biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Bugfix] Fix DP/EP Shared Expert With Monolithic Kernels (#36061)

Browse Source 
Signed-off-by: Robert Shaw <robshaw@redhat.com>
Co-authored-by: Robert Shaw <robshaw@redhat.com>
This commit is contained in:
Robert Shaw
2026-03-11 12:07:14 -04:00
committed by GitHub
parent 822e250ab7
commit b7e5a588d8
2 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/model_executor/layers/fused_moe/oracle/fp8.py
View File

@@ -567,7 +567,7 @@ def make_fp8_moe_kernel(
experts,
shared_experts=(
shared_experts
if moe_config.moe_parallel_config.use_all2all_kernels
if moe_config.moe_parallel_config.use_deepep_ll_kernels
else None
),
moe_parallel_config=moe_config.moe_parallel_config,

2
vllm/model_executor/layers/fused_moe/oracle/nvfp4.py
View File

@@ -433,7 +433,7 @@ def make_nvfp4_moe_kernel(
experts,
shared_experts=(
shared_experts
if moe_config.moe_parallel_config.use_all2all_kernels
if moe_config.moe_parallel_config.use_deepep_ll_kernels
else None
),
moe_parallel_config=moe_config.moe_parallel_config,