metadata
dataset_info:
features:
- name: id
dtype: int64
- name: object1
dtype: string
- name: object2
dtype: string
- name: original_img
dtype: image
- name: edited_img
dtype: image
- name: sc_score
struct:
- name: reasoning for Task1
dtype: string
- name: reasoning for Task2
dtype: string
- name: score
sequence: int64
- name: score for Task1
dtype: int64
- name: score for Task2
dtype: int64
- name: min_sc_score
dtype: int64
splits:
- name: train
num_bytes: 1189272715
num_examples: 500
download_size: 1189198312
dataset_size: 1189272715
configs:
- config_name: default
data_files:
- split: train
path: data/train-*