metadata
dataset_info:
features:
- name: qid
dtype: string
- name: questions
dtype: image
- name: choices
dtype: image
- name: answer
dtype: string
- name: difficulty_level
dtype: string
- name: gt_info
dtype: string
splits:
- name: train
num_bytes: 89501424
num_examples: 80
download_size: 79918792
dataset_size: 89501424
configs:
- config_name: default
data_files:
- split: train
path: data/train-*