trainer_state.json

{
  "best_metric": 0.8066666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0329/checkpoints/checkpoint-999",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 999,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.7263991832733154,
      "learning_rate": 0.0005,
      "loss": 0.8447,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7552,
      "eval_loss": 0.838362991809845,
      "eval_runtime": 11.9928,
      "eval_samples_per_second": 312.688,
      "eval_steps_per_second": 4.92,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.15366792678833,
      "learning_rate": 0.0005,
      "loss": 0.4934,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8029333333333334,
      "eval_loss": 0.7237253189086914,
      "eval_runtime": 12.025,
      "eval_samples_per_second": 311.85,
      "eval_steps_per_second": 4.906,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 0.3719,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8066666666666666,
      "eval_loss": 0.6865793466567993,
      "eval_runtime": 12.0132,
      "eval_samples_per_second": 312.157,
      "eval_steps_per_second": 4.911,
      "step": 999
    }
  ],
  "logging_steps": 500,
  "max_steps": 1332,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.94223961867776e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}