trainer_state.json

{
  "best_metric": 0.45936229825019836,
  "best_model_checkpoint": "./vit-base-beans/checkpoint-80",
  "epoch": 3.2,
  "global_step": 80,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4,
      "learning_rate": 0.00018,
      "loss": 0.653,
      "step": 10
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.00016,
      "loss": 0.4656,
      "step": 20
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00014,
      "loss": 0.4179,
      "step": 30
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00012,
      "loss": 0.2177,
      "step": 40
    },
    {
      "epoch": 1.6,
      "eval_accuracy": 0.6923076923076923,
      "eval_loss": 0.5713487267494202,
      "eval_runtime": 45.484,
      "eval_samples_per_second": 0.572,
      "eval_steps_per_second": 0.088,
      "step": 40
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0001,
      "loss": 0.2069,
      "step": 50
    },
    {
      "epoch": 2.4,
      "learning_rate": 8e-05,
      "loss": 0.1021,
      "step": 60
    },
    {
      "epoch": 2.8,
      "learning_rate": 6e-05,
      "loss": 0.0848,
      "step": 70
    },
    {
      "epoch": 3.2,
      "learning_rate": 4e-05,
      "loss": 0.058,
      "step": 80
    },
    {
      "epoch": 3.2,
      "eval_accuracy": 0.8076923076923077,
      "eval_loss": 0.45936229825019836,
      "eval_runtime": 45.4285,
      "eval_samples_per_second": 0.572,
      "eval_steps_per_second": 0.088,
      "step": 80
    }
  ],
  "max_steps": 100,
  "num_train_epochs": 4,
  "total_flos": 4.959487335333888e+16,
  "trial_name": null,
  "trial_params": null
}