Skip to content

Pull requests: vllm-project/vllm-ascend

New pull request New
150 Open 1,981 Closed
150 Open 1,981 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

[Misc] Refactor platform module:core
#3069 opened Sep 20, 2025 by wangxiyuan Loading…
3
[bugfix][torchair] fix wasted NPU memory buffer allocation for quantized deepseek with unquantized MTP layer ready read for review ready-for-test start test by label for PR
#3068 opened Sep 20, 2025 by linfeng-yuan Loading…
4
[CI] Upgrade vLLM to 20250920 (c60e613) and address config break module:tests ready read for review ready-for-test start test by label for PR vllm-break
#3067 opened Sep 20, 2025 by Yikun • Draft
3
[bugfix][torchair] fix kv_nz accuracy problem and remove redundant reshape_and_cache operation ready read for review ready-for-test start test by label for PR
#3066 opened Sep 20, 2025 by linfeng-yuan Loading…
4
【long_seq_optim】update main ci/build documentation Improvements or additions to documentation module:core module:ops module:quantization module:tests
#3065 opened Sep 20, 2025 by LookAround0301 Loading…
2
Refactor ci to reuse base workflow ready read for review ready-for-test start test by label for PR
#3064 opened Sep 20, 2025 by Yikun • Draft
1
Bump main
#3063 opened Sep 20, 2025 by Yikun • Draft
2
[Feature] qwen3 qk norm support multi_stream. module:core
#3060 opened Sep 20, 2025 by weijinqian0 Loading…
3
[Feature] qwen3_moe qk norm support multi_stream.
#3059 opened Sep 20, 2025 by weijinqian0 Loading…
3
flashcomm1 for shared fused moe module:core module:ops
#3058 opened Sep 20, 2025 by zhaozx-cn Loading…
4
Test on guide decoding merge-conflicts module:core module:ops module:quantization module:tests ready-for-test start test by label for PR
#3057 opened Sep 20, 2025 by Yikun • Draft
3
[Feature] qwen3_moe qk norm support multi_stream.
#3056 opened Sep 20, 2025 by weijinqian0 Loading…
3
fix e2e test module:tests
#3054 opened Sep 20, 2025 by loukong33 Loading…
3
Mmkwarg test merge-conflicts module:core module:ops module:quantization module:tests ready-for-test start test by label for PR
#3050 opened Sep 20, 2025 by MengqingCao • Draft
3
[Feature] Reduce host memory usage for attention mask generation ready read for review ready-for-test start test by label for PR
#3048 opened Sep 20, 2025 by jianzs Loading…
3
[Bugfix][LoRA] Fix LoRA bug after supporting Qwen3-Next ready read for review ready-for-test start test by label for PR
#3044 opened Sep 19, 2025 by paulyu12 Loading…
5
FlashLB algorithm
#3042 opened Sep 19, 2025 by Mercykid-bash Loading…
9
Add support for qwen3_vl and qwen3_vl_moe
#3040 opened Sep 19, 2025 by booker123456 Loading…
7
Support W8A8_dynamic on Step3 Model
#3038 opened Sep 19, 2025 by yifeilin1015 Loading…
4
[v0.9.1] Supprot W8A8_dynamic on Step3 Model
#3036 opened Sep 19, 2025 by yifeilin1015 Loading…
3
[EPLB] ut for EPLB module:tests
#3035 opened Sep 19, 2025 by Clorist33 Loading…
5
[Structured Output][CI] Update structured output config to sync with upstream merge-conflicts module:core module:tests
#3032 opened Sep 19, 2025 by shen-shanshan Loading…
3
splitting MTP into graph mode and non-graph mode merge-conflicts
#3030 opened Sep 19, 2025 by weisirui-eng Loading…
7
[Bugfix] Eliminate the redundant handling of index_select in get_splitfuse_att…
#3028 opened Sep 19, 2025 by tt545571022 Loading…
4
fix: explicitly setting the tensor shape of otp output to fix shape i… module:ops
#3027 opened Sep 19, 2025 by zzhx1 Loading…
3
ProTip! Exclude everything labeled bug with -label:bug.