A few months ago, I was working on a feature that had to be integrated with ML, I basically needed a simple way to run a small neural network in JavaScript.
I tried installing existing libraries like brain.js, but ran into a few problems; broken native build errors,
Reminder for this morning: Youβre doing better than you think, and youβre carrying a lot more than you give yourself credit for. Take a breath and take the next step
Excited to announce that STRIPE3 has just joined the @dev3pack Bridge Accelerator! π
We're building on @Solana and can't wait to show what we've been working on. https://t.co/n0s4hmmhs8
Today, that side quest became my first official npm package! Meet LiteGrad : a lightweight, zero-dependency Neural Network engine built in pure TypeScript.
You can try it out: npm i litegrad
#TypeScript#MachineLearning#NPM#100DaysOfCode
Sometimes the best way to build something is by getting stuck. I originally just wanted a simple ML library for a personal project, but when I couldn't get brain.js to work, so I decided to just write the matrix calculus and backpropagation myself.