I have tried to put randomstream in a model ages ago, which seems to work fine: https://nbviewer.jupyter.org/github/pymc-devs/resources/blob/master/BCM/CaseStudies/MemoryRetention.ipynb
Just to confirm, have you try putting a tt.print in the model block so it prints the rng value? I guess a risk here is that the gradient is broken when a rng is added to the model