Update clbaseimplementation.py

clbaseimplementation.py

@@ -21,6 +21,7 @@ from threading import Thread
 import chainlit as cl
 import torch
 import time
+import openai
 #load model 
 openai.api_key = "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"
 @cl.on_chat_start
@@ -34,9 +35,6 @@ async def main():
 
 
 
-    cl.user_session.set("llm", model)
-    cl.user_session.set("tokenizer", tokenizer)
-
     # Let the user know that the system is ready
     await msg.update(content=f"Chat has been loaded. You can now ask questions!")
 
@@ -46,16 +44,17 @@ async def main():
 @cl.on_message
 async def main(message: str):
     h = cl.user_session.get("history")
-    h.append({"role": "system", "content":message})
-    cl.user_session.set("history",h)
+    h.append({"role": "user", "content":message})
+
     resp = ""
-    async for stream_resp in await openai.ChatCompletion.create(
-          model="gpt-3.5-turbo",
-          messages=cl.user_session.get("history"),
-        stream = True
-        )
-        resp += stream_resp.get("choices")[0].get("text")
-        token = stream_resp.get("choices")[0].get("text")
-        await msg.stream_token(token)
-    cl.user_session.set("history",cl.user_session.get("history").append({"role": "assistant", "content":resp})
+    msg = cl.Message(content="")
+    async for stream_resp in await openai.ChatCompletion.acreate(model="gpt-3.5-turbo",messages=h,stream = True):
+        print(stream_resp)
+        token = stream_resp.get("choices")[0].get("delta").get("content")
+        if token:
+            resp += token
+            await msg.stream_token(token)
+    h.append({"role": "assistant", "content":resp})
+    cl.user_session.set("history",h)
+    print(h)
     await msg.send()