fix: fix linting errors

.pre-commit-config.yaml

@@ -1,12 +1,12 @@
 repos:
 -   repo: local
     hooks:
-    # -   id: mypy
-    #     name: mypy
-    #     entry: python3 -m mypy --ignore-missing-imports
-    #     language: system
-    #     types: [python]
-    #     files: ^litellm/
+    -   id: mypy
+        name: mypy
+        entry: python3 -m mypy --ignore-missing-imports
+        language: system
+        types: [python]
+        files: ^litellm/
     -   id: isort
         name: isort
         entry: isort

litellm/llms/azure.py

@@ -854,7 +854,6 @@ class AzureChatCompletion(BaseLLM):
                 additional_args={"complete_input_dict": data},
                 original_response=str(e),
             )
-            exception_mapping_worked = True
             raise e
         except asyncio.CancelledError as e:
             ## LOGGING
@@ -1029,9 +1028,10 @@ class AzureChatCompletion(BaseLLM):
                 openai_aclient = AsyncAzureOpenAI(**azure_client_params)
             else:
                 openai_aclient = client
-            response = await openai_aclient.embeddings.with_raw_response.create(
+            raw_response = await openai_aclient.embeddings.with_raw_response.create(
                 **data, timeout=timeout
             )
+            response = raw_response.parse()
             stringified_response = response.model_dump()
             ## LOGGING
             logging_obj.post_call(

litellm/llms/openai.py

@@ -2027,8 +2027,8 @@ class OpenAITextCompletion(BaseLLM):
                 else:
                     openai_client = client
 
-                response = openai_client.completions.with_raw_response.create(**data)  # type: ignore
-
+                raw_response = openai_client.completions.with_raw_response.create(**data)  # type: ignore
+                response = raw_response.parse()
                 response_json = response.model_dump()
 
                 ## LOGGING
@@ -2075,8 +2075,12 @@ class OpenAITextCompletion(BaseLLM):
             else:
                 openai_aclient = client
 
-            response = await openai_aclient.completions.with_raw_response.create(**data)
+            raw_response = await openai_aclient.completions.with_raw_response.create(
+                **data
+            )
+            response = raw_response.parse()
             response_json = response.model_dump()
+
             ## LOGGING
             logging_obj.post_call(
                 input=prompt,

litellm/tests/test_exceptions.py

@@ -855,7 +855,7 @@ def _pre_call_utils(
 ):
     if call_type == "embedding":
         data["input"] = "Hello world!"
-        mapped_target = client.embeddings.with_raw_response
+        mapped_target: Any = client.embeddings.with_raw_response
         if sync_mode:
             original_function = litellm.embedding
         else: