Update app.py

app.py

@@ -29,29 +29,34 @@ def respond(
     try:
         client = get_client(model_name)
     except ValueError as e:
-        return chat_history + [[message, str(e)]]
+        chat_history.append((message, str(e)))
+        return chat_history
 
     messages = [{"role": "system", "content": system_message}]
-
     for human, assistant in chat_history:
         messages.append({"role": "user", "content": human})
         messages.append({"role": "assistant", "content": assistant})
-
     messages.append({"role": "user", "content": message})
 
     try:
-        response = client.chat_completion(
+        stream = client.chat_completion(
             messages,
-            max_tokens=max_tokens,
+            max_new_tokens=max_tokens,
             temperature=temperature,
             top_p=top_p,
+            stream=True,
         )
-        assistant_message = response.choices[0].message.content
+        partial_message = ""
+        for response in stream:
+            if response.token.special:
+                continue
+            partial_message += response.token.text
+            chat_history = chat_history + [(message, partial_message)]
+            yield chat_history
     except Exception as e:
-        assistant_message = f"An error occurred: {str(e)}"
-
-    chat_history.append((message, assistant_message))
-    return chat_history
+        error_message = f"An error occurred: {str(e)}"
+        chat_history.append((message, error_message))
+        yield chat_history
 
 def clear_conversation():
     return []