---
dataset_info:
  features:
  - name: images
    dtype: image
  - name: questions
    sequence: string
  - name: answers
    sequence: string
  - name: question_answer_pairs
    list:
    - name: answer
      dtype: string
    - name: question
      dtype: string
  - name: base_question
    dtype: string
  - name: base_answer
    dtype: string
  splits:
  - name: train
    num_bytes: 1230407840.125
    num_examples: 9255
  download_size: 1217151520
  dataset_size: 1230407840.125
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---