zhangyiss/mlx
1
0
Fork 0
mirror of https://github.com/ml-explore/mlx.git synced 2025-10-28 22:28:11 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Align mlx::core::max op nan propagation with NumPy (#2339)

Browse Source 
* Make max op NaN propagation rules align with numpy

* Adding benchmarks and testing for max op nanpropagation

* Pre-commit formatting

* Fix max complex64 nan propagation and add test

* Improve the cpp unittest

* Only check nans on non-integral types in simd_reduce_impl.

* Cleanup using namespace alias

* Add cpu Max nanpropagation. Fix a small fib in cpu max dispatch data types for int8/int16.

* Make the max nanpropagation test more meaningful for integer types

* Remove tuple unpacking syntax to comply with earlier python versions. Add cuda skip to nanpropagation tests, fix cuda implementation in a separate PR.
This commit is contained in:
jhavukainen
2025-07-09 11:26:27 -07:00
committed by GitHub
parent fb4e8b896b
commit 8b9a3f3cea
7 changed files with 131 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
tests/ops_tests.cpp
View File

@@ -1024,6 +1024,10 @@ TEST_CASE("test reduction ops") {
x = array({true, true, true, false, true, false}, {2, 3});
CHECK(array_equal(min(x, 1), array({true, false})).item<bool>());
CHECK(array_equal(min(x, 0), array({false, true, false})).item<bool>());
x = array({1.0f, NAN, 3.0f, 4.0f, 5.0f, 6.0f}, {2, 3});
CHECK(array_equal(max(x, 0), array({4.0f, NAN, 6.0f}), true).item<bool>());
CHECK(array_equal(max(x, 1), array({NAN, 6.0f}), true).item<bool>());
}
// Test logsumexp