pass image in default format

app.py

@@ -36,7 +36,7 @@ generator = torch.manual_seed(0)
 def infer(prompt, negative_prompt, image):
     # implement your inference function here
 
-    cond_input = conditioning_image_transforms(Image.create(image))
+    cond_input = conditioning_image_transforms(image)
     #cond_input = T.ToPILImage(cond_input)
     
     output = pipe(
@@ -62,7 +62,7 @@ gr.Interface(
             max_lines=1,
             placeholder="low quality",
         ),
-        gr.Image(type="pil"),
+        gr.Image(),
     ],
     outputs=gr.Gallery().style(grid=[2], height="auto"),
     title="Generate controlled outputs with Categorical Conditioning on Waifu Diffusion 1.5 beta 2.",