Can ADVI fitting be parallelized

Questions
1

If I want to fit an ADVI model with a large number of iterations is it possible to reduce the total fit time by fitting on two cores in parallel? I assume the answer is no but want to check.

Something like this:

with pm.Model() as model:
    pm.Beta('beta', alpha=1, beta=1)
    advi = pm.ADVI(   **njobs=2** )
    advi.fit(1000000)
2

I believe massive parallelization is one of the goals of PyMC4.