Skip to content

Pull requests: sgl-project/sglang

New pull request New
327 Open 3,818 Closed
327 Open 3,818 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

fixing qwen3 chat template
#6409 opened May 19, 2025 by leeparkuky Loading…
1 of 6 tasks
Add intel_amx backend for Radix Attention for CPU
#6408 opened May 19, 2025 by yanbing-j • Draft
6 tasks
feat(draft_model): support draft_model for RemoteModelLoader
#6407 opened May 19, 2025 by DellCurry Loading…
6 tasks
Update extend/decode attention kernel for CPU in sgl-kernel and add UTs
#6405 opened May 19, 2025 by yanbing-j • Draft
6 tasks
Add fp8 fused_experts kernel for CPU in sgl-kernel and add UT
#6404 opened May 19, 2025 by chunyuan-w • Draft
Add environment flag for disabling message queue broadcaster
#6403 opened May 19, 2025 by Fridge003 Loading…
6 tasks
nit: fix bench moe fused gate test
#6402 opened May 19, 2025 by yuan-luo Loading…
6 tasks
Update CI tests with Qwen3 models
#6399 opened May 18, 2025 by ravi03071991 Loading…
6 tasks
4
[Misc] Replace log_info_on_rank0 with RankZeroFilter
#6398 opened May 18, 2025 by CatherineSue Loading…
1 of 6 tasks
@Fridge003
5
feat(kernel): fuse silu_and_mul with group quant fp8 high priority
#6394 opened May 18, 2025 by tanruixiang Loading…
4 of 6 tasks
@Alcanderian @BBuf
4
[Fix] [Vul] fix the unsafe usage of shell=True for subprocess.check_output
#6393 opened May 18, 2025 by shaoyuyoung Loading…
1
Add composed attention backend for two batch overlapping
#6390 opened May 18, 2025 by fzyzcjy Loading…
6 tasks
Fix All-gather after DP FFNs
#6389 opened May 18, 2025 by ch-wan Loading…
Support updating expert locations dynamically
#6388 opened May 18, 2025 by fzyzcjy Loading…
6 tasks
Support DeepSeek EPLB algorithm with static distributions
#6387 opened May 18, 2025 by fzyzcjy Loading…
6 tasks
Support loading weights when physical experts are different from logical experts
#6386 opened May 18, 2025 by fzyzcjy Loading…
6 tasks
1
Support dispatching logical to physical experts
#6385 opened May 18, 2025 by fzyzcjy Loading…
6 tasks
1
aiter attention-backend (default enabled on AMD/ROCm) high priority
#6381 opened May 18, 2025 by HaiShaw Loading…
6 tasks done
@HaiShaw @Alcanderian @saienduri
2
Disable compiling arch below sm_90 in aarch64 by default
#6380 opened May 18, 2025 by Qiaolin-Yu • Draft
6 tasks
@zhyncs
2
Implement gather before attn
#6378 opened May 18, 2025 by ch-wan Loading…
6 tasks
@fzyzcjy
Optimize server startup
#6375 opened May 17, 2025 by fzyzcjy Loading…
6 tasks
reduce torch.zeros overhead in moe align block size kernel
#6369 opened May 17, 2025 by BBuf Loading…
2 of 8 tasks
[VLM] Support chunk prefill for VLM high priority visIon-LM
#6355 opened May 16, 2025 by CatherineSue Loading…
2 of 6 tasks
@mickqian @JustinTong0323
4
feat: add sglang to runpod hub
#6350 opened May 16, 2025 by TimPietrusky Loading…
[Feature] Support Qserve high priority
#6349 opened May 16, 2025 by HandH1998 Loading…
6 tasks
@yundai424 @zhyncs
1
ProTip! Add no:assignee to see everything that’s not assigned.