Glad to hear progress is being made on this side!
I didn’t have time to do much more work on this (just some unpublished code for SGHMC), and I’ve been using TensorFlow recently for some of my research mainly due to its support of more statistical distributions.
I’m looking forward to trying out Pyro soon.
Hello, I assuming this wasn’t finished. Is there another MC or Variational tools for pytorch that anyone is aware of? I’m happy to contribute to developing projects also as bandwidth allows.
never mind. I found Pyro.ai might do the trick.