Zha0q1 commented on a change in pull request #19015:
URL: https://github.com/apache/incubator-mxnet/pull/19015#discussion_r478563458
##########
File path: src/operator/numpy/np_nonzero_op-inl.h
##########
@@ -43,6 +43,25 @@ namespace mxnet {
namespace op {
struct NonzeroForwardKernel {
+ // this is for cpu
+ template<int ndim>
+ MSHADOW_XINLINE static void Map(index_t i,
+ int64_t* out,
Review comment:
This int64_t is actually the output data type. According to the original
author this "must be int 64"
https://github.com/apache/incubator-mxnet/blob/a2b77fb311204488bd72f99b681fc0ff334bca8d/src/operator/numpy/np_nonzero_op.cc#L33.
I think the reasoning is because the tensor is flattened, even with
int64tensorsize=off we can have > int32_max many elements e.g.(2**20, 2**20)
----------------------------------------------------------------
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:
[email protected]
