I have now finished porting most of my WebPPL and BLOG models to Turing.jl. They are collected at My Julia Turing.jl (probabilistic programming) page and also available at my GitHub page: https://github.com/hakank/hakank/tree/master/julia/turing
It’s now about 170 models, many are (small/ish) probability problems, and just testing certain probabilistic programming concepts. Hopefully these models might be of some interest.
Last year I started to play with Turing.jl, but had then problems with a certain constructs that I was used to from WebPPL/BLOG, e.g. observing (discrete) variables in the model. This was much easier after Distributions.jl added the great Dirac distribution (which I perhaps tend to overuse now). The last week I’ve added 110 new models, since I got inspired to play with Turing.jl again (induced by an interesting off-line discussion with Ramon D-U as well as the ongoing JuliaCon).
Thanks to all here at Discourse that answered my newbie questions about Turing.jl!