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: 6762481
num_examples: 120
- name: test
num_bytes: 179724
num_examples: 4
download_size: 5488007
dataset_size: 6942205
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*