All Tools
S
AgentsFreeOpen Source
SHIMMY
Gymnasium and PettingZoo bindings for external RL environments
MIT
ABOUT
Many classic RL suites still use one-off APIs, so a Gymnasium trainer cannot load dm_control or OpenSpiel without a hand-written wrapper. Shimmy provides those bindings so researchers can keep a single trainer loop across Farama and external environments.
INSTALL
pip install shimmyINTEGRATION GUIDE
1. Wrap DeepMind Control Suite tasks as Gymnasium envs for Stable Baselines3
2. Run OpenSpiel games through PettingZoo bindings without a custom adapter
3. Plug DMLab or other legacy suites into a Gymnasium-only training pipeline
4. Standardize environment APIs when comparing algorithms across labs
TAGS
pythongymnasiumpettingzoofaramadm-controlopenspielwrappers