Update app.py

app.py

@@ -6,14 +6,17 @@ import os
 
 # Load the model - for Hugging Face deployment
 try:
-    # Load the MLflow model
-    model = mlflow.pyfunc.load_model(".")
+    # Get absolute path to current directory
+    current_dir = os.getcwd()
+    print(f"Loading model from: {current_dir}")
+    
+    # List all files to verify
+    print("Available files:", os.listdir(current_dir))
+    
+    # Load model from current directory
+    model = mlflow.sklearn.load_model(current_dir)
     print("Model loaded successfully")
     
-    # Print model info for debugging
-    print("Model type:", type(model))
-    if hasattr(model, '_classifier'):
-        print("Classifier type:", type(model._classifier))
 except Exception as e:
     print(f"Error loading model: {str(e)}")
     print(f"Current working directory: {os.getcwd()}")
@@ -38,14 +41,9 @@ def predict_stroke_risk(age, gender, hypertension, heart_disease, ever_married,
             'smoking_status': [smoking_status]
         })
         
-        # Make prediction using MLflow model
-        prediction = model.predict(model_input=data)  # Add model_input parameter
-        
-        # For probability, we need to access the underlying classifier
-        if hasattr(model, '_classifier'):
-            probability = model._classifier.predict_proba(data)[0][1]
-        else:
-            probability = 0.5  # Default if we can't get probability
+        # Make prediction
+        prediction = model.predict(data)
+        probability = model.predict_proba(data)[0][1]
         
         # Create result message
         if prediction[0] == 1: