metadata
dataset_info:
features:
- name: image
dtype: string
- name: conditioning
dtype: image
- name: caption
dtype: string
- name: hed
dtype: image
- name: thinning_hed
dtype: image
splits:
- name: train
num_bytes: 2077553129.313
num_examples: 12243
download_size: 2091159317
dataset_size: 2077553129.313
configs:
- config_name: default
data_files:
- split: train
path: data/train-*