Possible direct users

[gdalle]

Meta-issue to survey users of DifferentiationInterface and the related discussions.

Current users

Registered

ContinuousNormalizingFlows.jl

• Add Support for DifferentiationInterface impICNF/ContinuousNormalizingFlows.jl#397
• add DI support impICNF/ContinuousNormalizingFlows.jl#399

CTBase.jl and CTFlows.jl

• AD: change the backend transparently control-toolbox/CTBase.jl#25
• Start using DifferentiationInterface control-toolbox/CTBase.jl#140
• WIP: Differentiationinterface control-toolbox/CTBase.jl#141

Fronts.jl

• Switch to DifferentiationInterface gerlero/Fronts.jl#184

ImplicitDifferentiation.jl

• Switch to DifferentiationInterface JuliaDecisionFocusedLearning/ImplicitDifferentiation.jl#135

ItemResponseFunctions.jl

• Add 5PL model JuliaPsychometrics/ItemResponseFunctions.jl#14

MarginalLogDensities.jl

• Work out good approach for automatic Hessian sparsity detection ElOceanografo/MarginalLogDensities.jl#13
• Calculate sparse Hessians using DifferentiationInterface ElOceanografo/MarginalLogDensities.jl#25

OptimizationBase.jl and Optimization.jl

• Features necessary for Optimization.jl #89
• [WIP]Start experimenting with DI SciML/OptimizationBase.jl#31
• Optimization.LBFGS() can not compute the gradient SciML/Optimization.jl#744
• OptimizationBase v2.0 plan SciML/OptimizationBase.jl#62
• [WIP] Fresh attempt at DI integration SciML/OptimizationBase.jl#54

SimpleNonlinearSolve.jl

• Use DifferentiationInterface SciML/SimpleNonlinearSolve.jl#148

Not registered

Hyperelastics.jl

• Differentiation interface backend TRACER-LULab/Hyperelastics.jl#144

Future users

ADNLPModels.jl

• Useful packages for (sparse) differentiation? JuliaSmoothOptimizers/ADNLPModels.jl#229

AutoDiffOperators.jl

• DifferentiationInterface oschulz/AutoDiffOperators.jl#17

DataDrivenLux.jl

• Switch from AbstractDifferentiation to DifferentiationInterface? SciML/DataDrivenDiffEq.jl#519

DifferentialEquations.jl

• autodiff in default algorithm SciML/DifferentialEquations.jl#1027

Flux.jl

• Test against primal from within AD or outside of it #106
• Comparison of outputs between CPU and GPU #107

Lux.jl

• Externalize gradient computations to DifferentiationInterface.jl? LuxDL/Lux.jl#544
• Continuous Benchmarking  LuxDL/Lux.jl#573
• Capture DifferentiationInterface calls for efficient Nested AD LuxDL/Lux.jl#600

MuseInference.jl

• Switch from AbstractDifferentiation to DifferentiationInterface? marius311/MuseInference.jl#19

NNLib.jl

• Testing NNLib / Lux / Flux #105
• Start replacing tests with DifferentiationInterface FluxML/NNlib.jl#570

NonlinearSolve.jl

• Start using DifferentiationInterface.jl SciML/NonlinearSolve.jl#415

ProximalAlgorithms.jl

• Switch from AbstractDifferentiation to DifferentiationInterface JuliaFirstOrder/ProximalAlgorithms.jl#93

SciMLSensitivity.jl

• What would we need to become backend independent via using DifferentiationInterface? SciML/SciMLSensitivity.jl#1040

SNOW.jl

• Try out DifferentiationInterface for sparse autodiff? byuflowlab/SNOW.jl#17

Turing.jl through LogDensityProblemsAD.jl

• Can DifferentiationInterface be useful for Turing? TuringLang/Turing.jl#2187
• Can DifferentiationInterface.jl replace the AD backend extensions here? tpapp/LogDensityProblemsAD.jl#26
• Switch to DifferentiationInterface tpapp/LogDensityProblemsAD.jl#29

VoronoiFVM.jl

• Try out DifferentiationInterface for sparse autodiff? WIAS-PDELib/VoronoiFVM.jl#109

WaterLily.jl

• WaterLily using ForwardDiff and Enzyme through DifferentiationInterface #355

[prbzrg]

I like this package 👍 ContinuousNormalizingFlows will use it from now on 🚀
impICNF/ContinuousNormalizingFlows.jl#399

[ElOceanografo]

Currently working on integrating this into MarginalLogDensities...

[gerlero]

Fronts is about to switch to DifferentiationInterface: gerlero/Fronts.jl#184

[gdalle]

Closing because there are starting to be too many to keep track