https://github.com/bcorfman/pettingzoo-classicenvs
PettingZoo ConnectFour and TicTacToe examples, configured with Rye as dependency manager
https://github.com/bcorfman/pettingzoo-classicenvs
gymnasium-environment multi-agent-reinforcement-learning pettingzoo reinforcement-learning rye sb3-contrib stable-baselines3
Last synced: about 1 month ago
JSON representation
PettingZoo ConnectFour and TicTacToe examples, configured with Rye as dependency manager
- Host: GitHub
- URL: https://github.com/bcorfman/pettingzoo-classicenvs
- Owner: bcorfman
- License: mit
- Created: 2024-05-29T20:03:48.000Z (about 2 years ago)
- Default Branch: main
- Last Pushed: 2024-06-01T03:53:08.000Z (about 2 years ago)
- Last Synced: 2024-06-01T04:45:26.600Z (about 2 years ago)
- Topics: gymnasium-environment, multi-agent-reinforcement-learning, pettingzoo, reinforcement-learning, rye, sb3-contrib, stable-baselines3
- Language: Python
- Homepage:
- Size: 24.4 KB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# PettingZoo-ClassicEnvs
[PettingZoo](https://pettingzoo.farama.org/) ConnectFour and TicTacToe examples, configured with [Rye](https://rye.astral.sh/) as dependency manager. Tested with MacOS Sonoma and Ubuntu on WSL2.
## Prerequisites
* At a command prompt in the project directory, type `make devinstall` to set up Rye with the appropriate Python version plus the project-level dependencies in development mode, or just `make install` to set up the code as runnable only.
## To run the examples
* At a command prompt in the project directory, type `make run`.
* The first time through, the RL model will be trained and then saved to disk. After that, the trained model will be loaded from disk before evaluation.
* If retraining is desired, delete the model .zip files from the project directory.
## Notes for Visual Studio Code users
* I've included some extension recommendations that can make your development easier.
* [Run On Save](https://marketplace.visualstudio.com/items?itemName=emeraldwalk.RunOnSave)
* [Make support and task provider](https://marketplace.visualstudio.com/items?itemName=carlos-algms.make-task-provider)
* These recommendations will pop up when opening the project inside VSCode.
* Installing both extensions will
* Use the code in `settings.json` to run `make format` and `make lint` on each `File:Save`.
* Display available Make targets within the _Makefile Tasks_ sidebar pane and allow them to be run with a mouse click.