metadata
dataset_info:
features:
- name: image
dtype: image
- name: questions_and_answers
list:
- name: answer
dtype: string
- name: question
dtype: string
splits:
- name: train
num_bytes: 272363134
num_examples: 2596
download_size: 271278407
dataset_size: 272363134
configs:
- config_name: default
data_files:
- split: train
path: data/train-*