Update app.py

app.py

@@ -1,6 +1,5 @@
 import os
 import openai
-from openai.error import InternalServerError  # Import the error class
 import torch
 from transformers import pipeline
 
@@ -48,7 +47,6 @@ client = OpenAI(
 
 #     return "".join(full_response)
 
-
 def generate_llm_response(text, model_id="ccibeekeoc42/Llama3.1-8b-base-SFT-2024-11-09"):
     full_response = []
     try:
@@ -71,9 +69,12 @@ def generate_llm_response(text, model_id="ccibeekeoc42/Llama3.1-8b-base-SFT-2024
             if chunk.choices[0].delta.content:
                 full_response.append(chunk.choices[0].delta.content)
         return "".join(full_response)
-    except InternalServerError as e:
-        # This error is raised when the GPU is unavailable (e.g. scaling down to 0 after idle)
-        return "The GPU is currently booting up. Please wait about 10 minutes and try again."
+    except Exception as e:
+        # If the error has a response with status code 503, assume the GPU is booting up.
+        if hasattr(e, 'response') and e.response is not None and e.response.status_code == 503:
+            return "The GPU is currently booting up. Please wait about 10 minutes and try again."
+        else:
+            raise e
 
 generate_llm_response("Explain Deep Learning in Igbo")