test in latnent upcale

pipeline.py

@@ -811,12 +811,16 @@ class StableDiffusionLongPromptWeightingPipeline(StableDiffusionPipeline):
 
         # 7. Prepare extra step kwargs. TODO: Logic should ideally just be moved out of the pipeline
         extra_step_kwargs = self.prepare_extra_step_kwargs(generator, eta)
-        print("before denoise latents")
+        # print("before denoise latents")
         print(latents.shape)
         # 8. Denoising loop
         for i, t in enumerate(self.progress_bar(timesteps)):
             # expand the latents if we are doing classifier free guidance
             latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents
+            print("latent_model_input 1st step")
+            # print(latent_model_input)
+            print(latent_model_input.shape)
+
             latent_model_input = self.scheduler.scale_model_input(latent_model_input, t)
 
             # predict the noise residual
@@ -842,7 +846,7 @@ class StableDiffusionLongPromptWeightingPipeline(StableDiffusionPipeline):
                 if is_cancelled_callback is not None and is_cancelled_callback():
                     return None
         print("after first step denoise latents")
-        print(latents)
+        # print(latents)
         print(latents.shape)
         latents = torch.nn.functional.interpolate(
             latents, size=(int(height*resize_scale)//8, int(width*resize_scale)//8))
@@ -851,8 +855,8 @@ class StableDiffusionLongPromptWeightingPipeline(StableDiffusionPipeline):
             # expand the latents if we are doing classifier free guidance
             latent_model_input = torch.cat(
                 [latents] * 2) if do_classifier_free_guidance else latents
-            print("latent_model_input")
-            print(latent_model_input)
+            print("latent_model_input 2nd step")
+            # print(latent_model_input)
             print(latent_model_input.shape)
 
             print("2nd step timestep")
@@ -861,13 +865,13 @@ class StableDiffusionLongPromptWeightingPipeline(StableDiffusionPipeline):
             latent_model_input = self.scheduler.scale_model_input(
                 latent_model_input, t)
             print("latent_model_input after scheduler")
-            print(latent_model_input)
+            # print(latent_model_input)
             print(latent_model_input.shape)
             # predict the noise residual
             noise_pred = self.unet(latent_model_input, t,
                                    encoder_hidden_states=text_embeddings).sample
             print("noise_pred")
-            print(noise_pred)
+            # print(noise_pred)
             print(noise_pred.shape)
             # perform guidance
             if do_classifier_free_guidance: