metadata
dataset_info:
features:
- name: image
dtype: image
- name: label
dtype:
class_label:
names:
'0': AR_docs
'1': CN_docs
'2': Latin_docs
'3': non-text
splits:
- name: train
num_bytes: 27290843.67117117
num_examples: 754
- name: test
num_bytes: 4701416.328828828
num_examples: 134
download_size: 31849475
dataset_size: 31992260