---
dataset_info:
  features:
  - name: id
    dtype: string
  - name: image
    sequence:
      sequence:
        sequence: uint8
  - name: answers
    sequence: string
  - name: words
    sequence: string
  - name: boxes
    sequence:
      sequence: float32
      length: 4
  - name: question
    dtype: string
  - name: answer
    dtype: string
  splits:
  - name: train
    num_bytes: 160343399
    num_examples: 1000
  - name: test
    num_bytes: 31701447
    num_examples: 200
  download_size: 48014432
  dataset_size: 192044846
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---