Joe Penna
Senior Director, Creative Technology
Santa Barbara, California, United States
Summary
🤩
RockstarJoe Penna is a creative technology leader who blends hands-on machine learning engineering with senior product and studio leadership, currently serving as Senior Director, Creative Technology at Netflix. He has rapidly progressed through founding and executive roles at startups and AI companies—most recently VP of Applied ML at Stability AI—bringing a practical focus on deploying generative models and tooling into production. On GitHub he contributed a widely used Dreambooth implementation for Stable Diffusion, optimizing face, object, and style training and adding cloud notebooks and pruning to make models more usable at scale. Based in Santa Barbara, he pairs a filmmaker’s sensibility with engineering rigor to turn research ideas into polished creative experiences. Colleagues know him for moving quickly from prototype to product while keeping an eye for performance and usability details that non-engineers appreciate.
4 years of coding experience
Github Skills (10)
pytorch10
machine-learning10
artificial-intelligence10
stable-diffusion10
image-generation10
im10
python10
ai10
text2image10
jupyter-notebook9
Programming languages (1)
Github contributions (5)
JoePenna/Dreambooth-Stable-Diffusion
Sep 2022 - Dec 2022
Implementation of Dreambooth (https://arxiv.org/abs/2208.12242) by way of Textual Inversion (https://arxiv.org/abs/2208.01618) for Stable Diffusion (https://arxiv.org/abs/2112.10752). Tweaks focused on training faces, objects, and styles.
Role in this project:
ML Engineer
ML EngineerContributions:51 commits, 8 PRs, 42 pushes in 3 months
Contributions summary:Joe primarily contributed to the implementation and optimization of a Dreambooth model for Stable Diffusion, focusing on training faces, objects, and styles. They added a notebook to facilitate running the code on cloud platforms and provided instructions for generating regularization images and preparing training variables. Further contributions included adding a pruning option to reduce the checkpoint size and integrating improvements from other contributors.
inversionstablediffusionmodel-trainingtrainingdiffusion
JoePenna/para_chording
Feb 2022 - Mar 2022
Contributions:3 releases, 24 pushes, 1 branch in 8 days
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.