---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: seg
    dtype: image
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 31089648080.43
    num_examples: 20210
  download_size: 31282897836
  dataset_size: 31089648080.43
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---