add teamreward to decision step

ml-agents-envs/mlagents_envs/base_env.py

 
     obs: List[np.ndarray]
     reward: float
+    team_reward: float
     agent_id: AgentId
     action_mask: Optional[List[np.ndarray]]
     team_manager_id: int
         return DecisionStep(
             obs=agent_obs,
             reward=self.reward[agent_index],
+            team_reward=self.team_reward[agent_index],
             agent_id=agent_id,
             action_mask=agent_mask,
             team_manager_id=team_manager_id,
 
     obs: List[np.ndarray]
     reward: float
+    team_reward: float
     interrupted: bool
     agent_id: AgentId
     team_manager_id: int
         return TerminalStep(
             obs=agent_obs,
             reward=self.reward[agent_index],
+            team_reward=self.team_reward[agent_index],
             interrupted=self.interrupted[agent_index],
             agent_id=agent_id,
             team_manager_id=team_manager_id,