zhangyiss/mlx
1
0
Fork 0
mirror of https://github.com/ml-explore/mlx.git synced 2025-07-19 23:51:14 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Add a special case when not keeping the dims

Browse Source
This commit is contained in:
Angelos Katharopoulos 2025-06-24 10:39:53 -07:00
parent a57a75b992
commit a7faa04cd4
1 changed files with 5 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
mlx/backend/cuda/reduce/reduce_utils.cuh
View File

@ -108,6 +108,11 @@ inline void allocate_same_layout(
array& out,
const array& in,
const std::vector<int>& axes) {
if (out.ndim() < in.ndim()) {
out.set_data(allocator::malloc(out.nbytes()));
return;
}
// Calculate the transpositions applied to in in order to apply them to out.
std::vector<int> axis_order(in.ndim());
std::iota(axis_order.begin(), axis_order.end(), 0);