fix(utils.py): fix azure streaming logic

litellm/tests/test_streaming.py

@@ -222,14 +222,12 @@ tools_schema = [
 
 def test_completion_azure_stream_special_char():
     litellm.set_verbose = True
-    messages = [
-        {"role": "user", "content": "Respond with the '<' sign and nothing else."}
-    ]
+    messages = [{"role": "user", "content": "hi. respond with the <xml> tag only"}]
     response = completion(model="azure/chatgpt-v-2", messages=messages, stream=True)
     response_str = ""
     for part in response:
         response_str += part.choices[0].delta.content or ""
-
+    print(f"response_str: {response_str}")
     assert len(response_str) > 0
 
 

litellm/utils.py

@@ -8842,11 +8842,11 @@ class CustomStreamWrapper:
         Output parse <s> / </s> special tokens for sagemaker + hf streaming.
         """
         hold = False
-        # if (
-        #     self.custom_llm_provider != "huggingface"
-        #     and self.custom_llm_provider != "sagemaker"
-        # ):
-        #     return hold, chunk
+        if (
+            self.custom_llm_provider != "huggingface"
+            and self.custom_llm_provider != "sagemaker"
+        ):
+            return hold, chunk
 
         if finish_reason:
             for token in self.special_tokens: