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

Posted to commits@mxnet.apache.org by GitBox <gi...@apache.org> on 2021/07/21 10:59:57 UTC

[GitHub] [incubator-mxnet] bartekkuncer commented on pull request #19561: [operator] Integrate oneDNN matmul implementation to mxnet dot operator

bartekkuncer commented on pull request #19561:
URL: https://github.com/apache/incubator-mxnet/pull/19561#issuecomment-884098941


   > Please fix warning from sanity check:
   > src/operator/nn/mkldnn/mkldnn_batch_dot-inl.h:59: Add #include for shared_ptr<> [build/include_what_you_use] [4]
   > 
   > Do you find any pattern for worse results? Could you put in the table with results column with number of elements.
   
   The pattern I see is "thin" middle axis along which multiplication is being performed or, as you suggested above, number of elements in tensors. Running more tests to determine where to draw the line.


-- 
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.

To unsubscribe, e-mail: commits-unsubscribe@mxnet.apache.org

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