---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: caption
    dtype: string
  splits:
  - name: train
    num_bytes: 15942685868.77317
    num_examples: 53072
  - name: test
    num_bytes: 597799789.0
    num_examples: 2000
  download_size: 16531331722
  dataset_size: 16540485657.77317
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---