Skip to content

Pull requests: sgl-project/sglang

New pull request New
533 Open 5,160 Closed
533 Open 5,160 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

[ROCm] Use triton vision attention to replace FA3 on AMD GPU
#8656 opened Aug 1, 2025 by zhangnju Loading…
6 tasks
1
fix(cache): Respect sliding_window_size in SWAChunkCache eviction
#8655 opened Aug 1, 2025 by ppraneth Loading…
6 tasks
1
1
Support page first layout zero copy for mooncake store
#8651 opened Aug 1, 2025 by huangtingwei9988 Loading…
7 tasks
1
Support limiting max loaded loras in CPU.
#8650 opened Aug 1, 2025 by lifuhuang Loading…
6 tasks
1
Fix FP8 block quantization when N or K is not multiples of 128
#8648 opened Aug 1, 2025 by yanbing-j • Draft
6 tasks
1
disable tp for shared experts when enable ep moe for GLM4.5 model
#8647 opened Aug 1, 2025 by zminglei Loading…
6 tasks
6
Increase tolerance to address CI failures
#8643 opened Aug 1, 2025 by lifuhuang Loading…
6 tasks
@ispobock @ch-wan
1
fix compile_deep_gemm.py warmup
#8641 opened Aug 1, 2025 by Alcanderian Loading…
6 tasks
4
[router] introduce dp worker abstraction
#8639 opened Jul 31, 2025 by slin1237 Loading…
2 of 6 tasks
10
TRTLLM-MLA FP8 path
#8638 opened Jul 31, 2025 by farazkh80 • Draft
6 tasks
7
[router] Add Bearer Token Authentication API Support router
#8637 opened Jul 31, 2025 by slin1237 Loading…
3 of 6 tasks
6
Do layernorm before allgather for DP attention
#8631 opened Jul 31, 2025 by trevor-m Loading…
6 tasks
5
[PD metrics] Fix some uncompleted PD related metrics
#8627 opened Jul 31, 2025 by acelyc111 Loading…
1 of 6 tasks
6
[WIP] Optimizing vision encoder memory usage for PD and EPD modes
#8625 opened Jul 31, 2025 by ccw1996 • Draft
6 tasks
11
Use Tensor Core Decode when gqa group size >= 4
#8624 opened Jul 31, 2025 by Edenzzzz Loading…
6 tasks
3
[bugfix]: use correct cache location for cross attention in torch native backend
#8622 opened Jul 31, 2025 by MahmoudAshraf97 Loading…
1 of 6 tasks
4
[Quantization] Supported w8a8 int8 quantized Gemma3 and Qwen-VL models
#8619 opened Jul 31, 2025 by ichernob Loading…
1 of 6 tasks
5
[bugfix] Add 'disaggregation_mode' parameter to warmup function when compile deep_gemm manually
#8618 opened Jul 31, 2025 by lbh2001 Loading…
6 tasks
1
@Alcanderian
5
Support MHA with chunked prefix cache for flashinfer backend
#8616 opened Jul 31, 2025 by xu-yfei Loading…
1 of 6 tasks
@Fridge003
5
[feat] Support eagle3 for qwen2
#8612 opened Jul 31, 2025 by ChenQiaoling00 Loading…
6 tasks done
7
[feat] Enable Ascend profiling on SGLang
#8610 opened Jul 31, 2025 by ping1jing2 Loading…
6 tasks
6
[Feature] Support BurstGPT for server benchmark.
#8605 opened Jul 31, 2025 by VincentXWD Loading…
6 tasks
6
Adapt Eagle3 to Deepseek
#8602 opened Jul 31, 2025 by xuhaojie-2025 Loading…
6 tasks
4
Fix qwen2 audio not working bug
#8600 opened Jul 31, 2025 by byjiang1996 Loading…
6 tasks done
@JustinTong0323
7
[bugfix] Remove the invalid initialization of weight_scale and input_scale for W4AFp8 EPMoE
#8597 opened Jul 31, 2025 by huangzhilin-hzl Loading…
6 tasks
1
2
ProTip! Add no:assignee to see everything that’s not assigned.