Hi,
I have a model and it is composed of two parts. The first part is a DCGAN architecture, randomly initialized by me, and the second part is a pre-trained Image Recognition model (something like resnet52). I have joined these two parts in a nn.Module class. What I want to do is to freeze the resent52 part and just train the DCGAN one. Here is my sample code.
class Test(nn.Module):
def __init__(self, dis):
super().__init__()
# define generator here
self.gen = Generator()
# use the pretrained classifier
self.dis = dis
def forward(self, input):
...
return ...
I would like to know how is it possible?