---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: image_url
    dtype: string
  - name: caption
    dtype: string
  - name: clip_score
    dtype: float64
  splits:
  - name: train
    num_bytes: 1435616055
    num_examples: 500
  download_size: 1435214603
  dataset_size: 1435616055
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---