Skip to content

Issues: vllm-project/vllm-ascend

vLLM Ascend Roadmap Q2 2025
#448 opened Mar 31, 2025 by Yikun
Open 5
[Guide] Official Guide Index
#840 opened May 14, 2025 by wangxiyuan
Open
[v0.8.5rc1] FAQ / Feedback | 问题/反馈
#754 opened May 5, 2025 by Yikun
Open 2
Beta
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

114 Open 184 Closed
114 Open 184 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
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

[v0.7.3] FAQ / Feedback | 问题/反馈
#848 opened May 14, 2025 by wangxiyuan
[Bug]: 使用--lora-modules字段加载lora模型效果不好 bug Something isn't working
#843 opened May 14, 2025 by joyhhheee
1
[RFC]: P/D Disaggregation Support RFC Request For Comments
#841 opened May 14, 2025 by MengqingCao
2 of 10 tasks
[Guide] Official Guide Index
#840 opened May 14, 2025 by wangxiyuan
[Bug]: install mindie turbo fail to start DS-W8A8 bug Something isn't working mindie-turbo MindIE Turbo related
#832 opened May 13, 2025 by FrankMinions
[Bug]: modelscope.hub.errors.NotExistError: The model: Qwen/Qwen2.5-VL-7B-Instruct has no revision: main ! bug Something isn't working
#829 opened May 12, 2025 by nutriver
@shen-shanshan
14
[RFC]: vLLM Ascend Governance | Mechanics RFC Request For Comments
#828 opened May 12, 2025 by Yikun
[Bug]: Many unused UserWorkspaceSize0 log print bug Something isn't working
#825 opened May 12, 2025 by as12138
1
[Bug]: Qwen2-1.5B Inference Startup Hang on Huawei 910B Card under vNPU bug Something isn't working
#821 opened May 12, 2025 by fyuan1316
1
[Performance]: vllm-ascend + mindie-turbo Performance Optimization documentation Improvements or additions to documentation RFC Request For Comments
#815 opened May 12, 2025 by shen-shanshan
[Bug]: fail to start W8A8 deepseek-R1 with TP=8,PP=2 bug Something isn't working
#813 opened May 12, 2025 by gao12312
2
[Bug]: RuntimeError: shape '[-1, 3, 80, 1280]' is invalid for input size 1966080 bug Something isn't working
#809 opened May 12, 2025 by as12138
1
Re-enable vllm-empty/tests/benchmarks test
#808 opened May 11, 2025 by Yikun
1
[RFC]: Custom Ascendc Kernel Of 'Prepare Input' in Multi-Step Feature. RFC Request For Comments
#807 opened May 11, 2025 by wonderful199082
1
[Installation]: Failed to install vllm-ascend from source installation
#804 opened May 10, 2025 by tongtong0613
2
[Performance]: custom ascendc kernel(rotary_embedding) performance performance
#802 opened May 9, 2025 by ttanzhiqiang
5
[Bug]: vllm can't run deepseek 70b with Huawei ascend 910b npu card bug Something isn't working
#800 opened May 9, 2025 by huazq
1
[WIP]: Rack Scale Ascend Platform Large-scale MoE Deployment Support feature request
#798 opened May 8, 2025 by Zaragoto
[Feature]: Support Mixtral 8*22B & 8*7B feature request
#797 opened May 8, 2025 by wu330
1
[Accuracy]: vllm-ascend v0.7.3 release accuarcy report performance
#790 opened May 8, 2025 by hfadzxy
[Usage]: Usage on AscendScheduler in vLLM Ascend
#788 opened May 8, 2025 by MengqingCao
2
[Bug]: The program cannot exit normally when using offline inference bug Something isn't working
#787 opened May 8, 2025 by mengwei805
1
[Bug]: precision issue: V0 engine + deepseekR1 model + double G8600 + dp2tp16 bug Something isn't working
#785 opened May 7, 2025 by rfy48
2
[Bug]: Qwen3-235B cannot be run successfully with vllm v1 engine on version 0.8.5rc1 bug Something isn't working
#781 opened May 7, 2025 by BestKuan
2
[Bug]: bug Something isn't working
#780 opened May 7, 2025 by caomengqing
ProTip! Mix and match filters to narrow down what you’re looking for.