From Paper to Prod in 1 Business Day: Working with OS Foundation Models
When a new state-of-the-art open-source foundation model is released, the hype is instant. But how do you deploy and evaluate new models quickly?
This talk breaks down the exact process my team and I use to put ML models into production within hours of weights being released.
In an interactive session including slides, live coding, and Q&A, we’ll explore selecting appropriate hardware, packaging models and dependencies, and evaluating model results for recent models like Llama 2 and Stable Diffusion XL.
We’ll cover some pretty technical topics in detail, but this talk is designed to be accessible to anyone with an engineering background and an interest in AI.
Speaker: Philip Kiely
Philip writes code and words at Baseten, a Series A ML infrastructure startup backed by Greylock. He’s the author of Writing for Software developers and Life-Changing Email. A long-time Iowan, Philip is a graduate of Grinnell College and Valley High School and a former Iowa Senate legislative page.
Register