Update inline doc.

app.py

@@ -5,7 +5,9 @@ client = InferenceClient(
     "mistralai/Mixtral-8x7B-Instruct-v0.1"
 )
 
-
+# According to "Model Card for Mixtral-8x7B"
+# The instruction format must be strictly respected, otherwise the model will generate sub-optimal outputs.
+# Refer to: https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1#instruction-format
 def format_prompt(message, history):
   prompt = "<s>"
   for user_prompt, bot_response in history: