trainer_state.json

{
  "best_metric": 0.8992,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0175/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": 6.735155628520579e-05,
      "loss": 2.0865,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7637333333333334,
      "eval_loss": 0.9004030227661133,
      "eval_runtime": 6.2634,
      "eval_samples_per_second": 598.72,
      "eval_steps_per_second": 9.42,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.7780826091766357,
      "learning_rate": 5.9777905730703695e-05,
      "loss": 0.6087,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8242666666666667,
      "eval_loss": 0.6056052446365356,
      "eval_runtime": 5.9499,
      "eval_samples_per_second": 630.268,
      "eval_steps_per_second": 9.916,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.011209011077881,
      "learning_rate": 4.843204364097674e-05,
      "loss": 0.353,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8538666666666667,
      "eval_loss": 0.5070375800132751,
      "eval_runtime": 5.8288,
      "eval_samples_per_second": 643.359,
      "eval_steps_per_second": 10.122,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.29144492745399475,
      "learning_rate": 3.5041274668689884e-05,
      "loss": 0.2131,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8818666666666667,
      "eval_loss": 0.4232081472873688,
      "eval_runtime": 5.8118,
      "eval_samples_per_second": 645.239,
      "eval_steps_per_second": 10.152,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.866642713546753,
      "learning_rate": 2.16442220022508e-05,
      "loss": 0.1207,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8861333333333333,
      "eval_loss": 0.39914169907569885,
      "eval_runtime": 5.8255,
      "eval_samples_per_second": 643.723,
      "eval_steps_per_second": 10.128,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 19.929121017456055,
      "learning_rate": 1.0280465465539994e-05,
      "loss": 0.0667,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8914666666666666,
      "eval_loss": 0.395877867937088,
      "eval_runtime": 6.651,
      "eval_samples_per_second": 563.824,
      "eval_steps_per_second": 8.871,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.7404213547706604,
      "learning_rate": 2.6800339785621635e-06,
      "loss": 0.0393,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8992,
      "eval_loss": 0.3777335286140442,
      "eval_runtime": 6.224,
      "eval_samples_per_second": 602.502,
      "eval_steps_per_second": 9.479,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}