metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: image
dtype: image
- name: question
dtype: string
- name: answer
dtype: string
- name: explanation
dtype: string
- name: image_path
dtype: string
- name: image_id
dtype: int64
- name: question_id
dtype: string
- name: type
dtype: int64
- name: split
dtype: string
splits:
- name: train
num_bytes: 5121597
num_examples: 100
download_size: 5113628
dataset_size: 5121597