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 2021/05/07 00:18:50 UTC

[GitHub] [incubator-mxnet] Zha0q1 edited a comment on pull request #20249: [v1.x] ONNX Add an option to un-interleave BERT

Zha0q1 edited a comment on pull request #20249:
URL: https://github.com/apache/incubator-mxnet/pull/20249#issuecomment-833965061


   > It probably makes more sense to have gluonnlp export a BERT graph that doesn't involve interleaving than doing it in mxnet. framework shouldn't have knowledge about or rely on the implementation of its ecosystem packages.
   
   I agree. Alternatively we might be able to make mx2onnx support loading custom conversion functions dynamically and put these functions to gluonnlp? Or if even that is too hacky this pr can serve as a poc to evaluate the performance benefit with trt 8.0 by un-interleaving the matrix multiplication


-- 
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