Update app.py

app.py

@@ -39,7 +39,7 @@ def generate_response(
             response += token
             yield response
     else:
-        input_text = tokenizer(messages, return_tensors="pt").input_ids.to(device)
+        input_text = tokenizer.apply_chat_template(messages, return_tensors="pt").to(device)
         output = model.generate(input_text, max_length=max_tokens, temperature=temperature, top_p=top_p)
         response = tokenizer.decode(output[0], skip_special_tokens=True)
         yield response