Rapid training of deep neural networks without skip connections or normalization layers using Deep Kernel Shaping

Add code
Oct 05, 2021

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: