metadata
dataset_info:
features:
- name: image
dtype: image
- name: prefix
dtype: string
- name: subject
dtype: string
- name: illumination
dtype: string
- name: background
dtype: string
- name: prompt
dtype: string
- name: image_embedding
sequence: float64
splits:
- name: train
num_bytes: 679360
num_examples: 10
download_size: 681428
dataset_size: 679360
configs:
- config_name: default
data_files:
- split: train
path: data/train-*