Datasets:
metadata
license: apache-2.0
dataset_info:
features:
- name: image
dtype: image
- name: are_images_connected
dtype: bool
- name: difficulty_in_understanding
dtype: string
- name: is_text_in_left_image
dtype: bool
- name: is_text_in_right_image
dtype: bool
- name: left_image
dtype: string
- name: overall_description
dtype: string
- name: right_image
dtype: string
splits:
- name: train
num_bytes: 30501506
num_examples: 283
download_size: 30080016
dataset_size: 30501506
configs:
- config_name: default
data_files:
- split: train
path: data/train-*