---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: question
    dtype: string
  - name: answer
    dtype: string
  splits:
  - name: train
    num_bytes: 13979199515.88
    num_examples: 4320
  - name: test
    num_bytes: 1713198571.0
    num_examples: 540
  - name: val
    num_bytes: 1716902162.0
    num_examples: 540
  download_size: 17342263054
  dataset_size: 17409300248.879997
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
  - split: val
    path: data/val-*
---