metadata
dataset_info:
features:
- name: id
dtype: string
- name: input_image
dtype: image
- name: weak_prompt
dtype: string
- name: strong_prompt
dtype: string
- name: constraints_explained
dtype: string
- name: tool_augmentation
dtype: string
- name: verdict
dtype: string
- name: ground_truth
dtype: string
- name: image_src
dtype: string
splits:
- name: train
num_bytes: 923608
num_examples: 5
download_size: 929609
dataset_size: 923608
configs:
- config_name: default
data_files:
- split: train
path: data/train-*