Skip to content

Pull requests: sgl-project/sglang

New pull request New
178 Open 2,829 Closed
178 Open 2,829 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

Enable radix attention for qwen2 vl models
#4702 opened Mar 24, 2025 by yupbank Loading…
6 tasks
Added async_encode method to Engine
#4701 opened Mar 23, 2025 by shimizust Loading…
1 of 6 tasks
Support fine-grained control of requests that are run together
#4699 opened Mar 23, 2025 by fzyzcjy Loading…
6 tasks
Unify variable naming: replace is_in_free_group with is_not_in_free_group
#4698 opened Mar 23, 2025 by c1lovez1 Loading…
2 of 5 tasks
2
[Feature] use pytest for sgl-kernel
#4697 opened Mar 23, 2025 by adarshxs Loading…
2 tasks done
2
Speedup warmup when DP > 1
#4695 opened Mar 23, 2025 by fzyzcjy Loading…
6 tasks
2
Update MMMU Benchmark instructions
#4694 opened Mar 23, 2025 by ravi03071991 Loading…
6 tasks
[Model] Adding Qwen3 and Qwen3MoE
#4693 opened Mar 23, 2025 by yhyang201 Loading…
3
fix FlashMLA cudagraph config
#4691 opened Mar 23, 2025 by sleepcoo Loading…
Support controlling nsys start and end range programmatically
#4688 opened Mar 23, 2025 by fzyzcjy Loading…
6 tasks
Fix torch.cuda.MemPool() internal assertion failure
#4687 opened Mar 23, 2025 by fzyzcjy Loading…
6 tasks
[WIP] fix loading KV quantization scale; Enable modelopt kv cache high priority
#4686 opened Mar 23, 2025 by yundai424 • Draft
6 tasks
refactor: Veteran-level full-stack improvements to code quality
#4683 opened Mar 22, 2025 by ikhanganin Loading…
fix: Inappropriate lack of Optional type on OpenAI ChatCompletionRequest
#4681 opened Mar 22, 2025 by BroadbentJim Loading…
1 of 6 tasks
[WIP] Support FA3 as Attention backend high priority performance
#4680 opened Mar 22, 2025 by hebiao064 Loading…
5 of 13 tasks
@Ying1123 @merrymercy
@ispobock @zhyncs @Fridge003
24
check marlin format before attempting conversion
#4675 opened Mar 22, 2025 by qeternity Loading…
bump v0.4.4.post2 high priority
#4669 opened Mar 22, 2025 by zhyncs Loading…
6 tasks
1
Support profiling in bench_one_batch_server.py
#4667 opened Mar 22, 2025 by fzyzcjy Loading…
6 tasks
Allow benchmarking each forward pass in e2e systems
#4666 opened Mar 22, 2025 by fzyzcjy Loading…
6 tasks
fixed function call parser for llama 32
#4656 opened Mar 21, 2025 by kyle-pena-kuzco Loading…
3 of 6 tasks
3
Preliminary support for the CuDNN attention backend
#4650 opened Mar 21, 2025 by yhyang201 • Draft
Fix Engine error when enabling DP attention
#4648 opened Mar 21, 2025 by fzyzcjy Loading…
6 tasks
Tok boosting sampler -- TODO DRAFT PR
#4647 opened Mar 21, 2025 by tginart • Draft
6 tasks
Remove Unintended Capture Batch Sizes in AMD HIP Graph Runner
#4638 opened Mar 20, 2025 by gmlwns2000 Loading…
6 tasks done
4
fix save_sharded_state error with --max-file-size
#4634 opened Mar 20, 2025 by AllenXu93 Loading…
6 tasks
1
ProTip! What’s not been updated in a month: updated:<2025-02-23.