checkpoint-200/trainer_state.json

{
  "best_metric": 0.6430063247680664,
  "best_model_checkpoint": "output/checkpoint-50",
  "epoch": 0.16033670708487824,
  "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": 4.444444444444445e-07,
      "loss": 0.5542,
      "step": 1
    },
    {
      "epoch": 0.04,
      "learning_rate": 2e-05,
      "loss": 0.554,
      "step": 50
    },
    {
      "epoch": 0.04,
      "eval_accuracy": 0.6626865671641791,
      "eval_loss": 0.6430063247680664,
      "eval_runtime": 61.7692,
      "eval_samples_per_second": 5.423,
      "eval_steps_per_second": 1.36,
      "step": 50
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.222222222222222e-05,
      "loss": 0.5957,
      "step": 100
    },
    {
      "epoch": 0.08,
      "eval_accuracy": 0.6686567164179105,
      "eval_loss": 0.6705919504165649,
      "eval_runtime": 60.4388,
      "eval_samples_per_second": 5.543,
      "eval_steps_per_second": 1.39,
      "step": 100
    },
    {
      "epoch": 0.12,
      "learning_rate": 6.444444444444446e-05,
      "loss": 0.5668,
      "step": 150
    },
    {
      "epoch": 0.12,
      "eval_accuracy": 0.6805970149253732,
      "eval_loss": 0.7224913835525513,
      "eval_runtime": 56.4882,
      "eval_samples_per_second": 5.93,
      "eval_steps_per_second": 1.487,
      "step": 150
    },
    {
      "epoch": 0.16,
      "learning_rate": 8.666666666666667e-05,
      "loss": 0.5626,
      "step": 200
    },
    {
      "epoch": 0.16,
      "eval_accuracy": 0.6746268656716418,
      "eval_loss": 0.7115103006362915,
      "eval_runtime": 56.506,
      "eval_samples_per_second": 5.929,
      "eval_steps_per_second": 1.487,
      "step": 200
    }
  ],
  "logging_steps": 50,
  "max_steps": 1247,
  "num_train_epochs": 1,
  "save_steps": 50,
  "total_flos": 7.406401432129536e+16,
  "trial_name": null,
  "trial_params": null
}