Early exploration of proximal optimization algorithms in Julia (2020-2021).
This package implemented basic gradient descent methods (standard descent, FISTA) with proximal operators, built before I became familiar with Julia's optimization ecosystem (Zygote.jl, Optimisers.jl). It used a flat tuple-based structure rather than Functors for state management.
Status: Superseded by FluxOptics.jl, which integrates properly with the Julia ecosystem and includes improved implementations of these algorithms.
If you need proximal optimization with automatic differentiation, use FluxOptics.jl instead.