metadata
dataset_info:
features:
- name: caption
dtype: string
- name: coco_image
dtype: image
- name: sd3_image
dtype: image
- name: sd21_image
dtype: image
- name: sdxl_image
dtype: image
- name: dalle_image
dtype: image
- name: midjourney_image
dtype: image
splits:
- name: train
num_bytes: 3967077819
num_examples: 7000
download_size: 3963922174
dataset_size: 3967077819
configs:
- config_name: default
data_files:
- split: train
path: data/train-*