KLqp underline implementation

1

I’m new to Edward. I’m planing to use Edward for research purposes. I’m interested in knowing the concepts behind some of the underline implementation, especially about KLpq and KLqp. Can someone please answer the following questions.

  1. Does KLqp use stochastic variational inference ?
  2. What is the underline implementation of KLqp and KLpq ? Is it ADVI or Black-box variational inference?
  3. I found that KLqp support sub-sampling. If KLqp uses ADVI, what techniques can we used to extend it (compensate for dataset size - N) for streaming ML ?

I appreciate a lot if someone can answer these questions

Thanks!!