grain: mir, LLVM, GPU, CUDA, dynamic neural networks

[jmh530]

On Tuesday, 12 June 2018 at 11:10:30 UTC, Per Nordlöw wrote:
> I just discovered
>
> https://github.com/ShigekiKarita/grain
>
> which seems like a very ambitious and active project for making 
> dynamic neural networks run on the GPU using D in front of mir 
> and CUDA.
>
> Are there any long-term goals around this project except for 
> the title?
>
> It would great if someone (author) could write a little 
> background-knowledge (tutorial) around the subject of dynamic 
> neural networks that assists all the details in the examples at
>
> https://github.com/ShigekiKarita/grain/tree/master/example
>
> Further, could parts of grain be refactored out into some 
> generic CUDA-library for use in domains other than dynamic 
> neural networks?

Looks interesting, though it seems the author has only just 
recently tagged the first two releases (3-4 days ago). That 
doesn't mean that I don't agree with your suggestions (more 
examples/tutorials, separate GPU & autograd/NN library), just 
maybe the author has been more focused on basic functionality for 
now.