st--

Just so I understand better, what do we gain from defining e.g. ExpLink explicitly (rather than "only" having GenericLink(exp))?

To get a better understanding of all the moving pieces, it'd be really good to have some concrete examples (in code) of likelihoods both with additional parameters and multiple latent...

I can see the appeal, but then I think we should have a proper type for this. Dispatching on generic NamedTuple without specifying what fields that tuple might contain seems...

Hi @tanhevg, we're building a generally applicable Gaussian process API for Julia over at https://github.com/JuliaGaussianProcesses/AbstractGPs.jl/ (building on top of [KernelFunctions.jl](https://github.com/JuliaGaussianProcesses/KernelFunctions.jl/)), if you're interested in moving your package to build on...

Thanks for responding! Could you clarify what you mean by "enforcing the dimension" (what & where)?

(Just to clarify, I used GPyTorch==1.4.0 and botorch==0.4.0, as the code doesn't run with recent versions due to a bunch of breaking changes in those libraries.)

Well, changing the svgp constructor arguments from `.view(-1)` to provide either `.view(-1, 1)` or `.view(1, -1)` instead still doesn't work... would you be able to fix the script ?

We mainly coordinate on the Turing.jl slack - it has significantly less traffic than the official JuliaLang slack, and has correspondingly more useful history - I sent you a direct...

@SebastianAment we've now got a slack invite link, too: https://join.slack.com/t/turingjl/shared_invite/zt-ubbtg7r8-xtg4WjXEQTG_jvE~estuKw so that should work as well - join us at #kernelfunctions in particular

Ping, if you're still interested in this: )