Skip to content

This project is supervised by Tenuki, Brest's Go association, with the help of engineering students from IMT Atlantique. GoInsight provides Go game review with different feedback to help player of all level to learn.

License

Notifications You must be signed in to change notification settings

MomoWorlData/GoInsight

 
 

Repository files navigation

GoInsight

Partner Tenuki Partner IMT Atlantique Website KataGo GitHub KataGo License MIT

Logo

GoInsight is a tool to help Go player analyse their games. It uses a fine-tuned version of KataGo, alongside various pre and post-processings steps to provide a user-friendly feedbacks on Go games. This project is the result of the collaboration between Tenuki and IMT Atlantique.

Features

  • Better move highlights.
  • Blunder and mistake highlights.
  • Important square highlight to help the player understand the engine choice.
  • Ignorable area to be able to study a specific area of the game.

Installation

This project requires Python 3.7 or higher. Please ensure Python 3.7+ is installed and available on your PATH before running the setup.

Unix (Linux/MacOS)

First, open a terminal in the directory you'd like to clone the project in.

# Clone the repo
git clone https://github.com/leobeaumont/GoInsight.git
cd GoInsight
# Setup the environement
make setup
# Activate the virtual environment
source .venv/bin/activate
# Download KataGo model
make get-model

You're all setup !

Windows

First, open PowerShell in the directory where you'd like to clone the project.

# Clone the repo
git clone https://github.com/leobeaumont/GoInsight.git
cd ./GoInsight/
# Allow PowerShell scripts to run (first time only)
Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope CurrentUser
# Setup the environment
.\make.ps1 setup
# Activate the virtual environment
.\.venv\Scripts\Activate.ps1
# Download KataGo model
.\make.ps1 get-model

Commands

Run KataGo in terminal

This start an instance of KataGo, type gtp commands to interact with it.

  • Use quit command to close the instance
  • Use list_commands to get a list of all the commands

Unix (Linux/MacOs)

make run-model

Windows

.\make.ps1 run-model

Optimise KataGo for your device (optional)

Start a batch of tests to find the best parameters (This will take a few minutes)

Unix (Linux/MacOS)

make opt-model

Windows

.\make.ps1 opt-model

Tests

This will run all tests declared in the tests directory

Unix (Linux/MacOS)

make tests

Windows

.\make.ps1 tests

Documentation

This will open the project's documentation on your default web browser

Unix (Linux/MacOS)

make docs

Windows

.\make.ps1 docs

Clean project

Remove setup files and the virtual environment from the project

Unix (Linux/MacOS)

make clean

Windows

.\make.ps1 clean

Contributing

Contributions are not welcome yet, as this project is part of students cursus a IMT Atltantique. The project will be opened to contributors after the course ended.

Acknowledgements

Structures

  • Tenuki Brest Go association.
  • IMT Atlantique engineering school.
  • KataGo open ource Go engine.

Supervisors and clients

  • Coppin Gilles
  • Le Hir Mathieu
  • Peillard Étienne

Project members

Contacts

For any questions or supports, please contact leo.beaumont@imt-atlantique.net.

About

This project is supervised by Tenuki, Brest's Go association, with the help of engineering students from IMT Atlantique. GoInsight provides Go game review with different feedback to help player of all level to learn.

Resources

License

Contributing

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 83.6%
  • PowerShell 10.7%
  • Makefile 5.7%