(2) D4RL提供了非常简单的API接口,方便于学习者直接去获取数据集完成智能体的训练。 importd4rl# Import required to register environmentsenv=gym.make('maze2d-umaze-v1')dataset=env.get_dataset() (3) D4RL定义了标准的衡量指标 $$ normalized score =100\%* \frac{score- random_score}{export_...
You can use theenv.get_normalized_score(returns)function to compute a normalized score for an episode, wherereturnsis the undiscounted total sum of rewards accumulated during an episode. The individual min and max reference scores are stored ind4rl/infos.pyfor reference. ...