metadata
dataset_info:
features:
- name: image
dtype: image
- name: images
list: image
- name: question
dtype: string
- name: question_type
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 583127766
num_examples: 400
download_size: 517658955
dataset_size: 583127766
configs:
- config_name: default
data_files:
- split: train
path: data/train-*