Why didn't loss function receive the return from the layer "self.fc"?

[diogofreitasribeiro]

In file nn.py on line 79, the loss function receives the variable temp_out "loss_fn(temp_out, y)", but why didn't loss function receive the variable "out" returned from the layer "out = out + self.fc(temp_out)", instead of "temp_out", then "loss_fn(out, y)"?

[Victordmz]

I feel like your question is justified. However, couldn't it be that instead of temp_out it should be the value of self.fc(temp_out), as evaluated in line 77?

This is because if you choose out, you would do the addition twice and the averaging only once.