Weight decay in SGD and batch normalization layers

[shaden (Shaden Alshammari)]

Does the weight decay in optim.SGD includes applying penalty on the batch normalization parameters?