tsonglew / yuri

Deep Learning StarCraft II AIbot

Yuri

Macro actions based toy DQN StarCraft II AIbot, which beats Hard(Level 5) builtin bot with 95% win rate

Getting Started

Prerequisites

• Install Starcraft II from official site
• Install python package manager: pipenv
• Download Training data for attack actions and link to yuri/attack_train
• Create directory to save random victory data

$ pipenv install

Running the tests

To-do

Break down into end to end tests

To-do

And coding style tests

$ pylint *.py

Deployment

Configuration

Fill configs in yuri/yuri.json

Run game

$ pipenv run python -m yuri.main --type game 

Train model

$ pipenv run python -m yuri.main --type train

Built With

• pipenv

License

This project is licensed under the MIT License - see the LICENSE.md file for details

Acknowledgments

To-do