metadata
dataset_info:
features:
- name: image
dtype: image
- name: query
dtype: string
- name: answers
dtype: string
splits:
- name: train
num_bytes: 1496313092.09
num_examples: 56170
- name: test
num_bytes: 439826068.787
num_examples: 14043
download_size: 2668770177
dataset_size: 1936139160.8769999
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*