test latent fix

pipeline.py

@@ -617,7 +617,7 @@ class StableDiffusionLongPromptWeightingPipeline(StableDiffusionPipeline):
             extra_step_kwargs["generator"] = generator
         return extra_step_kwargs
 
-    def prepare_latents(self, image, timestep, batch_size, height, width, dtype, device, generator,upscale=False, latents=None):
+    def prepare_latents(self, image, timestep, batch_size, height, width, dtype, device, generator, latents=None):
         if image is None:
             shape = (
                 batch_size,
@@ -633,17 +633,13 @@ class StableDiffusionLongPromptWeightingPipeline(StableDiffusionPipeline):
                 else:
                     latents = torch.randn(shape, generator=generator, device=device, dtype=dtype)
             else:
-                print(latents)
                 if latents.shape != shape:
                     raise ValueError(f"Unexpected latents shape, got {latents.shape}, expected {shape}")
                 latents = latents.to(device)
 
-                # if upscale:
-                #         return latents, None, None
             # scale the initial noise by the standard deviation required by the scheduler
-                # else:
-                latents = latents * self.scheduler.init_noise_sigma
-                return latents, None, None
+            latents = latents * self.scheduler.init_noise_sigma
+            return latents, None, None
         else:
             init_latent_dist = self.vae.encode(image).latent_dist
             init_latents = init_latent_dist.sample(generator=generator)
@@ -677,7 +673,6 @@ class StableDiffusionLongPromptWeightingPipeline(StableDiffusionPipeline):
         eta: float = 0.0,
         generator: Optional[torch.Generator] = None,
         latents: Optional[torch.FloatTensor] = None,
-        upscale: bool = False,
         return_latents: bool = False,
         max_embeddings_multiples: Optional[int] = 3,
         output_type: Optional[str] = "pil",
@@ -812,9 +807,7 @@ class StableDiffusionLongPromptWeightingPipeline(StableDiffusionPipeline):
             dtype,
             device,
             generator,
-            upscale,
-            latents
-            
+            latents,
         )
 
         # 7. Prepare extra step kwargs. TODO: Logic should ideally just be moved out of the pipeline
@@ -1030,7 +1023,6 @@ class StableDiffusionLongPromptWeightingPipeline(StableDiffusionPipeline):
             generator=generator,
             latents=latents,
             return_latents=True,
-            upscale=False,
             max_embeddings_multiples=max_embeddings_multiples,
             output_type=output_type,
             return_dict=return_dict,
@@ -1040,7 +1032,7 @@ class StableDiffusionLongPromptWeightingPipeline(StableDiffusionPipeline):
         )
 
         latents = torch.nn.functional.interpolate(latents, size=(int(height*resize_scale)//8, int(width*resize_scale)//8))
-        co
+
         return self.__call__(
             prompt=prompt,
             negative_prompt=negative_prompt,
@@ -1054,7 +1046,6 @@ class StableDiffusionLongPromptWeightingPipeline(StableDiffusionPipeline):
             generator=generator,
             latents=latents,
             return_latents=False,
-            upscale=True,
             max_embeddings_multiples=max_embeddings_multiples,
             output_type=output_type,
             return_dict=return_dict,