metadata
dataset_info:
features:
- name: id
dtype: string
- name: image
struct:
- name: bytes
dtype: binary
- name: path
dtype: string
- name: question
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 472244978
num_examples: 709
download_size: 392154237
dataset_size: 472244978
configs:
- config_name: default
data_files:
- split: train
path: data/train-*