{
  "best_metric": 0.9517446176688938,
  "best_model_checkpoint": "./nlu_finetuned_models/sst2/roberta-base_lr1e-05/checkpoint-18945",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 37890,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.13196093956188967,
      "grad_norm": 16.203157424926758,
      "learning_rate": 2.198768689533861e-06,
      "loss": 0.6535,
      "step": 500
    },
    {
      "epoch": 0.26392187912377935,
      "grad_norm": 39.97472381591797,
      "learning_rate": 4.397537379067722e-06,
      "loss": 0.3243,
      "step": 1000
    },
    {
      "epoch": 0.39588281868566905,
      "grad_norm": 32.75914001464844,
      "learning_rate": 6.596306068601583e-06,
      "loss": 0.2817,
      "step": 1500
    },
    {
      "epoch": 0.5278437582475587,
      "grad_norm": 18.060081481933594,
      "learning_rate": 8.795074758135444e-06,
      "loss": 0.2591,
      "step": 2000
    },
    {
      "epoch": 0.6598046978094484,
      "grad_norm": 6.257637977600098,
      "learning_rate": 9.936545372866128e-06,
      "loss": 0.2599,
      "step": 2500
    },
    {
      "epoch": 0.7917656373713381,
      "grad_norm": 27.890222549438477,
      "learning_rate": 9.796159029649596e-06,
      "loss": 0.2496,
      "step": 3000
    },
    {
      "epoch": 0.9237265769332278,
      "grad_norm": 29.086610794067383,
      "learning_rate": 9.655772686433064e-06,
      "loss": 0.244,
      "step": 3500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9365998515219005,
      "eval_loss": 0.19152739644050598,
      "eval_runtime": 3.2031,
      "eval_samples_per_second": 2102.672,
      "eval_steps_per_second": 131.436,
      "step": 3789
    },
    {
      "epoch": 1.0556875164951174,
      "grad_norm": 32.525482177734375,
      "learning_rate": 9.515386343216533e-06,
      "loss": 0.2199,
      "step": 4000
    },
    {
      "epoch": 1.187648456057007,
      "grad_norm": 39.73225402832031,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.2081,
      "step": 4500
    },
    {
      "epoch": 1.3196093956188968,
      "grad_norm": 18.29433250427246,
      "learning_rate": 9.234613656783469e-06,
      "loss": 0.1982,
      "step": 5000
    },
    {
      "epoch": 1.4515703351807865,
      "grad_norm": 0.6866791248321533,
      "learning_rate": 9.094227313566937e-06,
      "loss": 0.2094,
      "step": 5500
    },
    {
      "epoch": 1.5835312747426762,
      "grad_norm": 2.060551166534424,
      "learning_rate": 8.953840970350406e-06,
      "loss": 0.2017,
      "step": 6000
    },
    {
      "epoch": 1.7154922143045659,
      "grad_norm": 26.714765548706055,
      "learning_rate": 8.813454627133872e-06,
      "loss": 0.1891,
      "step": 6500
    },
    {
      "epoch": 1.8474531538664554,
      "grad_norm": 6.763668060302734,
      "learning_rate": 8.67306828391734e-06,
      "loss": 0.2028,
      "step": 7000
    },
    {
      "epoch": 1.9794140934283453,
      "grad_norm": 0.7139686346054077,
      "learning_rate": 8.532681940700809e-06,
      "loss": 0.1984,
      "step": 7500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9406087602078693,
      "eval_loss": 0.22181583940982819,
      "eval_runtime": 3.2002,
      "eval_samples_per_second": 2104.543,
      "eval_steps_per_second": 131.553,
      "step": 7578
    },
    {
      "epoch": 2.1113750329902348,
      "grad_norm": 1.851331353187561,
      "learning_rate": 8.392295597484278e-06,
      "loss": 0.1615,
      "step": 8000
    },
    {
      "epoch": 2.2433359725521247,
      "grad_norm": 0.1076214462518692,
      "learning_rate": 8.251909254267747e-06,
      "loss": 0.1622,
      "step": 8500
    },
    {
      "epoch": 2.375296912114014,
      "grad_norm": 3.893419027328491,
      "learning_rate": 8.111522911051213e-06,
      "loss": 0.1506,
      "step": 9000
    },
    {
      "epoch": 2.507257851675904,
      "grad_norm": 0.19271089136600494,
      "learning_rate": 7.971136567834682e-06,
      "loss": 0.1713,
      "step": 9500
    },
    {
      "epoch": 2.6392187912377936,
      "grad_norm": 0.5250680446624756,
      "learning_rate": 7.83075022461815e-06,
      "loss": 0.1591,
      "step": 10000
    },
    {
      "epoch": 2.771179730799683,
      "grad_norm": 72.66580963134766,
      "learning_rate": 7.690363881401618e-06,
      "loss": 0.1688,
      "step": 10500
    },
    {
      "epoch": 2.903140670361573,
      "grad_norm": 12.931042671203613,
      "learning_rate": 7.549977538185086e-06,
      "loss": 0.1595,
      "step": 11000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9492204899777282,
      "eval_loss": 0.22260741889476776,
      "eval_runtime": 3.2317,
      "eval_samples_per_second": 2084.074,
      "eval_steps_per_second": 130.274,
      "step": 11367
    },
    {
      "epoch": 3.0351016099234625,
      "grad_norm": 0.4131239056587219,
      "learning_rate": 7.409591194968554e-06,
      "loss": 0.142,
      "step": 11500
    },
    {
      "epoch": 3.1670625494853524,
      "grad_norm": 18.098291397094727,
      "learning_rate": 7.269204851752022e-06,
      "loss": 0.1164,
      "step": 12000
    },
    {
      "epoch": 3.299023489047242,
      "grad_norm": 0.05766018107533455,
      "learning_rate": 7.12881850853549e-06,
      "loss": 0.1209,
      "step": 12500
    },
    {
      "epoch": 3.4309844286091318,
      "grad_norm": 11.65517807006836,
      "learning_rate": 6.988432165318958e-06,
      "loss": 0.131,
      "step": 13000
    },
    {
      "epoch": 3.5629453681710213,
      "grad_norm": 104.57723999023438,
      "learning_rate": 6.848045822102426e-06,
      "loss": 0.1415,
      "step": 13500
    },
    {
      "epoch": 3.694906307732911,
      "grad_norm": 0.9235928654670715,
      "learning_rate": 6.707659478885895e-06,
      "loss": 0.1359,
      "step": 14000
    },
    {
      "epoch": 3.8268672472948007,
      "grad_norm": 32.136817932128906,
      "learning_rate": 6.567273135669363e-06,
      "loss": 0.1356,
      "step": 14500
    },
    {
      "epoch": 3.9588281868566906,
      "grad_norm": 8.762619018554688,
      "learning_rate": 6.426886792452831e-06,
      "loss": 0.1225,
      "step": 15000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9501113585746103,
      "eval_loss": 0.24848563969135284,
      "eval_runtime": 3.2316,
      "eval_samples_per_second": 2084.138,
      "eval_steps_per_second": 130.278,
      "step": 15156
    },
    {
      "epoch": 4.0907891264185805,
      "grad_norm": 47.4286994934082,
      "learning_rate": 6.2865004492362994e-06,
      "loss": 0.1143,
      "step": 15500
    },
    {
      "epoch": 4.2227500659804695,
      "grad_norm": 17.368412017822266,
      "learning_rate": 6.146114106019767e-06,
      "loss": 0.1018,
      "step": 16000
    },
    {
      "epoch": 4.3547110055423595,
      "grad_norm": 8.772218704223633,
      "learning_rate": 6.005727762803235e-06,
      "loss": 0.1123,
      "step": 16500
    },
    {
      "epoch": 4.486671945104249,
      "grad_norm": 21.83694839477539,
      "learning_rate": 5.865341419586703e-06,
      "loss": 0.1039,
      "step": 17000
    },
    {
      "epoch": 4.618632884666139,
      "grad_norm": 24.351490020751953,
      "learning_rate": 5.7249550763701715e-06,
      "loss": 0.1156,
      "step": 17500
    },
    {
      "epoch": 4.750593824228028,
      "grad_norm": 0.13713432848453522,
      "learning_rate": 5.584568733153639e-06,
      "loss": 0.1225,
      "step": 18000
    },
    {
      "epoch": 4.882554763789918,
      "grad_norm": 0.23665529489517212,
      "learning_rate": 5.444182389937107e-06,
      "loss": 0.1061,
      "step": 18500
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9517446176688938,
      "eval_loss": 0.2689071297645569,
      "eval_runtime": 3.2131,
      "eval_samples_per_second": 2096.114,
      "eval_steps_per_second": 131.027,
      "step": 18945
    },
    {
      "epoch": 5.014515703351808,
      "grad_norm": 0.020952098071575165,
      "learning_rate": 5.303796046720575e-06,
      "loss": 0.0945,
      "step": 19000
    },
    {
      "epoch": 5.146476642913697,
      "grad_norm": 0.08160785585641861,
      "learning_rate": 5.163409703504043e-06,
      "loss": 0.0741,
      "step": 19500
    },
    {
      "epoch": 5.278437582475587,
      "grad_norm": 3.612468957901001,
      "learning_rate": 5.023023360287511e-06,
      "loss": 0.0848,
      "step": 20000
    },
    {
      "epoch": 5.410398522037477,
      "grad_norm": 0.16665929555892944,
      "learning_rate": 4.88263701707098e-06,
      "loss": 0.0856,
      "step": 20500
    },
    {
      "epoch": 5.542359461599366,
      "grad_norm": 14.307438850402832,
      "learning_rate": 4.7422506738544475e-06,
      "loss": 0.0919,
      "step": 21000
    },
    {
      "epoch": 5.674320401161256,
      "grad_norm": 81.50785827636719,
      "learning_rate": 4.601864330637916e-06,
      "loss": 0.091,
      "step": 21500
    },
    {
      "epoch": 5.806281340723146,
      "grad_norm": 0.545400857925415,
      "learning_rate": 4.461477987421384e-06,
      "loss": 0.0966,
      "step": 22000
    },
    {
      "epoch": 5.938242280285036,
      "grad_norm": 0.13066865503787994,
      "learning_rate": 4.321091644204852e-06,
      "loss": 0.0921,
      "step": 22500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9498144023756496,
      "eval_loss": 0.26830601692199707,
      "eval_runtime": 3.2279,
      "eval_samples_per_second": 2086.527,
      "eval_steps_per_second": 130.427,
      "step": 22734
    },
    {
      "epoch": 6.070203219846925,
      "grad_norm": 0.016320781782269478,
      "learning_rate": 4.18070530098832e-06,
      "loss": 0.0786,
      "step": 23000
    },
    {
      "epoch": 6.202164159408815,
      "grad_norm": 1.0383734703063965,
      "learning_rate": 4.040318957771789e-06,
      "loss": 0.0598,
      "step": 23500
    },
    {
      "epoch": 6.334125098970705,
      "grad_norm": 37.35941696166992,
      "learning_rate": 3.899932614555256e-06,
      "loss": 0.0741,
      "step": 24000
    },
    {
      "epoch": 6.466086038532595,
      "grad_norm": 0.22372713685035706,
      "learning_rate": 3.7595462713387242e-06,
      "loss": 0.0648,
      "step": 24500
    },
    {
      "epoch": 6.598046978094484,
      "grad_norm": 0.3024131953716278,
      "learning_rate": 3.6191599281221925e-06,
      "loss": 0.0742,
      "step": 25000
    },
    {
      "epoch": 6.730007917656374,
      "grad_norm": 0.5557472109794617,
      "learning_rate": 3.4787735849056607e-06,
      "loss": 0.0697,
      "step": 25500
    },
    {
      "epoch": 6.8619688572182636,
      "grad_norm": 10.842752456665039,
      "learning_rate": 3.338387241689129e-06,
      "loss": 0.0781,
      "step": 26000
    },
    {
      "epoch": 6.993929796780153,
      "grad_norm": 0.058014508336782455,
      "learning_rate": 3.1980008984725967e-06,
      "loss": 0.0749,
      "step": 26500
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9480326651818857,
      "eval_loss": 0.2953985631465912,
      "eval_runtime": 3.1998,
      "eval_samples_per_second": 2104.852,
      "eval_steps_per_second": 131.573,
      "step": 26523
    },
    {
      "epoch": 7.1258907363420425,
      "grad_norm": 0.021114373579621315,
      "learning_rate": 3.057614555256065e-06,
      "loss": 0.0606,
      "step": 27000
    },
    {
      "epoch": 7.257851675903932,
      "grad_norm": 0.09994231164455414,
      "learning_rate": 2.9172282120395328e-06,
      "loss": 0.0513,
      "step": 27500
    },
    {
      "epoch": 7.389812615465822,
      "grad_norm": 0.02594364993274212,
      "learning_rate": 2.776841868823001e-06,
      "loss": 0.0578,
      "step": 28000
    },
    {
      "epoch": 7.521773555027712,
      "grad_norm": 0.3120315968990326,
      "learning_rate": 2.6364555256064692e-06,
      "loss": 0.0622,
      "step": 28500
    },
    {
      "epoch": 7.653734494589601,
      "grad_norm": 0.04183578863739967,
      "learning_rate": 2.496069182389937e-06,
      "loss": 0.0644,
      "step": 29000
    },
    {
      "epoch": 7.785695434151491,
      "grad_norm": 3.4326181411743164,
