metadata
dataset_info:
features:
- name: image
dtype: image
- name: caption
dtype: string
- name: conditional_image
dtype: image
splits:
- name: train
num_bytes: 4302977033.5
num_examples: 8178
download_size: 4280369713
dataset_size: 4302977033.5
configs:
- config_name: default
data_files:
- split: train
path: data/train-*