Evaluating models without test data (Practical AI #194)
Changelog Master Feed · Changelog Media
Audio is streamed directly from the publisher (op3.dev) as published in their RSS feed. Play Podcasts does not host this file. Rights-holders can request removal through the copyright & takedown page.
Show Notes
WeightWatcher, created by Charles Martin, is an open source diagnostic tool for analyzing Neural Networks without training or even test data! Charles joins us in this episode to discuss the tool and how it fills certain gaps in current model evaluation workflows. Along the way, we discuss statistical methods from physics and a variety of practical ways to modify your training runs.
Changelog++ members support our work, get closer to the metal, and make the ads disappear. Join today!
Featuring:
- Charles Martin – GitHub, LinkedIn, X
- Chris Benson – Website, GitHub, LinkedIn, Bluesky, X
- Daniel Whitenack – Website, GitHub, X
Show Notes:
- WeightWatcher
- Talk from the Silicon Valley ACM meetup
- A deep dive into the theory behind WeightWatcher (a talk from ENS)
Something missing or broken? PRs welcome!