{
  "best_metric": 0.9545454382896423,
  "best_model_checkpoint": "/kaggle/working/ckpts/checkpoint-2453",
  "epoch": 14.0,
  "eval_steps": 500,
  "global_step": 3122,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.45,
      "grad_norm": 2.3771965503692627,
      "learning_rate": 9.701046337817639e-06,
      "loss": 1.5285,
      "step": 100
    },
    {
      "epoch": 0.9,
      "grad_norm": 2.858564615249634,
      "learning_rate": 9.402092675635277e-06,
      "loss": 1.1628,
      "step": 200
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7727272510528564,
      "eval_loss": 0.712559700012207,
      "eval_runtime": 9.2054,
      "eval_samples_per_second": 21.509,
      "eval_steps_per_second": 5.432,
      "step": 223
    },
    {
      "epoch": 1.35,
      "grad_norm": 12.418754577636719,
      "learning_rate": 9.106128550074738e-06,
      "loss": 0.8286,
      "step": 300
    },
    {
      "epoch": 1.79,
      "grad_norm": 7.768007755279541,
      "learning_rate": 8.807174887892378e-06,
      "loss": 0.6562,
      "step": 400
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8484848737716675,
      "eval_loss": 0.5068599581718445,
      "eval_runtime": 9.5278,
      "eval_samples_per_second": 20.781,
      "eval_steps_per_second": 5.248,
      "step": 446
    },
    {
      "epoch": 2.24,
      "grad_norm": 9.381482124328613,
      "learning_rate": 8.51121076233184e-06,
      "loss": 0.5053,
      "step": 500
    },
    {
      "epoch": 2.69,
      "grad_norm": 2.293752908706665,
      "learning_rate": 8.212257100149478e-06,
      "loss": 0.4199,
      "step": 600
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8989899158477783,
      "eval_loss": 0.356963574886322,
      "eval_runtime": 9.2472,
      "eval_samples_per_second": 21.412,
      "eval_steps_per_second": 5.407,
      "step": 669
    },
    {
      "epoch": 3.14,
      "grad_norm": 23.293209075927734,
      "learning_rate": 7.916292974588939e-06,
      "loss": 0.3121,
      "step": 700
    },
    {
      "epoch": 3.59,
      "grad_norm": 3.8754687309265137,
      "learning_rate": 7.617339312406578e-06,
      "loss": 0.325,
      "step": 800
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.939393937587738,
      "eval_loss": 0.20920716226100922,
      "eval_runtime": 9.2568,
      "eval_samples_per_second": 21.39,
      "eval_steps_per_second": 5.401,
      "step": 892
    },
    {
      "epoch": 4.04,
      "grad_norm": 62.81392288208008,
      "learning_rate": 7.318385650224216e-06,
      "loss": 0.2896,
      "step": 900
    },
    {
      "epoch": 4.48,
      "grad_norm": 35.08163833618164,
      "learning_rate": 7.019431988041854e-06,
      "loss": 0.2535,
      "step": 1000
    },
    {
      "epoch": 4.93,
      "grad_norm": 14.269490242004395,
      "learning_rate": 6.720478325859492e-06,
      "loss": 0.2217,
      "step": 1100
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9444444179534912,
      "eval_loss": 0.23924072086811066,
      "eval_runtime": 9.2044,
      "eval_samples_per_second": 21.511,
      "eval_steps_per_second": 5.432,
      "step": 1115
    },
    {
      "epoch": 5.38,
      "grad_norm": 0.41719043254852295,
      "learning_rate": 6.421524663677131e-06,
      "loss": 0.2165,
      "step": 1200
    },
    {
      "epoch": 5.83,
      "grad_norm": 1.484471321105957,
      "learning_rate": 6.1225710014947695e-06,
      "loss": 0.1831,
      "step": 1300
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9292929172515869,
      "eval_loss": 0.27538299560546875,
      "eval_runtime": 9.1435,
      "eval_samples_per_second": 21.655,
      "eval_steps_per_second": 5.468,
      "step": 1338
    },
    {
      "epoch": 6.28,
      "grad_norm": 0.09743738174438477,
      "learning_rate": 5.823617339312408e-06,
      "loss": 0.2059,
      "step": 1400
    },
    {
      "epoch": 6.73,
      "grad_norm": 0.3065042793750763,
      "learning_rate": 5.524663677130046e-06,
      "loss": 0.1598,
      "step": 1500
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9343434572219849,
      "eval_loss": 0.3294394910335541,
      "eval_runtime": 9.1064,
      "eval_samples_per_second": 21.743,
      "eval_steps_per_second": 5.491,
      "step": 1561
    },
    {
      "epoch": 7.17,
      "grad_norm": 0.05342373996973038,
      "learning_rate": 5.228699551569507e-06,
      "loss": 0.1455,
      "step": 1600
    },
    {
      "epoch": 7.62,
      "grad_norm": 1.5460679531097412,
      "learning_rate": 4.929745889387145e-06,
      "loss": 0.1676,
