Develop fix cumulative reward (#725)

* [Cold Fix] Split the way cummulative rewards and episode length are counted
The reward is appended at each step to the cummulative reward
The episode count is ONLY incremented when d_t+1 is false

python/unitytrainers/bc/trainer.py

                 continue
             else:
                 next_idx = next_info_student.agents.index(agent_id)
+                if agent_id not in self.cumulative_rewards:
+                    self.cumulative_rewards[agent_id] = 0
+                self.cumulative_rewards[agent_id] += next_info_student.rewards[next_idx]
-                    if agent_id not in self.cumulative_rewards:
-                        self.cumulative_rewards[agent_id] = 0
-                    self.cumulative_rewards[agent_id] += next_info_student.rewards[next_idx]
                     if agent_id not in self.episode_steps:
                         self.episode_steps[agent_id] = 0
                     self.episode_steps[agent_id] += 1

python/unitytrainers/ppo/trainer.py

                     self.training_buffer[agent_id]['rewards'].append(next_info.rewards[next_idx])
                     self.training_buffer[agent_id]['action_probs'].append(a_dist[idx])
                     self.training_buffer[agent_id]['value_estimates'].append(value[idx][0])
-
-                if not next_info.local_done[next_idx]:
+                if not next_info.local_done[next_idx]:
                     if agent_id not in self.episode_steps:
                         self.episode_steps[agent_id] = 0
                     self.episode_steps[agent_id] += 1