Skip to content

Comment on Optimizing ML training with metagradient descentparent

Comments

That's the advantage of deep learning over traditional ML: if you've got enough data, you don't need domain knowledge or hyperparameter tuning, just throw a large enough universal approximator at it. The challenge lies in generating good enough artificial data for domains without enough data, and getting deep models to perform competitively with simpler models.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.