Skip to content

Pull requests: sgl-project/sglang

New pull request New
344 Open 3,925 Closed
344 Open 3,925 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

Fix VLM CI and eliminate incorrect dependency on pixel values.
#6584 opened May 25, 2025 by lifuhuang Loading…
6 tasks
1
feat: Add warnings for invalid tool_choice and UTs
#6582 opened May 25, 2025 by CatherineSue Loading…
2 of 6 tasks
1
refactor qwen moe code, use communicator to support tp+dp high priority
#6581 opened May 25, 2025 by yizhang2077 Loading…
6 tasks
@fzyzcjy @ispobock @sleepcoo
4
[Feature] Add Logit Bias enhancement New feature or request feature
#6579 opened May 25, 2025 by b8zhong Loading…
1
2
chore: upgrade transformers 4.52.3 high priority
#6575 opened May 24, 2025 by zhyncs Loading…
6 tasks
@mickqian
fix: log avg_request_queue_latency metric
#6574 opened May 24, 2025 by jorgeantonio21 Loading…
1
Support custom_logit_processor for OpenAI API
#6572 opened May 24, 2025 by YorkSu Loading…
3 of 6 tasks
1
Support fake perfectly balanced EP dispatch algorithm
#6571 opened May 24, 2025 by fzyzcjy Loading…
6 tasks
Hint users when weight update timeouts
#6570 opened May 24, 2025 by fzyzcjy Loading…
6 tasks
1
add async api for kv-transfer high priority
#6565 opened May 24, 2025 by niqi-lyu Loading…
6 tasks
@whybeyoung @Alcanderian
@ByronHsu @yizhang2077 @hnyls2002 @ShangmingCai
1
[WIP]hybrid kv cache for LLaMA4 high priority
#6563 opened May 24, 2025 by tarinkk Loading…
6 tasks
@CatherineSue @ch-wan
4
[tool_call] fix: added double dumping, valid json check. Fixed structural tags for Qwen25 parser high priority
#6559 opened May 23, 2025 by shuaills Loading…
6 tasks
@CatherineSue @ispobock @zhyncs
7
Fix torch profiler bugs for bench_offline_throughput.py
#6557 opened May 23, 2025 by PanJason Loading…
1 of 6 tasks
@xiezhq-hermann
Tool call fix
#6556 opened May 23, 2025 by chen700564 Loading…
6 tasks
1
[PD] Decouple tokenize process to single process worker
#6555 opened May 23, 2025 by whybeyoung Loading…
1
Replace assignment (=) with comparison (==) in write() check
#6554 opened May 23, 2025 by 44670 Loading…
[FIX] fix wrong indexing for hidden states when prefix-cache take effect
#6551 opened May 23, 2025 by Dutch-voyage Loading…
1 of 6 tasks
[CPU] Bind threads and numa node for each TP rank
#6549 opened May 23, 2025 by chunyuan-w • Draft
1
[New Model] Devstral support
#6547 opened May 23, 2025 by JustinTong0323 Loading…
1 of 6 tasks
4
added support for tied weights in qwen pipeline parallelism
#6546 opened May 23, 2025 by FrankLeeeee Loading…
2 of 6 tasks
refactor apply_w8a8_block_fp8_linear in fp
#6545 opened May 23, 2025 by ChangyiYang Loading…
6 tasks
11
Support profile args in Engine API
#6539 opened May 23, 2025 by fzyzcjy Loading…
6 tasks
1
[sgl-kernel] add rotary embed kernel for trivial head_sizes
#6530 opened May 22, 2025 by mickqian Loading…
6 tasks
1
[PD] Fix prefill_servers in mini_lb
#6527 opened May 22, 2025 by wangxiyu191 Loading…
1 of 6 tasks
1
[WIP] Independent GPTQ
#6519 opened May 22, 2025 by yhyang201 • Draft
6 tasks
ProTip! Find all pull requests that aren't related to any open issues with -linked:issue.