# LuxDL Docs

> Elegant & Performant Deep Learning in JuliaLang

Model with the elegance of Julia, and the performance of XLA.

## Table of Contents

### Getting Started

- [Why we wrote Lux? {#Why-we-wrote-Lux?}](/dev/introduction/overview.md)
- [Resources to Get Started {#Resources-to-Get-Started}](/dev/introduction/resources.md)
- [Updating to Lux v1 {#updating-to-v1}](/dev/introduction/updating_to_v1.md)
- [Citation {#Citation}](/dev/introduction/citation.md)

### Tutorials

- [Tutorials](/dev/tutorials.md)

#### Beginner

- [Julia & Lux for the Uninitiated {#Julia-and-Lux-for-the-Uninitiated}](/tutorials/beginner/1_Basics.md)
- [Fitting a Polynomial using MLP {#Fitting-a-Polynomial-using-MLP}](/tutorials/beginner/2_PolynomialFitting.md)
- [Training a Simple LSTM {#Training-a-Simple-LSTM}](/tutorials/beginner/3_SimpleRNN.md)
- [MNIST Classification with SimpleChains {#MNIST-Classification-with-SimpleChains}](/tutorials/beginner/4_SimpleChains.md)
- [Fitting with Optimization.jl {#Optimization-Lux-Tutorial}](/tutorials/beginner/5_OptimizationIntegration.md)

#### Intermediate

- [MNIST Classification using Neural ODEs {#MNIST-Classification-using-Neural-ODEs}](/tutorials/intermediate/1_NeuralODE.md)
- [Training a HyperNetwork on MNIST and FashionMNIST {#Training-a-HyperNetwork-on-MNIST-and-FashionMNIST}](/tutorials/intermediate/3_HyperNet.md)
- [Training a PINN on 2D PDE {#Training-a-PINN-on-2D-PDE}](/tutorials/intermediate/4_PINN2DPDE.md)
- [Convolutional VAE for MNIST {#Convolutional-VAE-Tutorial}](/tutorials/intermediate/5_ConvolutionalVAE.md)
- [Graph Convolutional Networks on Cora {#GCN-Tutorial-Cora}](/tutorials/intermediate/6_GCN_Cora.md)
- [Normalizing Flows for Density Estimation {#Normalizing-Flows-for-Density-Estimation}](/tutorials/intermediate/7_RealNVP.md)
- [Building a LSTM Encoder-Decoder model using Lux.jl {#Building-a-LSTM-Encoder-Decoder-model-using-Lux.jl}](/tutorials/intermediate/8_LSTMEncoderDecoder.md)
- [ConvMixer on CIFAR-10 {#ConvMixer-on-CIFAR-10}](/tutorials/intermediate/9_CIFAR10_conv_mixer.md)
- [Convolutional Neural Network on CIFAR-10 {#Convolutional-Neural-Network-on-CIFAR-10}](/tutorials/intermediate/10_CIFAR10_simple_cnn.md)
- [ResNet20 on CIFAR-10 {#ResNet20-on-CIFAR-10}](/tutorials/intermediate/11_CIFAR10_resnet20.md)

#### Advanced

- [Training a Neural ODE to Model Gravitational Waveforms {#Training-a-Neural-ODE-to-Model-Gravitational-Waveforms}](/tutorials/advanced/1_GravitationalWaveForm.md)
- [Denoising Diffusion Implicit Model (DDIM) {#Denoising-Diffusion-Implicit-Model-DDIM}](/tutorials/advanced/2_DDIM.md)
- [ImageNet Classification using Distributed Data Parallel Training {#ImageNet-Classification-using-Distributed-Data-Parallel-Training}](/tutorials/advanced/3_ImageNet.md)
- [Qwen3 Implementation from Scratch {#Qwen3-Implementation-from-Scratch}](/tutorials/advanced/4_Qwen3.md)

### Manual

#### Basics

- [Lux Interface {#lux-interface}](/manual/interface.md)
- [Freezing Model Parameters {#freezing-model-parameters}](/manual/freezing_model_parameters.md)
- [GPU Management {#GPU-Management}](/manual/gpu_management.md)
- [Initializing Weights {#Initializing-Weights}](/manual/weight_initializers.md)
- [Visualizing Lux Models using Model Explorer {#Visualizing-Lux-Models-using-Model-Explorer}](/manual/visualize_lux_models.md)

#### Reactant Compilation

- [Compiling Lux Models using `Reactant.jl` {#reactant-compilation}](/manual/compiling_lux_models.md)
- [Exporting Lux Models to Jax (via EnzymeJAX & Reactant) {#exporting_to_stablehlo}](/manual/exporting_to_jax.md)
- [Profiling Lux Training Loops {#profiling-training-loop-reactant}](/manual/profiling_training_loop.md)

#### Automatic Differentiation

- [Automatic Differentiation {#autodiff-lux}](/manual/autodiff.md)
- [Nested Automatic Differentiation {#nested_autodiff}](/manual/nested_autodiff.md)

#### Debugging / Performance Enhancement Tools

- [Debugging Lux Models {#debug-lux-layers}](/manual/debugging.md)
- [Performance Pitfalls & How to Catch Them {#Performance-Pitfalls-and-How-to-Catch-Them}](/manual/performance_pitfalls.md)

#### Migration Guides

- [Migrating from Flux to Lux {#migrate-from-flux}](/manual/migrate_from_flux.md)
- [Supporting Both Flux and Lux {#flux-lux-interop}](/manual/flux_lux_interop.md)

#### Advanced Usage

- [Dispatching on Custom Input Types {#Dispatching-on-Custom-Input-Types}](/manual/dispatch_custom_input.md)
- [Preferences for Lux.jl {#Preferences-for-Lux.jl}](/manual/preferences.md)
- [Distributed Data Parallel Training {#Distributed-Data-Parallel-Training}](/manual/distributed_utils.md)

### API Reference

#### Lux

- [Built-In Layers {#Built-In-Layers}](/api/Lux/layers.md)
- [Automatic Differentiation Helpers {#autodiff-lux-helpers}](/api/Lux/autodiff.md)
- [Utilities {#Utilities}](/api/Lux/utilities.md)
- [Experimental Features {#Experimental-Features}](/api/Lux/contrib.md)
- [Interoperability between Lux and other packages {#Interoperability-between-Lux-and-other-packages}](/api/Lux/interop.md)
- [Distributed Utils {#Distributed-Utils}](/api/Lux/distributed_utils.md)
- [Serialization {#Serialization}](/api/Lux/serialization.md)

#### Accelerator Support

- [MLDataDevices {#MLDataDevices-API}](/api/Accelerator_Support/MLDataDevices.md)

#### NN Primitives

- [LuxLib {#LuxLib-API}](/api/NN_Primitives/LuxLib.md)
- [NNlib {#NNlib-API}](/api/NN_Primitives/NNlib.md)
- [Activation Functions {#NNlib-ActivationFunctions-API}](/api/NN_Primitives/ActivationFunctions.md)

#### Building Blocks

- [LuxCore {#LuxCore}](/api/Building_Blocks/LuxCore.md)
- [WeightInitializers {#WeightInitializers-API}](/api/Building_Blocks/WeightInitializers.md)

#### Testing Functionality

- [LuxTestUtils {#LuxTestUtils}](/api/Testing_Functionality/LuxTestUtils.md)

### Other

- [Bayesian Neural Network {#Bayesian-Neural-Network}](/dev/tutorials/intermediate/2_BayesianNN.md)
- [Introduction {#getting-started}](/dev/introduction.md)
- [References {#References}](/dev/references.md)