Fine-Tuning LLMs: Customising AI for Your Use Case

Public lesson preview

Learn how to take a pre-trained large language model and specialise it for a specific domain or task through fine-tuning.

Learning objectives

By the end of this lesson, you will understand:

• ✓🗺️ Three Ways to Specialise a Model
• ✓🔬 What Fine-Tuning Actually Does
• ✓⚡ Full Fine-Tuning vs Parameter-Efficient Fine-Tuning
• ✓📊 The Training Data Challenge
• ✓🔄 RLHF: Fine-Tuning With Human Feedback

Lesson outline

1. 🗺️ Three Ways to Specialise a Model
2. 🔬 What Fine-Tuning Actually Does
3. ⚡ Full Fine-Tuning vs Parameter-Efficient Fine-Tuning
4. LoRA (Low-Rank Adaptation)
5. QLoRA (Quantised LoRA)
6. 📊 The Training Data Challenge
7. 🔄 RLHF: Fine-Tuning With Human Feedback
8. 🛠️ Practical Steps
9. 🚫 When NOT to Fine-Tune