attempt fix

app.py

@@ -6,6 +6,7 @@ import torchvision.transforms as T
 import torchvision.transforms.v2 as T2
 import cv2
 from PIL import Image
+import numpy as np
 
 output_res = (768,768)
 
@@ -31,7 +32,6 @@ generator = torch.manual_seed(0)
 def infer(prompt, negative_prompt, image):
     # implement your inference function here
     
-
     cond_input = conditioning_image_transforms(image)
     
     output = pipe(
@@ -42,7 +42,7 @@ def infer(prompt, negative_prompt, image):
         num_inference_steps=20
             )
 
-    return output[0]
+    return torch.tensor(output[0])
 
     # you need to pass inputs and outputs according to inference function
 title = "Categorical Conditioning Controlnet for One-Shot Image Stylization."