---
dataset_info:
  features:
  - name: image_id
    dtype: string
  - name: question
    dtype: string
  - name: negative_passages
    dtype: string
  - name: gold_index
    dtype: string
  - name: gold_in_union
    dtype: bool
  splits:
  - name: train
    num_bytes: 1419880547
    num_examples: 228675
  download_size: 380233133
  dataset_size: 1419880547
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---