News for hackers. Stuff that matters.
Re:Empiricism and the limits of gradient descent by PeterisP on Monday May 28, 2018 @03:50AM (#17171144)
"An enormous hyperparameter search and lots of elbow grease to carefully tune that dynamical system" i.e. the classic GDGS (gradient descent by grad student) approach where you have a grad student train a system, decide in which direction the parameters should be updated (i.e. look at the gradient), tweak the system and repeat until convergence.

There are no comments on this comment yet.