1

Suppose I have two layers:

x1 = Conv3D(16, kernel_size=(3, 1, 1), activation='relu', dilation_rate=2)(x)
x2 = Conv3D(16, kernel_size=(3, 1, 1), activation='relu', dilation_rate=3)(x)

Is there a way to share weights between these two layers?

DikobrAz
  • 3,557
  • 4
  • 35
  • 53
  • 2
    I think you can try the solution [here](https://stackoverflow.com/q/45245396/1531463). I haven't tested on `Conv3D` and `dilation_rate`, but I assume that they should be the same. – Yu-Yang Sep 21 '17 at 03:54

0 Answers0