张宏纲245页ppt《大语言模型基础导论》（研究生课程）

澳门城市大学张宏纲教授（IEEE Fellow）本学期在澳门城市大学数据科学学院针对研究生开设了《人工智能前沿专题-大语言模型基础导论》课程，该课程的主要内容包括：LLM的基本概念、关键机制和方法、AI Agent、LLM与强化学习、大模型的数学原理、扩散模型、通信专属大模型和NetGPT、以及DeepSeek的若干技术内核。

该ppt共分为8各部分，具体如下：

1. The Background and Basics of Data-Driven Scaling Law, Transformer, andFoundation Models (LLMs/LMMs -Large Language/Multimodal Models);
2. Key Mechanisms and Methods in Data-Driven LLMs/LMMs - Prompt, Fine-Tuning, LoRA, Mixture of Experts (MoE) , Chain of Thoughts (CoT) and RAG;
3. Embodied Al Agents and Data-Driven LLMs/LMMs;
4. Mutual-Play between Data-Driven LLMs/LMMs and Reinforcement Learning;
5. The Mathematical Basics of Generative Pretrained Transformer (GPT) andData-Driven LLMs/LMMS;
6. Diffusion Models, Algorithms and Data-Driven LLMs/LMMS;
7. Telecom LLMs/LMMs, Networked GPT (NetGPT) and other Key Applications;
8. Summary.