trainer_state.json

{
  "best_metric": 0.7685333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0385/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": Infinity,
      "learning_rate": 0.0003,
      "loss": 1.98,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7074666666666667,
      "eval_loss": 0.9985540509223938,
      "eval_runtime": 13.8005,
      "eval_samples_per_second": 271.73,
      "eval_steps_per_second": 4.275,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.5064191818237305,
      "learning_rate": 0.0003,
      "loss": 0.8029,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.756,
      "eval_loss": 0.8293225169181824,
      "eval_runtime": 13.0713,
      "eval_samples_per_second": 286.889,
      "eval_steps_per_second": 4.514,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.225942373275757,
      "learning_rate": 0.0003,
      "loss": 0.522,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7685333333333333,
      "eval_loss": 0.8025628924369812,
      "eval_runtime": 12.8443,
      "eval_samples_per_second": 291.959,
      "eval_steps_per_second": 4.593,
      "step": 999
    }
  ],
  "logging_steps": 500,
  "max_steps": 999,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.94223961867776e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}