metadata
dataset_info:
features:
- name: query
dtype: string
- name: image
dtype: image
- name: ocr
dtype: string
splits:
- name: train
num_bytes: 938118470.658238
num_examples: 3986
- name: test
num_bytes: 234647294.34176198
num_examples: 997
download_size: 980633723
dataset_size: 1172765765
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*