Max Balandat
Max Balandat
Thanks for flagging this. This is something we should fix on our end, the sensitivity analysis should work regardless of which device the `input_function` operates on even in the `SobolSensitivity`...
Interesting. You're essentially trying to use time as a "contextual" variable. Unfortunately, I don't think this is easily supported in the current service API, but it's something that we want...
> Would it be possible to implement something using the Developer API, and informing the model of the correct time after a new trial has been suggested? Yes, that should...
@SebastianAment is the requirement that all inputs are contained in the unit cube critical for this kernel?
Yeah I think this is sensible. The constructor of `MixedSingleTaskGP` is kind of a beast already so we should think about if/how we can simplify that a bit. Maybe factoring...
> So we first brainstorm how to refactor it before I implement the categorical_kernel_factory? I don't think we need to do that - I think it's fine to implement this...
We should just upgrade to 3.9+ as py3.8 is EOL anyway. cc @jandylin, @SebastianAment re the Kronecker library
> FastKron performs orders of magnitude faster (0.9x to 21x) than current algorithm on both x86 CPUs and NVIDIA GPUs Can you share the benchmarks that you ran for this?
Hmm yeah `d` should just not be sampled from a normal distribution - you'll end up adding negative values to the diagonal. Something like `d = 1e-9 + torch.rand(1000)` should...