---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: question
    dtype: string
  - name: answer
    dtype: string
  - name: dataset
    dtype: string
  splits:
  - name: test
    num_bytes: 338910200.0
    num_examples: 886
  - name: train
    num_bytes: 1020339788.41
    num_examples: 2743
  download_size: 379500980
  dataset_size: 1359249988.4099998
configs:
- config_name: default
  data_files:
  - split: test
    path: data/test-*
  - split: train
    path: data/train-*
---