Allowed scoring multiple model/benchmark pairs per run via the endpoi…

…nt (optional)

brainscore_language/submission/endpoints.py

@@ -102,13 +102,15 @@ def get_models_and_benchmarks(args_dict: Dict[str, Union[str, List]]) -> Tuple[L
 
 def run_scoring(args_dict: Dict[str, Union[str, List]]):
     """ prepares parameters for the `run_scoring_endpoint`. """
-    model = _get_ids(args_dict, 'new_models')
-    benchmark = _get_ids(args_dict, 'new_benchmarks')
+    new_models = _get_ids(args_dict, 'new_models')
+    new_benchmarks = _get_ids(args_dict, 'new_benchmarks')
 
-    run_scoring_endpoint(domain="language", jenkins_id=args_dict["jenkins_id"], 
-        model_identifier=model, benchmark_identifier=benchmark, user_id=args_dict["user_id"], 
-        model_type="artificialsubject", public=args_dict["public"], 
-        competition=args_dict["competition"])
+    for model in new_models:
+        for benchmark in new_benchmarks:
+            run_scoring_endpoint(domain="language", jenkins_id=args_dict["jenkins_id"], 
+                model_identifier=model, benchmark_identifier=benchmark, user_id=args_dict["user_id"], 
+                model_type="artificialsubject", public=args_dict["public"], 
+                competition=args_dict["competition"])
 
 
 def parse_args() -> argparse.Namespace: