SVI-GP with HMC for hypers / Non-Gaussian likelihoods

vr308 / Generalised-Gaussian-Processes

Fully Bayesian Inference in GPs - Gaussian and Generic Likelihoods

21 stars 1 forks source link

Open vr308 opened 3 years ago

vr308 commented 3 years ago

I believe this is broken in gpytorch but would be good to have for completeness