[Bugfix] Fix quant RMS norm fusion for quantization with TMA-aligned scales (#33255)

Signed-off-by: ElizaWszola <ewszola@redhat.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
This commit is contained in:
ElizaWszola
2026-02-17 23:35:04 -08:00
committed by GitHub
co-authored by Luka Govedič
parent a49ea5a58f
commit a88b3be7c4
12 changed files with 234 additions and 75 deletions
+1 -1
View File
@@ -349,7 +349,7 @@ def _align(x: int, y: int) -> int:
# Taken from https://github.com/deepseek-ai/DeepGEMM/blob/v2.1.1/csrc/utils/math.hpp#L19
def get_tma_aligned_size(x: int, element_size: int):
def get_tma_aligned_size(x: int, element_size: int) -> int:
return _align(x, 16 // element_size)