trainer_state.json

{
  "best_metric": 0.8824,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0873/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": 3e-05,
      "loss": 2.4217,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7141333333333333,
      "eval_loss": 1.2633739709854126,
      "eval_runtime": 12.9727,
      "eval_samples_per_second": 289.069,
      "eval_steps_per_second": 4.548,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 3e-05,
      "loss": 0.8427,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8264,
      "eval_loss": 0.6709011197090149,
      "eval_runtime": 12.3964,
      "eval_samples_per_second": 302.507,
      "eval_steps_per_second": 4.759,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.9713053703308105,
      "learning_rate": 3e-05,
      "loss": 0.4676,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8613333333333333,
      "eval_loss": 0.502513587474823,
      "eval_runtime": 13.288,
      "eval_samples_per_second": 282.21,
      "eval_steps_per_second": 4.44,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.08915638923645,
      "learning_rate": 3e-05,
      "loss": 0.3052,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8573333333333333,
      "eval_loss": 0.485494464635849,
      "eval_runtime": 13.7924,
      "eval_samples_per_second": 271.888,
      "eval_steps_per_second": 4.278,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 43.37397384643555,
      "learning_rate": 3e-05,
      "loss": 0.21,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8661333333333333,
      "eval_loss": 0.45904648303985596,
      "eval_runtime": 14.3346,
      "eval_samples_per_second": 261.605,
      "eval_steps_per_second": 4.116,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.6002681255340576,
      "learning_rate": 3e-05,
      "loss": 0.1512,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.872,
      "eval_loss": 0.4403385818004608,
      "eval_runtime": 13.3038,
      "eval_samples_per_second": 281.875,
      "eval_steps_per_second": 4.435,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.13070310652256012,
      "learning_rate": 3e-05,
      "loss": 0.1096,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8824,
      "eval_loss": 0.42060381174087524,
      "eval_runtime": 13.4204,
      "eval_samples_per_second": 279.425,
      "eval_steps_per_second": 4.396,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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
}