trainer_state.json

{
  "best_metric": 0.544,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0890/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": 0.00047538820348083834,
      "loss": 3.208,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.21466666666666667,
      "eval_loss": 2.889378070831299,
      "eval_runtime": 5.9398,
      "eval_samples_per_second": 631.335,
      "eval_steps_per_second": 9.933,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00040639873814655834,
      "loss": 2.5821,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.34373333333333334,
      "eval_loss": 2.4301605224609375,
      "eval_runtime": 5.8945,
      "eval_samples_per_second": 636.182,
      "eval_steps_per_second": 10.009,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0003066152414593445,
      "loss": 2.1079,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.41386666666666666,
      "eval_loss": 2.0863778591156006,
      "eval_runtime": 5.7923,
      "eval_samples_per_second": 647.413,
      "eval_steps_per_second": 10.186,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.671649217605591,
      "learning_rate": 0.000195355683908512,
      "loss": 1.655,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.48773333333333335,
      "eval_loss": 1.8369499444961548,
      "eval_runtime": 5.8589,
      "eval_samples_per_second": 640.056,
      "eval_steps_per_second": 10.07,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 10.654561996459961,
      "learning_rate": 9.491910332815573e-05,
      "loss": 1.1332,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5170666666666667,
      "eval_loss": 1.7502703666687012,
      "eval_runtime": 5.9,
      "eval_samples_per_second": 635.59,
      "eval_steps_per_second": 10.0,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 13.987378120422363,
      "learning_rate": 2.519819627604078e-05,
      "loss": 0.5743,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5408,
      "eval_loss": 1.7337146997451782,
      "eval_runtime": 5.8473,
      "eval_samples_per_second": 641.327,
      "eval_steps_per_second": 10.09,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 4.116909503936768,
      "learning_rate": 2.0434646496314548e-09,
      "loss": 0.2567,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.544,
      "eval_loss": 1.7494640350341797,
      "eval_runtime": 6.1076,
      "eval_samples_per_second": 613.993,
      "eval_steps_per_second": 9.66,
      "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
}