Skip to content

[Question] Why are the rewards I get from training in OpenAI and PettingZoo's MPE environments so different? #1304

@xumanba

Description

@xumanba

Question

rewards from openai/multiagent-particle-envs/simple_spread_v3 is -200~-140
rewards from pettingzoo/mpe/simple_spread_v3 is around -70
When I use the same MAPPO algorithm to test the MPE environments of OpenAI and PettingZoo, the reward in OpenAI converges to a good value, but the same MPE environment in PettingZoo remains around -70 and does not rise.

Metadata

Metadata

Assignees

No one assigned

    Labels

    questionFurther information is requested

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions