-
Notifications
You must be signed in to change notification settings - Fork 179
Issues: vllm-project/vllm-ascend
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.
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
Author
Label
Projects
Milestones
Assignee
Sort
Issues list
[Bug]: deepseek-v2-lite tp=8 ep=8 accuracy is not correct
bug
Something isn't working
#1077
opened Jun 5, 2025 by
david6666666
[Bug]: vllm-ascend 离线1p1d分离报错,和vllm不兼容
bug
Something isn't working
#1074
opened Jun 5, 2025 by
1027866388
[Usage]: Question about cache capacity checking in the scheduler
#1069
opened Jun 4, 2025 by
Zazzle516
[Bug][CI Failure]: AttributeError: Something isn't working
help wanted
Extra attention is needed
ParallelConfig
object has no attribute expert_parallel_size
bug
#1059
opened Jun 4, 2025 by
shen-shanshan
[Bug][CI Failure]: TypeError: Something isn't working
help wanted
Extra attention is needed
InputBatch.__init__()
got an unexpected keyword argument block_size
bug
#1058
opened Jun 4, 2025 by
shen-shanshan
When deploying deepseek-R1-W8A8 with vLLM 0.8.5 rc1, report warning ArgSort to AiCPU due to int32/int64; and when stack model long time no response
bug
Something isn't working
#1057
opened Jun 4, 2025 by
ThebestLK
[Feature]: support V1 report_usage_stats in vllm-ascend
feature request
help wanted
Extra attention is needed
#1053
opened Jun 3, 2025 by
Yikun
[Bug][V1]: Failed to start openai api_server with exception "Parameter block_size has unsupported type list[int]"
bug
Something isn't working
#1048
opened Jun 3, 2025 by
farawayboat
[Usage]: 部署vllm-ascend报错
module:mindie-turbo
MindIE Turbo related
#1047
opened Jun 3, 2025 by
bottleofwater11
[Bug]: Attempted to assign 58 = 58 multimodal tokens to 59 placeholders
bug
Something isn't working
#1045
opened Jun 3, 2025 by
jojolee123
[Bug][V1]: Failed to start Qwen/Qwen2.5-VL-7B-Instruct accuracy serve
bug
Something isn't working
#1044
opened Jun 3, 2025 by
Yikun
[Bug][V1]: Qwen/Qwen2.5-7B-Instruct accuracy ceval-valid failed
bug
Something isn't working
#1043
opened Jun 3, 2025 by
Yikun
[Bug]: Failed to complete vllm benchmark after enable VLLM_USE_V1=1
bug
Something isn't working
#1038
opened May 30, 2025 by
Yikun
[Bug]: 在昇腾上使用api方式部署Qwen2-Audio-7B-Instruct模型,openai回复:{"object":"error","message":"The model does not support Transcriptions API","type":"BadRequestError","param":null,"code":400}
bug
Something isn't working
#1035
opened May 30, 2025 by
winni0
[Doc]: add a description after mindie-turbo install successfully
documentation
Improvements or additions to documentation
#1028
opened May 30, 2025 by
Yikun
[Guide][Performance]: vLLM Ascend v0.7.3.post1 benchmark for Qwen3
guide
guide note
performance
#1025
opened May 30, 2025 by
zhanglzu
[Bug]: v0.8.4rc2 - The inference performance of QwQ-32B-w8a8 is worse than fp16
bug
Something isn't working
#1015
opened May 29, 2025 by
tingyiz97
[Bug]: vllm0.8.4+vllm_ascend0.8.4rc2(驱动24.1rc2,cann8.1rc1(cann8.0也试过),torch_npu2.5.1),离线能够跑起来,并发运行在线推理服务,算子库链接不到aclnnNonzeroV2
bug
Something isn't working
#1006
opened May 29, 2025 by
towy98
[Feature]: Implement Eagle3 Acceleration on vllm-ascend
feature request
RFC
Request For Comments
#1004
opened May 29, 2025 by
umeiko
[Doc]: 能否提供一个国内的镜像
documentation
Improvements or additions to documentation
#1001
opened May 29, 2025 by
aicodex
[Bug]: Qwen2-VL-7B-Instruct oom on single card (64GB)
bug
Something isn't working
module:multimodal
#1000
opened May 29, 2025 by
jojolee123
[Bug]: 发送请求卡死 偶现
bug
Something isn't working
module:multimodal
#992
opened May 28, 2025 by
jojolee123
[Bug]: moe ep=4 etp=4, the result is abnormal
bug
Something isn't working
#991
opened May 28, 2025 by
ttanzhiqiang
Previous Next
ProTip!
Follow long discussions with comments:>50.