---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: id
    dtype: string
  - name: query
    struct:
    - name: question
      dtype: string
  - name: answers
    sequence: string
  - name: words
    sequence: string
  - name: bounding_boxes
    sequence:
      sequence: int64
  - name: ground_truth
    dtype: string
  splits:
  - name: train
    num_bytes: 750926137.5
    num_examples: 1100
  - name: test
    num_bytes: 346217097.0
    num_examples: 492
  download_size: 1093528693
  dataset_size: 1097143234.5
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---