RAI
ML framework with ergonomic APIs in Rust. Lazy computation and composable transformations like JAX.
Installation
Code snippets
Function transformations (jvp, vjp, grad, value_and_grad)
use ;
NN Modules, Optimizer and loss functions
Examples
- linear_regression
cargo run --bin linear_regression --release
- mnist
cargo run --bin mnist --release
cargo run --bin mnist --release --features=cuda
- mnist-cnn
cargo run --bin mnist-cnn --release
cargo run --bin mnist-cnn --release --features=cuda
- phi2
cargo run --bin phi2 --release
cargo run --bin phi2 --release --features=cuda
- phi3
cargo run --bin phi3 --release
cargo run --bin phi3 --release --features=cuda
- qwen2
cargo run --bin qwen2 --release
cargo run --bin qwen2 --release --features=cuda
- gemma
- accept license agreement in https://huggingface.co/google/gemma-2b
pip install huggingface_hub
- login to hf
huggingface-cli login
cargo run --bin gemma --release
cargo run --bin gemma --release --features=cuda
- vit
cargo run --bin vit --release
cargo run --bin vit --release --features=cuda
LICENSE
This project is licensed under either of
- Apache License, Version 2.0, (LICENSE-APACHE or http://www.apache.org/licenses/LICENSE-2.0)
- MIT license (LICENSE-MIT or http://opensource.org/licenses/MIT)
at your option.