demo.launch() fix

app.py

@@ -1,4 +1,4 @@
-## Importing Libraries
+# Importing Libraries
 import os
 import torch
 import numpy as np
@@ -6,10 +6,11 @@ import gradio as gr
 from model import Model
 import albumentations as A
 
-## Creating a model instance
-efficientnet_b5_model =  Model()
-efficientnet_b5_model = torch.nn.DataParallel(efficientnet_b5_model) ## Must wrap our model in nn.DataParallel()
-    ## if used multi-gpu's to train the model otherwise we would get state_dict keys mismatch error.
+# Creating a model instance
+efficientnet_b5_model = Model()
+efficientnet_b5_model = torch.nn.DataParallel(
+    efficientnet_b5_model)  # Must wrap our model in nn.DataParallel()
+# if used multi-gpu's to train the model otherwise we would get state_dict keys mismatch error.
 efficientnet_b5_model.load_state_dict(
     torch.load(
         f='efficientnet_b5_checkpoint_fold_0.pt',
@@ -17,7 +18,9 @@ efficientnet_b5_model.load_state_dict(
     )
 )
 
-## Predict on a single image
+# Predict on a single image
+
+
 def predict_on_single_image(img):
     """
     Function takes an image, transforms for 
@@ -28,13 +31,13 @@ def predict_on_single_image(img):
     having melanoma. 
     """
     img = np.array(img)
-    transforms = A.Compose([A.Resize(512,512),
-        A.Normalize(mean=(0.485, 0.456, 0.406), 
-                    std=(0.229, 0.224, 0.225), 
-                    max_pixel_value=255.0, 
-                    always_apply=True
-        )]
-    )
+    transforms = A.Compose([A.Resize(512, 512),
+                            A.Normalize(mean=(0.485, 0.456, 0.406),
+                                        std=(0.229, 0.224, 0.225),
+                                        max_pixel_value=255.0,
+                                        always_apply=True
+                                        )]
+                           )
     img = transforms(image=img)['image']
     image = np.transpose(img, (2, 0, 1)).astype(np.float32)
     image = torch.tensor(image, dtype=torch.float).unsqueeze(dim=0)
@@ -43,27 +46,28 @@ def predict_on_single_image(img):
         probs = torch.sigmoid(efficientnet_b5_model(image))
         prob_of_melanoma = probs[0].item()
         prob_of_not_having_melanoma = 1 - prob_of_melanoma
-        pred_label = {"Probability of Having Melanoma": prob_of_melanoma, 
-                      "Probability of Not having Melanoma": prob_of_not_having_melanoma} 
+        pred_label = {"Probability of Having Melanoma": prob_of_melanoma,
+                      "Probability of Not having Melanoma": prob_of_not_having_melanoma}
         return pred_label
-    
-## Gradio App
-import gradio as gr
 
-## Examples directory path
+
+# Gradio App
+
+# Examples directory path
 melanoma_app_examples_path = "examples"
 
-## Creating the title and description strings  
+# Creating the title and description strings
 title = "Melanoma Cancer Detection App"
 description = 'An efficientnet-b5 model that predicts the probability of a patient having melanoma skin cancer or not.'
-example_list = [["examples/" + example] for example in os.listdir(melanoma_app_examples_path)]
+example_list = [["examples/" + example]
+                for example in os.listdir(melanoma_app_examples_path)]
 
-## Create the Gradio demo
+# Create the Gradio demo
 demo = gr.Interface(fn=predict_on_single_image,
-                   inputs=gr.Image(type='pil'),
-                   outputs=[gr.Label(label='Probabilities')],
-                   examples=example_list, title=title,
-                   description=description)
+                    inputs=gr.Image(type='pil'),
+                    outputs=[gr.Label(label='Probabilities')],
+                    examples=example_list, title=title,
+                    description=description)
 
-## Launch the demo!
-demo.launch(debug=False, share=True)
+# Launch the demo!
+demo.launch()