llm-cookbook

Why we included this project

The Datawhale community took Andrew Ng's LLM course series, translated it into Chinese, and reworked it into a curriculum for developers who want a structured path into building with large language models. It's not a single tool but a sequenced set of eleven courses that move from prompt engineering through chat systems, LangChain, retrieval-augmented generation, and model fine-tuning, each with runnable Jupyter notebooks. The courses split into required and elective tracks, so a beginner can follow a clear order and then branch into whatever topics matter for their work. For Chinese-speaking developers it's especially valuable because the original material is hard to access locally, and the project ships Chinese prompts that behave comparably to the English originals. Teams bringing new members up to speed on LLM application development will find it a practical, self-paced starting point rather than a production dependency.