---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: query
    sequence: string
  - name: answers
    sequence: string
  - name: ground_truth
    dtype: string
  splits:
  - name: train
    num_bytes: 39363167.0
    num_examples: 600
  - name: test
    num_bytes: 227796.0
    num_examples: 3
  download_size: 33948680
  dataset_size: 39590963.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---