Course based on GitHub - rednote-hilab/dots.ocr: Multilingual Document Layout Parsing in a Single Vision-Language Model