metadata
dataset_info:
features:
- name: context
dtype: string
- name: keyword
dtype: string
- name: model_name
dtype: string
- name: base_prompt
dtype: string
- name: parse_prompt
dtype: string
- name: base_image
dtype: image
- name: our_image
dtype: image
- name: base_image_score
dtype: float64
- name: our_image_score
dtype: float64
splits:
- name: train
num_bytes: 99414620
num_examples: 103
download_size: 97261869
dataset_size: 99414620
configs:
- config_name: default
data_files:
- split: train
path: data/train-*