-
Notifications
You must be signed in to change notification settings - Fork 146
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
cannot train ma-gym environment with IQL #204
Comments
Could you please provide me with the complete error tracing log? |
I met a similar issue. To solve it, import a different ENV_REGISTRY This Is because IQL algorithm is a (fake) VD network that uses MARLlib/marllib/marl/algos/run_vd.py Line 30 in 368c617
In this add_new_env.py however, the ENV_REGISTRY comes from base_env. MARLlib/examples/add_new_env.py Line 40 in 368c617
|
Thank you very much for your reply. But for some reason after replacing that line its telling me that my environment is not registered. Do I have to change something in my config file? Thanks in advance. |
You can try to register both environment in the registries (base_env and global_reward_env). Also, If you can kindly provide the specific error screenshot you have encountered, I may get a better understanding of what's going on :) |
I solved it afterwards. I had to change the init.py file in the marl folder. I had to make the ENV_REGISTRY as COOP_ENV_REGISTRY. After that it worked. Thanks for your help though! |
Hello there!
I am trying to train environments with IQL. But I think I am doing something wrong. The add_new_env.py file utilizes ma-gym environment and trains that with mappo. But when I write iql instead of mappo it doesn't work. I have tried to solve it but wasn't able to. I think there is something I am missing.
Thanks in advance.
The text was updated successfully, but these errors were encountered: