fix

app.py

@@ -63,7 +63,7 @@ if not IS_SPACES_ZERO:
     # pipe.enable_xformers_memory_efficient_attention()
     pipe.enable_model_cpu_offload()
 pipe.enable_vae_tiling()
-canny_torch = SobelOperator(device=device)
+canny_torch = SobelOperator()
 
 
 def pad_image(image):
@@ -96,17 +96,20 @@ def predict(
     strength=1.0,
     controlnet_start=0.0,
     controlnet_end=1.0,
+    canny_low=0.01,
+    canny_high=0.2,
     progress=gr.Progress(track_tqdm=True),
 ):
     if IS_SPACES_ZERO:
         apply_hidiffusion(pipe)
+        canny_torch.to(device)
     if input_image is None:
         raise gr.Error("Please upload an image.")
     padded_image = pad_image(input_image).resize((1024, 1024)).convert("RGB")
     conditioning, pooled = compel([prompt, negative_prompt])
     generator = torch.manual_seed(seed)
     last_time = time.time()
-    canny_image = canny_torch(padded_image, 0.01, 0.2)
+    canny_image = canny_torch(padded_image, canny_low, canny_high)
     images = pipe(
         image=padded_image,
         control_image=canny_image,
@@ -126,7 +129,7 @@ def predict(
         eta=1.0,
     )
     print(f"Time taken: {time.time() - last_time}")
-    return (padded_image, images.images[0])
+    return (padded_image, images.images[0]), padded_image, canny_image
 
 
 css = """
@@ -145,6 +148,7 @@ with gr.Blocks(css=css) as demo:
 
 [HiDiffusion](https://github.com/megvii-research/HiDiffusion) enables higher-resolution image generation.  
 You can upload an initial image and prompt to generate an enhanced version. 
+SDXL Controlnet [TheMistoAI/MistoLine](https://huggingface.co/TheMistoAI/MistoLine)
 [Duplicate Space](https://huggingface.co/spaces/radames/Enhance-This-HiDiffusion-SDXL?duplicate=true) to avoid the queue.  
 
 <small>
@@ -217,11 +221,28 @@ You can upload an initial image and prompt to generate an enhanced version.
                     value=1.0,
                     label="ControlNet End",
                 )
+                canny_low = gr.Slider(
+                    minimum=0,
+                    maximum=1,
+                    step=0.001,
+                    value=0.0,
+                    label="Canny Low Threshold",
+                )
+                controlnet_end = gr.Slider(
+                    minimum=0.0,
+                    maximum=1.0,
+                    step=0.001,
+                    value=0.2,
+                    label="Canny High Threshold",
+                )
 
             btn = gr.Button()
         with gr.Column(scale=2):
             with gr.Group():
                 image_slider = ImageSlider(position=0.5)
+            with gr.Row():
+                padded_image = gr.Image(type="pil", label="Padded Image")
+                canny_image = gr.Image(type="pil", label="Canny Image")
     inputs = [
         image_input,
         prompt,
@@ -234,7 +255,7 @@ You can upload an initial image and prompt to generate an enhanced version.
         controlnet_start,
         controlnet_end,
     ]
-    outputs = [image_slider]
+    outputs = [image_slider, padded_image, canny_image]
     btn.click(lambda x: None, inputs=None, outputs=image_slider).then(
         predict, inputs=inputs, outputs=outputs, concurrency_limit=1
     )
@@ -303,9 +324,9 @@ You can upload an initial image and prompt to generate an enhanced version.
             ],
             [
                 "./examples/huggingface.jpg",
-                "photo realistic huggingface human+++ emoji costume, round, yellow, skin+++ texture+++",
+                "photo realistic huggingface human emoji costume, round, yellow, (human skin)+++ (human texture)+++",
                 "blurry, ugly, duplicate, poorly drawn, deformed, mosaic, emoji cartoon,  drawing, pixelated",
-                232312,
+                12312353423,
                 0.101,
                 25.206,
                 2,