{
  "best_metric": 0.9495412844036697,
  "best_model_checkpoint": "./electra-base-discriminator/fine_tuned_models/checkpoint-8420",
  "epoch": 10.0,
  "global_step": 21050,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 1.8398562688161604e-05,
      "loss": 0.2278,
      "step": 2105
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9380733944954128,
      "eval_loss": 0.190090149641037,
      "eval_runtime": 2.2145,
      "eval_samples_per_second": 393.765,
      "eval_steps_per_second": 49.221,
      "step": 2105
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.6354277945032536e-05,
      "loss": 0.1202,
      "step": 4210
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9403669724770642,
      "eval_loss": 0.19069908559322357,
      "eval_runtime": 2.2053,
      "eval_samples_per_second": 395.408,
      "eval_steps_per_second": 49.426,
      "step": 4210
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.4309993201903469e-05,
      "loss": 0.0872,
      "step": 6315
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9438073394495413,
      "eval_loss": 0.18960702419281006,
      "eval_runtime": 2.2065,
      "eval_samples_per_second": 395.202,
      "eval_steps_per_second": 49.4,
      "step": 6315
    },
    {
      "epoch": 4.0,
      "learning_rate": 1.2265708458774402e-05,
      "loss": 0.068,
      "step": 8420
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9495412844036697,
      "eval_loss": 0.15451420843601227,
      "eval_runtime": 2.1827,
      "eval_samples_per_second": 399.509,
      "eval_steps_per_second": 49.939,
      "step": 8420
    },
    {
      "epoch": 5.0,
      "learning_rate": 1.0221423715645335e-05,
      "loss": 0.0524,
      "step": 10525
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.944954128440367,
      "eval_loss": 0.2057093232870102,
      "eval_runtime": 2.2044,
      "eval_samples_per_second": 395.579,
      "eval_steps_per_second": 49.447,
      "step": 10525
    },
    {
      "epoch": 6.0,
      "learning_rate": 8.177138972516268e-06,
      "loss": 0.0408,
      "step": 12630
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9438073394495413,
      "eval_loss": 0.21199779212474823,
      "eval_runtime": 2.1989,
      "eval_samples_per_second": 396.567,
      "eval_steps_per_second": 49.571,
      "step": 12630
    },
    {
      "epoch": 7.0,
      "learning_rate": 6.132854229387201e-06,
      "loss": 0.031,
      "step": 14735
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.944954128440367,
      "eval_loss": 0.27671030163764954,
      "eval_runtime": 2.19,
      "eval_samples_per_second": 398.168,
      "eval_steps_per_second": 49.771,
      "step": 14735
    },
    {
      "epoch": 8.0,
      "learning_rate": 4.088569486258134e-06,
      "loss": 0.0219,
      "step": 16840
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9403669724770642,
      "eval_loss": 0.27350538969039917,
      "eval_runtime": 2.2049,
      "eval_samples_per_second": 395.478,
      "eval_steps_per_second": 49.435,
      "step": 16840
    },
    {
      "epoch": 9.0,
      "learning_rate": 2.044284743129067e-06,
      "loss": 0.018,
      "step": 18945
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9426605504587156,
      "eval_loss": 0.2893427312374115,
      "eval_runtime": 2.2181,
      "eval_samples_per_second": 393.129,
      "eval_steps_per_second": 49.141,
      "step": 18945
    },
    {
      "epoch": 10.0,
      "learning_rate": 0.0,
      "loss": 0.0125,
      "step": 21050
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9461009174311926,
      "eval_loss": 0.3249703049659729,
      "eval_runtime": 2.2049,
      "eval_samples_per_second": 395.491,
      "eval_steps_per_second": 49.436,
      "step": 21050
    },
    {
      "epoch": 10.0,
      "step": 21050,
      "total_flos": 2.831471218376448e+16,
      "train_loss": 0.06799061220218904,
      "train_runtime": 4411.4461,
      "train_samples_per_second": 152.669,
      "train_steps_per_second": 4.772
    }
  ],
  "max_steps": 21050,
  "num_train_epochs": 10,
  "total_flos": 2.831471218376448e+16,
  "trial_name": null,
  "trial_params": null
}