Join us for a hands-on workshop designed specifically for researchers and students who want to learn how to fine-tune Large Language Models using your own datasets. This workshop will introduce the fundamentals of model Fine-Tuning vs Retrieval Augmented Generation (RAG) and provide practical insights into how powerful LLMs can be tailored to your own specific data requirements.

Expect a casual pace as we work through Python Notebooks to demonstrate data wrangling and AI workflows. 

Participants will gain a practical understanding of:

• Differentiating Fine-Tuning from RAG: You'll clearly understand the core differences between fine-tuning large language models and using Retrieval Augmented Generation (RAG) for scientific applications, helping you choose the right approach for your research.

• Practical LLM Customisation: You'll learn how to adapt state-of-the-art (SOTA) models like Gemini and Gemma to your specific scientific datasets enabling you to unlock more accurate and relevant results.

• End-to-End AI Workflow for Scientific Data: You'll get practical experience with the entire process, from preparing your own scientific data for model consumption to implementing and evaluating fine-tuned LLMs, all within a Python Notebook environment.

• Optimising LLMs for Scientific Research: You'll gain insights into how fine-tuning can enhance the performance of LLMs on specialised scientific tasks, allowing you to leverage these models more effectively for data analysis, hypothesis generation, and more.

The course will be all hands-on exercises and live coding demonstrations to reinforce your learning and build confidence.

Requirements: 

1. A Google account to use Colab for Python https://colab.research.google.com/

2. An API key from https://aistudio.google.com/apikey to use web based models.

3. A Kaggle account (you can use your Google credentials) to download AI models https://www.kaggle.com/

4. Basic familiarity with Python is recommended - but we will walk through the details.

5. Working in a research related field at an Australian institution. Your ticket will be cancelled if this cannot be verified.