metadata
dataset_info:
features:
- name: target_image
dtype: image
- name: target_pose
dtype: image
- name: target_garment
dtype: image
- name: reference_image_1
dtype: image
- name: reference_image_2
dtype: image
- name: prompt
dtype: string
- name: target_group_key
dtype: string
- name: reference_group_key_1
dtype: string
- name: reference_group_key_2
dtype: string
- name: cluster_id
dtype: int32
- name: target_path
dtype: string
- name: target_pose_path
dtype: string
- name: target_seg_path
dtype: string
- name: ref1_path
dtype: string
- name: ref2_path
dtype: string
splits:
- name: train
num_bytes: 66873261714.12
num_examples: 91280
download_size: 19285546622
dataset_size: 66873261714.12
configs:
- config_name: default
data_files:
- split: train
path: data/train-*