周周周 5e54770b2e [Feature] 添加 MoE 层 latent mode 支持 (#7382)

2026-04-15 13:57:07 +08:00

..

__init__.py

support w4afp8 EP inference (#3044)

2025-08-25 11:27:45 +08:00

ep.py

[Cleanup] Replace torch proxy alias with public compat API (#7348)

2026-04-13 11:43:26 +08:00

flashinfer_cutedsl_moe.py

[BugFix] fix flashinfer-cutedsl moe nvfp4 (#7120)

2026-04-03 15:43:19 +08:00

fused_moe_backend_base.py

[Feature] 添加 MoE 层 latent mode 支持 (#7382)

2026-04-15 13:57:07 +08:00

fused_moe_blackwell_backend.py

[Feature] support blackwell gemm in ht (#7053)

2026-04-07 19:52:51 +08:00

fused_moe_cutlass_backend.py

[RL]moe bf16 ep support paddle batch_gemm (#7337)

2026-04-11 21:51:12 +08:00

fused_moe_deepgemm_backend.py

[TI-consistent] support quant use pow2scale (#7308)

2026-04-13 00:01:53 -07:00

fused_moe_marlin_backend.py

[Feature] support compute shared experts before combine for better overlap (#6697)

2026-03-17 15:18:51 +08:00

fused_moe_triton_backend.py

[Feature] 添加 MoE 层 latent mode 支持 (#7382)

2026-04-15 13:57:07 +08:00

fused_moe_wint2_backend.py

[Feature] support compute shared experts before combine for better overlap (#6697)

2026-03-17 15:18:51 +08:00

moe.py

[Feature] 添加 MoE 层 latent mode 支持 (#7382)

2026-04-15 13:57:07 +08:00

routing_indices_cache.py

[RL] add stream guard (#6814)

2026-03-13 11:22:26 +08:00

triton_moe_kernels.py

[Optimization] Use a separate driver when using Triton with Paddle (#6897)

2026-03-24 10:56:00 +08:00