---
dataset_info:
  features:
  - name: original_prompt
    dtype: string
  - name: original_image
    dtype: image
  - name: edit_prompt
    dtype: string
  - name: edited_prompt
    dtype: string
  - name: edited_image
    dtype: image
  - name: adversarial_image
    dtype: image
  splits:
  - name: train
    num_bytes: 24670304.0
    num_examples: 200
  download_size: 24669571
  dataset_size: 24670304.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---