You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@tvm.apache.org by GitBox <gi...@apache.org> on 2020/12/08 16:22:13 UTC

[GitHub] [tvm] Wheest commented on pull request #7050: Sparse Conv2D for CPU (NCHW)

Wheest commented on pull request #7050:
URL: https://github.com/apache/tvm/pull/7050#issuecomment-740733320


   Hi there @giuseros, I'm testing in both x86 and ARM, and as long as you're not doing autotuning it will use the same code.  Nothing x86 specific yet.   It's just that I've got the code defined in the `x86` TOPI directory for now, perhaps I should move it to the `sparse`, or `nn` one.
   
   See the [gist](https://gist.github.com/Wheest/94433f73ff3279669bf35adcc38b321d) for an example of how it is used.  You can also apply the `data_dep_optimization` to sparsify Conv2D on full networks, 


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org