This is the course page for PalmDrive Introduction to GPT and LLM for High School Students If you have interests for this course, please chat with PalmDrive
English | 简体中文
By reading literature, using online, analyzing and comparing different ChatGPT-like products, you can understand the hottest and most commercially valuable artificial intelligence direction. Through two different characteristic projects (report & python), students with different programming foundations can equip with (1) some knowledge about NLP, LLM and their applications and/or (2) Apps based on ChatGPT-like products.
-
Dr. Sherry Lixue Cheng (Full-time, Lecture + all projects). Research interests: AI4Sci(科学人工智能). Google scholar: Link
-
Jiace Sun (Part-time, only assist python projects). Research interests: Quantum computing, AI4Sci. Google scholar: Link
A ChatGPT account: Register at https://chat.openai.com/ (Note: OpenAI does not provide serivces in the Chinese area, you have to have an non-Chinese IP to access ChatGPT & other OpenAI products)
-
Report project: Basic understanding of scientific writing & comparative studies & data analysis.
-
Python project: Expertise in Python (need to take a coding test) & basic understanding of machine learning.
Introduction to basic usage of GitHub
Lec 0: GitHub usage guidelines and coding practices for open-source packages
Previous year archived course pages: None
| Topic | Content | Homework/Discussion |
|---|---|---|
| Introduction | Lec 1: Introduction to artificial intelligence (AI) and Deep Learning (DL) Lec 2: Introduction to natural language processing (NLP) & large language models (LLM) from different companies [GPT4 (OpenAI), Bard (Google), LLaMA (Meta) & Ernie (Baidu)] |
- Python project: Setup GitHub page & create a hello-world repo - Report project: Work on a shot survey on AGI (artificial general intelligence) |
| GPT & ChatGPT | Lec 3: Intro to OpenAI, GPT, ChatBot,& ChatGPT Lec 4.1: How to better use ChatGPT for different purposes: Step-by-step demonstrations & Prompts Lec 4.2: Coding practices |
Generating a paper summary using ChatGPT with series of prompts ** 1-1 meeting to discuss personal projects** Paper reading: Sparks of AGI, GPT3, Math ability, Translation ability, Integrity & Ethics, Other large models |
| GPT-based apps | Lec 5: Python & report project illustration Lec 6: Successful ChatGPT-based products & open-source applications: Microsoft products as examples (New Bing & Copilots) |
- Python project: Push the current progress of your project to GitHub - Report project: Write & analyze some selected GPT experiments (Refer to the style of Sparks of AGI paper) |
| Other LLMs | Lec 7: Limitations of current LLMs & future direction to AGI | ** 1-1 meeting to discuss the project progress** - Python project: Wrap-up codes & Debug - Report project: Half-done essay submission |
| Optional | * Poster/presentation: 5-10 mins to show your work | Submit your name by the end of Lec 7 |
Please see the suggested_reading folder. Optional and recommended for advanced students
- Choice 1: Using GPT-like models/apps to develop your own app (Advanced). Examples: ChatGPT Next Web, ChatBox, ChatGPT Academic, wechat-chatgpt, ...
- Choice 2: Use Chatterbot to design your own chatbot
- Any other ideas are welcome
- Examine the ability of GPT-like models/products in a selected fields. Analyze & compare their performances. Explore their advantages & limitations. Foresee the path to AGI.
- A potential publication if the report is in a high-quality.
- Any other ideas are welcome
Please see the document project_ideas.pdf
For the students who complete both poster/presentation and report/python projects, it is possible to work together with the instructors on developmenst of open-source software based on LLM or AI4Sci research.