---
dataset_info:
  features:
  - name: id
    dtype: string
  - name: input_images
    list: image
  - name: masked_target
    dtype: image
  - name: target
    dtype: image
  splits:
  - name: train
    num_bytes: 13688383453
    num_examples: 22316
  download_size: 17271077529
  dataset_size: 13688383453
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---