Ecosystem

This page lists models available for reuse in Flux, as well other useful projects in the ecosystem. To add your project please send a PR.

Modelling packages

There are a number of packages in the Flux ecosystem designed to help with creating different kinds of models.

• Transformers.jl provides components for Transformer models for NLP, as well as providing several trained models out of the box.

• DiffEqFlux provides tools for creating Neural Differential Equations.

• GeometricFlux makes it easy to build fast neural networks over graphs.

• Flux3D shows off machine learning on 3D data.

• AdversarialPrediction.jl provides a way to easily optimize generic performance metrics in supervised learning settings using the Adversarial Prediction framework.

• Metalhead.jl includes many state-of-the-art computer vision models which can easily be used for transfer learning.

• Augmentor.jl is a real-time library augmentation library for increasing the number of training images.

• MLDataUtils.jl is a utility package for generating, loading, partitioning, and processing Machine Learning datasets.

• UNet.jl is a generic UNet implentation.

• FluxArchitectures is a collection of slightly more advanced network achitectures.

Projects using Flux

Other projects use Flux under the hood to provide machine learning capabilities, or to combine Machine Learning with another domain.

• The Yao project uses Flux and Zygote for Quantum Differentiable Programming.

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

• ObjectDetector.jl provides ready-to-go image analysis via YOLO.

• TextAnalysis.jl provides several NLP algorithms that use Flux models under the hood.

• RayTracer.jl combines ML with computer vision via a differentiable renderer.

• Turing.jl extends Flux’s differentiable programming capabilities to probabilistic programming.

• Stheno provides flexible Gaussian processes.

• Omega is a research project aimed at causal, higher-order probabilistic programming.

• Mill helps to prototype flexible multi-instance learning models.

• Torch.jl exposes torch in Julia.

See also academic work citing Flux or Zygote.

–