Overview
Dive into the exciting world of Large Language Models (LLMs) with this comprehensive, hands-on programming tutorial series. Originally developed from university course materials at Shanghai Jiao Tong University, this free, public-access resource is designed to equip students and researchers with practical skills for understanding and building with LLMs. The series has recently expanded to include a dedicated curriculum on the full LLM development lifecycle, developed in collaboration with Huawei Ascend Community.
Key Features
- Practical, Code-First Approach: Learn by doing with guided tutorials, accompanying notebooks, and presentation slides for each topic.
- Broad LLM Curriculum: Covers essential LLM concepts from fine-tuning and deployment to prompt engineering, knowledge editing, mathematical reasoning, and even advanced topics like watermarking, jailbreaking, and multimodal models.
- Expanded "Full LLM Development Lifecycle" Series: A new, collaborative offering with Huawei Ascend, providing in-depth, step-by-step guidance on developing LLMs using Ascend's ecosystem. This series includes PPTs, lab manuals, and video resources for beginner to advanced users.
- Focus on Emerging Techniques: Explores cutting-edge areas like GUI Agents, Model Watermarking, LLM Steganography, and safety alignment (RLHF).
- Community-Driven: Actively welcomes contributions through Issues and Pull Requests, fostering a collaborative learning environment.
Typical Use Cases
- Academic Research & Course Projects: Provides a strong foundation for students in Natural Language Processing, AI Security, and related fields.
- LLM Development & Experimentation: Enables developers to quickly learn how to fine-tune, deploy, and interact with LLMs for specific tasks.
- Exploring LLM Capabilities & Limitations: Understand how LLMs reason, can be manipulated, and how to secure them.
- Gaining Proficiency in Chinese LLM Ecosystem: Leverage the "Full LLM Development Lifecycle" series for practical experience with Chinese AI hardware and software.
- Prototyping AI Applications: Build practical AI agents that can interact with graphical user interfaces.
