fix: fix test

litellm/proxy/_new_secret_config.yaml

@@ -7,6 +7,11 @@ model_list:
       model: azure/gpt-4o
       api_key: os.environ/AZURE_API_KEY
       api_base: os.environ/AZURE_API_BASE
-
+  - model_name: fake-openai-endpoint-5
+    litellm_params:
+      model: openai/my-fake-model
+      api_key: my-fake-key
+      api_base: https://exampleopenaiendpoint-production.up.railway.app/
+      timeout: 1
 litellm_settings:
   fallbacks: [{"gpt-3.5-turbo": ["gpt-4o"]}]

litellm/proxy/_types.py

@@ -1994,13 +1994,14 @@ class ProxyException(Exception):
         message: str,
         type: str,
         param: Optional[str],
-        code: Optional[Union[int, str]] = None,
+        code: Optional[Union[int, str]] = None,  # maps to status code
         headers: Optional[Dict[str, str]] = None,
+        openai_code: Optional[str] = None,  # maps to 'code'  in openai
     ):
         self.message = str(message)
         self.type = type
         self.param = param
-
+        self.openai_code = openai_code or code
         # If we look on official python OpenAI lib, the code should be a string:
         # https://github.com/openai/openai-python/blob/195c05a64d39c87b2dfdf1eca2d339597f1fce03/src/openai/types/shared/error_object.py#L11
         # Related LiteLLM issue: https://github.com/BerriAI/litellm/discussions/4834

litellm/proxy/proxy_server.py

@@ -3716,7 +3716,8 @@ async def chat_completion(  # noqa: PLR0915
             message=getattr(e, "message", error_msg),
             type=getattr(e, "type", "None"),
             param=getattr(e, "param", "None"),
-            code=getattr(e, "code", getattr(e, "status_code", 500)),
+            openai_code=getattr(e, "code", None),
+            code=getattr(e, "status_code", 500),
             headers=headers,
         )
 
@@ -3929,7 +3930,8 @@ async def completion(  # noqa: PLR0915
             message=getattr(e, "message", error_msg),
             type=getattr(e, "type", "None"),
             param=getattr(e, "param", "None"),
-            code=getattr(e, "code", getattr(e, "status_code", 500)),
+            openai_code=getattr(e, "code", None),
+            code=getattr(e, "status_code", 500),
         )
 
 
@@ -4138,7 +4140,8 @@ async def embeddings(  # noqa: PLR0915
                 message=getattr(e, "message", error_msg),
                 type=getattr(e, "type", "None"),
                 param=getattr(e, "param", "None"),
-                code=getattr(e, "code", getattr(e, "status_code", 500)),
+                openai_code=getattr(e, "code", None),
+                code=getattr(e, "status_code", 500),
             )
 
 
@@ -4257,7 +4260,8 @@ async def image_generation(
                 message=getattr(e, "message", error_msg),
                 type=getattr(e, "type", "None"),
                 param=getattr(e, "param", "None"),
-                code=getattr(e, "code", getattr(e, "status_code", 500)),
+                openai_code=getattr(e, "code", None),
+                code=getattr(e, "status_code", 500),
             )
 
 
@@ -4518,7 +4522,8 @@ async def audio_transcriptions(
                 message=getattr(e, "message", error_msg),
                 type=getattr(e, "type", "None"),
                 param=getattr(e, "param", "None"),
-                code=getattr(e, "code", getattr(e, "status_code", 500)),
+                openai_code=getattr(e, "code", None),
+                code=getattr(e, "status_code", 500),
             )
 
 
@@ -4667,7 +4672,8 @@ async def get_assistants(
                 message=getattr(e, "message", error_msg),
                 type=getattr(e, "type", "None"),
                 param=getattr(e, "param", "None"),
-                code=getattr(e, "code", getattr(e, "status_code", 500)),
+                openai_code=getattr(e, "code", None),
+                code=getattr(e, "status_code", 500),
             )
 
 

tests/test_fallbacks.py

@@ -193,7 +193,7 @@ async def test_chat_completion_with_timeout():
         start_time = time.time()
         response, headers = await chat_completion(
             session=session,
-            key="sk-1234",
+            key="sk-PIp1h0RekR",
             model=model,
             messages=messages,
             num_retries=0,