{
  "best_metric": 0.955902004454343,
  "best_model_checkpoint": "./nlu_finetuned_models/sst2/roberta-large_lr1e-05/checkpoint-26523",
  "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": 14.684767723083496,
      "learning_rate": 2.198768689533861e-06,
      "loss": 0.6452,
      "step": 500
    },
    {
      "epoch": 0.26392187912377935,
      "grad_norm": 6.864231109619141,
      "learning_rate": 4.397537379067722e-06,
      "loss": 0.2856,
      "step": 1000
    },
    {
      "epoch": 0.39588281868566905,
      "grad_norm": 19.211210250854492,
      "learning_rate": 6.596306068601583e-06,
      "loss": 0.2445,
      "step": 1500
    },
    {
      "epoch": 0.5278437582475587,
      "grad_norm": 21.7692813873291,
      "learning_rate": 8.795074758135444e-06,
      "loss": 0.2459,
      "step": 2000
    },
    {
      "epoch": 0.6598046978094484,
      "grad_norm": 1.4644837379455566,
      "learning_rate": 9.936545372866128e-06,
      "loss": 0.2531,
      "step": 2500
    },
    {
      "epoch": 0.7917656373713381,
      "grad_norm": 18.3651065826416,
      "learning_rate": 9.796159029649596e-06,
      "loss": 0.2392,
      "step": 3000
    },
    {
      "epoch": 0.9237265769332278,
      "grad_norm": 9.527914047241211,
      "learning_rate": 9.655772686433064e-06,
      "loss": 0.2195,
      "step": 3500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9458054936896808,
      "eval_loss": 0.2150438129901886,
      "eval_runtime": 7.1689,
      "eval_samples_per_second": 939.47,
      "eval_steps_per_second": 58.726,
      "step": 3789
    },
    {
      "epoch": 1.0556875164951174,
      "grad_norm": 67.29890441894531,
      "learning_rate": 9.515386343216533e-06,
      "loss": 0.2065,
      "step": 4000
    },
    {
      "epoch": 1.187648456057007,
      "grad_norm": 3.1265335083007812,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.1957,
      "step": 4500
    },
    {
      "epoch": 1.3196093956188968,
      "grad_norm": 2.3976995944976807,
      "learning_rate": 9.234613656783469e-06,
      "loss": 0.1817,
      "step": 5000
    },
    {
      "epoch": 1.4515703351807865,
      "grad_norm": 0.1834949404001236,
      "learning_rate": 9.094227313566937e-06,
      "loss": 0.1855,
      "step": 5500
    },
    {
      "epoch": 1.5835312747426762,
      "grad_norm": 48.085628509521484,
      "learning_rate": 8.953840970350406e-06,
      "loss": 0.1865,
      "step": 6000
    },
    {
      "epoch": 1.7154922143045659,
      "grad_norm": 0.10328803956508636,
      "learning_rate": 8.813454627133872e-06,
      "loss": 0.1766,
      "step": 6500
    },
    {
      "epoch": 1.8474531538664554,
      "grad_norm": 6.290451526641846,
      "learning_rate": 8.67306828391734e-06,
      "loss": 0.1789,
      "step": 7000
    },
    {
      "epoch": 1.9794140934283453,
      "grad_norm": 0.1234852746129036,
      "learning_rate": 8.532681940700809e-06,
      "loss": 0.1849,
      "step": 7500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.948626577579807,
      "eval_loss": 0.16994288563728333,
      "eval_runtime": 7.3255,
      "eval_samples_per_second": 919.392,
      "eval_steps_per_second": 57.471,
      "step": 7578
    },
    {
      "epoch": 2.1113750329902348,
      "grad_norm": 0.3948022723197937,
      "learning_rate": 8.392295597484278e-06,
      "loss": 0.1325,
      "step": 8000
    },
    {
      "epoch": 2.2433359725521247,
      "grad_norm": 0.054762326180934906,
      "learning_rate": 8.251909254267747e-06,
      "loss": 0.1393,
      "step": 8500
    },
    {
      "epoch": 2.375296912114014,
      "grad_norm": 0.13094937801361084,
      "learning_rate": 8.111522911051213e-06,
      "loss": 0.131,
      "step": 9000
    },
    {
      "epoch": 2.507257851675904,
      "grad_norm": 1.9226784706115723,
      "learning_rate": 7.971136567834682e-06,
      "loss": 0.1454,
      "step": 9500
    },
    {
      "epoch": 2.6392187912377936,
      "grad_norm": 6.964934349060059,
      "learning_rate": 7.83075022461815e-06,
