Oliver Schulz
Oliver Schulz
> Do you have a MWE I can try? At the moment, it's definitely not minimal, but I'll see what I can cook up. > (or are guarded with PyGILState_Ensure)...
> around the threaded part of your code, maybe it will work? The problem is, Python is calling an MCMC sampler written in Julia, which then calls back to evaluate...
Thanks a lot @cjdoris , will try it out!
> Unfortunately, in any case it would be a breaking change Argh, yes ... So maybe we change the specializations in "mvnormal.jl" to use things like `MvNormal{
Yes, that sounds good!
I recently ran into trouble too, trying to integrate Distributions into GPU code. Maybe we could make Distributions.jl more GPU friendly step-by-step? Currently, even simple things like ```julia using Distributions,...
I think it's because most distribution default ctors (like `Normal()`) check arguments and can throw an exception (involving string generation), which CUDA doesn't like. One can work around that by...
I think we'll find a few CUDA/GPU-incompatible things like that scattered throughout Distributions, but I hope many will be easy to fix in that regard (except calls to Rmath, obviously,...
> SciMLBase: Really expensive Interesting, how expensive, and due to what? There's not much in there I was very surprised as well, I had always assumed SciMLBase to be very...