put team reward in decision steps

ml-agents-envs/mlagents_envs/base_env.py

      this simulation step.
     """
 
-    def __init__(self, obs, reward, agent_id, action_mask, team_manager_id):
+    def __init__(
+        self, obs, reward, team_reward, agent_id, action_mask, team_manager_id
+    ):
+        self.team_reward: np.ndarray = team_reward
         self.agent_id: np.ndarray = agent_id
         self.action_mask: Optional[List[np.ndarray]] = action_mask
         self.team_manager_id: np.ndarray = team_manager_id
         return DecisionSteps(
             obs=obs,
             reward=np.zeros(0, dtype=np.float32),
+            team_reward=np.zeros(0, dtype=np.float32),
             agent_id=np.zeros(0, dtype=np.int32),
             action_mask=None,
             team_manager_id=np.zeros(0, dtype=np.int32),
      across simulation steps.
     """
 
-    def __init__(self, obs, reward, interrupted, agent_id, team_manager_id):
+    def __init__(
+        self, obs, reward, team_reward, interrupted, agent_id, team_manager_id
+    ):
+        self.team_reward: np.ndarray = team_reward
         self.interrupted: np.ndarray = interrupted
         self.agent_id: np.ndarray = agent_id
         self.team_manager_id: np.ndarray = team_manager_id
         return TerminalSteps(
             obs=obs,
             reward=np.zeros(0, dtype=np.float32),
+            team_reward=np.zeros(0, dtype=np.float32),
             interrupted=np.zeros(0, dtype=np.bool),
             agent_id=np.zeros(0, dtype=np.int32),
             team_manager_id=np.zeros(0, dtype=np.int32),