metadata
dataset_info:
features:
- name: image
dtype: image
- name: conditioning
dtype: image
- name: caption
dtype: string
splits:
- name: train
num_bytes: 3324491451
num_examples: 120000
download_size: 2654978833
dataset_size: 3324491451
configs:
- config_name: default
data_files:
- split: train
path: data/train-*