Finetuning Open Source LLMs

This course provides a practical introduction to finetuning open source large language models (LLMs, and focuses on applying models to custom and domain specific use cases. The course is intended for data scientists, machine learning engineers, and AI researchers aiming to excel in fine-tuning open-source LLMs tailored for products and applications. We'll delve into training techniques including direct preference optimization (DPO), supervised fine-tuning (SFT), and odds ratio preference optimization (ORPO). Moreover, the course frames training techniques with data-centric model evaluation, benchmarking, and model quantization, to prepare students for productizing models beyond the classroom.