test: minor fix

src/loaders.py

@@ -81,7 +81,7 @@ def load_leaderboard_datastore(file_path, version) -> LeaderboardDataStore:
     return ds
 
 
-def load_eval_results(file_path: str) -> Dict[str, LeaderboardDataStore]:
+def load_eval_results(file_path: Union[str, Path]) -> Dict[str, LeaderboardDataStore]:
     output = {}
     for version in BENCHMARK_VERSION_LIST:
         fn = f"{file_path}/{version}"