PyLoL is the Python component of the League of Legends v4.20 Learning Environment (using a modified version of the LeagueSandbox's GameServer project, not the original server from Riot). It exposes a custom machine learning API for the GameServer project as a Python RL Environment. PyLoL provides an interface for RL agents to interact with the League of Legends v4.20, getting observations and sending actions.
Disclaimer: This project is not affiliated with Riot Games in any way.
If you are interested in using this project or are just curious, send an email to raokosan@gmail.com.
The easiest way to try pylol at the moment is to visit this Google Colab:
Within the demonstration, you can try different agents by changing the "--agent" parameter to either random, base or scripted. Random will choose random actions from the action space uniformly, base will just issue no-ops (i.e. no operation, it will do nothing) and scripted will have the agents just attack each other.
The easiest way to get PyLoL is to use pip:
pip install pylol-rlThat will install the pylol package along with all the required dependencies.
virtualenv can help manage your dependencies. You may also need to upgrade pip:
pip install --upgrade pip for the pylol dependency.
You can install PyLoL from a local clone of the git repo:
git clone https://github.com/MiscellaneousStuff/pylol.git
pip install --upgrade pylol/Once you have PyLoL installed, you will need to create a config.txt file with the following format:
[dirs]
gameserver = .../LeagueSandbox-RL-Learning/GameServerConsole/bin/Debug/netcoreapp3.0/
lolclient = .../RADS/solutions/lol_game_client_sln/releases/0.0.1.68/deploy/
NOTE: You do not need to have the League of Legends client installed to use PyLoL,
you only need the GameServer installed. If you do not have the League of Legends
client installed, leave the lolclient option as follows
[dirs]
gameserver = .../LeagueSandbox-RL-Learning/GameServerConsole/bin/Debug/netcoreapp3.0/
lolclient =