Zixin HuangandClaude Fable 5.1 f447bb7080 [Kernel] Coalesce the KDA CuTe DSL decode state transpose: ~3x faster, bit-identical (#39680) ...

Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>

2026-09-18 10:53:13 +08:00

..

activation

fix(moe): cast filtered-activation expert_ids to int32 for torch.compile (#38085)

2026-09-05 17:13:07 +08:00

attention

[Kernel] Coalesce the KDA CuTe DSL decode state transpose: ~3x faster, bit-identical (#39680)

2026-09-18 10:53:13 +08:00

communication

dsv4.1: communication kernels and wrappers (#39653)

2026-09-16 02:09:40 -07:00

diffusion

[diffusion] Fix the XPU capability gates that broke the Wan2.2 A14B DiT path (#36825)

2026-09-17 12:15:07 +08:00

elementwise

support qwen 3.8 flash next (#37500)

2026-09-08 13:56:21 -07:00

embeddings

dsv4.1: standalone kernels and Python wrappers (#39646)

2026-09-15 22:31:18 -07:00

gemm

dsv4.1: standalone kernels and Python wrappers (#39646)

2026-09-15 22:31:18 -07:00

grammar

[CI][RFC] Replace black-jupyter with ruff-format (#37210)

2026-09-02 19:46:08 -07:00

kimi_k3

[Kimi-K3] Accept fp32 routing weights in the fused MoE finalize (#38612)

2026-09-09 01:33:39 -07:00

kv_canary

Fix KV-canary workspace accounting after graph capture (#38596)

2026-09-10 16:10:12 -07:00

kvcache

[ROCm] Raise HiCache JIT block quota for mapped-host throughput (#39036)

2026-09-11 00:12:24 -07:00

layernorm

dsv4.1: mHC computation and compensated projections (#39664)

2026-09-16 17:11:55 -07:00

lplb

[CI][RFC] Replace black-jupyter with ruff-format (#37210)

2026-09-02 19:46:08 -07:00

mamba

[XPU][Fix] Pack device-pointer tables as uint64 to avoid 64-bit address overflow (#35051)

2026-09-10 10:12:18 +08:00

memory

[PD] Add optional KV transfer checksums (#39500)

2026-09-16 23:53:30 +08:00

minicpm_sala

Refactor JIT kernel and expert-pack directory layout (#36704)

2026-08-29 07:41:25 +08:00

mm

feat(kernels): port standalone Kimi K3 kernels (#32890)

2026-08-01 13:26:47 +08:00

moe

[Intel XPU] Enable fused_moe_triton tuning on XPU and add tuned DeepSeek-OCR-2 configs (#28723)

2026-09-16 10:20:22 +08:00

quantization

dsv4.1: Hopper FP8 matmul kernels and tuning (#39657)

2026-09-16 16:54:19 -07:00

sampling

[CPU] Add native CPU kernel for MurmurHash32 (#35604)

2026-09-07 09:10:35 +08:00

speculative

[Kernel] Move CUDA and ROCm speculative kernels to JIT (#40033)

2026-09-17 17:27:26 -07:00

__init__.py

Refactor JIT kernel and expert-pack directory layout (#36704)

2026-08-29 07:41:25 +08:00

qwen4_ple.py

support qwen 3.8 flash next (#37500)

2026-09-08 13:56:21 -07:00