metadata
dataset_info:
features:
- name: file_name
dtype: string
- name: image
dtype: image
- name: caption_sd15
dtype: string
- name: caption_sdxl
dtype: string
- name: caption_flux
dtype: string
- name: caption_sd3
dtype: string
splits:
- name: train
num_bytes: 20879485
num_examples: 34
download_size: 18988855
dataset_size: 20879485
configs:
- config_name: default
data_files:
- split: train
path: data/train-*