vLLM/vllm/model_executor/layers/fused_moe at b87575d24b752ee2afffaec00bfb244570d5c95a - vLLM

configs

[ROCm] Enable VLLM triton FP8 moe for gfx1201, tuned for Qwen3-30B-A3B-FP8 tp=2 and Qwen/Qwen3.5-35B-A3B-FP8 tp=2 (#38086 )

2026-04-02 08:13:42 +00:00

experts

[Bug] Fix routing bias dtype for trtllm per-block fp8 moe (#38989 )

2026-04-08 19:42:43 -07:00

oracle

[MoE] Move DEEP_GEMM into experts/ subdirectory (#39005 )

2026-04-08 19:23:08 +00:00

prepare_finalize

[MoE Refactor] Make SharedExperts class for use with DefaultMoERunner (#35153 )

2026-04-01 09:44:08 -04:00

router

Revert "[Kernel] Add gpt-oss Router GEMM kernel (#37205 )" (#38778 )

2026-04-01 22:02:32 -07:00

runner

[compile] Allow strings in custom ops without regressing compilation times (#38123 )

2026-04-10 07:26:37 +00:00

__init__.py

[MoE] Move DEEP_GEMM into experts/ subdirectory (#39005 )

2026-04-08 19:23:08 +00:00

activation.py

[Refactor] Replace activation: str with MoEActivation enum (#33843 )

2026-02-11 17:29:32 -08:00

all2all_utils.py

[ROCm] Enable MORI EP for unquantized MoE with AITER backend (#37529 )

2026-03-30 15:19:33 +08:00

config.py

[MoE Refactor] Make SharedExperts class for use with DefaultMoERunner (#35153 )

2026-04-01 09:44:08 -04:00

cpu_fused_moe.py

[CPU] Support gelu act in cpu_fused_moe (#38770 )

2026-04-02 14:14:32 +08:00

cutlass_moe.py

[MoE Refactor] Make SharedExperts class for use with DefaultMoERunner (#35153 )

2026-04-01 09:44:08 -04:00

deep_gemm_utils.py

[Bugfix] ep_scatter kernel store-load race condition (#34991 )

2026-03-13 01:07:59 +00:00

fallback.py

[Feature]: Remove Chunking From FusedMoE (#34086 )

2026-03-12 14:24:38 -04:00

flashinfer_cutlass_moe.py

[NVIDIA] Bugfix NVFP4 DGX Spark and RTX50 (#38423 )

2026-03-30 09:36:18 -07:00

fused_batched_moe.py

[ROCm] Enable DeepEP ROCm as all2allbackend for AMD GPUs. (#34692 )

2026-03-21 00:32:31 -07:00

fused_marlin_moe.py

[Kernel] Add MXFP8 to Marlin GEMM/MoE and refactor Mxfp8LinearOp (#34664 )

2026-04-01 09:41:42 -07:00

fused_moe_method_base.py

[MoE Refactor] Make SharedExperts class for use with DefaultMoERunner (#35153 )

2026-04-01 09:44:08 -04:00

fused_moe_modular_method.py

[MoE Refactor] Make SharedExperts class for use with DefaultMoERunner (#35153 )

2026-04-01 09:44:08 -04:00

fused_moe.py

[ROCm] Enable VLLM triton FP8 moe for gfx1201, tuned for Qwen3-30B-A3B-FP8 tp=2 and Qwen/Qwen3.5-35B-A3B-FP8 tp=2 (#38086 )

2026-04-02 08:13:42 +00:00

gpt_oss_triton_kernels_moe.py

[Kernels][MoE] Fix legacy_routing to use bitmatrix-based routing path (#38504 )

2026-04-07 10:57:09 +08:00

layer.py

[MoE Refactor] Split of DefaultMoERunner class (#35326 )

2026-04-06 12:41:59 -04:00

modular_kernel.py

[MoE Refactor] Make SharedExperts class for use with DefaultMoERunner (#35153 )

2026-04-01 09:44:08 -04:00

moe_align_block_size.py

[Kernel][MoE] optimize moe_align_block_size (#29642 )

2025-12-07 01:58:47 -08:00

moe_permute_unpermute.py

[Refactor] Remove align block size logic in moe_permute (#33449 )

2026-02-06 10:57:06 -08:00

mori_prepare_finalize.py

[Bugfix][ROCm] Fix MoRI + AITER FP8 dispatch compatibility for defer_input_quant (#37418 )

2026-03-19 09:49:27 +00:00

nixl_ep_prepare_finalize.py

[2/N] Elastic EP Milestone 2: Integrating NIXL-EP (#35627 )

2026-03-13 09:25:33 -04:00

rocm_aiter_fused_moe.py

[Refactor] Move FusedMoE hidden_size roundup to quant_method (#34285 )

2026-03-26 23:38:26 -07:00

routed_experts_capturer.py

[Bugfix] fix device_name for routing replay (#34336 )

2026-02-26 12:18:38 +00:00

shared_fused_moe.py

[MoE Refactor] Make SharedExperts class for use with DefaultMoERunner (#35153 )

2026-04-01 09:44:08 -04:00

topk_weight_and_reduce.py

[MoE Refactor] Create MK for TRTLLM Kernels (#32564 )

2026-03-03 10:39:50 -08:00

triton_cutlass_moe.py

[MoE Refactor] Create MK for TRTLLM Kernels (#32564 )

2026-03-03 10:39:50 -08:00

triton_deep_gemm_moe.py

[MoE] Move DEEP_GEMM into experts/ subdirectory (#39005 )

2026-04-08 19:23:08 +00:00

unquantized_fused_moe_method.py

[XPU] Quick fix for TritonMLA to remove cuda hardcode (#39088 )

2026-04-08 00:17:58 +08:00

utils.py

[Quantization] Support Quark W8A8 INT8 MoE inference (#36320 )

2026-04-09 17:24:43 +00:00

xpu_fused_moe.py

[XPU][MoE Refactor] Refactor xpu mxfp4 support into oracle (#37784 )

2026-03-23 11:10:41 +00:00

zero_expert_fused_moe.py

[MoE Refactor][5/N] Isolate zero expert to LongCatFlash (#28891 )

2025-12-20 18:22:04 +00:00