recieves brain_name and identifier on python side

ml-agents/mlagents/trainers/trainer_util.py

     :return:
     """
     trainer_parameters = trainer_config["default"].copy()
-    brain_name, brain_name_identifiers = brain_parameters.brain_name.split("?")
     trainer_parameters["summary_path"] = "{basedir}/{name}".format(
         basedir=summaries_dir, name=str(run_id) + "_" + brain_name
     )