move from average to sum of rewards

ml-agents/mlagents/trainers/trajectory.py

             agent_buffer_trajectory["team_rewards"].append(teammate_rewards)
             team_reward = teammate_rewards + [exp.reward]
             agent_buffer_trajectory["average_team_reward"].append(
-                sum(team_reward) / len(team_reward)
+                sum(team_reward)
+            #    sum(team_reward) / len(team_reward)
+            #)
 
             # Next actions
             teammate_cont_next_actions = []