In recent years, Artificial Intelligence (AI) has experienced explosive growth, with Large Language Models (LLMs) emerging as one of the most transformative technologies. To help students and enthusiasts understand this revolutionary technology, the Computer Electronic and Software Society (CESS) is organizing an exciting workshop titled Introduction to Large Language Models. This workshop is designed to give participants a comprehensive understanding of how LLMs work, how to implement them, and their applications.
Workshop Schedule
Date:
Session 1: 19th October, 2024
Session 2: 20th October, 2024
Session 3: 26th October, 2024
Session 4: 27th October, 2024
Time:
5 PM to 7 PM
Location:
Main Campus:
Sessions 1-3: Mingde N201
Session 4: Mingde N401
Learning Objectives
Over the course of four comprehensive sessions, participants will acquire essential knowledge and skills, enabling them to work with LLMs effectively.
Week 1 (Sessions 1 & 2)
- Introduction to the fundamental concepts behind Large Language Models (LLMs).
- Essential text data processing techniques for training language models.
Week 2 (Session 3 & 4)
- Implementing the attention mechanism and building a GPT model.
- Training the GPT model and inferencing the trained model to generate outputs.
- Techniques for saving, loading, and reusing trained models.
- Loading pre-trained GPT model weights from OpenAI for practical applications.
Prerequisites
To get the most out of the workshop, participants should ensure they have:
A laptop (self-provided devices are encouraged).
Pre-installed software: Visual Studio Code and Anaconda Navigator.
Workshop Lead
The sessions will be conducted by Khant Hmu Paing, a 2nd-year PhD candidate in Computer Science and Technology. His research areas include learned image compression, super-resolution, and image denoising.
Key Details
Target Audience: Students in Computer Science and Artificial Intelligence-related majors.
Format: Hands-on coding with practical implementation.
Platforms: The workshop will be conducted offline only at the Main Campus. Only 120 students will be selected to join and selection will be based on a registration process after joining the WeChat group.
By the end of the workshop, attendees will have the confidence and technical expertise to:
Build, train, and perform inference with their LLMs.
Utilize pre-trained GPT models from OpenAI to generate text-based applications.
