metadata
dataset_info:
features:
- name: image
dtype: image
- name: conditioning_image
dtype: image
- name: text
dtype: string
- name: label
sequence: string
splits:
- name: train
num_bytes: 18091936016.3
num_examples: 194700
download_size: 21810993407
dataset_size: 18091936016.3
configs:
- config_name: default
data_files:
- split: train
path: data/train-*