How to use GLMM guides in SVI

How to use the guides defined in glmm/glmm.py and glmm/guides.py such as NormalInverseGammaGuide in SVI.

The example on OED oed/ab_test.py shows the use of a model with normal_guide for calculating posterior entropy.

Is there a way to use these guides in SVI for inference? Like the autoguides:

guide = AutoMultivariateNormal(model)
svi = SVI(model, guide, opt, loss, num_samples)