def step(self, action):
obs, reward, done, info = self.env.step(action)
if self.mode == 'Bob':
import ipdb
ipdb.set_trace()
#First visit done for Bob
if np.allclose(reward, 0.):
done = True
info['is_success'] = True
if info.get('TimeLimit.truncated'):
del info['TimeLimit.truncated']
return obs, reward, done, info
elif self.mode == 'Alice':
import ipdb
ipdb.set_trace()
info = AttrDict(info)
self.total_rewards += reward
if done:
done = False
info.done_observation = obs
#info.terminal_state = True
if info.get('TimeLimit.truncated'):
done = True
info.terminal_state = False
info.episodic_return = self.total_rewards
self.total_rewards = 0
else:
info.terminal_state = False
info.episodic_return = None
return obs, reward, done, info
def step(self, action):
obs, reward, done, info = self.env.step(action)
info = AttrDict(info)
self.total_rewards += reward
if done:
info.done_observation = obs
info.terminal_state = True
if info.get('TimeLimit.truncated'):
info.terminal_state = False
info.episodic_return = self.total_rewards
self.total_rewards = 0
else:
info.terminal_state = False
info.episodic_return = None
return obs, reward, done, info