      "loss": 0.1458,
      "step": 10000
    },
    {
      "epoch": 2.771179730799683,
      "grad_norm": 12.109170913696289,
      "learning_rate": 7.690363881401618e-06,
      "loss": 0.1413,
      "step": 10500
    },
    {
      "epoch": 2.903140670361573,
      "grad_norm": 6.971311569213867,
      "learning_rate": 7.549977538185086e-06,
      "loss": 0.134,
      "step": 11000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9556050482553823,
      "eval_loss": 0.20318306982517242,
      "eval_runtime": 7.253,
      "eval_samples_per_second": 928.576,
      "eval_steps_per_second": 58.045,
      "step": 11367
    },
    {
      "epoch": 3.0351016099234625,
      "grad_norm": 58.1721076965332,
      "learning_rate": 7.409591194968554e-06,
      "loss": 0.1312,
      "step": 11500
    },
    {
      "epoch": 3.1670625494853524,
      "grad_norm": 0.04744189232587814,
      "learning_rate": 7.269204851752022e-06,
      "loss": 0.1035,
      "step": 12000
    },
    {
      "epoch": 3.299023489047242,
      "grad_norm": 0.29440391063690186,
      "learning_rate": 7.12881850853549e-06,
      "loss": 0.0867,
      "step": 12500
    },
    {
      "epoch": 3.4309844286091318,
      "grad_norm": 14.616955757141113,
      "learning_rate": 6.988432165318958e-06,
      "loss": 0.105,
      "step": 13000
    },
    {
      "epoch": 3.5629453681710213,
      "grad_norm": 2.5657553672790527,
      "learning_rate": 6.848045822102426e-06,
      "loss": 0.1108,
      "step": 13500
    },
    {
      "epoch": 3.694906307732911,
      "grad_norm": 0.30954834818840027,
      "learning_rate": 6.707659478885895e-06,
      "loss": 0.1045,
      "step": 14000
    },
    {
      "epoch": 3.8268672472948007,
      "grad_norm": 0.18087419867515564,
      "learning_rate": 6.567273135669363e-06,
      "loss": 0.1003,
      "step": 14500
    },
    {
      "epoch": 3.9588281868566906,
      "grad_norm": 0.2143256664276123,
      "learning_rate": 6.426886792452831e-06,
      "loss": 0.1015,
      "step": 15000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9553080920564216,
      "eval_loss": 0.2478303164243698,
      "eval_runtime": 7.168,
      "eval_samples_per_second": 939.587,
      "eval_steps_per_second": 58.733,
      "step": 15156
    },
    {
      "epoch": 4.0907891264185805,
      "grad_norm": 0.6192390322685242,
      "learning_rate": 6.2865004492362994e-06,
      "loss": 0.0898,
      "step": 15500
    },
    {
      "epoch": 4.2227500659804695,
      "grad_norm": 7.8285417556762695,
      "learning_rate": 6.146114106019767e-06,
      "loss": 0.0786,
      "step": 16000
    },
    {
      "epoch": 4.3547110055423595,
      "grad_norm": 3.172856330871582,
      "learning_rate": 6.005727762803235e-06,
      "loss": 0.073,
      "step": 16500
    },
    {
      "epoch": 4.486671945104249,
      "grad_norm": 0.03749631717801094,
      "learning_rate": 5.865341419586703e-06,
      "loss": 0.0808,
      "step": 17000
    },
    {
      "epoch": 4.618632884666139,
      "grad_norm": 0.01692149229347706,
      "learning_rate": 5.7249550763701715e-06,
      "loss": 0.0789,
      "step": 17500
    },
    {
      "epoch": 4.750593824228028,
      "grad_norm": 0.09162786602973938,
      "learning_rate": 5.584568733153639e-06,
      "loss": 0.0977,
      "step": 18000
    },
    {
      "epoch": 4.882554763789918,
      "grad_norm": 0.0481809638440609,
      "learning_rate": 5.444182389937107e-06,
      "loss": 0.0823,
      "step": 18500
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9535263548626578,
      "eval_loss": 0.24863098561763763,
      "eval_runtime": 7.223,
      "eval_samples_per_second": 932.441,
      "eval_steps_per_second": 58.286,
      "step": 18945
    },
    {
      "epoch": 5.014515703351808,
      "grad_norm": 0.07389198243618011,
      "learning_rate": 5.303796046720575e-06,
      "loss": 0.0747,
      "step": 19000
    },
    {
      "epoch": 5.146476642913697,
      "grad_norm": 0.06816897541284561,
