manual cleaning after rebasing

src/helm/benchmark/scenarios/wildbench_scenario.py

@@ -58,6 +58,7 @@ def get_instances(self, output_path: str) -> List[Instance]:
         # Read all instances
         instances: List[Instance] = []
         for idx, row in enumerate(dataset):
+
             conversation = row["conversation_input"]
 
             # Following https://github.com/allenai/WildBench/blob/d6b8dcaf377d173d031980f97c16e1a82618c03d/src/eval.py

src/helm/benchmark/static/schema_lite_v2.yaml

@@ -215,20 +215,3 @@ run_groups:
       who: "real-world users"
       when: "2024"
       language: English
-
-  - name: wildbench
-    display_name: WildBench
-    description: WildBench
-    metric_groups:
-      - accuracy
-      - efficiency
-      - general_information
-    environment:
-      main_name: wildbench_score
-      main_split: test
-    taxonomy:
-      task: "?"
-      what: "?"
-      who: "?"
-      when: "?"
-      language: English

src/helm/clients/vertexai_client.py

@@ -202,7 +202,6 @@ def get_model(model_name: str) -> GenerativeModel:
     def make_request(self, request: Request) -> RequestResult:
         """Make a request"""
         contents = [request.prompt]
-        content_key = request.prompt
 
         # For the multimodal case, build up the content with the media objects of `request.multimodal_prompt`
         if request.multimodal_prompt is not None:
@@ -215,7 +214,6 @@ def make_request(self, request: Request) -> RequestResult:
                 contents.append(
                     Content(role=role_mapping.get(msg["role"], "user"), parts=[Part.from_text(msg["content"])])
                 )
-            content_key = "\n".join([msg["content"] for msg in request.messages])
 
         parameters = {
             "temperature": request.temperature,