"logvar" parameter in discriminator loss can't receive any gradient

[Jaeger416]

During train causalvae, It seems that the "logvar" parameter in discriminator loss can't receive any gradient, since opt1.zero_grad() clear the gradient from gan loss and logvar is not used in discloss

opt1.zero_grad()
        self.manual_backward(aeloss)
        self.clip_gradients(opt1, gradient_clip_val=1, gradient_clip_algorithm="norm")
        opt1.step()
        # ---- GAN Loss ----
        discloss, log_dict_disc = self.loss(
            inputs,
            reconstructions,
            posterior,
            1,
            self.global_step,
            last_layer=self.get_last_layer(),
            split="train",
        )
        self.log(
            "discloss",
            discloss,
            prog_bar=True,
            logger=True,
            on_step=True,
            on_epoch=True,
        )
        opt2.zero_grad()
        self.manual_backward(discloss)