Update app.py

app.py

@@ -6,7 +6,7 @@ client = InferenceClient(model="https://zgg3nzdpswxy4a-80.proxy.runpod.net/")
 def inference(message, history):
     partial_message = ""
     for token in client.text_generation(message, max_new_tokens=128, stream=True):
-        partial_message += token.replace("##", "\##")
+        partial_message += token.replace("\###", "\##")
         yield partial_message
 
 gr.ChatInterface(