Example implementation of an OpenAI Gym environment, to illustrate problem representation for RLlib use cases.
Clone the repo and connect into its top level directory.
To initialize and run the gym example:
pip install -r requirements.txt
pip install -e gym-example
python sample.py
To run Ray RLlib to train a policy based on this environment:
python train.py
h/t:
- https://github.com/DerwenAI/gym_trivial
- https://github.com/DerwenAI/gym_projectile
- https://github.com/apoddar573/Tic-Tac-Toe-Gym_Environment/
- https://medium.com/@apoddar573/making-your-own-custom-environment-in-gym-c3b65ff8cdaa
- https://github.com/openai/gym/blob/master/docs/creating-environments.md
- https://stackoverflow.com/questions/45068568/how-to-create-a-new-gym-environment-in-openai