trainer_state.json

{
  "best_metric": 0.8810666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0693/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": 9.507764069616767e-05,
      "loss": 1.8718,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.736,
      "eval_loss": 0.946401834487915,
      "eval_runtime": 12.5352,
      "eval_samples_per_second": 299.157,
      "eval_steps_per_second": 4.707,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7342416048049927,
      "learning_rate": 8.122714722185695e-05,
      "loss": 0.6214,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8096,
      "eval_loss": 0.6400879621505737,
      "eval_runtime": 12.3594,
      "eval_samples_per_second": 303.412,
      "eval_steps_per_second": 4.774,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.989327907562256,
      "learning_rate": 6.119173426442958e-05,
      "loss": 0.3383,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8285333333333333,
      "eval_loss": 0.5618894696235657,
      "eval_runtime": 12.4715,
      "eval_samples_per_second": 300.686,
      "eval_steps_per_second": 4.731,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.08769326657056808,
      "learning_rate": 3.893966107831872e-05,
      "loss": 0.176,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8656,
      "eval_loss": 0.4628290832042694,
      "eval_runtime": 12.3815,
      "eval_samples_per_second": 302.872,
      "eval_steps_per_second": 4.765,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 5.19697380065918,
      "learning_rate": 1.887822366180192e-05,
      "loss": 0.0759,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8730666666666667,
      "eval_loss": 0.454529732465744,
      "eval_runtime": 13.3654,
      "eval_samples_per_second": 280.575,
      "eval_steps_per_second": 4.414,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.22133997082710266,
      "learning_rate": 4.980835732604045e-06,
      "loss": 0.0283,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.876,
      "eval_loss": 0.45606639981269836,
      "eval_runtime": 12.5532,
      "eval_samples_per_second": 298.728,
      "eval_steps_per_second": 4.7,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.07335377484560013,
      "learning_rate": 4.5410380539179233e-11,
      "loss": 0.0167,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8810666666666667,
      "eval_loss": 0.4372062385082245,
      "eval_runtime": 12.588,
      "eval_samples_per_second": 297.904,
      "eval_steps_per_second": 4.687,
      "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
}