      "learning_rate": 5.163409703504043e-06,
      "loss": 0.0512,
      "step": 19500
    },
    {
      "epoch": 5.278437582475587,
      "grad_norm": 0.15339165925979614,
      "learning_rate": 5.023023360287511e-06,
      "loss": 0.0599,
      "step": 20000
    },
    {
      "epoch": 5.410398522037477,
      "grad_norm": 65.28050994873047,
      "learning_rate": 4.88263701707098e-06,
      "loss": 0.0733,
      "step": 20500
    },
    {
      "epoch": 5.542359461599366,
      "grad_norm": 6.615115642547607,
      "learning_rate": 4.7422506738544475e-06,
      "loss": 0.0625,
      "step": 21000
    },
    {
      "epoch": 5.674320401161256,
      "grad_norm": 40.05002212524414,
      "learning_rate": 4.601864330637916e-06,
      "loss": 0.0667,
      "step": 21500
    },
    {
      "epoch": 5.806281340723146,
      "grad_norm": 0.10902038961648941,
      "learning_rate": 4.461477987421384e-06,
      "loss": 0.0623,
      "step": 22000
    },
    {
      "epoch": 5.938242280285036,
      "grad_norm": 0.029639070853590965,
      "learning_rate": 4.321091644204852e-06,
      "loss": 0.062,
      "step": 22500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.955011135857461,
      "eval_loss": 0.31579306721687317,
      "eval_runtime": 7.3084,
      "eval_samples_per_second": 921.548,
      "eval_steps_per_second": 57.605,
      "step": 22734
    },
    {
      "epoch": 6.070203219846925,
      "grad_norm": 0.03687451034784317,
      "learning_rate": 4.18070530098832e-06,
      "loss": 0.0539,
      "step": 23000
    },
    {
      "epoch": 6.202164159408815,
      "grad_norm": 0.11739882826805115,
      "learning_rate": 4.040318957771789e-06,
      "loss": 0.0446,
      "step": 23500
    },
    {
      "epoch": 6.334125098970705,
      "grad_norm": 0.21322381496429443,
      "learning_rate": 3.899932614555256e-06,
      "loss": 0.0528,
      "step": 24000
    },
    {
      "epoch": 6.466086038532595,
      "grad_norm": 1.3615431785583496,
      "learning_rate": 3.7595462713387242e-06,
      "loss": 0.0551,
      "step": 24500
    },
    {
      "epoch": 6.598046978094484,
      "grad_norm": 0.005917022004723549,
      "learning_rate": 3.6191599281221925e-06,
      "loss": 0.0423,
      "step": 25000
    },
    {
      "epoch": 6.730007917656374,
      "grad_norm": 0.0038304822519421577,
      "learning_rate": 3.4787735849056607e-06,
      "loss": 0.0445,
      "step": 25500
    },
    {
      "epoch": 6.8619688572182636,
      "grad_norm": 0.23638126254081726,
      "learning_rate": 3.338387241689129e-06,
      "loss": 0.056,
      "step": 26000
    },
    {
      "epoch": 6.993929796780153,
      "grad_norm": 0.015073439106345177,
      "learning_rate": 3.1980008984725967e-06,
      "loss": 0.0547,
      "step": 26500
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.955902004454343,
      "eval_loss": 0.30895689129829407,
      "eval_runtime": 7.2906,
      "eval_samples_per_second": 923.787,
      "eval_steps_per_second": 57.745,
      "step": 26523
    },
    {
      "epoch": 7.1258907363420425,
      "grad_norm": 0.04011049121618271,
      "learning_rate": 3.057614555256065e-06,
      "loss": 0.0366,
      "step": 27000
    },
    {
      "epoch": 7.257851675903932,
      "grad_norm": 0.04726456105709076,
      "learning_rate": 2.9172282120395328e-06,
      "loss": 0.0283,
      "step": 27500
    },
    {
      "epoch": 7.389812615465822,
      "grad_norm": 0.0173865407705307,
      "learning_rate": 2.776841868823001e-06,
      "loss": 0.0353,
      "step": 28000
    },
    {
      "epoch": 7.521773555027712,
      "grad_norm": 0.007793079596012831,
      "learning_rate": 2.6364555256064692e-06,
      "loss": 0.035,
      "step": 28500
    },
    {
      "epoch": 7.653734494589601,
      "grad_norm": 0.014837320894002914,
      "learning_rate": 2.496069182389937e-06,
      "loss": 0.0378,
      "step": 29000
    },
    {
      "epoch": 7.785695434151491,
      "grad_norm": 0.007654995657503605,
