README.md

metadata

dataset_info:
  features:
    - name: image
      dtype: image
    - name: solution
      struct:
        - name: text
          dtype: string
    - name: original_question
      dtype: string
    - name: original_answer
      dtype: string
    - name: question
      dtype: string
    - name: language
      dtype: string
    - name: source
      dtype: string
  splits:
    - name: train
      num_bytes: 2915131981
      num_examples: 581
  download_size: 2914761513
  dataset_size: 2915131981
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*