README.md

metadata

dataset_info:
  features:
    - name: image
      sequence:
        sequence:
          sequence: uint8
    - name: labels
      dtype: int64
  splits:
    - name: train
      num_bytes: 42301183339
      num_examples: 27612
  download_size: 11266692714
  dataset_size: 42301183339
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*