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

[MoE Refactor] Introduce MoERunner abstraction and move execution logic from FusedMoE to DefaultMoERunner (#32344)

Browse Source 
Signed-off-by: Bill Nell <bnell@redhat.com>
This commit is contained in:
bnellnm
2026-02-10 19:51:07 -05:00
committed by GitHub
parent dc6de33c3d
commit d1481ba783
25 changed files with 913 additions and 753 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
docs/design/moe_kernel_features.md
View File

@@ -32,7 +32,7 @@ th {
| Backend | Output act. format | Quant. types | Quant. format | Async | Apply Weight On Input | Subclass |
|---------|--------------------|--------------|---------------|-------|-----------------------|-----------|
| naive | standard | all<sup>1</sup> | G,A,T | N | <sup>6</sup> | [layer.py][vllm.model_executor.layers.fused_moe.layer.FusedMoE.forward_impl] |
| naive | standard | all<sup>1</sup> | G,A,T | N | <sup>6</sup> | [layer.py][vllm.model_executor.layers.fused_moe.layer.FusedMoE |
| pplx | batched | fp8,int8 | G,A,T | Y | Y | [`PplxPrepareAndFinalize`][vllm.model_executor.layers.fused_moe.pplx_prepare_finalize.PplxPrepareAndFinalize] |
| deepep_high_throughput | standard | fp8 | G(128),A,T<sup>2</sup> | Y | Y | [`DeepEPLLPrepareAndFinalize`][vllm.model_executor.layers.fused_moe.deepep_ll_prepare_finalize.DeepEPLLPrepareAndFinalize] |
| deepep_low_latency | batched | fp8 | G(128),A,T<sup>3</sup> | Y | Y | [`DeepEPHTPrepareAndFinalize`][vllm.model_executor.layers.fused_moe.deepep_ht_prepare_finalize.DeepEPHTPrepareAndFinalize] |