metadata
dataset_info:
features:
- name: image
dtype: image
- name: conditioning_image
dtype: image
- name: text
dtype: string
- name: label
sequence: string
- name: box2d_pro
sequence:
sequence:
sequence: int64
splits:
- name: train
num_bytes: 24149199949.632
num_examples: 213792
download_size: 23239060092
dataset_size: 24149199949.632
configs:
- config_name: default
data_files:
- split: train
path: data/train-*