tennis-ai

A pair of reinforcement learning agents that can play tennis 🎾

Languages: Python 3.6 and Pytorch
Environment: Unity ML-Agents Toolkit

The Environment

Before Training: Reward: 0.0, 0.0, 0.0, 0.1, 0.0

After Training: Reward: +2.6

Actions

Actions are in the form of a continuous vector space of size 2, corresponding to horizontal and vertical movement.

States

Each agent receives an environment state of 8 dimensions, position and velocity of the ball and racket. They are stacked in 3 frames for each time step, giving a total vector of size 24.

Rewards

Each agent is given a reward of +0.1 whenever it hits the ball over the net, accumulated throughout the episode. When either agent lets the ball hit the ground or hits it out of bounds, it received a reward of -0.01. The environment is solved when the average maximum reward accumulated by either agent over 100 consecutive episodes reaches +0.5.

Installation (Windows 10 64-bit)

1. Install Anaconda if you don't have it already.
2. Open Anaconda Prompt/command line/terminal
3. Create a new environment (named tennis-env): conda create --name tennis-env python=3.6
4. Activate environment: activate tennis-env
5. Navigate to desired directory to download project file: cd path/to/desired/directory
6. Clone the repository: git clone https://github.com/albertlai431/tennis-ai
7. Go to dependencies directory: cd tennis-ai/python
8. Install dependencies (may take a while): pip install .
9. Install pytorch 0.4.0 with conda: conda install pytorch=0.4.0 -c pytorch
10. Create kernel with environment: python -m ipykernel install --user --name tennis-env --display-name "tennis-env"

How to Use

1. Launch jupyter-notebook and navigate to cloned repository directory
2. Open train.ipynb and run the code if you would like to train the agent 💪
3. Open test.ipynb and run the code if you would like to observe a fully trained agent! 😃
4. Important: Before running any code in either of the ipynb files, click Kernel on the top bar, Change kernel > tennis-env
5. Remember to deactivate the environment in the Anaconda Prompt/command line/terminal after you are done: conda deactivate

Potential Issues

• The folder Tennis_Windows_x86_64 may not always work; if you are getting a UnityTimeOutException, please go to this link and replace Tennis_Windows_x86_64 with the correct folder for your system. You may also need to modify the env declaration.