metadata
dataset_info:
features:
- name: image
dtype: image
- name: query
sequence: string
- name: answers
sequence: string
- name: ground_truth
dtype: string
splits:
- name: train
num_bytes: 12155468
num_examples: 80
- name: test
num_bytes: 1601968
num_examples: 10
download_size: 13721466
dataset_size: 13757436
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*