---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: caption
    dtype: string
  - name: cid
    dtype: int64
  - name: split
    dtype: string
  - name: clean_image
    dtype: image
  splits:
  - name: train
    num_bytes: 529700241.754
    num_examples: 1667
  - name: val
    num_bytes: 85417082.0
    num_examples: 267
  download_size: 608624569
  dataset_size: 615117323.754
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: val
    path: data/val-*
---