trainer_state.json

{
  "best_metric": 0.844,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0826/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "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": 7e-05,
      "loss": 1.5777,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7698666666666667,
      "eval_loss": 0.7691076993942261,
      "eval_runtime": 11.613,
      "eval_samples_per_second": 322.914,
      "eval_steps_per_second": 5.081,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 28.610206604003906,
      "learning_rate": 7e-05,
      "loss": 0.5664,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8234666666666667,
      "eval_loss": 0.6102097034454346,
      "eval_runtime": 11.2355,
      "eval_samples_per_second": 333.762,
      "eval_steps_per_second": 5.251,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 10.222493171691895,
      "learning_rate": 7e-05,
      "loss": 0.3783,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8344,
      "eval_loss": 0.5441902875900269,
      "eval_runtime": 12.3355,
      "eval_samples_per_second": 304.002,
      "eval_steps_per_second": 4.783,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 7e-05,
      "loss": 0.2828,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8394666666666667,
      "eval_loss": 0.5339456796646118,
      "eval_runtime": 11.2874,
      "eval_samples_per_second": 332.229,
      "eval_steps_per_second": 5.227,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 84.9881362915039,
      "learning_rate": 7e-05,
      "loss": 0.2175,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8394666666666667,
      "eval_loss": 0.5317642092704773,
      "eval_runtime": 11.1523,
      "eval_samples_per_second": 336.252,
      "eval_steps_per_second": 5.29,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.41221195459365845,
      "learning_rate": 7e-05,
      "loss": 0.1687,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.844,
      "eval_loss": 0.546898603439331,
      "eval_runtime": 11.2097,
      "eval_samples_per_second": 334.532,
      "eval_steps_per_second": 5.263,
      "step": 1998
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}