Skip to content

Pull requests: sgl-project/sglang

New pull request New
279 Open 3,503 Closed
279 Open 3,503 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

[Misc] add service discovery for sgl router
#5865 opened Apr 29, 2025 by slin1237 Loading…
4 of 6 tasks
[DRAFT] Test new runner for AMD Nightly.
#5863 opened Apr 29, 2025 by saienduri • Draft
6 tasks
[DRAFT] Adding AMD Nightly.
#5861 opened Apr 29, 2025 by saienduri • Draft
6 tasks
[WIP] Add qwen3 30b fused moe config
#5859 opened Apr 28, 2025 by JustinTong0323 Loading…
1 of 6 tasks
1
[fix]: PyO3 macOS linking and consolidate on tracing for logging
#5856 opened Apr 28, 2025 by slin1237 Loading…
4 of 6 tasks
1
[WIP] Llama 4: Hybrid KV buffer (disable radix attention)
#5853 opened Apr 28, 2025 by tarinkk • Draft
6 tasks
[WIP] Add qwen3-235B-FP8 fused moe config
#5844 opened Apr 28, 2025 by JustinTong0323 • Draft
6 tasks
Tables instead of bulletpoints for sampling doc
#5841 opened Apr 28, 2025 by simveit • Draft
2
[feature] offline compile_deep_gemm for PD and docker build
#5840 opened Apr 28, 2025 by Alcanderian Loading…
2 of 6 tasks
1
Add benchmark script integrating with lmm-evals
#5837 opened Apr 28, 2025 by ravi03071991 Loading…
6 tasks
[Docs] Update docs for Qwen3 and Qwen3MoE
#5836 opened Apr 28, 2025 by adarshxs • Draft
2
[PD] Vectorise group_concurrent_contiguous in NumPy
#5834 opened Apr 28, 2025 by yuan-luo Loading…
6 tasks
2
[PD] Fix Assertion failed: /DeepEP/csrc/kernels/internode.cu:483, condition: ibgda_get_state()->num_rc_per_pe >= num_channels #134
#5830 opened Apr 28, 2025 by whybeyoung Loading…
1
[Feature] Overlap DeepEP Combine Communication and Shared Experts Computation
#5829 opened Apr 28, 2025 by liz-badada Loading…
6 tasks
@ch-wan @liz-badada @xutizhou
1
[PD] Revert transfer thread pool to fix compatibility with other features
#5824 opened Apr 28, 2025 by ShangmingCai Loading…
6 tasks
opt flashinfer mla cat
#5822 opened Apr 28, 2025 by xu-yfei Loading…
2 of 6 tasks
@ispobock
3
chore: update vlm mmmu test accuracy
#5821 opened Apr 28, 2025 by mickqian Loading…
6 tasks
1
cutlass 3.9 supported to improve fp8_blockwise_gemm
#5820 opened Apr 28, 2025 by BBuf Loading…
6 tasks
@elfiegg
1
Enabling DeepGeMM for TP-based MoE
#5819 opened Apr 28, 2025 by ch-wan • Draft
6 tasks
@yinfan98 @sleepcoo @xutizhou
feat: realize Tokenizer API, which is a simple wrapper over HuggingFace-style tokenizers.
#5813 opened Apr 28, 2025 by minleminzui Loading…
6 tasks
feat: integrate deepgemm into EPMoE
#5805 opened Apr 28, 2025 by TianQiLin666666 Loading…
3 of 6 tasks
@ch-wan @sleepcoo @xutizhou
9
fix qwen2.5 vl processor default max pixels
#5803 opened Apr 28, 2025 by Titan-p Loading…
6 tasks
1
misc: hash on multimodal bytes instead of array
#5799 opened Apr 28, 2025 by mickqian • Draft
6 tasks
Auto set draft model path for MTP high priority
#5793 opened Apr 27, 2025 by ispobock Loading…
Remove unused method calculate_num_image_tokens from qwen2_vl.py
#5783 opened Apr 27, 2025 by JustinTong0323 Loading…
1 of 6 tasks
ProTip! no:milestone will show everything without a milestone.