frozenleaves/pytorch
1
0
Fork 0
mirror of https://github.com/pytorch/pytorch.git synced 2025-10-20 21:14:14 +08:00
Code Issues Packages Projects Releases Wiki Activity

Fix B200 test fails in scaled_mm (#165747)

Browse Source 
Summary:

PR #165528 changes some scale/swizzle inference behavior in scaled_mm
tests - mxfp8 tests on Blackwell can get incorrectly classified,
resulting in failures.

Fix the scale/swizzle inference code to prevent this.

Fixes https://github.com/pytorch/pytorch/issues/165743

Test Plan:

```
pytest -svv test/test_scaled_matmul_cuda.py
```

Reviewers:

@jagadish-amd @jeffdaily @drisspg

Subscribers:

@Aidyn-A

Tasks:

Tags:
Signed-off-by: Simon Layton <simonlaytonmeta.com>
Pull Request resolved: https://github.com/pytorch/pytorch/pull/165747
Approved by: https://github.com/eqy, https://github.com/drisspg, https://github.com/jeffdaily
This commit is contained in:
Simon Layton
2025-10-17 05:44:35 -07:00
committed by PyTorch MergeBot
parent dd3b48e85d
commit 39e0a832c9
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
test/test_scaled_matmul_cuda.py
View File

@ -154,8 +154,8 @@ def infer_scale_swizzle(mat, scale):
# MXFP4 w/o swizzle
if (
scale.numel() == 2 * math.ceil(mat.shape[0] // 32) * mat.shape[1]
or scale.numel() == 2 * math.ceil(mat.shape[1] // 32) * mat.shape[0]
(scale.numel() == 2 * math.ceil(mat.shape[0] // 32) * mat.shape[1]
or scale.numel() == 2 * math.ceil(mat.shape[1] // 32) * mat.shape[0])
and mat.dtype == torch.float4_e2m1fn_x2
and scale.dtype == torch.float8_e8m0fnu
):