Skip to content

Pull requests: vllm-project/vllm-ascend

New pull request New
121 Open 1,116 Closed
121 Open 1,116 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

[BugFix]fixed all_reduce_merge_allgather_ep bug module:core module:ops
#1818 opened Jul 15, 2025 by ttanzhiqiang Loading…
1
[BugFix]fixed rm_router_logits_allgather_ep bug module:core module:ops
#1817 opened Jul 15, 2025 by ttanzhiqiang Loading…
1
[Bugfix]Fix the performance gap between 0.9.2rc1 and 0.9.1
#1811 opened Jul 15, 2025 by lianyiibo Loading…
1
3
[Misc][V0 Deprecation] Remove draft model runner used for V0 spec decode merge-conflicts
#1810 opened Jul 15, 2025 by shen-shanshan Loading…
4
[0.9.1] Fix wheel glibc version incompatibility ci/build
#1808 opened Jul 15, 2025 by wxsIcey Loading…
1
[0.9.1][bugfix] fix disaggregate prefill hang issue in long output sc…
#1807 opened Jul 15, 2025 by liziyu179 Loading…
[main] Use AddRmsNormQuant ops in the custom model to optimize Qwen3's performance module:ops module:quantization
#1806 opened Jul 15, 2025 by rjg-lyh Loading…
1
[Perf][v0.9.1] Add TP2 for deepseek mla o_proj in pure DP/EP scenario for better performance. module:core
#1804 opened Jul 15, 2025 by whx-sjtu Loading…
[Bugfix] Fix num_hidden_layers when Qwen2-Audio 7B documentation Improvements or additions to documentation module:core
#1803 opened Jul 15, 2025 by zhangxinyuehfad Loading…
3
[Prefill Perf] Parallel Strategy Optimizations (VRAM-for-Speed Tradeoff) module:core module:ops module:quantization
#1802 opened Jul 15, 2025 by SlightwindSec Loading…
[V1][PP] Support pp with ray backend in V1
#1800 opened Jul 15, 2025 by MengqingCao • Draft
1
[Platform] Add support for Altlas A3 series ci/build module:core
#1794 opened Jul 15, 2025 by wxsIcey Loading…
1
5
Move deepseek_v3 from deepseek_v2.py
#1793 opened Jul 14, 2025 by xleoken Loading…
7
[CI] Switching to infra cache server to reduce network pressure
#1792 opened Jul 14, 2025 by pkking Loading…
4
add ut for device communitor file module:tests
#1789 opened Jul 14, 2025 by Agonixiaoxiao Loading…
Add graph mode for Qwen2.5 and Qwen3 module:core module:ops
#1787 opened Jul 14, 2025 by NicholasTao Loading…
11
[0.9.1]optmize rope in qwen2 merge-conflicts module:core module:ops module:tests
#1782 opened Jul 14, 2025 by David9857 Loading…
3
[CI] Speed up CI by switch to more stable registry
#1775 opened Jul 14, 2025 by pkking Loading…
1
chunkprefill mla support torchair graph
#1772 opened Jul 14, 2025 by haojiangzheng Loading…
1
add patch ut module:tests
#1766 opened Jul 14, 2025 by Pr0Wh1teGivee Loading…
[PD Disagg][CI] Upgrade vllm version to fix ci pd-test enable pd test for PR ready-for-test start test by label for PR
#1765 opened Jul 14, 2025 by MengqingCao Loading…
【main】 Support SP for qwen2.5 and qwen3 moe module:core module:ops module:tests
#1761 opened Jul 12, 2025 by lbk-sys Loading…
[V0.9.1] torchair_graph bugfix when chunked_prefill is true
#1748 opened Jul 11, 2025 by fems14 Loading…
2
[V0.9.1] Add support for flashcomm_v1 in Qwen2.5 module:core
#1745 opened Jul 11, 2025 by rjg-lyh Loading…
3
flashcomm3 multi stream of moe layer merge-conflicts module:core module:ops module:quantization
#1744 opened Jul 11, 2025 by wyhhyw123 Loading…
1
ProTip! Follow long discussions with comments:>50.