Skip to content

Pull requests: sgl-project/sglang

New pull request New
290 Open 3,578 Closed
290 Open 3,578 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
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

[CI] add 4-gpu test & update verlengine ci
#6007 opened May 4, 2025 by ocss884 Loading…
6 tasks
chore: upgrade cutlass 3.9.2 wip
#6004 opened May 4, 2025 by zhyncs Loading…
6 tasks
@yizhang2077
[WIP] zero-copy gpu-tensor in Req
#6003 opened May 4, 2025 by mickqian • Draft
6 tasks
1
[PD] Allow customizing reserved tokens to avoid KV cache waste
#6002 opened May 4, 2025 by fzyzcjy Loading…
6 tasks
fix: fix broadcast_pyobj breaking VerlEngine
#5997 opened May 3, 2025 by ocss884 Loading…
6 tasks
1
@Ying1123
feat: append more comprehensive fields in messages instead of merely role and content
#5996 opened May 3, 2025 by minleminzui Loading…
6 tasks
[refactor] move some rocm codes to stand-alone folder(1/N) high priority
#5993 opened May 3, 2025 by Alcanderian Loading…
6 tasks
@HaiShaw @saienduri
[Fix] Suppress dynamo logging when using flashinfer backend with torch compile
#5992 opened May 3, 2025 by Fridge003 Loading…
6 tasks
Add H20 TP16 profiled config files for fused_moe_triton
#5990 opened May 3, 2025 by guoyuhong Loading…
1 of 6 tasks
1
Support reduce-scatter when dp=tp
#5988 opened May 2, 2025 by EstherBear Loading…
1 of 6 tasks
@ch-wan
1
[Feat] Enable PDL automatically on Hopper architecture
#5981 opened May 2, 2025 by PopSoda2002 Loading…
1 of 6 tasks
4
support max-waiting-requests
#5976 opened May 2, 2025 by Bruce-x-1997 Loading…
6 tasks
1
vlm: speed up gpu-tensor hash
#5974 opened May 2, 2025 by mickqian Loading…
6 tasks
Follow-up to #5485 – Fix import errors in Glm4ForCausalLM integration
#5971 opened May 2, 2025 by Muqi1029 Loading…
2 of 6 tasks
fix: conv template matching func misnamed
#5970 opened May 2, 2025 by KivenChen Loading…
6 tasks done
Add endpoint /abort
#5966 opened May 2, 2025 by zhuzilin Loading…
2 of 6 tasks
1
2
[OAI] Support non-normalized logprobs in OpenAI server
#5961 opened May 2, 2025 by CatherineSue Loading…
1 of 6 tasks
@merrymercy @hnyls2002
2
fix: small typos and grammar in error messages
#5959 opened May 1, 2025 by davidxia Loading…
2 of 6 tasks
Fix flaky issues of lora and add multi batch tests ready-to-merge The PR is ready to merge after the CI is green.
#5957 opened May 1, 2025 by Qiaolin-Yu Loading…
6 tasks
6
Add triton version as a fused_moe_triton config search key to avoid performace decrease in different Triton version
#5955 opened May 1, 2025 by BBuf • Draft
6 tasks
[Feat][WIP] QWen-1M context support[2/2]: Update block sparse attention backend
#5949 opened May 1, 2025 by yinfan98 • Draft
4 of 6 tasks
[Misc] Add comment for sgl-kernel python interface
#5947 opened May 1, 2025 by sleepcoo Loading…
6 tasks
Online Rotations in SGLang
#5943 opened May 1, 2025 by amd-rcorzine Loading…
4 of 6 tasks
11
Add Cohere Aya Vision support
#5941 opened May 1, 2025 by vincentzed • Draft
6 tasks done
Updated the use of Type | None to Optional[Type] for support [email protected]
#5935 opened May 1, 2025 by vincentzed Loading…
6 tasks
ProTip! no:milestone will show everything without a milestone.