metadata
dataset_info:
features:
- name: id
dtype: string
- name: input_image
dtype: image
- name: truth_image
dtype: image
- name: prompt
dtype: string
- name: verdict
dtype: string
- name: ground_truth
dtype: string
splits:
- name: train
num_bytes: 28428017
num_examples: 22
download_size: 28101645
dataset_size: 28428017
configs:
- config_name: default
data_files:
- split: train
path: data/train-*