      "learning_rate": 2.3556828391734053e-06,
      "loss": 0.0583,
      "step": 29500
    },
    {
      "epoch": 7.917656373713381,
      "grad_norm": 0.20767995715141296,
      "learning_rate": 2.2152964959568735e-06,
      "loss": 0.0625,
      "step": 30000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9484780994803267,
      "eval_loss": 0.343423455953598,
      "eval_runtime": 3.2166,
      "eval_samples_per_second": 2093.807,
      "eval_steps_per_second": 130.882,
      "step": 30312
    },
    {
      "epoch": 8.04961731327527,
      "grad_norm": 0.006022294517606497,
      "learning_rate": 2.0749101527403413e-06,
      "loss": 0.0569,
      "step": 30500
    },
    {
      "epoch": 8.181578252837161,
      "grad_norm": 2.1597390174865723,
      "learning_rate": 1.9345238095238096e-06,
      "loss": 0.044,
      "step": 31000
    },
    {
      "epoch": 8.31353919239905,
      "grad_norm": 0.019608333706855774,
      "learning_rate": 1.7941374663072778e-06,
      "loss": 0.0509,
      "step": 31500
    },
    {
      "epoch": 8.445500131960939,
      "grad_norm": 0.008767606690526009,
      "learning_rate": 1.6537511230907458e-06,
      "loss": 0.0437,
      "step": 32000
    },
    {
      "epoch": 8.57746107152283,
      "grad_norm": 0.47362828254699707,
      "learning_rate": 1.513364779874214e-06,
      "loss": 0.0505,
      "step": 32500
    },
    {
      "epoch": 8.709422011084719,
      "grad_norm": 0.08651397377252579,
      "learning_rate": 1.372978436657682e-06,
      "loss": 0.0571,
      "step": 33000
    },
    {
      "epoch": 8.841382950646608,
      "grad_norm": 0.09942079335451126,
      "learning_rate": 1.2325920934411503e-06,
      "loss": 0.0466,
      "step": 33500
    },
    {
      "epoch": 8.973343890208499,
      "grad_norm": 33.968544006347656,
      "learning_rate": 1.092205750224618e-06,
      "loss": 0.048,
      "step": 34000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.948626577579807,
      "eval_loss": 0.3509092926979065,
      "eval_runtime": 3.1604,
      "eval_samples_per_second": 2131.066,
      "eval_steps_per_second": 133.211,
      "step": 34101
    },
    {
      "epoch": 9.105304829770388,
      "grad_norm": 0.12073567509651184,
      "learning_rate": 9.518194070080863e-07,
      "loss": 0.0404,
      "step": 34500
    },
    {
      "epoch": 9.237265769332277,
      "grad_norm": 0.8697101473808289,
      "learning_rate": 8.114330637915545e-07,
      "loss": 0.0445,
      "step": 35000
    },
    {
      "epoch": 9.369226708894168,
      "grad_norm": 0.02095695585012436,
      "learning_rate": 6.710467205750225e-07,
      "loss": 0.043,
      "step": 35500
    },
    {
      "epoch": 9.501187648456057,
      "grad_norm": 0.010166825726628304,
      "learning_rate": 5.306603773584906e-07,
      "loss": 0.0441,
      "step": 36000
    },
    {
      "epoch": 9.633148588017947,
      "grad_norm": 0.10797163844108582,
      "learning_rate": 3.902740341419587e-07,
      "loss": 0.0388,
      "step": 36500
    },
    {
      "epoch": 9.765109527579837,
      "grad_norm": 0.16056068241596222,
      "learning_rate": 2.4988769092542675e-07,
      "loss": 0.034,
      "step": 37000
    },
    {
      "epoch": 9.897070467141726,
      "grad_norm": 0.35128360986709595,
      "learning_rate": 1.095013477088949e-07,
      "loss": 0.0414,
      "step": 37500
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9499628804751299,
      "eval_loss": 0.3629242777824402,
      "eval_runtime": 3.1958,
      "eval_samples_per_second": 2107.443,
      "eval_steps_per_second": 131.735,
      "step": 37890
    },
    {
      "epoch": 10.0,
      "step": 37890,
      "total_flos": 1.114824255444396e+16,
      "train_loss": 0.12249788010960426,
      "train_runtime": 1979.2422,
      "train_samples_per_second": 306.249,
      "train_steps_per_second": 19.144
    }
  ],
  "logging_steps": 500,
  "max_steps": 37890,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.114824255444396e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}