Dedicated hf dynamic SUT fixes

src/modelgauge/suts/huggingface_chat_completion.py

@@ -124,9 +124,10 @@ def translate_response(
 class HuggingFaceChatCompletionDedicatedSUT(BaseHuggingFaceChatCompletionSUT):
     """A Hugging Face SUT that is hosted on a dedicated inference endpoint and uses the chat_completion API."""
 
-    def __init__(self, uid: str, inference_endpoint: str, token: HuggingFaceInferenceToken):
+    def __init__(self, uid: str, inference_endpoint: str, model: str, token: HuggingFaceInferenceToken):
         super().__init__(uid, token)
         self.inference_endpoint = inference_endpoint
+        self.model = model
 
     def _create_client(self):
         endpoint = get_inference_endpoint(self.inference_endpoint, token=self.token.value)
@@ -158,6 +159,7 @@ def translate_text_prompt(self, prompt: TextPrompt, options: SUTOptions) -> Hugg
         if options.top_logprobs is not None:
             logprobs = True
         return HuggingFaceChatCompletionRequest(
+            model=self.model,
             messages=[ChatMessage(role="user", content=prompt.text)],
             logprobs=logprobs,
             **options.model_dump(),
@@ -168,6 +170,7 @@ def translate_chat_prompt(self, prompt: ChatPrompt, options: SUTOptions) -> Hugg
         if options.top_logprobs is not None:
             logprobs = True
         return HuggingFaceChatCompletionRequest(
+            model=self.model,
             messages=[ChatMessage(role=p.role.lower(), content=p.text) for p in prompt.messages],
             logprobs=logprobs,
             **options.model_dump(),

src/modelgauge/suts/huggingface_sut_factory.py

@@ -104,23 +104,24 @@ def _find(sut_definition: SUTDefinition) -> str | None:
         try:
             endpoints = hfh.list_inference_endpoints()
             for e in endpoints:
-                if e.repository == model_name and e.status != "running":
-                    try:
-                        e.resume()
-                    except Exception as ie:
-                        logger.error(
-                            f"Found endpoint for {model_name} but unable to start it. Check your token's permissions. {ie}"
-                        )
-                    return e.name
+                if e.repository.lower() == model_name:
+                    if e.status != "running":
+                        try:
+                            e.resume()
+                        except Exception as ie:
+                            logger.error(
+                                f"Found endpoint for {model_name} but unable to start it. Check your token's permissions. {ie}"
+                            )
+                    return e.name, e.repository
         except Exception as oe:
             logger.error(f"Error looking up dedicated endpoints for {model_name}: {oe}")
         return None
 
     def make_sut(self, sut_definition: SUTDefinition) -> HuggingFaceChatCompletionDedicatedSUT:
-        endpoint_name = HuggingFaceChatCompletionDedicatedSUTFactory._find(sut_definition)
+        endpoint_name, model_name = HuggingFaceChatCompletionDedicatedSUTFactory._find(sut_definition)
         if not endpoint_name:
             raise ProviderNotFoundError(
                 f"No dedicated inference endpoint found for {sut_definition.external_model_name()}."
             )
         sut_uid = sut_definition.dynamic_uid
-        return HuggingFaceChatCompletionDedicatedSUT(sut_uid, endpoint_name, self.injected_secrets())
+        return HuggingFaceChatCompletionDedicatedSUT(sut_uid, endpoint_name, model_name, self.injected_secrets()[0])