metadata
dataset_info:
features:
- name: id
dtype: string
- name: mask
dtype: image
- name: caption
dtype: string
- name: generated_image
dtype: image
splits:
- name: train
num_bytes: 2231455539.5
num_examples: 1500
download_size: 1863976143
dataset_size: 2231455539.5
configs:
- config_name: default
data_files:
- split: train
path: data/train-*