Skip to content

Latest commit

 

History

History
40 lines (26 loc) · 3.44 KB

CONTRIBUTING.md

File metadata and controls

40 lines (26 loc) · 3.44 KB
Raw

Lux AI Challenge Season 2 Contributing Guide

If you find a bug or have a feature request, please open an issue here or let us know on our discord, the forums etc.

Want to make an contribution? Create an issue to this repository detailing what you want to change and if approved, make a pull request to this repository with your contribution! If you aren't sure about what to contribute on, check out our open issues. Make sure before you start contributing that you fork the repository and work off the main branch.

And of course, please be aware of and read the Code Of Conduct

Development Setup

For development, we recommend using conda to setup your python environment. Once you have conda (or the faster mamba), clone this repository and setup the environment

git clone https://github.com/Lux-AI-Challenge/Lux-Design-S2.git
conda env create -f environment.yml
conda activate luxai_s2
pip install -e luxai_s2
git submodule init
git submodule pull

This repository is organized as follows

  • luxai_s2/ - all code related to the Season 2 engine and CPU version
  • luxai_s2/luxai_runner/ - all code related to running episodes between different bots of different languages
  • kits/ - all code for the starter kits that allow competitors to develop strategies and compete
  • visualizer/ - the web based visualizer for the competition
  • juxai_s2/ - a submodule which contains code for the GPU/TPU optimized version of the luxai_s2 engine

Contributing Starter Kits

If you are interested in contributing a starter kit for a language that is not currently supported, please post an issue about it on our issue tracker so that people do not accidentally do the same things.

Here are a few things to be aware of. The competition servers currently run on Ubuntu 18.04, and has Python, NodeJS, and Java installed on the system, along with a whole ton of other Python packages. Hence, any language that can compile to machine code / binaries on Ubuntu 18.04 can be easily added to the competition.

If you want to get started, we recommend copying the structure of the folder kits/cpp or kits/python. We require you to provide a README similar to the other kit readmes, along with documentation on how to get started compiling code (if necessary), then running a match using the compiled code. To help understand what the raw data given by the engine are, see https://github.com/Lux-AI-Challenge/Lux-Design-S2/blob/main/kits/sample_env_cfg.json and https://github.com/Lux-AI-Challenge/Lux-Design-S2/blob/main/kits/sample_obs.json

Moreover, for compiled languages, we recommend also copying over the create_submission.sh script and Dockerfile in kits/cpp/ if the language's compiled binaries are OS dependent. For example, for the C++ kit, the dockerfile is used to compile the C++ agent code on Ubuntu 18.04 so then that code can be submitted to the competition servers.

If you have any questions or need help adding a starter kit, the Lux AI team is more than happy to help! Message us anywhere (preferably Github). You can also follow this document https://github.com/themmj/Lux-Design-2022/blob/documentation/kits/starter-kit-notes.md which has some important notes to be aware of when understanding and building a new starter kit.