---
dataset_info:
  features:
  - name: image_1
    dtype: image
  - name: image_2
    dtype: image
  - name: differences
    list: string
  - name: source
    dtype: string
  splits:
  - name: train
    num_bytes: 1083101366
    num_examples: 513
  download_size: 1082132807
  dataset_size: 1083101366
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---