Toggle navigation
OpenReview
.net
Login
×
Go to
ICML 2021
homepage
Better Training using Weight-Constrained Stochastic Dynamics
Benedict J. Leimkuhler
,
Tiffany J. Vlaar
,
Timothée Pouchon
,
Amos J. Storkey
2021 (modified: 23 Apr 2023)
ICML 2021
Readers:
Everyone
Abstract:
We employ constraints to control the parameter space of deep neural networks throughout training. The use of customised, appropriately designed constraints can reduce the vanishing/exploding gradie...
0 Replies
Loading