piojanu / humblerl

Straightforward reinforcement learning python framework

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Refactor environments wrappers (OpenAI Gym and Unity ML-Agent)

piojanu opened this issue · comments

Create environments wrappers (implementing core.Environment interface) and env mocks to test them. Use dummy environments with their original API, so we can truly check if wrappers are compatible with environments (with e.g. some chosen version of OpenAI Gym).

DoD.:

Wrappers for OpenAI Gym and Unity ML-Agent and tests that use original environment's API from appropriate package. Environments supported versions are included in README.md.

Next steps:

  • Implement OpenAI Gym wrapper
  • Implement Unity ML-Agent wrapper

It is outdated.