complex_shift_autoencoder.py [282:297]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
                N += bs
                losses.append(loss_batch)
        test_loss = torch.stack(losses).sum() / float(N)
        self.encoder.train()
        self.decoder.train()
        return test_loss

    def compute_mean_loss(self, loss_func, data_loader):
        """Computes RMSE based on given loss function."""
        self.encoder.eval()
        self.decoder.eval()
        losses = []
        with torch.no_grad():
            for i, (x1, x2, angles) in enumerate(data_loader):
                x1 = x1.to(device=self.device)
                x2 = x2.to(device=self.device)
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



weakly_complex_shift_autoencoder.py [349:366]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
                N += bs
                losses.append(loss_batch)
                
        test_loss = torch.stack(losses).sum() / float(N)
        
        self.encoder.train()
        self.decoder.train()
        return test_loss

    def compute_mean_loss(self, loss_func, data_loader):
        """Computes RMSE based on given loss function."""
        self.encoder.eval()
        self.decoder.eval()
        losses = []
        with torch.no_grad():
            for i, (x1, x2, angles) in enumerate(data_loader):
                x1 = x1.to(device=self.device)
                x2 = x2.to(device=self.device)
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -