Skip to content
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

2.6.2版本 + 2433 的bugfix,Vulkan Image模式(-DMNN_VULKAN_IMAGE=ON)的误差很小了,但是性能很差了,符合预期吗 #2553

Closed
cmelody999 opened this issue Aug 21, 2023 · 9 comments
Labels

Comments

@cmelody999
Copy link

平台(如果交叉编译请再附上交叉编译目标平台):

Platform(Include target platform as well if cross-compiling):

Github版本:

Github Version:

直接下载ZIP包请提供下载日期以及压缩包注释里的git版本(可通过7z l zip包路径命令并在输出信息中搜索Comment 获得,形如Comment = bc80b11110cd440aacdabbf59658d630527a7f2b)。 git clone请提供 git commit 第一行的commit id

Provide date (or better yet, git revision from the comment section of the zip. Obtainable using 7z l PATH/TO/ZIP and search for Comment in the output) if downloading source as zip,otherwise provide the first commit id from the output of git commit

编译方式:

Compiling Method

请在这里粘贴cmake参数或使用的cmake脚本路径以及完整输出
Paste cmake arguments or path of the build script used here as well as the full log of the cmake proess here or pastebin

编译日志:

Build Log:

粘贴在这里
Paste log here or pastebin

2.6.2版本 + #2433 的bugfix(#2542 ),Vulkan Image模式(-DMNN_VULKAN_IMAGE=ON)的误差很小了,但是性能很差了,符合预期吗
image

@cmelody999 cmelody999 changed the title 2.6.2版本 + https://github.com/alibaba/MNN/issues/2433 的bugfix(https://github.com/alibaba/MNN/pull/2542 ),Vulkan Image模式(-DMNN_VULKAN_IMAGE=ON)的误差很小了,但是性能很差了,符合预期吗 2.6.2版本 + 2433 的bugfix,Vulkan Image模式(-DMNN_VULKAN_IMAGE=ON)的误差很小了,但是性能很差了,符合预期吗 Aug 21, 2023
@jxt1234
Copy link
Collaborator

jxt1234 commented Aug 23, 2023

precision 是?
建议优先用 OpenCL ,优化投入较多

@jxt1234 jxt1234 added the vulkan label Aug 23, 2023
@cmelody999
Copy link
Author

16位的模型,cost是ms

@jxt1234
Copy link
Collaborator

jxt1234 commented Aug 24, 2023

运行性能与模型无关,主要是 precision 设的是 normal / high / low ?

@jxt1234
Copy link
Collaborator

jxt1234 commented Aug 24, 2023

有用 benchmark.out 测试过么?

@cmelody999
Copy link
Author

backendConfig.precision = MNN::BackendConfig::Precision_Low

image

@cmelody999
Copy link
Author

有用 benchmark.out 测试过么?

benchmark.out测试方法有文档么

@jxt1234
Copy link
Collaborator

jxt1234 commented Sep 4, 2023

@qiankl
Copy link

qiankl commented Sep 5, 2023

2.7.0 vulkan推理误差变小了,速度问题还没有解决是吗

@jxt1234
Copy link
Collaborator

jxt1234 commented Sep 7, 2023

这个模型到后面尺寸较大,被 vulkan image 拆成了很多份 image ,执行效率下降。速度问题由于 厂商加速相关扩展主要在 opencl ,mnn 主要基于 opencl 优化,建议都优先使用 opencl 。mnn vulkan 主要是启动性能上有优势。

@jxt1234 jxt1234 closed this as completed Sep 7, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

No branches or pull requests

3 participants