trainer_state.json

{
  "best_metric": 0.8842666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0040/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": 2.85232922088503e-05,
      "loss": 2.3893,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7514666666666666,
      "eval_loss": 1.2075742483139038,
      "eval_runtime": 6.1997,
      "eval_samples_per_second": 604.867,
      "eval_steps_per_second": 9.517,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.532941818237305,
      "learning_rate": 2.4368144166557088e-05,
      "loss": 0.8042,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8408,
      "eval_loss": 0.6513584852218628,
      "eval_runtime": 6.591,
      "eval_samples_per_second": 568.957,
      "eval_steps_per_second": 8.952,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 1.837722045066209e-05,
      "loss": 0.4334,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8536,
      "eval_loss": 0.5355240106582642,
      "eval_runtime": 7.0331,
      "eval_samples_per_second": 533.194,
      "eval_steps_per_second": 8.389,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.753833532333374,
      "learning_rate": 1.1701616683386336e-05,
      "loss": 0.2648,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8717333333333334,
      "eval_loss": 0.44873321056365967,
      "eval_runtime": 6.5761,
      "eval_samples_per_second": 570.245,
      "eval_steps_per_second": 8.972,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.25633430480957,
      "learning_rate": 5.679298183982636e-06,
      "loss": 0.1734,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8816,
      "eval_loss": 0.4188244938850403,
      "eval_runtime": 6.5646,
      "eval_samples_per_second": 571.247,
      "eval_steps_per_second": 8.988,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.789655923843384,
      "learning_rate": 1.5030589901472829e-06,
      "loss": 0.1202,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8802666666666666,
      "eval_loss": 0.40899574756622314,
      "eval_runtime": 6.3549,
      "eval_samples_per_second": 590.098,
      "eval_steps_per_second": 9.284,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.4903445839881897,
      "learning_rate": 5.4492431898478526e-11,
      "loss": 0.0996,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8842666666666666,
      "eval_loss": 0.4069821536540985,
      "eval_runtime": 6.8883,
      "eval_samples_per_second": 544.405,
      "eval_steps_per_second": 8.565,
      "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
}