Update app.py

app.py

@@ -1,13 +1,13 @@
 import gradio as gr
 import os
-import openai
+import aiohttp
 import tenacity
 import nest_asyncio
+import asyncio
 
 nest_asyncio.apply()
 
 ACCESS_TOKEN = os.getenv("HF_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))
@@ -23,39 +23,26 @@ async def respond(
         messages = [{"role": "system", "content": system_message},
                     {"role": "user", "content": message}]
 
-        response = ""
-        # Properly stream chat completions using dot notation
-        stream = openai.ChatCompletion.create(
-            model="NousResearch/Hermes-3-Llama-3.1-8B",
-            max_tokens=max_tokens,
-            temperature=temperature,
-            top_p=top_p,
-            messages=messages,
-            stream=True,
-        )
+        async with aiohttp.ClientSession() as session:
+            async with session.post(
+                "https://api.openai.com/v1/chat/completions",
+                headers={"Authorization": f"Bearer {ACCESS_TOKEN}"},
+                json={
+                    "model": "NousResearch/Hermes-3-Llama-3.1-8B",
+                    "max_tokens": max_tokens,
+                    "temperature": temperature,
+                    "top_p": top_p,
+                    "messages": messages,
+                    "stream": True,
+                },
+            ) as response:
+                response.raise_for_status()
+                response_text = await response.text()
+                return response_text
 
-        # Stream response and concatenate tokens
-        for chunk in stream:
-            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 openai.error.APIError as e:
-        # Handle both string and dict types of error bodies
-        error_details = e.body
-        if isinstance(error_details, dict):
-            error_type = error_details.get("type", "Unknown")
-            error_code = error_details.get("code", "Unknown")
-            error_param = error_details.get("param", "Unknown")
-            error_message = error_details.get("message", "An error occurred.")
-            error_str = f"{error_type}: {error_message} (code: {error_code}, param: {error_param})"
-        else:
-            error_str = f"Error: {error_details}"
-
-        print(f"APIError: {error_str}")
-        return error_str
+    except aiohttp.ClientError as e:
+        print(f"ClientError: {e}")
+        return "Error occurred. Please try again."
 
     except Exception as e:
         print(f"Exception: {e}")
@@ -64,7 +51,6 @@ async def respond(
 
 # Gradio function to handle user input and response generation without history
 def generate_response(message, system_message, max_tokens, temperature, top_p):
-    import asyncio
     loop = asyncio.new_event_loop()
     asyncio.set_event_loop(loop)
     response = loop.run_until_complete(respond(message, system_message, max_tokens, temperature, top_p))