trainer_state.json

{
  "best_metric": 0.7258666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0078/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": 0.00047538820348083834,
      "loss": 2.9279,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3448,
      "eval_loss": 2.417267084121704,
      "eval_runtime": 5.7578,
      "eval_samples_per_second": 651.294,
      "eval_steps_per_second": 10.247,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00040639873814655834,
      "loss": 1.9684,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5072,
      "eval_loss": 1.7074518203735352,
      "eval_runtime": 6.071,
      "eval_samples_per_second": 617.689,
      "eval_steps_per_second": 9.718,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.207923412322998,
      "learning_rate": 0.00030628700751103485,
      "loss": 1.3234,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6096,
      "eval_loss": 1.286794900894165,
      "eval_runtime": 5.8679,
      "eval_samples_per_second": 639.072,
      "eval_steps_per_second": 10.055,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.65076208114624,
      "learning_rate": 0.00019502694472310563,
      "loss": 0.8373,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6717333333333333,
      "eval_loss": 1.0933171510696411,
      "eval_runtime": 6.3843,
      "eval_samples_per_second": 587.383,
      "eval_steps_per_second": 9.241,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.41366034746170044,
      "learning_rate": 9.465496973304394e-05,
      "loss": 0.405,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7130666666666666,
      "eval_loss": 0.9878906011581421,
      "eval_runtime": 5.838,
      "eval_samples_per_second": 642.346,
      "eval_steps_per_second": 10.106,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.19928739964962006,
      "learning_rate": 2.505098316912138e-05,
      "loss": 0.1401,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7234666666666667,
      "eval_loss": 0.9728491902351379,
      "eval_runtime": 5.8545,
      "eval_samples_per_second": 640.533,
      "eval_steps_per_second": 10.078,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 5.151848316192627,
      "learning_rate": 9.082071983079754e-10,
      "loss": 0.0518,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7258666666666667,
      "eval_loss": 0.9867076873779297,
      "eval_runtime": 6.145,
      "eval_samples_per_second": 610.255,
      "eval_steps_per_second": 9.601,
      "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
}