trainer_state.json

{
  "best_metric": 0.8901333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0859/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": 6.655434848731736e-05,
      "loss": 1.6609,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8058666666666666,
      "eval_loss": 0.7088698148727417,
      "eval_runtime": 10.8786,
      "eval_samples_per_second": 344.713,
      "eval_steps_per_second": 5.423,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.710521221160889,
      "learning_rate": 5.685900305529986e-05,
      "loss": 0.4541,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8405333333333334,
      "eval_loss": 0.5497837066650391,
      "eval_runtime": 11.3274,
      "eval_samples_per_second": 331.057,
      "eval_steps_per_second": 5.209,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 11.671475410461426,
      "learning_rate": 4.28342139851007e-05,
      "loss": 0.2233,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8650666666666667,
      "eval_loss": 0.44137898087501526,
      "eval_runtime": 10.9465,
      "eval_samples_per_second": 342.575,
      "eval_steps_per_second": 5.39,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.27672886848449707,
      "learning_rate": 2.7257762754823102e-05,
      "loss": 0.1024,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8765333333333334,
      "eval_loss": 0.4443959891796112,
      "eval_runtime": 10.8175,
      "eval_samples_per_second": 346.661,
      "eval_steps_per_second": 5.454,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.13102540373802185,
      "learning_rate": 1.3214756563261342e-05,
      "loss": 0.0383,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8864,
      "eval_loss": 0.41161346435546875,
      "eval_runtime": 10.8471,
      "eval_samples_per_second": 345.714,
      "eval_steps_per_second": 5.439,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.2367280721664429,
      "learning_rate": 3.4865850128228307e-06,
      "loss": 0.0139,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8888,
      "eval_loss": 0.40637022256851196,
      "eval_runtime": 11.2138,
      "eval_samples_per_second": 334.41,
      "eval_steps_per_second": 5.261,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.47114044427871704,
      "learning_rate": 3.178726637742546e-11,
      "loss": 0.0078,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8901333333333333,
      "eval_loss": 0.40787577629089355,
      "eval_runtime": 11.153,
      "eval_samples_per_second": 336.233,
      "eval_steps_per_second": 5.29,
      "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
}