README.md

metadata

dataset_info:
  features:
    - name: image
      dtype: image
    - name: label
      dtype: string
    - name: bbox
      sequence:
        sequence: int64
  splits:
    - name: train
      num_bytes: 88352756
      num_examples: 8000
    - name: validation
      num_bytes: 22077430
      num_examples: 2000
  download_size: 109825041
  dataset_size: 110430186
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*