Starter files for the First School of AI Competition. The objective is to create an agent capable of playing Connect4
- Your agent must return a valid move within the time limit of 0.25 seconds. Failure to do so will result in losing the match.
- Your agent must be able to run on This Machine.
- Only the Libraries in the requirements.txt file will be available on the competition environment.
- Add all your necessary files into either a .zip or .rar archive.
- Fill out this Form and upload your archive on it.
-
Make sure you have Python 3.6 installed.
-
Clone the repository
git clone https://github.com/SchoolofAI-Vancouver/AI-Competition-Connect4.git
-
Use Conda to create a new environment and install dependencies.
Click Here if you need a detail guide on using conda.- Linux or Mac:
conda create --name tournament python=3.6 source activate tournament conda install numpy conda install cudnn pip install tensorflow-gpu==2.0.0alpha0 conda install pytorch torchvision cudatoolkit=10.0 -c pytorch conda install matplotlib conda install jupyter notebook
- Windows:
conda create --name tournament python=3.6 activate tournament conda install numpy conda install cudnn pip install tensorflow-gpu==2.0.0alpha0 conda install pytorch torchvision cudatoolkit=10.0 -c pytorch conda install matplotlib conda install jupyter notebook
-
Join us on Slack if you have any questions and make sure to join the #competitions channel.
Navigate to the directory and open Connect4.ipynb
jupyter notebook Connect4.ipynb
- test_agent.py is provided both as an opponent to test your agent against and as a starting point for implementing a minimax agent with alpha beta pruning and iterative deepening.
- If you wish to build an agent based on test_agent.py. Create a new copy of test_agent.py and heuristic.py modify heuristic.py to customize the behaviour of your agent.