---
dataset_info:
  features:
  - name: id
    dtype: string
  - name: manipulate
    dtype: int64
  - name: manipulation type
    dtype: string
  - name: caption
    dtype: string
  - name: image_path
    dtype: string
  splits:
  - name: train
    num_bytes: 6121391
    num_examples: 31000
  download_size: 3030397
  dataset_size: 6121391
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---