Skip to content

Pull requests: vllm-project/vllm-ascend

New pull request New
279 Open 3,070 Closed
279 Open 3,070 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

[WIP][Feat] Eagle Proposer support FULL_DECODE_ONLY graph mode module:core
#4763 opened Dec 6, 2025 by anon189Ty • Draft
6
Adapt dispatch_ffn_combine for decoding module:core module:ops module:quantization module:tests
#4762 opened Dec 6, 2025 by kiscad Loading…
8
[KVPOOl]Support pp
#4761 opened Dec 6, 2025 by baxingpiaochong Loading…
3
Top9 module:core module:ops
#4760 opened Dec 6, 2025 by raindaywhu Loading…
4
[Perf] Update causal conv1d fn for better perf module:ops
#4759 opened Dec 6, 2025 by SunnyLee151064 Loading…
3
refactor rejection sampler merge-conflicts
#4758 opened Dec 6, 2025 by realliujiaxu • Draft
1 of 3 tasks
5
[Feature] Support multi graphs for torchair
#4757 opened Dec 5, 2025 by jianzs Loading…
3
[Feature] Reduce the cost of torchair ready read for review ready-for-test start test by label for PR
#4756 opened Dec 5, 2025 by jianzs Loading…
1
[CI] Fix ngram & suffix test oom module:tests ready read for review ready-for-test start test by label for PR
#4755 opened Dec 5, 2025 by fluctlux Loading…
4
[Bugfix] Prevent engine hang during KVCacheSendingThread startup ready read for review ready-for-test start test by label for PR
#4754 opened Dec 5, 2025 by jianzs Loading…
3
Support DeepSeekV3.2 with MLAPO operator
#4753 opened Dec 5, 2025 by ZYang6263 • Draft
7
[Bugfix] Disable the dispatch_ffn_combine kernel in MTP path
#4751 opened Dec 5, 2025 by kiscad Loading…
1
[CustomOp][MM] Register AscendMMEncoderAttention CustomOp and remove related patch module:core module:ops
#4750 opened Dec 5, 2025 by shen-shanshan Loading…
4
[Bugfix] Add the check for a null VllmConfig module:core
#4749 opened Dec 5, 2025 by gcanlin Loading…
3
add qwen3_next ops: fused_qkvzba_split_reshape and rope_forward_triton module:ops
#4747 opened Dec 5, 2025 by ZT-AIA Loading…
4
[CI] Fix ut no space merge-conflicts
#4746 opened Dec 5, 2025 by Potabk Loading…
3
[Bugfix] Fix MoE MLP related issues (ref #4490) module:ops module:tests
#4743 opened Dec 5, 2025 by Clorist33 Loading…
3
[Bugfix] Fix eplb device transfer loader issues (ref #4490) module:tests
#4742 opened Dec 5, 2025 by Clorist33 Loading…
6
BugFix: Resolve PolicyFlashlb warm up function attribute error
#4741 opened Dec 5, 2025 by Mercykid-bash Loading…
4
[Misc] Rope optimize merge-conflicts module:core module:ops
#4740 opened Dec 5, 2025 by Angazenn Loading…
5
MLA Prefill performance optimization
#4739 opened Dec 5, 2025 by pichangping Loading…
10
[WIP][perf] replace all_reduce for kv_consumer and support different num_tokens among all ranks merge-conflicts module:ops module:quantization
#4736 opened Dec 5, 2025 by linfeng-yuan Loading…
6
[Fix] fix llava-1.5-7b-hf & Qwen2-Audio-7B-Instruct accuracy test module:tests
#4734 opened Dec 5, 2025 by zhangxinyuehfad Loading…
3
[Misc] Upgrade vllm commit to 12_05 documentation Improvements or additions to documentation ready read for review ready-for-test start test by label for PR
#4733 opened Dec 5, 2025 by Potabk Loading…
5
feat: implement high-performance Triton kernels for rejection sampling
#4732 opened Dec 5, 2025 by yuxingcyx Loading…
14
ProTip! Mix and match filters to narrow down what you’re looking for.