RTX5080显卡,使用最新的版本报错 #4710
Replies: 3 comments
-
|
是的,你遇到的问题确认是 RTX 5080 (Blackwell 架构, 几个可行的解决思路:
考虑到你在 Windows 上使用,Docker 或切换 To reply, just mention @dosu. How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other |
Beta Was this translation helpful? Give feedback.
-
|
blackwell用wsl+vllm比较稳定 |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.

Uh oh!
There was an error while loading. Please reload this page.
-
在使用最新版的 MinerU 尝试通过
vlm模式解析 PDF 时,底层lmdeploy的 Turbomind 引擎初始化崩溃。排查后确认,是由于 RTX 5080 (Blackwell 架构,
sm_120) 过于新,当前 pip 仓库中的lmdeploy==0.11.1Windows 预编译包似乎未包含sm_120的 CUDA kernel,导致报no kernel image is available致命错误。报错代码:
命令行中输出的致命错误截取:
Plaintext
2026-04-01 16:10:48.874 | INFO | mineru.backend.vlm.vlm_analyze:get_model:198 - lmdeploy device is: cuda, lmdeploy backend is: turbomind
...
Add dll path C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\bin, please note cuda version should >= 11.3 when compiled with cuda 11
torch_dtypeis deprecated! Usedtypeinstead![TM][FATAL] kernels\gemm\tuner\measurer.cu(83): Check failed: status == cudaSuccess no kernel image is available for execution on the device
随后 Gradio 抛出 502 Bad Gateway 异常:
Plaintext
click.exceptions.ClickException: Failed to query task status for text.pdf: 502 Bad Gateway
这个是我在gemini指导下安装过程和解决方法,是这个问题吗?
Beta Was this translation helpful? Give feedback.
All reactions