Fail to support different envs

[lingxiao-guo]

Hello! I am currently attempting to train robots in different environments using a set of networks, but I have encountered an issue where it seems that isaacgym does not support creating more than one environment. The following code and error message provide a brief explanation:

def train(args):
    env, env_cfg = task_registry.make_env(name=args.task, args=args) # args.task = 'anymal_c_flat'
    args_ = copy.deepcopy(args)
    args_.task = 'cassie'
    env_, env_cfg_ = task_registry.make_env(name=args_.task, args=args_)

And after running it, the console shows:

/buildAgent/work/99bede84aa0a52c2/source/foundation/FdFoundation.cpp (166) : invalid operation : Foundation object exists already. Only one instance per process can be created.

[dipampatel18]

@glx15534565855 Were you able to create the new env? I am also working on that.

[lingxiao-guo]

Yes I solved it. Using  gym.create_actor() instead of create_env. Check the docs/index.html located in your isaacgym packages for detailed usage. Also you can find the usage in the example "actor_scaling.py" in the docs.

Hope that helps

------------------ 原始邮件 ------------------ 发件人: "leggedrobotics/legged_gym" @.>; 发送时间: 2023年12月10日(星期天) 上午10:33 @.>; @.@.>; 主题: Re: [leggedrobotics/legged_gym] Fail to support different envs (Issue #46)

@glx15534565855 Were you able to create the new env? I am also working on that.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>

[yingchengyang]

@glx15534565855 Thanks a lot for your issue, I have met the same problem. How do you solve it? For example, if I hope to create two environments 'anymal_c_flat' and 'cassie', how can I achieve it?

[yaocj97]

@glx15534565855 Thanks a lot for your issue, I have met the same problem. How do you solve it? For example, if I hope to create two environments 'anymal_c_flat' and 'cassie', how can I achieve it?

@yingchengyang Have you solved the issue? I guess the solution by @glx15534565855 means diving into the lower level of code and creating multiple heterogenous actors (and before that, envs) in a single sim. But to my knowledge, envs in a single sim can only take steps simultaneously. Do you have any idea if I want to specify which env to step?

[yingchengyang]

No, I have not solved this problem. I would appreciate a lot if you have any ideas.