{
  "best_metric": 0.03547798842191696,
  "best_model_checkpoint": "logs/google-bert/bert-large-cased_accuracy-coverage/checkpoint-1737",
  "epoch": 12.0,
  "eval_steps": 500,
  "global_step": 2316,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 18.177562713623047,
      "learning_rate": 1.900518134715026e-05,
      "loss": 0.0796,
      "step": 193
    },
    {
      "epoch": 1.0,
      "eval_MAE": 0.16079820692539215,
      "eval_R2": 0.09031999984589933,
      "eval_RMSE": 0.20937980711460114,
      "eval_loss": 0.04383990541100502,
      "eval_runtime": 14.15,
      "eval_samples_per_second": 203.604,
      "eval_steps_per_second": 3.463,
      "step": 193
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.8847270011901855,
      "learning_rate": 1.800518134715026e-05,
      "loss": 0.0589,
      "step": 386
    },
    {
      "epoch": 2.0,
      "eval_MAE": 0.16792413592338562,
      "eval_R2": -0.035297909277749895,
      "eval_RMSE": 0.22336912155151367,
      "eval_loss": 0.04989376664161682,
      "eval_runtime": 14.0183,
      "eval_samples_per_second": 205.517,
      "eval_steps_per_second": 3.495,
      "step": 386
    },
    {
      "epoch": 3.0,
      "grad_norm": 7.614197731018066,
      "learning_rate": 1.7005181347150262e-05,
      "loss": 0.051,
      "step": 579
    },
    {
      "epoch": 3.0,
      "eval_MAE": 0.18675751984119415,
      "eval_R2": -0.032218313600698156,
      "eval_RMSE": 0.22303664684295654,
      "eval_loss": 0.0497453473508358,
      "eval_runtime": 14.1082,
      "eval_samples_per_second": 204.208,
      "eval_steps_per_second": 3.473,
      "step": 579
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.26663488149642944,
      "learning_rate": 1.6005181347150262e-05,
      "loss": 0.0368,
      "step": 772
    },
    {
      "epoch": 4.0,
      "eval_MAE": 0.14720453321933746,
      "eval_R2": 0.17138686484945131,
      "eval_RMSE": 0.19983261823654175,
      "eval_loss": 0.03993307799100876,
      "eval_runtime": 13.9915,
      "eval_samples_per_second": 205.91,
      "eval_steps_per_second": 3.502,
      "step": 772
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.195333480834961,
      "learning_rate": 1.500518134715026e-05,
      "loss": 0.0305,
      "step": 965
    },
    {
      "epoch": 5.0,
      "eval_MAE": 0.14292865991592407,
      "eval_R2": 0.23018361342873883,
      "eval_RMSE": 0.19261230528354645,
      "eval_loss": 0.037099506705999374,
      "eval_runtime": 14.2625,
      "eval_samples_per_second": 201.999,
      "eval_steps_per_second": 3.436,
      "step": 965
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.641444683074951,
      "learning_rate": 1.400518134715026e-05,
      "loss": 0.0251,
      "step": 1158
    },
    {
      "epoch": 6.0,
      "eval_MAE": 0.14436666667461395,
      "eval_R2": 0.21001704033333712,
      "eval_RMSE": 0.19511890411376953,
      "eval_loss": 0.03807138651609421,
      "eval_runtime": 14.0937,
      "eval_samples_per_second": 204.418,
      "eval_steps_per_second": 3.477,
      "step": 1158
    },
    {
      "epoch": 7.0,
      "grad_norm": 6.497671127319336,
      "learning_rate": 1.300518134715026e-05,
      "loss": 0.0211,
      "step": 1351
    },
    {
      "epoch": 7.0,
      "eval_MAE": 0.14056500792503357,
      "eval_R2": 0.25886664349880983,
      "eval_RMSE": 0.18898992240428925,
      "eval_loss": 0.0357171930372715,
      "eval_runtime": 14.0545,
      "eval_samples_per_second": 204.987,
      "eval_steps_per_second": 3.486,
      "step": 1351
    },
    {
      "epoch": 8.0,
      "grad_norm": 7.818419933319092,
      "learning_rate": 1.2005181347150261e-05,
      "loss": 0.0174,
      "step": 1544
    },
    {
      "epoch": 8.0,
      "eval_MAE": 0.15229639410972595,
      "eval_R2": 0.11496506128758999,
      "eval_RMSE": 0.2065240740776062,
      "eval_loss": 0.0426521971821785,
      "eval_runtime": 14.1469,
      "eval_samples_per_second": 203.649,
      "eval_steps_per_second": 3.464,
      "step": 1544
    },
    {
      "epoch": 9.0,
      "grad_norm": 4.021462917327881,
      "learning_rate": 1.100518134715026e-05,
      "loss": 0.0153,
      "step": 1737
    },
    {
      "epoch": 9.0,
      "eval_MAE": 0.13968582451343536,
      "eval_R2": 0.2638301636372278,
      "eval_RMSE": 0.18835601210594177,
      "eval_loss": 0.03547798842191696,
      "eval_runtime": 14.0576,
      "eval_samples_per_second": 204.942,
      "eval_steps_per_second": 3.486,
      "step": 1737
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.8230902552604675,
      "learning_rate": 1.0005181347150258e-05,
      "loss": 0.0127,
      "step": 1930
    },
    {
      "epoch": 10.0,
      "eval_MAE": 0.1468822956085205,
      "eval_R2": 0.2201558651457277,
      "eval_RMSE": 0.19386275112628937,
      "eval_loss": 0.037582769989967346,
      "eval_runtime": 14.0579,
      "eval_samples_per_second": 204.939,
      "eval_steps_per_second": 3.486,
      "step": 1930
    },
    {
      "epoch": 11.0,
      "grad_norm": 1.3326853513717651,
      "learning_rate": 9.005181347150261e-06,
      "loss": 0.0112,
      "step": 2123
    },
    {
      "epoch": 11.0,
      "eval_MAE": 0.15102224051952362,
      "eval_R2": 0.12659890818148722,
      "eval_RMSE": 0.2051621973514557,
      "eval_loss": 0.042091526091098785,
      "eval_runtime": 14.0985,
      "eval_samples_per_second": 204.348,
      "eval_steps_per_second": 3.476,
      "step": 2123
    },
    {
      "epoch": 12.0,
      "grad_norm": 4.036520481109619,
      "learning_rate": 8.005181347150259e-06,
      "loss": 0.0102,
      "step": 2316
    },
    {
      "epoch": 12.0,
      "eval_MAE": 0.15016663074493408,
      "eval_R2": 0.13732674296360936,
      "eval_RMSE": 0.20389831066131592,
      "eval_loss": 0.04157452657818794,
      "eval_runtime": 14.1218,
      "eval_samples_per_second": 204.011,
      "eval_steps_per_second": 3.47,
      "step": 2316
    }
  ],
  "logging_steps": 500,
  "max_steps": 3860,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 1.2883969633724826e+17,
  "train_batch_size": 60,
  "trial_name": null,
  "trial_params": null
}