Skip to content

Pull requests: vllm-project/vllm-ascend

New pull request New
60 Open 557 Closed
60 Open 557 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

[perf] Improve Prefill Performance by Removing Redundant Padding and Optimizing Alltoall Communication module:quantization
#949 opened May 25, 2025 by SlightwindSec Loading…
Improve Prefill Performance by Removing Redundant Padding and Optimizing Alltoall Communication module:quantization
#948 opened May 25, 2025 by SlightwindSec Loading…
[perf][WIP] Support MOE Multi-stream in Deepseek module:ops module:quantization
#947 opened May 24, 2025 by David9857 Loading…
[Bugfix] Adjust inputbatch to be compatible with latest vllm
#945 opened May 24, 2025 by MengqingCao Loading…
[BugFix] Fix a problem of failing to utilize custom kv cache dtype. module:core
#944 opened May 24, 2025 by whx-sjtu Loading…
[Scheduler][P/D] Add support for disaggregated prefill in AsecendScheduler.
#943 opened May 24, 2025 by whx-sjtu Loading…
[feat][WIP]: support multistream overlap(dbo) for deepseek module:ops
#941 opened May 23, 2025 by zxdukki • Draft
[BugFix] Fix data parallel
#940 opened May 23, 2025 by yiz-liu Loading…
fix: ascend_scheduler adapt to v0.9.0
#939 opened May 23, 2025 by zzzzwwjj Loading…
2
[MLA][Graph] Improve assertion on Graph mode with MLA
#933 opened May 22, 2025 by MengqingCao Loading…
2
[Do Not Merge]add grouped_matmul_swiglu_quant module:quantization
#930 opened May 22, 2025 by Angazenn Loading…
[CI] Add nightly CI
#926 opened May 22, 2025 by Potabk Loading…
2
[Performance] Add EPLB expert map import capabilities module:ops
#919 opened May 21, 2025 by songshanhu07 Loading…
5
[ModelRunner] Support embedding inputs module:tests ready read for review
#916 opened May 21, 2025 by Potabk Loading…
5
[WIP][Platform] Add support for Ascend 310P module:core module:ops
#914 opened May 21, 2025 by farawayboat Loading…
[Fix] Fix update_aclgraph_sizes when running MoE models module:core
#913 opened May 21, 2025 by yiz-liu Loading…
6
use group gemm nz module:quantization
#910 opened May 21, 2025 by ttanzhiqiang Loading…
2
[perf]: add NZ transformation for QuantMatmul and use dequant_swiglu_… module:quantization ready read for review
#907 opened May 20, 2025 by linfeng-yuan Loading…
2
[perf][WIP]: using NZ optimization for quantized GMM module:quantization
#906 opened May 20, 2025 by linfeng-yuan Loading…
1
[Bugfix] Fix deepseek V0 percision issue and add acc ci for it module:ops module:quantization module:tests
#905 opened May 20, 2025 by MengqingCao Loading…
1
[bugfix] some bugs maybe fail to run
#896 opened May 19, 2025 by NINGBENZHE Loading…
2
[1/N][UT][v1 MTP] add basic v1 mtp features module:ops module:tests
#890 opened May 17, 2025 by XWFAlone Loading…
7
[CI/UT][PD Disaggreate] Initialize PD Disaggreate UT module:pd PD disaggregation related module:tests
#889 opened May 17, 2025 by MengqingCao Loading…
3
Fix the device error when using ray as vllm-acend backend module:core module:ops module:tests
#884 opened May 16, 2025 by zhuo97 Loading…
5
[CI] Add upper time to distributed test module:tests
#880 opened May 16, 2025 by Potabk Loading…
ProTip! Follow long discussions with comments:>50.