Thanks @_eigenfoo, that’s a nice example.
If anyone wanted to see the non-minibatch way of doing this (i.e. iteratively running a whole new model on ~500 samples at a time, but with the priors being informed by the previous run of the model), see: https://nbviewer.jupyter.org/gist/ljmartin/f727560f49cb935ba1de811da859edaf
cheers,
lewis