I don't know why there will be errors, I have no problem testing with test_l2_normalization offline. @leezu I think using **Tensor** may slow down the calculation speed, after calculating **nrm2** , Is there any other good method? @haojin2 @piiswrong @leezu @anirudh2290 @szha
[ Full content available at: https://github.com/apache/incubator-mxnet/pull/12287 ] This message was relayed via gitbox.apache.org for [email protected]
