Kyurae Kim

Implement sparse vector, sparse matrix types using cuSPARSE ### preliminary - none

Implement 1D, 2D, 3D convolution operator CUDA kernel ### preliminary - none

Hi, Sorry for posting an issue for this silly question. Couldn't think of a better way. What is the emacs theme in the screenshot? The color theme, the look of...

On mingw, a variable declaration in a case label needs to be in a separate block. I added a block for the label. ```shell ------------------------------------------------- - Compiler: gcc - Configuration:...

`mapreduce` with accumulation inside is broken

10

Hi, the following use-case of `mapreduce` doesnt work: ```julia gradient(randn(10)) do x y₀ = Float64[] ∑x = 0.0 ys = mapreduce(vcat, x, 1:length(x); init = y₀) do xᵢ, r yᵢ...

This PR updates the internals of `Turing` to match the [rebooted](https://github.com/TuringLang/AdvancedVI.jl/pull/49) `AdvancedVI.jl`. ## Breaking? Changes The only change from the previous interface is that I propose to move the number...

`rrule` for `mean(f, x)` is not vectorized?

2

Hi, it seems that the `rrule` for `mean(f, x)` is not vectorized and thus does not place nicely with CUDA: ```julia using Zygote, CUDA, Statistics julia> gradient(y -> mean(x ->...

Zygote is broken for `Stacked` bijectors

5

Here's a MWE ```julia using Distributions, Bijectors, Zygote dists = [Beta(), InverseGamma()] ranges = [1:2, 3:3] bs = Bijectors.bijector.(dists) binvs = inverse.(bs) stacked = Bijectors.Stacked(binvs, ranges) display(stacked(randn(3))) Zygote.gradient(x -> Bijectors.with_logabsdet_jacobian(stacked,...

Question on simplex bijector implementation

9

Hi, It appears that `torch.probability` simply uses [softmax](https://github.com/pytorch/pytorch/blob/8472c24e3b5b60150096486616d98b7bea01500b/torch/distributions/constraint_registry.py#L246C41-L246C41) for the simplex bijector. Is there a reason our simplex transform is much more complicated? I was also thinking about a GPU-friendly...

Hi, The Turing ecosystem recently changed domain, but `juliapackages` seems to point towards the old domain name: - [Turing.jl](https://juliapackages.com/p/turing): [old](https://juliapackages.com/p/turing) -> [new](https://turinglang.org/stable/) And `DynamicPPL.jl` page points towards the wrong page:...