checkpoint-200/trainer_state.json

{
  "best_metric": 0.5592061877250671,
  "best_model_checkpoint": "output/checkpoint-200",
  "epoch": 0.201765447667087,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.5465,
      "step": 1
    },
    {
      "epoch": 0.05,
      "learning_rate": 7.5e-05,
      "loss": 0.7053,
      "step": 50
    },
    {
      "epoch": 0.05,
      "eval_accuracy": 0.6561679790026247,
      "eval_loss": 0.6764907836914062,
      "eval_runtime": 59.7766,
      "eval_samples_per_second": 6.374,
      "eval_steps_per_second": 1.606,
      "step": 50
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00015666666666666666,
      "loss": 0.6576,
      "step": 100
    },
    {
      "epoch": 0.1,
      "eval_accuracy": 0.7217847769028871,
      "eval_loss": 0.574826717376709,
      "eval_runtime": 59.8708,
      "eval_samples_per_second": 6.364,
      "eval_steps_per_second": 1.603,
      "step": 100
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.00023999999999999998,
      "loss": 0.5874,
      "step": 150
    },
    {
      "epoch": 0.15,
      "eval_accuracy": 0.7480314960629921,
      "eval_loss": 0.5501317381858826,
      "eval_runtime": 55.7487,
      "eval_samples_per_second": 6.834,
      "eval_steps_per_second": 1.722,
      "step": 150
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002948780487804878,
      "loss": 0.5775,
      "step": 200
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.7401574803149606,
      "eval_loss": 0.5592061877250671,
      "eval_runtime": 54.7265,
      "eval_samples_per_second": 6.962,
      "eval_steps_per_second": 1.754,
      "step": 200
    }
  ],
  "logging_steps": 50,
  "max_steps": 1000,
  "num_train_epochs": 2,
  "save_steps": 100,
  "total_flos": 3.444343869136896e+16,
  "trial_name": null,
  "trial_params": null
}