Skip to content

Pull requests: zejunchen-zejun/sglang

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
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort

Pull requests list

diasble flash infer rope
#189 opened Feb 6, 2026 by LiuYinfeng01 Loading…
4 tasks
[Wan] Add torch.compile for vae decode and use CL for conv
#184 opened Jan 28, 2026 by zhuyuhua-v Loading…
4 tasks
work for torch compile for aiter rmsnorm
#176 opened Jan 23, 2026 by XiaobingSuper Loading…
4 tasks
enable all2all overlap, and use rope overlap last v gemm
#173 opened Jan 22, 2026 by ganyi1996ppo Loading…
4 tasks
add offline generate lora qwen-image-edit script
#157 opened Jan 13, 2026 by zhuyuhua-v Loading…
4 tasks
Cuda Graph Capture WA for HIP Runtime
#147 opened Jan 9, 2026 by sammysun0711 Loading…
4 tasks
[Feat] add ttft measure for qwen3vl
#128 opened Dec 31, 2025 by ZLkanyo009 Loading…
4 tasks
[feat] Add ROCm ATOM model impl backend
#119 opened Dec 26, 2025 by zejunchen-zejun Loading…
Add tuned triton MOE config for Qwen3-Omni
#105 opened Dec 19, 2025 by sammysun0711 Loading…
4 tasks
Qwen3 next -- fixed conv update split q/k/v in decode phase
#87 opened Dec 10, 2025 by IzacharyI Loading…
4 tasks
Qwen3 next -- fixed sigmoid and mul broadcast issue
#86 opened Dec 10, 2025 by IzacharyI Loading…
6 tasks
[CI] Enable Qwen3-Omni Performance Benchmark
#85 opened Dec 10, 2025 by sammysun0711 Loading…
4 tasks
Increase _AITER_PARTITION_SIZE_ROCM
#84 opened Dec 10, 2025 by apinge Draft
4 tasks
CI: Debug Qwen3 Next issue
#48 opened Dec 2, 2025 by gyohuangxin Draft
ProTip! Add no:assignee to see everything that’s not assigned.