Burn is a next generation Deep Learning Framework that doesn't compromise on flexibility, efficiency and portability.

Deep learning in Rust, with shape checked tensors and neural networks

Tensors and differentiable operations (like TensorFlow) in Rust

l2 is a fast, Pytorch-style Tensor+Autograd library written in Rust

N-dimensional matrix class for Rust

A learning-focused, high-performance tensor computation library built from scratch in Rust, featuring automatic differentiation and CPU/CUDA backends.

A Rust Library for High-Performance Tensor Exchange with Python

A high performance N-dimensional array library for Rust

A user-friendly ML framework built in Rust

Simplistic API for deep learning tensor operations

A neural network, and tensor dynamic automatic differentiation implementation for Rust.

An n-dimensional rust tensor library

Rust accelerated contraction ordering primitives for tensor networks and einsums

High-performance Machine Learning, Dynamic Auto-Differentiation and Tensor Algebra crate for Rust

DLPack safe Rust binding

DEPRECATED: this repo has been moved into https://github.com/dmlc/tvm/blob/master/rust

An animal can do training and inference every day of its existence until the day of its death. A forward pass is all you need.

Automatic differentiation for tensor operations

A small, educational library for neural networks, implemented from zero in pure Rust.