Update app.py

app.py

@@ -47,8 +47,7 @@ def generate(query: Query):
 
     # Decode & extract assistant response
     output_string = tokenizer.batch_decode(outputs, skip_special_tokens=True)[0]
-    #response = output_string.split("assistant")[-1].strip()
-    response = output_string.strip()
+    response = output_string.split("assistant")[-1].strip()
     return {"response": response}
 
 @app.get("/")