This presentation offers a practical walkthrough of fine-tuning models in Azure AI Studio, providing AI developers with a comprehensive understanding of dataset preparation and its application to models such as Meta's Llama 2. Drawing from a collaborative effort with UC Berkeley on their influential paper "RAFT: RAFT: Adapting Language Model to Domain Specific RAG" (blog), we will demonstrate the application of fine tuning by applying RAFT and use AI Studio's fine-tuning feature which abstracts away fine-tuning and allows us to focus on the dataset preparation. Additionally, we explore methods for evaluating the results of fine-tuning.