Skip to content

Pull requests: sgl-project/sgl-kernel-npu

New pull request New
99 Open 499 Closed
99 Open 499 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Update
#681 opened Aug 10, 2026 by YZY00Raiser Contributor Loading…
Verify ops optimization.
#680 opened Aug 10, 2026 by hanwlax Loading…
A2 testcases 2tp
#677 opened Aug 10, 2026 by YZY00Raiser Contributor Loading…
[bugfix] Fix sgemmc multi-slice correctness and align LoRA ops with Ascend LoRABackend dtype protocol
#676 opened Aug 10, 2026 by CaiNot Loading…
2 tasks done
1
Delete redundant tests
#675 opened Aug 10, 2026 by YZY00Raiser Contributor Loading…
[feat] (deepep): add A5 normal multi-round ant migration
#673 opened Aug 8, 2026 by WenhaoGe Contributor Loading…
4 tasks done
4
FuseEP supports using the swiglu-oai activation method in DispatchFFNCombine mode
#671 opened Aug 7, 2026 by vstone-w Loading…
[NPU] Add AscendC mHC operators for TeleChat4
#670 opened Aug 7, 2026 by momaekar1 • Draft
add minimax m3 attention ops on A3
#668 opened Aug 7, 2026 by vstone-w Loading…
1
add weights h2d func and mamba pool trans func
#665 opened Aug 7, 2026 by jamesjxliu Loading…
third_party/pto-isa: Update reference to recent state
#663 opened Aug 6, 2026 by ChristosMatzoros Contributor Loading…
[feat] A5 deep fused moe supported
#661 opened Aug 6, 2026 by 666syh Contributor Loading…
4 tasks done
1
[NPU] Add LLaDA2 block routing and dLLM reduction kernels
#659 opened Aug 6, 2026 by leedongkun30-arch • Draft
Validate fused DeepEP MoE quant mode
#657 opened Aug 5, 2026 by WenhaoGe Contributor Loading…
2
[fix]Replace dist.gather with dist.all_gather for hccl backend compatibility
#655 opened Aug 5, 2026 by 1329009851 Contributor Loading…
4 tasks done
1
align speculative Mamba state with sequential decode
#651 opened Aug 4, 2026 by McZyWu Contributor Loading…
1
添加了缓存命中
#650 opened Aug 3, 2026 by huangxiaojun15 Loading…
1
[perf] Parallelize apply_token_bitmask across vocab tiles
#645 opened Aug 3, 2026 by vuuihc • Draft
3 of 8 tasks
1
Add causal_conv1d tests
#643 opened Jul 31, 2026 by Pairman Loading…
1
Fix causal_conv1d and CANN 9.1.0 build for A5
#642 opened Jul 31, 2026 by Pairman Loading…
1
A3 supports mega_moe activation by situ
#639 opened Jul 30, 2026 by 666syh Contributor Loading…
1
✨ [NPU] Add portable Gemma RMSNorm API
#638 opened Jul 30, 2026 by TallMessiWu Loading…
@ping1jing2
8
feat: add NPU primitives for sparsity-driven KV offloading
#636 opened Jul 29, 2026 by nzbej Loading…
5 tasks done
1
WIP: kdn decode pto-isa
#635 opened Jul 28, 2026 by fiskrt Contributor Loading…
2 of 4 tasks
2
[docs] Add submodule update instructions to README
#633 opened Jul 28, 2026 by ITCJ Loading…
1
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.