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 2017/11/09 22:46:16 UTC

[GitHub] kaylimekay opened a new issue #8602: Compute Jacobian using contrib.autograd.grad()

kaylimekay opened a new issue #8602: Compute Jacobian using contrib.autograd.grad()
URL: https://github.com/apache/incubator-mxnet/issues/8602
 
 
   The very nice grad() function in contrib.autograd computes the gradient of a scalar function with respect to an arbitrary number of inputs.
   
   Is there a recommended way to extend this so that it automatically computes the gradients of a vector-valued function, i.e. the Jacobian?  In other words, given a function that returns an array of size N and takes M inputs, grad() would give an NxM array, where each element of the array is the gradient of one of the outputs w.r.t one of the inputs?

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on 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