---
dataset_info:
  features:
  - name: question_id
    dtype: string
  - name: image_path
    dtype: string
  - name: text
    dtype: string
  - name: ground_truth
    dtype: string
  - name: image
    dtype: image
  splits:
  - name: train
    num_bytes: 440048210.0
    num_examples: 2100
  download_size: 417100334
  dataset_size: 440048210.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---