[Bugfix][Quantization] Fix MXFP8 startup crash on layers below mm_mxfp8 shape limits (#54223)

Signed-off-by: samuelkim7 <samuelmwkim@gmail.com>
Signed-off-by: mgoin <mgoin64@gmail.com>
Co-authored-by: mgoin <mgoin64@gmail.com>
This commit is contained in:
Samuel Kim
2026-09-25 02:54:11 +00:00
committed by GitHub
co-authored by mgoin
parent 6936e77ba4
commit f55419c43c
16 changed files with 221 additions and 57 deletions
@@ -66,7 +66,7 @@ from vllm.platforms import PlatformEnum
"MarlinMxfp8LinearKernel",
"EmulationMxfp8LinearKernel",
init_mxfp8_linear_kernel,
{},
{"weight_shape": (2048, 2048)},
),
(
B12xTensorFP8ScaledMMLinearKernel,
@@ -365,7 +365,7 @@ def test_b12x_tensor_fp8_apply_quantizes_and_uses_packed_weight(
def test_b12x_mxfp8_can_implement_supported_config() -> None:
can_implement, reason = B12xMxfp8LinearKernel.can_implement(
Mxfp8LinearLayerConfig()
Mxfp8LinearLayerConfig(weight_shape=(256, 512))
)
assert can_implement