Update app.py

app.py

@@ -1,15 +1,10 @@
 import gradio as gr
-from openai import OpenAI, APIError
 import os
+import openai
 import tenacity
-import asyncio
 
 ACCESS_TOKEN = os.getenv("HF_TOKEN")
-
-client = OpenAI(
-    base_url="https://api-inference.huggingface.co/v1/",
-    api_key=ACCESS_TOKEN,
-)
+openai.api_key = ACCESS_TOKEN
 
 # Retry logic with tenacity for handling API rate limits
 @tenacity.retry(wait=tenacity.wait_exponential(multiplier=1, min=4, max=10), stop=tenacity.stop_after_attempt(5))
@@ -27,24 +22,24 @@ async def respond(
 
         response = ""
         # Properly stream chat completions using dot notation
-        stream = client.chat.completions.create(
+        stream = openai.ChatCompletion.create(
             model="NousResearch/Hermes-3-Llama-3.1-8B",
             max_tokens=max_tokens,
-            stream=True,
             temperature=temperature,
             top_p=top_p,
             messages=messages,
+            stream=True,
         )
 
         # Stream response and concatenate tokens
         for chunk in stream:
-            if hasattr(chunk.choices[0].delta, 'content'):
-                token = chunk.choices[0].delta.content
+            if 'choices' in chunk and 'delta' in chunk['choices'][0] and 'content' in chunk['choices'][0]['delta']:
+                token = chunk['choices'][0]['delta']['content']
                 response += token
 
         return response
 
-    except APIError as e:
+    except openai.error.APIError as e:
         # Handle both string and dict types of error bodies
         error_details = e.body
         if isinstance(error_details, dict):
@@ -95,4 +90,4 @@ def launch_app():
         print("Please try again.")
 
 if __name__ == "__main__":
-    launch_app()
+    launch_app()