Upload app.py with huggingface_hub

app.py

@@ -12,23 +12,19 @@ model = load_model('melanoma_cancer_model.h5')
 
 # Define the function to make predictions
 def classify_image(img):
-    img = np.array(img)
     img = np.expand_dims(img, axis=0)
-    img = img/255
-    # Resize data for the model (128, 128,3)
-    resized_img = tf.image.resize(img, [128, 128])
-
-    prediction = model.predict(resized_img)
-    print(prediction)
-    # return melanuma or not melanoma
-    return {"melanoma": prediction[0][0], "not melanoma": 1 - prediction[0][0]}
+    # Resize image
+    resized_img = tf.image.resize(img, [160, 160])
+    # Predict the image
+    prediction = model.predict(resized_img)[0][0]
+    # Convert to float value
+    prediction = float(prediction)
+    # return dictionary for Gradio
+    return {"melanoma": prediction, "not melanoma": 1 - prediction}
     
 
-#image = gr.Image(shape=(128, 128), type='pil', image_mode='RGB')
-#label = gr.Label(num_top_classes=2)
-
-# Run on localhost
-gr.Interface(fn=classify_image, inputs="image", outputs="label").launch()
-#gr.Interface(fn=classify_image, inputs=image, outputs=label).launch(server_port=7860)
+# Launch the Gradio interface
+gr.Interface(fn=classify_image, inputs='image', outputs="label").launch()
+# Launch shareble Gradio interface 
+# gr.Interface(fn=classify_image, inputs='image', outputs="label").launch(share=True)
 
-# Run the app