---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: question
    dtype: string
  - name: ground_truth
    dtype: string
  splits:
  - name: train
    num_bytes: 1321519958.506
    num_examples: 11781
  download_size: 965596255
  dataset_size: 1321519958.506
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---