      "learning_rate": 2.3556828391734053e-06,
      "loss": 0.0425,
      "step": 29500
    },
    {
      "epoch": 7.917656373713381,
      "grad_norm": 0.2803705036640167,
      "learning_rate": 2.2152964959568735e-06,
      "loss": 0.039,
      "step": 30000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9536748329621381,
      "eval_loss": 0.3630528151988983,
      "eval_runtime": 7.1993,
      "eval_samples_per_second": 935.506,
      "eval_steps_per_second": 58.478,
      "step": 30312
    },
    {
      "epoch": 8.04961731327527,
      "grad_norm": 0.04258790612220764,
      "learning_rate": 2.0749101527403413e-06,
      "loss": 0.0369,
      "step": 30500
    },
    {
      "epoch": 8.181578252837161,
      "grad_norm": 0.008549284189939499,
      "learning_rate": 1.9345238095238096e-06,
      "loss": 0.0185,
      "step": 31000
    },
    {
      "epoch": 8.31353919239905,
      "grad_norm": 0.002880709245800972,
      "learning_rate": 1.7941374663072778e-06,
      "loss": 0.03,
      "step": 31500
    },
    {
      "epoch": 8.445500131960939,
      "grad_norm": 0.007983551360666752,
      "learning_rate": 1.6537511230907458e-06,
      "loss": 0.0264,
      "step": 32000
    },
    {
      "epoch": 8.57746107152283,
      "grad_norm": 0.005722792819142342,
      "learning_rate": 1.513364779874214e-06,
      "loss": 0.0354,
      "step": 32500
    },
    {
      "epoch": 8.709422011084719,
      "grad_norm": 0.006476383190602064,
      "learning_rate": 1.372978436657682e-06,
      "loss": 0.0292,
      "step": 33000
    },
    {
      "epoch": 8.841382950646608,
      "grad_norm": 0.029401870444417,
      "learning_rate": 1.2325920934411503e-06,
      "loss": 0.027,
      "step": 33500
    },
    {
      "epoch": 8.973343890208499,
      "grad_norm": 14.256837844848633,
      "learning_rate": 1.092205750224618e-06,
      "loss": 0.0234,
      "step": 34000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9542687453600593,
      "eval_loss": 0.38435572385787964,
      "eval_runtime": 7.155,
      "eval_samples_per_second": 941.299,
      "eval_steps_per_second": 58.84,
      "step": 34101
    },
    {
      "epoch": 9.105304829770388,
      "grad_norm": 0.0036281670909374952,
      "learning_rate": 9.518194070080863e-07,
      "loss": 0.017,
      "step": 34500
    },
    {
      "epoch": 9.237265769332277,
      "grad_norm": 0.009701136499643326,
      "learning_rate": 8.114330637915545e-07,
      "loss": 0.0247,
      "step": 35000
    },
    {
      "epoch": 9.369226708894168,
      "grad_norm": 0.07909628003835678,
      "learning_rate": 6.710467205750225e-07,
      "loss": 0.0197,
      "step": 35500
    },
    {
      "epoch": 9.501187648456057,
      "grad_norm": 0.00657819677144289,
      "learning_rate": 5.306603773584906e-07,
      "loss": 0.0176,
      "step": 36000
    },
    {
      "epoch": 9.633148588017947,
      "grad_norm": 0.009074714966118336,
      "learning_rate": 3.902740341419587e-07,
      "loss": 0.0241,
      "step": 36500
    },
    {
      "epoch": 9.765109527579837,
      "grad_norm": 0.026519833132624626,
      "learning_rate": 2.4988769092542675e-07,
      "loss": 0.0208,
      "step": 37000
    },
    {
      "epoch": 9.897070467141726,
      "grad_norm": 0.001519062090665102,
      "learning_rate": 1.095013477088949e-07,
      "loss": 0.0163,
      "step": 37500
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9535263548626578,
      "eval_loss": 0.3952884376049042,
      "eval_runtime": 7.2818,
      "eval_samples_per_second": 924.908,
      "eval_steps_per_second": 57.815,
      "step": 37890
    },
    {
      "epoch": 10.0,
      "step": 37890,
      "total_flos": 3.948673638307961e+16,
      "train_loss": 0.10002695248730424,
      "train_runtime": 3896.4569,
      "train_samples_per_second": 155.562,
      "train_steps_per_second": 9.724
    }
  ],
  "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": 3.948673638307961e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}