Deep Neural Network Library for Javascript

Train a neural network with your data & save its trained state!

Demo • Installation • Documentation • Contribute • Discord • License

Installation

CDN :

<script src="https://cdn.jsdelivr.net/gh/matiasvlevi/[email protected]/build/dann.min.js"></script>

Node :

npm i dannjs

dannjs on npmjs.com

Getting started

Require package

Components from the library can be imported like this

const { Dann } = require('dannjs');

Basic model construction

Setting up a small (4,6,6,2) neural network.

const nn = new Dann(4, 2);
nn.addHiddenLayer(6, 'leakyReLU');
nn.addHiddenLayer(6, 'leakyReLU');
nn.outputActivation('tanH');
nn.makeWeights();
nn.lr = 0.0001;
nn.log({details:true});

Train by backpropagation

Training with a dataset.

//XOR 2 inputs, 1 output
const dataset = [
    {
        input: [0, 0],
        output: [0]
    },
    {
        input: [1, 0],
        output: [1]
    },
    {
        input: [0, 1],
        output: [1]
    },
    {
        input: [1, 1],
        output: [0]
    }
];

//train 1 epoch
for (data of dataset) {
    nn.backpropagate(data.input, data.output);
    console.log(nn.loss);
}

Train by mutation

For neuroevolution simulations. Works best with small models & large population size.

const populationSize = 1000;
let newGeneration = [];

for (let i = 0; i < populationSize; i++) {

    // parentNN would be the best nn from past generation.
    const childNN = parentNN;
    childNN.mutateRandom(0.01, 0.65);

    newGeneration.push(childNN);
}

Standalone function

Convert a Neural Network to a JS function that can output predictions without the library.

let strfunc = nn.toFunction();
console.log(strfunc);

Save JSON

let json = nn.toJSON();
console.log(json);

Demo:

AI predicts San-francisco Housing prices.
more examples & demos here

Online editor:

https://dannjs.org/sandbox

Socials

Graph Dann models with this library

Dann-p5

Stickers

Get Dannjs stickers!

Contributors

Matias Vazquez-Levi 💻 📖 ⚠️ ✅	Francesco Ciulla 📢	Labnan 🐛 💻 ⚠️	sharkAce 💻	Hasnain Iqbal 💻 ⚠️	EL Ramos 🐛 ⚠️ 💻	viabhinav ✅
and1can 💻 ⚠️

Any contributions are welcome! See CONTRIBUTING.md.

License

MIT