README.md

metadata

dataset_info:
  features:
    - name: image
      dtype: image
    - name: question
      dtype: string
    - name: answer
      dtype: string
    - name: dataset
      dtype: string
  splits:
    - name: train
      num_bytes: 28625583143.45
      num_examples: 257835
    - name: test
      num_bytes: 10875586
      num_examples: 100
  download_size: 28613969788
  dataset_size: 28636458729.45
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*