---
dataset_info:
  features:
  - name: path
    dtype: string
  - name: question
    dtype: string
  - name: dictionary
    sequence: string
  - name: answer
    sequence: string
  - name: image
    dtype: image
  splits:
  - name: train
    num_bytes: 2830937489.75
    num_examples: 26074
  - name: test
    num_bytes: 479924998.25
    num_examples: 4102
  download_size: 2365703118
  dataset_size: 3310862488.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---