Comments (14)
maybe it's a driver bug
from mnn.
from mnn.
maybe your compute have two GPUs ,one is intel another is Nvidia, program default use device 0 , Intel's GPU is device 0 ,and Nvidia's GPU is device 1 .please check this
from mnn.
Thanks for the reply. The first two links don't help. I indeed have two GPUs, but both of them are NV's cards, the same model.
from mnn.
sorry ,the log "Invalide device for support vulkan" twice, so maybe enumeratePhysicalDevices
failed ,please modify const bool VulkanInstance::supportVulkan()
function ,and print error code
from mnn.
it's -3.
VK_ERROR_INITIALIZATION_FAILED = -3
so a vulkan driver compatible issue?
from mnn.
try run vulkaninfo
on terminal , is it success ?
from mnn.
try export VK_ICD_FILENAMES=/usr/share/vulkan/icd.d/nvidia_icd.json
maybe the dir is not this ,please change this ,then relaunch ./backendTest.out
from mnn.
vulkaninfo output:
# vulkaninfo
Xlib: extension "NV-GLX" missing on display ":0".
==========
VULKANINFO
==========
Vulkan Instance Version: 1.1.106
/home/lunarg/sdk-build/Vulkan-Tools/vulkaninfo/vulkaninfo.c:5523: failed with VK_ERROR_INITIALIZATION_FAILED
the json file is located at:
/etc/vulkan/icd.d/nvidia_icd.json
I tried:
export VK_ICD_FILENAMES=/etc/vulkan/icd.d/nvidia_icd.json
then ran ./backendTest.out
, still no lucky with the same error
from mnn.
if vulkaninfo
works incorrect , another application will not work too.
Check if you have vulkan-intel installed, it may prevent Nvidia's vulkan driver from being detected
from mnn.
following https://wiki.archlinux.org/index.php/Vulkan#Nvidia_-_vulkan_is_not_working_and_can_not_initialize
from mnn.
your computer CPU is Intel desktop or Xeon ? usually desktop CPU has an eGPU.
from mnn.
from mnn.
that works for Fedora ,maybe it works on CentOS too, but maybe dnf
is not exists on CentOS ,so try yum install that.
from mnn.
Related Issues (20)
- onnx转mnn,resize报错 HOT 3
- TorchScript to MNN转换出错 HOT 1
- ImageProcess,官方文档中的示例应用于 灰度图,从Tensor.print()出来的数值无法用matplotlib绘制出图像 HOT 10
- convert onnx model of decode part of SAM failed. HOT 2
- windows下是否支持GPU推理? HOT 2
- 【Mac上运行llm_demo报错:./llm_demo model/qwen-1.8b-int4 HOT 2
- mnn推断比pytorch推断耗时长 HOT 8
- 交叉编译报错 HOT 3
- ONNX模型成功转为MNN后,通过几个测试工具如MNNV2Basic.out测试均未发现错误,使用python加载运行出现段错误 HOT 2
- Error for concat size of op HOT 3
- AMD64 上运行正常的程序到Arm64上出现段错误,出现在创建session时 HOT 2
- 如何调用自带的 vulkan ,经试验,无论MNN_USE_SYSTEM_LIB怎么设置都还是去找系统的库 HOT 5
- 提个bug, Windows 使用 MSVC 编译时, 如果配置为 MT/MTd , BenchMark 的doBench函数在 析构智能指针崩溃 HOT 1
- 请问mnn框架在英伟达的orinX上可以用吗(使用GPU作为backend)?
- MNN 在 iOS平台上编译mnn_2.8.0_ios_llm.zip release版本的方法 HOT 2
- Android使用Vulkan推理模型,好像依然是跑在CPU上 HOT 1
- 请教关于mnncompress的使用问题 HOT 3
- 几乎必现的heap-use-after-free HOT 3
- debian系统下mnnquant量化出错:
- mnn推理时间异常 HOT 5
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from mnn.