How do you implement weight dropout (i.e., “DropConnect”) for linear and convolutional layers? Some simple example code would be extremely helpful.
2 Likes
Still waiting for help on this. Anybody?!
This topic links to an implementation and discusses a custom approach.
1 Like