/rust-neural-network

A neural network from scratch in Rust.

Primary LanguageRust

Overview

Badges

This is a simple POC that implements a neural network from scratch in Rust, its used for pure learning purposes

Thanks to the work done by mathletedev

Usage

Clone this repo

execute

cargo run

Testing

Ensure grcov and llvm tools-preview are installed

cargo install grcov 

rustup component add llvm-tools-preview

execute the tests

# add the -- --nocapture or --show-ouput flags to see println! statements
$ CARGO_INCREMENTAL=0 RUSTFLAGS='-Cinstrument-coverage' LLVM_PROFILE_FILE='cargo-test-%p-%m.profraw' cargo test

# for individual tests
$ CARGO_INCREMENTAL=0 RUSTFLAGS='-Cinstrument-coverage' LLVM_PROFILE_FILE='cargo-test-%p-%m.profraw' cargo test create_diff_tar_pass -- --show-output

check the code coverage

$ grcov . --binary-path ./target/debug/deps/ -s . -t html --branch --ignore-not-existing --ignore '../*' --ignore "/*" --ignore "src/main.rs" -o target/coverage/html