trainer_state.json

{
  "best_metric": 0.8594666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0132/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": 2.2607,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7304,
      "eval_loss": 1.0928750038146973,
      "eval_runtime": 13.7922,
      "eval_samples_per_second": 271.893,
      "eval_steps_per_second": 4.278,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 3e-05,
      "loss": 0.8139,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8069333333333333,
      "eval_loss": 0.6995651125907898,
      "eval_runtime": 14.5549,
      "eval_samples_per_second": 257.644,
      "eval_steps_per_second": 4.054,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.260736465454102,
      "learning_rate": 3e-05,
      "loss": 0.5264,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8293333333333334,
      "eval_loss": 0.5750405192375183,
      "eval_runtime": 13.4993,
      "eval_samples_per_second": 277.793,
      "eval_steps_per_second": 4.371,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 92.89977264404297,
      "learning_rate": 3e-05,
      "loss": 0.3783,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8421333333333333,
      "eval_loss": 0.53780198097229,
      "eval_runtime": 14.5847,
      "eval_samples_per_second": 257.119,
      "eval_steps_per_second": 4.045,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 68.60945892333984,
      "learning_rate": 3e-05,
      "loss": 0.2895,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8496,
      "eval_loss": 0.5136145949363708,
      "eval_runtime": 13.2173,
      "eval_samples_per_second": 283.718,
      "eval_steps_per_second": 4.464,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.627971172332764,
      "learning_rate": 3e-05,
      "loss": 0.215,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8541333333333333,
      "eval_loss": 0.48843345046043396,
      "eval_runtime": 13.3761,
      "eval_samples_per_second": 280.35,
      "eval_steps_per_second": 4.411,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 4.0504069328308105,
      "learning_rate": 3e-05,
      "loss": 0.1647,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8594666666666667,
      "eval_loss": 0.47765007615089417,
      "eval_runtime": 14.1859,
      "eval_samples_per_second": 264.347,
      "eval_steps_per_second": 4.159,
      "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
}