metadata
dataset_info:
features:
- name: index
dtype: string
- name: raw_prompt
dtype: string
- name: ground_truth
dtype: image
- name: gen_images_sd1_4
list: image
- name: rephrased_prompt_gpt
dtype: string
- name: gen_images_rephrased_sd1_4
list: image
- name: url
dtype: string
splits:
- name: train
num_bytes: 1564800318
num_examples: 500
download_size: 1560238374
dataset_size: 1564800318
configs:
- config_name: default
data_files:
- split: train
path: data/train-*