---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: image_id
    dtype: string
  - name: caption
    dtype: string
  - name: modality
    dtype: string
  splits:
  - name: train
    num_bytes: 5537217785.125
    num_examples: 28551
  - name: test
    num_bytes: 248315021.5
    num_examples: 7124
  download_size: 5781578185
  dataset_size: 5785532806.625
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---