Skip to content

Pull requests: sgl-project/sglang

New pull request New
387 Open 4,550 Closed
387 Open 4,550 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

npu support deepseek moe
#7564 opened Jun 26, 2025 by wangqian108 Loading…
6 tasks
10
[opt] replace extend_prefix_lens with extend_prefix_lens_cpu to avoid cudaStreamSync.
#7563 opened Jun 26, 2025 by zzt93 Loading…
2 of 6 tasks
2
Fix incomplete tool call capture issue in streaming response of DeepSeek-V3 when enable MTP
#7562 opened Jun 26, 2025 by xianzhiT Loading…
1 of 6 tasks
4
[WIP] feat: add dockerfile for gb200
#7556 opened Jun 26, 2025 by ishandhanani Loading…
7
Fix flakiness in LoRA batch test.
#7552 opened Jun 26, 2025 by lifuhuang Loading…
6 tasks
@Fridge003 @Qiaolin-Yu
7
[CI] Add CI Testing for Prefill-Decode Disaggregation with Router
#7540 opened Jun 25, 2025 by key4ng Loading…
1 of 6 tasks
12
[AMD] Add unit-test-sgl-kernel-amd to AMD CI
#7539 opened Jun 25, 2025 by hubertlu-tw Loading…
2 of 6 tasks
6
fix raise info in communicator get_fn
#7529 opened Jun 25, 2025 by JamesLim-sy Loading…
3
P/D load balancer forwards profiling requests to instances
#7525 opened Jun 25, 2025 by gronsti-amd Loading…
1 of 6 tasks
7
[CPU] add c++ kernel to bind CPU cores and memory node
#7524 opened Jun 25, 2025 by chunyuan-w Loading…
7
test: improve test for omni models
#7519 opened Jun 25, 2025 by mickqian Loading…
6 tasks
1
fix: incorrect dtype when load_model
#7517 opened Jun 25, 2025 by yudian0504 Loading…
6 tasks
2
load draft model fix
#7506 opened Jun 24, 2025 by yilian49 Loading…
@ch-wan
3
[CPU] remove process_group from inputs of shm_allreduce and shm_allgather cpu cpu backend performance optimization intel sgl-kernel
#7486 opened Jun 24, 2025 by chunyuan-w Loading…
4
[AMD] Remove vllm's scaled_fp8_quant and moe_sum when SGLANG_USE_AITER=1 high priority
#7484 opened Jun 23, 2025 by hubertlu-tw Loading…
3 of 6 tasks
@HaiShaw @Alcanderian
6
[Feature] dynamic server payload size limit
#7475 opened Jun 23, 2025 by khan-yin Loading…
4 of 6 tasks
16
fix(bench_serving): handle None tokenizer.bos_token when apply_chat_template==True
#7466 opened Jun 23, 2025 by renne444 Loading…
1 of 6 tasks
4
[BugFix] Destroy nccl Comm to fix cuda memory leak of destroy_model_parallel
#7465 opened Jun 23, 2025 by wcsjtu Loading…
2 of 6 tasks
3
Support non-contiguous query input for extend/decode attention cpu cpu backend performance optimization intel sgl-kernel
#7462 opened Jun 23, 2025 by yanbing-j Loading…
6 tasks
17
[benchmark] print final benchmark args in json format
#7455 opened Jun 23, 2025 by staugust Loading…
1 of 6 tasks
2
Fix for fp8 quantization failure of qwen 2.5 VL 7B model. high priority
#7448 opened Jun 22, 2025 by PanJason Loading…
2 of 6 tasks
@Alcanderian
5
Support dynamic LoRA loading / unloading in engine/server API ready-to-merge The PR is ready to merge after the CI is green.
#7446 opened Jun 22, 2025 by lifuhuang Loading…
2 of 6 tasks
22
fix: fix apply_shuffle_mul_sum
#7444 opened Jun 22, 2025 by mickqian Loading…
6 tasks
@Alcanderian @ch-wan
6
fix: minor fix cutlass moe
#7442 opened Jun 22, 2025 by mickqian Loading…
6 tasks
5
OPTForCasualLM Support (facebook/opt Series) new-model
#7440 opened Jun 22, 2025 by b8zhong Loading…
6
ProTip! Adding no:label will show everything without a label.