About

Style transfer of the author at the Grand Canyon. Style source: Way To Home by Leonid Afremov. I’m Benjamin Warner, a Machine Learning Engineer and AI Researcher. I write about deep learning topics which catch my interest.

I’m a founding employee at Answer.ai, where I work on AI research and development, open-source, and collaborations with other AI researchers and developers.

I have a formal background in economics with a masters from Duke University and a professional background in programming. I am self-taught in machine and deep learning.

# Open-Source Libraries

I created and maintain two open source libraries:

• fastxtend: train fastai models faster (and other useful tools)
• optimī: fast, modern, memory efficient, and low precision PyTorch optimizers.

# Website

This website is powered by Jekyll and is hosted on Cloudflare Pages.

The theme is a custom combination of the philosophy and format of Tufte-Jekyll remixed with Swiss Style. The codebase is primarily from Tufte-Jekyll, with some Swiss Jekyll code, along with numerous additions, customizations, and modifications.

The fonts are Inter by Rasmus Andersson, Neuton by Brian Zick, and Consolas.

# Contact Me

You can reach me on twitter or threads, where I am benjamin_warner and @_benjaminwarner, respectively.