trainer_state.json

{
  "best_metric": 0.7729406356811523,
  "best_model_checkpoint": "./vit-base/checkpoint-100",
  "epoch": 50.0,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 5.0,
      "learning_rate": 0.000182,
      "loss": 0.4121,
      "step": 10
    },
    {
      "epoch": 10.0,
      "learning_rate": 0.000162,
      "loss": 0.0007,
      "step": 20
    },
    {
      "epoch": 15.0,
      "learning_rate": 0.000142,
      "loss": 0.0002,
      "step": 30
    },
    {
      "epoch": 20.0,
      "learning_rate": 0.000122,
      "loss": 0.0001,
      "step": 40
    },
    {
      "epoch": 25.0,
      "learning_rate": 0.00010200000000000001,
      "loss": 0.0001,
      "step": 50
    },
    {
      "epoch": 30.0,
      "learning_rate": 8.2e-05,
      "loss": 0.0001,
      "step": 60
    },
    {
      "epoch": 35.0,
      "learning_rate": 6.2e-05,
      "loss": 0.0001,
      "step": 70
    },
    {
      "epoch": 40.0,
      "learning_rate": 4.2e-05,
      "loss": 0.0001,
      "step": 80
    },
    {
      "epoch": 45.0,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 0.0001,
      "step": 90
    },
    {
      "epoch": 50.0,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.0,
      "step": 100
    },
    {
      "epoch": 50.0,
      "eval_accuracy": 0.75,
      "eval_loss": 0.7729406356811523,
      "eval_runtime": 0.1172,
      "eval_samples_per_second": 34.137,
      "eval_steps_per_second": 8.534,
      "step": 100
    },
    {
      "epoch": 50.0,
      "step": 100,
      "total_flos": 7.7493378686976e+16,
      "train_loss": 0.04134651497006416,
      "train_runtime": 38.0787,
      "train_samples_per_second": 26.261,
      "train_steps_per_second": 2.626
    }
  ],
  "max_steps": 100,
  "num_train_epochs": 50,
  "total_flos": 7.7493378686976e+16,
  "trial_name": null,
  "trial_params": null
}