-
Notifications
You must be signed in to change notification settings - Fork 811
Pull requests: vllm-project/vllm-ascend
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
[refactor]Optimized the kvcache usage of Deepseek v3.2
#6610
opened Feb 7, 2026 by
kunpengW-code
Loading…
[Ops][Feature] Use add_rms_norm fusion operator on 310P
#6609
opened Feb 7, 2026 by
csoulndai
Loading…
[main][bugfix] Fix spec acceptance rate problem in vllm_0.15.0
merge-conflicts
module:core
#6606
opened Feb 6, 2026 by
lilinsiman
Loading…
[EPLB][dispatchFFNcombine][Bugfix] Bugfix for dispatchFFNcombine in d…
module:core
module:ops
#6605
opened Feb 6, 2026 by
shenchuxiaofugui
Loading…
[Quantization] Add GPTQ quantization support for Ascend NPU
merge-conflicts
module:core
module:quantization
module:tests
#6603
opened Feb 6, 2026 by
22dimensions
•
Draft
[draft]support cos_sin_cache
ready
read for review
ready-for-test
start test by label for PR
#6602
opened Feb 6, 2026 by
Angazenn
Loading…
[bugfix]Fix no attribute 'data' when MLAPO is enable
merge-conflicts
module:ops
ready
read for review
ready-for-test
start test by label for PR
#6601
opened Feb 6, 2026 by
Meihan-chen
Loading…
[BugFix] Fix actual_seq_lengths_q mismatch in eagle proposer first step
ready
read for review
ready-for-test
start test by label for PR
#6596
opened Feb 6, 2026 by
LICO1314
Loading…
[fix bug] fix tensor mismatch bug in sigmoid operate test case
module:tests
#6595
opened Feb 6, 2026 by
lhp-deep
Loading…
[Ops] Update fused_sigmoid_gating_delta_rule_update_kernel
module:ops
#6592
opened Feb 6, 2026 by
AyiStar
Loading…
[bugfix] Support pipeline parallellism for Deepseek V3.2 DSA-CP
merge-conflicts
module:ops
#6589
opened Feb 6, 2026 by
zzhx1
Loading…
[BugFix] Add support for rotary_dim parameter when using partial rope in rotary_embedding
module:ops
#6581
opened Feb 5, 2026 by
GoCHug
Loading…
[Misc]Wrap whole rotary embedding into a single fake_impl
merge-conflicts
module:core
module:ops
ready
read for review
ready-for-test
start test by label for PR
#6568
opened Feb 5, 2026 by
Angazenn
Loading…
[draft][feat] [Spec Decode] Unified Parallel Drafting
merge-conflicts
#6565
opened Feb 5, 2026 by
HF-001
Loading…
step3p5 migration to npu
merge-conflicts
module:ops
#6546
opened Feb 4, 2026 by
cywang250805
Loading…
[Feat] Add lightling indexer skipping for first 2048 tokens of each request
#6540
opened Feb 4, 2026 by
YzTongNiar
Loading…
[CI] Add long and short prompt tests for DeepSeek-V3.2
module:tests
#6536
opened Feb 4, 2026 by
starmountain1997
Loading…
[MOE Refactor] Remove QuantType in prepare_finalize.py
merge-conflicts
module:ops
module:tests
ready
read for review
ready-for-test
start test by label for PR
#6534
opened Feb 4, 2026 by
shenchuxiaofugui
Loading…
Previous Next
ProTip!
no:milestone will show everything without a milestone.