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 2019/12/19 06:22:50 UTC

[GitHub] [incubator-mxnet] xidulu opened a new issue #17118: Custom Gradient compatible with Hybridize

xidulu opened a new issue #17118: Custom Gradient compatible with Hybridize
URL: https://github.com/apache/incubator-mxnet/issues/17118
 
 
   ## Description
   Custom Backward is a feature supported by both tensorflow(https://www.tensorflow.org/api_docs/python/tf/custom_gradient?version=stable) and pytorch(https://pytorch.org/tutorials/beginner/examples_autograd/two_layer_net_custom_function.html). MXNet implements `autograd.Function` https://github.com/apache/incubator-mxnet/pull/7403/files , which is quite similar to the one in pytorch, however, it does not support either Symbolic mode or Hybridize
   

----------------------------------------------------------------
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


With regards,
Apache Git Services