教學大綱與進度

課程基本資料：

學年期	課號	課程名稱	階段	學分	時數	修	教師	班級	人	撤	備註
113-2	345313	生成式AI技術與應用	1	3.0	3	★	傅子恒	互動三	49	2	外生三與大三合開,限本系生35人

教學大綱與進度：

教師姓名	傅子恒　教師諮商時間(Office Hours)
Email	jomo0825@gmail.com
最後更新時間	2025-02-16 13:06:22
課程大綱	The breakthrough in Generative AI (GAI) technology has brought about a significant transformation in multimedia content design, with AI-generated videos, images, and text approaching the proficiency level of professionals. This course focuses on two major technologies: Large Language Models (LLM) and Diffusion Models, utilizing open-source tools such as Llama and Stable Diffusion for hands-on practice. In addition to providing an in-depth and comprehensive explanation of the working principles of LLMs and Diffusion Models, the course also teaches students the fundamental operation of AI and how to efficiently produce high-quality and creative AI-generated content. By mastering generative AI technology and its applications, students will be able to embed AI-generated multimedia content into various information systems, significantly enhancing system functionality, aesthetics, and user experience. This course aims to equip students with solid skills in learning, research, and application, achieving a human-AI co-creation of hypermedia content. * Both theories and applications will be covered. * Use Python + Keras for coding in W01~W06.
課程進度	02/20 W1 Overview of Generative AI 02/27 W2 Introduction to Diffusion Models 03/06 W3 Clip Prompt Engineering 03/13 W4 Fine-tuning a Diffusion Model – TI & DB 03/20 W5 ControlNet 03/27 W6 LLM applications - Reading/Writing/Coding 04/03 W7 Introduction to LLMs – Prompt engineering 04/10 W8 LLM Prompt Engineering 04/17 W9 Mid-term Week 04/24 W10 LLM Fine-tuning & RAG 05/01 W11 LLM AI Agent 05/08 W12 Basics of Generative AI (1) – Dense Network 05/15 W13 Basics of Generative AI (2) – Auto encoder, VAE 05/22 W14 Basics of Generative AI (3) – LSTM 05/29 W15 Basics of Generative AI (4) – Transformer 06/05 W16 Basics of Generative AI (5) – Diffusion model 06/12 W17 PBL 06/19 W18 Final Week
評量方式與標準	* Participation: 10% * Mid-term: 40% * Final: 50%
使用教材、參考書目或其他	【遵守智慧財產權觀念，請使用正版教科書，不得使用非法影印教科書】 使用外文原文書：是 https://www.amazon.com/Generative-Deep-Learning-Teaching-Machines-dp-1098134184/dp/1098134184/ref=dp_ob_image_bk
課程諮詢管道	Will launch an MS Teams group when the class starts.
課程對應SDGs指標	SDG9：產業創新與基礎設施（Industry, Innovation and Infrastructure）
課程是否導入AI
備註	* You'll need to run AI on GPU. For running on a local Windows system, a 8GB VRAM graphic card is required. RTX 2070 8GB * If you don't have a qualified GPU, you may consider to purchase Runpod or Google Colab. For each week roughly USD$10 expense is estimated. Actual expenses fluctuate according to personal exercise usage.