#7979th of 206 in ML frameworks and libraries
flux.jl
Pure-Julia machine learning library with layers, losses and pluggable autodiff
Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 4.8k
- License
- custom license
- Last commit
- Nov 2026
- Last release
- Oct 2026
- Language
- Julia
4,699 on 22 Apr 20264,755 on 11 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Flux is a machine learning library written entirely in Julia. It provides layers and losses for MLP, convolutional, recurrent and attention models, and trains with Zygote by default, or Enzyme or Mooncake. It runs on NVIDIA, AMD and Apple GPUs, can compile a training step through Reactant to XLA, and supports MPI and NCCL distributed training.
Who it is for: Julia users building and training neural networks, including researchers
Strengths
- Models are plain Julia: any parameterised callable, such as a closure, works as a model
- Selectable AD engine: Zygote default, Enzyme or Mooncake as alternatives
- GPU support for NVIDIA, AMD and Apple through CUDA, AMDGPU or Metal packages
- Reactant integration fuses the training step into one XLA program for CPU, GPU or TPU
Weaknesses
- Requires Julia, so Python-based model code and tooling do not carry over
- License is reported as NOASSERTION, so terms need manual checking
- GPU use needs a separate backend package loaded alongside Flux
- Model zoo and larger examples live in separate repos or folders, not the core package
What it needs
- GPU optional
- Needs Julia 1.10 or later
- Models: MLP, convolutional, recurrent, attention
What changed
From the release notes. No flag means the note does not say so, not that the upgrade is safe.
Also in ML frameworks and libraries
See all 206| Rank | Project | Score |
|---|---|---|
| 1 | transformersPython library defining open models for text, vision, audio and multimodal tasks | 94 out of 100 |
| 2 | ultralyticsPython library and CLI for training and running YOLO vision models | 94 out of 100 |
| 3 | pytorchPython tensor library with GPU support and tape-based autograd | 93 out of 100 |
| #4 | pytorch-image-modelsPyTorch library of image models, pretrained weights and training scripts | 93 out of 100 |
| #5 | supervisionPython toolkit for loading, annotating and counting computer vision detections | 92 out of 100 |