{
  "best_metric": 0.15142391622066498,
  "best_model_checkpoint": "./outputs/50_50/checkpoint-468",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 4680,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.943939393939394,
      "eval_loss": 0.15142391622066498,
      "eval_runtime": 2.5713,
      "eval_samples_per_second": 256.684,
      "eval_steps_per_second": 32.28,
      "step": 468
    },
    {
      "epoch": 1.07,
      "learning_rate": 1.7863247863247866e-05,
      "loss": 0.2863,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9303030303030303,
      "eval_loss": 0.1917150467634201,
      "eval_runtime": 2.7645,
      "eval_samples_per_second": 238.743,
      "eval_steps_per_second": 30.024,
      "step": 936
    },
    {
      "epoch": 2.14,
      "learning_rate": 1.5726495726495726e-05,
      "loss": 0.2377,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9333333333333333,
      "eval_loss": 0.17250454425811768,
      "eval_runtime": 2.6638,
      "eval_samples_per_second": 247.77,
      "eval_steps_per_second": 31.159,
      "step": 1404
    },
    {
      "epoch": 3.21,
      "learning_rate": 1.3589743589743592e-05,
      "loss": 0.2142,
      "step": 1500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9287878787878788,
      "eval_loss": 0.17816253006458282,
      "eval_runtime": 2.5386,
      "eval_samples_per_second": 259.984,
      "eval_steps_per_second": 32.695,
      "step": 1872
    },
    {
      "epoch": 4.27,
      "learning_rate": 1.1452991452991454e-05,
      "loss": 0.2058,
      "step": 2000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9272727272727272,
      "eval_loss": 0.17876102030277252,
      "eval_runtime": 2.5719,
      "eval_samples_per_second": 256.62,
      "eval_steps_per_second": 32.272,
      "step": 2340
    },
    {
      "epoch": 5.34,
      "learning_rate": 9.316239316239318e-06,
      "loss": 0.1899,
      "step": 2500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9318181818181818,
      "eval_loss": 0.18244825303554535,
      "eval_runtime": 2.5621,
      "eval_samples_per_second": 257.6,
      "eval_steps_per_second": 32.395,
      "step": 2808
    },
    {
      "epoch": 6.41,
      "learning_rate": 7.17948717948718e-06,
      "loss": 0.1838,
      "step": 3000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9333333333333333,
      "eval_loss": 0.1878737211227417,
      "eval_runtime": 2.7257,
      "eval_samples_per_second": 242.138,
      "eval_steps_per_second": 30.451,
      "step": 3276
    },
    {
      "epoch": 7.48,
      "learning_rate": 5.042735042735043e-06,
      "loss": 0.1757,
      "step": 3500
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9333333333333333,
      "eval_loss": 0.23907029628753662,
      "eval_runtime": 2.5302,
      "eval_samples_per_second": 260.853,
      "eval_steps_per_second": 32.804,
      "step": 3744
    },
    {
      "epoch": 8.55,
      "learning_rate": 2.9059829059829063e-06,
      "loss": 0.1852,
      "step": 4000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9409090909090909,
      "eval_loss": 0.17251791059970856,
      "eval_runtime": 2.7072,
      "eval_samples_per_second": 243.796,
      "eval_steps_per_second": 30.659,
      "step": 4212
    },
    {
      "epoch": 9.62,
      "learning_rate": 7.692307692307694e-07,
      "loss": 0.1634,
      "step": 4500
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9393939393939394,
      "eval_loss": 0.17622749507427216,
      "eval_runtime": 2.5223,
      "eval_samples_per_second": 261.67,
      "eval_steps_per_second": 32.907,
      "step": 4680
    },
    {
      "epoch": 10.0,
      "step": 4680,
      "total_flos": 2.898200411585741e+18,
      "train_loss": 0.20278241247193426,
      "train_runtime": 515.941,
      "train_samples_per_second": 72.489,
      "train_steps_per_second": 9.071
    }
  ],
  "logging_steps": 500,
  "max_steps": 4680,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 2.898200411585741e+18,
  "trial_name": null,
  "trial_params": null
}