-
Notifications
You must be signed in to change notification settings - Fork 544
Pull requests: vllm-project/vllm-ascend
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
[P/D]Add a heartbeat mechanism to PD separation
#4071
opened Nov 8, 2025 by
wangxiaoteng888
Loading…
【Draft】support triton chunk_gated_delta_rule ops
module:ops
#4070
opened Nov 8, 2025 by
shiyuan680
Loading…
[P/D][BugFix][v0.11.0-dev]Fix proxy format processing errors & Layerwise connector performance optimization
module:tests
#4069
opened Nov 8, 2025 by
wangxiaoteng888
Loading…
[Perf]enable prefill flashcommon3
module:core
module:ops
module:quantization
#4065
opened Nov 7, 2025 by
AlvisGong
Loading…
[Feat] shared expert dp for deepseek and deepseek_mtp for v0.11.0dev
module:core
module:ops
module:tests
#4064
opened Nov 7, 2025 by
dragondream-chen
Loading…
[WIP][Perf] Remove D2H operations to imporve performance
#4063
opened Nov 7, 2025 by
yiz-liu
Loading…
[CI] Integrate mooncake to vllm-ascend base image
module:tools
ready
read for review
ready-for-test
start test by label for PR
#4062
opened Nov 7, 2025 by
Potabk
Loading…
improve installation & usage instructions for ops‑mathImprove installation and usage instructions for ops-math
#4061
opened Nov 7, 2025 by
Luoyukeji
Loading…
[info][v0.11.0] Correct mistakes in source doc
documentation
Improvements or additions to documentation
module:core
#4059
opened Nov 7, 2025 by
lilinsiman
Loading…
[BugFix] Fixes Qwen3-Next enable nz accuracy problem
ready
read for review
ready-for-test
start test by label for PR
#4058
opened Nov 7, 2025 by
wxsIcey
Loading…
[0.11.0] Fixes Qwen3-Next enable nz accuracy problem
ready
read for review
ready-for-test
start test by label for PR
#4056
opened Nov 7, 2025 by
wxsIcey
Loading…
[WIP] mooncake connector support pipeline parallel & fix pp with flashcomm1
merge-conflicts
#4054
opened Nov 7, 2025 by
lidenghui1110
Loading…
[refactor]support gatingtopk operator generalization
module:core
module:ops
#4050
opened Nov 7, 2025 by
1092626063
Loading…
[BugFix]This PR aims to fix the precision issue of the LoRA feature i…
#4046
opened Nov 7, 2025 by
liuchenbing
Loading…
[Test] Add tests for the multi-node DeepSeek-V2-Lite network in GE Graph
module:tests
#4039
opened Nov 6, 2025 by
ForBetterCodeNine
Loading…
[Quantization] Support compressed tensors w8a8 static and w8a8 dynamic weight
module:core
module:quantization
#4036
opened Nov 6, 2025 by
LHXuuu
Loading…
[bugfix] Fixed the bug in retrieving the quantization method for mlp.experts (e.g., DeepSeek_v3.2_exp w8a8)
module:quantization
#4035
opened Nov 6, 2025 by
yangqinghao-cmss
Loading…
[Doc] add qwen3 w4a4 tutorial
documentation
Improvements or additions to documentation
#4034
opened Nov 6, 2025 by
22dimensions
Loading…
[bugfix] fix rmsnorm redundant memory usage
module:ops
#4029
opened Nov 6, 2025 by
hwhaokun
Loading…
[Doc] Remove extra MLAPO installation step for DeepSeek-V3.2.
documentation
Improvements or additions to documentation
#4024
opened Nov 6, 2025 by
menogrey
Loading…
[0.11.0][BugFix] Improve the performance of prefixcache features
module:tests
ready
read for review
ready-for-test
start test by label for PR
#4021
opened Nov 6, 2025 by
underfituu
Loading…
[0.11.0][Fix] Fix Qwen2-Audio-7B-Instruct accuracy test
module:tests
#4018
opened Nov 6, 2025 by
zhangxinyuehfad
Loading…
Previous Next
ProTip!
no:milestone will show everything without a milestone.