---
dataset_info:
  features:
  - name: image_id
    dtype: string
  - name: caption
    dtype: string
  - name: original_caption
    dtype: string
  - name: question
    dtype: string
  - name: answer
    dtype: string
  - name: match_type
    dtype: string
  - name: image
    dtype: image
  splits:
  - name: train
    num_bytes: 259293284.0
    num_examples: 5000
  download_size: 257739788
  dataset_size: 259293284.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---