trainer_state.json

{
  "best_metric": 0.5189333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0289/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "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": 0.0005,
      "loss": 3.225,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.22266666666666668,
      "eval_loss": 2.870443344116211,
      "eval_runtime": 11.1071,
      "eval_samples_per_second": 337.623,
      "eval_steps_per_second": 5.312,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.6154,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.31893333333333335,
      "eval_loss": 2.4606032371520996,
      "eval_runtime": 11.4715,
      "eval_samples_per_second": 326.897,
      "eval_steps_per_second": 5.143,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 14.095934867858887,
      "learning_rate": 0.0005,
      "loss": 2.2143,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.38293333333333335,
      "eval_loss": 2.192361354827881,
      "eval_runtime": 14.0583,
      "eval_samples_per_second": 266.746,
      "eval_steps_per_second": 4.197,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 11.138202667236328,
      "learning_rate": 0.0005,
      "loss": 1.9412,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.40986666666666666,
      "eval_loss": 2.106083631515503,
      "eval_runtime": 12.837,
      "eval_samples_per_second": 292.124,
      "eval_steps_per_second": 4.596,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 9.29428768157959,
      "learning_rate": 0.0005,
      "loss": 1.7204,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.45386666666666664,
      "eval_loss": 1.958842396736145,
      "eval_runtime": 10.8672,
      "eval_samples_per_second": 345.076,
      "eval_steps_per_second": 5.429,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 12.648058891296387,
      "learning_rate": 0.0005,
      "loss": 1.5102,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4856,
      "eval_loss": 1.817623496055603,
      "eval_runtime": 11.1352,
      "eval_samples_per_second": 336.771,
      "eval_steps_per_second": 5.299,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 1.3261,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5189333333333334,
      "eval_loss": 1.6764856576919556,
      "eval_runtime": 11.3173,
      "eval_samples_per_second": 331.352,
      "eval_steps_per_second": 5.213,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}