metadata
language:
- en
dataset_info:
features:
- name: qid
dtype: string
- name: question
dtype: string
- name: images
sequence: image
- name: answer
dtype: string
- name: question_info
dtype: string
- name: type
dtype: string
- name: choices
sequence: string
splits:
- name: train
num_bytes: 15217283
num_examples: 591
download_size: 7561357
dataset_size: 15217283
configs:
- config_name: default
data_files:
- split: train
path: data/train-*