---
dataset_info:
  features:
  - name: images
    sequence: image
  - name: question
    dtype: string
  - name: answer
    struct:
    - name: data
      sequence:
        sequence: string
  splits:
  - name: train
    num_bytes: 452817298.535
    num_examples: 2395
  download_size: 294213993
  dataset_size: 452817298.535
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---