4 ms·PaddleOCR layout works, and so do some open source large language vision modelsby ilaksh 2y agoPaddleOCR layout works, and so do some open source large language vision models