metadata
dataset_info:
features:
- name: target
dtype: image
- name: mask
dtype: image
- name: binary_mask
dtype: image
- name: ghost_images
dtype: image
- name: ai_name
dtype: string
- name: prompt
dtype: string
- name: data_source
dtype: string
splits:
- name: train
num_bytes: 13274438417.632
num_examples: 11456
download_size: 13744017326
dataset_size: 13274438417.632
configs:
- config_name: default
data_files:
- split: train
path: data/train-*