 101.3316,
      "eval_samples_per_second": 398.987,
      "eval_steps_per_second": 12.474,
      "step": 58000
    },
    {
      "epoch": 5.19,
      "learning_rate": 1.0236979239757876e-05,
      "loss": 0.0991,
      "step": 59000
    },
    {
      "epoch": 5.19,
      "eval_accuracy": 0.9139995053178332,
      "eval_combined_score": 0.8992518702473571,
      "eval_f1": 0.8845042351768809,
      "eval_loss": 0.37594878673553467,
      "eval_runtime": 101.3841,
      "eval_samples_per_second": 398.781,
      "eval_steps_per_second": 12.467,
      "step": 59000
    },
    {
      "epoch": 5.28,
      "learning_rate": 1.004986574606828e-05,
      "loss": 0.1035,
      "step": 60000
    },
    {
      "epoch": 5.28,
      "eval_accuracy": 0.9140242394261687,
      "eval_combined_score": 0.8998636256975642,
      "eval_f1": 0.8857030119689596,
      "eval_loss": 0.3245335817337036,
      "eval_runtime": 102.513,
      "eval_samples_per_second": 394.389,
      "eval_steps_per_second": 12.33,
      "step": 60000
    },
    {
      "epoch": 5.36,
      "learning_rate": 9.862752252378681e-06,
      "loss": 0.0967,
      "step": 61000
    },
    {
      "epoch": 5.36,
      "eval_accuracy": 0.9130101409844175,
      "eval_combined_score": 0.8977728972181029,
      "eval_f1": 0.8825356534517885,
      "eval_loss": 0.3583969473838806,
      "eval_runtime": 104.7234,
      "eval_samples_per_second": 386.065,
      "eval_steps_per_second": 12.07,
      "step": 61000
    },
    {
      "epoch": 5.45,
      "learning_rate": 9.675638758689083e-06,
      "loss": 0.1001,
      "step": 62000
    },
    {
      "epoch": 5.45,
      "eval_accuracy": 0.9139995053178332,
      "eval_combined_score": 0.8993361485592781,
      "eval_f1": 0.8846727918007231,
      "eval_loss": 0.3541596829891205,
      "eval_runtime": 101.634,
      "eval_samples_per_second": 397.8,
      "eval_steps_per_second": 12.437,
      "step": 62000
    },
    {
      "epoch": 5.54,
      "learning_rate": 9.488525264999486e-06,
      "loss": 0.0999,
      "step": 63000
    },
    {
      "epoch": 5.54,
      "eval_accuracy": 0.9132327479594361,
      "eval_combined_score": 0.8982896913513594,
      "eval_f1": 0.8833466347432828,
      "eval_loss": 0.36085495352745056,
      "eval_runtime": 103.5986,
      "eval_samples_per_second": 390.256,
      "eval_steps_per_second": 12.201,
      "step": 63000
    },
    {
      "epoch": 5.63,
      "learning_rate": 9.301411771309888e-06,
      "loss": 0.0961,
      "step": 64000
    },
    {
      "epoch": 5.63,
      "eval_accuracy": 0.9137026960178085,
      "eval_combined_score": 0.8989772147109578,
      "eval_f1": 0.884251733404107,
      "eval_loss": 0.3343105912208557,
      "eval_runtime": 101.6502,
      "eval_samples_per_second": 397.736,
      "eval_steps_per_second": 12.435,
      "step": 64000
    },
    {
      "epoch": 5.72,
      "learning_rate": 9.114298277620291e-06,
      "loss": 0.0992,
      "step": 65000
    },
    {
      "epoch": 5.72,
      "eval_accuracy": 0.9153846153846154,
      "eval_combined_score": 0.9020065310865193,
      "eval_f1": 0.8886284467884233,
      "eval_loss": 0.34594395756721497,
      "eval_runtime": 102.9906,
      "eval_samples_per_second": 392.56,
      "eval_steps_per_second": 12.273,
      "step": 65000
    },
    {
      "epoch": 5.8,
      "learning_rate": 8.927184783930693e-06,
      "loss": 0.0991,
      "step": 66000
    },
    {
      "epoch": 5.8,
      "eval_accuracy": 0.9145189215928766,
      "eval_combined_score": 0.9006334285510589,
      "eval_f1": 0.886747935509241,
      "eval_loss": 0.31652504205703735,
      "eval_runtime": 103.8354,
      "eval_samples_per_second": 389.366,
      "eval_steps_per_second": 12.173,
      "step": 66000
    },
    {
      "epoch": 5.89,
      "learning_rate": 8.740071290241096e-06,
      "loss": 0.1012,
      "step": 67000
    },
    {
      "epoch": 5.89,
      "eval_accuracy": 0.9160277022013357,
      "eval_combined_score": 0.9019111310689396,
      "eval_f1": 0.8877945599365435,
      "eval_loss": 0.3151375651359558,
      "eval_runtime": 100.9327,
      "eval_samples_per_second": 400.564,
      "eval_steps_per_second": 12.523,
      "step": 67000
    }
  ],
  "max_steps": 113710,
  "num_train_epochs": 10,
  "total_flos": 1.410189745636608e+17,
  "trial_name": null,
  "trial_params": null
}