chenxu214 88cfa6c11d [NPU]Releasing redundant memory of w13_weight and nz when the ascend_fuseep feature is enabled (#19813)

2026-03-04 19:26:29 +08:00

..

ep_moe

[NPU]Releasing redundant memory of w13_weight and nz when the ascend_fuseep feature is enabled (#19813)

2026-03-04 19:26:29 +08:00

fused_moe_triton

support fused_moe_triton and moe_sum_all_reduce kernel fusion[reduce … (#19672)

2026-03-04 10:30:33 +08:00

moe_runner

Fix nvfp4 weight update (#18085)

2026-02-27 14:55:08 -08:00

token_dispatcher

[AMD] MORI-EP support for EP4. (#19578)

2026-02-28 13:13:46 -08:00

__init__.py

Add --speculative-moe-runner-backend server arg (#10183)

2025-11-04 00:20:56 -08:00

cutlass_moe_params.py

[CUTLASS-FP4-MOE] Introduce CutlassMoEParams class for easy initialization of Cutlass Grouped Gems Metadata (#6887)

2025-06-05 13:13:14 -07:00

cutlass_moe.py

Add mxfp8 support for online quantization, Triton dense linear, and CUTLASS MoE (#17449)

2026-01-29 21:33:57 +08:00

cutlass_w4a8_moe.py

update pre-commit config (#18860)

2026-02-16 00:18:31 +08:00

flashinfer_cutedsl_moe.py

Replace [silu_and_mul_]scaled_fp4_group_quant by Flashinfer equivalent (#12376)

2025-11-13 00:26:00 -08:00

fused_moe_native.py

[11/N] MoE Refactor: Simplifying SBO Implementation with Dispatcher Hooks (#13327)

2025-11-21 01:11:37 -08:00

kt_ep_wrapper.py

[2/N] Quantization Refactor: Compressed tensors MoE schemes (#17503)

2026-02-16 18:03:51 +03:00

rocm_moe_utils.py

Integration mori backend for EP a2a data communication (#17012)

2026-01-28 19:07:34 -08:00

routed_experts_capturer.py

[Feature] Enable return routed experts (#12162)

2025-12-21 15:16:43 +08:00

router.py

move more files under srt/utils (#11285)

2025-10-09 16:46:15 -07:00

topk.py

Piecewise Cuda Graph set default (#16331)

2026-03-02 23:18:07 +08:00

utils.py

Clean up noisy startup log messages and refactor loader.py (#18531)

2026-02-11 16:12:57 -08:00