diff --git a/model.py b/model.py
index 9e42e9b..57d8929 100644
--- a/model.py
+++ b/model.py
@@ -72,4 +72,7 @@ def text_preprocessing(mystr):
 
 filename = 'NaiveModel.pkl'
 pickle.dump(model, open(filename, 'wb'))
-    
\ No newline at end of file
+    
+
+
+# Hi, I just cloned your repo and noticed something interesting. We're both using the same dataset and model, but I'm curious whether your model predicts accurately even though the data is imbalanced. You're using only predict() without any class weights or threshold tuning, right? Can you please let me know how it's still performing well on the spam class?"