Has anyone created a cloud based model that calls PyMC from Amazon Web Services? Can you share any solution architecture diagrams? thanks!
There is a (slightly outdated) docker image that might help: Docker
Otherwise, what specifically are you looking for? What type of deployment do you need?
I’d be curious too e.g how do you guys set up a production repeatable inference task. Maybe A/B testing, batch job, ecs framework, lambda functions etc…??