Can you add *If* statements in feedforward pass of NN - Pytorch

I am trying to test multiple different models of an MLP (dropouts or no dropouts, depth, size of layers) and was thinking instead of manually creating a different MLP for each run I would add if logics when defining my nn architecture, such as this snippet;

#Define Forwards Pass
def forward(self, x):
if Dropout = 0:
  if num_of_layers = 4:
    # flatten image input
    x = x.view(-1,28*28)
    x = self.fc1(x)
    x = F.relu(x)enter code here

Does pytorch let you do such a thing, because I got a Syntax Error for my first if statement?