Hello, I am writing to ask a question about RLlib. I am creating a Customized Environment in Ray. Whenever I start training of the agent, the simulation outputs show this message: “Your env doesn’t have a .spec.max_episode_steps attribute.” !! “_get_slice_indices has been deprecated. This will raise an error in the future!” Even though I have defined max_episode_steps in configuration (i.e., horizon or soft_horizon). I would appreciate it if you could help me to solve these problems. Thanks,

Max_episode_steps warning

avnishn January 19, 2023, 6:27pm 2

Topic		Replies	Views
Max_episode_steps attribute in customized environment RLlib	3	2238	April 14, 2023
When run PPO,it can not calculate episode reward Uncategorized	0	288	August 18, 2023
EnvCompatibility problem on the number of parameters returned by step RLlib	5	658	May 4, 2023
Inconsistent number of episodes with 'evaluate' RLlib	2	337	July 18, 2022
How to tell RLLIB tune to run that many number of episodes RLlib	1	252	August 14, 2021