trainer_state.json

{
  "best_metric": 0.8344,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0939/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": 9e-05,
      "loss": 1.656,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7650666666666667,
      "eval_loss": 0.8085070848464966,
      "eval_runtime": 7.1188,
      "eval_samples_per_second": 526.776,
      "eval_steps_per_second": 8.288,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.3395028114318848,
      "learning_rate": 9e-05,
      "loss": 0.6416,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8224,
      "eval_loss": 0.6327426433563232,
      "eval_runtime": 7.3964,
      "eval_samples_per_second": 507.005,
      "eval_steps_per_second": 7.977,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 33.761505126953125,
      "learning_rate": 9e-05,
      "loss": 0.4326,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8344,
      "eval_loss": 0.531470000743866,
      "eval_runtime": 7.5219,
      "eval_samples_per_second": 498.546,
      "eval_steps_per_second": 7.844,
      "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
}