Loading pretrained model on CPU using development branch

[NicoRenaud]

When loading a model that has been trained on GPU and loading it on CPU the line 230-240 of NeuralNet.py removes the state_dict keys and therefore we can't load the model.

        # load parameters of pretrained model if provided
        if self.pretrained_model:
            # a prefix 'module.' is added to parameter names if
            # torch.nn.DataParallel was used
            # https://pytorch.org/docs/stable/nn.html#torch.nn.DataParallel
            if self.state['cuda']:
                for paramname in list(self.state['state_dict'].keys()):
                    paramname_new = paramname.lstrip('module.')
                    self.state['state_dict'][paramname_new] = \
                        self.state['state_dict'][paramname]
                    del self.state['state_dict'][paramname]
            self.load_model_params()

I'm not even sure if adding the module. is necessary anymore