强化学习,tensorflow自定义python环境

时间:2020-09-24 17:19:58

标签: tensorflow typeerror reinforcement-learning

我试图为tensorflow实现自定义python环境。因此,我的_step方法在返回ts.transition(np.array(observation = self._state, reward=reward, discount=1.0))时抛出并出错了TypeError: array() missing required argument 'object' (pos 1)

0 个答案:

没有答案