Qiu 84f6cd741b [Mics] add pcp basic support to MoE model (#31003)

2025-12-30 20:01:29 -08:00

..

configs

Add Fused MoE Triton kernels for GLM-4.5-Air, GLM-4.5v, GLM-4.6v on 2x RTX Pro 6000 (#31407)

2025-12-28 08:38:33 -08:00

__init__.py

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

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

all2all_utils.py

[Feat][Perf] Enable deepep-low-latency with round-robin expert placement. (#28449)

2025-11-19 13:46:24 +01:00

batched_deep_gemm_moe.py

[Feature] Add SM103 (Blackwell Ultra) Support to vLLM (#30484)

2025-12-12 19:34:23 -08:00

config.py

[Mics] add pcp basic support to MoE model (#31003)

2025-12-30 20:01:29 -08:00

cpu_fused_moe.py

[CPU] Refactor CPU fused MOE (#30531)

2025-12-18 14:36:49 +08:00

cutlass_moe.py

[MoE Refactor][3/N] Deprecate cutlass block quant fp8 (b200) (#30990)

2025-12-19 13:09:54 -08:00

deep_gemm_moe.py

[Feat] Refactor for parallel_config in FusedMoEModularKernel (#30282)

2025-12-15 04:21:36 +00:00

deep_gemm_utils.py

[Perf] Optimize deepgemm experts initialization, 3.9% TTFT improvement (#30494)

2025-12-11 17:28:34 -05:00

deepep_ht_prepare_finalize.py

[Bug] Fix DBO IMA issue for DeepEPHT (#27666)

2025-10-29 16:28:27 -04:00

deepep_ll_prepare_finalize.py

[MoE] CuteDSL MoE with Nvfp4 DeepEP dispatch (#27141)

2025-11-30 16:05:32 -08:00

flashinfer_cutedsl_moe.py

[MoE] CuteDSL MoE with Nvfp4 DeepEP dispatch (#27141)

2025-11-30 16:05:32 -08:00

flashinfer_cutlass_moe.py

Feature: Support Relu2 in FusedMoE fp8 cutlass path (#27261)

2025-11-16 13:39:44 -05:00

flashinfer_cutlass_prepare_finalize.py

[Kernels] Enable FlashInfer FP8 Blockscale on SM90 (for TEP DSR1) (#27134)

2025-11-14 08:02:44 -08:00

flashinfer_trtllm_moe.py

[Performance] Support FP8 flashinfer TRTLLM MOE on Qwen3 and Qwen-3next (#27492)

2025-11-10 12:34:57 -05:00

fused_batched_moe.py

Update Optional[x] -> x | None and Union[x, y] to x | y (#26633)

2025-10-12 09:51:31 -07:00

fused_marlin_moe.py

[MoE Refactor][4/N] Marlin Fp8 Mk (#31036)

2025-12-21 12:37:42 -05:00

fused_moe_method_base.py

[Perf] Do FP4 quant before All gather on flashinfer trtllmgen MOE (#30014)

2025-12-16 13:01:48 -08:00

fused_moe_modular_method.py

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

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

fused_moe.py

CustomOp: Unify aiter impl into GroupedTopk (#31221)

2025-12-26 12:44:29 -05:00

gpt_oss_triton_kernels_moe.py

[LoRA] Support FusedMoE LoRA Triton kernel for mxfp4 (#29708)

2025-11-30 10:37:25 +08:00

layer.py

CustomOp: Unify aiter impl into GroupedTopk (#31221)

2025-12-26 12:44:29 -05:00

modular_kernel.py

[moe] Use enable_chunking func (to support disabling chunking) (#29935)

2025-12-18 09:02:38 +00:00

moe_align_block_size.py

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

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

moe_pallas.py

Convert formatting to use ruff instead of yapf + isort (#26247)

2025-10-05 07:06:22 -07:00

moe_permute_unpermute.py

Update Optional[x] -> x | None and Union[x, y] to x | y (#26633)

2025-10-12 09:51:31 -07:00

moe_torch_iterative.py

Convert formatting to use ruff instead of yapf + isort (#26247)

2025-10-05 07:06:22 -07:00

pplx_prepare_finalize.py

[Chore]:Extract math and argparse utilities to separate modules (#27188)

2025-10-26 04:03:32 -07:00

prepare_finalize.py

[MoE Refactor][7/N] AITER MK (#31102)

2025-12-22 16:42:58 -07:00

rocm_aiter_fused_moe.py

[Bugfix][ROCm] Fix Static Quant Issue (#31502)

2025-12-29 13:27:55 -08:00

routing_simulator.py

[Bugfix] Fix MoE Routing Simulation (#28002)

2025-11-03 22:26:49 +00:00

shared_fused_moe.py

[Feat] Enable eplb with default all2all backend (#30559)

2025-12-16 10:33:52 -05:00

topk_weight_and_reduce.py

Update Optional[x] -> x | None and Union[x, y] to x | y (#26633)

2025-10-12 09:51:31 -07:00

triton_deep_gemm_moe.py

[Feature] Migrate DeepGEMM API from get_m_alignment_for_contiguous_layout to get_mk_alignment_for_contiguous_layout (#26935)

2025-10-16 16:46:48 -04:00

trtllm_moe.py

[BugFix][Performance] Restore flashinfer autotuning for all scenarios (#27904)

2025-11-04 15:56:21 +08:00

unquantized_fused_moe_method.py

[MoE Refactor][9/N] Use modular kernel for unquantized Triton MoE (#31052)

2025-12-22 17:34:19 +00:00

utils.py

[Chore]:Extract math and argparse utilities to separate modules (#27188)

2025-10-26 04:03:32 -07:00

zero_expert_fused_moe.py

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

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