---
dataset_info:
  features:
  - name: qry
    dtype: string
  - name: qry_image_path
    sequence: string
  - name: pos_text
    dtype: string
  - name: pos_image_path
    dtype: 'null'
  - name: neg_text
    dtype: string
  - name: neg_image_path
    dtype: 'null'
  splits:
  - name: train
    num_bytes: 380252030
    num_examples: 80500
  download_size: 114381126
  dataset_size: 380252030
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---