Optimizing LLMs for Cost and Quality

Below-the-line response quality and prohibitively expensive inference are significant blockers to scaling LLMs today. This technical session will teach you a path using open source to achieve superior quality with cheaper/faster models to power your production applications.

In this session you’ll:

• Understand why fine tuning models is critically important

• Learn a proven “crawl, walk, run” path toward model quality optimization

• See what the continuous development cycle looks like for LLM-powered apps

• Watch a live demo showing how to fine-tune Llama3.1-8B to outperform a GPT-4o model at redacting personally identifiable information (PII) from your enterprise data

Session Resources:

• Slides

• Notebook

• LLM Comparison Dashboard

• Sample Dataset