---
language:
- en
dataset_info:
  features:
  - name: image_emb
    dtype:
      array3_d:
        shape:
        - 4
        - 8
        - 8
        dtype: float32
  - name: text_emb
    dtype:
      array2_d:
        shape:
        - 1
        - 768
        dtype: float32
  - name: text
    dtype: string
  - name: image
    dtype: image
  splits:
  - name: train
    num_bytes: 195358513.25
    num_examples: 14630
  - name: val
    num_bytes: 20025508.5
    num_examples: 1500
  download_size: 190138017
  dataset_size: 215384021.75
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: val
    path: data/val-*
---