trainer_state.json

{
  "best_metric": 0.8365333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0775/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": 0.0002627252252252252,
      "loss": 2.6257,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5242666666666667,
      "eval_loss": 1.633152723312378,
      "eval_runtime": 6.5176,
      "eval_samples_per_second": 575.363,
      "eval_steps_per_second": 9.052,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 26.28261947631836,
      "learning_rate": 0.0002252252252252252,
      "loss": 1.0953,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6912,
      "eval_loss": 0.9823421835899353,
      "eval_runtime": 7.0934,
      "eval_samples_per_second": 528.664,
      "eval_steps_per_second": 8.318,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.641635894775391,
      "learning_rate": 0.00018772522522522518,
      "loss": 0.5354,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7669333333333334,
      "eval_loss": 0.7766706347465515,
      "eval_runtime": 14.3256,
      "eval_samples_per_second": 261.769,
      "eval_steps_per_second": 4.119,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.4988974630832672,
      "learning_rate": 0.0001502252252252252,
      "loss": 0.2439,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7818666666666667,
      "eval_loss": 0.7360715866088867,
      "eval_runtime": 6.7064,
      "eval_samples_per_second": 559.171,
      "eval_steps_per_second": 8.798,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.21947646141052246,
      "learning_rate": 0.00011272522522522522,
      "loss": 0.1023,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.792,
      "eval_loss": 0.7378551959991455,
      "eval_runtime": 6.5074,
      "eval_samples_per_second": 576.264,
      "eval_steps_per_second": 9.067,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.5241366028785706,
      "learning_rate": 7.522522522522523e-05,
      "loss": 0.0362,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8064,
      "eval_loss": 0.7492461204528809,
      "eval_runtime": 6.4582,
      "eval_samples_per_second": 580.661,
      "eval_steps_per_second": 9.136,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.041464537382125854,
      "learning_rate": 3.7725225225225225e-05,
      "loss": 0.0078,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8365333333333334,
      "eval_loss": 0.686863362789154,
      "eval_runtime": 6.6324,
      "eval_samples_per_second": 565.407,
      "eval_steps_per_second": 8.896,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}