If we combine one trainable parameters with a non-trainable parameter, is the original trainable param trainable?

[Brando_Miranda (MirandaAgent)]

How does one delete layers in a Neural Network?

I was trying to delete parameters and then set them to something different but I get the following error: AttributeError: conv0.weight why is that? Why does pytorch return such an attribute if it doesn’t let you delete it later…? e.g. code: for name,param in params: if name in dict_params_place_holder: param_original = dict_params_original[name] delattr(self.net_place_holder, name)