Update app.py

app.py

@@ -48,7 +48,7 @@ def retrieve_documents(query, k=5):
     distances, indices = index.search(query_embedding, k)
     return [documents[i] for i in indices[0]]
 
-def respond(message, history, system_message, max_tokens, temperature, top_p):
+async def respond(message, history, system_message, max_tokens, temperature, top_p):
     # Retrieve relevant documents
     relevant_docs = retrieve_documents(message)
     context = "\n\n".join(relevant_docs[:3])  # Limit context to top 3 documents
@@ -67,9 +67,7 @@ def respond(message, history, system_message, max_tokens, temperature, top_p):
 
     messages.append({"role": "user", "content": message})
 
-    response = ""
-
-    for message in client.chat_completion(
+    async for message in client.chat_completion(
         messages,
         max_tokens=max_tokens,
         stream=True,
@@ -78,8 +76,7 @@ def respond(message, history, system_message, max_tokens, temperature, top_p):
     ):
         if message.choices and message.choices[0].delta and message.choices[0].delta.content:
             token = message.choices[0].delta.content
-            response += token
-            yield response
+            yield token
 
 demo = gr.ChatInterface(
     respond,