Datasets:

zhimeng
/

hico_det

dataset_info:
  features:
    - name: image
      dtype: image
    - name: label
      dtype:
        class_label:
          names:
            '0': test
            '1': train
  splits:
    - name: train
      num_bytes: 12810367262.884
      num_examples: 76236
    - name: test
      num_bytes: 3356442982.744
      num_examples: 19316
  download_size: 16452513857
  dataset_size: 16166810245.628
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*