denormalize latents with the mean and std if available (#7111)

* denormalize latents with the mean and std if available

* fix denormalize

* add latent mean and std in vae config

* address sayak's comment

src/diffusers/models/autoencoders/autoencoder_kl.py

@@ -80,6 +80,8 @@ class AutoencoderKL(ModelMixin, ConfigMixin, FromOriginalVAEMixin):
         norm_num_groups: int = 32,
         sample_size: int = 32,
         scaling_factor: float = 0.18215,
+        latents_mean: Optional[Tuple[float]] = None,
+        latents_std: Optional[Tuple[float]] = None,
         force_upcast: float = True,
     ):
         super().__init__()

src/diffusers/pipelines/stable_diffusion_xl/pipeline_stable_diffusion_xl.py

@@ -1313,7 +1313,22 @@ class StableDiffusionXLPipeline(
                 self.upcast_vae()
                 latents = latents.to(next(iter(self.vae.post_quant_conv.parameters())).dtype)
 
-            image = self.vae.decode(latents / self.vae.config.scaling_factor, return_dict=False)[0]
+            # unscale/denormalize the latents
+            # denormalize with the mean and std if available and not None
+            has_latents_mean = hasattr(self.vae.config, "latents_mean") and self.vae.config.latents_mean is not None
+            has_latents_std = hasattr(self.vae.config, "latents_std") and self.vae.config.latents_std is not None
+            if has_latents_mean and has_latents_std:
+                latents_mean = (
+                    torch.tensor(self.vae.config.latents_mean).view(1, 4, 1, 1).to(latents.device, latents.dtype)
+                )
+                latents_std = (
+                    torch.tensor(self.vae.config.latents_std).view(1, 4, 1, 1).to(latents.device, latents.dtype)
+                )
+                latents = latents * latents_std / self.vae.config.scaling_factor + latents_mean
+            else:
+                latents = latents / self.vae.config.scaling_factor
+
+            image = self.vae.decode(latents, return_dict=False)[0]
 
             # cast back to fp16 if needed
             if needs_upcasting: