This session provides a quick guide to essential tools for training and accelerating the inference of open-weight large language models (LLMs). We will cover an overview of open-weight LLMs, the Pytorch Profiler for performance optimisation, and techniques such as torch, compile, and quantisation to improve model efficiency. Attendees will gain practical insights into how to optimise their models using these tools. The session will include a Q&A session for further discussion. It is ideal for data scientists, machine learning engineers, and AI enthusiasts.
Ivan Gentile is a data scientist at the non-profit foundation IFAB in Bologna. He specialises in machine learning on HPC infrastructures and various aspects of LLM. With a background in physics, mathematics and complexity science, Ivan’s passion for philosophy led him to explore the vast depths of the IT world. Ivan was a Fellow of the Pi School of AI Session 13.
Follow us on Pi School’s social media channels and keep updated on upcoming events.