Why SVGD does not display average loss or any metric during the fitting?

Questions
1

The title is pretty much what I want to ask.
As an example here is the code in a tutorial in PyMC3 website:

w = pm.floatX([.2, .8])
mu = pm.floatX([-.3, .5])
sd = pm.floatX([.1, .1])
with pm.Model() as model:
    pm.NormalMixture('x', w=w, mu=mu, sigma=sd)
    approx = pm.fit(method=pm.SVGD(n_particles=200, jitter=1.))

I imagine I need to add a callback function, but adding pm.callbacks.CheckParametersConvergence did not solve the problem.