Matthew Schlegel

Coming kind of late to this. But it would be super neat if when autocompleting unicode commands like `\eta` it displayed the the unicode character in the right column of...

Ok. Neat. From some of the talks at JuliaCon it does seem like sparse operations are still in flux quite a bit in the ecosystem. I may think about this...

Also while increasing the number of threads (BLAS (for OMEinsum) and Julia threads (for Tullio)), Tullio starts approaching OMEinsum in performance.

@mcabbott You are probably correct about the CUDA.@sync. I'm not well versed in CUDAeise, so likely my simple benchmarking is pretty inadequate. Thank you for the tip. In any case,...

Unfortunately, while incorporating this into my system we get gradient code which does scalar operations on the gpu (so is painfully slow). The forward pass is fine, just the backward...

Again putting here because it is related: when doing the gradient the kernel abstraction that is generated for the gradient is: ```Julia KernelAbstractions.@kernel function var"##🇨🇺#473"(𝛥mid, 𝛥a_gpu, 𝛥ℛ::Zygote.Fill{𝒯}, ℛ, mid, a_gpu,...

That is very reasonable logic as this is a bit odd of an operation, and likely not used very often. I'm wondering if there is a way for me to...

After some testing, it seems if I manually activate recentf-mode in 'M-x' and then resize the buffer it is populated as expected once, and then returns to the behavior mentioned...

For sure. I'm just getting started on my custom config (moving away from Spacemacs), so it is pretty bare. https://github.com/mkschleg/.emacs.d

I've started working on the Folded interface. It actually should be pretty easy to add without disrupting the current api, but I haven't thought through how CuDNN fits yet. I'll...