{
  "best_metric": 0.029607130214571953,
  "best_model_checkpoint": "model/checkpoint-6500",
  "epoch": 2.138157894736842,
  "eval_steps": 500,
  "global_step": 6500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.16,
      "learning_rate": 4.7944078947368425e-05,
      "loss": 0.1494,
      "step": 500
    },
    {
      "epoch": 0.16,
      "eval_accuracy": 0.9837171052631579,
      "eval_loss": 0.0790267065167427,
      "eval_runtime": 113.0978,
      "eval_samples_per_second": 53.759,
      "eval_steps_per_second": 6.72,
      "step": 500
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.588815789473684e-05,
      "loss": 0.1072,
      "step": 1000
    },
    {
      "epoch": 0.33,
      "eval_accuracy": 0.9824013157894737,
      "eval_loss": 0.06455110013484955,
      "eval_runtime": 113.235,
      "eval_samples_per_second": 53.694,
      "eval_steps_per_second": 6.712,
      "step": 1000
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.383223684210527e-05,
      "loss": 0.0765,
      "step": 1500
    },
    {
      "epoch": 0.49,
      "eval_accuracy": 0.9899671052631579,
      "eval_loss": 0.04877911135554314,
      "eval_runtime": 113.1524,
      "eval_samples_per_second": 53.733,
      "eval_steps_per_second": 6.717,
      "step": 1500
    },
    {
      "epoch": 0.66,
      "learning_rate": 4.177631578947369e-05,
      "loss": 0.0742,
      "step": 2000
    },
    {
      "epoch": 0.66,
      "eval_accuracy": 0.9886513157894737,
      "eval_loss": 0.054583221673965454,
      "eval_runtime": 113.2769,
      "eval_samples_per_second": 53.674,
      "eval_steps_per_second": 6.709,
      "step": 2000
    },
    {
      "epoch": 0.82,
      "learning_rate": 3.972039473684211e-05,
      "loss": 0.0748,
      "step": 2500
    },
    {
      "epoch": 0.82,
      "eval_accuracy": 0.990625,
      "eval_loss": 0.04383059963583946,
      "eval_runtime": 113.3147,
      "eval_samples_per_second": 53.656,
      "eval_steps_per_second": 6.707,
      "step": 2500
    },
    {
      "epoch": 0.99,
      "learning_rate": 3.7664473684210526e-05,
      "loss": 0.0437,
      "step": 3000
    },
    {
      "epoch": 0.99,
      "eval_accuracy": 0.9901315789473685,
      "eval_loss": 0.05416030064225197,
      "eval_runtime": 113.1123,
      "eval_samples_per_second": 53.752,
      "eval_steps_per_second": 6.719,
      "step": 3000
    },
    {
      "epoch": 1.15,
      "learning_rate": 3.560855263157895e-05,
      "loss": 0.0134,
      "step": 3500
    },
    {
      "epoch": 1.15,
      "eval_accuracy": 0.990953947368421,
      "eval_loss": 0.05749928951263428,
      "eval_runtime": 113.4787,
      "eval_samples_per_second": 53.578,
      "eval_steps_per_second": 6.697,
      "step": 3500
    },
    {
      "epoch": 1.32,
      "learning_rate": 3.355263157894737e-05,
      "loss": 0.0277,
      "step": 4000
    },
    {
      "epoch": 1.32,
      "eval_accuracy": 0.9904605263157895,
      "eval_loss": 0.06314379721879959,
      "eval_runtime": 113.4629,
      "eval_samples_per_second": 53.586,
      "eval_steps_per_second": 6.698,
      "step": 4000
    },
    {
      "epoch": 1.48,
      "learning_rate": 3.1496710526315794e-05,
      "loss": 0.0231,
      "step": 4500
    },
    {
      "epoch": 1.48,
      "eval_accuracy": 0.9912828947368421,
      "eval_loss": 0.04834901914000511,
      "eval_runtime": 113.5144,
      "eval_samples_per_second": 53.561,
      "eval_steps_per_second": 6.695,
      "step": 4500
    },
    {
      "epoch": 1.64,
      "learning_rate": 2.944078947368421e-05,
      "loss": 0.0243,
      "step": 5000
    },
    {
      "epoch": 1.64,
      "eval_accuracy": 0.9901315789473685,
      "eval_loss": 0.06605446338653564,
      "eval_runtime": 113.6185,
      "eval_samples_per_second": 53.512,
      "eval_steps_per_second": 6.689,
      "step": 5000
    },
    {
      "epoch": 1.81,
      "learning_rate": 2.7384868421052633e-05,
      "loss": 0.0232,
      "step": 5500
    },
    {
      "epoch": 1.81,
      "eval_accuracy": 0.9932565789473684,
      "eval_loss": 0.037462268024683,
      "eval_runtime": 112.9475,
      "eval_samples_per_second": 53.83,
      "eval_steps_per_second": 6.729,
      "step": 5500
    },
    {
      "epoch": 1.97,
      "learning_rate": 2.5328947368421052e-05,
      "loss": 0.0198,
      "step": 6000
    },
    {
      "epoch": 1.97,
      "eval_accuracy": 0.9945723684210527,
      "eval_loss": 0.034890029579401016,
      "eval_runtime": 112.9353,
      "eval_samples_per_second": 53.836,
      "eval_steps_per_second": 6.73,
      "step": 6000
    },
    {
      "epoch": 2.14,
      "learning_rate": 2.3273026315789475e-05,
      "loss": 0.0122,
      "step": 6500
    },
    {
      "epoch": 2.14,
      "eval_accuracy": 0.9962171052631579,
      "eval_loss": 0.029607130214571953,
      "eval_runtime": 112.9987,
      "eval_samples_per_second": 53.806,
      "eval_steps_per_second": 6.726,
      "step": 6500
    }
  ],
  "logging_steps": 500,
  "max_steps": 12160,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 1.368177487872e+16,
  "trial_name": null,
  "trial_params": null
}