prev_action=prev_action,
interrupted=max_step,
memory=memory,
group_status=[],
group_reward=0,
)
steps_list.append(experience)
obs = []
interrupted=max_step_complete,
steps=steps_list, agent_id=agent_id, behavior_id=behavior_id, next_obs=obs
steps=steps_list,
agent_id=agent_id,
behavior_id=behavior_id,
next_obs=obs,
next_group_obs=[],