/DifferentiableWasserstein.jl

An automatically differentiable implementation of exact 1st-order Wasserstein distance between two vectors of samples from 1D distributions.

Primary LanguageJuliaOtherNOASSERTION

Watchers