Compiled Eager Code

Flux provides a single, intuitive way to define models, just like mathematical notation. Julia transparently compiles your code, optimising and fusing kernels for the GPU, for the best performance.

Differentiable Programming

Existing Julia libraries are differentiable and can be incorporated directly into Flux models. Cutting edge models such as Neural ODEs are first class, and Zygote enables overhead-free gradients.

First-class GPU support

GPU kernels can be written directly in Julia via CUDA.jl. Flux is uniquely hackable and any part can be tweaked, from GPU code to custom gradients and layers.

Probabilistic Programming

The Turing.jl and Stheno.jl libraries enable probabilistic programming, Bayesian inference and Gaussian processes on top of Flux.

Graph Neural Networks

GraphNeuralNetworks.jl is a graph neural network library for Flux and supports CUDA GPU acceleration.

Computer Vision

Metalhead.jl includes many state-of-the-art computer vision models with pre-trained weights.

SciML

The SciML ecosystem uses the FluxML stack to mix neural nets with differential equations, to get the best of black box and mechanistic modelling.

Natural Language Processing

Transformers.jl provides components for transformer architectures for language modeling, as well as providing several trained models out of the box.

Community team

Flux is maintained by community team (see our governance model). Join us or talk to us on Zulip! 👉

Slack

Official Julia Slack for casual conversation. See #flux-bridged and #machine-learning.

Zulip

Zulip server for the Julia programming language community. See #ml-contributors and #machine-learning.

Discourse forum

Machine Learning in Julia community.

Stack Overflow

Ask questions about Flux.jl.

Contribute!

Help us by contributing code!