---
dataset_info:
  features:
  - name: candidate_id
    dtype: string
  - name: caption
    dtype: string
  - name: group
    sequence: string
  - name: split
    dtype: string
  - name: target_id
    dtype: string
  - name: target
    dtype: image
  - name: candidate
    dtype: image
  splits:
  - name: val
    num_bytes: 902818814.275
    num_examples: 4181
  download_size: 608604226
  dataset_size: 902818814.275
configs:
- config_name: default
  data_files:
  - split: val
    path: data/val-*
---