🚀 Fine-Tuning Multimodal LLM with Unsloth AI

Presented by Google Developer Groups on Campus - Asia Pacific University Chapter

📅 Event Details

• Date: Sunday, November 16th, 2025

• Time: 1:30 PM - 3:00 PM (MYT)

• Format: Online Workshop

• Duration: 90 minutes

• Department: GDGoC AI Department

• Workshop Lead: Samantha Tee Ker Xin

🎯 About This Workshop

Unlock the power of efficient AI model fine-tuning! This hands-on workshop introduces participants to Unsloth AI, a revolutionary platform that makes fine-tuning large language models accessible, fast, and cost-efficient—even for those with limited computing resources.

Traditionally, fine-tuning LLMs requires significant computational power and resources, making it challenging for students and individual developers. Unsloth AI changes the game by offering a free and resource-efficient approach through quantization and adapter-based methods, democratizing access to advanced AI model customization.

👥 Who Should Attend?

This workshop is perfect for:

• Students & AI/ML Beginners seeking practical, hands-on exposure to LLM fine-tuning

• Developers looking for resource-efficient methods to adapt LLMs for vision and text tasks

• AI Enthusiasts interested in exploring Unsloth AI's advantages for real-world model training

• Anyone curious about making AI more accessible and efficient!

Expected Attendance: 60+ participants

Prerequisites:

• Basic Python knowledge (helpful but not mandatory)

• Some familiarity with AI/ML concepts (beginner-friendly!)

• A Google account for Colab access

• A Hugging Face account (Please Register Prior To The Workshop)

📚 What You'll Learn

By the end of this workshop, you will:

✅ Understand the landscape of LLM fine-tuning (Full Fine-Tuning vs. PEFT)
✅ Discover Unsloth AI and its benefits for efficient model training
✅ Gain hands-on experience fine-tuning Gemma 3n E2B with an image dataset
✅ Explore alternatives for LLM customization beyond fine-tuning

⏱️ Workshop Agenda

1:00 PM - 1:35 PM | Welcome & Introduction

• Workshop objectives and overview

• Introduction to the AI landscape

1:35 PM - 2:00 PM | Fine-Tuning & Unsloth AI Deep Dive

• Full fine-tuning vs. Parameter-Efficient Fine-Tuning (PEFT)

• Understanding Adapters, LoRA, and QLoRA

• What makes Unsloth AI special?

  • Speed optimization

  • Memory efficiency

  • Sensitivity-aware quantization

2:00 PM - 2:45 PM | Hands-On Coding Session 🛠️

Live walkthrough covering:

• Dataset preprocessing

• Model loading

• Fine-tuning with Unsloth AI

• Model inferencing

• Uploading your fine-tuned model

2:45 PM - 2:55 PM | Beyond Fine-Tuning

• Quick exploration of LLM customization alternatives

• Prompting strategies for optimized results

2:55 PM - 3:00 PM | Wrap-Up

• Key takeaways recap

• Open Q&A session

🎁 What You'll Receive

All participants will get:

• Workshop Notes

• Hands-On Colab Notebook

• Recording

🛠️ Tools & Platforms

• Google Colab (with free T4 GPU)

• Hugging Face

• Unsloth AI

📝 Registration

Space is limited! Register now to secure your spot in this exciting hands-on workshop.

Join us to discover how Unsloth AI is making advanced AI accessible to everyone, regardless of computational resources. Whether you're taking your first steps in AI or looking to level up your LLM skills, this workshop offers valuable insights and practical experience.

📧 Questions?

For inquiries, please contact the any of our social media handles!

See you there!

Organized by Google Developer Groups on Campus - Asia Pacific University Chapter