Update app/main.py

app/main.py

@@ -1,9 +1,8 @@
 import gradio as gr
-from text_generation import AsyncClient
-import text_generation.types.Message as Message
+from openai import AsyncOpenAI
 
-endpoint_url = "http://127.0.0.1:8080"
-client = AsyncClient(endpoint_url)
+base_url = "http://127.0.0.1:8080/v1"
+client = AsyncOpenAI(base_url=base_url, api_key="-")
 
 """
 frequency_penalty: Optional[float] = None,
@@ -32,18 +31,21 @@ def _translate_messages(history):
     messages = []
 
     for conv in history:
-        messages.append(Message(role="user", content=conv[0]))
-        messages.append(Message(role="assistant", content=conv[1]))
+        messages.append({"role":"user", "content":conv[0]))
+        messages.append({"role":"assistant", "content":conv[1]))
 
     return messages
 
 async def echo(message, history):
     parameters = _default_parameters()
     messages = _translate_messages(history)
+    messages.append()
     
     history.append([message, ""])
-    full_resp = ""
-    async for resp in client.chat(messages=messages, **parameters):
+    full_resp = ""    
+    async for resp in client.chat.completions.create(
+        model="tgi", messages=messages, **parameters
+    ):
         full_resp = full_resp + resp.choices[0].delta.content
         history[-1][1] = full_resp
         yield history