---
dataset_info:
  features:
  - name: image
    struct:
    - name: bytes
      dtype: binary
    - name: path
      dtype: string
  - name: question
    dtype: string
  - name: answer
    dtype: string
  splits:
  - name: train
    num_bytes: 48811901
    num_examples: 1400
  download_size: 42366000
  dataset_size: 48811901
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---