Digit Recognition

Web application that recognizes hand-drawn digits

Uses a custom from-scratch neural net written in Rust and compiled to WebAssembly

Screen.Recording.2024-01-20.mov

Updating model and development

(all snippets start in root of repo)

1. Build wasm for node

cd ./rust-neural-network
wasm-pack build --target node

2. Unpack data

cd ./trainer
unzip ./mnist.zip

3. Train model

cd ./trainer
yarn train

4. Copy model to web app

cp ./trainer/model.json ./frontend/public

5. Build wasm for browser

cd ./rust-neural-network
wasm-pack build

6. Start website

cd ./frontend
yarn install
yarn serve