Skip to content
This repository has been archived by the owner on Apr 7, 2024. It is now read-only.
/ NimNN Public archive

Genetic Neural Network from scratch written in Nim

License

Notifications You must be signed in to change notification settings

amaank404/NimNN

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

NimNN

Nim Neural Networks (genetic). This is a simple program written without any prior knowledge about neural network algorithms. This is a command line program and nnsim.exe --help should give out to you all available options for you to choose from.

Demo

I have a demo for the same available at my webpage. Click here, A video shall be presented to you under My Projects section.

Build

you can build the latest release by cloning this repository and you need choosenim installation (nim, nimble)

nimble -d:release build nnsim  # Using release significantly boosts performance
nnsim --help

For even faster builds. You can try this:

nimble -d:release --mm:orc --deepcopy:on build nnsim  # 2x better performance.
nnsim --help

If you want, you can try compiling it with danger mode and it should run fine and faster. But unexpected things can happen on unexpected command line arguments. Doing so is thus not recommended.

Download (Without build)

If you do not want to compile this package, release builds are already provided here.

Scripts

There are a few scripts written in python from my old Neural Network project that was written in python. Those scripts have been modified now to work with this program's output simdata.json.

view.py

You can view simulation in real time if you had saved the data with -f or -i while simulating. use python view.py --help for usage information

genomeview.py

You can view the brain of a neural network using this script, do note that using this requires you to have the /labels downloaded. use python genomeview.py --help for more information.