README.md

metadata

dataset_info:
  features:
    - name: image
      dtype: image
    - name: label
      dtype: string
  splits:
    - name: train
      num_bytes: 3341095987.16
      num_examples: 25160
    - name: validation
      num_bytes: 748548716.14
      num_examples: 5372
    - name: test
      num_bytes: 760439383.18
      num_examples: 5445
  download_size: 5129381297
  dataset_size: 4850084086.48
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*
task_categories:
  - image-classification