trainer_state.json

{
  "best_metric": 0.8430493273542601,
  "best_model_checkpoint": "vit-base-patch16-224-jvadlamudi2/checkpoint-7",
  "epoch": 3.0,
  "global_step": 21,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8430493273542601,
      "eval_loss": 0.4596026539802551,
      "eval_runtime": 3.3311,
      "eval_samples_per_second": 66.944,
      "eval_steps_per_second": 2.101,
      "step": 7
    },
    {
      "epoch": 1.43,
      "learning_rate": 3.055555555555556e-05,
      "loss": 0.5115,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8071748878923767,
      "eval_loss": 0.4634478688240051,
      "eval_runtime": 3.3321,
      "eval_samples_per_second": 66.924,
      "eval_steps_per_second": 2.101,
      "step": 14
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.777777777777778e-06,
      "loss": 0.4571,
      "step": 20
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8251121076233184,
      "eval_loss": 0.4669438898563385,
      "eval_runtime": 3.496,
      "eval_samples_per_second": 63.788,
      "eval_steps_per_second": 2.002,
      "step": 21
    },
    {
      "epoch": 3.0,
      "step": 21,
      "total_flos": 2.071360882398044e+17,
      "train_loss": 0.4825253642740704,
      "train_runtime": 229.1314,
      "train_samples_per_second": 11.666,
      "train_steps_per_second": 0.092
    }
  ],
  "max_steps": 21,
  "num_train_epochs": 3,
  "total_flos": 2.071360882398044e+17,
  "trial_name": null,
  "trial_params": null
}