Skip to content
This repository was archived by the owner on Sep 4, 2025. It is now read-only.

Commit 8678a69

Browse files
dsikkaElizaWszola
andauthored
[Kernel] Expand MoE weight loading + Add Fused Marlin MoE Kernel (vllm-project#7527)
Co-authored-by: ElizaWszola <[email protected]>
1 parent 5844017 commit 8678a69

File tree

15 files changed

+2374
-84
lines changed

15 files changed

+2374
-84
lines changed

CMakeLists.txt

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -286,7 +286,8 @@ define_gpu_extension_target(
286286

287287
set(VLLM_MOE_EXT_SRC
288288
"csrc/moe/torch_bindings.cpp"
289-
"csrc/moe/topk_softmax_kernels.cu")
289+
"csrc/moe/topk_softmax_kernels.cu"
290+
"csrc/moe/marlin_moe_ops.cu")
290291

291292
define_gpu_extension_target(
292293
_moe_C

0 commit comments

Comments
 (0)