---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: caption
    dtype: string
  - name: source
    dtype: string
  splits:
  - name: train
    num_bytes: 60542049225.0
    num_examples: 300000
  - name: val_small
    num_bytes: 258013573.0
    num_examples: 5000
  - name: val_medium
    num_bytes: 306308363.0
    num_examples: 5000
  - name: val_large
    num_bytes: 2466625345.0
    num_examples: 5000
  download_size: 63453556176
  dataset_size: 63572996506.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: val_small
    path: data/val_small-*
  - split: val_medium
    path: data/val_medium-*
  - split: val_large
    path: data/val_large-*
---