Update pipeline_semantic_stable_diffusion_xl_img2img_ddpm.py

pipeline_semantic_stable_diffusion_xl_img2img_ddpm.py

@@ -829,7 +829,7 @@ class SemanticStableDiffusionXLImg2ImgPipeline_DDPMInversion(DiffusionPipeline,
         prompt_2: Optional[Union[str, List[str]]] = None,
         height: Optional[int] = None,
         width: Optional[int] = None,
-        #num_inference_steps: int = 50,
+        num_inference_steps: int = 50,
         #denoising_end: Optional[float] = None,
         guidance_scale: float = 5.0,
         negative_prompt: Optional[Union[str, List[str]]] = None,
@@ -1011,7 +1011,7 @@ class SemanticStableDiffusionXLImg2ImgPipeline_DDPMInversion(DiffusionPipeline,
             `tuple`. When returning a tuple, the first element is a list with the generated images.
         """
         # eta = self.eta
-        num_inference_steps = self.num_inversion_steps
+        # num_inference_steps = self.num_inversion_steps
         num_images_per_prompt = 1
         # latents = self.init_latents
 
@@ -1023,10 +1023,15 @@ class SemanticStableDiffusionXLImg2ImgPipeline_DDPMInversion(DiffusionPipeline,
             self.smoothing = GaussianSmoothing(self.device)
 
         # 0. Default height and width to unet
-        height = self.height
-        width = self.width
-        original_size = self.original_size
-        target_size = self.target_size
+        # height = self.height
+        # width = self.width
+        # original_size = self.original_size
+        # target_size = self.target_size
+        
+        height = height or self.default_sample_size * self.vae_scale_factor
+        width = width or self.default_sample_size * self.vae_scale_factor
+        original_size = original_size or (height, width)
+        target_size = target_size or (height, width)
 
         # 1. Check inputs. Raise error if not correct
         self.check_inputs(