Manually specified posteriors and gradients or mixing with external packages #1685
Unanswered
RichardRedding
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Turing.jl looks great but I would like to apply it to existing models that have been developed independently of the Turing ecosystem. For example, in other projects we have built models with log-posterior and gradient functions and we want to apply the tools in Turing to these models (e.g. NUTS/VI,...). Is this possible and, if so, can someone point me to some resource so that we can give it a go?
It was not obvious to me how I would do it with Turing.jl but it was relatively easy to apply AdvancedHMC.jl to our models by following the documentation. Given that AdvancedHMC is in the same family of packages as Turing, this gave me hope that it was possible to plug our models into Turing somehow.
Beta Was this translation helpful? Give feedback.
All reactions