Max-Heinrich Laves
Max-Heinrich Laves
> @mlaves are you planning to update this? Yes, I will update my PR soon.
> > Implementation of explicit_gemm_conv_ND_cpu. > > However, using this seems to be considerably slower than the naive implementation. I guess that materializing the strided input view takes very long....
> > Should I remove this now or is there any chance the reshaping will be faster in the future? > > It's possible, we haven't looked at / optimized...
I incorporated all reviewer suggestions.
I rebased onto main, CI should pass now.
"We truly live in an age of wonders".
This repository has been abandoned by the creator. I have compiled the latest version as universal binary with support for both Intel and M1/M2 processors: https://1drv.ms/u/s!AiF3lPpzI9Tej5Y3JYnW-99Yt3H8IQ Let me know if...
RDM now compiles for macOS 12.2. Fixes #57
What is the need for the arch flags? It compiles just fine as a universal binary without the flags.
@Ritbit Okay, then I will add the arch flags back again.