Anton Lokhmotov
Anton Lokhmotov
Happy to find out about this project, and thanks for sharing benchmark results! Wondering where would your results lie on the [ReQuEST](http://cknowledge.org/request.html) [scoreboard](http://cknowledge.org/request-results)? Specifically for MobileNets v1/v2, are you using...
@llhe We may be able to help with [CK](https://github.com/ctuning/ck) integration for a couple of programs just to get started. Would you be interested in that?
We have already [added](https://github.com/ctuning/ck-math/pull/30) a CK package for MACE.
Has time come to use [GEMMBench](https://github.com/dividiti/gemmbench) and [CK-Lift](https://github.com/lift-project/ck-lift)? ;))
It appears that a third-party won't be able to build the Habana backend since the paths are hardcoded (` /usr/lib/habanalabs`)?
cc: @bertmaher
Thank you so much for the explanation @mciprian13! I have a prototype graph pass to fold MatMul and BachedAdd into a new atomic node (inspired by your PR https://github.com/pytorch/glow/pull/4280). But...
@mciprian13 I see. Thanks for the hint! My source has Gemms instead of FCs but may be those can be folded into FCs.
I've got another interesting observation comparing two Raspberry Pi 4's: 322 ms on a new Model B Rev 1.4 8GB @ 1800 MHz vs 315 ms on a Model B...
@MatthewARM, how do I dump event profiles? Can I do it from a release build?