trainer_state.json

{
  "best_metric": 0.5536,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0010/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": 0.0004287859287859288,
      "loss": 3.2501,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.21626666666666666,
      "eval_loss": 2.905996322631836,
      "eval_runtime": 6.5761,
      "eval_samples_per_second": 570.247,
      "eval_steps_per_second": 8.972,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00035757185757185754,
      "loss": 2.6027,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.34186666666666665,
      "eval_loss": 2.3999176025390625,
      "eval_runtime": 6.7584,
      "eval_samples_per_second": 554.865,
      "eval_steps_per_second": 8.73,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 15.549736022949219,
      "learning_rate": 0.00028614328614328614,
      "loss": 2.1371,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.40373333333333333,
      "eval_loss": 2.154214859008789,
      "eval_runtime": 6.851,
      "eval_samples_per_second": 547.364,
      "eval_steps_per_second": 8.612,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 20.237049102783203,
      "learning_rate": 0.0002147147147147147,
      "loss": 1.6575,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.46026666666666666,
      "eval_loss": 1.8864367008209229,
      "eval_runtime": 7.0404,
      "eval_samples_per_second": 532.639,
      "eval_steps_per_second": 8.38,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 9.874077796936035,
      "learning_rate": 0.0001432861432861433,
      "loss": 1.1309,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5122666666666666,
      "eval_loss": 1.7405973672866821,
      "eval_runtime": 6.9937,
      "eval_samples_per_second": 536.194,
      "eval_steps_per_second": 8.436,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 22.306495666503906,
      "learning_rate": 7.185757185757186e-05,
      "loss": 0.5517,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5442666666666667,
      "eval_loss": 1.7057732343673706,
      "eval_runtime": 8.1559,
      "eval_samples_per_second": 459.791,
      "eval_steps_per_second": 7.234,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.9594851732254028,
      "learning_rate": 4.2900042900042903e-07,
      "loss": 0.1386,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5536,
      "eval_loss": 1.7617971897125244,
      "eval_runtime": 6.4838,
      "eval_samples_per_second": 578.366,
      "eval_steps_per_second": 9.1,
      "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
}