---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 37319673874.68
    num_examples: 3016640
  download_size: 36766631030
  dataset_size: 37319673874.68
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---