-
Notifications
You must be signed in to change notification settings - Fork 1.5k
Pull requests: vllm-project/vllm-ascend
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
[CI] to reduce ubuntu_latest, change runner in image build workflow
ci/build
#11398
opened Jul 3, 2026 by
czc-unac
Contributor
Loading…
[Test][Misc] Merge two test cases for batch invariance into one
module:tests
#11397
opened Jul 3, 2026 by
wangx700
Contributor
Loading…
[BugFix][Ops][310p]:fix the accuracy issue caused by MoEGatingTopkSoftmax
module:tests
#11391
opened Jul 3, 2026 by
Tflowers-0129
Collaborator
Loading…
[Feature] Support IndexCache in context-parallel DSA path
#11386
opened Jul 3, 2026 by
GDzhu01
Contributor
Loading…
[Performance][950PR]Split ChunkedPrefill into separate decode and prefill FIA calls
#11385
opened Jul 3, 2026 by
FengDengcai
Loading…
[BugFix][Parser] Avoid partial MiniMax parameter arguments
module:tests
#11384
opened Jul 3, 2026 by
QwertyJack
Contributor
Loading…
[Feature][Prefix Caching][DSv4] Support selective prefix-cache retention #43447
#11383
opened Jul 3, 2026 by
Csrayz
Contributor
Loading…
Mtp vllm23
merge-conflicts
module:core
module:ops
module:tests
#11381
opened Jul 3, 2026 by
swimming2007-doge
Loading…
[BugFix][xlite] Performance improvement in xlite full mode by avoiding redundant memory allocation during profiling runs
#11380
opened Jul 3, 2026 by
SijieFu
Contributor
Loading…
[Test] Verify cpu-32-hk Vault/buildkitd fix
ci/build
#11379
opened Jul 3, 2026 by
JavaPythonAIForBAT
Loading…
3 tasks
[BugFix][Proxy] Refine retry logic in load balance proxy and preserve upstream error status code
#11378
opened Jul 3, 2026 by
xingzhang8023
Loading…
[BugFix]fix vllm use v2 but vllm-ascend use v1
merge-conflicts
module:core
#11377
opened Jul 3, 2026 by
wangx700
Contributor
Loading…
[Feature] Add DFX for PD Disaggregated
merge-conflicts
module:core
#11376
opened Jul 3, 2026 by
zzzzzmeng
Contributor
Loading…
[Performance][310P] Access FA approximation calculation
#11375
opened Jul 3, 2026 by
YangShuai52
Contributor
Loading…
[Doc][Misc] Update Qwen3.5-27B and Qwen3.6-27B documentation
documentation
Improvements or additions to documentation
#11372
opened Jul 3, 2026 by
AJF-cmd
Contributor
Loading…
Fix W4A8 dense bias handling
module:quantization
module:tests
#11371
opened Jul 3, 2026 by
gdgfd22
Loading…
[BugFix] Fix AscendStore eagle store mask propagation
module:tests
#11370
opened Jul 3, 2026 by
Pz1116
Collaborator
Loading…
[Doc][Misc] Update Qwen3.5-27B and Qwen3.6-27B documentation
documentation
Improvements or additions to documentation
#11369
opened Jul 3, 2026 by
AJF-cmd
Contributor
Loading…
[Feature] Add e5-mistral embedding model test
documentation
Improvements or additions to documentation
module:tests
#11364
opened Jul 2, 2026 by
EheinWang
Loading…
[Feature] Enable 4-stage multistream overlap for W4A4_MXFP4 shared experts
module:ops
module:quantization
module:tests
#11355
opened Jul 2, 2026 by
zhenwenqi2024
Collaborator
Loading…
[Test][Feature] Add comprehensive unit tests for Ascend KV transfer and store connector components
module:tests
#11354
opened Jul 2, 2026 by
Mango03111
Loading…
[BugFix] Avoid 310P Mamba align postprocess hang for MTP
module:tests
#11353
opened Jul 2, 2026 by
Alex-stack-hub
Contributor
Loading…
Previous Next
ProTip!
no:milestone will show everything without a milestone.