Model selection in Variational inference

Hi, thanks for your explanation on this topic. I’m also quite interested in model comparison in pyro. You mentioned that “it makes more sense to look at quantities like predictive log-likelihood on held-out data”. Could you please give a hint on how to calculate predictive log-likelihood for SVI in pyro?

I have opened a similar topic and it would be highly appreciated if you could take a look!
Bayesian Model Comparison for SVI