Merge branch 'master' into cleanup

train.py

@@ -43,7 +43,8 @@ def on_episode_end(info):
             episode_num_agents = agent_info["num_agents"]
         episode_steps = max(episode_steps, agent_info["agent_step"])
         episode_score += agent_info["agent_score"]
-        episode_num_swaps += agent_info["num_swaps"]
+        if "num_swaps" in agent_info:
+            episode_num_swaps += agent_info["num_swaps"]
         if agent_info["agent_done"]:
             episode_done_agents += 1