@tduncan / Nvidia-Optimization-Stack-VLLM Mirror · read-only / commits / fa028207aa
Fork 0

A high-throughput and memory-efficient inference and serving engine for LLMs

Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git

Synced Nov 7, 00:15 UTC

Fix/resupport nongated fused moe triton (#36412)

Signed-off-by: Shaun Kotek - Nvidia <skotek@nvidia.com> Signed-off-by: Natan Bagrov <nbagrov@nvidia.com> Signed-off-by: Daniel Serebrenik <daserebrenik@nvidia.com> Signed-off-by: zjy0516 <riverclouds.zhu@qq.com> Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com> Signed-off-by: yewentao256 <zhyanwentao@126.com> Signed-off-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com> Signed-off-by: liweiguang <codingpunk@gmail.com> Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io> Signed-off-by: wang.yuqi <noooop@126.com> Signed-off-by: Alex Brooks <albrooks@redhat.com> Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk> Signed-off-by: cong-or <conchubhar.gannon@gmail.com> Signed-off-by: Tushar Shetty <tushar.shetty@abbyy.com> Signed-off-by: Tushar Shetty <54362365+tusharshetty61@users.noreply.github.com> Signed-off-by: jiang1.li <jiang1.li@intel.com> Signed-off-by: zhenwei-intel <zhenwei.liu@intel.com> Signed-off-by: Xin Yang <xyangx@amazon.com> Signed-off-by: Kevin H. Luu <khluu000@gmail.com> Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn> Co-authored-by: nvnbagrov <nbagrov@nvidia.com> Co-authored-by: Sage <80211083+sagearc@users.noreply.github.com> Co-authored-by: danisereb <daserebrenik@nvidia.com> Co-authored-by: Jiangyun Zhu <riverclouds.zhu@qq.com> Co-authored-by: Kunshang Ji <kunshang.ji@intel.com> Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com> Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com> Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com> Co-authored-by: Weiguang Li <codingpunk@gmail.com> Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> Co-authored-by: Li, Jiang <jiang1.li@intel.com> Co-authored-by: wang.yuqi <yuqi.wang@daocloud.io> Co-authored-by: Alex Brooks <albrooks@redhat.com> Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk> Co-authored-by: cong-or <conchubhar.gannon@gmail.com> Co-authored-by: Tushar Shetty <54362365+tusharshetty61@users.noreply.github.com> Co-authored-by: liuzhenwei <zhenwei.liu@intel.com> Co-authored-by: Xin Yang <105740670+xyang16@users.noreply.github.com> Co-authored-by: Kevin H. Luu <khluu000@gmail.com> Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>

Shaun Kotek committed Mar 9, 2026 at 20:01 UTC fa028207aa9d4baa6cfc4863f6f54c4277884e6e
2 files changed +5 -2
vllm/model_executor/layers/fused_moe/fused_batched_moe.py
+1 -1
@@ -912,7 +912,7 @@ class BatchedTritonExperts(mk.FusedMoEExpertsModular):
912
913 @staticmethod
914 def _supports_no_act_and_mul() -> bool:
915 - return False
915 + return True
916
917 @staticmethod
918 def _supports_quant_scheme(
vllm/model_executor/layers/fused_moe/fused_moe.py
+4 -1
@@ -1944,7 +1944,7 @@ class TritonExperts(mk.FusedMoEExpertsModular):
1944
1945 @staticmethod
1946 def _supports_no_act_and_mul() -> bool:
1947 - return False
1947 + return True
1948
1949 @staticmethod
1950 def _supports_quant_scheme(
@@ -1983,6 +1983,9 @@ class TritonExperts(mk.FusedMoEExpertsModular):
1983 MoEActivation.GELU,
1984 MoEActivation.SWIGLUOAI,
1985 MoEActivation.SWIGLUSTEP,
1986 + MoEActivation.SILU_NO_MUL,
1987 + MoEActivation.GELU_NO_MUL,
1988 + MoEActivation.RELU2_NO_MUL,
1989 ]
1990
1991 @staticmethod