dive-into-llms
About
"Dive into LLMs" is a free, public welfare programming practice tutorial series, extended from Shanghai Jiao Tong University course materials, aiming to provide an introductory programming reference for large language models (LLMs). The tutorials cover core LLM technologies, including fine-tuning and deployment, prompt learning, knowledge editing, mathematical reasoning, model watermarking, jailbreak attacks, LLM steganography, multimodal models, GUI agents, agent safety, and RLHF alignment. This project also collaborates with the Huawei Ascend Community to offer a comprehensive "LLM Development Full Process" tutorial series, providing diverse teaching resources like PPTs, lab manuals, and videos, dedicated to guiding researchers and developers through the full process of LLM development based on domestic hardware and software.
Features
- Comprehensive LLM programming tutorials
- Covers core LLM technologies (fine-tuning, prompting, agents, security)
- Free, open-source with rich resources (slides, guides, scripts)
- Collaboration with Huawei Ascend for full-stack domestic LLM development guidance
- Derived from Shanghai Jiao Tong University course materials