Richard Zou 654a71fc3c [torch.compile] Improve Cold Start for MoEs (#32805) ...

Signed-off-by: Richard Zou <zou3519@gmail.com>

2026-01-22 10:44:40 -05:00

..

configs

[ROCm][CI] Fix AITER test flakiness by using explicit attention backend (#32346)

2026-01-22 13:55:25 +08:00

oracle

[MoE Refactor] Oracle Select FP8+NVFP4 Kernels In Priority (#32414)

2026-01-21 08:22:33 -05:00

router

[Kernel] Add topk_sigmoid kernel (#31246)

2026-01-21 22:49:51 +00:00

__init__.py

[AMD][ROCm] MoRI EP: a high-performance all2all backend (#28664)

2026-01-22 16:33:18 +08:00

all2all_utils.py

[AMD][ROCm] MoRI EP: a high-performance all2all backend (#28664)

2026-01-22 16:33:18 +08:00

batched_deep_gemm_moe.py

[MoE Refactor] Oracle Select FP8+NVFP4 Kernels In Priority (#32414)

2026-01-21 08:22:33 -05:00

config.py

[AMD][ROCm] MoRI EP: a high-performance all2all backend (#28664)

2026-01-22 16:33:18 +08:00

cpu_fused_moe.py

[Misc] Fix Current vLLM config is not set. warnings, assert to avoid issues in the future (#31747)

2026-01-08 15:20:49 -08:00

cutlass_moe.py

[MoE Refactor] Oracle Select FP8+NVFP4 Kernels In Priority (#32414)

2026-01-21 08:22:33 -05:00

deep_gemm_moe.py

[MoE Refactor] Oracle Select FP8+NVFP4 Kernels In Priority (#32414)

2026-01-21 08:22:33 -05:00

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

Add back missing DeepEP LL params (#31911)

2026-01-07 17:47:54 -05:00

fallback.py

[MoE Refactor] Oracle Select FP8+NVFP4 Kernels In Priority (#32414)

2026-01-21 08:22:33 -05:00

flashinfer_cutedsl_moe.py

[MoE Refactor] Oracle Select FP8+NVFP4 Kernels In Priority (#32414)

2026-01-21 08:22:33 -05:00

flashinfer_cutlass_moe.py

[MoE Refactor] Oracle Select FP8+NVFP4 Kernels In Priority (#32414)

2026-01-21 08:22:33 -05:00

flashinfer_cutlass_prepare_finalize.py

[MoE Refactor][16/N] Apply Refactor to NVFP4 (#31692)

2026-01-08 03:46:27 +00:00

flashinfer_trtllm_moe.py

Bump Flashinfer to v0.6.1 (#30993)

2026-01-21 08:49:50 -08:00

fused_batched_moe.py

[ROCm] fix import for on_gfx9 (#32783)

2026-01-21 18:41:11 +00:00

fused_marlin_moe.py

[MoE Refactor] Oracle Select FP8+NVFP4 Kernels In Priority (#32414)

2026-01-21 08:22:33 -05:00

fused_moe_method_base.py

[MoE Refactor] Separate Router into OO Classes (#30623)

2026-01-18 11:40:49 -05:00

fused_moe_modular_method.py

[MoE Refactor] Separate Router into OO Classes (#30623)

2026-01-18 11:40:49 -05:00

fused_moe.py

[ROCm] fix import for on_gfx9 (#32783)

2026-01-21 18:41:11 +00:00

gpt_oss_triton_kernels_moe.py

[MoE Refactor] Oracle Select FP8+NVFP4 Kernels In Priority (#32414)

2026-01-21 08:22:33 -05:00

layer.py

[torch.compile] Improve Cold Start for MoEs (#32805)

2026-01-22 10:44:40 -05:00

modular_kernel.py

[MoE Refactor] Oracle Select FP8+NVFP4 Kernels In Priority (#32414)

2026-01-21 08:22:33 -05:00

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

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

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

mori_prepare_finalize.py

[AMD][ROCm] MoRI EP: a high-performance all2all backend (#28664)

2026-01-22 16:33:18 +08:00

pplx_prepare_finalize.py

"refactor: refactor_repeated_interfaces" (#32486)

2026-01-18 22:07:01 +08:00

prepare_finalize.py

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

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

rocm_aiter_fused_moe.py

[AMD][ROCm] MoRI EP: a high-performance all2all backend (#28664)

2026-01-22 16:33:18 +08:00

routed_experts_capturer.py

[Bugfix] Refactor to support DP parallel in R3 (#32306)

2026-01-16 15:13:58 +08:00

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_cutlass_moe.py

[MoE Refactor] Oracle Select FP8+NVFP4 Kernels In Priority (#32414)

2026-01-21 08:22:33 -05:00

triton_deep_gemm_moe.py

[MoE Refactor] Oracle Select FP8+NVFP4 Kernels In Priority (#32414)

2026-01-21 08:22:33 -05:00

trtllm_moe.py

Bump Flashinfer to v0.6.1 (#30993)

2026-01-21 08:49:50 -08:00

unquantized_fused_moe_method.py

[MoE Refactor] Oracle Select FP8+NVFP4 Kernels In Priority (#32414)

2026-01-21 08:22:33 -05:00

utils.py

[Feat] Support non-gated MoE with Marlin, NVFP4 CUTLASS, FP8, INT8, compressed-tensors (#32257)

2026-01-15 16:15:05 -08:00

zero_expert_fused_moe.py

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

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