Bearer.

evaluations/models.py

@@ -25,7 +25,7 @@ class LocalLLM():
 class RemoteLLM():
   def __init__(self, model_name):
     token = os.getenv("hfToken")
-    self.client = InferenceClient(model_name, token=token)
+    self.client = InferenceClient(model_name, token=f"Bearer {token}")
 
   def predict(self, response_type, prompt):
     for message in self.client.chat_completion(