Installation

In this doc we are going to setup Social Gym’s config runner which allows you to run Social Gym using a configuration file. If all you need to do is train, evaluate, or test your installation of Social Gym, this is the only installations tep you have to follow. However, if you wish to contribute and develop on Social Gym, we recommend the advanced installation which goes over setting up a debugger.

Prerequisites

WARNING: The authors of this project have had numerous issues with Docker and Docker-Compose when installed via snap. We recommend uninstalling docker and docker-compose if you have GPU-related issues with the docker images and installing them via apt.

Step 1: Clone the repo and checkout main

git clone git@github.com:ut-amrl/SocialGym2.git
git checkout main

Although we use Docker to run Social Gym’s internals, we have scripts to set up Social Gym as well as scripts that kick off the process written in python (so you still need to install stuff)

Step 2: Install requirements and run the install script!

python3 ./scripts/install_config_runner.py

Expect this to take a while (20-40m depending on your machine, mostly for the last step.).

Step 3: Run the Config Runner!

python3 config_runner/run.py -c 1_31_23/door/sacadrl.json

NOTE: If you do NOT see windows, there’s a problem with the docker file.

First try running

sudo xhost +

and repeat the command. If it still fails to show windows then try adding

--network host to the file {PROJECT_ROOT}/config_runner/run.sh on lines where we are setting docker environment variables. Otherwise, look up stuff like “cannot display window from docker container” – this is critical for Social Gym to work. We will fill out more tips/tricks and common failures here as we experience them.