Update app.py

app.py

@@ -38,7 +38,9 @@ class_names = ["Benign", "Malignant"]
 
 # Load trained Neural Network model (TensorFlow/Keras)
 nn_model_path = "my_NN_BC_model.keras"
-nn_model = None
+
+nn_model = tf.keras.models.load_model(nn_model_path)
+
 if os.path.exists(nn_model_path):
     try:
         nn_model = tf.keras.models.load_model(nn_model_path)
@@ -80,7 +82,7 @@ def classify(model_choice, image=None, *features):
             output = vit_model(input_tensor)
             predicted_class = torch.argmax(output, dim=1).item()
 
-        return f"🔍 **Prediction:** {class_names[predicted_class]}"
+        return class_names[predicted_class]
 
     elif model_choice == "Neural Network":
         if any(f is None for f in features):
@@ -91,7 +93,7 @@ def classify(model_choice, image=None, *features):
         prediction = nn_model.predict(input_data_std) if nn_model else [[0, 1]]
         predicted_class = np.argmax(prediction)
 
-        return f"🔍 **Prediction:** {class_names[predicted_class]}"
+        return class_names[predicted_class]
 
 # Gradio UI
 with gr.Blocks() as demo:
@@ -117,8 +119,8 @@ with gr.Blocks() as demo:
         return {feature_inputs[i]: example[i] for i in range(len(feature_inputs))}
 
     with gr.Row():
-        example_btn_1 = gr.Button("🔵 Benign Example")
-        example_btn_2 = gr.Button("🔴 Malignant Example")
+        example_btn_1 = gr.Button("🔵 Malignant Example")
+        example_btn_2 = gr.Button("🔴 Benign Example")
 
     output_text = gr.Textbox(label="🔍 Model Prediction", interactive=False)