update UI

app.py

@@ -1444,8 +1444,9 @@ def run_fn(
         info2 = 'Or try use the Python package that powers this app: <a style="white-space: nowrap;text-underline-offset: 2px;color: var(--body-text-color)" href="https://ncut-pytorch.readthedocs.io/en/latest/">ncut-pytorch</a>'
         info = info1 + info2
         
-        message = "<b>HuggingFace: </b></br>" + e.message + "</br></br>---------</br>" + "<b>`ncut-pytorch` Developer: </b></br>" + info
-        raise gr.Error(message, duration=0)
+        # message = "<b>HuggingFace: </b></br>" + e.message + "</br></br>---------</br>" + "<b>`ncut-pytorch` Developer: </b></br>" + info
+        gr.Warning(info, duration=0)
+        raise gr.Error(e.message , duration=0)
 
 
 import torch
@@ -1856,7 +1857,7 @@ def make_input_images_section(rows=1, cols=3, height="450px", advanced=False, is
             save_profile_button.click(fn=lambda name, images: on_disk_profiles.save_profile(name, images), inputs=[profile_text, input_gallery])
             load_profile_button.click(fn=lambda name, existing_images: gr.update(value=on_disk_profiles.load_profile(name, existing_images)), inputs=[profile_text, input_gallery], outputs=[input_gallery])
             delete_profile_button.click(fn=lambda name: on_disk_profiles.delete_profile(name), inputs=profile_text)
-            list_profiles_button.click(fn=lambda: gr.Info(on_disk_profiles.list_profiles(), duration=0))
+            list_profiles_button.click(fn=lambda: gr.Info(on_disk_profiles.list_profiles(), duration=10))
     
     if advanced:
         advanced_block.visible = True
@@ -2314,7 +2315,7 @@ with demo:
                     fig, ax = plt.subplots(1, 1, figsize=(6, 6))
                     ax.scatter(tsne_embed[:, 0], tsne_embed[:, 1], s=20, c=fps_tsne3d_rgb)
                     # compute the length of the edges
-                    lengthes = np.linalg.norm(tsne_embed[edges[:, 0]] - tsne_embed[edges[:, 1]], axis=1)
+                    # lengthes = np.linalg.norm(tsne_embed[edges[:, 0]] - tsne_embed[edges[:, 1]], axis=1)
                     # max_length = lengthes[k:].max() 
                     # diag_length = np.linalg.norm(tsne_embed.max(axis=0) - tsne_embed.min(axis=0))