Getting Turing to work with a hierarchical multinomial regression model

[jfhawkin]

You might find my post on this topic helpful. I’m working on a hierarchical multinomial problem, as well. I was able to improve the performance with a few tricks from the Turing developers.

Porting multi-level regression with poststratification (MRP) tutorial from rstanarm

I ran my code above through a proper benchmark and Turing actually run in comparable time to rstanarm (about 45 minutes). I’ve marked my last post as the solution. Any improvements would simply add an extra speed boost.