      "step": 1700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9494949579238892,
      "eval_loss": 0.2668905258178711,
      "eval_runtime": 9.2118,
      "eval_samples_per_second": 21.494,
      "eval_steps_per_second": 5.428,
      "step": 1784
    },
    {
      "epoch": 8.07,
      "grad_norm": 17.537992477416992,
      "learning_rate": 4.630792227204783e-06,
      "loss": 0.1762,
      "step": 1800
    },
    {
      "epoch": 8.52,
      "grad_norm": 0.20349286496639252,
      "learning_rate": 4.3318385650224224e-06,
      "loss": 0.1566,
      "step": 1900
    },
    {
      "epoch": 8.97,
      "grad_norm": 15.300110816955566,
      "learning_rate": 4.03288490284006e-06,
      "loss": 0.1597,
      "step": 2000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9292929172515869,
      "eval_loss": 0.34383586049079895,
      "eval_runtime": 9.179,
      "eval_samples_per_second": 21.571,
      "eval_steps_per_second": 5.447,
      "step": 2007
    },
    {
      "epoch": 9.42,
      "grad_norm": 0.4512959420681,
      "learning_rate": 3.7339312406576984e-06,
      "loss": 0.1416,
      "step": 2100
    },
    {
      "epoch": 9.87,
      "grad_norm": 0.7455862760543823,
      "learning_rate": 3.4349775784753366e-06,
      "loss": 0.1132,
      "step": 2200
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9444444179534912,
      "eval_loss": 0.31586208939552307,
      "eval_runtime": 9.1631,
      "eval_samples_per_second": 21.608,
      "eval_steps_per_second": 5.457,
      "step": 2230
    },
    {
      "epoch": 10.31,
      "grad_norm": 0.25966259837150574,
      "learning_rate": 3.136023916292975e-06,
      "loss": 0.1654,
      "step": 2300
    },
    {
      "epoch": 10.76,
      "grad_norm": 0.45347365736961365,
      "learning_rate": 2.8370702541106134e-06,
      "loss": 0.1224,
      "step": 2400
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9545454382896423,
      "eval_loss": 0.29796990752220154,
      "eval_runtime": 9.1354,
      "eval_samples_per_second": 21.674,
      "eval_steps_per_second": 5.473,
      "step": 2453
    },
    {
      "epoch": 11.21,
      "grad_norm": 27.043094635009766,
      "learning_rate": 2.538116591928251e-06,
      "loss": 0.1021,
      "step": 2500
    },
    {
      "epoch": 11.66,
      "grad_norm": 72.37726593017578,
      "learning_rate": 2.2391629297458894e-06,
      "loss": 0.095,
      "step": 2600
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9444444179534912,
      "eval_loss": 0.2970119118690491,
      "eval_runtime": 9.1388,
      "eval_samples_per_second": 21.666,
      "eval_steps_per_second": 5.471,
      "step": 2676
    },
    {
      "epoch": 12.11,
      "grad_norm": 0.6068007946014404,
      "learning_rate": 1.940209267563528e-06,
      "loss": 0.1307,
      "step": 2700
    },
    {
      "epoch": 12.56,
      "grad_norm": 4.567564964294434,
      "learning_rate": 1.641255605381166e-06,
      "loss": 0.1087,
      "step": 2800
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.9343434572219849,
      "eval_loss": 0.34486597776412964,
      "eval_runtime": 9.3094,
      "eval_samples_per_second": 21.269,
      "eval_steps_per_second": 5.371,
      "step": 2899
    },
    {
      "epoch": 13.0,
      "grad_norm": 41.62958908081055,
      "learning_rate": 1.3423019431988044e-06,
      "loss": 0.0917,
      "step": 2900
    },
    {
      "epoch": 13.45,
      "grad_norm": 0.026164406910538673,
      "learning_rate": 1.0433482810164425e-06,
      "loss": 0.0904,
      "step": 3000
    },
    {
      "epoch": 13.9,
      "grad_norm": 52.47389221191406,
      "learning_rate": 7.443946188340807e-07,
      "loss": 0.1254,
      "step": 3100
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.9444444179534912,
      "eval_loss": 0.31978654861450195,
      "eval_runtime": 9.2595,
      "eval_samples_per_second": 21.384,
      "eval_steps_per_second": 5.4,
      "step": 3122
    },
    {
      "epoch": 14.0,
      "step": 3122,
      "total_flos": 7.9842219974856e+17,
      "train_loss": 0.30023268478028214,
      "train_runtime": 1742.0026,
      "train_samples_per_second": 15.336,
      "train_steps_per_second": 1.92
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.9545454382896423,
      "eval_loss": 0.29796990752220154,
      "eval_runtime": 8.9823,
      "eval_samples_per_second": 22.043,
      "eval_steps_per_second": 5.566,
      "step": 3122
    }
  ],
  "logging_steps": 100,
  "max_steps": 3345,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "total_flos": 7.9842219974856e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}