metadata
dataset_info:
features:
- name: original_image
dtype: image
- name: conditioning_image
dtype: image
- name: overlaid
dtype: image
- name: caption
dtype: string
splits:
- name: train
num_bytes: 1730245
num_examples: 21
download_size: 0
dataset_size: 1730245
task_categories:
- text-to-image
- keypoint-estimation
- image-to-image