kerrykerrykerry999 / actor-critic-rl-to-play-fifa

Actor Critic Reinforcement Learning algorithm to train the football agents to dribble, pass, score etc and win a FIFA game

Football playing AI agent

Demo

Requirements:

-python >=3.6 -pytorch (tested on version 1.3) [pip install torch torchvision] #Assuming cuda 10.2 on the system(refer pytorch website) -google research platform

Steps to install Google Football Research Environment

- sudo apt-get install git cmake build-essential libgl1-mesa-dev libsdl2-dev \
libsdl2-image-dev libsdl2-ttf-dev libsdl2-gfx-dev libboost-all-dev \
libdirectfb-dev libst-dev mesa-utils xvfb x11vnc libsdl-sge-dev python3-pip
-git clone https://github.com/google-research/football.git
-cd football
-pip3 install .

Test the installation of football platform

cd into the repo
Run "python environment_test.py"

It will continuously render the frames:

To train the Football playing agent

-Download my repo -cd into the repo -Run python train.py

To test the model

Requirement:A trained model for 6000 epochs is added in the repo in the name "PPO_act6000.pth' -Run python test.py