Skip to content

Pull requests: sgl-project/sglang

New pull request New
2,628 Open 16,088 Closed
2,628 Open 16,088 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

[Spec] Relay next_draft_input via GenerationBatchResult; unify scheduler install
#24862 opened May 9, 2026 by hnyls2002 Collaborator Loading…
1 task
1
Refactor device cache emptying run-ci
#24861 opened May 9, 2026 by hebiao064 Collaborator Loading…
@hebiao064 @ByronHsu
2
[Spec V2] Migrate V2 path to EagleDraftExtendInput; verify returns EagleVerifyOutput
#24860 opened May 9, 2026 by hnyls2002 Collaborator Loading…
2 tasks
1
[Spec V1] Introduce EagleDraftExtendInput; split draft-extend phase from EagleDraftInput bypass-fastfail high priority run-ci
#24859 opened May 9, 2026 by hnyls2002 Collaborator Loading…
2 tasks
1
multi_layer_eagle: add tracing hooks
#24858 opened May 9, 2026 by hnyls2002 Collaborator Loading…
1 task
1
Optimize SWA memory preallocation for disaggregated decode
#24857 opened May 9, 2026 by yhyang201 Collaborator • Draft
4 tasks
1
Fix TRTLLM MHA routing for draft extend blackwell SM100/SM120
#24856 opened May 9, 2026 by YAMY1234 Collaborator • Draft
1
[Model] Add MiniCPM-V 4.6 support run-ci
#24855 opened May 9, 2026 by AgainstEntropy Collaborator Loading…
3 of 5 tasks
4
[RL] Call torch.cuda.empty_cache() for in-place pause mode to avoid OOM
#24854 opened May 9, 2026 by ByronHsu Collaborator Loading…
2 tasks done
5
feat(multimodal): support per-request min_pixels/max_pixels for Qwen-VL
#24852 opened May 9, 2026 by CrepuscularIRIS Loading…
1
2
[Session R3] Add routed_experts_start_len for absolute routing slice control
#24851 opened May 9, 2026 by ByronHsu Collaborator Loading…
6 tasks
2
[MoE] Fix NaN in flashinfer TRT-LLM A2A dispatch by sanitizing padding slots run-ci
#24850 opened May 9, 2026 by minosfuture Contributor Loading…
3 of 5 tasks
2
fix(server-args): add validation to prevent trtllm_mla backend on unsupported SM120 GPUs
#24848 opened May 9, 2026 by KhanCold Loading…
1
2
fix(vit-cuda-graph): add fa4 support for B200 GPUs
#24847 opened May 9, 2026 by KhanCold Loading…
1
1
fix(mooncake): auto-detect EFA devices and use efa transport instead of hardcoded rdma
#24846 opened May 9, 2026 by KhanCold Loading…
1
3
fix(scheduler): add error handling for think_end_token encoding in reasoning parser
#24844 opened May 9, 2026 by KhanCold Loading…
1
1
Fix awq offload
#24836 opened May 9, 2026 by THINKER-ONLY Loading…
5 tasks done
1
4
[NPU] fix npu profiler
#24835 opened May 9, 2026 by zhaozx-cn Contributor Loading…
5 tasks done
1
[diffusion] docs: Add example Python Open AI Image edit API documentation Improvements or additions to documentation
#24831 opened May 9, 2026 by maxdswain Loading…
3 tasks done
1
Fix V2 trace filename collisions when DP/PP/EP enabled
#24830 opened May 9, 2026 by fzyzcjy Collaborator Loading…
2
Inject ParallelState into ProfilerV2
#24828 opened May 9, 2026 by fzyzcjy Collaborator Loading…
1
Bundle Scheduler rank/size fields into a frozen ParallelState
#24827 opened May 9, 2026 by fzyzcjy Collaborator Loading…
1
[spec decoding] support kimi-k2.5-eagle3-mla run-ci
#24826 opened May 9, 2026 by Qiaolin-Yu Collaborator Loading…
5 tasks
2
[AMD] DSv4 nightly hotfix + schedule-aware --continue-on-error in AMD CI amd deepseek
#24825 opened May 9, 2026 by bingxche Collaborator • Draft
2 of 5 tasks
2
Add mechanical-refactor-verify skill
#24824 opened May 9, 2026 by fzyzcjy Collaborator Loading…
1
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.